Switching of reactions between hydrogenation and epoxidation of propene over Au/Ti-based oxides in the presence of H2 and O2

Article abstract of DOI:10.1016/j.jcat.2011.03.028

In a H₂ and O₂ mixture, the reaction of propene switches between hydrogenation and epoxidation over Au/Ti-based oxides. Reaction pathways are strongly dependent on the size of Au particles and on the presence of alkalis. The hydrogenation prevails over Au clusters smaller than 2.0 nm (in the absence of alkalis) and over Au nanoparticles larger than 5.0 nm. Alkali contamination can switch hydrogenation to epoxidation over Au clusters and promote epoxidation over Au nanoparticles with diameters of 2.0-5.0 nm. It was also confirmed that the hydrogenation of propene is enhanced by the addition of a small amount of O₂. Switching of product by alkalis and enhancement of hydrogenation by oxygen are interpreted by parallel and competitive reaction pathways.

Full text of DOI:10.1016/j.jcat.2011.03.028

Journal of Catalysis 281 (2011) 12–20
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Switching of reactions between hydrogenation and epoxidation of propene
over Au/Ti-based oxides in the presence of H2 and O2
Caixia Qi a, , Jiahui Huang b,d, Shuangquan Bao b, Huijuan Su a, Tomoki Akita c,d, Masatake Haruta b,d,
⇑
⇑
a R & D Centre for Gold Industrial Application, Yantai University, Shandong Applied Research Center for Gold Nanotechnology (Au-SDARC), 32 Qingquan Road,
Laishan District, Yantai 264005, PR China
b Graduate School of Urban Environmental Sciences, Tokyo Metropolitan University, 1-1 Minami-osawa, Hachioji, Tokyo 192-0397, Japan
c Research Institute for Ubiquitous Energy Devices, National Institute of Advanced Industrial Science and Technology (AIST), 1-8-31 Midorigaoka, Ikeda, Osaka 563-8577, Japan
d Core Research for Exploratory Science and Technology (CREST), Japan Science and Technology Agency (JST), 4-1-8 Hon-cho, Kawaguchi, Saitama 332-0012, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
In a H2 and O2 mixture, the reaction of propene switches between hydrogenation and epoxidation over
Au/Ti-based oxides. Reaction pathways are strongly dependent on the size of Au particles and on the
presence of alkalis. The hydrogenation prevails over Au clusters smaller than 2.0 nm (in the absence of
alkalis) and over Au nanoparticles larger than 5.0 nm. Alkali contamination can switch hydrogenation
to epoxidation over Au clusters and promote epoxidation over Au nanoparticles with diameters of
2.0–5.0 nm. It was also confirmed that the hydrogenation of propene is enhanced by the addition of a
small amount of O2. Switching of product by alkalis and enhancement of hydrogenation by oxygen are
interpreted by parallel and competitive reaction pathways.
Received 21 December 2010
Revised 27 March 2011
Accepted 29 March 2011
Available online 4 May 2011
Keywords:
Au catalysts
Propene
Epoxidation
Hydrogenation
Reaction switching
Alkaline effect
Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction
One surprising feature of the Au/TiO2 (P-25) catalyst is a
switch-like change in product selectivity from PO to propane
depending on Au loadings [7]. The product at reaction tempera-
tures below 373 K sharply switched from PO (epoxidation) to
propane (hydrogenation) when Au loading was decreased below
0.1 wt.%. The majority of Au particles were smaller than 2.0 nm
in diameter. Propane formation at very low gold loadings was also
observed over the Au/Ti–MCM-48 catalyst in which the size of Au
particles was assumed to be below 2.0 nm [16]. Until now, exper-
imental results have been inconsistent concerning the dependence
of the reaction products on the particle size and loading of Au
[7,13,14,16–20]; the catalysts with very low Au loadings
(<0.05 wt.%) and Au clusters smaller than 2.0 nm prefer PO forma-
tion [13,14,17,19,20], while the hydrogenation reaction can take
place over larger Au particles with a mean diameter of 8.4 nm
and 1 wt.% Au loading [18]. Therefore, it is of great importance to
elucidate ruling factors that determine the reaction route to epox-
idation or to hydrogenation of propene over Au catalysts.
The co-adsorption of H2 and O2 on gold catalysts is an interest-
ing research topic because oxidation reactions can be promoted by
H2, such as preferential CO oxidation [21] and direct epoxidation of
propene, while the hydrogenation reaction of propene is enhanced
by O2 over an Au/SiO2 catalyst [22]. As a matter of fact, there are
four competitive reactions that occur simultaneously over Au/Ti-
based oxides in the mixture of propene with H2 and O2 as shown
Gold surfaces have long been regarded as being chemically
inert. However, they exhibit surprisingly high catalytic activity
for CO oxidation at a temperature as low as 200 K when Au nano-
particles are deposited on base metal oxides by co-precipitation
(CP) or deposition–precipitation (DP) techniques [1,2]. This finding
has stimulated catalysis research on Au nanoparticles since the
1990s, and it has been revealed that nanogold catalysts possess
excellent activity and selectivity for many chemical reactions
[3–6]. A typical example is that Au supported on Ti-containing oxi-
des can catalyze the epoxidation of propene by molecular O2 in the
gas phase with [7–9] or without H2 [10]. Commercially viable
catalytic activity close to 10% yield of propene oxide (PO) over
Au/Ti–SiO2 catalyst systems has recently been achieved by opti-
mizing catalyst formulations [11–14].
In principle, Au nanoparticles and highly dispersed Ti4+ in the
form of isolated tetrahedral coordination in a SiO2 matrix are indis-
pensable for PO formation in the presence of H2 and O2 [8,9,15].
⇑
Corresponding authors. Address: Fax: +86 535 6902233 (C. Qi). Graduate School
of Urban Environmental Sciences, Tokyo Metropolitan University, 1-1 Minami-
osawa, Hachioji, Tokyo 192-0397, Japan. Fax: +81 42 677 2852 (M. Haruta).
E-mail addresses: qicx@ytu.edu.cn (C. Qi), haruta-masatake@center.tmu.ac.jp
(M. Haruta).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.03.028

C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
13
Table 1
Four competitive reactions that occur simultaneously over Au/Ti-based oxides in the mixture of propene with H2 and O2.
No.
Type of reaction
Chemical equation
2 + 1/2O2 ? H2O
C3H6 + H2 ? C3H8
C3H6 + 9/2O2 ? 3CO2 + 3H2O
C3H6 + O2 + H2 ? C3H6O + H2O
DrHo (kJ molÀ1
)
DrGo (kJ molÀ1
)
Notes
1
2
3
4
Hydrogen combustion
H
À241.8
À124.3
À1925.9
À355.0
À228.6
À86.2
Major reaction
Second major
At high temperatures
Limited reaction
Hydrogenation of propene
Complete oxidation of propene
Epoxidation of propene
À1931.6
À317.1
in Table 1. Since the combustion of propene to CO2 and water
usually needs temperatures much higher than the optimum
temperature range for PO or propane formation, the combustion
of hydrogen, which can take place at lower temperatures, should
therefore be taken into account for a study of switching from epox-
idation to hydrogenation. Herein, an effort is mainly focused on
clarifying the factors which define the reactions of epoxidation
and hydrogenation of propene in the presence of H2 and O2 over
Au/Ti-based oxide catalysts. Three types of support TiO2, titanosili-
cates, and silicates without Ti and various preparation methods are
employed for a comprehensive comparison.
and the pH was adjusted to 7 by adding NaOH solution. The sup-
port was then added, and the suspension was stirred while the
pH was maintained at 7.0 for 1.0 h. Then, the solid was separated
by filtration with or without washing and dried under vacuum at
room temperature for overnight or further calcined in air at given
temperatures.
DP-NH4OH method: An aqueous solution of HAuCl4Á4H2O with
0.05 wt.% of nominal Au loading was added to Ti–TUD-1 support
by incipient wetness impregnation. The sample was aged at room
temperature for 2.0 h and then washed twice with an aqueous
ammonia solution (15 mL, 1.0 M, pH of 11) and twice with deion-
ized water (15 mL). The solid was centrifuged between each wash-
ing. The sample was dried under vacuum at room temperature for
overnight and then calcined at given temperatures.
2. Experimental
DP-urea method: A similar method reported by Zanella et al.
[25] was applied to prepare Ti–TUD-1 supported Au catalysts with
0.05 wt.% of nominal Au loading by using urea (CO(NH2)2) as a pre-
cipitating agent; 0.5 g of Ti–TUD-1 support was added to 30 mL of
an aqueous solution of HAuCl4 (9.73 Â 10À3 M) and 0.04 g of urea.
The suspension thermo-stated at 353 K was vigorously stirred for
4.0 h in the absence of light. The cooled sample was then washed
twice using 100 mL of deionized water with centrifuging between
each washing. The sample was then dried under vacuum at
room temperature for overnight and finally calcined at given
temperatures.
SG method [26]: Solid grinding was used to deposit Au on HPG,
TiO2 microsphere, and Ti–TUD-1. The powder of the supports and
the required amount of dimethyl Au(III) acetylacetonate
[(CH3)2Au(acac)] were ground in an agate mortar in air for
20 min at room temperature, followed by reduction in 10 vol.%
H2/Ar stream or by calcination at 573 K.
2.1. Preparation of support materials
As support materials, anatase TiO2 microsphere (specific surface
area: 54 m2 gÀ1), MCM-41 (a silicate with arrays of non-intersect-
ing hexagonal channels, specific surface area: 1011 m2 gÀ1), Ti–
TUD-1 (a 3-D mesoporous titanosilicate, specific surface area:
822 m2 gÀ1), and HPG (a hybrid porous organic/inorganic Ti–SiO2
material with a structure as shown in Scheme 1, specific surface
area: 627 m2 gÀ1) were chosen in the present study. Anatase TiO2
microsphere was provided by the Institute of Process Engineering,
Chinese Academy of Sciences, Beijing, China. It was calcined at
773 K for 4 h in air before the deposition of Au. HPG was provided
by Toyota Central R & D Labs., Inc. and used as received. MCM-41
supported Au catalyst was prepared by one-pot synthesis, i.e.
incorporating gold precursor compound in the process of MCM-
41 synthesis.
The Ti–TUD-1 support with Ti/Si = 2/100 was prepared by a
modified sol–gel method similar to that reported by Shan et al.
[23]. In a typical synthesis, 29.8 g of triethanolamine (TEA, Wako,
98.0%) was added dropwise into a mixture of 0.678 g of tetrabutyl
orthotitanate (TBOT, TCI, 99.0%) and 20.8 g of tetraethoxysilane
(TEOS, Wako, 95.0%) with vigorous stirring for 2.0 h. The mixture
was continuously stirred for another 1.0 h at room temperature;
19.8 g of H2O was then added dropwise, and 14.7 g of tetraethy-
lammonium hydroxide solution (TEAOH, Merck, 20%) was poured
subsequently. After stirring for 24 h, the clear solution was trans-
ferred into a Teflon-lined autoclave and aged statically at 373 K
for 40 h. The obtained solid was then finalized by calcination at
973 K for 10 h in air with a ramp rate of 1.0 K/min from room tem-
perature to 973 K.
2.3. Characterization of Au catalysts
The size of gold particles and its distribution for Au catalysts
with higher Au loadings were observed with a JEOL JEM-2100
transmission electron microscopy operated at 200 kV. A small
amount of powder sample was put into a test tube filled with
95% ethanol solution. After agitation under supersonic environ-
ment for a few minutes, one drop of the dispersion was dipped
onto a carbon coated copper mesh and dried under light at room
temperature. SEM-EDS was performed to estimate Au loadings of
the catalysts with limited quantity.
Gold size and location information of a 0.05 wt.% Au/Ti–TUD-1
catalyst prepared by DP-NH4OH method were obtained by high-
angle annular dark field scanning transmission electron micros-
copy (HAADF–STEM) with a JEOL JEM-3000F transmission electron
microscope equipped with a digitally processed STEM imaging sys-
tem and electron energy-loss spectroscopy (EELS). Operating volt-
age was 300 kV, and resolution was about 0.20 nm. The catalyst
sample was directly dispersed on a microgrid supported on a cop-
per mesh without solvent.
2.2. Preparation of Au catalysts
DP-NaOH method [24]: Typically, a solution of HAuCl4 was pre-
pared by dissolving the appropriate amount of HAuCl4Á4H2O in
100 mL of deionized water. The solution was heated to 343 K,
2.4. Catalytic tests
Powder catalyst sample of 0.15 g was loaded into a vertical
fixed-bed U-shaped quartz reactor. The feed gas was composed
Scheme 1. The chemical structure of HPG support.

14
C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
Table 2
Reaction of propene with H2 and O2 mixture over supported Au catalysts (C3H6:O2:H2:Ar = 1:1:1:7, GHSV = 8000 hÀ1 mL gÀ1 ).
cat:
a
Support
Catalyst
number
Preparation
method
Au loading
(wt.%)
DAu
(nm)
Contaminants
(wt.%)
Calcination
conditions
Temp.
(K)
Conversion
(%)
Selectivity
(%)
Na or K
No
H2
Pyb
PO
C3H8
HPG
C.1-1
C.1-2
C.1-3
SG
ꢀ0.31
1–13
–
7.8
RTc, air
423 K, H2/Ar
573 K, air
433
373
373
4
13
83
<1
5
84
25, 72 (CO2)
No
No
100
100
Ti–TUD-1
C.2-1
C.2-2
C.2-3
C.3-1
C.3-2
C.3-3
C.4
SG
ꢀ0.31
–
–
4.3
3–15
16
No
RT, air
373
383
373
453
413
453
433
448
443
3.8
6.7
16
22
18
38
73
43
38
0.77
1.4
1.3
4.5
11
21
35
22
30
89, oxygenates
76
76
41
423 K, H2/Ar
573 K, air
Vacuum dry
573 K, air
873 K, air
573 K, air
573 K, air
Vacuum dry
24
24
53
99
99
99
98
92
DP-NaOH
0.31d
Yes
0
21
Trace
–
Trace
4.7
DP-NH4OH
DP-urea
DP-NaOH
60.04e
60.05e
60.03e
<2.0
<2.0f
<2.0f
No
No
No
C.5
C.6
MTiO2
C.7
C.8
SG
One-pot
ꢀ1.0
<2.0f
<2.0f
No
No
423 K, H2/Ar
Vacuum dry
323–383
473
9–21
2.2
7–9
1.2
No
No
95–75
99
g
MCM-41
<0.01
a
b
c
d
e
f
Chloride contamination was negligible.
Propene.
Room temperature.
Au loading was measured by SEM-EDS analysis.
Au loadings were estimated by the previous samples.
Au sizes were estimated.
g
Anatase TiO2 microsphere.
of propene, H2, O2, and Ar with a volume ratio of 10/10/10/70 and
was passed through a catalyst bed under atmospheric pressure. All
catalyst samples were not pretreated before catalytic tests. The
hourly space velocity of the feed gas was 8000 hÀ1 mL g . Reac-
À1
cat:
tants and products were analyzed by online GCs equipped with
TCD (Porapak Q column) and FID (HR-20M column) detectors
and with an auto-injector.
3. Results
Our previous works [7,11,12,15–17,27–29] and those of others
[30–34] have clearly demonstrated that the presence of approxi-
mate 2.0–5.0 nm size Au nanoparticles in the proximity of isolated
tetrahedral Ti sites is an important requirement for preparing an
efficient propene epoxidation catalyst, regardless of the types of
Ti-containing supports including anatase TiO2 and titanosilicates
with varying porosities from non-porous to high surface area mes-
oporous structure.
Propane formation at non-ignorable rates has often been ob-
served in our recent experiments under the conditions of PO syn-
thesis irrespective of supports and preparation methods used, as
listed in Table 2. As a major product in the mixture of propene,
H2, and O2, propane can be formed over Ti–TUD-1, HPG, MCM-
41, and anatase TiO2 microsphere in this work as well as over
TiO2 (P-25) [7,18] and SiO2 [22] in the literatures, indicating that
Ti, a prerequisite for PO synthesis, is not necessary for the hydroge-
nation of propene.
Fig. 1. Temperature dependence of the reaction of propene with H2 and O2 over
0.31 wt.% Au/Ti–TUD-1 prepared by DP-NaOH and by calcination in air at: (a) 573 K
(C.3-2 in Table 2) and (b) 873 K (C.3–3 in Table 2). j PO, d CO2, N acetone, .
propanal, e C3H8, J CH3CHO, s H2, w C3H6.
was composed of Au particles with a mean diameter of 4.3 nm
and with 97% of Au particles smaller than 5.0 nm, indicating that
three Au/Ti–TUD-1 catalysts prepared by SG method had the right
size of Au particles for PO production.
3.1. Propane formation over Au particles larger than 5.0 nm
Over Au catalysts in upper half of Table 2, propane formation
greatly depends on calcination temperature of the catalyst precur-
sors, which usually changes the size of Au particles and the content
of residual chloride. Concerning Au on Ti–TUD-1 with an actual Au
loading of 0.31 wt.% which produces both PO and propane, SG
method (C.2) led to higher selectivity to PO than to propane while
DP-NaOH method (C.3 in Table 2) led to higher propane selectivity,
especially after high temperature calcination of the catalyst pre-
cursor in air. The SG catalyst calcined at 573 K (C.2-3 in Table 2)
Since a considerable amount of PO was formed simultaneously
with propane when the DP-NaOH catalyst was vacuum dried at
room temperature (C.3-1 in Table 2) and possessed a wider size
distribution of Au particles in the range of 3.0–15 nm, it can be as-
sumed that PO production could be correlated with Au particles
with diameters in the majority of 2.0–5.0 nm range. As depicted
in Fig. 1, the Au catalyst became more active and selective for
propane production when calcination temperature was elevated
to 573 and then to 873 K and the mean diameter of Au particles

C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
15
increased to 16 and 21 nm, respectively, implying that the larger
the Au particles were, the more active and selective they were to
propane. This is supported by the similar phenomena observed
on a 0.31 wt.% Au/HPG catalyst prepared by SG method (C.1), as
shown in the upper part of Table 2 and Fig. 2. It is noteworthy that
the 0.31 wt.% Au/HPG catalyst calcined at 573 K in air (C.1-3 in
Table 2) with a mean diameter of Au particles of 7.8 nm exposed
Au nanoparticles outside of HPG channels as shown in Fig. 3 and
presented a surprisingly high conversion of propene up to 84% with
100% selectivity to propane without the oxidation of H2 to H2O at
temperatures 6373 K.
selectivity of 72% to CO2 could be ascribed to a relatively high reac-
tion temperature of 443 K.
The catalyst precursor was simply separated by filtration and
not washed after deposition of Au for the above-mentioned
0.31 wt.% Au/Ti–TUD-1 DP-NaOH catalysts (C.2 in Table 2). The
presence of chloride is known to cause remarkable aggregation of
Au particles in the subsequent reduction treatment [35]. The larger
Au particles for 0.31 wt.% Au/HPG calcined at 573 K (C.1-3 in Table
2) might be due to the hydrophobic property of the support when
the SG method was employed to deposit Au via a weak interaction
between –OH groups on the surfaces of HPG support and O atoms
in Au complex (CH3)2Au(O2C5H7) [26]. High selectivity to propane
over these catalysts agrees with the result reported by Chou et al.
that Au particles of about 8.0 nm in diameter with 1.0 wt.% Au
loading deposited onto P-25 TiO2 favor the hydrogenation of pro-
pene in the presence of H2 and O2 [18].
In addition, air-dried SG Au/HPG catalyst (C.1-1 in Table 2)
showed a wider size distribution of Au particles in the range from
1.0 to 13 nm. Seventy-seven percent of Au particles dropped to the
2.0–5.0 nm range and might account for 25% PO selectivity. A
In contrast to DP-NaOH method, the method of solid grinding
does not bring contaminants (e.g. Na+, K+, ClÀ) in the catalyst and
Au particles are deposited on both Si and Ti sites via the Au com-
plex precursor [26]. High selectivity to propane over three DP-
NaOH Au/Ti–TUD-1 catalysts (C.3) and three SG Au/HPG catalysts
(C.1 in Table 2) indicates that propane formation over larger Au
particles is not sensitive to the presence of alkalis and the location
of Au particles.
3.2. Propane formation on Au clusters smaller than 2.0 nm
As shown in the lower part of Table 2, high selectivity to pro-
pane with moderate conversions above 10% of propene was again
obtained over the Au/Ti–TUD-1 catalysts with very low starting
Au loadings in solution, such as 0.05 wt.% for DP-NH4OH (C.4)
and DP-urea (C.5), and 0.5 wt.% DP-NaOH (C.6 in Table 2) catalysts.
Only very limited amount of alkalis could remain in these catalysts
because NH4OH solution and urea solution and extensive washing
were applied to remove chloride in the process of catalyst prepara-
tion. Based on ICP analyses of co-precipitated Au catalysts, Na+ and
ClÀ impurities remained in DP Au catalysts were estimated to be
below 100 ppm when extensive washing and calcination in air
were performed [7]. HAADF–STEM observation of statistical 60
particles for the corresponding DP-NH4OH catalyst confirmed that
75% of Au particles were smaller than 2.0 nm as shown in Fig. 4.
Fig. 2. Temperature dependence of the reaction of propene with H2 and O2 over
0.31 wt.% Au/HPG catalyst prepared by SG method and by: (a) reduction in 10 vol.%
H2/Ar at 423 K for 2.0 h (C.1-2 in Table 2) and (b) by calcination in air at 573 K for
3.0 h (C.1-3 in Table 2). j PO, d CO2, N acetone, .propanal, e C3H8, J CH3CHO, s
H2, w C3H6.
Fig. 3. HRTEM image and size distribution of Au particles for 0.31 wt.% Au/HPG catalyst prepared by SG method and by calcination in air at 573 K (C.1-3 in Table 2), DAu
denotes the mean diameter of Au particles.

16
C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
of Au catalysts supported on micro and mesoporous TS-1 and Ti–
TUD, the presence of Au clusters smaller than 2.0 nm was observed
by HRTEM [10,13,14,19,20]. Gold located on tetrahedral TiO4 sites
is also indispensable for PO formation on Au clusters [19,20]. We
noticed that there were some amount of alkali cations (Na+ or
K+) remained in these catalysts [10,13,14,17,19,20], coming from
the processing of supports and/or catalyst preparation. Table 3 lists
the important epoxidation results, which were mostly obtained by
our group and partly by other groups. It can be seen that PO can be
principally produced on 2.0–5.0 nm Au particles no matter
whether alkalis co-exist or not, while PO formation over Au clus-
ters with mean diameters below 2.0 nm needs the assistance by
alkali.
In order to investigate the role of alkali in reaction switching be-
tween hydrogenation and epoxidation of propene over gold clus-
ters in a mixture of H2 and O2, anatase TiO2 microsphere was
chosen as the fourth support material to obtain Au clusters
(<2.0 nm) with higher Au loadings. A given amount of anatase
TiO2 microsphere was firstly impregnated with NaOH solution to
introduce 0.50 wt.% Na into the final catalyst. The SG method
was applied to deposit 0.31 wt.% Au on Na-containing TiO2, and
the ground mixture was calcined at 573 K in air. As depicted in
Fig. 5, 100% selectivity to PO was observed at 303 K over the corre-
sponding catalyst. In fresh 0.31 wt.% Au/0.50 wt.% Na–TiO2 micro-
sphere catalyst, more than 76% of Au particles were smaller than
2.0 nm in diameter, suggesting that the key role of alkali (Na) is
to shift hydrogenation to epoxidation over Au clusters. At 353 K,
selectivity to PO decreased to 56% while selectivity to CO2 in-
creased considerably. When temperature was raised to 383 K,
82% selectivity to propane was obtained initially and propane
selectivity was maintained up to 90% in 4.0 h on stream. TEM
observation (Fig. 5b) showed a remarkable aggregation of Au par-
ticles over the catalyst after reaction. Twelve percentage of Au par-
ticles were smaller than 2.0 nm, 32% of gold particles were in the
2.0–5.0 nm range, and 56% of gold particles were larger than
5.0 nm, demonstrating the size dependence of Au particles in Au/
Ti-based oxides on the reaction pathways of propene in a H2 and
O2 mixture. The positive effect of surface alkalis (not only Na and
K) on PO productivity and selectivity has already been demon-
strated by a number of works [19,35–37] for Au catalysts com-
posed mainly of Au particles with mean diameters of 2.0–5.0 nm.
Fig. 4. HAADF–STEM image and the size distribution of Au particles for Au/Ti–TUD-
1 prepared by DP-NH4OH method and by calcination in air at 573 K (C.4 in Table 2).
The Au loading in starting solution was 0.05 wt.%.
Results obtained by these three catalysts (C.4, C.5, C.6 in Table
2) support our earlier work on Au clusters (<2.0 nm) on P-25
TiO2, which favors hydrogenation of propene in the absence of al-
kali ions (e.g. Na+) [7]. The simultaneous formation of PO and pro-
pane was observed for the medium Au loadings and mean
diameters of Au particles slightly smaller than 2.0 nm [7]). It can
be suggested that Au particles larger than 2.0 nm mainly led to
PO formation while those smaller than 2.0 nm to propane forma-
tion in the absence of alkalis.
EELS measurements of the oxide supports were carried out to
obtain Ti composition mapping and to define the deposition sites
of Au clusters on Ti–TUD-1 for the 0.05 wt.% Au/Ti–TUD-1 DP-
NH4OH catalyst (C.4 in Table 2). It was hard to find any correlation
between tiny Au clusters and Ti sites, implying that these Au clus-
ters were mostly located on the SiO2 matrix but not concentrated
in the proximity of Ti sites. This feature was different from that
of Au catalysts with high metal loadings prepared by the DP meth-
od [17]. It is likely that very small amounts of gold entities are
washed away from Ti sites. Au clusters on both Ti and Si sites
can proceed the hydrogenation of propene.
3.4. Enhancement of propane formation by O2
Over the catalyst of 1.0 wt.% Au deposited on anatase TiO2
microsphere (C.7 in Table 2) by SG method, the reaction of propene
with H2 and O2 was carried out after in situ reduction of the cata-
lyst precursor in a flowing stream of 10 vol.% H2 in Ar at 423 K for
2.0 h. As demonstrated in Fig. 6, propane was dominant in the tem-
perature range of 298–383 K, except that trace PO was observed in
the very beginning of the reaction. No product was found when H2
was removed from the feed (propene + O2) at 323 K, whereas pro-
pane was formed when O2 was replaced with H2 (propene + H2)
even at temperature down to 273 K. The mean size of Au particles
was not measured but assumed to be below 2.0 nm through a com-
parison with the catalyst with Na in section 3.3.
So far, there is no report on the production of propane from pro-
pene in a mixture of H2 and O2 over Au/Ti-based oxides with a
mean diameter of Au particles in 2.0–5.0 nm range. In principle,
the reaction of propene hydrogenation can mainly take place on
the Au/Ti-based oxides when Au particles are larger than 5.0 nm
or smaller than 2.0 nm in the presence of H2 and O2.
3.3. Necessity of alkalis for PO formation over Au clusters smaller than
2.0 nm
More propane was formed when oxygen was added again. A
continued increase in the conversion of propene with an increase
in reaction temperature from 273 to 353 K indicates that the reac-
tion of propene hydrogenation over Au catalyst is enhanced by O2,
in agreement with our result over an Au/MCM-41 with Au loading
lower than 0.01 wt.% (C.8 in Table 2) and the result reported by
Naito and Tanimoto over Au/SiO2 catalysts with small Au loadings
[22].
In contrast to the switching of the product from PO to propane
over Au/TiO2 (P-25) with Au loadings below 0.10 wt.% and Au clus-
ters smaller than 2.0 nm [7], high PO formation rate was observed
for low loadings of Au (<0.02 wt.%) on a series of modified Ti–
MCM-41 [17] and Au (<0.05 wt.%) on NH4OH-treated microporous
TS-1 [13,14] as well as Au (0.1 wt.%) on alkali-treated mesoporous
TS-1 [10,20] and on Ba-promoted Ti–TUD support [19]. In the case
As depicted in Fig. 7, 0.31 wt.% Au/HPG calcined at 573 K (C.1–3
in Table 2) shows much higher propane formation rate in the feed

C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
17
Table 3
Experimental results reported on the epoxidation of propene over Au catalysts (C3H6:O2:H2:Ar = 1:1:1:7).
a
Support
Au loading (wt.%)
DAu (nm)
Contaminants
Na or K
Temp. (K)
Conversion (%)
Selectivity (%)
Formation rate (mol/gAu h)
PO
Refs.
H2
Py
1.1
<0.6
0.46
0.71
4.5
PO
C3H8
TiO2 (P-25)
>0.4
5.0
0.67
0.84
0.60
0.1
0.01
0.04
1.37
0.02
0.11
0.05
0.10
2.2–2.4
4.6
3.4
3.8
3.2
4.1
<2.0
2.1
2.0
–
No
No
No
323
298–343
323
373
423
473
–
373
373
473
423
473
473
3.2
–
99
>83
95
>90
>80
92
Minor
>90
>90
85
No
No
No
No
No
No
Major
No
No
No
No
No
No
0.002
0.0008
0.012
0.014
0.11
0.19
–
0.27
0.02
0.68
0.45
4.0
[7]b
[34]c
[27]d
13.2
7.9
32.1
1.62
–
TS-1
TiO–SiO2
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
[28]d
[20]e
[16]d
[27]
Mesopore TS-1
Ti–MCM-48
TS-2
Ti–MCM-41
Ti–MCM-41
Ti–TUD
0.6
–
5.8
29.6
11
4.9
–
0.68
1.71
0.95
1.4
[17]d
[19]f
[13]f
[20]e
0.9
<2
1.6
99
81
85
Micropore TS-1
Mesopore TS-1
8.8
31.2
7.7
2.3
a
b
c
d
e
f
Cl contamination was negligible.
GHSV = 2000.
1800.
4000.
8000.
7000 hÀ1 mL g
.
À1
cat:
Fig. 5. (a) Product selectivity as a function of reaction temperature over 0.31 wt.% Au/0.50 wt.% Na–TiO2 microsphere prepared by SG method and by calcination in air at
573 K. The data were obtained at each temperature after 1.0 h. (b) The size distribution of Au particles for the catalysts before and after reaction.
of propene: H2:O2:Ar = 1:1:1:7 than in the feed of H2: propene:
Ar = 1:1:8 without oxygen, showing that oxygen can also enhance
the hydrogenation over larger Au nanoparticles by decreasing the
activation energy from 33.6 to 16.2 kJ molÀ1. However, the rate
markedly dropped when too much oxygen was co-fed.
producing propane mainly takes place over Au particles larger than
5 nm and over Au clusters smaller than 2.0 nm. In contrast, pro-
pane can scarcely formed over 2–5 nm Au particles, on which pro-
pene oxide is predominately produced. The hydrogenation over Au
clusters requires the absence of alkalis or the concentration at least
lower than 100 ppm [7]). The present work for the first time has
proved that a certain amount of alkali can turn the hydrogenation
of alkene to epoxidation reaction over Au clusters smaller than
2.0 nm.
4. Discussion
The most striking phenomenon related to Au catalysis is the
strong dependence of the catalytic property on the size of Au par-
ticles and on the type of support materials. The switching of reac-
tion between hydrogenation and epoxidation of propene again
embodies this unique feature of Au catalysis. In the co-presence
of H2 and O2, the occurrence of hydrogenation and epoxidation
over Au/Ti-based oxides varies with the size of Au particles, reveal-
ing that Au particles of different sizes possess different capabilities
in the chemical adsorption or dissociation of H2 and O2.
4.1. Propene epoxidation and hydrogenation against hydrogen
oxidation
Hydrogen oxidation to produce H2O (combustion) is the fastest
reaction among the reactions to occur in the gas feed containing
C3H6, H2, and O2. Hydrogen is also consumed by both epoxidation
and hydrogenation of propene with a stoichiometric ratio of 1:1.
The occurrence of hydrogen combustion can be estimated by the ra-
tio of H2 conversion to propene conversion (see Tables 2 and 3).
Hydrogen combustion takes place preferentially to the other
The experimental results reported previously and obtained in
the present study clearly show that in the co-presence of H2, O2,
and Au/Ti-based oxides as catalysts, the hydrogenation of propene

18
C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
Fig. 7. Arrhenius plots of the reaction rates for propane formation over 0.31 wt.%
Au/HPG catalyst prepared by SG method and by calcination at 573 K in air (C.1-3 in
Table 2). Space velocity is 8000 mL g hÀ1. The volume ratio of propene, H2, O2 and
Fig. 6. Time on stream change of the reaction of propene and H2 with or without O2
over 1.0 wt.% Au/TiO2 (anatase microsphere) catalyst prepared by SG method and
by reduction in a flowing stream of 10 vol.% H2 in Ar at 423 K for 2.0 h (C.7 in Table
2). e propane selectivity; j PO selectivity; h propane formation rate, – catalyst bed
temperature.
À1
cat:
Ar is 1/1/1/7, 1/1/0/8 and 1/1/8/0, respectively.
reactions in all cases, with only one exception that occurs at
temperatures 6373 K over a 0.31 wt.% Au/HPG catalyst prepared
by the SG method and followed by calcination in air at 573 K (C.1-3
in Table 2). Although hydrogen combustion can be appreciably
suppressed in the co-presence of propene [7], the rate of hydrogen
combustion is still higher than the PO formation rate by a factor of
10 when PO is a dominant product. It is worth noting that the
conversion ratio of H2 to propene is limited to being below 3.0 in
the case of propane formation no matter how much propene is
converted. This implies that the epoxidation of propene and the
combustion of H2 are parallel reactions, whereas the hydrogenation
of propene and the combustion of H2 are competitive reactions.
Generally, for both epoxidation and hydrogenation of propene
over Au catalysts, there is an optimum reaction temperature at
which each catalyst presents the highest PO or propane yield (see
Figs. 1 and 2). This optimal temperature is 6473 K as shown in
the 8th column in Table 2 and the 5th column in Table 3, because
H2 combustion is more enhanced with an increase in temperature.
Up to now, we could some way identify the features that deter-
mine product selectivity in a complex system containing propene
in a mixture of H2 and O2. A schematic representation for reaction
switching is shown in Fig. 8 as functions of the mean diameters of
Au particles, alkali content, and the reaction temperature.
Fig. 8. The schematic representation for the hydrogenation and epoxidation of
propene over Au/Ti-based oxides in the presence of H2 and O2.
hydrogenation reaction over both Au particles larger than 5.0 nm
and clusters smaller than 2.0 nm.
H2–D2–16O2–18O2 and C3H6–H2–16O2–18O2 reactions over Au/
SiO2 catalysts [22] suggest that oxygen molecules are not dissocia-
tively chemisorbed but are transformed into peroxo-like adsorbed
species, which enhance the dissociation of hydrogen molecule. The
formation of the peroxo-like species is the most accepted hypoth-
esis to explain the promotion of epoxidation by hydrogen or by
H2O over Au nanoparticles smaller 5.0 nm on Au/Ti catalysts
[10,13,19,20,30,39,40–44]. Formation of Ti–OOH species during
PO synthesis, one of the rate-determining steps in PO production
estimated by kinetic studies [45–46], was confirmed by in situ
UV–vis and EPR study [41]. One can assume that the hydroperoxy
species are formed over Au catalysts during both the epoxidation
and the hydrogenation of propene in a mixture of H2 and O2.
Table 4 presents the formation rates of PO and propane over the
catalysts in Table 2 except for the catalysts of C.7 and C.8, because
4.2. Mutual enhancement by H2 to epoxidation and by O2 to
hydrogenation
As a matter of fact, H2 as a sacrificial reductant is needed to ob-
tain high PO selectivity over Au particles smaller than 5.0 nm,
although recent work has shown that a certain amount of PO can
be obtained by Au nanoparticles or clusters with O2 alone
[10,38,39] with a help of H2O (replacing H2). On the other
hand, the addition of oxygen to some extent can enhance the

C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
19
Table 4
Formation rates of PO and propane over supported Au catalysts in Table 2 except for the catalysts of C.7 and C.8.
Catalyst number
Au loading (wt.%)
DAu (nm)
Temp. (K)
Conversion of propene (%)
Selectivity (%)
PO
Formation rate (mol/gAu h)
C3H8
PO
Propane
C.1-1
C.1-2
C.1-3
ꢀ0.31
1–13
–
7.8
433
373
373
<1
5
84
25, 72 (CO2)
No
No
0.29
100
100
–
–
–
0.58
9.7
C.2-1
C.2-3
0.31
0.31
–
4.3
373
373
0.77
1.3
89
76
Oxygenates
24
0.079
0.11
–
0.036
C.3-1
C.3-2
C.3-3
3–15
16
21
453
413
453
4.5
11
21
41
0
Trace
53
99
99
0.21
–
–
0.27
1.3
2.4
C.4
C.5
C.6
60.04
60.05
60.03
<2
<2
<2
433
448
443
35
22
30
–
99
98
92
–
–
1.7
31
15
33
Trace
4.7
there is high uncertainty in Au particle size for C.7 and the content
of Au in C.8. The PO formation rates listed in Tables 3 and 4, in gen-
eral, are greater over Au clusters than over Au nanoparticles if we
neglect those data with high gold loadings. This agrees with our re-
cent results on the better capability of Au clusters than that of Au
nanoparticles for O2 activation [10]. Interestingly, the PO formation
rate (1.7 mol/gAu h) over the catalyst of C.6 in Table 4 is compara-
ble with the highest rates reported for alkali-promoted Au clusters
(the last two columns in Table 3) [13,20], although the selectivity
to PO is as low as 4.7%. This PO formation rate is still lower by a fac-
tor of 10 than the rate of propane formation over the same catalyst,
implying that H2 dissociation is much easier than O2 activation on
Au clusters in the co-presence of H2 and O2.
The work reported by Fujitani et al. has demonstrated that
hydrogen dissociation may take place at the perimeter interfaces
between Au clusters and TiO2 [47]. The rate of HD formation by
H2 and D2 exchange per unit weight of Au increased with a de-
crease in the size of Au particles and markedly increased below
2.0 nm. It can be assumed that the rapidly dissociated hydrogen
on the perimeter interfaces of Au clusters and the support can
readily react with propene adsorbed on the surfaces of Au and
the support to form propane. Some part of dissociated hydrogen
atoms may react with adsorbed O2 at oxygen defect vacancies,
on which H2 dissociation is very slow [48], to form the hydroper-
oxy species that can in turn accelerate H2 dissociation. The OOH
species might be parallelly and competitively transferred at tetra-
hedrally coordinated Ti cation sites that can subsequently react
with C3H6 adsorbed on the support to form PO in Au/Ti-based
oxide system. One possibility for switching of hydrogenation to
epoxidation over Au clusters by a certain amount of alkali is that
more oxygen defect vacancies are created by the strong interaction
between the added alkali and the Ti-containing support, as pro-
posed by Panagiotopoulou and Kondarides [49] in the study on
the effects of alkali addition to TiO2 on the chemisorptive proper-
ties and water–gas shift (WGS) activity of supported noble metal
catalysts (NM/X–TiO2, NM = Ru, Pd, Pt, X = Li, Na, K, Cs).
In addition, the rate of propane formation by Au/HPG catalyst
with a mean diameter of Au particles of 7.8 nm (C.1-3) is higher
than those by two TUD-1 supported Au catalysts even though
the size of Au is in double. This can be ascribed to the hydrophobic-
ity of the HPG material, on which the hydrogen combustion is
markedly inhibited.
Hydrogen oxidation to water occurs at temperatures lower than
CO oxidation on Au fine powder without a support [50], indicating
that hydrogen can be dissociated on flat gold surfaces at tempera-
tures below 473 K. The surfaces of Au particles larger than 5.0 nm
may play a key role in H2 dissociation in the co-presence of H2 and
O2 over Au/Ti-based oxide catalysts. Part of propene molecules is
adsorbed on the Au surfaces and readily react with dissociated H
atoms to form propane. Over large Au particles, the number of
perimeter Au–O–Ti sites that are selective to the epoxidation of
propene markedly decreases.
4.3. Comparison with the epoxidation of higher alkenes in liquid phase
The epoxidation of alkenes higher than ethene with molecular
oxygen alone remains a great challenge because of the hard com-
petition against the allylic C–H activation. The epoxidation of pro-
pene [51], cyclohexene [52], cyclooctene [52,53], and stilbene [54]
has been carried out using H2O2 or t-butyl hydroperoxide in liquid
phase. Interestingly, 55 atoms of Au clusters were found to be a
critical size in the epoxidation of styrene by using oxygen alone
[55]. Liu et al. found that the epoxide yield over Au clusters with
a mean diameter of 1.4 nm was higher than that over Au nanopar-
ticles of 5.6 nm for the epoxidation of styrene using t-butyl hydro-
peroxide as the oxidant [56]. The catalytic performance of
supported Au catalysts for the liquid phase epoxidation reactions
shows some similarity to the results obtained in the present study
for gas phase propene epoxidation. The liquid phase epoxidation
rates reported so far mostly lie in the range of 0.1–1.5 mol/gAu h
[52–56], comparable with the rates of the gas phase propene
epoxidation over 2.0–5.0 nm Au nanoparticles, 0.1–4.0 mol/gAu h
(Tables 3 and 4).
Fujitani et al. have reported that the rate of H2 dissociation per
unit weight of Au is slowed down over 5.6 nm Au nanoparticles on
TiO2 by 30 times lower than that over 1.3 nm Au clusters [47].
Since propane is rarely reported as one of the products over the
supported 2.0–5.0 nm Au nanoparticles in a mixture of propene
with H2 and O2, one would speculate that the capability of 2.0–
5.0 nm Au nanoparticles in H2 dissociation is low and the forma-
tion of hydroperoxides might be the main driving force for H2 dis-
sociation. The limited isolated Ti4+ sites, compared with higher Au
loadings, can only catch partial hydroperoxides to form Ti–OOH
species which are indispensible for PO formation. The majority of
hydroperoxides will directly decompose to H2O, finally leading to
the poor H2 efficiency in many previously published works.
5. Conclusions
Over Au/Ti-based oxide catalysts in a mixture of propene, H2,
and O2, three reactions mainly take place as follows: H2 combus-
tion, propene epoxidation, and propene hydrogenation. The hydro-
gen oxidation to form H2O (combustion) can proceed under any
conditions while switching of reactions between hydrogenation
and epoxidation of propene occurs under specific circumstances.
The hydrogenation of propene and H2 combustion are competitive
reactions, and the latter is remarkably depressed by the former,

20
C. Qi et al. / Journal of Catalysis 281 (2011) 12–20
[18] J. Chou, N.R. Franklin, S-H. Baeck, T.F. Jaramillo, E.W. McFarland, Catal. Lett. 95
(2004) 107.
[19] J. Lu, X. Zhang, J.J. Bravo-Suárez, K.K. Bando, T. Fujitani, S.T. Oyama, J. Catal. 250
(2007) 350.
[20] J. Huang, T. Takei, T. Akita, H. Ohashi, M. Haruta, Appl. Catal. B 95 (2010) 430.
[21] S.C. Crisafulli, S. Minicò, G.G. Condorelli, A.D. Mauro, J. Mol. Catal. A 284 (2008)
24.
whereas the epoxidation of propene and H2 oxidation are parallel
reactions. The ruling factors that determine the reaction pathways
are investigated by employing different types of support materials
and several catalyst preparation methods as well as by combining
the experimental data available from the published work.
Over Au nanoparticles deposited in the proximity of tetrahedral
Ti species, epoxidation of propene predominantly takes places with
a size of 2.0–5.0 nm in a mixture of H2 and O2. The presence of
alkalis promotes the epoxidation. The hydrogenation of propene
usually prevails on Au nanoparticles larger than 5.0 nm no matter
the presence of alkalis and on Au clusters smaller than 2.0 nm in
the absence of alkalis irrespective of the location of Au clusters
deposited either on Ti sites or on Si sites. A certain amount of alka-
lis can switch hydrogenation to epoxidation over Au clusters lo-
cated on TiO4 sites. The hydrogenation of propene is enhanced by
the addition of oxygen, and the catalytic activity varies with the
gas composition, the size of Au particles, and the type of supports.
[22] S. Naito, M. Tanimoto, J. Chem. Soc. Chem. Commun. 12 (1988) 832.
[23] Z. Shan, J.C. Jansen, L. Marchese, Th. Maschmeyer, Microporous Mesoporous
Mater. 48 (2001) 181.
[24] G.R. Bamwenda, S. Tsubota, T. Nakamura, M. Haruta, Catal. Lett. 44 (1997) 83.
[25] R. Zanella, L. Delannoy, C. Louis, Appl. Catal. A 291 (2005) 62.
[26] M. Ishida, N. Kinoshita, H. Okatsu, T. Akita, T. Takei, M. Haruta, Angew. Chem.
Int. Ed. 47 (2008) 9265.
[27] B.S. Uphade, S. Tsubota, T. Hayashi, M. Haruta, Chem. Lett. (1998) 1277.
[28] M.P. Kapoor, A.K. Sinha, S. Seelan, S. Inagaki, S. Tsubota, H. Yoshida, M. Haruta,
Chem. Commun. (2002) 2902.
[29] C. Qi, T. Akita, M. Okumura, M. Haruta, Appl. Catal. A 218 (2001) 81.
[30] T.A. Nijhuis, H. Huizinga, M. Makkee, J.A. Moulijn, Ind. Eng. Chem. Res. 38
(1999) 884.
[31] E.E. Stangland, K.B. Stavens, R.P. Andres, W.N. Delgass, J. Catal. 191 (2000) 332.
[32] G. Mul, A. Zwijnenburg, B. Linden, M. van der Makkee, J.A. Moulijn, J. Catal. 201
(2001) 128.
[33] N. Yap, R.P. Andres, W.N. Delgass, J. Catal. 226 (2004) 156.
[34] J. Chou, E.W. McFarland, Chem. Commun. (2004) 1648.
[35] B.S. Uphade, M. Okumura, S. Tsubota, M. Haruta, Appl. Catal. A 190 (2000) 43.
[36] L. Cumaranatunge, W.N. Delgass, J. Catal. 232 (2005) 38.
[37] E.E. Stangland, B. Taylor, R.P. Andres, W.N. Delgass, J. Phys. Chem. B 109 (2005)
232.
[38] S. Lee, L.M. Molina, M.L. Pez, J.A. Alonso, B. Hammer, B. Lee, S. Seifert, R.E.
Winans, J.W. Elam, M.J. Pellin, S. Vajda, Angew. Chem. Int. Ed. 48 (2009) 1467.
[39] M. Ojeda, E. Iglesia, Chem. Commun. (2009) 352.
[40] A.K. Sinha, S. Seelan, M. Okumura, T. Akita, S. Tsubota, M. Haruta, J. Phys. Chem.
B 109 (2005) 3956.
[41] B. Chowdhury, J.J. Bravo-Suárez, N. Mimura, J. Lu, K.K. Bando, S. Tsubota, M.
Haruta, J. Phys. Chem. B 110 (2006) 22995.
[42] D.H. Wells, W.N. Delgass, K.T. Thomson, J. Catal. 225 (2004) 69.
[43] M. Okumura, Y. Kitagawa, K. Yamaguchi, T. Akita, S. Tsubota, M. Haruta, Chem.
Lett. 32 (2003) 822.
[44] C. Sivandinarayana, T.V. Choudhardy, L.L. Daemen, J. Eckert, D.W. Goodman, J.
Am. Chem. Soc. 126 (2004) 38.
[45] B. Taylor, J. Lauterbach, G.E. Blau, W.N. Delgass, J. Catal. 242 (2006) 142.
[46] J. Lu, X. Zhang, J.J. Bravo-Suárez, S. Tsubota, J. Gaudet, S.T. Oyama, Catal. Today
123 (2007) 189.
[47] T. Fujitani, N.T. Akita, M. Okumura, M. Haruta, Angew. Chem. Int. Ed. 48 (2009)
9515.
[48] Y. Yang, M. Sushchikh, G. Mills, H. Metiu, E. McFarland, Appl. Surf. Sci. 229
(2004) 346.
[49] P. Panagiotopoulou, D.I. Kondarides, J. Catal. 267 (2009) 57.
[50] M. Haruta, T. Kobayashi, H. Sano, N. Yamada, Chem. Lett. (1987) 405.
[51] E. Klemm, E. Dietzsch, T. Schwarz, T. Kruppa, A. Lange, de. Oliveira, F. Becker, G.
Markowz, S. Schirrmeister, R. Schütte, K.J. Caspary, F. Schüth, D. Hönicke, Ind.
Eng. Chem. Res. 47 (2008) 2086.
[52] M.D. Hughes, Y.-J. Xu, P. Jenkins, P. McMorn, P. Landon, D.I. Enache, A.F. Carley,
G.A. Attard, G.J. Hutchings, F. King, E.H. Stitt, P. Johnston, K. Griffin, C.J. Kiely,
Nature 437 (2005) 1132.
[53] S. Bawaked, N.F. Dummer, N. Dimitratos, D. Bethell, Q. He, C.J. Kiely, G.J.
Hutchings, Green Chem. 11 (2009) 1037.
[54] V. Mendez, K. Guillois, S. Daniele, A. Tuel, V. Caps, Dalton Trans. 39 (2010)
8457.
Acknowledgments
Financial supports by Japan Science and Technology Agency to a
CREST research project on gold catalysts and the Grow Programme
from the World Gold Council (R.P. Project 0408) are greatly
acknowledged. The authors also thank Dr. Shinji Inagaki and Dr.
Yasutomo Goto of Toyota Central R & D Labs., Inc. Nagakute, Japan
and Prof. Dan Wang of the Institute of Process Engineering, Chinese
Academy of Sciences, Beijing, China, for providing us HPG and
anatase TiO2 microsphere, respectively.
References
[1] M. Haruta, N. Yamada, T. Kobayashi, S. Iijima, J. Catal. 115 (1989) 301.
[2] M. Haruta, S. Tsubota, T. Kobayashi, H. Kageyama, M.J. Genet, B. Delmon, J.
Catal. 144 (1993) 175.
[3] G.J. Hutchings, M. Haruta, Appl. Catal. A: Gen. 291 (2005) 1.
[4] A. Stephen, K. Hashmi, G.J. Hutchings, Angew. Chem. Int. Ed. 45 (2006) 7896.
[5] G.C. Bond, C. Louis, D.T. Thompson, Catalysis by Gold, Imperial College Press,
London, 2006. p. 161.
[6] G.J. Hutchings, W. Goodman, Top. Catal. 44 (2007) 1.
[7] T. Hayashi, K. Tanaka, M. Haruta, J. Catal. 178 (1998) 566.
[8] T.A. Nijhuis, M. Makkee, J.A. Moulijn, Bert M. Weckhuysen, Ind. Eng. Chem. Res.
45 (2006) 3447.
[9] C. Qi, Gold Bull. 41 (2008) 24.
[10] J. Huang, T. Akita, J. Faye, T. Fujitani, T. Takei, M. Haruta, Angew. Chem. Int. Ed.
48 (2009) 7862.
[11] A.K. Sinha, S. Seelan, S. Tsubota, M. Haruta, Angew. Chem. Int. Ed. 43 (2004)
1546.
[12] B. Chowdhury, J.J. Bravo-Suárez, M. Date, S. Tsubota, M. Haruta, Angew. Chem.
Int. Ed. 45 (2006) 412.
[13] B. Taylor, J. Lauterbach, W.N. Delgass, Appl. Catal. A 291 (2005) 188.
[14] B. Taylor, J. Lauterbach, W.N. Delgass, Catal. Today 123 (2007) 50.
[15] A.K. Sinha, S. Seelan, S. Tsubota, M. Haruta, Top. Catal. 29 (2004) 95.
[16] B.S. Uphade, T. Akita, T. Nakamura, M. Haruta, J. Catal. 209 (2002) 331.
[17] C. Qi, T. Akita, M. Okumura, K. Kuraoka, M. Haruta, Appl. Catal. A 253 (2003)
75.
[55] M. Turner, V.B. Golovko, O.P.H. Vaughan, P. Abdulkin, A. Berenguer-Murcia,
M.S. Tikhov, B.F.G. Johnson, R.M. Lambert, Nature 454 (2008) 981.
[56] Y. Liu, H. Tsunoyama, T. Akita, T. Tsukuda, Chem. Comun. 46 (2010) 550.