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: adrian@fe.up.pt

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