c1
c1
c1
Electrochemical Thermodynamics
3.1
Equilibrium Cell Potential and Gibbs Energy
As discussed in Chap. 2 any chemical reaction involving charge exchange be-
tween two different redox reactants, see Eqs. (3.1 a,b) can be performed virtual-
ly – and very often also practically – by performing the redox reactions of the
two redox-couples separately at two different electrodes but jointly in a divided
electrochemical cell (Fig. 3.1):
A x + + B( ) → A ( x +1)+ + B y +
y +1 +
total reaction (3.1)
A x+ → A( ) + e−
x +1 +
redox couple A (3.1 a)
B( ) + e − → By+
y +1 +
redox couple B (3.1 b)
may be performed in two half cells, one containing the ferric/ferrous system
(Fe 2+ → Fe3+ + e − ) , the other containing the ceric/cerous system (Ce 4 + + e − → Ce3+ ).
Fig. 3.1. Schematic of a cell reaction composed of two separate redox reactions which jointly
would establish a homogeneous redox reaction between an oxidant and a reductant.
18 3 Electrochemical Thermodynamics
Fig. 3.2. Equilibrium cell potential U0 in a cell divided by a diaphragm composed of two in-
terfacial potentials ∆ϕi and a (mostly negligible) diffusion potential ∆ϕdiff .The diffusion
potential ∆ϕdiff vanishes if a salt bridge with an electrolyte whose cations and anions pos-
sess transference numbers of 0.5, each, connects the two separate cells. With aqueous KCl
solutions this is the case
Both being connected by a salt bridge1), they deliver an electrical potential dif-
ference U0 between two Pt-electrodes inserted into the two separated electro-
lytes (Fig. 3.2)
Likewise the formation of hydrochloric acid, dissolved at a given concentra-
tion in water, from gaseous chlorine and gaseous hydrogen, realized in two
half cells connected by an electrolyte bridge, generates a cell potential. If the
reaction is performed electrochemically, (H 2 + Cl 2 → → +
← 2HCl ; H 2 ← 2H + 2e
−
and
− → −
Cl 2 + 2e ← 2Cl ) , by dipping a chlorine electrode (made of platinized platinum
and being supplied and sparged with elemental chlorine) and a hydrogen elec-
trode (platinized, H2 -sparged platinum electrode) into an aqueous solution of
hydrochloric acid of the respective concentration, the equilibrium cell potential
can be measured between the two platinum electrodes at vanishing current with
a voltmeter of high internal resistance.
This cell potential is sometimes called “electromotive force”. Under condi-
tions of vanishing cell current and established reversibility of the two different
electrode reactions, (which is accomplished by effective electrocatalysis of the
two electrode reactions by applying for instance platinum black), the so called
open cell potential becomes the equilibrium cell voltage U0.
U0 equals the free energy ∆G of the cell reaction per mol of product divided
by the number νe of Faradays (1F=96,500 As) necessary for the electrochemical
generation of one mol of product.
1/2H 2 + 1/2Cl 2 →
← HCl diss (3.2 a)
1/2H 2 → +
← H diss +e
–
(H2-electrode) (3.2 b)
1/2Cl 2 + e − → −
←Cl diss (Cl2-electrode) (3.2 c)
1 The connecting electrolyte in the bridge should have equal transference numbers (0.5)
for cations and anions imposing equal distribution of cationic and anionic currents on
the cell.
3.1 Equilibrium Cell Potential and Gibbs Energy 19
∆G
U0 = (3.3)
νe F
where νe=1 for HCl.
If the reaction Eq. (3.2 a) proceeds spontaneously, the hydrogen electrode is the
anode and if the electronic current passes through a wire or a load from the anode to
the cathode, then ∆G is negative and the cell would spontaneously produce hydro-
chloric acid. For HCl electrolysis, however, the chlorine electrode is the anode, the hy-
drogen electrode becomes the cathode, and under these conditions where reaction at
Eq. (3.2 a) is reversed by imposing an external potential differentially greater than U0,
hydrochloric acid would be electrolytically decomposed and ∆G becomes positive.
The Gibbs energy of any chemical reaction depends on:
(1) the temperature;
(2) the concentrations of dissolved reactants; and
(3) the partial pressures of gaseous reactants.
Because of the latter two dependencies it is reasonable to define standard molar
Gibbs energies called standard chemical potentials µi0 and also standard cell volt-
ages U0 defined by standard concentrations and standard pressures of reactants
and products. Standard concentrations of dissolved substances – in particular for
aqueous solutions of electrolytes are defined by solute activities of 1 mol dm–3 and
for gaseous substances by partial fugacities of 0.1 MPa or 1 bar respectively.
Concentration related activities and fugacities are derived from actual con-
centrations and partial pressures by multiplying molar concentrations with ac-
tivity coefficients, γi,
ai=ci γi (3.4 a)
or partial pressures with fugacity coefficients fi,
p*=p
i ifi. (3.4 b)
These coefficients account for deviations from ideal behaviour which is de-
scribed by the two limiting equations, Eqs. (3.5 a, b), which describe chemical
potentials of gases and solutes at low – or more precisely – vanishing pressures
and concentrations.
µ i = µ 0i + RT ln(pi p0 ) (3.5 a)
ν ν
U 0 = ∑ νi , jµ i ,0j + RT ln Π i , jc j j pi i
( νe F ) (3.6)
with νi,j the stoichiometric factors of gaseous, index i, and dissolved, index j, re-
actants and products of the cell reaction; νi,j is positive for products and negative
for educts and νe is the stoichiometric factor of electrons for the reaction under
consideration..
For the hydrogen/chlorine cell this translates into
FU 0 = ∆G HCl
0
(
+ RT ln pH 2 pCl 2
)
12
c HCl
−1
(mol –1
dm3 bar
) (3.6 a)
so that the measurement of equilibrium cell potentials combined with their ex-
trapolation to standard condition is used to determine data for chemical equi-
libria, in particular for chemical equilibria which involve dissolved electrolytes.
For the chosen chlorine/hydrogen cell the corresponding relation, which al-
lows to determine the standard Gibbs energy, reads
(p )
1/2
0
U HCl F = ∆G HCl
0
= −RT ln K = −RT ln a HCl , diss H2 ⋅ pCl 2 (3.6 d)
equil
3.2 Electrode Potentials, Reference Electrodes, Voltage Series, Redox Schemes 21
For the electrochemical engineer it is most important to note that the equilib-
rium cell potential U0 determines the minimum amount of electrical energy
which has to be expended for any given electrochemical conversion or electrol-
ysis process (∆G >0) accounting for any given concentration and pressure of re-
actants and products or the maximal amount of electrical energy which can be
generated from a battery or a fuel cell (∆G <0) which is driven by the respective
reaction. Any irreversibility in a practically performed electrochemical process
increases the expended electrical energy for an energy consuming electrolysis
and decreases the extractable electrical energy from a fuel cell or a battery.
3.2
Electrode Potentials, Reference Electrodes, Voltage Series, Redox Schemes
For vanishing cell currents and neglecting diffusion potentials the electrical po-
tential in the electrolyte is constant between the two electrodes (Fig. 3.2). There-
fore the equilibrium cell voltage U0 is to a good approximation equal to the dif-
ference of the two so-called interfacial potentials ∆ϕa and ∆ϕc which exist at the
phase boundaries anode/electrolyte and cathode/electrolyte. Interfacial poten-
tials mean differences of internal potentials of neighboured phases:
U0=[∆ϕa–∆ϕc]i=0=[(ϕa–ϕel)–(ϕc–ϕel)]i=0 (3.7)
For principle reasons the internal potential differences ∆ϕa and ∆ϕc which are
called interfacial potentials cannot be measured whereas the difference of inter-
facial potentials occurring in electrochemical cells are well defined and easily
measurable quantities. If one of these interfacial potentials is fixed by a rapidly
established electrochemical equilibrium which does not vary with time, a
change in ∆ϕa–∆ϕc reflects any change in the second non-fixed interfacial poten-
tial.
This is the reason to introduce reference electrodes which possess themselves
well defined and easily reproducible, though principally unknown, interfacial
potentials ∆ϕreference (Fig. 3.3) in order to measure electrode potentials of single
electrodes vs. such fixed electrode potential of the reference electrode. As the
most popular example of a reference electrode we refer to the normal hydrogen
or standard hydrogen electrode (NHE). This reference electrode, which is by no
means a uniquely distinguished reference, is a hypothetical hydrogen electrode
which dips into an acidic aqueous solution in which protons possess the activity
(activity means activity coefficient corrected concentration) of 1 mol dm–3 and
is sparged with hydrogen of 0.1 MPa pressure (more precisely with a fugacity of
0.1 MPa). Other practical reference electrodes are for instance the 0.01n HCl/hy-
drogen electrode, the KCl-saturated calomel electrode (SCE: Hg, covered by a
Hg2Cl2/Hg mixture and being in contact with a saturated aqueous potassium
chloride solution) or 1 mol l–1 or also a KCl-saturated silver/silver chloride elec-
trodes 1 mol l–1 KCl Ag/AgCl or KClsat Ag/AgCl which are connected electrolyt-
ically by means of an electrolyte bridge most often in the form of a Luggin – cap-
illary, filled with aqueous saturated KCl-solution, Fig. 3.3, to the electrode under
22 3 Electrochemical Thermodynamics
Fig. 3.3. Schematic description of the use of the interfacial potential ∆ϕref of a reference elec-
trode in an electrochemical cell by means of an electrolyte bridge, most often filled by sat.
KCl solution. The electrolyte bridge is introduced in the form of a so called Luggin-capillary.
Index WE means working electrode
solution and the (unknown) value of the internal potential difference of the standard
hydrogen electrode, ∆ϕ H 2 , is deliberately and per definitionem taken to be zero.
0
The standard potential value of any other equilibrium electrode (all partici-
pating reactants possess either unit activity, if they are pure phases or all con-
centration activities of dissolved reactants equal 1 mol/dm3 or gas fugacities
equal 0.1 MPa or 1 bar respectively) are referred to the potential of the standard
hydrogen electrode and correspondingly ordered in the voltage series.
The voltage series for aqueous solutions of a number of metal/metal ion elec-
trodes of practical interest is collected in Table 3.2 a, for element/element-anions
electrodes in Table 3.2 b and for redox couple electrodes, where the reduced as
well as the oxidised species is dissolved in solution, in Table 3.2 c.
The concentration dependence of the equilibrium potential of a redox couple
A + e− →
←A
−
(3.8 a)
is given by the Nernst equation
RT c
E = E0 + ln A (3.8 b)
F c A−
For the more involved and more generally formulated redox-reaction which
is written with omission of the electric charge of species A and B
ν A A + νe e − →
← νB B (3.9 a)
one obtains the Nernst equation
ν
RT c A
E = E0 + ln A (3.9 b)
νe F ν
cB B
24 3 Electrochemical Thermodynamics
Table 3.2 c. Normal potentials in Volt of redox couples with soluble partners1)
[Co(CN)6]4–/ –0.83 [Mn(CN)6]4–/[Mn(CN)6]3– –0.22 Hg22+/2Hg2+ +0.906
[Co(CN)6]3–
Ga2+/Ga3+ –0.65 [Co(NH3)6]2+/[Co(NH3)6]3+ +0.1 IrCl63–/IrCl62– +1.02
In2+/In3+ –0.45 Sn2+/Sn4+ +0.15 3Br–/Br3– +1.05
Eu2+/Eu3+ –0.43 Cu+/Cu2+ +0.159 Ti+/Ti3+ +1.28
Cr2+/Cr3+ –0.41 [Fe(CN)6]4–/[Fe(CN)6]3– +0.36 Au+/Au3+ +1.29
WCl52–/WCl5– –0.4 3I–/I3– +0.535 Ce3+/Ce4+ +1.44
(in 0.1 M H2SO4)
Ti2+/Ti3+ –0.37 MnO42–/MnO4– +0.54 Mn2+/Mn3+ +1.51
In+/In2+ –0.35 Fe2+/Fe3+ +0.783 Pb2+/Pb4+ +1.69
V2+/V3+ –0.255 OsCl63–/OsCl62– +0.85 Co2+/Co3+ +1.842
For a metal/metal ion electrode where the redox reaction at Eq. (3.10 a) is poten-
tial determining, and the metal of the electrodes is the pure material of activity 1
M z+ + ze − →
←M (3.10 a)
the Nernst equation reads
RT
E = E0 + ln a M z + (3.10 b)
M z + /M zF
or if activity corrections may be neglected ( a M z + = c M z + ) one obtains
RT
E = E0 + ln c M z + (3.10 c)
M z + /M zF
For an element/element-anion electrode as the chlorine electrode with gase-
ous chlorine
1 2X 2 + e − →
←X
−
(3.11 a)
one obtains with neglect of activity coefficients
RT
E = E0 + ln p1X/2 / c X − (3.11 b)
X 2 /X − F 2
z* −
M x O zy− + 2nH + + νee − →
← M x O( y −n) + nH 2O (3.12 a)
c z− ⋅ c H + 2n
RT M x Oy
E = E0 + ln (3.12 b)
M x Oy
z− *
/M x Ozy −−n νe F c *
M x Ozy −−n
RT c CrO42− ⋅c8 +
E = E0 + ln H
CrO42− /Cr 3+ 3F c Cr 3+
(3.13 b)
RT c CrO42− RT
= E 0 2− 3 + + ln − 2.302 ⋅ 8 ⋅ pH
CrO4 /Cr 3F c Cr 3 + 3F
RT c Cr2O72− RT
E = E0 + ln − 2.302 ⋅14 ⋅ pH (3.13 d)
Cr2O72− /Cr 3+ 6F c 2 3+ 3F
Cr
Table 3.2 c lists some more prominent examples of this type of redox cou-
ples. Normal potentials refer in these cases to aH+=1 mol dm–3, that means to
a pH of 0 or if hydroxyl ions are involved in the reaction to aOH- = 1mol dm–3
or pH = 14 respectively.
The equilibrium cell potential for any process conditions may be obtained by
calculating the difference of the two involved single electrode potentials. For this
purpose starting from the normal potentials of Tables 3.2 a–d, single equilibri-
um electrode potentials are calculated by inserting concentrations (activities)
and pressures (fugacities) of dissolved or gaseous reactants as they prevail under
practical operating conditions into the respective Nernst equation.
In the membrane chloroalkali electrolysis, for instance, hydrogen (1 bar) is
evolved at the cathode from a 30 wt% KOH solution, whereas the chlorine anode
evolves chlorine (1 bar) from brine which contains approximately 200 g dm–3
NaCl (3.4 mol dm–3). Omitting in a first approach any activity and fugacity cor-
rections and calculating cNaOH with ρ30 wt% NaOH=1.3 g cm–3 to be cNaOH=6 mol
dm–3 one obtains for the single electrode potentials
ECl = ECl
2
0
2
(
− (RT F)ln cCl − = 1.40 − 0.032 V = +1.368 V ) (3 a)
( )
E H 2 = E 0H − 2.302 ⋅ RT F ⋅ pH = 0 − 0.887 V = −0.887 V
2
(3 b)
soda solution of the catholyte. For technical electrolysis processes which often
use highly concentrated electrolyte solutions, it is important to take into account
that the activity of the solvent, which usually is water, cannot be assumed to be
unity, but is very often considerably decreased.
28 3 Electrochemical Thermodynamics
3.3
Reaction Enthalpy, Reaction Entropy, Thermoneutral Cell Voltage and Heat
Balances of Electrochemical Reactions
Equation (3.3) defines the Gibbs free energy ∆G of any cell reaction as the min-
imal amount of electrical energy per mol of converted substrate which must be
expended in an electrolysis process. This equation differs from the total energy
balance and the heat balance of a given cell reaction which is performed under
equilibrium conditions. The total energy to be converted is the reaction enthalpy
∆H which differs from ∆G by the entropic term T∆S
∆G=∆H–T∆S (3.15)
For any electrolysis process (∆G>O) which involves a positive reaction entro-
py, ∆G is smaller than ∆H and the entropic term T∆S has to be supplied by in-
troducing thermal heat from the environment into the cell, if the reaction is per-
formed reversibly, but at constant temperature.
T∆S=∆H–∆G (3.16)
If ∆S is negative, T∆S must be transferred as heat from the cell to the environ-
ment.
In general all electrolysis processes which involve gas evolution from a liquid
electrolyte have a positive reaction entropy of the order of magnitude of R per
mol of evolved gas.
Table (3.3 a) collects for four important electrolysis processes which involve
gas evolution standard Gibbs enthalpies and reaction enthalpies together with
standard entropies at the given process temperatures.
Figure 3.5 depicts the temperature dependence of ∆G0 and ∆H0 with the dif-
ference representing T∆S0 of the most important water splitting/water forma-
tion reaction. For water formation ∆G0 and ∆H0 are negative for the formation
of water. At 373 K the transition from liquid water to water vapour causes a drop
in ∆H due to the heat of evaporation. Whereas ∆H0 is almost constant from 373–
1100 K, ∆G0 decreases by approximately 20%. Only the quantity ∆G0 can be con-
3.4 Heat Balances of Electrochemical Processes 29
Table 3.3. a Thermodynamic data. b Equilibrium voltage, U0, and thermoneutral cell voltage
Uth of some cell reactions with gas evolution
(a) (b)
Cell reaction ∆G0/kJmol–1 ∆H0/kJ mol–1 ∆S0/Jmol–1K–1 U0/V Uth/V
(H2O)1→H2+1/2O2 100 °C 224 277.5 +159 1.16 1.42
(H2O)g→H2+1/2O2 100 °C 224 243 +50 1.16 1.25
(Al2O3)s→2Al1+3/2O2 1300 K 1263 1688 +326 2.18 2.9
2(Al2O3)diss+3Cs→3CO2+4Al 1300 K 1763 2189 +431 1.89 1.89
Fig. 3.5. ∆G0, ∆H0 and U0 of the water splitting/water formation reaction from 300 to 1100 K
verted into electricity in a fuel cell and must at least be expended in water elec-
trolysis.
Whenever the electrolysis process has to be assumed to be performed under
adiabatic conditions, which is almost realised in large industrial electrolyzers,
the total reaction enthalpy must be provided for by expending electrical energy
so that instead of U0 – the minimal or equilibrium cell voltage – the so called
“thermoneutral” cell voltage Uth or ∆H cell voltage must be applied in order to
balance the input of electrical energy with the enthalpy demand of the cell reac-
tion.
( )
U th = ∆H νeF (3.17)
3.4
Heat Balances of Electrochemical Processes
Under technical conditions electrolysis processes are always performed irrevers-
ibly so that actual cell voltages exceed the equilibrium cell voltages often by far.
30 3 Electrochemical Thermodynamics
Fig. 3.6. Balance of electric power-input, electrochemical energy demand and generated
heat
For a fuel cell process the fuel cell voltage is correspondingly lower than U0. The
heat Q generated in excess per mol of product is calculated by
Qmolar=Ucell ve F – ∆H (3.18)
Since most technical electrolysis processes are performed at elevated temper-
atures, that means in the temperature range between 70 and 90 °C, the evolved
gases which, being by the very nature of their generation fully saturated with wa-
ter vapour, carry away substantial amounts of heat by the respective amount of
latent heat of vaporisation which is eventually removed from the product gas
stream in coolers by condensing of the water vapour.
In chloroalkali electrolysis according to the membrane process the catholyte
is 30 wt% NaOH and the anolyte is brine with 200 g NaCl dm–3. At a process
temperature of 90 °C, the chlorine gas leaving the cell contains approximately
60 mol% of water vapour whereas the evolved hydrogen – due to the diminished
water activity in 30 wt% NaOH solutions – contains approximately only
30 mol% of H2O. Per mol of H2 or Cl2 evolved, approximately 0.9 mol of water
evaporate in total, which with a molar heat of evaporation of 38 kJ mol–1 are able
to remove approximately 34 kJ mol–1 of chlorine from the cell or per two mol of
NaCl converted in the process.
The total energy balance of an operating cell (Fig. 3.6) comprises the follow-
ing terms per unit time:
(i) the reaction enthalpy:
ṅ p ∆H r
1 pH 2O
H 2O,vap ⋅ ∆H vap = I∑
˙
N ⋅ ∆H vap
j νe , jF P − pH 2O
(iv) enthalpies of dissolution or dilution for those cases where pure solvent or
solid or concentrated electrolyte is fed directly into the cell additionally to
the recirculated electrolyte:
˙ ∆H + N ˙
N salt diss H 2O ⋅ ∆H dil
In Eq. (3.19) all ∆H values are molar quantities (kJ/mol–1) whereas the cp-
values of the circulated electrolyte is meant to be a mass specific quantity
(kJkg–1k–1), hence multiplication of v̇ with density and specific heat cp of
the circulating electrolyte.
3.5
Retrieval of Thermodynamic Data and Activity Coefficients
Standard Gibbs free energies, enthalpies and entropies are collected for numer-
ous compounds in JANAF Tables2) and in other collections of thermodynamic
data of, for instance, Baring and Knacke3). Both explicitly mentioned data col-
lections and others are not consistent with each other insofar as they choose dif-
ferent reference states.
2 JANAF tables are setting the standard Gibbs energies, enthalpies and entropies of all ele-
ments in their respective thermodynamically stable modification at any given tempera-
ture to zero. [2]
3 Baring and Knacke are setting Gibbs energies and enthalpies of all elements at the refer-
ence temperature of 298 K to zero whereas entropies are calculated for the elements from
the third law [3]
32 3 Electrochemical Thermodynamics
Since most tables are referring only to pure elements or compounds respec-
tively, concentration or pressure corrections to calculate free enthalpies under
any condition of concentration and pressure must be applied. Neglecting to a
first approximation activity (fugacity) coefficients one obtains for gases:
and for ideal solid solutions and liquid mixtures of non-ionised components
with mol fractions extending from zero to unity, where the standard state is the
pure substance
with xk being the mole fraction of the kth component. Regard that µ 0j and µ 0k are
not identical as they refer to different concentration scales and reference states.
Since only the logarithm of activity and fugacity coefficients enter into the calcu-
lation of more refined data of free enthalpies, the approximate equations which ne-
glect the deviation from ideal behaviour are very often of sufficient accuracy. For
solutions of ionised compounds the situation is more complex because usually sol-
ute/solvent and solute/solute (interionic) interaction are non-negligible.
Molten mixtures of salts (for instance mixed NaCl/KCl melts), however, as far as
they do not exhibit miscibility gaps can often be assumed to behave as almost ideal
mixtures according to Eq. (3.20 c). For more refined calculations the empirical ap-
proximation given in Eq. (3.20 d) is used for the solvent or host melt, index 1, and
for the solute, index 2. α and I are adjusted parameters.
lgγ 1 = α x 2 (3.20 d)
2
lgγ 2 = α x1 +I (3 e)
dm–3 the approximation of Debye and Hückel for single ion activity coefficients
γ± reads
2
lg γ ± = A I1/2 ; I = ionic strength = 1/2 c i z i (3.22)
with c the molar concentration of the ions, and the numerical value of A given by
2
ce0
lnA = (3.22 a)
2.303 ⋅ 2 ⋅ ε ⋅ ε 0 RT
e0 is the elemental charge, and ε, ε0 the dielectric constant and induction coeffi-
cient and c the molar concentration of a 1:1 electrolyte.
Equation (3.22) allows one to calculate activity coefficients of single ions and
mean activity coefficients of the dissolved electrolytes for relatively diluted elec-
trolytes {
γ ± = (γ + ) ⋅ (γ − ) }1/2 for 1.1 electrolytes, and in general
1
( ) ( )
γ ± = γ − ⋅ γ − z + z .
z+ z−
+ −
But this equation fails at concentrations higher than 10–2 molar concentra-
tions and is therefore only of limited value for electrochemical engineers who
usually deal with concentrated electrolyte solution.
Since the electrochemical engineer has in most cases to apply his calculations
to electrolyte concentrations from 1 mol dm–3 upwards where these simplified
approximations based on calculations of the coulombic ion/ion interaction un-
der complete neglect of solute/solvent interactions fail, he has to account for ac-
tivity corrections mainly by evaluating empirical data. For those higher electro-
lyte concentrations it is meaningless to aim at calculating single ion activities
and the experimental determination of total activities and activity coefficients
of dissolved electrolytes should be performed instead. Nearly any method ap-
plied for the investigation of thermodynamic properties of higher electrolyte
concentration like for instance ebulloscopy, cryoscopy, osmometry or vapour
pressure determination determines solvent activities rather than solute activi-
ties and calculates solute activities by integrating the Gibbs–Duhem equation
for isothermal mixing under constant pressure
N i dµ i = 0 (3.23)
which results in
x solv .
x solv
lg γ el = ∫ 1 − x solv
d ln γ solv (3.24)
1
xsolv is the mole fraction of the solvent and γsolv is its activity coefficient which
can be measured relatively easily.
34 3 Electrochemical Thermodynamics
Figure 3.4 a,b is an example for the determination of water vapour pressures
of concentrated electrolyte solutions which allow to determine the activity and
activity coefficients respectively of water in these solutions by making use, for
instance, of Eq. (3.14). Solving Eq. (3.24) by measuring γsolv over an extended
concentration range of the electrolyte. This meets the difficulty that for low
electrolyte concentrations (xsolv → 1,(1–xsolv) → 0) the solvent activity ap-
proaches 1, log γsolv approaches zero and x solv 1− x solv( )
approaches ∞ so
that the integral cannot be evaluated.
Using the Gibbs Duhem equation (Eq. 3.23) in the form
nsolv d ln asolv=nelyte d ln aelyte (3.23 a)
multiplying both sides of the equation with 1000 Msolv n solv, applying concen-
tration definition on the molality scale, cm and making use of the osmotic coef-
ficient, fosm, which is easily measured e.g. by cyroscopy and defines the solvent
activity at low electrolyte concentration one arrives at Eq. (3.25) [4]:
c
fosm − 1
( )
ln γ ± = fosm − 1 + 2∫
1/ 2
dc m (3.25)
0 c1m/2
which can be integrated as the quantity (fosm–1) c1m/2 converges towards a fi-
nite value as c1m/2 approaches zero. The mean ion activity coefficient γ± is defined
by Eq. (3.26):
( γ ± ) z + z = ( γ i , + ) ( γ j, − )
i j zi zj
(3.26)
3.6
Thermodynamics of Electrosorption
The electrosorption of reaction educts, products or reaction intermediates that
means their adsorption on the electrode surface does not bear on the thermody-
namics of an electrochemical process because electrosorption is always an inter-
mediate step of the overall reaction. But electrosorption is important for elec-
trode processes and electrocatalysis and their kinetics. Very often electrochem-
ical reactions demand adsorption of the educt or some reaction intermediate
prior to the decisive, rate determining charge transfer. Therefore as an appendix
to “Electrochemical Thermodynamics” the thermodynamics of electrosorption
is treated in this chapter in a general way.
The thermodynamics of electrosorption of charged particles differ funda-
mentally from that of uncharged particles in so far as for charged particles the
Gibbs energy of adsorption ∆Gad is directly influenced by the interfacial poten-
tial difference at the phase boundary or the electrode potential respectively –
compare Eq. (3.29) below. For uncharged particles the electrode potential influ-
ences the free energy of adsorption more indirectly, as the difference between
the actual electrode potential and the respective potential of zero charge Epzc the
so called rational potential, Er=E–Epzc, determines the magnitude of ∆Gad of un-
charged particles. ∆Gad for uncharged particles frequently changes with the ra-
tional potential according to (E–Epzc)–2. For charged particles one cannot distin-
guish “true adsorption” that means incorporation of the species under consid-
eration into the inner Helmholtz layer from accumulation in the diffuse part of
the double layer, where at least two sheets of solvent molecules separate the re-
spective charged particle from the electrode surface. Uncharged particles are
not subject to this type of accumulation in the diffuse layer. The formal thermo-
dynamic description of electrosorptive processes as well as that of simple ad-
36 3 Electrochemical Thermodynamics
sorption also has to take into account that the free energy of adsorption usually
is influenced by the mutual interaction of the adsorbed particles, that means that
the degree of coverage in contrast to the simple model of the Langmuir isotherm
generally alters the adsorption enthalpy.
The Frumkin model postulates a linear dependence of the adsorption en-
thalpy ∆Had on the degree of coverage, Θ
∆Had=∆Had+ßΘ (3.27)
For charged as well as for uncharged particles “true” electrosorption in the
meaning of contact-adsorption or adsorption in the inner and/or outer Helm-
holtz layer is often described by the generalised formula (Langmuir isotherm)
which holds too for adsorption on solid surface from the gas phase.
Kc
Θ= with ∆G ad = −RT ln K (3.28)
1 + Kc
K is the adsorption equilibrium coefficient which depends on the electrode po-
tential E and the coverage; c is the concentration of the adsorbate in the bulk of
the electrolyte.
Correspondingly one obtains Eq. (3.29 a) for the adsorption equilibrium co-
efficient for charged particles:
∆µ 0 − zFη
K charged = K 0 exp− ad (3.29 a)
RT
For charged particles with charge ze and Frumkin dependence of ∆Had this
equilibrium constant reads
∆µ 0 − βθ − zFη
K charged = K 0 exp− ad (3.29 b)
RT
For uncharged particles the last term vanishes whereas ∆µ0ad very often con-
tains implicitly the interfacial potential as ∆µ0ad depends on the rational poten-
tial Er=(E–Epzc). The standard free enthalpy of adsorption ∆µ0ad of uncharged
species is often found to vary according to Eq. (3.30):
A
µ 0ad ,unch arg ed = (3.30)
( )
2
E − E pzc
References
1. D. Dobos, Electrochemical Data, Elsevier, Amsterdam, Oxford, New York, 1975
2. JANAF Thermochemical Tables, 2nd Edition, D.R. Stull. H. Prophet eds., Ntl. Bur. Stand.
(U.S.) NSRDS - NBS 37, 1971
3. I. Barin, O. Knacke, Thermochemical Properties of Inorganic Substances, Springer-Ver-
lag/Verlag Stahleisen, Berlin/Düsseldorf, 1973
4. G. Kortüm, Treatise on Electrochemistry, Elsevier Publishing Comp., Amsterdam, Lon-
don, New York, 1965, p. 169
http://www.springer.com/978-3-540-64386-9