Metal exchange method using Au25 nanoclusters as templates for alloy nanoclusters with atomic precision

Article abstract of DOI:10.1021/ja511635g

A metal exchange method based upon atomically precise gold nanoclusters (NCs) as templates is devised to obtain alloy NCs including Cu_xAu_25-x(SR)₁₈, Ag_xAu_25-x(SR)₁₈, Cd₁Au₂₄(SR)₁₈, and Hg₁Au₂₄(SR)₁₈ via reaction of the template with metal thiolate complexes of Cu^II, Ag^I, Cd^II, and Hg^II (as opposed to common salt precursors such as CuCl₂, AgNO₃, etc.). Experimental results imply that the exchange between gold atoms in NCs and those of the second metal in the thiolated complex does not necessarily follow the order of metal activity (i.e., galvanic sequence). In addition, the crystal structure of the exchange product (Cd₁Au₂₄(SR)₁₈) is successfully determined, indicating that the Cd is in the center of the 13-atom icosahedral core. This metal exchange method is expected to become a versatile new approach for synthesizing alloy NCs that contain both high- and low-activity metal atoms.

Full text of DOI:10.1021/ja511635g

Communication
pubs.acs.org/JACS
Metal Exchange Method Using Au25 Nanoclusters as Templates for
Alloy Nanoclusters with Atomic Precision
Shuxin Wang,†,§ Yongbo Song,†,§ Shan Jin,† Xia Liu,‡ Jun Zhang,† Yong Pei,‡ Xiangming Meng,†
Man Chen,† Peng Li,† and Manzhou Zhu*,†
†Department of Chemistry and Center for Atomic Engineering of Advanced Materials, Anhui University, Hefei, Anhui 230601,
People’s Republic of China
‡Department of Chemistry, Key Laboratory of Environmentally Friendly Chemistry and Applications of Ministry of Education,
Xiangtan University, Xiangtan, Hunan 411105, People’s Republic of China
S
* Supporting Information
Recent research has reported an anti-galvanic reduction
ABSTRACT: A metal exchange method based upon
atomically precise gold nanoclusters (NCs) as templates
is devised to obtain alloy NCs including CuxAu25‑x(SR)18,
AgxAu25‑x(SR)18, Cd1Au24(SR)18, and Hg1Au24(SR)18 via
reaction of the template with metal thiolate complexes of
CuII, AgI, CdII, and HgII (as opposed to common salt
precursors such as CuCl2, AgNO3, etc.). Experimental
results imply that the exchange between gold atoms in
NCs and those of the second metal in the thiolated
complex does not necessarily follow the order of metal
activity (i.e., galvanic sequence). In addition, the crystal
structure of the exchange product (Cd1Au24(SR)18) is
successfully determined, indicating that the Cd is in the
center of the 13-atom icosahedral core. This metal
exchange method is expected to become a versatile new
approach for synthesizing alloy NCs that contain both
high- and low-activity metal atoms.
(AGR) approach, suggesting that when Au/Ag nanoparticle
size is smaller than ∼3 nm (i.e., in the NC regime), their reducing
capability is enhanced;6 for example, Au NCs can reduce Ag+ and
Cu2+, which are more active in bulk metals than Au, and similarly
Ag NCs can reduce Cu2+.6a The order of metal activity was
summarized as Fe > Ni > [Au NCs] > Cu > Ag.6a These results
suggest that the AGR method can only be applied to synthesize
CuAu and AgAu NCs; therefore, the production of doped NCs is
quite limited, especially alloy NCs with the second metal having
high activity, which is particularly attractive for tuning catalytic,
magnetic, and other properties.2,7
Previous theoretical work indicates that the stability of
MxAu25‑x(SR)18 (X = Pd, Ag, and Cd) NCs should be similar
−
to that of Au25(SR)18
.
8,9c The stability of these clusters is in part
due to their 8e− shell closing structure.8,9 This electronic factor
indicates that, in the AGR process, Ag and Cu doping should
meet the 8e− shell closing. If this is true, metal doping may not
exclusively follow the order of metal activity, since Cd is a high-
activity metal in the series (Fe > Cd > Co > Ni > Pb > [Au NCs]
> Cu > Hg > Ag > Pd > Pt > Au), and Cd (5s2) might be doped
recisely doped metal nanoclusters (NCs) with a few tens of
P
metal atoms have emerged as a new class of nanomaterial.1
These small, doped NCs are attracting wide interest owing to
their synergistic properties that depend on the atomic
composition, such as catalytic activity, which is largely enhanced
for single Pd- or Pt-atom-doped M1Au24 NCs (M = Pd/Pt)
compared to the homo-gold counterpart,2 as well as
luminescence, which becomes stronger in Ag- or Cu-doped
gold NCs.3 Controlling the number of each type of metal atoms
and their doping sites in the alloy (or bimetallic) NC is crucial, as
the metal composition and doping sites are the keys to tailoring
the alloy NC’s properties. Controlled synthesis of doped NCs at
the atomic level is still quite challenging, and until now only a few
doped NCs have been attained with atomic precision and
molecular purity.2,4,5 A conventional synthetic strategy involves
reducing a mixture of metal precursors by NaBH4 in the presence
of ligands, such as the cases of MxAu25‑x(SR)18 (M = Ag or Cu)
NCs.4 To effectively produce alloy NCs, new methods should be
devised for synthesizing alloy NCs with designed metal
compositions, precise atom numbers, and metal doping site
control. Several fundamental questions remain about metal
doping:3c What metals can be doped into Au NCs? Can one
precisely control the number of doping atoms? Where would
these dopants go in the Au NC structure?
−
into the Au25(SR)18 NC to form Cd1Au24(SR)180. Consid-
eration of the electronic factor over the metal activity offers a new
perspective for synthetic control of alloy NCs, especially for
doping high-activity metal into Au NCs, which is not feasible via
the AGR method.
This work demonstrates a metal exchange method for
−
obtaining alloy NCs using Au25(SR)18 NCs as the template.
Reaction of the template with a series of metal−thiolate
complexes (as opposed to metal salts) of groups 10−12 gives
rise to AgxAu25‑x, CuxAu25‑x, Cd1Au24, and Hg1Au24, while
NixAu25‑x, PdxAu25‑x, and PtxAu25‑x are not formed. This distinctly
different behavior between group 10 and groups 11−12 metals
offers some important implications for the metal exchange
mechanism. The crystal structure of Cd1Au24 not only shows that
the charge of this cluster is neutral but also indicates that the Cd
atom is in the center of the M13 core; in other words, this Cd
atom is in the metallic form.
−
In a typical experiment, the Au25(SR)18 NC (Supporting
Information (SI), Figure S1) reacted with thiolated metal
Received: November 12, 2014
Published: March 23, 2015
© 2015 American Chemical Society
4018
DOI: 10.1021/ja511635g
J. Am. Chem. Soc. 2015, 137, 4018−4021

Journal of the American Chemical Society
Communication
Figure 2. (a) ESI mass spectrum of Cd1Au24(SC2H4Ph)18 NCs. Insets
Figure 1. MALDI-MS of MxAu25‑x(SR)18 (x as indicated in each
spectrum) after reaction with different metal ions (NiII, PdII, PtII, CuII,
AgI, CdII, and HgII, all in thiolate complexes). R = C2H4Ph. The reaction
time was set to 2 h. See Figure S3 for wider range mass spectra (2000−
20000 range).
are experimental and simulated isotope patterns of Cd1Au24-
2+
(SC2H4Ph)18Cs2 and Cd1Au24(SC2H4Ph)18Cs+, respectively. (b)
ESI-MS spectrum of Hg1Au24(PhC2H4S)18 NCs. Insets are experimental
and simulated isotope patterns of Hg1Au24(SC2H4Ph)18Cs22+ (together
with simulated Au25(SC2H4Ph)18Cs22+) and Hg1Au24(SC2H4Ph)18Cs+
(together with simulated Au25(SC2H4Ph)18Cs+), respectively.
complexes including AgISR, CuII(SR)2, CdII(SR)2, HgII(SR)2,
NiII(SR)2, PdII(SR)2, and PtII(SR)2, where R = C2H4Ph hereafter.
We note that ZnII(SR)2 could not be synthesized, as thiol cannot
react with Zn2+. Matrix-assisted laser desorption/ionization mass
spectrometry (MALDI-MS) and UV/vis spectroscopy were used
to monitor whether the metal exchange process occurred. As
shown in Figure 1, Ni-, Pd-, and Pt-doped Au25 NCs were not
detected in MALDI-MS. The Cu/Au (or Ag/Au) alloy NCs
showed a distribution of peaks corresponding to different
numbers of Ag (or Cu) atoms doped into the 25-atom cluster
(Figures 1 and S2), and the mass difference between peaks was
equal to the Au and Ag (or Cu) atomic mass difference (i.e., MAu
− MCu = 133.5 Da, MAu − MAg = 89 Da). The total number of
metal atoms is, however, preserved at 25 in all NCs, implying the
robustness of the 25-atom structure. The Cu/Ag doping results
are consistent with the previously reported AGR, in which Au
NCs could reduce Cu2+ and Ag+ ions, albeit Cu and Ag are more
active than Au in the form of bulk metals.
2+
these peaks are assigned to Cd1Au24(SR)18Cs2 (calcd m/z
3787.8) and Cd1Au24(SR)18Cs+ (calcd m/z 7442.8), respectively,
and the experimental and simulated isotope patterns match well
(Figure 2a inset). Taken together, ESI-MS indicates Cd1Au24-
(SR)18 is neutral since the number of adducted Cs+ ions equals
the charge number. For Hg1Au24(SR)18, the ESI spectrum
(Figure 2b) showed peaks at m/z 3831.9 and 7531.0, assigned to
Hg1Au24(SR)18Cs22+ (calcd m/z 3831.9) and Hg1Au24(SR)18Cs+
(calcd m/z 7530.8), confirmed by isotope pattern analysis
(Figure 2b inset). Similar to the case of Cd1Au24(SR)18,
Hg1Au24(SR)18 is also charge neutral. Considering the similar
masses of Hg and Au, we note that the experimental isotope
pattern and mass of Hg1Au24(SR)18 differ from those of
Au25(SR)18 (Figure 2b inset); hence, Au25(SR)18 as the product
is ruled out. The charge neutrality of Hg- or Cd-doped alloy NCs
is further confirmed by the absence of TOA+ signals in the
MALDI mass spectra (Figure S6) and NMR spectra (Figure S7).
0
Surprisingly, Au25 NCs were not able to reduce NiII/PdII/PtII,
but they reacted readily with CdII (note: Cd metal is more active
than Ni, Pd, and Pt), and the Cd-doping product was identified
to be mono-doped Cd1Au24(SR)18 (m/z = 7309, Figure 1). No
residual Au25(SR)18 was observed in the mass spectrum,
indicating a nearly complete conversion. These findings are
important because Pt and Pd are less active than Ag and Cu, while
Cd is more active than Ni and would not be doped into Au NCs
according to the AGR approach. But our results suggest that
Au25(SR)18− is capable of reacting with CdII. For mercury doping,
a single Hg atom was doped into the Au NC (Figure 1) in nearly
100% yield; of note, the Hg and Au masses are very close, and
thus the mass peak of Hg1Au24(SR)18 appears at the Au25(SR)18
position (Figure 1), but high-precision MS analysis rules out
Au25(SR)18 (vide infra).
The Cd1Au24(SC2H4Ph)18 NCs obtained by the metal
exchange method allow the growth of single crystals. X-ray
crystallography determined the structure of the doped NCs
0
(Figure 3, for details see SI ). The Cd1Au24 cluster (ligand
omitted for simplicity) adopts the same core−shell structure as
10
−
0
0
that of Au25 and Au25
.
The Cd1Au24 core consists of a
centered 13-atom icosahedral M13 core, further capped by a
second shell comprised of 12 Au atoms. The structure of the
Ag/Cu/Cd/Hg doping into the 25-atom NC perturbs the
optical absorption (Figure S4), but the spectral profile is
essentially retained in these cases, while when metal salts were
used, the NCs finally decomposed (Figure S5). To confirm the
newly attained single-Cd/Hg-doped M1Au24(SR)18 NCs and
probe their charge states, we further performed electrospray
ionization (ESI) MS analysis. For Cd1Au24(SR)18 NCs, the
positive mode ESI spectrum (Figure 2a) showed two intense
peaks at m/z 3787.9 and 7442.9 (the most abundant isotope
peaks); since cesium acetate was added to form Cs+ adducts,
0
Figure 3. (a) Total structure of Cd1Au24(PhC2H4S)18 NCs. Color
labels: dark green = Cd; yellow = Au; magenta = S; all C and H atoms are
shown in wireframe. (b) Superatomic orbital of [Cd1Au12]6+.
4019
DOI: 10.1021/ja511635g
J. Am. Chem. Soc. 2015, 137, 4018−4021

Journal of the American Chemical Society
Communication
Cd1Au24 NC closely resembles Au250, which has less distortion
the same as that of the Cu reduced by Au25(SR)180 NCs reported
by Wu et al.6a The Ag 3d peak (368.30 and 374.30 eV, Figure
S11b) of AgxAu25‑x(SR)18 indicates that the incorporated Ag is
neutral; i.e., the AgI complex was reduced by Au25(SR)18−. In the
case of Cd1Au24(SR)18, the Cd signal (Cd 3d, 405.8 eV) was
found (Figure S11c), but the binding energies of Cd0 and CdII are
not much different (404.9 vs 405.2 eV). The Hg 4f peak (101.2
eV, Figure S11d) of Hg1Au24(SR)18 was observed.
10c The distinct difference between these two NCs is
−
than Au25
.
the central atom in the M13 core, Cd in Cd1Au24 NCs instead of
Au. Hg is similar to Au in terms of the electronic structure, and it
is not easy to distinguish Hg1Au24 and Au25 NCs by X-ray
crystallography. Hg1Au24 should have the same structure as
Cd1Au24, evidenced in our results by both having the same
valence, the same NMR spectrum, and the loss of the same
M1Au4 fragment in MALDI-MS.
To understand the bonding interactions between Cd and the
Au24(SR)18 unit, deformation charge density analysis is carried
out. The deformation charge density is defined as Δρ = ρ(Cd@
Au24(SR)18) − ρ(Au24(SR)18) − ρ(Cd), which describes the
bonding interactions between the central Cd atom and the
surrounding Au24(SR)18 unit. The deformation charge density
displayed in Figure S8 clearly shows the depletion of electronic
charge density at the Cd atom, consistent with the Bader charge
analysis (computational results and details are given in the SI). At
the same time, the accumulation of electronic charge density is
found in the space between the Cd atom and the Au12 shell,
indicating relatively strong bonding interactions between the
central Cd atom and the surrounding Au12 cage. On basis of the
Bader charge analysis and the deformation charge density
analysis, we propose that the central Cd atom is positively
charged.
It is of interest to discuss whether Cd1Au24(SR)18 is a
superatomic species. Cd1Au24(SR)18 has the same number of
valence electrons as Au25(SR)18−. According to the proposed
electron counting rule for a superatom model,9a each staple motif
of the thiolate-protected Au cluster can bind one electron of the
inner metallic Au-core. In view of the structure of Cd1Au24-
(SR)18, the central Cd@Au12 core is protected by six dimeric
staple motifs. Therefore, the Cd@Au12 core can be considered as
a [Cd@Au12]6+ species. The total number of valence electrons of
[Cd@Au12]6+ is 8e−. Electronic structure analysis shows that the
quasi-icosahedron [Cd@Au12]6+ possesses distinct delocalized
1S, 1P, 1D, and 2S superatom orbitals. As shown in Figure 3b, the
1S and three 1P orbitals are occupied. The 1D and 2S orbitals are
all empty. The presence of delocalized superatomic orbitals in the
[Cd@Au12]6+ core suggests that Cd@Au24(SR)18 falls into the
category of superatomic cluster.
Detailed MALDI-MS analyses on MxAu25‑x(SR)18 offer more
structural insight. The structure of Au25(SR)18 consists of a 13-
atom core and six dimeric staples.10 For Ag- and Cu-doped
MxAu25‑x(SR)18, MxAu21−x(SR)14 fragments were formed in both
cases by losing a common Au4(SR)4 unit from the molecular ions
(Figure S9a,b), consistent with the previous work on Cu- or Ag-
doped MxAu25‑x(SR)18 NCs.4 Loss of the Au4(SR)4 unit is
associated with the surface Au2(SR)3 staples (i.e., the latter is
detached from the Au13 core),11a and relevant works have been
reported.11b−d According to the recently reported crystal
structure of AgxAu25‑x(SR)18, Ag atoms are at the surface of the
13-atom core,12 indicating that Ag atoms only exchange with the
Au atoms on the surface of the Au13 core. Unlike the cases of Ag-
and Cu-doped NCs, both Cd1Au24(SR)18 and Hg1Au24(SR)18
NCs are missing a unit of M1Au4(SR)4 (M = Cd/Hg) (Figure
S9c,d), and some other fragmentation pathways were observed in
the case of Hg1Au4(SR)4 (Figure S9d).
The reactions between Au25 NCs and different metal
complexes exhibit some interesting phenomena. The Au25-
−
(SR)18 reaction with CuII(SR)2 was rapid, resulting in
decomposition of Au25 NCs (Figure S12), and no CuxAu25‑x
NCs could be found, but after addition of solid NaBH4 to the
reaction system, CuxAu25‑x NCs resulted (Figure S13). By
extending the reaction time, more atoms of Cu could be doped
into the Au25 NCs (Figure S13); the same is true for Ag doping.
Of note, compared to the direct synthesis route for alloy NCs,4d,e
the present metal exchange method leads to more Cu atoms
doped into the cluster (Figure S13); for example, nine Cu atoms
were doped into the NCs in ∼24 h, while the direct synthesis of
CuxAu25‑x(SR)18 could only achieve a maximum five Cu
atoms.4d,e
The resulting MxAu25‑x NCs exhibit delocalized electron shell
closing (DESC) at 8e− (Scheme 1); this suggests that the
superatom rules for pure Au NCs also apply for alloy NCs.
[Note: The electron number is calculated as 25 − 18 − (−1) = 8
for AgxAu25‑x(SR)18− and CuxAu25‑x(SR)18−, and 24 + 2 − 18−0
= 8 for Cd1Au24(SR)180 and Hg1Au24(SR)180, since Cu and Ag are
of ns1 configuration, whereas Hg and Cd are of ns2.] The 8e−
Scheme 1. Number of Delocalized Electrons (ns) of Doped
Nanoclusters
requirement can explain our experimental observation: when
there is no NaBH4 in the reaction of CuII with Au25(SR)18−, the
product decomposes rapidly (Figure S12), because the exchange
between CuII and Au atoms in the NC alters the total valence of
Cu1Au24 to 0 as the Au atom leaves the cluster in the form of
AuISR, and its electronic number will become 25 − 18 − 0 = 7,
Scheme 2. Metal Exchange Process of Nanoclusters Using
a
Metal Complexes with the Metal Core Size Preserved
X-ray photoelectron spectroscopy (XPS) measurements were
carried out to confirm the presence of the doped metals in the
NCs (Figure S10 for survey scans and Figure S11 for narrow
scans). The Cu 2p peak (932.4 eV, Figure S11a) found in
CuxAu25‑x(SR)18 is close to that of pure Cu metal (932.6 eV) and
a
Only part of the Au25(SR)18 structure is shown: the icosahedral Au13
core surrounded by two Au2SR3 motifs. The full structure of Au25 NCs
is shown in Figure S15.10 R = C2H4Ph. Note that NaBH4 was needed
in the reaction of Cu(II) with Au25 NCs.
4020
DOI: 10.1021/ja511635g
J. Am. Chem. Soc. 2015, 137, 4018−4021

Journal of the American Chemical Society
Communication
hence destabilizing the resultant product. The DESC also
explains why only one Cd or Hg atom could be doped into the
Au25 NCs and why Au25(SR)18 could not react with Cd(II)/
REFERENCES
■
(1) (a) Negishi, Y.; Kurashige, W.; Niihori, Y.; Nobusada, K. Phys.
Chem. Chem. Phys. 2013, 15, 18736. (b) Jin, R.; Nobusada, K. Nano Res.
2014, 7, 285.
0
Hg(II) to form M1Au24 NCs, because their valence and
electronic numbers were changed. All of the obtained alloy
NCs meet the 8e− shell closing structure, suggesting that the
metal exchange process in the NC regime is more critically
dependent on the DESC and the 25-atom structural stability.
An important question arises naturally: What is the driving
force of the metal exchange process? In the above results, the
metals (in the thiolated complex form) that can be doped into
the Au NCs all have fully filled d orbitals with 10e−. The d10−d10
interaction is common in metals of groups 11 and 12. Thus, we
hypothesize that the metal exchange may be caused in part by the
d10−d10 interaction between metal ions and Au NCs. This means
that AgI can exchange for Au atoms in Au25 NCs, and the Ag
atoms in the product (AgxAu25‑x NCs) may also be exchanged by
AuI ions. This is indeed what we observed experimentally (Figure
S14). MALDI-MS spectra clearly show that the Ag atoms in
AgxAu25‑x NCs were exchanged by AuI gradually (Figure S14b,d).
Furthermore, after addition of AgISR, Ag atoms can be re-
exchanged into the NCs. All these results indicate that the metal
exchange is essentially reversible (Scheme 2).
(2) (a) Xie, S.; Tsunoyama, H.; Kurashige, W.; Negishi, Y.; Tsukuda, T.
ACS Catal. 2012, 2, 1519. (b) Qian, H.; Jiang, D.; Li, G.; Gayathri, C.;
Das, A.; Gil, R. R.; Jin, R. J. Am. Chem. Soc. 2012, 134, 16159.
(3) (a) Wang, Q.; Lee, Y.; Crespo, O.; Deaton, J.; Tang, C.; Gysling, H.
J.; Gimeno, C.; Larraz, C.; Villacampa, M. D.; Laguna, A.; Eisenberg, R. J.
Am. Chem. Soc. 2004, 126, 9488. (b) Jia, J.; Wang, Q. J. Am. Chem. Soc.
2009, 131, 16634. (c) Udayabhaskararao, T.; Sun, Y.; Goswami, N.; Pal,
S. K.; Balasubramanian, K.; Pradeep, T. Angew. Chem., Int. Ed. 2012, 51,
2155. (d) Pei, X.; Yang, Y.; Lei, Z.; Wang, Q. J. Am. Chem. Soc. 2013, 52,
6435. (e) Wang, S.; Meng, X.; Das, A.; Li, T.; Song, Y.; Cao, T.; Zhu, X.;
Zhu, M.; Jin, R. Angew. Chem., Int. Ed. 2014, 53, 2376.
(4) (a) Negishi, Y.; Kurashige, W.; Niihori, Y.; Iwasa, T.; Nobusada, K.
Phys. Chem. Chem. Phys. 2010, 12, 6219. (b) Negishi, Y.; Iwai, T.; Ide, M.
Chem. Commun. 2010, 46, 4713. (c) Qian, H.; Barry, E.; Zhu, Y.; Jin, R.
Acta Phys. Chim. Sin. 2011, 27, 513. (d) Negishi, Y.; Munakata, K.;
Ohgake, W.; Nobusada, K. J. Phys. Chem. Lett. 2012, 3, 2209.
(e) Gottlieb, E.; Qian, H.; Jin, R. Chem.Eur. J. 2013, 19, 4238.
(f) Kurashige, W.; Munakata, K.; Nobusada, K.; Negishi, Y. Chem.
Commun. 2013, 49, 5447.
(5) (a) Kumara, C.; Dass, A. Nanoscale 2011, 3, 3064. (b) Kumara, C.;
Dass, A. Nanoscale 2012, 4, 4084. (c) Negishi, Y.; Igarashi, K.; Munakata,
K.; Ohgake, W.; Nobusada, K. Chem. Commun. 2012, 48, 660.
(d) Kothalawala, N.; Kumara, C.; Ferrando, R.; Dass, A. Chem.
Commun. 2013, 49, 10850. (e) Yang, H.; Wang, Y.; Lei, J.; Shi, L.; Wu,
Our results illustrate that, in the metal exchange process, the
−
Au atoms in the Au25(SR)18 NC can be exchanged by metals
with different activities (e.g., Cu, Ag, Cd, Hg, etc.) to produce
new, stable NCs such as Cd1Au24(SR)180 and Hg1Au24(SR)180 in
very high yields (∼100%). This metal exchange method for alloy
NCs is, to a large extent, associated with electron shell closing
and the NC’s structural stability, but less on the metal activity.
These findings shed some new light on the metal exchange
process at the atomic level, and this approach holds promise in
future development as a versatile method for synthesizing alloy
NCs that contain both high- and low-activity metal atoms with
precise control of metal composition, doping site, and dopant
number for specific applications. Future work will also elucidate
how the Cd atom is doped into the center of the M13 core.
X.; Makinen, V.; Lin, S.; Tang, Z.; He, J.; Hakkinen, H.; Zheng, L.;
Zheng, N. J. Am. Chem. Soc. 2013, 135, 9568. (f) Yang, H. Y.; Wang, Y.;
̈
̈
Huang, H. Q.; Gell, L.; Lehtovaara, L.; Malola, S.; Hakkinen, H.; Zheng,
̈
N. F. Nat. Commun. 2013, 4, 2422. (g) Yang, H.; Wang, Y.; Yan, J.; Chen,
X.; Zhang, X.; Hakkinen, H.; Zheng, N. F. J. Am. Chem. Soc. 2014, 136,
̈
7197. (h) Puls, A.; Jerabek, P.; Kurashige, W.; Forster, M.; Molon, M.;
̈
Bollermann, T.; Winter, M.; Gemel, C.; Negishi, Y.; Frenking, G.;
Fischer, R. A. Angew. Chem., Int. Ed. 2014, 53, 4327. (i) Ganesamoorthy,
C.; Weßing, M. S. J.; Kroll, M. S. C.; Seidel, R. W.; Gemel, C.; Fischer, R.
A. Angew. Chem., Int. Ed. 2014, 53, 8077.
(6) (a) Wu, Z. Angew. Chem., Int. Ed. 2012, 51, 2934. (b) Li, H.; Yue, Y.;
Liu, T.; Li, D.; Wu, Y. J. Phys. Chem. C 2013, 117, 16159. (c) Sun, J.; Wu,
H.; Jin, Y. Nanoscale 2014, 6, 5449. (d) Wang, S.; Meng, X.; Feng, Y.;
Sheng, H.; Zhu, M. RSC Adv. 2014, 4, 9680.
(7) Reveles, J. U.; Claybome, P. A.; Reber, A. C.; Khanna, S. N.;
Pradhan, K.; Sen, P.; Pederson, R. Nat. Chem. 2009, 1, 310.
(8) Walter, M.; Moseler, M. J. Phys. Chem. C 2009, 113, 15834.
(9) (a) Walter, M.; Akola, J.; Lopez-Acevedo, O.; Jadzinsky, P. D.;
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental details and characterization data. This material is
available free of charge via the Internet at http://pubs.acs.org.
Calero, G.; Ackerson, C. J.; Whetten, R. L.; Gronbeck, H.; Hakkinen, H.
̈
̈
Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 9157. (b) Hakkinen, H. Chem. Soc.
̈
Rev. 2008, 37, 1847. (c) Jiang, D.-e.; Dai, S. Inorg. Chem. 2009, 48, 2720.
(d) Aikens, C. M. J. Phys. Chem. Lett. 2011, 2, 99.
AUTHOR INFORMATION
■
(10) (a) Heaven, M. W.; Dass, A.; White, P. S.; Holt, K. M.; Murray, R.
W. J. Am. Chem. Soc. 2008, 130, 3754. (b) Zhu, M.; Aikens, C. M.;
Hollander, F. J.; Schatz, G. C.; Jin, R. J. Am. Chem. Soc. 2008, 130, 5883.
(c) Zhu, M.; Eckenhoff, W. T.; Pintauer, T.; Jin, R. J. Phys. Chem. C 2008,
112, 14221.
(11) (a) Wu, Z.; Gayathri, C.; Gil, R. R.; Jin, R. J. Am. Chem. Soc. 2009,
131, 6535. (b) Jiang, D.; Walter, M.; Dai, S. Chem.Eur. J. 2010, 16,
4999. (c) Macdonald, M.; Cheverie, D.; Zhang, P.; Qian, H.; Jin, R. J.
Phys. Chem. C 2011, 115, 15282. (d) Krishna, K. S.; He, M.; Bruce, D. A.;
Kumar, C. S. S. R. Nanotechnol. Rev. 2014, 3, 311.
Corresponding Author
*zmz@ahu.edu.cn
Author Contributions
§S.W. and Y.S. contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
(12) Kumara, C.; Aikens, C. M.; Dass, A. J. Phys. Chem. Lett. 2014, 5,
461.
■
We acknowledge financial support by the NSFC (21072001,
21201005, 21372006), the Ministry of Education and Ministry of
Human Resources and Social Security, the Education Depart-
ment of Anhui Province, the Anhui Province International
Scientific and Technological Cooperation Project, and the 211
Project of Anhui University. Y.P. acknowledges financial support
by NSFC (21373176, 21422305).
4021
DOI: 10.1021/ja511635g
J. Am. Chem. Soc. 2015, 137, 4018−4021