CHEM465/865, 2004-3, Lecture 14-16, 20th Oct.

, 2004

Electrode Reactions and Kinetics

From equilibrium to non-equilibrium: We have a cell, two individual

electrode compartments, given composition, molar reaction Gibbs free
energy as the driving force, ∆ r G , determining EMF Ecell (open circuit

potential) between anode and cathode!

In other words: everything is ready to let the current flow! – Make the
external connection through wires.

Current is NOT determined by equilibrium thermodynamics (i.e. by the

composition of the reactant mixture, i.e. the reaction quotient Q), but it is
controlled by resistances, activation barriers, etc

Equilibrium is a limiting case – any kinetic model must give the correct
equilibrium expressions).

All the involved phenomena are generically termed

ELECTROCHEMICAL KINETICS!
 This involves:
Bulk diffusion
Ion migration
(Ohmic
resistance)
Adsorption
Charge transfer
Desorption

Let’s focus on charge transfer! Also called: Faradaic process.

The only reaction directly affected by potential!

An electrode reaction differs from ordinary chemical reactions in that at

least one partial reaction must be a charge transfer reaction – against
potential-controlled activation energy, from one phase to another, across
the electrical double layer. The reaction rate depends on the distributions
of species (concentrations and pressures), temperature and electrode
potential E.

Assumption used in the following: the electrode material itself (metal) is

inert, i.e. a catalyst. It is not undergoing any chemical transformation. It is
only a container of electrons.

The general question on which we will focus in the following is:

How does reaction rate depend on potential?
The electrode potential E of an electrode through which a current flows
differs from the equilibrium potential Eeq established when no current
flows.

The difference between these potentials is called overvoltage:

η = E − E eq > 0 (anodic current)

η = E − E eq < 0 (cathodic current)
Convention:
Current associated with reaction

Red Ox + e- (Me) (anodic reaction)

e.g. Had , Fe2+ e.g. H+ , Fe3+
anodic current, positive direction

anodic direction cathodic direction

e- e-
R R
O O
ϕM ϕS ϕM ϕS

Overall rate of electrochemical reaction (per unit electrode surface area,

units of [mol/(cm2s)]):
net rate = rate of oxidation – rate of reduction

vnet = K ox cred
s
− K red cox
s
(*)
 S S
cred , cox : surface concentrations of reduced and oxidized species

This definition of the net reaction rate is the same in anode and cathode.
Both partial reactions take place at both electrodes. However, in the anode
the anodic reaction dominates and in the cathode the cathodic reaction
dominates.
Remember: Boths processes occur at both electrodes!
anode: vnet > 0 , positive net rate

cathode: vnet < 0 , negative net rate

General observation, e.g. Arrhenius (T-dependence of reation rate):

Charge transfer reactions are kinetically hindered!

Since species is charged: strong dependence of activation energy on

electrode potential

E = ϕM −ϕS

Discuss: where to take the value of the solution potential ϕ S ????

close to metal surface, at position of compact layer
role of supporting electrolyte (problem sets!)
What do you actually control by applying a potential to an electrode?
The highest occupied electron level in the electrode is the Fermi-level.
Electrons are always transferred to or from this level.

Oxidation (anodic) Reduction (cathodic)

EF
E smaller
e-
EF
e- Eredox Eredox
EF
E larger
EF

larger potential (E or ϕM larger) smaller potential (E or ϕM smaller)

favors oxidation favors reduction

The energy of the redox couple in solution is not affected by the potential
variation (no double layer corrections). Under ideal conditions, a shift in
electrode potential only affects the position of the Fermi-level.
In order to understand, how the reaction proceeds in detail, we need to
determine the rate constants in Eq.(*): Use absolute rate theory!

 ∆Gox
†
(E) 
K ox = A exp  − 
 RT 
 ∆Gred
†
(E) 
K red = A exp  − 
 RT 
where ∆Gox
†
( E ) and ∆Gox
†
( E ) are molar Gibbs free energies of activation

depend on potential!

Which tendency do you expect upon variation of potential? Let’s say the
value of potential E is increased. Does the rate of the anodic reaction
increase or decrease? What about the cathodic reaction?
Let’s consider the free energy profile:

G activated
complex

reduction

E increases
oxidation

electron in metal, electron in solution, reaction

oxidized state reduced state coordinate

Charge transfer between reduced and oxidized state always proceeds via
an activated intermediate complex. The height of the potential barrier
depends on the electrode potential E. In the following pictures, the
potential is shifted from the equilibrium value (black curves) to a more
positive potential (blue curve). Note: only the state corresponding to the
electrons in the metal is affected by the potential shift, as discussed earlier.
An increase in E favors the anodic direction (oxidation) and disfavors the
cathodic direction (reduction).
Anodic direction
e-
G

−α F ( E − E 0 )

∆Gox
† ∆G †,0

0 0
Gox Gred
−F ( E − E 0 )

∆Gox
†
= ∆G †,0 − α F E − E 0 ( )
electron in metal, electron in solution, reaction
oxidized state reduced state coordinate

Cathodic direction
e-
G

∆G †,0

0 0
Gox Gred
−F ( E − E ) ∆Gred
0 †

∆Gred
†
(
= ∆G †,0 + (1 − α ) F E − E 0 )
electron in metal, electron in solution, reaction
oxidized state reduced state coordinate
Now, let’s expand ∆Gox
†
( E ) about the standard equilibrium potential E 0 .

Recall: E 0 is a characteristic value of the redox pair that we are

considering (at one particular electrode). At this potential, no net current
would flow in this electrode under standard conditions. The equilibrium
potential E eq and the standard equilibrium potential E 0 are related by
RT
E eq = E 0 − ln Q (Nernst-equation) .
νe F
Anodic reaction (in one electrode compartment):

∆Gox
†
( E ) = ∆Gox† ( E 0 ) − α F ( E − E 0 ) (#)
anodic transfer coefficient:

1 ∂∆Gox
†
α=− > 0 ( dimensionless )
F ∂E E 0

Cathodic reaction (in the same electrode compartment):

∆Gred
†
( E ) = ∆Gred
†
( E 0 ) + β F ( E − E 0 ) (##)
cathodic transfer coefficient:

1 ∂∆Gred
†
β= > 0 ( dimensionless )
F ∂E E 0
Now, the trends of changing the potential E are obvious:
Larger E ∆Gox
†
smaller anionic reaction faster, in other words:

at larger electrode potential, electrons are more easily transferred

from the solution to the metal

Smaller E ∆Gred
†
smaller cathodic reaction faster, in other words:

at larger electrode potential, electrons are more easily transferred

from the metal to the solution

In going from E 0 to E > E 0 the Gibbs free energy of electrons in the metal
is lowered which makes electron transfer to the metal more likely.

Note: Deviations from the linear approximations in Eqs.(#) and (##) arise at
RT
large values of E − E 0
F
Particular relation:
∆Gox
†
( )
E 0 = ∆Gred
†
( )
E 0 = ∆G †,0

Consider a case with a high concentration of an inert electrolyte (recall

your calculation involving a supporting electrolyte! – what was the effect
that you found!?):
this inert (i.e. non-reacting) electrolyte screens the electrode potential
there will be practically no potential drop in the diffuse layer on the
solution side
the electrostatic potential of reactants in solution is unchanged, when
the electrode potential is varied, a variation in electrode potential is equal
to a variation in the metal potential, ∆E = ∆ϕ M
in this case, the solution potential serves as a constant reference

∆Gox
†
( E ) − ∆Gred
†
( E ) = Gox − Gred = − F ( E − E 0 ) and α + β = 1
This simple relation between the Gibbs free energy of the charge transfer
and potential E is fulfilled, if the entire potential drop occurs in the compact
layer and reactants are not specifically adsorbed. The reaction is a so-
called “outer-sphere reaction”.
The observed current density, in units of [A/cm2], of the electrode reaction
can be expressed as:

j = Fν net = F  K ox cred
s
− K red cox
s


Hence, the observed net current can be split up in an oxidation current jox
and a reduction current jred:
jox = FK ox cred
s
and jred = FK redcox
s

Using the linear approximations for ∆Gox

†
(E) and ∆Gred
†
(E) in the

expressions for the rate constants we can write:

 ∆Gox
†
(E)   α F ( E − E 0 ) 
K ox = A exp  −  = k0 exp  
 RT  RT
 
 ∆Gred
†
(E)  (1 − α )F ( E − E 0 ) 
K red = A exp  −  = k0 exp  − 
 RT   RT 

 ∆G †,0 
where k0 = A exp  −  is the rate constant (in units of [cm/s]). It is a
 RT 
measure of the reaction rate at E0and where we used β = 1−α

Accordingly, we can now write an equation relating the electrode potential

E to the observed current density for the electrode reaction as:

  αF ( E − E0 )   ( 1 − α ) F ( E − E 0 )  
(BV 1) j = Fk  cred exp 
0 s

 − cox exp  −

s



 RT RT
   

This equation is widely known as the BUTLER-VOLMER equation.

However, in general, the deviation of the electrode potential E from the

situation of zero net current, i.e. at Eeq, under any experimental conditions
is considered. Hence, we have to use the Nernst equation to determine the
equilibrium potential Eeq for a given composition in the bulk to establish
the link between composition in the bulk and equilibrium
b
RT cox
E eq
=E + 0
ln b
F cred
Using this relation in (BV 1) and recalling the definition of overpotential,
η = E − E eq , we can rewrite after some rearrangement (exercise!!) the
Butler-Volmer equation in a different form, as a relation between current
density and overpotential:

0 b ( 1− α ) b α
 cred
s
 α Fη  cox
s
 ( 1 − α ) Fη  
(BV 2) j = Fk c red c ox  b exp  −
 b exp − 
c
 red  RT  cox  RT  

compact Note, that we explicitly distinguish surface

layer concentrations from bulk concentrations.
Bulk concentrations (far away from the
S b
c c ox,red interface) determine the equilibrium potential.
ox,red
Under equilibrium conditions, surface
s
concentrations, cred/ox (at the position of the
b
compact layer), are equal to bulk concentrations, cred/ox , and no fluxes of

species exist. At finite current, surface concentrations will generally be

different from bulk concentrations, due to mass transport from bulk to
surface.
In the particular case, when mass transport from bulk to surface is fast
enough (rate of diffusion much larger than rate of reaction, see section on
s
mass transport), we have cred = cred
b s
and cox = cox
b
. In this case, we can write

the simplest form of the Butler-Volmer equation (historically the first

version),

  α Fη   ( 1 − α ) Fη  
(BV 3) j = j 0 exp   − exp  − 
  RT   RT  

Now, we have three different ways to write the Butler-Volmer equation. You
will find all the different forms used throughout electrochemical literature.
Which one is most appropriate? (BV 3) is the least general form. It can only
be used, when all sorts of mass transport limitations can be neglected and
the current-potential relation is only determined by the kinetics of charge
transfer. Whenever you see people using (BV 3) in a situation with
significant mass transport limitations, something is screwed.

(BV 2) can be used for simple electrode structures, when the bulk
concentrations are controlled.

(BV 1) is the most general form. It explicitly displays all dependencies on

concentrations and electrode potential. All other quantities in it are
constants.

The factor in front of the brackets (pre-exponential factor) on the right hand
side of (BV 2) and (BV 3)
b ( 1 −α ) b α
j 0 = Fk 0 cred c ox

is called exchange current density and from the last equation, it can be
seen that it is concentration dependent. It is a measure of the current at
zero overpotential, when no net current is flowing.
One can also define a standard exchange current density as:

j 0 = j 00 ( cred ) (c )
(1−α )
s α
j 00 = Fk 0 and hence s
ox . Usually, these values are

reported in the literature (unless specific concentrations are explicitly

specified). Values of exchange current densities can widely differ for
different reactions. In most cases, they are known only with low accuracy
(within an order of magnitude). A few values are listed in the following:
O2 + 4 H + + 4e − 2 H 2O , j 00 ~ 10−10 A cm-2
Fe ( CN ) 6 + e − Fe ( CN ) 6 ,
3− 4−
j 00 ~ 10 A cm-2
Cl 2 + 2e − 2Cl − , j 00 ~ 10−1 A cm-2
2 H + + 2e − H2 , j 00 ~ 10−4 A cm-2
Now, let us explore the meaning of the different parameters of the Butler-
Volmer equation:
1) The exchange current density represents the current density, which
flows in each direction, anodic and cathodic, when the overall
current is zero, i.e. in the electrochemical equilibrium. Hence, at Eeq
we have a dynamic equilibrium where both oxidation and reduction
take place at the same rate, given by j 0 .

jred jred jred

Metal

jox jox jox

-
e flux
E = Eeq ; η = 0, j = 0 E < Eeq ; η < 0 ; j < 0 E < Eeq ; η < 0 ; j > 0

At positive overpotentials, the oxidation current prevails whereas at

negative overpotentials, the reduction rate is higher. As a pre-
exponential factor, the exchange current density largely determines
the overall kinetics:
a) j 0 large: Fast electrode kinetics. High current densities are
already reached at small overpotentials.
b) j 0 small: Sluggish electrode kinetics. A large overpotential is
needed to obtain a significant current density.

2) The transfer coefficient α is often termed symmetry factor because it

determines the dependence of the current density on overpotential.
a) If α = 0.5: The increase in the absolute value of j, |j|, is equal
for the oxidation as well as for the reduction current density.
The current density-overpotential graph is point-symmetric
about (E = Eeq, j = 0).
 b) If α > 0.5: |j| increases faster for oxidation than for reduction
η| .
with the same |η
c) If α < 0.5: |j| increases slower for oxidation than for reduction
η| .
with the same |η

The following picture shows free energy profiles for different values
of α.
α ¡ Ö1 α¡Ö
1/2 α ¡ Ö0

ox
red

red ox red ox

usually (for metals) α ≈ 1/2 → symmetric electron transfer!

activated complex has predominantly structure of oxidized state:
α≈1 anodic reaction strongly dependent on E
activated complex has predominantly structure of reduced state:
α≈0 anodic reaction weakly dependent on E

RT
3) Small overpotentials η :
F
Under these conditions mass transport limitations are negligible. We
can use (BV 3). Both exponential terms are important. We can
linearize the exponential terms in (BV 3) to get
Fη j0 F
j ≈ j0 = Rctη where Rct =
RT RT
A linear current-voltage dependence in the form of Ohm’s law is
found. Rct is the charge transfer resistance. We can use the low

overpotential regime to determine the exchange current density!

 RT
η|, η
4) For large values of |η (the so-called Tafel-regime), one of
F
the partial currents dominates so that the other one can be neglected
in (BV 3). We still assume, that mass transport limitations are
neglibible.
RT
a) η and η > 0 ⇒ j ≈ jox and jred ≈ 0 :
F
αF
log10 ( j ) = log10 ( j0 ) + η
2.3 RT
RT
b) η and η < 0 ⇒ j ≈ jred and jox ≈ 0 :
F
(1 − α )F
log10 ( j ) = log10 ( j0 ) − η
2.3 RT

We can draw so called Tafel-plots by plotting log|j| vs η. From such a

plot, we can extract the transfer coefficient α from the slope of a fit of
η| >> 0 as well as the
the experimental data to a straight line at |η
exchange current density from the crossing point with the ordinate
of this linear fit.
 log|j|
(1 − α ) F
Slope = −
2.3RT αF
Slope =
2.3RT

j0

5) The inverse of the oxidation current slope in a Tafel plot is usually

called the Tafel-constant and is commonly denoted as b.
2.3RT
b=
αF
6) At very large absolute overpotentials, the net current density reaches
a limiting value because the reaction rates become limited by
diffusion of reactant to the reaction zone – see later lectures.

7) Most charge transfer reactions are in fact inner-sphere reactions and

the Butler-Volmer equation does in general not apply anymore.
However, it could be observed that the current density still depends
on overpotential within certain ranges of η . It is convenient to retain
the form of the Butler-Volmer equation and define empirical transfer
coefficients α and β:

 α Fη   β Fη 
K ox = k 0 exp   and K red = k exp  −
0

 RT   RT 
However, now, α and β generally depend on electrode potential and
temperature. Also, α + β ≠ 1.