Enhancing the catalytic performance of Pt/ZnO in the selective hydrogenation of cinnamaldehyde by Cr addition to the support

Article abstract of DOI:10.1016/j.jcat.2008.05.025

An enhancement of the catalytic performance of platinum in the liquid phase hydrogenation of cinnamaldehyde has been achieved by modifying a ZnO support by addition of Cr(III) cations. The addition of chromium improves the structural properties of the support (larger surface area) and increases its reducibility, and resulted in a better Pt dispersion. The presence of chromium in the support enhanced the overall catalytic activity, and improved the high selectivity towards the unsaturated alcohol. The increase of the reduction temperature from 473 K to 623 K produces an impressive decrease in the turnover frequency for Pt/ZnO, whereas this value remains practically unmodified for Pt/Cr-Zn. The higher reduction temperature reduces the overall activity in both catalysts, but improves selectivity towards the unsaturated alcohol. The present 5 wt% Pt/Cr-ZnO catalyst showed the best results (in terms of selectivity) published until now without use of reaction promoters. The improved performance of this catalyst is ascribed to an adequate metal-support interaction and to the higher reducibility of the support that opens the possibility of alloy formation.

Full text of DOI:10.1016/j.jcat.2008.05.025

Journal of Catalysis 258 (2008) 52–60
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Enhancing the catalytic performance of Pt/ZnO in the selective hydrogenation
of cinnamaldehyde by Cr addition to the support
E.V. Ramos-Fernández a, A.F.P. Ferreira b, A. Sepúlveda-Escribano a,∗, F. Kapteijn b, F. Rodríguez-Reinoso a
a
Departamento de Química Inorgánica, Universidad de Alicante, Apartado 99, E-03080 Alicante, Spain
Catalysis Engineering, DelftChemTech, Delft University of Technology, Julianalaan, 136, 2628 BL Delft, Netherlands
b
a r t i c l e i n f o
a b s t r a c t
Article history:
An enhancement of the catalytic performance of platinum in the liquid phase hydrogenation of
cinnamaldehyde has been achieved by modifying a ZnO support by addition of Cr(III) cations. The
addition of chromium improves the structural properties of the support (larger surface area) and
increases its reducibility, and resulted in a better Pt dispersion. The presence of chromium in the
support enhanced the overall catalytic activity, and improved the high selectivity towards the unsaturated
alcohol. The increase of the reduction temperature from 473 K to 623 K produces an impressive decrease
in the turnover frequency for Pt/ZnO, whereas this value remains practically unmodified for Pt/Cr–Zn.
The higher reduction temperature reduces the overall activity in both catalysts, but improves selectivity
towards the unsaturated alcohol. The present 5 wt% Pt/Cr–ZnO catalyst showed the best results (in terms
of selectivity) published until now without use of reaction promoters. The improved performance of this
catalyst is ascribed to an adequate metal–support interaction and to the higher reducibility of the support
that opens the possibility of alloy formation.
Received 22 January 2008
Revised 8 April 2008
Accepted 29 May 2008
Available online 1 July 2008
Keywords:
Cinnamaldehyde hydrogenation
Pt
Cr–ZnO
SMSI
© 2008 Elsevier Inc. All rights reserved.
1. Introduction
is well known that the addition of a second, more electropositive
metal, like iron [9], tin [10–12] or zinc [13], to supported platinum
enhances it selectivity. This has been explained on the basis of al-
loy formation, which results in the weakening of the adsorption
through the olefinic bond and in the presence of acid Lewis cen-
tres (the more electropositive metal atoms) that interact with the
oxygen atom in the carbonyl bond and favours its hydrogenation
[11,14].
The search for chemoselective catalysts is an issue of great in-
terest for the production of many pharmaceutical, agrochemicals,
and fragrance compounds [1–3]. Thus, the hydrogenation of or-
ganic substrates containing several unsaturated functional groups
has been attracting a lot of interest for fundamental research in
catalysis. The selective hydrogenation of the carbonyl bond in α,β-
unsaturated aldehydes with supported metal catalysts is still a
challenge in heterogeneous catalyst. The selective reduction can be
achieved by means of properly designed organometallic catalysts,
through a Meerwein–Ponndorf–Verley reaction with alcohols as re-
ducing agents and solid Lewis acids as catalysts [4–6]. However,
the selective hydrogenation of the carbonyl bond in the presence
of an olefinic bond is not an easy task with metallic catalysts,
as the C=C bond hydrogenation is favoured both from thermody-
namic and kinetics considerations [7,8]. Therefore, it is necessary
to modify the intrinsic behaviour of the metal in order to increase
the selectivity toward the formation of the desired unsaturated al-
cohol.
Also, the choice of the support can play an important role.
The support can affect the catalytic behaviour of the noble metal
through several mechanisms. Thus, it has been shown that the
selectivity towards the hydrogenation of the carbonyl bond in cin-
namaldehyde can be increased by encapsulating Pt particles in
Y-type zeolite [15,16]. In this case, it has been claimed that molec-
ular constraints in the zeolite micropores favour the adsorption of
the cinnamaldehyde molecule on the metal particles through the
carbonyl bond, this facilitating its hydrogenation. However, diffu-
sional limitations induced by the microporous structure yielded
low reaction rates [16]. Electronic effects due to the support can
also be used. In this way, Giroir-Fendler et al. [17] compared the
behaviour of group VIII noble metals supported on activated car-
bon and graphite in the hydrogenation of cinnamaldehyde, and
observed that the graphite-supported systems were more selec-
tive, under the same experimental conditions. This higher selec-
tivity was explained on the basis of an electron transfer from the
graphitic support to the metal particles located at the edges of the
basal planes. The increased charge density on the metal particles
The metal promotion can be achieved by weakening the C=O
bond and/or by avoiding the adsorption through C=C bond, and
there are several ways of obtaining one or both of these effects. It
Corresponding author. Fax: +34 965 90 34 54.
E-mail address: asepul@ua.es (A. Sepúlveda-Escribano).
*
0021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.05.025

E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
53
would decrease the probability of adsorption via the C=C bond,
thus hindering its hydrogenation. More recently, several studies
have dealt with the promotion of platinum particles by graphitic
supports such as carbon nanotubes and nanofibres, for which the
promotion effect would be expected to be higher. However, the
selectivity values obtained up to now have been relatively low
[18,19].
with an INCA Energy TEM 100 analytical system and a SIS Mega-
View II camera. The samples reduced at 623 K were suspended
in methanol and placed on copper grids with a holey-carbon film
support.
Temperature-programmed reduction (TPR) measurements were
carried out in a U-shaped quartz reactor, using a 5% H2/He gas
flow of 50 cm3/min and 100 mg of catalyst, with a heating rate
of 10 K/min. The hydrogen consumption was monitored by mass
spectrometry.
Other interesting supports for achieving promotion in this kind
of reactions are partially reducible oxides such as TiO2 [20–22],
ZnO [13,14,23–25] CeO2 [26–29] or SnO2 [30,31]. In these cases,
the possible interaction of the noble metal with the partially re-
duced oxide (SMSI: strong metal–support interaction effect) or
even with the metal formed upon a reduction treatment, has been
proposed to be responsible for the improved selectivity. In the case
of zinc oxide, both effects can play a role. When platinum is sup-
ported on ZnO, the onset of electronic and ensemble effects is
well known [14,23]. The electronic effects are due to free electrons
produced upon the partial reduction of ZnO to Zn, these free elec-
trons being donated to Pt [32]. The resulting increase in electron
density of Pt enhances its catalytic properties, in terms of selec-
tivity, because the probability of adsorption via the C=C bond is
decreased [23]. The reducibility of ZnO can be improved through
valence induction by doping the oxide support with cations having
a formal charge larger than +2. The addition to the support of an
appropriate doping cation in the proper amount can also increase
the BET surface area by suppressing the ZnO crystal growth.
This paper reports the effect of the addition of Cr(III) to ZnO on
the catalytic behaviour of supported platinum in the liquid phase
selective hydrogenation of cinnamaldehyde.
X-ray photoelectron spectra (XPS) were acquired with a VG-
Microtech Multilab 3000 spectrometer equipped with a hemi-
spherical electron analyser and a MgKα (h = 1253.6 eV, 1 eV =
−19
1.6302 × 10
J) 300-W X-ray source. The powder samples were
pressed into small Inox cylinders and then mounted on a sam-
ple rod placed in a pretreatment chamber and reduced in flow-
ing H2 for 1 h at 473 and 623 K before being transferred to the
analysis chamber. Before recording the spectra, the sample was
maintained in the analysis chamber until a residual pressure of ca.
−7
5 × 10
N/m2 was reached. The spectra were collected at pass
energy of 50 eV. The intensities were estimated by calculating
the integral of each peak, after subtraction of the S-shaped back-
ground, and by fitting the experimental curve to a combination
of Lorentzian (30%) and Gaussian (70%) lines. All binding energies
(BE) were referenced to the C 1s line at 284.6 eV, which provided
binding energy values with an accuracy of 0.2 eV. The surface
atomic ratios were estimated from the integrated intensities cor-
rected by the atomic sensitivity factors [33].
The CO chemisorption experiments were carried out in a mano-
metric equipment. Before CO chemisorption, the samples were re-
duced at 473 K and 623 K under flowing hydrogen for 1 h, and
then cleaned up with a helium flow (50 ml/min) for 1 h at the
same temperature. Chemisorption data were collected by sequen-
tially introducing small doses (1–10 μmol) of CO (99.5%, with fur-
ther purification, Air Liquide) onto the sample at 298 K until it
became saturated. The amount of gas adsorbed (μmol) was deter-
mined manometrically from the dosis, equilibrium pressures, and
the system’s volumes and temperatures. The time required for the
pressure to equilibrate in the cell after each dosis was approxi-
mately 40 min.
2. Experimental
2.1. Catalysts preparation
Two catalyst supports were prepared by a homogeneous co-
precipitation method: pure ZnO (SBET: 5 m2/g) and Cr–ZnO (Cr/Zn
atomic ratio of 0.12, SBET: 10 m2/g). An aqueous solution (pH 9)
containing Zn(NO3)2·H2O, CO(NH2)2 and, for the Cr–ZnO support
also Cr(NO3)3·9H2O, was strongly stirred and then heated up to
365 K. The precipitate formed was centrifuged and calcined in air
at 773 K for 2 h with a heating rate of 10 K/min. For catalyst
preparation, the supports were immersed into an aqueous solution
of H2PtCl6·6H2O of the adequate concentration as to achieve a Pt
loading of 5 wt%, and stirred for 24 h. Then, the excess of solvent
was removed by evaporation in a rotary evaporator at 363 K and
the obtained solids were calcined in air at 773 K for 2 h with a
heating rate of 10 K/min. The catalysts obtained were labelled as
Pt/ZnO and Pt/Cr–ZnO.
2.4. Catalytic tests
The liquid phase hydrogenation of cinnamaldehyde was carried
out in a 300 ml stirred autoclave (Biometa, fitted with a system
for liquid sampling) at 383 K and at constant pressure of 7 MPa.
Pre-reduced catalysts (300 mg) were immersed into 90 ml of sol-
vent (isopropanol 99% previously out-gassed by bubbling a flow
of helium). After a first flush with helium and a second one with
hydrogen, the temperature was raised to 383 K under 3 MPa of
hydrogen to re-active the catalysts. Then, a mixture of substrate
(0.5 ml of cinnamaldehyde) and solvent (10 ml) was loaded into
the autoclave through a liquid container under a 7 MPa hydrogen
pressure. The final cinnamaldehyde concentration was 0.04 mol/l.
Zero time was taken at this moment and stirring was switched on.
Liquid samples were analysed on a gas chromatograph equipped
with a FID detector and a capillary Carbowax column, using he-
lium as carrier gas.
2.2. Supports characterisation
The textural properties of the prepared materials were deter-
mined by nitrogen adsorption at 77 K on a Coulter Omnisorp 610
system. Before measurements, the samples were dried at 383 K
for 12 h and out-gassed at 523 K under vacuum for 4 h. Their sur-
face areas were calculated following the BET method in the relative
pressure range from 0.05 to 0.25.
The X-ray diffraction patterns of the supports were obtained
with a JSO Debye-Flex 2002 system, from Seifert, fitted with a
◦
3. Results
Cu cathode and a Ni filter, and using a 2 /min scanning rate. The
room temperature Raman spectra were recorded on a Raman-laser
spectrometer using a 514.5 nm line of an argon-ion laser.
3.1. XRD analysis
Fig. 1 shows the XRD patterns of the ZnO and Cr–ZnO supports.
For the pure zinc oxide support all the diffraction peaks in the pat-
tern can be assigned to the hexagonal (würtzite-type) ZnO, with
lattice constants a = 3.25 Å and c = 5.20 Å. The comparison of
2.3. Catalysts characterisation
Conventional TEM analysis was carried out with a JOEL model
JEM-210 electron microscope working at 200 kV and equipped

54
E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
Fig. 1. XRD patterns of the supports.
Fig. 2. Raman spectra of the supports.
the observed and standard intensities [34] of the diffraction peaks
indicates that there are not preferred orientations. The diffrac-
tion peaks are narrow and intense, indicating a high crystallinity.
Two crystalline structures can be observed in the XRD pattern of
the Cr–ZnO support: hexagonal ZnO and, to a lesser extent, spinel
ZnCr2O4. The crystallinity of the former is poorer for this support,
and the wider diffraction peaks indicate a smaller crystallite size.
The presence of crystalline Cr2O3 was not detected.
ion radius is very similar to that of Zn(II), so in principle, the
shift should be lower. However, the amount of chromium added
is very high and this produces a large distortion in the ZnO struc-
−1
ture. The broad features between 1025–1150 cm
are assigned
to the two-phonon modes (2-LO), characteristic of this II–IV semi-
conductor [36]. The spectra for the Cr–ZnO support showed the
Raman active modes of würtzite-type structure and a peak centred
at 845 cm−1, which is ascribed to the ZnCr2O4 spinel structure.
These findings corroborate those obtained by XRD diffraction, i.e.,
the presence of a ZnCr2O4 spinel structure. The decrease in the E2
phonon frequency demonstrates that the chromium ions are taken
part in the ZnO structure forming zinc chromite. No indications of
the Cr2O3 phase were found.
3.2. Raman spectroscopy
Raman spectra of the Cr–ZnO and ZnO supports are collected in
Fig. 2. The spectra obtained for the pure ZnO support shows several
bands, which are assigned to the Raman active mode of the ma-
3.3. Transmission electron microscopy
terial with würtzite-type structure (ZnO). Thus, the Raman E2(high)
,
A1(LO) and A1(TO) phonon modes at 430, 569, and 383 cm−1, re-
spectively, were clearly observed, this indicating that the ZnO crys-
Fig. 3 shows TEM micrographs obtained for the two catalysts re-
duced at 623 K. Dark zones are due to platinum particles, whereas
lighter parts correspond to the support. Homogeneous particle size
distributions have been found in both catalysts, being especially
narrower for catalyst Pt/Cr–ZnO. Analysis of a set of high mag-
nification images allowed to obtain the particle size distributions
presented in Fig. 4. The particle size distribution for the Pt/ZnO
catalyst is centred at 4.1 1.2 nm, whereas for Pt/Cr–ZnO it is
centred at 1.7 0.5 nm. The addition of Cr generated a support
with larger surface area, this resulting in a better platinum disper-
sion. The low surface area of ZnO and the relative high platinum
loading (5 wt%) caused larger platinum particles to be formed on
Pt/ZnO. Catalyst reduced at 473 K showed the same platinum par-
−1
tals did not show any preferred orientation. The band at 320 cm
is the second order Raman spectrum arising from zone-boundary
phonons 2-E2 of ZnO [35]. In würtzite structure crystals, stress in-
duced in the crystals affects the E2 phonon frequency. An increase
in the E2 phonon frequency is ascribed to compressive stress,
whereas a decrease in the E2 phonon frequency is ascribed to ten-
sile stress [35]. The addition to the structure of a second ion with
a different atomic radius is considered to be the main factor that
would cause lattice distortion in the crystals. It can be seen that
after chromium addition, the E2 phonon frequency is shifted to
−1
lower wavenumbers (from 430 cm
large tensile stress remaining in the ZnO crystals. The chromium
to 425 cm−1), indicating a

E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
55
Fig. 3. TEM pictures of catalysts reduced at 623 K: (A) Pt/Cr–ZnO; (B) Pt/ZnO.
Fig. 4. Size distribution of Pt particles on catalysts reduced at 623 K: (A) Pt/Cr–ZnO, (B) Pt/ZnO.
Fig. 5. Pt 4f core-level photoelectron spectra of the Pt/Cr–ZnO catalyst (A) and Pt/ZnO catalysts (B) after reduction and calcination treatments.
ticle size distribution, this indicating that the reduction treatment
at 623 K did not result in any sintering of the metal particles.
pears at 70.5 eV, which is characteristic of metallic platinum. The
same behaviour was found for catalyst Pt/Cr–ZnO. These results in-
dicate that Pt species are fully reduced even after the reduction
treatment at low temperature (473 K).
3.4. X-ray phoelectron spectroscopy
The O 1s spectrum (not shown) for catalysts Pt/ZnO and Pt/Cr–
ZnO both calcined (un-reduced) and after reduction, showed the
presence of two oxygen components at BE 530 eV and 531.5 eV.
The first peak, centred at 530 eV, corresponds to oxygen taking
part in the Zn–O bond, and the second one is ascribed to –OH
surface groups [40]. The intensity of the peak at 531.5 eV decreased
after the reduction treatments, thus indicating the partial removal
of the –OH groups.
The Pt 4f level X-ray photoelectron spectra for catalyst Pt/ZnO
after reduction in flowing hydrogen at 473 K and 623 K are shown
in Fig. 5. The calcined unreduced sample presents two peaks at
73.8 eV (level Pt 4f7/2) and 77.1 eV (level Pt 4f5/2). Both peaks are
unambiguously ascribed to electron deficient Pt(II) species [37–39].
Both peaks shifted to lower binding energies (BE) after the reduc-
tion treatment. The peak corresponding to Pt 4f7/2 level now ap-

56
E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
Fig. 6. Zn 2p3/2 core-level XPS spectra of the Pt/Cr–ZnO and Pt/ZnO after calcination and reduction treatments.
Fig. 7. Zn LMM Auger transition for Pt/Cr–ZnO after calcination and reduction treatments.
The BE of Zn 2p3/2 transition does not permit to discern the
can also play its role [41]. However, it was not possible to distin-
guish between zinc species forming part of ZnO or Zn(OH)2 and
that forming ZnCr2O4.
oxidation state of zinc. In the catalysts studied in this work, a sin-
gle band appears centred at 1021.7 eV, which could be assigned to
Zn(II) species (Fig. 6). The band is quite broad, this indicating the
presence of zinc cations in different environments (ZnO, Zn(OH)2,
etc.). This assertion is consistent with the XPS O 1s data discussed
above. The XPS Zn 2p spectra for the Pt/ZnO catalyst obtained af-
ter the reduction treatments are also shown in Fig. 6. Reduction at
473 K and at 623 K produced an insignificant shift in binding en-
ergy. However, the band widths are lower than that of the calcined
sample, this probably being due to a lower influence by the Zn hy-
droxide feature. The XPS Zn 2p spectrum of calcined un-reduced
Pt/Cr–ZnO showed a peak centred at 1021.7 eV, which is wider
than that obtained with the Pt/ZnO catalyst. This indicates that the
peak broadness is not only due to the presence of zinc hydroxide
but, in this case, the Zn:Cr spinel structure located on the surface
The difficulty of assessing the oxidation state of zinc by XPS
measurements is well recognized in the literature, as the binding
energies of metallic and oxidised Zn species are very close. In this
case, it is necessary to take into account the analysis of the Zn
LMM Auger transition, which is more sensitive to oxidation state
of zinc [42]. For both the Pt/ZnO and the Pt/Cr–ZnO catalysts, the
Auger spectra showed the absence of metallic zinc in the calcined
un-reduced catalysts (band centred at 988.1 eV, assigned to Zn(II)
species). The reduction treatments resulted in a shoulder at a ki-
netic energy of 991.7 eV, attributed to metallic Zn (as an example,
Fig. 7 shows the Zn LMM Auger [13,25] transition of catalyst Pt/Cr–
ZnO after the different treatments). The XPS spectrum in the Cr
2p3/2 region (not shown) showed one peak at 576 eV, which is as-
signed to Cr(III) species [41,43–45].

E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
57
Table 1
Table 2
XPS characterisation of catalysts
CO chemisorption data of catalysts reduced at 473 K and 623 K
Reduction
temperature (K)
Pt/(Cr + Zn)
Cr/Zn
(at)
Reduction
temperature (K)
CO uptake
(μmol/g)
CO/Pt
(at)
Pt/Cr,ZnO
Pt/ZnO
Calcined
Red: 473 K
Red: 623 K
0.160
0.120
0.131
0.082
0.100
0.160
Pt/Cr–ZnO
Pt/ZnO
Red: 473 K
Red: 623 K
4.2
2.5
0.0160
0.0100
Red: 473 K
Red: 623 K
0.25
0.01
0.0010
0.0004
Calcined
Red: 473 K
Red: 623 K
0.040
0.050
0.072
–
–
–
Fig. 9. Cinnamaldehyde hydrogenation reaction network.
3.6. CO chemisorption
The CO uptakes and the CO/Pt ratios obtained by CO chemisorp-
tion at room temperature after reduction at 473 and 623 K are
reported in Table 2. The CO/Pt atomic ratios were very low for
both catalysts after the two reduction treatments, in spite of the
fact that the mean Pt particle size measured by TEM was very
low. This is especially significant for catalyst Pt/ZnO. The CO uptake
decreased after the reduction treatments, this being attributed to
the influence of a strong metal–support interaction effect, as TEM
studies revealed that no sintering of Pt particles had occurred. CO
chemisorption should therefore not be used to calculate Pt particle
sizes. Instead, it only will reflect the number of accessible Pt sur-
face sites. A strong interaction of the metal with the support, the
formation of alloy phase(s) between Pt and metallic Zn, and even
the decoration of Pt particles by patches of support after reduction
can be the origin of the low CO chemisorption capacity in these
catalysts. In this way, in spite of the small size of the Pt particles
present in the prepared catalysts, the amount of Pt atoms exposed
seems to be very low.
Fig. 8. TPR profiles of the Pt/Cr,ZnO and Pt/ZnO catalysts.
The XPS atomic ratios between the different components in the
catalysts are shown in Table 1. In principle, the Pt/Zn ratio for the
Pt/ZnO catalyst and the Pt/(Cr + Zn) ratio for the Pt/Cr–ZnO cat-
alyst can be indicative of platinum dispersion [37,38,46,47]. The
Pt/Zn XPS ratio in Pt/ZnO indicates a lower amount of surface Pt,
i.e., larger particles, than the Pt/Cr–ZnO catalyst for both calcined
and reduced catalysts (Table 1). These results agree with those
from TEM analysis (Figs. 3 and 4). On the other hand, the atomic
XPS Cr/Zn ratios in Pt/Cr–Zn increased with increasing the reduc-
tion temperature. This demonstrates that the reduction treatment
produced segregation of chromium species towards the surface.
This result has been previously reported in other studies carried
out by Raman and XPS [41,43,45].
3.7. Selective hydrogenation of cinnamaldehyde
3.5. Temperature-programmed reduction
Hydrogenation of cinnamaldehyde typically leads to three main
reaction products, among which the desired one is cinnamylalco-
hol. The reaction scheme shown in Fig. 9 is rather simplified. Other
by-products may appear, depending on the reaction conditions. For
instance, the formation of acetals is typical for hydrogenations car-
ried out over acidic catalysts. In accordance with Gallezot et al. [2],
who reported solely the three main reaction products (B, C, D),
there was no clear indication for other reactions. In the case of the
catalysts studied in this work, the main reaction product found
was cinnamylalcohol (unsaturated alcohol). Only a minor amount
(about 5%) of other products was detected.
Fig. 10A shows the cinnamaldehyde concentration as function
of time for the catalysts reduced at 473 K. Cinnamaldehyde was
completely converted over both catalysts in less than 500 min,
the Pt/Cr–ZnO catalyst being more active than Pt/ZnO. Fig. 10A
also reports the initial rates, expressed in terms of a first or-
der rate constant (min−1). Fig. 10B shows the integral selectiv-
ity towards cinnamylalcohol (the amount of unsaturated alcohol
present relative to the total amount of aldehyde converted) as
function of the conversion of cinnamaldehyde, for the catalysts re-
duced at 473 K. Both samples catalysed selectively the formation
Fig. 8 shows the temperature-programmed reduction (TPR) pro-
files (H2-consumption) for both catalysts. Both TPR profiles exhibit
three peaks centred at 400, 435 and one between 440–480 K. The
first hydrogen consumption at low temperature (400 K) is assigned
to the reduction of oxidised Pt species to metallic Pt. The second
reduction peak at 435 K is assigned to the breakdown of Pt–O–Zn
entities formed upon calcination and reduction of oxy-chlorinated
Pt species. The third peak, centred at around 460 K for Pt/Cr–ZnO
catalysts, is assigned to the surface reduction of ZnO in close con-
tact with the platinum particles. In the case of Pt/ZnO catalysts,
this peak is shifted at higher temperatures (475 K), this indicating
that the reduction temperature of the support has been increased.
These results are consistent with those previously reported by Con-
sonni et al. [14].
The main difference between both catalysts found in these TPR
experiments was the amount of H2 consumed. The Cr-containing
catalyst showed larger hydrogen consumption, this indicating
greater support reducibility. The distortion produced in the würzite
structure by the presence of chromium cations favours the support
reducibility.

58
E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
(a)
(b)
(c)
(d)
Fig. 10. Catalytic performance in the cinnamaldehyde hydrogenation (0.04 mol/L, 383 K and 7 MPa) after reduction at 473 K (a, b) and at 623 K (c, d): (a and c) cinnamalde-
hyde concentration as a function of time; (b and d) cinnamylalcohol selectivity as a function of conversion. First order rate constants are inserted.
of cinnamylalcohol (unsaturated alcohol), even up to high conver-
sion. A small amount of hydrocinnamaldehyde was also detected
(≈5%)
The catalytic performance of the catalysts reduced at 623 K is
shown in Figs. 10C and 10D. The catalytic activities of both cata-
lysts were lower than when reduced at 473 K, especially for cata-
lyst Pt/ZnO. Reduction at 623 K produced a loss of active sites, as
evidenced by CO chemisorption, which is attributed to a strong
metal-support interaction effect as discussed above [14,23]. The
selectivity towards the desired product (cinnamylalcohol) is very
high for both catalysts (>95%) for conversions above 15% up to
complete conversion of cinnamaldehyde (Fig. 10D).
Catalytic tests and characterisation results demonstrated that
at the lowest reduction temperature (473 K) the extent of the
metal–support interaction was adequate to generate active sites
that catalysed the formation of cinnamylalcohol. However, reduc-
tion at 623 K caused the extensive coverage of active sites by the
Fig. 11. Turnover frequencies for Pt/Cr,ZnO and Pt/ZnO catalysts reduced at 473 and
623 K in the cinnamaldehyde hydrogenation.
support (especially for the Pt/ZnO catalyst). This result agrees with
tive phase. Furthermore, the loss of active sites is lower than in
those previously reported for this kind of catalysts where it was
the Pt/ZnO catalyst.
shown that the SMSI effect can take place even after reduction at
If the catalytic activities are represented as turn-over frequen-
temperatures at low as 473 K [23]. Chromium addition increased
cies (TOF), some remarkable conclusions can be obtained. Fig. 11
the reducibility of the support, thus increasing the amount of Zn0.
shows the TOF for both catalysts after the two reduction tem-
The increased amount of Zn0 resulted in a higher probability of
peratures used in this study, which were calculated using the CO
alloy formation and in an increase of the support’s conductivity,
what can induce a change in the electronic properties of the ac-
chemisorption data (Table 2). After the reduction treatment at
473 K, the TOF value for Pt/ZnO was much higher than that for

E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
59
Table 3
Best results reported recently for this reaction over Pt catalysts
Catalyst
Solvent
Reaction conditions
Conversion (%)
Selectivity (%)
Reference
Pt/montmorillonite
Pt/Na-montmorillonite
PtRu/carbon nanotubes
Isopropanol
Ethanol
Isopropanol
298 K, 4.0 MPa
373 K, 7 MPa
373 K, 2 MPa
95
65
80
95
80
93
Szollosi et al., 1998 [52]
Manikandan et al., 2007 [51]
Vu et al., 2006 [50]
Pt/Cr–ZnO
Isopropanol
383 K, 7 MPa
100
96
This work
Pt/Cr–ZnO. This indicates that the active sites formed in Pt/ZnO af-
ter this reduction treatment at low temperature are much more ac-
tive. For Pt/Cr–ZnO, the number of active sites is higher and, thus,
the overall activity is higher, but the intrinsic activity of the active
sites is lower. However the TOF values are similar after reduction
treatment at 623 K. This may be ascribed to different factors. On
one hand, the surface segregation of chromium cations after the
reduction treatments seems to avoid the migration of patches of
the ZnO support over the metal particles and, consequently, the
number of active sites remains higher. Furthermore, chromium ad-
dition enhances the support reducibility producing an increase in
the amount of metallic zinc, which favours the formation of alloy
phases, this also affecting the catalytic activity [21,48,49].
The results found have special relevance in terms of selectiv-
ity. It has been recently reported that this kind of catalysts show
high activity and selectivity towards the production of the unsatu-
rated alcohol [23] in the vapour phase selective hydrogenation of
crotonaldehyde. But the use of vapour phase reaction limits their
industrial applications. Thereby, it has been found that their good
catalytic properties are kept also when they are used in the liquid
phase and with a complex reaction such as the selective hydro-
genation of cinnamaldehyde. When the results obtained with these
catalysts are compared with those recently reported for supported
platinum catalysts, it can be seen that these catalysts present the
best selectivity found without the use additives to the reaction
medium (Table 3) [50–52]. The high selectivity is attributed to the
electronic donating interaction of the metallic zinc to the platinum,
this lowering the interaction of the olefinic double bond of the
aldehyde with the catalyst [11,14].
and European Commission (Contract No. NMP3-CT2004-500895
‘InsidePores’ Network of Excellence).
References
[1] M. De Bruyn, S. Coman, R. Bota, V.I. Parvulescu, D. De Vos, P.A. Jacobs, Angew.
Chem. 42 (2003) 5333.
[2] P. Gallezot, D. Richard, Rev.-Sci. Eng. 40 (1998) 81.
[3] A. Corma, L.T. Nemeth, M. Renz, S. Valencia, Nature 412 (2001) 423.
[4] A. Corma, M.E. Comine, L.T. Nemeth, S. Valencia, J. Am. Chem. Soc. 124 (2002)
3194.
[5] A. Corma, M.E. Domine, S. Valencia, J. Catal. 215 (2003) 294.
[6] A. Corma, H. Garcia, Chem. Rev. 103 (2003) 4307.
[7] J. Jeak, J.E. Germain, J. Catal. 65 (1980) 133.
[8] A. Giroir-Fendler, D. Richard, P. Gallezot, Stud. Surf. Sci. Catal. 41 (1988) 171.
[9] J.C.S. Wu, T.-S. Cheng, C.-L. Lai, Appl. Catal. A Gen. 314 (2006) 233.
[10] T.B.L. Marinelli, V. Ponec, J. Catal. 156 (1995) 51.
[11] J.L. Margitfalvi, I. Borbáth, M. Hegedus, A. Tompos, Catal. A Gen. 229 (2002) 35.
[12] J.C. Serrano-Ruiz, A. Sepúlveda-Escribano, F. Rodríguez-Reinoso, J. Catal. 246
(2007) 158.
[13] J. Silvestre-Albero, J.C. Serrano-Ruiz, A. Sepúlveda-Escribano, F. Rodríguez-Rei-
noso, Appl. Catal. A Gen. 292 (2005) 244.
[14] M. Consonni, D. Jokic, D.Y. Murzin, R. Touroude, J. Catal. 188 (1999) 165.
[15] P. Gallezot, A. Giroir-Fendler, D. Richard, Catal. Lett. 5 (1990) 164.
[16] D.G. Blackmond, R. Oukaci, B. Blanc, P. Gallezot, J. Catal. 131 (1991) 401.
[17] A. Giroir-Fendler, D. Richard, P. Gallezot, Stud. Surf. Sci. Catal. 41 (1988) 171.
[18] M.L. Toebes, T.A. Nijhuis, J. Hájek, J.H. Bitter, A.J. Van Dellen, D. Murzin, K.P. De
Jong, Chem. Eng. Sci. 60 (2005) 5682.
[19] M.L. Toebes, Y. Zhang, J. Hájek, T. Nijhuis, J.B. Bitter, A. Jos Van Dillen, D.Y.
Murzin, D.C. Koningsberger, K.P. De Jong, J. Catal. 226 (2004) 215.
[20] A. Dandekar, M.A. Vannice, J. Catal. 183 (1998) 344.
[21] M.A. Vannice, Top. Catal. 4 (1997) 241.
[22] A. Huidobro, A. Sepúlveda-Escribano, F. Rodríguez-Reinoso, J. Catal. 212 (2002)
94.
[23] E.V. Ramos-Fernández, A. Sepúlveda-Escribano, F. Rodríguez-Reinoso, DOI:
10.1016/j.catcom.2007.11.010.
[24] A. Sepúlveda-Escribano, F. Coloma, J. Silvestre-Albero, F. Rodríguez-Reinoso,
Appl. Catal. A Gen. 304 (2006) 159.
4. Conclusions
[25] J. Silvestre-Albero, F. Rodríguez-Reinoso, A. Sepúlveda-Escribano, J. Catal. 210
(2002) 127.
[26] A. Sepúlveda-Escribano, F. Coloma, F. Rodríguez-Reinoso, J. Catal. 178 (1998)
649.
[27] A. Sepúlveda-Escribano, J. Silvestre-Albero, F. Coloma, F. Rodríguez-Reinoso,
Stud. Surf. Sci. Catal. 130 (2000) 1013.
[28] S. Bernal, M.A. Cauqui, G.A. Cifredo, J.M. Gatica, C. Larese, J.A. Pérez-Omil, Catal.
Today 29 (1996) 77.
[29] B. Campo, M. Volpe, S. Ivanova, R. Touroude, J. Catal. 242 (2006) 162.
[30] K. Liberková, L. Cerveny, R. Touroude, Res. Chem. Intermed. 26 (2003) 609.
[31] K. Liberková, R. Touroude, J. Mol. Catal. A Chem. 180 (2002) 221.
[32] M. Ohta, Y. Ikeda, A. Igarashi, Appl. Catal. A Gen. 258 (2003) 153.
[33] D.A. Shirley, Phys. Rev. B 5 (1972) 4709.
[34] R.B. Kale, H. Yung-Jung, L. Yi-Feng, L. Shih-Yuan, Solid State Commun. 142
(2007) 302.
[35] J. Wang, Y. Wang, S. Xie, M. Qiao, H. Li, K. Fan, Appl. Catal. A Gen. 272 (2004)
29.
The addition of Cr(III) cations to ZnO used as support for Pt cat-
alysts in the selective hydrogenation of cinnamaldehyde generated
several modifications both in the support and in the catalyst be-
haviour: (i) ZnCr2O4 spinel (mixed oxide) was generated; (ii) the
textural properties (larger surface area and resistance against sin-
tering) were improved; (iii) the presence of chromium favoured
the reducibility of ZnO; (iv) Pt dispersion was increased. The cat-
alytic performance was enhanced with these modifications, result-
ing in a very active and selective catalyst. About 95% selectivity
towards cinnamylalcohol was obtained up to complete conversion.
This enhancement is ascribed to a favourable metal–support inter-
action induced by chromium addition and the higher reducibility
of ZnO. The reduction temperature controls both the catalytic ac-
tivity and the selectivity. After reduction treatment at low temper-
ature (473 K), a very active and selective (90–95%) catalyst was
obtained, while reduction at high temperature (623 K) decreased
the activity, especially for the Cr-free catalyst, but the selectivity
was slightly improved to 95%.
[36] X. Wang, S. Yang, J. Wang, M. Li, X. Jiang, G. Du, X. Liu, R.P.H. Chang, J. Cryst.
Growth 19 (2001) 123.
[37] J.A. Rodríguez, M. Kuhn, J. Chem. Phys. 102 (1995) 4279.
[38] A. Sepúlveda-Escribano, J.C. Serrano-Ruiz, G.W. Huber, M.A. Sanchez-Castillo,
J.A. Dumesic, F. Rodríguez-Reinoso, J. Catal. 241 (2006) 378.
[39] P. Concepción, A. Corma, J. Silvestre-Albero, V. Franco, J.Y. Chane-Ching, J. Am.
Chem. Soc. 126 (2004) 5523.
Acknowledgments
[40] L. Jing, D. Wang, B. Wang, S. Li, B. Xin, H. Fu, J. Sun, J. Mol. Catal. A Chem. 244
(2006) 193.
[41] W.E. Epling, G.B. Hoflund, D.M. Minahan, J. Catal. 175 (1998) 175.
[42] J. Silvestre-Albero, A. Corma, A. Sepúlveda-Escribano, F. Rodríguez-Reinoso,
Appl. Catal. A Gen. 304 (2006) 159.
This work was supported from Ministerio de Educación y Cien-
cia, Spain (Projects BQU2003-06150 and NAN2004-09267-C03)

60
E.V. Ramos-Fernández et al. / Journal of Catalysis 258 (2008) 52–60
[43] W.E. Epling, G.B. Hoflund, W.M. Hart, D.M. Minahan, J. Catal. 172 (1997) 13.
[44] B. Liu, M. Terano, J. Mol. Catal. A Chem. 172 (2001) 227.
[45] W.E. Epling, G.B. Hoflund, W.M. Hart, D.M. Minahan, J. Catal. 169 (1997) 438.
[46] F.P.J.M. Kerkhof, J.A. Moulijn, J. Phys. Chem. 83 (1979) 1612.
[47] A. Huidobro, A. Sepúvelda-Escribano, F. Rodríguez-Reinoso, J. Catal. 121 (2002)
94.
[48] P. Claus, Top. Catal. 5 (1998) 51.
[49] V. Ponec, Appl. Catal. A Gen. 149 (1997) 27.
[50] H. Vu, F. Gonçalves, R. Philippe, E. Lamouroux, M. Corrias, Y. Kihn, D. Plee,
P. Kalck, P. Serp, J. Catal. 240 (2006) 18.
[51] D. Manikandan, D. Divakar, T. Sivakumar, Catal. Commun. 8 (2007) 1781.
[52] G. SzöllöSi, B. Török, L. Baranyi, M. Bartók, J. Catal. 179 (1998) 619.