RESEARCH PAPER
Pt nanoparticles supported over Ce–Ti–O: the solvothermaland photochemical approaches for the preparationof catalytic materials
Adrian M. T. Silva Æ Bruno F. Machado ÆHelder T. Gomes Æ Jose L. Figueiredo ÆGoran Drazic Æ Joaquim L. Faria
Received: 20 June 2008 / Accepted: 7 January 2009 / Published online: 4 February 2009
� Springer Science+Business Media B.V. 2009
Abstract Ce–Ti–O supports with different Ce/Ti
molar ratios were synthesized by the solvothermal
method using hexadecyltrimethylammonium bro-
mide. Pt nanoparticles were then supported by
photochemical deposition. The shape, size, and
structure of these materials were analyzed by high-
resolution transmission electron microscopy. The
single CeO2 support was also prepared, consisting
of agglomerated cubic particles ranging from *3 to
8 nm. When titania was combined with ceria, a
nanostructured architecture was produced, evidencing
the strong influence of Ti in the support structure.
Photodeposition of Pt nanoparticles is more efficient
on Ce–Ti–O supports than in pristine CeO2. Crystal-
line Pt nanoparticles (mainly of *2 to 4 nm) were
detected. The catalytic properties of the materials
were tested in the selective hydrogenation of cinna-
maldehyde to cinnamyl alcohol. It was observed that
Pt supported on Ce–Ti–O is more active and selective
than Pt on CeO2 or TiO2 separately. The catalyst with
40 mol% Ce leads to total conversion of cinnamal-
dehyde in a few minutes; however, higher selectivity
toward the desired product (cinnamyl alcohol) was
obtained with higher amounts of Ce (50 mol%).
Keywords Nanocatalysts � Solvothermolysis �Photodeposition � Platinum, cerium and titanium �Hydrogenation � Cinnamaldehyde
Introduction
Synthesis of nanostructured catalysts has gained
importance with the understanding of the relationships
between size-derived properties in nanoparticles at
molecular-level and catalytic performance. This
boosted the design and synthesis of new nanostruc-
tured catalysts (Bell 2003; Luisetto et al. 2008). The
advantage of using small particles in heterogeneous
catalysis arises from the larger fraction of atoms which
are available to participate in the catalytic process.
Especially in the liquid phase, the surface-to-volume
ratio increases significantly when particle size goes
nano (Rawle 2007). Surface atoms at the corners and
edges of these nanoparticles tend to be chemically
unsaturated and, thus, more active.
A. M. T. Silva (&) � B. F. Machado �H. T. Gomes � J. L. Figueiredo � J. L. Faria
Laboratorio de Catalise e Materiais (LCM), Laboratorio
Associado LSRE/LCM, Departamento de Engenharia
Quımica, Faculdade de Engenharia, Universidade do
Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal
e-mail: [email protected]
H. T. Gomes
Departamento de Tecnologia Quımica e Biologica,
Escola Superior de Tecnologia e de Gestao do Instituto
Politecnico de Braganca, Campus de Santa Apolonia,
5300-857 Braganca, Portugal
G. Drazic
Department of Nanostructured Materials, Jozef Stefan
Institute, Jamova 39, 1000 Ljubljana, Slovenia
123
J Nanopart Res (2010) 12:121–133
DOI 10.1007/s11051-009-9584-3
The catalysts reported in the hydrogenation of
several compounds, including cinnamaldehyde, have
been recently reviewed (Burda et al. 2005). Hydro-
genation reactions are considered as the most
common reactions where transition and noble metal
nanoparticles are applied in colloidal solution as
quasi-homogeneous catalysts. The main drawback in
the use of colloidal nanoparticles for catalysis is the
difficulty in recovery and reuse. This is especially
important if environmental issues are considered.
Moreover, it has been proved that, in some reactions,
the activity of platinum nanoparticles in colloidal
solution decreases with reaction time due to the
arrangement of the atoms toward a surface with
higher stability. Taking these reasons into consider-
ation, attention has been focused on heterogeneous
catalysts.
Impregnation, ion-exchange, adsorption, and
deposition–precipitation are typical preparation meth-
ods of supported metal catalysts, consisting of three
major steps: (i) preparation of the support, (ii)
introduction of the metal precursor onto the support,
and (iii) activation. In these methods a large number
of parameters are involved increasing the complexity
for proper control of metal particle size. Other
methods, such as chemical vapour deposition
(CVD), have gained especial attention due to their
reproducibility, simplicity, and flexibility. However,
specific equipment is required for high pressure and
temperature operation as well as for handling
hazardous and toxic compounds (Crisafulli et al.
2006). In this context, suitable alternatives are being
investigated, including the liquid-phase photochem-
ical deposition method, which allows integrated
preparation of the support and direct deposition of
active metallic species by using simple and inexpen-
sive equipment in liquid phase at room pressure and
temperature (Zhang et al. 2004; Crisafulli et al. 2006;
Giuffrida et al. 2007).
It is well known that the cinnamaldehyde hydro-
genation is strongly dependent on metal particle size.
Each metal deposition technique usually results in
different metal particle sizes. The particle sizes
normally obtained with the photochemical deposition
technique (2 nm [ d [ 50 nm) have associated
higher selectivities in hydrogenation reactions than
metal particles presenting a smaller size (Gallezot and
Richard 1998). Since curved metal surfaces of small
particles do not promote the repulsion of the aromatic
ring of the cinnamaldehyde molecule, the selectivity
to cinnamyl alcohol is lower when Pt particles are too
small (\2 nm). Accordingly, larger Pt particles
enhance the selectivity to cinnamyl alcohol due to
the steric effect created during the adsorption of the
planar cinnamaldehyde molecule onto the flat metal
surface. For instance, 98% of selectivity at 50%
conversion was observed on large graphite-supported
Pt particles (3–6 nm) produced by sintering 1.3 nm Pt
particles (Gallezot and Richard 1998).
Ceria (CeO2) is a rare-earth oxide extensively used
in catalysis, mainly as support or dopant (Ganduglia-
Pirovano et al. 2007; Martins et al. 2007; Laha and
Ryoo 2003; Silva et al. 2004a, b). This lanthanide
oxide, considered as the most abundant in the earth
crust, has attracted special attention due to its ability
to be easily and reversibly reduced from CeO2 to
nonstoichiometric oxides, CeO2-x. Beyond doubt, the
reducibility of the support plays an important role in
the properties of platinum catalysts used for selective
hydrogenation of aldehydes (Abid et al. 2006;
Gallezot and Richard 1998). Hydrothermal oxidation
of metals is a useful preparation method for ceria
nanoparticles with different shapes (Tok et al. 2007;
Lee and Choi 2004). When a solvent other than water
is used, the process is usually known as solvothermo-
lysis (Kobayashi et al. 2006; Verdon et al. 1995;
Zheng et al. 2005).
In what concerns noble metal deposition, the
photochemical method is gaining importance because
of its simplicity and its ability to control the particle
size and oxidation state of the deposited metal (Zhang
et al. 2004). It is well known (Englisch et al. 1997; da
Silva et al. 1997) that a high reduction temperature
improves the selectivity, and often also the activity,
this effect being mostly attributed to a Strong Metal
Support Interaction (SMSI) when reducible supports
are used (Tauster et al. 1978; Tauster 1987; Silvestre-
Albero et al. 2002).
Since solvothermal and photodeposition approaches
are interesting options in the synthesis of nanostruc-
tured catalysts, we focused our attention on the
production of Ce-based supports by the solvothermal
method and further deposition of Pt nanoparticles by a
photochemical approach. In particular, the effect of
combining titania with ceria was investigated. To the
best of our knowledge, this is the first report where
photochemical deposition is used to support Pt nano-
particles on cerium–titanium-oxide substrates.
122 J Nanopart Res (2010) 12:121–133
123
These catalysts were tested in the selective hydro-
genation of cinnamaldehyde (CAL). In this reaction,
cinnamyl alcohol (COL) is obtained by the hydroge-
nation of the carbonyl group of the unsaturated
aldehyde. Allyl alcohols are valuable intermediates
for the production of perfumes, flavoring additives,
pharmaceuticals, and agrochemicals. Unfortunately,
the selectivity toward these products is difficult to
achieve because thermodynamics favors the hydroge-
nation of the C=C bond over the C=O bond by about
35 kJ mol-1 (Mohr and Claus 2001), making the
production of the allyl alcohol very challenging.
Experimental
Reagents and materials
Cerium(III) nitrate hexahydrate (Ce(NO3)3 � 6H2O,
Fluka, 99%), tetraisopropyl orthotitanate (Ti[OCH
(CH3)2]4, Aldrich, 97%), cetyltrimethylammonium
bromide (CH3(CH2)15N(Br)(CH3)3-CTAB, Sigma,
C99%), potassium hydroxide (Aldrich), heptane
(Aldrich, 99%), cinnamaldehyde (Fluka, 98%) cin-
namyl alcohol (Fluka, 97%), hydrocinnamaldehyde
(Fluka, 90%), hydrocinnamyl alcohol (Aldrich, 98%),
3-cyclohexylpropan-1-ol (Aldrich, 99%), decane
(Fluka, 98%), hexachloroplatinic acid hexahydrate
(H2PtCl6 � 6H2O, Alfa Aesar, 99.9%), and methanol
(Chromanorm, C99.8%) were used as purchased.
Ultrapure water was produced in a Direct-Q millipore
system.
Solvothermal synthesis
The solvothermal synthesis of cerium and titanium
colloids was performed in a 160 ml 316-SS high-
pressure autoclave (Parr Instruments, USA Mod.
4564) equipped with a temperature controller (Mod.
4842). Initially, the cerium(III) precursor was slowly
dissolved in 75 ml of methanol containing the
cationic surfactant (CTAB) under continuous stirring
in order to obtain an equimolar solution with a metal
concentration of 0.10 M, which was prepared in
alkaline conditions by using KOH (3 M). The
solution obtained was transferred to a teflon vessel
inserted in the autoclave and heated up to the desired
temperature (423 K) under autogenous pressure. The
solution was then maintained at the desired
temperature for a duration of 150 min under a
continuous stirring speed of 500 rpm. At the end of
this stirring period, the autoclave was cooled to room
temperature. Colloids of Ce–Ti were prepared by
maintaining the total molar concentration equal to
0.10 M, but using different Ce:Ti molar percentages:
30:70, 40:60, and 50:50 (56:44, 66:34, and 75:25 in
wt.%, respectively). Colloids of Ti were synthesized
by the same approach but using the titanium precur-
sor instead of adding the cerium precursor. The
colloids were continuously washed in up-flow mode
with deionized water (ca. 3 ml min-1) for several
hours with the aid of a peristaltic pump. The colloids
were used as-obtained for the photodeposition of Pt.
Photodeposition of platinum
A solution was prepared by dissolving the adequate
amount of H2PtCl6 � 6H2O in methanol. Since
hydroxylated ceria is a basic precipitate which
dissolves in acidic solutions (Tok et al. 2007), KOH
(3 M) was added to the Pt solution. The cerium (or
the cerium–titanium or the titanium) colloids where
then transferred to a photocatalytic reactor and mixed
with the platinum precursor solution (5 wt.% Pt).
After 30 min of sonication, the suspension was
irradiated by a low-pressure mercury lamp with an
emission line at 253.7 nm (ca. 3 W of radiant flux)
for 4 h. The liquid-phase was evaporated at 400 K in
an oven. Finally, the catalysts were calcined (N2, 4 h,
100 cm3 min-1), reduced (H2, 2 h, 20 cm3 min-1),
and flushed (N2, 30 min) at 773 K prior to reaction.
The Pt loaded to the support was controlled indirectly
by monitoring the 261 nm band using an UV-Vis
spectrometer (Jasco V560).
Catalytic hydrogenation procedure
Hydrogenation of cinnamaldehyde (CAL) was car-
ried out in a 100 ml well-stirred stainless steel reactor
with a reaction mixture containing cinnamaldehyde
(ca. 1.4 g l-1), heptane, decane (as internal standard
for gas chromatography), and 2.5 g l-1 of catalyst
concentration. Traces of dissolved oxygen were
removed by bubbling N2 through the solution.
Immediately before starting the reaction, the reactor
was flushed with H2 (to purge the N2) and then
pressurized up to 10 bar (at 363 K). Small aliquots
were analyzed at different reaction times to monitor
J Nanopart Res (2010) 12:121–133 123
123
product distribution. The results of the reaction runs
were analyzed in terms of CAL conversion (XCAL,t):
XCAL;t ¼CCAL;0 � CCAL;t
CCAL;0� 100 ð1Þ
and product selectivity (Si,t):
Si;t ¼Ci;t
CCAL;0 � CCAL;t� 100 ð2Þ
where Ci,t represents the concentration of a generic
intermediate, namely cinnamyl alcohol (COL),
hydrocinnamaldehyde (HCAL), or hydrocinnamyl
alcohol (HCOL) at a given time t, which are the
main identifiable products in the hydrogenation of
cinnamaldehyde. For t = 0, CCAL,0 is the initial
concentration of CAL, all others being zero.
Analytical techniques
A JEOL 2010F analytical electron microscope,
equipped with a Schottky field-emission gun (FEG)
was used for transmission electron microscopy
(TEM) and high-resolution transmission electron
microscopy (HTREM) investigations. The micro-
scope was operated at 200 kV and an energy-dis-
persive X-ray spectrometer (EDXS) LINK ISIS-300
from Oxford Instruments, with an UTW Si–Li
detector was employed for the chemical analysis.
The samples for TEM were prepared from the diluted
suspension of nanoparticles in ethanol. A drop of
suspension was placed on a lacey carbon-coated Ni
grid and allowed to dry in air. Z-contrast images were
collected by using high-angle annular dark-field
detector (HAADF) in scanning transmission mode
(STEM). The crystal structures were determined by
the selected area electron diffraction (SAED)
technique.
Scanning electron microscopy (SEM) was per-
formed in a high performance FEI Quanta 200 FEG
SEM microscope equipped with a Schottky field
emission gun (FEG), for optimal spatial resolution,
and an Oxford Inca Energy Dispersive X-ray (EDX)
system for chemical analysis. The samples were
mounted on a double-sided adhesive tape made of
carbon. X-ray diffraction (XRD) analysis was carried
out in a Philips X0Pert MPD rotatory target diffrac-
tometer (k = 0.154 nm, 40 kV; data recorded at a
0.017� step size) in order to identify the crystallo-
graphic phases present and to calculate the crystallite
size from the XRD diffraction peaks by using the
modified Scherrer equation. Thermogravimetric anal-
ysis (TGA) was performed in air with a Mettler TA
4000 system, from 303 to 1,223 K at the rate of
10 K min-1. The particle size distribution was mon-
itored by light scattering in a Coulter LS230
instrument.
The liquid samples collected during the hydrogena-
tion experiments were analyzed in a DANI GC-1000
gas chromatograph, equipped with a split/splitless
injector, a capillary column (WCOT Fused Silica
30 m, 0.32 mm i.d., coated with CP-Sil 8 CB low
bleed/MS 1 lm film) and a flame ionization detector.
Results and discussion
CeO2 and Ce–Ti–O
Preliminary experiments were performed in order to
analyze the effect of the main parameters involved in
the solvothermal synthesis of ceria. The presence of a
base in the initial solution and the temperature used in
the solvothermal process had a strong effect in the
colloids. For synthesis carried out at 423 K, a dark
purple color was observed for solutions with excess
of KOH while a white color was characteristic of the
materials synthesized in the presence of low KOH
concentration (0.2 M). It was also observed that the
color of the material is dependent on the temperature
used. For instance, the particles are white for
temperatures lower than 423 K while a mixture of
purple and white colloids was obtained at 493 K. It
has been previously reported (Tang et al. 2005) that
the purple color is characteristic of cerium(III)
hydroxide, while the white color is related to
cerium(IV) hydroxide. Since cerium(III) is very
unstable under air atmosphere, the temperature was
set at 423 K.
A representative HRTEM micrograph of the
nanosized colloids (Fig. 1a) shows the as-obtained
Ce particles after solvothermal synthesis (suspended
in methanol), identified by EDXS. These colloids
appear to be ultrafine agglomerated crystallites,
mostly in the form of cubes with a body diagonal
*3 to 8 nm. These ceria nanoparticles are assembled
with different orientations and in some cases rounded
edges are observed. Similar HRTEM micrographs
124 J Nanopart Res (2010) 12:121–133
123
were observed for nanosized ceria obtained by the
hydrothermal method using higher temperatures
(523 K) and longer time of synthesis, between 6
and 24 h (Tok et al. 2007), by the microemulsion
method with toluene (Patil et al. 2002) as well as by
applying the solvothermal method at 423 K with
ethanol (Zheng et al. 2005).
The particle size distribution of the agglomerates
composed of Ce nanoparticles, measured in volume
percentage, was monitored using light scattering. The
solution containing the nanoparticles was allowed to
settle down for several days, deposition of the
agglomerates being observed at the bottom of the
flask. Figure 1b shows that the agglomerates’ size
distribution has a maximum at 30.1 lm (t1). When
ultrasonic irradiation was applied to the colloidal
solution, a shift in the maximum of the size
distribution toward lower particle sizes was observed,
explained by particle deagglomeration, tending to a
size distribution with maxima at 58 and 195 nm.
The colloids were slowly dried at ca. 353 K, pale
yellow particles being obtained. TGA analysis of
these precipitates resulted in a weight loss of 11.5%
between 303 and 1173 K, indicating dehydration of
the hydrous oxide form of CeO2 (CeO2 � xH2O) (Tok
et al. 2007).
Particle size (µm)0.1 1 10 100
Vol
ume
(%)
0
2
4
6
8
10
12
t1t2t3t4
(b)
(a)
Ce
Ce
O
2 3 4 5 6 7 8 91 keV
124
248
372
496
Counts
Fig. 1 a HRTEM
micrograph of CeO2 (inside
figure: EDXS analysis); (b)
Particle size distribution
before (t1) and after
(t2 \ t3 \ t4) different time
periods of ultrasound
irradiation
J Nanopart Res (2010) 12:121–133 125
123
The interaction between Ce and Ti when prepared
together was also investigated. Figure 2a shows the
HRTEM micrograph of Ce50–Ti50–O. In opposition
to CeO2, Ce–Ti–O particles are practically amor-
phous, with some crystalline nuclei (a few nm in size)
being observed. These nuclei were identified as cubic
CeO2 phase by comparing simulated (calculated) and
experimental selected area electron diffraction
(SAED) patterns (Fig. 2b).
Summarizing the previous observations, it is
concluded that, while CeO2 is essentially crystalline,
Ce–Ti–O supports are mostly amorphous, with some
CeO2 crystalline nanoparticles dispersed over the
materials.
Pt/CeO2 and Pt/Ce–Ti–O
The suitability of photodeposition to support Pt on the
nanostructured CeO2 was evaluated first. Figure 3a
shows HRTEM images of Pt/CeO2 produced by this
technique. It is possible to identify some small dark
spots (some of them marked by arrows) which
represent the deposited Pt nanoparticles (5 wt.%
confirmed by EDXS). The corresponding diffraction
patterns (experimental and simulated for cubic CeO2)
are shown in Fig. 3b.
The simulated pattern for cubic CeO2 agrees very
well with the rings observed in the experimental
SAED pattern. Apart from the rings belonging to
cubic ceria there are some spots (indicated by arrows)
that could be ascribed to cubic platinum. From the
fact that these spots are isolated (not forming a
continuous ring) and relatively intense, we could
assume that they originate from relatively large Pt
particles (�10 nm).
Figure 4a shows a typical HRTEM micrograph
of Pt photodeposited on the Ce–Ti–O amorphous
support.
Pt particles ranging in size from 2.5 to 4 nm are
clearly identified. A spherical Pt particle is shown at
higher magnification in Fig. 4b. This particle was
identified by means of EDXS (Fig. 5a). The compo-
sition of the support (Ce–Ti) was also estimated from
EDXS spectra (Fig. 5b). The average mass percent-
ages predicted for this sample by EDXS in terms of Ce
and Ti were 75 and 25 wt.%, respectively, which fully
agree with the nominal value of the Pt/Ce50–Ti50–O
catalyst. Moreover, a 5 wt.% Pt with respect to Ce–Ti
(95%) was estimated. When comparing the SAED
patterns of samples Pt/CeO2 and Pt/Ce50–Ti50–O
(Fig. 5c) with calculated patterns for Pt, CeO2, and
TiO2, it was found that a small amount of TiO2
crystalline anatase form could be present in the
Pt/Ce50–Ti50–O sample, since the spots correspond to
the position of the (101) lines of anatase and could not
be ascribed to cubic Pt or cubic CeO2.
SEM analysis was also performed for this
Pt/Ce50–Ti50–O sample (Fig. 6). The Pt particles were
clearly identified with the Back Scattered Electron
Detector (BSED), seen as small bright points distrib-
uted on the surface of Ce–Ti–O.Fig. 2 a HRTEM micrograph of Ce50–Ti50–O; (b) experi-
mental versus CeO2-simulated SAED patterns
126 J Nanopart Res (2010) 12:121–133
123
Figure 7 shows the XRD diffraction patterns of the
catalysts (the XRD of pure CeO2 before and after
calcination are also shown for comparison). The
diffraction peaks of pure CeO2 are indexed as 100%
of CeO2 with cubic structure. From the XRD spectra,
it can be seen that the particles have a lower degree of
crystallinity before calcination. These diffraction
peaks are slightly broader and with weaker intensities
when compared with the diffraction peaks obtained
for the calcined CeO2. Using the modified Scherrer
equation the calculated values of particle size are 3.6
and 6.5 nm, respectively, before and after calcination,
which are within the range observed in the TEM
images (Figs. 1a and 3a). The XRD diffraction peaks
identified as platinum are normally attributed to
2h = 39.7, 46.3, and 67.6�. They were clearly
identified in the Pt/Ce–Ti–O samples and evidence
of Pt was also found by XRD in the Pt/CeO2 catalyst,
in agreement with the previous HRTEM/SAED
analyses. The diffraction peaks of TiO2 anatase
(2h = 25.3, 48.0, and 54.7�) or TiO2 rutile
(2h = 27.4 and 41.3�) were not observed by XRD.
Some evidence of crystalline anatase was previously
detected by SAED (Fig. 5c) but it was not observed
by XRD probably due to the low content of this TiO2
phase in the Ce–Ti–O support.
Fig. 3 a HRTEM
micrographs of Pt/CeO2
with cubic CeO2 particles;
(b) Experimental versus
CeO2-simulated SAED
patterns
J Nanopart Res (2010) 12:121–133 127
123
In fact, Fang et al. (2007) who studied interfacial
structures of Ce–Ti–O oxides prepared by the sol-gel
technique found that, by increasing the amount of
titanium on ceria, the diffraction peaks in the XRD
patterns became broader and weaker than those
observed in pure cubic CeO2. This means that the
crystalline structure of cubic CeO2 is affected, decreas-
ing the crystalline size of cubic CeO2. One possible
explanation is that Ti4? substitutes Ce4? in the lattice of
cubic CeO2. Furthermore, Nakagawa et al. (2007)
studied various CeO2–TiO2 composite nanostructures.
When cerium was present in a higher content than
titanium, TiO2 peaks were not observed by XRD. These
authors concluded that the materials are not a simple
mixture of pure CeO2 and TiO2, instead, composite
materials were formed due to the good mixture of the
precursors in a molecular scale. Therefore, it is possible
to conclude that in the 40 mol% Ce support, Ce and Ti
are very well dispersed in small particles, originating a
nanostructured architecture.
The size of the Pt particles in most of the samples is
in the nanometer range (ca. 3 nm). Nevertheless, Pt was
clearly detected in the Ti-based materials by XRD
analysis, suggesting that some larger Pt particles are
present. From SAED results presented in Fig. 5c, in the
Fig. 4 a HRTEM micrograph of Pt/Ce50–Ti50–O; (b) HRTEM
at higher magnification showing a Pt particle
(a)
(b)
(c)
Pt
O
TiCe
0 3 6 9 12 15 18 21
0 3 6 9 12 15 18 21
keV
keV
0
15
30
45
60
75
90
105
120
Co
un
ts
0
200
400
600
800
Co
un
ts
Fig. 5 EDXS of Pt/Ce50–Ti50–O (a) spectrum from a platinum
particle and (b) from the Ce50–Ti50–O support; (c) Comparison
of (Pt/CeO2) and (Pt/Ce–Ti–O)-experimental versus (CeO2)
and (Pt)-simulated SAED patterns
128 J Nanopart Res (2010) 12:121–133
123
lower-left part where Pt/CeO2 pattern is shown, we can
detect just one strong and few weak spots at the
position of (111) planes of cubic Pt. In upper-right part
position corresponding to the Pt/Ce50–Ti50–O sample,
many strong spots are obvious. Since the number of the
diffraction spots belonging to Pt is higher in the case of
Pt/Ce50–Ti50–O, we can conclude that also the number
of Pt particles is higher in this sample. The HAADF/
STEM image (Z-contrast image) of the Pt/Ce40–Ti60–O
material in Fig. 8a shows a *30 nm Pt particle
(white contrast). The small Pt particles (*3 nm) are
labeled with arrows. Therefore, besides small Pt parti-
cles, there are also some larger Pt particles over the
Ce–Ti–O support. The HRTEM image is shown in
Fig. 8b, together with a detailed HRTEM image of a
3 nm-sized crystalline Pt particle on the amorphous
Ce–Ti–O matrix.
Hydrogenation of cinnamaldehyde
The liquid-phase selective hydrogenation of cinna-
maldehyde (CAL) was studied with the synthesized
nanostructured catalysts. Cinnamyl alcohol (COL) is
obtained by the hydrogenation of the carbonyl group
of the parent CAL. When the reaction follows the
thermodynamically favored hydrogenation of the C=C
bond, hydrocinnamaldehyde (HCAL) is produced.
Hydrogenation of both double bonds leads to the sat-
urated alcohol (hydrocinnamyl alcohol-HCOL). In addi-
tion, we also found traces of 3-cyclohexylpropan-1-ol
(CHP), resulting from the hydrogenation of the
aromatic ring of HCOL. The general reaction scheme
for the hydrogenation of cinnamaldehyde is presented
in Fig. 9. Since the reaction is carried out in heptane,
no products of dehydration are expected to occur, and
in fact they were not detected, and therefore not
included in the scheme.
The results obtained during the hydrogenation
experiments are shown in Fig. 10, in terms of
cinnamaldehyde conversion at different reaction
times, namely 10, 300, and 600 min (Fig. 10a) and
in terms of selectivity to the reaction products at
constant conversion of 50% (Fig. 10b). The conver-
sion levels achieved with Pt/CeO2 and Pt/TiO2
single-oxide catalysts were low when compared to
the mixed-oxide supports. Combination of Ce with Ti
had a positive effect on the conversion level which
increased significantly. The maximum activity was
obtained with Pt/Ce40–Ti60–O, 80% of cinnamalde-
hyde being converted in less than 10 min. The
behavior regarding the selectivity results was similar
to that observed for the conversion.
Fig. 6 SEM image of Pt/Ce50–Ti50–O with BSE detector
10 20 30 40 50 60 70 80 90 10
Inte
nsi
ty (
a.u
.)
2θ (degrees)
CeO2 dried
CeO2 calcined
Pt /CeO2
Pt / Ce50
-Ti50
-O
Pt / Ce40
-Ti60
-O
PtPt
Pt
PtCe
Ce
Ce
Ce
Ce Ce
Ce Ce
CeCe
Ce Ce
Pt
Fig. 7 X-ray diffraction patterns of the synthesized materials
(the numbers indicate Ce:Ti molar ratios)
J Nanopart Res (2010) 12:121–133 129
123
50 nm
50 nm
Pt small
Pt large
(a)
(b)
Fig. 8 a HAADF/STEM
image (Z-contrast image)
and (b) HRTEM
micrograph of
Pt/Ce40–Ti60–O
OHO
O OH OH
Cinnamaldehyde(CAL)
Cinnamyl alcohol(COL)
Hydrocinnamyl alcohol(HCOL)
Hydrocinnamaldehyde(HCAL)
3-cyclohexylpropan-1-ol(CHP)
H2
H2
H2 H2
3H2
2H2
Fig. 9 General reaction
pathway of the
cinnamaldehyde
hydrogenation reaction
130 J Nanopart Res (2010) 12:121–133
123
The use of single-oxide-supported catalysts
resulted in low amounts of desired COL, which
increased upon combination of Ce with Ti. For
Pt/CeO2, a lower amount of HCAL was produced
(15% selectivity, at 50% conversion) in comparison
to the amounts of COL and HCOL, which were
equivalent (around 43%). Hydrogenation of the C=C
bond was favored with Pt/TiO2, since larger amounts
of HCAL were produced. Moreover, similar values
were observed for HCAL and HCOL (nearly 40%).
Regarding the mixed-oxide-supported catalysts,
Pt/Ce50–Ti50–O provided the best results: a selecti-
vity of 63% was observed with respect to COL and a
low selectivity with respect to HCAL. The values for
HCOL were lower than those obtained for single-
oxide species. Hence, an interesting fact can be
evidenced: when used separately both single oxides
provide high amounts of HCOL, but when mixed
together the selectivity toward this product decreases
significantly, increasing the selectivity to the desired
COL. This may be explained by the occurrence
of two effects: (i) the structure created between Ce
and Ti when prepared together enhances the strong
metal-support interaction (SMSI) effect resulting
from the migration of reduced CeOx and TiOx
species onto the Pt particles. The occurrence of the
SMSI effect is well known in Pt/CeO2 and Pt/TiO2
catalysts reduced at high temperatures (Bernal et al.
2003) and affects the interaction between the metal
and the oxygen atom in the carbonyl group of
unsaturated aldehydes, activating and weakening the
C=O bond, enabling its preferential hydrogenation
(Concepcion et al. 2004; Vannice and Sen 1989);
(ii) the higher Pt particle sizes in Pt/Ce–Ti–O
catalysts promote the preferential hydrogenation of
the C=O bond (Abid et al. 2006). Therefore, in
general, it can be concluded that the combination of
Ce with Ti increases the conversion of CAL and the
selectivity toward COL, thus being possible to
enhance the catalytic performance of the catalyst.
Conclusions
Ultrafine nanometer-sized CeO2 particles with sizes
*3 to 8 nm were produced by the solvothermal
method. When Ce was combined with Ti, an amor-
phous Ce–Ti–O phase was obtained. It was proved
that crystalline spherical Pt nanoparticles (mainly *2
to 4 nm) can be efficiently photodeposited on the
amorphous Ce–Ti–O support. Enhancement in the
catalytic activity for cinnamaldehyde hydrogenation
besides higher selectivity for cinnamyl alcohol
production was observed when compared with the
single-oxide supports.
Acknowledgments The authors acknowledge Fundacao para
a Ciencia e Tecnologia (FCT/DREBM), Portugal, and the
Ministry of Higher Education, Science and Technology,
Slovenia, for financial support from the Portugal-Slovenia
Cooperation in Science and Technology (2008–2009) project
‘‘Synthesis and Characterization of Nanostructured Catalytic
Materials’’. AMTS acknowledges the financial support from
POCI/N010/2006. GD acknowledges the financial support
from the Slovenian Research Agency. BFM gratefully
acknowledges FCT for the PhD grant SFRH/BD/16565/2004.
This research was partially supported by project POCTI/58252/
EQU/2004 approved by FCT/POCTI and co-sponsored by
FEDER.
0
20
40
60
80
100
CA
L C
on
vers
ion
, %
0
20
40
60
80
100600 min 300 min 10 min
Ce mol %100 % 50 % 40 % 30 % 0 %
Sel
ecti
vity
, % (
XC
AL
50%
)
0
20
40
60
80
100
0
20
40
60
80
100HCOL HCAL COL
Ce mol %100 % 50 % 40 % 30 % 0 %
(a)
(b)
Fig. 10 Hydrogenation of cinnamaldehyde at 363 K and
10 bar of H2 pressure: (a) conversion results for different
reaction times (10, 300, and 600 min) and (b) selectivity
results (at constant conversion of 50%) obtained with Pt/CeO2
(Ce mol% = 100), Pt/TiO2 (Ce mol% = 0), and with Pt/Ce–
Ti–O catalysts presenting different values for Ce mol% (30, 40,
and 50%)
J Nanopart Res (2010) 12:121–133 131
123
References
Abid M, Paul-Boncour V, Touroude R (2006) Pt/CeO2 cata-
lysts in crotonaldehyde hydrogenation: selectivity, metal
particle size and SMSI states. Appl Catal Gen 297:48–59.
doi:10.1016/j.apcata.2005.08.048
Bell AT (2003) The impact of nanoscience on heterogeneous
catalysis. Science 299:1688–1691. doi:10.1126/science.
1083671
Bernal S, Calvino JJ, Cauqui MA, Gattica JM, Lopez Cartes C,
Perez Omil JA, Pintado JM (2003) Some contributions of
electron microscopy to the characterisation of the strong
metal–support interaction effect. Catal Today 77:385–
406. doi:10.1016/S0920-5861(02)00382-6
Burda C, Chen X, Narayanan R, El-Sayed MA (2005) Chem-
istry and properties of nanocrystals of different shapes.
Chem Rev 105:1025–1102. doi:10.1021/cr030063a
Concepcion P, Corma A, Silvestre-Albero J, Franco V, Chane-
Ching Y (2004) Chemoselective hydrogenation catalysts:
Pt on mesostructured CeO2 nanoparticles embedded
within ultrathin Layers of SiO2 binder. J Am Chem Soc
126:5523–5532. doi:10.1021/ja031768x
Crisafulli C, Scire S, Giuffrida S, Ventimiglia G, Lo Nigro R
(2006) An investigation on the use of liquid phase photo-
deposition for the preparation of supported Pt catalysts.
Appl Catal Gen 306:51–57. doi:10.1016/j.apcata.2006.03.
035
da Silva AB, Jordao E, Mendes MJ, Fouilloux P (1997) Effect
of metal–support interaction during selective hydrogena-
tion of cinnamaldehyde to cinnamyl alcohol on platinum
based bimetallic catalysts. Appl Catal Gen 148:253–264.
doi:10.1016/S0926-860X(96)00102-0
Englisch M, Jentys A, Lercher JA (1997) Structure sensitivity
of the hydrogenation of crotonaldehyde over Pt/SiO2 and
Pt/TiO2. J Catal 166:25–35. doi:10.1006/jcat.1997.1494
Fang J, Bi X, Si D, Jiang Z, Huang W (2007) Spectroscopy
studies of interfacial structures of CeO2–TiO2 mixed
oxides. Appl Surf Sci 253:8952–8961. doi:10.1016/
j.apsusc.2007.05.013
Gallezot P, Richard D (1998) Selective hydrogenation of a,
b-unsaturated aldehydes. Catal Rev Sci Eng 40:81–126.
doi:10.1080/01614949808007106
Ganduglia-Pirovano MV, Hofmann A, Sauer J (2007) Oxygen
vacancies in transition metal rare earth oxides: current
state of understanding and remaining challenges. Surf Sci
Rep 62:219–270. doi:10.1016/j.surfrep.2007.03.002
Giuffrida S, Condorelli GG, Costanzo LL, Ventimiglia G, Lo
Nigro R, Favazza M, Votrico E, Bongiorno C, Fragala IL
(2007) Nickel nanostructured materials from liquid phase
photodeposition. J Nanopart Res 9:611–619. doi:10.1007/
s11051-006-9089-2
Kobayashi T, Iwamoto S, Inoue M (2006) Properties of the
ceria particles prepared by the solvothermal oxidation of
cerium metal. J Alloy Compd 408–412:1149–1152. doi:
10.1016/j.jallcom.2004.12.188
Laha SC, Ryoo R (2003) Synthesis of thermally stable meso-
porous cerium oxide with nanocrystalline frameworks
using mesoporous silica templates. ChemComm 17:2138–
2139. doi:10.1039/b305524h
Lee J-S, Choi S-C (2004) Crystallization behaviour of nano-
ceria powders by hydrothermal synthesis using a mixture
of H2O2 and NH4OH. Mater Lett 58:390–393. doi:
10.1016/S0167-577X(03)00508-1
Luisetto I, Pepe F, Bemporad E (2008) Preparation and char-
acterization of nano cobalt oxide. J Nanopart Res 10:59–
67. doi:10.1007/s11051-008-9365-4
Martins TS, Hewer TLR, Freire RS (2007) Cerium: catalytic
properties and environmental applications. Quim Nova
30:2001–2006
Mohr C, Claus P (2001) Hydrogenation properties of supported
nanosized gold particles. Sci Prog 84:311–334. doi:
10.3184/003685001783238925
Nakagawa K, Murata Y, Kishida M, Adachi M, Hiro M, Susa
K (2007) Formation and reaction activity of CeO2 nano-
particles of cubic structure and various shaped CeO2–
TiO2 composite nanostructures. Mater Chem Phys
104:30–39. doi:10.1016/j.matchemphys.2007.02.047
Patil S, Kuiry SC, Seal S, Vanfleet R (2002) Synthesis of
nanocrystalline ceria particles for high temperature oxi-
dation resistant coating. J Nanopart Res 4:433–438. doi:
10.1023/A:1021696107498
Rawle AF (2007) Micron sized nano-materials. Powder
Technol 174:6–9. doi:10.1016/j.powtec.2006.10.012
Silva AMT, Marques RN, Quinta-Ferreira RM (2004a) Cata-
lysts based in cerium oxide for wet oxidation of
acrylic acid in the prevention of environmental risks. Appl
Catal Environ 47:269–279. doi:10.1016/j.apcatb.2003.09.
019
Silva AMT, Oliveira ACM, Quinta-Ferreira RM (2004b) Cat-
alytic wet oxidation of ethylene glycol: kinetics of
reaction on a Mn–Ce–O catalyst. Chem Eng Sci 59:5291–
5299. doi:10.1016/j.ces.2004.08.024
Silvestre-Albero J, Rodrıguez-Reinoso F, Sepulveda-Escribano
A (2002) Improved metal–support interaction in Pt/CeO2–
SiO2 catalysts after zinc addition. J Catal 210:127–136.
doi:10.1006/jcat.2002.3670
Tang C, Bando Y, Liu B, Golberg D (2005) Cerium oxide
nanotubes prepared from cerium hydroxide nanotubes. Adv
Mater 17:3005–3009. doi:10.1002/adma.200501557
Tauster SJ (1987) Strong metal–support interactions. Acc
Chem Res 20:389–394. doi:10.1021/ar00143a001
Tauster SJ, Fung SC, Garten RL (1978) Strong metal–support
interactions. Group 8 noble metals supported on TiO2. J
Am Chem Soc 100:170–175. doi:10.1021/ja00469a029
Tok AIY, Boey FYC, Dong Z, Sun XL (2007) Hydrothermal
synthesis of CeO2 nano-particles. J Mater Process
Technol 190:217–222. doi:10.1016/j.jmatprotec.2007.02.
042
Vannice MA, Sen B (1989) Metal–support effects on the
intramolecular selectivity of crotonaldehyde hydrogena-
tion over platinum. J Catal 115:65–78. doi:10.1016/0021-
9517(89)90007-9
Verdon E, Devalette M, Demazeau G (1995) Solvothermal
synthesis of cerium dioxide microcrystallites: effect of the
solvent. Mater Lett 25:127–131. doi:10.1016/0167-577X
(95)00161-1
Zhang F, Chen J, Zhang X, Gao W, Jin R, Guan N, Li Y (2004)
Synthesis of titania-supported platinum catalysts: the
132 J Nanopart Res (2010) 12:121–133
123
effect of pH on morphology control and valence state
during photodeposition. Langmuir 20:9329–9334. doi:
10.1021/la049394o
Zheng X, Wang S, Wang X, Wang S, Wang X, Wu S (2005)
Synthesis, characterization and catalytic property of ceria
spherical nanocrystals. Mater Lett 59:2769–2773. doi:
10.1016/j.matlet.2005.04.025
J Nanopart Res (2010) 12:121–133 133
123