A study of PtRuO2 catalysts thermally formed on titanium mesh for methanol oxidation

Article abstract of DOI:10.1016/j.electacta.2004.07.047

Platinum-based catalysts, for the electro-oxidation of methanol, have been made by thermal decomposition of chloride precursors onto titanium mesh. The catalysed electrodes were successfully operated in acidic methanol electrolytes. Electrochemical characterisation has been carried out using cyclic voltammetry, electrochemical impedance spectroscopy and galvanostatic polarisations. A complete analysis of the electrochemical results showed that the preliminary performance of the catalysed titanium mesh was comparable to that achieved with carbon-supported PtRu catalysts. The catalysts formed on titanium mesh by thermal decomposition also exhibited dimensional stability. Catalysed titanium mesh therefore appears to be a promising alternative to carbon-supported catalysts for certain fuel cell applications.

Full text of DOI:10.1016/j.electacta.2004.07.047

Electrochimica Acta 50 (2005) 1217–1223
A study of PtRuO2 catalysts thermally formed on
titanium mesh for methanol oxidation
L.X. Yang1, R.G. Allen, K. Scott∗, P.A. Christenson, S. Roy
School of Chemical Engineering and Advanced Materials, University of Newcastle, Merz Court, Newcastle upon Tyne NE1 7UR, UK
Received 5 May 2004; received in revised form 10 June 2004; accepted 31 July 2004
Available online 15 September 2004
Abstract
Platinum-based catalysts, for the electro-oxidation of methanol, have been made by thermal decomposition of chloride precursors onto
titanium mesh. The catalysed electrodes were successfully operated in acidic methanol electrolytes. Electrochemical characterisation has
been carried out using cyclic voltammetry, electrochemical impedance spectroscopy and galvanostatic polarisations. A complete analysis
of the electrochemical results showed that the preliminary performance of the catalysed titanium mesh was comparable to that achieved
with carbon-supported PtRu catalysts. The catalysts formed on titanium mesh by thermal decomposition also exhibited dimensional stability.
Catalysed titanium mesh therefore appears to be a promising alternative to carbon-supported catalysts for certain fuel cell applications.
© 2004 Elsevier Ltd. All rights reserved.
Keywords: Methanol oxidation; Thermal decomposition; Titanium mesh; Fuel cell; AC impedance; Platinum and ruthenium electrocatalysts
1. Introduction
back their widespread commercial applications [1]. Further-
more, appropriate carbon dioxide (the product of the anode
reaction) management is an important issue for good cell
performance. If the gas produced at the anode is not rapidly
removed, a gradual deterioration in electrical performance
can occur. Modifications to and improvements in the flow
bed design have been used to improve the removal of carbon
dioxide from the anode side of the DMFC [3]. Results have
shown promising behaviour for the DMFC using on stainless
steel mesh flow fields, in terms of gas removal characteristics
and electrical performance. Based on previous work [3], fab-
ricationandcharacterisationofelectrocatalystdirectlycoated
onto titanium mesh are of interest at Newcastle University.
The results of incorporating mesh into the MEA structure of
the DMFC are encouraging [4,5].
When operated at relatively low temperatures, direct
methanol fuel cells (DMFCs) are promising candidates for
portable and stationary power sources and transportation
[1,2]. A principal reason is that the fuel is stored and reacted
in a liquid form. In comparison to fuel cell systems using H2
reformed from methanol, the DMFC, when using methanol
as a liquid fuel, has the advantages of simple system design
and cell operation.
In spite of the improved performance of DMFCs, using
polymer electrolyte membranes (PEM), achieved over re-
cent years, obstacles such as the low activity of methanol
electro-oxidation catalysts and methanol crossover still hold
For the DMFC, catalyst syntheses can be achieved us-
ing different techniques such as impregnation, colloids, elec-
trodeposition, sol–gel and thermal decomposition. In gen-
eral, electrocatalytic anodes for industrial use should possess
properties such as high conductivity, thin catalyst layers (e.g.
low loading), low cost support, rapid and easy fabrication and
coating, high physical stability and low cost [6]. Takasu and
∗
Corresponding author. Tel.: +44 191 2228700;
fax: +44 191 2225292.
E-mail addresses: lianxi.yang@nrc.gc.ca (L.X. Yang),
k.scott@ncl.ac.uk (K. Scott).
1
Present address: Integrated Manufacturing Technologies Institute, 800
Collip Circle, London, Ont., Canada N6G 4X8. Tel.: +1 519 430 7133;
fax: +1 519 430 7064.
0013-4686/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2004.07.047

1218
L.X. Yang et al. / Electrochimica Acta 50 (2005) 1217–1223
Murakami have compared different methods, including ther-
mal decomposition and sol–gel methods, to prepare Ru-based
oxide-coated electrodes and recommended that thermal de-
composition was effective for the preparation of highly active
catalysts [7].
The present work considers the preparation of electrocata-
lysts onto titanium mesh, by thermal decomposition of chlo-
ride precursors, and their electrochemical performance for
the oxidation of methanol. This is the first step in our re-
search to develop titanium mesh catalytic electrodes, which
can be directly hot pressed onto a solid polymer electrolyte
membrane and used in a fuel cell.
each electrode. Calcination was then performed in an oven
(in air) at 400 ◦C for 1 h. The electrodes fabricated in this way
were designated PtO2/Ti, RuO2/Ti and PtRuO2/Ti (Pt:Ru =
1:1 in atomic ratio).
In certain experiments, the electrodes were subsequently
coated with 5 wt.% Nafion® solution (Aldrich). This coating
was achieved by dipping the catalysed titanium mesh into
the Nafion® solution and then drying it in air, in an oven, at
80 ◦C for half an hour. The resulting Nafion® loading was
approximately 0.5 mg cm−2. The electrodes were immersed
in distilled water for 24 h before electrochemical evaluation
in order to hydrate the Nafion layer.
Cyclic voltammetry (CV), electrochemical impedance
spectroscopic (EIS) and galvanostatic experiments were per-
formed using a Gill AC potentiostat (ACM Instruments), con-
trolled by a Windows Personal Computer loaded with ACM
Sequencing Software, version 3. All experiments were con-
ducted in an N2-deaired electrolyte containing 2 mol dm−3
CH3OH + 0.5 mol dm−3 H2SO4. A three-compartment cell,
with a reference electrode separated from the working and
counter-electrode compartment by a Luggin capillary, was
employed. The working electrodes were lacquered (Mic-
croshield, Hi-Tek Products Ltd.), leaving a window of 1 cm2
exposed for electrochemical tests. The counter-electrode was
a platinum mesh measuring 25 mm × 25 mm and the refer-
ence electrode was a mercury sulfate (Hg/Hg2SO4/H2SO4
(0.617 V versus RHE)). Unless otherwise specified, electrode
potentialsarereferencedagainstthereversiblehydrogenelec-
trode (RHE).
2. Experimental
In this study thermal decomposition, which is widely
used for preparing dimensionally stable anodes (DSA), was
employed to deposit catalyst onto a titanium mesh support
(supplied by Dexmet Corporation). The geometric charac-
teristics of the mesh were as follows: strand width (SW) =
0.14 mm, thickness = 0.2 mm, and opening size = 1.5 mm
(LWD)/0.67 mm (SWD) and is shown in Fig. 1. Prior to de-
positing the catalyst layer, the titanium mesh was cut into
1.2 cm × 1.2 cm pieces and etched in 10% oxalic acid at
80 ◦C for 1 h, to enable good adhesion of electrocatalyst.
The mesh was then thoroughly rinsed with distilled water
before coating with catalyst. To apply a catalyst layer by
thermal decomposition, the etched substrate was dipped five
times into a precursor solution (e.g. 50% of a 0.2 mol dm−3
of H2PtCl6·6H2O in ethanol + 50% of a 0.2 mol dm−3 of
RuCl3·xH2O in isopropanol). After each dip, the sample was
manipulated, and gently swirled to form a uniform coating,
which was then dried at 100 ◦C for 10 min. A mass of approx-
imately 0.2 mg (nominal thickness of 0.07 ␮m) on 1 cm2 of
substrate was formed on each dipping, yielding a final cat-
alyst loading of approximately 1 mg cm−2 of total metal on
3. Results and discussion
In this work, great care was taken to ensure that a stable
state of the electrode was reached prior to collecting electro-
chemical data. It has been reported [8] that for a thermally
prepared electro-catalyst, whilst the bulk characteristics were
unchanged for the duration of the electrochemical experi-
ments, the electrochemical behaviour did change, especially
in the early life of the electrode. The change in electrochem-
ical response was attributed to so-called “hydration effects”
or a maturing process [9] and could be evidenced from pro-
gressive changes in the voltammetric charges and capaci-
tances [8]. Hence the electrodes were cycled between 0.0 and
900 mV in 0.5 mol dm−3 H2SO4 until a negligible variation
in the current voltage behaviour was observed. At this point,
a stable electrode state was reached and the electrode was de-
scribed as mature. All results presented here were obtained
on mature electrodes.
Fig. 2 shows a typical voltammogram obtained in the
acidic methanol electrolyte at 20 ◦C for a PtO2/Ti electrode
prepared by thermal decomposition at 400 ◦C. The voltam-
mogram has a number of characteristic features that are gen-
erally seen for platinum metal electrodes (see inset of Fig. 2).
On the forward sweep, a polarisable region (double-layer
region) is observed at the potential below 0.5 V. With an
Fig. 1. Titanium mesh used for thermal deposition of electrocatalyst.

L.X. Yang et al. / Electrochimica Acta 50 (2005) 1217–1223
1219
Fig. 2. Cyclic voltammograms obtained, respectively, on PtO2/Ti thermally formed on titanium mesh at 400 ◦C for 1 h (sweep rate: 100 mV s−1) and on platinum
metal (the inset, sweep rate: 20 mV s−1). The electrolyte: 2 mol dm−3 CH3OH + 0.5 mol dm−3 H2SO4 (20 ◦C).
increase in potential, an anodic peak is observed (peak I in
Fig. 2) in the potential range 500–700 mV. This peak may
be attributed to dehydrogenation of methanol. This process
is slow on the surface of catalysts at lower potentials, and
exhibits an activation energy of 35 kJ mol−1 [10]. The co-
adsorption of water and methanol on platinum can also take
place and has been evidenced by kinetics isotope studies on
methanol oxidation [11].
of the sweep potential beyond 1000 mV. In particular, the
ruthenium component in the electrode may be oxidised
and reduced reversibly (e.g. RuOx(OH)y/RuOx−δ(OH)y+δ
)
through a mechanism involving proton exchange with so-
lution [12,13]. However, the purpose of applying the high
potential was to examine the overall electrochemical charac-
teristics of the thermally prepared electrodes. In practical fuel
cell applications, the potentials reached will be well below
1000 mV and such changes in the ruthenium oxidative state
will be less relevant.
Methanol oxidation began at approximately 800 mV and
reached its maximum peak current density at about 900 mV
(peak II). After this peak the current declined, as the surface
was “poisoned” by the formation of carbon oxides [11]. At a
potential greater than 1100 mV, methanol oxidation contin-
ues. On the reverse sweep, re-oxidation of methanol began
at approximately 800 mV and reached a peak current density
at around 600 mV (peak III); after which strongly bonded
surface intermediates began to block the catalyst surface.
The electrochemical behaviour of PtRuO2/Ti electrode
(Fig. 3) was similar to that of the PtO2/Ti electrode (Fig. 2),
although the peak current densities of the former were signif-
icantly higher than the latter (by a factor of approximately 2).
The inset in Fig. 3 shows that no peaks appeared in the cyclic
voltammogram for the RuO2/Ti electrode, i.e. ruthenium ox-
ide alone has little activity for methanol oxidation. Therefore,
the main role of ruthenium in the PtRuO2/Ti electrode is to
promote the electrochemical activity of the platinum catalyst
for methanol oxidation, via the so-called bi-functional mech-
anism [11]. It should be noted that, during the cyclic voltam-
metry measurements, the surface structure of the PtRuO2/Ti
electrode might be changed, especially at the upper limit
The traditional DMFC is also usually operated with a solid
polymer proton conducting membrane and with a Nafion®
coating on the carbon-supported electrocatalysts. To study
the effect of a Nafion® coating on the electrochemical per-
formance of the catalyst, galvanostatic polarisation was per-
formed on coated and uncoated PtO2/Ti in acidic methanol
solution at 60 ◦C; the results are shown in Fig. 4. Before the
measurement, the Nafion®-coated electrodes were dried at
80 ◦C for half an hour and then immersed in distilled water
for 24 h. A negative effect on methanol oxidation was antici-
pated due to the Nafion® coating presenting a possible mass
transfer barrier for methanol. However, Fig. 4 shows that the
Nafion® coating had an insignificant effect on the perfor-
mance of the electrodes: the electrochemical behaviour of
the Nafion®-coated electrodes is reproducible and compara-
ble with that of the uncoated electrodes. The slightly lower
current densities obtained with the Nafion-coated electrode
may be due a slightly lower methanol concentration at the
electrode surface, due to diffusion and/or a different pro-
ton concentration at the catalyst surface. In the following,

1220
L.X. Yang et al. / Electrochimica Acta 50 (2005) 1217–1223
Fig. 3. Cyclic voltammograms obtained, respectively, on PtRuO2/Ti thermally formed on titanium mesh at 400 ◦C for 1 h (sweep rate: 100 mV s−1) and on
RuO2/Ti thermally formed on titanium mesh at 400 ◦C for 1 h (the inset, sweep rate: 20 mV s−1). The electrolyte: 2 mol dm−3 CH3OH + 0.5 mol dm−3 H2SO4
(20 ◦C).
unless specified, the electrochemical data were collected on
the electrodes coated with Nafion®.
in voltammetry measurements, as shown in Fig. 5. There is
reported evidence that potential sweeps can increase poros-
ity or open sites in the catalyst [15], creating more hydrated
active sites for direct contact with reactant species [9]. Thus,
possible decreased electrochemical activity, due to oxidation
of the metal, is compensated for by the increased hydrated
surface area of the electrode during the cyclic sweeps, leading
to an almost constant activity. In addition, it is well known
that anodes prepared using thermal decomposition are highly
In order to assess the stability of the PtRuO2/Ti electrodes,
anodic linear sweeps were conducted in the acidic methanol
electrolyte at 60 ◦C. In the measurement, the sample was po-
larised at 1.44 V for 5 min between each test. Such a high
polarised potential may form a less hydrous and more closed
ruthenium film [14] or cause oxidation of titanium beneath
the PtRu oxide thin film; potentially reducing the electro-
chemical activity of the electrode. However, after repeated
tests, little change in activity of the electrode was observed,
Fig. 5. Anodic linear sweep (sweep rate: 20 mV s−1) of PtRu/Ti thermally
formed on titanium mesh at 400 ◦C for 1 h, in 2 mol dm−3 CH3OH +
0.5 mol dm−3 H2SO4 (60 ◦C), showing the reproducibility of the catalytic
activity. Between each series test, the sample was polarised at 1.44 V vs.
RHE for 5 min.
Fig. 4. Galvanostatic polarisation plots in 2 mol dm−3 CH3OH
+
0.5 mol dm−3 H2SO4 (60 ◦C) on PtO2/Ti thermally formed on titanium mesh
at 400 ◦C for 1 h, showing the effect of Nafion coating on the catalytic ac-
tivity.

L.X. Yang et al. / Electrochimica Acta 50 (2005) 1217–1223
1221
Fig. 7. Cyclic voltammograms obtained on PtRuO2/Ti catalysts thermally
formed on titanium mesh at 400 ◦C for 1 h in 2 mol dm−3 CH3OH +
0.5 mol dm−3 H2SO4 (20 ◦C), showing the effect of sweep rate on the
voltammetric behaviour (sweep rates: 10, 40, 60, 100 and 200 mV s−1).
Fig. 6. Galvanostatic polarisation plots in 2 mol dm−3 CH3OH
+
sweep rates between 10 and 200 mV s−1 were carried out, and
typical voltammograms obtained on PtRuO2/Ti are shown in
Fig. 7. From these voltammograms, plots of current density,
j, measured at 350 mV, versus sweep rate, ν, are derived,
and they show a linear relationship for the entire sweep rate
range investigated (Fig. 8). The pseudo-capacitance of the
electrodes, Cpseudo, can be obtained from the slopes of the
lines, using the following relationship between current den-
sity, j and sweep rate, ν [17–18]:
0.5 mol dm−3 H2SO4 (60 ◦C) on different catalysts thermally formed on
titanium mesh at 400 ◦C for 1 h: (ꢀ) Pt/Ti, (᭹) PtRuO2/Ti and (ꢁ) from
Hogarth [16].
dimensionally stable [6]. Since catalysts used in fuel cells are
usually operated at much lower potentials (e.g. 500 mV ver-
sus RHE) than the maximum potential tested in the current
study, the electrode is expected to be stable in real fuel cell
operation.
Fig. 6 compares the galvanostatic polarisation behaviour
of PtO2/Ti and PtRuO2/Ti. The data clearly indicate that the
presence of ruthenium increased the catalytic activity of the
Pt catalyst thermally formed at 400 ◦C; the onset potential
of methanol oxidation, on PtRuO2/Ti, is 100 mV lower than
that on PtO2/Ti. In Fig. 6, the catalytic activity on a carbon-
supported PtRu catalyst (1.5 mg loading with a ratio of Pt:Ru
= 1:0.5) reported from Hogarth [16] is also provided for com-
parison. The data show that the activity of the catalysed tita-
nium mesh (PtRuO2/Ti) is not as good as that of one of the
most active carbon-supported catalysts reported. The differ-
ence in activity is of the order of 50 mV. However, the results
quoted for the carbon-supported catalyst were obtained using
1.5 mg cm−2 of PtRu noble metals, while the amount of the
PtRu metals used in this study was 1.0 mg cm−2. This may
be a factor that explains the difference in performance, al-
though in general an optimisation of the catalyst loading and
indeed the precursor salt concentration and type is required.
It was also found that further improvements in the electrode
performance could be achieved by thermally depositing the
PtRu oxide catalyst onto different types and configurations
of titanium mesh. The effect of different types of mesh on
electrochemical activity of the catalysts and different com-
positions of precursor solution will be reported in a separate
paper.
j = Cpseudoν
(1)
The pseudo-capacitance thus obtained for the PtRuO2/Ti and
PtO2/Ti electrodes are approximately 26 and 4.5 mF cm−2
.
The different values of pseudo-capacitance may reflect dif-
ferent double-layer charging and adsorption processes in the
interface. However, a realistic interpretation of the pseudo-
capacitance results requires input from other experiments
(e.g. AC impedance).
Impedance spectroscopy can be used to distinguish elec-
trode processes according to their kinetics, i.e. their “relax-
ation times”. Detailed analysis of the Nyquist loci at different
ranges of frequency may reveal constituent relaxation pro-
cesses. A typical capacitance–frequency plot derived from
The electro-catalysts thermally formed on titanium mesh
can be further characterised by cyclic voltammetry in the
double-layer potential region, where the electrochemical be-
haviour of the electrode mainly corresponds to the double-
layercharging/dischargingprocessesattheinterface. Voltam-
metric measurements in the double-layer region at different
Fig. 8. Current density–sweep rate plots derived from the voltammograms
(e.g. Fig. 7) in 2 mol dm−3 CH3OH + 0.5 mol dm−3 H2SO4 (20 ◦C) on
PtO2/Ti and PtRuO2/Ti thermally formed on titanium mesh at 400 ◦C for
1 h.

1222
L.X. Yang et al. / Electrochimica Acta 50 (2005) 1217–1223
measurement was 5 mV (e.g. 10 mV peak to peak). This
means that a frequency of 1 Hz is equivalent to a sweep
rate of 10 mV s−1 in the voltammetry measurement, i.e.
the voltammograms performed at sweep rates from 10 to
200 mV s−1 correspond to the low-frequency range of about
1–20 Hz. From the capacitance–frequency curve (Fig. 9), a
ratio of the capacitance/frequency slope for the PtO2/Ti and
PtRuO2/Ti electrodes is approximately 5.3 (3.7/0.7). The
corresponding ratio for the pseudo-capacitance obtained
from the plots of current versus sweep rate (Fig. 8) was
approximately 26/4.5 = 5.8. Thus, results obtained from both
cyclic voltammetry and from impedance measurements,
over the low-frequency range, are similar. It has been
previously suggested that pseudo-capacitance could be used
as a relative measure of the electrochemically active surface
area of electrocatalytic oxide electrodes [17]. As such, the
higher pseudo-capacitance of the PtRuO2/Ti, compared to
that of the PtO2/Ti, suggests that the former possessed the
higher electrochemically active surface area, although the
true area of the PtRuO2/Ti was smaller than that of PtO2/Ti,
as evidenced from the capacitances over the high-frequency
range (Fig. 9). This is apparently attributed to the ruthenium
component in the PtRuO2/Ti electrode. Processes, such as ad-
sorption/desorption of methanol and water on the PtRuO2/Ti
catalysts, were much more efficient than on the PtO2/Ti. The
adsorption/desorption of water on the ruthenium oxide active
sites increased the activity of the Pt catalyst for methanol
oxidation and supports the bi-functional mechanism of
PtRu-based catalysts, as frequently reported [11].
Fig. 9. Capacitance–frequency curves derived from the Nyquist plots (in-
set) in 2 mol dm−2 CH3OH + 0.5 mol dm−3 H2SO4 (20 ◦C) on PtO2/Ti and
PtRuO2/Ti thermally formed on titanium mesh at 400 ◦C for 1 h. The bias
potential for the impedance measurement was at 350 mV vs. RHE.
theimpedancedataisdepictedin Fig. 9andthecorresponding
complex Nyquist plot is shown in the figure as an insertion.
In performing the impedance measurement, the bias potential
of 350 mV (versus RHE), located within the double-layer re-
gion, was employed to coincide the potential region used for
recording the cyclic voltammograms (Fig. 7). It has been pre-
viously reported that the capacitance obtained in the double-
layer region is independent of potential [19]; this makes it
plausible to use impedance to further characterise the elec-
trodes.
The capacitance–frequency curve (Fig. 9) showed dis-
tinct high (above 100 Hz) and low (below 10 Hz) frequency
processes. It is interesting to note that whilst the low-
frequency capacitance increased linearly with a decrease in
frequency, the high-frequency capacitance was quite flat. In
the high-frequency range, over 100 Hz, the adsorption and
desorption process cannot follow the frequency change. The
capacitance observed at high frequency can therefore only
be attributed to electric charging/discharging or specific
adsorption. The interactions of the solvent with the electrode
in the inner Holmholtz layer (IHL) were orientated according
to the electrode charge, where they formed a layer together
with specifically adsorbed ions [20]. It has been suggested
that the true surface area of electrocatalysts dominated the
high-frequency process at the interface in the polarisable
region [19]. Therefore, the high-frequency capacitances of
5 and 3 mF on PtO2/Ti and PtRuO2/Ti, as indicated in Fig. 9
are in the approximate ratio of the true surface areas of both
electrodes. In contrast, the low-frequency capacitance was
related to a slow process such as adsorption/desorption of
solvated ions, which was greatly influenced by the applied
frequency. Theslopesofthecapacitance–frequencyplot, over
the low-frequency range (below 10 Hz) were 0.7 and 3.7 mF
per decade Hz for PtO2/Ti and PtRuO2/Ti, respectively
(Fig. 9).
4. Summary
Preliminary electrochemical characterisation of Pt-based
electrocatalysts thermally deposited on titanium mesh has
been carried out. The electrodes were successfully used for
theoxidationofmethanolinacidicmethanolelectrolyteswith
and without a Nafion® coating. Further improvements in the
electrode performance by thermally depositing the PtRu ox-
ide catalyst onto different types and configurations of tita-
nium mesh and by using different precursor salt composi-
tions can be anticipated. Such electrodes, which exhibit di-
mensional stability, high conductivity and good resistance to
anodic corrosion of the titanium-supported catalyst, may lead
to a technically viable electrode to fabricate MEAs for fuel
cell applications.
Acknowledgements
EPSRC supported L.X. Yang and R.G. Allen. The MOD
supported the work through the Joint Grant Scheme No.
JGS/826. The work was performed in research facilities pro-
vided through an EPSRC/HEFCE Joint Infrastructure Fund
award No. JIF4NESCEQ. Mesh was supplied by DEXMET
Corporation, USA.
The pseudo-capacitances derived from the voltammo-
grams and the capacitances obtained from the impedance
measurements over the low-frequency range are comparable.
The amplitude of sinusoidal perturbations in the impedance

L.X. Yang et al. / Electrochimica Acta 50 (2005) 1217–1223
1223
References
[11] M. Watanabe, S. Motoo, Electroanalytical Chemistry and Interfacial
Electrochemistry 60 (1975) 275.
[12] S. Trasatti, G. Buzzanca, J. Electroanal. Chem. 29 (1971) 1.
[13] K. Doblhofer, M. Metikos, Z. Ogumi, H. Gerischer, Ber. Bunsenges.
Phys. Chem. 82 (1978) 1046.
[14] K. Kvastek, V. Horvat-Radosevic, J. Electroanal. Chem. 511 (2001)
65.
[15] V. Birss, R. Myers, H. Angerstein-Kozlowska, B.E. Conway, J. Elec-
trochem. Soc. 131 (1984) 1502.
[16] M.P. Hogarth, Ph.D. Thesis, University of Newcastle, 1995.
[17] V.A. Alves, L.A. Silva, J.F.C. Boodts, Electrochim. Acta 44 (1998)
1525.
[18] J.O.M. Bockris, T. Otogawa, J. Electrochem. Soc. 131 (1984)
290.
[19] J.P.G. Farr, L.X. Yang, Trans. IMF 79 (2001) 163.
[20] C.A. Brett, A.M.O. Brett, Electrochemistry Principles, Methods and
Applications, Oxford University Press, Oxford, 1994.
[1] X.M. Ren, E.T. Springer, T.A. Zawodzinski, S. Gottsfield, J. Elec-
trochem. Soc. (2000) 147.
[2] K. Scott, W.M. Taama, P. Argyropoulos, J. Power Sources 79 (1999)
43.
[3] K. Scott, P. Argyropoulos, P. Yiannopoulos, W.M. Taama, J. Appl.
Electrochem. 31 (2001) 823.
[4] E.H. Yu, K. Scott, Electrochem. Commun. 6 (2004) 361.
[5] L.X. Yang, R.G. Allen, K. Scott, P. Christensen, S. Roy, J. Fuel Cell
Sci. Technol., submitted for publication.
[6] G. Lodi, Electrodes of Conductive Metallic Oxides, Elsevier, Ams-
terdam, 1980, Part A.
[7] Y. Takasu, Y. Murakami, Electrochim. Acta 45 (2000) 4135.
[8] L.X. Yang, J.P.G. Farr, M.A. Ashworth, Trans. IMF 78 (2000) 231.
[9] S. Trasatti, Electrochim. Acta 36 (1991) 225.
[10] A. Hamnett, Catal. Today 38 (1997) 445.