Synthesis of bimetallic PdAu nanoparticles for formic acid oxidation

Article abstract of DOI:10.1016/j.electacta.2010.12.037

In this work, a simple co-deposition strategy for the synthesis of carbon-supported Pd-Au alloy was reported. Our approach involves the co-reduction of Au and Pd ions using ethylene glycol and sodium citrate as the reducing and stabilizing reagents. Both alloy and non-alloy bimetallic Pd-Au nanoparticles are produced using a right rate-limiting strategy. For example, when ethylene glycol and sodium citrate are the limiting reagent with Au and Pd ions in excess, the synthesis environment favors preferential nucleation and growth of Au nanoparticles followed by deposition of Pd either as the shell of Au core or as separate Pd clusters. On the other hand, if the supply of metal ions (not the reducing reagents) limits the reaction, it creates a synthesis condition for Pd-Au alloy particles. The as-prepared Pd-Au alloys exhibit higher Pd-specific activities towards formic acid oxidation compared with the non-alloy counterpart or individual Pd catalyst and an easier removal of adsorbed oxygen species (e.g., O_ads or OH_ads) was observed from the surface of Pd-Au alloy with a higher content of Au.

Full text of DOI:10.1016/j.electacta.2010.12.037

Electrochimica Acta 56 (2011) 2174–2183
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Synthesis of bimetallic PdAu nanoparticles for formic acid oxidation
Yange Suo, I-Ming Hsing∗
Department of Chemical and Biomolecular Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong SAR, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 October 2010
Received in revised form 9 December 2010
Accepted 9 December 2010
Available online 16 December 2010
In this work, a simple co-deposition strategy for the synthesis of carbon-supported Pd–Au alloy was
reported. Our approach involves the co-reduction of Au and Pd ions using ethylene glycol and sodium
citrate as the reducing and stabilizing reagents. Both alloy and non-alloy bimetallic Pd–Au nanoparticles
are produced using a right rate-limiting strategy. For example, when ethylene glycol and sodium citrate
are the limiting reagent with Au and Pd ions in excess, the synthesis environment favors preferential
nucleation and growth of Au nanoparticles followed by deposition of Pd either as the shell of Au core or
as separate Pd clusters. On the other hand, if the supply of metal ions (not the reducing reagents) limits the
reaction, it creates a synthesis condition for Pd–Au alloy particles. The as-prepared Pd–Au alloys exhibit
higher Pd-specific activities towards formic acid oxidation compared with the non-alloy counterpart or
individual Pd catalyst and an easier removal of adsorbed oxygen species (e.g., Oads or OHads) was observed
from the surface of Pd–Au alloy with a higher content of Au.
Keywords:
Ethylene glycol
Alloy
Non-alloy
Lattice parameter
© 2010 Elsevier Ltd. All rights reserved.
1. Introduction
radiolytic reduction method, through which, Pd–Au alloy nanopar-
ticles stabilized by poly(acrylic acid) were achieved. Mejía-Rosales
Structural or electronic properties of binary catalysts could
be very different from that of their individual components and
enhanced activity or stability of the binary catalysts were often
observed [1,2]. Among the various bimetallic materials, Pd–Au sys-
tem is an interesting one as it shows high activity towards useful
chemical reactions (e.g., hydrogenation of 4-pentenotic acid [3],
acetoxylation of ethylene to vinyl acetate [4], and oxidation of alco-
hols to aldehydes [5]) as well as electrochemical reactions (e.g.,
ethanol oxidation [6], 2-propanol oxidation [7], and formic acid
oxidation [8]). Investigations of catalysts for formic acid electro-
oxidation have attracted interests due to their use in direct formic
acid fuel cell, a promising power device for consumer electron-
ics. Aucore–Pdshell structured materials have been reported to show
enhanced activity for formic acid oxidation than the well-known
Pd [8,9]. Catalytic properties of bimetallic nanoscale catalysts are
highly dependent on their structure (e.g., alloy, core–shell struc-
ture) and surface composition [10]. Moreover, interesting and
unique electrochemical properties of Pd–Au alloys for formic acid
oxidation might be explored.
et al. [16] synthesized and characterized Pd–Au nanoparticles
passivated with poly(vinlpyrrolidone) (PVP) using a sequential
reduction protocol. Using hydrazine as the strong reductant, PVP-
protected Pd–Au alloy nanodendrites with an average particle size
of 15–26 nm have been synthesized [17].
In this work, we report a simple co-reduction method with ethy-
lene glycol and sodium citrate as the reducing and stabilizing agents
to synthesize Pd–Au bimetallic nanoparticles. Traditionally, the for-
mation of Pd–Au alloys through concurrent reduction of Pd and Au
ions is not easy because of the vast difference of redox potentials
between Au and Pd [15]. At the same reducing environment, Au ions
have much higher reduction rate than that of Pd ions. To overcome
this limitation, our approach adopts the principle of reaction engi-
neering and rate-limiting strategy to carefully control the synthesis
environment that would enable the production of alloy and non-
alloy Pd–Au bimetallic nanoparticles. Pd–Au alloy nanoparticles
with different metallic compositions were successfully synthesized
by this new approach. Electrochemical evaluation suggests that
the as-synthesized Pd–Au alloys exhibit higher Pd-specific activ-
ities towards formic acid electro-oxidation than Pd/C. Moreover,
Pd–Au catalysts consisting of a higher content of Au show an easier
removal of adsorbed oxygen species (e.g., Oads or OHads).
Although efforts have been made to study the synthesis of Pd–Au
bimetallic nanocatalysts, preparation of alloy structured Pd–Au
nanoparticles is still considered to be a big challenge [11,12]. The
synthesis of core–shell structured Pd–Au catalysts can be easily
achieved using either sequential [13,14] or co-reduction methods
[11,12,15]. To make alloy catalysts, Ksar et al. [6] have reported a
2. Experimental
2.1. Chemicals
∗
Ammonium tetrachloropalladate ((NH4)2PdCl4, 99.995%),
hydrogen tetrachloroaurate trihydrate (HAuCl4·3H2O, 99.9%),
Corresponding author. Tel.: +852 23587131; fax: +852 31064857.
E-mail address: kehsing@ust.hk (I-M. Hsing).
0013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2010.12.037

Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
2175
Table 1
Metallic loadings of catalysts based on ICP-AES characterization.
Catalysts
Metallic loadingsbasedon ICP-AEScharacterization (%)
Atomic ratio (Pd/Au)
Experimental
Au
Pd
Precursor
Pd
N/A
4.8
4.9
8.1
10.3
12.7
10.9
12.8
8.8
N/A
N/A
Pd82Au18/C I
Pd82Au18/C II
Pd67Au33/C II
Pd50Au50/C II
81/19
83/17
67/33
51/49
82/18
82/18
67/33
50/50
5.9
Table 4
Atomic compositions of five nanoparticles of the Pd82Au18/C I catalyst based on EDX
characterization.
Nanoparticles
Pd (atom %)
Au (atom %)
# 1
# 2
100
87
N/A
13
# 3
57
43
# 4
# 5
100
100
82
N/A
N/A
18
Stoichiometric value
to obtain the precursor solution. Then, 48 mg Vulcan XC-72
carbon was poured into the solution to get the precursor sus-
pension, which was stirred and ultrasonically mixed for 2 h. At
the same time, 0.283 g sodium citrate was dissolved into 2 ml
water + 10 ml ethylene glycol mixture solution to get the sodium
citrate water/ethylene glycol solution. After removing air with Ar
gas bubbling for 30 min, the precursor suspension was refluxed
at 170 ◦C oil bath for 5 min and then the prepared sodium citrate
water/ethylene glycol solution was added into the heated precursor
suspension drop by drop. Another 40 ml ethylene glycol was added
into the reaction system, which was then continued to be heated
for another 2 h. The reaction product was filtered and washed with
water and ethanol. The residue was dried at 60 ◦C oven for 12 h
and then grounded in an agate mortar. Catalyst synthesized by this
protocol is denoted as Pd82Au18/C I in the following discussion.
Fig. 1. XRD patterns of Au/C, Pd/C, and Pd82Au18/C (prepared by protocols I and II)
catalysts.
Table 2
Crystal size and lattice parameter of the Pd/C, Au/C and Pd82Au18/C (prepared by
protocols I and II) based on XRD characterization.
Catalysts
Crystal size/nm
Lattice parameter/nm
Pd/C (protocol II)
3.6
0.3893
Au/C (protocol II)
7.2
0.4064
4.2 (Pd)
6.8 (Au)
3.0
0.3890 (Pd)
0.4054 (Au)
0.3940
Pd82Au18/C (protocol I)
Pd82Au18/C (protocol II)
2.2.2. Synthesis protocol II for Pd82Au18/C catalyst
Firstly, 0.283 g sodium citrate was dissolved into 50 ml
water/ethylene glycol mixture solution (volume/volume = 1:1), and
then 48 mg Vulcan XC-72 carbon was poured into the above
solution to obtain the sodium citrate suspension, which was
stirred and ultrasonically mixed for 2 h. At the same time, 21.9 mg
ammonium tetrachloropalladate ((NH4)2PdCl4) and 0.65 ml hydro-
gen tetrachloroaurate trihydrate (HAuCl4·3H2O) aqueous solution
(1 g/100 ml) were dissolved into another 10 ml water/ethylene gly-
col solution (volume/volume = 1:1) to obtain the precursor solution.
After removing air with Ar gas bubbling for 30 min, the sodium cit-
rate suspension was refluxed at 170 ◦C oil bath. After 5 min heating,
the precursor solution was added into the heated sodium citrate
suspension drop by drop. Another 40 ml water/ethylene glycol
solution (volume/volume = 1:1) was added into the reaction sys-
tem, which was then continued to be heated for another 2 h. The
reaction product was filtered and washed with water and ethanol.
The residue was dried at 60 ◦C for 12 h and then grounded in an
agate mortar. Catalyst synthesized by this protocol is denoted as
Pd82Au18/C II in the following discussion.
ethylene glycol (C2H6O2, 99.5%), perchloric acid (HClO4, 69%) were
purchased from Sigma Aldrich. Sodium citrate (C6H5Na3O7·2H2O,
99%) was ordered from BDH. Formic acid (HCOOH, 98%) was
ordered from Riedel-de-Haën. Vulcan XC-72 carbon was pur-
chased from E-TEK. Nafion solution (5%) was received from
Dupont. Ethanol (C2H5OH, 99.5%) was purchased from Merck
KGaA.
2.2. Synthesis of catalysts
2.2.1. Synthesis protocol I for Pd82Au18/C catalyst
Firstly, 21.9 mg ammonium tetrachloropalladate ((NH4)2PdCl4)
and 0.65 ml hydrogen tetrachloroaurate trihydrate (HAuCl4·3H2O)
aqueous solution (1 g/100 ml) were dissolved into 48 ml water
Table 3
Atomic compositions of five nanoparticles of the Pd82Au18/C II catalyst based on
EDX characterization.
Protocol II was also used for the synthesis of Pd/C (same pal-
ladium/carbon ratio with that in the Pd82Au18/C catalysts), Au/C
(same gold/carbon ratio with that in the Pd82Au18/C catalysts),
Pd67Au33/C (metal loading of 20 wt% with atomic ratio of Pd to
Au being 67:33 in the precursor solution) and Pd50Au50/C (metal
loading of 20 wt% with atomic ratio of Pd to Au being 50:50 in the
precursor solution).
Nanoparticles
Pd (atom %)
Au (atom %)
# 1
# 2
# 3
# 4
89.30
85.81
82.38
86.35
88.50
82
10.70
14.19
17.62
13.65
11.50
18
# 5
Stoichiometric value

2176
Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
Fig. 2. TEM images and the size distribution of (a) Pd/C, (b) Pd82Au18/C I and (c) Pd82Au18/C II catalysts.
2.3. Physicochemical characterization
Powder X-ray diffraction (XRD) was carried out with a PAN-
alytical X-ray diffractometer (Model X’pert Pro) using a Cu K␣
radiation source operating at 40 kV and 40 mA. For crystalline size
and lattice parameter analysis, (2 2 0) diffraction peak was fitted to
a Lorentzian line shape to obtain the values of diffraction angle and
breadth of diffraction line at half-maximum intensity.
Transmission electron microscopy (TEM) was carried out with
a JEOL 2010F high resolution TEM system operated with LaB6 fil-
ament at 200 kV. The samples were prepared by dropping catalyst
suspensions (prepared by dispersing the catalysts ultrasonically in
ethanol) onto carbon coated Cu grids and then dried at room tem-
Inductively coupled plasma atomic emission spectroscopy (ICP-
AES) was carried out using an Optima 3000XL system (Perkin
Elmer) to measure the real metallic loadings of the catalysts. The
following procedure was followed to prepare the samples. Firstly,
5 mg catalyst was poured into 50 ml aqua regia and the mixture was
left at room temperature for 42 h to completely dissolve metallic
nanoparticles. The obtained suspension was then centrifuged with
a rotating rate of 13,400 rpm for 2 min. The supernatant was taken
out and then diluted to do ICP-AES characterization.

Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
2177
Fig. 3. (a) EDX spectrum of one nanoparticle of the Pd82Au18/C II catalyst and (b) HRTEM image of one nanoparticle of the Pd82Au18/C II catalyst and the inset picture is the
corresponding Fast Fourier Transfer (FFT) image.
perature. Number average particle size was recorded by taking the
size average of more than 300 particles. Atomic percentages of Pd
and Au were measured by energy-dispersive X-ray spectroscopy
(EDX), which was integrated with TEM equipment. To get rid of
contamination by other nanoparticles, we chose single particles
which are not closely surrounded by other metallic nanoparticles.
The spot size of the light beam for EDX analysis was a few nanome-
ters.
the activity measurement, the working electrode was cleaned by
CV sweeping in Ar-saturated 1 M HClO4 solution. Then a certain
amount (in order to keep the concentration in the electrolyte as
1 M) of formic acid was added into the electrolyte to do the activity
measurements. Pd-specific activities were obtained through nor-
malizing formic acid oxidation currents with the electrochemical
active surface area of Pd, measured by analyzing the Pd sur-
face oxide reduction and CO stripping charge. As the peak area
of the surface oxide reduction in the cyclic voltammogram (CV)
is proportional to the electrochemical surface area of the cata-
lyst, peak areas of Pd surface oxide reduction were simply used
to denote the electrochemical surface areas. Pd-specific activi-
ties were obtained by simply normalizing formic acid oxidation
currents with the peak areas of the Pd surface oxide reduction
(potential range of CV scan: 0.06–1.16 V) of the catalysts in Ar sat-
urated 1 M HClO4. The unit of Pd-specific activities obtained by
this method is not mA cm−2, but denoted as arbitrary unit (a.u.)
The quantified Pd active surface areas of the catalysts were mea-
sured by CO stripping experiments and the following procedures
were followed. Firstly, the working electrode was cleaned by CV
sweeping in Ar saturated 1 M HClO4 solution. CO gas was then
bubbled into the electrolyte solution for 20 min. The working elec-
trode was held at 0.1 V for 15 min for CO adsorption. Ar gas was
then bubbled into the electrolyte for 30 min to remove excess CO
dissolved in the solution. CV sweeping was conducted to do CO
stripping. The active surface areas were obtained by calculating
X-ray photoelectron spectroscopy (XPS) was carried out with
a Physical Electronics PHI 5600 multi-technique system, using Al
monochromatic X-ray at a power of 350 W.
2.4. Electrochemical evaluation
Electrochemical measurements were carried out using an Auto-
lab potentiostat (PGSTAT20, Eco Chemie, The Netherlands) in a
conventional three-electrode cell with a catalyst coated glassy car-
bon electrode (working electrode), a Pt coil (counter electrode) and
a saturated Ag/AgCl electrode (reference electrode). All the poten-
tials in our data are converted to be referred to standard hydrogen
electrode (SHE). The glassy carbon substrate was polished with alu-
mina suspension prior to use. To prepare the working electrode,
7 mg catalyst was dispersed ultrasonically in 1 ml diluted nafion
solution (0.05 wt% in ethanol) for 30 min and 10 ␮l of the sus-
pension was pipetted onto the glassy carbon electrode (d = 5 mm)
by a microsyringe and left to dry at room temperature. Prior to

2178
Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
Fig. 5. Cyclic voltammograms of Au/C, Pd/C and Pd82Au18/C (prepared by protocols
I and II) catalysts in Ar-saturated 1 M HClO4 solution at room temperature with scan
rate of 50 mV/s.
the stripping charges of pre-adsorbed CO (assuming 420 ␮C cm−2).
Through normalizing formic acid oxidation currents by the active
surface areas, the Pd-specific activities were then obtained. All
electrochemical experiments were conducted at room tempera-
ture.
3. Results and discussion
3.1. Synthesis of Pd–Au bimetallic catalysts
ICP-AES was carried out to analyze the real metallic loadings of
the Pd82Au18/C and Pd/C catalysts. As shown in Table 1, the mea-
sured atomic ratios of Pd to Au in Pd82Au18 catalysts are consistent
with the stoichiometic value (82:18) in the precursor solution.
XRD patterns of the Pd/C, Au/C and bimetallic Pd82Au18/C cat-
alysts are shown in Fig. 1. Positions of the diffraction peaks of
Pd/C and Au/C catalysts fit well with their characteristic face cen-
tered cubic (fcc) patterns. Pd82Au18/C I exhibits XRD patterns with
superimposed patterns of Pd/C and Au/C, suggesting a structure
of non-alloy catalysts. This result is consistent with the literature
report of Pd–Au catalyst prepared with microwave-assisted syn-
thesis protocol using ethylene glycol as the reducing reagent [11].
On the other hand, Pd82Au18/C II reveals new diffraction peaks with
positions between that of pure Pd/C and Au/C. The lattice parame-
ter, calculated based on Bragg’s law [18] using (2 2 0) diffraction
peak (at 67.15◦), is 0.3940 nm (Table 2), which is similar to the
empirical value of Pd82Au18 (0.3924 nm) by Vegard’s law [19]. XRD
data suggests that the bimetallic Pd–Au particles by protocol II
could form an alloy structure.
Fig. 4. XPS spectra of Pd3d and Au4f of Pd82Au18/C catalysts (prepared by protocols
I and II).
TEM images of the Pd/C, Pd82Au18/C I and Pd82Au18/C II cata-
lysts indicate that all the catalysts are well dispersed on the carbon
surface (Fig. 2). The histogram of particle size distribution of each
catalyst was obtained by recording more than 300 nanoparticles at
different areas. The average particle size of Pd82Au18/C II (4.2 nm) is
close to that of Pd/C (4.5 nm). Pd82Au18/C I exhibits a larger average
size of 6.9 nm.
EDX spectrum (Fig. 3a) of one particle of the Pd82Au18/C II cat-
alyst shows that a single particle is composed of both Pd and Au
elements. Moreover, atomic compositions of the five nanoparti-
cles are roughly consistent with the stoichiometric value (Table 3).
Fig. 3b shows the high resolution TEM (HRTEM) image of one spe-
cific particle of the Pd82Au18/C II catalyst and the inset photo shows
its corresponding Fast Fourier Transfer (FFT) image. Three groups of
diffraction spots are observed. By measuring the distance between
Scheme 1. Illustration of two protocols for the preparation of bimetallic Pd–Au cat-
alysts supported on carbon. Protocol I: drop by drop addition of ethylene glycol
and sodium citrate (the reducing and stabilizing reagents) to the abundant mix-
ture of metallic precursors; protocol II: drop by drop addition of the Pd and Au ion
precursors to the abundant mixture of ethylene glycol and sodium citrate.

Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
2179
Fig. 7. (a) XRD patterns of the Pd–Au/C catalysts with different metallic compo-
sitions synthesized by protocol II and (b) comparison of the experimental and
empirical lattice parameters of the carbon supported Pd–Au alloys with different
metallic compositions synthesized by protocol II.
on EDX spectrum of five nanoparticles of the Pd82Au18/C I suggest
that there are individual Pd and Pd–Au bimetallic nanoparticles
(Table 4) in the catalyst.
XPS characterization was carried out to get surface informa-
tion. Fig. 4 shows the Pd3d and Au4f peaks of the Pd82Au18/C I
and Pd82Au18/C II catalysts. Surface atomic ratio of Pd to Au in
Pd82Au18/C I is 93:7, which is much larger than the stoichiometic
value (82:18), indicating enrichment of Pd element on the surface
of the nanoparticles. On the contrary, surface atomic ratio of Pd to
Au in Pd82Au18/C II is 77:23, which is close to the stoichiometric
value (82:18).
The above results have strongly suggested that most of the
Pd–Au nanoparticles in Pd82Au18/C II catalyst form alloy structure,
while non alloy nanoparticles are generated using protocol I. As
illustrated in Scheme 1, when the rate-limiting reactant (supplied
in drop-by-drop) of the reducing agent (ethylene glycol and sodium
citrate) is added to the metallic ion solutions in excess, it creates
a synthesis environment for faster deposition of Au than Pd due
to higher reduction potential of Au ions than that of Pd ions [20]
and thus leads to core–shell, as well as other non-alloy structure of
Pd–Au. On the other hand, the environment of a limited supply of
Pd and Au ion precursors (supplied in drop-by-drop) and an abun-
dant amount of reducing reagents at high temperature, as shown
in protocol II, accelerates the reduction rates of both Pd and Au
ions and drives the synthesis process to be controlled by the mass
transport of ion reactants (i.e., Pd and Au precursors). Therefore, the
drop-by-drop arrival of Pd and Au ions determines catalyst synthe-
sis condition and simultaneous reduction of Pd and Au ions to form
alloy structure is thus achievable.
Fig. 6. (a) Cyclic voltammogram of Au/C, Pd/C and Pd82Au18/C (prepared by proto-
cols I and II) catalysts in Ar-saturated 1 M HCOOH + 1 M HClO4 at room temperature
with scan rate of 50 mV/s, (b) Pd-specific activities (through normalizing formic acid
oxidation currents by the electrochemical active surface areas of Pd) at 0.2 V on Pd/C
and Pd82Au18/C (prepared by protocols I and II) and (c) potentials (averaged by three
measurements) of the abrupt current increase in the backward sweep on Pd/C and
Pd82Au18/C (prepared by protocols I and II).
the two spots in each group, the inter-planer distances of 0.226 nm,
0.140 nm and 0.120 nm are obtained. The measured values are
consistent with the empirical values of Pd82Au18 (1 1 1), (2 2 0),
(3 1 1) crystallographic planes, which are 0.227 nm, 0.139 nm and
0.118 nm respectively based on Vegard’s law [19]. EDX and high
resolution TEM image confirm that Pd82Au18/C II form an alloy
structure. Unlike that of Pd82Au18/C II, atomic compositions based

2180
Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
Fig. 8. TEM images and the size distribution of (a) Pd67Au33/C II and (b) Pd50Au50/C II.
3.2. Electrochemical properties of Pd82Au18/C catalyst in
comparison with Pd/C and Au/C
from Pd82Au18/C alloy catalyst compared with that from non-alloy
counterpart or Pd/C. Size and electronic effects could lead to the
phenomenon of this easier removal of adsorbed oxygen species
from the substrate. On larger particles, easier removal of adsorbed
oxygen species from Pt or Pd surface was observed [21,22]. How-
ever, based on previous XRD and TEM analysis (Table 2 and Fig. 2),
Pd82Au18/C II has a relatively smaller particle size compared with
Pd82Au18/C I or Pd/C. Therefore, the observed phenomenon of
easier removal of adsorbed oxygen species from Pd82Au18/C II com-
pared with that from Pd82Au18/C I or Pd could not be explained by
the particle size effect.
Cyclic voltammetric investigations of binary Pd82Au18/C pre-
pared by protocols I and II in benchmark comparison with Au/C
and Pd/C are reported in Figs. 5 and 6a. Other than Au/C, Pd/C and
Pd82Au18/C catalysts exhibit high surface reaction currents in the
measured potential range (Fig. 5). Current peaks of hydrogen des-
orption/adsorption at low potentials (i.e., 0.06–0.4 V), and surface
oxidation/reduction at high potentials (i.e., higher than 0.6 V) reveal
typical CV characteristics of a “palladium-like” catalyst. Moreover,
the reduction peak of surface oxide reveals an anodic shift on the
Pd82Au18/C catalyst compared with that on Pd/C. Pd82Au18/C II, in
particular, shows the highest shift (as indicated by the arrow in
Fig. 5), indicating an easier removal of adsorbed oxygen species
Fig. 6a shows the polarization curves of formic acid oxidation on
Au/C, Pd/C and Pd82Au18/C catalysts. Au/C shows no catalytic activ-
ity for formic acid oxidation in the measured potential range. The
activity of Pd82Au18/C is higher than that of Pd/C, with Pd82Au18/C II
Table 5
Table 6
Atomic compositions of five nanoparticles of the Pd67Au33/C II catalyst based on
EDX characterization.
Atomic compositions of five nanoparticles of the Pd50Au50/C II catalyst based on
EDX characterization.
Nanoparticles
Pd (atom %)
Au (atom %)
Nanoparticles
Pd (atom %)
Au (atom %)
# 1
# 2
# 3
# 4
91.27
76.55
69.88
61.89
66.70
67
8.73
# 1
# 2
# 3
# 4
53.90
56.07
56.73
56.43
71.88
50
46.10
43.93
43.27
43.57
28.12
50
23.45
30.12
38.11
33.30
33
# 5
# 5
Stoichiometric value
Stoichiometric value

Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
2181
Fig. 10. (a) Cyclic voltammogram of Au/C, Pd/C and Pd–Au/C catalysts with differ-
ent metallic compositions synthesized by protocol II in Ar-saturated 1 M HCOOH
at room temperature with scan rate of 50 mV/s and (b) peak potentials of the sur-
face oxide reduction of the Au/C, Pd/C and Pd–Au/C catalysts with different metallic
compositions synthesized by protocol II.
Fig. 9. XPS spectra of Pd3d and Au4f of the Pd67Au33/C II and Pd50Au50/C II.
exhibiting the highest activity. Fig. 6b shows the Pd-specific activi-
ties at 0.2 V on different catalysts. It can be seen that Pd82Au18/C II
shows a higher activity compared with Pd82Au18/C I and Pd/C. In
addition, the sharp increase of the current during the backward
scan on the Pd82Au18/C II occurs at a more positive potential (as
indicated by the arrow sign) than that of the other catalysts (Fig. 6a
and c). As discussed in our previous work [22], the adsorbed oxy-
gen species could block the active surface area of Pd. Removal
of these adsorbed oxygen species could trigger a sharp current
increase during the backward sweep. Therefore, consistent with
the CV results shown in Fig. 5, an anodic shift of sharp current
increase on Pd82Au18/C II catalyst suggests that it is easier to
remove adsorbed oxygen species from carbon supported Pd–Au
alloy surface than that from the non-alloy counterpart or single Pd
surface. Interestingly, Pt surface has also been reported to become
more resistant towards surface oxidized species formation through
manipulation by small Au clusters [1]. The d-band theory proposed
by Hammer and Nøskov [23] provides a theoretical analysis for
estimation of surface reactivity. Based on their model, the surface
reactivity is related to the position of the center of metal d-band.
Shift of d-band center has been demonstrated for Pd overlayers
on Au and Au overlayers on Pd compared with pure Au or Pd
substrates by theoretical calculation [24]. Therefore, the observed
phenomenon of easier removal of adsorbed oxygen species from
carbon supported Pd–Au alloy in comparison with that from non-
alloy counterpart or carbon supported Pd in this work could be
caused by the adjusted electronic properties due to alloying Au to
Pd.
3.3. Effect of metallic composition on the electrochemical
property of Pd–Au/C
Pd–Au/C catalysts with other metallic compositions
(Pd67Au33/C, Pd50Au50/C) were also synthesized using our protocol
II. The metallic loadings of the Pd–Au/C catalysts were character-
ized by ICP-AES. The metallic ratios of Pd to Au are consistent with
that in the precursor solutions, as shown in Table 1.
XRD patterns of the Pd–Au/C catalysts with different metallic
compositions synthesized by protocol II are compared with that of
Pd/C (Fig. 7a). All the Pd–Au/C catalysts exhibit characteristic fcc
diffraction peaks with positions between that of Pd/C and Au/C,
indicating the formation of alloy structures. The corresponding
diffraction peaks shift towards lower diffraction angles with the
increase of Au content. Fig. 7b shows experimental lattice param-
eters (calculated based on Bragg’s law [18] using (2 2 0) diffraction
peak) and empirical lattice parameters (based on Vegard’s law [19])
of the catalysts with different metallic compositions. Clearly, the
experimental lattice parameters are consistent with the empirical
values and with the increase of Au content, lattice parameter of the
Pd–Au catalyst increases.
Nanoparticles in Pd67Au33/C II and Pd50Au50/C II catalysts are
well dispersed on carbon surface, as shown in TEM images (Fig. 8).
However, a wide size distribution is observed. With the increase
of Au content, the average particle size of the Pd–Au catalysts
increases. Atomic compositions (measured by EDX) of five single
particles in Pd67Au33/C II and Pd50Au50/C II catalysts are consistent
with their stoichiometric values (Tables 5 and 6).

2182
Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
peaks of surface oxide on Pd/C and Au/C catalysts are at about 0.6 V
and 1.17 V respectively, consistent with the literature report [25].
Two surface oxide reduction peaks are observed on Pd–Au/C cata-
lysts, with one peak at Pd oxide reduction region (lower potential)
and the other at Au oxide reduction region (higher potential). Large
shiftofreduction peaks on Pd–Au/C catalysts withdifferentmetallic
compositions can be observed. Fig. 10b compares the peak poten-
tials of the Pd–Au surface oxide reduction on different catalysts.
Clearly, both of the two reduction peaks on each Pd–Au/C catalysts
are between that of pure Pd and Au, which could indicate alloy
structure of Pd–Au [25].
Fig. 11a compares the polarization curves (the oxidation cur-
rents were normalized by Pd contents characterized by IPC-AES)
of formic acid oxidation on Pd/C and Pd–Au/C alloys with different
metallic compositions. All carbon supported Pd–Au alloys exhibit
higher Pd-mass activities compared with pure Pd. Fig. 11b com-
pares the Pd-specific activities of Pd–Au catalysts with different
metallic compositions for formic acid oxidation at 0.2 V. Clearly, the
Pd-specific activities are higher on catalysts with higher content of
Au. Interestingly, a shift of sharp current increase in the backward
sweep was observed on catalysts with different metallic compo-
sitions. Larger shift towards more positive potential was clearly
observed on Pd–Au catalysts with higher content of Au (Fig. 11a
and c), suggesting an easier removal of adsorbed oxygen species
from Pd–Au alloy catalysts with higher content of Au. Consistent
with the previous discussion (Section 3.2), the higher alloying com-
ponent of Au on Pd contributes to the easier removal of adsorbed
oxygen species from Pd–Au alloy surface.
4. Conclusions
A simple co-deposition strategy for the synthesis of carbon-
supported Pd–Au alloy was developed. This approach involves the
co-reduction of Au and Pd ions in a controlled environment of
ethylene glycol and sodium citrate as the reducing and stabiliz-
ing reagents. Both alloy and non-alloy Pd–Au nanoparticles could
be achieved by choosing the right rate-limiting reactant during
the reduction process. Higher Pd-specific activity was achieved on
Pd–Au alloy catalyst compared with the non-alloy counterpart or its
individual component, e.g., Pd. Easier removal of adsorbed oxygen
species from Pd–Au alloy with a higher content of Au was observed.
Acknowledgements
The authors gratefully acknowledge the Research Grants Coun-
cil of the Hong Kong SAR Government for the funding support.
The postgraduate scholarship to YangeSuofrom the Nanoscience
and Nanotechnology program of School of Engineering is also
appreciated. Technical support on surface analysis from the staff
of Materials Characterization and Preparation Facility (MCPF) of
HKUST is also appreciated.
Fig. 11. (a) Polarization curves (currents normalized by Pd contents based on ICP-
AES characterization) of 1 M formic acid oxidation on Pd–Au/C alloys with different
metallic compositions synthesized by protocol II in Ar-saturated 1 M HClO4 solution
at room temperature with scan rate of 50 mV/s, (b) Pd-specific activities (through
normalizing formic acid oxidation currents by the electrochemical active surface
areas of Pd) at 0.2 V on Pd–Au/C catalysts with different metallic compositions syn-
thesized by protocol II and (c) potentials (averaged by three measurements) of the
abrupt current increase in the backward sweep on Pd–Au/C catalysts with different
metallic compositions synthesized by protocol II.
References
[1] J. Zhang, K. Sasaki, E. Sutter, R.R. Adzic, Science 315 (2007) 220.
[2] Y.G. Suo, L. Zhuang, J.T. Lu, Angew. Chem. Int. Ed. 46 (2007) 2862.
[3] Y. Mizukoshi, T. Fujimoto, Y. Nagata, R. Oshima, Y. Maeda, J. Phys. Chem. B 104
(2000) 6028.
[4] M.S. Chen, D. Kumar, C.W. Yi, D.W. Goodman, Science 310 (2005) 291.
[5] D.I. Enache, J.K. Edwards, P. Landon, B. Solsona-Espriu, A.F. Carley, A.A. Herz-
ing, M. Watanabe, C.J. Kiely, D.W. Knight, G.J. Hutchings, Science 311 (2006)
362.
[6] F. Ksar, L. Ramos, B. Keita, L. Nadjo, P. Beaunier, H. Remita, Chem. Mater. 21
(2009) 3677.
[7] C.W. Xu, Z.Q. Tian, Z.C. Chen, S.P. Jiang, Electrochem. Commun. 10 (2008) 246.
[8] W.J. Zhou, J.Y. Lee, Electrochem. Commun. 9 (2007) 1725.
[9] I.-S. Park, K.-S. Lee, S.J. Yoo, Y.-H. Cho, Y.-E. Sung, Electrochim. Acta 55 (2010)
4339.
Fig. 9 shows the Pd3d and Au4f XPS spectra of the Pd67Au33/C II
and Pd50Au50/C II catalysts. Surface atomic ratios of Pd to Au in
Pd67Au33/C II and Pd50Au50/C II are 67:33 and 52:48 respectively.
The values are similar to the stoichiometric values of Pd67Au33/C II
(67:33) and Pd50Au50/C II (50:50).
Fig. 10a shows CV of Pd/C, Au/C and Pd–Au/C catalysts with
different metallic compositions in Ar saturated perchloric acid solu-
tion with a high upper limit potential (i.e., 1.6 V). The reduction
[10] S. Alayoglu, A.U. Nilekar, M. Mavrikakis, B. Eichhorn, Nat. Mater. 7 (2008) 333.

Y. Suo, I-M. Hsing / Electrochimica Acta 56 (2011) 2174–2183
2183
[11] R. Harpeness, A. Gedanken, Langmuir 20 (2004) 3431.
[12] Y.W. Lee, M. Kim, Z.H. Kim, S.W. Han, J. Am. Chem. Soc. 131 (2009) 17036.
[13] A. Henglein, J. Phys. Chem. B 104 (2000) 6683.
[14] J.W. Hu, J.F. Li, B. Ren, D.Y. Wu, S.G. Sun, Z.Q. Tian, J. Phys. Chem. C 111 (2007)
1105.
[18] R. Jenkins, R.L. Snyder, Introduction to X-ray Powder Diffractometry, Wiley,
New York, 1996.
[19] A.R. Denton, N.W. Ashcroft, Phys. Rev. A 43 (1991) 3161.
[20] M.S. Antelman, F.J. Harris, The Encyclopedia of Chemical Electrode Potentials,
Plenum Press, New York, 1982.
[15] S. Nath, S. Praharaj, S. Panigrahi, S.K. Ghosh, S. Kundu, S. Basu, T. Pal, Langmuir
21 (2005) 10405.
[16] S.J. Mejía-Rosales, C. Fernández-Navarro, E. Pérez-Tijerina, D.A. Blom, L.F.
Allard, M. José-Yacaman, J. Phys. Chem. C 111 (2007) 1256.
[17] Y.W. Lee, M. Kim, Y. Kim, S.W. Kang, J.H. Lee, S.W. Han, J. Phys. Chem. C 114
(2010) 7689.
[21] X. Li, Xi. Qiu, H. Yuan, L. Chen, W. Zhu, J. Power Sources 184 (2008) 353.
[22] Y. Suo, I.M. Hsing, Electrochim. Acta 55 (2009) 210.
[23] B. Hammer, J.K. Norskov, Surf. Sci. 343 (1995) 211.
[24] J. Greeley, J.K. Norskov, Surf. Sci. 592 (2005) 104.
[25] F.A. Al-Odail, A. Anastasopoulos, B.E. Hayden, Phys. Chem. Chem. Phys. 12
(2010) 11398.