Draft version
on a single feature. In grey, values for known references have been included for comparison. The error bar are based on varying
the integral range in the first moment equation.
In Figure 4a a slight change in the Mn K-edge can be observed for the gradual increase in potential. at more
positive potentials, the overall edge shifts towards higher energies, which is consistent with a small degree of
oxidation of the Mn atoms, as expected from applying an oxidizing potential. For the samples with Au
incorporated, Figures 4b and c, the same overall trend can be seen; at increasing potentials the edge shifts
towards higher energies. However, the change is significantly more drastic for the mixed films, compared to
the pure MnOx. Two specific features should be noted. The shoulders visible around 6548 and 6552,5 eV for
the initial measurements disappear for the films with Au, while it is still visible for pure MnOx, even at 1.65 V.
These two shoulders are normally seen for Mn oxides in lower oxidation states, Mn2+.44,45 Another feature is
the pre-edge located around 6540.5 eV for the dry measurements. For the pure MnOx there is little to no
change in the pre-edge, whereas for the mixed films a very clear splitting into two peaks is seen for both
concentrations after around 1 V. Farges attributed this pre-edge splitting, which he observed on several
different types of Mn compounds, to oxidation states of 3+ or higher.45 However, Mn compounds with the
same oxidation state can exhibit very different K-edge features dependent on the structure and due to the fact
that mixed oxidation states can be present at the same time.46 To account for such a possibility, we analyze the
overall shift of the edge using a moment method where we calculate the weighted integral of the spectrum.47
The results of that analysis are shown in Figure 4d, where data for each of the films are included together with
literature values for two known Mn oxidation states. Consistent with the Farges’ interpretations described
above, this analysis shows that the incorporation of Au leads to a significant shift of the Mn K-edge towards
higher energies. Even at the open circuit conditions a clear difference is observed between the pure MnOx and
the mixed films. At the highest potential, 1.65 V, the mixed films have a Mn K-edge shifted 1 eV higher than the
pure Mn3O4. It is unlikely that the Mn exists in a single oxidation state based on the data presented here. We
emphasize, that since the Mn K-edge energies are in the hard X-ray regime the entire bulk of the film is
contributing to the signal, together with the surface atoms. We expect the electrochemically active Mn atoms
to be more oxidized when the potential is increased, than the bulk atoms.
The Au LIII-edge was also measured for the samples containing gold, and the resulting spectra are shown in
figure 5a and b. At OCV, figure 5a, the spectra for Au(30%)-MnOx and Au(50%)-MnOx are almost identical.
However, at 1.65 VRHE there is a distinct in the white line for the lower Au concentration. A higher white line
indicates that the Au is oxidized.48,49 From the Pourbaix diagram of gold the oxidation to a +3 state can occur at
potentials close to 1.46 VRHE so it is not a surprising feature. Nevertheless it is a feature that is missing for the
9