Effective alkene epoxidation with dilute hydrogen peroxide on amorphous silica-supported titanium catalysts

Article abstract of DOI:10.1039/b000929f

The use of amorphous silica-supported titanium catalysts in which the titanium ions display a chemical environment similar to that of Ti- substituted zeolites, afforded excellent activity in the epoxidation of terminal linear and bulky alkenes with dilute solutions of hydrogen peroxide.

Full text of DOI:10.1039/b000929f

Effective alkene epoxidation with dilute hydrogen peroxide on amorphous
silica-supported titanium catalysts
M. C. Capel-Sanchez,a J. M. Campos-Martin,a J. L. G. Fierro,*a M. P. de Frutosb and A. Padilla Polob
a Instituto de Catálisis y Petroleoquímica, CSIC, Cantoblanco, 28049 Madrid, Spain. E. mail: jlgfierro@icp.csic.es
b Centro de Investigacion y Desarrollo REPSOL-YPF, Embajadores 183, 28045 Madrid, Spain
Received (in Cambridge, UK) 2nd February 2000, Accepted 5th April 2000
Table 1 Characteristics of the catalysts
The use of amorphous silica-supported titanium catalysts in
which the titanium ions display a chemical environment
Titanium
content
(%) (XPS)
XPS ratio
(460 eV/
458.5 eV)
similar to that of Ti-substituted zeolites, afforded excellent
activity in the epoxidation of terminal linear and bulky
alkenes with dilute solutions of hydrogen peroxide.
Catalyst
Synthesis
C1
C2
C3
TS-11
TiF/SiO2
Cyclohexanol
Toluene
n-Decane
0.8
2.6
2.6
1.7
1.1
2.1
0.5
0.1
2.5
Despite numerous reports in the literature, the epoxidation of
terminal alkenes remains a challenge in petrochemistry. Many
different methods have been developed for the preparation of
epoxides. However, in recent years maximum interest has
focused on titanium-substituted zeolites in the framework,
6
a
—
a Not determined owing to the presence of fluorinated species.
2
including TS-1,1 Ti-b and Ti-incorporated mesoporous silica,3
which are claimed to be good catalysts for epoxidation with
hydrogen peroxide. Although in the presence of methanol as a
solvent all these catalysts display high activity and selectivity
levels towards epoxide, they have poor mechanical strength and
low thermal stability. Non-zeolitic TiO2–SiO2-supported cata-
lysts are effective in the epoxidation of alkenes with organic
hydroperoxides,4,5 but it is generally believed that they are not
effective in the epoxidation of alkenes with hydrogen peroxide.
Nevertheless, titanium silica-supported catalysts have been
reported to be active to the epoxidation of alkenes with
hydrogen peroxide, but with low selectivity towards epoxide.6
Here we report a very simple route for the preparation of
TiO2/SiO2 catalysts with highly dispersed titanium atoms.
These catalysts are relatively inexpensive, easy to synthesise
and regenerate, and at the same time show high conversion rates
without the requirement of methanol as a solvent. Hitherto,
these materials have not been used as catalysts in the
epoxidation of alkenes (oct-1-ene, cyclohexene and norbor-
nene) with hydrogen peroxide solutions.
210 nm, which can be taken as a clear indication of isolated
Ti(IV) in tetrahedrally coordinated sites. The slight shift in band
position and the increase in bandwidth point to a distorted
tetrahedral environment of the titanium. The broad adsorption
band of the silica-supported samples shifted to high wavelength
values; this can be explained in terms of: (i), the presence of
titanium atoms in octahedral coordination (values between 290
and 333 nm),9 as observed in catalysts C2 and C3, and as a small
component in TiF/SiO2; (ii) the hydrophilic nature of the silica
surface, which favours the presence of hydrated titanium in
tetrahedral coordination (232 nm),9 as already seen for catalyst
C1 and the major proportion of TiF/SiO2. However, the absence
of a band at 370–410 nm rules out the presence of free TiO2 in
all the catalysts.9
High resolution photoelectron spectra of Ti 2p core-levels of
the in situ outgassed samples (Fig. 2) displayed the character-
istics Ti 2p1/2/Ti 2p3/2 doublet. Although chemical information
can be derived from each component, discussion should be
based on the most intense Ti 2p3/2 component. Curve fitting of
The catalysts were prepared in the following manner:7 0.75 g
of titanium isopropoxide was dispersed in the solvent (either n-
decane, toluene or cyclohexanol to produce catalysts C3, C2
and C1, respectively) (150 ml). The solution was heated to
423 K (the boiling point of toluene solution) under stirring and
then 5 g of silica (Grace Davison G-952, surface area:
310 m2 g21, pore volume: 1.5 ml g21) was added to the solution,
which was stirred at 423 K for 2 h. The solid obtained was
filtered off and washed twice with 150 ml of hot solvent and
subsequently dried at 383 K and calcined at 773 K for 5 h. As
reference catalysts, a titanium-supported silica (TiF/SiO2), as
described by Jorda et al.,6 and a titanium silicalite (TS-1), as
reported by Taramaso et al.,1 were prepared.
The total amount of titanium introduced into the synthesis
solution was incorporated to catalysts TiF/SiO2, C2 and C3
(Table 1). However, when cyclohexanol was employed as the
solvent (C1), only part of the titanium was incorporated into the
catalyst. For TS-1, the degree of titanium incorporation was
similar to that reported in the literature. The structural FTIR
spectra of the catalysts diluted in KBr were inconclusive
because the silica support displayed a broad peak at 960 cm21
,
which overshadowed the Ti–O–Si vibration mode
(960 cm21).8
The DRS UV–VIS spectra (Fig. 1) of catalysts C1, C2, C3
and TiF/SiO2 were clearly different from that of the TS-1
sample. The spectrum of the latter showed a single peak at
Fig. 1 DRS UV–VIS spectra of solids in ambient conditions recorded with
a Shimadzu UV-2100 spectrometer.
DOI: 10.1039/b000929f
Chem. Commun., 2000, 855–856
This journal is © The Royal Society of Chemistry 2000
855

Table 2 Data for alkene epoxidation with hydrogen peroxide after 1 h of
reaction (T = 353 K)
Conversion
of H2O2 (%)
Selectivity to
epoxidea (%)
Catalyst
Alkene
TS-1
TiF/SiO2
C3
C2
C1
Oct-1-ene
Oct-1-ene
Oct-1-ene
Oct-1-ene
Oct-1-ene
Cyclohexene
Norbornene
96
75
96
97
97
98
98
70
65
32
56
95
91
98
C1
C1
a Selectivity to epoxide = mol of epoxide formed/mol of H2O2 con-
sumed.
values towards epoxide, which is consistent with the require-
ment of the presence of methanol to obtain high selectivity
values.11 The method for preparing silica-supported titanium
had a strong effect on performance for the epoxidation of oct-
1-ene. The TiF/SiO2 catalyst showed similar conversion and
selectivity values to those reported in previous work,6 but lower
than those of TS-1. Catalysts C3 and C2 showed very low
selectivity towards epoxide, in agreement with the high
proportion of octahedrally coordinated titanium, as revealed by
DRS UV–VIS and XPS spectroscopic techniques. Never-
theless, catalyst C1 had very high (95%) selectivity values
towards epoxide, related to the high proportion of titanium in
tetrahedral coordination as revealed by the DRS UV–VIS
technique, and specifically by XPS spectroscopy. Finally,
catalyst C1 also displayed very good performance in the
epoxidation of bulky alkenes with hydrogen peroxide. In
cyclohexene and norbornene, 91 and 98% selectivity towards
epoxide, respectively, were reached. Used catalyst C1 was
analysed and the Ti content found to be the same as in the fresh
sample, thus excluding leaching.
In short, using a very simple route, we have prepared
titanium-supported amorphous silica catalysts in which the
chemical environment of titanium atoms is very similar to Ti-
substituted zeolites. The use of cyclohexanol is pivotal to the
production of very active and selective catalysts in the reaction
of alkene epoxidation with hydrogen peroxide. These catalysts
also display excellent performance with linear as well as larger
or bulky alkenes.
Fig. 2 Ti 2p core-level spectra of outgassed samples in situ at 473 K
acquired with a VG Escalab 200R spectrometer.
the Ti 2p3/2 core level to two components shows the highest
binding energy component (460.0 eV) can be attributed to
titanium in tetrahedral coordination,10 while the lowest binding
energy component (458.5 eV) is usually assigned to titanium in
octahedral coordination10 or in interaction with surface hydroxy
groups. Catalyst C3 shows a large part of the titanium in
octahedral coordination (Table 1). For catalyst C2, the propor-
tion of titanium in tetrahedral sites is higher than in C3, but most
of the titanium still remains in octahedral coordination
(Table 1). Catalysts C1 and TS-1 exhibit a titanium peak in
tetrahedral coordination that is clearly more intense than the
lowest binding energy component. This indicates that titanium
ions are in a very similar environment in catalysts C1 and TS-1,
although the synthesis route of silica-supported titanium
catalyst (C1) is considerably easier than that for the Ti-silicalite
sample. Both DRS UV–VIS and XPS spectroscopic techniques
pointed to the necessity of carefully controlling the incorpora-
tion of titanium into silica (preparation of C1 in cyclohexanol)
to obtain titanium in tetrahedral coordination. This is due to the
fact that cyclohexanol interacts strongly with the titanium
precursor which favours the formation of isolated titanium
species, and also that only a fraction of the titanium added is
incorporated as a consequence of the equilibrium of adsorp-
tion.
These catalysts were tested in the epoxidation of alkenes with
dilute aqueous hydrogen peroxide solutions. Epoxidation was
performed under ambient pressure in a round-bottomed flask
equipped with a condenser and magnetic stirrer. The reaction
procedure was as follows: 0.2 mol of alkene, 11 g of tert-butyl
alcohol and 1 g of catalyst were stirred and heated to the
reaction temperature (353 K). Then, 4 g of a dilute solution of
hydrogen peroxide (6 wt% in 1-phenylethanol) was added
dropwise under stirring over 2 h. Aliquots were taken at regular
intervals. H2O2 consumption was evaluated by iodometric
titration, and organic compounds were analysed by GC to
determine their selectivity. The results are summarised in
Table 2. The reference catalyst TS-1 showed low selectivity
Notes and references
1 M. Taramaso, G. Perego and B. Notari, US Pat., 4410501, 1983.
2 P. J. Saxton, W. Chester, J. G. Zajacek, G. L. Crocco and K. S.
Wijesekera, US Pat., 5621122, 1997.
3 J. S. Reddy, A. Dicko and A. Sayari, Chem. Ind., 1997, 69, 405.
4 H. P. Wulff, US Pat., 3923843, 1975; M. Tamura, K. Yamauchi and K.
Uchida, Eur. Pat., 734764, 1996.
5 J. M. Fraile, J. I. Garcia, J. A. Mayoral, L. C. De Menorval and F.
Rachdi, J. Chem. Soc., Chem. Commun., 1995, 539.
6 E. Jorda, A. Tuel, R. Teissier and J. Kerneval, J. Chem. Soc., Chem.
Commun., 1995, 1775.
7 J. M. Campos-Martin and M. P. de Frutos, WO9948884, 1999.
8 E. Astorino, J. B. Peri, R. J. Willey and G. Busca, J. Catal., 1995, 157,
482.
9 F. Geobaldo, S. Bordiga, A. Zecchina, E. Giamello, G. Leofanti and G.
Petrini, Catal. Lett., 1992, 16, 109.
10 T. Blasco, M. A. Camblor, J. L. G. Fierro and J. Perez-Pariente,
Microporous Mater., 1994, 3, 259.
11 M. G. Clerici, G. Bellussi and U. Romano, J. Catal., 1991, 129, 159.
856
Chem. Commun., 2000, 855–856