CHEMELECTROCHEM
ARTICLES www.chemelectrochem.org
Electrochemical Tests
The electrochemical tests on RuO2 were performed in a glass cell
with 0.05m N2-saturated H2SO4 (Merck Suprapur 96%, diluted with
18.2 MW Millipore water) at room temperature. The tests on MnOx
were performed in a two-compartment Teflon cell with 1m N2-saturated
KOH (Merck Suprapur 99.995, diluted with 18.2 MW Millipore
water) at room temperature. For the tests in sulfuric acid, a Hg/
HgSO4 reference electrode was used, whereas for the tests in KOH,
a Hg/HgO in 20 wt% KOH reference electrode was used. The reference
electrode potentials were measured with respect to a reversible
hydrogen electrode (RHE) by bubbling 1 bar hydrogen over
a clean Pt surface in the same electrolyte. In both electrolytes,
graphite rods were used as counter electrodes. All the data is presented
using the RHE scale and corrected for Ohmic losses, found
from the fitted high-frequency intercept measured using electrochemical
impedance spectroscopy over the range of 1–200000 Hz
at a DC potential of 10 mV. The Ohmic drop for tests in 0.05m
H2SO4 was in the range of 15–18 W, whereas the tests in 1m KOH
comprised an Ohmic drop in the range of 3–5 W. For chronoamperometry
measurements, the Ohmic drop was compensated at
85% using the Bio-Logic software EC-lab method MIR. Mass
changes from EQCM measurements were calculated using the Sauerbrey
equation with Cf=(56.62.8) Hzcm2mg1;52 this value was
calibrated by the electrodeposition of silver onto the Au-coated
quartz crystal, repeated six times.53 Each stability test performed
by both EQCM and ICP–MS was repeated four times.
Since gold is not expected to be stable at high oxidative potentials,
40 we did not perform reference measurements with the bare
EQCM crystals. Instead, the gold was coated by the catalyst thin
film, as described above, and we measured the amount of gold in
the electrolyte after OER tests. Based on two-hour tests at 1.9 VRHE
for three MnOx samples, the increase
of gold was less than
9 ngcm2 or equivalent to less
than 0.6 Hz. We therefore assumed
that Au is sufficiently masked from
the electrolyte.
At high current densities, gas formation
on the electrode affects
the frequency measurement; however,
since the electrode was oriented
vertically, the bubbles
moved upwards and did not accumulate
on the active area. While
the bubble formation could cause
some noise in the measurement, it
would not have an effect on the
trends observed over two-hour
experiments.
Characterisation Methods
ICP–MS experiments were performed
with equipment from
Thermo Fisher Scientific, model
iCAP-QC ICP-MS. Samples were
taken out of the electrolyte before
and after each measurement using
a pipette. For tests in 0.05m H2SO4,
the samples were analysed without
further dilution, whereas for
1m KOH, the samples were diluted to 0.1m to protect the ICP–MS
components. For the quantitative analysis, calibration tests were
performed using diluted solutions of Mn or Ru, made from standards
with 1000 mgmetalmL1 purchased from SCP Science. Calibrations
were made with at least three concentrations. These were
prepared in the range of 0.1 to 10 mgL1 since the concentrations
of Mn and Ru in the investigated electrolytes are all within that
range. The calibration curves obtained could all be fitted to
a linear curve with an R2 of 0.99 or better. To calculate the total
mass loss with ICP–MS, the volume of the electrolyte was measured
for each experiment. For measurements with the two-compartment
Teflon cell, only the volume of the compartment containing
the working electrode was used. It was confirmed with a separate
ICP–MS test that the amount of metal in the reference-electrode
compartment was negligible. The thin films were evaluated
by X-ray photoelectron spectroscopy using a Theta Probe instrument
(Thermo Scientific) where the base pressure was 5
1010 mbar. The X-ray source was monochromatized AlKa
(1486.7 eV). Furthermore, the thin films were analysed by X-ray diffraction
(XRD) using a PANanalytical X’pert PRO equipment with an
X-ray wavelength of 1.54 for the CuKa line.
2. Results and Discussion
To determine the structure of the two oxides, we performed
glancing-angle XRD measurements. In Figure 2 a, the diffractograms
for RuO2 and MnOx are shown, together with literature
references and a measurement for the glass substrate. For the
RuO2 film, the diffractogram obtained was consistent with
a rutile RuO2 structure,54 whereas for MnOx, no significant
peaks were found, indicating that the film is amorphous. To
Figure 2. a) XRD diffractograms for RuO2, MnOx and the substrate, together with literature references. The literature
data for RuO2 is from Ref. 54 and that for MnO2 from Ref. 59. b) XPS spectra of the Mn 3s region for
a 40 nm MnOx on EQCM sample. The red arrow indicates the difference in binding energy for the Mn 3s multiplet
splitting. c) XPS spectra of the Mn 2p region for a 40 nm MnOx on EQCM sample. The red arrow indicates the distance
from the Mn 2p1/2 peak to its corresponding satellite. d) XPS spectra of the Ru 3d core level region for
a 40 nm RuO2 on EQCM sample.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim ChemElectroChem 0000, 00, 1–8 &3&
These are not the final page numbers!