Comparison of the Catalytic Properties of 25-Atom Gold Nanospheres and Nanorods

Article abstract of DOI:10.1016/S1872-2067(10)60238-0

The catalytic properties of two nanocluster catalysts with atomically precisely known structures, icosahedral two-shelled Au₂₅(SC₂H₄Ph)₁₈ nanospheres and biicosahedral Au₂₅(PPh₃)₁₀(SC₂H₄Ph)₅Cl₂ nanorods, were compared. Their catalytic performance in the two reactions of the selective oxidation of styrene and chemoselective hydrogenation of α,β-unsaturated benzalacetone was investigated. The catalytic activities of icosahedral Au₂₅(SC₂H₄Ph)₁₈ nanospheres were superior to those of the bi-icosahedral Au₂₅(PPh₃)₁₀(SC₂H₄Ph)₅Cl₂ nanorods for both reactions. The better catalytic performance of the Au₂₅(SC₂H₄Ph)₁₈ nanospheres can be attributed to their unique core-shell (Au₁₃/Au₁₂) geometric structure that has an open exterior atomic shell and to their electronic structure with an electron-rich Au₁₃ core and an electron-deficient Au₁₂ shell.

Full text of DOI:10.1016/S1872-2067(10)60238-0

CHINESE JOURNAL OF CATALYSIS
Volume 32, Issue 7, 2011
Online English edition of the Chinese language journal
Cite this article as: Chin. J. Catal., 2011, 32: 1149–1155.
RESEARCH PAPER
Comparison of the Catalytic Properties of 25-Atom Gold
Nanospheres and Nanorods
Yan ZHU, Huifeng QIAN, Anindita DAS, Rongchao JIN*
Department of Chemistry, Carnegie Mellon University, Pittsburgh, PA 15213, USA
Abstract: The catalytic properties of two nanocluster catalysts with atomically precisely known structures, icosahedral two-shelled
Au25(SC2H4Ph)18 nanospheres and biicosahedral Au25(PPh3)10(SC2H4Ph)5Cl2 nanorods, were compared. Their catalytic performance in the
two reactions of the selective oxidation of styrene and chemoselective hydrogenation of Į,ȕ-unsaturated benzalacetone was investigated.
The catalytic activities of icosahedral Au25(SC2H4Ph)18 nanospheres were superior to those of the bi-icosahedral Au25(PPh3)10(SC2H4Ph)5Cl2
nanorods for both reactions. The better catalytic performance of the Au25(SC2H4Ph)18 nanospheres can be attributed to their unique
core-shell (Au13/Au12) geometric structure that has an open exterior atomic shell and to their electronic structure with an electron-rich Au13
core and an electron-deficient Au12 shell.
Key words: nanocluster catalyst; icosahedral Au25; biicosahedral Au25; selective oxidation; selective hydrogenation
The importance of atomic level information on the nature of
nanoparticle catalysts has long been recognized in heteroge-
neous catalysis, but getting this information poses major chal-
lenges due to the unavailability of atomically defined nano-
catalysts. To gain deeper insight into the fundamental origin of
nano-metal catalysis and to explore new catalytic applications,
the preparation of well-defined metal catalysts is of critical
importance. Recently, major advances have been made in the
solution phase synthesis of atomically precise gold nano-
clusters by protecting these with thiolate ligands (referred to as
Aun(SR)m, where n is the number of gold atoms in the particle
and m is the number of protecting ligands) [1–10]. Examples
include Au25(SR)18, Au38(SR)24, Au144(SR)60, and so forth
[11–22]. A remarkable feature of these Aun(SR)m nanoclusters
is that they can be dispersed as individual clusters composed of
a specific number (n) of gold atoms with n tunable from tens to
hundreds [7–22]. Relatively small Aun(SR)m nanoclusters (n <
ca. 100) exhibit strong quantum size effects [1]. The
non-metallic behavior of Aun(SR)m nanoclusters in the range
(n) from ~ten to a few hundreds is of particular interest to
catalysis [23–28]. More importantly, the atom packing struc-
tures and unique electronic properties of these Aun(SR)m
nanoclusters [19,29,30] can permit the correlation of particle
structure with catalytic properties, which would allow the
identification of the catalytically active sites on gold nano-
clusters [31].
In our previous work, we have made important advances in
gold nanocatalysis using well defined Aun(SR)m nanoclusters
as catalysts for selective oxidation and hydrogenation proc-
esses, such as the selective oxidation of styrene and the
chemoselective hydrogenation of Į,ȕ-unsaturated ketones (or
aldehydes) [23,31–33]. Concerning the 25 gold atom nano-
clusters, two different structures have been reported  an
icosahedral two-shelled [Au25(SC2H4Ph)18]q (q = –1, 0) nano-
sphere [29,30] and a biicosahedral [Au25(PPh3)10(SC2H5)5Cl2]2+
rod [34]. Recently, we have developed a size-focusing method
that employed phosphine-capped, polydispersed Au nanopar-
ticles as
a common starting material to synthesize
[Au25(SG)18]ꢀ nanospheres (1.27 nm diameter, where SG =
glutathione) and [Au25(PPh3)10(SC2H4Ph)5Cl2]2+ nanorods (0.8
nm u 1.38 nm), respectively, by two-phase and one-phase thiol
etching [35]. The aim was to use these two Au25 structures
(sphere vs. rod) in a comparative study of their catalytic prop-
erties in the hope of finding the factors that affect their catalytic
performance.
Here, we chose the selective oxidation of styrene and selec-
Received 8 February 2011. Accepted 6 April 2011.
*Corresponding author. Tel: +1-412-268-9448; Fax: +1-412-268-1061; E-mail: rongchao@andrew.cmu.edu
This work was supported by CMU, AFOSR, and NIOSH.
Copyright © 2011, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier BV. All rights reserved.
DOI: 10.1016/S1872-2067(10)60238-0

Yan ZHU et al. / Chinese Journal of Catalysis, 2011, 32: 1149–1155
tive hydrogenation of Į,ȕ-unsaturated benzalacetone as the
mmol, dissolved in 5 ml ethanol) was injected rapidly. The
solution immediately turned dark. The reaction was allowed to
proceed for 2 h at room temperature in air. A black product
(containing PPh3-protected Au nanoparticles) was obtained
after the rotary evaporation of the solvent (toluene). The black
product was washed several times with water and hexane to
remove excess PPh3 and TOAB. Then, the black product was
dissolved in chloroform. Unreacted gold salt (i.e. Au(PPh3)Cl)
was removed by precipitating the black product by adding
pentane to the chloroform solution. The black product (~100
mg) obtained was collected. Thermogravimetric (TG) analysis
showed that the organic content in the nanoparticles was about
50 wt%. For the second step, size-focusing conversion of
polydisperse Au nanoparticles (1–3 nm) into biicosahedral
Au25 nanoclusters (formula: [Au25(PPh3)10(SC2H4Ph)5Cl2]2+,
counterion: Clí) was achieved by adding PhC2H4SH to
phosphine-protected Au nanoparticles (~20 mg, dissolved in
20 ml CH2Cl2). The molar ratio of PhC2H4SH to Au atoms was
controlled at PhC2H4SH/Au = 6:1; where it should be noted
that PhC2H4SH was in large excess. The solution was vigor-
ously stirred for 12 h at room temperature. The product (de-
noted as bi-Au25) was obtained after drying the solution by
rotary evaporation, followed by further purification by washing
with hexane and extraction with ethanol.
reactions for comparing the catalytic properties of icosahedral
two-shelled Au25(SC2H4Ph)18 spheres and biicosahedral
[Au25(PPh3)10(SC2H4Ph)5Cl2]2+ rods. The results clearly dem-
onstrated that the icosahedral Au25(SC2H4Ph)18 nanospheres
had a significantly higher catalytic activity than the biicosa-
hedral Au25 nanorods. Based on this, we discuss the factors that
are potentially important for catalytic properties, including
cluster shape (sphere vs. rod), arrangement of surface atoms,
and electronic structures. This work provides the first example
of looking at how the atomic structure of nanoparticle catalysts
affects their catalytic properties.
1 Experimental
1.1 Preparation of Au25 nanocluster
The detailed synthesis and characterization of
Au25(SC2H4Ph)18 spherical nanoclusters were reported in our
previous work [29]. Briefly, HAuCl4·3H2O (0.4 mmol) was
dissolved in water (5 ml). Tetraoctylammonium bromide
(TOAB, 0.47 mmol) was dissolved in toluene (10 ml). The two
solutions were combined in a 25 ml three-neck round bottom
flask. The solution was vigorously stirred with a magnetic bar
to facilitate the phase transfer of the gold(III) salt into the
toluene phase. After ~15 min, the phase transfer was com-
pleted, leaving a clear aqueous phase at the bottom of the flask.
The aqueous phase was removed using a glass pipette. The
toluene solution of gold(III) was purged with N2 and cooled to
1.2 Preparation of catalyst
The CeO2 nanopowder was purchased from Aldrich and
used as received. CeO2 supported Au25 nanoclusters were ob-
tained by mixing a specific amount of CeO2 with a CH2Cl2
solution of the Au25 nanoclusters, followed by stirring for 24 h.
The solution was then dried under a N2 gas stream. To calcine
the Au25/CeO2 catalysts, a thermal treatment in a vacuum oven
was performed.
o
0 C in an ice bath over a period of ~30 min under magnetic
stirring. Phenylethylthiol (PhC2H4SH, 0.17 ml) was added. The
deep red solution turned to faint yellow over a period of 5 min,
and finally to clear over 1 h. After the solution turned clear, a
freshly made aqueous solution of NaBH4 (4 mmol in ~7 ml ice
cold water) was quickly added. The reaction was allowed to
proceed overnight. Then, ethanol (20 ml) was added to separate
Au25 nanoclusters from TOAB and the reaction byproducts.
The Au25(SC2H4Ph)18 clusters (denoted as i-Au25) were col-
lected after removing the supernatant.
1.3 Catalytic tests
The catalytic selective oxidation of styrene reaction was
carried out at atmospheric pressure. CeO2 supported Au25
catalysts (100 mg powder, 1 wt% loading of gold clusters),
styrene (10 mmol), and t-butyl hydroperoxide (TBHP, 38
mmol) were mixed in CH3CN (solvent, 15 ml) in a 50 ml sealed
The details of the two-step synthesis of [Au25(PPh3)10-
(SC2H4Ph)5Cl2]2+ rod nanoclusters were reported in [35].
Briefly, the first step was to prepare triphenylphosphine
(PPh3)-protected Au nanoparticles as follows. HAuCl4·3H2O
(0.118 g, 0.3 mmol, dissolved in 5 ml H2O) and TOAB (0.190
g, 0.348 mmol, dissolved in 10 ml toluene) were combined in a
three-neck round bottom flask. After the solution was vigor-
ously stirred for about 15 min to effect the phase transfer of
Au(III) into the toluene phase, the aqueous phase became clear
and the toluene phase turned dark red, indicating that Au(III)
was completely transferred to the toluene phase. The aqueous
phase was removed and 0.235 g (0.9 mmol) of PPh3 was added
to the toluene solution under vigorous stirring. The toluene
solution became whitish cloudy. Then, NaBH4 (0.034 g, 0.9
o
glass reactor. Then the mixture was heated to 82 C under
constant vigorous stirring in an oil bath for ~20 h. The solvent
was removed by rotary evaporation and the products were
analyzed by nuclear magnetic resonance (1H NMR) spectros-
copy. 1H NMR spectra were collected on a Bruker AvanceTM
300 MHz spectrometer. The conversion and selectivity were
calculated by integrating the NMR peak areas.
For the selective hydrogenation of benzalacetone, benzal-
acetone (0.1 mmol) was dissolved in a mixed solvent (5 ml
toluene and 5 ml ethanol) in a 50-ml three-neck glass flask.
CeO2 supported Au25 catalysts (100 mg powder, 1 wt% loading

Yan ZHU et al. / Chinese Journal of Catalysis, 2011, 32: 1149–1155
of gold clusters) were added. The reaction was initiated by
trons of a Au25 rod are 25–7–2 = 16e after taking the 2+ charge
state into consideration. Therefore, each icosahedral unit of the
rod may be viewed formally as having 8 valence electrons, and
is expected to be relatively electron-rich due to these highly
delocalized valence electrons. The difference in electron dis-
tribution as compared to the Au25 sphere could have important
effects on their catalytic performance.
introducing a H2 flow (under atmospheric pressure). The reac-
tion temperature was 0 oC. The reaction was typically allowed
to proceed under continuous H2 flow for ~4 h. The products
were analyzed by 1H NMR spectroscopy.
2 Results and discussion
For the catalytic experiments, both types of clusters were
supported on CeO2 (powder) with a 1 wt% loading of
ligand-protected clusters. We investigated both calcined and
uncalcined catalysts for the oxidation and hydrogenation reac-
tions. The ultra-small Au25 spheres and rods were barely ob-
servable under conventional bright field TEM due to the low
contrast caused by the ultra-small size. In addition, these ul-
tra-small particles were susceptible to damage by the intense
high energy electron beam, thus high resolution TEM charac-
terization of the supported Au25 particles was difficult. Never-
theless, preliminary characterization by X-ray absorption
spectroscopy indicated no changes in the structure and size of
Au25 after its deposition onto ceria as well as after the catalytic
2.1 Atomic structure of the Au25 nanospheres and nano-
rods
The crystal structures of i-Au25 and bi-Au25 have been re-
ported previously [29,30,34]. These are shown in Fig. 1, where
for clarity, the thiolate and phosphine ligands are represented
only by S and P atoms, respectively.
The spherical Au25 cluster (1.27 nm diameter, Au sur-
face-to-surface distance) adopts a core-shell structure and can
be viewed as an icosahedral Au13 core (Fig. 1(a), magenta)
encapsulated by an exterior shell consisting of the 12 remain-
ing gold atoms (Fig. 1(a), green) [29,30]. The entire particle is
protected by eighteen thiolate ligands. Both experiment and
theory indicated that the Au13 icosahedral core is relatively
electron-rich due to the presence of 8e (for anionic i-Au25ꢀ) or
7e (for neutral i-Au250), while the exterior Au12 shell is rela-
tively electron-deficient due to the covalent bonding of Au-S
[36]. It should be noted that each thiolate ligand formally lo-
o
reaction. However, calcination at above 180 C caused the
sintering of the Au25 nanoclusters due to ligand loss.
2.2 Selective oxidation of styrene
We first discuss the catalytic results of the selective oxida-
tion of styrene on the Au25/CeO2 catalysts. The catalytic reac-
tion temperature was maintained at 82 oC, which is well below
the thiolate desorption temperature of i-Au25 (onset tempera-
calizes one valence electron of Au 6s, that is,
ꢀ
25(Au)-18(ligands)+1 = 8e for Au25(SR)18
.
The Au25 rod (0.8 nm diameter u 1.38 nm length, Au sur-
face-to-surface distance) structure may be viewed as two ico-
sahedrons fused together by sharing a common vertex (i.e.
13+13–1 = 25 atoms, Fig. 1(b)). Its overall shape resembles a
rod. The entire structure is protected by 5 thiolates at the waist,
10 phosphines evenly distributed on the two icosahedrons, and
two chloride ligands on the rod ends. Concerning the electronic
properties of the Au25 rod, phosphine only forms dative bonds
with Au and hence does not localize the valence electrons of
Au 6s as thiolate does, while five thiolate and two chloride
ligands localize 7e in total, thus, the number of valence elec-
o
ture § 190 C, Fig. 2(a)) and of bi-Au25 (onset temperature §
150 oC, Fig. 2(b)), as determined by TG analysis. Note that in
bi-Au25 the loss of triphenylphosphine and Cl ligands starts at
o
175 C (Fig. 2(b)). We have also investigated the use of dif-
ferent atmospheres (e.g. N2, O2) for the TGA analysis and
found that the desorption of the ligands did not depend on the
atmosphere. Thus, in the case of the uncalcined catalysts, all
the ligands remained on the Au25 spheres and rod clusters
during the course of the catalytic reaction. A detailed X-ray
absorption spectroscopic analysis will be given in a future
P
S
(a)
(b)
S
Au
Au
Cl
S
Fig. 1. Atom packing structures of i-Au25 (a) and bi-Au25 (b). The core atoms of gold are shown in magenta. Surface gold atoms and ligand atoms are as
labeled.

Yan ZHU et al. / Chinese Journal of Catalysis, 2011, 32: 1149–1155
(a)
(b)
100
90
80
70
60
50
100
SC2H4Ph loss
90
80
70
60
50
SC2H4Ph loss
PPh3 and Cl loss
50
100
150
200
250
300
350
50 100 150 200 250 300 350 400 450
Temperature (oC)
Temperature (oC)
Fig. 2. TG analysis of i-Au25 (a) and bi-Au25 (b).
work.
focus on comparing the catalytic performances of the i-Au25
and bi-Au25 catalysts.
The selective oxidation of styrene can be carried with either
peroxide (e.g. TBHP) or O2. Here we chose TBHP as the oxi-
dant. The catalytic reaction of styrene oxidation was carried out
in the liquid phase (acetonitrile as the solvent). As shown in
Table 1, the catalytic activity of i-Au25/CeO2 was significantly
higher than that of bi-Au25/CeO2 (Table 1, entries 1 and 3). 95%
conversion of styrene (TOF: 1.3 u 10ꢀ4 mol/(g·s)) was obtained
with the uncalcined i-Au25/CeO2 catalyst, but there was only
57% conversion (TOF: 8 u 10ꢀ5 mol/(g·s)) with the uncalcined
bi-Au25/CeO2 catalyst. Interestingly, both catalysts gave similar
selectivities for epoxide (97% vs. 95%), which indicated that
the different structures of the 25-gold-atom cluster catalysts
had little effect on the selectivity, although a significant effect
on the catalytic activity was observed. It is worth noting that
the activity (95%) and epoxide selectivity (97%) obtained on
the i-Au25/CeO2 catalyst were much higher than those reported
for conventional gold catalysts [37ꢀ42]. We have previously
reported for i-Au25-catalyzed oxidation of styrene in toluene by
O2 [23] or TBHP [32] that the catalytic activity with TBHP as
the oxidant was significantly higher than that with O2. This is
because TBHP is much easier to activate than O2. On the other
hand, the difference in selectivity (97% epoxide in the present
work vs. ~100% for benzaldehyde in the previous work [33])
seems to be due to a promotion effect of acetonitrile. Here, we
During the catalytic reaction, the thiolate ligands will remain
on both the i-Au25 and bi-Au25 catalysts since the reaction
o
temperature (82 C) was well below the ligand desorption
temperature. Reactant molecules such as styrene can readily
penetrate the ligand shell of the Au25 clusters to reach the
cluster surface where styrene and TBHP are adsorbed and
subsequently react on the surface. To further investigate the
two types of Au25 catalysts, we performed a thermal pretreat-
ment of the catalysts before the catalytic reaction. The thermal
treatment step removes surface ligands at least partially (de-
pending on the temperature), so that the Au25 surface becomes
more accessible to reactant molecules. This would give a
higher catalytic activity, but the negative effect is that the
(partial) loss of protecting ligands can cause cluster sintering if
the cluster/support interaction is not strong enough to retain the
clusters. If the latter effect dominates, the resulting larger Au
nanoparticles would be less catalytically active. According to
o
the TG results (Fig. 2), calcination at 260 C leads to a com-
plete removal of ligands from both types of Au25 cluster sur-
faces. Catalytic tests with these catalysts thermally treated at
260 oC showed a drop in activity (Table 1, entries 2 and 5). A
significant drop of activity from 95% to 49% was observed for
the calcined i-Au25/CeO2 catalyst, while the calcined
bi-Au25/CeO2 catalyst showed a smaller drop from 57% to
42%. The selectivity for styrene epoxide was essentially un-
changed and both catalysts still gave 95% selectivity. The drop
in activity for both types of Au25 catalysts indicated that cluster
sintering had occurred, which was also indicated by the color
change of the catalysts to red. The negative effect of thermal
pretreatment was dominant for the calcined catalysts. This is
reasonable since large Au nanoparticles (> 3–5 nm) are known
to be less catalytically active. The reason the i-Au25 catalyst
was much more sensitive to the thermal pretreatment than
bi-Au25 was because of their stability difference. The majority
of i-Au25 clusters have aggregated into larger nanoparticles
Table 1 Catalytic performance of two types of Au25/CeO2 catalysts for
the selective oxidation of styrene by peroxide TBHP
Conversion
Selectivity (%)
Entry
Catalyst
(%)
95
49
57
46
42
14
Epoxide Benzaldehyde Acetophenone
a
1
2
3
4
5
6
i-Au25/CeO2
i-Au25/CeO2
97
95
95
93
95
81
3
4
trace
b
1
2
a
c
bi-Au25/CeO2
bi-Au25/CeO2
3
5
2
b
bi-Au25/CeO2
CeO2
5
trace
5
14
Reaction conditions: Au25/CeO2 100 mg, styrene 10 mmol, TBHP 38
mmol, solvent CH3CN 15 ml, 82 oC, 20 h.
aCatalysts without thermal treatment before catalytic tests.
bCatalysts with thermal treatment at 260 oC for 2 h.
o
after the 260 C pretreatment, while with bi-Au25, sintering
occurred to a lesser extent.
cCatalysts with thermal treatment (in vacuum oven) at 170 oC for 2 h.
For the bi-Au25 catalyst, the TG results (Fig. 2) showed a two

Yan ZHU et al. / Chinese Journal of Catalysis, 2011, 32: 1149–1155
stage desorption of the surface ligands. Thus, one can selec-
large difference in their catalytic activity (95% vs. 57%, Table
o
tively remove the thiolate ligands (at ~175 C) while keeping
1). After ruling out the influence of surface ligands, the re-
maining factors is the 25-atom metal core. A feature of bi-Au25
is that all the surface atoms are relatively electron-rich as the 16
valence electrons are evenly delocalized on the two icosahedra.
As we discussed above in the case of the core-shell i-Au25
catalyst, we believe that the electron-rich bi-Au25 particle
should be capable of similarly adsorbing TBHP since the
bi-Au25 surface is electron-rich, but is less suited for activating
the C=C bond of styrene since the C=C bond activation prefers
positively charged Au sites, which are not available in bi-Au25.
Some surface Au atoms in the bi-Au25 cluster (i.e. those Au
atoms coordinated to Cl and SC2H4Ph ligands) are positive, but
the highly delocalized 16 valence electrons would compensate
for these surface Au atoms bonded to Cl and SC2H4Ph ligands,
and render them less positive, hence, they would have diffi-
culty activating C=C bonds. Apart from the difference in the
electronic effects between i-Au25 and bi-Au25, we also stress
that the hole-like surface site in the i-Au25 cluster is important
for adsorbing and activating reactant molecules. These unique
surface structures are not present in bi-Au25 (Fig. 3(b)). Theo-
retical calculations [45] are underway to provide a deeper
insight into how the surface atomic structure facilitates ad-
sorption and activation of reactant molecules.
PPh3 and Cl ligands, as opposed to the 260 oC pretreatment that
removed all the ligands (SR, PPh3, and Cl). We thus performed
o
a 170 C thermal pretreatment of the bi-Au25/CeO2 catalyst.
The as-treated bi-Au25/CeO2 catalyst gave 46% conversion of
styrene with 93% selectivity for epoxide (Table 1, entry 4),
which was somewhat lower than the activity of the uncalcined
o
catalyst (Table 1, entry 3) but slightly higher than the 260 C
pretreated bi-Au25/CeO2 catalyst (Table 1, entry 5). The partial
removal of surface thiolate ligands at 170 oC should not cause
aggregation of Au25 rod clusters since the PPh3 ligands can still
protect the cluster well. In our previous work, we found that the
structure of i-Au25 clusters remained unaffected after a thermal
o
treatment up to 160 C [43]. The above catalytic results indi-
cated that the removal of thiolate ligands led to an increased
exposure of surface Au atoms but did not have a distinct effect
on the activity. This is consistent with our conclusion above
that the presence of ligands is not a critical factor for the
catalytic activity, since reactant molecules can readily pene-
trate the ligand shell [31,32].
Taken together, the above results clearly showed the major
influences of the structure (sphere vs. rod) of Au25 clusters on
their catalytic activity. It is worth noting that both types of
Au25/CeO2 catalysts only showed a slight decrease (< 5%) in
activity after three cycles and no change in selectivity (not
shown). The icosahedral Au25 cluster adopts a quasi-D2h
symmetry and is protected by 18 thiolate ligands (Fig. 1(a))
[29,30]. This structure possesses an open exterior Au shell
since the Au13 core has 20 triangular faces but only twelve of
these are capped by the 12 exterior Au atoms, leaving eight
triangular Au3 faces uncapped. This can be clearly seen in the
space-filled model (Fig. 3(a)). The eight uncapped sites form
volcano-type structures and may act as the catalytically active
sites [31–33]. This unique core-shell Au13/Au12 structure was
primarily responsible for the facile adsorption and activation of
both oxidant (e.g. TBHP) and the ꢀCH=CH2 group of styrene.
Previous work [36,44] has shown distinct differences in the
charge distribution on the Au13 core and the Au12 shell of a
i-Au25 particle. The presence of partial positive charges on the
surface Au atoms (i.e. AuG+, 0 < G < 1) in Au25(SR)18 would
facilitate the activation of the nucleophilic ꢀCH=CH2 group of
styrene since the positive Au atoms are electrophilic, while the
electron-rich Au13 core would facilitate TBHP activation.
Overall, the unique geometric and electronic properties of the
Au25(SR)18 clusters make them highly active in the selective
oxidation of styrene.
2.3 Selective hydrogenation of benzalacetone
Here, we further investigate the catalytic performance of
i-Au25 and bi-Au25 catalysts in the reaction of the selective
hydrogenation of Į,ȕ-unsaturated benzalacetone with H2. The
reaction was performed in the liquid phase at 0 oC by bubbling
H2 into the solution. The results further showed the importance
of the atomic packing structure and electronic properties of the
Au25 clusters on their catalytic properties. For the i-Au25 clus-
ters we have previously reported 100% selectivity for unsatu-
rated alcohols in the hydrogenation reaction of Į,ȕ-unsaturated
ketones (or aldehydes) with H2 [31].
Table 2 shows a comparison of i-Au25/CeO2 and bi-Au25/
CeO2 catalysts for the selective hydrogenation of benzalace-
tone. The best conversion (34%) was obtained with the uncal-
cined i-Au25/CeO2 catalyst with 100% selectivity for unsatu-
(a)
(b)
Compared to the icosahedral Au25 sphere, the biicosahedral
Au25 rod has a much lower activity as shown above (Table 1).
The bi-Au25 core is constructed by the vertex sharing of two
icosahedral Au13 units. The surface PPh3 ligands in the bi-Au25
structure are much bulkier than the phenylethiolate (SC2H4Ph)
on the i-Au25 surface but this factor should not cause such a
Fig. 3. Space filling models of i-Au25 (a) and bi-Au25 (b). The arrow
indicates one of the unique volcano-like surface sites of the two-shelled
Au25 icosahedral sphere.

Yan ZHU et al. / Chinese Journal of Catalysis, 2011, 32: 1149–1155
spheres was attributed to their unique atomic packing structure
(the presence of volcano-like surface sites that act as active
sites) and unique electronic properties (electron-rich Au13 core
and electron-deficient Au12 shell). This work provided the first
demonstration of the influence of the structure of nanogold
catalysts on their catalytic activity. For the interaction between
Au and the CeO2 support, X-ray absorption spectroscopic
analysis should be able to reveal the nature of the metal-support
interaction and its role in enhancing the catalytic activity.
Theoretical calculations in future work should reveal more
fundamental details of the structure-catalytic property rela-
tionship.
Table 2 Catalytic performance of two types of Au25/CeO2 catalysts for
the selective hydrogenation of benzalacetone with H2
Selectivity (%)
Unsaturated Saturated Saturated
Conversion
(%)
Entry
Catalyst
alcohol
100
100
100
100
100
0
alcohol
ketone
a
1
2
3
4
5
6
i-Au25/CeO2
i-Au25/CeO2
34
4
0
0
0
0
0
0
0
0
0
0
0
0
b
a
c
bi-Au25/CeO2
bi-Au25/CeO2
9
4
b
bi-Au25/CeO2
CeO2
3
0
Reaction conditions: Au25/CeO2 100 mg, benzalacetone 0.1 mmol,
toluene 5 ml, ethanol 5 ml, 0 oC, 4 h.
aCatalysts without thermal treatment before catalytic tests.
bCatalysts with thermal treatment at 260 oC for 2 h.
References
cCatalysts with thermal treatment at 170 oC for 2 h. The activity was
1
2
Jin R C. Nanoscale, 2010, 2: 343
averaged over two independent measurements.
Negishi Y, Nobusada K, Tsukuda T. J Am Chem Soc, 2005, 127:
5261
rated alcohol (Table 2, entry 1). In contrast, uncalcined bi-Au25/
CeO2 catalyst only showed 9% conversion (Table 2, entry 3). It
is worth noting that the recycled i-Au25/CeO2 catalyst only
showed a slight decrease in activity with the same 100% se-
lectivity (not shown).
3
4
Schaaff T G, Whetten R L. J Phys Chem B, 2000, 104: 2630
Lee D, Donkers R L, Wang G, Harper A S, Murray R W. J Am
Chem Soc, 2004, 126: 6193
5
6
7
Zhu M Z, Lanni E, Garg N, Bier M E, Jin R. J Am Chem Soc,
2008, 130: 1138
After the 260 oC thermal pretreatment, i-Au25/CeO2 showed
a significant drop in activity (from 34% to 4%), indicating
aggregation of Au25 clusters into larger particles that were
essentially inactive in hydrogenation reactions, although they
still showed some activity in the selective oxidation reaction
(Table 1). As for the bi-Au25/CeO2 catalyst, a thermal pre-
treatment either at 170 or 260 oC resulted in a drop of catalytic
activity (Table 2, entries 4 and 5). Overall, the bi-Au25 catalyst
gave a very low activity in the hydrogenation reaction.
With respect to the relation between the Au25 structure and
catalytic activity, the exterior low coordination AuG+ atoms in
the i-Au25/CeO catalyst can readily adsorb H2 [31,32], while
the C=O bond of benzalacetone is activated by the elec-
tron-rich Au13 core. The volcano-like surface structure in
i-Au25 was again identified as critical for the hydrogenation. In
contrast, the bi-Au25 structure and surface charge distribution
are not favorable for activating H2 and C=O bonds, hence, the
low activity in the hydrogenation.
Jin R C, Qian H F, Wu Z, Zhu Y, Zhu M Z, Mohanty A, Garg
N. J Phys Chem Lett, 2010, 1: 2903
Liu Y, Tsunoyama H, Akita T, Xie S, Tsukuda T. ACS Catal,
2011, 1: 2
8
9
Zhu M, Qian H F, Jin R C. J Am Chem Soc, 2009, 131: 7220
Zhu M, Qian H F, Jin R C. J Phys Chem Lett, 2010, 1: 1003
10 Tsunoyama H, Tsukuda T. J Am Chem Soc, 2009, 131:
18216
11 Wu Z, Suhan J, Jin R C. J Mater Chem, 2009, 19: 622
12 Qian H F, Sfeir M Y, Jin R C. J Phys Chem C, 2010, 114:
19935
13 Negishi Y, Kurashige W, Niihori Y, Iwasa T, Nobusada K. Phys
Chem Chem Phys, 2010, 12: 6219
14 Fields-Zinna C A, Sardar R, Beasley C A, Murray R W. J Am
Chem Soc, 2009, 131: 16266
15 Angel L A, Majors L T, Dharmaratne A C, Dass A. ACS Nano,
2010, 4: 4691
16 Muhammed M A H, Pradeep T. Small, 2011, 7: 204
17 Qian H F, Zhu M Z, Andersen U N, Jin R C. J Phys Chem A,
2009, 113: 4281
18 Qian H F, Zhu Y, Jin R C. ACS Nano, 2009, 3: 3795
19 Qian H F, Eckenhof W T, Zhu Y, Pintauer T, Jin R C. J Am
Chem Soc, 2010, 132: 8280
3 Conclusions
The catalytic properties of two types of Au25 structures,
namely, icosahedral Au25 nanospheres and biicosahedral Au25
rods were compared. These two types of nanocluster catalysts
have the same number of gold atoms but different structures.
The results clearly demonstrated that the icosahedral
Au25(SC2H4Ph)18 cluster catalyst was superior to the biicosa-
hedral [Au25(PPh3)10(SC2H4Ph)5Cl2]2+ cluster catalyst for both
the selective oxidation of styrene to epoxide and chemoselec-
tive hydrogenation of Į,ȕ-unsaturated benzalacetone to un-
saturated alcohol. The activity of the icosahedral Au25 nano-
20 van Wijngaarden J T, Toikkanen O, Liljeroth P, Quinn B M,
Meijerink A. J Phys Chem C, 2010, 114: 16025
21 Qian H F, Jin R C. Nano Lett, 2009, 9: 4083
22 MacDonald M A, Zhang P, Qian H F, Jin R C. J Phys Chem
Lett 2010, 1: 1821
23 Zhu Y, Qian H F, Zhu M Z, Jin R C. Adv Mater, 2010, 22:
1915
24 Liu Y, Tsunoyama H, Akita T, Tsukuda T. Chem Commun,
2010, 46: 550
25 Corma A, Boronat M, Gonzalez S, Illas F. Chem Commun,

Yan ZHU et al. / Chinese Journal of Catalysis, 2011, 32: 1149–1155
2007, 3371
Chem C, 2009, 113: 17599
26 Turner M, Golovko V B, Vaughan O P H, Abdulkin P, Beren-
guer-Murcia A, Tikhov M S, Johnson B F G, Lambert R M.
Nature, 2008, 454: 981
36 Zhu M, Aikens C M, Hendrich M P, Gupta R, Qian H, Schatz
G C, Jin R C. J Am Chem Soc, 2009, 131: 2490
37 Deng Y H, Cai Y, Sun Z K, Liu J, Liu C, Wei J, Li W, Liu C,
Wang Y, Zhao D Y. J Am Chem Soc, 2010, 132: 8466
38 Kumar S B, Mirajkar S P, Pais G C G, Kumar P, Kumar R. J
Catal, 1995, 156: 163
27 Lopez-Acevedo O, Kacprzak K A, Akola J, Häkkinen H. Nat
Chem, 2010, 2: 329
28 Pei Y, Shao N, Gao Y, Zeng X C. ACS Nano, 2010, 4: 2009
29 Zhu M, Aikens C M, Hollander F J, Schatz G C, Jin R C. J Am
Chem Soc, 2008, 130: 5883
39 Xu R, Wang D S, Zhang J T, Li Y D. Chem-Asian J, 2006, 1:
888
30 Zhu M Z, Eckenhoff W T, Pintauer T, Jin R C. J Phys Chem C,
2008, 112: 14221
40 Liu J H, Wang F, Xu T, Gu Z G. Catal Lett, 2010, 134: 51
41 Patil N S, Uphade B S, Jana P, Bharagava S K, Choudhary V R.
J Catal, 2004, 223: 236
31 Zhu Y, Qian H F, Drake B A, Jin R C. Angew Chem, Int Ed,
2010, 49: 1295
42 Yin D H, Qin L S, Liu J F, Li C Y, Jin Y. J Mol Catal A, 2005,
240: 40
32 Zhu Y, Wu Z K, Gayathri C, Qian H F, Gil R R, Jin R C. J
Catal, 2010, 271: 155
43 Wu Z K, Jin R C. ACS Nano, 2009, 3: 2036
44 Jiang D E, Tiago M L, Luo W D, Dai S. J Am Chem Soc, 2008,
130: 2777
33 Zhu Y, Qian H F, Jin R C. Chem Eur J, 2010, 16: 11455
34 Shichibu Y, Negishi Y, Watanabe T, Chaki N K, Kawaguchi H,
Tsukuda T. J Phys Chem C, 2007, 111: 7845
45 Jiang D E, Nobusada K, Luo W, Whetten R L. ACS Nano, 2009,
3: 2351
35 Qian H F, Zhu M Z, Lanni E, Zhu Y, Bier M E, Jin R C. J Phys