Investigation of carbon-supported Pt and PtCo catalysts for oxygen reduction in direct methanol fuel cells

Article abstract of DOI:10.1016/j.electacta.2009.03.070

Carbon-supported Pt and Pt₃Co catalysts with a mean crystallite size of 2.5 nm were prepared by a colloidal procedure followed by a carbothermal reduction. The catalysts with same particle size were investigated for the oxygen reduction in a di

Full text of DOI:10.1016/j.electacta.2009.03.070

Electrochimica Acta 54 (2009) 4844–4850
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Investigation of carbon-supported Pt and PtCo catalysts for oxygen reduction in
direct methanol fuel cells
S. Siracusanoa, A. Stassia, V. Baglioa,1, A.S. Aricòa,1, F. Capitaniob, A.C. Tavaresb,∗,1
a CNR-ITAE, Via Salita Santa Lucia Sopra Contesse 5, 98126 Messina, Italy
b Institut National de la Recherche Scientifique-Énergie, Matériaux et Télécommunications, 1650 Boulevard Lionel Boulet, Varennes, Québec, Canada, J3X 1S2
a r t i c l e i n f o
a b s t r a c t
Article history:
Carbon-supported Pt and Pt3Co catalysts with a mean crystallite size of 2.5 nm were prepared by a colloidal
procedure followed by a carbothermal reduction. The catalysts with same particle size were investigated
for the oxygen reduction in a direct methanol fuel cell (DMFC) to ascertain the effect of composition.
The electrochemical investigations were carried out in a temperature range from 40 to 80 ◦C and the
methanol concentration feed was varied in the range 1–10 mol dm−3 to evaluate the cathode performance
in the presence of different conditions of methanol crossover. Despite the good performance of the Pt3Co
catalyst for the oxygen reduction, it appeared less performing than the Pt catalyst of the same particle size
for the cathodic process in the presence of significant methanol crossover. Cyclic voltammetry analysis
indicated that the Pt3Co catalyst has a lower overpotential for methanol oxidation than the Pt catalyst,
and thus a lower methanol tolerance. Electrochemical impedance spectroscopy (EIS) analysis showed that
the charge transfer resistance for the oxygen reduction reaction dominated the overall DMFC response
in the presence of high methanol concentrations fed to the anode. This effect was more significant for
the Pt3Co/KB catalyst, confirming the lower methanol tolerance of this catalyst compared to Pt/KB. Such
properties were interpreted as the result of the enhanced metallic character of Pt in the Pt3Co catalyst
due to an intra-alloy electron transfer from Co to Pt, and to the adsorption of oxygen species on the more
electropositive element (Co) that promotes methanol oxidation according to the bifunctional theory.
© 2009 Elsevier Ltd. All rights reserved.
Received 15 January 2009
Received in revised form 26 March 2009
Accepted 28 March 2009
Available online 7 April 2009
Keywords:
Direct methanol fuel cells
Oxygen reduction
Methanol tolerance
Polarizations
ac-impedance spectroscopy
Cyclic voltammetry
1. Introduction
gen reduction and methanol oxidation occurring simultaneously,
reduces the cell voltage [1].
Portable power sources are one of the most promising appli-
cations of direct methanol fuel cells (DMFCs) [1]. Methanol is a
liquid fuel, easy to handle, with very high specific energy density
(∼6 kWh/kg), and DMFCs have intrinsically higher energy density
and improved autonomy than batteries. In addition, DMFCs have
the potential for instantaneous re-charge (refuelling). High energy
density requires the utilization of high concentration of methanol
so the size of the fuel reservoir in the DMFC can be reduced [2].
However, the fuel cell operation with concentrated methanol is
seriously limited by the permeation of methanol across the poly-
mer electrolyte membrane (methanol crossover) which results in a
significant loss in the fuel and voltage efficiencies. Current DMFCs
operating at low temperature use high loadings of Pt catalyst in
the cathode to enhance the oxygen reduction reaction (ORR) in the
presence of methanol crossover. The methanol adsorbs on the Pt
sites at the cathode, and a mixed potential, resulting from the oxy-
One approach to circumvent this problem is to develop novel
Pt-based catalysts capable of enhancing the oxygen reduction reac-
tion but with limited activity, or even inert, for methanol oxidation.
Pt-based alloy catalysts, such as Pt-M (M = Fe, Ni, Co, Cr, Cu) have
been proposed and tested as methanol tolerant cathode catalysts
for DMFCs based on the following: (i) the second metal could block
the methanol adsorption on Pt due to a dilution effect [3–10] since
the dissociative adsorption of methanol requires the existence of
several adjacent Pt ensembles [1,11], and (ii) binary Pt-based alloy
catalysts exhibit an enhanced electrocatalytic activity for the ORR
in comparison with Pt alone [12–15] due to the increased Pt d-
band vacancy (electronic factor) [16] and by a favourable Pt–Pt
interatomic distance (geometric effect) [17].
There is a vast literature on methanol tolerant Pt-M alloy cath-
ode catalysts for DMFCs and it has been recently reviewed by
Antolini et al. [7,18]. There is some disagreement in the literature
over the improvement of methanol tolerance and DMFC perfor-
mance through the formation of Pt-M alloy catalysts [1,7,19], and
the reasons could be in the different alloys compositions, or in the
use of different synthetic methods resulting in different degrees of
alloying, surface composition and particle size. The large majority
∗
Corresponding author. Tel.: +1 450 929 8947; fax: +1 450 929 8102.
E-mail address: tavares@emt.inrs.ca (A.C. Tavares).
1
ISE member.
0013-4686/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2009.03.070

S. Siracusano et al. / Electrochimica Acta 54 (2009) 4844–4850
4845
of the DMFC studies on bimetallic catalysts for ORR were carried
out with diluted methanol at the anode (1–2 mol dm−3, except for
one case where the authors went up to 4 mol dm−3 [4]), with pres-
surized oxygen feed at a cathode and at temperatures higher than
60 ◦C. Although some of these experimental conditions are more
favourable for the ORR, they are often far from those envisaged
for portable applications. Finally, it is difficult to draw conclusions
about the effect of the second metal on the methanol tolerance of a
Pt catalyst when both composition and particle size in the catalytic
systems are different.
the powder X-ray diffraction (XRD) patterns on a Philips X-pert 3710
X-ray diffractometer using Cu K␣ radiation operating at 40 kV and
30 mA. The peak profile of the (2 2 0) reflection in the face centered
cubic structure was obtained by using the Marquardt algorithm and
used to calculate the crystallite size by using the Debye–Sherrer
equation. The morphological characterization was carried out by
transmission electron microscopy (TEM) analysis using a FEI CM12
microscope.
2.3. Electrodes fabrication and characterization
The aims of the present work were: (i) to evaluate the perfor-
mance of a state of the art Pt-Co catalyst for PEFC fuel cells [12]
in DMFCs operating at low temperature and with concentrated
methanol, and (ii) to try to clarify the effect of the second metal
(Co) in the alloy on the methanol tolerance of Pt cathode catalyst.
For this purpose a Pt-Co alloy catalyst characterized by a high con-
centration of metallic phase (50 wt.%) on carbon black (Ketjenblack)
and an optimum mean crystallite size for the ORR [20] was syn-
thesised using the carbothermal reduction method. A 50 wt.% Pt
on Ketjenblack catalyst with the identical particle size was also
prepared using the same method and used for comparison. Gas
diffusion electrodes with Pt loading of 3 mg cm−2 were fabricated
using the two catalysts, assembled with Nafion 117 membrane and
a PtRu anode, and tested in a DMFC. Under such circumstances, any
difference in the methanol tolerance or DMFC performance can be
directly ascribed to the cathode composition effect.
2.3.1. DMFC
The in-house prepared Pt/KB or Pt3Co/KB catalysts were
mixed with 15 wt.% Nafion ionomer (Ion Power, 5 wt.% solution)
and deposited by a doctor blade technique onto an LT-ELAT
gas-diffusion layer (E-TEK, USA). For the anode a commercial unsup-
ported Pt-Ru (1:1 atomic ratio) catalyst (Johnson-Matthey) was
mixed with 15 wt.% Nafion ionomer (Ion Power, 5 wt.% solution)
and deposited by a doctor blade technique onto the HT-ELAT gas-
diffusion layer (E-TEK, USA). The Pt loading was 3 mg cm−2 on both
anode and cathode. Nafion 117 (Ion Power) was used as electrolyte
membrane. Membrane-electrode assemblies (MEAs) were formed
by a hot-pressing procedure [25] and subsequently installed in a
fuel cell test fixture of 5 cm2 active area (Fuel Cell Technologies).
This latter was connected to a fuel cell test station (model 850C
from Scribner Associates) equipped with a FRA unit. The single
cells were equilibrated with the humidified gases at room temper-
ature, and for each MEA two cycles of galvanostatic polarizations
were recorded between 25 and 80 ◦C with 1 mol dm−3 of aque-
ous methanol solution fed to the anode chamber of the DMFC
through a peristaltic pump (Gilson); humidified air, pre-heated at
the same temperature of the cell, was fed to the cathode. Atmo-
spheric pressure in the anode and cathode compartments was
used for all experiments. Reactant flow rates were 2 ml min−1 and
350 ml min−1 for methanol and air stream, respectively. The MEAs
performance improved during the first 2 days of operation up to
reach a steady-state behaviour. Only the data-set recorded under
steady-state conditions is presented.
Galvanostatic polarization curves and short-term stability tests
(ca. 2 h) were also recorded for different methanol concentrations
(1, 2, 5 and 10 mol dm−3) at 40 ◦C. Electrochemical impedance spec-
troscopy (EIS) measurements were carried out as a function of the
methanol concentration and were used to evaluate differences in
the series and charge transfer resistances between the two MEAs;
additional EIS measurements were done with lower and higher air
flow rate (220 and 420 ml min−1) and with pure oxygen fed at the
cathode (81 ml min−1 to keep the same oxygen stoichiometry with
respect to air at 350 ml min−1). EIS spectra were recorded in the
potentiostatic mode, applying a sinusoidal signal with an amplitude
of 10 mV and a frequency in the 10 kHz to 10 mHz range.
2. Experimental
2.1. Catalysts preparation
A 50 wt.% Pt/Ketjenblack (Pt/KB) was synthesised through the
sulfite-complex route [20,21]. For this purpose, chloroplatinic acid
(Engelhard) was used to prepare the Na6Pt(SO3)4 precursor. Chloro-
platinic acid was dissolved in distilled water and the pH of the
solution was adjusted to 7 by adding Na2CO3 (Aldrich). Subse-
quently, NaHSO3 (Aldrich) was added to the solution to obtain
a white precipitate of Na6Pt(SO3)4, which was filtered, washed
copiously with hot distilled water, and dried in an oven at 80 ◦C.
Ketjenblack EC carbon black (BET area 850 m2/g) was suspended
in distilled water and agitated in an ultrasonic water bath at about
80 ◦C to form a slurry. The appropriate amount of Na6Pt(SO3)4 was
successively added to the slurry. The Pt sulfite complex was decom-
posed by adding drop wise a 40% H2O2 solution (Carlo Erba) at
a temperature of 80 ◦C, which resulted in vigorous gas evolution.
A colloidal PtOx/KB was obtained in this way. The metallic Plat-
inum supported on carbon (50 wt.% Pt/KB) was finally obtained
by carbothermal reduction in inert (Ar) atmosphere at 600 ◦C. A
50 wt.% Pt-Co/Ketjenblack catalyst with a Pt3Co atomic composi-
tion (Pt3Co/KB) was prepared using the same colloidal PtOx/KB
followed by impregnation with cobalt nitrate [12,22,23]. After the
cobalt precursor impregnation step, a high temperature carbother-
mal reduction at 600 ◦C in inert (Ar) atmosphere was used to obtain
the carbon-supported PtCo alloy [12]. The resulting catalyst was
pre-leached in perchloric acid (0.5 mol dm−3) at 85 ◦C to remove
the non-alloyed cobalt [24].
2.4. Half cell characterization
For these studies another series of electrodes were prepared as
described above but the catalyst inks were applied directly onto
carbon cloth substrate (Ballard 1071 HCB) to investigate methanol
oxidation under flooded electrode conditions. The Pt loading was
3 mg cm−2 for both electrodes.
2.2. Catalysts characterization
The half cell characterization was performed in a conventional
three-electrode cell (Metrohm) with a platinum wire as counter
electrode and a Ag/AgCl reference electrode. All experiments were
carried out at room temperature. The electrochemical cell was con-
nected to an Autolab PGSTAT30 Potentiostat/Galvanostat equipped
with FRA2 module for impedance spectroscopy. The Pt/KB and
Pt3Co/KB electrodes were characterized by means of cyclic voltam-
metry (CV) in H2SO4 0.5 mol dm−3 saturated with nitrogen (high
X-ray fluorescence (XRF) analysis was used to determine the
Pt/Co atomic ratio for the Pt3Co/KB catalyst. The XRF analysis was
carried out by a Bruker AXS S4 Explorer spectrometer operating at a
power of 1 kW and equipped with a Rh X-ray source, a LiF 220 crystal
analyzer and a 0.12◦ divergence collimator. The structural character-
ization of the Pt/KB and Pt3Co/KB catalysts was done by recording

4846
S. Siracusano et al. / Electrochimica Acta 54 (2009) 4844–4850
Fig. 1. X-ray diffractograms of Pt/KB and Pt3Co/KB catalysts.
purity 5.0), and in the presence of methanol 0.5 mol dm−3. Cyclic
voltammograms were recorded at a sweep rate of 20 mV s−1
.
3. Results and discussion
3.1. Catalysts characterization
The XRF analysis of the Pt3Co/KB catalyst confirmed a Pt/Co
atomic ratio of 3. Fig. 1 reports the XRD patterns of the Pt/KB
and Pt3Co/KB catalysts. Both diffractograms show the typical face
centered cubic (fcc) crystallographic structure of platinum and
characteristic of the Pt (75 at.%)–Co (25 at.%) alloys [26–28]; no
extra peaks related to other crystalline phases were detected for
the Pt3Co/KB catalyst. Assuming only the presence of a fcc structure
the degree of alloying was calculated from the lattice contraction
determined by XRD and by using the Vegard’s law for the solid solu-
tion between these two metals [26,28]. A high degree of alloying
was found for the Pt3Co/KB catalyst, as indicated by the signifi-
cant shift of the reflection peaks to higher (2ꢀ) values due to the
decrease of the lattice parameter (Table 1). This is, however, slightly
smaller than that reported in the literature for similar composition
and crystallographic structure [8,29]. A good agreement between
the Pt/Co atomic ratio determined independently by XRF and XRD
methods was found for the in-house prepared PtCo/KB catalyst.
The XRD patterns show broad reflection peaks indicating that
both catalysts are formed by fine metal particles. The mean crys-
tallite sizes were determined by line broadening analysis (Table 1).
The mean diameters of Pt and Pt3Co crystallites are 2.6 and 2.5 nm,
respectively. The morphology of both Pt/KB and Pt3Co/KB catalysts
was analysed by TEM. As shown in Fig. 2a and b, the two cata-
lysts consist of fine particles with almost identical dimensions, in
accordance to the XRD analysis, distributed homogeneously on the
carbon support, even for such high metal loading on carbon. Simi-
lar particle size and morphology in the catalysts under investigation
are important aspects since any difference in the electrochemical
activity towards the oxygen reduction and methanol tolerance can
be related mainly to the compositional effect. In this work it will
be possible to do a straightforward comparison between the com-
positions of the two catalysts since both have the same particle
size.
Fig. 2. TEM micrographs of (a) 50% Pt/KB and (b) 50% Pt3Co/KB catalysts.
3.2. Electrochemical characterization: direct methanol fuel cell
The performance of the Pt/KB and Pt3Co/KB cathode catalysts
was thus investigated in a DMFC with 1 mol dm−3 methanol fed at
the anode and air fed at the cathode. Fig. 3a compares the polar-
ization behaviour of two single cells equipped with the Pt/KB and
Pt3Co/KB cathode catalysts at 40 and 80 ◦C. At 40 ◦C and at low
current density (in the activation region) the Pt3Co/KB catalyst per-
forms better than the Pt/KB catalyst, as expected from the enhanced
catalytic activity for ORR of this alloy catalyst with respect to plat-
inum [12,14,29]. The overall performance of both fuel cells is similar
within the experimental error (ca. 27 mW cm−2). At 80 ◦C and low
current density the Pt3Co/KB catalyst appears to show a higher
activity for ORR than Pt/KB, whereas the overall performance of the
fuel cell with Pt/KB catalyst is superior (45.0 mW cm−2 for Pt3Co/KB
and 56.4 mW cm−2 for Pt/KB). Since both fuel cells have the same
anode and membrane, and since both catalysts have the same mor-
phology and particle size, all variations between the polarization
curves can be attributed to differences in the chemical composition
between the two oxygen reduction catalysts. The rate of methanol
crossover increases with the fuel cell operating temperature [1],
and a lower methanol tolerance of Pt3Co/KB catalyst with respect to
Pt/KB catalyst could be the possible cause for the distinct behaviour
observed with increasing temperature.
In order to investigate this hypothesis, additional polarization
curves were recorded at 40 ◦C for various methanol concentrations,
and up to 10 mol dm−3 of methanol. The results are reported in
Fig. 4a–c for 2, 5 and 10 mol dm−3 methanol solutions. In the activa-
tion region the Pt3Co/KB catalyst performs better for 1–5 mol dm−3
methanol concentrations, but for 10 mol dm−3 the two catalysts
Table 1
Physicochemical characteristics of Pt/KB and PtCo/KB catalysts.
Catalysts
A220 (nm)
PSD (nm)
Pt/Co ratio (XRD)
Pt/Co ratio (XRF)
50% CoPtKB
50% PtKB
0.383
0.392
2.5
2.6
3.2
–
3.0
–

S. Siracusano et al. / Electrochimica Acta 54 (2009) 4844–4850
4847
Fig. 3. (a) Polarization and (b) power density curves of single DMFCs with Pt/KB and
Pt3Co/KB cathode catalysts; measurements done at 40 and 80 ◦C, with 1 mol dm−3
MeOH solution and air (atmospheric pressure).
have similar performance. This could be due to a higher cat-
alytic activity for ORR of the alloy catalyst compared to bare Pt,
although the methanol tolerance characteristics have also to be
taken into account, in particular at high methanol concentration.
However, at high current densities the performance of the fuel
cell equipped with the Pt/KB catalyst is superior as indicated by
the higher cell voltage and lower slope in the E–j plots. A simi-
lar result was observed by Baglio et al. [23] with a DMFC equipped
with a Pt-Co/Vulcan XC-72 alloy catalyst operating at 60 ◦C and with
1 mol dm−3 methanol solution.
Fig. 5 reports on the variation of the open circuit voltages and
maximum power densities with the methanol concentration. The
open circuit voltage (OCV) of the DMFC equipped with the Pt3Co/KB
catalyst is always higher with respect to the values recorded for
the Pt/KB based cell confirming the enhanced activity for the ORR
of the alloy catalyst. The OCV decreases as a function of methanol
concentration due to the increase of methanol crossover. The max-
imum power density of the DMFC equipped with the Pt3Co/KB
catalyst decreases monotonically with the methanol concentra-
tion from 27.6 to 11.6 mW cm−2. For the DMFC equipped with
the Pt/KB cathode catalyst the maximum power density is sig-
nificantly higher in the 2–10 mol dm−3 methanol concentration
range, and reaches a maximum value of ca. 31 mW cm−2 with 2 and
5 mol dm−3 methanol solutions. The results shown in Figs. 4 and 5
seem to support our hypothesis on the lower methanol tolerance
of the Pt3Co/KB catalyst with respect to the Pt/KB catalyst.
Fig. 4. Polarization curves of single DMFCs with Pt/KB and Pt3Co/KB cathode cat-
alysts recorded at 40 ◦C, with air (atmospheric pressure) and different methanol
concentrations (a) 2 mol dm−3, (b) 5 mol dm−3 and (c) 10 mol dm−3
.
air at the cathode. Fig. 6a shows the current density–time profiles
recorded for the DMFC equipped with the Pt3Co/KB catalyst, and
Fig. 6b compares the current density values recorded at 100 min
as a function of the methanol concentration and cathode catalyst
type. The chrono-amperograms of the fuel cell with Pt/KB cathode
catalysts are equivalent to those reported in Fig. 6a and therefore
not shown. For both DMFCs, the current density decreases slightly
for 1 and 2 mol dm−3 methanol solutions, while it increases slightly
for 5 and 10 mol dm−3 methanol solutions.
It is hypothesized that, in the presence of low methanol concen-
tration in the anode feed, the cathode is negatively affected by the
continuous methanol permeation through the membrane during
operation. The oxidation of methanol permeated through the mem-
brane produces a mixed potential at the cathode. In the presence
of large methanol concentration, the elevated methanol crossover
Both methanol fuel cells were subjected to short-term stability
tests at 40 ◦C. The cell voltage was set to 0.3 V and tests were run
with 1, 2, 5 and 10 mol dm−3 methanol solutions at the anode and

4848
S. Siracusano et al. / Electrochimica Acta 54 (2009) 4844–4850
Fig. 5. Open circuit voltage and maximum power density recorded for two sin-
gle DMFCs with Pt/KB and Pt3Co/KB cathode catalysts in function of the methanol
concentration.
causes the occurrence of large oxidation current at the cathode;
this counteracts the oxygen reduction process producing a further
lowering of the cathode potential. Thus, the performance decreases
significantly when the methanol concentration is increased from
1–2 to 10 mol dm−3. The comparison between the current density
values recorded at 100 min confirms what found previously: the
performance of both DMFCs in 1 mol dm−3 methanol is practically
identical, but the DMFC equipped with Pt/KB catalyst performs bet-
ter in more concentrated methanol (higher methanol tolerance of
this catalyst) than the DMFC equipped with the Pt3Co/KB catalyst.
Fig. 7. (a) Nyquist plots of two single DMFCs equipped with the Pt/KB and Pt3Co/KB
cathode catalysts under a cell voltage Ecell = 0.3 V, with 5 and 10 mol dm−3 methanol
solutions and with air or oxygen (atmospheric pressure) and at 40 ◦C; (b) Nyquist
plots of the single DMFCs equipped with the Pt3Co/KB cathode catalysts in function of
the cell voltage and air flow (atmospheric pressure), and with 5 mol dm−3 methanol
solutions at 40 ◦C; the dashed lines are just a guide line for the readers and a few
frequencies are indicated in a qualitative way.
The maximum performance for the fuel cell equipped with the
Pt/KB catalyst is in the 2 and 5 mol dm−3 methanol concentration
range.
The polarization studies were complemented with EIS analysis
and cyclic voltammetry to investigate more in-depth the catalysts’
behaviour. The impedance of the two single DMFC cells was mea-
sured as a function of the methanol concentration with both air and
oxygen feed. Fig. 7a shows the typical impedance spectra of the two
DMFCs recorded at a cell potential of 0.3 V, under different condi-
tions of methanol concentration (5 and 10 mol dm−3) and cathode
feed (air and pure oxygen).
The series resistance (Rs) measured with the different methanol
concentrations and varying the cathode fed was similar for both
DMFCs. The total DMFC impedance spectra profiles result from the
overlapping of two distorted semicircles. The high frequency semi-
circle occurs in the 10 kHz to 10 Hz frequency range and the low
frequency semicircle occurs in the 10 Hz to 10 mHz frequency range.
The amplitude of the second semicircle at low frequency is always
larger in the Pt3Co/KB based cell (Fig. 7a). Since both DMFCs are
identical except for the cathode catalyst, it can be concluded that the
low frequency arc is related to the cathode. As shown in Fig. 7a, the
amplitude of the low frequency arc decreased significantly when
the cell impedance was recorded with pure oxygen. The impedance
spectrum of a DMFC cathode includes the contribution of the charge
transfer impedance of ORR at higher frequency, oxygen diffusion at
low frequency and methanol crossover through the inductive loop
at very low frequency [30–32]. Additional tests performed under
different cell voltages demonstrated that the amplitude of the low
Fig. 6. (a) Chrono-amperometric experiments carried out with the DMFC equipped
with the Pt3Co/KB catalyst for various methanol concentrations (1, 2, 5,
10 mol dm−3); (b) current density recorded after 100 min in function of the methanol
concentration and cathode catalyst type; the cell voltage (Ecell) was 0.3 V, all exper-
iments were performed at 40 ◦C and air (atmospheric pressure) was used for the
cathode.

S. Siracusano et al. / Electrochimica Acta 54 (2009) 4844–4850
4849
Fig. 8 shows that for the Pt3Co/KB alloy catalyst the methanol
oxidation peak in the forward and reverse scans are positioned at a
significant lower potential with respect to the Pt/KB catalyst indi-
cating that the Pt3Co/KB has a much higher activity for methanol
oxidation and therefore a lower methanol tolerance. In addition,
the current density for methanol oxidation is significantly higher
for the Pt3Co/KB electrode due to the promoting effect of the elec-
tropositive element in adsorbing oxygen species necessary for the
removal of the adsorbed methanolic residues.
The Pt-Co alloy catalysts are promising catalysts for the ORR in
PEMFCs. The catalysts investigated in this work have already been
examined for the ORR; the PtCo catalyst has shown superior per-
formance than Pt and better resistance to degradation [24]. PtCo
catalysts have been sometime proposed as methanol tolerant cath-
ode catalyst based on dilution effect of Pt to decrease methanol
chemisorption. Yet, the presence of electropositive elements (like
Co) alloyed to Pt decreases the Pt 5 d-band vacancy [34]. An elec-
tron transfer from Co to Pt in a carbon-supported PtCo alloy has
been reported in a previous paper [35]. As clearly evidenced by the
present cyclic voltammetry results, no methanol oxidation occurs
when Pt is covered by a strong oxide layer. Furthermore, the onset of
methanol adsorption in the reverse scan occurs first (higher poten-
tials) for the Pt3Co alloy. This indicates that the reduction process of
the oxide layer on the alloy catalyst is shifted to higher potentials.
In fact, without the reduction of the oxide layer, no methanol oxi-
dation may occur. Thus, Pt is less covered by strong oxide species on
the Pt3Co as previously observed by surface analysis studies [34].
The larger metallic character of Pt in the alloy catalyst favours the
adsorption of methanolic species and the cleavage of the C H bonds
[1]. Besides, according to the bifunctional mechanism for methanol
oxidation, the role of the second element in the Pt alloy is to promote
OH adsorption on the catalyst surface at lower potentials favouring
methanol oxidation [1]. In the literature, it has been observed that
the addition of Co to Pt has two opposite effects [7], and the Co
content in the alloy seems to determine which effect is preponder-
ant: low Co content has a negative effect on the methanol oxidation
(thus a positive effect on the methanol tolerance), instead high Co
content increases the activity of the catalysts for methanol oxida-
tion. However, a significant enhancement of oxygen reduction is
observed for PtCo catalystswithsuitable Co contentand high degree
of alloying [12,24].
Fig. 8. Cyclic voltammograms of Pt/KB and Pt3Co/KB catalyst electrodes in
0.5 mol dm−3 H2SO4 + 1 mol dm−3 methanol (25 ◦C, 20 mV s−1); current normalised
to the electrodes geometric area.
frequency arc decreases with decreasing cell voltage; for cell volt-
age higher than 0.4 V, the amplitude decreases with increasing air
flow while, for cell voltage lower than 0.3 V, the amplitude is less
and almost not affected by the air flow rate (Fig. 7b). The ampli-
tude of the low frequency semicircle increases with the methanol
concentration likely due to an increasing methanol crossover, and
increases more for the Pt3Co/KB cathode catalyst as expected from
a less methanol tolerant catalyst.
It was also seen that the inductive loop related to methanol
crossover disappears at low cell voltage. It is well known that
methanol crossover in a DMFC decreases significantly with an
increase of current density due to the decrease of methanol con-
centration gradient at the anode–electrolyte interface [1]. Under
significant methanol crossover, a high flow rate of air is necessary to
regenerate the catalyst sites [31] and this could explain the decrease
in the amplitude of the semicircle with the flow rate for high cell
voltage (higher crossover) and its independence at low cell voltage
(lower crossover).
Although spectra fitting, quantification and interpretation are
still under investigation, a comparison and qualitative interpre-
tation of the present ac-impedance spectra have shown that the
charge transfer resistance for the ORR occurring in the low fre-
quency range dominates the overall DMFC impedance response. In
addition it can be hypothesized that the impedance for the ORR in
the presence of methanol is higher for the Pt3Co/KB cathode cata-
lyst due to a large effect of the mixed potential being the oxidation
of the methanol permeated through the membrane faster on this
catalyst.
The Pt3Co/KB alloy catalyst used in this work, although promis-
ing for the ORR [24], is more active for methanol oxidation and
thus less tolerant to methanol during the ORR, than the Pt catalyst
prepared in the same way.
4. Conclusions
Fig. 8 shows the cyclic voltammograms of both Pt/KB and
Pt3Co/KB catalysts based gas diffusion electrodes, recorded in
nitrogen saturated 0.5 mol dm−3 H2SO4 + 0.5 mol dm−3 methanol
solution. In the forward scan, the electrode coverage by CO-like
residues strongly adsorbed on the surface occurs at low potentials,
and these species are subsequently stripped off the surface only at
high potentials [33]. The anodic peak occurring between 1.0 and
1.2 V vs. Ag/AgCl in the forward scan is related to the diffusion-
controlled oxidation process of methanol to CO2 [33]. At potentials
of 1.5–1.6 V vs. Ag/AgCl, the oxygen evolution becomes prevailing
and the electrode surface is covered by a strong oxide layer. The
anodic peak occurring at about 0.5 V vs. Ag/AgCl in the reverse scan
is also related to the methanol oxidation; however, it occurs on
a Pt-oxide free surface after the PtO reduction to metallic Pt. In
the reverse scan, and after the PtO reduction, a labile adsorption
of OH species occurs on the surface and the coverage of methano-
lic species is smaller with respect to the forward scan since the
electrode was previously covered by a strong oxide layer [33].
The performance of carbon-supported Pt and Pt3Co cathode cat-
alysts has been investigated in direct methanol fuel cells equipped
with a conventional Nafion 117 membrane. The Pt3Co catalyst
shows better performance than a Pt catalyst of similar crystallite
size and morphology only at low temperatures and in the presence
of a low methanol concentration feed at the anode. Such conditions
correspond to low methanol permeation through the membrane.
This indicates poor methanol tolerance for this catalyst. This aspect
is further confirmed by the cyclic voltammetry analysis, previous
studies of the electronic properties of the Pt3Co alloy as compared
to Pt, and EIS experiments under different conditions. From ac-
impedance, it is derived that the charge transfer resistance for the
ORR dominates the overall DMFC impedance response. This effect
is larger for the Pt3Co/KB catalyst, indicating a lower methanol
tolerance of this catalyst compared to Pt/KB. However, the develop-
ment of methanol impermeable membranes could provide better
chances of using PtCo catalyst as cathodes in DMFCs due to the
promising activity towards oxygen reduction.

4850
S. Siracusano et al. / Electrochimica Acta 54 (2009) 4844–4850
[14] U.A. Paulus, A. Wokaun, G.G. Scherer, T.J. Schmidt, V. Stamenkovic, V. Rad-
Acknowledgements
milovic, N.M. Markovic, P.N. Ross, J. Phys. Chem. B 106 (2002) 4181.
[15] A.M. Castro Luna, A. Bonesi, W.E. Triaca, V. Baglio, V. Antonucci, A.S. Aricò, J.
Solid State Electrochem. 12 (2008) 643.
S. Siracusano acknowledges the Italian National Council of
Research for the financial support of a short-term-mobility period
at the Institut National de la Recherche Scientifique in Varennes
(Québec, Canada). A.C. Tavares acknowledges the financial support
from NSERC (Discovery Grant) and FQRNT (Nouveaux Chercheurs)
agencies.
[16] A. Hamnett, Catal. Today 38 (1997) 445.
[17] S. Mukerjee, S. Srinivasan, J. Electroanal. Chem. 357 (1993) 201.
[18] E. Antolini, T. Lopes, E.R. Gonzalez, J. Alloy Compd. 461 (2008) 253.
[19] H.G. Petrow, R.J. Allen, US Patent 3, 992, 331 (1976).
[20] N. Giordano, E. Passalacqua, L. Pino, A.S. Aricò, V. Antonucci, M. Vivaldi, K.
Kinoshita, Electrochim. Acta 36 (1991) 1979.
[21] A.S. Aricò, V. Baglio, A. Di Blasi, E. Modica, P.L. Antonucci, V. Antonucci, J. Elec-
troanal. Chem. 557 (2003) 167.
[22] V. Baglio, A.S. Aricò, A. Stassi, C. D’Urso, A. Di Blasi, A.M. Castro Luna, V.
Antonucci, J. Power Sources 159 (2006) 900.
References
[23] V. Baglio, A. Stassi, A. Di Blasi, C. D’Urso, V. Antonucci, A.S. Aricò, Electrochim.
Acta 53 (2007) 1361.
[24] A. Stassi, E. Modica, V. Antonucci, A.S. Aricò, Fuel Cells, doi:10.1002/
fuce.200800062, available on-line 20 January 2009.
[25] A.S. Aricò, A.K. Shukla, K.M. el-Khatib, P. Cretì, V. Antonucci, J. Appl. Electrochem.
29 (1999) 671.
[26] B.C. Beard, P.N. Ross, J. Electrochem. Soc. 137 (1990) 3368.
[27] L. Xiong, A. Manthriram, J. Mater. Chem. 14 (2004) 1454.
[28] M. Lucariello, N. Penazzi, E. Arca, G. Mulas, S. Enzo, Mater. Chem. Phys. 114
(2009) 227.
[1] A.S. Aricò, S. Srinivasan, V. Antonucci, Fuel Cells 1 (2001) 133.
[2] S.K. Kamarudin, W.R.W. Daud, S.L. Ho, U.A. Hasran, J. Power Sources 163 (2007)
743.
[3] A.K. Shukla, R.K. Raman, N.A. Choudhury, K.R. Priolkar, P.R. Sarode, S. Emura, R.
Kumashiro, J. Electroanal. Chem. 563 (2004) 181.
[4] K. Scott, W. Yuan, H. Cheng, J. Appl. Electrochem. 37 (2007) 21.
[5] A. Stassi, C. D’Urso, V. Baglio, A. Di Blasi, V. Antonucci, A.S. Arico, A.M. Castro
Luna, A. Bonesi, W.E. Triaca, J. Appl. Electrochem. 36 (2006) 1143.
[6] H. Yang, C. Coutanceau, J.-M. Leger, N. Alonso-Vante, C. Lamy, J. Electroanal.
Chem. 576 (2005) 305.
[7] E. Antolini, J.R.C. Salgado, E.R. Gonzalez, Appl. Catal. B 63 (2006) 137.
[8] J.R.C. Salgado, E. Antolini, E.R. Gonzalez, Appl. Catal. B 57 (2005) 283.
[9] W. Li, W. Zhou, H. Li, Z. Zhou, B. Zhou, G. Sun, Q. Xin, Electrochim. Acta 49 (2004)
1045.
[29] H.A. Gasteiger, S.S. Kocha, B. Sompalli, F.T. Wagner, Appl. Catal. B 56 (2005)
9.
[30] J.T. Mueller, P.M. Urban, J. Power Sources 75 (1998) 139.
[31] P. Piela, R. Fields, P. Zelenay, J. Electrochem. Soc. 153 (2006) A1902.
[32] K. Furukawa, K. Okajima, M. Sudoh, J. Power Sources 139 (2005) 9.
[33] B.D. McNicol, D.A.J. Rand, K.R. Williams, J. Power Sources 83 (1999) 15.
[34] S. Mukerjee, S. Srinivasan, M.P. Soriaga, J. Electrochem. Soc. 142 (1995) 1409.
[35] A.S. Aricò, A.K. Shukla, H. Kim, S. Park, M. Min, V. Antonucci, Appl. Surf. Sci. 172
(2001) 33.
[10] M. Neergat, A.K. Shukla, K.S. Gandhi, J. Appl. Electrochem. 31 (2001) 373.
[11] H.A. Gasteiger, N.M. Markovic, P.N. Ross, E.J. Cairns, Electrochim. Acta 39 (1994)
1825.
[12] A.S. Aricò, A. Stassi, E. Modica, R. Ornelas, I. Gatto, E. Passalacqua, V. Antonucci,
J. Power Sources 178 (2008) 525.
[13] R.C. Koffi, C. Coutanceau, E. Garnier, J.M. Leger, C. Lamy, Electrochim. Acta 50
(2005) 4117.