Letter

pubs.acs.org/NanoLett

Shape-Controlled Synthesis of Palladium Nanocrystals: A

Mechanistic Understanding of the Evolution from Octahedrons to
Tetrahedrons
Yi Wang,†,‡ Shuifen Xie,† Jingyue Liu,§ Jinho Park,† Cheng Zhi Huang,‡ and Younan Xia*,†
†
The Wallace H. Coulter Department of Biomedical Engineering, School of Chemistry and Biochemistry, and School of Chemical and
Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332, United States
‡
Education Ministry Key Laboratory on Luminescence and Real-Time Analysis, College of Chemistry and Chemical Engineering,
Southwest University, Chongqing 400715, People’s Republic of China
§
Department of Physics, Arizona State University, Tempe, Arizona 85287, United States
*
S Supporting Information

ABSTRACT: Palladium octahedrons and tetrahedrons en-

closed by eight and four {111} facets have been synthesized
from cuboctahedral Pd seeds by using Na2PdCl4 and
Pd(acac)2, respectively, as the precursors. Our mechanistic
studies indicate that the cuboctahedral seeds were directed to
grow into octahedrons, truncated tetrahedrons, and then
tetrahedrons when Pd(acac)2 was used as a precursor. In
contrast, the same batch of seeds only evolved into
octahedrons with increasing sizes when the precursor was
switched to Na2PdCl4. The difference in growth pattern could
be attributed to the different reduction rates of these two
precursors. The fast reduction of Pd(acac)2 led to a quick drop in concentration for the precursor in the very early stage of a
synthesis, forcing the growth into a kinetically controlled mode. In comparison, the slow reduction of Na2PdCl4 could maintain
this precursor at a relatively high concentration to ensure thermodynamically controlled growth. This work not only advances
our understanding of the growth mechanism of tetrahedrons but also offers a new approach to controlling the shape of metal
nanocrystals.
KEYWORDS: Nanocrystal, palladium, tetrahedron, precursor, reaction kinetics

S hape control has proven to be a powerful means for

tailoring and controlling the properties of metal nanocryst-
als and thus optimizing their performance in a broad range of
applications related to photonics, electronics, catalysis, sensing,
and biomedicine.1 To this end, metal nanocrystals with a
variety of different shapes have been achieved over the past
decade,2 and the growth mechanisms for some of these shapes
have been established through the use of seed-mediated
synthesis. For example, starting from single-crystal cubic or
cuboctahedral seeds, nanocrystals with a variety of shapes
including cubes, truncated cubes, cuboctahedrons, truncated
octahedrons, and octahedrons could all be obtained by altering
the surface free energies of different facets with a capping agent
and thus the growth rates along different directions.3 As for
pentagonal rods or wires, they could be obtained via axial
growth along the 5-fold axes of decahedral seeds.4 In addition,
nanocrystals with concave and convex surfaces could be
generated by confining the growth selectively to the corners/
edges and side faces of seeds, respectively.5 Despite these
successful demonstrations, it is still challenging to elucidate the
growth pathways that lead to the formation of metal
nanocrystals with a variety of shapes and the possible
correlations between the nanocrystals with different shapes.
In contrast to many other basic shapes (e.g., cubic,
octahedral, cuboctahedral, spherical, and rod- or barlike)
commonly taken by single-crystal metal nanocrystals, the
tetrahedral shape has been elusive and rarely explored. Like
an octahedron, the surface of a tetrahedron is also completely
covered by {111} facets. However, the much larger (by 1.3
times) surface area to volume ratio relative to an octahedron
makes the tetrahedral shape less favorable in a thermodynami-
cally controlled synthesis. There are a few reports on the
synthesis of metal nanocrystals with a tetrahedral shape but
their growth mechanisms are yet to be elucidated. For example,
El-Sayed and co-workers reported the first synthesis of Pt
tetrahedrons as a mixture with Pt cubes when K2PtCl4 was
reduced by H2 in the presence of polyacrylate as a colloidal
stabilizer.6 Kaneda and co-workers reported the synthesis of Pd
tetrahedrons by using tetranuclear Pd cluster as the precursor
Received: March 9, 2013
Revised: April 2, 2013
Published: April 9, 2013

© 2013 American Chemical Society 2276 dx.doi.org/10.1021/nl400893p | Nano Lett. 2013, 13, 2276−2281
Nano Letters
Figure 1. (a) Schematic illustration showing the formation of Pd octahedrons and tetrahedrons, respectively, by using Na2PdCl4 and Pd(acac)2 as
the precursors for seed-mediated growth. (b−d) Typical TEM images of the Pd cuboctahedral seeds, octahedrons, and tetrahedrons. The insets
show the corresponding samples at a higher magnification with the scale bars being 5 nm. The 20 nm scale bar applies to all other images.
without the involvement of any reductant and stabilizer.7
Huang and co-workers recently synthesized Pt tetrahedrons by
taking advantage of the selective capping effect of a peptide
engineered with a specific sequence.8 Zheng and co-workers
and Yan and co-workers prepared Pd concave tetrahedrons and
Pt−Pd tetrahedrons, respectively, by introducing formaldehyde
into a hydrothermal reaction.9 All of these studies were unable
to decipher the growth pathway that led to the formation of
tetrahedrons because it was essentially impossible to separate
the nucleation and growth steps in these one-pot syntheses.
Here we demonstrate that seed-mediated growth could be
used to obtain Pd nanocrystals with a tetrahedral or octahedral
shape depending on the precursor. The use of presynthesized
seeds with the same size, shape, crystallinity, surface capping
agent, dispersion medium, and particle concentration allowed
us to single out one specific parameter for study.10 As a model
system, Pd cuboctahedrons of 5 nm in size were synthesized
using a polyol method and then used as seeds for growth.
Except for the use of different precursors, all other experimental
conditions including temperature, reductant, solvent, and the
concentrations of seeds/reagents were kept the same. When
Pd(acac)2 was used as a precursor, we obtained octahedrons,
truncated tetrahedrons, and then tetrahedrons at different
stages of growth. In contrast, the same seeds only evolved into
octahedrons with increasing sizes when we switched the
precursor from Pd(acac)2 to Na2PdCl4. We also investigated
the mechanism responsible for the formation of Pd nanocryst-
als with a tetrahedral shape and found that it was related to the
much faster reduction rate of Pd(acac)2 relative to Na2PdCl4.
Figure 1a shows a schematic that compares the evolution
pathways from single-crystal, cuboctahedral seeds of Pd to
octahedral and tetrahedral nanocrystals when Na2PdCl4 and
Pd(acac)2 were used as the precursors, respectively. The Pd
cuboctahedrons were 5 nm in size and synthesized by reducing
Na2PdCl4 with ethylene glycol as a solvent and reductant at 160
°C in the presence of poly(vinyl pyrrolidone) (PVP) as a
stabilizer. Figure 1b shows a typical transmission electron
2277
Letter
microscopy (TEM) image of the as-obtained cuboctahedral
seeds of Pd with a purity approaching 100%. The Pd
cuboctahedrons were enclosed by a mix of both {111} and
{100} facets on the surface.
We conducted the seed-mediated growth of Pd nanocrystals
by introducing different precursors into a polyol system
containing both the Pd seeds and PVP. In a typical synthesis,
the Pd seeds and PVP were first dissolved in tetraethylene
glycol (TTEG) at 140 °C under magnetic stirring. Here TTEG
acted as both a solvent and a reductant while PVP served as a
colloidal stabilizer. The two different Pd precursors, Na2PdCl4
and Pd(acac)2, were dissolved in TTEG and then quickly
injected into the reaction mixtures with a pipet, respectively.
The color of the solution was gradually darkened with time
after the injection of both precursors, indicating that the
precursor compounds were reduced to Pd atoms by TTEG at
the used temperature. After 1 h, we obtained Pd nanocrystals
with distinct shapes in the two syntheses conducted in parallel.
As shown by TEM image in Figure 1c, Pd octahedrons
enclosed by eight {111} facets were obtained in the presence of
Na2PdCl4 as a precursor with an average edge length of 12.8 ±
1.3 nm. In comparison, Pd tetrahedrons enclosed by four {111}
facets were obtained under the same condition, except for the
use of Pd(acac)2 as precursor (Figure 1d). The average edge
length of the Pd tetrahedrons was 14.5 ± 1.0 nm. Clearly, the
difference in final shape for the Pd nanocrystals obtained
through the seed-mediated growth should be attributed to the
different precursors involved since all other parameters,
including the seeds, temperature, reductant/solvent, and the
concentrations of reagents, were kept the same.
In order to understand how Pd nanocrystals with distinct
shapes were formed when the two different precursors were
used, we analyzed the products sampled at different reaction
times by TEM. In this case, aliquots were taken out from the
reaction mixture at different time points in the early stage (≤10
min) of a standard synthesis after the introduction of Pd
precursor. Figure 2a−c shows TEM images and the
dx.doi.org/10.1021/nl400893p | Nano Lett. 2013, 13, 2276−2281
Nano Letters Letter

Figure 2. TEM images of Pd nanocrystals obtained in the early stages of standard syntheses when (a−c) Na2PdCl4 and (d−f) Pd(acac)2 were used
as the precursors, respectively. (a,d) 0.5, (b,e) 1.5, and (c,f) 10 min. The insets show 3D models of the corresponding products obtained at different
time points. The 20 nm scale bar applies to all images.

Figure 3. Structural analyses of the Pd tetrahedrons and truncated tetrahedrons obtained in a standard synthesis: (a) HAADF-STEM image of a Pd
tetrahedron; (b) HAADF-STEM image of a Pd truncated tetrahedron; (c) typical TEM images of individual Pd truncated tetrahedrons at different
orientations; and (d) 3D models of truncated tetrahedrons at different orientations matching the TEM images in (c). The insets in (a) and (b) show
the corresponding FT patterns for the Pd tetrahedron and truncated tetrahedron.

corresponding three-dimensional (3D) models of the Pd
nanocrystals obtained at different time points when Na2PdCl4
was used as a precursor. At t = 0.5 min, Pd octahedrons with
slight truncation at corners were observed. This result
confirmed that the growth was mainly limited to the ⟨100⟩
directions in the initial stage due to the higher surface free
2278
energy of {100} facets relative to {111} facets.11 Once the
nanocrystals were fully enclosed by {111} facets, the octahedral
shape would not change any more with the increase of reaction
time, while the edge length gradually increased from 5.5 ± 0.5
nm at t = 0.5 min to 6.7 ± 0.6 and 9.0 ± 0.8 nm for t = 1.5 and
10 min, respectively.
dx.doi.org/10.1021/nl400893p | Nano Lett. 2013, 13, 2276−2281
Nano Letters Letter

When Pd(acac)2 rather than Na2PdCl4 was used as precursor,

the cuboctahedral seeds were immediately transformed into
octahedrons with an average edge length of 7.1 ± 0.5 nm at t =
0.5 min (Figure 2d). The growth mode during this period of
time was similar to the case of Na2PdCl4 in which the {100}
facets grew at a relatively faster rate than the {111} facets.
Different from the case of Na2PdCl4, only four of the {111}
facets of an octahedron could further grow in the following step
to generate a truncated tetrahedron enclosed by eight {111}
facets with unequal surface areas at t = 1.5 min (Figure 2e).
Afterward, sharp corners started to appear on the truncated
tetrahedron to form a perfect tetrahedron at t = 10 min (Figure
2f). It should be pointed out that the edge length of a
tetrahedron obtained at the end of a synthesis was exactly twice
that of an octahedron obtained at t = 0.5 min (14.5 ± 1.0 nm vs
7.1 ± 0.5 nm). This result corresponds to an ideal situation for
the formation of the smallest tetrahedron from an octahedral
seed, as illustrated in Figures S1 and S2 of the Supporting
Information. It indicates that the overgrowth only occurred on
four of the eight {111} facets of an octahedron formed at the
initial stage as there was essentially no growth for the other four
{111} facets. In addition to the mechanistic study of growth
pathway, we have also optimized the amount of precursor
added into the solution in an effort to obtain Pd tetrahedrons
with the highest possible purity and uniformity, as well as to
avoid the generation of Pd nanoparticles via homogeneous
nucleation (Supporting Information Figure S3).
We further characterized the Pd tetrahedrons and truncated
tetrahedrons shown in Figures 1d and 2e by high-resolution
high-angle annular dark-field scanning-transmission electron
microscopy (HAADF-STEM). As shown in Figure 3a,b, the
atomic lattices of a tetrahedron and truncated tetrahedron can
be clearly resolved. The lattice fringe spacing of 1.9 and 2.3 Å
marked on their surfaces can be indexed to the {200} and
{111} reflections of face-centered cubic Pd, respectively. The
Fourier transform (FT) patterns (insets) obtained from
selected areas of the corresponding nanocrystals indicated Figure 4. Comparisons of the reaction kinetics when the two different
that they were single crystals sitting against a plane precursors were used: (a) plots showing the percentages (determined
perpendicular to the [110] zone axis, confirming that the by ICP-MS) of Pd2+ remaining in the reaction solutions as a function
{111} facets were exposed on the surfaces of both Pd of reaction time; and (b) photographs showing changes to the solution
tetrahedrons and truncated tetrahedrons. In order to better color with increasing reaction time for standard syntheses with no
appreciate the shape of a truncated tetrahedron, we obtained cuboctahedral Pd seeds.
TEM images of individual nanocrystals at different orientations
relative to the electron beam (Figure 3c,d, together with their 2.2% within 10 min. This result was also consistent with the
corresponding 3D models). Owing to the unequal surface areas solution colors observed at different time points. As shown in
of the eight {111} facets on a truncated tetrahedron, its profiles Figure 4b (top trace), the solution gradually changed from a
were quite different under TEM when it was sitting on the grid deep yellow color to light brown, deep brown, and black within
with different orientations. Figure S4 in the Supporting 10 min when Na2PdCl4 was reduced by the polyol. However,
Information also shows TEM images of individual Pd the solution quickly turned to light brown at 0.5 min, deep
tetrahedrons at different orientations. brown at 1.5 min, and dark at 3 min when Pd(acac)2 was
To better understand how the reaction kinetics of a synthesis reduced by the polyol (Figure 4b, bottom trace). The Pd seeds
was correlated with the type of precursor, we measured the were not added in these control experiments to better observe
percentages of metal ions remaining in the reaction solution as the color changes as the precursors were reduced. These
a function of time using inductively coupled plasma mass observations confirmed that Pd(acac)2 could be reduced at a
spectrometry (ICP-MS). As shown in Figure 4a, the reduction much faster rate than Na2PdCl4 under the same condition,
of Na2PdCl4 to Pd atoms was relatively slow and the percentage which could be attributed to a relatively weak binding force of
of Pd2+ ions remaining in the solution only decreased to 79.1% the [acac]− ligand to Pd2+ relative to Cl−, and thus a faster
at t = 0.5 min and then to 63.6, 44.7, 19.7, and 4.8% at 1.5, 3.0, release rate of Pd2+ from the Pd(acac)2. In another set of
5.0, and 10 min, respectively. In contrast, the conversion of experiments (Supporting Information Figures S5 and S6), we
Pd(acac)2 into Pd atoms was much faster than the Na2PdCl4 also demonstrated that halides such as Cl−, Br−, and I− could all
under the same condition, particularly in the early stage. The replace the [acac]− ligand in the precursor of Pd(acac)2 by
percentage of Pd2+ ions in the solution quickly dropped to forming relatively more stable complexes and thus slow down
22.9% at t = 0.5 min and was then maintained at a level of ca. the reduction of Pd2+ ions into Pd atoms. As a result, Pd
2279 dx.doi.org/10.1021/nl400893p | Nano Lett. 2013, 13, 2276−2281
Nano Letters
nanocrystals with shapes other than tetrahedrons were obtained
in these cases. A similar observation was also reported by
Zheng and co-workers, where the stability of [PdI4]2− was
found to be higher than Pd(acac)2 and the morphologies of Pd
nanocrystals generated by reducing Pd(acac)2 with DMF were
drastically varied with the addition of I−.12 It was also found
that [acac]− did not have any effect on the reaction kinetics
when Na2PdCl4 was used as a precursor, further suggesting the
stronger binding energy of Cl− with Pd2+ ions than [acac]−. In
another experiment, we introduced Cl− into the reaction
system at t = 1.5 min after most of the Pd(acac)2 had already
been reduced. As shown in Supporting Information Figure S7,
most of the nanocrystals in the products were still Pd
tetrahedrons. This result indicated that the ligand exchange
between [acac]− and Cl− might occur on a slower time scale
relative to the reduction of Pd(acac)2. As such, the growth
pathway from octahedron to tetrahedron was not altered.
A tetrahedron can only be formed under a kinetically
controlled condition because it has a much higher (1.3 times)
surface area to volume ratio than an octahedron. As we
discussed before, the newly formed Pd atoms resulting from the
reduction of a Pd precursor tended to be deposited on the
{100} facets of a Pd cuboctahedral seed due to the difference in
surface free energy between the {100} and {111} facets. It is
worth noting that the edge length of the octahedrons obtained
in the early stage of a synthesis involving Pd(acac)2 was
relatively larger than the octahedrons obtained with Na2PdCl4
as the precursor (7.1 ± 0.5 nm vs 5.5 ± 0.5 nm at t = 0.5 min,
as shown in Figure 2). This difference suggests that more
precursor had been reduced to Pd atoms and subsequently
included into the Pd octahedrons for the case of Pd(acac)2 due
to its faster reduction rate. This observation was consistent with
the ICP-MS data shown in Figure 4a. For the syntheses
involving Pd(acac)2 and Na 2 PdCl 4 as the precursors,
respectively, the percentages of remaining Pd precursor in the
solutions were 22.9 and 79.1% at t = 0.5 min. As such, the Pd
atoms newly formed through the reduction of Pd(acac)2 would
not be able to nucleate on all of the eight {111} facets of a Pd
octahedron. Instead, only four of the eight {111} facets of an
octahedron could be involved in the heterogeneous nucleation
and growth (as illustrated in Figure 1a and Figure S1 in the
Supporting Information). Once a cluster (or nucleus) of Pd
atoms had been created on a certain face of a seed, the
reduction of precursor in the following step would preferen-
tially occur at this site rather than on other regions due to the
lower energy barrier.13 The site-localized growth could be
retained as long as the reduction rate is faster than the surface
diffusion rate of adatoms.14 Further growth of the four {111}
facets eventually resulted in the formation of Pd tetrahedrons.
In contrast, when Na2PdCl4 was used as a precursor, the
availability of this precursor around a Pd octahedral seed at a
relatively high concentration in the early stage of a synthesis
allowed the nucleation to occur on all eight facets of the Pd
seed. As a result, all of the eight {111} facets could undergo
further growth, leading to the formation of Pd octahedrons
with gradually increasing sizes. Even after the concentration of
Pd2+ ions had dropped to a very low level at a later stage of the
synthesis, the slow reduction rate of this precursor relative to
surface diffusion of adatoms still could not lead to localized
growth and thus formation of octahedrons.14
The formation of a tetrahedron from an octahedron can also
be explained using another possible mechanism in which the
newly formed Pd atoms were preferentially deposited at the
2280
Letter
corners/edges rather than the side faces of the Pd octahedron
formed at the early stage owing to the higher surface energy of
these sites than the side faces. Then, the adatoms could migrate
to the side faces through surface diffusion due to the relatively
high temperature used.14 When Pd(acac)2 was used as a
precursor, the adatoms could only migrate to some of the eight
side faces of an octahedron due to the fast reduction rate and
the relatively low concentration of available Pd2+ ions, resulting
in the formation of a tetrahedral shape enclosed by four {111}
facets. In contrast, when the precursor was switched to
Na2PdCl4, the slow reduction rate and the availability of Pd2+
ions around a Pd octahedral seed at a relatively high
concentration allowed them to migrate to all eight facets of
the Pd seed, leading to the formation of a Pd octahedron.
Taken together, we can conclude that it was the difference in
reduction rate for Na2PdCl4 and Pd(acac)2 that led to the
formation of products with thermodynamically and kinetically
controlled shapes (i.e., octahedrons and tetrahedron), respec-
tively.
In summary, we have demonstrated the effect of precursor on
the reaction kinetics and thus growth pathway of a seed-
mediated synthesis. When Pd(acac)2 was used as a precursor,
Pd cuboctahedral seeds could be directed to evolve into
octahedrons, truncated tetrahedrons, and finally tetrahedrons as
the reaction proceeded. In contrast, the same Pd seeds could
only evolve into octahedrons when Na2PdCl4 was added into
the reaction system as precursor. This study clearly
demonstrates that nanocrystals enclosed by the same type of
facet but in different shapes could be obtained by manipulating
the reduction kinetics of a precursor. This work greatly
advances our understanding of the growth mechanism for
nanocrystals with a tetrahedral shape. We believe this strategy
based upon the use of an appropriate precursor to manipulate
the reaction kinetics could also be extended to cover other
metals and even other types of inorganic materials.
■
ASSOCIATED CONTENT
*
S Supporting Information
Experimental details and additional figures. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: younan.xia@bme.gatech.edu.
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
This work was supported in part by a grant from NSF (DMR-
1215034) and startup funds from Georgia Institute of
Technology. Y.X. was partially supported by the World Class
University (WCU) program through the National Research
Foundation of Korea funded by the Ministry of Education,
Science and Technology (R32-20031). As a jointly supervised
Ph.D. student from Southwest University, Y.W. was also
partially supported by a Fellowship from the China Scholarship
Council (CSC).
■
REFERENCES
(1) (a) Jin, R.; Cao, Y.; Mirkin, C. A.; Kelly, K. L.; Schatz, G. C.;
Zheng, J. G. Science 2001, 294, 1901. (b) Sun, Y.; Xia, Y. Science 2002,
298, 2176. (c) Murphy, C. J.; Jana, N. R. Adv. Mater. 2002, 14, 80.
(d) Burda, C.; Chen, X.; Narayanan, R.; El-Sayed, M. A. Chem. Rev.
dx.doi.org/10.1021/nl400893p | Nano Lett. 2013, 13, 2276−2281
Nano Letters Letter

2005, 105, 1025. (e) Tao, A.; Sinsermsuksakul, P.; Yang, P. Nat.
Nanotechnol. 2007, 2, 435. (f) Anker, J. N.; Hall, W. P.; Lyandres, O.;
Shah, N. C.; Zhao, J.; Van Duyne, R. P. Nat. Mater. 2008, 7, 442.
(g) Wang, C.; Daimon, H.; Onodera, T.; Koda, T.; Sun, S. Angew.
Chem., Int. Ed. 2008, 47, 3588. (h) McEachran, M.; Keogh, D.;
Pietrobon, B.; Cathcart, N.; Gourevich, I.; Coombs, N.; Kitaev, V. J.
Am. Chem. Soc. 2011, 133, 8066. (i) Liu, X.; Wang, D.; Li, Y. Nano
Taday 2012, 7, 448.
(2) (a) Tao, A. R.; Habas, S.; Yang, P. Small 2008, 4, 310. (b) Xia, Y.;
Xiong, Y.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int. Ed. 2009, 48, 60.
(3) (a) Zeng, J.; Zheng, Y.; Rycenga, M.; Tao, J.; Li, Z.-Y.; Zhang, Q.;
Zhu, Y.; Xia, Y. J. Am. Chem. Soc. 2010, 132, 8552. (b) Xia, X.; Zeng, J.;
Oetjen, L. K.; Li, Q.; Xia, Y. J. Am. Chem. Soc. 2012, 134, 1793.
(4) (a) Pietrobon, B.; McEachran, M.; Kitaev, V. ACS Nano 2009, 3,
21. (b) Langille, M. R.; Zhang, J.; Mirkin, C. A. Angew. Chem., Int. Ed.
2011, 50, 3543.
(5) (a) Zhang, Q.; Li, W.; Moran, C.; Zeng, J.; Chen, J.; Wen, L.-P.;
Xia, Y. J. Am. Chem. Soc. 2010, 132, 11372. (b) Xia, X.; Zeng, J.;
McDearmon, B.; Zheng, Y.; Li, Q.; Xia, Y. Angew. Chem., Int. Ed. 2011,
50, 12542. (c) Zhang, H.; Li, W.; Jin, M.; Zeng, J.; Yu, T.; Yang, D.;
Xia, Y. Nano Lett. 2011, 11, 898. (d) Xie, S.; Lu, N.; Xie, Z.; Wang, J.;
Kim, M. J.; Xia, Y. Angew. Chem., Int. Ed. 2012, 51, 10266.
(6) Ahmadi, T. S.; Wang, Z. L.; Green, T. C.; Henglein, A.; El-Sayed,
M. A. Science 1996, 272, 1924.
(7) Norimatsu, F. Y.; Mizokoshi, Y.; Mori, K.; Mizugaki, T.; Ebitani,
K.; Kaneda, K. Chem. Lett. 2006, 35, 276.
(8) Chiu, C.-Y.; Li, Y.; Ruan, L.; Ye, X.; Murray, C. B.; Huang, Y. Nat.
Chem. 2011, 3, 393.
(9) (a) Huang, X.; Tang, S.; Zhang, H.; Zhou, Z.; Zheng, N. J. Am.
Chem. Soc. 2009, 131, 13916. (b) Yin, A.-X.; Min, X.-Q.; Zhang, Y.-W.;
Yan, C.-H. J. Am. Chem. Soc. 2011, 133, 3816.
(10) (a) Jana, N. R.; Gearheart, L.; Murphy, C. J. Adv. Mater. 2001,
13, 1389. (b) Nikoobakht, B.; El-Sayed, M. A. Chem. Mater. 2003, 15,
1957. (c) Chen, Y.-H.; Hung, H.-H.; Huang, M. H. J. Am. Chem. Soc.
2009, 131, 9114. (d) Langille, M. R.; Personick, M. L.; Zhang, J.;
Mirkin, C. A. J. Am. Chem. Soc. 2012, 134, 14542.
(11) Vitos, L.; Ruban, A. V.; Skriver, H. L.; Kollár, J. Surf. Sci. 1998,
411, 186.
(12) Huang, X.; Zhang, H.; Guo, C.; Zhou, Z.; Zheng, N. Angew.
Chem., Int. Ed. 2009, 48, 4808.
(13) (a) Zeng, J.; Zhu, C.; Tao, J.; Jin, M.; Zhang, H.; Li, Z.-Y.; Zhu,
Y.; Xia, Y. Angew. Chem., Int. Ed. 2012, 51, 2354. (b) Zhu, C.; Zeng, J.;
Tao, J.; Johnson, M. C.; Schmidt-Krey, I.; Blubaugh, L.; Zhu, Y.; Gu,
Z.; Xia, Y. J. Am. Chem. Soc. 2012, 134, 15822. (c) Xia, X.; Xia, Y. Nano
Lett. 2012, 12, 6038.
(14) Xia, X.; Xie, S.; Liu, M.; Peng, H.-C.; Lu, N.; Wang, J.; Kim, M.
J.; Xia, Y. Proc. Natl. Acad. Sci. U.S.A. 2013, DOI: 10.1073/
pnas.1222109110.

2281 dx.doi.org/10.1021/nl400893p | Nano Lett. 2013, 13, 2276−2281