CHEY0016 Lecture 15
Corrosion Science and the application of the Butler-Volmer
equation to measure the rate of corrosion
Corrosion
is one of the most important examples of surface chemistry. Corrosion is
thought to cost around 3.5 % of the G.D.P. of industrialised societies and is
thought to cost the US Air-Force more than $1 bn per year. Corrosion is an
interesting phenomenon, because it is an example of surface chemistry, with
electrochemical mechanisms. Indeed, without understanding electrochemistry and
especially electrode kinetics it is not really possible to understand
corrosion.
Basic concepts
Corrosion
involves (using a rather specific definition) the oxidation of a metal at a
surface. Since an oxidation process must necessarily be supplied with electrons,
in most corrosion
situations a second reduction process takes place in the vicinity of the
oxidation site.
For example:
It is important to remember that both anodic and cathodic (oxidative and
reductive) processes must be present for corrosion to take place. This fact
though allows for control of corrosion, by suppression of one or both of the
processes.
An
overly simplistic picture might look like this:
An example is crevice corrosion, where the separate anodic
and cathodic sites are clearly identified and result in morphological change of
the surface.
Naturally, corrosion processes, like everything else are controlled both by thermodynamics and kinetics. Both are important, since thermodynamics tells us the fundamental feasibility of the reaction, kinetics the rate.
A brief revision of 1st year thermodynamics:
The standard free energy, DGº tells us the thermodynamic feasibility of a reaction. Since
the standard electrode potential of a reaction, E° and DG° are intimately linked; one can use the standard
electrode potentials of a reaction to calculate the thermodynamic feasibility
of a reaction.
DGo
= -nFEo
For the example of iron corroding in deoxygenated acid solution:
Fe + 2H+ → Fe2+
+ H2
Standard
reduction potentials (from Atkins)
So, we have a
Galvanic cell and can write the above in terms of an overall cell reaction. The
convention is to place the oxidising system on the Left Hand Side (LHS) and the
system being reduced on the RHS:
Fe | Fe2+
|| 2H+ | H2
The cell potential is
the RHS – LHS or more formally: E°
(cell) = E° (RHS) - E° (LHS).
So for Fe corroding
in 1M acid can write:
E° = 0.00 – -0.44
E° = +0.44 V.
So, this reaction is
thermodynamically feasible since:
DG = -nFE°
To correct for
different concentrations of acid, use the Nernst equation.
Moreover, it can be
seen that the reduction of oxygen to water also forms a convenient half
reaction with iron oxidation:
O2 + 4e-
+ 4H+ ® 2H2O
It can already' be seen that the
reduction of H+ to H2 is a thermodynamically feasible way to allow the
oxidation of Fe to take place. This also implies that (especially in deoxygenated solutions) the concentration of
H+ is very important.
Since pH = -log [H+], then:
using the Nernst equation one can write:
So, E for the hydrogen evolution (H+ reduction) changes by 59 mV for every change in pH unit.
The Pourbaix diagram is commonly used by corrosion engineers to
determine the pH and potential where a metal will either be stable to
corrosion, will corrode or form a passivating layer. Note, this is a very
simplified diagram, with
Looking at the above diagram, it can be seen that at applied potentials
more negative than -0.6 V vs. SHE (Standard Hydrogen Electrode), Fe will not
corrode (at all but very high pH's). At potentials more positive than -0.6 V Fe
will corrode, but to different products depending on pH, however at pH's> 4,
Fe will form a blocking passive layer of Fe203.
This diagram is important, since it shows that by controlling the
applied potential on a piece of metal, and or the pH of the immersing solution
one can control the corrosion rate. This leads to two methods of protecting
metal from corrosion known as Cathodic protection and Anodic
protection.
Cathodic Protection
This involves applying a negative potential to the
metal to be protected, so taking the metal into the immunity region on the
Pourbaix diagram. This can be done via an external electrical circuit, or by
attaching a sacrificial anode to the metal, which corrodes more readily,
donating its electrons to the host metal hence reducing its potential. This
method is commonly used
on pipelines.
Zinc galvanisation is a good example of cathodic protection, where the
Zn layer as well as forming a protective coating forms a sacrificial
anode which cathodically protects the underlying metal.
Anodic Protection
This involves applying a positive potential to a piece of metal, to form a passivating layer. This is risky, since if the potential and pH are not finely controlled, the corrosion rate can actually increase (see Pourbaix diagram).
Since the kinetics of corrosion essentially the
kinetics of electron transfer, the methodology and arguments discussed in
lecture 14 can be used.
In corrosion studies, it is important to be able to
accurately calculate the rate of corrosion. The corrosion rate is directly
proportional to the exchange current density at zero applied potentials.
In corrosion studies, the exchange current density, io
is referred to as the corrosion current:
io = icorr µ k
where k is the corrosion rate.
Hence, a determination of icorr is of fundamental importance.
Fortunately, icorr can be quite easily determined (in most cases).
If we treat a corroding
piece of metal as an electrochemical system under electrode kinetic control, one can apply the Butler- Volmer equation to the system, which was derived from first principals in last lecture.
This equation
can be simplified as shown below to:
A corroding system has both anodic and cathodic processes taking place,
with the electrons being liberated by the oxidation process being used up in
the reduction process. Hence, one would expect the value of icorr in both the
anodic and cathodic reactions to be identical.
Experimental
determination of icorr
Experimentally, the easiest way to determine icorr is to
connect the corroding system to a potentiostat via a 3 electrode cell. The
corroding sample forms the working electrode. The system is polarised, that is
both anodic and cathodic over-potentials (h) applied
to the sample by sweeping the applied potential and the current measured. This
is essentially a linear
sweep voltammetry experiment, but potential sweep should be slow to prevent
mass transfer effects.
Clearly then for both anodic and cathodic half reactions, by plotting
log i vs. h will yield a
straight line, with intercept at h = 0 giving log icorr. Moreover, both
straight lines will intercept at h = O. In fact, one has to extrapolate the linear part
of the B-V equation to the h= 0 case, since at very low
applied potentials the exponential relationship between h and i breaks down.
In the following
example, a steel nail was placed in 0.1 M HCI and the corrosion rate determined
by sweeping a potential each side of the open circuit potential (-0.495 V vs.
Ag/AgCI) at 5 mV s-1.
graph shows how Tafel analysis can
be used to determine corrosion rate:
So, by extrapolation
of the h vs. log Iii
plot, the corrosion current could be determined:
log icorr =
-3.13
icorr = 7.41 x 10-4
A.
Now, 1 A = 1 coulomb
s-1.
And 1 F = 96485 C,
moreover, F = NAe (NA = Avagadro constant), e = charge on
electron. That is, 1 F is 1 mole of charge.
So, a current of 7.41
x 10-4 A =
7.41 x 10-4 / 96485 =
7.68 x 10-9 moles of Fe corroding per second for this screw of surface area x.
Note, for simplicity, we have largely ignored the surface
area, but this is clearly of crucial importance when measuring the relative
corrosion rate of different metals, or effects of different environments.
Ideally, one should quote the corrosion current in units of A cm-2. In this
case, it was a bit difficult to measure the surface area of the screw.
Molecular weight of
Fe = 56 g mol-1, so corrosion rate of Fe in 0.1 M HCI = 0.037 g / day.
So, if this rate of
corrosion is linear with time (which is most unlikely since the surface area will
change with time) and no passive layer is formed, then a screw weighing 5 g
will
entirely corrode in 135 days.
This calculation is
somewhat approximate, but hopefully shows how engineers when designing
structures can use electrochemical measurements and how important the problem
of corrosion is. Obviously, when designing an oil rig, a ship or a nuclear
reactor such calculations become even more important.
The study of
corrosion is a good example of how the analytical methods described in this
lecture course can be combined to build up a complete picture of a process. For
example, corrosion in the stainless steel tanks used to hold radioactive waste
at Sellafield, Cumbria is a big problem. Research at Sellafield into this
problem combines electrochemical methods as show here with surface infra-red,
AFM, STM, XPS and other methods.
Check out: http://www.hse.gov.uk/nsd/ilrws1.htm
;
for more on corrosion
problems in radioactive waste storage.
Further reading
Electrochemistry, Principles,
Methods and Applications by Brett and Brett has a chapter on corrosion.
Physical Chemistry by Atkins talks about it, but in a
confusing way.
Corrosion and Environmental
Degradation edited
by Michael Schotze is very comprehensive.