Catalytic aerogel-like materials dried at ambient pressure for liquid-phase epoxidation

Article abstract of DOI:10.1039/a809012b

Titanium-containing hydrophobic silica with a pore volume and pore sizes in the range of aerogels was prepared by silylation before drying of well-dispersed mixed oxides obtained via a sol-gel method; it showed remarkable properties as a catalyst in the epoxidation of styrene with Bu¹OOH.

Full text of DOI:10.1039/a809012b

Catalytic aerogel-like materials dried at ambient pressure for liquid-phase
epoxidation
J. L. Sotelo,* R. Van Grieken and C. Martos
Chemical Engineering Department, Faculty of Chemistry, Complutense University of Madrid, 28040 Madrid, Spain.
E-mail: rada@eucmax.sim.ucm.es
Received (in Cambridge, UK) 18th November 1998, Accepted 19th February 1999
Titanium-containing hydrophobic silica with a pore volume
and pore sizes in the range of aerogels was prepared by
silylation before drying of well-dispersed mixed oxides
obtained via a sol-gel method; it showed remarkable
properties as a catalyst in the epoxidation of styrene with
ButOOH.
oxides with adequate textural properties and highly dispersed Ti
atoms, both features leading to active catalysts in epoxidation
reactions with tert-butyl hydroperoxide (TBHP).
The catalysts were prepared following a two-step sol-gel
method. In the first step, a solution of tetraethoxysilane (TEOS)
was prehydrolyzed with aqueous HCl (0.05 m) for 45 min at
room temperature (samples 1–3). Other catalysts were prepared
by diluting TEOS in EtOH (samples 4–6); the solution was
prehydrolized with aqueous HCl for 90 min, since less water is
added to these samples. Then, the Ti precursor [titanium(iv)
butoxide, TNBT] diluted in PriOH was added to give gel molar
ratios (TEOS:H2O:HCl:EtOH:PriOH:TNBT) of 1:4:3.6 3
Environmental concern surrounding chemical processes for the
production of commodity and fine chemicals has led to the
development of environmentally-friendly processes based on
heterogeneous catalysts. Among them, the epoxidation of
olefins is a major target since the epoxide functional group is
one of the most useful intermediates in organic synthesis.1
Heterogeneous catalysts such as silica-supported titania are
used on an industrial scale for propylene epoxidation using
alkyl hydroperoxides as oxidant.2
1023 :0:1:0.023 for samples 1–3 and 1:1.25:9
3
1024 :2:1:0.023 for samples 4–6, with additional hydrolysis
times of 20 and 75 min, respectively. The second step of the
synthesis procedure increased the condensation rates by
changing the pH to higher values with aqueous NH3 (1 m) for
samples 1–3 and dilute NH4OH (1 m) in EtOH (1 NH3 :5.6
H2O:0.7 EtOH) for samples 4–6. Once the gel point of the sol
had been reached, the solid gels were aged for 48 h and samples
of 1 and 4 were dried at 110 °C overnight, whereas the other
catalysts were washed with n-hexane and treated with a TMSCl
solution in n-hexane before drying. For comparison, a silica
obtained by an analogous sol-gel procedure was impregnated
with a solution of TiCl4 in EtOH (Si/Ti = 54) and calcined at
550 °C for 5 h. Also, a Ti-containing HMS was provided by Dr
Tuel from CNRS13 with a Si/Ti molar ratio of 87.3. All the
catalyst samples were tested in the oxidation of styrene with
TBHP at 60 °C in N2 atmosphere. Conversions and selectivities
The discovery that the Ti-substituted silicalite molecular
sieve TS-1 (an isomorphous ZSM-5 zeolite) is a catalyst for
3
selective oxidation reactions with H2O2 has promoted in-
vestigation of its use in several oxidation reactions.4 To
overcome the restriction imposed by the relatively small
average diameter (ca. 0.55 nm) of the channel system, Ti has
been incorporated into materials with larger pore sizes such as
zeolite beta,5 or hexagonal mesoporous silica (HMS).6 Other
attempts to obtain suitable epoxidation catalysts have been
based on grafting of Ti compounds onto the silica surface7 and
the use of the sol-gel technique.8 The latter has been used to
obtain epoxidation catalysts ranging from amorphous micro-
porous9 to mesoporous titania–silica mixed oxides,8 and even
solid precursors with the appropriate properties for Ti-contain-
ing zeolite synthesis.10
1
were determined by GC, H NMR analysis and iodometric
titration.
The many potential advantages inherent to the use of sol-gel
techniques in preparing epoxidation catalysts are strongly
dependent on the drying method. Supercritical drying prevents
the network collapse induced by capillary forces arising from
the liquid–vapour interface, leading to materials with high
porosity called aerogels with suitable properties to be used as
catalysts,11 although supercritical processing has numerous
drawbacks due to the extreme conditions used. Recently a
simpler method at ambient pressure for the preparation of
aerogel-like films has been reported12 based on the use of a
silylating agent which is reacted with the hydroxy groups on the
surface of an inorganic gel, preventing the irreversible shrink-
age of the porous structure during drying.
All samples were characterised by diffuse reflectance UV–
VIS spectra (DR UV–VIS) in the range 190 to 500 nm. They
presented a maximum absorption band centered at 220 nm, and
no absorption around 330 nm was detected. Nitrogen adsorp-
tion–desorption isotherms of the samples obtained on an ASAP
2010 from Micromeritics showed a hysteresis loop correspond-
ing to a type IV isotherm, typically assigned to mesopores
present in the materials. Samples 1 and 4 present similar textural
properties independent of the presence of EtOH as a diluting
agent (Table 1), with BET surface areas above 600 m2 g21 pore
volumes around 1 cm3 g21 and average pore sizes in the range
of 50 Å as the result of the BJH analysis on the adsorption
branch of the isotherm. The silylation procedure leads to a
decrease in the final Ti content of the catalysts as a consequence
of the reaction of TMSCl with the OH groups at the surface of
We report here a simple route, based on the above mentioned
silylation procedure, for the preparation of titania–silica mixed
Table 1 Textural properties and chemical composition of the different tested samples
Diluted
in EtOH?
SBET
/
VP (ADS)/
cm3 g21
DP (ADS)/
Å
Sample
Silylated?
Calcined?
Si/Ti
m2 g21
1
2
3
4
5
6
NO
NO
NO
YES
YES
YES
NO
YES
NO
YES
YES
NO
43
73
69
43
73
69
682
871
849
604
723
688
1.0
1.8
1.3
1.0
3.0
2.3
50
71
68
57
132
125
YES
YES
NO
YES
YES
YES
Chem. Commun., 1999, 549–550
549

View Article Online
groups, with partial recovery of the surface OHs shown by the
increase of the shoulder at 2100 ppm. This behaviour is also
seen in the IR spectra of the samples, since in the silylated
catalysts there is a large decrease in the band corresponding to
Si–OH vibrations (3000–3750 cm21), which is partly recovered
after calcination.
The catalytic activities of these samples and the materials
used as reference were tested in the epoxidation of styrene with
TBHP. Besides the conversion of styrene and TBHP, Table 2
shows the turnover numbers (TON), relative to the total amount
of Ti in the catalyst, and the selectivity of styrene towards the
different products for the different catalysts, including a blank
test. Different products were obtained besides the primary
product, styrene oxide (SO): benzaldehyde (BADH) formed
from oxidative cleavage of styrene by TBHP and 2-phenyl-
2-tert-butoxyethanol (22ET) formed from nucleophilic attack
of ButOH on oxirane ring of the primary product.
The Ti-containing silica xerogels prepared without the use of
the silylation procedure (samples 1 and 4) show similar TON to
that of the TiO2 impregnated silica xerogel, although the
selectivity towards SO is higher for the latter. Silylated samples
show very high selectivity towards the epoxide with negligible
formation of the product coming from the opening of the
oxirane ring. The reuse of the silylated sample 5 shows a slight
decrease in styrene conversion (XST = 14.2%) without any
change in selectivity (SSO = 98.5%). Once the silylated samples
are calcined a decrease in SO selectivity is observed due to the
regeneration of HO groups on the sample surface, which
promotes the catalytic reaction of the epoxide with ButOH
formed from TBHP. The pore volume of the Ti-containing silica
is an important factor in the preparation of these catalysts, since
for comparable BET surface areas (samples 2 and 5, for
example), the higher the pore volume the thinner the pore walls;
therefore there is a higher probability of the Ti atoms being
located at the surface. Although at first sight the silylation
procedure could lead to the complete blocking of Ti atoms at the
catalyst surface, hindering their oxidation activity, not all this
surface is covered by the silylating agent, as was proved by the
catalytic activity of the silylated samples.
Fig. 1 Pore volume distribution (BJH analysis of the adsorption branch of
the isotherm) of (a) 4 (xerogel), (b) 5 (silylated) and (c) 6 (silylated and
calcined). Inset: 29Si MAS NMR spectra of these samples are shown.
the gel, which prevents irreversible shrinkage during drying.
There is a remarkable increase in the BET surface area, pore
volume and mean pore size in the samples prepared with EtOH.
It is interesting to note how the silylation procedure is capable
of preserving at least partially the different gel structures
derived from the use of increasing amounts of EtOH as solvent,
something that is not shown in the textural properties of the
xerogels when conventional drying is used (samples 1 and 4).
Similarly, after calcination, the differences are clearly shown
not only in comparison to the xerogels but also between the
materials prepared with different dilution levels, i.e. EtOH
content. The calcination of the silylated samples leads to a
decrease mainly in the pore volume of the samples compared
with the uncalcined ones, showing some evidence of pore
collapse. In any case, the pore volumes are clearly higher than
those corresponding to xerogels and are in accordance with the
textural properties of the starting non-calcined silylated sam-
ples. The used samples, after reaction, showed similar textural
properties to the fresh ones.
The method reported in this work is an easy and reproducible
way to obtain a highly hydrophobic Ti-containing material with
aerogel properties, oxidation activity comparable to Ti-HMS
and higher selectivity towards epoxide formation, without the
use of templates or surfactants as in the case of zeolites or
micelle structured zeolitic materials.
We are grateful to Dr Tuel (CNRS) for supplying Ti-HMS,
CEPSA for the X-ray fluorescence measurements. This work
was supported by the EU through the Brite-Euram Project (No.
Br.-1169).
Fig. 1 shows the pore size distribution coming from BJH
analysis of the N2 isotherms of samples 4–6. There is a shift in
the pore size distribution towards higher values when the
silylation procedure is used, with a decrease in the size of the
largest pores after calcination. The 29Si MAS NMR spectra of
4–6 are shown in Fig. 1 (inset), recorded in a Varian 300 MHz
with a spinning rate of 4 kHz. It is clearly shown how the
silylation procedure decreases the large Q3 contribution at
2100 ppm in the non-silylated sample, showing up as a peak 15
ppm upfield corresponding to trimethylsilyl groups (TMS)
formed in the reaction of TMSCl with the silanol groups of the
gel surface.14 The calcination procedure removes the TMS
Notes and references
1 K. A. Jorgensen, Chem. Rev., 1989, 89, 431.
2 R. C. Ragers, Br. Pat. 1 249 079, 1971.
Table 2 Activity of the different titanium-containing materials in the
epoxidation of styrenea
Conversion (%)
Selectivity (%)
3 M. Taramasso, G. Perego and B. Notari, US Pat. 4 410 501, 1983.
4 B. Notari, Catal. Today, 1993, 18, 163.
5 A. Camblor, A. Corma, A. Martinez and J. Pe´rez-Pariente, J. Chem.
Soc., Chem. Commun., 1992, 583.
6 T. Tanev, M. Chibwe and T. J. Pinnavaia, Nature, 1994, 368, 321.
7 M. Fraile, J. I. Garc´ıa, J. A. Mayoral, L. C. De Me`norval and F. Rachdi,
J. Chem. Soc., Chem. Commun., 1995, 539.
8 M. Schneider and A. Baiker, Catal. Today, 1997, 35, 339.
9 S. Klein, J. A. Martens, R. Parton, K. Vercruysse, P. A. Jacobs and W. F.
Maier, Catal. Lett., 1996, 38, 209.
10 M. A. Uguina, G. Ovejero, R. Van Grieken, D. P. Serrano and M.
Camacho, J. Chem. Soc., Chem. Commun., 1994, 27.
11 G. M. Pajonk, Catal. Today, 1997, 35, 319.
12 S. Sai, C. Prakash, J. Brinker, A. J. Hurd Sudeep and M. Rao, Nature,
1995, 374, 439.
13 S. Gonier and A. Tuel, Zeolites, 1995, 15, 601.
14 H. Yokogawa and M. Yokoyama, J. Non-Cryst. Solids, 1995, 186,
23.
Sample
Blank
TBHP
3.9
ST
0.6
TONb
BADH
SO
22ET
—
35
77
35
50
48
37
63
66
13.6
1.4
1.0
1.4
1.9
1.1
1.4
1.4
1.2
86.4
81.5
88.3
59.9
98.1
83.9
65.7
98.6
87.8
0.0
17.1
10.7
38.7
0.0
15.0
32.9
0.0
Impregnated 30.2
9.8
16.2
15.0
12.9
13.0
15.9
16.1
17.6
Ti-HMS
45.0
39.7
35.3
35.6
41.8
45.4
41.6
1
2
3
4
5
6
a
11.0
Reaction conditions: T = 60 °C, t = 3 h, autogenous pressure, no
solvent, stirring speed = 300 rpm, inert atmosphere, catalyst loading = 400
mg, styrene = 100 mmol, TBHP = 40 mmol; ST = styrene; TBHP = tert-
butyl hydroperoxide; SO = styrene oxide; BADH = benzaldehyde; 22ET
= 2-phenyl-2-tert-butoxyethanol. b In moles of styrene per mole of titanium
per hour.
Communication 8/09012B
550
Chem. Commun., 1999, 549–550