Trimethylsilylation of ordered and disordered titanosilicates: Improvements in epoxidation with aqueous H2O2 from micro- to meso-pores and beyond

Article abstract of DOI:10.1039/a809396b

A novel method for trimethylsylilation of micro- and mesoporous titanosilcates using BSTFA [N,O-bis(trimethylsilyl)trifluoroacetamide] renders Ti-MCM-41 and SiO₂/TiO₂ aerogels active for olefin epoxidation with aqueous H₂O₂, and even improves the activity of TS-1.

Full text of DOI:10.1039/a809396b

Trimethylsilylation of ordered and disordered titanosilicates: improvements in
epoxidation with aqueous H2O2 from micro- to meso-pores and beyond†
Michael B. D’Amore* and Stephan Schwarz*
DuPont Central Research and Development, Experimental Station, Wilmington, DE 19880-0262, USA.
E-mail: michael.b.damore@usa.dupont.com.stephan.schwarz@usa.dupont.com
Received (in Bloomington, IN, USA) 3rd August 1998, Revised manuscript received 19th November 1998,
Accepted 20th November 1998
A novel method for trimethylsylilation of micro- and meso-
porous titanosilcates using BSTFA [N,O-bis(trimethylsilyl-
)trifluoroacetamide] renders Ti-MCM-41 and SiO2/TiO2
aerogels active for olefin epoxidation with aqueous H2O2,
and even improves the activity of TS-1.
in Gelest catalogue = PSITI-019), [(EtO)2SiO][(EtO)2TiO],
Si : Ti ≈ 15. The general method for the preparation of aerogel
catalysts was that of Dutoit et al.,10 with acetylacetone-
modified titanium isopropoxide. The sol–gel was continuously
extracted with supercritical CO2 at 40 °C at 24.2 MPa for 5 h.
The resulting fluffy yellow powder was calcined at 400 °C for
1 h in N2 followed by 600 °C for 5 h in air.
The importance of hydrophobicity for titanosilicate catalyzed
oxidation reactions with aqueous H2O2 has been the subject of
much recent research. The hydrophobicity of the interior pores
of titanosilicalite (TS-1) and Al-free Ti-beta renders them active
for aqueous H2O2 oxidation of alkanes and alkenes.1, 2 By
comparison, Al containing Ti-beta is less efficient, mainly
because of framework Al induced hydrophilicity.3 Because of
their mesoporous nature (25–100 Å pores) Ti-MCM types4,5
would be useful as oxidation catalysts for larger, higher value
molecules. However, Corma and coworkers5 observed that
because of its hydrophilic nature, Ti-MCM-41 was much less
effective for oxidation with aqueous H2O2 than even Al-Ti-beta.
Therefore it is not surprising that efforts to render Ti-MCM’s
hydrophobic and hence, more active, were undertaken: Tatsumi
et al.6 trimethylsilylated Ti-MCM-41 and Ti-MCM-48 using
trimethylsilyl chloride and hexamethyldisiloxane and observed
increased oxidation activity.4 Amorphous silica/titanias with
even wider pores present opportunity for further increasing the
range of oxidation substrates. Recently non-catalytic and
hydrophilic amorphous silica/titanias have been made hydro-
phobic and catalytic by substituting alkoxy groups in the silica
precursors with alkyl or phenyl groups during the sol–gel
synthesis.7,8
A typical procedure for the trimethylsilylation of catalysts is
as follows: a mixture of N,O-bis(trimethylsilyl)trifluoroaceta-
mide [CF3COSiMe3)3]NNSiMe3 or BSTFA] (1.0 g) and toluene
(8 g) was added to the catalyst (0.5 g). The mixture was stirred
for 2 h at room temperature, filtered, and the solids washed with
toluene and air dried. The dried materials’ hydrophobicity was
evidenced by their floating on water. In addition, using a TGA
method derived from Anderson and Klinowski17 we obtained
the following hydrophobicities (mg water adsorbed silylated/
mg water adsorbed non-silylated at 50 °C) for TS-1, Ti-MCM,
and TiAgel respectively: 1.04, 0.41, 0.68. As expected, the ratio
for trimethylsilylated TS-1 was near unity because trimethyl-
silylation is not expected to affect the interior pore system.
However, for the mesoporous materials the ratio is less than
unity consistent with trimethylsilylation of the accessible
surface. Other trimethylsilylating agents such as trimethylsilyl
chloride and hexamethydisilazane were also used successfully.
However, these required more severe conditions and, in the case
of the chloride, resulted in loss of Ti from the aerogels. While
BSTFA has been used to derivatize organics and end group cap
chromatographic columns,11 there is no reference to using it to
modify catalysts. Characterization data for the catalysts are
given in Table 1.
Herein we report a simple and effective technique for
trimethylsilylation of microporous crystalline TS-1, mesopor-
ous Ti-MCM-41(Ti-MCM), and SiO2/TiO2 aerogels (Ti-Agel).
The resultant materials are hydrophobic and have enhanced
activity for olefin epoxidation with aqueous H2O2.
TS-1 was synthesized by the method of Thangaraj and
Sivasanker.9 Corma’s method5 for the preparation of Ti-MCM-
41 was modified by using a single precursor for titanium and
silicon: Gelest diethoxysiloxane–ethyltitanate copolymer (no.
BSTFA treatment of TS-1 introduces IR bands consistent
with trimethylsilylation. Because of the strong asymmetric
stretch of the silicate lattice at ca. 1225–1230 cm21, we are
unable to detect the diagnostic symmetric methyl deformation
in the 1250–1260 cm21 region.12 We do see the antisymmetric
methyl deformation at ca. 1400 cm21, as well as bands at ca.
725 and 700 cm21 that may be assigned to the antisymmetric
Si–C and the symmetric Si–C stretching modes, respectively.
Table 1 Characterization data for TS-1, Ti-MCM and Ti-Agela
CF3C(OSiMe3)NNSiMe3 + 2 [O3Si–OH]s ? CF3CONH2 + 2 [O3Si–O–SiMe3]s
N2 adsorption
SA PV
A1, A2
(UV–VIS/nm)
A1/A2
(UV–VIS/nm)
Sample name
XRDb
IRc/cm21
I960/I1090
Si/Ti ICP
APD
TS-1
MFI
MFI
MCM-41
—
Amorphous
—
962
972
960
950
959
955
0.10
0.05
0.37
0.23
0.34
0.32
215, 258d
207, 260
216, 268
214, 269
214, 255
214, 260
1.34
1.45
0.97
0.99
1.21
1.15
26
—
8e
—
12
—
485 0.11
453 0.14
810 0.12
52
74
27
24
TS-1-syl
Ti-MCM
Ti-MCM-syl
Ti-Agel
Ti-Agel-syl
a
727
493
318
0.14
1.50 148
1.33 166
Elemental analysis by inductively coupled plasma spectroscopy (ICP): MCM: Ti-MCM-41, Ti-Agel: SiO2/TiO2 Aerogel, syl: trimethylsilylated I960
:
assumed to be n(Si–O–Ti):10 I1090: n(Si–O–Si).10 A1: absorbance of band due to isolated Ti. A2: absorbance of band due to (Ti–O)n oligomers. SA: BET N2
surface area/m2 g21. PV: BJH N2 pore volume ml g21. APD: BJH N2 av. pore diameter/Å. No TiO2 phases observed by XRD in any of the samples.
b
c Position of 960 band, KBr pellet technique. d Si/Ti charged = 15. e Shoulder at ≈ 320 nm indicates anatase.
Chem. Commun., 1999, 121–122
121

View Online
Bands near 1680 and 1750 cm21 are assigned to the amide
group of trifluoroacetamide (the silylation byproduct) and
residual BSTFA, respectively. The IR spectrum of Ti-Agel is
very similar to that of Ti-MCM, which is to be expected since
the latter is composed of amorphous walls. For both Ti-MCM
and Ti-Agel, silylation added bands to the IR spectra that can be
attributed to SiMe3 groups, especially the very distinctive band
at 1250–1260 cm21, and the CH3 rocking modes at ca. 840 and
Table 2 Effect of trimethylsilylation on the epoxidation activity of Ti-Agel,
Ti-MCM, and TS-1 titanosilicatesa
Catalyst
%H2O2
Olefinb
Epoxide yieldc
Ti-Agel
Ti-Agel
Ti-Agel-syl
Ti-Agel-syl
Ti-MCM
Ti-MCM-syl
TS-1
10
10
10
10
3
1-Oct
COT
1-Oct
COT
1-Oct
1-Oct
1-Oct
COT
1-Oct
COT
0.2
0.5
6
32
755 cm21 12
.
0.5
23.6
2.4
0
12.9
0.5
The IR band at ca. 960 cm21 (the 960 band) has been
attributed to Si–O–Ti bonds as a fingerprint for tetrahedral Ti
substitution in silicate lattices,10 or to Si–OH groups.13, 14 We
hypothesize that for our TS-1 loss of silanol groups after
trimethylsilylation results in a decrease in the 960 band
intensity. The intensity of the 960 band of Ti-MCM decreased
on silylation, in agreement with Tatsumi et al.6 However, the
960 band of Ti-Agel decreases only slightly upon trimethylsily-
lation, consistent with a smaller fraction of silanols available to
the silylating agent compared to MCM. Based upon the
pronounced hydrophobic nature of the silylated material these
must be the ones which confer hydrophilicity to the base
aerogel.
3
10
10
10
10
TS-1
TS-1-syl
TS-1-syl
a
Reaction conditions: olefin (3 ml), aq. H2O2 (1 ml), catalyst (50–100
mg)d stirred at room temperature for 24 h. 1-Oct: oct-1-ene, COT: cis-
b
cyclooctene. c Yield of the corresponding epoxide was based on initial H2O2
concentration. The only significant byproduct with oct-1-ene was octan-1-al
(5–10%); with cyclooctene no byproducts were observed. d 100 mg with Ti-
Agel and TS-1, 50 mg with Ti-MCM.
epoxidizing molecules too large to fit in the pores of TS-1 has
been realized for the Ti-aerogel after the silylation treatment.
cis-Cyclooctene is epoxidized readily by Ti-Agel-syl compared
to almost no reaction with either TS-1 or TS-1-syl.
In summary we have demonstrated a general method for
significantly increasing the activity of titanosilicate oxidation
catalysts using aqueous hydrogen peroxide by rendering their
surfaces hydrophobic by treatment with the trimethylsilylating
agent BSTFA.
Analysis of the UV–VIS data15 for all of the base materials
indicates the presence of isolated Ti and oligomeric (Ti–O)n
species; in addition, TS-1 and Ti-MCM are contaminated by
anatase. The non-isolated Ti-species may have a detrimental
effect on the epoxidation reaction, since they have been reported
to decompose H2O2 homolytically.16 Ti-isolation, as measured
roughly by the ratio of the absorbance of band due to isolated Ti
to the band due to (Ti–O)n oligomers, (A1/A2), is better for Ti-
Agel compared with Ti-MCM. This may be ascribed to the acac
modification performed on the Ti-precursor for Ti-Agel, which
slows down the condensation rate of the hydrolyzed Ti
species.10 The effect of trimethylsilylation on the UV–VIS
spectra of the samples is not great, indicating that the silylation
procedure probably did not result in reaction with the isolated Ti
sites or H2O2 destroying sites.
Notes and references
† Contribution number 7801.
1 C. B. Khouw, C. B. Dartt, J. A. Labinger and M. E. Davis, J. Catal.,
1994, 149, 195.
N2 adsorption measurements on Ti-MCM confirmed that
surface area and pore size decrease on trimethylsilylation.6 Our
Ti-MCM contains microporosity, probably related to the titania
or silica/titania impurities resulting from the high Ti-concentra-
tion in the synthesis gel. For Ti-Agel, trimethylsilylation
decreases surface area and mesopore volume as expected.
Because of limitations in the N2 adsorption techniques, the
effect of trimethylsilylation on the microporosity of TS-1 could
not be determined.
2 T. Blasco, M. A. Camblor, A. Corma, P. Esteve, J. M. Guil, A. Martinez,
J. A. Perdigon-Melon and S. Valencia, J. Phys. Chem. B, 1998, 102,
75.
3 A. Corma, P. Esteve and A. Martinez, J. Catal., 1996, 161, 11.
4 J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge,
K. D. Schmitt, C. T.-W. Chu, D. H. Olson, E. W. Sheppard, S. B.
McCullen, J. B. Higgins and J. L. Schlenker, J. Am. Chem. Soc., 1992,
114, 10 834.
5 T. Blasco, A. Corma, M. T. Navarro and J. Perez Pariente, J. Catal.,
1995, 156, 65.
Trimethylsilylation of the available surfaces of TS-1, Ti-
MCM, and Ti-Agel increased the efficiency of olefin epoxida-
tion reactions with dilute aqueous hydrogen peroxide (Table 2).
Especially striking were the results with Ti-MCM and the
aerogels where epoxidation yields were dramatically increased
after the treatment. Rendering the catalyst surfaces hydrophobic
has either improved access of the olefin to the active sites or has
diminished the inhibitory effect of water.1 Either scenario
should result in increased catalytic activity.
In the case of the TS-1 the improved activity after
trimethylsilylation may be attributed to better wetting of the
microcrystallite surface with the olefin. While it is possible that
the increase in activity is related to loss of oligomeric (Ti-O)n
H2O2 decomposition sites after trimethylsilylation, lack of
changes in the UV–VIS and the probably very low concentra-
tion of reaction sites for silylation argue against this.
It is important to note that the trimethylsilylation technique
described in this communication results in nearly equivalent
epoxidation activity for any of the titanosilicates studied, the
only requirement being the presence of isolated, tetrahedral Ti
sites. Especially striking is the treatment of normally inert
aerogel; we believe this to be the first report of post synthesis
activation of this material. In addition the expectation of
6 T. Tatsumi, K. A. Koyano and N. Igarashi, Chem. Commun., 1998,
325.
7 H. Kochkar and F. Figueras, J. Catal., 1997, 171, 420.
8 S. Klein and W. F. Maier, Angew. Chem., Int. Ed. Engl., 1996, 35,
2230.
9 A. Thangaraj and S. Sivasanker, J. Chem. Soc., Chem. Commun.,1992,
123.
10 D. C. M. Dutoit, M. Schneider and A. Baiker, J. Catal., 1995, 153,
165.
11 K. D. McMurtrey, J. Liq. Chromatogr., 1988, 11, 3375; M. G. Horning,
E. A. Boucher, and A. M. Moss, J. Gas Chromatogr., 1967, 5, 297.
12 L. J. Bellamy, The infrared spectra of complex molecules vol. 2:
Advances in infrared group frequencies, Chapman and Hall, New York,
2nd edn., 1980.
13 F. Boccuzzi, S. Coluccia, G. Ghiotti, C. Morterra and A. Zecchina,
J. Phys. Chem., 1978, 82, 1298.
14 M. Decottignies, J. Phalippou and J. Zarzycki, J. Mater. Sci., 1978, 13,
2605.
15 T. Blasco, M. A. Camblor, A. Corma and J. Perez-Pariente, J. Am.
Chem. Soc., 1993, 115, 11 806.
16 M. G. Clerici and P. Ingallina, J. Catal., 1993, 140, 71.
17 M. W. Anderson and J. J. Klinowski, J. Chem. Soc., Faraday Trans. 1,
1986, 82, 1449.
Communication 8/09396B
122
Chem. Commun., 1999, 121–122