Pt nanoparticles supported on highly dispersed TiO2 coated on SBA-15 as an efficient and recyclable catalyst for liquid-phase hydrogenation

Article abstract of DOI:10.1016/j.jcat.2012.12.007

A series of TiO₂@SBA-15 composites with different TiO ₂ loadings have been synthesized through hydrolysis of tetrabutyl orthotitanate via a facile sol-gel method in the presence of SBA-15 mesoporous silica. The TiO₂@SBA-15 composites retain the mesostructure of the SBA-15 host, and TiO₂ is highly dispersed and uniformly coated. The TiO₂@SBA-15 composites serve as remarkable supports for Pt nanoparticles, which can be applied efficiently to the liquid-phase hydrogenation of benzaldehyde, its derivatives, and other unsaturated compounds under mild conditions. Of particular note is that the Pt/TiO₂@SBA-15 catalysts can be reused several times without distinct loss in activity or selectivity. The more electron-deficient surface state, together with the structural and physicochemical features of Pt/TiO₂@SBA-15 catalysts, would improve the catalytic activity and reusability toward the liquid-phase hydrogenation of unsaturated compounds.

Full text of DOI:10.1016/j.jcat.2012.12.007

Journal of Catalysis 300 (2013) 9–19
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Pt nanoparticles supported on highly dispersed TiO2 coated on SBA-15
as an efficient and recyclable catalyst for liquid-phase hydrogenation
⇑
Xiaohong Li , Wenli Zheng, Huiyan Pan, Yin Yu, Li Chen, Peng Wu
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, Department of Chemistry, East China Normal University, 3663 North Zhongshan Rd., Shanghai 200062, China
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of TiO2@SBA-15 composites with different TiO2 loadings have been synthesized through hydro-
lysis of tetrabutyl orthotitanate via a facile sol–gel method in the presence of SBA-15 mesoporous silica.
The TiO2@SBA-15 composites retain the mesostructure of the SBA-15 host, and TiO2 is highly dispersed
and uniformly coated. The TiO2@SBA-15 composites serve as remarkable supports for Pt nanoparticles,
which can be applied efficiently to the liquid-phase hydrogenation of benzaldehyde, its derivatives,
and other unsaturated compounds under mild conditions. Of particular note is that the Pt/TiO2@SBA-
15 catalysts can be reused several times without distinct loss in activity or selectivity. The more elec-
tron-deficient surface state, together with the structural and physicochemical features of Pt/TiO2@SBA-
15 catalysts, would improve the catalytic activity and reusability toward the liquid-phase hydrogenation
of unsaturated compounds.
Received 28 September 2012
Revised 6 December 2012
Accepted 7 December 2012
Keywords:
TiO2@SBA-15 composites
Pt nanoparticles
Liquid-phase hydrogenation
Benzaldehyde
Ó 2012 Elsevier Inc. All rights reserved.
Unsaturated compounds
1. Introduction
hydrogen transfer, using isopropanol as the hydrogen donor in
the presence of potassium hydroxide at 349 K [16].
The reduction of carbonyl groups to obtain the corresponding
alcohols is an important step in the preparation of fine chemicals
and pharmaceuticals. The transformation can be achieved through
phase transfer catalysis [1], hydrogenation by hydrogen transfer
[2], or direct hydrogenation [3]. Among metallic active centers that
can realize the hydrogenation of carbonyl compounds, Pt catalysts
have been focused on for several years, due to their high catalytic
performance [4–10].
Recently, TiO2 has been attracting much attention as a catalyst
or as a catalyst support [11,12]. When Tauster and Fung found that
a strong metal–support interaction (SMSI) existed between noble
metals and TiO2 in the late 1970s, great interest in TiO2 support
was aroused [13]. For instance, the hydrogenation of methyl oleate
over 2%Ru/TiO2, 4.7%Sn/TiO2, and 2%Ru/4.7%Sn/TiO2 catalysts was
studied [14]. Ni/TiO2 was investigated in catalytic crotonaldehyde
hydrogenation and significant enhancement in selectivity to crotyl
alcohol was observed [4]. Pt/TiO2 catalyst was applied to the hydro-
genation of aldehydes and exhibited high activity and excellent
selectivity to the primary alcohol due to the SMSI between Pt par-
ticles and TiO2 [5,15]. Platinum (0) nanoparticles supported on tita-
nia, prepared by impregnation of a commercial TiO2 support with
an aqueous solution of H2PtCl6 and further reduction under flowing
hydrogen, effectively catalyzed the reduction of acetophenone by
However, TiO2 has serious disadvantages when used as a cata-
lyst support for noble metals. First, TiO2 commonly has small sur-
face area, so that the adsorption and concentration of reactants
around the active centers is low [17]. Second, the required li-
quid–solid separation is expensive, due to the formation of milky
dispersions after the TiO2 powder catalyst is mixed in water [12].
For these reasons, the synthesis of mesoporous-material-sup-
ported TiO2 has been investigated as a way to facilitate the catalyst
recovery stage. Depositing TiO2 onto a large internal surface [18] or
incorporating Ti species into mesoporous frameworks such as mes-
oporous silicate [19] could produce some novel materials, which
not only take advantage of mesoporous materials with large spe-
cific surface area and pore volume, but also exhibit properties
not found in TiO2 alone. This could be explained as a synergetic ef-
fect [20].
Nanocrystalline TiO2/SBA-15 materials, prepared by a postsyn-
thesis step via hydrolysis of Ti-alkoxide, showed much higher pho-
todegradation ability for methylene blue than P-25 commercial
TiO2 nanoparticles. Besides much higher specific surface area and
pore volume of the TiO2/SBA-15 composites than in pure TiO2,
the smaller particle size of TiO2 with higher crystallization of
TiO2/SBA-15 composites leads to improvement of the photocata-
lytic activity [21]. Another interesting example was TiO2 nanocrys-
tals embedded inside SBA-15 after being loaded with Pt
nanoparticles, which displayed 1.7–3 times higher activity than
Pt catalyst supported on bulk TiO2 oxides, and exhibited two orders
⇑
Corresponding author. Fax: +86 21 6223 8590.
E-mail address: xhli@chem.ecnu.edu.cn (X. Li).
of magnitude higher activity than Pt/SBA-15 and Pt/c-Al2O3
0021-9517/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2012.12.007

10
X. Li et al. / Journal of Catalysis 300 (2013) 9–19
catalysts in ethyl acetate combustion. The reasonable explanation
is that insertion of TiO2 inside the SBA-15 host could stabilize
the small nanocrystals of active TiO2 polymorphs at high tempera-
tures during the preparation process [22].
Additionally, TiO2 as
a reducible support would interact
strongly with noble metal such as platinum after treatment at an
elevated temperature (normally higher than 773 K) under hydro-
gen [23], so the TiO2 layer incorporated on silica would improve
the dispersion and stability of noble metals. The strong interaction
would influence the electronic state of noble metals in turn and
likely causes considerable changes in their activity in catalytic
reactions. It was proved that the strong interaction between plati-
num and TiO2 suppressed the migration and aggregation of Pt par-
ticles on the TiO2 surface [24]. Therefore, the Pt catalysts
impregnated on TiO2-incorporated silica showed higher catalytic
activity in methane combustion than Pt catalysts impregnated on
silica [24,25].
In the present work, we introduced TiO2 with different loadings
onto the SBA-15 surface to prepare TiO2@SBA-15 composites and
investigated the catalytic performance of Pt/TiO2@SBA-15 catalysts
toward the liquid-phase hydrogenation of benzaldehyde, its deriv-
atives, and other unsaturated compounds. The Pt/TiO2@SBA-15
catalysts showed performance superior to that of the Pt/TiO2 cata-
lyst and Pt/SBA-15 catalyst in water under mild conditions. The
excellent performance of the Pt/TiO2@SBA-15 catalysts could be
attributed to their more electron-deficient surface state and their
structural and physicochemical properties.
Scheme 1. Preparation of the xTS composites and the Pt/xTS catalysts.
pressure range from 0.05 to 0.35. The pore size distribution curves
were calculated from analysis of the adsorption branch of the iso-
therm using the Barrett–Joyner–Halenda (BJH) algorithm. The dif-
fuse reflectance UV–visible spectra were recorded on a Shimadzu
UV-2400PC spectrophotometer using BaSO4 as a reference. The
morphology and TiO2 size were measured by scanning electron
microscopy (SEM) on a Hitachi S-4800 microscope. Transmission
electron microscopy (TEM) and scanning transmission electron
microscopy (STEM) images were taken with an FEI Tecnai G2-
TF30 microscope at an acceleration voltage of 300 kV.
2. Experimental
2.1. Catalyst preparation
The CO chemisorption of samples was measured at 308 K on a
Quantachrome CHEMBET-3000 pulse chemisorption analyzer after
the samples were pretreated in a 5 vol.% H2/95 vol.% Ar flow at
673 K for 2 h. The degree of dispersion and the mean particle size
(cubic model) were estimated from the measured CO uptake,
assuming a cross-sectional area for a surface platinum atom of
8.0 ꢀ 10ꢁ20 m2 and a stoichiometric factor of one, using the nomi-
nal platinum concentrations. The surface electronic state of Pt was
examined using X-ray photoelectron spectroscopy (XPS) measure-
ments with a Thermo Fisher Scientific ESCALAB 250 spectrometer
SBA-15 was synthesized using the triblock copolymer Pluronic
123 and tetraethyl orthosilicate (TEOS) according to Ref. [26].
The TiO2@SBA-15 composites were prepared by hydrolysis of tetra-
butyl orthotitanate (TBOT) onto SBA-15 via a sol–gel method [21].
A typical procedure was as follows: SBA-15 (1.0 g) was sonicated in
15 mL isopropanol for 2 h, and then, the required amount of TBOT
was added dropwise with stirring for 1 h. Then, water was slowly
added to the mixture to hydrolyze TBOT. The stirring was main-
tained for an extra 2 h to complete the hydrolysis of TBOT. The
as-prepared materials were recovered by filtration, rinsed with
deionized water and ethanol, dried at 353 K overnight, and finally
calcined at 773 K for 2 h. The obtained TiO2@SBA-15 composites
were denoted as xTS, where x represents the TiO2 loading by
weight. For comparison, pure TiO2 was also prepared using the
same method without SBA-15.
The 5 wt% Pt/xTS catalysts were prepared according to Ref. [27].
The xTS composites were impregnated with an ethanolic solution
of H2PtCl6. After 4–6 h of stirring, the mixture was evaporated to
remove the excess solvent, followed by drying at 393 K overnight.
Then, the catalyst precursors were reduced in an aqueous solution
of sodium formate. The 5 wt% Pt/TiO2 catalyst and Pt/SBA-15 cata-
lyst were prepared using the same procedures. The procedures for
preparing the xTS composites and the Pt/xTS catalysts are graphi-
cally illustrated in Scheme 1.
with Al K
a radiation (1486.6 eV) as incident beam with a mono-
chromator. All the spectra were obtained at room temperature,
and the binding energies of elements were referenced to the C1s
peak at 284.8 eV.
The electronic state of Pt was also characterized by IR spectros-
copy using CO as a probe. All infrared spectra were collected on a
Fourier transform infrared spectrometer (Nicolet Nexus 470) with
a resolution of 4 cmꢁ1 and 64 scans in the region 4000–1000 cmꢁ1
.
The sample was pressed into a self-supporting wafer (ca. 15 mg/
cm2) and put into a quartz IR cell with CaF2 windows. It was then
pretreated under hydrogen flow at 673 K for 2 h. The IR measure-
ments were carried out as follows: the pretreated sample was
evacuated at room temperature for about 20 min, and then, 20 Torr
of CO was introduced. The IR spectra were obtained after the gas-
eous CO was removed by evacuation. The amount of Pt atoms lea-
ched into solution after reaction and the actual Pt content of the
catalysts were detected with Thermo Elemental IRIS Intrepid II
XSP inductively coupled plasma-atomic emission spectroscopy
(ICP-AES).
2.2. Characterization
The X-ray diffraction (XRD) patterns of the samples were col-
lected on a Bruker D8 ADVANCE instrument using Cu Ka radiation.
The nitrogen adsorption–desorption isotherms of samples were
measured at 77 K on a Quantachrome Autosorb-3B system after
the samples were evacuated for 10 h at 473 K. The BET-specific
surface area was calculated using adsorption data in the relative
2.3. Catalytic tests
For the hydrogenation of unsaturated compounds, Pt/xTS, Pt/
SBA-15, or Pt/TiO2 catalyst (0.1 g) was pretreated under hydrogen

X. Li et al. / Journal of Catalysis 300 (2013) 9–19
11
flow (40 mL minꢁ1) at 673 K for 2 h before use. The catalyst was
then mixed with solvent (20 mL) and substrate (21 mmol). The
mixture was subsequently transferred to a 100-mL autoclave.
The hydrogenation reaction began with stirring (1200 rpm) at a
designated temperature after hydrogen (4.0 MPa) was introduced
into the autoclave. The reaction was stopped after a proper
time, and the products were analyzed by GC-FID (GC-2014,
Shimadzu Co.) equipped with a capillary column (DM-WAX,
introduced, the more surface area and the more pore volume of
the SBA-15 were occupied. Accordingly, the BET surface areas of
5TS, 10TS, 15TS, and 20TS were 638, 625, 618, and 600 m2 gꢁ1
,
respectively. The pore sizes of 5TS, 10TS, 15TS, and 20TS were
7.4, 7.3, 6.7, and 6.5 nm, respectively.
Further evidence about the size of TiO2 nanoparticles was ob-
tained from the diffuse reflectance UV–visible spectra of xTS com-
posites (Fig. 4). It is clearly observed that there is a distinct blue
shift in the absorption band edge of xTS composites with decreas-
ing TiO2 content, which can be attributed to the well-known quan-
tum size effect for semiconductors when the particle size decreases
[28]. The adsorption band edge is 380 nm for the pure TiO2, while it
is blue-shifted to about 290 nm for 5TS composites. This is consis-
tent with the SEM characterizations and strongly demonstrates
that the mesopores of SBA-15 can effectively suppress the growth
of TiO2 particles, resulting in high dispersion of TiO2.
Fig. 5 shows the low-angle and wide-angle XRD patterns of the
Pt/xTS catalysts. All the catalysts possessed a typical mesoporous
structure similar to that of the SBA-15 host, demonstrating that
loading of Pt nanoparticles did not destroy the mesostructure of
the xTS composites, which was further confirmed by the nitrogen
adsorption–desorption isotherms (Fig. S2 in the Supplementary
information). The specific surface area of Pt/xTS catalysts de-
creased with increased TiO2 loadings (Table 1). Compared with
the SBA-15 host, the specific surface area of the Pt/xTS catalysts
was strongly decreased, mainly because of occupation of titanium
oxide and Pt nanoparticles. Occasionally, the Pt nanoparticles also
blocked the mesopores of TiO2@SBA-15 composites. The broad and
weak Pt(111) diffraction peak (inset of Fig. 5) indicates that the Pt
particles were well dispersed on the composites. The Pt particle
size and dispersion were also calculated according to the measured
CO chemisorption (also listed in Table 1). The average Pt particle
sizes for Pt/5TS, Pt/10TS, and Pt/15TS catalysts were around 9.6,
8.3, and 7.0 nm, respectively. With the increase in TiO2 loading in
the xTS composites, the Pt particle size was decreased, consistent
with the results from the wide-angle XRD. This also proves that
the introduction of TiO2 inside the SBA-15 mesopores was helpful
for the dispersion of Pt particles.
The morphology and Pt particle size were also characterized by
TEM. As displayed in the TEM images of Pt/xTS (Fig. 6a–c), the Pt
nanoparticles were highly dispersed, mostly inside the mesopores
of SBA-15, and the distribution of Pt particle size is centered at 7.1–
7.8, 5.2–6.3, and 4.4–5.2 nm for the Pt/5TS, Pt/10TS, and Pt/15TS
catalysts, respectively. The differences between the sizes of the
Pt particles measured from the CO chemisorption and TEM can
be interpreted as showing that the Pt nanoparticles were partly
embedded in TiO2, so that CO cannot be adsorbed during the CO
chemisorption measurements. Hence, the result calculated from
CO chemisorption is slightly higher than that measured from
TEM. As for the Pt/TiO2 catalyst, the Pt particle size was around
2.3 nm with 48.1% dispersion according to the CO chemisorption,
whereas from the TEM image of the Pt/TiO2 catalyst (Fig. 6d), the
mean Pt particle size was centered at 3–3.4 nm. Owing to addi-
tional CO chemisorption onto TiO2 [29], the Pt particle size calcu-
lated from CO chemisorption was smaller than that from the
TEM image for the Pt/TiO2 catalyst. The actual Pt content in Pt/
15TS and Pt/TiO2 catalysts was also measured by ICP-AES. The
nominal 5.0 wt.% Pt/15TS catalyst actually contained 4.2 wt.% Pt,
while for the Pt/TiO2 catalyst, a slightly higher Pt amount of about
4.8 wt.% was detected.
30 m ꢀ 0.25 mm ꢀ 0.25
lm).
3. Results and discussion
3.1. Characterization of xTS composites and Pt/xTS catalysts
The xTS composites were first investigated by XRD and N2 sorp-
tion measurements. All the xTS composites displayed the typical
(100), (110), and (200) diffraction peaks in the region of
2h = 0.5–2°, characteristic of the same hexagonal structure as the
SBA-15 host (Fig. 1). The wide-angle XRD patterns (inset in
Fig. 1) did not show the distinct TiO2 crystalline peaks until the
TiO2 loading reached 20 wt.%. The grain size of TiO2 in 20TS com-
posites from XRD was about 5 nm according to the Scherrer equa-
tion. The morphology and TiO2 particle size were also measured by
SEM (Fig. 2). For 5TS composites, the morphology of SBA-15 was
retained and the external surface of SBA-15 was smooth. With
TiO2 loading increased to 10TS, there were a few TiO2 particle
aggregates smaller than 10 nm located outside the SBA-15 mesop-
ores. From the SEM images of 15TS, many well-dispersed TiO2 par-
ticle aggregates smaller than 30 nm located outside the SBA-15
mesopores were clearly observed. For pure TiO2, sphere-like aggre-
gates larger than 500 nm were detected by SEM. The TiO2 aggre-
gate size increased with the increase in TiO2 loading in xTS
composites. Nevertheless, the TiO2 aggregate size in xTS compos-
ites was much smaller than that of the pure TiO2. In other words,
the growth of TiO2 particles was greatly restrained when they were
coated onto the SBA-15.
To further characterize the distribution of TiO2 in the xTS com-
posites, the STEM mapping of 15TS composites was conducted. As
shown in Fig. 3, the TiO2 was uniformly and highly dispersed on
the SBA-15 surface, including inside the mesoporous channels
and on the outer surface.
The typical type IV N2 adsorption–desorption isotherms of xTS
composites (Fig. S1 in the Supplementary information) indicate
that xTS composites preserved the mesostructure of the SBA-15
host. The specific surface area, the pore volume, and the pore size
of the xTS composites decreased with increasing TiO2 loading com-
pared with the SBA-15 host (Table 1), because the more TiO2 was
(100)
20TS
15TS
10TS
5TS
SBA-15
TiO2
20
30
40
50
2θ (°)
60
70
80
(110)
(200)
SBA-15
5TS
10TS
15TS
20TS
3.2. Liquid-phase hydrogenation with the Pt/xTS catalysts
1
2
3
4
5
3.2.1. Hydrogenation of benzaldehyde with the Pt/xTS catalysts
Before we investigated different Pt catalysts for the liquid-
phase hydrogenation of benzaldehyde, we wanted to exclude the
2θ (°)
Fig. 1. Low-angle and wide-angle (inset) XRD patterns of the xTS composites.

12
X. Li et al. / Journal of Catalysis 300 (2013) 9–19
Fig. 2. SEM images of (a and b) 5TS, (c and d) 10TS, (e and f) 15TS, and (g and h) TiO2.
mass transfer limitation in advance, experimentally and theoreti-
cally. To investigate the mass transfer limitations in a batch reac-
tor, the stirring rate was changed for the hydrogenation of
benzaldehyde with the Pt/15TS catalyst as a model catalyst in
water. Fig. S3 in the Supplementary information shows the plot
of conversion versus the stirring rate. The conversion of benzalde-
hyde increased linearly with increasing stirring rate when the
stirring rate was less than 900 rpm and remained almost
constant after the stirring rate reached 1000 rpm. Hence, in the
following studies, 1200 rpm was used for the hydrogenation of
benzaldehyde, so that the mass transfer limitation could be ex-
cluded. In addition, a Weisz–Prater analysis and a Mears’ analysis
were also made [30,31]. The Weisz–Prater criterion gave
indicating no external diffusion limitations (see the Supplementary
information).
The Pt/xTS catalysts were first applied to the liquid-phase
hydrogenation of benzaldehyde at ambient temperature. For com-
parison, the Pt/SBA-15 catalyst was also used. As listed in Table 2,
the Pt/SBA-15 catalyst gave only 22.0% conversion of benzaldehyde
in neat water within 1 h (Table 2, entry 1). After 5 wt.% TiO2 was
introduced onto the SBA-15 surface, the catalytic performance
was greatly improved, obtaining 73.0% conversion of benzaldehyde
(Table 2, entry 2). When the TiO2 loading increased to 10 wt.% in
the composites, the reaction went much faster and 92.6% conver-
sion was achieved (Table 2, entry 3). If the TiO2 loading further in-
creased to 15 wt.%, the catalyst attained 100% conversion (Table 2,
entry 4). When the TiO2 loading in the composites increased from
5 wt.% to 15 wt.%, the average Pt particle size decreased with
robsq
cR2p
CWP
¼
¼ 1:4 ꢀ 10ꢁ6 < 1, indicating no internal diffusion lim-
itationsD. effTChs e Mears’ criterion gave CM
¼
¼ 0:011 < 0:15,
robsqbRp
n
k
cCAb

X. Li et al. / Journal of Catalysis 300 (2013) 9–19
13
Fig. 3. STEM images of 15TS composites.
Table 1
Relevant physicochemical parameters of supports and Pt catalysts.
Sample
SBET (m2 gꢁ1
)
Vpore (cm3 gꢁ1
)
Pt size (nm) Pt dispersion (%)
5TS
10TS
SBA-15
5TS
10TS
15TS
20TS
Pt/5TS
Pt/10TS
Pt/15TS
Pt/SBA-15 294
Pt/TiO2 67
830
638
625
618
600
400
367
355
1.10
0.84
0.84
0.77
0.74
0.89
0.87
0.77
0.89
0.37
–
–
–
–
–
–
–
–
15TS
–
–
9.6
8.3
7.0
14.1
2.3
11.4
13.6
16.1
8.0
48.1
20TS
increased dispersion. Consequently, the performance of the Pt/xTS
catalysts was improved with the increased TiO2 loading. It is worth
noting that higher than 99% selectivity to benzyl alcohol was
achieved in these cases.
200
300
400
500
600
700
800
Wavelength (nm)
Fig. 4. Diffuse reflectance UV–visible spectra of the xTS composites.
Besides water as solvent, the hydrogenation of benzaldehyde
with the Pt/15TS catalyst was also carried out in ethanol. Only
64.3% conversion was obtained in ethanol (Table 2, entry 5), much
lower than that obtained in water. The noticeable solvent effect on
benzaldehyde hydrogenation with the Pt/15TS catalyst can also be
found by comparison of the kinetic profiles in ethanol and in water
(Fig. 7a). The reaction in water ran much faster than that in etha-
nol. The hydrogenation of benzaldehyde was nearly complete in
water within 45 min, while the same reaction in ethanol needed al-
most 2 h to make a full conversion.
In order to understand the striking water effect, we investigated
the soaking properties of the Pt/15TS catalyst in water and in eth-
anol containing benzaldehyde and benzyl alcohol, respectively (see
Fig. S4 in the Supplementary information). As benzaldehyde is
insoluble in water, benzaldehyde was located on the lower layer
to form two distinct phases when benzaldehyde was added into
water. After the Pt/15TS catalyst was added to the mixture contain-
ing water and benzaldehyde, the catalyst was dispersed well in the
upper layer of water. Even though the hydrogenation product of
benzaldehyde, benzyl alcohol, has slightly higher solubility in
water than benzaldehyde, the organic mixture containing benzal-
dehyde and benzyl alcohol was still located on the lower layer sep-
arately. Thus, the hydrogenation of benzaldehyde in water with the
Pt/15TS catalyst actually took place at the interface of aqueous and
organic phases. However, for the same reaction conducted in eth-
anol, the catalyst was dispersed totally in a single phase containing
reactant, product, and solvent.

14
X. Li et al. / Journal of Catalysis 300 (2013) 9–19
derivatives. Moreover, it is worthy of note that for any substituents
TiO2
Pt/5TS
at the phenyl ring, the meta-substituted benzaldehydes with the
Pt/TiO2 catalyst were more active than their ortho- and para-
isomers.
The reason for the differences in the electronic effect of substi-
tuted groups at different positions of the phenyl ring on the cata-
lytic performance of the Pt/15TS and the Pt/TiO2 catalysts is still
unclear at the moment. Although the Pt/TiO2 catalyst showed low-
er activity toward the hydrogenation of benzaldehyde derivatives,
it furnished selectivities to the primary alcohols as high as the Pt/
15TS catalyst did.
(100)
Pt/10TS
Pt/15TS
Pt/TiO2
20
30
40
50
2θ (°)
60
70
80
(110)
(200)
SBA-15
Pt/5TS
Pt/10TS
Pt/15TS
3.2.3. Hydrogenation of other unsaturated compounds with the Pt/
15TS catalyst
Besides the benzaldehyde derivatives, we also tested the Pt/
15TS catalyst for the hydrogenation of other unsaturated com-
pounds containing C@O or C@C double bonds. Considering the ex-
tremely low solubility of these organic substrates in neat water,
especially for olefins, a mixed solvent containing 18 mL water
and 2 mL ethanol was used initially. Subsequently, taking into ac-
count that neat water can accelerate the hydrogenation of benzal-
dehyde, some substrates in this section were also hydrogenated in
neat water. In addition, the hydrogenation of these compounds
was carried out at 333 K due to the relatively low hydrogenation
ability of Pt catalyst toward the C@C bond.
As listed in Table 3, for the substrates tested in this work, the Pt/
15TS catalyst showed higher activity than the Pt/TiO2 catalyst. For
instance, styrene was fully hydrogenated to ethylbenzene within
1 h with the Pt/15TS catalyst (Table 3, entry 1), while it took about
1.5 h to realize a similar conversion with the Pt/TiO2 catalyst (Ta-
ble 3, entry 2). The corresponding TOFs (defined as the number
of moles of consumed H2 per mole of exposed Pt per second) are
1.40 and 0.31 sꢁ1 for the Pt/15TS and the Pt/TiO2 catalysts, respec-
tively. It is worthy of note that the hydrogenation of cyclohexene
to cyclohexane could be completed within 18 min, with a TOF
reaching over 4.7 sꢁ1, the fastest hydrogenation obtained with
the Pt/15TS catalyst in this work (Table 3, entry 3). For the same
reaction with the Pt/TiO2 catalyst, the full conversion needed half
an hour (Table 3, entry 4).
The Pt/15TS catalyst also affords 95.0% conversion of cyclohex-
anone within 1 h (Table 3, entry 5), whereas the conversion of
98.0% afforded by the Pt/TiO2 catalyst needed about 1.5 h (Table 3,
entry 7). For the hydrogenation of acetophenone, the Pt/15TS and
Pt/TiO2 catalysts gave 40.0% and 35.3% conversion within 2 h,
respectively (Table 3, entries 8 and 10).
To further test whether the striking water effect still exists, the
hydrogenation of acetophenone and cyclohexanone with the Pt/
15TS catalyst was conducted in neat water. The activities were
greatly enhanced in neat water (Table 3, entries 6 and 9), indicating
again that water accelerates the hydrogenation. It seems that the
hydrogenation ability of the Pt catalysts toward the conjugated
olefins or carbonyl compounds is inferior to that for the nonconju-
gated cyclic olefins or ketones under our conditions.
1
2
3
4
5
2θ (°)
Fig. 5. Low-angle and wide-angle (inset) XRD patterns of the Pt/xTS catalysts.
These observations have been already reported by several
groups [32–34]. They elucidated this phenomenon as a hydropho-
bic effect. The repulsive interactions between hydrophobic mole-
cules and water lead to the formation of hydrophobic aggregates
that allow reduction of the contact surface between them. To
maintain the network of hydrogen bonds, water wraps itself
around these aggregates, thus providing internal pressure to accel-
erate reactions with negative activation volume [35].
We also compared the kinetic profiles of the Pt/TiO2 and the Pt/
15TS catalysts for benzaldehyde hydrogenation in water under the
same conditions (Fig. 7b). The Pt/15TS catalyst showed higher
activity than the Pt/TiO2 catalyst. About 10–30% higher conver-
sions were obtained with Pt/15TS catalyst than with Pt/TiO2 cata-
lyst. Based on the conversion within 5 min, the TOF of the Pt/15TS
catalyst could reach 2.02 sꢁ1, even higher than that obtained with a
water-soluble, poly(diallyldimethylammonium chloride) sup-
ported, {[Pt30(CO)60]
2ꢁ}-derived Pt catalyst under similar condi-
tions [8]. For the Pt/TiO2 catalyst, only 0.54 sꢁ1 TOF was obtained
under the same conditions, less than one-third of that afforded
by the Pt/15TS catalyst.
3.2.2. Hydrogenation of benzaldehyde derivatives with the Pt/15TS
catalyst
With these findings in hand, we extended the liquid-phase
hydrogenation to a series of benzaldehyde derivatives with differ-
ent substituents at the phenyl ring. The Pt/15TS catalyst was cho-
sen as the standard catalyst and the hydrogenation was performed
in water at room temperature. For comparison, the Pt/TiO2 catalyst
was also applied. Fig. 8 illustrates the TOFs obtained in the hydro-
genation of benzaldehyde derivatives with the Pt/15TS and the Pt/
TiO2 catalysts in water.
For the comparatively electron-donating groups such as meth-
oxy-substituted benzaldehyde, the benzaldehydes substituted at
the ortho- and para-positions are more active than their meta-
substituted isomer in hydrogenation with Pt/15TS catalyst. The
para-substituted isomer was the most active, with a TOF of
1.04 sꢁ1. For the hydrogenation of benzaldehydes substituted at
the phenyl ring with electron-withdrawing groups such as Cl or
F, the benzaldehydes substituted at ortho- and para-positions were
less active than their meta-substituted isomer with the Pt/15TS
catalyst. The hydrogenation of 3-F-benzaldehyde with the Pt/
15TS catalyst afforded the highest TOF in this section, over
1.41 sꢁ1. The substituted groups at different positions of the phenyl
ring have distinct electronic effects.
3.2.4. Reusability of the Pt/15TS catalyst in water
The reusability of Pt/xTS catalyst is also important to consider.
Therefore, we investigated the reusability of the Pt/15TS catalyst
toward the hydrogenation of benzaldehyde in water at ambient
temperature. To our delight, the catalyst showed no distinct loss
in activity or selectivity after eight time cycles (Fig. 9). The activity
decreased slightly when the catalyst was subjected to the ninth
use. This is partially owing to the catalyst powder gradually wash-
ing out from the reaction system and to a trace amount of Pt leach-
ing during the recycling processes. After the first cycle, only 0.9% of
the total amount of Pt was leached for the Pt/15TS catalyst, as
detected by ICP-AES. This may be attributed to the interaction
Compared with the Pt/15TS catalyst, the Pt/TiO2 catalyst
showed lower activity toward the hydrogenation of benzaldehyde

X. Li et al. / Journal of Catalysis 300 (2013) 9–19
15
a
25
20
15
10
5
0
5.7-6.4 6.4-7.1 7.1-7.8 7.8-8.5 8.5-9.2 9.2-9.9 9.9-10.6
Particle Size (nm)
b
35
30
25
20
15
10
5
0
3-4.1
4.1-5.2 5.2-6.3 6.3-7.4 7.4-8.5 8.5-9.6 9.6-10.7
Particle size (nm)
c
35
30
25
20
15
10
5
0
2.8-3.6 3.6-4.4 4.4-5.2 5.2-6.0 6.0-6.8 6.8-7.6 7.6-8.4
Particle size (nm)
d
30
25
20
15
10
5
0
1.8-2.2 2.2-2.6 2.6-3
3-3.4 3.4-3.8 3.8-4.2 4.2-4.6
Particle size (nm)
Fig. 6. TEM images and particle sizes of (a) the Pt/5TS catalyst, (b) the Pt/10TS catalyst, (c) the Pt/15TS catalyst, and (d) the Pt/TiO2 catalyst and their particle size distribution.

16
X. Li et al. / Journal of Catalysis 300 (2013) 9–19
Table 2
of the Pt/TiO2 catalyst decreased dramatically during the second
run. The poor reusability of the Pt/TiO2 catalyst can be partially
attributed to the separation difficulty. As already mentioned above,
owing to formation of milky dispersions, the Pt/TiO2 catalyst can-
not be recovered easily after the reaction, while for the Pt/15TS cat-
alyst, the separation and recycling becomes more facilitated. The
photographs in Fig. S5 in the Supplementary information clearly
show the different dispersion states of these two Pt catalysts in
water. Even after 27 h natural subsidence, the milky dispersion of
TiO2 powder in water still existed, whereas the whole 15TS powder
in water subsided to the bottom.
CHO
CH2OH
Pt/xTS
20 mL Solvent, 4.0 MPa H2
Hydrogenation of benzaldehyde with different Pt catalysts.a
Entry
Catalyst
Solvent
Time (h)
Conv. (%)
Sele. (%)
1
2
3
4
5
6
7
Pt/SBA-15
Pt/5TS
H2O
H2O
H2O
H2O
H2O
EtOH
H2O
1
1
1
0.75
1
22.0
73.0
92.6
93.9
100
99
99
99
99
99
99
99
Pt/10TS
Pt/15TS
Pt/15TS
Pt/15TS
Pt/TiO2
3.3. Spectroscopic characterizations of the Pt/15TS catalyst
1
0.75
64.3
84.5
To make clear the differences between the Pt/TiO2 and the Pt/
xTS catalysts, the chemical composition of the catalyst surface
and the surface electronic state of the Pt/15TS and the Pt/TiO2 cat-
alysts were also evaluated by XPS. Before the XPS measurements,
the samples were pretreated under hydrogen flow at 673 K for
2 h. For the Pt/15TS and the Pt/TiO2 catalysts, the surface Pt/Ti
atomic ratios are 0.0265 and 0.0505, respectively, which is consis-
tent with their Pt particle dispersions observed from TEM images.
The Pt4f7/2 peak observed at 71.0 eV in the XP spectrum of Pt/TiO2
catalyst was assigned to platinum in the metallic state [36], while
the Pt4f7/2 peaks at 72.0 eV and 74.6 eV could be assignable to PtO
and PtO2 species, respectively, which were caused by reoxidation
with air (Fig. 10). As for the Pt4f XPS of the Pt/15TS catalyst, a sin-
gle peak centered at 71.6 eV appears. Compared with the binding
energy of the Pt4f7/2 level for the metallic platinum in the Pt/TiO2
catalyst, there is a 0.6-eV shift to higher energy for the Pt4f7/2 level
in the Pt/15TS catalyst, indicating that the Pt nanoparticles in the
Pt/15TS catalyst bear somewhat positive charges. Namely, the
Ptd+ species dominates on the Pt/15TS catalyst surface, while for
the Pt/TiO2 catalyst, Pt0 species are prevalent.
To reveal the intrinsic difference between these two catalysts
under the reaction conditions, the Pt/TiO2 and Pt/15TS catalysts
were further characterized by FT-IR spectroscopy using CO as
probe molecules after in situ pretreatment under hydrogen flow
at 673 K for 2 h. The IR band of linearly adsorbed CO was observed
at 2083 cmꢁ1 for the Pt/TiO2 catalyst (Fig. 11), while for CO linearly
adsorbed onto the Pt/15TS catalyst, the IR band centered at
2069 cmꢁ1. The linearly adsorbed CO stretch bands on TiO2-sup-
ported metallic Pt catalysts were reported at frequencies varying
between 2050 and 2090 cmꢁ1, which is sensitive to various factors,
such as catalyst dispersion, Pt particle microstructure (steps, crys-
tal phases, etc.), and charge transfer in the vicinity of Pt adsorption
sites [37]. In our studies, these two Pt catalysts had different dis-
persions, first. Second, the surface electronic states of the two Pt
catalysts are also different, which has already been proved by the
XPS results.
a
Reaction conditions: 100 mg Pt catalyst; 21 mmol benzaldehyde; 20 mL water
or EtOH; 4.0 MPa H2; r.t., 1200 rpm.
100
a
80
60
40
20
0
Ethanol
Water
0
20
40
60
80
100
120
Reaction time (min)
100
80
60
40
20
0
b
Pt/TiO2
Pt/15TS
3.4. Further discussion
0
10
20
30
40
50
60
Reaction time (min)
Compared with the TiO2 support, the ordered mesoporous
material SBA-15 possesses mesoporous structures with much lar-
ger surface area and pore volume. These structural features of
the SBA-15 host make the guest titanium oxide disperse easily.
Furthermore, the mesoporous channels of SBA-15 can also prevent
the growth of TiO2 particles, so that nanocrystals of TiO2 with less
than 5 nm can be stabilized inside the SBA-15 mesopores. As for
TiO2 alone, the particle size is above 500 nm.
The difference in TiO2 particle size for TiO2 alone and for the xTS
composites may cause different interactions with Pt nanoparticles.
The TiO2 particle size is about two orders of magnitude larger than
that of Pt nanoparticles in the Pt/TiO2 catalyst, while in the Pt/xTS
catalysts, the TiO2 particles have size similar to that of the Pt
Fig. 7. The kinetic profiles of benzaldehyde hydrogenation (a) with the Pt/15TS
catalyst in ethanol and in water and (b) with the Pt/TiO2 catalyst and the Pt/15TS
catalyst in water. The detailed conditions are as follows: 100 mg catalyst; 21 mmol
substrate; 20 mL solvent; 4.0 MPa H2; 1200 rpm; r.t.
between platinum and TiO2 dispersed on the SBA-15 surface,
which would enhance the stability of Pt particles and effectively
prevent leaching or aggregation of Pt particles during the recycling
processes.
Compared with the Pt/15TS catalyst, the reusability of the Pt/
TiO2 catalyst was rather poor. As also shown in Fig. 9, the activity

X. Li et al. / Journal of Catalysis 300 (2013) 9–19
17
1.8
1.5
1.2
0.9
0.6
0.3
0.0
Pt/15TS
Pt/TiO2
Cl
F
OMe
CHO
Cl
F
F
OMe
OMe
CHO
Cl
CHO
CHO
CHO
CHO
CHO
CHO
CHO
Fig. 8. Hydrogenation of benzaldehyde derivatives with the Pt/15TS and the Pt/TiO2 catalysts. The detailed conditions are as follows: 100 mg catalyst; 21 mmol substrate;
20 mL water; 4.0 MPa H2; 1200 rpm; r.t.; 1 h.
Table 3
Hydrogenation of unsaturated compounds with the Pt/15TS catalyst.a
b
Entry
Catalyst
Substrate
Time (h)
Conv. (%)
Product
TOF (sꢁ1
)
1
2
Pt/15TS
Pt/TiO2
1
1.5
100
98.6
1.4
0.31
3
4
Pt/15TS
Pt/TiO2
0.3
0.5
100
100
4.70
0.95
5
6c
7
Pt/15TS
Pt/15TS
Pt/TiO2
Pt/15TS
1
95.0
39.8
98.0
40.0
1.35
1.70
0.31
0.28
0.33
1.5
2
OH
O
8
9c
10
Pt/15TS
Pt/TiO2
2
2
89.6
35.3
0.63
0.084
OH
O
a
b
c
Reaction conditions: 100 mg catalyst; 21 mmol substrate; 18 mL water + 2 mL EtOH; 4.0 MPa H2; 1200 rpm; 333 K.
Defined as the number of moles of consumed H2 per mole of exposed Pt per second.
20 ml water was used as solvent.
nanoparticles. Because the Pt catalysts were only pretreated under
hydrogen at 673 K, the SMSI would not exist in this circumstance,
so that Pt nanoparticles would not embed to TiO2 crystals. There-
fore, the interaction between Pt nanoparticles with TiO2 would oc-
cur at the contact interface. The TiO2@SBA-15 composites with
smaller TiO2 particle size could provide more contact interface
with Pt nanoparticles; consequently, more electrons from Pt nano-
particles would donate to TiO2 particles. As a result, the Pt/15TS
catalyst is more electron-deficient than the Pt/TiO2 catalyst, which
has already been proved by the XPS analyses.
for benzaldehyde hydrogenation with the Pt/xTS catalyst is dem-
onstrated in Scheme 2. The H2 activation is fast by dissociation
adsorption on the Pt surface. Furthermore, the spillover of H from
Pt surface to TiO2 also promotes the H2 activation [38]. Then, the
activated benzaldehyde is hydrogenated to benzyl alcohol on the
Pt surface.
Therefore, the Pt/15TS catalyst exhibited excellent catalytic per-
formance toward the liquid-phase hydrogenation of benzaldehyde
and its derivatives and other unsaturated compounds containing
C@O or C@C double bonds. In addition, with regard to the superior
reusability of the Pt/15TS catalyst over the Pt/TiO2 catalyst toward
the liquid-phase hydrogenation of benzaldehyde in water, the
more facilitated separation of the Pt/15TS catalyst from water also
plays a key role.
This also implies that the Pt/15TS catalyst can more easily acti-
vate the reactant, benzaldehyde, through the adsorption of benzal-
dehyde via the carbonyl oxygen atoms bonding to the Ptd+ sites.
The proposed adsorption or activation and hydrogenation model

18
X. Li et al. / Journal of Catalysis 300 (2013) 9–19
Pt/TiO2
Pt/15TS
100
80
60
40
20
0
Scheme 2. Proposed adsorption/activation and hydrogenation model for benzal-
dehyde hydrogenation on Pt/xTS catalyst.
recyclable toward the liquid-phase hydrogenation of benzalde-
hyde, its derivatives, and other unsaturated compounds in water
under mild conditions.
1
2
3
4
5
6
7
8
9
10
Run number
Fig. 9. Reusability of the Pt/15TS and the Pt/TiO2 catalysts toward the hydrogena-
tion of benzaldehyde in water.
4. Conclusions
A series of TiO2@SBA-15 composites with different TiO2 load-
ings were prepared via a facile sol–gel method. The Pt/TiO2@SBA-
15 catalysts not only retained the ordered mesostructure of the
SBA-15 host, but also showed different surface electronic states
from the Pt/TiO2 catalyst. The Pt/TiO2@SBA-15 catalysts were
proved active and selective for the hydrogenation of benzalde-
hydes and other unsaturated compounds under mild conditions.
The most remarkable feature of the Pt/15TS catalyst is that it can
be reused at least eight times without distinct deactivation for
the liquid-phase hydrogenation of benzaldehyde in water. The
interaction of Pt particles with well-dispersed TiO2 small nanocrys-
tals coated on the SBA-15 surface causes the more electron-defi-
cient Pt nanoparticles to easily activate the reactant. The more
facilitated separation from water also greatly improved the reus-
ability of the Pt/xTS catalyst toward the liquid-phase hydrogena-
tion of benzaldehydes.
Pt4f
71.0
74.6
500
72.0
Pt0 : 57.8%
PtO: 29.4%
PtO2 : 12.8%
Pt/TiO2
Ptδ+ : 100%
71.6
Pt/15TS
84
82
80
78
76
74
72
70
68
Binding Energy (eV)
Acknowledgments
Fig. 10. XPS Pt4f spectra for the Pt/15TS and Pt/TiO2 catalysts after pretreatment
under hydrogen at 673 K for 2 h.
This work was supported by the National Natural Science Foun-
dation of China (21273076 and 20925310), the Shanghai Rising-
Star Program (08QA1402700), and the fund of the State Key Labo-
ratory of Catalysis in DICP, CAS (N-11-05) and Shanghai Leading
Academic Discipline Project (B409). The authors thank Dr. Shuai
Shen for his help with the IR measurements. Finally, helpful com-
ments have been made by the reviewers, and some of these were
followed in the revision.
2083
0.1
2069
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2012.12.007.
Pt/TiO2
Pt/15TS
References
[1] H.A. Zahalka, H. Alper, Organometallics 5 (1986) 1909.
[2] X. Wu, J. Liu, X. Li, A. Zanotti-Gerosa, F. Hancock, D. Vinci, J. Ruan, J. Xiao,
Angew. Chem. Int. Ed. 45 (2006) 6718.
[3] K. Taniya, H. Jinno, M. Kishida, Y. Ichihashi, S. Nishiyama, J. Catal. 288 (2012)
84.
[4] A. Dandekar, M.A. Vannice, J. Catal. 183 (1999) 344.
[5] M.A. Vannice, D. Poondi, J. Catal. 178 (1998) 386.
[6] S. Handjani, E. Marceau, J. Blanchard, J.-M. Krafft, M. Che, P. Maki-Arvela, N.
Kumar, J. Warna, D.Y. Murzin, J. Catal. 282 (2011) 228.
[7] S. Bhogeswararao, D. Srinivas, J. Catal. 285 (2012) 31.
[8] P. Maity, S. Basu, S. Bhaduri, G.K. Lahiri, Adv. Synth. Catal. 349 (2007) 1955.
[9] X. Li, Y. Shen, L. Song, H. Wang, H. Wu, Y. Liu, P. Wu, Chem. Asian J. 4 (2009)
699.
2200
2100
2000
1900
1800
1700
Wavenumbers (cm-1)
Fig. 11. IR spectra of CO linearly adsorbed onto the Pt/15TS and Pt/TiO2 catalysts.
In a word, the more electron-deficient surface state of Pt parti-
cles, in combination with their structural and physicochemical
features, determines that the Pt/15TS catalyst is efficient and

X. Li et al. / Journal of Catalysis 300 (2013) 9–19
19
[10] P.D. Zgolicz, J.P. Stassi, M.J. Yanez, O.A. Scelza, S.R. de Miguel, J. Catal. 290
(2012) 37.
[25] M.Y. Kim, S.B. Jung, M.G. Kim, Y.S. You, J.H. Park, C.H. Shin, G. Seo, Catal. Lett.
129 (2009) 194.
[11] G. Yang, Z. Yan, T. Xiao, Appl. Surf. Sci. 258 (2012) 8704.
[12] H. Chun, Y. Wang, H. Tang, Appl. Catal. B 30 (2001) 277.
[13] S.J. Tauster, S.C. Fung, J. Catal. 55 (1978) 29.
[14] S.A.d.S. Corradini, G.G. Lenzi, M.K. Lenzi, C.M.F. Soares, O.A.A. Santos, J. Non-
Cryst. Solids 354 (2008) 4865.
[15] P. Claus, S. Schimpf, R. Schodel, P. Kraak, W. Morke, D. Honicke, Appl. Catal. A
165 (1997) 429.
[16] F. Alonso, P. Riente, F. Rodriguez-Reinoso, J. Ruiz-Martinez, A. Sepulveda-
Escribano, M. Yus, J. Catal. 260 (2008) 113.
[26] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Science 279 (1998) 548.
[27] H. Wang, X. Li, Y.M. Wang, P. Wu, ChemCatChem 2 (2010) 1303.
[28] A. Henglein, Chem. Rev. 89 (1989) 1861.
[29] J. Silvestre-Albero, A. Sepulveda-Escribano, F. Rodrıguez-Reinoso, J.A.
Anderson, J. Catal. 223 (2004) 179.
[30] S.T. Oyama, X. Zhang, J. Lu, Y. Gu, T. Fujitani, J. Catal. 257 (2008) 1.
[31] A. Jia, G. Hu, L. Meng, Y. Xie, J. Lu, M. Luo, J. Catal. 289 (2012) 199.
[32] R. Breslow, Acc. Chem. Res. 24 (1991) 159.
[17] P. Cheng, M. Zheng, Y. Jin, Q. Huang, M. Gu, Mater. Lett. 57 (2003) 2989.
[18] Y. Zhu, L. Zhang, W. Yao, L. Cao, Appl. Surf. Sci. 158 (2000) 32.
[19] T. Maschmeyer, F. Rey, G. Sanker, J.M. Thomas, Nature 378 (1995) 159.
[20] J. Ma, J. Chu, L. Qiang, J. Xue, RSC Adv. 2 (2012) 3753.
[21] J. Yang, J. Zhang, L. Zhu, S. Chen, Y. Zhang, Y. Tang, Y. Zhu, Y. Li, J. Hazard. Mater.
B 137 (2006) 952.
[33] S. Narayan, J. Muldoon, M.G. Finn, V.V. Fokin, H.C. Kolb, K.B. Sharpless, Angew.
Chem. Int. Ed. 44 (2005) 3275.
[34] X. Wu, J. Xiao, Chem. Commun. (2007) 2449.
[35] M.-O. Simon, C.-J. Li, Chem. Soc. Rev. 41 (2012) 1415.
[36] J. Ruiz-Martinez, A. Sepulveda-Escribano, J.A. Anderson, F. Rodriguez-Reinoso,
Phys. Chem. Chem. Phys. 11 (2009) 917.
[22] X.G. Wang, M.V. Landau, H. Rotter, L. Vradman, A. Wolfson, A. Erenburg, J.
Catal. 222 (2004) 565.
[37] S. Zafeiratos, G. Papakonstantinou, M.M. Jacksic, S.G. Neophytides, J. Catal. 232
(2005) 127.
[23] W. Xu, K.D. Schierbaum, W. Goepel, J. Solid State Chem. 119 (1995) 237.
[24] M.Y. Kim, Y.S. You, H.S. Han, G. Seo, Catal. Lett. 120 (2008) 40.
[38] L.R. Baker, G. Kennedy, M. Van Spronsen, A. Hervier, X. Cai, S. Chen, L.-W.
Wang, G.A. Somorjai, J. Am. Chem. Soc. 134 (2012) 14208.