Palladium-based electrodes: A way to reduce platinum content in polymer electrolyte membrane fuel cells

Article abstract of DOI:10.1016/j.electacta.2010.11.101

To decrease the Pt content, a polymer electrolyte membrane fuel cell (PEMFC) was formed using a carbon supported Pd₉₆Pt₄ catalyst as the anode material, and a carbon supported Pd₄₉Pt ₄₇Co₄ catalyst as the cathode material. The as-obtained Pd-based PEMFC with an overall Pd:Pt:Co atomic composition of electrodes (anode + cathode) = 72:26:2 exhibited a performance not too far from that of the fuel cell with the conventional 100% Pt electrodes. With a Pt content of 35 wt% of that of the cell with full Pt electrodes, at a current density of 1 A cm ^-2 the performance loss of the cell with the Pd-based catalysts was only 11%, with 6% ascribed to the anode catalyst and 5% to the cathode catalyst. The maximum power density of the Pd-based cell was 76% of that of the cell with Pt catalysts.

Full text of DOI:10.1016/j.electacta.2010.11.101

Electrochimica Acta 56 (2011) 2299–2305
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Palladium-based electrodes: A way to reduce platinum content in polymer
electrolyte membrane fuel cells
a,b,∗
b
c
b
Ermete Antolini
, Sabrina C. Zignani , Sydney F. Santos , Ernesto R. Gonzalez
a
Scuola di Scienza dei Materiali, Via 25 aprile 22, 16016 Cogoleto, Genova, Italy
Instituto de Química de São Carlos, USP, C. P. 780, São Carlos, SP 13560-970, Brazil
CECS - UFABC, Rua Santa Adélia, 166, Santo André - SP, Brazil
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 August 2010
Received in revised form
To decrease the Pt content, a polymer electrolyte membrane fuel cell (PEMFC) was formed using a car-
bon supported Pd96Pt4 catalyst as the anode material, and a carbon supported Pd49Pt47Co4 catalyst as
the cathode material. The as-obtained Pd-based PEMFC with an overall Pd:Pt:Co atomic composition of
electrodes (anode + cathode) = 72:26:2 exhibited a performance not too far from that of the fuel cell with
the conventional 100% Pt electrodes. With a Pt content of 35 wt% of that of the cell with full Pt electrodes,
2
7 November 2010
Accepted 30 November 2010
Available online 8 December 2010
−2
at a current density of 1 A cm the performance loss of the cell with the Pd-based catalysts was only
1%, with 6% ascribed to the anode catalyst and 5% to the cathode catalyst. The maximum power density
of the Pd-based cell was 76% of that of the cell with Pt catalysts.
1
Keywords:
PEMFC
Palladium
©
2010 Elsevier Ltd. All rights reserved.
PdPtCo/C catalyst
Hydrogen oxidation
Oxygen reduction
1
. Introduction
case, as discussed by Wee et al. in a recent review [2], the reduc-
tion of Pt loading in electrocatalysts can be achieved through an
enhancement of the Pt utilization by increasing the active Pt sites,
thinning the active layer thickness and introducing smaller, carbon-
supported, nanometer-sized, Pt particles. Regarding the point (2),
the partial or total substitution of platinum with palladium seems
a promising way to reduce Pt content [3]. Pt and Pd have very sim-
ilar properties (same group of the periodic table, same fcc crystal
structure, similar atomic size). The cost of palladium is about three
times lower than that of platinum, so it could be a good substitute
for Pt as electrode catalyst in fuel cells. Moreover, Pd is interesting
as it is at least fifty times abundant on the earth than Pt. Pd, as other
platinum-group metals, presents electrocatalytic activity for both
the hydrogen oxidation reaction (HOR) and the oxygen reduction
reaction (ORR), but the HOR and ORR activities of Pd are consider-
ably lower than those of Pt [4–6]. The performance of Pd/C for the
HOR as anode catalyst in PEMFC is very poor [7,8]. It has be found,
however, that by addition of a very low amount (5–10 at.%) of Pt, the
HOR activity of Pd attains that of pure Pt [8,9]. On the other hand, it
has been reported that by addition of a suitable metal, as Co and Fe,
the ORR activity of Pd becomes comparable to that of Pt [10–13].
The activity of these Pt-free Pd-based catalysts, however, seems to
depend on the synthesis method. Indeed, in various papers a con-
siderably lower ORR activity of Pd-based catalysts than Pt was also
reported [14–16]. Raghuveer et al. [14] found that, in the absence
of thermal treatment, the ORR activity of Pd–Co–Au/C, prepared
Recently, high operating efficiency and environmental friendli-
ness polymer electrolyte fuel cells (PEMFCs) have begun to move
from the demonstration phase to commercialization due to the
impressive research effort in recent years. Nevertheless, several
outstanding cost reduction problems and technological challenges
remain to be solved. Among the technological challenges in terms of
PEMFC electrodes, the development of anode electrocatalysts toler-
ant to carbon monoxide and cathode electrocatalysts able to reduce
the overpotential encountered under open circuit conditions and
significantly enhance the exchange current density are the most
significant. One of the barriers to commercialization remains the
prohibitive cost of this technology. The U.S. Department of Energy
set long-term goals for PEMFC performance in a 50 kW stack that
included operation with cathode loadings of 0.05 mg cm or less
of precious metals [1]. The limited supply and high cost of the
Pt used in PEMFC electrocatalysts necessitate a reduction in the
Pt level. Generally, there are two ways to reduce the use of Pt
in PEMFCs, that is (1) Pt electrodes with low Pt content and (2)
total or partial substitution of Pt with other metals. In the former
−
2
∗ Corresponding author at: Scuola di Scienza dei Materiali, Department of Chem-
istry, Via 25 aprile 22, 16016 Cogoleto, Genova, Italy.
E-mail address: ermantol@libero.it (E. Antolini).
0
013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2010.11.101

2
300
E. Antolini et al. / Electrochimica Acta 56 (2011) 2299–2305
both by the borohydride method and the reverse microemulsion
method, was lower than that of Pt/C. Wang et al. [15] found the
ORR overpotential on all of Pd–Co/C electrocatalysts with differ-
ent Pd/Co atomic ratios prepared by a modified polyol reduction is
higher than that on the Pt/C catalyst, indicating that the ORR activity
on Pd-based catalysts is lower than that on the Pt-based catalysts
in pure acidic solution. Finally, Liu and Manthiram [16] observed
Pd–Pt–Co/C), a palladium chloride (PdCl ·2H O, MERCK) solution
2
2
(in the case of Pd–Co/C, Pd–Pt/C and Pd–Pt–Co/C) and cobalt
nitrate (Co(NO ) ·6H O, Aldrich) (for Pd–Co/C and Pd–Pt–Co/C)
3
2
2
were slowly added to the carbon suspension. The suspension was
left to cool at room temperature and the solid filtered and dried
◦
in an oven at 80 C for 1 h. The material was 20 wt% metal on car-
bon. The Pd–Pt–Co/C catalysts in the nominal atomic ratio Pd:Pt:Co
50:30:20 were prepared by two ways: (1) simultaneous impregna-
tion of Pd, Pt and Co precursors on the formic acid impregnated
carbon support (catalyst named PdPtCoS), and (2) impregnation of
Pd and Pt precursors on the support, then impregnation of Co, fol-
that the as-prepared Pd Co/C catalyst, synthesized by a modified
4
polyol reduction process, has lower catalytic activity for ORR than
the as-prepared Pt/C, although all of them have a similar particle
size.
◦
The addition of Pt is another way to increase the ORR of Pd.
Indeed, many works showed that platinum addition increases the
ORR activity of palladium [17–23] and that the dependence of the
ORR activity on the Pt content goes through a maximum. In partic-
ular, Guerin et al. [21] and Ye and Crooks [22] reported an increase
in the activity of Pd with the amount of Pt, following a parabolic
trend with a maximum activity in the composition range of 50–90%
Pt that is potential-dependent. The presence of platinum should
improve the stability of palladium in acid media as well as reinforce
the electrocatalytic activity for the ORR through the synergistic
effects of the metals.
lowed by thermal treatment at 700 C under flowing N2 (catalyst
named PdPtCoT).
2.2. Structural characterization
The atomic ratio of the catalysts was determined by the energy
dispersive X-ray (EDX) technique coupled to a scanning electron
microscopy LEO Mod. 440 with a silicon detector with Be window
and applying 20 keV.
X-ray diffractograms (XRD) of the electrocatalysts were
obtained in a universal diffractometer Carl Zeiss-Jena, URD-6, oper-
ating with Cu k␣ radiation (ꢀ = 0.15406 nm) generated at 40 kV and
Generally the particle size of carbon supported metals increases
going from Pt to Pt–Pd to Pd, independently of the preparation
method [8,17,24–27]. The increase in metal particle size is more
significant for Pd contents >50 at.%. The different dispersion of the
metal nanoparticles on the carbon support is mainly due to their
different nucleus-growth mechanisms of metal nanoparticles in
Pt/C and Pd/C catalysts. Recently Beard et al. [28] used surface mod-
ification of a carbon support preceding deposition of catalytic Pd
seed nuclei to prepare Pt–Pd catalysts by the electroless deposi-
tion (ED) of Pt onto the Pd surface. Functionalization of a carbon
support with nitric acid to form surface carboxylic acid groups is fol-
lowed by pre-treatment in a pH 14 bath to convert the acid moieties
◦
−1
◦
20 mA. Scans were done at 3 min
for 2ꢁ values between 20
◦
and 100 . It has to be remarked that the XRD measurements on
as-prepared and thermally treated catalysts were carried out on
the catalyst powder, whereas the XRD analysis on the catalysts
submitted to RPC was carried out on the electrode.
The in situ X-ray absorption spectroscopy (XAS) measure-
ments were performed in the Pt LIII absorption edges at 0.8 V
vs. the reversible hydrogen electrode (RHE), using a spectro-
electrochemical cell [22]. The working electrodes consisted of
pellets formed with the dispersed catalysts agglutinated with PTFE
−
2
(ca. 40 wt%) and containing 6 mg cm of catalyst. The counter elec-
trode was a Pt screen. This electrode was cut in the center, in order
to allow the free passage of the X-ray beam. Prior to the experi-
ments, the working electrodes were soaked in the electrolyte for at
least 48 h. XAS experiments were made at 0.8 V vs. RHE, after cycling
the electrodes in the range defined by these potentials. Results pre-
sented here correspond to the average of at least two independent
measurements.
−
to the corresponding carboxylate (RCOO ) groups. The negative
surface charge results in an electrostatic attraction between the
carboxylate groups and positively charged Pd2 cations during wet
impregnation to produce smaller catalytic Pd seed nuclei which
act as deposition sites for Pt in the ED process. Using this method
Ohashi et al. [23] prepared various Pd–Pt/C catalysts with a max-
imum Pd/Pt atomic ratio of ca. 1 and a maximum metal (Pd + Pt)
loading on carbon of 13.4 wt%, having a particle size of 3.2 nm.
This preparation method could be valid also for the achievement of
Pd–Pt/C catalysts with small particle size in more severe conditions,
that is, Pd/Pt atomic ratio > 1 and metal loading of 20 wt%.
+
All the XAS experiments were conducted at the D04B-XAS1
beam line in the Brazilian Synchrotron Light Laboratory (LNLS),
Brazil. The data acquisition system for XAS comprised three ion-
ization detectors (incidence I , transmision It and reference Ir). The
0
Palladium-based electrocatalysts have been tested as anode
materials for alcohol oxidation in alkaline direct alcohol fuel cells
reference channel was employed primarily for internal calibration
of the edge positions by using a foil of the pure metal. Nitrogen was
[
29]. In this work we have tested single PEMFCs with Pd-based elec-
used in the I , It and Ir chambers.Transmission electron micrograph
0
trodes, that is, a Pd catalyst with a very low Pt content (5 at.% Pt)
as anode material, and a Pt-free Pd–Co and ternary Pd–Pt–Co cat-
alysts as cathode materials. The performance of the best Pd-based
PEMFC was compared with that of a PEMFC with conventional Pt
electrodes.
(TEM) analysis was carried out at 120 kV using a JEOL JSM-5900LV
microscope.
2.3. Electrochemical characterization
Tests in PEMFCs: For the PEMFC studies, the electrodes were hot
®
◦
pressed on both sides of a Nafion 115 membrane at 125 C and
−
2
for 2 min. The Nafion® membranes were pre-treated
2
. Experimental part
50 kg cm
with a 3 wt% solution of H O , washed and then treated with a
2
2
2.1. Synthesis methods
0.5 M solution of H SO . The geometric area of the electrodes was
2 4
2
−2
4
.6 cm . The metal loading was 0.4 mg cm both at the anode and
Carbon supported Pt, Pd–Pt in the nominal atomic compo-
at the cathode. The Pt/C and Pd–Pt/C (95:5) samples were tested
as anode catalysts, using a commercial 20 wt% Pt/C by E-TEK as the
cathode catalyst. The Pd–Co/C, Pd–Pt–Co/C and Pt/C catalysts were
tested as cathode materials, using the Pd–Pt/C (95:5) sample as the
anode catalyst.
sition 95:5, Pd–Co (75:25) and Pd–Pt–Co (50:30:20) catalysts
were prepared by reduction of metal precursors with formic acid.
An appropriate amount of carbon powder (Vulcan XC-72, Cabot,
2
−1
2
40 m g ) was suspended in 2 M formic acid solution and the
◦
suspension heated to 80 C. Chloroplatinic acid (H PtCl ·6H O,
The polarization experiments in the PEMFC were carried out
2
6
2
◦
Johnson Matthey) solution (in the case of Pt/C, Pd–Pt/C and
galvanostatically (Electronic Load HP 6050A) with the cell at 90 C,

E. Antolini et al. / Electrochimica Acta 56 (2011) 2299–2305
2301
◦
Fig. 2. Polarization curves in single PEMFCs operating at 90 C and 3 atm O2 pressure
Fig. 1. XRD diffractograms of Pd95Pt5/C and Pt/C catalysts.
with Pd96Pt4/C and Pt/C as anode electrocatalysts. Cathode: 20 wt% Pt/C. Anode and
−
2
cathode metal loading 0.4 mg cm
.
◦
using O2 saturated with water at 90 C and 3 atm in the cathode,
amount of Pt remarkably increased the activity of Pd, attaining
the value of pure Pt. The slightly lower performance of the cell
with Pd95Pt5/C as anode catalyst than that of the PEMFC with Pt/C
could be ascribed to the higher particle size of the Pd-based cata-
lyst than Pt. Moreover, as expected, Pd95Pt5/C presented a higher
CO tolerance than Pt/C.
and either pure hydrogen or a mixture of H /100 ppm CO saturated
2
◦
with water at 95 C and 1 atm in the anode. The gases were primary
mixture of 100 ppm carbon monoxide in hydrogen balance, nitro-
gen (99.996%), carbon monoxide (99.5%) and hydrogen or oxygen,
all from White Martins. Before cell test, to activate the catalysts,
the electrodes have been submitted to a repetitive potential cycling
(
1500 cycles) in the range 0.5–1.0 V. The cell polarization data were
◦
collected at 90 C/3 atm O2 pressure.
3.2. Pd-based cathode catalysts
CO stripping experiments were performed on the gas diffusion
electrodes using a potentiostat–galvanostat (Solartron 1285). In
these experiments, the cathode was used as the working electrode,
while the anode was used as both the counter and the RHE. The CO
stripping experiments were carried out in the following way: after
recording a CV in a N2 purged system, CO was admitted to the cell
and adsorbed at 0.075 V for 20 min. The excess CO was eliminated
with N2 gas and the stripping charges determined between 0.075
A Pt-free Pd–Co/C and ternary Pd–Pt–Co/C catalysts have been
used as cathode materials in PEMFCs and compared with a con-
ventional Pt/C. The EDX compositions of Pd–Co/C and Pd–Pt–Co/C
catalysts are reported in Table 1. The amount of Pt in the PdPtCoS
catalyst was lower than the nominal one, as well as the amount
of Co in the PdPtCoT catalyst. Fig. 3 shows the XRD patterns of the
carbon supported Pd–Co/C, Pd–Pt–Co/C and Pt/C electrocatalysts,
and the lattice parameters and the crystallite size of these catalysts
are given in Table 1. Assuming that the lattice parameter of PdPt
alloys obey to Vegard’s law in the entire range of compositions
−
1
and 1.0 V vs. RHE using a scan rate of 10 mV s after correcting the
currents for the background. Being the Pd49Pt47Co used as cathode
4
catalyst, CO was bubbled only on the cathode. Linear sweep curves
were recorded in the range 0.1–1.0 V vs. RHE.
3
. Results
3.1. Pd-based anode catalysts
A PEMFC using a Pd–Pt/C catalyst with a very low Pt content as
anode material was compared with a cell using a conventional Pt/C
catalyst as anode material. As reported by Papageorgopoulos et al.
[
7] and Garcia et al. [8], the performance of Pd/C for the HOR as
anode catalyst in PEMFC is very poor. For this reason we have cho-
sen the nominal Pd:Pt composition 95:5. The value of Pd:Pt atomic
ratio of the Pd–Pt/C catalyst was 96:4, near to the nominal com-
position. Fig. 1 shows the XRD patterns of the carbon supported
Pd–Pt and Pt electrocatalysts. The lattice parameters and the par-
ticle size of these catalysts are given in Table 1. The value of the
lattice parameter (0.3894 nm, slightly higher than that of pure Pd
(
a = 0.3890 nm)) indicates the formation of a PdPt alloy. The parti-
cle size of the Pd–Pt/C catalyst was more than twice with respect
to Pt/C. Fig. 2 shows the current–potential curves for single PEM-
FCs using as anode materials the Pd96Pt /C and Pt/C catalysts, and
4
conventional Pt/C as cathode material. As can be seen in Fig. 2,
in agreement with previous works [8,9], the presence of a small
Fig. 3. XRD diffractograms of Pd75Co25/C, Pd53Pt15Co32/C, Pd49Pt47Co4/C and Pt/C
catalysts.

2
302
E. Antolini et al. / Electrochimica Acta 56 (2011) 2299–2305
Table 1
EDX composition, lattice parameter, crystallite size by XRD and particle size by TEM of Pd–Co/C, Pd–Pt–Co/C and Pt/C catalysts.
Catalyst
Pt/C
EDX composition Pd:Pt:Co
0:100:0
Lattice parameter (2 2 0) (nm)
0.3922
Particle size by XRD (2 2 0) (nm)
4.1
Particle size by TEM (nm)
dn 4.4
dv 5.2
PdPtCoT/C (50:30:20)
49:47:4
0.3905
18
dn 7.7
dv 15* (without the particle
with size 53 nm)
PdPtCoS/C (50:30:20)
Pd75Co25/C
Pd95Pt5/C
53:15:32
75:0:25
96:4:0
0.3879
0.3894
0.3894
9.2
12.0
11.3
from pure Pt to pure Pd, we have calculated the lattice param-
eter of a PdPt alloy in the Pd:Pt composition = 1:1, ca. the Pd:Pt
composition in the PdPtCoT catalyst. The calculated lattice param-
eter (0.3908 nm) was in good agreement with that experimental
value of PdPtCoT (0.3905 nm). Co likely do not alloy or its effect
on the lattice parameter is negligible due to its low amount. The
lattice parameter of the PdPtCoS catalyst, instead, was lower than
that calculated applying Vegard’s law to the PdPt alloy in the Pd:Pt
composition = 3.5:1, that is, the Pd:Pt atomic ration in the PdPtCoS
catalyst. This means that likely a ternary PdPtCo alloy was formed.
The very large particle size of the PdPtCoT/C catalyst is due to the
◦
◦
thermal treatment at 700 C. As the thermal treatment at 700 C has
been carried out under flowing N , cobalt is present in the reduced
2
form Co(0).
Fig. 4 shows the current–potential and power density curves for
a single PEMFC using as cathode materials the activated Pd75Co25/C,
PdPtCoS/C, PdPtCoT/C and Pt/C catalysts, and the Pd96Pt /C cat-
4
alyst as anode material. The cathode performance was in the
order Pt/C > PdPtCoT/C > PdPtCoS/C ꢀ Pd75Co25/C. The performance
of the Pt-free Pd75Co2 /C was poor: the maximum power density
5
−
2
Fig. 5. Dependence of the PEMFC potential at 0.01 A cm on Pt content in the
catalyst.
(
MPD) of the cell with Pd75Co25/C as cathode catalyst was 25% of
MPD of the cell with Pt/C as cathode material. The cell with PdPt-
CoT/C as cathode catalyst, instead, showed a good performance,
with a value of MPD which attained 85% of that of the cell with Pt/C
3.2.1. Comparison of structural characteristics of Pd49Pt47Co /C
4
−
2
as cathode catalyst. In a fuel cell, the potential at 0.01 mA cm
,
and Pt/C
V0.01, is a useful parameter to evaluate the ORR activity [30]. Fig. 5
shows the dependence of V0.01 on Pt content in the catalyst. The
data follow a parabolic law with a maximum for a Pt content of ca.
Among the catalysts tested as cathode material in PEMFCs, PdPt-
CoT/C was the most interesting, so we have investigated more in
detail its characteristics and compared with those of Pt/C.
Fig. 6 shows the X-ray absorption near edge structure (XANES)
spectra at the Pt L3 edge at 0.8 V vs. RHE for the as-prepared PdPt-
CoT/C and Pt/C electrocatalysts, and for the reference Pt foil. A
7
0%, in agreement with the literature results [17,21].
◦
Fig. 4. Polarization and power density curves in single PEMFCs operating at 90
C
and 3 atm O2 pressure with Pd75Co25/C (squares), PdPtCoS/C (circles), PdPtCoT/C (up
triangles) and Pt/C (down triangles) as cathode electrocatalysts. Full symbols: polar-
ization data; open symbols: power density data. Anode: 20 wt% Pd96Pt4/C. Anode
Fig. 6. Pt L3 XANES spectra at 0.8 V vs. RHE for carbon supported Pt and PtPdCoT
electrocatalysts relative to a Pt reference foil in 0.5 mol L H2SO4.
−2
−1
and cathode: catalyst loading 0.4 mg cm
.

E. Antolini et al. / Electrochimica Acta 56 (2011) 2299–2305
2303
Fig. 7. TEM image (a) and histogram of particle size distribution (b) of the Pt/C
Fig. 8. TEM image (a) and histogram of particle size distribution (b) of the PdPtCoT/C
catalyst.
catalyst.
The number averaged particle size, dn, and the volume averaged
significant increase in the intensity of the Pt L3 white line of Pt/C
compared to PdPtCoT/C and the reference Pt foil was observed. Nor-
mally, the intensity of the Pt L3 edge increases with the increase in
Pt d-band vacancy, in this case due to the adsorption of oxygenated
species [31]. The affinity of OH chemisorption on Pt depends on
both the metal particle size and alloying. It is known that there is an
increase in the affinity of OH chemisorption on Pt as the metal par-
ticle size decreases [32] and the amount of M in Pt–M also decrease
in Pt alloys [31]. The lower intensity of the Pt L3 white line of PdPt-
CoT/C compared to that of Pt/C was mainly due to the higher particle
size of PdPtCoT/C than Pt/C, but also to alloying effect.
particle size, d , have been calculated using Eqs. (1) and (2):
v
ꢀ
(
k)nkd
k
dn =
ꢀ
(1)
(2)
(k)nk
ꢀ
(
k)nkd4
k
dv = ꢀ
(k)nkd3
k
where n is the frequency of occurrence of particles with size d . The
k
k
particle sizes of Pt/C obtained from the TEM images were in agree-
ment with those calculated from the XRD (2 2 0) peak, as reported
in Table 1, and the values of dn and dv were similar. In the case
of the PdPtCoT/C catalyst, the discrepancy of the number averaged
size by TEM with the crystallite size by XRD is due to the fact that
the Scherrer formula yields a volume averaged size, to which the
largest particles contribute disproportionately. The higher value of
dv than that of dn confirms the effect of the large particles on the
main particle size. The chemical surface area (CSA) of the metal
particles was calculated using the following equation:
Figs. 7 and 8 show the results of TEM analysis of the as-prepared
Pt/C (Fig. 7) and PdPtCo/C (Fig. 8) catalysts. As can be seen in Fig. 7a,
the distribution of Pt particles on the carbon support is in a narrow
particle size range. The histogram of the particle size distribution
(
Fig. 7b) reflects quantitatively the size distribution in the Pt/C cat-
alyst, with a narrow particle distribution centered at a particle size
of 4.5 nm. Conversely, the image of the PdPtCoT/C catalyst (Fig. 8a)
shows a non-uniform distribution of the particle size, with the
presence of some giant particles (size > 30 nm) together with par-
ticles having a smaller size and in a narrow particle size range. The
histogram of the PdPtCoT/C (Fig. 8b) shows a wide particle distribu-
tion, with the presence of a tail at large particles, with one particle
large than 50 nm, likely formed during the thermal treatment.
3
CSA (m g−1) = 6
× 10
(3)
2
ꢂdn
where ꢂ is the density of Pt (21.4 g cm−3) or the Pd–Pt (1:1) alloy
−
3
(16.8 g cm ). The values of CSA are reported in Table 2.

2
304
E. Antolini et al. / Electrochimica Acta 56 (2011) 2299–2305
Table 2
−
2
Chemical surface area (CSA) and electrochemical surface area (ECSA) of PdPtCoT/C and Pt/C, and potential at 0.01 A cm (ꢃV0.01) and maximum power density (MPD), in
terms of both geometric area and ECSA, of PEMFCs with PdPtCoT/C and Pt/C as cathode catalysts.
CSA (TEM) (m2
g
−1
)
ECSA (CO stripping) (m2
g
−1
)
ꢃV0.01 (mV)
MPD (geometric area) (mW cm
−2
)
−2
MPD (ECSA) (mW cm )
Catalyst
PdPtCoT/C
Pt/C
46.4
63.7
35.6
48.5
904
900
598
710
4.2
3.7
Fig. 9 shows the results of the linear sweep CO stripping exper-
iments. In agreement with the results of Garcia et al. [8], the peak
potential for PdPtCoT/C occurred at a higher value than for Pt/C,
indicative of the stronger CO bonding to Pd. While CO adsorption
on PtPdCoT/C can occur on both Pt and Pd atoms, the oxidation
involves only a single uniform peak. A synergistic effect is thus
evident; the peak is not simply an addition of fractional contribu-
tions of Pt and Pd sites. Electronic modification of the CO adsorption
characteristics on the two metals, due to alloy formation, is clearly
supported by the results shown in Fig. 9. The electrochemical chem-
ical surface area (ECSA) of the metal particles obtained by CO
stripping is reported in Table 2. From the electrochemical sur-
face area (ECSA, by CO stripping) and the chemical surface area
(
CSA, by TEM) the Pt utilization efficiency, i.e. the ratio between
the ECSA and CSA, can be calculated [33,34]. The catalyst utiliza-
tion efficiency was similar, 0.76 and 0.77 for Pt/C and PdPtCoT/C,
respectively, in agreement with the value of 0.72 reported in the
literature [33,34]. As shown in Table 2, the PdPtCoT/C catalyst pre-
sented a specific activity, in terms of V0.01, slightly higher than that
of Pt/C. In agreement with the values of V0.01, while the MPD of the
cell with PdPtCoT/C, expressed in terms of the geometric area, that
is, in terms of mass activity, is lower than that of the cell with Pt/C,
the MPD of the PEMFC with PdPtCoT/C, expressed in terms of the
electrochemical surface area, that is, in terms of specific activity, is
slightly higher than that of the cell with Pt/C (see Table 2).
◦
Fig. 10. Polarization and power density curves of single PEMFCs operating at 90
C
and 3 atm O2 pressure with Pd96Pt4/C and Pd49Pt47Co4/C as anode and cathode
catalysts, respectively, and with conventional Pt/C catalysts. Full symbols: polariza-
tion data; open symbols: power density data. Anode and cathode: catalyst loading
0.4 mg cm−2.
ascribed to the anode and 5% to the cathode. The maximum power
density (MPD) of the Pd-based cell was 76% of the MPD of the cell
with Pt/C catalysts. The lower performance has to be ascribed to
the higher particle size of both the Pd-based anode and cathode
catalysts than Pt/C.
3.3. Comparison of PEMFCs with Pd-based and conventional Pt
electrodes
The polarization and power density curves of PEMFCs with
Pd96Pt /C and PtPdCoT/C as anode and cathode catalysts, respec-
4
tively, and with conventional Pt/C catalysts are shown in Fig. 10. As
can be seen in Fig. 10, the performance of the cell with Pd-based
electrodes was only slightly lower than that of the cell with con-
4
. Conclusions
The performances of PEMFCs with Pd-based catalysts were com-
−
2
ventional Pt/C electrodes. The power densities at 1 A cm
were
pared with that of a PEMFC with conventional Pt/C catalysts. The
power density at 1 A cm
−2
6
23 and 554 mW cm for the cells with Pt/C and Pd-based cata-
−2
of a cell with Pd96Pt /C and Pt/C as
4
lysts, respectively. A performance loss of only 11% occurred, with 6%
anode and cathode catalysts, respectively, was only slightly lower
94%) with respect to the conventional PEMFC with Pt/C catalysts
(
−
2
at both electrodes. The performance at 1 A cm of a quasi Pt-free
PEMFC with Pd96Pt /C as anode catalyst and Pd75Co25/C as cath-
4
ode catalyst (overall Pd:Pt:Co catalyst composition 85.5:2:12.5),
instead, was considerably lower (22%) than that of the cell with
−
2
Pt/C catalysts. Finally, the performance at 1 A cm
of a PEMFC
with Pd96Pt /C as anode catalyst and PdPtCoT/C as cathode cata-
4
lyst (overall Pd:Pt:Co catalyst composition 72:26:2) was 89% than
that of the conventional PEMFC, with a performance loss of 6%
ascribed to the anode catalysts and 5% to the cathode catalyst.
Thus, this work demonstrates that Pd-based catalysts can substi-
tute conventional Pt catalysts in PEMFCs with only a slightly loss
in cell performance, particularly at high current densities. For both
electrodes, the performance loss is due to the higher particle size
of Pd-based catalysts than Pt, decreasing the active surface area
of the catalyst, and negatively influencing their catalytic activity.
To reduce cell performance loss, it is necessary to synthesize Pd-
based catalysts with low particle size. So, further works should
be addressed to develop new synthesis methods of Pd-based cata-
lysts with low Pt content and low particle size, maintaining a high
catalytic activity.
Fig. 9. COads stripping voltammograms of Pt/C and PdPtCoT/C electrocatalysts.

E. Antolini et al. / Electrochimica Acta 56 (2011) 2299–2305
2305
Acknowledgment
[15] W. Wang, D. Zheng, C. Du, Z. Zou, X. Zhang, B. Xia, H. Yang, D.L. Akins, J. Power
Sources 167 (2007) 243.
[
[
16] H. Liu, A. Manthiram, Energy Environ. Sci. 2 (2009) 124.
17] H. Li, G. Sun, N. Li, S. Sun, D. Su, Q. Xin, J. Phys. Chem. C 111 (2007) 5605.
The authors thank the Conselho Nacional de Desenvolvimento
Científico e Tecnológico (CNPq, Proc. 310151/2008-2) for financial
assistance to the project.
[18] T. Lopes, E. Antolini, E.R. Gonzalez, Int. J. Hydrogen Energy 33 (2008) 5563.
[
19] J. Yang, J.Y. Lee, Q. Zhang, W. Zhou, Z.L. Yang, J. Electrochem. Soc. 155 (2008)
B776.
[
20] Y. Xu, X. Lin, J. Power Sources 170 (2007) 13.
References
[21] S. Guerin, B.E. Hayden, C.E. Lee, C. Mormiche, A.E. Russell, J. Phys. Chem. B 110
2006) 14355.
(
[
[
22] H. Ye, R.M. Crooks, J. Am. Chem. Soc. 129 (2007) 3627.
23] M. Ohashi, K.D. Beard, S. Ma, D.A. Blom, J. St-Pierre, J.W. Van Zee, J.R. Monnier,
Electrochim. Acta 55 (2010) 7376.
24] S.A. Grigoriev, E.K. Lyutikova, S. Martemianov, V.N. Fateev, Int. J. Hydrogen
Energy 32 (2007) 4438.
[
1] A.T. Haug, R.E. White, J.W. Weidner, W. Huang, S. Shi, T. Stoner, N. Rana, J.
Electrochem. Soc. 149 (2002) A280.
[
[
[
[
2] J.-H. Wee, K.-Y. Lee, S.H. Kim, J. Power Sources 165 (2007) 667.
3] E. Antolini, Energy Environ. Sci. 2 (2009) 915.
4] M. Maximov, O.A. Petrii, Elektrokhimiya 10 (1974) 1721.
5] M.S. Rau, P.M. Quaino, M.R.G. de Chialvo, A.C. Chialvo, Electrochem. Commun.
[
[
[
25] Z. Liu, L. Hong, M.P. Tham, T.H. Lim, H. Jiang, J. Power Sources 161 (2006) 831.
26] W. Wang, Q. Huang, J. Liu, Z. Zou, Z. Li, H. Yang, Electrochem. Commun. 10
1
0 (2008) 208.
6] J.J. Salvador-Pascual, S. Citalán-Cigarroa, O. Solorza-Feria, J. Power Sources 172
2007) 229.
7] D.C. Papageorgopoulos, M. Keijzer, J.B.J. Veldhuis, F.A. De Bruijn, J. Electrochem.
Soc. 149 (2002) A1400.
(
2008) 1396.
27] J.B. Joo, Y.J. Kim, W. Kim, N.D. Kim, P. Kim, Y. Kim, J. Yi, Korean J. Chem. Eng. 25
2008) 770.
[
[
[
(
(
[
[
[
28] K.D. Beard, J.W. Van Zee, J.R. Monnier, Appl. Catal. B: Environ. 88 (2009) 185.
29] C. Bianchini, P.K. Shen, Chem. Rev. 109 (2009) 4183.
30] S. Mukerjee, S. Srinivasan, M.P. Soriaga, J. Electrochem. Soc. 142 (1995)
[
[
8] A.C. Garcia, V.A. Paganin, E.A. Ticianelli, Electrochim. Acta 53 (2008) 4309.
9] Y.-H. Cho, B. Choi, Y.-H. Cho, H.-S. Park, Y.-E. Sung, Electrochem. Commun. 9
1
409.
31] S. Mukerjee, S. Srinivasan, M.P. Soriaga, J. McBreen, J. Phys. Chem. 99 (1995)
577.
(
2007) 378.
[
[
[
10] J.L. Fernandez, D.A. Walsh, A.J. Bard, J. Am. Chem. Soc. 127 (2005) 357.
11] M.H. Shao, T. Huang, P. Liu, J. Zhang, K. Sasaki, M.B. Vukmirovic, R.R. Adzic,
Langmuir 22 (2006) 10409.
12] M.-H. Shao, K. Sasaki, R.R. Adzic, J. Am. Chem. Soc. 128 (2006) 3526.
13] S. Song, Y. Wang, P. Tsiakaras, P.K. Shen, Appl. Catal. B 78 (2008) 381.
14] V. Raghuveer, P.J. Ferreira, A. Manthiram, Electrochem. Commun. 8 (2006) 807.
4
[
[
32] S. Mukerjee, J. McBreen, J. Electroanal. Chem. 448 (1998) 163.
33] W. Li, W. Zhou, H. Li, Z. Zhou, B. Zhou, G. Sun, Q. Xin, Electrochim. Acta 49 (2004)
[
[
[
1045.
[
34] F. Colmati, E. Antolini, E.R. Gonzalez, Electrochim. Acta 50 (2005) 5496.