Pchem Student Answers PDF

Student Solutions
Manual to
Accompany Atkins’
Physical Chemistry
INTERNATIONAL EDITION
Peter Bolgar
Haydn Lloyd
Aimee North
Vladimiras Oleinikovas
Stephanie Smith
and
James Keeler
Department of Chemistry
University of Cambridge
UK
Table of contents
Preface vii
1 The properties of gases 1

1A The perfect gas 1
1B The kinetic model 12
1C Real gases 23
2 Internal energy 39
2A Internal energy 39
2B Enthalpy 45
2C Thermochemistry 48
2D State functions and exact differentials 57
2E Adiabatic changes 63
3 The second and third laws 69

3A Entropy 69
3B Entropy changes accompanying specific processes 75
3C The measurement of entropy 86
3D Concentrating on the system 96
3E Combining the First and Second Laws 102
4 Physical transformations of pure substances 113

4A Phase diagrams of pure substances 113
4B Thermodynamic aspects of phase transitions 115
5 Simple mixtures 131

5A The thermodynamic description of mixtures 131
5B The properties of solutions 142
5C Phase diagrams of binary systems: liquids 159
5D Phase diagrams of binary systems: solids 166
5E Phase diagrams of ternary systems 172
5F Activities 177
6 Chemical equilibrium 193

6A The equilibrium constant 193
iv TABLE OF CONTENTS
6B The response of equilibria to the conditions 202

6C Electrochemical cells 215
6D Electrode potentials 221
7 Quantum theory 237

7A The origins of quantum mechanics 237
7B Wavefunctions 243
7C Operators and observables 247
7D Translational motion 256
7E Vibrational motion 271
7F Rotational motion 281
8 Atomic structure and spectra 291

8A Hydrogenic Atoms 291
8B Many-electron atoms 300
8C Atomic spectra 303
9 Molecular Structure 311

9A Valence-bond theory 311
9B Molecular orbital theory: the hydrogen molecule-ion 315
9C Molecular orbital theory: homonuclear diatomic molecules 319
9D Molecular orbital theory: heteronuclear diatomic molecules 323
9E Molecular orbital theory: polyatomic molecules 329
10 Molecular symmetry 345

10A Shape and symmetry 345
10B Group theory 355
10C Applications of symmetry 366
11 Molecular Spectroscopy 377

11A General features of molecular spectroscopy 377
11B Rotational spectroscopy 387
11C Vibrational spectroscopy of diatomic molecules 400
11D Vibrational spectroscopy of polyatomic molecules 413
11E Symmetry analysis of vibrational spectroscopy 416
11F Electronic spectra 418
11G Decay of excited states 428
TABLE OF CONTENTS v
12 Statistical thermodynamics 439

12A The Boltzmann distribution 439
12B Partition functions 443
12C Molecular energies 453
12D The canonical ensemble 461
12E The internal energy and entropy 463
12F Derived functions 476
13 Molecules in motion 487

13A Transport properties of a perfect gas 487
13B Motion in liquids 494
13C Diffusion 499
14 Chemical kinetics 509

14A The rates of chemical reactions 509
14B Integrated rate laws 515
14C Reactions approaching equilibrium 531
14D The Arrhenius equation 535
14E Reaction mechanisms 539
14F Examples of reaction mechanisms 545
14G Photochemistry 551
15 Reaction dynamics 569

15A Collision theory 569
15B Diffusion-controlled reactions 574
15C Transition-state theory 577
15D The dynamics of molecular collisions 589
15E Electron transfer in homogeneous systems 591
16 Magnetic resonance 599

16A General principles 599
16B Features of NMR spectra 602
16C Pulse techniques in NMR 612
16D Electron paramagnetic resonance 620
17 Molecular Interactions 625

17A Electric properties of molecules 625
vi TABLE OF CONTENTS
17B Interactions between molecules 637

17C Liquids 643
17D Macromolecules 646
17E Self-assembly 657
18 Solids 663
18A Crystal structure 663
18B Diffraction techniques 666
18C Bonding in solids 673
18D The mechanical properties of solids 678
18E The electrical properties of solids 679
18F The magnetic properties of solids 681
18G The optical properties of solids 684
19 Processes at solid surfaces 689

19A An introduction to solid surfaces 689
19B Adsorption and desorption 694
19C Heterogeneous catalysis 706
19D Processes at electrodes 710
Preface
This manual provides detailed solutions to the (a) Exercises and the odd-numbered Discus-
sion questions and Problems from the international edition of Atkins’ Physical Chemistry.
Conventions used is presenting the solutions
We have included page-specific references to equations, sections, figures and other features
of the main text. Equation references are denoted [17B.3b–685], meaning eqn 17B.3b located
on page 685 (the page number is given in italics). Other features are referred to by name,
with a page number also given.
Generally speaking, the values of physical constants (from the first page of the main text)
are used to 5 significant figures except in a few cases where higher precision is required.
In line with the practice in the main text, intermediate results are simply truncated (not
rounded) to three figures, with such truncation indicated by an ellipsis, as in 0.123...; the
value is used in subsequent calculations to its full precision.
The final results of calculations, generally to be found in a box , are given to the precision
warranted by the data provided. We have been rigorous in including units for all quantities
so that the units of the final result can be tracked carefully. The relationships given on
the back of the front cover are useful in resolving the units of more complex expressions,
especially where electrical quantities are involved.
Some of the problems either require the use of mathematical software or are much easier
with the aid of such a tool. In such cases we have used Mathematica (Wolfram Research,
Inc.) in preparing these solutions, but there are no doubt other options available. Some of
the Discussion questions relate directly to specific section of the main text in which case we
have simply given a reference rather than repeating the material from the text.
Acknowledgements
In preparing this manual we have drawn on the equivalent volume prepared for the 10th edi-
tion of Atkins’ Physical Chemistry by Charles Trapp, Marshall Cady, and Carmen Giunta. In
particular, the solutions which use quantum chemical calculations or molecular modelling
software, and some of the solutions to the Discussion questions, have been quoted directly
from the solutions manual for the 10th edition, without significant modification. More
generally, we have benefited from the ability to refer to the earlier volume and acknowledge,
with thanks, the influence that its authors have had on the present work.
This manual has been prepared by the authors using the LATEX typesetting system, in
the implementation provided by MiKTEX (miktex.org); the vast majority of the figures
and graphs have been generated using PGFPlots. We are grateful to the community who
maintain and develop these outstanding resources.
Finally, we are grateful to the editorial team at OUP, Jonathan Crowe and Roseanna
Levermore, for their invaluable support in bringing this project to a conclusion.
viii PREFACE
Errors and omissions

In such a complex undertaking some errors will no doubt have crept in, despite the authors’
best efforts. Readers who identify any errors or omissions are invited to pass them on to us
by email to [email protected].
1 The properties of gases
1A The perfect gas

Answers to discussion questions
D1A.1 The partial pressure of gas J, p J , in a mixture of gases is given by [1A.6–9], p J =
x J p, where p is the total pressure and x J is the mole fraction of J.
If the gases are perfect, the partial pressure is also the pressure the gas would
exert if it occupied on its own the same container as the mixture at the same
temperature. This leads to Dalton’s law, which is that the pressure of a mixture
of gases is the sum of the pressures that each one would exert if it occupied the
container alone.
Dalton’s law is a limiting law because it holds exactly only in the limit that there
are no interactions between the molecules, which for real gases will be in the
limit of zero pressure.
Solutions to exercises
E1A.1(a) Consider 1 m3 of air: the mass of gas is therefore 1.146 kg. If perfect gas be-
haviour is assumed, the amount in moles is given by n = pV /RT
pV (0.987 × 105 Pa) × (1 m3 )

n= = = 39.5... mol
RT (8.3145 J K−1 mol−1 ) × ([27 + 273.15] K)
(i) The total amount in moles is n = n O2 + n N2 . The total mass m is computed

from the amounts in moles and the molar masses M as
m = n O2 × M O2 + n N2 × M N2
These two equations are solved simultaneously for n O2 to give the follow-
ing expression, which is then evaluated using the data given
m − M N2 n
n O2 =
M O2 − M N2
(1146 g) − (28.02 g mol−1 ) × (39.5... mol)
= = 9.50... mol
(32.00 g mol−1 ) − (28.02 g mol−1 )
The mole fractions are therefore

n O2 9.50... mol
x O2 = = = 0.240 x N2 = 1 − x O2 = 0.760
n 39.5... mol
2 1 THE PROPERTIES OF GASES
The partial pressures are given by p i = x i p tot

p O2 = x O2 p tot = 0.240(0.987 bar) = 0.237 bar
p N2 = x N2 p tot = 0.760(0.987 bar) = 0.750 bar
(ii) The simultaneous equations to be solved are now
n = n O2 + n N2 + n Ar m = n O2 M O2 + n N2 M N2 + n Ar M Ar
Because it is given that x Ar = 0.01, it follows that n Ar = n/100. The two
unknowns, n O2 and n N2 , are found by solving these equations simultane-
ously to give
100m − n(M Ar + 99M O2 )
n N2 =
100(M N2 − M O2 )
100×(1146 g)−(39.5... mol)×[(39.95 g mol−1 )+99×(32.00 g mol−1 )]
=
100 × [(28.02 g mol−1 ) − (32.00 g mol−1 )]
= 30.8... mol
From n = n O2 + n N2 + n Ar it follows that
n O2 = n − n Ar − n N2
= (39.5... mol) − 0.01 × (39.5... mol) − (30.8... mol) = 8.31... mol
The mole fractions are
nN 30.8... mol n O2 8.31... mol
x N2 = 2 = = 0.780 x O2 = = = 0.210
n 39.5... mol n 39.5... mol
The partial pressures are
p N2 = x N2 p tot = 0.780 × (0.987 bar) = 0.770 bar
p O2 = x O2 p tot = 0.210 × (0.987 bar) = 0.207 bar
Note: the final values are quite sensitive to the precision with which the inter-
mediate results are carried forward.
E1A.2(a) The vapour is assumed to be a perfect gas, so the gas law pV = nRT applies. The
task is to use this expression to relate the measured mass density to the molar
mass.
First, the amount n is expressed as the mass m divided by the molar mass M to
give pV = (m/M)RT; division of both sides by V gives p = (m/V )(RT/M).
The quantity (m/V ) is the mass density ρ, so p = ρRT/M, which rearranges
to M = ρRT/p; this is the required relationship between M and the density.
ρRT
M=
p
(1.23 kg m−3 ) × (8.3145 J K−1 mol−1 ) × (330 K)
=
20.0 × 103 Pa
= 0.169 kg mol−1
The relationships 1 J = 1 kg m2 s−2 and 1 Pa = 1 kg m−1 s−2 have been used.

SOLUTIONS MANUAL TO ACCOMPANY ATKINS’ PHYSICAL CHEMISTRY 3
E1A.3(a) Charles’ law [1A.3b–7] states that V ∝ T at constant n and p, and p ∝ T at

constant n and V . For a fixed amount the density ρ is proportional to 1/V , so it
follows that 1/ρ ∝ T. At absolute zero the volume goes to zero, so the density
goes to infinity and hence 1/ρ goes to zero. The approach is therefore to plot
1/ρ against the temperature (in ○ C) and then by extrapolating the straight line
find the temperature at which 1/ρ = 0. The plot is shown in Fig 1.1.
θ/○ C ρ/(g dm−3 ) (1/ρ)/(g−1 dm3 )

−85 1.877 0.532 8
0 1.294 0.772 8
100 0.946 1.057 1
1.0
(1/ρ)/(g−1 dm3 )
0.5
0.0
−300 −200 −100 0 100
θ/○ C
Figure 1.1
The data are a good fit to a straight line, the equation of which is
(1/ρ)/(g−1 dm3 ) = 2.835 × 10−3 × (θ/○ C) + 0.7734
The intercept with 1/ρ = 0 is found by solving
0 = 2.835 × 10−3 × (θ/○ C) + 0.7734
This gives θ = −273 ○ C as the estimate of absolute zero.
E1A.4(a) (i) The mole fractions are
n H2 2.0 mol
x H2 = = = 2
x N2 = 1 − x H2 = 1
n H2 + n N2 2.0 mol + 1.0 mol 3 3
(ii) The partial pressures are given by p i = x i p tot . The total pressure is given
by the perfect gas law: p tot = n tot RT/V
2 (3.0 mol) × (8.3145 J K−1 mol−1 ) × (273.15 K)

p H2 = x H2 p tot = ×
3 22.4 × 10−3 m3
= 2.0 × 105 Pa
1 (3.0 mol) × (8.3145 J K−1 mol−1 ) × (273.15 K)
p N2 = x N2 p tot = ×
3 22.4 × 10−3 m3
5
= 1.0 × 10 Pa
Expressed in atmospheres these are 2.0 atm and 1.0 atm, respectively.
(iii) The total pressure is
(3.0 mol) × (8.3145 J K−1 mol−1 ) × (273.15 K)

= 3.0 × 105 Pa
22.4 × 10−3 m3
or 3.00 atm.
Alternatively, note that 1 mol at STP occupies a volume of 22.4 dm3 , which is
the stated volume. As there are a total of 3.0 mol present the (total) pressure
must therefore be 3.0 atm.
E1A.5(a) From the inside the front cover the conversion between pressure units is: 1 atm
≡ 101.325 kPa ≡ 760 Torr; 1 bar is 105 Pa exactly.
(i) A pressure of 108 kPa is converted to Torr as follows
1 atm 760 Torr
108 kPa × × = 810 Torr
101.325 kPa 1 atm
(ii) A pressure of 0.975 bar is 0.975 × 105 Pa, which is converted to atm as
follows
1 atm
0.975 × 105 Pa × = 0.962 atm
101.325 kPa
E1A.6(a) The perfect gas law [1A.4–8], pV = nRT, is rearranged to give the pressure,
p = nRT/V . The amount n is found by dividing the mass by the molar mass of
Xe, 131.29 g mol−1 .
n
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
(131 g) (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (298.15 K)
p=
(131.29 g mol−1 ) 1.0 dm3
= 24.4 atm
So no , the sample would not exert a pressure of 20 atm, but 24.4 atm if it were
a perfect gas.
E1A.7(a) Because the temperature is constant (isothermal) Boyle’s law applies, pV =

const. Therefore the product pV is the same for the initial and final states
p f Vf = p i Vi hence p i = p f Vf /Vi
The initial volume is 2.20 dm3 greater than the final volume so Vi = 4.65+2.20 =
6.85 dm3 .
Vf 4.65 dm3
pi = × pf = × (5.04 bar) = 3.42 bar
Vi 6.85 dm3
(i) The initial pressure is 3.42 bar
(ii) Because a pressure of 1 atm is equivalent to 1.01325 bar, the initial pressure
expressed in atm is
1 atm
× 3.40 bar = 3.38 atm
1.01325 bar
E1A.8(a) If the gas is assumed to be perfect, the equation of state is [1A.4–8], pV = nRT.
In this case the volume and amount (in moles) of the gas are constant, so it
follows that the pressure is proportional to the temperature: p ∝ T. The ratio
of the final and initial pressures is therefore equal to the ratio of the temper-
atures: p f /p i = Tf /Ti . The pressure indicated on the gauge is that in excess
of atmospheric pressure, thus the initial pressure is 24 + 14.7 = 38.7 lb in−2 .
Solving for the final pressure p f (remember to use absolute temperatures) gives
Tf
pf = × pi
Ti
(35 + 273.15) K
= × (38.7 lb in−2 ) = 44.4... lb in−2
(−5 + 273.15) K
The pressure indicated on the gauge is this final pressure, minus atmospheric
pressure: 44.4... − 14.7 = 30 lb in−2 . This assumes that (i) the gas is behaving
perfectly and (ii) that the tyre is rigid.
E1A.9(a) The perfect gas law pV = nRT is rearranged to give the pressure
nRT
p=
V
n
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
255 × 10−3 g (8.3145 × 10−2 dm3 bar K−1 mol−1 ) × (122 K)
= ×
20.18 g mol−1 3.00 dm3
= 0.0427 bar
Note the choice of R to match the units of the problem. An alternative is to

use R = 8.3154 J K−1 mol−1 and adjust the other units accordingly, to give a
pressure in Pa.
[(255 × 10−3 g)/(20.18 g mol−1 )] × (8.3145 J K−1 mol−1 ) × (122 K)

p=
3.00 × 10−3 m3
5
= 4.27 × 10 Pa
where 1 dm3 = 10−3 m3 has been used along with 1 J = 1 kg m2 s−2 and 1 Pa =
1 kg m−1 s−2 .
E1A.10(a) The vapour is assumed to be a perfect gas, so the gas law pV = nRT applies. The
task is to use this expression to relate the measured mass density to the molar
mass.
First, the amount n is expressed as the mass m divided by the molar mass M to
give pV = (m/M)RT; division of both sides by V gives p = (m/V )(RT/M).
The quantity (m/V ) is the mass density ρ, so p = ρRT/M, which rearranges
to M = ρRT/p; this is the required relationship between M and the density.
ρRT (3.710 kg m−3 ) × (8.3145 J K−1 mol−1 ) × ([500 + 273.15] K)

M= =
p 93.2 × 103 Pa
= 0.255... kg mol−1
where 1 J = 1 kg m2 s−2 and 1 Pa = 1 kg m−1 s−2 have been used. The molar mass
of S is 32.06 g mol−1 , so the number of S atoms in the molecules comprising
the vapour is (0.255... × 103 g mol−1 )/(32.06 g mol−1 ) = 7.98. The result is
expected to be an integer, so the formula is likely to be S8 .
E1A.11(a) The vapour is assumed to be a perfect gas, so the gas law pV = nRT applies; the
task is to use this expression to relate the measured data to the mass m. This
is done by expressing the amount n as m/M, where M is the the molar mass.
With this substitution it follows that m = MPV /RT.
The partial pressure of water vapour is 0.60 times the saturated vapour pressure
M pV
m=
RT
(18.0158 g mol−1 ) × (0.60 × 0.0356 × 105 Pa) × (400 m3 )
=
(8.3145 J K−1 mol−1 ) × ([27 + 273.15] K)
= 6.2 × 103 g = 6.2 kg
Solutions to problems
P1A.1 (a) The expression ρgh gives the pressure in Pa if all the quantities are in
SI units, so it is helpful to work in Pa throughout. From the front cover,
760 Torr is exactly 1 atm, which is 1.01325×105 Pa. The density of 13.55 g cm−3
is equivalent to 13.55 × 103 kg m−3 .
p = p ex + ρgh
= 1.01325 × 105 Pa + (13.55 × 103 kg m−3 ) × (9.806 m s−2 )
× (10.0 × 10−2 m) = 1.15 × 105 Pa
(b) The calculation of the pressure inside the apparatus proceeds as in (a)
p = 1.01325 × 105 Pa + (0.9971 × 103 kg m−3 ) × (9.806 m s−2 )

× (183.2 × 10−2 m) = 1.192... × 105 Pa
The value of R is found by rearranging the perfect gas law to R = pV /nT
pV (1.192... × 105 Pa) × (20.000 × 10−3 m3 )

R= =
nT [(1.485 g)/(4.003 g mol−1 )] × ([500 + 273.15] K)
= 8.315 J K−1 mol−1
P1A.3 The perfect gas law pV = nRT implies that pVm = RT, where Vm is the molar
volume (the volume when n = 1). It follows that p = RT/Vm , so a plot of p
against T/Vm should be a straight line with slope R.
However, real gases only become ideal in the limit of zero pressure, so what is
needed is a method of extrapolating the data to zero pressure. One approach is
to rearrange the perfect gas law into the form pVm /T = R and then to realise
that this implies that for a real gas the quantity pVm /T will tend to R in the limit
of zero pressure. Therefore, the intercept at p = 0 of a plot of pVm /T against p
is an estimate of R. For the extrapolation of the line back to p = 0 to be reliable,
the data points must fall on a reasonable straight line. The plot is shown in
Fig 1.2.
p/atm Vm /(dm3 mol−1 ) (pVm /T)/(atm dm3 mol−1 K−1 )

0.750 000 29.8649 0.082 001 4
0.500 000 44.8090 0.082 022 7
0.250 000 89.6384 0.082 041 4
(pVm /T)/(atm dm3 mol−1 K−1 )
0.08206
0.08204
0.08202
0.08200
0.0 0.2 0.4 0.6 0.8

p/atm
Figure 1.2
The data fall on a reasonable straight line, the equation of which is
(pVm /T)/(atm dm3 mol−1 K−1 ) = −7.995 × 10−5 × (p/atm) + 0.082062
The estimate for R is therefore the intercept, 0.082062 atm dm3 mol−1 K−1 .
The data are given to 6 figures, but they do not fall on a very good straight line
so the value for R has been quoted to one fewer significant figure.
P1A.5 For a perfect gas pV = nRT which can be rearranged to give p = nRT/V . The
amount in moles is n = m/M, where M is the molar mass and m is the mass of
the gas. Therefore p = (m/M)(RT/V ). The quantity m/V is the mass density
ρ, and hence
p = ρRT/M
It follows that for a perfect gas p/ρ should be a constant at a given temperature.
Real gases are expected to approach this as the pressure goes to zero, so a
suitable plot is of p/ρ against p; the intercept when p = 0 gives the best estimate
of RT/M. The plot is shown in Fig. 1.3.
p/kPa ρ/(kg m−3 ) (p/ρ)/(kPa kg−1 m3 )

12.22 0.225 54.32
25.20 0.456 55.26
36.97 0.664 55.68
60.37 1.062 56.85
85.23 1.468 58.06
101.30 1.734 58.42
58
(p/ρ)/(kPa kg−1 m3 )
56
54
0 20 40 60 80 100
p/kPa
Figure 1.3
(p/ρ)/(kPa kg−1 m3 ) = 0.04610 × (p/kPa) + 53.96
The intercept is (p/ρ)lim p→0 , which is equal to RT/M.
RT (8.3145 J K−1 mol−1 ) × (298.15 K)

M= = = 4.594×10−2 kg mol−1
(p/ρ)lim p→0 53.96 × 103 Pa kg−1 m3
The estimate of the molar mass is therefore 45.94 g mol−1 .

P1A.7 (a) For a perfect gas pV = nRT so it follows that for a sample at constant
volume and temperature, p 1 /T1 = p 2 /T2 . If the pressure increases by
∆p for an increase in temperature of ∆T, then with p 2 = p 1 + ∆p and
T2 = T1 + ∆T is follows that
p 1 p 1 + ∆p p 1 ∆T
= hence ∆p =
T1 T1 + ∆T T1
For an increase by 1.00 K, ∆T = 1.00 K and hence
p 1 ∆T (6.69 × 103 Pa) × (1.00 K)

∆p = = = 24.5 Pa
T1 273.16 K
Another way of looking at this is to write the rate of change of pressure
with temperature as
∆p p 1 6.69 × 103 Pa
= = = 24.5... Pa K−1
∆T T1 273.16 K
(b) A temperature of 100.00 ○ C is equivalent to an increase in temperature

from the triple point by 100.00 + 273.15 − 273.16 = 99.99 K
∆p 6.69 × 103 Pa
∆p′ = ∆T ′ × ( ) = (99.99 K) × = 2.44... × 103 Pa
∆T 273.16 K
The final pressure is therefore 6.69 + 2.44... = 9.14 kPa .

(c) For a perfect gas ∆p/∆T is independent of the temperature so at 100.0 ○ C
a 1.00 K rise in temperature gives a pressure rise of 24.5 Pa , just as in (a).
P1A.9 The molar mass of SO2 is 32.06+2×16.00 = 64.06 g mol−1 . If the gas is assumed
to be perfect the volume is calculated from pV = nRT
n
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
nRT 200 × 106 g (8.3145 J K−1 mol−1 ) × ([800 + 273.15] K)
V= =( )
p 64.06 g mol−1 1.01325 × 105 Pa
= 2.7 × 105 m3
Note the conversion of the mass in t to mass in g; repeating the calculation for
300 t gives a volume of 4.1 × 105 m3 .
The volume of gas is therefore between 0.27 km3 and 0.41 km3 .
P1A.11 Imagine a column of the atmosphere with cross sectional area A. The pressure
at any height is equal to the force acting down on that area; this force arises
from the gravitational attraction on the gas in the column above this height –
that is, the ‘weight’ of the gas.
Suppose that the height h is increased by dh. The force on the area A is reduced
because less of the atmosphere is now bearing down on this area. Specifically,
the force is reduced by that due to the gravitational attraction on the gas con-
tained in a cylinder of cross-sectional area A and height dh. If the density of
the gas is ρ, the mass of the gas in the cylinder is ρ × A dh and the force due to
gravity on this mass is ρgA dh, where g is the acceleration due to free fall. The
change in pressure dp on increasing the height by dh is this force divided by
the area, so it follows that
dp = −ρgdh
The minus sign is needed because the pressure decreases as the height increases.
The density is related to the pressure by starting from the perfect gas equation,
pV = nRT. If the mass of gas is m and the molar mass is M, it follows that
n = m/M and hence pV = (m/M)RT. Taking the volume to the right gives
p = (m/MV )RT. The quantity m/V is the mass density ρ, so p = (ρ/M)RT;
this is rearranged to give an expression for the density: ρ = M p/RT.
This expression for ρ is substituted into dp = −ρgdh to give dp = −(M p/RT)gdh.
Division by p results in separation of the variables (1/p) dp = −(M/RT)gdh.
The left-hand side is integrated between p 0 , the pressure at h = 0 and p, the
pressure at h. The right-hand side is integrated between h = 0 and h
p 1 h Mg
∫ dp = ∫ − dh
p0 p 0 RT
p Mg h
[ln p] p 0 = − [h]0
RT
p M gh
ln =−
p0 RT
The exponential of each side is taken to give
RT
p = p 0 e−h/H with H=
Mg
It is assumed that g and T do not vary with h.
(a) The pressure decrease across such a small distance will be very small be-
cause h/H ≪ 1. It is therefore admissible to expand the exponential and
retain just the first two terms: ex ≈ 1 + x
p = p 0 (1 − h/H)
This is rearranged to give an expression for the pressure decrease, p − p 0
p − p 0 = −p 0 h/H
If it is assumed that p 0 is one atmosphere and that H = 8 km,
(1.01325 × 105 Pa) × (15 × 10−2 m)

p − p 0 = −p 0 h/H = − = −2 Pa
8 × 103 m
(b) The pressure at 11 km is calculated using the full expression
p = p 0 e−h/H = (1 atm) × e−(11 km)/(8 km) = 0.25 atm

P1A.13 Imagine a volume V of the atmosphere, at temperature T and pressure p tot .

If the concentration of a trace gas is expressed as X parts per trillion (ppt), it
means that if that gas were confined to a volume X × 10−12 × V at temperature
T is would exert a pressure p tot . From the perfect gas law it follows that n =
pV /RT, which in this case gives
p tot (X × 10−12 × V )
n trace =
RT
Taking the volume V to the left gives the molar concentration, c trace
n trace X × 10−12 × p tot

c trace = =
V RT
An alternative way of looking at this is to note that, at a given temperature and

pressure, the volume occupied by a gas is proportional to the amount in moles.
Saying that a gas is present at X ppt implies that the volume occupied by the
gas is X × 10−12 of the whole, and therefore that the amount in moles of the gas
is X × 10−12 of the total amount in moles
n trace = (X × 10−12 ) × n tot
This is rearranged to give an expression for the mole fraction x trace

n trace
x trace = = X × 10−12
n tot
The partial pressure of the trace gas is therefore
p trace = x trace p tot = (X × 10−12 ) × p tot
The concentration is n trace /V = p trace /RT, so
n trace X × 10−12 × p tot

c trace = =
V RT
(a) At 10 ○ C and 1.0 atm
X CCl3 F × 10−12 × p tot

c CCl3 F =
RT
261 × 10−12 × (1.0 atm)
=
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × ([10 + 273.15] K)
= 1.1 × 10−11 mol dm−3
X CCl2 F2 × 10−12 × p tot
c CCl2 F2 =
RT
509 × 10−12 × (1.0 atm)
=
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × ([10 + 273.15] K)
= 2.2 × 10−11 mol dm−3
(b) At 200 K and 0.050 atm

X CCl3 F × 10−12 × p tot
c CCl3 F =
RT
261 × 10−12 × (0.050 atm)
=
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (200 K)
= 8.0 × 10−13 mol dm−3
X CCl2 F2 × 10−12 × p tot
c CCl2 F2 =
RT
509 × 10−12 × (0.050 atm)
=
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (200 K)
= 1.6 × 10−12 mol dm−3
1B The kinetic model

Answer to discussion questions
D1B.1 For an object (be it a space craft or a molecule) to escape the gravitational
field of the Earth it must acquire kinetic energy equal in magnitude to the
gravitational potential energy the object experiences at the surface of the Earth.
The gravitational potential between two objects with masses m 1 and m 2 when
separated by a distance r is
Gm 1 m 2
V =−
r
where G is the (universal) gravitational constant. In the case of an object of
mass m at the surface of the Earth, it turns out that the gravitational potential
is given by
GmM
V =−
R
where M is the mass of the Earth and R its radius. This expression implies that
the potential at the surface is the same as if the mass of the Earth were localized
at a distance equal to its radius.
As a mass moves away from the surface of the Earth the potential energy in-
creases (becomes less negative) and tends to zero at large distances. This change
in potential energy must all be converted into kinetic energy if the mass is to
escape. A mass m moving at speed υ has kinetic energy 12 mυ 2 ; this speed will
be the escape velocity υ e when
√
GmM 2GM
1
mυ e =
2
hence υe =
2
R R
The quantity in the square root is related to the acceleration due to free fall,
g, in the following way. A mass m at the surface of the Earth experiences
a gravitational force given GMm/R 2 (note that the force goes as R−2 ). This
force accelerates the mass towards the Earth, and can be written mg. The two
expressions for the force are equated to give
GMm GM
= mg hence = gR
R2 R
This expression for GM/R is substituted into the above expression for υ e to give
√
2GM √
υe = = 2Rg
R
The escape velocity is therefore a function of the radius of the Earth and the
acceleration due to free fall.
The radius of the Earth is 6.37×106 m and g = 9.81 m s−2 so the escape velocity
is 1.11×104 m s−1 . For comparison, the mean speed of He at 298 K is 1300 m s−1
and for N2 the mean speed is 475 m s−1 . For He, only atoms with a speed in
excess of eight times the mean speed will be able to escape, whereas for N2 the
speed will need to be more than twenty times the mean speed. The fraction of
molecules with speeds many times the mean speed is small, and because this
fraction goes as e−υ it falls off rapidly as the multiple increases. A tiny fraction
2
of He atoms will be able to escape, but the fraction of heavier molecules with
sufficient speed to escape will be utterly negligible.
D1B.3 The mean free path is given by [1B.14–18], λ = kT/σ p. In a container of con-
stant volume, the mean free path is directly proportional to temperature and
inversely proportional to pressure. The former dependence can be rationalized
by noting that the faster the molecules travel, the farther on average they go
between collisions. The latter also makes sense in that the lower the pressure,
the less frequent are collisions, and therefore the further the average distance
between collisions.
Perhaps more fundamental than either of these considerations is the depen-
dence on the size of the container and on the size of the molecules. The ratio
T/p is directly proportional to volume for a perfect gas, so the average distance
between collisions is directly proportional to the size of the container holding a
given number of gas molecules. Finally, the mean free path is inversely propor-
tional to the size of the molecules as given by the collision cross section (and
therefore inversely proportional to the square of the radius of the molecule).
E1B.1(a) The most probable speed is given by [1B.10–16], υ mp = (2RT/M)1/2 , the mean
speed is given by [1B.9–16], υ mean = (8RT/πM)1/2 , and the mean relative
speed
√ between two molecules of the same mass is given by [1B.11a–16], υ rel =
2υ mean .
M CO2 = 12.01 + 2 × 16.00 = 44.01 g mol−1 .
1/2
2RT 1/2 2 × (8.3145 J K−1 mol−1 ) × (293.15 K)
υ mp =( ) =( ) = 333 m s−1
M 44.01 × 10−3 kg mol−1
1/2
8RT 1/2 8 × (8.3145 J K−1 mol−1 ) × (293.15 K)
υ mean = ( ) =( ) = 376 m s−1
πM π × (44.01 × 10−3 kg mol−1 )
√ √
υ rel = 2υ mean = 2 × (376 m s−1 ) = 531 m s−1
E1B.2(a) The collision frequency is given by [1B.12b–17], z = σ υ rel p/kT, with the
√ relative
speed for two molecules of the same type given by [1B.11a–16], υ rel = 2υ mean .
The mean speed is given by [1B.9–16], υ mean = (8RT/πM)1/2 . From the Re-
source section the collision cross-section σ is 0.27 nm2 .
σ υ rel p σ p √ 8RT 1/2

z= = × 2×( )
kT kT πM
(0.27 × 10−18 m2 ) × (1.01325 × 105 Pa) √
= × 2
(1.3806 × 10−23 J K−1 ) × (298.15 K)
1/2
8 × (8.3145 J K−1 mol−1 ) × (298.15 K)
×( )
π × (2 × 1.0079 × 10−3 kg mol−1 )
= 1.7 × 1010 s−1
where 1 J = 1 kg m2 s−2 and 1 Pa = 1 kg m−1 s−2 have been used. Note the
conversion of the collision cross-section σ to m2 : 1 nm2 = (1 × 10−9 )2 m2 =
1 × 10−18 m2 .
E1B.3(a) The mean speed is given by [1B.9–16], υ mean = (8RT/πM)1/2 . The collision
frequency is given by [1B.12b–17], z = σ υ rel p/kT, with the
√ relative speed for
two molecules of the same type given by [1B.11a–16], υ rel = 2υ mean . The mean
free path is given by [1B.14–18], λ = kT/σ p
(i) The mean speed is calculated as
1/2
8RT 1/2 8 × (8.3145 J K−1 mol−1 ) × (298.15 K)
υ mean = ( ) =( ) = 475 m s−1
πM π × (2 × 14.01 × 10−3 kg mol−1 )
(ii) The collision cross-section σ is calculated from the collision diameter d

as σ = πd 2 = π × (395 × 10−9 m)2 = 4.90... × 10−19 m2 . With this value
the mean free path is calculated as
kT (1.3806 × 10−23 J K−1 ) × (298.15 K)

λ= = = 82.9×10−9 m = 82.9 nm
σ p (4.90... × 10−19 m2 ) × (1.01325 × 105 Pa)
where 1 J = 1 kg m2 s−2 and 1 Pa = 1 kg m−1 s−2 have been used.

(iii) The collision rate is calculated as

σ υ rel p σ p √ 8RT 1/2
z= = × 2×( )
kT kT πM
(4.90... × 10−19 m2 ) × (1.01325 × 105 Pa) √
= × 2
(1.3806 × 10−23 J K−1 ) × (298.15 K)
1/2
8 × (8.3145 J K−1 mol−1 ) × (298.15 K)
×( )
π × (2 × 14.01 × 10−3 kg mol−1 )
= 8.10 × 109 s−1
An alternative for the calculation of z is to use [1B.13–18], λ = υ rel /z,
rearranged to z = υ rel /λ
√ √
υ rel 2υ mean 2 × (475 m s−1 )
z= = = = 8.10 × 109 s−1
λ λ 82.9 × 10−9 m
E1B.4(a) The container is assumed to be spherical with radius r and hence volume V =
4
3
πr 3 . This volume is expressed in terms the the required diameter d = 2r as
V = 16 πd 3 . Rearrangement of this expression gives d
1/3
6V 1/3 6 × 100 cm3
d=( ) =( ) = 5.75... cm
π π
The mean free path is given by [1B.14–18], λ = kT/σ p. This is rearranged to
give the pressure p with λ equal to the diameter of the vessel
kT (1.3806 × 10−23 J K−1 ) × (298.15 K)
p= = = 0.20 Pa
σ d (0.36 × 10−18 m2 ) × (5.75... × 10−2 m)
Note the conversion of the diameter from cm to m.
E1B.5(a) The mean free path is given by [1B.14–18], λ = kT/σ p.
kT (1.3806 × 10−23 J K−1 ) × (217 K)
λ= =
σ p (0.43 × 10−18 m2 ) × (0.05 × 1.01325 × 105 Pa)
= 1.4 × 10−6 m = 1.4 µm
E1B.6(a) The mean speed is given by [1B.9–16], υ mean = (8RT/πM)1/2 , so υ mean ∝

(i) √
1/M. The ratio of the mean speeds therefore depends on the ratio of
the molar masses
1/2
200.59 g mol−1
1/2
υ mean,H2 M Hg
=( ) =( ) = 9.975
υ mean,Hg M H2 2 × 1.0079 g mol−1
(ii) The mean translational kinetic energy ⟨E k ⟩ is given by 12 m⟨υ 2 ⟩, where

⟨υ 2 ⟩ is the mean square speed, which is given by [1B.7–15], ⟨υ 2 ⟩ = 3RT/M.
The mean translational kinetic energy is therefore
1 1 3RT
⟨E k ⟩ = m⟨υ 2 ⟩ = m ( )
2 2 M
The molar mass M is related to the mass m of one molecule by M = mN A ,

where N A is Avogadro’s constant, and the gas constant can be written R =
kN A , hence
1 3RT 1 3kN A T 3
⟨E k ⟩ = m ( ) = m( ) = kT
2 M 2 mN A 2
The mean translational kinetic energy is therefore independent of the
identity of the gas, and only depends on the temperature: it is the same
for H2 and Hg.
This result is related to the principle of equipartition of energy: a molecule
has three translational degrees of freedom (x, y, and z) each of which
contributes 12 kT to the average energy.
E1B.7(a) The rms speed is given by [1B.8–15], υ rms = (3RT/M)1/2 .

1/2
3 × (8.3145 J K−1 mol−1 ) × (293.15 K)
1/2
3RT
υ rms,H2 =( ) =( )
M H2 2 × 1.0079 × 10−3 kg mol−1
= 1.90 km s−1
where 1 J = 1 kg m2 s−2 has been used. Note that the molar mass is in kg mol−1 .
1/2
3 × (8.3145 J K−1 mol−1 ) × (293.15 K)
υ rms,O2 = ( ) = 478 m s−1
2 × 16.00 × 10−3 kg mol−1
E1B.8(a) The Maxwell–Boltzmann distribution of speeds, f (υ), is given by [1B.4–14].

The fraction of molecules with speeds between υ 1 and υ 2 is given by the integral
υ2
∫ f (υ) dυ
υ1
If the range υ 2 − υ 1 = δυ is small, the integral is well-approximated by

f (υ mid ) δυ
where υ mid is the mid-point of the velocity range: υ mid = 12 (υ 2 + υ 1 ). In this
exercise υ mid = 205 m s−1 and δυ = 10 m s−1 .
M 3/2 2 −Mυ 2mid

fraction = f (υ mid ) δυ = 4π × ( ) υ mid exp ( ) δυ
2πRT 2RT
3/2
2 × 14.01 × 10−3 kg mol−1
= 4π × ( ) × (205 m s−1 )2
2π × (8.3145 J K−1 mol−1 ) × (400 K)
−(2 × 14.01 × 10−3 kg mol−1 ) × (205 m s−1 )2
× exp ( ) × (10 m s−1 )
2 × (8.3145 J K−1 mol−1 ) × (400 K)
= 6.87 × 10−3
where 1 J = 1 kg m2 s−2 has been used. Thus, 0.687% of molecules have veloci-
ties in this range.
E1B.9(a) The mean relative speed is given by [1B.11b–16], υ rel = (8kT/πµ)1/2 , where
µ = m A m B /(m A + m A ) is the effective mass. Multiplying top and bottom of
the expression for υ rel by N A and using N A k = R gives υ rel = (8RT/πN A µ)1/2
in which N A µ is the molar effective mass. For the relative motion of N2 and H2
this effective mass is
M N2 M H2 (2 × 14.01 g mol−1 ) × (2 × 1.0079 g mol−1 )
NA µ = = = 1.88... g mol−1
M N2 + M H2 (2 × 14.01 g mol−1 ) + (2 × 1.0079 g mol−1 )
1/2
8 × (8.3145 J K−1 mol−1 ) × (298.15 K)
1/2
8RT
υ rel = ( ) =( ) = 1832 m s−1
πN A µ π × (1.88... × 10−3 kg mol−1 )
The value of the effective mass µ is dominated by the mass of the lighter molecule,
in this case H2 .
P1B.1 A rotating slotted-disc apparatus consists of a series of disks all mounted on a
common axle (shaft). Each disc has a narrow radial slot cut into it, and the slots
on successive discs are displaced from one another by a certain angle. The discs
are then spun at a constant angular speed.
Detector
Source
Selector
Imagine a molecule moving along the direction of the axle with a certain veloc-
ity such that it passes through the slot in the first disc. By the time the molecule
reaches the second disc the slot in that disc will have moved around, and the
molecule will only pass through the slot if the speed of the molecule is such
that it arrives at the second disc at just the time at which the slot appears in
the path of the molecule. In this way, only molecules with a specific velocity
(or, because the slot has a finite width, a small range of velocities) will pass
through the second slpt. The velocity of the molecules which will pass through
the second disc is set by the angular speed at which the discs are rotated and
the angular displacement of the slots on successive discs.
The angular velocity of the discs is 2πv rad s−1 so in time t the discs move
through an angle θ = 2πvt. If the spacing of the discs is d, a molecule with
velocity υ x will take time t = d/υ x to pass from one disc to the next. If the
second slit is set at an angle α relative to the first, such a molecule will only pass
through the second slit if
d 2πvd
2πv ( )=α hence υx =
υx α
If the angle α is expressed in degrees, α = π(α ○ /180○ ), this rearranges to

2πvd 360○ vd
υx = ○ ○
=
π(α /180 ) α○
With the values given the velocity of the molecules is computed as
360○ vd 360○ v(0.01 m)
υx = = = 180v(0.01 m)
α○ 2○
The Maxwell–Boltzmann distribution of speeds in one dimension is given by
[1B.3–13]
m 1/2 −mυ 2x /2k T
f (υ x ) = ( ) e
2πkT
The given data on the intensity of the beam is assumed to be proportional to
f (υ x ): I ∝ f (υ x ) = Af (υ x ). Because the constant of proportionality is not
known and the variation with υ x is to be explored, it is convenient to take
logarithms to give
m 1/2 mυ 2x
ln I = ln[Af (υ x )] = ln A + ln ( ) −
2πkT 2kT
A plot of ln I against υ 2x is expected to be a straight line with slope −m/2kT;
such a plot is shown in Fig. 1.4.
ν/Hz υ x /m s−1 υ 2x /(104 m2 s−2 ) I(40 K) ln I(40 K) I(100 K) ln I(100 K)

20 36 0.13 0.846 −0.167 0.592 −0.524
40 72 0.52 0.513 −0.667 0.485 −0.724
80 144 2.07 0.069 −2.674 0.217 −1.528
100 180 3.24 0.015 −4.200 0.119 −2.129
120 216 4.67 0.002 −6.215 0.057 −2.865
At both temperatures the data fall on reasonable straight lines, with slope −1.33
at 40 K and −0.516 at 100 K.
If the Maxwell–Boltzmann distribution applies the expected slope at 40 K is
computed as
m M 83.80 × 10−3 kg mol−1

− =− =− = −1.26 × 10−4 m−2 s2
2kT 2RT 2 × (8.3145 J K−1 mol−1 ) × (40 K)
where R = N A k has been used. The expected slope of the above graph is there-
fore −1.26, which compares reasonably well with that found experimentally.
At 100 K the expected slope is
83.80 × 10−3 kg mol−1

− −1 −1 = −5.04 × 10−5 m−2 s2
2 × (8.3145 J K mol ) × (100 K)
Again, the expected slope −0.504 compares reasonably well with that found
experimentally.
0.0
T = 40 K
T = 100 K
−2.0
ln I
−4.0
−6.0
0 1 2 3 4 5
2 −2
υ 2x /(104 m s )
Figure 1.4
P1B.3 The Maxwell–Boltzmann distribution of speeds in one dimension (here x) is

given by [1B.3–13]
m 1/2 −mυ 2x /2k T
f (υ x ) = ( ) e
2πkT
The first task is to find an expression for the mean speed, which is found using
∞
[1B.6–15], ⟨υ n ⟩ = ∫0 υ n f (υ) dυ. In this case
∞ m 1/2 −mυ 2x /2k T
⟨υ x ⟩ = ∫ υx ( ) e dυ
0 kT
The required integral is of the form of G.2 from the Resource section
∞ 1
xe−ax dx =
2
∫
0 2a
With a = m/2kT the mean speed is
m 1/2 1 kT 1/2
υmean = ⟨υ x ⟩ = ( ) ( )=( )
kT 2(m/2kT) 2πm
After the beam emerges from the velocity selector, f (υ x ) is zero for υ x > υmean .
The probability distribution is therefore changed and so needs to be re-normalized
such that υ mean
e−mυ x /2k T dυ x = 1
2
Kx ∫
0
This integral is best evaluated using mathematical software which gives
υ mean πkT 1/2 1
e−mυ x /2k T dυ = (
2
∫ ) erf( √ )
0 2m 2 π
where erf(x) is the error function. The normalized distribution is therefore
2m 1/2 1
e−mυ x /2k T
2
f new (υ x ) = ( ) 1
πkT erf( 2√π )
The new mean speed is computed using this distribution; again this intergral is
best evaluated using mathematical software. Note that the integral extends up
to υmean
2m 1/2 1 υ mean
υ x e−mυ x /2k T dυ x
2
υmean, new = ( ) 1 ∫
πkT erf( 2√π ) 0
υ mean
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
1/2
2kT 1 kT 1/2 1
= (1 − e1/4π ) ( ) = (1 − e 1/4π
)2 ( )
πm erf( 2√1 π ) 2πm erf( 2√1 π )
1
= (1 − e1/4π )2υmean
erf( 2√1 π )
The error function is evaluated numerically to give υmean, new ≈ 0.493 υmean .
P1B.5 The Maxwell–Boltzmann distribution of speeds in three dimensions is given

by [1B.4–14]
M 3/2 2 −M υ 2 /2RT
f (υ) = 4π ( ) υ e
2πRT
with M the molar mass. The most probable speed is given by [1B.10–16], υmp =
(2RT/M)1/2 . If the interval of speeds, ∆υ is small, the fraction of molecules
with speeds in this range, centred at speed υmp is well-approximated by f (υmp )∆υ.
The required fraction of molecules with speeds in the range ∆υ around n × υmp
compared to that centred around υmp is given by
f (n × υmp )∆υ (n × υmp )2 e−M(nυmp ) /2RT

2
= n 2 e−Mυmp (n −1)/2RT
2 2
= −M υ 2 /2RT
f (υmp )∆υ υmp
2
e mp
In taking the ratio, with the exception of the term υ 2 , all of the terms in f (υ)
which multiply the exponential cancel. In this expression the term υmp is re-
placed by (2RT/M)1/2 to give
f (n × υmp )∆υ
= n 2 e−M υmp (n −1)/2RT = n 2 e−M(2RT/M)(n −1)/2RT = n 2 e(1−n )
2 2 2 2
f (υmp )∆υ
For n = 3 this expression evaluates to 3.02 × 10−3 and for n = 4 it evaluates

to 4.89 × 10−6 . These numbers indicate that very few molecules have speeds
several times greater than the most probable speed.
P1B.7 The key idea here is that for an object to escape the gravitational field of the
Earth it must acquire kinetic energy equal in magnitude to the gravitational
potential energy the object experiences at the surface of the Earth. The grav-
itational potential energy between two objects with masses m 1 and m 2 when
separated by a distance r is
Gm 1 m 2
V =−
r
where G is the (universal) gravitational constant. In the case of an object of

mass m at the surface of the Earth, it turns out that the gravitational potential
energy is given by
GmM
V =−
R
where M is the mass of the Earth and R its radius. This expression implies that
the potential at the surface is the same as if the mass of the Earth were localized
at a distance equal to its radius.
As a mass moves away from the surface of the Earth the potential energy in-
creases (becomes less negative) and tends to zero at large distances. If the mass
is to escape its kinetic energy must be greater than or equal to this change in
potential energy. A mass m moving at speed υ has kinetic energy 12 mυ 2 ; this
speed will be the escape velocity υ e when
GmM 2GM 1/2

1
mυ 2e = hence υe = ( )
2
R R
The quantity in the square root is related to the acceleration due to free fall,
g, in the following way. A mass m at the surface of the Earth experiences
a gravitational force given GMm/R 2 (note that the force goes as R −2 ). This
force accelerates the mass towards the Earth, and can be written mg. The two
expressions for the force are equated to give
GMm GM
= mg hence = gR (1.1)
R2 R
This expression for GM/R is substituted into the above expression for υ e to give
2GM 1/2
υe = ( ) = (2Rg)1/2
R
The escape velocity is therefore a function of the radius of the Earth and the
acceleration due to free fall.
The quoted values for the Earth give
√ √
υ e = 2Rg = 2 × (6.37 × 106 m) × (9.81 m s−2 ) = 1.12 × 104 m s−1
For Mars, data is not given on the acceleration due to free fall. However, it
follows from eqn 1.1 that g = GM/R 2 , and hence
gMars MMars REarth 2

= ( )
gEarth MEarth RMars
The acceleration due to freefall on Mars is therefore computed as
MMars REarth 2
gMars = gEarth ( )
MEarth RMars
2
6.37 × 106 m
= (9.81 m s−2 ) × (0.108) × ( ) = 3.76... m s−2
3.38 × 106 m
The escape velocity on Mars is therefore

√ √
υ e = 2Rg = 2 × (3.38 × 106 m) × (3.76... m s−2 ) = 5.04 × 103 m s−1
The mean speed is given by [1B.9–16], υmean = (8RT/πM)1/2 . This expression

is rearranged to give the temperature T at which the mean speed is equal to the
escape velocity
υ 2 πM
T= e
8R
For H2 on the Earth the calculation is
(1.12 × 104 m s−1 )2 × π × (2 × 1.0079 × 10−3 kg mol−1 )

T= = 1.19 × 104 K
8 × (8.3145 J K−1 mol−1 )
The following table gives the results for all three gases on both planets
planet υ e /m s−1 T/104 K (H2 ) T/104 K (He) T/104 K (O2 )

Earth 1.12 × 104 1.19 2.36 18.9
Mars 5.04 × 103 0.242 0.481 3.84
The fraction of molecules with speed greater than υ e is found by integrating the
Maxwell–Boltzmann distribution from this speed up to infinity:
∞ M 3/2 2 −M υ 2 /2RT
fraction with speed ≥ υ e = F = ∫ 4π ( ) υ e dυ
υe 2πRT
This integral is best computed using mathematical software, to give the fol-
lowing results for the fraction F; an entry of zero indicates that the calculated
fraction is zero to within the machine precision.
planet T/K F(H2 ) F(He) F(O2 )

Earth 240 0 0 0
1500 1.49 × 10−4 9.52 × 10−9 0
Mars 240 1.12 × 10−5 5.09 × 10−11 0
1500 0.025 4.31 × 10−2 4.61 × 10−14
These results indicate that the lighter molecules have the greater chance of
escaping (because they are moving faster on average) and that increasing the
temperature increases the probability of escaping (again becuase this increases
the mean speed). Escape from Mars is easier than from the Earth because of
the lower escape velocity, and heavier molecules are seemingly very unlikely to
escape from the Earth.
P1B.9 The Maxwell–Boltzmann distribution of speeds in three dimensions is given

by [1B.4–14]
M 3/2 2 −M υ 2 /2RT
f (υ) = 4π ( ) υ e
2πRT
The fraction with speed between υ 1 and υ 2 is found by integrating the distri-
bution between these speeds; this is best done using mathematical software
υ2 M 3/2 2 −Mυ 2 /2RT
fraction with speed between υ 1 and υ 2 = ∫ 4π ( ) υ e dυ
υ1 2πRT
At 300 K and with M = 2 × 16.00 g mol−1 the fraction is 0.0722 and at 1000 K
the fraction is 0.0134 .
P1B.11 Two hard spheres will collide if their line of centres approach within 2r of one
another, where r is the radius of the sphere. This distance defines the collision
diameter, d = 2r, and the collision cross-section is the area of a circle with this
radius, σ = πd 2 = π(2r)2 . The pressure is computed from the other parameters
using the perfect gas law: p = nRT/V .
The collision frequency is given by [1B.12b–17], z = σ υ rel p/kT, with the
√ relative
speed for two molecules of the same type given by [1B.11a–16], υ rel = 2υ mean .
The mean speed is given by [1B.9–16], υ mean = (8RT/πM)1/2 .
Putting this all together gives
σ υ rel p π(2r)2 √ 8RT 1/2 nRT

z= = × 2×( ) ×
kT kT πM V
√ 8RT 1/2
nN A
= π(2r)2 × 2 × ( ) ×
πM V
where to go to the second line R = N A k has been used. The expression is
evaluated to give
1/2
−9
√ 8×(8.3145 J K−1 mol−1 )×(298.15 K)
z = π(2×(0.38 × 10 2
m)) × 2×( )
π×(16.0416 × 10−3 kg mol−1 )
(0.1 mol) × (6.0221 × 1023 mol−1 )
× = 9.7 × 1010 s−1
1 × 10−3 m3
1C Real gases
D1C.1 The critical constants represent the state of a system at which the distinction
between the liquid and vapour phases disappears. This situation is usually
described by saying that above the critical temperature the liquid phase cannot
be produced by the application of pressure alone. The liquid and vapour phases
can no longer coexist, though supercritical fluids have both liquid and vapour
characteristics.
D1C.3 The van der Waals equation is a cubic equation in the volume V . Every cubic
equation has some values of the coefficients for which the number of real roots
passes from three to one. In fact, any equation of state of odd degree n > 1 can
in principle account for critical behavior because for equations of odd degree
in V there are necessarily some values of temperature and pressure for which
the number of real roots of V passes from n to 1. That is, the multiple values of
V converge from n to 1 as the temperature approaches the critical temperature.
This mathematical result is consistent with passing from a two phase region
(more than one volume for a given T and p) to a one phase region (only one V
for a given T and p), and this corresponds to the observed experimental result
as the critical point is reached.
E1C.1(a) The relation between the critical constants and the van der Waals parameters
is given by [1C.6–26]
a 8a
Vc = 3b pc = Tc =
27b 2 27Rb
All three critical constants are given, so the problem is over-determined: any
pair of the these expressions is sufficient to find values of a and b. It is con-
venient to use R = 8.2057 × 10−2 dm3 atm K−1 mol−1 and volumes in units of
dm3 .
If the expressions for Vc and p c are used, a and b are found in the following
way
Vc = 3b hence b = Vc /3 = (0.0987 dm3 mol−1 )/3 = 0.0329 dm3 mol−1
a a
pc = 2
= hence a = 27(Vc /3)2 p c
27b 27(Vc /3)2
a = 27(Vc /3)2 p c = 27([0.0987 dm3 mol−1 ]/3)2 × (45.6 atm)
= 1.33 atm dm6 mol−2
There are three possible ways of choosing two of the expressions with which to
find a and b, and each choice gives a different value. For a the values are 1.33,
1.74, and 2.26, giving an average of 1.78 atm dm6 mol−2 . For b the values are
0.0329, 0.0329, and 0.0429, giving an average of 0.0362 dm3 mol−1 .
In Section 1C.2(a) on page 23 it is argued that b = 4Vmolec N A , where Vmolec is
the volume occupied by one molecule. This volume is written in terms of the
radius r as 4πr 3 /3 so it follows that r = (3b/16πN A )1/3 .
1/3
3b 1/3
3 × (0.0362 dm3 mol−1 )
r=( ) =( ) = 1.53×10−9 dm = 153 pm
16πN A 16π × (6.0221 × 1023 mol−1 )
E1C.2(a) (i) In Section 1C.1(b) on page 20 it is explained that at the Boyle temperature
Z = 1 and dZ/dp = 0; this latter condition corresponds to the second
virial coefficient, B or B′ , being zero. The task is to find the relationship
between the van der Waals parameters and the virial coefficients, and the
starting point for this is the expressions for the product pVm is each case
([1C.5b–24] and [1C.3b–21])
RT a RT Vm a
van der Waals: p = − 2 hence pVm = −
(Vm − b) Vm (Vm − b) Vm
B
virial: pVm = RT (1 + )
Vm
The van der Waals expression for pVm is rewritten by dividing the denom-
inator and numerator of the first fraction by Vm to give
RT a
pVm = −
(1 − b/Vm ) Vm
The dimensionless parameter b/Vm is likely to be ≪ 1, so the approxima-

tion (1 − x)−1 ≈ 1 + x is used to give
a 1 a
pVm = RT(1 + b/Vm ) − = RT [1 + (b − )]
Vm Vm RT
Comparison of this expression with the virial expansion shows that
a
B=b−
RT
It therefore follows that the Boyle temperature, when B = 0, is Tb = a/Rb.
For the van der Waals parameters from the Resource section
a 6.260 atm dm6 mol−2

Tb = =
Rb (8.2057 × 10−2 dm atm K−1 mol−1 ) × (5.42 × 10−2 dm3 mol−1 )
3
= 1.41 × 103 K
(ii) In Section 1C.2(a) on page 23 it is argued that b = 4Vmolec N A , where

Vmolec is the volume occupied by one molecule. This volume is written
in terms of the radius r as 4πr 3 /3 so it follows that r = (3b/16πN A )1/3 .
1/3
3b 1/3
3 × (5.42 × 10−2 dm3 mol−1 )
r=( ) =( )
16πN A 16π × (6.0221 × 1023 mol−1 )
= 1.75 × 10−9 dm = 175 pm
E1C.3(a) The reduced variables are defined in terms of the critical constants,[1C.8–26]
Vr = Vm /Vc p r = p/p c Tr = T/Tc
If the reduced pressure is the same for two gases (1) and (2) it follows that
p(1) p(2) p(1)

(1)
= (2)
hence p(2) = (1)
× p(2)
c
pc pc pc
and similarly
T (1)
T (2) = (1)
× Tc(2)
Tc
These relationships are used to find the pressure and temperature of gas (2)
corresponding to a particular state of gas (1); it is necessary to know the critical
constants of both gases.
(i) From the tables in the Resource section, for H2 p c = 12.8 atm, Tc = 33.23 K,
and for NH3 p c = 111.3 atm, Tc = 405.5 K. Taking gas (1) as H2 and
gas (2) as NH3 , the pressure and temperature of NH3 corresponding to
p(H2 ) = 1.0 atm and T (H2 ) = 298.15 K is calculated as
p(H2 ) 1.0 atm

p(NH3 ) = (H )
× p(NH
c
3)
= × (111.3 atm) = 8.7 atm
pc 2 12.8 atm
T (H2 ) 298.15 K
T (NH3 ) = (H )
× Tc(NH3 ) = × (405.5 K) = 3.6 × 103 K
Tc 2 33.23 K
(ii) For Xe p c = 58.0 atm, Tc = 289.75 K.
p(H2 ) 1.0 atm

p(Xe) = (H )
× p(Xe)
c = × (58.0 atm) = 4.5 atm
pc 2 12.8 atm
T (H2 ) 298.15 K
T (Xe) = (H )
× Tc(Xe) = × (289.75 K) = 2.6 × 103 K
Tc 2 33.23 K
(iii) For He p c = 2.26 atm, Tc = 5.2 K.
p(H2 ) 1.0 atm

p(He) = (H )
× p(He)
c = × (2.26 atm) = 0.18 atm
pc 2 12.8 atm
T (H2 ) 298.15 K
T (He) = (H )
× Tc(He) = × (5.2 K) = 47 K
Tc 2 33.23 K
E1C.4(a) The van der Waals equation of state in terms of the molar volume is given by
[1C.5b–24], p = RT/(Vm − b) − a/Vm2 . This relationship is rearranged to find b
RT a a RT
p= − hence p + 2 =
Vm − b Vm2 Vm Vm − b
pVm2 + a RT Vm2 Vm − b
hence = hence =
Vm 2 Vm − b pVm + a
2 RT
RT Vm2
hence b = Vm −
pVm2 + a
With the data given
RT Vm2
b = Vm − = (5.00 × 10−4 m3 mol−1 )
pVm2 + a
(8.3145 J K−1 mol−1 ) × (273 K) × (5.00 × 10−4 m3 mol−1 )2
−
(3.0 × 106 Pa) × (5.00 × 10−4 m3 mol−1 )2 + (0.50 m6 Pa mol−2 )
= 4.6 × 10−5 m3 mol−1
where 1 Pa = 1 kg m−1 s−2 and 1 J = 1 kg m2 s−2 have been used.

The compression factor Z is defined in [1C.1–20] as Z = Vm /Vm○ , where Vm○ is

the molar volume of a perfect gas under the same conditions. This volume is
computed from the equation of state for a perfect gas, [1A.4–8], as Vm○ = RT/p,
hence Z = pVm /RT, [1C.2–20]. With the data given
pVm (3.0 × 106 Pa) × (5.00 × 10−4 m3 mol−1 )

Z= = = 0.66
RT (8.3145 J K−1 mol−1 ) × (273 K)
E1C.5(a) The van der Waals equation of state in terms of the volume is given by [1C.5a–
23], p = nRT/(V − b) − an 2 /V 2 . The parameters a and b for ethane are
given in the Resource section as a = 5.507 atm dm6 mol−2 and b = 6.51 ×
10−2 dm3 mol−1 .
With these units it is convenient to use R = 8.2057 × 10−2 dm3 atm K−1 mol−1 .
(i) T = 273.15 K, V = 22.414 dm3 , n = 1.0 mol
nRT an 2
p= − 2
V − nb V
(1.0 mol) × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (273.15 K)
=
(22.414 dm3 ) − (1.0 mol) × (6.51 × 10−2 dm3 mol−1 )
(5.507 atm dm6 mol−2 ) × (1.0 mol)2
− = 0.99 atm
(22.414 dm3 )2
(ii) T = 1000 K, V = 100 cm3 = 0.100 dm3 , n = 1.0 mol

nRT an 2
p= − 2
V − nb V
(1.0 mol) × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (1000 K)
=
(0.100 dm3 ) − (1.0 mol) × (6.51 × 10−2 dm3 mol−1 )
(5.507 atm dm6 mol−2 ) × (1.0 mol)2
− = 1.8 × 103 atm
(0.100 dm3 )2
E1C.6(a) Recall that 1 atm = 1.01325 × 105 Pa, 1 dm6 = 10−6 m6 , and 1 Pa = 1 kg m−1 s−2
1.01325 × 105 Pa 10−6 m6
a = (0.751 atm dm6 mol−2 ) × × = 0.0761 Pa m6 mol−2
1 atm 1 dm6
= 0.0760 kg m−1 s−2 m6 mol−2 = 0.0761 kg m5 s−2 mol−2
10−3 m3
b = (0.0226 dm3 mol−1 ) × = 2.26 × 10−5 m3 mol−1
1 dm3
E1C.7(a) The compression factor Z is defined in [1C.1–20] as Z = Vm /Vm○ , where Vm○ is

the molar volume of a perfect gas under the same conditions. This volume is
computed from the equation of state for a perfect gas, [1A.4–8], as Vm○ = RT/p,
hence Z = pVm /RT [1C.2–20].
(i) If Vm is 12% smaller than the molar volume of a perfect gas, it follows that
Vm = Vm○ (1 − 0.12) = 0.88Vm○ . The compression factor is then computed
directly as
Vm 0.88 × Vm○
Z= ○ = = 0.88
Vm Vm○
(ii) From [1C.2–20] it follows that Vm = ZRT/p
ZRT 0.88 × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (250 K)

Vm = =
p 15 atm
= 1.2 dm3 mol−1
Because Z < 1, implying that Vm < Vm○ , attractive forces are dominant.
E1C.8(a) The van der Waals equation of state in terms of the volume is given by [1C.5a–
23], p = nRT/(V −b)−an 2 /V 2 . The molar mass of N2 is M = 2×14.01 g mol−1 =
28.02 g mol−1 , so it follows that the amount in moles is
n = m/M = (92.4 kg)/(0.02802 kg mol−1 ) = 3.29... × 103 mol
The pressure is found by substituting the given parameters into [1C.5a–23],

noting that the volume needs to be expressed in dm3
nRT an 2
p= − 2
V − nb V
(3.29... × 103 mol) × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (500 K)
=
(1000 dm3 ) − (3.29... × 103 mol) × (0.0387 dm3 mol−1 )
(1.352 atm dm6 mol−2 ) × (3.29... × 103 mol)2
− = 140 atm
(1000 dm3 )2
E1C.9(a) (i) The pressure is computed from the equation of state for a perfect gas,
[1A.4–8], as p = nRT/V
nRT (10.0) × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × ([27 + 273.15] K)
p= =
V 4.860 dm3
= 50.7 atm
(ii) The van der Waals equation of state in terms of the volume is given by
[1C.5a–23], p = nRT/(V − b) − an 2 /V 2 . This is used to calculate the
pressure
nRT an 2
p= − 2
V − nb V
(10.0 mol) × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × ([27 + 273.15] K)
=
(4.860 dm3 ) − (10.0 mol) × (0.0651 dm3 mol−1 )
(5.507 atm dm6 mol−2 ) × (10.0 mol)2
− = 35.2... = 35.2 atm
(4.860 dm3 )2
The compression factor Z is given in terms of the molar volume and pres-
sure by [1C.2–20], Z = pVm /RT. The molar volume is V /n
pVm pV
Z= =
RT nRT
(35.2... atm) × (4.860 dm3 )
= = 0.695
(10.0 mol) × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (300.15 K)
P1C.1 The virial equation is given by [1C.3b–21], pVm = RT(1 + B/Vm + . . .), and
from the Resource section the second virial coefficient B for N2 at 273 K is
−1
−10.5 cm3 mol . The molar mass of N2 is 2 × 14.01 = 28.02 g mol−1 , hence the
molar volume is
V V 2.25 dm3
Vm = = = = 13.8... dm3 mol−1
n m/M (4.56 g)/(28.02 g mol−1 )
This is used to calculate the pressure using the virial equation. It is convenient
to use R = 8.2057 × 10−2 dm3 atm K−1 mol−1 and express all the volumes in
dm3
RT B
p= (1 + )
Vm Vm
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (273 K) −1.05 × 10−2 dm3 mol−1
= (1 + )
13.8... dm3 mol−1 13.8... dm3 mol−1
= 1.62 atm
P1C.3 The virial equation is [1C.3b–21], pVm = RT(1 + B/Vm + C/Vm2 + . . .). The com-
pression factor is defined in [1C.1–20] as Z = Vm /Vm○ , and the molar volume of
a perfect gas, Vm○ is given by Vm○ = RT/p.
It follows that
Vm = (RT/p)(1 + B/Vm + C/Vm2 ) = Vm○ (1 + B/Vm + C/Vm2 )
Vm B C
hence Z = ○ = 1 + +
Vm Vm Vm2
To evaluate this expression, the molar volume is approximated by the molar
volume of a perfect gas under the prevailing conditions
RT (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (273 K)

Vm○ = = = 0.224... dm3 mol−1
p 100 atm
This value of the molar volume is then used to compute Z; note the conversion
of all the volume terms to dm3
B C
Z =1+ + 2
Vm Vm
−21.3 × 10−3 dm3 mol−1 1200 × 10−6 dm6 mol−2
=1+ + = 0.928... = 0.929
0.224... dm3 mol−1 (0.224... dm3 mol−1 )2
The molar volume is computed from the compression factor

Vm Vm
Z= =
Vm○ RT/p
ZRT 0.928... × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (273 K)
hence Vm = =
p 100 atm
= 0.208 dm3 mol−1
P1C.5 In Section 1C.1(b) on page 20 it is explained that at the Boyle temperature

Z = 1 and dZ/dp = 0; this latter condition corresponds to the second virial
coefficient, B or B′ , being zero. The Boyle temperature is found by setting the
given expression for B(T) to zero and solving for T
0 = a + be−c/T hence − a/b = e−c/T

2 2
Taking logarithms gives ln(−a/b) = −c/T 2 hence

1/2 1/2
−c −1131 K2
T =( ) =( )
ln(−a/b) ln[−(−0.1993 bar−1 )/(0.2002 bar−1 )]
= 501.0 K
P1C.7 (a) The molar mass M of H2 O is 18.02 g mol−1 . The mass density ρ is related
to the molar density ρ m by ρ m = ρ/M, and the molar volume is simply
the reciprocal of the molar density Vm = 1/ρ m = M/ρ
M 18.02 × 10−3 kg mol−1

Vm = = = 1.352... × 10−4 m3 mol−1
ρ 133.2 kg m−3
The molar volume is therefore 0.1353 dm3 mol−1

(b) The compression factor Z is given by [1C.2–20], Z = pVm /RT
pVm (327.6 atm) × (0.1352... dm3 mol−1 )

Z= = = 0.6957
RT (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (776.4 K)
(c) The virial equation (up to the second term) in terms of the molar volume
is given by [1C.3b–21]
B
pVm = RT (1 + )
Vm
Division of each side by p gives
RT B
Vm = (1 + )
p Vm
The quantity RT/p is recognised as the molar volume of a perfect gas, Vm○ ,
so it follows that
B Vm B
Vm = Vm○ (1 + ) hence ○ = Z = (1 + )
Vm Vm Vm
In Problem P1C.4 it is shown that B is related to the van der Waals con-
stants by B = b − a/RT; using this, it is then possible to compute the
compression factor
a
B=b− = (0.03049 dm3 mol−1 )
RT
(5.464 atm dm6 mol−2 )
−
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (776.4 K)
= −0.552... dm3 mol−1
B −0.552... dm3 mol−1

Z =1+ =1+ = 0.5914
Vm 0.1352... dm3 mol−1
P1C.9 According to Table 1C.4 on page 25, for the Dieterici equation of state the
critical constants are given by
a a
pc = Vc = 2b Tc =
4e2 b 2 4bR
From the Resource section the values for Xe are Tc = 289.75 K, p c = 58.0 atm,
−1
Vc = 118.8 cm3 mol . The coefficient b is computed directly from Vc
b = Vc /2 = (118.8 × 10−3 dm3 mol−1 )/2 = 0.0594 dm3 mol−1
The expressions for p c and Vc are combined to eliminate b

a a
pc = =
4e2 b 2 4e2 Vc2 /4
This is then rearranged to find a
a = p c e2 Vc2 = (58.0 atm) × e2 × (118.8 × 10−3 dm3 mol−1 )2

= 6.049 atm dm6 mol−2
Alternatively, the expressions for Tc and Vc are combined to eliminate b

a a
Tc = =
4bR 4RVc /2
This is then rearranged to find a
a = 2Tc Vc R
= 2 × (289.75 K) × (118.8 × 10−3 dm3 mol−1 )
× (8.2057 × 10−2 dm3 atm K−1 mol−1 ) = 5.649 atm dm6 mol−2
The two values of a are not the same; their average is 5.849 atm dm6 mol−2 .
From Table 1C.4 on page 25 the expression for the pressure exerted by a Di-
eterici gas is
nRT exp(−a/[RT V /n])
p=
V − nb
With the parameters given the exponential term evaluates to
−(5.849 atm dm6 mol−2 )

exp ( )
(8.2057 × 10−2 dm3 atm K−1 mol−1 )×(298.15 K)×(1.0 dm3 )/(1.0 mol)
= 0.787...
and hence the pressure evaluates to
(1.0 mol)×(8.2057 × 10−2 dm3 atm K−1 mol−1 )×(298.15 K)×(0.787...)

p=
(1.0 dm3 ) − (1.0 mol)×(0.0594 dm3 mol−1 )
= 20.48 atm
P1C.11 The van der Waals equation in terms of the molar volume is given by [1C.5b–
24], p = RT/(Vm − b) − a/Vm2 . Multiplication of both sides by Vm gives
RT Vm a
pVm = −
(Vm − b) Vm
and then division of the numerator and denominator of the first fraction by Vm
gives
RT a
pVm = −
(1 − b/Vm ) Vm
The approximation (1−x)−1 ≈ 1+x+x 2 is the used to approximate 1/(1−b/Vm )
to give
b b2 a
pVm = RT (1 + + 2)−
Vm Vm Vm
The terms in 1/Vm and 1/Vm2 are gathered together to give
1 a b2
pVm = RT (1 + [b − ]+ 2)
Vm RT Vm
This result is then compared with the virial equation in terms of the molar
volume, [1C.3b–21]
B C
pVm = RT (1 + + )
Vm Vm2
This comparison identifies the virial coefficients as
a
B=b− C = b2
RT
√
From the given value C = 1200 cm6 mol−2 it follows that b = C = 34.64 cm3 mol−1 .
Expressed in the usual units this is b = 0.03464 dm3 mol−1 . The value of a is
found by rearranging B = b − a/RT to
a = RT(b − B) = (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (273 K)×

[(0.03464 dm3 mol−1 ) − (−21.7 × 10−3 dm3 mol−1 )]
= 1.262 atm dm6 mol−2
P1C.13 In Section 1C.2(b) on page 24 it is explained that critical behaviour is associated

with oscillations in the isotherms predicted by a particular equation of state,
and that at the critical point there is a point of inflexion in the isotherm. At this
point it follows that
dp d2 p
=0 =0
dVm dVm2
The procedure is first to find expressions for the first and second derivatives.
Then these are both set to zero give two simultaneous equations which can be
solved for the critical pressure and volume.
dp RT 2B 3C d2 p 2RT 6B 12C
=− 2 + 3 − 4 =0 = 3 − 4 + 5 =0
dVm Vm Vm Vm dVm2 Vm Vm Vm
The first of these equations is multiplied through by Vm4 and the second by Vm5
to give
−RT Vm2 + 2BVm − 3C = 0 2RT Vm2 − 6BVm + 12C = 0
The first equation is multiplied by 2 and added to the second, thus eliminating
the terms in Vm2 and giving
4BVm − 6C − 6BVm + 12C = 0 hence Vm = 3C/B
This expression for Vm is then substituted into −RT Vm2 + 2BVm − 3C = 0 to give
(3C)2 3C
−RT + 2B − 3C = 0
B2 B
A term 3C is cancelled and the equation is multiplied through by B 2 to give
−RT(3C) + 2B 2 − B 2 = 0 hence T = B 2 /3RC
Finally the pressure is found by substituting Vm = 3C/B and T = B 2 /3RC into

the equation of state
RT B C
p= − +
Vm Vm2 Vm3
B2 R B B3 CB 3 B3 B3 B3 B3
= − 2
+ 3
= 2
− 2
+ 2
=
3RC 3C 9C 27C 9C 9C 27C 27C 2
In summary, the critical constants are
Vm = 3C/B T = B 2 /3CR p = B 3 /27C 2
P1C.15 The virial equation in terms of the pressure, [1C.3a–21], is (up to the second
term)
pVm = RT (1 + B′ p)
The mass density ρ is given by m/V , and the mass m can be written as nM,
where n is the amount in moles and M is the molar mass. It follows that
ρ = nM/V = M/Vm , where Vm is the molar volume. Rearranging gives Vm =

M/ρ: measurements of the mass density therefore lead to values for the molar
volume.
With this substitution for the molar volume the virial equation becomes
pM p RT
= RT (1 + B′ p) hence = (1 + B′ p)
ρ ρ M
Therefore a plot of p/ρ against p is expected to be a straight line whose slope is
related to B′ ; such a plot is shown in Fig. 1.5.
p/kPa ρ/(kg m−3 ) (p/ρ)/(kPa kg−1 m3 )

12.22 0.225 54.32
25.20 0.456 55.26
36.97 0.664 55.68
60.37 1.062 56.85
85.23 1.468 58.06
101.30 1.734 58.42
58
(p/ρ)/(kPa kg−1 m3 )
56
54
0 20 40 60 80 100
p/kPa
Figure 1.5
(p/ρ)/(kPa kg−1 m3 ) = 0.04610 × (p/kPa) + 53.96
The slope is B′ RT/M

B′ RT
= 0.04610 kg−1 m3
M
For methoxymethane, CH3 OCH3 , M = 2 × 12.01 + 6 × 1.0079 + 16.00 =
46.0674 g mol−1 .
(0.04610 kg−1 m3 ) × (46.0674 × 10−3 kg mol−1 )

B′ = = 8.57 × 10−7 m3 J−1
(8.3145 J K−1 mol−1 ) × (298.15 K)
The units of the result can be simplified by noting that 1 J = 1 kg m2 s−2 , so

1 m3 J−1 = 1 m kg−1 s2 . Recall that 1 Pa = 1 kg m−1 s−2 , so the units of the B′ are
Pa−1 , an inverse pressure, as expected: B′ = 8.57×10−7 Pa−1 or B′ = 0.0868 atm−1 .
The virial coefficient B is found using the result from Problem P1C.14, B = B′ RT
B = B′ RT
= (0.0868 atm−1 ) × (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (298.15 K)
= 2.12 dm3 mol−1
P1C.17 A gas can only be liquefied by the application of pressure if the temperature is
below the critical temperature, which for N2 is 126.3 K.
P1C.19 The compression factor is given by [1C.1–20], Z = Vm /Vm○ = Vm p/RT. The

given equation of state is rearranged to give an expression for Vm after putting
n=1
RT
p(V − nb) = nRT becomes p(Vm − b) = RT hence Vm = +b
p
It follows that the compression factor is given by
Vm p (RT/p + b)p bp
Z= = = 1+
RT RT RT
If Vm = 10b it follows from the previous equation that
Vm p 10bp bp RT
= =1+ hence b=
RT RT RT 9p
With this expression for b the compression factor is computed from Z = 1 +

bp/RT as
bp RT p 1
Z =1+ =1+ = 1 + = 1.11
RT 9p RT 9
P1C.21 The virial equation in terms of the molar volume, [1C.3b–21], is (up to the third
term)
B C
pVm = RT (1 + + )
Vm Vm2
For part (a) only the first two terms are considered, and it then follows that a
plot of pVm against 1/Vm is expected to be a straight line with slope BRT; such
a plot is shown in Fig. 1.6.
p/MPa Vm /(dm3 mol−1 ) (pVm )/(MPa dm3 mol−1 ) (1/Vm )/(dm−3 mol)
0.400 0 6.220 8 2.488 3 0.160 75
0.500 0 4.973 6 2.486 8 0.201 06
0.600 0 4.142 3 2.485 4 0.241 41
0.800 0 3.103 1 2.482 5 0.322 26
1.000 2.479 5 2.479 5 0.403 31
1.500 1.648 3 2.472 5 0.606 69
2.000 1.232 8 2.465 6 0.811 16
2.500 0.983 57 2.458 9 1.016 7
3.000 0.817 46 2.452 4 1.223 3
4.000 0.609 98 2.439 9 1.639 4
linear
(pVm )/(MPa dm3 mol−1 )
2.48 quadratic
2.46
2.44
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

−3
(1/Vm )/(dm mol)
Figure 1.6
(pVm )/(MPa dm3 mol−1 ) = −0.03302 × (1/Vm )/(dm−3 mol) + 2.4931
The slope is BRT
BRT = (−0.03302 MPa dm6 mol−2 )
It is convenient to convert to atm giving BRT = (−0.3259 atm dm6 mol−2 )

hence
(−0.3259 atm dm6 mol−2 )

B=
RT
(−0.3259 atm dm6 mol−2 )
=
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (300 K)
= −0.01324 dm3 mol−1
For part (b) the data points are fitted to polynomial of order 2 in 1/Vm using
mathematical software; the data are a slightly better fit to such a function (see
the dashed line in the graph above) which is
(pVm )/(MPa dm3 mol−1 ) =

0.002652 × (1/Vm )2 /(dm−6 mol2 ) − 0.03748 × (1/Vm )/(dm−3 mol) + 2.494
The coefficient of the term in (1/Vm )2 is CRT
CRT = (0.002652 MPa dm9 mol−3 )
It is convenient to convert to atm giving CRT = (0.02617 atm dm9 mol−3 )

hence
(0.02617 atm dm9 mol−3 )
C=
RT
(0.02617 atm dm9 mol−3 )
=
(8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (300 K)
= 1.063 × 10−3 dm6 mol−2
P1C.23 The van der Waals equation of state in terms of the molar volume is given by
[1C.5b–24], p = RT/(Vm − b) − a/Vm2 . This equation is a cubic in Vm , as is seen
by multiplying both sides by (Vm −b)Vm2 and then gathering the terms together
pVm3 − Vm2 (pb + RT) + aVm − ab = 0
From the Resource section the van der Waals parameters for Cl2 are
a = 6.260 atm dm6 mol−2 b = 5.42 × 10−2 dm3 mol−1
It is convenient to convert the pressure to atm
p = (150 × 103 Pa) × (1 atm)/(1.01325 × 105 Pa) = 1.4804 atm
and to use R = 8.2057 × 10−2 dm3 atm K−1 mol−1 ; inserting all of these values
and the temperature gives the polynomial
1.4804Vm3 − 20.5946Vm2 + 6.260Vm − 0.3393 = 0
The roots of this polynomial are found numerically using mathematical soft-
ware and of these roots only Vm = 13.6 dm3 mol−1 is a physically plausible
value for the molar volume.
The molar volume of a perfect gas under corresponding conditions is
RT (8.2057 × 10−2 dm3 atm K−1 mol−1 ) × (250 K)

Vm = = = 13.9 dm3 mol−1
p 1.48 atm
The molar volume of the van der Waals gas is about 2% smaller than that of
the perfect gas.
Answers to integrated activities

I1.1 In Section 1C.2(a) on page 23 it is argued that b = 4Vmolec N A , where Vmolec is
the volume occupied by one molecule. The collision cross-section σ is defined
in terms of a collision diameter d as σ = πd 2 , and in turn the diameter is
interpreted as twice the radius of the colliding spheres: d = 2r. It follows that
r = (σ/4π)1/2
b = 4Vmolec N A
4 16πN A σ 3/2
= 4 ( πr 3 ) N A = ( )
3 3 4π
16π(6.0221 × 1023 mol−1 ) 0.46 × 10−18 m2
3/2
= ( )
3 4π
= 7.1 × 10−5 m3 mol−1 = 0.071 dm3 mol−1
I1.3 According to the equipartition theorem (The chemist’s toolkit 7 in Topic 2A),
each quadratic contribution to the energy of a molecule contributes 12 kT to
the average energy per molecule. Translational kinetic energy is a quadratic
term, and because translation is possible in three dimensions, the quoted en-
ergy density of 0.15 J cm−3 is the result of three such contributions.
The rotation of a molecule about an axis is also a quadratic contribution to the
energy, and in general three such contributions are expected corresponding to
rotation about three mutually perpendicular axes. However, linear molecules,
such as diatomics, do not show rotation about their long axes, so there are
only two contributions. The contribution of rotation to the energy density will
therefore be 32 of that due to translation
trans. rot.
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
total energy denisty = (0.15 J cm−3 ) + 32 × (0.15 J cm−3 ) = 0.25 J cm−3
2 Internal energy
2A Internal energy
D2A.1 Table 2A.1 on page 37 lists four varieties of work: expansion, surface expan-
sion, extension, and electrical. There is also work associated with processes in
magnetic and gravitational fields which we will not describe in detail.
D2A.3 An isothermal expansion of a gas may be achieved by making sure that the gas
and its container are in thermal contact with a large ‘bath’ which is held at a
constant temperature – that is, a thermostat.
D2A.5 Work is done when a body is moves against an opposing force. For an infinites-
imal displacement in the x-direction, dx, against a force F along that direction
the work done by the body is F dx.
When the energy of a system changes as a result of a temperature difference
between the system and its surroundings, the resulting energy transfer from the
hotter to the cooler body is described as heat. In thermodynamic terms, both
heat and work cause the internal energy of an object to change: if heat ‘flows in’
the internal energy of the body rises, if the body ‘does work’, its internal energy
decreases.
If the internal energy of an object increases, this is interpreted in molecular
terms as the molecules moving up to higher energy levels. If the molecules
drop down to lower levels the resulting energy is available as heat or work.
E2A.1(a) The system is expanding against a constant external pressure, hence the expan-
sion work is given by [2A.6–38], w = −pex ∆V . The change in volume is the
cross-sectional area times the linear displacement
∆V = (50 cm2 ) × (15 cm) = 750 cm3 = 7.5 × 10−4 m3
The external pressure is 1.0 atm = 1.01325 × 105 Pa, therefore the expansion
work is
w = −(1.01325 × 105 Pa) × (7.5 × 10−4 m3 ) = −76 J
Note that the volume is expressed in m3 . The relationships 1 Pa = 1 kg m−1 s−2

and 1 J = 1 kg m2 s−2 are used to verify the units of the result.
40 2 INTERNAL ENERGY
E2A.2(a) For all cases ∆U = 0, because the internal energy of a perfect gas depends on
the temperature alone.
(i) The work of reversible isothermal expansion of a perfect gas is given by

[2A.9–39]
Vf
w = −nRT ln ( )
Vi
30.0 dm3
= −(1.00 mol) × (8.3145 J K−1 mol−1 ) × (293.15 K) × ln ( )
10.0 dm3
= −2.68 × 103 J = −2.68 kJ
Note that the temperature is expressed in K in the above equation. Using

the First Law of thermodynamics, [2A.2–36], gives
q = ∆U − w = 0 − (−2.68 kJ) = +2.68 kJ
(ii) The final pressure of the expanding gas is found using the perfect gas law,
[1A.4–8]
nRT (1.00 mol) × (8.3145 J K−1 mol−1 ) × (293.15 K)

pf = =
Vf (30.0 × 10−3 m3 )
= 8.12... × 104 Pa
This pressure equals the constant external pressure against which the gas
is expanding, therefore the work of expansion is
w = −pex × ∆V = (8.12... × 104 Pa) × (30.0 × 10−3 m3 − 10.0 × 10−3 m3 )

= −1.62 × 103 J = −1.62 kJ
and hence q = +1.62 kJ

(iii) Free expansion is expansion against zero force, so w = 0 and therefore
q = 0 as well.
E2A.3(a) For a perfect gas at constant volume pi /Ti = pf /Tf therefore,
Tf 400 K
pf = pi × = (1.00 atm) × ( ) = 1.33 atm
Ti 300 K
The change in internal energy at constant volume is given by [2A.15b–43]
3
∆U = nC V ,m ∆T = (1.00 mol) × ( × 8.3145 J K−1 mol−1 ) × (400 K − 300 K)
2
= +1.25 × 103 J = +1.25 kJ
The volume of the gas is constant, so the work of expansion is zero, w = 0 . The
First Law of thermodynamics gives q = ∆U − w = +1.25 kJ − 0 = +1.25 kJ .
E2A.4(a) (i) The work of expansion against constant external pressure is given by [2A.6–
38]
w = −pex ∆V
133.3 Pa
= −(200 Torr) × ( ) × (3.3 × 10−3 m3 ) = −88 J
1 Torr
Note that the pressure is expressed in Pa and the change in volume in m3 ,
to give the work in J.
(ii) The work done in a reversible isothermal expansion is given by [2A.9–39],
w = −nRT ln(Vf /Vi ). The amount in moles of methane is
m (4.50 g)
n= = = 0.280... mol
M (16.0416 g mol−1 )
w = −(0.280... mol) × (8.3145 J K−1 mol−1 ) × (310 K)
[12.7 + 3.3] dm3
× ln ( ) = −1.7 × 102 J
12.7 dm3
Note that the modulus of the work done in a reversible expansion is greater
than the work for expansion against constant external pressure because
the latter is an irreversible process.
E2A.5(a) The chemist’s toolkit 7 in Topic 2A gives an explanation of the equipartition
theorem. The molar internal energy is given by
Um = 1
2
× (νt + νr + 2νv ) × RT
where νt is the number of translational degrees of freedom, νr is the number
of rotational degrees of freedom and νv is the number of vibrational degrees of
freedom. As each gas molecule can move independently along the x, y and z
axis, the number of translational degrees of freedom is three.
(i) Molecular iodine is a diatomic molecule, therefore it has two degrees of
rotational freedom. On account of its heavy atoms, molecular iodine is
likely to have one degree of vibrational freedom at room temperature.
Therefore, the molar internal energy of molecular iodine gas at room
temperature is
Um = 1
2
× (3 + 2 + 2) × RT = 7
2
× (8.3145 J K−1 mol−1 ) × (298.15 K)
= 8.7 kJ mol−1
(ii) and (iii) Both methane (tetrahedral) and benzene (planar) have three degrees of
rotational freedom. At room temperature it is unlikely that any of their
vibrational modes would be excited, therefore both are expected to have
approximately the same internal energy at room temperature:
Um = 1
2
× (3 + 3 + 0) × RT = 3 × (8.3145 J K−1 mol−1 ) × (298.15 K)
= 7.4 kJ mol−1
E2A.6(a) A state function is a property with a value that depends only on the current state
of the system and is independent of how the state has been prepared. Pressure,
temperature and enthalpy are all state functions.
P2A.1 From the equipartition theorem (The chemist’s toolkit 7 in Topic 2A), the molar
internal energy is given by
Um = 1
2
× (νt + νr + 2νv ) × RT

of rotational degrees of freedom and νv is the number of vibrational degrees
of freedom. Each gas molecule can move independently along the x, y and z
axis giving rise to three translational degrees of freedom. Carbon dioxide is
a linear molecule therefore it has two rotational degrees of freedom. None of
the vibrational modes of carbon dioxide are likely to be significantly excited at
room temperature.
U= 1
2
× (3 + 2 + 0) × RT = 5
2
× (8.3145 J K−1 mol−1 ) × (298.15 K)
= 6.2 kJ mol−1
P2A.3 The definition of work is given by [2A.4–37], dw = −∣F∣dz. Integrating both

sides gives:
l
∫ dw = ∫ F(x) dx
0
l l 1
w=∫ kf (x)x dx = ∫ (a − bx 2 ) x dx
0 0
1 2 5 l 1 2 5
= [ ax 2 − bx 2 ] = al 2 − bl 2
2 5 0 2 5
Note that the second term arises from the non Hooke’s Law behaviour of the
elastomer, reducing the overall work done.
P2A.5 (a) The natural logarithm can be expanded using the Taylor series as ln(1 +
ν) ≈ ν + ν 2 /2! + ν 3 /3! + ..., which, for ν << 1, can be approximated as
ln(1 + ν) ≈ ν, and similarly, ln(1 − ν) ≈ −ν. Therefore,
kT kT νkT
F= [ln(1 + ν) − ln(1 − ν)] ≈ [ν − (−ν)] =
2l 2l l
Because ν = n/N, it follows that
νkT nkT
F= =
l Nl
(b) Hooke’s law predicts F = const × x, that is the restoring force is directly
proportional to the displacement. Using n = x/l, the expression for the
force obtained in part (a) is rewritten as
nkT kTx
= ≡ const × x
Nl Nl2
Therefore Hooke’s law applies and kT/N l 2 is the force constant.
P2A.7 The van der Waals equation of state is given by [1C.5b–24],

nRT n2 a
p= − 2
V − nb V
The expansion work is given by [2A.6–38], dw = −pex dV . For a reversible
expansion pex is always equal to the pressure of the gas so
nRT n 2 a
dw = −pgas dV = − ( − 2 ) dV
V V
Integrating both sides give
VfnRT n2 a
w = −∫ − 2 dV
Vi V − nb V
Vf dV Vf dV
= −nRT ∫ + n2 a ∫
Vi V − nb Vi V 2
Vf − nb 1 1
= −nRT ln ( ) − n2 a ( − )
Vi − nb Vf Vi
The work done during the isothermal reversible expansion in the various cases
is calculated below and the results portrayed in Fig. 2.1.
(a) A perfect gas.
Vf
w = −nRT ln ( ) = −(1.0 mol) × (8.3145 J K−1 mol−1 ) × (298 K)
Vi
2.0 dm3
× ln ( ) = −1.7 kJ
1.0 dm3
(b) A van der Waals gas in which repulsive forces dominate.
Using the general expression derived above with a = 0 and b = 5.11 ×
10−2 dm3 mol−1
w = −(1.0 mol) × (8.3145 J K−1 mol−1 ) × (298 K)
(2.0 dm3 ) − (1.0 mol) × (5.11 × 10−2 dm3 mol−1 )
× ln ( ) = −1.8 kJ
(1.0 dm3 ) − (1.0 mol) × (5.11 × 10−2 dm3 mol−1 )
(c) A van der Waals gas in which attractive forces dominate.
−2
Using the general expression derived above with a = 4.2 dm6 atm mol
and b = 0. Constant a in SI units is
−2 1 m6 1.01325 × 105 Pa
a = (4.2 dm6 atm mol ) × ( 6)×( )
6
10 dm 1 atm
−2
= 0.425... m6 Pa mol
2.0 dm3
w = −(1.0 mol) × (8.3145 J K−1 mol−1 ) × (298 K) × ln ( )
1.0 dm3
−2 1 1
− (1.0 mol)2 × (4.25... × 105 dm6 Pa mol ) × ( 3 − )
2.0 dm 1.0 dm3
= −1.71...kJ + 0.212...kJ = −1.5 kJ
30
perfect
repulsion only
25 attraction only
p/atm 20
15
10
0.5 1.0 1.5 2.0 2.5
3
V /dm
Figure 2.1
P2A.9 (a) The virial equation of state is given by [1C.3b–21]. The first three terms
are
1 B C
p = RT ( + + )
Vm Vm2 Vm3
Using Vm = V /n gives
1 nB n 2 C
p = nRT ( + + 3 )
V V2 V
The work of expansion is calculated as

Vf Vf 1 nB n 2 C
w = −∫ pdV = − ∫ nRT ( + + 3 ) dV
Vi Vi V V2 V
Vi 1 1 1 1 1
= −nRT ln + n 2 RTB ( − ) + n 3 RTC ( 2 − 2 )
Vf Vf Vi 2 Vf Vi
−1
From Table 1C.1, at 273 K, B = −21.7 cm3 mol and from the problem
−2
C = 1.2 × 103 cm6 mol , therefore
nRT = (1.0 mol) × (8.3145 J K−1 mol−1 ) × (273 K) = 2.26... × 103 J

n 2 RTB = n × (nRT) × B = (1.0 mol) × (2.26... × 103 J)
−1
× (−21.7 cm3 mol )
= −4.92... × 104 J cm3
2
n RTC
1 3
= 1
2
× n 2 × (nRT) × C
−2
= 1
2
× (1.0 mol)2 × (2.26... × 103 J) × (1.2 × 103 cm6 mol )
= 1.36... × 106 J cm6
1000 cm3
w = −(2.26... × 103 J) × ln ( )
500 cm3
1 1
+ (−4.92... × 104 J cm3 ) × ( − )
1000 cm3 500 cm3
1 1
+ (1.36... × 106 J cm6 ) × ( 3 2
− )
(1000 cm ) (500 cm3 )2
= −1.5 kJ
(b) For a perfect gas, the work of reversible isothermal expansion is given by
[2A.9–39]
Vf
w = −nRT ln ( )
Vi
1000 cm3
= −(1.0 mol) × (8.3145 J K−1 mol−1 ) × (273 K) × ln ( )
500 cm3
= −1.6 × 103 J = −1.6 kJ
2B Enthalpy
D2B.1 If a substance is heated at constant volume all of the energy as heat is trans-
formed into internal energy of the substance. If the same process is carried out
under conditions of constant pressure some of the energy as heat will be used
to expand the substance against the external pressure and so less of the energy
as heat is transformed into internal energy. This effect is largest for gases whose
volumes change much more rapidly with temperature than do solids or liquids.
The temperature of a substance is related to its internal energy, and therefore

as energy as heat is supplied the temperature increases. For a given amount
of energy as heat this increase in internal energy is smaller for the constant-
pressure case than for the constant-volume case. Therefore the rise in temper-
ate is smaller for the constant-pressure process, and this implies that the heat
capacity is greater.
E2B.1(a) (i) The heat capacity can be expressed as C p = a + bT where a = 20.17 J K−1
and b = 0.3665 J K−2 . Integrating the relationship dH = C p dT on both
sides gives
T2 T2 T2 T2
∫ dH = ∫ C p dT = ∫ (a + bT)dT = [aT + 21 bT 2 ]T
T1 T1 T1 1
H(T2 )−H(T1 ) = na(T2 − T1 ) + 1

2
nb(T22 − T12 )
−1
= (20.17 J K ) × (373.15 K − 298.15 K)
+ 21 × (0.3665 J K−2 ) × [(373.15 K)2 − (298.15 K)2 ]
= +10.7... kJ = +10.7 kJ
Under constant pressure conditions ∆H = qp = +10.7 kJ .
The work of expansion against constant pressure pex is given by [2A.6–
38], w = −pex ∆V = −pex (Vf − Vi ). Assume that the gas is in mechanical
equilibrium with its surroundings, therefore pex is the same as the pres-
sure of the gas, p. The initial and final volumes are calculated from Tf and
Ti by Vf = nRTf /p and Vi = nRTi /p, therefore Vf − Vi = (Tf − Ti )nR/p.
Hence
nR
w = −p × (Tf − Ti ) = −nR∆T
p
= −(1.00 mol) × (8.3145 J K−1 mol−1 ) × (373.15 K − 298.15 K)
= −6.23... × 102 J = −624 J
∆U = q + w = (+10.7... kJ) + (−0.623... kJ) = +10.1 kJ
(ii) The energy and enthalpy of a perfect gas depends on the temperature
alone, hence ∆H and ∆U is the same as above, that is ∆H = +10.7 kJ and
∆U = +10.1 kJ . Under constant volume conditions there is no expansion
work, w = 0 , therefore the heat is equal to the change in internal energy,
qV = +10.1 kJ .
E2B.2(a) Under constant pressure qp = ∆H, therefore
qp = ∆H = nC p,m ∆T = (3.0 mol) × (29.4 J K mol−1 ) × (285 K − 260 K)
= 2.20... kJ = +2.2 kJ
The definition of enthalpy is given by [2B.1–44], H = U + pV . For a change
at constant pressure, it follows that ∆H = ∆U + p∆V , where ∆V = Vf − Vi .
If the gas is assumed to be perfect, then Vf = nRTf /p and Vi = nRTi /p, so
∆V = (Tf − Ti )nR/p. Hence p∆V = nR(Tf − Ti ) = nR∆T.
∆U = ∆H − nR∆T
= (2.20... kJ) − (3.0 mol) × (8.3145 J K−1 mol−1 ) × (285 K − 260 K)
= +1.6 kJ
E2B.3(a) The heat transferred under constant pressure equals the change in enthalpy
of the system, [2B.2b–45], qp = ∆H. The relationship between the change in
enthalpy, change in temperature and the heat capacity is given by [2B.6b–47]
∆H (229 J)
C p,m = = = 29.9... J K mol−1 = 30 J K−1 mol−1
n∆T (3.0 mol) × (2.55 K)
For a perfect gas C p,m − C V ,m = R, [2B.9–47], therefore
C V ,m = C p,m −R = (29.9... J K−1 mol−1 )−(8.3145 J K−1 mol−1 ) = 22 J K−1 mol−1
E2B.4(a) In the reaction 2 moles of gas are formed from 4 moles of gas. The relationship
between ∆H and ∆U is given by [2B.3–46]
∆Hm −∆Um = ∆ng RT = (−2.0)×(8.3145 J K−1 mol−1 )×(298 K) = −5.0 kJ mol−1
P2B.1 10 g of benzene is n = m/M = (10 g)/(78.1074 g mol−1 ) = 0.128... mol. The
heat needed to vaporize 10 g of benzene under constant pressure is qp = n ×
∆ vap Hm = (0.128... mol) × (30.8 kJ mol−1 ) = 3.94... kJ. A current I flowing
through a potential ∆ϕ corresponds to a power of I∆ϕ. If the current flows for
a time interval ∆t the energy is power × time , that is q = I∆ϕ∆t. Therefore
∆t = q/(I∆ϕ).
qp (3.94... × 103 J)
∆t = = = 657 s = 11 min
I∆ϕ (0.50 A) × (12 V)
The relationships 1 A = 1 C s−1 and 1 V = 1 J C−1 are used in the last step.
P2B.3 Fitting the data set using a computer program to an expression in the form of
C −p,m
○
(T) = a+bT+cT −2 yields a = 48.0 J K−1 mol−1 , b = 6.49×10−3 J K−2 mol−1
and c = −9.33 × 105 J K mol−1 . The change in enthalpy in response to a change
in temperature at constant pressure is given by [2B.6a–47], dHm = C p,m dT.
Substituting in the expression for C p,m and integrating both sides gives
dHm = (a + bT + cT −2 ) dT
(T2 ) T2
∫ dHm = ∫ (a + bT + cT −2 ) dT
(T1 ) T1
By using Integral A.1 in the Resource section for each term, it follows that
1 1
Hm (T2 ) − Hm (T1 ) = a(T2 − T1 ) + 21 b(T22 − T12 ) − c ( − )
T2 T1
= (48.0 J K−1 mol−1 ) × (1500 K − 298.15 K)
+ 21 × (6.49 × 10−3 J K−2 mol−1 )
× [(1500 K)2 − (298.15 K)2 ]
1 1
− (−9.33 × 105 J K mol−1 ) × ( − )
1500 K 298.15 K
= 62.2 kJ
P2B.5 Since the volume is fixed no expansion work is done, hence w = 0 and
∆U = qV = +2.35 kJ . From H = U + pV it follows that ∆H = ∆U + V ∆p at
constant volume. Using the van der Waals equation of state, [1C.5b–24],
RT a R∆T
p= − 2 it follows that ∆p =
Vm − b Vm Vm − b
Note that the term in a cancels because the volume is constant. Therefore
RV ∆T
∆H = ∆U + v∆p = ∆U +
Vm − b
From the given data Vm = V /n = (15.0 dm3 )/(2.0 mol) = 7.5 dm3 mol−1 ,
∆T = 341 K − 300 K = 41 K and from Table 1C.3 b = 4.29 × 10−2 dm3 mol−1 .
Therefore
(8.3145 J K−1 mol−1 ) × (15.0 dm3 ) × (41 K)
∆H = (+2.35 × 103 J) +
(7.5 dm3 mol−1 ) − (4.29 × 10−2 dm3 mol−1 )
= 3.03... × 103 J = 3.0 kJ
2C Thermochemistry
D2C.1 When a system is subjected to constant pressure conditions, and only expan-
sion work can occur, the energy supplied as heat is the change in enthalpy of the
system. Thus enthalpy changes in the system can be determined by measuring
the amount of heat supplied under constant-pressure conditions.
A very simple example often encountered in elementary laboratory classes is a
thermally insulated vessel (for example, a foam plastic coffee cup) left open to
the atmosphere: the heat released in the reaction is determined by measuring
the change in temperature of the contents.
For a combustion reaction a constant-pressure flame calorimeter (Section 2B.1(b)
on page 45) may be used. In this apparatus a certain amount of substance
burns in a supply of oxygen and the rise in temperature is monitored. More
sophisticated methods include isothermal titration calorimetry and differential
scanning calorimetry, both described in Section 2C.4 on page 54.
D2C.3 The main objection to the use of the term ‘heat’ to describe the energy change
associated with a physical or chemical process is that heat is not a state function.
The value of the heat therefore depends on the path chosen.
If, in fact, the processes being described takes place at constant pressure, the
heat is equal to the enthalpy change. Because enthalpy is a state function, the
heat measured under these circumstance is a property of the physical or chem-
ical change itself, and not affected by the path taken, and so is a meaningful and
useful quantity to discuss.
It is more appropriate to talk about the enthalpy change, the change in a state
function, rather to talk about the heat which, because of an unstated restriction,
just so happens to have the same value.
E2C.1(a) The relationship between ∆ r H and ∆ r U is given by [2B.3–46], ∆ r H = ∆ r U +

∆ng RT where ∆ng is the change in the amount of gas molecules in the reaction.
For this reaction ∆ng = 2 mol + 3 mol − 3 mol = +2 mol
∆ r H −○ = ∆ r U −○ + ∆ng RT
= (−1373 × 103 J mol−1 ) + (+2.0 mol) × (8.3145 J K−1 mol−1 ) × (298 K)
= −1368 kJ mol−1
E2C.2(a) (i) The standard reaction enthalpy at 298 K is calculated from the standard
enthalpies of formation at 298 K using [2C.5a–53]
∆ r H −○ (298 K) = −
○
∑ ν∆ f H − ∑ ν∆ f H −○
products reactants
= ∆ f H −○ (CO, g) + ∆ f H −○ (H2 , g) − ∆ f H −○ (H2 O, g) − ∆ f H −○ (graphite, s)

= (−110.53 kJ mol−1 ) + 0 − (−241.82 kJ mol−1 ) − 0
= +131.29 kJ mol−1
The relationship between ∆ r H and ∆ r U is given by [2B.3–46], ∆ r H =

∆ r U + ∆ng RT where ∆ng is the change in the amount of gas molecules
in the reaction. For this reaction ∆ng = 1 mol + 1 mol − 1 mol = +1 mol
∆ r U −○ (298 K) = ∆ r H −○ (298 K) − ∆ng RT

= (+131.29 × 103 J mol−1 )
− (+1.0 mol) × (8.3145 J K−1 mol−1 ) × (298 K)
= +128.81 kJ mol−1
(ii) The difference of the molar heat capacities of products and reactants is
calculated using [2C.7b–53] and the data from Table 2C.4
∆ r C −p○ = −
○
∑ νC p,m − ∑ νC −p,m
○
products reactants
= C −p,m
○
(CO, g) + C −p,m
○
(H2 , g) − C −p,m
○
(H2 O, g) − C −p,m
○
(graphite, s)
= (29.14 J K−1 mol−1 ) + (28.824 J K−1 mol−1 ) − (33.58 J K−1 mol−1 )
− (8.527 J K−1 mol−1 ) = +15.86 J K−1 mol−1
It is assumed that all heat capacities are constant over the temperature
range of interest, therefore the integrated form of Kirchhoff ’s Law is ap-
plicable, [2C.7d–54]
∆ r H −○ (478 K) = ∆ r H −○ (298 K) + ∆T∆ r C −p○

= (+131.29 × 103 J mol−1 )
+ (478 K − 298 K) × (+15.86 J K−1 mol−1 )
= +134.1 kJ mol−1
∆ r U (478 K) = ∆ r H −○ (478 K) − ∆ng RT
−
○
= (+134.1 × 103 J mol−1 )

− (+1.0 mol) × (8.3145 J K−1 mol−1 ) × (478 K)
= +130 kJ mol−1
E2C.3(a) The reaction equation C(graphite, s) + O2 (g) ÐÐ→ CO2 (g) represents the for-
mation of CO2 (g) from its elements in their reference states, therefore ∆ r H −○ (298 K) =
∆ f H −○ (CO2 , g) = −393.51 kJ mol−1 . The variation of standard reaction en-
thalpy with temperature is given by Kirchhoff ’s Law, [2C.7a–53]
T2
∆ r H −○ (T2 ) = ∆ r H −○ (T1 ) + ∫ ∆ r C −p○ dT
T1
The difference of the molar heat capacities of products and reactants is given
by [2C.7b–53]
∆ r C −p○ = −
○
○
products reactants
= C −p,m
○
(CO2 , g) − C −p,m
○
(O2 , g) − C −p,m
○
(graphite, s)
The heat capacities in Table 2B.1 are expressed in the form of C −p,m
○
= a + bT +
−
○
c/T therefore ∆ r C p = ∆a + ∆bT + ∆c/T where ∆a = a(CO2 , g) − a(O2 , g) −
2 2
a(graphite, s) and likewise for ∆b and ∆c.
∆a = (44.22 J K−1 mol−1 ) − (29.96 J K−1 mol−1 ) − (16.86 J K−1 mol−1 )

= −2.60 J K−1 mol−1
∆b = (8.79 × 10−3 J K−2 mol−1 ) − (4.18 × 10−3 J K−2 mol−1 )
− (4.77 × 10−3 J K−2 mol−1 ) = −0.16 × 10−3 J K−2 mol−1
∆c = (−8.62 × 105 J K mol−1 ) − (−1.67 × 105 J K mol−1 )
− (−8.54 × 105 J K mol−1 ) = +1.59 × 105 J K mol−1
Integrating Kirchhoff ’s Law gives

T2 ∆c
∆ r H −○ (T2 ) = ∆ r H −○ (T1 ) + ∫ (∆a + ∆bT + ) dT
T1 T2
−
○ 1 1
= ∆ r H (T1 ) + ∆a(T2 − T1 ) + 21 ∆b(T22 − T12 ) − ∆c ( − )
T2 T1
= (−393.51 × 103 J mol−1 ) + (−2.60 J K−1 mol−1 ) × (500 K − 298 K)
+ 12 × (−0.16 × 10−3 J K−2 mol−1 ) × [(500 K)2 − (298 K)2 ]
1 1
− (1.59 × 105 J K mol−1 ) × ( − )
500 K 298 K
= −394 kJ mol−1
E2C.4(a) Tetrachloromethane is vaporized at constant pressure, therefore q = ∆H.
q = ∆H = n∆ vap H −○ = (0.75 mol) × (30.0 kJ mol−1 ) = +22.5 kJ
The work of expansion under constant pressure is given by [2A.6–38], w =

−pex ∆V . Note that ∆V = Vf because the final state (gas) has a much larger
volume than the initial state (liquid). The perfect gas law is used to calculate
Vf .
nRT
w = −pex ∆V = −pex (Vf − Vi ) ≈ −pex Vf = −pex × = −nRT
pex
= −(0.75 mol) × (8.3145 J K−1 mol−1 ) × (250 K) = −1.55... kJ = −1.6 kJ
The First Law of thermodynamics [2A.2–36] gives
∆U = q + w = (+22.5 kJ) + (−1.55... kJ) = +21 kJ
E2C.5(a) The equation for combustion of ethylbenzene is C8 H10 (l) + 10.5 O2 (g) ÐÐ→
8 CO2 (g) + 5 H2 O(l). The standard enthalpy of combustion is calculated us-
ing [2C.5a–53] and using values for the standard enthalpies of formation from
Table 2C.7.
∆ c H −○ = −
○
∑ ν∆ f H − ∑ ν∆ f H −○
products reactants
= 8∆ f H −○ (CO2 , g) + 5∆ f H −○ (H2 O, l) − 21
∆ H −○ (O2 , g) −
2 f
∆ f H −○ (C8 H10 , l)
= 8 × (−393.51 kJ mol−1 ) + 5 × (−285.83 kJ mol ) − 0 − (−12.5 kJ mol−1 )
−1
= −4.57 × 103 kJ mol−1
E2C.6(a) The standard enthalpy of formation of HCl(aq) is ∆ r H −○ for the reaction

1
H (g) + 21 Cl2 (g)
2 2
ÐÐ→ HCl(aq)
Because HCl is a strong acid, the reaction is effectively

1
H (g) + 12 Cl2 (g)
2 2
ÐÐ→ H+ (aq) + Cl− (aq)
By definition, ∆ f H −○ (H+ , aq) = 0, so ∆ r H −○ for this reaction is ∆ f H −○ (Cl− , aq).
∆ r H −○ = ∆ f H −○ (Cl− , aq) = −167 kJ mol−1
E2C.7(a) The equation for the combustion of naphthalene is C10 H8 (s) + 12 O2 (g) ÐÐ→
10 CO2 (g) + 4 H2 O(l). ∆ c H −○ is computed using the thermochemical data from
Tables 2C.6 and 2C.7 and equation [2C.5a–53]
∆ c H −○ = −
○
∑ ν∆ f H − ∑ ν∆ f H −○
products reactants
= 10∆ f H −○ (CO2 , g) + 4∆ f H −○ (H2 O, l)

− 12∆ f H −○ (O2 , g) − ∆ f H −○ (C10 H8 , s)
= 10 × (−393.51 kJ mol−1 ) + 4 × (−285.83 kJ mol−1 )
− 0 − (+78.53 kJ mol−1 ) = −5.16 × 103 kJ mol−1
In a bomb calorimeter the heat is at constant volume and is given by qv =

n∆ c U −○ . ∆ c U −○ is related to ∆ c H −○ by [2B.3–46], ∆ r H −○ = ∆ r U −○ + ∆ng RT, where
∆ng is the change in the amount of gas molecules in the reaction. In this case
∆ng = 10 mol − 12 mol = −2 mol, hence
∆ c U −○ = ∆ c H −○ − ∆ng RT
= (−5.15... × 106 J) − (−2.0 mol) × (8.3145 J K−1 mol−1 ) × (298 K)
= −5.15... × 103 kJ mol−1
The heat released in the bomb calorimeter on combustion of 120 mg of naph-

thalene (M = 128.1632 g mol−1 ) is
0.120 g
qv = n∆ c U −○ = ( ) × (−5.15... × 103 kJ mol−1 ) = −4.82 ...kJ
128.1632 g mol−1
Therefore the calorimeter constant is

∣qv ∣ (4.82... kJ)
C= = = 1.58... kJ K−1 = 1.58 kJ K−1
∆T (3.05 K)
The chemical equation for combustion of phenol is C6 H6 O(s) + 7 O2 (g) ÐÐ→

6 CO2 (g)+3 H2 O(l). Following the same logic as above, ∆ c H −○ = −3054 kJ mol−1 ,
∆ng = −1 mol and ∆ c U −○ = −3.05... × 103 kJ mol−1 . The heat released on
combustion of 150 mg of phenol (M = 94.10 g mol−1 ) is
0.150 g
qv = n∆ c U −○ = ( ) × (−3.05... × 103 kJ mol−1 ) = −4.86... kJ
94.1074 g mol−1
Therefore ∆T = ∣qv ∣/C = (4.86... kJ)/(1.58...kJ K−1 ) = +3.07 K

E2C.8(a) (i) Reaction(3) is reaction(2) − 2 × reaction(1), therefore

∆ r H −○ (3) = ∆ r H −○ (2) − 2∆ r H −○ (1)
= (−483.64 kJ mol−1 ) − 2 × (−184.62 kJ mol−1 )
= −114.40 kJ mol−1
The relationship between ∆ r H and ∆ r U is given by [2B.3–46], ∆ r H =
∆ r U + ∆ng RT where ∆ng is the change in the amount of gas molecules
in the reaction. For this reaction ∆ng = 2 mol + 2 mol − 4 mol − 1 mol =
−1 mol
∆ r U −○ = ∆ r H −○ − ∆ng RT
= (−114.40 × 103 J mol−1 )
− (−1.0 mol) × (8.3145 J K−1 mol−1 ) × (298 K) = −112 kJ mol−1
(ii) Reaction(1) represents the formation of 2 moles of HCl(g) from its ele-
ments in their reference states, therefore the standard enthalpy of forma-
tion of HCl(g) is
∆ f H −○ (HCl, g) = 12 ∆ r H −○ (1) = 21 ×(−184.62 kJ mol−1 ) = −92.31 kJ mol−1

Reaction(2) represents the formation of 2 moles of H2 O(g) from its ele-
ments in their reference states, therefore
∆ f H −○ (H2 O, g) = 12 ∆ r H −○ (2) = 12 ×(−483.64 kJ mol−1 ) = −241.82 kJ mol−1
P2C.1 At constant pressure the temperature rise is given by [2B.7–47], qp = C p ∆T.
The heat capacity is approximated as nC −p,m
○
(H2 O, l) where n is the amount in
moles. Therefore
qp (1.0 × 107 J)
∆T = = = 36.8... K = 37 K
nC −p,m
○
( 18.0158 g mol−1 ) × (75.29 J K−1 mol−1 )
65000 g
Assuming that H2 O (298 K, l) ÐÐ→ H2 O (298 K, g) is the main process respon-

sible for heat loss, and using data from Table 2C.1, the amount of water to
evaporate is
qp (1.0 × 107 J)
n= = = 2.27... × 102 mol
∆ vap H −○ (44.016 × 103 J mol−1 )
The corresponding mass is
m = nM = (2.27... × 102 mol) × (18.0158 g mol−1 ) = 4.1 kg
P2C.3 (a) The combustion equation for cyclopropane is C3 H6 (g) + 4.5 O2 (g) ÐÐ→
3 CO2 (g) + 3 H2 O(l). Using data from Table 2C.7 and applying [2C.5a–53]
for this case
∆ c H −○ = 3∆ f H −○ (CO2 , g) + 3∆ f H −○ (H2 O, l) − ∆ f H −○ (C3 H6 , g)
Therefore
∆ f H −○ (C3 H6 , g) = 3∆ f H −○ (CO2 , g) + 3∆ f H −○ (H2 O, l) − ∆ c H −○

= 3 × (−393.51 kJ mol−1 ) + 3 × (−285.83 kJ mol−1 ) − (2091 kJ mol−1 )
= +52.9... kJ mol−1 = +52.98 kJ mol−1
(b) The reaction is cyclopropane ÐÐ→ propene
∆ r H −○ = ∆ f H −○ (propene, g) − ∆ f H −○ (cyclopropane, g)
= (+20.42 kJ mol−1 ) − (+52.9... kJ mol−1 ) = −32.56 kJ mol−1
P2C.5 The combustion of 0.825 g of benzoic acid (M = 122.1174 g mol−1 ) is used to

determine the calorimeter constant. The heat released in the calorimeter is
0.825 g
qv = n∆ c U −○ = ( ) × (−3251 kJ mol−1 ) = −21.9... kJ
122.1174 g mol−1
Therefore the calorimeter constant is
∣qv ∣ (21.9... kJ)
C= = = 11.3... kJ K−1
∆T (1.940 K)
It follows that the heat released during the combustion of ribose is ∣qv ∣ = C∆T =
(11.3... kJ K−1 ) × (0.910 K) = 10.3... kJ, and so the standard internal energy of
combustion is
−∣qv ∣ −(10.3... kJ)
∆ c U −○ = = = −2.12... × 103 kJ mol−1
n (0.727 g)/(150.129 g mol−1 )
The combustion reaction is C5 H10 O5 (s) + 5 O2 (g) ÐÐ→ 5 CO2 (g) + 5 H2 O(l).
There is no change in the number of gaseous species when going from reactants
to products, therefore ∆ c U −○ = ∆ c H −○ = −2.12... × 103 kJ mol−1 . Applying
[2C.5a–53] for this reaction and using the values for the standard enthalpies
of formation from Table 2C.7 gives
∆ c H −○ = 5∆ f H −○ (CO2 , g) + 5∆ f H −○ (H2 O, l) − 5∆ f H −○ (O2 , g)

− ∆ f H −○ (C5 H10 O5 , s)
∆ f H −○ (C5 H10 O5 , s) = 5∆ f H −○ (CO2 , g) + 5∆ f H −○ (H2 O, l)
− 5∆ f H −○ (O2 , g) − ∆ c H −○
= 5 × (−393.51 kJ mol−1 ) + 5 × (−285.83 kJ mol−1 )
− 0 − (−2.12... × 103 kJ mol−1 )
= −1.27 × 103 kJ mol−1
P2C.7 The chemical reaction for the combustion is C60 (s) + 60 O2 (g) ÐÐ→ 60 CO2 (g).
The standard internal energy of combustion is
∆ c U −○ = (−36.0334 kJ g−1 ) × (60 × 12.01 g mol−1 ) = −2.59... × 104 kJ mol−1

There is no change in the number of gaseous species when going from re-
actants to products, therefore ∆ c U −○ = ∆ c H −○ = −2.59... × 104 kJ mol−1 =
−25966 kJ mol−1 . Applying [2C.5a–53] for this case gives
∆ c H −○ = 60∆ f H −○ (CO2 , g) − 60∆ f H −○ (O2 , g) − ∆ f H −○ (C60 , s)
Therefore
∆ f H −○ (C60 , s) = 60∆ f H −○ (CO2 , g) − 60∆ f H −○ (O2 , g) − ∆ c H −○

= 60 × (−393.51 kJ mol−1 ) − 0 − (−2.59... × 104 kJ mol−1 )
= +2355.1 kJ mol−1
P2C.9 The reaction equation for combustion of methane is CH4 (g) + 2 O2 (g) ÐÐ→
CO2 (g) + 2 H2 O(g). The standard enthalpy of combustion can be calculated
using [2C.5a–53] and using the values for the standard enthalpies of formation
from Tables 2C.6 and 2C.7
∆ c H −○ = −
○
∑ ν∆ f H − ∑ ν∆ f H −○
products reactants
= ∆ f H −○ (CO2 , g) + 2∆ f H −○ (H2 O, g) − 2∆ f H −○ (O2 , g) − ∆ f H −○ (CH4 , g)

= (−393.51 kJ mol−1 ) + 2 × (−241.82 kJ mol−1 ) − 0 − (−74.81 kJ mol−1 )
= −802.3 kJ mol−1
The variation of standard reaction enthalpy with temperature is given by Kirch-

hoff ’s Law, [2C.7a–53]
T2
∆ r H −○ (T2 ) = ∆ r H −○ (T1 ) + ∫ ∆ r C −p○ dT
T1
The difference of the molar heat capacities of products and reactants is given
by [2C.7b–53]
∆ r C −p○ = −
○
○
products reactants
= C −p,m
○
(CO2 , g) + 2C −p,m
○
(H2 O, g) − 2C −p,m
○
(O2 , g) − C −p,m
○
(CH4 , g)
The heat capacities are expressed in the form of C −p,m

○
= α + βT + γT 2 therefore
−
○
∆ r C p = ∆α + ∆βT + ∆γT where ∆α = α(CO2 , g) + 2α(H2 O, g) − 2α(O2 , g) −
2
α(CH4 , g) and likewise for ∆β and ∆γ.
∆α = [(26.86) + 2 × (30.36) − 2 × (25.72) − (14.16)] J K−1 mol−1

= +21.98 J K−1 mol−1
∆β = [(6.97) + 2 × (9.61) − 2 × (12.98) − (75.5)] × 10−3 J K−2 mol−1
= −75.27 × 10−3 J K−2 mol−1
∆γ = [(−0.82) + 2 × (1.184) − 2 × (−3.862) − (−17.99)] × 10−6 J K−3 mol−1
= +27.262 × 10−6 J K−3 mol−1
Integrating Kirchhoff ’s Law gives

T2
∆ c H −○ (T2 ) = ∆ c H −○ (T1 ) + ∫ (∆α + ∆βT + ∆γT 2 ) dT
T1
= ∆ c H −○ (T1 ) + ∆α(T2 − T1 ) + 21 ∆β(T22 − T12 ) + 13 ∆γ(T23 − T13 )
= (−802.3 kJ mol−1 ) + (+21.98 J K−1 mol−1 ) × (500 K − 298 K)
+ 21 × (−75.27 × 10−3 J K−2 mol−1 ) × [(500 K)2 − (298 K)2 ]
+ 13 × (27.262 × 10−6 J K−3 mol−1 ) × [(500 K)3 − (298 K)3 ]
= −803 kJ mol−1
P2C.11 (a) The heat released in the calorimeter by the combustion of glucose is
∣qv ∣ = c∆T = (641 J K−1 ) × (7.793 K) = 4.99... kJ
Therefore the standard internal energy of combustion is
−∣qv ∣ −(4.99... kJ)
∆ c U −○ = = = −2.80... × 103 kJ mol−1
n (0.3212 g)/(180.1548 g mol−1 )
= −2.80 × 103 kJ mol−1
The chemical equation for the combustion of glucose is C6 H12 O6 (s) +
6 O2 (g)
ÐÐ→ 6 CO2 (g) + 6 H2 O(l). There is no change in the number of gaseous
species when going from reactants to products, therefore ∆ c U −○ = ∆ c H −○ =
−2.80... × 103 kJ mol−1 = −2.80 × 103 kJ mol−1 . Applying [2C.5a–53] for
this reaction and using the values for the standard enthalpies of formation
from Table 2C.7 gives
∆ c H −○ = 6∆ f H −○ (CO2 , g) + 6∆ f H −○ (H2 O, l) − 6∆ f H −○ (O2 , g)
− ∆ f H −○ (C6 H12 O6 , s)
∆ f H −○ (C6 H12 O6 , s) = 6∆ f H −○ (CO2 , g) + 6∆ f H −○ (H2 O, l)
− 6∆ f H −○ (O2 , g) − ∆ c H −○
= 6 × (−393.51 kJ mol−1 ) + 6 × (−285.83 kJ mol−1 )
− 0 − (−2.80... × 103 kJ mol−1 )
= −1.27 × 103 kJ mol−1
(b) The reaction equation corresponding to anaerobic oxidation is C6 H12 O6 (s)
ÐÐ→ 2 C3 H6 O3 (s). The standard enthalpy of reaction is
∆ r H −○ = 2∆ f H −○ (C3 H6 O3 (s)) − ∆ f H −○ (C6 H12 O6 (s))
= 2 × (−694.0 kJ mol−1 ) − (−1.27 × 103 kJ mol−1 )
= −1.13... × 102 kJ mol−1
The amount of energy released in aerobic oxidation exceeds that of anaer-
obic glycolysis by
2.80... × 103 kJ mol−1 − 1.13... × 102 kJ mol−1 = 2.69 × 103 kJ mol−1

2D State functions and exact differentials

D2D.1 An inversion temperature is the temperature at which the Joule–Thomson co-
efficient µ changes sign from negative to positive or vice-versa. For a perfect
gas µ is always zero, thus it cannot have an inversion temperature. As explained
in detail in Section 2D.4 on page 62, the existence of the Joule–Thomson effect
depends upon intermolecular attractions and repulsions. A perfect gas has by
definition no intermolecular attractions and repulsions, so it cannot exhibit the
Joule–Thomson effect.
E2D.1(a) The volume of the liquid can be written as
V = V ′ (a + bT + cT 2 )
where a = 0.75, b = 3.9 × 10−4 K−1 and c = 1.48 × 10−6 K−2 . The expansion
coefficient is defined in [2D.6–60] as α = (1/V )(∂V /∂T) p . The derivative with
respect to T is
∂V
( ) = V ′ (b + 2cT)
∂T p
Therefore
1 b + 2cT
α= × [V ′ (b + 2cT)] =
V ′ (a + bT + cT 2 ) a + bT + cT 2
Evaluating this expression at 320 K gives
(3.9 × 10−4 K−1 ) + 2 × (1.48 × 10−6 K−2 ) × (320 K)
α 320 =
(0.75) + (3.9 × 10−4 K−1 ) × (320 K) + (1.48 × 10−6 K−2 ) × (320 K)2
= +1.3 × 10−3 K−1
E2D.2(a) The isothermal compressibility is defined in [2D.7–60], κ T = −(1/V )(∂V /∂p)T ,

therefore at constant temperature dV /V = −κ T dp. This question is concerned
with changes in density, so the next step is to rewrite the volume in terms of
the density, ρ. If the mass is m, V = m/ρ, and therefore dV = (−m/ρ 2 )dρ.
Therefore
dV 1 m ρ m 1
= (− 2 dρ) = − ( ) ( 2 ) dρ = − dρ
V V ρ m ρ ρ
It therefore follows that
dρ 1
= κ T dp and hence dp = dρ
ρ κT ρ
This expression gives the relationship between the change in pressure and the
change in density. Approximating dρ by δρ and dp by δ p for sufficiently small
changes gives
1 δρ 1
δp = × =( ) × (1.0 × 10−3 ) = +20 atm
κT ρ 4.96 × 10−5 atm−1
E2D.3(a) The difference C p,m − C V ,m is given by [2D.11–61], C p,m − C V ,m = α 2 T Vm /κ T .

In this expression the molar volume is found from the mass density ρ and the
molar mass M by Vm = M/ρ. The values of α and κ are available in the Resource
section, as is the mass density.
α 2 T Vm α 2 TM
C p,m − C V ,m = =
κT κT ρ
(12.4 × 10−4 K−1 )2 × (298 K) × (78.1074 g mol−1 )
=
[(92.1 × 10−6 bar−1 ) × (1 bar/105 Pa)] × (0.879 × 106 g m−3 )
= +44.2 J K−1 mol−1
The units are K−1 Pa m3 mol−1 = K−1 (N m−2 ) m3 mol−1 = K−1 N m mol−1
= J K−1 mol−1
E2D.4(a) The molar volume of a perfect gas at 400 K is calculated as
RT (8.3145 J K−1 mol−1 ) × (400 K)

Vm = = = 3.32... × 10−2 m3 mol−1
p (1.00 bar) × [(105 Pa)/(1.00 bar)]
The van der Waals parameter a of water vapour is found in Table 1C.3, and
needs to be converted to SI units
−2 10−6 m6 1.01325 × 105 Pa

a = (5.464 dm6 atm mol ) × ( 6 )×( )
1 dm 1 atm
−2
= 0.553... m6 Pa mol
Therefore the internal pressure is

−2
a (0.553... m6 Pa mol )
πT = = = 501 Pa
Vm (3.32... × 10−2 m3 mol−1 )2
2
E2D.5(a) The internal energy of a closed system of constant composition is a function of

temperature and volume. For a change in V and T, dU is given by [2D.5–59],
dU = π T dV + C V dT. At constant temperature, this reduces to dU = π T dV .
Substituting in the given expression for π T for a van der Waals gas and using
molar quantities
a
dUm = 2 dVm
Vm
This expression is integrated between Vm,i and Vm,f to give
Vm,f Vm,f a
∫ dU m = ∫ dVm
Vm,i Vm,i Vm2
hence
Vm,f
a 1 1
∆U m = − ∣ = −a ( − )
Vm2 Vm,i Vm,f Vm,i
The van der Waals parameter a for nitrogen gas is found in Table 1C.3, and
needs to be converted to SI units
−2 10−6 m6 1.01325 × 105 Pa

a = (1.352 dm6 atm mol ) × ( ) × ( )
1 dm6 1 atm
−2
= 0.136... m6 Pa mol
−2 1 1
∆Um = −(0.136... m6 Pa mol ) ( − )
20.00 × 10−3 m3 mol−1 1.00 × 10−3 m3 mol−1
= +1.30... × 102 J mol−1 = +130 J mol−1
The work done by an expanding gas is given by [2A.5a–37], dw = −pex dV . For

a reversible expansion pex is the pressure of the gas, hence
w = − ∫ pdVm
Substituting in the expression for the pressure of a van der Waals gas, [1C.5b–
24]
RT a RT a
w = −∫ − dVm = − ∫ dVm + ∫ dVm
Vm − b Vm2 Vm − b Vm2
RT
= −∫ dVm + ∆Um
Vm − b
The second term is identified as ∆U m from the above. According to the First
Law, ∆U = q + w, the first term in the expression above must be −q, therefore
Vm,f RT V Vm,f − b
q=∫ dVm = RT ln(Vm − b)∣Vm,f = RT ln ( )
Vm,i Vm − b m,i
Vm,i − b
20.00 dm3 − 3.87 × 10−2 dm3
= (8.3145 J K−1 mol−1 ) × (298 K) × ln ( )
1.00 dm3 − 3.87 × 10−2 dm3
= +7.51... × 103 J mol−1 = +7.52 kJ mol−1
where the value for b is taken from the Resource section. From the First Law
the work done is w = −q + ∆Um , hence
w = −q + ∆Um = −7.51... ×103 J mol−1 +1.30... ×102 J mol−1 = −7.39 kJ mol−1
P2D.1 The expansion coefficient is defined in [2D.6–60] as α = (1/V )(∂V /∂T) p . For
sufficiently small changes the derivatives are approximated by finite changes to
give
1 δV
α= ( )
V δT
Using the relationship given in the hint, δV = Aδr

1 Aδr αV
α= ( ) therefore δr = δT
V δT A
From Table 2D.1 α = 2.1 × 10−4 K−1 . For a temperature rise of 1.0○ C
(2.1 × 10−4 K−1 ) × (1.37 × 109 km3 )

∆r = × (1.0 K) = 7.96... × 10−4 km
(361 × 106 km2 )
= 0.80 m
As ∆r ∝ ∆T, for a temperature rise of 2.0 ○ C, ∆r = 1.6 m , and for a tempera-
ture rise of 3.5 ○ C, ∆r = 2.8 m is expected. This calculation assumes that the
total surface area of the oceans remains constant and that α(H2 O) = α(ocean).
P2D.3 (a) If V = V (p, T) then
∂V ∂V
dV = ( ) dp + ( ) dT
∂p T ∂T p
If p = p(V , T) then
∂p ∂p
dp = ( ) dV + ( ) dT
∂V T ∂T V
(b) Dividing the expression for dV by V gives
1 1 ∂V 1 ∂V
dV = ( ) dp + ( ) dT
V V ∂p T V ∂T p
Noting the definition of κ T , [2D.7–60], κ T = −(1/V )(∂V /∂p)T and that
of α, [2D.6–60], α = (1/V )(∂V /∂T) p , and rewriting (1/V )dV as d ln V
gives
d ln V = −κ T dp + αdT
Dividing the expression for dp by p gives
1 1 ∂p 1 ∂p
dp = ( ) dV + ( ) dT
p p ∂V T p ∂T V
1 ∂p 1 ∂p ∂V
d ln p = ( ) dV − ( ) ( ) dT
p ∂V T p ∂V T ∂T p
1 ∂p ∂V
= ( ) [dV − ( ) dT]
p ∂V T ∂T p
1 1⎛ 1 ⎞ ∂V
= (− ) ⎜ ⎟ [dV − ( ) dT]
V p ⎝− ( ) ⎠
1 ∂V ∂T p
V ∂p T
1 dV 1 ∂V
= [− + ( ) dT]
pκ T V V ∂T p
1 dV
= (− + αdT)
pκ T V
Going from line 1 to 2, the identity (1/x)dx = d ln x and Euler’s chain

relationship are used. Going from line 3 to 4 the reciprocal relationship
is used. The definition of κ T is used to go from line 4 to 5 and to go from
line 5 to 6 the definition of α is used.
P2D.5 The isothermal compressibility is defined in [2D.11–61], κ T = −(1/V )(∂V /∂p)T .

Using the reciprocal identity this is equivalent to
1 1
κT = −
V (∂p/∂V )T
The pressure of a van der Waals gas is p = nRT/(V − nb) − n 2 a/V 2 , therefore
∂p nRT 2n 2 a −nRT V 3 + 2n 2 a(V − nb)2

( ) =− + =
∂V T (V − nb) 2 V3 (V − nb)2 V 3
It follows that
∂V (V − nb)2 V 3
( ) =
∂p T −nRT V 3 + 2n 2 a(V − nb)2
and hence
1 ∂V V 2 (V − nb)2
κT = − ( ) =
V ∂p T nRT V 3 − 2n 2 a(V − nb)2
The expansion coefficient is defined in [2D.6–60] as α T = (1/V )(∂V /∂T) p ,
which is rewritten using the reciprocal identity as
1 1
α=
V (∂T/∂V ) p
Rearranging the van der Waals equation of state gives

p na
T= (V − nb) + (V − nb)
nR RV 2
hence
A
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
∂T p na 2na T 2na
( ) = + − (V − nb) = − (V − nb)
∂V p nR RV 2 RV 3 V − nb RV 3
RT V 3 − 2na(V − nb)2
=
(V − nb)RV 3
where the term A is identified from the previous equation as being equal to
T/(V − nb).Using the reciprocal identity
∂V (V − nb)RV 3
( ) =
∂T p RT V 3 − 2na(V − nb)2
therefore
1 ∂V (V − nb)RV 2
α= ( ) =
V ∂T p RT V 3 − 2na(V − nb)2
It follows that
κT V 2 (V − nb)2 (V − nb)RV 2
= ×
α nRT V 3 − 2n 2 a(V − nb)2 RT V 3 − 2na(V − nb)2
(V − nb)
=
nR
For a molar quantity V → Vm and n → 1 to give κ T /α = (Vm − b)/R or
κ T R = α(Vm − b) .
P2D.7 The equation of state of this gas is rearranged to give

nRT
V= + nb
p
From this it follows that
∂V nR
( ) =
∂T p p
Substituting this result in the expression for µ gives
1 ∂V
µ= [T ( ) − V]
Cp ∂T p
1 nR nRT nb
= [T ( ) − ( + nb)] = −
Cp p p Cp
The heat capacity and the van der Waals parameter b are always positive, there-
fore the Joule–Thomson coefficient µ is negative for this gas. This means that
the temperature of the gas will increase when expanding in an isenthalpic pro-
cess.
P2D.9 From [2D.13–61], dH = −µC p dp+C p dT, it follows that at constant temperature
1 ∂H
µ=− ( )
C p ∂p T
The partial derivative (∂H/∂p)T is the slope of the plot of H against p measured
at constant temperature.
(a) The plot for 300 K is shown in Fig. 2.2.
The slope of the plot is −17.93. The Joule–Thomson coefficient is deter-
mined from the slope,
(−17.93 kJ kg−1 MPa−1 )

µ=− = 23 K MPa−1
(0.7649 kJ K−1 kg−1 )
(b) The plot for 350 K is shown in Fig. 2.3.

The slope of the plot is −14.46. The Joule–Thomson coefficient is deter-
mined from the slope,
(−14.46 kJ kg−1 MPa−1 )

µ=− = 14 K MPa−1
(1.0392 kJ K−1 kg−1 )
426.6
426.4
H/kJ kg−1
426.2
426.0
425.8
425.6
0.07 0.08 0.09 0.10 0.11 0.12 0.13
p/MPa
Figure 2.2
462
460
−1
H/kJ kg
458
456
0.8 0.9 1.0 1.1 1.2
p/MPa
Figure 2.3
2E Adiabatic changes
D2E.1 In an adiabatic expansion the system does work but as no energy as heat is
permitted to enter the system, the internal energy of the system falls and so
consequently does the temperature. From the First Law, ∆U = w because q =
0. However, the change in internal energy is also related to the temperature
change and the heat capacity: ∆U = C V ∆T. Equating these two expressions
for ∆U gives w = C V ∆T, or dw = C V dT for an infinitesimal change.
For a reversible expansion, the work is dw = −pdV . Equating these two ex-
pressions for the work gives −pdV = C V dT. This equation is the point from
which the relationships between pressure, volume and temperature for a re-
versible adiabatic expansion are found: the heat capacity comes into the final
expressions via this route.
In words, the key thing here is that in an adiabatic process there is a change
in temperature, so it is not surprising that the properties of such a process are

related to the heat capacity because this quantity relates the energy and the
temperature rise.
E2E.1(a) For a reversible adiabatic expansion the initial and final states are related by
γ γ
[2E.3–66], p i Vi = p f Vf , where γ is the ratio of heat capacities, γ = C p,m /C V ,m .
The initial volume of the sample is
nRT (1.0 mol) × (8.3145 J K−1 mol−1 ) × (300 K)

Vi = = = 5.79... × 10−3 m3
p (4.25 atm) × [(1.01325 × 105 Pa)/(1 atm)]
= 5.79... dm3
For a perfect gas C p,m − C V ,m = R, hence
C p,m C V ,m + R (20.8 J K−1 mol−1 ) + (8.3145 J K−1 mol−1 )

γ= = = = 1.39...
C V ,m C V ,m (20.8 J K−1 mol−1 )
The final volume is given by Vf = (5.79... dm3 )×(4.25/2.50)1/1.39... = 8.46... dm3 .

Thus Vf = 8.46 dm3 .
The initial and final states are also related by (Tf /Ti ) = (Vi /Vf )1/c where c =
C V ,m /R. For this gas c = (20.8 J K−1 mol−1 )/(8.3145 J K−1 mol−1 ) = 2.50...
and hence
1/2.50...
Vi 1/c 5.79... dm3
Tf = Ti ( ) = (300 K) × ( ) = 2.57... × 102 K = 258 K
Vf 8.46 dm3
The work done by a perfect gas during adiabatic expansion is given by [2E.1–65]
w ad = C V ∆T = C V (Tf − Ti ) = (20.8 J K−1 )×(2.57... ×102 K−300 K) = −877 J
E2E.2(a) The work done in a reversible adiabatic expansion is given by [2E.1–65], w ad =

C V ∆T. The task is to find ∆T. The initial and final states in a reversible adia-
batic expansion are related by [2E.2b–66], Vi Tic = Vf Tfc , where c = C V ,m /R. If
the gas is assumed to be perfect, C p,m − C V ,m = R [2B.9–47], and so
C V ,m C p,m − R (37.11 J K−1 mol−1 ) − (8.3145 J K−1 mol−1 )

c= = = = 3.46...
R R (8.3145 J K−1 mol−1 )
The temperature change is given by
Vi 1/c Vi 1/c
∆T = Tf − Ti = Ti ( ) − Ti = Ti [( ) − 1]
Vf Vf
Therefore the work done is
Vi 1/c
w ad = C V ∆T = nC V ,m ∆T = nC V ,m Ti [( ) − 1]
Vf
Vi 1/c
= n(C p,m − R)Ti [( ) − 1]
Vf
2.45 g
=( ) × [(37.11 J K−1 mol−1 )
44.01 g mol−1
− (8.3145 J K−1 mol−1 )] × (300.15 K)
⎡ 1/3.46... ⎤
⎢ 500 cm3 ⎥
× ⎢⎢( 3 3
) − 1⎥⎥ = −194 J
⎢ 3.00 × 10 cm ⎥
⎣ ⎦
E2E.3(a) The initial and final states in a reversible adiabatic expansion are related by
γ γ
[2E.3–66], p i Vi = p f Vf , therefore
1.4
Vi γ 0.50 dm3
p f = p i ( ) = (67.4 kPa) × ( ) = 9.7 kPa
Vf 2.00 dm3
E2E.4(a) Ammonia and methane are nonlinear polyatomic molecules which have three
degrees of translational and three degrees of rotational freedom. From the
equipartition theorem (The chemist’s toolkit 7 in Topic 2A)
C V ,m = 1
2
× (νt + νr + 2νv ) × R

of rotational degrees of freedom and νv is the number of vibrational degrees of
freedom.
(i) Without any vibrational contribution the calculation is the same for both
molecules. C V ,m = 21 × (3 + 3 + 0) × R = 3R. For a perfect gas, [2B.9–47],
C p,m = C V ,m + R, therefore
C V ,m + R 3R + R 4
γ= = =
C V ,m 3R 3
(ii) If the vibrational contribution is included, the molecules have different

values of γ. The number of vibrational modes for a nonlinear polyatomic
molecule is νv = 3N − 6, where N is the number of atoms in the molecule.
Therefore, for ammonia νv = 3N − 6 = 6 and for methane νv = 3N − 6 =
9. This gives C V ,m (NH3 ) = 12 × (3 + 3 + 2 × 6) × R = 9R, therefore
C p,m (NH3 ) = 10R and γ = 10/9. For methane C V ,m (CH4 ) = 21 × (3 + 3 +
2 × 9) × R = 12R, C p,m (CH4 ) = 13R and γ = 13/12.
The experimental values of γ for ammonia and methane are the same, γ = 1.31,
which is closer to the value calculated without taking the vibrational degrees of
freedom into account.
E2E.5(a) For a reversible adiabatic expansion the initial and final states are related by
[2E.2a–66], (Tf /Ti ) = (Vi /Vf )1/c , where c = C V ,m /R. For a monoatomic ideal
gas C V ,m = 32 R, so c = 32 . Therefore
1 2
Vi c 1.0 dm3 3
Tf = Ti ( ) = (273.15 K) × ( ) = 1.3 × 102 K
Vf 3.0 dm3
P2E.1 The work done in a reversible adiabatic expansion is given by [2E.1–65], w ad =
C V ∆T. The initial and final states in a reversible adiabatic expansion are related
by [2E.2b–66], Vi Tic = Vf Tfc , where c = C V ,m /R. If the gas is assumed to be
perfect, C p,m − C V ,m = R [2B.9–47], and so
C V ,m C p,m − R (35.06 J K−1 mol−1 ) − (8.3145 J K−1 mol−1 )

c= = = = 3.21...
R R (8.3145 J K−1 mol−1 )
The final temperature is
1/3.21...
Vi 1/c 0.50 dm3
Tf = Ti ( ) = (298 K) × ( ) = 1.93... × 102 K = 194 K
Vf 2.00 dm3
Therefore the work done is
w ad = C V ∆T = nC V ,m ∆T = n(C p,m − R)(Tf − Ti )1/c
= (1.00 mol) × [(35.06 J K−1 mol−1 ) − (8.3145 J K−1 mol−1 )]
× [(1.93... × 102 K) − (298 K)]1/3.21... = −2.79 kJ
∆U = w ad = −2.79 kJ
Integrated activities
I2.1 The change in reaction enthalpy with respect to temperature is described by
Kirchhoff ’s Law, [2C.7a–53]. Assuming that the heat capacities are indepen-
dent of temperature, the integrated form of Kirchoff ’s Law is applicable and is
given by [2C.7d–54]
∆ r H −○ (T2 ) = ∆ r H −○ (T1 ) + (T2 − T1 )∆ r C −p○
If ∆ r C −p○ is negative, then the reaction enthalpy will decrease with increasing
temperature, whereas if ∆ r C −p○ is positive, then the reaction enthalpy will in-
crease with increasing temperature.
(a)
∆ r C −p○ = −
○
○
products reactants
= 2C −p,m
○
(H2 O, l) − C −p,m
○
(O2 , g) − 2C −p,m
○
(H2 , g)
= 2 × (9R) − ( 27 R) − 2 × ( 27 R) = +7 21 R
∆ r C −p○ is positive, therefore the standard reaction enthalpy of reaction will

increase with increasing temperature.
(b)
∆ r C −p○ = C −p,m
○
(CO2 , g) + 2C −p,m
○
(H2 O, l) − 2C −p,m
○
(O2 , g) − C −p,m
○
(CH4 , g)
= ( 27 R) + 2 × (9R) − 2 × ( 27 R) − (4R) = +10 21 R
I2.3 (a) From the equipartition principle (The chemist’s toolkit 7 in Topic 2A),
and assuming that only translational and rotational levels contribute, the
constant volume heat capacity of a gas of diatomic molecules is C V ,m =
5
2
R. For a perfect gas C p,m − C V ,m = R, [2B.9–47], therefore C p,m =
C V ,m + R = 72 R and so γ = C p,m /C V ,m = 57 . It follows that
γRT 1/2 7RT 1/2

cs = ( ) =( )
M 5M
(b) A linear triatomic has the same number of translational and rotational
modes of motion (three translational, two rotational) as a diatomic, so
the calculation is the same as in (a).
(c) A gas of non-linear triatomic molecules has one extra mode of rotational
motion when compared to a linear triatomic, so the constant volume heat
capacity is C V ,m = 3R. Because C p,m − C V ,m = R it follows that C p,m =
C V ,m + R = 4R and so γ = C p,m /C V ,m = 34 .
γRT 1/2 4RT 1/2

cs = ( ) =( )
M 3M
The air is mostly composed of diatomic and linear triatomic molecules, N2 , O2
and CO2 , hence γ ≈ 57 . The average molecular weight of air is approximately
29.0 g mol−1 , therefore
1/2
7RT 1/2 7 × (8.3145 J K−1 mol−1 ) × (298.15 K)
cs = ( ) =( ) = 346 m s−1
5M 5 × (29.0 × 10−3 kg mol−1 )
I2.5 A state function is a thermodynamic property, the value of which is indepen-

dent of the history of the system. Examples of state functions are the properties
of pressure, temperature, internal energy, enthalpy as well as the properties of
entropy, Gibbs energy, and Helmholtz energy to be discussed fully in Focus 3.
The differentials of state functions are exact differentials. Hence, the mathemat-
ical properties of exact differentials can be used to draw far-reaching conclu-
sions about the relations between physical properties and establish connections
that were unexpected but turn out to be very significant.
One practical importance of these results is that the value of a property of inter-
est can be obtained from the combination of measurements of other properties
without actually having to measure the required property itself, the measure-
ment of which might be very difficult.
I2.7 The change in reaction enthalpy with respect to temperature is described by

Kirchhoff ’s Law, [2C.7a–53]. Assuming that the heat capacities are indepen-
dent of temperature, the integrated form of Kirchoff ’s Law is applicable and is
given by [2C.7d–54]
∆ r H −○ (T2 ) = ∆ r H −○ (T1 ) + (T2 − T1 )∆ r C −p○
If ∆ r C −p○ is negative, then the reaction enthalpy will decrease with increasing
temperature, whereas if ∆ r C −p○ is positive, then the reaction enthalpy will in-
crease with increasing temperature.
(a)
∆ r C −p○ = −
○
○
products reactants
= 2C −p,m
○
(H2 O, g) − C −p,m
○
(O2 , g) − 2C −p,m
○
(H2 , g)
= 2 × (4R) − ( 27 R) − 2 × ( 27 R) = −2 21 R
∆ r C −p○ is negative, therefore the standard reaction enthalpy of reaction will

decrease with increasing temperature.
(b)
∆ r C −p○ = C −p,m
○
(CO2 , g) + 2C −p,m
○
(H2 O, g) − 2C −p,m
○
(O2 , g) − C −p,m
○
(CH4 , g)
= ( 27 R) + 2 × (4R) − 2 × ( 27 R) − (4R) = + 21 R

(c)
∆ r C −p○ = 2C −p,m
○
(NH3 , g) − C −p,m
○
(N2 , g) − 3C −p,m
○
(H2 , g)
= 2 × (4R) − ( 72 R) − 3 × ( 27 R) = −6R
∆ r C −p○ is negative, therefore the standard reaction enthalpy of reaction will

decrease with increasing temperature.
3 The second and third laws
3A Entropy
D3A.1 Everyday experience indicates that the direction of spontaneous change in an
isolated system is accompanied by the dispersal of the total energy of the sys-
tem. For example, for a gas expanding freely and spontaneously into a vacuum,
the process is accompanied by a dispersal of energy and matter. For a perfect
gas this entropy change of such an expansion is derived in Section 3A.2(a) on
page 76 as ∆S = nR ln(Vf /Vi ). The entropy change is clearly positive if Vf is
greater than Vi .
The molecular interpretation of this thermodynamic result is based on the iden-
tification of entropy with molecular disorder. An increase in disorder results
from the chaotic dispersal of matter and energy and the only changes that can
take place within an isolated system (the universe) are those in which this kind
of dispersal occurs. This interpretation of entropy in terms of dispersal and
disorder allows for a direct connection of the thermodynamic entropy to the
statistical entropy through the Boltzmann formula S = k ln W, where W is
the number of microstates, the number of ways in which the molecules of the
system can be arranged while keeping the total energy constant.
The concept of the number of microstates makes quantitative the more ill-
defined qualitative concepts of ‘disorder’ and ‘the dispersal of matter and en-
ergy’ used above to give a physical feel for the concept of entropy. A more
‘disorderly’ distribution of energy and matter corresponds to a greater number
of microstates associated with the same total energy.
D3A.3 The Second Law of thermodynamics states only that the total entropy of both
the system (here, the molecules organizing themselves into cells) and the sur-
roundings (here, the medium) must increase in a naturally occurring process.
It does not state that entropy must increase in a portion of the universe that
interacts with its surroundings. In this case, the cells grow by using chemical
energy from their surroundings (the medium) and in the process the increase
in the entropy of the medium outweighs the decrease in entropy of the system.
Hence, the Second Law is not violated.
E3A.1(a) The efficiency is defined in [3A.7–80], η = ∣w∣/∣q h ∣, and for a Carnot cycle
efficiency is given by [3A.9–80], η = 1 − (Tc /Th ). These two are combined
70 3 THE SECOND AND THIRD LAWS
and rearranged into an expression for the temperature of the cold sink
∣w∣/∣q h ∣ = 1 − (Tc /Th )

∣w∣
hence Tc = (1 − ) × Th
∣q h ∣
∣3.00 kJ∣
= (1 − ) × (273 K) = 191 K .
∣ − 10.00 kJ∣
E3A.2(a) The efficiency of a Carnot cycle is given by [3A.9–80], η = 1 − (Tc /Th ). Thus
Tc (273.15 K + 10 K)
η =1− =1− = 0.241 = 24.1% .
Th (273.15 K + 100 K)
Note that the temperatures must be in kelvins.
E3A.3(a) For the process to be spontaneous it must be irreversible and obey the Clausius
inequality [3A.12–82] implying that ∆S tot = ∆S + ∆S sur > 0. In this case,
∆S tot = 125 J K−1 + (−125 J K−1 ) = 0, thus the process is not spontaneous in
either direction and is at equilibrium.
E3A.4(a) The thermodynamic definition of entropy is [3A.1a–76], dS = dq rev /T or for

a finite change at constant temperature ∆S = q rev /T. The transfer of heat is
specified as being reversible, which can often be assumed for a large enough
metal block, therefore q rev = 100 kJ.
(i)
q rev 100 kJ
∆S = = = 0.366 kJ = +366 J
T 273.15 K
(ii)
q rev 100 kJ
∆S = = = 0.309 kJ = +309 J
T (273.15 K + 50 K)
E3A.5(a) As explained in Section 3A.2(a) on page 76 the change in entropy for an isother-
mal expansion of a gas is calculated using
Vf m Vf
∆S = nR ln ( ) = R ln ( )
Vi M Vi
15 g −1 −1 3.0 dm3
=( ) × (8.3145 J K mol ) × ln ( )
44.01 g mol−1 1.0 dm3
= +3.1 J K−1 .
E3A.6(a) The change in entropy for an isothermal expansion of a gas is ∆S = nR ln (Vf /Vi )
as explained in Section 3A.2(a) on page 76. For a doubling of the volume
Vf /Vi = 2.
(i) Isothermal reversible expansion
14 g −1 −1 −1
∆S = ( −1 ) × (8.3145 J K mol ) × ln (2) = +2.9 J K .
28.02 g mol
Because the process is reversible ∆S tot = 0 .

Because ∆S tot = ∆S + ∆S sur
∆S sur = ∆S tot − ∆S = −2.9 J K−1 .
(ii) Isothermal irreversible expansion against p ex = 0 Because entropy is a

state function and the initial and final states of the system are the same as
in (a), ∆S is the same.
∆S = +2.9 J K−1 .
Expansion against an external pressure of 0 does no work, and for an
isothermal process of an ideal gas ∆U = 0. From the First Law if follows
that q = 0 and therefore ∆S sur = 0 .
∆S tot = ∆S + ∆S sur = +2.9 J K−1 .

(iii) Adiabatic reversible expansion For an adiabatic expansion there is no heat
flowing to or from the surroundings, thus ∆S sur = 0 . For a reversible
process ∆S tot = 0 , therefore it follows that ∆S = 0 as well.
P3A.1 (a) Isothermal reversible expansion
The work of a reversible isothermal expansion of an ideal gas is given
by [2A.9–39], w = −nRT ln (Vf /Vi ). Because at fixed temperature p ∝
(1/V ) as given by Boyle’s law, an equivalent expression is
pi
w = −nRT ln ( )
pf
= −(1.00 mol) × (8.3145 J K−1 mol−1 )
3.00 atm
× (273.15 K + 27 K) × ln ( )
1.00 atm
= −2.74 × 103 ... J = −2.74 kJ .
For an isothermal process of a perfect gas ∆U = 0 and ∆H = 0 . The

First Law is defined in [2A.2–36], ∆U = q + w, hence
q = ∆U − w = 0 − (−2.74... kJ) = +2.74 kJ .
The heat transfer is reversible, therefore q rev = q.
q rev 2.74 × 103 ... J

∆S = =
T 273.15 K + 27 K
= +9.13... J K−1 = +9.13 J K−1 .
The process is reversible, therefore ∆S tot = 0 . Finally because ∆S tot =

∆S + ∆S sur
∆S sur = ∆S tot − ∆S = 0 − (+9.13... J K−1 ) = −9.13 J K−1 .

(b) Isothermal expansion against p ex = 1.00 atm The expansion work against
a constant external pressure is given by [2A.6–38], w = −p ex (Vf − Vi ).
The volumes are written in terms of pressures by using the perfect gas law
[1A.4–8], pV = nRT.
w = −p ex (Vf − Vi )
nRT nRT p ex p ex
= −p ex ( − ) = −nRT × ( − )
pf pi pf pi
= −(1.00 mol) × (8.3145 J K−1 mol−1 )
1.00 atm 1.00 atm
× (273.15 K + 27 K) × ( − )
1.00 atm 3.00 atm
= −1.66... × 103 J = −1.66 kJ .
For an isothermal process in perfect gas ∆U = 0 and ∆H = 0 . Using the

First Law
q = ∆U − w = 0 − (−1.66 kJ) = +1.66 kJ .
Because entropy is a state function and the initial and final states of the
system are the same, the entropy change of the system is as in (a), ∆S =
+9.13 J K−1 .
The entropy change of the surroundings in terms of the heat of the sur-
roundings, q sur , is given by [3A.2b–77], ∆S sur = q sur /T. This heat is
simply the opposite of the heat of the system: q sur = −q, therefore
q sur −q
∆S sur = =
T T
−1.66... × 103 J
= = −5.54... J K−1 = −5.54 J K−1 .
(273.15 K + 27 K)
∆S tot = ∆S + ∆S sur
= (+9.13... J K−1 ) + (−5.53... J K−1 ) = +3.59 J K−1 .
P3A.3 (a) After Stage 1 the volume doubles, thus VB = 2 × VA = 2 × (1.00 dm3 ) =
2.00 dm3 . Assuming V T 3/2 = constant for the adiabatic stages, the
volume after Stage 2 is
Th 3/2 373 K 3/2

VC = VB × ( ) = (2.00 dm3 ) × ( )
Tc 273 K
= 3.19... dm3 = 3.19 dm3 .
(b) Again assuming V T 3/2 = constant for the adiabatic stage, the volume
after Stage 3 can be related to the initial volume
Th 3/2 373 K 3/2

VD = VA × ( ) = (1.00 dm3 ) × ( )
Tc 273 K
= 1.59... dm3 = 1.60 dm3 .
(c) As shown in Section 3A.3(a) on page 78 the heat transferred reversibly

during an isothermal gas expansion is q rev = nRT ln (Vf /Vi ), thus the
heats for Stage 1 and Stage 3 are, respectively
VB
q 1 = q h = nRTh ln ( )
VA
= (0.100 mol) × (8.3145 J K−1 mol−1 ) × (373 K) × ln (2)
= +2.14... × 102 J = +215 J .
VD
q 3 = q c = nRTc ln ( )
VC
1.59... dm3
= (0.100 mol) × (8.3145 J K−1 mol−1 ) × (273 K) × ln ( )
3.19... dm3
= −1.57... × 102 J = −157 J .
Because there is no heat exchange during adiabatic processes, he heat

transfer for Stages 2 and 4 are q 2 = 0 and q 4 = 0 , respectively.
(d) At the beginning and end of the cycle the temperature is the same. Be-
cause the working substance is a perfect gas, ∆U = 0 over the cycle. The
First Law [2A.2–36], ∆U = w + q, therefore implies that w = −q, that
is, the net heat over the cycle is converted to work. This net heat is the
difference between that extracted from the hot source and deposited into
the cold sink.
(e) The efficiency is defined in [3A.7–80], η = ∣w∣/∣q h ∣. As has been explained,
∣w∣ is the net heat.
∣w∣ = ∣q h ∣ − ∣q c ∣ = ∣ + 2.14... × 102 J∣ − ∣ − 1.57... × 102 J∣

= +5.7... × 101 J = +58 J .
∣w∣ ∣ + 5.7... × 101 J∣
hence η= = = 0.268... = 27% .
∣q h ∣ ∣ + 2.14... × 102 J∣
(f) The Carnot efficiency is given by [3A.9–80],
Tc 273 K
η =1− =1− = 0.268 = 26.8% .
Th 373 K
the result is the same as the above (the difference is due to the use of fewer
significant figures in the previous calculation).
Using the values of the heat transfer calculated above in equation [3A.6–
80] gives
q c q c 214... J −157... J
+ = +
Th Tc 373 K 273 K
= 0.0 J .
the result is zero, as expected from a Carnot cycle.

P3A.5 (a) Consider a process in which heat dq c is extracted from the cold source at
temperature Tc , and heat dq h is discarded into the hot sink at temperature
Th . The overall entropy change of such process is
dq c dq h
dS = +
Tc Th
Assume that dq c = −dq and dq h = +dq, where dq is a positive quantity.
It follows that
+dq −dq 1 1
dS = + = dq × ( − )
Th Tc Th Tc
Because Th > Tc , the term in parentheses is negative, therefore dS is
negative. The process is therefore not spontaneous and not allowed by
the Second Law. If work is done on the engine, ∣dq h ∣ will become greater
than ∣dq c ∣ and eventually dS will be greater than zero.
(b) Assuming q c = −∣q∣ and q h = ∣q∣ + ∣w∣ the overall change in entropy is
−∣q∣ ∣q∣ + ∣w∣
∆S = +
Tc Th
For the process to be permissible by the Second Law the Clausius inequal-
ity defined in [3A.12–82], dS ≥ 0, must be satisfied. Therefore
−∣q∣ ∣q∣ + ∣w∣
+ ≥0
Tc Th
which implies
Th Th
∣w∣ ≥ ∣q∣ × ( − 1) = ∣q∣ × ( − 1) .
Tc Tc
P3A.7 Suppose two adiabatic paths intersect at point A as shown in the figure. Two
remote points corresponding to the same temperature on each adiabat, A and
B, are then connected by an isothermal path forming a cycle.
Consider energy changes for each Stage of the cycle. Stage 1 (A → B) is adiabatic
and, thus, no heat exchange takes place q 1 = 0. Therefore, the total change in
internal energy is ∆U 1 = w 1 + q 1 = w 1 . Stage 2 (B → C) is an isothermal change
and assuming that the system energy is a function of temperature only (e.g.
ideal gas): ∆U 2 = w 2 + q 2 = 0. Stage 3 (C → A) is again adiabatic, q 3 = 0, with
∆U 3 = w 3 + q 3 = w 3 . Because the system energy is a function of temperature
only, U B = U C and, thus
∆U 3 = U A − U C = U A − U B = −∆U 1
This implies that w 1 = −w 3 .

Because internal energy is a state function and the cycle is closed:
U cycle = w cycle + q cycle = 0

= ∆U 1 + ∆U 2 + ∆U 3
Pressure, p
Stage 1
Stage 3
C
B
Stage 2
Volume, V
Figure 3.1
Finally, analyse the net work done, w cycle = w 1 + w 2 + w 3 = w 2 , and the net heat
absorbed, q cycle = q 1 + q 2 + q 3 = q 2 , over the cycle. It is apparent that the sole
result of the process is the absorption of heat q 2 and its convertion to work w 2 ,
which directly contradicts the statement of the Second Law by Kelvin, unless
the q 2 = w 2 = 0, i.e. points B and C are the same and correspond to the same
path. Therefore, no two such adiabatic paths exist.
3B Entropy changes accompanying specific processes
Answer to discussion question
D3B.1 The explanation of Trouton’s rule is that a comparable change in volume is

expected whenever any unstructured liquid forms a vapour; accompanying this
will be a comparable change in the number of accessible microstates. Hence, all
unstructured liquids can be expected to have similar entropies of vaporization.
Liquids that show significant deviations from Trouton’s rule do so on account of
strong molecular interactions that restrict molecular motion. As a result there
is a greater dispersal of matter and energy when such liquids vaporize.
Water is an example of a liquid with strong intermolecular interactions (hy-

drogen bonding) which tend to organize the molecules in the liquid, hence its
entropy of vaporization is expected to be greater than the value predicted by
Trouton’s rule. The same is true for ethanol, which is also hydrogen bonded in
the liquid.
Mercury has quite strong interactions between the atoms, as evidenced by its
cohesiveness, and so its entropy of vaporization is expected to be greater than
that predicted by Trouton’s rule.
E3B.1(a) Two identical blocks must come to their average temperature. Therefore the
final temperature is
Tf = 12 (T1 + T2 ) = 1
2
× (50 ○ C + 0 ○ C) = 25 ○ C = 298 K .
Although the above result may seem self-evident, the more detailed explaina-
tion is as follows. The heat capacity at constant volume is defined in [2A.14–
41], C V = (∂U/∂T)V . As shown in Section 2A.4(b) on page 41, if the heat
capacity is constant, the internal energy changes linearly with the change in
temperature. That is ∆U = C V ∆T = C V (Tf − Ti ). For the two blocks at
the initial temperatures of T1 and T2 , the change in internal energy to reach
the final temperature Tf is ∆U 1 = C V ,1 (Tf − T1 ) and ∆U 2 = C V ,2 (Tf − T2 ),
respectively. The blocks of metal are made of the same substance and are of
the same size, therefore C V ,1 = C V ,2 = C V . Because the system is isolated
the total change in internal energy is ∆U = ∆U 1 + ∆U 2 = 0. This means that
∆U = C V ((Tf − T1 ) − (Tf − T2 )) = C V × (2Tf − (T1 + T2 )) = 0, which implies
that the final temperature is Tf = 21 (T1 + T2 ), as stated above.
The temperature variation of the entropy at constant volume is given by [3B.7–
86], ∆S = C V ln (Tf /Ti ), with C p replaced by C V . Expressed with the specific
heat C V ,s = C V /m it becomes
Tf
∆S = mC V ,s ln ( ).
Ti
Note that for a solid the internal energy does not change significantly with the
volume or pressure, thus it can be assumed that C V = C p = C. The entropy
change for each block is found using this expression
Tf
∆S 1 = mC V ,s ln ( )
T1
298 K
= (1.00 × 103 g) × (0.385 J K−1 g−1 ) × ln ( )
50 K + 273.15 K
= −31.0... J K−1 = −31.0 J K−1 .
Tf
∆S 2 = mC V ,s ln ( )
T2
298 K
= (1.00 × 103 g) × (0.385 J K−1 g−1 ) × ln ( )
273.15 K
= 33.7... J K−1 = +33.7 J K−1 .
The total change in entropy is
∆S tot = ∆S 1 + ∆S 2 = (−31.0... J K−1 ) + (33.7... J K−1 )

= 27.2... J K−1 = +2.7 J K−1 .
Because ∆S tot > 0 the process is spontaneous, in accord with experience.

E3B.2(a) Because entropy is a state function, ∆S between the initial and final states is
the same irrespective of the path taken. Thus the overall process can be broken
down into steps that are easier to evaluate. First consider heating the initial
system at constant pressure to the final temperature. The variation of entropy
with temperature at constant pressure is given by [3B.7–86], S(Tf ) = S(Ti ) +
C p ln (Tf /Ti ). Thus the change in entropy, ∆S = S(Tf ) − S(Ti ), of this step is
Tf Tf
∆S 1 = C p ln ( ) = nC p,m ln ( )
Ti Ti
Next consider an isothermal change in pressure. As explained in Section 3A.2(a)

on page 76 the change in entropy of an isothermal expansion of an ideal gas
is given by ∆S = nR ln (Vf /Vi ). Because for a fixed amount of gas at fixed
temperature p ∝ (1/V ) an equivalent expression for this entropy change is
pi
∆S 2 = nR ln ( )
pf
Therefore the overall entropy change for the system is
Tf pi
∆S = ∆S 1 + ∆S 2 = nC p,m ln ( ) + nR ln ( )
Ti pf
273.15 K + 125 K
= (3.00 mol) × ( 52 × 8.3145 J K−1 mol−1 ) × ln ( )
273.15 K + 25 K
1.00 atm
+ (3.00 mol) × (8.3145 J K−1 mol−1 ) × ln ( )
5.00 atm
= (+18.0... J K−1 ) + (−40.1... J K−1 ) = −22.1 J K−1 .
E3B.3(a) Because entropy is a state function, ∆S between the initial and final states is
down into steps that are easier to evaluate. First consider heating the ice at
constant pressure from the initial temperature to the melting point, Tm . The
variation of entropy with temperature at constant pressure is given by [3B.7–
86], S(Tf ) = S(Ti ) + C p ln (Tf /Ti ). Thus the change in entropy, ∆S = S(Tf ) −
S(Ti ), for this step is
Tm Tm
∆S 1 = C p ln ( ) = nC p,m (H2 O(s)) ln ( )
Ti Ti
Next consider the phase transition from solid to liquid at the melting temper-
ature. The entropy change of a phase transition is given by [3B.4–85], ∆ trs S =
∆ trs H/Ttrs , thus
−
○
∆ fus H m
∆S 2 = n
Tm
Then the liquid is heated to the boiling temperature, Tb . In analogy to the first
step
Tb
∆S 3 = nC p,m (H2 O(l)) ln ( )
Tm
The next phase transition is from liquid to gas

−
○
∆ vap H m
∆S 4 = n
Tb
Finally, the vapour is heated from Tb to Tf
Tf
∆S 5 = nC p,m (H2 O(g)) ln ( )
Tb
∆S/n = ∆S 1 + ∆S 2 + ∆S 3 + ∆S 4 + ∆S 5
−
○
Tm ∆ fus H m Tb
= C p,m (H2 O(s)) ln ( )+ + C p,m (H2 O(l)) ln ( )
Ti Tm Tm
−
○
∆ vap H m Tf
+ + C p,m (H2 O(g)) ln ( )
Tb Tb
273.15 K
= (37.6 J K−1 mol−1 ) × ln ( )
273.15 K − 10.0 K
6.01 × 103 J mol−1
+
273.15 K
273.15 K + 100.0 K
+ (75.3 J K−1 mol−1 ) × ln ( )
273.15 K
40.7 × 103 J mol−1
+
273.15 K + 100.0 K
273.15 K + 115.0 K
+ (33.6 J K−1 mol−1 ) × ln ( )
273.15 K + 100.0 K
= (+1.40... J K−1 mol−1 ) + (+22.0... J K−1 mol−1 )
+ (+23.4... J K−1 mol−1 ) + (+1.09... × 102 J K−1 mol−1 )
+ (+1.32... J K−1 mol−1 )
= +1.57... × 102 J K−1 mol−1
Hence
10.0 g
∆S = × (+1.57... × 102 J K−1 ) = +87.3 J K−1 .
18.02 g mol−1
E3B.4(a) The entropy change of a phase transition is given by [3B.4–85], ∆ trs S = ∆ trs H/Ttrs .
As discussed in Section 3B.2 on page 85 because there is no hydrogen bonding
in liquid benzene it is safe to apply Trouton’s rule. That is ∆ vap S −○ = +85 J K−1 mol−1 .
It follows that
∆ vap H −○ = Tb × ∆ vap S −○
= (273.15 K + 80.1 K) × (+85 J K−1 mol−1 )
= 3.00... × 104 J mol−1 = +30 kJ mol−1 .
E3B.5(a) (i) The entropy change of a phase transition is given by [3B.4–85], ∆ trs S =
∆ trs H/Ttrs . For vaporisation this becomes
∆ vap H −○ + 29.4 × 103 J mol−1

∆ vap S −○ = =
Tb 334.88 K
= +87.8 J K−1 mol−1 .
(ii) Because the system at the transition temperature is at equilibrium, ∆S tot =

0, thus
∆S sur = −∆ vap S −○ = −87.8 J K−1 mol−1 .
E3B.6(a) The change in entropy with temperature is given by [3B.6–86],

Tf dT
∆S = S(Tf ) − S(Ti ) = ∫ Cp
Ti T
Assuming that C p is constant in the temperature range Ti to Tf , this becomes
∆S = C p ln (Tf /Ti ) as detailed in Section 3B.3 on page 86. Thus, the increase
in the molar entropy of oxygen gas is
348 K
∆S m = S m (348 K) − S m (298 K) = (29.355 J K−1 mol−1 ) × ln ( )
298 K
= +4.55 J K−1 mol−1 .
E3B.7(a) As explained in Section 3B.3 on page 86 the temperature variation of the en-
tropy at constant volume is given by
Tf dT
∆S = S(Tf ) − S(Ti ) = ∫ CV
Ti T
Assuming that C V = 32 R, the ideal gas limit, for the temperature range of
interest, the molar entropy at 500 K is given by
500 K
3 dT
S m (500 K) = S m (298 K) + ∫ R
298 K T 2
500 K
= S m (298 K) + 32 R × ln ( )
298 K
= (146.22 J K−1 mol−1 )
500 K
+ ( 23 × 8.3145 J K−1 mol−1 ) × ln ( )
298 K
= 153 J K−1 mol−1 .
P3B.1 Because entropy is a state function, ∆S between the initial and final states is
down into steps that are easier to evaluate.
First consider heating the water at constant pressure from the initial tempera-
ture T to the melting point. The variation of the entropy with temperature at
constant pressure is given by [3B.7–86], S(Tf ) = S(Ti ) + C p ln (Tf /Ti ). Thus
the change in entropy for this step is
Tm Tm
∆S 1 = C p ln ( ) = nC p,m (H2 O(l)) ln ( )
T T
Next consider the phase transition from liquid to solid at the melting temper-
−
○
ature; note that freezing is just the opposite of fusion, thus ∆H 2 = n(−∆ fus H m ).
The entropy change of a phase transition is given by [3B.4–85], ∆ trs S = ∆ trs H/Ttrs ,
thus −
○
∆H 2 −∆ fus H m
∆S 2 = =n
Tm Tm
The ice is then cooled to the final temperature, T. Similarly to the first step
T
∆S 3 = nC p,m (H2 O(s)) ln ( )
Tm
∆S = ∆S 1 + ∆S 2 + ∆S 3
Tm −∆ fus H −m
○
T
= nC p,m (H2 O(l)) ln ( )+n + nC p,m (H2 O(s)) ln ( )
T Tm Tm
−1 −1 273.15 K
= (1.00 mol) × (75.3 J K mol ) × ln ( )
273.15 K − 5.00 K
−6.01 × 103 J mol−1
+ (1.00 mol) ×
273.15 K
273.15 K − 5.00 K
+ (1.00 mol) × (37.6 J K−1 mol−1 ) × ln ( )
273.15 K
= (+1.39... J K−1 ) + (−22.0... J K−1 ) + (−0.694... J K−1 )
= −21.3... J K−1 = −21.3 J K−1 .
Consider enthalphy change for the same path. The variation of enthalpy with
temperature at constant pressure is given by [2B.6b–47], ∆H = C p ∆T. Thus for
the first and third steps, respectively
∆H 1 = nC p,m (H2 O(l))(Tm − T) and ∆H 3 = nC p,m (H2 O(s))(T − Tm )
Therefore the overall enthalpy change for the system is

∆H = ∆H 1 + ∆H 2 + ∆H 3
−
○
= nC p,m (H2 O(l))(Tm − T) + n(−∆ fus H m ) + nC p,m (H2 O(s))(T − Tm )
= (1.00 mol) × (75.3 J K−1 mol−1 ) × (+5.00 K)
+ (1.00 mol) × (−6.01 × 103 J mol−1 )
+ (1.00 mol) × (37.6 J K−1 mol−1 ) × (−5.00 K)
= (+3.76... × 102 J) + (−6.01... × 103 J) + (−1.88... × 102 J)
= −5.82... × 103 J
At constant pressure the heat released by the system is the enthalpy change of
the system, q = ∆H. Because q sur = −q, the entropy change of the surroundings
is
−q −(−5.82... × 103 J)
∆S sur = =
T 273.15 K − 5.00 K
= +21.7... J K−1 = +21.7 J K−1 .
Therefore the total entropy change is

∆S tot = ∆S + ∆S sur = (−21.3... J K−1 ) + (+21.7... J K−1 )
= +0.403... J K−1 = +0.4 J K−1 .
Because the total entropy change is positive, the Second Law implies that the
process is spontaneous .
A similar method is used to find the entropy change when the liquid evaporates
at T2 . Consider heating the liquid to the boiling temperature Tb , then the phase
transition taking place, followed by cooling of the gas back to the temperature
T2 . The entropy changes are calculated in an analagous way
∆S = ∆S 1 + ∆S 2 + ∆S 3
−
○
Tb ∆ vap H m T2
= nC p,m (H2 O(l)) ln ( )+n + nC p,m (H2 O(g)) ln ( )
T2 Tb Tb
−1 −1 273.15 K + 100 K
= (1.00 mol) × (75.3 J K mol ) ln ( )
273.15 K + 95.0 K
4.07 × 104 J mol−1
+ (1.00 mol) ×
273.15 K + 100 K
273.15 K + 95.0 K
+ (1.00 mol) × (33.6 J K−1 mol−1 ) ln ( )
273.15 K + 100 K
= (+1.01... J K−1 ) + (+1.09... × 102 J K−1 ) + (−0.453... J K−1 )
= +1.09... × 102 J K−1 = +110 J K−1 .
−∆H 1
∆S sur = = − × (∆H 1 + ∆H 2 + ∆H 3 )
T2 T2
−
○
Tb − T2 −∆ vap H m T2 − Tb
= − (nC p,m (H2 O(l)) +n + nC p,m (H2 O(g)) )
T2 T2 T2
5.00 K
= −(1.00 mol) × (75.3 J K−1 mol−1 )
273.15 K + 95.0 K
4.07 × 104 J mol−1
− (1.00 mol) ×
273.15 K + 95.0 K
−5.00 K
− (1.00 mol) × (33.6 J K−1 mol−1 ) ×
273.15 K + 95.0 K
= −(+1.02... J K−1 ) − (+1.10... × 102 J K−1 ) − (−0.456... J K−1 )
= −1.11... × 102 J K−1 = −111 J K−1 .
Therefore the total entropy change is

∆S tot = ∆S + ∆S sur = (+1.09... × 102 J K−1 ) + (−1.11... × 102 J K−1 )
= −1.48... J K−1 = −1.5 J K−1 .
Because the change in the entropy is negative, the Second Law implies that the
process is not spontaneous .
P3B.3 Consider heating trichloromethane at constant pressure from the initial to fi-
nal temperatures. The variation of the entropy with temperature is given by
T
[3B.6–86], S(Tf ) = S(Ti ) + ∫Ti f (C p /T)dT. The contant-pressure molar heat
capacity is given as a function of temperature of a form C p,m = a + bT, with
a = +91.47 J K−1 mol−1 and b = +7.5 × 10−2 J K−2 mol−1 . Thus the change in
molar entropy, ∆S m = S m (Tf ) − S m (Ti ), of this process is
Tf Tf a + bT
∆S m = ∫ (C p,m /T)dT = ∫ dT
Ti Ti T
Tf 1 Tf
= a×∫ dT + b × ∫ dT
Ti T Ti
Tf
= a × ln ( ) + b × (Tf − Ti )
Ti
300 K
= (+91.47 J K−1 mol−1 ) × ln ( )
273 K
+ (+7.5 × 10−2 J K−2 mol−1 ) × (300 K − 273 K)
= (+8.62... J K−1 mol−1 ) + (+2.02... J K−1 mol−1 )
= +10.7 J K−1 mol−1 .
P3B.5 Two identical blocks must come to their average temperature. Therefore the
final temperature is
T = 21 (Tc + Th )
Although the above result may seem self-evident, the more detailed explaina-
tion is as follows. The heat capacity at constant volume is defined in [2A.14–
41], C V = (∂U/∂T)V . As shown in Section 2A.4(b) on page 41, if the heat
capacity is constant, the internal energy changes linearly with the change in
temperature. That is ∆U = C V ∆T = C V (Tf − Ti ). For the two blocks at
the initial temperatures of Tc and Th , the change in internal energy to reach
the final temperature T is ∆U c = C V ,c (T − Tc ) and ∆U h = C V ,h (T − Th ),
respectively. The blocks of metal are made of the same substance and are of
the same size, therefore C V ,c = C V ,h = C V . Note that for a given solid the
internal energy does not change significantly on the volume or pressure, thus
it can be assumed that C V = C p . Assuming the system is isolated the total
change in internal energy is ∆U = ∆U c + ∆U h = 0. This means that ∆U =
C p ((T − Tc ) − (T − Th )) = C p × (2T − (T1 + T2 )) = 0, which implies that the
final temperature is T = 21 (Tc + Th ), as stated above.
At constant pressure the temperature dependence of the entropy is given by
[3B.7–86],
Tf
∆S = nC p,m ln ( )
Ti
Therefore for the two blocks

T T
∆S c = nC p,m ln ( ) and ∆S h = nC p,m ln ( )
Tc Th
The total change in entropy is
∆S tot = ∆S c + ∆S h
T T
= nC p,m ln ( ) + nC p,m ln ( )
Tc Th
T2
= nC p,m × ln ( )
Tc × Th
2
m ⎛ [ 1 (Tc + Th )] ⎞
= C p,m × ln 2
M ⎝ Tc × Th ⎠
m (Tc + Th )2
= C p,m ln ( ).
M 4(Tc × Th )
where m is the mass of the block and M is the molar mass.
In the case given
500 g −1 −1 (250 K + 500 K)2
∆S tot = × (24.4 J K mol ) × ln ( )
63.55 g mol−1 4 × (250 K × 500 K)
= +22.6 J K−1 .
P3B.7 The heat produced by the resistor over a time period ∆t is q = power × ∆t =
IV ∆t = I 2 R∆t, where the last expression was obtained using Ohm’s law, V =
IR. Note that care is needed handling the units. From the inside of the front
cover of the textbook use (1 A) ≡ (1 Cs−1 ) and (1 V) ≡ (1 JC−1 ), so that
(1 Ω) ≡ (1 JsC−2 ). Therefore the units of the final expression for the heat are
as expected
A2 × Ω × s ≡ (C2 s−2 ) × (JsC−2 ) × (s) ≡ J
Assuming that all the heat is absorbed by the large metal block at constant
pressure, this heat is the change of enthalpy of the system, ∆H = q. The enthalpy
change on heating is given by [2B.6b–47], ∆H = C p ∆T. This is rearranged to
give an expression for a temperature change
∆H q I 2 R∆t I 2 R∆t
∆T = = = =
Cp Cp Cp (m/M)C p,m
where m is the mass, M the molar mass and C p,m the molar heat capacity. Thus
the final temperature of the metal block is
I 2 R∆t
Tf = Ti + ∆T = Ti +
(m/M)C p,m
(1.00 A)2 × (1.00 × 103 Ω) × (15.0 s)
= (293 K) +
[(500 g)/(63.55 g mol−1 )] × (24.4 J K−1 mol−1 )
= (293 K) + 78.1... K = 3.71... × 102 K.
The variation of entropy with temperature at constant pressure is given by [3B.7–

86], S(Tf ) = S(Ti ) + C p ln (Tf /Ti ). Therefore the change in entropy is
Tf m Tf
∆S = S(Tf ) − S(Ti ) = C p ln ( ) = ( ) C p,m ln ( )
Ti M Ti
500 g −1 −1 3.71... × 102 K
=( −1 ) × (24.4 J K mol ) × ln ( )
63.55 g mol 293 K
= +45.4 J K−1 .
For the second experiment, the initial and final states of the metal block is the
same, therefore ∆S = 0 . All the heat is released into surroundings, that is water
bath, which can be assumed to be large enough to retain constant temperature.
Thus
q I 2 R∆t
∆S sur = =
Tsur Tsur
(1.00 A)2 × (1.00 × 103 Ω) × (15.0 s)
= = +51.2 J K−1 .
293 K
P3B.9 As suggested in the hint, first consider heating the folded protein at constant
pressure to from the initial temperature T to that of the transition, Ttrs . The
variation of entropy with temperature at constant pressure is given by [3B.7–
86], S(Tf ) = S(Ti ) + C p ln (Tf /Ti ). Thus the change in molar entropy, ∆S m =
S m (Tf ) − S m (Ti ), of this step is
Ttrs
∆S 1,m = C p,m (folded) ln ( )
T
Next consider the unfolding step. The entropy change of such a transition is
given by [3B.4–85], ∆ trs S = ∆ trs H/Ttrs , thus
∆ trs H −m
○
∆S 2,m =
Ttrs
The final step is cooling the unfolded protein to the initial temperature
T Ttrs
∆S 3,m = C p,m (unfolded) ln ( ) = −C p,m (unfolded) ln ( ).
Ttrs T
The overall entropy change is the sum of above steps
∆S m = ∆S 1,m + ∆S 2,m + ∆S 3,m

Ttrs ∆ trs H −m
○
Ttrs
= C p,m (folded) ln ( )+ − C p,m (unfolded) ln ( )
T Ttrs T
∆ trs H −m
○
Ttrs
= + [C p,m (folded) − C p,m (unfolded)] × ln ( )
Ttrs T
Given that C p,m (unfolded) − C p,m (folded) = 6.28 × 103 J K−1 mol−1 , the molar
entropy of unfolding at 25.0 ○ C is thus
5.09 × 105 J mol−1

∆S m =
273.15 K + 75.5 K
273.15 K + 75.5 K
+ (−6.28 × 103 J K−1 mol−1 ) × ln ( )
273.15 K + 25.0 K
= (1.45... × 103 J K−1 mol−1 ) + (−9.82 × 102 J K−1 mol−1 )
= +4.77... × 102 J K−1 mol−1 = +477 J K−1 mol−1 .
P3B.11 (a) Consider a process in which heat ∣dq∣ is extracted from the cold source at
temperature Tc , and heat q h = ∣dq∣ + ∣dw∣ is discarded into the hot sink at
temperature Th . The overall entropy change of such process is
−∣dq∣ ∣dq∣ + ∣dw∣

dS = +
Tc Th
For the process to be permissible by the Second Law, the Clausius inequal-
ity defined in [3A.12–82], dS ≥ 0, must be satisfied. Therefore
−∣dq∣ ∣dq∣ + ∣dw∣

+ ≥0
Tc Th
the equality implies the minimum amount of work for which the process
is permissible. Hence it follows that
∣dq∣ ∣dq∣ + ∣dw∣

= .
Tc Th
(b) The expression in (a) is rearranged to find ∣dw∣ and the given relation,
dq = CdTc , is used to give
∣dq∣
∣dw∣ = Th − ∣dq∣
Tc
dTc
∣dw∣ = CTh ∣ ∣ − C∣dTc ∣
Tc
Integration of both sides between the appropriate limits gives
w Tf dTc Tf
∫ ∣dw ′ ∣ = CTh ∫ ∣ ∣ − C ∫ ∣dTc ∣
0 Ti Tc Ti
which evaluates to
Tf
∣w∣ = CTh ∣ln ( )∣ − C∣Tf − Ti ∣ .
Ti
(c) Using C = (m/M)C p,m , the work needed is

250 g −1 −1 273 K
∣w∣ = −1 × (75.3 J K mol ) × (293 K) × ∣ln ( )∣
18.02 g mol 293 K
250 g −1 −1
− −1 × (75.3 J K mol ) × ∣273 K − 293 K∣
18.02 g mol
= ∣ − 2.16... × 104 ∣ J − ∣ − 2.08... × 104 ∣ J
= +7.47... × 102 J = +7.5 × 102 J .
(d) Assuming constant temperature, for finite amounts of heat and work, the
expression derrived in (a) becomes
∣q∣ ∣q∣ + ∣w∣
=
Tc Th
This is rearranged to give the work as
Th
∣w∣ = ( − 1) × ∣q∣
Tc
The heat transferred during freezing is equal to the enthalpy of the tran-
sition, which is the opposite of fusion, q = ∆ trs H = (m/M)(−∆ fus H −○ ).
Therefore the work needed is
293 K 250 g
∣w∣ = ( − 1) × ∣ × (−6.01 × 103 J K−1 mol−1 )∣
273 K 18.02 g mol−1
= 6.10... × 103 J = 6.11 × 103 J .
(e) The total work is the sum of the two steps described in (c) and (d). There-
fore
w tot = (+7.47... × 102 J) + (6.10... × 103 J)

= +6.85... × 103 J = +6.86 kJ .
(f) Assuming no energy losses, power is the total work divided by the time
interval over which the work is done, P = w tot /∆t, hence
w tot 6.85... × 103 J

∆t = = = 68.6 s .
P 100 W
3C The measurement of entropy

D3C.1 Because solutions of cations cannot be prepared in the absence of anions, the
standard molar entropies of ions in solution are reported on a scale in which,
by convention, the standard entropy of H+ ions in water is taken as zero at all
temperatures: S −○ (H+ , aq) = 0.
Because the entropies of ions in water are values relative to the hydrogen ion in
water, they may be either positive or negative. A positive entropy means that
an ion has a higher molar entropy than H+ in water, and a negative entropy
means that the ion has a lower molar entropy than H+ in water. An ion with
zero entropy in fact has that same entropy as H+ .
E3C.1(a) Consider the chemical equation
1
N (g) + 32 H2 (g)
2 2
Ð→ NH3 (g)
The standard reaction entropy is given by [3C.3b–90], ∆ r S −○ = ∑J ν J S m
−
○
(J),
where ν J are singed stoichiometric coefficients for a given reaction equation.
Therefore, using data from the Resource section
∆ r S −○ = nS m
−
○ −
○
(NH3 , (g)) − 23 nS m −
○
(H2 , (g)) − 12 nS m (N2 , (g))
= (1.00 mol) × (192.45 J K−1 mol−1 )
− ( 23 × 1.00 mol) × (130.684 J K−1 mol−1 )
− ( 21 × 1.00 mol) × (191.61 J K−1 mol−1 )
= −99.38 J K−1 .
E3C.2(a) Assuming that the Debye extrapolation is valid, the constant-pressure molar
heat capacity is C p,m (T) = aT 3 . The temperature dependence of the entropy
T
is given by [3C.1a–88], S(T2 ) = S(T1 ) = ∫T12 (C p,m /T)dT. For a given temper-
ature T the change in molar entropy from zero temperature is therefore
T C p,m T aT ′ 3
′
S m (T) − S m (0) = ∫ dT = ∫ dT ′
0 T′ 0 T′
T aT 3 C p,m (T)
= a ∫ T ′ dT ′ =
2
=
0 3 3
Hence
C p,m (4.2 K) 0.0145 J K−1 mol−1
S m (4.2 K) − S m (0) = =
3 3
= 4.8 × 10−3 J K−1 mol−1 .
E3C.3(a) The standard reaction entropy is given by [3C.3b–90], ∆ r S −○ = ∑J ν J S m

−
○
(J),
where ν J are the signed stoichiometric numbers.
(i)
∆ r S −○ = 2S m
−
○ −
○
(CH3 COOH, (l)) − 2S m −
○
(CH3 CHO, (g)) − S m (O2 , (g))
= 2 × (159.8 J K−1 mol−1 ) − 2 × (250.3 J K−1 mol−1 )
− (205.138 J K−1 mol−1 )
= −386.1 J K−1 mol−1 .
(ii)
∆ r S −○ = 2S m
−
○ −
○
(AgBr, (s)) + S m −
○
(Cl2 , (g)) − 2S m −
○
(AgCl, (s)) − S m (Br2 , (l))
= 2 × (107.1 J K−1 mol−1 ) + (223.07 J K−1 mol−1 )
− 2 × (96.2 J K−1 mol−1 ) − (152.23 J K−1 mol−1 )
= +92.6 J K−1 mol−1 .
(iii)
∆ r S −○ = S m
−
○ −
○
(HgCl2 , (s)) − S m −
○
(Hg, (l)) − S m (Cl2 , (g))
= (146.0 J K−1 mol−1 ) − (76.02 J K−1 mol−1 )
− (223.07 J K−1 mol−1 )
= −153.1 J K−1 mol−1 .
P3C.1 Consider the process of determining the calorimetric entropy from zero to
the temperature of interest. Assuming that the Debye extrapolation is valid,
the constant-pressure molar heat capacity at the lowest temperatures is of a
form C p,m (T) = aT 3 . The temperature dependence of the entropy is given
T
by [3C.1a–88], S(T2 ) = S(T1 ) = ∫T12 (C p,m /T)dT. Thus for a given (low)
temperature T the change in molar entropy from zero is
T C p,m T aT ′ 3
′
S m (T) − S m (0) = ∫ dT = ∫ dT ′
0 T′ 0 T′
T a
= a ∫ T ′ dT ′ = T 3 = 31 C p,m (T)
2
0 3
Hence
−
○
Sm −
○
(10 K) − S m (0) = 1
3
× (4.64 J K−1 mol−1 ) = 1.54... J K−1 mol−1
−
○
The increase in entropy on raising the temperature to the melting point is S m (234.4 K)−
−
○ −1 −1
S m (10 K) = 57.74 J K mol . The entropy change of a phase transition is
given by [3C.1b–88], ∆ trs S(Ttrs ) = ∆ trs H(Ttrs )/Ttrs . Thus
2322 J mol−1
−
○
∆ fus S m (234.4 K) = = 9.90... J K−1 mol−1
234.4 K
Further raising the temperature to 298.0 K gives an increase in the entropy of

−
○
Sm −
○
(298 K) − S m (234.4 K) = 6.85 J K−1 mol−1 .
The Third-Law standard molar entropy at 298 K is the sum of the above con-
tributions.
−
○ −
○ −
○ −
○ −
○ −
○
Sm (298 K) − S m (0) = (S m (10 K) − S m (0)) + (S m (234.4 K) − S m (10 K))
−
○ −
○ −
○
+ ∆ fus S m (234.4 K) + (S m (298 K) − S m (234.4 K))
= (1.54... J K−1 mol−1 ) + (57.74 J K−1 mol−1 )
+ (9.90... J K−1 mol−1 ) + (6.85 J K−1 mol−1 )
= 76.04 J K−1 mol−1 .
P3C.3 (a) Assuming that the Debye extrapolation is valid, the constant-pressure
molar heat capacity is of a form C p,m (T) = aT 3 . The temperature depen-
T
dence of the entropy is given by [3C.1a–88], S(T2 ) = S(T1 ) = ∫T12 (C p,m /T)dT.
Thus for a given (low) temperature T the change in the molar entropy
from zero is
T C p,m T aT ′ 3
′
S m (T) − S m (0) = ∫ dT = ∫ dT ′
0 T′ 0 T′
T a
= a ∫ T ′ dT ′ = T 3 = 13 C p,m (T)
2
0 3
Hence
−
○
Sm −
○
(10 K) − S m (0) = 1
3
× (2.8 J K−1 mol−1 ) = 0.933... J K−1 mol−1
= 0.93 J K−1 mol−1 .
(b) The change in entropy is determined calorimetrically by measuring the

area under a plot of (C p,m /T) against T, as shown in Fig. 3.2.
The plot is rather irregular and is best fitted by two polynomials of order
3: one in the range 10 K to 30 K and the other in the range 30 K to 298 K.
Define y = (C p,m /T)/(J K−2 mol−1 ) and x = T/K, so that the fitted
function is expressed
y = c3 x 3 + c2 x 2 + c1 x + c0
where the best fitted coefficients c i for the respective temperature ranges
are
ci 10 K to 30 K 30 K to 298 K
c3 +5.0222 × 10−5 −5.2881 × 10−8
c2 −4.3010 × 10−3 +3.5425 × 10−5
c1 +1.2025 × 10−1 −8.1107 × 10−3
c0 −5.4187 × 10−1 +7.5533 × 10−1
The integral of the fitted functions over the range x i to x f is
xf
I=∫ c 3 x 3 + c 2 x 2 + c 1 x + c 0 dx
xi
c3 c2 c1
= (x f 4 − x i 4 ) + (x f 3 − x i 3 ) + (x f 2 − x i 2 ) + c 0 (x f − x i )
4 3 2
T/K C p,m /(J K−1 mol−1 ) (C p,m /T)/(J K−2 mol−1 )

10 2.8 0.280 0
15 7.0 0.466 7
20 10.8 0.540 0
25 14.1 0.564 0
30 16.5 0.550 0
50 21.4 0.428 0
70 23.3 0.332 9
100 24.5 0.245 0
150 25.3 0.168 7
200 25.8 0.129 0
250 26.2 0.104 8
298 26.6 0.089 3
0.6
(C p,m /T)/(J K−2 mol−1 )
0.4
0.2
0.0
0 50 100 150 200 250 300
T/K
Figure 3.2
Using the appropriate coefficients and limits the integrals are evaluated to
give the respective changes in entropy
−
○
Sm −
○
(30 K) − S m (10 K) = 10.0... J K−1 mol−1
−
○
Sm −
○
(298 K) − S m (30 K) = 53.8... J K−1 mol−1
The total entropy change is the sum of the two integrals. Therefore
−
○
Sm −
○
(298 K) − S m (10 K) = (10.0... J K−1 mol−1 ) + (53.8... J K−1 mol−1 )
= 63.9... J K−1 mol−1 = 63.9 J K−1 mol−1 .
(c) The standard Third-Law entropy at 298 K is the sum of the above calcu-
lated contributions. Thus

−
○ −
○ −
○ −
○
Sm (298 K) − S m (0) = (S m (298 K) − S m (10 K))
−
○ −
○
+ (S m (10 K) − S m (0))
= (63.9... J K−1 mol−1 ) + (0.933... J K−1 mol−1 )
= 64.8 J K−1 mol−1 .
For the standard Third-Law entropy at 273 K, the second integral in part
(b) needs to be repeated with Tf = 273 K. Therefore
−
○
Sm −
○
(273 K) − S m (30 K) = 51.4... J K−1 mol−1
The other contributions are the same, hence
−
○
Sm −
○
(273 K) − S m (0) = (10.0... J K−1 mol−1 + 51.4... J K−1 mol−1 )
+ (0.933... J K−1 mol−1 )
= 62.4 J K−1 mol−1 .
P3C.5 The standard reaction entropy is given by [3C.3b–90], ∆ r S −○ = ∑J ν J S m

−
○
(J),
∆ r S −○ (298 K) = S m
−
○ −
○
(CO, (g)) + S m (H2 O, (g))
−
○ −
○
− Sm (CO2 , (g)) − S m (H2 , (g))
= (197.67 J K−1 mol−1 ) + (188.83 J K−1 mol−1 )
− (213.74 J K−1 mol−1 ) − (130.684 J K−1 mol−1 )
= +42.07... J K−1 mol−1 = +42.08 J K−1 mol−1 .
Similarly, the standard reaction enthalpy is given by [2C.5b–53], ∆ r H −○ = ∑J ν J ∆ f H −○ (J).

∆ r H −○ (298 K) = ∆ f H −○ (CO, (g)) + ∆ f H −○ (H2 O, (g))
− ∆ f H −○ (CO2 , (g)) − ∆ f H −○ (H2 , (g))
= (−110.53 kJ mol−1 ) + (−241.82 kJ mol−1 )
− (−393.51 kJ mol−1 ) − 0
= +41.16 kJ mol−1 .
The temperature dependence of the reaction entropy is given by [3C.5a–91],

T
∆ r S −○ (T2 ) = ∆ r S −○ (T1 ) + ∫T12 (∆ r C −p○ /T)dT. Similarly, the enthalpy depen-
dence on temperature is given by Kirchhoff ’s law [2C.7a–53], ∆ r H −○ (T2 ) =
T
∆ r H −○ (T1 ) + ∫T12 ∆ r C −p○ dT. The quantity ∆ r C −p○ is defined in [3C.5b–91], ∆ r C −p○ =
∑J ν J C −p,m
○
(J). For the reaction at 298 K
∆ r C −p○ = C −p,m
○
(CO, (g)) + C −p,m
○
(H2 O, (g))
− C −p,m
○
(CO2 , (g)) − C −p,m
○
(H2 , (g))
= (29.14 J K−1 mol−1 ) + (33.58 J K−1 mol−1 )
− (37.11 J K−1 mol−1 ) − (28.824 J K−1 mol−1 )
= −3.21... J K−1 mol−1
Assuming that ∆ r C −p○ is constant over the temperature range involved, the stan-
dard entropy and enthalpy changes of the reaction is given by, respectively,
[3C.5b–91], ∆ r S −○ (T2 ) = ∆ r S −○ (T1 )+∆ r C −p○ ln(T2 /T1 ), and [2C.7d–54], ∆ r H −○ (T2 ) =
∆ r H −○ (T1 ) + ∆ r C −p○ (T2 − T1 ).
398 K
∆ r S −○ (398 K) = ∆ r S −○ (298 K) + ∆ r C −p○ × ln ( )
298 K
= (+42.0... J K−1 mol−1 )
398
+ (−3.21... J K−1 mol−1 ) × ln ( )
298
= +41.15 J K−1 mol−1 .
∆ r H −○ (398 K) = ∆ r H −○ (298 K) + ∆ r C −p○ × (398 K − 298 K)

= (+41.1... × 103 J mol−1 )
+ (−3.21... J K−1 mol−1 ) × (100 K)
= +40.8... × 103 J mol−1 = +40.8 kJ mol−1 .
P3C.7 Assuming that the Debye extrapolation is valid, the constant-pressure molar
heat capacity is of a form C p,m (T) = aT 3 . The temperature dependence of the
T
entropy is given by [3C.1a–88], S(T2 ) = S(T1 ) = ∫T12 (C p,m /T)dT. Thus for a
given (low) temperature T the change in the molar entropy from zero is
C p,m
T T aT ′ 3
′
S m (T) − S m (0) = ∫ dT = ∫ dT ′
0 T′ 0 T′
T a
= a ∫ T ′ dT ′ = T 3 = 31 C p,m (T)
2
0 3
Hence
−
○
Sm −
○
(14.14 K) − S m (0) = 1
3
× (9.492 J K−1 mol−1 ) = 3.16... J K−1 mol−1
The change in entropy is determined calorimetrically by measuring the area

under a plot of (C p,m /T) against T.
T/K C p,m /(J K−1 mol−1 ) (C p,m /T)/(J K−2 mol−1 )

14.14 9.492 0.671 29
16.33 12.70 0.777 71
20.03 18.18 0.907 64
31.15 32.54 1.044 62
44.08 46.86 1.063 07
64.81 66.36 1.023 92
100.90 95.05 0.942 02
140.86 121.3 0.861 14
183.59 144.4 0.786 54
225.10 163.7 0.727 23
262.99 180.2 0.685 20
298.06 196.4 0.658 93
1.2
(C p,m /T)/(J K−2 mol−1 )
1.0
0.8
0.6
0 50 100 150 200 250 300

T/K
The plot is rather irregular and is best fitted by two polynomials of order 2
and 3, respectively, in the ranges 14.14 K to 44.08 K and 44.08 K to 298.06 K.
Define y = (C p,m /T)/(J K−2 mol−1 ) and x = T/K, so that the fitted function
is expressed
y = c3 x 3 + c2 x 2 + c1 x + c0
where the best fitted coefficients c i for the respective temperature ranges are
ci 14.14 K to 44.08 K 44.08 K to 298.06 K

c3 0 +7.1979 × 10−9
c2 −7.6119 × 10−4 −3.0830 × 10−7
c1 +5.6367 × 10−2 −2.2415 × 10−3
c0 +5.1090 × 10−2 +1.1644 × 100
The integral of the fitted functions over the range x i to x f is

xf
I=∫ c 3 x 3 + c 2 x 2 + c 1 x + c 0 dx
xi
c3 c2 c1
= (x f 4 − x i 4 ) + (x f 3 − x i 3 ) + (x f 2 − x i 2 ) + c 0 (x f − x i )
4 3 2
The temperatures of interest are beyond 44.08 K. Thus the contribution cor-
responding to the integral of the quadratic is included in the estimates of the
entropy. Using the appropriate coefficients and limits the integral gives
−
○
Sm −
○
(44.08 K) − S m (14.14 K) = 29.6... J K−1 mol−1
Finally, the remaining contribution is found by estimating the integral of the

cubic polynomial to the temperature of interest. Therefore for T = 100 K, 200 K
and, after small extrapolation, 300 K
−
○
Sm −
○
(100 K) − S m (44.08 K) = 56.1... J K−1 mol−1
−
○
Sm −
○
(200 K) − S m (44.08 K) = 1.41... × 102 J K−1 mol−1
−
○
Sm −
○
(300 K) − S m (44.08 K) = 2.11... × 102 J K−1 mol−1
The standard Third-Law molar entropy at the temperatures of interests is the

sum of all the contributions up to that point. Thus, the entropy at three tem-
peratures are
−
○ −
○ −
○ −
○
Sm (100 K) − S m (0) = S m (14.14 K) − S m (0)
−
○ −
○
+ (S m (44.08 K) − S m (14.14 K))
−
○ −
○
+ (S m (100 K) − S m (44.08 K))
= (3.16... J K−1 mol−1 ) + (29.6... J K−1 mol−1 )
+ (56.1... J K−1 mol−1 )
= 89.0 J K−1 mol−1 .
Similarly it is found for 200 K and 300 K, respectively
−
○
Sm −
○
(200 K) − S m (0) = (3.16... J K−1 mol−1 ) + (29.6... J K−1 mol−1 )
+ (1.41... × 102 J K−1 mol−1 )
= 173.8 J K−1 mol−1 .
−
○
Sm −
○
(300 K) − S m (0) = (3.16... J K−1 mol−1 ) + (29.6... J K−1 mol−1 )
+ (2.11... × 102 J K−1 mol−1 )
= 243.9 J K−1 mol−1 .
P3C.9 (a) Given the expression for the constant-pressure molar heat capacity, C p,m (T) =
aT 3 + bT, consider C p,m /T.
C p,m aT 3 + bT
= = aT 2 + b
T T
This expression is of the form of a straight line, y = (slope)×x+(intercept),

if y = C p,m (T) and x = T 2 . It follows that (slope) = a and (intercept) = b.
(b) The data below are plotted in Fig. 3.3.
T/K C p,m /(J K−1 mol−1 ) T 2 /(K2 ) (C p,m /T)/(J K−2 mol−1 )
0.20 0.437 0.040 2.185 0
0.25 0.560 0.063 2.240 0
0.30 0.693 0.090 2.310 0
0.35 0.838 0.123 2.394 3
0.40 0.996 0.160 2.490 0
0.45 1.170 0.203 2.600 0
0.50 1.361 0.250 2.722 0
0.55 1.572 0.303 2.858 2
3.0
(C p,m /T)/(J K−2 mol−1 )
2.8
2.6
2.4
2.2
2.0
0.05 0.10 0.15 0.20 0.25 0.30 0.35
T/K
Figure 3.3
The data lie on a good straight line, the equation of which is
(C p,m /T)/(J K−2 mol−1 ) = 2.569 × (T/K)2 + 2.080
Thus a = 2.569 JK−4 mol−1 and b = 2.080 JK−2 mol−1 .

(c) The dependence of the entropy on temperature is given by [3C.1a–88],
T
S(T2 ) = S(T1 ) + ∫T12 (C p,m /T)dT. Thus for a given (low) temperature T
the molar entropy change from the zero temperature is
T C p,m ′
T
(aT ′ + b) dT ′
2
S m (T) = S m (0) + ∫ dT = S m (0) + ∫
0 T′ 0
a
= S m (0) + T 3 + bT .
3
(d) Assuming that the expression derived above can be extrapolated to 2.0 K
(2.569 JK−4 mol−1 )

S m (2.0 K) − S m (0) = × (2.0 K)3
3
+ (2.080 JK−2 mol−1 ) × (2.0 K)
= 11.01 J K−1 mol−1 .
3D Concentrating on the system
D3D.1 As is discussed in detail in Topic 3D the criteria for spontaneity at constant

volume and temperature is expressed in terms of the Helmholtz energy, dA ≤ 0,
and at constant pressure and temperature in terms of the Gibbs energy, dG ≤ 0.
Both the Helmholtz and Gibbs energies refer to properties of the system alone.
However, because of the way they are defined these quantities effectively allow
the entropy change of the system plus surroundings to be evaluated. For exam-
ple, at constant volume and temperature the change in the Helmholtz energy
is expressed in terms of the internal energy change and the entropy change
of the system: dA = dU − TdS. If this expression is divided by −T to give
−dA/T = −dU/T + dS the two terms on the right can bothe be identified as
entropy changes.
The first term, −dU/T, is equal to the entropy change of the surroundings
because dq sur = −dq, and at constant volume dq = dU. The second term is
the entropy change of the system. Thus the sum of the two is the total en-
tropy change, which the Second Law shows must be positive in a spontaneous
process. Therefore, the change in the Helmholtz energy is an indicator of the
total entropy change, even though the former refers only to the system. Similar
considerations can be applied to the Gibbs energy.
It is also possible to express the criterion for spontaneity in terms of the change
in H, U or S for the system. For example, as shown in Topic 3D, dS U ,V ≥ 0.
However, the variables which are being held constant (here U and V ) do not
correspond to such easily realizable conditions such as constant temperature
and volume (or pressure) so such criteria are less applicable to chemical sys-
tems.
E3D.1(a) The maximum non-expansion work is equal to the Gibbs free energy as ex-
plained in Section 3D.1(e) on page 96. The standard reaction Gibbs energy is
given by [3D.10b–97], ∆ r G −○ = ∑J ν J ∆ f G −○ (J), where ν J are the signed stoichio-
metric numbers. For the reaction CH4 (g) + 3O2 (g) Ð→ CO2 (g) + 2H2 O(l)
∆ r G −○ = ∆ f G −○ (CO2 ,(g)) + 2∆ f G −○ (H2 O ,(l)) − ∆ f H −○ (CH4 ,(g))
− 3∆ f H −○ (O2 ,(g))
= (−394.36 kJ mol−1 ) + 2 × (−237.13 kJ mol−1 )
− (−50.72 kJ mol−1 ) − 3 × 0
= −817.90 kJ mol−1 .
Therefore, ∣w add,max ∣ = ∣∆ r G −○ ∣ = 817.90 kJ mol−1 .

E3D.2(a) The standard reaction Gibbs energy is given by [3D.10b–97], ∆ r G −○ = ∑J ν J ∆ f G −○ (J),
(i)
∆ r G −○ = 2∆ f G −○ (CH3 COOH ,(l)) − 2∆ f G −○ (CH3 CHO ,(g)) − ∆ f G −○ (O2 ,(g))
= 2 × (−389.9 kJ mol−1 ) − 2 × (−128.86 kJ mol−1 ) − 0
= −522.1 kJ mol−1 .
(ii)
∆ r G −○ = 2∆ f G −○ (AgBr ,(s)) + ∆ f G −○ (Cl2 ,(g)) − 2∆ f G −○ (AgCl ,(s))
− ∆ f G −○ (Br2 ,(l))
= 2 × (−96.90 kJ mol−1 ) + 0 − 2 × (−109.79 kJ mol−1 ) − 0
= +25.78 kJ mol−1 .
(iii)
∆ r G −○ = ∆ f G −○ (HgCl2 ,(s)) − ∆ f G −○ (Hg ,(l)) − ∆ f G −○ (Cl2 ,(g))
= (−178.6 kJ mol−1 ) − 0 − 0 = −178.6 kJ mol−1 .
E3D.3(a) Consider the reaction

CH3 COOC2 H5 (l) + 5O2 (g) Ð→ 4CO2 (g) + 4H2 O(l)
The standard reaction enthalpy is [2C.5b–53], ∆ r H −○ = ∑J ν J ∆ f H −○ (J), where ν J
are the signed stoichiometric numbers.
∆ r H −○ = 4∆ f H −○ (CO2 ,(g)) + 4∆ f H −○ (H2 O ,(l))
− ∆ f H −○ (CH3 COOC2 H5 ,(l)) − 5∆ f H −○ (O2 ,(g))
when rearranged this gives
∆ f H −○ (CH3 COOC2 H5 ,(l)) = 4∆ f H −○ (CO2 ,(g)) + 4∆ f H −○ (H2 O ,(l))
− 5∆ f H −○ (O2 ,(g)) − ∆ r H −○
= 4 × (−393.51 kJ) + 4 × (−285.83 kJ)
− 5 × 0 − (−2231 kJ mol−1 )
= −486.36 kJ mol−1
The standard reaction entropy is given by [3C.3b–90], ∆ r S −○ = ∑J ν J S m

−
○
(J).
Therefore, for the formation of the compound
∆ f S −○ (CH3 COOC2 H5 ,(l)) = S m

−
○ −
○
(CH3 COOC2 H5 ,(l)) − 4S m (C ,(s))
−
○ −
○
− 4S m (H2 ,(g)) − S m (O2 ,(g))
= (259.4 J K−1 mol−1 ) − 4 × (5.740 J K−1 mol−1 )
− 4 × (130.684 J K−1 mol−1 ) − (205.138 J K−1 mol−1 )
= −4.91... × 102 J K−1 mol−1
The standard reaction Gibbs energy is defined in [3D.9–96], ∆ r G −○ = ∆ r H −○ −

T∆ r S −○ , thus
∆ f G −○ (CH3 COOC2 H5 ,(l)) = (−486.36 kJ mol−1 )

− (298.15 K)(−0.491... kJ K−1 mol−1 )
= −340 kJ mol−1 .
E3D.4(a) The standard reaction Gibbs energy is given by [3D.9–96], ∆ r G −○ = ∆ r H −○ −

T∆ r S −○ . The standard reaction enthalpy is given in terms of the enthalpies
of formation by [2C.5b–53], ∆ r H −○ = ∑J ν J ∆ f H −○ (J), where ν J are the signed
stoichiometric numbers.
(i)
∆ r H −○ = 2∆ f H −○ (CH3 COOH ,(l)) − 2∆ f H −○ (CH3 CHO ,(g))

− ∆ f H −○ (O2 ,(g))
= 2 × (−484.5 kJ mol−1 ) − 2 × (−166.19 kJ mol−1 ) − 0
= −636.62 kJ mol−1 = −636.6 kJ mol−1 .
Given the result for the previous execise, ∆ r S −○ = −386.1 J K−1 mol−1
∆ r G −○ = (−636.62 kJ mol−1 ) − (298.15 K) × (−0.3861 kJ K−1 mol−1 )

= −521.5 kJ mol−1 .
(ii)
∆ r H −○ = 2∆ f H −○ (AgBr ,(s)) + ∆ f H −○ (Cl2 ,(g)) − 2∆ f H −○ (AgCl ,(s))

− ∆ f H −○ (Br2 ,(l))
= 2 × (−100.37 kJ mol−1 ) + 0 − 2 × (−127.07 kJ mol−1 ) − 0
= +53.40 kJ mol−1 .
Given the result for the previous execise, ∆ r S −○ = +92.6 J K−1 mol−1
∆ r G −○ = (+53.40 kJ mol−1 ) − (298.15 K) × (+0.0926 kJ K−1 mol−1 )

= +25.8 kJ mol−1 .
(iii)
∆ r H −○ = ∆ f H −○ (HgCl2 ,(s)) − ∆ f H −○ (Hg ,(l)) − ∆ f H −○ (Cl2 ,(g))

= (−224.3 kJ mol−1 ) − 0 − 0 = −224.3 kJ mol−1 .
Given the result for the previous execise, ∆ r S −○ = −153.1 J K−1 mol−1 .
∆ r G −○ = (−224.3 kJ mol−1 ) − (298.15 K) × (−0.1531 kJ K−1 mol−1 )

= −178.7 kJ mol−1 .
E3D.5(a) The standard reaction entropy is given by [3C.3b–90], ∆ r S −○ = ∑J ν J S m

−
○
(J),
∆ r S −○ = 2S m
−
○ −
○
(I2 ,(s)) + 2S m −
○
(H2 O ,(l)) − 4S m −
○
(HI ,(g)) − S m (O2 ,(g))
= 2 × (116.135 J K−1 mol−1 ) + 2 × (69.91 J K−1 mol−1 )
− 4 × (206.59 J K−1 mol−1 ) − (205.138 J K−1 mol−1 )
= −659.40... J K−1 mol−1
The standard reaction enthalpy is given by [2C.5b–53], ∆ r H −○ = ∑J ν J ∆ f H −○ (J).
∆ r H −○ = 2∆ f H −○ (I2 ,(s)) + 2∆ f H −○ (H2 O ,(l)) − 4∆ f H −○ (HI ,(g))

− ∆ f H −○ (O2 ,(g))
= 2 × 0 + 2 × (−285.83 kJ mol−1 ) − 4 × (+26.48 kJ mol−1 ) − 0
= −677.58 kJ mol−1
The standard reaction Gibbs energy is given by [3D.9–96], ∆ r G −○ = ∆ r H −○ −

T∆ r S −○ .
∆ r G −○ = (−677.58 kJ mol−1 ) − (298.15 K) × (−0.65940... kJ K−1 mol−1 )

= −480.98 kJ mol−1 .
P3D.1 (a) From the perfect gas law, pV = nRT, the final pressure is
n B RTB (2.00 mol) × (8.3145 × 10−2 dm3 bar K−1 mol−1 ) × (300 K)
p= =
VB,f 1.00 dm3
= 49.8... bar = 49.9 bar .
The final volume of Section A is VA,f = (VA + VB ) − VB,f = 3.00 dm3 .

Therefore the final temperature of Section A is
pVA,f (49.8... bar) × (3.00 dm3 )

TA,f = =
nR (2.00 mol) × (8.3145 × 10−2 dm3 bar K−1 mol−1 )
= 900 K .
(b) Taking the hint, first consider the heating at constant volume. The entropy
dependence on temperature at constant volume is given by [3B.7–86],
∆S = nC V ,m ln (Tf /Ti ), with C p replaced by nC V ,m . The volume is then
allowed to expand to the final. The entropy change for the isothermal
expansion of a perfect gas is given by [3B.2–84], ∆S = nR ln(Vf /Vi ).
Therefore the total change in the entropy for the gas in Section A is
TA,f VA,f
∆S A = nC V ,m ln ( ) + nR ln ( )
TA VA
900 K
= (2.00 mol) × (20.0 J K−1 mol−1 ) × ln ( )
300 K
3.00 dm3
+ (2.00 mol) × (8.3145 J K−1 mol−1 ) × ln ( )
2.00 dm3
= +50.6... J K−1 = +50.7 J K−1 .
(c) Section B is kept at the constant temperature throughout the process, thus
only the change in the volume needs to be considered
VB,f
∆S B = nR ln ( )
VB
1.00 dm3
= (2.00 mol) × (8.3145 J K−1 mol−1 ) × ln ( )
2.00 dm3
= −11.5... J K−1 = −11.5 J K−1 .
(d) The change in internal energy as a result of a change in temperature as-

suming constant heat capacity is given by [2A.15b–43], ∆U = C V ∆T. Be-
cause the internal energy of a perfect gas depends only on the temperature
∆U A = (2.00 mol) × (20.0 J K−1 mol−1 ) × (900 K − 300 K)

= +2.40 × 104 J = +24.0 kJ .
∆U B = 0 .
(e) The Helmholtz energy is defined in [3D.4a–93], A = U −T S. For the finite

changes it becomes ∆A = ∆U − ∆(T S) = ∆U − T∆S − S∆T. Because
Section B is kept at constant temperature ∆T = 0 and so
∆A B = ∆U B − TB ∆S B
= 0 − (300 K) × (−11.5... J K−1 ) = +3.46 × 103 J .
Because ∆T is not zero and S is not given for the Section A, the equivalent
expression cannot be evaluated.
(f) Because the process is reversible, the total ∆A = ∆A A + ∆A B = 0 . This
implies ∆A A = −∆A B .
P3D.3 Consider the thermodynamic cycle shown in Fig. 3.4.

H+ (g) + e – (g) + I (g)
∆ ec G −○ (I)
= −E a (I)
∆ ion G −○ (H) = I(H)
H+ (g) + I – (g)
H (g) + I (g)
∆ f G −○ (H (g))
+∆ f G −○ (I (g)) ∆ solv G −○ (H+ )
+∆ solv G −○ (I− )
1
H
2 2
(g) + 12 I2 (g)
−∆ f G −○ (H+ (aq))
−∆ f G −○ (I− (aq))
H+ (aq) + I – (aq)
Figure 3.4
The sum of the Gibbs energies for all the steps around a closed cycle is zero.
0 = − [∆ f G −○ (H+ (aq)) + ∆ f G −○ (I− (aq))] + ∆ f G −○ (H (g)) + ∆ f G −○ (I (g))

I(H) + (−E a (I)) + ∆ solv G −○ (H+ ) + ∆ solv G −○ (I− )
Because ∆ f G −○ (H+ (aq) = 0, by convention, the Gibbs energy of formation for

the I – (aq) is
∆ f G −○ (I− (aq)) = ∆ f G −○ (H (g)) + ∆ f G −○ (I (g)) + I(H)

− E a (I) + ∆ solv G −○ (H+ ) + ∆ solv G −○ (I− )
= (203.25 kJ mol−1 ) + (70.25 kJ mol−1 ) + (1312.0 kJ mol−1 )
− (295.3 kJ mol−1 ) + (−1090 kJ mol−1 ) + (−247 kJ mol−1 )
= -47 kJ mol−1 .
P3D.5 The standard reaction Gibbs energy is given by [3D.9–96], ∆ r G −○ = ∆ r H −○ −

T∆ r S −○ . The standard reaction entropy is [2C.5b–53], ∆ r S −○ = ∑J ν J S m
−
○
(J), where
ν J are the signed stoichiometric numbers. Therefore
∆ r S 1−○ = S m
−
○
(Li+ ,(g)) + S m
−
○
(F− ,(g)) − S m
−
○
(LiF ,(s))
= (133 J K−1 mol−1 ) + (145 J K−1 mol−1 ) − (35.6 J K−1 mol−1 )
= 242.4 J K−1 mol−1
And so
∆ r G 1−○ = (1037 kJ mol−1 ) − (298 K) × (0.2424 kJ K−1 mol−1 )

= +9.64... × 102 kJ mol−1 = +965 kJ mol−1 .
For the second step ∆ r G 2−○ = ∆ solv G −○ (Li+ ) + ∆ solv G −○ (F− ). The Gibbs en-
ergy of solvation in water is given by Born equation [3D.12b–99], ∆ solv G −○ =
−(z i 2 /[r i /pm]) × 6.86 × 104 kJ mol−1 , thus
(+1)2 (−1)2
∆ r G 2−○ = − ( + ) × 6.86 × 104 kJ mol−1
[r(Li )/pm] [r(F− )/pm]
+
1 1
= −( + ) × 6.86 × 104 kJ mol−1 = −9.61... × 102 kJ mol−1
127 163
= −961 kJ mol−1
Therefore the total Gibbs energy change of the process
∆ r G −○ = ∆ r G 1−○ + ∆ r G 2−○ = (+9.64... × 102 kJ mol−1 ) + (−9.61... × 102 kJ mol−1 )

= +3.74... kJ mol−1 = +4 kJ mol−1 .
The change in positive implying that the reverse process is spontaneous.
3E Combining the First and Second Laws

D3E.1 The relation (∂G/∂p)T = V , combined with the fact that the volume is always
positive, shows that the Gibbs function of a system increases as the pressure
increases (at constant temperature).
E3E.1(a) As explained in Section 3E.2(c) on page 104, the change in Gibbs energy of a
p
phase transition varies with pressure as ∆ trs G m (p f ) = ∆ trs G m (p i ) ∫ p i f ∆ trs Vm dp.
Assuming that ∆ trs Vm changes little over the range of pressures considered
∆G m = ∆ trs G m (p f ) − ∆ trs G m (p i ) = (p f − p i )∆ trs Vm

= [(1000 × 105 Pa) − (1 × 105 Pa)] × (−1.6 × 10−6 m3 mol−1 )
= −1.6 × 102 J mol−1 .
E3E.2(a) The Gibbs energy dependence on pressure for a perfect gas is given by [3E.14–
105], G m (p f ) = G m (p i ) + RT ln(p f /p i ), thus
pf
∆G m = RT ln ( )
pi
100.0 atm
= (8.3145 J K−1 mol−1 ) × (298 K) × ln ( )
1.0 atm
= +11.4... × 103 J mol−1 = +11 kJ mol−1 .
E3E.3(a) The Gibbs energy dependence on temperature for a perfect gas is given by
[3E.14–105], G m (p f ) = G m (p i ) + RT ln(p f /p i ). From the perfect gas law p ∝
(1/V ). This allows rewriting the previous equation for the change in Gibbs
energy due to isothermal gas expansion
Vi
∆G = nRT ln ( )
Vf
42 cm3
= (2.5 × 10−3 mol) × (8.3145 J K−1 mol−1 ) × (300 K) × ln ( )
600 cm3
= −17 J .
E3E.4(a) The variation of the Gibbs energy with pressure is given by [3E.8–103], (∂G/∂T) p =
−S. The change in entropy is thus
∂G f ∂G i ∂(G f − G i )
∆S = S f − S i = − ( ) +( ) = −( )
∂T p ∂T p ∂T p
∂∆G ∂[(−85.40 J) + T × (36.5 J K−1 )]

= −( ) = −( )
∂T p ∂T p
= −(36.5 J K−1 ) = −36.5 J K−1 .
E3E.5(a) The Gibbs-Helmholtz relation for the change in Gibbs energy is given by [3E.11–
104], (∂[∆G/T]/∂T) p = −∆H/T 2 . Expressing for the change in enthalpy gives
∂[∆G/T] ∂[(−85.40 J)/T + (36.5 J K−1 )]

∆H = −T 2 ( ) = −T 2 ( )
∂T p
∂T p
−85.40 J
= T2 ( ) = −85.40 J .
T2
E3E.6(a) The molar Gibbs energy dependence on pressure for an incompressible sub-
stance is given by [3E.13–104], G m (p f ) = G m (p i )+(p f − p i )Vm . Assuming that
the volume of liquid octane changes little over the range of pressures considered
∆G = n[G m (p f ) − G m (p i )] = (p f − p i )nVm = (p f − p i )V
1.01325 × 105 Pa
= (100 atm − 1.0 atm) × × (1.0 × 10−3 m3 )
1 atm
= +1.00... × 104 J = +10 kJ .
For the molar Gibbs energy

∆G ∆G M∆G
∆G m = = =
n m/M ρV
(114.23 g mol−1 ) × (+10.0... kJ)
= = +1.6 kJ mol−1 .
(0.703 g cm−3 ) × (1.0 × 103 cm3 )
P3E.1 (a) The Gibbs-Helmholtz relation for the change in Gibbs energy is given by
[3E.11–104], (∂[∆G/T]/∂T) p = −∆H/T 2 . Integrating the equation be-
tween the temperatures T1 and T2 and assuming that ∆H is temperature
independent gives
T2 ∂ ∆G(T) T2 1
∫ ( ) dT = ∆H ∫ − 2 dT
T1 ∂T T p
T1 T
∆G(T2 ) ∆G(T1 ) 1 1
− = ∆H ( − )
T2 T1 T2 T1
Therefore
∆G(T2 ) ∆G(T1 ) 1 1
= + ∆H ( − )
T2 T1 T2 T1
(b) The standard reaction entropy is given by [3C.3b–90], ∆ r G −○ = ∑J ν J ∆ f G −○ (J),
∆ r G −○ (298 K) = 2∆ f G −○ (CO2 (g)) − 2∆ f G −○ (CO (g)) − ∆ f G −○ (O2 (g))
= 2 × (−394.36 kJ mol−1 ) − 2 × (−137.17 kJ mol−1 ) − 0
= −514.38 kJ mol−1 .
Similarly, the standard reaction enthalpy is given by [2C.5b–53], ∆ r H −○ =
∑J ν J ∆ f H −○ (J).
∆ r H −○ (298 K) = 2∆ f H −○ (CO2 (g)) − 2∆ f H −○ (CO (g)) − ∆ f H −○ (O2 (g))
= 2 × (−393.51 kJ mol−1 ) − 2 × (−110.53 kJ mol−1 ) − 0
= −565.96 kJ mol−1 .
(c) The above derived expression is rearranged to give

T2 T2
∆G(T2 ) = ∆G(T1 ) + ∆H (1 − )
T1 T1
Hence
375 K
∆G(375 K) = (−514.38 kJ mol−1 )
298 K
375 K
+ (−565.96 kJ mol−1 ) (1 − )
298 K
= (−6.47 × 102 kJ mol−1 ) + (+1.46 × 102 kJ mol−1 )
= −501 kJ mol−1 .
P3E.3 The given expression for the reaction Gibbs energy dependence on temperature
is rearranged for ∆G(T2 ) and becomes
T2 T2
∆ r G −○ (T2 ) = ∆ r G −○ (T1 ) + ∆ r H −○ (1 − )
T1 T1
Hence at 37 ○ C = 310 K
273.15 K + 37 K
∆ r G −○ (310 K) = (−6333 kJ mol−1 )
298 K
−1 273.15 K + 37 K
+ (−5797 kJ mol ) (1 − )
298 K
= −6355 kJ mol−1
The extra non-expansion work that is obtained by raising the temperature is

the difference
∆ r G −○ (310 K) − ∆ r G −○ (298 K) = (−6355 kJ mol−1 ) − (−6333 kJ mol−1 )

= −22 kJ mol−1
Therefore the result is an extra 22 kJ mol−1 of energy that is available for non-
expansion work.
P3E.5 Consider the exact differential of the Enthalpy, H = U + pV
dH = dU + d(pV ) = dU + V dp + pdV
The exact differential of the internal energy is given by the fundamental equa-
tion [3E.1–100], dU = TdS − pdV , hence
dH = TdS − pdV + V dp + pdV = TdS + V dp
Because dH is the exact differential this implies
∂H ∂H
( ) =T and ( ) =V
∂S p ∂p S
The mixed partial derivatives are equal irrespective of the order
∂ ∂H ∂ ∂H
( ( ) ) =( ( ) )
∂p ∂S p S ∂S ∂p S p
Therefore
∂T ∂V
( ) =( )
∂p S ∂S p
Similarly consider the exact differentials of the Helmholtz energy, A = U − T S,

and the Gibbs energy, G = H − T S. Starting with the Helmholtz energy
dA = dU − d(T S) = dU − TdS − SdT

= TdS − pdV − TdS − SdT = −pdV − SdT
It follows that
∂p ∂S
( ) =( )
∂T V ∂V T
For the Gibbs energy, the above derived result for dH is used
dG = dH − d(T S) = TdS + V dp − TdS − SdT

= V dp − SdT
It follows that
∂V ∂S
( ) = −( ) .
∂T p ∂p T
P3E.7 (a) Assuming that a = 0 and b ≠ 0, the van der Waals equation becomes
p = RT/(Vm − b). The molar volume is thus
RT
Vm = +b
p
Consider the exact differential of the molar Gibbs energy at constant tem-
perature, dG m = (∂G m /∂p)T dp. Integrating this gives
G m,f pf ∂G m pf p f RT
∫ dG m = ∫ ( ) dp = ∫ Vm dp = ∫ ( + b) dp
G m,i pi ∂p T pi pi p
Therefore
pf
G m (p f ) = G m (p i ) + RT ln ( ) + b(p f − p i )
pi
The change in Gibbs energy energy increases more rapidly with pressure
than the perfect gas due to the last term originating from the repulsion.
(b) Assuming that a ≠ 0 and b = 0, the van der Waals equation becomes
p = RT/Vm ) + a/Vm2 . This is rearranged into a quadratic equation in Vm
pVm2 − RT Vm + a = 0
The solutions for Vm are

√
(−RT)2 − 4pa
−(−RT) ±
Vm =
2p
√
RT RT 4pa
= ± 1− 2 2
2p 2p R T
Because the van der Waals equation is a correction to the ideal gas, the
result should be approximately similar. Considering 4pa/(RT)2 ≪ 1, it
is obvious that only a positive root reproduces the perfect gas and hence
is physically relevant. This solution is used further to apply the suggested

approximate expansion
√
RT RT 4pa
Vm = + 1−
2p 2p R2 T 2
RT RT 4pa RT a
≈ + (1 − 12 [ 2 2 ]) = − .
2p 2p R T p pRT
Integrating this as before gives the Gibbs energy dependence on pressure

G m,f pf pf RT a
∫ dG m = ∫ Vm dp = ∫ − dp
G m,i pi pi p pRT
Therefore
pf a pf
G m (p f ) = G m (p i ) + RT ln ( )− ln ( )
pi RT pi
The change in Gibbs energy energy decreases more rapidly with pressure
than the perfect gas due to the last term originating from the attractive
interaction between the molecules term.
(c) Using the given data the change in molar Gibbs energy, ∆G m (p) = G m (p)−
G m (p−○ ), is plotted against (p/p−○ ) at 298 K using the requested units
(Fig. 3.5). A zoomed version of the same plot is shown in Fig. 3.6.
150.0
∆G m (p)/atm dm3 mol−1
100.0
50.0 perfect gas

vdW, repulsive
vdW, attractive
0.0
0 50 100 150 200 250 300
p/p−○
Figure 3.5

I3.1 (a) The variation of entropy with volume at constant temperature is given by
one of the Maxwell relations from Table 3E.1 on page 101, (∂S/∂V )T =
(∂p/∂T)V . Working with molar quantities, the van der Waals equation
of state is p = RT/(Vm − b) − a/Vm2 , therefore (∂p/∂T)V = R/(Vm − b).
100.0
∆G m (p)/atm dm3 mol−1

95.0
90.0
perfect gas
85.0
vdW, repulsive
vdW, attractive
80.0
25 30 35 40 45 50
−
○
p/p
Figure 3.6
The integration is then straightforward

Vm,f R Vm,f − b
∆S m = ∫ dVm = R ln
Vm,i Vm − b Vm,i − b
= (8.3145 J K−1 mol−1 )
(10.0 dm3 mol−1 ) − (4.29 × 10−2 dm3 mol−1 )
× ln
(1.00 dm3 mol−1 ) − (4.29 × 10−2 dm3 mol−1 )
= +19.5 J K−1 mol−1
where the value of b is taken from the tables in the Resource section; note
the conversion of the molar volumes to dm3 mol−1 so as to match the
units of b.
(b) The variation of entropy with temperature at constant volume and pres-
sure are given by
∆S m = C V ,m ln(T2 /T1 ) and ∆S m = C p,m ln(T2 /T1 )
respectively; both relationships assume that the heat capacities do not

change in the temperature interval.
The equipartition theorem, The chemist’s toolkit 7 in Topic 2A, is used to
estimate the value of C V ,m ; for a perfect gas C p,m = C V ,m + R. For atoms
there are just three translational degrees of freedom therefore C V ,m = 23 R
and C p,m = 52 R. For linear rotors there are in addition two rotational
degrees of freedom, therefore C V ,m = 25 R and C p,m = 72 R. For non-
linear rotors there are three rotational degrees of freedom, C V ,m = 3R
and C p,m = 4R.
Figures 3.7 and 3.8 show plots of ∆S m /R against ln(T2 /T1 ) for the con-
stant volume and constant pressure cases, respectively.
(c) The change in entropy as a function of temperature is given by [3B.6–86];
20
atoms
linear rotors
15 non-linear rotors
∆S m /R
10
0
0 1 2 3 4 5
ln(T2 /T1 )
Figure 3.7
20
atoms
linear rotors
15 non-linear rotors
∆S m /R
10
0
0 1 2 3 4 5
ln(T2 /T1 )
Figure 3.8
this is integrated for the particular form of the heat capacity suggested
CTf Tf a c
∆S = ∫ dT = ∫ ( + b + 3 ) dT
Ti T Ti T T
Tf 1 1
= a ln + b(Tf − Tf ) − 21 c ( 2 − 2 )
Ti ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶ Tf Ti
´¹¹ ¹ ¹ ¸ ¹ ¹ ¹ ¶ term 2 ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶
term 1 term 3
A convenient way of exploring this result is to choose a specific temper-

ature range, say from 273 K to 473 K, and then plot the contribution of
each of the three terms as a function of the relevant parameter, a, b, or c.
Referring to the data in the Resource section it is seen that the ranges of
these parameters are: for a, between 15 J K−1 mol−1 and 80 J K−1 mol−1 ;
for b between 0 and 50 × 10−3 J K−2 mol−1 ; and for c between −10 ×
105 J K mol−1 and +2.0 × 105 J K mol−1 .
Figure 3.9 compares the contributions made by three terms over this tem-
perature range; from the plots it is clear that the first term makes by far
the greatest contribution. Terms 1 and 2 both result in an increase in the
entropy with temperature, but term 3 will make a negative contribution
(term 1)/(J K−1 mol−1 ) to the entropy change if c < 0, which is commonly the case.
(term 2)/(J K−1 mol−1 )

40 40
20 20
0 0
20 40 60 80 0.00 0.02 0.04
−1 −1 −2 −1
a/(J K mol ) b/(J K mol )
(term 3)/(J K−1 mol−1 )
−2
−4
−10 −5 0
−1
c/(10 J K mol )
5
Figure 3.9
(d) The variation of G with p at constant T is given by [3E.8–103], (∂G/∂p)T =

V . The physical significance of the derivative is therefore that it is equal to
the volume of the system. For a perfect gas, V = nRT/p, which makes the
integration straightforward to give ∆G = nRT ln(p f /p i ) ([3E.14–105]).
Figure 3.10 shows a plot of ∆G/nRT as a function of p f /p i . The Gibbs
energy increases with pressure at constant temperature.
(e) The fugacity coefficient is given in terms of the compression factor Z by
p Z−1 pVm
ln ϕ = ∫ dp Z=
0 p RT
The pressure, volume and temperature can be expressed in terms of the
reduced variables p r , Vr , and Tr , given by
p r = p/p c Vr = Vm /Vc Tr = T/Tc
where the critical values of p, V , and T are given in terms of the van der
Waals parameters by
p c = a/27b 2 Vc = 3b Tc = 8a/27bR
2.0
1.5
∆G/nRT
1.0
0.5
0.0
1 2 3 4 5
p f /p i
Figure 3.10
The compression factor can therefore be written

pVm p c Vc p r Vr 1 a 27bR p r Vr 3 p r Vr
Z= = = × 3b × × =
RT RTc Tr R 27b 2 8a Tr 8 Tr
The aim is to write Z in terms of just Vr and Tr , therefore p r is substituted
by
8Tr 3
pr = −
3Vr − 1 Vr2
to give
3 Vr 8Tr 3 3Vr 9
Z= ( − )= −
8 Tr 3Vr − 1 Vr2 3Vr − 1 8Tr Vr
The variable of integration is p, and it is desired to change this to Vr , for
which the following derivative is required
dp dp r d 8Tr 3
= pc = pc ( − 2)
dVr dVr dVr 3Vr − 1 Vr
−24Tr 6
= pc ( + )
(3Vr − 1)2 Vr3
The lower limit of the integral is p = 0, which corresponds to Vr = ∞. The
integral therefore becomes
p Z−1
ln ϕ = ∫ dp
0 p
Vr Z−1 −24Tr 6
=∫ pc ( + ) dVr
∞ pc pr (3Vr − 1)2 Vr3
Vr Z−1 −24Tr 6
=∫ ( + ) dVr
∞ pr (3Vr − 1)2 Vr3
−1
Vr 3Vr 9 8Tr 3 −24Tr 6
=∫ ( − − 1) ( − ) ( + ) dVr
∞ 3Vr − 1 8Tr Vr 3Vr − 1 Vr2 (3Vr − 1)2 Vr3
Mathematical software may be able to evaluate this integral analytically,

or failing that it will be necessary to resort to numerical methods. Some
representative results are shown in Fig. 3.11.
1.2
Tr = 1
Tr = 2
Tr = 3
1.0
ϕ
0.8
0.6
1.0 1.5 2.0 2.5 3.0
Vr
Figure 3.11
I3.3 The statistical definition of entropy is given by [3A.4–77], S = k ln W, where W

is the number of microstates, the number of ways in which the molecules of a
system can be distributed over the energy states for a specified total energy. As
explained in Section 3A.2(a) on page 76 the molecular interpretation of helps
to explain why, in the thermodynamic definition given by [3A.1a–76],dS =
q rev /T, the entropy change depends inversely on the temperature. In a system
at high temperature the molecules are spread out over a large number of energy
states. Increasing the energy of the system by the transfer of heat makes more
states accessible, but given that very many states are already occupied the pro-
portionate change in W is small. In contrast, for a system at a low temperature
fewer states are occupied, and so the transfer of the same energy results in a
proportionately larger increase in the number of accessible states, and hence
a larger increase in W. This argument suggests that the change in entropy for
a given transfer of energy as heat should be greater at low temperatures than
at high, as in the thermodynamic definition. As discussed in Section 3C.2(a)
on page 89, the statistical definition of entropy also justifies the Third Law
of thermodynamics. The law states that the entropy of all perfect crystalline
substances is zero at T = 0. At a molecular level the absence of thermal motion
in a perfectly localized crystalline solid is interpreted as there is only one way
to arrange the molecules like that. Thus, W = 1 and from S = k ln W it follows
that S = 0 as stated by the law.
4 Physical transformations
of pure substances
4A Phase diagrams of pure substances

D4A.1 For two phases to be in equilibrium, the chemical potentials of each component
must be equal in the two phases. In a one-component system, this means that
the chemical potential of that one component must be the same in all phases
that are in equilibrium. The chemical potential is a function of two variables,
say p and T (and not of composition in a one-component system). Thus, if there
are four phases α, β, γ, and δ in equilibrium the chemical potentials would need
to satisfy
µ α (p, T) = µ β (p, T) = µ γ (p, T) = µ δ (p, T)
This is a set of three independent equations in only two variables (p and T),
which are not compatible.
D4A.3 Chemical potential is the single function that governs phase stability. The phase
whose chemical potential is least under a set of given conditions is the most
stable. Conditions under which two or more phases have equal chemical poten-
tials are conditions under which those phases are in equilibrium. Understand-
ing how chemical potential varies with physical conditions such as temperature,
pressure, and composition makes it possible to compute chemical potentials for
various phases and to map out the conditions for stability of those phases and
for equilibrium between them.
E4A.1(a) Use the phase rule [4A.1–116], F = C − P + 2, with C = 1 (one component).
Inserting P = 1 gives F = 1 − 1 + 2 = 2. The condition P = 1 therefore
represents an area . An area has F = 2 because it is possible to vary pressure
and temperature independently (within limits) and stay within the area. P = 1
indicates that a single phase is present, so this result confirms that a single phase
is represented by an area in a phase diagram.
E4A.2(a) (i) 200 K and 2.5 atm lies on the boundary between solid and gas phases.
Two phases , solid and gas, would therefore be present in equilibrium
under these conditions.
114 4 PHYSICAL TRANSFORMATIONS OF PURE SUBSTANCES
(ii) 300 K and 4 atm lies in the vapour region, so only one phase , vapour,
will be present.
(iii) 310 K is greater than the critical temperature, which means that there
is no distinction between gas and liquid. Therefore only one phase (a
supercritical fluid) will be present at all pressures.
E4A.3(a) In a phase diagram, a single phase is represented by an area, while a line repre-
sents a phase boundary where two phases coexist in equilibrium. Point a lies
within an area and therefore only one phase is present. Points b and d each
lie on the boundary between two areas, and therefore in each case two phases
are present. Point c lies at the intersection of three phase boundaries, where
three phases are present in equilibrium.
E4A.4(a) The change in Gibbs energy when an infinitesimal amount dn of substance is

moved from location 1 to location 2 is given by (Section 4A.1(c) on page 113)
dG = (µ 2 − µ 1 )dn
Assuming that 0.1 mmol is a sufficiently small amount to be regarded as in-

finitesimal, the Gibbs energy change in this case is
∆G = (µ 2 − µ 1 )∆n = (7.1 × 103 J mol−1 ) × (0.1 × 10−3 mol) = 0.71 J
E4A.5(a) Use the phase rule [4A.1–116], F = C − P + 2, with C = 2 (for two components).
Rearranging for the number of phases gives
P = C−F +2=2−F +2=4−F
The number of variables that can be changed arbitrarily, F, cannot be smaller

than zero so the maximum number of phases in this case is 4 .
P4A.1 (a) 100 K and 1 atm lies in the solid region of the phase diagram, so initially
only solid carbon dioxide (dry ice) will be present. When the temper-
ature reaches 194.7 K, the sublimation point of CO2 at 1 atm, solid and
gas phases will be present in equilibrium. Above this temperature only
gaseous CO2 is present.
(b) 100 K and 70 atm lies in the solid region of the phase diagram, so again
CO2 will initially be a solid. On heating, a point is reached at which the
solid melts; at this temperature solid and liquid phases are both present in
equilibrium. Above this temperature only a liquid phase is present until
the boiling temperature is reached, at which point liquid and gas will be in
equilibrium. Above this temperature, only the gas phase will be present.
P4A.3 A schematic phase diagram is shown in Fig 4.1. Note that in reality the phase
boundaries may be curved rather than straight. There are two triple points
which are marked with dots.

δ
s Triple points

β
p s (((
(
s
α
γ

T
Figure 4.1
4B Thermodynamic aspects of phase transitions

D4B.1 Formally, the pressure derivative of the chemical potential is (∂µ/∂p)T = Vm .
Because the molar volume is always positive, the slope of the change in chemical
potential with respect to change in pressure is positive: that is, the chemical
potential increases with increasing pressure.
D4B.3 Formally, the temperature derivative of the chemical potential is (∂µ/∂T) p =

−S m . Because the molar entropy is always positive for all pure substances, the
slope of the change in chemical potential with respect to change in temperature
is negative: that is, the chemical potential decreases with increasing tempera-
ture.
E4B.1(a) The relationship between pressure and temperature along the solid–liquid bound-
ary is given by [4B.7–124], p = p∗ + (∆ fus H/T ∗ ∆ fus V )(T − T ∗ ). In this case
p∗ = 1 atm (corresponding to the normal melting point, T ∗ = 273.15 K) and
p = 1 bar (corresponding to the standard melting point). Rearranging for
(T − T ∗ ), the difference in melting points, gives
T ∗ ∆ fus V
(T − T ∗ ) = (p − p∗ )
∆ fus H
1.01325 × 105 Pa
= (1 × 105 Pa − 1 atm × )
1 atm
(273.15 K) × (−1.6 × 10−6 m3 mol−1 )
× = 9.6 × 10−5 K
6.008 × 103 J mol−1
This result shows that the standard melting point of ice is slightly higher than
the normal melting point, but the difference is negligibly small for most pur-
poses.
E4B.2(a) Since 1 W = 1 J s−1 , the rate at which energy is absorbed is (1.2 kW m−2 ) ×
(50 m2 ) = 60 kJ s−1 . The rate of vaporization is then
rate of energy absorption 60 kJ s−1
= = 1.36... mol s−1
∆ vap H 44 kJ mol−1
Multiplication by the molar mass of water gives the rate of loss of water as
(1.36... mol s−1 ) × (18.0158 g mol−1 ) = 25 g s−1 .
E4B.3(a) The perfect gas equation [1A.4–8], pV = nRT, is used to calculate the amount
as n = pV /RT. V is the volume of the laboratory (75 m3 ) and p is the vapour
pressure. The mass is found from m = nM, where M is the molar mass; hence
m = pV M/RT
pV M (3.2 × 103 Pa) × (75 m3 ) × (18.0158 g mol−1 )
Water: m= = = 1.7 kg
RT (8.3145 J K−1 mol−1 ) × ([25 + 273.15] K)
pV M (13.1 × 103 Pa)×(75 m3 )×(78.1074 g mol−1 )
Benzene: m= = = 31 kg
RT (8.3145 J K−1 mol−1 ) × ([25 + 273.15] K)
pV M (0.23 Pa) × (75 m3 ) × (200.59 g mol−1 )
Mercury: m= = = 1.4 g
RT (8.3145 J K−1 mol−1 ) × ([25 + 273.15] K)
1 Pa = 1 kg m−1 s−2 and 1 J = 1 kg m2 s−2 have been used. Note that an typically
sized bottle of benzene (containing less than 31 kg of benzene) would evaporate
completely before saturating the air of the laboratory with benzene vapour.
E4B.4(a) (i) The integrated form of the Clausius–Clapeyron equation [4B.10–125] is
p ∆ vap H 1 1
ln ∗
=− ( − ∗)
p R T T
Rearranging for ∆ vap H and substituting in the numbers, taking p∗ , T ∗ at
85.8 ○ C and p,T at 119.3 ○ C, gives
1 1 −1 p
∆ vap H = −R ( − ∗ ) ln ∗
T T p
−1
−1 −1 1 1
= −(8.3145 J K mol ) × ( − )
[119.3 + 273.15] K [85.5 + 273.15] K
5.3 kPa
× ln ( ) = 4.86... × 104 J mol−1 = 49 kJ mol−1
1.3 kPa
(ii) The integrated form of the Clausius–Clapeyron equation is now rearranged
for T. Substituting in p = 1 atm, or 1.01325 × 105 Pa, corresponding to
the normal boiling point, together with the value of ∆ vap H from above
and the same values for p∗ , T ∗ as before, gives
−1
1 R p
T=( ∗
− ln ∗ )
T ∆ vap H p
−1
1 8.3145 J K−1 mol−1 1.01325 × 105 Pa
=( − × ln )
[85.5 + 273.15] K 4.86... × 104 1.3 × 103 Pa
= 4.89... × 102 K = 4.9 × 102 K or 2.2 × 102 ○ C
(iii) To find ∆ vap S at the boiling temperature, use [3B.4–85]:

∆ vap H 4.86... × 104 J mol−1
∆ vap S = = = 99 J K−1 mol−1
T 4.89... × 102 K
ary is given by [4B.7–124], p = p∗ + (∆ fus H/T ∗ ∆ fus V )(T − T ∗ ). The molar
volume is Vm = M/ρ where M is the molar mass and ρ is the mass density
∆ fus V = Vm (l) − Vm (s) = M/ρ(l) − M/ρ(s)
This expression is inserted into [4B.7–124], which is then rearranged for T. T ∗ ,
p∗ , and ∆ vap H are taken as the values corresponding to the normal melting
point of ice, that is, 0 ○ C (273.15 K) and 1 atm (101.325 kPa). It is assumed that
∆ vap H is constant over the temperature range of interest.
T∗ M M
T = T ∗ + (p − p∗ ) ( − )
∆ fus H ρ(l) ρ(s)
273.15 K
= (273.15 K) + ([50 × 105 − 1.01325 × 105 ] Pa) ×
6.008 × 103 J mol−1
18.0158 g mol−1 18.0158 g mol−1
×( − ) = 273 K or −0.35 ○ C
1.00 × 106 g m−3 0.92 × 106 g m−3
E4B.6(a) The variation of chemical potential with temperature is given by [4B.1a–120],

(∂µ/∂T) p = −S m . For a finite change this gives ∆µ = −S m ∆T.
∆µ(liquid) = −(65 J K−1 mol−1 ) × (1 K) = −65 J mol−1

∆µ(solid) = −(43 J K−1 mol−1 ) × (1 K) = −43 J mol−1
The chemical potentials of both solid and liquid are decreased at the higher
temperature, but the chemical potential of the liquid is decreased by a greater
amount. As they were at equilibrium before it follows that the liquid is the
more stable phase at the higher temperature, so melting will be spontaneous.
E4B.7(a) The variation of chemical potential with temperature is given by [4B.1a–120],
(∂µ/∂T) p = −S m . For a finite change this gives ∆µ = −S m ∆T, assuming that
S m is constant over the temperature range.
∆µ = −(69.9 J K−1 mol−1 ) × ([35 − 25] K) = −699 J mol−1
E4B.8(a) The variation of chemical potential with pressure is given by [4B.1b–120], (∂µ/∂p)T =
Vm . For a finite change, and assuming that Vm is constant over the pressure
range, this gives ∆µ = Vm ∆p. The molar volume Vm is given by M/ρ where M
is the molar mass of copper and ρ is the mass density.
∆µ = Vm ∆p = (M/ρ)∆p
63.55 × 10−3 kg mol−1
= × ([10 × 106 − 100 × 103 ] Pa) = +70 J mol−1
8960 kg m−3
Note that 1 Pa m3 = 1 J.
E4B.9(a) The variation of vapour pressure with applied pressure is given by [4B.2–122],
p = p∗ eVm (l)∆ p/RT .
(18.1 × 10−6 m3 mol−1 )×([20 × 106 − 1 × 105 ] Pa)

p = (2.34 × 103 Pa)×exp ( )
(8.3145 J K−1 mol−1 ) × ([20 + 273.15] K)
= 2710 Pa = 2.71 kPa
ary is given by [4B.7–124], p = p∗ + (∆ fus H/T ∗ ∆ fus V )(T − T ∗ ), which is
rearranged to give ∆ fus H. In this case p∗ = 1.00 atm, T ∗ = 350.75 K, p =
100 atm and T = 351.26 K.
p − p∗ ∗
∆ fus H = T ∆ fus V
T − T∗
([100 − 1] atm) × (1.01325 × 105 Pa/1 atm)
= × (350.75 K)
([351.26 − 350.75] K)
× ([163.3 − 161.0] × 10−6 m3 mol−1 )
= 1.58... × 104 J mol−1 = 15.9 kJ mol−1
The entropy of transition is given by [3B.4–85], ∆ fus S = ∆ fus H/T, where T is

the transition temperature. At the melting temperature the entropy of fusion is
1.58... × 104 J mol−1

∆ fus S = = 45.2 J K−1 mol−1
350.75 K
E4B.11(a) The integrated version of the Clausius–Clapeyron equation [4B.10–125] is given

by ln(p/p∗ ) = −(∆ vap H/R)(1/T − 1/T ∗ ). Rearranging for T gives
−1
1 R p
T =( − ln )
T ∗ ∆ vap H p∗
−1
1 8.3145 J K−1 mol−1 70.0 kPa
=( − × ln )
[24.1 + 273.15] K 28.7 × 103 J mol−1 53.3 kPa
= 304 K or 31.2 ○ C
E4B.12(a) The Clausius–Clapeyron equation [4B.9–125] is d ln p/dT = ∆ vap H/RT 2 . This

equation is rearranged for ∆ vap H, and the expression for ln p is differentiated.
It does not matter that the pressure is given in units of Torr because only the
slope of ln p is required.
d ln p d 2501.8 K 2501.8 K
∆ vap H = RT 2 = RT 2 (16.255 − ) = RT 2 ( )
dT dT T T2
= (2501.8 K)R = (2501.8 K) × (8.3145 J K−1 mol−1 ) = 20.801 kJ mol−1
E4B.13(a) (i) The Clausius–Clapeyron equation [4B.9–125] is d ln p/dT = ∆ vap H/RT 2 .

This equation is rearranged for ∆ vap H and the expression for ln p is dif-
ferentiated, noting from inside the front cover that ln x = (ln 10) log x. It
does not matter that the pressure is given in units of Torr because only
the slope of ln p is required.
d ln p d log p d 1780 K
∆ vap H = RT 2 = RT 2 ln 10 = RT 2 ln 10 (7.960 − )
dT dT dT T
1780 K
= RT 2 ln 10 ( ) = (1780 K)R ln 10
T2
= (1780 K) × (8.3145 J K−1 mol−1 ) × ln 10 = 34.08 kJ mol−1
(ii) The normal boiling point refers to the temperature at which the vapour
pressure is 1 atm which is 760 Torr. The given expression, log(p/Torr) =
7.960 − (1780K)/T, is rearranged for T and a pressure of 760 Torr is
substituted into it to give
1780 K 1780 K
T= = = 350.4 K or 77.30 ○ C
7.960 − log(p/Torr) 7.960 − log 760
Note that this temperature lies outside the range 10 ○ C to 30 ○ C for which
the expression for log(p/Torr) is known to be valid, and is therefore an
estimate.
ary is given by [4B.7–124], p = p∗ + (∆ fus H/T ∗ ∆ fus V )(T − T ∗ ). The value of
∆ fus V is found by using Vm = M/ρ where M is the molar mass and ρ is the
mass density:
M M
∆ fus V = Vm (l) − Vm (s) = −
ρ(l) ρ(s)
78.1074 g mol−1 78.1074 g mol−1
= − = 1.19... × 10−6 m3 mol−1
0.879 × 106 g m−3 0.891 × 106 g m−3
Equation [4B.7–124] is then rearranged to find T:
T ∗ ∆ fus V
T = T ∗ + (p − p∗ )
∆ fus H
1.01325 × 105 Pa
= ([5.5 + 273.15] K) + (([1000 − 1] atm) × )
1 atm
([5.5 + 273.15] K) × (1.19... × 10−6 m3 )
× = 2.8 × 102 K or 8.7 ○ C
10.59 × 103 J mol−1
P4B.1 The work done in expanding against a constant external pressure is given by
equation [2A.6–38], w = −p ex ∆V . Because the molar volume of a gas is so
much greater the molar volume of a liquid, ∆ vap V ≈ Vm (g). In addition, if
the gas behaves perfectly, Vm = RT/p (from the perfect gas law, [1A.4–8]) with
p = p ex as the gas expands against constant external pressure. The work of
expansion is therefore
RT
w = −p ex × = −RT = −(8.3145 J K−1 mol−1 ) × ([100 + 273.15] K)
p ex
= −3.10... × 103 J mol−1 = −3.10 kJ mol−1
The negative sign indicates that the system has done work on the surround-
ings, so the internal energy of the system falls. The fraction of the enthalpy of
vaporization spent on expanding the vapour is
3.10... kJ mol−1
× 100 % = 7.62 %
40.7 kJ mol−1
P4B.3 The variation of vapour pressure with temperature is given by [4B.10–125], p =

p∗ exp[(−∆ vap H/R)(1/T − 1/T ∗ )]. The values of T ∗ and p∗ corresponding to
the normal boiling point are used
∆ vap H 1 1
p = p∗ exp (− ( − ∗ ))
R T T
= (1 atm)
20.25 × 103 J mol−1 1 1
× exp (− −1 −1 × ( − ))
8.3145 J K mol (40 + 273.15) K (−29.2 + 273.15) K
= 9.08 atm or 920 kPa
P4B.5 (a) From the variation of chemical potential with temperature (at constant
pressure) [4B.1a–120], (∂µ/∂T) p = −S m , the slope of the chemical po-
tential against temperature is equal to the negative of the molar entropy.
The difference in slope on either side of the normal freezing point of water
is therefore
∂µ(l) ∂µ(s)
( ) −( ) = −S m (l) − (−S m (s))
∂T p ∂T p
= −∆ fus S = −22.0 J K−1 mol−1
(b) In a similar way, the difference in slope on either side of the normal boil-
ing point of water is
∂µ(g) ∂µ(l)
( ) −( ) = −S m (g) − (−S m (l))
∂T p ∂T p
= −∆ vap S = −109.9 J K−1 mol−1
(c) From part (a)
∂µ(l) ∂µ(s) ∂[µ(l) − µ(s)]

( ) −( ) = −∆ fus S hence ( ) = −∆ fus S
∂T p ∂T p ∂T p
For a finite change ∆[µ(l) − µ(s)] = −∆ fus S × ∆T. For a 5 ○ C drop in

temperature:
∆[µ(l) − µ(s)] = −(22.0 J K−1 mol−1 ) × (−5 K) = +110 J mol−1
Therefore, since water and ice are in equilibrium (µ(l) − µ(s) = 0) at

0 ○ C it follows that the chemical potential of liquid water exceeds that of
ice by +110 J mol−1 at −5 ○ C. The fact that µ(l) > µ(s) indicates that
supercooled water at −5, ○ C has a tendency to freeze to ice.
P4B.7 The total pressure at the bottom of the column is
p = ρgd + 1 atm
10−3 kg 106 cm3
= (13.6 g cm−3 × × ) × (9.81 m s−2 ) × (10 m)
1g 1 m3
+ (1.01325 × 105 Pa) = 1.44... × 106 Pa
To find the freezing point, use [4B.7–124], p = p∗ +(∆ fus H/T ∗ ∆ fus V )(T −T ∗ ).
Rearranging for T gives
T ∗ ∆ fus V
T = T∗ + (p − p∗ )
∆ fus H
(234.3 K) × (0.517 × 10−6 m3 mol−1 )
= (234.3 K) +
2.292 × 103 J mol−1
× ([1.44... × 10 − 1.01325 × 105 ] Pa) = 234.4 K
6
Note that this is not a very large difference from the normal freezing point,
reflecting the fact that the slope of the solid-liquid boundary is generally very
steep compared to the liquid-vapour boundary. Large changes in pressure are
therefore needed to bring about significant changes in freezing point.
P4B.9 The integrated form of the Clausius–Clapeyron equation [4B.10–125], ln(p/p∗ ) =

−(∆ vap H/R)(1/T − 1/T ∗ ), is rewritten
p ∆ vap H 1 ∆ vap H
ln ∗
=− +
p R T RT ∗
This implies that a plot of ln(p/p∗ ) against 1/T should be a straight line of slope
−∆ vap H/R and intercept ∆ vap H/RT ∗ ; such a plot is shown in Fig. 4.2. If p∗ is
taken to be 1 atm, or 101.325 kPa, then T ∗ corresponds to the normal boiling
point which can then be obtained from the intercept.
θ/○ C p/kPa T −1 /K−1 ln(p/p∗ )

0 1.92 0.003 66 −3.966
20 6.38 0.003 41 −2.765
40 17.70 0.003 19 −1.745
50 27.70 0.003 09 −1.297
70 62.30 0.002 91 −0.486
80 89.30 0.002 83 −0.126
90 124.90 0.002 75 0.209
100 170.90 0.002 68 0.523
0
ln(p/p∗ )
−2
−4
0.0025 0.0030 0.0035
T −1 /K−1
Figure 4.2
The data fall on a good straight line, the equation of which is
ln(p/p∗ ) = (−4.570 × 103 ) × (T −1 /K−1 ) + 12.81
The values of ∆ vap H and T ∗ are obtained from the slope and intercept respec-
tively:
∆ vap H = −slope × R = −(−4.570 × 103 K) × (8.3145 J K−1 mol−1 )

= 3.79... × 104 J mol−1 = 38.0 kJ mol−1
∆ vap H 3.79... × 104 J mol−1
T∗ = = = 357 K or 84 ○ C
R × intercept (8.3145 J K−1 mol−1 ) × 12.81
P4B.11 (a) The Clapeyron equation [4B.4a–123] is dp/dT = ∆ trs S/∆ trs V . For subli-
mation, and with ∆ trs S = ∆ trs H/T this becomes
dp ∆ sub H
=
dT T∆ sub V
Since the molar volume of a gas is much greater than that of a solid, ∆ sub V
can be approximated as ∆ sub V = Vm (g) − Vm (s) ≈ Vm (g), and if the gas
behaves perfectly, Vm = RT/p. Substituting these into the above equation

gives
dp ∆ sub H p∆ sub H
= =
dT T(RT/p) RT 2
Using dx/x = d ln x this becomes written as d ln p/dT = ∆ sub H/RT 2

(b) Integration of the equation derived in (a) under the assumption that ∆ vap H
is independent of T gives
∆ sub H 1
ln p = − + constant
R T
This implies that a plot of ln p against 1/T should be a straight line of
slope −∆ sub H/R; such a plot is shown in Fig. 4.3.
T/K p/Pa T −1 /K−1 ln(p/Pa)

145.94 13.07 0.006 852 2.570
147.96 18.49 0.006 759 2.917
149.93 25.99 0.006 670 3.258
151.94 36.76 0.006 582 3.604
153.97 50.86 0.006 495 3.929
154.94 59.56 0.006 454 4.087
4.5
4.0
ln(p/Pa)
3.5
3.0
2.5
0.0064 0.0066 0.0068

−1 −1
T /K
Figure 4.3
ln(p/Pa) = (−3.816 × 103 ) × (T −1 /K−1 ) + 28.71
The slope is equal to −∆ vap H/R, so:
∆ vap H = −slope × R = −(−3.816 × 103 K) × (8.3145 J K−1 mol−1 )

= 31.7 kJ mol−1
P4B.13 (a) If the mass of the liquid decreases by m, then the amount in moles of
vapour formed is n vap = m/M. The amount in moles of the input gas
is given by n gas = PV /RT (from the perfect gas equation) so the mole
fraction of the vapour is
n vap n vap m/M mRT
x vap = = = =
n tot n vap + n gas PV /RT + m/M MPV + mRT
(b) If the total pressure remains at P, the partial pressure of the vapour is
mRT mRTP
p = x vap × P = ×P =
MPV + mRT MPV + mRT
(c) Dividing top and bottom of this expression by MPV gives
AmP RT
p= where A=
1 + Am MPV
(d) For geraniol, noting that P = 760 Torr = 1.01325 × 105 Pa,
(8.3145 J K−1 mol−1 ) × ([110 + 273.15] K)

A= = 0.0407... g−1
(154.2 g mol−1 )×(1.01325 × 105 Pa)×(5.00 × 10−3 m3 )
hence
AmP (0.0407... g−1 ) × (0.32 g) × (1.01325 × 105 Pa)
p= = = 1.31 kPa
1 + Am 1 + (0.0407... g−1 ) × (0.32 g)
P4B.15 The integrated form of the Clausius–Claypeyron equation [4B.10–125] is

p ∆ vap H 1 1
ln ∗
=− ( − ∗)
p R T T
Taking p∗ and T ∗ as corresponding to the pressure and boiling point at sea
level, p 0 and T0 , and inserting p/p 0 = e−a/H from the barometric formula gives
−1
∆ vap H 1 1 1 R a
ln (e−a/H ) = − ( − ) hence T =( + )
R T T0 T0 ∆ vap H H
For water at 3 km, a = 3000 m, the boiling point is

−1
1 8.3145 J K−1 mol−1 3 km
T=( + × ) = 363 K or 89.6 ○ C
373.15 K 40.7 × 103 J mol−1 8 km
Solutions to integrated activities

I4.1 (a) The expressions are plotted on the graph shown in Fig. 4.4. Note that the
liquid-vapour line is only plotted for T3 ≤ T ≤ Tc because the liquid phase
does not exist below the triple point and there is no distinction between
liquid and vapour above the critical point. The solid-liquid line is plotted
for T ≥ T3 .
50
40
30
p/bar
Solid Liquid
20
10
Vapour
0
0 100 200 300 400 500 600 700
T/K
Figure 4.4
(b) The standard melting point is the temperature corresponding to a pres-

sure of 1 bar on the solid-liquid boundary. Setting p = 1 bar in the equa-
tion for the solid-liquid boundary and substituting in the value of p 3
gives:
1 = 0.4362 × 10−6 + 1000(5.60 + 11.727x)x
This equation is rearranged to the standard quadratic form
11727x 2 + 5600x − 0.9999995638 = 0
which on solving for x gives x = 1.78... × 10−4 or x = −0.477.... Then,
since x = T/T3 − 1 where T3 = 178.15 K, it follows that
T = 178.15(1.78... × 10−4 + 1) = 178.18 K
or T = 178.15(−0.477... + 1) = 93.11 K
The 93.11 K solution is rejected since it lies below T3 where the liquid, and
therefore the solid-liquid boundary, does not exist. The standard melting
point is therefore estimated to be 178.18 K .
(c) The standard boiling point is the temperature at the point on the liquid-
vapour phase boundary corresponding to p = 1 bar. Substituting this
value of p into the equation for the liquid-vapour boundary and noting
that ln 1 = 0 gives
0 = −10.418/y+21.157−15.996y+14.015y 2 −5.0120y 3 +4.7334(1− y)1.70
Solving numerically gives
y = 0.645... and so T = y × Tc = 0.645... × 593.95 = 383.54 K
(d) Use the Clapeyron equation for the liquid-vapour boundary [4B.8–125]:
dp ∆ vap H dp
= which rearranges to ∆ vap H = T∆ vap V ×
dT T∆ vap V dT
To find dp/dT, use d ln x = dx/x so that dp/dT = p × d ln p/dT. The

expression for ln p is inserted and differentiated, and then evaluated at
the standard boiling point found above. For the evaluation of d ln p/dT it
does not matter that the expression has p in bar not Pa because the slope
of ln p is independent of the units of p due to the logarithm.
dp d ln p d ln p dy
=p = p× ×
dT dT dy dT
d 10.418
= p× (− + 21.157 − 15.996y + 14.015y 2 − 5.0120y 3
dy y
d T
+ 4.7334(1 − y)1.70 ) × ( )
dT Tc
p 10.418
= ×( − 15.996 + 28.030y − 15.0360y 2
Tc y2
− 8.04678(1 − y)0.70 )
105 Pa 10.418
= (
593.95 K (0.645...)2
− 15.996 + 28.030(0.645...) − 15.0360(0.645...)2
− 8.04678(1 − 0.645...)0.70 ) = 2.84... × 103 Pa K−1
Then
dp
∆ vap H = T∆ vap V ×
dT
= (383.54 K) × ([30.3 − 0.12] × 10−3 m3 mol−1 ) × (2.84... × 103 Pa K−1 )
= 33.0 kJ mol−1
I4.3 (a) The data are plotted in Fig. 4.5.

These data fit well to the cubic
p/MPa = (4.989×10−6 K−3 )T 3 −(1.452×10−3 K−2 )T 2 +(0.1461 K−1 )T−5.058
This equation is used to plot the line on the graph.
(b) The standard boiling point corresponds to the temperature at which p =
1 bar or 0.1 MPa. This value is substituted into the fitted function to give
0.1 = (4.989×10−6 K−3 )T 3 −(1.452×10−3 K−2 )T 2 +(0.1461 K−1 )T−5.058
which, on solving numerically using mathematical software, yields
T = 1.11... × 10−2 K = 112 K
(c) The Clapeyron equation for the liquid-vapour boundary is [4B.8–125]:

dp ∆ vap H dp
= hence ∆ vap H = T∆ vap V ×
dT T∆ vap V dT
4.0
p/MPa
2.0
0.0
100 120 140 160 180 200
T/K
Figure 4.5
The value of dp/dT is obtained by differentiating the fitted function and

substituting in the value of T found above.
d(p/MPa)
= (14.967 × 10−3 K−3 )T 2 − (2.904 × 10−3 K−2 )T + (0.1461 K−1 )
dT
= (14.967 × 10−3 K−3 ) × (1.11... × 102 K)2
− (2.904 × 10−3 K−2 ) × (1.11... × 102 K) + (0.1461 K−1 )
= 8.57... × 10−3 K−1
so that
dp
= 8.57... × 10−3 MPa K−1 = 8.57... × 103 Pa K−1
dT
Therefore
dp
∆ vap H = T∆ vap V ×
dT
10−3 m3
= (1.11... × 102 K) × ([8.89 − 3.80 × 10−2 ] dm3 mol−1 ) ×
1 dm3
× (8.57 × 103 Pa K−1 ) = 8.49 kJ mol−1
I4.5 The relationship between p and T along the solid-liquid boundary is given by
equation [4B.7–124]:
∆ fus H
p = p∗ + (T − T ∗ )
T ∗ ∆ fus V
Using Vm = M/ρ, ∆ fus V is calculated as
M M 78.1074 g mol−1 78.1074 g mol−1

∆ fus V = Vm (l) − Vm (s) = − = −
ρ(l) ρ(s) 0.879 × 106 g m−3 0.891 × 106 g m−3
= 1.19... × 10−6 m3 mol−1
This is used in equation [4B.7–124] together with the value of ∆ fus H quoted.
Taking p∗ and T ∗ as corresponding to the triple point, p∗ = 36 Torr = 4.80 kPa
and T ∗ = 5.50 ○ C = 278.65 K gives the equation of the solid-liquid boundary
as
10.6 × 103 J mol−1

p = (4.80 × 103 Pa) + (T − 278.65 K)
(278.65 K) × (1.19... × 10−6 m3 mol−1 )
= (4.80 × 103 Pa) + (3.18 × 107 Pa K−1 ) × (T − 278.65 K)
so that (p/kPa) = 4.80 + (3.18 × 104 ) × [(T/K) − 278.65]
This takes the form of a steep straight line with a positive gradient extending
upwards from the triple point. This is plotted in Fig. 4.6. The line is only drawn
for T ≥ T ∗ (p ≥ p∗ ) because the liquid does not exist below the triple point.
The relationship between p and T along the liquid-vapour boundary is given by
equation [4B.10–125]; p∗ and T ∗ are again taken as corresponding to the triple
point.
∆ vap H 1 1
p = p∗ exp (− ( − ∗ ))
R T T
30.8 × 103 J mol−1 1 1
= (4.80 × 103 Pa) × exp (− ( − ))
8.3145 J K−1 mol−1 T 278.65 K
1 1
or (p/kPa) = 4.80 × exp [−3.70 × 103 ( − )]
T/K 278.65
This equation is also plotted in Fig. 4.6, again only for values of T in the range
T ≥ 278.65 K since the liquid does not exist below this temperature.
The relationship between p and T along the solid-vapour boundary is given by
an equation that is analogous to the liquid-vapour one except that ∆ vap H is
replaced by ∆ sub H. ∆ sub H = ∆ fus H + ∆ vap H so the required equation is
∆ sub H 1 1
p = p∗ exp (− ( − ∗ ))
R T T
[10.6 + 30.8] × 103 J mol−1 1 1
= (4.80 × 103 Pa) × exp (− ( − ))
8.3145 J K−1 mol−1 T 278.65 K
1 1
or (p/kPa) = 4.80 × exp [−4.98 × 103 ( − )]
T/K 278.65
This equation is plotted Fig. 4.6 for values in the range T ≤ 278.65 K since the
solid and vapour phases are only in equilibrium at the triple point and below.
25
20
Liquid
15 Solid
p/kPa
10
5 Vapour
0
250 260 270 280 290 300 310 320
T/K
Figure 4.6
5 Simple mixtures
5A The thermodynamic description of mixtures

D5A.1 Perfect gases spontaneously mix in all proportions. There are, however, con-
ceivable circumstances under which two real gases might not mix spontaneously.
Consider allowing two gases initially at the same pressure p to mix (so that
mixing them would not change the pressure) under conditions of constant tem-
perature. Mixing is spontaneous if ∆ mix G < 0, and this Gibbs energy change
has an entropic and an enthalpic contribution
∆ mix G = ∆ mix H − T∆ mix S
The entropy change, ∆ mix S, is always positive, so mixing is always favoured

entropically. The only circumstances under which mixing might not be sponta-
neous would be if ∆ mix H > T∆ mix S, that is if the change in enthalpy on mixing
was so unfavourable as to outweigh the entropic term.
For perfect gases, ∆ mix H = 0, so mixing always occurs. However, there are
liquids for which unfavourable interactions prevent mixing at least in some pro-
portions and at some temperatures. If two such species were taken above their
critical temperatures and held at a pressure high enough to make their densities
more typical of liquids than gases, then it is possible to imagine that mixing
might not occur. Because the temperature is above the critical temperatures
the species are technically gases, although the term supercritical fluid might
be more appropriate. In conclusion, there might be examples of immiscibility
among supercritical fluids.
D5A.3 Raoult’s law, [5A.22–141] defines the behaviour of ideal solutions. Like perfect
gases, what makes the behaviour ideal can be expressed in terms of intermolec-
ular interactions. Unlike perfect gases, however, the interactions in an ideal
solution cannot be neglected. Instead, ideal behaviour amounts to having the
same interactions between molecules of the different components of the mix-
ture as there are between molecules of the same type.
In short, ideal behaviour consists of A–B interactions being the same as A–A
and B–B interactions. If that is the case, then the cohesive forces that would
keep a molecule in the liquid phase would be the same in the solution as in a
pure liquid, and the vapour pressure of a component will differ from that of a
pure liquid only in proportion to its abundance (mole fraction). Thus, Raoult’s
132 5 SIMPLE MIXTURES
law is expected to be valid for mixtures of components that have very simi-
lar chemical structures. Similar structures imply both similar intermolecular
interactions and similar sizes.
In an ideal dilute solution, on the other hand, Raoult’s law holds for the solvent
in the limit as x A approaches 1, not because A–B interactions are like A–A
interactions, but because there are so many more A–A interactions than A–
B interactions that A–A interactions dominate the behaviour of the solvent.
For the solute, on the other hand, there are many more A–B interactions than
B–B interactions in the limit as x B approaches zero. Thus, only one kind of
interaction (A–B) is important in determining the affinity of the solute for the
solution.
D5A.5 The change in Gibbs energy at constant temperature is equal to the maximum
additional (non-expansion) work that the system can do dG = dw add,max , [3D.8–
96]. Changing the composition of a mixture gives rise to a change in Gibbs
energy, given by [5A.7–135], dG = µ A dn A + µ B dn B . . .. It therefore follows that
dw add,max = µ A dn A + µ B dn B . . .
and so non-expansion work can arise from the changing composition of a sys-
tem.
E5A.1(a) The partial pressure of gas A, p A above a liquid mixture is given by Raoult’s
Law, [5A.22–141], p A = x A p∗A , where x A is the mole fraction of A in the liquid
and p∗A is the vapour pressure over pure A. The total pressure over a mixture
of A and B is p A + p B .
The first step is to calculate the mole fractions. If the molar mass of A is M A
and the mass of A is m, then the amount in moles of A is m/M A , and likewise
because the mass of B is the same, the amount of B is m/M B . The mole fraction
of A is therefore
m/M A 1/M A 1/M B

xA = = likewise x B =
m/M A + m/M B 1/M A + 1/M B 1/M A + 1/M B
These mole fractions are used with Raoult’s law to give the total pressure
1/M A 1/M B
p = x A p∗A + x B p∗B = p∗ + p∗
1/M A + 1/M B A 1/M A + 1/M B B
If A is benzene, M A = 6×12.01 g mol−1 +6×1.0079 g mol−1 = 78.1074 g mol−1 ,

and if B is methylbenzene M B = 7 × 12.01 g mol−1 + 8 × 1.0079 g mol−1 =
92.1332 g mol−1 .
1/M A 1/M B
p= p∗ + p∗
1/M A + 1/M B A 1/M A + 1/M B B
1/(78.1074 g mol−1 )
= × (10 kPa)
1/(78.1074 g mol−1 ) + 1/(92.1332 g mol−1 )
1/(92.1332 g mol−1 )
+ × (2.8 kPa)
1/(78.1074 g mol−1 ) + 1/(92.1332 g mol−1 )
= 5.41... kPa + 1.28... kPa = 6.7 kPa
E5A.2(a) The total volume is calculated from the partial molar volumes of the two com-
ponents using [5A.3–134], V = n A VA + n B VB . The task is therefore to find the
amount in moles, n A and n B , of A and B in a given mass m of solution. If the
molar masses of A and B are M A and M B then it follows that
m = nA MA + nB MB
The mole fraction of A is defined as x A = n A /(n A + n B ), hence n A = x A (n A +

n B ) and likewise for B. With these substitutions for n A and n B the previous
equation becomes
m
m = x A M A (n A + n B ) + x B M B (n A + n B ) hence (n A + n B ) =
xA MA + xB MB
This latter expression for the total amount in moles, (n A + n B ), is used with
n A = x A (n A + n B ) to give
mx A
n A = x A (n A + n B ) =
xA MA + xB MB
and likewise
mx B
nB =
xA MA + xB MB
With these expressions for n A and n B the total volume is computed from the
partial molar volumes
mx A VA mx B VB
V = n A VA + n B VB = +
xA MA + xB MB xA MA + xB MB
m
= [x A VA + x B VB ]
xA MA + xB MB
m
= [x A VA + (1 − x A )VB ]
x A M A + (1 − x A )M B
where on the last line x B = (1 − x A ) is used.

Taking A as trichloromethane and B as propanone the molar masses are M A =
12.01+1.0079+3×35.45 = 119.3679 g mol−1 and M B = 3×12.01+6×1.0079+
16.00 = 58.0774 g mol−1 . With these values, the expression for the volume of
1.000 kg evaluates as
1000 g
V=
0.4693 × (119.3679 g mol ) + (1 − 0.4693) × (58.0774 g mol−1 )
−1
−1 −1
× [0.4693 × (80.235 cm3 mol ) + (1 − 0.4693) × (74.166 cm3 mol )]
= 886.8 cm3
E5A.3(a) Consider a solution of A and B in which the fraction (by mass) of A is α (here
α = 21 ). The total volume of a solution of A and B is calculated from the partial
molar volumes of the two components using [5A.3–134], V = n A VA + n B VB . In
this exercise V and VA are known, so the task is therefore to find the amount
in moles, n A and n B , of A and B in the solution of known mass density ρ.
The mass of a volume V of the solution is ρV , so the mass of A is αρV . If
the molar mass of A is M A , then the amount in moles of A is n A = αρV /M A .
Similarly, n B = (1 − α)ρV /M B . The volume is expressed using these quantities
as
αρV VA (1 − α)ρV VB
V = n A VA + n B VB = +
MA MB
The term V cancels between the first and third terms to give
αρVA (1 − α)ρVB
1= +
MA MB
This equation is rearranged to give an expression for VA
MA (1 − α)VB ρ
VA = (1 − )
αρ MB
In this exercise let B be H2 O and A be ethanol, and as the mixture is 50%

by mass, α = 12 . The molar mass of B (H2 O) is M B = 16.00 + 2 × 1.0079 =
18.0158 g mol−1 and the molar mass of A (ethanol) is M A = 2 × 12.01 + 16.00 +
6 × 1.0079 = 46.0674 g mol−1 . The above expression for VA evaluates as
MA (1 − α)VB ρ
VA = (1 − )
αρ MB
(46.0674 g mol−1 )
=
0.5 × (0.914 g cm−3 )
−1
⎛ (1 − 0.5) × (17.4 cm3 mol ) × (0.914 g cm−3 ) ⎞
× 1−
⎝ 18.0158 g mol−1 ⎠
−1
= 56.3 cm3 mol
E5A.4(a) Henry’s law gives the partial vapour pressure of a solute B as p B = K B x B ,

[5A.24–142]. A test of this law is to make a plot of p B against x B which is
expected to be a straight line with slope K B ; such a plot is shown in Fig. 5.1.
100
p HCl /(kPa)
50
0
0.000 0.005 0.010 0.015 0.020
x HCl
Figure 5.1

p HCl /(kPa) = 6.41 × 103 × (x HCl ) − 0.071
If Henry’s law is obeyed the pressure should go to zero as x HCl goes to zero,
and the graph shows that this is almost achieved. Overall the conclusion is that
these data obey Henry’s law quite closely. The Henry’s law constant K HCl is
computed from the slope as 6.4 × 103 kPa .
E5A.5(a) In Section 5A.3(b) on page 142 it is explained that for practical applications
Henry’s law is often expressed as p B = K B b B , where b B is the molality of the
solute, usually expressed in mol kg−1 . The molality is therefore calculated from
the partial pressure as b B = p B /K B .
Molality is the amount of solute per kg of solvent. The mass m of a volume V
of solvent is given by m = ρV , where ρ is the mass density of the solvent. If the
amount of solute in volume V is n B , the molar concentration c B is related to
the molality by
bB
«
nB nB nB
cB = = =ρ = ρb B
V m/ρ m
Using Henry’s law the concentration is therefore given by
bB
«
pB ρx B p
c B = ρb B = ρ =
KB KB
where the partial pressure p B is expressed in terms of the mole fraction and the
total pressure p, p B = x B p.
The mole fraction of N2 in air is 0.780, the Henry’s law constant for N2 in
benzene is 1.87 × 104 kPa kg mol−1 and the density of benzene is 0.879 g cm−3 .
If it is assumed that the total pressure is 1 atm then
ρx N2 p (0.879 × 103 kg m−3 )×(0.780)×(101.325 kPa)
c N2 = = = 3.71... mol m−3
K N2 1.87 × 104 kPa kg mol−1
The molar concentration is therefore 3.7 × 10−3 mol dm−3 .

E5A.6(a) In Section 5A.3(b) on page 142 it is explained that for practical applications
Henry’s law is often expressed as p B = K B b B , where b B is the molality of the
solute, usually expressed in mol kg−1 . The molality is therefore calculated from
the partial pressure as b B = p B /K B . The Henry’s law constant for CO2 in water
is 3.01 × 103 kPa kg mol−1 .
For the case where the pressure of CO2 is 0.10 atm
p CO2 (0.10 atm) × (101.325 kPa/1 atm)
b CO2 = = −1 = 3.4 × 10−3 mol kg−1
K CO2 3
3.01 × 10 kPa kg mol
When the pressure is ten times greater at 1.00 atm the solubility is increased by
the same factor to 3.37 × 10−2 mol kg−1
E5A.7(a) As explained in Exercise E5A.9(a) the concentration of a solute is estimated
as c B = ρp B /K B where ρ is the mass density of the solvent. The Henry’s law
constant for CO2 in water is 3.01 × 103 kPa kg mol−1 and the density of water
is 0.997 g cm−3 or 997 kg m−3 .
ρp CO2 (997 kg m−3 ) × (5.0 atm) × (101.325 kPa/1 atm)

c CO2 = =
K CO2 3.01 × 103 kPa kg mol−1
= 1.67... × 102 mol m−3
The molar concentration is therefore 0.17 mol dm−3 .

E5A.8(a) The partial molar volume of B is defined from [5A.1–133] as
∂V
VB = ( )
∂n B p,T ,n ′
The polynomial given relates υ to x, and so from this it is possible to compute
the derivative dυ/dx. This required derivative is dV /dn B (where the partials
are dropped for simplicity), which is related to dυ/dx in the following way
dV dV dυ dx
( )=( )( )( )
dn B dυ dx dn B
Because x = n B /mol, dx/dn B = mol−1 , and because υ = V /cm3 , dυ/dV =
cm−3 and so dV /dυ = cm3 . Hence
dV dV dυ dx dυ
( )=( )( )( ) = ( ) cm3 mol−1
dn B dυ dx dn B dx
The required derivative is
dυ
( ) = 35.677 4 − 0.918 46 x + 0.051 975 x 2
dx
hence
VB = (35.677 4 − 0.918 46 x + 0.051 975 x 2 ) cm3 mol−1
E5A.9(a) The partial molar volume of solute B (here NaCl) is defined from [5A.1–133] as
∂V
VB = ( )
∂n B p,T ,n ′
The total volume is given as a function of the molality, but this volume is de-
scribed as that arising from adding the solute to 1 kg of solvent. The molality of
a solute is defined as (amount in moles of solute)/(mass of solvent in kg), there-
fore because in this case the mass of solvent is 1 kg, the molality is numerically
equal to the amount in moles, n B .
The polynomial given relates υ to x, and so from this it is possible to compute
the derivative dυ/dx. This required derivative is dV /dn B (where the partials
are dropped for simplicity), which is related to dυ/dx in the following way
dV dV dυ dx
( )=( )( )( )
dn B dυ dx dn B
The quantity x is defined as b/b −○ , but it has already been argued that the molal-
ity can be expressed as n B /(1 kg), hence x = n B /(mol) and therefore dx/dn B =
mol−1 . Because υ = V /cm3 , dυ/dV = cm−3 and so dV /dυ = cm3 . Hence
dV dV dυ dx dυ
( )=( )( )( ) = ( ) cm3 mol−1
dn B dυ dx dn B dx
The required derivative is
dυ
( ) = 16.62 + 2.655 x 1/2 + 0.24 x
dx
Hence the expression for the partial molar volume of B (NaCl) is
VB = (16.62 + 2.655 x 1/2 + 0.24 x) cm3 mol−1
The partial molar volume when b/b −○ = 0.1 is given by
VB /(cm3 mol−1 ) = (16.62 + 2.655 x 1/2 + 0.24 x)
= (16.62 + 2.655(0.100)1/2 + 0.24 × 0.100) = 17.4...
Therefore VB = 17.5 cm3 mol−1 .

The total volume is calculated from the partial molar volumes of the two com-
ponents, [5A.3–134], V = n A VA + n B VB . In this case V and VB are known,
so VA , the partial molar volume of the solvent water, can be found from VA =
(V − n B VB )/n A . The total volume when b/b −○ = 0.1 is given by
V = 1003 + 16.62 × 0.100 + 1.77 × 0.1003/2 + 0.12 × 0.1002 = 1004.7... cm3
The amount in moles of 1 kg of water is (1000 g)/[(16.00+2×1.0079) g mol−1 ] =
55.5... mol, hence
V − n B VB (1004.7... cm3 ) − (0.100 mol) × (17.4... cm3 mol−1 )
VA = =
nA 55.5... mol
= 18.1 cm3 mol−1
where, as before, for this solution a molality of 0.100 mol kg−1 corresponds to
n B = 0.100 mol.
E5A.10(a) For a binary mixture the Gibbs–Duhem equation, [5A.12b–136], relates changes
in the chemical potentials of A and B
n A dµ A + n B dµ B = 0
If it is assumed that the differential can be replaced by the small change
(0.1 n B ) × (+12 J mol−1 ) + n B δµ B = 0

(0.1 n B )
hence δµ B = − (+12 J mol−1 ) = −1.2 J mol−1
nB
E5A.11(a) Because the gases are assumed to be perfect and are at the same temperature
and pressure when they are separated, the pressure and temperature will not
change upon mixing. Therefore [5A.18–139], ∆ mix S = −nR(x A ln x A +x B ln x B ),
applies. The amount in moles is computed from the total volume, pressure
and temperature using the perfect gas equation: n = pV /RT. Because the
separate volumes are equal, and at the same pressure and temperature, each
compartment contains the same amount of gas, so the mole fractions of each
gas in the mixture are equal at 0.5.
∆ mix S = −nR(x A ln x A + x B ln x B ) = −(pV /T)(x A ln x A + x B ln x B )

(1.01325 × 105 Pa) × (5.0 × 10−3 m3 )
=− (0.5 ln 0.5 + 0.5 ln 0.5)
298.15 K
= +1.2 J K−1
Note that the pressure in expressed in Pa and the volume in m3 ; the units of the
result are therefore (N m−2 ) × (m3 ) × (K−1 ) = N m K−1 = J K−1 .
Under these conditions the Gibbs energy of mixing is given by [5A.17–138],
∆ mix G = nRT(x A ln x A + x B ln x B ); as before n = pV /RT.
∆ mix G = nRT(x A ln x A + x B ln x B ) = (pV )(x A ln x A + x B ln x B )

= [(1.01325 × 105 Pa) × (5.0 × 10−3 m3 )](0.5 ln 0.5 + 0.5 ln 0.5)
= −3.5 × 102 J
The units of the result are (N m−2 )×(m3 ) = N m = J. As expected, the entropy
of mixing is positive and the Gibbs energy of mixing is negative.
P5A.1 This problem is similar to the Example given in Section 5A.1(d) on page 136.
The Gibbs–Duhem equation [5A.12b–136], expressed in terms of partial molar
volumes is n A dVA + n B dVB = 0 which is rearranged to
nB
dVA = − dVB
nA
If the variation of the solute partial molar volume VB with concentration is
described by a known function, then integration of this equation gives an ex-
pression for how the solvent partial molar volume VA varies.
The range of integration of VA is from pure A, at which the partial molar volume
is equal to the molar volume of the pure solvent VA∗ , up to some arbitrary
concentration. The corresponding range for VB is from 0, the molar volume
of B in the limit of of no B being present (that is pure A), up to some arbitrary
concentration.
VA VB n
B
∫ ∗ dVA = − ∫ dVB
VA 0 nA
The expression for VB is given as a function of the molality, which is the amount
in moles divided by the mass of the solvent in kg. In 1 kg of solvent the amount
in moles is n A = (1 kg)/M A , where M A is the molar mass of the solvent A.
With this expression the ratio n B /n A is rewritten
nB nB nB MA
= =
n A (1 kg)/M A (1 kg)
The quantity n B /(1 kg) is recognised as the molality b of solute B, hence n B /n A =

bM A . The expression for VB is given in terms of x = b/b−○ , thus b = b −○ x and
hence n B /n A = M A b −○ x, With this, the integral to be evaluated becomes
VA VB
∫ dVA = − ∫ M A b −○ x dVB
VA∗ 0
The partial molar volumes VJ are replaced throughout by the dimensionless

−1
quantities υ J = VJ /(cm3 mol ) to give
υA υB
∫ dυ A = − ∫ M A b −○ x dυ B
υ ∗A 0
The next step is to change the variable of integration on the right from υ B to x;
this is done by differentiating the relationship between these two quantities
υ B = 5.117 + 19.121 x 1/2 hence dυ B = 9.5605 x −1/2 dx
The integral is then

υA x x
∫ dυ A = −M A b −○ ∫ x(9.5605 x −1/2 ) dx = −M A b −○ ∫ 9.5605 x 1/2 dx
υ ∗A 0 0
Evaluating the integrals gives
υ A − υ∗A = −M A b −○ × 23 × 9.5605 x 3/2
The molar mass of the solvent H2 O is 18.0158 g mol−1 ; for compatibility with
the units of molality this needs to be expressed as 1.80158 × 10−2 kg mol−1 . The
value of υ∗A is given as 18.079; with these values the expression for υ A becomes
υ A = 18.079 − 0.11483 x 3/2
P5A.3 The required molar masses are: N2 28.02 g mol−1 ; O2 32.00 g mol−1 ; Ar 39.95 g mol−1 ;
CO2 44.01 g mol−1 .
Consider 100 g of the mixture. Of this 75.5 g is N2 so the amount in moles of this
gas is n N2 = (75.5 g)/(28.02 g mol−1 ) = 2.69... mol. Similar calculations are
made for the other cases to give the results shown below in the table. The total
amount in moles n is found by summing these individual contributions and
this is then used to compute the mole fractions from x J = n J /n: the resulting
values are also shown in the table.
gas N2 O2 Ar CO2 total

mass % 75.5 23.2 1.3
n J /mol in 100 g 2.69... 0.725 0.0325... 3.45...
xJ 0.780... 0.210... 9.42... × 10−3
mass % 75.52 23.15 1.28 0.046
n J /mol in 100 g 2.69... 0.723... 0.0320... 1.04... × 10−3 3.45...
xJ 0.780... 0.209... 9.28... × 10−3 3.02... × 10−4
The entropy of mixing (at constant pressure and temperature) is given by a

generalisation of [5A.18–139]
∆ mix S = −nR ∑ x J ln x J
J
The entropy of mixing per mole is (∆ mix S)/n is given by
(∆ mix S)/n = −R ∑ x J ln x J
J
This expression is used togther with the values given in the table to compute
the entropy of mixing for the first set of data as +4.70 J K−1 mol−1 and for the
second set of data as +4.711 J K−1 mol−1 . The difference is of the order of
0.01 J K−1 mol−1 .
P5A.5 The definition of the partial molar volume VB is

∂V
VB = ( )
∂n B n A
which is interpreted as the slope of a plot of V against n B , at constant n A . Let
B be the solute CuSO4 and A be the solvent H2 O.
The task is therefore to calculate the volume of a solution with a fixed amount
of A as a function of the amount of B. The data given refer to a particular mass
of the solution, whereas what is required is data for a particular mass of solvent,
so some manipulation is required. Imagine a solution created from a fixed mass
m A /(g) of solvent and which contains a mass m B /(g) of solute; the total mass is
therefore m A /(g) + m B /(g). From the data supplied 100 g of solution contains
m/(g) of CuSO4 , so it follows that
multiples of 100 g
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
m A /(g) + m B /(g)
×m/(g) = m B /(g)
100
This equation is rearranged to give an expression for m B /(g)
m A /(g) × m/(g)
m B /(g) =
100 − m/(g)
The amount in moles of B is found using m B /M B , where M B is the molar mass

of B, which in this case is 159.61 g mol−1 .
The volume of this solution is computed from the mass density as (m A +m B )/ρ.
The following table of data is drawn up using m A = 1000 g as the fixed mass
of solvent, and using this a plot of V against n B is made, as shown in Fig. 5.2.
Note that m tot /(g) = 1000 + m B /(g).
m(CuSO4 )/g ρ/g cm−3 m B /g n B /mol m tot /g V /cm3

5 1.051 52.6 0.330 1 053 1 001.6
10 1.107 111.1 0.696 1 111 1 003.7
15 1.167 176.5 1.106 1 176 1 008.1
20 1.230 250.0 1.566 1 250 1 016.3
1 015
V /cm3
1 010
1 005
1 000
0.4 0.6 0.8 1.0 1.2 1.4 1.6
n B /mol
Figure 5.2
The data fit well to the polynomial (shown as the smooth curve on the plot)
V /(cm3 ) = 7.2249(n B /mol)2 − 1.8512(n B /mol) + 1001.4
The partial molar volume is the slope of this curve which is the derivative with
respect to n B
−1
VB /(cm3 mol ) = 14.450(n B /mol) − 1.8512
The following table gives values of VB for each of the data points. These are
plotted in Fig. 5.3; the line is the function above.
−1
m(CuSO4 )/g ρ/g cm−3 n B /mol VB /cm3 mol
5 1.051 0.330 2.91
10 1.107 0.696 8.21
15 1.167 1.106 14.13
20 1.230 1.566 20.78
20
−1
15
VB /cm3 mol
10
0.4 0.6 0.8 1.0 1.2 1.4 1.6

n B /mol
Figure 5.3
P5A.7 In Example 5A.1 on page 134 the partial molar volume of ethanol is found to be
given by
υ = 54.6664 − 0.72788 z + 0.084768 z 2
−1
where υ = VE /(cm3 mol ) and z = n E /mol. The value of z at which υ is a
minimum or maximum is found by setting the derivative dυ/dz = 0
dυ 0.72788
= −0.72788 + 0.169536 z = 0 hence z= = 4.2934
dz 0.169536
This value of z corresponds to 4.2934 mol in 1.000 kg of solvent water (specified
in the Example). The molality is the amount in moles divided by the mass of the
solvent in kg, thus the corresponding molality is 4.2934 mol kg−1 . The plot in
the text confirms that this is indeed the position of the minimum in the partial
molar volume.
5B The properties of solutions

Answers to discussion question
D5B.1 The boiling-point constant is given by [5B.9b–150], K = RT ∗2 /∆ vap H, where
T ∗ is the boiling point of the pure liquid and ∆ vap H is its enthalpy of vapori-
sation. However, by Trouton’s rule (Section 3B.2 on page 85), ∆ vap H/T ∗ is ap-
proximately constant, so the boiling-point constant is simply ∝ T ∗ . Differences
in boiling-point constants are therefore identified as being due to differences in

the boiling points of the pure liquids. Water and benzene have different boiling
points and so have different boiling-point constants.
D5B.3 The typical experimental arrangement for observing osmosis involves a pure
solvent being separated from a solution by a semipermeable membrane through
which only the solvent can pass. The chemical potential of the solvent in the
solution is lower than that of the pure solvent, therefore there is a tendency for
the solvent to pass through the membrane from the side on which it is pure
into the solution because this results in a reduction in Gibbs energy.
At a molecular level the process involves an increase in ‘randomness’ as in-
creasing the amount of solvent in the solution increases the number of possible
arrangements of solvent and solute molecules.
D5B.5 A regular solution has excess entropy S E of zero, but an excess enthalpy H E that
is non-zero. A regular solution of A and B can be thought of as one in which the
different molecules of A and B are distributed randomly, as in an ideal solution,
but where the energy of A–A, B–B, and A–B interactions are different.
In real solutions both S E and H E are non-zero, and in general both are likely
to vary with composition. The non-zero value for S E is interpreted as arising
from the non-random distribution of molecules. This is exemplified by ionic
solutions, in which ions of one charge are more likely to be surrounded by ions
of the opposite charge than of the same charge (Topic 5F).
D5B.7 All of the colligative properties result from the lowering of the chemical poten-
tial of the solvent due to the presence of the solute. For an ideal solution, this
reduction is predicted by µ A = µ ∗A + RT ln x A . The relationship shows that as
the amount of solute increases, the mole fraction of the solvent x A decreases
and hence the chemical potential the solvent A decreases.
If the chemical potential of the solvent is lowered, then the chemical potential
of the vapour in equilibrium with the solvent is also lowered because at equi-
librium these two chemical potentials must be equal. The chemical potential
of a perfect gas is given by µ A = µ −A○ + RT ln p A , so a lowering of the chemical
potential results in a reduction in the pressure.
The overall result is that addition of a solute reduces the vapour pressure of
the solvent, and therefore the temperature at which the solvent boils is raised
because a greater increase in temperature is needed to make the vapour pres-
sure equal to the external pressure. Similarly, the freezing point of the solvent
is decreased because the chemical potential of the solid will equal that of the
solvent at a lower temperature.
At a molecular level the decrease in vapour pressure can be thought of as being
due to the solute molecules getting in the way of the solvent molecules, thus
reducing their tendency to escape. Another way of looking at this is that the
presence of a solute increases the ‘randomness’, and hence the entropy, of the
solution, thus reducing the tendency for the formation of the (pure) vapour or
solid.
E5B.1(a) In Exercise E5A.8(a) it is found that the vapour pressure obeys
p HCl /(kPa) = 6.41 × 103 × (x HCl ) − 0.071 (5.1)
The task is to work out the mole fraction that corresponds to the given molality.
The molality of HCl is defined as b HCl = n HCl /m GeCl4 , where n HCl is the amount
in moles of HCl and m GeCl4 is the mass in kg of solvent GeCl4 . The mole
fraction of HCl is n HCl /(n HCl + n GeCl4 ), where n GeCl4 is the amount in moles of
GeCl4 , which is given by n GeCl4 = m GeCl4 /M GeCl4 , where M GeCl4 is the molar
mass of GeCl4 . These relationships allow the mole fraction to be rewritten as
follows
n HCl n HCl
x HCl = =
n HCl + n GeCl4 n HCl + m GeCl4 /M GeCl4
The amount in moles of HCl is written is n HCl = b HCl m GeCl4 ; using this the
above expression for the mole fraction becomes
n HCl b HCl m GeCl4 b

x HCl = = =
n HCl + m GeCl4 /M GeCl4 b HCl m GeCl4 + m GeCl4 /M GeCl4 b + 1/M GeCl4
The molar mass of GeCl4 is 214.44 g mol−1 , therefore the mole fraction corre-
sponding to b = 0.10 mol kg−1 is
b (0.10 mol kg−1 )

x HCl = = = 0.0209...
b + 1/M (0.10 mol kg ) + 1/(214.44 × 10−3 kg mol−1 )
−1
The pressure is found by inserting this value into eqn 5.1
p HCl /(kPa) = 6.41 × 103 × (0.0209...) − 0.071 = 1.34... × 102
The vapour pressure of HCl is therefore 1.3 × 102 kPa .
E5B.2(a) Raoult’s law, [5A.22–141], p A = x A p∗A relates the vapour pressure to the mole
fraction of A, therefore from the given data is it possible to compute x A . The
task is to relate the mole fraction of A to the masses of A (the solvent) and B
(the solute), and to do this the molar masses M A and M B are introduced. With
these n A = m A /M A , where m A is the mass of A, and similarly for n B . It follows
that
nA m A /M A MB mA
xA = = =
n A + n B m A /M A + m B /M B M B m A + M A m B
The final form of this expression for x A is rearranged to given an expression for
M B , which is the desired quantity; then x A is replaced by p A /p∗A
xA MA mB (p A /p∗A )M A m B
MB = =
m A (1 − x A ) m A [1 − (p A /p∗A )]
The molar mass of the solvent benzene, A, is 78.1074 g mol−1 , hence

(p A /p∗A )M A m B
MB =
m A [1 − (p A /p∗A )]
[(51.5 kPa)/(53.3 kPa)] × (78.1074 g mol−1 ) × (19.0 g)
=
(500 g) × [1 − (51.5 kPa)/(53.3 kPa)]
= 84.9 g mol−1
E5B.3(a) The freezing point depression ∆Tf is related to the molality of the solute B, b B ,
by [5B.12–151], ∆Tf = K f b B , where K f is the freezing-point constant. From the
data and the known value of K f it is possible to calculate b B . The task is then to
relate this to the given masses and the desired molar mass of the solute, M B .
The molality of B is defined as b B = n B /m A , where m A is the mass of the solvent
A in kg. It follows that
nB m B /M B
bB = =
mA mA
where m B is the mass of solute B. From the freezing point data b B = ∆Tf /K f ,
therefore
∆Tf m B /M B mB Kf
= hence M B =
Kf mA m A ∆Tf
With the data given and the value of the freezing-point constant from the Re-
source section
(100 g) × (30 K kg mol−1 )
MB = = 381 g mol−1
(0.750 kg) × (10.5 K)
Note that because molality is defined as (amount in moles)/(mass of solvent in
kg), the mass of solvent m A is used as 0.750 kg.
E5B.4(a) The freezing point depression ∆Tf is related to the molality of the solute B,
b B , by [5B.12–151], ∆Tf = K f b B , where K f is the freezing-point constant. The
molality of the solute B is defined as b B = n B /m A , where n B is the amount in
moles of B and m A is the mass in kg of solvent A. The amount is related to the
mass of B, m B , using the molar mass M B : n B = m B /M B . It therefore follows
that
Kf mB
∆Tf = K f b B =
MB mA
The molar mass of sucrose C12 H22 O11 is 342.2938 g mol−1 . A volume 200 cm3 of
water has mass 200 g to a good approximation. Using these values with the data
given and the value of the freezing-point constant from the Resource section
gives the freezing point depression as
Kf mB (1.86 K kg mol−1 ) × (2.5 g)

∆Tf = = = 0.0679... K
M B m A (342.2938 g mol−1 ) × (0.200 kg)
Note that because molality is defined as (amount in moles)/(mass of solvent
in kg), the mass of solvent m A is used as 0.200 kg. The new freezing point is
therefore 273.15 K − 0.0679... K = 273.08 K
E5B.5(a) The osmotic pressure Π is related to the molar concentration of solute B, [B],
by [5B.16–153], Π = [B]RT. The freezing point depression ∆Tf is related to the
molality of B, b B , by [5B.12–151], ∆Tf = K f b B , where K f is the freezing-point
constant. The task is to relate [B] to b B so that these two relationships can be
used together.
The molar concentration [B] is given by [B] = n B /V , where n B is the amount
in moles of B and V is the volume of the solvent A. This volume is related to
the mass of A, m A , using the mass density ρ: V = m A /ρ. It therefore follows
that
bB
«
nB nB nB
[B] = = = ρ = bB ρ
V m A /ρ m A
With this the osmotic pressure is related to the molality
Π Π Π
[B] = hence bB ρ = and so bB =
RT RT ρRT
The freezing point depression for a solution exerting this osmotic pressure is
therefore
Kf Π
∆Tf = K f b B =
ρRT
Note that because molality is defined as (amount in moles)/(mass of solvent in
kg), the mass of solvent m A must be in kg and therefore the mass density must
be used in kg volume−1 .
With the data given, the value of the freezing-point constant from the Resource
section, and taking the mass density of water as 1 g cm−3 = 1000 kg m−3 gives
the freezing point depression as
Kf Π (1.86 K kg mol−1 ) × (120 × 103 Pa)

∆Tf = =
ρRT (1000 kg m−3 ) × (8.3145 J K−1 mol−1 ) × (300 K)
= 0.0894... K
In this expression all of the quantities are in SI units therefore the temperature
is expected to be in K, which is verified as follows
(K kg mol−1 ) × (Pa) Pa
−3 −1 −1 = −3 × K−1
(kg m ) × (J K mol ) × (K) J × m
kg m−1 s−2
= =K
(kg m2 s−2 ) × m−3 × K−1
The freezing point is therefore 273.15 K − 0.0894... K = 273.06 K
E5B.6(a) The Gibbs energy of mixing is given by [5B.3–145], ∆ mix G = nRT(x A ln x A +

x B ln x B ), the entropy of mixing by [5B.4–145], ∆ mix S = −nR(x A ln x A +x B ln x B ).
∆ mix H for an ideal solution is zero .
The total amount in moles is 0.50 mol+2.00 mol = 2.50 mol. With A as hexane
and B as heptane the thermodynamic quantities are calculated as
∆ mix G = nRT(x A ln x A + x B ln x B )
0.50 0.50 2.00 2.00
= (2.50 mol) × (8.3145 J K−1 mol−1 ) × (298 K) × ( ln + ln )
2.50 2.50 2.50 2.50
= −3.10 × 103 J
∆ mix S = −nR(x A ln x A + x B ln x B )
0.50 0.50 2.00 2.00
= −(2.50 mol) × (8.3145 J K−1 mol−1 ) × ( ln + ln )
2.50 2.50 2.50 2.50
= +10.4 J K−1
E5B.7(a) The entropy of mixing is given by [5B.4–145], ∆ mix S = −nR(x A ln x A +x B ln x B ),

and is a maximum when x A = x B = 12 . This is evident from Fig. 5B.2 on page
146.
The task is to relate the mole fraction of A (heptane) to the masses of A and
B (hexane), and to do this the molar masses M J are introduced. With these
n J = m J /M J , where m J is the mass of J. It follows that
nA m A /M A MB mA
xA = = =
n A + n B m A /M A + m B /M B M B m A + M A m B
This is rearranged to give an expression for m B /m A
MB mA MB mB MB 1
xA = = hence = ( − 1)
M B m A + M A m B M B + M A (m B /m A ) mA MA xA
The molar mass of A (heptane) is 100.1964 g mol−1 , and that of B (hexane) is

86.1706 g mol−1 . With these values and x A = 21
mB MB 1 86.1706 g mol−1 1
= ( − 1) = ( − 1) = 0.8600
mA MA xA 100.1964 g mol−1 1/2
More simply, if equal amounts in moles of A and B are required, the ratio of
the corresponding masses of A and B must be equal to the ratio of their molar
masses: m B /m A = M B /M A .
E5B.8(a) The ideal solubility of solute B at temperature T is given by [5B.14–152], ln x B =

(∆ fus H/R)(1/Tf − 1/T), where ∆ fus H is the enthalpy of fusion of the solute,
and Tf is the freezing point of the pure solute.
∆ fus H 1 1
ln x B = ( − )
R Tf T
28.8 × 103 J mol−1 1 1
= −1 −1 ( − ) = −4.55...
8.3145 J K mol (217 + 273.15) K (25 + 273.15) K
hence x B = 0.0105....
The mole fraction is expressed in terms of the molality, b B = n B /m A , where m A

is the mass of the solvent in kg, in the following way
nB nB n B /m A bB
xB = = = =
n A + n B m A /M A + n B 1/M A + n B /m A 1/M A + b B
xB
hence bB =
(1 − x B )M A
where M A is the molar mass of A, expressed in kg mol−1 . The molar mass of
solvent benzene is 78.1074 g mol−1 or 78.1074 × 10−3 kg mol−1 , therefore
xB 0.0105...
bB = = = 0.136... mol kg−1
(1 − x B )M A (1 − 0.0105...) × (78.1074 × 10−3 kg mol−1 )
The molality of the solution is therefore 0.137 mol kg−1 . The molar mass of
anthracene (C14 H10 ) is 178.219 g mol−1 , so the mass of anthracene which is
dissolved in 1 kg of solvent is (0.136... mol kg−1 )×(1 kg)×(178.219 g mol−1 ) =
24.3 g .
E5B.9(a) Let the solvent CCl4 be A and the solute Br2 be B. The vapour pressure of the
solute in an ideal dilute solution obeys Henry’s law, [5A.24–142], p B = K B x B ,
and the vapour pressure of the solvent obeys Raoult’s law, [5A.22–141], p A =
p∗A x A .
p B = K B x B = (122.36 Torr) × 0.050 = 6.11... Torr
p A = p∗A x A = (33.85 Torr) × (1 − 0.050) = 32.1... Torr
p tot = p A + p A = (6.11... Torr) + (32.1... Torr) = 38.2... Torr
Therefore the pressure are p B = 6.1 Torr , p A = 32 Torr , and p tot = 38 Torr .
The partial pressure of the gas is given by p A = y A p tot , where y A is the mole
fraction in the vapour
pA 32.1... Torr
yA = = = 0.84
p tot 38.2... Torr
pB 6.11... Torr
yB = = = 0.16
p tot 38.2... Torr
E5B.10(a) Let methylbenzene be A and 1,2-dimethylbenzene be B. If the solution is ideal

the vapour pressure obeys Raoult’s law, [5A.22–141], p J = p∗J x J . The mixture
will boil when the sum of the partial vapour pressures of A and B equal the
external pressure, here 0.50 atm.
p ext = p A + p B = x A p∗A + x B p∗B = x A p∗A + (1 − x A )p∗B
p ext − p∗B p ext − p∗A
hence x A = ∗ and by analogy x B =
p A − p∗B p∗B − p∗A
(0.50 atm) × [(101.325 kPa)/(1 atm)] − (20.0 kPa)

xA = = 0.920... = 0.92
(53.3 kPa) − (20.0 kPa)
(0.50 atm) × [(101.325 kPa)/(1 atm)] − (53.3 kPa)
xB = = 0.0792... = 0.08
(20.0 kPa) − (53.3 kPa)
The partial pressure of the gas is given by p J = y J p ext , where y J is the mole
fraction in the gas, and p J is given by p J = p∗J x J , hence y J = x J p∗J /p ext
x A p∗A (0.920...) × (53.3 kPa)

yA = = = 0.97
p ext (0.50 atm) × [(101.325 kPa)/(1 atm)]
x B p∗B (0.0792...) × (20.0 kPa)
yB = = = 0.03
p ext (0.50 atm) × [(101.325 kPa)/(1 atm)]
E5B.11(a) The vapour pressure of component J in the solution obeys Raoult’s law, [5A.22–
141], p J = p∗J x J , where x J is the mole fraction in the solution. In the gas the
partial pressure is p J = y J p tot , where y J is the mole fraction in the vapour.
These relationships give rise to four equations
p A = p∗A x A p B = p∗B (1 − x A ) p A = p tot y A p B = p tot (1 − y A )
where x A + x B = 1 is used and likewise for the gas. In these equations x A and
p tot are the unknowns to be found. The expressions for p A are set equal, as are
those for p B , to give
p∗A x A
p∗A x A = p tot y A hence p tot =
yA
p∗B (1 − x A )
p∗B (1 − x A ) = p tot (1 − y A ) hence p tot =
1 − yA
These two expressions for p tot are set equal and the resulting equation rear-
ranged to find x A
p∗A x A p∗B (1 − x A ) p∗B y A

= hence xA =
yA 1 − yA p∗A (1 − y A ) + p∗B y A
With the data given
p∗B y A (52.0 kPa) × (0.350)

xA = =
p∗A (1 − y A ) + p∗B y A (76.7 kPa)∗ (1 − 0.350) + (52.0 kPa) × (0.350)
= 0.267... and x B = 1 − 0.267... = 0.732...
The composition of the liquid is therefore x A = 0.267 and x B = 0.733 .

The total pressure is computed from p A = p tot y A and p A = p∗A x A to give p tot =
x A p∗A /y A
x A p∗A (0.267...) × (76.7 kPa)

p tot = = = 58.6 kPa
yA 0.350
E5B.12(a) If the solution is ideal, the vapour pressure of component J in the solution
obeys Raoult’s law, [5A.22–141], p J = p∗J x J , where x J is the mole fraction in
the solution. In the gas the partial pressure is p J = y J p tot , where y J is the mole
fraction in the vapour.
Assuming ideality, the total pressure is computed as
p tot = p A + p B = p∗A x A + p∗B (1 − x A )

= (127.6 kPa) × (0.6589) + (50.60 kPa) × (1 − 0.6589) = 101 kPa
The normal boiling point is when the total pressure is 1 atm, and this is exactly
the pressure found by assuming Raoult’s law applies. The solution is therefore
ideal .
The composition of the vapour is computed from p A = p tot y A and p A = p∗A x A
hence
pA p∗ x A (127.6 kPa) × (0.6589)
yA = = A = = 0.829...
p tot p tot 101.325 kPa
It follows that y B = 1 − 0.829... = 0.170.... The composition of the vapour is
therefore y A = 0.830 and y B = 0.170 .
P5B.1 The freezing point depression in terms of mole fraction is predicted by [5B.11–
151]
RT ∗2
∆T = x B K ′ K′ =
∆ fus H
With the data given
RT ∗2 (8.3145 J K−1 mol−1 ) × (290 K)2

K′ = = = 61.3... K
∆ fus H 11.4 × 103 J mol−1
The data are given in terms of molality, which is n B /m A , where n B is the amount
in moles of solute and m A is the mass of solvent in kg. The mole fraction x B is
related to the molality by using the molar mass of the solvent, M A
nB nB n B /m A bB
xB = = = =
n A + n B m A /M A + n B 1/M A + n B /m A 1/M A + b B
The molar mass of ethanoic acid CH3 COOH is 60.0516 g mol−1 . Because m A
must be in kg the molar mass must be expressed in kg volume−1 , M A = 60.0516×
10−3 kg mol−1 . For the data given 1/M A ≫ b B therefore the expression for the
mole fraction is well approximated by x B = b B M A . With this, the freezing point
depression is given by
xB

∆T = b B M A K ′ hence b B = ∆T/M A K ′
The table below gives values of b B calculated from the given ∆T and this ex-
pression; to distinguish these values for the experimental values of b B , the cal-
culated values are termed apparent molalities, b B,app
b B /(mol kg−1 ) ∆T/K b B,app /(mol kg−1 ) b B,app /b B M B,app /(g mol−1 )
0.015 0.115 0.031 2.081 27.9
0.037 0.295 0.080 2.165 26.8
0.077 0.470 0.128 1.657 35.1
0.295 1.381 0.375 1.271 45.7
0.602 2.67 0.725 1.204 48.3
The apparent molar mass of B, M B,app , is computed using
M B,app bB
=
MB b B,app
with M B = 58.01 g mol−1 , the molar mass of KF. The argument that leads
to this is that the greater the apparent molality the smaller the molar mass:
M B,app ∝ 1/b B,app .
The data in the table show that the molality predicted from the experimental
freezing point depression using the value of the freezing-point constant deter-
mined by [5B.11–151] is always greater than the molality know from the way
the solution was prepared. Presumably this latter molality is based on adding
a know mass of KF to a known mass of solvent, and assuming that the molar
mass of KF is 58.1 g mol−1 . The fact that the apparent molality is higher than
the molality of the prepared solution implies that the number of solute species
is greater than expected.
The freezing point depression depends on the mole fraction of the solute, re-
gardless of its identity. Therefore if the added KF were to dissociate completely
on dissolution in ethanoic acid the mole fraction of the solute would be twice as
large as expected on the basis of the amount of added KF, and in turn this would
mean that the apparent molality (based on the freezing-point depression) is
twice as large as expected.
The data in the table can be interpreted as indicating that there is dissociation
of the KF, and that this dissociation is greater at lower molalities. However, this
only part of the picture as it does no explain why b B,app /b B is greater than 2 at
some molalities.
P5B.3 Let the two components of the mixture be labelled 1 (propionic acid) and 2
(THP). The definition of the partial molar volume of 1, V1 , is
∂V
V1 = ( )
∂n 1 n 2
To use this definition an expression for V as a function of the n J is required.

The excess volume V E is defined as V E = ∆V −∆V ideal , where ∆V is the volume
of mixing and ∆V ideal is the volume of mixing of the ideal solution, which is
zero. Therefore V E = ∆V .
The volume of mixing ∆V is written ∆V = V − Vseparated , which from the above

is also written V E = V − Vseparated . The expression given in the problem for V E
is per mole, so for mixing n 1 moles of 1 is mixed with n 2 moles of 2 the excess
volume is in fact (n 1 + n 2 )V E .
The volume of the separated components is computed from their mass densi-
ties: n 1 moles corresponds to a mass n 1 M 1 , where M 1 is the molar mass, which
has volume n 1 M 1 /ρ 1 , where ρ 1 is the mass density. It follows that
n1 M1 n2 M2 n1 M1 n2 M2
(n 1 + n 2 )V E = V − − hence V = (n 1 + n 2 )V E + +
ρ1 ρ2 ρ1 ρ2
The second equation above is the required expression for V as a function of n 1

and n 2 . The expression given in the problem for V E is a function of the mole
fractions, which are easily written in terms of the amounts.
To compute the partial molar volume it is necessary to compute the derivative

of V with respect to n 1 , keeping in mind that V E is a function of n 1
n1 M1 n2 M2
V = (n 1 + n 2 )V E + +
ρ1 ρ2
∂V ∂V E M1
hence V1 = ( ) = (n 1 + n 2 ) ( ) + VE + (5.2)
∂n 1 n 2 ∂n 1 n 2 ρ1
To compute the derivative it has been recognised that (n 1 + n 2 )V E is a product

of two functions of n 1 .
The first step is to compute (∂V E /∂n 1 )n 2 , and this requires rewriting the mole
fractions in terms of the n i
V E = x 1 x 2 a 0 + x 12 x 2 a 1 − x 1 x 22 a 1
n1 n2 a0 n 12 n 2 a 1 n 1 n 22 a 1
= + −
(n 1 + n 2 )2 (n 1 + n 2 )3 (n 1 + n 2 )3
∂V E n2 a0 2n 1 n 2 a 0 2n 1 n 2 a 1
( ) = − +
∂n 1 n 2 (n 1 + n 2 ) 2 (n 1 + n 2 ) 3 (n 1 + n 2 )3
3n 12 n 2 a 1 n 22 a 1 3n 1 n 22 a 1
− − +
(n 1 + n 2 )4 (n 1 + n 2 )3 (n 1 + n 2 )4
The quantity required is (n 1 +n 2 )(∂V E /∂n 1 )n 2 , so the above expression is mul-

tiplied by (n 1 + n 2 ). This cancels a term (n 1 + n 2 ) in each of the denominators
and allows the expression to be rewritten in terms of the mole fractions
(n 1 + n 2 )(∂V E /∂n 1 )n 2 = x 2 a 0 −2x 1 x 2 a 0 +2x 1 x 2 a 1 −3x 12 x 2 a 1 − x 22 a 1 +3x 1 x 22 a 1

The parts of eqn 5.2 are now assembled
∂V E M1
V1 = (n 1 + n 2 ) ( ) + VE +
∂n 1 n 2 ρ1
= (x 2 a 0 − 2x 1 x 2 a 0 + 2x 1 x 2 a 1 − 3x 12 x 2 a 1 − x 22 a 1 + 3x 1 x 22 a 1 )
+ (x 1 x 2 a 0 + x 12 x 2 a 1 − x 1 x 22 a 1 ) + M 1 /ρ 1
= a 0 x 2 (1 − x 1 ) + a 1 x 2 (2x 1 − 2x 12 − x 2 + 2x 1 x 2 ) + M 1 /ρ 1
= a 0 x 2 (x 2 ) + a 1 x 2 (2x 1 [1 − x 1 ] − x 2 + 2x 1 x 2 ) + M 1 /ρ 1
= a 0 x 22 + a 1 x 2 (2x 1 x 2 − x 2 + 2x 1 x 2 ) + M 1 /ρ 1
= a 0 x 22 + a 1 x 2 (4x 1 x 2 − x 2 ) + M 1 /ρ 1
= a 0 x 22 + a 1 x 22 (4x 1 − 1) + M 1 /ρ 1
The last four lines involve repeated use of x 1 + x 2 = 1 in order to simplify the
expression. A similar process is used to find an expression for V2 . In principle
all that is required is to swap the indices 1 and 2, however when this is done for
the expression for V E the result is
V E = x 2 x 1 a 0 + x 22 x 1 a 1 − x 2 x 1 a 1
which, when compared with the original expression, shows that the sign of the
term in a 1 is reversed: this change needs to be carried through to the end. In
summary
V1 = a 0 x 22 + a 1 x 22 (4x 1 − 1) + M 1 /ρ 1
V2 = a 0 x 12 − a 1 x 12 (4x 2 − 1) + M 2 /ρ 2
The molar mass of propionic acid CH3 CH2 COOH is M 1 = 74.0774 g mol−1
and that of THP C5 H10 O is M 1 = 86.129 g mol−1 . For an equimolar mixture
x 1 = x 2 = 12 and therefore
V1 = a 0 x 22 + a 1 x 22 (4x 1 − 1) + M 1 /ρ 1 = 41 a 0 + 14 a 1 + M 1 /ρ 1
−1 −1
= 0.25 × (−2.4697 cm3 mol ) + 0.25 × (0.0608 cm3 mol )
−1
+ (74.0774 g mol−1 )/(0.97174 g cm−3 ) = 75.6 cm3 mol
V2 = a 0 x 12 − a 1 x 12 (4x 2 − 1) + M 2 /ρ 2 = 41 a 0 − 14 a 1 + M 2 /ρ 2
−1 −1
= 0.25 × (−2.4697 cm3 mol ) − 0.25 × (0.0608 cm3 mol )
−1
+ (86.129 g mol−1 )/(0.86398 g cm−3 ) = 99.1 cm3 mol
P5B.5 The excess Gibbs energy G E is defined in [5B.5–146], G E = ∆ mix G − ∆ mix G ideal .
The ideal Gibbs energy of mixing (per mole) is given by [5B.3–145], ∆ mix G ideal =
RT(x A ln x A + x B ln x B ). Let A by MCH and B be THF. The Gibbs energy of

mixing of n A moles of A with n B moles of B is therefore given by
∆ mix G = ∆ mix G ideal + G E

= (n A + n B )RT(x A ln x A + x B ln x B )
+ (n A + n B )RTx A (1 − x A )[0.4857 − 0.1077(2x A − 1)
+ 0.0191(2x A − 1)2 ]
With the values given (n A + n B ) = 4 mol, x A = 41 , and x B = 3

4
∆ mix G = (4.00 mol) × (8.3145 J K−1 mol−1 ) × (303.15 K) × ( 41 ln 14 + 34 ln 34 )

+ (4.00 mol) × (8.3145 J K−1 mol−1 ) × (303.15 K) × 41 × (1 − 41 )
× [0.4857 − 0.1077(2 × 14 − 1) + 0.0191(2 × 14 − 1)2 ]
= −4.64 kJ
P5B.7 The osmotic pressure Π is related to the molar concentration [J] through a
virial-type equation, [5B.18–153], Π = [J]RT(1 + B[J]). The data are given
in terms of the mass concentration, so the first task is to relate this to the molar
concentration. If the amount in moles of solute dissolved in volume V is n J and
the molar mass is M J it follows that
cJ
«
n J m J /M J m J 1 cJ
[J] = = = =
V V V MJ MJ
where c J is the mass concentration. With this the virial equation is rewritten
cJ cJ
Π = [J]RT(1 + B[J]) = RT (1 + B )
MJ MJ
Division of both sides by c J gives an equation of a straight line

Π RT BRT
= + cJ
cJ MJ M J2
A plot of Π/c J against c J will have intercept RT/M J when c J = 0, and from this
it is possible to determine the molar mass.
The pressure is given by Π = hρg; for the pressure to be in Pa the height needs
to be in m and ρ in kg m−3 ; for the present case ρ = 1 g cm−3 = 1000 kg m−3 .
The data are plotted in Fig. 5.4.
c/(mg cm−3 ) h/(cm) Π/Pa (Π/c)/(Pa mg−1 cm3 )

3.221 5.746 563.7 175.002 4
4.618 8.238 808.1 174.999 5
5.112 9.119 894.6 174.994 9
6.722 11.990 1 176.2 174.980 5
175.02
(Π/c)/(Pa mg−1 cm3 )

175.00
174.98
0 1 2 3 4 5 6 7
−3
c/(mg cm )
Figure 5.4
The data are a modest fit to a straight line, the equation of which is
(Π/c)/(Pa mg−1 cm3 ) = −6.628 × 10−3 × c/(mg cm−3 ) + 175.03
The intercept is RT/M J ; before using the intercept in this expression it is best
to convert it from (Pa mg−1 cm3 ) to SI units, (Pa kg−1 m3 )
1 m3 1 mg
(175.03 Pa mg−1 cm3 ) × 6 3
× −6 = 175.03 Pa kg−1 m3
10 cm 10 kg
RT (8.3145 J K−1 mol−1 ) × (293.15 K)

MJ = = = 13.92... kg mol−1
intercept 175.03 Pa kg−1 m3
The molar mass of the protein is therefore 1.39 × 104 g mol−1 .
P5B.9 The osmotic pressure Π is related to the molar concentration [J] through a
virial-type equation, [5B.18–153], Π = [J]RT(1 + B[J]). As is shown in Problem
P5B.7 this equation can be rewritten in terms of the mass concentration c J and
the molar mass of J, M J
Π RT BRT
= + cJ
cJ MJ M J2
A plot of Π/c J against c J will have intercept RT/M J when c J = 0: from this it is
possible to determine the molar mass. The second virial coefficient is obtained
from the slope. The plot is shown in Fig. 5.5.
c/(mg cm−3 ) Π/Pa (Π/c)/(Pa mg−1 cm3 )

1.33 30 22.6
2.10 51 24.3
4.52 132 29.2
7.18 246 34.3
9.87 390 39.5
40
(Π/c)/(Pa mg−1 cm3 )

35
30
25
20
0 2 4 6 8 10
−3
c/(mg cm )
Figure 5.5
(Π/c)/(Pa mg−1 cm3 ) = 1.975 × (c/(mg cm−3 )) + 20.09
to convert it from (Pa mg−1 cm3 ) to SI units, (Pa kg−1 m3 )
1 m3 1 mg
(20.09 Pa mg−1 cm3 ) × 6 3
× −6 = 20.09 Pa kg−1 m3
10 cm 10 kg
RT (8.3145 J K−1 mol−1 ) × (303.15 K)
MJ = = = 1.25... × 102 kg mol−1
The molar mass of the polymer is therefore 1.25 × 105 g mol−1 .
The second virial coefficient is found by taking the ratio (slope)/(intercept)
slope B slope
= hence B = MJ ×
intercept M J intercept
1.975 Pa mg−2 cm6

B = (1.25... × 102 kg mol−1 ) ×
20.09 Pa mg−1 cm3
1.975 Pa mg−2 cm6
= (1.25... × 108 mg mol−1 ) × = 1.23... × 107 cm3 mol−1
20.09 Pa mg−1 cm3
Given that [J] is usually in mol dm−3 it is convenient to quote the value of the
second virial coefficient as B = 1.23 × 104 mol−1 dm3 .
P5B.11 The excess enthalpy of mixing for this particular regular solution is given by
[5B.6–147], H E = nRT ξx A x B . The plot in Fig. 5.6 shows H E /(nRT) as a func-
tion of x A for different values of ξ; recall that x A +x B = 1, so x A x B = x A (1−x A ).
If ξ is fixed, the temperature dependence is explored by plotting H E /(nRξ) as
a function of x A : H E /(nRξ) = Tx A x B = Tx A (1− x A ). This is shown in Fig. 5.7.
Evidently the strongest temperature dependence is once more at x A = 21 .
0.4 ξ = −2
ξ = −1
0.2 ξ=0
H E /(nRT)
ξ = +1
0.0 ξ = +2
−0.2
−0.4
0.0 0.2 0.4 0.6 0.8 1.0

xA
Figure 5.6
200 K
80
250 K
300 K
(H E /(nRξ))/K
60 350 K
40
20
0
0.0 0.2 0.4 0.6 0.8 1.0
xA
Figure 5.7
P5B.13 The osmotic pressures Π is related to the molar concentration [J] through a
virial-type equation, [5B.18–153], Π = [J]RT(1 + B[J]). As is shown in Problem
P5B.7 this equation can be rewritten in terms of the mass concentration c J and
the molar mass of J, M J
Π RT BRT
= + cJ
cJ MJ M J2
A plot of Π/c J against c J will have intercept RT/M J when c J = 0: from this it is
possible to determine the molar mass. The second virial coefficient is obtained
from the slope. Such a plot is shown in Fig. 5.8.
c/(g dm−3 ) Π/Pa (Π/c)/(Pa g−1 dm3 )

1.00 27 27.0
2.00 70 35.0
4.00 197 49.3
7.00 500 71.4
9.00 785 87.2
80
(Π/c)/(Pa g−1 dm3 )
60
40
20
0 2 4 6 8 10
−3
c/(g dm )
Figure 5.8
(Π/c)/(Pa g−1 dm3 ) = 7.466 × (c/(g dm−3 )) + 19.64
to convert it from (Pa g−1 dm3 ) to SI units, (Pa kg−1 m3 )
1 m3 1g
(19.64 Pa g−1 dm3 ) × × = 19.64 Pa kg−1 m3
103 dm3 10−3 kg
RT (8.3145 J K−1 mol−1 ) × (298 K)
MJ = = = 1.26... × 102 kg mol−1
The molar mass of the polymer is therefore 1.26 × 105 g mol−1 . The second
virial coefficient is found by taking the ratio (slope)/(intercept)
slope B slope
= hence B = MJ ×
intercept M J intercept
7.466 Pa g−2 dm6

B = (1.26... × 102 kg mol−1 ) ×
19.64 Pa g−1 dm3
7.466 Pa g−2 dm6
= (1.26... × 105 g mol−1 ) × = 4.79... × 104 dm3 mol−1
19.64 Pa g−1 dm3
The second virial coefficient is therefore B = 4.80 × 104 mol−1 dm3 .

5C Phase diagrams of binary systems: liquids

D5C.1 The principal factor is the shape of the two-phase liquid-vapour region in the
phase diagram (usually a temperature-composition diagram). The closer the
liquid and vapour lines are to each other, the more steps of the sort illustrated in
Fig. 5C.9 on page 161 are needed to move from a given mixed composition to an
acceptable enrichment in one of the components. However, the presence of an
azeotrope could prevent the desired degree of separation from being achieved.
D5C.3 A low-boiling azeotrope has a boiling temperature lower than that of either
component, so it is easier for the molecules to move into the vapour phase
than in a ‘normal’ (non-azeotropic) mixture. Therefore, the liquid phase has
less favorable intermolecular interactions than in a ‘normal’ mixture, a sign
that the components are less attracted to each other in the liquid phase than to
molecules of their own kind. These intermolecular interactions are determined
by factors such as dipole moment (polarity) and hydrogen bonding.
Conversely, a high-boiling azeotrope has a boiling temperature higher than that
of either component, so it is more difficult for the molecules to move into the
vapour phase. This reflects the relatively unusual situation of components that
have more favorable intermolecular interactions with each other in the liquid
phase than with molecules of their own kind.
E5C.1(a) The molar masses of phenol and water are 94.1074 g mol−1 and 18.0158 g mol−1 ,
respectively. The mole fraction of phenol (P) is
(7.32 g)/(94.1074 g mol−1 )

xP = = 0.149...
(7.32 g)/(94.1074 g mol−1 ) + (7.95 g)/(18.0158 g mol−1 )
Hence x P = 0.150 . Let the two phases be α (x P = 0.042) and β (x P = 0.161).

The proportions of these two phases, n β /n α is given by the level rule, [5C.6–160]
n β l α 0.149... − 0.042
= = = 9.68
n α l β 0.161 − 0.149...
The phenol-rich phase is more abundant by a factor of almost 10.
E5C.2(a) An approximate phase diagram is shown in Fig. 5.9; the given data points are
shown with dots and the curve is a quadratic which is a modest fit to these
points. The shape conforms to the expected phase diagram for such a system.
(i) A temperature of 23 ○ C is above the highest temperature at which par-

tial miscibility occurs, and therefore the expectation is that hexane and
perfluorohexane mix in all proportions to give a single phase.
(ii) At 22 ○ C the possibility of phase separation exists; as the mole fraction
of perfluorohexane increases the phase diagram is traversed along the
23.0
22.5
θ/○ C
22.0
21.5
21.0
0.0 0.2 0.4 0.6 0.8 1.0
x C6 F14
Figure 5.9
dashed line. When the mole fraction of perfluorohexane is low a single

phase forms, but as the mole fraction goes beyond 0.24 phase separa-
tion occurs. Initially, according to the lever rule, the proportion of the
perfluorohexane-rich phase is very small, but as more and more perfluo-
rohexane is added the proportion of this phase increases. When the mole
fraction is just under 0.48, there is very little of the perfluorohexane-poor
phase present, and as the mole fraction increases further a one-phase zone
is reached in which there is complete miscibility.
E5C.3(a) The temperature–composition phase diagram is a plot of the boiling point against
(1) composition of the liquid, x M and (2) composition of the vapour, y M . The
horizontal axis is labelled z M , which is interpreted as x M or y M according to
which set of data are being plotted. In addition to the data in the table, the
boiling points of the pure liquids are added. The plot is shown in Fig 5.10; in
this plot, the lines are best-fit polynomials of order 3.
θ/○ C xM yM θ/○ C xM yM
110.6 1 1 117.3 0.408 0.527
110.9 0.908 0.923 119.0 0.300 0.410
112.0 0.795 0.836 121.1 0.203 0.297
114.0 0.615 0.698 123.0 0.097 0.164
115.8 0.527 0.624 125.6 0 0
(i) The vapour composition corresponding to a liquid composition of x M =

0.250 is found by taking the vertical line at this composition up to the
intersection with the liquid curve, and then moving across horizontally
to the intersection with the vapour curve; occurs at y M = 0.354 , which
gives the composition of the vapour. The exact points of intersection can
be found either from the graph or by using the fitted functions.
125 liquid
vapour
120
θ/○ C
y M = 0.354
x M = 0.250
115
110
0.0 0.2 0.4 0.6 0.8 1.0
zM
Figure 5.10
(ii) A composition x O = 0.250 corresponds to x M = 0.750; from the graph

this corresponds to a vapour composition y M = 0.811 .
E5C.4(a) At the lowest temperature shown in the diagram the mixture is in the two-
phase region, and the two phases have composition of approximately x B =
0.88 and x B = 0.05. The level rule shows that there is about 9 times more of
the B-rich than of the B-poor phase. As the temperature is raised the B-rich
phase becomes slightly less rich in B, and the other phase becomes richer in B.
The lever rule implies that the proportion of the B-rich phase increases as the
temperature rises.
At temperature T1 the vertical line intersects the phase boundary. At this point
the B-poor phase disappears and only one phase, with x B = 0.8, is present.
P5C.1 If it is assumed that Raoult’s law applies, [5A.22–141], the partial vapour pres-
sures of benzene (B) and methylbenzene (M) are
p B = x B p∗B p M = x M p∗M
where x J are the mole fractions and p∗J are the vapour pressures over the pure
liquids. The total pressure is taken to be p tot = p B + p M .
The mole fraction in the vapour, y J , is related to the total pressure by p J = y J p tot ,
so it follows that
pJ x J p∗J
yJ = =
p tot p tot
Therefore
x B p∗B 0.75 × (75 Torr)
yB = = = 0.91
p tot 0.75 × (75 Torr) + 0.25 × (21 Torr)
x M p∗M 0.25 × (21 Torr)
yM = = = 0.085
p tot 0.75 × (75 Torr) + 0.25 × (21 Torr)
P5C.3 It is convenient to construct a pressure–composition phase diagram in order

to answer this question. If it is assumed that Raoult’s law applies, [5A.22–141],
the total pressure is computed from the sum of the partial vapour pressures of
benzene (B) and methylbenzene (M)
p tot = p B + p M = x B p∗B + x M p∗M
where x J are the mole fractions and p∗J are the vapour pressures over the pure
liquids. This equation is used to construct the liquid line on the graph shown
in Fig. 5.11, where z B is interpreted as x B .
The mole fraction in the vapour, y J , is related to the total pressure by p J = y J p tot .
Using this it can be shown that the total pressure in terms of the mole fraction
in the vapour in given by [5C.5–157],
p∗B p∗M
p tot =
p∗B + (p∗M − p∗B )y B
This equation is used to construct the vapour line on the phase diagram, where
z B is interpreted as y B .
10.0 liquid
vapour
8.0
p tot /kPa
a1 a ′1
6.0
a ′′2 ′
a2 a2
a′′3
4.0 a3
2.0
0.0 0.2 0.4 0.6 0.8 1.0
zB
Figure 5.11
(a) A mixture with equal amounts of B and M has mole fractions x B = x M =

1
2
. The total pressure is therefore
p tot = x B p∗B + x M p∗M = 12 (9.9 kPa) + 12 (2.9 kPa) = 6.4 kPa
This is the pressure at which boiling first occurs, point a 1 in the diagram.
(b) The composition of the vapour is given by
x B p∗B 2 (9.9 kPa)

1
pB
yB = = = = 0.773... = 0.77
p tot p tot 6.4 kPa
and therefore y M = 1 − 0.773... = 0.23 . This is point a ′1 on the diagram:

the lever-rule also indicates that the fraction of the phase with composi-
tion a′1 (the vapour) is very small.
(c) As the pressure is reduced further, say to point a 2 , the tie line indicates
that the liquid will have composition a ′′2 and the vapour will have com-
position a′2 , the latter being richer in the more volatile component B.
The level rule indicates that the proportion of the vapour phase is now
significant.
The process continues until point a 3 is reached. At this pressure the com-
position of the liquid is given by point a ′′3 , and the level rule indicates that
the proportion of the liquid phase is very small. It is also evident from the
diagram that at point a 3 the vapour composition is y B = y M = 21 , therefore
p B = 21 p tot and p M = 21 p tot . Raoult’s law gives the partial vapour pressures
of B and M are p B = p∗B x B and p M = p∗M x M . It follows that
1
p
2 tot
= x B p∗B and 1
p
2 tot
= x M p∗M = (1 − x B )p∗M
These two equations are combined to give
p∗M (2.9 kPa)

xB = = = 0.226... = 0.23
p∗B + p∗M (9.9 kPa) + (2.9 kPa)
and x M = 1 − x B = 1 − 0.226... = 0.77 . This is point a ′′3 on the phase

diagram. The vapour pressure of a mixture with this composition is
p tot = x B p∗B + x M p∗M = 0.226... × (9.9 kPa) + (1 − 0.226...) × (2.9 kPa)

= 4.5 kPa
P5C.5 The annotated phase diagrams are shown in Fig. 5.12. Given that the normal
boiling point of hexane is certainly lower than that of heptane the horizontal
scale should presumably be mole fraction of heptane; however, the solution
provided follows the labelling of the diagram in the text.
(a) The phases present are indicated on the diagrams above.

(b) For an equimolar mixture x H = 0.5; the vertical line at this composition
intersects the liquid line at a, and reading across the pressure is 625 Torr .
(c) At this pressure the composition of the vapour is given by the intersection
of the horizontal line with the vapour line, which occurs at point d. The
vapour is less rich in hexane than the liquid. As the solution continues
to evapourate the composition of the liquid moves along the liquid line
to point c. This is at the pressure at which the composition of the vapour
matches the original composition of the liquid (x H = 0.5, point b). The
composition of the liquid is read off the scale as x H = 0.7 , and the pres-
sure is 500 Torr .
(d) From part (b) the composition of the liquid is x H = 0.5 , and the compo-
sition of the vapour is read off from where the horizontal line at 625 Torr
intersects the vapour curve, point d, which is at y H = 0.3 .
(e) From part (c) the composition of the vapour is y H = 0.5 and the com-
position of the liquid can be read off from where the horizontal line at
500 Torr intersects the liquid curve, point c, which is at x H = 0.7 .
900
liquid line
70°C
700
liquid
Pressure, p /Torr
d a
625
500 c 500
b
300 vapour vapour line
0 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1

Mole fraction of hexane, z H
100
vapour line
vapour
Temperature, θ /°C
90
f e g
v l
80
liquid
70
liquid line 760 Torr

60
0 0.2 0.4 0.6 0.8 1

0.44 0.74
Figure 5.12 Mole fraction of hexane, zH
(f) Refer to the temperature–composition phase diagram; the stated compo-

sition is z heptane = 0.40 which corresponds to z H = 0.60. The vertical line
at z H = 0.60 intersects the horizontal line at 85 ○ C at point e. From the tie
line the composition of the vapour is read off from point f, y H = 0.44; the
composition of the liquid is read off from point g, y H = 0.74. From the
lever rule
n l v 0.60 − 0.44
= = = 1.1
n v l 0.74 − 0.60
The two phases are roughly equally abundant.
P5C.7 The relationship between y A and x A is given in [5C.4–156]

x A p∗A x A (p∗A /p∗B )
yA = =
p∗B + (p∗A − p∗B )x A 1 + (p∗A /p∗B − 1)x A
The form of the function on the right gives y A as a function of x A and the ratio
(p∗A /p∗B ) as required. The plot if shown in Fig. 5.13
P5C.9 If the excess enthalpy is modelled as H E = ξRTx A2 x B2 then, by anaolgy with
[5B.7–147], the expression for for Gibbs energy of mixing is
∆ mix G = nRT (x A ln x A + x B ln x B + ξx A2 x B2 )
The minima and maxima in this function are located by setting the derivative
with respect to x A to zero; it is convenient to take the derivative of ∆ mix G/nRT
1.0
0.8
0.6
yA 0.4
(p∗A /p∗B ) = 1
(p∗A /p∗B ) = 4
0.2
(p∗A /p∗B ) = 50
0.0
0.0 0.2 0.4 0.6 0.8 1.0
xA
Figure 5.13
and before doing this x B is replaced by (1 − x A )
∆ mix G/nRT = x A ln x A + (1 − x A ) ln(1 − x A ) + ξx A2 (1 − x A )2

= x A ln x A + ln(1 − x A ) − x A ln(1 − x A ) + ξx A2 (1 − x A )2
d(∆ mix G/nRT) 1 xA
= 1 + ln x A − + − ln(1 − x A )
dx A 1 − xA 1 − xA
+ 2ξx A (1 − x A )2 − 2ξx A2 (1 − x A )
1 − xA − 1 + xA xA
= + ln + 2ξx A (1 − x A )(1 − 2x A )
1 − xA 1 − xA
xA
= ln + 2ξx A (1 − x A )(1 − 2x A )
1 − xA
As before, the derivative is zero at x A = 0.5 for all values of ξ; this corresponds
either to a minimum when ξ is small, or to a maximum when ξ is sufficiently
large. Qualitatively the behaviour is similar to that shown in Fig. 5B.5.
Apart from this solution at x A = 0.5, there are no analytical solutions for when
this derivative is zero. However, solutions can be found by graphically by look-
ing for the intersection between ln (x A /[1 − x A ]) and −2ξx A (1− x A )(1−2x A ).
This is done with the aid of Fig. 5.3b. From the graph it is evident that for ξ = 1
there are no values of x A at which the curves intersect, and so no minima, but
at sufficiently high values of ξ (such as ξ = 6) such intersections do occur and
lead to two minima.
Overall, as is seen in Fig. 5.14, the behaviour is qualitatively similar to that

for H E = ξRTx A x B . If ξ is below some particular positive value ∆ mix G is
always negative and shows a single minimum at x A = 0.5. Above some critical
value, ∆ mix G may become positive for some values of x A , and the plot shows
a maximum at x A = 0.5, flanked symmetically by two minima. This indicates
that phase separation will occur.
ξ=0
ξ=3
∆ mix G/nRT 0.0 ξ=6
ξ = 15
−0.5
0.0 0.2 0.4 0.6 0.8 1.0

xA
Figure 5.14
5D Phase diagrams of binary systems: solids

D5D.1 The schematic phase diagram is shown in Fig. 5.15. Incongruent melting means
that the compound AB2 does not occur in the liquid phase.
liquid
A(s) + B(l)
+ A(l) AB2(s),B(l)
B(l) + A(l)
A(l)
Temperature
+ B(s)
AB2(s),B(l)
A(l)
A(s) + AB2(s)
AB2(s) + B(s)
0 0.2 0.4 0.6 0.8 1

Figure 5.15 Mole fraction of B, xB AB2
E5D.1(a) The feature that indicates incongruent melting is the intersection of the two
liquid curves at around x B = 0.6. The incongruent melting point is marked as
T1 ≈ 350 ○ C. The composition of the eutectic is x B ≈ 0.25 and its melting point
is labelled T2 ≈ 190 ○ C .
E5D.2(a) The cooling curves are shown in Fig 5.16; the break points are shown by the
short horizontal lines. For isopleth a the first break point is at 380 ○ C where
the isopleth crosses the liquid curve, there is a second break point where the
isopleth crosses the boundary at T1 ; there is then a eutectic halt at 190 ○ C. For
isopleth b the first break point is at 450 ○ C where the isopleth crosses the liquid
curve, there is a second break point where the isopleth crosses the boundary at
T1 , and then a eutectic halt at 190 ○ C.
500
400
T/K
300
a b
200
time
Figure 5.16
E5D.3(a) Figure 5.17 shows the relevant phase diagram to which dotted horizontal lines
have been added at the relevant temperatures.
1000
c1
800
Temperature, θ/°C
Liquid
600
c2
Ag 3Sn
400
c3
200
0
0 20 40 60 80 100
Figure 5.17 Mass percentage of Ag/%
(i) The solubility of silver in tin at 800 ○ C is determined by the point c 1 . (At
higher proportions of silver, the system separates into two phases, a liquid
and a solid phase rich in silver.) The point c 1 corresponds to 76% silver
by mass.
(ii) The compound Ag3 Sn decomposes at this temperature. Three phases are
in equilibrium here: a liquid containing atomic Ag and Sn about 52% Ag
by mass; a solid solution of Ag3 Sn in Ag; and solid Ag3 Sn. See point c 2 .
(iii) At point c 3 , two phases coexist: solid Ag3 Sn and a solid solution of the
compound and metallic silver. Because this point is close to the Ag3 Sn
composition, the solid solution is mainly Ag3 Sn, at least when measured
in mass terms. The composition of the solid solution is expressed as a ratio
of moles of compound (n c ) to moles of atomic silver (n a ). These quantities
are related to the silver mass fraction c Ag by employing the definition of
mass fraction, namely the mass of silver (from the compound and from
atomic silver) over the total sample mass
m Ag (3n c + n a )M Ag
c Ag = =
m Ag + m Sn (3n c + n a )M Ag + n c M Sn
This relationship is rearranged, collecting terms in n c on one side and n a

on the other
n c [3M Ag (c Ag − 1) + M Sn c Ag ] = n a M Ag (1 − c Ag )
The mole ratio of compound to atomic silver is given by
nc M Ag (1 − c Ag )
=
n a 3M Ag (c Ag − 1) + M Sn c Ag
At 460 ○ C, c Ag = 0.78 (point c 3 on the coexistence curve), so
nc (107.9 g mol−1 ) × (1 − 0.78)

= = 1.11
n a 3 × (107.9 g mol−1 ) × (0.78 − 1) + (118.7 g mol−1 ) × 0.78
At 300 ○ C, c Ag = 0.77 (point c 2 on the coexistence curve), so
nc (107.9 g mol−1 ) × (1 − 0.77)

= = 1.46
n a 3 × (107.9 g mol−1 ) × (0.77 − 1) + (118.7 g mol−1 ) × 0.77
E5D.4(a) The schematic phase diagram is shown in Fig 5.18. The solid points are the data
given in the Exercise, and the lines are simply plausible connections between
these points; it is assumed that the compound in 1:1. Note that to the right of
x B = 0.5 the solids are AB and B, whereas to the left of this composition the
solids are AB and A.
130 A(s) AB(s)

+AB(l) B(s)
+A(l)
+A(l) +AB(l)
+AB(l)
+B(l)
120
T/K
AB(s)+AB(l)
+B(l)
110 A(s)+AB(s)
B(s)+AB(s)
100
0.0 0.2 0.4 0.6 0.8 1.0
xB
Figure 5.18
(i) (ii) (iii) liquid

(iv) (v)
94
92
90
two-phase liquid
T/K
88
86
84 CF4 (s)+CH4 (s)
82
0.0 0.2 0.4 0.6 0.8 1.0
x CF4
Figure 5.19
E5D.5(a) The schematic phase diagram is shown in Fig 5.19. The solid points are the data
given in the Exercise, and lines are simply plausible connections between these
points. (The dash-dotted lines are referred in to Exercise E5D.6(a).)
E5D.6(a) The compositions at which the cooling curves are plotted are indicated by the
vertical dash-dotted lines on the phase diagram for Exercise E5D.5(a), Fig. 5.19.
The cooling curves are shown in Fig 5.20. The break points, where solid phases
start to form are shown by the short horizontal lines, and the dotted lines in-
dicate the temperatures of the two eutectics (86 K and 84 K). The horizontal
segments correspond to solidification of a eutectic.
94
92
90 (i) (ii) (iii) (iv) (v)
T/K
88
86
84
82
time
Figure 5.20
P5D.1 The schematic phase diagram is shown in Fig 5.21. The solid points are the
data given in the Exercise, and in the absence of any further information and
because there are so few points, these have just been joined by straight lines.
1,000 solid
liquid
θ/○ C
900
800
0.0 0.2 0.4 0.6 0.8 1.0

x ZrF4
Figure 5.21
On cooling a liquid with composition x ZrF4 = 0.4 solid first starts to appear
when the isopleth intersects the liquid line (at about 870 ○ C). The composition
of the small amount of solid that forms is given by the left-hand open circle in
the diagram (about x = 0.29; containing less ZrF4 than the liquid). As the
temperature drops further more solid is formed and its composition moves
along the solid line to the right becoming richer in ZrF4 until it reaches the
point where the isopleth crosses the solid line. At this point what remains of
the liquid has composition given by the right-hand open circle (about x = 0.48).
A further drop in temperature results in complete solidification.
P5D.3 The phase diagram is shown in Fig. 5.22, along with the relevant cooling curves.
The fact that there is a phase boundary indicated by the vertical line at x B = 0.67
is taken to indicate the formation of compound AB2 which has x B = 23 . By
analogy with the phase diagram shown in Fig. 5D.5 on page 169, the form of
the given phase diagram indicates that AB2 does not exist in the liquid phase.
The number of distinct chemical species (as opposed to components) and phases
present at the indicated points are, respectively
b(3, 2), d(2, 2), e(4, 3), f (4, 3), g(4, 3), k(2, 2)
Liquid A and solid A are considered to be distinct species.
P5D.5 (a) Note that, as indicated on the diagram, Ca2 Si, CaSi, CaSi2 appear at mole
fractions of Si 13 , 12 and 23 , as expected.
eutectics: x Si = 0.056 at approximately 800 ○ C, x Si = 0.402 at 1268 ○ C,
x Si = 0.694 at 1030 ○ C
congruent melting compounds: Ca2 Si Tf = 1314 ○ C, CaSi Tf = 1324 ○ C
incongruent melting compound: CaSi2 Tf = 1040 ○ C (melts into CaSi(s)
and Si-rich liquid with x Si around 0.69)
(b) For an isopleth at x Si = 0.2 and at 1000 ○ C the phases in equilibrium are
CaSi2 and a Ca-rich liquid (x Si = 0.11). The lever rule, [5C.6–160], gives
A(l)+B(l)+B(s)
Temperature, T
f
A(l)+B(l)
+A(s) g
c A(l)+B(l)+AB2(s)
k
e
d AB2(s)+A(s) AB2(s)+B(s)
0.23
0.57
0.67
0.84
0.16
0 0.2 0.4 0.6 0.8 1
Mole fraction of B, xB
xB = 0.16 xB = 0.57 xB = 0.84
xB = 0.23
xB = 0.67
Figure 5.22
the relative amounts:

n Ca2 Si l liq 0.2 − 0.11
= = = 0.7
n liq l Ca2 Si 0.333 − 0.2
(c) For an isopleth at x Si = 0.8 Si(s) begins to appear at about 1300 ○ C. Fur-
ther cooling causes more Si(s) to freeze out of the melt so that the melt
becomes more concentrated in Ca. There is a eutectic at x Si = 0.694 and
1030 ○ C.
At a temperature just above the eutectic point the liquid has composition
x Si = 0.694 and the lever rule gives that the relative amounts of the Si(s)
and liquid phases as:
n Si l liq 0.80 − 0.694

= = = 0.53
n liq l Si 1.0 − 0.80
At the eutectic temperature a third phase appears, CaSi2 (s). As the melt
cools at this temperature, both Si(s) and CaSi2 (s) freeze out of the melt
while the composition of the melt remains constant. At a temperature
slightly below the eutectic point all the melt will have frozen to Si(s) and
CaSi2 (s) with the relative amounts:
n Si l CaSi2 0.80 − 0.667
= = = 0.67
n CaSi2 l Si 1.0 − 0.80
P5D.7 The data are plotted as the phase diagram shown in Fig 5.23; the filled and open
circles are the data points and the solid/dashed line is a best-fit cubic function.
A mixture of 0.750 mol of N,N-dimethylacetamide with 0.250 mol of heptane
has mole fraction of the former of x 1 = 0.750/(0.750 + 0.250) = 0.750. The
tie line at 296.0 ○ C is shown on the diagram, and this intersects with the two
curves at x 1 = 0.167 and x 2 = 0.805 (determined from the best-fit polynomial -
310 x1
x2
305
T/K
300
295
0.0 0.2 0.4 0.6 0.8 1.0

x 1 or x 2
Figure 5.23
an alternative would be to use the data points given for this temperature). The
lever rule, [5C.6–160] gives the proportion of the two phases as
n x=0.805 0.750 − 0.167
= = 10.6
n x=0.167 0.805 − 0.750
The N,N-dimethylacetamide-rich phase is therefore more than ten times more
abundant than the other phase.
A mixture of this composition will become a single phase at the temperature at
which the x 1 = 0.750 isopleth intersects the right-hand phase boundary. Using
the fitted function, this intersection is at 302.5 ○ C .
5E Phase diagrams of ternary systems

D5E.1 The composition represented by point c is approximately x Ni = 0.73, x Fe = 0.20,
x Cr = 0.07. This is a three-phase region, with Fe, Ni and γFeNi present.
D5E.3 The lever rule, [5C.6–160], applies in a ternary system, but with an important
caveat. For binary systems the tie lines to which the rule appplies are always
horizontal and so can be added to the phase diagram at will. In contrast, for
a ternary system the tie lines have no such simple orientation and have to be
determined experimentally. Thus the lever rule applies, but in order to use
it additional information is needed about the tie lines at the composition of
interest.
E5E.1(a) The points corresponding to the given compositions are marked with letters on
the phase diagram shown in Fig. 5.24. Composition (i) is in a two-phase region,
(ii) is in a three-phase region, (iii) is in a region where there is only one phase.
Composition (iv) corresponds to the point at which the phase boundaries meet
and all of the phases are in equilibrium
0 1 H2O
S1
0.2 P = 1 0.8
iii S2
0.4 0.6
P=2 P=2
0.6 i 0.4
iv
0.8 P=3 ii 0.2
NH4Cl 1 0 (NH4)2SO4
0 0.2 0.4 0.6 0.8 1
Figure 5.24
E5E.2(a) (i) The phase equilibrium between NH4 Cl and H2 O is indicated by the left-
hand edge of the phase diagram shown in Fig. 5.24. At the point S 1 the
system moves from two phases to one – in other words from a system of
solid NH4 Cl in equilibrium with a solution of NH4 Cl in H2 O, to a system
in which there is just one phase, a solution of NH4 Cl in H2 O. This point
therefore marks the solubility of NH4 Cl, and from the diagram is occurs
at x NH4 Cl = 0.19.
The task is to convert this mole fraction to a molar concentration. Imagine
a solution made from a mass m of H2 O: the amount in moles of H2 O is
m/M, where M is the molar mass of H2 O. The mole fraction of NH4 Cl
is therefore
n NH4 Cl n NH4 Cl
x NH4 Cl = =
n NH4 Cl + n H2 O n NH4 Cl + m/M
This rearranges to n NH4 Cl = x NH4 Cl (m/M)/(1 − x NH4 Cl ). If it is assumed
that the mass density of the solution is approximately the same as the mass
density of water, ρ, the volume of this solution is given by V = m/ρ. With
this, the molar concentration of NH4 Cl is computed as
n NH4 Cl 1 x NH4 Cl (m/M) ρ x NH4 Cl (m/M)
[NH4 Cl] = = =
V V 1 − x NH4 Cl m 1 − x NH4 Cl
x NH4 Cl ρ
=
M(1 − x NH4 Cl )
With the data from the diagram and assuming ρ = 1000 g dm−3
x NH4 Cl ρ 0.19 × (1000 g dm−3 )
[NH4 Cl] = =
M(1 − x NH4 Cl ) (18.0158 g mol−1 ) × (1 − 0.19)
= 13 mol dm−3
This high concentration rather casts doubt on assuming the density of the
solution is the same as that of water.
(ii) The solubility of (NH4 )2 SO4 is indicated by point S 2 on the right-hand

edge, at x(NH4 )2 SO4 = 0.3. An analogous calculation to that in part (a)
gives the concentration as [(NH4 )2 SO4 ] = 24 mol dm−3 .
E5E.3(a) The ternary phase diagram is shown in Fig 5.25.
A
0.0
1.0
0.2
0.8
xC 0.4 xA
0.6
0.6
0.4
(iii)
0.8
(i) 0.2
1.0 (ii)
C 0.0 B
0.0 0.2 0.4 0.6 0.8 1.0
xB
Figure 5.25
E5E.4(a) The composition by mass needs to be converted to mole fractions, which re-
quires the molar masses: M NaCl = 58.44 g mol−1 , M H2 O = 18.016 g mol−1 , and
M Na2 SO4 ⋅ 10 H2 O = 322.20 g mol−1 . Imagine that the solution contains 25 g NaCl,
25 g Na2 SO4 ⋅ 10 H2 O and hence (100 − 25 − 25) = 50 g H2 O. The mole fraction
of NaCl is
m NaCl /M NaCl
x NaCl =
m NaCl /M NaCl + m Na2 SO4 ⋅ 10 H2 O /M Na2 SO4 ⋅ 10 H2 O + m H2 O /M H2 O
(25 g)/(58.44 g mol−1 )
=
(25 g)/(58.44 g mol ) + (25 g)/(322.20 g mol−1 ) + (50 g)/(18.016 g mol−1 )
−1
= 0.13
Likewise, x Na2 SO4 ⋅ 10 H2 O = 0.024 and x H2 O = 0.85; this point is plotted in the
ternary phase diagram shown in Fig 5.26.
The line with varying amounts of water but the same relative amounts of the
two salts (in this case, equal by mass), passes through this point and the vertex
corresponding to x H2 O = 1. This line intersects the NaCl axis at a mole fraction
corresponding to a 50:50 mixture (by mass) of the two salts
(50 g)/(58.44 g mol−1 )

x NaCl = = 0.85
(50 g)/(58.44 g mol−1 ) + (50 g)/(322.20 g mol−1 )
The line is shown on the diagram.

NaCl
0.0
1.0
0.2
0.8
x H2 O 0.4 x NaCl
0.6
0.6
0.4
0.8
0.2
1.0
H2 O 0.0 Na2 SO4
0.0 0.2 0.4 0.6 0.8 1.0
x Na2 SO4 ⋅ 10 H2 O
Figure 5.26
E5E.5(a) The composition by mass needs to be converted to mole fractions, which re-
quires the molar masses: M CHCl3 = 119.37 g mol−1 , M H2 O = 18.016 g mol−1 ,
and M CH3 COOH = 60.052 g mol−1 . The mole fraction of CHCl3 is
m CHCl3 /M CHCl3
x CHCl3 =
m CHCl3 /M CHCl3 + m CH3 COOH /M CH3 COOH + m H2 O /M H2 O
9.2 g
119.37 g mol−1
= 9.2 g 3.1 g 2.3 g = 0.30
119.37 g mol−1 60.052 g mol−1 18.016 g mol−1
+ +
Likewise, x CH3 COOH = 0.20 and x H2 O = 0.50 . This point in marked with the
open circle on the phase diagram shown in Fig. 5.27; it falls clearly in the two-
phase region. The point almost lies on the tie line a ′2 –a ′′2 , and using this as a
guide the lever rule indicates that the phase with composition a′2 , (x W = 0.57,
x T = 0.20, x E = 0.23), is approximately 5 times more abundant than the phase
with composition a′′2 , (x W = 0.06, x T = 0.82, x E = 0.12).
(i) When water is added to the mixture the composition moves along the
dashed line to the lower-left corner. The system will pass from the two-
phase to the one-phase region when the line crosses the phase boundary,
which is at approximately (x W = 0.75, x T = 0.14, x E = 0.10).
(ii) When ethanoic acid is added to the mixture the composition moves along
the dashed line to the vertex. The system will pass from the two-phase to
the one-phase region when the line crosses the phase boundary, which is
at approximately (x W = 0.44, x T = 0.26, x E = 0.30).
P5E.1 The given points are shown by filled dots in the phase diagram shown in Fig. 5.28,
and they are connected by a straight line, as indicated in the problem. Beneath
CH3 COOH
0.0
1.0
0.2
0.8
xW 0.4 xE
0.6
0.6
0.4
0.8 a′2
0.2
1.0 a ′′2
H2 O 0.0 CHCl3
0.0 0.2 0.4 0.6 0.8 1.0
xT
Figure 5.27
this line lies a one-phase region because CO2 and nitrobenzene are miscible
in all proportions. Addition of I2 eventually causes phase separation into a
two-phase region for all compositions about the line.
0 1 I2
0.2 0.8
0.4 0.6
0.6 P=2 0.4
0.8 0.2
P=1
CO2 1 0 nitrobenzene
0 0.2 0.4 0.6 0.8 1
Figure 5.28
P5E.3 Consider the construction shown in Fig. 5.29. The line is interest is AW, where
as indicated the position of W is determined by the mole fractions of B and
C in the binary mixture; to avoid confusing these particular mole fractions
with others, they are denoted y B and y C . The point P lies on this line and its
perpendicular distance from each of the edges of the triangle gives the mole
fractions of B and C, x B and x C .
Construct the line UV passing through P and parallel to the base of the triangle.
It follows that AWB and APV are similar triangles, therefore there exists the
following relationship between the ratios of the sides.
WB PV
= (5.3)
AW AP
N
M xC
xB
U V
P
C B
yB W yC
Figure 5.29
The indicated angle is 60○ , so it follows that x C = PV sin 60○ .

Now consider the triangles AWC and APU, which are also similar and therefore
WC PU
= , (5.4)
AW AP
and as before x B = PU sin 60○ .
Dividing eqn 5.3 by eqn 5.4 gives
WB PV y C PV
= hence =
WC PU y B PU
Dividing x C = PV sin 60○ by x B = PU sin 60○ gives x C /x B = PV/PU. Equating

the two expressions for PV/PU gives the required result
yC xC
=
yB xB
Alternatively (and avoiding the use of the angle explicitly) PNV and PMU are
also recognised as similar triangles, so that it follows directly that x C /PV =
x B /PU.
5F Activities
D5F.1 The Debye–Hückel theory of electrolyte solutions formulates deviations from
ideal behaviour (essentially, deviations due to electrostatic interactions between
the ions) in terms of the work of charging the ions. The assumption is that
the solute particles would behave ideally if they were not charged, and the
difference in chemical potential between real and ideal behaviour amounts to
the work of putting electrical charges onto the ions.
To find the work of charging, the distribution of ions must be found, and that is
done using the shielded Coulomb potential which takes into account the ionic
strength of the solution and the dielectric constant of the solvent. The Debye–
Hückel limiting law, [5F.27–178], relates the mean ionic activity coefficient to
the charges of the ions involved, the ionic strength of the solution, and depends
on a constant that takes into account solvent properties and temperature.
D5F.3 If a solvent or solute has a certain chemical potential, then this is related to
the activity of the solvent or solute through [5F.1–173] and [5F.9–174]. At first
sight it seems odd to think of the chemical potential determining the activity,
but this approach is in fact logical as the chemical potential is the experimen-
tally measurable quantity. For example, measurements of cell potentials or the
vapour pressure of liquids provide (slightly indirect, it must be admitted) ways
of measuring the chemical potential.
For ideal systems expressions are available which relate the chemical potential
to the concentration (in the form of its various measures, such as mole fraction
or molality). It therefore follows that the difference between the measured
chemical potential and that predicted from idealised models is attributable to
factors not taken into account in the ideal systems. Such factors are collectively
described as non-ideal interactions; the difference between activity and con-
centration is therefore ascribed to the presence of such interactions
Non-ideal interactions are no different from the usual interactions between
molecular species. For example, they may include interactions between charged
or polar species, hydrogen bonding or more specific interactions.
D5F.5 The main way of measuring activities described in this Topic is from measure-
ments of partial vapour pressures, as given in [5F.2–173] for the solvent and in
[5F.10–174] for the solute activity. Other measurements from which the value
of the chemical potential can be inferred (for example, cell potentials) are used
to determine activities via the general relationship µ J = µ −J○ + RT ln a J .
E5F.1(a) Ionic strength is defined in [5F.28–178]
I = 21 ∑ z 2i (b i /b −○ )
i
where the sum runs over all the ions in the solution, z i is the charge number
on ion i, and b i is its molality. For KCl the molality of K+ and Cl – are both
0.10 mol kg−1 ; z K+ = +1 and z Cl− = −1. For CuSO4 the molality of Cu2+ and
SO4 2 – are both 0.20 mol kg−1 ; z Cu2+ = +2 and z SO4 2− = −2. The ionic strength
is therefore
I = 12 [1/(1 mol kg−1 )] [(+1)2 × (0.10 mol kg−1 ) + (−1)2 × (0.10 mol kg−1 )
+(+2)2 × (0.20 mol kg−1 ) + (−2)2 × (0.20 mol kg−1 )] = 0.9
E5F.2(a) Ionic strength is defined in [5F.28–178]
I = 12 ∑ z 2i (b i /b −○ )
i
where the sum runs over all the ions in the solution, z i is the charge number
on ion i, and b i is its molality.
(i) The aim here is to increase the ionic strength by 0.250−0.150 = 0.100; the
task is therefore to compute the mass m of Ca(NO3 )2 which, when added
to a mass m w of water, gives this increase in the ionic strength.
A solution of Ca(NO3 )2 of molality b contributes Ca2+ at molality b and
NO3 – at molality 2b. The contribution to the ionic strength is therefore
1
2
[(+2)2 × b + (−1)2 × 2b]/b −○ = 3b/b−○ .
The molality arising from dissolving mass m of Ca(NO3 )2 in a mass m w
of solvent is (m/M)/m w , where M is the molar mass. It therefore follows
that to achieve the desired increase in ionic strength
m 1
3× × = 0.100 hence m= 1
× 0.100 × Mm w b −○
Mm w b−○ 3
The molar mass of Ca(NO3 )2 is 164.10 g mol−1 ; using this, and recalling
that the molality is expressed in mol kg−1 , gives
m= 1
3
× 0.100 × (164.10 × 10−3 kg mol−1 ) × (0.500 kg) × (1 mol kg−1 )
= 2.73... × 10−3 kg
Hence the 2.74 g of Ca(NO3 )2 needs to be added to achieve the desired

ionic strength.
(ii) The argument is as in (a) except that the added solute is now NaCl which
contributes singly-charged ions at the same molality as the solute so the
contribution to the ionic strength is simply b/b−○ . It therefore follows that
to achieve the desired increase in ionic strength
m 1
× = 0.100 hence m = 0.100 × Mm w b −○
Mm w b −○
The molar mass of NaCl is 58.44 g mol−1 , hence
m = 0.100 × (58.44 × 10−3 kg mol−1 ) × (0.500 kg) × (1 mol kg−1 )

= 2.92... × 10−3 kg
Hence the 2.92 g of NaCl needs to be added to achieve the desired ionic
strength.
E5F.3(a) The Debye–Hückel limiting law, [5F.27–178], is used to estimate the mean ac-
tivity coefficient, γ± , at 25 ○ C in water
log γ± = −0.509 ∣z+ z− ∣ I 1/2 I = 12 ∑ z 2i (b i /b −○ )

i
where z± are the charge numbers on the ions from the salt of interest and I is
the ionic strength, defined in [5F.28–178]. In the definition of I the sum runs
over all the ions in the solution, z i is the charge number on ion i, and b i is its
molality.
A solution of CaCl2 of molality b contributes Ca2+ at molality b and Cl – at

molality 2b. The contribution to the ionic strength is therefore 12 [(+2)2 × b +
(−1)2 × 2b]/b −○ = 3b/b −○ . A solution of NaF of molality b ′ contributes Na+
at molality b′ and F – at molality b ′ . The contribution to the ionic strength is
therefore 12 [(+1)2 × b ′ + (−1)2 × b ′ ]/b −○ = b ′ /b −○ . The ionic strength of the
solution is therefore
(3b+b ′ )/b −○ = [3×(0.010 mol kg−1 )+1×(0.030 mol kg−1 )]/(1 mol kg−1 ) = 0.060
For solute CaCl2 z+ = +2 and z− = −1 so the limiting law evaluates as
log γ± = −0.509 ∣z+ z− ∣ I 1/2 = −0.509 ∣(+2) × (−1)∣ (0.060)1/2 = −0.249...

The mean activity coefficient is therefore γ± = 10−0.249 ... = 0.563... = 0.56 .
E5F.4(a) The Davies equation is given in [5F.30b–179]
−A ∣z+ z− ∣ I 1/2
log γ± = + CI
1 + BI 1/2
Because the electrolyte is 1:1 with univalent ions, the ionic strength is simply
I = b HBr /b −○ . There is no obvious straight-line plot using which the data can
be tested against the Davies equation, therefore a non-linear fit is made us-
ing mathematical software and assuming that A = 0.509; remember that the
molalities must be expressed in mol kg−1 . The best-fit values are B = 1.96 and
C = 0.0183; With these values the predicted activity coefficients are 0.930, 0.907
and 0.879, which is very good agreement.
E5F.5(a) The activity in terms of the vapour pressure p is given by [5F.2–173], a = p/p∗ ,
where p∗ is the vapour pressure of the pure solvent. With the data given a =
p/p∗ = (1.381 kPa)/(2.3393 kPa) = 0.5903 .
E5F.6(a) On the basis of Raoult’s law, the activity in terms of the vapour pressure p A
is given by [5F.2–173], a A = p A /p∗A , where p∗A is the vapour pressure of the
pure solvent. With the data given a A = p A /p∗A = (250 Torr)/(300 Torr) =
0.833... = 0.833 . The activity coefficient is defined through [5F.4–173], a A =
γ A x A , therefore γ A = a A /x A = 0.833.../0.900 = 0.926 .
For the solute, Henry’s law is used as the basis and the activity is given by [5F.10–
174], a B = p B /K B , where K B is the Henry’s law constant expressed in terms of
mole fraction. In this case a B = p B /K B = (25 Torr/200 Torr) = 0.125 .
E5F.7(a) On the basis of Raoult’s law, the activity in terms of the vapour pressure p J is
given by [5F.2–173], a J = p J /p∗J , where p∗J is the vapour pressure of the pure
solvent. The partial vapour pressure of component J in the gas is given by p J =
y J p tot . In this case
p P y P p tot 0.516 × (1.00 atm) × [(101.325 kPa)/(1 atm)]
aP = = = = 0.497...
p∗P p∗P 105 kPa
The activity of propanone is therefore a P = 0.498 . The activity coefficient is
defined through [5F.4–173], a J = γ J x J , therefore γ P = a P /x P = 0.497.../0.400 =
1.24 .
For the other component the mole fractions are y M = 1 − y P = 0.484 and
x M = 1 − x P = 0.600. The rest of the calculation follows as before
p M y M p tot 0.484 × (1.00 atm) × [(101.325 kPa)/(1 atm)]

aM = = = = 0.667...
p∗M p∗M 73.5 kPa
The activity of methanol is therefore a M = 0.667 and its activity coefficient is

given by γ M = a M /x M = 0.667.../0.600 = 1.11 .
E5F.8(a) For this model of non-ideal solutions the vapour pressures are given by [5F.18–
176], p A = p∗A x A exp(ξ[1 − x A ]2 ) and likewise for p B ; the total pressure is given
by p tot = p A + p B . The vapour pressures are plotted in Fig. 5.30.
20
pA
pB
15 p tot
p/kPa
10
0
0.0 0.2 0.4 0.6 0.8 1.0
xA
Figure 5.30
P5F.1 In the Raoult’s law basis the activity is given by [5F.2–173], a J = p J /p∗J , and the
activity coefficient by [5F.4–173], a J = γ J x J . The partial pressure of the vapour
is given by p J = y J p, where y J is the mole fraction of J in the vapour and p the
total pressure.
It follows that γ J = a J /x J = y J p/x J p∗J . In this binary system the activity coeffi-
cient for 1,2-epoxybutane (E) is found using the given data for trichloromethane
(T) by using and x E = 1 − x T , and likewise for y E . It follows that γ E = (1 −
y T )p/(1 − x T )p∗E . The resulting activity coefficients are shown in the table.
p/kPa xT yT γT γE
23.40 0 0 1
21.75 0.129 0.065 0.417 0.998
20.25 0.228 0.145 0.490 0.958
18.75 0.353 0.285 0.576 0.885
18.15 0.511 0.535 0.723 0.738
20.25 0.700 0.805 0.885 0.563
22.50 0.810 0.915 0.966 0.430
26.30 1 1 1
P5F.3 The Debye–Hückel limiting law, [5F.27–178], is
log γ± = −0.509 ∣z+ z− ∣ I 1/2 I = 12 ∑ z 2i (b i /b −○ )

i
where z± are the charge numbers on the ions from the salt of interest and I is
the ionic strength, defined in [5F.28–178]. For a 1:1 electrolyte of univalent ions
at molality b, I = b/b −○ (recall that b −○ = 1 mol kg−1 so b must also be in units
of mol kg−1 ). A test of this equation is to plot log γ± against I 1/2 , such a plot is
shown in Fig. 5.31.
b/(mmol kg−1 ) γ± log γ± I 1/2

1.00 0.964 9 −0.015 52 0.031 6
2.00 0.951 9 −0.021 41 0.044 7
5.00 0.927 5 −0.032 69 0.070 7
10.0 0.902 4 −0.044 60 0.100 0
20.0 0.871 2 −0.059 88 0.141 4
The data fit to quite a good straight line, the equation of which is
log γ± = −0.404 × I 1/2 − 3.395 × 10−3
The Debye–Hückel limiting law is verified to the extent that log γ± is linear
in I 1/2 , however the slope is −0.404 rather than the expected value of −0.509.
According to the limiting law the intercept at I 1/2 = 0 should be zero, which is
not the case. Overall, the conclusion is that the limiting law predicts the correct
functional dependence of γ± on ionic strength but fails to predict the correct
values.
The Davies equation is given in [5F.30b–179]
−A ∣z+ z− ∣ I 1/2
log γ± = + CI
1 + BI 1/2
The data can be fitted to this equation using mathematical software to imple-
ment a non-linear fit; the value of A is fixed as 0.509 for the fit. The best-fit
0.00
−0.02
log γ±
−0.04
−0.06
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

I 1/2
Figure 5.31
parameters are B = 1.2975, C = −0.0470. Using these, a table is drawn up

comparing the experimental values of γ± with those predicted by the limiting
law and by the Davies equation.
b/(mmol kg−1 ) γ± γ± (DH) error (%) γ± (Davies) error (%)

1.00 0.964 9 0.963 6 −0.13 0.965 1 0.02
2.00 0.951 9 0.948 9 −0.31 0.951 9 0.00
5.00 0.927 5 0.920 5 −0.76 0.927 4 −0.01
10.00 0.902 4 0.889 4 −1.44 0.902 4 0.00
20.00 0.871 2 0.847 3 −2.75 0.871 2 0.00
The table shows that the values predicted by the limiting law are increasingly in
error as the ionic strength increases, whereas the Davies equation reproduces
most of the experimental data to within the stated precision. However, it must
be kept in mind that the B and C parameters in the Davies equation have been
adjusted specifically to fit these data.

I5.1 (a) To develop the expression for K into the form requested it is useful to
rewrite [MA] and [M]free in terms of the total concentration of macro-
molecule, [M]. The total amount of A in the dialysis bag is [A]in = [A]free +
[A]bound , but the amount of A bound is equal to the amount of the macro-
molecule ligand complex, MA: [A]bound = [MA], therefore
[A]in = [A]free + [MA] hence [MA] = [A]in − [A]free
Recall that [A]free = [A]out and that, by definition ν = ([A]in −[A]out )/[M],
it therefore follows that
[MA] = [A]in − [A]out = ν[M]
Now consider the macromolecule, the total concentration of which is

[M]. It follows that [M] = [MA] + [M]free . The expression just derived
for [MA], [MA] = ν[M] is substituted in to give [M] = ν[M] + [M]free ,
from which it follows that [M]free = [M](1 − ν)
With these expressions for [M]free and [MA], the expression for K is
developed into the requested form
[MA]c −○ ν[M]c −○ νc −○
K= = =
[M]free [A]free [M](1 − ν)[A]out (1 − ν)[A]out
where [A]free = [A]out is also used.
(b) The equilibrium constant K ′ describes the equilibrium between a maco-
molecule with a single binding site, S, and the bound complex, SA
[SA]c −○
K′ =
[S]free [A]free
In part (a) ν is defined as the average number of bound ligands per macro-
molecule, and is therefore given by ν = [A]bound /[M]. Whereas M has N
binding sites, S only has one site, so the average number of ligands bound
per S is ν/N. This number is also expressed (by analogy with the earlier
discussion) as [A]bound /[S], so it follows that ν/N = [A]bound /[S]. The
final step is to realise that the concentration of bound ligand is equal to
the concentration of the S–A complex, so [A]bound = [SA]. It therefore
follows that ν/N = [SA]/[S]. This is rearranged to [SA] = [S]ν/N, which
is one of the terms needed in the expression for K ′ .
The other term is [S]free which is related to [S] as follows. The total con-
centration of S is given by [S] = [S]free + [SA], hence [S]free = [S] − [SA].
Substituting [SA] = [S]ν/N gives [S]free = [S] − [S]ν/N = (1 − ν/N)[S].
The expression for the equilibrium constant is now developed as
[SA]c −○ ([S]ν/N)c −○ νc −○
K′ = = =
[S]free [A]free (1 − ν/N)[S][A]free [A]free (N − ν)
where to go to the final expression the numerator and demoninator are
multiplied by N and [S] is cancelled. The Scatchard equation follows by
taking the factor (N − ν) to the left
νc −○
K′ N − K′ν =
[A]free
(c) The straight line plot is ν/[A]free against ν. The task is therefore to deter-
mine ν from the data. Note that the data given are the total concentrations
of EB in and outside the bag. It therefore follows that
[EB]total,in = [EB]bound + [EB]free = [EB]bound + [EB]out
It follows that [EB]bound = [EB]total,in − [EB]out . Recall that ν is defined
as ν = [EB]bound /[M], therefore
[EB]total,in − [EB]out
ν=
[M]
The given data and the derived values of ν are shown in the following table
(1 µM = 1 µmol dm−3 ), and plotted in Fig. 5.32.
[EB]out /µM [EB]in /µM ν (ν/[EB]out )/(µM−1 )

0.042 0.292 0.250 5.95
0.092 0.590 0.50 5.41
0.204 1.204 1.00 4.90
0.526 2.531 2.01 3.81
1.150 4.150 3.00 2.61
6.0
(ν/[EB]out )/(µM−1 )
4.0
2.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
ν
Figure 5.32
The data fit to quite a good straight line, the equation of which is
(ν/[EB]out )/(µM−1 ) = −1.17 × ν + 6.12
The slope is −K ′ /c −○ where c −○ is the standard concentration, 1 mol dm−3 ,

but because [EB]out is used in µM, c −○ = 106 µM. It follows that the
(dimensionless) equilibrium constant is K ′ = 1.17 × 106 . The intercept
gives K ′ N/c −○ , hence N = 5.23 : this is the average number of binding
sites per DNA molecule. The graph is a good straight line, indicating that
the data fit the model quite well.
I5.3 The dissolution of the protein according to the given equilibrium is described
by a solubility constant K s
PXv (s) ↽ ⇀ Pv+ (aq) + v X− (aq)

ÐÐÐ K s = a P a Xv
where the solubility constant is written in terms of the activities. Introducing

the activity coefficients and molalities, b, gives
K s = γ±v+1 b P b Xv
At low to moderate ionic strengths the Debye–Hückel limiting law, [5F.27–178],

log γ± = −A∣z− z+ ∣I 1/2 , is a reasonable approximation for the activity coeffi-
cients.
Addition of a salt, such as (NH4 )2 SO4 , causes I to increase, log γ± to become

more negative, and hence γ± to decrease. However, K s is an equilibrium con-
stant and remains unchanged. Therefore, the molality of Pv+ increases and the
protein solubility increases proportionately.
This effect is also explicable in terms of Le Chatelier’s principle. As the ionic
strength increases by the addition of an inert electrolyte such as (NH4 )2 SO4 ,
the ions of the protein that are in solution attract one another less strongly, so
that the equilibrium is shifted in the direction of increased solubility.
The explanation of the salting out effect is somewhat more complicated and
can be related to the failure the Debye–Hückel limiting law at higher ionic
strengths. At high ionic strengths the Davies equation, [5F.30b–179], is a better
approximation
−A ∣z+ z− ∣ I 1/2
log γ± = + CI
1 + BI 1/2
At low concentrations of inert salt, I 1/2 > I, the first term dominates, γ± de-
creases with increasing I, and salting in occurs; however, at high concentra-
tions, I > I 1/2 , the second term dominates, γ± increases with increasing I, and
salting out occurs. The Le Chatelier’s principle explanation is that the water
molecules are tied up by ion-dipole interactions and become unavailable for
solvating the protein, thereby leading to decreased solubility.
I5.5 In Section 5B.2(e) on page 152 the derivation of the expression for the osmostic
pressure starts by equating the chemical potential of A as a pure liquid subject
to pressure p with that of A in a solution of mole fraction x A containing solute B
and subject to pressure p+Π: µ ∗A (p) = µ A (x A , p+Π). The chemical potential of
A in the solution is then expressed as µ A (x A , p + Π) = µ∗A (p + Π) + RT ln x A ,
which assumes ideality. If the solution is not ideal, then the mole fraction is
replaced by the activity a A to give
µ ∗A (p) = µ ∗A (p + Π) + RT ln a A
The derivation then proceeds as before yielding the intermediate result −RT ln a A =
Vm Π for the non-ideal solution.
The osmotic coefficient ϕ is defined as ϕ = −(x A /x B ) ln a A , hence ln a A =
−ϕx B /x A . This expression for ln a A is substituted into −RT ln a A = Vm Π to
give RTϕx B /x A = Vm Π. The final steps assume that the solution is dilute so
that x A ≈ 1 and x B = n B /(n A + n B ) ≈ n B /n A
xB
RTϕ = Vm Π
xA
nB
hence RTϕ = Vm Π
nA
nB
RTϕ =Π
n A Vm
RTϕ[B] = Π
To go to the last line V = n A Vm and [B] = n B /V are used.

I5.7 On the basis of Raoult’s law, the activity in terms of the vapour pressure p J
is given by [5F.2–173], a J = p J /p∗J , where p∗J is the vapour pressure of the
pure solvent. The activity coefficient is defined through [5F.4–173], a J = γ J x J ,
therefore γ J = p J /p∗J x J .
The data as given do not include values for the vapour pressure over the pure
liquids, so the first task is to plot p J against x J and extrapolate to x J = 1 to
find p∗J . The vapour pressures are plotted in this way Fig. 5.33, and the lin-
ear extrapolations to find the vapour pressures of the pure substances are also
shown. These give the values p∗E = 7.45 kPa and p∗B = 35.41 kPa; using these
values the activity coefficients are computed as shown in the table.
pE
pB
30
p J /kPa
20
10
0
0.0 0.2 0.4 0.6 0.8 1.0
xJ
Figure 5.33
xE p E /kPa xB p B /kPa γE γB γ E (Henry) G E /kJ mol−1

0.016 0 0.484 0.984 0 35.05 4.06 1.01 1.70 0.08
0.043 9 0.967 0.956 1 34.29 2.96 1.01 1.24 0.16
0.083 5 1.535 0.916 5 33.28 2.47 1.03 1.03 0.26
0.113 8 1.89 0.886 2 32.64 2.23 1.04 0.93 0.34
0.171 4 2.45 0.828 6 30.90 1.92 1.05 0.80 0.42
0.297 3 3.31 0.702 7 28.16 1.49 1.13 0.63 0.55
0.369 6 3.83 0.630 4 26.08 1.39 1.17 0.58 0.59
0.583 4 4.84 0.416 6 20.42 1.11 1.38 0.47 0.53
0.660 4 5.36 0.339 6 18.01 1.09 1.50 0.46 0.52
0.843 7 6.76 0.156 3 10.00 1.07 1.81 0.45 0.41
0.993 1 7.29 0.006 9 0.47 0.98 1.92 0.41 −0.03
On the basis of Henry’s law, the activity in terms of the vapour pressure p J is
given by [5F.10–174], a J = p J /K J , where K J is the Henry’s law constant for J as
a solute. The activity coefficient is defined as before, a J = γ J x J , and therefore

γ J = p J /K J x J .
To find the Henry’s law constant for E, the limiting slope of a plot of p E against
x E is taken. The three data points given at the lowest values of x E do not ex-
trapolate back to the origin, which is not in accord with Henry’s law. Arguably
there are several equally valid ways of proceeding here, but one is to force the
best-fit line to pass through the origin and then use the first three data points;
this leads to the slope is shown by the dashed line in Fig. 5.33. The limiting
slope, taken in this way is 17.77 and so K E = 17.77 kPa. This value is used to
compute the activity coefficients for E based on Henry’s law, and the results are
shown in column headed γ E (Henry) in the table above. The outcome is not
satisfactory because the expecting limiting behaviour γ E → 1 as x E → 0 is not
evidenced.
The excess Gibbs energy is define in [5B.5–146] as G E = ∆ mix G − ∆ mix G ideal .

As explained in Section 5F.3 on page 175, the Gibbs energy of mixing is given
in terms of the activities as ∆ mix G = nRT (x A ln a A + x B ln a B ), whereas the
ideal Gibbs energy of mixing is ∆ mix G ideal = nRT (x A ln x A + x B ln x B ). The
activities are written as a A = γ A x A and hence
G E = ∆ mix G − ∆ mix G ideal

= nRT (x A ln a A + x B ln a B ) − nRT (x A ln x A + x B ln x B )
= nRT (x A ln γ A x A + x B ln γ B x B ) − nRT (x A ln x A + x B ln x B )
= nRT (x A ln γ A + x B ln γ B )
Using the final expression G E /n is computed from the given data and using
the activity coefficients (based on Raoult’s law) already derived. The computed
values are given in the table.
I5.9 On the basis of Raoult’s law, the activity in terms of the vapour pressure p J is
given by [5F.2–173], a J = p J /p∗J , where p∗J is the vapour pressure of the pure
substance. The activity coefficient is defined through [5F.4–173], a J = γ J x J ,
therefore γ J = p J /p∗J x J . The partial pressure in the gas phase is determined
from the mole fraction in the gas phase, y J , p J = y J p tot , so the final calculation
is γ J = y J p tot /p∗J x J .
The total pressure is given in kPa, whereas the vapour pressure over pure oxy-
gen is given in Torr. The conversion is
(101.325 kPa)/(1 atm)

(p kPa) = (p′ Torr) ×
(760 Torr)/(1 atm)
The temperature-composition phase diagram is shown in Fig. 5.34 and the

computed values of the activity coefficient are given in the table below. The
fact that the activity coefficient is close to 1 indicates near-ideal behaviour.
95
vapour
liquid
90
T/K
85
80
75
0.0 0.2 0.4 0.6 0.8 1.0
x O2 or y O2
Figure 5.34
T/K x O2 y O2 p∗O2 /Torr γ O2

77.3 0 0 154
78 0.10 0.02 171 0.88
80 0.34 0.11 225 1.08
82 0.54 0.22 294 1.04
84 0.70 0.35 377 0.99
86 0.82 0.52 479 0.99
88 0.92 0.73 601 0.99
90.2 1.00 1.00 760 0.99
I5.11 In Section 5B.2(a) on page 148 the derivation of the expression for the freezing
point depression starts by equating the chemical potential of A as a pure solid
with that of A in a solution of mole fraction x A containing solute B: µ ∗A (s) =
µ A (l, x A ). The latter chemical potential is written, for an ideal solution, as
µ A (l, x A ) = µ∗A (l)+ RT ln x A . If the solution is not ideal, then the mole fraction
is replaced by the activity a A to give
µ∗A (s) = µ ∗A (l) + RT ln a A
The derivation then proceeds as before. First, µ ∗A (s) − µ ∗A (l) is identified as
−∆ fus G to give
µ ∗ (s) − µ∗A (l) −∆ fus G
ln a A = A =
RT RT
Next, both sides are differentiated with respect to T and the Gibbs–Helmholtz
equation, d(G/T)/dT = −H/T 2 , is applied to the right-hand side
d ln a A −1 d ∆ fus G ∆ fus H
= ( )=
dT R dT T RT 2
The freezing point depression ∆T is defined as ∆T = T ∗ − T, where T is the
freezing point of the solution and T ∗ is the freezing point of the pure solvent.
It follows that d∆T = −dT. Finally, because the freezing point depression is
small, T on the right-hand side of the previous equation can be replaced by T ∗
to give
d ln a A ∆ fus H
=− (5.5)
d∆T RT ∗ 2
The empirical freezing-point constant K f is introduced in [5B.12–151], ∆T =
K f b B , where b B is the molality of the solvent. This expression is developed
by writing b B = n B /m A , where m A is the mass of solvent A (in kg) and then
m A = n A M, where M is the molar mass of the solvent. It follows that ∆T =
K f n B /(n A M). For dilute solutions x B ≈ n B /n A so ∆T = K f x B /M.
The expression for the freezing point depression given in [5B.11–151] is
RT ∗ 2
∆T = xB
∆ fus H
Comparison of this with the expression just derived, ∆T = K f x B /M, gives
Kf RT ∗ 2
=
M ∆ fus H
Using this expression for the right-hand side in eqn 5.5 gives the required form
d ln a A M
=− (5.6)
d∆T Kf
Start with the Gibbs–Duhem equation, [5A.12b–136], n A dµ A + n B dµ B = 0 and

divide both sides by (n A + n B ) to give x A dµ A + x B dµ B = 0. Next introduce the
general dependence of the chemical potential on activity, µ A = µ −A○ + RT ln a A ,
from which it follows that dµ A = RTd ln a A . Introducing this into the Gibbs–
Duhem equation, and dividing both sides by RT gives
xB
x A d ln a A + x B d ln a B = 0 hence d ln a A = − d ln a B (5.7)
xA
It follows that
d ln a A x B d ln a B
=−
d∆T x A d∆T
This expression for (d ln a A )/d∆T is substituted into eqn 5.6
x B d ln a B M
− =−
x A d∆T Kf
This expression is developed further by using the approximations x A ≈ 1 and
x B ≈ n B /n A which are appropriate for dilute solutions
d ln a B x A M
=
d∆T xB Kf
1 M 1 nA M 1 mA
= = =
n B /n A K f n B K f nB Kf
1
=
bB Kf
On the penultimate line m A = n A M, where m A is the mass of solvent A, is used,

and to go to the final line the molality of B, b B = n B /m A is introduced.
Recall the definition of the osmotic coefficient ϕ = −(x A /x B ) ln a A and the

result from eqn 5.7 that d ln a A = −(x B /x A )d ln a B . It follows that
xB xB
∫ d ln a A = ∫ − d ln a B hence ln a A = ∫ − d ln a B
xA xA
where the integration is from pure A to some arbitrary composition. This result
is used with the definition of ϕ to give
xA xA xB
ϕ=− ln a A = − ∫ − d ln a B
xB xB xA
The terms in x A and x B cannot be cancelled because those inside the integral
are functions of the variable of integration. For dilute dilute solutions x A ≈ 1
and x B ≈ n B /n A = n B /(m A /M) = M(n B /m A ) = Mb B , where M is the molar
mass of A, m A is the mass of the solvent, and b B is the molality of B. The integral
for ϕ therefore becomes
xA 1
ϕ=− ln a A = ∫ b B d ln a B (5.8)
xB bB
For a 1:1 univalent electrolyte the Debye–Hückel limiting law [5F.27–178], log γ± =
−A ∣z+ z− ∣ I 1/2 , becomes log γ± = −A(b B /b −○ )1/2 , and changing to naural loga-
rithms this becomes ln γ± = −A′ (b B /b −○ )1/2 , with A′ = 2.303 × A. Using this,
the activity and its derivative are developed as
ln a B = ln(b B /b−○ ) + ln γ = ln(b B /b −○ ) − A′ (b B /b−○ )1/2

1 1 1/2
hence d ln a B = [ − 12 A′ ( ) ] db B
bB b B b −○
This expression for d ln a B is used in eqn 5.8 and the integral is then evaluated
1
ϕ= ∫ b B d ln a B
bB
1 1 1 ′ 1 1/2
= ∫ b B [ − A ( ) ] db B
bB bB 2 b B b −○
1/2 ⎤
1 ′ bB
1 ⎡
⎢ ⎥
= ∫ ⎢⎢1 − 2 A ( −○ ) ⎥⎥ db B
bB ⎣ b ⎦
bB
1 1 1/2 3/2
= [b B − 12 × 23 × A′ ( −○ ) b B ]
bB b b B =0
b B 1/2
= 1 − 13 A′ ( )
b −○
6 Chemical equilibrium
6A The equilibrium constant

D6A.1 The terms appearing in the equilibrium constant are the activities of the species
involved in the equilibrium, and these terms arise because the chemical po-
tential of each species depends on its activity. If a pure liquid or pure solid
is part of the equilibrium, its chemical potential contributes to the value of
∆ r G −○ . However, as the substance is in its pure form there is no composition
dependence of its chemical potential and hence no term in the equilibrium
constant. Put another way, such species have unit activity.
ν
E6A.1(a) The equilibrium constant is defined by [6A.14–193], K = (∏J a J J )equilibrium .
The ‘equilibrium’ subscript indicates that the activities are those at equilibrium
rather than at an arbitrary stage in the reaction; however this subscript is not
usually written explicitly. In this case
a PH
4
3 (g)
K = a−1 −6
P 4 (s) a H 2 (g) a PH 3 (g) =
4
a P4 (s) a H
6
2 (g)
The activity of P4 (s) is 1, because it is a pure solid. Furthermore if the gases

are treated as perfect then their activities are replaced by a J = p J /p−○ . The
equilibrium constant becomes
(p PH3 /p−○ )4 p4PH3 p−○
2
K= =
(p H2 /p−○ )6 p6H2
E6A.2(a) The standard reaction Gibbs energy is given by [6A.13a–193]

∆ r G −○ = −
○
∑ ν∆ f G − ∑ ν∆ f G −○
Products Reactants
The relationship between ∆ r G −○ and K, [6A.15–194], ∆ r G −○ = −RT ln K, is then

used to calculate the equilibrium constant.
(i) For the oxidation of ethanal
∆ r G −○ = 2∆ f G −○ (CH3 COOH, l) − {2∆ f G −○ (CH3 CHO, g) + ∆ f G −○ (O2 , g)}
= 2∆ f G −○ (CH3 COOH, l) − 2∆ f G −○ (CH3 CHO, g)
= 2(−389.9 kJ mol−1 ) − 2(−128.86 kJ mol−1 ) = −5.22... × 105 J mol−1
194 6 CHEMICAL EQUILIBRIUM
Then
−
○ −5.22... × 105 J mol−1
K = e−∆ r G /RT
= exp (− ) = 3.24 × 1091
(8.3145 J K−1 mol−1 ) × (298 K)
(ii) For the reaction of AgCl(s) with Br2 (l)
∆ r G −○ = 2∆ f G −○ (AgBr, s) + ∆ f G −○ (Cl2 , g)
− {2∆ f G −○ (AgCl, s) + ∆ f G −○ (Br2 , l)}
= 2∆ f G −○ (AgBr, s) − 2∆ f G −○ (AgCl, s)
= 2(−96.90 kJ mol−1 ) − 2(−109.79 kJ mol−1 ) = +25.7... kJ mol−1
Then
−
○ 25.7... × 103 J mol−1
K = e−∆ r G /RT
= exp (− ) = 3.03 × 10−5
(8.3145 J K−1 mol−1 ) × (298 K)
Of these two reactions, the first has K > 1 at 298 K.
E6A.3(a) The relationship between ∆ r G −○ and the equilibrium constant is given by [6A.15–
194], ∆ r G −○ = −RT ln K. The ratio of the equilibrium constants for the two
reactions is
−
○
K 1 e−∆ r G 1 /RT ∆ r G 1−○ − ∆ r G 2−○
= −∆ G −○ /RT = exp (− )
K2 e r 2 RT
(−320 × 103 J mol−1 ) − (−55 × 103 J mol−1 )
= exp (− ) = 1.4 × 1046
(8.3145 J K−1 mol−1 ) × (300 K)
E6A.4(a) The reaction Gibbs energy at an arbitrary stage is given by [6A.11–193], ∆ r G =

∆ r G −○ + RT ln Q. In this case ∆ r G −○ = −32.9 kJ mol−1 . The values of ∆ r G for at
value of Q are:
(i) At Q = 0.010
∆ r G = (−32.9 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 )×(298 K)×ln(0.010)

= −4.43... × 104 J mol−1 = −44 kJ mol−1
(ii) At Q = 1.0
∆ r G = (−32.9 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × (298 K) × ln(1.0)

= −3.29... × 104 J mol−1 = −33 kJ mol−1 (= ∆ r G −○ )
(iii) At Q = 10
∆ r G = (−32.9 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × (298 K) × ln(10)

= −2.71... × 104 J mol−1 = −27 kJ mol−1
(iv) At Q = 105
∆ r G = (−32.9 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × (298 K) × ln(105 )

= −4.37... × 103 J mol−1 = −4.4 kJ mol−1
(v) At Q = 106
∆ r G = (−32.9 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × (298 K) × ln(106 )

= +1.33... × 103 J mol−1 = +1.3 kJ mol−1
The equilibrium constant K is the value of Q for which ∆ r G = 0. From the

above values, K will therefore be somewhere between 105 and 106 . To find
exactly where by linear interpolation, note that according to ∆ r G = ∆ r G −○ +
RT ln Q, a plot of ∆ r G against ln Q should be a straight line. Consider the two
points on either side of zero, that is, (iv) and (v). The point ∆ r G = 0 occurs a
fraction (4.37...)/(1.33... + 4.37...) = 0.766... of the way between points (iv)
and (v), so is at
ln K = ln 105 + (0.766...) × (ln 106 − ln 105 ) = 13.2...
Hence K = e13.2 ... = 5.84 × 105

The value is calculated directly by setting ∆ r G = 0 and Q = K in ∆ r G = ∆ r G −○ +
RT ln Q and rearranging for K
−
○ −32.9 × 103 J mol−1
K = e−∆ r G /RT
= exp (− ) = 5.84 × 105
(8.3145 J K−1 mol−1 ) × (298 K)
which is the same result as obtained from the linear interpolation.
E6A.5(a) For the reaction 2H2 O(g) ⇌ 2H2 (g) + O2 (g) the following table is drawn up
by supposing that there are n moles of H2 O initially and that at equilibrium a
fraction α has dissociated.
2H2 O ⇌ 2H2 + O2
Initial amount n 0 0
Change to reach equilibrium −αn +αn + 21 αn
Amount at equilibrium (1 − α)n αn 1
2
αn
1−α α
1
α
Mole fraction, x J 2
1 + 12 α 1 + 12 α 1 + 12 α
(1 − α)p αp
1
αp
Partial pressure, p J 2
1 + 12 α 1 + 12 α 1 + 12 α
The total amount in moles is n tot = (1 − α)n + αn + 12 αn = (1 + 21 α)n. This

value is used to find the mole fractions. In the last line, p J = x J p has been
used. Treating all species as perfect gases so that a J = (p J /p−○ ), the equilibrium
constant is
αp 2 1
αp
(p H2 /p−○ )2 (p O2 /p−○ ) p2H2 p O2 ( 1+ 12 α ) ( 1+ 12 α )
2
aH
2
aO
K = 22 2 = = =
a H2 O (p H2 O /p−○ )2 p2H2 O p−○ (1−α)p 2
( 1+ 1 α ) p−○
2
2
α p (1 + 12 α)2
1 3 3
α3 p
= =
(1 − α)2 p2 (1 + 12 α)3 p−○ (1 − α)2 (2 + α) p−○
In this case α = 1.77% (= 0.0177) and p = 1.00 bar; recall that p−○ = 1 bar.
0.01773 1.00 bar
K= 2
× = 2.85 × 10−6
(1 − 0.0177) × (2 + 0.0177) 1 bar
The relationship between K and K c is [6A.18b–195], K = K c × (c −○ RT/p−○ ) .

∆ν
E6A.6(a)
For the reaction H2 CO(g) ⇌ CO(g) + H2 (g)
∆ν = ν CO + ν H2 O − ν H2 CO = 1 + 1 − 1 = +1 hence K = K c × (c −○ RT/p−○ )
p−○ /c −○ R evaluates to 12.03 K so the relationship can alternatively be written as

K = K c × (T/K)/(12.03).
E6A.7(a) The following table is drawn up:
2A + B ⇌ 3C + 2D
Initial amount, n J,0 /mol 1.00 2.00 0 1.00
Change, ∆n J /mol −0.60 −0.30 +0.90 +0.60
Equilibrium amount,
0.40 1.70 0.90 1.60
n J /mol
Mole fraction, x J 0.0869... 0.369... 0.195... 0.347...
Partial pressure, p J (0.0869...)p (0.369...)p (0.195...)p (0.347...)p
To go to the second line, the fact that 0.90 mol of C has been produced is used to
deduce the changes in the other species given the stoichiometry of the reaction.
For example, 2 mol of A is consumed for every 3 mol of C produced so ∆ν A =
− 32 ∆ν C = 32 ×+0.90 mol = −0.60 mol. The total amount in moles is (0.40 mol)+
(1.70 mol) + (0.90 mol) + (1.60 mol) = 4.6 mol. This value has been used to
find the mole fractions. In the last line, p J = x J p has been used.
(i) The mole fractions are given in the above table.

(ii) Treating all species as perfect gases so that a J = p J /p−○ the equilibrium
constant is
a C3 a D
2
(p C /p−○ )3 (p D /p−○ )2 p3C p2D x C3 x D2 p2
K= = = =
a A2 a B (p A /p−○ )2 (p B /p−○ ) p2A p B p−○ 2 x A2 x B p−○ 2
3 2
(0.195...) (0.347...) (1.00 bar)2
= 2 × = 0.324... = 0.32
(0.0869...) (0.369...) (1 bar)2
(iii) The relationship between ∆ r G −○ and K [6A.15–194], ∆ r G −○ = −RT ln K, is

used to calculate ∆ r G −○ :
∆ r G −○ = −(8.3145 J K−1 mol−1 ) × ([25 + 273.15] K) × ln 0.324...

= +2.8 kJ mol−1
E6A.8(a) The reaction Gibbs energy for an arbitrary reaction quotient is given by [6A.11–
193], ∆ r G = ∆ r G −○ + RT ln Q. Treating borneol and isoborneol as perfect gases
so that a J = p J /p−○ , the reaction quotient Q is
a isoborneol p isoborneol /p−○ p isoborneol

Q= = =
a borneol p borneol /p−○ p borneol
Because p J = x J p = (n J /n)p ∝ n J it follows that
n isoborneol 0.30 mol

Q= = = 2.0
n borneol 0.15 mol
Hence
∆ r G = ∆ r G −○ + RT ln Q
= (+9.4 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × (503 K) × ln 2.0
= +12 kJ mol−1
E6A.9(a) The reaction corresponding to the standard Gibbs energy change of formation
of NH3 is
1
N (g) + 32 H2 (g) ⇌ NH3 (g)
2 2
This is the reaction in question. The reaction Gibbs energy for an arbitrary
reaction quotient is given by [6A.11–193], ∆ r G = ∆ r G −○ + RT ln Q. All species
are treated as perfect gases so that a J = p J /p−○ . Therefore the reaction quotient
Q is
a NH3 (p NH3 /p−○ ) p NH3 p−○

Q= = =
1/2 3/2
a N2 a H2 (p N2 /p−○ )1/2 × (p H2 /p−○ )3/2 p1/2 p3/2
N2 H2
(4.0 bar) × (1 bar)
= = 2.30...
(3.0 bar)1/2 × (1.0 bar)3/2
Hence
∆ r G = ∆ r G −○ + RT ln Q
= (−16.5 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × (298 K) × ln(2.30...)
= −14 kJ mol−1
Because ∆ r G < 0 the spontaneous direction of the reaction under these condi-
tions is from left to right.
E6A.10(a) The standard Gibbs energy change for the reaction is given in terms of the
standard Gibbs energies of formation by [6A.13a–193]:
∆ r G −○ = ∆ f G −○ (CaF2 , aq) − ∆ f G −○ (CaF2 , s)
This is rearranged for ∆ f G −○ (CaF2 , aq) and ∆ r G −○ is replaced by −RT ln K [6A.15–

194] to give
∆ f G −○ (CaF2 , aq) = ∆ r G −○ + ∆ f G −○ (CaF2 , s) = −RT ln K + ∆ f G −○ (CaF2 , s)

= −(8.3145 J K−1 mol−1 ) × ([25 + 273.15] K) × ln(3.9 × 10−11 )
+ (−1167 × 103 J mol−1 ) = −1.1 × 103 kJ mol−1
E6A.11(a) In general if the extent of a reaction changes by an amount ∆ξ then the amount
of a component J changes by ν J ∆ξ where ν J is the stoichiometric number for
species J (positive for products, negative for reactants). In this case ν A = −1
and ν B = +2.
n A = n A,0 + ∆n A = n A,0 + ν A ∆ξ = (1.50 mol) + (−1) × (0.60 mol) = 0.90 mol

n B = n B,0 + ∆n B = n B,0 + ν B ∆ξ = 0 + 2 × (0.60 mol) = 1.20 mol
E6A.12(a) The reaction Gibbs energy ∆ r G is defined by [6A.1–190], ∆ r G = (∂G/∂ξ) p,T .

Approximating the derivative by finite changes gives
∂G ∆G −6.4 kJ
∆r G = ( ) ≈ = = −64 kJ mol−1
∂ξ p,T ∆ξ +0.1 mol
E6A.13(a) A reaction is exergonic if ∆ r G < 0 and endergonic if ∆ r G > 0. From the Re-
source section the standard Gibbs energy change for the formation of methane
from its elements in their reference states at 298 K is ∆ f G −○ = −50.72 kJ mol−1 .
This is negative so the reaction is exergonic .
ν
E6A.14(a) The reaction quotient is defined by [6A.10–193], Q = ∏J a J J . For the reaction
A + 2B → 3C, ν A = −1, ν B = −2, and ν C = +3. The reaction quotient is then
a C3
Q = a−1 −2 3
A aB aC =
a A a B2
P6A.1 (a) The relationship between the equilibrium constant and ∆ r G −○ is [6A.15–
194], ∆ r G −○ = −RT ln K.
∆ r G −○ = −(8.3145 J K−1 mol−1 ) × (298.15 K) × ln 0.164 = +4.48 kJ mol−1
(b) The following table is drawn up. Iodine is not included in the calculations
as it is a solid.
I2 (s) + Br2 (g) ⇌ 2IBr(g)

Initial amount — n 0
Change to reach equilibrium — −αn +2αn
Amount at equilibrium — (1 − α)n 2αn
1−α 2α
Mole fraction, x J —
1+α 1+α
(1 − α)p 2α p
Partial pressure, p J —
1+α 1+α
The total amount in moles is n tot = (1 − α)n + 2αn = (1 + α)n. This
value is used to find the mole fractions. Treating Br2 (g) and IBr(g) as
perfect gases, so that a J = p J /p−○ , and I2 as a pure solid, so that a I2 = 1,
the equilibrium constant is:
2α p 2
a IBr
2
(p IBr /p−○ )2 p2IBr ( 1+α ) 4α 2 p
K= = −
○
= −
○
= =
a I2 a Br2 1 × (p Br2 /p ) p Br2 p (1−α)p
p −
○ (1 − α)(1 + α) p−○
1+α
Note that (1 − α)(1 + α) = 1 − α . With this, an expression for α is found

2
by straightforward algebra.
1
K p−○ 2
α=( )
4p + K p−○
1
2
0.164 × (1 bar)
=( )
4 × (0.164 atm) × (1.01325 bar)/(1 atm) + 0.164 × (1 bar)
= 0.444...
Hence
2α p 2 × 0.444... × (0.164 atm)
p IBr = = = 0.101 atm or 0.102 bar
1+α 1 + 0.444...
(c) The issue here is that the reaction under discussion is that with I2 (s). If
the partial pressure of I2 is not zero then p Br2 and p IBr no longer sum to
the total pressure p but rather to p − p I2 where p I2 is the partial pressure
of iodine. The partial pressures in the last line of the above table therefore
become
1−α 2α
p Br2 = (p − p I2 ) and p IBr = (p − p I2 )
1+α 1+α
The equilibrium constant K for the I2 (s) + Br2 (g) ⇌ 2IBr(g) is still K =
p2IBr /p Br2 p−○ but now with the new partial pressures:
2
2α
[( 1+α ) (p − p I2 )] 4α 2 p − p I2
K= = ( )
( 1−α
1+α
) (p − p I2 )p−○ (1 + α)(1 − α) p−○
Given the partial pressure of I2 this equation can be solved for α, and p IBr
calculated as before.
P6A.3 The following table is drawn up for the reaction, assuming that to reach equi-
librium the reaction proceeds by an amount z in the direction of the products.
H2 (s) + I2 (g) ⇌ 2HI(g)
Initial amount n H2 ,0 n I2 ,0 n HI,0

Change to reach
−z −z +2z
equilibrium
Amount at
n H2 ,0 − z n I2 ,0 − z n HI,0 + 2z
equilibrium
n H2 ,0 − z n I2 ,0 − z n HI,0 + 2z
Mole fraction, x J
n tot n tot n tot
(n H2 ,0 − z)p (n I2 ,0 − z)p (n HI,0 + 2z)p
Partial pressure, p J
n tot n tot n tot
where n tot = n H2 ,0 + n I2 ,0 + n HI,0 . Treating all species as perfect gases, so that

a J = p J /p−○ , the equilibrium constant is
a HI
2
(p HI /p−○ )2 p2HI (n HI,0 + 2z)2
K= = = =
a H2 a I2 (p H2 /p−○ )(p I2 /p−○ ) p H2 p I2 (n H2 ,0 − z)(n I2 ,0 − z)
Rearranging gives
K(n H2 ,0 − z)(n I2 ,0 − z) = (n HI,0 + 2z)2

Hence (K − 4)z 2 − ([n H2 ,0 + n I2 ,0 ]K + 4n HI,0 )z + (n H2 ,0 n I2 ,0 K − n HI
2
)=0
Substituting in the values for n J and K, dividing through by mol2 and writing
x = z/mol yields the quadratic
866x 2 − 609.8x + 104.36 = 0
which has solutions x = 0.410... and x = 0.293... implying z = (0.410... mol)

or z = (0.293... mol). The solution z = (0.410... mol) is rejected because z
cannot be larger than n H2 ,0 or n I2 ,0 . The amounts of each substance present at
equilibrium are therefore
n H2 = n H2 ,0 − z = (0.300 mol) − (0.293... mol) = 6.67 × 10−3 mol

n I2 = n I2 ,0 − z = (0.400 mol) − (0.293... mol) = 0.107 mol
n HI = n HI,0 + 2z = (0.200 mol) + 2 × (0.293... mol) = 0.787 mol
P6A.5 If the extent of reaction at equilibrium is ξ, then from the stoichiometry of the
reaction the amounts of A and B that have reacted are ξ and 3ξ respectively and
the amount of C that has been formed is 2ξ. If the initial amounts of A, B and
C are n, 3n and 0, the following table is drawn up.
A + 3B ⇌ 2C
Initial amount n 3n 0
Change to reach equilibrium −ξ −3ξ +2ξ
Amount at equilibrium n−ξ 3(n − ξ) 2ξ
n−ξ 3(n − ξ) 2ξ
Mole fraction, x J
4n − 2ξ 4n − 2ξ 4n − 2ξ
(n − ξ)p 3(n − ξ)p 2ξp
4n − 2ξ 4n − 2ξ 4n − 2ξ
The total amount in moles is n tot = (n − ξ) + 3(n − ξ) + 2ξ = 4n − 2ξ. This value

is used to find the mole fractions. Treating all species as perfect gases, so that
a J = p J /p−○ , the equilibrium constant is
2ξ p 2
( 4n−2ξ ) p−○
2
(p C /p−○ )2 p2C p−○
2
a2
K = C3 = = =
a A a B (p A /p−○ )(p B /p−○ )3 p A p3B (n−ξ)p 3(n−ξ)p 3
( 4n−2ξ ) ( 4n−2ξ )
2
16ξ 2 (2n − ξ)2 p−○
= ( )
27(n − ξ)4 p
Rearranging and then taking the square root gives
ξ 2 (2n − ξ)2 27K p2 ξ(2n − ξ) 1 √

= hence = 4 27K(p/p−○ )
(n − ξ)4 16p−○ 2 (n − ξ)2
The negative square root is rejected because 0 ≤ ξ ≤ n. This requirement arises

because if ξ < 0 this would imply a negative amount of C, while if ξ > n this
would imply negative amounts of A and B. Because 0 ≤ ξ ≤ n the left hand side
of the square rooted expression is always ≥ 0. Because p/p−○ cannot be negative
either it follows that the positive square root is required.
Rearranging further yields the quadratic
√
2
1
4
27K(p/p−○ )
(ξ/n) − 2(ξ/n) + √ =0
1 + 14 27K(p/p−○ )
which is solved to give

1
2
⎛ 1 ⎞
(ξ/n) = 1 − √
⎝ 1 + 4 27K(p/p−○ ) ⎠
1
The positive square root is rejected in order to ensure that 0 ≤ (ξ/n) ≤ 1.

Inspection of this expression shows that ξ → 0 as p → 0, indicating that the
reactants are favoured at low pressures. On the other hand (ξ/n) → 1 as p → ∞
indicating that the products are favoured at high pressure. (ξ/n) is plotted
against p/p−○ in the graph shown in Fig. 6.1, using three different values of K.
1.0
K = 100
0.8
K=1
0.6
ξ/n
0.4
K = 0.01
0.2
0.0
0 10 20 30 40 50
−
○
p/p
Figure 6.1
6B The response of equilibria to the conditions

D6B.1 This is discussed in Section 6B.2(a) on page 199.
D6B.3 This is discussed in Section 6B.1 on page 198 and in Section 6B.2 on page 199.
E6B.1(a) The relationship between ∆ r G −○ and K is given by [6A.15–194], ∆ r G −○ = −RT ln K.
Hence if K = 1, ∆ r G −○ = −RT ln 1 = 0. Furthermore ∆ r G −○ is related to ∆ r H −○
and ∆ r S −○ by [3D.9–96], ∆ r G −○ = ∆ r H −○ − T∆ r S −○ , so if K = 1
∆ r H −○
∆ r H −○ − T∆ r S −○ = 0 hence T=
∆ r S −○
Values of ∆ r H −○ and ∆ r S −○ at 298 K are calculated using data from the Resource
section.
∆ r H −○ = ∆ f H −○ (CaO, s) + ∆ f H −○ (CO2 , g) − ∆ f H −○ (CaCO3 , s, calcite)
= (−635.09 kJ mol−1 ) + (−393.51 kJ mol−1 ) − (−1206.9 kJ mol−1 )
= +178.3 kJ mol−1
∆ r S −○ = S m
−
○ −
○
(CaO, s) + S m −
○
(CO2 , g) − S m (CaCO3 , s, calcite)
= (39.75 J K−1 mol−1 ) + (213.74 J K−1 mol−1 ) − (92.9 J K−1 mol−1 )
= 160.59 J K−1 mol−1
Substituting these values into the equation found above, assuming that ∆ r H −○
and ∆ r S −○ do not vary significantly with temperature over the range of interest,
gives:
∆ r H −○ 178.13 × 103 J mol−1
T= = = 1109 K
∆ r S −○ 160.59 J K−1 mol−1
E6B.2(a) Treating the vapour as a perfect gas, so that a J = p J /p−○ , and noting that a A2 B = 1
because it is a pure solid, the equilibrium constant for the dissociation A2 B(s) ⇌
A2 (g) + B(g) is
a A2 ,g a B,g (p A2 /p−○ )(p B /p−○ ) p A2 p B2

K= = =
a A2 B,s 1 p−○ 2
Furthermore, because A2 and B are formed in a 1 ∶ 1 ratio, they each have a

mole fraction of 1/2 and the partial pressure of each is half the total pressure:
p A2 = p B = 12 p. The equilibrium constant is therefore
( 21 p)( 12 p) p2
K= =
p−○ 2 4p−○ 2
The variation of K with temperature, assuming that ∆ r H −○ does not vary with
T over the temperature range of interest, is given by [6B.4–201]:
∆ r H −○ 1 1 R ln(K 2 /K 1 )
ln K 2 − ln K 1 = − ( − ) hence ∆ r H −○ = −
R T2 T1 (1/T2 ) − (1/T1 )
Noting that the above expression for K implies that ln(K 2 /K 1 ) = ln(p22 /p21 ),
∆ r H −○ is calculated as
(8.3145 J K−1 mol−1 ) × ln ((547 kPa)2 /(208 kPa)2 )

∆ r H −○ = −
[1/(477 + 273.15) K] − [1/(367 + 273.15) K]
= 7.01... × 104 J mol−1 = 70.2 kJ mol−1
The standard entropy of reaction, ∆ r S −○ is found by rearranging ∆ r G −○ = ∆ r H −○ −

T∆ r S −○ [3D.9–96] and replacing ∆ r G −○ by ∆ r G −○ = −RT ln K [6A.15–194]:
∆ r H −○ − ∆ r G −○ ∆ r H −○ + RT ln K ∆ r H −○ p2
∆ r S −○ = = = + R ln ( −○ )
T T T 4p
Using the data for 367 ○ C (both temperatures give the same result) gives:
7.01... × 104 kJ mol−1 (208 kPa)2

∆ r S −○ = + (8.3145 J K−1 mol−1 ) × ln ( )
(367 + 273.15) K 4 × (100 kPa)2
= 1.10... × 102 J K−1 mol−1 = 110 J K−1 mol−1
An alternative (but equivalent) approach to finding ∆ r H −○ and ∆ r S −○ is to first

calculate ∆ r G −○ at both temperatures and hence obtain two equations of the
form ∆ r G −○ = ∆ r H −○ − T∆ r S −○ . These can then be solved simultaneously to find
the two unknowns ∆ r H −○ and ∆ r S −○ , assuming them to be constant.
The values of ∆ r H −○ and ∆ r S −○ are then used with ∆ r G −○ = ∆ r H −○ − T∆ r S −○
and ∆ r G −○ = −RT ln K to calculate ∆ r G −○ and K at the temperature of interest,
422 ○ C or 695.15 K. In making this calculation it is again assumed that ∆ r H −○
and ∆ r S −○ do not vary with temperature.

∆ r G −○ = ∆ r H −○ − T∆ r S −○
= (7.01... × 104 J mol−1 ) − (695.15 K) × (1.10... × 102 J K−1 mol−1 )
= −6.48... × 103 J mol−1 = −6.48 kJ mol−1
−
○ −6.48... × 103
K = e−∆ r G /RT
= exp (− ) = 3.07
(8.3145 J K−1 mol−1 ) × (695.15 K)
E6B.3(a) For the reaction N2 O4 (g) ⇌ 2NO2 (g) the following table is drawn up by
supposing that there are n moles of N2 O4 initially and that at equilibrium a
fraction α has dissociated.
N2 O4 ⇌ 2NO2
Initial amount n 0
Change to reach equilibrium −αn +2αn
Amount at equilibrium (1 − α)n 2αn
1−α 2α
Mole fraction, x J
1+α 1+α
(1 − α)p 2α p
1+α 1+α
The total amount in moles is n tot = (1 − α)n + 2αn = (1 + α)n. This value
is used to find the mole fractions. In the last line, p J = x J p [1A.6–9] has been
used. Treating all species as perfect gases so that a J = (p J /p−○ ), the equilibrium
constant is
2α p 2
a NO
2
(p NO2 /p−○ )2 p2NO2 ( 1+α ) 4α 2 p
K= 2
= −
○
= −
○
= = −
○
a N2 O4 (p N2 O4 /p ) p N2 O4 p (1−α)p
( 1+α ) p −
○ (1 − α)(1 + α) p
In this case α = 0.1846 and p = 1.00 bar; recall that p−○ = 1 bar.
4 × 0.18462 1.00 bar
K= × = 0.141... = 0.141
(1 − 0.1846) × (1 + 0.1846) 1 bar
The temperature dependence of K is given by [6B.4–201],

∆ r H −○ 1 1
ln K 2 − ln K 1 = − ( − )
R T2 T1
assuming that ∆ r H −○ is constant over the temperature range of interest. Taking
T1 = 25 ○ C (= 298.15 K) and T2 = 100 ○ C (= 373.15 K) gives
56.2 × 103 J mol−1 1 1
ln K 2 = ln(0.141...) − −1 −1 ( − ) = 2.59...
8.3145 J K mol 373.15 K 298.15 K
That is, K 2 = 13.4 , a larger value than at 25 ○ C, as expected for this endothermic
reaction.
E6B.4(a) The data in the Resource section is used to calculate ∆ r G −○ and ∆ r H −○ at 298 K
∆ r G −○ = ∆ f G −○ (CO2 , g) − ∆ f G −○ (PbO, s, red) − ∆ f G −○ (CO, g)

= (−394.36 kJ mol−1 ) − (−188.93 kJ mol−1 ) − (−137.17 kJ mol−1 )
= -68.26 kJ mol−1
∆ r H −○ = ∆ f H −○ (CO2 , g) − ∆ f H −○ (PbO, s, red) − ∆ f H −○ (CO, g)
= (−393.51 kJ mol−1 ) − (−218.99 kJ mol−1 ) − (−110.53 kJ mol−1 )
= −63.99 kJ mol−1
The equilibrium constant at 298 K is calculated from ∆ r G −○ using [6A.15–194],

∆ r G −○ = −RT ln K
∆ r G −○ −68.26 × 103 J mol−1

ln K = − =− = 27.5...
RT (8.3145 J K−1 mol−1 ) × (298 K)
hence K = e27.5 ... = 9.21... × 1011 = 9.22 × 1011
The temperature dependence of K is given by [6B.4–201],

∆ r H −○ 1 1
ln K 2 − ln K 1 = − ( − )
R T2 T1
assuming that ∆ r H −○ is constant over the temperature range of interest. This is
used to calculate the equilibrium constant at 400 K
−63.99 × 103 J mol−1 1 1

ln K 2 = ln(9.21... × 1011 ) − ( − ) = 20.9...
8.3145 J K−1 mol−1 400 K 298 K
That is, K 2 = 1.27 × 109 , a smaller value than at 298 K, as expected for this
exothermic reaction.
E6B.5(a) Assuming that ∆ r H −○ is constant over the temperature range of interest, the
temperature dependence of K is given by [6B.4–201],
∆ r H −○ 1 1
ln K 2 − ln K 1 = − ( − )
R T2 T1
Using ∆ r G −○ = −RT ln K to substitute for K 1 and setting ln K 2 = ln 1 = 0 (the
crossover point) gives
∆ r G −○ (T1 ) ∆ r H −○ 1 1
=− ( − )
RT1 R T2 T1
Rearranging for T2 gives
T1 ∆ r H −○ (1280 K) × (+224 kJ mol−1 )

T2 = = = 1.5 × 103 K
∆ r H −○ − ∆ r G −○ (T1 ) (+224 kJ mol−1 ) − (+33 kJ mol−1 )
Note that this temperature is outside the range over which ∆ r H −○ is known to
be constant and is therefore an estimate.
E6B.6(a) The van ’t Hoff equation [6B.2–200], d ln K/dT = ∆ r H −○ /RT 2 , is rearranged to

obtain an expression for ∆ r H −○
d ln K
∆ r H −○ = RT 2
dT
2 d B C B 2C 2C
= RT (A + + 2 ) = RT 2 (− 2 − 3 ) = −R (B + )
dT T T T T T
2 × (1.51 × 105 K2 )
= −(8.3145 J K−1 mol−1 ) × ((−1088 K) + )
400 K
= +2.76... × 103 J mol−1 = +2.77 kJ mol−1
The standard reaction entropy is obtained by first finding an expression for
∆ r G −○ using [6A.15–194]
B C C
∆ r G −○ = −RT ln K = −RT (A + + ) = −R (AT + B + )
T T2 T
The equation ∆ r G −○ = ∆ r H −○ −T∆ r S −○ [3D.9–96] is then rearranged to find ∆ r S −○
⎛ ⎞
−
○ ∆ r H −○ − ∆ r G −○ 1 ⎜ 2C C ⎟ C
∆r S = = ⎜−R (B +
⎜ ) + R (AT + B + )⎟ = R (A − 2 )
T T ⎜ T T ⎟ ⎟ T
⎝´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸−○¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹−○¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶⎠
∆r H −∆ r G
1.51 × 105 K2
= (8.3145 J K−1 mol−1 ) × (−1.04 − ) = −16.5 J K−1 mol−1
(400 K)2
An alternative approach to finding ∆ r S −○ is to use the variation of G with T

which is given by [3E.8–103], (∂G/∂T) p = −S. This implies that d∆ r G −○ /dT =
−∆ r S −○ where the derivative is complete (not partial) because ∆ r G −○ is indepen-
dent of pressure. Using the expression for ∆ r G −○ from above it follows that
⎛ ⎞
−
○ d∆ r G −○ d ⎜−R (AT + B + )⎟ = R (A − C )
⎜ C ⎟
∆r S = − =− ⎜
dT dT ⎜ T ⎟ ⎟ T2
⎝ ´¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸−
○
¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶ ⎠
∆r G
which is the same expression obtained above.

E6B.7(a) Treating all species as perfect gases so that a J = p J /p−○ , the equilibrium constant
for the reaction H2 CO(g) ⇌ CO(g) + H2 (g) is
a H2 a CO (p H2 /p−○ )(p CO /p−○ ) p H2 p CO (x H2 p)(x CO p)
K= = = =
a H2 CO (p H2 CO /p−○ ) p H2 CO p−○ (x H2 CO p)p−○
x H x CO p p
= 2 −
○
= K x × −○
x H2 CO p p
where K x is the part of the equilibrium constant expression that contains the
equilibrium mole fractions of reactants and products. Because K is indepen-
dent of pressure, if p doubles K x must halve in order to preserve the value of
K. In other words, K x is reduced by 50% .
E6B.8(a) The following table is drawn up for the borneol ⇌ isoborneol reaction, denoting
the initial amounts of borneol and isoborneol by n b,0 and n iso,0 and supposing
that in order to reach equilibrium an amount z of borneol has converted to
isoborneol.
Borneol ⇌ Isoborneol
Initial amount n b,0 n iso,0
Change to reach equilibrium −z +z
Amount at equilibrium n b,0 − z n iso,0 + z
n b,0 − z n iso,0 + z
Mole fraction, x J
n b,0 + n iso,0 n b,0 + n iso,0
(n b,0 − z)p (n iso,0 + z)p

n b,0 + n iso,0 n b,0 + n iso,0
The total amount in moles is (n b,0 − z) + n iso,0 + z = n b,0 + n iso,0 . This value is
used to find the mole fractions. Treating both species as perfect gases so that
a J = p J /p−○ the equilibrium constant is
a borneol p borneol n iso,0 + z
K= = =
a isoborneol p isoborneol n b,0 − z
Rearranging for z gives z = (Kn b,0 − n iso,0 )/(1 + K). Noting that n = m/M
where M = 154.2422 g mol−1 is the molar mass of borneol and isoborneol,
gives
n b,0 = (7.50 g)/(154.2422 g mol−1 ) = 0.0486... mol

n iso,0 = (14.0 g)/(154.2422 g mol−1 ) = 0.0907... mol
Kn b,0 − n iso,0 0.106 × (0.0486... mol) − (0.0907... mol)
z= = = −0.0774... mol
1+K 1 + 0.106
The negative value of z indicates that in order to reach equilibrium there is a net
conversion of isoborneol to borneol. Using this value of z, and the expressions
for x borneol and x isoborneol in the above table, the mole fractions at equilibrium
are calculated as
n b,0 − z (0.0486... mol) − (−0.0774... mol)
x borneol = = = 0.904...
n b,0 + n iso,0 (0.0486... mol) + (0.0907... mol)
= 0.904
Then x isoborneol = 1 − x borneol = 1 − 0.904... = 0.096
E6B.9(a) The temperature dependence of K is given by [6B.4–201]
∆ r H −○ 1 1 R ln(K 2 /K 1 )
ln K 2 − ln K 1 = − ( − ) hence ∆ r H −○ = −
R T2 T1 (1/T2 ) − (1/T1 )
(i) If the equilibrium constant is doubled then K 2 /K 1 = 2
(8.3145 J K−1 mol−1 ) × ln 2

∆ r H −○ = − = +52.9 kJ mol−1
[1/(308 K)] − (1/[298 K])
(ii) If the equilibrium constant is halved then K 2 /K 1 = 1/2
(8.3145 J K−1 mol−1 ) × ln(1/2)

∆ r H −○ = − = −52.9 kJ mol−1
[1/(308 K)] − (1/[298 K])
P6B.1 Assuming ∆ r H −○ to be constant over the temperature range of interest, the tem-
perature dependence of K is given by [6B.4–201]
∆ r H −○ 1 1 R ln(K 2 /K 1 )
ln K 2 − ln K 1 = − ( − ) hence ∆ r H −○ = −
R T2 T1 (1/T2 ) − (1/T1 )
Therefore
(8.3145 J K−1 mol−1 ) × ln [(1.75 × 105 )/(2.13 × 106 )]

∆ r H −○ = −
[1/(308 K)] − [1/(288 K)]
−1
= −92.2 kJ mol
P6B.3 The reaction for which ∆ r H −○ is the standard enthalpy of formation of UH3 is:
U(s) + 23 H2 (g) ⇌ UH3 (s)
Treating H2 as a perfect gas (so that a H2 = p H2 /p−○ ) and noting that pure solids
have a J = 1, the equilibrium constant for this reaction is written
3/2 −3/2
a UH3 1 p−○ p
K = 3/2 = −
○ 3/2
= ( ) = ( −○ )
a H2 a U (p/p ) × 1 p p
where p is the pressure of H2 . The standard reaction enthalpy, which corre-

sponds to ∆ f H −○ (UH3 , s), is obtained by rearranging the van ’t Hoff equation
[6B.2–200], d ln K/dT = ∆ r H −○ /RT 2 , for ∆ r H −○
−3/2
d ln K d p d p
∆ f H −○ (UH3 , s) = RT 2 = RT 2 ln ( −○ ) = − 32 RT 2 ln ( −○ )
dT dT p dT p
d d
= − 32 RT 2 ( ln p − ln p−○ ) = − 23 RT 2 ln p
dT dT
d B B C
= − 32 RT 2 (A + + C ln T) = − 32 RT 2 (− 2 + )
dT T T T
= − 23 R(CT − B)
Heat capacity at constant pressure is defined by [2B.5–46], C p = (∂H/∂T) p ,

which implies that ∆ f C −p○ = d∆ f H −○ /dT where the derivative is complete (not
partial) because ∆ f H −○ does not depend on pressure. Therefore
d
∆ f C −p○ = ( − 23 R[−B + CT]) = − 32 RC
dT
= − 32 × (8.3145 J K−1 mol−1 ) × (−5.65) = +70.5 J K−1 mol−1
P6B.5 The van ’t Hoff equation [6B.2–200] is:

d ln K ∆ r H −○ d ln K ∆ r H −○
= which can also be written − =
dT RT 2 d(1/T) R
The second form implies that a graph of − ln K against 1/T should be a straight
line of slope ∆ r H −○ /R. It is first necessary to relate K to α. To do this, the
following table is drawn up for the CO2 (g) ⇌ CO(g) + 12 O2 (g) equilibrium
1
CO2 (g) ⇌ CO(g) + O (g)
2 2
Initial amount n 0 0
Change to reach equilibrium −αn +αn + 12 αn
Amount at equilibrium (1 − α)n αn 1
2
αn
1−α α
1
α
1 + 21 α 1 + 12 α 1 + 12 α
(1 − α)p α
1
α
1 + 21 α 1 + 12 α 1 + 12 α
The total amount in moles is n tot = (1 − α)n + αn + 12 αn = (1 + 21 α)n. This

value is used to find the mole fractions. Treating all species as perfect gases (so
that a J = p J /p−○ ) the equilibrium constant is
1 1/2
αp αp
( ) ( 2 )
1/2 1/2
a CO a O2 (p CO /p−○ )(p O2 /p−○ )1/2 p CO p O2 1
1+ α
1
1+ α
K= = = = 2 2
a CO2 (p CO2 /p−○ ) p CO2 p−○ 1/2 (1−α)p
( 1 ) p−○
1+ α
2
1/2 1/2
( 21 α 3 ) p
= ( −○ )
(1 − α)(1 + 21 α)1/2 p
Using this expression, with p = 1 bar, K is calculated at each temperature
and − ln K is plotted against 1/(T/K) as described above; the plot is shown
in Fig. 6.2.
T/K α K 104 /(T/K) − ln K

1 395 1.44 × 10−4 1.222 × 10−6 7.168 13.62
1 443 2.50 × 10−4 2.794 × 10−6 6.930 12.79
1 498 4.71 × 10−4 7.224 × 10−6 6.676 11.84
14.0
13.0
− ln K
12.0
6.6 6.8 7.0 7.2

4
10 /(T/K)
Figure 6.2
ln K = 3.607 × 104 /(T/K) − 12.23
∆ r H −○ /R is determined from the slope
∆ r H −○ = R × (slope × K) = (8.3145 J K−1 mol−1 ) × (3.607 × 104 K)

= +2.99... × 105 J mol−1 = +3.00 × 102 kJ mol−1
The equilibrium constant K has already been calculated; from the table above
the value of K at 1443 K is 2.79 × 10−6 .
The standard reaction Gibbs energy is then calculated using ∆ r G −○ = −RT ln K,
[6A.15–194], and the standard reaction entropy from ∆ r G −○ = ∆ r H −○ − T∆ r S −○
[3D.9–96].
∆ r G −○ = −RT ln K = −(8.3145 J K−1 mol−1 ) × (1443 K) × ln(2.79... × 10−6 )

= +1.53... × 105 J mol−1 = +153 kJ mol−1
∆ r H −○ − ∆ r G −○ (2.99... × 105 J mol−1 ) − (1.53... × 105 J mol−1 )
∆ r S −○ = =
T 1443 K
= +102 J K−1 mol−1
P6B.7 The equilibrium is 2CH3 COOH(g) ⇌ (CH3 COOH)2 (g), the dimer being
held together by hydrogen bonds. The following table is drawn up, assuming
that the initial amount in moles of ethanoic acid is n and that at equilibrium a
fraction α of the ethanoic acid has dimerised.
2CH3 COOH ⇌ (CH3 COOH)2

Initial amount n 0
Change to reach equilibrium −αn + 21 αn
Amount at equilibrium (1 − α)n 1
2
αn
1−α
1
α
1 − 12 α 1 − 12 α
(1 − α)p
1
αp
1 − 12 α 1 − 12 α
The total amount in moles is n tot = (1 − α)n + 21 αn = (1 − 21 α)n. This value is

used to find the mole fractions.
The total amount in moles present at equilibrium is found from the pressure by
using the perfect gas law [1A.4–8], pV = n tot RT
2pV 2pV
pV = (1 − 12 α)nRT hence α =2− =2−
nRT (m/M)RT
where M = 60.0516 g mol−1 is the molar mass of ethanoic acid. The value of
α is then used to calculate K. Assuming that both species present are perfect
gases (so that a J = p J /p−○ ) and using the expressions for p J from the above table,
the equilibrium constant is
1
αp
( 2
1 ) p−○
a(CH COOH)2 (p(CH3 COOH)2 /p−○ ) p(CH3 COOH)2 p−○ 1− α
K= 2 3 = = = 2
a CH3 COOH (p CH3 COOH /p−○ )2 p2CH3 COOH (1−α)p
2
( 1 )
1− α
2
1
2
α(1 − 12 α)p−○
=
(1 − α)2 p
The values of α and K at the two temperatures are then calculated using these
formulae as
At 437 K
2 × (101.9 × 103 Pa) × (21.45 × 10−6 m3 )

α =2− = 0.607...
(0.0519 g/60.0516 g mol−1 ) × (8.3145 J K−1 mol−1 ) × (437 K)
1
α(1 − 12 α)p−○ 1
× (0.607...) × (1 − 12 × 0.607...) × (100 kPa)
K= 2
= 2
= 1.35...
(1 − α)2 p (1 − 0.607...)2 × (101.9 kPa)
= 1.35
At 471 K
2 × (101.9 × 103 Pa) × (21.45 × 10−6 m3 )
α =2− = 0.235...
(0.0380 g/60.0516 g mol−1 ) × (8.3145 J K−1 mol−1 ) × (471 K)
1
α(1 − 21 α)p−○ 1
× (0.235...) × (1 − 12 × 0.235...) × (100 kPa)
K= 2
= 2
= 0.174...
(1 − α)2 p (1 − 0.235...)2 × (101.9 kPa)
= 0.175
The standard enthalpy of the dimerization reaction is found using the temper-
ature dependence of K [6B.4–201]
∆ r H −○ 1 1 R ln(K 2 /K 1 )
ln K 2 − ln K 1 = − ( − ) hence ∆ r H −○ = −
R T2 T1 (1/T2 ) − (1/T1 )
Taking T1 = 437 K and T2 = 471 K gives
R ln(0.174.../1.35...)
∆ r H −○ = − = −103 kJ mol−1
[1/(471 K)] − [1/(437 K)]
P6B.9 The relationship between K and ∆ r G −○ [6A.15–194], ∆ r G −○ = −RT ln K, implies

that −
○ −
○ −
○ −
○ −
○
K = e−∆ r G /RT = e−(∆ r H −T ∆ r S )/RT = e−∆ r H /RT e∆ r S /R
The ratio of the values of K that would be obtained using the lowest and highest
values of ∆ r H −○ is
−
○ −
○ −
○ −
○
K lowH e−∆ r H low /RT e∆ r S /R ∆ r H low − ∆ r H high
= −∆ H −○ /RT = exp (− )
K highH e r high e∆ r S −○ /R RT
For the given data, the value of this factor is
At 298 K:
K lowH (243 − 289) × 103 J mol−1
= exp (− ) = 1.2 × 108
K highH (8.3145 J K−1 mol−1 ) × (298 K)
At 700 K:
K lowH (243 − 289) × 103 J mol−1
= exp (− ) = 2.7 × 103
K highH (8.3145 J K−1 mol−1 ) × (700 K)
P6B.11 The standard reaction Gibbs energy is related to the standard reaction enthalpy
and entropy according to [3D.9–96], ∆ r G −○ = ∆ r H −○ − T∆ r S −○ . However, ∆ r H −○
and ∆ r S −○ themselves vary with temperature according to Kirchhoff ’s law [2C.7a–
53] for ∆ r H −○ and the analogous equation [3C.5a–91] for ∆ r S −○
T2
∆ r H −○ (T2 ) = ∆ r H −○ (T1 ) + ∫ ∆ r C −p○ dT
T1
T2 ∆ r C −p○
∆ r S −○ (T2 ) = ∆ r S −○ (T1 ) + ∫ dT
T1 T
∆ r C −p○ is defined by [2C.7b–53], ∆ r C −p○ = ∑products νC −p,m

○
− ∑reactants νC −p,m
○
.
Because each species has C p,m given by C p,m = a + bT + c/T it is convenient
2
to write ∆ r C −p○ = A + BT + C/T 2 where A is defined by A = ∑products νa −

∑reactants νa and similarly for B and C. Expressions for ∆ r H −○ and ∆ r S −○ at
temperature T2 are then obtained by performing the integrations
T2
∆ r H −○ (T2 ) = ∆ r H −○ (T1 ) + ∫ A + BT + C/T 2 dT
T1
1 1
= ∆ r H −○ (T1 ) + A(T2 − T1 ) + 21 B (T22 − T12 ) − C ( − )
T2 T1
A + BT + C/T 2 T2
∆ r S −○ (T2 ) = ∆ r S −○ (T1 ) + ∫ dT
T1 T
T2 A C
= ∆ r S −○ (T1 ) + ∫ + B + 3 dT
T1 T T
T 1 1
= ∆ r S −○ (T1 ) + A ln
2
+ B(T2 − T1 ) − 21 C ( 2 − 2 )
T1 T2 T1
The standard reaction Gibbs energy at temperature T2 is then given by
∆ r G −○ (T2 ) = ∆ r H −○ (T2 ) − T2 ∆ r S −○ (T2 )

1 1
= [∆ r H −○ (T1 ) + A(T2 − T1 ) + 21 B (T22 − T12 ) − C ( − )]
T2 T1
T2 1 1
− T2 [∆ r S −○ (T1 ) + A ln + B(T2 − T1 ) − 21 C ( 2 − )]
T1 T2 T12
T
= ∆ r H −○ (T1 ) − T2 ∆ r S(T1 ) + A [T2 − T1 − T2 ln ( )]
2
T1
+ B [ 12 (T22 − T12 ) − T2 (T2 − T1 )]
T2 1 1 1 1
+C[ ( 2 − 2 ) − ( − )]
2 T2 T1 T2 T1
In order to obtain an expression that contains ∆ r G −○ (T1 ), it is necessary to

write the first part of the above expression as ∆ r H −○ (T1 ) − T1 ∆ r S −○ (T1 ) −(T2 −
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
∆ r G −○ (T1 )
T1 )∆ r S −○ (T1 ) so that:
T2
∆ r G −○ (T2 ) = ∆ r G −○ (T1 ) − (T2 − T1 )∆ r S −○ (T1 ) + A [T2 − T1 − T2 ln ( )]
T1
+ B [ 12 (T22 − T12 ) − T2 (T2 − T1 )]
T2 1 1 1 1
+C[ ( − ) − ( − )]
2 T22 T12 T2 T1
The standard Gibbs energy for the formation of H2 O(l) is ∆ r H −○ for the reaction
H2 (g) + 12 O2 (g) → H2 O(l). From the data in the Resource section, at 298 K,
∆ f H −○ (H2 O, l) = −285.83 kJ mol−1 and ∆ f S −○ (H2 O, l) is calculated as
∆ f S −○ = S m
−
○ −
○
(H2 O, l) − S m −
○
(H2 , g) − 21 S m (O2 , g)
= (69.91 − 130.684 − 12 × 205.138) J K−1 mol−1 = −163.343 J K−1 mol−1
The quantities A, B, and C are also calculated using the data from the Resource
section:
A = a H2 O,l − a H2 ,g − 12 a O2 ,g
= (75.29 − 27.28 − 12 × 29.96) J K−1 mol−1 = 33.03 J K−1 mol−1
B = b H2 O,l − b H2 ,g − 12 b O2 ,g
= (0 − 3.26 − 21 × 4.18) × 10−3 J K−2 mol−1 = −5.35 × 10−3 J K−2 mol−1
C = c H2 O,l − c H2 ,g − 21 c O2 ,g
= (0 − 0.50 − 12 × (−1.67)) × 105 J K mol−1 = 0.335 × 105 J K mol−1
Hence, using the expressions derived above with T1 = 298 K and T2 = 372 K:
∆ f H −○ (298 K)
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
∆ f H −○ (372 K) = (−285.83 × 103 J mol−1 )
+ (33.03 J K−1 mol−1 ) × ([372 − 298] K)
+ 12 (−5.35 × 10−3 J K−2 mol−1 ) × [(372 K)2 − (298 K)2 ]
1 1
− (0.335 × 105 J K mol−1 ) × ( − )
372 K 298 K
= −283.49... kJ mol−1
∆ f S −○ (298 K)
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ 372 K
∆ f S −○ (372 K) = (−163.343 J K−1 mol−1 ) +(33.03 J K−1 mol−1 ) × ln ( )
298 K
+ (−5.35 × 10−3 J K−2 mol−1 ) × [(372 K) − (298 K)]
1 1
− 12 (0.335 × 105 J K mol−1 ) × ( − )
(372 K)2 (298 K)2
= −156.34... J K−1 mol−1
∆ f G −○ (372 K) = ∆ f H −○ (372 K) − (372 K) × ∆ f S −○ (372 K)

= (−283.49... × 103 J mol−1 )
− (372 K) × (−156.34... J K−1 mol−1 )
= −225.34 kJ mol−1
This compares to −237.13 kJ mol−1 at 298 K (from the Resource section). Note
that ∆ f H −○ and ∆ r S −○ do not change very much between 298 K and 372 K in
this case. In fact, assuming that they are constant gives almost the same value
of ∆ f G −○ (372 K), as is seen by calculating ∆ f G −○ at 372 K using the values of
∆ f H −○ and ∆ f S −○ at 298 K:
∆ f G −○ (372 K) ≈ ∆ f H −○ (298 K) + (372 K) × ∆ f S −○ (298 K)

= (−285.83 × 103 J mol−1 ) − (372 K) × (−163.343 J K−1 mol−1 )
= −225.07 kJ mol−1
which differs from the value obtained above by less than 0.3 kJ mol−1 .
6C Electrochemical cells
D6C.1 The role of a salt bridge is to minimise the liquid junction potential which
would otherwise occur as a result of the contact between the electrolytes in
the two half cells. For a cell to generate a potential these solutions must be
in electrical contact: the salt bridge achieves this without involving a physical
contact between the two solutions.
D6C.3 When a current is being drawn from an electrochemical cell, the cell potential
is altered by the formation of charge double layers at the surface of electrodes
and by the formation of solution chemical potential gradients (concentration
gradients). Resistive heating of the cell circuits may occur and junction poten-
tials between dissimilar materials both external and external to the cell may
change.
D6C.5 A galvanic cell is an electrochemical cell that produces electricity as a result

of the spontaneous reaction occuring inside it. An electrolytic cell is an elec-
trochemical cell in which a non-spontaneous reaction is driven by an external
source of current.
E6C.1(a) The reduction half-reactions for the cell in question are:
R: Cd2+ (aq) + 2e− → Cd(s) E −○ (R) = −0.40 V

L: AgBr(s) + e− → Ag(s) + Br− (aq) E −○ (L) = +0.0713 V
The cell reaction is obtained by subtracting the left-hand half-reaction from the
right-hand half-reaction, after first multiplying the left-hand half-reaction by
two so that both half-reactions involve the same number of electrons.
Cd2+ (aq) + 2Ag(s) + 2Br− (aq) → Cd(s) + 2AgBr(s)
The cell potential is given by the Nernst equation [6C.4–207]

−
○
E cell = E cell − (RT/νF) ln Q
In this case ν = 2 and the reaction quotient Q is

1
1
³¹¹2¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ µ
a Cd(s) a AgBr(s) 1
Q= =
a Cd2+ (aq) a Ag(s)
2
a Br
2
− (aq) a Cd2+ (s) a Br
2
− (aq)
²
1
where a J = 1 for pure solids has been used. For ions in solution the activity is
written as a = γ± b/b −○ , where γ± is the mean activity coefficient, as established
in Section 5F.4 on page 177. The Debye–Hückel limiting law [5F.27–178], which
applies at low molalities, is
log γ± = −A∣z+ z− ∣I 1/2
where A = 0.502 for an aqueous solution at 298 K, z+ and z− are the charges
on the ions, and I is the dimensionless ionic strength of the solution which
for a solution containing two types of ion at molality b+ and b− is given by
[5F.29–178], I = 12 (b+ z+2 + b− z−2 )/b −○ .
For the cell in question, the right-hand electrode contains a solution of Cd(NO3 )2
of molality b R = 0.010 mol kg−1 . In this case z+ = +2 (for Cd2+ ), z− = −2 (for
NO−3 ), b+ = b Cd2+ = b R and b− = b NO−3 = 2b R . The ionic strength is
I R = 21 (22 × b R + (−1)2 × (2b R )) /b −○ = 3b R /b −○

and the mean activity coefficient for the right-hand electrode is therefore given
by
1 1
1/2 3b R 2 3b R 2
log γ±,R = −A∣z+ z− ∣I R = −A∣(+2)(−1)∣ ( −○ ) = −2A ( −○ )
b b
1
3 × (0.010 mol kg−1 ) 2
= −2 × 0.509 × ( ) = −0.176...
1 mol kg−1
Hence γ±,R = 10−0.176 ... = 0.666..., and so
b Cd2+ 0.010 mol kg−1
a Cd2+ = γ±,R = (0.666...) × = 6.66... × 10−3
b −○ 1 mol kg−1
In a similar way, the left-hand electrode contains a solution of KBr of molality
b L = 0.050 mol kg−1 , so that z+ = +1 (for K+ ), z− = −1 (for Br− ), and b+ = b− =
b L . It follows that
I L = 12 (b+ z+2 + b− z−2 )/b −○ = 12 (12 × b L + (−1)2 × b L ) /b−○ = b L /b−○
and therefore
1 1
1/2 bL 2 bL 2
log γ±,L = −A∣z+ z− ∣I L = −A × ∣(+1) × (−1)∣ × ( −○ ) = −A ( −○ )
b b
1
0.050 mol kg−1 2
= −0.509 × ( ) = −0.113...
1 mol kg−1
Hence γ±,L = 10−0.113 ... = 0.769..., and so
b Br− 0.050 mol kg−1

a Br− = γ±,L = (0.769...) × = 0.0384...
b −○ 1 mol kg−1
Putting these activities into the Nernst equation for the cell with ν = 2 and the
expression for Q obtained above gives
−
○ RT RT 1
E cell = E cell − ln Q = [E −○ (R) − E −○ (L)] − ln ( )
νF ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸¹ ¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¹¶ 2F a a2
Cd 2+ Br−
E cell −
○
(8.3145 J K−1 mol−1 ) × (298 K)

= [(−0.40 V) − (+0.0713 V)] −
2 × (96485 C mol−1 )
1
× ln ( ) = −0.619 V
(6.66... × 10−3 ) × (0.0384...)2
E6C.2(a) The reduction half-reactions for the cell in question are
R: Cu2+ (aq) + 2e− → Cu(s)

L: Zn2+ (aq) + 2e− → Zn(s)
which reveal that ν = 2 for the given cell reaction. The relationship between
∆ r G −○ and E cell
−
○
is given by [6C.2–203], ∆ r G −○ = −νFE cell
−
○
∆ r G −○ = −2 × (96485 C mol−1 ) × (+1.10 V) = −212 kJ mol−1
where 1 C V = 1 J is used.
−○
E6C.3(a) The Nernst equation [6C.4–207] is E cell = E cell − (RT/νF) ln Q. If Q changes
from Q 1 to Q 2 then the change in cell potential is given by
−
○ RT −
○ RT RT Q2
E cell,1 − E cell,2 = [E cell − ln Q 2 ] − [E cell − ln Q 1 ] = − ln ( )
νF νF νF Q1
For ν = 2 and Q 2 /Q 1 = 1/10 the change in cell potential is
(8.3145 J K−1 mol−1 ) × (298 K)

E cell,1 − E cell,2 = − 1
× ln 10 = +0.030 V
2 × (96485 C mol−1 )
where 1 J C−1 = 1 V is used.
E6C.4(a) (i) The reduction half-reactions for the cell Zn(s)|ZnSO4 (aq)||AgNO3 |Ag(s),
together with their standard electrode potentials from the Resource sec-
tion, are
R: Ag+ (aq) + e− → Ag(s) E −○ (R) = +0.80 V
L: Zn2+ (aq) + 2e− → Zn(s) E −○ (L) = −0.76 V
The cell reaction is obtained by subtracting the left-hand reduction half-

reaction from the right-hand reduction half-reaction, after first multiply-
ing the right-hand half-reaction by two so that the numbers of electrons
in both half-reactions are the same
2Ag+ (aq) + Zn(s) → 2Ag(s) + Zn2+ (aq)
The standard cell potential is calculated as the difference of the two stan-
−
○
dard electrode potentials, [6D.3–210], E cell = E −○ (R) − E −○ (L)
−
○
E cell = (+0.80 V) − (−0.76 V) = +1.56 V
(ii) Following the same approach as part (i), and noting that the Pt(s) is an
‘inert metal’ that is only present to act as a source or sink of electrons,
the half-reactions for the Cd(s)|CdCl2 (aq)||HNO3 (aq)|H2 (g)|Pt(s) cell
and their electrode potentials are
R: 2H+ (aq) + 2e− → H2 (g) E −○ (R) = 0 (by definition)
L: Cd2+ (aq) + 2e− → Cd(s) E −○ (L) = −0.40 V
The cell reaction (R − L) is therefore
2H+ (aq) + Cd(s) → H2 (g) + Cd2+ (aq)
and the standard cell potential is
−
○
E cell = 0 − (−0.40 V) = +0.40 V
(iii) For the Pt(s)|K3 [Fe(CN)6 ](aq),K4 [Fe(CN)6 ](aq)||CrCl3 (aq)|Cr(s) cell
the reduction half-reactions are:
R: Cr3+ (aq) + 3e− → Cr(s) E −○ (R) = −0.74 V
L: [Fe(CN)6 ] (aq) + e → [Fe(CN)6 ] (aq) E −○ (L) = +0.36 V
3− − 4−
The cell reaction is obtained by subtracting the left-hand half-reaction

from the right-hand half-reaction, after first multiplying the right-hand
half reaction by three so that both half-reactions involve the same number
of electrons.
Cr3+ (aq) + 3[Fe(CN)6 ]4− (aq) → Cr(s) + 3[Fe(CN)6 ]3− (aq)
The standard cell potential is
−
○
E cell = (−0.74 V) − (+0.36 V) = −1.10 V
E6C.5(a) (i) The required reduction half-reactions are

R: Cu2+ (aq) + 2e− → Cu(s) E −○ (R) = +0.34 V
L: Zn2+ (aq) + 2e− → Zn(s) E −○ (L) = −0.76 V
The cell reaction (R − L) generated from these reduction half-reactions is
Zn(s)+Cu2+ (aq) → Zn2+ (aq)+Cu(s) which is equivalent to the required
reaction. The cell required is
Zn(s)∣ZnSO4 (aq)∣∣CuSO4 (aq)∣Cu(s)
and the standard cell potential is
−
○
E cell = E −○ (R) − E −○ (L) = (+0.34 V) − (−0.76 V) = +1.10 V
(ii) The required reduction half-reactions are:
R: 2AgCl(s) + 2e− → 2Ag(s) + 2Cl− (aq) E −○ (R) = +0.22 V

L: 2H+ (aq) + 2e− → H2 (g) E −○ (L) = 0 (by definition)
2AgCl(s) + H2 (g) → 2Ag(s) + 2Cl− (aq) + 2H+ (aq) which is equivalent
to the required reaction. The required cell is:
Pt(s)∣H2 (g)∣HCl(aq)∣AgCl(s)∣Ag(s)
The Pt(s) electrode is an ‘inert metal’ that acts as an electron source or

sink. Note that there is no interface between the two half cells because
both electrodes have a common electrolyte (HCl). The standard cell po-
tential is
−
○
E cell = E −○ (R) − E −○ (L) = +0.22 V
(iii) The required reduction half reactions are
R: O2 (g) + 4H+ (aq) + 4e− → 2H2 O(l) E −○ (R) = +1.23 V
L: 4H+ (aq) + 4e− → 2H2 (g) E −○ (L) = 0 (by definition)
the required reaction, O2 (g) + 2H2 (g) → 2H2 O(l). The required cell is
Pt(s)∣H2 (g)∣HCl(aq)∣O2 (g)∣Pt(s)
As in part (ii) the platinum electrode is an ‘inert metal’ and there is no

interface between the half cells because they have a common electrolyte.
−
○
E cell = E −○ (R) − E −○ (L) = +1.23 V
An alternative combination of reduction half-reactions is

R: O2 (g) + 2H2 O(l) + 4e− → 4OH− (aq) E −○ (R) = +0.40 V
L: 4H2 O(l) + 4e− → 4OH− (aq) + 2H2 (g) E −○ (L) = −0.83 V
which uses alkaline instead of acidic conditions. The cell required is
Pt(s)∣H2 (g)∣NaOH(aq)∣O2 (g)∣Pt(s)
The overall cell reaction (R − L) is the same and so, therefore, is the stan-
dard cell potential:
−
○
E cell = (+0.40 V) − (−0.83V) = +1.23 V
P6C.1 (a) The reaction of hydrogen and oxygen, 2H2 (g) + O2 (g) → 2H2 O(l), can
be broken down into the reduction half-reactions
R: O2 (g) + 4H+ (aq) + 4e− → 2H2 O(l) E −○ (R) = +1.23 V
L: 4H+ (aq) + 4e− → 2H2 (g) E −○ (L) = 0 (by definition)
The standard cell potential is given by

−
○
E cell = E −○ (R) − E −○ (L) = +1.23 V
(b) The balanced chemical equation for the combustion of butane, C4 H10 (g)+
13
2
O2 (g) → 4CO2 (g) + 5H2 O(g) can be broken down into the reduction
half-reactions
R: 13
2
O2 (g) + 26H+ (aq) + 26e− → 13H2 O(l)
L: 4CO2 (g) + 26H+ (aq) + 26e− → C4 H10 (g) + 8H2 O(l)
The standard electrode potential for the left-hand reduction half-reaction
−
○
is not in the Resource section, so E cell cannot be calculated from E −○ (R) −
−
○ −
○
E (L) as in part (a). Instead ∆ r G is calculated for the cell reaction by
−
○
first using standard Gibbs energies of formation and then using E cell =
−
○ −○
−∆ r G /νF [6C.3–207] to calculate E cell . Note from the above half-reactions
that ν = 26.
∆ r G −○ = 4∆ f G −○ (CO2 , g) + 5∆ f G −○ (H2 O, l) − ∆ f G −○ (C4 H10 , g)

= 4 × (−394.36 kJ mol−1 ) + 5 × (−237.13 kJ mol−1 )
− (−17.03 kJ mol−1 ) = −2746.06 kJ mol−1
Hence
−
○ ∆ r G −○ −2746.06 × 103 J mol−1
E cell =− =− = +1.09 V
νF 26 × (96485 C mol−1 )
P6C.3 The reduction half-reactions for the cell are
R: Q(aq) + 2H+ (aq) + 2e− → QH2 (aq) E −○ (R) = +0.6994 V

L: Hg2 Cl2 (s) + 2e− → 2Hg(l) + 2Cl− (aq) E −○ (L) = +0.27 V
for which ν = 2. The value of E −○ (L) is taken from the Resource section. The cell
reaction (R − L) is
Q(aq) + 2H+ (aq) + 2Hg(l) + 2Cl− (aq) → QH2 (aq) + Hg2 Cl2 (s)
−
○
E cell = E −○ (R) − E −○ (L) = (+0.6994 V) − (+0.27 V) = +0.4294 V
Noting that ν = 2 and that a J = 1 for pure solids and liquids, the Nernst
equation is
−
○ RT a QH2
E cell = E cell − ln ( )
2F aQ aH
2 a2
+ Cl−
Taking a QH2 = a Q , because Q and QH2 are present at the same concentration,
and a H+ = a Cl− gives:
−
○ RT 1 −
○ 2RT −
○ 2RT
E cell = E cell − ln ( 4 ) = E cell + ln a H+ = E cell + ln 10 × log a H+
2F a H+ F F
−
○ 2RT ln 10
= E cell − × pH
F
where pH = − log a H+ and ln x = ln 10 × log x (from inside the front cover)

have been used. Rearranging for pH gives:
F
pH = (E −○ − E cell )
2RT ln 10 cell
96485 C mol−1
=
2 × (8.3145 J K−1 mol−1 ) × (298 K) × ln 10
× [(+0.4294 V) − (+0.190 V)] = 2.0
6D Electrode potentials
D6D.1 This is discussed in Impact 10.
E6D.1(a) The reduction half-reactions for the given cell are
R: Ag+ (aq) + e− → Ag(s)

L: AgI(s) + e− → Ag(s) + I− (aq)
The cell reaction (R − L) is Ag+ (aq) + I− (aq) → AgI(s), with ν = 1. The equi-
−
○
librium constant for this reaction is calculated from E cell using [6C.5–207],
−
○
E cell = (RT/νF) ln K. Rearranging gives
νF −○ 1 × (96485 C mol−1 )
ln K = E cell = × (0.9509 V) = 37.0...
RT (8.3145 J K−1 mol−1 ) × (298.15 K)
where 1 V = 1 J C−1 is used. Hence K = 1.18... × 1016 .

The dissolution reaction, AgI(s) → Ag+ (aq) + I− (aq), corresponds to the re-
verse of the cell reaction as written above. The required equilibrium constant
is therefore the reciprocal of the one just calculated
1
K diss = = 8.445 × 10−17
1.18... × 1016
E6D.2(a) (i) The reduction half-reactions for the specified cell and their corresponding
electrode potentials from the Resource section are
R: Cu2+ (aq) + 2e− → Cu(s) E −○ (R) = +0.34 V
L: 2Ag+ (aq) + 2e− → 2Ag(s) E −○ (L) = +0.80 V
The overall cell reaction is
Cu2+ (aq) + 2Ag(s) → Cu(s) + 2Ag+ (aq) ν=2

−
○
E cell = E −○ (R) − E −○ (L) = (+0.34 V) − (+0.80 V) = -0.46 V
The reaction Gibbs energy is related to the cell potential according to

[6C.3–207], ∆ r G −○ = −νFE cell
−
○
. Therefore
∆ r G −○ = −νFE cell
−
○
= −2 × (96485 C mol−1 ) × (−0.46 V) = 88.7... kJ mol−1
= +89 kJ mol−1
The standard reaction enthalpy is calculated using standard enthalpies of

formation from the Resource section, noting that elements in their refer-
ence states have ∆ f H −○ = 0.
∆ r H −○ = 2∆ f H −○ (Ag+ , aq) − ∆ f H −○ (Cu2+ , aq)

= 2 × (+105.58 kJ mol−1 ) − (+64.77 kJ mol−1 ) = +146.39 kJ mol−1
(ii) The standard entropy change of reaction is obtained from ∆ r G −○ and ∆ r H −○

using [3D.9–96], ∆ r G −○ = ∆ r H −○ − T∆ r S −○ . Rearranging for ∆ r S −○ gives
∆ r H −○ − ∆ r G −○ (146.39 × 103 J mol−1 ) − (88.7... × 103 J mol−1 )

∆ r S −○ = =
T 298 K
2 −1 −1
= +1.93... × 10 J K mol
The value of ∆ r G −○ at 308 K is then calculated using [3D.9–96], ∆ r G −○ =

∆ r H −○ − T∆ r S −○ , assuming that ∆ r H −○ and ∆ r S −○ do not vary significantly
with temperature over this range
∆ r G −○ = (+146.39 × 103 J mol−1 ) − (308 K) × (+1.93... × 102 J K−1 mol−1 )

= +87 kJ mol−1
E6D.3(a) Assuming that the mercury forms Hg2 SO4 (s) in the reaction, the required
reduction half-equations and the corresponding standard electrode potentials
are
R: Zn2+ (aq) + 2e− → Zn(s) E −○ (R) = −0.76 V

L: Hg2 SO4 (s) + 2e− → 2Hg(l) + SO2−
4 (aq) E −○ (L) = +0.62 V
The cell reaction is Zn2+ (aq) + SO2−

4 (aq) + 2Hg(l) → Zn(s) + Hg2 SO4 (s), and
the standard cell potential is
−
○
E cell = E −○ (R) − E −○ (L) = (−0.76 V) − (+0.62 V) = −1.38 V
−
○
The negative value of E cell indicates that the cell reaction as written will not be
spontaneous. This means that no , mercury cannot produce zinc metal from
aqueous zinc sulfate under standard conditions.
E6D.4(a) (i) The following electrodes are combined

R: Sn4+ (aq) + 2e− → Sn2+ (aq) E −○ (R) = +0.15 V
L: Sn2+ (aq) + 2e− → Sn(s) E −○ (L) = −0.14 V
The overall cell reaction (R−L) is therefore Sn4+ (aq)+Sn(s) → 2Sn2+ (aq),
which is the required reaction, and has ν = 2. The standard cell potential
−
○
is given by [6D.3–210], E cell = E −○ (R) − E −○ (L)
−
○
E cell = (+0.15 V) − (−0.14 V) = +0.29V
The relationship between the equilibrium constant and the standard cell
−
○
potential is given by [6C.5–207], E cell = (RT/νF) ln K. Rearranging gives
νF −○ 2 × (96485 C mol−1 )
ln K = E cell = × (+0.29 V) = 22.5...
RT (8.3145 J K−1 mol−1 ) × (298 K)
where 1 V = 1 J C−1 is used. Hence K = 6.4 × 109 .

(ii) The following electrodes are combined
R: 2AgCl(s) + 2e− → 2Ag(s) + 2Cl− (aq) E −○ (R) = +0.22 V
L: Sn2+ (aq) + 2e− → Sn(s) E −○ (L) = −0.14 V
The cell reaction is 2AgCl(s) + Sn(s) → 2Ag(s) + 2Cl− (aq) + Sn2+ (aq)
which is equivalent to the required reaction, and has ν = 2. Therefore,
using the same equations as in part (i)
−
○
E cell = E −○ (R) − E −○ (L) = (+0.22 V) − (−0.14 V) = +0.36 V
νF −○ 2 × (96485 C mol−1 )
ln K = E cell = × (+0.36 V) = 28.0...
RT (8.3145 J K−1 mol−1 ) × (298 K)
Hence K = 1.5 × 1012
P6D.1 The given reaction can be broken down into the following reduction half equa-
tions
R: 2Fe3+ (aq) + 2e− → 2Fe2+ (aq)
L: Ag2 CrO4 (s) + 2e− → 2Ag(s) + CrO2−
4 (s)
where the K+ and Cl− spectator ions have been ignored. These half-equations
show that ν = 2 for the given reaction.
(a) The standard potential is calculated from the standard reaction Gibbs
−
○
energy using [6C.3–207], E cell = −∆ r G −○ /νF.
−
○ −∆ r G −○ −62.5 × 103 J mol−1
E cell =− =− = 0.323... V = +0.324 V
νF 2 × (96485 C mol−1 )
−
○
(b) The standard potential of the Ag2 CrO4 /Ag,CrO2− 4 couple, equal to E (L)
−
○
of the cell considered above, is calculated from E cell and the known value
of E −○ (R) using [6D.3–210], E cell
−
○
= E −○ (R) − E −○ (L). The value of E −○ (R),
−
○
in this case E (Fe /Fe ), is taken from the Resource section.
3+ 2+
E −○ (L) = E cell
−
○
− E −○ (R) = (+0.77 V) − (+0.323... V) = +0.45 V
P6D.3 (a) The equilibrium HCO−3 (aq) ⇌ CO2− +

3 (aq) + H (aq) is broken down into
the following reduction half-reactions
R: HCO−3 (aq) + e− → 21 H2 (g) + CO2−
3 (aq)
L: H+ (aq) + e− → 21 H2 (g)
The standard cell potential for this cell is given by
−
○
E cell = E −○ (R) − E −○ (L) = E −○ (HCO−3 /CO2− −
○ +
3 , H2 ) − E (H /H2 )
The standard electrode potential of the H+ /H2 electrode is zero by defi-

nition, so it follows that E −○ (HCO−3 /CO2− −
○ −
○
3 , H2 ) = E cell . The value of E cell
−
○ −
○
is calculated using [6C.3–207], E cell = −∆ r G /νF, noting that ν = 1. The
value of ∆ r G −○ is calculated using the data in the question, noting from
Section 3D.2(a) on page 97 that ∆ f G −○ (H+ , aq) = 0.
∆ r G −○ = ∆ f G −○ (CO2− −
○ −
3 , aq) − ∆ f G (HCO3 , aq)
= (−527.81 kJ mol−1 ) − (−586.77 kJ mol−1 ) = +58.96 kJ mol−1
Hence
−
○ ∆ r G −○ 58.96 × 103 J mol−1
E cell =− =− = −0.6111 V
νF 1 × (96485 C mol−1 )
As shown above, this is equal to E −○ (HCO−3 /CO2−

3 , H2 ).
(b) The reaction Na2 CO3 (aq) + H2 O(l) → NaHCO3 (aq) + NaOH(aq) is bro-
ken down into the following reduction half-equations, in which the Na+
counterions are ignored because they play no part in the reaction. The
value of E −○ (L) is as calculated in part (a), and E −○ (R) is taken from the
Resource section.
R: H2 O(l) + e− → 21 H2 (g) + OH− (aq) E −○ (R) = −0.83 V
L: HCO−3 (aq) + e− → 21 H2 (g) + CO2−
3 (aq) E −○ (L) = −0.611... V
The standard cell potential is given by
−
○
E cell = E −○ (R)−E −○ (L) = (−0.83 V)−(−0.611... V) = −0.218... V = −0.22 V
(c) The cell reaction for the cell considered in part (b) is
− −
CO2−
3 (aq) + H2 O(l) → HCO3 (aq) + OH (aq) ν=1
It is assumed that a H2 O = 1 because solvent water is close to being in its

standard state. Therefore the Nernst equation is
−
○ RT ⎛ a HCO−3 a OH− ⎞
E cell = E cell − ln
F ⎝ a CO2−
3
⎠
(d) The standard cell potential corresponds to all species involved in the cell
reaction, which includes OH− , being present at unit activity. This means
that the pH will need to be approximately 14, in order to give a OH− = 1. At
pH 7.0, the concentration of OH− will be lower than at pH 14, which will
mean that the cell reaction as written above will have a greater tendency
to move in the forward direction. As a result E cell is predicted to be larger
at pH 7.0 than when a OH− = 1.
Assuming that the activities of all other species remain the same, the
change in cell potential on going from a OH− = 1 to pH 7 is
⎡ ⎤ ⎡ ⎤
⎢ −○ RT ⎛ a HCO−3 a OH− ⎞⎥⎥ ⎢⎢ −○ RT ⎛ a HCO−3 × 1 ⎞⎥⎥
∆E cell = ⎢⎢E cell − ln − E − ln
⎢ F ⎝ a CO2− ⎠⎥⎥ ⎢⎢ cell F ⎝ a CO2− ⎠⎥⎥
⎣ 3 ⎦ ⎣ 3 ⎦
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
E cell at pH=7 E cell at a OH− =1
RT RT ln 10
=− ln a OH− = − log a OH−
F F
where ln x = ln 10 × log x from inside the front cover is used. To relate
a OH− to the pH, use the relation K w = a H+ a OH− so that
Kw
a OH− = hence log a OH− = log K w − log a H+ = −pK w + pH
a H+
where pH = − log a H+ and pK w = − log K w . Taking K w = 1.00 × 10−14 , or
pK w = 14.0, the change in cell potential when the pH is changed to 7 is
therefore
RT ln 10
∆E cell = − (pH − pK w )
F
(8.3145 J K−1 mol−1 ) × (298 K) × ln 10
=− × (−14.0 + 7.0)
(96485 C mol−1 )
= +0.4139 V
Therefore the cell potential has increased on going from a OH− = 1 to pH 7.
P6D.5 The relationship between ∆ r S −○ and E cell
−
○ −○
is given by [6C.6–208], dE cell /dT =
−
○ −
○
∆ r S /νF. If it is assumed that ∆ r S is independent of temperature over the
−
○
range of interest, integration of dE cell = (∆ r S −○ /νF)dT between T1 and T2 gives
∆ r S −○ −
○ −
○
(T2 − T1 ) E cell (T2 ) − E cell (T1 ) = −
νF
−
○
where E cell (T) is the potential at temperature T. This equation is conveniently
written as ∆ r S −○ = νF∆E cell
−
○
/∆T.
−
○
∆E cell
∆ r S −○ = νF ×
∆T
(+1.2251 V)−(+1.2335 V)
= 4×(96485 C mol−1 )× = −324 J K−1 mol−1
(303 K)−(293 K)
The standard reaction enthalpy is then calculated from [3D.9–96], ∆ r G −○ =
∆ r H −○ − T∆ r S −○ , with ∆ r G −○ being given by [6C.3–207], ∆ r G −○ = −νFE cell
−
○
.
∆ r H −○ = ∆ r G −○ + T∆ r S −○ = −νFE cell
−
○
+ T∆ r S −○
= −4 × (96485 C mol−1 ) × (+1.2335 V)
+ (293 K) × (−3.24... × 102 J K−1 mol−1 ) = −571 kJ mol−1
In calculating ∆ r H −○ , the value of E cell

−
○
at 293 K has been used. However, because
−
○
∆ r S has been assumed to be constant over the temperature range, the data at
303 K will give the same value for ∆ r H −○ .

I6.1 The reduction half-reactions, and the overall cell reaction, for the specified cell
(R − L) are:
R: AgCl(s) + e− → Ag(s) + Cl− (aq)

L: H+ (aq) + e− → 12 H2 (g)
R − L: AgCl(s) + 21 H2 (g) → Ag(s) + Cl− (aq) + H+ (aq) ν=1
Noting that a J = 1 for pure solids and that in this cell a H2 = 1 because the
hydrogen is at standard pressure, the Nernst equation is
−
○ RT
E cell = E cell − ln (a Cl− a H+ )
F
In addition, the base B and its conjugate acid are in equilibrium:
a B a H+
BH+ (aq) ⇌ B(aq) + H+ (aq) Ka =
a BH+
The expression for K a is rearranged to give a H+ = K a a BH+ /a B and this is sub-
stituted into the Nernst equation to give
−
○ RT −
○ RT a Cl− a BH+ K a
E cell = E cell − ln (a Cl− a H+ ) = E cell − ln ( )
F F aB
Replacing activities by a J = γ J (b J /b −○ ) [5F.14–175] gives
−
○ RT (γ Cl− b Cl− /b −○ )(γ BH+ b BH+ /b −○ )K a
E cell = E cell − ln ( )
F (γ B b B /b −○ )
In this case b Cl− = b BH+ = b B so the Nernst equation simplifies to
−
○ RT γ Cl− γ BH+ bK a −
○ RT γ 2 bK a
E cell = E cell − ln ( × −○ ) = E cell − ln ( ± −○ )
F γB b F b
where the mean activity coefficient of the BH+ and Cl – ions is given by [5F.22–
177], γ± = (γ Cl− γ BH+ )1/2 and the neutral base B is assumed to be an ideal solute
so that γ B = 1. Noting from inside the front cover that ln x = ln 10 log x, the
Nernst equation becomes
−
○ RT ln 10 γ 2 bK a
E cell = E cell − log ( ± −○ )
F b
−
○ RT ln 10 b
= E cell − (2 log γ± + log ( −○ ) − pK a )
F b
where pK a = − log K a has been used. Next the Davies equation [5F.30b–179],
log γ± = −A∣z+ z− ∣I 1/2 /(1+BI 1/2 )+CI, is used to substitute for log γ± . The ionic
strength I is given by [5F.28–178], I = 21 ∑ i z 2i (b i /b −○ ), where z i is the charge
on ion species i and the sum extends over all the ions present in the solution.
In this case, b BH+ = b Cl− = b, and b H+ is neglected because it will be much
smaller on account of the equilibrium involving the base B. Therefore the ionic
strength is
−
○
I = 21 (z BH
2
+ b + z Cl− b) /b
2
= 21 (12 × b + (−1)2 × b) /b −○ = b/b−○
and therefore
A∣z BH+ × z Cl− ∣I 1/2 A(b/b −○ )1/2 b

log γ± = − + CI = − −○
+ C ( −○ )
1 + BI 1/2 1 + B(b/b ) 1/2 b
Substitution of this expression into the Nernst equation derived above gives
−
○ RT ln 10 A(b/b−○ )1/2 b b
E cell = E cell − (2 [− −○
+ C ( −○ )] + log ( −○ ) − pK a )
F 1 + B(b/b ) 1/2 b b
which rearranges to
−○
F(E cell − E cell ) 2A(b/b−○ )1/2 b b
= − 2C ( −○ ) − log ( −○ ) + pK a
RT ln 10 1 + B(b/b −○ )1/2 b b
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
y
Defining (b/b −○ )1/2 as x and the left-hand side as y, and introducing A = 0.5091
gives
1.0182x
y= − 2Cx 2 − 2 log x + pK a
1 + Bx
which is fitted to the data using mathematical software to give the following val-
ues for the parameters: B = 2.54 , C = −0.204 , and pK a = 6.74 . These values
have been used to draw the line on the graph shown in Fig. 6.3.
b/mmol kg−1 E cell /V (b/b−○ )1/2 y

0.01 0.744 52 0.100 8.823 73
0.02 0.728 53 0.141 8.553 44
0.03 0.719 28 0.173 8.397 08
0.04 0.713 14 0.200 8.293 30
0.05 0.708 09 0.224 8.207 94
0.06 0.703 80 0.245 8.135 42
0.07 0.700 59 0.265 8.081 16
0.08 0.697 90 0.283 8.035 69
0.09 0.695 71 0.300 7.998 67
0.10 0.693 38 0.316 7.959 29
8.8
)/RT ln 10
8.6
F(E cell − E cell

−
○ 8.4
8.2
8.0
0.05 0.10 0.15 0.20 0.25 0.30 0.35
−
○ 1/2
(b/b )
Figure 6.3
I6.3 From Impact 9 the reaction for the hydrolysis of ATP to ADP and inorganic
phosphate P−i is
ATP(aq) + H2 O(l) → ADP(aq) + P−i (aq) + H3 O+ (aq)
Under biological standard conditions, that is, pH = 7, the standard reaction

Gibbs energy at 37 ○ C is given in Impact 9 as ∆ r G ⊕ = −31 kJ mol−1 .
In an environment in which pH = 7.0 and the ATP, ADP and P−i concentrations
are all 1.0 mmol dm−3 , the reaction Gibbs energy is given by [6A.11–193],
∆ r G = ∆ r G ⊕ + RT ln Q ⊕
where Q ⊕ is the reaction quotient calculated relative to the biological standard

state. Because pH is defined by pH = − log a H3 O+ , pH 7 corresponds to a H3 O+ =
10−7 so that when computing Q ⊕ the activity of H3 O+ is measured relative to
an activity of 10−7 rather than an activity of 1 as is usually the case. In practice
this means that a H3 O+ is replaced by (a H3 O+ /10−7 ) in the expression for Q ⊕ .
For the ATP hydrolysis reaction this gives
a ADP × a P−i × (a H3 O+ /10−7 )

∆ r G = ∆ r G ⊕ + RT ln ( )
a ATP × a H2 O
Water is a pure liquid so a H2 O = 1, and for the environment specified in the

question, pH = 7 so a H3 O+ = 10−7 . For the other species activities are approxi-
mated by concentrations according to a J = [J]/c −○ where c −○ = 1 mol dm−3 .
([ADP]/c −○ )([P−i ]/c −○ )(10−7 /10−7 )

∆ r G = ∆ r G ⊕ + RT ln ( )
([ATP]/c −○ )
[ADP][P−i ]
= ∆ r G ⊕ + RT ln ( )
[ATP]c −○
= (−31 × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × ([37 + 273.15] K)
(1.0 × 10−3 mol dm−3 ) × (1.0 × 10−3 mol dm−3 )
× ln ( )
(1.0 × 10−3 mol dm−3 ) × (1 mol dm−3 )
= −49 kJ mol−1
This is to be compared with the value under standard biological conditions,

which is −31 kJ mol−1 , and also with the value under the usual standard con-
ditions. The difference between ∆ r G −○ and ∆ r G ⊕ is that the former has a H+ = 1
and the latter has a H+ = 10−7 . Given that
a ADP a Pi a H3 O+
∆ r G = ∆ r G −○ + RT ln ( )
a ATP a H2 O
setting all the activities to 1 except for that for H3 O+ which is set to 10−7 gives
∆r G ⊕
∆ r G ⊕ = ∆ r G −○ + RT ln 10−7
hence
∆ r G −○ = ∆ r G ⊕ − RT ln 10−7
= (−31 × 103 J mol−1 )−(8.3145 J K−1 mol−1 )×([37 + 273.15] K)×ln 10−7
= +11 kJ mol−1
I6.5 A bacterium could potentially oxidise ethanol to ethanal, ethanoic acid, or

CO2 (g), while nitrate, NO−3 (aq), could potentially be reduced to a number
of possible species including NO2 (g), NO−2 (aq), NO(g), N2 (g), or NH+4 (aq).
Assuming complete oxidation of ethanol to CO2 and complete reduction of
NO−3 to NH+4 the reduction half-reactions are
R: NO−3 (aq) + 10H+ + 8e− → NH+4 (aq) + 3H2 O(l)

L: 2CO2 (g) + 12H+ + 12e− → CH3 CH2 OH(aq) + 3H2 O(l)
The right-hand half reaction is multiplied by three and the left-hand half re-
action by two in order that both involve the same number of electrons. The
overall reaction is
2CH3 CH2 OH(aq) + 3NO−3 (aq) + 6H+ (aq) → 4CO2 (g) + 3NH+4 (aq) + 3H2 O(l)
The data in the Resource section is used to calculate ∆ r G −○ for this reaction:
∆ r G −○ = 4∆ f G −○ (CO2 , g) + 3∆ f G −○ (NH+4 , aq) + 3∆ f G −○ (H2 O, l)
− 2∆ f G −○ (CH3 CH2 OH, aq) − 3∆ f G −○ (NO−3 , aq) − 6∆ f G −○ (H+ , aq)
= 4 × (−394.36 J mol−1 ) + 3 × (−79.31 J mol−1 ) + 3 × (−237.13 J mol−1 )
− 2 × (−174.78 J mol−1 ) − 3 × (−108.74 J mol−1 ) = −1851 kJ mol−1
The negative value of ∆ r G −○ indicates that the reaction is exergonic, so yes , a
bacterium could evolve to use this reaction to drive endergonic processes such
as the formation of ATP for use in cellular processes.
The calculation is valid under standard conditions, which includes a H+ = 1
(pH = 0). As explained in Impact 9 on the website of this text, pH = 0 is
not normally appropriate for biological conditions so it is common to adopt
the biological standard state in which pH = 7.0. The reaction Gibbs energy
for the oxidation of ethanol by nitrate under standard biological conditions is
calculated by using the appropriate value of a H+ in [6A.11–193], ∆ r G = ∆ r G −○ +
RT ln Q, leaving all other species with a J = 1. The reaction consumes six
moles of H+ so under standard biological conditions Q = 1/a H 6
+ . Noting from
inside the front cover that ln x = ln 10 log x, and also that pH = − log a H+ , and
assuming T = 298 K, gives
1
∆ r G = ∆ r G −○ + RT ln Q = ∆ r G −○ + RT ln ( )
aH
6
+
= ∆ r G −○ − 6RT ln 10 log(a H+ ) = ∆ r G −○ + 6RT ln 10×pH

= (−1.85... × 106 J mol−1 ) + 6×(8.3145 J K−1 mol−1 )×(298 K)×ln 10×7
= −1611 kJ mol−1
Thus the reaction remains exergonic under standard biological conditions.
I6.7 (a) The standard reaction enthalpy is found using the van ’t Hoff equation
[6B.2–200]:
d ln K ∆ r H −○ d ln K ∆ r H −○
= which can also be written − =
dT RT 2 d(1/T) R
The second form implies that a graph of − ln K against 1/T should be a
straight line of slope ∆ r H −○ /R, from which ∆ r H −○ can be determined. The
value of ∆ r S −○ is found by combining ∆ r G −○ = ∆ r H −○ − T∆ r S −○ [3D.9–96]
and ∆ r G −○ = −RT ln K [6A.15–194]. Equating these expressions for ∆ r G −○
gives
∆ r H −○ 1 ∆ r S −○
−RT ln K = ∆ r H −○ − T∆ r S −○ hence − ln K = −
R T R
Assuming that ∆ r H −○ and ∆ r S −○ do not very significantly over the temper-
ature range of interest this equation implies that a plot of − ln K against
1/T should be a straight line of intercept −∆ r S −○ /R, from which ∆ r S −○ can
be determined; such a plot is shown in Fig. 6.4. The plot will have a slope
of ∆ r H −○ /R, as already deduced above.
T/K K 1/(T/K) − ln K
233 4.13 × 108 0.004 29 −19.8
248 5.00 × 107 0.004 03 −17.7
258 1.45 × 107 0.003 88 −16.5
268 5.37 × 106 0.003 73 −15.5
273 3.20 × 106 0.003 66 −15.0
280 9.62 × 105 0.003 57 −13.8
288 4.28 × 105 0.003 47 −13.0
295 1.67 × 105 0.003 39 −12.0
303 6.02 × 104 0.003 30 −11.0
−10
− ln K
−15
−20
0.0032 0.0036 0.0040 0.0044
1/(T/K)
Figure 6.4
− ln K = −8 787 × 1/(T/K) + 17.62
∆ r H −○ /R is determined from the slope
∆ r H −○ = R × (slope × K) = (8.3145 J K−1 mol−1 ) × (−8787 K)

= −73.0... kJ mol−1 = −73.1 kJ mol−1
−∆ r S −○ /R is determined from the intercept
∆ r S −○ = −R × intercept = −(8.3145 J K−1 mol−1 ) × (+17.62)

= −1.46... × 102 J K−1 mol−1 = −147 J K−1 mol−1
(b) The standard reaction enthalpy for the reaction 2ClO(g) → (ClO)2 (g) is
expressed in terms of standard enthalpies of formation
∆ r H −○ = ∆ f H −○ [(ClO)2 , g] − 2∆ f H −○ (ClO, g)
Hence
∆ f H −○ [(ClO)2 , g] = ∆ r H −○ + 2∆ f H −○ (ClO, g)
= (−73.0... kJ mol−1 ) + 2 × (+101.8 kJ mol−1 )
= +131 kJ mol−1
Similarly
∆ r S −○ = S m
−
○ −
○
[(ClO)2 , g] − 2S m (ClO, g)
Hence
−
○
Sm [(ClO)2 , g] = ∆ r S −○ + 2S m
−
○
(ClO, g)
= (−1.46... × 102 J K−1 mol−1 ) + 2 × (226.6 J K−1 mol−1 )
= 307 J K−1 mol−1
I6.9 (a) The ionic strength is given by [5F.29–178], I = 12 (b+ z+2 + b− z−2 ) /b −○ , where
z+ and z− are the charges on the ions. For the CuSO4 compartment,
z+ = 2, z− = −2, and b+ = b− = b CuSO4 :
I = 21 (b+ z+2 + b− z−2 ) /b−○ = 12 [b CuSO4 × (+2)2 + b CuSO4 × (−2)2 ] /b −○

1.00 × 10−3 mol kg−1
= 4(b CuSO4 /b−○ ) = 4 × = 4.00 × 10−3
1 mol kg−1
Because the charges are the same for ZnSO4 it follows that I = 4(b ZnSO4 /b−○ )
= 1.20 × 10−2 .
(b) According to the Debye–Hückel limiting law (Section 5F.4(b) on page
177), the mean activity coefficient is given by [5F.27–178], log γ± = −A∣z+ z− ∣I 1/2 ,
where A = 0.509 for aqueous solutions at 25 ○ C. For the CuSO4 solution
log γ±,CuSO4 = −(0.509) × ∣(2) × (−2)∣ × (4.00 × 10−3 )1/2 = −0.128...
Hence γ±,CuSO4 = 10−0.128 ... = 0.743... = 0.743 . For the ZnSO4 solution
log γ±,ZnSO4 = −(0.509) × ∣(2) × (−2)∣ × (1.20 × 10−2 )1/2 = −0.223...
Hence γ±,ZnSO4 = 10−0.223 ... = 0.598... = 0.598 .

(c) Noting that pure solids have a J = 1 and writing the activities of ions in
solution as a = γ± (b/b−○ ), the reaction quotient for the reaction
Zn(s) + Cu2+ (aq) → Zn2+ (aq) + Cu(s)
is given by
a Zn2+ γ±,ZnSO4 (b Zn2+ /b −○ ) γ±,ZnSO4 b Zn2+

Q= = = ×
a Cu2+ γ±,CuSO4 (b Cu2+ /b −○ ) γ±,CuSO4 b Cu2+
0.598... 3.00 × 10−3 mol kg−1
= × = 2.41... = 2.41
0.743... 1.00 × 10−3 mol kg−1
(d) The reaction is thought of as being composed of the reduction half-reactions

R: Cu2+ (aq) + 2e− → Cu(s)
L: Zn2+ (aq) + 2e− → Zn(s)
which show that ν = 2 for this reaction. The standard cell potential is
calculated from ∆ r G −○ using [6C.3–207], E cell
−
○
= −∆ r G −○ /νF:
−
○ ∆ r G −○ −212.7 × 103 J mol−1
E cell =− =− = +1.102 V
νF 2 × (96485 C mol−1 )
Note that 1 J C−1 = 1 V.

(e) The cell potential is given by the Nernst equation [6C.4–207]:
−
○ RT
E cell = E cell − ln Q
νF
(8.3145 J K−1 mol−1 ) × ([25 + 273.15] K)
= (+1.102 V) − × ln(2.41...)
2 × (96485 C mol−1 )
= +1.09 V
I6.11 The reaction for the autoprotolysis of liquid water is
H2 O(l) → H+ (aq) + OH− (aq)
This reaction is split into the reduction half-reactions
R: H2 O(l) + e− → 21 H2 (g) + OH− (aq)

L: H+ (aq) + e− → 21 H2 (g)
Because the standard electrode potential for the left-hand half-reaction is zero
by definition, the standard cell potential for this cell is equal to the standard
electrode potential of the H2 O/H2 ,OH− electrode. The equilibrium constant
−
○
K w for the cell reaction is then given by [6C.5–207], E cell = (RT/νF) ln K.
Rearranging for ln K and using ν = 1, K = K w , and E cell = E −○ (H2 O/H2 , OH− )
−
○
gives
F
ln K w = × E −○ (H2 O/H2 , OH− )
RT
Noting from inside the front cover that ln x = ln 10 × log x, and also that pK w =
− log K w allows the above equation to be rewritten as
ln K w F
pK w = − log K w = − =− × E −○ (H2 O/H2 , OH− )
ln 10 RT ln 10
The task is therefore to find E −○ (H2 O/H2 , OH− ) from the given data. To do
this, the specified cell is written in terms of its reduction half-reactions
R: AgCl(s) + e− → Ag(s) + Cl− (aq)

L: H2 O(l) + e− → 21 H2 (g) + OH− (aq)
The cell reaction, which has ν = 1, is
AgCl(s) + 12 H2 (g) + OH− (aq) → Ag(s) + Cl− (aq)
Noting that a J = 1 for pure solids, and that in this cell a H2 = 1 because the
hydrogen is at standard pressure, the Nernst equation for the cell is
−
○ RT −
○ RT a Cl−
E cell = E cell − ln Q = E cell − ln ( )
νF F a OH−
Writing the activities as a = γ± (b/b −○ ), the Nernst equation becomes
−
○ RT γ± (b Cl− /b −○ ) −
○ RT b Cl−
E cell = E cell − ln ( −○
) = E cell − ln ( )
F γ± (b OH− /b ) F b OH−
The standard cell potential is split into contributions from the two electrodes
−
○
using [6D.3–210], E cell = E −○ (R) − E −○ (L)
RT b Cl−
E cell = E −○ (AgCl/Ag, Cl− ) − E −○ (H2 O/H2 , OH− ) − ln ( )
F b OH−
Hence
RT b Cl−
E −○ (H2 O/H2 , OH− ) = E −○ (AgCl/Ag, Cl− ) − E cell − ln ( )
F b OH−
This equation is used with b Cl− = 0.01125 mol kg−1 , b OH− = 0.0100 mol kg−1 ,
and the values of E cell and E −○ (AgCl/Ag, Cl− ) to calculate E −○ (H2 O/H2 , OH− )
at each temperature. The relation derived earlier
pK w = −(F/RT ln 10)E −○ (H2 O/H2 , OH− )
is then used to calculate pK w . The results are given in the following table.
θ/○ C T/K E cell /V E −○ (AgCl/Ag, Cl− )/V E −○ (H2 O/H2 , OH− )/V pK w
20.0 293.15 1.047 74 0.225 02 −0.825 70 14.20
25.0 298.15 1.048 64 0.222 30 −0.829 37 14.02
30.0 303.15 1.049 42 0.219 59 −0.832 91 13.85
To find ∆ r S −○ for the autoprotolysis, the relationship between E cell

−○
and temper-
−
○ −○ −
○
ature [6C.6–208], dE cell /dT = ∆ r S /νF is used. If ∆ r S is constant over the
−
○
temperature range this equation implies that a plot of E cell against T should
be a straight line of slope ∆ r S −○ /νF. In this case E cell
−○
for the autoprotolysis
reaction is equal to E −○ (H2 O/H2 , OH− ) as explained earlier. The plot is shown
in Fig. 6.5.
E −○ (H2 O/H2 , OH− )/V = −7.229 × 10−4 × (T/K) − 0.6137

−0.825
E −○ (H2 O/H2 , OH− )/V

−0.830
−0.835
292 294 296 298 300 302 304
T/K
Figure 6.5
∆ r S −○ /νF is determined from the slope
∆ r S −○ = νF × slope × V K−1
= 1 × (96485 C mol−1 ) × (−7.229 × 10−4 V K−1 )
= −68.5... J K−1 mol−1 = −68.6 J K−1 mol−1
The standard enthalpy change for the autoprotolysis is calculated from [3D.9–
96], ∆ r G −○ = ∆ r H −○ − T∆ r S −○ , with ∆ r G −○ being given by [6C.2–203], ∆ r G −○ =
−
○ −
○
−νFE cell . In this case E cell = E −○ (H2 O/H2 , OH− ) and ν = 1. Using the value
○
for 25.0 C gives
∆ r H −○ = ∆ r G −○ + T∆ r S −○ = −νFE −○ (H2 O/H2 , OH− ) + T∆ r S −○

= −1 × (96485 C mol−1 ) × (−0.829...V) + (298.15 K) × (−68.5... J K−1 mol−1 )
= +59.6 kJ mol−1
7 Quantum theory
7A The origins of quantum mechanics

Answers to discussion question
D7A.1 By wave-particle duality it is meant that in some experiments an entity be-
haves as a wave while in other experiments the same entity behaves as a par-
ticle. Electromagnetic radiation behaves as a wave in diffraction experiments
but it behaves as particulate photons in absorption and emission spectroscopy.
Electrons behave as waves in diffraction experiments, but as particles in the
photoelectric effect.
The development of quantum theory is much concerned with the need to em-
brace this wave-particle duality and, as is explained in the following Topics,
this is exemplified by the introduction of the wavefunction to describe the
properties of a particles and the notion of ‘complementary variables’ such as
position and momentum.
D7A.3 The ultimately unsuccessful classical approach to the description of black-body

radiation involved assuming that the radiation resulted from oscillating electric
charges in the walls of the body, and that each oscillator has the same average
energy as predicted by the equipartition principle. This view results in the
ultra-violet catastrophe, in which the radiation increases without limit as the
wavelength becomes shorter.
Planck assumed two things: first, that the oscillators could only have energies
given by E = nhν, where ν is the frequency and n is 0, 1, 2, . . .; second, that the
probability of an individual oscillator having a particular energy is described
by the Boltzmann distribution. As the frequency of the oscillator or the value
of n increases, so does its energy and the Boltzmann distribution predicts that
such a state is less likely. In addition, the highest frequency oscillations may
not be excited at all, that is have n = 0, on the grounds that the resulting state
has too high an energy to be populated. Planck’s theory therefore avoids the
ultraviolet catastrophe by having no excitation of highest frequency oscillators.
E7A.1(a) If the power, P, is constant, the total energy emitted in time ∆t is P∆t. The
energy of each emitted photon is E photon = hν = hc/λ. The total number of
photons emitted in this time period is therefore the total energy emitted divided
238 7 QUANTUM THEORY
by the energy per photon
N = P∆t/E photon = P∆tλ/hc
The conservation of linear momentum requires that the loss of a photon must
impart an equivalent momentum in the opposite direction to the glow-worm,
hence the total momentum p imparted to the glow-worm in time ∆t is
p = N p photon = N h/c = (P∆tλ/hc) × (h/λ) = P∆t/c
Because p = (mυ)glow-worm , the final speed of the glow-worm is
υ = P∆t/cmglow-worm
(0.10 W) × (10 y) × (3.1536 × 107 s y−1 )
= = 21 m s−1
(2.9979 × 108 m s−1 ) × (0.0050 kg)
Noting that the number of seconds in one year is
365 × 24 × 60 × 60 = 3.1536 × 107
E7A.2(a) The de Broglie relation is [7A.11–230], λ = h/p = h/(mυ). Therefore,
h 6.6261 × 10−34 J s
υ= = = 7.27 × 106 m s−1
m e λ (9.1094 × 10−31 kg) × (100 × 10−12 m)
The kinetic energy acquired by an electron accelerated through a potential E is

eE: E k = 21 m e υ 2 = eE. Solving for the potential gives
m e υ 2 (9.1094 × 10−31 kg) × (7.27 × 106 m s−1 )2

E= = = 150 V
2e 2 × (1.6022 × 10−19 C)
E7A.3(a) The de Broglie relation is [7A.11–230] λ = h/p = h/(mυ).
h 6.6261 × 10−34 J s
υ= = = 2.4 × 10−2 m s−1
m e λ (9.1094 × 10−31 kg) × (3 × 10−2 m)
E7A.4(a) The de Broglie relation is [7A.11–230] λ = h/p = h/(mυ). Therefore
h α −1 h 137 × (6.6261 × 10−34 J s)

λ= = = = 332 pm
m e αc me c (9.1094 × 10−31 kg) × (2.9979 × 108 m s−1 )
E7A.5(a) The de Broglie wavelength is [7A.11–230], λ = h/p = h/(mυ)

(i)
6.6261 × 10−34 J s
λ= = 6.6 × 10−29 m
(1.0 × 10−3kg) × (1.0 × 10−2 m s−1 )
(ii)
6.6261 × 10−34 J s
λ= = 6.6 × 10−36 m
(1.0 × 10−3 kg) × (100 × 103 m s−1 )
(iii)
6.6261 × 10−34 J s
λ=
((4.00 × 10−3 kg mol−1 )/(6.0221 × 1023 mol−1 )) × (1.00 × 103 m s−1 )
= 99.8 pm
E7A.6(a) Wien’s law [7A.1–224], λ max T = 2.9 × 10−3 m K, is rearranged to give the
wavelength at which intensity is maximised
λ max = (2.9 × 10−3 m K)/T = (2.9 × 10−3 m K)/(298 K) = 9.7 × 10−6 m
E7A.7(a) Assuming that the object is a black body is equivalent to assuming that Wien’s
law [7A.1–224], λ max T = 2.9 × 10−3 m K, holds. Using ν̃ = λ−1 , Wien’s law is
expressed in terms of the wavenumber of maximum intensity (ν̃ max )
T/ν̃ max = 2.9 × 10−3 m K
This is rearranged to give the temperature
T = (2.9 × 10−3 m K) × ν̃ max

= (2.9 × 10−3 m K) × (2000 × 102 m−1 ) = 580 K
E7A.8(a) Molar heat capacities of monatomic non-metallic solids obey the Einstein re-
lation [7A.8a–227],
2
θ E 2 eθ E /2T
C V ,m (T) = 3R f E (T), f E (T) = ( ) ( θ /T )
T e E −1
where the solid is at temperature T and is characterized by an Einstein temper-
ature θ E . Thus, for a solid at 298 K with an Einstein temperature of 2000 K
2
2000 K 2 e(2000 K)/2(298 K)
f E (298 K) = ( ) ( (2000 K)/298 K ) = 5.49... × 10−2
298 K e −1
Hence, C V ,m (298 K) = (5.49 × 10−2 ) × 3R
E7A.9(a) The energy of the quantum is given by the Bohr frequency condition [7A.9–
227], ∆E = hν, and the frequency is ν = 1/T. The energy per mole is ∆E m =
N A ∆E.
(i) For T = 1.0 fs
∆E = (6.6261 × 10−34 J s)/(1.0 × 10−15 s) = 6.6 × 10−19 J

∆E m = (6.6... × 10−19 J) × (6.0221 × 1023 mol−1 ) = 4.0 × 102 kJ mol−1
(ii) For T = 10 fs
∆E = (6.6261 × 10−34 J s)/(10 × 10−15 s) = 6.6 × 10−20 J

∆E m = (6.6... × 10−20 J) × (6.0221 × 1023 mol−1 ) = 40 kJ mol−1
(iii) For T = 1.0 s
∆E = (6.6261 × 10−34 J s)/(1.0 s) = 6.6 × 10−34 J

∆E m = (6.6... × 10−34 J) × (6.0221 × 1023 mol−1 ) = 4.0 × 10−13 kJ mol−1
E7A.10(a) The energy of a photon with wavelength λ is given by
E = hν = hc/λ = (6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )/λ

= (1.9825 × 10−25 J)/(λ/m)
The energy per mole is then given by
E m = N A E = (0.11939 J mol−1 )/(λ/m)
Hence, the following table is drawn up
λ/nm E/zJ E m /kJ mol−1

(a) 600 330 199
(b) 550 360 217
(c) 400 496 298
E7A.11(a) When a photon is absorbed by a free hydrogen atom, the law of conservation of
energy requires that the kinetic energy acquired by the atom is E k , the energy
of the absorbed photon. Assuming relativistic corrections are negligible the
kinetic energy is E k = E photon = 12 m H υ 2 . The atom is accelerated to the speed
2E photon 1/2 2N A E photon 1/2

υ=( ) =( )
mH MH
1/2
2 × (6.0221 × 1023 mol−1 ) × E photon
=( )
(1.0079 × 10−3 kg mol−1 )
1/2
= (3.45... × 1013 m s−1 ) × (E photon /J)
The photon energies have been calculated in Exercise E7A.10(a), and thus the
following table is drawn up
λ/nm E/zJ υ/km s−1

(a) 600 330 19.9
(b) 550 360 20.8
(c) 400 496 24.4
E7A.12(a) The energy emitted from a lamp at (constant) power P in a time interval ∆t is
P∆t. The energy of a single photon of wavelength λ is E = hc/λ. Hence, the
total number of photons emitted in this time interval is the total energy emitted
divided by the energy per photon, N = P∆t/E photon = P∆tλ/hc. Thus, for a
time interval of 1 s and a wavelength of 550 nm
(i) P = 1 W
(1 W) × (1 s)(550 × 10−9 m)
N= = 2.77 × 1018
(6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )
(ii) P = 100 W
(100 W) × (1 s)(550 × 10−9 m)

N= = 2.77 × 1020
(6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )
E7A.13(a) As described in Section 7A.2 on page 228, photoejection can only occur if the
energy of the incident photon is greater than or equal to the work function of
the metal ϕ. If this condition is fulfilled, the energy of the emitted photon is
given by [7A.10–229], E k = hν − Φ = hc/λ − Φ. To convert the work function to
Joules, multiply through by the elementary charge, as described in Section 7A.2
on page 228,
Φ = (2.14 eV) × e = (2.14 eV) × (1.602 × 10−19 J eV−1 ) = 3.42... × 10−19 J

√
and since E k = 1/2m e υ 2 , υ = 2E k /m e
(i) For λ = 700 nm
hc (6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )

E photon = =
λ 700 × 10−9 m
−19
= 2.84... × 10 J
This is less than the threshold energy, hence no electron ejection occurs.
(ii) For λ = 300 nm
hc (6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )

E photon = =
λ 300 × 10−9 m
−19
= 6.62... × 10 J
This is greater than the the threshold frequency, and so photoejection can
occur. The kinetic energy of the electron is,
E k = hc/λ − Φ = 6.62... × 10−19 J − 3.42... × 10−19 J = 3.19 × 10−19 J
√
υ= 2E photon /m e
√
= 2 × (3.19 × 10−19 J)/(9.109 × 10−31 kg) = 837 km s−1
P7A.1 A cavity approximates an ideal black body, hence the Planck distribution [7A.6a–
225], applies
8πhc
ρ(λ, T) =
λ 5 (e hc/λk T − 1)
Because the wavelength range is small (5 nm), the energy density is approxi-
mated by
∆E(T) = ρ(λ, T)∆λ
Taking λ = 652.2 nm gives
hc (6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )

= = 2.20... × 104 K
λk (652.5 × 10−9 m) × (1.3806 × 10−23 J K−1 )
and
8πhc 8π × (6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )
= = 4.22... × 107 J m−4
λ5 (652.5 × 10−9 m)5
It follows that
1
∆E(T) = (4.22... × 107 J m−4 ) × × (5 × 10−9 m)
e(2.20 ...×104 K)/T −1
0.211... J m−3
=
e(2.20 ...×104 K)/T − 1
(a)
0.211... J m−3
∆E(298 K) = = 1.54 × 10−33 J m−3
e(2.20 ...×104 K)/(298 K) − 1
(b)
0.211... J m−3
∆E(3273 K) = = 2.51 × 10−4 J m−3
e(2.20 ...×104 K)/(3273 K) − 1
P7A.3 As λ increases, hc/λkT decreases, and at very long wavelengths hc/λkT ≪ 1.
Hence, the exponential can be expanded in a power series. Let x = hc/λkT,
then ex = 1 + x + 2!1 x 2 + 3!1 x 3 ..., and the Planck distribution becomes
8πhc
ρ=
λ 5 (1 + x + 1 2
2!
x
+ 1 3
3!
x ... − 1)
When x << 1 the second and higher order terms in x become negligibly small
compared to x. Consequently
8πhc 8πhc 1 8πkT

lim ρ = = 5 ( )=
λ→∞ λ5 x λ hc/λkT λ4
This is the Rayleigh–Jeans law
P7A.5 Each data point is used to find a value for λ max T, and then the mean of these is
used with λ max T = hc/(4.965k) to find a value of h, h = (4.965k/c) (λ max T)mean .
The following table is drawn up
θ/ ○ C 1000 1500 2000 2500 3000 3500

T/ K 1273 1773 2273 2773 3273 3773
λ max / nm 2181 1600 1240 1035 878 763
−3
λ max T/(10 m K) 2.776 2.837 2.819 2.870 2.874 2.879
The mean is 2.84... × 10−3 m K with a standard deviation of 4.01... × 10−5 m K.

Therefore
4.965 × (1.3806 × 10−23 J K−1 ) × (2.84... × 10−3 m K)

h= = 6.54 × 10−34 J s
2.9979 × 108 m s−1
P7A.7 The total energy density is given by [7A.7–226],

∞ ∞ 1 1
E(T) = ∫ ρ(λ, T) dλ = 8πhc ∫ hc/λk T −1
dλ
0 0 λ e
5
Let x = hc/λkT, or λ = hc/xkT. Then, dλ = −hc/x 2 kT dx

∞ 1 1 hc (kT)4 ∞ x3
E(T) = 8πhc ∫ dx = 8π ∫ dx
0 (hc/xkT)5 ex − 1 x 2 kT (hc)3 0 ex − 1
= 8π(kT)4 /(hc)3 × π 4 /15 = 8π 5 k 4 T 4 /15(hc)5
This is the Stefan–Boltzmannn law, that the total energy density is proportional
to T 4 , with a constant of proportionality of 8π 5 k 4 /15(hc)5 .
P7A.9 Assuming the Sun approximates an ideal black body, Wien’s law, λ max T = 2.9 ×
10−3 m K, applies. Hence, λ max = (2.9×10−3 m K)/(5800 K) = 500 nm which
corresponds to blue-green .
7B Wavefunctions
D7B.1 This is discussed in Section 7B.2(b) on page 235.
D7B.3 These terms are best illustrated by referring to a one-dimensional system. The
probability density P(x) is defined so that P(x) dx is the probability of finding
the system between x and x + dx. The value of the wavefunction itself, ψ(x),
is the probability amplitude. The probability density is given in terms of this
amplitude as P(x) = ψ ∗ (x)ψ(x).
E7B.1(a) (i) The function ψ = x 2 cannot be normalized as the area under ψ ∗ ψ = x 4
is infinite when the integral is evaluated over all space. However, if the
function were restricted to a finite region of space, it could be normalized.
(ii) The function ψ = 1/x cannot be normalized as it goes to ∞ as x goes to
0 so the integral of ψ ∗ ψ over all space is infinite. However, if the function
were restricted to a finite region of space which did not include x = 0 it
could be normalized.
(iii) The function ψ = exp(−x 2 ) can be normalized as ψ ∗ ψ = exp(−2x 2 )
goes to 0 at x = ±∞ so that the integral of ψ ∗ ψ over all space is finite.
A function is an acceptable wavefunction if it: (1) is not infinite over a finite

region; (2) is single-valued; (3) is continuous; (4) has a continuous first deriva-
tive. The functions ψ = x 2 and ψ = exp(−x 2 ) satisfy all these conditions. The
function ψ = 1/x satisfies these conditions provided that the wavefunction is
restricted to a region which excludes x = 0.
E7B.2(a) The normalized wavefunction is ψ(x) = (2/L)1/2 sin(2πx/L), and so the prob-
ability density is P(x) = ∣ψ(x)∣2 = (2/L) sin2 (2πx/L). This is maximized when
sin2 (2πx/L) = 1, and so when sin(2πx/L) = ±1. These values occur when
2πx/L = π/2, 3π/2 and hence x = L/4, 3L/4 .
Nodes occur when the wavefunction goes through zero: sin(2πx/L) = 0. The
sine function is zero when 2πx/L = π, hence x = L/2 . The wavefunction goes
to zero at x = 0 and x = L, but these do not count as nodes as the wavefunction
does not pass through zero.
E7B.3(a) The task is to find N such that ψ = N sin(2πx/L) satisfies the normalization
condition [7B.4c–234], ∫ ψ ∗ ψ dτ = 1. In this case the integration is over x and
the range is 0 to L; the function is real, so ψ = ψ ∗ .
=sin 4π=0
L ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
N2 ∫ sin2 (2πx/L)dx = N 2 [L/2 − (L/8π) sin(4πL/L)] = N 2 (L/2)
0
The integral is of the form of Integral T.2 with a = L and k = 2π/L. For
the wavefunction to be normalized, the integral must be 1 and therefore the
normalizing factor is N = (2/L)1/2 .
E7B.4(a) The task is to find N such that ψ = N exp(−ax 2 ) satisfies the normalization
condition [7B.4c–234], ∫ ψ ∗ ψ dτ = 1. In this case the integration is over x
and the range is −∞ to ∞; the function is real, so ψ = ψ ∗ and the integral is
∞
therefore N 2 ∫−∞ exp(−2ax 2 ) dx. The integrand is even, meaning that it has
the same value at +x and −x, so the integral between −∞ and +∞ is simply
twice that between 0 and +∞. The integral is then of the form of Integral G.1
with k = 2a
∞
2N 2 ∫ exp(−2ax 2 )dx = 2N 2 [ 21 (π/2a)1/2 ]
0
Setting this equal to 1 gives N = (2a/π)1/4 .
E7B.5(a) (i) The function ψ = exp(−ax 2 ) can be normalized as ψ ∗ ψ = exp(−2ax 2 )

goes to 0 at x = ±∞ so that the integral over all space of ψ ∗ ψ is finite.
(ii) The function ψ = exp(−ax) cannot be normalized as it goes to ∞ as x
goes to −∞, therefore the integral over all space of ψ ∗ ψ is infinite. If the
wavefunction were limited to a finite region of space it could, however, be
normalized.
A function is an acceptable wavefunction if it: (1) is not infinite over a finite re-
gion; (2) is single-valued; (3) is continuous; (4) has a continuous first derivative.
Both functions satisfy all of these conditions and so are acceptable wavefunc-
tions.
E7B.6(a) The probability of finding an electron in an infinitesimal region dx around x

is P(x)dx = ψ ∗ (x)ψ(x)dx, provided that ψ(x) is a normalized wavefunction.
The wavefunction is real so that ψ ∗ = ψ, hence the probability is given by
2
P(x)dx = [(2/L)1/2 sin(2πx/L)] dx = (2/L) sin2 (2πx/L)dx
Hence, at x = L/2, P(L/2)dx = (2/L) sin(π)dx = 0 .

E7B.7(a) The normalized wavefunction is ψ(x) = (2/L)1/2 sin(2πx/L). The probability
of finding the electron between x = L/4 and x = L/2 is, using Integral T.2
L/2
L/2
∫ (2/L) sin2 (2πx/L) dx = (2/L) [x/2 − (L/8π) sin(4πx/L)∣L/4 ]
L/4
= (2/L)([L/4 − (L/8π) sin(2π)] − [L/8 − (L/8π) sin(π)])

= (2/L)[L/4 − L/8] = 1/4
E7B.8(a) In two dimensions ∣ψ(x, y)∣2 dxdy is the probability of finding the particle in
the infinitesimal area dxdy at the position (x, y). The probability is a dimen-
sionless quantity, and the area dxdy has units of (length)2 . Hence, ∣ψ(x, y)∣2
must have units of (length)−2 , implying that ψ has units of length−1 .
P7B.1 (a) The task is to find N such that ψ = Neiϕ satisfies the normalization condi-
tion [7B.4c–234], ∫ ψ ∗ ψ dτ = 1. In this case the integration is over ϕ and
the range is 0 to 2π. The function is complex so ψ ∗ = N exp(−iϕ); note
that the normalization factor N is always real. The integrand is therefore
ψ ∗ ψ = e−iϕ eiϕ = e−iϕ+iϕ = e0 = 1
The integral to evaluate is therefore
2π
2π
N2 ∫ 1 dϕ = N 2 ϕ∣0 = 2πN 2
0
Setting this equal to 1 gives N = (2π)−1/2 .

(b) For the function eim l ϕ exactly the same argument applies and so the nor-
malization factor is the same.
P7B.3 (a) The task is to find N such that ψ = N sin(πx/L x ) sin(πy/L y ) satisfies the
normalization condition [7B.4c–234], ∫ ψ ∗ ψ dτ = 1. The integration is
over x from 0 to L x , and y from 0 to L y . The wavefunction is real, and so
the integral to evaluate is
Lx Ly
N2 ∫ ∫ sin2 (πx/L x ) sin2 (πx/L y )dxdy
0 0
This separates into the product of two integrals

Lx Ly
N 2 (∫ sin2 (πx/L x )dx) (∫ sin2 (πy/L y )dy)
0 0
Each of these is evaluated using Integral T.2

Lx
∫ sin2 (πx/L x )dx = L x /2 − (L x /4π) sin(2πL/L) = L x /2
0
and similarly the integral over y is equal to L y /2. Hence the integral is
√
evaluated as N 2 (L x /2)(L y /2). Setting this equal to 1 gives N = 2/ L x L y .
(b) In the case when L x = L y = L this reduces to N = 2/L .
P7B.5 The task is to find N such that ψ(x) = Ne−ax , satisfies the normalization
condition [7B.4c–234], ∫ ψ ∗ ψ dτ = 1. The integration is over x ranging from 0
to ∞.
∞ ∞
N2 ∫ e−2ax dx = −(N 2 /2a) e−2ax ∣0 = −(N 2 /2a)(e−∞ − e0 ) = N 2 /2a
0
where e−∞ = 0 and e0 = 1 are used. Setting this equal to 1 gives N = (2a)1/2 .
Hence, the total probability of finding the particle at a distance x ≥ x 0 is
∞ ∞ ∞
∫ ∣ψ(x)∣2 dx = 2a ∫ e−2ax dx = −(2a/2a) e−2ax ∣x = e−2ax 0
x0 x0 0
With a = 2 m−1 and x 0 = 1 m, the probability is exp(−2 × (2 m−1 ) × (1 m))

= exp(−4) = 0.0183
P7B.7 The probability of finding the particle in the range x = a to x = b is

b b
P(a → b) = ∫ ∣ψ(x)∣2 dx = ∫ (2/L) sin2 (πx/L) dx
a a
This is evaluated using Integral T.2 to to give

b
P(a → b) = (2/L) [ x/2 − (L/4π) sin(2πx/L)∣ a ]
b−a 1 2πb 2πa
= − [sin ( ) − sin ( )]
L 2π L L
Hence,
(a) L = 10.00 nm, a = 4.95 nm, b = 5.05 nm, P(a → b) = 2.00 × 10−2
(b) L = 10.00 nm, a = 1.95 nm, b = 2.05 nm, P(a → b) = 6.91 × 10−3
(c) L = 10.00 nm, a = 9.90 nm, b = 10.00 nm, P(a → b) = 6.58 × 10−6
(d) L = 10.00 nm, a = 5.00 nm, b = 10.00 nm, P(a → b) = 0.5 . This is
expected as for this wavefunction the probability density is symmetric
around x = 5.00 nm.
P7B.9 The probability of being in a small volume δV at distance r from the nucleus
is ψ(r)2 δV . The probability of the electron being in a small sphere of radius r ′
centred at r from the nucleus is
2
P(r) = [(πa 03 )−1/2 e−r/a 0 ] × [(4/3)π(r ′ )3 ] = [4(r ′ )3 /3a 03 ] e−2r/a 0
where a 0 = 53 pm is the Bohr radius. If the sphere has radius 1.0 pm then the
expression evaluates to
4 × (1.0 pm)3
× e−2r/a 0 = 8.95... × 10−6 × e−2r/a 0
3 × (53 pm)3
(a) If the sphere is centered at the nucleus r = 0, P(0) = 8.95 × 10−6

(b) If the sphere is centered at r = a 0 , P(a 0 ) = 8.95...×10−6 ×e−2 = 1.21 × 10−6
P7B.11 The probability finding the atom between x and x +dx, where dx is an infinites-
imal displacement, is given by
P(x)dx = ψ(x)2 dx = N 2 x 2 e−x /a 2

2
dx
The task is to find the position of the maximum in P(x), which is done by
setting dP(x)/dx = 0. To compute the derivative it is necessary to use the
product rule.
d e−x /a ⎞
2 2
dP(x) d ⎛ d x2
(N 2 x 2 e−x /a ) = N 2 × e−x /a + x 2 ×
2 2 2 2
=
dx dx ⎝ dx dx ⎠
2x −x 2 /a 2 x 2
= N 2 (2xe−x /a 2 2 −x 2 /a 2
2
− x2 × e ) = 2xN e [1 − ( ) ]
a2 a
The derivative goes to zero at x = 0, ±a, ±∞. Inspection of the form of P(x)
shows that x = 0 is a minimum (P(x) = 0), and at x = ±∞, P(x) also goes
asymptotically to zero. The maxima are therefore at x = ±a .
7C Operators and observables

D7C.1 In quantum mechanics an observable quantity (such as energy, position or
momentum) is represented by a particular operator Ω̂. If the wavefunction
is ψ the average value of the quantity represented by the operator Ω̂ is given by
⟨Ω⟩ = ∫ ψ ∗ Ω̂ψ dτ, called the expectation value. For the special case that ψ is
an eigenfunction of Ω̂, the expectation value is the eigenvalue corresponding
to this eigenfunction.
D7C.3 The operator which represents kinetic energy is T̂ = (−ħ 2 /2m)d2 /dx 2 . The
expectation value of the kinetic energy is therefore related to the average value
of d2 ψ/dx 2 , that is the average of the second derivative or curvature of the
wavefunction. Sharply curved regions of the wavefunction will make a larger
contribution to the kinetic energy than less sharply curved regions. The expec-
tation value of the kinetic energy will have contributions from all parts of the
wavefunction.
E7C.1(a) Two wavefunctions ψ i and ψ j are orthogonal if ∫ ψ ∗i ψ j dτ = 0, [7C.8–240].
In this case the integration is from ϕ = 0 to ϕ = 2π. Let ψ i = exp(iϕ), the
wavefunction with m l = +1, and ψ j = exp(2iϕ), the wavefunction with m l =
+2. Note that the functions are complex, so ψ ∗i = exp(−iϕ). The integrand is
therefore
ψ ∗i ψ j = exp(−iϕ) exp(2iϕ) = exp(iϕ)
and the integral evaluates as
=1 =1
2π
2π
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
∫ exp(iϕ) dϕ = (1/i) exp(iϕ)∣0 = (1/i)[exp(i × 2π) − exp(i × 0)] = 0
0
The identity exp(ix) = cos x + i sin x (The chemist’s toolkit 16 in Topic 7C on

page 242) is used to evaluate exp(i2π) = cos(2π) + i sin(2π) = 1 + 0 = 1. The
integral is zero, so the functions are indeed orthogonal.
E7C.2(a) The normalized wavefunction is ψ(x) = (2/L)1/2 sin(2πx/L). The operator for
position is x̂ = x, therefore the expectation value of the position of the electron
is [7C.11–242]
L
⟨x⟩ = ∫ ψ ∗ x̂ψ dτ = (2/L) ∫ x sin2 (2πx/L) dx
0
This integral is of the form of Integral T.11 with k = 2π/L and a = L
=sin 4π=0 =cos 4π=1

⎡ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ⎧ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ⎫⎤
2 ⎢⎢ L 2 L 4πL 1 ⎪
⎪ 4πL ⎪
⎪⎥
= ⎢ − sin ( )− ⎨ cos ( ) −1⎬⎥⎥
L ⎢ 4 4 × 2π/L L 2
8 × (2π/L) ⎪⎪ L ⎪
⎣ ⎩ ⎭⎥⎦
⎪
2 L 2
= × = L/2
L 4
Because the probability density ∣ψ(x)∣2 is symmetric about x = L/2, the ex-
pected result is ⟨x⟩ = L/2.
E7C.3(a) The normalized wavefunction is ψ(x) = (2/L)1/2 sin(2πx/L). The expectation

value of the momentum is ∫ ψ ∗ p̂ x ψ dx, and the momentum operator is p̂ x =
(ħ/i)d/dx, therefore
L
⟨p x ⟩ = (2/L) ∫ sin(2πx/L) p̂ x sin(2πx/L) dx
0
L
= (2ħ/iL) ∫ sin(2πx/L)(d/dx) sin(2πx/L) dx
0
Using (d/dx) sin(2πx/L) = (2π/L) cos(2πx/L) gives

L
⟨p x ⟩ = (4πħ/iL 2 ) ∫ sin(2πx/L) cos(2πx/L) dx
0
The integral is of the form of Integral T.7 with a = L, k = 2π/L

4πħ 1 2πL
⟨p x ⟩ = × sin2 ( )= 0
iL 2 2 × 2π/L L
This result is interpreted as meaning that there are equal probabilities of having
momentum in the positive and negative x directions.
E7C.4(a) For the case when m l = +1 the normalized wavefunction is ψ+1 (ϕ) = (2π)−1/2 eiϕ .
∗
This is complex, and so ψ+1 = (2π)−1/2 e−iϕ . The expectation value of the
position, specified by the operator ϕ, is
2π 2π
∗
⟨ϕ⟩ = ∫ ψ+1 ϕψ+1 dϕ = (1/2π) ∫ e−iϕ ϕ eiϕ dϕ
0 0
2π 2π
= (1/2π) ∫ ϕ dϕ = (1/2π) × (ϕ 2 /2)∣0 = π
0
where e−iϕ eiϕ = 1 is used.

For the general case the integral is
2π
(1/2π) ∫ e−im l ϕ ϕeim l ϕ dϕ
0
which evaluates to give the same result, ⟨ϕ⟩m l = π. This result is interpreted
by noting that the probability density around the ring is (1/2π)e−im l ϕ eim l ϕ =
1/2π, which is constant. Therefore the average position is half way round the
ring at ϕ = π.
E7C.5(a) The uncertainty in the momentum is given by ∆p = m∆υ, where m is the mass
and ∆υ is the uncertainty in the velocity. The uncertainties in position (∆q)
and momentum (∆p) must obey the Heisenberg uncertainty principle [7C.13a–
244], ∆p q ∆q ≥ (ħ/2), which in this case is expressed as m∆υ∆q ≥ (ħ/2). This
is rearranged to give the uncertainty in the velocity, ∆υ ≥ ħ/(2m∆q). The
minimum uncertainty in the speed is therefore ∆υ min = ħ/(2m∆q), which is
evaluated as
1.0546 × 10−34 J s
= 1.05 × 10−28 m s−1
2 × (500 × 10−3 kg) × (1.0 × 10−6 m)
The uncertainty principle can be rearranged for the uncertainty in the position
∆q ≥ ħ/2m∆υ, and so the minimum uncertainty is ħ/(2m∆υ). The uncer-
tainty in the position of the bullet is 1 × 10−5 m s−1 , and hence
1.0546 × 10−34 J s
∆q min = = 1.05 × 10−27 m
2 × (5.0 × 10−3 kg) × (1 × 10−5 m s−1 )
E7C.6(a) The desired uncertainty in the momentum is
∆p = 0.0100 × 10−2 p = 1.00 × 10−4 m p υ

= 1.00 × 10−4 × (1.6726 × 10−27 kg) × (0.45 × 106 m s−1 )
= 7.52... × 10−26 kg m s−1
The Heisenberg uncertainty principle, [7C.13a–244] is rearranged to give the

uncertainty in the position as ∆q ≥ ħ/(2∆p), which gives a minimum uncer-
tainty of ∆q min = ħ/(2∆p). This is evaluated as
1.0546 × 10−34 J s
= 7.01 × 10−10 m
2 × (7.52... × 10−26 kg m s−1 )
E7C.7(a) To construct the potential energy operator, replace the position x in the clas-
sical expression by the operator for position x̂ to give V̂ = 12 k f x̂ 2 . However,
because x̂ = x× the potential energy operator is V̂ = 21 k f x 2 .
E7C.8(a) A function ψ is an eigenfunction of an operator Ω̂ if Ω̂ψ = ωψ where ω is a

constant called the eigenvalue.
(i) (d/dx) cos(kx) = −k sin(kx). Hence cos kx is not an eigenfunction of

the operator d/dx.
(ii) (d/dx)eik x = ikeik x . Hence eik x is an eigenfunction of the operator
d/dx, with eigenvalue ik .
(iii) (d/dx)kx = k. Hence kx is not an eigenfunction of the operator d/dx.
(iv) (d/dx)e−ax = −2axe−ax . Hence e−ax is not an eigenfunction of the
2 2 2
operator d/dx.
E7C.9(a) Wavefunctions ψ 1 and ψ 2 are orthogonal if ∫ ψ ∗1 ψ 2 dτ = 0, [7C.8–240]. Here

ψ 1 (x) = sin(πx/L), ψ 2 (x) = sin(2πx/L), and the region is 0 ≤ x ≤ L. The
integral is evaluated using Integral T.5
L
∗
∫ ψ 1 ψ 2 dτ = ∫ sin(πx/L) sin(2πx/L) dx
0
= [(−L/2π) sin(−π) − (L/6π) sin(3π)] = 0
where sin(nπ) = 0 for integer n is used. Thus, the two wavefunctions are
orthogonal.
E7C.10(a) Wavefunctions ψ 1 and ψ 2 are orthogonal if ∫ ψ ∗1 ψ 2 dτ = 0, [7C.8–240]. Here

ψ 1 (x) = cos(πx/L), ψ 2 (x) = cos(3πx/L), and the region is −L/2 ≤ x ≤ L/2.
The integral is evaluated using Integral T.6
L/2
∗
∫ ψ 1 ψ 2 dτ = ∫ cos(πx/L) cos(3πx/L) dx
−L/2
L/2
= (−L/4π) sin(−2πx/L) + (L/8π) sin(4πx/L)∣−L/2
= [(−L/4π) sin(−π) + (L/8π) sin(2π)]
− [(−L/4π) sin(π) + (L/8π) sin(−2π)] = 0
where sin(nπ) = 0 for integer n is used. Thus, the two wavefunctions are
orthogonal.
P7C.1 Operate on each function f with i,ˆ the inversion operator, which has the effect
of making the replacement x → −x. If the result of the operation is f multiplied
by a constant, f is an eigenfunction of iˆ and the constant is the eigenvalue.
(a) i(x
ˆ 3 − kx) = (−x)3 − k(−x) = −(x 3 − kx). Yes , f is an eigenfunction,
eigenvalue −1 .
(b) i(cos(kx))
ˆ = cos(k(−x)) = cos(kx). Yes , f is an eigenfunction, eigen-
value +1 .
(c) i(x
ˆ 2 + 3x − 1) = (−x)2 + 3(−x) − 1 = x 2 − 3x − 1. No , f is not an
eigenfunction.
∗
P7C.3 An operator Ω̂ is hermitian if ∫ ψ ∗i Ω̂ψ j dτ = [∫ ψ ∗j Ω̂ψ i dτ] , [7C.7–239]. For
the kinetic energy operator, it is necessary to integrate by parts (The chemist’s
toolkit 15 in Topic 7C on page 240) twice
∗ d ψj
ħ 2 d2 ħ2 ∞ 2
∗
∫ ψ i (− ) ψ j dx = − ∫ ψ i dx
2m dx 2 2m −∞ dx 2
⎛ ⎞
∞
ħ2 ⎜ dψ ∞ dψ ∗ dψ ⎟ ħ2 ∞ dψ ∗ dψ
⎜ ∗ j i j ⎟ i j
=− ⎜ ψi ∣ −∫ dx ⎟ = dx
2m ⎜ dx −∞ dx dx ⎟ 2m ∫−∞ dx dx
⎜ −∞ ⎟
⎝´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸
A
¹¹¹¹¹¹¹¹¹¹¹¶ ⎠
Term A evaluates to zero because, as stated in the problem, the wavefunction

and its derivative can be assumed to go to zero at x = ±∞. Then integrate the
remaining term by parts once again,
⎛ ⎞
∞ dψ ∗ dψ ∗ ∞ ∞ d2 ψ ∗
ħ2 i j ħ 2 ⎜
dψ
⎜ i ψ j∣ − ∫ i
⎟
∫ dx = ψ j dx ⎟
2m −∞ dx dx 2m ⎜⎜ dx −∞ −∞ dx 2 ⎟
⎟
⎝´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶
B
⎠
As before, term B evaluates to zero, so the final result is
ħ 2 d2 ∞ ħ 2 d2
∗
∫ ψ i (− ) ψ j dx = ∫ ψ j (− ) ψ ∗i dx
2m dx 2 −∞ 2m dx 2
The term on the right can be rewritten as a complex conjugate to give

∗
∗ ħ 2 d2 ∗ ħ 2 d2
∫ ψ i (− ) ψ j dx = [∫ ψ j (− ) ψ i dx]
2m dx 2 2m dx 2
Note that because the complex conjugate of the whole term is taken, to com-
pensate for this the complex conjugate of the terms inside the bracket need to
be taken too ψ = [ψ ∗ ]∗ . This final equation is consistent with [7C.7–239] and
so demonstrates that the kinetic energy operator is hermitian.
P7C.5 (a) Consider the sum of two arbitrary hermitian operators Â and B̂,
∗ ∗ ∗
∫ ψ i (Â + B̂)ψ j dτ = ∫ ψ i Âψ j dτ + ∫ ψ i B̂ψ j dτ
∗
As Â is hermitian, ∫ ψ ∗i Âψ j dτ = [∫ ψ ∗j Âψ i dτ] and similarly for B̂.
∗ ∗
∗ ∗ ∗
∫ ψ i (Â + B̂)ψ j dτ = [∫ ψ j Âψ i dτ] + [∫ ψ j B̂ψ i dτ]
∗
= [∫ ψ ∗j (Â + B̂)ψ i dτ]
which demonstrates that the sum of two hermitian operators is also her-
mitian.
(b) As Ω̂ψ j = ψ k is also a function, the integral can be rewritten
∗ ∗
∫ ψ i Ω̂ Ω̂ψ j dτ = ∫ ψ i Ω̂ψ k dτ
As Ω̂ is a hermitian operator it follows that

∗
∗ ∗
∫ ψ i Ω̂ψ k dτ = [∫ ψ k Ω̂ψ i dτ]
Next realise that Ω̂ψ i is a function which will be called ψ l : Ω̂ψ i = ψ l .

With this substitution
∗
∗ ∗ ∗
∫ ψ i Ω̂ψ k dτ = [∫ ψ k ψ l dτ] = ∫ ψ k ψ l dτ
= ∫ ψ ∗l ψ k dτ
= ∫ ψ ∗l Ω̂ψ j dτ
To go to the second line the two functions have been re-ordered, which
is always permitted, and to go to the third line the definition Ω̂ψ j = ψ k
has been used. Because Ω̂ is hermitian the final term can be rewritten as
[∫ ψ ∗j Ω̂ψ l dτ]∗ and so
∗
∗ ∗
∫ ψ i Ω̂ψ k dτ = [∫ ψ j Ω̂ψ l dτ]
In the term on the right the definition Ω̂ψ i = ψ l is used, and in the term
on the left Ω̂ψ j = ψ k is used to give
∗
∗ ∗
∫ ψ i Ω̂ Ω̂ψ j dτ = [∫ ψ j Ω̂ Ω̂ψ i dτ]
This is exactly the property which proves that Ω̂ Ω̂ is hermetian.

P7C.7 The operator for position is multiplication by x, so it follows that the expecta-
tion value of the position operator is
∞ 1!
⟨x⟩ = ∫ ψ ∗ x̂ψdτ = a ∫ xe−ax dx = a × = 1/a
0 a2
The integral is of the form of Integral E.3 with n = 1 and k = a.
P7C.9 (a) The expectation value is given by [7C.11–242], ⟨Ω⟩ = ∫ ψ ∗ Ω̂ψ dτ, where
ψ is normalized. A hermitian operator has the property ∫ ψ ∗i Ω̂ψ j dτ =
[∫ ψ ∗j Ω̂ψ i dτ]∗ , which for the case ψ i = ψ j = ψ becomes
A B
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ⎡ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ⎤∗
∗
⎢ ∗
⎥
∫ ψ Ω̂ψ dτ = ⎢⎢ ∫ ψ Ω̂ψ dτ ⎥⎥
⎢ ⎥
⎣ ⎦
The quantities A and B and the same, so it follows that A = B∗ = A∗ ,
which is only true if A is real. The quantity A is the expectation value, so
it follows that the expectation value of a hermitian operator is real.
(b) The expectation value of the square of the hermitian operator Ω̂ Ω̂ is
⟨ΩΩ⟩ = ∫ ψ ∗ Ω̂ Ω̂ψ dτ
Recognise that Ω̂ψ is a function, which will be written ϕ: Ω̂ψ = ϕ; with

this, and using the fact that Ω̂ is hermitian it follows that
∗
∗ ∗ ∗
∫ ψ Ω̂ Ω̂ψ dτ = ∫ ψ Ω̂ϕ dτ = [∫ ϕ Ω̂ψ dτ]
Once again use Ω̂ψ = ϕ in the term on the right to give

∗
∗ ∗
∫ ψ Ω̂ Ω̂ψ dτ = [∫ ϕ ϕ dτ]
The integral ∫ ϕ∗ ϕ dτ is the sum over all space of the square magnitude
of a wavefunction. Using the Born interpretation, this is proportional to
the total probability and therefore must be a real positive number. It is
therefore shown that the expectation value of the square of a hermitian
operator is real and positive.
(a) The normalized wavefunction is ψ(x) = (2a/π)1/4 e−ax , leading to a

2
P7C.11
probability density of ∣ψ(x)∣2 = (2a/π)1/2 e−2ax . The expectation value
2
of the position is therefore

∞
xe−2ax dx = 0
2
⟨x⟩ = (2a/π)1/2 ∫
−∞
as the integrand is odd and the integration is over a symmetric range. For
⟨x 2 ⟩ the required integral is
∞ ∞
x 2 e−2ax dx = 2(2a/π)1/2 ∫ x 2 e−2ax dx
2 2
⟨x 2 ⟩ = (2a/π)1/2 ∫
−∞ 0
The integrand is an even function, and so its integral between −∞ and ∞

is twice that from 0 to ∞. This integral is of the form of Integral G.3 with
k = 2a
⟨x 2 ⟩ = 2(2a/π)1/2 × 14 (π/(2a)3 )1/2 = 1/4a
The effect of the momentum operator, (ħ/i)(d/dx), on the wavefunction

is (ħ/i)(d/dx)(2a/π)1/4 e−ax = (ħ/i)(2a/π)1/4 × (−2axe−ax ). Hence
2 2
the expectation value of the momentum is
ħ 2a 1/2 ∞
⟨p x ⟩ = ∫ ψ ∗ p̂ x ψ dτ = −2axe−2ax dx = 0
2
( ) ∫
i π −∞
as the integrand is odd.

To evaluate the effect of the operator p̂2x = −ħ 2 (d2 /dx 2 ) on the wavefunc-
tion the product rule is used
ħ d
((ħ/i)(2a/π)1/4 × [−2axe−ax ])
2
p̂2x ψ = p̂ x ( p̂ x )ψ =
i dx
This is equal to
(4a 2 x 2 e−ax − 2ae−ax )

1/4 2 2
−ħ 2 (2a/π)
which gives
⟨p2x ⟩ = ∫ ψ ∗ p̂2x ψ dτ
+∞
(4a 2 x 2 e−2ax − 2ae−2ax )
1/2 2 2
= −ħ 2 (2a/π) ∫
−∞
⎡ 1/2 ⎤
2a 1/2 ⎢ π π 1/2 ⎥⎥
= −ħ 2 ( ) × ⎢⎢2a 2 ( ) − 2a ( ) ⎥
π ⎢ (2a)3 2a ⎥
⎣ ⎦
= ħ a
2
where Integrals G.3 and G.1 are used.

(b) The uncertainty in the position is given by
1/2
∆x = [⟨x 2 ⟩ − ⟨x⟩2 ] = (1/4a − 02 )1/2 = (4a)−1/2
and the uncertainty in the momentum is given by

1/2 √
∆p x = [⟨p2x ⟩ − ⟨p x ⟩2 ] = (ħ 2 a − 02 )1/2 = ħ a
The product of the uncertainties is

√
∆x∆p x = (4a)−1/2 × ħ a = ħ/2
This uncertainty is ≥ ħ/2 and so is consistent with the uncertainty prin-

ciple; in fact it is the smallest possible uncertainty.
P7C.13 To evaluate the commutator of the operators consider the effect of the commu-
tator on an arbitrary function ψ; use the product rule to evaluate the derivatives
of products as necessary.
(a)
d 1 d ψ 1 dψ 1 dψ dx −1 1 dψ
[ , ]ψ = ( )− = +ψ −
dx x dx x x dx x dx dx x dx
1
=− 2 ψ
x
The commutator is therefore identified as the term multiplying ψ, −1/x 2 .
(b)
d 2 d(x 2 ψ) dψ dψ dx 2 dψ
[ , x ]ψ = − x2 = x2 +ψ − x2
dx dx dx dx dx dx
= 2x ψ
The commutator is therefore identified as the term multiplying ψ, 2x .
P7C.15 To complete this problem it is useful to establish some general properties of
commutators. First, an operator Â commutes with powers of itself: for example,
[Â, Â2 ] = 0. This is proved by simply multiplying out the commutator:
[Â, Â2 ] = Â(ÂÂ) − (ÂÂ)Â = 0
Note that multiplying either of the terms by a constant does not alter this prop-
erty. Second, any operator commutes with a constant: [Â, c] = 0, which follows
trivially on multiplying out the commutator. Finally, the useful property that
[Â + B̂, Ĉ] = [Â, Ĉ] + [B̂, Ĉ]. Again, this follows by multiplying out the com-
mutator
[Â + B̂, Ĉ] = (Â + B̂)Ĉ − Ĉ(Â + B̂) = ÂĈ − Ĉ Â + B̂Ĉ − Ĉ B̂ = [Â, Ĉ] + [B̂, Ĉ]
(a) For the case V̂ = V0 :
A B
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ µ
[Ĥ, p̂ x ] = [ p̂2x /2m + V0 , p̂ x ] = [ p̂2x /2m, p̂ x ] + [V0 , p̂ x ] = 0
Term A is zero because it is a commutator of an operator with a power of
itself, and term B is zero because it is a commutator with a constant.
For the case V̂ = 21 k f x 2 :
[Ĥ, p̂ x ] = [ p̂2x /2m + 12 k f x 2 , p̂ x ] = [ p̂2x /2m, p̂ x ] + [ 12 k f x 2 , p̂ x ]
The first commutator is zero. To evaluate the second commutator con-
sider its effect on an arbitrary wavefunction ψ; to simplify the calculation
set aside the constant multiplying term 12 k f and re-introduce it at the end
[x 2 , p̂ x ]ψ = (x 2 p̂ x − p̂ x x 2 )ψ
ħ 2 dψ d(x 2 ψ) ħ dψ dψ
= (x − ) = (x 2 − x2 − ψ × 2x)
i dx dx i dx dx
= −2x(ħ/i)ψ
Reintroducing the constants gives the result [Ĥ, p̂ x ] = −(k f ħ/i)x .

(b) For the case V̂ = V0 :
[Ĥ, x̂] = [ p̂2x /2m + V0 , x̂] = [ p̂2x /2m, x̂] + [V0 , x̂]
The second term is zero because it is a commutator with a constant. The

first term is evaluated in the usual way, first setting aside the constants
[ p̂2x , x̂]ψ = ( p̂2x x − x p̂2x )ψ

d2 (xψ) d2 ψ d dψ d2 ψ
= −ħ 2 [ − x ] = −ħ 2
[ (ψ + x ) − x ]
dx 2 dx 2 dx dx dx 2
dψ d2 ψ dψ d2 ψ dψ
= −ħ 2 [ +x 2 + − x 2 ] = −2ħ 2
dx dx dx dx dx
Reintroducing the constants gives the result [Ĥ, x̂] = −(ħ 2 /m)(d/dx)
For the case V̂ = 12 k f x 2 :
[Ĥ, x̂] = [ p̂2x /2m + 12 k f x 2 , x] = [ p̂2x /2m, x] + [ 12 k f x 2 , x]
The second commutator is zero. The first has already been evaluated in
the preceding calculation hence [Ĥ, x̂] = −(ħ 2 /m)(d/dx)
7D Translational motion
D7D.1 The physical origin of tunnelling is related to the probability density of the
particle, which according to the Born interpretation is the square of the wave-
function that represents the particle. This interpretation requires that the wave-
function of the system be everywhere continuous, even at barriers. Therefore,
if the wavefunction is non zero on one side of a barrier it must be non zero on
the other side of the barrier and this implies that the particle has tunnelled into
the barrier. The transmission probability depends upon the mass of the par-
ticle: the greater the mass the smaller the probability of tunnelling. Electrons
and protons have small masses, molecular groups large masses; therefore, tun-
nelling effects are more observable in process involving electrons and protons.
D7D.3 The hamiltonian for a particle in a two- or three-dimensional box is separa-

ble into a sum of terms each of which depends on just one of the variables
x, y or z. Therefore the solutions to the Schrödinger equation are products
of separate functions of each of these variables, Ψ(x, y, z) = ψ(x)ψ(y)ψ(z).
In the hamiltonian the terms in the three variables are all of the same form,
so the wavefunctions are likewise of the same form. Similarly, the boundary
conditions along each direction are the same, so the quantization which this
imposes is the same. As a result the wavefunctions are simply a product of the
wavefunctions that would be found by solving the Schrödinger equation along
each direction separately.
The energy levels of two- and three-dimensional boxes can be degenerate, that
is there are different wavefunctions which have the same energy. This is usually
a consequence of the box having a certain kind of symmetry, for example in a
two-dimensional box in which the two sides are the same length the levels with
quantum numbers n 1 = l, n 2 = m and n 1 = m, n 2 = l are degenerate.
E7D.1(a) The energy levels of a particle in a box with length L are given by [7D.6–249],
E n = h 2 n 2 /8mL 2 and when the length is increased to 1.1 × L the energies
become E n′ = h 2 n 2 /8m(1.1 × L)2 = h 2 n 2 /8m(1.21 × L 2 ) giving a fractional
change of
E n′ − E n [h 2 n 2 /8m(1.21 × L 2 )] − [h 2 n 2 /8mL 2 ] 1
= = − 1 = −0.174
En h n /8mL
2 2 2 1.21
E7D.2(a) The energy levels of a particle in a box of length L are given by [7D.6–249],
E n = h 2 n 2 /8mL 2 . Hence the spacing between adjacent levels (n and n + 1) is
given by
h 2 (n + 1)2 h 2 n 2 h2
∆E(n) = E n+1 − E n = − = ((n + 1)2 − n 2 )
8mL 2 8mL 2 8mL 2
h2 h2
= 2
(n 2 + 2n + 1 − n 2 ) = (2n + 1)
8mL 8mL 2
Setting ∆E(n) to the average thermal energy gives h 2 (2n + 1)/8mL 2 = kT/2,
and so (2n + 1) = 4mkTL 2 /h 2 , leading to
2mkTL 2
n= − 1
h2 2
For a helium atom, mass 4.00 m u in a box for length 1 cm,
n = [2(4.00 × 1.6605 × 10−27 kg) × (1.3806 × 10−23 J K−1 ) × (298 K)

(1 × 10−2 m)2 1
× ]− = 1.24 × 1016
(6.6261 × 10−34 J s)2 2
E7D.3(a) The wavefunction of a particle in a square box of side length L with quan-
tum numbers n 1 = 2, n 2 = 2 is ψ 2,2 (x, y) = (2/L) sin (2πx/L) sin (2πy/L).
2
The probability density is P2,2 (x, y) = (2/L) sin2 (2πx/L) sin2 (2πy/L). The
probability density is maximized when sin (2πx/L) × sin2 (2πy/L) = 1 which
2
occurs only when each sin term is equal to ±1,
sin (2πx/L) = ±1
2πx/L = π/2, 3π/2 hence x = L/4, 3L/4
and similarly for y. Hence the maxima in probability density occurs at (x, y) =
(L/4, L/4), (L/4, 3L/4), (3L/4, L/4), (3L/4, 3L/4)
Nodes occur when the wavefunction passes through zero, which is when either
of the sin terms are zero, excluding the boundaries at x = 0, L y = 0, L because
at these points the wavefunction does not pass through zero. There is thus a
node when sin (2πx/L) = 0, corresponding to x = L/2 and any value of y. This
node is therefore a line at x = L/2 and parallel to the y axis . Similarly there is
another nodal line at y = L/2 and parallel to the x axis .
E7D.4(a) The energy levels for a 2D rectangular box, side lengths L 1 , L 2 are
h 2 n 12 n 22
E n 1 ,n 2 = ( + )
8m L 21 L 22
where n 1 and n 2 are integers greater than or equal to 1.
For the specific case where L 1 = L, L 2 = 2L,
h 2 n 12 n 22 h2 n 22
E n 1 ,n 2 = ( 2+ ) = (n 2
+ )
8m L (2L)2 8mL 2 1
4
The energy of the state with n 1 = n 2 = 2 is then
h2 22 h2
E 2,2 = 2
(22 + ) = (5)
8mL 4 8mL 2
This is degenerate with the state with n 1 = 1, n 2 = 4
h2 42 h2
E(1,4) = (1 2
+ ) = (5)
8mL 2 4 8mL 2
The question notes that degeneracy frequently accompanies symmetry, and

suggests that one might be surprised to find degeneracy in a box with unequal
lengths. Symmetry is a matter of degree. This box is less symmetric than a
square box, but it is more symmetric than boxes whose sides have a non-integer
or irrational ratio. Every state of a square box except those with n 1 = n 2 is
degenerate (with the state that has n 1 and n 2 reversed). Only a few states in this
rectangular box are degenerate. In this system, a state (n 1 , n 2 ) is degenerate
with a state (n 2 /2, 2n 1 ) as long as the latter state (a) exists (that is, n 2 /2 must
be an integer) and √ (b) is distinct from (n 1 , n 2 ). A box with incommensurable
sides, say, L and 2L, would have no degenerate levels.
E7D.5(a) The energy levels of a cubic box are given by [7D.13b–253]
h2
E n 1 ,n 2 ,n 3 =
(n 2 + n 22 + n 32 )
8mL 2 1
where n 1 , n 2 , n 3 are integers greater than or equal to 1. Hence the lowest energy
state is that with n 1 = n 2 = n 3 = 1, with energy
h2 h2
E(1,1,1) = (1 2
+ 1 2
+ 1 2
) = (3)
8mL 2 8mL 2
and so the energy of the level with energy three times that of the lowest is
9h 2 /8mL 2 , which will be produced by states for which n 12 + n 22 + n 32 = 9. There
are three states with this energy, (n 1 , n 2 , n 3 ) = (1, 2, 2), (2, 1, 2), (2, 2, 1), and
so the degeneracy is 3 .
E7D.6(a) The transmission probability [7D.20a–255] depends on the energy of the tun-
nelling particle relative to the barrier height (ε = E/V0 = (1.5 eV)/(2.0 eV) =
0.75), the width of the barrier, (L = 100 pm), and the decay parameter of the
wavefunction inside the barrier (κ)
(2m(E − V0 ))1/2
κ=
ħ
(2 × (9.1094 × 10−31 kg) × [(2.0 − 1.5) eV × 1.6022 × 10−19 J eV−1 ])1/2
=
1.0546 × 10−34 J s
9 −1
= 3.68... × 10 m
The product κL = (3.68... × 109 m−1 ) × (100 × 10−12 m) = 0.368..., and so the
transmission probability is given by
−1 −1
(eκ L − e−κL )2 (e0.368 ... − e−0.368 ... )2
T = [1 + ] = [1 + ] = 0.84
16ε(1 − ε) 16 × 0.75 × (1 − 0.75)
E7D.7(a) The linear momentum of a free electron is given by
p = ħk = (1.0546 × 10−34 J s) × (3 × 109 m−1 ) = 3 × 10−25 kg m s−1
where 1 J = 1 kg m2 s−2 is used; note that 1 nm−1 = 1 × 109 m−1 . The kinetic
energy is given by [7D.2–247]
(ħk)2 ((1.0546 × 10−34 J s) × (3 × 109 m−1 ))2
Ek = = = 5 × 10−20 J
2m 2 × (9.1094 × 10−31 kg)
E7D.8(a) The electron is travelling in the negative x direction hence the momentum,
and the value of k, are both negative. The kinetic energy is related to k through
[7D.2–247], E k = k 2 ħ 2 /2m, which is rearranged to give k
1/2
2mE k 1/2 2 × (2.0 × 10−3 kg) × (20 J)
k = −( ) = − ( ) = −2.7 × 1033 m−1
ħ2 (1.0546 × 10−34 J s)2
The wavefunction is then, with x is measured in metres,
m−1 )x
ψ(x) = eik x = e−i(2.7×10
33
E7D.9(a) The energy levels of a particle in a box are given by [7D.6–249], E n = n 2 h 2 /8mL 2 ,
where n is the quantum number. With the mass equal to that of the electron
and the length as 1.0 nm, the energies are
n2 h2 (6.6261 × 10−34 J s)2
En = = n 2
×
8mL 2 8 × (9.1094 × 10−31 kg) × (1.0 × 10−9 m)2
= n 2 × (6.02... × 10−20 J)
To convert to kJ mol−1 , multiply through by Avogadro’s constant and divide by

1000.
6.0221 × 1023 mol−1

= n 2 × (6.02... × 10−20 J) × = n 2 × (36.2... kJ mol−1 )
1000
To convert to electronvolts divide through by the elementary charge
1
= n 2 × (6.02... × 10−20 J) × = n 2 × (0.376... eV)
1.6022 × 10−19 C
To convert to reciprocal centimetres, divide by hc, with c in cm s−1
(6.02... × 10−20 J)
= n2 × = n 2 × (3.03... × 103 cm−1 )
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
The energy separation between two levels with quantum numbers n 1 and n 2 is
∆E(n 1 , n 2 ) = E n 2 − E n 1 = (n 22 − n 12 ) × (6.02... × 10−20 J)
The values in the other units are found by using the appropriate value of the
constant, computed above.
(i) ∆E(1, 2) = 1.8 × 10−19 J , 1.1 × 102 kJ mol−1 , 1.1 eV , or 9.1 × 103 cm−1
(ii) ∆E(5, 6) = 6.6 × 10−19 J , 4.0 × 102 kJ mol−1 , 4.1 eV , or 3.3 × 104 cm−1
E7D.10(a) The wavefunctions are ψ 1 (x) = (2/L)1/2 sin(πx/L) and ψ 2 (x) = (2/L)1/2 ×
sin(2πx/L). They are orthogonal if ∫ ψ ∗1 ψ 2 dτ = 0. In this case the integral is
taken from x = 0 to x = L as outside this range the wavefunctions are zero. The
required integral is of the form of Integral T.5, with A = π/L, B = 2π/L, and
a=L
2 L πx 2πx 2 sin(−πL/L) sin(3πL/L)
∫ sin ( ) sin ( ) dx = [ − ]
L 0 L L L −2π/L 6π/L
2 sin(−π) sin(3π)
= [ − ]=0
L −2π/L 6π/L
where sin nπ = 0 for integer n is used. The two wavefunctions are orthogonal.
E7D.11(a) The particle in a box wavefunction with quantum number n is given by ψ n (x) =
(2/L)1/2 sin(nπx/L) The probability of finding the electron in a small region of
space δx centred on position x is approximated as ψ 2 (x)δx. For this Exercise
x = 0.50L, δx = 0.02L.
For the case where n = 1
√ 2
ψ 1 (0.50L)2 × 0.02L = [ 2/L sin (π(0.50L)/L)] × 0.02L
= (2/L) sin2 (0.50π) × 0.02L = 0.04
For the case where n = 2

√ 2
ψ 2 (0.50L)2 × 0.02L = [ 2/L sin (2 × π(0.50L)/L)] × 0.02L
= (2/L) sin2 (1.00π) × 0.02L = 0
ψ1
ψ 12
x/L
0.2 0.4 0.6 0.8 1.0
Figure 7.1
E7D.12(a) The wavefunction √ of the lowest energy state for a particle in a box has n = 1
and is ψ 1 (x) = 2/L sin(πx/L), which leads to a probability density P1 (x) =
∣ψ 1 (x)∣2 = (2/L) sin2 (πx/L). Graphs of these functions are shown in Fig 7.1.
The probability density is symmetric about x = L/2. Therefore, there is an equal
probability of observing the particle at an arbitrary position x ′ and at L − x ′ , so
it follows that the average position of the particle must be at L/2.
√
E7D.13(a) The wavefunction for the state with n = 1 is ψ 1 (x) = 2/L sin(πx/L), which
leads to a probability density P1 (x) = ∣ψ 1 (x)∣2 = (2/L) sin2 (πx/L); these are
plotted in Fig 7.1.
The plotted probability density gives the probability of the particle being found
in an interval at a particular value of x – that is, it is the probability density of
x. It is not the probability density of x 2 , so there is no simple way of inferring
from the plot the value of ⟨x 2 ⟩. The fact that the probability density of x is
symmetric about L/2 does not imply that the probability density of x 2 has the
same property. There is therefore no reason to assume that ⟨x 2 ⟩ = (L/2)2 .
E7D.14(a) For an electron in a square well of side L the energy of the state characterized
by quantum numbers n 1 and n 2 is given by [7D.12b–253], E n x ,n y = h 2 (n 2x +
n 2y )/8m e L 2 , where n x , n y are integers greater than or equal to one. Hence, the
minimum or zero-point energy has n x = n y = 1,
E 1,1 = h 2 (12 + 12 )/8m e L 2 = h 2 /4m e L 2
Setting this equal to the rest energy m e c 2 , and rearranging for the length, L
h 2 /4m e L 2 = m e c 2
L = h/2m e c = λ C /2
where λ C = h/(m e c) is the Compton wavelength.
√in a one-dimensional box with length L and in

E7D.15(a) The wavefunction for a particle
the state with n = 3 is ψ(x) = 2/L sin(3πx/L) giving a probability density of
P(x) = ∣ψ(x)∣2 = (2/L) sin2 (3πx/L).
As this is a trigonometric function, the function is maximized when sin2 (3πx/L) =

1, which is when sin(3πx/L) = ±1; these values occur when the argument of
the sine function is an odd multiple of π/2
3πx/L = π/2, 3π/2, 5π/2... hence x = L/6, L/2, 5L/6
The probability density is zero when the wavefunction is zero, which is when
the argument of the sine function is a multiple of π
3πx/L = 0, π, 2π, ... hence x = 0, L/3, 2L/3, L
P7D.1 The energy levels of a particle in a one dimensional box are E n = h 2 n 2 /8mL 2 .
The mass of an O2 molecule is 2 × 16.00 amu × 1.6605 × 10−27 kg = 5.3136 ×
10−26 kg. The separation in energy of the lowest two levels, being n = 1 and
n = 2, is given by
h 2 22 h 2 12 h2
∆E = E 2 − E 1 = − = (4 − 1)
8mL 2 8mL 2 8mL 2
(6.6261 × 10−34 J s)2
= ×3
8 × (5.3136 × 10−26 kg) × (5.0 × 10−2 m)
= 6.2 × 10−41 J
To find the value of n that makes the energy equal to the average thermal energy,
set E n = 12 kT and solve for n
√
2L mkT
n=
h
1/2
2×(5.0 × 10−2 m)×[(5.3136 × 10−26 kg)×(300 K)×(1.3806 × 10−23 J K−1 )]
=
6.6261 × 10−34 J s
= 2.23... × 109 = 2.2 × 109
The separation of level n from the one immediately below it, n − 1, is E n − E n−1
h 2 n 2 h 2 (n − 1)2 h 2 (2n − 1)
E n − E n−1 = − =
8mL 2 8mL 2 8mL 2
−34
(6.6261 × 10 J s) × (2.23... × 109 − 1)
2
=
8 × (5.3136 × 10−26 kg) × (5.0 × 10−2 m)2
= 1.8 × 10−30 J
P7D.3 The normalized wavefunction with n = 1 is ψ 1 (x) = (2/L)1/2 sin(πx/L), and

this is used in [7C.11–242], ⟨Ω⟩ = ∫ ψ ∗ Ω̂ψ dτ, to compute the expectation value
for the cases Ω̂ = x and Ω̂ = x 2 .
L L
⟨x⟩ = ∫ ψ ∗1 xψ 1 dx = (2/L) ∫ x sin2 (πx/L) dx
0 0
The integral is of the form of Integral T.11 with a = L and k = π/L
=sin 2π=0 =cos(2π)=1

³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
= (2/L)[L /4 − (L /4π) sin(2πL/L) −(L 2 /8π 2 ){cos(2πL/L) −1}]
2 2
= L/2
L L
⟨x 2 ⟩ = ∫ ψ ∗1 x 2 ψ 1 dx = (2/L) ∫ x 2 sin2 (πx/L) dx
0 0
The integral is of the form of Integral T.12 with a = L and k = π/L
= (2/L) [L 3 /6 − (L 3 /4π − L 3 /8π 3 ) sin(2πL/L) − (L 3 /4π 2 ) cos(2πL/L)]

= L 2 /3 − 1/2π 2
P7D.5 (a) Assuming that the system can be modelled as a 1D particle in a box, the
energy levels of the system are given by [7D.6–249], E n = h 2 n 2 /8mL 2 .
Hence, the separation between the levels n and n + 1 is
∆E = E n+1 − E n = h 2 [(n + 1)2 − n 2 ]/8mL 2 = h 2 (2n + 1)/8mL 2
The conjugated system consists of 12 atoms, and so there are 11 bonds,

therefore the length of the box is 11 times the average internuclear distance
L = 11×140 pm = 1.54 nm. Hence the separation between the levels with
n = 6 and n = 7 is
(6.6261 × 10−34 J s)2 × (2 × 6 + 1)

∆E = = 3.30 × 10−19 J
8 × (9.1094 × 10−31 kg) × (1.54 × 10−9 m)2
(b) The energy and frequency of a transition are related by [7A.9–227], ∆E =

hν, rearranging for the frequency gives
ν = ∆E/h = (3.30... × 10−19 J)/(6.6261 × 10−34 J s) = 4.98 × 1014 Hz
(c) The terms in the energy expression that change with the number of con-
jugated atoms N are the quantum number of the highest energy occupied
state, n, and the length of the box, L. The energy, and hence frequency, of
the transition is directly proportional to 2n +1 and inversely proportional
to L 2 . Because the there are N electrons which occupy the states in pairs,
the quantum number of the highest occupied state is n = N/2 (rounded
up if N is odd). The length L is proportional to the number of bonds
in the molecule, which is N − 1. Hence, the frequency of the transition is
proportional to (2(N/2)+1)/(N −1)2 = (N +1)/(N −1)2 ≈ N −1 , and so
the absorption spectrum of a linear polyene shifts to a lower frequency
as the number of conjugated atoms increases .
P7D.7 The expectation values are ⟨x⟩ = L/2, ⟨x 2 ⟩ = L 2 (1/3−1/2n 2 π 2 ), ⟨p x ⟩ = 0, ⟨p2x ⟩ =

n 2 h 2 /4L 2 .
(a) The uncertainty in the position is
∆x = [⟨x 2 ⟩ − ⟨x⟩2 ]1/2 = [L 2 (1/3 − 1/2n 2 π 2 ) − (L/2)2 ]1/2

= L(1/12 − 1/2n 2 π 2 )1/2
and in the momentum
∆p x = [⟨p2x ⟩ − ⟨p x ⟩2 ]1/2 = [n 2 h 2 /4L 2 − 02 ]1/2 = nh/2L
(b) The product of these is then
∆x∆p x = L(1/12 − 1/2n 2 π 2 )1/2 × nh/2L

= (nh/2)(1/12 − 1/2n 2 π 2 )1/2
(c) The Heisenberg uncertainty principle states that ∆x∆p x ≥ ħ/2. Rewriting
the expression in (b) in terms of ħ gives ∆x∆p x = (ħ/2) × 2πn(1/12 −
1/2n 2 π 2 )1/2 . For n = 1
∆x∆p x = (ħ/2) × 2π(1/12 − 1/2π 2 )1/2 = 1.13... × (ħ/2)
and for n = 2,
∆x∆p x = (ħ/2) × 4π(1/12 − 1/8π 2 )1/2 = 3.34... × (ħ/2)
In both cases ∆x∆p x ≥ (ħ/2), which is consistent with the uncertainty

principle. It is evident that ∆x∆p x is an increasing function of n and
therefore it can be inferred that the uncertainly principle is obeyed for all
n.
P7D.9 (a) In 3D the Schrödinger equation with a potential energy function V (x, y, z)
is given by
ħ2 ∂2 ∂2 ∂2
− ( 2 + 2 + 2 ) ψ + V (x, y, z)ψ = Eψ
2m ∂x ∂y ∂z
In this case V (x, y, z) = 0 inside the box and infinite elsewhere. It fol-
lows that the wavefunction is zero outside the box. Inside the box the
Schrödinger equation is
ħ2 ∂2 ∂2 ∂2
− ( 2 + 2 + 2 ) ψ = Eψ
2m ∂x ∂y ∂z
Assume that the solution is a product of three functions of a single vari-

able, that is let ψ = X(x)Y(y)Z(z). Substituting this into the Schrödinger
equation gives
ħ2 d2 X d2 Y d2 Z
− (Y Z 2 + XZ 2 + Y Z 2 ) = EXY Z
2m dx dy dz
Dividing through by the function XY Z gives

ħ 2 1 d2 X 1 d2 Y 1 d2 Z
− ( + + )=E
2m X dx 2 Y dy 2 Z dz 2
Now rearrange such that the terms that depend on x only are on the left
hand side
ħ 2 d2 X ħ 2 1 d2 Y 1 d2 Z
− =E+ ( + )
2mX dx 2 2m Y dy 2 Z dz 2
The left hand side depends only on x, whilst the right side depends only
on y and z. The only way that this can be true for all x, y, z is if both sides
are equal to a constant, arbitrarily called E x , such that
ħ 2 d2 X ħ 2 1 d2 Y 1 d2 Z
− = Ex Ex = E + ( + )
2mX dx 2 2m Y dy 2 Z dz 2
The same procedure is followed with the functions of y and z to give
ħ 2 d2 Y ħ 2 d2 Z
− = Ey − = Ez
2mY dy 2 2mZ dz 2
where E = E x + E y + E z . The assumption that the wavefunction can be
written as a product of single-variable functions is a valid one, as ordinary
differential equations can be found for the assumed factors. That is what
it means for a partial differential equation to be separable.
(b) The differential equation involving the function X is recognized as being
the same as the Schrödinger equation for a particle in a one-dimensional
box, whose solutions are already known: X(x) = (2/L)1/2 sin(nπx/L)
with energies E n = n 2 h 2 /8mL 2 . The product X(x)Y(y)Z(z) which is
the solution to this three-dimensional problem will therefore be a product
with three such one-dimensional functions along x, y and z, each with the
appropriate length L i and quantum number n i
ψ(x, y, z) = (2/L 1 )1/2 sin(n x πx/L 1 ) × (2/L 2 )1/2 sin(n y πy/L 2 )

× (2/L 3 )1/2 sin(n z πz/L 3 )
Likewise, the total energy will be the sum of energies along each direction
h 2 n 2x n y n 2z
2
E = Ex + E y + Ez = ( + + )
8m L 21 L 22 L 23
(c) For a cubic box, L 1 = L 2 = L 3 = L and so

E = h 2 (n 2x + n 2y + n 2z )/8mL 2
Generally states with all three quantum numbers the same are non degen-
erate, if two quantum numbers are the same they are triply degenerate,
and if all three quantum numbers are the different they are six-fold de-
generate. In addition, there may be ‘accidental’ degeneracies, such as that
between (n x , n y , n z ) = (3, 3, 3) and (5, 1, 1). The energy level diagram is
shown in Fig. 7.2; in the diagram the states are labelled (n x , n y , n z ).
30
(4,3,2)
(3,3,3),(5,1,1)
(4,3,1)
(4,2,2)
(3,3,2)
(4,2,1)
20
(3,3,1)
E n x ,n y ,n z /(h 2 /8ml 2 ) (4,1,1)
(3,2,2)
(3,2,1)
(2,2,2)
(3,1,1)
10
(2,2,1)
(2,1,1)
(1,1,1)
0
Figure 7.2
(d) Comparing the 3D and 1D energy diagrams, the one dimensional case is
more sparse than the three dimensional case, as the fifteenth level is not
reached until E n /(h 2 /8ml 2 ) = 225. In addition, none of the levels in the
one-dimensional case are degenerate.
P7D.11 The rate of tunnelling is proportional to the transmission probability, given by

[7D.20b–256], T = 16ε(1 − ε)e−2κL . Therefore the ratio of tunnelling rates
is equal to the corresponding ratio of transmission probabilities. The desired
factor is T1 /T2 where the subscripts denote different tunnelling distances.
T1 16ε(1 − ε)e−2κ L 1
= = e−2κ(L 1 −L 2 )
T2 16ε(1 − ε)e−2κ L 2
With κ = 7 nm−1 , L 1 = 1 nm and L 1 = 2 nm

T1 −1
= e−2κ(L 1 −L 2 ) = e−2(7 nm )[(1 nm)−(2 nm)] = 1.20 × 106
T2
P7D.13 (a) The wavefunctions in each region are ψ 1 (x) = eik 1 x + B 1 ei k 1 x , ψ 2 (x) =
A 2 e k 2 x + B 2 e−k 2 x , ψ 3 (x) = A 3 eik 3 x . With this choice of A 1 = 1, the
transmission probability is simply T = ∣A 3 ∣3 .
The wavefunction coefficients are determined by the criteria that both the
wavefunction and its first derivative with respect to x are continuous at
the potential boundaries, ψ 1 (0) = ψ 2 (0), dψ 1 (0)/dx = dψ 2 (0)/dx, ψ 2 (W) =
ψ 3 (W), dψ 3 (W)/dx = dψ 2 (W)/dx. These establish the algebraic rela-
tionships
(a) 1 + B1 = A2 + B2
(b) ik 1 − ik 1 B 1 = A 2 k 2 − B 2 k 2
(c) A 2 e k 2 W + B 2 e−k 2 W = A 3 eik 3 W
(d) A 2 k 2 e k 2 W − B 2 k 2 e−k 2 W = ik 3 A 3 eik 3 W
These equations need to be solved for A 3 and hence B 1 , A 2 , B 2 need to be
eliminated. A 3 appears in (c) and (d) only. Solving these equations for
A 3 eik 3 W and setting the results equal to each other yields
A 2 e k 2 W + B 2 e−k 2 W = (k 2 /ik 3 )(A 2 k 2 e k 2 W − B 2 e−k 2 W )
This equation can be solved for B 2 in terms of A 2
A 2 e2k 2 W (k 2 /ik 3 − 1) A 2 e2k 2 W (k 2 − ik 3 )

B2 = =
k 2 /ik 3 + 1 k 2 + ik 3
Note that B 1 appears only in (a) and (b). Solving these equations for B 1
and setting the results equal to each other yields
B 1 = A 2 + B 2 − 1 = 1 − (1/ik 1 )(A 2 k 2 − B 2 k 2 )
Substituting the expression for B 2 into this equation yields
e2k 2 W (k 2 − ik 3 ) e2k 2 W (k 2 − ik 3 )
A 2 [1 + ] = 2ik 1 − A 2 k 2 [1 − ]
k 2 + ik 3 k 2 + ik 3
This can be rearranged to give
−1
1 e2k 2 W (k 2 − ik 3 ) k 2 (k 2 − ik 3 ) k 2 e2k 2 W (k 2 − ik 3 )
A2 = [ + + − ]
2 2(k 2 + ik 3 ) 2ik 1 (k 2 + ik 3 ) 2ik 1 (k 2 + ik 3 )
Hence,
−1
A 2 e2k 2 W (k 2 − ik 3 ) k 2 + ik 3
B2 = = A 2 [ 2k W ]
k 2 + ik 3 e 2 (k 2 − ik 3 )
−1
k 2 + ik 3 1 k2 k2
=[ 2k W
+ + 2k W
− ]
2e 2 (k 2 − ik 3 ) 2 2ik 1 e 2 2ik 1
This gives
4k 1 k 2 eik 2 W
A 3 = e−ik 3 W (A 2 e k 2 W + B 2 e−k 2 W ) =
(ia + b)e k 2 W − (ia − b)e−k 2 W
Since sinh z = 12 (ez − e−z ), substitute e k 2 L = 2 sinh k 2 L + e k 2 L
2k 1 k 2 eik 2 L
A3 =
(ia + b) sinh(k 2 W) + be−k 2 W
The transmission coefficient is then

4k 12 k 22
T = ∣A 3 ∣2 =
(a 2 + b 2 ) sinh2 (k 2 L) + b 2
where a 2 + b 2 = (k 12 + k 22 )(k 22 + k 32 ), and b 2 = k 22 (k 1 + k 3 )2 .

(b) In the special case, for which V1 = V3 = 0, k 1 = k 3 . Additionally, k 1 /k 2 =
E/(V2 − E) = ε/(1 − ε), where ε = E/V2 . Also
a 2 + b 2 = (k 12 + k 22 )2 = k 24 [1 + (k 1 /k 2 )2 ]2
b 2 = 4k 12 k 22
Hence
−1
b2 a2 + b2
T= = (1 + sinh2 k 2 W)
2
b 2 + (a 2 + b 2 ) sinh (k 2 W) b2
−1 −1
sinh2 (k 2 W) [ 1 (e k 2 W − e−k 2 W )]2
= (1 + ) = (1 + 2 )
4ε(1 − ε) 4ε(1 − ε)
−1
(e k 2 W − e−k 2 W )2
= (1 + ) (7.1)
16ε(1 − ε)
Which is eqn 7D.20a as expected. In the case of a high, wide barrier,

k 2 W ≫ 1. This implies that e−k 2 W is negligibly small in comparison to
e k 2 W , and that 1 is negligibly small compared to e2k 2 W /16ε(1 − ε), such
that −1
T ≈ (e2k 2 W /16ε(1 − ε)) = 16ε(1 − ε)e−2k 2 W
(c) The function plotted in Fig. 7.3 is eqn 7.1 with the parameters given and
k 2 = [2m(V2 − E)]1/2 /ħ ε = E/V2
√
P7D.15 (a) The particle in a box wavefunctions are given by ψ n (x) = 2/L sin(nπx/L),
giving a probability density of Pn (x) = ∣ψ n (x)∣2 = (2/L) sin2 (nπx/L).
This probability density is plotted in Fig. 7.4 for n = 1 . . . 5, and for n = 50
in Fig. 7.5.
The correspondence principle is that as the quantum number becomes
large the quantum result must converge with the classical result. For a
particle in a box, the classical result is an even probability. As n increases
the oscillations become more rapid leading to a more uniform distribu-
tion especially if it is averaged over short distance which is a small fraction
of the width of the box.
(b) The transmission probability of a particle of mass m passing through a
barrier of height V0 , length W is given by, [7D.20a–255]
−1
(eκW − e−κW )2
T = (1 + )
16ε(1 − ε)
0.2
T
0.1
0.0
10 12 14 16 18 20
−1
V2 /kJ mol
Figure 7.3
1.0 n=1
n=2
n=3
n=4
ψ 2n /(2/L)
n=5
0.5
0.0
0.0 0.2 0.4 0.6 0.8 1.0
x/L
Figure 7.4
√ √
where ε = E/V0 and κ = 2m(V0 − E)/ħ = 2mV0 (1 − ε)/ħ. A plot of
T versus ε is shown in Fig. 7.6 for the passage by a H2 molecule, a proton,
and an electron.
The curves in the figure differ in the value of W(mV0 )1/2 /ħ , a measure of
the obstruction that the barrier represents taking into account its height,
width and the type of particle. The curves can be thought of as having the
same value of W and V0 , but differing only in m. The values of W and V0
were chosen such that the proton and hydrogen molecule exhibit ‘typical’
tunnelling behaviour: if the incident energy is small enough, there is
practically no transmission, and if the incident energy is high enough,
transmission is virtually complete. A barrier through which a proton and
hydrogen molecule exhibit tunnelling behaviour is practically no barrier
for an electron on account of the much smaller mass of the latter. On this
plot, T for the electron is essentially 1.
(c) The wavefunction of a 2D particle in a box with quantum numbers n 1 and
1.0
ψ 50
2
/(2/L) 0.5
0.0
0.0 0.2 0.4 0.6 0.8 1.0
x/L
Figure 7.5
1.0 H2
p+
e–
T
0.5
0.0
0.0 1.0 2.0 3.0 4.0
ε
Figure 7.6
n 2 is given by [7D.12a–253]
ψ n 1 ,n 2 (x, y) = (2/L x )1/2 (2/L y )1/2 sin(n 1 πx/L x ) sin(n 2 πy/L y )
This is plotted for n 1 = 1, n 2 = 1, and for n 1 = 1, n 2 = 2 in Fig. 7.7; the

wavefunctions with n 1 = 2, n 2 = 1, and with n 1 = 2, n 2 = 2 are plotted in
Fig. 7.8.
As can be seen from the plots, the function sin(n 1 πx/L x ) has n 1 −1 nodes
where the function passes through zero; the boundaries do not count
as nodes because the wavefunction does not pass through zero at these
points. Similarly the function sin(n 2 πy/L y ) has n 2 − 1 nodes. In two
dimensions a node in the wavefunction along x or y leads to a nodal line,
and the total number of these is (n 1 − 1) + (n 2 − 1) = n 1 + n 2 − 2 .
n 1 = 1, n 2 = 1
ψ/(2/L x )1/2 (2/L y )1/2

1
0.5
1
0
0 0.2 0.5
0.4 0.6 0.8 1 0 y/L y
x/L x
n 1 = 1, n 2 = 2
psi/(2/L x )1/2 (2/L y )1/2
1
−1
0 0.2 0.5
0.4 0.6 0.8 1 0 y/L y
x/L x
Figure 7.7
7E Vibrational motion
D7E.1 If the harmonic oscillator were to have zero energy the particle would be at
rest and located at the bottom of the potential energy well (x = 0). Such
an arrangement has no uncertainty in either the position or the momentum,
which is not in accord with the uncertainty principle. By giving the oscillator
zero-point energy, there is some uncertainty in the position as the particle
oscillates back and forth, and then it is possible for there to be the appropriate
uncertainty in the momentum such that the uncertainty principle is satisfied.
The existence of zero-point energy is thus traced to the need to satisfy the
uncertainty principle.
D7E.3 For the harmonic oscillator the spacing of the energy levels is constant. There-
fore, relative to the energy of the oscillator, the spacing becomes progressively
smaller as the quantum number increases. In the limit of very high quantum
numbers this spacing becomes negligible compared to the total energy, and
effectively the energy can take any value, as in the classical case.
n 1 = 2, n 2 = 1
psi/(2/L x )1/2 (2/L y )1/2

1
1
−1
0 0.2 0.5
0.4 0.6 0.8 1 0 y/L y
x/L x
n 1 = 2, n 2 = 2
psi/(2/L x )1/2 (2/L y )1/2
1
−1
0 0.2 0.5
0.4 0.6 0.8 1 0 y/L y
x/L x
Figure 7.8
As is shown in Fig. 7E.7 on page 263, as the quantum number becomes large the
probability density clusters more and more around the classical turning points
of the classical harmonic oscillator (that is, the points at which the kinetic
energy is zero). Because the classical oscillator is moving most slowly near
these points, they are the displacements at which it is most probable that the
oscillator will be found. Again, the quantum and classical results converge at
high quantum numbers
E7E.1(a) The wavefunctions are depicted in Fig. 7E.6 on page 262. The general form of
the harmonic oscillator wavefunctions is ψ υ = N υ H υ (y)e−y /2 , where N υ is
2
a normalization constant and H υ (y) is the Hermite polynomial of order υ in

the reduced position variable y. Nodes occur when the wavefunction passes
through zero. The wavefunction asymptotically approaches zero at y = ±∞,
but as the function does not pass through zero these limits do not count as
nodes. The nodes in the wavefunction therefore correspond to the solutions of
H υ (y) = 0. H υ (y) is a polynomial of order υ, meaning that the highest power
of y that occurs is y υ ; such polynomials in general have υ solutions and hence

there are υ nodes. Therefore (a) the wavefunction with υ = 3 has 3 nodes; (b)
the wavefunction with υ = 4 has 4 nodes.
The wavefunction with υ = 2 is ψ 2 (y) = N 2 (y 2 − 1)e−y /2 . Nodes occur when

2
E7E.2(a)
the wavefunction passes through zero; the wavefunction approaches zero at
y = ±∞, but these do not count as nodes as the wavefunction does not pass
through zero. It is evident that the nodes occur when y 2 − 1 = 0, which solves
to give nodes at y = −1, +1 .
The wavefunction with υ = 1 is ψ 1 (y) = N 1 ye−y /2 , which gives a probability

2
E7E.3(a)
density of P(y) = ∣ψ 2 (y)∣2 = N 12 y 2 e−y . The extrema are located by differentiat-
2
ing the wavefunction, setting the result to 0 and solving for y. The differential
is evaluated using the product rule
de−y ⎞
2
dψ 1 (y) ⎛ dy 2 −y 2
= N 12 [2ye−y + y 2 × (−2ye−y )]
2 2
= N 12 e + y2
dy ⎝ dy dy ⎠
= N 12 2y(1 − y 2 )e−y
2
This function is equal to zero when y = 0, or 1 − y 2 = 0 or when e−y = 0. These

2
correspond to extrema at y = 0, ±1, ±∞.

To identify the maxima, consider the form of the probability density. This goes
to zero at y = 0, ±∞, and so these must be minima, implying that y = ±1
correspond to maxima.
E7E.4(a) The zero-point energy of a harmonic oscillator is given by [7E.5–260], E 0 =

1
2
ħω, where the frequency ω is given by [7E.3–260], ω = (k f /m)1/2 . For this
system,
1/2
E0 = 1
2
× (1.0546 × 10−34 J s) × [(155 N m−1 )/(2.33 × 10−26 kg)]
= 4.30 × 10−21 J
E7E.5(a) The separation between adjacent energy levels of a harmonic oscillator is [7E.4–
260], ∆E = ħω, where the frequency, ω is given by [7E.3–260], ω = (k f /m)1/2 .
This is rearranged for the force constant as k f = m(∆E/ħ)2 . Evaluating this
gives
2
k f = (1.33 × 10−25 kg) × [(4.82 × 10−21 J)/(1.0546 × 10−34 J s)] = 278 N m−1
E7E.6(a) The separation between adjacent energy levels of a harmonic oscillator is [7E.4–
260], ∆E = ħω, where the frequency, ω is given by [7E.3–260], ω = (k f /m)1/2 .
The Bohr frequency condition [7A.9–227], ∆E = hν, can be rewritten in terms
of the wavelength as ∆E = hc/λ. The wavelength of the photon corresponding
to a transition between adjacent energy levels is therefore given by ħω = hc/λ,
or ħ(k f /m)1/2 = hc/λ. Solving for λ gives λ = 2πc/(k f /m)1/2 ; with the data
given
2π × (2.9979 × 108 m s−1 )

λ= 1/2
[(855 N m−1 )/(1.0078 × 1.6605 × 10−27 kg)]
= 2.64 × 10−6 m
E7E.7(a) The wavefunctions are depicted in Fig. 7E.6 on page 262; they are real. Two
wavefunctions are orthogonal if ∫ ψ ∗i ψ j dτ = 0. In this case the wavefunctions
are ψ 0 (y) = N 0 e−y /2 and ψ 1 (y) = N 1 ye−y /2 , and the integration is from
2 2
y = −∞ to +∞. The integrand ψ 0 ψ 1 is N 0 N 1 ye−y , which is an odd function,

2
meaning that its value at −y is the negative of its value at +y. The integral of
an odd function over a symmetric range is zero, hence these wavefunctions are
orthogonal.
E7E.8(a) The zero-point energy of a harmonic oscillator is given by [7E.5–260], E 0 =

1
2
ħω, where the frequency ω is given by [7E.3–260], ω = (k f /µ)1/2 . The effec-
tive mass µ of a diatomic AB is given by [7E.6–260], µ = (m A m B )/(m A + m B ).
In the case of a homonuclear diatomic A2 this reduces to µ = m A /2. With the
data given
E0 = 1
2
× (1.0546 × 10−34 J s)
1/2
× [(329 N m−1 )/( 12 × 34.9688 × 1.6605 × 10−27 kg)]
= 5.61 × 10−21 J
E7E.9(a) For the state with υ = 0 the energy of a harmonic oscillator is given by [7E.5–
260], E 0 = 12 ħω, where the frequency ω is given by [7E.3–260], ω = (k f /m)1/2 .
(i) For the system with k f = 1000 N m−1 the energy of the state with υ = 0 is
E0 = 1
2
× (1.0546 × 10−34 J s) × [(1000 N m−1 )/(1 × 1.6605 × 10−27 kg)]1/2
= 4.09... × 10−20 J = 4.09 × 10−20 J
The classical turning points of this state occur when E 0 = 12 k f x tp

2
. Solving
√
this for x tp leads to x tp = ± 2E 0 /k f , giving a separation of
√ √
2 2E 0 /k f = 2 2 × (4.09... × 10−20 J)/(1000 N m−1 ) = 18.1 pm
(ii) For the system with k f = 100 N m−1 the energy of the state with υ = 0 is
E0 = 1
2
× (1.0546 × 10−34 J s) × [(100 N m−1 )/(1 × 1.6605 × 10−27 kg)]1/2
= 1.29 × 10−20 J
The classical turning points of this state occur when E 0 = 12 k f x tp

2
. Solving
√
this for x tp leads to x tp = ± 2E 0 /k f , giving a separation of
√ √
2 2E 0 /k f = 2 2 × (1.29... × 10−20 J)/(100 N m−1 ) = 32.2 pm
P7E.1 The intermolecular potential is electrostatic in origin and arises from the in-
teraction between the protons and electrons, the number and distribution of
which remain the same on isotopic substitution and so therefore does the po-
tential. The force constant, which reflects the shape of the intermolecular po-
tential, is therefore unaffected by such a change.
(a) The force constants for the molecules are the same therefore ω A′ B /ω AB =
(k f /µ A′ B )1/2 /(k f /µ AB )1/2 = (µ AB /µ A′ B )1/2 ; multiplying through by ω AB ,
gives ω A′ B = ω AB (µ AB /µ A′ B )1/2 . The vibrational frequency in Hz, ν, is
related to ω by ν = ω/2π, hence ν A′ B = ν AB (µ AB /µ A′ B )1/2 .
(b) The effective mass of 1H 35Cl is µ 1 = (1 × 35)/(35 + 1) = 0.972... m u , that
of 2H 35Cl is µ 2 = (2 × 35)/(35 + 2) = 1.89... m u , and that of 1H 37Cl is
µ 3 = (1 × 37)/(1 + 37) = 0.973... m u .
The vibrational frequency of 2H 35Cl is
ν 2 = ν 1 (µ 1 /µ 2 )1/2 = (5.63 × 1014 Hz)[(0.972... m u )/(1.89... m u )]1/2

= 4.04 × 1014 Hz
and for 1H 37Cl
ν 3 = ν 1 (µ 1 /µ 3 )1/2 = (5.63 × 1014 Hz)[(0.972... m u )/(0.973... m u )]1/2

= 5.63 × 1014 Hz
P7E.3 Assuming that the force constant is the same √ for all √ the molecules,
√ then the
ratio of frequencies is ω 2 /ω 1 = ν 2 /ν 1 = k f /µ 2 / k f /µ 1 = µ 1 /µ 2 . For a
homonuclear diatomic molecule A2 the effective mass is µ A2 = m A2 /2m A =
m A /2. Hence, µ 1H2 = 1/2 = 0.5 m u , µ 2H2 = 2/2 = 1 m u , µ 3H2 = 3/2 = 1.5 m u .
√ √
ν 2H2 = ν 1H2 µ 1H2 /µ 2H2 = 131.9 THz × 0.5 m u /1 m u = 93.27 THz
√ √
ν 3H2 = ν 1H2 µ 1H2 /µ 3H2 = 131.9 THz × 0.5 m u /1.5 m u = 76.15 THz
P7E.5 (a) The wavenumber, ν̃ is given by ν̃ = ν/c, where c is the speed of light (see
The chemist’s toolkit 13). The frequency in Hz is related to the angular
frequency by ν = ω/2π, hence ν̃ = ω/2πc .
(b) The vibrational frequency of 1H 35Cl is 5.63×1014 s−1 , which gives a wave-
number of
ν̃ = (5.63 × 1014 s−1 )/[2π × (2.9979 × 1010 cm s−1 )] = 2.99 × 103 cm−1
Note that in order to express the wavenumber in cm−1 , the speed of light
is used in cm s−1 .
1/2
(c) The frequency ω is given by [7E.3–260], ω = (k f /µ) , and so the vibra-
1/2
tional frequency expressed as a wavenumber is ν̃ = (1/2πc) (k f /µ) .
This rearranges to give k f = µ(2πν̃c)2 .
(d) The effective mass of 12C 16O is µ = (12 × 16)/(12 + 16) = 6.85... m u .
Using the previous result, and converting the mass to kg, gives the force
constant
k f = [6.85... × (1.6605 × 10−27 kg)]

× [2π × (2170 cm−1 ) × (2.9979 × 1010 cm s−1 )]2 = 1902 N m−1
For an isotopic substitution the vibrational√frequencies are related by the

effective masses (Problem P7E.1) ν̃ 2 /ν̃ 1 = µ 1 /µ 2 . The effective mass of
C O is µ 14C 16O = (14 × 16)/(14 + 16) = 7.46... m u . This gives the
14 16
wavenumber for 14C 16O as

√
ν̃ 14C 16O = ν̃ 12C 16O µ 12C 16O /µ 14C 16O
√
= (2170 cm−1 ) × 6.85... m u /7.46 m u = 2080 cm−1
P7E.7 If the C atom is immobilized, only the O will be moving, and the force constant
will be the same as that in the CO molecule. The wavenumber of the CO
vibration is 2170 cm−1 , as given in Problem P7E.6, and as the force constant is
the√same this is linked to the wavenumber of the vibration when bound as ν̃ 2 =
ν̃ 1 µ 1 /µ 2 , where 1 refers to the free molecule and 2 that bound to the haem
group. The effective mass of 12C 16O is µ 1 = (12×16)/(12+16) = 6.85... m u , and
in the bound case the effective mass is simply the mass of the O atom, 16 m u ,
as this is the only atom which is moving. Hence
√
ν̃ 2 = (2170 cm−1 ) × (6.85... m u )/(16 m u ) = 1420 cm−1
Expressed as a frequency this is ν = 4.259 × 1013 Hz.
(a) The second derivative of the function ψ = e−gx is evaluated using the
2
P7E.9
chain rule
de−gx ⎞
2
d2 ψ d2 −gx 2 d −gx 2 ⎛ dx −gx 2
= (e ) = (−2gxe ) = −2g e + x
dx 2 dx 2 dx ⎝ dx dx ⎠
= −2g (e−gx + x × −2gxe−gx ) = 2g(2gx 2 − 1)e−gx

2 2 2
(b) The Hamiltonian for the harmonic oscillator is Ĥ = −(ħ 2 /2m)d2 /dx 2 +
k x 2 . To find the condition for e−gx to be an eigenfunction of Ĥ, first
2
1
2 f
allow the operator to act on the function
Ĥe−gx = [−(ħ 2 /2m)d2 /dx 2 + 12 k f x 2 ] e−gx

2 2
= −(ħ 2 /2m)2g(2gx 2 − 1)e−gx + 21 k f x 2 e−gx

2 2
(7.2)
Evidently e−gx is not an eigenfunction of Ĥ because when this operator

2
acts on e−gx the result is not a constant times that function. However, this
2
condition would be satisfied if the terms in x 2 were to cancel one another,

that is
−(ħ 2 /2m)2g(2gx 2 ) + 21 k f x 2 = 0
Solving this for g gives g = (mk f )1/2 /2ħ ; this is the value of g needed
for e−gx to be an eigenfunction of the hamiltonian. The resulting wave-
2
function is ψ = e−(mk f ) x /2ħ , which is in agreement with that quoted in

1/2 2
[7E.8b–261].
(c) Returning to eqn 7.2, if the terms in x 2 cancel the remaining terms are
−(ħ 2 /2m)2g(−1)e−gx
2
which is of the form of a constant times the original function. The con-
stant is the eigenvalue, which in this case is the energy
E = (ħ 2 /2m)2g = (ħ 2 /2m) × 2 × (mk f )1/2 /2ħ = 1

2
ħ(k f /m)1/2
The energy is indeed the same as that of the lowest level of the harmonic
oscillator as described in the text.
P7E.11 (a) The variable y is defined as y = x/α where α = (ħ 2 /mk f )1/4 , and the ki-
netic energy operator in terms of x is T̂ = −(ħ 2 /2m)d2 /dx 2 . Substituting
in y leads to
ħ 2 d2 ħ 2 d2
T̂ = − =− 2
2m d(α y) 2 2α m dy 2
ħ 2 mk f 1/2 d2 k f 1/2 d2 d2
=− ( 2 ) = − 1
ħ ( ) = − 1
ħω
2m ħ dy 2 2
m dy 2 2
dy 2
where for the last line the definition of the frequency ω is used, ω =
(k f /m)1/2 .
(b) The expectation value of the kinetic energy of the state with quantum
number υ is given by
∞ d2
H υ e−y /2
H υ e−y /2 dy
2 2
⟨T⟩υ = − 12 ħωN υ2 ∫ 2
−∞ dy
To evaluate the derivative, the product rule is employed,
d e−y /2
2
d2 −y 2 /2 d ′ −y 2 /2 d
(H ′ − yH υ )e−y /2
2
H υ e = (H υ e + H υ )=
dy 2 dy dy dy υ
The product rule is applied once more
d(H ′υ − yH υ ) −y 2 /2 d e−y /2
2
d2
H υ e−y /2 = + (H ′υ − yH υ )
2
2
e
dy dy dy
= (H ′′υ − yH ′υ − H υ )e−y /2
− y(H ′υ − yH υ )e−y /2
2 2
= [H ′′υ − 2yH ′υ + (y 2 − 1)H υ ]e−y /2

2
From Table 7E.1 on page 261 one of the properties of the Hermite polyno-
mials is H ′′υ − 2yH ′υ + 2υH υ = 0, so it follows that H ′′υ − 2yH ′υ = −2υH υ .
Using this in the last line above gives
d2
H υ e−y /2 = −(2υ + 1)H υ e−y /2 + y 2 H υ e−y /2
2 2 2
dy 2
(c) A further property of the Hermite polynomials is H υ+1 −2yH υ +2υH υ−1 =
0, which rearranges to yH υ = υH υ−1 + 21 H υ+1 . Using this, the term y 2 H υ
is rewritten
y 2 H υ = y(yH υ ) = y(υH υ−1 + 12 H υ+1 ) = yυH υ−1 + 12 yH υ+1
The same relationship is used to substitute for the two terms on the right,
but rewritten firstly by making the substitution υ → υ − 1 to give yH υ−1 =
(υ − 1)H υ−2 + 21 H υ , and secondly with with the substitution υ → υ + 1 to
give yH υ+1 = (υ + 1)H υ + 12 H υ+2 . With these substitutions
yυH υ−1 + 21 yH υ+1 = υ[(υ − 1)H υ−2 + 12 H υ ] + 12 [(υ + 1)H υ + 12 H υ+2 ]

= υ(υ − 1)H υ−2 + (υ + 12 )H υ + 14 H υ+2
Hence, the second derivative is given by
d2
H υ e−y /2 = [−(2υ + 1)H υ + y 2 H υ ] e−y /2
2 2
dy 2
= [−(2υ + 1)H υ + υ(υ − 1)H υ−2 + (υ + 12 )H υ + 14 H υ+2 ] e−y /2
2
= [υ(υ − 1)H υ−2 − (υ + 12 )H υ + 14 H υ+2 ] e−y /2

2
(d) The expectation value of the kinetic energy is therefore given by

∞
H υ e−y /2
2
⟨T⟩υ = − 12 ħωN υ2 ∫ [υ(υ − 1)H υ−2
−∞
− (υ + 12 )H υ + 14 H υ+2 ]e−y /2
2
dy
∞ ∞
H υ H υ−2 e−y dy + 14 ∫ H υ H υ+2 e−y dy
2 2
= − 12 ħωN υ2 [υ(υ − 1) ∫
−∞ −∞
∞
−y 2
− (υ + 1
)
2 ∫
Hυ Hυ e dy]
−∞
The first two integrals are zero due to orthogonality of the Hermite poly-
nomials, leaving
∞
H υ2 e−y dy
2
⟨T⟩υ = 12 (υ + 12 )ħωN υ2 ∫ (7.3)
−∞
The next step to to evaluate the normalization constant N υ which is found

from ∞
H υ2 e−y dy = 1
2
N υ2 ∫
−∞
A property of the Hermite polynomials is (Table 7E.1 on page 261)
∞
H υ2 e−y dy = π 1/2 2υ υ!
2
∫
−∞
If follows that N υ2 = 1/(π 1/2 2υ υ!). However, by the same property of

the Hermite polynomials the integral in eqn 7.3 is also equal to π 1/2 2υ υ!,
therefore this term cancels with N υ2 to leave
⟨T⟩υ = 21 (υ + 12 )ħω = 21 E υ
P7E.13 As is shown in Example 7E.3 on page 265, in terms of the dimensionless variable
y the classical turning points are at y tp = ±(2υ + 1)1/2 , where υ is the quantum
√
number of the state. For the first excited state υ = 1, and so y tp = ± 3.
The wavefunction is ψ 1 = N 1 ye−y /2
2
, which is normalized by solving
+∞ +∞
y 2 e−y dy = 2N 12 ∫ y 2 e−y dy = 1
2 2
N 12 ∫
−∞ 0
The integral is of the form of Integral G.3 with k = 1 and evaluates to 41 π 1/2 ,
thus N 12 = 2π−1/2 . The probability of finding the particle outside the range of
the turning points is then
∞ 4 ∞
P = 2 × 2π−1/2 ∫√ y 2 e−y dy = 2 −y
2 2
∫√ y e dy
3 π 1/2 3
Evaluating this integral numerically gives P = 0.112 .

∞
P7E.15 The integral to be evaluated is I = ∫−∞ ψ ∗υ′ xψ υ dx. This can be written in terms
of the dimensionless variable y, defined as x = α y where α = (ħ 2 /mk f )1/4 ,
∞ ∞
as I = ∫−∞ ψ ∗υ′ (α y)ψ υ d(α y) = α 2 ∫−∞ ψ ∗υ′ yψ υ dy. The wavefunctions are
ψ υ = N υ H υ e−y /2 , where N υ is a normalization constant, and H υ is a Hermite
2
polynomial of order υ. With these substitutions I is

∞
(N υ′ H υ′ e−y /2
)y(N υ H υ e−y /2
2 2
I = α 2∫ ) dy
−∞
∞
H υ′ (yH υ )e−y dy
2
= α 2 N υ N υ′ ∫
−∞
Using the properties of the Hermite polynomials given in Table 7E.1 on page
261, the term yH υ can be rewritten as yH υ = 12 H υ+1 +υH υ−1 , and so the integral
can be rewritten as
∞ ∞
H υ′ H υ+1 e−y dy + υ ∫ H υ′ H υ−1 e−y dy]
2 2
I = α 2 N υ N υ′ [ 12 ∫ (7.4)
−∞ −∞
∞
Due to orthogonality of the Hermite polynomials, the integral ∫−∞ H υ′ H υ e−y dy
2
is only non-zero when υ′ = υ. Hence, the integral I is only non-zero if υ′ = υ+1

or υ′ = υ − 1. Hence, the only transitions with intensities not equal to zero are
those with υ′ = υ ± 1 .
The normalization constant N υ is found from
∞
H υ2 e−y dy = 1
2
N υ2 ∫
−∞
∞
It is a property of the Hermite polynomials that ∫−∞ H υ2 e−y dy = π 1/2 2υ υ! so
2
it follows that N υ = (π 1/2 2υ υ!)−1/2 .

For υ′ = υ+1 only the first integral in eqn 7.4 is non-zero. From the properties of
∞
e−y dy = [π 1/2 2υ+1 (υ + 1)!]1/2 .
2
Hermite polynomials it follows that ∫−∞ H υ+1 2
Using this, and the normalization factors already quoted, gives

∞
e−y dy
2
I = α 2 N υ N υ+1 21 ∫ H υ+1
2
−∞
1 1 1/2 υ+1 α2
= 12 α 2 1/2 υ 1/2 1/2 υ+1 (π 2 (υ + 1)!) = (υ + 1)1/2
(π 2 υ!) [π 2 (υ + 1)!]1/2 21/2
In the case that υ′ = υ − 1 only the second integral in eqn 7.4 is non-zero. Using
∞
e−y dy = [π 1/2 2υ−1 (υ − 1)!]1/2 ,
2
the same procedure as before with ∫−∞ H υ−1 2
gives
∞
e−y dy
2
I = α 2 N υ N υ−1 υ ∫ H υ−1
2
−∞
1 1 1/2 υ−1 α2 √
= υα 2 1/2 υ 1/2 1/2 υ−1 [π 2 (υ − 1)!] = υ
[π 2 υ!] (π 2 (υ − 1)!)1/2 21/2
P7E.17 (a) The harmonic oscillator potential energy is V (x) = 21 k f x 2 which is sym-
metric about x = 0. It therefore follows that the probability density must
also be symmetric about x = 0, for all wavefunctions. Hence the proba-
bility of finding the particle at x = x ′ is the same as finding it at x = −x ′
and so, for all υ, ⟨x⟩υ = 0 .
(b) As the potential is symmetric about x = 0 the probability of finding the
particle moving to the right with a particular positive momentum is equal
to that of finding the particle moving to the left with same momentum,
but with opposite sign. It therefore follows that ⟨p x ⟩υ = 0 .
(c) From the virial theorem, ⟨E k ⟩υ = 12 E υ = 21 (υ + 12 )ħω. As the kinetic
energy is E k = p2x /2m, where m is the mass of the particle, it follows that
⟨p2x ⟩υ = 2m⟨E k ⟩υ = 2m × 12 (υ + 12 )ħω = m(υ + 12 )ħω .
(d) The value of ⟨x 2 ⟩υ is given in [7E.12b–264] as ⟨x 2 ⟩υ = (υ + 21 )ħ/(mk f )1/2 .
∆x υ = [⟨x 2 ⟩υ − ⟨x⟩2υ ]1/2 = [(υ + 12 )ħ/(mk f )1/2 − 02 ]1/2

= (υ + 12 )1/2 (ħ 2 /mk f )1/4
∆p υ = [⟨p2x ⟩υ − ⟨p x ⟩2υ ]1/2 = [m(υ + 12 )ħω − 02 ]1/2
= (υ + 12 )1/2 (ħmω)1/2
(e) The product of the uncertainties is
∆x υ ∆p υ = (υ + 12 )(ħ 2 /mk f )1/4 × (ħmω)1/2 = (υ + 12 )(ħ 4 mω 2 /k f )1/4
This can be simplified by noting that ω = (k f /m)1/2 ; using this in the

expression above gives ∆x υ ∆p υ = (υ + 12 )ħ . This function increases
linearly with υ and its minimum value is when υ = 0: ∆x 0 ∆p 0 = ħ/2.
This value satisfies the uncertainty principle, ∆x∆p ≥ ħ/2, as do wave-
functions with higher values of υ.
(f) The minimum product of the uncertainties is for υ = 0 .
7F Rotational motion
D7F.1 The vector model of angular momentum is described in Section 7F.2(c) on page
274. The model captures the key idea that the magnitude and the z-component
of the angular momentum can be known simultaneously and precisely. How-
ever, if this is so there is no knowledge of the other components of the angular
momentum other than that the sum of their squares must be equal to a partic-
ular value.
D7F.3 Rotational motion on a ring and on a sphere share the following features: (a)
quantization arising as a result of the need to satisfy a cyclic boundary con-
dition; (b) energy levels which go inversely with the moment of inertia; (c)
the lack of zero-point energy; (d) degeneracy; (e) quantization of the angular
momentum about one axis.
E7F.1(a) The diagrams shown in Fig. 7.9 are drawn by forming a vector of length [l(l +
1)]1/2 and√with a projection m l on the z-axis. For l = 1, m l = +1 the vector is
of length √2 and has projection +1 on the z-axis; for l = 2, m l = 0 the vector is
of length 6 and has projection 0 on the z-axis. Each vector may lie anywhere
on a cone described by rotating the vector about the z-axis.
z
+1 l = 1, m l = 1
0
l = 2, m l = 0
Figure 7.9
E7F.2(a) The spherical harmonic is
Y3,0 = 14 (7/π)1/2 (5 cos3 θ − 3 cos θ) = 1

4
× (7/π)1/2 cos θ(5 cos2 θ − 3)
The nodes occur when Y3,0 passes through zero, which √ happens when either
cos θ = 0 or 5 cos2 θ − 3 = 0, that is when cos θ = ± 3/5; recall that θ is in the
range 0 to π. Hence, the nodes are θ = π/2, 0.684, 2.46 . There are 3 angular
nodes.
√
E7F.3(a) The real part of the spherical harmonic Y1,+1 is − 12 3/π sin θ cos ϕ. When ϕ
varies, an angular node therefore occurs when cos ϕ = 0, i.e. at ϕ = π/2, 3π/2 .
This plane corresponds to the plane x = 0, i.e. the yz plane.
√
The imaginary part of the same spherical harmonic is − 12 3/π sin θ sin ϕ. When
ϕ varies, an angular node therefore occurs when sin ϕ = 0, i.e. at ϕ = 0, π. This
plane corresponds to the plane y = 0, i.e. the xz plane.
E7F.4(a) The rotational energy depends only on the quantum number l [7F.10–273], but
there are distinct states for every allowed value of m l , which can range from −l
to +l in integer steps. There are 2l + 1 such states, as there are l of these with
m l > 0, l of these with m l < 0 and m l = 0. Hence l = 3 has a degeneracy of 7 .
E7F.5(a) The magnitude of the angular momentum associated with a wavefunction with
angular momentum quantum number l is given by [7F.11–274], magnitude =
ħ[l(l + 1)]1/2 . Hence for l = 1 the magnitude is ħ[1(1 + 1)]1/2 = 21/2 ħ .
The projection of the angular momentum onto the z-axis is given by [7F.6–270],
ħm l , where m l is a quantum number that takes values between −l and +l in
integer steps, m l = −l , −l + 1, . . . + l. Hence the possible projections onto the
z-axis are −ħ, 0, ħ .
E7F.6(a) The wavefunction of a particle on a ring, with quantum number m l is ψ m l =
eim l ϕ = cos(m l ϕ) + i sin(m l ϕ) in the range 0 ≤ ϕ ≤ 2π. The real and imaginary
parts of the wavefunction are therefore cos(m l ϕ) and sin(m l ϕ) respectively.
Nodes occur when the function passes through zero, which for trigonometric
functions are the same points at which the function is zero. Hence in the real
part, nodes occur when cos(m l ϕ) = 0, and so when m l ϕ = (2n + 1)π/2 for
integer n, which gives ϕ = (2n + 1)π/2m l . In the imaginary part, nodes occur
when sin(m l ϕ) = 0 and so when m l ϕ = nπ for an integer n, which gives ϕ =
nπ/m l .
(i) With m l = 0 the real part is a constant and has no nodes ; the imaginary
part is zero everywhere.
(ii) With m l = 3, nodes in the real part occur at π/6, π/2, 5π/6, 7π/6, 3π/2 ,
11π/6 . In the imaginary part nodes occur at 0, π/3, 2π/3, π, 4π/3, 5π/3 .
There are 6 nodes in each of the parts.
E7F.7(a) The normalization condition is ∫ ψ ∗m l ψ m l dτ = 1. In this case the integral is
over ϕ in the range 0 ≤ ϕ ≤ 2π, and the wavefunction is ψ m l = Neim l ϕ . Hence,
noting that the wavefunction is complex
2π 2π
N 2 ∫ ψ ∗m l ψ m l dτ = N 2 ∫ e−im l ϕ eim l ϕ dϕ = N 2 ∫ dϕ = 2πN 2
0 0
where e−iθ e+iθ = e0 = 1 is used. Setting this integral to 1 gives N = (2π)−1/2 .
The integral to evaluate is ∫0 ψ ∗m ′ ψ m l dϕ, with ψ m l = eim l ϕ . This wavefunction

2π
E7F.8(a)
l
′
is complex and so ψ ∗m ′ = e−im l ϕ . Note that both m l and m′l are integers and
l
therefore m l ± m′l is also an integer. The integral is then
2π ′ 2π ′
∫ e−im l ϕ eim l ϕ dϕ = ∫ ei(m l −m l )ϕ dϕ
0 0
′
= (i[m l − m′l ])−1 (e2πi(m l −m l ) − e0 )
= (i[m l − m′l ])−1 (1 − 1) = 0
where e2πin = 1 for any integer n is used. Hence, wavefunctions for a particle
on a ring with different quantum numbers are orthogonal.
E7F.9(a) The energy levels of a particle on a ring are [7F.4–269], E m l = m 2l ħ 2 /2I where I
is the momentum of inertia of the system, I = mr 2 , see The chemist’s toolkit 20
in Topic 7F on page 268. The minimum excitation is therefore ∆E = E 1 − E 0 =
(ħ 2 /2I)(12 − 02 ) = ħ 2 /2I Evaluating this gives
(1.0546 × 10−34 J s)2

∆E = = 3.32 × 10−22 J
2 × (1.6726 × 10−27 kg) × (100 × 10−12 m)2
E7F.10(a) The energy levels are [7F.10–273], E l = ħ 2 l(l + 1)/2I, where I is the moment
of inertia. The minimum energy to start it rotating is the minimum excitation
energy, the energy to take it from the motionless l = 0 to the rotating l = 1
state, ∆E = E 1 − E 0 = (ħ 2 /2I)(1(1 + 1) − 0(0 + 1)] = ħ 2 /I. Evaluating this gives
∆E = (1.0546 × 10−34 J s)2 /(5.27 × 10−47 kg m2 ) = 2.11 × 10−22 J
E7F.11(a) The energy levels are [7F.10–273], E l = ħ 2 l(l + 1)/2I, where I is the moment
of inertia. So that the excitation energy is ∆E = E 2 − E 1 = (ħ 2 /2I)[2(2 + 1) −
1(1 + 1)] = 2ħ 2 /I. Evaluating this gives
∆E = 2(1.0546 × 10−34 J s)2 /(5.27 × 10−47 kg m2 ) = 4.22 × 10−22 J
E7F.12(a) The energy levels are [7F.10–273], E l = ħ 2 l(l + 1)/2I, where I is the
√moment
of inertia. The corresponding angular momentum is ⟨l 2 ⟩1/2 = ħ l(l + 1).
Hence, the minimum energy allowed is 0, through this corresponds to zero
angular momentum, and so rest and not motion. So the minimum energy of
rotation occurs for the state that has l = 1. The angular momentum of that state
√ √ √
is ⟨l 2 ⟩1 = ħ 1(1 + 1 = 2ħ = 2 × (1.0546 × 10−34 J s) = 1.49 × 10−34 J s .
1/2
E7F.13(a) The diagrams shown in Fig. 7.10 are drawn by forming a vector of length [l(l +
1)]
√
1/2
and with a projection m l on the z-axis. For l = 1 the vector is of length
√ 2 and has projection −1, 0, +1 on the z-axis. For l = 2 the vector is of length
6 and has projection −2, . . . + 2 in integer steps on the z-axis. Each vector
may lie anywhere on a cone described by rotating the vector about the z-axis.
z
+2 +2
+1 +1
z
+1 +1 0 0
0 0 −1 −1
−1 −1 −2 −2
Figure 7.10
E7F.14(a) The angle in question is that between the z-axis and the vector representing the
angular momentum. The projection
√ of the vector onto the z-axis is m l ħ, and
the length of the vector is ħ l(l + 1). Therefore the angle θ that the vector
√
makes to the z-axis is given by cos θ = m l / l(l + 1).
√ √
When m l = l, cos θ = l/ l(l + 1), which for l = 1 gives cos θ = 1/ 2, and so
√
θ = π/4 , and for l = 5 gives cos θ = 5/ 30, and so θ = 0.420 .
P7F.1 The angular momentum states have quantum numbers m l = 0, ±1, ±2.... The
energy levels for a particle on a ring are given by [7F.4–269], E m l = m 2l ħ 2 /2I,
and have angular momentum [7F.6–270], J z = m l ħ. The moment of inertia for
an electron on this ring is I = mr 2 = (9.1094 × 10−31 kg) × (440 × 10−12 m)2 =
1.76... × 10−49 kg m2
(a) If there are 22 electrons in the system the highest occupied state will be
the degenerate levels m l = ±5. These states have an energy of E±5 =
(±5)2 (1.0546 × 10−34 J s)2 /2(1.76... × 10−49 kg m2 ) = 7.88 × 10−19 J , and
angular momenta of
J z = ±5ħ = ±5 × (1.0546 × 10−34 J s) = 5.273 × 10−34 J s
(b) The lowest unoccupied levels are those with m l = ±6, and so the dif-
ference in energy between the highest occupied and lowest unoccupied
levels is
∆E = E±6 − E±5 = (ħ 2 /2I)(62 − 52 )

= 11 × 1.0546 × 10−34 J s)2 /2(1.76... × 10−49 kg m2 ) = 3.46... × 10−19 J
The Bohr frequency condition, [7A.9–227] states that the frequency of

radiation that will excite such a transition is
3.46... × 10−19 J
ν = ∆E/h = = 5.23 × 1014 Hz
6.6261 × 10−34 J s
P7F.3 In Cartesian coordinates, the equation defining the ellipse is x 2 /a 2 + y 2 /b 2 = 1.

An appropriate change of variables can transform this ellipse into a circle. That
change of variable is most conveniently described in terms of new Cartesian
coordinates (X, Y) where X = x and Y = ay/b. In these coordinates, the
equation for the ellipse can be rewritten as X 2 + Y 2 = a 2 , which is the equation
of a circle radius a centered on the origin. The text found the eigenfunctions
and eigenvalues for a particle on a circular ring by transforming from Cartesian
coordinates to plane polar coordinates. A similar transformation can be made
by defining coordinates (R, Φ) such that X = R cos Φ, Y = R sin Φ.
In these coordinates, this is simply a particle on a ring, as described in the text,
for which the Schrödinger equation is separable .
P7F.5 The Schrödinger equation for a particle on a sphere is −(ħ 2 /2I)Λ̂ 2 ψ = Eψ,
where
1 ∂2 1 ∂ ∂
Λ̂ 2 = + sin θ
sin2 θ ∂ϕ 2 sin θ ∂θ ∂θ
(a) Y0,0 = (1/2)π−1/2 , which is a constant, and so its derivatives with respect
to all θ and ϕ are zero, so Λ̂ 2 Y0,0 = 0 implying that E = 0 and lˆz Y0,0 = 0,
so that J z = 0 .
(b) Y2,−1 = N sin θ cos θe−iϕ , thus
∂Y2,−1 /∂θ = Ne−iϕ (cos2 θ−sin2 θ) ∂Y2,−1 /∂ϕ = −iN sin θ cos θe−iϕ = −iY2,−1
In addition, ∂ 2 Y2,−1 /∂ϕ 2 = N sin θ cos θe−iϕ
1 ∂ 2 Y2,−1 1 ∂ ∂Y2,−1
Λ̂ 2 Y2,−1 = + sin θ
sin θ ∂ϕ
2 2 sin θ ∂θ ∂θ
N cos θe−iϕ Ne−iϕ ∂
= + sin θ(cos2 θ − sin2 θ)
sin θ sin θ ∂θ
The derivative is evaluated using the product rule
N cos θe−iϕ
=
sin θ
Ne−iϕ
+ [sin θ(−4 cos θ sin θ) + cos θ(cos2 θ − sin2 θ)]
sin θ
as cos3 θ = cos θ cos2 θ = cos θ(1 − sin2 θ)
N cos θe−iϕ cos θ

= + Ne−iϕ (−6 sin θ cos θ + )
sin θ sin θ
= −6Ne−iϕ sin θ cos θ
This has an eigenvalue of −6, giving an energy eigenvalue of 6ħ 2 /2I . For

angular momentum, lˆz Y2,−1 = (ħ/i)×−iY2,−1 = ħY2,−1 , giving an angular
momentum eigenvalue of J z = −ħ .
(c) Y3,+3 = Ne3iϕ sin3 θ, thus
∂Y3,+3 /∂θ = 3Ne3iϕ sin2 θ cos θ ∂Y3,+3 /∂ϕ = 3iNe3iϕ sin3 θ = 3iY3,+3
In addition, ∂ 2 Y3,+3 /∂ϕ 2 = −9Ne3iϕ sin3 θ. Hence,
1 ∂ 2 Y3,+3 1 ∂ ∂Y3,+3
Λ̂ 2 Y3,+3 = + sin θ
sin2 θ ∂ϕ 2 sin θ ∂θ ∂θ
−9Ne sin θ
3iϕ 3
1 ∂
= + 3Ne3iϕ sin3 θ cos θ
sin2 θ sin θ ∂ϕ
The derivative is evaluated using the product rule
3Ne3iϕ
= −9Ne3iϕ sin θ + (3 sin2 θ cos2 θ − sin4 θ)
sin θ
= −9Ne3iϕ sin θ + 3Ne3iϕ (3 sin θ cos2 θ − sin3 θ)
= −12Ne3iϕ sin3 θ = −12Y3,+3
This has an eigenvalue of −12, giving an energy eigenvalue of 12ħ 2 /2I .

For angular momentum, lˆz Y2,−1 = (ħ/i) × 3iY2,−1 = 3ħY2,−1 , giving an
angular momentum eigenvalue of J z = 3ħ .
The energies are given by [7F.10–273], E l = ħ 2 l(l + 1)/2I, and therefore E 0 =
0, E 2 = 2(3)ħ 2 /2I = 6ħ 2 /2I and E 3 = 4(3)ħ 2 /2I = 12ħ 2 /2I, all of which
are consistent with the calculated eigenvalues.
√ √
P7F.7 The function Y1,+1 = − 12 3/2π sin θeiϕ and Y1,0 = 12 3/π cos θ. The integral
to evaluate is
π 2π
∗
∫ ∫ Y1,0 Y1,+1 sin θ dθ dϕ
θ=0 ϕ=0
Y1,0 is real and so the integrand is

√ √ √
− 12 3/2π sin θeiϕ × 12 3/π cos θ × sin θ = −(3/4π 2) sin2 θ cos θeiϕ
This gives the integral as
π 2π 3 π 2π
∗
I=∫ ∫ Y1,0 Y1,+1 sin θ dθ dϕ = − √ ∫ ∫ sin2 θ cos θeiϕ dθ dϕ
θ=0 ϕ=0 4π 2 θ=0 ϕ=0
The integrand is the product of separate functions of θ and ϕ, and so the integral
can be separated
3 2π π
I=− √ ∫ eiϕ dϕ ∫ sin2 θ cos θ dθ
4π 2 0 0
Evaluating the first integral gives

2π 2π
∫ eiϕ dϕ = (1/i) eiϕ ∣0 = (1/i)[e2πi − e0 ] = (1/i)(1 − 1) = 0
0
Hence
π 2π
∗
∫ ∫ Y1,0 Y1,+1 sin θ dθ dϕ = 0
θ=0 ϕ=0
so the two functions are orthogonal .

P7F.9 (a) Multiplying out the brackets, noting that the derivatives in the left brack-
ets act on the whole term in the right brackets, e.g. x∂ f /∂x
∂ ∂ ∂f ∂f
lˆx lˆy f = −ħ 2 (y − z ) (z −x )
∂z ∂y ∂x ∂z
∂ ∂f ∂ ∂f ∂ ∂f ∂ ∂f
= −ħ 2 [y (z ) − y (x ) − z (z ) + z (x )]
∂z ∂x ∂z ∂z ∂y ∂x ∂y ∂z
To evaluate the first term in this, the product rule is used
∂z ∂ f ∂2 f ∂2 f ∂2 f ∂2 f
= −ħ 2 [y + yz − x y 2 − z2 + zx ]
∂z ∂x ∂z∂x ∂z ∂y∂x ∂y∂z
∂f ∂2 f ∂2 f ∂2 f ∂2 f
= −ħ 2 [y + yz − x y 2 − z2 + zx ]
∂x ∂z∂x ∂z ∂y∂x ∂y∂z
(b) Similarly,
∂ ∂ ∂f ∂f
lˆy lˆx f = −ħ 2 (z − x ) (y −z )
∂x ∂z ∂z ∂y
∂ ∂f ∂ ∂f ∂ ∂f ∂ ∂f
= −ħ 2 [z (y ) − z (z ) − x (y ) + x (z )]
∂x ∂z ∂x ∂y ∂z ∂z ∂z ∂y
To evaluate the final term in this, the product rule must be used
∂2 f ∂2 f ∂2 f ∂z ∂ f ∂2 f
= −ħ 2 [yz − z2 − xy 2 + x + xz ]
∂x∂z ∂x∂y ∂z ∂z ∂y ∂z∂y
∂2 f ∂2 f ∂2 f ∂f ∂2 f
= −ħ 2 [yz − z2 − xy 2 + x + xz ]
∂x∂z ∂x∂y ∂z ∂y ∂z∂y
(c) Due to the symmetry of mixed partial derivatives, the only terms that are
not repeated in both of these terms are the first derivatives. Hence,
∂f ∂f ħ ∂f ∂f
lˆx lˆy f − lˆy lˆx f = ħ 2 (x − y ) = iħ × (x − y ) = iħ lˆz f
∂y ∂x i ∂y ∂x
where the definition of the lˆz operator given in [7F.13–275] is used. It
follow that [ lˆx , lˆy ] = lˆz .
(d) Applying cyclic permutation to lˆx gives (ħ/i)(z∂/∂x − x∂/∂z) = lˆy . Like-
wise lˆy gives (ħ/i)(x∂/∂y − y∂/∂x) = lˆz , and lˆz gives (ħ/i)(y∂/∂z −
z∂/∂y) = lˆx .
(e) Applying this permutation to the expression [ lˆx , lˆy ] = iħ lˆz gives [ lˆy , lˆz ] =
iħ lˆx . Permuting this expression gives [ lˆz , lˆx ] = iħ lˆy , and permuting that
expression gives [ lˆx , lˆy ] = iħ lˆz .
P7F.11 The Cartesian coordinates expressed in terms of the spherical polar coordinates
are x = r sin θ cos ϕ, y = r sin θ sin ϕ, z = r cos θ, see The chemist’s toolkit 21 in
Topic 7F on page 272. The chain rule is therefore used to express ∂/∂ϕ in terms
of derivatives of x, y and z.
∂ ∂x ∂ ∂y ∂ ∂z ∂
= + +
∂ϕ ∂ϕ ∂x ∂ϕ ∂y ∂ϕ ∂z
Evaluating the derivatives gives
∂ ∂ ∂ ∂ ∂
= −r sin θ sin ϕ + r sin θ cos ϕ + 0 = −y +x
∂x ∂y ∂z ∂x ∂y
where it has been noted that the factors multiplying the derivatives are Carte-
sian coordinates. Hence, lˆz = (ħ/i)(x∂/∂y − y∂/∂x) = (ħ/i)∂/∂ϕ

I7.1 (a) The first step is to compute the total energy of the system of N A particles,
which is identified as the internal energy U. The energy levels for a parti-
cle in a cubic box of side L are given by [7D.13b–253], E n = h 2 n 2 /8mL 2 ,
where n 2 = n 12 + n 22 + n 32 . If there are N A particles, all occupying the
level corresponding to a particular value of n, the internal energy of the
system is U = N A E n = N A h 2 n 2 /8mL 2 . Using V = L 3 the length is
written in terms of the volume as L = V 1/3 , hence L 2 = V 2/3 and therefore
U = N A h 2 n 2 /8mV 2/3 .
If the expansion is adiabatic (that is, not heat flows into or out of the
system) then from the First Law, dU = dq + dw, it follows that dU = dw.
The work done on expansion is therefore computed by finding how U
changes with volume, specifically by finding ∂U/∂V .
∂U ∂ NA h2 n2 −2 N A h 2 n 2 NA h2 n2
= ( ) = × = −
∂V ∂V 8mV 2/3 adia 3 8mV 5/3 12mV 5/3
The change in internal energy on expansion through dV will therefore be

A
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
∂U NA h2 n2
dU = ( ) dV = − dV (7.5)
∂V adia 12mV 5/3
The work is equal to this change in internal energy. For a finite change
the expression is integrated with respect to V between limits V1 and V2 ,
with ∆V = V2 − V1 .
(b) It is evident from eqn 7.5 that dU, and hence the work, goes as n 2 .
(c) The work of expansion against an external pressure p ex is given by [2A.5a–
37], dw = −p ex dV . In eqn 7.5 the term A which multiplies dV refers to the
sample itself, and so must presumably in some way reflect the pressure
of the sample, not the external pressure. However, if the expansion is
reversible, the external pressure is equal to the internal pressure and the
term A can then be identified as the pressure. Therefore, if it is assumed
that the expansion is both adiabatic and reversible
NA h2 n2
p=
12mV 5/3
The expression can be rewritten in terms of the average energy of each
particle which, because they all occupy the same level, is simply E av =
n 2 h 2 /8mL 2 = n 2 h 2 /8mV 2/3 , hence
E av
³¹¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹µ
N A h 2 n 2 8N A n 2 h 2 2N A E av
p= 5/3
= 2/3
=
12mV 12V 8mV 3V
This expression is reminiscent of the form of the pressure derived using
the kinetic theory of gases (Topic 1B): pV = 31 nMυ 2rms , where n is the
amount in moles, M is the molar mass, and υ rms is the root-mean-square

speed. Because M = mN A , where m is the mass of a molecule, the ex-
pression can be rewritten
Ek
³¹¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ µ
pV = 1
3
nmN A υ 2rms = n 23 N A 21 mυ 2rms
The term 12 mυ 2rms is identified as the average kinetic energy of one molecule
and, because in the kinetic theory the only energy a molecule possesses is
kinetic, E k can further be identified as the average energy, E av . Thus, for
one mole (n = 1)
2N A E av
pV = 32 N A E av hence p=
3V
The two expressions for the pressure are therefore directly comparable
within the restrictions imposed.
(d) For an isothermal expansion heat would have to enter the system in order
to maintain its temperature, and this would involve promoting particles
to higher energy levels. As the volume increases the energy levels move
closer together, so the promotion of particles to higher levels needs to
offset this effect as well.
I7.3 (a) In Problem P7D.6 and Problem P7D.7 it is shown that for a particle in a
box in a state with quantum number n
∆x = L(1/12 − 1/2n 2 π 2 )1/2 and ∆p x = nh/2L
hence
∆x∆p x = L(1/12 − 1/2n 2 π 2 )1/2 × nh/2L = (nh/2)(1/12 − 1/2n 2 π 2 )1/2
For n = 1
∆x∆p x = L(1/12 − 1/2π 2 )1/2 × h/2L = (h/2)(1/12 − 1/2π 2 )1/2 ≈ 0.57ħ
and for n = 2 ∆x∆p x ≈ 1.7ħ. The Heisenberg uncertainty principle

is satisfied in both cases, and it is evident that ∆x∆p x is an increasing
function of n. The principle is therefore satisfied for all n > 1.
(b) In Problem P7E.17 it is shown that for a harmonic oscillator in a state with
quantum number υ
∆x υ ∆p υ = (υ + 21 )ħ
Therefore, for the ground state with υ = 0, ∆x∆p = 12 ħ: the Heisenberg
uncertainty principle is satisfied with the smallest possible uncertainty. It
follow that for υ > 0 the principle is also satisfied because ∆x υ ∆p υ is an
increasing function of υ.
I7.5 Macroscopic synthesis and material development always contains elements of

randomness at the molecular level. Crystal structures are never perfect. A
product of organic synthesis is never absolutely free of impurities, although

impurities may be at a level that is lower than measurement techniques make
possible. Alloys are grainy and slightly non-homogeneous within any particu-
lar grain. Furthermore, the random distribution of atomic/molecular positions
and orientations within, and between, macroscopic objects causes the conver-
sion of energy to non-useful heat during manufacturing processes. Production
efficiencies are difficult to improve. Nanometre technology on the 1 nm to
100 nm scale may resolve many of these problems. Self-organization and pro-
duction processes by nanoparticles and nanomachines may be able to exclude
impurities and greatly improve homogeneity by effective examination and se-
lection of each atom/molecule during nanosynthesis and nanoproduction pro-
cesses. Higher efficiencies of energy usage may be achievable as nanomachines
produce idealized materials at the smaller sizes and pass their products to larger
nanomachines for production of larger scale materials.
The directed, non-random, use of atoms and molecules by nanotechniques
holds the promise for the production of smaller transistors and wires for the
electronics and computer industries. Unusual material strengths, optical prop-
erties, magnetic properties, and catalytic properties may be achievable. Higher
efficiencies of photo-electronic conversion would have a great impact.
8 Atomic structure and
spectra
8A Hydrogenic Atoms
D8A.1 (i) A boundary surface for a hydrogenic orbital is drawn so as to contain
most (say 90%) of the probability density of an electron in that orbital. Its
shape varies from orbital to orbital because the electron density distribu-
tion is different for different orbitals.
(ii) The radial distribution function P(r) gives the probability density that the
electron will be found at a distance r from the nucleus. It is defined such
that P(r) dr is the probability of finding the electron in a shell of radius r
and thickness dr. Because the radial distribution function gives the total
density, summed over all angles, it has no angular dependence and, as a
result, perhaps gives a clearer indication of how the electron density varies
with distance from the nucleus.
D8A.3 (i) The principal quantum number n determines the energy of a hydrogenic
atomic orbital through [8A.8–286].
(ii) The azimuthal quantum number l determines the magnitude of the or-
bital angular momentum, given by [l(l + 1)]1/2 ħ.
(iii) The magnetic quantum number m l determines the z-component of the
orbital angular momentum, given by m l ħ.
(iv) The spin quantum number s determines the magnitude of the spin angu-
lar momentum, given by [s(s + 1)]1/2 ħ; for hydrogenic atomic orbitals s
can only be 21 .
(v) The quantum number m s determines the z-component of the spin angu-
lar momentum, given by m s ħ; for hydrogenic atomic orbitals m s can only
be ± 21 .
E8A.1(a) The radial wavefunction of a 3s orbital is given in Table 8A.1 on page 286 as
R 3,0 = N(6 − 6ρ + ρ 2 )e−ρ/2 , where ρ = 2Zr/3a 0 . Radial nodes occur when the
wavefunction passes through 0, which is when 6 − 6ρ + ρ 2 = 0. The roots of this
292 8 ATOMIC STRUCTURE AND SPECTRA
√
quadratic equation are at ρ = 3 ± 3 and hence the nodes are at
√
r = (3 ± 3)(3a 0 /2Z)
The wavefunction goes to zero as ρ → ∞, but this does not count as a node as
the wavefunction does not pass through zero.
E8A.2(a) Angular nodes occur when cos θ sin θ cos ϕ = 0, which occurs when any of
cos θ, sin θ, or cos ϕ is equal to zero; recall that the range of θ is 0 → π and of
ϕ is 0 → 2π.
Although the function is zero for θ = 0 this does not describe a plane, and so
is discounted. The function is zero for θ = π/2 with any value of ϕ: this is the
x y plane. The function is also zero for ϕ = π/2 with any value of θ: this is the
yz plane. There are two nodal planes, as expected for a d orbital.
E8A.3(a) The radial distribution function is defined in [8A.17b–292], P(r) = r 2 R(r)2 .

For the 2s orbital R(r) is given in Table 8A.1 on page 286 as R 2,0 = N(2 −
ρ)e−ρ/2 where ρ = 2Zr/na 0 , which for n = 2 is ρ = Zr/a 0 . With the sub-
stitution r 2 = ρ 2 (a 0 /Z)2 , the radial distribution function is therefore P(ρ) =
N 2 (a 0 /Z)2 ρ 2 (2 − ρ)2 e−ρ .
Mathematical software is used √ to find the values of ρ for which dP(ρ)/dρ = 0,
giving the results ρ = 0, 2, 3 ± 5. The simplest way to identify which of these
is a maximum is to plot√ P(ρ) against ρ, from which it is evident that ρ = 2 is a
minimum√ and ρ = 3± 5 are both maxima, with the principal maximum being
at ρ = 3 + 5. The maximum in the radial distribution function is therefore at
√
r = (3 + 5)(a 0 /Z) .
E8A.4(a) The radius at which the electron is most likely to be found is that at which the
radial distribution function is a maximum. The radial distribution function is
defined in [8A.17b–292], P(r) = r 2 R(r)2 . For the 2p orbital R(r) is given in
Table 8A.1 on page 286 as R 2,1 = N ρe−ρ/2 where ρ = 2Zr/na 0 , which for n = 2
is ρ = Zr/a 0 . With the substitution r 2 = ρ 2 (a 0 /Z)2 , the radial distribution
function is therefore P(ρ) = N 2 (a 0 /Z)2 ρ 4 e−ρ .
To find the maximum in this function the derivative is set to zero; the multi-
plying constants can be discarded for the purposes of this calculation
d 4 −ρ
ρ e = 4ρ 3 e−ρ − ρ 4 e−ρ
dr
Setting this derivative to zero gives the solutions ρ = 0 and ρ = 4. P(ρ) is zero
for ρ = 0 and as ρ → ∞, therefore ρ = 4 must be a maximum. This occurs at
r = 4a 0 /Z .
E8A.5(a) The M shell has n = 3. The possible values of l (subshells) are 0, corresponding
to the s orbital, l = 1 corresponding to the p orbitals, and l = 2 corresponding
to the d orbital; there are therefore 3 subshells . As there is one s orbital, 3 p
orbitals and 5 d orbitals, there are 9 orbitals in total.
E8A.6(a) The magnitude

√ of the orbital angular momentum of an orbital with quantum
number l is l(l + 1)ħ. The total number of nodes for an orbital with quantum
number n is n − 1, l of these are angular and so the number of radial nodes is
n − l − 1.
orbital n l ang. mom. angular nodes radial nodes

1s 1 0 0 0 0
3s 3 0 0 0 2
√
3d 3 2 6ħ 2 0
E8A.7(a) All the 2p orbitals have the same value of n and l, and hence have the same
radial function, which is given in Table 8A.1 on page 286 as R 2,1 = N ρe−ρ/2
where ρ = 2Zr/na 0 , which for n = 2 is ρ = Zr/a 0 . Radial nodes occur when
the wavefunction passes through zero. The function goes to zero at ρ = 0 and as
ρ → ∞, but it does not pass through zero at these points so they are not nodes.
The number of radial nodes is therefore 0 .
E8A.8(a) The energy of the level of the H atom with quantum number n is given by
[8A.8–286], E n = −hc R̃ H /n 2 . As described in Section 8A.2(d) on page 289,
the degeneracy of a state with quantum number n is n 2 .
The state with E = −hc R̃ H has n = 1 and degeneracy (1)2 = 1 ; that with
E = −hc R̃ H /9 has n = 3 and degeneracy (3)2 = 9 ; and that with E = −hc R̃ H /25
has n = 5 and degeneracy (5)2 = 25 .
E8A.9(a) The task is to find the value of N such that the integral ∫ ψ ∗ ψ dτ = 1, where
ψ = Ne−r/a 0 . The integration is over the range r = 0 to ∞, θ = 0 to π, and
ϕ = 0 to 2π; the volume element is r 2 sin θ dr dθ dϕ. The required integral is
therefore
∞ π 2π
N2 ∫ ∫ ∫ r 2 e−2r/a 0 sin θ dr dθ dϕ
0 0 0
The integrand is a product of functions of each of the variables, and so the

integral separates into three
∞ π 2π
N2 ∫ r 2 e−2r/a 0 dr ∫ sin θ dθ ∫ dϕ
0 0 0
π 2π
= N [2!/(2/a 0 )3 ] × (− cos θ)∣0 × ϕ∣0
2
= N 2 [2!/(2/a 0 )3 ] × 2 × 2π = N 2 a 03 π
The integral over r is evaluated using Integral E.3 with n = 2 and k = 2/a 0 .
Setting the full integral equal to 1 gives N = (a 03 π)−1/2 .
E8A.10(a) The wavefunction is given by [8A.12–287], ψ n,l ,m l = Yl ,m l (θ, ϕ)R n,l (r); for the
state with n = 2, l = 0, m l = 0 this is
ψ 2,0,0 = Y0,0 (θ, ϕ)R 1,0 (r) = (4π)−1/2 (Z/2a 0 )3/2 (2 − ρ)e−ρ/2
where the radial wavefunction is taken from Table 8A.1 on page 286, the angular
wavefunction (the spherical harmonic) is taken from Table 7F.1 on page 272,
and ρ = 2Zr/na 0 . The probability density is therefore
P2,0,0 = ∣ψ 2,0,0 ∣2 = (4π)−1 (Z/2a 0 )3 (2 − ρ)2 e−ρ
The probability density at the nucleus, ρ = 0, is then (1/4π)Z 3 /(8a 03 )(2 −

0)2 e0 = Z 3 /(8πa 03 ) .
E8A.11(a) The radial wavefunction of a 2s orbital is taken from Table 8A.1 on page 286,
R 2,0 (r) = N(2−ρ)e−ρ/2 , where ρ = 2Zr/na 0 ; for n = 2, ρ = Zr/a 0 . The extrema
are located by finding the values of ρ for which dR 2,0 /dρ = 0; the product rule
is required
dR 2,0 d(2 − ρ) −ρ/2 de−ρ/2

=N e + N(2 − ρ)
dρ dρ dρ
= N(−1)e−ρ/2 + N(2 − ρ)(− 12 e−ρ/2 ) = N(ρ/2 − 2)e−ρ/2
The derivative is zero when ρ = 4, which corresponds to r = 4a 0 /Z . The wave-

function is positive at ρ = 0, negative at ρ = 4, and asymptotically approaches
zero as ρ → ∞; ρ = 4 must therefore correspond to a minimum.
E8A.12(a) Assuming that the electron is in the ground state, the wavefunction is ψ =
Ne−r/a 0 , and so the probability density is P(r) = ψ 2 = N 2 e−2r/a 0 . P(r) is a
maximum at r = 0 and then simply falls off as r increases; it falls to 21 its initial
value when P(r ′ )/P(0) = 12
′
P(r ′ )/P(0) = e−2r /a 0 = 1
2
Hence r ′ = − 21 ln 12 a 0 = 0.347a 0 .
P8A.1 The 2p orbitals only differ in the axes along which they are directed. Therefore,
the distance from the origin to the position of maximum probability density
will be the same for each.
The radial function for the 2p orbitals is R 2,1 = N ρe−ρ/2 , where ρ = Zr/a 0 . The
probability density is the square of the radial function, R 2,1
2
= N 2 ρ 2 e−ρ , and the
2
maximum in this is found by setting dR /dρ = 0
2
dR 2,1
= 2N 2 ρe−ρ − N 2 ρ 2 e−ρ = 0
dρ
Turning points occur at ρ = 0, 2, and it is evident from a plot of R(ρ) that ρ = 2
is the maximum. This corresponds to r = 2a 0 /Z. For the 2pz orbital, for which
the angular part goes as cos θ, the maximum will be at θ = 0, which corresponds
to x = y = 0. The position of maximum probability density is therefore at
x = 0, y = 0, z = 2a 0 /Z . The corresponding positions for the other 2p orbitals
are found by permuting the x, y and z coordinates.
P8A.3 The energy levels of a hydrogenic atom with atomic number Z are given by
[8A.13–288], E n = −hcZ 2 R̃ N /n 2 , where the Rydberg constant for the atom is
given by [8A.14–288], R̃ N = (µ/m e )R̃∞ ; µ is the reduced mass of the atom.
For the D atom, with nuclear mass m D , the reduced mass is
mD me
µ=
mD + me
and therefore the Rydberg constant for D is
µ mD me mD
R̃ D = R̃∞ = R̃∞ = R̃∞
me m e (m D + m e ) mD + me
2.01355 × (1.660 539 × 10−27 kg)
=
[2.01355 × (1.660 539 × 10−27 kg)]+(9.109 382 × 10−31 kg)
× (109 737 cm−1 ) = 109 707 cm−1
where the constants have been used to sufficient precision to match the data.
The energy of the ground state is
E 1 = −hc R̃ D = −(6.626 070 × 10−34 J s) × (2.997 925 × 1010 cm s−1 )
× (109 707 cm−1 ) = −2.179 27 × 10−18 J
Expressed as a molar quantity this is −1312.39 kJ mol−1 .
P8A.5 (a) By analogy with [8A.21–294], the wavefunction for a 3px orbital is ψ 3p x =
R 3,1 (r) × [Y1,+1 (θ, ϕ) − Y1,−1 (θ, ϕ)](2)−1/2 . The required integral to
verify normalization is
∞ π 2π
2−1 ∣Y1,+1 (θ, ϕ) − Y1,−1 (θ, ϕ)∣ sin θ dθ dϕ
2
∫ R 3,1
2
r 2 dr ∫ ∫
0 0 0
Consider first the integral over r. From Table 8A.1 on page 286 R 3,1 (ρ) =
(486)−1/2 (Z/a 0 )3/2 (4 − ρ)ρe−ρ/2 , where ρ = 2Zr/na 0 which is this case
is 2Zr/3a 0 . It is convenient to calculate the integral over ρ, noting that
r = ρ(3a 0 /2Z) so that r 2 = ρ 2 (3a 0 /2Z)2 and dr = dρ (3a 0 /2Z). The
integral becomes
1 Z3 ∞
3 ∫
(4 − ρ)2 ρ 2 (3a 0 /2Z)2 ρ 2 e−ρ (3a 0 /2Z) dρ
486 a 0 0
1 33 ∞ 1 ∞
2 4 −ρ 6 −ρ
∫ (4 − ρ) ρ e dρ = ∫ (16ρ − 8ρ + ρ )e dρ
4 5
= 3
486 2 0 144 0
1
= [16(4!/15 ) − 8(5!/16 ) + (6!/17 )] = 1
144
where Integral E.3 is used, with k = 1 and the appropriate value of n. The
radial part of the function is therefore normalized.
The angular function is found using the explicit form of the spherical
harmonics listed in Table 7F.1 on page 272
1 3 1/2 1
Yx = 1/2
(Y1,+1 − Y1,−1 ) = − ( ) (sin θ eiϕ + sin θ e−iϕ )
2 8π 21/2
3 1/2 1 3 1/2
= −( ) × 2 sin θ cos ϕ = − ( ) sin θ cos ϕ
8π 21/2 4π
The normalization integral over the angles becomes

3 π 2π
( )∫ ∫ sin3 θ cos2 ϕ dθ dϕ
4π 0 0
Using Integral T.3 with a = π and k = 1 gives the integral over θ as 4/3.
2π
The term cos2 ϕ is written as 1 − sin2 ϕ. The integral ∫0 1 dϕ = 2π and,
2π
using Integral T.2 with a = 2π and k = 1, gives ∫0 sin2 ϕ dϕ = π. Hence
the integral over ϕ is 2π − π = π. The overall integral over the angles is
therefore (3/4π) × (4/3) × (π) = 1; the angular part is also normalized,
and as a result the complete wavefunction is normalized.
The next task is to show that ψ 3p x and ψ 3d x y are mutually orthogonal;
taking the hint from question, attention is focused on the angular parts,
because if these are orthogonal the overall wavefunctions will also be
orthogonal. Setting aside all of the normalization factors, which will not
be relevant to orthogonality, the angular part of the wavefunction for px
is (Y1,+1 − Y1,−1 ). In a similar way, it is expected that the angular parts of
a d orbital can be constructed from spherical harmonics with l = 2.
To find the combination that represents dx y recall that this function is of
the form x y f (r) and that in spherical polar coordinates x = r sin θ cos ϕ
and y = r sin θ sin ϕ, therefore
x y f (r) = (r sin θ cos ϕ)(r sin θ sin ϕ) f (r) = r 2 f (r) sin2 θ cos ϕ sin ϕ
= 21 r 2 f (r) sin2 θ sin 2ϕ
The spherical harmonics Y2,±2 have the form (again, omitting the normal-
ization factors) sin2 θ e±2iϕ . The angular part of dx y is therefore obtained
by the combination
Y2,+2 − Y2,−2 = sin2 θ [e2iϕ − e−2iϕ ]

= sin2 θ [cos 2ϕ + i sin 2ϕ − cos 2ϕ + i sin 2ϕ] = 2i sin2 θ sin 2ϕ
It therefore follows that, to within some numerical factors, the angular

part of dx y is given by Y2,+2 − Y2,−2 . The orthogonality of the angular
parts of dx y and px therefore involves the following integral
π 2π
∫ ∫ (Y2,+2 − Y2,−2 )∗ (Y1,+1 − Y1,−1 ) sin θ dθ dϕ (8.1)
0 0
Concentrating on the integral over ϕ, this will involve terms such as

2π 2π 2π
∗
∫ Y2,+2 Y1,+1 dϕ = ∫ sin2 θ e−2iϕ sin θ eiϕ dϕ = ∫ sin3 θ e−iϕ dϕ
0 0 0
This integral is zero because ∫0 eniϕ dϕ is zero for integer n. All of the
2π
terms in eqn 8.1 follow this pattern and therefore the overall integral is
zero; the orbitals are therefore orthogonal.
(b) The radial nodes for the 3s, 3p and 3d orbitals are found by examining
the radial wavefunctions, which are listed in Table 8A.1 on page 286, ex-
pressed as functions of ρ = 2Zr/3a 0 . These functions all go to zero as
ρ → ∞ and, in some cases they are also zero at ρ = 0; these do not

count as radial nodes as the wavefunction does not pass through zero
at these points. The nodes are located by finding the values of ρ at which
the polynomial part of the radial function is zero.
The positions of the nodes in the 3s√orbital are given by the solutions to
6 − 6ρ + ρ 2 = 0, which are at ρ = 3 ± 3. In terms of r these nodes occur at
√
r = (3a 0 /2Z)(3 ± 3) . The positions of the nodes in the 3p orbital are
given by the solutions to (4 − ρ)ρ = 0; there is just one node t ρ = 4 which
corresponds to r = 6a 0 /Z . For the 3d orbital the polynomial is simply
ρ 2 , which does not lead to any nodes.
The 3s orbital has no angular nodes , as it has no angular variation. The
3px orbital has an angular node when x = 0, that is the yz plane. The
3dx y orbital has angular nodes when x = 0 or y = 0, corresponding to the
yz and xz planes.
(c) The mean radius is calculated as
∞ π 2π
⟨r⟩ = ∫ ψ ∗3s rψ 3s dτ = ∫ ∫ ∫ ψ ∗3s rψ 3s r 2 sin θ dr dθ dϕ
0 0 0
The wavefunction is written in terms of its radial and angular parts: ψ 3s =

R 3,0 (r)Y0,0 (θ, ϕ). The angular part, the spherical harmonic Y0,0 (θ, ϕ), is
normalized with respect to integration over the angles
π 2π
∫ ∫ Y0,0 (θ, ϕ)∗ Y0,0 (θ, ϕ) sin θ dθ dϕ = 1
0 0
All that remains is to compute the integral over r

∞
⟨r⟩ = ∫ R 3,0 (r)2 r 3 dr
0
The form of R(ρ) is given in Table 8A.1 on page 286, where ρ = 2Zr/3a 0 .
It is convenient to compute the integral over ρ using r 3 = ρ 3 (3a 0 /2Z)3
and dr = (3a 0 /2Z)dρ
∞ ∞
⟨r⟩ = ∫ R 3,0 (r)2 r 3 dr = (3a 0 /2Z)4 ∫ R 3,0 (ρ)2 ρ 3 dρ
0 0
1 3a 0 4 Z 3 ∞
= ( ) ( ) ∫ (6 − 6ρ + ρ 2 )2 ρ 3 e−ρ dρ
243 2Z a0 0
1 3 a0
4 ∞
2 2 3 −ρ
= ∫ (6 − 6ρ + ρ ) ρ e dρ
243 24 Z 0
1 a0 ∞
7 6 5 4 3 −ρ
= ∫ (ρ − 12ρ + 48ρ − 72ρ + 36ρ ) e dρ
48 Z 0
1 a0 1 a0
= [7! − 12(6!) + 48(5!) − 72(4!) + 36(3!)] = × 648
48 Z 48 Z
= (27a 0 )/(2Z)
where the integrals are evaluated using Integral E.3 with k = 1 and the
appropriate value of n.
(d) The radial distribution function is defined in [8A.17b–292], P(r) = r 2 R(r)2 .

It is convenient to express this in terms of ρ = 2Zr/3a 0 using R(ρ) from
Table 8A.1 on page 286, and with r 2 = ρ 2 (3a 0 /2Z)2
1 Z 3 3a 0 2
P3s = ( ) ( ) (6 − 6ρ + ρ 2 )2 ρ 2 e−ρ
243 a 0 2Z
1 Z 6
= (ρ − 12ρ 5 + 48ρ 4 − 72ρ 3 + 36ρ 2 ) e−ρ
108 a 0
1 Z 3 3a 0 2
P3p = ( ) ( ) (4 − ρ)2 ρ 2 ρ 2 e−ρ
486 a 0 2Z
1 Z 6
= (ρ − 8ρ 5 + 16ρ 4 ) e−ρ
216 a 0
1 Z 3 3a 0 2 2 2 2 −ρ
P3d = ( ) ( ) (ρ ) ρ e
2430 a 0 2Z
1 Z 6 −ρ
= ρ e
1080 a 0
Plots of these three functions are shown in Fig. 8.1
0.10 3s
3p
3d
P/(Z/a 0 )
0.05
0.00
0 5 10 15 20
ρ
Figure 8.1
The radial distribution function for the 3s orbital has two subsidiary max-
ima which lie close in to the nucleus, and that for 3p has one such max-
imum. In multi-electron atoms this density close to the nucleus results
in the energies of the 3s, 3p and 3d orbitals no longer being equal: see
Section 8B.3 on page 299.
P8A.7 The probability of finding an electron within a sphere of radius σ is found by

integrating the probability density over all angles and from r = 0 to r = σ
σ π 2π
P(σ) = ∫ ∫ ∫ ∣ψ(r, θ, ϕ)∣2 r 2 sin θ dr dθ dϕ
0 0 0
The ground state of the H atom is the 1s orbital for which the wavefunction is
ψ 1s = (πa 03 )−1/2 e−r/a 0 therefore P(r) = (πa 03 )−1 e−2r/a 0 . Because P(r) does not
depend on the angles, the integral over the angles can be evaluated separately
to give 4π. The expression for P(σ) therefore becomes
σ
P(σ) = (4/a 03 ) ∫ r 2 e−2r/a 0 dr
0
The integral is evaluated using integration by parts

σ σ σ
∫ r 2 e−2r/a 0 dr = −(a 0 /2) r 2 e−2r/a 0 ∣0 + a 0 ∫ re−2r/a 0 dr
0 0
= −(a 0 σ 2 /2)e−2σ/a 0
σ σ
+ a 0 [−(a 0 /2) re−2r/a 0 ∣0 + (a 0 /2) ∫ e−2r/a 0 dr]
0
σ
= −(a 0 σ 2 /2)e−2σ/a 0 + a 0 [−(a 0 σ/2)e−2σ/a 0 − (a 02 /4) e−2r/a 0 ∣0 ]
= a 03 /4 − e−2σ/a 0 [(a 0 σ 2 /2) + (a 02 σ/2) + a 03 /4]
hence
P(σ) = 1 − e−2σ/a 0 [2(σ/a 0 )2 + 2(σ/a 0 ) + 1]
To find the radius at which P(σ) = 0.9 needs the solution to the equation
0.9 = 1 − e−2σ/a 0 [2(σ/a 0 )2 + 2(σ/a 0 ) + 1], which is found numerically to be
σ = 2.66a 0 . Figure 8.2 is a plot of P(σ) as a function of σ; it is a sigmoid curve
and shows, as expected, that the the radius of the sphere increases as the total
enclose probability increases.
1.0
0.8
0.6
P(σ)
0.4
0.2
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
σ/a 0
Figure 8.2
P8A.9 The repulsive centrifugal force of an electron travelling with an angular mo-
mentum J in a circle radius r is J 2 /m e r 3 . Bohr postulated that J = nħ where
n can only take integer values, making the centrifugal force (nħ)2 /m e r 3 . The
atom is in a stationary state when repulsive force is balanced by the attractive
Coulombic force Ze 2 /4πε 0 r 2 , that is Ze 2 /4πε 0 r 2 = (nħ)2 /m e r 3 . This relation-

ship is rearranged to give an expression for the radius r n of the orbit for an
electron in state n, r n = 4π(nħ)2 ε 0 /Ze 2 m e .
The total energy of the state with an electron orbiting at radius r n is the sum
of the kinetic and potential energies. The kinetic energy is written in terms of
the angular momentum as J 2 /2I = J 2 /2m e r n2 , with I the moment of inertia and
J = nħ. The potential energy depends only on r n
J2 Ze 2
E n = Ek + V = −
2m e r n2 4πε 0 r n
(nħ)2 Ze 2
= −
2m e [4π(nħ) ε 0 /Ze m e ]
2 2 2 4πε 0 [4π(nħ)2 ε 0 /Ze 2 m e ]
Z 2 e 4 me 1
= − × 2
32π ε 0 ħ
2 2 2 n
P8A.11 The Bohr radius a 0 is given by [8A.9–286] and the Hartree is defined as E h =
2hc R̃∞ , where the Rydberg constant is given by [8A.14–288]
4πε 0 ħ 2 me e 4
a 0,H = E h,H =
me e 2 4ε 20 h 2
These constants are based on the approximation that the nucleus is infinitely
heavy. If this is not the case, then the mass of the electron m e must be replaced
by the reduced mass of the atom, µ = m e m N /(m e + m N ), where m N is the mass
of the nucleus.
In the case of positronium the ‘nucleus’ has the same mass as the electron, so
µ = m e /2 and hence
4πε 0 ħ 2 (m e /2)e 4
a 0,pos = = 2a 0,H E h,pos = = 1
E
2 h,H
e 2 (m e /2) 4ε 20 h 2
8B Many-electron atoms
D8B.1 See Section 8B.4 on page 305.
D8B.3 This is covered in any introductory or general chemistry text.
E8B.1(a) All configurations have the [Ar] core.
Sc Ti V Cr Mn
2 1 2 2 2 3 1 5
4s 3d 4s 3d 4s 3d 4s 3d 4s2 3d5
Fe Co Ni Cu Zn
2 6 2 7 2 8 1 10
4s 3d 4s 3d 4s 3d 4s 3d 4s 3d10
2
E8B.2(a) [Ar] 3d8
E8B.3(a) Across the period the energy of the orbitals generally decreases as a result of
the increasing nuclear charge. Therefore Li is expected to have the lowest
ionization energy as its outer electron has the highest orbital energy.
E8B.4(a) Hydrogenic orbitals are written in the form [8A.12–287], R n, l (r)Yl ,m l (θ, ϕ),
where the appropriate radial function R n, l is selected from Table 8A.1 on page
286 and the appropriate angular function Yl ,m l is selected from Table 7F.1 on
page 272. Using Z = 2 for the 1s and Z = 1 for the 2s gives
ψ 1s (r) = R 1,0 Y0,0 = 2(2/a 0 )3/2 e−2r/a 0 × (4π)−1/2
ψ 2s (r) = R 2,0 Y0,0 = (8)−1/2 (1/a 0 )3/2 [2 − (r/a 0 )]e−r/2a 0 × (4π)−1/2

The overall wavefunction is simply the product of the orbital wavefunctions
Ψ(r 1 , r 2 ) = ψ 1s (r 1 )ψ 2s (r 2 )
E8B.5(a) For a subshell with angular momentum quantum number l there are 2l + 1
values of m l , each of which corresponds to a separate orbital. Each orbital can
accommodate two electrons, therefore the total number of electrons is 2×(2l +
1). The subshell with l = 3 can therefore accommodate 2(6+1) = 14 electrons.
P8B.1 The radial distribution function for a 1s orbital is given by [8A.18–292], P(r) =
(4Z 3 /a 03 )r 2 e−2Zr/a 0 . This gives the probability density of finding the electron
in a shell of radius r. The most probable radius is found by finding the maxi-
mum in P(r), when dP(r)/dr = 0. In finding this maximum the multiplying
constants are not relevant and can be discarded
d 2 −2Zr/a 0
r e = [2r − (2Z/a 0 )r 2 ] e−2Zr/a 0 = 0
dr
It follows that r max = a 0 /Z; that this is a maximum is most easily seen by
plotting P(r). For Z = 126 the most probable radius will be a 0 /126 .
P8B.3 Toward the middle of the first transition series (Cr, Mn, and Fe) elements ex-
hibit the widest ranges of oxidation states. This is due to the large number
of electrons in the 3d and 4s subshells that have similar energies, and as the
3d electrons that are generally removed provide very little shielding to the 4s
orbitals, the effective nuclear charge does not increase significantly between
adjacent oxidation states, meaning that the ionization energies of these levels
are close, and as these are the outermost electrons the ionization levels are
relatively small, meaning that large numbers of reactions will release enough
energy to lose many electrons.
However, it should be noted that the higher oxidation states of the middle
transition metals do not exist as cations, but only in compounds or compound
ions where there is a significant stabilization of this ion by electron rich atoms,
for example the MnVII state only exists in the MnO4 – ion where there is large
electron donation from bonding with four O2 – ions, as here the effective nu-
clear charge has increased a lot over the neutral atom.
This phenomenon is related to the availability of both electrons and orbitals
favourable for bonding. Elements to the left (Sc and Ti) of the series have few
electrons and relatively low effective nuclear charge leaves d orbitals at high
energies that are relatively unsuitable for bonding. To the far right (Cu and
Zn) effective nuclear charge may be higher but there are few, if any, orbitals
available for bonding. Consequently, it is more difficult to produce a range of
compounds that promote a wide range of oxidation states for elements at either
end of the series. At the middle and right of the series the +2 oxidation state is
very commonly observed because normal reactions can provide the requisite
ionization energies for the removal of 4s electrons.
P8B.5 The first, second and third ionization energies for the group 13 elements are
plotted in Fig. 8.3
40
first IE
second IE
Ionization energy/eV
30 third IE
20
10
0
0 20 40 60 80
Atomic number, Z
Figure 8.3
The following trends are identified.
(a) In all cases, I 1 < I 2 < I 3 because of decreased nuclear shielding as each
successive electron is removed.
(b) The ionization energies of boron are much larger than those of the re-
maining group elements because the valence shell of boron is very small
and compact with little nuclear shielding. The boron atom is much smaller
than the aluminum atom.
(c) The ionization energies of Al, Ga, In, and Tl are comparable even though
successive valence shells are further from the nucleus because the ioniza-
tion energy decrease expected from large atomic radii is balanced by an
increase in effective nuclear charge.
8C Atomic spectra
D8C.1 This is discussed in Section 8C.2(b) on page 309.
D8C.3 The selection rules are given in [8C.8–315]. In part these can be rationalised
by noting that a photon has one unit of (spin) angular momentum and that
in the spectroscopic transition this angular momentum must be conserved.
The selection rule for l, ∆l = ±1, can be understood as a single electron in
the atom changing angular momentum by one unit in order to accommodate
the angular momentum from the photon. This selection is derived in How
is that done? 8C.1 on page 307 by considering the relevant transition dipole
moment. The selection rule for the total spin, ∆S = 0, stems from the fact that
the electromagnetic radiation does not affect the spin directly.
The selection rules for multi-electron atoms are harder to rationalise not least
because the change in the overall angular momentum (L and J) is affected both
by changes in the angular momenta of individual electrons and by the way in
which these couple together.
E8C.1(a) For a d electron l = 2 and s = 21 . Using the Clebsh–Gordon series, [8C.5–312],
the possible values of j are l + s, l + s − 1, . . . ∣l − s∣, which in this case are
j = 25 , 32 .
For an f electron l = 3 and s = 1

2
hence j = 27 , 52 .
E8C.2(a) The Clebsch–Gordan series [8C.5–312], in the form j = l + s, l + s − 1, . . . ∣l − s∣,

with s = 12 implies that there are two possible values of j, j = l ± 21 . Hence, given
that j = 23 , 12 it follows that l = 1 .
E8C.3(a) The symbol D implies that the total orbital angular momentum L = 2 , the su-
perscript 1 implies that the multiplicity 2S + 1 = 1, so that the total spin angular
momentum S = 0 . The subscript 2 implies that the total angular momentum
J =2.
E8C.4(a) The Clebsch–Gordan series, [8C.5–312], is used to combine two spin angular
momenta s 1 and s 2 to give S = s 1 + s 2 , s 1 + s 2 − 1 ..., ∣s 1 − s 2 ∣. (i) For two
electrons, each with s = 12 , S = 1, 0 with multiplicities, 2S = 1, of 3, 1 . (ii)
Three electrons are treated by first combining the angular momenta of two of
them to give S ′ = 0, 1 and then combining each value of S ′ with s 3 = 12 for
the third spin. Therefore, for S ′ = 1, S = 1 + 12 , ∣1 − 12 ∣ = 32 , 12 . Combining
S ′ = 0 with s = 21 simply results in S = 21 . The overall result is S = 23 , 1
2
with
corresponding multiplicities 4, 1 .
E8C.5(a) The valence electron configuration of the Ni2+ is [Ar] 3d8 . In principle the
same process could be adopted as in Exercise E8C.8(a), in which the spin an-
gular momenta of all eight electrons are coupled together in successive steps to
find the overall spin angular momentum. Such an approach would be rather
tedious and would also run the risk of generating values of S which come from
arrangements of electrons which violate the Pauli principle. A quicker method,
and one which ensures that the Pauli principle is not violated, is to consider
combinations of the quantum number m s which gives the z-component of the
spin angular momentum and which takes values ± 12 . The total z-component of
the spin angular momentum is found by simply adding together the m s values:
M S = m s 1 + m s 2 + . . ..
With 8 electrons in the 5 d orbitals, 6 of these electrons must doubly occupy
three of the orbitals, and the Pauli principle requires that the two electrons in
each orbital are spin paired: one has m s = + 12 and one has m s = − 21 . These six
electrons therefore make no net contribution to M S , in the sense that the sum
of the individual m s values is 0. The remaining two electrons can either occupy
the same orbital with spins paired, giving M S = + 21 − 12 = 0, or they can occupy
different orbitals with either their spins paired, giving M S = 0 once more, or
with their spins parallel, giving M S = + 21 + 12 = +1 or M S = − 12 − 12 = −1. Recall
that a total spin S gives M S values of S, (S − 1) . . . − S. Therefore the first
arrangement with just M S = 0 is interpreted as arising from S = 0 , and the
second arrangement with M S = 0, ±1 is interpreted as arising from S = 1 .
E8C.6(a) These electrons are not equivalent, as they are in different subshells, hence all
the terms that arise from the vector model and the Clebsch–Gordan series are
allowed. The orbital angular momentum of the s and d electrons are l 1 = 0 and
l 2 = 2 respectively, and these are combined using L = l 1 +l 2 , l 1 +l 2 −1, ... ∣l 1 −l 2 ∣
which in this case gives L = 2 only. The spin angular momenta of each electron
is s 1 = s 2 = 21 , and these combine in the same way to give S = 1, 0; these values
of S have spin multiplicities of 2S +1 = 3, 1. The terms which arise are therefore
3
D and 1 D.
The possible values of J are given by J = L+S, L+S −1, ..., ∣L−S∣, and hence for
S = 1, L = 2 the values of J are 3, 2, and 1. For S = 0, L = 2 only J = 2 is possible.
The term symbols are therefore 3 D3 , 3 D2 , 3 D1 , and 1 D2 . From Hund’s rules,
described in Section 8C.2(d) on page 315, the lowest energy state is the one with
the greatest spin and then, because the shell is less than half full, the smallest J.
This is 3 D1 .
E8C.7(a) (i) 1 S has L = 0, S = 0 and so J = 0 only; there are 2J + 1 values of M J , which

for J = 0 is just 1 state. (ii) 2 P has L = 1, S = 12 , and so J = 32 , 12 ; the former has
4 states and the latter has 2 states. (iii) 3 P has L = 1, S = 1, and so J = 2, 1, 0 ,
with 5, 3, 1 states, respectively.
E8C.8(a) Closed shells have total spin and orbital angular momenta of zero, and so do
not contribute to the overall values of S and L. (i) For the configuration 2s1
there is just one electron to consider with l = 0 and s = 12 , so L = 0, S = 12 , and
J = 21 . The term symbol is 2 S1/2 . (ii) For the configuration 2p1 there is just one
electron to consider with l = 1 and s = 12 , so L = 1, S = 12 , and J = 32 , 1
2
. The
term symbols are therefore 2 P3/2 and 2 P1/2 .
E8C.9(a) The two terms arising from a d1 configuration are 2 D3/2 , 2 D5/2 , which have
S = 21 , L = 2 and J = 32 , 52 . The energy shift due to spin-orbit coupling is given
by [8C.4–311], E L,S , J = 21 hc Ã[J(J + 1) − L(L + 1) − S(S + 1)], where Ã is the
spin-orbit coupling constant. Hence, E 2,1/2,3/2 = −(3/2)hc Ã , and E 2,1/2,5/2 =
+hc Ã .
E8C.10(a) The selection rules for a many-electron atom are given in [8C.8–315].
(i) 3 D2 (S = 1, L = 2, J = 2) → 3 P1 (S = 1, L = 1, J = 1) has ∆S = 0, ∆L = −1,
∆J = −1 and so is allowed .
(ii) 3 P2 (S = 1, L = 1, J = 2) → 1 S0 (S = 0, L = 0, J = 0) has ∆S = −1, ∆L = −1,
∆J = −2 and so is forbidden by the S and J selection rules.
(iii) 3 F4 (S = 1, L = 3, J = 4) → 3 D3 (S = 1, L = 2, J = 3) has ∆S = 0, ∆L = −1,
∆J = −1 and so is allowed .
E8C.11(a) The spectral lines of a hydrogen atom are given by [8A.1–284], ν̃ = R̃ H (n−2
1 −
n−2
2 ), where R̃ H is the Rydberg constant and ν̃ is the wavenumber of the tran-
sition.
The Lyman series corresponds to n 1 = 1. The lowest energy transition, which
would involve a photon with the longest wavelength, is to the next highest
energy level which has n 2 = 2 . Transitions to higher energy levels involve more
an more energy, and the limit of this is the transition to n 2 = ∞ which involves
the greatest possible energy change and hence the shortest wavelength.
E8C.12(a) The energy levels of a hydrogenic atom are E n = −hcZ 2 R̃ N n−2 , where Z is the
atomic number; for all but the most precise work it is sufficient to approximate
R̃ N by R̃∞ . The wavenumber of the transition between states with quantum
numbers n 1 and n 2 in the He+ ion is given by a modified version of [8A.1–284],
ν̃ = Z 2 R̃∞ (n−2 −2
1 − n 2 ). For the 2 → 1 transition and with Z = 2
ν̃ = 22 × (1.0974 × 105 cm−1 ) × (1−2 − 2−2 ) = 3.29 × 105 cm−1

λ = ν̃ −1 = 1/[22 × (1.0974 × 105 cm−1 ) × (1−2 − 2−2 )]
= 3.03... × 10−6 cm = 30.4 nm
ν = c/λ = (2.9979 × 108 m s−1 )/(3.03... × 10−8 m) = 9.87 PHz
E8C.13(a) The selection rules for a many-electron atom are given in [8C.8–315]. For a
single electron these reduce to ∆l = ±1; there is no restriction on changes in n.
(i) 2s (n = 2, l = 0) → 1s (n = 1, l = 0) has ∆l = 0, and so is forbidden .
(ii) 2p (n = 2, l = 1) → 1s (n = 1, l = 0) has ∆l = −1, and so is allowed .
(iii) 3d (n = 3, l = 2) → 2p (n = 2, l = 1) has ∆l = −1, and so is allowed .
E8C.14(a) The single electron in a p orbital has l = 1 and hence L = 1, and s = 12 hence
S = 21 . The spin multiplicity is 2S + 1 = 2. Using the Clebsh–Gordon series,
[8C.5–312], the possible values of J are J = L + S, L + S − 1, . . . ∣L − S∣ = 32 , 21 .
Hence, the term symbols for the levels are 2 P1/2 , 2 P3/2 .
P8C.1 The wavenumbers of the spectral lines of the H atom for the n 2 → n 1 transition
are given by [8A.1–284], ν̃ = R̃ H (n−2 −2
1 − n 2 ), where R̃ H is the Rydberg constant
−1
for Hydrogen, R̃ H = 109677 cm . Hence, the wavelength of this transition is
λ = ν̃−1 = R̃−1 −2 −2 −1
H (n 1 − n 2 ) .
The lowest energy, and therefore the longest wavelength transition (the one at
λ max = 12368 nm = 1.2368 × 10−3 cm) corresponds to the transition from
n 1 + 1 → n 1 , therefore
1 1 1 (n 1 + 1)2 − n 12 2n 1 + 1
= − = = 2
λ max R̃ H n 1 (n 1 + 1)
2 2 n 1 (n 1 + 1)
2 2 n 1 (n 1 + 1)2
From the given data (λ max R̃ H )−1 = [(1.2368 × 10−3 cm) × (109677 cm−1 )]−1 =
(135.6...)−1 . The value of n 1 is found by seeking an integer value of n 1 for
which n 12 (n 1 + 1)2 /(2n 1 + 1) = 135.6.... For n 1 = 6 the fraction on the left is
62 × 72 /13 = 135.6.... Therefore, the Humphreys series is that with n 1 = 6 .
The wavelengths of the transitions in the Humphreys series are therefore given
by λ = (109677 cm−1 )−1 × (6−2 − n−22 )
−1
for n 2 = 7, 8, .... The next few lines,
with n 2 = 8, 9, and 10 are at 7502.5 nm, 5908.3 nm, 5128.7 nm, respectively.
The convergence limit, corresponding to n 2 = ∞ is 3282.4 nm, as given in the
data.
P8C.3 The wavenumbers of transitions between energy levels in hydrogenic atoms are
given by a modified version of [8A.1–284]
ν̃ = Z 2 R̃ N (n−2 −2
1 − n2 ) (8.2)
where Z is the nuclear charge and R̃ N is the Rydberg constant for the nucleus
in question. In turn this is given by [8A.14–288]
µ me mN
R̃ N = R̃∞ µ=
me me + mN
where m N is the mass of the nucleus. The spectra of 4 He+ and 3 He+ differ
because R̃ N is different for the two atoms. However, this difference is very
small because the value of the reduced mass µ is dominated by the mass of
the electron (m e ≪ m N ). It is therefore necessary to work at high precision.
The first step is to compute R̃ N for each nucleus, using m 4 He = 4.002 602m u
and m3 He = 3.016 029m u .
µ mN
R̃4 He = R̃∞ = R̃∞
me me + mN
4.002 602 × (1.660 539 × 10−27 kg)
=
(9.109 383 × 10−31 kg) + 4.002 602 × (1.660 539 × 10−27 kg)
× (1.097 373 × 105 cm−1 ) = 1.097 223 × 105 cm−1
A similar calculation gives R̃ 3 He = 1.097 173 × 105 cm−1 . With these values of
the Rydberg constant the wavenumber of the relevant transitions is computed
using eqn 8.2; the results are given in the table.
ν̃ 3→2 /cm−1 ν̃ 2→1 /cm−1

4
He+ 60 956.8 329 167
3 +
He 60 954.1 329 152
difference 2.8 15
If the spectrometer has sufficient resolution these differences are detectable; the
greatest difference is for the higher wavenumber transition.
P8C.5 The three transitions originate from the same level, the 2 P, with energy E 2 P ,
and if it is assumed that the nd 2 D states are hydrogenic their energies may be
written E n = −A/n 2 , where A is some constant. It follows that the wavenumber
of the transitions can be written
ν̃ n = E 2 P /hc − (A/hc)/n 2
A plot of ν̃ n against 1/n 2 is therefore expected to a straight line with y-intercept
(at 1/n 2 = 0) E 2 P /hc. The data are tabulated below and the graph is given in
Fig. 8.4.
n 1/n 2 λ/nm ν̃ n /cm−1

3 0.111 610.36 16 384
4 0.063 460.29 21 725
5 0.040 413.23 24 200
30 000
ν̃∞
25 000
ν̃ n /cm−1
20 000
15 000
0.00 0.02 0.04 0.06 0.08 0.10 0.12
2
1/n
Figure 8.4
The data fall on a good straight line which has y-intercept ν̃∞ = E 2 P /hc =
28 595 cm−1 . The transition from the 2 S ground state to the 2 P state is at a
wavelength of 670.78 nm, which corresponds to a wavenumber of 14 908 cm−1 .
Therefore the transition from 2 S to the ionization limit of the 2 P–2 D series will
be at wavenumber
14 908 cm−1 + ν̃∞ = 14 908 cm−1 + 28 595 cm−1 = 43 503 cm−1
This corresponds to the ionization energy of the ground state, which can be
expressed in eV as 5.39 eV . Although the data are given to high precision,
quite a long extrapolation is needed to find the energy of the 2 P state and it also
has been assumed that the constant A is independent of n, which may not be
the case. As a result, the ionization energy is quoted to more modest precision.
P8C.7 The outer electron in K can occupy an s, p or d orbital and such configura-
tions gives rise to 2 S1/2 , 2 P3/2,1/2 , and 2 D5/2,3/2 states, respectively. Taking into
account the selection rules and the effect of spin-orbit coupling, two closely
spaced lines are expected as a result of the transitions 2 S1/2 → 2 P3/2 and 2 S1/2 →
2
P1/2 . The separation of the two lines will reflect the separation of the 2 P3/2 and
2
P1/2 levels, which is computed using [8C.4–311]; the terms in L and S cancel
as they take the same value for the two states
∆E = E 1,1/2,3/2 − E 1,1/2,1/2
= 21 hc Ã[ 32 ( 32 + 1) − L(L + 1) − S(S + 1)]
− 12 hc Ã[ 12 ( 12 + 1) − L(L + 1) − S(S + 1)] = 32 hc Ã
The wavenumber of the separation between the two lines is therefore 23 Ã, hence
Ã = 23 [(766.70 × 10−7 cm)−1 − (770.11 × 10−7 cm)−1 ]

= 32 (57.7... cm−1 ) = 38.5 cm−1
P8C.9 The Rydberg constant for positronium is [8A.14–288] R̃ Ps = R̃∞ × (µ Ps /m e ),

where the reduced mass of the positron–electron system is µ Ps = m e2 /(2m e ) =
m e /2, as the mass of the nucleus is equal to that of the electron. Hence, R̃ Ps =
R̃∞ /2 = (109737 cm−1 )/2 = 54868.5 cm−1 . The spectral lines of the positron-
ium atom are given by ν̃ = R̃ Ps (n−2 −2
1 − n 2 ).
The Balmer series are those lines with n 1 = 2, and so the wavenumbers of these
are
ν̃ = R̃ Ps (2−2 − n−2 −1
2 ) = (54868.5 cm ) × (2
−2
− n−2
2 ) n 2 = 3, 4 . . .
The first three lines have n 2 = 3, 4, 5 and are at 7 621 cm−1 , 10 288 cm−1 , and
11 522 cm−1 , respectively. The ionization energy is simply the binding energy
of the ground state, which is hc R̃ Ps . Hence I = hc R̃ Ps /e = 6.803 eV
P8C.11 The derivation follows the method used in How is that done? 8C.1 on page
307. For a transition to be allowed the transition dipole moment µfi must
be non-zero. It is convenient to explore this condition by examining the x-,
y-, and z-components of the moment: if any of these are non-zero, the over-
all moment will also be non-zero. The x- and y-components are given by
µ x ,fi = −e ∫ ψ ∗f xψ i dτ and µ y,fi = −e ∫ ψ ∗f yψ i dτ, respectively. The limits of
integration are r = 0 to ∞, θ = 0 to π, and ϕ = 0 to 2π, and the volume element
is dτ = r 2 sin θ dr dθ dϕ.
The first step is to express the Cartesian co-ordinates in spherical polar coordi-
nates and then in terms of the spherical harmonics. From The chemist’s toolkit
21 in Topic 7F on page 272 it is seen that x = r sin θ cos ϕ, and y = r sin θ sin ϕ.
From Table 7F.1 on page 272 the spherical harmonics Y1,±1 are ∓N sin θe±iϕ ,
where N is the normalization constant. Using the identity e±iϕ = cos ϕ ± i sin ϕ
it follows that
Y1,+1 + Y1,−1 = −N sin θ(cos ϕ + i sin ϕ) + N sin θ(cos ϕ − i sin ϕ)

= −2iN sin θ sin ϕ
Similarly
Y1,+1 − Y1,−1 = −N sin θ(cos ϕ + i sin ϕ) − N sin θ(cos ϕ − i sin ϕ)

= −2N sin θ cos ϕ
With these relationships x and y are expressed as
x = r sin θ cos ϕ = −r(Y1,+1 − Y1,−1 )/2N

y = r sin θ sin ϕ = −r(Y1,+1 + Y1,−1 )/2iN
The wavefunctions of the atomic orbitals are expressed in terms of a radial and
an angular part: ψ n, l ,m l = R n,l (r)Yl ,m l (θ, ϕ). As is seen in the calculation in
the text, the selection rule is derived by considering only the integral over the
angles. Focusing just on this, and setting aside all the normalization and other
factors, the integral to consider for the x-component of the transition moment
is
π 2π
µ x ,fi ∝ ∫ ∫ Yl∗f ,m l ,f (Y1,+1 − Y1,−1 )Yl i ,m l ,i sin θ dθ dϕ
0 0
It is a property of spherical harmonics that the ‘triple integral’
π 2π
∫ ∫ Yl∗f ,m l ,f Yl ,m Yl i ,m l ,i sin θ dθ dϕ
0 0
vanishes unless l f = l i ± l and m l ,f = m l ,i ± m. In this case the integrals of

interest have l = 1 and m = ±1, therefore they are non-zero only if l f = l i ± 1
and m l ,f = m l ,i ± 1. It follows that the integral on which the x-component
of the transition depends is only non-zero if these conditions are satisfied. A
similar argument applies to the y-component. The selection rules are therefore
∆l = ±1, ∆m l = ±1 .
In the text it is seen that the selection rule deriving from the z-component of
the transition moment is ∆l = ±1, ∆m l = 0; when the x- and y- components
are considered as well, transitions with ∆m l = ±1 are also allowed.

I8.1 (a) The ground state of the He+ ion is 1s1 with S = 21 , L = 0 and hence J = 12 .
The term symbol is therefore 2 S1/2 . The excited state configuration is 4p1
which has S = 12 , L = 1 and hence J = 23 or 12 ; the term symbols are
2
P3/2 and 2 P1/2 , the lowest of which is that with J = 12 . According to the
selection rules the transitions from 2 S1/2 to both 2 P states are allowed:
hence, the transitions are 2 S1/2 → 2 P1/2 and 2 S1/2 → 2 P3/2 .
(b) The wavenumber the spectral line corresponding to the n 1 → n 2 transi-
tion is given by a modified version of [8A.1–284] which takes into account
the nuclear charge Z: ν̃ = Z 2 R̃ H (n−2 −2
1 − n 2 ), where R̃ H is the Rydberg
constant for hydrogen, 109 677 cm ; Z = 2 for He+ . In principle the
−1
Rydberg constant is different for He, but the change is so small that it can
safely be ignored. Hence for a transition for n = 1 → 4, the wavenumber
is ν̃ = 4 × (109 677 cm−1 ) × (1−2 − 4−2 ) = 411 289 cm−1 .
This corresponds to a wavelength of λ = ν̃−1 = (411 289 cm−1 )−1 =
2.43... × 10−6 cm = 24.313 8 nm . The corresponding frequency is
ν = cλ−1 = c ν̃ = (2.997 925 × 1010 cm s−1 ) × (411 289 cm−1 )

= 1.233 01 × 1016 Hz
(c) The mean radius of a hydrogenic orbital, characterized by quantum num-

bers, n, l , m l is given by
n2 a0 l(l + 1)
⟨r⟩n, l ,m l = [1 + 12 (1 − )]
Z n2
For the ground state orbital, with Z = 2, n = 1 and l = 0 in He+
(12 )a 0 0(0 + 1) 3a 0
⟨r⟩1,0,0 = [1 + 12 (1 − 2
)] =
2 1 4
For the upper state with n = 4 and l = 1
(4)2 a 0 1(1 + 1) 23a 0

⟨r⟩4,1,0 = [1 + 12 (1 − )] =
2 42 2
Hence, the mean radius of the atom increases by 23a 0 /2−3a 0 /4 = 43a 0 /4 .
I8.3 Because the beam splits into two, with deflections ±(µ B L 2 /4E k )dB/dz, a split-
ting between the two beams of ∆x is achieved by satisfying the condition ∆x =
(µ B L 2 /2E k )dB/dz, which is rearranged to give an expression for the field gra-
dient dB/dz = 2E k ∆x/µ B L 2 . A reasonable estimate for the mean kinetic energy
is to take the equipartition value of 23 kT.
dB 2E k ∆x 3kT∆x
= =
dz µB L2 µB L2
3 × (1.3806 × 10−23 J K−1 ) × (1000 K) × (1.00 × 10−3 m)
=
(9.2740 × 10−24 J T−1 ) × (50 × 10−2 m)2
= 17.9 T m−1
9 Molecular Structure
9A Valence-bond theory
D9A.1 See Section 9A.3(b) on page 327 for details on hybridization applied to simple
carbon compounds. The carbon atoms in alkanes are sp3 hybridized. This
explains the nearly tetrahedral bond angles about the carbon atoms in such
molecules.
The double-bonded carbon atoms in alkenes are sp2 hybridized. This explains
the bond angles of approximately 120○ about these atoms. The simultaneous
overlap of sp2 hybridized orbitals and unhybridized p orbitals in C=C double
bonds explains the resistance of such bonds to torsion and the co-planarity of
the atoms attached to those atoms.
The triple-bonded carbon atoms in alkynes are sp hybridized, which explains
the 180○ bond angles about these atoms. The central carbon atom in allene is
also sp hybridized. Each of its C=C double bonds involves one of its sp hybrids
and one unhybridized p orbital. The two resulting π orbitals are oriented per-
pendicular to one another, which is why the two CH2 groups are rotated by
90○ relative to one another. This arrangement of orbitals also accounts for the
resistance to the two CH2 groups being rotated relative to one another about
the long axis.
D9A.3 Resonance refers to the superposition of the wave functions representing dif-
ferent electron distributions in the same nuclear framework. The wavefunction
resulting from the superposition is called a resonance hybrid.
Resonance allows for a more refined description of the electron distribution,
and hence bonding, than is given by a single valence bond wavefunction. Dif-
ferent valence bond structures are allowed to contribute to different extents,
meaning that the overall wavefunction is built up from contributions from
different valence-bond wavefunctions.
This approach makes it possible to describe polar bonds as a combination of a
purely covalent and a purely ionic structure, and delocalized bonding in terms
of combinations of valence-bond structures in which, for example, a double
bond is located in different parts of a molecule. Resonance is a device for
calculating an improved wavefunction: it does not imply that wavefunction
flickers between those for the different structures.
D9A.5 Promotion and hybridization are two modifications to the simplest version of
valence-bond (VB) theory, adopted to overcome obvious mismatches between
312 9 MOLECULAR STRUCTURE
predictions of that theory and observations. In its simplest form VB theory

assumes that the functions ψ A and ψ B that appear in a VB wavefunction, [9A.2–
324], are orbitals in free atoms occupied by unpaired electrons. For exam-
ple, such a theory would predict that carbon, with the electronic configuration
2s2 2p2 , would form two bonds on account if it having two unpaired electrons.
This prediction is at odds with the characteristic valency of four shown by
carbon.
To account for the tetravalence of carbon it is supposed that one of the 2s
electrons is excited (‘promoted’) to the empty 2p orbital, giving a configuration
of 2s1 2p3 . There are now four unpaired electrons (in the 2s and 2p orbitals)
available for forming four valence bonds.
Hybrid orbitals are invoked to account for the fact that valence bonds formed
from atomic orbitals would have different orientations in space than are com-
monly observed. For instance, the four bonds in CH4 are observed to be equiv-
alent and directed toward the corners of a regular tetrahedron. By contrast,
bonds made from the three distinct 2p orbitals in carbon would be expected to
be oriented at 90○ angles from each other, and those three bonds would not be
equivalent to the bond made from a 2s orbital. Hybrid atomic orbitals, in this
case sp3 hybrids, are formed by combining the atomic orbitals in such a way
that the hybrid orbitals have the required directional properties.
E9A.1(a) The ammonium ion is iso-electronic with methane, therefore the two species
are expected to have the same description of bonding. Four sp3 hybrid atomic
orbitals are formed from the 2s and the three 2p orbitals of the nitrogen atom;
each hybrid then forms a σ bond by overlapping with a hydrogen 1s orbital.
E9A.2(a) All the carbon atoms in 1,3-butadiene are sp2 hybridized. The σ framework of
the molecule consists of C–H and C–C σ bonds. Each C–H σ bond is formed
by the overlap of an sp2 hybrid atomic orbital on a carbon atom with a 1s atomic
orbital on a neighbouring hydrogen atom. Similarly, C–C σ bonds are formed
by the overlap of sp2 hybrid atomic orbitals on neighbouring carbon atoms.
The two π bonds are formed by the side-by-side overlap of unhybridized 2p
orbitals on carbon atoms C1 and C2, and likewise between C3 and C4.
E9A.3(a) The carbon and nitrogen atoms in methylamine are sp3 hybridized. The C–N
bond is formed by the overlap of an sp3 orbital on carbon with an sp3 orbital
on nitrogen. The C–H bonds are formed by the overlap of a carbon sp3 hybrid
atomic orbital with a hydrogen 1s atomic orbital. Similarly, the N–H bonds are
formed by the overlap of a nitrogen sp3 hybrid atomic orbital with a hydrogen
1s atomic orbital. The lone pair on nitrogen resides on an sp3 hybrid atomic
orbital.
E9A.4(a) The condition of orthogonality is given by [7C.8–240], ∫ Ψi∗ Ψ j dτ = 0 for i ≠ j.

The atomic orbitals are all real, therefore Ψi∗ = Ψi . The orthogonality condition
becomes
∗
∫ h 1 h 2 dτ = ∫ (s + px + p y + pz )(s − px − p y + pz ) dτ
1 0 0 0
³¹¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
= ∫ s2 dτ − ∫ spx dτ − ∫ sp y dτ + ∫ spz dτ
1 1 1
³¹¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ µ
2
+ ... − ∫ px dτ +.... − ∫ p y dτ +... + ∫ p2z dτ 2
=1−1−1+1=0
All the integrals of the form ∫ sp i dτ are zero because the s and p orbitals are
orthogonal, and all the integrals of the form ∫ s2 dτ and ∫ p2i dτ are 1 because
the orbitals are normalized. The condition for the orthogonality of h 1 and h 2
is satisfied.
E9A.5(a) A normalized wavefunction satisfies [7B.4c–234], ∫ Ψ ∗ Ψ dτ = 1. The wave-

function is normalized by finding the value of N for which h = N(s + 21/2 p)
satisfies this condition. The orbital wavefunctions s and p are real as is N,
therefore
∗
∫ h h dτ = N ∫ (s + 2 p) dτ
2 1/2 2
⎡ 1 1 0 ⎤
⎢³¹¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ µ⎥⎥
⎢
⎢ ⎥
= N 2 ⎢∫ s2 dτ +2 ∫ p2 dτ +23/2∫ sp dτ ⎥ = 3N 2
⎢ ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
The integral ∫ sp dτ is zero because the s and p orbitals are orthogonal, and the
integrals ∫ s2 dτ and ∫ p2 dτ are 1 because the orbitals are normalized. From
the normalization condition it follows that 3N 2 = 1 and hence N = 1/31/2 .
E9A.6(a) Using [9A.2–324] and assuming that the valence-bond in HF is formed be-
tween the H1s and F2pz atomic orbitals, the spatial part of the valence-bond
wavefunction is written as Ψ(1, 2) = ψF2pz (1)ψH1s (2) + ψF2pz (2)ψH1s (1). The
overall wavefunction must be antisymmetric to satisfy the Pauli principle, there-
fore the symmetric spatial part has to be combined with the antisymmetric two-
electron spin wavefunction given by [8B.3–299], σ− (1, 2). The (unnormalized)
complete two-electron wavefunction is therefore
Ψ(1, 2) = [ψF2pz (1)ψH1s (2) + ψF2pz (2)ψH1s (1)] × [α(1)β(2) − β(1)α(2)]
E9A.7(a) The resonance hybrid wavefunction constructed from one two-electron wave-
function corresponding to the purely covalent form of the bond and one two-
electron wavefunction corresponding to the ionic form of the bond is given
in [9A.3–326] as Ψ = Ψcovalent + λΨionic . Therefore the (unnormalized) reso-
nance hybrid wavefunction of HF with two ionic structures is written as ΨHF =
ΨH–F + λΨH+ F− + κΨH− F+ . ΨH–F is written as in Exercise E9A.1(a), ΨH–F =
[ψF2pz (1)ψH1s (2) + ψF2pz (2)ψH1s (1)] × σ− (1, 2). The wavefunction ΨH+ F− de-
scribes the electron distribution when both electrons reside on the F2pz or-
bital. The spatial part of this wavefunction is given by ψF2pz (1) ψF2pz (2) , which
is symmetric, therefore it has to be combined with the antisymmetric spin
wavefunction resulting in ΨH+ F− = [ψF2pz (1) ψF2pz (2) ] × σ− (1, 2). Similarly, the
other ionic structure has ΨH− F+ = [ψH1s(1) ψH1s(2) ] × σ− (1, 2).
E9A.8(a) Both phosphorus and nitrogen are in Group 15, therefore the valence bond
description of the bonding in P2 is similar to that of N2 . There is a triple bond
between the two sp hybridized phosphorus atoms. A σ bond is formed by the
overlap of two sp hybrid atomic orbitals projecting towards each other along
the internuclear axis. The two π bonds are the result of the side-by-side overlap
of 3px with 3px and 3p y with 3p y orbitals. There is one lone pair on each
phosphorus atom, contained in the sp orbital projecting outwards along the
internuclear axis.
Consider the equilibrium P4 ↽ ⇀ 2 P2 . In the tetrahedral P4 there are six σ
ÐÐÐ
bonds, whereas in two molecules of P2 there are two σ and four π bonds overall.
π bonds are generally weaker than σ bonds, therefore the equilibrium favors P4 .
P9A.1 The wavefunction in terms of the polar coordinates of each electron is given in
Brief illustration 9A.1 on page 324 as
1
Ψ(1, 2) = [e−(rA1 +rB2 )/a 0 + e−(rA2 +rB1 )/a 0 ]
πa 03
Given that the internuclear separation along the z-axis is R, in Cartesian coor-
dinates rAi and rBi becomes
rAi = (x i2 + y 2i + z 2i )1/2 and rBi = (x i2 + y 2i + (z i − R)2 )1/2
Therefore the wavefunction is
1
Ψ(1, 2) =
πa 03
× [e−[(x 1 +y 1 +z 1 ) +(x 22 +y 22 +(z 2 −R)2 )1/2 ]/a 0
+ e−[(x 2 +y 2 +z 2 ) +(x 12 +y 12 +(z 1 −R)2 )1/2 ]/a 0
2 2 2 1/2 2 2 2 1/2
]
P9A.3 For the purposes of this problem, the px and p y orbitals are represented by
unit vectors along the x-and y-axes, respectively. The given hybrid atomic
orbitals are created by the linear combination of the s, px and p y orbitals. The s
orbital is spherically symmetric about the origin, therefore it does not modify
the directions in which the hybrids point. The vector representations of the
hybrid atomic orbitals are
√ √ √ √ √
h1 = 2j h2 = 3/2 i − 1/2 j h3 = − 3/2 i − 1/2 j
y
h1
√
3
2
α α √x
1
2
h3 h2
√
From the diagram it is evident that α = tan−1 (1/ 3) = 30○ . It follows that the
angle between adjacent hybrids is 120○ .
9B Molecular orbital theory: the hydrogen molecule-ion

D9B.1 As described in Section 9B.1(b) on page 333, the reason why the bonding molec-
ular orbital is lower in energy than the atomic orbitals is not entirely clear.
However, it is clear that bonding character correlates strongly with molecular
orbitals that have an accumulation of electron density between nuclei due to
overlap and constructive interference of their component atomic orbitals. A
simple and plausible explanation of this correlation is that enhanced electron
probability between nuclei lowers the potential energy by putting electrons in
a position where they can be attracted to two nuclei at the same time; however,
the source of the reduced energy may be more complicated.
D9B.3 The Born–Oppenheimer approximation treats the nuclei of the multi-particle

system of electrons and nuclei as if they were fixed. The dependence of en-
ergy on nuclear positions is then obtained by solving the Schrödinger equa-
tion at many different (fixed) nuclear geometries. Molecular potential energy
curves and surfaces are plots of molecular energy (computed under the Born–
Oppenheimer approximation) as a function of nuclear coordinates.
E9B.1(a) The energy of the σ bonding orbital in H2 + is given by [9B.4–333], E σ = EH1s +
j 0 /R − ( j + k)/(1 + S). Molecular potential energy curves are usually plotted
with respect to the energy of the separated atoms, therefore the energies to be
plotted are E σ − EH1s = j 0 /R − ( j + k)/(1 + S).
Using [9B.5d–333], j 0 /a 0 = 27.21 eV = 1 Eh the energy for R/a 0 = 1 is com-
puted as
(1 E h ) (0.729 E h ) + (0.736 E h )
E σ − EH1s = − = +0.211 E h
1 (1 + 0.858)
Similar calculations give the following energies

R/a 0 1 2 3 4
−2 −2
(E σ − EH1s )/Eh +0.211 −5.32 × 10 −5.88 × 10 −3.76 × 10−2
(E σ − EH1s )/E h 0.2
0.1
R/a 0
0.0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
−0.1
Figure 9.1
These data are plotted in Fig. 9.1; with so few data points it is difficult to locate
the minimum. The data are fitted well by the following cubic
(E σ − EH1s )/E h = −0.0386(R/a 0 )3 + 0.3611(R/a 0 )2 − 1.0771(R/a 0 ) + 0.9656
Note that this cubic equation has no physical meaning, it is only used to draw
the line on the plot above and to locate the minimum by setting the derivative
to zero; in particular the maximum close to the final data point has no physical
basis. This minimum is found to be at R = 2.5 a 0 , which corresponds to the
predicted equilibrium bond length. The depth of the potential energy well at
this distance is about −0.073 E h which is 2.0 eV .
E9B.2(a) A sketch of the bonding and the antibonding molecular orbitals resulting from
the side-by-side overlap of two p orbitals is shown in Fig. 9C.5 on page 339. The
bonding molecular orbital is antisymmetric with respect to inversion, therefore
it is denoted as a πu orbital. The antibonding molecular orbital is symmetric
with respect to inversion, therefore it is a πg orbital.
E9B.3(a) The normalization condition is given by [7B.4c–234], ∫ ψ ∗ ψ dτ = 1. The wave-

function is normalized by finding N such that ψ = N(ψA + λψB ) satisfies this
condition. The wavefunctions ψA and ψB are real, as is N, therefore
∗
∫ ψ ψ dτ = N ∫ (ψA + λψB ) dτ
2 2
⎡ 1 1 S ⎤
⎢³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ⎥⎥
⎢
⎢ ⎥
= N 2 ⎢∫ ψA2 dτ +λ 2 ∫ ψB2 dτ +2λ∫ ψA ψB dτ ⎥
⎢ ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
= N (1 + λ + 2λS)
2 2
The integrals ∫ ψA2 dτ and ∫ ψB2 dτ are 1 because the wavefunctions ψA and ψB
are normalized. It follows that N = 1/(1 + λ 2 + 2λS)1/2 .
E9B.4(a) The condition of orthogonality is given by [7C.8–240], ∫ ψ ∗i ψ j dτ = 0 for i ≠ j.

The given molecular orbital, ψ i = 0.145A + 0.844B is real, therefore ψ ∗i = ψ i .
The new linear combination for A and B, which is orthogonal to ψ i must have
the form of ψ j = A + βB, where the coefficient of wavefunction A is chosen
to be 1 for simplicity. Substitution of these wavefunctions in the condition of
orthogonality gives
∗
∫ ψ i ψ j dτ = ∫ (0.145A + 0.844B) × (A + βB) dτ
1 1 S
³¹¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
= 0.145∫ A dτ +0.844β ∫ B dτ +(0.145β + 0.844)∫ AB dτ
2 2
= 0.145 + 0.844β + (0.145β + 0.844)S
Using S = 0.250 the value of the integral becomes 0.356 + 0.88025β. This value
must be zero for the two wavefunctions to be orthogonal, therefore β = −0.404
and hence ψ j = A − 0.404B.
Normalization of ψ i follows the same logic as in Exercise E9B.3(a). First the

wavefunction is written as ψ i = N(0.145A + 0.844B) and then the normaliza-
tion constant N is found such that ∫ ψ ∗ ψ dτ = 1.
∗
∫ ψ i ψ i dτ = ∫ [N(0.145A + 0.844B)] dτ
2
1 1 S
³¹¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
= N (0.145 ∫ A dτ +0.844 β∫ B dτ +(2 × 0.145 × 0.844)∫ AB dτ )
2 2 2 2 2
= N 2 (0.733 + 0.245S)
√
Using S = 0.250 gives a value of 0.794N 2 for the integral, therefore N = 1/ 0.794 =
1.12. Therefore the normalized wavefunction is
ψ i = 1.12 × (0.145A + 0.844B) = 0.163A + 0.947B
Normalization of ψ j follows a similar procedure as for ψ i , giving N = 1.02 and

therefore ψ j = 1.02A − 0.412B .
P9B.1 Inspection of [9B.1–331] reveals that the repulsion energy between two hydro-
gen nuclei is given by e 2 /4πε 0 R, where R is the internuclear separation. In
molar quantities, the repulsion energy is N A e 2 /4πε 0 R, which, at an equilibrium
separation of R = 74.1 pm becomes
(6.0221 × 1023 mol−1 ) × (1.6022 × 10−19 C)2

= 1.87 × 106 J mol−1
4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (74.1 × 10−12 m)
The molar gravitational potential energy between two hydrogen nuclei is

N A Gmp2
R
(6.0221 × 1023 mol−1 ) × (6.6738 × 10−11 N m2 kg−2 ) × (1.6726 × 10−27 kg)2
=
(74.1 × 10−12 m)
= 1.52 × 10−30 J mol−1
Therefore the gravitational attraction is entirely negligible compared to the elec-

trostatic repulsion between the two nuclei.
P9B.3 Refer to the data presented in Exercise E9B.1(a). The energy of the σ bonding
orbital in H2 + is given by [9B.4–333], E σ = EH1s + j 0 /R − ( j + k)/(1 + S). This
energy in usually measured with respect to the energy of the separated atoms,
therefore the energy is E σ − EH1s = j 0 /R − ( j + k)/(1 + S). Likewise for the
σ∗ antibonding orbital the energy is given by [9B.7–335], E σ∗ = EH1s + j 0 /R −
( j − k)/(1 − S). Relative to the separated atoms, the energy is E σ∗ − EH1s =
j 0 /R − ( j − k)/(1 − S). With the data given, and using j 0 /a 0 = 27.21 eV = 1 Eh ,
the energies of these molecular orbitals are
R/a 0 1 2 3 4
−2 −2
(E σ − EH1s )/Eh +0.211 −5.32 × 10 −5.88 × 10 −3.76 × 10−2
(E σ∗ − EH1s )/Eh +1.05 +0.340 +0.132 +5.52 × 10−2
It is evident that at each distance the antibonding molecular orbital is raised in

energy by more than the bonding molecular orbital is lowered. This appears to
be generally true for any reasonable internuclear separation.
P9B.5 The bonding and antibonding MOl wavefunctions are ψ± = N± (ψ A ± ψ B ),

where N± is the normalizing factor, given by (Example 9B.1 on page 332)
1
N± =
[(2(1 ± S)]1/2
where for two 1s AOs separated by a distance R the overlap integral is given
by [9B.5a–333], S = (1 + R/a 0 + 13 (R/a 0 )2 ) e−R/a 0 . The form of ψ A and ψ B are
given in Brief illustration 9B.1 on page 332
ψ A = (1/πa 03 )1/2 e−r A1 /a 0 ψ B = (1/πa 03 )1/2 e−r B1 /a 0
Without loss of generality, it is assumed that atom A is located at z A1 = 0 and

atom B at z B1 = R, the internuclear separation. The requirement is to plot the
wavefunction along the z-axis, so x A1 = y A1 = 0, and likewise for orbital B.
With all of these conditions imposed the function to plotted is
1 1
ψ± = (e−∣z∣/a 0 ± e−∣(z−R)∣/a 0 )
[(2(1 ± S)]1/2 (πa 03 )1/2
The modulus signs are needed because the argument of the exponential is the
distrance from the nucleus, which is always positive.
Figure 9.2 shows plots of (a) the bonding and antibonding wavefunctions, and
(b) the squares of these functions (the probability density) for the case R = 2a 0 .
The quantity plotted in (a) is (a 0 )3/2 ψ± , and in (b) it is (a 0 )3 ψ±2 ; the same scale
is used for each orbital.
The antibonding orbital has a node at the mid-point of the bond, whereas the
bonding orbital has significant electron density at this point. From the dia-
grams it appears that the antibonding orbital has greater overall probability, but
this is not in fact the case – both orbitals are normalized. It is just that when
the functions are plotted along the z-axis there appears to be such a difference
due to the different distribution of electron density elsewhere in the orbitals.
The difference density is the difference in electron density between the molecu-
lar orbital and two non-interacting 1s orbitals, one on each atom. It is a measure
of the way in which the electron density is changed when the molecular orbitals
are compared to non-interacting atomic orbitals. The difference density is given
by
ψ±2 − 21 (ψ A2 + ψ B2 )
Also shown in Fig. 9.2 is this difference density for (c) the bonding and (d)
the antibonding molecular orbital; the same scale is used for each plot. If
the difference density is positive the implication is that the electron density
is greater than for two non-interacting atomic orbitals, whereas if the differ-
ence density is negative the implication is that there is a reduction in electron
density. It is evident from the plots that in the bonding molecular orbital there
is an increase in the electron density in the internuclear region, whereas for
the antibonding orbital the density in this region is reduced, but the density is
increased further away. These observations account (partially, at least) for the
fact that occupying the bonding molecular orbital promotes bond formation.
The apparent difference in size between the difference densities between (c) and
(d) is a result of simply plotting the function along the z-axis.
9C Molecular orbital theory: homonuclear diatomic molecules

D9C.1 The building-up principle for homonuclear diatomic molecules is essentially
the same as for atoms, but the diatomic molecular orbitals used in the former
are different in name and in nature than the atomic orbitals used in the latter.
A diagram of energy levels (orbitals) and degeneracies is needed. For diatomic
molecules, these energy levels are either nondegenerate (for σ bonds) or doubly
degenerate (for all others). The orbitals are populated with electrons, placing
each successive electron in the lowest-energy orbital available, no more than
two electrons per orbital. Hund’s rule indicates that different degenerate or-
bitals should be populated first, with electrons that have parallel spins, before
pairing two electrons in the same degenerate orbital.
D9C.3 The bond strength is related to the extent to which the occupied bonding molec-
ular orbitals are lowered in energy compared to the constituent atomic orbitals.
(a) (b)
antibonding
bonding
–4 –2 0 2 4 6
z/a0
–4 –2 0 2 4 6
z/a0
(c) (d)
antibonding
bonding
–4 –2 0 2 4 6 –4 –2 0 2 4 6
z/a0 z/a0
Figure 9.2
As described in Topic 9B for the case of H2 + , this lowering in energy depends

on the size of the term k, [9B.5c–333], which is a measure of the interaction
between a nucleus and the excess electron density in the internuclear region
arising from overlap.
The overlap integral, S, is a different quantity than k, but its behaviour with
(for example) internuclear distance is quite similar. Thus the overlap integral
is often taken as a proxy for k, not least as it much easier to imagine how
the overlap varies when the orbital or the internuclear distance is varied. It is
therefore common to speak of a bond being strong when ‘there is good overlap’.
The fact that there is a correlation between overlap and bond strength may,
however, simply be fortuitous as the theory does not indicate such a connection.
E9C.1(a) The molecular orbital energy level diagram for Li2 , Be2 , B2 , C2 and N2 is shown
in Fig. 9C.12 on page 341, and for O2 , F2 and Ne2 in Fig. 9C.11 on page 341.
Following the same logic as in Exercise E9C.4(a) and Exercise E9C.5(a) gives
Li2 1 + 1 = 2 VE 1σ 2g b = 21 (2 − 0) = 1
Be2 2 + 2 = 4 VE 1σ 2g 1σ∗2
u b = 12 (2 − 2) = 0
B2 3 + 3 = 6 VE 1σ 2g 1σ∗2 2
u 1π u b = 12 (4 − 2) = 1
C2 4 + 4 = 8 VE 1σ 2g 1σ∗2 4
u 1π u b = 12 (6 − 2) = 2
N2 5 + 5 = 10 VE 1σ 2g 1σ∗2 4 2
u 1π u 2σ g b = 12 (8 − 2) = 3
O2 6 + 6 = 12 VE 1σ 2g 1σ∗2 2 4 ∗2
u 2σ g 1π u 1π g b = 21 (8 − 4) = 2
F2 7 + 7 = 14 VE 1σ 2g 1σ∗2 2 4 ∗4
u 2σ g 1π u 1π g b = 21 (8 − 6) = 1
Ne2 8 + 8 = 16 VE 1σ 2g 1σ∗2 2 4 ∗4 ∗2
u 2σ g 1π u 1π g 2σ u b = 21 (8 − 8) = 0
E9C.2(a) The molecular orbital energy level diagram for Li2 , Be2 , B2 , C2 , N2 and their
ions is shown in Fig. 9C.12 on page 341, and for O2 , F2 , Ne2 and their ions in
Fig. 9C.11 on page 341. The highest occupied molecular orbital (HOMO) is the
molecular orbital which is the highest in energy and is at least singly occupied.
The HOMO of each of the listed ions is indicated by a box around it.
Li2 + 1 + 1 − 1 = 1 VE 1σ 1g
Be2 + 2 + 2 − 1 = 3 VE 1σ 2g 1σ∗1
u
B2 + 3 + 3 − 1 = 5 VE 1σ 2g 1σ∗2 1
u 1π u
C2 + 4 + 4 − 1 = 7 VE 1σ 2g 1σ∗2 3
u 1π u
N2 + 5 + 5 − 1 = 9 VE 1σ 2g 1σ∗2 4 1
u 1π u 2σ g
O2 + 6 + 6 − 1 = 11 VE 1σ 2g 1σ∗2 2 4 ∗1
u 2σ g 1π u 1π g
F2 + 7 + 7 − 1 = 13 VE 1σ 2g 1σ∗2 2 4 ∗3
u 2σ g 1π u 1π g
Ne2 + 8 + 8 − 1 = 15 VE 1σ 2g 1σ∗2 2 4 ∗4 ∗1
u 2σ g 1π u 1π g 2σ u
Li2 − 1 + 1 + 1 = 3 VE 1σ 2g 1σ∗1
u
Be2 − 2 + 2 + 1 = 5 VE 1σ 2g 1σ∗2 1
u 1π u
B2 − 3 + 3 + 1 = 7 VE 1σ 2g 1σ∗2 3
u 1π u
C2 − 4 + 4 + 1 = 9 VE 1σ 2g 1σ∗2 4 1
u 1π u 2σ g
N2 − 5 + 5 + 1 = 11 VE 1σ 2g 1σ∗2 4 2 ∗1
u 1π u 2σ g 1π g
O2 − 6 + 6 + 1 = 13 VE 1σ 2g 1σ∗2 2 4 ∗3
u 2σ g 1π u 1π g
F2 − 7 + 7 + 1 = 15 VE 1σ 2g 1σ∗2 2 4 ∗4 ∗1
u 2σ g 1π u 1π g 2σ u
Ne2 − 8 + 8 + 1 = 17 VE 1σ 2g 1σ∗2 2 4 ∗4 ∗2 1
u 2σ g 1π u 1π g 2σ u 3σ g
Note that the extra electron in Ne2 is accommodated on a bonding molecular

orbital resulting from the overlap of the 3s atomic orbitals.
E9C.3(a) The energy of the incident photon must equal the sum of the ionization energy
of the orbital and the kinetic energy of the ejected photoelectron, [9C.5–342],
hν = I + 21 me υ 2 . The energy of the incident photon is given by hν = hc/λ =
(6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )/(100 × 10−9 m) = 1.98... × 10−18 J.
Rearranging the equation to give the speed of the ejected electron gives
1/2
2 hc
υ=[ ( − I)]
me λ
2
=[ × [(1.98... × 10−18 J)
(9.1094 × 10−31 kg)
1/2
− (12.0 eV) × (1.6022 × 10−19 J eV−1 )]] = 3.70 × 105 m s−1
E9C.4(a) The molecular orbital diagram for the homonuclear diatomic molecules Li2 ,
Be2 , and C2 is shown in Fig. 9C.12 on page 341. According to the Pauli principle,
up to two valence electrons can be placed in each of the molecular orbitals. First
the lowest energy orbital is filled up, then the next lowest and so on, until all
the valence electrons are used up.
(i) Li2 has 1+1 = 2 valence electrons (VE) overall, therefore the ground-state
electron configuration is 1σ 2g . The bond order is defined in [9C.4–341] as
b = 21 (N − N ∗ ), therefore b = 12 (2 − 0) = 1 .
(ii) Be2 : 2 + 2 = 4 VE; 1σ 2g 1σ∗2
u ; b = 2 (2 − 2) = 0 .
1
(iii) C2 : 4 + 4 = 8 VE; 1σ 2g 1σ∗2

u 1π u ; b = 2 (6 − 2) = 2 .
4 1
E9C.5(a) The molecule with the greater bond order is expected to have the larger dis-
sociation energy. Qualitatively B2 and C2 share the same molecular orbital
energy level diagram, shown in Fig. 9C.12 on page 341. B2 has 3 + 3 = 6
valence electrons overall, therefore its ground-state electron configuration is
1σ 2g 1σ∗2
u 1π u . The bond order is defined in [9C.4–341] as b = 2 (N − N ),
2 1 ∗
therefore b = 21 (4 − 2) = 1.
C2 has 4 + 4 = 8 valence electrons, its configuration is 1σ 2g 1σ∗2 4
u 1π u , and the
bond order is b = 21 (6 − 2) = 2. C2 has greater bond order than B2 , therefore
C2 is expected to have the larger bond dissociation energy.
E9C.6(a) The molecule with the greater bond order is expected to have the larger dissoci-
ation energy. The molecular orbital energy level diagram of F2 and F2 + is shown
in Fig. 9C.11 on page 341. F2 has 7 + 7 = 14 valence electrons overall, therefore
the ground-state electron configuration is 1σ 2g 1σ∗2 2 4 ∗4
u 2σ g 1π u 1π g . The bond
∗
order is defined in [9C.4–341] as b = 2 (N − N ), therefore b = 2 (8 − 6) = 1.
1 1
Removing one electron from F2 gives F2 + , which has one fewer electron in the
antibonding π∗g orbital, therefore the bond order is b = 21 (8 − 5) = 32 . F2 + has
greater bond order than F2 , therefore F2 + is expected to have the larger bond
dissociation energy.
P9C.1 (a) Figure 9.3 shows a plot of the overlap integral (Z = 1 is assumed)
1 R 1 R 2 1 R 4
S(2s, 2s) = [1 + ( )+ ( ) + ( ) ] e−R/2a 0
2 a0 12 a 0 240 a 0
1.0
S(2s,2s)
0.8 S(2p,2p)
0.6
S
0.4
0.2
0.0
0 5 10 15 20
R/a 0
Figure 9.3
(b) The overlap integral S(2s, 2s) reaches a value of 0.50 at R/a 0 = 8.03 ; this
value can be read off a graph or found by using mathematical software.
(c) Figure 9.3 shows a plot of the overlap integral (Z = 1 is assumed)
1 R 1 R 2 1 R 3
S(2p, 2p) = [1 + ( )+ ( ) + ( ) ] e−R/2a 0
2 a0 10 a 0 120 a 0
(d) The value of the overlap integral at R/a 0 = 8.03 is
S(2p, 2p) = [1 + 12 ×8.03 + 1
10
×(8.03)2 + 1
120
×(8.03)3 ]×e−8.03/2 = 0.29
P9C.3 Figure 9.4 shows contour plots of the bonding and antibonding 2pσ and 2pπ
molecular orbitals for a representative internuclear distance of R = 6a 0 ; nega-
tive amplitude is indicated by dashed contours, and the locations of the nuclei
are shown by the black dots. Density plots, in which the intensity of the shading
is proportional to the square of the wavefucntion, are shown in Fig. 9.5.
These plots are useful as they identify aspects of the symmetry of the wavefunc-
tions and the positions of nodal planes. In addition, they illustrate that for the
bonding orbitals electron density is accumulating in the internuclear region.
9D Molecular orbital theory: heteronuclear diatomic molecules

D9D.1 The Coulomb integral is essentially the energy of an electron when it occupies
an atomic orbital in the molecule.
2pσg 2pσu
x / a0 4 4
0 0
−4 −4
−12 −8 −4 0 4 8 12 −12 −8 −4 0 4 8 12
z / a0 z / a0
2pπu 2pπg
8 8
4 4
x / a0
0 0
−4 −4
−8 −8
−12 −8 −4 0 4 8 12 −12 −8 −4 0 4 8 12
z / a0 z / a0
Figure 9.4
The resonance integral is a contribution to the energy of a molecule that can be

associated with an electron interacting with more than one nucleus at once.
D9D.3 In forming a bond an atom must, to some extent, give up electron density to
be shared with other atoms in the molecule. The energy needed to do this is
connected with the value of the ionization energy of the orbital. Equally, the
atom will to some extent acquire additional electron density to interact with,
and the energy gained from acquiring this density is connected in some way to
the electron affinity. Thus both electron gain and electron loss, in the loosest
sense, are involved in the process of bonding. It is for this reason that ionization
energy and electron affinity are involved in the estimation of atomic orbital
energies for participation in bonding. See Section 9D.2(a) on page 347.
E9D.1(a) A suitable MO diagram in shown in the solution to Exercise E9D.6(a). The ion
with the greater bond order is expected to have the shorter bond length. NO+
has 5 + 6 − 1 = 10 valence electrons, just enough to completely fill up all the
bonding molecular orbitals, leading to a ground state electron configuration
of 1σ 2 2σ 2 3σ 2 1π 4 . NO− has two more electrons, both accommodated in the
antibonding 2π orbital. It follows that NO+ has a greater bond order than NO− ,
2pσg 2pσu
4 4
x / a0
0 0
−4 −4
−12 −8 −4 0 4 8 12 −12 −8 −4 0 4 8 12
z / a0 z / a0
8 2pπu 8 2pπg
4 4
x / a0
0 0
−4 −4
−8 −8
−12 −8 −4 0 4 8 12 −12 −8 −4 0 4 8 12
z / a0 z / a0
Figure 9.5
therefore NO+ is expected to have the shorter bond length.
E9D.2(a) The relationship between the Pauling and Mulliken electronegativities is given
1/2
by [9D.4–346], χPauling = 1.35χMulliken − 1.37. A plot of the Pauling electroneg-
ativities of Period 2 atoms against the square root of their Mulliken electroneg-
ativities is shown in Fig. 9.6.
1/2
The equation of the best fit line is χPauling = 3.18χMulliken − 2.57, which is very
far from the expected relationship.
E9D.3(a) The orbital energy of an atomic orbital in a given atom is estimated using the
procedure outlined in Brief illustration 9D.2 on page 349, and using data from
the Resource section. The orbital energy of hydrogen is
αH = − 21 [I + Eea ]
(1 eV)
= − 21 ×[(1312.0 kJ mol−1 )+(72.8 kJ mol−1 )]× = −7.18 eV
(96.485 kJ mol−1 )
The conversion factor between kJ mol−1 and eV is taken from inside the front
4.0
3.0
χPauling
2.0
1.0
1.0 1.2 1.4 1.6 1.8 2.0 2.2
1/2
χMulliken
Figure 9.6
cover . Similarly for chlorine
αCl = − 21 [I + Eea ]
(1 eV)
= − 12 ×[(1251.1 kJ mol−1 )+(348.7 kJ mol−1 )]× = −8.29 eV
(96.485 kJ mol−1 )
E9D.4(a) The orbital energies of hydrogen (αH = −7.18 eV) and chlorine (αCl = −8.29 eV)
are calculated in Exercise E9D.3(a). Taking β = −1.0 eV as a typical value and
setting S = 0 for simplicity, substitution into [9D.9c–348] gives
2 1/2
2β
E± = 21 (αH + αCl ) ± 12 (αH − αCl ) [1 + ( ) ]
αH − αCl
= 12 [(−7.18 eV) + (−8.29 eV)]
⎡ 2 ⎤1/2
1
⎢
⎢ (−2.0 eV) ⎥
± 2
eV) − (−8.29 eV)] ⎢1 + (
[(−7.18 ) ⎥⎥
⎢ (−7.18 eV) − (−8.29 eV) ⎥
⎣ ⎦
= (−7.73... eV) ± (1.14... eV)
Therefore the energy of the bonding molecular orbital is E− = (−7.73... eV) −

(1.14... eV) = −8.88 eV , and the antibonding orbital is at an energy level of
E+ = (−7.73... eV) + (1.14... eV) = −6.59 eV .
E9D.5(a) The orbital energies of hydrogen (αH = −7.18 eV) and chlorine (αCl = −8.29 eV)
are calculated in Exercise E9D.3(a). Taking β = −1.0 eV as a typical value, and
setting S = 0.2, substitution into [9D.9a–348] gives
αH + αCl − 2βS ± [(2βS − (αH + αCl ))2 − 4(1 − S 2 )(αH αCl − β 2 )]1/2
E± =
2(1 − S 2 )
(−15.0... eV) ± (1.54... eV)
= = (−7.84... eV) ± (0.803... eV)
(1.92)
Therefore the energy of the bonding molecular orbital is E− = (−7.48... eV) −

(0.803... eV) = −8.65 eV , and the antibonding orbital is at an energy level of
E+ = (−7.48... eV) + (0.803... eV) = −7.05 eV .
E9D.6(a) The molecular orbital energy level diagram for a heteronuclear diatomic AB
is similar to that for a homonuclear diatomic A2 (Fig. 9C.11 on page 341 or
Fig. 9C.12 on page 341) except that the atomic orbitals on A and B are no longer
at the same energies. As a result the molecular orbitals no longer have equal
contributions from the orbitals on A and B; furthermore, it is more likely that
the 2s and 2p orbitals will mix. From simple considerations it it not possible to
predict the exact ordering of the resulting molecular orbitals, so the diagram
shown in Fig. 9.7 is simply one possibility. Note that because the heteronuclear
diatomic no longer has a centre of symmetry the g/u labels are not applicable.
The electronic configurations are: (i) CO (10 valence electrons) 1σ 2 2σ 2 3σ 2 1π 4 ;

(ii) NO (11 valence electrons) 1σ 2 2σ 2 3σ 2 1π 4 2π 1 ; (iii) CN – is isoelectronic
with CO and therefore has the same configuration.
A Molecule B
4σ
2π
2p
2p
1π
3σ
2σ
2s
2s
1σ
Figure 9.7
E9D.7(a) The molecular orbital energy level diagram of the heteronuclear diatomic molecule
XeF is similar to the one shown in the solution to Exercise E9D.6(a) except that
the orbitals on atom A are 5s and 5p. It is not possible to predict the precise
energy ordering of the orbitals from simple considerations, so this diagram is
simply a plausible suggestion.
XeF has 8 + 7 = 15 valence electrons, therefore the ground state electron con-
figuration is 1σ 2 2σ 2 3σ 2 1π 4 2π 4 4σ 1 . The configuration of XeF+ is the same
except that, as there is one fewer electrons, the antibonding 4σ orbital is not
occupied. This means that the bond order in XeF+ (b = 1) is greater than the
bond order in XeF (b = 21 ), therefore XeF+ is likely to have a shorter bond
length than XeF.
P9D.1 (a) A normalized wavefunction satisfies the condition given by [7B.4c–234],
∗ ∗
∫ ψψ dτ = 1. The given wavefunction is real, therefore ψ = ψ .
∗
∫ ψψ dτ = ∫ (ψA cos θ + ψB sin θ) dτ
2
1 1 0
³¹¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
= cos θ ∫ ψA dτ + sin θ ∫ ψB dτ +2 cos θ sin θ ∫ ψA ψB dτ
2 2 2 2
= cos2 θ + sin2 θ = 1
The values of the integrals come from the fact that ψA and ψB are or-
thonormal.
(b) The wavefunction which describes the bonding molecular orbital is formed
by the in-phase interference of the atomic orbitals ψA and ψB , therefore
the coefficients of ψA and ψB must have the same sign. Similarly, the
antibonding orbital is the result of the out-of-phase interference of the
basis atomic orbitals, therefore the corresponding coefficients must have
opposite signs. A plot of the coefficients, cos θ and sin θ as a function of
θ is shown in Fig. 9.8.
1.0 cos θ
sin θ
value of function
0.5
0.0
−0.5
−1.0
0.0 0.2 0.4 0.6 0.8 1.0
θ/π
Figure 9.8
Therefore ψ describes a bonding molecular orbital for 0 < θ < π/2, and
an antibonding molecular orbital for π/2 < θ < π.
P9D.3 The energy of ψA and ψC are kept constant in the following, but the energy of
ψB is progressively lowered.
(a) Taking the energy of the orbital ψB to be −12.0 eV, the secular determi-
nant becomes
RRR (−7.2 eV) − E (−1.0 eV) (−0.8 eV) RRRR
RRR RRR
RRR (−1.0 eV) (−12.0 eV) − E 0
RR
RRR (−0.8 eV)
R 0 (−8.4 eV) − E RRRR
= −E 3 − (27.6 eV)E 2 − (246.04 eV2 )E − (709.68 eV3 )
Setting the polynominal to zero and solving the cubic gives the following
energies: E 1 = −12.2 eV , E 2 = −8.75 eV and E 3 = −6.65 eV . The ma-
trix which diagonalizes the hamiltonian matrix is
⎛ 0.202 0.394 0.897 ⎞

⎜ 0.978 −0.121 −0.168 ⎟
⎝ 0.0425 0.911 −0.409 ⎠
The entries in each column of the matrix above give the coefficients of the
atomic orbitals for the corresponding molecular orbital. Note that E 1 is
close to the energy of ψB , and that in the first column the orbital with the
largest coefficient by far is ψB .
(b) If the energy of ψB is lowered further to −15.0 eV, the secular determinant
becomes
RRR (−7.2 eV) − E (−1.0 eV) (−0.8 eV) RRRR
RRR RRR
RRR (−1.0 eV) (−15.0 eV) − E 0 RR
RRR (−0.8 eV)
R 0 (−8.4 eV) − E RRRR
= −E 3 − (30.6 eV)E 2 − (292.84 eV2 )E − (889.2 eV3 )
The energies are E 1 = −15.1 eV , E 2 = −8.77 eV and E 3 = −6.70 eV , and

the matrix which diagonalizes the hamiltonian is
⎛ 0.127 0.419 0.899 ⎞

⎜ 0.992 −0.0672 −0.108 ⎟
⎝ 0.0151 0.906 −0.424 ⎠
As before E 1 is close in energy to the energy of ψB , and in the first col-

umn the coefficient of that atomic orbital is close to 1. Furthermore, in
columns 2 and 3 the other molecular orbitals are seen to have only small
contributions from ψB .
The interpretation is that as the energy of ψB becomes more and more
separate from the other orbitals, one of the molecular orbitals becomes
very much like ψB and would be classed as non-bonding, and the other
two molecular orbitals have little contribution from ψB .
9E Molecular orbital theory: polyatomic molecules

D9E.1 See Section 9E.3 on page 357.
D9E.3 In ab initio methods an attempt is made to evaluate all integrals that appear
in the secular determinant. Approximations are still employed, but these are
mainly associated with the construction of the wavefunctions involved in the
integrals. In semi-empirical methods, many of the integrals are expressed in
terms of spectroscopic data or physical properties. Semi-empirical methods
exist at several levels. At some levels, in order to simplify the calculations, many
of the integrals are set equal to zero.
Density functional theory (DFT) is different from the Hartree-Fock (HF) self-
consistent field (HF-SCF) methods, the ab initio methods, in that DFT focuses
on the electron density while HF-SCF methods focus on the wavefunction.
However, they both attempt to evaluate integrals from first principles, so DFT
methods are in that sense ab initio methods: both are iterative self-consistent
methods in that the calculations are repeated until the energy and wavefunc-
tions (HF-SCF) or energy and electron density (DFT) are unchanged to within
some acceptable tolerance.
D9E.5 These are all terms originally associated with the Hückel approximation used
in the treatment of conjugated π electron molecules, in which the π electrons
are considered independent of the σ electrons. The π electron binding energy is
the sum of the energies of each π electron in the molecule. The delocalization
energy is the difference in energy of the π electrons between the conjugated
molecule with n π bonds and the energy of n ethene molecules, each of which
has one π bond. The π bond formation energy is the energy released when a
π bond is formed. It is obtained from the total π electron binding energy by
subtracting the contribution from the Coulomb integrals, α.
E9E.1(a) (i) Following the same logic as in Exercise E9E.4(a) and applying the Hückel
approximations as explained there the secular determinant for anthracene
is written as (the numbers in bold refer to the numbering of the carbon
atoms in the molecule)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
1 α−E β 0 0 0 0 0 0 0 0 0 0 0 β
2 β α−E β 0 0 0 0 0 0 0 0 0 0 0
3 0 β α−E β 0 0 0 0 0 0 0 0 0 0
4 0 0 β α−E β 0 0 0 0 0 0 0 0 0
5 0 0 0 β α−E β 0 0 0 0 0 0 0 β
6 0 0 0 0 β α−E β 0 0 0 0 0 0 0
7 0 0 0 0 0 β α−E β 0 0 0 β 0 0
8 0 0 0 0 0 0 β α−E β 0 0 0 0 0
9 0 0 0 0 0 0 0 β α−E β 0 0 0 0
10 0 0 0 0 0 0 0 0 β α−E β 0 0 0
11 0 0 0 0 0 0 0 0 0 β α−E β 0 0
12 0 0 0 0 0 0 β 0 0 0 β α −E β 0
13 0 0 0 0 0 0 0 0 0 0 0 β α−E β
14 β 0 0 0 β 0 0 0 0 0 0 0 β α−E
(ii) Similarly for phenanthrene
1 2 3 4 5 6 7 8 9 10 11 12 13 14
1 α−E β 0 0 0 0 0 0 0 0 0 0 0 β
2 β α−E β 0 0 0 0 0 0 0 0 0 0 0
3 0 β α−E β 0 0 0 0 0 0 0 0 0 0
4 0 0 β α−E β 0 0 0 0 0 0 0 0 0
5 0 0 0 β α−E β 0 0 0 0 0 0 0 β
6 0 0 0 0 β α−E β 0 0 0 0 0 0 0
7 0 0 0 0 0 β α−E β 0 0 0 0 0 0
8 0 0 0 0 0 0 β α−E β 0 0 0 β 0
9 0 0 0 0 0 0 0 β α−E β 0 0 0 0
10 0 0 0 0 0 0 0 0 β α−E β 0 0 0
11 0 0 0 0 0 0 0 0 0 β α−E β 0 0
12 0 0 0 0 0 0 0 0 0 0 β α−E β 0
13 0 0 0 0 0 0 0 β 0 0 0 β α−E β
14 β 0 0 0 β 0 0 0 0 0 0 0 β α−E
E9E.2(a) To calculate the π-electron binding energy of the given systems, it is necessary
to calculate the energies of the occupied molecular orbitals. This is done by di-
agonalising the hamiltonian matrix: the diagonal elements of the resulting ma-
trix are the energies of the molecular orbitals. The hamiltonian matrix has the
same form as the secular matrix except that the diagonal elements are α instead
of α − E. Alternatively, the energies can be found my finding the eigenvalues of
the hamiltonian matrix, or by multiplying out the secular determinant, setting
the resulting polynomial in E to zero and then finding the roots. Mathematical
software is needed for all of these approaches.
The secular determinants are derived in Exercise E9E.1(a), and from these the
form of the hamiltonain matrix is easily found.
(i) The orbital energies for anthracene are E = α + 2.41β, α + 2β, α + 1.41β
(doubly degenerate), α + β (doubly degenerate), α + 0.414β, α − 0.414β,
α − β (doubly degenerate), α − 1.41β (doubly degenerate), α − 2β, α −
2.41β. The π system of anthracene accommodates 14 electrons, therefore
the 7 lowest energy π molecular orbitals are filled. The π-electron binding
energy is therefore E π = 2(α + 2.41β) + 2(α + 2β) + 4(α + 1.41β) + 4(α +
β) + 2(α + 0.414β) = 14α + 19.3β .
(ii) The orbital energies for phenanthrene are E = α + 2.43β, α + 1.95β, α +
1.52β, α +1.31β, α +1.14β, α +0.769β, α +0.605β, α −0.605β, α −0.769β,
α − 1.14β, α − 1.31β, α − 1.52β, α − 1.95β, α − 2.43β. The π system of
anthracene accommodates 14 electrons, therefore the 7 lowest energy π
molecular orbitals are filled. The π-electron binding energy is therefore
E π = 2(α + 2.43β) + 2(α + 1.95β) + 2(α + 1.52β) + 2(α + 1.31β) + 2(α +
1.14β) + 2(α + 0.769β) + 2(α + 0.605β) = 14α + 19.5β
E9E.3(a) The hamiltonian for a single electron in H2 + is given by [9B.1–331]. It has a

kinetic energy term, T̂ = −(ħ 2 /2me )∇21 , and a potential energy term, V̂ . The
species HeH+ has two electrons, therefore the kinetic energy term is written as
ħ2 2 ħ2 2
T̂ = − ∇1 − ∇
2me 2me 2
The energy of interaction between an electron and a nucleus with charge num-
ber Z at distance r is given by −Ze 2 /4πε 0 r. The potential energy operator
consists of terms for each electron interacting with the H nuclues (Z = 1) and
the He nucleus (Z = 2)
e2 1 1 2 2 1
V̂ = − ( + + + − )
4πε 0 r 1H r 2H r 1He r 2He r 12
The first term represent the interaction between electron 1 and the H nucleus,
and the second is for electron 2 with the same nucleus. The third and fourth
terms represent the interactions of the two electrons with the He nucleus. The
last term accounts for the repulsion between the two electrons. The complete
electronic hamiltonian is Ĥ elec = T̂ + V̂ . Because only the electronic hamilto-
nian is required, the repulsion between the two nuclei is not included.
E9E.4(a) (i) Without making the Hückel approximations, the secular determinant of
the H3 molecule is written as
RRR α 1 − E β 12 − S 12 E β 13 − S 13 E RRR
RRR RRR
RRR β 21 − S 21 E α2 − E β 23 − S 23 E RRR
RRR β − S E β 32 − S 32 E α3 − E RRR
R 31 31 R
where α n is the Coulomb integral of the orbital on atom n, β nm is the
resonance integral accounting for the interaction between the orbitals on
atoms n and m, E is the energy of the molecular orbital and S nm is the
overlap integral between the orbtials on atoms n and m.
Within the Hückel approximations the energy of the basis atomic orbitals
is taken to be independent of the position of the corresponding atoms in
the molecule, therefore all Coulomb integrals are set equal to α (given
that there is only one type of basis atomic orbital and only one type of
atom is involved in the problem). Interaction between orbitals on non-
neighbouring atoms is neglected, that is β nm = 0 if atoms n and m are
not neighbouring. All other resonance integrals are set equal to β. The
overlap between atomic orbitals is also neglected, therefore all overlap
integrals S nm with n ≠ m are set to zero. Hence the secular determinant
for linear H3 is
RRR α − E β 0 RRRR
RRR
RRR β α−E β RRRR
R
RRR 0
R β α − E RRRR
(ii) In this case hydrogen atoms 1 and 3 are neighbours, therefore β 13 = β,
and the secular determinant is
RRR α − E β β RRR
RRR RRR
RRR β α−E β RRR
RRR β β α−E RRR
R R
E9E.5(a) The energies of the molecular orbitals in benzene are given by [9E.13–357] as
E = α ± 2β, α ± β, α ± β. Note that α and β are negative quantities, therefore the
a2u molecular orbital is the lowest in energy with energy of α + 2β, as shown in
Fig. 9E.4 on page 357.
(i) The benzene anion has 6 + 1 = 7 electrons in its π system, its electronic
e1g e2u and the total π-electron binding energy is E π =
2 4 1
configuration is a2u
2(α + 2β) + 4(α + β) + (α − β) = 7α + 7β .
(ii) The benzene cation has 6 − 1 = 5 electrons in its π system, its electronic
2 3
configuration is a2u e1g and the total π-electron binding energy is E π =
2(α + 2β) + 3(α + β) = 5α + 7β .
E9E.6(a) The delocalization energy is the energy difference between the π-electron bind-
ing energy E π in the given species and the hypothetical π-electron binding
energy if the given species had isolated π bonds. Therefore the delocalization
energy is given by Edeloc = E π − N π (α + β), where N π is the number of π
electrons. The π-bond formation energy is defined in [9E.12–356] as Ebf =
E π − N π α.
(i) The benzene anion has 7 π electrons and its π-electron binding energy
is calculated in Exercise E9E.5(a) as E π = 7α + 7β. Therefore Edeloc =
(7α + 7β) − 7(α + β) = 0 and Ebf = (7α + 7β) − 7α = 7β .
(ii) The benzene cation has 5 π electrons and its π-electron binding energy
is calculated in Exercise E9E.5(a) as E π = 5α + 7β. Therefore Edeloc =
(5α + 7β) − 5(α + β) = 2β and Ebf = (5α + 7β) − 5α = 7β .
P9E.1 Number the oxygen atoms from 1 to 3 and the carbon atom as number 4. Taking
the Hückel approximations the secular determinant of the carbonate ion is
written as
RRRR αO −E 0 0 β RRRR
RRR R
RRR 0 αO −E 0 β RRRR
= (αO − E)2 [(αO − E)(αC − E) − 3β 2 ]
RRR 0 0 αO −E β RRRR
RRR R
RR β β β αC −E RRRR
Hence the energies of the molecular orbitals of the carbonate ion are the solu-
tions of the equation (αO − E)2 [(αO − E)(αC − E) − 3β 2 ] = 0. The solution
E = αO occurs twice and represents a degenerate pair of orbitals. The other two
roots are the solutions of the equation (αO − E)(αC − E) − 3β 2 = 0, which gives
the energies √
E± = 12 [αO + αC ± (αO − αC )2 + 12β 2 ]
where E− < E+ . Because oxygen is the more electronegative element it is likely
that the three lowest energy molecular orbitlas will be those with E = αO and
E = E− . The carbonate ion has 6 π electrons which will occupy these orbitals
and hence give a π-electron binding energy of
√ √
E π = (αO +αC − (αO − αC )2 + 12β 2 )+4αO = 5αO +αC − (αO − αC )2 + 12β 2
In a localized description of the bonding two electrons occupy a π-bonding

molecular orbital between C and O, and four electrons will be localized on oxy-
gen atoms in orbitals with energy α O . Using the orbital energies from [9D.9c–
348], the total
√ energy of the two electrons in the bonding molecular orbital is
αO + αC + (αO − αC )2 + 4β 2 . Hence, in a localized structure the π-electron
binding energy is
√
E loc = 5αO + αC + (αO − αC )2 + 4β 2
The delocalization energy is given by E π − E loc
Edeloc = E π − E loc
√ √
= [5αO + αC + (αO − αC )2 + 12β 2 ] − [5αO + αC + (αO − αC )2 + 4β 2 ]
√ √
= (αO − αC )2 + 12β 2 − (αO − αC )2 + 4β 2
P9E.3 The secular equations are
(αA − E)cA + (βAB − SAB E)cB + (βAC − SAC E)cC = 0

(βBA − SBA E)cA + (αB − E)cB + (βBC − SBC E)cC = 0
(βCA − SCA E)cA + (βCB − SCB E)cB + (αC − E)cC = 0
In this case, orbitals B and C are on the same atom. It follows that the resonance
integral βBC and the overlap integral SBC are zero, as the atomic orbitals on one
atom are orthogonal to each other. Therefore the secular equations simplify to
(αA − E)cA + (βAB − SAB E)cB + (βAC − SAC E)cC = 0

(βBA − SBA E)cA + (αB − E)cB = 0
(βCA − SCA E)cA + (αC − E)cC = 0
and hence the secular determinant is

RRR
RRR αA − E βAB − SAB E βAC − SAC E RRR
RRR
RRR βBA − SBA E αB − E 0 RRR
RRR β − S E 0 αC − E RRR
R CA CA R
P9E.5 Within the Hückel approximations, the secular determinant of cyclobutadiene

is
RRR α − E β 0 β RRRR
RRRR R
RRR β α − E β 0 RRRR
RRR 0 β α−E β RRRR
RRR R
RR β 0 β α − E RRRR
The hamiltonian matrix is of the same form, but with diagonal elements α
⎛ α β 0 β ⎞ ⎛ 0 1 0 1 ⎞
⎜ β α β 0 ⎟ ⎜ 1 0 1 0 ⎟
H=⎜ ⎟ H = α1 + β ⎜ ⎟
⎜ 0 β α β ⎟ ⎜ 0 1 0 1 ⎟
⎝ β 0 β α ⎠ ⎝ 1 0 1 0 ⎠
As explained in the text, in order to find the energies it is sufficient to diagonal-

ize the matrix on the right, and this is convenient because most mathematical
software packages are only able to diagonalize numerical matrices. The diag-
onal elements in the resulting diagonalized matrix are +2, 0, 0, −2 giving the
energies as α + 2β, α (doubly degenerate), and α − 2β.
Similarly, the secular determinant and hamiltonian matrix for benzene is
RRR α − E β 0 0 0 β RRR ⎛0 1 0 0 0 1⎞
RRR R
RRRR β α − E β 0 0 0 RRRR
R
⎜1
⎜ 0 1 0 0 0⎟
⎟
RRR 0 β α−E β 0 0 RRRR ⎜0 1 0 1 0 0⎟
RRR R H = α1 + β ⎜ ⎟
RRR 0 0 β α−E β 0 RRRR ⎜0
⎜ 0 1 0 1 0⎟
⎟
R ⎜0
RRR 0
RRR 0 0 β α − E β RRRR ⎜ 0 0 1 0 1⎟
⎟
RR β 0 0 0 β α − E RRRR ⎝1 0 0 0 1 0⎠
The diagonal elements in the resulting diagonalized matrix are +2, +1, +1, −1,
−1, and −2, giving the energies α + 2β, α + β (doubly degenerate), α − β (doubly
degenerate), and α − 2β.
The secular determinant and hamiltonian matrix for cyclooctatetraene is
RRR α − E β 0 0 0 0 0 β RRRR
RRR ⎛0 1 0 0 0 0 0 1⎞
RRRR β α − E β 0 0 0 0 0 RRRR ⎜1 0 1 0 0 0 0 0⎟
R ⎜ ⎟
RRR 0 β α−E β 0 0 0 0 RRRR ⎜0 1 0 1 0 0 0 0⎟
⎜ ⎟
RRR R
RRRR 0 0 β α−E β 0 0 0 RRRR ⎜0 0 1 0 1 0 0 0⎟
R H = α1+β ⎜ ⎟
RRR 0 0 0 β α−E β 0 0 RRRR ⎜0 0 0 1 0 1 0 0⎟
⎜ ⎟
RRR R ⎜0 0 0 0 1 0 1 0⎟
RRR 0 0 0 0 β α−E β 0 RRRR ⎜ ⎟
⎜ ⎟
RRR 0 0 0 0 0 β α − E β RRRR ⎜0 0 0 0 0 1 0 1⎟
RRR R
RR β β α − E RRRR
0 0 0 0 0 ⎝1 0 0 0 0 0 1 0⎠
The diagonal elements in the resulting diagonalized matrix are +2.00, +1.41,
+1.41, 0, 0, −1.41, −1.41, and −2, giving energies α + 2β, α + 1.41β (doubly
degenerate), α (doubly degenerate), α − 1.41β (doubly degenerate), and α − 2β.
For all three molecules the lowest and highest energy molecular orbital is not
degenerate, and all the other orbitals occur as degenerate pairs.
P9E.7 (a) Within the Hückel approximations, the secular determinant of the trian-
gular species H3 is
RRR α − E β β RRRR
RRR
RRR β α − E β RRRR
RRR β R
R β α − E RRRR
The hamiltonian matrix is of the same form, but with diagonal elements
α
⎛ α β β ⎞ ⎛ 0 1 1 ⎞
H=⎜ β α β ⎟ H = α1 + β ⎜ 1 0 1 ⎟
⎝ β β α ⎠ ⎝ 1 1 0 ⎠
As explained in the text, in order to find the energies it is sufficient to
diagonalize the matrix on the right, and this is convenient because most
mathematical software packages are only able to diagonalize numerical
matrices. The diagonal elements in the resulting diagonalized matrix are

+2, −1, and −1 giving the energies as α+2β, and α−β (doubly degenerate).
The molecular orbital energy level diagram is shown below.
E =α−β
energy
E = α + 2β
The number of valence electrons (VE) and electron binding energies of

each species are
H3 + 2 VE Etot = 2α + 4β
H3 3 VE Etot = 3α + 3β
H3 − 4 VE Etot = 4α + 2β
(b) Consider the following set of equations
H3 + (g) ÐÐ→ 2 H(g) + H+ (g) ∆ r U −○ (1) = +849 kJ mol−1

H2 (g) ÐÐ→ 2 H(g) ∆ r U −○ (2) = +432.1 kJ mol−1
H+ (g) + H2 (g) ÐÐ→ H3 + (g) ∆ r U −○ (3)
Reaction(3) is reaction(2) − reaction(1), therefore
∆ r U −○ (3) = ∆ r U −○ (2) − ∆ r U −○ (1) = (+432.1 kJ mol−1 ) − (+849 kJ mol−1 )

= −4.16... × 102 kJ mol−1 = −417 kJ mol−1
(c) The change in the total electron binding energy directly gives the change
in the internal energy in the reaction H+ (g)+H2 (g) ÐÐ→ H3 + (g). There-
fore ∆ r U −○ is expressed in terms of the resonance and Coulomb integrals
as
∆ r U −○ = Etot (products) − Etot (reactants)

H3 + H+ H2
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ © ³¹¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
= (2α + 4β) − 0 − 2(α + β) = 2β
Therefore β = (−4.16... × 102 kJ mol−1 )/2 = −208 kJ mol−1 . The electron

binding energies are 2α − 834 kJ mol−1 for H3 + , 3α − 625 kJ mol−1 for H3
and 4α − 416 kJ mol−1 for H3 − .
P9E.9 For H2 the 6-31G* basis set is equivalent to the 6-31G basis set because the star
indicates that the basis set adds d-type polarization functions for each atom
other than hydrogen. Consequently, the basis sets (a) 6-31G* and (b) 6-311+G**
were chosen. Since the calculated energy is with respect to the energy of widely
separated stationary electrons and nuclei, the experimental ground electronic
energy of dihydrogen is calculated as D e + 2I.
H2
6-31G* 6-311+G** experimental
R/pm 73.0 73.5 74.1
ground-state energy, E 0 /eV −30.6626 −30.8167 −32.06
F2
R/pm 134.5 132.9 141.8
ground-state energy, E 0 /eV −5406.30 −5407.92
Both computational basis sets give satisfactory bond length agreement with the
experimental value for H2 . However the 6-31G* basis set is not as accurate as the
larger basis set as illustrated by consideration of both its higher ground-state
energy and the variation principle that the energy of a trial wavefunction is
never less than the true energy. That is, the energy provided by the 6-311+G**
basis set is closer to the true energy.
P9E.11 (a) Linear conjugated polyenes do not have degenerate energy levels, so each
value of k corresponds to a molecular orbital which can be occupied by
up to two electrons. Therefore for ethene, with 2 electrons, the HOMO
has k = 1, and the LUMO has k = 2. The HOMO–LUMO energy gap is
2π π
∆E = (α + 2β cos ) − (α + 2β cos ) = (α − β) − (α + β) = −2β
3 3
This energy gap corresponds to 61500 cm−1 , therefore β = −3.07... ×
104 cm−1 .
Butadiene has 4 electrons, the HOMO has k = 2 and the LUMO has k = 3;
the HOMO–LUMO energy gap is given by
3π 2π
∆E = (α + 2β cos ) − (α + 2β cos ) = −1.23β
5 5
∆E corresponds to 46080 cm−1 , therefore β = −3.72... × 104 cm−1 .
Hexatriene has 6 electrons, the HOMO has k = 3 and the LUMO has
k = 4; the HOMO–LUMO energy gap is given by
4π 3π
∆E = (α + 2β cos ) − (α + 2β cos ) = −0.890β
7 7
Octatetraene has 8 electrons, the HOMO has k = 4 and the LUMO has
k = 5; the HOMO–LUMO energy gap is given by
5π 4π
∆E = (α + 2β cos ) − (α + 2β cos ) = −0.695β
9 9
The average value of β is −4.00... × 104 cm−1 , which in electronvolts is
(−4.00... × 104 cm−1 )×(6.6261 × 10−34 J s)×(2.9979 × 1010 cm s−1 )
β=
(1.6022 × 10−19 J eV−1 )
= −4.96 eV
(b) Octatetraene has 8 π electrons, therefore the orbitals with quantum num-
bers k = 1, 2, 3 and 4 are all fully occupied. Hence the π-electron binding
energy is
π 2π 3π
E π = 2 (α + 2β cos ) + 2 (α + 2β cos ) + 2 (α + 2β cos )
9 9 9
4π
+ 2 (α + 2β cos ) = 8α + 9.52β
9
Therefore the delocalization energy is Edeloc = 8α + 9.52β − 8(α + β) =
1.52β .
(c) The energies and the orbital coefficients are calculated according to the
given formulae and presented in the following tabe; k is the index for the
molecular orbital.
k Ek c k,1 c k,2 c k,3 c k ,4 c k,5 c k,6

1 α + 1.80β 0.232 0.418 0.521 0.521 0.418 0.232
2 α + 1.25β 0.418 0.521 0.232 −0.232 −0.521 −0.418
3 α + 0.445β 0.521 0.232 −0.418 −0.418 0.232 0.521
4 α − 0.445β 0.521 −0.232 −0.418 0.418 0.232 −0.521
5 α − 1.25β 0.418 −0.521 0.232 0.232 −0.521 0.418
6 α − 1.80β 0.232 −0.418 0.521 −0.521 0.418 −0.232
For the lowest energy molecular orbital (k = 1) all the coefficients of the
atomic orbitals are positive, therefore this molecular orbital is bonding
between all pairs of carbon atoms. As the energy of the molecular orbitals
increase, the number of nodes increases as indicated by the number of
sign changes of the coefficients of neighbouring atomic orbitals. In the
highest energy molecular orbital (k = 6) the sign of the neighbouring
coefficients alternates, hence it is a fully antibonding molecular orbital.
I9.1 (a) The calculated and the measured values of the standard enthalpy of for-
mation (∆ f H −○ /kJ mol−1 ) of ethene, butadiene, hexatriene and octatetraene
are shown in the table below, together with the relative error in the calcu-
lated values.
molecule computed experimental % error
C2 H4 69.58 52.46694 32.6
C4 H6 129.8 108.8 ± 0.79 19.3
C6 H8 188.5 168 ± 3 12.2
C8 H10 246.9 295.9 16.6
The experimental values are taken from webbook.nist.gov/chemistry/ and
book Thermodynamic Data of Organic Compounds by Pedley, Naylor and
Kirby.
(b) The % errors shown in the table above are calculated using the expression
∣∆ f H −○ (calc) − ∆ f H −○ (expt)∣
% error =
∆ f H −○ (expt)
(c) For all of the molecules, the computed enthalpies of formation deviate
from the experimental values by much more than the uncertainty in the
experimental value. It is clear that molecular modeling software is not a
substitute for experimentation when it comes to quantitative measures.
I9.3 (a) The energies of the LUMOs of the given molecules are calculated using the
DF/B3LYP/6-31G* method. The results along with the standard reduction
potentials are listed in the table below.
molecule ELUMO /eV E −○ /V

A −3.54 0.078
B −3.39 0.023
C −3.24 −0.067
D −3.11 −0.165
E −3.01 −0.260
The plot of ELUMO against E −○ is shown in Fig. 9.9
−3.0
ELUMO /eV
−3.2
−3.4
−3.6
−0.3 −0.2 −0.1 0.0 0.1
E −○ /V
Figure 9.9
The data points are a moderate fit to a straight line, the equation of which
is
ELUMO /eV = (−1.53) × E −○ /V − 3.38
(b) The energy of the LUMO of this molecule is calculated using the same
method as above as −2.99 eV. Hence the predicted reduction potential is
ELUMO /eV + (3.38...) (−2.99) + (3.38...)

E −○ /V = = = −0.25
(−1.53...) (−1.53...)
(c) The energy of the LUMO of the given molecule is calculated as −3.11 eV.
Hence the predicted reduction potential is
ELUMO /eV + (3.38...) (−3.11) + (3.38...)

E −○ /V = V= = −0.18
(−1.53...) (−1.53...)
Plastoquinone has less negative reduction potential, therefore it is the

better oxidizing agent.
∗
I9.5 The energy of a normalized trial wavefunction Ψtrial is E = ∫ Ψtrial ĤΨtrial dτ.
The hamiltonian operator for the hydrogen atom can be inferred from [8A.4–
285] as
ħ2 e2
Ĥ = − ∇2 −
2µ 4πε 0 r
The Laplacian operator ∇2 is given in Section 7F.2(a) on page 271 as
1 ∂2 1
∇2 = r + 2 Λ2
r ∂r 2 r
but an entirely equivalent form, which is more convenient here, is
∂2 2 ∂ 1
∇2 = + + 2 Λ2
∂r 2 r ∂r r
The legendrian operator Λ 2 contains derivatives with respect to angles only. As
the given trial wavefunction is independent of angles, Λ 2 Ψ = 0 and therefore
the laplacian operating on the trial wavefunction gives
∂2 2 ∂
[Ne−αr ] + [Ne−αr ] + 0 = 4N α 2 r 2 e−αr − 6N αe−αr
2 2 2 2
∇2 Ψtrial =
∂r 2 r ∂r
∗
The wavefunction Ψtrial is real, therefore Ψtrial = Ψtrial . Therefore the hamilto-
nian operating on the trial wavefunction gives
N αħ 2 Ne 2 −αr 2
[2αr 2 e−αr − 3e−αr ] −
2 2
ĤΨtrial = − e
µ 4πε 0 r
Hence the energy of the trial wavefunction is given by
∗
E = ∫ Ψtrial ĤΨtrial dτ
∞ N 2 αħ 2 N 2 e 2 −2αr 2
[2αr 2 e−2αr − 3e−2αr ] −
2 2
=∫ − e dτ
0 µ 4πε 0 r
The volume element dτ in polar coordinates is r 2 sin θ dθ dϕ dr. There is no

angular dependence in the integrand, hence integrating over all angles gives
4π. Thus the integral becomes
∞ αħ 2 e 2 r −2αr 2
[2αr 4 e−2αr − 3r 2 e−2αr ] −
2 2
E = 4πN 2 ∫ − e dr
0 µ 4πε 0
The integral is best evaluated term by term. To evaluate the first term, Integral
G.5 is used from the Resource section to give
∞ 2α 2 ħ 2 4 −2αr 2 3ħ 2 π 1/2
∫ − r e dr = − ( )
0 µ 16µ 2α
Using Integral G.3, the second term gives
∞ 3αħ 2 2 −2αr 2 3ħ 2 π 1/2
∫ r e dr = ( )
0 µ 8µ 2α
The last term is evaluated using Integral G.2
∞ e 2 r −2αr 2 e2
∫ − e dr = −
0 4πε 0 16πεα
Therefore the energy is given by
3ħ 2 π π 1/2 e2
E = N2 [ ( ) − ]
4µ 2α 4ε 0 α
The value of N 2 is found using the normalization condition given by [7B.4c–
∗
234], ∫ Ψtrial Ψtrial dτ = 1. Again, note that in polar coordinates the volume
element dτ is given by r 2 sin θ dθ dϕ dr, and because the wavefunction is spher-
ically symmetric, integration over all angles gives 4π. Therefore the normaliza-
tion condition becomes
∞
r 2 e−2αr dr = 1
2
4πN 2 ∫
0
The integral is evaluated using Integral G.3 to give

1 π 1/2 2α 2α 1/2
4πN 2 ( 3 ) = 1 and therefore N2 = ( )
4 8α π π
Hence the energy corresponding to the trial wavefunction is
3ħ 2 α e2
E= − 1/2 3/2 α 1/2
2µ 2 π ε0
According to the variation principle the minimum energy is obtained by taking
the derivative of the trial energy with respect to the adjustable parameter, which
is in this case α. Setting the derivative equal to zero and then solving the
equation for α gives the value of α which minimises the energy of the trial
wavefunction.
dE 3ħ 2 e2 1
= − 3/2 3/2 =0
dα 2µ 2 π ε 0 α 1/2
Hence α is given by
µ2 e 4
α=
18ħ 4 π 3 ε 20
Substituting this back into the energy expression yields the minimum energy
for this trial wavefunction.
µe 4 µe 4 µe 4
E= − = −
12π 3 ħ 2 ε 20 6π 3 ħ 2 ε 20 12π 3 ħ 2 ε 20
I9.7 (a) All the coefficients of the atomic orbitals for the molecular orbital ψ 1 are
positive, therefore it is a fully bonding molecular orbital. The molecular
orbital ψ 3 has alternating positive, negative and then positive coefficients,
therefore it is a fully antibonding molecular orbital. The molecular orbital
ψ 2 is non-bonding, with no interaction between atomic orbitals on adja-
cent atoms.
(b) The oxygen, carbon and nitrogen atoms all contribute one p orbital (the
2pz atomic orbital) to the π system, therefore they must be sp2 hybridized.
All the σ bonds to an sp2 hybridized atom must lie in the same plane,
therefore all the atoms connected to the oxygen, carbon and nitrogen
atoms must lie in the same plane together with the oxygen, carbon and
the nitrogen atoms.
(c) The molecular orbital energy level diagram is shown in part (a). There
are four electrons in the π system of the peptide link. Two electrons come
from the formal π bond between the oxygen and the carbon atoms, and
two from the formal lone pair on the nitrogen atom. Hence the two low-
est energy molecular orbitals, that is the bonding and the non-bonding
molecular orbitals, are filled.
(d) This picture of bonding in the peptide group does not invoke delocal-
ization. The lowest energy, bonding molecular orbital, ψ 4 , represents a
localized π bond between the oxygen and the carbon atoms. According
to this representation, the next highest energy molecular orbital, ψ 5 , is
the lone pair on nitrogen, a non-bonding molecular orbital. The high-
est energy molecular orbital, ψ 6 is the antibonding π∗ molecular orbital
between the oxygen and the carbon atoms. The four electrons in the π
system occupy the two lowest energy molecular orbitals.
E O
C N ψ6
O ψ5
C N
O ψ4
C N
(e) Assume that the oxygen, carbon and nitrogen atoms are sp2 hybridized.
The p orbital on nitrogen is in the plane in which the oxygen, carbon and
their neighbouring atoms are, therefore the two other atoms connected
to nitrogen must lie in a plane perpendicular to this one. Therefore the
oxygen, carbon, nitrogen and the neighbouring atoms cannot all lie in the
same plane.
(f) The bonding molecular orbital ψ 1 must be lower in energy than ψ 4 , be-
cause ψ 1 extends over three atoms and is bonding between two pairs of
atoms, whereas ψ 4 extends over two atoms only and is bonding between
one pair of atoms only. Following the same logic, ψ 3 must be higher in
energy than ψ 6 . Disregarding the small difference in the Coulomb inte-
gral of nitrogen and oxygen, the nonbonding orbitals have approximately
the same energies.
(g) Only the two lowest energy molecular orbitals are occupied in either case.
Occupation of the non-bonding orbital does not result in any difference
in the π-electron binding energies. The bonding molecular orbital in the
case of the planar peptide link is lower in energy, hence occupation of
this orbital results in greater π-electron binding energy when compared
with the occupation of the bonding molecular orbital in the non-planar
peptide link. Therefore planar geometry is the lower energy conformation
of the peptide link.
10 Molecular symmetry
10A Shape and symmetry

D10A.1 A molecule may be chiral (optically active) only if it does not possess an axis of
improper rotation, S n .
D10A.3
symmetry operation symmetry element

identity, E the entire object
n-fold rotation n-fold axis of symmetry, C n
reflection mirror plane, σ
inversion centre of symmetry, i
n-fold improper rotation n-fold improper rotation axis, S n
E10A.1(a) The molecules to be assigned are shown in Fig. 10.1. For clarity not all symmetry
elements are shown.
C3,S3
C3
σv σv
F H C2',σv
C2,σv σh
C2 F σh
F P C2,i
σv F
N Cl Cl F F C2',σv
O O F Cl
NO2 C 2v PF5 D 3h CHCl3 C 3v 1,4-difluorobenzene D 2h
Figure 10.1
In each case the point group is identified using the flow diagram in Fig. 10A.7 on
page 369 having identified the symmetry elements, or by drawing an analogy
with one of the shapes in the summary table in Fig. 10A.8 on page 369.
346 10 MOLECULAR SYMMETRY
(i) Nitrogen dioxide, NO2 , is a V-shaped molecule shape with a bond angle
of approximately 134○ . It possesses
• the identity, E
• a C 2 axis that lies in the plane of the molecule and passes through
the nitrogen atom.
• two different vertical mirror planes σv , both containing the C 2 axis
but with one lying in the plane of the molecule and the other per-
pendicular to it.
The point group is C 2v .
(ii) PF5 has a trigonal bipyramidal shape with two axial and three equatorial
fluorine atoms. It possesses
• the identity, E
• a C 3 axis passing through the phosphorus and the two axial chlorine
atoms
• a horizontal mirror plane σh perpendicular to the C 3 axis and con-
taining the phosphorus and the three equatorial fluorine atoms
• three C 2′ axes perpendicular to the C 3 axis, each passing through
the phosphorus and one of the equatorial fluorine atoms. Only one
of the C 2′ axes is shown in Fig. 10.1.
• a S 3 axis coincident with the C 3 axis
• three vertical mirror planes σv , each containing the phosphorus, the
two axial fluorine atoms, and one of the three equatorial fluorine
atoms. Only one of the σv mirror planes is shown in Fig. 10.1.
The point group is D 3h
(iii) CHCl3 , chloroform, possesses
• the identity E
• a C 3 axis passing through the hydrogen and the carbon atom
• three vertical mirror planes σv , each containing the hydrogen, car-
bon, and one of the chlorine atoms; only one of these is shown in
Fig. 10.1.
The point group is C 3v
(iv) 1,4-difluorobenzene possesses
• the identity, E
• a C 2 axis perpendicular to the plane of the molecule and passing
through the centre of the benzene ring
• two different C 2′ axes, both perpendicular to the C 2 axis and lying
in the plane of the molecule. One C 2′ axis passes through the two
fluorine atoms, while the other perpendicular to this.
• a horizontal mirror plane σh lying in the plane of the molecule
• two different vertical mirror planes σv , both containing the C 2 axis
and perpendicular to the plane of the molecule. One contains the two
fluorine atoms while the other is perpendicular to the line joining the
two fluorine atoms.
• a centre of inversion i , which lies at the centre of the benzene ring

The point group is D 2h
E10A.2(a) The cis and trans isomers are shown in Fig. 10.2.
C2,σv
σv σh
Cl
Cl Cl C2
Cl
cis isomer C 2v trans isomer C 2h
Figure 10.2
(i) cis-dichloroethene possesses a C 2 axis lying in the plane of the molecule

and bisecting the C=C double bond. It also has two σv mirror planes
containing this axis, one lying in the plane of the molecule and the other
perpendicular to it. The flow diagram in Fig. 10A.7 on page 369 is then
used to establish that the point group is C 2v .
(ii) trans-dichloroethene possesses a C 2 axis that is perpendicular to the plane
of the molecule and passes through the centre of the C=C double bond. It
also has a horizontal mirror plane σh perpendicular to this axis and lying
in the plane of the molecule, and also has a centre of inversion i. The flow
diagram in Fig. 10A.7 on page 369 is then used to establish that the point
group is C 2h .
E10A.3(a) The molecules are shown in Fig. 10.3 along with their point groups. For clarity
not all symmetry elements are shown.
C2,σv C2
σv
H
σ O H H
C2 B
H i B H
N N H
i C∞ C2
O
H H
BeH2 , D∞h (Be not
Pyridine, C 2v Nitroethane, C s B2 H6 , D 2h
shown for clarity)
Figure 10.3
As explained in Section 10A.3(a) on page 372, only molecules belonging to the

groups C n , C nv , and C s may be polar.
(i) Pyridine has point group C 2v , so it may be polar . The dipole must lie
along the C 2 axis, which passes through the nitrogen and the carbon atom
opposite it in the ring.
(ii) Nitroethane, CH3 CH2 NO2 , belongs to point group C s , so it may be polar .
(iii) Linear BeH2 belongs to point group D∞h , so it may not be polar.
(iv) Diborane, B2 H6 , belongs to point group D 2h , so it may not be polar.
E10A.4(a) There are 10 distinct isomers of dichloronaphthalene, shown in Fig. 10.4 to-
gether with their point groups. All isomers have a mirror plane in the plane of
the paper; additional symmetry elements are marked.
Cl Cl Cl
8 Cl
Cl 2 1
7 C2, σv C2
3 6
4 5 Cl Cl Cl
1,2 isomer C s 1,3 isomer C s 1,4 isomer C 2v 1,5 isomer C 2h

Cl Cl
Cl Cl
Cl Cl
C2, σv
Cl C2, σv Cl
1,6 isomer C s 1,7 isomer C s 1,8 isomer C 2v 2,3 isomer C 2v
Cl Cl
Cl
C2
Cl C2, σv
2,6 isomer C 2h 2,7 isomer C 2v
Figure 10.4
E10A.5(a) As explained in Section 10A.3(b) on page 373, to be chiral a molecule must not
possess an axis of improper rotation, S n . This includes mirror planes, which
are the same as S 1 , and centres of inversion, which are the same as S 2 . The
table in Section 10A.2(c) on page 371 shows that a molecule belonging to D 2h
possesses three mirror planes, so such a molecule may not be chiral . Similarly
the table in Section 10A.2(b) on page 370 shows that a molecule belonging to
C 3h possesses a σh mirror plane, so such a molecule may not be chiral .
E10A.6(a) From the table in Section 10A.2(b) on page 370 a molecule belonging to the
point group C 3v , such as chloromethane, possesses
• the identity E
• a C 3 axis passing through the chlorine and carbon atoms
• three vertical mirror planes σv , each containing the chlorine, carbon and
one of the hydrogen atoms
C3
σv
Cl
H H
H
Figure 10.5
The C 3 axis and one of the σv mirror planes is shown in Fig. 10.5
E10A.7(a) The point group of the naphthalene molecule is identified by following the flow
diagram in Fig. 10A.7 on page 369. Firstly, naphthalene is not linear, which
leads to the question “Two or more C n , n > 2?” Naphthalene has three two-fold
axes of rotation but no higher order axes, so the answer to this question is ‘No’.
However the fact that it does have three C 2 axes means that the answer to the
next question “C n ?” is ‘Yes’.
This leads to the question “Select C n with the highest n; then, are there nC 2 per-
pendicular to C n ?” As already identified there are three mutually perpendicular
C 2 axes and therefore no one axis with highest n, that is, no principal axis. As
explained in Section 10A.1 on page 366 in the case of a planar molecule such as
naphthalene with more than one axis competing for the title of principal axis,
it is common to choose the one perpendicular to the plane of the molecule.
However, because all three axes are mutually perpendicular, whichever of the
three is selected it still has two other C 2 axes perpendicular to it, so the answer
to this question is necessarily ‘Yes’.
This leads to the question “σh ?” to which the answer is ‘Yes’ because, whichever
C 2 axis is selected, there is a mirror plane perpendicular to it. In the case of the
C 2 axis perpendicular to the plane of the molecule, the σh mirror plane lies in
the plane of the molecule. Answering ‘Yes’ to this question leads to the result
D nh , and because the C n axis with highest n in this molecule is C 2 , it follows
that the point group is D 2h .
Alternatively, the point group may be identified from the table in Fig. 10A.8 on
page 369 by drawing an analogy between the planar naphthalene molecule and
the rectangle which belongs to D 2h .
From the table in Section 10A.2(c) on page 371 a molecule belonging to the
point group D 2h possesses
• the identity E
• a C 2 axis , which in the case of a planar molecule such as naphthalene is
commonly taken to be the axis perpendicular to the plane of the molecule.
This axis passes through the mid-point of the central bond joining the two
rings.
• two C 2′ axes , perpendicular to the C 2 axis and lying in the plane of the
molecule. One of these passes along the line of the central bond joining
the two rings, while the other bisects this bond.

• a horizontal mirror plane σh in the plane of the molecule
In addition
• The presence of the C 2 axis and the two C 2′ axes jointly imply the presence
of two vertical mirror planes σv , both containing the principal axis. One
of these σv planes also contains the central bond joining the two rings,
while the other bisects this bond.
• The C 2 and σh elements jointly imply a centre of inversion i , which lies
at the midpoint of the central bond.
These symmetry elements are shown in Fig. 10.6.
C2',σv
σh
C2,i C2',σv
Figure 10.6
E10A.8(a) The objects to be assigned are shown in Fig. 10.7. For clarity not all symmetry
elements are shown.
C2,σv
C2,σv
σv C3,S3
C2,σv C2,σv i C∞
Figure 10.7
(i) As explained in Section 10A.2(f) on page 372 a sphere possesses an infinite

number of rotation axes with all possible values of n, and belongs to the
full rotation group R 3 .
(ii) The point group of an isosceles triangle is identified using the flow dia-
gram in Fig. 10A.7 on page 369. It has a C 2 axis lying in the plane of the
paper on which the shape is drawn which bisects the non-equal side and
passes through the vertex opposite this side. There are no other C n axes,
and there is not a σh mirror plane perpendicular to the C 2 axis. There
are, however, two σv mirror planes which contain the C 2 axis, one in the
plane of the paper and the other perpendicular to it. These considerations
establish the point group as C 2v .
(iii) The point group of an equilateral triangle is readily identified as D 3h by
reference to the table of shapes in Fig. 10A.8 on page 369.
(iv) Modelling the unsharpened cylindrical pencil as a cylinder (Fig. 10.7) and
assuming no lettering or other pattern on it, its point group is determined
using the flow diagram in Fig. 10A.7 on page 369. It is linear, in the sense
that rotation by any angle around the long axis of the pencil is a symmetry
operation so that the pencil possesses a C∞ axis. Because the pencil has
not been sharpened, both ends are the same and this means that the pencil
possesses a centre of inversion i. Using the flow diagram this establishes
the point group as D∞h .
P10A.1 The molecules concerned are shown in Fig. 10.8. For clarity not every symmetry
element is shown in each diagram.
(a) C2 σd
H H H H
H H H H
i C3 C3 H
C2
C3,S6 H H H H
σd H H H σd H σd
H
C2
(b) σd C2
C2
C3 C3
i C2
σd σv
σd σd σv
C2 C2
C3,S6
C2
C2 (d)
(c) H2N
H2
H N NH2
C2 H H Co C3
B i B C2
H H N NH2
H H2 H N
C2 2
C2
(e)
C2 C2 σd
σd σd
C2
C2
S S C4 C4 σd
S
S S
S
σd S S
C4,S8
Figure 10.8
(a) The staggered conformation of ethane possesses the following symmetry

elements
• The identity E
• A C 3 axis along the the C–C bond
• Three C 2 axes perpendicular to the C 3 axis. These are best seen in
the Newman projection also shown in Fig. 10.8(a), that is, the view
along the C 3 axis
• Three dihedral mirror planes σd , each containing the C 3 principal
axis and two hydrogen atoms. These mirror planes are ‘dihedral’
rather than ‘vertical’ because they bisect the angles between the C 2
axes. The three σd mirror planes are seen most easily in the Newman
projection; for clarity only one σd plane is shown
• An S 6 axis coincident with the C 3 axis
• A centre of inversion i at the midpoint of the C–C bond.
Using the flow diagram in Fig. 10A.7 on page 369 the point group is D 3d .
(b) The chair conformation of cyclohexane possesses the following symmetry
elements
• The identity E
• A C 3 axis perpendicular to the approximate plane of the molecule
• Three C 2 axis perpendicular to the C 3 axis, each passing through
the midpoint of two opposite C–C bonds
• Three dihedral mirror planes σd , each containing the C 3 axis as well
as two opposite carbon atoms
• A centre of inversion i at the intersection of the C 3 and C 2 axes.
The boat conformation of cyclohexane possesses
• The identity E
• A C 2 axis passing vertically through the centre of the boat where the
mast would be
• Two vertical mirror planes σv , both containing the C 2 axis but one
also containing the bow and stern of the boat and the other perpen-
dicular to this
Using the flow diagram in Fig. 10A.7 on page 369 the point group is C 2v .
(c) Diborane, B2 H6 , possesses
• The identity E
• Three different C 2 axes , all perpendicular to each other. One passes
through the two bridging hydrogen atoms, one passes through the
two boron atoms, and the third is perpendicular to the plane defined
by the boron and bridging hydrogen atoms.
• Three different mirror planes σ , one perpendicular to each of the
three C 2 axes. One mirror plane contains the boron atoms and the
four terminal hydrogen atoms. The second contains the two bridging
hydrogen atoms and forms the perpendicular bisector of a line join-

ing the two boron atoms. The third contains the boron atoms and the
two bridging hydrogen atoms. For clarity these mirror planes are not
shown in Fig. 10.8(c).
• A centre of inversion i at the intersection of the C 2 axes.
Using the flow diagram in Fig. 10A.7 on page 369 the point group is D 2h .
(d) Ignoring the detailed structure of the en ligands, [Co(en)3 ]3+ can be
drawn in a way that resembles a propeller. It possesses the following
symmetry elements
• The identity E
• A C 3 axis corresponding to the axis of the propeller
• Three C 2 axes , each passing through the cobalt and the centre of one
of the en ligands
Using the flow diagram in Fig. 10A.7 on page 369 the point group is D 3 .
(e) Crown-shaped S8 possesses
• The identity E
• A C 4 axis perpendicular to the approximate plane of the molecule
• Four C 2 axes perpendicular to the C 4 axis, analogous to the C 2 axes
in the chair conformation of cyclohexane
• Four dihedral mirror planes σd , again analogous to the σd planes in
the chair conformation of cyclohexane
(i) As explained in Section 10A.3(a) on page 372, only molecules belonging to
the groups C n , C nv , or C s may have a permanent electric dipole moment.
Therefore of the molecules considered only boat cyclohexane (C 2v ) can
be polar.
(ii) As explained in Section 10A.3(b) on page 373, a molecule may be chi-
ral only if it does not possess an axis of improper rotation S n . This in-
cludes mirror planes, which are equivalent to S 1 , and a centre of inversion,
which is equivalent to S 2 . Therefore of the molecules considered only
[Co(en)3 ]3+ is chiral.
P10A.3 The molecules are shown in Fig. 10.9 along with some of their key symmetry
elements.
(a) In addition to the identity element, ethene possesses three C 2 axes, three
mirror planes, and a centre of inversion i. As explained in Section 10A.1
on page 366, in the case of a planar molecule with several C n axes com-
peting for the title of principal axis it is common to choose the axis per-
pendicular to plane of the molecule to be the principal axis. This axis is
labelled C 2 in Fig. 10.9, while the other C 2 axes are labelled C 2′ . The mirror
plane that lies in the plane of the molecule is denoted σh while the other
(a) Ethene
C2',σv
Allene σd
σh C2'
H σd H
C2,i H H
H C2
C C C H H σ
C2',σv C2,S4 H d
H H H H
C2'
(b) (i) 0o (ii) 90o

C2',σv σd C2'
σh σd
C2,i C2
σd
C2',σv C2,S4
C2'
(iii) 45o (iv) 60o
C2' C2'
C2 45o C2 60o
C2' C2'
Figure 10.9
two are denoted σv . Using the flow diagram in Fig. 10A.7 on page 369 the
point group is established as D 2h .
In the case of allene there is a C 2 axis, coincident with an S 4 axis, that
passes through all three carbon atoms. There are also two other C 2 axes,
denoted C 2′ , that are perpendicular to C 2 and which pass through the
central carbon; these are most clearly seen in the Newman projection
shown in Fig. 10.9(a). Finally there are two dihedral mirror planes σd
which bisect the angle between the two C 2′ axes and which each pass
through all three carbon atoms and two of the hydrogen atoms. Using the
flow diagram in Fig. 10A.7 on page 369 these symmetry elements establish
the point group as D 2d .
(b) (i) The biphenyl molecule with a dihedral angle of 0○ , that is, with both
phenyl groups in the same plane, has a shape analogous to that of the
ethene molecule from part (a). Like ethene it belongs to the point
group D 2h and possesses the symmetry elements E, C 2 , 2C 2′ , 2σv ,
σh and i.
(ii) The biphenyl molecule with a dihedral angle of 90○ , that is, with the
two rings perpendicular, has a shape analogous to that of allene. Like
allene it belongs to the point group D 2d and possesses the symmetry
elements E, C 2 , 2C 2′ , 2σd , and S 4 .
(iii) If the dihedral angle is changed from 90○ to 45○ the C 2 and C 2′ axes
are retained but the σd mirror planes and the S 4 axis are no longer
present. Using the flow diagram in Fig. 10A.7 on page 369 the sym-
metry elements establish the point group as D 2 .
(iv) The biphenyl molecule with a dihedral angle of 60○ possesses the
same set of symmetry elements as when the dihedral angle is 45○
and therefore belongs to the same point group of D 2 .
P10A.5 The ion is shown in Fig. 10.10. In each diagram only the mirror plane referred
to in the question is shown; any other mirror planes are omitted for clarity.
C2 C2
σ σv
σh F F
F F CN
i CN F
F3C CF3 i CN
NC
C2 F F
C2 F F F
F C2 F
Structureless CF3 D 2h Staggered CF3 C 2h Eclipsed CF3 C 2v
Figure 10.10
(a) If the CF3 groups are treated as structureless ligands then in addition to
the identity E the ion possesses three C 2 axes, a centre of inversion, and
three mirror planes, only one of which is shown in Fig. 10.10. The point
group is D 2h .
(b) Fig. 10.10 shows the staggered and eclipsed conformations of the CF3 groups.
In the staggered arrangement the centre of inversion i is retained, as is the
C 2 axis that lies along the CN–Ag–CN bond and the mirror plane perpen-
dicular to this, but the other C 2 axes and mirror planes are lost. The point
group is C 2h . In the eclipsed arrangement, the C 2 axis perpendicular to
the plane of the Ag and the four ligands is retained, as are the two mirror
planes containing this axis, only one of which is shown in Fig. 10.10. The
other C 2 axes, the other mirror plane, and the centre of inversion are lost.
The point group is C 2v .
10B Group theory

D10B.1 The top row of the table lists the operations of the group; operations in the
same class are grouped together. Subsequent rows list the characters of each
of the irreducible representations (symmetry species), the symbols for which
are shown if the far left column. Operations in the same class have the same
character, which is why they can conveniently be grouped together.
One-dimensional irreducible representations are labelled A or B, and have char-
acter 1 under the operation E. Two- and three-dimensional irreducible repre-
sentations are labelled E and T, respectively, and have characters 2 and 3 under
the operation E.
On the right of the table various cartesian functions are listed on the row cor-
responding to the irreducible representation as which they transform. Where
a pair (or more) of functions transform as a two (or higher) dimensional rep-
resentation, the functions are bracketed together.
The order of the group, h, is equal to the number of operations. Alternatively,
as the operations are grouped into classes, the order is the sum of the number
of operations belonging to each class.
The number of irreducible representations equals the number of classes. Also,
apart from the groups C n with n > 2, the sum of the squares of the dimensions
of each irreducible representation is equal to the order.
D10B.3 The letters and subscripts of a symmetry species provide information about
the symmetry behaviour of the species. An A or a B is used to denote a one-
dimensional representation; A is used if the character under the principal ro-
tation is +1 (symmetric behaviour), and B is used if the character is −1 (anti-
symmetric behaviour). Subscripts are used to distinguish the irreducible rep-
resentations if there is more than one of the same type: A1 is reserved for the
representation with the character 1 under all operations.
E denotes a two-dimensional irreducible representation, and T a three dimen-
sional irreducible representation. For groups with an inversion centre, a sub-
script g (gerade) indicates symmetric behaviour under the inversion operation;
a subscript u (ungerade) indicates antisymmetric behaviour. If a horizontal
mirror plane is present, ′ or ′′ superscripts are added if the behaviour is sym-
metric or antisymmetric, respectively, under the σ h operation.
D10B.5 A representative is a mathematical operator, usually taking the form of a matrix,

that represents the effect of a particular symmetry operation. It is constructed
by considering the effect of the operation on a particular set of basis functions.
The set of all these mathematical operators corresponding to all the operations
of the group is called a representation.
E10B.1(a) BF3 belongs to the point group D 3h . The σh operation corresponds to a re-
flection in the plane of the molecule, while the C 3 operation corresponds to
rotation by 120○ around the C 3 axis which passes through the boron atom and
is perpendicular to the plane of the molecule (Fig. 10.11).
Using the orbital numbering shown in Fig. 10.11, the matrix representatives
for the σh and C 3 operations were found in Exercise E10B.7(a) and Exercise
E10B.7(b) to be
⎛ −1 0 0 0 ⎞ ⎛1 0 0 0⎞
⎜ 0 −1 0 0 ⎟ ⎜0 0 0 1⎟
D(σh ) = ⎜ ⎟ and D(C 3 ) = ⎜ ⎟
⎜ 0 0 −1 0 ⎟ ⎜0 1 0 0⎟
⎝ 0 0 0 −1 ⎠ ⎝0 0 1 0⎠
C3
p1
p3 p2
σh
p4
Figure 10.11
The matrix representative of the operation σh C 3 is found by multiplying the

matrix representatives of σh and C 3 . Basic information about how to handle
matrices is given in The chemist’s toolkit 24 in Topic 9E on page 353.
⎛ −1 0 0 0 ⎞⎛ 1 0 0 0⎞ ⎛ −1 0 0 0 ⎞
⎜ 0 −1 0 0 ⎟⎜ 0 0 0 1⎟ ⎜ 0 0 0 −1 ⎟
D(σh )D(C 3 ) = ⎜ ⎟⎜ ⎟= ⎜ ⎟
⎜ 0 0 −1 0 ⎟⎜ 0 1 0 0⎟ ⎜ 0 −1 0 0 ⎟
⎝ 0 0 0 −1 ⎠ ⎝ 0 0 1 0⎠ ⎝ 0 0 −1 0 ⎠
The operation corresponding to this representative is found by considering its

effect on the starting basis
⎛ −1 0 0 0 ⎞
⎜ 0 0 0 −1 ⎟
( p1 p2 p3 p4 ) ⎜ ⎟ = ( −p1 −p3 −p4 −p2 )
⎜ 0 −1 0 0 ⎟
⎝ 0 0 −1 0 ⎠
The operation σh C 3 therefore changes the sign of p1 , and converts p2 , p3 and
p4 into −p3 , −p4 , and −p2 respectively. This is precisely the same outcome as
achieved by the S 3 operation , that is, a 120○ rotation around the C 3 axis fol-
lowed by a reflection in the σh plane. Thus, D(σh )D(C 3 ) = D(S 3 ), as expected
from the fact that by definition the S 3 operation corresponds to a C 3 rotation
followed by a reflection in the plane perpendicular to the C 3 axis.
E10B.2(a) Fig. 10.12 shows BF3 , an example of a molecule with D 3h symmetry. The three
C 2 axes are labelled C 2 , C 2′ and C 2′′ , and likewise for the σv mirror planes.
C2',σv'
C2',σv'
C2',σv'
4 F 1
2 F 3 F
2
1 3 4
F F F F F F
C2'',σv'' 3 C2,σv C2'',σv'' 4 C2,σv C2'',σv'' C2,σv
2 1
σv′′−1 C 2 σv′′ = C 2′ σv−1 C 2′ σv = C 2′′ σv′−1 C 2′′ σv′ = C 2
Figure 10.12
The criteria for two operations R and R ′ to be in the same class is given by
[10B.1–376], R′ = S −1 RS where S is another operation in the group. The task is
therefore to find an operation S in D 3h such that this equation is satisfied when

R and R ′ are two of the C 2 axes.
Referring to the first diagram in Fig. 10.12, to show that C 2 and C 2′ are in the
same class consider the operation σv′′−1 C 2 σv′′ . Start at the arbitrary point 1, and
recall that the operations are applied starting from the right. The operation
σv′′ moves the point to 2, and then C 2 moves the point to 3. The inverse of a
reflection is itself, σv′′−1 = σv′′ , so the effect of σv′′−1 is to move the point to 4.
From the diagram it can be seen that 4 can be reached by applying C 2′ to point
1, thus demonstrating that σv′′−1 C 2 σv′′ = C 2′ and hence that C 2 and C 2′ belong to
the same class.
In a similar way the second diagram in Fig. 10.12 shows that σv−1 C 2′ σv = C 2′′ and
hence that C 2′ and C 2′′ belong to the same class, while the third diagram shows
that C 2′′ and C 2 belong to the same class.
E10B.3(a) The orthonormality of irreducible representations is defined by [10B.7–380],
1 Γ (i) Γ ( j) 0 for i ≠ j
∑ N(C)χ (C)χ (C) = {
h C 1 for i = j
where the sum is over all classes of the group, N(C) is the number of operations
(i)
in class C, and h is the number of operations in the group (its order). χ Γ (C)
(i)
is the character of class C in irreducible representation Γ , and similarly for
( j)
χ Γ (C).
The character table for the point group C 2h is available in the Online resource
centre. The operations of the group are {E, C 2 , i, σh }, so h = 4. There are four
classes and in this group each class has just one member, so all N(C) = 1. The
irreducible representations have the following characters
Ag {1, 1, 1, 1} Bg {1, −1, 1, −1}

Au {1, 1, −1, −1} Bu {1, −1, −1, 1}
To prove orthogonality, [10B.7–380] is used with each pair of irreducible rep-

resentations; the result is zero in each case which shows that each pair are
orthogonal
1
Ag and Bg 4
{1×1×1 + 1×1×(−1) + 1×1×1 + 1×1×(−1)} = 0
1
Ag and Au 4
{1×1×1 + 1×1×1 + 1×1×(−1) + 1×1×(−1)} = 0
1
Ag and Bu 4
{1×1×1 + 1×1×(−1) + 1×1×(−1) + 1×1×1} = 0
1
Bg and Au 4
{1×1×1 + 1×(−1)×1 + 1×1×(−1) + 1×(−1)×(−1)} = 0
1
Bg and Bu 4
{1×1×1 + 1×(−1)×(−1) + 1×1×(−1) + 1×(−1)×1} = 0
1
Au and Bu 4
{1×1×1 + 1×1×(−1) + 1×(−1)×(−1) + 1×(−1)×1} = 0
To prove that each irreducible representation is normalised, [10B.7–380] is used

with i = j for each irreducible representation Γ(i) ; the result is 1 in each case.
1
Ag and Ag 4
{1×1×1 + 1×1×1 + 1×1×1 + 1×1×1} = 1
1
Bg and Bg 4
{1×1×1 + 1×(−1)×(−1) + 1×1×1 + 1×(−1)×(−1)} = 1
1
Au and Au 4
{1×1×1 + 1×1×1 + 1×(−1)×(−1) + 1×(−1)×(−1)} = 1
1
Bu and Bu 4
{1×1×1 + 1×(−1)×(−1) + 1×(−1)×(−1) + 1×1×1} = 1
E10B.4(a) The D 3h character table is given in the Resource section. As explained in Sec-
tion 10B.3(a) on page 380, the symmetry species of s, p and d orbitals on a
central atom are indicated at the right hand side of the character table. The
position of z in the D 3h character table shows that pz , which is proportional
to z f (r), has symmetry species A′′2 . Similarly the positions of x and y in
the character table shows that px and p y , which are proportional to x f (r)
and y f (r) respectively, jointly span the irreducible representation of symmetry
species E′ .
In the same way the positions of z 2 , x 2 − y 2 , x y, yz and zx in the character table
show that dz 2 has symmetry species A′1 , dx 2 −y 2 and dx y jointly span E′ , and
d yz and dzx jointly span E′ .
E10B.5(a) As explained in Section 10B.3(c) on page 382, the highest dimensionality of
irreducible representations in a group is equal to the maximum degree of de-
generacy in the group. The highest dimensionality is found by noting the max-
imum value of χ(E) in the character table. An octahedral hole in a crystal
has O h symmetry, and from the O h character table in the Resource section the
maximum value of χ(E) is three, corresponding to the T irreducible represen-
tations. Therefore the maximum degeneracy of a particle in an octahedral hole
is three .
E10B.6(a) As explained in Section 10B.3(c) on page 382, the highest dimensionality of
any irreducible representation in a group is equal to the maximum degree of
degeneracy in the group. The highest dimensionality is found by noting the
maximum value of χ(E) in the character table. Benzene has D 6h symmetry, the
character table for which is available in the Online resource centre. The maxi-
mum value of χ(E) is two, corresponding to the E irreducible representations.
Therefore the maximum degeneracy of the orbitals in benzene is two .
E10B.7(a) BF3 belongs to the point group D 3h . The σh operation corresponds to reflection
in the plane of the molecule.
z z
C3 C3
p1
p3 p2 σh
σh σh
p4
Figure 10.13
As shown in Fig. 10.13 the effect of this operation is to change the sign of all
four p orbitals. This effect is written as ( −p1 −p2 −p3 −p4 ) ← ( p1 p2 p3 p4 )
which is expressed using matrix multiplication as

D(σ h )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ −1 0 0 0 ⎞
⎜ 0 −1 0 0 ⎟
( −p1 −p2 −p3 −p4 ) = ( p1 p2 p3 p4 ) ⎜ ⎟ = ( p1 p2 p3 p4 )D(σh )
⎜ 0 0 −1 0 ⎟
⎝ 0 0 0 −1 ⎠
The representative of this operation is therefore
⎛ −1 0 0 0 ⎞
⎜ 0 −1 0 0 ⎟
D(σh ) = ⎜ ⎟
⎜ 0 0 −1 0 ⎟
⎝ 0 0 0 −1 ⎠
P10B.1 Fig. 10.14 shows trans-difluoroethene, an example of a molecule with C 2h sym-
metry.
C2
1
2
H i F
σh F H
Figure 10.14
The group multiplication table is constructed by considering the effect of suc-

cessive transformations. For example, Fig. 10.14 shows the effect on an arbitrary
point of a C 2 rotation followed by a σh reflection. The C 2 operation moves
the point from 1 to 2, and the σh reflection moves the point to 3. The overall
effect, 1 → 3, is equivalent to carrying out the i operation. Thus, σh C 2 = i.
Considering all other pairs of operations in the same way gives the following
table.
R ↓ R′ → E C2 σh i
E E C2 σh i
C2 C2 E i σh
σh σh i E C2
i i σh C2 E
P10B.3 Fig. 10.15 shows that a water molecule together with the specified axis system
and the six basis orbitals, which are labelled sA , sB , sO , px , p y , and pz . The σv
plane is the xz plane, and the σv′ plane is the yz plane.
z z z
C2 pz
sO
O y y py y
x HA HB x x px
sA sB
Figure 10.15
The matrix representatives are obtained using the approach described in Sec-
tion 10B.2(a) on page 376. Beginning with C 2 , this operation exchanges sA and
sB , leaves sO and pz unchanged, and changes the sign of px and of p y . Its effect
is written ( sB sA sO −px −p y pz ) ← ( sA sB sO px p y pz ) which is expressed
using matrix multiplication as
D(C 2 )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 1 0 0 0 0 ⎞
⎜ 1 0 0 0 0 0 ⎟
⎜ ⎟
⎜ 0 0 1 0 0 0 ⎟
( sB sA sO −px −p y pz ) = ( sA sB sO px p y pz )⎜
⎜ 0 0 0 −1 0 0 ⎟
⎟
⎜ ⎟
⎜ 0 0 0 0 −1 0 ⎟
⎜ ⎟
⎝ 0 0 0 0 0 1 ⎠
Similarly, the σv operation exchanges sA and sB , leaves sO , px and pz unchanged,

and changes the sign of p y : ( sB sA sO px −p y pz ) ← ( sA sB sO px p y pz ) This
is expressed using matrix multiplication as
D(σ v )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
⎛ 0 1 0 0 0 0 ⎞
⎜ 1 0 0 0 0 0 ⎟
⎜ ⎟
⎜ 0 0 1 0 0 0 ⎟
( sB sA sO px −p y pz ) = ( sA sB sO px p y pz )⎜
⎜ 0 0 0 1 0 0 ⎟
⎟
⎜ ⎟
⎜ 0 0 0 0 −1 0 ⎟
⎜ ⎟
⎝ 0 0 0 0 0 1 ⎠
The σv′ operation leaves all orbitals unchanged except px , which changes sign.
D(σ v′ )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
⎛ 1 0 0 0 0 0 ⎞
⎜ 0 1 0 0 0 0 ⎟
⎜ ⎟
⎜ 0 0 1 0 0 0 ⎟
( sA sB sO −px p y pz ) = ( sA sB sO px p y pz )⎜
⎜ 0 0 0 −1 0 0 ⎟
⎟
⎜ ⎟
⎜ 0 0 0 0 1 0 ⎟
⎜ ⎟
⎝ 0 0 0 0 0 1 ⎠
Finally, the E operation, ‘do nothing’, leaves all orbitals unchanged.
D(E)
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 1 0 0 0 0 0 ⎞
⎜ 0 1 0 0 0 0 ⎟
⎜ ⎟
⎜ 0 0 1 0 0 0 ⎟
( s A s B s O p x p y p z ) = ( s A s B s O p x p y p z )⎜
⎜ 0 0 0 1 0 0 ⎟
⎟
⎜ ⎟
⎜ 0 0 0 0 1 0 ⎟
⎜ ⎟
⎝ 0 0 0 0 0 1 ⎠
(a) The representative of the operation C 2 σv is found by multiplying together

the matricies D(C 2 ) and D(σv ) found above. Basic information about
how to handle matrices is given in The chemist’s toolkit 24 in Topic 9E on
page 353.
D(C 2 ) D(σ v ) D(σ v′ )

³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
⎛ 0 1 0 0 0 0 ⎞⎛ 0 1 0 0 0 0 ⎞ ⎛ 1 0 0 0 0 0 ⎞
⎜ 1 0 0 0 0 0 ⎟⎜ 1 0 0 0 0 0 ⎟ ⎜ 0 1 0 0 0 0 ⎟
⎜ ⎟⎜ ⎟ ⎜ ⎟
⎜ 0 0 1 0 0 0 ⎟⎜ 0 0 1 0 0 0 ⎟ ⎜ 0 0 1 0 0 0 ⎟
D(C 2 )D(σv ) = ⎜
⎜ 0 0 0 −1 0 0 ⎟ ⎜ 0 0 0 1 0 0 ⎟ = ⎜ 0 0 0 −1 0 0 ⎟
⎟⎜ ⎟ ⎜ ⎟
⎜ ⎟⎜ ⎟ ⎜ ⎟
⎜ 0 0 0 0 −1 0 ⎟ ⎜ 0 0 0 0 −1 0 ⎟ ⎜ 0 0 0 0 1 0 ⎟
⎜ ⎟⎜ ⎟ ⎜ ⎟
⎝ 0 0 0 0 0 1 ⎠⎝ 0 0 0 0 0 1 ⎠ ⎝ 0 0 0 0 0 1 ⎠
= D(σv′ )
This establishes that C 2 σv = σv′ . Similarly, the representative of the oper-

ation σv σv′ is found by representative matricies of σv and σv ’; the result is
the representative of C 2 , establishing that σv σv′ = C 2 .
D(σ v ) D(σ v′ ) D(C 2 )

³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 1 0 0 0 0 ⎞⎛ 1 0 0 0 0 0 ⎞ ⎛ 0 1 0 0 0 0 ⎞
⎜ 1 0 0 0 0 0 ⎟⎜ 0 1 0 0 0 0 ⎟ ⎜ 1 0 0 0 0 0 ⎟
⎜ ⎟⎜ ⎟ ⎜ ⎟
′
⎜ 0 0 1 0 0 0 ⎟⎜ 0 0 1 0 0 0 ⎟ ⎜ 0 0 1 0 0 0 ⎟
D(σv )D(σv ) = ⎜
⎜ 0 0 0 1 0 0 ⎟ ⎜ 0 0 0 −1 0 0 ⎟ ⎜ 0 0 0 −1 0 0 ⎟
⎟⎜ ⎟=⎜ ⎟
⎜ ⎟⎜ ⎟ ⎜ ⎟
⎜ 0 0 0 0 −1 0 ⎟ ⎜ 0 0 0 0 1 0 ⎟ ⎜ 0 0 0 0 −1 0 ⎟
⎜ ⎟⎜ ⎟ ⎜ ⎟
⎝ 0 0 0 0 0 1 ⎠⎝ 0 0 0 0 0 1 ⎠ ⎝ 0 0 0 0 0 1 ⎠
= D(C 2 )
(b) The representation is reduced using the method in Section 10B.2(c) on

page 378. Inspection of the matrix representatives reveales that they are
all of block-diagonal format
⎛ 0 0 0 0 ⎞
⎜
⎜ 0 0 0 0 ⎟
⎟
⎜ 0 0 0 0 0 ⎟
D=⎜
⎜
⎟
⎟
⎜ 0 0 0 0 0 ⎟
⎜ 0 0 0 0 0 ⎟
⎜ ⎟
⎝ 0 0 0 0 0 ⎠
This implies that the symmetry operations of C 2v never mix the sO , px ,

p y and pz orbitals together, nor do they mix these orbitals with sA and sB ,
but sA and sB are mixed together by the operations of the group. Conse-
quently the basis can be cut into five parts: four one-dimensional bases
for the individual orbitals on the oxygen and a two-dimensional basis for
the two hydrogen s orbitals. The representations corresponding to the
four one-dimensional bases are
For sO : D(E) = 1 D(C 2 ) = 1 D(σv ) = 1 D(σv′ ) = 1
For px : D(E) = 1 D(C 2 ) = −1 D(σv ) = 1 D(σv′ ) = −1
For p y : D(E) = 1 D(C 2 ) = −1 D(σv ) = −1 D(σv′ ) = 1
For pz : D(E) = 1 D(C 2 ) = 1 D(σv ) = 1 D(σv′ ) = 1
The characters of one-dimensional representatives are just the represen-
tatives themselves. Therefore, inspection of the C 2v character table from
the Resource section indicates that the one-dimensional representations
for these orbitals correspond to A1 , B1 , B2 , and A1 respectively. The
oxygen-based orbitals therefore span 2A1 + B1 + B2 .
The remaining two orbitals, sA and sB , are a basis for a two-dimensional
representation
⎛1 0⎞ ⎛0 1⎞ ⎛0 1⎞ ⎛1 0⎞
D(E) = D(C 2 ) = D(σv ) = D(σv′ ) =
⎝0 1⎠ ⎝1 0⎠ ⎝1 0⎠ ⎝0 1⎠
where the matrices correspond to the top left-hand corners of the 6 ×
6 matricies found above. That this two-dimensional representation is
reducible is demonstrated by considering the linear combinations s1 =
sA + sB and s2 = sA − sB . The C 2 and σv operations both exchange sA and
sB : ( sB sA ) ← ( sA sB ) which means that (sB + sA ) ← (sA + sB ), corre-
sponding to (s1 ) ← (s1 ). Similarly, (sB − sA ) ← (sA − sB ), corresponding
to (−s2 ) ← (s2 ). On the other hand, the E and σv′ operations leave sA and
sB unchanged, which leads to the results (s1 ) ← (s1 ) and (s2 ) ← (s2 ).
The representation in the basis ( s1 s2 ) is therefore
⎛1 0⎞ ⎛1 0 ⎞ ⎛1 0 ⎞ ⎛1 0⎞
D(E) = D(C 2 ) = D(σv ) = D(σv′ ) =
⎝0 1⎠ ⎝ 0 −1 ⎠ ⎝ 0 −1 ⎠ ⎝0 1⎠
The new representatives are all in block diagonal format, in this case in
0
the form ( ), which means that the two combinations are not mixed
0
with each other by any operation of the group. The two dimensional basis
( s1 s2 ) can therefore be cut into two one-dimensional bases correspond-
ing to s1 and s2 . The representations for these bases are
For s1 : D(E) = 1 D(C 2 ) = 1 D(σv ) = 1 D(σv′ ) = 1
For s2 : D(E) = 1 D(C 2 ) = −1 D(σv ) = −1 D(σv′ ) = 1
Because the characters of one-dimensional representatives are just the
representatives themselves, inspection of the C 2v character table imme-
diately indicates that these one-dimensional representations correspond
to A1 and B2 respectively. Combining these results with those from the

oxygen orbitals found earlier, which span 2A1 + B1 + B2 , the original six-
dimensional representation therefore spans 3A1 + B1 + 2B2 .
P10B.5 The ethene molecule and axis system are shown in Fig. 10.16.
y
y
C2,σyz
σxy
sB H H s
C2z A
x
x
sC sD C2,σzx
H H
Figure 10.16
The C 2x operation exchanges sA with sD , and sB with sC :
( sD sC sB sA ) ← ( sA sB sC sD )
This is written using matrix multiplication as

D(C 2x )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 0 0 1 ⎞
⎜ 0 0 1 0 ⎟
( s D s C s B s A ) = ( s A s B s C s D )⎜ ⎟
⎜ 0 1 0 0 ⎟
⎝ 1 0 0 0 ⎠
The matrix D(C 2x ) is the representative of the C 2x operation in this basis. The
representatives of the other seven operations are obtained in the same way.
y
D(C 2 )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 1 0 0 ⎞
y ⎜ 1 0 0 0 ⎟
For C 2 : ( sB sA sD sC ) = ( sA sB sC sD )⎜ ⎟
⎜ 0 0 0 1 ⎟
⎝ 0 0 1 0 ⎠
D(C 2z )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 0 1 0 ⎞
⎜ 0 0 0 1 ⎟
For C 2z : ( s C s D s A s B ) = ( s A s B s C s D )⎜ ⎟
⎜ 1 0 0 0 ⎟
⎝ 0 1 0 0 ⎠
D(i)
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 0 1 0 ⎞
⎜ 0 0 0 1 ⎟
For i: ( s C s D s A s B ) = ( s A s B s C s D )⎜ ⎟
⎜ 1 0 0 0 ⎟
⎝ 0 1 0 0 ⎠
D(σ x y )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 1 0 0 0 ⎞
⎜ 0 1 0 0 ⎟
For σ x y : ( sA sB sC sD ) = ( sA sB sC sD )⎜ ⎟
⎜ 0 0 1 0 ⎟
⎝ 0 0 0 1 ⎠
D(σ yz )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 1 0 0 ⎞
⎜ 1 0 0 0 ⎟
For σ yz : ( sB sA sD sC ) = ( sA sB sC sD )⎜ ⎟
⎜ 0 0 0 1 ⎟
⎝ 0 0 1 0 ⎠
D(σ zx )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 0 0 1 ⎞
⎜ 0 0 1 0 ⎟
For σ zx : ( sD sC sB sA ) = ( sA sB sC sD )⎜ ⎟
⎜ 0 1 0 0 ⎟
⎝ 1 0 0 0 ⎠
D(E)
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 0 1 0 ⎞
⎜ 0 0 0 1 ⎟
For E: ( s C s D s A s B ) = ( s A s B s C s D )⎜ ⎟
⎜ 1 0 0 0 ⎟
⎝ 0 1 0 0 ⎠
P10B.7 For one-dimensional representatives, the characters as just the representatives

themselves. Thus for the first representation, D(C 3 ) = 1 and D(C 2 ) = 1, the
characters are χ(C 3 ) = 1 and χ(C 2 ) = 1. Inspection of the C 6v character table
in the Resource section shows that this combination of characters corresponds
to either A1 or A2 . Both of these irreducible representations have χ(C 6 ) = +1,
as expected from the matrix multiplication D(C 6 ) = D(C 3 )D(C 2 ) = 1 × 1 = 1
which implies that χ(C 6 ) = +1. The character table gives the characters of
σv and σd as both +1 for A1 or both −1 for A2 , implying that D(σv ) = +1 ,
D(σd ) = +1 in the case of A1 , or D(σv ) = −1, D(σd ) = −1 in the case of A2 .
For the second representation, D(C 3 ) = 1 and D(C 2 ) = −1, the characters are
χ(C 3 ) = 1 and χ(C 2 ) = −1. Inspection of the character table shows that this
corresponds to either B1 or B2 ; as expected these both have χ(C 6 ) = −1. The
characters of σv and σd are +1 and −1 in the case of B1 , or −1 and +1 in the
case of B2 . Hence D(σv ) = +1, D(σd ) = −1 in the case of B1 , or D(σv ) = −1 ,
D(σd ) = +1 in the case of B2 .
P10B.9 (a) The C 2v character table is given in the Resource section. As explained in
the question, r 2 is invariant to all operations of the group, and further-
more the C 2v character table shows that both z and z 2 belong to the totally
symmetric symmetric representation A1 which means that they are also
invariant to all operations of the group. Because all parts of the func-
tion z(5z 2 − 3r 2 ) are invariant to all operations, it follows that the entire
function and therefore the f orbital that it represents is invariant to all
operations. Consequently this f orbital belongs to the totally symmetric

representation A1 .
(b) The function y(5y 2 − 3r 2 ) is considered as a product of the functions
y and (5y 2 − 3r 2 ). The C 2v character table shows that the function y 2
belongs to the totally symmetric representation and is therefore invariant
to all operations of the group; because this is also true of r 2 it follows that
the factor (5y 2 −3r 2 ) is similarly invariant to all operations. The character
table also shows that the function y belongs to the B2 representation, for
which the characters are +1 under E and σv′ and −1 under C 2 and σv . This
indicates that the function y changes sign under C 2 and σv and, therefore,
because (5y 2 −3r 2 ) is invariant to all operations, the product y(5y 2 −3r 2 )
behaves in the same way as y. The f orbital therefore belongs to the B2
representation.
(c) In the same way, the function x(5x 2 − 3r 2 ) behaves in the same way as
x, because (5x 2 − 3r 2 ) is invariant to all operations. The character table
shows that x, and therefore this f orbital, belongs to the B1 representa-
tion.
(d) The function z(x 2 − y 2 ) belongs to the totally symmetric A1 representa-
tion, because as shown in the character table each of z, x 2 and y 2 belongs
to A1 and are therefore invariant to all operations of the group. It follows
that the product z(x 2 − y 2 ) is also invariant to all operations and hence
belongs to A1 .
(e) The function y(x 2 − z 2 ) behaves in the same way as y, because the factor
(x 2 − z 2 ) is invariant to all operations of the group. Hence this f orbital
belongs to B2 , the same as y.
(f) Similarly the function x(z 2 − y 2 ) behaves the same way as x and therefore
belongs to B1 .
(g) The function x yz is considered as the product x y × z. The character
table shows that z belongs to the totally symmetric representation A1 , so
the function x yz behaves in the same way as x y which as shown in the
character table belongs to A2 .
The seven f orbitals therefore span 2A1 + A2 + 2B1 + 2B2 .
10C Applications of symmetry

D10C.1 The key point is that only orbitals (or combinations of orbitals) which transform
as the same symmetry species (irreducible representation) can overlap to form
molecular orbitals. The first step is therefore to classify the valence orbitals
according to symmetry. Usually, it is possible to identify sets of such orbitals
which are interconverted by the operations of the group and so can be consid-
ered separately from other sets. Having classified a set of orbitals according to
symmetry the projection operator is then used to construct symmetry-adapted
linear combinations (SALCs) each of which transforms as a single symmetry
species (irreducible representation). The molecular orbitals are formed from

the overlap of these SALCs.
E10C.1(a) The D 2h character table shows that x y spans B1g in D 2h . To show this explicitly,
the direct product is formed between the irreducible representations spanned
by x and y individually. The character table shows that x spans B3u and y spans
B2u ; the direct product B3u × B2u is calculated using the method described in
Section 10C.1(a) on page 385
y
E C 2x C 2 C 2z i σ x y σ yz σ zx
B3u 1 −1 −1 1 −1 1 −1 1
B2u 1 −1 1 −1 −1 1 1 −1
product 1 1 −1 −1 1 1 −1 −1
The characters in the product row are those of symmetry species B1g , thus
confirming that the function x y has symmetry species B1g in D 2h .
E10C.2(a) As explained in Section 10C.2(a) on page 387, only orbitals of the same symme-
try species may have a nonzero overlap. Inspection of the C 2v character table
shows that the oxygen 2s and 2pz orbitals both have A1 symmetry, so they
can interact with the A1 combination of fluorine orbitals. The oxygen 2p y
orbital has B2 symmetry, so it can interact with the B2 combination of fluorine
orbitals. The oxygen 2px orbital has B1 symmetry, so cannot interact with either
combination of fluorine orbitals and therefore remains nonbonding.
In SF2 the same interactions with the sulfur s and p orbitals are possible but
there is now the possibility of additional interactions involving the d orbitals.
The C 2v character table shows that dz 2 and dx 2 −y 2 have A1 symmetry, so they
can interact with the A1 combination of fluorine orbitals. The d yz orbital
has B2 symmetry so can interact with the B2 combination. The dx y and dzx
orbitals have A2 and B1 symmetry respectively, so they cannot interact with
either combination of fluorine orbitals and therefore remain nonbonding.
E10C.3(a) As explained in Section 10C.2(a) on page 387, only orbitals of the same symme-
try species may have a nonzero overlap. Inspection of the C 2v character table
shows that the nitrogen 2s, px , p y , and pz orbitals span A1 , B1 , B2 , and A1
respectively. Because none of these orbitals have A2 symmetry, none of them
can interact with the A2 combination of oxygen orbitals.
The character table also shows that the dz 2 , dx 2 −y 2 , dx y , d yz , and dzx orbitals
of the sulfur in SO2 transform as A1 , A1 , A2 , B2 and B1 respectively. Therefore
the dx y orbital has the correct symmetry to interact with the A2 combination
of oxygen p orbitals.
E10C.4(a) As explained in Section 10C.3 on page 389, a transition from a state with sym-
metry Γ(i) to one with symmetry Γ(f) is only allowed if the direct product
Γ(f) × Γ(q) × Γ(i) contains the totally symmetric irreducible representation,

which in C 2v is A1 . If the ground state is A1 , then the direct product becomes
Γ(f) × Γ(q) × A1 . This is simply Γ(f) × Γ(q) because, from the first simplifying
feature of direct products listed in Section 10C.1(a) on page 385, the direct
product of the totally symmetric irreducible representation A1 with any other
representation is the latter representation itself.
If Γ(f) × Γ(q) is to be A1 , then Γ(f) must equal Γ(q) because, from the second
simplifying feature of direct products listed in Section 10C.1(a) on page 385,
the direct product of two irreducible representations only contains the totally
symmetric irreducible representation if the two irreducible representations are
identical. The C 2v character table shows that Γ(q) = B1 for x polarized light (q =
x), B2 for y polarised light, and A1 for z polarised light. It follows that x, y and
z polarised light can excite the molecule to B1 , B2 , and A1 states respectively.
E10C.5(a) The number of times n(Γ) that a given irreducible representation Γ occurs in a
representation is given by [10C.3a–386], n(Γ) = (1/h) ∑C N(C)χ(Γ) (C)χ(C),
where h is the order of the group, N(C) is the number of operations in class C,
χ(Γ) is the character of class C in the irreducible representation Γ, and χ(C) is
the character of class C in the representation being reduced.
n(A1 ) = 18 (1× χ(A1 ) (E)× χ(E) + 1× χ(A1 ) (C 2 )× χ(C 2 ) + 2× χ(A1 ) (C 4 )× χ(C 4 )
+ 2× χ(A1 ) (σv )× χ(σv ) + 2× χ(A1 ) (σd )× χ(σd ))
= 18 (1×1×5 + 1×1×1 + 2×1×1 + 2×1×3 + 2×1×1) = 2
Similarly
n(A2 ) = 81 (1×1×5 + 1×1×1 + 2×1×1 + 2×(−1)×3 + 2×(−1)×1) = 0
n(B1 ) = 81 (1×1×5 + 1×1×1 + 2×(−1)×1 + 2×1×3 + 2×(−1)×1) = 1
n(B2 ) = 81 (1×1×5 + 1×1×1 + 2×(−1)×1 + 2×(−1)×3 + 2×1×1) = 0
n(E) = 81 (1×2×5 + 1×(−2)×1 + 2×0×1 + 2×0×3 + 2×0×1) = 1
The representation therefore spans 2A1 + B1 + E .
E10C.6(a) The number of times n(Γ) that a given irreducible representation Γ occurs in a
representation is given by [10C.3a–386], n(Γ) = (1/h) ∑C N(C)χ(Γ) (C)χ(C),
the character of class C in the representation being reduced. In the case of D 4h ,
h = 16.
Because the representation being reduced has characters of zero for all classes
except E, C 2′ , σh , and σv , only these latter four classes make a non-zero con-
tribution to the sum and therefore only these classes need be considered. The
number of times that the irreducible representation A1g occurs is therefore
n(A1 ) = 1
16
(N(E)× χ(A1g ) (E)× χ(E) + N(C 2′ )× χ(A1g ) (C 2′ )× χ(C 2′ )
+ N(σh )× χ(A1g ) (σh )× χ(σh ) + N(σv )× χ(A1g ) (σv )× χ(σv ))
1
= 16
(1×1×4 + 2×1×2 + 1×1×4 + 2×1×2) = 1
Similarly
n(A2g ) = 1
16
(1×1×4 + 2×(−1)×2 + 1×1×4 + 2×(−1)×2) = 0
n(B1g ) = 16 (1×1×4 + 2×1×2 + 1×1×4 + 2×1×2) = 1
1
n(B2g ) = 161
(1×1×4 + 2×(−1)×2 + 1×1×4 + 2×(−1)×2) = 0
n(Eg ) = 16 (1×2×4 + 2×0×2 + 1×(−2)×4 + 2×0×2) = 0
1
n(A1u ) = 161
(1×1×4 + 2×1×2 + 1×(−1)×4 + 2×(−1)×2) = 0
n(A2u ) = 16 (1×1×4 + 2×(−1)×2 + 1×(−1)×4 + 2×1×2) = 0
1
n(B1u ) = 161
(1×1×4 + 2×1×2 + 1×(−1)×4 + 2×(−1)×2) = 0
n(B2u ) = 16 (1×1×4 + 2×(−1)×2 + 1×(−1)×4 + 2×1×2) = 0
1
n(Eu ) = 16
1
(1×2×4 + 2×0×2 + 1×0×4 + 2×1×2) = 1
The representation therefore spans A1g + B1g + Eu .
E10C.7(a) As explained in Section 10C.3 on page 389, a transition from a state with sym-
metry Γ(i) to one with symmetry Γ(f) is only allowed if the direct product Γ(f) ×
Γ(q) × Γ(i) contains the totally symmetric irreducible representation, which
for both molecules is A1g . The ground state is totally symmetric, implying
that it transforms as A1g . Therefore the direct product becomes Γ(f) × Γ(q) ×
A1g . This is simply Γ(f) × Γ(q) because, from the first simplifying feature of
direct products listed in Section 10C.1(a) on page 385, the direct product of the
totally symmetric representation A1g with any other representation is the latter
representation itself.
If Γ(f) × Γ(q) is to be A1g , then Γ(f) must equal Γ(q) because, from the second
simplifying feature of direct products listed in Section 10C.1(a) on page 385,
the direct product of two irreducible representations only contains the totally
symmetric irreducible representation if the two irreducible representations are
identical.
(i) Benzene belongs to point group D 6h . The D 6h character table in the On-
line resource centre shows that shows that z transforms as A2u , and x and
y together transform as E1u . Therefore light polarized along z can excite
benzene to an A2u state, and x or y polarised light can excite it to an E1u
state.
(ii) Naphthalene belongs to point group D 2h . The D 2h character table in the
Resource section shows that x transforms as B3u , y transforms as B2u , and z
transforms as B1u . Therefore naphthalene can be excited to B3u , B2u , or B1u
states by x, y, and z polarised light respectively.
E10C.8(a) As explained in Section 10C.1 on page 384 an integral can only be non-zero if
the integrand spans the totally symmetric irreducible representation, which in
C 2v is A1 . From Section 10C.1(a) on page 385 the symmetry species spanned by
the integrand px zpz is found by the forming the direct product of the symmetry
species spanned by px , z, and pz separately. These are read off the C 2v character
table by looking for the appropriate Cartesian functions listed on the right of
the table: x and hence px spans B1 , while z and pz both span A1 . The direct
product required is therefore B1 × A1 × A1 .
The order does not matter, so this is equal to A1 × A1 × B1 , which is equal to B1

because, from the first simplifying feature described in Section 10C.1(a) on page
385, the direct product of the totally symmetric representation with any other
representation is the latter representation itself: A1 × Γ(i) = Γ(i) . The integrand
therefore spans B1 . This is not the totally symmetric irreducible representation,
therefore the integral is zero .
E10C.9(a) As explained in Section 10C.3 on page 389, an electric dipole transition is for-
bidden if the electric transition dipole moment µ q,fi is zero. The transition
dipole moment is given by [10C.6–389], µ q,fi = −e ∫ ψ ∗f qψ i dτ where q is x,
y, or z. The integral is only non-zero if the integrand contains the totally sym-
metric representation, which from the C 3v character table is A1 .
For a transition A1 → A2 , the symmetry species of the integrand is given by

the direct product A2 × Γ(q) × A1 . The order does not matter so this is equal
to A1 × A2 × Γ(q) , which is simply equal to A2 × Γ(q) because, from the first
simplifying feature listed in Section 10C.1(a) on page 385, the direct product of
the totally symmetric irreducible representation with any other representation
is the latter representation itself. Therefore A1 × A2 = A2 .
The direct product A2 × Γ(q) contains the totally symmetric irreducible rep-
resentation only if Γ(q) spans A2 because, according to the second simplifying
feature listed in Section 10C.1(a) on page 385, the direct product of two irre-
ducible representations contains the totally symmetric irreducible representa-
tion only if the two irreducible representations are identical. The C 3v character
table shows that none of x, y or z span A2 , so it follows that the integrand does
not contain A1 and hence the transition is forbidden .
P10C.1 Methane belongs to point group Td . The methane molecule and its H1s orbitals
are shown in Fig. 10.17, along with one operation of each class. It is sufficient to
consider just one operation in each class because, by definition, all operations
the same class have the same character.
The C 3 operation shown in Fig. 10.17 leaves sA unchanged but converts sB into
sC , sC into sD , and sD into sB : ( sA sC sD sB ) ← ( sA sB sC sD ). This is written
C3
C2,S4 σd
sA
H
sB
sD
H H
H
sC
Figure 10.17
using matrix multiplication as
D(C 3 )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 1 0 0 0 ⎞
⎜ 0 0 0 1 ⎟
( s A s C s D s B ) = ( s A s B s C s D )⎜ ⎟
⎜ 0 1 0 0 ⎟
⎝ 0 0 1 0 ⎠
The matrix D(C 3 ) is the representative of the C 3 operation in this basis. Sim-
ilarly, the C 2 operation shown in Fig. 10.17 exchanges sA and sB , and also ex-
changes sC and sD . This is written using matrix multiplication as
D(C 2 )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 1 0 0 ⎞
⎜ 1 0 0 0 ⎟
( s B s A s D s C ) = ( s A s B s C s D )⎜ ⎟
⎜ 0 0 0 1 ⎟
⎝ 0 0 1 0 ⎠
The σd operation shown in Fig. 10.17 leaves sA and sB unchanged and exchanges
sC and sD ; this gives
D(σ d )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 1 0 0 0 ⎞
⎜ 0 1 0 0 ⎟
( s A s B s D s c ) = ( s A s B s C s D )⎜ ⎟
⎜ 0 0 0 1 ⎟
⎝ 0 0 1 0 ⎠
The S 4 operation shown in Fig. 10.17 converts sA to sD , sB to sC , sC to sA , and

sD to sB , giving
D(S 4 )
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 0 0 1 0 ⎞
⎜ 0 0 0 1 ⎟
( s D s C s A s B ) = ( s A s B s C s D )⎜ ⎟
⎜ 0 1 0 0 ⎟
⎝ 1 0 0 0 ⎠
Finally, the E operation leaves all orbitals unchanged, meaning that its repre-
sentative is simply the identity matrix
D(E)
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
⎛ 1 0 0 0 ⎞
⎜ 0 1 0 0 ⎟
( s A s B s C s D ) = ( s A s B s C s D )⎜ ⎟
⎜ 0 0 1 0 ⎟
⎝ 0 0 0 1 ⎠
The representatives found above have the following characters
χ(E) = 4 χ(C 3 ) = 1 χ(C 2 ) = 0 χ(σd ) = 2 χ(S 4 ) = 0
This result can be arrived at much more quickly by noting that: (1) only the
diagonal elements of the representative matrix contribute to the trace; (2) or-
bitals which are unmoved by an operation will result in a 1 on the diagonal; (3)
orbitals which are moved to other positions by an operation will result in a 0 on
the diagonal. The character is found simply by counting the number of orbitals
which do not move. In the present case 4 are unmoved by E, 1 is unmoved by
C 3 , none are unmoved by C 2 , 2 are unmoved by σd , and none are unmoved by
S 4 . The characters are thus {4, 1, 0, 2, 0}.
This representation is decomposed using the method described in Section 10C.1(b)
on page 386. The number of times n(Γ) that a given irreducible representation
Γ occurs in a representation is given by [10C.3a–386],
1 (Γ)
n(Γ) = ∑ N(C)χ (C)χ(C)
h C
the character of class C in the representation being reduced. In the case of the
group Td , h = 24. The number of times that the irreducible representation A1
occurs is
n(A1 ) = 1
24
(N(E)× χ(A1 ) (E)× χ(E) + N(C 3 )× χ(A1 ) (C 3 )× χ(C 3 )
+ N(C 2 )× χ(A1 ) (C 2 )× χ(C 2 ) + N(σd )× χ(A1 ) (σd )× χ(σd )
+ N(S 4 )× χ(A1 ) (S 4 )× χ(S 4 ))
1
= 24
(1×1×4 + 8×1×1 + 3×1×0 + 6×1×2 + 6×1×0) = 1
Similarly
n(A2 ) = 1
24
(1×1×4 + 8×1×1 + 3×1×0 + 6×(−1)×2 + 6×(−1)×0) = 0
n(E) = 24
1
(1×2×4 + 8×(−1)×1 + 3×2×0 + 6×0×2 + 6×0×0) = 0
n(T1 ) = 24 (1×3×4 + 8×0×1 +
1
3×(−1)×0 + 6×(−1)×2 + 6×1×0) = 0
n(T2 ) = 24
1
(1×3×4 + 8×0×1 + 3×(−1)×0 + 6×1×2 + 6×(−1)×0) = 1
The four H1s orbitals in methane therefore span A1 + T2 .

As explained in Section 10C.2(a) on page 387, only orbitals of the same sym-
metry species may have a nonzero overlap. The carbon 2s orbital spans the
totally symmetric irreducible representation A1 as it is unchanged under all
symmetry operations in the group. It can therefore form molecular orbitals
with the A1 combination of hydrogen orbitals. Inspection of the Td character
table shows that the carbon px , p y , and pz orbitals jointly span T2 , so they can
form molecular orbitals with the T2 combinations of hydrogen orbitals.
The character table also shows that the dz 2 and dx 2 −y 2 orbitals on the silicon in
SiH4 jointly span E; note that the dz 2 orbital appears as (3z 2 − r 2 ) in the char-
acter table. These d orbitals therefore cannot form molecular orbitals with the
A1 or T2 combinations of hydrogen orbitals. However, the dx y , d yz , and dzx
orbitals on the silicon span T2 and can therefore form molecular orbitals with
the T2 combinations of hydrogen orbitals.
P10C.3 As explained in Section 10C.1 on page 384, an integral over a region will neces-
sarily vanish unless the integrand, or part thereof, spans the totally symmetric
irreducible representation of the point group of the region. In the case of the
function 3x 2 − 1, the −1 part will always span the totally symmetric represen-
tation because it is invariant to all symmetry operations. Therefore the integral
will not necessarily vanish in any of the ranges.
P10C.5 The p 1 symmetry adapted linear combination is shown in Fig. 10.18, along with
some of the symmetry elements.
C2',σv C2'',σd
σh
A
B
x
D
C4,C2,i,S4
C
Figure 10.18
This SALC is unaffected by any of the operations E, C 2 , C 2′′ , S 4 , and σv , but

changes sign under C 4 , C 2′ , i, σh , and σd . The representatives, which are the
same as the characters because the representatives are one-dimensional, are
therefore
class C E C 4 C 2 C 2′ C 2′′ i S 4 σh σv σd
D(C) 1 −1 1 −1 1 −1 1 −1 1 −1
The D 4h character table shows that this representation corresponds to the B2u
irreducible representation. The s orbital on the xenon spans the totally sym-
metric representation A1g , and the D 4h character table indicates that px and p y
jointly span Eu ; pz spans A2u . It also shows that d yz and dzx jointly span Eg .
The orbitals dz 2 , dx 2 −y 2 and dx y span respectively A1g , B1g and B2g . Because
only orbitals of the same symmetry may have a non-zero overlap it follows that
none of the s, p and d orbitals on the xenon can form molecular orbitals with
the B2u orbital p 1 .
P10C.7 The ethene molecule is shown in Fig. 10.19.
y
y
C2,σyz
σxy
sB H H s
C2z A
x
x
sC sD C2,σzx
H H
Figure 10.19
The SALCs are generated using the method described in Section 10C.2(b) on
page 387, applying each operation to sA . The results are given in the following
table.
y
Row E C 2z C2 C 2x i σxy σ yz σ zx
1 effect on sA sA sC sB sD sC sA sB sD
2 characters for Ag 1 1 1 1 1 1 1 1
3 product of rows 1 and 2 sA sC sB sD sC sA sB sD
4 characters for B2u 1 −1 1 −1 −1 1 1 −1
5 product of rows 1 and 4 sA −sC sB −sD −sC sA sB −sD
6 characters for B3u 1 −1 −1 1 −1 1 −1 1
7 product of rows 1 and 6 sA −sC −sB sD −sC sA −sB sD
8 characters for B1g 1 1 −1 −1 1 1 −1 −1
9 product of rows 1 and 8 sA sC −sB −sD sC sA −sB −sD
10 characters for B1u 1 1 −1 −1 −1 −1 1 1
11 product of rows 1 and 10 sA sC −sB −sD −sC −sA sB sD
The SALCs are formed by summing rows 3, 5, 7, 9 and 11 and dividing each by
the order of the group (h = 8).
Row 3: ψ (A1g ) = 81 (sA + sC + sB + sD + sC + sA + sB + sD ) = 1

(s
4 A
+ sB + sC + sD )
Row 5: ψ (B2u ) = 81 (sA − sC + sB − sD − sC + sA + sB − sD ) = 1
(s
4 A
+ sB − sC − sD )
Row 7: ψ (B3u ) = 81 (sA − sC − sB + sD − sC + sA − sB + sD ) = 1
(s
4 A
− sB − sC + sD )
Row 9: ψ (B1g ) = 18 (sA + sC − sB − sD + sC + sA − sB − sD ) = 1
(s
4 A
− sB + sC − sD )
Row 11: ψ (B1u ) = 81 (sA + sC − sB − sD − sC − sA + sB + sD ) = 0
The results from row 11 show that attempting to project out a SALC with sym-
metry B1u gives zero . This is because the four hydrogen 1s orbitals do not span
B1u .
11 Molecular Spectroscopy
11A General features of molecular spectroscopy

D11A.1 This is discussed in Section 11A.3 on page 401.
D11A.3 Doppler broadening. This contribution to the linewidth is due to the Doppler
effect which shifts the frequency of the radiation emitted or absorbed when the
molecules involved are moving towards or away from the detecting device. In
a gas, molecules have a wide range of speeds in all directions and the detected
spectral line is the absorption or emission profile arising from the resulting
Doppler shifts. The shape of a Doppler-broadened spectral line reflects the
Maxwell distribution of speeds in the sample.
Lifetime broadening. This kind of broadening is a quantum mechanical effect
which predicts that for a state with a lifetime τ there is an energy uncertainty
δE given by δE ≈ ħ/τ. This uncertainty in the energy translates to absorption
(or emission) over a range of frequencies and hence a linewidth. The lifetime
of a state may be limited by the rate of spontaneous emission from the state, in
which case the resulting broadening is called natural line broadening.
Collisions between molecules are efficient at changing their rotational and vi-
brational energies, and therefore the lifetimes of such states are limited by the
the collision rate. The resulting line broadening is called collisional or pressure
line broadening.
The rate of spontaneous emission cannot be changed; hence its contribution is
the same regardless of phase. Doppler broadening is expected to contribute in a
similar way for both gases and liquids. The higher density of liquids compared
to gases implies that collisions will be more frequent and hence the collisional
line broadening will be greater for a liquid.
E11A.1(a) If a light source of frequency ν 0 is approached at a speed s, the Doppler shifted
frequency ν a is [11A.11a–399],
1/2
1 + s/c
νa = ν0 ( )
1 − s/c
378 11 MOLECULAR SPECTROSCOPY
Writing the frequencies in terms of the wavelength as ν = c/λ and then invert-
ing before sides gives
1/2
1 − s/c
λa = λ0 ( )
1 + s/c
1/2
At nonrelativistic speeds, s ≪ c, this simplifies to λ a = λ 0 (1 − s/c) . Hence
λ a = (680 nm) × [1 − (60 km h−1 ) × (1 h/3600 s)

× (1000 m/1 km)/(2.9979 × 108 m s−1 )]1/2 = 680 nm
Within the precision of the data given, the Doppler shift is insignificant.
E11A.2(a) The uncertainty in the energy of a state with lifetime τ is δE ≈ ħ/τ. Therefore a
spectroscopic transition involving this state has an uncertainty in its frequency,
and hence a linewidth, of the order of δν = δE/h ≈ (2πτ)−1 . This expression
is rearranged to give the lifetime as τ = (2πδν)−1 ; expressing the linewidth as
a wavenumber gives τ = (2πδ ν̃c)−1 .
(i) For δν̃ = 0.20 cm−1
τ = [2π×(0.20 cm−1 )×(2.9979×1010 cm s−1 )]−1 = 2.65...×10−11 s = 27 ps
(ii) For δ ν̃ = 2.0 cm−1
τ = [2π×(2.0 cm−1 )×(2.9979×1010 cm s−1 )]−1 = 2.65...×10−12 s = 2.7 ps
E11A.3(a) The uncertainty in the energy of a state with lifetime τ is δE ≈ ħ/τ. Therefore a
and hence a linewidth, of the order of δν = δE/h ≈ (2πτ)−1 . If the linewidth is
expressed as a wavenumber the expression becomes δν̃ = δE/hc ≈ (2πτc)−1 .
If each collision deactivates the molecule, the lifetime is 1/(collision frequency),
but if only 1 in N of the collisions deactivates the molecule, the lifetime is
N/(collision frequency). Thus τ = N/z, where z is the collision frequency. The
linewidth is therefore δ ν̃ = (2πcN/z)−1 .
(i) If each collision is effective at deactivation, N = 1 and with the data given
δ ν̃ = [2π × (2.9979 × 1010 cm s−1 ) × 1/(1.0 × 1013 s−1 )]−1 = 53 cm−1
(ii) If only 1 in 100 collisions are effective at deactivation, N = 100
δ ν̃ = [2π × (2.9979 × 1010 cm s−1 ) × 100/(1.0 × 1013 s−1 )]−1 = 0.53 cm−1
E11A.4(a) The ratio A/B is given by [11A.6a–396], A/B = 8πhν 3 /c 3 ; the frequency ν is
related to the wavelength though ν = c/λ, and to the wavenumber through
ν = ν̃c.
(i) For X-rays with λ = 70.8 pm
A 8πh(c/λ)3 8πh 8π × (6.6261 × 10−34 J s)

= = 3 = = 0.0469 J s m−3
B c3 λ (70.8 × 10−12 m)3
(ii) For visible light with λ = 500 nm
A 8πh 8π × (6.6261 × 10−34 J s)

= 3 = = 1.33 × 10−13 J s m−3
B λ (500 × 10−9 m)3
(iii) For infrared radiation with ν̃ = 3000 cm−1

A 8πh(c ν̃)3
= = 8πh ν̃ 3
B c3
= 8π × (6.6261 × 10−34 J s) × (3000 × 102 m−1 )3 = 4.50 × 10−16 J s m−3
Note the conversion of the wavenumber from cm−1 to m−1 .
E11A.5(a) The Beer–Lambert law [11A.8–397], I = I 0 10−ε[J]L relates the intensity of the
transmitted light I to that of the incident light I 0 .
mol−1 cm−1 )×(4.25×10−3 mol dm−3 )×(0.250 cm)
I/I 0 = 10−ε[J]L = 10−(723 dm
3
= 0.171...
Using this, the percentage reduction in intensity is calculated as 100(I 0 −I)/I 0 =

100(1 − I/I 0 ) = 100(1 − 0.171...) = 82.9% . Note the conversion of L to cm and
[J] to mol dm−3 in order to match the units of ε.
transmitted light I to that of the incident light I 0 . If a fraction T of the incident
light passes through the sample, I = TI 0 and hence I/I 0 = T; T is the transmit-
tance. It follows that log T = −ε[J]L hence ε = −(log T)/[J]L. If 18.1% of the
light is transmitted, T = 0.181
ε = −(log T)/[J]L = −[log(0.181)]/(0.139 × 10−3 mol dm−3 ) × (1.00 cm)

= 5.34 × 103 dm3 mol−1 cm−1
Note the use of L in cm and the conversion of [J] to mol dm−3 in order to give
the usual units of ε.
transmitted light I to that of the incident light I 0 . If a fraction α is absorbed,
then a fraction T = 1−α of the incident light passes through the sample, I = TI 0
and hence I/I 0 = T; T is the transmittance. It follows that log T = −ε[J]L
hence [J] = −(log T)/εL. If 38.5% of the light is absorbed, α = 0.385 and
T = 1 − 0.385 = 0.615
[J] = −(log T)/εL = −[log(0.615)]/(386 dm3 mol−1 cm−1 ) × (0.500 cm)

= 1.09... × 10−3 mol dm−3 = 1.09 mM
Note the use of L in cm.
E11A.8(a) The transmittance T is the ratio I/I 0 , hence the Beer–Lambert law [11A.8–397]
can be written T = I/I 0 = 10−ε[J]L . It follows that log T = −ε[J]L and hence
ε = −(log T)/[J]L. With this, the following table is drawn up, with L = 0.20 cm
[dye]/(mol dm−3 ) 0.0010 0.0050 0.0100 0.0500

T/% 81.4 35.6 12.7 3.0 × 10−3
T 0.814 0.356 0.127 3.0 × 10−5
ε/(dm3 mol−1 cm−1 ) 447 449 448 452
The average value of ε from these measurements is 449 dm3 mol−1 cm−1 .
E11A.9(a) The transmittance T is the ratio I/I 0 , hence the Beer–Lambert law [11A.8–397]
can be written T = I/I 0 = 10−ε[J]L . It follows that log T = −ε[J]L and hence
ε = −(log T)/[J]L. With the given data, T = 0.48 and L = 0.20 cm, the molar
absorption coefficient is calculated as
ε = −(log 0.48)/[(0.010 mol dm−3 )×(0.20 cm)] = 1.59...×102 dm3 mol−1 cm−1
The molar absorption coefficient is therefore ε = 1.6 × 102 dm3 mol−1 cm−1 .
For a path length of 0.40 cm the transmittance is
dm 3 mol−1 cm−1 )×(0.40 cm)×(0.01 mol dm−3 )

T = 10−(1.59 ...×10
2
= 0.230
Hence T = 23% .
E11A.10(a) The ratio of the incident to the transmitted intensities of light after passing
through a sample of length L, molar concentration [H2 O], and molar absorp-
tion coefficient ε is given by [11A.8–397], T = I/I 0 = 10−ε[H2 O]L . It follows that
log T = −ε[H2 O]L, which rearranges to give L = −(log T)/ε[H2 O].
The molar concentration of H2 O is calculated by noting that its mass density
is ρ = 1000 kg m−3 and its molar mass is M = 18.016 g mol−1 . The con-
centration is therefore ρ/M = (1000 kg m−3 )/(18.016 × 10−3 kg mol−1 ) =
55.5... × 103 mol m−3 = 55.5... mol dm−3 .
The light intensity is half that at the surface when T = 0.5, hence the depth is
calculated as
L = −(log 0.5)/[(6.2 × 10−5 dm3 mol−1 cm−1 ) × (55.5... mol dm−3 )]

= 87.4... cm = 0.875 m
The light intensity reaches a tenth of at the surface when T = 0.1
L = −(log 0.1)/[(6.2 × 10−5 dm3 mol−1 cm−1 ) × (55.5... mol dm−3 )]

= 290... cm = 2.90 m
E11A.11(a) The integrated absorption coefficient is given by [11A.10–399], A = ∫band ε(ν̃) dν̃,
where the integration is over the band and ν̃ = λ−1 is the wavenumber. The ini-
tial, peak, and final wavenumbers of the lineshape are given by (220×10−7 cm)−1 =
4.54...×104 cm−1 , (270×10−7 cm)−1 = 3.70...×104 cm−1 and (300×10−7 cm)−1 =
3.33... × 104 cm−1 .
1
Assuming that the lineshape is triangular the area under it is 2
× base × height
A= 1
2
× [(4.54... − 3.33...) × 104 cm−1 ] × (2.21 × 104 dm3 mol−1 cm−1 )
= 1.33... × 108 dm3 mol−1 cm−2
The integrated absorption coefficient is therefore 1.34 × 108 dm3 mol−1 cm−2 .
E11A.12(a) The Doppler linewidth is given by [11A.12a–400], δν obs = (2ν 0 /c)(2kT ln 2/m)1/2 .
Because λ = c/ν this may be rewritten δν obs = (2/λ 0 )(2kT ln 2/m)1/2 . Taking
the mass of a hydrogen atom as 1 m u gives the linewidth as
δν obs = (2/λ 0 )(2kT ln 2/m)1/2

1/2
2 × (1.3806 × 10−23 J K−1 ) × (300 K) × ln 2
= [2/(821 × 10−9 m)] × ( )
1.6605 × 10−27 kg
= 4.53... × 109 Hz
where 1 J = 1 kg m2 s−2 is used. Expressed as a wavenumber the linewidth is

(4.53... × 109 Hz)/(2.9979 × 1010 cm s−1 ) = 0.151 cm−1 .
P11A.1 The fraction of the incident photons that reach the retina is (1 − 0.30) × (1 −
0.25)×(1−0.09)×(1−0.43) = 0.272.... Hence the number of photons reaching
the retina in 0.1 s is
(4.0 × 103 mm−2 s−1 ) × (40 mm2 ) × (0.1 s) × 0.272... = 4.4 × 103
P11A.3 The absorbance at λ 1 and λ 2 are A 1 and A 2 , respectively
A 1 = ε A1 [A]L + ε B1 [B]L (11.1)

A 2 = ε A2 [A]L + ε B2 [B]L (11.2)
At each wavelength the absorbance depends on the concentration of each species

and the relevant molar absorption coefficient. Equation 11.1 is multiplied by ε A2
and eqn 11.2 is multiplied by ε A1 to give
ε A2 A 1 = ε A2 ε A1 [A]L + ε A2 ε B1 [B]L
ε A1 A 2 = ε A1 ε A2 [A]L + ε A1 ε B2 [B]L
Subtracting the two equations eliminates [A], and rearrangement gives the re-
quired expression for [B]
ε A2 A 1 − ε A1 A 2
ε A2 A 1 − ε A1 A 2 = ε A2 ε B1 [B]L − ε A1 ε B2 [B]L [B] =
(ε A2 ε B1 − ε A1 ε B2 )L
Simply exchanging the labels A and B gives the corresponding expression for
[A]
ε B2 A 1 − ε B1 A 2
[A] =
(ε B2 ε A1 − ε B1 ε A2 )L
P11A.5 Following the hint, a plot is made of ln ε against ν̃; the data are shown in the
following table and the plot is shown in Fig. 11.1.
λ/nm ε/(dm3 mol−1 cm−1 ) ln[ε/(dm3 mol−1 cm−1 )] ν̃/(104 cm−1 )

292.0 1 512 7.32 3.425
296.3 865 6.76 3.375
300.8 477 6.17 3.324
305.4 257 5.55 3.274
310.1 135.9 4.91 3.225
315.0 69.5 4.24 3.175
320.0 34.5 3.54 3.125
7
ln[ε/(dm3 mol−1 cm−1 )]
3.10 3.15 3.20 3.25 3.30 3.35 3.40 3.45

ν̃/(104 cm−1 )
Figure 11.1
The data are quite a good fit to the line
ln(ε/dm3 mol−1 cm−1 ) = 12.609 × [ν̃/(104 cm−1 )] − 35.793
This can be expressed as ln(ε/dm3 mol−1 cm−1 ) = a(ν̃/cm−1 ) + b with a =

1.2609 × 10−3 and b = −35.793. It follows that ε = e a ν̃ eb , where the units have
been omitted for clarity. With this expression for ε, the integrated absorption
coefficient is found by evaluating the integral
ν̃ max ν̃ max ν̃ max
A=∫ ε dν̃ = ∫ e a ν̃ eb dν̃ = eb (1/a)e a ν̃ ∣ν̃
ν̃ min ν̃ min min
b a ν̃ max a ν̃ min
= e (1/a) (e −e )
1 −3 −3
= e−35.793 (e(1.2609×10 )×(3.425×10 ) − e(1.2609×10 )×(3.125×10 ) )
4 4
1.2609 × 10−3
= 1.26 × 106 dm3 mol−1 cm−2
Again, units have been omitted for clarity.

1
P11A.7 (a) The area of a triangle is 2
× base × height, so the integrated absorption
coefficient is
A= 1
2
× [(34483 − 31250) cm−1 ] × (150 dm3 mol−1 cm−1 )
= 2.42 × 105 dm3 mol−1 cm−2
(b) Assume that the equilibrium involved is 2 M ↽ ⇀ M2 , where M is CH3 I.

ÐÐÐ
The total pressure is known, and from this it is possible to compute the
total concentration of M and M2 together; the fraction of the total present
as M2 is also known. Using these data it is possible to find the concentra-
tion of M, and hence the absorbance.
Suppose that initially there are n 0 moles of M which then come to equi-
librium by forming n moles of M2 : the amount in moles of M is then
n M = n 0 − 2n, and the total amount in moles of all species is n tot = n 0 − n.
Let the fraction that is present as dimer be α, α = n/n tot .
The aim is to express n M = n 0 − 2n in terms of the known quantities n tot
and α
=n tot +n =αn tot
© ©
n M = n 0 −2n = n tot − n = n tot (1 − α)
Assuming that the perfect gas law applies
n tot p
c tot = =
V RT
where c tot is the total concentration of both M and M2 . It follows that the
concentration of M is
n M n tot (1 − α) p
[M] = = = c tot (1 − α) = (1 − α)
V V RT
With the data given
(2.4 Torr)×[(1 atm)/(760 Torr)]×[(1.01325 × 105 Pa)/(1 atm)]

[M] =
(8.3145 J K−1 mol−1 ) × (373 K)
× (1 − 0.01) = 0.102... mol m−3 = 1.02... × 10−4 mol dm−3
The absorbance at the mid-point is
A = ε[M]L
= (150 dm3 mol−1 cm−1 ) × (1.02... × 10−4 mol dm−3 ) × (12.0 cm)
= 0.18
(c) With the data at 100 Torr and 18% dimers, the concentration of the monomer
is
(100 Torr)×[(1 atm)/(760 Torr)]×[(1.01325 × 105 Pa)/(1 atm)]
[M] =
(8.3145 J K−1 mol−1 ) × (373 K)
× (1 − 0.18) = 3.52... mol m−3 = 3.52... × 10−3 mol dm−3
The absorbance at the mid-point is
A = ε[M]L
= (150 dm3 mol−1 cm−1 ) × (3.52... × 10−3 mol dm−3 ) × (12.0 cm)
= 6.34...
The absorbance at the mid-point is therefore A = 6.35 . From this value

the molar absorption coefficient would be inferred as ε = A/c tot L and c tot
is computed as before from the pressure
(100 Torr)×[(1 atm)/(760 Torr)]×[(1.01325 × 105 Pa)/(1 atm)]

c tot =
(8.3145 J K−1 mol−1 ) × (373 K)
= 4.29... mol m−3 = 4.29... × 10−3 mol dm−3
Hence
ε = A/c tot L = (6.34...)/[(4.29... × 10−3 mol dm−3 ) × (12.0 cm)]

= 123 dm3 mol−1 cm−1
P11A.9 The line from the star is at longer wavelength, and hence lower frequency, than
for the Earth-bound observation, therefore the object is receding. The Doppler
shift is given by [11A.11a–399]
1/2
1 − (s/c)
f = ν r /ν 0 = ( )
1 + (s/c)
The ratio f is equal to λ 0 /λ r because the frequency is inversely proportional to

the wavelength. Writing x = s/c gives
1 − x 1/2 1− f2
f =( ) hence f 2 (1 + x) = (1 − x) hence x=
1+x 1+ f2
It follows that s = c[1 − (λ 0 /λ r )2 ]/[1 + (λ 0 /λ r )2 ].
1 − [(654.2 nm)/(706.5 nm)]2

s = (2.9979×108 m s−1 )× = 2.301 × 106 m s−1
1 + [(654.2 nm)/(706.5 nm)]2
The Doppler linewidth is given by [11A.12a–400], δν/ν 0 = (2/c)(2kT ln 2/m)1/2 .

Provided that the linewidth is small compared to the absolute frequency of the
line (which is the case here), δν/ν 0 is well approximated by δλ/λ 0
δλ 2 2kT ln 2 1/2 δλ 2 c 2 m
= ( ) hence T =( )
λ0 c m λ 0 8k ln 2
With the data given
0.0618 nm 2 (2.9979 × 108 m s−1 )2 × 47.95 × (1.6605 × 10−27 kg)

T=( )
706.5 nm 8 × (1.3806 × 10−23 J K−1 ) × ln 2
= 7.15 × 105 K
P11A.11 If each collision is effective at changing the energy of a state, the lifetime is
simply the inverse of the collision rate: τ = 1/z .
The uncertainty in the energy of a state with lifetime τ is δE ≈ ħ/τ. Therefore a
and hence a linewidth, of the order of δν = δE/h ≈ (2πτ)−1 . Using τ = 1/z
and the given expression for z gives the linewidth as
1/2
4σ kT 1/2 p 4σ 2
δν = 1/2πτ = z/2π = ( ) =( 3 ) p
2π πm kT π mkT
With the given data and taking m = 36 m u

1/2
4 × (0.30 × 10−18 m2 )2
δν = ( 3 )
π ×(36)×(1.6605 × 10−27 kg)×(1.3806 × 10−23 J K−1 ) × (298 K)
× (1.01325 × 105 Pa) = 0.70 GHz
The Doppler linewidth is given by [11A.12a–400], δν/ν 0 = (2/c)(2kT ln 2/m)1/2 ;

with the data given
2ν 0 2kT ln 2 1/2 2ν̃ 0 c 2kT ln 2 1/2 2kT ln 2 1/2

δν = ( ) = ( ) = 2ν̃ 0 ( )
c m c m m
1/2
2 × (1.3806 × 10−23 J K−1 ) × (298 K) × ln 2
−1
= 2×(6356 m )×( )
(36)×(1.6605 × 10−27 kg)
= 3.93 MHz
Note that ν̃ 0 is used in m−1 . For the collisional broadening to be equal to

the Doppler broadening the former must be reduced by a factor 700/3.93 =
178; because the linewidth is proportional to the pressure, this means that the
pressure must be reduced by this factor to (1.01325 × 105 Pa)/178 = 569 Pa or
4.27 Torr .
P11A.13 The best way to approach this is to generate the interferogram in a numerical
form, that is as a table of data points. As is seen in the previous Problem it is nec-
essary to have at least two data points per cycle in order to represent the wave-
number correctly, which implies that the distance by which the mirror must
be moved in one step is δ = 1/2ν̃ max , where ν̃ max is the highest wavenumber
which will be represented correctly. The i th data point in the interferogram is
constructed using the expression
apodization
³¹¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹µ
I i = ∑ {a j [1 + cos(2πν̃ j iδ)]} e−α(i δ)
2
where a j and ν̃ j are the intensity and wavenumber, respectively, of the jth peak
in the spectrum; i runs from 0 to N, the number of data points. The value of
N is a matter of choice, but a sensible starting value might be 256; the reason
for this apparently odd choice is that some numerical implementations of the
Fourier transform require that the number of points be a power of 2 (256 = 28 ).
The apodization term is there in order to force the interferogram to go smoothly

to zero (or at least near to zero) for the largest value of the pathlength difference
N δ. If this is not done, the peaks in the spectrum will have ‘wiggles’ around
their bases, as seen in Fig. 11A.2 on page 397. In a practical spectrometer this
term might not be required because with radiation passing through the inter-
ferometer covers a wide range of frequencies and interference between these
will naturally drive the interferogram to zero. In this simulation, with only a
few frequencies present, apodization is required. The parameter α is adjusted
to achieve the desired smoothing of the envelope.
Figure 11.2 shows an interferogram computed using the following parameters;

the data points have been joined up by a continuous line
N = 256 ν̃ max = 100 cm−1 δ = 1/(2 × 100 cm−1 ) = 0.005 cm α = 2.5 cm−2
a 1 = 0.25 ν̃ 1 = 5.0 cm−1 a 2 = 1.00 ν̃ 2 = 15 cm−1 a 3 = 0.75 ν̃ 3 = 50 cm−1
2
I
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
p/cm
Figure 11.2
To find the spectrum it is necessary to compute the Fourier transform of the in-
terferogram. There are many variants of the way this transform is implemented
as a numerical procedure, and the one needed here is usually referred to as the
discrete cosine Fourier transform. As can be seen from Fig. 11.2, the interfer-
ogram is always positive and decays to zero; this will give a large peak in the
spectrum at a wavenumber of zero, in addition to the peaks corresponding to
the wavenumbers of the oscillating terms that have been introduced. Figure 11.3
shows the spectrum obtained by Fourier transformation of the interferogram;
the peak at zero wavenumber has been truncated.
0 20 40 60 80 100
ν̃/cm
Figure 11.3
11B Rotational spectroscopy

D11B.1 This is discussed in Section 11B.3 on page 413.
12
D11B.3 C has spin zero and so is a boson; 13C and 1H have spin half and so are
fermions; 2H has spin 1 and so is a boson. All the molecules are linear so the
same considerations as described in Section 11B.4 on page 415 apply.
For 1H 12C ≡ 12C 1H the 12C have no effect as they are spin 0, so the rotational
levels behave in just the same way as 1 H2 : the (odd J)/(even J) statistical weight
ratio is therefore 3/1. Similarly, the rotational levels of 2H 12C ≡ 12C 2H behave in
just the same way as 2 H2 : the (odd J)/(even J) statistical weight ratio is therefore
I/(I + 1) = 1/2.
For 1H 13C ≡ 13C 1H there are four nuclear spin wavefunctions arising from the
1
H nuclei, three symmetric and one antisymmetric with respect to exchange of
the nuclei. In addition there are four nuclear spin wavefunctions arising from
the 13C nuclei, three symmetric and one antisymmetric. Overall, there are 16
nuclear spin wavefunctions. Of these, 9 arise from combining a symmetric
wavefunction for 1H2 and a symmetric wavefunction for 13C2 , giving overall
symmetric wavefunctions. In addition there is one more overall symmetric
wavefunction obtained by combining the antisymmetric wavefunction for 13C2
with that for 1H2 . The total number of symmetric wavefunctions is therefore
10, and the remaining 6 are therefore antisymmetric.
The ratio of symmetric to antisymmetric nuclear spin functions is therefore
10/6, therefore the (odd J)/(even J) statistical weight ratio is 10/6.
D11B.5 Symmetric rotor: The energy depends on J and K 2 , hence each level except the
K = 0 level is doubly degenerate. In addition, states of a given J have (2J + 1)
values of the component of their angular momentum along an external axis,
characterized by the quantum number M J . The energy is not affected by M J ,
so there is a degeneracy of 2J + 1 for each J. It follows that a symmetric rotor

level is 2(2J + 1)-fold degenerate for K ≠ 0, and 2J + 1 degenerate for K = 0.
Linear rotor: A linear rotor has K fixed at 0, but there are still 2J + 1 values of
M J , so the degeneracy is 2J + 1.
Spherical rotor: A spherical rotor can be regarded as a version of a symmetric
rotor in which A = B; consequently the energy is independent of the 2J + 1
values that K can assume. Hence, there is a degeneracy of 2J + 1 associated
with both K and M J , resulting in a total degeneracy of (2J + 1)2 .
If a decrease in rigidity affects the symmetry of the molecule, the rotational
degeneracy could be affected also.
D11B.7 A molecule has three principal moments of inertia about perpendicular axes:
these moments are labelled I a , I b , and I c , with I c ≥ I b ≥ I a . A prolate symmetric
rotor has I a ≠ I b = I c ; examples include a thin rod, any linear molecule, CH3 F
and CH3 CN. An oblate symmetric rotor has I a = I b ≠ I c ; examples include a
flat disc, benzene and BF3 . In terms of I∣∣ and I , prolate rotors have I∣∣ < I and
oblate tops have I∣∣ > I .
E11B.1(a) The wavenumbers of the lines in the rotational spectrum are given by [11B.20a–
412], ν̃(J) = 2B̃(J + 1); the J = 3 ← 2 transition is therefore at ν̃(2) = 2B̃(2 +
1) = 6B̃. The rotational constant is given by [11B.7–408], B̃ = ħ/4πcI, and the
moment of inertia is given by m eff R 2 , where m eff = m 1 m 2 /(m 1 + m 2 ).
(14.0031 × 15.9949)m u2 1.6605 × 10−27 kg

I= × × (115 × 10−12 m)2
(14.0031 + 15.9949)m u 1 mu
= 1.63... × 10−46 kg m2
ħ 1.0546 × 10−34 J s
B̃ = =
4πcI 4π × (2.9979 × 1010 cm s−1 ) × (1.63... × 10−46 kg m2 )
= 1.70... cm−1
The transition occurs at 6B̃ = 6 × (1.70... cm−1 ) = 10.2 cm−1 . Expressed in

frequency units this is 6c B̃ = 6 × (2.9979 × 1010 cm s−1 ) × (1.70... cm−1 ) =
3.07... × 1011 Hz = 307 GHz .
Centrifugal distortion will lower the frequency.
412], ν̃(J) = 2B̃(J + 1). The J = 3 ← 2 transition is therefore at ν̃(2) = 2B̃(2 +
1) = 6B̃, hence B̃ = (63.56/6) cm−1 . The rotational constant is given by [11B.7–
408], B̃ = ħ/4πcI, and the moment of inertia is given by m eff R 2 , where m eff =
m 1 m 2 /(m 1 + m 2 ). It follows that R = (ħ/4πcm eff B̃)1/2 .
(1.0078 × 34.9688)m u2 1.6605 × 10−27 kg

m eff = × = 1.62 . . . × 10−27 kg
(1.0078 + 34.9688)m u 1 mu
1/2
1.0546 × 10−34 J s
R=( )
4π×(2.9979 × 1010 cm s−1 )×(1.62... × 10−27 kg)×[(65.36/6)cm−1 ]
= 125.7 pm
412], ν̃(J) = 2B̃(J + 1); the lines are therefore spaced by 2B̃, it therefore follows
that B̃ = (12.604/2) cm−1 . The rotational constant is given by [11B.7–408],
B̃ = ħ/4πcI, and the moment of inertia is given by m eff R 2 , where m eff =
m 1 m 2 /(m 1 + m 2 ). It follows that I = ħ/4πc B̃ and R = (I/m eff )1/2 .
I = ħ/4πc B̃
1.0546 × 10−34 J s
= = 4.4420 × 10−47 kg m2
4π × (2.9979 × 1010 cm s−1 ) × [(12.604/2] cm−1 )
(1.0078 × 26.9815)m u2 1.6605 × 10−27 kg

m eff = × = 1.61... × 10−27 kg
(1.0078 + 26.9815)m u 1 mu
R = (I/m eff )1/2 = [(4.44... × 10−47 kg m2 )/(1.61... × 10−27 kg)]1/2

= 165.9 pm
E11B.4(a) The most occupied J state is given by [11B.21–413], J max = (kT/2hc B̃)1/2 − 12 .
(i) At 25 ○ C, 298 K, this gives
J max =
1/2
(1.3806 × 10−23 J K−1 )×(298 K) 1
( ) −
2×(6.6261 × 10−34 J s)×(2.9979 × 1010 cm s−1 )×(0.244 cm−1 ) 2
= 20
(ii) At 100 ○ C, 373 K, this gives
J max =
1/2
(1.3806 × 10−23 J K−1 )×(373 K) 1
( ) −
2×(6.6261 × 10−34 J s)×(2.9979 × 1010 cm s−1 )×(0.244 cm−1 ) 2
= 23
E11B.5(a) For a molecule to show a pure rotational Raman spectrum it must have an
anisotropic polarizability. With the exception of spherical rotors, all molecules
satisfy this requirement. Therefore H2 , HCl, CH3 Cl all give rotational Raman
spectra.
E11B.6(a) The Stokes lines appear at wavenumbers given by [11B.24a–414], ν̃(J + 2 ← J) =

ν̃ i − 2B̃(2J + 3), where the wavenumber of the incident radiation is ν̃ i , and J is
the quantum number of the initial state. With the given data
ν̃(2 ← 0) = 20 487 cm−1 − 2 × (1.9987 cm−1 )(2 × 0 + 3) = 20 475 cm−1

E11B.7(a) The Stokes lines appear at wavenumbers given by [11B.24a–414], ν̃(J + 2 ← J) =

ν̃ i − 2B̃(2J + 3), where the wavenumber of the incident radiation is ν̃ i , and J is
the quantum number of the initial state. It therefore follows that the separation
between adjacent lines is 4B̃, hence B̃ = (0.9752/4) cm−1 .
The rotational constant is given by [11B.7–408], B̃ = ħ/4πcI, and the moment
of inertia is given by m eff R 2 , where m eff = m 1 m 2 /(m 1 + m 2 ). It follows that
I = ħ/4πc B̃ and R = (I/m eff )1/2 .
I = ħ/4πc B̃
1.0546 × 10−34 J s
= = 1.14... × 10−45 kg m2
4π × (2.9979 × 1010 cm s−1 ) × [(0.9752/4) cm−1 ]
For a homonuclear diatomic the effective mass is simply m eff = 21 m
1/2
1.14... × 10−45 kg m2
R = (I/m eff )1/2 = ( 1 )
2
× 34.9688 × (1.6605 × 10−27 kg)
= 198.9 pm
E11B.8(a) The ratio of the weights for (odd J)/(even J) is given by [11B.25–416]. For 35Cl,
I = 32 and the nucleus is therefore a fermion. The ratio is (odd J)/(even J)
= (I + 1)/I = ( 32 + 1)/( 23 ) = 5
3
.
E11B.9(a) The moment of inertia I of a molecule about a specified axis is given by [11B.2–
406], I = ∑ i m i r 2i where the sum is over all the atoms, m i is the mass of atom
i and r i is its perpendicular distance to the axis. For the calculation of the
moment of inertia about the bisector, the central atom makes no contribution.
R θ/2
Each of the other atoms is at a perpendicular distance R sin(θ/2), where θ is

the bond angle and R the bond length. The moment of inertia is therefore
I = 2 × m O R 2 sin2 (θ/2)
= 2 × (15.9949) × (1.6605 × 10−27 kg)
× [(128 × 10−12 m) × sin(117○ /2)]2
= 6.32... × 10−46 kg m2 = 6.33 × 10−46 kg m2
The corresponding rotational constant is given by [11B.7–408],
ħ 1.0546 × 10−34 J s
B̃ = =
4πcI 4π × (2.9979 × 1010 cm s−1 ) × (6.32... × 10−46 kg m2 )
= 0.442 cm−1
E11B.10(a) The required expressions are the first listed under symmetric rotors in Table 11B.1
on page 407
m A (m B + m A )
I = m A f 1 (θ)R 2 + f 2 (θ)R 2
m
mC
+ {(3m A + m B )R ′ + 6m A R[ 13 f 2 (θ)]1/2 } R′
m
I∣∣ = 2m A f 1 (θ)R 2
Note that the molecule described by these relationships is BA3 C, which be-
comes BA4 by letting C=A; the question refers to a molecule AB4 , but for con-
sistency with the main text the exercise will be continued with BA4 . Let m C =
m A and R ′ = R to give
m A (m B + m A )
I = m A f 1 (θ)R 2 + f 2 (θ)R 2
m
mA
+ {(3m A + m B ) + 6m A [ 31 f 2 (θ)]1/2 } R 2
m
with m = m B + 4m A . To simplify the expression somewhat let m B = αm A . This
gives m = αm A + 4m A = (4 + α)m A
m A2 (1 + α)
I = m A f 1 (θ)R 2 + f 2 (θ)R 2
(4 + α)m A
mA
+ {m A (3 + α) + 6m A [ 13 f 2 (θ)]1/2 } R 2
(4 + α)m A
(1 + α) 1
I /(m A R 2 ) = f 1 (θ) + f 2 (θ) + {(3 + α) + 6[ 31 f 2 (θ)]1/2 }
(4 + α) (4 + α)
I∣∣ /(m A R 2 ) = 2 f 1 (θ)
The variation of the moments of inertia with θ are shown in Fig. 11.4; I is shown
for three representative values of α. Not surprisingly, I and I∣∣ converge onto
the same value when θ is the tetrahedral angle (shown by the vertical dotted
line). This is because in this limit the molecule becomes tetrahedral and is then
a spherical rotor, for which all the moments of inertia are the same.
At the tetrahedral angle cos θ tet = − 13 ; hence f 1 (θ tet ) = 4
3
and f 2 (θ tet ) = 1
3
(1 + α) 1 1
I /(m A R 2 ) = 4
+ ×3+ {(3 + α) + 6[ 31 × 13 ]1/2 }
3
(4 + α) (4 + α)
(1 + α) 1 1
= 43 + ×3+ {(3 + α) + 2}
(4 + α) (4 + α)
4(4 + α) + (1 + α) + 3(5 + α) 32 + 8α
= = =8
3(4 + α) 3(4 + α) 3
The moment of inertia for a tetrahedral molecule is, from the table, I/(m A R 2 ) =
8
3
, in agreement with the result just derived. In this limit the moment of inertia
does not depend on the mass of B (the central atom), as the axes pass through
this atom.
3.0
I∣∣
I α = 1
2.8
I α = .2
I α = 5
2.6
I/m A R 2
2.4
2.2
2.0
90.0 95.0 100.0 105.0 110.0
θ/○
Figure 11.4
E11B.11(a) To be a symmetric rotor a molecule most possess an n-fold axis with n > 2.
(i) O3 is bent (like H2 O), it has a two-fold axis and so is an asymmetric rotor.
(ii) CH3 CH3 has a three-fold axis and so is a symmetric rotor. (iii) XeO4 is
tetrahedral, and so is a spherical rotor. (iv) Ferrocene has a five-fold axis and
so is a symmetric rotor.
E11B.12(a) In order to determine two unknowns, data from two independent experiments
are needed. In this exercise two values of B for two isotopologues of HCN are
given; these are used to find two moments of inertia. The moment of inertia of
a linear triatomic is given in Table 11B.1 on page 407, and if it is assumed that
the bond lengths are unaffected by isotopic substitution, the expressions for
the moment of inertia of the two isotopologues can be solved simultaneously
to obtain the two bond lengths.
The rotational constant in wavenumber is given by [11B.7–408], B̃ = ħ/4πcI;
multiplication by the speed of light gives the rotational constant in frequency
units B = ħ/4πI, which rearranges to I = ħ/4πB
I HCN = (1.0546 × 10−34 J s)/[4π × (44.316 × 109 Hz)] = 1.89... × 10−46 kg m2

I DCN = (1.0546 × 10−34 J s)/[4π × (36.208 × 109 Hz)] = 2.31... × 10−46 kg m2
It is somewhat more convenient for the subsequent manipulations to express

the moments of inertia in units of the atomic mass constant m u and nm.
2
109 nm 1 mu
I HCN = (1.89... × 10−46 kg m2 ) × ( ) ×
1m 1.6605 × 10−27 kg
= 0.114... m u nm2
2
109 nm 1 mu
I DCN = (2.31... × 10−46 kg m2 ) × ( ) ×
1m 1.6605 × 10−27 kg
= 0.139... m u nm2
Using the expressions from Table 11B.1 on page 407, the moments of inertia
are expressed in terms of the masses and bond lengths, where the former are
expressed as multiples on m u . In this case A = 1H or 2H, B = 12C and C = 14N.
(m H R − m N R ′ )2
I HCN = m H R 2 + m N R′2 −
mH + mC + mN
(1.0078R − 14.0031R ′ )2
= 1.0078R 2 + 14.0031R′2 −
1.0078 + 12.0000 + 14.0031
(1.0078R − 14.0031R′ )2
= 1.0078R 2 + 14.0031R′2 −
27.0109
′2 (m D R − m N R′ )2
I DCN = mD R + mN R −
2
mD + mC + mN
(2.0141R − 14.0031R ′ )2
= 2.0141R 2 + 14.0031R′2 −
2.0141 + 12.0000 + 14.0031
(2.0141R − 14.0031R′ )2
= 2.0141R 2 + 14.0031R′2 −
28.0172
These two equations need to be solved simultaneously for R and R ′ , but be-
cause they are quadratics this is a very laborious process by hand: it is best
achieved using mathematical software. This gives the resulting bond lengths as
R = R CH = 0.1062 nm and R′ = R CN = 0.1157 nm .
E11B.13(a) The centrifugal distortion constant is given by [11B.16–410], D̃ J = 4B̃ 3 /ν̃ 2 . With
the given data D̃ J = 4(6.511 cm−1 )3 /(2308 cm−1 )2 = 2.073 × 10−4 cm−1 .
The rotational constant is inversely proportional to the moment of inertia of
the molecule, I = m eff R 2 where R is the bond length and m eff is the effective
mass. Assuming that isotopic substitution does not affect the bond length, it
follows that B̃ ∝ m−1 eff . Assuming that isotopic substitution does not affect
−1/2
the force constant, the vibrational frequency is proportional to m eff . Thus
−1/2
D̃ ∝ (m−1 −2
eff ) /(m eff ) = m eff . For this estimation it is sufficient to use integer
3 2
masses, and because a ratio is involved these can be expressed as multiples of

mu .
2 1 × 127 2 + 127 2
D̃ 2HI /D̃ 1HI = (m eff 1HI /m eff 2HI ) = ( × ) = 0.25
1 + 127 2 × 127
E11B.14(a) For a molecule to show a pure rotational (microwave) absorption spectrum is

must have a permanent dipole moment. Of the molecules given, the only ones
to satisfy this requirement are HCl, CH3 Cl and CH2 Cl2 .
P11B.1 Suppose that the bond length is R and that the centre of mass is at a distance x
from mass m 1 and therefore (R − x) from mass m 2 . Balancing moments gives
m 1 x = m 2 (R − x), hence x = m 2 R/(m 1 + m 2 ). Using this result it follows that
(R − x) = R − m 2 R/(m 1 + m 2 ) = m 1 R/(m 1 + m 2 ). The moment of inertia is
therefore
m 1 m 22 R 2 m 2 m 12 R 2
I = m 1 x 2 + m 2 (R − x)2 = +
(m 1 + m 2 )2 (m 1 + m 2 )2
m 1 m 2 (m 2 + m 1 )R 2
m1 m2 R2
= = = m eff R 2
(m 1 + m 2 )2 (m 1 + m 2 )
P11B.3 The rotational terms for a symmetric rotor are given by [11B.13a–409], F̃(J, K) =
B̃J(J+1)+(Ã− B̃)K 2 . The selection rules are ∆J = ±1 and ∆K = 0, and therefore
the term in K does not affect the wavenumber of the lines in the spectrum; the
result is that the lines are at exactly the same wavenumbers as for a linear rotor,
[11B.20a–412], ν̃(J) = 2B̃(J + 1). The separation of the lines is 2B̃.
In frequency units the spacing is 2B = 2 × (298 GHz) = 596 GHz . Expressed
as a wavenumber this spacing is (596 × 109 Hz)/(2.9979 × 1010 cm s−1 ) =
19.9 cm−1 .
The rotational constant is given by [11B.7–408], B̃ = ħ/4πcI. Expressed in
frequency units this is B = ħ/4πI. It follows that I = ħ/4πB
1.0546 × 10−34 J s
I = ħ/4πB = = 2.82 × 10−47 kg m2
4π × (298 × 109 Hz)
Expressions for the moment of inertia are given in Table 11B.1 on page 407;
NH3 is a symmetric rotor and the second entry under symmetric rotors is the
required one. The moment of inertia corresponding to the rotational constant
B is I . With the data given
mH mN
I = m H (1 − cos θ)R 2 + (1 + 2 cos θ)R 2
m N + 3m H
It is convenient to work with the masses as multiples of m u and R in nm
I = (0.1014 nm)2 × [(1.0078) × (1 − cos 106.78○ )

1.0078 × 14.0031
+ (1 + 2 cos 106.78○ )] × m u
14.0031 + 3 × 1.0078
= 0.0169... m u nm2
Converting to the usual units gives
10−18 m2 1.6605 × 10−27 kg

I = (0.0169... m u nm2 ) × × = 2.815 × 10−47 kg m2
1 nm2 1 mu
This value is consistent with the moment of inertia determined from the given
rotational constant.
P11B.5 Bonding is essentially the result of electrostatic interactions so to a very good

approximation it is expected that adding an uncharged neutron will have no
effect on the bond length.
The wavenumbers of the lines expected for a diatomic are given by [11B.20a–
412], ν̃(J) = 2B̃(J + 1); the separation of the lines is 2B̃. The rotational constant
is inversely proportional to the effective mass, therefore if the bond length is
unaffected by isotopic substitution the ratio of the rotational constants should
be equal to the inverse ratio of the effective masses. With the data given
B̃ 1H 35Cl /B̃ 2H 35Cl = (20.8784 cm−1 )(10.7840 cm−1 ) = 1.93605
m eff , 2H 35Cl m 1H m 35Cl m 2 + m 35Cl

= × H
m eff , 1H 35Cl m 1H + m 35Cl m 2H m 35Cl
1.007 825 + 34.968 85 2.0140 × 34.968 85
= ×
1.007 825 × 34.968 85 2.0140 + 34.968 85
= 1.93440
These two quantities differ by less than 0.1% so the hypothesis that the bond
length is invariant to isotopic substitution is confirmed to quite a high level of
precision; with the accuracy of the data given there is, however, some percep-
tible change.
P11B.7 Note: there is an error in the problem; for the 34

S isotopologue the line at
47.462 40 GHz is for J = 3.
The wavenumbers of the lines expected for a linear rotor are given by [11B.20a–
412], ν̃(J) = 2B̃(J + 1); the separation of the lines is 2B̃. For OC32 S the average
spacing of the lines is 12.16272 GHz, so the best estimate for the rotational
constant is B OCS = 21 × (12.16272 GHz) = 6.08136 GHz. For OC34 S there
are just two lines, one for J = 1 and one for J = 3; these are separated by
23.73007 GHz, which is 4B. The best estimate for the rotational constant is
B OCS′ = 14 × (23.73007 GHz) = 5.93252 GHz.
The rotational constant in wavenumber is given by [11B.7–408], B̃ = ħ/4πcI;
multiplication by the speed of light gives the rotational constant in frequency
units B = ħ/4πI, hence I = ħ/4πB
I OCS = (1.0546 × 10−34 J s)/[4π × (6.08136 × 109 Hz)] = 1.37... × 10−45 kg m2

I OCS′ = (1.0546 × 10−34 J s)/[4π × (5.93252 × 109 Hz)] = 1.41... × 10−45 kg m2
where for short S implies 32S and S′ implies 34S. It is somewhat more convenient
for the subsequent manipulations to express the moments of inertia in units of
the atomic mass constant m u and nm.
2
109 nm 1 mu
I OCS = (1.37... × 10−45 kg m2 ) × ( ) ×
1m 1.6605 × 10−27 kg
= 0.831... m u nm2
2
109 nm 1 mu
I OCS′ = (1.41... × 10−45 kg m2 ) × ( ) ×
1m 1.6605 × 10−27 kg
= 0.851... m u nm2
Using the expressions from Table 11B.1 on page 407, the moments of inertia
are expressed in terms of the masses and bond lengths, where the former are
expressed as multiples on m u . In this case A = 16O, B = 12C and C = 32S or 32S.
(m O R − m S R′ )2
I OCS = m O R 2 + m S R′2 −
mO + mC + mS
(15.9949R − 31.9721R′ )2
= 15.9949R 2 + 31.9721R ′2 −
15.9949 + 12.0000 + 31.9721
(15.9949R − 31.9721R ′ )2
= 15.9949R 2 + 31.9721R ′2 −
59.967
′ 2
(m R − m ′R )
= m O R 2 + m S′ R ′2 −
O S
I OCS′
m O + m C + m S′
(15.9949R − 33.9679R′ )2
= 15.9949R 2 + 33.9679R ′2 −
15.9949 + 12.0000 + 33.9679
(15.9949R − 33.9679R ′ )2
= 15.9949R 2 + 33.9679R ′2 −
61.9628
These two equations need to be solved simultaneously for R and R ′ , but be-
cause they are quadratics this is a very laborious process by hand: it is best
achieved using mathematical software. This gives the resulting bond lengths as
R = R OC = 0.1167 nm and R′ = R CS = 0.1565 nm .
P11B.9 The wavenumbers of the lines expected for a linear rotor are given by [11B.20a–
412], ν̃(J) = 2B̃(J +1); the separation of the lines is 2B̃. However, the separation
between adjacent lines in the given data is not constant, but increases along the
series. To account for this, the effects of centrifugal distortion are included,
and in this case the frequencies of the lines are given by [11B.20b–412], ν(J) =
2B(J+1)−4D J (J+1)3 (written with the constants in frequency units). Division
of both side of this expression by 2(J +1) indicates that a plot of [ν(J)]/2(J +1)
against (J + 1)2 should be a straight line with slope −2D J and intercept B. The
data are tabulated below; δ is the difference between successive lines. The plot
is shown in Fig. 11.5.
J ν(J)/MHz δ/MHz [ν(J)/2(J + 1)]/MHz (J + 1)2

24 214 777.7 4 295.6 625
25 223 379.0 8 601.3 4 295.8 676
26 231 981.2 8 602.2 4 296.0 729
27 240 584.4 8 603.2 4 296.2 784
28 249 188.5 8 604.1 4 296.4 841
29 257 793.5 8 605.0 4 296.6 900
The data are a good fit to the line
{ν(J)]/2(J + 1)}/MHz = 3.652 × 10−3 × (J + 1)2 + 4293.28
The value of the rotational constant is found from the intercept: (B/MHz) =
intercept. Some elementary statistics on the best-fit line indicates an error of
4 296.5
[ν(J)/2(J + 1)]/MHz
4 296.0
4 295.5
600 650 700 750 800 850 900
2
(J + 1)
Figure 11.5
about 0.03 MHz in the intercept, so the best estimate for the rotational constant
is B = 4293.28 ± 0.03 MHz or, expressed as a wavenumber, B̃ = 0.1432 cm−1 .
It is somewhat unusual that the centrifugal distortion constant appears to be
negative.
The most occupied J state is given by [11B.15–410], J max = (kT/2hc B̃)1/2 − 21 .
At 298 K
J max =
1/2
(1.3806 × 10−23 J K−1 )×(298 K) 1
( ) −
2×(6.6261 × 10 J s)×(2.9979 × 1010 cm s−1 )×(0.1432 cm−1 )
−34 2
= 26
A similar calculation at 100 K gives J max = 15 .
P11B.11 The population of level J, N J , is given by N J ∝ g J e−E J /k T . In this expression g J

is the degeneracy of level J, g J = (2J + 1), and E J is the energy of that level, E J =
hc B̃J(J+1). To find the level with the greatest population the derivative dN J /dJ
is computed and then set to zero; it is not necessary to know the constant of
proportion, which will be written A. To compute the derivative requires the
product rule and the chain rule
d
A(2J + 1)e−hc B̃ J(J+1)/k T
dJ
= A×2×e−hc B̃ J(J+1)/k T − A(2J + 1)×(2J + 1)×(hc B̃/kT)e−hc B̃ J(J+1)/k T
setting the derivative to zero and gathering terms gives
0 = Ae−hc B̃ J(J+1)/k T [2 − (2J + 1)2 (hc B̃/kT)]
The exponential term goes to zero as J → ∞, but this is not a maximum; rather,
the maximum is when the term in square brackets is zero
0 = [2 − (2J max + 1)2 (hc B̃/kT)]

(2J max + 1)2 = 2kT/hc B̃ hence (2J max + 1) = (2kT/hc B̃)1/2
J max = 1
2
× (2kT/hc B̃)1/2 − 1
2
The level with the greatest population is therefore J max = (kT/2hc B̃)1/2 − 12 .
With the given data
1/2
(1.3806 × 10−23 J K−1 )×(298 K)
J max =( ) − 1
2(6.6261 × 10−34 J s)×(2.9979 × 1010 cm s−1 )×(0.1142 cm−1 ) 2
= 30
For a spherical rotor the degeneracy of each level is (2J + 1)2 . Finding the most
populated level proceeds as before
d
A(2J + 1)2 e−hc B̃ J(J+1)/k T
dJ
= A × 4(2J + 1) × e−hc B̃ J(J+1)/k T
− A(2J + 1)2 × (2J + 1) × (hc B̃/kT)e−hc B̃ J(J+1)/k T
setting the derivative to zero and gathering terms gives
0 = Ae−hc B̃ J(J+1)/k T (2J + 1) [4 − (2J + 1)2 (hc B̃/kT)]
As before the maximum occurs when the term in square brackets is zero
0 = [4 − (2J + 1)2 (hc B̃/kT)]

(2J max + 1)2 = 4kT/hc B̃ hence (2J max + 1) = (4kT/hc B̃)1/2
J max = 1
2
× (4kT/hc B̃)1/2 − 1
2
The level with the greatest population is therefore J max = (kT/hc B̃)1/2 − 12 .
With the given data
1/2
(1.3806 × 10−23 J K−1 )×(298 K)
J max =( ) − 1
(6.6261 × 10−34 J s)×(2.9979 × 1010 cm s−1 )×(5.24 cm−1 ) 2
= 6
In such calculations it may be helpful to use kT/hc = 207.225 cm−1 at 298 K

(from inside the front cover).
P11B.13 Temperature effects. At extremely low temperatures (10 K) only the lowest rota-
tional states are populated. No emission spectrum is expected for the CO in the
cloud and star-light microwave absorptions by the CO in the cloud are from the
lowest rotational states. At higher temperatures additional high-energy lines

appear because higher energy rotational states are populated. Circumstellar
clouds may exhibit infrared absorptions due to vibrational excitation as well as
electronic transitions in the ultraviolet. Ultraviolet absorptions may indicate
the photodissocation of carbon monoxide. High temperature clouds exhibit
emissions.
Density effects. The density of an interstellar cloud may range from one particle
to a billion particles per cm3 . This is still very much a vacuum compared
to the laboratory high vacuum of a trillion particles per cm3 . Under such
extreme vacuum conditions the half-life of any quantum state is expected to
be extremely long and absorption lines should be very narrow. At the higher
densities the vast size of nebulae obscures distant stars. High densities and high
temperatures may create conditions in which emissions stimulate emissions of
the same wavelength by molecules. A cascade of stimulated emissions greatly
amplifies normally weak lines – the maser phenomena of microwave amplifi-
cation by stimulated emission of radiation.
Particle velocity effects. Particle velocity can cause Doppler broadening of spec-
tral lines. The effect is extremely small for interstellar clouds at 10 K but is ap-
preciable for clouds near high temperature stars. Outflows of gas from pulsing
stars exhibit a red Doppler shift when moving away at high speed and a blue
shift when moving toward us.
There will be many more transitions observable in circumstellar gas than in

interstellar gas, because many more rotational states will be accessible at the
higher temperatures. Higher velocity and density of particles in circumstellar
material can be expected to broaden spectral lines compared to those of inter-
stellar material by shortening collisional lifetimes. (Doppler broadening is not
likely to be significantly different between circumstellar and interstellar mate-
rial in the same astronomical neighbourhood. The relativistic speeds involved
are due to large-scale motions of the expanding universe, compared to which
local thermal variations are insignificant.)
A temperature of 1000 K is not high enough to significantly populate elec-

tronically excited states of CO; such states would have different bond lengths,
thereby producing transitions with different rotational constants. Excited vi-
brational states would be accessible, though, and ro-vibrational transitions with
P and R branches as detailed in this following Topic would be observable in
circumstellar but not interstellar material. The rotational constant for CO is
1.691 cm−1 . The first excited rotational energy level, J = 1, with energy 2hc B̃, is
thermally accessible at about 6 K (based on the rough equation of the rotational
energy to thermal energy kT). In interstellar space, only two or three rotational
lines would be observable; in circumstellar space (at about 1000 K) the number
of transitions would be more like 20.
11C Vibrational spectroscopy of diatomic molecules

D11C.1 The rotational constant depends inversely on the moment of inertia, which in
turn depends on the square of the bond length. However, because the molecule
is vibrating, the bond length is constantly changing. Vibration is much faster
than rotation, so for the purposes of calculating the moment of inertia it is
generally a good approximation to take an average over the vibrational motion
and use ⟨R 2 ⟩ in place of R 2 . It follows that B ∝ 1/⟨R 2 ⟩.
If the vibration is assumed to be harmonic ⟨R 2 ⟩ increases with increasing vi-
brational energy. However, for a typical anharmonic vibration there is a much
greater effect on ⟨R 2 ⟩ arising from the asymmetry of the potential. Put simply,
instead of the molecule oscillating symmetrically about the equilibrium posi-
tion, the bond stretches more than it is compressed, resulting in the average
bond length increasing. As the vibrational energy increases the potential curve
becomes shallower for bond extension and the average bond length increases
further.
The value of 1/⟨R 2 ⟩ therefore decreases as the vibrational quantum number υ
increases, and as a result the rotational constant B is a decreasing function of
the υ. For typical molecules, this effect of the anharmonicity is dominant, and
it is not unusual for the rotational constant to decrease by 1–2 per cent when
going from the υ = 0 to the υ = 1 vibrational level.
D11C.3 Because bonding is principally a matter resulting from electrostatic interac-
tions, the addition of a neutral particle to the nucleus is not expected to alter the
geometry of a molecule (bond lengths, bond angles), nor is it expected to alter
the force constants which describe the stretching of bonds. However, rotational
spectra, and the rotational fine structure which is associated with vibrational
spectra, depend on the rotational constants, and in turn these depend on the
effective mass. Likewise, vibrational frequencies also depend on the effective
mass, and so they too will be affected.
Different isotopes may have different nuclear spins and this can affect the pat-
tern of intensities of lines arising from different rotational states.
E11C.1(a) The wavenumber of the fundamental vibrational transition is simply equal to
the vibrational frequency expressed as a wavenumber. This is given by [11C.4b–
419], ν̃ = (1/2πc)(k f /m eff )1/2 , where m eff is the effective mass, given by m eff =
m 1 m 2 /(m 1 + m 2 ). It follows that k f = m eff (2πc ν̃)2 . With the data given the
following table is drawn up.
1
H 19F 1
H 35Cl 1
H 81Br 1
H 127I
ν̃/cm−1 4141.3 2988.9 2649.7 2309.5
m eff /m u 0.9570 0.9796 0.9954 0.9999
−1
k f /N m 967.0 515.6 411.7 314.2
E11C.2(a) The terms (energies expressed as wavenumbers) of the harmonic oscillator are
given by [11C.4b–419], G̃(υ) = (υ + 12 )ν̃; these are wavenumbers and so can be
converted to energy by multiplying by hc to give E(υ) = (υ + 12 )hc ν̃. The
ground state has υ = 0, and the first excited state has υ = 1. The relative
population of these levels is therefore given by the Boltzmann distribution,
n 1 /n 0 = e−(E 1 −E 0 )/k T . The energy difference E 1 − E 0 = hc ν̃, and hence n 1 /n 0 =
e−hc ν̃/k T . It is convenient to compute the quantity hc ν̃/k first to give
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (559.7 cm−1 )

hc ν̃/k =
1.3806 × 10−23 J K−1
= 805.3... K
It follows that n 1 /n 0 = e−(805.3 ... K)/T
(i) At 298 K, n 1 /n 0 = e−(805.3 ... K)/(298 K) = 0.0670

(ii) At 500 K, n 1 /n 0 = e−(805.3 ... K)/(500 K) = 0.200
As expected, the relative population of the upper level increases with tempera-
ture.
E11C.3(a) Taking y e = 0 is equivalent to using the terms for the Morse oscillator, which
are given in [11C.8–420], G̃(υ) = (υ + 12 )ν̃ − (υ + 12 )2 ν̃x e . The transition υ ← 0
has wavenumber
∆G̃(υ) = G̃(υ) − G̃(0)

= [(υ + 12 )ν̃ − (υ + 12 )2 ν̃x e ] − [(0 + 21 )ν̃ − (0 + 12 )2 ν̃x e ]
= υ ν̃ − υ(υ + 1)ν̃x e
Data on three transitions are provided, but only two are needed to obtain values
for ν̃ and x e . The ∆G̃(υ) values for the first two transitions are
1←0 ν̃ − 2ν̃x e = 1556.22 cm−1

2←0 2ν̃ − 6ν̃x e = 3088.28 cm−1
Multiplying the first expression by 3 and subtracting the second gives
3(ν̃ − 2ν̃x e ) − (2ν̃ − 6ν̃x e ) = ν̃

hence ν̃ = 3 × (1556.22 cm−1 ) − (3088.28 cm−1 ) = 1580.4 cm−1
This value for ν̃ is used in the first equation, which is then solved for x e to give
x e = 12 − (1556.22 cm−1 )/[2 × (1580.4 cm−1 )] = 7.65 × 10−3 .
E11C.4(a) Following the discussion in Section 11C.3(b) on page 421, D̃ 0 is given by the area
under a plot of ∆G̃ υ+1/2 against (υ + 12 ), where ∆G̃ υ+1/2 = G̃(υ + 1) − G̃(υ).
The data are shown in the table and the plot in Fig. 11.6.
υ G̃ υ /cm−1 ∆G̃ υ+1/2 /cm−1 υ + 12

0 1 481.86 2 885.64 0.5
1 4 367.50 2 781.54 1.5
2 7 149.04 2 677.44 2.5
3 9 826.48 2 573.32 3.5
4 12 399.80
3 000
∆G̃ υ+1/2 /cm−1
2 000
1 000
0
0 5 10 15 20 25 30
υ+ 1
2
Figure 11.6
∆G̃ υ+1/2 /cm−1 = −104.11 × (υ + 21 ) + 2 937.7
This line intercepts the horizontal axis when
0 = −104.11 × (υ + 12 )max + 2 937.7 hence (υ + 12 )max = 28.22
The area under the line is simply the area of a triangle, 12 × base × height, which
in this case is 12 × (28.22) × (2 937.7) = 4.14 × 104 . The dissociation energy
is therefore D̃ 0 = 4.14 × 104 cm−1 ; only modest precision is quoted because a
long extrapolation is made on the basis of few data points.
The fact that the data fall on a good straight line indicates that the Morse levels
apply, in which case, according to [11C.9b–421], ∆G̃ υ+1/2 = ν̃ − 2(υ + 1)x e ν̃.
This expression is rewritten
∆G̃ υ+1/2 = ν̃ − 2(υ + 12 )x e ν̃ − x e ν̃
which implies that a plot of ∆G̃ υ+1/2 against (υ + 12 ) will have slope −2x e ν̃ and
intercept (ν̃ − x e ν̃). Hence, using the slope of the plot already made
x e ν̃/cm−1 = − 21 (−104.11) hence x e ν̃ = 52.06 cm−1

and then using the intercept
(ν̃ − x e ν̃)/cm−1 = 2 937.7

hence ν̃ = (2 937.7 cm−1 ) + (52.06 cm−1 ) = 2989.8 cm−1
The depth of the well, D̃ e is then found using [11C.8–420], x e = ν̃/4 D̃ e rear-
ranged to D̃ e = ν̃/4x e = ν̃ 2 /4ν̃x e . The dissociation energy is D̃ 0 = D̃ e − G̃(0)
(Fig. 11C.3 on page 420), hence
ν̃ 2
D̃ 0 = D̃ e − G̃(0) = − G̃(0)
4ν̃x e
(2989.8 cm−1 )2
= − (1481.86 cm−1 ) = 4.14 × 104 cm−1
4 × (52.06 cm−1 )
To within the precision quoted, both methods give the same result.
To convert to eV, the conversion 1 eV = 8065.5 cm−1 from inside the front cover
is used to give D 0 = 5.14 eV .
E11C.5(a) The wavenumber of the transition arising from the rotational state J in the R
branch (∆J = +1) of the fundamental transition (υ = 1 ← υ = 0) is given
by [11C.13c–423], ν̃ R (J) = ν̃ + 2B̃(J + 1). In this case ν̃ = 2308.09 cm−1 and
B̃ = 6.511 cm−1 hence
ν̃ R (2) = (2308.09 cm−1 ) + 2 × (6.511 cm−1 ) × (2 + 1) = 2347.2 cm−1
E11C.6(a) The vibrational frequency of a harmonic oscillator is given by [7E.3–260], ω =

(k f /m)1/2 ; ω is an angular frequency, so to convert to frequency in Hz, ν, use
ω = 2πν. Therefore 2πν = (k f /m)1/2 . Rearranging this gives the force constant
as k f = m(2πν)2
k f = (0.100 kg) × (2π × 2.0 Hz)2 = 16 N m−1
where 1 N = 1 kg m s−2 and 1 Hz = 1 s−1 are used.
E11C.7(a) The vibrational frequency, expressed as a wavenumber, of a harmonic oscillator

is given by [11C.4b–419], ν̃ = (1/2πc)(k f /m eff )1/2 , where m eff is the effective
mass, given by m eff = m 1 m 2 /(m 1 + m 2 ). Assuming that the force constants of
the two isotopologues are the same, ν̃ simply scales as (m eff )−1/2 . The fractional
change is therefore
1/2
ν̃ Na 35Cl − ν̃ Na 37Cl ν̃ 37 m 35
= 1 − Na Cl = 1 − ( eff Na Cl )
ν̃ Na 35Cl ν̃ Na 35Cl m eff Na 37Cl
22.9898 × 34.9688 22.9898 + 36.9651 1/2
=1−( × ) = 0.0107...
22.9898 + 34.9688 22.9898 × 36.9651
The fractional change, expressed as a percentage, is therefore 1.077% .

E11C.8(a) The wavenumber of the fundamental vibrational transition is simply equal to

the vibrational frequency expressed as a wavenumber. This is given by [11C.4b–
419], ν̃ = (1/2πc)(k f /m eff )1/2 , where m eff is the effective mass, given by m eff =
m 1 m 2 /(m 1 + m 2 ). It follows that k f = m eff (2πc ν̃)2 ; for a homonuclear di-
atomic m eff = 21 m 1 . With the data given
k f = ( 12 × [34.9688 × (1.6605 × 10−27 kg)])

× [2π × (2.9979 × 1010 cm s−1 ) × (564.9 cm−1 )]2
= 328.7 N m−1
Note the conversion of the mass to kg.
P11C.1 (a) Figure 11.7 shows plot of the total electronic energy (with respect to the
free atoms) as a function of the bond length for each of the hydrogen
halides. Calculations are performed with Spartan 10 using the MP2 method
with the 6-311++G** basis set.
10 HI
HBr
8
6
HCl
4
2
V / eV
0 HF
–2
–4
–6
–8
50 100 150 200 250 300 350
Figure 11.7 R / pm
The plot clearly shows that in going down the halogen group from HF to
HI the equilibrium bond length increases and the depth of the potential
well decreases. The equilibrium properties of each molecule are sum-
marized in the following table. The force constants are computed in the
harmonic approximation using [11C.4b–419], ν̃ = (1/2πc)(k f /m eff )1/2 ,
with m eff = m 1 m 2 /(m 1 + m 2 ). It follows that k f = m eff (2πc ν̃)2 . The
calculated bond lengths are in good agreement with the experimental
values, but the vibrational frequencies do not agree very well at all.
property H 19F H 35Cl H 81Br H 127I

R e /pm 91.7 127.3 141.3 161.2
R e, expt /pm 91.680 127.45 141.44 160.92
ν̃/cm−1 4198.162 3086.560 2729.302 2412.609
−1
ν̃ expt /cm 4138.29 2990.95 2648.98 2309.01
k f /(N m−1 ) 993.7 549.8 436.7 342.9
(b) The force constants decrease steadily down the series, as expected.
P11C.3 Figure 11.8 shows a plot of V (x)/V0 as a function of x/a; the minimum at x = 0
is clearly rather ‘flat’.
0.0
V (x)/V0
−0.5
−1.0
−3 −2 −1 0 1 2 3
x/a
Figure 11.8
For a given potential the force constant is defined in [11C.2b–418] in terms of

the second derivative as k f = (d2 V /dx 2 )0 .
d 2a 2 2 2
V0 (e−a /x − 1) = V0 3 e−a /x
2 2
dx x
d2 d 2a 2 −a 2 /x 2 2V0 a 2 3 2a 2 −a 2 /x 2
V (x) = V 0 e = (− + 3 )e
dx 2 dx x3 x3 x x
The second derivative, and hence the force constant, goes to zero at x = 0 on
account of the argument of the exponential term going to −∞; this dominates
the other terms. Thus, for small displacements there is no restoring force and
harmonic motion will not occur.
The potential is confining so it is expected that there will be quantized energy
levels. By loose analogy with the harmonic case the ground state wavefunction
is expected to have a maximum at x = 0 and then decay away to zero as x →
±∞. The first excited state is likely to have a node at x = 0, increase to a
maximum at some positive value of x and then decay away to zero. The function
will be odd with respect to x = 0, and so will show a symmetrically placed
minimum at a negative value of x.
P11C.5 (a) The dissociation energy is D̃ 0 = D̃ e − G̃(0) (Fig. 11C.3 on page 420), where
G̃(0) is the energy of the lowest vibrational term. For the Morse energy
levels given by [11C.8–420], G̃(υ) = (υ + 12 )ν̃ − (υ + 12 )2 x e ν̃, it follows
that G̃(0) = 12 ν̃ − 14 x e ν̃. The conversion between cm−1 and eV is achieved
using 1 eV = 8065.5 cm−1 from inside the front cover. For 1H 35Cl
hc D̃ 0 = hc D̃ e − G̃(0) = hc D̃ e − ( 21 ν̃ − 14 x e ν̃)
= (5.33 eV) − ( 21 × 2989.7 − 14 × 52.05) × [(1 eV)/(8065.5 cm−1 )]
= 5.15 eV
(b) The task is to calculate the values of ν̃ and x e ν̃ for the isotopologue 2H 35Cl.
The potential energy curve, and hence the value of the depth of the well
D̃ e , is the same for the two isotopologues.
In the harmonic limit the vibrational frequency is given by [11C.4b–419],
ν̃ = (1/2πc)(k f /m eff )1/2 , with m eff = m 1 m 2 /(m 1 + m 2 ). Assuming that
−1/2
the force constants of the two isotopologues are the same, ν̃ ∝ m eff .
From [11C.8–420] it is seen that x e = ν̃/4 D̃ e which rearranges to D̃ e =
−1/2 −1/2
ν̃/4x e . Because ν̃ ∝ m eff it follows that x e ∝ m eff also in order for D̃ e
to be unaffected by isotopic substitution. Thus x e ν̃ ∝ m−1eff .
1/2 1/2
ν̃ 2H 35Cl m 1 m eff , 1H 35Cl
= ( eff , HX ) hence ν̃ 2H 35Cl = ν̃ 1H 35Cl × ( )
ν̃ 1H 35Cl m eff , 2H 35Cl m eff , 2H 35Cl
1.0078 × 34.9688 2.0140 + 34.9688 1/2
ν̃ 2H 35Cl = (2989.7 cm−1 ) ( × )
1.0078 + 34.9688 2.02140 × 34.9688
= 2144.25 cm−1
Similarly
m eff , 1H 35Cl
x e ν̃ 2H 35Cl = x e ν̃ 1H 35Cl × ( )
m eff , 2H 35Cl
1.0078 × 34.9688 2.0140 + 34.9688
= (52.05 cm−1 ) ( × )
1.0078 + 34.9688 2.02140 × 34.9688
= 26.77 cm−1
Hence for 2H 35Cl
hc D̃ 0 = hc D̃ e − G̃(0) = hc D̃ e − ( 12 ν̃ − 14 x e ν̃)
= (5.33 eV) − ( 21 × 2144.25 − 14 × 26.77) × [(1 eV)/(8065.5 cm−1 )]
= 5.20 eV
The term 14 x e ν̃ evaluates to 8.3 × 10−4 eV, so at the precision to which

hc D̃ e is quoted this term has no effect.
P11C.7 (a) The dissociation energy D̃ 0 and the well depth D̃ e are related by D̃ e =
D̃ 0 + G̃(0) (Fig. 11C.3 on page 420), where G̃(0) is the vibrational term of
the ground vibrational state. In the harmonic approximation G̃(0) = 12 ν̃,

so it follows that
ν̃ = 2(D̃ e − D̃ 0 ) = 2(D e /hc − D 0 /hc)
With the data given
ν̃ = 2[(1.51 × 10−23 J) − (2 × 10−26 J)]

/[(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )] = 1.5 cm−1
In the harmonic limit the vibrational frequency is given by [11C.4b–419],

ν̃ = (1/2πc)(k f /m eff )1/2 , with m eff = m 1 m 2 /(m 1 + m 2 ). For a homo-
nuclear diatomic m eff = 21 m. It follows that k f = 12 m(2πc ν̃)2 .
kf = 1
2
× (4.0026) × (1.6605 × 10−27 kg)
× [2π × (2.9979 × 1010 cm s−1 ) × (1.5 cm−1 )]2
= 2.7 × 10−4 N m−1
The moment of inertia is I = m eff R 2 = 12 mR 2
I= 1
2
× (4.0026) × (1.6605 × 10−27 kg) × (297 × 10−12 m)2
= 2.93... × 10−46 kg m2 = 2.93 × 10−46 kg m2
hence
ħ
B̃ =
4πcI
(1.0546 × 10−34 J s)
= = 0.955 cm−1
4π × (2.9979 × 1010 cm s−1 ) × (2.93... × 10−46 kg m2 )
(b) If the Morse energy levels are assumed G(0) = 1

2
ν̃ − 41 x e ν̃, and from
[11C.8–420] x e = ν̃/4 D̃ e . It follows that
D̃ e = D̃ 0 + 12 ν̃ − 14 x e ν̃ = D̃ 0 + 12 ν̃ − ν̃ 2 /16 D̃ e
The result is a quadratic in ν̃ which is solved in the usual way

1
± [ 14 − ( D̃ e − D̃ 0 )/4 D̃ e ]1/2
ν̃ 2 /16 D̃ e − 12 ν̃+( D̃ e − D̃ 0 ) = 0 hence ν̃ = 2
1/8 D̃ e
With the data given
(D̃ e − D̃ 0 )/4D̃ e = [(1.51 × 10−23 J) − (2 × 10−26 J)]/[4 × (1.51 × 10−23 J)]

= 0.2497
D̃ e = (1.51 × 10−23 J)/[(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )]
= 0.760 cm−1
hence
ν̃ = 8 × (0.760 cm−1 ) × [ 12 ± ( 14 − 0.2497)1/2 ] = 2.93 cm−1 or 3.15 cm−1

With these values the anharmonicity constant is computed using x e =
ν̃/4 D̃ e
2.93 cm−1 3.15 cm−1
xe = = 0.96 or xe = = 1.04
4 × (0.760 cm−1 ) 4 × (0.760 cm−1 )
The anharmonicity constant is expected to be < 1, so the plausible values
are ν̃ = 2.9 cm−1 and x e = 0.96 . These values are very approximate given
the data used to derive them.
P11C.9 The data are shown in the table and the plot in Fig. 11.9.
υ ∆G̃ υ+1/2 /cm−1 υ+1

0 2 143.1 1
1 2 116.1 2
2 2 088.9 3
3 2 061.3 4
4 2 033.5 5
2 150
∆G̃ υ+1/2 /cm−1
2 100
2 050
1 2 3 4 5
υ+1
Figure 11.9

∆G̃ υ+1/2 /cm−1 = −27.40 × (υ + 1) + 2 170.7
From the slope
x e ν̃/cm−1 = − 12 × slope = − 12 (−27.40) hence x e ν̃ = 13.7 cm−1

and from the intercept
ν̃/cm−1 = intercept = 2 170.70 hence ν̃ = 2 170.7 cm−1

P11C.11 The data provided allow the calculation of two independent moments of inertia.
If it is assumed that the bond lengths are unaffected by isotopic substitution,
then it is possible to set up two equations and solve them simultaneously for
the CC and CH bond lengths.
Expressions for the wavenumbers of the lines in the P and R branches are given
by [11C.13a–423] and [11C.13c–423]; from these it follows that the spacing be-
tween the lines is 2B̃. The rotational constant is given by [11B.7–408], B̃ =
ħ/4πcI; it follows that I = ħ/4πc B̃.
I H = ħ/4πc B̃ H
1.0546 × 10−34 J s
= = 2.38... × 10−46 kg m2
4π × (2.9979 × 1010 cm s−1 ) × [(2.352/2) cm−1 ]
I D = ħ/4πc B̃ D
1.0546 × 10−34 J s
= = 3.30... × 10−46 kg m2
4π × (2.9979 × 1010 cm s−1 ) × [(1.696/2) cm−1 ]
The moment of inertia is defined as I = ∑ i m i r 2i , where r i is the perpendicular

distance from the atom with mass m i to the axis. In HCCH the axis passes
through the mid-point of the CC bond and is perpendicular to the long axis of
the molecule. Thus
I H = 2m C (r CC /2)2 + 2m H (r CH + r CC /2)2
I D = 2m C (r CC /2)2 + 2m D (r CH + r CC /2)2
These are the two equations which need to be solved simultaneously. Finding
the solution is much simplified by letting p = (r CC /2)2 and q = (r CH + r CC /2)2
to give
I H = 2m C p + 2m H q I D = 2m C p + 2m D q
It follows that
IH − ID
q=
2(m H − m D )
(2.38... × 10−46 kg m2 ) − (3.30... × 10−46 kg m2 )
= = 2.75... × 10−20 m2
2(1.0078 − 2.0140) × (1.6605 × 10−27 kg)
mD IH − mH ID
p=
2m C (m D − m H )
(2.0140) × (2.38... × 10−46 kg m2 ) − (1.0078) × (3.30... × 10−46 kg m2 )
=
2 × (12.0000) × (2.0140 − 1.0078) × (1.6605 × 10−27 kg)
= 3.65... × 10−20 m2
If follows that r CC = 2× p1/2 = 121.0 pm , and r CH = q 1/2 −r CC /2 = q 1/2 − p1/2 =

105.5 pm .
P11C.13 The variable x is the displacement from the equilibrium separation R e . The
fact that the potential is symmetric about R e means that ⟨R⟩ = R e and ⟨x⟩ = 0.
On the other hand ⟨x 2 ⟩ is definitely non-zero, as was seen for the case of the
harmonic oscillator in Topic 7E.
It follows straightforwardly that 1/⟨R⟩2 = 1/R e2 .
⟨R 2 ⟩ is found in the following way
⟨R 2 ⟩ = ⟨(R e + x)2 ⟩ = ⟨(R e2 + 2xR e + x 2 )⟩

A B
¬ ³¹¹ ¹ ¹ · ¹ ¹ ¹ µ
= ⟨R e2 ⟩ + ⟨2xR e ⟩ +⟨x 2 ⟩
Term A is simply the average of a constant term, which is equal to the term itself,
in this case R e2 . Term B is rewritten 2R e ⟨x⟩ by taking constant terms outside
the averaging; this term is zero because ⟨x⟩ = 0. Therefore ⟨R 2 ⟩ = R e2 + ⟨x 2 ⟩.
Using this 1/⟨R 2 ⟩ is found in the following way
1 1 1 1
= 2 = 2×
⟨R ⟩ R e + ⟨x ⟩ R e 1 + ⟨x 2 ⟩/R e2
2 2
1 ⟨x 2 ⟩
≈ (1 − )
R e2 R e2
where to go to the last line the expansion (1+ y)−1 ≈ 1− y is used. The resulting
expression includes the lowest power of ⟨x 2 ⟩/R e2 , as required
⟨1/R 2 ⟩ is found in the following way
1 1 1 1
⟨ ⟩=⟨ ⟩= 2 ⟨ ⟩
R2 (R e + x)2 R e (1 + x/R e )2
1 1
≈ 2 ⟨1 − 2x/R e + 3x 2 /R e2 ⟩ = 2 (⟨1⟩ − (2/R e )⟨x⟩ + (3/R e2 )⟨x 2 ⟩)
Re Re
1 3⟨x 2 ⟩
= (1 + )
R e2 R e2
On the penultimate line the expansion (1 + y)−2 ≈ 1 − 2y + 3y 2 is used. To go

to the final line the fact that ⟨x⟩ = 0 is used; the final expression has the lowest
non-zero power of ⟨x 2 ⟩/R e2 , as required.
It is evident that none of the averages are the same and that
1 1 1
⟨ ⟩> >
R2 ⟨R⟩2 ⟨R 2 ⟩
P11C.15 The rotational constant B̃ 0 is computed from
B̃ 0 = B̃ e − 21 a = (0.27971 cm−1 ) − 12 (0.187 × 10−2 cm−1 ) = 0.27877 cm−1
and similarly for B̃ 1
B̃ 1 = B̃ e − 23 a = (0.27971 cm−1 ) − 32 (0.187 × 10−2 cm−1 ) = 0.27691 cm−1

The wavenumber of the lines in the P and R branches are given by [11C.14–424]
ν̃ P (J) = ν̃ 0 − (B̃ 1 + B̃ 0 )J + (B̃ 1 − B̃ 0 )J 2

ν̃ R (J) = ν̃ 0 + (B̃ 1 + B̃ 0 )(J + 1) + (B̃ 1 − B̃ 0 )(J + 1)2
In these expressions ν̃ 0 is the wavenumber of the pure vibrational transition. If

the Morse levels are assumed, and if it is assumed that it is the υ = 1 ← υ = 0
transition which is being observed, the wavenumber is given by [11C.9b–421],
ν̃ 0 = ν̃ − 2x e ν̃.
For the line in the P branch from J = 3
ν̃ P (J)/cm−1 = 610.258 − 2 × 3.141 − (0.27691 + 0.27877) × 3

+ (0.27691 − 0.27877) × 32 = 602.292
and for the corresponding line in the R branch
ν̃ R (J)/cm−1 = 610.258 − 2 × 3.141 + (0.27691 + 0.27877) × 4

+ (0.27691 − 0.27877) × 42 = 606.170
The depth of the well, D̃ e is found using [11C.8–420], x e = ν̃/4 D̃ e rearranged to

D̃ e = ν̃/4x e = ν̃ 2 /4ν̃x e . The dissociation energy is D̃ 0 = D̃ e − G̃(0) (Fig. 11C.3
on page 420), and for the Morse oscillator G̃(0) = 12 ν̃ − 14 ν̃x e .
ν̃ 2
D̃ 0 = − 1 ν̃ + 14 ν̃x e
4ν̃x e 2
(610.258 cm−1 )2 1
= − (610.258 cm−1 ) + 14 (3.141 cm−1 )
4 × (3.141 cm−1 ) 2
= 2.93 × 104 cm−1
To convert to eV, the conversion 1 eV = 8065.5 cm−1 from inside the front cover
is used to give D 0 = 3.64 eV .
P11C.17 The features centred about 2143.26 cm−1 are the P and R branches. From [11C.13a–
423] and [11C.13c–423] the first line in the R branch occurs at ν̃ + 2B̃, and the
first line in the P branch is ν̃ − 2B̃. The separation of these two, 7.655 cm−1 , is
therefore 4B̃.
(a) The centre of the band is at the vibrational wavenumber, ν̃ = 2143.26 cm−1
(b) In the harmonic approximation the vibrational terms are G̃(υ) = (υ+ 12 )ν̃,
and so the lowest term is G̃(0) = 12 ν̃. The molar zero-point energy is
therefore N A × hc × 12 ν̃
E zpe = (6.0221 × 1023 mol−1 ) × (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )

× 21 × (2143.26 cm−1 ) = 12.82 kJ mol−1
(c) The harmonic frequency is given by [11C.4b–419], ν̃ = (1/2πc)(k f /m eff )1/2 ,

with m eff = m 1 m 2 /(m 1 + m 2 ). It follows that k f = m eff (2πc ν̃)2 . With the
data given
12.0000 × 15.9949
kf = × (1.6605 × 10−27 kg)
12.000 + 15.9949
× [2π × (2.9979 × 1010 cm s−1 ) × (2143.26 cm−1 )]2
= 1856 N m−1
(d) As noted at the start of the answer, 4B̃ = 7.655 cm−1 , hence B̃ = 1.914 cm−1 .
(e) The rotational constant is given by [11B.7–408], B̃ = ħ/4πcI, and the mo-
ment of inertia is given by m eff R 2 . It follows that R = (ħ/4πcm eff B̃)1/2 .
12.0000 × 15.9949
m eff = × (1.6605 × 10−27 kg) = 1.13... × 10−26 kg
12.0000 + 15.9949
R = (ħ/4πcm eff B̃)1/2
1/2
1.0546 × 10−34 J s
=( )
4π(2.9979 × 1010 cm s−1 )×(1.13... × 10−26 kg)×(1.914 cm−1 )
= 113.3 pm
Although the data are given to quite high precision the assumption that the
harmonic oscillator/rigid rotor models apply means that the derived values of
the bond length and so on are likely to have systematic errors which are higher
than the apparent precision of the data.
P11C.19 The method of combination differences, described in Section 11C.4(b) on page
424, involves taking the difference between two transitions which share a com-
mon lower rotational level or a common upper rotational level. In the case of
O and S branches, which correspond to ∆J = −2 and ∆J = +2, respectively,
the two transitions which share a common lower level are ν̃ O (J) and ν̃ S (J):
these are the transitions from J to J − 2, and from J to J + 2. As is evident from
Fig. 11.10, the difference in wavenumber between these two transitions is the
interval indicated by the dashed arrow which is simply G̃(J + 2) − G̃(J − 2) for
the upper vibrational state (assumed to be υ = 1)
G̃(J + 2) − G̃(J − 2) = B̃ 1 (J + 2)(J + 3) − B̃ 1 (J − 2)(J − 1) = B̃ 1 (8J + 4)
hence ν̃ S (J) − ν̃ O (J) = 8B̃ 1 (J + 12 )
The two transitions sharing a common upper level are ν̃ O (J + 2) and ν̃ S (J − 2):
these are the transitions from J + 2 to J, and from J − 2 to J. As is evident from
Fig. 11.10, the difference in wavenumber between these two transitions is the
interval indicated by the dotted arrow which is simply G̃(J + 2) − G̃(J − 2) for
the lower vibrational state (assumed to be υ = 0). This is the same interval as
above, with the exception that the rotational constant is B̃ 0 .
G̃(J + 2) − G̃(J − 2) = B̃ 0 (J + 2)(J + 3) − B̃ 0 (J − 2)(J − 1) = B̃ 0 (8J + 4)
hence ν̃ S (J − 2) − ν̃ O (J + 2) = 8B̃ 0 (J + 21 )
J+2
J υ=1
J−2
ν̃ O (J +2)
ν̃ S (J −2)
ν̃ O (J)
ν̃ S (J)
J+2
J υ=0
J−2
Figure 11.10
11D Vibrational spectroscopy of polyatomic molecules

D11D.1 The gross selection rule for vibrational Raman scattering is that for a particular
normal mode to be active the vibration must result in a change in the polar-
izability as the molecule vibrates about the equilibrium position. The origin
of this rule is that such an change in the polarizability will result in the dipole
induced in the molecule by the electric field of the incident radiation being
altered as the molecule vibrates. It is this modulation of the dipole that results
in Raman scattering.
D11D.3 The gross selection rule for infrared spectroscopy is that for a particular normal
mode to be active the vibration must result in a change in the dipole moment
as the molecule vibrates about the equilibrium position. The origin of this rule
is that such an oscillating dipole is needed to stir the electromagnetic field into
oscillation (and vice versa for absorption).
E11D.1(a) According to [11D.1–427], a linear molecule has 3N − 5 normal modes, where
N is the number of atoms in the molecule. There are 44 atoms in this linear
molecule, and so there are 3(44) − 5 = 127 normal modes.
E11D.2(a) According to [11D.1–427], a non-linear molecule has 3N − 6 vibrational normal
modes, where N is the number of atoms in the molecule; therefore H2 O has
3 normal modes. The terms (energies expressed as wavenumbers) for normal
mode q are given by [11D.2–428], G̃ q (υ) = (υ q + 21 )ν̃ q , where υ q is the quantum
number for that mode and ν̃ q is the wavenumber of the vibration of that mode.
These terms are additive, so the ground state term corresponds to each mode
having υ q = 0
G̃ 1 (0) + G̃ 2 (0) + G̃ 3 (0) = 1

2
(ν̃ 1 + ν̃ 2 + ν̃ 3 )
E11D.3(a) A mode is infrared active if the vibration results in a dipole which changes
as the molecule oscillates back and forth about the equilibrium geometry. A
mode is Raman active if the vibration results in the polarizability changing as

the molecule oscillates back and forth about the equilibrium geometry.
(i) The three normal modes of an angular AB2 molecule are analogous to
those of H2 O illustrated in Fig. 11D.3 on page 428; all three modes are
both infrared and Raman active.
(ii) The four normal modes of a linear AB2 molecule are analogous to those
of CO2 illustrated in Fig. 11D.2 on page 428. Of these, three are infrared
active (the asymmetric stretch and the doubly degenerate bend), and one
(the symmetric stretch) is Raman active. The molecule has a centre of
symmetry, so the rule of mutual exclusion applies and no mode is both
Raman and infrared active.
E11D.4(a) The benzene molecule has a centre of symmetry, so the rule of mutual exclusion
applies. The molecule has no permanent dipole moment and if the ring expands
uniformly this situation does not change: such a vibration does not lead to a
changing dipole and so the mode is infrared inactive .
This kind of ‘breathing’ vibration does lead to a change in the polarizability, so
the mode is Raman active .
E11D.5(a) The exclusion rule applies only to molecules with a centre of symmetry. H2 O
does not possess such symmetry, and so the exclusion rule does not apply .
E11D.6(a) With the exception of homonuclear diatomics, all molecules have at least one
infrared active normal mode. Of the molecules listed, HCl, CO2 , and H2 O
have infrared active modes.
E11D.7(a) According to [11D.1–427], a non-linear molecule has 3N − 6 vibrational normal

modes, where N is the number of atoms in the molecule; a linear molecule has
3N − 5 normal modes. All of the molecules listed are non-linear.
(i) H2 O has N = 3 and hence 3 normal modes.

(ii) H2 O2 has N = 4 and hence 6 normal modes.
(iii) C2 H4 has N = 6 and hence 12 normal modes.
P11D.1 Figure 11.11 shows a plot of V (h)/V0 as a function of hb 1/4 .
For a given potential the force constant is defined in [11C.2b–418] in terms of
the second derivative of the potential with respect to the displacement from
equilibrium; in this case h is the varaible which describes the displacement.
Thus k f = (d2 V /dh 2 )0 .
d
V0 (1 − e−bh ) = V0 4bh 3 e−bh
4 4
dh
d2 d
V0 4bh 3 e−bh = 4V0 b (3h 2 − 4bh 6 ) e−bh
4 4
2
V (h) =
dh dh
1.0
0.8
V (h)/V0
0.6
0.4
0.2
0.0
−2 −1 0 1 2
hb 1/4
Figure 11.11
The second derivative, and hence the force constant, goes to zero at h = 0. Thus,
for small displacements there is no restoring force.
The potential is confining so it is expected that there will be quantized vibra-
tional energy levels. By loose analogy with the harmonic case the ground state
wavefunction is expected to have a maximum at h = 0 and then decay away to
zero as h → ±∞.
P11D.3 (a) Calculations on SO2 are performed with Spartan 10 using the MP2 method
with the 6-311++G** basis set. The calculated equilibrium structure is
shown in Fig. 11.12 where it is seen that the two S–O bonds have equal
length, as expected. The calculated bond length and angle agrees quite
well with the experimental values of 143.21 pm and 119.54○ .
(b) The calculated values of the fundamental vibrational wavenumbers, and
illustrations of the displacements involved in the normal modes, are also
shown in Fig. 11.12. They correlate well with experimental values but
are about 35–80 cm−1 lower. SCF calculations often yield systematically
lower or higher values than experiment while approximately paralleling
the experimental to within an additive constant.
C2v, 119.26°
146.9 pm 146.9 pm
A1, 489.086 cm−1 A1, 1072.813 cm−1 B2, 1285.263 cm−1
Figure 11.12
11E Symmetry analysis of vibrational spectroscopy

D11E.1 Because benzene has a centre of symmetry the rule of mutual exclusion applies.
Therefore a particular normal mode will be observed in either the infrared or
in the Raman spectrum (or, possibly, in neither). The most complete character-
ization of the normal modes therefore requires the observation of both kinds
of spectra.
E11E.1(a) A mode is infrared active if it has the same symmetry species as one of the
functions x, y, and z; in this point group these span B1 + B2 + A1 . A mode is
Raman active if it has the same symmetry as a quadratic form; in this group
such forms span A1 + A2 + B1 + B2 . Therefore all of the normal modes are both
Raman and infrared active.
E11E.2(a) (i) H2 O belongs to the point group C 2v . Rather than considering all 9 dis-
placement vectors together it is convenient to consider them in sub-sets
of displacement vectors which are mapped onto one another by the op-
erations of the group. The x, y, and z vectors on the oxygen are not
mapped onto the displacements of the H atoms and so can be considered
separately. In fact, because these displacement vectors are attached to the
principal axis, they transform as the cartesian functions x, y, and z as
listed in the character table: that is as B1 + B2 + A1 .
Assuming the same axis system as in Fig. 11E.1 on page 432, the two x
displacements on the H atoms map onto one another, as do the two y
displacements, as do the two z displacements: however, the x, y, and z
displacements are not mixed with one another. For the two z displace-
ments the operation E leaves both unaffected so the character is 2; the C 2
operation swaps the two displacements so the character is 0; the σv (xz)
operation swaps the two displacements so the character is 0; the σv′ (yz)
operation leaves the two displacements unaffected so the character is 2.
The representation is therefore (2, 0, 0, 2), which is easily reduced by in-
spection to A1 + B2 . For the two y displacements the argument is essen-
tially the same, resulting in the representation (2, 0, 0, 2), which reduces
to A1 + B2 .
For the two x displacements the operation E leaves both unaffected so
the character is 2; the C 2 operation swaps the two displacements so the
character is 0; the σv (xz) operation swaps the two displacements so the
character is 0; the σv′ (yz) operation leaves the two displacements in the
same position by changes their direction, so the character is −2. The rep-
resentation is therefore (2, 0, 0, −2), which is easily reduced by inspection
to A2 + B1 .
The 9 displacements therefore transform as 3A1 + A2 + 2B1 + 3B2 .
The displacements include translations and rotations. For the point group
C 2v , x, y, and z transform as B1 , B2 , and A1 , respectively. The rotations
R x , R y , and R z transform as B2 , B1 , and A2 , respectively. Taking these

symmetry species away leaves just the normal modes as 2A1 + B2 .
A mode is infrared active if it has the same symmetry species as one of
the functions x, y, and z; in this point group these span B1 + B2 + A1 .
Therefore all of the normal modes are infrared active.
(ii) H2 CO is a straightforward extension of the case of H2 O as both molecules
belong to the point group C 2v . The H2 C portion lies in the same position
as H2 O, with the carbonyl O atom lying on the z axis (the principal axis).
The analysis therefore includes three more displacement vectors for the
O, and as they are connected to the principal axis they transform as the
cartesian functions x, y, and z, that is as B1 + B2 + A1 . The tally of
normal modes is therefore those for H2 O plus these three in addition:
3A1 + B1 + 2B2 . All these modes are infrared active.
E11E.3(a) The displacements include translations and rotations. For the point group C 2v ,
x, y, and z transform as B1 , B2 , and A1 , respectively. The rotations R x , R y , and
R z transform as B2 , B1 , and A2 , respectively. Taking these symmetry species
away leaves just the normal modes as 4A1 + A2 + 2B1 + 2B2 . These correspond
to 9 normal modes, which is the number expected for CH2 Cl2 .
P11E.1 (a) CH3 Cl has a C 3 axis (the principal axis) along the C–Cl bond, and three
σv planes, one passing along each C–H bond and containing the principal
axis. The point group is therefore C 3v .
(b) The molecule is non-linear and has N = 5, there are thus 3 × 5 − 6 = 9
normal modes.
(c) The task is to find the symmetry species spanned by the set of (x, y, z)
displacement vectors on each atom. The (x, y, z) displacement vectors on
the Cl can be considered separately and, as these vectors are connected to
the principal axis, their symmetry species is simply read from the char-
acter table as E + A1 . The (x, y, z) displacement vectors on the C behave
in the same way and so transform as E + A1 .
Consider the set of 9 (x, y, z) displacement vectors on the H atoms. Un-
der the operation E these are all unaffected so the character is 9; under
the C 3 operation they are all moved to new positions so the character is 0.
Next consider one of the σv planes: the (x, y, z) vectors on the H atoms
which do not lie in this plane are all moved and so contribute 0 to the
character. Now consider the H atom which lies in the plane, and arrange
a local axis system such that z points along the C–H bond, y lies in the
plane and x lies perpendicular to the plane. The effect of σv on (x, y, z)
is to transform it to (−x, +y, +z); two of the basis functions remain the
same and one changes sign, so the character is −1 + 1 + 1 = +1. The
representation formed from the 9 (x, y, z) displacement vectors on the
H atoms is thus (9, 0, 1). This is reduced using the reduction formula,
[10C.3a–386], to give 2A1 + A2 + 3E.
In total, the displacements therefore transform as 4A1 + A2 + 5E. Taking

away the translations, A1 +E, and the rotations, A2 +E, leaves the symme-
try species of the vibrations as 3A1 + 3E . As expected, there are 9 normal
modes (recall that the E modes are doubly degenerate).
(d) A mode is infrared active if it has the same symmetry species as one of
the functions x, y, and z; in this point group these span A1 + E. All the
normal modes are infrared active.
(e) A mode is Raman active if it has the same symmetry as a quadratic form;
in this point group these span A1 + E. All the normal modes are Raman
active.
11F Electronic spectra

D11F.1 A simple model for the energy of the HOMO–LUMO transition in a polyene
is discussed in Example 7D.1 on page 251. In this the energy levels of the π
electrons in a polyene are modelled by those of a particle in a one-dimensional
box of length L. If the polyene consists of n conjugated double bonds, the length
L may be written as L = nd, where d is the length of a single conjugated bond.
A molecule with n conjugated double bonds will have 2n π electrons which
will occupy the energy levels pairwise. Therefore the HOMO is the level with
quantum number n and the LUMO has quantum number n + 1. The energy of
the HOMO–LUMO transition is therefore
h2 (2n + 1)h 2
E n+1 − E n = [(n + 1)2 − n 2 ] =
8mL 2 8mL 2
For large n, (E n+1 − E n ) goes as n. However, L = nd, therefore overall (E n+1 −
E n ) goes as 1/n, and hence the wavelength of the transition goes as n. Thus,
increasing the number of conjugated double bonds will increase the wavelength
of the absorption; that is, shift it to the red.
The intensity of the transition will depend on the square of the transition dipole
moment, given by
L L
∫ ψ n+1 µ̂ x ψ n dx = (2/L) ∫ sin[(n + 1)πx/L] µ̂ x sin[nπx/L] dx
0 0
where µ̂ x is the operator for the dipole moment along x, and the normalized
wavefunctions ψ n = (2/L)1/2 sin(nπx/L) are used. The dipole moment oper-
ator is proportional to x, so setting aside the constants of proportion and using
(n + 1) ≈ n for large n, the required integral is
L
I = (2/L) ∫ x sin2 (nπx/L) dx
0
As noted above, L = nd therefore

nd
I = (2/nd) ∫ x sin2 (πx/d) dx
0
The integral is of the form of Integral T.11 with k = π/d and a = nd; it evaluates
to n 2 d 2 /4. Taking into account the normalization factor gives the final result
nd/2. The conclusion is therefore that the transition dipole moment increases
with n, the number of conjugated double bonds.
In summary, the expectation is that increasing the number of conjugated dou-
ble bonds will increase the wavelength of the absorption and also the intensity
of the absorption.
D11F.3 This is explained in Section 11F.1(a) on page 435.
D11F.5 The wavenumbers of the lines in the P, Q and R branches are given in [11F.7–
441]. Consider first the P branch ν̃ P (J) = ν̃ − (B̃′ + B̃)J + (B̃′ − B̃)J 2 , and recall
that (B̃′ + B̃) ≫ ∣(B̃′ − B̃)∣. As J increases the lines move to lower wavenumber
on account of the term −(B̃′ + B̃)J. However, as J becomes larger still the term in
J 2 becomes proportionately more and more important. If (B̃′ − B̃) > 0 this term
contributes to an increase in the wavenumber of the lines, and for sufficiently
large J it will overcome the −(B̃′ + B̃)J term and cause the lines to start to move
to higher wavenumber as J increases further. There will therefore be a lowest
wavenumber at which any line appears: this is the band head. If (B̃′ − B̃) < 0
the term in J 2 simply causes the lines to move to lower wavenumber and no
band head is formed.
Next consider the R branch ν̃ R (J) = ν̃ + (B̃′ + B̃)(J + 1) + (B̃′ − B̃)(J + 1)2 .
As J increases the lines move to higher wavenumber on account of the term
+(B̃′ + B̃)(J +1). However, as J becomes larger still the term in (J +1)2 becomes
proportionately more and more important. If (B̃′ − B̃) < 0 this term contributes
to an decrease in the wavenumber of the lines, and for sufficiently large J it
will overcome the +(B̃′ + B̃)(J + 1) term and cause the lines to start to move
to lower wavenumber as J increases further. There will therefore be a highest
wavenumber at which any line appears: this is the band head. If (B̃′ − B̃) > 0
the term in (J + 1)2 simply causes the lines to move to higher wavenumber and
no band head is formed.
The lines in the Q branch, ν̃ Q (J) = ν̃ + (B̃′ − B̃)J(J + 1), all appear at higher or
lower wavenumber than ν̃ depending on the sign of (B̃′ − B̃); no band heads
are formed.
If a parameter α is defined as α = B̃′ /B̃, so that B̃′ = α B̃, the wavenumbers of
the lines in the P and R branches can be written
[ν̃ P (J)−ν̃]/B̃ = −(1+α)J−(1−α)J 2 [ν̃ R (J)−ν̃]/B̃ = (1+α)(J+1)−(1−α)(J+1)2
These functions are plotted in Fig. 11.13 for representative values of α. If α < 1,
meaning that (B̃′ − B̃) < 0, the band head occurs in the R branch, but if α > 1,
meaning that (B̃′ − B̃) > 0, the band head occurs in the P branch.
The Gaussian functions are written e−α x /2 and e−α(x−a/2) /2 , where the param-
2 2
E11F.1(a)
eter α determines the width. To evaluate the normalizing factor for the function
20
P branch, α = 1.2
P branch, α = 0.8
10 R branch, α = 1.2
[ν̃(J)P or R − ν̃]/B̃
R branch, α = 0.8
0
−10
−20
0 2 4 6 8 10 12
J or (J + 1)
Figure 11.13
e−α x /2
2
requires the integral
+∞
e−α x dx
2
∫
−∞
which is of the form of Integral G.1 with k = α and evaluates to (π/α)1/2 . The
normalizing factor is therefore N 0 = (α/π)1/4 . The same factor applies to the
other Gaussian function as this is simply the same Gaussian shifted to a/2: the
area under the square of this function is the same.
The transition moment is given by the integral
+∞
xe−α x /2 −α(x−a/2)2 /2
2
I = (α/π)1/2 ∫ e dx
−∞
+∞
xe−α[x /2+(x−a/2)2 /2]
2
= (α/π)1/2 ∫ dx
−∞
+∞
xe−α[x /2+x 2 /2−x a/2+a 2 /8]
2
= (α/π)1/2 ∫ dx
−∞
+∞
xe−α[x −x a/2+a 2 /8]
2
= (α/π)1/2 ∫ dx
−∞
+∞
xe−α(x−a/4) e−α a /16
2 2
= (α/π)1/2 ∫ dx
−∞
The final equality above is verified by expanding out the square and recombin-
ing the terms. Taking out the constant factor and then writing x as (x − a/4) +
a/4 gives
+∞
I = (α/π)1/2 e−α a /16
[(x − a/4)e−α(x−a/4) + (a/4)e−α(x−a/4) ] dx
2 2 2
∫
−∞
The first term in the integral is an odd function, and so evaluates to zero. The
second term is simply a shifted Gaussian and, as before, the integral is form of
Integral G.1 with k = α and evaluates to (π/α)1/2
I = (α/π)1/2 e−α a /16

(a/4)(π/α)1/2 = (a/4)e−α a /16
2 2
The ‘width’ w of a Gaussian can be defined as the distance between the values
of the coordinates (±w/2) at which the height falls to half its maximum value.
Because the maximum of the function is 1, the value of α for a given width is
found by solving
= e−α(w/2) /2
2
1
2
hence α = (8 ln 2)/w 2
The exercise specifies that the Gaussian should have width a, so α = (8 ln 2)/a 2 .
With this the transition moment becomes
(a/4)e−α a /16
= (a/4)e−(8 ln 2)a /16a 2
= (a/4)e−(ln 2)/2
2 2
= (a/4)e−(ln 2 )
1/2
= (a/4)(1/21/2 ) = a/(4 × 21/2 )
E11F.2(a) A simple model for the wavelength of the HOMO–LUMO transition in a poly-
ene is discussed in Example 7D.1 on page 251. In this model the energy levels
of the π electrons in a polyene are modelled by those of a particle in a one-
dimensional box of length L. If the polyene consists of n conjugated double
bonds, the length L may be written as L = nd, where d is the length of a single
conjugated bond. A molecule with n conjugated double bonds will have 2n π
electrons which will occupy the energy levels pairwise. Therefore the HOMO
is the level with quantum number n and the LUMO has quantum number n+1.
The energy of the HOMO–LUMO transition is therefore
h2 (2n + 1)h 2
E n+1 − E n = [(n + 1)2 − n 2 ] =
8mL 2 8mL 2
For large n, E n+1 − E n goes as n. However, L = nd, therefore overall E n+1 −
E n goes as 1/n, and hence the wavelength of the transition goes as n. Thus,
increasing the number of conjugated double bonds will increase the wavelength
of the absorption; that is, shift it to the red. The transition at 243 nm is therefore
likely to be from 4, and that at 192 nm is likely to be from 5.
E11F.3(a) The electronic configuration of H2 is σ 2g . The two electrons are in the same
orbital and so must be spin paired, hence S = 0, and (2S + 1) = 1. Each σ
electron has λ = 0, thus Λ = 0 + 0 = 0, which is represented by Σ. Two electrons
with g symmetry have overall symmetry g × g = g. The σ orbital is symmetric
with respect to reflection in a plane containing the internuclear axis, therefore
the two electrons in this orbital are also overall symmetric with respect to this
mirror plane; this is indicated by a right-superscript +. The term symbol is
therefore 1 Σ+g .
E11F.4(a) The electronic configuration of Li2 + is 1σ 2g 1σ 2u 2σ 1g . The filled orbitals, 1σ 2g and

1σ 2u , make no contribution to Λ and S, so can be ignored. Therefore, only the
single electron in the 2σ g orbital needs to be considered: it has λ = 0 and
s = 21 , hence Λ = 0 and S = 12 (giving a left superscript of 2S + 1 = 2). The
symmetry with respect to inversion is g and with respect to reflection is +. The
term symbol is therefore 2 Σ+g .
E11F.5(a) The electronic configuration given is 1σ 2g 1σ 2u 1π 3u 1π 1g . The filled orbitals, 1σ 2g

and 1σ 2u , make no contribution to Λ and S, and so can be ignored. With three
electrons in a pair of degenerate π u orbitals, two of the spins must be paired
leaving one unpaired. There is another electron in a π g orbital. These two
electrons can be paired, giving S = 0 (a singlet, 2S + 1 = 1), or parallel, giving
S = 1 (a triplet, 2S + 1 = 3).
The three electrons in the π u have overall symmetry with respect to inversion
(parity) u × u × u = g × u = u. The remaining electron has g symmetry, so over-
all the state has symmetry u × g = u.
In summary, the multiplicity is 1 or 3 , and the parity is u .
E11F.6(a) (i) Allowed
(ii) Allowed
(iii) No allowed, ∆Σ = 2
(iv) Not allowed, + ↮ −
(v) Allowed
To evaluate the normalizing factor for the function e−ax /2
2
E11F.7(a) requires the integral
+∞
e−ax dx
2
∫
−∞
which is of the form of Integral G.1 with k = a and evaluates to (π/a)1/2 . The
normalizing factor is therefore N 0 = (a/π)1/4 . The same factor applies to the
function e−a(x−x 0 ) /2 as this is simply a Gaussian shifted to x 0 : the area under
2
the square of this function is the same.

The Franck–Condon factor is given by [11F.5–440] and involves the square of
integral of the product of the two wavefunctions
+∞ +∞
e−ax /2 −a(x−x 0 )2 /2
e−a[x /2+(x−x 0 )2 /2]
2 2
I = N 02 ∫ e dx = (a/π)1/2 ∫ dx
−∞ −∞
+∞
e−a[x /2+x 2 /2−x x 0 +x 02 /2]
2
= (a/π)1/2 ∫ dx
−∞
+∞ +∞
e−a[x −x x 0 +x 02 /2]
e−a(x−x 0 /2) e−ax 0 /4 dx
2 2 2
= (a/π)1/2 ∫ dx = (a/π)1/2 ∫
−∞ −∞
the final equality above is verified by expanding out the square and recombining
the terms. Taking out the constant factors gives
+∞
I = (a/π)1/2 e−ax 0 /4 ∫ e−a(x−x 0 /2) dx
2 2
−∞
as before, the integral is form of Integral G.1 with k = a and evaluates to (π/a)1/2
I = (a/π)1/2 e−ax 0 /4 (π/a)1/2 = e−ax 0 /4

2 2
The Franck–Condon factor is therefore I 2 = e−ax 0 /2 . As expected this factor is

2
a maximum of 1 when x 0 = 0, that is when the two functions are aligned, and
falls off towards zero as x 0 increases.
E11F.8(a) The Franck–Condon factor is given by [11F.5–440] and involves the square of
integral of the product of the two wavefunctions. The region over which both
wavefunctions are non-zero is from L/4 to L: this is the domain of integration
L
I = (2/L) ∫ sin(πx/L) sin(π[x − L/4]/L) dx
L/4
applying the identity sin A sin B = 21 [cos(A − B) − cos(A + B)] gives
L πx π(x − L/4) πx π(x − L/4)

I = (1/L) ∫ cos (
− ) − cos ( + ) dx
L/4 L L L L
L π 2πx π
= (1/L) ∫ cos ( ) − cos ( − ) dx
L/4 4 L 4
Next the identity cos(A − B) = cos A cos B + sin A sin B is used
L π 2πx π 2πx π
= (1/L) ∫ cos ( ) − [cos ( ) cos ( ) + sin ( ) sin ( )] dx
L/4 4 L 4 L 4
Recognising that cos π/4 = sin π/4 = (2)−1/2 allows this factor to be taken
L 2πx 2πx
= (1/L)2−1/2 ∫ 1 − cos ( ) − sin ( ) dx
L/4 L L
Each term is now integrated and evaluated between the limits
L 2πx L 2πx L
= (1/L)2−1/2 ∣x − sin ( )+ cos ( )∣
2π L 2π L L/4
L 2πL 2π(L/4)
= (1/L)2−1/2 [L − L/4 + { − sin ( ) + sin ( )
2π L L
2πL 2πL/4
+ cos ( ) − cos ( ) }]
L L
L
= (1/L)2−1/2 [3L/4 + { − sin(2π) + sin(π/2) + cos(2π) − cos(π/2)}]
2π
3 + 4/π
= (1/L)2−1/2 (3L/4 + L/π) = 2−1/2 (3/4 + 1/π) =
4 × 21/2
The Franck–Condon factor is I 2 = (1/32)(3 + 4/π)2 ; numerically this is 0.134.
E11F.9(a) The wavenumbers of the lines in the P branch are given in [11F.7–441], ν̃ P (J) =
ν̃ − (B̃′ + B̃)J + (B̃′ − B̃)J 2 . The band head is located by finding the value
of J which gives the smallest wavenumber, which can be inferred by solving
dν̃ P (J)/dJ = 0.
d
[ν̃ − (B̃′ + B̃)J + (B̃′ − B̃)J 2 ] = −(B̃′ + B̃) + 2J(B̃′ − B̃)
dJ
Setting the derivative to zero and solving for J gives
B̃′ + B̃
J head =
2(B̃′ − B̃)
A band head only occurs in the P branch if B̃′ > B̃.

E11F.10(a) Because B̃′ < B̃ a band head will occur in the R branch .
The wavenumbers of the lines in the R branch are given in [11F.7–441], ν̃ R (J) =
ν̃ + (B̃′ + B̃)(J + 1) + (B̃′ − B̃)(J + 1)2 . The band head is located by finding
the value of J which gives the largest wavenumber, which can be inferred by
solving dν̃ R (J)/dJ = 0.
d
[ν̃ + (B̃′ + B̃)(J + 1) + (B̃′ − B̃)(J + 1)2 ] = (B̃′ + B̃) + 2(J + 1)(B̃′ − B̃)
dJ
Setting the derivative to zero and solving for J gives
−(B̃′ + B̃) B̃ − 3B̃′

J head = ′
−1=
2(B̃ − B̃) 2(B̃′ − B̃)
With the data given

0.3540 − 3 × 0.3101
J head = = 6.56
2(0.3101 − 0.3540)
Assuming that it is satisfactory simply to round this to the nearest integer the
band head occurs at J = 7 .
E11F.11(a) The fact that a band head is seen in the R branch implies that B̃′ < B̃. It is shown
in Exercise E11F.7(b) that the band head in the R branch occurs at
B̃ − 3B̃′
J head = (11.3)
2(B̃′ − B̃)
This rearranges to
2J + 1
B̃′ = B̃ × (11.4)
2J + 3
A band head at J = 1 might arise from a value of J determined from eqn 11.3
anywhere in the range 0.5 to 1.5, followed by subsequent rounding. Using these
non-integer values of J in eqn 11.4 gives B̃′ in the range 30 cm−1 to 40 cm−1 .
The bond length in the upper state is longer than that in the lower state (a
longer bond means a larger moment of inertia and hence a smaller rotational
constant).
E11F.12(a) Assuming that the transition corresponds to that between the two sets of d
orbitals which are split as a result on the interaction with the ligands (Sec-
tion 11F.2(a) on page 443), the energy of the transition is the value of ∆ o . Hence
˜ o = 1/(700 × 10−7 cm) = 1.43 × 104 cm−1 or 1.77 eV . This value is very
∆
approximate as it does not take into account the energy involved in rearranging
the electron spins.
E11F.13(a) A rectangular wavefunction with value h between x = 0 and x = a is normal-

ized if the area under the square of the wavefunction is equal to 1: in this case
1 = ah 2 , hence h = a −1/2 . For the wavefunction which is non-zero between
x = a/2 and x = b the height is h ′ = (b − a/2)−1/2 . The region where the
wavefunctions are both non-zero is x = a/2 to x = a (because b > a); this is
the domain of integration. The transition moment is
1/2
a 1 a
∫ ψ i xψ f dx = ( ) ∫ x dx
a/2 a(b − a/2) a/2
1/2 1/2
1 a 1
=( ) ∣ 12 x 2 ∣ a/2 = ( ) 1
(a 2 − a 2 /4)
a(b − a/2) a(b − a/2) 2
1/2 1/2
1 3a 2 3 a3
=( ) = ( )
a(b − a/2) 8 8 b − a/2
P11F.1 The first transition is not allowed because it violates the spin selection rule,
∆S = 0; this transition has ∆S = 1. The second transition is not allowed because
it violates the selection rule for Λ, ∆Λ = 0, ±1: this transition has ∆Λ = 2.
P11F.3 Figure 11.14 is helpful in understanding the various quantities involved in this
problem. The pure electronic energy of the ground electronic state, that is the
energy at the equilibrium separation, is (as a wavenumber) T̃e (X). Likewise,
the pure electronic energy of the excited state is T̃e (B). The vibrational terms
of the ground state (in the harmonic approximation) are G̃ X (υ X ) = (υ X + 12 )ν̃ X ,
with these terms measured from T̃e (X). Likewise, those of the excited state are
G̃ B (υ B ) = (υ B + 12 )ν̃ B measured from T̃e (B).
The wavenumber of the 0–0 transition, ν̃ 00 , is therefore
ν̃ 00 = T̃e (B) − T̃e (X) + G̃ B (0) − G̃ X (0)

= T̃e (B) − T̃e (X) + 21 ν̃ B − 21 ν̃ X
The difference T̃e (B) − T̃e (X) is the value quoted as 6.175 eV; this is converted
to a wavenumber using the factor from inside the front cover
8065.5 cm−1 1
ν̃ 00 = (6.175 eV) × + 2 × (700 cm−1 ) − 12 × (1580 cm−1 )
1 eV
= 4.936 × 104 cm−1
P11F.5 (a) The photoelectron spectrum involves a transition from the ground state
of the molecule to an electronic state of the molecular ion. The energy
needed for the transition is measured indirectly by measuring the energy
of the ejected electron, but in all other respects the spectrum is interpreted
in the same way as electronic absorption spectra.
T̃e (B) υ′ = 0
ν̃ 00
T̃e (X) υ=0
Figure 11.14
In HBr only the ground vibrational state of the ground electronic state
will be significantly populated, so only transitions from this level need
be considered. In principle there can be transitions from this vibrational
level to a range of different vibrational levels of the upper electronic state
(a υ = 0 progression) and the intensities of these transitions will be gov-
erned by the Franck–Condon factors. As described in Section 11F.1(c)
on page 438, if the two electronic states have similar equilibrium bond
lengths the υ = 0 → υ′ = 0 transition will be the strongest, and then the
intensity will drop of quickly for higher values of υ′ . On the other hand, if
the upper state is displaced to the left or right, several transitions will have
significant Franck–Condon factors and several lines in the progression
will be observed.
In the photoelectron spectrum of HBr the band centred at about 16 eV
shows extensive structure which is interpreted as being due to several
lines of a vibrational progression. This is consistent with this band being
due to the removal of a bonding electron, resulting in the upper state (of
the ion) having a significantly longer bond length than the ground state.
(b) The band at around 11.5 eV shows two peaks: these are not due to vi-
brational fine structure, but to spin-orbit coupling in the molecular ion,
specifically such coupling associated with the Br atom. In the ion there
are unpaired electrons, so spin-orbit coupling in manifested in the spec-
trum. Removal of a nonbonding electron from a lone pair on the Br is
not expected to give a change in equilibrium bond length, so only the
υ = 0 → υ′ = 0 transition has significant intensity: there is no vibrational
progression.
P11F.7 The intensity of a transition depends on the transition dipole moment, in this
case given by the integral
L L
∫ ψ n µ̂ x ψ 1 dx = −e(1/L) ∫ sin(nπx/L) x sin(πx/L) dx
0 0
where the normalized wavefunctions ψ n = (2/L)1/2 sin(nπx/L) are used, and

n = 2 or 3. The integral is most easily evaluated by first using the given identity
L nπx πx nπx πx
I = −e(1/L) ∫ x [cos (
− ) − cos ( + )] dx
0 L L L L
L (n − 1)πx (n + 1)πx
= −e(1/L) ∫ x [cos ( ) − cos ( )] dx
0 L L
The integral is of the form of Integral T.13 with a = L
⎧
⎪
⎪ L2 L2
= −e(1/L)⎨ 2 2
[cos(n − 1)π − 1] + sin(n − 1)π
⎪
⎪ (n − 1) π (n − 1)π
⎩
⎫
L2 L2 ⎪
⎪
− 2 2
[cos(n + 1)π − 1] − sin(n + 1)π ⎬
(n + 1) π (n + 1)π ⎪
⎪
⎭
For the case n = 2
L2 L2 16Le
I 1,2 = −e(1/L) { 2
[−1 − 1] − 2 [−1 − 1]} =
π 9π 9π 2
For the case n = 3
L2 L2
I 1,3 = −e(1/L) { [1 − 1] + − [1 − 1]} = 0
4π 2 16π 2
Thus, as was to be shown, the transition dipole is non-zero for the transition
1 → 2, but zero for 1 → 3.
This result is obtained much more simply by rewriting the integral using x =
(x − a/2) + a/2
L
I = −e(1/L) ∫ sin(nπx/L) [(x − a/2) + a/2] sin(πx/L) dx
0
L
= −e(1/L) ∫ sin(nπx/L) (x − a/2) sin(πx/L) dx
0
L
− e(1/L) ∫ sin(nπx/L) (a/2) sin(πx/L) dx
0
The second integral is zero because the two eigenfunctions sin(nπx/L) and
sin(πx/L) are orthogonal for n ≠ 1. For the first integral the integrand is
a product of three functions which can all be classified as odd or even with
respect to x = a/2. For n = 3, sin(nπx/L) is even, (x − a/2) is odd, and
sin(πx/L) is even: the integrand is therefore odd overall, and hence when
integrated over a symmetrical interval the result is necessarily zero.
For n = 2, sin(nπx/L) is odd, (x − a/2) is odd, and sin(πx/L) is even: the inte-
grand is therefore even overall, and hence when integrated over a symmetrical
interval the result is not necessarily zero. What is not shown by this argument
is that the integral is non-zero: however, a quick sketch of the integrand shows
that it is negative everywhere, so the integral is non-zero.
P11F.9 The overlap integral for two 1s hydrogen (Z = 1) orbitals separated by a distance
R is
R 1 R 2 −R/a 0
S = [1 + + ( ) ]e
a0 3 a0
where a 0 is the Bohr radius. The transition moment, given as −eRS, is therefore
R 1 R 2 −R/a 0
µ = −eR [1 + + ( ) ]e
a0 3 a0
hence
R R 1 R 2 −R/a 0
−µ/ea 0 = [1 + + ( ) ]e
a0 a0 3 a0
Figure 11.15 shows a plot of −µ/ea 0 against R/a 0 . The maximum occurs at
R/a 0 ≈ 2.1, and the transition moment tends to zero at large distances simply
because the overlap also goes to zero in this limit.
1.0
−µ/ea 0
0.5
0.0
0 2 4 6 8 10
R/a 0
Figure 11.15
P11F.11 The adsorption at 189 nm is due to a π∗ ← π transition of the carbonyl group.

The weaker adsorption at 280 nm is due to a π∗ ← n transition of the carbonyl
group.
11G Decay of excited states

D11G.1 Intersystem crossing (ISC) is the process by which the excited singlet state (S1 )
makes a radiationless transition to a triplet state, T1 . Following this process
spontaneous emission may occur from T1 down to the ground electronic state
– this is phosphorescence. Both ISC and the phosphorescent transition are spin
forbidden and so only occur relatively slowly, and to the extent that the spin
selection rule is broken. Spin-orbit coupling is one of the effects that leads to
this rule being broken, and it is observed that the rates of these spin forbidden
transitions is enhanced by the presence of heavy atoms which have significant
spin orbit coupling.
In the present case, if the added iodide ion is able to interact with the chro-
mophore for a significant period of time (a time comparable with ISC or phos-
phorescence) then the spin-orbit coupling in the transient species may increase
the rate of ISC and/or of the phosphorescence, leading to an increase in the
intensity of the latter.
D11G.3 This is described in Section 11G.3 on page 449.
D11G.5 The characteristics of fluorescence which are consistent with the accepted mech-
anism are: (1) it ceases as soon as the source of illumination is removed; (2)
the time scale of fluorescence, 10−9 s, is typical of a process in which the rate
determining step is a spontaneous radiative transition between states of the
same multiplicity; (3) it occurs at longer wavelength (lower frequency) than
the exciting radiation; (4) its vibrational structure is characteristic of that of a
transition from the ground vibrational level of the excited electronic state to the
vibrational levels of the ground electronic state; and (5) the observed shifting,
and in some instances quenching, of the fluorescence spectrum by interactions
with the solvent.
E11G.1(a) This observed increase in the linewidth is a result of predissociation, as il-
lustrated in Fig. 11G.8 on page 449. Where the dissociative 5 Π u state crosses
the bound upper electronic state the possibility exists that molecules in the
upper electronic state will undergo radiationless transitions to the dissociative
state leading to subsequent dissociation. This process reduces the lifetime of
the excited states and so increases the linewidth of the associated transitions
(lifetime broadening, see Section 11A.2(b) on page 401).
E11G.2(a) (i) The vibrational fine structure of the fluorescent transition is determined
by the vibrational energy levels of the ground electronic state because the
transitions observed are from the ground vibrational level of the upper
electronic state to various vibrational levels of the ground electronic state.
(ii) No information is available about the vibrational levels of the upper elec-
tronic state because the spectrum only shows transitions from the ground
vibrational level of this state.
P11G.1 The anthracene fluorescence spectrum reflects the vibrational levels of the lower
electronic state. The given wavelengths correspond to wavenumbers 22 730 cm−1 ,
24 390 cm−1 , 25 640 cm−1 , and 27 030 cm−1 ; these indicate spacings of the
vibrational levels of (24 390 cm−1 ) − (22 730 cm−1 ) = 1660 cm−1 , and, by
similar calculations, 1250 cm−1 , and 1390 cm−1 .
The vibrational fine structure in the absorption spectrum reflects the vibra-
tional levels of the upper electronic state. The wavenumbers of the absorption
peaks are 27 800 cm−1 , 29 000 cm−1 , 30 300 cm−1 , and 32 800 cm−1 . The vibra-
tional spacings are therefore 1200 cm−1 , 1300 cm−1 , and 2500 cm−1 .
The fact that the fluorescent transitions are all at longer wavelength (lower
energy) that the absorption transitions is consistent with the loss of vibrational
energy (by collision induced radiationless decay) after the initial excitation of
the molecule. The vibrational fine structure in the absorption and fluorescence
spectra do not mirror one another, suggesting that the bonding in the ground
and excited electronic states are dissimilar.
P11G.3 (a) The resonant modes satisfy nλ/2 = L therefore λ = 2L/n and ν = nc/2L.
With the data given
n × (2.9979 × 108 m s−1 )

ν= = n × 150 MHz
2 × (1.00 m)
(b) The spacing of the modes is therefore 150 MHz .
P11G.5 The peak power is energy/(duration of pulse)

0.10 J
Ppeak = = 33 MW
3.0 × 10−9 s
where 1 W = 1 J s−1 is used. Consider a total time of 1 s: because the repetition
frequency is 10 Hz, there will be 10 pulses in this time.
total energy 10 × (0.10 J)

Pav = = = 1.0 W
total time 1s
The average power is very much less than the peak power.
I11.1 The electronic ground state of a closed-shell molecule transforms as the totally
symmetric irreducible representation. This is because all of the orbitals are
doubly occupied by spin paired electrons. The overall symmetry of a filled
orbital is found from the direct product Γ(i) × Γ(i) , where Γ(i) is the irreducible
representation as which the orbital transforms. When multiplied out, such a
product always includes the totally symmetric irreducible representation and it
is this symmetry species which needs to be combined with the anti-symmetric
spin function for two spin-paired electrons.
(a) Ethene belongs to the point group D 2h : assume that the molecule lies
in the x y-plane, with the C–C bond along x. The π bonding molecular
orbital transforms in the same way as the cartesian function z, that is as
B1u . The π∗ anti-bonding molecular orbital transforms in the same way
as the cartesian function xz, that is as B2g . The excited state (π)1 (π∗ )1
therefore has symmetry B1u × B2g = B3u (the direct product is found by
multiplying together the characters of the two irreducible representations,

as described in Section 10C.1(a) on page 385).
The transition moment is given by ∫ ψ f ∗ µ̂ψ i dτ, and the symmetry species
of the integrand is found using the direct product; µ̂ transforms as x, y, or
z, which in this case is B3u , B2u , or B1u . The integral is only non-zero if the
integrand transforms as the totally symmetric irreducible representation,
which is the case here when the component of the dipole is along x
ψf µ̂ x ψi
« « ª
B3u × B3u × Ag = Ag
Thus the π → π∗ transition is symmetry allowed , with the transition

dipole along x .
(b) A carbonyl group (as exemplified by that in methanal) is assumed to be-
long to the point group C 2v : assume that the H2 CO fragment lies in the
xz-plane, with the C–O bond along z. The π∗ anti-bonding molecular
orbital transforms in the same way as the cartesian function y, that is as
B2 .
A non-bonding electron on oxygen is usually considered to be in a 2px
orbital which transforms as B1 . When an electron is promoted from n to
π∗ , the excited state has symmetry B2 × B1 = A2 . The symmetry of the
integrand for the transition moment is
ψf µ̂ ψi
ª ª
A2 × Γx , y,z × A1 = A2 × Γx , y,z
Because Γx , y,z ≠ A2 this product is never A1 , so the integral is zero and

the transition is forbidden .
I11.3 The energy of the HOMO, E HOMO , is reported in the table below, based on
calculations performed with Spartan 10 using the DFT/B3LYP/6-31G* method.
The experimentally determined energy of the I2 –aromatic hydrocarbon charge
transfer bands is also given.
Figure 11.16 shows a plot of E HOMO against hν max ; the best-fit straight line is
also shown. There is a modest correlation between the two quantities.
hydrocarbon hν max /eV E HOMO /eV

benzene 4.184 −6.70
biphenyl 3.654 −5.91
naphthalene 3.452 −5.78
phenanthrene 3.288 −5.73
pyrene 2.989 −5.33
anthracene 2.890 −5.23
−5.5
E HOMO /eV −6.0
−6.5
2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2

hν max /eV
Figure 11.16
R/2
d
R
R/2
Figure 11.17
I11.5 (a) The geometry of the molecule is illustrated in Fig. 11.17.

The moment of inertia about the axis parallel to the symmetry axis, that
is the axis passing through the middle of the molecule and perpendicular
to the plane of the molecule, is given by
I∣∣ = ∑ m i r 2i = 3m H d 2
i
where r i is the perpendicular distance from the√axis to mass m i . From

√
the diagram is follows that (R/2)/d = cos 30○ = 3/2, hence d = R/ 3
√
I∣∣ = 3m H (R/ 3)2 = m H R 2
The moment of inertia perpendicular to the symmetry axis is indepen-
dent of where this axis is located provided it passes through the centre of
the molecule and lies in the plane of the molecule: one such convenient
choice is the dashed line shown in the figure.
I = 2m H (R/2)2 = 1
m R2
2 H
As expected, I∣∣ = 2I .

(b) The moment of inertia is related to the rotational constant through [11B.7–
408], B̃ = ħ/4πcI ; note that the rotational constant B̃ is always associated
with the moment of inertia perpendicular to the symmetry axis. It follows
that I = ħ/4πc B̃ and hence, using the result above, 12 m H R 2 = ħ/4πc B̃.
This rearranges to give the following expression for R
1/2
ħ
R=( )
2πm H c B̃
1/2
1.0546 × 10−34 J s
=( )
2π(1.67... × 10 kg) × (2.9979 × 1010 cm s−1 ) × (43.55 cm−1 )
−27
= 87.64 pm
where the mass in kg is given by (1.0079 m u )×(1.6605×10−27 kg)/(1 m u )

= 1.67... × 10−27 kg. An alternative value for R is found from the other
rotational constant. Here C̃ = ħ/4πcI∣∣ and hence m H R 2 = ħ/4πc C̃.
1/2
ħ
R=( )
4πm H c C̃
1/2
1.0546 × 10−34 J s
=( )
4π(1.67... × 10 kg) × (2.9979 × 1010 cm s−1 ) × (20.71 cm−1 )
−27
= 89.87 pm
(c) With the given value of R, the rotational constant is computed from B̃ =
ħ/4πcI = ħ/2πm H cR 2
1.0546 × 10−34 J s
B̃ =
2π(1.67... × 10−27 kg) × (2.9979 × 1010 cm s−1 ) × (87.32 × 10−12 m)2
= 43.87 cm−1
The other rotational constant is just half of this, C̃ = 21.94 cm−1

−1/2
(d) For a harmonic oscillator the vibrational frequency goes as m eff , where
m eff is the effective mass. The value of the effective mass depends on the
vibration in question, but in this case as all the atoms are the same it is
reasonable to assume that the effective mass of H3 + will be proportional
to m H , and that of D3 + will be proportional to m D . It therefore follows
that
m H 1/2
ν̃ D = ν̃ H ( )
mD
1.0079 1/2
= (2521.6 cm−1 ) × ( ) = 1783.8 cm−1
2.0140
The rotational constants are inversely proportional to the moment of in-

ertia, and for this molecule all the atoms are the same so the rotational
constant goes as m−1 .

mH
B̃ D = B̃ H
mD
1.0079
= (43.55 cm−1 ) × = 21.79 cm−1
2.0140
A similar calculation gives C̃ D = 10.36 cm−1 .
−1/2
I11.7 (a) For a harmonic oscillator the vibrational frequency goes as m eff , where
m eff is the effective mass; for a diatomic the effective mass is simply 12 m.
It therefore follows that
m16 O 1/2
ν̃ 18O2 = ν̃ 16O2 ( )
m18 O
15.9949 1/2
= (844 cm−1 ) × ( ) = 796 cm−1
17.9992
(b) The bond order of O2 is 2, and to form the anions electrons are added
to the anti-bonding π g molecular orbital, thereby decreasing the bond
order to 1.5 for O2 – and 1 for O2 2 – . The steady decrease in the bond order
is matched by the steady decrease in the vibrational frequency. There is
thus a correlation between bond strength and vibrational frequency, as
expected.
(c) The observed vibrational frequency of O2 bound to haemerythrin most
closely matches that for O2 2 – , so of the alternatives offered Fe3+ 2 O2
2–
seems the most likely.

(d) The observation of two bands attributable to O–O stretching implies that
the O2 is bound is such a way that the two oxygen atoms are no longer
equivalent. If this is the case, when the isotopologue 16O 18O is used,
two different frequencies will result because the two ends of the oxygen
molecule are now distinguished. This eliminates structures 7 and 8.
I11.9 Expressed in terms of the absorbance A the Beer–Lambert law is given by [11A.9c–
398], A = ε[J]L. To convert the given absorbance into a molar absorption
coefficient requires [J] in mol dm−3 , which is computed using the perfect gas
law, pV = nRT, rearranged to n/V = p/RT. Assuming that the quoted com-
position of CO2 , ‘2.1 per cent’, refers to a mole per cent, x CO2 = 0.021, so
p CO2 = 0.021×(1.00 bar) = 0.021 bar. In the calculation of n/V it is convenient
to use R = 8.3145 × 10−2 dm3 bar K−1 mol−1
n p 0.021 bar
= =
V RT (8.3145 × 10−2 dm3 bar K−1 mol−1 ) × (298 K)
= 8.47... × 10−4 mol dm−3
The given expression for A(ν̃) is therefore converted to an expression for the
molar absorption coefficient using
ε(ν̃) = A(ν̃)/[J]L = A(ν̃)/[(8.47... × 10−4 mol dm−3 ) × (10 cm)]
= (1.17... × 102 mol−1 dm3 cm−1 ) × A(ν̃)
(a) Graphs of A(ν̃) and ε(ν̃) are shown in Fig. 11.18. This band is likely to be
due to the fundamental transition of the antisymmetric stretching normal
mode of CO2 , which has ν̃ 2 = 2349 cm−1 . The broad features are the
unresolved P and R branches; no Q branch is expected for this mode. The
principal contribution to the linewidth of an infrared transition is likely
to be pressure broadening.
ε(ν̃)/mol−1 dm3 cm−1

2.0
1.5 200
A(ν̃)
1.0
100
0.5
0.0 0
2 300 2 350 2 400 2 300 2 350 2 400
−1 −1
ν̃/cm ν̃/cm
Figure 11.18
(b) Expressions for the wavenumbers of the lines in the P and R branches are
given in [11C.13a–423] and [11C.13c–423]
ν̃ P (J) = ν̃ − 2B̃J ν̃ R (J) = ν̃ + 2B̃(J + 1)
Here ν̃ = 2349 cm−1 . As described in Section 11B.4 on page 415, because

16
O is a boson, only even rotational states are occupied (and, in fact, odd
rotational states in the first excited vibrational state of the anti-symmetric
stretch), so in the above expression J takes the values 0, 2, 4 . . .. The
rotational constant is computed from [11B.7–408], B̃ = ħ/4πcI, where
I = 2m O R 2 . With the data given
I = 2 × (15.9949 m u )×(1.6605 × 10−27 kg)×(116.2 × 10−12 m)2 /(1 m u )

= 7.17... × 10−46 kg m2
1.0546 × 10−34 J s
B̃ = = 0.3903 cm−1
4π×(2.9979 × 1010 cm s−1 ) × (7.17... × 10−46 kg m2 )
The intensity of the transition from level J is proportional to the popula-
tion of that level, taking into account the degeneracy
intensity ∝ (2J + 1)e−hc B̃/k T
Using these expressions the positions and intensities of the lines in the P
and R branches are computed and a synthetic spectrum constructed by
assuming a linewidth and lineshape (here arbitrarily taken as a Gaussian).
Two such spectra are shown in Fig. 11.19: in (a) the linewidth has been
chosen so that the lines are well resolved; in (b) a much wider line is used
so that the lines in the P and R branches merge into a broad contour. The
spectrum in (b) is roughly similar to that in Fig. 11.18, but the asymmetry
between the two branches is not reproduced by the calculation.
(a) (b)
2300 2320 2340 2360 2380 2400 2300 2320 2340 2360 2380 2400
~ ~
v / cm−1 v / cm−1
Figure 11.19
(c) The transmittance T is defined as I/I 0 and hence from the Beer–Lambert
law [11A.8–397], T = e−ε[J]L ; it follows that log T = −A. The concentration
of CO2 is computed as in (a) using [CO2 ] = x CO2 p atm /RT; both the
atmospheric pressure and temperature vary with the height, and therefore
so will the concentration. The pressure varies with height according to
p(h) = p 0 e−h/H , where the scale height H is about 8000 m. The total
absorbance up to height h 0 is therefore given by integral
h0 h0 x CO2 p
A=∫ ε[CO2 ] dh = ε ∫ dh
0 0 RT
h0 x CO2 p 0 e−h/H
= ε∫ dh
0 R(288 − 0.0065h)
h0 e−h/H
= εx CO2 p 0 ∫ dh
0 R(288 − 0.0065h)
where h is in m.
This integral cannot be solved by hand, but some work with mathematical
software and a typical value of ε shows that the absorbance exceeds 1 by
the time h 0 = 30 m. Over such a small height it is safe to assume that the
pressure and temperature are constant, in which case the calculation of
the absorbance is much simpler
x CO2 p
A = ε[CO2 ]h 0 = ε h0
RT
= [ε/(mol−1 dm3 cm−1 )]
(3.3 × 10−4 ) × (1 bar)
× × (h 0 /cm)
(8.3145 × 10−2 dm3 bar K−1 mol−1 ) × (288 K)
= (1.378 × 10−5 ) × [ε/(mol−1 dm3 cm−1 )] × (h 0 /cm)
The transmittance is T = 10−A .

Figure 11.20 shows plots of the transmittance as a function of height for
some representative values of the molar absorption coefficient. For the
maximum value of ε seen in this absorption band (refer to Fig. 11.18) the
transmittance drops to 0.1 after less than 5 m. For a value of ε more typical
of the average (say ε = 100 mol−1 dm3 cm−1 ), the transmission drops to
0.1 after about 7.5 m. Even for values of ε typical of the extremities of
the band, the transmission has fallen to 0.1 within 20 m or so. A surface
plot of the transmission as a function of both wavenumber and height is
shown in Fig. 11.21.
1.0
ε = 200 mol−1 dm3 cm−1
0.8 ε = 100
ε = 50
0.6 ε = 25
T
0.4
0.2
0.0
0 5 10 15 20 25 30
h 0 /m
Figure 11.20
~
v /cm−1
h0 /m
Figure 11.21
I11.11 In Problem P11A.6 it is shown that the integrated absorption coefficient for a
Gaussian line shape is A = 1.0645 ε max ∆ν̃ 1/2 , where ∆ν̃ 1/2 is the width at half
height. Interpolating, by eye, a smooth curve across the band centred at about
280 nm, gives ε max = 250 mol−1 dm3 cm−1 . The molar absorbance drops to half
this value at about 270 nm, which is 3.70×104 cm−1 and at about 310 nm, which
is 3.23 × 104 cm−1 , giving a width of 4700 cm−1 . The integrated absorption
coefficient is therefore
A = 1.0645 × (250 mol−1 dm3 cm−1 ) × (4700 cm−1 )

= 1.25 × 106 mol−1 dm3 cm−2
The transition moment is given by ∫ ψ ∗f µ̂ψ i dτ; µ̂ transforms as x, y, or z, which

in this case is B1 , B2 , or A1 . The integral is only non-zero if the integrand trans-
forms as the totally symmetric irreducible representation, which is determined
by computing the direct product
ψf µ̂ x ψi
© ª
Γf × Γx , y,z × A1 = Γf × Γx , y,z
where that fact that the direct product with the totally symmetric irreducible
representation has no effect is used. The only way this product can contain
the totally symmetric irreducible representation is if the irreducible represen-
tation of the final state, Γf , is equal to the irreducible representation of Γx , y,z .
Thus, transitions from the A1 ground state to A1 , B1 , or B2 excited states are
allowed.
12 Statistical
thermodynamics
12A The Boltzmann distribution
D12A.1 In terms of molecular energy levels the thermodynamic temperature is the one
quantity that determines the most probable populations of the states of the
system at thermal equilibrium, as discussed in Section 12A.1(b) on page 461.
The equipartition theorem allows a connection to be made between the temper-

ature as understood in statistical thermodynamics and the empirical concept
of temperature which arises in classical thermodynamics. Temperature is a
measure of the intensity of thermal energy, and is directly proportional to the
mean energy for each quadratic contribution to the energy (provided that the
temperature is sufficiently high).
D12A.3 The population of a state is the number of molecules in a sample that are in
that state. The configuration of a system is a list of populations in order of the
energy of the corresponding states. For example, {N−3, 2, 1, 0, . . .} is a possible
configuration of a system of N molecules in which all but three molecules are
in the ground state, two are in the next highest state, and one in the state above
that. The weight of a configuration is the number of ways a given configuration
can be achieved, and is given by [12A.1–461]. When N is large (as it is for
any macroscopic sample), the most probable configuration has a much greater
weight, that is it is more probable, than any other configuration. Under such
circumstances it can be assumed that the configuration adopted by the system
is this most probable configuration.
E12A.1(a) The Boltzmann population ratio is given by [12A.13a–464], N i /N j = e−β(ε i −ε j ) .

This is rearranged to β = − ln(N i /N j )/∆ε , where ∆ε = (ε i − ε j ). Substituting
440 12 STATISTICAL THERMODYNAMICS
β = 1/(kT) and rearranging for T gives

∆ε hc ν̃
T =− =−
k ln (N 1 /N 0 ) k ln (N 1 /N 0 )
(6.6261 × 10 J s) × (2.9979 × 1010 cm s−1 ) × (400 cm−1 )
−34
=−
(1.3806 × 10−23 J K−1 ) × ln(1/3)
= 524 K
E12A.2(a) The Boltzmann population ratio for degenerate energy levels is given by [12A.13b–
464], N i /N j = (g i /g j )e−β(ε i −ε j ) . The rotational term of a linear rotor is given
by [11B.14–410], F̃(J) = B̃J(J + 1) and, as explained in Section 11B.1(c) on page
410, its degeneracy is given as g J = 2J + 1. The rotational energy is related to
the rotational term as ε J = hc F̃(J). Therefore
N5 2 × 5 + 1
= × e−hc B̃[5×(5+1)−0×(0+1)]/k T = 11 × e−30 B̃hc/k T
N0 2 × 0 + 1
using kT/hc = 207.224 cm−1 at 298.15 K (from inside the front cover)
N5 −1 −1
= 11 × e−30×(2.71 cm )/(207.224 cm ) = 7.43
N0
E12A.3(a) The Boltzmann population ratio is given by [12A.13a–464], N i /N j = e−β(ε i −ε j ) .

This is rearranged to β = − ln(N i /N j )/∆ε , where ∆ε = (ε i − ε j ). Substituting
β = 1/(kT) and rearranging for T gives
∆ε hc ν̃
T =− =−
k ln (N 1 /N 0 ) k ln (N 1 /N 0 )
(6.6261 × 10 J s) × (2.9979 × 1010 cm s−1 ) × (540 cm−1 )
−34
=−
(1.3806 × 10−23 J K−1 ) × ln(10%/90%)
= 354 K
E12A.4(a) The weight of a configuration is given by [12A.1–461], W = N!/(N 0 !N 1 !N 2 !...),

thus
16!
W= = 21 621 600
0! × 1! × 2! × 3! × 8! × 0! × 0! × 0! × 0! × 2!
E12A.5(a) (i) 8! = 8 × 7 × 6 × 5 × 4 × 3 × 2 × 1 = 40 320 .
(ii) Stirling’s approximation for x ≫ 1 is given by [12A.2–461], ln (x!) ≈
x ln x − x. This is rearranged to x! ≈ e(x ln x−x) , thus
8! ≈ e(8×ln 8−8) = 5.63 × 103 .
(iii) Using the more accurate version of Stirling’s approximation
8! ≈ (2π)(1/2) × 8(8+1/2) × e−8 = 3.99 × 104
E12A.6(a) The Boltzmann population ratio is given by [12A.13a–464], N i /N j = e−β(ε i −ε j ) ,

where β = 1/(kT). At infinite temperature β becomes zero, therefore the
relative populations of two levels N 1 /N 0 = e−0 = 1 .
P12A.1 (a) There is no configuration in which the molecules are distributed evenly
over the states and which, at the same time, satisfies the constraint that
the energy is 5ε.
(b) The energy of a configuration is E/ε = N 1 + 2N 2 + 3N 3 . . ., and the weight
of a configuration is given by [12A.1–461], W = N!/(N 0 !N 1 !N 2 !...). The
configurations satisfying the total energy constraint E = 5ε are
N0 N1 N2 N3 N4 N5 W
4 0 0 0 0 1 5
3 1 0 0 1 0 20
3 0 1 1 0 0 20
2 2 0 1 0 0 30
2 1 2 0 0 0 30
1 3 1 0 0 0 20
0 5 0 0 0 0 1
Hence the most probable configurations are {2,2,0,1,0,0} and {2,1,2,0,0,0} .
P12A.3 The energy of a configuration is E/ε = N 1 + 2N 2 + 3N 3 . . ., and the weight of a

configuration is given by [12A.1–461], W = N!/(N 0 !N 1 !N 2 !...). There are a very
large number of possible configurations, but one way of selecting an interesting
and diverse set of these is to consider configurations in which N 0 is 19, then 18,
then 16 and so on.
N 0 N 1 N 2 N 3 N 4 N 5 N 6 N 7 N 8 N 9 N 10 W
19 0 0 0 0 0 0 0 0 0 1 20
18 1 0 0 0 0 0 0 0 1 0 380
17 1 1 0 0 0 0 0 1 0 0 6 840
16 1 1 1 1 0 0 0 0 0 0 116 280
15 2 2 1 0 0 0 0 0 0 0 465 120
14 3 2 1 0 0 0 0 0 0 0 2 325 600
13 4 3 0 0 0 0 0 0 0 0 2 713 200
12 6 2 0 0 0 0 0 0 0 0 7 054 320
11 8 1 0 0 0 0 0 0 0 0 1 511 640
10 10 0 0 0 0 0 0 0 0 0 184 756
Of the ones listed in the table the configuration with the greatest weight is
{12, 6, 2, 0, 0, 0, 0, 0, 0, 0, 0} .
The Boltzmann population ratio is given by [12A.13a–464], N i /N 0 = e−βε i ,
where it is assumed that ε 0 = 0. It follows that ln N i /N 0 = −βε i = −iβε, and
hence βε = (ln N i /N 0 )/(−i). For i = 1, βε = (ln 6/12)/(−1) = 0.693; for
i = 2, βε = (ln 2/12)/(−2) = 0.896. Taking an average of these two values gives
βε = 0.795 and hence T = ε/(0.795k) .
With this value for the temperature the populations predicted by the Boltz-
mann distribution are
N i /N 0 = e−βε i = e−i βε = e−0.795×i
Therefore N 1 /N 0 = 0.452, N 2 /N 0 = 0.204, N 3 /N 0 = 0.092. For the most

probable distribution given above these ratios are N 1 /N 0 = 0.500, N 2 /N 0 =
0.167, N 3 /N 0 = 0, which are roughly comparable.
P12A.5 The Boltzmann population ratio for degenerate energy levels is given by [12A.13b–
464], N i /N j = (g i /g j )e−β(ε i −ε j ) . Taking logarithms gives
ln(N i /N j ) = ln(g i /g j ) − β(ε i − ε j ) hence β = − ln[(N i /N j )(g j /g i )]/∆ε
where ∆ε = (ε i − ε j ). Substituting β = 1/(kT) and rearranging for T gives
∆ε hc ν̃
T=− =−
k ln[(N 1 /N 0 )(g 0 /g 1 )] k ln [(N 1 /N 0 )(g 0 /g 1 )]
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (450 cm−1 )
=−
(1.3806 × 10−23 J K−1 ) × ln [(30%/70%) × (2/4)]
= 420 K
The populations of the electronic states do not correspond to the translational

temperature; therefore the electronic states are not in equilibrium with the
translational states.
P12A.7 The Boltzmann population ratio is given by [12A.13a–464], N i /N j = e−β(ε i −ε j ) .

The energy ε i is interpreted as that at height h, and ε j is interpreted as that at
height 0, giving
N(h)/N(0) = e−β(m g h−m g×0) = e−βm g h
From the perfect gas law, pV = nRT, and because N ∝ n, it follows that p ∝ N,
and therefore p(h)/p 0 = N(h)/N(0). Because β = 1/(kT), m = M/N A , and
k = R/N A it follows that
mgh (M/N A )gh M gh h
βmgh = = = =
kT (R/N A )T RT H
where H = RT/M g is used. Hence
p(h)/p 0 = e−h/H and so p(h) = p 0 e−h/H
From the perfect gas law N = N/V ∝ p, therefore for O2

M gh
N(h)/N0 = exp (− )
RT
(32.00 × 10−3 kg mol−1 ) × (9.807 m s−2 ) × (8.0 × 103 m)
= exp (− )
(8.3145 J K−1 mol−1 ) × (298 K)
= 0.36
and for H2 O
N(h) M gh
= exp (− )
N0 RT
(18.02 × 10−3 kg mol−1 ) × (9.807 m s−2 ) × (8.0 × 103 m)
= exp (− )
(8.3145 J K−1 mol−1 ) × (298 K)
= 0.57
In these calculations the temperature is taken as 298 K and is assumed to be

constant with height, which is in fact not the case.
12B Partition functions

D12B.1 As discussed in Section 12B.2(b) on page 470, the symmetry number is the
number of indistinguishable orientations that the molecule can be rotated into.
This factor is needed in order to avoid counting contributions to the rotational
partition function which are forbidden by symmetry considerations arising
from the effects of nuclear spin (Section 11B.4 on page 415).
If the partition function is computed term-by-term then the symmetry number
is not needed because those terms which are forbidden are simply omitted.
However, in the high-temperature limit in which many terms are included, it
is convenient to allow all terms to contribute to the sum and then compensate
for those which should not have been included by division by the symmetry
number.
D12B.3 It is possible for there to be different wavefunctions which have the same en-
ergy: such wavefunctions are said to be degenerate. If this is the case, for a
given ‘energy level’, that is a given value of the energy, there are several ‘states’
each of which is distinct but has the same energy.
The partition function is computed as a sum over the states. However, because
degenerate states have the same energy, the sum may be computed as a sum over
energy levels, as long as the degeneracy g i of each level is taken into account.
q = ∑ e−βε i = ∑ g i e−βε i
states i levels i
E12B.1(a) The rotational partition function of a heteronuclear diatomic is given by [12B.11–
470], q R = ∑ J (2J + 1)e−β hc B̃ J(J+1) . This is evaluated explicitly by summing
successive terms until they become too small to affect the result to a given level
of precision. The partition function in the high-temperature limit is given by
[12B.12a–471], q R = kT/hc B̃. For the data given it follows that
k×T (1.3806 × 10−23 J K−1 ) × T

qR = =
hc B̃ (6.6261 × 10 J s) × (2.9979 × 1010 cm s−1 ) × (1.931 cm−1 )
−34
= (0.359... K−1 ) × T
The values of q R computed in these two different ways are compared in Fig. 12.1.
The high temperature limit becomes accurate to within 5 % of the exact solution
at around 18 K .
10
qR
exact
high T limit
0
0 5 10 15 20 25 30
T/K
Figure 12.1
E12B.2(a) The partition function is given by [12B.1b–466], q R = ∑ J g J e−βε J , where the

degeneracy is given as g J = (2J + 1)2 , as explained in Section 11B.1(c) on page
410, and ε J is given by [12B.1b–466], ε J = hc B̃J(J+1). This is evaluated explicitly
by summing successive terms until they become too small to affect the result
to a given level of precision.
The partition function in the high-temperature limit is given by [12B.12b–472],
q R = (kT/hc)3/2 (π/ÃB̃ C̃)1/2 = π 1/2 (kT/hc B̃)3/2 , because for a spherical rotor
B̃ = Ã = C̃. Ignoring the role of the nuclear spin means that all J states are
accessible and have equal weight. For the data given it follows that
k × T 3/2
q R = π 1/2 ( )
hc B̃
3/2
(1.3806 × 10−23 J K−1 )
= π 1/2 ( ) × T 3/2
(6.6261 × 10 J s)×(2.9979 × 1010 cm s−1 )×(5.241 cm−1 )
−34
= (0.0855... K−3/2 ) × T 3/2
The values of q R computed in these two different ways are compared in Fig. 12.2.
The high temperature limit becomes accurate to within 5 % of the exact solution
at around 37 K .
E12B.3(a) (i) CO is a heteronuclear diatomic and so has σ = 1 .

30
20
qR 10
exact
high T limit
0
0 10 20 30 40 50
T/K
Figure 12.2
(ii) O2 is a homonuclear diatomic and so has σ = 2 .

(iii) H2 S has a twofold rotational axis bisecting the H–S–H angle; rotation
about this axis interchanges two identical hydrogen atoms, therefore σ = 2 .
(iv) SiH4 is tetrahedral and so has the same symmetry number as CH4 : σ = 12 .
(v) CHCl3 has a threefold rotational axis along the C–H bond; rotation about
this axis interchanges three identical chlorine atoms, therefore σ = 3 .
E12B.4(a) The rotational partition function of an asymmetric rotor is given by [12B.14–
473], q R = (1/σ)(kT/hc)3/2 (π/ÃB̃C̃)1/2 , where σ is the symmetry number.
For a molecule with high symmetry the simplest was to determine the symme-
try number is to count the total number of rotational symmetry operations, C n ,
listed in the character table of the point group to which the molecule belongs,
which in this case is D 2h . For this group the rotational operations are (E, C 2x ,
y
C 2 , C 2z ), and therefore σ = 4.
1 kT 3/2 π 1/2
qR = ( ) ( )
4 hc ÃB̃C̃
3/2
1 (1.3806 × 10−23 J K−1 ) × (298.15 K)
= ×( )
4 (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
1/2
π
×( )
(4.828 cm−1 ) × (1.0012 cm−1 ) × (0.8282 cm−1 )
= 660.6
E12B.5(a) The vibrational partition function is given by [12B.15–474], q V = 1/(1−e−β hc ν̃ ),

where β = 1/kT. The high-temperature approximation is given by [12B.16–
475], q V ≈ kT/hc ν̃.
k×T (1.3806 × 10−23 J K−1 ) × T
=
hc ν̃ (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (323.2 cm−1 )
= (2.15... × 10−3 K−1 ) × T
The values of q V computed using these two different expressions are compared
in Fig. 12.3. The high temperature limit becomes accurate to within 5 % of the
exact solution at 4500 K .
qV 10
exact
high T limit
0
0 1 000 2 000 3 000 4 000 5 000
T/K
Figure 12.3
E12B.6(a) The vibrational partition function for each mode is given by [12B.15–474], q V =
1/(1 − e−β hc ν̃ ), where β = 1/kT. The overall vibrational partition function
is the product of the partition functions of the individual modes; the bend is
included twice as it is doubly degenerate.
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )

hcβ =
(1.3806 × 10−23 J K−1 ) × (500 K)
= 2.87... × 10−3 cm
−1
q V1 = (1 − e−hc β ν̃ 1 )
−3 −1
cm)×(658 cm−1 )
= (1 − e−(2.87 ...×10 ) = 1.17...
Similarly
−3 −1
cm)×(397 cm−1 )
q V2 = (1 − e−(2.87 ...×10 ) = 1.46...
−3 −1
cm)×(1535 cm−1 )
q V3 = (1 − e−(2.87 ...×10 ) = 1.01...
q V = q V1 × (q V2 )2 × q V3 = (1.17...) × (1.46...)2 × (1.01...) = 2.57
E12B.7(a) The vibrational partition function for each mode is given by [12B.15–474], q V =
1/(1−e−β hc ν̃ ), where β = 1/kT. The overall vibrational partition function is the
product of the partition functions of the individual modes, taking into account
the stated degeneracies.
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )

hcβ =
(1.3806 × 10−23 J K−1 ) × (500 K)
= 2.87... × 10−3 cm
−1
q V1 = (1 − e−hc β ν̃ 1 )
−3 −1
cm)×(459 cm−1 )
= (1 − e−(2.87 ...×10 ) = 1.36...
Similarly
−3 −1
cm)×(217 cm−1 )
q V2 = (1 − e−(2.87 ...×10 ) = 2.15...
−3 −1 −1
q V3 = (1 − e−(2.87 ...×10 cm)×(776 cm )
) = 1.12...
−3 −1
cm)×(314 cm−1 )
q V4 = (1 − e−(2.87 ...×10 ) = 1.68...
q V = q V1 × (q V2 )2 × (q V3 )3 × (q V4 )3
= (1.36...) × (2.15...)2 × (1.12...)3 × (1.68...)3 = 42.1 .
E12B.8(a) The partition function is given by [12B.1b–466], q = ∑ i g i e−βε i , where g i is

degeneracy and the corresponding energy is given as ε i = hc ν̃ i . At T = 1900 K
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )

βhc = = 7.57... × 10−4 cm
(1.3806 × 10−23 J K−1 ) × (1900 K)
Therefore the electronic partition function is
q E = g 0 + g 1 × e−βε 1 + g 2 × e−βε 2
−4 −1 −4
cm)×(3500 cm−1 )
= 4 + 1×e−(7.57 ...×10 cm)×(2500 cm ) + 2×e−(7.57 ...×10
= 4 + 0.150... + 0.141... = 4.291 (12.1)
The population of level i with degeneracy g i is N i = (N g i /q)e−βε i , therefore

the relative populations of the levels are proportional to g i e−βε i , which are the
terms in eqn 12.1. Thus the populations, relative the ground state are
N 0 /N 0 ∶ N 1 /N 0 ∶ N 2 /N 0 = 4/4 ∶ (0.150.../4) ∶ (0.141.../4)

= 1 ∶ 0.0376 ∶ 0.0353
E12B.9(a) (i) The thermal wavelength is defined in [12B.7–469], Λ = h/(2πmkT)1/2 .

Because the mass of a molecule m is m = M/N A and k = R/N A it follows
that
h hN A
Λ= =
[2π(M/N A )(R/N A )T]1/2 (2πMRT)1/2
(6.6261 × 10−34 J s) × (6.0221 × 1023 mol−1 )

Λ(300 K) =
[2π×(0.150 kg mol−1 )×(8.3145 J K−1 mol−1 ) × (300 K)]1/2
= 8.22... × 10−12 m = 8.23 × 10−12 m
Similarly, Λ(3000 K) = 2.60 × 10−12 m
(ii) The translational partition function in three dimensions is given by [12B.10b–

469], q T = V /Λ 3 .
q T (300 K) = (1.00 × 10−6 m3 )/(8.22... × 10−12 m)3 = 1.78 × 1027

q T (3000 K) = 5.67 × 1028
E12B.10(a) The translational partition function in three dimensions is given by [12B.10b–

469], q T = V /Λ 3 , where Λ is the thermal wavelength defined in [12B.7–469],
Λ = h/(2πmkT)1/2 .
3 3
qHT 2 V /Λ H2 3 Λ He h/(2πm He kT)1/2 m H2 3/2
= = ( ) = ( ) = ( )
T
qHe V /Λ He 3 Λ H2 h/(2πm H2 kT)1/2 m He
Because the mass of a molecule m is m = M/N A it follows that
3/2
qHT 2 M H2 3/2 2 × 1.0079 g mol−1
= ( ) = ( ) = 0.358
T
qHe M He 4.00 g mol−1
E12B.11(a) The rotational partition function of a symmetric linear rotor is given by [12B.13a–
472], q R = kT/(2hc B̃), where the rotational constant is defined in [11B.7–408],
B̃ = ħ/(4πcI). The moment of inertia of a diatomic is I = µR 2 , where R is the
bond length and µ = m A m B /(m A +m B ). For a homonuclear diatomic m A = m B
so it follows that µ = m B /2. Using m = M/N A , this becomes µ = M B /2N A .
M B R 2 (0.01600 kg mol−1 ) × (120.75 × 10−12 m)2

I = µR 2 = =
2N A 2 × (6.0221 × 1023 mol−1 )
= 1.93... × 10−46 kg m2
kT kT 4πcI kT 4πcI kT
qR = =( )( )=( )( )= 2 I
2hc B̃ 2hc ħ 4πcħ ħ ħ
−23 −1
(1.3806 × 10 J K ) × (300 K)
= × (1.93... × 10−46 kg m2 ) = 72.1
(1.0546 × 10−34 J s)2
E12B.12(a) The rotational partition function of a non-linear rotor is given by [12B.14–473],

q R = (1/σ)(kT/hc)3/2 (π/ÃB̃C̃)1/2 , where σ is the symmetry number. NOF is
not centro-symmetric so σ = 1.
(i) At 25 ○ C which is 298.15 K
kT 3/2 π 1/2
qR = ( ) ( )
hc ÃB̃C̃
3/2
(1.3806 × 10−23 J K−1 ) × (298.15 K)
=( )
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
1/2
π
×( )
(3.1752 cm ) × (0.3951 cm−1 ) × (0.3505 cm−1 )
−1
= 7.97 × 103
(ii) At 100 ○ C which is 373.15 K

3/2
(1.3806 × 10−23 J K−1 ) × (373.15 K)
qR = ( )
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
1/2
π
×( )
(3.1752 cm ) × (0.3951 cm−1 ) × (0.3505 cm−1 )
−1
= 1.12 × 104
P12B.1 The partition function is given by [12B.1b–466], q = ∑ i g i e−βε i , where g i is de-
generacy and the corresponding energy is ε i . Therefore the partition function
is
q = 1 × e−β0 + 1 × e−βε + 1 × e−β(2ε) = 1 + e−ε/k T + e−2ε/k T
This is plotted in Fig 12.4. As expected, the partition function rises from a
value of 1 at low temperatures, where only the ground state is occupied, and
approaches a value of 3 at high temperatures, when all three states are nearly
equally populated.
P12B.3 As discussed in Section 11C.3(a) on page 420, the Morse oscillator has a finite
number of bound levels between the ground level and the dissociation limit.
The number of these is found by noting that E υ reaches a maximum value at
the dissociation limit, and therefore this limit is found by solving dE υ /dυ = 0
d
[(υ + 12 )hc ν̃ − (υ + 12 )2 hcx e ν̃] = hc ν̃ − 2(υ + 12 )hcx e ν̃
dυ
solving 0 = hc ν̃ − 2(υ max + 12 )hcx e ν̃ gives υ max = 1/2x e − 1
2
The energy of the lowest state is E 0 = 21 hc ν̃ − 14 hcx e ν̃, therefore the energies
used to evaluate the partition function are
E υ′ = E υ − E 0 = [(υ + 21 )hc ν̃ − (υ + 12 )2 hcx e ν̃] − [ 12 hc ν̃ − 41 hcx e ν̃]

= υhc ν̃ − (υ 2 + υ)hcx e ν̃
2
q
0 5 10 15 20 25 30
kT/ε
Figure 12.4
The partition function is evaluated from the sum
υ max
q VM = ∑ e−(υhc ν̃−(υ +υ)hc x e ν̃)/k T
2
υ=0
Defining the characteristic vibrational temperature as θ V = hc ν̃/k gives
υ max
q VM = ∑ e−(υ−(υ +υ)x e )θ V /T
2
υ=0
For the harmonic oscillator the partition is given by the exact expression [12B.15–
474], q VHO = (1 − e−θ /T )−1 .
V
Figure 12.5 compares the partition functions for various values of x e with that
for the harmonic case. For the smallest value of x e the partition function is
initially larger than that for the harmonic oscillator. This can be attributed
to more energy levels contributing at these temperatures as they are closer in
energy for the Morse oscillator than for the harmonic case. However, at higher
temperatures the partition function for the Morse oscillator starts to level off
because there are a finite number of levels, whereas for the harmonic case the
partition function continues without limit as there are an infinite number of
levels.
This behaviour is even more pronounced for x e = 0.05 and x e = 0.10. In these
two cases υ max is 10 and 5, respectively, and these values set the limiting high-
temperature value of the partition function.
P12B.5 The partition function is given by [12B.1b–466], q = ∑ i g i e−βε i , where g i is

degeneracy and the corresponding energy is given as ε i = hc ν̃ i , and β = 1/kT.
q E = g 0 + g 1 × e−hc ν̃ 1 /k T + g 2 × e−hc ν̃ 2 /k T + g 3 × e−hc ν̃ 3 /k T

harmonic
15
x e = 0.02
x e = 0.05
x e = 0.10
10
qV
0
0 2 4 6 8 10 12 14 16
T/θ V
Figure 12.5
(a) (i) At T = 298 K

(6.6261 × 10−34 J s)×(2.9979 × 1010 cm s−1 )
hcβ = = 4.82...×10−3 cm
(1.3806 × 10−23 J K−1 )×(298 K)
−3 −1
q E = 5 + 1 × e−(4.82 ...×10 cm)×(4707 cm )
−3 −1
+ 3 × e−(4.82 ...×10 cm)×(4751 cm )
−3
cm)×(10559 cm−1 )
+ 5 × e−(4.82 ...×10
= 5 + 1.34... × 10−10 + 3.27... × 10−10 + 3.61... × 10−22
= 5.00
(ii) At T = 5000 K
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
hcβ =
(1.3806 × 10−23 J K−1 ) × (5000 K)
= 2.87... × 10−4 cm
−4 −1
q E = 5 + 1 × e−(2.87 ...×10 cm)×(4707 cm )
−4 −1
+ 3 × e−(2.87 ...×10 cm)×(4751 cm )
−4
cm)×(10559 cm−1 )
+ 5 × e−(2.87 ...×10
= 5 + 0.258... + 0.764... + 0.239... = 6.262... = 6.262
(b) The population of level (term) i is N i /N = (g i × e−hc ν̃ i /k T )/q E

(i) T = 298 K
N 0 /N = 5/5.00... = 1.00
N 2 /N = (3.26... × 10−10 )/5.00... = 6.54 × 10−11
(ii) T = 5000 K
N0 5 N 2 0.764...
= = 0.798 = = 0.122
N 6.262... N 6.262...
P12B.7 The partition function is given by [12B.1b–466], q = ∑ i g i e−βε i , where g i is
degeneracy and the corresponding energy is given as ε i = hc ν̃ i , and β = 1/kT.
Here g i = 2J +1, where J is the right subscript in the term symbol. At T = 298 K
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
hcβ = = 4.82... × 10−3 cm
(1.3806 × 10−23 J K−1 ) × (298 K)
−3 −1
q E = 1 + 3 × e−(4.82 ...×10 cm)×(557.1 cm )
−3 −1 −3
cm)×(7125.3 cm−1 )
+ 5 × e−(4.82 ...×10 cm)×(1410.0 cm ) + 5 × e−(4.82 ...×10
−3
cm)×(16367.3 cm−1 )
+ 1 × e−(4.82 ...×10
= 1 + 0.203... + 5.52... × 10−3 + 5.72... × 10−15 + 4.78... × 10−35
= 1.209
Similarly, at T = 1000 K
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
hcβ = = 1.43... × 10−3 cm
(1.3806 × 10−23 J K−1 ) × (1000 K)
−3 −1
q E = 1 + 3 × e−(1.43 ...×10 cm)×(557.1 cm )
−3 −1 −3
cm)×(7125.3 cm−1 )
+ 5 × e−(1.43 ...×10 cm)×(1410.0 cm ) + 5 × e−(1.43 ...×10
−3
cm)×(16367.3 cm−1 )
+ 1 × e−(1.43 ...×10
= 1 + 1.34... + 0.657... + 1.76... × 10−4 + 5.92... × 10−11
= 3.003
P12B.9 The partition function is given by [12B.1b–466], q R = ∑ i g i e−βε i . Where g i is

degeneracy and ε i = hc F̃ i . The rotational terms of a symmetric rotor is given
by [11B.13a–409], F̃(J, K) = B̃J(J + 1) + (Ã − B̃)K 2 , with J = 0, 1, 2, ... and
K = 0, ±1, ..., ±J. Thus, the partition function is
∞ ⎡ +J ⎤
q R = ∑(2J + 1)e−hc β B̃ J(J+1) ⎢ ∑ e−hc β( Ã−B̃)K ⎥
⎢ 2⎥
J=0
⎢K=−J ⎥
⎣ ⎦
Using mathematical software the terms are evaluated and summed until con-
vergence is achieved to within the required precision.
In the high-temperature limit the partition function is given by [12B.12b–472],
q R = (kT/hc)3/2 (π/ÃB̃C̃)1/2 ; for a symmetric rotor B̃ = C̃, therefore q R =
(kT/hc)3/2 (π/ÃB̃ 2 )1/2 . With the given data
q R = (kT/hc)3/2 (π/ÃB̃ 2 )1/2 = (1.02... × 10 K) × T 3/2
The two forms of the partition function are plotted in Fig. 12.6; the high tem-
perature limit is accurate to within 5 % of the exact solution at 4.5 K .
30
20
qR 10
exact
high T limit
0
0 2 4 6 8 10
T/K
Figure 12.6
12C Molecular energies

D12C.1 This is described in Brief illustration 12C.1 on page 478.
E12C.1(a) The mean vibrational energy is given by [12C.8–479], ⟨ε V ⟩ = hc ν̃/(e β hc ν̃ − 1);
this result is exact. The equipartition value is ⟨ε V ⟩ = kT, because there are two
quadratic terms for a harmonic oscillator. These two expressions for the energy
are plotted as a function of T in Fig. 12.7. The value from the equipartition
theorem comes within 5 % of the exact value at 4.80 × 103 K .
6
⟨ε V ⟩ × 1020 /J
2
exact
equipartition
0
0 1 000 2 000 3 000 4 000 5 000
T/K
Figure 12.7
E12C.2(a) The mean vibrational energy per vibrational mode is given by [12C.8–479],
⟨ε Vi ⟩ = hc ν̃ i /(e β hc ν̃ i − 1); this result is exact. The overall vibrational energy

is the sum of the contributions from each normal mode, taking into account
the degeneracy of each
⟨ε V ⟩ = ⟨ε V1 ⟩ + 2 × ⟨ε V2 ⟩ + ⟨ε V3 ⟩
The equipartition value is ⟨ε V ⟩ = 4kT, because there are two quadratic terms
for a harmonic oscillator, and four modes in total. These two expressions for the
energy are plotted as a function of T in Fig. 12.8. The value from the equiparti-
tion theorem comes within 5% of the exact value at 1.10 × 104 K .
60
⟨ε V ⟩ × 1020 /J
40
20
exact
equipartition
0
0 2 000 4 000 6 000 8 000 10 000 12 000
T/K
Figure 12.8
E12C.3(a) The mean vibrational energy per vibrational mode is given by [12C.8–479],
⟨ε Vi ⟩ = hc ν̃ i /(e β hc ν̃ i − 1); this result is exact. The overall vibrational energy
is the sum of the contributions from each normal mode, taking into account
the degeneracy of each
⟨ε V ⟩ = ⟨ε V1 ⟩ + 2 × ⟨ε V2 ⟩ + 3 × ⟨ε V3 ⟩ + 3 × ⟨ε V4 ⟩
The equipartition value is ⟨ε V ⟩ = 9kT, because there are two quadratic terms
for a harmonic oscillator, and nine modes in total. These two expressions for
the energy are plotted as a function of T in Fig. 12.9. The value from the equipar-
tition theorem comes within 5% of the exact value at 6.85 × 103 K .
E12C.4(a) The mean molecular energy is given by [12C.4a–477], ⟨ε⟩ = −(1/q)(∂q/∂β)V ,

where β = 1/kT, and q is the partition function given by [12B.1b–466], q =
∑ i g i e−βε i , where g i is degeneracy and the corresponding energy is given as
ε i = hc ν̃ i . At T = 1900 K
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
βhc = = 7.57... × 10−4 cm
(1.3806 × 10−23 J K−1 ) × (1900 K)
q E = g 0 + g 1 e−β hc ν̃ 1 + g 2 e−β hc ν̃ 2 = 4.29...

100
⟨ε V ⟩ × 1020 /J
50
exact
equipartition
0
0 2 000 4 000 6 000 8 000
T/K
Figure 12.9
Therefore the mean energy is
1 ∂q E hc
⟨ε E ⟩ = − ( ) = E (g 1 ν̃ 1 e−β hc ν̃ 1 + g 2 ν̃ 2 e−β hc ν̃ 2 )
q E ∂β V q
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
=
4.29...
−4 −1
× [1 × (2500 cm ) × e−(7.57 ...×10 cm)×(2500 cm )
−1
−4
cm)×(3500 cm−1 )
+2 × (3500 cm−1 ) × e−(7.57 ...×10 ]
= 4.03 × 10−21 J
E12C.5(a) The mean energy of a molecule is given by [12C.2–477], ⟨ε⟩ = (1/q) ∑ i ε i e−βε i ,
where ε i = hc ν̃ i , β = 1/kT, and q is the partition function given by [12A.11–
463], q = ∑ i e−βε i . Therefore for the two-level system
0 + εe−βε ε hc ν̃
⟨ε⟩ = −βε
= βε = hc ν̃/k T
1+e e +1 e +1
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (500 cm−1 )
= (6.6261×10−34 J s)×(2.9979×10 10 cm s−1 )×(500 cm−1 )
e (1.3806×10−23 J K−1 )×(298 K) +1
= 8.15 × 10−22 J

where β = 1/kT and q is the partition function. The rotational partition func-
tion of a heteronuclear diatomic is given in terms of the rotational constant B̃
by [12B.11–470], q R = ∑ J (2J + 1)e−β hc B̃ J(J+1) .
1 ∂q R 1
⟨ε R ⟩ = − ( ) = R ∑ hc B̃J(J + 1)(2J + 1)e−β hc B̃ J(J+1)
q R ∂β V q J
The terms of the sum above, and also of the sum needed to compute q R , are
evaluated and summed until the result has converged to the required precision.
The equipartition value is ⟨ε R ⟩ = kT. These two expressions for the energy
are plotted as a function of T in Fig. 12.10. The value from the equipartition
theorem comes within 5 % of the exact value at 19.6 K .
2
⟨ε R ⟩ × 1022 /J
exact
equipartition
0
0 5 10 15 20
T/K
Figure 12.10

where β = 1/kT, and q is the partition function given by [12B.1b–466], q R =
∑ J g J e−βε J . The energy levels of a spherical rotor are given in [11B.8–408], ε J =
hc B̃J(J + 1) and, as is explained in Section 11B.1(c) on page 410, each has a
degeneracy g J = (2J + 1)2 . It follows that
q R = ∑(2J + 1)2 e−β hc B̃ J(J+1)

J
1 ∂q R 1 2 −β hc B̃ J(J+1)
⟨ε R ⟩ = − ( ) = ∑ hc B̃J(J + 1)(2J + 1) e
q R ∂β V q R J
The terms in the sum needed to compute q R and ⟨ε R ⟩ are evaluated and summed
until the result has converged to the required precision. The equipartition value
is ⟨ε R ⟩ = 32 kT, because for this non-linear molecule there are three rotational
degrees of freedom. These two expressions for the energy are plotted as a func-
tion of T in Fig. 12.11. The value from the equipartition theorem comes within
5 % of the exact value at 26.4 K .
P12C.1 The mean molecular energy is given by [12C.4a–477], ⟨ε⟩ = −(1/q)(∂q/∂β)V ,
where β = 1/kT, and q is the partition function given by [12B.1b–466], q =
∑ i g i e−βε i . For a symmetric rotor the rotational terms are given in [11B.13a–
409], F̃(J, K) = B̃J(J + 1) + (Ã − B̃)K 2 , with J = 0, 1, 2, ... and K = 0, ±1, ..., ±J;
the corresponding energies are hc F̃(J, K) and the degeneracy is (2J + 1). The
⟨ε R ⟩ × 1022 /J
4
2
exact
equipartition
0
0 5 10 15 20 25 30
T/K
Figure 12.11
rotational partition function is therefore given by

∞ ⎡ +J ⎤
q R = ∑(2J + 1)e−β hc B̃ J(J+1) ⎢ ∑ e−β hc( Ã−B̃)K ⎥
⎢ 2⎥
J=0
⎢K=−J ⎥
⎣ ⎦
The mean energy is therefore
1 ∂q R
⟨ε R ⟩ = − ( )
q R ∂β V
1 ⎛∞ ⎡ +J ⎤
−β hc B̃ J(J+1) ⎢ −β hc( Ã− B̃)K 2 ⎥
= ∑ hc B̃J(J + 1)(2J + 1)e ⎢ ∑ e ⎥
q R ⎝ J=0 ⎢K=−J
⎣
⎥
⎦
∞ ⎡ +J ⎤
2 ⎥⎞
+ ∑(2J + 1)e−β hc B̃ J(J+1) ⎢ ∑ hc(Ã − B̃)K 2 e−β hc( Ã−B̃)K ⎥
⎢
J=0
⎢K=−J ⎥⎠
⎣ ⎦
1 ∞ −β hc B̃ J(J+1)
= ∑(2J + 1)e
q R J=0
⎡ +J ⎤
× ⎢hc ∑ (B̃J(J + 1) + (Ã − B̃)K 2 ) e−β hc( Ã−B̃)K ⎥
⎢ 2⎥
⎢ K=−J ⎥
⎣ ⎦
The terms in the sum in this expression, and the terms in q R , are evaluated and
summed until the value converges to the required precision. The equipartition
value of the energy is ⟨ε R ⟩ = 32 kT because there are three rotational degrees
of freedom. The two expressions for the energy are plotted as a function of
temperature in Fig. 12.12. The equipartition value is within 5% of the exact
solution at 4.59 K .
P12C.3 H2 O has three translational and three rotational degrees of freedom. It there-
fore follows from the equipartition principle that the molar internal energy is
U m = (3 + 3) × 12 RT = 3RT. The constant-volume molar heat capacity is
therefore C V ,m = (∂U m /∂T)V = 3R.
1.0
⟨ε R ⟩ × 1022 /J
0.5
exact
equipartition
0.0
0 1 2 3 4 5 6
T/K
Figure 12.12
The energy needed to raise the temperature of n mol of H2 O by ∆T is equal

to the change in the internal energy which is ∆U = nC V ,m ∆T. In this case
∆U = (1.0 mol) × (3 × 8.3145 J K−1 mol−1 ) × (100 K) = 2.5 kJ .
P12C.5 The energy levels for a spin in a magnetic field are given by [16A.4d–633], E m I =
−g I µ N B0 m I , where m I = +1, 0, −1. These energies are conveniently written as
E m I = −m I δ, with δ = g I µ N B0 . If the energy of the lowest state is defined as the
energy zero, then the three levels have energies E 0 = 0, E 1 = δ, and E 2 = 2δ.
The partition function is
q = 1 + e−βδ + e−2βδ
The mean molecular energy is given by [12C.4a–477]
1 ∂q
⟨ε⟩ = ε gs − ( )
q ∂β V
1
= ε gs − (−δe−βδ − 2δe−2βδ )
1 + e−βδ + e−2βδ
δe−βδ + 2δe−2βδ δe−βδ + 2δe−2βδ
= ε gs + = −δ +
1 + e−βδ + e−2βδ 1 + e−βδ + e−2βδ
In the last step the fact that the energy of ground state (the one with m I = +1)
is −δ is used. With the data given
δ = g I µ N B0 = 2.0 × (5.0508 × 10−27 J T−1 ) × (2.5 T) = 2.52... × 10−26 J
and assuming that T = 298 K
βδ = δ/kT = (2.52... × 10−26 J)/[(1.3806 × 10−23 J K−1 ) × (298 K)]

= 6.13... × 10−6
⟨ε⟩ = (−2.52... × 10−26 J)

−6 −6
(2.52... × 10−26 J)e−6.13 ...×10 + 2 × (2.52... × 10−26 J)e−2×(6.13 ...×10 )
+
1 + e−6.13 ...×10−6 + e−2×6.13 ...×10−6
−31
= −1.03 × 10 J
The separation of the energy levels is very much smaller than kT, therefore the
three levels have almost equal populations giving a mean energy of very close
to zero.
P12C.7 The partition function given by [12B.1b–466], q = ∑ i g i e−βε i , where g i is de-

generacy, ε i = hc ν̃ and β = 1/kT. Therefore
q E = 2e0 + 2e−β hc ν̃ = 2 + 2e−hc ν̃/k T
This function in plotted in Fig. 12.13.
4.0
3.5
qE
3.0
2.5
2.0
0 200 400 600 800 1 000
T/K
Figure 12.13
(a) The ratio of the populations is given by [12A.13b–464]
N i /N j = (g i /g j )e−β(ε i −ε j )
At 300 K this ratio is
N 1 /N 0 = (g 1 /g 0 ) × e−β hc ν̃ = (2/2) × e−β hc ν̃

(6.6261×10−34 J s)×(2.9979×10 10 cm s−1 )×(121.1 cm−1 )
− (1.3806×10−23 J K−1 )×(300 K)
=e = 0.559....
Thus the populations expressed as a fraction of N are
N 0 /N = N 0 /(N 0 + N 1 ) = 1/(1 + 0.559...) = 0.641

N 1 /N = N 1 /(N 0 + N 1 ) = 0.559.../(1 + 0.559...) = 0.359
(b) The mean molecular energy is given by [12C.4a–477], ⟨ε⟩ = −(1/q)(∂q/∂β)V .

Thus
1 dq E e−hc ν̃/k T hc ν̃
⟨ε E ⟩ = − = hc ν̃ × =
E
q dβ 1 + e−hc ν̃/k T e hc ν̃/k T + 1
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (121.1 cm−1 )
= (6.6261×10−34 J s)×(2.9979×10 10 cm s−1 )×(121.1 cm−1 )
e (1.3806×10−23 J K−1 )×(300 K) +1
= 8.63 × 10−22 J
P12C.9 Mean values of any observable are given by a sum of the observed value over
all the possible states weighted by the probability of each state. Thus, the mean
of the square of energy is given by ⟨ε 2 ⟩ = (1/q) ∑ j ε j 2 e−βε j , where q is the
partition function given by [12A.11–463], q = ∑ i e−βε i .
It is useful to consider the first and second derivatives of the term e−βε j with
respect to β. The first derivative is
d −βε j
e = −ε j e−βε j
dβ
and hence the second is
d2 −βε j d d −βε j
2e = (−ε j e−βε j ) = (−ε j ) × e
dβ dβ dβ
= (−ε j ) × (−ε j ) × e−βε j = ε j 2 × e−βε j
This latter expression is used to rewrite the definition of ⟨ε 2 ⟩ as
1 2 −βε
⟨ε 2 ⟩ = ∑ε e j
q j j
⎡ ⎤ 2
1 d2 −βε 1 d2 ⎢∑ e−βε j ⎥ = 1 d q
⎢ ⎥
= ∑ 2e j = ⎢ ⎥ q 2
q j dβ q dβ 2 ⎢ j ⎥ dβ
⎣ ⎦
Therefore
1/2
2 1/2 1 d2 q
⟨ε ⟩ = ( )
q dβ 2
These results are also used to find the root mean square of the deviation from
the mean as
2 1/2 2 1/2
2 2 1/2 ⎛ 1 d2 q 1 dq ⎞ 1 ⎛ d2 q dq ⎞
(⟨ε ⟩ − ⟨ε⟩ ) = 2 −( ) = q 2 −( )
⎝ q dβ q dβ ⎠ q ⎝ dβ dβ ⎠
For a harmonic oscillator the partition function is given by [12B.15–474], q =

(1 − e−β hc ν̃ )−1 . Therefore
dq −hc ν̃ e−β hc ν̃
= = −hc ν̃ e−β hc ν̃ q 2
dβ (1 − e−β hc ν̃ )2
d2 q −hc ν̃ e−β hc ν̃ −2hc ν̃ e−β hc ν̃ e−β hc ν̃
= −hc ν̃ [ + ]
dβ 2 (1 − e−β hc ν̃ )2 (1 − e−β hc ν̃ )3
1 2e−β hc ν̃
= (hc ν̃)2 e−β hc ν̃ [ + ]
(1 − e−β hc ν̃ )2 (1 − e−β hc ν̃ )3
= (hc ν̃)2 e−β hc ν̃ [q 2 + 2e−β hc ν̃ q 3 ]
= q 2 (hc ν̃)2 e−β hc ν̃ [1 + 2e−β hc ν̃ q]
Using these results and making the substitution x = hc ν̃

2
2 1 d2 q 1 dq
⟨ε 2 ⟩ − ⟨ε⟩ = 2 −( )
q dβ q dβ
2
= q x 2 e−βx [1 + 2e−βx q] − [−x e−βx q]
x 2 e−βx x 2 e−2βx
= q x 2 e−βx + q 2 x 2 e−2βx = −βx
+
1−e (1 − e−βx )2
x 2 e−βx (1 − e−βx ) + x 2 e−2βx x 2 e−βx
= =
(1 − e−βx )2 (1 − e−βx )2
hence
2 1/2 hc ν̃ e−β hc ν̃/2

(⟨ε 2 ⟩ − ⟨ε⟩ ) =
1 − e−β hc ν̃
In terms of the vibrational temperature this is
kθ R e−θ /2T
R
1 − e−θ R /T
At high temperatures, T ≫ θ R , the exponential in the denominator is approx-

imated as 1 − θ R /T, and the exponential in the numerator goes to 1, giving
(⟨ε 2 ⟩ − ⟨ε⟩2 )1/2 = kT. At low temperatures, T ≪ θ R , (⟨ε 2 ⟩ − ⟨ε⟩2 )1/2 tends to
0. This is as expected because if all the particles are in the ground state there is
no uncertainty in their average energy.
12D The canonical ensemble

D12D.1 An ensemble is a set of a large number of imaginary replications of the actual
system. These replications are identical in some, but not all, respects. For
example, in the canonical ensemble, all replications have the same number of
particles, the same volume, and the same temperature, but they need not have
the same energy.
Ensembles are useful in statistical thermodynamics because it is mathemati-

cally more tractable to perform an ensemble average to determine the (time
averaged) thermodynamic properties than it is to perform an average over time
to determine these properties. Recall that macroscopic thermodynamic prop-
erties are averages over the time dependent properties of the particles that com-
pose the macroscopic system. In fact, it is taken as a fundamental principle
of statistical thermodynamics that the (sufficiently long) time average of every
physical observable is equal to its ensemble average. This principle is connected
to a famous assumption of Boltzmann’s called the ergodic hypothesis.
D12D.3 In the context of ensembles, the thermodynamic limit is achieved as the num-
ber of replications Ñ approaches infinity. In this limit, the dominating config-
uration is overwhelmingly the most probable configuration, and its properties
are essentially the same as those of the system.
E12D.1(a) It is essential to include the factor 1/N! when considering indistinguishable
particles which are free to move. Thus, such a factor is always needed for gases.
In the solid state, particles are distinguished by their positions in the lattice
and therefore the particles are regarded as distinguishable on the basis that
their locations are distinguishable. For the cases mentioned, the factor 1/N!
is needed for all but solid CO.
P12D.1
∂ ln Q ∂ ln(q N /N!) ∂(N ln q − ln N!)

p = kT ( ) = kT ( ) = kT ( )
∂V T ∂V T
∂V T
The ln N! term is volume independent and thus p = N kT (∂ ln q/∂V )T . The

molecular partition function, q, for the perfect gas is just the translational par-
tition function given by [12B.10b–469], q T = V /Λ 3 , where Λ is the thermal
wavelength which is independent of volume. Therefore
∂ ln(V /Λ 3 ) ∂(ln V − ln Λ 3 )
p = N kT ( ) = N kT ( )
∂V T
∂V T
∂ ln V N kT nRT
= N kT ( ) = =
∂V T V V
where for the last step N = nN A and R = kN A are used.

12E The internal energy and entropy

D12E.1 The statistical entropy is defined by Boltzmann’s formula, S = k ln W, in terms
of the number of configurations or microstates consistent with a given total
energy. The thermodynamic entropy is defined by dS = dq rev /T that is, in
terms of reversible heat transfer.
The concept of the number of microstates makes quantitative the qualitative
concepts of ‘disorder’ and ‘dispersal of matter and energy’ that are often used
to introduce the concept of entropy: a more ‘disorderly’ distribution of energy
and matter corresponds to a greater number of microstates consistent with the
same total energy. The more molecules that can participate in the distribution
of the energy, the more microstates there are for a given total energy and hence
the greater the entropy.
The molecular interpretation of entropy embodied in the Boltzmann formula
also suggests the thermodynamic definition. At high temperatures, where the
molecules of a system can occupy a large number of energy levels, a small
additional transfer of energy as heat will cause only a small change in the num-
ber of accessible energy levels, whereas at low temperatures the transfer of the
same quantity of heat will increase the number of accessible energy levels and
microstates significantly. Hence the change in entropy upon heating will be
greater when the energy is transferred to a cold body than when it is transferred
to a hot body, as required by the thermodynamic definition.
D12E.3 The entropy of a monatomic perfect gas is given by the Sackur–Tetrode equa-
tion [12E.9a–491]
Vm e5/2
S m = R ln ( ) Λ = h/(2πmkT)1/2 Vm = RT/p
NA Λ3
Because the molar volume appears in the numerator, the molar entropy in-
creases with the molar volume. In terms of the Boltzmann distribution, this
relationship is expected: large containers have more closely spaced energy lev-
els than do small ones, so more states are thermally accessible. Temperature
appears in the numerator of the expression (through the denominator of Λ),
so the molar entropy increases with the temperature. Again, this is consistent
with the Boltzmann distribution, because more states are accessible at higher
temperatures than at lower ones.
The fact that diatomic and polyatomic gases have rotational and vibrational
modes of motion as well does not change the above arguments. The parti-
tion functions of those modes are independent of volume, so the volume de-
pendence of the entropy is as described above. At most temperatures, rota-
tional modes of motion are active and contribute to the entropy, as expressed
in [12E.11a–492]; the contribution increases with temperature. Finally, most
vibrational modes contribute little if at all to the entropy, but as with rotation
the contribution increases with temperature.
D12E.5 The temperature is always high enough for the mean translational energy to
be 32 kT, the equipartition value. Therefore, the molar constant-volume heat
capacity for translation is C VT ,m = 32 R.
When the temperature is high enough for the rotations of the molecules to
be highly excited (when T ≫ θ R ) the equipartition value kT for the mean
rotational energy (for a linear rotor) can be used to obtain C VR ,m = R. For non-
linear molecules, the mean rotational energy is 23 kT, so the molar rotational
heat capacity rises to 32 R when T ≫ θ R . At intermediate temperatures the total
heat capacity takes a value between that due to translation, 32 R, and 52 R (for a
linear molecule) when both translation and rotation contribute fully.
Molecular vibrations contribute to the heat capacity, but only when the temper-
ature is high enough for them to be significantly excited. For each vibrational
mode, the equipartition mean energy is kT, so the maximum contribution to
the molar heat capacity is R. However, it is unusual for the vibrations to be so
highly excited that equipartition is valid, and in general the contribution to the
heat capacity has to be calculated using [12E.3–488].
E12E.1(a) For atoms with filled shells the only contribution to the entropy is translational.
The standard molar entropy of a monatomic perfect gas is given by the Sackur–
Tetrode equation [12E.9b–491]
−
○ kTe5/2
Sm = R ln ( ) Λ = h/(2πmkT)1/2
p−○ Λ 3
(i) Taking the mass of a He atom as 4.00 m u
6.6261 × 10−34 J s
Λ= 1/2
[2π(4.00×1.6605 × 10−27 kg)×(1.3806 × 10−23 J K−1 )×(298 K)]
= 5.05... × 10−11 m
(1.3806 × 10−23 J K−1 ) × (298 K) × e5/2
−
○
Sm = (8.3145 J K−1 mol−1 ) × ln [ ]
(105 N m−2 ) × (5.05 . . . × 10−11 m)3
= 126 J K−1 mol−1
(ii) Taking the mass of a Xe atom as 131.29 m u
6.6261 × 10−34 J s
Λ= 1/2
[2π(131.29×1.6605 × 10−27 kg)×(1.3806 × 10−23 J K−1 )×(298 K)]
= 8.82... × 10−12 m
(1.3806 × 10−23 J K−1 ) × (298 K) × e5/2
−
○
Sm = (8.3145 J K−1 mol−1 ) × ln [ ]
(105 N m−2 ) × (8.82... × 10−12 m)3
= 169.7 J K−1 mol−1
E12E.2(a) For atoms with filled shells the only contribution to the entropy is translational.
The standard molar entropy of a monatomic perfect gas is given by the Sackur–
Tetrode equation [12E.9b–491]
−
○ kTe5/2
Sm = R ln ( ) Λ = h/(2πmkT)1/2
p−○ Λ 3
−
○
It follows that S m = A ln(T 5/2 m 3/2 ), where A is a constant. Therefore
−
○ −
○ 5/2 3/2 5/2 3/2
Sm (He, T1 ) − S m (Xe, T2 ) = A ln(T1 m He ) − A ln(T2 m Xe )
−
○ −
○
If S m (He, T1 ) = S m (Xe, 298)
5/2 3/2 3/2

0 = A ln[T1 m He ] − A ln[(298 K)5/2 m Xe ]
5/2 3/2 3/2
hence T1 m He = (298 K)5/2 m Xe
2/5
(298 K)5/2 × (131.29)3/2
T1 = [ ] = 2.42 × 103 K
(4.00)3/2
E12E.3(a) The rotational partition function for a non-linear molecule is given by [12B.14–
473]
1 kT 3/2 π 1/2
qR = ( ) ( )
σ hc ÃB̃C̃
For H2 O the symmetry factor σ = 2. At 298 K
(1.3806 × 10−23 J K−1 ) × (298 K)

kT/hc =
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
= 207.1... cm−1
q R = 21 (207.1... cm−1 )3/2 ×
1/2
π
( )
(27.878 cm−1 ) × (14.509 cm−1 ) × (9.827 cm−1 )
= 43.1... = 43.1
The entropy is given in terms of the partition function by [12E.8a–490]
S m = [U m (T) − U m (0)]/T + R ln q
This is the appropriate form for the rotational contribution; for the translational
contribution the ln term is ln qe/N. At 298 K kT/hc = 207 cm−1 which is signif-
icantly greater than any of the rotational constants, therefore the equipartition
theorem can be used to find U m (T): there are three rotational modes, therefore
U m (T) − U m (0) = 32 RT.
Sm
R
= ( 23 RT)/T + R ln q R = R( 23 + ln q R )
= (8.3145 J K−1 mol−1 ) × [ 23 + ln(43.1...)]
= 43.76 J K−1 mol−1
E12E.4(a) Only the ground electronic state contributes to the electronic partition func-
tion, which is therefore simply the degeneracy of the ground state q E = g 0 .
For a given term the degeneracy is given by the value of J, which is the right
subscript: g 0 = (2J + 1) = (2 × 29 + 1) = 10. The entropy is given in terms of the
partition function by [12E.8a–490]
S m = [U m (T) − U m (0)]/T + R ln q
This is the appropriate form for the electronic contribution; for the translational
contribution the ln term is ln qe/N. In this case U m (T) − U m (0) = 0 as only
the ground state is considered
S m = R ln q = R ln 10 = 19.14 J K−1 mol−1
E12E.5(a) The contribution of a collection of harmonic oscillators to the standard molar

entropy is given by [12E.12b–492] (note that there is an error in the expression
in the text: the argument of the exponential term in the ln should be negative)
θ V /T
− ln(1 − e−θ /T )]
V
Sm
V
= R[ θ V = hc ν̃/k
eθ V /T − 1
The following table shows the vibrational temperatures and the contribution to
the molar entropy for each of the normal modes
298 K 500 K
−1
ν̃/cm θ /K
V
θ /T
V
Sm
V
/R θ /TV
Sm
V
/R
625 899 3.02 0.205 1.80 0.538
638 918 3.08 0.195 1.84 0.522
1033 1486 4.988 0.04110 2.973 0.2128
1105 1590 5.335 0.03066 3.180 0.1805
1229 1768 5.934 0.01841 3.537 0.1356
1387 1996 6.697 9.515 × 10−3 3.991 0.09378
1770 2547 8.546 1.855 × 10−3 5.093 0.03761
2943 4234 14.21 1.026 × 10−5 8.469 1.988 × 10−3
3570 5137 17.24 5.957 × 10−7 10.27 3.895 × 10−4
The molar entropy is obtained by summing the contributions from each normal
mode. Thus at 298 K S m
V
= 4.18 J K−1 mol−1 and at 500 K S m
V
= 14.3 J K−1 mol−1 .
E12E.6(a) The equipartition value for C V ,m is expressed in [12E.6–488]: each transla-

tional or rotational mode contributes 12 R, and each active vibrational mode
contributes R.
(i) I2 : three translational modes, two rotational modes (linear) and, because
the vibrational frequency of the molecule is rather low, one vibrational
mode: C V ,m /R = 3 × 12 + 2 × 21 + 1 = 72 .
(ii) CH4 : three translational modes, three rotational modes (non-linear), and
no active vibrational modes: C V ,m /R = 3 × 21 + 3 × 12 = 3 .
(iii) C6 H6 : three translational modes, three rotational modes (non-linear),
and no active vibrational modes: C V ,m /R = 3 × 12 + 3 × 12 = 3 . There
are four low-frequency normal modes which, if active, will contribute a
further 4R.
E12E.7(a) The equipartition value for C V ,m is expressed in [12E.6–488]: each transla-

tional or rotational mode contributes 12 R, and each active vibrational mode
contributes R. The number of vibrational modes is (3N −6), which is 6 for NH3
and 9 for CH4 . For NH3 there are three translational modes, three rotational
modes (non-linear) giving C V ,m = 3R; if the 6 vibrations are included, C V ,m =
9R. For CH4 there are three translational modes, three rotational modes (non-
linear) giving C V ,m = 3R; if the 9 vibrations are included, C V ,m = 12R.
γ = C p,m /C V ,m = (C V ,m + R)/C V ,m = 1 + R/C V ,m

γ NH3 = 1 + R/3R = 1.33 no vibrational contribution
γ NH3 = 1 + R/9R = 1.11 with vibrational contribution
γ CH4 = 1 + R/3R = 1.33 no vibrational contribution
γ CH3 = 1 + R/12R = 1.08 with vibrational contribution
The experimental value for γ is 1.31 for both gases: evidently the vibrational
modes are not active.
E12E.8(a) The partition function of this two-level system is
q = g 0 + g 1 e−β hc ν̃
where g 0 and g 1 are the degeneracies of the ground and first excited state,
respectively. The mean energy is given by [12C.4a–477], ⟨ε⟩ = −(1/q)(∂q/∂β)V
g 1 hc ν̃ e−β hc ν̃ g 1 hc ν̃
⟨ε⟩ = =
g 0 + g 1 e−β hc ν̃ g 0 e β hc ν̃ + g 1
N A g 1 hc ν̃
hence U m = N A ⟨ε⟩ =
g 0 e β hc ν̃ + g 1
By definition C V ,m = (∂U m /∂T)V , therefore
∂U m ∂U m dβ ∂U m −1
C V ,m = ( ) =( ) =( ) ×
∂T V ∂β V dT ∂β V kT 2
1 g 0 hc ν̃e β hc ν̃
= × N A g 1 hc ν̃
kT 2 (g 0 e β hc ν̃ + g 1 )2
N A (hc ν̃)2 g 0 g 1 e β hc ν̃
=
kT 2 (g 0 e β hc ν̃ + g 1 )2
For an electronic term the degeneracy is 2J + 1, hence g 0 = 2 × 3

2
+ 1 = 4 and
g 1 = 2 × 12 + 1 = 2. With the data given
hc ν̃ = (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (881 cm−1 )

= 1.75... × 10−20 J
at 500 K βhc ν̃ = hc ν̃/kT
= (1.75... × 10−20 J)/[(1.3806 × 10−23 J K−1 ) × (500 K)] = 2.53...
at 900 K βhc ν̃ = hc ν̃/kT
= (1.75... × 10−20 J)/[(1.3806 × 10−23 J K−1 ) × (900 K)] = 1.40...
N A (hc ν̃)2 8 e β hc ν̃
C V ,m (500) =
kT 2 (4 e β hc ν̃ + 2)2
(6.0221 × 1023 mol−1 )×(1.75... × 10−20 J)2 8 e2.53 ...
= −1
−23
(1.3806 × 10 J K ) × (500 K) 2 (4 e2.53 ... + 2)2
= 1.96 J K−1 mol−1
(6.0221 × 1023 mol−1 )×(1.75... × 10−20 J)2 8 e1.40 ...
C V ,m (900) = −1
−23
(1.3806 × 10 J K ) × (900 K) 2 (4 e1.40 ... + 2)2
= 1.60 J K−1 mol−1
E12E.9(a) The contribution of a collection of harmonic oscillators to the molar heat ca-
pacity is given by [12E.3–488]
2 2
θ V ⎛ e−θ /2T ⎞
V
C V ,m = R( ) θ V = hc ν̃/k
T ⎝ 1 − e−θ /T ⎠
V
This function is plotted in Fig 12.14.

The following table shows the vibrational temperatures and the contribution to
the heat capacity for each of the normal modes
298 K 500 K
ν̃/cm−1 θ V /K T/θ V C V ,m /R T/θ V C V ,m /R
612 881 0.338 0.506 0.568 0.777
729 1049 0.284 0.390 0.477 0.702
1974 2840 0.1049 6.593 × 10−3 0.1760 0.1109
3287 4729 0.06301 3.226 × 10−5 0.1057 6.980 × 10−3
3374 4855 0.06139 2.233 × 10−5 0.1030 5.725 × 10−3
The heat capacity is obtained by summing the contributions from each normal
mode, taking into account the double degeneracy of the modes at 612 cm−1 and
729 cm−1 by counting each twice. Thus at 298 K C V ,m = 14.95 J K−1 mol−1 and
at 500 K C V ,m = 25.62 J K−1 mol−1 .
1.0
0.8
0.6
C V ,m /R
0.4
0.2
0.0
0.0 0.2 0.4 0.6 0.8 1.0
T/θ V
Figure 12.14
P12E.1 The electronic levels of NO form a two-level system, an expression for the heat
capacity of which is derived in the solution to Exercise E13E.8(a).
N A (hc ν̃)2 g 0 g 1 e β hc ν̃ g 0 g 1 e β hc ν̃
C V ,m = β hc ν̃
= N A k(βhc ν̃)2
kT 2 (g 0 e + g1 ) 2 (g 0 e β hc ν̃ + g 1 )2
respectively. For NO g 0 = 2 and g 1 = 2. With the data given
βhc ν̃ = hc ν̃/k × T −1
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (121.1 cm−1 )
= × T −1
1.3806 × 10−23 J K−1
= (174.2... K) × T −1
A plot of C V ,m as a function of T is shown in Fig. 12.15.
P12E.3 The energy levels for a particle on a ring are given by [7F.4–269], E m = m 2 ħ 2 /2I
where I is the moment of inertia and m = 0, ±1, ±2, . . .. In the high-temperature
limit the partition function is well-approximated by an intergral
+∞
q R = ∑ e−βE m = ∫ e−βm ħ /2I
2 2
dm
m −∞
The integral is of the form of Integral G.1 with k ′ = βħ 2 /2I; the value needed is
twice that given for G.1 as that integral is from 0 to +∞.
1/2
R 2πI
q = ( 2)
βħ
These energy levels contribute one quadratic contribution to the energy, that is
there is one rotational mode. In the high-temperature limit the equipartition
theorem applies and hence the internal energy is U m = 12 RT and C VR ,m = 21 R .
0.4
C V ,m /R
0.2
0.0
0 100 200 300 400 500
T/K
Figure 12.15

S m = [U m (T) − U m (0)]/T + R ln q
This is the appropriate form for the rotational contribution; for the translational
contribution the ln term is ln qe/N. As has already been explained, U m (T) −
U m (0) = 12 RT, therefore
Sm
R
= ( 12 RT)/T + R ln q R = R( 21 + ln q R )
With the data given
1/2
2πI 2πkTI 1/2
qR = ( ) =( )
βħ 2 ħ2
1/2
2π(1.3806 × 10−23 J K−1 ) × (298 K) × (5.341 × 10−47 kg m2 )
=( )
(1.0546 × 10−34 J s)2
= 11.1...
Sm
R
= (8.3145 J K−1 mol−1 ) × [ 21 + ln(11.1...)] = 24.1 J K−1 mol−1
This calculation is for a particle on a ring. When used as a model for a rotating
CH3 group a symmetry factor of 3 is needed, so that q R is one third of the value
calculated here, giving S mR
= 15.1 J K−1 mol−1 .
P12E.5 The characteristic vibrational temperature is defined as θ V = hc ν̃/k. It follows
that ν̃ = kθ V /hc, so the vibrational frequency for a characteristic temperature
of 1000 K is
(1.3806 × 10−23 J K−1 ) × (1000 K)
ν̃ = = 695 cm−1
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
The following modes have vibrational frequencies of 695 cm−1 or less (the de-
generacies are given in parentheses)
525(3) 578(3) 354(3) 345(4) 403(5) 525(5) 667(5)
Thus in total there are 28 modes with characteristic vibrational temperatures

of less than 1000 K.
The contributions to C V ,m are three translational, three rotational and 28 vi-
brational modes, giving a heat capacity of C V ,m = 32 R + 32 R + 28R = 31R
P12E.7 The partition function of this three-level system is
q = 1 + e−βε + e−2βε
The mean energy is given by [12C.4a–477], E mean = −(1/q)(∂q/∂β)V
ε e−βε + 2ε e−2βε ε e βε + 2ε
E mean = =
1 + e−βε + e−2βε e2βε + e βε + 1
e βε + 2
hence U m = N A E mean = N A ε 2βε
e + e βε + 1
S m = [U m (T) − U m (0)]/T + R ln q
NA ε e βε + 2
= + R ln(1 + e−βε + e−2βε )
T e + e βε + 1
2βε
βε(e βε + 2)
= R ( 2βε + ln(1 + e−βε + e−2βε ))
e + e βε + 1
where to go to the last line N A /T = Rβ is used. At high temperatures, βε → 0

and S m tends to R ln 3. At low temperatures, βε → ∞ and S m tends to 0, as
expected.
P12E.9 The data in Problem P12B.8 fit very well to the terms F̃(J) = B̃J(J + 1) with B̃ =
10.593 cm−1 . The rotational contribution to the entropy is given by [12E.8a–
490]
Sm
R
= [U m (T) − U m (0)]/T + R ln q R
where the internal energy is given by [12E.2a–487]
N A ∂q R
U m (T) − U m (0) = − ( )
q R ∂β V
and the partition function is
q R = ∑(2J + 1)e−hc β B̃ J(J+1)

J
In order to compute the entropy down to low temperatures it is necessary to

evaluate the sums term by term rather than approximating them by an integral.
The derivative of q R is
∂q R
( ) = −hc B̃ ∑(2J + 1)[J(J + 1)]e−hc β B̃ J(J+1)
∂β V J
It is convenient to rewrite these expressions in terms of the characteristic vi-

brational temperature θ R = hc B̃/k: using this hc B̃ = kθ R and hcβ B̃ = θ R /T.
With these, the expressions for q R and its derivative become
∂q R
q R = ∑(2J +1)e−θ J(J+1)/T
) = −kθ R ∑(2J +1)[J(J +1)]e−θ J(J+1)/T
R R
(
J ∂β V J
The internal energy is therefore
Rθ R −θ R J(J+1)/T
U m (T) − U m (0) = ∑(2J + 1)[J(J + 1)]e
qR J
The sums are best evaluated using mathematical software and the results are
expressed in terms of the dimensionless parameter T/θ R . The result of such a
calculation is shown in Fig. 12.16
2
Sm
R
/R
0
0 1 2 3 4 5
T/θ R
Figure 12.16
P12E.11 Contributions to the entropy from translation, rotation and vibration are ex-
pected. The molecule has a doubly-degenerate ground electronic state, so this
will also contribute to the entropy. However, the excited electronic states are at
energies very much greater than kT at 298 K (kT at 298 K is 0.026 eV), so their
contribution is negligible.
The translational contribution to the standard molar entropy is given by the

Sackur–Tetrode equation [12E.9b–491]
kTe5/2
Sm
T
= R ln ( ) Λ = h/(2πmkT)1/2
p−○ Λ 3
Taking the mass of F−2 as 38.00 m u

6.6261 × 10−34 J s
Λ= 1/2
[2π(38.00×1.6605 × 10−27 kg)×(1.3806 × 10−23 J K−1 )×(298 K)]
= 1.64... × 10−11 m
(1.3806 × 10−23 J K−1 ) × (298 K) × e5/2
Sm
T
= (8.3145 J K−1 mol−1 ) × ln [ ]
(105 N m−2 ) × (1.64... × 10−11 m)3
= 1.54... × 102 J K−1 mol−1
The rotational constant is given by [11B.7–408], B̃ = ħ/4πcI, with I = µR 2 and

µ = 12 m for a homonuclear diatomic.
ħ ħ
B̃ = =
4πcI 2πcmR 2
1.0546 × 10−34 J s
=
2π(2.9979 × 1010 cm s−1 )×[19.00×1.6605 × 10−27 kg)]×(190.0 × 10−12 m)2
= 0.491... cm−1
The rotational contribution to the entropy is given by [12E.11a–492]; this high-
temperature form is applicable at 298 K because this temperature is much higher
than the characteristic rotational temperature, θ R = hc B̃/k = 0.707 K.
kT
Sm
R
= R (1 + ln )
σ hc B̃
= (8.3145 J K−1 mol−1 ) × (1+
(1.3806 × 10−23 J K−1 ) × (298 K)
ln )
2(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (0.491... cm−1 )
= 52.7... J K−1 mol−1
The characteristic vibrational temperature is

θ V = hc ν̃/k
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (450.0 cm−1 )
=
1.3806 × 10−23 J K−1
= 647.4... K
The vibrational contribution to the standard molar entropy is given by [12E.12b–
492] (note that there is an error in the expression in the text: the argument of
the exponential term in the ln should be negative)
θ V /T
− ln(1 − e−θ /T )]
V
Sm
V
= R[ θ V /T
e −1
= (8.3145 J K−1 mol−1 )×
(647.4... K)/(298 K)
[ − ln[1 − e(−647.4 ... K)/(298 K) ]]
e(647.4 ... K)/(298 K) − 1
= 3.32... J K−1 mol−1
The electronic partition function is q E = g 0 = 2, therefore the electronic con-

tribution to the molar entropy is
Sm
E
= R ln q E = (8.3145 J K−1 mol−1 ) × ln 2 = 5.76... J K−1 mol−1
The molar entropy is therefore
−
○
Sm = Sm
T R
+S m V
+S m E
+S m = 1.54...×102 +52.7...+3.32...+5.76... = 216.1 J K−1 mol−1
P12E.13 It is convenient to calculate the entropy using the approach set out in Problem
P12E.8 in which the two quantities q (the partition function) and q̇ are defined
q = ∑ e−βε j q̇ = ∑ βε j e−βε j
j j
It is shown in the solution to that problem that the entropy can be written in
terms of these as
q̇
S m = R ( + ln q) (12.2)
q
As discussed in Section 11C.3(a) on page 420, the Morse oscillator has a finite
number of bound levels between the ground level and the dissociation limit.
The number of these is found by noting that E υ reaches a maximum value at
the dissociation limit, and therefore this limit is found by solving dE υ /dυ = 0
d
[(υ + 12 )hc ν̃ − (υ + 12 )2 hcx e ν̃] = hc ν̃ − 2(υ + 12 )hcx e ν̃
dυ
solving 0 = hc ν̃ − 2(υ max + 12 )hcx e ν̃ gives υ max = 1/2x e − 1
2
The energy of the lowest state is E 0 = 21 hc ν̃ − 14 hcx e ν̃, therefore the energies
used to evaluate the partition function are
E υ′ = E υ − E 0
= [(υ + 12 )hc ν̃ − (υ + 12 )2 hcx e ν̃] − [ 12 hc ν̃ − 14 hcx e ν̃]
= υhc ν̃ − (υ 2 + υ)hcx e ν̃
The partition function is evaluated from the sum

υ max
q = ∑ e−(υhc ν̃−(υ +υ)hc x e ν̃)/k T
2
υ=0
Defining the characteristic vibrational temperature as θ V = hc ν̃/k gives

υ max
q = ∑ e−(υ−(υ +υ)x e )θ V /T
2
υ=0
The sum needed to compute the quantity q̇ is written, by analogy, as

υ max
θ V −(υ−(υ 2 +υ)x e )θ V /T
q̇ = ∑ (υ − (υ 2 + υ)x e ) e
υ=0 T
These results are used with eqn 12.2 to compute the entropy. For comparison,
for a harmonic oscillator the entropy is given by [12E.12b–492] (note that there
is an error in the expression in the text: the argument of the exponential term
in the ln should be negative)
θ V /T
− ln(1 − e−θ /T )
V
S m /R =
eθ V /T −1
Figure 12.17 compares the entropy for various values of x e with that for the
harmonic case. For the smallest value of x e the entropy is initially larger than
that for the harmonic oscillator. This can be attributed to fact that the energy
levels are more closely spaced for the Morse oscillator than for the harmonic
oscillator. However, at higher temperatures the entropy for the Morse oscillator
starts to level off because there are a finite number of levels, whereas for the
harmonic case the entropy continues to increase without limit as there are an
infinite number of levels. This behaviour is even more pronounced for x e = 0.05
and x e = 0.10, with the plateau at high temperatures being evident.
These plots are, however, somewhat unrealistic. For a typical molecule θ V ≈
1000 K, so at 298 K T/θ V ≈ 0.1, and x e is around 0.001. With these parameters
the contribution to the entropy determined using the Morse levels is 5.0 ×
10−4 × R; the result obtained using the harmonic levels is the same. This is
because there is very little contribution from excited vibrational states, so the
small difference between these low-lying states for the Morse and harmonic
oscillators has no significant effect on the partition function.
harmonic
3.0 x e = 0.02
x e = 0.05
x e = 0.10
2.0
S m /R
1.0
0.0
0 2 4 6 8 10
T/θ V
Figure 12.17
P12E.15 The partition function for a particle confined to a box of length X in one di-
mension is given by [12B.7–469]
q X = X/Λ Λ = h/(2πmkT)1/2
Therefore the partition function for a particle confined to a two-dimensional
box of dimensions X and Y is
q XY = q X qY = XY/Λ 2 = A/Λ 2
where A is the area. Because there are two translational modes, the internal
energy of such a system is given by the equipartition theorem as U = nRT,
where n is the amount in moles. The entropy of n moles is given by [12E.8b–
490]
S = U/T + N k ln qe/N
Ae Ae
= nR + N k ln 2 = nR + nN A k ln 2
Λ N Λ nN A
Am e
= nR + nR ln 2
Λ NA
where in the last step the molar area, A m = A/n, is introduced. The molar
entropy is therefore
Am e Am e
Sm
2D
= R + R ln = R ln e + R ln 2
Λ2 NA Λ NA
A m e2
= R ln
Λ2 NA
The translational molar entropy in three dimensions is given by [12E.9a–491]
Vm e5/2
Sm
3D
= R ln
Λ3 NA
Therefore the molar entropy of condensation is
∆S cond. = S m
2D
− Sm
3D
A m e2 Vm e5/2
= R ln − R ln
Λ2 NA Λ3 NA
Am Λ
= R ln
Vm e1/2
P12E.17 If there are N binucleotides of four different kinds then W = 4 N and
S = k ln W = N k ln 4 = (5 × 108 ) × (1.3806 × 10−23 J K−1 ) × ln 4

= 9.6 × 10−15 J K−1
12F Derived functions

D12F.1 This is discussed in Section 12F.2(c) on page 499.
D12F.3 This is discussed in Section 12F.2(c) on page 499.

E12F.1(a) The equilibrium constant for this dissociation reaction is computed using [12F.12–
499]
g I2 kTΛ 3I2
K= e−hc D̃ 0 /k T
g I2 p−○ qIR2 qIV2 Λ 6I
where g I = 4 and g I2 = 1. The various factors are computed separately. Λ is

given by [12B.7–469]
Λ = h/(2πmkT)1/2
6.6261 × 10−34 J s
Λ I2 =
[2π(253.8×1.6605 × 10−27 kg)×(1.3806 × 10−23 J K−1 )×(1000 K)]1/2
= 3.46... × 10−12 m
Λ I = 4.90... × 10−12 m
The rotational partition function for a homonuclear diatomic in the high-temperature

limit is given by [12B.13a–472]
kT
qR =
2hc B̃
(1.3806 × 10−23 J K−1 ) × (1000 K)
= = 9.31... × 103
2(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (0.0373 cm−1 )
The vibrational partition function is given by [12B.15–474], q V = (1−e−hc ν̃/k T )−1
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (214.36 cm−1 )

hc ν̃/kT =
(1.3806 × 10−23 J K−1 ) × (1000 K)
= 0.308...
q = (1 − e−0.308 ... )−1 = 3.76...
V
The dissociation energy is computed from the well depth (the conversion factor
from eV to cm−1 from inside the front cover is used), using the energy of the
ground state of the harmonic oscillator ε̃ 0 = 21 ν̃
D̃ 0 = D̃ e − ε̃ 0
8065.5 cm−1 1
= (1.5422 eV) × − 2 × (214.36 cm−1 )
1 eV
= 1.23... × 104 cm−1
hc D̃ 0 (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (1.23... × 104 cm−1 )
=
kT (1.3806 × 10−23 J K−1 ) × (1000 K)
= 17.7...
e−hc D̃ 0 /k T
= e−17.7 ... = 1.96... × 10−8
With these results the equilibrium constant is computed as

42 × (1.3806 × 10−23 J K−1 ) × (1000 K) × (3.46... × 10−12 m)3
K=
1 × (105 Pa) × (9.31... × 103 ) × (3.76...) × (4.90... × 10−12 m)6
× (1.96... × 10−8 )
= 3.72 × 10−3
E12F.2(a) The Gibbs energy is computed from the partition function using [12F.8–496],
G(T) = G(0) − nRT ln q/N. As usual, the partition function is factored into
separate contributions from translation, rotation and so on. The factor of 1/N
is usually taken with the translational contribution, so that, for example, the
rotational contribution to the Gibbs energy is −nRT ln q R , or −RT ln q R for the
molar quantity.
For a centro-symmetric linear molecule the rotational partition function in the
high-temperature limit is given by [12B.13a–472]
kT
qR =
2hc B̃
(1.3806 × 10−23 J K−1 ) × (298 K)
=
2(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (0.3902 cm−1 )
= 265.3...
Gm
R
= −RT ln q R
= −(8.3145 J K−1 mol−1 ) × (298 K) × ln(265.3...) = −13.83 kJ mol−1
The vibrational partition function for each mode is given by [12B.15–474], q V =

1/(1−e−β hc ν̃ ), where β = 1/kT. The overall vibrational partition function is the
product of the partition functions of the individual modes, taking into account
any degeneracy. In this case the contribution from the mode at 667.4 cm−1 is
included twice.
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
hcβ =
(1.3806 × 10−23 J K−1 ) × (298 K)
= 4.82... × 10−3 cm
−1
q V1 = (1 − e−hc β ν̃ 1 )
−3 −1
cm)×(1388.2 cm−1 )
= (1 − e−(4.82 ...×10 ) = 1.00...
−3 −1
cm)×(2349.2 cm−1 )
q V2 = (1 − e−(4.82 ...×10 ) = 1.00...
−3 −1
cm)×(667.4 cm−1 )
q V3 = (1 − e−(4.82 ...×10 ) = 1.04...
q V = q V1 × q V2 × (q V3 )2 = (1.00...) × (1.00...) × (1.04...)2 = 1.08...
Hence
Gm
V
= −RT ln q V
= −(8.3145 J K−1 mol−1 ) × (298 K) × ln(1.08...) = −0.204 kJ mol−1
E12F.3(a) The Gibbs energy is computed from the partition function using [12F.8–496],
G(T) = G(0) − nRT ln q/N. As usual, the partition function is factored into
separate contributions from translation, rotation and so on. The factor of 1/N
is usually taken with the translational contribution, therefore the electronic
contribution to the Gibbs energy is −nRT ln q E , or −RT ln q E for the molar
quantity.
The electronic partition function of this two-level system is
q E = g 0 + g 1 e−β hc ν̃
respectively. For an electronic term the degeneracy is 2J + 1, hence g 0 = 2 × 23 +
1 = 4 and g 1 = 2 × 12 + 1 = 2. With the data given
βhc ν̃ = hc ν̃/k × T −1
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (881 cm−1 )
= × T −1
1.3806 × 10−23 J K−1
= (1.26... × 103 K) × T −1
at 500 K
q E = 4 + 2e−(1.26 ...×10
3
K)/(500 K)
= 4.15...
Gm
E
= −RT ln q E
= −(8.3145 J K−1 mol−1 ) × (500 K) × ln(4.15...) = −5.92 kJ mol−1
at 900 K
q E = 4 + 2e−(1.26 ...×10
3
K)/(900 K)
= 4.48...
Gm
E
= −(8.3145 J K −1
mol ) × (900 K) × ln(4.48...) = −11.2 kJ mol−1
−1
P12F.1 The equilibrium constant for this reaction is given by [12F.10b–498]
−
○
qCHD q −○
3 ,m DCl,m −∆ r E 0 /RT
K= −
○
e
qCD q −○
4 ,m HCl,m
It is convenient to consider the contribution of each mode to the fraction in the

above expression separately.
The standard molar translational partition function is qm−○ = Vm−○ /Λ 3 , with Λ =
h/(2πmkT)1/2 , therefore qm−○ goes as m 3/2 . In the fraction all of the other con-
stants cancel to leave
−
○
q −○
3/2
qCHD 3 ,m DCl,m
m CHD3 m DCl
( −
○ −
○
) =( )
qCD4 ,m qHCl,m trans m CD4 m HCl
19.06 × 37.46 3/2
=( ) = 0.964...
20.07 × 36.46
Assuming the high-temperature limit, the rotational partition function for a

heteronuclear diatomic is given by [12B.13b–472], q R = kT/hc B̃, and for a
nonlinear molecule by [12B.14–473], q R = (1/σ)(kT/hc)3/2 (π/ÃB̃C̃)1/2 ; for
CD4 Ã = C̃ = B̃ and for CHD3 C̃ = B̃. The symmetry number is 12 for CD4 and
3 for CHD3 . In the fraction the terms in kT/hc cancel to leave
1/2
qCHD3 qDCl 12 ⎛ B̃ 3CD4 ⎞ B̃ HCl
( ) =
qCD4 qHCl rot 3 ⎝ Ã CHD3 B̃ CHD3 ⎠
2 B̃ DCl
1/2
12 (2.63)3 10.59
= ( ) = 6.23...
3 (2.63) × (3.28)2 5.445
The vibrational partition function is given by [12B.15–474] q V = (1−e−hc ν̃/k T )−1 ,

which is conveniently expressed as q V = (1 − e−(1.4388 cm K)ν̃/T )−1 . This term
is temperature dependent and so needs to be re-evaluated at each temperature.
The vibrational partition function for CHD3 and CD4 is the product of the
partition function for each normal mode, raised to the power of its degeneracy.
For example
V V V 2 V 3 V 3
qCD 4
= q(2109 cm−1 ) × (q(1092 cm−1 ) ) × (q(2259 cm−1 ) ) × (q(996 cm−1 ) )
The term ∆ r E 0 is computed as
∆ r E 0 = E 0 (CHD3 ) + E 0 (DCl) − E 0 (CD4 ) − E 0 (HCl)
To a good approximation it can be assumed that the pure electronic energy

of a species is unaffected by isotopic substitution, however the vibrational zero
point energy will be affected. For a harmonic oscillator the energy of the ground
state is 21 hc ν̃, therefore to compute the total vibrational zero point energy of
CHD3 and CD4 the contribution from each normal mode has to be taken into
account; a mode with degeneracy g contributes g × 21 hc ν̃.
E 0 (CHD3 )vib = 12 N A hc(2993 + 2142 + 3 × 1003 + 2 × 1291 + 2 × 1036)

= N A hc(6399 cm−1 )
E 0 (CD4 )vib = 12 N A hc(2109 + 2 × 1092 + 3 × 2259 + 3 × 996)
= N A hc(7029 cm−1 )
E 0 (HCl)vib = N A hc(1495.5 cm−1 ) E 0 (DCl)vib = N A hc(1072.5 cm−1 )
∆ r E 0 = N A hc(6399 + 1072.5 − 7029 − 1495.5) = N A hc(−1053 cm−1 )
Thus the term −∆ r E 0 /RT evaluates as
−∆ r E 0 −N A hc(−1053 cm−1 )
=
RT RT
= −(6.0221 × 1023 mol−1 ) × (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
(−1053 cm−1 )
× = (1515 K)/T
(8.3145 J K−1 mol−1 ) × T
With these expressions the equilibrium constant is evaluated using mathemati-

cal software and the results are plotted as a function of temperature in Fig 12.18.
At 300 K K = 945 and at 1000 K K = 36.9; the value of the equilibrium constant
is dominated by the symmetry factors and the e−∆ r E 0 /RT term.
800
600
K
400
200
300 400 500 600 700 800 900 1 000

T/K
Figure 12.18
P12F.3 In the absence of a magnetic field the ground state of the I atom, with term
symbol 2 P3/2 , has a degeneracy given by (2J + 1) = (2 × 32 + 1) = 4. When a
magnetic field is applied this level splits into four states characterised by M J =
+ 32 , + 21 , − 12 , − 32 . The energy of these states is given, by analogy with [16A.11c–
636], by E M J = g µ B BM J , where the g-value is given as 43 , µ B is the Bohr mag-
neton, and B is the applied magnetic field. The energies are therefore E±3/2 =
±2µ B B, E±1/2 = ± 23 µ B B, giving the partition function as
q E = e2µ B B/k T + e(2/3)µ B B/k T + e−2µ B B/k T + e−(2/3)µ B B/k T
Because it is expected that µ B B ≪ kT the exponentials are expanded to second

order; letting x = µ B B/kT gives
q E = [1 + 2x + 2x 2 ] + [1 + (2/3)x + (2/9)x 2 ]
+ [1 − 2x + 2x 2 ] + [1 − (2/3)x + (2/9)x 2 ]
= 4 + (40/9)x 2 = 4 [1 + (10/9)(µ B B/kT)2 ]
As is seen, the linear terms cancel which is why it is necessary to expand to

second order. In the absence of a magnetic field q E = 4, and because the
equilibrium is I2 ↽ ⇀ 2 I, the electronic partition function appears squared
ÐÐÐ
in the numerator of the expression for K. Therefore
2 2
K(B) ⎛ 4 [1 + (10/9)(µ B B/kT) ] ⎞
=
K(0) ⎝ 4 ⎠
2
= [1 + (10/9)(µ B B/kT)2 ] ≈ 1 + (20/9)(µ B B/kT)2
where to go to the final expression only the squared term is retained.

For a change of 1%
(20/9)(µ B B/kT)2 = 0.01

hence B 2 = 0.01 × (9/20) × (kT/µ B )2
9 1/2 (1.3806 × 10−23 J K−1 ) × (1000 K)
B=( ) ×
2000 9.2740 × 10−24 J T−1
= 100 T
This is a very strong magnetic field which at present can only be generated by
special techniques and only then for a very short times.
P12F.5 The standard molar Gibbs energy is computed from the partition function us-
−
○ −
○
ing [12F.9b–497], G m (T) = G m (0) − RT ln qm−○ /N A . As usual, the partition
function is factored into separate contributions from translation, rotation and
so on. The factor of 1/N A is usually taken with the translational contribution.
The standard molar translational partition function is given by qm−○ = Vm−○ /Λ 3 =
RT/p−○ Λ 3 . Taking the mass of Cl2 O2 as 2(35.45+16.00) = 102.9 m u , Λ is given
by [12B.7–469]
Λ = h/(2πmkT)1/2
6.6261 × 10−34 J s
=
[2π(102.9×1.6605 × 10−27 kg)×(1.3806 × 10−23 J K−1 )×(200 K)]1/2
= 1.21... × 10−11 m
RT
qm−○ /N A = −○ 3
p Λ NA
(8.3145 J K−1 mol−1 ) × (200 K)
=
(105 N m−2 ) × (1.21... × 10−11 m)3 × (6.0221 × 1023 mol−1 )
= 1.53... × 107
For a nonlinear molecule the rotational partition function is given by [12B.14–

473], q R = (1/σ)(kT/hc)3/2 (π/ÃB̃C̃)1/2 ; here σ = 2. If the rotational con-
stants are expressed in frequency units this expression becomes
q R = (1/σ)(kT/h)3/2 (π/ABC)1/2
3/2
(1.3806 × 10−23 J K−1 ) × (200 K)
= 21 ( )
6.6261 × 10−34 J s
1/2
π
×( )
(1.31094 × 1010 Hz) × (2.4098 × 109 Hz) × (2.1397 × 109 Hz)
= 2.89... × 104
The vibrational partition function for each mode is given by [12B.15–474], q V =

1/(1 − e−β hc ν̃ ), where β = 1/kT.
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )

hcβ =
(1.3806 × 10−23 J K−1 ) × (200 K)
= 7.19... × 10−3 cm
−1
q V1 = (1 − e−hc β ν̃ 1 )
−3 −1
cm)×(753 cm−1 )
= (1 − e−(7.19 ...×10 ) = 1.00...
The partition functions for the other normal modes evaluate to 1.02..., 1.12...,
1.66..., 1.00..., 1.05... in order of the given modes. The overall vibrational par-
tition function is the product of these individual contributions: q V = 2.03....
The overall partition function is the product of these contributions from the
different modes, therefore
−
○ −
○
Gm (200)−G m (0) = − RT ln qm−○ /N A
= −(8.3145 J K−1 mol−1 ) × (200 K)
×ln [(1.53... × 107 )×(2.89... × 104 )×(2.03...)]
= −45.8 kJ mol−1

I12.1 Note that there is an error in the question: the expression for ξ(β) should
include an additional factor of g(J). To make the notation more compact the
energy levels will be written ε J and the degeneracies g J ; derivatives with respect
to β will be assumed to be at constant V .
First, an expression for C V is developed.
1 dq 1 d −βε
U = −N = −N ∑ gJ e J
q dβ q dβ J
1
= N ∑ g J ε J e−βε J
q J
Noting that d/dT = −kβ 2 (d/dβ)
dU dU
CV = = −kβ 2
dT dβ
⎡
d ⎢1 ⎤
⎢ ∑ g J ε J e−βε J ⎥
⎥
= −N kβ 2
dβ ⎢⎣ q J ⎥
⎦
⎡ −1 dq 1 ⎤
g J ε J e−βε J − ∑ g J ε 2J e−βε J ⎥
⎢ ⎥
= −N kβ 2 ⎢ 2 ∑
⎢ q dβ J
⎣ q J
⎥
⎦
⎡ ⎤
2⎢ 1 ⎛ 1
⎢ −βε J ′ ⎞ ⎛ −βε J ⎞
⎥
2 −βε J ⎥
= −N kβ ⎢ 2 ∑ g J ′ ε J ′ e ∑ gJ εJ e − ∑ gJ εJ e ⎥
⎢ q ⎝ J′ ⎠⎝ J ⎠ q J ⎥
⎣ ⎦
The numerator and denominator of the final term in the bracket are both mul-
tiplied by q, and then a factor of 1/q 2 is taken outside the bracket to give
⎡ ⎤
−N kβ 2 ⎢⎛ ⎥
⎢ ∑ g J ′ ε J ′ e−βε J′ ⎞ ⎛∑ g J ε J e−βε J ⎞ − q ∑ g J ε 2 e−βε J ⎥
CV = ⎢⎝ J ⎥
q2 ⎢ J′ ⎠⎝ J ⎠ J ⎥
⎣ ⎦
⎡
−N kβ ⎢
2⎢⎛
−βε ⎞ ⎛ −βε ⎞
= ∑ g J ′ ε J ′ e J′ ∑ g J ε J e J
q 2 ⎢⎢ ⎝ J ′ ⎠⎝ J ⎠
⎣
⎤
⎛ ⎞⎛ ⎞⎥
− ∑ g J ′ e−βε J′ ∑ g J ε 2J e−βε J ⎥⎥
⎝ J′ ⎠⎝ J ⎠⎥
⎦
The product of the sums are next rewritten as double sums
−N kβ 2 ⎡ ⎤
⎢∑ g J g J ′ ε J ε J ′ e−β(ε J +ε J′ ) − ∑ g J g J ′ ε 2J e−β(ε J +ε J′ ) ⎥
⎢ ⎥
CV =
q2 ⎢ J, J ′
⎣ J, J ′
⎥
⎦
Taking a hint from the final result, consider the double sum
−β(ε J +ε J ′ )
∑(ε J − ε J ′ ) g J g J ′ e
2
J, J ′
= ∑ ε 2J g J g J ′ e−β(ε J +ε J′ ) + ∑ ε 2J ′ g J g J ′ e−β(ε J +ε J′ ) − 2 ∑ ε J ε J ′ g J g J ′ e−β(ε J +ε J′ )

J, J ′ J, J ′ J, J ′
The first two sums only differ by swapping the indices J and J ′ , so they are in
fact identical. Hence the last line may be written
= 2 ∑ ε 2J g J g J ′ e−β(ε J +ε J′ ) − 2 ∑ ε J ε J ′ g J g J ′ e−β(ε J +ε J′ )
J, J ′ J, J ′
Apart from an overall sign and a factor of 21 , these two terms are the same as
those in the bracket in the expression for C V above, hence
−N kβ 2 ⎡ ⎤
⎢∑ g J g J ′ ε J ε J ′ e−β(ε J +ε J′ ) − ∑ g J g J ′ ε 2J e−β(ε J +ε J′ ) ⎥
⎢ ⎥
CV =
q2 ⎢ J, J ′
⎣ J, J ′
⎥
⎦
N kβ 2 −β(ε J +ε J ′ )
∑(ε J − ε J ′ ) g J g J ′ e
2
=
2q 2 J, J ′
which is the required expression.
For a diatomic βε J = βhc B̃J(J + 1) = hc B̃J(J + 1)/kT = θ R J(J + 1)/T, where

θ R = hc B̃/k; the degeneracy is g J = (2J + 1). For the molar quantity N A kβ 2 =
N A k/k 2 T 2 = R/k 2 T 2 . The molar hear capacity is therefore given by
1 1
C V ,m /R = (hc B̃)2
k2 T 2 2q 2
[J(J+1)+J ′ (J ′ +1)]/T
× ∑[J(J + 1) − J ′ (J ′ + 1)]2 (2J + 1)(2J ′ + 1) e−θ
R
J, J ′
2
θR 1
=( )
T 2q 2
[J(J+1)+J ′ (J ′ +1)]/T
× ∑[J(J + 1) − J ′ (J ′ + 1)]2 (2J + 1)(2J ′ + 1) e−θ
R
J, J ′
This expression is used to generate the curves in Fig. 12.19 for particular pairs of
values of J and J ′ , that is just one term from the double sum. However, the term
for J = 0, J ′ = 1 is identical to that for J = 1, J ′ = 0, so the curves plotted in the
figure are twice the value for the particular combination of J and J ′ indicated.
This double sum is not a particularly efficient method for computing the heat
capacity, but it can be evaluated using mathematical software to give the curve
also shown in Fig. 12.19. For a plot up to T/θ R = 5 if is sufficient to consider
contributions from levels with J ≤ 10; this makes the calculation more tractable.
J = 0, J ′ = 1
1.0 J = 0, J ′ = 2
J = 1, J ′ = 2
J = 1, J ′ = 3
C V ,m /R
J = 0, J ′ = 3
0.5 total
0.0
0 1 2 3 4 5
T/θ R
Figure 12.19
I12.3 (a) In the high-temperature limit, the rotational partition function of an asym-
metric rotor is given by [12B.14–473], q R = (1/σ)(kT/hc)3/2 (π/ÃB̃C̃)1/2 ,
where σ is the symmetry number. The point group for ethene is D 2h
y
which contains the rotational operations (E, C 2x , C 2 , C 2z ); therefore σ = 4.
1 kT 3/2 π 1/2
qR = ( ) ( )
4 hc ÃB̃C̃
3/2
1 (1.3806 × 10−23 J K−1 ) × (298.15 K)
= ×( )
4 (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
1/2
π
×( )
(4.828 cm−1 ) × (1.0012 cm−1 ) × (0.8282 cm−1 )
= 660.6
(b) Pyridine belongs to the point group C 2v which contains the rotational
operations (E, C 2 ); therefore σ = 2.
1 kT 3/2 π 1/2
qR = ( ) ( )
2 hc ÃB̃C̃
3/2
1 (1.3806 × 10−23 J K−1 ) × (298.15 K)
= ×( )
2 (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
1/2
π
×( )
(0.2014 cm ) × (0.1936 cm−1 ) × (0.0987 cm−1 )
−1
= 4.26 × 104
13 Molecules in motion
13A Transport properties of a perfect gas

D13A.1 The diffusion constant is given by [13A.9–512], D = 13 λυ mean . The mean free
path λ decreases as the pressure is increased ([13A.1a–508]), so D decreases
with increasing pressure and, as a result, the gas molecules diffuse more slowly.
The mean speed υ mean increases with the temperature ([13A.1b–508]), so D also
increases with temperature. As a result, molecules in a hot gas diffuse more
quickly than those when the gas is cool (for a given concentration gradient).
Because the mean free path increases when the collision cross-section σ of
the molecules decreases, the diffusion coefficient is greater for small molecules
than for large molecules.
The viscosity is given by [13A.11c–514], η = 13 υ mean λmN . The mean free path
is inversely proportional to the pressure and N is proportional to the pres-
sure, therefore the product λN , and hence the viscosity, is independent of
pressure. The physical reason for this pressure-independence is that as the
pressure increases more molecules are available to transport the momentum,
but they carry it less far on account of the decrease in mean free path. The
mean speed goes as T 1/2 (at constant volume) and so the viscosity increases
with temperature. This is because at high temperatures the molecules travel
more quickly, so the flux of momentum is greater.
E13A.1(a) The rate of effusion, r is given by [13A.12–515], r = pA 0 N A /(2πMRT)1/2 ; this
rate is the number of molecules escaping through the hole in a particular period
of time, divided by that time. The mass loss ∆m in period ∆t is therefore ∆m =
∆t pA 0 N A /(2πMRT)1/2 × m, where m is the mass of a molecule. This mass is
written m = M/N A and so it follows ∆m = ∆t pA 0 M 1/2 /(2πRT)1/2 . This is
rearranged to give an expression for p
∆m(2πRT)1/2 ∆m 2πRT 1/2

p= = ( )
∆tA 0 M 1/2 ∆tA 0 M
1/2
2.85 × 10−4 kg 2π × (8.3145 J K−1 mol−1 ) × (673.15 K)
= ( )
(400 s) × π × (2.5 × 10−4 m)2 0.100 kg mol−1
= 2.15 × 103 Pa
488 13 MOLECULES IN MOTION

of time, divided by that time. In this experiment the pressure changes so the rate
of effusion changes throughout the experiment; nevertheless, the rate is always
proportional to M −1/2 . The two experiments involve comparing the time for
the same drop in pressure, therefore the only factor that affects this time is the
molar mass of the effusing gas. Because the rate is proportional to M −1/2 the
time for a given fall in pressure will be proportional to the inverse of this, that
is M 1/2 . It follows that
rate for gas A time for gas B M B 1/2

= =( )
rate for gas B time for gas A MA
Therefore
42 s M N2 1/2 52 2
=( ) hence M A = (28.02 g mol−1 ) ( ) = 43.0 g mol−1
52 s MA 42
E13A.3(a) The rate of effusion is given by [13A.12–515], dN/dt = pA 0 N A /(2πMRT)1/2 ;

this is the rate of change of the number of molecules. If it is assumed that the
gas is perfect, the equation of state pV = N kT allows the number to be written
as N = pV /kT, and therefore dN/dt = (V /kT)dp/dt. The rate of change of
the pressure is therefore
dp kT pA 0 N A RTA 0 A 0 RT 1/2
=− = − × p = − ( ) ×p
dt V (2πMRT)1/2 V (2πMRT)1/2 V 2πM
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
α
The minus sign is needed because the pressure falls with time. This differen-
tial equation is separable and can be integrated between p = p i and p = p f ,
corresponding to t = 0 and t = t.
pf t
∫ (1/p) dp = ∫ −α dt hence ln(p f /p i ) = −αt
pi 0
The time for the pressure to drop by the specified amount is therefore
V 2πM 1/2
t = − ln(p f /p i )/α = ln(p i /p f ) ( )
A 0 RT
1/2
8.0 × 104 Pa (3.0 m3 ) 2π × (3.200 × 10−2 kg mol−1 )
= ln ( ) ( )
7.0 × 10 Pa [π(10 m) ] (8.3145 J K−1 mol−1 ) × (298 K)
4 −4 2
= 1.15 × 105 s = 1.3 days
E13A.4(a) For a perfect gas, the collision flux Z w is [13A.7a–511], Z w = p/(2πmkT)1/2 .

The number of argon molecule collisions within area A in time interval t is
therefore N = Z w At . The mass m is written in terms of the molar mass M:
m = M/N A .
1/2
p pN A
N = Z w At = 1/2
At = At
(2πmkT) (2πMkT)1/2
(90 Pa) × (6.0221 × 1023 mol−1 )1/2 × [(2.5 × 3.0) × 10−6 m2 ] × (15 s)
=
[2π × (0.03995 kg mol−1 ) × (1.3806 × 10−23 J K−1 ) × (500 K)]1/2
= 1.9 × 1020 collisions
E13A.5(a) The diffusion constant is given by [13A.9–512], D = 31 λυmean , where λ is the

mean free path length λ = kT/σ p [13A.1a–508], and υmean is the mean speed
υmean = (8RT/πM)1/2 [13A.1b–508].
1 kT 8RT 1/2
D= [ ]
3 σ p πM
(1.3806 × 10−23 J K−1 )×(293.15 K)
=
3 × (3.6 × 10−19 m2 ) × (p/Pa)
1/2
8×(8.3145 J K−1 mol−1 )×(293.15 K)
×[ ]
π×(0.03995 kg mol−1 )
1
= (1.477... m2 s−1 ) ×
p/Pa
The flux of argon atoms J z is related to the diffusion coefficient D and the
concentration gradient dN /dz by [13A.4–509], J z = −DdN /dz. From the
perfect gas equation, pV = N kT, the number density is expressed in terms
of the pressure as N = N/V = p/kT. With this, the concentration gradient is
written in terms of the pressure gradient: dN /dz = (1/kT)dp/dz, and hence
the flow is J z = −(D/kT)dp/dz
−D dp −1 (1.47... m2 s−1 ) × (1.0 × 105 Pa m−1 )

Jz = = ×
kT dz p/Pa (1.3806 × 10−23 J K−1 ) × (293.15 K)
1
= −(3.64... × 1025 m−2 s−1 ) ×
p/Pa
p/Pa D/(m2 s−1 ) J z /(m−2 s−1 ) (J z /N A )/(mol m−2 s−1 )

1.00 1.48 −3.65 × 1025 −60.6
1.00 × 105 1.48 × 10−5 −3.65 × 1020 −6.06 × 10−4
1.00 × 107 1.48 × 10−7 −3.65 × 1018 −6.06 × 10−6
E13A.6(a) The thermal conductivity is given by [13A.10c–513], κ = νpD/T, where the

diffusion coefficient D is given by [13A.9–512], D = λυ mean /3. The mean free
path λ is given by [13A.1a–508], λ = kT/σ p, and the mean speed υ mean is
given by [13A.1b–508], υmean = (8RT/πM)1/2 . The quantity ν is the number
of quadratic contributions to the energy, and this is related to the heat capacity
by C V ,m = νkN A , hence ν = C V ,m /kN A . The thermal conductivity is therefore

expressed as
νpD νpλυ mean C V ,m p kT 8RT 1/2 C V ,m 8RT 1/2

κ= = = ( ) = ( )
T 3T kN A 3T σ p πM 3σ N A πM
12.5 J K−1 mol−1
hence κ =
3 × (3.6 × 10−19 m2 ) × (6.0221 × 1023 mol−1 )
1/2
8 × (8.3145 J K−1 mol−1 ) × (298 K)
×( )
π × (3.995 × 10−2 kg mol−1 )
= 7.6 × 10−3 J K−1 m−1 s−1
E13A.7(a) The thermal conductivity is given by [13A.10c–513], κ = νpD/T, where the

diffusion coefficient D is given by [13A.9–512], D = λυ mean /3. The mean free
path λ is given by [13A.1a–508], λ = kT/σ p, and the mean speed υ mean is
given by [13A.1b–508], υmean = (8RT/πM)1/2 . The quantity ν is the number
of quadratic contributions to the energy, and this is related to the heat capacity
by C V ,m = νkN A , hence ν = C V ,m /kN A . The thermal conductivity is therefore
expressed as
νpD νpλυ mean C V ,m p kT 8RT 1/2 C V ,m 8RT 1/2

κ= = = ( ) = ( )
T 3T kN A 3T σ p πM 3σ N A πM
Rearranging gives an expression for σ in terms of the thermal conductivity
C V ,m 8RT 1/2
σ= ( )
3κN A πM
The value of C p,m is given in the Resource section; C V ,m is found using C p,m −
C V ,m = R for a perfect gas.
(20.786 J K−1 mol−1 ) − (8.3145 J K−1 mol−1 )

σ=
3 × (4.65 × 10−2 J K−1 m−1 s−1 ) × (6.0221 × 1023 mol−1 )
1/2
8 × (8.3145 J K−1 mol−1 ) × (273 K)
×( ) = 0.0795 nm2
π × (2.018 × 10−2 kg mol−1 )
The value reported in Table 1B.2 on page 17 is 0.24 nm2 .
E13A.8(a) The flux of energy is given by [13A.3–509], J z = −κ dT/dz. The value of the
thermal conductivity κ for Ar at 298 K is determined in Exercise E16A.3(a) as
7.6 × 10−3 J K−1 m−1 s−1 . As is seen in that Exercise, κ ∝ T 1/2 provided that
the heat capacity is constant over the temperature range of interest. It therefore
follows that κ 280 K = (280 K/298 K) κ 298 K = 7.40... × 10−3 J K−1 m−1 s−1 .
1/2
With these data the flux is computed as
Jz = −κ dT/dz = −(7.40... × 10−3 J K−1 m−1 s−1 ) × (10.5 K m−1 )

= −0.078 J m−2 s−1
E13A.9(a) The flux of energy is given by [13A.3–509], J z = −κ dT/dz, where κ is the

thermal conductivity and the negative sign indicates flow of heat is towards
the lower temperature. The rate of energy transfer is r = J z A, where A is the
cross-sectional area. The temperature gradient is approximated as dT/dz =
∆T/∆z; because 1 W = 1 J s−1 it follows that 24 mW K−1 m−1 is equivalent to
2.4 × 10−2 J K−1 m−1 s−1
∆T
r = J z A = −κA
∆z
[(−15) − (28)] K
= −(2.4 × 10−2 J K−1 m−1 s−1 ) × (1.0 m2 ) ×
0.010 m
= 103 J s−1 = 103 W
Hence a heater of power 103 W is required to make good the loss of heat.
E13A.10(a) The viscosity η is given by [13A.11c–514], η = pMD/RT. In turn the diffu-

sion constant is given by [13A.9–512], D = 31 λυmean , where λ is the mean free
path length λ = kT/σ p [13A.1a–508], and υmean is the mean speed υmean =
(8RT/πM)1/2 [13A.1b–508]. The first step is to find an expression for η as a
function of temperature
pMD pM kT 8RT 1/2 M 8RT 1/2 1 8RM 1/2 1/2

η= = ( ) = ( ) = ( ) T
RT RT 3σ p πM 3σ N A πM 3σ N A π
1
=
3 × (4.0 × 10 m ) × (6.0221 × 1023 mol−1 )
−19 2
1/2
8 × (8.3145 J K−1 mol−1 ) × (0.029 kg mol−1 )
×( ) T 1/2
π
= (1.08... × 10−6 kg K−1/2 m−1 s−1 ) × (T/ K)1/2
where 1 J = 1 kg m2 s−2 has been used to arrive at the units on the final line. Us-
ing this expression the following table is drawn up (recall that 10−7 kg m−1 s−1 =
1 µP)
T/K η/(kg m−1 s−1 ) η/(µP)

273 1.79 × 10−5 179
298 1.87 × 10−5 187
1000 3.43 × 10−5 343
E13A.11(a) In the solution to Exercise E13A.10(a) it is shown that
1 8RMT 1/2 1 8RMT 1/2

η= ( ) hence σ= ( )
3σ N A π 3ηN A π
Recalling that 10−7 kg m−1 s−1 = 1 µP, the cross section is computed as
1
σ=
3 × (2.98 × 10−5 kg m−1 s−1 ) × (6.0221 × 1023 mol−1 )
1/2
8 × (8.3145 J K−1 mol−1 ) × (0.02018 kg mol−1 ) × (273 K)
×( )
π
= 0.201 nm2

of time, divided by that time. The mass loss ∆m in period ∆t is therefore ∆m =
∆t pA 0 N A /(2πMRT)1/2 × m, where m is the mass of a molecule. This mass is
written m = M/N A and so it follows ∆m = ∆t pA 0 M 1/2 /(2πRT)1/2 . Evaluating
this with the values given
∆t pA 0 M 1/2
∆m =
(2πRT)1/2
(7200 s) × (0.835 Pa) × π × ( 12 × 2.50 × 10−3 m)2 × (0.260 kg mol−1 )1/2
=
[2 × π × (8.3145 J K−1 mol−1 ) × (400 K)]1/2
= 1.04 × 10−4 kg = 104 mg
P13A.1 In the solution to Exercise E13A.10(a) it is shown that
1 8RMT 1/2 1 8RMT 1/2

η= ( ) hence σ= ( )
3σ N A π 3ηN A π
At 270 K and 1.00 bar

1
σ=
3 × (9.08 × 10−6 kg m−1 s−1 ) × (6.0221 × 1023 mol−1 )
1/2
8 × (8.3145 J K−1 mol−1 ) × (0.0170 kg mol−1 ) × (270 K)
×( )
π
= 6.00... × 10−19 m2
The collision cross-section is σ = π(2r)2 , where r is the molecular radius of

NH3 and d = 2r is the effective molecular diameter. With the value of σ
determined above d is found as 437 pm . A similar calculation at 490 K and
10.0 bar gives σ = 4.21... × 10−19 m2 and d = 366 pm .
P13A.3 In the solution to Exercise E13A.5(a) it is shown that the diffusion constant is
given by
1 kT 8RT 1/2
D= [ ]
3 σ p πM
If the gas is assumed to be perfect then the equation of state pV = N kT can be

used to find the number density N as N = N/V = p/kT. The collision cross
section is estimated as σ = π(2a 0 )2 where a 0 is the Bohr radius. A density of
1 atom cm−3 corresponds to N = 1 × 106 m−3 .
1 kT 8RT 1/2 1 8RT 1/2

D= [ ] = [ ]
3 σ p πM 3[π(2a 0 )2 ]N πM
1
=
3 × [π(2 × 5.2918 × 10−11 m)2 ] × (1 × 106 m−3 )
1/2
8 × (8.3145 J K−1 mol−1 ) × (10 × 103 K)
×[ ]
π(1.0079 × 10−3 kg mol−1 )
= 1.37 × 1017 m2 s−1
The thermal conductivity is given in terms of the diffusion constant by [13A.10c–

513], κ = νpD/T, which is rewritten using N = p/kT as κ = νN kD. For an
atom there are just three degrees of translational freedom, ν = 32 .
κ = νN kD = 23 ×(1 × 106 m−3 )×(1.3806 × 10−23 J K−1 )×(1.37... × 1017 m2 s−1 )

= 2.84 J K−1 m−1 s−1
For a gas at ambient temperature and pressure a typical value for the diffusion
coefficient is D = 1.5 × 10−5 m2 s−1 , and a typical value for the thermal con-
ductivity is κ = 0.025 J K−1 m−1 s−1 . The diffusion constant is much higher in
interstellar space when compared to ambient conditions because in interstellar
space the much higher temperature results in a higher mean speed, and the
much lower pressure results in a longer mean free path. Molecules move more
quickly and experience fewer collisions, resulting in more rapid diffusion.
Because κ ∝ N D and D ∝ 1/N , the value of the thermal conductivity is
unaffected by the change in number density in going from ambient pressure to
interstellar conditions. The higher thermal conductivity in the latter is therefore
attributable to the higher mean speed.
The kinetic theory of gases assumes that the rate of atomic collisions is very
high such that thermal equilibrium is established quickly. However, at such
a dilute concentration, the timescales on which particles exchange energy by
collision make this assumption questionable. In fact, atoms are more likely to
interact with photons from stellar radiation than with other atoms.
P13A.5 The rate of effusion, r is given by [13A.12–515], r = pA 0 N A /(2πMRT)1/2 . The

area of the slit is A 0 = (10 mm) × (1.0 × 10−2 mm) = 0.1 mm2 = 1.0 × 10−7 m2 .
pA 0 N A
r=
(2πMRT)1/2
(p/ Pa) × (1.0 × 10−7 m2 ) × (6.0221 × 1023 mol−1 )
=
[2π × (M/ kg mol−1 ) × (8.3145 J K−1 mol−1 ) × (380 K)]1/2
(p/ Pa)
= (4.27... × 1014 s−1 ) ×
(M/ kg mol−1 )1/2
For cadmium, r = (4.27... × 1014 s−1 ) × 0.13/(0.11241)1/2 = 1.7 × 1014 s−1 .

Hence there are 1.7 × 1014 atoms per second in the beam.
For mercury, r = (4.27... × 1014 s−1 ) × 12/(0.20059)1/2 = 1.1 × 1016 s−1 . Hence
there are 1.1 × 1016 atoms per second in the beam.
13B Motion in liquids

D13B.1 The Grotthuss mechanism for conduction by protons in water is described
in Section 13B.2(a) on page 519 and illustrated in Fig. 13B.2 on page 521. It
seems plausible that such a mechanism could also occur in the relatively open
hydrogen bonded structure of ice.
E13B.1(a) The ion molar conductivity λ is given in terms of the mobility u by [13B.10–521],
λ = zuF, where z is the charge number of the ion (unsigned) and F is Faraday’s
constant; it follows that u = λ/zF. Note that 1 S = 1 C V−1 s−1 .
3.87 mS m2 mol−1
u Li+ = = 4.01×10−5 mS m2 C−1 = 4.01 × 10−8 m2 V−1 s−1
(1)(96485 C mol−1 )
5.01 mS m2 mol−1
u Na+ = = 5.19×10−5 mS m2 C−1 = 5.19 × 10−8 m2 V−1 s−1
(1)(96485 C mol−1 )
7.35 mS m2 mol−1
u K+ = = 7.62×10−5 mS m2 C−1 = 7.62 × 10−8 m2 V−1 s−1
(1)(96485 C mol−1 )
E13B.2(a) The ion molar conductivity λ is given in terms of the mobility u by [13B.10–521],
λ = zuF, where z is the charge number of the ion (unsigned) and F is Faraday’s
constant. Note that 1 S = 1 C V−1 s−1 .
λ = zuF = (1)×(7.91×10−8 m2 V−1 s−1 )×(96485 C mol−1 ) = 7.63 mS m2 C−1
E13B.3(a) The drift speed s of an ion is given by [13B.8b–520], s = uE, where E is the
electric field strength. This field strength is given by E = ∆ϕ/l where ∆ϕ is the
potential difference between two electrodes separated by distance l.
∆ϕ 25.0 V
s = uE = u = (7.92 × 10−8 m2 V−1 s−1 ) ×
l 7.00 × 10−3 m
−4 −1 −1
= 2.83 × 10 m s = 283 µm s
E13B.4(a) The Einstein relation, [13B.13–522], u = zDF/RT, gives the relationship be-
tween the mobility u, the charge number of the ion z, and the diffusion coeffi-
cient D.
uRT (7.40 × 10−8 m2 V−1 s−1 ) × (8.3145 J K−1 mol−1 ) × (298 K)

D= =
zF (1) × (96485 C mol−1 )
= 1.90 × 10−9 m2 s−1
E13B.5(a) The temperature dependence of the viscosity η is given by [13B.1–517], η =

η 0 eE a /RT , where η 0 is viscosity in the limit of high temperature and E a is the
associated activation energy. Taking the natural logarithm gives ln η = ln η 0 +
E a /RT. Hence
Ea 1 1
ln η 1 − ln η 2 = (ln η 0 + E a /RT1 ) − (ln η 0 + E a /RT2 ) = ( − )
R T1 T2
Rearranging gives an expression for the activation energy
ln (η 1 /η 2 )
Ea = R
(T1−1 − T2−1 )
ln [(1.002 cP)/(0.7975 cP)]
= (8.3145 J K−1 mol−1 )
[(293.15 K)−1 − (303.15 K)−1 ]
= 16.9 kJ mol−1
E13B.6(a) According to the law of independent migration of ions, the limiting molar
conductivity Λ○m of an electrolyte is given by the sum of the limiting molar
conductivities λ i of the ions present, [13B.6–519], Λ○m = ν+ λ+ + ν− λ− ; in this
expression ν+ and ν− are the numbers of cations and anions provided by each
formula unit of electrolyte. For each of the given electrolytes it follows that
Λ○AgI = λ Ag+ + λ I− Λ○NaNO3 = λ Na+ + λ NO3 − Λ○AgNO3 = λ Ag+ + λ NO3 −
These expressions are manipulated to give Λ○AgI
Λ○AgI = λ Ag+ + λ I−
= (Λ○AgNO3 − λ NO3 − ) + (Λ○NaI − λ Na+ ) = Λ○AgNO3 + Λ○NaI − Λ○NaNO3
= (13.34 + 12.69 − 12.16) mS m2 mol−1 = 13.87 mS m2 mol−1
P13B.1 The temperature dependence of the viscosity η is given by [13B.2–517], η =
η 0 eE a /RT , where E a is the activation energy . Taking the natural logarithm gives
ln η = ln η 0 +E a /RT. A plot of ln η against (1/T) therefore has slope E a /R; such
a plot is shown in Fig. 13.1.
θ/○ C T/K η/cP 1/(T/K) ln (η/cP)

10 283 0.758 0.003 53 −0.277
20 293 0.652 0.003 41 −0.428
30 303 0.564 0.003 30 −0.573
40 313 0.503 0.003 19 −0.687
50 323 0.442 0.003 10 −0.816
60 333 0.392 0.003 00 −0.936
70 343 0.358 0.002 92 −1.027
0.0
−0.5
ln (η/cP)
−1.0
−1.5
0.0028 0.0030 0.0032 0.0034 0.0036
1/(T/K)
Figure 13.1
The data are a good fit to a straight line with equation
ln (η/ cP) = (1.2207 × 103 ) × 1/(T/K) − 4.5939
The activation energy is computed from the slope
E a = R × (slope)
= (8.3145 J K−1 mol−1 )(1.2207 × 103 K) = 10.15 kJ mol−1
P13B.3 The molar conductivity Λm is defined by [13B.4–518], Λm = κ/c, where κ is

conductivity and c is concentration. The Kohlrausch law, [13B.5–518], gives the
variation of the molar conductivity with concentration as Λm = Λ○m − Kc 1/2 .
Hence a plot of Λm against c 1/2 has slope equal to −K and y-intercept equal to
the limiting molar conductivity Λ○m . In computing Λm the concentration needs
to be converted from mol dm−3 to mol m−3 . The graph is shown in Fig. 13.2.
κ/S m−1 c/mol dm−3 Λm /mS m2 mol−1 c 1/2 /(mol dm−3 )1/2
13.1 1.334 9.82 1.155
13.9 1.432 9.71 1.197
14.7 1.529 9.61 1.237
15.6 1.672 9.33 1.293
16.4 1.725 9.51 1.313
9.8
Λm /mS m2 mol−1
9.6
9.4
1.10 1.15 1.20 1.25 1.30

−3 1/2
c 1/2
/(mol dm )
Figure 13.2
The two points corresponding to the highest concentrations seem to be anomolous

and are ignored in finding the best-fit line, the equation of which is
Λm /mS m2 mol−1 = (−2.5262) × c 1/2 /(mol dm−3 )1/2 + 12.737
Therefore, K = 2.53 mS m2 (mol dm−1 )−3/2 and Λ○m = 12.7 mS m2 mol−1
P13B.5 (a) The molar conductivity Λm is given by [13B.4–518], Λm = κ/c, where

κ is conductivity and c is concentration. The Kohlrausch law, [13B.5–
518], gives the dependence of the molar conductivity on concentration
for strong electrolytes, Λm = Λ○m − Kc 1/2 . According to this law a plot of
Λm against c 1/2 will be a straight line with slope −K and y-intercept Λ○m .
Given that κ = C/R where C = 0.2063 cm−1 , the molar conductivity is
computed from Λm = C/cR. The plot is shown in Fig. 13.3.
R/Ω c/mol dm−3 Λm /mS m2 mol−1 c 1/2 /(mol dm−3 )1/2

3 314 0.000 50 12.45 0.022
1 669 0.001 0 12.36 0.032
342 0.005 0 12.06 0.071
174 0.010 11.85 0.100
89 0.020 11.58 0.141
37 0.050 11.11 0.224
12.5
Λm /mS m2 mol−1
12.0
11.5
11.0
0.00 0.05 0.10 0.15 0.20 0.25
−3 1/2
c 1/2
/(mol dm )
Figure 13.3
The data fall on a good straight line, as predicted by the Kohlrausch law,
and the equation for the best-fit line as
Λm /mS m2 mol−1 = (−6.6551) × c 1/2 /(mol dm−3 )1/2 + 12.5558
Thus K = 6.655 mS m2 (mol dm−1 )−3/2 and Λ○m = 12.56 mS m2 mol−1 .

(b) The law of independent migration of ions, [13B.6–519], allows the limit-
ing molar conductivity to be calculated from the values for the individ-
ual ions, and this is then converted to the molar conductivity using the
Kohlrausch law
Λ○m = ν+ λ(Na+ ) + ν− λ(I− )

= (1) × (5.01 mS m2 mol−1 ) + (1) × (7.68 mS m2 mol−1 )
= 12.69 mS m2 mol−1
Λm = Λ○m − Kc 1/2
= (12.69 mS m2 mol−1 )
− [6.655 mS m2 (mol dm−1 )−3/2 ] × (0.010 mol dm−3 )1/2
= 12.02 mS m2 mol−1
The conductivity is found using [13B.4–518], and the resistance using the
given cell constant
κ = cΛm = (0.010 mol dm−3 ) × (12.02 mS m2 mol−1 )

= (0.010 × 103 mol m−3 ) × (12.02 mS m2 mol−1 ) = 120 mS m−1
C 0.2063 × 102 m−1
R= = = 172 Ω
κ 120 × 10−3 S m−1
where 1S−1 = 1 Ω is used.

P13B.7 A spherical particle of radius a and charge ze travelling at a constant speed

through a solvent of viscosity η has mobility u given by [13B.9–520], u = ze/ f ,
where f is the frictional coefficient with Stokes’ law value f = 6πηa. Hence
ze (1) × (1.6022 × 10−19 C)

a= =
6πηu 6π × (0.93 × 10−3 kg m−1 s−1 ) × (1.1 × 10−8 m2 V−1 s−1 )
= 8.30... × 10−10 m = 0.83 nm
This is substantially larger than the 0.5 nm van der Waals radius of a Buck-
minsterfullerene (C60 ) molecule because the anion attracts a considerable hy-
dration shell through the London dispersion attraction to the nonpolar solvent
molecules and through the ion-induced dipole interaction. The Stokes radius
reflects the larger effective radius of the combined anion and its solvation shell.
P13B.9 (a) The initial concentration of AB is c AB . After a fraction α has dissociated

the concentration of AB is c = (1 − α)c AB and the concentration of A is
c A = αc AB , which is equal to the concentration of B, c B . Therefore, the
equilibrium constant K is given by
(c A /c −○ ) (c B /c −○ ) c A c B (αc AB )2 α 2 c AB
K= = = =
c/c −○ cc −○ (1 − α)c AB c −○ (1 − α)c −○
(b) As the solution becomes more dilute, the degree of dissociation increases
and, in the limit of infinite dilution, α = 1. This is a consequence of the
form of K derived in part (a): because K is constant, a decrease in c AB
requires an increase in α towards 1.
The concentration of ions in the solution scales directly with α, therefore
the conductivity, and hence the molar conductivity, will be proportional
to α: Λ m ∝ α. At infinite dilution the molar conductivity takes the value
Λ m,l and α = 1, therefore α = Λm /Λm,l .
(c) Substitution of this expression for α into the equilibrium expression gives
α 2 c AB Λm2
c AB
K= −
○
=
(1 − α)c Λm,l [1 − (Λm /Λm,l )]c −○
2
α2 Λm2 α 2 Λm,1
2
Λm
= 2 Λm
hence =
(1 − α) Λm,1 (1 − Λ m,l
) (1 − α)Λm (1 − ΛΛm )
m,l
Λm
(1 − α)Λm (1 − Λm,1 ) 1 1 1 1 (1 − α)Λm
= = − hence = +
α 2 Λm,1
2 Λm Λm Λm,l Λm Λm,l α 2 Λm,l
2
13C Diffusion
D13C.1 See the text following [13C.6–527].
E13C.1(a) The Einstein–Smoluchowski equation [13C.15–531], D = d 2 /2τ, relates the dif-
fusion coefficient D to the jump distance d and time τ required for a jump.
Approximating the jump length as the molecular diameter, then d ≈ 2a where
a is the effective molecular radius. This is estimated using the Stokes–Einstein
equation [13C.4b–526], D = kT/6πηa, to give 2a = 2kT/6πηD.
Combining these expressions and using the value for viscosity of benzene from
the Resource section gives
2 2
d2 1 2kT 1 kT
τ= = ( ) = ( )
2D 2D 6πηD 18D 3 πη
2
1 (1.3806 × 10−23 J K−1 ) × (298 K)
= ( )
−9 2 −1
18 × (2.13 × 10 m s ) 3
π × (0.601 × 10−3 kg m−1 s−1 )
= 2.73 × 10−11 s = 27.3 ps
E13C.2(a) The root mean square displacement in one dimension is given by [13C.13a–529],
⟨x 2 ⟩1/2 = (2Dt)1/2 , where D is the diffusion coefficient and t is the time period.
For an iodine molecule in benzene, D = 2.13 × 10−9 m2 s−1
⟨x 2 ⟩1/2 = (2Dt)1/2 = [2 × (2.13 × 10−9 m2 s−1 ) × (1.0 s)]1/2 = 6.5 × 10−5 m

= 65 µm
For a sucrose molecule in water, D = 0.5216 × 10−9 m2 s−1
⟨x 2 ⟩1/2 = [2 × (0.5216 × 10−9 m2 s−1 ) × (1.0 s)]1/2 = 3.2 × 10−5 m

= 32 µm
E13C.3(a) The root mean square displacement in three dimensions is given by [13C.13b–
530], ⟨r 2 ⟩1/2 = (6Dt)1/2 , where D is the diffusion coefficient and t is the time
period.
⟨r 2 ⟩ (5.0 × 10−3 m)2
t= = = 6.2 × 103 s
6D 6 × (6.73 × 10−10 m2 s−1 )
E13C.4(a) The diffusion in one dimension from a layer of solute is described by [13C.10–
528]
n0
e−x /4D t
2
c(x, t) =
A(πDt)1/2
where c(x, t) is the concentration at time t and distance x from the layer, and
n 0 is the amount in moles in the layer of area A placed at x = 0. If the mass of
sucrose is m, then n 0 = m/M, where M is the molar mass (342.30 g mol−1 ).
m
e−x /4D t
2
c(x, t) =
MA(πDt) 1/2
−2 −9 2 −1
(0.020 kg) × e−(10×10 m) /4×(5.216×10 m s )t
2
c(10 cm, t)=

(342.30 g mol−1 )×(5.0 × 10−4 m2 )×[π(5.216 × 10−9 m2 s−1 )t]1/2
= [(9.12... × 102 mol dm−3 ) × (t/ s)−1/2 ]e−4.79 ...×10 /(t/ s)
5
c(10 cm, 10 s) = (9.12... × 102 mol dm−3 ) × (10)−1/2 × e−4.79 ...×10 /(10)
5
= 0.00 mol dm−3

c(10 cm, 24 h) = (9.12... × 102 mol dm−3 )[24(3600)]−1/2 × e−4.79 ...×10 /[24(3600)]
5
= 0.0121 mol dm−3
Diffusion is a very slow process: after 10 s the concentration at a height of 10 cm

is zero to within the precision of the calculation. Even after 24 hours only a very
small amount of the sucrose has moved up into the liquid.
E13C.5(a) The thermodynamic force F is given by [13C.3b–524]
RT ∂c
F =− ( )
c ∂x T , p
Substituting c(x) = c 0 − αc 0 x into the above expression gives
RT αRT
F =− (−αc 0 ) =
c 0 − αc 0 x 1 − αx
The constant α is found by noting that c = c 0 /2 at x = 10 cm = 0.10 m. Hence

c 0 /2 = c 0 − αc 0 × (0.10 m) and therefore α = 5.0 m−1 . At T = 298 K and
x = 10 cm the force is
(5 m−1 ) × (8.3145 J K−1 mol−1 ) × (298 K)

F= = 25 kN mol−1
1 − (5 m−1 )(10 × 10−2 m)
A similar calculation at x = 15 cm gives F = 50 kN mol−1 . The force is greater

at the larger distance, even though the gradient is the same.
E13C.6(a) The thermodynamic force F is given by [13C.3b–524]
RT ∂c
F =− ( )
c ∂x T , p
Substituting c(x) = c 0 e−α x into the above expression gives

2
RT
(−2αc 0 xe−α x ) = 2αxRT
2
F =−
c 0 e−α x 2
The constant α is found by noting that c = c 0 /2 at x = 5 cm = 0.05 m.

Hence c 0 /2 = c 0 e−α(0.05 m) and therefore α = ln 2/(0.05 m)2 = 277 m−2 The
2
thermodynamic force at T = 293 K and x = 5.0 cm is
F = 2(277 m−2 )×(0.050 m)×(8.3145 J K−1 mol−1 )×(293 K) = 67.5 kN mol−1

E13C.7(a) The root mean square displacement in three dimensions is given by [13C.13b–
530], ⟨r 2 ⟩1/2 = (6Dt)1/2 , where D is the diffusion coefficient and t is the time
period. Hence,
⟨r 2 ⟩ (5.0 × 10−3 m)2

t= = = 1.3 × 103 s
6D 6 × (3.17 × 10−9 m2 s−1 )
E13C.8(a) The Stokes–Einstein equation [13C.4b–526], D = kT/6πηa, relates the diffu-

sion coefficient D to the viscosity η and the radius a of the diffusing particle,
which is modelled as a sphere. Recall that 1 cP = 10−3 kg m−1 s−1 .
kT (1.3806 × 10−23 J K−1 ) × (298 K)

a= = = 0.42 nm
6πηD 6π × (1.00 × 10−3 kg m−1 s−1 ) × (5.2 × 10−10 m2 s−1 )
P13C.1 Thermodynamic force, F, is given by [13C.3b–524].
RT ∂c
F =− ( )
c ∂x T , p
where c is the concentration.For a linear gradation of intensity, that is concen-
tration, down the tube
dc/dx = ∆c/∆x = [(0.050 − 0.100) × 103 mol m−3 ]/(0.10 m) = 500 mol m−4
RT dc (8.3145 J K−1 mol−1 ) × (298 K)

F =− =− × (−500 mol m−4 )
c dx c
1.23... × 106 N mol−1
=
(c/ mol m−3 )
At the left face, c = 0.100 mol dm−3 :
F = (1.23... × 103 kN mol−1 )/(0.100 × 103 ) = 12.4 kN mol−1

−1
= 2.1 × 10−20 N (molecule)
In the middle, c = 0.075 mol dm−3 :
F = (1.23... × 103 kN mol−1 )/(0.075 × 103 ) = 16.5 kN mol−1

−1
= 2.7 × 10−20 N (molecule)
Close to the left face, c = 0.050 mol dm−3 :
F = (1.23... × 103 kN mol−1 )/(0.050 × 103 ) = 24.8 kN mol−1

−1
= 4.1 × 10−20 N (molecule)
P13C.3 The thermodynamic force F is given by [13C.3b–524]

RT ∂c
F =− ( )
c ∂x T , p
Substituting c(x) = c 0 (1 − e−ax ) into the above expression gives

2
2axRTe−ax
2
RT −ax 2 2axRT
F =− (2ac 0 xe ) = − =
c 0 (1 − e−ax )
2
(1 − e−ax 2 ) (1 − e ax 2 )
The final step involves multiplying top and bottom of the fraction by e ax . For
2
thermodynamic force for a = 0.10 cm−2 = 1000 m−2 and T = 298 K is

(5.0 MN mol−1 ) × (x/m)
F=
(1 − e1000×(x/m)2 )
(8.2 × 10−18 N molecule−1 ) × (x/m)
=
(1 − e1000×(x/m)2 )
2 000
F/kN mol−1
−2 000
−2.0 −1.0 0.0 1.0 2.0

x/cm
Figure 13.4
A plot of the thermodynamic force per mole against x is shown in Fig. 13.4. It
demonstrates that the force is directed such that mass is pushed by the ther-
modynamic force toward the centre of the tube to where the concentration is
lowest. A negative force pushes mass toward the left (x > 0) and a positive
force pushes mass toward the right (x < 0).
At x = 0 the gradient of the concentration is zero, so the thermodynamic force
is also zero. However, as x approaches zero the modulus of the thermodynamic
force increases without limit on account of the concentration becoming smaller
and smaller.
P13C.5 The generalised diffusion equation is [13C.6–527], where c is concentration, t
is time, D is the diffusion coefficient and x is displacement.
∂c ∂2 c
=D 2
∂t ∂x
An expression for c(x, t) is a solution of the diffusion equation if substitution

of the expression for c(x, t) into each side of the diffusion equation gives the
same result. The proposed solution is
n0
e−x /4D t
2
c(x, t) =
A(πDt) 1/2
LHS
∂c n 0 e−x /4D t
2
x2 1
= ( − )
∂t A(πDt) 1/2 4Dt 2 2t
RHS
∂2 c ∂ ⎡ n e−x 2 /4D t −x ⎤ n 0 e−x /4D t

2
x 2 1
⎢ 0 ⎥
D 2 =D ⎢ ( )⎥ = D [( ) − ]
∂x ∂x ⎢ A(πDt)1/2 2Dt ⎥
⎣ ⎦ A(πDt) 1/2 2Dt 2Dt
n 0 e−x /4D t
2
x2 1
= ( − )
A(πDt)1/2 4Dt 2 2t
As required the LHS = the RHS, hence the proposed form of c(x, t) is indeed
a solution to the diffusion equation.
As t → 0 the exponential term e−x /4D t falls off more and more rapidly, imply-
2
ing that in the limit t = 0 all the material is at x = 0. The exponential function
dominates the term t 1/2 in the denominator.
P13C.7 As discussed in Section 13C.2(c) on page 528 the probability of finding a molecule
in an interval dx at distance x from the origin at time t is P(x, t)dx, where
P(x, t) is given by
1
e−x /4D t
2
P(x, t) = 1/2
(πDt)
The mean value of x 4 is found by integrating P(x, t)x 4 dx over the full range
of x, which in this case is 0 to ∞
∞ 1 ∞
x 4 e−x /4D t dx
2
⟨x 4 ⟩ = ∫ x 4 P(x) dx = 1/2 ∫
0 (πDt) 0
1
= × 3 (4Dt)2 × (4πDt)1/2 = 12D 2 t 2
(πDt)1/2 8
where to go to the final line Integral G.5 is used with k = 1/4Dt. Hence,
⟨x 4 ⟩1/4 = (12D 2 t 2 )1/4 .
A similar calculation is used to find ⟨x 2 ⟩
∞ 1 ∞
x 2 e−x /4D t dx
2
⟨x 2 ⟩ = ∫ x 2 P(x) dx = 1/2 ∫
0 (πDt) 0
1
= × 1 π 1/2 (4Dt)3/2 = 2Dt
(πDt)1/2 4
where to go to the final line Integral G.3 is used with k = 1/4Dt. Hence,
⟨x 2 ⟩1/2 = (2Dt)1/2 .
The ratio of the ⟨x 4 ⟩1/4 to ⟨x 2 ⟩ is
⟨x 4 ⟩1/4 (12D 2 t 2 )1/4 12 1/4

= = ( ) = 31/4
⟨x 2 ⟩1/2 (2Dt)1/2 4
The result is independent of the time.
P13C.9 The probability of being n steps from the origin is P(nd) = N!/(N−N R )!N R !2 N
where N R is the number of steps taken to the right and N is the total number
of steps. Note that n = N R − N L and N = N R + N L , where N L is the number of
steps taken to the left.
N−n
NR = N − NL = NL + n hence NL =
2
N+n
NL = N − NR = NR − n hence NR =
2
N! N!
therefore P(nd) = = N+n
[N − ( N−n
2
)]! ( N−n
2
)! 2 N ( 2
)! ( N−n
2
)! 2 N
The probability of being six paces away from the origin (x = 6d) is
N!
Pexact (6d) =
( N+6
2
)! ( N−6
2
)! 2 N
This is the ‘exact’ value of the probability according to the random walk model.
In the limit of large N the probability density of being at distance x and time t
is given by [13C.14–531]
2τ 1/2 −x 2 τ/2td 2
P(x, t) = ( ) e
πt
For the present case the value of x is taken as nd, and t/τ is taken as N because
the time to take N steps is N τ. With these substitutions
2 1/2 −n 2 /2N
Plim (nd, N) = ( ) e
πN
The following table compares the exact values of the probability with those
predicted for large N. The discrepancy between the two values falls to less than
0.1% when N is greater than about 53.
N Pexact Plim 100(Pexact − Plim )/Pexact

6 0.015 6 0.016 2 −3.79
10 0.043 9 0.041 7 5.09
14 0.061 1 0.059 0 3.51
18 0.070 8 0.069 2 2.30
22 0.076 2 0.075 1 1.55
26 0.079 2 0.078 3 1.07
30 0.080 6 0.079 9 0.75
34 0.081 0 0.080 6 0.54
38 0.080 9 0.080 6 0.38
42 0.080 4 0.080 2 0.27
46 0.079 7 0.079 5 0.19
50 0.078 8 0.078 7 0.13
54 0.077 9 0.077 8 0.08
58 0.076 8 0.076 8 0.05
60 0.076 3 0.076 3 0.03
P13C.11 The Stokes–Einstein relation [13C.4b–526], shows that D ∝ T/η where D is

the diffusion coefficient and η is the viscosity. The temperature dependence of
viscosity is given by [13B.2–517], η = η 0 eE a /RT , it therefore follows that D ∝
Te−E a /RT . The activation energy E a can therefore be determined from the ratio
of the diffusion constants at two temperatures
D T1 T1 e−E a /RT1 T1 ERa (1/T2 −1/T1 )

= = e
D T2 T2 e−E a /RT2 T2
Solving for E a gives
R D T T2
Ea = ln 1
1/T2 − 1/T1 D T2 T1
(8.3145 J K−1 mol−1 ) (298 K) × (2.05 × 10−9 m2 s−1 )
= × ln ( )
1/(298 K) − 1/(273 K) (273 K) × (2.89 × 10−9 m2 s−1 )
= 6.9 kJ mol−1
The activation energy associated with diffusion of therefore 6.9 kJ mol−1 .

I13.1 (a) The diffusion equation is [13C.6–527].
∂c ∂2 c
=D 2
∂t ∂x
The proposed solution is
c(x, t) = c 0 + (c s − c 0 )[1 − erf(ξ)]

ξ
where ξ(x, t) = x/(4Dt)1/2 and erf(ξ) = 2π−1/2 ∫0 e−y dy. Note that, by
2
definition, ∂[erf(ξ)]/∂ξ = 2π−1/2 e−ξ .

2
The first step is to examine the initial and boundary conditions. At t = 0,

∞
ξ(x, 0) = ∞. Hence erf(ξ) = 2π−1/2 ∫0 e−y dy = (2π−1/2 )×( 21 π 1/2 ) = 1
2
and c(x, 0) = c 0 + (c s − c 0 )[1 − 1] = c 0 for 0 < x < ∞. At x = 0,

ξ(0, t) = 0. Hence erf(ξ) = 2π−1/2 ∫0 e−y dy = (2π−1/2 ) × 0 = 0 and
0 2
c(0, t) = c 0 +(c s −c 0 )[1−0] = c s for 0 ≤ t ≤ ∞. Therefore, this expression

for c(x, t) satisfies the initial and boundary conditions.
To determine whether or not the proposed solution solves the diffusion
equation it is substituted into the each side. For the LHS
∂c ∂(c 0 + (c s − c 0 )[1 − erf(ξ)]) ∂[erf(ξ)] ∂ξ

= = −(c s − c 0 )
∂t ∂t ∂ξ ∂t
2 ∂[x/(4Dt)
1/2
]
= −(c s − c 0 )(2π−1/2 e−ξ )
∂t
x x(c s − c 0 ) x 2 /4D t
= −(c s − c 0 )(2π−1/2 e−x /4D t ) [
2
]= e
(4D) t
1/2 3/2 (4πD)1/2 t 3/2
For the RHS

∂2 c ∂ ∂[erf(ξ)] ∂ξ
D =D [−(c s − c 0 ) ]
∂x 2 ∂x ∂ξ ∂x
2D(c s − c 0 ) ∂[e−x /4D t ]
2
2D(c s − c 0 ) −x −x 2 /4D t
=− =− ( e )
(4πDt)1/2 ∂x (4πDt)1/2 2Dt
x(c s − c 0 )
ex /4D t
2
=
(4πD)1/2 (t)3/2
Therefore the LHS is equal to the RHS and the proposed form of c(x, t)
does indeed satisfy the diffusion equation, as well as the initial and bound-
ary conditions.
(b) Diffusion through aveoli sites (about 1 cell thick) of oxygen and carbon
dioxide between lungs and blood capillaries (also about 1 cell thick) oc-
curs through about 75 µm (the diameter of a red blood cell). Thus, the
range 0 ≤ x ≤ 100 µm is reasonable for concentration profiles for the
diffusion of oxygen into water. Given the maximum distance, the longest
time is estimated using [13C.12–529].
2
πxmax π(0.1 × 10−3 m)2
tmax ≈ = = 3.74 s
4D 4(2.10 × 10−9 m2 s−1 )
The plots shown in Fig. 13.5 are with c 0 = 0, c s = 2.9 × 10−4 mol dm−3 ,
and D = 2.10 × 10−9 m2 s−1 .
3.0
t = 0.05 s
t = 0.25 s
104 c/(mol dm−3 )
t = 0.75 s
2.0 t=2s
1.0
0.0
0.00 0.02 0.04 0.06 0.08 0.10
x/mm
Figure 13.5
14 Chemical kinetics
14A The rates of chemical reactions

D14A.1 A reaction order for a particular species can only be ascribed when the rate
is simply proportional to a power of the concentration of that species. For
example, if the rate law is of the form υ = k r [A] a [B]b . . . an order is ascribable
to both A and B, but if the rate law is of the form υ = (k 1 [A])/(k 2 + k 3 [B]), an
order is ascribable to A, but not to B.
D14A.3 This is discussed in Section 14A.1 on page 539.
E14A.1(a) For a homogeneous reaction in a constant volume system the rate of reaction
is given by [14A.3b–542], υ = (1/ν J )d[J]/dt, which is rearranged to d[J]/dt =
ν J υ. In these expressions ν J is the stoichiometric number of species J, which
is negative for reactants and positive for products. For this reaction ν A = −1,
ν B = −2, ν C = +3 and ν D = +1.
For A d[A]/dt = ν A υ = (−1) × (2.7 mol dm−3 s−1 ) = −2.7 mol dm−3 s−1
For B d[B]/dt = ν B υ = (−2) × (2.7 mol dm−3 s−1 ) = −5.4 mol dm−3 s−1
For C d[C]/dt = ν C υ = (+3) × (2.7 mol dm−3 s−1 ) = +8.1 mol dm−3 s−1
For D d[D]/dt = ν D υ = (+1) × (2.7 mol dm−3 s−1 ) = +2.7 mol dm−3 s−1
The rate of consumption of A is 2.7 mol dm−3 s−1 , the rate of consumption of
B is 5.4 mol dm−3 s−1 , the rate of formation of C is 8.1 mol dm−3 s−1 , and the
rate of formation of D is 2.7 mol dm−3 s−1 .
E14A.2(a) For a homogeneous reaction in a constant volume system the rate of reaction
is given by [14A.3b–542], υ = (1/ν J )d[J]/dt, where ν J is the stoichiometric
number of species J which is negative for reactants and positive for products.
For species C, which has ν C = +2, this gives
1 d[C] 1
υ= = × (2.7 mol dm−3 s−1 ) = 1.35... mol dm−3 s−1
ν C dt +2
= 1.4 mol dm−3 s−1
510 14 CHEMICAL KINETICS
Rearranging [14A.3b–542] then gives
For A d[A]/dt = ν A υ = (−2) × (1.35... mol dm−3 s−1 ) = −2.70... mol dm−3 s−1
For B d[B]/dt = ν B υ = (−1) × (1.35... mol dm−3 s−1 ) = −1.35... mol dm−3 s−1
For D d[D]/dt = ν D υ = (+3) × (1.35... mol dm−3 s−1 ) = +4.05... mol dm−3 s−1
The rate of consumption of A is 2.7 mol dm−3 s−1 , the rate of consumption of
B is 1.4 mol dm−3 s−1 , and the rate of formation of D is 4.1 mol dm−3 s−1 .
E14A.3(a) As explained in Section 14A.2(b) on page 542 the units of k r are always such as
to convert the product of concentrations, each raised to the appropriate power,
into a rate expressed as a change in concentration divided by time. In this
case the rate is given in mol dm−3 s−1 , so if the concentrations are expressed
in mol dm−3 the units of k r will be dm3 mol−1 s−1 because
kr [A] [B]
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
(dm3 mol−1 s−1 ) × (mol dm−3 ) × (mol dm−3 ) = mol dm−3 s−1
The rate of reaction is given by [14A.3b–542], υ = (1/ν J )(d[J]/dt), where ν J

is the stoichiometric number of species J. Rearranging gives d[J]/dt = ν J υ. In
this case ν C = +3, ν A = −1, and υ = k r [A][B] so
d[C] d[A]
= ν C υ = 3k r [A][B] = ν A υ = −k r [A][B]
dt dt
The rate of formation of C is therefore d[C]/dt = 3k r [A][B] and the rate of

consumption of A is −d[A]/dt = k r [A][B] .
E14A.4(a) The rate of reaction is given by [14A.3b–542], υ = (1/ν J )(d[J]/dt). In this case
ν C = +2 so
1 d[C] 1
υ= = k r [A][B][C] = 1
k [A][B][C]
2 r
ν J dt +2
As explained in Section 14A.2(b) on page 542 the units of k r are always such as
into a rate expressed as a change in concentration divided by time. In this
case the rate is given in mol dm−3 s−1 , so if the concentrations are expressed
in mol dm−3 the units of k r will be dm6 mol−2 s−1 because
kr [A] [B] [C]

³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
(dm6 mol−2 s−1 ) × (mol dm−3 ) × (mol dm−3 ) × (mol dm−3 ) = mol dm−3 s−1
E14A.5(a) As explained in Section 14A.2(b) on page 542 the units of k r are always such as
into a rate expressed as a change in concentration divided by time.
(i) A second-order reaction expressed with concentrations in moles per cu-

bic decimetre is one with a rate law such as υ = k r [A][B]. If the rate is
given in mol dm−3 s−1 then the units of k r will be dm3 mol−1 s−1 because
kr [A] [B]
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
(dm3 mol−1 s−1 ) × (mol dm−3 ) × (mol dm−3 ) = mol dm−3 s−1
A third-order reaction expressed with concentrations in moles per cubic
decimetre is one with a rate law such as v = k r [A][B][C]. The units of k r
will then be dm6 mol−2 s−1 because
kr [A] [B] [C]
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
(dm6 mol−2 s−1 )×(mol dm−3 )×(mol dm−3 )×(mol dm−3 ) = mol dm−3 s−1
(ii) If the rate laws are expressed with pressures in kilopascals then a second-
order reaction is one with a rate law such as υ = k r p A p B and a third-order
reaction is one with a rate law such as υ = k r p A p B p C . If the rate is given in
kPa s−1 then the units of k r will be kPa−1 s−1 and kPa−2 s−1 respectively.
kr pA pB
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
For second-order (kPa−1 s−1 ) × (kPa) × (kPa) = kPa s−1
kr pA pB pC
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
For third-order (kPa−2 s−1 ) × (kPa) × (kPa) × (kPa) = kPa s−1
E14A.6(a) (i) In the rate law υ = k r1 [A][B]/(k r2 + k r3 [B]1/2 ) the concentration of A

appears raised to the power +1, so the reaction is first order in A, and
hence can be assigned an order with respect to A, under all conditions .
(ii) The concentration of B does not appear as a single term raised to a power,
so the reaction has an indefinite order with respect to B. However, if k r2 ≫
k r3 [B]1/2 , which might occur at very low concentrations of B, then the
term k r3 [B]1/2 in the denominator is negligible compared to the k r2 term
and so the rate law becomes
υ = k r1 [A][B]/k r2 = k r,eff [A][B] where k r,eff = k r1 /k r2
In this effective rate law the concentration of B appears raised to the power
+1, so under these conditions the reactions is first order in B. Similarly, if
k r2 ≪ k r3 [B]1/2 , which might occur at very high concentrations of B, the
term k r2 in the denominator is negligible compared to the k r3 [B]1/2 term
and so the rate law becomes
υ = k r1 [A][B]/k r3 [B]1/2 = k r,eff [A][B]1/2 where k r,eff = k r1 /k r3
In this effective rate law the order with respect to B is + 21 .
To summarize, an order can only be assigned with respect to B if either
k r2 ≫ k r3 [B]1/2 , in which case the order is +1, or k r2 ≪ k r3 [B]1/2 , in
which case the order is + 12 .
(iii) An overall order can be assigned only if all of the individual orders can
be assigned. Consequently the reaction can only be assigned an overall
order if k r2 ≫ k r3 [B]1/2 or k r2 ≪ k r3 [B]1/2 . The overall order in these
two cases is +2 and + 32 .
E14A.7(a) The gaseous species is denoted A and the order with respect to A as a. The rate
law expressed in terms of partial pressure is then υ = k r p Aa . Taking (common)
logarithms gives
log υ = log k r + log p Aa = log k r + a log p A
where the properties of logarithms log(x y) = log x + log y and log x a = a log x
are used.
This expression implies that a graph of log υ against log p A will be a straight line
of slope a, from which the order can be determined. However, because there
are only two data points a graph is not necessary so an alternative approach is
used.
If the initial partial pressure of the compound is p A,0 then the partial pressure
when a fraction f has reacted, so that a fraction 1 − f remains, is (1 − f )p A,0 .
Data are given for two points, f 1 = 0.100 and f 2 = 0.200. Denoting the rates
at these points by υ 1 and υ 2 and using the expression log υ = log k r + a log p A
from above gives the equations
log υ 1 = log k r + a log [(1 − f 1 )p A,0 ] log υ 2 = log k r + a log [(1 − f 2 )p A,0 ]
Subtracting the second equation from the first gives
log υ 1 − log υ 2 = a log [(1 − f 1 )p A,0 ] − a log [(1 − f 2 )p A,0 ]
Hence
υ1 (1 − f 1 )p A,0 1 − f1
log ( ) = a log ( ) = a log ( )
υ2 (1 − f 2 )p A,0 1 − f2
where the property of logarithms log x −log y = log(x/y) is used and the factor
of p A,0 is cancelled.
Rearranging for a gives
log(υ 1 /υ 2 ) log [(9.71 Pa s−1 )/(7.67 Pa s−1 )]

a= = = 2.00
log [(1 − f 1 )/(1 − f 2 )] log [(1 − 0.100)/(1 − 0.200)]
E14A.8(a) Assuming perfect gas behaviour the total pressure is proportional to the total
amount in moles of gas present, provided that the temperature is constant and
the volume of the container is fixed. The reaction 2 ICl(g) + H2 (g) → I2 (g) +
2 HCl(g) involves the same number of gas molecules on both sides of the re-
action arrow and therefore the total amount in moles of gas present does not
change as the reaction proceeds. Consequently there is no change in the total
pressure during the reaction. This means that the composition of the reaction
mixture cannot be determined by measuring the total pressure in this case.
E14A.9(a) The stoichiometry of the reaction shows that one mole of Br2 is formed for
every two moles of NO formed. Therefore the rate of formation of Br2 is half
the rate of formation of NO.
d[Br2 ] d[NO]
= 1
2
× = 1
2
× (0.24 mmol dm−3 s−1 ) = 0.12 mmol dm−3 s−1
dt dt
P14A.1 The rate law is assumed to take the form υ 0 = k r [C6 H12 O6 ] a where υ 0 is
the initial rate and a is the order with respect to glucose. Taking (common)
logarithms gives
log υ 0 = log k r + log[C6 H12 O6 ] a = log k r + a log[C6 H12 O6 ]
where the properties of logarithms log(x y) = log x + log y and log x a = a log x
are used.
This expression implies that a graph of log υ 0 against log[C6 H12 O6 ] will be a
straight line of slope a and intercept log k r . The data are plotted in Fig. 14.1.
[C6 H12 O6 ] υ 0 log([C6 H12 O6 ] log(υ 0

/mol dm−3 /mol dm−3 s−1 /mol dm−3 ) /mol dm−3 s−1 )
1.00 × 10−3 5.0 −3.000 0.699
1.54 × 10−3 7.6 −2.812 0.881
−3
3.12 × 10 15.5 −2.506 1.190
−3
4.02 × 10 20.0 −2.396 1.301
1.4
log(υ 0 /mol dm−3 s−1 )
1.2
1.0
0.8
0.6
−3.0 −2.8 −2.6 −2.4
−3
log([C6 H12 O6 ]/mol dm )
Figure 14.1
The data fall on a good straight line, the equation for which is
log(υ 0 /mol dm−3 s−1 ) = 1.00 × log([CH6 H12 O6 ]/mol dm−3 ) + 3.692
(a) Identifying the order a with the slope gives a = 1.00; that is, the reaction
is first order in glucose.
(b) The intercept at log([C6 H12 O6 ]/mol dm−3 ) = 0 is log(υ 0 /mol dm−3 s−1 ) =
3.692, which corresponds to υ 0 = 4.92×103 mol dm−3 s−1 when [C6 H12 O6 ]
= 1 mol dm−3 . Because a = 1, the rate law is υ 0 = k r [C6 H12 O6 ]1 , which
is rearranged to give
υ0 4.92 × 103 mol dm−3 s−1

kr = = = 4.92 × 103 s−1
[C6 H12 O6 ] 1 mol dm−3
P14A.3 (a) Experiments 1 and 2 both have the same initial H2 concentration, but
experiment 2 has an ICl concentration twice that of experiment 1. Because
the rate of experiment 2 is also twice that of experiment 1, it follows that
the rate is proportional to [ICl] and hence that the reaction is first order
in ICl.
Experiments 2 and 3 both have the same initial ICl concentration, but
experiment 3 has an H2 concentration three times that of experiment
2. Because the rate of experiment 3 is approximately three times that of
experiment 2, it follows that the rate is proportional to [H2 ] and hence
that the reaction is first order in H2 .
Therefore the rate law is υ = k r [ICl][H2 ] .
(b) The rate law υ = k r [ICl][H2 ] implies that a graph of υ 0 against [ICl][H2 ]
should be a straight line of slope k r and intercept zero. The data are plotted
in Fig. 14.2.
[ICl]0 [H2 ]0 [ICl]0 [H2 ]0 υ0

Expt.
/mol dm−3 /mol dm−3 /mol2 dm−6 /mol dm−3 s−1
1 1.5 × 10−3 1.5 × 10−3 2.25 × 10−6 3.7 × 10−7
2 3.0 × 10−3 1.5 × 10−3 4.50 × 10−6 7.4 × 10−7
3 3.0 × 10−3 4.5 × 10−3 1.35 × 10−5 2.2 × 10−6
υ 0 /(10−7 mol dm−3 s−1 )
20
10
0
0 2 4 6 8 10 12 14
−6 2 −6
[ICl]0 [H2 ]0 /(10 mol dm )
Figure 14.2
υ 0 /mol dm−3 s−1 = 0.163 × {[ICl]0 [H2 ]0 /(mol2 dm−6 )} + 6.19 × 10−9
Identifying the slope with k r gives k r = 0.16 dm3 mol−1 s−1 .

(c) The initial reaction rate for experiment 4 is predicted from the rate law
υ 0 = k r [ICl][H2 ]
= (0.163... dm3 mol−1 s−1 ) × (4.7 × 10−3 mol dm−3 )
× (2.7 × 10−3 mol dm−3 ) = 2.1 × 10−6 mol dm−3 s−1
14B Integrated rate laws

D14B.1 In general the rate of a reaction depends on the concentration of the various
species involved. If all but one of the species is placed in large excess, such
that their concentrations do not vary with time, the rate law may reduce to the
simple form υ = k r′ [A] a , where A is the species not in excess. If this is the case,
the power a is the order with respect to A: if a = 1 the reaction is said to be
pseudofirst order with respect to A, if a = 2 it is pseudosecond order. It is also
possible that under certain conditions a more complex rate law will simplify to
pseudofirst or pseudosecond order. For example if the rate law is of the form
υ = (k 1 [A]2 )/(k 2 +k 3 [A]), when k 3 [A] ≫ k 2 the rate law becomes pseudofirst
order in A, but when k 3 [A] ≪ k 2 , the rate law is pseudosecond order in A.
D14B.3 The determination of a rate law is simplified by the isolation method in which
the concentrations of all the reactants except one are in large excess. If B is
in large excess, for example, then to a good approximation its concentration
is constant throughout the reaction. Although the true rate law might be υ =
k r [A][B], we can approximate [B] by [B]0 and write
υ = k r′ [A] where k r′ = k r [B]0
which has the form of a first-order rate law. Because the true rate law has been
forced into first-order form by assuming that the concentration of B is constant,
it is called a pseudofirst-order rate law. The dependence of the rate on the
concentration of each of the reactants may be found by isolating them in turn
(by having all the other substances present in large excess), and so constructing
the overall rate law.
In the method of initial rates, which is often used in conjunction with the
isolation method, the rate is measured at the beginning of the reaction for
several different initial concentrations of reactants. Suppose that the rate law
for a reaction with A isolated is υ = k r [A] a ; then its initial rate, υ 0 , is given by
the initial values of the concentration of A and is written υ 0 = k r [A]0a . Taking
logarithms gives
log υ 0 = log k r + a log[A]0
For a series of initial concentrations, a plot of the logarithms of the initial rates
against the logarithms of the initial concentrations of A should be a straight
lime with slope a.
The method of initial rates might not reveal the full rate law, for the prod-
ucts may participate in the reaction and affect the rate. For example, products
participate in the synthesis of HBr, where the full rate law depends on the
concentration of HBr. To avoid this difficulty, the rate law should be fitted to the
data throughout the reaction. The fitting may be done, in simple cases at least,
by using a proposed rate law to predict the concentration of any component at
any time, and comparing it with the data.
Because rate laws are differential equations, they must be integrated in order to
find the concentrations as a function of time. Even the most complex rate laws
may be integrated numerically. However, in a number of simple cases analytical
solutions are easily obtained and prove to be very useful. These are summarized
in Table 14B.3 on page 551. Experimental data can be tested against an assumed
rate law by manipulating the integrated rate law into a form which will give a
straight line plot. If the data do indeed fall on a good straight line, then the data
are consistent with the assumed rate law.
E14B.1(a) Using [14A.3b–542], υ = (1/ν J )(d[J]dt), the rate of the reaction 2A → P is
υ = − 12 (d[A]/dt). Combining this with the rate law υ = k r [A]2 gives
1 d[A] d[A]
− = k r [A]2 hence = −2k r [A]2
2 dt dt
This is essentially the same as [14B.4a–549], d[A]/dt = −k r [A]2 except with
k r replaced by 2k r . The integrated rate law is therefore essentially the same as
that for [14B.4a–549], that is, [14B.4b–549] 1/[A] − 1/[A]0 = k r t, except with
k r replaced by 2k r . Hence for the reaction in question
1 1
− = 2k r t
[A] [A]0
Rearranging for t gives
1 1 1
t= ( − )
2k r [A] [A]0
1 1 1
= −1 −1 × ( −3 − )
−4 3
2 × (4.30 × 10 dm mol s ) 0.010 mol dm 0.210 mol dm−3
= 1.1 × 105 s or 1.3 days
E14B.2(a) The integrated rate law for a second-order reaction of the form A + B → P is
given by [14B.7b–550],
[B]/[B]0
ln = ([B]0 − [A]0 ) k r t
[A]/[A]0
(i) In 1 hour the concentration of B falls from 0.060 mol dm−3 to 0.030 mol dm−3 ,
so the change in the concentration of B in this time period is −0.030 mol dm−3 .
It follows from the reaction stoichiometry that the concentration of A
must fall by the same amount, so the concentration of A after 1 hour is
[A]0
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
[A] = (0.080 mol dm−3 ) −(0.030 mol dm−3 ) = 0.050 mol dm−3
The rate constant is then found by rearranging the integrated rate equa-
tion for k r and using the values of [A] and [B] at 1 hour, which corre-
sponds to 1 h × (602 s h−1 ) = 3600 s.
1 [B]/[B]0
kr = ln
([B]0 − [A]0 ) t [A]/[A]0
1
=
[(0.060 mol dm ) − (0.080 mol dm−3 )] × (3600 s)
−3
−3 −3
⎛ (0.030 mol dm ) / (0.060 mol dm ) ⎞
× ln
⎝ (0.050 mol dm−3 ) / (0.080 mol dm−3 ) ⎠
= 3.09... × 10−3 dm3 mol−1 s−1 = 3.1 × 10−3 dm3 mol−1 s−1
(ii) The half-life of a particular reactant is the time taken for the concentration
of that reactant to fall to half its initial value. The half-life of B is 1 hour
because it is given in the question that after 1 hour the concentration of
B had fallen from 0.060 mol dm−3 to 0.030 mol dm−3 , half the original
value.
The initial concentration of A is 0.080 mol dm−3 so the half-life is the
time at which the concentration of A has dropped by 0.040 mol dm−3 to
0.040 mol dm−3 . It follows from the stoichiometry of the reaction that the
concentration of B must also fall by 0.040 mol dm−3 during this period,
so the concentration of B will be
[B] = 0.060 mol dm−3 − 0.040 mol dm−3 = 0.020 mol dm−3
Rearranging the integrated rate equation then gives
1 1 [B]/[B]0
t= ln ( )
k r ([B]0 − [A]0 ) [A]/[A]0
1 1
= ×
3.09... × 10−3 dm3 mol−1 s−1 (0.060 mol dm−3 ) − (0.080 mol dm−3 )
−3 −3
⎛ (0.020 mol dm ) / (0.060 mol dm ) ⎞
× ln
⎝ (0.040 mol dm−3 ) / (0.080 mol dm−3 ) ⎠
= 6.5 × 103 s or 1.8 hours
E14B.3(a) (i) The integrated rate law for a zeroth-order reaction is given by [14B.1–547],
[A] = [A]0 − k r t where in this case A is NH3 . If concentrations are
expressed in terms of partial pressures, this becomes p NH3 = p NH3 ,0 − k r t.

Rearranging for k r and using p NH3 = 10 kPa when t = 770s with p NH3 ,0 =
21 kPa gives
p NH3 ,0 − p NH3 (21 × 103 Pa) − (10 × 103 Pa)

kr = = = 14.2... Pa s−1
t 770 s
= 14 Pa s−1
(ii) When all the ammonia has been consumed, p NH3 = 0. Rearranging the
rate law for t gives
p NH3 ,0 − p NH3 (21 × 103 Pa) − 0

t= = = 1.5 × 103 s
kr 14.2... Pa s−1
E14B.4(a) The fact that the two half-lives are not the same establishes that the reaction is
not first-order because, as explained in Section 14B.2 on page 547, a first-order
reaction has a constant half-life. For orders n ≠ 1 the half-life is given by [14B.6–
550], t 1/2 = (2n−1 − 1)/[(n − 1)k r [A]0n−1 ]. Denoting the two measurements by
t 1/2,i and t 1/2,ii and expressing concentration in terms of partial pressure gives
the two equations
2n−1 − 1 2n−1 − 1
t 1/2,i = n−1
t 1/2,ii = n−1
(n − 1)k r p A,i (n − 1)k r p A,ii
The second equation is divided by the first to give

n−1
t 1/2,ii p A,i t 1/2,ii p A,i
=( ) hence log ( ) = (n − 1) log ( )
t 1/2,i p A,ii t 1/2,i p A,ii
where log x a = a log x is used. Rearranging for n gives
log (t 1/2,ii /t 1/2,i ) log [(880 s)/(410 s)]

n= +1= + 1 = 2.00
log (p A,i /p A,ii ) log [(363 Torr)/(169 Torr)]
Therefore the reaction is second-order .
E14B.5(a) For the reaction 2N2 O5 (g) → 4NO2 (g) + O2 (g) the rate, as given by [14A.3b–
542], υ = (1/ν J )(d[J]/dt), is
1 dp N2 O5
υ=
−2 dt
where concentrations are expressed in terms of partial pressures. It is given that
the reaction is first-order in N2 O5 , so υ = k r p N2 O5 . Combining this with the
above expression for υ gives
1 dp N2 O5 dp N2 O5
= k r p N2 O5 hence = −2k r p N2 O5
−2 dt dt
This has the same form, except with 2k r instead of k r , as [14B.2a–547], (d[A]/dt) =
−k r [A], for which it is shown in Section 14B.2 on page 547 that the half-life and
the integrated rate law are
ln 2
t 1/2 = [A] = [A]0 e−k r t
kr
The expressions for the reaction in question are analogous, but with k r replaced
by 2k r .
ln 2
t 1/2 = p N2 O5 = (p N2 O5 ,0 )e−2k r t
2k r
The half-life is
ln 2 ln 2
t 1/2 = = = 1.03 × 104 s
2k r 2 × 3.38 × 10−5 s−1
The partial pressures at the specified times are calculated from the above inte-
grated form of the rate law. Hence
−5
s−1 )×(50 s)
t = 50 s p N2 O5 = (500 Torr) × e−2×(3.38×10 = 489 Torr
−5 −1
t = 20 min p N2 O5 = (500 Torr) × e−2×(3.38×10 s )×([20×60] s)
= 461 Torr
E14B.6(a) The reaction is of the form A + B → products. Assuming that it has rate law
υ = k r [A][B], the integrated rate law is given by [14B.7b–550]
[B]/[B]0
ln = ([B]0 − [A]0 )k r t
[A]/[A]0
Suppose that after time t the concentration of A has fallen by an amount x so

that [A] = [A]0 − x. Because of the stoichiometry of the reaction the concen-
tration of B must fall by the same amount, so [B] = [B]0 − x. Therefore
([B]0 − x)/[B]0
ln = ([B]0 − [A]0 )k r t
([A]0 − x)/[A]0
Hence
([B]0 − x)[A]0
= e([B]0 −[A]0 )k r t
([A]0 − x)[B]0
Rearranging gives
[B]0 [A]0 − x[A]0 = [B]0 [A]0 e([B]0 −[A]0 )k r t − x[B]0 e([B]0 −[A]0 )k r t
Hence
[B]0 [A]0 (e([B]0 −[A]0 )k r t − 1) [B]0 [A]0 (e λ − 1)
x= =
[B]0 e([B]0 −[A]0 )k r t − [A]0 [B]0 e λ − [A]0
where λ = ([B]0 − [A]0 )k r t. Taking A and B as OH− and CH3 COOC2 H5
respectively, the concentrations at the specified times are
For t = 20 s
λ = ([B]0 − [A]0 )k r t
= [(0.110 mol dm−3 ) − (0.060 mol dm−3 )] × (0.11 dm3 mol−1 s−1 ) × (20 s)
= 0.11
[B]0 [A]0 (e λ − 1) (0.110 mol dm−3 ) × (0.060 mol dm−3 ) × (e0.11 − 1)
x= =
[B]0 e λ − [A]0 (0.110 mol dm−3 ) × e0.11 − (0.060 mol dm−3 )
= 0.0122... mol dm−3
Hence the concentration of ester is
[B] = [B]0 − x
= (0.110 mol dm−3 ) − (0.0122... mol dm−3 ) = 0.0978 mol dm−3
For t = 15 min
λ = ([B]0 − [A]0 )k r t
= [(0.110 mol dm−3 ) − (0.060 mol dm−3 )] × (0.11 dm3 mol−1 s−1 )
× ([15 × 60] s) = 4.95...
[B]0 [A]0 (e λ − 1) (0.110 mol dm−3 ) × (0.060 mol dm−3 ) × (e4.95 ... − 1)
x= =
[B]0 e λ − [A]0 (0.110 mol dm−3 ) × e4.95 ... ) − (0.060 mol dm−3 )
= 0.0598... mol dm−3
Hence the concentration of ester is
[B] = [B]0 − x
= (0.110 mol dm−3 ) − (0.0598... mol dm−3 ) = 0.0502 mol dm−3
P14B.1 The concentration of B is given in the question as
[B] = n[A]0 (1 − e−k r t ) hence [B]/[A]0 = n(1 − e−k r t )
The concentration of A for a first-order reaction is given by [14B.2b–548],
[A] = [A]0 e−k r t hence [A]/[A]0 = e−k r t
These expressions are plotted in Fig. 14.3
P14B.3 The first task is to convert the masses of urea into concentrations of ammonium
cyanate A. Because the only fate of the ammonium cyanate is to be converted
into urea, the mass of ammonium cyanate m A remaining at any given time
is equal to the original mass of ammonium cyanate minus the mass of urea,
m A = m A,0 − m urea . In this case m urea = 22.9 g. Dividing by the molar mass
of the ammonium cyanate, M A = 60.0616 g mol−1 , gives the amount of A in
moles, and division by the volume of the solution then gives the concentration
in mol dm−3 .
B, n = 2
[A]/[A]0 or [B]/[A]0
2.0
1.0 B, n = 1
B, n = 1
2
A
0.0
0 1 2 3 4 5
kr t
Figure 14.3
t/min t/s m urea /g m A /g [A]/mol dm−3

0 0 0.0 22.9 0.381
20 1 200 7.0 15.9 0.265
50 3 000 12.1 10.8 0.180
65 3 900 13.8 9.1 0.152
150 9 000 17.7 5.2 0.087
The order is determined by testing the fit of the data to integrated rate law
expressions. A zeroth-order reaction of the form A → P has an integrated rate
law given by [14B.1–547], [A] = [A]0 − k r t, so if the reaction is zeroth-order
then a plot of [A] against t will be a straight line of slope −k r . On the other
hand, a first-order reaction has an integrated rate law given by [14B.2b–548],
ln([A]/[A]0 ) = −k r t, so if the reaction is first-order then a plot of ln [A]/[A]0
against t will be a straight line of slope −k r . Finally, if the order is n ≥ 2 the
integrated rate law is given in Table 14B.3 on page 551 as
1 1 1 1 1
kr t = ( − n−1
) hence = (n−1)k r t+
n − 1 ([A]0 − [P]) n−1 [A]0 [A] n−1 [A]0n−1
where to obtain the second expression the relation [P] = [A]0 − [A] is substi-
tuted and the equation rearranged. This expression implies that if the reaction
has order n ≥ 2 a plot of 1/[A]n−1 against t will be a straight line of slope
(n − 1)k r .
The data are plotted assuming zeroth-, first-, second-, and third-order in Fig. 14.4
and using the data in the table below. The second-order plot shows a good
straight line, while the other three plots show the data lying on distinct curves.
It is therefore concluded that the reaction is second-order .
[A] [A] 1/[A] 1/[A]2

t/s ln
/mol dm−3 [A]0 /dm3 mol−1 /dm6 mol−2
0 0.381 0.000 2.623 6.879
1 200 0.265 −0.365 3.777 14.269
3 000 0.180 −0.752 5.561 30.928
3 900 0.152 −0.923 6.600 43.562
9 000 0.087 −1.482 11.550 133.410
In an alternative approach visual examination of the concentration data indi-

cates that the half-life is not constant, and comparison of the 0–50 min data and
the 50-150 min data suggests that the second half-life is approximately double
the first. That is, the half-life starting from half the initial concentration is about
twice the initial half-life, suggesting that the half-life is inversely proportional
to the initial concentration. According to [14B.6–550] the half-life of a reaction
with order n > 1 is given by t 1/2 ∝ 1/[A]0n−1 , so this result suggests that the
reaction may have n = 2 as this gives t 1/2 ∝ 1/[A]0 . Because it is suspected
on this basis that the reaction may be second-order, only the second-order plot
from Fig. 14.4 is made, and the fact that it gives a good straight line confirms
that the reaction is indeed second-order.
0.4 zeroth-order 0.0 first-order

[A]/mol dm−3
ln([A]/[A]0 )
0.3
−0.5
0.2
−1.0
0.1
−1.5
0.0
0 4 000 8 000 0 4 000 8 000
t/s t/s
150
second-order third-order
[A]−1 /dm3 mol−1
[A]−2 /dm6 mol−2
10 100
5 50
0
0
0 4 000 8 000 0 4 000 8 000
t/s t/s
Figure 14.4
The equation of the line in the second-order plot is

[A]−1 /dm3 mol−1 = 9.95 × 10−4 × (t/s) + 2.62
Identifying the slope with (n − 1)k r as discussed above, and noting that n = 2
for a second-order reaction, gives k r = 9.95 × 10−4 dm3 mol−1 s−1 .
The concentration of ammonium cyanate left after 300 min, which is (300min)×
(60s/1 min) = 18000 s, is calculated using the integrated rate law for a second-
order reaction [14B.4b–549], [A] = [A]0 /(1 + k r t[A]0 )
(0.381... mol dm−3 )
[A] =
1 + (9.95... × 10−4 dm3 mol−1 s−1
)
× (18000 s) × (0.381... mol dm−3 )
= 0.0487... mol dm−3
Multiplication by the volume gives the amount in moles, and multiplication of
this by the molar mass gives the mass of A in g.
m A = MV [A] = (60.0616 g mol−1 )×(1.00 dm)×(0.0487... mol dm−3 ) = 2.9 g
P14B.5 The order is determined by testing the fit of the data to integrated rate law
expressions. A zeroth-order reaction of the form A → P has an integrated rate
law given by [14B.1–547], [A] = [A]0 −k r t, so if the reaction is zeroth-order then
a plot of [A] against t will be a straight line of slope −k r . In this case, A is the
organic nitrile. On the other hand, a first-order reaction has an integrated rate
law given by [14B.2b–548], ln([A]/[A]0 ) = −k r t, so if the reaction is first-order
then a plot of ln [A]/[A]0 against t will be a straight line of slope −k r . Finally,
if the order is n ≥ 2 the integrated rate law is given in Table 14B.3 on page 551
as
1 1 1 1 1
kr t = ( − n−1
) hence = (n−1)k r t+
n − 1 ([A]0 − [P]) n−1 [A]0 [A] n−1 [A]0n−1
tuted and the equation rearranged. This expression implies that if the reaction
has order n ≥ 2 a plot of 1/[A]n−1 against t will be a straight line of slope
(n − 1)k r .
The data are plotted assuming zeroth-, first-, second-, and third-order in Fig. 14.5.
The second-order plot shows the best fit to a straight line, so it is concluded
that the reaction is likely to be second-order . However, the first-order and
third-order plots also give a reasonable fit to a straight line, so experimental
data over a wider range of concentrations would be needed to establish the
order with greater confidence.
[A] [A] 1/[A] 1/[A]2

t/103 s ln
/mol dm−3 [A]0 /dm3 mol−1 /dm6 mol−2
0 1.500 0 0.000 0.67 4.444 × 10−1
2 1.260 0 −0.174 0.79 6.299 × 10−1
4 1.070 0 −0.338 0.93 8.734 × 10−1
6 0.920 0 −0.489 1.09 1.181 × 100
8 0.810 0 −0.616 1.23 1.524 × 100
10 0.720 0 −0.734 1.39 1.929 × 100
12 0.650 0 −0.836 1.54 2.367 × 100

[A]/mol dm−3
ln([A]/[A]0 )
1.0 −0.5
0.5 −1.0
0 5 10 0 5 10
t/10 s
3
t/10 s 3
1.5 second-order third-order

[A]−1 /dm3 mol−1
[A]−2 /dm6 mol−2

2
1.0
1
0.5
0
0 5 10 0 5 10
t/10 s
3
t/10 s 3
Figure 14.5
The equation of the line in the second-order plot is
[A]−1 /dm3 mol−1 = 7.33 × 10−5 × (t/s) + 0.652
Identifying the slope with (n − 1)k r as discussed above and noting that n = 2
for a second-order reaction gives k r = 7.33 × 10−5 dm3 mol−1 s−1 .
P14B.7 The order is determined by testing the fit of the data to integrated rate law
expressions. A first-order reaction has an integrated rate law given by [14B.2b–
548], ln([A]/[A]0 ) = −k r t, or ln[A] − ln[A]0 = −k r t, so if the reaction is
first-order then a plot of ln [A] against t will be a straight line of slope −k r .
On the other hand, a second-order reaction has an integrated rate law given
by [14B.4b–549], 1/[A] − 1/[A]0 = k r t, which implies that if the reaction is
second-order then a plot of 1/[A] against t will be a straight line of slope k r .
The data are plotted in Fig. 14.6. The first-order plot shows a good straight line
while in the second-order plot the data lie on a curve. It is therefore concluded
that the reaction is first-order .
t/min c/ng cm−3 ln(c/ng cm−3 ) 1/(c/ng cm−3 )

30 699 6.550 1.43 × 10−3
60 622 6.433 1.61 × 10−3
120 413 6.023 2.42 × 10−3
150 292 5.677 3.42 × 10−3
240 152 5.024 6.58 × 10−3
360 60 4.094 1.67 × 10−2
480 24 3.178 4.17 × 10−2
7
first-order second-order
0.04
1/(c/ng cm−3 )
6
ln(c/ng cm−3 )
5
0.02
4
3
0.00
0 200 400 0 200 400
t/min t/min
Figure 14.6
The equation of the line in the first-order plot is
ln(c/ng cm−3 ) = −7.65 × 10−3 × (t/min) + 6.86
Identifying the slope with −k r as discussed above gives the first-order rate con-
stant as k r = 7.65 × 10−3 min−1 .
The half-life of a first-order reaction is given by [14B.3–548], t 1/2 = ln 2/k r .
ln 2 ln 2
t 1/2 = = = 91 min
kr 7.65 × 10−3 min−1
P14B.9 The units of the rate constants show that both reactions are first-order, so their
rate equations are assumed to be
d[CH4 ] d[CH2 CO]
υ1 = = k 1 [CH3 COOH] and υ2 = = k 2 [CH3 COOH]
dt dt
where [14A.3b–542], υ = (1/ν J )(d[J]/dt), is used to express the rates in terms
of rate of formation of CH4 and CH2 CO. The ratio of the rate of ketene forma-
tion to the total rate of product formation is therefore
k 2 [CH3 COOH] k2 4.65 s−1

= =
k 1 [CH3 COOH] + k 2 [CH3 COOH] k 1 + k 2 (3.74 s−1 ) + (4.65 s−1 )
= 0.554...
Because this ratio is independent of the CH3 COOH concentration it will be

constant throughout the duration of the reaction and will be equal to the ratio of
ketene formed to total product formed. The maximum possible yield of ketene
is therefore 0.554... or 55.4 % .
Similarly the ratio of the rate of formation of ketene and methane is
k 2 [CH3 COOH] k 2 4.65 s

= = = 1.24...
k 1 [CH3 COOH] k 1 3.74 s
This ratio is independent of the CH3 COOH concentration so it will be constant

throughout the duration of the reaction and equal to the ratio of the total ketene
and methane formed up to any given time. Hence [CH2 CO]/[CH4 ] = 1.24...
and is constant over time.
P14B.11 The first task is to calculate the concentrations of the reactant A at each time.
The stoichiometry of the reaction 2A → B means that the initial concentration
of A is twice the final concentration of B, [A]0 = 2[B]∞ . In addition, the
amount of A that has reacted at any given time is equal to twice the amount
of B that has been formed. It follows that
A that has reacted
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
[A]0 − [A] = 2[B] hence [A] = [A]0 − 2[B]
Substituting [A]0 = 2[B]∞ from above gives [A] = 2([B]0 − [B]); this expres-
sion is used to calculate the concentration of [A] at each of the times.
The order is determined by testing the fit of the data to integrated rate law
expressions. If the rate law is υ = k r [A]n , where n is the order to be deter-
mined, expressing υ in terms of the rate of change of concentration of [A] using
[14A.3b–542], υ = (1/ν J )(d[J]/dt) gives
1 d[A] d[A]
υ= = k r [A]n hence = −2k r [A]n
−2 dt dt
Integrated rate laws are given in Table 14B.3 on page 551, but care is needed
because these are for reactions of the form A → P but here the reaction is 2A →
B.
For n = 0, Table 14B.3 on page 551 shows that a reaction A → P with rate law
υ = d[P]/dt = k r has integrated rate law A = A0 − k r t. To adapt this expression
for the reaction in question, the rate law for the reaction in the table is first
written as d[A]/dt = −k r using d[P]/dt = −d[A]/dt for a reaction of the form
A → P. This rate law matches that found above, d[A]/dt = −2k r [A]n , for n = 0
except that k r is replaced by 2k r . The integrated rate law will therefore be the
same except with k r replaced by 2k r , that is, [A] = [A]0 − 2k r t. This expression
implies that if the reaction is zeroth-order a plot of [A] against t will give a
straight line of slope −2k r .
Similarly Table 14B.3 on page 551 gives the integrated rate law for a first-order
reaction A → P with rate law υ = d[P]/dt = k r [A] as ln([A]0 /[A]) = k r t,
equivalent to [14B.2b–548], ln([A]/[A]0 ) = −k r t. By the same reasoning as
above the integrated rate law for the reaction will therefore be ln([A]/[A]0 ) =
−2k r t, implying that a plot of ln([A]/[A]0 ) against t will give a straight line of
slope −2k r .
Finally, if the order is n ≥ 2 the integrated rate law for a reaction A → P with
rate law υ = d[P]/dt = k r [A]n is given in Table 14B.3 on page 551 as
1 1 1 1 1
kr t = ( − ) hence = (n−1)k r t+
n − 1 ([A]0 − [P])n−1 [A]0n−1 [A]n−1 [A]0n−1
tuted and the equation rearranged. Adapting this expression for the reaction
in question gives 1/[A]n−1 = 2(n − 1)k r t + 1/[A]0n−1 . This expression implies
that if the reaction has order n ≥ 2 a plot of 1/[A]n−1 against t will be a straight
line of slope 2(n − 1)k r .
The data are plotted assuming zeroth-, first-, second-, and third-order in Fig. 14.7.
The first-order plot shows a good fit to a straight line, while the other plots are
curved, so it is concluded that the reaction is first-order .
[B] [A] [A] 1/[A] 1/[A]2

t/min ln
/mol dm−3 /mol dm−3 [A]0 /dm3 mol−1 /dm6 mol−2
0 0.000 0.624 0.000 1.603 2.57
10 0.089 0.446 −0.336 2.242 5.03
20 0.153 0.318 −0.674 3.145 9.89
30 0.200 0.224 −1.025 4.464 19.93
40 0.230 0.164 −1.336 6.098 37.18
∞ 0.312 0.000
The equation of the line in the first-order plot is

ln([A]/[A]0 ) = −0.03361 × (t/min) − 1.896 × 10−3
Identifying the slope with −2k r as discussed above gives the first-order rate
constant as
k r = − 21 × (−0.03361 min−1 ) = 0.0168 min−1
P14B.13 The order is determined by fitting the data to integrated rate laws. Table 14B.3
on page 551 gives integrated rate laws for the reaction A → P with rate law
υ = d[P]/dt = k r [A]n as
n=0 [A] = [A]0 − k r t hence k r = ([A]0 − [A]) /t
[A]0 1 [A]0
n=1 k r t = ln hence kr = ln
[A] t [A]
1 1 1
n≥2 kr t = ( − )
n − 1 ([A]0 − [P]) n−1 [A]0n−1
1 1 1
hence kr = ( − )
(n − 1)t [A]n−1 [A]0n−1

[A]/mol dm−3
ln([A]/[A]0 )
−0.5
0.4
−1.0
0.2
−1.5
0.0
0 20 40 0 20 40
t/min t/min
40
6 second-order third-order
[A]−1 /dm3 mol−1
[A]−2 /dm6 mol−2

30
4
20
2 10
0 0
0 20 40 0 20 40
t/min t/min
Figure 14.7
where in the n ≥ 2 case, [A] = [A]0 − [P] is used.

These expressions imply that if the reaction is zeroth-order the quantity ([A]0 −
[A])/t should be a constant, equal to k r , while if it is first-order [ln([A]0 /[A])]/t
should be constant, and if the order is n ≥ 2 the quantity
([A]−(n−1) − [A]0n−1 )/[(n − 1)t]
should be constant. In this case A is cyclopropane and the concentrations are

expressed in terms of partial pressures. Results assuming n = 0, 1, 2 are shown
in the following table.
n=0 n=1 n=2

p A,0 pA (p A,0 − p A )/t [ln(p A,0 /p A )]/t (1/p A − 1/p A,0 )/t
/Torr t/s /Torr /Torr s−1 /s−1 /Torr−1 s−1
200 100 186 0.140 7.26 × 10−4 3.76 × 10−6
200 200 173 0.135 7.25 × 10−4 3.90 × 10−6
400 100 373 0.270 6.99 × 10−4 1.81 × 10−6
400 200 347 0.265 7.11 × 10−4 1.91 × 10−6
600 100 559 0.410 7.08 × 10−4 1.22 × 10−6
600 200 520 0.400 7.16 × 10−4 1.28 × 10−6
The values assuming n = 1 are approximately constant, while those assum-

ing n = 0 or n = 2 are not. It is therefore concluded that the reaction is
first-order . The average value of [ln(p A,0 /p A )]/t, and hence of k r , from the
table is 7.1 × 10−4 s−1 .
P14B.15 A reaction of the form A → P that is nth order in A has rate law υ = k r [A]n .
Combining this with [14A.3b–542], υ = (1/ν J )(d[J]/dt) gives
1 d[A]
= k r [A]n hence − [A]−n d[A] = k r dt
−1 dt
Initially, at t = 0, the concentration of A is [A]0 , and at a later time t it is [A].
These are used as the limits of the integration to give
Integral A.1
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ t
[A] t [A]
1
∫ −[A]−n d[A] = ∫ k r dt hence [A]−(n−1) ∣ = k r t∣
[A]0 0 n−1 [A]0 0
1 1 1
hence ( − ) = kr t (for n ≠ 1)
n − 1 [A]n−1 [A]0n−1
This integrated rate law is equivalent to that given in Table 14B.3 on page 551.
(a) When t = t 1/2 , [A] = 12 [A]0 . Therefore, on dividing through by k r , the

above integrated rate law gives
1 ⎛ 1 1 ⎞ 1 2n−1 1
t 1/2 = n−1 − n−1
= ( n−1
− )
(n − 1)k r ⎝ ( [A]0 )
1 [A]0 ⎠ (n − 1)k r [A]0 [A]0n−1
2
2n−1 − 1
=
(n − 1)k r [A]0n−1
(b) The time taken for the concentration of a substance to fall to one-third its
initial value is denoted t 1/3 . Thus, at t = t 1/3 , [A] = 31 [A]0
1 ⎛ 1 1 ⎞ 1 3n−1 1
t 1/3 = − = ( − )
(n − 1)k r ⎝ ( 1 [A]0 )n−1 [A]0n−1 ⎠ (n − 1)k r [A]0n−1 [A]0n−1
3
3n−1 − 1
=
(n − 1)k r [A]0n−1
P14B.17 The stoichiometry of the reaction 2A + B → P implies that when the concen-
tration of P has increased from 0 to x, the concentration of A has fallen to
[A]0 − 2x and the concentration of B has fallen to [B]0 − x. This is because
each P that forms entails the disappearance of two A and one B. The rate law
υ = d[P]/dt = k r [A]2 [B] then becomes
d[P] dx
= k r ([A]0 − 2x)2 ([B]0 − x) hence = k r ([A]0 − 2x)2 ([B] − x)
dt dt
where to go to the second expression [P] = x is used, which implies that d[P]dt =
dx/dt. The expression is rearranged and the initial condition x = 0 when t = 0
is applied. This gives the integrations required as
x 1 t
∫ dx = ∫ k r dt
0 ([A]0 − 2x) ([B]0 − x)
2 0
The right-hand side evaluates to k r t. The left-hand side is evaluated below.
(a) If [B]0 = 12 [A]0 the left-hand side becomes
x 1 x 1
∫ dx = ∫ dx
0 ([A]0 − 2x)2 ( [A]0 − x)
1 1
0 ([A]0 − 2x)2 × ([A]0 − 2x)
2 2
x x
−3 1 −2 1 1
= ∫ 2([A]0 − 2x) dx = ([A]0 − 2x) ∣ = −
0 2 0 2([A]0 − 2x)2 2[A]20
Combining this with the right-hand side from above gives the integrated
rate law as
1 1
− = kr t
2([A]0 − 2x)2 2[A]20
(b) If [B]0 = [A]0 the left-hand side is integrated using the method of partial
fractions described in The chemist’s toolkit 28 in Topic 14B. The integrand
is first written as
1 A B C
= + +
([A]0 − 2x) ([A]0 − x) ([A]0 − 2x)
2 2 [A]0 − 2x [A]0 − x
where A, B, and C are constants to be found. This expression is multiplied

through by ([A]0 − 2x)2 ([A]0 − x) to give
1 = A([A]0 − x) + B([A]0 − 2x)([A]0 − x) + C([A]0 − 2x)2
This expression must be true for all x, so the values of A, B and C are most
conveniently found by substituting particular values of x.
When x = [A]0 1 = C(−[A]0 )2 hence C = 1/[A]20

When x = 21 [A]0 1 = A( 12 [A]0 ) hence A = 2/[A]0
When x = 0 1 = A[A]0 + B[A]20 + C[A]20
2 1
= [A]0 + B[A]20 + [A]20
[A]0 [A]20
hence 1 = 3 + B[A]20 hence B = −2/[A]20
The required integral is therefore

x 1
∫ dx
0 ([A]0 − 2x)2 ([A]0 − x)
x 2 2 1
=∫ − + dx
0 [A]20 ([A]0 − 2x) [A]20 ([A]0 − 2x) [A]20 ([A]0 − x)
x
1 1 1
= + ln([A]0 − 2x) − ln([A]0 − x)∣
[A]0 ([A]0 − 2x) [A]20 [A]20 0
1 1 1
=( + 2
ln([A]0 − 2x) − ln([A]0 − x))
[A]0 ([A]0 − 2x) [A]0 [A]20
1 1 1
−( 2
+ 2
ln[A]0 − ln[A]0 )
[A]0 [A]0 [A]20
1 1 [A]0 − 2x 1
= + ln −
[A]0 ([A]0 − 2x) [A]0 2 [A]0 − x [A]20
Combining this with the right-hand side integral found above gives the
integrated rate law as
1 1 [A]0 − 2x 1
+ ln − = kr t
[A]0 ([A]0 − 2x) [A]20 [A]0 − x [A]20
14C Reactions approaching equilibrium

D14C.1 If the equilibrium position shifts with pressure, a pressure jump can be used
to alter the rate of the reaction. For such an effect on equilibrium, the volume
change of the reaction must be non-zero.
E14C.1(a) The relaxation time in a jump experiment is given by [14C.9a–555], τ = 1/(k r +
k r′ ). This equation is rearranged for k r′ . It is convenient to convert τ to ms.
1 1
k r′ = − kr = − (12.4 ms−1 ) = 23.8 ms−1
τ 27.6 × 10−3 ms
E14C.2(a) The equilibrium constant in terms of rate constants is given by [14C.8–554], K =

k r /k r′ . However because the forward and backward reactions are of different
order it is necessary to include a factor of c −○ so that the ratio of k r , with units
dm3 mol−1 s−1 , to k r′ , with units s−1 , is turned into a dimensionless quantity.
The equation required is
k r c −○ (5.0 × 106 dm3 mol−1 s−1 ) × (1 mol dm−3 )

K= = = 2.5 × 102
k r′ 2.0 × 104 s
P14C.1 The expression for [A] in [14C.4–553] is differentiated
′
k r′ + k r e−(k r +k r )t d[A] ′
[A] = [A]0 hence = −k r [A]0 e−(k r +k r )t
k r + k r′ dt
According to [14C.3–553], d[A]/dt = −(k r + k r′ )[A]+ k r′ [A]0 . To verify that the
two expressions for d[A]/dt are the same, the expression for [A] from [14C.4–
553] is substituted into [14C.3–553]
[A]
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
′
d[A] k ′ + k r e−(k r +k r )t
= −(k r + k r′ ) [ r [A]0 ] +k r′ [A]0
dt k r + k r′
′ ′
= −k r′ [A]0 − k r e−(k r +k r )t [A]0 + k r′ [A]0 = −k r [A]0 e−(k r +k r )t
Therefore the two expressions for d[A]/dt are the same and so the equation is
satisfied.
P14C.3 (a) The forward and backward reactions are
d[A] d[A]
A→B = −k r [A] B→A = +k r′ [B]
dt dt
The overall rate of change of [A] is therefore
d[A]
= −k r [A] + k r′ [B]
dt
The stoichiometry of the reaction A ⇌ B means that the amount of B
present at any time is equal to the initial amount plus the amount of A
that has reacted. Hence [B] = [B]0 + ([A]0 − [A]). This is substituted
into the above expression to give
d[A]
= −k r [A] + k r′ ([B]0 + [A]0 − [A])
dt
Rearranging and integrating with the initial condition that [A] = [A]0
when t = 0 gives
[A] d[A] t
∫ = ∫ dt
[A]0 k r′ ([A]0 ′
+ [B]0 ) − (k r + k r )[A] 0
Performing the integration gives

[A]
ln [k r′ ([A]0 + [B]0 ) − (k r + k r′ )[A]]
∣ =t
−(k r + k r′ ) [A]0
1 k r′ ([A]0 + [B]0 ) − (k r + k r′ )[A]
Hence ln =t
−(k r + k r′ ) k r′ ([A]0 + [B]0 ) − (k r + k r′ )[A]0
Rearranging for [A] yields
′
k r′ ([A]0 + [B]0 ) + (k r [A]0 − k r′ [B]0 )e−(k r +k r )t
[A] =
k r + k r′
(b) As t → ∞, the exponential term in the expression for [A] decreases to

zero and the concentrations reach their equilibrium values. The equilib-
rium concentration of A is therefore
k r′ ([A]0 + [B]0 )
[A]eq =
k r + k r′
Noting from above that the concentration of B is given by [B] = [B]0 +

[A]0 − [A], the equilibrium concentration of B is therefore
k r′ ([A]0 + [B]0 )
[B]eq = [B]0 + [A]0 − [A]eq = [B]0 + [A]0 −
k r + k r′
(k r + k r′ )([B]0 + [A]0 ) − k r′ ([A]0 + [B]0 ) k r′ ([A]0 + [B]0 )
= =
k r + k r′ k r + k r′
This result is alternatively and more simply obtained by noting that at

equilibrium the rates of the forward and backward reactions are equal,
implying that
k r [A]eq = k r′ [B]eq hence k r [A]eq = k r′ ([B]0 + [A]0 − [A]eq )

k r′ ([B]0 + [A]0 )
hence [A]eq = as before
k r + k r′
P14C.5 (a) Application of [14A.3b–542], υ = (1/ν J )(d[J]/dt), to the forward and

backward reactions gives
1 d[A] d[A]
Forward 2A → A2 υ = k a [A]2 = hence = −2k a [A]2
−2 dt dt
1 d[A] d[A]
Backward A2 → 2A υ = k a′ [A2 ] = hence = 2k a′ [A2 ]
2 dt dt
The overall rate of change of A is therefore
d[A]
= −2k a [A]2 + 2k a′ [A2 ]
dt
If the deviation of [A] from its new equilibrium value is denoted 2x, so
that [A] = [A]eq + 2x, the stoichiometry of the reaction implies that
[A2 ] = [A2 ]eq − x. These are substituted into the above expression to
give
d[A] 2
= −2k a ([A]eq + 2x) + 2k a′ ([A2 ]eq − x)
dt
= −2k a ([A]2eq + 4x[A]eq + 4x 2 ) + 2k a′ ([A2 ]eq − x)
0 Neglect
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
= −2k a [A]2eq + 2k a′ [A2 ]eq −8k a x[A]eq + 2k a′ x + 8k a x 2
= − (8k a [A]eq + 2k a′ ) x
In the third line the first two terms cancel because at equilibrium the rates
of the forward reaction k a [A]2eq and the backward reaction k a′ [A2 ]eq are
equal. The last term is neglected because x is assumed to be small. Next,
because [A] = [A]eq + 2x it follows that d[A]/dt = 2 dx/dt. This is
substituted into the above expression to give
dx dx
2 = − (8k a [A]eq + 2k a′ ) x hence = − (4k a [A]eq + k a′ ) x
dt dt
Rearranging, and integrating with the condition that x = x 0 when t = 0
gives
x
dx t x
∫ = ∫ − (4k a [A]eq + k a′ ) dt hence ln = − (4k a [A]eq + k a′ ) t
x0 x 0 x0
′ 1
hence x = x 0 e−(4k a [A]eq +k a )t = x 0 e−t/τ where = 4k a [A]eq + k a′
τ
Squaring both sides of the expression for 1/τ gives
1 2
= (4k a [A]eq + k a′ ) = 16k a2 [A]2eq + 8k a k a′ [A]eq + k a′2
τ2
k a′ [A 2 ]eq
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
= 16k a (k a [A]2eq ) +8k a k a′ [A]eq + k a′2 = 16k a k a′ [A2 ]eq + 8k a k a′ [A]eq + k a′2
[A]tot
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
= 8k a k a′ (2[A2 ]eq + [A]eq ) +k a′2 = 8k a k a′ [A]tot + k a′2
In the second line, k a [A]2eq = k a′ [A2 ]eq is used; these quantities are equal
because as explained above the rates of the forward and backward reac-
tions are equal at equilibrium. In the third line, the relationship [A]tot =
[A] + 2[A2 ] is used; this expression is valid at all stages of the reaction
including at equilibrium.
(b) The result 1/τ 2 = 8k a k a′ [A]tot + k a′2 implies that a plot of 1/τ 2 against
[A]tot should give a straight line of intercept k a′2 and slope 8k a k a′ ; from
these quantities k a′ and k a are determined.
(c) The data are plotted in Fig. 14.8.
[P]/mol dm−3 τ/ns τ −2 /ns−2

0.500 2.3 0.189
0.352 2.7 0.137
0.251 3.3 0.092
0.151 4.0 0.063
0.101 5.3 0.036
The data fall on a reasonable straight line, the equation for which is
τ −2 /ns−2 = 0.380 × ([P]/mol dm−3 ) + 2.87 × 10−4

0.20
0.15
τ −2 /ns−2
0.10
0.05
0.00
0.0 0.1 0.2 0.3 0.4 0.5 0.6
−3
[P]/mol dm
Figure 14.8
Identifying the intercept with k a′2 gives k a′2 = 2.87 × 10−4 ns−2 , hence
√
k a′ = 2.87 × 10−4 ns−2 = 0.0169... ns−1 = 1.7 × 107 s−1
Identifying the slope with 8k a k a′ gives
8k a k a′ = 0.380 dm3 mol−1 ns−2

0.380 dm3 mol−1 ns−2 0.380 dm3 mol−1 ns−2
hence k a = =
8k a′ 8 × (0.0169... ns−1 )
= 2.80... dm3 mol−1 ns−1 = 2.8 × 109 dm3 mol−1 s−1
The equilibrium constant is given by [14C.8–554], K = (k a /k a′ )×(k b /k b′ )×

..., but it is necessary to include a factor of c −○ because the forward reaction
is second-order while the backward reaction is first-order.
k a c −○ (2.80 ... dm3 mol−1 ns−1 ) × (1 mol dm−3 )
K= = = 1.7 × 10−2
k a′ (0.0169... ns−1 )
It is noted that the points in Fig. 14.8 do not lie on a perfect straight line,
and the intercept is closer to zero than some of the points are to the line.
In fact, mathematical software gives the standard error in the intercept as
4 × 10−3 ns−2 , which is an order of magnitude larger than the intercept
itself. This indicates that there is considerable uncertainty in the intercept
and therefore in the values of the rate constants and equilibrium constant
deduced from it.
14D The Arrhenius equation

D14D.1 The temperature dependence of some reactions is not Arrhenius-like, in the
sense that a straight line is not obtained when ln k r is plotted against 1/T.
However, it is still possible to define an activation energy using [14D.3–558],

E a = RT 2 (d ln k r /dT). This definition reduces to the earlier one (as the slope
of a straight line) for a temperature-independent activation energy. However,
this latter definition is more general, because it allows E a to be obtained from
the slope (at the temperature of interest) of a plot of ln k r against 1/T even if
the Arrhenius plot is not a straight line. Non Arrhenius behaviour is sometimes
a sign that quantum mechanical tunnelling is playing a significant role in the
reaction. A reaction with a very small or zero activation energy, so that k r = A,
such as for some radical recombination reactions in the gas phase, has a rate
that is largely temperature independent.
E14D.1(a) The relationship between the values of a rate constant at two different temper-
atures is given by [14D.2–558], ln(k r,2 /k r,1 ) = (E a /R)(1/T1 − 1/T2 ). Hence,
taking T1 = 37 ○ C and T2 = 15 ○ C,
k r,2 Ea 1 1
= exp [ ( − )]
k r,1 R T1 T2
87 × 103 J mol−1 1 1
= exp [ ×( − )]
8.3145 J K−1 mol−1 [37+273.15] K [15+273.15] K
= 0.076
The rate constant therefore drops to about 7.6 % of its original value when the
temperature is lowered for 37 ○ C to 15 ○ C.
E14D.2(a) As explained in Section 14D.2(a) on page 559 the fraction f of collisions that are
sufficiently energetic to be successful is given by the exponential factor e−E a /RT .
Rearranging f = e−E a /RT for T and setting f = 0.10 gives
Ea 50 × 103 J mol−1
T =− =− = 2.6 × 103 K
R ln f (8.3145 J K−1 mol−1 ) × ln 0.10
E14D.3(a) The Arrhenius equation is given by [14D.4–559], k r = Ae−E a /RT . In this case
23 × 103 J mol−1
k r = (8.1 × 10−10 dm3 mol−1 s−1 ) × exp (− )
(8.3145 J K−1 mol−1 ) × (500 K)
= 3.2 × 10−12 dm3 mol−1 s−1
E14D.4(a) The relationship between the values of a rate constant at two different tempera-
tures is given by [14D.2–558], ln(k r,2 /k r,1 ) = (E a /R)(1/T1 − 1/T2 ). Rearrang-
ing for E a gives
−1 −1 −2 −3
R ln(k r,2 /k r,1 ) (8.3145 J K mol ) × ln [(2.67 × 10 )/(3.80 × 10 )]
Ea = =
1/T1 − 1/T2 1/([35 + 273.15] K) − 1/([50 + 273.15] K)
= 1.07... × 105 J mol−1 = 108 kJ mol−1
The frequency factor is found by rearranging the Arrhenius equation [14D.4–

559], k r = Ae−E a /RT , for A. The data for both temperatures gives the same result.
At T1 A = k r eE a /RT1
1.07...×105 J mol−1
= (3.80×10−3 dm3 mol−1 s−1 )×exp
(8.3145 J K−1 mol−1 )×([35+273.15] K)
= 6.62 × 1015 dm3 mol−1 s−1
At T2 A = k r eE a /RT2
1.07...×105 J mol−1
= (2.67×10−2 dm3 mol−1 s−1 )×exp
(8.3145 J K−1 mol−1 )×([50+273.15] K)
= 6.62 × 1015 dm3 mol−1 s−1
E14D.5(a) The relationship between the values of a rate constant at two different tempera-
tures is given by [14D.2–558], ln(k r,2 /k r,1 ) = (E a /R)(1/T1 − 1/T2 ). Rearrang-
ing for E a , and using k r,2 /k r,1 = 3 because the rate constant triples between the
two temperatures, gives
R ln(k r,2 /k r,1 ) (8.3145 J K−1 mol−1 ) × ln 3

Ea = = = 35 kJ mol−1
1/T1 −1/T2 1/([24+273.15] K)−1/([49+273.15] K)
P14D.1 The definition of E a in [14D.3–558], E a = RT 2 (d ln k r /dT), is rearranged and
integrated.
Ea d ln k r Ea
= hence ∫ d ln k r = ∫ dT
RT 2 dT RT 2
The left-hand side integral is simply ln k r . If E a does not vary with temperature
then the integral on the right is evaluated by taking E a /R outside the integral
to give
Ea 1 Ea
ln k r = ∫ dT = − +c
R T 2 RT
This is [14D.1–557], ln k r = ln A − E a /RT, once the constant of integration c is
identified with ln A.
P14D.3 The Arrhenius equation [14D.1–557], ln k r = ln A−E a /RT, implies that a plot of
ln k r against 1/T should give a straight line of slope −E a /R and intercept ln A.
T/K k r /dm3 mol−1 s−1 1/(T/K) ln(k r /dm3 mol−1 s−1 )

1 000 8.35 × 10−10 0.001 000 −20.90
1 200 3.08 × 10−8 0.000 833 −17.30
1 400 4.06 × 10−7 0.000 714 −14.72
1 600 2.80 × 10−6 0.000 625 −12.79
ln(k r /dm3 mol−1 s−1 )

−15
−20
0.0007 0.0009 0.0011

1/(T/K)
Figure 14.9
ln(k r /dm3 mol−1 s−1 ) = (−2.165 × 104 ) × 1/(T/K) + 0.7457
Identifying the slope with −E a /R gives the activation energy as
E a = −slope × R = −(−2.165 × 104 K) × (8.3145 J K−1 mol−1 ) = 180 kJ mol−1
Identifying the intercept with ln A gives the frequency factor as
A = e0.7457 dm3 mol−1 s−1 = 2.11 dm3 mol−1 s−1
The units of A are the same as the units of k r .
P14D.5 The Arrhenius equation [14D.1–557], ln k r = ln A−E a /RT, implies that a plot of
ln k r against 1/T should give a straight line of slope −E a /R and intercept ln A.
T/K k r /dm3 mol−1 s−1 1/(T/K) ln(k r /dm3 mol−1 s−1 )

295 3.55 × 106 0.003 39 15.08
223 4.94 × 105 0.004 48 13.11
218 4.52 × 105 0.004 59 13.02
213 3.79 × 105 0.004 69 12.85
206 2.95 × 105 0.004 85 12.59
200 2.41 × 105 0.005 00 12.39
195 2.17 × 105 0.005 13 12.29
The data fall on a reasonable straight line, the equation for which is
ln(k r /dm3 mol−1 s−1 ) = (−1.642 × 103 ) × 1/(T/K) + 20.59

15
ln(k r /dm3 mol−1 s−1 )

14
13
12
0.0035 0.0040 0.0045 0.0050
1/(T/K)
Figure 14.10
Identifying the slope with −E a /R gives the activation energy as
E a = −slope × R = −(−1.642 × 103 K) × (8.3145 J K−1 mol−1 ) = 13.7 kJ mol−1
Identifying the intercept with ln A gives the frequency factor as
A = e20.59 dm3 mol−1 s−1 = 8.75 × 108 dm3 mol−1 s−1
The units of A are the same as the units of k r .
14E Reaction mechanisms

D14E.1 In the pre-equilibrium assumption an intermediate is assumed to be in equilib-
rium with the reactants. For this assumption to apply it is necessary for the rate
at which the intermediate returns to reactants to be fast compared to the rate at
which the intermediate goes to products. The rate-determining step is between
the intermediate and the products. The rate at which the intermediate is formed
from the reactants is not relevant to the establishment of pre-equilibrium pro-
vided the other criteria are satisfied.
In the steady-state assumption an intermediate is formed from the reactants
and the moment it is formed it goes on to products. The step between reactants
and intermediate is therefore the rate-determining step, and necessarily the
concentration of the intermediate is low.
The two approximations differ in that in the steady-state approximation the in-
termediate is necessarily at low concentration, whereas in the pre-equilibrium
approximation this condition does not hold – indeed, the intermediate may ac-
cumulate. However, if the intermediate reacts immediately by either returning
to reactants or going on to products, then the pre-equilibrium assumption also
results in a low concentration of the intermediate.
In the pre-equilibrium assumption the apparent rate constant is a product of a

rate constant and an equilibrium constant and so the activation energy may be
negative (Section 14E.5 on page 566). For the steady-state approximation the
activation energy is positive.
D14E.3 Suppose that a reactant R can give alternative products P and Q by different
reactions. If the rate constant for the formation of P is greater than that for
forming Q, then to start with more P will be formed. However, as time proceeds
it may be that the reverse reactions from P and Q back to R start to become
significant, and eventually the reactions reach equilibrium. It may be that at
equilibrium the amount of Q exceeds that of P, even though initially the amount
of P exceeded that of Q.
If the relative proportions of the products are determined by the rate at which
they are formed, the reaction is said to be under kinetic control. If the amounts
are determined by the relevant equilibrium constants, the reaction is said to
be under thermodynamic control. The latter will only occur if the reverse
reactions are significant.
D14E.5 The overall reaction order is the sum of the powers of the concentrations of
all of the substances appearing in the experimental rate law for the reaction;
hence, it is the sum of the individual orders (exponents) associated with a each
reactant. Reaction order is an experimentally determined quantity.
Molecularity is the number of reactant molecules participating in an elemen-
tary reaction. Molecularity has meaning only for an elementary reaction, but
reaction order applies to any reaction. In general, reaction order bears no
necessary relation to the stoichiometry of the reaction, with the exception of
elementary reactions, where the order of the reaction corresponds to the num-
ber of molecules participating in the reaction; that is, to its molecularity. Thus
for an elementary reaction, overall order and molecularity are the same and are
determined by the stoichiometry.
E14E.1(a) The steady-state approximation is applied to the intermediate species O.
d[O]
= k a [O3 ] − k a′ [O2 ][O] − k b [O][O3 ] = 0
dt
Rearranging for [O] gives
k a [O3 ]
(k a′ [O2 ] + k b [O3 ]) [O] = k a [O3 ] hence [O] =
k a′ [O2 ] + k b [O3 ]
The rate of decomposition of O3 is
d[O3 ]
= −k a [O3 ]+k a′ [O2 ][O]−k b [O][O3 ] = −k a [O3 ]+[O]{k a′ [O2 ]−k b [O3 ]}
dt
because O3 is consumed in steps 1 and 3, but produced in step 2. Inserting the

steady-state expression for [O] gives
d[O3 ] k a [O3 ]{k a′ [O2 ] − k b [O3 ]}
= −k a [O3 ] +
dt k a′ [O2 ] + k b [O3 ]
−k a k a′ [O2 ][O3 ] − k a k b [O3 ]2 + k a k a′ [O3 ][O2 ] − k a k b [O3 ]2
=
k a′ [O2 ] + k b [O3 ]
−2k a k b [O3 ]2
= ′
k a [O2 ] + k b [O3 ]
If step 3 is rate limiting, such that k a′ [O2 ][O] ≫ k b [O][O3 ], and hence k a′ [O2 ] ≫
k b [O3 ], the denominator simplifies to k a′ [O2 ] and hence
d[O3 ] −2k a k b [O3 ]2

=
dt k a′ [O2 ]
As required, the rate of decomposition of O3 is second order in O3 and order
−1 in O2 .
E14E.2(a) The overall activation energy for a reaction consisting of a pre-equilibrium
followed by a rate-limiting elementary step is given by [14E.13–567], E a = E a,a +
E a,b − E a,a′ , where E a,a and E a,a′ are the forward and reverse activation ener-
gies for the pre-equilibrium and E a,b is the activation energy for the following
elementary step. In this case
E a = (25 kJ mol−1 ) + (10 kJ mol−1 ) − (38 kJ mol−1 ) = −3 kJ mol−1

As explained in Section 14E.5 on page 566, negative activation energies such as
this are possible for composite reactions.
E14E.3(a) (i) A pre-equilibrium A2 ⇌ 2A between A2 and A is described by the equi-
librium constant K given by
([A]/c −○ )2 [A]2
K= = hence [A] = (Kc −○ [A2 ])1/2
([A2 ]/c −○ ) [A2 ]c −○
The equilibrium constant K is written in terms of rate constants using
[14C.8–554], K = (k a /k a′ ) × (k b /k b′ ) × .... However, in order to make
K dimensionless it is necessary in this case to include a factor of 1/c −○
because k a is a first-order rate constant with units s−1 while k a′ is a second-
order rate constant with units dm3 mol−1 s−1 . Thus K = k a /k a′ c −○ , which,
on substituting into the above expression for [A] yields
1/2 1/2
ka ka
[A] = ( ′ −○ c −○ [A2 ]) = ( ′ [A2 ])
ka c ka
This expression is alternatively obtained by noting that at equilibrium the
rates of the forward and reverse reactions are the same (provided that the
step to P can be ignored)
1/2
ka
k a [A2 ] = k a′ [A]2 hence [A] = ( [A2 ])
k a′
The rate of formation of P is given by d[P]/dt = k b [A][B]; substituting

the above expression for [A] into this gives
1/2 1/2
d[P] ka ka
= k b [A][B] = k b ( ′ ) [A2 ][B] = k b ( ′ ) [A2 ]1/2 [B]
dt ka ka
(ii) The net rate of change in the concentration of A is
d[A]
= 2k a [A2 ] − 2k a′ [A]2 − k b [A][B]
dt
In the steady-state approximation this is assumed to be zero
2k a [A2 ] − 2k a′ [A]2 − k b [A][B] = 0
Hence 2k a′ [A]2 + k b [B][A] − 2k a [A2 ] = 0. This is a quadratic equation in

[A], for which the solution is
1/2
−k b [B] + (k b2 [B]2 + 16k a′ k a [A2 ])
hence [A] =
4k a′
where the positive square root is chosen in order to avoid obtaining a
negative value for [A]. The rate of formation of P is given by d[P]/dt =
k b [A][B]; substituting the above expression for [A] into this gives
1/2
d[P] −k b [B] + (k b2 [B]2 − 16k a′ k a [A2 ])
= k b [A][B] = k b [B] ×
dt 4k a′
⎡ 1/2 ⎤
k b [B] ⎢⎢ 16k a′ k a [A2 ] ⎥
= ⎢−k b [B] + k b [B] (1 + ) ⎥
⎥
4k a′ ⎢ k b [B]2
2
⎥
⎣ ⎦
2 ⎡ ′ 1/2 ⎤
k [B] ⎢
2
16k a k a [A2 ] ⎥
= b ′ ⎢⎢−1 + (1 + ) ⎥⎥
4k a ⎢ k b [B]
2 2
⎥
⎣ ⎦
Under certain circumstances this rate law simplifies. If 16k a′ k a [A2 ]/k b2 [B]2 ≫
1 then
⎡ ′ 1/2 ⎤
d[P] k b2 [B]2 ⎢−1 + ( 16k a k a [A2 ] ) ⎥
⎢ ⎥
≈ ⎢ ⎥
dt 4k a′ ⎢ k b [B]
2 2
⎥
⎣ ⎦
1/2 1/2
k b2 [B]2 16k a′ k a [A2 ] ka
≈ ′
×( ) = kb ( ) [A2 ]1/2 [B]
4k a k b2 [B]2 k a′
which is the same as the rate law derived in part (i) assuming a pre-
equilibrium. The condition 16k a′ k a [A2 ]/k b2 [B]2 ≫ 1 corresponds to the
A2 ⇄ A + A steps being much faster than the step involving B and k b ;
this is precisely the situation corresponding to a pre-equilibrium because
the removal of A in the reaction with B is then too slow to affect the
maintenance of the equilibrium.
On the other hand, if 16k a′ k a [A2 ]/k b2 [B]2 ≪ 1 then the square root is
approximated by the expansion (1 + x)1/2 ≈ 1 + 21 x to give
d[P] k b2 [B]2 1 16k a′ k a [A2 ] k b2 [B]2 8k a′ k a [A2 ]

≈ [−1 + (1 + × )] = ×
dt 4k a′ 2
k b2 [B]2 4k a′ k b2 [B]2
= 2k a [A2 ]
This rate law corresponds to the step A2 → A + A being rate-determining:

once A has formed from A2 in this step it immediately goes on to form
product. The factor of 2 arises because each molecule of A2 that dis-
sociates forms two A molecules which react with two B molecules to
form two molecules of product. Hence the rate of product formation is
twice the rate of A2 dissociation. This situation does not correspond to a
pre-equilibrium because the immediate removal of A by its reaction with
B does not allow A2 and A to come to equilibrium.
ka kb
P14E.1 The concentration of I in the reaction mechanism A → I → P is given by
[14E.4b–563],
ka
[I] = (e−k a t − e−k b t ) [A]0
kb − ka
This expression is plotted in Fig. 14.11 for [A]0 = 1 mol dm−3 , k b = 1 s−1 , and
various values of k a . The line for k a = 10 s−1 corresponds to part (a) of the
question.
0.8 ka = 10 s−1
ka = 2 s−1
ka = 0.5 s−1
0.6 ka = 0.2 s−1
[I]/mol dm−3
ka = 0.1 s−1
0.4 ka = 0.02 s−1
0.2
0.0
0 1 2 3 4 5
t/s
Figure 14.11
If k b ≫ k a , the concentration of I remains low and, apart from the initial induc-
tion period, approximately constant during the reaction. Thus the steady-state
approximation that d[I]/dt = 0 becomes increasingly valid as the ratio k b /k a

increases.
P14E.3 It is shown in Example 14E.1 on page 564 that for the case of two consecutive
unimolecular reactions the concentration of the intermediate is greatest at a
time given by t max = (ln[k a /k b ])/(k a − k b ). The half-life of a first-order reac-
tion is related to the rate constant according to [14B.3–548], t 1/2 = ln 2/k r . This
is rearranged to k r = ln 2/t 1/2 and used to substitute for the rate constants in
the expression for t max .
1 ka 1 ln 2/t 1/2,a
t max = ln = ln
k a − k b k b (ln 2/t 1/2,a ) − (ln 2/t 1/2,b ) ln 2/t 1/2,b
1 t 1/2,b
= ln
ln 2 [1/t 1/2,a − 1/t 1/2,b ] t 1/2,a
1 33.0 d
Hence t 1/2 = × ln = 39.1 d
ln 2 × [1/(22.5 d) − 1/(33.0 d)] 22.5 d
ka kb kc
P14E.5 For the scheme A ⇄ B ⇄′ C ⇄ D the rates of change of the intermediates B and
k a′ kb k c′
C are
d[B] d[C]
= k a [A]−k a′ [B]−k b [B]+k b′ [C] = k b [B]−k b′ [C]−k c [C]+k c′ [D]
dt dt
In the steady-state approximation, both of these expressions are equal to zero.
Furthermore, because D is removed as soon as it is formed, [D] = 0 and so the
expression for d[C]/dt becomes
(k b′ + k c )[C]
k b [B] − k b′ [C] − k c [C] = 0 hence [B] =
kb
The expression for d[B]/dt becomes
k a [A] − (k a′ + k b )[B] + k b′ [C] = 0

(k b′ + k c )[C]
hence k a [A] − (k a′ + k b ) + k b′ [C] = 0
kb
(k ′ + k b )(k b′ + k c ) − k b k b′
hence ( a ) [C] = k a [A]
kb
k a k b [A] k a k b [A]
hence [C] = ′ ′ ′
= ′ ′
(k a + k b )(k b + k c ) − k b k b k a k b + k a′ k c + k b k c
where on the second line the expression for [B] derived above is substituted in.
Finally, the rate of formation of D is
d[D] k a k b k c [A]
= k c [C] − k c′ [D] = k c [C] = ′ ′
dt k a k b + k a′ k c + k b k c
where [D] = 0 is used and the expression for [C] is substituted in.
P14E.7 The equilibrium constants for the two pre-equilibria are

[(HCl)2 ]c −○ K 1 [HCl]2
K1 = hence [(HCl)2 ] =
[HCl] 2 c −○
[complex]c −○ K 2 [HCl][CH3 CH−CH2 ]
K2 = hence [complex] =
[HCl][CH3 CH−CH2 ] c −○
The factors of c −○ are needed to make K 1 and K 2 dimensionless. The rate of
product formation is
d[CH3 CHClCH3 ]
υ= = k r [(HCl)2 ][complex]
dt
K 1 [HCl]2 K 2 [HCl][CH3 CH−CH2 ]
= kr × ×
c −○ c −○
kr K1 K2
= [HCl]3 [CH3 CH−CH2 ]
c −○ 2
Thus the reaction is predicted to be first-order in propene and third-order in
HCl, as required.
P14E.9 Applying the steady-state approximation to the intermediates OF and F gives
d[OF]
= k a [F2 O]2 + k b [F][F2 O] − 2k c [OF]2 = 0
dt
d[F]
= k a [F2 O]2 − k b [F][F2 O] + 2k c [OF]2 − 2k d [F]2 [F2 O] = 0
dt
On adding together these two equations the k b and k c terms cancel to give
2k a [F2 O]2 − 2k d [F]2 [F2 O] = 0 hence k a [F2 O] = k d [F]2
1/2
ka
hence [F] = ( [F2 O])
kd
Steps a and b lead to the net consumption of one F2 O, while steps d and c lead
to no net change. The rate of consumption of F2 O is therefore
1/2
d[F2 O] ka
− = k a [F2 O]2 + k b [F2 O][F] = k a [F2 O]2 + k b [F2 O] ( [F2 O])
dt kd
k a 1/2
= k a [F2 O]2 + k b ( ) [F2 O]3/2
® kd
kr ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶
k r′
√
which is the required expression with k r = k a and k r′ = k b k a /k d .
14F Examples of reaction mechanisms

D14F.1 In the analysis of stepwise polymerization, the rate constant for the second-
order condensation is assumed to be independent of the chain length and to
remain constant throughout the reaction. It follows, then, that the degree of
polymerization is given by [14F.12b–572], ⟨N⟩ = 1 + k r t[A]0 . Therefore, the
average molar mass can be controlled by adjusting the initial concentration of
monomer and the length of time that the polymerization is allowed to proceed.
As discussed in Section 14F.2(b) on page 572, chain polymerization involves

initiation, propagation, and termination steps. The derivation of the overall
rate equation utilizes the steady-state approximation and leads to the following
expression for the average number of monomer units in the polymer chain
([14F.15–573]) ⟨N⟩ = 2k r [M][In]−1/2 where k r = k p (4 f k i k t )−1/2 , and where
k p , k i , and k t are the rate constants for the propagation, initiation, and ter-
mination steps respectively, and f is the fraction of radicals that successfully
initiate a chain. It is seen that the average molar mass of the polymer is directly
proportional to the monomer concentration, and inversely proportional to the
square root of the initiator concentration, and to the rate constant for initiation.
Therefore, the slower the initiation of the chain, the higher the average molar
mass of the polymer.
D14F.3 As temperature increases, the rate of an enzyme-catalyzed reaction is expected

to increase. However, at a sufficiently high temperature the enzyme denatures
and a decrease in the reaction rate is observed. Temperature related denatura-
tion is caused by the action of vigorous vibrational motion, which destroys sec-
ondary and tertiary protein structure. Electrostatic, internal hydrogen bond-
ing, and van der Waals interactions that hold the protein in its active, folded
shape are broken with the protein unfolding into a random coil. The active site
and enzymatic activity is lost.
The rate of a particular enzyme-catalyzed reaction may also appear to decrease

at high temperature in the special case in which an alternative substrate re-
action, which has a relatively slow rate at low temperature, has the faster rate
increase with increasing temperature. A temperature may be reached at which
the alternative reaction predominates.
E14F.1(a) The Michaelis–Menten equation for the rate of an enzyme-catalysed reaction
is given by [14F.18a–575], υ = υ max /(1 + K M /[S]0 ). Rearranging for υ max gives
KM 0.046 mol dm−3

υ max = υ (1 + ) = (1.04 mmol dm−3 s−1 ) × (1 + )
[S]0 0.105 mol dm−3
= 1.50 mmol dm−3 s−1
E14F.2(a) Example 14F.2 on page 576 gives the values K M = 10.0 mmol dm−3 and υ max =
0.250 mmol dm−3 s−1 for an enzyme concentration of [E]0 = 2.3 nmol dm−3 .
The catalytic efficiency is defined in the exercise as k b /K M , and υ max is related to
k b according to [14F.17b–575], υ max = k b [E]0 , hence k b = υ max /[E]0 . Therefore,
the catalytic efficiency is
kb υ max 0.250 × 10−3 mol dm−3 s−1

= =
K M K M [E]0 (10.0 × 10−3 mol dm−3 ) × (2.3 × 10−9 mol dm−3 )
= 1.1 × 107 dm3 mol−1 s−1
E14F.3(a) The effective rate constant for the Lindemann–Hinshelwood mechanism is given
by [14F.8–570], 1/k r = k a′ /k a k b + 1/k a [A]. The difference between the effective
rate constant at two pressures is therefore
1 1 1 1 1 1/[A]2 − 1/[A]1
− = ( − ) hence ka =
k r,2 k r,1 k a [A]2 [A]1 1/k r,2 − 1/k r,1
The rate constant for the activation step, k a , is therefore
1/(12 Pa) − 1/(1.30 × 103 Pa)
ka = = 1.9 × 10−6 Pa−1 s−1
1/(2.10 × 10−5 s−1 ) − 1/(2.50 × 10−4 s−1 )
or 1.9 MPa−1 s−1 .

E14F.4(a) The fraction of condensed groups at time t of a stepwise polymerisation is given
by [14F.11–571], p = k r t[A]0 /(1 + k r t[A]0 ). Hence, after 5.00 h, or 5.00 h ×
(602 s)/(1 h) = 1.80 × 104 s,
k r t[A]0
p=
1 + k r t[A]0
(1.39 dm3 mol−1 s−1 ) × (1.80 × 104 s) × (10.0 × 10−3 mol dm−3 )
=
1 + (1.39 dm3 mol−1 s−1 ) × (1.80 × 104 s) × (10.0 × 10−3 mol dm−3 )
= 0.996... = 0.996
The degree of polymerisation in a stepwise polymerisation is given by [14F.12a–
571], ⟨N⟩ = 1/(1 − p).
1 1
⟨N⟩ = = = 251
1 − p 1 − 0.996...
E14F.5(a) The kinetic chain length in a chain polymerisation reaction is given by [14F.14c–
573], λ = k r [M][In]−1/2 . The ratio of chain length under the two different sets
of conditions is therefore
−1/2 −1/2
λ 2 k r [M]2 [In]2 [M]2 [In]2 1
= =( )×( ) = × 3.6−1/2 = 0.13
λ 1 k r [M]1 [In]−1/2 [M]1 [In]1 4.2
1
P14F.1 The effective rate constant k r in the Lindemann–Hinshelwood mechanism is
given by [14F.8–570], 1/k r = k a′ /k a k b + 1/k a [A]. This expression implies that
a plot of 1/k r against 1/[A] should be a straight line. The data are plotted in
Fig. 14.12, using pressure as a measure of concentration.
p/Torr 104 k r /s−1 1/(p/Torr) 1/(104 k r /s−1 )

84.1 2.98 0.011 9 0.336
11.0 2.23 0.090 9 0.448
2.89 1.54 0.346 0.649
0.569 0.857 1.76 1.167
0.120 0.392 8.33 2.551
0.067 0.303 14.9 3.300
3
1/(104 k r /s−1 )
0
0 5 10 15
1/(p/Torr)
Figure 14.12
The data lie on a curve rather than on a straight line, so it is concluded that the
Lindemann–Hinshelwood mechanism does not fit these data.
P14F.3 Each molecule of hydroxyacid has one OH group and one COOH group (A),
so [OH] = [A]. Hence the given rate expression, d[A]/dt = −k r [A]2 [OH],
becomes
d[A] 1
= −k r [A]3 hence − d[A] = k r dt
dt [A]3
Integration of this expression, with the limits that the concentration is [A]0 at
time t = 0 and [A] at some later time t, gives
[A]
[A] 1 t 1 t
∫ − d[A] = ∫ k r dt hence ∣ = k r t∣0
[A]0 [A]3 0 2[A]2 [A]0
1 1
hence − = 2k r t
[A] 2 [A]20
Rearranging gives
1 1 1 [A]20
= 2k r t + hence [A]2 = =
[A]2 [A]20 2k r t + 1/[A]20 2k r t[A]20 + 1
To go to the final expression the top and bottom of the fraction are multiplied by
[A]20 . Taking the square root gives [A] = [A]0 /(2k r t[A]20 + 1)1/2 . As explained
in Section 14F.2(a) on page 571, the degree of polymerisation ⟨N⟩ is the ratio of
the initial concentration of A, [A]0 , to the concentration of end groups, [A], at
the time of interest. Hence
[A]0 [A]0
⟨N⟩ = = = (2k r t[A]20 + 1)1/2
[A] [A]0 /(2k r t[A]20 + 1)1/2
P14F.5 The Michaelis–Menten equation [14F.18a–575] is

υ max
υ=
1 + K M /[S]0
This equation is plotted for fixed υ max with a range of K M values in Fig. 14.13,
and for fixed K M with a range of υ max values in Fig. 14.14.
1.0
K M = 0.01 mol dm−3
0.8
K M = 0.1 mol dm−3
υ/mmol dm−3 s−1
0.6
0.4 K M = 0.5 mol dm−3
0.2
υ max = 1 mmol dm−3 s−1
0.0
0.0 0.2 0.4 0.6 0.8 1.0
−3
[S]0 /mol dm
Figure 14.13
P14F.7 The Lineweaver-Burk equation, [14F.18b–575], expresses the reciprocal of the

velocity as 1/υ = 1/υ max + (K M /υ max )(1/[S]0 ). This expression implies that a
plot of 1/υ against 1/[S]0 will be a straight line of slope K M /υ max and intercept
1/υ max . Such a plot is shown in Fig. 14.15.
[ATP]/ υ/ 1 1
µmol dm−3 µmol dm−3 s−1 [ATP]/(µmol dm−3 ) υ/(µmol dm−3 s−1 )
0.6 0.81 1.67 1.23
0.8 0.97 1.25 1.03
1.4 1.30 0.71 0.77
2.0 1.47 0.50 0.68
3.0 1.69 0.33 0.59
1.0
K M = 0.1 mol dm−3
0.8
υ/mmol dm−3 s−1

υ max = 1 mmol dm−3 s−1
0.6
0.4
υ max = 0.5 mmol dm−3 s−1
0.2
υ max = 0.1 mmol dm−3 s−1
0.0
0.0 0.2 0.4 0.6 0.8 1.0
−3
[S]0 /mol dm
Figure 14.14
1/(υ/µmol dm−3 s−1 )
1.0
0.5
0.0 0.5 1.0 1.5

−3
1/([ATP]/µmol dm )
Figure 14.15
1/(υ/µmol dm−3 s−1 ) = 0.48 × 1/([ATP]/µmol dm−3 ) + 0.43
The intercept is identified with 1/υ max so that
1
υ max = = 2.32... µmol dm−3 s−1 = 2.3 µmol dm−3 s−1
0.43 µmol dm−3 s−1
The slope is identified with K M /υ max so that
slope υ max
¬ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
K M = ( 0.48 s) × (2.32... µmol dm−3 s−1 ) = 1.1 µmol dm−3
14G Photochemistry
D14G.1 The time scales of atomic processes are rapid indeed. Note that the times given
here are in some way typical values for times that may vary over two or three
orders of magnitude. For example, vibrational wavenumbers can range from
about 4400 cm−1 (for H2 ) to 100 cm−1 (for I2 ) and even lower, with a corre-
sponding range of associated times. Radiative decay rates of electronic states
can vary even more widely: times associated with phosphorescence can be in
the millisecond and even second range. A large number of time scales for phys-
ical, chemical, and biological processes on the atomic and molecular scale are
reported in Figure 2 of A. H. Zewail, ‘Femtochemistry: Atomic-Scale Dynamics
of the Chemical Bond’, Journal of Physical Chemistry A, 104, 5660 (2000).
Radiative decay of excited electronic states can range from about 10−9 s to
10−4 s, and even longer for phosphorescence involving ‘forbidden’ decay paths.
Molecular rotational motion takes place on a scale of 10−12 s to 10−9 s. Molecu-
lar vibrations are faster still, about 10−14 s to 10−12 s. Proton transfer reactions
occur on a timescale of about 10−10 s to 10−9 s, although protons can hop from
molecule to molecule in water even more rapidly (1.5 × 10−12 s).
Harvesting of light during plant photosynthesis involves very fast time scales
of several energy-transfer and electron-transfer steps in photosynthesis. Initial
energy transfer (to a nearby pigment) has a time scale of around 10−13 s to
5 × 10−12 s, with longer-range transfer (to the reaction centre) taking about
10−10 s. Immediate electron transfer is also very fast (about 3 ps), with ultimate
transfer (leading to oxidation of water and reduction of plastoquinone) taking
from 10−10 s to 10−3 s. The mean time between collisions in liquids is similar
to vibrational periods, around 10−13 s.
E14G.1(a) The efficiency of resonance energy transfer η T is defined by [14G.6–583], η T =
1 − ϕ F /ϕ F,0 , and the distance-dependence of the efficiency is given by [14G.7–
583], η T = R 06 /(R 06 + R 6 ), where R is the donor–acceptor distance and R 0 is a
constant characteristic of the particular donor–acceptor pair.
In this case a decrease of the fluorescence quantum yield by 10% implies that
ϕ F = 0.9ϕ F,0 . Hence the efficiency is η T = 1 − ϕ F /ϕ F,0 = 1 − 0.9 = 0.1.
Rearranging [14G.7–583] for R, and taking R 0 = 4.9 nm from Table 14G.3 on
page 583, gives
1/6
1 − ηT 1 − 0.1 1/6
R = R0 ( ) = (4.9 nm) × ( ) = 7.1 nm
ηT 0.1
E14G.2(a) The primary quantum yield is defined by [14G.1a–579], ϕ = N events /N abs . In

this equation N abs is the number of photons absorbed and N events is, in this
case, the number of molecules of A that decompose, N decomposed . Rearranging
gives
N decomposed n decomposed N A (n formed /2)N A
N abs = = =
ϕ ϕ ϕ
In the final expression, n formed is the amount in moles of B that is formed. The
stoichiometry of the reaction A → 2B + C implies that the amount of A that
decomposes is half the amount of B that is formed, n decomposed = n formed /2.
The quantum yield is 210 mmol einstein−1 , or 0.210 mol mol−1 = 0.210, hence
(n formed /2)N A (2.28 × 10−3 mol)/2 × (6.0221 × 1023 mol−1 )

N abs = =
ϕ 0.210
= 3.27 × 1021
E14G.3(a) The fluorescence quantum yield is given by [14G.4–581], ϕ F,0 = k F τ 0 . The

observed lifetime τ 0 is given by [14G.3b–580], τ 0 = 1/(k F + k ISC + k IC ), which
is written as τ 0 = 1/k r where k r = k F + k ISC + k IC is the effective first-order
rate constant for the decay of the excited state of the fluorescing species. For a
first-order process k r is related to the half-life according to [14B.3–548], t 1/2 =
ln 2/k r , and combining this expression with τ 0 = 1/k r gives t 1/2 = ln 2/(1/τ 0 ) =
(ln 2)τ 0 . Hence τ 0 = t 1/2 / ln 2.
Rearranging [14G.4–581] then gives
ϕ F,0 ϕ F,0 ϕ F,0 ln 2 0.35 × ln 2

kF = = = = = 4.3 × 107 s−1
τ0 t 1/2 / ln 2 t 1/2 5.6 × 10−9 s
E14G.4(a) The Stern–Volmer equation [14G.5–581] is ϕ F,0 /ϕ F = 1 + τ 0 k Q [Q], where ϕ F

and ϕ F,0 are the fluorescence quantum yields with and without the quencher.
The rate of fluorescence υ, and hence the fluorescence intensity, is directly pro-
portional to the fluorescence quantum yield according to [14G.1b–579], ϕ =
υ/I abs . Therefore to reduce the fluorescence intensity to 50% of the unquenched
value requires ϕ F = 12 ϕ F,0 and hence ϕ F,0 /ϕ F = 2. Rearranging the Stern–
Volmer equation then gives
ϕ F,0 /ϕ F − 1 2−1
[Q] = =
τ0 kQ (6.0 × 10 s)×(3.0 × 108 dm3 mol−1 s−1 )
−9
= 0.56 mol dm−3
P14G.1 The quantum yield is given by [14G.1a–579], ϕ = N events /N abs where N abs is
the number of photons absorbed and N events is, in this case, the number of
molecules of the absorbing substance that decomposed. The latter is equal
to n decomposed N A , where n decomposed is the amount in moles of substance that
decomposed.
The number of photons absorbed is found by noting that the energy transferred
by each photon is given by [7A.9–227], ∆E = hν = hc/λ. Therefore the total
energy absorbed is E abs = N abs hc/λ. This energy is also given by E abs = f Pt,
where P is the incident power, t is the time of exposure, and f is the fraction
of incident radiation that is absorbed. In this case f = 1 − 0.257 = 0.743.
Combining these expressions gives
N abs hc f Ptλ
f Pt = hence N abs =
λ hc
This is substituted into ϕ = N events /N abs , together with N events = n decomposed N A ,
to give
N events n decomposed N A n decomposed N A hc

ϕ= = =
N abs f Ptλ/hc f Ptλ
(0.324 mol)×(6.0221×1023 mol−1 )
=
(0.743) × (87.5 W) × (28.0 min) × (60 s)/(1 min)
(6.6261×10−34 J s)×(2.9979×108 m s−1 )
× = 1.11
(320 × 10−9 m)
Note that 1 W = 1 J s−1 . The fact that the quantum yield is greater than 1
indicates that each absorbed photon can lead to the decomposition of more
than one molecule of absorbing material.
P14G.3 (a) The concentration of the excited dansyl chloride decays with time accord-
ing to [14G.3a–580], [S∗ ] = [S∗ ]0 e−t/τ 0 , or [S∗ ]/[S∗ ]0 = e−t/τ 0 . The rate
of fluorescence is given by υ = k F [S∗ ] so the rate of fluorescence, and
hence the fluorescence intensity I F , is proportional to [S∗ ]. Therefore
I F /I 0 = [S∗ ]/[S∗ ]0 , and hence
IF [S∗ ] IF t
= ∗ = e−t/τ 0 hence ln ( )=−
I 0 [S ]0 I0 τ0
This expression implies that a plot of ln(I F /I 0 ) against t should be a straight

line of slope −1/τ 0 and intercept zero. The data are plotted in Fig. 14.16.
t/ns I F /I 0 ln(I F /I 0 )
5.0 0.45 −0.799
10.0 0.21 −1.561
15.0 0.11 −2.207
20.0 0.05 −2.996
The data lie on a good straight line that passes close to the origin. The
equation of the line is
ln(I F /I 0 ) = −0.145 × (t/ns) − 0.081
Identifying the slope with −1/τ 0 gives
1 1
− = −0.145 ns−1 hence τ0 = = 6.89... ns = 6.9 ns
τ0 0.145 ns−1
−1
ln(I F /I 0 )
−2
−3
0 5 10 15 20 25
t/ns
Figure 14.16
(b) The fluorescence quantum yield is given by [14G.4–581], ϕ F,0 = k F τ 0 . This

equation is rearranged for k F
ϕ F,0 0.70
kF = = = 1.0 × 108 s−1
τ0 6.89... × 10−9 s
P14G.5 The Stern–Volmer equation [14G.5–581] is ϕ F,0 /ϕ F = 1+τ 0 k Q [Q]. As explained

in Section 14G.4 on page 581, the ratio τ 0 /τ, where τ is the lifetime in the
presence of the quencher, is equal to ϕ F,0 /ϕ F , so the Stern–Volmer equation
becomes
τ0 τ 0 /τ − 1
= 1 + τ 0 k Q [Q] hence kQ =
τ τ 0 [Q]
In order to use the equation to calculate k Q it is necessary to find τ 0 and τ. This

is done as follows.
The concentration of an excited species such as Hg∗ varies with time according
to [14G.3a–580], [Hg∗ ] = [Hg∗ ]0 e−t/τ . Rearranging and taking logarithms
gives
[Hg∗ ] [Hg∗ ] t
= e−t/τ hence ln ( )=−
[Hg∗ ]0 [Hg∗ ]0 τ
The rate of fluorescence is υ = k F [Hg∗ ], so the fluorescence intensity I is pro-

portional to the Hg∗ concentration. Hence I/I 0 = [Hg∗ ]/[Hg∗ ]0 , and from the
above equation a plot of ln(I/I 0 ) against t should therefore be a straight line of
slope −1/τ and intercept zero. The fluorescence intensity data are given relative
to the value at t = 0 and therefore represent I/I 0 .

p N2 = 0 p N2 = 9.74 × 10−4 atm

t/µs I/I 0 ln(I/I 0 ) t/µs I/I 0 ln(I/I 0 )
0.0 1.000 0.000 0.0 1.000 0.000
5.0 0.606 −0.501 3.0 0.585 −0.536
10.0 0.360 −1.022 6.0 0.342 −1.073
15.0 0.220 −1.514 9.0 0.200 −1.609
20.0 0.135 −2.002 12.0 0.117 −2.146
0
p N2 = 0
0 p N2 = 9.74 × 10−4 atm
ln(I/I 0 )
ln(I/I 0 )
−1 −1
−2 −2
0 10 20 0 10 20
t/µs t/µs
Figure 14.17
The data fall on good straight lines, the equations of which are
p N2 = 0 ln(I/I 0 ) = −0.100 × (t/µs) + −4.18 × 10−3
p N2 = 9.74 × 10−4 atm ln(I/I 0 ) = −0.179 × (t/µs) + 7.02 × 10−5
The slopes are identified with −1/τ 0 and −1/τ respectively
1 1
τ0 = − = 10.0... µs τ=− = 5.59 ... µs
−0.100 µs−1 −0.179 µs−1
The rearranged form of the Stern–Volmer equation found earlier, k Q = (τ 0 /τ −
1)/τ 0 [Q] is then used to calculate k Q . The concentration of the N2 quencher
is calculated from the partial pressure using the perfect gas equation [1A.4–8],
pV = nRT.
n N2 p N2
[N2 ] = =
V RT
9.74 × 10−4 atm 1.01325 × 105 Pa 1 m3
= × ×
(8.3145 J K−1 mol−1 ) × (300 K) 1 atm 103 dm3
= 3.95... × 10−5 mol dm−3
where 1 Pa = 1 kg m−1 s−2 and 1 J = 1 kg m2 s−2 are used. Hence

τ 0 /τ − 1 (10.0... × 10−6 s)/(5.59... × 10−6 s) − 1
kQ = =
τ 0 [N2 ] (10.0 × 10−6 s) × (3.95... × 10−5 mol dm−3 )
= 2.00 × 109 dm3 mol−1 s−1
P14G.7 The efficiency of resonance energy transfer is given by [14G.6–583],
ϕF τ
ηT = 1 − =1−
ϕ F,0 τ0
where the second expression comes from the fact that, according to [14G.4–
581], ϕ F,0 = k F τ 0 , the lifetime is proportional to quantum yield. The efficiency
of resonance energy transfer in terms of donor–acceptor distance is given by
[14G.7–583], η T = R 06 /(R 06 + R 6 ). Equating the two expressions for η T gives
1/6
τ R6 1 R6 1
1− = 6 0 6 hence = 1 + 6 hence R = R 0 ( − 1)
τ0 R0 + R 1 − τ/τ 0 R0 1 − τ/τ 0
The distance required to give τ = 10 ps is therefore

1/6
1
R = (5.6 nm) × ( − 1) = 2.6 nm
1 − (10 × 10−12 s)/(1 × 10−9 s)

I14.1 The rate of the forward and backward steps are
d[B] d[B]
A→B = IA B→A = −k r [B]2
dt dt
The overall rate of change of [B] is therefore d[B]/dt = I A − k r [B]2 . In the

steady state, d[B]/dt = 0, hence
I a 1/2
k r [B]2 = I a hence [B] = ( )
kr
This concentration can differ significantly from an equilibrium distribution

because changing the illumination may change the rate of the forward reaction
without affecting the reverse reaction. This is in contrast to corresponding
equilibrium expression, in which the ratio [B]/[A] depends on a ratio of rate
constants for the forward and reverse reactions as explained in Section 14C.1
on page 553.
I14.3 (a) The expressions [A] = [A]0 − x and [P] = [P]0 + x are substituted into
the rate law to give
d[A]
υ=− = k r [A][P] = k r ([A]0 − x)([P]0 + x)
dt
The expression [A] = [A]0 − x implies that d[A]/dt = −dx/dt so the
expression becomes
dx dx
= k r ([A]0 − x)([P]0 + x) hence = k r dt
dt ([A]0 − x)([P0 ] + x)
Integration of this expression, using x = 0 at time t = 0 gives

x dx t
∫ = ∫ k r dt
0 ([A]0 − x)([P0 ] + x) 0
The left-hand side is evaluated using Integral A.3

Integral A.3 with A = [A]0 and B = −[P]0
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
x dx x dx
∫ = −∫
0 ([A]0 − x)([P0 ] + x) 0 ([A]0 − x)(−[P0 ] − x)
1 (−[P]0 − x)[A]0
=− ln ( )
(−[P]0 ) − [A]0 ([A]0 − x)(−[P]0 )
1 [A]0 ([P]0 + x)
= ln ( )
[A]0 + [P]0 [P]0 ([A]0 − x)
The right-hand side is k r t, hence the integrated rate law is
1 [A]0 ([P]0 + x)
ln ( ) = kr t
[A]0 + [P]0 [P]0 ([A]0 − x)
The expression [P] = [P]0 + x is rearranged to x = [P] − [P]0 . This is used

to replace x in the integrated rate law
1 [A]0 [P]
ln ( ) = kr t
[A]0 + [P]0 [P]0 ([A]0 − [P] + [P]0 )
a
[A]0 [P] ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
hence ln ( ) = ([A]0 + [P]0 )k r t
[P]0 ([A]0 − [P] + [P]0 )
hence [A]0 [P] = [P]0 ([A]0 − [P] + [P]0 )e at
hence [P]([A]0 + [P]0 e at ) = [P]0 ([A]0 + [P]0 )e at

b
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
[P] ([A]0 + [P]0 )e at (1 + [P]0 /[A]0 )e at (1 + b)e at
hence = = =
[P]0 [A]0 + [P]0 e at 1 + ([P]0 /[A]0 ) e at 1 + be at
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
b
(b) The quantity [P]/[P]0 is plotted against at in Fig. 14.18 for various values
of b.
The plots are sigmoid in shape because the reaction is initially slow be-
cause only a small amount of P is present. As more product is formed, the
rate of the reaction υ = k r [A][P] increases and the curve becomes steeper,
until the reaction slows down towards the end due to the reactant A being
used up.
The curves level off at different values because [P]/[P]0 is being plotted.
In each case the final concentration of P is given by the initial concentra-
tion of A, because all the A is eventually converted to P, plus the concen-
tration of P that was present at the start, that is, [P]∞ = [A]0 + [P]0 . The
100
b = 0.01
[P]/[P]0
b = 0.02
50
b = 0.1
0
0 2 4 6 8 10
at
Figure 14.18
final value of [P]/[P]0 is therefore
[P]∞ [A]0 + [P]0 [A]0 1

= = +1= +1
[P]0 [P]0 [P]0 b
Changing b therefore changes the final value of [P]/[P]0 .

The integrated rate equation for a first-order process A → P is given in
Table 14B.3 on page 551 as [P]/[A]0 = 1 − e−k r t . In order to facilitate
comparison to the autocatalytic reaction it is instructive to re-plot the au-
tocatalytic curves as [P]/[A]0 rather than as [P]/[P]0 . Furthermore it is
convenient to consider ([P]−[P]0 )/[A]0 rather than [P]/[A]0 , because in
this way the plot reflects the amount of P that is produced in the reaction
rather than including any P that was present at the start. The expression
for [P]/[P]0 derived above is adapted to give ([P] − [P]0 )/[A]0
[P] (1 + b)e at (1 + b)e at [P]0

= hence [P] =
[P]0 1 + be at 1 + be at
Hence
b
b ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
[P] − [P]0 [P] ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹·¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ (1 + b)e at [P]0 /[A]0
= − [P]0 /[A]0 = −b
[A]0 [A]0 1 + be at
b(1 + b)e at
= −b
1 + be at
This expression is plotted against t in Fig. 14.19 for various values of b, tak-
ing a = 1 s−1 in each case. The quantity [P]/[A]0 = 1−e−k r t is also plotted,
taking k r = 1 s−1 . As already noted, the autocatalytic curves are sigmoid,
in contrast to the first-order curve which is not. The autocatalytic curves
with larger b, that is a greater initial amount of P relative to the initial
amount of A, reach their maximum value faster than those with smaller
b. This is because, if less P is present to begin with, the autocatalytic step
is initially slower and the amount of P present builds up more slowly.
1.0 first-order
b = 0.1
([P] − [P]0 )/[A]0

b = 0.03
b = 0.01
0.5 b = 0.002
0.0
0 2 4 6 8 10
t/s
Figure 14.19
(c) The rate law found in part (a), [P]/[P]0 = (1 + b)e at /(1 + be at ), is rear-
ranged to [P] = (1 + b)e at [P]0 /(1 + be at ) and differentiated to give an
expression for the rate.
d[P] d (1 + b)e at [P]0
υ= = ( )
dt dt 1 + be at
(1 + be at ) × a(1 + b)e at [P]0 − (1 + b)e at [P]0 × abe at
=
(1 + be at )2
a(1 + b)e at [P]0
=
(1 + be at )2
The maximum rate is found by differentiating υ with respect to t and
setting the derivative equal to zero
dυ (1 + be at )2 × (a 2 (1 + b)e at [P]0 ) − a(1 + b)e at [P]0 × 2abe at (1 + be at )
=
dt (1 + be at )4
At the maximum, when dυ/dt = 0, the numerator of this expression is
zero
a 2 (1 + b)(1 + be at )2 e at [P]0 − 2a 2 b(1 + b)(1 + be at )e2at [P]0 = 0
Cancelling of terms followed by rearrangement gives
1 + be at − 2be at = 0 hence be at = 1 hence t = −(1/a) ln b
(d) As in part (a), [A] and [P] are written as [A]0 −x and [P]0 +x respectively.
The rate law is
d[P]
υ= = k r [A]2 [P] = k r ([A]0 − x)2 ([P]0 + x)
dt
The expression [P] = [P]0 + x implies that d[P]/dt = dx/dt so the rate
law becomes
dx dx
= k r ([A]0 − x)2 ([P]0 + x) hence = k r dt
dt ([A]0 − x)2 ([P]0 + x)
Integration of this expression, using x = 0 at time t = 0, gives

x dx x
∫ = ∫ k r dt
0 ([A]0 − x) ([P]0 + x)
2 0
The right-hand side is k r t. The left-hand side is evaluated using the method
of partial fractions described in The chemist’s toolkit 28 in Topic 14B. The
fraction is expressed as a sum
1 A B C
= + +
([A]0 − x) ([P] + x) ([A]0 − x)
2 2 [A] − x [P]0 + x
where A, B, and C are constants to be found. This expression is multiplied

through by ([A]0 − x)2 ([P] + x)
1 = A([P]0 + x) + B([A]0 − x)([P]0 + x) + C([A]0 − x)2
The brackets are expanded and the terms are collected
1 = (C−B)x 2 +(A+B[A]0 −B[P]0 −2C[A]0 )x+(A[P]0 +B[A]0 [P]0 +C[A]20 )
Equating coefficients gives the three equations
C−B=0
A + B[A]0 − B[P]0 − 2C[A]0 = 0
A[P]0 + B[A]0 [P]0 + C[A]20 = 1
The first equation implies that C = B. Substituting this into the second
two equations gives
A − B([A]0 + [P]0 ) = 0
A[P]0 + B([A]0 [P]0 + [A]20 ) = 1
The first equation of these two equations is multiplied by [P]0 and sub-
tracted from the second to give
B([A]0 [P]0 + [A]20 ) + B([A]0 [P]0 + [P]20 ) = 1
Rearranging gives
1
B([A]20 + 2[A]0 [P]0 + [P]20 ) = 1 hence B=C=
([A]0 + [P]0 )2
This is substituted back into the equation A − B([A]0 + [P]0 ) = 0 from

above to give
1 1
A− × ([A]0 + [P]0 ) = 0 hence A=
([A]0 + [P]0 )2 [A]0 + [P]0

x 1
kt = ∫ dx
0 ([A]0 − x)2 ([P]0 + x)
x 1 1
=∫ ( +
0 ([A]0 + [P]0 )([A]0 − x)2 ([A]0 + [P]0 )2 ([A]0 − x)
1
+ ) dx
([A]0 + [P]0 )2 ([P]0 + x)
x
1 ln([A]0 − x) ln([P]0 + x)
=[ − + ]
([A]0 + [P]0 )([A]0 − x) ([A]0 + [P]0 ) 2 ([A]0 + [P]0 )2 0
1 1 [P]0 + x
=( + ln )
([A]0 + [P]0 )([A]0 − x) ([A]0 + [P]0 )2 [A]0 − x
1 1 [P]0
−( + 2
ln )
([A]0 + [P]0 )[A]0 ([A]0 + [P]0 ) [A]0
1 1 1 1 [A]0 ([P]0 + x)
= ( − )+ ln
[A]0 + [P]0 [A]0 − x [A]0 ([A]0 + [P]0 )2 [P]0 ([A]0 − x)
Substituting x = [P] − [P]0 gives the integrated rate law as
1 1 1
kt = ( − )
[A]0 + [P]0 [A]0 + [P]0 − [P] [A]0
1 [A]0 [P]
+ ln
([A] + [P]0 )2 [P]0 ([A]0 + [P]0 − [P])
It is not possible to rearrange this equation to give a simple expression for
[P].
To find the time at which the rate reaches a maximum, the expression
for the rate, υ = k r [A]2 [P] = k r ([A]0 − x)2 ([P]0 + x), is differentiated
and the derivative is set equal to zero. The chain rule for differentiation
implies that dυ/dt = (dυ/dx) × (dx/dt), hence
dυ d dx
= [k r ([A]0 − x)2 ([P]0 + x)] ×
dt dx dt
dx
= [−2k r ([A]0 − x)([P]0 + x) + k r ([A]0 − x)2 ]
dt
Setting this equal to zero implies that
dx
−2k r ([A]0 − x)([P]0 + x) + k r ([A]0 − x)2 = 0 or =0
dt
Because x = [P] − [P]0 , dx/dt = d[P]/dt = υ and so the solution dx/dt =
0 corresponds to υ = 0. This represents a minimum rate rather than a
maximum and so is rejected. Examining the other solution gives
−2k r ([A]0 − x)([P]0 + x) + k r ([A]0 − x)2 = 0
hence ([A]0 − x) [([A]0 − x) − 2([P]0 + x)] = 0
Hence
x = [A]0 or [A]0 − x = 2([P]0 + x)
Because x = [A]0 − [A], the solution x = [A]0 corresponds to [A] = 0.
From the rate law υ = k r [A]2 [P] this corresponds to v = 0 and therefore to
a minimum rate rather than a maximum. The maximum rate is therefore
given by the second expression, which is rearranged to yield x = 13 ([A]0 −
2[P]0 ).
This is then substituted into the integrated rate law from above
1 1 1 1 [A]0 ([P]0 + x)
kr t = ( − )+ ln
[A]0 + [P]0 [A]0 − x [A]0 ([A]0 + [P]0 )2 [P]0 ([A]0 − x)
1 1 1
= (2 2 − )
[A]0 + [P]0 3 [A]0 + 3 [P]0 [A]0
1 [A]0 ( 13 [P]0 + 13 [A]0 )
+ ln
([A]0 + [P]0 )2 [P]0 ( 23 [A]0 + 23 [P]0 )
1 3 [A]0 + [P]0 [A]0
= ( − + ln )
([A]0 + [P]0 )2 2 [A]0 2[P]0
The time at which the rate is at a maximum is therefore
1 1 [P]0 [A]0
t= ( + + ln )
k r ([A]0 + [P]0 ) 2 [A]0
2 2[P]0
(e) The rate law is integrated as in part (d). Writing [A] = [A] − x and [P] =
[P]0 + x the rate law is
d[P]
υ= = k r [A][P]2 = k r ([A]0 − x)([P]0 + x)2
dt
The expression [P] = [P]0 + x implies that d[P]/dt = dx/dt. Therefore
dx x dx t
= k r ([A]0 −x)([P]0 +x)2 hence ∫ = ∫ k dt
dt 0 ([A]0 − x)([P]0 + x)2 0
The right-hand side is k r t. The left-hand side is evaluated using the method
of partial fractions, as in part (d). The fraction is expressed as a sum
1 A B C
= + +
([A]0 − x)([P]0 + x)2 [A]0 − x [P]0 + x ([P]0 + x)2
where A, B, and C are constants to be found. The expression is multiplied

through by ([A]0 − x)([P]0 + x)2 .
1 = A([P]0 + x)2 + B([A]0 − x)([P]0 + x) + C([P]0 + x)

= (A − B)x 2 + (2A[P]0 + B[A]0 − B[P]0 − C)x
+ (A[P]20 + B[A]0 [P]0 + C[A]0 )
Equating coefficients gives
A−B = 0
2A[P]0 + B[A]0 − B[P]0 − C) = 0
A[P]20 + B[A]0 [P]0 + C[A]0 = 1
The first equation implies that A = B. Substituting this into the second
two equations gives
A([A]0 + [P]0 ) − C = 0
A([P]20 + [A]0 [P]0 ) + C[A]0 = 1
The first of these two equations is multiplied by [A]0 and added to the
second equation to give
1
A([A]20 + 2[A]0 [P]0 + [P]20 ) = 1 hence A=B=
([A]0 + [P]0 )2
This is substituted back into the equation A([A]0 + [P]0 ) − C = 0 from

above to give
1 1
× ([A]0 + [P]0 ) − C = 0 hence C=
([A]0 + [P]0 )2 ([A]0 + [P]0 )
x 1 1
kr t = ∫ ( +
0 ([A]0 + [P]0 ) ([A]0 − x) ([A]0 + [P]0 )2 )([P]0 + x)
2
1
+ ) dx
([A]0 + [P]0 )([P]0 + x)2
x
− ln([A]0 − x) ln([P]0 + x) 1
=[ + − ]
([A]0 + [P]0 ) 2 ([A]0 + [P]0 ) 2 ([A]0 + [P]0 )([P]0 + x) 0
1 [P]0 + x 1
=( ln − )
([A]0 + [P]0 )2 [A]0 − x ([A]0 + [P]0 )([P]0 + x)
1 [P]0 1
−( 2
ln − )
([A]0 + [P]0 ) [A]0 ([A]0 + [P]0 )[P]0
1 [A]0 ([P]0 + x) 1 1 1
= ln + ( − )
([A]0 + [P]0 ) 2 [P]0 ([A]0 − x) [A]0 + [P]0 [P]0 [P]0 + x
Substituting x = [P] − [P]0 gives the integrated rate law as
1 [A]0 [P] 1 1 1
kr t = 2
ln + ( − )
([A]0 +[P]0 ) [P]0 ([A]0 +[P]0 −[P]) [A]0 +[P]0 [P]0 [P]
As in part (d) it is not possible to rearrange this equation to give a simple

expression for [P].
The time at which the rate reaches a maximum is found by the same
method as used in part (d).
dυ dv dx d dx
= × = [k r ([A]0 − x)([P]0 + x)2 ] ×
dt dx dt dt dt
dx
= [2k r ([A]0 − x)([P]0 + x) − k r ([P]0 + x) ] ×
2
dt
dx
= k r ([P]0 + x)(2[A]0 − [P]0 − 3x)
dt
At the maximum, this expression is equal to zero. The solution dx/dt = 0
is discarded for the same reason as in part (d), and the solution x = −[P]0
is discarded because x must be positive. The remaining solution is
2[A]0 − [P]0 − 3x = 0 hence x = 31 (2[A]0 − [P]0 )
This is substituted into the integrated rate law from above
1 [A]0 ([P]0 + x) 1 1 1
kr t = ln + ( − )
([A]0 + [P]0 )2 [P]0 ([A]0 − x) [A]0 + [P]0 [P]0 [P]0 + x
1 [A]0 ( 23 [A]0 + 32 [P]0 )
= ln
([A]0 + [P]0 )2 [P]0 ( 13 [A]0 + 13 [P]0 )
1 1 1
+ ( − 2 )
[A]0 + [P]0 [P]0 3 [A]0 + 23 [P]0
1 2[A]0 [A]0 + [P]0 3
= (ln + − )
([A]0 + [P]0 )2 [P]0 [P]0 2
Hence the maximum rate is reached at
1 2[A]0 [A]0 1
t= (ln + − )
k r ([A]0 + [P]0 )2 [P]0 [P]0 2
I14.5 (a) Because the second step is rate-determining, the first step and its reverse
are treated as a pre-equilibrium because the rate of reaction of A− with AH
to form product is assumed to be too slow to affect the maintenance of the
pre-equilibrium. As explained in Section 14E.5 on page 566 it follows that
k a [BH+ ][A− ] k a [AH][B]
K= = hence [A− ] =
k a′ [AH][B] k a′ [BH+ ]
The rate formation of product is υ = d[P]/dt = k b [A− ][AH]. The expres-
sion for [A− ] is substituted into this to give
k a [AH][B] k a k b [AH]2 [B]

υ = k b [A− ][AH] = k b [AH] =
k a′ [BH+ ] k a′ [BH+ ]
The same result is alternatively derived using the steady-state approxima-
tion. Applying the steady-state approximation to A− gives
d[A− ]
= k a [AH][B] − k a′ [BH+ ][A− ] − k b [A− ][AH] = 0
dt
This expression is rearranged
k a [AH][B]
[A− ] = ′
k a [BH+ ] + k b [AH]
Substituting this into the rate of formation of product gives
k a [AH][B] k a k b [AH]2 [B]

υ = k b [A− ][AH] = k b [AH] =
k a′ [BH+ ] + k b [AH] k a′ [BH+ ] + k b [AH]
Finally it is noted that because the second step is rate-determining the

rate of conversion of A− to products, k b [A− ][AH], is much slower than
the rate of its reversion to reactants, k a′ [A− ][BH+ ].
k a′ [A− ][BH+ ] ≫ k b [A− ][AH] hence k a′ [BH+ ] ≫ k b [AH]
The term k b [AH] is therefore neglected in the denominator of the rate

law which then becomes υ = k a k b [AH]2 [B]/k a′ [BH+ ] as before.
(b) Because the second step is rate-determining, the formation of HAH+ from
HA and H+ forms a pre-equilibrium in which the rates of the forward and
backward steps are considered to be equal because the gradual removal of
HAH+ to form products is assumed to be too slow to significantly affect
the maintenance of the equilibrium. Equating the rates of the forward
and backward steps gives
k a [HA][H+ ]
k a [HA][H+ ] = k a′ [HAH+ ] hence [HAH+ ] =
k a′
The rate of product formation is equal to the rate of the second step
k a [HA][H+ ] ka kb
v = k b [HAH+ ][B] = k b ′
[B] = [HA][H+ ][B]
ka k a′
I14.7 A polymer consisting of N monomer units has a molar mass of N M 1 , where M 1

is the molar mass of a single monomer unit. The mean molar mass is therefore
⟨N M 1 ⟩ = M 1 ⟨N⟩, and likewise the mean square molar mass and mean cube
molar mass are
⟨M 2 ⟩ = ⟨(N M 1 )2 ⟩ = M 12 ⟨N 2 ⟩ and ⟨M 3 ⟩ = ⟨(N M 1 )3 ⟩ = M 13 ⟨N 3 ⟩
The task is therefore to find ⟨N 2 ⟩ and ⟨N 3 ⟩.

It is supposed that each monomer has one end group A with which it can join
to another monomer. In a polymer, only the terminal monomer unit in the
chain has a free end group.
The probability PN that a polymer consists of N monomers is equal to the
probability that it has N − 1 reacted end groups and one unreacted end group.
The fraction of end groups that have reacted is p and the fraction of free end
groups remaining is 1 − p, so the probability that a polymer contains N − 1
reacted groups and one unreacted group is p N−1 × (1 − p).
It is convenient to begin by evaluating the average value of N.

∞ ∞ ∞
⟨N⟩ = ∑ N PN = ∑ N p N−1 (1 − p) = (1 − p) ∑ N p N−1
N=1 N=1 N=1
To evaluate the sum, it is noted that N p N−1 corresponds to the derivative of p N .

Hence
∞ ∞
N−1 d N d ∞ N d
∑ Np = ∑ p = [∑ p ] = [p + p2 + p3 + ...]
N=1 N=1 dp dp N=1 dp
The expression in square brackets is a geometric series with first term p and
common ratio p; the sum to infinity of this series is therefore p/(1 − p). Hence
∞
N−1 d p (1 − p) + p 1
∑ Np = [ ]= =
N=1 dp 1 − p (1 − p) 2 (1 − p)2
The average value of N is therefore

∞
1 1
⟨N⟩ = (1 − p) ∑ N p N−1 = (1 − p) × =
N=1 (1 − p) 2 1− p
This is the same result as [14F.12a–571] which is derived in Section 14F.2(a) on

page 571 by a different method. However the approach used here is more easily
generalised to find an expression for ⟨N 2 ⟩ and ⟨N 3 ⟩.
∞ ∞ ∞
⟨N 2 ⟩ = ∑ N 2 PN = ∑ N 2 p N−1 (1 − p) = (1 − p) ∑ N 2 p N−1
N=1 N=1 N=1
The sum ∑∞ 2 N−1

N=1 N p is evaluated by noting that N p N−1 is the derivative of
N
p
∞ ∞ ∞
2 N−1 d N d ∞
∑N p = ∑ N × N p N−1 = ∑ N × p = ∑ Np
N
N=1 N=1 N=1 dp dp N=1

∞
d
= [p ∑ N p N−1 ]
dp N=1
The sum ∑∞
N=1 N p
N−1
was already evaluated above; its value is 1/(1 − p)2 .
Hence
∞
2 N−1 d 1 1+ p
∑N p = [p × ]=
N=1 dp (1 − p) 2 (1 − p)3
The mean value of N 2 is therefore

∞
1+ p 1+ p
⟨N 2 ⟩ = (1 − p) ∑ N 2 p N−1 = (1 − p) × =
N=1 (1 − p)3 (1 − p)2
The mean value of N 3 is evaluated in a similar way, using the result already
deduced that ∑∞ 2 N−1

N=1 N p = (1 + p)/(1 − p)3 .
∞ ∞ ∞
⟨N 3 ⟩ = ∑ N 3 PN = ∑ N 3 p N−1 (1 − p) = (1 − p) ∑ N 2 × N p N−1
N=1 N=1 N=1
∞ ∞
d N d 2 N
= (1 − p) ∑ N 2 p = (1 − p) ∑N p
N=1 dp dp N=1
(1+p)/(1−p)3
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ·¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ
∞
d d 1+ p
= (1 − p) [p ∑ N 2 p N−1 ] = (1 − p) [p × ]
dp N=1 dp (1 − p)3
p2 + 4p + 1 p2 + 4p + 1
= (1 − p) × =
(1 − p)4 (1 − p)4
Using the results ⟨M 2 ⟩ = M 12 ⟨N 2 ⟩ and ⟨M 3 ⟩ = M 13 ⟨N 3 ⟩ deduced at the start of

the question, together with the expressions for ⟨N 2 ⟩ and ⟨N 3 ⟩, the mean square
and cube molar masses are
M 12 (1 + p) M 13 (p2 + 4p + 1)
⟨M 2 ⟩ = ⟨M 3 ⟩ =
(1 − p)2 (1 − p)3
(a) The required ratio is
⟨M 3 ⟩ M 13 (p2 + 4p + 1)/(1 − p)3 M 1 (p2 + 4p + 1)

= =
⟨M 2 ⟩ M 12 (1 + p)/(1 − p)2 (1 + p)(1 − p)
(b) The average number of monomers per polymer, that is the chain length,
is given by [14F.12a–571], ⟨N⟩ = 1/(1 − p). This expression is rearranged
to p = 1 − 1/⟨N⟩. This is then substituted into the expression derived in
(b) to give the ratio in terms of chain length.
⟨M 3 ⟩ M 1 (p2 + 4p + 1) (1 − 1/⟨N⟩)2 + 4(1 − 1/⟨N⟩) + 1

= = M 1
⟨M 2 ⟩ (1 + p)(1 − p) (1 + [1 − 1/⟨N⟩])(1 − [1 − 1/⟨N⟩])
M 1 (6⟨N⟩2 − 6⟨N⟩ + 1)
=
2⟨N⟩ − 1
15 Reaction dynamics
15A Collision theory
D15A.1 Reactions between complex molecules might be expected to have strong steric
requirements (small steric factors) as a result of the reaction requiring a par-
ticular orientation and approach of the reacting parts of the molecule: the
more complex the molecules, the smaller the fraction of collisions which are
potentially reactive.
In the RRK theory of unimolecular reactions molecular complexity plays a dif-

ferent role in that it governs the distribution of energy in the excited molecule.
In this theory the rate constant for the unimolecular decay of an energized
molecule A* is given by [15A.11–597],
E ∗ s−1
k b (E) = (1 − ) kb
E
Here E ∗ is the minimum energy that must be accumulated in a bond for it to

break, E is the total energy, and s is the number of modes of motion (modelled
as harmonic oscillators) that the molecule possesses. The term in parentheses is
less than 1, therefore the expression implies that the more complex the molecule
(the greater s), the smaller the rate constant becomes.
D15A.3 To the extent that real gases deviate from perfect gas behaviour, they do so
because of intermolecular interactions. Interactions tend to be more important
at high pressures, when the size of the molecules themselves is not negligible
compared to the average intermolecular distance (mean free path). Attractive
interactions might enhance a reaction rate compared to the predictions of col-
lision theory, particularly if the parts of the molecules that are attracted to each
other are the reactive sites. Similarly, repulsive interactions might reduce the
frequency of collisions compared to what would be predicted for perfect gases.
570 15 REACTION DYNAMICS
E15A.1(a) The collision theory expression for the rate constant is given in [15A.9–595].
1/2
8kT
kr = σ NA ( ) e−E a /RT
πµ
= (0.36 × 10−18 m2 ) × (6.0221 × 1023 mol−1 )
1/2
8 × (1.3806 × 10−23 J K−1 ) × (650 K)
×( )
π(3.32 × 10−27 kg)
J mol−1 )/[(8.3145 J K−1 mol−1 )×(650 K)]
× e−(171×10
3
= 1.0 × 10−5 mol−1 m3 s−1
The units are best resolved by realising that (8kT/πµ)1/2 is a speed, with units
m s−1 . Note that 0.36 nm2 is 0.36 × 10−18 m2 .
E15A.2(a) As described in Section 15A.1(b) on page 593, the reactive cross section may be
estimated from the (non-reactive) collision cross sections of A and B: σest =
1/2 1/2
1
4
(σA + σB )2 . The steric factor is given by the ratio of the experimental
reactive cross section, σexp , to the estimated cross section
σexp 9.2 × 10−22 m2

P= =
σest [(0.95 × 10−18 m2 )1/2 + (0.65 × 10−18 m2 )1/2 ]2 /4
= 1.2 × 10−3
E15A.3(a) In the RRK theory the rate constant for the unimolecular decay of an energized
E ∗ s−1
k b (E) = (1 − ) k b = (1 − x)s−1 k b
E
where x = E ∗ /E. For a non-linear molecule with 5 atoms there are 3N − 6 =
3 × 5 − 6 = 9 normal modes, so s = 9. This expression is rearranged for x to give
x = 1 − [k b (E)/k b ]1/(s−1)
= 1 − [3.0 × 10−5 ]1/(9−1) = 0.73
E15A.4(a) In the RRK theory the rate constant for the unimolecular decay of an energized
k b (E) E ∗ s−1
= (1 − )
kb E
where E ∗ is the minimum energy needed to break the bond, and E is the energy
available from the collision. With the data given
10−1
k b (E) 200 kJ mol−1
= (1 − ) = 5.12 × 10−7
kb 250 kJ mol−1
E15A.5(a) The collision frequency is given by [1B.12b–17], z = σ υ rel p/kT, where σ is the
collision cross-section, given in terms of the collision diameter d as σ = πd 2 ,
and υ rel is the mean relative speed of the colliding molecules. This speed is
given by [1B.11b–16], υ rel = (8kT/πµ)1/2 , with µ = m A m B /(m A + m B ). For
collisions between like molecules µ = m/2 and υ rel = (16kT/πm)1/2 .
σ υ rel p πd 2 p 16kT 1/2 π 1/2

z= = ( ) = 4d 2 p ( )
kT kT πm mkT
= 4 × (380 × 10−12 m)2 × (120 × 103 Pa)
1/2
π
×( )
(17.03 × 1.6605 × 10−27 kg) × (1.3806 × 10−23 J K−1 ) × (303 K)
= 1.12 × 1010 s−1
To confirm the units of z it is useful to recall that 1 J = 1 kg m2 s−2 and 1 Pa =

1 kg m−1 s−2 .
The collision density between identical molecules is given by [15A.4b–593]
4kT 1/2 2
Z AA = σ ( ) N A [A]2
πm
where [A] is the molar concentration of the gas. In turn, this is expressed in
terms of the pressure using the perfect gas equation to give [A] = n A /V =
p A /RT.
4kT 1/2 N A2 p2A 2 πkT

1/2
p2A π 1/2
Z AA = πd 2 ( ) = 2d ( ) = 2d 2
( ) p2A
πm R2 T 2 m k2 T 2 mk 3 T 3
= 2 × (360 × 10−12 m)2 × (120 × 103 Pa)2
1/2
π
×( )
(17.03 × 1.6605 × 10−27 kg) × (1.3806 × 10−23 J K−1 )3 × (303 K)3
= 1.62 × 1035 m−3 s−1
The above expression shows that z ∝ pT −1/2 , but at constant volume p ∝ T,

therefore the overall temperature dependence is z ∝ T 1/2 . The percentage
increase in z on increasing T by 10 K is therefore
3131/2 − 3031/2
= 0.0163... = 1.6%
3031/2
Similarly the final expression for the collision density shows Z AA ∝ p2 T −3/2
which, with p ∝ T, gives Z AA ∝ T 2 T −3/2 ∝ T 1/2 . This is the same dependence
as z, so the same percentage increase will result.
E15A.6(a) The collision theory expression for the rate constant is given in [15A.9–595].
In this expression, the factor e−E a /RT is identified as the fraction of collisions f
having at least kinetic energy E a along the flight path. For example with E a =
20 kJ mol−1 and T = 350 K
Ea 20 × 103 J mol−1
= = 6.87... f = e−6.87 ... = 1.04 × 10−3
RT (8.3145 J K−1 mol−1 ) × (350 K)
A similar calculation gives f = 0.069 at T = 900 K. With E a = 100 kJ mol−1

the result is f = 1.19 × 10−15 at T = 350 K, and f = 1.57 × 10−6 at T = 900 K.
E15A.7(a) The method for calculating the fractions is shown in the solution to Exercise
E15A.6(a). For E a = 20 kJ mol−1 and T = 350 K it is found that f = 1.03...×10−3
and increasing the temperature to 360 K gives f = 1.25...×10−3 . The percentage
increase is
(1.25... × 10−3 ) − (1.03... × 10−3 )
100 × = 21%
1.03... × 10−3
A similar calculation gives an increase by 3.0% at 900 K. With E a = 100 kJ mol−1

the result is 160% at T = 350 K, and 16% at T = 900 K.
P15A.1 The collision theory expression for the rate constant is given in [15A.9–595]
1/2
∗ 8kT
πµ
Here σ ∗ is interpreted as the reactive cross-section, related to the collision

cross-section σ by σ ∗ = Pσ, where P is the steric factor. Comparison of the
above expression for k r with the Arrhenius equation, k r = Ae−E a /RT , gives
the frequency factor as A = σ ⋆ N A (8kT/πµ) ; this is rearranged to give
1/2
∗
an expression for σ . It is convenient to express the given frequency factor
2.4 × 1010 dm3 mol−1 s−1 as 2.4 × 107 m3 mol−1 s−1 . The mass of a CH3 radical
is 15.03 m u , therefore the reduced mass of the collision is µ = 12 × 15.03 m u =
1.24... × 10−26 kg.
A πµ 1/2
σ∗ = ( )
N A 8kT
1/2
2.4 × 107 m3 mol−1 s−1 π(1.24... × 10−26 kg)
= ( )
6.0221 × 1023 mol−1 8 × (1.3806 × 10−23 J K−1 ) × (298 K)
= 4.34... × 10−20 m2 = 0.043 nm2
m s−1 .
To estimate the collision cross-section assume that d is twice the C–H bond
length and compute σ = πd 2 = π(2 × 154 × 10−12 m)2 = 2.98... × 10−19 m2 . The
steric factor is P = σ ∗ /σ = (4.34... × 10−20 m2 )/(2.98... × 10−19 ) = 0.15 .
P15A.3 The collision theory expression for the rate constant is given in [15A.9–595]
1/2
8kT
πµ
The maximum value for the rate constant is when E a = 0. The collision cross
section is taken as σ = πd 2 = π(308 × 10−12 m)2 = 2.98... × 10−19 m2 . The
mass of a CH3 radical is 15.03 m u , therefore the reduced mass of the collision
is µ = 12 × 15.03 m u = 1.24... × 10−26 kg
1/2
8kT
kr = σ NA ( ) = (2.98... × 10−19 m2 ) × (6.0221 × 1023 mol−1 )
πµ
1/2
8 × (1.3806 × 10−23 J K−1 ) × (298 K)
×( )
π(1.24... × 10−26 kg)
= 1.64 × 108 mol−1 m3 s−1
m s−1 .
For a second-order reaction the integrated rate law is [14B.4b–549], 1/[CH3 ] −
1/[CH3 ]0 = k r t. Suppose that initially an amount in moles n 0 of C2 H6 is in-
troduced into the vessel, and that a fraction α dissociates. The amount of C2 H6
remaining is n 0 (1 − α) and the amount of CH3 produced is 2n 0 α. The total
amount of gas is n 0 (1+α), therefore the mole fraction of CH3 is 2α/(1+α) and
hence the partial pressure of CH3 is 2α p tot /(1 + α). The molar concentration
corresponding to this pressure is found using the perfect gas law as
n CH3 p CH3 2α p tot
[CH3 ] = = =
V RT RT(1 + α)
With the data given this evaluates as
2 × 0.1 × (100 × 103 Pa)
[CH3 ] = = 7.33... mol m−3
(8.3145 J K−1 mol−1 )×(298 K)×(1 + 0.1)
1
If recombination proceeds to 90%, the amount of CH3 remaining is 10
of the
initial. The time for this to take place is found by solving
10 1
− = kr t
[CH3 ]0 [CH3 ]0
Hence
9 9
t= =
[CH3 ]0 k r (7.33... mol m−3 ) × (1.64 × 108 mol−1 m3 s−1 )
= 7.5 ns
P15A.5 The collision theory expression for the rate constant, including the steric factor
P, is given in [15A.10–596]
1/2
8kT
k r = Pσ N A ( ) e−E a /RT
πµ
As described in Section 15A.1(b) on page 593, the collision cross-section be-

tween A and B may be estimated from the collision cross sections of A and
1/2 1/2
B: σ = 14 (σA + σB )2 . From the Resource section the cross section for O2 is
0.40 nm2 . No values are given for the ethyl and cyclohexyl radicals, so these will
be approximated by the values for ethene (0.64 nm2 ) and benzene (0.88 nm2 ),
respectively. The reactive cross sections are therefore
σethyl = 41 [(0.40)1/2 + (0.64)1/2 ]2 = 0.512... nm2
A similar calculation gives σhexyl = 0.616... nm2

The mass of O2 is 32.00 m u , that of the C2 H5 radical is 29.06 m u , and that of
the C6 H11 radical is 83.15 m u . The reduced mass of the O2 –C2 H5 collision is
m O2 m C2 H5 32.00 × 29.06
µ= = × (1.6605 × 10−27 kg) = 2.52... × 10−26 kg
m O2 + m C2 H5 32.00 + 29.06
For the O2 –C6 H11 collision the reduced mass is 3.83... × 10−26 kg.
Taking the activation energy as E a = 0, the steric factor is given by
kr πµ 1/2
P= ( )
σ N A 8kT
For this calculation it is convenient to express the rate constants in units of

m3 mol−1 s−1 . For the reaction with C2 H5
4.7 × 106 m3 mol−1 s−1

P=
(0.512... × 10−18 m2 ) × (6.0221 × 1023 mol−1 )
1/2
π(2.52... × 10−26 kg)
×( )
8 × (1.3806 × 10−23 J K−1 ) × (298 K)
= 0.024
A similar calculation for the reaction with C6 H11 gives P = 0.043 .
15B Diffusion-controlled reactions

D15B.1 In the cage effect, a pair of molecules may be held in close proximity for an
extended period of time by the presence of other neighbouring molecules, typ-
ically solvent molecules. Such a pair is called an encounter pair, and their
time near each other is called an ‘encounter’ as opposed to a simple collision.
An encounter may include a series of collisions. Furthermore, as a result of
collisions with neighbouring molecules, an encounter pair may pick up enough
energy to react, even though the pair may not have had enough energy when
first formed.
E15B.1(a) For a diffusion-controlled reaction the rate constant is approximated by [15B.4–
601], k d = 8RT/3η, where η is the viscosity. Recall that 1 P = 10−1 kg m−1 s−1 ,
so that 1 cP = 10−3 kg m−1 s−1 . Therefore the rate constant is
8 × (8.3145 J K−1 mol−1 ) × (320 K)

kd =
3 × (0.89 × 10−3 kg m−1 s−1 )
= 7.97... × 106 m3 mol−1 s−1 = 8.0 × 106 m3 mol−1 s−1
The half-life of a second-order reaction is given by [14B.5–550], t 1/2 = 1/k r [A]0 .
The initial concentration is 1.5 mmol dm−3 which is 1.5 mol m−3 . With the data
given
1
t 1/2 = = 84 ns
(7.97... × 106 m3 mol−1 s−1 ) × (1.5 mol m−3 )
E15B.2(a) The second-order rate constant for a diffusion-controlled reaction is given by

[15B.3–601], k d = 4πR∗ DN A , where R∗ is the critical distance and D is the
diffusion constant. As explained in the text D is the sum of the diffusion con-
stants of the two species. The value of D is estimated using the Stokes–Einstein
equation, D = kT/6πηR, and with the data given separate values of D are
computed for the two species. The critical distance is taken as R ∗ = R A + R B .
k d = 4π(R A + R B )(D A + D B )N A
kT 1 1
= 4πN A (R A + R B ) ( + )
6πη R A R B
= 4π × (6.0221 × 1023 mol−1 ) × (655 + 1820)
(1.3806 × 10−23 J K−1 ) × (313 K) 1 1
× −1 −1 ( + )
−3
6π × (2.93 × 10 kg m s ) 655 1820
= 3.04... × 106 m3 mol−1 s−1
The initial concentrations are [A] = 0.170 mol dm−3 = 0.170 × 103 mol m−3
and [B] = 0.350 mol dm−3 = 0.350 × 103 mol m−3 . The initial rate is therefore
d[P]
= k d [A][B]
dt
= (3.04... × 106 m3 mol−1 s−1 )
× (0.170 × 103 mol m−3 ) × (0.350 × 103 mol m−3 )
= 1.81 × 1011 mol m−3 s−1
Using [15B.4–601], k d = 8RT/3η, the rate constant is
8 × (8.3145 J K−1 mol−1 ) × (313 K)

kd = = 2.37 × 106 m3 mol−1 s−1
3 × (2.93 × 10−3 kg m−1 s−1 )
This value would result in a significantly slower initial rate, casting doubt there-
fore on the validity of the approximations used.
E15B.3(a) The second-order rate constant for a diffusion-controlled reaction is given by

[15B.3–601], k d = 4πR∗ DN A , where R∗ is the critical distance and D is the
diffusion constant. As explained in the text, D is the sum of the diffusion
constants of the two species, therefore in this case D is twice the value given.
With the data given
k d = 4π × (0.5 × 10−9 m) × (2 × 6 × 10−9 m2 s−1 ) × (6.0221 × 1023 mol−1 )

= 4.5 × 107 m3 mol−1 s−1
E15B.4(a) For a diffusion-controlled reaction the rate constant is approximated by [15B.4–

601], k d = 8RT/3η, where η is the viscosity.
(i) For water
8 × (8.3145 J K−1 mol−1 ) × (298 K)

kd = = 6.61 × 106 m3 mol−1 s−1
3 × (1.00 × 10−3 kg m−1 s−1 )
In sorting out the units it is useful to recall 1 J = 1 kg m2 s−2 .

(ii) For pentane
8 × (8.3145 J K−1 mol−1 ) × (298 K)

kd = = 3.0 × 107 m3 mol−1 s−1
3 × (2.2 × 10−4 kg m−1 s−1 )
∗
P15B.1 To simplify the notation the dependence of [J] and [J] on x and t will not
∗
be written explicitly. The proposed solution, [15B.8–602], [J] = [J]e−k r t , is
substituted into the right-hand side of [15B.7–602]
∗
∂ 2 [J] ∗ ∂2
D − k r [J] = D 2 [J]e−k r t − k r [J]e−k r t
∂x 2 ∂x
∂ 2 [J] −k r t
=D e − k r [J]e−k r t
∂x 2
The solution is now substituted into the left-hand side of [15B.7–602]
∗
∂[J] ∂ ∂[J] −k r t
= [J]e−k r t = e − k r e−k r t [J]
∂t ∂t ∂t
The left-and right-hand sides are now set equal
∂ 2 [J] −k r t ∂[J] −k r t
D e − k r [J]e−k r t = e − k r e−k r t [J]
∂x 2 ∂t
The term k r e−k r t [J] cancels to give
∂ 2 [J] −k r t ∂[J] −k r t ∂ 2 [J] ∂[J]

D e = e hence D =
∂x 2 ∂t ∂x 2 ∂t
As specified in the problem, [J] is a solution of [15B.7–602] when k r = 0, and
indeed this is precisely the differential equation which has just been generated.
∗
P15B.3 It is first convenient to compute the derivative of [J] with respect to t and its
second derivative with respect to x.
∗
∂[J] ∂ t
= [k r ∫ [J]e−k r t dt + [J]e−k r t ]
∂t ∂t 0
∂[J] −k r t ∂[J] −k r t
= k r [J]e−k r t + e − k r e−k r t [J] = e (15.1)
∂t ∂t
∗
∂ 2 [J] ∂2 t
= 2 [k r ∫ [J]e−k r t dt + [J]e−k r t ]
∂x 2 ∂x 0
t ∂ 2 [J] ∂ 2 [J] −k r t
−k r t
= kr ∫ e dt + e (15.2)
0 ∂x 2 ∂x 2
Recall that [J] is a solution to [15B.7–602] with k r = 0
∂[J] ∂ 2 [J]
=D
∂t ∂x 2
This is used to substitute for ∂ 2 [J]/∂x 2 in eqn 15.2
∗
∂ 2 [J] t ∂[J] ∂[J] −k r t
D = kr ∫ e−k r t dt + e
∂x 2 0 ∂t ∂t
where the factor of D has been taken over to the left. Next use of made of the
∗
result in eqn 15.1, ∂[J] /∂t = (e−k r t )∂[J]/∂t to rewrite the last expression as
∗ ∗ ∗
∂ 2 [J] t ∂[J] ∂[J]
D = kr ∫ dt +
∂x 2 0 ∂t ∂t
∗
∗ ∗ ∂[J]
= k r {[J] (t) − [J] (0)} +
∂t
∗
∗ ∂[J]
= k r [J] (t) +
∂t
∗
where the initial condition that [J], and hence [J] , must be zero at t = 0 is
used. Rearranging the final equation gives the required differential equation,
∗
thus demonstrating that the proposed form of [J] is indeed a solution.
∗ ∗
∂[J] ∂ 2 [J] ∗
=D − k r [J]
∂t ∂x 2
15C Transition-state theory

D15C.1 This is described in Section 15C.2(b) on page 609. If the solvent were altered
to one with a lower dielectric constant the interaction between the ions would
be greater and this would be manifested in an increased value of A, and hence
steeper slopes for the plots of the rate constant against ionic strength.
D15C.3 The discarded mode would be the anti-symmetric stretch in which the B–C
distance lengthens and the A–B distance decreases.
E15C.1(a) The variation of the rate constant with ionic strength is given by [15C.23–609],
lg k r = lg k r○ + 2Az A z B I 1/2 ; at 298 K and for aqueous solutions A = 0.509. In
the absence of further information assume z A = +1 and z A = −1. Rearranging
for lg k r○ gives
lg k r○ = lg k r − 2Az A z B I 1/2
= lg(12.2 dm6 mol−2 min−1 ) − 2 × (0.509) × (+1) × (−1) × (0.0525)1/2
= 1.31...
Therefore k r○ = 20.9 dm6 mol−2 min−1 .
E15C.2(a) The effect of deuteration on the rate constant is given by [15C.25–611]
⎧ 1/2 ⎫
k r (C–D) ħω(C–H) ⎪
⎪ µ CH ⎪
⎪
= e−ζ ζ= ⎨1 − ( ) ⎬
k r (C–H) 2kT ⎪
⎪ µ CD ⎪
⎪
⎩ ⎭
In this expression ω(C–H) = (k f /µ CH )1/2 . It can be adapted for other pairs of

isotopes by changing the effective masses and the force constant.
The effective mass for 12 C–1 H is
mC mH 12 × 1.0078
µ CH = = m u = 0.929... m u = 1.54... × 10−27 kg
m C + m H 12 + 1.0078
Likewise for 12 C–3 H (denoted C–T) the effective mass is
mC mT 12 × 3.016
µ CT = = m u = 2.41... m u
m C + m T 12 + 3.016
With the given force constant
1/2
450 N m−1
ω(C–H) = ( ) = 5.39... × 1014 s−1
1.54... × 10−27 kg
At 298 K
(1.0546 × 10−34 J s) × (5.39... × 1014 s−1 ) 0.929... 1/2

ζ= −23 −1 {1 − ( ) }
2 × (1.3806 × 10 J K ) × (298 K) 2.41...
= 2.62...
k r (C–T)
= e−ζ = e−2.62 ... = 0.073
k r (C–H)
Raising the temperature will decrease ζ which will have the effect of increasing
the ratio k r (C–T)/k r (C–H) and thus moving it closer to 1. That is, the isotope
effect will be reduced.
E15C.3(a) The empirical expression is compared to the Arrhenius equation k r = Ae−E a /RT ,
allowing the activation energy to be determined from E a /R = 8681 K; hence
E a = (8.3145 J K−1 mol−1 ) × (8681 K) = 72.1... kJ mol−1 . The frequency factor
is A = 2.05 × 1013 dm3 mol−1 s−1 = 2.05 × 1010 m3 mol−1 s−1 .
The relationship between E a and ∆‡ H for a bimolecular solution-phase reac-
tion is given by [15C.17–608], ∆‡ H = E a − RT = (72.1... × 103 J mol−1 ) −
(8.3145 J K−1 mol−1 ) × (303 K) = 69.7 kJ mol−1 . The relationship between
A and ∆‡ S for a bimolecular solution-phase reaction is given by [15C.19b–608]
kT RT ∆‡ S/R
A=e e
h p−○
Ap−○ h
hence ∆‡ S = R ln
ekRT 2
= (8.3145 J K−1 mol−1 )
(2.05 × 1010 m3 mol−1 s−1 ) × (105 Pa) × (6.6261 × 10−34 J s)
× ln
e(1.3806 × 10−23 J K−1 ) × (8.3145 J K−1 mol−1 ) × (303 K)2
= −25.3 J K−1 mol−1
Note the conversion of the units of A to m3 mol−1 s−1 .
E15C.4(a) The empirical expression is compared to the Arrhenius equation k r = Ae−E a /RT ,
allowing the activation energy to be determined from E a /R = 9134 K; hence
E a = (8.3145 J K−1 mol−1 ) × (9134 K) = 75.9... kJ mol−1 . The frequency factor
is A = 7.78 × 1014 dm3 mol−1 s−1 = 7.78 × 1011 m3 mol−1 s−1 .
The relationship between E a and ∆‡ H for a bimolecular solution-phase reac-
tion is given by [15C.17–608], ∆‡ H = E a − RT = (75.9... × 103 J mol−1 ) −
(8.3145 J K−1 mol−1 ) × (303 K) = +73.4... kJ mol−1 . The relationship between
A and ∆‡ S for a bimolecular solution-phase reaction is given by [15C.19b–608]
kT RT ∆‡ S/R
A=e e
h p−○
Ap−○ h
hence ∆‡ S = R ln
ekRT 2
= (8.3145 J K−1 mol−1 )
(7.78 × 1011 m3 mol−1 s−1 ) × (105 Pa) × (6.6261 × 10−34 J s)
× ln
e(1.3806 × 10−23 J K−1 ) × (8.3145 J K−1 mol−1 ) × (303 K)2
= +4.88... J K−1 mol−1
Note the conversion of the units of A to m3 mol−1 s−1 . ∆‡ G is found by com-

bining the values of ∆‡ H and ∆‡ S in the usual way
∆‡ G = ∆‡ H − T∆‡ S
= (+73.4... × 103 J mol−1 ) − (303 K) × (+4.88... J K−1 mol−1 )
= +71.9 kJ mol−1
E15C.5(a) The rate constant for a bimolecular gas phase reaction is given by [15C.18a–608]
kT RT ∆‡ S/R −E a /RT
k r = e2 e e
h p−○
To use this expression the given rate constant needs to be converted to units
of m3 mol−1 s−1 . From the perfect gas law [J] = n/V = p/RT, therefore to
convert the rate constant from units of pressure−1 to units of concentration−1
units requires multiplication by RT
k r = 7.84 × 10−3 kPa−1 s−1 = 7.84 × 10−6 Pa−1 s−1

= (7.84 × 10−6 Pa−1 s−1 ) × (8.3145 J K−1 mol−1 ) × (338 K)
= 0.0220... m3 mol−1 s−1
The units can be deduced using 1 Pa = 1 kg m−1 s−2 and 1 J = 1 kg m2 s−2 .

The above equation rearranges to
hp−○ E a /RT k r hp−○ Ea

∆‡ S = R ln (k r e ) = R ln ( )+
e2 kRT 2 e2 kRT 2 T
= (8.3145 J K−1 mol−1 )
(0.0220... m3 mol−1 s−1 ) × (6.6261 × 10−34 J s) × (105 Pa)
× ln ( )
e2 (1.3806 × 10−23 J K−1 ) × (8.3145 J K−1 mol−1 ) × (338 K)2
58.6 × 103 J mol−1
+
338 K
= −91.2 J K−1 mol−1
E15C.6(a) In Example 15C.1 on page 606 the following expression for the rate constant for
a reaction between structureless particles is derived
1/2
8kT
kr = NA ( ) σ ∗ e−∆E 0 /RT
πµ
The activation energy is obtained from its usual definition, [14D.3–558]
d ln k r
E a = RT 2
dT
d ⎪⎪ ⎡⎢
⎧
8kT
1/2 ⎤
⎥ ∆E 0 ⎫
⎪
⎪
= RT 2 ⎨ln ⎢⎢N A ( ) σ ∗ ⎥⎥ − ⎬
dT ⎪⎪ ⎢
⎩ ⎣
πµ ⎥ RT ⎪ ⎪
⎭
⎦
1 ∆E 0
= RT 2 ( + ) = 21 RT + ∆E 0
2T RT 2
Therefore ∆E 0 = E a − 12 RT and hence

1/2
8kT
kr = NA ( ) σ ∗ e1/2 e−∆E a /RT
πµ
The rate constant for a bimolecular gas phase reaction is given by [15C.18a–608]
kT RT ∆‡ S/R −E a /RT
k r = e2 e e
h p−○
Comparing these two expressions gives
1/2
8kT kT RT ∆‡ S/R
NA ( ) σ ∗ e1/2 = e2 e
πµ h p−○
This is rearranged to give ∆‡ S, noting that for a collision between like molecules
µ = 12 m
1/2
⎛ 8kT hp−○ ⎞
∆ S = R ln N A (
‡
) σ∗
⎝ πµ e3/2 kRT 2 ⎠
= (8.3145 J K−1 mol−1 )
⎡ 1/2
⎢ 8 × (1.3806 × 10−23 J K−1 ) × (300 K)
× ln ⎢⎢(6.0221 × 1023 mol−1 ) × ( )
⎢ π × 12 × 65 × (1.6605 × 10−27 kg)
⎣
× (0.35 × 10−18 m2 )
⎤
(6.6261 × 10−34 J s) × (105 Pa) ⎥
⎥
× −1 −1 −1 ⎥
e3/2 (1.3806 × 10−23 J K ) × (8.3145 J K mol ) × (300 K)2 ⎥⎦
= −74 J K−1 mol−1
E15C.7(a) It is convenient to convert the units of the frequency factor and express it as
A = 4.6 × 109 m3 mol−1 s−1 . The relationship between E a and ∆‡ H for a bi-
molecular gas-phase reaction is given by [15C.17–608], ∆‡ H = E a − 2RT =
(10.0 × 103 J mol−1 ) − 2 × (8.3145 J K−1 mol−1 ) × (298 K) = +5.04... kJ mol−1 =
+5.0 kJ mol−1 . The relationship between A and ∆‡ S for a bimolecular gas-
phase reaction is given by [15C.19a–608]
kT RT ∆‡ S/R
A = e2 e
h p−○
Ap−○ h
hence ∆‡ S = R ln 2
e kRT 2
= (8.3145 J K−1 mol−1 )
(4.6 × 109 m3 mol−1 s−1 ) × (105 Pa) × (6.6261 × 10−34 J s)
× ln
e2 (1.3806 × 10−23 J K−1 ) × (8.3145 J K−1 mol−1 ) × (298 K)2
= −45.8... J K−1 mol−1 = −46 J K−1 mol−1
∆‡ G is found by combining the values of ∆‡ H and ∆‡ S in the usual way
∆‡ G = ∆‡ H − T∆‡ S
= (+5.04... × 103 J mol−1 ) − (298 K) × (−45.8... J K−1 mol−1 )
= +19 kJ mol−1
P15C.1 It is convenient to convert the units of the frequency factor and express it as
A = 4.07 × 102 m3 mol−1 s−1 . The relationship between E a and ∆‡ H for a
bimolecular gas-phase reaction is given by [15C.17–608], ∆‡ H = E a − 2RT =
(65.4 × 103 J mol−1 ) − 2 × (8.3145 J K−1 mol−1 ) × (300 K) = +60.4... kJ mol−1 =
+60.4 kJ mol−1 . The relationship between A and ∆‡ S for a bimolecular gas-
phase reaction is given by [15C.19a–608]
kT RT ∆‡ S/R
A = e2 e
h p−○
Ap−○ h
hence ∆‡ S = R ln 2
e kRT 2
= (8.3145 J K−1 mol−1 )
(4.073 × 102 m3 mol−1 s−1 ) × (105 Pa) × (6.6261 × 10−34 J s)
× ln
e2 (1.3806 × 10−23 J K−1 ) × (8.3145 J K−1 mol−1 ) × (300 K)2
= −1.80... × 102 J K−1 mol−1 = −181 J K−1 mol−1
∆‡ G is found by combining the values of ∆‡ H and ∆‡ S in the usual way
∆‡ G = ∆‡ H − T∆‡ S
= (+60.4... × 103 J mol−1 ) − (300 K) × (−1.80... × 102 J K−1 mol−1 )
= +115 kJ mol−1
The relationship between ∆U and ∆H is given by [2B.4–46], ∆H = ∆U +
∆ν g RT, where ∆ν g is the change in stoichiometric coefficients for gaseous species.
In this case ∆ν g = 1 − 2 = −1, hence
∆‡ U = ∆‡ H − ∆ν g RT = ∆‡ H + RT
= (+60.4... × 103 J mol−1 ) + (8.3145 J K−1 mol−1 ) × (300 K)
= +62.9 kJ mol−1
P15C.3 The starting point is the expression

kT RT N A Λ 3A Λ 3B 2IkT −∆E 0 /RT
kr = κ ( 3 −○ ) 2 e
h p−○ Λ C‡ Vm ħ
The first step is to realise that because pV = nRT, RT/p−○ Vm−○ = 1; the N A is also
taken out of the bracket to give
kT Λ A Λ B 3 2IkT −∆E 0 /RT
kr = κ NA ( ) e
h Λ C‡ ħ2
The thermal wavelength is Λ = h/(2πmkT)1/2 ; substituting this and cancelling

over the fraction gives
3/2
kT h 2 m C‡ 2IkT −∆E 0 /RT
=κ NA ( ) e
h 2πkTm A m B ħ2
Now note that m C‡ = m A + m B , so that m C‡ /m A m B = 1/µ. In addition, the

moment of inertia is written as I = µr 2
3/2
kT h2 2µr 2 kT −∆E 0 /RT
=κ NA ( ) e
h 2πkT µ ħ2
The reactive cross section is identified as σ ∗ = κπr 2 ; using this, and tidying up
all the constants gives the required expression
1/2
8kT
kr = NA ( ) σ ∗ e−∆E 0 /RT
πµ
P15C.5 The rate constant is given by the Eyring equation, [15C.10–606]

−
○
kT RT N A q HD‡2 −∆E 0 /RT
kr = κ e
h p−○ qH−○ qD−○2
To simplify the notation, the overline, double dagger and standard symbols will
be omitted. With such a complex calculation it is best to break it down into
parts and then assemble them to give the final result. First, the pre-multiplying
constants (assuming κ = 1)
kT RT (1.3806 × 10−23 J K−1 ) × (400 K)

N A =
h p−○ 6.6261 × 10−34 J s
(8.3145 J K−1 mol−1 ) × (400 K)
× × (6.0221 × 1023 mol−1 )
105 Pa
= 1.66... × 1035 m3 mol−2 s−1
Next consider the ratio of the translational partition functions: each is given
by q −○ = Vm−○ /Λ 3 = RT/p−○ Λ 3 , with Λ = h/(2πmkT)1/2 . For the purpose of this
approximate calculation it is sufficient to use integer masses
3/2 3/2
qHD2 p−○ h2 m HD2
= ( ) ( )
qH qD2 trans RT 2πkT m H m D2
105 Pa
=
(8.3145 J K mol−1 ) × (400 K)
−1
3/2
(6.6261 × 10−34 J s)2
×( )
2π(1.3806 × 10−23 J K−1 ) × (400 K)
3/2
5
×( )
1 × 4 × 1.6605 × 10−27 kg
= 2.79... × 10−29 mol
Next consider the ratio of the vibrational partition functions: each is given by
q = (1 − e−hc ν̃/k T )−1 . The vibrational frequency of D2 is so high that q = 1.
Each normal mode of the activated complex has ν = 1000 cm−1 n thus the
contribution to the partition function is computed as
hc ν̃ (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (1000 cm−1 )

=
kT (1.3806 × 10−23 J K−1 ) × (400 K)
= 3.59...
q = (1 − e−3.59 ... )−1 = 1.02...
The ratio of vibrational partition functions is therefore

qHD2
= (1.02...)n vib
qH qD2 vib
where n vib is the number of vibrational normal modes of the activated complex.
To find the rotational partition function for D2 requires a knowledge of the
rotational constant, which is computed from the given bond length of H2 . The
effective mass for D2 is 1 m u thus
h
B̃ =
8π 2 cµR 2
6.6261 × 10−34 J s
=
8π 2 (2.9979 × 1010 cm s−1 ) × (1 × 1.6605 × 10−27 kg) × (74 × 10−12 m)2
= 30.7... cm−1
The rotational partition function is therefore
kT (1.3806 × 10−23 J K−1 ) × (400 K)

q= =
σ hc B̃ 2 × (6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 ) × (30.7... cm−1 )
= 4.51...
The rotational partition function for the activated complex depends on the
model chosen, so for the moment it is simply written qDR2 H .
The exponential term e−∆E 0 /RT evaluates to 2.68... × 10−5 . The H atom has a
doublet ground state (one unpaired electron) as does the activated complex
(three electrons in total, one unpaired); the ground state of D2 is not generate.
The electronic partition functions therefore cancel. Putting this all together
gives the following expression for the rate constant
k r = (1.66... × 1035 m3 mol−2 s−1 ) × (2.79... × 10−29 mol) × (1.02...)n vib

× [qDR2 H /(4.51...)] × (2.68... × 10−5 )
= (27.7... m3 mol−1 s−1 ) × (1.02...)n vib × qDR2 H
(a) The next step is to compute the moments of intertia, and hence the ro-
tational constants, for the isoceles geometry of the activated complex.
From the data given the D–D distance is 88.8 pm and the H–D distance
is 96.2 pm. Figure 15.1 shows the geometry: the filled circles are D and
the open circle is H. For this approximate calculation integer masses are
used.
96.2 pm y
b
x
44.4 pm
Figure 15.1
The centre of mass lies along the vertical bisector a, and the moment of
inertia about this axis is
I a = 2 × 2 × m u × (44.4 × 10−12 m)2 = 1.30... × 10−47 kg m2
The corresponding rotational constant is computed as
h (6.6261 × 10−34 J s)
Ã = =
8π 2 cI 8π 2 (2.9979 × 1010 cm s−1 ) × (1.30... × 10−47 kg m2 )
= 21.3... cm−1
The centre of mass lies at the intersection of the axes a and b. Taking
moments about axis b gives 4x = 1 × (y − x). By Pythagoras y = (96.22 −
44.42 )1/2 = 85.3... pm, using which gives x = 17.0... pm. The moment of
inertia about the b axis is therefore
I b = 2 × 2 × m u × x 2 + 1 × m u (y − x)2
= 4 × (1.6605 × 10−27 kg) × (44.4 × 10−12 m)2
+ 1 × (1.6605 × 10−27 kg) × ([85.3... − 17.0...] × 10−12 m)2
= 9.67... × 10−47 kg m2
The corresponding rotational constant is B̃ = 28.9... cm−1 . The third

moment of inertia is most easily found using the property of planar bodies
that the moment of inertia about the axis perpendicular to the plane of
the body is equal to the sum of the other two moments of inertia. Using
this I c = 2.27... × 10−47 kg m2 and C̃ = 12.2... cm−1 . The rotational
partition function is given by [12B.14–473]; the molecule has a two-fold
axis therefore σ = 2.
1 kT 3/2 π 1/2
q= ( ) ( )
σ hc ÃB̃C̃
3/2
(1.3806 × 10−23 J K−1 ) × (400 K)
= 12 ( )
(6.6261 × 10−34 J s) × (2.9979 × 1010 cm s−1 )
1/2
π
×( )
(21.3... cm−1 ) × (28.9... cm−1 ) × (12.2... cm−1 )
= 47.1...
For this triangular activated complex there are three normal modes, one
of which corresponds to the reaction coordinate. Thus, the rate constant
is
k r = (27.7... m3 mol−1 s−1 ) × (1.02...)n vib × qDR2 H

= (27.7... m3 mol−1 s−1 ) × (1.02...)2 × (47.1...)
= 1.4 × 103 m3 mol−1 s−1 = 1.4 × 106 dm3 mol−1 s−1
(b) Now consider a linear geometry for the activated complex, H–D–D, with
the H–D distance as 96.2 pm and the D–D distance as 88.8 pm. The first
task is to locate the centre of mass. Assuming that this is a distance x from
the right-hand D it follows that 2x = 2(88.8 − x) + 1(96.2 + 88.8 − x).
which gives x = 72.52 pm. The moment of inertia is therefore
I = {2 × (72.52 pm)2 + 2 × ([88.8 − 72.52] pm)2
+ 1 × ([96.2 + 88.8 − 72.52] pm)2 }

× (1.6605 × 10−27 kg) × [(10−12 m)/(1 pm)]2
= 3.93... × 10−47 kg m2
The corresponding rotational constant is B̃ = 7.11... cm−1 . The rotational

partition function is
kT
q=
hc B̃
(1.3806 × 10−23 J K−1 )×(400 K)
= = 39.0...
(6.6261 × 10−34 J s)×(2.9979 × 1010 cm s−1 )×(7.11... cm−1 )
For this linear activated complex there are four normal modes, one of
which corresponds to the reaction coordinate. Thus, the rate constant is
k r = (27.7... m3 mol−1 s−1 ) × (1.02...)n vib × qDR2 H

= (27.7... m3 mol−1 s−1 ) × (1.02...)3 × (39.0...)
= 1.2 × 103 m3 mol−1 s−1 = 1.2 × 106 dm3 mol−1 s−1
The effect of changing the geometry is rather small.

(c) The calculations performed so far both give rate constants in excess of the
target (by a factor of about 3). The only contribution which can feasibly
change by this much is the rotational partition function of the activated
complex. To make this smaller, the rotational constants need to be in-
creased, which is achieved by generally shrinking the size of the complex.
Some trial calculations suggest that shrinking the H–D distance to 80%
of the H2 distance achieves the desired result, but this seems a rather
implausible structure of the activated complex.
P15C.7 The variation of the rate constant with ionic strength is given by [15C.23–609],
lg k r = lg k r○ + 2Az A z B I 1/2 ; at 298 K and for aqueous solutions A = 0.509. A
plot of lg(k r /k r○ ) against I 1/2 is used to explore whether or not this relationship
applies.
I I 1/2 k r /k r○ lg(k r /k r○ )
0.010 0 0.100 0 8.10 0.908
0.015 0 0.122 5 13.30 1.124
0.020 0 0.141 4 20.50 1.312
0.025 0 0.158 1 27.80 1.444
0.030 0 0.173 2 38.10 1.581
0.035 0 0.187 1 52.00 1.716
1.8
1.6
lg(k r /k r○ )
1.4
1.2
1.0
0.8
0.08 0.10 0.12 0.14 0.16 0.18 0.20
I 1/2
Figure 15.2
The plot is shown in Fig 15.2: the data fall on a good straight line with slope
+9.18. Such a value implies
2 × (0.509) × (z A z B ) = +9.18 hence (z A z B ) = +9.02
A plausible interpretation is z A = +3 and z B = +3, which is consistent with the

protein being cationic.
P15C.9 Some experimentation with various graphs shows that a plot of lg k r against I
(not I 1/2 ) gives a good straight line, as shown in Fig 15.3.
I k r /(dm3 mol−1 s−1 ) lg[k r /(dm3 mol−1 s−1 )]

0.020 7 0.663 −0.178 5
0.052 5 0.670 −0.173 9
0.092 5 0.679 −0.168 1
0.157 5 0.694 −0.158 6
lg[k r /(dm3 mol−1 s−1 )]
−0.16
−0.17
−0.18
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
I
Figure 15.3
The equation of the best-fit line is
lg[k r /(dm3 mol−1 s−1 )] = 0.1451 × I − 0.1815
In the limit of zero ionic strength lg[k r○ /(dm3 mol−1 s−1 )] = −0.1815, hence
k r○ = 0.658 dm3 mol−1 s−1 .
From the text [15C.21b–609] gives the dependence of lg k r on the activity coef-
ficients
γ C‡
lg k r = lg k r○ − lg
γA γB
In this case one reactant (say A) is known to have a charge of −1, and the other
reactant (say B) is neutral. Therefore, the charge on the activated complex is
also −1 and hence, if the Debye–Hückel limiting law applies to these species,
γ C‡ = γ A . It therefore follows that
lg k r = lg k r○ + lg γ B
The data given show a linear dependence of lg k r on I, therefore it is con-

cluded that lg γ B ∝ I. The constant of proportion is the slope of the graph:
lg γ B = 0.145 I .
P15C.11 The effect of deuteration on the rate constant is given by [15C.25–611]

⎧ 1/2 ⎫
k r (C–D) ħω(C–H) ⎪
⎪ µ CH ⎪
⎪
= e−ζ ζ= ⎨1 − ( ) ⎬
k r (C–H) 2kT ⎪
⎪ µ CD ⎪
⎪
⎩ ⎭
In this expression ω(C–H) = (k f /µ CH )1/2 .
In the case described in the problem the H (or D) is attached to a much heavier
fragment (the rest of the molecule, X), thus the effective mass for the vibration
is well approximated by the mass of the lighter atom (H or D), giving
k r (D) ħω(X–H) m H 1/2

= e−ζ ζ= {1 − ( ) }
k r (H) 2kT mD
Given that k r (D)/k r (H) = 1/6.4 it follows that ζ = 1.86.... With this value, and
using integer masses for H and D, the expression for ζ becomes
(1.0546 × 10−34 J s)ω

1.86... = {1 − (1/2)1/2 }
2(1.3806 × 10−23 J K−1 ) × (298 K)
hence ω = 4.95... × 1014 s−1
The force constant is therefore
k f = ω 2 m H = (4.95... × 1014 s−1 ) × (1 × 1.6605 × 10−27 kg) = 408 N m−1
15D The dynamics of molecular collisions

D15D.1 The saddle point on the potential energy surface corresponds to the transition
state of a reaction. The saddle-point energy is the minimum energy required
for reaction; it is the minimum energy for a path on the potential energy sur-
face that leads from reactants to products. Because many paths on the surface
between reactants and products do not pass through the saddle point, they
necessarily pass through points of greater energy, so the activation energy can
be greater than the saddle-point energy. Thus, the saddle-point energy is a
lower limit to the activation energy.
D15D.3 Molecular beams may be used to prepare molecules in specific rotational and
vibrational states, and then to examine the results of collisions between such
precisely prepared species. Section 15D.1(a) on page 613 describes how molec-
ular beams are prepared such that the molecules in them have a very nar-
row range of velocities and therefore relatively few collisions to redistribute
their energies. Molecules in such beams can be prepared in specific vibrational
states, for example, by using lasers to excite vibrations. Crossing two molecular
beams allows collisions to be staged between two sets of precisely characterized
molecules. Detectors can then be used to study the results of those collisions,
recording the number of molecules in which particular states that are scattered
in a given direction.
E15D.1(a) The numerator of [15D.6–621] is integrated as follows with the assumption that
P(E) is independent of E and can therefore be written P
∞ ∞
∫ P(E) e−E/k T dE = P ∫ e−E/k T dE
0 0
∞
= −PkT e−E/k T ∣0 = PkT
The effect of this term is to make the rate constant increase with temperature.
E15D.2(a) Refer to Fig. 15D.18 on page 620, which shows an attractive potential energy
surface as well as trajectories of both a successful reaction and an unsuccessful
one. The trajectories begin in the lower right, representing reactants. The
successful trajectory passes through the transition state (marked as ‡ ○). This
trajectory is fairly straight from the lower right through the transition state,
indicating little or no vibrational excitation in the reactant. Therefore most
of its energy is in translation. Since it has enough total energy to reach the
transition state, the reactant can be described as being high in translational
energy and low in vibrational energy.
This successful trajectory moves from side to side along the valley representing
products, so the product is high in vibrational energy and relatively lower in
translational energy. The unsuccessful trajectory, by contrast, has a reactant
high in vibrational energy; it moves from side to side in the reactant valley
without reaching the transition state.
P15D.1 The change in intensity of the beam, dI, is proportional to the number of scat-
terers per unit volume, N , the intensity of the beam, I, and the path length dL.
The constant of proportionality is the collision cross-section σ, the ‘target area’
of each scatterer.
1
dI = −σN I dL hence dI = −σN dL hence d ln I = −σN dL
I
If the incident intensity at L = 0 is I 0 , and the intensity after scattering through
length L is I, integration gives
I L
∫ d ln I = ∫ −σN dL
I0 0
hence ln I/I 0 = −σN L
The result may be expressed as I = I 0 e−σN L .
P15D.3 Following Brief illustration 15D.1 on page 617
k r○ −λυ −υhν/k T k○ ′ ′ k○ ′
kr = ∑ δ υυ′ e e = r ∑ e−λυ e−υ hν/k T = r ∑ e−υ (λ+hν/k T)
q υ,υ′ q υ′ q υ′
Writing out the first few terms of the sum shows that it is a geometric progres-
sion with common ratio e−(λ+hν/k T)
k r○
= [1 + e−(λ+hν/k T) + e−2(λ+hν/k T) + . . .]
q
k○ 1
= r
q 1 − e−(λ+hν/k T)
where to go to the final line the fact that the sum to infinity of a geometric series
is a/(1 − r), where r is the common ratio and a is the first term, is used. This
relationship requires r < 1, which is the case provided λ > 0.
15E Electron transfer in homogeneous systems

D15E.1 Electron tunnelling plays an important role in electron transfer. As is discussed
in Section 7D.4 on page 254, it is expected that tunnelling will be more impor-
tant for electrons than any other particles that participate in chemical reactions
because electrons are so much lighter than atoms or ions. Tunnelling is respon-
sible for the exponential distance dependence of the factor H et (d)2 , given by
[15E.4–624], and the electron-transfer rate constant, given by [15E.5–624], is
directly proportional to H et (d)2 . A full discussion is found in Section 15E.2 on
page 623.
D15E.3 This is discussed in detail in Section 15E.2 on page 623 and Section 15E.3 on
page 624.
E15E.1(a) The rate constant for electron-transfer is given by [15E.5–624] together with
[15E.6–625]
1/2
1 π3 (∆ r G −○ + ∆E R )2
) H et (d)2 e−∆ G/RT
‡
k et = ( ∆‡ G =
h RT∆E R 4∆E R
For the two reactions given, ∆E R and ∆‡ G are assumed to be the same. The
○2 −βd
distance dependence of H et (d)2 is given by [15E.4–624], H et (d)2 = H et e ,
therefore
k et,2 (H et (d)2 )2
= = e−β(d 2 −d 1 )
k et,1 (H et (d)2 )1
hence ln(k et,2 /k et,1 ) = −β(d 2 − d 1 )
ln(k et,2 /k et,1 )
therefore β = −
(d 2 − d 1 )
ln[(4.51 × 104 s−1 )/(2.02 × 105 s−1 )]
β=−
(1.23 nm) − (1.11 nm)
= 12.5 nm−1
○2 −βd
E15E.2(a) The distance dependence of H et (d)2 given by [15E.4–624], H et (d)2 = H et e .
H et (d 2 )2
= e−β(d 2 −d 1 )
H et (d 1 )2
−1
= e−(9 nm )[(2.0 nm)−(1.0 nm)]
= 1.23... × 10−4
Increasing the distance from 1.0 nm to 2.0 nm reduces H et (d)2 to about 0.01%
of its initial value.
E15E.3(a) The rate constant for electron-transfer is given by [15E.5–624] together with
[15E.6–625]
1/2
1 π3 (∆ r G −○ + ∆E R )2
) H et (d)2 e−∆ G/RT
‡
k et = ( ∆‡ G =
h RT∆E R 4∆E R
With the given data there is only one unknown quantity, ∆E R , but it is not pos-
sible to find an analytical expression for this in terms of the other parameters.
However, mathematical software is able to find a solution numerically. Before
embarking on such a calculation it is important to make sure that the units of
the various quantities are consistent.
The choice is made to express the energies as molar quantities (J mol−1 ). It
therefore follows that the term in parentheses has units (J−1 mol), and given
that the units of 1/h are (J−1 s−1 ), it follows that for the rate constant to have
the expected units of s−1 , H et (d)2 must be in J2 mol−1 . Using the conversion
factors from inside the front cover
2
1.9864 × 10−23 J
H et (d)2 = ((0.04 cm−1 ) × ) × (6.0221 × 1023 mol−1 )
1 cm−1
= 3.80... × 10−25 J2 mol−1
96.485 × 103 J mol−1
∆ r G −○ = (−0.185 eV) × = −1.78... × 104 J mol−1
1 eV
The constant factor evaluates to
1/2 1/2
1 π3 1 π3
( ) H et (d)2 = ( )
h RT 6.6261 × 10−34 J s (8.3145 J K−1 mol−1 )×(298 K)
× (3.80... × 10−25 J2 mol−1 )
= 6.41... × 107 J1/2 mol−1/2 s−1
The equation to be solved is therefore
−1/2 G/(2.47 ...×10 3 J mol−1 )
(37.5 s−1 ) = (6.41... × 107 J1/2 mol−1/2 s−1 ) × ∆E R × e−∆
‡
[(−1.78... × 104 J mol−1 ) + ∆E R ]2

with ∆‡ G =
4∆E R
In solving this equation it is helpful to know that the result is likely to be of the
order of tens of kJ mol−1 so as to guide the numerical solution. An alternative
is to plot the right-hand side of the above expression for over a range of values
of ∆E R and look for the value which gives the required k r . The final result is
∆E R = 2 kJ mol−1 .
P15E.1 This Problem is somewhat ill-posed and some of the references to equations
in the text are incorrect. The key missing item is that it should be assumed
that the electron-transfer process is rate limiting so that k r = Kk et , where K
is the equilibrium constant for the diffusive encounter; this point is discussed
in Section 15E.1 on page 622. In addition, the first-order rate constant for the
electron-transfer process is written, using transition-state theory, as
k 1st = κν‡ K ‡ = κν‡ e−∆ G/RT

‡
(15.3)
This relationship is analogous to those developed in Section 15C.1 on page 604

for second-order reactions. In parts (c) and (d) of the problem eqn 15.3 should
be used. The description of the dependence of f is somewhat misleading.
(a) The Gibbs energy of activation is given by [15E.6–625], ∆‡ G = (∆ r G −○ +

∆E R )2 /4∆E R . For the DA electron transfer
∆‡ G DA = (∆ r G −○ + ∆E R,DA )2 /4∆E R,DA (15.4)
For the AA self exchange ∆ r G −○ = 0 therefore
∆‡ G AA = (∆E R,AA )2 /4∆E R,AA = 41 ∆E R,AA
and likewise ∆‡ G DD = 14 ∆E R,DD

(b) The square on the right-hand side of eqn 15.4 is expanded and the terms
separated to give
∆ r G −○ 2∆ r G −○ ∆E R,DA ∆E R,DA 2
2
∆‡ G DA = + +
4∆E R,DA 4∆E R,DA 4∆E R,DA
∆ r G −○
2
= + 1 ∆ r G −○ + 14 ∆E R,DA ≈ 21 ∆ r G −○ + 41 ∆E R,DA
4∆E R,DA 2
where the approximation ∣∆ r G −○ ∣ ≪ ∆E R,DA is used in the last step. As
indicated in the Problem, ∆E R,DA may be written
∆E R,DA = 21 (∆E R,AA + ∆E R,DD )
Using the above relationships ∆‡ G DD = 14 ∆E R,DD and ∆‡ G AA = 14 ∆E R,AA ,

the expression for ∆E R,DA becomes
∆E R,DA = 21 (4∆‡ G AA + 4∆‡ G DD ) = 2(∆‡ G AA + ∆‡ G DD )
Using this in the above expression for ∆‡ G DA gives the required expres-
sion
∆‡ G DA = 21 ∆ r G −○ + 41 ∆E R,DA = 21 ∆ r G −○ + 21 (∆‡ G AA + ∆‡ G DD ) (15.5)
(c) The overall rate constant for the self-exchange process are written k AA =
K AA k et,AA and likewise for DD. The first-order rate constant for the elec-
tron transfer process is written using eqn 15.3 as k et,AA = κν‡ e−∆ G AA /RT
‡
and likewise for k et,DD . As indicated in the Problem, it is assumed that

κν‡ is the same for all reactions. Therefore, the expressions for the overall
rate constants are
k AA = K AA κν‡ e−∆ G AA /RT

k DD = K DD κν‡ e−∆ G DD /RT
‡ ‡
(d) By analogy with the expression for k AA
k r = k AD = K AD κν‡ e−∆ G AD /RT

‡
(e) Next, eqn 15.5 is used to substitute for ∆‡ G AD in the expression for k r
k r = K AD κν‡ e−∆ G AD /RT

‡
−
○
= K AD κν‡ e−(∆ r G +∆‡ G AA +∆ ‡ G DD )/2RT
−
○ 1/2 1/2
= K AD e−∆ r G /2RT
[κν‡ e−∆ G AA /RT
[κν‡ e−∆ G DD /RT
‡ ‡
] ]
The term is the first bracket is recognised as k AA /K AA , and that in the

second as k DD /K DD to give
−
○
k r = K AD e−∆ r G /2RT 1/2 1/2
[k AA /K AA ] [k DD /K DD ]
−
○
The equilibrium constant for the overall reaction is written K = e−∆ r G /RT
,
−
○
therefore e−∆ r G /2RT = K 1/2 and hence
K AD
kr = 1/2
(k AA k DD K)1/2
(K AA K DD )
which is of the required form.
P15E.3 The variation of the electron-transfer rate constant with ∆ r G −○ is given by [15E.8–
626]
2
RT ∆ r G −○ ∆ r G −○
ln k et = − ( ) − 12 ( ) + const.
4∆E R RT RT
A plot of ln k et against −∆ r G −○ is expected to be an inverted parabola and, as
described in the text, the maximum occurs at −∆ r G −○ = ∆E R . The plot is shown
in Fig 15.4.
The data are a good fit to the second-order polynomial
lg k et = −2.828 × (−∆ r G −○ /eV)2 + 5.942 × (−∆ r G −○ /eV) + 7.129
which is shown on the plot. The maximum of this function occurs when the
derivative is zero, that is when 2×−2.828×(−∆ r G −○ /eV)+5.942 = 0; this occurs
at (−∆ r G −○ /eV) = 1.05.... Therefore ∆E R = 1.05 eV .
P15E.5 The variation of the electron-transfer rate constant with distance is given by
[15E.7–625], ln k et = −βd + const. This relationship is tested by plotting ln k et
against d; the plot is shown in Fig 15.5.
10
lg k et
6
0.5 1.0 1.5 2.0 2.5

−
○
−∆ r G /eV
Figure 15.4
d/nm k et /s−1 ln(k et /s−1 )

0.48 1.58 × 1012 28.1
0.95 3.98 × 109 22.1
0.96 1.00 × 109 20.7
1.23 1.58 × 108 18.9
1.35 3.98 × 107 17.5
2.24 63.1 4.14
30
lg(k et /s−1 )
20
10
0.5 1.0 1.5 2.0

d/nm
Figure 15.5
The data are a reasonable fit to a straight line with slope −13.43. The slope is
identified as −β, therefore β = 13 nm−1 .

I15.1 Both the Marcus theory of photo-induced electron transfer and the Förster
theory of resonance energy transfer examine interactions between a molecule
excited by absorption of electromagnetic energy (the chromophore S) and an-
other molecule Q. They explain different mechanisms of quenching, that is,
different ways that the chromophore gets rid of extra energy after absorbing a
photon through intermolecular interactions. Another common feature of the
two is that they depend on physical proximity of S and Q: they must be close
for action to be efficient.
In the Marcus theory, the rate of electron transfer depends on the reaction
Gibbs energy of electron transfer, ∆ r G, and on the energy cost to S, Q, and
the reaction medium of any concomitant molecular rearrangement. The rate is
enhanced when the driving force (∆ r G) and the reorganization energy are well
matched.
Resonant energy transfer in the Förster mechanism is most efficient when Q
can directly absorb electromagnetic radiation from S. The oscillating dipole
moment of S is induced by the electromagnetic radiation it absorbed. It trans-
fers the excitation energy of the radiation to Q via a mechanism in which its
oscillating dipole moment induces an oscillating dipole moment in Q. This
energy transfer can be efficient when the absorption spectrum of the acceptor
(Q) overlaps with the emission spectrum of the donor (S).
I15.3 Typical orders of magnitudes are qmT /N A ≈ 107 , q R ≈ 10 per rotational degree of
freedom, q V ≈ 1 per vibrational degree of freedom, and q E ≈ 1. Vibrational and
electronic contributions will therefore be ignored from now on. According to
transition-state theory the rate constant is given by ([15C.10–606] and [15C.9–
606])
kT RT N A qC−○‡ −∆E 0 /RT
kr = κ e
h p−○ qC−○ qB−○
At 298 K the factors in from the the ratio of partition functions evaluate to
1.5 × 1011 m3 mol−1 s−1 , assuming κ = 1.
For a reaction between structureless particles A, B, and C‡ all have contribu-
tions from translation; in addition, C‡ has two rotational degrees of freedom,
therefore
N A × qmT × (q R )2 −∆E 0 /RT
k r = (1.5 × 1011 m3 mol−1 s−1 ) × e
qmT × qmT
(q R )2 −∆E 0 /RT
= (1.5 × 1011 m3 mol−1 s−1 ) × e
qmT /N A
(10)2
= (1.5 × 1011 m3 mol−1 s−1 ) ×
107
= (1.5 × 106 m3 mol−1 s−1 ) × e−∆E 0 /RT
In collision theory the rate constant is given by [15A.9–595]

1/2
8kT
πµ
For a typical value σ = 0.4 nm2 and a mass of 2 × 10−26 kg, at 298 K
k r = (1.7 × 108 m3 mol−1 s−1 ) × e−E a /RT
Assuming that E a ≈ ∆E 0 , collision theory gives a rate constant greater by about

a factor of 100, implying a steric factor of about 0.01; this is a plausible result.
If A and B are non-linear triatomics, then A, B and C‡ all have three rotational
degrees of freedom
N A × qmT × (q R )3
k r = (1.5 × 1011 m3 mol−1 s−1 ) × e−∆E 0 /RT
qmT × (q R )3 × qmT × (q R )3
1
= (1.5 × 1011 m3 mol−1 s−1 ) × R 3 e−∆E 0 /RT
(q ) × qmT /N A
1
= (1.5 × 1011 m3 mol−1 s−1 ) ×
(10) × 107
3
= (15 m3 mol−1 s−1 ) × e−∆E 0 /RT
The steric factor is now 9 × 10−8 – very much smaller than for the reaction
between structureless particles.
16 Magnetic resonance
16A General principles

D16A.1 The Larmor frequency is the rate of precession of a magnetic moment (electron
or nuclear) in a magnetic field. Resonance occurs when the frequency of the
applied radiation matches the Larmor frequency.
D16A.3 The energy levels associated with the interaction between magnetic nuclei and
an applied magnetic field scale directly with the field, with the constant of
proportion depending on the identity of the nucleus. The interaction between
an electron and an applied magnetic field behaves in the same way, but the
interaction is much greater, by a factor of the order of 1000.
E16A.1(a) The energy level separation is ∆E = hν where ν = γ N B0 /2π, [16A.6–633].
Hence, in megahertz, the frequency separation is
γ N B0 (6.73 × 107 T−1 s−1 ) × (15.4 T)
ν = 10−6 × = 10−6 × = 165 MHz
2π 2π
E16A.2(a) The energy level separation is ∆E = hν where ν = γ N B0 /2π, [16A.6–633].
Hence, for a given magnetic field, ∆E ∝ γ N . Using γ N (15 N) = −2.712 ×
107 T−1 s−1 and γ N (31 P) = 10.84 × 107 T−1 s−1 , it follows that ∣γ N (31 P)∣ >
∣γ N (15 N)∣ and so the separation of energy levels is larger for 31 P than for 15 N.
E16A.3(a) The ground state has m I = + 21 (α spin) and population N α , and the upper
state has m I = − 12 (β spin) and population N β . The total population N is N =
N α + N β , and the population difference is N α − N β . The Boltzmann distribution
gives N β /N α = e−∆E/k T , where ∆E is the energy difference between the two
states: ∆E = γ N ħB0 . It follows that N β = N α e−∆E/k T . With these results
N α − N β N α − N β N α (1 − e−∆E/k T ) 1 − e−∆E/k T
= = =
N N α + N β N α (1 + e−∆E/k T ) 1 + e−∆E/k T
Because ∆E ≪ kT the exponential e−∆E/k T is approximated as 1 − ∆E/kT to

give
N α − N β 1 − (1 − ∆E/kT) ∆E/kT ∆E γ N ħB0
≈ = = =
N 1 + (1 − ∆E/kT) 2 + ∆E/kT 2kT 2kT
600 16 MAGNETIC RESONANCE
For a 1 H nucleus and at 298 K

N α − N β γ N ħB0 (26.75 × 107 T−1 s−1 ) × (1.0546 × 10−34 J s) × B0
= =
N 2kT 2 × (1.3806 × 10−23 J K−1 ) × (298 K)
= 3.42... × 10−6 × (B0 /T)
For B0 = 0.30 T, (N α − N β )/N = 1.0 × 10−6 ; for B0 = 1.5 T, the ratio is

5.1 × 10−6 ; for B0 = 10 T, the ratio is 3.4 × 10−5 .
E16A.4(a) The population difference for a collection of N spin- 12 nuclei is given by [16A.8b–
635], (N α − N β ) ≈ Nγ N ħB0 /2kT, where N α is the number of spins in the
lower energy state and N β is the number of spins in the higher energy state. At
constant temperature, (N α − N β )/N ∝ B0 . Hence, for the relative population
difference to be increased by a factor of 5, the applied magnetic field must
increase by a factor of 5 . This is independent of the type of nucleus.
E16A.5(a) The EPR resonance frequency ν is given by [16A.12b–636], hν = g e µ B B0 , where

g e is the magnetogyric ratio of the electron and B0 is the magnetic field strength.
With ν = c/λ it follows that
hν hc
B0 = =
ge µB ge µB λ
(6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )
= = 1.3 T
2.0023 × (9.2740 × 10−24 J T−1 ) × (8 × 10−3 m)
E16A.6(a) The nuclear g-factor g I is given by [16A.4c–633], g I = γ N ħ/µ N , where µ N is the

nuclear magneton (5.051 × 10−27 J T−1 ) and γ N is the nuclear magnetogyric
ratio, the value of which depends on the identity of the nucleus. The units of ħ
are J s and g I is a dimensionless number, so the nuclear magnetogyric ratio γ N
has units (J T−1 )/(J s) = T−1 s−1 .
E16A.7(a) The magnitude of the angular momentum is given by [I(I + 1)]1/2 ħ where I is
the nuclear spin quantum number. For a proton, I = 21 , hence the magnitude
√
of the angular momentum is [ 12 ( 21 + 1)]1/2 ħ = 3ħ/2 .
The component of the angular momentum along the z-axis is m I ħ where m I =
I, I − 1, ..., −I. For a proton, the components along the z-axis are ± 12 ħ and the
angle between angular momentum vector and the z-axis takes the values
⎛ ± 12 ħ ⎞ 1
θ = cos−1 √ = cos−1 ( √ ) = ±0.9553 rad = ±54.74○
⎝ 3ħ⎠ 3
2
E16A.8(a) The NMR frequency is equal to the Larmor precession frequency, ν L , which
is given by [16A.7–633], ν L = γ N B0 /2π, where B0 is the magnitude of the
magnetic field and γ N is the nuclear magnetogyric ratio. Use Table 16A.2 on
page 433 in the Resource section for the value of γ N . Hence,
γ N B0 (26.752 × 107 T−1 s−1 ) × (13.5 T)

νL = = = 5.75 × 108 Hz = 575 MHz
2π 2π
E16A.9(a) The energies of the nuclear spin states in a magnetic field are given by [16A.4d–
633], E m I = −g I µ N B0 m I where g I is the nuclear g-factor, µ N is the nuclear
magneton, B0 is the magnitude of the magnetic field, and the component of
the angular momentum on a specified axis is m I ħ where m I = I, I − 1, ..., −I.
Therefore, since the possible values of m I are ± 23 , ± 21 , the energies of nuclear
spin states are
E m I = −g I µ N B0 m I
= −(0.4289) × (5.0508 × 10−27 J T−1 ) × (6.800 T) × m I
= (−1.473... × 10−26 J) × m I
Hence E±3/2 = ∓2.210 × 10−26 J and E±1/2 = ∓7.365 × 10−27 J .
P16A.1 The resonance condition in NMR is [16A.6–633], hν = γ N ħB0 , where ν is
the resonance frequency, γ N is the nuclear magnetogyric ratio, and B0 is the
magnetic field strength. Hence, assuming the same magnetic field for both
neutrons and 1H,
hν n γ N (n)ħB0 γ N (n)
= =
hν 1H γ N (1 H)ħB0 γ N (1 H)
With the data given, and taking the modulus of the negative magnetogyric ratio
of the neutron,
∣γ N (n)∣ (18.324 × 107 T−1 s−1 )

ν n = ν 1H × = (300 × 10 6
Hz) ×
γ N (1 H) (26.752 × 107 T−1 s−1 )
= 2.1 × 108 Hz = 210 MHz
The energy of the state with quantum number m I is E m I = −m I γ N ħB0 . The

neutron has spin 12 and a negative magnetogyric ratio. Therefore the state with
m I = − 12 (the β state) has the lower energy.
The relative population difference for spin- 21 nuclei with positive γ N is given by
[16A.8b–635], N α − N β ≈ Nγ N ħB0 /2kT. This applies equally well to negative
γ N : the term on the right is then negative, implying that (N α − N β ) is negative,
which is expected because β is the lower energy, more populated, level.
If the 1H resonance frequency is ν 1H , it follows from the resonance condition
hν 1H = γ N ( 1H)ħB0 that B0 = 2πν 1H /γ N ( 1H). The population difference is
computed as (N β − N α ), requiring a reversal of the sign on both sides
N β − N α −γ N (n)ħB0
=
N 2kT
The above expression for B0 is used to give
N β − N α −γ N (n)ħ 2πν 1H −γ N (n)hν 1H
= × =
N 2kT γ N ( 1H) 2kTγ N ( 1H)
With the data given
N β − N α −(−18.324 × 107 T−1 s−1 ) × (6.6261 × 10−34 J s) × (300 × 106 Hz)

=
N 2(1.3806 × 10−23 J K−1 ) × (298 K) × (26.752 × 107 T−1 s−1 )
= 1.65 × 10−5
P16A.3 (a) Absorption intensity is given by [16A.8c–635], intensity ∝ Nγ N

2 2
B0 /T. For
the same N, B0 and T, the intensity is simply ∝ γ N . For the intensity to
2
be equal for 13 C with natural abundance A, and 15 N with an enriched

abundance A′
A′ × [γ N (15 N)]2 = A × [γ N (13 C)]2

A[γ N (13 C)]2
hence A′ =
[γ N (15 N)]2
1.108% × (6.7272 × 107 T−1 s−1 )2
=
(−2.7126 × 107 T−1 s−1 )2
= 6.81%
Therefore, a 6.8 per cent enrichment in 15 N is needed.

(b) The intensity I relative to natural abundance 13 C, for 100 per cent enrich-
ment of 17 O is
A′ × [γ N (17 O)]2
I= × I13 C
A × [γ N (13 C)]2
100% × (−3.627 × 107 T−1 s−1 )2
= × I13 C
1.108% × (6.7272 × 107 T−1 s−1 )2
= 26.2 I13 C
16B Features of NMR spectra

D16B.1 This is discussed in Section 16B.3(c) on page 646.
D16B.3 These are discussed in Section 16B.2 on page 640.
D16B.5 This is discussed in detail in Section 16B.3(d) on page 647. In summary, two
nuclei are chemically equivalent if they are related by a symmetry operation of
the molecule. Such nuclei will have the same resonance frequency, that is, the
same chemical shift. Examples are the protons in benzene, and H-2 and H-6 in
4-nitrophenol. In benzene the protons are related by a C 6 operation (as well as
others) and in 4-nitrophenol the protons are related by reflection in a mirror
plane perpendicular to the plane of the ring and passing through C-1 and C-4.
Two chemically equivalent nuclei A and A′ are magnetically equivalent if the
coupling between A and any other magnetic nucleus in the molecule, say Z,
is the same as the coupling between A′ and Z. If Z itself is to be magnetically

equivalent to Z′ , then the couplings A–Z, A–Z′ , A′ –Z, and A′ –Z′ must all be
equal: that is, the couplings between any member of a group of equivalent spins
with any member of another group of equivalent spins must be the same in
order for the groups each to be magnetically equivalent.
For example, in PF5 , which has a trigonal bi-pyramidal structure, the three
equatorial fluorines are magnetically equivalent, as are the two axial fluorines.
A second example is 1,2,3-trichlorobenzene: H–4 and H–6 are magnetically
equivalent as they both have the same coupling to H-5. Chemical equivalence
does not imply magnetic equivalence. In the case of 4-nitrophenol, protons
H-2 and H-6, though chemically equivalent, are not magnetically equivalent
because the coupling of H-2 to H-3 is different from the coupling of H-6 to
H-3.
E16B.1(a) The resonance frequency ν is given by [16B.5–638], ν = ν○ + (ν spect /106 )δ,
where ν○ is the resonance frequency of the standard. The frequency separation
of the two peaks, ∆ν, is
ν spect ν spect ν spect
∆ν = ν 2 − ν 1 = [ν○ + ( ) δ 2 ] − [ν○ + ( 6 ) δ 1 ] = ( 6 ) (δ 2 − δ 1 )
106 10 10
6
400.130000 × 10 Hz
=( ) (9.80 − 2.20) = 3040 Hz
106

where ν○ is the resonance frequency of the standard. The frequency separation
of the two peaks, ∆ν, is
ν spect ν spect ν spect
∆ν = ν 2 − ν 1 = [ν○ + ( ) δ 2 ] − [ν○ + ( 6 ) δ 1 ] = ( 6 ) (δ 2 − δ 1 )
106 10 10
This is rearranged to give ∆δ, the separation of the two peaks on the shift scale
106 106
∆δ = δ 2 − δ 1 = ∆ν × ( ) = (550 Hz) × = 1.37
ν spec 400.130000 × 106 Hz
E16B.3(a) The combination of [16B.1–638], Bloc = B0 + δB, and [16B.2–638], δB = −σB0 ,

gives the relationship Bloc = (1 − σ)B0 . For ∣σ ○ ∣ << 1, [16B.6–639] gives δ =
(σ ○ − σ) × 106 . Hence,
∆δ = δ 2 −δ 1 = (σ ○ −σ2 )×106 −(σ ○ −σ1 )×106 = −(σ2 −σ1 )×106 = −(∆σ)×106
∆Bloc = (1 − σ2 )B0 − (1 − σ1 )B0 = −(σ2 − σ1 )B0 = −(∆σ)B0

Substituting for ∆σ gives ∆Bloc = (∆δ/106 )B0 .
(i) For B0 = 1.5 T, ∆Bloc = [(9.80−2.20)/106 ]×(1.5 T) = 1.1×10−5 T = 11 µT .
(ii) For B0 = 15 T, ∆Bloc = [(9.80 − 2.20)/106 ] × (15 T) = 1.1 × 10−4 T =
110 µT .

where ν○ is the resonance frequency of the standard. The resonance from TMS
is taken as the origin so ν○ = 0.
For 1 H nuclei, I = 21 and so each signal is split into (2nI + 1) = 2 × 1 × 21 + 1 =
2 peaks, with a splitting equal to the coupling constant J AX = 10 Hz. For a
spectrometer operating at 250 MHz the resonance frequencies of the two spins
are
ν spect 250 × 106 Hz
νA = ν○ + ( 6 ) δ = 0 + × 1.00 = 250 Hz
10 106
νX = ν○ + ( 6
)δ = 0+ × 2.00 = 500 Hz
10 106
There will therefore be peaks of equal height at 245 Hz, 255 Hz, 495 Hz, and
505 Hz.
νX νA
J AX = 10 Hz J AX = 10 Hz
250 Hz
ν increasing
For a spectrometer operating at 800 MHz the resonance frequencies of the two
spins are

νA = ν○ + ( ) δ = 0 + × 1.00 = 800 Hz
106 106

νX = ν○ + ( ) δ = 0 + × 2.00 = 1600 Hz
106 106
There will therefore be peaks of equal height at 795 Hz, 805 Hz, 1595 Hz, and
1605 Hz.
νX νA
J AX = 10 Hz J AX = 10 Hz
800 Hz
ν increasing
19
E16B.5(a) F NMR: the four 19 F nuclei are equivalent and so all have the same shift. This
resonance is split by coupling to the single 10 B nucleus into 2nI + 1 lines; here
n = 1 and I = 3 so there are 7 lines of equal intensity. The splitting between
each is the B–F coupling constant, and the multiplet is centred at the shift of
the fluorine.
10
B NMR: the 10 B nucleus is coupled to four equivalent 19 F nuclei, giving a
multiplet with 2nI + 1 lines; here n = 4 and I = 12 so there are 5 lines. These
lines have intensities in the ratio 1:4:6:4:1, each line is separated from the next
by the B–F coupling constant, and the whole multiplet is centred at the shift of
the boron.
E16B.6(a) The 31 P nucleus is coupled to six equivalent 19 F nuclei which have I = 12 . This
splits the resonance into 2nI + 1 = 2 × 6 × 21 + 1 = 7 lines (a septet) with intensity
ratio 1:6:15:20:15:6:1. The separation between adjacent lines in the multiplet is
the P–F coupling constant.
E16B.7(a) The multiplet resulting from coupling to four equivalent spin- 21 nuclei orig-
inates from the resonance of nucleus A being split into two by coupling with
one X nucleus, and then each of those two lines being split into two by coupling
to the second X nucleus. These lines are then each further split into two by
coupling to the third X nucleus, and finally each of these lines is split into two
by the fourth X nucleus.
As a result of the X nuclei being equivalent, each causes the same splitting and
thus some lines are coincident and give rise to absorption lines of increased
intensity. As shown in Fig. 16.1, the overall result is a 1:4:6:4:1 quintet .
J AX
ν
Figure 16.1
E16B.8(a) The multiplet resulting from coupling to two non-equivalent spin- 12 nuclei orig-
inates from the resonance of nucleus A being split into two by coupling with
the X1 nucleus, and then each of these lines being split into two by coupling to
the X2 nucleus.
Because the coupling between A and X1 is not the same as that between A and
X2 , no lines are necessarily coincident. Coupling to two non-equivalent spin- 12
nuclei thus results in a multiplet with four lines of equal intensity, in this context
usually referred to as a doublet of doublets, as shown in Fig. 16.2.
J AX1
J AX2
ν
Figure 16.2
E16B.9(a) Coupling to one I = 52 nucleus results in the original line being split into (2I +
1) = 6 lines. Each of these lines is then split into six in the same way by coupling
to the second I = 52 nucleus.
As a result of each nucleus being equivalent, the splittings are the same and
thus some lines are coincident and give rise to absorption lines of increased
intensity. The result is a 1:2:3:4:5:6:5:4:3:2:1 multiplet , as is shown in Fig. 16.3.
J AX
ν
Figure 16.3
E16B.10(a) A group of nuclei are chemically equivalent if they are related by a symmetry
operation of the molecule and therefore have the same chemical shift. Two
chemically equivalent nuclei A and A′ are magnetically equivalent if the cou-
pling between A and any other magnetic nucleus in the molecule, say Z, is
the same as the coupling between A′ and Z. If Z itself is to be magnetically
equivalent to Z′ , then the couplings A–Z, A–Z′ , A′ –Z, and A′ –Z′ must all be
equal: that is, the couplings between any member of a group of equivalent spins
with any member of another group of equivalent spins must be the same in
order for the groups each to be magnetically equivalent.
In 1-chloro-4-bromobenzene, H2 and H6 are chemically equivalent because
they are related by a 180○ degree rotation about the C1 –C4 axis. Similarly, H3
and H5 are chemically equivalent because they are related by the same opera-
tion.
The coupling between H2 and H3 is not the same as between H2 and H5 , there-
fore the test for magnetic equivalence is not satisfied: H2 and H6 are not mag-
netically equivalent, nor are H3 and H5 .
Cl
H6 H2
H5 H3
Br
E16B.11(a) The molecule PF5 is trigonal bipyramidal. The three equatorial fluorine atoms,
Fe , are related by a 120○ rotation about the Fa –P–Fa axis: they are therefore
chemically equivalent. The two axial fluorine atoms, Fa , are related by reflection
in the plane formed by the three equatorial fluorine atoms: the axial fluorine
atoms are therefore chemically equivalent.
Each axial fluorine has the same coupling to each equatorial fluorine so the test
for magnetic equivalence is satisfied: the three Fe are magnetically equivalent,
as are the two Fa .
Fa
Fe
P Fe
Fe
Fa
E16B.12(a) Coalescence of two NMR lines due to rapid nuclei exchange occurs when the
condition given in [16B.16–649], τ = 21/2 /πδν, is satisfied; in this expression τ
is the lifetime of an environment, and δν is the difference between the Larmor
frequencies of the two environments. Using [16B.5–638],
νspec νspec νspec

δν = ν 2 − ν 1 = [ν○ + ( 6
) δ 2 ] − [ν○ + ( 6 ) δ 1 ] = ( 6 ) (δ 2 − δ 1 )
10 10 10
For a proton jumping between two sites with first-order rate constant k, k = 1/τ
νspec 550 × 106

δν = ( ) (δ 2 − δ 1 ) = ( ) (4.8 − 2.7) = 1155 Hz
106 106
1 πδν π × (1155 Hz)

k= = = = 2.6 × 103 s−1
τ 21/2 21/2
E16B.13(a) The δ scale is defined by [16B.4a–638], δ = (ν − ν○ ) × 106 /ν○ , where δ is the

chemical shift of the peak, ν is the resonance frequency of the peak, and ν○ is
the resonance frequency of the standard. Hence
ν − ν○ (500.132500 MHz) − (500.130000 MHz)

δ= ○
× 106 = × 106 = 5.0
ν (500.130000 MHz)
E16B.14(a) The δ scale is defined by [16B.4a–638], δ = (ν − ν○ ) × 106 /ν○ , where δ is the

chemical shift of the peak, ν is the resonance frequency of the peak, and ν○ is
the resonance frequency of the standard. Hence
ν − ν○ 750 Hz
δ= ○
× 106 = × 106 = 1.5
ν (500.130000 × 106 Hz)
129
P16B.1 Xe NMR spectrum: the signal is split by coupling to the 19F atom (100%
natural abundance, spin 12 ) into a doublet with coupling constant J = 7600 Hz.
19
F NMR spectrum: this consists of a superposition of spectra from the dif-
ferent isotopes of Xe, and the intensity of these spectra reflects the abundance
of each isotope. A doublet, with splitting 7600 Hz, arises from those 26% of
molecules containing 129Xe. The remaining 74% of molecules do not contain
Xe nuclei which give rise to splitting, so a single line appears.
The observed spectrum is not a triplet, in the sense that it does not arise from
coupling to two equivalent spin- 12 nuclei. Rather it is a superposition of a singlet
and a doublet.
P16B.3 Three possible structures for SF4 are shown below.
.. Fa ..
Fe Fe
S Fe S : F S F
Fe Fe F F
Fa Fa
A B C
Structure A. There are two environments for fluorine atoms: the three Fe are
equivalent as they are related by a threefold rotation about the S–Fa bond; Fa
is unique. Each Fe has the same coupling to Fa , so the Fe are magnetically
equivalent. The spectrum will therefore consist of a doublet from Fe and a
quartet from Fa .
Structure B. There are two environments for fluorine atoms: the two Fe are
equivalent as they are related by a twofold rotation about the S–lone pair axis;
the two Fa atoms are equivalent because they are related by reflection in the
mirror plane formed by the two Fe and the S. Each Fe has the same coupling to
each Fa , so the Fe are magnetically equivalent, as are the Fa . The spectrum will
therefore consist of a triplet from Fe , and likewise a triplet from Fa .
Structure C. All the fluorine atoms are equivalent as they are related by a four-
fold rotation about the S–lone pair axis. A single line will be seen in the spec-
trum, therefore.
P16B.5 Figure 16.4 shows the three possible staggered conformations of XYCHCHR3 R4 .
When a staggered conformation is adopted, the angle between adjacent C–H
bonds is ϕ = 60○ . Conformation a is trans (ϕ = 180○ , 3 J HH = 3 J t ) and confor-
mations b and c are gauche (ϕ = 60○ , 3 J HH = 3 J g ).
R3 H R4
X H X H X H
H R4 R4 R3 R3 H
Y Y Y
a b c
Figure 16.4
For R3 = R4 = H, all conformations have the same energy and each occurs with
equal probability (P = 13 ). Thus
3
J HH = 13 (3 J t + 2 × 3 J g ) = 7.3 Hz (16.1)
For R3 = CH3 and R4 = H, the steric bulk of the methyl group makes con-
formation c very unfavourable. Thus, only equivalent conformations a and
b are relevant for calculating 3 J HH : these have equal energy and hence equal
probability (P = 21 ). Thus
3
J HH = 21 (3 J t + 3 J g ) = 8.0 Hz (16.2)
For R3 = R4 = CH3 , the steric bulk of the methyl group makes conformations
b and c very unfavourable due to close proximity to the bulky X and Y groups.
Therefore only conformation a is relevant for calculating 3 J HH . Thus
3
J HH = 3 J t = 11.2 Hz (16.3)
These three simultaneous equations, with two unknown variables 3 J t and 3 J g ,

can be solved to estimate the values of the unknowns; note that with three
equations and two unknowns the system is overdetermined.
Using eqn 16.1 and eqn 16.2, 3 J g = 5.9 Hz and 3 J t = 10.1 Hz.
The latter two results are used to give an average value 3 J g = 5.1 Hz and 3 J t =
11.2 Hz.
Using the original Karplus equation, 3 J HH = A cos2 ϕ + B, the gauche and trans
couplings are calculated as
3
J g = A cos2 (60○ ) + B = 14 A + B 3
J t = A cos2 (180○ ) + B = A + B
With the values 3 J g = 5.1 Hz and 3 J t = 11.2 Hz these simultaneous equations

yield A = 8.1 Hz and B = 3.1 Hz. The original Karplus equation is consistent
with these data.
The modern form of the Karplus equation quoted in the text is [16B.14–645],
3
J HH = A + B cos ϕ + C cos 2ϕ. Using the values A = 7 Hz, B = −1 Hz, and
C = 5 Hz, quoted in the text, gives
3
J g = A + B cos (60○ ) + C cos (2 × 60○ ) = A + 21 B − 12 C
= [7 + 12 (−1) − 12 (5)] Hz = 4 Hz
3
J t = A + B cos (180○ ) + C cos (2 × 180○ ) = A − B + C
= [7 − (−1) + (5)] Hz = 13 Hz
There is poor agreement between these values and the ones determined from
the experimental data.
P16B.7 The Karplus equation is given in [16B.14–645] as 3 J HH = A+ B cos ϕ + C cos 2ϕ.

From Fig. 16B.15 on page 646 it is evident that with typical values of the con-
stants A, B, and C the coupling constant is a minimum around ϕ = 90○ .
Differentiating the Karplus equation with respect to ϕ, and then setting the
derivative to zero allows the position of the minimum to be found.
∂ 3 J HH ∂
= [A + B cos ϕ + C cos 2ϕ]
∂ϕ ∂ϕ
= −B sin ϕ − 2C sin 2ϕ
The trigonometric identity sin 2ϕ = 2 sin ϕ cos ϕ is then used to give
∂ 3 J HH
= −B sin ϕ − 4C sin ϕ cos ϕ
∂ϕ
= −(B + 4C cos ϕ) sin ϕ
The derivative is zero when either sin ϕ = 0 or B + 4C cos ϕ = 0. The former of

these solutions corresponds to ϕ = 0 or π, and it is evident from the plot that
these are maxima. For the other solution it follows that cos ϕ = −B/4C, which
is the required result. Given that typically C ≈ 5∣B∣, it follows that ∣ cos ϕ∣ ≪ 1,
making ϕ close to 90○ , as expected.
P16B.9 (a) The structure of cyclohexane, in its most stable chair conformation, is
shown below. The H atoms occupy either ‘axial’ or ‘equatorial’ positions,
and these two environments have different chemical shifts. However, at

room temperature, the molecule undergoes a conformational change in
which this chair form converts to another chair, but in the process the
axial and equatorial H atoms are interchanged. The process is sufficiently
fast at this temperature that the H atoms experience an average environ-
ment and only a single resonance is seen. In other words, the rate constant
for the exchange is large compared to the frequency difference between
the axial and equatorial environments (fast exchange).
Hax Hax
Hax
Heq
Heq Heq
Heq Heq
Heq
Hax
Hax Hax
At lower temperatures, the rate constant for the interconversion of the

two chair forms is reduced, and at sufficiently low temperatures separate
resonances for the axial and equatorial H atoms are seen (slow exchange).
Between the fast and slow exchange limits the system passes through an
intermediate rate regime where the lines are broadened.
(b) The molecule PF5 is trigonal bipyramidal, as shown below. The axial and
equatorial F atoms are not equivalent, but there is a very fast process
which interconverts them (the Berry pseudorotation), and even at the
lowest accessible temperatures the rate constant for this process is suffi-
ciently large that the two fluorine environments are in fast exchange. The
two lines seen in the spectrum are due to coupling to 31P.
Fa
Fe
P Fe
Fe
Fa
(c) Unless water is excluded rigorously from the sample there will exchange
between the OH of the alcohol and the residual water in the solvent. This
process leads to a collapse of the splitting due to any coupling to the
OH proton because its spin state is scrambled as it leaves and returns
to the molecule. The spectrum then consists of a triplet and a quartet,
characteristic of a CH3 CH2 group.
If water is excluded, the exchange is inhibited and then splittings due to
coupling to the OH proton are seen, giving more complex multiplets.
16C Pulse techniques in NMR

D16C.1 Longitudinal relaxation is the result of fluctuating local magnetic fields, and
for such fields to be effective at causing relaxation they must be fluctuating at
frequencies near to the Larmor frequency. In liquids, such fluctuations arise
from the random tumbling motion of the molecules. Typically, such motion
result in fluctuations covering a range of frequenies. As the temperature in-
creases the fluctuations spread over a wider range of frequencies and out to
higher frequencies.
At the lowest temperatures there is little motion at the Larmor frequency and
relaxation is slow. As the temperature increases the frequency range covered
by the motion increases, resulting in more motion at the Larmor frequency,
and hence faster relaxation. However, as the temperature is increased further
still the motion continues to spread over a wider range of frequencies and the
amount of motion at the Larmor frequency decreases, thus leading to slower
relaxation. Thus, there is a temperature at which the relaxation is most rapid
(and the relaxation time at a minimum, see Fig. 16C.9 on page 658).
Small molecules tumble rapidly in non-viscous solvents and at normal tem-
peratures the resulting fluctuations are spread over a wide frequency range,
well beyond the Larmor frequency. Increasing the temperature results in the
fluctuations spreading out even more, such that the amount at the Larmor
frequency is decreased, resulting in slower relaxation, and hence longer lon-
gitudinal relaxation times. In terms of the graph shown in Fig. 16C.9 on page
658 the molecule is to the right of the minimum.
In contrast, large molecules tumble much more slowly, and at normal tem-
peratures the tumbling rate is such that the fluctuations are confined to lower
frequencies. As the temperature increases the fluctuations spread to higher fre-
quencies, therefore increasing the component at the Larmor frequency, leading
to faster relaxation. In terms of the graph, the molecule is to the left of the
minimum.
D16C.3 Homogeneous broadening is the contribution to the linewidth which results

from transverse relaxation. It is the result of local fields modulated by random
molecular motion, and its effects cannot be undone by the application of a spin
echo. Inhomogeneous broadening is the contribution to the linewidth from
the spatial inhomogeneity of the applied magnetic field. If the field varies from
location to location in the sample, so does the Larmor frequency and therefore
the spread of magnetic fields results in a contribution to the linewidth. This
contribution arises from time-independent magnetic fields and, as a result, its
effects can be reversed by the application of a spin echo.
D16C.5 Relaxation results from the interaction between individual spins and local mag-
netic fields which arise in various ways. As is described in Topic 12A, the inter-
action between a spin and a magnetic field depends on the magnetogyric ratio
of the spin: the greater the magnetogyric ratio, the stronger the interaction.
Thus, a given local field has a greater interaction with a nucleus with a greater
magnetogyric ratio and, all other things being equal, the relaxation caused by
these fields will therefore be faster the greater the magnetogyric ratio.
The magnetogyric ratio for 1H is about four times that of 13C, hence the inter-
actions between 1H and local fields is stronger leading to faster relaxation and
shorter relaxation times. A somewhat separate issue is that in typical molecules
H atoms tend to be more accessible than C atoms, and so if the local field is from
outside the molecule it may be felt more strongly by the more exposed H atoms.
E16C.1(a) Ethanoic acid has two carbon environments: a methyl carbon and a carboxylic
acid carbon. The 13C NMR spectrum shows a signal at δ = 21, which is due
to the methyl carbon. This C atom is adjacent to three protons and couples to
them, giving rise to a quartet. The splitting of the quartet is the one-bond C–H
coupling: 1 J = 130 Hz.
The signal at δ = 178 is due to the carboxylic acid carbon. This C atom also
couples to the three methyl protons, but this time over two bonds which results
in a much smaller coupling. As with the methyl carbon, a quartet is seen; the
splitting is the much smaller two-bond C–H coupling.
When the 13C NMR spectrum is recorded with proton decoupling, the split-
tings due to C–H couplings are collapsed, leaving two singlets.
E16C.2(a) The maximal NOE enhancement is given by [16C.8–661], η = γ X /2γ A , where
γ X and γ A are the magnetogyric ratios of nuclei X and A, respectively.
γ1 H (26.752 × 107 T−1 s−1 )
η= = = 1.234
2γ 31 P 2 × (10.840 × 107 T−1 s−1 )
E16C.3(a) From the discussion in Section 16C.1(a) on page 654, the flip angle ϕ of a pulse
of duration ∆τ is given by ϕ = γN B1 ∆τ, where B1 is the strength of the applied
field. It follows that B1 = ϕ/γN ∆τ. With the data given
ϕ π
B1 = = = 9.40 × 10−4 T
γN ∆τ (26.752 × 107 T−1 s−1 ) × (12.5 × 10−6 s)
The corresponding 90○ pulse has half the flip angle so, for the same B1 , ∆τ is
halved: ∆τ = 21 × (12.5 µs) = 6.25 µs .
E16C.4(a) The effective transverse relaxation time, T2∗ , is given by [16C.6–658], T2∗ =
1/π∆ν 1/2 , where ∆ν 1/2 is the width of the (assumed) Lorentzian signal mea-
sured at half the peak height. Hence
1 1
T2∗ = = = 0.21 s
π∆ν 1/2 π × (1.5 Hz)
E16C.5(a) The amplitude of the free induction decay, S(t), is proportional to e−t/T2 , where
t is time and T2 is the transverse relaxation time . If S(t 2 ) = 21 S(t 1 ) then
S(t 2 ) 1 e−t 2 /T2

= = = e−(t 2 −t 1 )/T2 = e−∆t/T2
S(t 1 ) 2 e−t 1 /T2
taking logarithms gives
−∆t ∆t
ln 12 = −∆t/T2 therefore T2 = =
ln 12 ln 2
With the data given
(1.0 s)
T2 = = 1.4 s
ln 2
P16C.1 (a) The flip angle ϕ of the pulse is given by ϕ = ∆τγ N B1 (Section 16C.1(a) on
page 654). Therefore, increasing the pulse duration ∆τ increases the flip
angle. The strongest signal will be observed after application of a 90○ pulse
because such a pulse rotates the magnetization completely to the xy-plane.
A 180○ pulse rotates the magnetization from +z to −z and therefore creates
no transverse magnetization, and hence no observable signal.
From the description of the experiment it is clear that pulses of duration
2.5 µs and 5.0 µs correspond to flip angles of less than 90○ , and about 90○ ,
respectively. The 7.5 µs pulse must correspond to a flip angle of more than
90○ because the signal is weaker than for a shorter pulse. The 10.0 µs pulse
gives no signal and so must correspond to a 180○ pulse.
(b) The 10.0 µs pulse gives no signal and so corresponds to a flip angle of 180○ .
A 90○ pulse has half the duration of a 180○ pulse, that is ∆τ 90 = 5.0 µs .
(c) The magnetic field strength B1 is determined by rearranging ϕ = ∆τγ N B1
to B1 = ϕ/∆τγ N . The length of the 180○ pulse is normally used in the
calculation of B1 because such a pulse gives a null signal, which is easier
to determine than the maximum seen for a 90○ pulse. If the duration of
the 180○ pulse is ∆τ 180 it follows that B1 = π/∆τ 180 γ N . Substituting this
expression into that for the B1 Larmor frequency gives
γ N B1 γ N π 1
ν′L = = × =
2π 2π ∆τ 180 γ N 2∆τ 180
1
= = 5.00 × 104 Hz
2 × (10.0 × 10−6 s)
An alternative way at looking at this is to say that the magnetization com-
pletes a full rotation in 20.0 µs (a 360○ pulse), so the period of the rotation
is 20.0 µs. The corresponding frequency is 1/period = 1/(20.0×10−6 s) =
50.0 kHz. This is the B1 Larmor frequency, the frequency at which the
magnetization rotates about B1 .
P16C.3 To use the discrete Fourier transform the function needs to be generated as an
array of points, regularly sampled in time. For the plots shown here 512 time-
domain points were generated out to a maximum time of 3.0 s (some Fourier
transform routines will only transform numbers of data points which are a
power of 2, 29 = 512); the interval between the points is therefore (3.0 s)/512 =
5.86 × 10−3 s. This means that in the spectrum the maximum frequency which
is represented correctly is 1/(2 × 5.86 × 10−3 s) = 85 Hz; this determines the
scale on the spectra.
Figure 16.5 shows the free-induction decay and spectrum for the parameters
given in the Problem. The peak at 10 Hz is higher than that at 50 Hz both because
its intensity is greater and also because it is narrower on account of having a
longer T2 ; recall that the area under the peak is constant, so if it broadens the
peak height decreases.
In Fig. 16.6 the high-frequency peak has been moved from 50 Hz to 60 Hz,
and its intensity has been increased from 1 to 3; its peak height is still lower
than the peak at 10 Hz on account of the higher-frequency resonance having a
shorter T2 and hence a broader line. In Fig. 16.7 the T2 of the high-frequency
peak is reduced from 0.5 s to 0.2 s, as is evident from the broadening of the
line and the reduction in its peak height. In all three cases it is evident that
the time-domain signal is changing, but only the frequency-domain signal, the
spectrum, is readily interpretable.
0.0 0.5 1.0 1.5 2.0 2.5 3.0

t/s
0 20 40 60 80 f / Hz
Figure 16.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0

t/s
0 20 40 60 80 f / Hz
Figure 16.6
0.0 0.5 1.0 1.5 2.0 2.5 3.0

t/s
0 20 40 60 80 f / Hz
Figure 16.7
P16C.5 The shape of a spectral line, I(ω), is given by
∞
I(ω) = Re [∫ G(t)eiωt dt]
0
Using the suggested substitution, the free-induction decay G(t) = ae−t/τ cos ω 0 t
is rewritten
G(t) = 12 ae−t/τ (e−iω 0 t + eiω 0 t )
∞
1 −t/τ
I(ω) = aRe [∫ 2
e (e−iω 0 t + eiω 0 t ) eiωt dt]
0
∞
= 21 aRe [∫ eit(ω−ω 0 +i/τ) + eit(ω+ω 0 +i/τ) dt]
0
∞
1 1
= 1
aRe ∣ eit(ω−ω 0 +i/τ) + eit(ω+ω 0 +i/τ) ∣
2
i(ω − ω 0 + i/τ) i(ω + ω 0 + i/τ) 0
1 1
= − 12 aRe [ + ]
i(ω − ω 0 + i/τ) i(ω + ω 0 + i/τ)
i i
= 21 aRe [ + ]
(ω − ω 0 + i/τ) (ω + ω 0 + i/τ)
i(ω − ω 0 − i/τ) i(ω + ω 0 − i/τ)
= 21 aRe [ + ]
(ω − ω 0 + i/τ)(ω − ω 0 − i/τ) (ω + ω 0 + i/τ)(ω + ω 0 − i/τ)
i(ω − ω 0 ) + 1/τ i(ω + ω 0 ) + 1/τ
= 21 aRe [ + ]
(ω − ω 0 ) + 1/τ
2 2 (ω + ω 0 )2 + 1/τ 2
a 1 1
= [ + ]
2τ (ω − ω 0 )2 + 1/τ 2 (ω + ω 0 )2 + 1/τ 2
aτ 1 1
= [ 2 + 2 ]
2 τ (ω − ω 0 ) + 1 τ (ω + ω 0 )2 + 1
2
The Lorentzian lineshape is of the form I L = S 0 T2 /[1 + T22 (ω − ω 0 )2 ], and this

lineshape has a width at half-height of 2/T2 in angular frequency units. Com-
parison with the result above shows that the spectrum contains two Lorentzian
peaks at ω = ±ω 0 , with width 2/τ. Two peaks arise in the spectrum because
the original free-induction decay, being of the form of a cosine, is invariant to
the sign of ω 0 .
P16C.7 Note that there is an error in the Problem: the initial condition should be M z (0) =
−M 0
(a) Separate the differential equation and integrate between t = 0 and t = τ,

and the corresponding limits M z (0) = −M 0 and M z (τ).

M z (τ) 1 τ 1
∫ dM z (t) = − ∫ dt
−M 0 M z (t) − M 0 0 T1
M (τ) t τ
[ln (M z (t) − M 0 )]−Mz 0 = [− ]
T1 0
M z (τ) − M 0 τ
ln ( )=−
−2M 0 T1
hence M z (τ) = M 0 (1 − 2e−τ/T1 )
(b) Figure 16.8 shows plots of M z (τ)/M 0 = 1 − 2e−τ/T1 for different values
of T1 . It is evident that increasing T1 increases the time required for the
z-component of magnetisation to relax back to its equilibrium value, M 0 .
1.0 T1 = 0.2 s
T1 = 1.0 s
0.5 T1 = 2.0 s
M z (τ)/M 0
0.0
−0.5
−1.0
0 1 2 3 4 5
τ/ s
Figure 16.8
(c) From the result above, M z (τ) = M 0 (1−2e−τ/T1 ), it follows that M z (τ)/M 0
= (1−2e−τ/T1 ), and hence − 21 [M z (τ)/M 0 −1] = e−τ/T1 . Taking logarithms
gives ln {[M 0 − M z (τ)]/2M 0 } = −τ/T1 . It therefore follows that a plot of
ln {[M 0 − M z (τ)]/2M 0 } against τ is expected to be a straight line of slope
= 1/T1 .
(d) The data are shown in the table below and the plot in Fig 16.9.
τ/ s M z (τ)/M 0 ln {[M 0 − M z (τ)]/2M 0 }
0.000 −1.000 0.000
0.100 −0.637 −0.200
0.200 −0.341 −0.400
0.300 −0.098 −0.600
0.400 0.101 −0.800
0.600 0.398 −1.201
0.800 0.596 −1.599
1.000 0.729 −1.999
1.200 0.819 −2.402
ln {[M 0 − M z (τ)]/2M 0 }
−1
−2
0.0 0.2 0.4 0.6 0.8 1.0 1.2

τ/ s
Figure 16.9
The data are a good fit to the equation
ln {[M 0 − M z (τ)]/2M 0 } = (−2.001) × (τ/ s) + 3.206 × 10−4
The value of T1 is found from the slope
1 1
T1 = − =− = 0.500 s
(slope) (−2.001 s−1 )
P16C.9 (a) Separate the differential equation and integrate between t = 0 and t = τ,
and the corresponding limits M xy (0) and M xy (τ).
M xy (τ) 1 τ 1
∫ dM xy (t) = − ∫ dt
M xy (0) M xy (t) 0 T2
M xy (τ) t τ
∣ln M xy (t)∣ M (0) = ∣− ∣
xy T2 0
M xy (τ) τ
ln ( )=−
M xy (0) T2
hence M xy (τ) = M xy (0)e−τ/T2
(b) Divide both sides of this expression by M xy (0) and take the natural log-
arithm to give
M xy (τ) τ
ln =−
M xy (0) T2
Hence a plot of ln [M xy (τ)/M xy (0)] against τ is expected to be a straight

line with slope = −1/T2 .
(c) The data are shown in the table below and the plot in Fig 16.10.
τ/ ms M xy (τ)/M 0 ln [M xy (τ)/M xy (0)]

10.0 0.819 −0.200
20.0 0.670 −0.400
30.0 0.549 −0.600
50.0 0.368 −1.000
70.0 0.247 −1.398
90.0 0.165 −1.802
110.0 0.111 −2.198
130.0 0.074 −2.604
0
ln [M xy (τ)/M xy (0)]
−1
−2
−3
0 20 40 60 80 100 120 140
τ/ ms
Figure 16.10
The data are a good fit to the equation
ln [M xy (τ)/M xy (0)] = (−0.0200) × (τ/ ms) + 8.1662 × 10−4
The value of T2 is found from the slope,
1 1
T2 = − =− = 50.0 ms
slope −0.0200 ms−1
P16C.11 The magnitude of the local magnetic field is given by [16B.17a–650]
γ N ħµ 0 m I
Bloc = − (1 − 3 cos2 θ)
4πR 3
where Bloc is the local magnetic field, γ N is the nuclear magnetogyric ratio,
µ 0 is the vacuum permeability, m I ħ is the z-component of the spin angular
momentum, R is the internuclear separation, and θ is the angle between the
direction of local field vector and the direction of the applied magnetic field.
Rearrange this expression to give R, assume that θ = 0 so that (1 − 3 cos2 θ) =
−2, and take m I = + 12

1/3
γ N ħµ 0 m I
R = [− (1 − 3 cos2 θ)]
4πB0
γ N ħµ 0 1/3
=[ ]
4πB0
1/3
(26.752 × 107 T−1 s−1 )×(1.0546 × 10−34 J s)×(1.257 × 10−6 T2 J−1 m3 )
=[ ]
4π×(0.715 × 10−3 T)
= 1.58 × 10−10 m = 158 pm
16D Electron paramagnetic resonance

D16D.1 The hyperfine parameter a for the coupling between a nucleus and the elec-
tron in a radical is easily measured from the splittings of the lines in the EPR
spectrum caused by the nucleus in question. If the quantitative relationship
between a and ρ is know (for example, as it is for aromatic systems, the Mc-
Connell equation) it is possible to map out the spin density in the molecule, thus
revealing the spatial distribution of the molecular orbital which is occupied by
the unpaired electron.
E16D.1(a) The total local magnetic field is given by [16D.3–664], Bloc = B0 + am I , where
m I = ± 12 and a is the hyperfine coupling constant. The centre of the spectrum is
at 332.5 mT. Proton 1 splits the centre line into two components with separation
a 1 = 2.0 mT.
Bloc = B0 + a 1 m I = (332.5 mT) ± 12 × (2.0 mT) = (332.5 ± 1.0) mT
Hence, B loc,+1/2 = 331.5 mT or B loc,−1/2 = 333.5 mT. Coupling to proton 2

splits each of these lines into two with a 2 = 2.6 mT, so the four lines occur at
Bloc = B loc,+1/2 +a 2 m I = (331.5 mT)± 21 ×(2.6 mT) = 330.2 mT and 332.8 mT
Bloc = B loc,−1/2 +a 2 m I = (333.5 mT)± 21 ×(2.6 mT) = 332.2 mT and 334.8 mT

The four lines have equal intensity .
E16D.2(a) (i) The three protons in the ⋅CH3 radical are equivalent, each with spin quan-
tum number I = 12 . If a radical contains N equivalent nuclei with spin I,
then there are 2N I + 1 hyperfine lines with an intensity distribution given
by Pascal’s triangle. Hence for the CH3 radical, there are (2×3× 12 +1) = 4
hyperfine lines with intensity distribution 1 ∶ 3 ∶ 3 ∶ 1 .
(ii) The three 2 H atoms in the ⋅CD3 radical are equivalent, each with spin
quantum number I = 1. The number of hyperfine lines is (2×3×1+1) = 7,
and their intensity distribution is the same as that shown in Fig. 16D.5 on
page 665, 1 ∶ 3 ∶ 6 ∶ 7 ∶ 6 ∶ 3 ∶ 1 .
E16D.3(a) The resonance frequency is related to the g-value by [16D.2–663], hν = g µB B0 .

Rearranging this gives B0 = hν/µB .
(i) For a spectrometer operating at 9.313 GHz,
hν (6.6261 × 10−34 J s) × (9.313 × 109 Hz)

B0 = = = 332.3 mT
g µB 2.0025 × (9.2740 × 10−24 J T−1 )
(ii) For a spectrometer operating at 33.80 GHz,
hν (6.6261 × 10−34 J s) × (33.80 × 109 Hz)

B0 = = = 1.206 T
g µB 2.0025 × (9.2740 × 10−24 J T−1 )
E16D.4(a) If a radical contains N equivalent nuclei with spin quantum number I, then
there are 2N I + 1 hyperfine lines. The radical contains a single magnetic nu-
cleus, N = 1, and there are a total of four hyperfine lines. Therefore 2N I + 1 =
2 × 1 × I + 1 = 4 hence I = 32 .
E16D.5(a) The EPR resonance condition is given by [16D.2–663], hν = g µ B B0 , where ν is

the spectrometer operating frequency, g is the g-value of the radical, µ B is the
Bohr magneton, and B0 is the magnetic field. Rearranging the expression gives
hν (6.6261 × 10−34 J s) × (9.2231 × 109 Hz)

g= = = 2.0022
µ B B0 (9.2740 × 10−24 J T−1 ) × (329.12 × 10−3 T)
E16D.6(a) As described in Section 16D.2 on page 664, the splitting between the lines in
the multiplet is the hyperfine coupling constant. In this case the splitting is
(334.8 − 332.5) mT = 2.3 mT and (332.5 − 330.2) mT = 2.3 mT, therefore
a = 2.3 mT .
The central line is unaffected by coupling and so is used to compute the g-value
using [16D.2–663], hν = g µB B0
hν (6.6261 × 10−34 J s) × (9.319 × 109 Hz)

g= = = 2.0025
µB B0 (9.2740 × 10−24 J T−1 ) × (332.5 × 10−3 T)
P16D.1 The EPR resonance condition is given by [16D.1–663], hν = g e µ B B0 , where ν
is the spectrometer operating frequency, g e is the g-value of the electron, µ B is
the Bohr magneton and B0 is the magnetic field. The expression is rearranged
to make ν the subject.
g e µ B B0 2.0023×(9.2740 × 10−24 J T−1 )×(1.0 × 103 T)

ν= = = 2.8 × 1013 Hz
h 6.6261 × 10−34 J s
Expressed as a wavenumber this is ν̃ = 924 cm−1 . This frequency is in the
infrared region of the electromagnetic spectrum and hence is comparable to
the separation between the energy levels associated with molecular vibrations .
It is much greater than the separations of rotational levels, but much less than
the separations of electronic energy levels.
P16D.3 (a) The hyperfine coupling is a result of an interaction between the magnetic
moment of a nucleus and that of an electron. The magnetic moment of a
nucleus is proportional to the magnetogyric ratio γ N , so it is not unrea-
sonable to expect the hyperfine coupling a to have the same dependence.
(b) Assuming that a ∝ γ N it follows that
γ N,D 4.1067 × 107 T−1 s−1

a⋅CD3 = × a⋅CH3 = × (2.3 mT) = 0.35 mT
γ N,H 26.752 × 107 T−1 s−1
The ⋅CH3 spectrum consists of 4 lines (quartet, 1:3:3:1) and so has width
3a H . As shown in Fig. 16D.5, the ⋅CD3 spectrum consists of 7 lines (septet,
1:3:6:7:6:3:1) and so has width 6a D . Hence the overall width of the ⋅CH3
spectrum is 3a H = 3 × (2.3 mT) = 6.9 mT , and the overall width of the
⋅CD3 spectrum is 6a D = 6 × (0.35 mT) = 2.1 mT .
P16D.5 The McConnell equation is given in [16D.5–665], a = Qρ, where a is the hy-
perfine coupling constant observed for a H atom in a given molecule, Q is a
constant, and ρ is the spin density on the adjacent carbon atom.
a 0.274 mT
carbon 1 ρ= = = 0.122
Q 2.25 mT
a 0.151 mT
carbon 2 ρ= = = 0.067
Q 2.25 mT
a 0.534 mT
carbon 9 ρ= = = 0.237
Q 2.25 mT
P16D.7 The proportion of time the unpaired electron spends in the N 2s orbital is given
by the ratio of the hyperfine couplings: P(N2s) = (5.7 mT)/(55.2 mT) =
0.103..., that is 10% of the time. The proportion of time the unpaired electron
spends in the N 2p orbital is P(N2pz ) = (1.3 mT)/(3.4 mT) = 0.382..., that is
38% of the time.
The total probability that the electron will be found on the N atom is P(N) =
10% + 38% = 48% . The total probability that the electron will be found on the
O atoms is P(O) = 100% − 48% = 52% .
In Problem P9A.4 it is shown that the angle θ between two sp λ hybrids is related
to the hybridisation ratio λ by λ = ±(−1/ cos θ)1/2 . The hybridization ratio is
given by λ 2 = (p character)/(s character), therefore in the present case λ 2 =
(38%)/(10%) = 3.8, hence λ = 1.95 . The angle between the hybrids is found
by rearranging the above expression to cos θ = −1/λ 2 , hence cos θ = −1/3.8
giving θ = 105○ . Such a hybrid is suitable for bonding an angular arrangement,
and so supports the view that this is indeed the geometry for NO2 .

I16.1 (a) The spin densities, computed by molecular modelling software (ab initio,
density functional theory, Gaussian 98), are shown below
O 0.409
− 0.082
0.299 0.271
− 0.135 − 0.129
0.369
CH3
− 0.024
The computed spin densities for the the two protons ortho to the oxygen,
and to the two protons meta to the oxygen are different. This is undoubt-
edly a computational artefact, a result of the minimum-energy structure
having one methyl proton in the plane of the ring, which makes the right
and left side of the ring slightly non-equivalent. Fast internal rotation
is expected to make the two meta protons, and the two ortho protons,
equivalent. The spin density at the ortho carbons is therefore taken as
0.285, and that at the meta carbons as −0.132.
(b) The hyperfine coupling constants resulting from coupling to the ring pro-
tons are computed using the McConnell equation, [16D.5–665]: a = Qρ,
with Q = 2.25 mT. Hence a ortho = 0.285 × (2.25 mT) = 0.641 mT,
and a meta = 0.132 × (2.25 mT) = 0.297 mT. The hyperfine coupling for
interaction with the methyl protons cannot be calcuated in the same way
but is given by the software as 1.045 mT. The EPR spectrum will show
a rather complex multiplet arising from splitting into a 1:2:1 triplet as a
result of the coupling to the ortho protons, each line is then split into a
1:2:1 triplet as a result of the coupling to the meta protons, and each line
is split once more into a 1:3:3:1 quartet as a result of the coupling to the
methyl protons.
I16.3 Terms such as [I], [EI], [E], νI , νEI and ν need to be eliminated from the ex-
pressions. First, note that total I (free I plus bound I) is the same as initial I;
similarly, total E (free E plus bound E) is the same as initial E.
[I]tot = [I]0 = [I] + [EI] and [E]tot = [E]0 = [E] + [EI]
Using this, the expression for KI is rewritten
[E][I] ([E]0 − [EI])([I]0 − [EI])
KI = =
[EI] [EI]
Because [I]0 >> [E]0 it follows that [I]0 must also be much greater than [EI],
even if all of the E binds I. Hence [I]0 − [EI] ≈ [I]0 . This is substituted into the
above expression which is then solved for [EI].
([E]0 − [EI])[I]0 [E]0 [I]0
KI = hence [EI] = (16.4)
[EI] KI + [I]0
The expression for ν is written using [I] = [I]0 − [EI] to give
[I]νI + [EI]νEI νI ([I]0 − [EI]) + νEI [EI]

ν = fI νI + fEI νEI = =
[I] + [EI] [I]0
This is rearranged to given an expression for [EI], and then [EI] is substituted
using eqn 16.4
(ν − νI )[I]0 δν[I]0 [E]0 [I]0 δν[I]0

[EI] = = hence =
νEI − νI ∆ν KI + [I]0 ∆ν
where δν and ∆ν as as defined in the text. Finally, [I]0 is made the subject to
give
∆ν
[I]0 = [E]0 − KI
δν
It follows from this that a plot of [I]0 against 1/(δν) is a straight line with slope
[E]0 ∆ν and y-intercept −KI .
17 Molecular Interactions
17A Electric properties of molecules

D17A.1 Dipole moments are not measured directly, but are calculated from a mea-
surement of the relative permittivity, ε r = ε/ε 0 of the medium as measured by
comparing the capacitance of a capacitor with and without the sample present
using ε r = C/C 0 . From [17A.10–680] and [17A.11–680] it follows that the dipole
moment can be determined from a measurement of ε r as a function of tem-
perature (see Example 17A.2 on page 681). An alternative method involves
measuring the refractive index and then using [17A.13–682] to relate this to
ε r . It is also possible to infer a value of the dipole moment by measuring the
splitting seen in the rotational (microwave) spectra of some molecules when a
field is applied (the Stark effect).
D17A.3 When the applied field changes direction slowly, the permanent dipole moment
has time to reorientate and so follows the field: the whole molecule rotates
into a new direction in response to a change in the field. However, when the
frequency of the field is high, a molecule cannot change direction fast enough to
follow the change in direction of the applied field and the dipole moment then
makes no contribution to the polarization of the sample. Because a molecule
takes about 1 ps to turn through about 1 radian in a fluid, the loss of this contri-
bution to the polarization occurs when measurements are made at frequencies
greater than about 1011 Hz (in the microwave region). It is said that orientation
polarization, the polarization arising from the permanent dipole moments, is
lost at such high frequencies
The next contribution to the polarization to be lost as the frequency is raised
is the distortion polarization, the polarization that arises from the distortion of
the positions of the nuclei by the applied field. The molecule is distorted by
the applied field, and the molecular dipole moment changes accordingly. The
time taken for a molecule to bend is approximately the inverse of the molec-
ular vibrational frequency, so the distortion polarization disappears when the
frequency of the radiation is increased through the infrared.
At even higher frequencies, in the visible region, only the electrons are mobile
enough to respond to the rapidly changing direction of the applied field. The
polarization that remains is now due entirely to the distortion of the electron
distribution, and the surviving contribution to the molecular polarizability is
called the electronic polarizability.
626 17 MOLECULAR INTERACTIONS
E17A.1(a) The relationship between the induced dipole moment µ∗ and the electric field
strength E is given by [17A.5a–677], µ ∗ = αE, where α is the polarizability.
The polarizability volume α ′ is related to the polarizability α by [17A.6–677],
α ′ = α/4πε 0 . Combining these equations, rearranging for E, and using 1 V =
1 J C−1 gives
µ∗ (1.0 × 10−6 D) × [(3.3356 × 10−30 C m)/(1 D)]

E= ′
=
4πε 0 α 4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (2.6 × 10−30 m3 )
= 1.2 × 104 V m−1
E17A.2(a) The molar polarization Pm is defined by [17A.11–680], Pm = (N A /3ε 0 )(α +

µ 2 /3kT), where α is the polarizability of the molecule and µ is its dipole mo-
ment. This equation is written as
NA α NA µ2 1
Pm = +
3ε 0 9ε 0 k T
which implies that a graph of Pm against 1/T should be a straight line with
slope N A µ 2 /9ε 0 k and intercept N A α/3ε 0 . However, as there are only two data
points it is convenient to calculate the required quantities directly from the
data. Writing the molar polarization at the two temperatures as Pm (T1 ) and
Pm (T2 ) and considering Pm (T2 ) − Pm (T1 ) gives
NA µ2 1 1
Pm (T2 ) − Pm (T1 ) = ( − )
9ε 0 k T2 T1
which is rearranged to give
1/2
9ε 0 k Pm (T2 ) − Pm (T1 )
µ=( × )
NA 1/T2 − 1/T1
9 × (8.8542 × 10−12 J−1 C2 m−1 ) × (1.3806 × 10−23 J K−1 )
=(
6.0221 × 1023 mol−1
1/2
(62.47 × 10−6 m3 mol−1 ) − (70.62 × 10−6 m3 mol−1 )
× )
1/(423.2 K) − 1/(351.0 K)
= 5.53... × 10−30 C m = 1.659 D
The value of α is found using this value of µ together with one of the data points;
both give the same answer. Rearranging [17A.11–680], Pm = (N A /3ε 0 )(α +
µ 2 /3kT), for α and using the data for 351.0 K gives
3ε 0 Pm µ2
α= −
NA 3kT
3 × (8.8542 × 10−12 J−1 C2 m−1 ) × (70.62 × 10−6 m3 mol−1 )
=
6.0221 × 1023 mol−1
(5.53... × 10−30 C m)2
− = 1.008 × 10−39 C2 m2 J−1
3 × (1.3806 × 10−23 J K−1 ) × (351.0 K)
E17A.3(a) The relationship between relative permittivity and molar polarization is given
by the Debye equation, [17A.10–680], (ε r − 1)/(ε r + 2) = ρPm /M. Rearranging
gives
ρPm ρPm 2ρPm
εr − 1 = (ε r + 2) hence ε r − 1 = ε r ( )+
M M M
ρPm 2ρPm 1 + 2ρPm /M
hence ε r (1 − )=1+ hence ε r =
M M 1 − ρPm /M
The molar mass of ClF3 is M = 92.45 g mol−1 , which gives
1 + 2ρPm /M
εr =
1 − ρPm /M
1 + 2 × (1.89 g cm−3 ) × (27.18 cm3 mol−1 )/(92.45 g mol−1 )
= = 4.75
1 − (1.89 g cm−3 ) × (27.18 cm3 mol−1 )/(92.45 g mol−1 )
E17A.4(a) The relationship between the refractive index n r at a specified wavelength and
the relative permittivity ε r at the same wavelength is given by [17A.13–682], n r =
1/2
ε r , hence ε r = n r2 . In addition the relationship between relative permittivity
and the polarizability α is given by the Clausius–Mossotti equation, [17A.12–
680], (ε r − 1)/(ε r + 2) = ρN A α/3Mε 0 . In using this equation it is assumed
that there are no contributions from permanent electric dipole moments to
the polarization, either because the molecules are nonpolar, which is not the
case for CH2 I2 , or because the frequency of the applied field is so high that the
molecules cannot orientate quickly enough to follow the change in direction of
the field. Replacing ε r by n r2 in the Clausius–Mossotti equation gives
n r2 − 1 ρN A α 3Mε 0 n r2 − 1
= hence α= × 2
n r2 + 2 3Mε 0 ρN A nr + 2
The molar mass of CH2 I2 is M = 267.8258 g mol−1 which gives
3 × (267.8258 g mol−1 ) × (8.8542 × 10−12 J−1 C2 m−1 ) 1.7322 − 1

α= ×
(3.32 × 106 g m−3 ) × (6.0221 × 1023 mol−1 ) 1.7322 + 2
= 1.42 × 10−39 C2 m2 J−1
E17A.5(a) The relationship between the refractive index n r at a specified wavelength and
the relative permittivity ε r at the same wavelength is given by [17A.13–682],
1/2
n r = ε r . In order to find ε r the Clausius–Mossotti equation, [17A.12–680],
(ε r − 1)/(ε r + 2) = ρN A α/3Mε 0 , is used, with the value of the polarizability
α being determined from the polarizability volume α ′ using [17A.6–677], α ′ =
α/4πε 0 .
In using the Clausius–Mossotti equation it is assumed that there are no con-
tributions from permanent electric dipole moments to the polarization, either
because the molecules are nonpolar or because the frequency of the applied
field is so high that the molecules cannot orientate quickly enough to follow
the change in direction of the field.
The first step is to rearrange the Clausius–Mossotti equation for ε r
ε r − 1 ρN A α ρN A α
= hence εr − 1 = (ε r + 2)
ε r + 2 3Mε 0 3Mε 0
ρN A α 2ρN A α 1 + 2ρN A α/3Mε 0
hence ε r (1 − )=1+ hence ε r =
3Mε 0 3Mε 0 1 − ρN A α/3Mε 0
Replacing α by 4πε 0 α ′ and ε r by n r2 gives
1 + 2ρN A (4πε 0 α ′ )/3Mε 0 1 + 8πρN A α ′ /3M 1 + 2C

n r2 = = =
1 − ρN A (4πε 0 α ′ )/3Mε 0 1 − 4πρN A α ′ /3M 1−C
where C = 4πρN A α ′ /3M. Taking the mass density of water from the Resource
section as 0.997 g cm−3 and the molar mass as M = 18.0158 g mol−1
C = 4πρN A α ′ /3M
4π × (0.997 g cm−3 ) × (6.0221 × 1023 mol−1 ) × (1.5 × 10−24 cm3 )
=
3 × (18.0158 g mol−1 )
= 0.209...
1 + 2C 1/2 1 + 2 × 0.209... 1/2
nr = ( ) =( ) = 1.3
1−C 1 − 0.209...
E17A.6(a) The Debye equation, [17A.10–680], is (ε r −1)/(ε r +2) = ρPm /M, where Pm is the
molar polarizability. The latter is defined by [17A.11–680], Pm = (N A /3ε 0 )(α +
µ 2 /3kT), where µ is the dipole moment and α is the polarizability, which is
related to the polarizability volume α ′ according to [17A.6–677], α ′ = α/4πε 0 .
Replacing α in the expression for Pm by 4πε 0 α ′ gives
NA µ2
Pm = (4πε 0 α ′ + )
3ε 0 3kT
6.0221 × 1023 mol−1 ⎛
= −12 −1 2 −1
× 4π × (8.8542 × 10−12 J−1 C2 m−1 )
3 × (8.8542 × 10 J C m ) ⎝
2
−29 3
[(1.57 D) × (3.3356 × 10−30 C m)/(1 D)] ⎞
× (1.23 × 10 m )+ −1
3 × (1.3806 × 10−23 J K ) × ([25 + 273.15] K) ⎠
= 8.13... × 10−5 m3 mol−1
The Debye equation is then rearranged for ε r
ε r − 1 ρPm 1 + 2ρPm /M
= hence εr =
εr + 2 M 1 − ρPm /M
The molar mass of chlorobenzene, C6 H5 Cl is 112.5495 g mol−1 . Noting that ρ

needs to be in g m−3 in order for the units to cancel appropriately, the relative
permittivity is
1 + 2ρPm /M
εr =
1 − ρPm /M
1 + 2 × (1.173 × 106 g m−3 ) × (8.13... × 10−5 m3 mol−1 )/(112.5495 g mol−1 )
=
1 − (1.173 × 106 g m−3 ) × (8.13... × 10−5 m3 mol−1 )/(112.5495 g mol−1 )
= 17.8
E17A.7(a) The molecules are shown in Fig. 17.1.
δ-
F I
δ- δ-
C δ+ δ+ O O
O
3δ+ δ- δ- δ+
F F H δ+ H
δ- δ- O O
Figure 17.1
In none of the cases do the dipole moments associated with each bond cancel,
so all three molecules are polar . Note that the O–O bonds in ozone are polar
because the central atom is different from the other two. In the case of CIF3
the dipole moment associated with the C–I bond is much less than that for the
C–F bonds, which results in the dipole moments not cancelling.
An alternative approach is use a symmetry argument. As explained in Sec-
tion 10A.3(a) on page 372, only molecules belonging to point groups C n , C nv or
C s may have a permanent electric dipole moment. CIF3 , O3 and H2 O2 belong
to point groups C 3v , C 2v and C 2 respectively, so all three are polar.
E17A.8(a) The magnitude of the resultant dipole moment, µres , is given by [17A.3a–676],
µres = (µ 12 + µ 22 + 2µ 1 µ 2 cos θ)1/2 .
1/2
µres = ((1.5 D)2 + (0.8 D)2 + 2 × (1.5 D) × (0.8 D) × cos 109.5○ ) = 1.4 D
E17A.9(a) The arrangement of charges is shown on the left of Fig. 17.2

y/nm
6
−2e µx
u
µθ µ y

0! 3
.2 ! !
!
!○ 9
?
3e u
!! 20
−e
u -
0.32 x/nm
Figure 17.2
The x component of the dipole are given by [17A.4a–676], µ x = ∑J Q J x J , and

similarly for the y and z components; note that in this case µ z = 0 because
all of the point charges have a z coordinate of zero. The components are then
combined using [17A.4b–676], µ = (µ 2x + µ 2y + µ z2 )1/2 , which is represented
graphically on the right half of Fig. 17.2.
µ x = ∑ Q J x J = [ − (1.6022 × 10−19 C) × (0.32 × 10−9 m)

J
− 2 × (1.6022 × 10−19 C) × (0.23 × 10−9 m) × cos 20○ ]

1D
× = −36.1... D
3.3356 × 10−30 C m
where 1 D = 3.3356 × 10−30 C m is used from inside the front cover.
µ y = ∑ QJ yJ
J
= −2 × (1.6022 × 10−19 C) × (0.23 × 10−9 m) × sin 20○

1D
× = −7.55... D
3.3356 × 10−30 C m
The magnitude of the resultant is
1/2
µ = (µ 2x + µ 2y ) = [(−36.1... D)2 + (−7.55... D)2 ] = 37 D
and, from the diagram on the right of Fig. 17.2, the direction is given by
7.55... D
θ = tan−1 ( ) = 12○
31.6... D
P17A.1 The molecules are shown in Fig. 17.3.
CH3
CH3
CH3 CH3
CH3
CH3 CH3
toluene 1,2 isomer 1,3 isomer 1,4 isomer
Figure 17.3
The dipole moments of the three isomers of dimethylbenzene are estimated

using the method described in Section 17A.1 on page 675. In each case the
resultant dipole moment is assumed to be that of a vector sum of two toluene
dipole moments µ 1 arranged at the appropriate angle to each other. In the case
of two contributing dipoles of equal magnitude at angle Θ to each other, the
size of the resultant dipole is given by [17A.3b–676], µres = 2µ 1 cos( 21 Θ). In

this case µ 1 = 0.4 D, the dipole moment in toluene, and the angle Θ is 60○ ,
120○ or 180○ for the 1,2-, 1,3- and 1,4- isomers respectively.
1,2 isomer: µres = 2µ 1 cos( 12 Θ) = 2 × (0.4 D) × cos( 21 × 60○ ) = 0.7 D

1,3 isomer: µres = 2µ 1 cos( 21 Θ) = 2 × (0.4 D) × cos( 21 × 120○ ) = 0.4 D
1,4 isomer: µres = 2µ 1 cos( 21 Θ) = 2 × (0.4 D) × cos( 12 × 180○ ) = 0
It is possible to be certain about the result for the 1,4 isomer, because the sym-
metry of the molecule means that the dipoles must cancel and so µres is neces-
sarily zero.
P17A.3 The individual ethanoic acid molecules have a permanent dipole moment but
the dimers do not because of their symmetry which cause the individual dipoles
associated with polar bonds to cancel. As the temperature increases the dimer
⇌ monomer equilibrium shifts in favour of the monomers because the pro-
cess of breaking the hydrogen bonds between the molecules is endothermic.
Consequently a greater proportion of the species present are monomers and
therefore have a dipole, and consequently the apparent dipole moment per
molecule appears to increase.
P17A.5 The relationship between induced dipole moment µ ∗ and electric field strength
E is given by [17A.5a–677], µ ∗ = αE where α is the polarizability. The polar-
izability volume α ′ is related to α by [17A.6–677], α ′ = α/4πε 0 , so the induced
dipole is
µ ∗ = αE = 4πε 0 α ′ E
= 4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (2.22 × 10−30 m3 )×(15.0 × 103 V m−1 )
= 3.70... × 10−36 C m = 1.11... × 10−6 D = 1.11 µD
P17A.7 The temperature-dependence of molar polarization is given by [17A.11–680],
NA α NA µ2 1 NA µ2
Pm = + × or Pm = 43 πN A α ′ +
3ε 0 9ε 0 k T 9ε 0 k
where in the second form the definition of polarizability volume, which is given
by [17A.6–677], α ′ = α/4πε 0 , is used to replace α by 4πε 0 α ′ . This equation im-
plies that a graph of Pm against 1/T should be a straight line of slope N A µ 2 /9ε 0 k
and intercept 34 πN A α ′ .
The molar polarization is calculated using the Debye equation [17A.10–680],

(ε r −1)/(ε r +2) = ρPm /M, which is rearranged to Pm = (M/ρ)(ε r −1)/(ε r +2).
Taking the molar mass of trichloromethane as M = 119.3679 g mol−1 the values
of Pm are as shown in the table below and are plotted in Fig. 17.4.
θ/○ C ε r ρ/g cm−3 1/(T/K) Pm /cm3 mol−1

−80 3.1 1.65 0.005 18 29.79
−70 3.1 1.64 0.004 92 29.97
−60 7.0 1.64 0.004 69 48.52
−40 6.5 1.61 0.004 29 47.97
−20 6.0 1.57 0.003 95 47.52
0 5.5 1.53 0.003 66 46.81
20 5.0 1.50 0.003 41 45.47
50
Pm /cm3 mol−1
liquid
40
←Ð increasing T
solid
30
0.000 0.001 0.002 0.003 0.004 0.005 0.006

1/(T/K)
Figure 17.4
The first thing to note is that the molar polarization Pm at −80 ○ C and −70 ○ C
is much lower than for the other temperatures. These are the temperatures for
which trichloromethane (freezing point −64 ○ C) is a solid. The lower Pm is
consistent with the molecules in the solid being fixed in position so that they
cannot change orientation to respond to the electric field and therefore their
permanent dipole moment makes no contribution to the polarization of the
sample; that is, there is no orientation polarization. In the liquid state there is
a contribution from orientation polarization and therefore the values of Pm are
larger.
For the liquid state, the plot of Pm against 1/T should be linear with a slope of
N A µ 2 /9ε 0 k and intercept 34 πN A α ′ as deduced earlier. In fact, Fig. 17.4 shows
that the data do not lie very close to a straight line. The line drawn on Fig. 17.4
uses only the points corresponding to the three highest temperatures and has
equation
(Pm /cm3 mol−1 ) = 3.76 × 103 × 1/(T/K) + 32.8
Identifying the slope with N A µ 2 /9ε 0 k gives
N A µ 2 /9ε 0 k = (3.76 × 103 cm3 mol−1 K) = (3.76 × 10−3 m3 mol−1 K)

and hence
1/2
9ε 0 k × (3.76 × 10−3 m3 mol−1 K)
µ=( )
NA
1/2
9 × (8.8542 × 10−12 J−1 C2 m−1 ) × (3.76 × 10−3 m3 mol−1 K)
=( )
6.0221 × 1023 mol−1
= 2.62... × 10−30 C m = 0.79 D
Similarly, identifying 43 πN A α ′ with the intercept gives

4
3
πN A α ′ = (32.8 cm3 mol−1 )
Rearranging gives
3
α′ = × (32.8 cm3 mol−1 )
4πN A
3 3 −1 −23
= −1 × (32.8 cm mol ) = 1.3 × 10 cm3
23
4π × (6.0221 × 10 mol )
Note that the intercept of Fig. 17.4 corresponds approximately to Pm for the
solid. This is interpreted as arising because the intercept gives Pm in the limit
of high temperature, under which conditions the molecules then have so much
thermal energy that they cannot be oriented by the applied electric field. Con-
sequently there is no orientation polarization contribution to Pm , just as is the
case in the solid where the molecules are prevented from rotating.
P17A.9 The temperature-dependence of molar polarization is given by [17A.11–680],
NA α NA µ2 1 NA µ2 1
Pm = + × or Pm = 34 πN A α ′ + ×
3ε 0 9ε 0 k T 9ε 0 k T
where in the second form the definition of polarizability volume, which is given
by [17A.6–677], α ′ = α/4πε 0 , is used to replace α by 4πε 0 α ′ . This equation im-
plies that a graph of Pm against 1/T should be a straight line of slope N A µ 2 /9ε 0 k
and intercept 34 πN A α ′ . The data are plotted in Fig. 17.5.
T/K 1/(T/K) Pm /cm3 mol−1

292.2 0.003 42 57.57
309.0 0.003 24 55.01
333.0 0.003 00 51.22
387.0 0.002 58 44.99
413.0 0.002 42 42.51
446.0 0.002 24 39.59
(Pm /cm3 mol−1 ) = (1.5236 × 104 ) × 1/(T/K) + 5.5432

60
Pm /cm3 mol−1
50
40
0.0024 0.0028 0.0032

1/(T/K)
Figure 17.5
Identifying the slope with N A µ 2 /9ε 0 k gives
N A µ 2 /9ε 0 k = (1.5236 × 104 cm3 mol−1 K) = (1.5236 × 10−2 m3 mol−1 K)
and hence
1/2
9ε 0 k
µ=( × (1.5236 × 10−2 m3 mol−1 K))
NA
9 × (8.8542 × 10−12 J−1 C2 m−1 ) × (1.3806 × 10−23 J K−1 )
=(
6.0221 × 1023 mol−1
1/2
×(1.5236 × 10−2 m3 mol−1 K)) = 5.27... × 10−30 C m = 1.582 D
Similarly, identifying 34 πN A α ′ with the intercept gives
3
α′ = × (5.5432 cm3 mol−1 )
4πN A
3
= × (5.5432 cm3 mol−1 ) = 2.197 × 10−24 cm3
4π × (6.0221 × 1023 mol−1 )
In the case of polarization measurements made with radiation of relatively high

frequency, such as visible light, the permanent electric dipole moment of the
molecules does not contribute to the polarization because the electric field os-
cillates too quickly for the molecules to orientate themselves to follow the field
direction. The difference in molar polarization recorded under these condi-
tions from the total molar polarization therefore corresponds to the permanent
dipole contribution, that is, the term N A µ 2 /9ε 0 kT in the expression for total
molar polarization.
The molar polarisation under the high frequency conditions is determined from
the refractive index measurement using the Debye equation [17A.10–680], (ε r −
1)/(ε r + 2) = ρPm /M, together with the relationship between refractive index
1/2
and relative permittivity which is given by [17A.13–682], n r = ε r . Combining
these equations gives
n 2r − 1 ρPm M n 2r − 1
= hence Pm = ( )
n 2r + 2 M ρ n 2r + 2
The mass density ρ is given by M/Vm , where M is the molar mass and Vm is
the molar volume. Assuming perfect gas behaviour, Vm = RT/p and hence
M n 2r − 1 M n2 − 1 n2 − 1 RT n 2r − 1
Pm = ( 2 )= ( 2r ) = Vm ( 2r )= ( )
ρ nr + 2 M/Vm n r + 2 nr + 2 p n 2r + 2
Noting that 1 Pa = 1 J m−3 , the high frequency molar polarization is therefore
RT n 2r − 1 (8.3145 J K−1 mol−1 ) × (273 K) 1.0003792 − 1

Pm = ( 2 )= × ( )
p nr + 2 100 × 103 Pa 1.0003792 + 2
= 5.73... × 10−6 m3 mol−1 = 5.73 cm3 mol−1
This corresponds to the term N A α/3ε 0 in the expression
NA α NA µ2
Pm = +
3ε 0 9ε 0 kT
because as discussed above the term involving µ does not contribute at high fre-
quency. Assuming the high frequency contribution to be the same at 292 K, the
difference between this value and the static polarization Pm = 57.57 cm3 mol−1
at this temperature corresponds to the term N A µ 2 /9ε 0 kT.
NA µ2
= Pm,static − Pm,high frequency
9ε 0 kT
Hence
1/2
9ε 0 kT
µ=( (Pm,static − Pm,high frequency ))
NA
9 × (8.8542 × 10−12 J−1 C2 m−1 ) × (1.3806 × 10−23 J K−1 ) × (292 K)
=( ×
6.0221 × 1023 mol−1
1/2
[(57.57 × 10−6 m3 mol−1 ) − (5.73... × 10−6 m3 mol−1 )] )
= 5.24... × 10−30 C m = 1.57 D
The result is in good agreement with that obtained from the graph.
P17A.11 The molar polarization is given by [17A.11–680], Pm = (N A /3ε 0 )(α + µ 2 /3kT).

However, assuming that the frequency of oscillation of the electric field is high,
as it will be for visible light, there will be no contribution from orientation po-
larization because the molecules cannot reorientate themselves quickly enough
to follow the field. Therefore the term µ 2 /3kT in the expression for Pm does
not contribute, and the equation reduces to Pm = N A α/3ε 0 .
Taking the polarizability of methanol as α = 3.59 × 10−40 J−1 C2 m2 from Ta-
ble 17A.1 on page 675 gives
N A α (6.0221 × 1023 mol−1 ) × (3.59 × 10−40 J−1 C2 m2 )

Pm = =
3ε 0 3 × (8.8542 × 10−12 J−1 C2 m−1 )
= 8.13... × 10−6 m3 mol−1 = 8.14 cm3 mol−1
The relationship between molar polarization and the relative permittivity is

given by the Debye equation [17A.10–680], (ε r − 1)/(ε r + 2) = ρPm /M. Rear-
ranging for ε r and taking the molar mass of methanol as M = 32.0416 g mol−1
gives
1 + 2ρPm /M
εr =
1 − ρPm /M
1 + 2 × (0.7914 g cm−3 ) × (8.13... cm3 mol−1 )/(32.0416 g mol−1 )
=
1 − (0.7914 g cm−3 ) × (8.13... cm3 mol−1 )/(32.0416 g mol−1 )
= 1.75... = 1.75
The refractive index n r is related

√ to the relative permittivity according to [17A.13–
√
682], n r = ε r . Hence n r = 1.75... = 1.32 .
P17A.13 The monomer–dimer equilibrium is shown in Fig. 17.6.
O O HO
2
OH OH O
Figure 17.6
The relationship between relative permittivity and molar polarization is given

by the Debye equation, [17A.10–680]
ε r − 1 ρPm 1 + 2ρPm /M
= hence εr =
εr + 2 M 1 − ρPm /M
This equation implies that if the molar polarization Pm increases then ε r will
increase, because the numerator in the expression for ε r will increase while the
denominator will decrease. The molar polarization in turn is given by [17A.11–
680], Pm = (N A /3ε 0 )(α + µ 2 /3kT).
The individual molecules are polar but the dimers are not, because their sym-
metry means the individual dipoles associated with their polar bonds cancel
out. At low temperatures, a significant proportion of molecules exist as the
non-polar dimer and therefore do not contribute to the µ 2 /3kT term in the
expression for Pm . As the temperature is raised, the hydrogen bonds hold-

ing the dimers together are broken and the equilibrium shifts in favour of the
monomers. Consequently more molecules contribute to the µ 2 /3kT term in
the expression for Pm and consequently Pm increases. In turn this leads to an
increase in ε r as shown above.
If the sample is diluted, then the equilibrium will also shift in favour of the
monomers, according to Le Chatelier’s principle as discussed in Section 6B.1
on page 198. Consequently the molar polarization and hence the relative per-
mittivity will increase in the same way as when the temperature is increased.
17B Interactions between molecules

D17B.1 There are three van der Waals type interactions that depend upon distance as
1/r 6 ; they are the Keesom interaction between rotating permanent dipoles, the
permanent-dipole–induced-dipole interaction, and the induced-dipole–induced-
dipole, or London dispersion, interaction. In each case, the distance depen-
dence of the potential energy can be thought of as arising from the 1/r 3 de-
pendence of the field (and hence the magnitude of the induced dipole) and the
1/r 3 dependence of the potential energy of interaction of the dipoles (either
permanent or induced).
D17B.3 The monomer unit as shown is expected to be planar and rigid on account of
the benzene rings and the amide linkage between them; the π system will be
delocalised across both rings and the amide linkage, resulting in a significant
barrier to rotation. Hydrogen bonds can be formed between the NH hydrogens
on one polymer chain and the carbonyl groups of another, resulting in strong
interactions between the chains.
The flatness of the Kevlar polymeric molecule makes it possible to process the
material so that many molecules with parallel alignment form highly ordered,
untangled crystal bundles. The alignment makes possible both considerable
van der Waals attractions between adjacent molecules and for strong hydrogen
bonding between the polar amide groups on adjacent molecules. These bond-
ing forces create the high thermal stability and mechanical strength observed
in Kevlar.
D17B.5 The arrangement of charges corresponding to the various multipoles are shown
in Fig. 17A.2 on page 677. According to [17B.4–686], the potential energy of
interaction between an n-pole and an m-pole goes as 1/r n+m+1 . Two examples
of this relationship are derived in the text for the interaction between: (i) a point
charge (n = 0) and a dipole (m = 1), giving a 1/r 0+1+1 = 1/r 2 dependence; and
(ii) a dipole (n = 1) and a dipole (m = 1) giving a 1/r 1+1+1 = 1/r 3 dependence.
In both cases this form of the interaction arises by assuming that the distance
r is much greater than the separation between charges which form the dipole.
Consider the interaction between a point charge and a dipole, with the point
charge along the line of the dipole. The point charge interacts with the two
charges which form the dipole: these interactions are opposite in sign but dif-
ferent in magnitude because the distances between the point charge and the
two charges which form the dipole are not the same. As the point charge moves
further away, these interactions both decrease in magnitude. However, because
the distance between these charges is becoming less significant compared to
the distance of the point charge, the magnitude of the interaction with each
becomes more similar. As a result, the terms with opposite sign come closer
to cancelling one another out and the overall interaction therefore goes to zero
faster than does the simple interaction between point charges.
Effectively, as r increases the two charges which form the dipole merge and
begin to cancel one another. This is why the potential energy of interaction
falls off more quickly than it does for the interaction between two point charges.
For quadrupoles and high n-poles the effect is greater, leading to an even faster
fall off of the interaction, presumably because the effective cancellation of the
charges happens more quickly when there are more charges forming the n-pole.
E17B.1(a) The shape of a linear quadrupole is given in Fig. 17A.2 on page 677; an example
of such an arrangement is a CO2 molecule which has negative charges on the
oxygen atoms and a balancing positive charge on the central carbon. Two such
quadrupoles are shown in Fig. 17.7, arranged so that they are collinear.
r
A B C D E F
-Q +2Q -Q -Q +2Q -Q
l l l l
Figure 17.7
The interaction energy is derived in a similar way to that used in Section 17B.1(b)
on page 684 for dipole–dipole interactions. The total interaction energy is the
sum of nine pairwise terms, one for each combination of a point charge in one
quadrupole with a point charge in the other quadrupole. Each term has the
form Q 1 Q 2 /4πε 0 r 12 where Q 1 and Q 2 are the charges being considered in that
term and r 12 is the distance between them.

A–D A–E A–F B–D
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ · ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
1 (−Q)(−Q) (−Q)(2Q) (−Q)(−Q) (2Q)(−Q)
V =− ( + + +
4πε 0 r r+l r + 2l r−l
(2Q)(2Q) (2Q)(−Q) (−Q)(−Q) (−Q)(2Q) (−Q)(−Q)
+ + + + + )
r r+l r − 2l r−l r
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
B–E B–F C–D C–E C–F
1 Q 2Q Q 2
2Q 4Q 2Q Q 2Q Q22 2 2 2 2 2 2
= ( − + − + − + − + )
4πε 0 r r + l r + 2l r − l r r + l r − 2l r − l r
1 6Q 2 4Q 2 4Q 2 Q2 Q2
= ( − − + + )
4πε 0 r r + l r − l r + 2l r − 2l
Q2 4 4 1 1
= (6 − − + + )
4πε 0 r 1 + x 1 − x 1 + 2x 1 − 2x
where x = l/r. Assuming that r ≫ l, so that x ≪ 1, the fractions can be
expanded using the series
1 1
= 1 − x + x 2 − x 3 + x 4 − ... and = 1 + x + x 2 + x 3 + x 4 + ...
1+x 1+x
from inside the front cover. The first term that survives is the one in x 4 ; higher
order terms are much smaller and so are not included.
Q2
V= (6 − 4 (1 − x + x 2 − x 3 + x 4 − ...) − 4 (1 + x + x 2 + x 4 + ...)
4πε 0 r
+ (1 − (2x) + (2x)2 − (2x)3 + (2x)4 − ...)
+ (1 + (2x) + (2x)2 + (2x)3 + (2x)4 + ...) )
Q2
≈ (6 − 4 − 4 + 1 + 1 + 4x − 4x − 2x + 2x − 4x 2 − 4x 2 + 4x 2 + 4x 2
4πε 0 r
+ 4x 3 − 4x 3 − 8x 3 + 8x 3 − 4x 4 − 4x 4 + 16x 4 + 16x 4 )
Q2 Q2 l 4 6Q 2 l 4
= × 24x 4 = × 24 ( ) =
4πε 0 r 4πε 0 r r πε 0 r 5
E17B.2(a) The average energy of interaction between rotating polar molecules is given by
the Keesom interaction [17B.4–686].
C 2µ 12 µ 22
⟨V ⟩ = − C=
r6 3(4πε 0 )2 kT
In this case µ 1 = µ 2 , so
2µ 14
C=
3(4πε 0 )2 kT
4
2 × [(1 D) × (3.3356 × 10−30 C m)/(1 D)]
=
3 × [4π × (8.8542 × 10−12 J−1 C2 m−1 )]2 × (1.3806 × 10−23 J K−1 ) × (298 K)
= 1.62... × 10−78 J m6
C 1.62... × 10−78 J m6
Hence ⟨V ⟩ = − = − = −1.0 × 10−22 J . This energy
r 6 (0.5 × 10−9 m)6
corresponds, after multiplication by N A , to −62 J mol−1 . This is very much
smaller than the average molar kinetic energy of the molecules which, as ex-
plained in Section 2A.2(a) on page 35, is given by
3
2
RT = 3
2
× (8.3145 J K−1 mol−1 ) × (298 K) = 3.7 kJ mol−1
E17B.3(a) The dipole–induced dipole interaction between a polar molecule such as wa-
ter and a polarizable molecule such as benzene is given by [17B.6–687], V =
−µ 12 α 2′ /4πε 0 r 6 . From the data in the Resource section the dipole moment of
water is 1.85 D and the polarizability volume of benzene is 10.4 × 10−30 m3 .
µ 12 α 2′
V =−
4πε 0 r 6
2
[(1.85 D) × (3.3356 × 10−30 C m)/(1 D)] × (10.4 × 10−30 m3 )
=−
4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (1.0 × 10−9 m)6
= −3.55... × 10−24 J
This interaction energy corresponds, after multiplication by Avogadro’s con-

stant, to −2.1 J mol−1 .
E17B.4(a) The London formula for the energy of the dispersion interaction is given by
[17B.7–688]
C I1 I2
V = − 6 C = 32 α 1′ α 2′
r I1 + I2
In the case that the two interacting species are the same, with polarizability
volume α ′ and ionisation energy I, this expression becomes
I 2 3 ′2 3α ′2 I
C = − 23 α ′2 = α I hence V =−
2I 4 4r 6
Table 17A.1 on page 675 in the Resource section gives the polarizability volume
of helium as α ′ = 0.20 × 10−30 m3 and Table 8B.4 on page 305 gives the first
ionisation energy as I = 2372.3 kJ mol−1 , therefore
3α ′2 I 3×(0.20 × 10−30 m3 )2 ×(2372.3 × 103 J mol−1 )

V =− = − = 0.071 J mol−1
4r 6 4 × (1.0 × 10−9 m)6
E17B.5(a) The interaction between a point charge and a point dipole orientated directly
away the charge is given by [17B.1–683], V = −µ 1 Q 2 /4πε 0 r 2 . In this case the
lithium ion has a charge of +e = +1.6022 × 10−19 C.
µ1 Q2 (1.85 D)×[(3.3356 × 10−30 C m)/(1 D)]×(1.6022 × 10−19 C)

V =− = −
4πε 0 r 2 4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (100 × 10−12 m)2
= −8.88... × 10−19 J
The potential energy if the dipole has the reverse orientation is +8.88...×10−19 J
so the energy required to reverse the direction is
∆V = (8.88... × 10−19 J) × 2 = 1.77... × 10−19 J = 1.77 × 10−18 J
The energy per mole is found by multiplying by N A
∆Vm = (1.77... × 10−21 kJ) × (6.0221 × 1023 mol−1 ) = 1.07 × 103 kJ mol−1
E17B.6(a) The potential energy interaction between two parallel point dipoles separated
by distance r at angle Θ is given by [17B.3b–685], V = µ 1 µ 2 (1−3 cos2 Θ)/4πε 0 r 3 .
µ 1 µ 2 (1 − 3 cos2 Θ)
V=
4πε 0 r 3
2
(2.7 D) × (2.7 D) × [(3.3356 × 10−30 C m)/(1 D)] × (1 − 3 cos2 45○ )
=
4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (3.0 × 10−9 m)3
= −1.3 × 10−23 J which corresponds to −8.1 J mol−1
P17B.1 (a) In a vacuum the interaction energy is
Q 1 Q 2 (−0.36 × 1.6022 × 10−19 C) × (+0.45 × 1.6022 × 10−19 C)

V= =
4πε 0 r 4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (3.0 × 10−9 m)
= −1.2 × 10−20 J
This interaction energy corresponds, after multiplication by Avogadro’s

constant, to −7.5 kJ mol−1 .
(b) If instead the medium is bulk water, ε 0 should be replaced by ε 0 ε r where
ε r is the relative permittivity of water.
Q1 Q2 (−0.36 × 1.6022 × 10−19 C) × (+0.45 × 1.6022 × 10−19 C)

V= =
4πε 0 ε r r 4π × (8.8542 × 10−12 J−1 C2 m−1 ) × 80 × (3.0 × 10−9 m)
= −1.6 × 10−22 J
This interaction energy corresponds, after multiplication by Avogadro’s

constant, to −94 J mol−1 . Note that the presence of an intervening medium
significantly reduces the energy of the interaction in this case.
P17B.3 The energy of a dipole µ in an electric field E is given in How is that done? 17A.2
on page 679 by E(θ) = −µE cos θ, where θ is the angle between the direction
of the dipole and the field. If the dipole is parallel to the field, θ = 0, this energy
is simply −µE, while if it is at θ = 90○ the energy is zero. The dipole–induced
dipole interaction between a polar molecule such as water with dipole µ and a
polarizable species such as an argon atom with polarizability volume α ′ is given
by [17B.6–687], V = −µ 2 α ′ /4πε 0 r 6 . The task is to find the distance r for which
this energy is equal to that of the dipole aligned parallel to the field, that is, for
which
µ2 α′
−µE = −
4πε 0 r 6
Rearranging, and noting that 1 V = 1 J C−1 , gives
1/6
µα ′
r=( )
4πε 0 E
1/6
(1.85 D) × (3.3356 × 10−30 C m)/(1 D) × (1.66 × 10−30 m3 )
=( )
4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (1.0 × 103 V m−1 )
= 2.12... × 10−9 m = 2.1 nm
For distances closer than this the arrangement with the dipole directed towards
the argon atom is lower in energy than that with the dipole parallel to the field.
P17B.5 The dipole–induced dipole interaction between a dipole and a polarizable molecule
is given by [17B.6–687], V = −µ 12 α 2′ /4πε 0 r 6 .
µ 12 α 2′
V =−
4πε 0 r 6
2
[(2.7 D) × (3.3356 × 10−30 C m)/(1 D)] × (1.04 × 10−29 m3 )
=−
4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (0.4 × 10−9 m)6
= −1.85... × 10−21 J
This interaction energy corresponds, after multiplication by Avogadro’s con-

stant, to −1.1 kJ mol−1 .
P17B.7 The London formula for the energy of the dispersion interaction is given by
[17B.7–688]
C I1 I2
V = − 6 C = 32 α 1′ α 2′
r I1 + I2
Differentiation with respect to r gives
dV d C 6C I1 I2 1
F=− = − (− 6 ) = − 7 = −9α 1 α 2
dr dr r r I1 + I2 r 7
The negative sign indicates that the force is attractive.
P17B.9 One way of rewriting the Lennard-Jones potential with an exponential repul-
sive term (the exp-6 potential) is
r 0 12 r0 6 r0 6
4ε {( ) −( ) } → Aε {Be−r/r 0 − ( ) }
r r r
where the constants A and B are introduced to scale the overall potential and to
scale the repulsive term relative to the attractive term; their values can be found
by introducing constraints to the potential. This modified potential immedi-

ately presents a problem. In the limit that r → 0 the repulsive term, Be−r/r 0 ,
tends to B, whereas the attractive term becomes more and more negative, with-
out limit. This is not the required behaviour: a realistic potential is expected to
be repulsive for small r.
The Lennard-Jones potential crosses zero at r = r 0 . If this condition is applied
to the modified potential the value of B is found by solving
Aε{Be−1 − 1} = 0
which gives B = e and hence the potential becomes
r0 6
V (r) = Aε {e1−r/r 0 − ( ) }
r
The function in the curly braces is positive for r > r 0 and negative for r <
r 0 . Therefore, for it to represent a realistic potential with the attractive part
dominating at large distances, the constant A must be negative. Assuming this
to be so, the potential shows a minimum at a distance somewhat beyond r =
r 0 ; the minimum is located by solving dV (r)/dr = 0. There is no analytical
solution to the resulting equation, but mathematical software is able to locate
the minimum numerically at r = 1.360 r 0 .
If, like the Lennard-Jones potential, the depth of the well is to be ε at this point,
then A = −1.853. A plot of the potential with this value of the constant is shown
in Fig. 17.8. For comparison, the Lennard-Jones potential is also plotted.
3
exp-6
Lennard-Jones
2
V (r)/ε
−1
1.0 1.5 2.0 2.5 3.0
r/r 0
Figure 17.8
17C Liquids
D17C.1 This is discussed in Section 17C.4 on page 701.
E17C.1(a) The height climbed by a liquid in a capillary tube of radius r is given by [17C.8–
697], h = 2γ/ρg acc r, assuming that the contact angle is zero. Rearranging for
γ, taking ρ = 0.9982 g cm−3 = 998.2 kg m−3 , and noting that 1 N = 1 kg m s−2
gives
γ = 12 ρg acc rh
= 12 ×(998.2 kg m−3 ) × (9.807 m s−2 ) × (0.300 × 10−3 m) × (4.96 × 10−2 m)
= 0.0728... kg s−2 = 72.8 mN m−1
E17C.2(a) The pressure difference between the inside and outside of a spherical droplet
is given by the Laplace equation [17C.7–696], p in = p out + 2γ/r. Hence, taking
the surface tension of water as 72.75 mN m−1 from Table 17C.1 on page 695, and
noting that 1 Pa = 1 N m−2 ,
2γ 2 × (72.75 × 10−3 N m−1 )
∆p = p in − p out = = = 728 kPa
r (200 × 10−9 m)
E17C.3(a) The height climbed by a liquid in a capillary tube of radius r is given by [17C.8–
697], h = 2γ/ρg acc r, assuming that the contact angle is zero. Rearranging for
γ, replacing r by 12 d where d is the diameter of the tube, and noting that 1 N =
1 kg m s−2 gives
γ = 21 ρg acc rh = 14 ρg acc dh
= 41 ×(997.0 kg m−3 )×(9.807 m s−2 )×(0.500 × 10−3 m)×(5.89 × 10−2 m)
= 0.0719... kg s−2 = 72.0 mN m−1
E17C.4(a) The vapour pressure of a liquid when it is dispersed as spherical droplets of

radius r is given by the Kelvin equation [17C.15–701], p = p∗ e2γVm (l)/rRT , where
p∗ is the vapour pressure of bulk liquid to which no additional pressure has
been applied. Because the mass density of a substance with molar volume
Vm and molar mass M is given by ρ = M/Vm , it follows that Vm = M/ρ.
Substituting this into the Kelvin equation gives p = p∗ e2γ(M/ρ)/rRT .
The surface tension γ of water at 20 ○ C is taken as 72.75 mN m−1 , which is equal
to 72.75 × 10−3 J m−2 (Table 17C.1 on page 695), and the molar mass of water is
M = 18.0158 g mol−1 . Hence, taking p∗ as 2.3 kPa,
2γM/ρ
p = p∗ exp ( ) = (2.3 kPa)
rRT
2×(72.75 × 10−3 J m−2 )×(18.0158 g mol−1 )/(0.9982 × 106 g m−3 )
×exp ( )
(10 × 10−9 m)×(8.3145 J K−1 mol−1 )×([20 + 273.15] K)
= 2.6 kPa
P17C.1 (a) The function is plotted in Fig. 17.9
g(r)
1
0
0 2 4 6 8
r/r 0
Figure 17.9
The plot resembles Fig. 17C.1 on page 693 in that the function oscillates
for short values of r, corresponding to short-range order, but approaches
1 for large separations.
(b) The virial v 2 (r) = r(dV /dr) is obtained by differentiating the Lennard-
Jones expression for V
d r 0 12 r0 6 d
v 2 (r) = r 4ε [( ) − ( ) ] = r 4ε (r 012 r −12 − r 06 r −6 )
dr r r dr
= r × 4ε (−12r 012 r −13 + 6r 06 r −7 ) = 4ε (−12r 012 r −12 + 6r 0 r −6 )
r 0 12 r0 6
= −24ε [2 ( ) −( ) ]
r r
The quantity v 2 (r)/ε is plotted in Fig. 17.10; it falls away steeply to the left
of the maximum.
2
v 2 (r)/ε
−2
0 1 2 3 4
r/r 0
Figure 17.10
17D Macromolecules

D17D.1 The freely jointed random coil model of a polymer chain of ‘units’ or ‘residues’
gives the simplest possibility for the conformation of the polymer that is not
capable of forming hydrogen bonds or any other type of non-linkage bond. In
this model, a bond that links adjacent units in the chain is free to make any
angle with respect to the preceding one. The residues are assumed to occupy
zero volume, so different parts of the chain can occupy the same region of space.
It is also assumed in the derivation of the expression for the probability of the
ends of the chain being a distance nl apart, that the chain is compact in the
sense that n ≪ N. This model is obviously an oversimplification because a
bond is actually constrained to a cone of angles around a direction defined by
its neighbour and it is impossible for one section of a chain to overlap with
another. Constrained angles and self-avoidance tend to swell the coil, so it is
better to regard the R rms and R g values of a random coil as lower bounds to the
actual values.
The freely jointed chain is improved by constraining each successive individual

bond to a single cone of angle θ relative to its neighbour. This constrained
chain reduces R rms and R g values of a freely jointed random coil by a factor of
F = [(1 − cos θ)/(1 + cos θ)]1/2 .
D17D.3 Polymer melting occurs at a specific melting temperature, Tm , above which the
crystallinity of polymers can be destroyed by thermal motion. Higher melting
temperatures correspond to increased strength and number of intermolecular
interactions in the material. Polymers undergo a transition from a state of
high to low chain mobility at the glass transition temperature, Tg . There is
sufficient energy available to an elastomer at normal temperatures for limited
bond rotation to occur and the flexible chains writhe. At lower temperatures,
the amplitudes of the writhing motion decrease until a specific temperature, Tg ,
is reached at which motion is frozen completely and the sample forms a glass.
D17D.5 The primary structure of a macromolecule is the sequence of small molecular

residues making up the polymer, whereas the secondary structure is the (of-
ten local) spatial arrangement of a chain. The tertiary structure is the overall
three-dimensional structure of a single macromolecule. Two or more macro-
molecules may form an aggregation that yields an overall quaternary structure.
E17D.1(a) As explained in Section 17D.3(c) on page 708, the root mean square separation
of the ends of a partially rigid chain with persistence length l p is given by the
value for a free chain, N 1/2 l, multiplied by a factor F, where F is given by
[17D.10–709], F = (2l p /l − 1)1/2 . The contour length is given by [17D.5–707],
R c = N l, so a persistence length of 5.0% of the contour length corresponds to
l p = 0.050R c = 0.050N l. Hence, for N = 1000,

1/2 1/2
2l p 2 × (0.050N l) 1/2
F=( − 1) =( − 1) = (0.100N − 1)
l l
1/2
= (0.100 × 1000 − 1) = 9.94 . . .
This corresponds to a percentage increase of [(9.94...) − 1] × 100% = +895% .

The volume is proportional to the cube of the radius, so the volume of the par-
tially rigid chain is related to that of a free chain by a factor of F 3 = (9.94...)3 =
9.85... × 102 . This corresponds to a percentage increase of [(9.85... × 102 ) − 1] ×
100% = +(9.84 × 104 )% .
E17D.2(a) By analogy with [17D.10–709], the radius of gyration R g of a partially rigid coil
is related to that of a freely jointed chain according to R g = F × R g,free where
F = (2l p /l −1)1/2 . The radius of gyration for a three-dimensional freely jointed
chain is given by [17D.7b–708], R g,free = (N/6)1/2 l, so
1/2
2l p N 1/2
R g = F × R g,free = ( − 1) ×( ) l
l 6
Rearranging gives
2l p N 6R g2 2l p 2
l 6R g
R g2 = ( − 1) ( ) l 2 hence = −1 hence lp = ( 2 + 1)
l 6 Nl2 l 2 Nl
Therefore for the polymer in question, taking l = 0.150 nm,

2
l 6R g 0.150 nm 6 × (2.1 nm)2
lp = ( 2 + 1) = ( + 1) = 0.16 nm
2 Nl 2 1000 × (0.150 nm)2
E17D.3(a) Modelling the polyethene as a 1D random coil perfect elastomer, the restoring
force is given by [17D.12a–710], F = (kT/2l) ln[(1 + λ)/(1 − λ)] where λ =
x/N l.
The monomer of polyethene –[CH2 CH2 ]n – is taken to be CH2 CH2 , so the
number of monomers in the chain is given by
M polymer 65 × 103 g mol−1

N= = = 2.31... × 103
M CH2 CH2 28.0516 g mol−1
The length of each CH2 CH2 unit is estimated as the length of a two C–C bonds:
one C–C bond in the centre and half a bond length either side where the unit
connects to carbons in adjacent units. From Table 9C.2 on page 342 a C–C
bond length is approximately 154 pm, so the monomer length l is taken as 2 ×
(154 pm) = 308 pm = 0.308 nm. The value of λ corresponding to a 1.0 nm
extension is therefore
x 1.0 nm
λ= = = 1.40... × 10−3
N l (2.31... × 103 ) × (0.308 nm)
Because λ ≪ 1 the simplified equation for the restoring force, [17D.12b–711],

F = (kT/N l 2 )x is used. Noting that 1 J m−1 = 1 N gives
kT (1.3806 × 10−23 J K−1 ) × ([20 + 273.15] K)
F= x = × (1.0 × 10−9 nm)
Nl2 (2.31... × 103 ) × (0.308 × 10−9 m)2
= 1.8 × 10−14 N
E17D.4(a) The entropy change when a 1D random coil is stretched or compressed by a

distance x is given by [17D.11–710], ∆S = − 12 kN ln[(1 + λ)(1+λ) (1 − λ)(1−λ) ]
where λ = x/R c . The contour length R c is given by [17D.5–707], R c = N l, so it
follows that λ = x/N l.
N= = = 2.31... × 103
M CH2 CH2 28.0516 g mol−1
(154 pm) = 308 pm = 0.308 nm. The value of λ corresponding to a 1.0 nm
extension is therefore
x 1.0 nm
λ= = = 0.00140...
N l (2.31... × 103 ) × (0.308 nm)
The entropy change is therefore
∆S = 21 kN ln [(1 + λ)(1+λ) (1 − λ)(1−λ) ]

= 1
2
× (1.3806 × 10−23 J K−1 ) × (2.31... × 103 )
× ln [(1 + 0.00140...)(1+0.00140 ...) × (1 − 0.00140...)(1−0.00140 ...) ]
= −3.14... × 10−26 J K−1
The molar entropy change is obtained by multiplying by Avogadro’s constant
∆S m = (−3.14... × 10−26 J K−1 ) × (6.0221 × 1023 mol−1 ) = −0.019 J K−1 mol−1
E17D.5(a) The number-average molar mass is given by [17D.1a–703], M n = (1/N total )

∑ i N i M i . Denoting the polymers as 1 and 2 gives
1 1
Mn = ∑ Ni Mi = (N 1 M 1 + N 2 M 2 )
N total i N1 + N2
Because the two polymers are present in equal amounts (equal amounts in
moles), N 1 = N 2 = N and hence
1
Mn = (N M 1 + N M 2 ) = 12 (M 1 + M 2 ) = 21 [(62 kg mol−1 ) + (78 kg mol−1 )]
2N
= 70 kg mol−1
The weight-average molar mass is given by [17D.1b–703], M W = (1/m total )

∑ i m i M i , where m i is the mass of polymer i present. Using m i = n i M i and
m total = ∑ i m i gives
∑ i m i M i ∑ i (n i M i )M i ∑ i n i M i2
MW = = =
∑i m i ∑i n i M i ∑i n i M i
The two polymers are present in equal amounts n 1 = n 2 = n, therefore
nM 12 + nM 22 M 12 + M 22 (62 kg mol−1 )2 + (78 kg mol−1 )2

MW = = =
nM 1 + nM 2 M 1 + M 2 (62 kg mol−1 ) + (78 kg mol−1 )
= 71 kg mol−1
E17D.6(a) The root mean square separation of the ends of a freely jointed one-dimensional
chain is given by [17D.6–707], R rms = N 1/2 l, where N is the number of monomer
units and l is the length of each unit. In this case
R rms = N 1/2 l = 7001/2 × (0.9 nm) = 24 nm
E17D.7(a) The contour length R c of a polymer is given by [17D.5–707], R c = N l, and the

root mean square separation of the ends of a freely jointed one-dimensional
chain is given by [17D.6–707], R rms = N 1/2 l. In both cases N is the number of
monomer units and l is the length of each unit.
The monomer of polyethene –(CH2 CH2 )n – is taken to be CH2 CH2 . The num-
ber of monomers in the chain is given by

N= = = 9.98... × 103
M CH2 CH2 28.0516 g mol−1
connects to carbons in adjacent units. From Table 9C.2 on page 342 in the
Resource section a C–C bond length is approximately 154 pm, so the monomer
length l is taken as 2 × (154 pm) = 308 pm. The contour length and root mean
square separation are then given by
R c = N l = (9.98... × 103 ) × (308 pm) = 3.07... × 106 pm = 3.07 µm
R rms = N 1/2 l = (9.98... × 103 )1/2 × (308 pm) = 3.07... × 104 pm = 30.8 nm
E17D.8(a) The radius of gyration R g of a one-dimensional random coil is given by [17D.7a–

708], R g = N 1/2 l. Rearranging, and taking the length of each C–C link as
154 pm (Table 9C.2 on page 342) gives
2
Rg 2 7.3 × 10−9 m
N =( ) =( ) = 2.2 × 103
l 154 × 10−12 m
E17D.9(a) The probability that the ends of a one-dimensional random coil are a distance
nl apart is given by [17D.3–706], P = (2/πN)1/2 e−n /2N where N is the total
2
number of monomers in the chain and l is the length of each monomer unit.
N= = = 2.31... × 103
M CH2 CH2 28.0516 g mol−1
(154 pm) = 308 pm.
If the end-to-end distance is d, then d = nl and hence n = d/l. In this case
d = 10 nm and l = 308 pm hence
d 10 × 10−9 m
n= = = 32.4...
l 308 × 10−12 m
The probability of the ends being this distance apart is therefore
2 1/2 −n 2 /2N
P=( ) e
πN
1/2
2
× e−(32.4 ...) /2×(2.31 ...×10 3 )
2
=( ) = 0.013
π × (2.31... × 103 )
E17D.10(a) The probability distribution function for a three-dimensional freely jointed chain
is given by [17D.4–706]
a 3
3 1/2
) r 2 e−a r
2 2
f (r) = 4π ( a=( )
π 1/2 2N l 2
where N is the number of monomers in the chain, l is the length of each monomer,
and f (r) dr is the probability that the ends of the chain are a distance between
r and r + dr apart.
N= = = 2.31... × 103
M CH2 CH2 28.0516 g mol−1
(154 pm) = 308 pm = 0.308 nm. Therefore
1/2
3 1/2 3
a=( ) =( ) = 0.0826... nm−1
2N l 2 2 × (2.31... × 10 ) × (0.308 nm)2
3
a 3
) r 2 e−a r
2 2
f (10.0 nm) = 4π (
π 1/2
3
0.0826... nm−1 −1 2
) × (10.0 nm)2 × e−(0.0826 nm ) ×(10.0 nm)
2
= 4π (
π 1/2
= 0.0642... nm−1
The probability that the ends will be found in a narrow range of width δr =
0.1 nm at 10.0 nm is therefore
f (10.0 nm)δr = (0.0642... nm−1 ) × (0.1 nm) = 6.4 × 10−3
E17D.11(a) As explained in Section 17D.3(b) on page 708, the radius of gyration of a con-
strained chain is given by the value for a free chain multiplied by a factor F,
1/2
where F is given by [17D.8–708], F = [(1 − cos θ)/(1 + cos θ)] . For θ =
○
109 ,
1 − cos θ 1/2 1 − cos 109○ 1/2
F=( ) =( ) = 1.40...
1 + cos θ 1 + cos 109○
This corresponds to a percentage increase of [(1.40...) − 1] × 100% = +40.1% .
The volume is proportional to the cube of the radius, so the volume of the
constrained chain is related to that of a free chain by a factor of F 3 = (1.40...)3 =
2.75.... This corresponds to a percentage increase of [(2.75...) − 1] × 100% =
+176% .
P17D.1 There is some lack of clarity in the text over the definition of the radius of
gyration, R g . For a polymer consisting of N identical monomer units, R g is
defined as
N
R g2 = (1/N) ∑ r 2i (17.1)
i=1
where r i is the distance of monomer unit i from the centre of mass. In other
words, the radius of gyration is the root-mean-square of the distance of the
monomer units from the centre of mass.
A related quantity is the moment of inertia I about an axis, which is defined in
the following way
N
I = ∑ md i2 (17.2)
i=1
where m is the mass of the monomer unit and d i is the perpendicular distance
from the monomer to the axis. In general, the distance d i is not the same as r i :
the first is the perpendicular distance to the axis, the second is the distance to
the centre of mass.
A radius of gyration can be related to a moment of inertia by imagining a rigid
rotor consisting of a mass m tot equal to the total mass of the polymer held at a
distance R g from the origin; the moment of inertia of this rotor is I = m tot R g2 ,
and hence R g2 = I/m tot . However, note that this radius of gyration is associated
by the rotation about a particular axis.
(a) For a solid sphere the mass is distributed continuously rather than at
discrete points as for the simple polymer. The definition in eqn 17.1 is
adapted by imagining that the monomer unit at r i is a volume element dV
located at distance r from the centre of mass (the origin). The sum over r 2i
becomes the integral of r 2 dV over the sphere, and division by N becomes
division by the volume of the sphere Vs . Hence R g2 = (1/Vs ) ∫sphere r 2 dV .
It is convenient to complete the calculation using spherical polar coordi-
nates
1 r=a θ=π ϕ=2π
R g2 = ∫ ∫ ∫ r 2 × r 2 sin θ dr dθ dϕ
Vs r=0 θ=0 ϕ=0
3 r=a θ=π ϕ=2π
= ∫ r 4
dr ∫ sin θ dθ ∫ dϕ
4πa 3 r=0 θ=0 ϕ=0
3 a5
= × × 2 × 2π
4πa 3 5
= 53 a 2
Hence the radius of gyration is (3/5)1/2 a .

(b) For a solid rod it is more convenient to use eqn 17.2 for rotation about (i)
the long axis of the rod, and (ii) an axis perpendicular to this and which
passes through the centre of mass. The long axis of the rod defines the
z-axis and the centre of mass is at z = 0; the rod therefore extends from
−l/2 to +l/2 along z. It is convenient to use cylindrical polar coordinates
described in The chemist’s toolkit 19 in Topic 7F on page 267. In such a
coordinate system the volume element is r dr dϕ dz, and ϕ ranges from 0
to 2π.
Equation 17.2 is adapted for a solid object by replacing the mass by a
volume element dV which has mass ρ dV , where ρ is the mass density; the
summation becomes an integration over the relevant coordinates which
describe the rod: z = −l/2 to +l/2, ϕ = 0 to 2π, and r = 0 to a. To compute
the moment of inertia about the long axis note that the perpendicular
distance to the axis is r so the integral is
z=+l /2 r=a ϕ=2π
I∣∣ = ∫ ∫ ∫ r 2 × ρ dV
z=−l /2 r=0 ϕ=0
z=+l /2 r=a ϕ=2π
=ρ ∫ ∫ ∫ r 2 × r dz dr dϕ
z=−l /2 r=0 ϕ=0
z=+l /2 r=a ϕ=2π
=ρ ∫ dz ∫ r 3 dr ∫ dϕ
z=−l /2 r=0 ϕ=0
a4
=ρ × l × × 2π = ρl a 4 π/2 = m tot a 2 /2
4
where on the last line the total mass is given by m tot = ρ × πa 2 × l. A rigid
rotor with the same total mass has moment of inertia I = m tot R g,∣∣
2
, hence
R g,∣∣ = (2)−1/2 a .
To compute the moment of inertia perpendicular to the long axis, say
about the x-axis, it is necessary to know the perpendicular distance d
between an arbitrary point (x, y, z) and that axis. This distance is that
between the points (x, y, z) and (x, 0, 0); by Pythagoras’ theorem d 2 =
y 2 + z 2 = r 2 sin2 ϕ + z 2 . The moment is inertia is therefore found from the
integral
I=∫ (r 2 sin2 ϕ + z 2 ) × ρ dV = ρ ∫ r 2 sin2 ϕ dV + ρ ∫ z 2 dV

cyl. cyl. cyl.
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶ ´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
A B
where the integration is over the complete cylinder. The integrals A and
B are conveniently evaluated separately.
z=+l /2 r=a ϕ=2π

A =ρ ∫ ∫ ∫ r 2 sin2 ϕ × r dz dr dϕ
z=−l /2 r=0 ϕ=0
z=+l /2 r=a ϕ=2π
=ρ ∫ dz ∫ r 3 dr ∫ sin2 ϕ dϕ
z=−l /2 r=0 ϕ=0
a4
=ρ × l × × π = ρl a 4 π/4 = m tot a 2 /4
4
where the integral over ϕ is found using Integral T.2 with k = 1 and a =
2π.
z=+l /2 r=a ϕ=2π

B =ρ ∫ ∫ ∫ z 2 × r dz dr dϕ
z=−l /2 r=0 ϕ=0
z=+l /2 r=a ϕ=2π
=ρ ∫ z 2 dz ∫ r dr ∫ dϕ
z=−l /2 r=0 ϕ=0
l3
a2
=ρ × × × 2π = ρl 3 a 2 π/12 = m tot l 2 /12
12 2
The moment of inertia about the perpendicular axis is therefore
I = A + B = m tot a 2 /4 + m tot l 2 /12 = m tot (a 2 /4 + l 2 /12)
A rigid rotor with the same total mass has moment of inertia I = m tot R g,
2
,
hence R g, = (a 2 /4 + l 2 /12)1/2 .
(c) The specific volume is the volume divided by the mass, υ s = V /m, and
the mass of an individual macromolecule is given by M/N A , where M is
the molar mass.
V (4/3)πa 3 3υ s M
υs = = hence a3 =
m M/N A 4πN A
The radius of gyration is therefore
3 1/2 3υ s M 1/3
Rg = ( ) ( )
5 4πN A
3 1/2 3 1/3
1/3
=( ) ( ) (υ s M)
5 4πN A
1/3
3 1/2 3
R g /m = ( ) ( )
5 4π × (6.0221 × 1023 mol−1 )
1/3
× [(υ s /cm3 g−1 )(M/g mol−1 ) × 10−6 ]
1/3
= 5.6902 × 10−11 × [(υ s /cm3 g−1 )(M/g mol−1 )]
1/3
R g /nm = 0.056902 × [(υ s /cm3 g−1 )(M/g mol−1 )]
The factor of 10−6 on the fourth line is there to convert cm3 to m3 . For
the given data
1/3
R g /nm = 0.056902 × [(0.750) × (100 × 103 )] = 2.40
The radius of gyration is therefore 2.40 nm .
For a rod R g,∣∣ = (1/2)1/2 a = (1/2)1/2 × (0.50 nm) = 0.35 nm . To find the
radius of gyration about the perpendicular axis requires a knowledge of l and
this is found from the specific volume in a similar way to the method used for
a sphere
V πa 2 l υs M
υs = = hence l =
m M/N A πa 2 N A
υs M
l=
πa 2 N A
(0.750 × 10−6 m3 g−1 ) × (100 × 103 g mol−1 )
=
π × (0.50 × 10−9 m)2 × (6.0221 × 1023 mol−1 )
= 1.585... × 10−7 m = 158.5... nm
Hence
R g, = (a 2 /4 + l 2 /12)1/2 = [(0.50 nm)2 /4 + (158.5... nm)2 /12]1/2 = 46 nm
Because l ≫ a the radius of gyration is dominated by the term in l and R g, ≈

(12)−1/2 l.
P17D.3 The problem should be stated as ⟨r 2 ⟩ = N l 2 . The probability distribution for

the separation of the ends in a 3D random coil is given by [17D.4–706]
a 3
3 1/2
) r 2 e−a r
2 2
f (r) = 4π ( where a=( )
π 1/2 2N l 2
The mean-square separation is calculated as

∞ a 3
) r 2 e−a r
2 2
⟨r 2 ⟩ = ∫ r 2 f (r) dr = ∫ r 2 × 4π ( dr
0 π 1/2
a 3 ∞ a 3 3 π 1/2
r 4 e−a r dr = 4π ( 1/2 ) ×
2 2
= 4π ( ) ∫ ( 2)
π 1/2 0 π 8(a ) a
2 2
´¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¸ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¶
Integral G.5 with k = a 2
3 3 2N l 2
= = ( ) = Nl2
2a 2 2 3
P17D.5 The walk is constructed by starting at the origin and taking steps of unit length
in a direction specified by a randomly generated angle θ between 0 and 360○ .
Each step then involves incrementing the x coordinate by cos θ and the y co-
ordinate by sin√θ. The final distance r reached at the end of the walk is found
calculating r = x 2 + y 2 where x and y are the final x and y coordinates.
10 10
5 5
0 0
−5 −5
−10 −10
−4 −2 0 2 4 −4 −2 0 2 4
Figure 17.11
Two such walks, of 50 and 100 steps, are shown in Fig. 17.11. The final values of
r in these cases are 6.24 and 7.13.
To investigate whether the mean and most probable values of r vary as N 1/2 ,
where N is the number of steps, a large number of random walks with vary-
ing numbers of steps are generated and the value of r found for each. The
table shows the mean and most probable values of r estimated from samples
of 100 random walks carried out with each of 10, 20, 30, 40, 50, 60, 70, 80, 90
and 100 steps; the most probable values have been estimated by constructing a
histogram of the values of r, fitting a curve to the histogram, and finding the
maximum of the curve.
If r mean and rmost probable vary as N 1/2 then plots of these values against N 1/2
should give a straight line passing through the origin. The data are plotted in
Fig. 17.12. In both cases the the data fall on a reasonable straight line that almost
passes through the origin, thus indicating that rmean and rmost probable do indeed
vary as N 1/2 .
N N 1/2 r mean rmost probable

10 3.16 2.49 3.44
20 4.47 3.75 4.81
30 5.48 4.54 7.03
40 6.32 6.32 8.94
50 7.07 6.30 8.81
60 7.75 7.05 8.77
70 8.37 7.10 8.94
80 8.94 7.83 10.99
90 9.49 8.40 11.80
100 10.00 8.70 12.94
10 15
8
rmost probable
10
6
rmean
4
5
2
0 0
0 2 4 6 8 10 0 2 4 6 8 10
N 1/2 N 1/2
Figure 17.12
P17D.7 The formula given in the text for the radius of gyration of a sphere of radius R is
R g = (3/5)1/2 R. If the molecules given are globular, that is, roughly spherical,
their specific volume υ s should be given by υ s = V /m = (4/3)πR 3 /m where R
is the radius of the sphere and m is the mass of one molecule. Replacing m by
M/N A and rearranging gives
4
πR 3 3υ s M 1/3 3 1/2 3υ s M 1/3
υs = 3
hence R=( ) hence Rg = ( ) ( )
M/N A 4πN A 5 4πN A
In the last step R g = (3/5)1/2 R is used. Using this expression, the value of R g
expected for each of the molecules if they are spherical is calculated from the
υ s and M data, and compared to the experimental value of R g . If the values are
similar then there is evidence that the molecules are globular.
For serum albumin
1/3
3 1/2 3υ s M 1/3 3 1/2 3×(0.752 cm3 g−1 )×(66 × 103 g mol−1 )
Rg = ( ) ( ) =( ) ( )
5 4πN A 5 4π × (N A )
= 2.09... × 10−7 cm = 2.09 nm
For bushy stunt virus

1/3
3 1/2 3υ s M 1/3 3 1/2 3×(0.741 cm3 g−1 )×(10.6 × 106 g mol−1 )
Rg = ( ) ( ) =( ) ( )
5 4πN A 5 4π×(N A )
= 1.13... × 10−6 cm = 11.3 nm
For DNA
1/3
3 1/2 3υ s M 1/3 3 1/2 3×(0.556 cm3 g−1 )×4 × 106 g mol−1 )
Rg = ( ) ( ) =( ) ( )
5 4πN A 5 4π × (N A )
= 7.42... × 10−7 cm = 7.43 nm
For serum albumin and bushy stunt virus the experimental radii of gyration
(2.98 nm and 12.0 nm) are similar to the values that would be expected if these
macromolecules were spherical, thus suggesting that they are globular. In the
case of DNA the experimental radius of gyration (117.0 nm) is much greater
than the value expected if it were spherical, suggesting that DNA is not globular
and therefore more rod-like.
P17D.9 As explained in the How is that done? 17D.4 on page 710 the restoring force
for an extended elastomer is given by F = −T(∂S/∂x)T . This restoring force
is equal to the tension t required to keep the sample at a particular length,
hence t = −T(∂S/∂x)T . The restoring force therefore depends on (∂S/∂x).
Extension of a polymer reduces the disorder and hence entropy of the chains,
so there is a tendency to revert to the more disordered non-extended state.
17E Self-assembly
D17E.1 The formation of micelles is favoured by the interaction between hydrocarbon
tails and is opposed by charge repulsion of the polar groups which are placed
close together at the micelle surface. As the salt concentration is increased, the
repulsion of head groups is reduced because their charges are partly shielded
by the ions of the salt. This favours micelle formation causing the micelles
to be formed at a lower concentration and hence reducing the critical micelle
concentration.
D17E.3 Lipids with unsaturated chains ‘freeze’ at lower temperatures, so membranes

can remain fluid at lower ambient temperatures.
D17E.5 This is discussed in Section 17E.2(a) on page 716.
E17E.1(a) The isoelectric point of a protein is the pH at which the protein has no net
charge and therefore is unaffected by an electric field. This is the pH at which
the velocity is zero; solving for this gives
0 = 0.50 − 0.10(pH) − (3.0 × 10−3 )(pH)2 + (5.0 × 10−4 )(pH)3
This equation is solved numerically to yield the solutions pH = −13.8, pH =

14.9, and pH = 4.9. The −13.8 and 14.9 solutions are rejected as they are
outside the pH range 3.0–7.0 over which the expression is valid. Therefore the
isoelectric point is 4.9 .
P17E.1 (a) The data show that π increases by 0.5 for every additional CH2 group.
The R group in question, (CH2 )6 CH3 , has two more CH2 groups than
(CH2 )4 CH3 which has π = 2.5, so the predicted value for (CH2 )6 CH3 is
2.5 + 2 × 0.5 = 3.5 .
(b) The data are plotted in Fig. 17.13. The points fall on a reasonable straight
line, the equation of which is
log K I = −1.49π − 1.95
The slope and intercept of the line are therefore −1.49 and −1.95 .
−1.5
−2.0
π
−2.5
−3.0
−3.5
−0.2 0.0 0.2 0.4 0.6 0.8 1.0
log K I
Figure 17.13
(c) The definition of the hydrophobicity constant π is given by [17E.5–717],

π = log [s(RX)/s(HX)]. This definition implies that π = log 1 = 0 for the
case R = H. It follows that for this case
log K I = −1.95 hence K I = 0.011
P17E.3 The equilibrium constant for the formation of micelles containing N monomers,
M N , is given by [17E.6b–717],
[M N ]
K=
([M]total − N[M N ]) N
where the factors of 1/c −○ are omitted for clarity. For the case that K = 1 and
N = 2 this equation becomes
[M2 ] 2
1= hence ([M]total − 2[M2 ]) = M2
([M]total − 2[M2 ])2
hence 4[M2 ]2 − (1 + 4[M]total ) [M2 ] + [M]2tot = 0

√
(1 + 4[M]total ) ± (1 + 4[M]total )2 − 4 × 4 × [M]2total
hence [M2 ] =
8
√
1 + 4[M]total − 1 + 8[M]total
hence [M2 ] =
8
The positive square root is rejected because this predicts [M2 ] ≠ 0 when [M]total =
0.

I17.1 (a) The charges on the atoms as calculated using a 6-31G* basis are shown in
the table below, and the diagram shows the numbering of the atoms and
the charges on the heteroatoms.
–0.61
O5 H 1, 2, 3
H +0.75 6 –0.29
C –0.79 C
H 4 H
C8 N
–0.58 7 H
9, 10, 11
H H
4
atom type number q/e x/Å y/Å z/Å

C 4 0.746 0.469 −0.168 −0.583
O 5 −0.611 1.228 −0.554 −1.430
C 6 −0.288 0.674 −2.216 0.767
N 7 −0.790 0.143 −0.896 0.509
C 8 −0.577 −0.186 1.194 −0.685
H 3 0.231 1.339 −2.475 −0.041
H 1 0.168 1.227 −2.235 1.700
H 2 0.168 −0.122 −2.951 0.819
H 9 0.161 −0.848 1.425 0.141
H 10 0.207 0.591 1.946 −0.738
H 11 0.207 −0.746 1.236 −1.611
H 12 0.380 −0.492 −0.509 1.167
The dipole moment along x is computed as µ x = ∑ i q i x i , where i is

the index for the atom, q i is its charge, and x i its coordinate. Using the
data in the table the components of the dipole moment are easily com-
puted in units of the elementary charge times Å, and then these values
are converted to Debye in the usual way. The total dipole moment is
µ = (µ 2x + µ 2y + µ z2 )1/2 .
µ x = −0.461 e Å
= (−0.461 × 10−10 m) × (1.6022 × 10−19 C)/(3.3356 × 10−30 C m)
= −2.212 D
Similarly µ y = 0.607 D and µ z = 2.897 D, giving µ = 3.695 D .

The energy of interaction of two dipoles is given by [17B.3b–685]
µ1 µ2
V= × (1 − 3 cos2 Θ)
4πε 0 r 3
With the data given, and converting to molar units
[(3.695 D) × (3.3356 × 10−30 C m)/(1 D)]2 )

V=
4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (3.0 × 10−9 nm)3
× (1 − 3 cos2 Θ) × (6.0221 × 1023 mol−1 )
= (30.4... J mol−1 ) × (1 − 3 cos2 Θ)
A plot of this function is shown in Fig. 17.14.
20
V /(J mol−1 )
−20
−40
−60
0 50 100 150 200 250 300 350
Θ/○
Figure 17.14
(b) The maximum of the dipole–dipole interaction is −61 J mol−1 which is

only 0.3% of the energy of the hydrogen bond.
I17.3 Starting from G = U − T S − tl, the differential is formed and the basic equation
dU = TdS + tdl is introduced to give
dG = dU − d(T S) − d(tl) = dU − TdS − SdT − tdl − ldt

= TdS + tdl − TdS − SdT − tdl − ldt = −SdT − ldt
The Helmholtz energy A is given by [3D.4a–93], A = U − T S. Forming the

differential gives
dA = dU − d(T S) = dU − TdS − SdT = TdS + tdl − TdS − SdT = −SdT + tdl
The Maxwell relations are derived using the method described in Section 3E.1(a)
on page 100. Because G is a state function, dG = −SdT − ldt is an exact
differential. Therefore according to the criterion in [3E.4–100], (dg/dy)x =
(dh/dx) y if d f = gdx + hdy is an exact differential, it follows that
∂S ∂l ∂S ∂l
−( ) = −( ) hence ( ) =( )
∂t T ∂T t ∂t T ∂T t
Applying the same approach to dA = −SdT + tdl gives
∂S ∂t ∂S ∂t
−( ) =( ) hence ( ) = −( )
∂l T ∂T l ∂l T ∂T l
To deduce the equation of state, the basic equation dU = TdS + tdl is divided
by dl and the condition of constant T is imposed
∂U ∂S ∂t ∂t
( ) = T ( ) + t = −T ( ) + t = t − T ( )
∂l T ∂l T ∂T l ∂T l
In the second step the Maxwell relation (∂S/∂l)T = −(∂T/∂t) l , derived above,
is used.
I17.5 The osmotic pressure Π is expressed in terms of the concentration [J] by the
virial-like expression of [5B.18–153], Π = RT ([J] + B[J]2 ). In this equation
the units of B are the inverse of the units of [J].
The osmotic virial coefficient arises largely from the effect of excluded volume.
If a solution of a macromolecule is imagined as being built by the successive
addition of macromolecules of effective radius a to the solvent, each one being
excluded by the ones that preceded it, then B is the excluded volume per mole
of molecules. The volume of a molecule is v mol = (4/3)πa 3 , but the excluded
volume is determined by the smallest distance possible between centres of two
molecules, which is 2a. Therefore the excluded volume is (4/3)π(2a)3 = 8v mol
for a pair of molecules. The volume excluded per molecule is one-half this
volume or 4v mol . Thus, for an effective radius of a = γR g = 0.85R g the osmotic
virial coefficient is
16π 16π
B = 4N A v mol = NA a3 = N A γ 3 R g3
3 3
(a) For a freely jointed chain R g = (N/6)1/2 l hence
16π N 3/2
B= NA γ3 ( )
3 6
16π 4000 3/2
= (6.0221 × 1023 mol−1 ) × [0.85×(154 × 10−12 m)]3 ( )
3 6
= 0.39 m3 mol−1 .
(b) For a chain with tetrahedral bond angles R g = (N/3)1/2 l hence
16π N 3/2
B= NA γ3 ( )
3 3
16π 4000 3/2
= (6.0221 × 1023 mol−1 ) × [0.85×(154 × 10−12 m)]3 ( )
3 3
= 1.1 m3 mol−1 .
I17.7 (a) The Lennard-Jones potential is given by [17B.12–690], VLJ (r) = 4ε{(r 0 /r)12 −
(r 0 /r)6 }. As shown in Fig. 17B.8 on page 690, the depth of the potential
well is given by ε and the position of the minimum is given by r e = 21/6 r 0 .
̃ e = 1.51 × 10−23 J and
The results for He2 give the depth of the well as hc D
the position of the minimum as R = 297 pm so it follows that
re 297 pm
ε = 1.51 × 10−23 J and r0 = 1/6
= = 265 pm
2 21/6
A plot of the potential with these values is shown in Fig. 17.15.
(b) The Morse potential is given by [11C.7–420], VM (x) = hc D ̃ e (1 − e−ax )2 ,
where x = r − r e and hc D e is the depth of the well. With this poten-
̃
tial VM (0) = 0, in contrast to the Lennard-Jones potential for which
VLJ (∞) = 0. To compare the two, the Morse potential is replotted as
VM (x) = hc D̃ e (1 − e−a(r−r e ) )2 − hc D
̃ e ; this is shown plotted in Fig. 17.15
using the given values of hc D ̃ e = 1.51 × 10−23 J and a = 5.79 × 1010 m−1 .
4
Lennard-Jones
Morse
2
V /(10−23 J)
−2
250 300 350 400 450 500
r/pm
Figure 17.15
18 Solids
18A Crystal structure

D18A.1 Lattice planes are labelled by their Miller indices h, k, and l, where h, k, and
l refer respectively to the reciprocals of the smallest intersection distances (in
units of the lengths of the unit cell, a, b and c) of the plane along the a, b, and
c axes. These axes may be non-orthogonal.
D18A.3 A space lattice is the three-dimensional structural pattern formed by lattice
points representing the locations of motifs which may be atoms, molecules, or
groups of atoms, molecules, or ions within a crystal. All points of the space
lattice have identical environments and they define the crystal structure. The
unit cell is an imaginary parallelepiped from which the entire crystal structure
can be generated, without leaving gaps, using translations of the unit cell alone.
Each unit cell is defined in terms of lattice points and the unit cell is commonly
formed by joining neighbouring lattice points by straight lines. The smallest
possible unit cell is called the primitive unit cell. Non-primitive unit cells may
exhibit lattice points within the cell, at the cell centre, on cell faces, or on cell
edges.
E18A.1(a) The separation of (hkl) planes d hk l of a cubic lattice is give by [18A.1a–733],
d hk l = a/(h 2 + k 2 + l 2 )1/2 .
(562 pm) (562 pm)
d 112 = = 229 pm d 110 = = 397 pm
(12 + 12 + 22 )1/2 (12 + 12 + 02 )1/2
(562 pm)
d 224 = = 115 pm
(22 + 22 + 42 )1/2
E18A.2(a) The separation of (hkl) planes d hk l of an orthorhombic lattice is given by

l = h /a + k /b + l /c . Therefore
2 2 2 2 2 2 2
[18A.1b–733], 1/d hk
d hk l = (h 2 /a 2 + k 2 /b 2 + l 2 /c 2 )−1/2
−1/2
32 22 12
d 321 = [ + + ]
(812 pm)2 (947 pm)2 (637 pm)2
= (2.05... × 1019 )−1/2 m = 220 pm
664 18 SOLIDS
E18A.3(a) The volume of an orthorhombic unit cell is given by V = abc, and the mass
of the unit cell m is given by m = ρV , where ρ is the mass density. Using the
estimate of mass density ρ = 3.9 g cm−3
m = abcρ = [(634 × 784 × 516) × 10−36 m3 ] × (3.9 × 106 g m−3 )

= 1.00... × 10−21 g
The mass of a unit cell is also related to the molar mass by m = nM = N M/N A ,
where n is the amount in moles of NiSO4 in a unit cell, M is the molar mass,
and N is the number of formula units per unit cell.
mN A (1.00... × 10−21 g) × (6.0221 × 1023 mol−1 )

N= = = 3.89...
M 154.75 g mol−1
If it is assumed that there are no defects in the crystal lattice N is expected to be

an integer and hence N = 4 . With this value a more precise value of the mass
density is calculated as
m NM
ρ= =
V NA V
4 × (154.75 g mol−1 )
= = 4.01 g cm−3
(6.0221 × 1023 mol−1 ) × [(634 × 784 × 516) × 10−36 m3 ]
E18A.4(a) Miller indices are of the form (hkl) where h, k, and l are the reciprocals of
the intersection distances along the a, b and c axes, respectively. If the recipro-
cal intersection distances are fractions then the Miller indices are achieved by
multiplying through by the lowest common denominator.
intersect axes at (2a, 3b, 2c) (2a, 2b, ∞c)

remove cell dimensions (2,3,2) (2,2,∞)
take reciprocals ( 12 , 13 , 21 ) ( 12 , 21 ,0)
Miller indices (323) (110)
P18A.1 A face-centred cubic unit cell has lattice points at its 8 corners and also at the
centres of its six faces. Therefore there are (8 × 81 + 6 × 12 ) = 4 lattice points per
unit cell. The mass density ρ is therefore ρ = 4m/V , where m is the mass per
lattice point and the volume V is a 3 , where a is the unit cell dimension. The
molar mass M is calculated from M = mN A
M = N A ρa 3 /4
= (6.0221 × 1023 mol−1 ) × (1.287 × 106 g m−3 ) × (12.3 × 10−9 m)3 × 1
4
= 3.61 × 105 g mol−1

P18A.3 From Fig. 18A.8 on page 731 it is seen that the unit cell can be envisaged as a
prism of height c whose base is rhombus with sides a and interior angle 120○ ,
which is depicted below.
60○

@
a@ 120○ @a@ x
a
@ @
The area of the rhombus is ax = a × a sin(60○ ), hence the volume is V = cax =

√
c × [a × a sin(60○ )] = ( 3/2)a 2 c .
P18A.5 For a monoclinic unit cell, V = abc sin β. From the information given, a =
1.377b and c = 1.436b. Because there are two napthalene molecules within
the unit cell it follows that the mass density is ρ = 2m/V , where m is the mass
per molecule given by m = M/N A , where M is the molar mass of napthalene
(128.1... g mol−1 ).
Using ρV = 2m and V = abc sin β, it follows that abc sin β = 2m/ρ and hence
abc = 2m/(ρ sin β) = 2M/(N A ρ sin β). The product abc = 1.377 × 1.436 × b 3
and so
1/3
2M
b=[ ]
N A ρ sin β×1.377×1.436
1/3
2×(128.1... g mol−1 )
=[ ]
(6.0221 × 1023 mol−1 )×(1.152 × 106 g m−3 )×sin (122.82○ )×1.377×1.436
= 605.8 pm
Thus a = 834.2 pm and c = 870.0 pm
P18A.7 The mass of the unit cell m is given by m = N M/N A , where N is the number
of monomer units per unit cell and M is the molar mass of a monomer unit.
The mass is also written in terms of the mass density ρ and the volume V as
m = ρV . Hence N M/N A = ρV and so N = ρN A V /M. The molar mass is
M = 63.55+7×12.01+13×1.0079+5×14.01+8×16.00+32.06 = 390.8... g mol−1
For a monoclinic unit cell,
V = abc sin β
= (1.0427 nm) × (0.8876 nm) × (1.3777 nm) × sin(93.254○ )
= 1.27... × 10−27 m3
hence
ρN A V (2.024 × 106 g m−3 ) × (6.0221 × 1023 mol−1 ) × (1.27... × 10−27 m3 )

N= =
M (390.8... g mol−1 )
= 3.97...
There are 4 monomer units per unit cell.

666 18 SOLIDS
P18A.9 Consider the two-dimensional lattice and planes shown in Fig. 18A.12 on page
732. The (hk0) planes intersect the a, and b axes at distances a/h and b/k from
the origin, respectively. Using trigonometry,
d hk0 d hk0 h d hk0 d hk0 k
sin ϕ = = cos ϕ = =
a/h a b/k b
Because sin2 θ + cos2 θ = 1 it follows that
d hk0 h 2 d hk0 k 2
( ) +( ) =1
a b
Rearranging gives
1 h2 k2
= +
d hk0
2 a2 b2
Because the third side of the cell is mutually perpendicular to the other two,
the extension to three dimensions simply involves adding an additional term,
as in the derivation for a cubic lattice
1 h2 k2 l 2
= + +
d hk
2
l a2 b2 c2
18B Diffraction techniques

D18B.1 This is discussed in Section 18B.1(e) on page 740.
D18B.3 The scattering factor determines how strongly an atom scatters the X-rays, and
hence how strong the contribution from a particular atom is to a reflection. It is
defined and described in Section 18B.1(c) on page 737. For forward scattering,
the scattering factor is equal to the number of electrons in the atom.
E18B.1(a) The Bragg law [18B.1b–736], λ = 2d sin(θ), is rearranged to give the glancing
angle as 2θ = 2 sin−1 (λ/2d), where d is the plane separation and λ is the
wavelength of the X-rays. For the case where λ = 154.433 pm,
2θ = 2 × sin−1 [(154.433 pm)/2 × (77.8 pm)] = 165.9...○
For the case where λ = 154.051 pm,
2θ = 2 × sin−1 [(154.051 pm)/2 × (77.8 pm)] = 163.8...○
The difference in the glancing angles is 165.9...○ − 163.8...○ = 2.14○ .
E18B.2(a) In Section 18B.1(c) on page 737 it is shown that the scattering factor in the
forward direction, f (0), is equal to the total number of electrons in the species,
N e . Thus for Br – f (0) = 36 .
E18B.3(a) The structure factor is given by [18B.3–738]

Fhk l = ∑ f j eiϕ hk l ( j)
j
where f j is the scattering factor of species j and ϕ hk l ( j) = 2π(hx j + ky j + l z j )

is the phase of the scattering from that species.
Species at the corners of the unit cell are shared between eight adjacent unit
cells so they have weight of 81 and so, if all the atoms are the same and have the
same scattering factor f , the contribution from each is 18 f . The structure factor
is
j
= f [1+e2iπk +e2iπl +e2iπ(k+l ) +e2iπh +e2iπ(h+k) +e2iπ(h+l ) +e2iπ(h+k+l ) ]

1
8
The indices h, k and l are all integers, and einπ = (−1)n for integer n. All the
exponents in the sum are even multiples of iπ, so all the exponential terms are
equal to +1. Hence Fhk l = f .
E18B.4(a) The orthorhombic C unit cell is shown in Fig. 18A.8 on page 731. The structure
factor is given by [18B.3–738]
j
where f j is the scattering factor of species j and ϕ hk l ( j) = 2π(hx j + ky j + l z j )

is the phase of the scattering from that species.
The ions at the corners of the unit cell are shared between eight adjacent unit
cells so they have weight 81 and therefore, if they all have the same scattering
factor f , the contribution from each is 81 f . As is shown in Exercise E18B.3(a),
these ions together contribute + f to the structure factor.
The ions on the faces have positions ( 12 a, 12 a, 0) and ( 12 a, 21 a, a) and are shared
between two adjacent unit cells. Each face ion thus contributes 12 × (2 f ) = f ,
where (2 f ) is the scattering factor for the face ions, given as twice that of the
other ions. The contribution to the scattering factor from the face ions is
f e2iπ( 2 h+ 2 k) + f e2iπ( 2 h+ 2 k+l ) = f (1 + e2iπ l ) eiπ(h+k) = 2 f (−1)(h+k)
1 1 1 1
The structure factor is therefore Fhk l = f + 2 f (−1)(h+k) . Therefore for (h + k)

odd, Fhk l = f − 2 f = − f , and for (h + k) even, Fhk l = f + 2 f = 3 f .
E18B.5(a) The electron density distribution ρ(r) in the unit cell is given by [18B.4–739],
ρ(r) = (1/V ) ∑ hk l Fhk l e−2πi(hx+k y+l z) , where V is the volume of the unit cell.
In this case the structure factors are only given for the x direction so the sum
is just over the index h. Furthermore, because Fh = F−h the summation can be
taken from h = 0 to h = +∞
∞ ∞
V ρ(x) = ∑ Fh e−2πihx = F0 + ∑ (Fh e−2πihx + F−h e2πihx )
h=−∞ h=1
∞ ∞
= F0 + ∑ Fh (e−2πihx + e2πihx ) = F0 + 2 ∑ Fh cos(2πhx)
h=1 h=1
668 18 SOLIDS
In this case there are a total of ten terms to include, h = 0 to 9. Figure 18.1 shows
a plot of V ρ(x) against x; the electron density is at a maximum of 110/V at
x = 0.5, the centre of the unit cell.
V ρ(x) 100
50
0.0 0.2 0.4 0.6 0.8 1.0

x
Figure 18.1
E18B.6(a) The Patterson synthesis is given by [18B.5–741],
1 2 −2πi(hx+k y+l z)
P(r) = ∑ ∣Fhk l ∣ e
V hk l
In this case the structure factors are only given for the x direction so the sum
is just over the index h. Furthermore, because Fh = F−h the summation can be
taken from from h = 0 to h = +∞. Using a similar line of argument to that in
Exercise E18B.4(a), the Patterson synthesis is
∞
V P(x) = F02 + 2 ∑ Fh2 cos(2πhx)
h=1
In this case there are a total of ten terms to include, h = 0 to 9. Figure 18.2 shows
a plot of V P(x) against x. As expected, there strong feature at the origin; this
arises from the separation between each atom and itself. There is also a strong
feature at x = 1 which indicate that atoms are separated by 1 × a unit along the
x-axis.
E18B.7(a) To constructor the Patterson map, choose the position of one atom to be the
origin (here, the boron). Then add peaks to the map corresponding to vec-
tors joining each pair of atoms (Fig. 18.3). Heavier atoms give more intense
contributions than light atoms, so peaks arising from F and F separations are
shown with greater diameter than those representing B and F separations. The
vector between atom A and atom B has the same magnitude as that between
B and A, but points in the opposite direction; the map therefore includes two
symmetry related peaks on either side of the origin. The vectors between each
atom and itself give a peak at the centre point of the Patterson map, and the
many contributions at this position create an intense peak.
1 000
V P(x)
500
−1.0 −0.5 0.0 0.5 1.0

x
Figure 18.2
BF3 Patterson map
F
RBF
RBF
B
RFF
F F
RFF
Figure 18.3
E18B.8(a) Using the de Broglie relation [7A.11–230], λ = h/p = h/(mυ), where p is the
momentum, m is the mass of a neutron and υ its speed, it follows that
h 6.6261 × 10−34 J s
υ= = = 6.1 km s−1
λm (65 × 10−12 m) × (1.6749 × 10−27 kg)
E18B.9(a) From the equipartition principle the kinetic energy is E k = 12 kT. This en-
ergy can be written in terms of the momentum as p2 /(2m) and hence p =
(mkT)1/2 . The de Broglie relation [7A.11–230], λ = h/p, is then used to find
the wavelength
h 6.6261 × 10−34 J s
λ= =
(mkT)1/2 [(1.6749 × 10−27 kg) × (1.3806 × 10−23 J K−1 ) × (350 K)]1/2
= 233 pm
E18B.10(a) Bragg’s law [18B.1b–736], λ = 2d sin θ, describes the relationship between wave-
length of the X-rays λ, the Bragg angle θ, and the plane separation d. Thus
λ = 2 × (99.3 pm) × sin (20.85○ ) = 70.7 pm
670 18 SOLIDS
E18B.11(a) As shown in Fig. 18B.10 on page 740, for the cubic I lattice reflections from
planes with h + k + l = odd are absent from the diffraction pattern. Hence the
first three possible reflections occur for planes (110), (200) and (211). Using the
Bragg law [18B.1b–736], λ = 2d hk l sin θ, and the expression for the spacing of
the planes [18A.1a–733], d hk l = a/(h 2 + k 2 + l 2 )1/2 , the following table is drawn
up
Miller indices (110) (200) (211)

d hk l a/(1 + 1 )
2 2 1/2
a/(2 ) 2 1/2
a/(22 + 12 + 12 )1/2
d hk l /pm 205.7... 145.5 118.8...
sin θ 0.174... 0.247... 0.303...
○
θ/ 10.1 14.3 17.6
E18B.12(a) The separation of the (hkl) planes of an orthorhombic lattice is given by [18A.1b–
733], 1/d hk
2
l = h /a + k /b + l /c . This distance is used with [18B.1b–736]
2 2 2 2 2 2
to compute the angle of reflection as θ = sin−1 (λ/2d hk l ).
Miller indices (100) (010) (111)

2 −2 2 2 2 2
(1/d hk l )/pm 1 /542 1 /917 12 /5422 + 12 /9172 + 12 /6452
d hk l /pm 542 917 378.0...
○
θ/ 8.17 4.82 11.8
P18B.1 The NaCl unit cell is depicted in Fig. 18B.9 on page 738. The unit cell is a cube
with volume V = a 3 where a is the unit cell side length. There are eight Na+
ions at vertices, six Na+ ions on faces, one Cl – ion at centre and 12 Cl – ions
at the edges of the unit cell. Thus in total there are four NaCl units per unit
cell. The mass density ρ is ρ = m/V where m is the total mass per unit cell. It
follows that m = 4m NaCl = 4M/N A where M is the molar mass of an NaCl unit
(M = 58.44 g mol−1 ).
The spacing of the planes is given by [18A.1a–733], d hk l = a/(h 2 + k 2 + l 2 )1/2 ;

for the (100) reflection, this evaluates to d 100 = a. The angle of refraction,
the spacing and the wavelength are related by the Bragg law, [18B.1b–736], λ =
2d sin θ which rearranges to d 100 = a = λ/(2 sin θ). Using this, the density can
be expressed as
4M 4M 32M sin3 θ
ρ= = =
NA V NA a3 NA λ3
In turn this expression is rearranged to give the wavelength in terms of the

known parameters
1/3
32M
λ=( ) sin θ
NA ρ
1/3
32 × (58.44 g mol−1 )
=( ) sin(6.0○ )
(6.0221 × 1023 mol−1 ) × (2.17 × 106 g m−3 )
= 118 pm
P18B.3 Combining Bragg’s law [18B.1b–736], λ = 2d sin θ, with the expression for the
the separation of planes for a cubic lattice [18A.1a–733], d hk l = a/(h 2 + k 2 +
l 2 )1/2 , gives sin θ = (λ/2a) (h 2 + k 2 + l 2 )1/2 .
The first three reflections for a cubic P lattice are (100), (110) and (200). Con-
sider the ratio of sin θ for the first two of these compared to ratio of sin θ for
the first two observed lines:
sin θ 110 (12 + 12 )1/2 sin θ 1st sin 22.171○
= = 1.41... = = 1.15...
sin θ 100 (12 )1/2 sin θ 2nd sin 19.076○
These do not match up, so the lattice is not cubic P. For cubic I the first three
reflections are (110), (200) and (211); making the same comparison gives
sin θ 200 (22 )1/2 sin θ 1st sin 22.171○

= 2 = 1.41... = = 1.15...
sin θ 110 (1 + 12 )1/2 sin θ 2nd sin 19.076○
These do not match up, so the lattice is not cubic I. For cubic F the first three
reflections are (111), (200) and (220).
sin θ 200 (22 )1/2 sin θ 1st sin 22.171○

= 2 = 1.15... = = 1.15...
sin θ 111 (1 + 12 + 12 )1/2 sin θ 2nd sin 19.076○
This matches well. The same procedure is used for the second and third reflec-
tions
sin θ 220 (22 + 22 )1/2 sin θ 2nd sin 32.256○
= = 1.41... = = 1.41...
sin θ 200 (22 )1/2 sin θ 3rd sin 22.171○
Again,there is a good match. Therefore silver adopts a cubic F lattice .

The lattice parameter is computed from a = λ(h 2 + k 2 + l 2 )1/2 /(2 sin θ). With
the data for the (111) reflection this gives
(154.18 pm)(12 + 12 + 12 )1/2

a= = 408.55 pm
2 sin(19.076○ )
Cubic F has four atoms per unit cell and so the mass density is ρ = 4m/V =
4M/N A V , where M is the molar mass of silver.
4M 4 × (107.87 g mol−1 )
ρ= = = 10.51 g cm−3
N A a 3 (6.0221 × 1023 mol−1 ) × (408.55 × 10−12 m)3
672 18 SOLIDS
P18B.5 The scattering factor is given by [18B.2–737]

∞ sin kr 2 4π
f (θ) = 4π ∫ ρ(r) r dr k= sin θ
0 kr λ
For ρ(r) = 3Z/4πR 3 where 0 ≤ r ≤ R and ρ(r) = 0 for r ≥ R, the integral is
evaluated by parts to give
3Z R 3Z −r cos kr R 3Z R
f (θ) = ∫ r sin kr dr = ∣ + 2 3 ∫ cos kr dr
kR 0
3 kR 3 k 0 k R 0
−3Z cos kR 3Z sin kr R 3Z
= + ∣ = (sin kR − kR cos kR)
k2 R2 k2 R3 k 0 k3 R3
6
R = 100 pm Z=3
R = 100 pm Z=6
4 R = 400 pm Z=3
f (θ)
0.000 0.002 0.004 0.006 0.008 0.010

−1
sin(θ)/(λ pm )
Figure 18.4
A plot of f (θ) against (sin θ)/λ is shown in Fig. 18.4. In the forward direction,
θ = 0, the scattering factor is equal to Z, which is the expected result because
Z is equal to the number of electrons. The scattering oscillates with increasing
angle but superimposed on this is an overall decay that becomes faster as R
increases. Thus, the larger the atom, the more concentrated is the scattering in
the forward direction.
P18B.7 The structure factor is given by [18B.3–738], Fhk l = ∑ j f j e2πi(hx j +k y j +l z j ) . Each
atom A is shared between 8 unit cells and therefore has weight 18 , whereas the
B atom contributes to just one unit cell and so has weight 1.
Fhk l = 18 f A [1+e2πik +e2πil +e2πi(k+l ) +e2πih +e2πi(h+k) +e2πi(h+l ) +e2πi(h+k+l ) ]

+ f B eπi(h+k+l )
= f A + f B (−1)(h+k+l )
where to go to the last line the relationship eπin = (−1)n for integer n is used.
The intensity of the diffraction pattern is directly proportional to the square
modulus of the structure factor.
(a) For f A = f , f B = 0 then Fhk l = f for all (hkl) planes so the diffraction
pattern will display no systematic absences.
(b) For f B = 21 f A then Fhk l = f A [1 + 12 (−1)(h+k+l ) ]. For all reflections with
(h + k + l) odd the intensity will be proportional to the square of 21 f A , and
for all reflections with (h + k + l) even the intensity will be proportional
to the square of 32 f A .
(c) For f A = f B = f then Fhk l = f [1 + (−1)(h+k+l ) ]. For all reflections with
(h + k + l) odd the structure factor is zero, and for all reflections with
(h + k + l) even the structure factor is 2 f . The diffraction pattern will
show systematic absences for lines where (h + k + l) is odd.
18C Bonding in solids

D18C.1 In a face-centred cubic close-packed lattice, there is an octahedral hole in the
centre. The rock-salt structure can be thought of as being derived from an fcc
structure of Cl – ions in which Na+ ions have filled the octahedral holes.
The caesium-chloride structure can be considered to be derived from the ccp
structure by having Cl – ions occupy all the primitive lattice points and octahe-
dral sites, with all tetrahedral sites occupied by Cs+ ions. This is rather difficult
to visualize and describe without carefully constructed figures or models: refer
to S.-M. Ho and B.E. Douglas, J. Chem. Educ. 46, 208, (1969), for the appropri-
ate diagrams.
E18C.1(a) The lattice enthalpy ∆H L is the change in standard molar enthalpy for the pro-
cess MX(s) → M+ (g) + X− (g) and its equivalent. The value of the lattice en-
thalpy is determined indirectly using a Born–Haber cycle, as shown in Fig. 18.5
(all quantities are given in kJ mol−1 ). From the cycle it follows that
−635 kJ mol−1 + ∆H L = (178 + 1735 + 249 − 141 + 844) kJ mol−1
Thus ∆H L = 3500 kJ mol−1

E18C.2(a) The densest packing arrangement possible for cylinders is the hexagonal pack-
ing shown in Fig. 18.6; the unit cell is the rhombus indicated, and the internal
angles in this rhombus are 60○ and 120○ . The centre-to-centre spacing of the
cylinders is 2R, where √R is the radius of one cylinder. The distance h is given
by h =√2R sin 60○ = R 3, therefore the area of the rhombus is base × height
= 2R 2 3.
If the depth of the unit cell is z then
√ the volume of the unit cell is the area of the
rhombus times this depth, V = 2 3R 2 z. Each cylinder occupies volume πR 2 z
and there is a total of one cylinder per unit cell. The packing density, f , is
πR 2 z π
f = √ = √ = 0.9069
2 3R z 2 3
2
674 18 SOLIDS
Ca2+ (g) + O2− (g)
Ca2+ (g)+O(g)
249 844
Ca2+ (g)+ 12 O2 (g) −141 Ca2+ (g) + O – (g)
∆H L
1735
Ca(g)+ 12 O2 (g)
178 Ca(s)+ 12 O2 (g)

−635 CaO(s)
Figure 18.5
Figure 18.6
E18C.3(a) The packing fraction is f = NVa /Vc where N is the number of spheres per unit
cell, Va = 4πR 3 /3 is the volume of each sphere of radius R, and Vc is the volume
of the unit cell.
(i) For a primitive cubic unit cell the spheres touch along the edges of the
cell, so the edges of the cube have length 2R and hence Vc = (2R)3 . There
is one sphere per unit cell, N = 1, and therefore
4πR 3 /3 π
f = = = 0.5236
8R 3 6
(ii) For a bcc unit cell, the spheres touch along the body diagonal so the
length of this diagonal is 4R. Imagine a right-angle triangle in which the
hypotenuse is the body diagonal, and the other
√ two sides are an edge of the
cube, length a, and a face diagonal,
√ length 2a. It follows that (4R) 2
√ 3=
a + 2a and hence a = 4R/ 3. The volume is therefore Vc = (4R/ 3) ,
2 2
and as N = 2 it follows
√
that side of the cube is 4R/ 3 so N = 2, Va = 4πR 3 /3 and
It follows √
Vc = (4R/ 3)3 . Thus
√
2 × 4πR 3 /3 3π
f = √ = = 0.6802
(4R/ 3)3 8
(iii) For a fcc unit cell, the spheres touch along a face diagonal which therefore
has length 4R. If the edge of the cube has length a it√follows, by consid-
ering a face, that (4R)2 = a 2 + a 2 and hence a = 2 2R. The volume is
√
therefore Vc = (2 2R)3 , and because N = 4 the packing fraction is
4 × 4πR 3 /3 π
f = √ = √ = 0.7405
16 2R 3 3 2
E18C.4(a) The coordination number N of an ionic lattice depends on the radius ratio of
the cation and anion of the lattice. The radius-ratio rule, which considers the
maximum possible packing density of hard spheres of a given radius around a
hard sphere of a different radius, provides a method to determine the structure
type. The radius ratio is γ = r s /r l where r s is the radius of the smallest ion and
r l is the radius of the largest ion. If γ ≤ (21/2 − 1) then N < 6; for (21/2 − 1) <
γ < (31/2 − 1) then N = 6; for γ ≥ (31/2 − 1) then N = 8.
The range for sixfold coordination is therefore 0.414 < γ < 0.732, and hence
r l × 0.414 < r s < r l × 0.732. For the case of the Cl – anion (181 pm) × 0.414 =
75.0 pm and (181 pm) × 0.732 = 132.5 pm. Therefore for sixfold coordination
the smallest radius for the cation is 75.0 pm , whilst for eightfold coordination
the smallest radius is 133 pm .
E18C.5(a) The unit cell volume V is related to the packing density f and the atomic vol-
ume v by f V = v. Assuming the atoms can be approximated as spheres then
v = 4πR 3 /3 where R is the atomic radius. Using the packing densities calculated
in Exercise E18C.3(a) and the given data
Vbcc v bcc f hcp (R bcc )3 f hcp 142.53 × 0.7405...

= × = = = 1.016...
Vhcp v hcp f bcc (R hcp )3 f bcc 145.83 × 0.6802...
Thus transformation from hcp to bcc causes cell volume to expand by 1.6%
P18C.1 The packing fraction is f = NVa /Vc , where N is the number of atoms per unit
cell, Va = 4πR 3 /3 is the volume of an atom of radius R, and Vc is the unit cell
volume. The structure of diamond is shown in Fig. 18C.15 on page 752: there
are 8 atoms at the vertices of the cell (weight 18 ), 6 atoms at the face-centres
(weight 12 ), and 4 atoms within the unit cell (weight 1), giving a total of 8 atoms
per unit cell.
The two nearest-neighbour atoms which touch along the body diagonal are at
locations (0,0,0) and ( 41 , 14 , 41 ), where the coordinates are expressed as fractions
of the length of the side of the unit cell, a. These two atoms are at the√ opposite
corners of a small cube with edge a/4. The body diagonal of a cube is 3 times
the length of the
√ edge, so it follows that the length of the body diagonal of this
small cube is 3a/4. As the two atoms √ touch along this diagonal, this distance
is also equal to 2R, hence a = 8R/ 3. The packing fraction is therefore given
by
√
8 × 4πR 3 /3 8 × 4πR 3 /3 3π
f = = √ 3 = = 0.3401
a 3
(8R/ 3) 16
676 18 SOLIDS
P18C.3 (a) Close-packed spheres form a face-centred cubic structure, which is shown
in Exercise E18C.3(b) to have a packing density of f = 0.7405. A sam-
ple of volume V of diamond therefore contains f V /(4πR 3 /3) carbon
atoms, where R is the atomic radius. The mass of these carbon atoms is
f V /(4πR 3 /3)×(M/N A ), where M is the molar mass of carbon, therefore
the mass density is
mass f V /(4πR 3 /3) × (M/N A ) 3f M

ρ= = =
volume V 4πN A R 3
With the data given
3f M 3 × (0.7405) × (12.01 g mol−1 )

ρ= =
4πN A R 3 4π × (6.0221 × 1023 mol−1 ) × ( 21 × 154.45 pm)3
= 7.655 g cm−3
(b) The experimentally determined density is significantly lower than that

calculated on the assumption of a fcc structure. This implies that atoms
which are assumed to be in contact in the fcc structure are in fact further
apart, and in turn this can be ascribed to the highly directional (tetra-
hedal) bonding known to occur in diamond.
In Problem P18C.1 it is shown that the packing density for the diamond
structure is f = 0.3401. With this value the predicted density is reduced
to
(7.655 g cm−3 ) × (0.3401/0.7405) = 3.516 g cm−3
which is in close agreement with the experimental value.
P18C.5 The formation of a band in one dimension results in a set of states which spread,
to a finite extent, above and below the energy of the original atomic orbital from
which the band is created (Fig. 18C.6(e) on page 746). If the system is extended
to two dimensions, each one of these original states itself becomes the starting
point for a band arising from overlap in the second dimension; this is illustrated
in a highly schematic way in Fig. 18.7.
The original atomic orbital is indicated on the left, and overlap of these results
in a one-dimensional band, indicated by the dotted lines. Then, each state in
this band itself gives rise to a further band when interactions are allowed in a
second dimension. This is illustrated for the states at the very top and bottom
of the band (shown by dotted lines), and a selection of levels between. The band
clearly increases in overall width, but in addition the density of states increases
in the centre of the two-dimensional band as many of one-dimensional bands
overlap here. In contrast, at the extremities of the band, fewer one-dimensional
bands are overlapping. These are only qualitative arguments, but they are in-
dicative of the origin for the change in the density of states which is indeed
observed.
P18C.7 (a) The rock salt (NaCl) structure, shown in Fig. 18C.10 on page 748, exhibits
sixfold coordination. Let the radius of the chloride ions be r l and that of
atomic
1D
2D
Figure 18.7
the sodium ions be r s ; the radius ratio γ is defined as γ = r s /r l . The lowest

value for γ occurs when r l is as large as possible, and the limit of its value
will be when the chloride ions just touch along the face diagonal. In this
limit, the length of the face diagonal will be 4r l ; if the edge of the unit cell
is of length a, it follows that (4r l )2 = 2a 2 .
When the chloride ions are just touching, γ is further minimized by mak-
ing the sodium ions as large as possible, and the limit of this is set by the
point at which the sodium ions and chloride ions touch along the edge of
a the cube. In this limit 2r l + 2r s = a.
Combining these two results establishes the relationship between the two
radii
(4r l )2 = 2a 2 = 2(r l + r s )2 hencer s2 + 2r l r s − r l2 = 0
√
The quadratic in r s is solved to give r s = (−1 ± 2)r l ; of these solutions,
√
only the one with the positive √ sign is physically reasonable so r s = ( 2 −
1)r l and hence γ = r s /r l = 2 − 1 = 0.414.
(b) Eightfold coordination is shown in Fig. 18C.9 on page 748. The limit on
the size of the chloride ions is when they touch along the edge, which is
when 2r l = a. The limit on the size of the caesium ion is when it touches
the chloride ions√ along the body diagonal. Given√ that the length of the
body diagonal is 3a, the condition is 2r l + 2r s = 3a. It follows that
√ √
2r l + 2r s = 3a = 3(2r l )
√ √
Solving this gives r s = ( 3 − 1)r l and hence γ = r s /r l = 3 − 1 = 0.732.
P18C.9 The contribution of the Coulomb interaction to lattice energy, E p , is given by
equation [18C.3–750] and the positive contribution due to overlap of atomic
orbitals, E p∗ , is given by equation [18C.4–750].
∣z A z B ∣N A e 2 ∗
E p,tot = E p + E p∗ = −A + N A C ′ e−d/d (18.1)
4πε 0 d
The minimum in this is found by differentiating E p,tot with respect to d and
setting the result equal to zero.
dE p,tot ∣z A z B ∣N A e 2 N A C ′ −d/d ∗
=A − e =0
dd 4πε 0 d 2 d∗
678 18 SOLIDS
∗ ∣z A z B ∣e 2 d ∗
hence C ′ e−d/d = A
4πε 0 d 2
Note that in this expression the distance d is now that which gives the minimum
∗
potential energy. Substituting this expression for C ′ e−d/d into eqn 18.1 gives
∣z A z B ∣N A e 2 d∗
E p,min = −A (1 − )
4πε 0 d d
18D The mechanical properties of solids

D18D.1 If, when the stress is removed, an object returns to the same shape it had before
the stress was applied, the deformation is said to be elastic. If the result of
the stress is a change in shape which remains even after the stress has been
removed, the deformation is said to be plastic.
E18D.1(a) Poisson’s ratio, ν P , is defined in [18D.2–755], ν P = εtrans /εnorm , where εtrans is
the transverse strain and εnorm is the normal (uniaxial) strain. If the normal
strain is 1.0%, it follows that the change in length ∆Lnorm is
∆Lnorm = εnorm Lnorm = 0.01 × (1.0 × 10−2 m) = 1.0 × 10−4 m
The transverse strain is εtrans = ν P εnorm , so the change in dimension in the

transverse direction ∆Ltrans is
∆Ltrans = εtrans Ltrans = ν P εnorm Ltrans = 0.45×0.01×(1.0×10−2 m) = 4.5×10−5 m
It is expected that the result of applying the stress will be to decrease the size
of the cube in the transverse dimension (that is ∆Ltrans is negative), and that
the decrease will be the same in each transverse direction. The volume after the
stress has been applied is therefore
(1.0×10−2 m+1.0×10−4 m)×(1.0×10−2 m−4.5×10−5 m)2 = 1.000930...×10−6 m3
The change in volume is 1.000930...×10−6 m3 −1.0×10−6 m3 = 9.3 × 10−4 cm3
E18D.2(a) The relationship between the applied pressure p, the bulk modulus K, and the
fractional change in volume ∆V /V is given by [18D.1b–754], K = p/(∆V /V ).
For a fractional change of 1%, ∆V /V = 0.01, the pressure is p = 0.01 × 3.43 ×
109 Pa = 34.3 MPa .
E18D.3(a) The Young’s modulus E is related to the stress σ and the strain ε by [18D.1a–754],
E = σ/ε. The stress is given by σ = F/A where F is the force applied and A is
the cross-sectional area. Hence
F 500 N
σ= = = 1.59... × 108 Pa = 1.6 × 102 MPa
A π(1.0 × 10−3 m)2
∆L σ 1.59... × 108 Pa
ε= = = = 0.036...
L E 4.42 × 109 Pa
Hence the percentage increase in length L is 3.6%
P18D.1
µ(3λ + 2µ) 3λ + 2µ λ
E= [i] K= [ii] G = µ [iii] νP = [iv]
λ+µ 3 2(λ + µ)
Rearranging [iv] to make λ the subject gives
2ν P µ
λ= [v]
1 − 2ν P
Substituting [v] into [i], and then using [iii] in the final step gives
6ν P µ+2µ(1−2ν P )
µ( 1−2ν P
) 6ν P µ + 2µ(1 − 2ν P )
E= 2ν µ+µ(1−2ν )
= = 2µ(1 + ν P ) = 2G(1 + ν P )
( P 1−2ν P P ) 2ν P + (1 − 2ν P )
It therefore follows that G = E/[2(1 + ν P )], which is the first relationship to be

shown. Substituting [v] into [ii], and recalling that E = 2µ(1 + ν P ), gives
3(2ν P µ) + 2µ(1 − 2ν P ) 2µ(1 + ν P ) E

K= = =
3(1 − 2ν P ) 3(1 − 2ν P ) 3(1 − 2ν P )
as required
18E The electrical properties of solids

D18E.1 The Fermi–Dirac distribution takes into account the effect of the Pauli exclu-
sion principle, which is that no more than two electrons may occupy any one
state. In contrast, the Boltzmann distribution places no restriction on the num-
ber of particles that can occupy a given state; such a distribution cannot, in
general, be used to described the behaviour of electrons.
In both the Boltzmann and Fermi–Dirac distributions the probability of a state
being occupied depends on its energy and the temperature, and this probability
tails off exponentially as the energy is increased. However, in the Fermi–Dirac
distribution an additional parameter, the chemical potential µ, appears. At T =
0 the probability of states with energy < µ being occupied is 1, and states at
higher energies are not occupied. At a finite temperature, the probability of the
state with energy µ being occupied is 12 .
680 18 SOLIDS
E18E.1(a) The Fermi–Dirac distribution is given by [18E.2b–759], f (E) = 1/[e(E−E F )/k T +
1], where f (E) is the probability of occupation of a state with energy E, and
E F is the Fermi energy. In this case E F = 1.00 eV = 1.60... × 10−19 J, using the
conversion factor from inside of the front cover. With some rearrangement of
the expression for f (E) it follows that
E = kT ln[1/ f (E) − 1] + E F
= (1.3806 × 10−23 J K−1 ) × (298 K) × ln (1/0.25 − 1) + (1.60... × 10−19 J)
= 1.64... × 10−19 J = 1.03 eV
E18E.2(a) Arsenic is a Group 15 element and germanium is a Group 14 element. Thus,

an electron can be transferred from an arsenic atom into the otherwise empty
conduction band, thereby increasing the conductivity of the material relative
to pure germanium. This type of doping results in an n-type semi-conductor.
E18E.3(a) Assuming that the temperature, T, is not so high that many electrons are excited
to states above the Fermi energy, E F , the Fermi–Dirac distribution can be writ-
ten as [18E.2b–759], f (E) = 1/[e(E−E F )/k T + 1], where f (E) is the probability
of occupation of a state with energy E.
For E = E F + kT, f (E F + kT) = 1/[e(E F +k T−E F )/k T + 1] = 1/[e1 + 1] = 0.269
P18E.1 The Fermi–Dirac distribution is given by [18E.2b–759], f (E) = 1/[e(E−E F )/k T +
1], where f (E) is the probability of occupation of a state with energy E, and
E F is the Fermi energy. Let x = (E − E F )/E F and y = E F /kT so that f (E) can
be written as f (x, y) = 1/(ex y + 1). Note that x can be negative for energies
below the Fermi energy, but y must always be positive.
A set of curves for different combinations of x and y are shown in Fig. 18.8.
Note that as T → ∞, y → 0 and f → 12 since all available energy states have the
same probability of 12 of being occupied. Also, as T → 0, y → ∞ and f tends
towards a step distribution for which f = 1 for x < 0 and f = 0 for x > 0.
P18E.3 Substituting eqn [18E.2a–758] into eqn [18E.1–758] and integrating over the full
energy range gives
∞ ∞ ∞ ρ(E)
N =∫ dN(E) = ∫ ρ(E) f (E) dE = ∫ dE
0 0 0 e(E−µ)/k T + 1
Assuming that ρ is independent of temperature then in order for N to remain

constant as the temperature is increased from T = 0 it follows that e(E−µ)/k T
must remain constant. Hence as T is increased, (E − µ) must increase and
therefore the value of µ must decrease.
P18E.5 The arrangement of bands in a semiconductor is shown in Fig. 18E.4 on page

758. An n-type semiconductor consists of a host sample of a Group 14 element
f (E) 0.5
0 4
−4 −2 2
0 2 EF
4 0 kT
E−E F
EF
Figure 18.8
doped with Group 15 atoms. The presence of the Group 15 atoms results in
occupied donor levels at energies just below the bottom of the conductance
band, as shown in Fig. 18E.6 on page 759. The energy gap between the donor
levels and the conductance band is significantly smaller than that between the
valence band and the conduction band.
At T = 0 the the valence band is full and the conductance band empty: the ma-
terial is therefore an insulator. When the temperature is increased to the point
where kT is comparable to the energy separation between the donor states and
the conduction band, electrons will be promoted from these states into the
conduction band and the material will start to conduct. As the temperature
is raised further a point will be reached when just about all the electrons from
the donor levels have been promoted. Now the conductivity no longer increases
with temperature and a plateau is reached.
If the temperature is raised much higher, electrons will start to be excited from
the valence band into the conduction band, and the conductivity will start to
rise once more.
18F The magnetic properties of solids

D18F.1 Suppose that the molecules in a sample possess a permanent magnetic dipole
moment. In the absence of an applied magnetic field, these dipoles will point in
random directions because there is no energetic preference for them to point in
any particular direction. The sample therefore has no net magnetic moment.
If a magnetic field is applied, then it will be energetically favourable for the
dipoles to point in certain directions. When averaged over the whole sample the
contributions from the dipoles will not cancel and the result is that the sample
682 18 SOLIDS
will have a net magnetic moment. This is the origin of the magnetization of the
sample.
The same idea applies when considering the interaction between a permanent
electric dipole and an applied electric field. In the presence of the field the
dipoles favour certain directions and so when averaged over the sample they
do not cancel. The result is a net electric moment of the sample, called the
polarization. In both cases the magnetization or polarization depends on the
competition between the randomizing effect of thermal motion and the order-
ing effect of the applied field.
E18F.1(a) The molar susceptibility χ m of a substance is given by [18F.4b–763], the Curie
law,
C N A g e2 µ 0 µ B2 S(S + 1)
χm = where C =
T 3k
This is rearranged to give the spin quantum number as
3kT χ m
S(S + 1) =
N A g e2 µ 0 µ B2
3 × (1.3806 × 10−23 J K−1 ) × (294.53 K)
=
(6.0221 × 1023 mol−1 ) × (2.0023)2
1.463 × 10−7 m3 mol−1
×
(1.2566 × 10−6 J s2 C−2 m−1 ) × (9.2740 × 10−24 J T−1 )2
= 6.839...
Note the conversion of the molar magnetic susceptibility from units of cm3 mol−1
to m3 mol−1 . To sort out the units the relations 1 T = 1 kg s−2 A−1 and 1 A =
1 C s−1 , hence 1 C = 1 A s, are useful. The value of S is found by solving the
quadratic
√
S 2 + S − 6.839... = 0 S = 21 (−1 ± 1 + 4 × 1 × 6.839...) = −0.500 ± 2.662...
The root S = −3.16 is non-physical. The other root, S = 2.16, implies an

effective number of electrons of 2 × 2.16 = 4.3 (higher precision is not justified
because the expected result is an integer).
The high-spin arrangement of electrons in Mn2+ has 5 unpaired electrons. The
discrepancy arises because the analysis here considers only the contribution
from the electron spins and does not include any possible orbital contribution;
in addition, the effect of interactions between the spins is not considered.
E18F.2(a) The spin contribution to the molar magnetic susceptibility is given by equation
the Curie law, [18F.4b–763]
C N A g e2 µ 0 µ B2 S(S + 1)
χm = where C=
T 3k
If octahedral coordination is assumed, the 9 d electrons in Cu2+ are arranged

as t62g e3g so theoretically there is one unpaired electron and S = 12
N A g e2 µ 0 µ B2 (6.0221 × 1023 mol−1 ) × (2.0023)2

=
3k 3 × (1.3806 × 10−23 J K−1 )
× (1.2566 × 10−6 J s2 C−2 m−1 ) × (9.2740 × 10−24 J T−1 )2
= 6.302... × 10−6 m3 K mol−1
To sort out the units the relations 1 T = 1 kg s−2 A−1 and 1 A = 1 C s−1 , hence
1 C = 1 A s, are useful. The molar susceptibility follows as
(6.302... × 10−6 m3 K mol−1 ) × ( 12 )( 21 + 1)
χm = = 1.59 × 10−8 m3 mol−1
298 K
E18F.3(a) Superconductors classed as Type I show abrupt loss of superconductivity when
an applied magnetic field exceeds a critical value Hc characteristic of the mate-
rial. The dependence of Hc on T is given by [18F.5–764], Hc (T) = Hc (0) [1 − T 2 /Tc2 ],
provided T ≤ Tc . For Nb,
(6 K)2
Hc (6.0 K) = (158 kA m−1 ) [1 − ] = 95 kA m−1
(9.5 K)2
The material is superconducting at 6.0 K for 95 kA m−1 and weaker applied

field strengths.
E18F.4(a) The magnetic moment m is given by [18F.3–763], m = g e [S(S +1)]1/2 µ B , where
g e = 2.0023 and µ B = eħ/(2m e ). For CrCl3 , 3.81µ B = g e [S(S + 1)]1/2 µ B , the
constant µ B cancels leaving a quadratic which is solved for S
2 √
S 2 +S−(3.81/2.0023) = 0 S = 12 (−1± 1 + 4 × 1 × 3.620...) = −0.500±1.967...
Of the two solutions, S = −2.47 is non-physical, and the solution S = 1.47
is close to S = 23 . A reasonable conclusion is therefore that CrCl3 has three
unpaired electrons.
E18F.5(a) The molar susceptibility χ m of a substance is given by [18F.2–762], χ m = χVm ,
where χ is the volume magnetic susceptibility and Vm is the molar volume.
The mass density ρ can be written ρ = M/Vm , hence Vm = M/ρ. With the data
given
χM (−7.2 × 10−7 ) × (6 × 12.01 + 6 × 1.0079) g mol−1
χ m = χVm = =
ρ 0.879 g cm−3
= −6.4 × 10−11 m3 mol−1
P18F.1 The spin contribution to the molar magnetic susceptibility is given by the Curie
lawm [18F.4b–763],
C N A g e2 µ 0 µ B2 S(S + 1)
χm = where C=
T 3k
684 18 SOLIDS
N A g e2 µ 0 µ B2 (6.0221 × 1023 mol−1 ) × (2.0023)2

=
3k 3 × (1.3806 × 10−23 J K−1 )
× (1.2566 × 10−6 J s2 C−2 m−1 ) × (9.2740 × 10−24 J T−1 )2
= 6.302... × 10−6 m3 K mol−1
To sort out the units the relations 1 T = 1 kg s−2 A−1 and 1 A = 1 C s−1 , hence
1 C = 1 A s, are useful.
For S = 2
(6.302... × 10−6 m3 K mol−1 ) × (2)(3)
χm = = 1.27 × 10−7 m3 mol−1
298 K
For S = 3
(6.302... × 10−6 m3 K mol−1 ) × (3)(4)
χm = = 2.54 × 10−7 m3 mol−1
298 K
For S = 4
(6.302... × 10−6 m3 K mol−1 ) × (4)(5)
χm = = 4.23 × 10−7 m3 mol−1
298 K
The Boltzmann factor, e−E/RT , represents the probability of a state of energy E

relative to the probability of a state with energy E = 0. If the state S = 3 has
relative energy E = 0 with Boltzmann factor f = 1 then states S = 2 and S = 4
have probability
J mol−1 )/(8.3145 J K−1 mol−1 )×(298 K)
f = e(−50×10 = 1.72 × 10−9
3
The populations of the states with S = 2 and S = 4 is therefore negligible, and

hence the molar susceptibility is well-approximated by the molar susceptibility
of the state with S = 3, 2.54 × 10−7 m3 mol−1 .
18G The optical properties of solids

D18G.1 See Section 18G.1 on page 766.
E18G.1(a) The energy gap is given by
hc (6.6261 × 10−34 J s) × (2.9979 × 108 m s−1 )
∆ε = hν = =
λ (350 × 10−9 m)
= 5.67... × 10−19 J
Converting to eV the band gap is
(5.67... × 10−19 J) × [1 eV/(1.6022 × 10−19 J eV−1 )] = 3.54 eV
P18G.1 (a) To find whether or not a vector is an eigenvector of the hamiltonian ma-
trix, the matrix is allowed to act on the vector
ν̃mon β̃ 1 ν̃ + β̃ 1
( ) ( ) = ( mon ) = (ν̃mon + β̃) ( )
β̃ ν̃mon 1 ν̃mon + β̃ 1
Acting on the vector with the hamiltonian matrix regenerates the original
vector times a constant, which is the eigenvalue ν̃+ = ν̃mon + β̃. Similarly,
for the second proposed eigenvector
ν̃mon β̃ 1 ν̃ − β̃ 1
( )( ) = ( mon ) = (ν̃mon − β̃) ( )
β̃ ν̃mon −1 −ν̃mon + β̃ −1
The vector is indeed an eigenvector with eigenvalue ν̃− = ν̃mon − β̃.

(b) The normalisation factor N is calculated by evaluating the following in-
tegral
I = ∫ Ψ+∗ Ψ+ d τ = ∫ [Ψb∗ (1) + Ψb∗ (2)][Ψb (1) + Ψb (2)] dτ
= ∫ Ψb∗ (1)Ψb (1) dτ + ∫ Ψb∗ (2)Ψb (2) dτ + 2 ∫ Ψb∗ (1)Ψb (2) dτ
Assuming that Ψb (i) is normalised the first and second integrals are = 1,
and with the definition S = ∫ Ψb∗ (1)Ψb (2) dτ the third term is 2S; overall
I = 2(1+S). Division of the wavefunction by I 1/2 therefore normalizes the
function, so the normalization constant is N+ = [2(1 + S)]−1/2 . A similar
calculation gives the normalization constant for Ψ− as N− = [2(1−S)]−1/2
(c) The integral µ dim = ∫ Ψ±∗ µ̂Ψ0 dτ is evaluated by substituting in the given
forms of Ψ± and Ψ0 and using the definition µ mon = ∫ Ψb∗ (i) µ̂Ψa (i) dτ.
Note that ∫ Ψb∗ (2) µ̂Ψa (1) dτ = ∫ Ψb∗ (1) µ̂Ψa (2) dτ = 0 because these
correspond to transitions from a level of one monomer to a level of the
other monomer.
µ dim = ∫ Ψ±∗ µ̂Ψ0 d τ
1 1
=∫ [Ψb (1) ± Ψb (2)]∗ µ̂ 1/2 [Ψa (1) + Ψa (2)] dτ
[2(1 ± S)]1/2 2
µ mon µ mon
⎡ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ
1 ⎢
⎢ ∫ Ψ ∗ (1) µ̂Ψa (1) dτ ± ∫ Ψ ∗ (2) µ̂Ψa (2) dτ
= b b
2(1 ± S) ⎢⎢
1/2
⎣
=0 =0
³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ³¹¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹· ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹ ¹µ ⎤
⎥
± ∫ Ψb∗ (2) µ̂Ψa (1) dτ + ∫ Ψb∗ (1) µ̂Ψa (2) dτ ⎥⎥
⎥
⎦
1
= (µ mon ± µ mon )
2(1 ± S)1/2
Hence for the excited state wavefunction Ψ+ , µ dim = (1 + S)−1/2 µ mon

and for the wavefunction Ψ− , µ dim = 0 .
686 18 SOLIDS
P18G.3 An incident electric field E induces a dipole moment µ in a material. If the

response is non-linear then, according to [18G.1–768], µ = αE + 12 βE 2 . If there
are two electric fields applied at frequencies ω 1 and ω 2 , the total electric field is
E = E 1 cos ω 1 t + E 2 cos ω 2 t. Expansion of the E 2 factor in the non-linear term
1
2
βE 2 gives the non-linear response
E 2 = (E 1 cos ω 1 t + E 2 cos ω 2 t)2

= E 12 cos2 (ω 1 t) + E 22 cos2 (ω 2 t) + 2E 1 E 2 cos(ω 1 t) cos(ω 2 t)
= 12 E 12 [1 + cos(2ω 1 t)] + 12 E 22 [1 + cos(2ω 2 t)] + E 1 E 2 cos([ω 1 + ω 2 ]t)
+ E 1 E 2 cos([ω 1 − ω 2 ]t)
The trigonometric identities
cos2 A = 21 (1 + cos 2A) and cos A cos B = 12 [cos (A − B) + cos (A + B)]
are used to generate the final expression.

The induced dipole therefore has components oscillating at 2ω 1 , 2ω 2 , (ω 1 +ω 2 )
and (ω 1 − ω 2 ), and each of these can result in radiation at that frequency. Thus,
a medium with a non-linear response may result in the generation of sum and
difference frequencies (as well as harmonics).

I18.1 The scattering factor f (θ) is given by [18B.2–737]
∞ sin kr 2 4π
f (θ) = 4π ∫ ρ(r) r dr k= sin θ
0 kr λ
The quantity 4πr 2 ρ(r) is identified as the radial distribution function P(r),
which is given in terms of the radial wavefunction of the orbital, P(r) = R(r)2 r 2
∞ sin kr
f (θ) = ∫ P(r) dr
0 kr
The requested plot is of f (θ) as a function of ξ = sin θ/λ, hence k = 4πξ and
∞ sin(4πξr) 1 ∞
f (θ) = ∫ P(r) dr = ∫ R(r)2 sin(4πξr) r dr
0 4πξr 4πξ 0
From Table 8A.1 on page 286 the 1s hydrogenic orbital the radial wavefunction
is R(r) = 2(Z/a 0 )3/2 e−Zr/a 0 , and so the radial distribution function is P(r) =
R(r)2 r 2 = 4(Z/a 0 )3 r 2 e−2Zr/a 0 . The integral has to be evaluated numerically
using mathematical software. The parameter ξ is some fraction of 1/λ, where
λ is the wavelength of the X-rays used. Typically λ = 100 pm so ξ is less
than 1010 m−1 . The upper limit of the integration can conveniently be set to
a modest multiple of the Bohr radius, say 100a 0 , because beyond this distance
the electron density will be negligible. The computed scattering factor is shown
in Fig. 18.9 for Z = 1 and Z = 2. As expected, for forward scattering, ξ = 0,
the scattering factor is equal to the number of electrons, in this case 1. As ξ
increases the scattering factor falls off, but the decay is slower for the more
compact orbital with Z = 2.
For a 2s hydrogenic orbital
R(r) = 8−1/2 (Z/a 0 )3/2 (2 − Zr/a 0 )e−Zr/2a 0

P(r) = 8−1 (Z/a 0 )3 (2 − Zr/a 0 )2 r 2 e−Zr/a 0
The computed scattering factor is shown in Fig. 18.10 for Z = 1 and Z = 2.

As for the 1s orbital the scattering factor is equal to 1 for forward scattering.
Compared to the 1s the scattering from 2s falls off more quickly and also shows
some small negative excursions.
1.0
Z=1
0.8 Z=2
0.6
f
0.4
0.2
0.0
0.000 0.002 0.004 0.006 0.008 0.010
ξ/(1012 m)
Figure 18.9
1.0
Z=1
Z=2
0.5
f
0.0
0.000 0.001 0.002 0.003 0.004 0.005
ξ/(10 12
m)
Figure 18.10
I18.3 The spacing of the {hkl} planes in a cubic lattice is given by [18A.1a–733],
d hk l = a/(h 2 + k 2 + l 2 )−1/2 . This is used with the Bragg law [18B.1b–736],
688 18 SOLIDS
λ = 2d sin θ, to give
2a sin θ λ(h 2 + k 2 + l 2 )1/2

λ= hence a=
(h 2 + k 2 + l 2 )1/2 2 sin θ
For the (111) reflection this becomes a = 31/2 λ/2 sin θ.
at 100 K a(100 K) = 31/2 ×(154.0562 pm)/2 sin(22.0403○ ) = 355.53... pm
at 300 K a(300 K) = 31/2 ×(154.0562 pm)/2 sin(21.9664○ ) = 356.66... pm

The volume is V = a 3 , thus the change in the volume is δV = a 3 (300 K) −
a 3 (100 K). The thermal expansion coefficient is therefore
1 δV 1 a 3 (300 K) − a 3 (100 K)
α= = 3
V δT a (100 K) δT
1 (356.66... pm)3 − (355.53... pm)3
= = 4.811 × 10−5 K−1
(355.53... pm)3 (300 − 100) K
If the average volume is used in the denominator, α = 4.788 × 10−5 K−1 .

19 Processes at solid surfaces
19A An introduction to solid surfaces

D19A.1 The Auger effect is described in Section 19A.3(b) on page 780. In Auger electron
spectroscopy the electron beam is only able to affect species on the surface
and perhaps a few layers into the bulk, and therefore the emitted electrons are
characteristic of the surface and the species adsorbed onto it. The energies of
the emitted electrons are characteristic of the material present and so provide
a fingerprint of the sample.
In scanning Auger microscopy (SAM) the electron beam is narrowly focused so

that electrons are excited only from a small region (of dimension about 50 nm).
By scanning the beam across the sample it is possible to build up a map of the
composition.
In scanning tunnelling microscopy (STM) a topological map of the surface is

built up, and under favourable circumstances individual atoms may be identi-
fied; the resolution is thus much greater than that achieved in SAM. However,
STM gives no clues as to the identity of the atoms on the surface, in contrast to
SAM.
E19A.1(a) For a process to be spontaneous it must be accompanied by a reduction in the
Gibbs energy, that is ∆G < 0. The adsorption of a gas on a surface is likely to
be accompanied by a significant reduction in entropy on account of the loss of
translational degrees of freedom, therefore ∆S < 0. Given that ∆G = ∆H−T∆S,
a process with ∆S < 0 can only have ∆G < 0 if ∆H is sufficiently negative, that
is the process must be exothermic.
E19A.2(a) The collision flux, Z w , is given by [19A.1–777], Z w = P/(2πMkT/N A )1/2 where

p is the pressure of gas, M is the molar mass of the molecule, k is Boltzmann’s
constant, T is the temperature and N A is Avogadro’s constant. From inside the
front cover, 760 Torr = 1 atm = 1.01325 × 105 Pa, therefore 1 Torr is 133.32 Pa.
(i) For a hydrogen molecule, the molar mass M = 2 × (1.0079 g mol−1 ) =

690 19 PROCESSES AT SOLID SURFACES
2.0158 g mol−1 .
p
Zw =
(2πMkT/N A )1/2
(0.10 × 10−6 Torr)×(133.32 Pa Torr−1 )×(6.0221 × 1023 mol−1 )1/2
= 1/2
[2π×(2.0158 × 10−3 kg mol−1 )×(1.3806 × 10−23 J K−1 )×(298.15 K)]
= 1.43... × 1018 m−2 s−1 = 1.4 × 1014 cm−2 s−1
(ii) For propane, the molar mass M = 3×(12.011 g mol−1 )+8×(1.0079 g mol−1 ) =
44.096 g mol−1 .
p
Zw =
(2πMkT/N A )1/2
(0.10 × 10−6 Torr)×(133.32 Pa Torr−1 )×(6.0221 × 1023 mol−1 )1/2
= 1/2
[2π×(44.096 × 10−3 kg mol−1 )×(1.3806 × 10−23 J K−1 )×(298.15 K)]
= 3.06... × 1017 m−2 s−1 = 3.1 × 1013 cm−2 s−1
E19A.3(a) The collision flux, Z w , is given by [19A.1–777], Z w = p/(2πMkT/N A )1/2 where

p is the pressure of gas, M is the molar mass of the molecule, k is Boltzmann’s
constant, T is the temperature and N A is Avogadro’s constant.
The collision rate, z, is given by z = AZ w where A is the surface area. Hence,
Ap
z = AZ w =
(2πMkT/N A )1/2
For an argon atom, the molar mass M = 39.95 g mol−1 . Thus, for A = π(d/2)2 ,
where d is the diameter of the circular surface, rearranging the above expres-
sion gives
r(2πMkT/N A )1/2
p=
A
4.5 × 1020 s−1
=
π×(0.5 × 1.5 × 10−3 m)2
1/2
2π×(39.95 × 10−3 kg mol−1 )×(1.3806 × 10−23 J K−1 )×(425 K)
×( )
6.0221 × 1023 mol−1
= 1.25... × 104 Pa = 0.13 bar
E19A.4(a) For a perfect gas, and at constant temperature, p ∝ 1/V , where V is the volume
occupied by the gas at pressure p. Therefore
p 2 V1 V1 p 1
= hence V2 =
p 1 V2 p2
The surface coverage θ is given by θ = V /V∞ where V is the volume of gas

adsorbed at a particular pressure p and V∞ is the volume of gas which gives
a complete monolayer, but where the volume has been corrected to the same
pressure p.
At 5.0 bar, the volume adsorbed is a complete monolayer and thus V∞ = 22 cm3
at 5.0 bar. At 0.1 bar, this same volume is
(22 cm3 ) × (5.0 bar)
V∞,0.1 bar = = 1100 cm3
0.1 bar
Hence the surface coverage is
10 cm3
θ= = 9.1 × 10−3
1100 cm3
P19A.1 The Coulombic energy of interaction of the test ion with a section of lattice is
determined by summing the interaction energy of this ion with each of the ions
in the section of the lattice. Interactions between ions of opposite charge make
a negative contribution to the energy of −C/r, where C is a positive constant
and r is the distance between the test ion and an ion in the lattice. Similarly,
interactions between ions of the same charge make a positive contribution to
the energy of +C/r. Define a 0 as the distance between nearest neighbours in
the lattice, that is the lattice spacing.
(a) For a Type 2 section, and considering the nearest 10 ions only, the inter-
action energy with the test ion is
C 1 1 1 C 10 (−1)n C
E2 = × (−1 + − + ... + ) = ( ) ∑ = −0.646 ( )
a0 2 3 10 a 0 n=1 n a0
(b) For a Type 1 section of lattice, with 10 atoms in each direction, the inter-
action energy with the test ion is
C 10 10 (−1)n+m C
E1 = ( )∑ ∑ 2 = +0.259 ( )
a 0 n=1 m=1 (n + m )
2 1/2 a0
(c) To calculate the energy of interaction between the test ion and the lattice
in arrangement (a), observe that there is one Type 2 interaction and two
Type 1 interactions. Hence the interaction energy is given by
C C
E(a) = E 2 + 2E 1 = ( ) × (−0.646 + 2 × 0.259) = −0.128 ( )
a0 a0
To calculate the energy of interaction between the test ion and the lattice
in arrangement (b), observe that there are two Type 2 interactions and
three Type 1 interactions. Hence the interaction energy is given by
C C
E(b) = 2E 2 + 3E 1 = ( ) × [2(−0.646) + 3(0.259)] = −0.516 ( )
a0 a0
The energy of interaction of the probe cation is much lower for (b) than
for (a), therefore (b) is the more favourable arrangement .
P19A.3 From inside the front cover, 760 Torr = 1 atm = 1.01325 × 105 Pa, therefore
1 Torr is 133.32 Pa. The unit cell for a face-centred cubic lattice is shown in
Fig. 18A.8 on page 731 (cubic F) and how the planes are identified using Miller
indices is described in Section 18A.2(a) on page 731.
(a) The (100) plane is the face of the cube and the arrangement of the atoms
in the plane is shown in Fig. 19.1. There two atoms in this face, being the
total of one in the centre and a quarter of each of the four atoms at the
corners. Each atom has surface area πr 2 , where r is the atomic radius,
and the area of the face is a 2 = (352 × 10−12 pm)2 = 1.24 × 10−15 cm2 .
(100) surface
Figure 19.1
The surface number density n is the number of atoms divided by the area
of the face
2
n= = 1.61 × 1015 cm−2 = 1.61 × 1019 m−2
1.24 × 10−15 cm2
A = 1/n is therefore the area occupied by atoms within this face.
The collision flux, Z w , is given by [19A.1–777], Z w = p/(2πMkT/N A )1/2 .
For a hydrogen molecule, the molar mass M = 2 × (1.0079 g mol−1 ) =
2.0158 g mol−1 , so at T = 298.15 K and p = 100 Pa the frequency, f , of
molecular collisions with the atoms exposed on this face is
Ap
f = AZ w =
(2πMkT/N A )1/2
(1.61 × 1019 m−2 )−1 ×(100 Pa)×(6.0221 × 1023 mol−1 )1/2
= 1/2
[2π×(2.0158 × 10−3 kg mol−1 )×(1.3806 × 10−23 J K−1 )×(298.15 K)]
= 6.7 × 105 s−1
At p = 0.1 × 10−6 Torr, which is (0.1 × 10−6 Torr) × (133.32 Pa Torr−1 ) =

1.33... × 10−5 Pa
Ap
f = AZ w =
(2πMkT/N A )1/2
(1.61 × 1019 m−2 )−1 ×(1.33... × 10−5 Pa)×(6.0221 × 1023 mol−1 )1/2
= 1/2
[2π×(2.0158 × 10−3 kg mol−1 )×(1.3806 × 10−23 J K−1 )×(298.15 K)]
= 8.9 × 10−2 s−1
For propane, the molar mass M = (3 × 12.01 + 8 × 1.0079) g mol−1 =

44.09 g mol−1 . Similar calculations give the corresponding collision rates
as f = 1.42 × 105 s−1 at 100 Pa, and f = 1.9 × 10−2 s−1 at 0.10 µTorr.
(b) The (110) plane is a diagonal plane taken from corner to corner along one
face, and perpendicular to the face. It has the surface structure shown in
Fig. 19.2. There are again two atoms in this face.
(110) surface
r
√
2a
Figure 19.2
√
The area of the face is 2a 2 , therefore the surface number density n is
2 2
n= √ =√
2a 2 2 × (352 × 10−12 pm)2
= 1.14 × 1019 m−2 = 1.14 × 1015 cm−2
Similar calculations to those above give, for H2 , f = 9.4 × 105 s−1 and
f = 0.13 s−1 at p = 100 Pa and p = 0.10 µTorr, respectively. For propane
the rates are f = 2.0 × 105 s−1 and f = 2.7 × 10−2 s−1 .
(c) The (111) plane has the surface structure shown in Fig. 19.3; there is one
atom on this face.
(111) surface
2r
2r
Figure 19.3
√ √ 2
The area of this rhombus is 2r × 3r = 2 3r √ . Because the atoms touch
along the face diagonals it follows that
√ 2 √ 2 4r = 2a, hence r 2
= a 2 /8. Using
this the area of the face is 2 3r = 3a /4, and thus the surface number
density is
1 1
n= √ =√
3a 2 /4 3 × (352 × 10−12 pm)2 /4
= 1.86 × 1019 m−2 = 1.86 × 1015 cm−2
Similar calculations to those above give, for H2 , f = 5.8 × 105 s−1 and
f = 7.7 × 10−2 s−1 at p = 100 Pa and p = 0.10 µTorr, respectively. For
propane the rates are f = 1.2 × 105 s−1 and f = 1.6 × 10−2 s−1 .
19B Adsorption and desorption

D19B.1 The assumptions made in deriving the Langmuir isotherm are:
(1) Adsorption cannot proceed beyond monolayer coverage.

(2) All sites are equivalent and the surface is uniform.
(3) The ability of a molecule to adsorb at a given site is independent of the
occupation of neighbouring sites.
For the BET isotherm assumption (1) is removed so that multi-layer coverage
is possible. In the derivation of this isotherm a distinction is made between the
energetics involved in forming the first and subsequent layers.
E19B.1(a) The isosteric enthalpy of adsorption is define as [19B.5b–786]
∂ ln(α p−○ ) ∆ ad H −○
( ) =−
∂(1/T) θ R
From the Langmuir isotherm is follows that α = θ/p(1 − θ) but, because an

isosteric process is being considered (θ is constant), this reduces to α = C/p,
where C is a constant. With just two sets of data the derivative is approximated
as the finite interval to give
ln(C p−○ /p 2 ) − ln(C p−○ /p 1 ) ∆ ad H −○

=−
(1/T2 ) − (1/T1 ) R
∆ ad H −○ 1 1
hence ln(p 1 /p 2 ) = − ( − )
R T2 T1
R ln(p /p )
hence ∆ ad H −○ =
2 1
1/T2 − 1/T1
With the data given
(8.3145 J K−1 mol−1 ) × ln[(3.2 × 106 Pa)/(490 × 103 Pa)]

∆ ad H −○ =
1/(250 K) − 1/(190 K)
= −12.4 kJ mol−1
E19B.2(a) The rate constant for desorption is assumed to follow an Arrhenius law, k des =
Ae−E a,des /RT . Recall that for a first order process the half life is simply propor-
tional to the inverse of the rate constant, therefore the time needed for a certain
amount to desorb is also inversely proportional to the rate constant. Thus
τ 1 /τ 2 = e−(E a,des /R)(1/T2 −1/T1 )

E a,des 1 1
hence ln(τ 1 /τ 2 ) = − ( − )
R T2 T1
−R ln(τ 1 /τ 2 )
hence E a,des =
(1/T2 − 1/T1 )
With the data given
−(8.3145 J K−1 mol−1 ) ln[(2.0 min)/(27 min)]

E a,des =
1/(1978 K) − 1/(1856 K)
= 6.51... × 105 J mol−1 = 651 kJ mol−1
The times for desorption at different temperatures are computed using
τ 1 /τ 2 = e−(E a,des /R)(1/T2 −1/T1 ) hence τ 2 = τ 1 e(E a,des /R)(1/T2 −1/T1 )

The time needed at 298 K is related to that at 1856 K
J mol−1 )/(8.3145 J K−1 mol−1 )][1/(298 K)−1/(1856 K)]
τ 2 = (27 min) e[(6.51 ...×10
5
= 1.7 × 1097 min

Effectively, the gas does not desorb at this temperature. Repeating the calcula-
tion at 3000 K
J mol−1 )/(8.3145 J K−1 mol−1 )][1/(3000 K)−1/(1856 K)]
τ 2 = (27 min) e[(6.51 ...×10
5
= 2.8... × 10−6 min = 0.17 µs

At the higher temperature the gas leaves very rapidly indeed.
E19B.3(a) The average time that a species remains adsorbed is proportional to its half-life,
given by [19B.14–791], t 1/2 = τ 0 eE a,des /RT . Therefore, if the two times are τ 1 and
τ 2 at temperatures T1 and T2
τ 2 /τ 1 = e(E a,des /R)(1/T2 −1/T1 )
E a,des 1 1
hence ln(τ 2 /τ 1 ) = ( − )
R T2 T1
R ln(τ 2 /τ 1 )
hence E a,des =
1/T2 − 1/T1
With the data given

(8.3145 J K−1 mol−1 ) ln[(3.49 s)/(0.36 s)]
E a,des =
1/(2362 K) − 1/(2548 K)
= 611 kJ mol−1
E19B.4(a) The half-life for a species on the surface is given by [19B.14–791], t 1/2 = τ 0 eE a,des /RT .
(i) With E a,des = 15 kJ mol−1

J mol−1 )/[(8.3145 J K−1 mol−1 )×(400 K)]
at 400 K t 1/2 = (0.1 ps) e(15×10
3
= 9.1 ps
J mol−1 )/[(8.3145 J K−1 mol−1 )×(1000 K)]
at 1000 K t 1/2 = (0.1 ps) e(15×10
3
= 0.61 ps
(ii) With E a,des = 150 kJ mol−1

J mol−1 )/[(8.3145 J K−1 mol−1 )×(400 K)]
at 400 K t 1/2 = (0.1 ps) e(150×10
3
= 3.86... × 1018 ps = 3.9 × 106 s

J mol−1 )/[(8.3145 J K−1 mol−1 )×(1000 K)]
at 1000 K t 1/2 = (0.1 ps) e(150×10
3
= 6.83... × 106 ps = 6.8 µs
E19B.5(a) The Langmuir isotherm is [19B.2–785], θ = α p/(1 + α p), with α = k a /k d . The

surface coverage may be written in terms of the volume of gas adsorbed V ,
θ = V /V∞ , where V∞ is the volume corresponding to complete coverage. For
two different pressures
V1 α p1 V2 α p2
= =
V∞ 1 + α p 1 V∞ 1 + α p 2
Inverting both sides

V∞ 1 V∞ 1
= +1 = +1
V1 α p1 V2 α p2
To eliminate α first multiply the left-hand equation by 1/p 2 and the right-hand
equation by 1/p 1
V∞ 1 1 V∞ 1 1
= + = +
p 2 V1 α p 1 p 2 p 2 p 1 V2 α p 1 p 2 p 1
Subtracting the two equations gives then eliminates α

V∞ V∞ 1 1
− = −
p 2 V1 p 1 V2 p 2 p 1
1/p 2 − 1/p 1 p1 − p2
hence V∞ = =
1/p 2 V1 − 1/p 1 V2 p 1 /V1 − p 2 /V2
where for the last step top and bottom are multiplied by p 1 p 2 .
With the data given
p1 − p2
V∞ =
p 1 /V1 − p 2 /V2
(145.4 Torr) − (760 Torr)
= = 33.6 cm3
(145.4 Torr)/(0.286 cm3 ) − (760 Torr)/(1.443 cm3 )
E19B.6(a) The residence half-life is given by [19B.14–791], t 1/2 = τ 0 eE a,des /RT . The acti-
vation energy for desorption, E a,des , is approximated as minus the enthalpy of
adsorption.
J mol−1 )/[(8.3145 J K−1 mol−1 )×(400 K)]
t 1/2 = (1.0 × 10−14 s) e(120×10
3
= 47 s
E19B.7(a) The Langmuir isotherm is [19B.2–785], θ = α p/(1 + α p). The surface coverage
may be written in terms of the volume of gas adsorbed V , θ = V /V∞ , where
V∞ is the volume corresponding to complete coverage. Equivalently, θ may be
expressed in terms of the mass adsorbed, θ = m/m∞ , where m∞ is the mass
corresponding to complete coverage. For two different pressures
m1 α p1 m2 α p2
= =
m∞ 1 + α p 1 m∞ 1 + α p 2
The aim is to find m∞ , and the algebra to do this is just the same as the method
for finding V∞ in Exercise E19B.5(a) with volumes replaced by masses. The
result is
p1 − p2
m∞ =
p 1 /m 1 − p 2 /m 2
(26.0 kPa) − (3.0 kPa)
= = 0.531... mg
(26.0 kPa)/(0.44 mg) − (3.0 kPa)/(0.19 mg)
The surface coverage at the first pressure is therefore
m1 0.44 mg
θ1 = = = 0.83
m∞ 0.531... mg
At the second pressure θ 2 = (0.19 mg)/(0.531... mg) = 0.36 .
E19B.8(a) The Langmuir isotherm is [19B.2–785], θ = α p/(1 + α p), inverting both sides
gives
1 1 1 1 1−θ
= +1 hence = −1=
θ αp αp θ θ
Inverting again gives
θ θ
αp = hence p=
1−θ α(1 − θ)
With the data given

0.15
p= = 0.24 kPa
(0.75 kPa−1 ) × (1 − 0.15)
A similar calculation for θ = 0.95 gives 25 kPa .
E19B.9(a) The isosteric enthalpy of adsorption is define as [19B.5b–786]
∂ ln(α p−○ ) ∆ ad H −○
( ) =−
∂(1/T) θ R
From the Langmuir isotherm is follows that α = θ/p(1 − θ) but, because an

isosteric process is being considered (θ is constant), this reduces to α = C/p,
where C is a constant. With just two sets of data the derivative is approximated
as the finite interval to give
ln(C p−○ /p 2 ) − ln(C p−○ /p 1 ) ∆ ad H −○
=−
(1/T2 ) − (1/T1 ) R
∆ ad H −○ 1 1
hence ln p 1 /p 2 = − ( − )
R T2 T1
∆ ad H −○ 1 1
hence ln p 2 = ln p 1 + ( − )
R T2 T1
The data gives the enthalpy of desorption as +10.2 J for 1.00 mmol of gas,
therefore the molar enthalpy of adsorption is −10.2 kJ mol−1 .
−10.2 kJ mol−1 1 1
ln (p 2 /kPa) = ln(12 kPa) + −1 −1 ( − ) = 2.68...
8.3145 J K mol 313 K 298 K
Therefore (p 2 /kPa) = e2.68 ... , giving p 2 = 15 kPa .
P19B.1 (a) The Langmuir isotherm is [19B.2–785], θ = α p/(1 + α p), inverting both
sides gives 1/θ = 1/α p + 1. Figure 19.4 shows a plot of 1/θ against 1/p for
three different values of α.
(b) The Langmuir isotherm for adsorption with dissociation is [19B.4–785],
θ = (α p)1/2 /[1 + (α p)1/2 ]. Figure 19.5 shows a plot of 1/θ against 1/p for
the same values of α used in Fig. 19.4. In contrast to the straight lines
seen in Fig. 19.4, for the case of adsorption with dissociation the 1/θ
against 1/p plot shows pronounced curvature. Such a plot may there-
fore in principle make it possible to distinguish between dissociative and
non-dissociative adsorption.
(c) In [19B.7–788] the BET isotherm is manipulated into a straight-line plot
z 1 (c − 1) zVmon 1 (c − 1)
= + z or = + z
(1 − z)V cVmon cVmon (1 − z)V c c
100
α = 2 atm−1
80 α = 1 atm−1
α = 0.5 atm−1
60
1/θ
40
20
0
0 20 40 60 80 100
−1
(1/p)/atm
Figure 19.4
20
α = 2 atm−1
α = 1 atm−1
15
α = 0.5 atm−1
1/θ
10
0
0 20 40 60 80 100
(1/p)/atm−1
Figure 19.5
Thus a plot of zVmon /(1 − z)V against z is expected to be a straight line.

Figure 19.6 shows such a plot for three different values of c. Note that
when c ≫ 1, the slope becomes independent of c, and tends to 1.
P19B.3 In [19B.7–788] the BET isotherm is manipulated into a straight-line plot
z 1 (c − 1)
= + z z = p/p∗
(1 − z)V cVmon cVmon
Thus a plot of z/(1 − z)V against z is expected to be a straight line with slope
(c − 1)/cVmon and intercept 1/cVmon ; note that (slope)/(intercept) = c − 1. For
brevity the term z/(1 − z)V is denoted y.
(a) The data are shown in the table below and the plot is shown in Fig. 19.7.
10
c=2
c=4
8
c = 10
zVmon /(1 − z)V
6
0
0 2 4 6 8 10
z
Figure 19.6
p/kPa V /cm3 z y/(cm−3 )

14.0 11.1 0.0326 0.00303
37.6 13.5 0.0875 0.00711
65.6 14.9 0.153 0.0121
79.2 16.0 0.184 0.0141
82.7 15.5 0.193 0.0154
100.7 17.3 0.234 0.0177
106.4 16.5 0.248 0.0200
0.020
0.015
y/(cm−3 )
0.010
0.005
0.000
0.00 0.05 0.10 0.15 0.20 0.25
z
Figure 19.7
The data are a reasonable fit to a straight line with equation
y/(cm−3 ) = 0.07612 × (z) + 4.638 × 10−4
The parameter c is found using

slope 0.07612
= c−1 hence c =1+ = 165
intercept 4.638 × 10−4
The intercept is 1/cVmon , therefore

1 1
Vmon = = = 13.1 cm3
c × (intercept) 165 × (4.638 × 10−4 cm−3 )
(b) The data are shown in the table below and the plot is shown in Fig. 19.8.
p/kPa V /cm3 z y/(cm−3 )

5.3 9.2 0.0065 0.00071
8.4 9.8 0.0102 0.00106
14.4 10.3 0.0176 0.00174
29.2 11.3 0.0356 0.00327
62.1 12.9 0.0758 0.00635
74.0 13.1 0.0903 0.00758
80.1 13.4 0.0977 0.00808
102.0 14.1 0.1244 0.01008
0.010
y/(cm−3 )
0.005
0.000
0.00 0.02 0.04 0.06 0.08 0.10 0.12
z
Figure 19.8
The data are a reasonable fit to a straight line with equation
y/(cm−3 ) = 0.07953 × (z) + 3.036 × 10−4
The parameter c is found using
slope 0.07953
= c−1 hence c =1+ = 263
intercept 3.036 × 10−4
The intercept is 1/cVmon , therefore

1 1
Vmon = = = 12.5 cm3
c × (intercept) 263 × (3.036 × 10−4 cm−3 )
P19B.5 The Langmuir isotherm is [19B.2–785], θ = α p/(1+α p); the fractional coverage
can be expressed as n/n∞ , where n is the amount in moles covering the surface,
and n∞ is the amount corresponding to a monolayer. The same argument as

developed in Example 19B.1 on page 785 then applies, but with n instead of V .
A suitable plot to fit data to the Langmuir isotherm is therefore of p/n against
p; such a plot has intercept 1/αn∞ and slope 1/n∞ . The table of data is given
below and the plot is shown in Fig. 19.9.
p/kPa n/(mol kg−1 ) (p/n)/(kPa mol−1 kg)

31.00 1.00 31.00
38.22 1.17 32.67
53.03 1.54 34.44
76.38 2.04 37.44
101.97 2.49 40.95
130.47 2.90 44.99
165.06 3.22 51.26
182.41 3.30 55.28
205.75 3.35 61.42
219.91 3.36 65.45
(p/n)/(kPa mol−1 kg)
60
40
20
50 100 150 200
p/kPa
Figure 19.9
The data plainly fall on a curve, rather than the expected straight line. On the
assumption that the isotherm is more likely to hold at low pressure, the first six
data points are used to construct the line, the equation of which is
(p/n)/(kPa mol−1 kg) = 0.1368 × (p/kPa) + 27.08
The limiting coverage is n∞ = 1/slope = 1/(0.1368) mol kg−1 = 7.3 mol kg−1 .
The value of α is found from (slope)/(intercept), α = (0.1368)/(27.08) kPa−1 =

5.1 × 10−3 kPa−1 .
P19B.7 The isosteric enthalpy of adsorption is define as [19B.5b–786]
∂ ln(α p−○ ) ∆ ad H −○
( ) =−
∂(1/T) θ R
For the data in this Problem p−○ is replaced by c −○ . This equation implies that a
plot of ln(αc −○ ) against 1/T will have slope −∆ ad H −○ /R, for data at constant θ.
T/K (10−11 α)/(mol−1 dm3 ) (1/T)/(10−3 K−1 ) ln (αc −○ )

283 2.642 3.53 26.30
298 2.078 3.36 26.06
308 1.286 3.25 25.58
318 1.085 3.14 25.41
26.5
26.0
ln (αc −○ )
25.5
25.0
3.1 3.2 3.3 3.4 3.5 3.6
−3 −1
(1/T)/(10 K )
Figure 19.10
The data are a modest fit to a straight line, the equation of which is
ln (αc −○ ) = 2.42 × (1/T)/(10−3 K−1 ) + 17.81
The value of ∆ ad H −○ is obtained from the slope (the scatter means that high
precision on the result is not warranted)
∆ ad H −○ = −R × (slope) = −(8.3145 J K−1 mol−1 ) × 103 × (2.42 K−1 )

= −20 kJ mol−1
With the given data
∆ ad G −○ = ∆ ad H −○ − T −○ ∆ ad S −○
= (−20 kJ mol−1 ) − (300 K) × (+0.146 kJ K−1 mol−1 ) = −64 kJ mol−1
P19B.9 The Freundlich isotherm is given in [19B.11–789], θ = c 1 p1/c 2 . For the pur-
poses of analysing this data set the isotherm is rewritten by assuming that the
surface coverage is proportional to the mass adsorbed, and by replacing the
pressure by the concentration divided by the standard concentration, to give
w a = c 1 ([A]/c −○ )1/c 2 . The units of c 1 are adjusted accordingly. Taking loga-
rithms gives ln w a = ln c 1 + (1/c 2 ) ln([A]/c −○ ), implying that a plot of ln w a
against ln([A]/c −○ ) should be a straight line of slope 1/c 2 and intercept ln c 1 .
The data are given below and the plot is shown in Fig. 19.11.
[A]/c −○ w a /g ln([A]/c −○ ) ln(w a /g)

0.05 0.04 −3.00 −3.22
0.10 0.06 −2.30 −2.81
0.50 0.12 −0.69 −2.12
1.00 0.16 0.00 −1.83
1.50 0.19 0.41 −1.66
−1.5
−2.0
ln(w a /g)
−2.5
−3.0
−3.5
−3 −2 −1 0
−
○
ln([A]/c )
Figure 19.11
ln(w a /g) = 0.450 × ln([A]/c −○ ) − 1.83
The slope is 1/c 2 , therefore c 2 = 1/slope = 1/(0.450) = 2.22 . The intercept

gives ln c 1 and hence c 1 = 0.16 g .
P19B.11 The Langmuir isotherm is [19B.2–785], θ = α p/(1+α p); the fractional coverage
can be expressed as s/s∞ , where s is the amount in moles covering the surface
(per g of charcoal), and s∞ is the amount corresponding to a monolayer. For
adsorption from solution the pressure is replaced by the concentration c. The
same argument as developed in Example 19B.1 on page 785 then applies, but
with s instead of V , and c instead of p. A suitable plot to fit data to the Langmuir
isotherm is therefore of c/s against c; such a plot has intercept 1/αs∞ and slope
1/s∞ .
The table of data is shown below and the plot is shown in Fig. 19.12; for brevity
the units of c and s are omitted throughout. It is clear that the data fall on a
curve and so do not conform to the Langmuir isotherm.
c s c/s
15.0 0.60 25.0
23.0 0.75 30.7
42.0 1.05 40.0
84.0 1.50 56.0
165 2.15 76.7
390 3.50 111.4
800 5.10 156.9
150
100
c/s
50
0
0 200 400 600 800
c
Figure 19.12
The Freundlich isotherm [19B.11–789] is written s = Kc 1/n ; taking logarithms

gives ln s = ln K + (1/n) ln c. This implies that a plot of ln s against ln c should
be a straight line of slope 1/n and intercept ln K. The data are given below and
the plot is shown in Fig. 19.13.
c s ln c ln s
15.0 0.60 2.71 −0.511
23.0 0.75 3.14 −0.288
42.0 1.05 3.74 0.049
84.0 1.50 4.43 0.405
165 2.15 5.11 0.765
390 3.50 5.97 1.253
800 5.10 6.68 1.629

ln s = 0.539 × ln c − 1.975
ln s
3 4 5 6
ln c
Figure 19.13
The slope is 1/n, therefore n = 1/slope = 1/(0.539) = 1.9 . The intercept gives
ln K and hence K = 0.14 mmol acetone/g charcoal .
The Temkin isotherm [19B.10–789] is written in this case as s = K ln(nc).
There is no straight-line plot for testing the data against this isotherm, but
mathematical software can be used to find the best-fit parameters K and n.
The result of such a fit is s = 1.083 ln(0.0738 c). The table below compares the
values of s predicted by this relationship and those predicted by the Freundlich
isotherm (using the best-fit parameters found above). It is evident that the Fre-
undlich isotherm reproduces the data far more precisely than does that Temkin
isotherm.
c s sTemkin sFreundlich
15.0 0.60 0.11 0.60
23.0 0.75 0.57 0.75
42.0 1.05 1.23 1.04
84.0 1.50 1.98 1.51
165.0 2.15 2.71 2.18
390.0 3.50 3.64 3.46
800.0 5.10 4.42 5.09
19C Heterogeneous catalysis

D19C.1 The two relevant equations derived in How is that done? 19C.1 on page 794 are
(α A p A + 1)θ A + α A p A θ B = α A p A α B p B θ A + (α B p B + 1)θ B = α B p B
The first is multiplied by α B p B and the second by (α A p A + 1) to give

α B p B (α A p A + 1)θ A + α B p B α A p A θ B = α B p B α A p A
(α A p A + 1)α B p B θ A + (α A p A + 1)(α B p B + 1)θ B = (α A p A + 1)α B p B
The terms in θ A are now the same, and so will disappear when the two equations
are subtracted to give
α B p B α A p A θ B − (α A p A + 1)(α B p B + 1)θ B = α B p B α A p A − (α A p A + 1)α B p B

The expression for θ B then follows with some rearrangement and simplification
α B p B α A p A − (α A p A + 1)α B p B
θB =
α B p B α A p A − (α A p A + 1)(α B p B + 1)
−α B p B
=
αB pB αA pA − αA pA αB pB − αA pA − αB pB − 1
αB pB
=
αA pA + αB pB + 1
which is the required expression. That for θ B is simply found by swapping the
indices A and B: the equations all remain valid under such a change.
E19C.1(a) The amount in moles of N2 gas is found using the perfect gas law; close attention
to the units is needed.
pV (760 Torr)×(1.01325 × 105 Pa) 3.86 × 10−6 m3
n= = ×
RT 760 Torr (8.3145 J K−1 mol−1 ) × (273.15 K)
= 1.72... × 10−4 mol
which corresponds to N A n = (6.0221 × 1023 mol−1 ) × (1.72 × 10−4 mol) =

1.03... × 1020 molecules.
A rough calculation of the surface area notes that the collision cross section
is σ = πd 2 , where d is the diameter of the colliding spheres. Therefore d =
(σ/π)1/2 , and hence r = 21 (σ/π)1/2 . The area of one molecule is πr 2 = π 14 σ/π =
1
4
σ. The surface area is therefore (1.03... × 1020 ) × 14 × (0.43 × 10−18 m2 ) =
11 m2 .
In fact circles do not cover a plane completely, and it can be shown that the
highest coverage which can be achieved is one in which the circles cover 0.91
of the area of the plane. The estimate of the area therefore needs to be scaled
up by a factor of 1/0.91 ≈ 1.1 to give 12 m2 .
P19C.1 (a) The Langmuir–Hinshelwood rate law is given in [19C.2b–794]
kr αA αB pA pB
υ=
(1 + α A p A + α B p B )2
(b) When the partial pressures of the reactants are low, α A p A ≪ 1 and α B p B ≪
1, and therefore these terms may be ignored in the denominator to give
υlow = k r α A α B p A p B
The order with respect to A and B is now 1, and the overall order is 2.
(c) If, compared to B, A is strongly adsorbed or is present at high pressure,
then α A p A ≫ 1 and α A p A ≫ α B p B . The denominator simplifies to
(α A p A )2 and
kr αA αB pA pB kr αB pB
υhigh A = =
(α A p A )2 αA pA
In this limit the rate law is −1 order in A, first order in B, and therefore
overall zeroth order. It does not appear to be possible to achieve zeroth
order for either A or B alone.
P19C.3 (a) The Langmuir isotherm is [19B.2–785], θ = α p/(1 + α p). The fraction of
uncovered sites, θ u , is θ u = 1 − θ
αp 1 + αp − αp
θu = 1 − =
1 + αp 1 + αp
1 1
= ≈
1 + αp αp
where the approximation holds in the limit α p ≫ 1, that is strong adsorp-

tion of the gas.
(b) If a gas is weakly adsorbed, meaning α p ≪ 1, the fractional coverage is
θ ≈ α p.
If hydrogen is strongly adsorbed, the fraction of uncovered surface sites
goes as p−1
H 2 : these are the sites available for ammonia to bind to and to
react, therefore the rate is expected to go as p−1
H 2 . The ammonia only binds
weakly, so its surface coverage, and hence the rate of reaction, goes as
p NH3 . Overall, the rate law is expected to go as p NH3 /p H2 .
(c) The stoichiometric equation is NH3 ÐÐ→ 21 N2 + 32 H2 . Assume that the
initial pressure is p 0 and due solely to NH3 . After some time, suppose
that the partial pressure of NH3 has fallen to p NH3 = p 0 − δ; the partial
pressure of H2 is then 32 δ. It follows that δ = p 0 − p NH3 and hence p H2 =
3
2
δ = 32 (p 0 − p NH3 ). The rate law can therefore be written in terms of p 0
and p NH3
dp NH3 p NH3 p NH3

= −k c 3 = − 23 k c
dt 2
(p 0 − p NH3 ) p 0 − p NH3
The differential equation is separable and can be integrated in a straight-

forward way; to simplify the notation p NH3 is written as p

p p0 − p t
∫ dp = − 32 k c ∫ dt
p0 p 0
p p0 t
∫ [ − 1] dp = − 23 k c ∫ dt
p0 p 0
p
[p 0 ln p − p]∣ p 0 = − 23 k c t
p 0 ln(p/p 0 ) − (p − p 0 ) = − 32 k c t
(d) A plot of y = p 0 ln(p/p 0 ) − (p − p 0 ) against t is expected to be a straight

line. The table of data is shown below and the plot is shown in Fig. 19.14.
t/s p/kPa y/kPa
0 13.3 0.000
30 11.7 −0.105
60 11.2 −0.186
100 10.7 −0.293
160 10.3 −0.400
200 9.9 −0.527
250 9.6 −0.636
0.0
−0.2
y/kPa
−0.4
−0.6
0 50 100 150 200 250
t/s
Figure 19.14
The data fall on a reasonable straight line with equation

y/kPa = −2.474 × 10−3 × (t/s) − 0.0237
A simple statistical analysis suggests an error of about 8 × 10−5 on the
slope and 0.01 on the intercept. The intercept should be zero, and given
the estimated errors the data are pretty much consistent with this.
From the slope it follows that k c = (− 23 ) × (−2.474 × 10−3 kPa s−1 ) hence
k c = 3.7 × 10−3 kPa s−1 .
19D Processes at electrodes

D19D.1 This is discussed in Section 19D.3 on page 802.
E19D.1(a) (i) The Butler–Volmer equation is [19D.2–800], j = j 0 (e(1−α) f η −e−α f η ). For
H+ on Ni j 0 = 6.3×10−6 A cm−2 and α = 0.58; at 298.15 K, f = 38.921 V−1 .
For an overpotential of +0.20 V the current density is
j = (6.3 × 10−6 A cm−2 )

−1 −1
× (e(1−0.58)×(38.921 V )×(0.20 V)
− e−0.58×(38.921 V )×(0.20 V)
)
= 1.7 × 10−4 A cm−2
(ii) If the current is entirely anodic, only the first term is needed
−1
j = (6.3 × 10−6 A cm−2 ) × e(1−0.58)×(38.921 V )×(0.20 V)
= 1.7 × 10−4 A cm−2
The result confirms that the current is indeed dominated by the anodic
term, which is the term for which the power of the exponential is positive.
E19D.2(a) The Butler–Volmer equation is [19D.2–800], j = j 0 (e(1−α) f η −e−α f η ); at 298.15 K,

f = 38.921 V−1 .
(i) With the given data and for an η = +0.010 V the current density is
j = (0.79 mA cm−2 )
−1 −1
× (e(1−0.5)×(38.921 V )×(0.010 V)
− e−0.5×(38.921 V )×(0.010 V)
)
= 0.31 mA cm−2
(ii) For η = +0.100 V the current density is
j = (0.79 mA cm−2 )
−1 −1
× (e(1−0.5)×(38.921 V )×(0.100 V)
− e−0.5×(38.921 V )×(0.100 V)
)
= 5.4 mA cm−2
(iii) For η = −5.0 V the current density is
j = (0.79 mA cm−2 )
−1 −1
× (e(1−0.5)×(38.921 V )×(−5.0 V)
− e−0.5×(38.921 V )×(−5.0 V)
)
= −1.4 × 1042 mA cm−2
Such a current density would be quite impossible to achieve in practice.

E19D.3(a) At equilibrium, only the exchange current flows, therefore for an electrode with
area A the current is j 0 A, and thus the charge passing in time t is (current ×
time): q = j 0 At. If each species passing through the double layer carries one
fundamental change, the number of charges is N = q/e = j 0 At/e. Thus the
number per second through an area of 1.0 cm2 is, for H+ /Pt,
N/t = j 0 A/e = (7.9 × 10−4 A cm−2 ) × (1 cm2 )/(1.6022 × 10−19 C)

= 4.93... × 1015 s−1 = 4.9 × 1015 s−1
A similar calculation for Fe3+ /Pt gives 1.6 × 1016 s−1 , and for H+ /Pb the result
is 3.1 × 107 s−1 .
The number of atoms covering 1 cm2 of electrode is (10−4 m2 )/(280×10−12 m)2
= 1.27... × 1015 . Therefore for H+ /Pt the number of times per second that each
atom is involved in a electron transfer event is (number of such events)/(number
of atoms) = (4.93... × 1015 s−1 )/(1.27... × 1015 ) = 3.9 s−1 . Similar calculations
for Fe3+ /Pt and H+ /Pb give 12 s−1 and 2.4 × 10−8 s−1 , respectively. For H+ /Pb
the time between events is more than 1 year.
E19D.4(a) In the linear region the current density and overpotential are related by [19D.4–
801], η = RT j/F j 0 , therefore the current density is j = ηF j 0 /RT. For an
electrode of area A the current is I = jA, and therefore the resistance is
η η RT
r= = =
I ηF j 0 A/RT F j 0 A
For H+ /Pt
(8.3145 J K−1 mol−1 ) × (298 K)

r= = 33 Ω
(96485 C mol−1 ) × (7.9 × 10−4 A cm−2 ) × (1.0 cm2 )
The units are resolved by using (from inside the front cover) 1 V = 1 J C−1 and
1 Ω = 1 V A−1 . A similar calculation for H+ /Hg gives 3.3 × 1010 Ω .
E19D.5(a) Because the standard potential of Zn2+ /Zn is −0.76 V, under standard condi-
tions Zn metal will only be deposited when the applied potential is more nega-
tive than −0.76 V. The current density is given by [19D.2–800], j = j 0 (e(1−α) f η −
e−α f η ), but under these conditions only the second term (the cathodic current)
is significant. Using the data given for H+ , assuming α = 0.5, and recalling that,
at 298.15 K, f = 38.921 V−1
j H+ = − j 0 e−α f η
−1
= −(50 × 10−12 A cm−2 ) e−0.5×(38.921 V )×(−0.76 V)
= −1.3 × 10−4 A cm−2
It is usually considered that the metal can be deposited if the current density
for discharge of H+ is less than about 1 mA cm−2 , which is satisfied in this case,
but not by a large margin. The expectation is that zinc metal will be deposited,
but accompanied by significant evolution of H2 due to discharge of H+ .
E19D.6(a) If the anodic process is dominant, the current density is given by [19D.5a–802],
ln j = ln j 0 + (1 − α) f η, where f = F/RT. At 298.15 K
f = (96485 C mol−1 )/[(8.3145 J K−1 mol−1 ) × (298.15 K)] = 38.921 V−1
where the units are resolved by recalling 1 V = 1 J C−1 . Taking the difference of
two expressions for ln j for different overpotentials gives
ln( j 2 / j 1 ) = (1 − α) f (η 2 − η 1 )
ln( j 2 / j 1 )
hence η 2 = + η1
(1 − α) f
ln(75/55.0)
= + 0.125 V = 0.14 V
(1 − 0.39) × (38.921 V−1 )
j = j 0 e(1−α) f η , where f = F/RT. At 298.15 K, f = 38.921 V−1 . Rearranging for
j 0 and then using the data given
j 0 = j e−(1−α) f η
−1
= (55.0 mA cm−2 ) e−(1−0.39)×(38.921 V )×(0.125 V)
= 2.8 mA cm−2
j = j 0 e(1−α) f η , where f = F/RT. At 298.15 K, f = 38.921 V−1 . Taking the ratio
of two expressions for j for different overpotentials gives
j 2 / j 1 = j 0 e(1−α) f η 2 / j 0 e(1−α) f η 1
hence j 2 = j 1 e(1−α) f (η 2 −η 1 )
−1
= (1.0 mA cm−2 ) e(1−0.5)×(38.921 V )×[(0.60−0.40) V)]
= 49 mA cm−2
The current density increases dramatically with this increase in overpotential.
P19D.1 (a) The current density is given by [19D.2–800], j = j 0 (e(1−α) f η − e−α f η ),
but for positive η the second term (the anodic current) dominates and
therefore ln j = ln j 0 + (1 − α) f η. A plot of ln j against η will have slope
(1 − α) f and intercept ln j 0 . Such a plot is shown in Fig. 19.15.
η/V j/(mA cm−2 ) ln[ j/(mA cm−2 )]

0.050 2.66 0.978
0.100 8.91 2.19
0.150 29.9 3.40
0.200 100 4.61
0.250 335 5.81
ln[ j/(mA cm−2 )]

4
0
0.00 0.05 0.10 0.15 0.20 0.25
η/V
Figure 19.15
The data fall on a good straight line with equation
ln[ j/(mA cm−2 )] = 24.18 × (η/V) − 0.230
From the slope it follows that (1 − α) × (38.921 V−1 ) = 24.18 V−1 hence
α = 0.38 . The exchange current density is computed from the intercept
as j 0 = 0.79 mA cm−2 .
(b) For negative overpotentials the cathodic current dominates and j = − j 0 e−α f η .
The following table is drawn up using the results from (a).
η/V j/(mA cm−2 )

−0.050 −1.65
−0.100 −3.47
−0.150 −7.26
−0.200 −15.2
−0.250 −31.9
P19D.3 (a) The Nernst equation [6C.4–207] for the half cell is
RT
E(Fe2+ /Fe) = E(Fe2+ /Fe)−○ + ln a Fe2+
2F
Therefore with the given concentration, the potential is
(8.3145 J K−1 mol−1 ) × (298 K)

E(Fe2+ /Fe) = (−0.44 V) +
2 × (96485 C mol−1 )
1.70 × 10−6 mol dm−3
× ln = −0.611 V
1 mol dm−3
The overpotential is thus computed as η = E ′ − E = E ′ − (−0.611 V) =
E ′ + (0.611 V).
(b) The current density is the rate of deposition in moles, multiplied by the
Faraday constant (to give the charge) and divided by the area of the elec-
trode: j = 2υF/A; the factor of two is needed as a divalent ion is being
discharged. For the first data point
j = 2 × (1.47 × 10−12 mol s−1 ) × (96485 C mol−1 )/(9.1 cm2 )
= 3.11... × 10−8 A cm−2 = 31.1... nA cm−2
The current density is given by [19D.2–800], and can be separated into an
anodic and cathodic part: j = j a + j c = j 0 (e(1−α) f η − e−α f η ). Thus
j = j 0 e−α f η (e f η − 1) = − j c (e f η − 1)
j
hence j c =
1 − ef η
For the first data point
31.1... nA cm−2
jc = = 32.1... nA cm−2
1 − e(38.921 V−1 )×[(−0.702+0.611) V]
The remaining values are given in the table in part (c).
(c) The cathodic current density is ∣ j c ∣ = j 0 e−α f η , therefore a plot of ln ∣ j c ∣
against η should have slope −α f and intercept ln j 0 . The data are tabulated
below and such a plot is shown in Fig. 19.16.
E ′ /V η/V υ/(pmol s−1 ) j/(nA cm−2 ) ∣ j c ∣/(nA cm−2 ) ln[∣ j c ∣/(nA cm−2 )]
−0.702 −0.091 1.47 31.2 32.1 3.47
−0.727 −0.116 2.18 46.2 46.7 3.84
−0.752 −0.141 3.11 65.9 66.2 4.19
−0.812 −0.201 7.26 154 154 5.04
5.0
ln[∣ j c ∣/(nA cm−2 )]
4.5
4.0
3.5
−0.22 −0.20 −0.18 −0.16 −0.14 −0.12 −0.10 −0.08
η/V
Figure 19.16

ln[∣ j c ∣/(nA cm−2 )] = −14.20 × (η/V) + 2.187
From the slope it follows that −α × (38.921 V−1 ) = −14.20 V−1 hence
α = 0.365 . The exchange current density is computed from the intercept
as j 0 = 8.91 nA cm−2 .
P19D.5 The data given correspond to positive overpotentials, so the anodic current will
dominate and hence ln j = ln j 0 + (1 − α) f η. A plot of ln j against η will have
slope (1 − α) f and intercept ln j 0 . Such a plot is shown in Fig. 19.17.
η/V j/(mA m−2 ) ln[ j/(mA m−2 )]

0.60 2.9 1.06
0.65 6.3 1.84
0.73 28 3.33
0.79 100 4.61
0.84 250 5.52
0.89 630 6.45
0.93 1 650 7.41
0.96 3 300 8.10
8
ln[ j/(mA m−2 )]
0.6 0.7 0.8 0.9 1.0

η/V
Figure 19.17

ln[ j/(mA m−2 )] = 19.55 × (η/V) − 10.83
From the slope it follows that (1−α)×(38.921 V−1 ) = 19.55 V−1 hence α = 0.50 .
The exchange current density is j 0 = 1.99 × 10−5 mA m−2 (computed from the
intercept).
P19D.7 Because the overpotential is always positive and ‘high’, the current is entirely
anodic and given by j = j 0 e(1−α) f η . Imagine that the potential ramps linearly
from η− to η+ in a time τ: over this period the time-dependent overpotential
is η(t) = η− + (η+ − η− )t/τ. For this rising part of the ramp the current is
j t=0→τ = j 0 e(1−α) f [η− +(η+ −η− )t/τ] = j 0 e(1−α) f η− e(1−α) f (η+ −η− )t/τ
This is a current which rises from j 0 e(1−α) f η− at t = 0 to j 0 e(1−α) f η+ at t = τ

with an exponential dependence on time.
For the falling part of the ramp the time-dependent overpotential is η(t) =
(2η+ − η− ) + (η− − η+ )t/τ, giving a current
j t=τ→2τ = j 0 e(1−α) f [(2η+ −η− )+(η− −η+ )t/τ] = j 0 e(1−α) f (2η+ −η− ) e(1−α) f (η− −η+ )t/τ
This is a current which falls from j 0 e(1−α) f η+ at t = τ to j 0 e(1−α) f η− at t = 2τ

with an exponential dependence on time. After 2τ the system is back at its
starting position.

I19.1 The Coulombic potential between two charges Q 1 and Q 2 at a distance r is
Q1 Q2
V=
4πε 0 r
The force is given by F = −dV /dr
d Q1 Q2 Q1 Q2
F=− =
dr 4πε 0 r 4πε 0 r 2
(1.6022 × 10−19 C)2
=
4π × (8.8542 × 10−12 J−1 C2 m−1 ) × (2.00 × 10−9 m)2
= 57.7 pN
I19.3 The approach is to compute the standard reaction Gibbs energy of the com-
bustion reaction using tabulated standard Gibbs energies of formation. The
standard cell potential is then computed using ∆ r G −○ = −νFE −○ ; the number
of electrons involved in the reaction is identified by considering the oxidation
numbers of the products and reactants. For brevity the phases of the species
are omitted from the chemical equations.
(a) The reaction is H2 + 12 O2 → H2 O. The oxygen goes from oxidation

number 0 in O2 , to −2 in H2 O, that is a change of 2. Because there is just
one oxygen atom involved, ν = 1 × 2 = 2. For this reaction ∆ r G −○ is equal
to ∆ f G −○ (H2 O), which is −237.13 kJ mol−1 . Hence E −○ = −∆ r G −○ /νF =
−(−237.13 × 103 J mol−1 )/[(2) × (96485 C mol−1 )] = +1.23 V .
(b) The reaction is CH4 + 2 O2 ÐÐ→ CO2 + 2 H2 O. The oxygen goes from
oxidation number 0 in O2 , to −2 in both CO2 and H2 O, that is a change
of 2. Because there are 4 oxygen atoms in total, ν = 4 × 2 = 8.
∆ r G −○ = ∆ f G −○ (CO2 ) + 2∆ f G −○ (H2 O) − ∆ f G −○ (CH4 )

(∆ r G −○ /kJ mol−1 ) = (−394.36) + 2 × (−237.13) − (−50.72) = −817.9
E −○ = −∆ r G −○ /νF
= −(−817.9 × 103 J mol−1 )/[8 × (96485 C mol−1 )]
= +1.06 V
(c) The reaction is C3 H8 + 5 O2 ÐÐ→ 3 CO2 + 4 H2 O. The oxygen goes from

oxidation number 0 in O2 , to −2 in both CO2 and H2 O, that is a change
of 2. Because there are 10 oxygen atoms in total, ν = 10 × 2 = 20.
∆ r G −○ = 3∆ f G −○ (CO2 ) + 4∆ f G −○ (H2 O) − ∆ f G −○ (C3 H8 )

(∆ r G −○ /kJ mol−1 ) = 3 × (−394.36) + 4 × (−237.13) − (−23.49)
= −2108.11
E = −∆ r G −○ /νF
−
○
= −(−2108.11 × 103 J mol−1 )/[20 × (96485 C mol−1 )]

= +1.09 V
I19.5 The relativistic correction term is

e∆ϕ (1.6022 × 10−19 C) × (50 × 103 V)
= = 0.0489...
2m e c 2 2 × (9.1094 × 10−31 kg) × (2.9979 × 108 m s−1 )2
The non-relativistic wavelength is
h
λnon-rel =
(2m e e∆ϕ)1/2
6.6261 × 10−34 J s
=
[2×(9.1094 × 10−31 kg)×(1.6022 × 10−19 C)×(50 × 103 V)]1/2
= 5.48 pm
With the relativistic correction

6.6261 × 10−34 J s
λrel =
[2×(9.1094 × 10−31 kg)×(1.6022 × 10−19 C)×(50 × 103 V)]1/2
1
× = 5.36 pm
(1 + 0.0489...)1/2
There is a significant difference of about 2%. Whether or not this will affect
any particular measurement depends on whether it is necessary to know the
wavelength to high precision.

Pchem Student Answers PDF

Uploaded by

Copyright:

Available Formats

Pchem Student Answers PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pchem Student Answers PDF

Uploaded by

Copyright:

Available Formats

Student Solutions

1 The properties of gases 1

3 The second and third laws 69

4 Physical transformations of pure substances 113

5 Simple mixtures 131

6 Chemical equilibrium 193

6B The response of equilibria to the conditions 202

7 Quantum theory 237

8 Atomic structure and spectra 291

9 Molecular Structure 311

10 Molecular symmetry 345

11 Molecular Spectroscopy 377

12 Statistical thermodynamics 439

13 Molecules in motion 487

14 Chemical kinetics 509

15 Reaction dynamics 569

16 Magnetic resonance 599

17 Molecular Interactions 625

17B Interactions between molecules 637

19 Processes at solid surfaces 689

Conventions used is presenting the solutions

Errors and omissions

1A The perfect gas

pV (0.987 × 105 Pa) × (1 m3 )

(i) The total amount in moles is n = n O2 + n N2 . The total mass m is computed

The mole fractions are therefore

The partial pressures are given by p i = x i p tot

The relationships 1 J = 1 kg m2 s−2 and 1 Pa = 1 kg m−1 s−2 have been used.

E1A.3(a) Charles’ law [1A.3b–7] states that V ∝ T at constant n and p, and p ∝ T at

θ/○ C ρ/(g dm−3 ) (1/ρ)/(g−1 dm3 )

(1/ρ)/(g−1 dm3 ) = 2.835 × 10−3 × (θ/○ C) + 0.7734

The intercept with 1/ρ = 0 is found by solving

0 = 2.835 × 10−3 × (θ/○ C) + 0.7734

This gives θ = −273 ○ C as the estimate of absolute zero.

E1A.4(a) (i) The mole fractions are

by the perfect gas law: p tot = n tot RT/V

2 (3.0 mol) × (8.3145 J K−1 mol−1 ) × (273.15 K)

(3.0 mol) × (8.3145 J K−1 mol−1 ) × (273.15 K)

E1A.7(a) Because the temperature is constant (isothermal) Boyle’s law applies, pV =

Note the choice of R to match the units of the problem. An alternative is to

[(255 × 10−3 g)/(20.18 g mol−1 )] × (8.3145 J K−1 mol−1 ) × (122 K)

ρRT (3.710 kg m−3 ) × (8.3145 J K−1 mol−1 ) × ([500 + 273.15] K)

p = 1.01325 × 105 Pa + (0.9971 × 103 kg m−3 ) × (9.806 m s−2 )

The value of R is found by rearranging the perfect gas law to R = pV /nT

pV (1.192... × 105 Pa) × (20.000 × 10−3 m3 )

p/atm Vm /(dm3 mol−1 ) (pVm /T)/(atm dm3 mol−1 K−1 )

0.0 0.2 0.4 0.6 0.8

The data fall on a reasonable straight line, the equation of which is

(pVm /T)/(atm dm3 mol−1 K−1 ) = −7.995 × 10−5 × (p/atm) + 0.082062

p/kPa ρ/(kg m−3 ) (p/ρ)/(kPa kg−1 m3 )

The data fall on a reasonable straight line, the equation of which is

(p/ρ)/(kPa kg−1 m3 ) = 0.04610 × (p/kPa) + 53.96

The intercept is (p/ρ)lim p→0 , which is equal to RT/M.

RT (8.3145 J K−1 mol−1 ) × (298.15 K)

The estimate of the molar mass is therefore 45.94 g mol−1 .

p 1 ∆T (6.69 × 103 Pa) × (1.00 K)

(b) A temperature of 100.00 ○ C is equivalent to an increase in temperature

The final pressure is therefore 6.69 + 2.44... = 9.14 kPa .

The exponential of each side is taken to give