Influence of the support composition and acidity on the catalytic properties of mesoporous SBA-15, Al-SBA-15, and Al2O 3-supported Pt catalysts for cinnamaldehyde hydrogenation

Article abstract of DOI:10.1016/j.jcat.2011.06.017

Three Pt catalysts exhibiting similar metal particles size (<2 nm) were prepared on mesoporous supports differing by their chemical composition and tested in the liquid phase hydrogenation of cinnamaldehyde. The hydrogenation of the CC bond was favored with Pt/SBA-15 and Pt/Al-SBA-15, with significantly higher reaction rates on the latter catalyst. The acidic support changes the adsorptive properties of the platinum nanoparticles, modifying their surface electron density and possibly their morphology. In contrast, Pt/Al ₂O₃ was the most selective catalyst of the series toward cinnamyl alcohol, with CC and CO hydrogenations taking place at the same rate. The presence of Lewis acidic sites close to the particles is assumed to favor the adsorption of molecules via their polar CO bond. Even if cinnamaldehyde hydrogenation does not require acidic sites to be performed, the composition and acidity of the support significantly influence the reaction rate and selectivity.

Full text of DOI:10.1016/j.jcat.2011.06.017

Journal of Catalysis 282 (2011) 228–236
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Influence of the support composition and acidity on the catalytic properties
of mesoporous SBA-15, Al-SBA-15, and Al O -supported Pt catalysts for
2
3
cinnamaldehyde hydrogenation
Soraya Handjani a,b, Eric Marceau a,b, , Juliette Blanchard , Jean-Marc Krafft a,b, Michel Che a,b,c
⇑
a,b
,
d
d
d
d
Päivi Mäki-Arvela , Narendra Kumar , Johan Wärnå , Dmitry Yu. Murzin
a
Laboratoire de Réactivité de Surface, UMR 7197 CNRS, Case 178, UPMC, 4 Place Jussieu, 75252 Paris Cedex 05, France
b
CNRS, 4 Place Jussieu, 75252 Paris Cedex 05, France
Institut Universitaire de France, France
c
d
Laboratory of Industrial Chemistry and Reaction Engineering, Åbo Akademi, Biskopsgatan 8, 20500 Åbo, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
Three Pt catalysts exhibiting similar metal particles size (<2 nm) were prepared on mesoporous supports
differing by their chemical composition and tested in the liquid phase hydrogenation of cinnamaldehyde.
The hydrogenation of the C@C bond was favored with Pt/SBA-15 and Pt/Al-SBA-15, with significantly
higher reaction rates on the latter catalyst. The acidic support changes the adsorptive properties of the
platinum nanoparticles, modifying their surface electron density and possibly their morphology. In con-
Received 26 April 2011
Revised 14 June 2011
Accepted 16 June 2011
Available online 20 July 2011
2 3
trast, Pt/Al O was the most selective catalyst of the series toward cinnamyl alcohol, with C@C and C@O
Keywords:
hydrogenations taking place at the same rate. The presence of Lewis acidic sites close to the particles is
assumed to favor the adsorption of molecules via their polar C@O bond. Even if cinnamaldehyde hydro-
genation does not require acidic sites to be performed, the composition and acidity of the support signif-
icantly influence the reaction rate and selectivity.
Cinnamaldehyde hydrogenation
Selective hydrogenation
Acidity
Platinum
Mesoporous supports
Ó 2011 Elsevier Inc. All rights reserved.
1
. Introduction
6 5 2 2
the saturated aldehyde C H ACH ACH ACH@O (Fig. 1). Because
the hydrogenation of the C@C bond is thermodynamically favored
and the desorption of the alcohol is slower [8], the major product is
usually the saturated aldehyde. However, the selectivity to cin-
namyl alcohol has been found to be highly dependent on the cata-
lyst composition and structure. The catalytic properties of the
systems described in the literature are influenced by three main
parameters:
The ever increasing demand for specialty chemicals has raised
the selective hydrogenation of
a,b-unsaturated aldehydes to one
of the current challenges to be tackled both from the fundamental
and from the industrial standpoints [1]. Within this family of
reactions, the selective production of cinnamyl alcohol (C
6 5
H A
CH@CHACH OH) from cinnamaldehyde (C ACH@CHACH@O)
2
6 5
H
is highly sought for, because this unsaturated alcohol is an impor-
tant additive in the food and pharmaceutical industries. Cinammyl
alcohol has been synthesized through the Meerwein–Ponndorf–
Verley reaction [2,3], with selectivities up to 85–90%, but this
procedure also leads to the production of aluminum-containing
by-products [4,5]. To overcome this problem, new and more direct
reaction pathways involving noble metal-based heterogeneous
catalysts have been investigated over the past years and reviewed
(i) the adjunction of promoters to the catalysts, such as a sec-
ond metal (Fe, Ge, Sn, Ga) [7,9–11] or, in zeolites, a promoter
of basicity (Ba, Sr) [12], which tunes the catalytic properties
of the system with excellent results in terms of selectivity.
(ii) the size of the metal particles. Larger, faceted Pt or Ru parti-
cles would favor the adsorption of cinnamaldehyde via its
C@O bond and thus the production of cinnamyl alcohol
[13,14]. However, a reverse effect was observed with Au cat-
alysts, on which smaller particles (1 nm) were more selec-
tive [15]. An enhanced back donation between the C@O
bond and the electron-enriched surface of the Au nanoparti-
cles has been invoked to explain this observation.
[
6,7].
The initial hydrogenation of cinnamaldehyde can yield two
compounds: the unsaturated alcohol C ACH@CHACH OH and
H
6 5
2
⇑
Corresponding author at: Laboratoire de Réactivité de Surface, UMR 7197 CNRS,
Case 178, UPMC, 4 Place Jussieu, 75252 Paris Cedex 05, France. Fax: +33 1 44 27 60
(
iii) the nature of the support. On non-silicoaluminic supports,
SMSI effects [16,17] or variations of the support hydrophilicity
[18,19] have been shown to influence the catalyst properties.
3
3.
E-mail address: eric.marceau@upmc.fr (E. Marceau).
0
021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.06.017

S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
229
The alumination procedure of SBA-15 was realized following a
postsynthesis method described in the literature [24,25]. Prior to
grafting, the SBA-15 support was dehydrated at 450 °C for 2 h
Reaction path A
ꢀ1
(
heating rate 4 °C min ) and added to a solution of 3.18 g of
H
H
2
2
1
aluminum isopropoxide (Aldrich) in 60 mL of dry cyclohexane
(Fluka), under a flow of dry argon. The resulting suspension was
left under stirring for 16 h. The final gel-like material was recov-
ered by filtration and washed several times with cyclohexane.
The resulting solid was then calcined at 500 °C (heating rate of
k’ , r
hydrocinnamaldehyde
k’ , r
2 2
1
(
HALD)
H
ꢀ1
1
°C min ) with a plateau of 4 h at this temperature. A Si/Al atom-
ic ratio of 6 was determined by chemical analysis performed using
ICP at the Vernaison Center of Chemical Analysis of CNRS.
Because mesostructured aluminas may exhibit a poor stability
upon introduction of the active phase in an aqueous solution
Hydrocinnamyl alcohol
HALC)
cinnamaldehyde
CALD)
(
(
H
H
2
2
[
26], a non-mesostructured, mesoporous, and stable alumina was
H
k’ , r
k’ , r
4 4
3
3
preferred to prepare the Pt/Al catalyst. The procedure has been
2 3
O
initially described by Xu et al. [27] and later reviewed by Handjani
et al. [28]. It involves an equimolar mixture of aluminum isoprop-
oxide as a precursor and glucose as a porogen. After precipitation
at pH 5 and standing at room temperature for 5 h, the milky mix-
ture was heated at 100 °C in air to remove water and other volatile
components, leading to a brown solid gel, which was calcined at
Cinnamyl alcohol
CALC)
(
Reaction path B
Fig. 1. Reaction pathways and definition of the rate constants in the catalytic
hydrogenation of cinnamaldehyde.
ꢀ1
6
00 °C for 6 h (heating rate 1 °C min ) to completely remove the
organic additive.
On silicoaluminic supports, high selectivities have been
reported for Pt catalysts supported on zeolites Y and BEA
2.2. Catalysts preparation
[
5,20].
Supported 1 wt.% Pt catalysts were prepared on SBA-15 and
Al-SBA-15 according to an electrostatic adsorption method de-
scribed earlier [25,29]. The Pt precursor was Pt(NH ) (OH) . Two
The acidity or the basicity of the support is often put forward to
explain these results, but with conflicting interpretations [5,12,14].
In fact, any rationalization is made difficult because modifications
of the support (parameter (iii)) are often strongly correlated with
the two other parameters: alkaline earth ions may have been intro-
duced onto the support to change the catalyst properties (parame-
ter (i) [12]); the metal dispersion may considerably change
depending on the support nature, composition, and surface proper-
ties (parameter (ii) [4,21]). The influence of the support acid–base
properties on a reaction that does not require acidic sites still re-
mains puzzling.
In the present paper, we have attempted to consider this prob-
lem from a fundamental point of view and investigated how a
change of the support composition and acidity would modify the
catalytic properties of Pt-based systems whose structural charac-
teristics (particles size, support porosity) would be as close as pos-
sible. The choice of mesoporous supports (SBA-15, Al-SBA-15,
3
4
2
ꢀ1
hundred microliter of the Pt solution (Alfa Aesar, 0.52 mol L
)
was dissolved into 400 mL of deionized water. Two gram of the sup-
port was added to this solution whose pH was adjusted to 8 by drop-
ꢀ1
wise addition of a 0.1 mol L ammonia solution. The suspension
was left under stirring for 3 h at room temperature. The solid was
recovered by filtration, washed with deionized water and centri-
fuged three times, dried first at room temperature under vacuum
for 16 h and under oxygen flow at 150 °C for 2 h afterward (flow rate
ꢀ
1
ꢀ1
ꢀ1
1 L min
g , heating rate 1 °C min ). Platinum was then reduced
ꢀ1
under hydrogen flow at 500 °C for 2 h (flow rate 0.1 L min , heating
ꢀ
1
rate 2 °C min ). It was shown by CO adsorption followed by infra-
red spectroscopy that 500 °C was the lowest temperature allowing
one to attain a complete reduction in Pt on Pt/SBA-15 (absence of
x+
ꢀ1
CO/Pt band above 2100 cm at the lowest coverages).
Due to a poor electrostatic interaction between alumina and
2
+
[Pt(NH ) ] at pH 8 and to the partial transformation of mesopor-
3
4
2
Al O
3
), till now less studied than microporous ones [4,21,22], al-
ous alumina into Al(OH) upon exposure to large amounts of water
3
lowed us to prepare catalysts with different acidic surface proper-
ties and similar platinum dispersions. Catalytic results have been
complemented by an infrared study of CO adsorption on the metal
particles, which may provide some hints on the way each support
influences the metal nanoparticles properties.
[28], platinum was deposited on Al O by incipient wetness
2
3
ꢀ1
impregnation, using a 0.043 mol L solution of [Pt(NH ) ](OH)
2
3
4
prepared from the commercial solution mentioned above. Because
alumina exhibits a specific surface area lower than that of SBA-15
and Al-SBA-15, the Pt content was limited to 0.5 wt.% Pt, in order to
prevent an increase in the particles size that might be consecutive
to a poor distribution of Pt on the support. After one night of drying
2
. Experimental
at 90 °C, the catalyst was reduced in a flow of H
2
at 400 °C for 2 h
ꢀ1
ꢀ1
(
flow rate 0.1 L min , heating rate 2 °C min ). This temperature
2.1. Supports preparation
of reduction was chosen in order to obtain a Pt dispersion similar
to that of Pt/SBA-15 and Pt/Al-SBA-15.
The synthesis of purely siliceous SBA-15 was carried out as re-
The three catalysts will be referred to as: Pt/SBA-15; Pt/Al-
ported in the literature, using TEOS (Fluka) as a precursor and
block-copolymer Pluronic P123 (Aldrich) as a porogen [23]. The so-
lid obtained after precipitation and aging, first at 38 °C under stir-
ring (24 h) then at 95 °C in static conditions (72 h), was separated
by filtration, washed several times with deionized water, dried at
2 3
SBA-15; and Pt/Al O . Pt contents, Al content, and reduction temper-
atures will not be recalled.
2.3. Catalysts characterization
ꢀ1
9
5 °C for 16 h, and calcined (heating rate 1 °C min ) at 550 °C in
Elemental analysis for Pt was performed by ICP at the Vernaison
Center of Chemical Analysis of CNRS. X-ray diffraction analyses
a muffle furnace, with a plateau of 6 h at this temperature.

2
30
S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
(
XRD) were carried out at low angles on a Bruker D8 diffractome-
filled with hydrogen. 160–310 mg cinnamaldehyde (98%, Aldrich)
dissolved in 200 mL of 2-pentanol (Merck, p.a.) was then intro-
duced into the autoclave. The reaction took place for 3 h at 95 °C
(above 100 °C, cracking side reactions may be favored), under
kinetic control (stirring rate: 1500 rpm). 2-Pentanol was chosen
because as a polar solvent it increases the reaction rate, while its
ter, using Cu K radiation (0.15418 nm). Specific surface areas and
a
pore characteristics of the catalysts were determined by the BET
and BJH models from nitrogen adsorption/desorption experiments
carried out at ꢀ196 °C on samples outgassed at 150 °C for 4 h prior
to analysis, using an automatic Micromeritics ASAP 2010 M instru-
ment. After reduction under a flow of 5% H
chemisorbed hydrogen in Ar at the temperature of reduction, plat-
inum dispersion was measured at ambient temperature by H –O
titrations for Pt/SBA-15 and Pt/Al-SBA-15 (H/Pt @O/Pt = 1) and
by CO chemisorption for Pt/Al (CO/Pt = 1), using an Autochem
910 (Micromeritics).
Transmission Electron Micrographs (TEM) were collected on the
reduced catalysts using a 200 kV JEOL JEM 2010 microscope. This
instrument is equipped with a system for EDX analysis (IMIXPRISM
detector) with an analysis domain of 100 nm. Ultramicrotomed
slices (ca. 50 nm thick) were used to obtain extensive electron-
transparent regions. The samples were prepared by embedding
the material in a polymer resin subsequently cured at 70 °C for
2
/Ar and desorption of
5
C alkyl chain limits side reactions of acetalization compared to
alcohols with shorter chain [21]. Aliquots of the reaction mixture
(1–1.5 mL) were withdrawn periodically and analyzed on a gas
chromatograph (GC Agilent 6890 N) equipped with autoinjector,
FID detector, and Agilent DB-1 column. It should be noted that
the initial time reported on the conversion curves corresponds to
the first analysis time, 5–10 min after reactant and catalyst have
been contacted. The hydrogenation of cinnamaldehyde (CALD)
can result in hydrocinnamaldehyde (HALD), cinnamyl alcohol
(CALC), and the fully hydrogenated hydrocinnamyl alcohol (HALC)
(Fig. 1). Reaction pathway A leads to HALD and then HALC (hydro-
genation of the C@C bond, then of the C@O bond); reaction path-
way B leads to CALC and then to HALC (hydrogenation of the
C@O bond, then of the C@C bond). The conversion of CALD and
selectivity to CALC, HALD, and HALC were calculated as described
before [19]. The kinetic experiments were modeled using the
Odessa solver in the Modest 6.1 program package (Profmath Oy,
Helsinki, 2002).
2
2
s
s
2
O
3
s
2
2
days. Slices were cut using a diamond knife.
Transmission FTIR spectra of adsorbed CO were collected on a
Bruker Vector 22 spectrometer using a DTGS detector (resolution
ꢀ1
2
cm , 64 scans per spectrum). The setup, described earlier [30],
allows one to raise or lower the sample vertically between the
two sections of the cell: the upper oven section in which thermal
treatment takes place and the lower section equipped with a per-
3
. Results
2
pendicular body mounted with CaF windows, cooled to liquid
nitrogen temperature (LNT), where measurements take place.
Dried samples were pressed into self-supported wafers of 15 mg
and 1 cm . The wafer was first heated in situ at the temperature
3
.1. Catalysts characterization
2
X-ray diffractograms of the catalysts are similar to those of the
of reduction for 1 h under
a
flow of 5%
H
2
/Ar (flow rate
ꢀ
1
ꢀ1
three pristine supports (not shown), confirming the preservation of
the ordered mesoporosity of the SBA-15 and Al-SBA-15 systems
5
0 mL min , heating rate 2 °C min ), before being outgassed
ꢀ6
(
5 ꢁ 10 Torr) at the same temperature for 1 h. The wafer was
(
reflections (1 0 0), (1 1 0), and (2 0 0)) and of the
2 3
c-Al O structure
then moved to the lower part of the cell, which was cooled down
to ca. 100 K with liquid nitrogen. Pulses of CO were successively
of the aluminic system.
The main structural and physicochemical characteristics of the
three catalysts and corresponding supports are summarized in
Table 1. The presence of a mesoporosity is evidenced by N
physisorption isotherms (not shown). The steep uptake of N for
Pt/SBA-15 and Pt/Al-SBA-15 indicates the presence of cylindrical
mesopores whose diameter (6.2–6.5 nm) is identical to that of
the corresponding support. For these two catalysts, the specific
surface area decreases by about 15% compared with the supports,
mostly due to a clogging of the microporosity upon platinum intro-
duction. It was shown that the alumina BET specific surface area
does not decrease when Pt is introduced by incipient wetness
impregnation [28]. In contrast to the SBA-15 and Al-SBA-15 sys-
ꢀ
8
added (five pulses of 8.9 ꢁ 10 mol CO, followed by 13 pulses of
ꢀ
7
4
.5 ꢁ 10 mol). The difference spectra were obtained by subtract-
2
ing the spectrum of the dehydrated sample recorded at LNT from
those recorded after successive increments in CO. In case of Pt/
2
2 3
Al O , CO adsorption was carried out at room temperature after
reduction and outgassing, because at LNT the CO/Pt band was
not discernible due to the very intense bands resulting from the
2 3
competitive adsorption of CO on the Al O surface sites (Lewis
ꢀ1
ꢀ1
sites, 2187 cm ; Al–OH sites, 2156 cm [31]). At room tempera-
ture, only the CO/Pt bands were present.
2.4. Catalytic test
2 3
tems, the porosity of Pt/Al O is intergranular and the shape of
the hysteresis suggests an ‘‘ink-bottle’’ structure of the pores with
smaller apertures. The diameter of the pores calculated from the
adsorption branch of the isotherm was evaluated to be 5.6 nm [28].
Pt dispersion measurements led to similar results on the
three catalysts: 62–69%, i.e., an average particle diameter of
1.6–1.8 nm, assuming that the particles are hemispherical. No
particle larger than 2 nm was observed on TEM images of bulk
The catalysts were tested for the cinnamaldehyde hydrogena-
tion reaction in a 500 mL stainless steel stirred semi-batch reactor
Autoclave Engineers, USA) equipped with a sample port, a reagent
injection port, gas inlet, and vent, at P(H ) = 30 bar [10,18,19]. The
prereduced catalyst samples (0.3–0.5 g, 25–90 m) were activated
(4 bar) at 350 °C for 2 h before transfer into the reactor
(
2
l
in situ in H
2
Table 1
Physicochemical characteristics of the three catalysts and their supports.
Support/
catalyst
BET specific surface
Mesopore surface
Micropore surface
Mesopore diameter
(nm)
Mesopore volume
Pt dispersion
(%)
Ø Pt particles
(nm)
area (m2
g
ꢀ1
)
area (m
2
g
ꢀ1
)
area (m2
g
ꢀ1
)
(cm3
g
ꢀ1
)
SBA-15
867
749
525
438
367
356
531
500
342
318
367
356
233
152
94
15
–
6.5
6.5
6.3
6.2
5.6
5.3
0.90
0.88
0.57
0.56
0.60
0.55
–
69
–
69
–
62
–
1.6
–
1.6
–
1.8
Pt/SBA-15
Al-SBA-15
Pt/Al-SBA-15
Al
2
O
3
Pt/Al
2
O
3
–

S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
0.08
231
catalysts grains or of ultramicrotomed slices. Furthermore, for
(
a)
Pt/SBA-15 and Pt/Al-SBA-15, no Pt particle could be seen on the
external surface of the ordered support. EDX analysis over 10 dif-
ferent support grains confirmed that the distribution of Pt over
the three supports was homogeneous.
.
..
0
0
0
0
0
0
.07
.06
.05
.04
.03
.02
Pt CO
079
2
.
..
Si-OH CO
2157
3.2. CO adsorption/IR spectroscopy
At LNT, CO is a molecule able to interact with cationic sites
Lewis acids) and perturb the proton of oxidic surface OH groups
Brønsted acids) by bonding via its carbon atom. Moreover, for a gi-
(
(
ven type of site, the position of the perturbed m(CO) band depends
.
..
Pt2 CO
1
on the strength of the site: the stronger the acidic site, the higher
the energy [25,31]. CO is also a well-known probe molecule for
metallic sites, especially because the position of the CO band is
sensitive to the particles electronic density. CO adsorption fol-
lowed by infrared spectroscopy thus appears as a suitable tech-
nique to compare both the acidic properties of the supports and
the adsorptive properties of the Pt nanoparticles. Spectra recorded
on the three catalysts are presented in Fig. 2.
In terms of acidity, former results have shown that neither Pt/
2 3
SBA-15 nor Pt/Al O exhibits any acidic character toward two test
reactions, cumene cracking and isopropanol dehydration [25]. For
the Pt/SBA-15 sample, the only band related to the adsorption onto
0.01
0
860
2300
2200
2100
2000
1900
1800
.
..
Si/Al-OHB CO
169
0.08
.
..
2
Pt CO
2087
(b)
.
..
AlL CO
230
0
0
0
.07
.06
.05
2
2196
ꢀ1
the support is CO/silanol groups (2157 cm ) (Fig. 2a). It is recalled
here that at room temperature, the bands corresponding to CO/
weak Lewis sites (2187–2192 cm ) and CO/aluminol groups
0.04
0.03
ꢀ1
ꢀ1
(
2156 cm ) are not present for Pt/Al
In contrast, several types of acidic sites are revealed by CO
adsorption on Pt/Al-SBA-15 (Fig. 2b). Their analysis has been de-
tailed in Ref. [25]. Besides CO adsorbed on Al -like weak Lewis
sites (2196 cm ), present in low quantity, and on SiO -like silanol
groups (2158 cm ), the spectra present bands typical of silica-
aluminas and assigned to CO/strong Lewis sites (2230 cm ; iso-
lated Td Al atoms) and CO/strong and medium Brønsted sites
2173 and 2168 cm ; presumably SiAOH .•. .Al groups [32,33]).
The latter are distinguished from SiAOH and AlAOH groups by
their ability to crack cumene and dehydrate isopropanol into iso-
propene [25]. Pt/Al-SBA-15 is thus the only catalyst that exhibits
strong Lewis and Brønsted acidities after thermal activation. Pt/
2 3
O (Fig. 2c).
0
.02
...
Pt2 CO
0.01
0
1
874
2 3
O
ꢀ1
2
ꢀ1
2300
2200
2100
2000
1900
1800
ꢀ1
0
.04
0.035
.03
ꢀ
1
(c)
(
.
..
Pt CO
2
061
0
0
0
0
.025
0.02
.015
2 3
Al O exhibits a large number of weak Lewis acid sites, while Pt/
SBA-15 exhibits only the very weak Brønsted acidity of silanols.
Two regions can be distinguished in the zone of the spectra cor-
responding to the CO/Pt band. An intense band between 2000 and
ꢀ1
2
100 cm is assigned to CO linearly adsorbed on the Pt nanoparti-
0.01
...
ꢀ1
Pt2 CO
835
cles. At lower energies (1800–1900 cm ), one can find a broad and
weak band attributed to bridging CO adsorbed on two Pt atoms.
This mode of adsorption is usually favored on larger faceted parti-
cles. The low intensity of this band confirms that facets are scarce
on the nanoparticles of the three catalysts, in line with the high
dispersion measured by chemisorption. Beside this, it has also been
reported that the quantity of bridging CO/Pt is low for supports pre-
senting poor basic properties [34,35], which is the case here.
As can be seen on Fig. 2, the position of the two CO/Pt bands
varies depending on the support. A zoom on the region of linear
CO/Pt is presented in Fig. 3 at iso-coverage, after adsorption of
the first pulses of CO on Pt/SBA-15 and Pt/Al-SBA-15. Due to com-
petitive adsorption on strong Lewis and Brønsted acidic sites, CO
coverage of nanoparticles is smaller on Pt/Al-SBA-15 than on Pt/
SBA-15 for the same amount of CO introduced; the number of
pulsed required to reach the same coverage of the metallic sites
is therefore larger on Pt/Al-SBA-15. The position of the CO/Pt band
1
.005
0
2
300
2200
2100
2000
1900
1800
-1
Wavenumber (cm )
Fig. 2. FTIR spectra of: (a) Pt/SBA-15; (b) Pt/Al-SBA-15; and (c) Pt/Al
2
O
3
, after
successive introductions of CO pulses. Recording of the spectra was carried out at
LNT for Pt/SBA-15 and Pt/Al-SBA-15 and at ambient temperature for Pt/Al
except for the spectrum in dotted lines recorded after cooling down to LNT. For Pt/
2 3
O ,
ꢀ
1
Al-SBA-15, the drift of the baseline in the 1850–1950 cm region, seen from the
first pulse, results from the initial subtraction of the spectra.
two catalysts in terms of band position is significant and reproduc-
ible and larger than the initial coverage-related shift detected on
Pt/Al-SBA-15. Finally, the position of the band does not change
after heating to room temperature.
does not change with coverage for Pt/SBA-15
079 cm ), whereas for Pt/Al-SBA-15 it shifts by about +2 cm
(
m
(CO/Pt) =
ꢀ
1
ꢀ1
2
The highest position in energy of m(CO/Pt) is thus observed for
between pulse 3 and pulse 7 and remains at the same position
for the next CO pulses (2086 cm ). The difference between the
Pt/Al-SBA-15, while for Pt/SBA-15 the band looks broader with
possibly high- and low-energy contributions (Fig. 2a and b,
ꢀ
1

2
32
S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
2
086 2079
0
.04
.035
.03
.025
.02
.015
.01
0.8
(( aa ))
0
0
0
0
0
0
0
0
0
0
0
.7
.6
.5
.4
.3
.2
.1
0
0
0
0
.005
0
2
150
2100
2050
2000
-
1
Wavenumber (cm )
0
10
20
30
40
50
60
45
16
time*mPt
Fig. 3. Zoom on the linear CO/Pt region for Pt/Al-SBA-15 (full lines, pulses 3, 5, 7, 8,
and 9) and Pt/SBA-15 (dotted lines, pulses 1, 2, and 3).
0.9
(( bb ))
Fig. 3). Despite similar dispersions, Pt nanoparticles on Pt/SBA-15
and Pt/Al-SBA-15 present different adsorptive properties, with a
globally lower back-donation toward CO when particles are lo-
cated on the more acidic support.
0.8
0
0
.7
.6
2 3
The position of the linear CO/Pt band on Pt/Al O is located at a
ꢀ1
lower energy than the two others, 2061 cm (Fig. 2c). We would
like to remind here that at room temperature, the surface sites of
the alumina support do not compete with Pt for CO adsorption.
Consequently, the coverage of the Pt nanoparticles is high from
the first CO pulse. Therefore, the position of the bands cannot be
compared at iso-coverage with the two other catalysts. We will as-
0.5
0
0
0
.4
.3
.2
2 3
sume that for Pt/Al O the position of the band does not change
significantly with coverage as was observed for Pt/SBA-15 and Pt/
Al-SBA-15. Concerning an effect of the temperature, we checked
that the position of the band associated with CO linearly adsorbed
on Pt was not modified after outgassing and decreasing the tem-
perature of the cell to LNT. We can therefore conclude that the
adsorptive properties of Pt nanoparticles also vary when the sup-
port is neither silica nor silica–alumina, but alumina.
0.1
0
0
5
10
15
20
25
30
35
40
time*mPt
1
0
.9
0.8
0.7
(( cc ))
Finally, the position of the bridging CO/Pt band follows the
same trend as the linear CO/Pt band, with even larger shifts. Pt/
ꢀ1
Al-SBA-15 presents the band at the highest energy (1874 cm ),
ꢀ
1
ꢀ1
2 3
followed by Pt/SBA-15 (1860 cm ) and Pt/Al O (1835 cm ).
0.6
3.3. Catalytic results
0.5
0
0
0
0
.4
.3
.2
.1
0
Conversion curves of cinnamaldehyde vs. time and selectivity
plots vs. conversion are given in Figs. 4 and 5 respectively, while
the parameters of each test are presented in Table 2. Unlike what
has been observed during a test experiment on 1 wt.% Pt/H-beta
catalyst (Si/Al = 75), no product of cracking or acetalization reac-
tion was detected for any of the three catalysts. This shows that
the Brønsted acidity of Pt/Al-SBA-15 is not as strong as that of a
protonic zeolite. This type of mesoporous system is thus suitable
to catalyze a reaction of selective hydrogenation without causing
side reactions.
Reaction has started between the moment when the mixing and
heating of catalyst and reactants starts and the first analysis, which
explains why curves shown in Fig. 4 do not start at a molar fraction
of 1 in CALD.
0
2
4
6
8
10
12
14
time*mPt
Fig. 4. Cinnamaldehyde conversion and product distribution as a function of time
for: (a) Pt/SBA-15; (b) Pt/Al-SBA-15; and (c) Pt/Al O . O: CALD; x: HALD; h: HALC;
+: CALC.
2
3
On Pt/SBA-15, the main product of reaction is HALD, resulting
from the hydrogenation of the C@C bond. After 3 h of reaction,
the conversion of CALD is not complete and a plateau seems to
be reached in the production of HALD. The fully hydrogenated
HALC comes second, with a selectivity increasing with time and
conversion. Cinnamyl alcohol (CALC), whose production tends to
slowly increase with time, is a minor product on this catalyst (less
than 10% of selectivity at 70% of conversion and after 3 h). Similar
results have been reported for Pt particles supported on commer-
cial silica [11].

S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
233
1
00
of the reaction after 2 h only. During the first 50 min, HALD re-
mains the major product of hydrogenation, indicating that the
hydrogenation of the C@C bond is favored. But the conversion pro-
file differs from that of Pt/SBA-15, since the production of HALD
starts decreasing after 50 min and its selectivity drops sharply,
while the production of HALC, initially low, constantly increases till
reaching 80% of selectivity after 3 h. The S shape of the HALC curve
suggests that HALD is an intermediate product, ultimately leading
to HALC by hydrogenation of the C@O bond in agreement with
pathway A of the reaction mechanism (Fig. 1). On Pt/Al-SBA-15
as well, CALC is the minor product, with a selectivity of 11% at
70% of conversion and a final selectivity of 8% after 3 h. Selectivity
in CALC is almost constant with increasing conversion.
(
a)
8
6
4
2
0
0
0
0
0
Compared with the two other systems, Pt/Al
terms of selectivity. Conversion of CALD is almost complete after
h of reaction, despite the fact that the Pt content of the catalyst
2 3
O stands out in
0
20
40
60
80
100
3
is lower than that of the two other systems (0.5 wt.% compared with
1 wt.%, Table 2). If initially the main product of reaction is HALD, the
trend changes after 15 min for two reasons: (i) the production of
HALC starts increasing constantly following reaction pathway A,
and the maximum in HALD production at about 60 min shows that
on this catalyst as well, HALD is an intermediate compound con-
sumed by hydrogenation of its C@O bond;(ii) the productionofCALC
increases progressively following reaction pathway B. After 50 min
of reaction, CALC is the main product of reaction and will remain
so till the end of the run (46% of selectivity at 70% of conversion
and after 3 h of reaction). Without a kinetic analysis of the curves,
it is difficult to assert whether the hydrogenation of the C@C bond
in CALC also contributes to the production of HALC. However, no
maximum in the production of CALC is seen during the course of
the reaction, unlike what is observed for HALD; the C@O reduction
in HALD is probably faster that the C@C hydrogenation in CALC.
We have aimed at extracting apparent kinetic constants from
these experimental curves. Data were fitted following the model
developed for the most hydrophilic Pt/carbon nanofibers catalyst
described by Toebes et al. [18]. According to this model: (i) the
reaction proceeds through a Langmuir–Hinshelwood mechanism;
1
00
(
b)
8
6
4
2
0
0
0
0
0
0
20
40
60
80
100
1
00
(
c)
8
6
4
2
0
0
0
0
0
(
ii) no competition takes place with hydrogen and dissociative
hydrogen adsorption is assumed.
The rate equations (notation is given in Fig. 1) then take the fol-
lowing form
k
1
ꢂ cCALD ꢂ KCALD
r
r
r
1
2
3
¼
¼
¼
ð1 þ cCALD ꢂ KCALD þ cHALD ꢂ KHALD þ cCALC ꢂ KCALC þ cHALC ꢂ KHALC
Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
c
H
ꢂ K
H
2
ꢂ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð1Þ
1
þ
c
H
ꢂ K
H
2
k
2
ꢂ cHALD ꢂ KHALD
0
20
40
60
80
100
ð1 þ cCALD ꢂ KCALD þ cHALD ꢂ KHALD þ cCALC ꢂ KCALC þ cHALC ꢂ KHALC
Þ
Cinnamaldehyde conversion (%)
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
c
H
ꢂ K
H
2
ꢂ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð2Þ
Fig. 5. Selectivity plots as a function of conversion. Open circles, Pt/SBA-15; full
circles, Pt/Al-SBA-15; full squares, Pt/Al
1
þ
c
H
ꢂ K
H
2
2 3
O .
k
3
ꢂ cCALD ꢂ KCALD
ð1 þ cCALD ꢂ KCALD þ cHALD ꢂ KHALD þ cCALC ꢂ KCALC þ cHALC ꢂ KHALC
Þ
Compared with the siliceous system, Pt/Al-SBA-15 exhibits a
much higher activity, as evidenced by the higher conversion of
CALD measured during the first analysis and by the completion
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
c
H
ꢂ K
H
2
ꢂ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð3Þ
1 þ
c
H
ꢂ K
H
2
Table 2
Experimental parameters for the catalytic tests in cinnamaldehyde hydrogenation.
Catalyst
Pt wt.%
Pt dispersion (%)
Amount of
catalyst (g)
Amount of
cinnamaldehyde (g)
Concentration in
Molar ratio
cinnamaldehyde/Pt
s
cinnamaldehyde (mol Lꢀ1)
3
Pt/SBA-15
Pt/Al-SBA-15
1.0
1.0
0.5
69
69
62
0.510
0.383
0.280
0.307
0.228
0.168
11 ꢁ 10ꢀ
129
142
285
ꢀ
3
3
8 ꢁ 10
Pt/Al
2
O
3
6 ꢁ 10ꢀ

2
34
S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
Table 3
0
3
ꢀ1
Apparent kinetic constants k (ꢁ10 min ) extracted from the fits of the curves presented in Fig. 4 (inside brackets: estimation of the 95% confidence interval).
1
Reaction pathway A
Reaction pathway B
0
0
0
0
k
k
k
k
1
2
3
4
CALD ? HALD (C@C)
HALD ? HALC (C@O)
CALD ? CALC (C@O)
CALC ? HALC (C@C)
Pt/SBA-15
Pt/Al-SBA-15
12 (9–15)
34 (31–37)
8 (7–9)
1 (0–2)
1 (0–2)
10 (9–11)
3 (2–4)
16 (14–18)
9 (8–10)
0
2 (0–4)
1 (0–2)
Pt/Al
2
O
3
k
4
ꢂ cCALC ꢂ KCALC
can influence the properties of metal nanoparticles of small size
(<2 nm) [34,37–39]. CO adsorption results detailed by Koningsber-
ger’s group on various types of Pt catalysts are in excellent agree-
ment with our results, whether in terms of respective positions of
the bands (at lower energy both for linear and bridging CO/Pt on
r
4
¼
ð1 þ cCALD ꢂ KCALD þ cHALD ꢂ KHALD þ cCALC ꢂ KCALC þ cHALC ꢂ KHALC
Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
c
H
ꢂ K
H
2
ꢂ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð4Þ
1
þ
c
H
ꢂ K
H
2
Several simplifications can be introduced in these kinetic equa-
2 3
Pt/Al O [39]) or in terms of shifts toward higher energies when
tions. At high H
2
pressure, the order associated with H
2
is zero.
the acidity of the support increases [34]. Concurrently, activities
in hydrogenation and hydrogenolysis reactions increase when
the acidity of the support increases [38,39]. On a more acidic sup-
port, i.e., on a support containing oxygen atoms whose formal neg-
ative charge (or ionic character) is low, the electron density of the
Pt nanoparticles is driven toward the metal – support interface.
The electron density at the surface of the metal and thus electron
back-donation toward adsorbates decrease [37,39], leading to less
strongly adsorbed molecules [39]. This results in a higher activity
of the acidic catalysts in reactions involving hydrogen. This model
applies well to two of our catalysts. On both Pt/SBA-15 and Pt/Al-
SBA-15, the hydrogenation of the C@C bond is favored, as
commonly observed for Pt catalysts, but the rate of the reactions
is significantly higher when the support is acidic.
It should be noted that some structural differences have been
observed on Pt particles dispersed on acidic, non-acidic, or basic
supports. Pt particles have been reported to flatten over the surface
of acidic supports, as a way to increase the interaction between
surface oxygens and metal atoms [34,39]. These changes are small
and detected by X-ray absorption spectroscopy only. A difference
in the particles morphology cannot be ruled out on our systems,
as suggested by the broader bands for linear CO/Pt seen on Pt/
SBA-15, which may correspond to the presence of several contribu-
tions instead of one on Pt/Al-SBA-15.
Moreover, preliminary modeling indicated that the reaction orders
in each organic molecule in the four reactions listed in Fig. 1 are
close to unity; thus, terms containing adsorption constants in the
denominators of Eqs. (1)–(4) could be neglected leading finally to:
0
0
r
r
1
3
¼ k
¼ k
1
ꢂ cCALD ꢂ KCALD ¼ k ꢂ cCALD; r
ꢂ cCALD ꢂ KCALD ¼ k ꢂ cCALD; r
2
4
¼ k
¼ k
2
ꢂ cHALD ꢂ KHALD ¼ k ꢂ cHALD
1
2
0
0
3
3
4
ꢂ cCALC ꢂ KCALC ¼ k ꢂ cCALC
4
ð5Þ
It should be noted that both on hydrophobic supports attracting
the ring of cinnamaldehyde [18,19] and for very high cinnamalde-
hyde/Pt ratios [36], the inhibiting contribution of cinnamaldehyde
p
s
should be taken into account, e.g., the term cCALD ꢂ KCALD should be
included in denominators of Eqs. (1)–(4). In contrast, molar ratios
(
cinnamaldehyde/surface Pt atoms) are close in the present work
and in the work by Toebes et al. (129–285 in this work, 74–300 for
an initial concentration in CALD of 14–34 10 mol L in [18]). Four
kinetic constants (labeled k to k ) are consequently obtained from
the fits. Values are presented in Table 3.
The apparent rate constants increase considerably between Pt/
SBA-15 and Pt/Al-SBA-15, i.e., the two catalysts on which reaction
pathway A is the major path. The acidic catalyst displays the high-
est values of the series for k and k , with k much larger than k , in
agreement with the experimental measurements of selectivities.
By comparison, the kinetic constants associated with the second-
ꢀ3
ꢀ1
0
0
1
4
0
0
0
0
1
3
1
3
In contrast, it is the selectivity of the reaction that is completely
modified in case of Al O compared to SBA-15. Kinetically speak-
2 3
ing, all the steps involving the hydrogenation of the C@O bond
are on a par with that involving the initial hydrogenation of the
C@C bond. Four main differences exist between Pt/Al O and Pt/
2 3
SBA-15:
ary reactions are quite low. Pt/Al
alysts in two aspects: (i) constants k and k have similar values, in
line with the selectivities in HALD and CALC up to 40 min of reac-
tion; (ii) constant k corresponding to the reduction in the C@O
2
O
3
differs from the two other cat-
0
0
1
3
0
2
0
bond in HALD is almost equal to k , corresponding to the reduction
3
in the C@O bond in CALD. On this catalyst, the hydrogenation of
(
i) the presence of smaller pore apertures in alumina. However,
cinnamaldehyde, a rigid molecule, has been shown to diffuse
in micropores as well as mesopores [14].
the C@O bond proceeds at the same rate whatever the organic sub-
0
strate. In contrast, k (hydrogenation of the C@C bond in cinnamyl
4
alcohol) remains very low.
(
ii) a slightly lower dispersion. Cinnamaldehyde hydrogenation
is a reaction sensitive to the particles size [40]. However,
except in case of gold catalysts [15], the literature usually
concludes that no significant effect of the particles size is
expected on cinnamaldehyde hydrogenation in the 1–2 nm
range [17].
4
. Discussion
The results presented earlier show that the hydrogenation of
cinnamaldehyde takes place differently on Pt catalysts prepared
on mesoporous acidic supports, compared with non-acidic SBA-
(iii) the energy of the CO/Pt band, which is lower in case of Pt/
Al , denoting a more important back donation from the
1
5 that we will consider as the reference system. Since the charac-
2 3
O
teristics of the catalysts are very close in terms of porosity and Pt
dispersion, the differences in catalytic behavior should be attrib-
uted to the effect of the support composition and/or acidity,
though, apparently, cinnamaldehyde hydrogenation does not re-
quire acidic sites to be performed.
metal toward the adsorbates. This observation can be
explained assuming either:
ꢃ
specific morphologies for the Pt particles. Hydrogenation
of the C@O bond is considered to be favored on particles
facets, which repel the electron-rich aromatic ring and
C@C bond of the molecule [41–43]. However, the ratio
between the quantities of linear CO and bridging CO is
X-ray absorption and infrared spectroscopies correlated with
catalytic data have provided some insight into the way supports

S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
235
about the same for Pt/Al
ficult to sustain that particles would be considerably
more faceted on the former.
2
O
3
and Pt/SBA-15, making it dif-
(ii) by providing adsorption sites to the organic substrates,
allowing them to interact with the catalyst via their polar
C@O bond, as suggested for the Lewis sites on the surface
ꢃ
or a perturbation of the Pt nanoparticles by the support
underneath. A higher selectivity in cinnamyl alcohol has
been reported on particles whose surface was enriched
in electrons, most often on basic supports [12,44]. An
increase in the particles surface electron density is evi-
denced here by the stronger interaction of CO with the
metal. One might infer that this favors the desorption of
molecules adsorbed via the electron-rich AOH function,
considered to be the rate-determining step for the pro-
duction of alcohols [8].
2 3
of Al O . An enrichment of the particles surface electron
density, deriving from the ionic character of alumina, might
also favor the production of alcohols. Consequently, on alu-
mina, the hydrogenation of cinnamaldehyde occurs through
two parallel reaction pathways at the same rate, yielding
cinnamyl alcohol on the one hand and CAC saturated mole-
cules on the other hand.
Although cinnamaldehyde hydrogenation does not require
acidic sites, the composition and acidity of the oxidic support can
thus significantly influence the reaction rate or the selectivity of
the process when Pt nanoparticles are small enough.
(
iv) the presence of numerous weak Lewis sites all over the sup-
port, including in the vicinity of the Pt nanoparticles, which
would be involved in the adsorption of CALD and HALD via
their polar C@O function [29,45]. This hypothesis has been
also proposed in the literature to account for the selectivity
of catalysts presenting a SMSI effect (partial covering of the
metallic surface by oxidic species [16,17]) or by a second
metallic element in an ionic state such as tin [10]. If this
hypothesis is correct, the activity of a Pt catalyst can be
strongly modified by a direct intervention of the support
acidic groups, here Lewis sites acting on the adsorption of
the organic substrates.
Acknowledgments
This work was part of the project Nanocat (No. 506621) funded
through the European 6th Framework Program. The authors ex-
press their gratitude to Patricia Beaunier for her help in TEM
measurements.
References
[
1] G. Neri, L. Bonaccorsi, L. Mercadante, S. Galvagno, Ind. Eng. Chem. Res. 36
(1997) 3554.
It can be noted that the selectivity in CALC measured on our Pt/
Al
ously published by other groups on alumina-supported systems.
2 3
Actually, with respect to cinnamaldehyde hydrogenation, Pt/Al O
2
O
3
catalyst is quite unexpected, when compared to results previ-
[2] Y. Zhu, G.K. Chuah, S. Jaenicke, J. Catal. 241 (2006) 25.
[
[
3] G.K. Chuah, S. Jaenicke, Y.Z. Zhu, S.H. Liu, Curr. Org. Chem. 10 (2006) 1639.
4] J. Hájek, N. Kumar, P. Mäki-Arvela, T. Salmi, D.Yu. Murzin, I. Paseka, T. Heikkilä,
E. Laine, P. Laukkanen, J. Väyrynen, Appl. Catal. A 251 (2003) 385.
catalystsreported intheliterature canbedivided intotwo categories
as follows: catalysts that are selective in CALC because Pt particles
are large and faceted (4–13 nm) [41,46]; catalysts on which Pt parti-
cles are well dispersed and behave poorly in CALC production
[5] J. Hájek, N. Kumar, T. Salmi, D.Yu. Murzin, Catal. Today 100 (2005) 349.
[
[
[
6] P. Gallezot, D. Richard, Catal. Rev. Sci. Eng. 40 (1998) 81.
7] P. Mäki-Arvela, J. Hájek, T. Salmi, D.Yu. Murzin, Appl. Catal. A 292 (2005) 1.
8] D. Loffreda, F. Delbecq, F. Vigné, P. Sautet, J. Am. Chem. Soc. 128 (2006) 1316.
[9] S. Galvagno, A. Donato, G. Neri, R. Pietropaolo, D. Pietropaolo, J. Mol. Catal. A 49
1989) 223.
(
[
47,48]. In this latter category, one can observe that the aluminas
[
10] A.J. Plomp, D.M.P. van Asten, A.M.J. van der Eerden, P. Mäki-Arvela, D.Yu.
Murzin, K.P. de Jong, J.H. Bitter, J. Catal. 263 (2009) 146.
[11] J.M. Ramallo-López, G.F. Santori, L. Giovanetti, M.L. Casella, O.S. Ferretti, F.G.
Requejo, J. Phys. Chem. B 107 (2003) 11441.
12] G. Li, T. Li, Y. Xu, Chem. Commun. (1996) 497.
13] L. Mercadante, G. Neri, C. Milone, A. Donato, S. Galvagno, J. Mol. Catal. A 105
(1996) 93.
[14] M. Lashdaf, V.V. Nieminen, M. Tiita, T. Venäläinen, H. Österholm, O. Krause,
Micropor. Mesopor. Mater. 75 (2004) 159.
15] E. Bus, R. Prins, J.A. van Bokhoven, Catal. Commun. 8 (2007) 1397.
16] A.B. Boffa, C. Lin, A.T. Bell, G.A. Somorjai, Catal. Lett. 27 (1994) 243.
[17] M. Englisch, A. Jentys, J.A. Lercher, J. Catal. 166 (1997) 25.
18] M.L. Toebes, T.A. Nijhuis, J. Hájek, J.H. Bitter, A.J. van Dillen, D.Yu. Murzin, K.P.
de Jong, Chem. Eng. Sci. 60 (2005) 5682.
19] A.J. Plomp, H. Vuori, A.O.I. Krause, K.P. de Jong, J.H. Bitter, Appl. Catal. A 351
employed are well crystallized, since their specific surface area is
2
ꢀ1
low (100–130 m g ), and presumably exhibita low surface density
of Lewis sites. This observation leads us to favor an explanation
based on the action of Lewis sites, rather than on the enhanced sur-
face electron density of Pt nanoparticles. The benefit of using a mes-
oporous alumina as a support would thus be dual: increasing the
catalyst activity by bringing a high specific surface area and thus
promoting a high dispersion of Pt nanoparticles and increasing the
catalyst selectivity in CALC by supplying Lewis acid sites that favor
the adsorption of CALD via the aldehyde function.
[
[
[
[
[
[
(
2008) 9.
[
[
20] P. Gallezot, A. Giroir-Fendler, D. Richard, Catal. Lett. 5 (1990) 169.
21] J. Hájek, N. Kumar, D. Francová, I. Paseka, P. Mäki-Arvela, T. Salmi, D.Yu.
Murzin, Chem. Eng. Technol. 27 (2004) 1290.
5
. Conclusions
In the present paper, three Pt catalysts exhibiting similar pore
[22] M. Chatterjee, Y. Ikushima, T. Yokohama, T. Suzuki, Micropor. Mesopor. Mater.
17 (2009) 201.
23] D. Zhao, Q. Huo, J. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120 (1998)
025.
[24] R. Mokaya, Angew. Chem., Int. Edit. Engl. 38 (1999) 2930.
1
size and metal dispersion, but prepared on mesoporous supports
differing by their chemical composition and acidic properties, were
tested in the liquid phase hydrogenation of cinnamaldehyde in 2-
pentanol. When cinnamaldehyde hydrogenation is performed on
Pt catalysts exhibiting particles size smaller than 2 nm, the support
can influence the catalytic properties of the system in two ways:
[
6
[
[
[
[
25] S. Handjani, S. Dzwigaj, J. Blanchard, E. Marceau, J.M. Krafft, M. Che, Top. Catal.
2 (2009) 334.
26] N. Bejenaru, C. Lancelot, P. Blanchard, C. Lamonier, L. Rouleau, E. Payen, F.
Dumeignil, S. Royer, Chem. Mater. 21 (2009) 522.
27] B. Xu, T. Xiao, Z. Yan, X. Sun, J. Sloan, S.L. Gonzáles-Cortés, F. Alshahrani, M.L.H.
Green, Micropor. Mesopor. Mater. 91 (2006) 293.
28] S. Handjani, J. Blanchard, E. Marceau, P. Beaunier, M. Che, Micropor. Mesopor.
Mater. 116 (2008) 14.
5
(
i) by modifying the electronic properties of the nanoparticles
or perhaps their morphology. This is evidenced for mesopor-
ous silicoaluminic supports exhibiting Brønsted and Lewis
acidities (Al-SBA-15), on which Pt particles display a lower
surface electron density than their homologs on siliceous
SBA-15. In both cases, the hydrogenation of the C@C bond
is favored, but rate constants increase dramatically when
Al is introduced on the support. On Al-SBA-15, the major
final product of reaction is the saturated alcohol.
[29] F. Letellier, J. Blanchard, K. Fajerwerg, C. Louis, M. Breysse, D. Guillaume, D.
Uzio, Catal. Lett. 110 (2006) 115.
[
[
30] K. Hadjiivanov, H. Knözinger, M. Mihaylov, J. Phys. Chem. B 106 (2002) 2618.
31] W. Daniell, U. Schubert, R. Glöckler, A. Meyer, K. Noweck, H. Knözinger, Appl.
Catal. A 196 (2000) 247.
[
[
32] G. Crépeau, F. Montouillout, A. Vimont, L. Mariey, T. Cseri, F. Maugé, J. Phys.
Chem. B 110 (2006) 15172.
33] C. Chizallet, P. Raybaud, ChemPhysChem 11 (2009) 105.
[34] B.L. Mojet, J.T. Miller, D.C. Koningsberger, J. Phys. Chem. B 103 (1999) 2724.

2
36
S. Handjani et al. / Journal of Catalysis 282 (2011) 228–236
[
35] C.Y. Hou, T.C. Chiu, M.H. Shih, W.J. Tsai, W.Y. Chen, C.H. Lin, J. Phys. Chem. C
14 (2010) 4502.
[42] J.C. Serrano-Ruiz, A. López-Cudero, J. Solla-Gullón, A. Sepúlveda-Escribano, A.
Aldaz, F. Rodríguez-Reinoso, J. Catal. 253 (2008) 159.
1
[
[
36] T. Vergunst, F. Kapteijn, J.A. Moulijn, Catal. Today 66 (2001) 381.
37] B.L. Mojet, J.T. Miller, D.E. Ramaker, D.C. Koningsberger, J. Catal. 186 (1999)
[43] E.V. Ramos-Fernández, J.M. Ramos-Fernández, M. Martínez-Escandell, A.
Sepúlveda-Escribano, F. Rodríguez-Reinoso, Catal. Lett. 133 (2009) 267.
[44] E.V. Ramos-Fernández, A.F.P. Ferreira, A. Sepúlveda-Escribano, F. Kapteijn, F.
Rodríguez-Reinoso, J. Catal. 258 (2008) 52.
[45] G. Szöllösi, B. Török, L. Baranyi, M. Bartók, J. Catal. 179 (1998) 619.
[46] G. Szöllösi, B. Török, G. Szakonyi, I. Kun, M. Bartók, Appl. Catal. A 172 (1998) 225.
[47] M. Arai, K. Usui, Y. Nishiyama, J. Chem. Soc., Chem. Commun. (1993) 1853.
[48] M. Lashdaf, J. Lahtinen, M. Lindblad, T. Venäläinen, A.O.I. Krause, Appl. Catal. A
276 (2004) 129.
3
73.
38] D.E. Ramaker, J. de Graaf, J.A.R. van Veen, D.C. Koningsberger, J. Catal. 203
2001) 7.
[
[
(
39] A.Yu. Stakheev, Y. Zhang, A.V. Ivanov, G.N. Baeva, D.E. Ramaker, D.C.
Koningsberger, J. Phys. Chem. C 111 (2007) 3938.
40] M. Che, C.O. Bennett, Adv. Catal. 36 (1989) 55.
[
[
41] A. Giroir-Fendler, D. Richard, P. Gallezot, Catal. Lett. 5 (1990) 175.