Gold nanoparticles deposited on mesoporous alumina for epoxidation of styrene: Effects of the surface basicity of the supports

Article abstract of DOI:10.1016/j.molcata.2005.06.044

Three kinds of mesoporous alumina (denoted as meso-Al₂O ₃-x (x = a, b, c)) supports with different surface basicity were synthesized by different structure-direct agent and different assembly pathway, respectively. The surface basicity of the meso-Al₂O₃-x and the commercial γ-Al₂O₃ was measured by means of CO₂-TPD and IR spectroscopy. The profiles of CO₂-TPD and IR spectra indicated that the mesoporous alumina owned more abundant surface basic sites than that of γ-Al₂O₃. Au nanoparticles deposited on the above supports via the method of homogeneous deposition precipitation (HDP) using urea as the precipitation agent. The supported Au catalysts were characterized by N₂ adsorption-desorption isotherms, TEM, XPS and testing the catalytic performances in the epoxidation of styrene. The dispersion and average size of the gold particles are dependent of the number of surface basic sites on the supports. Transmission electron microscopy (TEM) observations show a homogeneous distribution of gold particles lower than 4 nm on meso-Al₂O₃-b and meso-Al₂O ₃-c, which owned more basic sites on the surface. XPS spectra reveal that only metallic state of Au is presented on the supports, which is independent of surface properties. The different Al₂O₃ supported Au catalysts were employed as highly active/selective and reusable catalysts for the epoxidation of styrene by anhydrous t-butyl hydroperoxide (TBHP). Au nanoparticles as well as the surface basic sites of the support are responsible for the epoxidation.

Full text of DOI:10.1016/j.molcata.2005.06.044

Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
Gold nanoparticles deposited on mesoporous alumina for epoxidation
of styrene: Effects of the surface basicity of the supports
a,b,∗
a
a
a
a,b
Donghong Yin
, Liangsheng Qin , Jianfu Liu , Chengyong Li , Yong Jin
a
Technology Center, Changsha Cigarette Factory, Hunan, Changsha 410014, China
b
Institute of Fine Catalysis and Synthesis, College of Chem. and Chem. Eng., Hunan Normal University, Changsha 410081, China
Received 8 April 2005; received in revised form 21 June 2005; accepted 23 June 2005
Available online 26 July 2005
Abstract
Three kinds of mesoporous alumina (denoted as meso-Al
different structure-direct agent and different assembly pathway, respectively. The surface basicity of the meso-Al
-Al was measured by means of CO -TPD and IR spectroscopy. The profiles of CO -TPD and IR spectra indicated that the mesoporous
alumina owned more abundant surface basic sites than that of ␥-Al . Au nanoparticles deposited on the above supports via the method
of homogeneous deposition precipitation (HDP) using urea as the precipitation agent. The supported Au catalysts were characterized by
adsorption–desorption isotherms, TEM, XPS and testing the catalytic performances in the epoxidation of styrene. The dispersion and
average size of the gold particles are dependent of the number of surface basic sites on the supports. Transmission electron microscopy (TEM)
observations show a homogeneous distribution of gold particles lower than 4 nm on meso-Al -b and meso-Al -c, which owned more
basic sites on the surface. XPS spectra reveal that only metallic state of Au is presented on the supports, which is independent of surface
properties. The different Al supported Au catalysts were employed as highly active/selective and reusable catalysts for the epoxidation of
2
O
3
-x (x = a, b, c)) supports with different surface basicity were synthesized by
2
O
3
-x and the commercial
␥
2
O
3
2
2
2
O
3
N
2
2
O
3
2 3
O
2
O
3
styrene by anhydrous t-butyl hydroperoxide (TBHP). Au nanoparticles as well as the surface basic sites of the support are responsible for the
epoxidation.
©
2005 Elsevier B.V. All rights reserved.
Keywords: Mesoporous alumina; Surface basicity; Gold; Styrene; Epoxidation
1
. Introduction
HAuCl4 is used as the typical metal precursor. The chloroau-
ric anion hydrolyzes in solution to form [Au(OH)nCl4−n]
−
Supported gold catalysts are very active in catalyzing
(n = 1–4) which the interaction with the support proceeds
by an anionic adsorption to precipitate gold as Au(OH)3
[12–14].The surface of the support acts as a nucleating agent,
and therefore the properties of the support material play a pre-
dominant role in the formation and stabilization of particle
size of gold, which finally influences the catalytic activity
of the resulting catalyst [10,15–19]. As a result, the iso-
electric points of the supports with IEP ≈ 7, such as TiO2
(IEP = 6), CeO2 (IEP = 6.75) [20], ZrO2 (IEP = 6.7) [21] and
Fe2O3 (IEP = 6.5–6.9) [22] produce highly active catalysts.
However, the acidic support, such as SiO2 (IEP = 1–2), is
found to be an unsuitable support material for gold catalysts
prepared via deposition–precipitation due to the highly nega-
many types of reactions like the low temperature oxidation of
CO [1], oxidation of ethylene glycol [2] and d-glucose [3], the
epoxidationofpropene[4,5]andstyrene[6,7], hydrogenation
of crotonaldehyde [8] and the reduction of nitrogen oxides
[9], etc. It is known that the activity of gold catalysts is related
to the Au particle size, Au oxidation state and the nature of
the support material. Recently, versatile methods have been
reported to prepare the gold nanoparticles with high disper-
sion on different supports. Among the available methods,
the method of deposition–precipitation is one of the most
successfully used [1,10,11]. In such a preparation method,
∗
tively charged surface of SiO does not allow the adsorption
of [Au(OH)nCl4−n] species onto the support surface [15].
Corresponding author. Tel.: +86 731 8872576; fax: +86 731 8872531.
E-mail address: yindh@hunnu.edu.cn (D. Yin).
2
−
1
381-1169/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2005.06.044

D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
41
Al2O3 (IEP = 8–9) with the amphoteric character used as a
support to deposit gold has been reported as being either
very active, or as expressing a certain degree of activity or
evenbeinginactiveduringCOoxidationatroomtemperature,
Method A [36]: a 6.078 g of triblock poly(ethylene
glycol)-poly(propylene glycol)-poly(ethylene glycol)
(EO20PO70EO20 (P123), BASF) together with 0.060 g
LaCl3 was dissolved in 25 mL sec-butanol, and then 5.4 mL
aluminum sec-butoxide (Aldrich) was added with stirring.
After stirring for 1 h at ambient temperature, a dilute solution
of water in sec-butanol (1.14 mL H2O/10 mL sec-butanol)
was added with dropwise. The molar composition of the
[23,24]. For example, significant differences in the catalytic
activity of CO oxidation were obtained over the gold catalysts
prepared on ␦-Al2O3 from two different supplies at the same
conditions [15]. It is suggested that the physical and chemical
state of the different alumina supports could influence the par-
ticles size of gold, and therefore the catalytic activity of the
resulting catalysts. However, no systematic study concern-
ing the surface basic sites of the supports on the formation
of nanometer gold particles and on the catalytic activity has
been published yet.
3
+
s
final reaction mixture was 0.01 La :1.0 Al(Bu O)3:0.10
s
P123:3.0 H2O:15.5 Bu OH. The mixture was stirred at
45 C for another 48 h. The product obtained was filtered,
dried in air and calcined at 500 C for 6 h in flowing air. The
mesoporous alumina prepared via a neutral N I assembly
using P123 as a structure-directing agent is designed as
meso-Al2O3-a.
◦
◦
0
0
The selective and efficient epoxidation of olefins to obtain
epoxides has become very important, since the epoxides are
versatile intermediates in organic synthesis of fine chemi-
cals and pharmaceuticals. Epoxidation has traditionally been
carried out using peracids [25], and the procedures are very
costly and usually produce huge amounts of carboxylic acids
as side products. Hence, it is highly desirable to replace the
conventional process with an environmentally benign pro-
cedure. Heterogeneous catalysts, such as Ti-silicalite-1 [26],
Ti-␤ zetolite [27], Ti-SiO2 [28], TS-2 [29] and ␥-Al2O3 [30],
are proposed as alternative catalysts because of easy sepa-
ration from the reaction mixture and reusability. However,
these catalysts showed either poor activity or low selectivity
for the epoxides. Recently, Choudhary and co-workers have
reported liquid-phase epoxidation of styrene by anhydrous t-
butyl hydroperoxide (TBHP) over gold supported on various
supports. [6,7]. They found that the activity and the epoxide
selectivity were depended on Au particles size and the nature
of supports.
Organized mesoporous alumina represents a very interest-
ing molecular sieve exhibiting a narrow pore size distribution
with worm-like structure, higher surface areas and abundant
of surface hydroxyl groups compared to conventional alu-
mina [31]. Recently, there are some reports regarding the
formation of Au nanoparticle in the meso-silica and meso-
titania [19,32–35]. However, using mesoporous alumina as a
support for deposition of nanometer Au catalysts has never
been reported. The objective of this work is to understand the
effect of surface basicity of mesoporous alumina synthesized
by different structure-direct agent on the distribution and
on the size of Au nanoparticles, aiming to improve catalyst
activity for liquid-phase epoxidation of styrene by anhydrous
TBHP.
Method B [37]: (1) a solution containing 0.8 g NaOH in
4 mL H2O was added to 80 mL triethanolamine and heated
at 120 C for 5 min to evaporate the water. Over this solution,
◦
21.8 mL aluminum sec-butoxide was added dropwise with
◦
stirring. The resulting solution was then heated at 150 C
for 10 min (solution I). (2) 14.56 g cetyltrimethyl ammonium
bromide (CTABr, SCR) was dissolved in 240 mL H2O at
◦
60 C (solution II). (3) Solution I was slowly added to solu-
◦
tion II with vigorous stirring at 60 C, and the mixture was
allowed to age for another 72 h. The precipitate was filtered,
◦
◦
washed with ethanol, dried at 30 C and calcined at 500 C
for 5 h in flowing air. The mesoporous alumina prepared via
+
−
an electrostatic S T pathway using CTABr as a structure-
directing agent is designed as meso-Al2O3-b.
Method C: a solution of 4.60 g aluminum nitrate
(Al(NO3)3·9H20, Fluka) dissolved in 20 mL H2O was added
to the solution of 1.50 g chitosan (DAC degree > 91.3%, vis-
−
1
cosity 75 mPa s , Yuhuan Ocean Biochem. Co. Ltd. China)
in 50 mL of CH3COOH solution (5%, v/v) under stirring for
1 h. And then, the mixture was added dropwise to a 800 mL
of aqueous ammonia solution (50%, v/v) with vigorous stir-
ring for 1 h. The precipitate was filtered and washed with
◦
◦
water, dried at 30 C for 72 h and calcined at 550 C for
1.5 h in flowing air. The obtained mesoporous alumina by
the structure-directing agent of chitosan is designed as meso-
Al2O3-c.
2.2. Preparation of catalysts
meso-Al2O3-a, meso-Al2O3-b, meso-Al2O3-c and the
commercial ␥-Al2O3 (provided by Changling Oil Refine Cor-
◦
poration, calcined at 550 C for 6 h before use) were used
as the supports for deposition of gold, respectively. The
supported gold catalysts were prepared by the method of
homogeneous deposition precipitation (HDP) using urea as
a precipitation reagent. In a typical procedure, 1 g of alumina
support was added to 100 mL of an aqueous solution contain-
2
. Experimental
2
.1. Preparation of mesoporous alumina
−
3
ing HAuCl4 (9.71 × 10 M, 10.5 mL). The initial pH was
The different mesoporous alumina was prepared by the
about 4. Excess urea (Acros, p.a.) was added. The suspen-
◦
following three kinds of method with different structure-
directing agent.
sion was vigorously stirred and heated to 70–75 C in order
todecomposetheurea. Thistemperaturewasmaintaineduntil

4
2
D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
the pH of the solution reached gradually from 4 to 7.8. The
suspension was then cooled down and filtered, washed with
hot deionized water several times until no Cl was detected
determined by examination of several micrographs. The aver-
age particle diameter ds was calculated using the following
ꢀ
ꢀ
−
3
i i
2
formula: ds =
nid / nid , where ni is the number of
by silver nitrate solution (0.1 M) in the filtrate. The result-
particles of diameter di. At least 200 particles were chosen
to determine the mean diameter of gold particles. X-ray pho-
toelectron spectroscopy (XPS) analysis was performed using
an Escalab MK2 spectrometer. The spectra were acquired
using a monochromatic Mg K␣ radiation source and at the
◦
ing Au/Al2O3 was dried overnight at 80 C in air, followed
◦
by calcined in air at 330 C for 1 h under a heating rate of
◦
−1
◦
5
4
C min and then calcined at 300 C for 0.5 h in a flow of
% (v/v) H2/He.
−
7
pressure below 7 × 10 Pa. The binding energy scale was
2
.3. Catalyst characterization
referred to the C 1s line at 284.6 eV.
The BET surface area and textural structure were mea-
2.4. Catalytic activity
sured by nitrogen adsorption at liquid nitrogen temperature
in a Micomeritics ASAP-2400 apparatus. Before analysis, the
samples were degassed 4 h at 250 C in vacuum. The surface
areas were calculated using the BET method. The isoelectric
point (IEP) of alumina suspension was determined using a
Brookhaven Zeta 90Plus Instruments. The ground alumina
samples (0.01 wt%) were dispersed in an electrolyte solution
containing 1 mM KCl as supporting electrolyte. The pH value
was adjusted carefully by adding HCl (0.01 M) or NaOH
The styrene epoxidation reactions over the supported
Au catalysts were carried out at atmospheric pressure by
contacting 0.1 g catalyst with 1.2 mL (10 mmol) styrene,
2.8 mL (15.4 mmol, Aldrich) anhydrous t-butyl hydroper-
oxide (5.5 M in decane, Fluka) and 5 mL benzene in
◦
3
a magnetically stirred glass reactor (capacity: 15 cm )
◦
at ca. 82–83 C for 12 h. The reaction was monitored
by gas chromatography with flame ionization detector
(Agilent Technologies 6890N, DB-35 capillary column,
30 mm × 0.25 mm), Helium as a carrier gas, injector tem-
(0.01 M). Zeta potential was measured following the proce-
dures described in the instrument manual. The pH of the sus-
pension was measured at the end of agitation to an accuracy
of ± 0.02 with a REX Model PHS-3C pH meter. The isoelec-
tric points of the samples obtained by reading the intercept of
the trend lines with the horizontal line of 0 mV are given in
the plot of pH–zeta potential. The temperature-programmed
desorption of CO2 (CO2-TPD) was measured in a homemade
◦
◦
perature 250 C, the detector temperature 250 C, oven tem-
◦
◦
◦
perature program 110 C (2 min) → 15 C min → 130 C
◦
−1
◦
(0.5 min) → 1 C min → 140 C (2 min), using dodecane
as an internal standard substance. The products were iden-
tified by GC–MS using an HP5790 Series mass selective
detector. The used catalyst was washed with benzene, and
then it was reused for the subsequent epoxidation of styrene
for testing the catalyst reusability.
TM
equipment with mass spectrum (Balzers Omnistar 200) as
◦
a detector. The CO2 adsorption temperature was 30 C, and
◦
◦
−1
.
the temperature was ramped 500 C at a rate of 10 C min
FTIR analysis was carried out using a commercial instrument
Thermo Nicolet, Avatar 370) equipped with a controlled-
atmosphere chamber. The IR spectrum was measured in the
(
3. Results and discussion
−
1
reflection mode at 32 scans with a resolution of 2 cm and
corrected for the background spectrum. Pretreatment of the
3.1. Catalyst characterization
−
2
self-supportingwaferwasperformedbyevacuation(10 Pa)
at 573 K for 1 h. Bright-field images of transmission electron
microscopy (TEM) were carried out in a JEOL JEM2010
microscope operating at 200 kV. The samples were dispersed
in ethanol by sonication and dropped on a copper grid coated
with a carbon film. Micrographs at various positions were
taken and the size of the preferentially present particles was
3.1.1. Textural properties of the supports
The textural properties of the mesoporous Al2O3-x
(x = a, b, c) and ␥-Al2O3 were characterized by the N2
adsorption–desorption method. The specific surface area,
pore volume and the mean pore diameter of the samples
are summarized in Table 1 for comparison. Obviously, the
meoporous Al2O3 have a higher specific surface area and
Table 1
Textural properties of the supports and the supported gold catalystsa
2
−1
3
−1
Sample
BET surface area (m g
)
Pore volume (cm g
)
Mean pore diameter (nm)
IEP (pH)
meso-Al2O3-a
meso-Al2O3-b
meso-Al2O3-c
325
302
271
108
259
355
310
92
0.93
0.78
0.50
0.43
0.49
0.59
0.48
0.46
7.5
7.2
4.9
29
4.9
4.3
4.5
28
8.4
9.4
9.2
8.4
–
–
–
–
␥-Al2O3
Au/meso-Al2O3-a
Au/meso-Al2O3-b
Au/meso-Al2O3-c
Au/␥-Al2O3
a
The Au loading over the support is 2.0 wt.%.

D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
43
larger pore volume than that of the commercial ␥-Al2O3.
Furthermore, the specific surface area and pore volume of the
mesoporous Al2O3 depend on the different structure-direct
agent during the synthesized procedure. meso-Al2O3-a and
0
0
meso-Al2O3-b obtained via the N I assembly pathway with
+
−
P123 as the structure-direct agent and the S T assembly
pathway with CTABr as the structure-direct agent have a
higherspecificsurfacearea, lagerporevolumeandwiderpore
diameter than that of meso-Al2O3-c prepared with chitosan
as the structure-direct agent. All the samples of mesoporous
Al2O3 have a narrow pore size distribution, but the pore size
distribution of the ␥-Al2O3 is rather broad due to the aggre-
gate of nanosized nonporous particles.
The specific surface area, the pore volume and mean pore
diameter decrease after Au particles deposited on the sup-
ports of meso-Al2O3-a and ␥-Al2O3, which are in agreement
with previously publish when gold particles are dispersed
on the support of SBA-15 [32]. Interestingly, the surface
areas increase as gold particles deposited on the supports of
meso-Al2O3-b and meso-Al2O3-c. Somorjai and co-workers
obtained the similar results when the gold nanoparticles
incorporated into the silicate materials of MCM-41 and
MCM-48 [33]. This is probably caused by the fact that
more gold particles are small enough and arrange on the
pore surface of the supports as spherical and half-spherical
forms to accommodate more nitrogen molecules. On the
other hand, the pore volumes and pore diameters of the sam-
ples decreased. Similar observation is obtained when organic
groups are anchored onto the pore surface of MSU [38]. The
difference of the surface area should be resulted from the
different surface properties of the supports which leading
to different Au particles size when Au species deposited on
them.
Fig. 1. CO2-TPD profile of meso-Al2O3-b (a), meso-Al2O3-c (b), meso-
Al2O3-a (c) and ␥-Al2O3 (d).
orption amount of CO2, the number of surface basic sites on
the mesoporous alumina follows the order of meso-Al2O3-
b > meso-Al2O3-c > meso-Al2O3-a > ␥-Al2O3. The compari-
son of isoelectric points value of all the supports listed in
Table 1 also indicated that the surface basic properties of
meso-Al2O3-b and meso-Al2O3-c are stronger than that of
meso-Al2O3-a and ␥-Al2O3.
Fig. 2 shows FTIR spectra in the region of the OH stretch-
ing mode of the different alumina samples after evacuation
under vacuum at 573 K. An intense broad band centered
−
1
at 3738 cm accompanied by shoulders at the region of
−
1
3
681–3618 cm appears on the samples of mesoporous alu-
−
1
mina. The OH band of 3738 cm corresponding to the over-
lapping of bands is assigned to OH species bridging two and
three Al atoms in octahedral sites as well as two Al atoms
in 4- and 6-fold coordination [39], which is related to the
surface basic sites. Apparently, the hydroxyl group vibration
of mesoporous alumina differs from that of ␥-Al2O3. Only
a weak band centered around 3738 cm is observed on the
commercial␥-Al2O3. TherelativeintensitiesoftheOHgroup
vibrations of mesoporous alumina are evidently higher than
those of ␥-Al2O3, indicating that the number of surface basic
3
.1.2. Surface basicity measurement
−
1
From the results of the textural properties of the sup-
ports and the supported Au catalysts, we can conclude that
the mesoporous Al2O3 and ␥-Al2O3 should have different
surface properties. Therefore, we use the means of CO2-
TPD to measure their surface basic properties. Fig. 1 shows
only one asymmetric broad peak on the profiles of the four
kinds of Al2O3. The maximum temperature of CO2 des-
◦
orption peaks are close each other and near 120 C. Fur-
thermore, the number of surface basic sites on the four
kinds of Al2O3 corresponding to the areas of CO2 desorp-
tion peaks are different. The desorption amount of CO2
−
1
−1
are 0.0369 mmol g for meso-Al2O3-a, 0.0606 mmol g
−
1
for meso-Al2O3-b, 0.0503 mmol g for meso-Al2O3-c and
.0130 mmol g for ␥-Al2O3, respectively. Apparently, the
−
1
0
mesoporous alumina with wormhole structure possesses
more abundant surface basic sites owing to their high sur-
face area. However, among the three kinds of mesoporous
Al2O3, the support of meso-Al2O3-a with the highest sur-
face area shows the least number of surface basic sites. This
should be attributed to the different structure-directing agent
used during their synthesis process. By comparing the des-
Fig. 2. IR spectra of meso-Al2O3-b (a), meso-Al2O3-a (b), meso-Al2O3-c
(c) and ␥-Al2O3 (d).

4
4
D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
sites on the mesoporous alumina samples is greater than that
on ␥-Al2O3.
and distribution on the different Al2O3 supports are mainly
affected by their surface basicity.
3
.1.3. TEM observation
3.1.4. XPS characterization
The morphology and particle size of the gold on the dif-
Although the catalytic activity of gold catalysts in the low
temperature CO oxidation has been intensively studied dur-
ing the last decade, the nature of the active species is still
discussed controversially. It has been suggested that the role
of the metal oxide is the stabilization of the gold nanoparti-
cles and that the reaction takes place on the gold surface [1].
Some authors proposed that the reaction takes place at the
gold/metal oxide interface and that the metal oxide could act
as a source of oxygen [16]. Also the electronic structure of
gold in active catalysts is unclear, but most of other authors
suggest metallic gold to be the active species. The XPS spec-
tra of the Au 4f (region Au 4f7/2 and Au 4f5/2) are employed
to investigate the state of gold on the different supports. The
results indicated that only a binding energy of 83.8 eV for
the Au 4f7/2 electron and 87.5 eV for the Au 4f5/2 electron
(∆ = 3.7 eV) with the change of supports for the fresh sam-
ples. ThebindingenergyindicatesexclusivelyAu( -particles
without any trace of higher oxidized species on the four kinds
of Al2O3 surface. The surface basicity of the support has no
effect on the state of gold if the Au catalysts prepared at the
same conditions.
ferent Al2O3 supports have been determined by TEM (see
Fig. 3). The gold particles are seen as dark contrasts on the
surface of the supports. TEM micrographs show that the dif-
ference of the average size of the metallic gold particles with
the change of supports. A homogeneous distribution of gold
particles with the size of lower than 4 nm is observed all over
the surface of meso-Al2O3-b and meso-Al2O3-c, which pos-
sessed more abundant surface basic sites and higher surface
areas (Fig. 3b and c). The Au particles with the diameters
less than 2 nm are also seen on the meso-Al2O3-b support
as well as those with diameters of 3–4 nm. The average size
of the gold particles on the support of meso-Al2O3-b and
meso-Al2O3-c is 3.2 and 3.6 nm, respectively. However, the
distributions of gold particles on supports of meso-Al2O3-
a and ␥-Al2O3 are not so uniform as that on meso-Al2O3-b
even that the supports of meso-Al2O3-a has the largest surface
area (Fig. 3a and d). The average size of the gold particles on
the support of meso-Al2O3-a and ␥-Al2O3 is 4.4 and 4.9 nm,
respectively. This observation demonstrates that the differ-
ences in the final dispersion and average size of the metallic
gold particles deposited on the surface of the four supports are
mainly a consequence of their different surface basic proper-
ties, although a positive influence of a higher surface area of
the supports cannot be ruled out. The more the number of the
surface basic sites has, the better of the gold particles distri-
bution and the smaller of the mean diameter are obtained.
It is known that with increasing the basicity of supports,
the pH at the support solution interface is increased, result-
ing in more precipitation of hydroxylated gold species on
0)
3.2. Catalytic properties
3.2.1. Effect of supports
Table 2 summarizes the results of epoxidation of styrene
using anhydrous TBHP as an oxidizing agent over Au cata-
lysts supported on different Al2O3. The effect of supports on
the conversion of styrene and the selectivity of the styrene
oxide was investigated. As can be seen from this table, the
epoxidation of styrene is very slow in the absence of catalyst.
However, the different support of Al2O3 used as a catalyst
shows the different catalytic activity, which is different from
the reports of Choudhary and co-workers [7]. They did not
find apparently conversion of styrene over Al2O3 support.
Our results suggest the surface basic sites of the supports
facilitating to the epoxidation should not be ruled out com-
pletely. Furthermore, the catalytic activity increased with
increasing the surface basicity of the supports. Kirm et al.
also observed that the basic sites of hydrotalcites with dif-
ferent atomic ratio of Mg/Al play an important role in the
epoxidation of styrene [40], and also the catalytic activity
increased with the surface Br o¨ sted basic sites.
For the Au/meso-Al2O3-a, Au/meso-Al2O3-b, Au/meso-
Al2O3-c and Au/␥-Al2O3 catalysts, the difference of selectiv-
ity to styrene oxide is not significant. However, the catalytic
activity follows the order of Au/meso-Al2O3-b > Au/meso-
Al2O3-c > Au/meso-Al2O3-a > Au/␥-Al2O3, which is coin-
cident with the Au particle size and the surface basic sites on
the supports. The catalyst of Au/meso-Al2O3-b shows 84.3%
of conversion, which is higher than that the results obtained
over the catalyst of Au/Al2O3 in the reference [7]. It should be
−
the support surface due to an increase of OH ion concen-
−
tration at the interface [15]. The OH ions on the surface
of support can exchange the chloride ions in the species of
−
−
[
Au(OH)nCl4−n] (n = 1–4) to decrease the residual Cl and
−
to form [AuCl(OH)3] precursor, resulting in small and high
dispersion of gold particles on the supports. Furthermore, in
the preparation of Au/Al2O3 catalysts by HDP method, the
gradual and homogenous addition of hydroxide ions through-
out the solution avoids local increasing in pH value in the
solution, which is always below the IEP value of the supports.
It can be inferred that the surface-charged alumina support
is always positive, which should be in favor of attracting
−
negatively charged species of [Au(OH)nCl4−n] (n = 1–4)
−
in solution, especially for the complex of [AuCl(OH)3] .
The obtained average gold particle size should be smaller at
the surface of the support with higher IEP value. Therefore,
the supports of meso-Al2O3-b and meso-Al2O3-c with more
number of the surface basic sites are advantageous for the for-
mation and stabilization of small gold particles. Combining
with the results of surface areas and IEP value of the supports
obtained in Table 1, as well as the determination of surface
basicity in Fig. 1, it is proposed that the gold particles size

D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
45
Fig. 3. TEM micrographs and size histogram of Au particles supported on meso-Al2O3-a (a), meso-Al2O3-b (b), meso-Al2O3-c (c), and ␥-Al2O3 (d).

4
6
D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
Table 2
Performance of Au catalyst with different supports in styrene epoxidationa
Catalysts
Styrene conversion (%)
Selectivity (%)
Styrene oxide
Benzaldehyde
Phenyl acetaldehyde
Othersb
No
8.3
38.6
50.3
43.8
36.0
71.1
84.3
76.1
64.5
◦
75.9
65.0
66.9
65.8
69.9
65.6
69.0
67.4
67.8
14.6
18.0
21.0
20.1
16.4
24.5
23.0
22.3
25.6
4.7
5.2
5.2
2.0
4.7
6.4
3.6
5.5
3.3
4.8
11.8
6.9
12.1
9.0
3.5
4.4
4.7
3.3
meso-Al2O3-a
meso-Al2O3-b
meso-Al2O3-c
␥-Al2O3
Au/meso-Al2O3-a
Au/meso-Al2O3-b
Au/meso-Al2O3-c
Au/␥-Al2O3
a
The reaction was performed at 82 C for 12 h over 0.10 g of the supported catalyst with 2.0% of Au loading.
Others are benzoic acid and phenyl acetic acid.
b
Table 3
a
The effect of Au loading on the epoxidation of styrene
Gold loading (%)
Styrene conversion (%)
Selectivity (%)
Styrene oxide
Benzaldehyde
Phenyl acetaldehyde
Othersb
1
2
3
5
.0
.0
.5
.0
a
68.5
84.3
100
100
◦
74.7
69.0
60.3
58.8
19.3
23.0
30.1
28.0
3.6
3.6
3.6
4.0
2.4
4.4
6.0
9.2
The reaction was performed at 82 C for 12 h over 0.10 g of the supported catalyst.
Same as in Table 2.
b
attributed to more homogeneous distribution and the smaller
average Au particle size (3.1 nm) in our experimental. It is
known that the epoxidation activity of the supported gold cat-
alysts is mainly attributed to the low coordinated Au atoms
at the corners and edges of nanosize gold particles even that
the catalytic reaction mechanism is not yet clear [6,7]. There-
fore, the small Au particles are presented homogeneously on
the surface of the support should exhibit the high activity for
the epoxidation of styrene. The smaller size of Au particles
is, the higher catalytic activity can be obtained. Therefore,
both the Au nanoparticles and the surface basic sites of the
supports can catalyze the epoxidation of styrene, but the Au
particles dispersed homogeneously onto support with small
particle size should play major role for the epoxidation.
workers reported that the activity of the Au/TiO2 catalyst for
CO oxidation is mainly affected by the amount of moisture
adsorbed on the catalyst [41]. They found that the suppres-
sion of the reaction by the excess amount of water could be
explained by blocking of the active sites. A series of exper-
iments is carried out using Au/meso-Al2O3-b as a catalyst
to examine the effect of water content on the catalytic prop-
erties. The results are shown in Fig. 4. It can be seen that
the conversion of styrene and selectivity of styrene oxide
remain almost unchanged with addition of water less than
2%. The results suggest that the low concentration of water
3
.2.2. Effect of Au loading
Table 3 summarizes the results of epoxidation of styrene
over the catalyst of Au/meso-Al2O3-b at various Au loading.
It is expected that the conversion of styrene increased with an
increase of the Au loading. However, the selectivity to styrene
epoxide decreased due to the increase of benzaldehyde selec-
tivity. When Au loading is up to 3.5 wt%, the conversion of
styrene is up to 100%, but the selectivity of benzaldehyde,
benzoic acid and phenyl acetic acid increase. These results
indicate that Au nanoparticles on the support of Al2O3 are
the main active sites for the epoxidation of styrene.
3
.2.3. Effect of water
It is significant that the activity of Au nanoparticles is
sensitive to the water content on the catalyst. Haruta and co-
Fig. 4. Effect of amounts of water on conversion of styrene (a) and selectivity
of styrene oxide (b).

D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
47
Fig. 5. TEM micrographs and size histogram of Au particles in the catalyst of Au/meso-Al2O3-b separated from reaction by addition of 20% water.
in the reaction has not significant effect on the catalytic activ-
ity and selectivity. With an increase of water content from 5
to 20%, the catalytic activity decreases (from 84 to 72%).
In this case, the color of the catalyst turned from wine to
pink, which is suggested that high content of water result
in aggregation of Au particles on the support. It was con-
firmed by the TEM image of the Au/Al2O3-b catalyst (Fig. 5),
where the catalyst was separated from the reaction system
3.2.4. Reuse of the catalyst
The used catalyst of Au/meso-Al2O3-b was washed with
benzene, and then reused for the subsequent epoxidation of
styrene. The above procedure was repeated for eight times
and the results are shown in Fig. 6. It was observed that there
was no noticeable decrease in activity even up to five sub-
sequent runs. Only 3% catalytic activity (from 84 to 81%)
is lost after the catalyst reused nine times. Interestingly, the
trend of the selectivity of styrene oxide increases slightly.
These results indicate that the catalyst is relatively stable in
the epoxidation of styrene using anhydrous TBHP as an oxi-
dant.
◦
with addition of 20% water and then dried at 80 C in vac-
uum. It is observed that the gold particles become large with
the average gold particle size of 8.7 nm. Only 70% of con-
version was obtained if the dried catalyst was used again.
The lost activity is attributed to the adsorption of water on
the catalyst leading the unstable small Au nanoparticles to
aggregate together. When the content of water exceeds 25%,
the change of conversion of styrene and the selectivity of
styrene oxide are not apparent. In this case, the remained
activity of the catalyst is mainly from the surface basic
sites.
4. Conclusions
The surface basic sites on the support of mesoporous
Al2O3 influence the final dispersion and average size
of the obtained gold particles during the homogeneous
deposition–precipitation process. The more number of the
surface basic sites are advantageous for to form uniform and
smaller size of Au nanoparticles. The supported nanometer
Au catalyst shows high catalytic activity and selectivity for
the liquid-phase epoxidation of styrene by anhydrous t-butyl
hydroperoxide. Smaller size of Au particles and surface basic
sites on the support are responsible for the catalytic reac-
tion. Due to the possibility to separate easily and reuse in the
reaction, the gold nanoparticles supported on meso-Al2O3-b
should be a promising heterogeneous catalyst for the epoxi-
dation of alkylene.
Acknowledgements
The project was sponsored by the Scientific Research
Foundation for the Returned Overseas Chinese Scholars,
State Education Ministry [2003-406].
Fig. 6. Effect of Au/meso-Al2O3-b reuse on the conversion of styrene (a)
and selectivity of styrene oxide (b).

4
8
D. Yin et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 40–48
[21] A. Knell, P. Barnickel, A. Baiker, A. Wokaun, J. Catal. 137 (1992)
References
306.
[
22] M. Haruta, N. Yamada, T. Kobayashi, S. Ijima, J. Catal. 115 (1989)
[
1] M. Haruta, S. Tsubota, T. Kobayashi, H. Kageyama, M.J. Genet, B.
Delmon, J. Catal. 144 (1993) 175.
301.
[
[
23] R.J.H. Grisel, B.E. Nieuwenhuys, J. Catal. 199 (2001) 48.
24] S.J. Lee, A. Gavriilidis, J. Catal. 206 (2002) 305.
[2] H. Berndt, I. Pitsch, S. Evert, K. Struve, M.-M. Pohl, J. Radnik, A.
Martin, Appl. Catal. A 244 (2003) 169.
¨
[25] D. Swen, in: R. Adams (Ed.), Organic Reactions, Wiley, New York,
953, p. 378.
[
[
[
3] Y. Onal, S. Schimpf, P. Claus, J. Catal. 223 (2004) 122.
1
4] T. Hayashi, K. Tanaka, M. Haruta, J. Catal. 178 (1998) 566.
5] E.E. Stangland, K.B. Stavens, R.P. Andres, W.N. Delgass, Stud. Surf.
Sci. Catal. 130 (2000) 827.
[
[
26] M. Taramasso, G. Perego, B. Notari, US Patent, 4,410,501 (1983).
27] A. Camblor, A. Corma, A. Mart ˇı y´ nez, J. P e´ rez-Pariente, J. Chem.
Soc., Chem. Commun. (1992) 583.
[
[
[
[
6] N.S. Patil, B.S. Uphade, P. Jana, S.K. Bharagava, V.R. Choudhary,
[
[
[
[
[
[
28] Q. Yang, S. Wang, J. Lu, G. Xiong, Z. Feng, X. Xin, C. Li, Appl.
J. Catal. 223 (2004) 236.
Catal. A 194–195 (2000) 507.
29] Z. Fu, D. Yin, D. Yin, Q. Li, L. Zhang, Y. Zhang, Microporous
Mesoporous Mater. 29 (1999) 351.
7] N.S. Patil, R. Jha, B.S. Uphade, S.K. Bhargava, V.R. Choudhary,
Appl. Catal. A 275 (2004) 87.
8] R. Zanella, C. Louis, S. Giorgio, R. Touroude, J. Catal. 223 (2004)
28.
9] T. Salama, R. Ohnishi, T. Shido, M. Ichikawa, Chem. Commun.
1997) 105.
30] V.R. Choudhary, N.S. Patil, S.K. Bhargava, Catal. Lett. 89 (2003)
3
5
5.
31] S.A. Bagshaw, T.J. Pinnavaia, Angew. Chem. Int. Ed. Engl. 35 (1996)
102.
32] C.M. Yang, M. Kalwei, F. Sch u¨ th, K.J. Chao, Appl. Catal. A 254
2003) 289.
(
1
[
[
10] G.C. Bond, D.T. Thompson, Catal. Rev. Sci. Eng. 41 (1999) 319.
11] S. Tsubota, D.A. Cunningham, Y. Bando, M. Haruta, in: G. Poncelet,
et al. (Eds.), Preparation of Catalysts, vol. VI, Elsevier, Amsterdam,
(
33] Z. Ko’nya, V.F. Puntes, I. Kiricsi, J. Zhu, J.W. Ager, M.K. Ko,
H. Frei, P. Alivisatos, G.A. Somorjai, Chem. Mater. 15 (2003)
1995, pp. 227–235.
[
[
[
12] S. Ivanova, C. Petit, V. Pitchon, Appl. Catal. A 267 (2004) 191.
13] H.H. Kung, M.C. Kung, C.K. Costello, J. Catal. 216 (2003) 425.
14] M.A. Centeno, I. Carrizosa, A. Jos e´ , Odriozola, Appl. Catal. A 246
2003) 365.
1242.
´
[34] H. Araki, A. Fukuoka, Y. Sakamoto, S. Inagaki, N. Sugimoto, Y.
Fukushima, M. Ichikawa, J. Mol. Catal. A 199 (2003) 95.
(
[
[
[
35] J.J. Pietron, R.M. Stroud, D.R. Rolison, Nano. Lett. 2 (2002) 545.
36] W. Zhang, J. Pinnavaia, Chem. Commun. 11 (1998) 1185.
37] S. Cabrara, J.E. Haskouri, J. Alamo, A. Beltran, D. Beltran, S. Men-
dioroz, M.D. Marcos, P. Amoros, Adv. Mater. 11 (1999) 379.
38] N. Yu, Y. Gong, D. Wu, Y. Sun, Q. Luo, W. Liu, F. Deng, Microp-
orous Mesoporous Mater. 72 (2004) 25.
[
[
15] A. Wolf, F. Sch u¨ th, Appl. Catal. A 226 (2002) 1.
16] M.M. Schubert, S. Hackenberg, A.C. van Veen, M. Muhler, V. Plzak,
R.J. Behm, J. Catal. 197 (2001) 113.
[
[
[
[
17] J.-D. Grunwaldt, M. Maciejewski, O.S. Becker, P. Fabrizioli, A.
Baiker, J. Catal. 186 (1999) 458.
18] K. Mallick, M.J. Witcomb, M.S. Scurrell, J. Mol. Catal. A 215
[
[
[
[
39] V. Gonz a´ lez-Pe n˜ a, I. D ´ı az, C. M a´ rquez-Alvarez, E. Sastre, J. P e´ rez-
Pariente, Microporous Mesoporous Mater. 44–45 (2001) 203.
40] I. Kirm, F. Medina, X. Rodriguez, Y. Cesteros, P. Salagre, J. Sueiras,
Appl. Catal. A 272 (2004) 175.
(2004) 103.
19] W. Yan, B. Chen, S.M. Mahurin, E.W. Hagaman, S. Dai, S.H. Over-
bury, J. Phys. Chem. B 108 (2004) 2793.
20] D. Andreeva, V. Idakiev, T. Tabakova, L. Ilieva, P. Falaras, A. Bourli-
nos, A. Travlos, Catal. Today 72 (2002) 51.
41] M. Date, M. Haruta, J. Catal. 201 (2001) 221.