Effect of concentration on the rate of chemical reactions

C. A. Ward

Citation: The Journal of Chemical Physics 79, 5605 (1983); doi: 10.1063/1.445681
View online: http://dx.doi.org/10.1063/1.445681
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/79/11?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in

Concentration effects on the rates of irreversible diffusion-influenced reactions
J. Chem. Phys. 141, 084101 (2014); 10.1063/1.4893340

Concentration fluctuations in chemical reactions

J. Chem. Phys. 63, 5037 (1975); 10.1063/1.431204

The effect of reagent energy on chemical reaction rates: An information theoretic analysis
J. Chem. Phys. 63, 4280 (1975); 10.1063/1.431198

Effects of Chemical Reactions on the Bimolecular Rate Constant

J. Chem. Phys. 46, 1316 (1967); 10.1063/1.1840850

The Absolute Rate of a Chemical Reaction

J. Chem. Phys. 3, 242 (1935); 10.1063/1.1749644

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 Effect of concentration on the rate of chemical reactions
C. A. Warda)
Laboratoire d'Electrochimie Interfaciale du C.N.R.S., 92195 Meudon Principal Cedex, France
(Received 24 May 1983; accepted 16 August 1983)

A method that has recently been used to examine the rate of molecular transport across a phase
boundary-statistical rate theory- is extended to obtain the expression for the rate of a single step chemical
reaction. The concentration dependence of the newly derived rate expression leads to the prediction of a
higher reaction rate in the period immediately after the reaction has begun than any of the previous theories
and gives a different physical interpretation of the "rate constant". When ions are reacting in solution under
near-equilibrium conditions, the concentration dependence of the newly derived rate expression reduces to the
conventional expression; further under these conditions, the new fate expression leaps to an expression for the
conventional rate constant that is equivalent to the Brtlnsted primary salt effect for the change in the rate
constant with ionic strength. To eltamine the predicted behavior in the initial period, the predictions are
compared with the measurements of ion concentration during electron exchange reactions. When these
experiments were originally conducted, using radioactive labeling techniques, substantial disagreement was
found between the conventional theory and the measurements. To bring them together the concept of "zero-
time exchange" was introduced which amounts to adding a constant to the predictions of the conventional
theory. The physical basis for this concept has remained undefined. When the statistical rate theory
expression is compared with the data from six experiments available in the literature, it is found for five of the
six that there is no measurable disagreement between the data and the theory and only small disagreement for
the other case. From the point of view of statistical rate theory it is unnecessary to introduce the concept of
zero-time exchange.

I. INTRODUCTION If A and B of Eq. (1) represent ions in solution that

are exchanging charge, the usual approach to obtaining
Suppose it is known that the following reaction occurs
the expreSSion for the rate has been through the absolute
as a single step reaction:
rate theory (Refs. 5 and 6 give reviews of these theo-
(1) ries). Thus all of these approaches lead to an expres-
and it is proposed to predict the net rate at which the sion for the concentration dependence of the reaction
reaction proceeds. To obtain the concentration depen- rate as given in Eq. (2), but with different approaches
dence of the expression for the forward rate of the for calculating the rate constant k f • 5,6
chemical reaction J f , a number of theories have adopted A related approach is the reaction-elementary-act
the concept of the reaction proceeding through an acti- theory aSSOCiated primarily with Dogonadze and Le-
vated complex and assumed equilibrium to exist be- vich. 7,8 Although the approach is quantum mechanical
tween the activated complexes and the reactants. This in nature, they abandon the formalism of quantum me-
assumption waS introduced by Eyring in 1935 and it chanics and define the rate to be of the form [see Ref.
leads to following expression for the forward rate of 8, Eq. (2.1))
the chemical reaction!:
Jf= kfCAC B , (2) Jf=CAC B i""
TO
W(R)P(R)dR , (3)

where C; is the concentration of species i and k f is the where P(R) is the probability that ions approach to a
rate constant in the forward direction. distance R and W(R) is the probability (per unit time)
The assumption of equilibrium between the activated that an electron performs the tranSition between A and
complexes and the reactants has been questioned be- B when they are separated by the distance R. Note
fore, 2,3 but the approach to examining this assumption that by proceeding in this fashion they have hypothesized
has been through the Boltzmann equation and this equa- the concentration dependence of the rate (rather than
tion contains the assumption of molecular chaos 4 (which deriv ing it from quantum mechanics) and the hypothes ized
amounts to assuming that there is no correlation between dependence is the same as that of absolute rate theory.
reacting molecular pairs). Such an assumption is also Thus all of the previous theories for a chemical reaction
questionable when a reaction is occurring at a high rate. of the type shown in Eq. (1) would predict the same Con-
If one considers the level of approximation at which k f centration dependence for the reaction rate as shown in
is time independent and assumes that the distribution Eq. (1), although the basiS for this dependence would
over the internal states of the reactant molecules is the vary.
Maxwellian distribution, then the approach through the In the 1950's it became possible to examine experi-
Boltzmann equation leads to Eq. (2) as well. 2,3
mentally the dependence of the rate expression on the
concentration of the constituents for a single step reac-
aJpermanent address: Department of Mechanical Engineering, tion using radioactively labeled components. 5 This
University of Toronto, Toronto M5S lA4, Canada. method was used to study electron exchange between

J. Chern. PhY5., 79(11), 1 Dec. 1983 0021·9606/83/235605-11 $02.10 © 1983 American Institute of Physics 5605

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 5606 C. A. Ward: Effect of concentration on chemical reactions

ions of the same basic chemical component and it was concentration dependence of the derived expression is
possible to follow a reaction sequentially from shortly different from that of any of the previous theories, and
after the reactants were brought together until the sys- the "rate constant" has a different physical interpreta-
tem was near equilibrium. tion. However when the derived expression is applied
to consider the rate of reaction between ions in weak
To predict the concentration as function of time, it
solution under near-equilibrium conditions, the concen-
is necessary to take into account the reverse reaction:
tration dependence of the rate expression reduces to
(4) the conventional expreSSion [Le., Eq. (2)], and when
it is assumed that under equilibrium conditions the ions
The potential between such ions is said to be symmetric 9
fluctuate between associated ion pairs and free ions,
so that kT and kl are taken to be equal. The concentra-
the Brpnsted primary kinetic salt effect20 is predicted
tion of the radioactively labeled product may then be
for the change in the rate constant with ionic strength.
predicted as a function of time from Eqs. (2) and (4).5
As a result of the derived concentration dependence
When this predicted concentration was compared with
of the new reaction rate expreSSion, it is predicted that
the measurements strong disagreement was found during
the rate will be initially higher than that predicted by
the initial period of the reaction (e. g., see Refs. 10-12).
the previous theories. To examine this prediction we
The experimental values being considerably larger than
compare the concentration dependence predicted from
those predicted. However by this time the rate expres-
the new reaction rate with the available measurements
sion in Eq. (2) occupied such a hallowed pOSition in
for electron exchange between ions in solution. Close
chemical kinetics that the experimentalists chose to de-
agreement is found even in the initial period, and in
cide that there was something wrong with the experi-
particular it is found that there is no need to introduce
mental technique, rather than to decide that there was the (still unexplained) concept of zero-time exchange. 21
something wrong with the rate expression given in Eq.
(2); they therefore chose to add a constant to the pre-
dicted concentration obtained from Eqs. (2) and (4). II. TRANSITION PROBABILITY AND THE RATE OF
This constant was referred to as the "zero-time ex- CHEMICAL REACTIONS
change". The idea was that this amount of product was Consider an isolated system that is spatially homo-
being produced in some unspecified way in the experi- geneous and consists of r chemical components. Sup-
ment. By adding this constant, they could bring their pose that four of the components react according to the
experimental values into agreement with the theoretical scheme of Eq. (1). Let the number of molecules of com-
expression obtained from Eqs. (2) and (4). This amounts ponent s be denoted N s ; then the number of molecules of
to ignoring the data in the initial period and determining each component at the instant t1 may be denoted as A1
the rate constant approximately from the data in the where
near-equilibrium period; this procedure is discussed
further below. Al : N~, N~, N~, ... ,N; . (5)

We would wish to emphasize the similarity between We propose first to construct the expreSSion for the
this experimental situation and that which was noted probability of the system making a change in molecular
recently for gas adsorption on a solid surface. Many distribution corresponding to it undergoing a single step
experimental studies had indicated a much more rapid in the reaction defined in Eq. (1). We shall adopt the
rate of adsorption immediately after exposure of the quantum mechanical formalism. In determining this
gas to solid than was predicted on the basis of absolute rate from quantum mechanics, it should be noted that
rate theory. 13 Adsorption during this early period was it is impossible to specify the path by which the system
sometimes referred to as nonactivated. 14-16 goes from one molecular distribution to another. One
is concerned only with the molecular distribution before
When a recently developed theoretical procedure for the change and after.
predicting the rate of molecular transport across the
Suppose the molecules of the system are constrained
interface between two phases 17 was applied to predict
to be in a particular distribution Al and let the Hamil-
the rates of nondissociative13 and dissociative adsorp-
tonian operator for the system in this constrained con-
tion,18 it was found to give predictions consistent with
dition be denoted as li o• When the molecules of the sys-
the measurements, even in the period immediately after
tem are free to react (1. e., subjected only to their
the exposure of the solid surface to gas. The same the-
physical potentials), the Hamiltonian operator is de-
ory was also applied to predict the rate of gas absorp-
noted as iI. We may then define a potential operator V
tion by a liquid. 19 For this latter case, it is possible
as follows:
to make a complete prediction of the rate and when the
predictions were compared with a set of measurements (6)
they were found to be in close agreement. 19
and the Schrodinger equation may then be written
Using this same theoretical procedure-called sta-
tistical rate theory16-we propose t'o develop the ex- (7)
pression for the rate at which a chemical reaction of
the type indicated in Eq. (1) proceeds. We adopt the where if! is the wave function for the system; Ii is Planck's
formalism of quantum mechanics and allow it to lead constant divided by 21T and i is the unit imaginary num-
to concentration dependence of the rate expression. The ber. Physically, when both iIo and V are acting, the

J. Chem. Phys., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 C. A. Ward: Effect of concentration on chemical reactions 5607

components can react at their normal rate. We shall where we have made use of the Boltzmann definition of
assume that I V1/J I is small compared to Iilol/ll; hence entropy. 11
one can apply perturbation theory to construct the ex-
At that same instant that there is a probability for a
pression for the wave function. 22 The eigenfunctions
U::(g) and eigenvalues E~ of flo are obtained from
transition from molecular distribution Ai to Ak, there
is a probability for a transition from Ai to Ai where Ai
If u~(g) = E~ u~(g) , (8) is the molecular distribution resulting from the reac-
tion taking place one complete step in the reverse di-
where g are the generalized coordinates for the system.
rection:
In principle the eigenfunctions for the operators flO could
be constructed corresponding to all possible molecular AI : N~ + 1, N~ + 1, Nb -1, Nb -1, ... , N: . (18)
distributions of the system and the wave function ex- Following the same procedure as that outlined above,
pressed as one finds for this rate, i. e., J~, that

I/I(g, t) =L bv(t) u~ exp( - iE~ tlli) . (9) (19)

v

Among the eigenfunctions are those corresponding to where K{A j , Ai) is

the molecular distribution A/
K(Aj' AI) = (21T/1l) w/E) VJi I 12 (20)
(10)
Finally, we take as the expression for the net rate at
These eigenfunctions will be denoted u~(g) and their total which the chemical reaction proceeds in the forward di-
number by n(A i ). After one forward step of the reaction rection J, the difference in the two unidirectional rates
the distribution will be given in Eqs. (17) and (19):
Ak : N~ -1, N~ -1, Nb + 1, Nt + 1, ... , N~ • (11) J =K(A , Ak ) exp [S(A k) ~ S(A j ) ]
j
The eigenfunctions corresponding to this distribution
will be denoted u~(g) and we suppose there to be n(Ak) _ &(A. A ) [S(A I) - S(A 1) ]
I exp k (21)
such functions. Let J'

t=tk - tj . (12) In the following section this expreSSion for the net rate
at which the chemical reaction proceeds will be put in
To describe the fact that the system is initially in mo-
lecular distribution AJ we adopt the ensemble description a more usable form by making use of thermodynamic
relations to evaluate the difference in the entropies.
so that the initial value of the coefficient b,,(ti ) satisfy
the relation: III. RATE OF A CHEMICAL REACTION EXPRESSED
(13) IN TERMS OF CHEMICAL POTENTIALS AND THE
EQUILIBRIUM EXCHANGE RATE
where the double bar refers to the ensemble average of
a quantity. One may now follow the standard procedure The system will be assumed to be close enough to
of first order perturbation analysis to construct the ex- equilibrium so that the thermodynamic properties are
pression for the probability of a transition from Ai to well defined and the entropy may be expressed:
Ak at any instant in time, r(Aj' Ak)22'23: ~

S =(liT) U + (piT) V - Ls (J1. slT) Ns , (22)

r(Ai' Ak) =K(Ai' Ak ) n(Ak)/n(A i ) , (14)
where where T is the temperature, P the pressure, and J1. s
is the chemical potential of component s. We shall as-
(15) sume that none of the intensive thermodynamic proper-
and it has been assumed that the energy of the system is ties are changed as a result of the system undergoing
constant to within its uncertainty, ~E; wi(E) is the mi- one reaction step; thus for the isolated system we con-
sider one finds:
croscopic state density evaluated at the mean energy of
the system, and I Vik I is the average of the absolute S(Ak) - S(A i ) =(liT) [J1. A(Aj) + J1.B(Aj) - J1.C<~) - J1.D(Aj)l , (23)
values of the transition matrix elements of the operator
V for the transition from the microscopic states of the S(AI) - S(Ai) =(liT) [- J1.A(Ai) - J1.B(A i )+ J1.c(A i ) + J1.D(A i )] ,
molecular distribution Ai to states of molecular distri- (24)
bution Ak • and the instantaneous net rate of the reaction in the
~ter neglecting the difference between wJ(E) and
positive direction when the system is in molecular dis-
wk(E) and taking advantage of the Hermitian property of
tribution Ai can be written from Eq. (21) as
the transition matrix elements one finds J i =K(Aj, Ak)~i - K(A i , AI)~? , (25)
(16) where ~j is defined by
We now follow the statistical rate theory procedure17 ~i =exp{[J1. A(A J) + J1.B(A J) - J1.C(A i ) - J1.D(A J) l/kT}. (26)
to construct the expression for the unidirectional rate
of transitions from AJ to Ak, i. e., J,(A j , Ak), and find It remains to investigate K(AJ' Ak), K(AJ' AI), but note
that ~J only depends on the instantaneous molecular dis-
J,(AJ,Ak)=K(AJ,Ak)exp[S(Ak)-S(ANk] , (17) tribution AJ •

J. Chern. Phys., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 5608 C. A. Ward: Effect of concentration on chemical reactions

First, we consider the expression for the rate when constraints we consider, there is one stable equilibrium
the system has reached the stable equilibrium state. state for the system; thus Kx is a constant. However,
A necessary condition for equilibrium of a system in to determine its value will require speCification of the
which the reaction of Eq. (1) occurs is that mechanism by which the exchange takes place under
equilibrium conditions; i. e., such quantities as the num-
(27)
ber of encounters between the reactants and the prob-
where Ae denotes the equilibrium molecular distribution: ability of a reaction occurring in an encounter. However,
since we are only concerned with equilibrium conditions,
Ae: ~,N"s,N2,~, ... ,N;. ; (28)
a number of simplifications are applicable.
N".is the equilibrium value for the number of molecules
From Eqs. (17), (19), (23), (24), and (26), one can
of component s. Let AI' Ad denote the molecular distri-
see that the rate at which the reaction takes place in the
butions on either side of equilibrium:
forward direction is Kx.:l and in the reverse direction is
Af : N;. - 1, ~ - 1, m+ 1, ~ + 1, ... , N'. , (29) Kx.:l- 1 • The non equilibrium effects are contained in the
term.:l. Note that the non equilibrium effects do not by
Ad: N;.+I,~+I,m-l,~-l, ... ,N'.. (30)
themselves create a net chemical reaction. If under
According to Eq. (27), .:le is unity and since the equi- equilibrium conditions the exchange rate for the reaction
librium state is supposed stable, the net rate of decay is zero, then under nonequilibrium conditions it is zero
from this state must vanish so that from Eq. (25) one as well.
finds
Equation (34) gives the basic expression for the pre-
K(Ae, Ad) =K(Ae, AI) • (31) dicted rate of the chemical reaction. Below we apply
For later use it is important to have a physical under- it to the special case of electron exchange between ions
standing of K(Ae, Ad), K(A e, Af). We find they are the in solution with the aim of assessing its validity.
equilibrium exchange rates for the reaction. This may
be seen by noting that according to Eqs. (17), (19), (23), IV. REACTIONS IN SOLUTION
(24), and (27) the unidirectional rates for the reaction
Suppose the reactants are ions in solution; the chemi-
J(A., Ad) and J(Ae, Af) reduce to K(Ae, Ad), K(Ae, AI) respec- cal potentials may then be written:
tively in equilibrium, and according to Eq. (31) these
unidirectional rates are equal. Thus they are the equi- IL" := IL ~ (T, P) + k T In a", , (35)
librium exchange rate for the reaction and will be de- where 0' can be equal to A, B, C, or D; a", is the activity
noted Kx. In Ref. 17 it is pointed out that Kx may also of component 0' and IL~ is the reference chemical poten-
be examined at the microscopic level. It is found to be tial of that component in the solution.
the average rate at which the system makes a transition
between microscopic states of neighboring molecular For a system in which a reaction of the type shown in
distributions. Eq. (1) takes place, the condition for equilibrium is
given in Eq. (27); thus for equilibrium to exist
From Eq. (16) we may rewrite Eq. (31) as
K - aCe aD. (36)
K(Ae, Aa) =K(A e , Af) =Kx • (32)
e-aAeaBe '
Then following Ref. 17 we shall introduce the hypothesis where the subscript e on a property indicates an equi-
that an equation of the type shown in Eq. (32) is valid librium value; Ke is the equilibrium constant and is
for transitions between any two distributions; thus given by
Kx:= K(Ad' Ae):= K("Ae, Af) Ke = exp[(IL~ + IL~ - IL~ -1L~)/kTl . (37)
(33) Then .:l from Eq. (26) may be written
At the microscopic. level, Eq. (33) amounts to assuming
that the average rate at which the system makes transi- (38)
tions between microscopic states of neighboring molecu-
lar distributions is the same for all distributions; thus Thus the unidirectional rate at which the reaction pro-
this rate is independent of the number of states corre- ceeds to the right J/(A j , Ai) is from Eqs. (17), (23), (26),
sponding to a distribution. 17 At the macroscopic level and (33) given by
this rate must, according to Eqs. (31) and (32), be (39)
equal to the equilibrium exchange rate for the reaction.
One finds Similarly for the unidirectional rate at which
One of the immediate consequences of Eq. (33) is that
the reaction of Eq. (1) proceeds to the left, i. e. ,
the expression for the net rate at which the chemical
Jr(A j , AI) that
reaction proceeds may be written for any distribution as
Jr(A j , AI) :=KxaDaC/KeaAaB , (40)
(34)
and for the net rate of the reaction J from Eq. (34) that
and since J j and .:lJ only depend on the instantaneous
molecular distribution we have dropped the subscript j. J=Kx [(KeaAaB/aDaC> - (aDaC/KeaAaB)] • (41)
The important considerations of the equilibrium ex- The expression in Eq. (41) is the expression for the
change rate Kx remain to be made. For the macroscopic rate obtained from statistical rate theory when the re-

J. Chern. Phys., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 C. A. Ward: Effect of concentration on chemical reactions 5609

action of Eq. (1) takes place in the solution. Below we um constant given in Eq. (36) and using Eq. (45) in Eq.
shall use it to predict the concentration as a function of (41), we may write the chemical reaction rate expres-
time for a particular set of reactions, but first we com- sion as
pare a limit of this relation with the conventional expres- (46)
sion.
Since E is small compared to unity, one may expand
Eq. (46) and write
v. RELATION BETWEEN THE STATISTICAL RATE
THEORY EXPRESSION FOR THE CHEMICAL (47)
REACTIONIRATE ANDITHATIOF PREVIOUS THEORIES
and retaining terms to second power in E one finds that
From Eqs. (2) and (4) one finds that the expression the near equilibrium limit of the rate J. is
for the net rate of the reaction, J', according to all
J. =Kx(4E) (48)
previous theories is of the form
(42) and thus

The previous theories would disagree as to the approach J =2K

"
(~-~)
x aAeaBe
aDeaCe
(49)
to be used for the calculation of ki' k .. but they would
::tgree that the concentration dependence of the reaction Finally on examining Eq. (49) one may identify the uni-
rate is as shown in Eq. (42).5 direction rates. Thus the near-equilibrium unidirec-
tional rate in the positive direction is given by J'n:
Because of the assumptions in the absolute rate theory
approach one might expect it to be valid when the sys- (50)
tem is close to equilibrium, and in this limit there is where we have introduced the activity coefficients Ys
considerable empirical support for that approach. Thus that are defined from the relation
we shall examine the expression for the chemical reac-
tion rate obtained from statistical rate theory in the near (51)
equilibrium limit to see if it reduces to the conventional After comparing Eq. (49) with Eq. (3) one may identify
expression. The case we shall consider is that of re- the expression for the conventional rate constant k"
actions in solution. For this particular case, the Brpn- i. e.,
sted primary salt effect has been established empiri-
call14.25 and it forms a part of the basis for the idea of (52)
an activated complex in the conventional approaches to and with this identification the near-equilibrium limit
chemical reaction rates. It is of interest to see in what of the statistical rate theory expression is seen to re-
way the Brpnsted primary salt effect is obtained from duce to the conventional expression.
the approach developed above.
Note that in Eq. (46) the ratio 2Kx/aAeaBe is a ratio
It should be emphasized that in obtaining the expres- of equilibrium properties and is strictly constant. For
sion for the reaction rate in Eq. (34) from the statisti- example, suppose one considers a reaction of the type
cal rate theory approach no concept has been intrOduced indicated in Eq. (1) in which ions 'lZI, ~2 are reacting
of the reaction path being through an activated complex. in solution where ZI and Zz are the charges on the ions.
Indeed, no concept of the reaction path has been intro- Under equilibrium conditions, we suppose that the sys-
duced. This results from the fact that the basis for the tem fluctuates between associated ion pairs and free
statistical rate theory approach is quantum mechanical ions so that Eq. (1) may be expanded to write:
and the path by which a change takes place is not intro-
duced in quantum mechanics. (53)

As the definition of the near-equilibrium condition, where <5 denotes the associated ion pair and Zl +Z2 is its
we suppose that the parameter E is small compared to charge. Since we are considering equilibrium condi-
unity where E is defined by tions we may write without assumption:
(54)
(43)
where K~ is the equilibrium constant between the ions
and we shall suppose that the displacement from equi- a·l , ~2 and the complex (0)·1··2. Then equation (52)
librium by both reactants and products is symmetrical may be written
so that the following sum has the same value in the non-
(55)
equilibrium states as it does in equilibrium, i. e. ,
where Y~e is the activity coefficient for the ion com-
-1 (a. a.. + ---"'--"'-
----"'--"- an aro) = 1 (44) plexes and c 6e is their concentration.
2 a"'eaBe aDeaCe .
To understand this relation physically we consider it
Equations (43) and (44) may then be solved simultaneous-
in the dilute limit in which the Debye-Hiickel expres-
ly to give
sion for the activity coefficients is valid. If we suppose
a",aB ~ _ 1+E the ions are in aqueous solution at 25 ee, then the
(45)
a"'e aBe aD ac - 1- E Debye-Hiickel relation may be written 26
and after making use of the expression for the equilibri- (56)

J. Chem. PhY5., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 5610 C. A Ward: Effect of concentration on chemical reactions

where I is the ionic strength; for the associated com- We consider electron exchange between ions of the
plex: same basic component in solution and suppose that the
reaction is followed by radioactively labeling one of
(57)
them; thus as a special case of Eq. (1) we may write
After taking the logarithm of Eq. (55), inserting Eq.
*A+A+=*N+A, (60)
(57), and simplifying, one finds the result may be
written in the form: where the superscript • denotes a radioactively labeled
component; A is an ion and A+ is an ion of the same ele-
log10kf=B+1.018z1zzI1/Z, (58) ment in one higher state of ionization. We suppose that
where B is defined by the activity coefficients for the labeled and unlabeled
ions are equal (which would be exactly true within the
B =10g10 (~K,,/ c 6e ) • (59) Debye-Huckel approximation); then according to statis-
Note that B is constant for a given system. The rela- tical rate theory, the net rate may be written from Eq.
tion given in Eq. (58) is equivalent to the well-known (41):
Brpnsted primary salt effect showing the effect of the J =Kx (C A+ C.AKe/CAC. A+ - C AC.A+/K.,CA+C.A). (61)
ionic strength on the conventional rate constant. 20 It
has substantial experimental support. Z4,25 Note that in After introducing the reaction variable Y that is equal to
obtaining this relation no concept is introduced of equi- the concentration of the labeled product, one finds that
librium between an "activated complex" and the reac- Eq. (61) leads to
tants while the reaction is taking place. (dy/dt) =Kx {(a - y) ({3- y)/l- [l/(a - y)({3- y)]} ,
(62)
VI. ELECTRON EXCHANGE REACTIONS BETWEEN where a, (3 are the initial concentrations of components
IONS IN SOLUTION • A and A+, respectively, and the equilibrium constant
has been taken as unity because of the similarity of the
It has been seen above that the statistical rate theory labeled and unlabeled ions. From Eq. (36) one then
expression for the rate of a chemical reaction reduces finds that at equilibrium:
to the conventional rate expression in the near-equi-
librium limit. However, comparing the statistical rate
Ye =('l!{3/a +(3) . (63)
theory expression (41) with the conventional expres- Equation (62) may be integrated exactly. One finds
sion (4) shows that they predict very different re- from separation of variables and expansion in partial
sults when the amount of products is small, i. e., in fractions that if y vanishes initially and if the initial
the early period of the reaction. In this section we pro- ratio of the concentrations satisfies the condition
pose to examine this difference in the predictions by
0.1715<{3/a<5.82 , (64)
comparing the two rate equations with the previously
reported experimental results. then

Kxt = [a{3( a Z+ (3z)/ 4( a + (3)3] (Y/Ye) - [a Z{32 /4( a + (3)3](y/ Ye)Z - [aZf /2( a + (3)3] In(1 - Y/ Y.) + [(a + (3)/ 16] In[2a{3(1/ a + (3)2
X (y/y.)z _ (Y/ Ye) + 1] + (a - (3)2/8'; 6a {3 - (a Z+ (32) [tan -1({[ 4a{3y/ Y.( a + (3)] - (a + (3)}/v' 6a13 - (a Z + (3Z»
- tan- 1 [- (a + (3)J" 6a{3 - (a 2 + (32)]] . (65)
When (3/ a does not satisfy the condition of relation (64) one finds
~t = ya{3(a z + f)/4Ye( a + (3)3 - aZ{3V/4y~(a + (3)3 - [a zr:J2j2(a + (3)3] In(1 - y/Ye) + (a z - {3zW640(a + (3)z

+ [(az _ fW640(a + (3)Z] In { [4( a{3/ Q! + (3)(y/ Ye) - (0' + (3 + 40)] (40 - 0' - (3) } + (0' + (3/16) In{2[a{3/(O' + (3)Z]
[4(O'{3/a +(3) (Y/Ye) - (0' +(3- 40)] (-41i - 0' - (3).
(66)

i. e., when the predictions from statistical rate theory or from the
conventional rate expression, it is necessary to choose
{3/O'<0.1715 or {3/O'> 5.82 . (67) a value of either Kx or k. We shall choose values of
If one adopts the conventional rate expression (42), each of these such as to minimize the square of the total
then it may be simply integrated as well. One finds, error in each of the respective cases.
since k f and kr are said to be equa1 9: Suppose that a set of data is available so that a set of
kt =(0' + (3)"11n(1 - y'/Ye)"1 , (68) n values of Y/Ye are given at times tl> then we take as
the definition of the square of the error:
where Y' is the concentration of the labeled product and
it has also been required in this case that y' is initially (69)
zero.
To compare the available measurements with either where t~ is the predicted time at which the concentra-

J. Chern. Phys., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 C. A. Ward: Effect of concentration on chemical reactions 5611

~----------~------------,

tion ratio has the value Y/Ye'

1.0
First consider the predictions from statistical rate ----
theory. Equations (65) or (66) may be viewed as being 0.9
of the form ~0.8
(70) ~07 /
/'
/
where F(y/ Ye) is the right-hand side of either Eq. (65) 206
o . /
a:: /
or (66), then Eq. (68) may be written as c: 0.5 /
o /
/
(71)
'E 0.4 /
-=~ 0.3 I
/
u I
After requiring that the first derivative of E2 with re- § 0.2 /
spect to K" vanishes, one finds the value of K" which u
0.1 /
I
minimizes the error to be given by

Xx= t F(YJYe)/t. tj • (72)

2 3 4

FIG. 2. Same comparison as that of Fig. 1 and same condi-

5 6 7
Time (Min)
8 9 10 11 12 13 14

The value of Xx determined in this way will be used be- tions except the hydrogen ion concentration was 0.4 M. See
low to compare the predicted concentration with the Table I.
measurements.
For the conventional rate expression one may also A. Comparison of predicted concentration as a function
use this procedure to obtain a value of k which mini- of time with available experimental data
mizes the error. If Eq. (68) is viewed as being of the
form: Previous studies of electron exchange between ions in
solution provide some data on the concentration as a
kf=!(y'/Ye) , (73) function of time for a reaction of the type shown in Eq.
then the value of k which minimizes the total error (60). The usual procedure is to label an ion radioac-
squared is tively, mix the components together, and measure the
rate of charge transfer by measuring the appearance of
(74) a product ion, using separation techniques. 27.28
Although these reactions have been extensively stud-
In comparing the predictions from the statistical rate ied,5.6 very little of the data has been reported in suffi-
theory approach and from the conventional rate theory cient detail to allow it to be examined from a new theo-
expression we shall choose Xx and k on the same basis, retical point of view. Rather, the usual procedure is
i. e., from Eq. (72) and (74), respectively. to give only the rate constant inferred from absolute
rate theory. The principal exception to this is Wahl
and his co-workers who usually report the concentration
as a function of time for at least one experiment. 10-12
1.0 This will allow us to compare our predictions with their
experiments for three different types of ions. It was
0.9
also Wahl and his co_workers10- 12 ,29 who most clearly
~0.8 pointed out the difficulty of the zero-time eXChange, al-
~0.7 though the concept was later briefly discussed by Rey-
o nolds and Lumry in their reviewS and by Lipkowski et
ii 0.6
a: . al. 21
c 0.5,
~ 0.4
1 A second method that has been used to study these re-
actions is nuclear magnetic resonance. 30,31 In this
-=~ 0.3
u
method there is no change in the concentration as a func-
§ 0.2 / tion of time; however as will be discussed below these
u / methods measure K" directly.
0.1 I

2 3 4 5 6 7 8 9 10 11 12 13 14 1. Electron exchange between vanadium ions in solution

Time (Min)
Krishnamurty and Wahl12 have reported the concentra-
FIG. 1. Comparison of the measured concentration (solid dots tion as a function of time for the following reaction:
with 10% error bars) for the V2+, V3+ electron exchange reac-
tion with the calculations from statistical rate theory (solid (75)
line) and with the conventional theoretical expression (dashed
line). Both theories are required to be initially zero. The for three different hydrogen ion concentrations. The
data is that of Ref. 12. The medium was an aqueous solution reaction took place in 1. 00 f HCIO, with ionic strength
of HCIO, with 1.00 M If" and the ionic strength adjusted to 2.00 of 2. 00 and the hydrogen ion concentration was varied
by addition NaClO,. The initial conditions and the inferred at constant ionic strength by addition of NaCIO.. Their
equilibrium exchange rate are given in Table I. data is shown in Figs. 1-3. Since the initial concentra-

J. Chern. Phys., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 5612 C. A. Ward: Effect of concentration on chemical reactions

1-------- ~------------~----

period. Note that the measurements are above the pre-

1.0 ----------- dictions in each case; thus the discrepancy could not be
0.9 explained by incomplete mixing or difficulties with flow.
These difficulties would tend to decrease the rate and
:.: 0.8
thus would give measurements less than the theory.
~0.7 The discrepancy is too large to be explained as inaccu-
,go 06
. racies in measurements in time. One possible explana-
a: tion is an ineffective quenching procedure, but this ex-
c: 0.5
o planation has not been established by further experimen-
'E 0.4 tation. 12
t 0.3
u When first introduced the zero-time exchange was
§ 0.2 a small percentage of the total. 29 However for some
u
0.1 systems studied later that had higher rates of exchange,
the percentage was significantly increased. For exam-
2 3 4 5 6 7 8 9 10 11 12 13 14 ple, for the case shown in Fig. 2, the zero-time ex-
Time (Min) change was greater than 40% of the total. Hudis and
FIG. 3. Same comparison as that of Fig. 1 and same condi- Wahl 29 point out that the apparent zero-time exchange
tions except the hydrogen ion concentration was 0.19 M. See is reproducible for a given set of experimental condi-
Table 1. tions.
If one accepts the validity of the statistical rate theory
tions are equal in each case, Eq. (65) gives the pre- approach for the moment, then it is clear that there is
dicted concentrations as a function of time according to no need to introduce the zero-time exchange concept.
As may be seen in Figs. 1-3, the results obtained from
statistical rate theory, and the solid line in each of these
statistical rate theory are in agreement with the data.
figures shows the results calculated from Eq. (65) where
Also one can understand why the previous theories gave
the value of Kx has been chosen so as to minimize the
consistent results for the rate constant k: by introducing
total error. The values of ~ are given in Table I. The
the zero-time exchange, the previous theories are in
error in this type of experiment has been estimated as
essence being fitted to the data only at later times. As
10%28 and 13%.31 We have shown error bars of 10%,
has been pointed out above, the previous expressions for
based on these estimates. No repeated experiments
the rate of the chemical reaction are in agreement with
were reported.
statistical rate theory in this limit, and thus should
Note that in each case there is close agreement be- lead to consistent results for k.
tween the predictions from statistical rate theory and
The value of k reported by Krishnamurty and Wahl for
the measurements, even in the period initially after the
each experiment is listed in Table I. These values were
reactants were brought into contact.
inferred by assuming a zero-time exchange of 30%, 30%,
This agreement between statistical rate theory and the and 43% of the equilibrium exchange. If one does not
measurements is in contrast to what one finds if the re- assume a zero-time exchange, but requires that the
sults from the conventional rate theories are compared previous theories have zero amount of product at zero
with the results. This comparison can be seen by com- time, then the value of rate constant inferred from the
paring the dashed line in Figs. 1-3 with the experimen- data is changed. The value of the rate constant inferred
tal results. In each case, there is strong disagreement on the basis of minimum total error and no product ini-
between these theories and the experiment in the initial tially present is listed in Table I as k o. Note that the

TABLE 1. Comparison of rate constants.

(A) Electron exchange between y2+ and y3+ at 25°C; see Figs. I, 2, and 3
W/M a/f {3/f K:r Xl04/f min-l k o/(fminr1 k/(fminr1
1.00 0.086 0.086 3.87 1.273 0.79
0.40 0.086 0.086 6.5 1.415 1, 06
0.19 0.086 0.086 11.44 2.7 2.01

(B) Electron exchange between Mn04 and MnO~-; see Fig. 4

k XlO-3/(Msrl

0.1 0.4 0.96 3.6 1.48 1. 38

(C) Electron exchange between Ag+ and Ai+; see Figs. 5 and 6

0.2 10.61 1. 793 1.627 263

11. 4 3.366 4.14 0.1462 ? 648

J. Chern. Phys., Vol. 79, No. II, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 C. A. Ward: Effect of concentration on chemical reactions 5613

value of ko is 60%, 35%, and 30% larger than the value

1.0
of k for the three cases shown in this Table.
0.9
~0.8
2. Electron exchange between manganate-permanganate ;:'0.7

The kinetics of electron exchange between manganate

and permanganate have been studied both by radioactive
labeling10 and nuclear magnetic resonance. 30,31 In the
case of labeling, the reaction is
*Mn~- + MnO;;: *MnO; + Mn~- . (76)
10
For the only case reported by Sheppard and Wahl the 0.2 0.3 of. 0.5 O.s 0:7 0.8 0.9
Time Isec)
reaction took place in a weak aqueous solution with sev-
eral ions present. The data obtained for this case and FIG. 5. Comparison of the measured concentration as a func-
tion of time for the Ai+, Ag+ electron exchange reaction with
the conditions are given in Fig. 4, along with the cal-
the calculation from statistical rate theory (solid line) and with
culations from statistical rate theory [Eq. (65)] and the the conventional theoretical expression (dashed line). Both
value of ~ is given in Table I. The statistical rate theories are for the mechanism of Eq. (83) and both are re-
theory approach leads to an overprediction of the data quired to be initially zero. The data is from Ref. 11. The
points during the first 3 s. Since this is the only ex- medium was 5.87 f HCI0 4 , at 0.2 ·C. The initial conditions
periment reported, it is not possible to know whether and inferred value of the equilibrium exchange rate are given
this is significant because explanations are easily in Table I.
imagined, such as a lack of complete mixing during this
early period.
The conventional theory again is significantly below parison of the results obtained for Kx from the two dif-
the data and there is no explanation available other than ferent methods.
the zero-time exchange.
3. Electron exchange between silver ions in solution
When the reaction is studied with nuclear magnetic
resonance, there is no change in concentration, i. e., Although more difficult chemically, one of the fastest
the reaction is exchange reactions studied with the radioactive labeling
technique is
Mn~- + MnO;;: MnO; + Mn~- , (77)
(79)
hence the expreSSion for the unidirectional rates as ob-
ll
tained from statistical rate theory, L e., Eqs. (44) and Gordon and Wahl examined this reaction in 5.87 f
(45) reduce to the equilibrium exchange rate: HCI04 and have reported two experiments in which con-
centration as a function of time was measured. This
(78) reaction is of particular interest for two reasons: first,
Several values ~ have been measured in Refs. 29 and because its kinetics are very rapid and thus one would
30, but there is not sufficient data reported from the expect it to be a challenge to the conventional rate the-
radioactive labeling technique for us to make a com- ories if they are only valid in the equilibrium limit, as
has been claimed above. Second, Gordon and Wahll l
found it necessary to abandon the mechanism of the re-
. -_ ... __ ._----------------, action being first order in each constituent, i. e., Eq .
(1), and adopted a mechanism corresponding to Ag2+
1.0
0.9
---- , disproportionating, L e.,
~0.8 2Ag2+;: Ag3+ + Ag+ (80)
~0.7
~ 0.6 and took the rate expression to be
~ 0.5
rate =k( Ci'"g)2 •
.2 (81)
2 0.4
~ 0.3 Their main reason for propOSing this mechanism was
§ 0.2
u
because the usual expression for the rate as being first
0.1 order in Ag+ and Ag2+, Le., Eq. (2), did not lead to an
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1·5 16 17 18 19 adequate fit of the data. However, no measurement of
Time (sec) the presence of Ag3+ was reported and it is a necessary
FIG. 4. Comparison of the measured concentration (solid dots constituent of the mechanism in Eq. (80).
with 10% error bars) for the MnOl-, MnO, electron exchange Gordon and Wahll l report measurements of concen-
reaction with the calculation from statistical rate theory (solid
tration as a function of time for only two of their many
line) and with the conventional theoretical expression (dashed
line). Both theories are required to be initially zero. The experiments. Their data for these two experiments are
data is that of Ref. 10. The medium was an aqueous solution: shown in Figs. 5 and 6. For the values of ~ given in
0.16 f NaOH, 0.08 f Na2S0h 4x 10'" f KRe04' The initial con- Table I, the predictions from statistical rate theory ap-
ditions and the inferred equilibrium exchange rate are given in proach are also shown in these Figs. (solid line). The
Table I. calculations in this case are from Eq. (65) because the

J. Chern. Phys., Vol. 79, No. 1" 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 5614 C. A. Ward: Effect of concentration on chemical reactions

mechanism by which the exchange rate is established;

1.0
0,9
however, since it is the equilibrium exchange rate one
~O.8
can take advantage of the substantial simplifications
:;0.7 that become available in the equilibrium state.
,g 0.6 (2) The expression for the rate of the chemical reac-
Ii
!5 0.5 tion may be written in terms of the concentrations and
'6 0.4 the equilibrium exchange rate by conSidering a special
i 0.3 case and making use of the expressions for the chemical

J
15 0.2 potentials. For the case of a reaction between ions in
u
0.1
solution one finds that the concentration dependence of
0.2 D.3 0.4 0.5 O.S 0.7 DB 0.9 the reaction rate is different than that predicted by ab-
Time Isec)
solute rate theory or any of the previous theories, with
FIG. 6. The same comparisons as in Fig. 5 and under the the statistical rate theory expression predicting a higher
same conditions except the temperature was 11. 4 ·C and the rate during the initial period.
initial conditions are as given in Table I.
When the statistical rate theory expression is exam-
ined in the near-equilibrium limit, it is found to reduce
initial concentrations do not satisfy Eq. (64). Note that to the same concentration dependence as used in previous
no disagreement is observed between the calculations theories and an expression is obtained for the conven-
from statistical rate theory and the data of Gordon and tional rate constant k in terms of quantities ariSing from
Wahl. the statistical rate theory approach.
By contrast, the calculations from previous theories To evaluate the expreSSion obtained for the conven-
for the mechanism that assumes first order kinetics tional rate constant, a dilute electrolyte system is con-
for both Ag+, Agz+ [i. e., from Eq. (68)] are shown in sidered and it is assumed that in equilibrium the ions
these figures as well (dashed line) and they show little fluctuate between associated ion pairs and free ions.
agreement with the data. This is particularly true in When the derived expreSSion for the conventional rate
Fig. 6 where the shape of the prediction from the pre- constant is evaluated for this system, the well known
vious theories for the first order mechanism does not empirical relation between the ionic strength and the
have the same shape as the data. Thus one can see why conventional rate constant (called the Brpnsted primary
Gordon and Wahl found it necessary to abandon this salt effect) is obtained.
mechanism.
(3) To examine the prediction of a higher reaction rate
From the point of view of statistical rate theory then, during the initial period made from statistical rate the-
so far as one can tell from the data given, there is no ory, this approach was applied to examine the electron
reason to abandon the mechanism of Eq. (79), but it exchange reaction between ions in solution. Data avail-
is the rate expreSSion that assumes first order in each able in the literature that gave the concentration as a
reacting component, i. e., Eq. (2), that should be function of time for six different experiments was ex-
abandoned. amined. This data was obtained by Wahl and his co-
workers using radioactive labeling techniques for vana-
VII. SUMMARY AND CONCLUSIONS dium, manganate-permanganate, and silver.
(1) In the preceding development, the formalism of According to statistical rate theory, Kx should be con-
quantum mechanics is used to obtain the expreSSion for stant for these experiments. After integrating the the-
the probability of a system undergoing a change in mo- oretical expreSSion for the reaction rate and chOOSing
lecular distribution corresponding to a single step of a the value of K" to be that which gives a minimum total
chemical reaction in an element of time, This formal- error between the theory and experiment, one finds
ism does not require that the path by which the system that there is no measurable disagreement between the
makes the change in distribution be specified. One only theory and experiment for five of the six cases examined,
considers the molecular distribution before and after and that the disagreement is small for the other case.
the step. From the expression for this probability the This general agreement exists throughout the experi-
statistical rate theory approach is used to obtain the mental period, even in the period initially after the re-
expression for the net rate of the chemical reaction. actants are brought together.
The form of the expression is that of a product of the This is in contrast to the results obtained from pre-
equilibrium exchange rate Kx with a factor ariSing from vious reaction rate theories. All of these have the same
nonequilibrium effects that is ultimately expressed in
concentration dependence. They all disagree with the
terms of the chemical potentials of the products and re- data in the initial period. To obtain agreement in this
actants. H under equilibrium conditions, the exchange period, it has become conventional to introduce the con-
rate for a reaction vanishes; then the reaction rate is cept of zero-time exchange. This amounts to adding a
zero at all times; thus non equilibrium effects do not constant to the relation expreSSing the concentration as
create a reaction but they only accelerate it in one di- a function of time that is obtained by integrating the rate
rection or the other. expressions. Although adding this constant does lead
To determine the value of the equilibrium exchange to agreement between the calculations and the measure-
rate one must consider a speCial case and specify the ments except in the case of silver, the zero-time ex-

J. Chern. Phys., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15
 C. A. Ward: Effect of concentration on chemical reactions 5615

change has not been explained on a physical basis. In ular Theory of Gases and Liquids (Wiley, New York, 1954),
the case of silver, even the addition of the zero-time p. 449.
exchange was not sufficient to obtain agreement between 'w. L. Reynolds and R. W. Lumry, Mechanisms of Charge
Transfer (Ronald, New York, 1966), p. 74.
the previous theories and the experimental data. It
6R. D. Cannon, Electron Transfer Reaction (Butterworth,
was necessary to suggest a change in mechanism. Toronto, 1980), p. 175.
From the point of view of statistical rate theory, it 1R. R. Dogonadze, A. M. Kuznetsov, and T. A. Marsagishvili,
Electrochim. Acta 25, 1 (1980).
is unnecessary to introduce the concept of zero-time
Bv. G. Levich in Physical Chem., edited by H. Eyring, D.
exchange, since the predictions are in agreement with Henderson, and W. Jost (Academic, New York, 1970), p. 986.
the theory, even in the period immediately after the re- 9F. A. Cotton and G. Wilkinson, Basic Inorganic Chemistry
actants are brought together; nor is there any need to (Wiley, Toronto, 1976), p. 154.
suggest a change in mechanism on the basis of the in- 10J. C. Sheppard and A. C. Wahl, J. Am. Chern. Soc. 79,
formation presently available in the case of silver, 1020 (1957).
since there is agreement in this case as well. JIB. M. Gordon and A. C. Wahl, J. Am. Chern. Soc. 80,
273 (1958).
This improvement in the agreement between the theory 12K. Y. Krishnamurty and A. C. Wahl, J. Am. Chern. Soc.
and experiment (seen in Figs. 1-6) results from the con- 80, 5921 (1958).
centration dependence of the rate expreSSion derived 13C. A. Ward and R. D. Findlay, J. Chern. Phys. 76,5615
(1981).
from the statistical rate theory approach.
14A. Clark, The Theory of Adsorption and Catalysis (Academic,
(4) We note that the agreement between the results New York, 1970), Chap. 9.
from the statistical rate theory approach and the avail- 15D. O. Hayward and B. M. W. Trapnell, Chemisorption, 2nd
able data for the rate of chemical reactions discussed ed. (Butterworths, London, 1964), Chap. 3.
16S. J. RothenburgandY. S. Cheng, J. Phys. Chern. 84,
above is matched by the agreement between this theory
1644 (1980).
and the measured rate of gas absorption by a liquid, 19 nC. A. Ward, R. D. Findlay, andM. Rizk, J. Chern. Phys.
and the measured rates of nondissociative13 and disso- 76, 5599 (1982).
ciative18 adsorption of a gas on a solid surface. Thus 18R. D. Findlay and C. A. Ward, J. Chern. Phys. 76, 5624
the statistical rate theory approach appears to lead to (1982).
an improved expreSSion for the rate of a broad range 19C. A. Ward, M. Rizk, and A. S. Tucker, J. Chern. Phys.
of processes. 76, 5606 (1982).
2OW. J. Moore, Physical Chemistry (Longman, London, 1972),
p. 465.
ACKNOWLEDGMENTS 21J. Lipkowski, A. Czerwinski, E. Cieszynski, Z. Galus,
I wish to thank Roger Parsons and Jacek Lipkowski and J. Sobkowski, J. Electroanal. Chern. 119, 261 (1981).
for reading the manuscript and for many helpful sug- 22W. Pauli, Collected Scientific Papers, edited by R. Kronig
and Y. S. Weisskopt (Interscience, New York, 1964), Vol. 1,
gestions; the Laboratoire d'Electrochimie Interfaciale
p. 549.
du CNRS, France for one year of residence and the Na- 23R. C. Tolman, Principle of Statistical Mechanics (Oxford
tural Science and Engineering Research Council of University, London, 1938), p. 395.
Canada for support of this work. 24J. N. Br~nsted, Z. Phys. Chern. 102, 169 (1922).
2,y. K. La Mer, Chern. Rev. 10, 179 (1932).
26W. J. Moore, Ref. 20, p. 455.
21N. Sutin, Annu. Rev. Nucl. Sci. 12, 285 (1962).
28E. F. Caldin, Fast Reaction in Solution (Blackwell, Oxford,
1964), p. 24.
29J. Hudis and A. C. Wahl, .J. Am. Chern. Soc. 75, 4153
tH. Eyring, J. Chern. Phys. 3, 107 (1935). (1953).
2M. A. Elliason and J. O. Hirschfelder, J. Chern. Phys. 300. E. Myers and J. C. Sheppard, J. Am. Chern. Soc. 83,
30, 1426 (1959). 4739 (1961).
3J . RossandP. Mazur, J. Chern. Phys. 35,19(1961). 31A. D. Britt and W. M. Yen, J. Am. Chern. Soc. 83, 4516
'J. O. Hirschfelder, C. F. Curtiss, and R. Bird, The Molec- (1961).

J. Chern. Phys., Vol. 79, No. 11, 1 December 1983

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.24.51.181 On: Sun, 30 Nov 2014 01:56:15