72 Benchmarking the stability of OER catalysts
the two methods and the overall trend is the same.
Figure 4.7: Comparison between EQCM (in blue) and ICP-MS (in green) results
for the mass losses of MnOx lms. The two methods are compared with chronoamperometry
(CA) at 1.8 and 1.9 VRHE and chronopotentiometry (CP) at 20 mA/cm2. The
error bars indicate 1 standard deviation based on four independent measurements.
With the corrosion rate established as function of potential and time it is interesting
to analyse the magnitude of current density that it represents. If we
assume that the loss of MnOx at the surface proceeds as the following reaction
138.
MnO2 + 2H2O ! MnO��
4 + 4H+ + 3e�� (4.2)
with three electrons transferred per Mn atom dissolved, a current density can
be calculated from the dissolution rate. The dissolution rate is 1128 ng/cm2
over two hours at 1.9 VRHE based on the EQCM results, which corresponds
to 0.5 A/cm2. The dissolution current is therefore more than four orders
of magnitude lower than the total measured current density for the electrode.
Such small currents are challenging to measure accurately with electrochemical
methods. As an example the Faradaic eciency towards oxygen evolution can be
measured with rotating ring disk electrode, RRDE, systems 109,182. However,
slow processes such as the anodic MnO��
4 formation would be extremely dicult
to identify since the rate corresponds to an eciency of less than 0.001 %. In
fact, even if the entire lm of 40 nm dissolved in one hour the dissolution current
would only be around 20 A/cm2 and the Faradaic eciency would be 0.07 %
which would be close to the accuracy limit of RRDE mesaurements. In table
4.1 the mass loss rates can be seen.
The dissolution rates can also be used to predict a lifetime for the thin lms.
At 1.9 VRHE MnOx lost 1128 ng/cm2 over two hours and the lifetime of 40 nm