281
EtOMe
T3
Si(OEt)4
Me3OBF4
(b)
(a)
0
O-H st
Et O Si(OEt)3
T2
T1
-50
EtO Si(OEt)3
Me
4
Q4 silica
(EtO)3Si O Si(OEt)3
-100
-150
4000
3500
3000
2500
Et
5
Wavenumber/cm-1
Si(OEt)4
Chemical shift/ppm
Figure 3. 29Si MAS NMR (a) and FT-IR (ATR, b) spectra for
biphenylylene silica prepared by MR-catalyzed (red line) or conven-
tional sol-gel synthesis (black line).
Scheme 2. Plausible pathway for MR-catalyzed sol-gel polycon-
densation of TEOS.
Additional experiments were performed for a thorough
investigation by varying the mol % of water in anhydrous MeCN
(Entries 3-5).11 Entry 3 provided 98% silica with Q4 content up
to 94%. Polycondensation in the presence of trimethyloxonium
tetrafluoroborate (Me3OBF4) also resulted in highly-condensed
silica without altering the Q4 content (Entry 6). Et3OBF4 is labile
to hydrolysis with the formation of HBF4. Thus, in order to
ascertain any acid-catalyzed sol-gel polycondensation, we per-
formed the reaction in the presence of 60 mol % of HBF4 (45%
solution in water). The polycondensation proceeded incompletely
resulting in Q2 and Q3 together with Q4 silicon species (see: SI).
During the course of our studies on the MR-catalyzed sol-gel
process, Asefa et al. reported the use of Et3OBF4 for the
fluorination of mesoporous silicas.12 In accordance with his and
previous reports by Sakurai7a and Olah,7b,7c we propose that during
the initial stage of polycondensation, Et3OBF4 dissociates into Et+,
After achieving highly-condensed silica gels from TEOS
using MR, we extended our idea to generate organosilica gels.
BTEBPh as a model compound successfully provided highly
condensed amorphous biphenylylene silica in 94% yield in the
presence of Me3OBF4 (Figures 3 and SI20).
We thank Kyoto Advanced Nanotechnology Network for
generous support.
References and Notes
1
2
J. J. Ebelman, Ann. Chim. Phys. 1846, 16, 129.
M. A. Brook, Silicon in Organic, Organometallic, and Polymer
Chemistry, John Wiley & Sons, Inc., New York, 2000, Chap. 10.
G. W. Scherer, Sol-Gel Science: The Physics and Chemistry of Sol-Gel
Processing, Academic Press, New York, 1990. c) K. J. Shea, D. A.
Bourget, R. J. P. Corriu, D. Leclercq, P. H. Mutin, A. Vioux, J. Non-
3
¹ 13
¹
Et2O, and BF4
.
Formation of BF4 raised a question about the
ÔSi-F containing species in the obtained silica. Consequently,
19F MAS NMR was measured, which showed a sharp peak at
4
5
a) T. Shimada, K. Aoki, Y. Shinoda, T. Nakamura, N. Tokunaga, S.
M. P. Kapoor, S. Inagaki, S. Ikeda, K. Kakiuchi, M. Suda, T. Shimada,
2260. c) K. Su, D. R. Bujalski, K. Eguchi, G. V. Gordon, D.-L. Ou, P.
Thakar, T. J. Schildhauer, W. Buijs, F. Kapteijn, J. A. Moulijn,
¹
¹150 ppm (assigned for BF4 ) and a broad peak ranging from
¹140 to ¹150 ppm (assigned for ÔSi-F). These results suggest
that nucleophilic fluorination14 to the Si center occurred to some
extent in the course of the condensation reaction with TEOS.
Furthermore, ion chromatography analysis data for the silica
sample showed a residual fluoride content of 1.1 mass % (SI).
In order to elucidate the plausible reaction mechanism, we
next performed additional reaction of TEOS (1.0 equiv) with
Me3OBF4 (0.3 equiv) using an NMR tube with J Young valve (see
6
7
W. K. Pang, I. M. Low, J. V. Hanna, J. Aust. Ceram. Soc. 2009, 45, 39.
For mono-, di-, and trisilyloxonium intermediates, see: a) M. Kira, T.
1
SI). The formation of EtOMe observed by H NMR indicates the
presence of silyloxonium intermediate 4.7a,15,16 Moreover, we
observed two new quartets at 4.0 and 4.54 ppm corresponding to
-CH2 protons adjacent to -CH3 groups at 1.27 and 1.45 ppm,
respectively. As there is no previous report and no NMR evidence
regarding such intermediates, we speculate that these two quartets
correspond to silyloxonium intermediates. On the basis of these
findings, we propose a plausible reaction pathway in Scheme 2.
As we discussed, Me3OBF4 dissociates into Me+ which
presumably alkylate the -OEt group of TEOS to form silyloxo-
nium 4. According to reports by Asefa12 and Maier,17 nucleophilic
8
9
H. Meerwein, Org. Synth. 1966, 46, 113.
10 Powder XRD reveals amorphous architecture without any periodicity.
¹1
The sample has a BET surface area of 109.3m2 g (SI).
11 For experimental procedure, 29Si and 19F MAS NMR, see: SI.
12 C. T. Duncan, A. V. Biradar, S. Rangan, R. E. Mishler, II, T. Asefa,
¹
attack by F at the Si center likely results in a hypervalent silicon
intermediate that stretches and weakens the Si-OEt bonds. At a
later stage, silyloxonium 5 will also likely be generated with an
additional mol of TEOS. Independent reports by Sakurai7a and
Olah7b,7c on the detection of silyloxoniums 1-3 (Figure 1) support
our claim for proposed silyloxoniums 4 and 5. In addition to the
above reports, Charpentier also recently observed the presence of a
silyloxonium in the polycondensation of TMOS.18 Furthermore,
our protocol need 10-18 mol % of water in dry MeCN19 however,
role of water is unclear at this stage. We speculate that the use of
water may possibly help in either solubilizing the oxonium salts or
in shifting the equilibrium from silyloxonium intermediate 5 to
carry forward further condensation of TEOS.
16 Diethylsilyloxonium has been detected by 1H NMR in methylene
chloride-d2. For details, see: ref. 7a.
17 For hypervalent silicon species, see: I. C. Tilgner, P. Fischer, F. M.
19 For the reactions of MeCN with Me+, see: J. E. Gordon, G. C. Turrell,
20 Supporting Information is available electronically on the CSJ-Journal
Chem. Lett. 2012, 41, 280-281
© 2012 The Chemical Society of Japan