6.2 Experimental investigation of Au-MnOx catalysts 103
RF plasma and a bias on the substrate holder. After the cleaning substrate, Mn
and Au were co-sputtered with rates calibrated by the in chamber QCM. The
sputter deposition was carried out dierently than for the lms in chapter 4 and
5, where a slow rate of MnOx deposition was used. The Au sputter yield is high
compared to that of Mn in the presence of oxygen, so the argon to oxygen ratio
was changed from 25:5 to 25:3. This ensured a higher rate of MnOx, which could
then be matched with the Au rate. Furthermore, with a low overall deposition
rate there would be a greater chance of the Au forming large domains due to
high mobility of Au atoms. The substrate temperature during deposition was
kept at 200 oC. The focus of this study will be on comparing two dierent Au
concentrations (30 and 50 % on a total metal basis) in MnOx with pure MnOx
lms.
6.2.1 Characterization
The prepared thin lms were characterized to investigate any dierences due
to the changed preparation procedure and Au modication. First X-ray Photoelectron
Spectroscopy was used to quantify the Mn:Au ratio in the lms and
to get an initial estimation of the Mn:O stoichiometry. The stoichiometry can
be evaluated with the Mn2p1
2 satellite distance and Mn3s multiplet splitting, as
explained in section 3.2.1. In gure 6.5a and b the Mn2p and Mn3s spectra can
be seen for the MnOx lms prepared with a lower oxygen partial pressure. For
the pure MnOx lm the distance between the Mn2p 1
2 peak and its satellite is
found to be 10.1 eV. The distance between the two peaks in the Mn3s spectrum
is 5.4 eV. These values match with reference values for the Mn3O4 oxide as seen
in table 3.1. The Au concentrations were found from integrating the Mn2p and
Au4d peaks, shown in gures 6.5a and c. The Au concentrations are reported
on a total metal basis, i.e. Au
Au+Mn.
In order to investigate the structure of the mixed lms, Glancing Angle X-ray
Diraction experiments were carried out. For these measurements the lms
were deposited with double thickness, to increase signal from the lms, on SiO2
windows. The resulting diractograms can be seen in gure 6.6a and b, for pure
MnOx and Au modied lms respectively.
The pure MnOx lm matches well with a Mn3O4 structure based on nine observed
peaks. From this point the pure Mn oxide lms will therefore be denoted
Mn3O4. This Mn:O stoichiometry is consistent with a lower amount of oxygen
available during deposition and is in contrast to the disordered MnO2 prepared
at higher oxygen ow used previously. Interestingly, the ratios of peak heights
do not match the powder standard reference, which indicates preferential growth
of some planes, which is not unusual for thin lms. However, from 6.6b, where
the results from mixed lms are shown, the Mn3O4 phase vanishes upon introduction
of Au. Instead, the four most pronounced Au peaks are observed.
For both concentrations the peaks are broad which indicates small Au domains
located in the Mn3O4 lm. The peaks are slightly broader in the lm with 30 %