Chemoselective hydrogenation of cinnamaldehyde over Pd/CeO 2-ZrO2 catalysts

Article abstract of DOI:10.1007/s10562-010-0423-z

Selective hydrogenation of cinnamaldehyde over Pd (2 wt%) supported on CeO_2, ZrO₂ and CeO₂-ZrO₂ catalysts is reported for the first time. In general, the olefinic (C=C) group of cinnamaldehyde is preferentially hydrogenated compared to the carbonyl (C=O) group. This selectivity preference could, however, be altered or reversed by adding alkali additives to the catalyst. The influence of additive on the structure and redox properties of the active sites and correlation of that with selective hydrogenation activity is investigated. Graphical Abstract: Selective hydrogenation of cinnamaldehyde over Pd (2 wt%) supported on CeO_2, ZrO₂ and CeO₂-ZrO₂ catalysts is reported for the first time. Alkali additives influence the selectivity for hydrogenation of olefinic (C=C) and carbonyl (C=O) groups of cinnamaldehyde. The influences of additive on the structure and redox properties of the active sites and correlation with selective hydrogenation activity are reported.[Figure not available: see fulltext.]

Full text of DOI:10.1007/s10562-010-0423-z

Catal Lett (2010) 140:55–64
DOI 10.1007/s10562-010-0423-z
Chemoselective Hydrogenation of Cinnamaldehyde
over Pd/CeO –ZrO Catalysts
2
2
S. Bhogeswararao D. Srinivas
•
Received: 3 June 2010 / Accepted: 30 July 2010 / Published online: 14 August 2010
Ó Springer Science+Business Media, LLC 2010
Abstract Selective hydrogenation of cinnamaldehyde
over Pd (2 wt%) supported on CeO ZrO and CeO –ZrO
2
catalysts [3–8] were reported to selectively hydrogenate the
carbonyl group resulting in cinnamyl alcohol. On the con-
trary, supported Pd catalysts were selective for hydrogena-
tion of olefinic group resulting in hydrocinnamyl aldehyde
[9]. In view of the industrial importance of the hydrogena-
tion products, design and development of more efficient and
selective catalysts is still remaining as a challenging task.
For this, a detailed study revealing the influence of the
nature of metal and support metal-interaction is essential.
We report here, for the first time, a preliminary study on the
catalytic activity of Pd (2 wt%) supported on CeO , ZrO
2
,
2
2
catalysts is reported for the first time. In general, the ole-
finic (C=C) group of cinnamaldehyde is preferentially
hydrogenated compared to the carbonyl (C=O) group. This
selectivity preference could, however, be altered or
reversed by adding alkali additives to the catalyst. The
influence of additive on the structure and redox properties
of the active sites and correlation of that with selective
hydrogenation activity is investigated.
2
2
Keywords Hydrogenation ꢀ Cinnamaldehyde ꢀ a, b-
Unsaturated aldehyde ꢀ Ceria–zirconia ꢀ Palladium catalyst
and CeO –ZrO . The influence of various reaction param-
2 2
eters and in particular, alkali (NaOH) addition on product
selectivity is investigated. By incorporation of Zr ions into
ceria lattice, the redox behaviour of the ceria is enhanced
[10]. Zirconia incorporation also increases the thermal sta-
bility of CeO and improves the metal–CeO interactions.
1
Introduction
2
2
Selective hydrogenation of a, b-unsaturated aldehydes is
one of the most important areas in catalysis research [1].
Cinnamaldehyde is a vinyologue of benzaldehyde. In such
compounds, hydrogenation can occur either at olefinic
CeO –ZrO -supported Pd has been used as three-way cat-
2 2
alysts [11–16], in natural gas combustion [17], total oxida-
tion of hydrocarbons [18], methanol decomposition [19],
vapor phase hydrogenation of phenol [20] and in several
other reactions [21]. We report here the chemoselective
hydrogenation activity of these novel catalysts.
(
C=C) or carbonyl (C=O) groups. While the former leads to
saturated aldehydes of industrial and biological impor-
tance, the latter results in unsaturated alcohols (Scheme 1).
All these selective hydrogenation products of cinnamal-
dehyde are widely used in the perfumery industry. Recently
hydrocinnamaldehyde was found be an important inter-
mediate in the preparation of pharmaceuticals used in the
treatment of HIV [2]. Supported Pt, Ru, Co, Au and Ir
2 Experimental
2.1 Catalyst Preparation
2.1.1 CeO –ZrO
2 2
CeO –ZrO composites were prepared by co-precipitation
2
2
S. Bhogeswararao ꢀ D. Srinivas (&)
Catalysis Division, National Chemical Laboratory,
Pune 411 008, India
method [3]. Ce(NO ) ꢀ6H O (7.35 g, Semco) and ZrO
3
3
2
(
NO ) ꢀxH O (2.8 g, Loba Chemie) were dissolved sepa-
3 2
2
e-mail: d.srinivas@ncl.res.in
rately in distilled water (168 and 120 mL, respectively).
123

5
6
S. Bhogeswara Rao, D. Srinivas
Scheme 1 Hydrogenation
products of cinnamaldehyde
These solutions were mixed together and added drop-wise
to a continuously stirred 0.1 M aq. solution (1 L) of NaOH
taken in a 2 L round bottom flask, placed in an oil bath.
Stirring was continued for 3 h. During addition, the tem-
perature of the flask was maintained at 80 °C and pH at 10.
The cations were precipitated in the form of their
hydroxides. The mixture was digested at 80 °C for another
Xpert Pro PANalytical X-ray diffractometer with Ni-fil-
tered Cu Ka radiation (40 kV, 30 mA) in the 2h range of
10–80° at a scan rate of 2.3°/min. The specific surface area
(BET) of the samples was determined using a NOVA 1200
Quanta Chrome instrument. The micropore volume was
estimated from the t-plot and the pore diameter was esti-
mated using the Barret-Joyner-Halenda (BJH) model.
Morphological characteristics of the samples were deter-
mined using a high resolution transmission electron
microscope (HRTEM, JEOL, model 1200 EX) operating at
100 kV.
3
h and then cooled to 25 °C. The precipitate was sepa-
rated, washed with copious amounts of distilled water until
?
all the Na ions were removed. It was then air-dried,
powdered and calcined at 450 °C, for 2.5 h, to get the final
product. The CeO to ZrO molar ratios in the composites
X-ray photoelectron spectra (XPS) of the samples were
acquired on a VG Microtech Multilab ESCA 3000 with
Mg Ka radiation (hm = 1253.6 eV). Base pressure in the
2
2
were 1:1, 1:2 and 1:4. The Na content in the samples is
below the detection limit of inductively couple plasma
-
10
(
ICP) technique. ‘‘Neat’’ ceria and ‘‘neat’’ zirconia supports
analysis chamber was maintained at 3–6 9 10
mbar.
were also prepared by the precipitation method.
The peak corresponding to carbon 1s (at 284.2 eV) was
taken as the reference in estimating the binding energy
(BE) values of various elements in the catalyst. The error in
BE values is ± 0.1 eV. Samples were prepared by pressing
the powders into pellets. In order to accelerate outgassing
and remove C from the surface, samples were heated at
2
.1.2 CeO –ZrO -Supported Pd Catalysts
2 2
Pd (2 wt%) supported on CeO –ZrO was prepared by dry-
2
2
impregnation method. In a typical preparation, 0.570 g of
tetraamminepalladium(II) nitrate [Pd(NH ) (NO ) , 10 wt%
450 °C for 2 h in flowing O before performing the XPS
3
4
3 2
2
solution in water, Aldrich] was added drop-wise to 1 g of
CeO –ZrO taken in a beaker. It was mixed thoroughly at
measurements. The measurements included a survey
spectrum and the Ce 3d, Zr 3d and Pd 3d core-level spectra,
which were analyzed following the standard procedures.
Temperature-programmed reduction experiments were
carried on a Micromeritics Auto Chem 2910 instrument:
2
2
2
5 °C, dried at 120 °C for 5 h till all the water evaporated and
then, calcined at 400 °C for 2 h. Pd (2 wt%)/CeO and Pd
2
(
2 wt%)/ZrO catalysts were prepared in a similar manner.
2
0
.1 g of the catalyst was placed in a quartz tube and treated
3
with a O (22%)–He (78%) gas mixture (30 cm /min) at
2
.1.3 Alkali Impregnated CeO –ZrO -Supported Pd
2
2
2
Catalysts
250 °C for 2 h. A gas mixture of H (5%)–Ar (95%) was
2
3
then passed (50 cc /min) through the quartz reactor at
In a typical synthesis, 0.262 g of NaOH dissolved in 1 mL
of water and 0.570 g of 10 wt% aqueous solution of tet-
raamminepalladium(II) nitrate were added slowly to 1 g of
CeO –ZrO while mixing thoroughly at 25 °C. The slurry
was dried at 120 °C for 5 h till the water got evaporated. It
was then, calcined at 400 °C for 2 h.
25 °C for 1 h. The temperature was raised to 950 °C at a
heating rate of 5°/min and held at 950 °C for 0.5 h. The
amount of hydrogen consumed was estimated and the
reduction capacity of the catalyst was determined. A
standard CuO powder was used to calibrate the amount of
2
2
H consumption.
2
The type and density of the acid sites were determined
by diffuse reflectance infrared Fourier transform (DRIFT)
spectroscopy of adsorbed pyridine and temperature-pro-
2
.2 Characterization Techniques
Crystallinity and phase purity of the samples were deter-
mined from X-ray diffractograms (XRD) recorded on a
grammed ammonia desorption (NH -TPD) techniques. In
3
DRIFT measurements, the samples were initially activated
1
23

Chemoselective Hydrogenation of Cinnamaldehyde
57
at 450 °C. The temperature was brought down to 50 °C and
pyridine was adsorbed. Then, the temperature of the sam-
ple was raised and held at a desired value (100 °C) for
x C 0.85, a monoclinic phase is produced [22]. At inter-
0
00
mediate compositions, various phases (t, t , t , j and t*) have
been identified [22–24]. Oxygen deficient structures, such
as the pyrochlore structure (A B X ) may also be formed
0
.5 h before starting the measurements (spectral resolu-
2
2 7
-
1
tion = 2 cm ; number of scans = 200). Difference FTIR
spectra were obtained by subtracting the spectrum of the
catalyst from that of the sample adsorbed with pyridine.
[25–29]. For ceria–zirconia mixed oxides, the pyrochloric
structure is generally reported to be formed under severe
reduction conditions. However, there is an evidence of its
formation even during low temperature reduction of the
sample [29]. In other words, the ceria–zirconia material
exhibit complex structural characteristics. Neat ceria of the
present study showed XRD peaks in the 2h range of 28–80°
indicating a distinct, cubic, fluorite structure [30]. On the
contrary, the CeO –ZrO composites exhibited XRD pattern
NH -TPD measurements performed on a Micromeritics
3
AutoChem 2910 instrument. In a typical experiment, 0.1 g
of catalyst was taken in a U-shaped, flow-thru, quartz
sample tube. Prior to measurements, the catalyst was pre-
3
treated in He (30 cm /min) at 700 °C for 1 h. A mixture of
3
NH in He (10%) was passed (30 cm /min) at 25 °C for
3
2
2
1
h. Then, the sample was, subsequently flushed with He
3
30 cm /min) at 100 °C for 1 h. The TPD measurements
typical of a mixture of cubic and tetragonal phases with the
former being more predominant (Fig. 1; the peaks due to the
tetragonal phase are marked by asterisk). Recently, Li et al.
[31] have also reported such a mixed crystallite phase in
CeO –ZrO materials prepared by calcining at lower tem-
(
were carried out in the range 100–800 °C at a heating rate
of 10 °C/min. Ammonia concentration in the effluent was
monitored with a gold-plated, filament thermal conductiv-
ity detector. The amount of desorbed ammonia was
determined based on the area under the peak.
2
2
peratures (\500 °C). In the materials prepared by us, the
presence of pyrochloric structure was not detected. The unit
cell parameter of the cubic phase (acubic) of CeO –ZrO is
2
2
2
.3 Reaction Procedure: Hydrogenation
of Cinnamaldehyde
smaller (0.537 nm) than CeO2 (0.542 nm), suggesting
the formation of a CeO –ZrO solid solution [note: ionic
2
2
4
radius of Zr (0.084 nm) is smaller than that of Ce
?
4?
3
?
Prior to reactions, the catalyst was reduced under a flow of
hydrogen (30 mL/min) at 200 °C for 2 h. Upon reduction,
the color of the catalyst changed from dark brown to black.
It was cooled to 25 °C. Without exposing to atmosphere,
(0.097 nm)/Ce (0.112 nm)] [32]. Palladium and sodium
impregnation did not alter the structure except for shift in
peaks to higher 2h value by 0.1–0.3°. At 2 wt% of Pd
loading, no separate PdO phase was detected indicating that
it is highly dispersed. Representative XRD profiles of
CeO –ZrO , Pd/CeO –ZrO and Na–Pd/CeO –ZrO are
0
.05 g of the reduced catalyst was transferred into a Parr
reactor (300 mL) containing cinnamaldehyde (4 g) and
solvent (40 mL). The reactor was initially flushed and then
pressurized with hydrogen to a desired value (2, 5, 10 or
2
2
2
2
2
2
depicted in Fig. 1. The average crystallite size of the
materials was determined from the linewidth of the XRD
peak corresponding to (111) reflection (2h & 28.7°) using
2
0 bar). The temperature of the reactor was also raised to a
desired value (50, 80, 120 or 150 °C) and the reaction was
conducted while stirring (500 revolutions per minute) for
1
1
1
400
200
000
8
h. Samples were collected at intermittent times, diluted
with the solvent and then analyzed and quantified by gas
chromatographic technique (Varian 3400; CP-SIL8CB
column; 30 m-long and 0.53 mm-i.d.). The products were
identified by GC-MS (Shimadzu GCMS-QP5050A; HP-5
column; 30 m-long 9 0.25 mm i.d. 9 0.25 lm thickness).
800
00
400
6
Na-Pd/CeO -ZrO
2
2
200
0
1
1
1
400
200
000
8
00
00
00
00
0
6
4
2
*
Pd/CeO -ZrO
2
2
*
*
3
3
3
Results and Discussion
1200
(111)
.1 Structural and Textural Properties
.1.1 X-Ray Diffraction
1000
8
6
00
00
(220)
(
200)
(311)
400
CeO -ZrO
2
2
200
0
20
30
40
50
60
70
80
The ceria–zirconia mixed oxides are formed in different
crystalline structures depending on their chemical compo-
2
θ (degree)
sition. According to the Ce1-xZr O phase diagram, for
2
x
Fig. 1 X-ray diffractograms of CeO –ZrO (1:1 M ratio) and
2
2
x B 0.15 a cubic fluorite-type phase is formed, while for
supported Pd catalysts
123

5
8
S. Bhogeswara Rao, D. Srinivas
Table 1 Physicochemical properties of CeO
2
–ZrO
2
-supported Pd catalysts
adsorption–desorption
Sample
XRD
N
2
Acidity (NH
TPD, mmol/g) (TPR, mmol/g)
3
-
2
H uptake–Pd
Lattice
parameter
(
Average
crystallite (m /g) (cm /g)
acubic, nm) size (nm)
S
BET
Pore volume Average pore
2
3
diameter (nm)
CeO
2
0.542
0.512
0.537
0.540
0.510
0.538
6.9
11.5
6.7
–
–
–
–
–
ZrO
2
–
–
–
CeO
2
–ZrO
2
115
–
0.17
–
6.1
–
0.179
0.231
0.390
Pd (2 wt%)/CeO
2
2
10.0
11.1
10.6
8.5
Pd (2 wt%)/ZrO
2
–
–
–
Pd (2 wt%)/CeO
–ZrO
2
(1:1)
65
121
0.11
0.14
7.2
4.3
0.308
0.034
0.076
0.583
Na–Pd (2 wt%)/CeO
2
2
–ZrO (1:1) 0.538
Debye-Scherrer equation (Table 1). It varied for different
lower (4.3 nm) than CeO –ZrO (6.1 nm) and Pd/CeO –
2
2
2
supports in the order: ZrO (11.5 nm) [ CeO (6.9 nm)
ZrO (7.2 nm). It is worth noting that the variation in S
2 BET
2
2
[
CeO –ZrO (6.7 nm) indicating that zirconia incorpora-
(obtained from N adsorption–desorption measurements)
2
2
2
tion renders materials with low crystallite size. The crys-
tallite size of CeO –ZrO -supported Pd was higher
follows the variation in crystallite size of these materials
(determined from XRD studies).
2
2
(
10.6 nm) than the corresponding support oxide (6.7 nm,
respectively). Alkali impregnation (Na–Pd/CeO –ZrO )
2
3.1.3 High Resolution Transmission Electron Microscopy
2
decreased the crystallite size to 8.5 nm.
.1.2 N Adsorption–Desorption
Representative HRTEM images of Pd/CeO , Pd/ZrO and
2
2
3
Pd/CeO –ZrO are shown in Fig. 3. Palladium is in a dis-
2 2
2
persed state on CeO –ZrO support. Agglomeration of
2
2
The textural properties of CeO –ZrO , Pd/CeO –ZrO and
2
particles was detected on neat ceria and zirconia supports.
In spite of the fact that HRTEM usually only measures a
small area of the sample, it is interesting to note that the
crystallite size of the support (5–12 nm) estimated with
HRTEM) agrees well with that determined using XRD. CO
chemisorption could provide more information on metal
dispersion. However, these experiments could not be done
due to lack of this facility.
2
2
2
Na–Pd/CeO –ZrO2 were determined by N2 adsorption–
2
desorption technique (Fig. 2; Table 1). All these materials
showed typical type IV isotherms with H3 hysteresis loops.
The specific surface area (SBET) and pore volume of sup-
2
3
ported Pd catalysts are lower (65 m /g and 0.11 cm /g,
respectively) than those of the support composite oxide
2
115 m /g and 0.17 cm /g, respectively). When sodium
3
(
was also present the material exhibited highest S
BET
2
121 m /g). However, the average pore diameter of it was
(
3.1.4 X-Ray Photoelectron Spectroscopy
Representative X-ray photoelectron spectra of Pd/CeO2–
ZrO catalysts in Pd 3d region are shown in Fig. 4. The
2
1
1
1
20
CeO -ZrO2
2
9
6
3
0
0
0
0
binding energy values are reported in Table 2. The binding
0
2?
4?
energies of Pd , Pd and Pd states are 335 and 340.6,
37.5 and 342.9, and 338.4 and 343.5 eV, respectively
3
20
Pd/CeO -ZrO
2
2
corresponding to the 3d5/2-3/2 spin–orbit doublet [33–36].
The binding energy of Zr 3p also falls in the same range.
Therefore, the B.E. peaks of Zr 3p overlap with those of Pd
3d. From the estimated binding energy values (Table 2) it
is concluded that Pd on CeO –ZrO is in a ?2 oxidation
9
6
3
0
0
0
0
20
Na-Pd/CeO -ZrO
2
9
6
3
0
0
0
0
2
2
2
state. The binding energy value of Pd (3d5/2) on pure ZrO2
is higher than that observed over pure CeO support con-
2
0
.0
0.2
0.4
0.6
0.8
1.0
sistent with the facile reducibility of the latter than the
former support. The 3d5/2 peak of Pd in Na–Pd/CeO –ZrO
P/P
0
2
2
appeared at lower binding energy value (336.5 eV) than
that of Pd/CeO –ZrO (336.8 eV) revealing that Pd in the
2 2 2
Fig. 2 N adsorption–desorption isotherms of CeO –ZrO (1:1 M
ratio) and supported Pd catalysts
2
2
1
23

Chemoselective Hydrogenation of Cinnamaldehyde
59
Fig. 3 High-resolution transmission electron micrographs: a Pd (2 wt%)/CeO
ratio = 1:1)
2
, b Pd (2 wt%)/ZrO
2
, and c Pd (2 wt%)/CeO
2
–ZrO
2
(Ce:Zr molar
Fig. 4 XPS of Pd 3d region of
supported palladium catalysts
Pd3d
Zr3p
Zr3p
Pd 3d
Pd/CeO -ZrO
Na-Pd/CeO -ZrO
2
2
2
2
3
25
330
335
340
345
325
330
335
340
345
B.E. (eV)
B.E. (eV)
The sample showed IR peaks of adsorbed pyridine at 1592,
Table 2 Binding energy values (in eVs) for supported Pd catalysts
-
575, 1484 and 1440 cm (Fig. 5). While the peaks at
1
1
Sample
Pd/CeO
Pd 3d5/2
Ce 3d5/2
Zr 3d5/2
Na 1s
-1
592 and 1440 cm are attributed to H-bonded pyridine,
1
-
1
2
2
336.4
337.9
336.8
336.5
881.3
–
–
–
those at 1575 and 1484 cm are corresponded to pyridine-
coordinated to weak Lewis acid sites [37]. It is known that
strong Lewis acid sites show adsorbed pyridine-IR peaks at
Pd/ZrO
2
181.5
181.3
181.0
–
–
Pd/CeO
–ZrO
2
881.0
881.2
-
1
and Br o¨ nsted sites at 1639 and
546 cm . However, these peaks are not observed in the
spectrum of CeO –ZrO indicating that such strong acidic
Na–Pd/CeO
2
–ZrO
2
1070.1
1623 and 1455 cm
-
1
1
2
2
former is in a reduced ?(2 - d) oxidation state while that
in the latter is in a ‘‘formal’’ ?2 oxidation state. In other
words, alkali affected the electronic structure of Pd. The
peaks corresponding to Pd in Na–Pd/CeO –ZrO are
sites are absent. The pyridine-IR peaks broadened and
decreased in intensity as the desorption temperature
increased from 200 to 400 °C (Fig. 5). At higher temper-
atures (ca., 400 °C) strong additional peaks appeared at
2
2
-
1
broader than those for Pd/CeO –ZrO due to the presence
2
1540 and 1580 cm , the origin of which is not clear at this
2
of different metal oxidation states (?2 and ?(2 - d)). This
led to differences in the relative intensity of Zr and Pd
peaks in these two samples (Fig. 4). The B.E. values of Ce
point of time.
The number of acidic sites on CeO –ZrO and supported
2
2
Pd catalysts were quantified by NH -TPD measurements
3
3
d5/2 and Zr 3d5/2 are also included in Table 2 and are
(Fig. 6; Table 1). CeO –ZrO showed a broad desorption
2
2
consistent with the reported values [33].
peak in the temperature region 100–500 °C (Fig. 6, right
panel). Based on the knowledge gained from pyridine-IR
3
.1.5 Acidic Properties: DRIFT Spectroscopy of Adsorbed
Pyridine and NH -TPD
studies, this broad NH desorption is attributed to surface
3
3
acidic hydroxyl and weak Lewis acid (Ce and Zr) sites. In
the case of Pd catalysts a shoulder peak was also observed
at around 110–150 °C which could be attributed to NH3
desorption from PdO species (Fig. 6). The amount of NH3
desorbed from different supported Pd catalysts decreased in
CeO –ZrO is a weakly acidic oxide. The nature and type
2
2
of acid sites on CeO –ZrO were determined by DRIFT
2
2
spectroscopy using pyridine as probe molecule (Fig. 5).
123

6
0
S. Bhogeswara Rao, D. Srinivas
0
.015
0.010
.005
3
37
Pd/CeO -ZrO
CeO -ZrO2
2
2
2
4
49
0
5
2
°
2
00 C
0.000
0.005
-
0
0
0
0
.09
.06
.03
.00
°
3
00 C
6
1
Na-Pd/CeO -ZrO
2
2
9
7
°
4
00 C
0
100 200 300 400 500 600 700 800 900
°
1700
1650
1600
1550
1500
1450
1400
Temperature ( C)
-1
Wavenumber (cm )
Fig. 5 Pyridine-IR of CeO –ZrO support
2
Fig. 7 H —temperature-programmed reduction of supported Pd
catalysts
2
2
the following order: Pd/ZrO (0.390 mmol/g) [ Pd/CeO –
[39]. Indeed Pd/CeO –ZrO sample showed such a com-
2 2
2
2
ZrO2 (0.308 mmol/g) [ Pd/CeO2 (0.231 mmol/g). The
plicated TPR profile (Fig. 7). The peaks at around 52 °C
could be attributed to reduction of PdO. The negative peaks
at higher temperatures (100–179 °C) are attributed to
decomposition of PdHx species. Two intense peaks
observed 337 and 449 °C are attributed to reduction of
amount of NH desorbed from support CeO –ZrO was
3
2
2
found to be 0.179 mmol/g. These results indicate that ZrO2
is more acidic than CeO . Incorporation of ZrO into the
2
2
CeO lattice enhances the acidity of the support. The alkali
2
impregnated supported Pd catalyst (Na–Pd/CeO –ZrO )
2
surface and bulk ceria. Neat CeO shows reduction peak at
2
2
did not show NH desorption peak. In other words, alkali
3
around 550 °C [37]. Its shift to low temperature (337 °C)
as observed in the present study is due to interaction with
zirconia in the solid solution and hydrogen spillover and
metal (Pd)-support interactions [37]. When Na is also
present the reduction peak due to Pd shifted to higher
temperatures (Fig. 7). Reduction of PdO occurred at 61 and
97 °C in Na–Pd/CeO –ZrO instead of 52 °C in Pd/CeO –
impregnation suppressed the acidity of the catalysts.
3
.1.6 Temperature-Programmed Reduction
The H2 consumption profiles during temperature-pro-
grammed reduction (TPR) of supported Pd catalysts are
shown in Fig. 7. According to the published reports [19,
2
2
2
ZrO . This shift in the reduction peak to higher temperature
2
3
8], PdO on ceria–zirconia supports is reduced at low
is a consequence of metal support interactions. Thus, Na
impregnation not only alters the textural and acidic
properties but also influences the reduction behaviour.
The peaks due to CeO2 reduction which occurred at
337–449 °C in Pd/CeO –ZrO are not seen in the case of
temperatures (60–90 °C). Hence, the H2 consumption
profiles of the present samples were recorded from room
temperature onwards up to 950 °C. The interaction of
hydrogen with Pd–ceria based catalyst system is very
complex as (1) Pd can be reduced at temperatures as low as
room temperature, (2) Pd forms hydrides with complex and
unknown stoichiometry, (3) the amount of hydrogen stored
2
2
Na–Pd/CeO –ZrO . H uptakes exclusively by palladium
2
2
2
in Pd/CeO –ZrO and Na–Pd/CeO –ZrO were estimated to
2
2
2
2
be 0.076 and 0.583 mmol/g, respectively. The higher
amount of H consumption in Na–Pd/CeO –ZrO than in Pd/
in these PdH depends on their size, and (4) these hydrides,
x
2
2
2
except the surface Pd–H ones, decompose at around
CeO –ZrO indicates higher dispersion of palladium in the
2 2
7
0–80 °C, H spillover occurs on ceria or ceria–zirconia
former catalyst. The H consumption for Na–Pd/CeO –ZrO
2
2
2
Fig. 6 NH -TPD profiles of
supports and supported Pd
catalyst
3
0.9
197
0.9
0.6
1
95
157
0.6
0.3
0.0
0
0
.3
.0
CeO -ZrO
2
2
Pd/CeO2
Pd/ZrO2
0
0
.9
.6
110
258
0.3
.0
Pd/CeO -ZrO2
2
0.9
0.6
0.3
0.0
0
0.9
0.6
Na-Pd/CeO -ZrO
2
2
0.3
0.0
1
00
200
300
400
500
600
700
100
200
300
400
500
600
700
Temperature (°C)
Temperature (
°
C)
1
23

Chemoselective Hydrogenation of Cinnamaldehyde
61
is higher (0.583 mmol/g) than that expected for theoretical
PdO reduction (0.188 mmol/g). This difference may be
composition of CeO in the mixed oxide. In other words,
2
while ZrO support leads to higher conversion and HCA
2
4
?
attributed to concurrent reduction of surface Ce , which
occurs at this lower temperature [40].
product formation (via 3,4 and 1,4 additions), CeO sup-
2
port results in higher amount of PPL formation. CeO is an
2
easily reducible oxide and hence a part of Pd can be in d?
state. On the other hand, Pd on ZrO (a weakly reducible
d?
3
.2 Catalytic Activity
2
oxide) will be in zero state only. An electron depleted Pd
Hydrogenation of cinnamaldehyde (CA) yields different
products as shown in Scheme 1. Hydrogenation of the
carbonyl group (1,2-addition) yields the unsaturated alco-
hol—cinnamyl alcohol (CAL). Hydrogenation of the ole-
finic bond (3,4-addition) gives the saturated aldehyde—
hydrocinnamaldehyde (HCA). The 1, 4-addition gives enol
which isomerizes into HCA. Further hydrogenation leads
to the formation of 3-phenylpropanol (PPL) and subse-
quently, propylbenzene (PB). Also CAL can get reduced to
PPL. Hydrogenation of a, b unsaturated aldehydes on metal
surfaces proceeds via the Horiuti–Polayni mechanism
on CeO will polarize the electron rich carbonyl group
2
instead of electron deficient olefinic group and thereby
results in preferential 1,2-addition of hydrogen and for-
mation of PPL. Pdd? would be less reactive than Pd and
hence lower conversion of CA was observed over Pd/CeO2
than over Pd/ZrO2.
0
3.2.2 Effect of Temperature
Hydrogenation of CA increased with increasing reaction
temperature (Table 4). Complete conversion of CA was
observed at 80 °C itself. Hydrocinnamaldehyde (HCA) was
the major product (82.5–90%). 3-Phenylpropanol (PPL)
was detected with selectivity in the range 10.0–16.7%. A
small amount of other products including ethers and acetals
(0–0.7 wt%) was also detected. With increasing tempera-
ture the selectivity of HCA increased at the expense of PPL
and other products.
2
2
involving adsorbed states: di-rC = Og , di-rC = Cg , or
2
4
di-pg (g ) [41–43]. In the present study, over Pd/CeO –
2
ZrO catalysts, 3,4 and 1,4-additions are more predominant
2
than the 1,2-addition. At our reaction conditions, formation
of PB could not be detected.
3
.2.1 Influence of Support
The type and the nature of the support have a marked
influence on CA conversion and product selectivity
3.2.3 Effect of Solvent
(
Table 3). CA conversion is higher with Pd supported on
The reaction was conducted in non-polar toluene and
cyclohexane and polar ethanol solvents (Table 4). In gen-
eral, CA conversion is much higher in ethanol than in tol-
uene and cyclohexane. Figure 8 shows the conversion
verses time plots in different solvents. While HCA is the
selective product (82.5–84.6%) in non-polar solvents, HCA
(50.7%) along with significant amounts of PPL (10.8%) and
other products (acetals ? ethers; 38.5%) formed in polar
ethanol solvent. Unlike in non-polar solvents, hydrogena-
tion of the carbonyl group of CA yielding PPL is more
pronounced in ethanol. As PPL is formed in higher amounts,
it participates in further reactions at higher temperatures
leading to acetals and ethers (selectivity = 38.5% at 50 °C
and 74% at 150 °C).
ZrO than on CeO (compare entry nos. 4 and 5; Table 3).
2
2
In the case of Pd supported on CeO –ZrO , an increase in
2
2
conversion of CA was observed with increasing amount of
ZrO in the mixed oxide composition (compare entry nos. 1
2
and 3, Table 3). However, PPL selectivity (obtained via
hydrogenation of carbonyl group) was relatively higher
over Pd supported on CeO (18%) than on ZrO (11.3%).
2
2
Even in the case of CeO –ZrO catalysts, the PPL selec-
2
2
tivity was found higher when there is an increase in the
Table 3 Hydrogenation of cinnamaldehyde over supported Pd
catalysts
Catalyst (Ce:Zr
molar ratio)
CA conversion Product selectivity (%)
(wt%)
H2 pressure influenced CA conversion and products
selectivity (Table 5; Fig. 9). The conversion of CA
increased from 59.1 to 89.9% by increasing the pressure
from 2 to 20 bar. Also a significant reduction in the for-
mation of other products (from 73.6 to 38.5%) was
observed at higher temperatures. The reaction occurs even
on ‘‘bare’’ CeO –ZrO . But CA conversion and HCA and
HCA
PPL
Others
Pd/CeO
Pd/CeO
Pd/CeO
Pd/CeO
2
2
2
2
–ZrO
–ZrO
–ZrO
2
2
2
(1:1) 69.6
(1:2) 68.1
(1:4) 79.6
91.0
82.5
86.0
90.0
82.0
88.7
16.7
13.9
10.0
18.0
11.3
0.7
0.1
0
0
2
2
Pd/ZrO
2
95.5
0
PPL selectivity were much higher when Pd was also
present. Pd to some extent blocks the acidic sites leading to
Reaction conditions: Catalyst (0.05 g), cinnamaldehyde (CA, 4 g),
solvent—toluene (40 mL), H pressure = 20 bar, reaction time =
h. HCA hydrocinnamaldehyde, PPL 3-phenylpropanol, other
products include ethers and acetals
2
side reactions on the CeO –ZrO surface and thereby
2
8
2
enhances the selectivity of the desired products.
123

6
2
S. Bhogeswara Rao, D. Srinivas
2 2
Table 4 Influence of temperature on hydrogenation of cinnamaldehyde over Pd (2 wt%)/CeO –ZrO (Ce:Zr molar ratio = 1:1)
Temperature (°C) Solvent = toluene, additive = nil
Solvent = ethanol,
additive = nil
Solvent = ethanol,
additive = 16 mg NaOH
in 10 mL water
CA conversion Product selectivity (%)
(wt%)
CA Product
conversion selectivity (%)
CA
Product
conversion selectivity (%)
(wt%)
(
wt%)
89.9
HCA
PPL
Others
HCA PPL Others
HCA PPL Others
a
a
82.5 (84.6) 16.7 (15.0) 0.7 (0.4)
a
a
5
8
1
1
1
1
0
69.6 (60.4)
50.7 10.8 38.5
37.0 19.9 43.1
21.0 14.5 64.5
12.3 12.3 74.0
99.4
100
67.3 29.7
60.1 32.7
2.9
7.3
0
100
100
100
87.7
88.1
90.0
11.6
11.3
10.0
0.7
0.6
0
100
100
100
20
50
50
50
100
100
42.1 30.7 27.2
29.5 37.9 32.6
b
b
35.0 40.4 23.3
b
b
c
100
c
c
c
20.0 63.4 16.6
100
Reaction conditions: Catalyst (0.05 g), cinnamaldehyde (CA, 4 g), solvent (40 mL), H
cinnamaldehyde, PPL 3-phenylpropanol, other products include ethers and acetals
2
pressure = 20 bar, reaction time = 8 h. HCA hydro-
a
Data in parentheses corresponds to experiments in cyclohexane instead of toluene
b
H
2
pressure = 10 bar, additive = 18 mg of NaOH in 30 mL water
c
H
2
pressure = 10 bar, additive = 54 mg of NaOH in 30 mL water
100
(d)
c)
Table 5 Effect of pressure and alkali on hydrogenation of cinna-
maldehyde over Pd (2 wt%)/CeO –ZrO (Ce:Zr molar ratio = 1:1)
(
2
2
8
0
0
0
0
Catalyst H pressure CA
2
Product selectivity
(
b)
(
bar)
conversion (%)
(wt%)
6
4
2
(
a)
HCA PPL Others
CeO –ZrO
2
2
20
20
58.8
89.9
86
30.4
4.3 65.3
Pd/CeO
–ZrO
2 2
50.7 10.8 38.5
1
0
5
2
39.2
38.4
24.5
89.0
9.5 51.3
8.8 52.8
1.9 73.6
8.9 2.1
0
2
4
6
8
62.6
59.1
72.3
Reaction time (h)
Fig. 8 Conversion versus time plots for hydrogenation of cinnamal-
dehyde (CA) in different solvents. (a) Cyclohexane, (b) toluene,
2 2
Na–Pd/CeO –ZrO 20
Reaction conditions: Catalyst (0.05 g), cinnamaldehyde (CA, 4 g),
solvent—ethanol (40 mL), reaction temperature = 50 °C, reaction
time = 8 h. HCA hydrocinnamaldehyde, PPL 3-phenylpropanol,
other products include ethers and acetals
(
c) ethanol and (d) ethanol ? additive (10 mg of NaOH in 10 mL
water). Reaction conditions: Pd (2 wt%)/CeO –ZrO (Ce:Zr molar
pressur-
2
2
ratio = 1:1) = 0.05 g, CA = 4 g, solvent = 40 mL, H
e = 20 bar, reaction temperature = 50 °C
2
3
.2.4 Effect of Alkali Addition
in ethanol conducted at 150 °C and 10 bar H pressure
2
(
Table 4). Suppression of other products and change in
The formation of acetals and ethers (other products) are
catalyzed by acid sites on the catalyst surface. Pyridine-IR
selectivity for hydrogenation of C=C (yielding HCA) to
C=O (yielding PPL) was noted. An enhancement in PPL
product selectivity from 37.9 to 63.4% was observed when
54 mg of NaOH in 30 mL water was added. Alkali has a
positive effect on the rate of hydrogenation (Fig. 8). It
activates the C=O group of a, b-unsaturated aldehydes and
facilitates hydrogenation yielding unsaturated alcohols
(CAL). The enhanced activation of the C–O bond could be
interpreted by the polarization of C=O bond resulting from
the interaction of the alkali cation with the lone-pair elec-
trons of the oxygen atom of the C–O group. These unsatu-
rated alcohols (CAL) are further hydrogenated to saturated
alcohols (PPL) in presence of Pd/CeO –ZrO catalyst by 1,4
(
Fig. 5) and NH -TPD (Fig. 6) experiments revealed the
3
presence of Lewis acid sites on CeO –ZrO -based catalysts.
2
2
In order to suppress these side reactions forming undesired
products we have impregnated the catalysts with alkali and
then calcined to prepare Na–Pd/CeO –ZrO . NH desorp-
2
2
3
tion is little (0.034 mmol/g) from Na–Pd/CeO –ZrO indi-
2
2
cating that the catalyst is nearly neutral. The selectivity for
the acetals and ethers decreased from 38.5 to 2.1% by using
this alkali-modified catalyst (Table 5).
In other experiments, we have added NaOH solution of
different concentrations to the reactions over Pd/CeO –ZrO
2
2
2
2
1
23

Chemoselective Hydrogenation of Cinnamaldehyde
63
1
00
time. Cinnamaldehyde has two functional groups (C=C and
C=O) for hydrogenation. In general, the hydrogenation of
C=C is preferred over Pd catalysts. However, by adding a
small quantity of alkali the selectivity for hydrogenation
can be changed/altered. Alkali addition influenced the
dispersity, electronic property and reducibility of Pd. These
changes in molecular electronic structure of Pd are
responsible for the changes in the chemoselectivity brought
about by alkali addition.
2
0 bar
8
0
0
0
0
1
0 bar
5
bar
bar
6
4
2
2
0
2
4
6
8
Reaction time (h)
Acknowledgments S.B. acknowledges the Council of Scientific and
Industrial Research, New Delhi for the award of Junior Research
Fellowship. Authors thank Dr. K.V.R. Chary, IICT, Hyderabad for the
help in TPR measurements.
Fig. 9 Conversion of cinnamaldehyde (CA) at various pressures over
Pd (2 wt%)/CeO –ZrO (Ce:Zr molar ratio = 1:1) catalyst. Reaction
conditions: catalyst = 0.05 g, CA = 4 g, ethanol = 40 mL, reaction
temperature = 50 °C
2
2
References
addition. Finally, the saturated alcohol percentage increases
with increasing alkali percentage and the formation of ace-
tals and ethers is reduced. Mahmoud et al. [2] reported that
hydrogenation of HCA to PPL does not occur significantly at
1. Boudart M (1994) Nature 372:320
. Mahmoud S, Hammoudeh A, Gharaibeh S, Melsheimer J (2002) J
Mol Catal A 178:161
2
3
. Serrano-Ruiz JC, Luettich J, Sep u´ lveda-Escribano A, Rodr ´ı guez-
Reinoso F (2006) J Catal 241:45
4. Kluson P, Cerveny L (1995) Appl Catal 128:13
normal temperatures over Pd/SiO catalyst. However, the
2
hydrogenation of CAL proceeds very fast (30 times faster
than the hydrogenation of CA to CAL) [2]. These sequential
reactions thereby enhance the selectivity of PPL in the
product. A similar behaviour occurs even in our experi-
ments. The Na ions in Na–Pd/CeO –ZrO are involved
5
. Chen XF, Li HX, Xu YP, Wang MH (2002) Chin Chem Lett
3:107
1
6
. Claus P, Br u¨ lckner A, Mohr C, Hofmeister H (2000) J Am Chem
Soc 122:11430
7. Rylander PN (1979) Catalytic hydrogenation in organic synthesis.
Academic Press, New York, p 72
2
2
mainly in the reduction of acid sites on the support surface.
On the other hand, in the experiments with alkali solutions,
the alkali ions have not only suppressed the acidity of the
catalyst but at the same time involved in polarising the C=O
group of CA leading to differences in product selectivity.
8
. Zhang L, Winterbottom JM, Boyes AP, Raymahasay S (1998) J
Chem Technol Biotechnol 72:264
. Delbecq F, Sautet P (1995) J Catal 152:217
9
10. Ranga Rao G, Ka sˇ par J, Di Monte R, Meriani S, Graziani M
1994) Catal Lett 24:107
1. Zhao B, Li G, Ge C, Wang Q, Zhou R (2010) Appl Catal B
6:338
(
1
1
1
1
The catalytic activity of Pd (2 wt%)/CeO –ZrO2 is
2
9
significantly higher than that of carbon nanofiber and
activated carbon-supported Pd (5 wt%) catalysts. With less
amount of Pd in the former catalyst, complete conversion
of CA at 80 °C was achieved in just 1 h while it requires
nearly 30 h over the latter catalysts [44]. Pd (1 wt%)/SiO2
required higher temperature (100 instead of 80 °C), pres-
sure (30 instead of 20 bar) and amount of catalyst (200
instead of 50 mg) to achieve similar conversions of CA. In
other words, the Pd catalysts of the present study are more
efficient and the support has a significant role in the
improved activity of these catalysts. Pd catalysts are known
for selective hydrogenation of C=C. However, in this
study, it is demonstrated that by adding a small quantity of
alkali, the electronic and redox properties of Pd can be fine-
tuned and the selectivity for hydrogenation (between C=C
and C=O of cinnamaldehyde) can be controlled.
2. Shen M, Wang J, Shang J, An Y, Wang J, Wang W (2009) J Phys
Chem C 113:1543
3. Shen M, Yang M, Wang J, Wen J, Zhao M, Wang W (2009) J
Phys Chem C 113:3212
4. Graham GW, Jen HW, McCabe RW, Straccia AM, Haack LP
(2000) Catal Lett 67:99
15. Zhao M, Chen S, Zhang X, Gong M, Chen Y (2009) J Rare Earths
7:728
2
1
1
1
1
6. Bera P, Patil KC, Jayaram V, Subbanna GN, Hegde MS (2000) J
Catal 186:293
7. Specchia S, Palmisano P, Finocchio E, Busca G (2010) Chem
Eng Sci 65:186
8. Wang G, Meng M, Zha Y, Ding T (2010) Fuel. doi:10.1015/
j.fuel.2010.03.031
9. Wang H, Chen Y, Zhang Q, Zhu Q, Gong M, Zhao M (2009) J
Nat Gas Chem 18:211
20. Velu S, Kapoor MP, Inagaki S, Suzuki K (2003) Appl Catal A
45:317
2
2
2
1. Vivier L, Duprez D (2010) ChemSusChem 3:654
2. Yashima M, Hirose T, Katano S, Suzuki Y, Kakihana M,
Yoshimura M (1995) Phys Rev B 51:8018
2
2
2
3. Otsuka-Yao-Matsuo S, Omata T, Izu N, Kishimoto H (1998) J
Solid State Chem 138:47
4. Omata T, Kishimoto H, Otsuka-Yao-Matsuo S, Ohtori N, Ume-
saki N (1999) J Solid State Chem 147:573
5. Thomson JB, Armstrong AR, Bruse PG (1996) J Am Chem Soc
118:11129
4
Conclusions
Chemoselective hydrogenation of cinnamaldehyde over
CeO –ZrO supported Pd catalysts is reported for the first
2
2
123

6
4
S. Bhogeswara Rao, D. Srinivas
2
2
2
6. Montini T, Speghini A, De Rogatis L, Lorenzut B, Bettinelli M,
Graziani M, Fornasiero P (2009) J Am Chem Soc 131:13155
7. Shah PR, Kim T, Zhou G, Fornasiero P, Gorte RJ (2006) Chem
Mater 18:5363
8. Montini T, Hickey N, Fornasiero P, Graziani M, Ba n˜ ares MA,
Martinez-Huerta MV, Alessandri I, Depero LE (2008) Chem
Mater 17:1157
36. Roy S, Marimuthu A, Hegde MS, Madras G (2008) Catal Commun
9:811
37. Bekyarova E, Fornasiero P, Ka sˇ par J, Graziani M (1998) Catal
Today 45:179
38. Luo MF, Zheng XM (1999) Appl Catal A 189:15
39. Gatica JM, Baker RT, Fornasiero P, Bernal S, Ka sˇ par J (2001) J
Phys Chem B 105:1191
2
9. Montini T, Ba n˜ ares MA, Hickey N, Di Monte R, Fornasiero P,
Ka sˇ par J, Graziani M (2004) Phys Chem Chem Phys 6:1
0. Trovarelli A (1996) Catal Rev 38:439
40. Formasiero P, Di Monte R, Ranga Rao G, Ka sˇ par J, Trovarelli A,
Graziani M (1995) J Catal 151:168
41. Bauer K, Garbe D (1985) Common fragrance and flavor mate-
rials. VCH, Weinheim
42. Fukuzawa SI, Fujuami T, Yamuchi S, Sakai S (1986) J Chem Soc
Perkin Trans I 1929
43. Touroude R (1980) J Catal 65:110
44. Pham-Huu C, Keller N, Ehret G, Charbonniere LJ, Ziessel R,
Ledoux MJ (2001) J Mol Catal A 170:155
3
3
3
1. Li G, Zhao B, Wang Q, Zhou R (2010) Appl Catal B 97:41
2. Mart ´ı nez-Arias A, Fern a´ ndez-Garc ´ı a M, Ballesteros V, Salamanca
LN, Conesa JC, Otero C, Soria J (1999) Langmuir 15:4796
3. Franchini CA, Cesar DV (2010) Catal Lett 137:45
4. Dolbecq A, Compain JD, Mialane P, Marrot J, Secheresse F,
Keita B, Holzie LRB, Miserque F, Nadjo L (2009) Chem Eur J
3
3
1
5:733
3
5. Parvulescu VI, Parvulescu V, Endruschat U, Filoti G, Wagner FE,
Kubel C, Richards R (2006) Chem Eur J 12:2343
1
23