Synthesis of Pd-Pt bimetallic nanocrystals with a concave structure through a bromide-induced galvanic replacement reaction

Article abstract of DOI:10.1021/ja201156s

This article describes a systematic study of the galvanic replacement reaction between PtCl₆^2- ions and Pd nanocrystals with different shapes, including cubes, cuboctahedrons, and octahedrons. It was found that Br^- ions pl

Full text of DOI:10.1021/ja201156s

ARTICLE
pubs.acs.org/JACS
Synthesis of PdꢀPt Bimetallic Nanocrystals with a Concave Structure
through a Bromide-Induced Galvanic Replacement Reaction
Jinguo Wang, Weiyang Li, Pedro H. C. Camargo, Moon J. Kim,
|
|
,†,‡
||,†,§
^
,†
†
†
^
Hui Zhang, Mingshang Jin,
‡
§
Deren Yang, Zhaoxiong Xie, and Younan Xia*
†
Department of Biomedical Engineering, Washington University, St. Louis, Missouri 63130, United States
‡
State Key Laboratory of Silicon Materials and Department of Materials Science and Engineering, Zhejiang University,
Hangzhou, Zhejiang 310027, People’s Republic of China
§
State Key Laboratory for Physical Chemistry of Solid Surfaces and Department of Chemistry, College of Chemistry and Chemical
Engineering, Xiamen University, Xiamen, Fujian 361005, People’s Republic of China
^
Department of Materials Science, University of Texas at Dallas, Richardson, Texas 75083, United States
S Supporting Information
b
ABSTRACT: This article describes a systematic study of the
2
ꢀ
galvanic replacement reaction between PtCl6 ions and Pd
nanocrystals with different shapes, including cubes, cuboctahe-
ꢀ
drons, and octahedrons. It was found that Br ions played an
important role in initiating, facilitating, and directing the re-
ꢀ
placement reaction. The presence of Br ions led to the selective
initiation of galvanic replacement from the {100} facets of Pd
ꢀ
nanocrystals, likely due to the preferential adsorption of Br ions
on this crystallographic plane. The site-selective galvanic re-
placement resulted in the formation of PdꢀPt bimetallic nano-
crystals with a concave structure owing to simultaneous dissolution of Pd atoms from the {100} facets and deposition of the
resultant Pt atoms on the {111} facets. The PdꢀPt concave nanocubes with different weight percentages of Pt at 3.4, 10.4, 19.9, and
3
4.4 were also evaluated as electrocatalysts for the oxygen reduction reaction (ORR). Significantly, the sample with a 3.4 wt.% of Pt
exhibited the largest specific electrochemical surface area and was found to be four times as active as the commercial Pt/C catalyst for
the ORR in terms of equivalent Pt mass.
1
2
’
INTRODUCTION
an overgrowth approach. We have demonstrated the syntheses
of Au nanocages and nanoframes with cavities on the surfaces
using a galvanic replacement reaction between Ag nanocrystals
Noble-metal nanocrystals with a concave structure on the
surface have physical/chemical properties different from those
enclosed by a flat or convex surface.
facets and negative curvature, nanocrystals with a concave
structure have also received attention for a number of applica-
tions related to catalysis, electrocatalysis, optical sensing, and
13
1
ꢀ5
and a Au(I) or Au(III) salt precursor. We have also demon-
strated the synthesis of Pd nanocages using an oxidative etching
process. Zheng and co-workers reported the synthesis of
tetrahedral and trigonal bipyramidal nanoframes of Pd using a
Owing to their unusual
14
1
5
6
ꢀ8
solvothermal process in the presence of formaldehyde.
localized surface plasmon resonance.
In general, overgrowth
Berkovitch and co-workers demonstrated a technique based on
electron beam lithography (EBL) for fabricating arrays of con-
cave hyperbolic Au particles on glass substrates with controllable
concavity via an etching strategy. Mirkin and co-workers
demonstrated the synthesis of Au concave nanocubes via seed-
mediated overgrowth with cetyltrimethylammonium chloride
(CTAC) as a capping agent. Most recently, we reported a
facile process for the formation of Rh, Pt, and Pt/Rh concave
nanocubes by manipulating the reaction kinetics with a syringe
and etching represent two powerful routes to the synthesis of
nanocrystals with a concave structure. For overgrowth, the
formation of a concave structure is typically achieved under
kinetic control to overcome the requirement of minimizing the
total surface free energy. In comparison, the formation of a
concave structure by etching has to rely on an anisotropic
process, in which a specific set of facets of the template are
preferentially dissolved over others. Compared to the remarkable
success in the synthesis of nanocrystals enclosed by a flat or
convex surface (with examples including cube, octahedron,
16
17
1
8,19
pump to alter the injection rate of a salt precursor.
remains a grand challenge to extend these strategies to other
It still
9
ꢀ11
tetrahedron, decahedron, icosahedron, plate, bar, and rod),
there are only a few reports on the synthesis of metal nanocrystals
with a concave structure. To this end, we have demonstrated the
synthesis Pt tetrahedra and octahedra with concave facets using
Received: February 6, 2011
Published: March 25, 2011
r 2011 American Chemical Society
6078
dx.doi.org/10.1021/ja201156s
|J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
metals to generate nanocrystals with a concave structure and
relatively small sizes (below 20 nm).
with high specific surface areas. Our electrochemical data clearly
demonstrate that the PdꢀPt concave nanocubes have substan-
tially enhanced ORR mass activity as compared to the commer-
cial Pt/C catalyst. The ORR activity also showed a strong
dependence on the compositions of the PdꢀPt bimetallic
nanocrystals, with the highest activity being observed for a Pt
weight percentage of 3.4.
As a strategy based on etching, galvanic replacement repre-
sents a remarkably simple and versatile route to the synthesis of
bimetallic nanostructures with hollow interiors as well as cavities
20ꢀ22
and thus negative curvatures on the surface.
In a galvanic
replacement reaction, the electrochemical potential difference
between a sacrificial template and the metal ions in a solution
drives the reaction, i.e., oxidation and dissolution of the template
accompanied by reduction of ions of another metal and deposi-
tion of the resultant atoms on the surface of the template. The
size and morphology of the final product can be readily manipu-
lated by using sacrificial templates with different sizes and shapes
and/or by controlling the extent of replacement. As demon-
strated in our previous work, Ag nanocrystals with a variety of
shapes, including nanocubes, nanospheres, nanorods, and nano-
wires, can all react with Au(III) or Au(I), Pt(IV) or Pt(II), and
Pd(II) salt precursors to generate bimetallic nanostructures in
’ EXPERIMENTAL SECTION
Chemicals and Materials. Sodium tetrachloropalladate (II)
2 4 2 6
(Na PdCl , Sigma-Aldrich, 99.998%), hexachloroplatinic acid (H PtCl ,
Sigma-Aldrich, 99.995%), potassium bromide (KBr, Fisher Scientific),
potassiumchloride (KCl, T. J. Baker), poly(vinylpyrrolidone) (PVP, MW
≈
55 000, Sigma-Aldrich), ascorbic acid (AA, Sigma-Aldrich), formalde-
hyde (Fisher Scientific), acetone (EMD), and ethanol (Pharmco Pro-
ducts, 200 proof) were all used as received. All aqueous solutions were
prepared using deionized water with a resistivity of 18.2 MΩ cm. All
23ꢀ25
3
the form of box, cage, frame, shell, rattle, and tube.
As a
syntheses were carried out in glass vials (20 mL, VWR international).
sacrificial template for the synthesis of bimetallic or alloyed
nanostructures, Ag nanocrystals have a number of limitations:
For example, their sizes are typically larger than 20 nm; the
remaining Ag may cause instability and poisoning issues for a
catalyst; and the remaining Ag may cause toxicity for biomedical
applications. In order to avoid these limitations or potential
problems, Pd nanocrystals have been recognized as an alternative
candidate for the galvanic reaction due to their availability in a
myriad of well-defined shapes and with sizes controllable down
to a few nanometers. This is particularly useful in the synthesis of
Pt-based catalysts for various applications. For the system invol-
ving Pd nanocrystals and a Pt precursor, however, no galvanic
replacement seemed to occur between them under conventional
conditions despite their large difference in electrochemical
Synthesis of Pd Nanocubes. The Pd nanocubes were synthe-
2 4
sized by adding a Na PdCl solution into a mixture of AA, KBr, and KCl
35
according to our previous report. In a typical synthesis, 8.0 mL of an
aqueous solution containing 105 mg of PVP, 60 mg of AA, and different
amounts of KBr and KCl was hosted in a vial and preheated to 80 ꢀC in
an oil bath under magnetic stirring for 10 min. Subsequently, 3.0 mL of
an aqueous solution containing 57 mg of Na PdCl was added with a
2
4
pipet. After the vial had been capped, the reaction was allowed to
continue at 80 ꢀC for 3 h. The size of Pd nanocubes was controlled by
varying the amounts of KBr and KCl: For example, the use of 600 mg of
KBr, 300 mg of KBr, and 5 mg of KBr together with 185 mg of KCl
resulted in the formation of Pd nanocubes about 18, 10, and 6 nm in
size, respectively. The product was collected by centrifugation, washed
three times with water to remove excess PVP, and redispersed in
11 mL water.
2
6
potential. In addition, compared with the system involving
Ag nanocrystals and a Au precursor, very little is known about the
mechanistic details associated with the heterogeneous nucleation
and growth of Pt on templates made of Pd nanocrystals during a
galvanic replacement process.
Synthesis of Pd Cuboctahedrons and Octahedrons. These
nanocrystals were prepared by adding 3 mL of aqueous Na
solution (9.5 mg/mL) into an aqueous solution (8 mL) containing
05 mg of PVP, 100 μL of HCHO, and different amounts of the Pd
PdCl
2
4
1
nanocubes (18 nm in size) at 60 ꢀC under magnetic stirring. The
morphology of resultant nanocrystals was determined by the amount of
Pd nanocubes added as the seeds. The use of 1.0 and 0.3 mL of the
suspension of Pd nanocubes resulted in the formation of cuboctahe-
drons of 25 nm in size and octahedrons of 33 nm in size, respectively. For
each synthesis, the vial was removed from the oil bath after 3 h. After
collection by centrifugation and washing three times with water, the
product was redispersed in 11 mL water. For the synthesis of Pd
octahedrons of 16 nm in size, the procedure was exactly the same as
Herein we report a bromide-induced galvanic replacement
2
ꢀ
reaction between PtCl6 ions and Pd nanocrystals with different
shapes for generating PdꢀPt bimetallic nanocrystals with a
concave structure. The PdꢀPt concave nanocrystals are of
particular interest and importance for catalytic or electrocatalytic
applications owing to the following attractive features: (i) they
can offer much higher specific surface areas and thus improved
27,28
activity relative to their solid counterparts;
(ii) the presence
of a cavity in the Pt layer and a Pd core in the interior can help
11
what we used in a previous study.
reduce the loading of Pt, an extremely scarce and expensive noble
2
9,30
Synthesis of PdꢀPt Bimetallic Nanocrystals via Galvanic
Replacement. In a standard procedure, 1 mL of the aqueous suspen-
sion of Pd nanocubes (18 nm in size) and 7 mL of an aqueous solution
containing PVP (33.3 mg) and KBr (300 mg) were mixed in a glass vial.
The mixture was heated to 90 ꢀC in air under magnetic stirring.
Meanwhile, 3.5 mg of H PtCl was dissolved in 3 mL deionized water.
2 6
This aqueous solution was then injected into the solution containing
PVP, KBr, and Pd nanocrystals using a syringe pump at a rate of
metal;
metals, such as Ag or Co, eliminates the corrosion by an acidic in
and (iii) the use of Pd in the core instead of other
31,32
a proton-exchange membrane (PEM) fuel cell.
In addition,
Pt nanostructures supported on Pd substrates may exhibit
enhanced activity for the oxygen reduction reaction (ORR) at
33,34
the cathode of a PEM fuel cell in comparison with pure Pt.
Specifically in this work, we have evaluated Pd nanocrystals with
various shapes, including cubes, octahedrons, and cuboctahe-
drons, as templates for the galvanic replacement reaction with
1
mL/min. The reaction mixture was then heated to 90 ꢀC for 12 h in air.
Finally, the solution was centrifuged and washed three times with water
to remove PVP before characterization. We systematically investigated
the effects of various parameters, including the duration of reaction, the
amount of H PtCl , the shape of the Pd nanocrystals, and the tempera-
2ꢀ
PtCl6 ions in an aqueous solution. The use of Pd nanocrystals
with different shapes (and thus facets) as templates allowed us to
systematically investigate the selectivity of facet during a galvanic
replacement reaction. Using this simple approach, we have been
able to routinely produce PdꢀPt bimetallic concave nanocrystals
2
6
ture as well as the amounts of KBr and KCl on the morphology of
resultant PdꢀPt nanocrystals.
6
079
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
Figure 1. TEM images of PdꢀPt bimetallic nanocrystals prepared via galvanic replacement using the standard procedure, except for different periods of
time: (A) 0.5, (B) 4, (C) 9, and (D) 20 h. The insets show TEM images of individual nanocrystals at a higher magnification. The scale bars in the insets
are 10 nm.
Morphological, Structural, and Elemental Characteriza-
tions. Transmission electron microscopy (TEM) images were taken
using a Tecnai G2 Spirit Twin microscope (FEI, Hillsboro, OR)
operated at 120 kV. High-resolution TEM (HRTEM), high-angle
annular dark-field scanning TEM (HAADF-STEM), and energy dis-
persive X-ray (EDX) analyses were performed using a JEOL 2100F
microscope (JEOL, Tokyo, Japan) operated at 200 kV. Scanning
electron microscopy (SEM) images were captured with a Nova Nano-
SEM 230 field-emission microscope (FEI, Hillsboro, OR) operated
at 30 kV. The percentages of Pd and Pt in the samples were
determined using inductively coupled plasma mass spectrometry
suspension of the catalyst was transferred to the glassy carbon rotating
disk electrode. For all PdꢀPt nanocrystals and Pt/C catalyst, the loading
of the catalyst was 3 μg based on the mass of the metal(s). Upon drying
under air for 2 h, the electrode was covered with 15 μL Nafion dispersed
in water (0.05%). The electrolyte was 0.1 M HClO prepared by diluting
4
a commercial solution (70%, ACS reagent grade, T. J. Baker) using
Millipore ultrapure water. The cyclic voltammetry curves were recorded
at room temperature in an N -purged (ultrahigh purity, Airgas) 0.1 M
2
aqueous HClO4 solution at a sweep rate of 50 mV/s. The ORR
measurements were performed at a sweep rate of 10 mV/s from 0.05
to 1.1 V at 1600 rpm under continuous O flow. The KouteckꢀLevich
2
(ICP-MS) (Perkin-Elmer Elan DRC II ICP-MS).
equation was applied to calculate the kinetic current density, which can
3
6,37
Electrochemical Measurements. The electrochemical measure-
be described as
ments were performed at room temperature using a rotating disk
electrode (Pine Research Instrumentation, United States) connected
to a PARSTAT 283 potentialstat (Princeton Applied Research, United
States). A hydrogen electrode (Gaskatel, Germany) was used as the
1=j ¼ 1=jk þ 1=jd
where j is the experimentally measured current, j
d
is the diffusion-
2
reference. The counter electrode was a Pt mesh (1 ꢁ 1 cm ) connected
limiting current, and j is the kinetic current. For each catalyst, the kinetic
current was normalized to the loading of metals (both Pt and Pd for the
k
to a Pt wire. To prepare the working electrode, 15 μL of an aqueous
6
080
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
Figure 2. TEM images of PdꢀPt nanocrystals obtained using the standard procedure, except for the addition of different amounts of H PtCl : (A) 0.5,
2
6
(B) 1.4, (C) 3.5, and (D) 7.0 mg. The insets show TEM images of individual nanocrystals at a higher magnification. The scale bars in the insets are 10 nm.
Pm da ꢀssPa tc tb i iv mi t ye .t allic nanocrystals or only Pt for Pt/C) in order to obtain
the differences in reaction time. In the initial stage of the reaction
(t = 0.5 h), the morphology of the Pd templates was essentially
maintained. The magnified TEM image (inset of Figure 1A)
clearly shows that the nanocube was slightly truncated at corners
’
RESULTS AND DISCUSSION
and bounded by a mix of {100} and {111} facets, which can be
ꢀ
Morphological Changes during the Galvanic Re-
attributed to minor oxidative etching caused by the Cl /O
2
and
ꢀ
38,39
placement Reaction. We monitored the evolution of different
morphologies during galvanic replacement by imaging the
products obtained at different reaction times. In a set of experi-
ments, Pd nanocubes with an average edge length of 18 nm
Br /O
2
pairs.
After the reaction had proceeded to t = 4 h
(Figure 1B), nanocubes with obvious concave faces were ob-
tained due to the galvanic replacement between Pd nanocubes
and H PtCl . The magnified TEM image (inset of Figure 1B)
2 6
(
Figure S1, Supporting Information) were employed as tem-
further reveals that each side face of the Pd nanocube was
excavated in the center, resulting in the formation of a concave
structure. A careful analysis indicated that the truncated corners
of the Pd nanocube were sharpened due to the deposition of
newly formed Pt atoms at these sites. These images suggest a
transition in morphology from Pd nanocubes with relatively
sharp corners to slightly truncated Pd nanocubes and then
PdꢀPt concave nanocubes. When the reaction further pro-
ceeded to t = 9 h (Figure 1C), most of the Pd nanocubes were
transformed into PdꢀPt concave nanocubes. We found that a
plates for the galvanic replacement reaction with H PtCl . A
2
6
close examination shows that a small portion of the Pd nano-
cubes were somewhat elongated along one of the axes to form
nanobars with aspect-ratios (length to width) slightly larger than
one. Since both the nanocubes and nanobars were enclosed by
{
100} facets, they are collectively called “nanocubes” in the
present work for the purpose of simplicity. Figure 1 shows
representative TEM images of the PdꢀPt nanocrystals prepared
via galvanic replacement using the standard procedure, except
6
081
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
Figure 3. The morphology, structure, and composition of the PdꢀPt concave nanocubes that were prepared using the standard procedure. (A) SEM
image, (B) HADDF-STEM image, (C) dark-field TEM and EDX mapping/analysis, and (D) HRTEM image. The inset in (A) shows an SEM image of
the nanocrystals at a higher magnification. The scale bars in the inset is 20 nm. The yellow and green colors in (C) correspond to Pd and Pt elements,
respectively.
reaction time of 12 h was sufficient to generate PdꢀPt concave
nanocubes in high quality. When the reaction time was extended
to 20 h (Figure 1D), all the side faces of the Pd nanocubes were
found to disappear due to the galvanic replacement. The newly
formed Pt atoms preferred to continuously grow these sites close
to the corners of a Pd template in order to facilitate the exchange
or {100} facets were excavated and tiny tips of Pt started to appear
at the corners of each Pd nanocube. When the amount of H PtCl
2
6
was increased to 1.4 mg (Figure 2B), the side faces were further
etched due to galvanic replacement, together with deposition of
more Pt at the corners of the Pd nanocube, resulting in the
formation of PdꢀPt concave nanocubes. After adding 3.5 mg of
H PtCl (Figure 2C), high-quality PdꢀPt concave nanocubes
2
ꢀ
of electrons between Pd and PtCl6 , eventually leading to the
formation of an octapod characterized by slightly elongated arms
with enlarged bumps at the tips. Taken together, these results
2
6
with an average edge length of 18 nm were obtained. Notably,
when the amount of H PtCl was further increased to 7.0 mg
2
6
2
ꢀ
indicate that the galvanic replacement between Pd and PtCl6
(Figure 2D), the PdꢀPt concave nanocubes were etched into
octapods with large bumps of Pt due to the complete removal of
Pd from the {100} faces and extensive lateral deposition of Pt at
the corners. This branched structure was also confirmed by SEM
imaging (Figure S2, Supporting Information). Taken together, we
can conclude that PdꢀPt concave nanocubes with different weight
percentages of Pt could be readily obtained by controlling the
amount of H PtCl dialed into the reaction.
ꢀ
could readily occur in the presence of Br . Furthermore, the
oxidation and dissolution of Pd seemed to be preferentially
initiated and continued on the {100} side faces.
We could easily control the extent of galvanic replacement
2ꢀ
between Pd nanocubes and PtCl6 ions by varying the amount of
H PtCl . Figure 2 shows typical TEM images of PdꢀPt nano-
2
6
crystals that were prepared via galvanic replacement with the
2
6
standard procedure, except the difference in H PtCl amount.
We also used SEM, HRTEM, HAADF-STEM, and EDX to
characterize the morphology, structure, and composition of the
2
6
When reacted with 0.5 mg of H PtCl (Figure 2A), the side faces
2
6
6
082
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
Figure 4. TEM images of PdꢀPt nanocrystals obtained using the standard procedure, except the addition of different amounts of KBr: (A) 0, (B) 8, (C)
0, (D) 60, (E) 150, and (F) 600 mg. The insets show TEM images of individual nanocrystals at a higher magnification. The scale bars in the insets
2
are 10 nm.
PdꢀPt concave nanocubes prepared using the standard proce-
dure. The SEM images in Figure 3A show that most of the PdꢀPt
nanocrystals still had a cubic shape, with a size similar to that of
the original Pd nanocubes. The side faces of the PdꢀPt
nanocubes were concaved due to the galvanic replacement.
The concave structure was also supported by HAADF-STEM
imaging (Figure 3B). The distribution of Pt and Pd in the
concave nanocube was determined using EDX analysis
major elements in the sample: Pd, Pt, Br, C, and Cu. Obviously,
Pd and Pt came from the PdꢀPt concave nanocubes, while Br, C,
and Cu came from the reactant KBr, the PVP coating, and the
carbon-coated Cu grid, respectively. The EDX mapping clearly
shows a color difference between the interior (mainly Pd, yellow)
and the surface (mainly Pt, green), confirming a bimetallic
structure. Moreover, most of the Pt was deposited at the corners
of the concave nanocube. Figure 3D shows a typical HRTEM
image of an individual PdꢀPt concave nanocube along the [100]
(Figure 3C). As shown by the EDX spectrum, there were five
6
083
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
Figure 5. TEM images of PdꢀPt nanocrystals prepared via galvanic replacement using the standard procedure, except the difference in reaction
temperature: (A) 50, (B) 60, (C) 70, and (D) 80 ꢀC. The insets show TEM image of individual nanocrystals at a higher magnification. The scale bars in
the insets are 10 nm.
zone axis. The HRTEM image clearly shows well-resolved,
continuous fringes in the same orientation, indicating that the
concave nanocube was a single crystal. We believe that the same
crystal structure and negligible lattice mismatch between Pd and
Pt (0.77%) were responsible for the formation of a single-crystal,
PdꢀPt bimetallic nanostructure during galvanic replacement.
The fringes with lattice spacing of 2.0 and 1.4 Å can be indexed to
the {200} and {220} planes of a face-centered cubic (fcc) lattice,
respectively. As a result, the side surface of the concave region is
likely bounded by a mix of both {100} and {110} facets.
concave nanocubes. Figure 4 shows TEM images of PdꢀPt
nanocrystals prepared using the standard procedure except for
the amount of KBr added into the reaction solution. When no
KBr was added (Figure 4A), the side faces of Pd nanocubes were
slightly excavated due to oxidation and dissolution of Pd via
2
ꢀ
galvanic replacement reaction with PtCl6 . In this case, we
ꢀ
believe that the small amount of Br adsorbed on the {100} side
faces of Pd nanocubes during their synthesis was responsible for
the galvanic replacement. To support this argument, Pd octahe-
drons with edge length of 16 nm were also employed as templates
ꢀ
The Effect of Bromide on the Initiation of Galvanic
for the galvanic replacement reaction in the absence of Br
ꢀ
Replacement. From the viewpoint of potential difference, the
(Figure S3, Supporting Information). In this case, no Br was
2ꢀ
galvanic replacement between Pd and PtCl6 should be able to
occur spontaneously. However, no obvious sign of galvanic
replacement between Pd nanocubes and H PtCl was observed
used during the synthesis of Pd octahedrons. As a result, no
galvanic replacement took place in the case of Pd octahedrons if
ꢀ
2ꢀ
no additional Br ions were added. In this case, PtCl6 could be
slightly reduced by PVP to generate small Pt nanoparticles as
marked by arrows in Figure S3, Supporting Information. As the
amount of KBr increased, the resultant PdꢀPt nanocrystals
became more concaved on the side faces (Figure 4BꢀF). We
2
6
when they were mixed in an aqueous solution and heated up to
00 ꢀC. In the present work, the galvanic replacement between
Pd nanocubes and H PtCl seems to be induced and then
1
2
6
ꢀ
facilitated by Br ions, including the formation of the PdꢀPt
6
084
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
Figure 6. (A) TEM and (B) SEM images of Pd cuboctahedrons with an average edge length of 25 nm. (C and E) TEM and (D and F) SEM images of
PtCl being added, respectively. The insets show TEM image of
concave PdꢀPt cuboctahedrons synthesized at 90 ꢀC for 12 h, with 1.4 and 3.5 mg of H
2
6
individual nanocrystals at a higher magnification. The scale bars in the insets are 10 nm.
ꢀ
can conclude that Br ions played a critical role in both initiating
and promoting the galvanic replacement between Pd templates
pairs. At 60 ꢀC (Figure 5B), atoms started to be removed from the
side faces of a Pd nanocube due to galvanic replacement. As the
temperature was further increased to 70 ꢀC (Figure 5C), the side
faces of a Pd nanocube became more excavated. High-quality
PdꢀPt concave nanocubes could not be obtained until the reaction
temperature had been increased to at least 80 ꢀC (Figure 5D). We
believe that the use of an elevated temperature could speed up the
2
ꢀ
and PtCl6
.
The Effect of Reaction Temperature. The reaction tempera-
ture was also found to play an important role during the formation of
PdꢀPt concave nanocubes via galvanic replacement because the
rate of galvanic replacement was strongly dependent on this
parameter. Figure 5 shows TEM images of PdꢀPt nanocrystals
obtained at different reaction temperatures, with all other param-
eters being the same as in the standard procedure. When the
reaction was performed at 50 ꢀC (Figure 5A), the Pd nanocubes
were slightly truncated at the corners (which are more active than
2ꢀ
galvanic replacement between Pd nanocubes and PtCl
ions by
6
2ꢀ
accelerating the diffusion of PtCl6 ions to the Pd nanocubes and
that the dissolution of Pd from the surface both are beneficial to the
formation of a concave structure.
Galvanic Replacement with Pd Nanocrystals in Different
Sizes and Shapes. We also examined Pd nanocrystals with
ꢀ
ꢀ
the side faces) due to oxidative etching by Cl /O and Br /O
2
2
6
085
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
other different sizes and shapes as the sacrificial templates for
any two metals with proper difference in redox potentials.
2
ꢀ
2ꢀ
the galvanic replacement with PtCl6 ions. Figure S4, Sup-
porting Information shows TEM images of Pd nanocubes of 10
and 6 nm in size that were used as the sacrificial templates and
their corresponding PdꢀPt concave nanocubes. From the
images, it is clear that high-quality PdꢀPt concave nanocubes
with roughly the same size as the starting Pd nanocubes were
also obtained via the galvanic replacement reaction. Combined
with the results obtained from Pd nanocubes of 18 nm in size,
we can conclude that the size of PdꢀPt concave nanocubes
could be controlled by simply using Pd nanocubes with
different sizes as the templates.
Because the reduction potential of PtCl6 /Pt (0.74 V versus
2
ꢀ
RHE) is more positive than that of PdCl /Pd (0.62 V versus
4
2
ꢀ
RHE), Pd can be oxidized by PtCl6 according to the following
equation:
2
f Pt þ 2PdCl42ꢀ
2
Pd þ PtCl6 ꢀ þ 2Clꢀ
ð1Þ
ðsÞ
ðaqÞ
ðaqÞ
ðsÞ
ðaqÞ
However, our observations indicate that the galvanic re-
2ꢀ
placement between Pd nanocrystals and PtCl6 ions could
ꢀ
not occur in the absence of Br . From eq 1, we expect that the
addition of Cl ions could promote galvanic replacement
between Pd and PtCl6 by shifting the equilibrium to the right
ꢀ
2ꢀ
Both octahedral and cuboctahedral Pd nanocrystals enclosed
by {111} or a mix of {111} and {100} facets were also employed
as sacrificial templates for the galvanic replacement in an effort to
better understand the selectivity in initiation of the reaction on
side. Figure S6A, Supporting Information, shows TEM image of
the product obtained using the standard procedure except that
the 300 mg KBr was replaced by 188 mg KCl (to obtain the same
ꢀ
ꢀ
molar concentration for Cl and Br ). This image indicates that
galvanic replacement did occur due to the presence of a large
ꢀ
the {100} facets in the presence of Br . Figure 6A and B shows
TEM and SEM images of Pd cuboctahedrons of about 25 nm in
edge length that served as the sacrificial templates. Careful
examination (inset of Figure 6A) indicates that the projected
image of a cuboctahedron along the [100] axis consisted of a dark
square in the center region corresponding to the {100} facets,
four light isoceles triangles corresponding to the {111} facets,
and four edges corresponding to the {100} facets. After galvanic
replacement reaction with 1.4 mg H PtCl , the dark square in the
ꢀ
amount of Cl . However, the galvanic replacement induced by
Cl showed no selectivity toward different facets of the Pd
ꢀ
templates, leading to the formation of PdꢀPt nanocrystals with
an irregular morphology.
ꢀ
When excess Br is used, the galvanic replacement between
2ꢀ
Pd and PtCl6 should occur according to the following equation
due to much larger stability constants for [PdBr ] and
PtBr6 ] relative to [PdCl4 ] and [PtCl6 ]:
2ꢀ
4
2ꢀ
2ꢀ
2ꢀ 40,41
2
6
[
center region was slightly thinned, while the four isosceles
triangles were thickened, as shown in Figure 6, C and D. The
selective etching at the {100} facets due to galvanic replacement
and deposition of the formed Pt atoms on the {111} facets was
responsible for the changes of contrast in the projected image. As
the amount of H PtCl was further increased to 3.5 mg, PdꢀPt
2PdðsÞ þ PtBr6 ꢀðaqÞ þ 2BrꢀðaqÞ f PtðsÞ þ 2PdBr ðaqÞ ð2Þ
2
2ꢀ
4
2ꢀ
The reduction potential of PtBr6 /Pt (0.61 V versus RHE) is also
more positive than that of PdBr4 /Pd (0.49 V versus RHE). As a
result, Pd can be oxidized by PtBr6 according to eq 2, which is
expected to be further accelerated in the presence of excess Br
2ꢀ
2ꢀ
2
6
ꢀ
concave cuboctahedrons were obtained (Figure 6E and F). From
these images, the dark square at the center and four edges,
corresponding to the {100} facets, were concaved due to
selective galvanic replacement, together with the deposition of
the formed Pt atoms at the four isosceles triangles corresponding
to the {111} facets. These results indicate that the galvanic
ꢀ
ions. Due to the selective adsorption of Br on the {100} facets of
a Pd nanocrystal, the {100} facets (or at least in the vicinity of the
surface) of the template is supposed to be populated by the
2ꢀ
PtBr6 complex, which is a key component for both the initiation
and continuation of the galvanic replacement reaction. We believe
that this chemical difference was responsible for the selective
2
ꢀ
replacement between Pd nanocrystals and PtCl6 ions exhib-
galvanic replacement on the {100} facets of a Pd nanocrystal.
ited a selectivity for the {100} facets of Pd nanocrystals in the
ꢀ
ꢀ
In order to exclude the role of oxidative etching by Br /O
2
presence of Br .
ꢀ
and Cl /O in the formation of concave nanocrystals, KCl
instead of H PtCl with the same molar concentration of Cl
2
The facet selectivity was also observed in the case of Pd
octahedrons. Figure S5, Supporting Information shows TEM
and SEM images of the Pd octahedrons and their corresponding
products obtained with the addition of different amounts of
H PtCl . When Pd octahedrons of 33 nm in size with a smooth
ꢀ
2
6
was used as a reactant, while keeping other conditions un-
changed. Figure S6B, Supporting Information shows TEM image
of the product obtained using the standard procedure except that
2
6
3
.8 mg of KCl instead of 3.5 mg of H PtCl was added. In this
2 6
surface (Figure S5A and B, Supporting Information) reacted with
case, only the sharp corners of the Pd nanocubes were removed
due to oxidative etching by Br /O and Cl /O , no sign of
1
.4 mg H PtCl , the six sharp corners terminated in the {100}
2
6
ꢀ
ꢀ
2
2
facets were selectively removed, while the eight {111} facets were
essentially retained (Figure S5C and D, Supporting In-
formation). Meanwhile, the contrast difference between faces
galvanic replacement was observed in the absence of H PtCl . To
2
6
ꢀ
2ꢀ
ꢀ
2ꢀ
further identify the role of Br ions, PtCl4 instead of PtCl6
was employed as a precursor, because Br ions were supposed
not to strongly promote the galvanic reaction between Pd and
(bright) and edges (dark) indicated that the formed Pt atoms
were mainly deposited on the {111} facets, making the surface
appear rough under SEM imaging (inset of Figure S5D, Support-
ing Information). When the amount of H PtCl was increased to
2
ꢀ
PtCl4 according to the following equation:
Pd þ PtBr42ꢀ
Because PVP can also reduce PtBr4 , no PVP was added into
this reaction. Figure S7, Supporting Information shows TEM
f Pt þ PdBr42ꢀ
ð3Þ
2
6
ðsÞ
ðaqÞ
ðsÞ
ðaqÞ
3
.5 mg (Figure S5E and F, Supporting Information), the {100}
2
ꢀ
facets became increasingly concaved, and the {111} facets were
further thickened and roughened due to the continuous deposi-
tion of Pt. Obviously, no galvanic replacement reaction was
observed on the {111} facets, indicating the selectivity of facet for
2
ꢀ
images of the product obtained with PtCl4 as a reactant in the
2ꢀ
absence of PVP. Based on their potential difference (PtBr
/
4
ꢀ
2ꢀ
the Br -induced galvanic replacement.
Pt = 0.70 V, PdBr4 /Pd = 0.49 V versus RHE), Pd could be
ꢀ
Mechanism for the Formation of PdꢀPt Concave Nano-
oxidized according to eq 3 in the presence of Br ions, and
cubes. In principle, the galvanic replacement can occur between
PdꢀPt octapods should be formed when 7.0 mg of K PtCl was
2
4
6
086
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
Figure 7. Schematic illustrating a plausible mechanism for the formation of a PdꢀPt concave nanocube through the galvanic replacement between a Pd
2
ꢀ
nanocube and PtCl6 ions.
added. However, only the surface of the Pd nanocubes was
Figure 7 illustrates a plausible mechanism responsible for the
formation of PdꢀPt concave nanocubes and summarizes the
evolution of morphologies during the galvanic replacement
reaction. For a nanocrystal with a cubic shape, the reactivity is
supposed to decrease in the order of corner, edge, and side face.
This difference in reactivity can be further amplified due to the
2
ꢀ
oxidized by PtBr4 due to galvanic replacement. In this case,
ꢀ
Br ions did not obviously promote the galvanic replacement
2
ꢀ
between Pd and PtBr4 , which was in agreement with the
stoichiometric argument shown in eq 3.
The role of galvanic replacement in the formation of PdꢀPt
ꢀ
concave nanocubes was also confirmed by ICP-MS analysis of the
selective adsorption of Br on the {100} facets of side faces. As
such, in the initial stage, Pd nanocubes were preferentially
2þ
molar ratio of Pd (in the solution) to Pt (in the concave
2þ
ꢀ
ꢀ
ꢀ
nanocubes). Obviously, Pd could be generated in two different
etched by Br /O and Cl /O at the corners (the most active
2 2
2
ꢀ
ꢀ
ways: the galvanic replacement between Pd and PtCl
and the
sites) in the presence of Br and Cl (step 1 in Figure 7). After
adding a small amount of H PtCl , the PtBr
formed through replacement of Cl in PtCl
stronger binding between Br and Pt
6
ꢀ
ꢀ
2ꢀ
oxidative etching of Pd by Br /O and Cl /O . We determined
the amount of Pd in the solution due to oxidative etching by
using KCl instead of H PtCl to react with Pd nanocubes. Table
S1, Supporting Information, shows the amounts of Pd in the
solution resulting from oxidative etching in the presence of
different amounts of KCl (1.5, 3.1, 3.8, and 7.6 mg) with the
complex was
2
2
2
6
6
2þ
ꢀ
2ꢀ
ꢀ
by Br due to a
6
.
ꢀ
4þ 40
Owing to the
2
6
2þ
ꢀ
2ꢀ
selective adsorption of Br on the {100} facets, the PtBr6
complex was also expected to have a higher concentration in the
vicinity of the {100} facets. As a result, Pd atoms on the {100}
facets were preferentially oxidized and dissolved by reacting
ꢀ
same molar concentration of Cl as H PtCl used in the galvanic
2
6
2
ꢀ
replacement experiments (1.4, 2.8, 3.5, and 7.0 mg). It is clear that
with PtBr6 . The oxidation of Pd led to the production of
2þ
the Pd in the solution obtained through oxidative etching was
electrons, which could quickly migrate to the entire surface of a
2
ꢀ
almost identical in the presence of different amounts of KCl,
Pd nanocube. The PtBr6 ions in the solution were then
reduced to Pt(0) atoms by these electrons and subsequently
deposited at the corners of a Pd nanocube (step 2 in Figure 7).
This is in contrast to what is involved in the galvanic re-
ꢀ
indicatingthatthe Br from300 mgof KBr was mainly responsible
for the oxidative etching. Table S2, Supporting Information, shows
2þ
the amounts of Pd in the solution and Pt in the concave
nanocubes resulting from galvanic replacement in the presence
of different amounts of H PtCl . In these cases, we had subtracted
placement between truncated Ag nanocubes and HAuCl , in
4
which the reaction started simultaneously from all corners of a
truncated nanocube, while the formed Au atoms were mainly
deposited on the side faces due to the strong protection of PVP
2
6
2þ
the contributions to Pd due to oxidative etching. It can be seen
2þ
that the molar ratio of Pd /Pt was approximately 2, in good
agreement with eq 2 for the galvanic replacement reaction except
for the sample that involved 7.0 mg of H PtCl . Careful analysis
2
4
for Ag {100} facets. As the galvanic reaction proceeded, the
Pd {100} facets were gradually removed together with the
deposition of more Pt atoms at the corners, resulting in the
formation of a PdꢀPt concave nanocube (step 3 in Figure 7).
Obviously, the bottom of cavity on the surface of a PdꢀPt
concave nanocube was still terminated in Pd atoms, and the
newly formed Pt atoms were largely deposited at the corners.
Meanwhile, the Pt atoms could also grow laterally, leading to
the formation of PdꢀPt octapods with slightly increased arms
and large bumps at the tips after complete removal of the Pd
{100} facets (step 4 in Figure 7).
2
6
showed that the Pd nanocubes had been etched into PdꢀPt
octapods together with a small number of Pt fragments
(
Figure 2D) in the presence of 7.0 mg of H PtCl . These Pt
2 6
fragments with small sizes still exited in the solution after
centrifugation, and thus the amount of Pt in the concave nano-
cubes was underestimated, resulting in a larger than expected
2þ
molar ratio of Pd /Pt. Taken together, the formation of PdꢀPt
concave nanocubes can be ascribed to the facet-selective galvanic
ꢀ
replacement reaction in the presence of Br ions.
6
087
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
catalyst was calculated in the standard region from 0.02 to 0.4 V.
The specific ECSA (i.e., ECSA per unit weight of a metal) of the
PdꢀPt catalysts with weight percentages of 19.9 and 34.3 was
2
7
9.1 and 86.4 m /g , respectively, which was comparable to that
Pt
2
of the Pt/C catalyst (89.5 m /g , see Figure S8B, Supporting
Pt
Information). In comparison, the PdꢀPt catalysts with low
weight percentages (3.4 and 10.4) of Pt exhibited relatively high
2
specific ECSA (185.8 and 125.8 m /g ).
Pt
The ORR measurements were performed in O -saturated
2
0
.1 M aqueous HClO solutions using a glassy carbon RDE at
4
room temperature at a sweep rate of 10 mV/s and a rotation rate
of 1600 rpm. Figure 8A shows the ORR polarization curves for
the PdꢀPt concave nanocubes with Pt weight percentages of 3.4,
1
0.4, 19.9 and 34.4, and Pt/C catalysts. On the basis of an
equivalent total mass of Pd and Pt, PdꢀPt concave nanocubes
with different percentages of Pt showed relatively lower activities
than Pt/C due to the inactive ORR component of Pd. The kinetic
current was calculated from the polarization curve by considering
the mass-transport correction and normalized with respect to the
loading amount of Pt in order to compare the Pt mass activities
for different catalysts (Figure 8B). At 0.9 V versus RHE, the Pt
mass activities of the PdꢀPt concave nanocubes decreased with
weight percentages of Pt, which was in agreement with the trend
of specific ECSA. Moreover, the PdꢀPt concave nanocubes with
a Pt weight percentage of 3.4 show Pt mass activities four times
larger than that of the Pt/C catalyst. Obviously, only those Pt
atoms on the surface of a catalyst were directly involved in ORR.
Those in the bulk would merely increase the loading of Pt and
thus reduce the mass specific activity. For the catalyst with a low
weight percentage of Pt, most of the Pt atoms were deposited on
the surface of Pd nanocrystals and directly contributed to the
ORR activity. In addition, as previously reported by Adzic
Figure 8. (A) ORR polarization curves for Pt/C and PdꢀPt concave
nanocubes (prepared from 18 nm Pd nanocubes) with four different
weight percentages for Pt at room temperature in an O -saturated 0.1 M
2
4
HClO solution with an equivalent total mass of metals. The data were
recorded at a sweep rate of 10 mV/s and a rotation rate of 1600 rpm. For
the PdꢀPt concave nanocubes and Pt/C catalysts, the total loading of
2
metals on the RDE was 15.3 μg/cm . The current densities were
2
normalized relative to the geometric area of RDE (0.196 cm ). (B)
44,45
The mass activity of these five catalysts normalized by the weight of Pt at
et al.,
a Pt monolayer deposited on a Pd surface could have
0.9 V versus RHE.
enhanced ORR activity due to an electronic coupling between
these two metals. When the amount of Pt atoms exceeded a
monolayer, the extra Pt atoms would be deposited on the outer
surface with a reduced coupling to the Pd support. As such, the
mass specific ORR activity would decrease with increasing Pt
weigh percentage. These results suggest that the methodology
based on the galvanic replacement between Pd nanocrystals and
Electrocatalytic Activity toward the ORR. The PdꢀPt
concave nanocubes with Pt weight percentages of 3.4, 10.4,
1
9.9, and 34.4 were obtained via a galvanic replacement reaction
with 0.5, 1.4, 3.5, and 7.0 mg of H PtCl and then tested as
2
6
ꢀ
electrocatalysts for the ORR. We benchmarked the electrocata-
lytic activity of these PdꢀPt nanocrystals against the commercial
Pt/C catalyst (E-TEK, 20 wt % of 3.2 nm Pt nanoparticles on
Vulcan XC-72 carbon support). Figure S8A, Supporting Infor-
mation, shows CV curves for these five different catalysts
a Pt salt precursor in the presence of Br ions could provide a
simple and versatile route to the preparation of highly active
ORR electrocatalysts for proton-exchange membrane (PCM)
fuel cells.
recorded at room temperature in N -purged 0.1 M aqueous
2
’
CONCLUSIONS
HClO solutions at a sweep rate of 50 mV/s. For all samples, the
4
total loading of metals (both Pd and Pt) on the glassy carbon
We have systematically investigated the galvanic replacement
2
2ꢀ
electrode was 15.3 μg
/cm . The geometric area of each RDE
reaction between PtCl6 ions and Pd nanocrystals with con-
trollable sizes and well-defined shapes, including cubes, cubocta-
hedrons, and octahedrons. We have shown that the addition of
metal
2
was 0.196 cm . The CV curves exhibited both potential regions
associated with H adsorption and desorption processes
upd
þ
ꢀ
ꢀ
(
H þ e = H ) from 0 to 0.4 V and reversible adsorption
Br ions could promote the galvanic replacement reaction.
upd
þ
3
ꢀ
ꢀ
of OH (2H O = OH þ H O þ e ) beyond 0.6 V, where
Specifically, the Br -induced galvanic replacement reaction
ad
2
ad
Hupd and OH refer to the underpotentially deposited hydrogen
showed high selectivity toward the Pd {100} facets, resulting in
the formation PdꢀPt concave nanocrystals, for cubes, cubocta-
hedrons, and octahedrons. It was proposed that the preferential
ad
42
and the adsorbed hydroxyl species, respectively. Due to the
intense peak of Hupd adsorption and desorption at ∼0.05 V
43
ꢀ
originating from the Pd surface uncovered by the Pt shell, the
electrochemically active surface area (ECSA) of the PdꢀPt
catalyst was underestimated by only measuring the charges
collected in the Hupd adsorption region from 0.1 to 0.4 V after
adsorption of Br ions on the Pd {100} facets and their strong
4
þ
complexing interaction with Pt ions were responsible for the
selective galvanic replacement. The PdꢀPt concave nanocubes
with a 3.4 wt. % of Pt exhibited the largest specific ECSA and
were four times more active for the ORR (on the basis of
equivalent Pt mass) than the commercial Pt/C catalyst. This
2
double-layer correction and assuming a value of 210 μC/cm for
the adsorption of a hydrogen monolayer. The ECSA of the Pt/C
6
088
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089

Journal of the American Chemical Society
ARTICLE
work not only greatly advances our understanding of the galvanic
(12) Herricks, T.; Chen, J. Y.; Xia, Y. Nano Lett. 2004, 4, 2367.
(13) Au, L.; Lu, X. M.; Xia, Y. Adv. Mater. 2008, 20, 2517.
2ꢀ
replacement between Pd nanocrystals and PtCl
ion in the
6
ꢀ
ꢀ
(
14) Xiong, Y. J.; Wiley, B.; Chen, J. Y.; Li, Z. Y.; Yin, Y. D.; Xia, Y.
Angew. Chem., Int. Ed. 2005, 44, 7913.
15) Huang, X. Q.; Tang, S. H.; Zhang, H. H.; Zhou, Z. Y.; Zheng,
N. F. J. Am. Chem. Soc. 2009, 131, 13916.
16) Berkovitch, N.; Ginzburg, P.; Orenstein, M. Nano Lett. 2010,
presence of halogen ions (e.g., Br and Cl ) but also provides a
facile and versatile approach to the production of PdꢀPt concave
nanocrystals with improved electrocatalytic activity for fuel cell
applications.
(
(
1
0, 1405.
’
ASSOCIATED CONTENT
(17) Zhang, J. A.; Langille, M. R.; Personick, M. L.; Zhang, K.; Li,
S. Y.; Mirkin, C. A. J. Am. Chem. Soc. 2010, 132, 14012.
(18) Zhang, H.; Li, W. Y.; Jin, M. S.; Zeng, J.; Yu, T.; Yang, D. R.; Xia,
Y. Nano Lett. 2011, 11, 898.
S
Supporting Information. TEM images of Pd nanocubes,
b
SEM images of PdꢀPt concave nanocubes, TEM images of Pd
octahedrons and their corresponding PdꢀPt nanocrystals pre-
pared in the absence of KBr, TEM images of PdꢀPt concave
nanocubes with different sizes, SEM and TEM images of Pd
octahedrons and their corresponding PdꢀPt concave structure,
TEM images of PdꢀPt nanocrystals synthesized in the presence
of KCl, TEM images of PdꢀPt nanocrystals using K PtCl as a
(19) Yu, T.; Kim, D.; Zhang, H.; Xia, Y. Angew. Chem., Int. Ed. 2011,
50, 2773.
(20) Sun, Y.; Mayers, B. T.; Xia, Y. Nano Lett. 2002, 2, 481.
(21) Skrabalak, S. E.; Chen, J. Y.; Sun, Y. G.; Lu, X. M.; Au, L.;
Cobley, C. M.; Xia, Y. Acc. Chem. Res. 2008, 41, 1587.
(22) Seo, D.; Song, H. J. Am. Chem. Soc. 2009, 131, 18210.
(23) Sun, Y.; Xia, Y. J. Am. Chem. Soc. 2004, 126, 3892.
(24) Chen, J.; McLellan, J. M.; Siekkinen, A.; Xiong, Y.; Li, Z. Y.; Xia,
2
4
precursor, ICP-MS data of Pd and Pt, and CV curves and specific
ECSA. This material is available free of charge via the Internet at
http://pubs.acs.org.
Y. J. Am. Chem. Soc. 2006, 128, 14776.
(25) Lu, X.; Tuan, H. Y.; Chen, J.; Li, Z. Y.; Korgel, B. A.; Xia, Y.
J. Am. Chem. Soc. 2007, 129, 1733.
’
AUTHOR INFORMATION
(26) Huang, X.; Zhang, H.; Guo, C.; Zhou, Z.; Zheng, N. Angew.
Chem., Int. Ed. 2009, 48, 4808.
(27) Chen, Z.; Waje, M.; Li, Z.; Yan, Y. Angew. Chem., Int. Ed. 2007,
Corresponding Author
xia@biomed.wustl.edu
46, 4060.
(28) Liang, H. P.; Zhang, H. M.; Hu, J. S.; Guo, Y. G.; Wan, L. J.; Bai,
Author Contributions
These authors contributed equally.
C. L. Angew. Chem., Int. Ed. 2004, 43, 1540.
(29) Wang, L.; Yamauchi, Y. J. Am. Chem. Soc. 2010, 132, 13636.
(30) Shao, M. H.; Shoemaker, K.; Peles, A.; Kaneko, K.; Protsailo, L.
J. Am. Chem. Soc. 2010, 132, 9253.
(31) Peng, Z. M.; Yang, H. J. Am. Chem. Soc. 2009, 131, 7542.
’
ACKNOWLEDGMENT
(
32) Wang, J. X.; Inada, H.; Wu, L. J.; Zhu, Y. M.; Choi, Y. M.; Liu, P.;
Zhou, W. P.; Adzic, R. R. J. Am. Chem. Soc. 2009, 131, 17298.
33) Lim, B.; Jiang, M. J.; Camargo, P. H. C.; Cho, E. C.; Tao, J.; Lu,
X. M.; Zhu, Y. M.; Xia, Y. Science 2009, 324, 1302.
34) Zhang, J.; Vukmirovic, M. B. V.; Xiu, Y.; Mavrikakis, M.; Adzic,
R. R. Angew. Chem., Int. Ed. 2005, 44, 2132.
35) Jin, M. S.; Liu, H. Y.; Zhang, H.; Xie, Z. X.; Liu, J. Y.; Xia, Y.
Nano Res. 2011, 4, 83.
(36) Bard, A. J.; Faulkner, L. R. Electrochemical Methods Fundamen-
tals and Application, 2nd ed.; John Wiley & Sons: New York, 2001.
This work was supported in part by a grant from the NSF
(
DMR-0804088), a DOE subcontract from the University of
(
Delaware (DE-FG02-03 ER15468), and startup funds from
Washington University in St. Louis. As a visiting scholar from
Zhejiang University, H.Z. was partially supported by the “New
Star Program” of Zhejiang University. J.G.W. was also supported
by a grant from CNMT (2010K000336) under the 21st Frontier
R&D Program of the MEST, Korea. Part of the research was
performed at the Nano Research Facility (NRF), a member of
the National Nanotechnology Infrastructure Network (NNIN),
which is supported by the NSF under award ECS-0335765.
(
(
(
37) Schmidt, T. J.; Gasteiger, H. A.; St €a b, G. D.; Urban, P. M.; Kolb,
D. M.; Behm, R. J. J. Electrochem. Soc. 1998, 145, 2354.
38) Xiong, Y. J.; Chen, J. Y.; Wiley, B.; Xia, Y. J. Am. Chem. Soc.
005, 127, 7332.
39) Xiong, Y. J.; Cai, H. G.; Wiley, B. J.; Wang, J. G.; Kim, M. J.; Xia,
(
2
’
REFERENCES
(
(
1) Burda, C.; Chen, X. B.; Narayanan, R.; El-Sayed, M. A. Chem.
Rev. 2005, 105, 1025.
2) Tian, N.; Zhou, Z. Y.; Sun, S. G.; Ding, Y.; Wang, Z. L. Science
007, 316, 732.
Y. J. Am. Chem. Soc. 2007, 129, 3665.
(40) Srivastava, S. C.; Newman, L. Inorg. Chem. 1966, 5, 1506.
(41) Feldberg, S.; Klotz, P.; Newman, L. Inorg. Chem. 1972,
11, 2860.
(42) Gomez, R.; Orts, J. M.; Alvarez-Ruiz, B. L.; Feliu, J. J. Phys.
Chem. B 2004, 108, 228.
(
2
(
(
3) Lim, B.; Xia, Y. Angew. Chem., Int. Ed. 2011, 50, 76.
4) Tian, N.; Zhou, Z. Y.; Sun, S. G. J. Phys. Chem. C 2008,
1
12, 19801.
5) Camden, J. P.; Dieringer, J. A.; Zhao, J.; Van Duyne, R. P. Acc.
Chem. Res. 2008, 41, 1653.
6) Ma, Y. Y.; Kuang, Q.; Jiang, Z. Y.; Xie, Z. X.; Huang, R. B.; Zheng,
L. S. Angew. Chem., Int. Ed. 2008, 47, 8901.
7) Yu, Y.; Zhang, Q. B.; Lu, X. M.; Lee, J. Y. J. Phys. Chem. C 2010,
14, 11119.
8) Lu, C. L.; Prasad, K. S.; Wu, H. L.; Ho, J. A. A.; Huang, M. H.
J. Am. Chem. Soc. 2010, 132, 14546.
9) Xia, Y.; Xiong, Y. J.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int.
Ed. 2009, 48, 60.
(43) Zhang, J.; Huang, M.; Ma, H.; Tian, F.; Pan, W.; Chen, S.
Electrochem. Commun. 2007, 9, 63.
(
(44) Adzic, R. R.; Zhang, J.; Sasaki, K.; Vukmirovic, M. B.; Shao, M.;
Wang, J. X.; Nilekar, A. U.; Mavrikakis, M.; Valerio, J. A.; Uribe, F. Top.
Catal. 2007, 46, 249.
(45) Zhou, W. P.; Yang, X. F.; Vukmirovic, M. B.; Koel, B. E.; Jiao, J.;
Peng, G. W.; Mavrikakis, M.; Adzic, R. R. J. Am. Chem. Soc. 2009,
131, 12755.
(
(
1
(
(
(
10) Tao, A. R.; Habas, S.; Yang, P. D. Small 2008, 4, 310.
(11) Lim, B.; Jiang, M. J.; Tao, J.; Camargo, P. H. C.; Zhu, Y. M.; Xia,
Y. Adv. Funct. Mater. 2009, 19, 189.
6
089
dx.doi.org/10.1021/ja201156s |J. Am. Chem. Soc. 2011, 133, 6078–6089