Completely selective hydrogenation of trans-cinnamaldehyde to cinnamyl alcohol promoted by a Ru-Pt bimetallic catalyst supported on MCM-48 in supercritical carbon dioxide

Article abstract of DOI:10.1039/b210161k

Supercritical carbon dioxide (scCO₂) is a unique reaction medium for the selective hydrogenation of cinnamaldehyde using a well-defined Ru-Pt bimetallic catalyst supported on MCM-48. The reaction is completely selective to the unsaturated alcohol in supercritical carbon dioxide. The selectivity to the desired unsaturated alcohol shows a significant pressure dependence. By varying the pressure at the constant reaction temperature of 323 K, we are able to optimize the selectivity of cinnamyl alcohol to 100%. This is a particularly dramatic enhancement compared to that in organic solvents and under solventless conditions, where Ru-Pt bimetallic catalyst produces cinnamyl alcohol along with hydrocinnamaldehyde and 3-phenylpropanol. More importantly, the excellent physicochemical properties of the scCO₂ medium provide a significant influence on the selectivity, which goes beyond that of simple solvent replacement only.

Full text of DOI:10.1039/b210161k

View Article Online / Journal Homepage / Table of Contents for this issue
Completely selective hydrogenation of trans-cinnamaldehyde to
cinnamyl alcohol promoted by a Ru–Pt bimetallic catalyst
supported on MCM-48 in supercritical carbon dioxide
Maya Chatterjee,* Yutaka Ikushima and Fengyu Zhao
Supercritical Fluid Research Center, National Institute of Advanced Industrial Science and
Technology, 4-2-1 Nigatake, Miyagino-ku, Sendai 983-8551, Japan. E-mail: c-maya@aist.go.jp;
Fax: +81-22-237-5217; Tel: +81-22-237-5211
Received (in Montpellier, France) 15th October 2002, Accepted 3rd December 2002
First published as an Advance Article on the web 11th February 2003
Supercritical carbon dioxide (scCO2) is a unique reaction medium for the selective hydrogenation of
cinnamaldehyde using a well-defined Ru–Pt bimetallic catalyst supported on MCM-48. The reaction is
completely selective to the unsaturated alcohol in supercritical carbon dioxide. The selectivity to the desired
unsaturated alcohol shows a significant pressure dependence. By varying the pressure at the constant reaction
temperature of 323 K, we are able to optimize the selectivity of cinnamyl alcohol to 100%. This is a particularly
dramatic enhancement compared to that in organic solvents and under solventless conditions, where Ru–Pt
bimetallic catalyst produces cinnamyl alcohol along with hydrocinnamaldehyde and 3-phenylpropanol. More
importantly, the excellent physicochemical properties of the scCO2 medium provide a significant influence on
the selectivity, which goes beyond that of simple solvent replacement only.
the use of green and clean technology. Hence, it is of practical
Introduction
importance to develop an environmentally friendly process
based on heterogeneous catalysis. Although the hydrogenation
of a-b unsaturated aldehydes using heterogeneous catalysts
has been reviewed by several authors,10 the design of catalysts
Supercritical carbon dioxide (Tc ¼ 31 ꢀC, Pc ¼ 7.375 MPa,
dc ¼ 0.468 g mLꢁ1) has received significant attention during
the past decades as a prospective solvent for organic synthetic
reactions.1–6 It is considered to be an environmentally benign
reaction medium that can be easily tuned by changing the pres-
sure and temperature. From a fundamental point of view it
represents a new class of solvents with some properties
between those of the gas and liquid phases. It can dissolve
easily many reactant gases such as H2 , N2 , and O2 , etc. The
scCO2 medium not only enhances the miscibility of reactant
gases, but it also reduces the viscosity and increases the diffu-
sion rate, so that mass transfer is no longer the limiting factor.7
The particular advantage of this medium is the ability to con-
trol conditions with great precision, to manipulate the selectiv-
ity of the reaction.
=
to adapt the selectivity to C O remains a challenging task.
Several studies have searched for highly active and selective
catalysts and it has been found that noble metal catalysts are
active and selective to P1.10a,11 In order to further improve
the activity and selectivity, several suitable catalytic systems
have been reported;12 these often contain a second metal along
with the noble metal. These bimetallic systems, like Pt–Sn,12a
Pt–Fe,13 Ru–Sn,14 etc., are supported on conventional silica,
alumina, zeolites or active carbon. Among these, Pt-containing
bimetallic catalysts have attracted the most attention because
of their activity and chemoselectivity. Several attempts were
made to identify the factors that control the selectivity in
hydrogenation of unsaturated carbonyl compounds. The most
studied factors are the nature of the support,15 shape selectiv-
ity,16 electronic effects,17 reaction conditions such as hydrogen
pressure,11 and solvent.18 For instance, noble metal encapsu-
lated zeolites have been considered as shape-selective catalysts
in selective hydrogenation reactions. High selectivity to unsa-
turated alcohol at 50% conversion was observed on a platinum
cluster supported zeolite Y (96%)17 and zeolite beta (87%).16
However, these catalysts suffer from quick inactivation, which
slows down the reaction rate. It is well established that, com-
pared to the zeolites, mesoporous materials can act as good
supports19 because of their incredible surface properties such
as large surface area and tuneable pore size. The selective
hydrogenation of a-b unsaturated carbonyl compounds on
mesoporous supports has hardly been studied. Parvulesue
et al.20 compared the selective hydrogenation of R utilizing Ru
supported on Beta zeolite MCM-41 and MCM-48 supported
Ru catalyst in iso-propanol medium with Ru is the active spe-
cies. The result showed lower selectivity (50% selectivity to P1
Selective hydrogenation of a-b unsaturated aldehydes to
unsaturated alcohols is a stepping-stone for the production
of fine chemicals, particularly used in the fragrance and fla-
vor industry. It is well known that selective hydrogenation of
=
the C O double bond is much more difficult because reac-
tion kinetics and thermodynamics both favor hydrogenation
=
of the C C double bond. This is due to the fact that the
difference in negative free reaction enthalpy of ꢂ35 kJ molꢁ1
=
=
favors the hydrogenation of C C rather than C O. Cinnamal-
dehyde (R), an example of a a-b unsaturated aldehyde, is of
great practical importance; both the hydrogenated and semi-
hydrogenated products find wide application in theperfume
industry. Besides this, the unsaturated alcohol, cinnamyl
alcohol (P1), is an important building block in organic synth-
esis.8 Traditionally, the hydrogenation of R to P1 has been car-
ried out in organic solvents using homogeneous catalysts.9 The
main drawbacks of this method lie in: (i) difficulty in separat-
ing the catalyst from the product and (ii) the use of organic
solvents as reaction medium, which are not compatible with
510
New J. Chem., 2002, 27, 510–513
DOI: 10.1039/b210161k
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2002

View Article Online
at 60% conversion) and the products were not restricted to P1
or the saturated aldehyde but also gave a number of side pro-
ducts depending on the solvent used.
conducted at 323 K, with 7.5 mmol of reactant in the presence
of a constant hydrogen pressure of 4 MPa and with varying
CO2 pressure in the range of 7.0 to 17.0 MPa. From naked
eye observations R is immiscible at 7.0–10.0 MPa of pressure.
But the solubility increases with pressure to form a clearly visi-
ble uniform single phase at 12 MPa of CO2 .
To avoid the disadvantage of using conventional organic
p
solvent as a reaction medium, we have chosen supercritical
carbon dioxide as the solvent. Recently, we have reported
the successful use of MCM-48 supported Pt as a suitable cat-
alyst for the highly selective hydrogenation of R to P1 in
scCO2 .21 Herein we report the successful application of a
Ru–Pt bimetallic catalyst supported on MCM-48 in scCO2
as a reaction medium for the hydrogenation of R. In particu-
lar, we elucidate the behavior of the Ru–Pt catalyst in scCO2
and the results are also compared with those in organic sol-
vents to assist in the development of an environmentally
friendly process.
Results and discussion
The XRD (Fig. 1) pattern of the bimetallic catalyst clearly
implied the cubic crystallographic space group Ia3d, indicating
that the mesoporosity of the material remain unaltered after
the in situ addition of Ru and Pt solutions. The TEM (Fig. 2)
also directly reveals the intact mesoporosity of the support
after the encapsulation of Ru and Pt. Massive agglomerations
of smaller particles are observed on the surface of the meso-
porous support. These agglomerates consist of many small
crystallites clustered together [Fig. 2(a)] to form larger aggre-
gates that are distributed randomly on the support surface.
The larger particles are manily located at the surface and the
smaller particles reside inside the pores. In addition to this,
the XPS study (results not shown) suggested the presence of
Ru and Pt as the +3 and metallic states, respectively, in the
calcined material.
Scheme 1 shows the general pathway and the products
detected during the hydrogenation of R in scCO2 under the
studied reaction conditions using supported noble metal cata-
lysts. Detailed results are summarized in Table 1. To keep the
H2 pressure and temperature of the reaction medium constant
the CO2 pressure has been changed from 7.0–17 MPa. Quanti-
tative formation of P1 was observed over the Ru–Pt bimetallic
catalyst in scCO2 medium, depending on the CO2 pressure.
Table 1 exemplifies the interesting CO2 pressure dependence
on P1 selectivity. At a constant temperature of 323 K the selec-
tivity to P1 rises to a maxima at around 7–7.5 MPa pressure of
CO2 , which then falls off as the pressure is increased further.
Clifford et al.22 found a similar maximum for the ratio of endo
to exo products in Diels–Alder reactions. From a theoretical
calculation they suggested that this maximum can be asso-
ciated with tuning of the average distance between the solute
molecules and the transition state as the pressure density
Experimental
Materials and methods
The bimetallic catalyst was prepared by an in situ addition of a
1 wt % solution of Ru and Pt salts in a mixture of surfactant
and silica precursor. The sources of silicon, Pt and Ru were
tetraethylorthosilicate (TEOS; Nacalai Tesque, Japan),
chloroplatinic acid (Aldrich) and ruthenium acetylacetonate
(Aldrich), respectively. Cetyltrimethylammonium bromide
(Merck) was used as a template in order to obtain the meso-
porous structure. A 1:1 ratio of Ru and Pt solutions has been
added to the starting gel containing template, sodium hydro-
xide and TEOS with constant stirring at room temperature.
Finally, the resultant gel was autoclaved at 413 K for 48 h.
The molar composition of the gel was
1 TEOS:0.27
CTABr:0.34 Na2O:62.5 H2O. The resultant material was fil-
tered, dried and then calcined at 823 K for 10 h in air. The cat-
alyst was primarily characterized by X-ray diffraction (XRD),
transmission electron microscopy (TEM) and X-ray photoelec-
tron spectroscopy (XPS).
Hydrogenation reaction
The hydrogenation of cinnamaldehyde (R) was carried out in a
50 ml stainless steel, high pressure, temperature-controlled,
stirred batch reactor. Catalyst (0.1 g) and 7.5 mmol of the reac-
tant was loaded into the reactor. Then the reactor was sealed
and flushed with CO2 to remove air. After flushing the reactor
was heated to the reaction temperature of 323 K. A prescribed
amount of hydrogen was first loaded into the reactor. Liquid
CO2 was charged into the reactor using a high- pressure liquid
pump and then compressed to the chosen pressure. Pressure
was kept constant by a back-pressure regulator in the system.
The hydrogen and the reaction mixture were stirred continu-
ously with a Teflon coated magnetic bar throughout the reac-
tion. After the reaction was complete, the reactor was cooled
by ice water and depressurized carefully using the back-pres-
sure regulator. The liquid mixture was identified by GC/MS
and analyzed quantitatively by GC (HP 5890) equipped with
a flame ionization detector. Quantification of the products
was obtained by a multipoint calibration curve for each pro-
duct. For the recycling studies the catalyst was separated from
the reactant and product and then mixed with fresh reactant.
Visual observation of the phase behavior in scCO2
As the reactor used in this reaction was closed observation of
the number of phase present during the reaction was not pos-
sible. A viewing cell was used to observe the phase behavior of
R and scCO2 . The cell (internal volume 10 ml) was equipped
with windows and connected to a temperature-controlled bath.
A magnetic bar was used to mix the contents of the cell from
the bottom. The phase behavior experiment of R in scCO2 was
Fig. 1 X-Ray diffraction patterns of calcined Ru–Pt-MCM-48 repre-
senting the structural order of cubic space group Ia3d (d211 ¼ 3.84 nm,
d220 ¼ 3.32 nm, d321 ¼ 2.79 nm, d400 ¼ 2.29 nm, d420 ¼ 2.40 nm,
d421 ¼ 2.05 nm, d332 ¼ 1.96 nm, d422 ¼ 1.87 nm.
New J. Chem., 2002, 27, 510–513
511

View Article Online
Table 1 Hydrogenation of cinnamaldehyde at different CO2 pressures
using Ru–Pt supported on MCM-48a
% Selectivity
Run
P(CO2)/MPa
% Conversion
P1
P2
P3
2.4
1
2
–
9.6
25.2
40.6
46.5
45.7
44.8
40.9
38.9
35.6
20.2
38.2
59.3
9.6
6.0
6.5
7.0
7.5
8.0
10.0
12.0
17.0
–
90.2
99.0
100.0
100.0
99.7
95.8
91.1
88.9
54.7
0.2
0.0
0.0
0.0
0.0
0.0
0.4
0.9
10.7
3
1.0
0.0
4
5
0.0
0.3
6
7
4.2
8.5
8
9
10b
10.2
34.6
a
Reaction conditions: catalyst ¼ 0.1 g (Ru ¼ 0.18 wt %, Pt ¼ 0.25 wt
%), cinnamaldehyde ¼ 7.5 mmol, pressure of H2 ¼ 4.0 MPa, time ¼
b
2 h, temperature ¼ 323 K. Isopropanol is used as reaction medium
instead of scCO2 .
presence of metallic platinum. Chemisorbed hydrogen can be
transferred from metallic Pt to a carbonyl group (as the surface
=
exposed metallic Pt is more favorable for C O adsorption),
Fig. 2 Transmission electron micrograph of calcined Ru–Pt-MCM-
48 (a); a closer view of the larger crystal is also shown (b).
=
resulting in the hydrogenation of C O, followed by formation
of the unsaturated alcohol. Compared to previous data21,25 the
activity and selectivity to P1 increases upon the addition of Ru.
Concerning the role of Ru, it can be suggested that the more
electropositive Ru transfers an electron to Pt26 to cause elec-
varies; for a compressed liquid such as a supercritical fluid the
variation of density causes a change in physical or chemical
equilibrium and affects the activity and selectivity of a reaction.
Oakes et al.23 also observed a significant enhancement of dis-
tereoselectivity from ꢂ35% to ꢂ95% for sulfoxidation of a
cysteine derivative upon varying pressure (ꢂ6–18 Mpa).
Now, it is well established that the density of scCO2 changes
with pressure. For instance, for a change of pressure from 6
=
trophilic activation to C O, leading to selective hydrogenation
of that bond. Further hydrogenation of the unsaturated
alcohol to saturated alcohol (P3) remains very low under the
applied reaction conditions. Above a CO2 pressure of 10
MPa (runs 7 to 10) the increasing amount of CO2 dilutes the
R concentration.22 It has already been said that the mode of
adsorption on the catalyst surface can depend on the R con-
centration27 At lower concentrations more space is available
to 8 MPa the density changes from 160.0 to 520.8 kg mꢁ3 24
.
Therefore, CO2 pressure depending complete selectivity to
the cinnamyl alcohol in scCO2 can be appropriately related
to the density of the medium. The actual reason for this effect
may be debatable but there is no doubt that pressure effects
have a strong influence on the selectivity.
The fact that marked decrease in activity and selectivity of
the reaction was observed in organic solvent (Table 1, Run
10) strongly suggests that the reaction medium plays an impor-
tant role in the activity and selectivity of the reaction. The
effect of scCO2 becomes more prominent when the reaction
is carried out only in the presence of hydrogen, resulting in a
much lower conversion as well as selectivity (Run 1).
The nature of the metal can also be an important factor to
increase the selectivity of the reaction, since lower selectivity
to P1 was observed with Ru alone (substrate conversion was
slow and gave 56.5% selectivity). For the studied catalyst the
physicochemical characterization gives firm evidence for the
=
for adsorption. Hence, R preferably adsorbs through C C
on the catalyst surface and the selectivity is shifted to P2.
Supercritical carbon dioxide is completely miscible with
hydrogen,28 leading to an increase in hydrogen concentration
at the catalyst. Thus, the influence of hydrogen pressure has
been considered to be an important factor to direct the activity
and selectivity of the reaction. At a constant pressure of CO2 (7
MPa), an enhancement in the activity has been observed with
increasing pressure of hydrogen (Fig. 3), increasing to 68.7% at
6 MPa of H2 . Moreover, considering the effect of temperature
the catalyst is unimpaired at the studied reaction temperature,
because higher selectivity and conversion are observed. Similar
Fig. 3 Hydrogen pressure dependence of the activity and selectivity
of cinnamaldehyde hydrogenation in scCO2 .
Scheme 1
512
New J. Chem., 2002, 27, 510–513

View Article Online
Chemistry of Double Bonded Functional Groups, Wiley, New York,
1997.
(a) Z. Poltarzewski, S. Galvagno, R. Pietropaolo and P. Staiti,
J. Catal., 1986, 102, 190; (b) F. Coloma, A. Sepu´lveda-Escarbino,
J. L. G. Fierro and F. Rodrignez, Appl. Catal. A, 1996, 148, 63.
conversion and selectivity were obtained during the additional
two cycles of recycling studies.
9
Conclusion
10 (a) G. Cordier, Y. Colleuille and P. Fouilloux, in Catalyse par les
´
Metaux, eds. B. Imelik et al., Editions du CNRS, Paris, 1984,
Supercritical carbon dioxide is shown to be a versatile reaction
medium for the selective hydrogenation of R to P1 using a
MCM-48 supported Ru–Pt bimetallic catalyst. Most signifi-
cantly, the product distribution obtained in scCO2 differs
remarkably from that in organic solvents. To the best of our
knowledge, compared to other heterogeneous catalysts in
scCO2 medium, this is the best selectivity to the P1 that has
ever been achieved. Particular advantages can be obtained by
exploiting the unique properties of the scCO2 medium. This
pertains to a simplified and convenient workup for selective
hydrogenation reactions by simply tuning the density.
Together with its environmentally and toxicologically benign
character, these characteristics sum up to a very attractive sol-
vent profile for CO2 . Therefore, the simple catalyst methodol-
ogy presented here could be highly relevant for the further
development of clean chemical process based on heteroge-
neous catalysts in scCO2 medium.
p. 349; (b) P. Gallezot and D. Richard, Catal. Rev. Sci. Eng.,
1998, 40, 81; (c) P. Claus, Top. Catal., 1988, 5, 51.
11 D. V. Sokol’skii, N. V. Anisimova, A. K. Zharmagambetova and
A. Ualikhanova, React. Kinet. Catal. Lett., 1981, 17, 419.
12 (a) S. Galvagno, Z. Poltarzewski, A. Donato, G. Neri and R.
Pietropaolo, J. Chem. Soc., Chem. Commun., 1986, 1729; (b) M.
Englisch, A. Jentys and J. A. Lercher, J. Catal., 1997, 166, 25.
13 D. Goupil, P. Fouilloux and R. Maurel, React. Kinet. Catal. Lett.,
1987, 35, 185.
14 (a) S. Galvagno, A. Donato, G. Neri and R. Pietropaolo, Catal.
Lett., 1991, 8, 9; (b) S. Galvagno, A. Donato, G. Neri, G.
Campannelli and R. Pietropaolo, J. Mol. Catal., 1993, 78, 237.
15 M. K. Vennice and B. Sen, J. Catal., 1989, 115, 65.
16 P. Gallezot, B. Blanc and M. I. Pais da Silva, Stud. Surf. Sci.
Catal., 1994, 84, 1433.
17 P. Gallezot, A. Giroir-Fender and D. Richard, Catal. Lett., 1990,
4, 169.
18 G. Neri, L. Mercadante, A. Donato, A. M. Visco and S.
Galvagno, Catal. Lett., 1994, 29, 379.
´
´
´
19 (a) A. Corma, A. Martınez, V. Martınez-Soria and J. B. Monton,
J. Catal., 1995, 153, 25; (b) A. Sayari, Chem. Mater., 1996,
8, 1840.
ˆ
ˆ
20 V. I. Parvulescu, V. Parvulescu, S. Kaliaguine, U. Endruschat, B.
Tesche and H. Bonnemann, Catalysis of Organic Reactions,
Chemistry and Industry series, ed. M. E. Ford, Marcel Dekker,
NY, 2001, vol. 82, p. 301.
References
1
2
3
4
B. Subramaniam and M. A. Mchuge, Ind. Eng. Chem. Proc. Des.
Dev., 1986, 25, 1.
P. E. Savage, S. Gopalan, T. I. Mizan, C. J. Martino and E. E.
Brock, AIChE, 1995, 41, 1723.
Y. Ikushima, N. Saito and M. Arai, Bull. Chem. Soc. Jpn., 1991,
64, 282.
Y. Ikushima, N. Saito and M. Arai, J. Phys. Chem., 1992, 96,
2293.
21 M. Chatterjee, Y. Ikushima and F.-Y. Zhao, Catal. Lett., 2002,
82, 141.
22 A. A. Clifford, K. Pople, W. J. Gaskill, K. D. Bartle and C. M.
Rayner, J. Chem. Soc., Faraday Trans., 1998, 94, 1451.
23 R. S. Oakes, A. A. Clifford, K. D. Bartle, M. T. Pett and C. M.
Rayner, Chem. Commun., 1999, 247.
24 M. Arai, Y. Nishiyama and Y. Ikushima, J. Supercrit. Fluids,
1998, 13, 149.
25 B. M. Bhanage, Y. Ikushima, M. Shirai and M. Arai, Catal. Lett.,
1999, 62, 175.
5
6
M. E. Sigman and J. E. Leffler, J. Org. Chem., 1987, 52, 3123.
J. B. Ellington and J. F. Brennecke, J. Chem. Soc., Chem. Com-
mun., 1993, 1094.
7
(a) P. G. Jessop, T. Ikariya and R. Noyori, Nature (London),
1994, 368, 231; (b) P. G. Jessop, Y. Hsiano, T. Ikariya and R.
Noyori, J. Am. Chem. Soc., 1996, 118, 344; (c) M. J. Bank, S.
Feng, M. F. Gross and W. Tumas, J. Am. Chem. Soc., 1995,
117, 8277.
26 D. Richard, J. Couves and J. M. Thomas, Fraday Discuss., 1991,
92, 109.
27 X. Jin, Z. Ma, J. Talbot and N.-H. L. Wang, Langmuir, 1999, 15,
3321.
28 L. J. P. van den Broeke, E. L. V. Goetheer, A. W. Verkerk, E. de
Wolf, B.-J. Deelman, G. van Koten and J. T. F. Keurentjes,
Angew. Chem., Int. Ed., 2001, 40, 4473.
´
´
M. Bartok and A. Molnar, Heterogeneous Catalytic Hydrogena-
tions in Chemistry of Functional Groups, Supplement A3: The
´
8
New J. Chem., 2002, 27, 510–513
513