Journal of Applied Electrochemistry 34: 61–66, 2004.

61
 2004 Kluwer Academic Publishers. Printed in the Netherlands.

Preparation of Pt–Ru bimetallic anodes by galvanostatic pulse electrodeposition:

characterization and application to the direct methanol fuel cell

C. COUTANCEAU*, A.F. RAKOTONDRAINIBÉ, A. LIMA, E. GARNIER, S. PRONIER, J-M. LÉGER and

C. LAMY
Equipe Electrocatalyse, UMR CNRS 6503, Universite´ de Poitiers, 40, avenue du Recteur Pineau, 86022 Poitiers,
France
(*author for correspondence, e-mail: [email protected])

Received 25 November 2002; accepted in revised form 24 June 2003

Key words: carbon electrode, DMFC, electrodeposition, galvanostatic pulses, platinum–ruthenium catalysts

Abstract

The electrodeposition of platinum and ruthenium was carried out on carbon electrodes to prepare methanol anodes
with different Pt/Ru atomic ratios using a galvanostatic pulse technique. Characterizations by XRD, TEM, EDX
and atomic absorption spectroscopy indicated that most of the electrocatalytic anodes consisted of 2 mg cm)2 of
Pt–Ru alloy particles with the desired composition and with particle sizes ranging from 5 to 8 nm. Electrochemical
tests in a single DMFC show that these electrodes are very active for methanol oxidation and that the best Pt/Ru
atomic ratio in the temperature range used (50–110
C) is 80:20. The influence of the relaxation time toff was also
studied and it appeared that a low toff led to smaller particle sizes and higher performances in terms of current
density and power density.

1. Introduction to clean the catalyst particles from surfactant contami-

nation [11, 12]. These treatments can greatly alter the
The main challenge for the development of DMFCs is to surface structure of the particles [13].
reduce the noble metal catalyst loading of the electrodes One way to avoid these problems is to prepare
and the associated cost without decreasing the efficiency electrodes by electrodeposition of metals at carbon
of the fuel cell [1–3]. To achieve this goal, it is necessary electrodes. In a previous paper [14], we have shown that
to increase the effective surface area of the catalysts, that it is possible to deposit platinum, ruthenium and
is, to increase the surface contact between the catalyst, molybdenum at carbon electrodes in a reproducible
the electronic conductor (carbon), the electrolyte (Naf- way by applying a constant potential. It has also been
ion) and the reactant (methanol). The electrochemical shown that these anodes could be used in DMFCs. Since
reaction occurs in this active part of the electrodes and it is generally recognized that Pt/Ru alloys are the most
thus the performances depend greatly on the kinetics of active catalysts for the oxidation of methanol [5–18], this
interfacial phenomena [4, 5]. paper deals with the preparation, characterization and
Usually, electrodes for PEMFCs (polymer exchange fuel cell tests of Pt–Ru anodes of different atomic
membrane fuel cells) are constituted of black carbon composition prepared by galvanostatic pulse electrode-
powder which acts as a catalyst support and solid position on a carbon support.
electrolyte such as Nafion [6–10]. In this case, to Some papers concern the electrodeposition of bime-
increase the performance of the electrodes (i.e., the true tallic systems by galvanostatic pulses [19] but, to our
surface area of the catalyst) one needs either to increase knowledge, none deals with Pt–Ru deposition. How-
the thickness of the active layer, for a given catalyst ever, some authors have described the electrochemical
loading, or to increase the amount of catalyst in the deposition, at constant potential, of bimetallic Pt–Ru
catalytic powder. Increasing the thickness of the active catalysts [14, 20]. Some articles describing the electro-
layer leads to a decrease in the diffusion rate of the deposition of platinum on a carbon support are avail-
reactant towards the catalytic sites, whereas increasing able [3, 21–25]. Gloaguen et al. [26] and Thomson et al.
the weight loading generally leads to an increase in the [27] have studied effective cathodes for hydrogen/oxygen
particle size of the catalysts, thus decreasing their fuel cells prepared by potentiostatic pulse electrodepo-
efficiency. Moreover, the most used methods of prepa- sition of platinum on carbon electrodes, whereas Choi
ration of carbon supported catalysts by the colloidal et al. [28] used galvanostatic pulse electrodeposition.
route, necessitate heating or oxidative treatment in order The galvanostatic technique is more convenient to
62
prepare high geometric surface area electrodes, because formation of ruthenium oxides at the counter electrode,
the electrical set up requires only two electrodes and an a pretreated Nafion 112 membrane was inserted
arbitrary waveform generator, whereas the potentiosta- between the deposition compartment containing the
tic set up requires a three electrode configuration. The solution of metallic salts in 1.0 M H2SO4 and the
potentiostatic technique is not convenient for electrodes counter electrode compartment containing only 1.0 M
with large geometric surface area (greater than about H2SO4. After deposition of the metals, the electrodes
10 cm2). were thoroughly rinsed with ultrapure water.
The main goal of this work is to show that Pt–Ru Prior to the preparation of the membrane electrode
bimetallic anodes can be prepared reproducibly using assembly (MEA), a mixture of water and 5 wt. %
the technique of metal electrodeposition by galvano- Nafion in alcohol was spread on the electrode surface,
static pulses and that these anodes display good perfor- which was then dried at 90
C in an oven and heated at
mance in a DMFC. 150
C to recast the Nafion film. The Nafion loading
of the electrode was 0.8 mg cm)2. The MEAs were
prepared by hot pressing (130
C, 90 s, 35 kg cm)2) an
2. Experimental details E-TEK cathode (2.0 mg cm)2 Pt loading, 40% metal/C,
40% PTFE, 0.8 mg cm)2 Nafion) and the home made
All solutions were prepared using Alfa Aesar salts of anode on a Nafion 117 membrane.
platinum (K2PtCl6) and ruthenium (K2RuCl5) dissolved XRD patterns were recorded using a Siemens diffrac-
in 1.0 M H2SO4 (suprapur from Merck) in ultrapure tometer in the Bragg–Brentano geometry with a CuKa
water (MilliQ Millipore). Carbon gas diffusion elec- X-ray radiation with about 10 mg of a powder com-
trodes of 10 cm2 geometric surface area were made using posed of PTFE, C and Pt/Ru. The EDX measurements
a carbon cloth from Electrochem Inc. on which an ink were performed with a Philips CM 120 microscope/
made of Vulcan XC72 carbon powder and PTFE EDX analyser equipped with a La B6 filament. For the
dissolved in isopropanol was painted. The gas diffusion transmission electron microscopy (TEM) the sample
electrodes were loaded with a 4 mg cm)2 mixture of was embedded in a polymeric resin and cut into a
carbon powder and 15 wt.% PTFE. The galvanostatic section as small as 50 nm with an ultramicrotome. A
pulse deposition of platinum and ruthenium was carried diamond knife was used to cut the sample. The
out in a two electrode cell using a high power poten- localization of the Pt and Ru by secondary electron
tiostat (Wenking model HP 88) and an arbitrary microscopy (SEM) was determined by X-ray mapping.
waveform generator (Hewlett Packard 33120A). The The fuel cell tests in a single DMFC with a 5 cm2
pulse programs were processed by a home made geometric surface area were carried out with a Globe
software. Figure 1 shows the scheme of the electrical Tech test bench. The E against j and P against j curves
set-up. The output current applied between the two were recorded using a high power potentiostat (Wenking
electrodes was monitored using a Tektronix digital model HP88) interfaced with a PC to apply the current
storage oscilloscope. The time program for the current sequences and to store the data, and a variable resis-
pulses was 0.1 s (ton) at j ¼ 20 mA cm)2, and 2.5 s or tance in order to fix the applied current to the cell.
0.3 s (toff) of relaxation time at j ¼ 0. To avoid the

3. Results and discussion

Computer
3.1. Electrode preparation and characterization

Several electrodes were prepared with the different

Control Input
CE WE Ref
Potentiostat compositions listed in Table 1. According to Choi et al.
[28], the best conditions for preparing the most effective
fuel cell cathodes by galvanostatic pulse electrodeposi-
R = 10 Ω
tion of platinum at carbon were: j ¼ 20 mA cm)2,
ton ¼ 0.1 s, toff ¼ 0.3 s. Under these conditions, Choi
et al. obtained well-defined Pt particles with a distribu-
Carbon electrode tion size around 2 nm and excellent performance in an
H2/O2 fuel cell. Thus, our study on the anode prepara-
Membrane tion was carried out under conditions close to those of
[28].
Digital storage
oscilloscope
According to Cheh [29], the rate determining step of
pulse electrodeposition is controlled by mass transport.
Waveform Then, two different relaxation times (toff ¼ 0.1 s and
generator
2.5 s) were used to determine the effect of reequilibrating
Fig. 1. Scheme of the electrical set-up used for the galvanostatic pulse the concentration of metal salts at the electrode surface
electrodeposition. on the particle size and composition. Theoretically, the
 63
deposition of 2.0 mg cm)2 of metals on 10 cm2 carbon 1400

electrodes requires 4.2, 4.4 and 4.6 C cm)2 for Pt/Ru 1200
catalysts with an atomic ratio of 80/20, 65/35 and 50/50, 1000
respectively. The calculation of the time necessary for

Intensity / a.u.
800
metal deposition, assuming a faradaic yield of 100%,
gives about 90 and 15 min, with toff ¼ 2.5 s and 0.3 s, 600

respectively, and ton ¼ 0.1 s. However, it took several 400

hours to obtain metal deposition at the electrode 200
surface, indicating that the faradaic yield is very low,
0
i.e. about 10%. Two explanations are proposed: first, 30 40 50 60 70 80 90 100
the variation of the electrode potential during the 2 θ / degree
current pulse leads to a nonfaradaic current due to the
charge of the double layer capacity (which is important Fig. 2. XRD pattern of 80–20 (- - -); 65–35 (  ) and 50–50 (—) Pt–Ru/
C anodes prepared by galvanostatic pulse electrodeposition.
in the case of a carbon support), so that only a part of
the overall applied current is involved in the electro-
chemical reaction [30]; secondly, as soon as some
1200
platinum particles are formed on the surface of the
Cu grid
carbon electrode, hydrogen evolution takes place during 1000
Pt
the pulse of negative current. Both these effects are
responsible for the decrease in the faradaic yield. It is Intensity / a.u. 800
then difficult to estimate the deposition time necessary Pt

for preparing electrodes containing the desired metal 600

loading. To control the catalyst composition and

400 Pt
loading, the concentration of metallic salts in solution
was chosen to lead to a theoretical loading of 200 Pt
Ru Ru

2.5 mg cm)2 if all the metal is deposited. In that way, Pt

Pt

the catalyst composition should be close to the ratio of 0

the metal salt concentrations. The remaining solution 0 5 10 15 20
Energy / keV
was then analysed by atomic absorption spectroscopy in
order to determine the amount of platinum and ruthe- Fig. 3. EDX measurements on a 65–35 Pt–Ru/C anode.
nium which did not react and further to evaluate the
loading and composition of the catalysts at the carbon
electrode. Results gave catalyst loading values between then clear that the choice of toff is very important.
1.92 and 2.1 mg cm)2, which is very close to that Applying the Vegard law for a true alloy and the values
expected. The Pt/Ru atomic ratios listed in Table 1 of the cell parameter for bulk alloys as determined by
indicate that the experimental values are again in good Vogel et al. [33], we succeeded in evaluating the atomic
agreement with those expected. ratio. In Table 1, the XRD results are compared with
The XRD patterns of different anodes (Figure 2) those obtained using EDX (Figure 3) and atomic
show characteristic peaks of a platinum ccp structure. It absorption spectroscopy (AAS) showing a good agree-
can also be seen that the diffraction peaks shift towards ment in the analytical results.
higher 2h values when the Pt/Ru ratio is decreasing. The The TEM image (Figure 4(a)) shows the formation of
value of the lattice parameter as obtained from the aggregates with sizes between 20 and 50 nm. These
refinement of the whole pattern [31] indicates that a Pt– aggregates are in fact comprised of smaller particles as
Ru alloy was formed. Using the Williamson and Hall shown in Figure 4(b). The average size of the particles is
plot [32], the particle size of the catalysts could be close to 8 and 5 nm for toff ¼ 2.5 s and toff ¼ 0.3 s,
evaluated. It was found to be close to d
7 nm and respectively, in good agreement with the XRD results.
d
5 nm with toff ¼ 2.5 s and 0.3 s, respectively. It is These particle sizes are rather greater than those

Table 1. Electrodes preparation and characterization

Electrode Pt/Ru atomic ratio toff Atomic ratio Particle size

in solution /s /nm
XRD EDX AAS

PtRu 50–50 50/50 2.5 55/45 60/40 – 7–8

PtRu 50–50 50/50 0.3 57/43 53/47 48/52 5
PtRu 65–35 65/35 2.5 69/31 69/31 – 7–8
PtRu 65–35 65/35 0.3 66/34 – 69/31 5
PtRu 80–20 80/20 2.5 75/25 82/18 – 7–8
PtRu 80–20 80/20 0.3 77/23 78/22 78/22 5
64

Fig. 4. TEM photographs of 65–35 Pt–Ru/C catalyst from an anode

prepared by galvanostatic pulse electrodeposition with toff ¼ 2.5 s.
Scale: (a) 100 nm, (b) 20 nm. Fig. 5. Secondary electron microscopy (SEM) of a 65–35 Pt–Ru/C
catalyst from an anode prepared by galvanostatic pulse electrodepo-
sition with toff ¼ 2.5 s. (a) Pt La, (b) Ru Ka.

obtained by Choi et al. [28], indicating that the nucle-

ation or growth mechanism is different with Pt/Ru than 120
0.7
with Pt. In spite of this, the bigger size of our particles is
not necessary bad for methanol oxidation, since Frelink 0.6 100
et al. [34] showed that the best specific activity for 0.5
P / mW cm-2

methanol oxidation on a Pt/C catalyst was obtained 80

E/V

with particle sizes greater than 4.5 nm. 0.4

60
Secondary electron microscopy (SEM) was used to 0.3
determine the spatial repartition of platinum (Fig- 40
0.2
ure 5(a)) and ruthenium (Figure 5(b)). Both these met-
als are located in the same place on the electrode surface. 0.1 20

This result can be correlated with the data obtained

0 0
from XRD which indicate that the metallic phase is 0 100 200 300 400 500 600 700 800 900
mainly an alloy of platinum and ruthenium. j / mA cm-2

3.2. Fuel cell tests Fig. 6. Cell voltage (E) and power density (P) against current density
(j) in a single 5 cm2 surface area DMFC with a 80–20 Pt–Ru/C anode
prepared with toff ¼ 0.3 s, at different temperatures (Nafion 117
The electrodes were tested in a single 5 cm2 surface area membrane, 2 M CH3OH). Temperature: ()) 50, (n) 70, (n) 90, (*) 100
DMFC in order to determine their electrochemical and (s) 110
C.
behaviour under operating conditions. Figure 6 shows
the performances of a 80–20 Pt–Ru/C anode as a
function of temperature. Increasing the temperature reaches 110
C. This fact confirms the difficulty of
leads to improved performance. The maximum power oxidizing methanol and the necessity to work at
density at 50
C, close to 12 mW cm)2, is greatly temperatures higher than 100
C to enhance the elec-
enhanced (about ten times higher) when the temperature trode kinetics, and thus, the DMFC performance.
 65
0.8 120 0.7 120
0.7
100 0.6 100
0.6

-2
80 0.5

P / mW cm

P / mW cm-2
0.5

-
80
E/V

0.4 60 0.4

E/V
0.3 60
40
0.3
0.2
20 40
0.1 0.2
toff = 0,3 s
0 0 toff = 2,5 s
0.1 20
0 200 400 600 800 1000
-2
j / mA cm 0 0
0 200 400 600 800 1000
-2
Fig. 7. Cell voltage (E) and power density (P) against current density j / mA cm
(j) in a single 5 cm2 surface area DMFC with a 80–20 Pt–Ru/C anode
prepared with toff ¼ 0.3 s (¤) and commercial E-TEK anode 50–50 Fig. 8. Cell voltage (E) and power density (P) against current density
(h), at 110
C (Nafion 117 membrane, 2 M CH3OH, PO2 ¼ 2:5 bar; (j) in a single 5 cm2 DMFC with a 80–20 Pt–Ru/C anode prepared
PMeOH ¼ 2.0 bar). with toff ¼ 0.3 s (d) and toff ¼ 2.5 s ()) at 110
C (Nafion 117
membrane, 2 M CH3OH, PO2 ¼ 2:5 bar; PMeOH ¼ 2.0 bar).

Comparison of the 80–20 Pt–Ru/C anode with a

commercial 50–50 Pt–Ru/C anode from E-TEK (Fig- 0.7
(a)
90

ure 7) shows that the former electrode gives best 0.6 80

70
performance in term of power density, the maximum 0.5
of power density being close to 100 mW cm)2 with the 60

P / mW cm-2
0.4
E-TEK anode and to 110 mW cm)2 with the 80–20 Pt–
E/V

50

Ru/C anode. Power density values of 100 mW cm)2 0.3 40

30
achieved with E-TEK catalysts are rather low when 0.2
20
compared with data of Shukla et al. [35]. These authors 0.1
10
achieved 180 mW cm)2 at 90
C under ambient pressure 0 0
of oxygen. But, they used cathodes with a 4.6 mg cm)2 0 100 200 300 400 500 600 700
j / mA cm-2
platinum loading which minimizes the methanol cross-
over effect and likely enhances the cathode efficiency. In 0.7 120
our experiments, all parameters being kept constant, 0.6
(b)
100
comparison between the commercial and home made
0.5
anodes is consistent. For lower current densities, the

P / mW cm-2
80

commercial anode displays the best performance where-

E/V

0.4
60
as higher current densities are achieved with the home 0.3
made anode. This difference can be explained by catalyst 0.2
40

composition of the anode which is 50/50 for the 20

0.1
commercial anode and 80/20 for the home made anode.
0 0
In addition, the PTFE loading in the commercial anode 0 200 400 600 800 1000
is twice higher than that in the home made anode (i.e., j / mA cm-2

30 and 15 wt.%, respectively). The PTFE being hydro-

Fig. 9. Cell voltage (E) and power density (P) against current density
phobic, the diffusion of the methanol/water mixture
(j) in a single 5 cm2 surface area DMFC with different Pt–Ru/C
towards the catalytic sites is easier with a low PTFE anodes, at 110
C (Nafion 117 membrane, 2 M CH3OH, PO2 ¼ 2:5
loading, lowering the mass transfer limitation and then bar, PMeOH ¼ 2.0 bar). Key: (¤) Pt–Ru: 50–50; (h) Pt–Ru: 65–35;
increasing the performance at high current densities. (d) Pt–Ru: 80–20. (a) toff ¼ 2.5 s; (b) toff ¼ 0.3 s.
In Figure 8, the importance of toff is clearly demon-
strated. Best performance in terms of maximum power model of methanol electrooxidation involves four plati-
density was obtained with toff ¼ 0.3 s, as claimed by num sites for the dissociative adsorption of methanol
Choi et al. [28]. This fact can be correlated with the size and one ruthenium site for the water molecule activation
distribution of the catalyst particles, as determined by as shown by in situ FTIR spectroscopic studies [37].
EDX and XRD. Indeed, the particle sizes are smaller Other authors [38–40] also claimed that the best
and better distributed with toff ¼ 0.3 s than with platinum/ruthenium atomic ratio for methanol oxida-
toff ¼ 2.5 s. It is also clear from Figure 9(a) and (b) tion was between 20 and 30%. But it is possible that
that, in the temperature range 50 to 110
C, the best increasing the temperature involves some changes in the
performance in terms of maximum power density is oxidation mechanism of methanol and then in the
obtained with a Pt/Ru atomic ratio of 80/20, whatever optimal Pt/Ru atomic ratio. Thermal activation of
the conditions used for the preparation of the anodes. ruthenium may give it the ability to adsorb and
This result confirms those already obtained in a classical dehydrogenate methanol at temperatures greater than
three-electrode electrochemical cell [36]. Indeed, the 40
C [41].
66
4. Conclusions 7. S.C. Thomas, X. Ren and S. Gottesfeld, J. Electrochem. Soc. 146
(1999) 4354.
8. P. Argyropoulos, K. Scott and W.M. Taama, J. Power Sources 87
The main goal of this study was to prepare reproducibly (2000) 152.
new Pt–Ru bimetallic electrocatalysts by galvanostatic 9. A. Fisher, J. Jindra and H. Wendt, J. Appl. Electrochem. 28 (1998)
pulse electrodeposition. We obtained interesting results 277.
concerning the anode preparation and the behaviour of 10. P.L. Antonucci, A.S. Arico, P. Creti, E. Ramunni and V.
these anodes in a DMFC that can be summarised as Antonucci, Solid State Ionics 125 (1999) 431.
11. T.J. Schmidt, M. Noeske, H.A. Gasteiger, R.J. Behm, P. Britz and
follows: H. Bönnemann, J. Electrochem. Soc. 145 (1998) 925.
ii(i) The galvanostatic pulse electrodeposition method 12. T.J. Schmidt, M. Noeske, H.A. Gasteiger, R.J. Behm, P. Britz, W.
is convenient in terms of controlling the catalyst Brijoux and H. Bönnemann, Langmuir 13 (1997) 2591.
loading and composition (atomic ratio). 13. S. Mukerjee, J. Appl. Electrochem. 20 (1990) 537.
i(ii) The catalysts show an alloy character as determined 14. A. Lima, C. Coutanceau, J-M. Léger and C. Lamy, J. Appl.
Electrochem. 31 (2001) 379.
by XRD. 15. W.T. Napporn, J-M. Léger and C. Lamy, J. Electroanal. Chem.
(iii) The performance in a DMFC of such new bime- 404 (1996) 153.
tallic electrodes are better than that obtained with a 16. X. Ren, M.S. Wilson and S. Gottesfeld, J. Electrochem. Soc. 143
commercial E-TEK anode. (1996) L12A.
(iv) In the temperature range used to carry out this 17. S. Wasmus and W. Vielstich, J. Appl. Electrochem. 23 (1993)
120.
study, the best platinum/ruthenium atomic ratio is 18. P. Waszczuk, A. Wieckowski, P. Zelenay, S. Gottesfeld,
close to 80/20. C. Coutanceau, J-M. Léger and C. Lamy, J. Electroanal. Chem.
i(v) The influence of toff on the structure and perfor- 511 (2001) 55.
mance of the electrodes is crucial. 19. N.P. Leisner, C.B. Nielsen, P.T. Tang, T.C. Dörge and P. Moller,
Although the faradaic yield is low, this method of J. Mater. Process. Techn. 58 (1996) 39 and references therein.
20. C. Catteneo, M.I. Sanchez de Pinto, H. Mishima, B.A. Lopez de
preparing supported catalysts on a high surface area Mishima and D. Lescano, J. Electroanal. Chem. 461 (1999) 32.
carbon electrode is very convenient for industrial 21. E.J. Taylor, E.B. Anderson and N.R. Vilambi, J. Electrochem. Soc.
application: no organic solvent is required, the deposi- 139 (1992) 145.
tion by consumption of the metal salts in solution allows 22. T. Napporn, M-J. Croissant, J-M. Léger and C. Lamy, Electro-
control of the metal loading on the electrode and the chim. Acta 43 (1998) 2447.
23. A.D. Jannakoudakis, E. Theodoridou and D. Jannakoudakis,
same batch can be reused by adding metal salts and Synth. Met. 10 (1984/85) 131.
sulfuric acid to adjust the concentrations. Moreover, by 24. D-L. Lu and K-I. Tanaka, Surf. Sci. 373 (1997) L339.
varying different parameters (ton, toff, concentration of 25. A.A. Mikhaylova, O.A. Khazova and V.S. Bagotzky, J. Electro-
metallic salts, pulse current density, etc.), the structure anal. Chem. 480 (2000) 225.
and nature of the bimetallic catalyst can be changed and 26. F. Gloaguen, J-M. Léger, C. Lamy, A. Marmann, U. Stimming
and R. Vogel, Electrochim. Acta 44 (1999) 1805.
adapted to the desired application. 27. S.D. Thomson, L.R. Jordan and M. Forsyth, Electrochim. Acta 46
(2001) 1657.
28. K.H. Choi, H.S. Kim and T.H. Lee, J. Power Sources 75 (1998)
Acknowledgement 230.
29. H.Y. Cheh, J. Electrochem. Soc. 118 (1971) 1132.
30. A.J. Bard and L.R. Faulkner, ‘Electrochemical Methods: Funda-
This work was carried out under the framework of the mentals and Applications’ (J. Wiley & Sons, New York, 1980),
Commission of the European Union through the p. 258.
NEMECEL program (Joule III, JOE3-CT97-0063). 31. W. Kraus and G. Nolze, ‘CPD International Union of Crystal-
lography’, Newsletter No. 20 (1998).
32. G.K. Williamson and W.H. Hall, Acta Metall. 1 (1953) 22.
33. W. Vogel, P. Britz, H. Bönnemann, J. Rothe and J. Hormes,
References J. Phys. Chem. B 101 (1997) 11029.
34. T. Frelink, W. Visscher and J.A.R. Veen, J. Electroanal. Chem. 382
1. C. Lamy, J-M. Léger and S. Srinivasan, in J.O’M. Bockris, B.E. (1995) 65.
Conway and R.E. White (Eds), ‘Modern Aspects of Electrochem- 35. A.K. Shukla, C.L. Jackson, K. Scott and R.K. Raman, Electro-
istry’, vol. 34, (Kluwer Academic/Plenum, New York, 2001), chim. Acta 47 (2002) 3401.
chapter 3, pp. 53–118. 36. L. Dubau, C. Coutanceau, E. Garnier, J-M. Léger and C. Lamy,
2. B. Höhlein, P. Biedermann, T. Grube and R. Menzer, J. Power J. Appl. Electrochem. (2003), in press.
Sources 84 (1999) 203. 37. A. Kabbabi, R. Faure, R. Durand, B. Beden, F. Hahn, J-M. Léger
3. C. Coutanceau, M-J. Croissant, T. Napporn and C. Lamy, and C. Lamy, J. Electroanal. Chem. 444 (1998) 41.
Electrochim. Acta 46 (2000) 579. 38. R. Ianiello, V.M. Schmidt, U. Stimming, J. Stumper and
4. S. Motoo, M. Watanabe and N. Furuya, J. Electroanal. Chem. 160 A. Wallau, Electrochim. Acta 39 (1994) 1863.
(1984) 351. 39. V.S. Entina and O.A. Petrii, Sov. Electrochem. 4 (1968) 97.
5. M.S. Wilson and S. Gottesfeld, J. Appl. Electrochem. 22 (1992) 1. 40. T. Iwasita, H. Hoster, A. John-Anaker, W.F. Lin and W. Vielstich,
6. A.K. Shukla, P.A. Christensen, A.J. Dickinson and A. Hamnett, Langmuir 16 (2000) 522.
J. Power Sources 76 (1998) 54. 41. D. Chu and S. Gilman, J. Electrochem. Soc. 143 (1996) 5.