6.3 Conclusion 117
This is due to a larger fraction of undercoordinated sites as the particle size
decrease 241. For Au it has further been shown that the (211) surface exhibits
stronger binding to O, OH and OOH, which would also be consistent with
earlier oxidation of the Au atoms on the surface 133. However, a more systematic
study is needed, which due to time constraints have not been part of the
thesis work. As future work, it is expected that investigations into size eects
could result in a more detailed model. Such a study could encompass a combination
of DFT calculations on Au nanoparticles and experimental characterization
of a larger set of thin lms with varied gold domain sizes.
Another possibility that has not been discussed so far is whether the increased
gold content could signicantly increase electronic conductivity of the
thin lms. At this stage experiments have not been carried out to pursue this
option. A way to nd out would be to investigate thinner lms where conductivity
is not expected to play an important role or alternatively follow a procedure
similar to the Ti-MnO2 lms reported in chapter 5, where only the top 5 nm
is modied. It should also be noted that the benecial interaction between Au
and Mn has been reported for other systems where conductivity enhancements
are less likely to play a role. As example, MnOx nanoparticles deposited onto
Au nanoparticles exhibited great enhancement in activity for OER 225.
6.3 Conclusion
To sum up, this chapter encompassed interactions between gold and manganese
oxide from both a theoretical and experimental point of view. The theoretical
model was briey introduced and was based on a bifunctional surface, where
active Mn sites could benet from neighbouring proton accepting Au sites. The
proton acceptor scheme is benecial due to a selective stabilization of the OOH
binding, which is normally too weak on Mn oxides. It was shown, from a thermodynamic
analysis, that including the proton acceptor pathway results in a
signicant lowering of the predicted overpotential, with 100 mV decrease for
MnO2. From the experimental studies on mixed lms, an enhancement of up to
5.5 times the activity of pure Mn3O4 could be reported. This activity enhancement
was measured for a high concentration of Au, 50 %, which consisted of 3
nm Au domains dispersed in the Mn oxide lms. For a lower concentration, 30
%, the activity enhancement was more modest at approximately 2 times the activity
of Mn3O4. The activity enhancements could not be explained by higher
porosity or increase in surface area judging from capacitance measurements.
The lms were also characterized with XPS, GA-XRD and SEM for structural
and compositional analyses. Furthermore an in-situ study of the Mn K-edge
revealed that Mn atoms oxidise at an earlier potential when Au is nearby, which
is in contrast to previously reported ex-situ studies in the literature 225, 226.
The measurements on Au L3-edge at the same time showed that the Au domain
size had an impact on the Au oxidation. For the lower Au concentration, with