Pt-Co alloy nanoparticles synthesized on sulfur-modified carbon nanotubes as electrocatalysts for methanol electrooxidation reaction

Article abstract of DOI:10.1016/j.jcat.2012.05.001

Several Pt-Co catalyst nanoparticles supported on sulfur-modified carbon nanotubes (CNTs) have been synthesized and used for methanol oxidation reaction (MOR) in acidic solutions. The catalysts were prepared by impregnation of the catalyst precursors on the CNTs and then alloying at 800 °C under an atmosphere of H₂/Ar. The results indicated that in spite of high annealing temperature, the catalyst nanoparticles were highly distributed on the support with an average particle size of ～3 nm and narrow size distribution. Among the catalysts, Pt₃Co/CNT showed highest MOR activity. Moreover, the high temperature annealing process resulted in the formation of faceted cubo-octahedral nanocrystals with high stability in acid solutions and electrochemical environments. The results demonstrated the effectiveness of sulfur-modified CNTs as a precursor for the synthesis of catalyst nanoparticles, without the need for tedious pretreatment procedures to modify CNTs or complex equipments to achieve high dispersion of nanoparticles on the CNT support.

Full text of DOI:10.1016/j.jcat.2012.05.001

Journal of Catalysis 292 (2012) 81–89
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Pt–Co alloy nanoparticles synthesized on sulfur-modified carbon nanotubes
as electrocatalysts for methanol electrooxidation reaction
b
R. Ahmadi a, M.K. Amini a, , J.C. Bennett
⇑
a Chemistry Department, University of Isfahan, Isfahan 81744-73441, Iran
b Department of Physics, Acadia University, Wolfville, Nova Scotia, Canada B4P 2R6
a r t i c l e i n f o
a b s t r a c t
Article history:
Several Pt–Co catalyst nanoparticles supported on sulfur-modified carbon nanotubes (CNTs) have been
synthesized and used for methanol oxidation reaction (MOR) in acidic solutions. The catalysts were pre-
pared by impregnation of the catalyst precursors on the CNTs and then alloying at 800 °C under an atmo-
sphere of H2/Ar. The results indicated that in spite of high annealing temperature, the catalyst
nanoparticles were highly distributed on the support with an average particle size of ꢀ3 nm and narrow
size distribution. Among the catalysts, Pt3Co/CNT showed highest MOR activity. Moreover, the high tem-
perature annealing process resulted in the formation of faceted cubo-octahedral nanocrystals with high
stability in acid solutions and electrochemical environments. The results demonstrated the effectiveness
of sulfur-modified CNTs as a precursor for the synthesis of catalyst nanoparticles, without the need for
tedious pretreatment procedures to modify CNTs or complex equipments to achieve high dispersion of
nanoparticles on the CNT support.
Received 9 March 2012
Revised 29 April 2012
Accepted 1 May 2012
Available online 4 June 2012
Keywords:
Sulfur-modified carbon nanotubes
Pt–Co alloy nanoparticles
Methanol oxidation reaction
Direct methanol fuel cell (DMFC)
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
[3–10] systems, among which Pt–Co seems to be more promising.
Methanol oxidation activity and stability of these catalysts, besides
One of the primary concerns associated with direct methanol
fuel cell (DMFC) applications is the development of electrocatalysts
to promote methanol oxidation reaction (MOR). Supported
Pt-based catalysts are currently the most active electrocatalysts
for MOR in practical DMFCs, but pure Pt is susceptible to surface
poisoning by strongly adsorbed CO-like intermediates, the byprod-
ucts of methanol electrooxidation at low temperatures, which
severely limits the conversion efficiency of these fuel cells [1,2].
The electrocatalysts free from poisoning effects are thus desired
in order to realize high conversion efficiency in DMFCs.
It has been shown that alloying of Pt with other transition met-
als results in a remarkable enhancement of the electrocatalytic
activity and poison tolerance of Pt catalysts for MOR [3–6]. Within
this context, various alloys of Pt with other group VIII transition
metals, including the remainder of precious Pt-group metals (Rh,
Pd, Os, Ru, and Ir) and non-precious ones (Ni, Co, and Fe) have been
investigated [1]. The Pt-group elements are, however, among the
most expensive metals, thus, particular interest has been devoted
to alloying Pt with the non-precious metals of group VIII with
the aim of facilitating the commercialization of DMFCs. Promising
results have been reported for Pt–Fe [4,7], Pt–Co [4–9], and Pt–Ni
the type and proportion of the base metal (M), depend on the
degree of alloying M with Pt, electrode pretreatment, and type of
the substrate [6,11].
The catalytic performance and utilization efficiency of Pt-based
electrocatalysts in fuel cells are strongly related to the type of
support [12,13]. The support can promote the formation of highly
distributed catalyst nanoparticles with small size and narrow size
distribution. A good support not only enables high dispersion of cat-
alyst nanoparticles, but also facilitates electron transfer, increases
long-term stability of the catalyst nanoparticles by slowing down
their sintering rate and allows fast mass transport of reactants
and products at the fuel cell electrodes, resulting in better device
performance [1,14]. Carbon blacks are currently the most widely
used supports for the preparation of Pt-based electrocatalysts for
DMFCs. However, the utilization efficiency of the catalysts on these
supports is relatively low because some of the Pt nanoparticles are
trapped in their micropores, rendering them inaccessible to the
reactants [13]. Moreover, due to their low graphitization degree,
carbon blacks are prone to undergo electrochemical corrosion [15].
Carbon nanotubes (CNTs), because of their unique surface struc-
ture, high electronic conductivity, and high-accessible surface area,
together with high mechanical, chemical and electrochemical sta-
bilities, are superior to carbon blacks as catalyst supports for
DMFCs [10,16,17]. Moreover, highly graphitized CNTs are far more
resistant toward oxidation catalyzed by platinum. However, due to
⇑
Corresponding author. Fax: +98 311 6736055.
E-mail address: mkamini@chem.ui.ac.ir (M.K. Amini).
0021-9517/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2012.05.001

82
R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
their chemical inertness and hydrophobic nature, the CNT surfaces
need to be modified to increase the anchoring sites for high and
uniform dispersion of catalyst nanoparticles [18–20], especially
when the conventional impregnation method is used. Generally,
CNTs have been modified by covalent attachment of functional
groups or non-covalent adsorption of modifier on their surfaces
[21,22]. Most of the chemical modifications are based on complex
and/or harsh reaction conditions including corrosive reactants and
tedious work up, which may impair the interesting properties of
CNTs such as their high conductivity and remarkable mechanical
properties [20]. Therefore, developing simpler alternative for mod-
ification of CNTs that provides well-dispersed catalyst nanoparti-
cles of uniform particle size and narrow size distribution without
damaging their unique electronic and structural characteristics
are still a major concern [13,23,24].
at the same experimental conditions, as well as a Pt3Co/CNT cata-
lyst prepared on unmodified nanotubes and a commercial Pt/C
catalyst.
2. Experimental
2.1. Materials
Hexachloroplatinic acid (37 wt% Pt), sulfur, Co(NO3)2ꢁ6H2O, and
methanol were obtained from Merck. MWCNTs from Green Nano-
tech Co., Tehran, Iran (purity P95%, 10–15 nm in diameters, and
10–30 lm in length) were used as the support. Nafion solution
(5 wt% in lower aliphatic alcohols and water) was supplied by
Aldrich. All other chemicals used in this investigation were of
analytical grade from Merck. Double-distilled, deionized water
was used for the preparation of all solutions.
Our previous work involving the use of sulfur-modified CNTs for
the synthesis of Pt/CNT electrocatalysts showed that the dispersion
and size of Pt nanoparticles are remarkably affected by the amount
of sulfur modifier [25]. The modification was based on impregnation
of CNTs with sulfur from toluene solution followed by a melt-coat
step. Then, Pt nanoparticles were supported on the modified CNTs
by wet impregnation method, followed by heat treatment under a
flow of H2/Ar (10% H2) at 350 °C. The Pt/CNT nanocomposite synthe-
sized at S/CNT ratio of 0.3 afforded well-dispersed Pt nanoparticles
on CNTs with an average particle size of less than 3 nm and a narrow
size distribution, as well as high electrochemically active surface
area (EASA) and catalytic activity toward methanol oxidation. Our
results demonstrated the effectiveness of a relatively simple route
for preparation of sulfur-modified CNTs as a precursor for the syn-
thesis of Pt/CNTs, without the need for tedious pretreatment proce-
dures to modify CNTs or complex equipments to achieve high
dispersion of Pt nanoparticles on the support. The common practice
for preparation of the carbon-supported Pt alloy catalyst nanoparti-
cles from the impregnated Pt–M salt solutions is annealing/reduc-
tion at high temperatures under a reducing atmosphere. While
heat treatment generally helps to increase the degree of alloying
of Pt–M catalysts and consequently improves their corrosion resis-
tance [11,26], it often results in undesired particle size growth due
to sintering of the nanoparticles, and a consequent decrease in their
surface area and catalytic activity [27–29]. Therefore, development
of novel synthetic methods based on the simple classical approach
of impregnation and high temperature annealing/reduction method
for the preparation of alloy catalysts with reduced sintering behav-
ior are still in demand.
In the present study, highly dispersed nanoparticle alloys of Pt
with Co (with atomic ratios of 3:1, 1:1 and 1:3) are synthesized
by wet impregnation of the corresponding metal salts on sulfur-
modified CNTs followed by thermal annealing/reduction at
800 °C to achieve a high degree of alloying. This temperature was
selected because the critical temperature for phase transitions to
solid solution (disordered phase) for Pt3Co, which was found to
be the most stable catalyst with highest electrochemical activity
among all the catalysts investigated in this work, is ca. 750 °C
[30]. It should be noted that Watanabe et al. [30] in studying the
stability of ordered and disordered Pt–Co/C catalysts, observed
up to 47% activity degradation in the ordered alloys, compared
with only 1% degradation in the disordered alloys. Also, dissolution
of cobalt from the ordered catalyst was several percentage points
higher than from the disordered alloy. Modification of the support
in the present work is based on sulfur impregnation of CNTs from
toluene solution followed by a melt-coat step. The catalysts are
characterized by high-resolution transmission electron microscopy
(HRTEM), X-ray diffraction (XRD), X-ray photoelectron spectros-
copy (XPS) and inductively-coupled plasma optical emission
spectrometry (ICP-OES). The electrochemical performances of the
Pt–Co/CNTs are investigated and compared with Pt/CNT prepared
2.2. Synthesis of the catalysts
The sulfur-modified CNTs with S/CNT ratios of 0.3 were pre-
pared according to our previously reported method [25]. Briefly
elemental sulfur was ultrasonically dissolved in toluene, and then,
CNTs were added. The mixture was sonicated for 1 h to break
aggregates and disperse the CNTs. Then, the solvent was evapo-
rated at room temperature, and the sample was ground thoroughly
in an agate mortar. The mixture in a sealed vial was heated to
155 °C in an oven for 6 h.
The modified CNTs (S/CNT) were impregnated with Pt(IV) and
Co(II) solutions with different nominal atomic ratios of 3:1, 1:1,
and 1:3 by ultrasonically dispersing the nanotubes in mixture solu-
tion of the metal ions in methanol–water (50/50 by volume).
Typically, for preparing Pt3Co/CNT, 0.186 mmol H2PtCl6ꢁ6H2O and
0.062 mmol Co(NO3)2ꢁ6H2O (corresponding to 40 mg metals to
afford 20 wt% Pt3Co on CNTs) were dissolved in methanol–water
(1:1) solution. Then, 160 mg CNTs were added to the solution,
and the solvent was evaporated in a vacuum oven at 60 °C over-
night. The resulting mixture in a quartz tube was heated under a
flow of H2/Ar (10% H2) to reduce the metal ions and to remove
the sulfur content. The heating was carried out by ramping from
room temperature to 800 °C with a heating rate of 5 °C minꢂ1
and holding at that temperature for 1 h, followed by quenching
to room temperature during a period of 2 h. Pt/CNT and Co/CNT
with 20 wt% of the respective metals were also synthesized on sul-
fur-modified CNTs in a similar way for comparison with the alloy
electrocatalysts. For comparison, also a Pt3Co catalyst was synthe-
sized on unmodified CNTs (Pt3Co/CNTum) under exactly the same
experimental conditions.
2.3. Physicochemical characterization
The XRD analysis of the samples was performed with a Bruker
D8 Advance powder diffractometer using Ni filtered CuK
a radia-
tion (k = 1.54056 Å). The XPS spectra of the supported catalysts
were recorded on a VG Microtech Twin anode XR3E2 X-ray source
and a concentric hemispherical analyzer operated at a base pres-
sure of 5 ꢃ 10ꢂ10 mbar using AlK
a (hm = 1486.6 eV). Peak fits of
all spectra were performed using the Shirley background correc-
tion and Gaussian–Lorentzian peak shapes. Binding energies
(BEs) were referenced to C 1s peak at 284.5 eV. Low-resolution sur-
vey spectra, as well as higher-resolution spectra for C, O, Co, and Pt,
were collected. The morphologies and size of the catalyst nanopar-
ticles were characterized by HRTEM on Jeol 2010F TEM. The chem-
ical compositions of the alloy nanoparticles were measured by
energy-dispersive X-ray spectroscopy (EDS) with the same TEM
instrument using a field emission gun (FEG) in spot analysis mode.

R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
83
Sample preparation for HRTEM examination involved the ultra-
Pt(111)
Pt(200)
C(002)
sonic dispersion of the sample in ethanol and placing a drop of
the suspension on a copper grid covered with perforated carbon
film. The mean particle size and distribution were obtained from
randomly chosen areas in the HRTEM images containing about
250 particles. The exact Pt and Co loading of the Pt–Co/CNT cata-
lysts were determined by ICP-OES (Optima 7300 DV, Perkin-Elmer)
after dissolving the metals in hot aqua regia and centrifuging to re-
move the CNTs. The residual sulfur content of the catalysts was
determined using an elemental analyzer (CHNS-932, LECO).
Electrochemical measurements were taken with an Autolab
PGSTAT-101 instrument (Eco Chemie BV, the Netherland) using
NOVA 1.5 software in a conventional one-compartment three-
electrode glass cell with a modified glassy carbon (GC) working
electrode (0.385 cm2), a Pt gauze counter electrode and a double-
junction Ag/AgCl (sat’d KCl) reference electrode. All potentials are
quoted vs. this reference electrode. The salt bridge was daily filled
with 0.5 M H2SO4 solution. The GC disk electrode was polished to a
Pt(220)
Pt(311)
*
Pt3Co/CNT
Pt/CNT
Pt3Co/CNT
PtCo/CNT
PtCo /CNT
3
Co/CNT
20
30
40
50
2θ
60
70
80
90
/ degree
mirror finish with 0.3 and 0.05 lm alumina pastes, respectively,
and sonicated in acetone/water (50/50, v/v) for 5 min. For prepara-
tion of the catalyst ink, 5 mg Pt/CNTs or Pt–Co/CNTs, 25 mg Nafion
solution, 800 lL absolute ethanol, and 200 lL water were added to
Fig. 1. XRD spectra of Pt and Pt–Co catalyst nanoparticles synthesized on sulfur-
modified CNTs at various Pt to Co atomic ratios after treatment at 800 °C. The Pt3Co
catalyst prepared on pristine CNTs (no sulfur modification) is shown at the top as
Pt3Co/CNTꢄ.
a vial and sonicated for 1 h. For comparison, a catalyst ink was also
prepared with a commercially available 20 wt% Pt/C (Alfa-Aesar).
Then, 20
with 0.385 cm2 surface area to yield a nominal metal loading of
20 g (corresponding to 52
g cmꢂ2 geometric electrode area) on
lL of the catalyst ink was dripped onto the GC electrode
3.2. Physical characterization and structural studies of Pt–Co/CNT
electrocatalysts
l
l
all the working electrodes to facilitate a better comparison be-
tween platinum and different alloy catalysts. After drying the ink
under infrared lamp, a smooth film was formed on the electrode
surface. Then, the electrode was subjected to potential cycling in
deaerated 0.5 M H2SO4 solution between ꢂ0.25 and +1.2 V at a
Fig. 1 shows the XRD spectra of Pt–Co catalysts at various atom-
ic ratios deposited on CNTs compared to that of Pt/CNT. The peak
around 2h of 26° corresponds to (002) planes of graphitized carbon
from CNTs. The pure Pt/CNT catalyst has a face-centered cubic
(FCC) crystallographic structure. It shows diffraction peaks located
at 2h values of 39.6°, 46.3°, 67.4°, and 81.4° which are assigned to
(111), (200), (220), and (311) planes, respectively. Fig. 1 also
shows the diffraction pattern of pure Co/CNT, indicating the origi-
nal metallic peak positions. The diffraction peaks in all the Pt–Co/
CNT catalysts, which shift to higher 2h angles compared to Pt/CNT,
are also indexed to (111), (200), (220), and (311) reflections of
FCC crystals (Fig. 1). The shift of the reflection peaks to higher 2h
angles, which increase with increasing the Co content of the cata-
lysts, indicate lattice contraction due to the incorporation of smal-
ler Co atoms into the FCC crystal structure of Pt and formation of
Pt–Co alloys. For instance, the lattice parameter for Pt3Co was
3.825 Å (compared to 3.915 Å for Pt/CNT and 3.543 Å for Co/CNT
prepared in a similar way), which is in reasonable agreement with
the lattice constant for the corresponding bulk Pt3Co solid solution
(3.831 Å [31]) and indicates a high degree of alloying. No charac-
teristic peaks of metallic Co or its oxides were detected, but their
presence cannot be disregarded because they may be present in
small amounts or even in amorphous forms. Also, no distinctive
superlattice diffraction peaks were observed around 2h of 23.1°,
32.8°, 53.2°, and 58.6°, confirming the formation of disordered
Pt–Co alloys with FCC symmetry.
sweep rate of 50 mV sꢂ1 until
observed.
a stable voltammogram was
The charge for underpotentially deposited hydrogen (Hupd) was
used to evaluate the active surface area of the Pt catalyst, assuming
a value of 2100 mC mꢂ2 for crystalline Pt. The electrocatalytic
activities toward MOR were investigated by recording cyclic vol-
tammograms (CVs) from 0.0 to 1.0 V at a scan rate of 20 mV sꢂ1
in deaerated 1.0 M CH3OH + 0.5 M H2SO4 solution.
3. Results and discussion
3.1. Chemical composition of the catalysts
The exact metal loading and concentration of Pt and Co for dif-
ferent catalysts, as determined by ICP in solutions prepared accord-
ing to Section 2.3, are reported in Table 1. As can be seen, the Pt–Co
atomic ratios obtained from ICP analysis are in good agreement
with the nominal values. The catalysts also showed a residual sul-
fur concentration of 60.4 wt%, as determined by the CHNS ana-
lyzer, indicating a S/CNT wt ratio of 0.005 compared to the initial
value of 0.3; that is, ca. 98.3% of the sulfur has been removed dur-
ing the heat treatment step under H2/Ar (10% H2), which otherwise
may cause undesirable effect on performance of the catalysts.
The broad peaks observed in Fig. 1 for Pt and Pt–Co samples
synthesized on the surface of sulfur-modified CNTs demonstrate
Table 1
Composition of the catalysts with different Pt/Co atomic ratios as measured with ICP-OES, and their crystallite sizes estimated from XRD data.
Catalyst
Nominal Pt:M atomic ratio
Composition from ICP analysis
Crystallite size (nm)
Pt (wt%)
M (wt%)
Pt:M (atomic)
Pt/CNT
–
18.04
16.77
13.63
8.46
–
–
3.60
2.66
2.93
2.95
Pt3Co/CNT
PtCo/CNT
PtCo3/CNT
75:25
50:50
25:75
1.63
4.62
7.84
75.6:24.3
47.2:52.8
24.6:75.4

84
R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
0:9kK
a1
B2h cos hmax
that all the diffracting phases are nanocrystalline in nature. The
average crystallite size for Pt and Pt–Co electrocatalysts synthe-
sized with various atomic ratios was determined from full width
at half maximum (FWHM) of Pt (111) peak which has the highest
intensity, using Scherrer’s equation:
d ¼
where d is the average crystallite size (nm), kk
wavelength (1.54056 Å), hmax is the angle of the considered Bragg
is the X-ray
1
a
a
b
c
80
e
d
60
40
20
0
2
3
4
5
Particle size (nm)
Fig. 2. HRTEM images (a) and (b) showing faceted cubo-octahedral Pt3Co nanoparticles of ca. 3.2 0.6 nm particle size. (c) HRTEM image showing the distribution of Pt3Co
nanoparticles on CNTs. (d) Histogram of particle size distribution of Pt3Co nanoparticles on CNTs. (e) Typical EDS spectrum of Pt3Co nanoparticle on CNT support (the Pt–Co
atomic ratio, 75.8:24.2). (f) SAED pattern of Pt3Co/CNT together with the filament being used for measurement. The inset figure in (b) shows the TEM image of the catalyst
synthesized on unmodified CNTs.

R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
85
reflection in radians, and B2h is the FWHM of the diffraction peak
(radians). The obtained crystallite sizes, presented in Table 1, are
between 2.6 and 3.0 nm for Pt–Co/CNTs synthesized on sulfur-mod-
ified CNTs compared with the value of 3.6 for Pt/CNT and 7.5 nm for
the catalyst prepared on unmodified nanotubes under exactly the
same experimental conditions (Pt3Co/CNTum, the top spectrum in
Fig. 1). Considering that all the catalysts were prepared by impreg-
nation and annealing/reduction at 800 °C, the resulting values,
except that of Pt3Co/CNTum, are considerably smaller than the par-
ticle size expected for the samples treated at such a high tempera-
ture. The relatively larger particles for Pt3Co/CNTum can be related
to instability of Pt nanoparticles on unmodified CNTs.
diffraction (SAED) pattern of Pt3Co/CNT is shown in Fig. 2f for the
aperture being centered on a long filament (ꢀ200 nm length shown
as the inset) of carbon containing approximately 50 nanoparticles.
The ring pattern is consistent with an FCC crystal structure in which
the major diffraction peaks can be indexed to (111), (200), (220),
and (311) reflections. From the TEM images in Fig. 2, it can be seen
that the nanoparticles were uniformly dispersed on the surface of
CNTs with very small particle size and narrow size distribution,
indicating a significantly enhanced nucleation density for the cata-
lyst nanoparticles synthesized on the sulfur-modified nanotubes
and the highly beneficial role of sulfur in generating anchoring sites
on the relatively inert surface of CNTs. In fact, sulfur modification
provides a uniform functional surface, which can effectively pro-
mote the nucleation process on the CNTs, and prevents migration
and coalescence of Pt particles. It is worthwhile to point out that
pristine CNTs do not possess a high number of functional groups
on their surface, and mainly surface defects such as vacancies or
pentagone–heptagon pairs (Stone–Wales defects) formed during
their synthesis, which results in locally enhanced chemical reactiv-
ity of the graphitic nanostructures, can serve as anchoring sites for
metals [20]. This causes Pt to have a very low nucleation density
and, consequently, larger particles which result from aggregation
of smaller nanoparticles during the heat treatment of the composite
under H2/Ar atmosphere.
To investigate the oxidation state and possible changes in BEs of
different catalysts, XPS analyses were performed on Pt3Co/CNT. As
shown in Fig. 3a, the main peaks observed in quick survey scan are
due to Pt 4f, Pt 4d, C 1s, and O 1s. There was no peak for the resid-
ual sulfur S in the survey or high-resolution measurements. Fig. 3b
and c show Pt 4f region of the high-resolution XPS spectra for
Pt3Co/CNT and Pt/CNT catalysts. Each spectrum exhibits a broad
band that can be curve-fitted into three pairs of peaks, all of which
have a spin–orbit splitting of 4f7/2 and 4f5/2 states of ca. 3.3 eV and
relative peak area close to the expected ratio of 4:3. In the case of
Pt3Co/CNT, the majority is present at BE of 71.4 eV (71.0%),
together with the 4f5/2 component at 74.7 eV, corresponding to
Pt in the zero-valent state [39]. As can be seen, Pt is predominately
metallic in Pt3Co/CNT, but the presence of two minor peaks with
4f7/2 at 72.5 eV (16.0%) and 74.2 eV (13.0%), attributed to PtO and
PtO2, respectively, indicates some surface oxidation [40]. The BEs
of the 4f7/2 signals for Pt3Co/CNT and Pt/CNT catalysts are given
in Table 2 together with their respective relative intensities. From
the data in this table, it is found that Pt3Co/CNT has larger Pt0 com-
ponent relative to Pt/CNT, which seems to be responsible, at least
partly, for its enhanced electrochemical activity as will be seen in
Section 3.3. The BE for 4f7/2 peak of Pt/CNT (Fig. 3c and Table 2) cat-
alyst is 0.4–0.5 eV higher than that of bulk Pt (71.1–71.2 eV) [41].
This positive shift may be attributed to final state relaxation (par-
ticle size effect) and Pt–CNT interaction [42,43]. However, the cor-
responding peak for the alloyed Pt3Co/CNT catalyst shifts to lower
BE compared to Pt/CNT. This can be elucidated in terms of the elec-
tronic effect of Co on Pt [44]. This element being more electropos-
itive than Pt would donate electrons to Pt, causing a negative shift
in the BE of the 4f7/2 signal of the Pt0 species. In fact, as can be seen
in Table 2, the electron withdrawing effect of Pt on Co increases the
amount of Pt0 in Pt3Co/CNT catalysts (71.0%) compared with that in
Pt/CNT (60.8%), indicating the oxide-cleansing action of Co on Pt
[44,45]. These effects are expected to enhance the catalytic activity
of Pt for MOR. The ill-defined peak of Co, shown in Fig. 3d at
778.6 eV, can be related to Co 2p3/2 in metallic state, which seems
to have a positive shift to higher BE compared to Co bulk (778.3)
[41].
It is generally known that a noble metal such as Pt can be easily
reduced at lower temperatures than transition metals. Tempera-
ture-programmed reduction (TPR) investigation of Pt–Co/C catalyst
by Hernández-Fernández et al., [32] has shown that Pt species are
reduced to Pt0 at temperatures <300 °C, but Co species are reduced
at higher temperatures. We believe that during the temperature
ramp, first PtCl26ꢂ ions are reduced to form metallic nuclei. The rate
of nucleation is higher for the modified CNTs, because sulfur species
on the CNT surfaces act as anchoring sites for the metal species, and
specially, they have high tendency for noble metals [33–35], which
arise from soft acid–soft base interactions. This behavior was evi-
denced by the smaller particle size obtained for Pt3Co/CNT than
Pt3Co/CNTum. In fact, sulfur species prevent migration and coales-
cence of Pt particles. By increasing temperature to higher values,
which is required to form the alloy, reduction of cobalt and further
reduction of platinum species occur resulting in particle growth.
The observation of some splitting in the XRD pattern of the catalyst
prepared on unmodified CNTs can be related to the presence of
some unalloyed Pt nanoparticles on the CNT surface, which is due
to the lowering of the ratio between nucleation and growth rates.
Many reports indicate that while annealing at high tempera-
tures generally helps to increase the degree of alloying, it also leads
to a growth in crystallite size and a consequent decrease in surface
area of the catalyst nanoparticles [36,37]. For instance, Wang et al.
[29] observed that annealing of Pt3Co/C nanoparticles at 800 °C un-
der Ar/H2 results in agglomerated particles with sizes over 20 nm;
Liu et al. [27] also have shown that particle size of Pt/C was in-
creased from about 3 nm to 6.0 and 7.7 nm upon annealing at
200 and 500 °C, respectively. Similar results have been observed
by other groups [28,32–38]. However, the results of the present
work clearly show that the use of sulfur-modified CNTs in the syn-
thesis of Pt and Pt–Co alloys have a considerable effect in reducing
the sintering behavior of the catalyst nanoparticles. Furthermore,
Pt–Co/CNTs, specially Pt3Co, seem to be more stable toward sinter-
ing than Pt on the same support. This may be due to less uniform
spatial distribution of pure Pt particles on CNTs or reduced sinter-
ing behavior of Pt–Co compared to pure Pt nanoparticles [27].
The results of XRD were further proved by TEM measurements.
Fig. 2 shows typical HRTEM images of Pt3Co/CNT together with par-
ticle size distribution histogram of the nanoparticles. The inset of
Fig. 2c shows the TEM image of the catalyst synthesized on unmod-
ified CNTs, indicating that the nanoparticles are agglomerated and
not well dispersed on the support surface. The nanostructure of typ-
ical Pt3Co particles illustrated in Fig. 2a and b is indicative of faceted
cubo-octahedral nanocrystals. Histogram of particle size distribu-
tion was constructed using about 250 randomly chosen particles
in the HRTEM images. A typical histogram for Pt3Co/CNT is shown
in Fig. 2d. The average particle size was 3.2 0.6 nm for Pt3Co, which
is in agreement with that of XRD (see Table 1). Combined HRTEM/
EDS analysis was performed on several single particles of the sup-
ported Pt3Co catalyst, and a typical EDS spectrum is shown in
Fig. 2e. The results revealed an average Pt–Co atomic ratio of
77.6:22.4, which is close to the composition determined by bulk
analysis of this catalyst using ICP (Table 1). A selected area electron
The XPS spectra of O 1s for the CNT-supported electrocatalysts
are shown in Fig. 3e and f, and the corresponding BEs are presented
in Table 2. The O 1s spectra for both Pt3Co/CNT and Pt/CNT could
be deconvoluted into several components, which can be identified

86
R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
Observed
Fitted
71.4
72.5
74.2
74.7
75.9
(a)
(b)
C1s
O1s
Pt4d
77.3
Pt4f
1000
800
600
400
200
0
84
80
76
72
68
Binding energy (eV)
Binding energy (eV)
Observed
Fitted
71.6
(c)
(d)
72.9
74.8
74.9
76.3
78.0
810
800
790
780
770
760
750
84
80
76
72
68
Binding energy (eV)
Binding energy / eV
Observed
Fitted
530.6 eV
532.7 eV
534.9 eV
Observed
Fitted
530.4 eV
532.5 eV
534.3 eV
(f)
(e)
540
536
532
528
524
540
536
532
528
524
Binding energy (eV)
Binding energy / eV
Fig. 3. XPS spectra of Pt3Co/CNT catalyst, (a) survey scan of the spectral region from 0 to 1000 eV, and the high-resolution spectra for (b) Pt 4f, (d) Co 2p, and (e) O 1s regions.
The Pt 4f and O 1s spectra of Pt/CNT catalyst are shown in (c) and (f), respectively. The spectra were obtained by calibration based on C 1s peak at 284.5 eV.
Table 2
The BEs and relative intensities of different platinum and oxygen species observed in Pt 4f7/2, and O 1s spectra of 20 wt% Pt3Co/CNT and Pt/CNT
electrocatalysts.
Electrocatalyst
Pt3Co/CNT
Pt 4f7/2 BE (eV)
Relative intensity (%)
O 1s BE (eV)
Relative intensity (%)
71.4
72.5
74.2
71.0
16.0
13.0
530.4
532.5
534.4
14.1
70.7
15.3
Pt/CNT
71.6
72.9
74.9
60.8
28.6
10.6
530.6
532.7
534.9
17.3
62.9
19.7
as Pt–O or physically adsorbed oxygen near 530 eV, together with
carbon–oxygen species and functional groups at BEs close to 532
and 534 eV [45,46]. It is interesting to note that the fraction of O
1s peak at 530.6 eV for Pt/CNT is significantly higher than that of
the alloyed electrocatalyst (17.3% vs. 14.1%) (Table 2), which
further confirms the oxide-cleansing action of Co on Pt. High-
resolution C 1s XPS spectrum was observed around 285 eV (not
shown). The principal deconvoluted component at 284.5 eV
(85.9%) indicated primarily graphitic carbon corresponding to
non-functionalized CNT support. Besides the main peak, the peaks
observed at BEs of 285.6 eV (8.8%) and 286.5 eV (6.3%) were due to
carbon species bound to oxygen [47].

R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
87
3.3. Electrochemical performance
The stabilized voltammograms for all the electrodes are shown
in Fig. 4B. All the CV features are characteristic of polycrystalline Pt
catalysts in acidic electrolytes regarding the hydrogen and oxide
regions. A couple of broad oxidation–reduction peaks appeared be-
tween +0.05 and ꢂ0.25 V, which originate from underpotential
adsorption (deposition) and desorption of a monatomic hydrogen
layer on Pt surfaces, followed by sharp cathodic and anodic peaks
for massive hydrogen evolution and oxidation. The intensities of
the adsorption/desorption peaks initially increased with number
of cycles, but remained unchanged after a few cycles. The Hupd
charges (QH) for all the supported catalysts are listed in Table 3
(column 3). As can be seen, the area in the Hupd region decreases
with decreasing Pt content of Pt–Co/CNTs compared to Pt/CNT.
For instance, the value of QH for Pt3Co/CNT is lower than that of
Pt/CNT by 7.5%, because the Pt loading of this catalyst is also lower
than Pt/CNT by about the same amount (7%, as determined by ICP,
see Table 3).
The Pt–Co and Pt catalysts synthesized on sulfur-modified CNTs
were characterized by cyclic voltammetry in order to investigate
their electrocatalytic activity for MOR and calculate their electro-
chemically active surface areas (EASAs). The catalyst-coated elec-
trodes were inserted into deaerated 0.5 M H2SO4 solution, and
cyclic voltammetry was performed to clean the working electro-
catalysts and to activate catalytic sites. Electrodes were cycled in
the range between ꢂ0.25 and 1.2 V at 50 mV sꢂ1 until stable
voltammograms were obtained. Fig. 4A shows the first CVs of
Pt–Co/CNTs, Pt/CNT, and Pt/C catalysts in 0.5 M H2SO4 solution. It
can be seen that among different supported Pt–Co catalysts, only
PtCo3/CNT shows a strong broad oxidation peak during the anodic
potential sweep. This peak which is centered at ꢀ0.9 V is due to
dissolution of the base metal Co, as evidenced by the CV of Co/
CNT shown in the inset of Fig. 4A. No such peak was observed
for PtCo/CNT or Pt3Co/CNT, signifying much less dissolution of Co
metal from these catalysts. Generally, stable CVs with features like
Pt/CNT were seen on Pt3Co/CNT and PtCo/CNT electrodes after the
first few cycles (Fig. 4B). The inset of Fig. 4B typically shows that
the total anodic to cathodic charge ratios (Qa/Qc) for Pt3Co/CNT,
PtCo/CNT, and Pt/CNT are very close to unity after the first few
cycles, indicating stabilization of these electrodes as well as their
higher stability with respect to PtCo3/CNT electrode which shows
higher Qa/Qc in all cases. This may be taken as an indication of
the high alloying degree of Pt3Co/CNT and PtCo/CNT as well as
crystalline structure promoted by the annealing process applied
to these catalysts. These parameters are the key factors in enhanc-
ing the stability of the catalysts [11].
Since the stabilized CVs resemble that of polycrystalline Pt and
Pt/CNT electrodes, it can be assumed that alloying with Co does not
affect the binding of Hupd to the Pt surface atoms in Pt3Co/CNT cat-
alysts. Therefore, the EASA can be calculated from the Hupd charges
determined for the corresponding electrodes using the following
equation:
QH
EASA ðm2 gꢂ1Þ ¼
2100 ꢁ LPt
where LPt represents the platinum loading (g) on the electrode, QH is
the charge associated with hydrogen monolayer adsorption (mC),
and the value of 2100 (mC mꢂ2) is the charge required to oxidize
a monolayer of atomic hydrogen on a unit surface area of polycrys-
talline Pt [48]. The calculated EASAs, expressed as specific values for
different electrocatalysts, are listed in Table 3. It is interesting to
note that the EASA value decreased considerably for PtCo/CNT and
PtCo3/CNT with respect to Pt/CNT, but it did not change significantly
for Pt3Co/CNT. Although the Pt loading for Pt3Co/CNT was lower
15
Pt/CNT
Pt
45
Co/CNT
3
30
15
0
PtCo/CNT
PtCo
/CNT
3
Pt/C com.
10
5
than that of Pt/CNT (16.77 compared to 18.04 lg for Pt/CNT, Table
3), their behaviors toward upd of hydrogen and the corresponding
EASA values are close to that of Pt/CNT. This can be related, at least
in part, to the smaller crystallite size of Pt3Co/CNT with respect to
Pt/CNT (see Table 1). Besides this, it is also possible that Pt3Co/
CNT do expose more Pt to the surface than expected from its
stoichiometry, in which case, the presence of Pt skin or partial
Pt-skin layer should also be considered. Unfortunately, we were
not able to confirm this by HRTEM, and the XPS signal of the base
metal was too weak to be used for this purpose. The Pt/CNT and
Pt3Co/CNT synthesized with sulfur-modified nanotubes have higher
EASA compared to the commercial Pt/C (Table 3). These observa-
tions, in line with TEM and XRD results, clearly indicate the benefi-
cial effect of sulfur-modified nanotubes for the preparation of Pt
nanoparticles with smaller particle size and more uniform disper-
sion, in spite of using a high temperature (800 °C) for the anneal-
ing/reduction of the catalysts. This is in contrast to previous
reports in which, generally, high temperature treatment resulted
in undesirable increase in mean particle size and in many cases
broad particle size distribution [26–38], which generally resulted
in a significant reduction in the active electrochemical surface area
and, consequently, decreased Pt mass–based activity of the catalyst
nanoparticles.
-0.4
0.0
0.4
0.8
1.2
1.6
E (V)
0
(A)
-5
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
E (V vs Ag/AgCl)
PtCo /CNT
3
2.5
2.0
1.5
1.0
(B)
Pt/CNT
6
2
PtCo/CNT
Pt Co/CNT
3
0
4
8
16
Cycle no.12
Pt/CNT
-2
-6
Pt Co/CNT
3
PtCo/CNT
PtCo /CNT
3
The electrocatalytic activity of Pt–Co/CNTs for MOR was inves-
tigated by recording CVs in deaerated 0.5 M H2SO4 aqueous
solution containing 1 M CH3OH at room temperature. The CVs of
Pt–Co/CNTs, normalized by the Pt mass (mass activity), are shown
in Fig. 5. For comparison, the CVs of Pt/CNT and the commercial Pt/
C catalysts are also included. The CV features are typical of meth-
anol oxidation on Pt/C electrocatalysts in acidic media with two
anodic peaks, one peak in the forward sweep between 0.6 and
Pt/C com.
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
E (V vs Ag/AgCl))
Fig. 4. The first (A) and stabilized (B) CVs of Pt–Co/CNT catalysts synthesized at
different Pt–Co atomic ratios, Pt/CNT, and commercial Pt/C in deaerated 0.5 M
H2SO4 solution. Scan rate: 50 mV sꢂ1. Insets: (A) CV of Co/CNT (20% Co) and (B) ratio
of total anodic to cathodic charges for different electrodes during cycling.

88
R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
Table 3
Electrochemical parameters obtained from cyclic voltammetry for Pt–Co/CNT, Pt/CNT, and Pt/C electrocatalysts.
Catalysts
Pt loading (lg)
QHupd (mC)
EASA (m2 gꢂ1
)
Mass activity (A gꢂ1
)
If/Ib ratio
Eonset (V)
Pt/CNT
18.04
16.77
13.63
8.46
4.93
4.56
2.99
1.80
2.69
130.1
129.5
104.5
101.3
64.0
971.3
1302.1
1113.5
1084.0
399.3
0.70
0.83
0.92
0.98
0.75
0.240
0.212
0.231
0.268
0.298
Pt3Co/CNT
PtCo/CNT
PtCo3/CNT
Pt/C com.
20.0
Pt Co
/CNT
3
1600
Pt Co/CNT
Pt/CNT
Pt/C Com.
3
PtCo
PtCo
/CNT
3
1800
/CNT
Pt/CNT
Pt/C Com.
1200
800
400
0
1400
1000
600
0.0
0.2
0.4
0.6
0.8
1.0
200
0
100
200
300
400
500
E (V vs. Ag/AgCl)
Cycle number
Fig. 5. The CVs of Pt–Co/CNTs synthesized at various Pt–M atomic ratios, together
with those of Pt/CNT and commercial Pt/C in deaerated 0.5 M H2SO4 + 1 M CH3OH
Fig. 6. Long-term electrochemical stabilities of different electrodes in deaerated
solution. Scan rate: 20 mV sꢂ1
.
0.5 M H2SO4 + 1.0 M CH3OH solution shown as peak current vs. cycle number. Scan
rate: 50 mV sꢂ1
.
0.7 V corresponding to methanol oxidation, and another one in the
reverse sweep between 0.4 and 0.6 V due to oxidative removal of
incompletely oxidized intermediates formed in the forward sweep.
The onset potential and forward peak current density in a cyclic
voltammetry curve are often used to estimate the performance of
an electrocatalyst for MOR. Higher forward peak current density
indicates faster rate of an electrochemical reaction, and lower onset
potential causes smaller overpotential and faster electrochemical
kinetics. The alloying effect of Co in Pt3Co/CNT electrocatalyst is
demonstrated by lower onset potential for MOR as presented in
Table 3. As can be seen, the onset for Pt3Co/CNT occurs at an anodic
potential of 212 mV which is significantly lower than those of Pt/
CNT (240 mV) and commercial Pt/C (298 mV). Moreover, the effect
of Co metal serving as Pt promoter is duly reflected by the higher
MOR mass activity of Pt3Co/CNT compared to Pt/CNT and Pt/C. As
can be seen in Fig. 5 and Table 3, the mass activity of Pt3Co/CNT is
higher than that of Pt/CNT by 34% and much higher than commer-
cial Pt/C by 226%. According to XPS results (Table 2), Pt is in more
reduced state in Pt3Co/CNT (71.0%) than in Pt/CNT (60.8%) and
would accordingly lead to a more efficient activation of adsorbed
methanol [9,49]. Highest electrocatalytic activity of Pt3Co/CNT is
in line with the more reduced state of Pt in this catalyst due to
alloying with Co metal. Considerable enhancement of the catalytic
activity of Pt3Co/CNT with respect to Pt/C is attributed to the high
dispersion of the nanoparticles on CNTs and indicates the beneficial
effect of using sulfur-modified CNTs as a precursor for the synthesis
of this electrocatalyst.
than both Pt/CNT and Pt/C. The higher If/Ib value indicates that
more intermediate carbonaceous species are oxidized to CO2
during the forward scan, suggesting that the alloying of Pt with
Co leads to less poisoning of Pt by the CO-like intermediates
formed in MOR. The positive role of Co metal in increasing the cat-
alyst resistance to poisoning effects can be clearly seen in Table 3,
where higher concentrations of Co in PtCo/CNT and PtCo3/CNT
increase the If/Ib ratio, although these catalysts have much less
overall MOR activities than Pt3Co/CNT, which can be related to
much lower concentration of Pt in these catalysts. The observation
that the peak potentials for both forward and reverse scans shifted
to less anodic potentials for the alloy catalysts also support the
positive role of the base metal on performance of the catalysts.
Among different catalysts, Pt3Co/CNT showed higher electrocat-
alytic activity for MOR; therefore, its long-term stability was inves-
tigated in 1.0 M CH3OH + 0.5 M H2SO4 solution using CV method,
and the results were compared with those of Pt/CNT and the com-
mercial Pt/C electrocatalysts. From the results shown in Fig. 6, it
can be observed that the peak currents for all of the electrocata-
lysts decrease gradually with successive scans. The mass activity
of Pt3Co/CNT is generally higher than that of Pt/CNT and commer-
cial Pt/C even after 500 scans.
4. Conclusions
In this paper, we have demonstrated the use of sulfur-modified
CNTs for the synthesis of highly stable bimetallic Pt–Co alloy nano-
particles as catalysts for methanol electrooxidation. The catalysts
were prepared by the simple wet impregnation method followed
by heat treatment at high temperature (800 °C) in a reducing atmo-
sphere to promote complete alloying. The modified support pro-
vided highly distributed metal anchoring sites for nucleation of
metal nanoparticles on the surface of CNTs, affording high and
uniform dispersion of Pt and Pt alloy nanoparticles. The high
temperature annealing process resulted in the formation of faceted
cubo-octahedral alloy nanocrystals with an average particle size of
An important factor for evaluating an electrocatalyst is its toler-
ance to poisoning effects of carbonaceous species formed during
MOR, which is defined as the ratio of the forward anodic peak cur-
rent (If) to the reverse anodic peak current (Ib) [8]. A higher value of
the If/Ib implies better oxidation of methanol to carbon dioxide in
the forward scan and less poisoning of the electrocatalysts by the
CO-like intermediates. The If/Ib values for all the electrocatalysts
are listed in Table 3. The ratio is 0.75 for the commercial Pt/C
electrocatalyst, which is in good agreement with that reported pre-
viously [8,50], but its value is significantly higher for Pt3Co/CNT

R. Ahmadi et al. / Journal of Catalysis 292 (2012) 81–89
89
[17] W. Yang, X.L. Wang, F. Yang, C. Yang, X.R. Yang, Adv. Mater. 20 (2008) 2579–
2587.
ꢀ3 nm and narrow size distribution by virtue of the modified sup-
port which prevents their sintering during thermal annealing at
800 °C and increases their stability in acid solutions and electro-
chemical environments. Among all the investigated supported cat-
alysts, Pt3Co/CNT exhibited highest MOR activity, CO tolerance, and
long-term cyclability. The improved electrochemical response and
facile synthesis make the present method promising for the prepa-
ration of various CNT-supported well-dispersed pure metal and al-
loy nanoparticles of small size and narrow size distribution. The
modification process of the support can also be extended to other
types of substrates such as carbon nanofibers and mesoporous car-
bon materials.
[18] A.L.M. Reddy, N. Rajalakshmi, S. Ramaprabhu, Carbon 46 (2008) 2–11.
[19] M.S. Saha, A. Kundu, J. Power Sources 195 (2010) 6255–6261.
[20] N. Karousis, N. Tagmatarchis, D. Tasis, Chem. Rev. 110 (2010) 5366–5397.
[21] V. Selvaraj, M. Alagar, K.S. Kumar, Appl. Catal. B: Environ. 75 (2007) 129–138.
[22] X. Peng, S.S. Wong, Adv. Mater. 21 (2009) 625–642.
[23] B. Yue, Y.W. Ma, H.S. Tao, L.S. Yu, G.Q. Jian, X.Z. Wang, X.S. Wang, Y.N. Lu, Z. Hu,
J. Mater. Chem. 18 (2008) 1747–1750.
[24] S. Jiang, Y. Ma, G. Jian, H. Tao, X. Wang, Y. Fan, Y. Lu, Z. Hu, Y. Chen, Adv. Mater.
21 (2009) 4953–4956.
[25] R. Ahmadi, M.K. Amini, Int. J. Hydrogen Energy 12 (2011) 7275–7283.
[26] C.W.B. Bezerra, L. Zhang, H. Liu, K. Lee, A.L.B. Marques, E.P. Marques, H. Wang,
J. Zhang, J. Power Sources 173 (2007) 891–908.
[27] Z. Liu, C. Yu, I.A. Rusakova, D. Huang, P. Strasser, Top Catal. 49 (2008) 241–250.
[28] H. Schulenburg, E. Müller, G. Khelashvili, T. Roser, H. Bönnemann, A. Wokaun,
G.G. Scherer, J. Phys. Chem. C 113 (2009) 4069–4077.
[29] C. Wang, G. Wang, D. van der Vliet, K.C. Chang, N.M. Markovic, V.R.
Stamenkovic, Phys. Chem. Chem. Phys. 12 (2010) 6933–6939.
[30] M. Watanabe, K. Tsurumi, T. Mizukami, T. Nakamura, P. Stonehart, J.
Electrochem. Soc. 141 (1994) 2659–2668.
[31] W.B. Pearson (Ed.), A Handbook of Lattice Spacings and Structures of Metals
and Alloys, vol. 1, Pergmon, Oxford, 1964.
[32] P. Hernández-Fernández, S. Rojas, P. Ocón, J.L. Gómez de la Fuente, P. Terreros,
M.A. Peña, J.L. García-Fierro, Appl. Catal. B: Environ. 77 (2007) 19–28.
[33] J.T. Miller, D.C. Koningsberger, J. Catal. 162 (1996) 209–219.
[34] T. Nakamura, T. Ohana, M. Ishihara, M. Hasegawa, Y. Koga, Diamond Rel.
Mater. 16 (2007) 1091–1094.
Acknowledgments
This work was financially supported by the Research Council
and Center of Excellence for Catalyst and Fuel Cell of the University
of Isfahan, and Iranian Nanotechnology Initiative Council.
References
[35] J. Liu, A.G. Rinzler, H. Dai, J.H. Hafner, R.K. Bradley, P.J. Boul, A. Lu, T. Iverson, K.
Shekimov, C.B. Huffman, F. Rodriguez-Macias, Y.S. Shon, T.R. Lee, D.T. Colbert,
R.E. Smalley, Science 280 (1998) 1253–1256.
[36] H. Liu, W. Li, A. Manthiram, Appl. Catal. B: Environ. 90 (2009) 184–194.
[37] L. Xiong, A. Manthiram, J. Electrochem. Soc. 152 (2005) A697–A703.
[38] M.K. Jeon, Y. Zhang, P.J. McGinn, Electrochim. Acta 54 (2009) 2837–2842.
[39] D.Q. Yang, B. Hennequin, E. Sacher, Chem. Mater. 18 (2006) 5033–5038.
[40] J.R. Croy, S. Mostafa, L. Hickman, H. Heinrich, B.R. Cuenya, Appl. Catal. A 350
(2008) 207–216.
[1] J. Zhang (Ed.), PEM Fuel Cell Electrocatalysts and Catalyst Layers, Springer-
Verlag, London, 2008.
[2] Y.J. Gu, W.T. Wong, Langmuir 22 (2006) 11447–11452.
[3] K.W. Park, J.H. Choi, B.K. Kwon, S.A. Lee, Y.E. Sung, H.Y. Ha, S.A. Hong, H. Kim, A.
Wieckowski, J. Phys. Chem. B 106 (2002) 1869–1877.
[4] E. Antolini, J.R.C. Salgado, E.R. Gonzalez, Appl. Catal. B: Environ. 63 (2006) 137–
149.
[5] H. Yang, J. Zhang, K. Sun, S. Zou, J. Fang, Angew. Chem. Int. Ed. 49 (2010) 6848–
6851.
[6] S. Papadimitriou, S. Armyanov, E. Valova, A. Hubin, O. Steenhaut, E. Pavlidou, G.
Kokkinidis, S. Sotiropoulos, J. Phys. Chem. C 14 (2010) 5217–5223.
[7] C.T. Hsieh, J.Y. Lin, J. Power Sources 188 (2009) 347–352.
[8] L. Liu, E. Pippel, R. Scholz, U. Gösele, Nano Lett. 9 (2009) 4352–4358.
[9] J. Zeng, J.Y. Lee, Int. J. Hydrogen Energy 32 (2007) 4389–4396.
[10] Z.C. Wang, Z.M. Ma, H.L. Li, Appl. Surf. Sci. 254 (2008) 6521–6526.
[11] E. Antolini, J.R.C. Salgado, E.R. Gonzalez, J. Power Sources 160 (2006) 957–968.
[12] T. Hyeon, S. Han, Y.E. Sung, K.W. Park, Y.W. Kim, Angew. Chem., Int. Ed. 42
(2003) 4352–4356.
[13] Y. Mu, H. Liang, J. Hu, L. Jiang, L. Wan, J. Phys. Chem. B 109 (2005) 22212–
22216.
[14] C. Coutanceau, S. Brimaud, C. Lamy, J.M. Leger, L. Dubau, S. Rousseau, F. Vigier,
Electrochim. Acta 53 (2008) 6865–6880.
[41] C.D. Wagner, A.V. Naumkin, A. Kraut-Vass, J.W. Allison, C.J. Powell, J.R. Rumble
Jr., NIST X-ray Photoelectron Spectroscopy Database 20, Version 3.5 (web
version), updated 2007.
[42] Y.T. Kim, T. Mitani, J. Catal. 238 (2006) 394–401.
[43] C. Bittencourt, M. Hecq, A. Felten, J.J. Pireaux, J. Ghijsen, M.P. Felicissimo, P.
Rudolf, W. Drube, X. Ke, G. Van Tendeloo, Chem. Phys. Lett. 462 (2008) 260–264.
[44] A.K. Shukla, M. Neergat, P. Bera, V. Jayaram, M.S. Hegde, J. Electroanal. Chem.
504 (2001) 111–119.
[45] A.S. Arico, A.K. Shukla, H. Kim, S. Park, M. Min, V. Antonucci, Appl. Surf. Sci. 172
(2001) 33–40.
[46] G. Zhang, S. Sun, D. Yang, J.P. Dodelet, E. Sacher, Carbon 46 (2008) 196–205.
[47] S. Banerjee, S.S. Wong, J. Phys. Chem. B 106 (2002) 12144–12151.
[48] S. Trasatti, O.A. Petrii, Pure Appl. Chem. 63 (1991) 711–734.
[49] H.C.S. Aislinn, M.H. Josephine, K.Y.K. Sherman, I.B. Viola, J. Phys. Chem. B 108
(2004) 689–695.
[15] J. Qi, S. Yan, Q. Jiang, Y. Liu, G. Sun, Carbon 48 (2010) 163–169.
[16] Y.T. Kim, K. Ohshima, K. Higashimine, T. Uruga, M. Takata, H. Suematsu, H.
Mitani, Angew. Chem., Int. Ed. 45 (2006) 407–411.
[50] X.B. Ge, R.Y. Wang, P.P. Liu, Y. Ding, Chem. Mater. 19 (2007) 5827–5829.