Mono and bifunctional catalysts for styrene oxide isomerization or hydrogenation

Article abstract of DOI:10.1007/s10562-010-0543-5

Styrene oxide can be effectively isomerized to phenyl-acetaldehyde (98%) over amorphous silica alumina catalysts under very mild liquid phase conditions. On the other hand, a copper catalyst prepared using a silica zirconia support gave up to 80% yield in the hydrogenation of styrene oxide to 2-phenyl-ethanol. Graphical Abstract: [Figure not available: see fulltext.]

Full text of DOI:10.1007/s10562-010-0543-5

Catal Lett (2011) 141:587–591
DOI 10.1007/s10562-010-0543-5
Mono and Bifunctional Catalysts for Styrene Oxide Isomerization
or Hydrogenation
•
Federica Zaccheria Rinaldo Psaro
•
•
Nicoletta Ravasio Laura Sordelli
•
Federica Santoro
Received: 5 October 2010 / Accepted: 30 December 2010 / Published online: 20 January 2011
Ó Springer Science+Business Media, LLC 2011
Abstract Styrene oxide can be effectively isomerized to
phenyl-acetaldehyde (98%) over amorphous silica alumina
catalysts under very mild liquid phase conditions. On the
other hand, a copper catalyst prepared using a silica zir-
conia support gave up to 80% yield in the hydrogenation of
styrene oxide to 2-phenyl-ethanol.
loading Al2O3 and ZrO2 on silica can effectively substitute
ZnCl2 in the isomerization of a-pinene oxide to canpho-
lenic aldehyde, an intermediate in the synthesis of santalol,
a possible substitute for environmentally unfriendly musks,
while the two isomers of limonene oxide gave two different
products [2].
We also reported that copper deposited on this kind of
solids is an effective catalyst in bifunctional transforma-
tions such as intra- [3] and inter-molecular [4] ether for-
mation starting from a ketone and an alcohol.
Keywords Solid acids Á Bifunctional catalysts Á Cu
catalysts Á Epoxides ring opening Á Lewis acids Á
Hydrogenation
Here we wish to report our results in the isomerization
of styrene oxide to phenylacetaldehyde over amorphous
mixed oxides and in its hydrogenation to 2 phenyl-ethanol
over a bifunctional supported Cu catalyst.
1 Introduction
In specialty chemicals synthesis, the need of low waste
technologies has recently increased the interest in solid
acids as substitutes for both mineral acids and homoge-
neous Broensted and Lewis ones. On the other hand,
epoxides are among the more versatile and important
intermediates in pharmaceutical and agrochemical indus-
tries. Most useful transformations are isomerization to
carbonyl compounds, hydrogenation and nucleophilic
addition of nitrogen, oxygen and sulfur compounds.
The most commonly used Lewis acid to promote the
acid catalyzed rearrangement of epoxides is BF3, usually as
the etherate, which is rather a reagent than a catalyst.
We already reported on the use of amorphous meso-
porous mixed cogels and of Fe based catalysts [1] in the
isomerization of epoxides. In particular, we found that low
Phenylacetaldehyde, used as semiochemical and in
perfume compositions for hyacinth notes, can be obtained
in high yield by vapor phase isomerization of styrene oxide
over alkali treated silica alumina [5] or hydrotalcite derived
basic solids [6]. It can also be obtained by oxidation of
2-phenylethanol but the product purity never exceeds 85%
as the aldehyde readily polymerizes under reaction condi-
tions. In the liquid phase Ti–silicalite TS-1 gives a 95%
yield in 1 h at 70 °C, showing that this reaction requires
only weak Lewis acid sites [7].
2-phenylethanol is also a valuable molecule for the
Flavours and Fragrances industry and it can be produced on
an industrial scale through Friedel–Crafts reaction of ben-
zene and ethylene oxide in the presence of AlCl3 and
through hydrogenation of styrene oxide over Raney Ni and
NaOH. In particular, 99.3% selectivity can be achieved
using Raney Nickel with 48% aqueous alkali at 20 atm of
H2 and 40–50 °C [8].
F. Zaccheria Á R. Psaro Á N. Ravasio (&) Á L. Sordelli Á
F. Santoro
Istituto di Scienze e Tecnologie Molecolari del CNR (CNR
ISTM), via C.Golgi, 19, 20133 Milan, Italy
e-mail: n.ravasio@istm.cnr.it
Both methods rise concerns from the environmental
point of view, therefore a catalyst based on a non toxic
metal, not requiring basic additives should be an interesting
123

588
F. Zaccheria et al.
one. However, though a number of patents are available,
the published information is conspicuously absent. Plati-
num and palladium catalysts gave excellent results both in
gas [9, 10] and in liquid phase [11, 12] but the use of bases
as promoter or of superatmospheric pressure of H2 is
required, while a catalyst prepared by calcining a borosil-
icate with Cu(NO3)2 was found to give 78.4% of PEA
under gas phase conditions [13].
TPR profiles were recorded with a modified version of
the Micromeritics Pulse Chemisorb 2700 apparatus; cata-
lysts (25 mg) were diluted with an equal amount of quartz,
calcined at 500 °C under O2 (40 mL/min) and then reduced
at 8 °C/min with a 8% H2/Ar mixture at 15 mL/min.
The X-ray characterization was performed at the XAFS
beamline of the Elettra Synchrotron facility in Trieste
(Italy) with a Si(311) double crystal monochromator. EX-
AFS and XANES spectra of Cu catalysts and of Cu ref-
erence systems (Cu foil for constant angle/energy
calibration, Cu2O and CuO) were collected at the Cu K
edge (8979 eV) in transmission mode during in situ cal-
cination and reduction treatments (carried out inside an
EXAFS-catalysis cell, designed to work in transmission
mode over powder sample under controlled gas flow and
temperature). At the end of each treatment spectra were
recorded at both room and liquid nitrogen temperature.
XANES spectra have been recorded with a sampling of
0.25 eV. EXAFS spectra have been recorded over a
900 eV range. The spectra were analysed with the IFEFFIT
software package and the best fitting results are reported in
Table 2 (coordination numbers, distance and Debye–Wal-
ler factors of Cu–Cu first neighbours shell) together with
the estimated metal particle size.
A preliminary investigation on the potential of sup-
ported copper catalyst in this reaction and on the influence
of some experimental conditions may then be useful.
2 Experimental Section
All the mixed oxides used were kindly supplied by Grace
Davison (Worms, Germany). Their textural properties are
reported in Table 1.
Solids acids were pretreated at 270 °C for 20 min in air
and for 20 min under reduced pressure at the same
temperature.
Copper catalysts, with a 8–9% metal loading, were
prepared as follows: the support (10 g) was added to a
0.7 M [Cu(NH3)4]2? solution prepared starting from
Cu(NO3)2Á3H2O (4 g) dissolved in 25 mL of water and
adding NH4OH until pH 9 [14]. After 20 min under stir-
ring, the slurry, held in an ice bath at 0 °C, was slowly
diluted in order to allow hydrolysis of the copper complex
and deposition of the finely dispersed product to occur.
Under these conditions, no dissolution of silica was
detected. The solid was separated by filtration, washed with
0.5 L of water, dried in oven overnight at 120 °C, and
calcined in static in air at 350 °C for 4 h. In all cases the Cu
content was ca. 8 wt% (Table 1). Other relevant surface
and textural properties of the samples, BET specific surface
area, pore volume and size are listed in Table 1. Before
reaction, catalysts were treated for 20 min in air at 270 °C
and for 20 min under rotary vacuum and thereafter reduced
in static at the same temperature with 1 atm of pure H2.
Catalytic tests Styrene oxide (C97%, 200 mg), obtained
from Aldrich was dissolved in the solvent (8 mL) and the
solution transferred under N2 into a glass reaction vessel in
which the catalyst had been previously activated. In the
case of hydrogenation reactions N2 was removed under
vacuum and substituted with H2. Reactions were carried
out under atmospheric pressure and magnetic stirring
(1400 rpm).
Reaction mixtures were analyzed by GC–MS using a
HP5 (5% phenyl)-methyl-polysiloxane capillary column,
length 30 m (injection T = 60 °C).
Recycling tests were carried out as follows: 200 mg SiAl
0.6, 2 mL styrene oxide, 8 mL toluene, 90 °C, N2, 0.5 h.
After each run the solution is withdrawn by means of a
syringe, the catalyst washed with 2 mL toluene under
Table 1 Textural features of
Catalyst
Co-oxide
loading (wt%)
Acronym
BET
(m2/g)
PV
(mL/g)
APR
˚
(A)
different cogels used as acids
and of the corresponding Cu
catalysts
SiO2
–
Si
480
363
483
430
400
412
453
382
304
331
0.75
0.68
1.43
0.86
0.77
0.75
0.91
0.75
1.62
1.05
60
72
8% Cu/SiO2
Cu/Si
SiO2–Al2O3
0.6
13
1
SiAl 0.6
Cu/SiAl 0.6
SiAl 13
Cu/SiAl 13
SiZr 1
117
80
8% Cu/SiO2Al2O3 0.6
SiO2–Al2O3
33
8% Cu/SiO2Al2O3 13
SiO2–ZrO2
37
80
8% Cu/SiO2ZrO2 1
SiO2–ZrO2
Cu/SiZr 1
SiZr 4.7
Cu/SiZr 4.7
79
4.7
106
64
8% Cu/SiO2ZrO2 4.7
123

Mono and Bifunctional Catalysts
589
O
Table 2 Curve fitting results of Cu K edge EXAFS data (at the end
of in situ reduction, first Cu–Cu shell)
O
a
D (A)
˚
R (A)
˚
˚
SAMPLE
CN
Dr (A)
DEo (eV)
H
1
2
CuSi
7.6
9.8
8.7
8.8
2.560
2.547
2.548
2.549
0.074
0.068
0.066
0.076
-0.3
0.1
17
30
21
22
CuSiAl 0.6
CuSiZr 1
CuSiZr 4.7
a
Scheme 1 Isomerization of styrene oxide
0.3
0.4
2-methoxy-2-phenylethanol but produces only very low
amount of phenylacetaldehyde and the corresponding
acetal [20]. However, the activity of SiZr materials was
much higher than that of pure silica.
Estimated Cu particle size diameter
stirring for 5 min, the solvent removed and a fresh solution
of the substrate added.
The silica alumina catalyst is also reusable up to four
times without re-activation, as it is shown in Fig. 1. After
four cycles a dramatic drop in activity was observed
although selectivity kept constant.
3 Results and Discussion
The high selectivity observed toward formation of
phenylacetaldehyde prompted us to use this kind of mate-
rials as support for a catalyst to be used in the hydroge-
nation of styrene oxide to 2-phenylethanol. All the 8% Cu
(by weight) catalysts prepared by using the acidic materials
as supports, were found to show a reduction profile very
similar to that of Cu/SiO2 but the one supported on SiAl 13
(Fig. 2).
Results obtained in the isomerization of styrene oxide to
phenylacetaldehyde over a series of amorphous, commer-
cial mixed cogels are summed up in Table 3. All the silica
alumina catalysts showed high activity, in agreement with
the well known Broensted and Lewis acidity of these
materials [15–19] but the most selective one was a silica
with a very low content in alumina giving up to 98%
phenylacetaldehyde under very mild experimental condi-
tions, in agreement with the high dispersion of Al Lewis
sites on the surface. These results are comparable to those
obtained by the ANIC group by using a Ti containing
zeolite under very similar experimental conditions [7] and
ascribed to the presence of weak Lewis acid sites. The
main difference is represented by the solvent, TS-1
requiring the use of polar solvents like alcohols and
ketones (Scheme 1).
This is in agreement with the fact that Cu on this catalyst
is mainly ionic [21] therefore this material was not tested in
the hydrogenation reaction.
XANES spectra of the reduced catalysts show that only
metallic Cu is present on the surface of all materials.
Table 2 reports the curve fitting results of the Cu K edge
EXAFS data that allow one to estimate the metal particle
size on in situ reduced samples. Very small Cu particles,
similar in size to those supported on pure silica are present
on all the catalysts tested.
Productivity values reported in Table 3 show on the
contrary that silica zirconia mixed oxides, even with a very
low content of ZrO2, are poorly active in this reaction. This
is consistent with the observation that zirconium doped
mesoporous silica promotes the methanolysis of styrene to
Results on the hydrogenation reaction (Scheme 2) are
reported in Table 4. In this reaction, Cu/SiAl 0.6 gave very
poor results, selectivity to phenetyl alcohol never exceed-
ing 68%. Significant amounts of products with high MW,
Table 3 Isomerization of styrene oxide over different acidic catalysts
Entry
Catalyst
C/Sa (wt)
t (h)
Conv. (%)
Sel to 2 (%)
Productivity
(mmol prod/gcat h)
1
2
4
4
5
6
7
8
SiAl 13
SiAl 0.6
SiAl 0.6
SiAl 0.6
SiZr 4.7
SiZr 1
1/10
1/10
1/20
1/30
1/10
1/10
1/1
0.5
0.5
0.5
3.5
6
100
100
100
78
92
95
98
97
72
95
65
85
153.3
158.3
326.7
51.5
10.0
11.3
11.8
3.1
100
100
100
26
7
SiZr 1
0.5
6
SiO2
1/10
a
90 °C, 1 bar N2, toluene; Catalyst to substrate ratio
123

590
F. Zaccheria et al.
conv.%
% 2
experimental conditions, show that in fact the same product
distribution was obtained over Cu/SiAl 0.6. A similar
product distribution was found also with Cu/SiZr 4.7. This
may be due to relevant penetration of the metal in the
mesopores of these two materials, significantly reducing
the average pore radius (Table 1). Work is in progress in
order to elucidate relationships between C–C and C–O
coupling activity and the presence of Cu in the bigger
pores. Silica was also tested for the sake of comparison as
the catalyst support, but in this case the reaction gave huge
amounts of ethylbenzene, likely formed through dehydra-
tion of 3. This catalyst has already shown very high activity
in hydrogenation reactions often leading to deoxygenated
products, particularly in the case of aromatic compounds
[23]. Intermediate formation of 1-phenylethanol was ruled
out as this alcohol, under the same experimental conditions
and over both Cu/SiO2 and Cu/SiZr catalysts gave small
amount of styrene whereas the symmetric ether was found
to be the main product.
100
80
60
40
20
0
1
2
3
4
5
run
Fig. 1 Conversion and selectivity in recycling tests (see Sect. 2)
The best catalyst appeared to be Cu/SiZr 1. This catalyst
at 90 °C in toluene gives 70% of the desired product,
limited hydrogenolysis to ethyl-benzene and limited for-
mation of high MW byproducts. In order to increase the
selectivity to 2-Ph-ethanol we examined four parameters:
Cu loading, reaction T, catalyst activation T and solvent
effect. Lowering the metal loading to 5% gave a catalyst
with poor hydrogenation activity that for long reaction
times converted 2 into high MW products but not into 3.
However it is interesting to note that adding Cu to SiZr 1 a
catalyst more selective than the pure support in styrene
epoxide isomerization was obtained (Table 4, entry 8 vs.
Table 3, entry 7). Lowering reaction and activation T had
both a negative effect on this reaction, whereas the only
positive effect was found by changing the solvent.
Cu/SiAl 13
Cu/SiZr 4.7
Cu/Si
Cu/SiZr 1
Cu/SiAl0.6
0
50
100 150 200 250 300 350 400 450 500
T (°C)
Fig. 2 TPR profiles of copper supported catalysts prepared over
different mixed oxides
in particular C–O coupling dimers such as 2,5-diphenyl-
1,4-dioxane were formed, as already observed when the
reaction was carried out in the presence of homogeneous
cationic Rh complexes [22]. Parallel experiments carried
out using phenylacetaldehyde as substrate under the same
In acetonitrile no reaction occurred, in agreement with
the Lewis base character of this solvent. In 2-propanol
Table 4 Hydrogenation of styrene oxide to phenetyl alcohol
Entry
Catalyst
Solvent
t (h)
3 (%)
2 (%)
4 (%)
5 (%)
Others (%)
6 (%)
1
Cu/SiAl 0.6
Cu/SiAl 0.6
Cu/Si
Toluene
1.5
2.5
3
59
68
32
70
20
64
80
–
1
0
6.5
2.2
43
33.5
27.4
4.5
14.4
16.3
8.4
5.8
8
–
2
Dioxane
Toluene
2
0
–
3
14
1
6.8
0.2
0
–
4
Cu/SiZr 1
Cu/SiZr 1
Cu/SiZr 1
Cu/SiZr 1
5% Cu/SiZr 1
Toluene
4.25
3
13.8
4.2
6.9
12.3
0.2
1.0
1.2
7
–
5
2-PrOH
0
58.7
18
–
6
2-PrOH ?diox.
Dioxane
Dioxane
Dioxane
Toluene
2.5
5
3
0
7
2
0
8
0.5
4
89
33
35
66
1.6
0.98
0.6
1.4
–
9
8
57
–
10
11
Cu/SiZr 4.7
0.5
3
10.8
14
52
–
Dioxane
11
–
8% Cu loading, 90 °C, 1 bar H2, C/S = 1 by weight, 1000 rpm
123

Mono and Bifunctional Catalysts
591
OH
O
3. Ravasio N, Leo V, Babudri F, Gargano M (1997) Tetrahedron
Lett 38:7103
4. Zaccheria F, Psaro R, Ravasio N (2009) Tetrahedron Lett
50:5221
5. Matsuzawa Y, Yamashita T, Ninagawa S (1972) JP 47042278 to
Ajinomoto. CAN 78:135878
6. Kochkar H, Clacens JM, Figueras F (2002) Catal Lett 78:91
7. Neri C, Buonomo F (1985) US Patent 4,495,371 to ANIC SpA.
CAN 101:23130
8. Hirako Y (1987) JP 62,286,939 to Daicel Chemical Industries.
CAN 108:152560
9. Kirm I, Medina F, Rodriguez X, Cesteros Y, Salagre P, Sueiras
JE (2005) J Mol Catal A Chem 239:215
10. Kirm I, Medina F, Sueiras JE, Salagre P, Cesteros Y (2007) J Mol
Catal A Chem 261:98
O
+
OH
+
+
6
4
5
1
3
Scheme 2 Products of the hydrogenation reaction
mainly the corresponding b-alkoxy alcohol was formed due
to a nucleophilic addition mechanism at the more substi-
tuted C atom, while in dioxane a significant increase in
selectivity was obtained and the product was formed in
80% yield.
11. Chaudari RV, Telkar MM, Rode CV (2000) US Patent 6,166,269
to CSIR, India. CAN 134:41976
4 Conclusions
12. Yadav VG, Chandalia SB (1998) Org Proc Res Dev 2:294
13. Hoelderich W, Goetz N, Hupfer L (1990) US patent 4,943,667 to
BASF AG. CAN 112:55217
Commercial mixed oxides can be conveniently used for the
opening of styrene oxide to phenylacetaldehyde with good
selectivity and productivity. Copper catalysts supported on
the same acidic materials show moderate selectivity in the
hydrogenation of styrene oxide to phenetylalcohol. How-
ever, reaction conditions were very mild with respect to
other reported systems, particularly as far as H2 pressure is
concerned, and no basic additives were needed. Moreover,
the study of different supports and different experimental
conditions made available a number of informations that
may be useful also in other transformations.
14. Boccuzzi F, Martra G, Coluccia S, Ravasio N (1999) J Catal
184:316
15. Chizallet C, Raybaud P (2010) Chem Phys Chem 11:105
16. Fihri A, Durand R, di Renzo F, Fajula F, Coq B, Roussel TM,
Vuillemin B (2010) Appl Catal B Environ 98:224
17. Dragoi B, Gervasini A, Dumitriu E, Auroux A (2004) Thermo-
chim Acta 420:127
18. Gervasini A, Fenyvesi J, Auroux A (1997) Catal Lett 43:219
19. Bevilacqua M, Montanari T, Finocchio E, Busca G (2006) Catal
Today 116:1
20. Barreca D, Copley MP, Graham AE, Holmes JD, Morris MA,
Seraglia R, Spalding TR, Tondello E (2006) Appl Catal A Gen-
eral 304:14
21. Gervasini A, Manzoli M, Martra G, Ponti A, Ravasio N, Sordelli
L, Zaccheria F (2006) J Phys Chem B 110:7851
22. Fujitsu H, Shirahama S, Matsumura E, Takeshita K, Mochida I
(1981) J Org Chem 46:2287
Acknowledgments The ELETTRA synchrotron light laboratory in
Basovizza (Trieste, Italy) is gratefully acknowledged for providing
the beamtime.
23. Zaccheria F, Ravasio N, Ercoli M, Allegrini P (2005) Tetrahe-
dron Lett 46:7743
References
1. Ravasio N, Zaccheria F, Gervasini A, Messi C (2008) Catal
Commun 9:1125
2. Ravasio N, Zaccheria F, Guidotti M, Psaro R (2004) Top Catal
27:157
123