
6.1 Theoretical model of Au-MnOx interaction 101
Figure 6.3: Free energy diagrams for the oxygen evolution reaction on Mn2O3,
to the left, and MnO2, to the right. Left: In green no H-transfer is assumed which
results in a high energy for OOH compared to O. In red the H-transfer is included
which results in a lowering of the OOH step. In blue the results are shown for proton
transfer to a nearby Au site, which indicates a further lowering of the OOH step.
Right: In green no H-transfer is included and the OOH step is high in free energy.
In red the H transfer is included, in purple a nearby Au NP is considered to facilitate
H-transfer and in blue an incorporated Au atom is considered for H-transfer. Note
that the red, purple and blue lines coincide and all lead to a siginicant lowering of
the OOH binding energy. For both diagrams the dotted line indicate the ideal OER
catalyst, where all steps require 1.23 V to become at.
those potentials 131. For MnO2 the proton transfer mechanism could in theory
be possible on two neighbouring Mn sites, however, since both sites are active
for OER, it is unlikely that one of the sites will be ecient as continuous proton
acceptor site. Instead, it is more likely to happen with a nearby Au site, which is
inactive for water oxidation at low overpotentials. At a potential of around 1.4
VRHE it is thermodynamically favourable to transfer and remove protons to such
a site. The predicted overpotentials for the two Mn oxides together with similar
analysis for Co oxides are shown in gure 6.4 as a function of the descriptor
GO GOH. From this model materials on the left side of the peak, at
1.6 eV, are predicted to benet from the proton transfer mechanism, whereas
materials on the right side are not. On the right side the potential determining
step is the oxidation of water to form O, meaning that a stabilization of OOH
has minimum impact on the activity.
In summary, these DFT calculations provide a possible explanation for the
activity enhancement. In terms of absolute decrease in overpotential it is di-
cult to directly compare to experimentally measured values. This is due to an
unknown fraction of Mn sites with Au neighbours as well as an often unknown
roughness factor. On the other hand the DFT calculations are not accurate
enough to fully predict absolute values of overpotential, although the trends
described are expected to be robust.