7.2 Outlook 121
trolyzers.
The strategy for stabilizing the MnO2 surface is still in its infancy and several
directions for optimization can be undertaken. It would be of great importance
to establish and observe how the Ti atoms are located on the MnO2 surface.
Such a characterization could be done with cross-sectional Transmission Electron
Microscopy imaging. Atomic resolution of the mixed Ti-MnO2 layer could
provide a lot of insight. It would be challenging to distinguish Ti from Mn,
especially if the Ti is located as single atoms on the surface. Another possible
route is to image the surface with Scanning Tunnelling Microscopy under UHV
while depositing small quantities of Ti onto MnO2. This requires very at and
well dened surfaces which are dicult to prepare using the sputter deposition
technique. It may instead be possible with annealing treatments for a Mn single
crystal and slow evaporation of Ti. Given that deposition and annealing parameters
can be utilized to block all the undercoordinated sites of MnO2 with
Ti, it would be highly relevant to further develop the concept. For commercial
electrolyzers the catalyst material must be applied with a very high surface
area. Therefore, it would be important to prepare nanoparticulate Ti-MnO2
with high activity on a geometric basis combined with optimal long term stability.
Finally, it is possible that other stability promoting materials could be even
more suited for blocking the undercoordinated sites. Tantalum, niobium and
zirconium are among the most stable materials in acidic environment and the
reason they have not been included here is that they do not form rutile dioxides.
In a larger screening study it could be interesting to mix these materials into
the surface layers of MnO2.
Regarding the Mn-Au interactions, it would be interesting to investigate the
Au size dependence in greater detail. The sputter deposition could be tuned
to deliver Au domains in dierent sizes. Alternatively, size selected Au clusters
could be added to a MnO2 surface, with the cluster source available in our
laboratory. Ideally, such studies should be combined with more detailed DFT
calculations so that the properties of gold clusters could be related to observed
activities. In the end, the ultimate goal would be to use the understanding
of Mn-Au interactions to propose and design Mn oxides modied with another
abundant element or compound.