Neutral alkoxysilanes from silica

Article abstract of DOI:10.1021/ja001885h

Silica (SiO₂) is found to react readily with ethylene glycol (EGH₂) to form neutral glycoxysilanes in the presence of catalytic amounts of high-boiling organic amines, such as triethylenetetramine (TETA), trishydroxymethyleneaminomethane [H₂NC(CH₂OH)₃, THAMH3], and triethanolamine [N(CH₂CH₂OH)₃, TEAH₃]. Kinetic studies show that these amines offer similar catalytic efficiencies although their pK_b values differ by 3 orders of magnitude. In addition, silica dissolution is found to be pseudo-zero order in silica. These kinetic data can be explained by a rate-limiting step involving release of free base from an intermediate pentacoordinated silicate coincident with the formation of a tetraalkoxysilane. The products from these reactions were characterized by ¹H, ¹³C, and ²⁹Si solution and solid-state NMR, thermal gravimetric analysis, and mass spectroscopy. Depending on the type and amount of base used, different products form: either neutral tetraalkoxysilanes, such as Si(OCH₂CH₂OH)₄ and its soluble oligomers, or neutral pentacoordinate silanes, such as N(CH₂CH₂O)₃SiOCH₂CH₂OH and H₃N⁺C(CH₂O)₃Si^-(OCH ₂CH₂O). Comparative studies demonstrate that Group I metal hydroxides also catalyze silica dissolution in ethylene glycol with better catalytic efficiencies than the amine bases. The products of silica dissolution using Group I metal hydroxide catalysts were also identified by ²⁹Si solution NMR and mass spectroscopy and found to consist primarily of Si(OCH₂CH₂OH)₄ and its oligomeric derivatives.

Full text of DOI:10.1021/ja001885h

J. Am. Chem. Soc. 2000, 122, 10063-10072
10063
Neutral Alkoxysilanes from Silica
Hengqin Cheng,† Ryo Tamaki,†,‡ Richard M. Laine,*,† Florence Babonneau,§
Yoshiki Chujo,‡ and David R. Treadwell†
Contribution from the Departments of Chemistry and Materials Science and Engineering, Macromolecular
Science and Engineering Center, The UniVersity of Michigan, Ann Arbor, Michigan 48109-2136,
Department of Polymer Chemistry, Graduate School of Engineering, Kyoto UniVersity, Yoshida, Sakyo-ku,
Kyoto 606-8501, Japan, and Laboratoire Chimie de la Matiere Condense´e, UniVersite´ de Pierre et Marie
Curie, Tour 54 E5, 4 Place Jussieu, 75252 Paris Cedex 05, France
ReceiVed May 30, 2000
Abstract: Silica (SiO2) is found to react readily with ethylene glycol (EGH2) to form neutral glycoxysilanes
in the presence of catalytic amounts of high-boiling organic amines, such as triethylenetetramine (TETA),
trishydroxymethyleneaminomethane [H2NC(CH2OH)3, THAMH3], and triethanolamine [N(CH2CH2OH)3,
TEAH3]. Kinetic studies show that these amines offer similar catalytic efficiencies although their pKb values
differ by 3 orders of magnitude. In addition, silica dissolution is found to be pseudo-zero order in silica. These
kinetic data can be explained by a rate-limiting step involving release of free base from an intermediate
pentacoordinated silicate coincident with the formation of a tetraalkoxysilane. The products from these reactions
were characterized by 1H, 13C, and 29Si solution and solid-state NMR, thermal gravimetric analysis, and mass
spectroscopy. Depending on the type and amount of base used, different products form: either neutral
tetraalkoxysilanes, such as Si(OCH2CH2OH)4 and its soluble oligomers, or neutral pentacoordinate silanes,
such as N(CH2CH2O)3SiOCH2CH2OH and H3N+C(CH2O)3Si-(OCH2CH2O). Comparative studies demonstrate
that Group I metal hydroxides also catalyze silica dissolution in ethylene glycol with better catalytic efficiencies
than the amine bases. The products of silica dissolution using Group I metal hydroxide catalysts were also
identified by 29Si solution NMR and mass spectroscopy and found to consist primarily of Si(OCH2CH2OH)4
and its oligomeric derivatives.
Introduction
to novel materials that offer ion conducting,11 liquid crystalline,12
charge transfer,13 or ceramic precursor properties.14 Compound
1 has also been recently reported to play a key role in the
nonaqueous synthesis of silica sodalite.15
A primary challenge in silicon chemistry is to use silica (SiO2)
as a direct source of chemicals and polymers. To this end,
several groups have explored dissolution of silica in basic,
aqueous, or nonaqueous environments to produce anionic alkoxy
and aryloxy silicates.1-7 Most recently Kinrade and co-workers
demonstrated that aliphatic polyols will react with aqueous
silicates to form high concentrations of stable polyolate com-
plexes containing penta- and hexacoordinated alkoxysilanes.8
In earlier studies, we explored silica dissolution in excess
ethylene glycol (EGH2) and reported the syntheses of penta-
and hexacoordinate anionic glycolato silicate complexes from
reactions of silica gel, fused silica, or sand with equivalent
amounts of Group I metal hydroxides (reaction 1) or Group II
metal oxides (reaction 2).9,10 Compounds 1 and 2 provide access
† The University of Michigan.
‡ Kyoto University.
Kinetic studies on the mechanism of silica dissolution in
ethylene glycol promoted by equivalent amounts of alkali
hydroxides indicate that silica dissolution is linearly dependent
on both [MOH] and silica surface area. We proposed a reaction
§ Universite´ de Pierre et Marie Curie.
(1) (a) Boudin, A.; Cerveau, G.; Chuit, C.; Corriu, R. J. P.; Reye, C.
Angew. Chem. 1986, 98, 473. (b) Boudin, A.; Cerveau, G.; Chuit, C.; Corriu,
R. J. P.; Reye, C. Angew. Chem., Int. Ed. 1986, 25, 473.
(2) Corriu, R. J. P.; Perz, R.; Reye, C. Organometallics 1988, 7, 1165.
(3) Frye, C. L. J. Am. Chem. Soc. 1964, 86, 3170.
(4) Rosenheim, A.; Raibmann, B.; Schendel, G. Z. Anorg. Allg. Chem.
1931, 196, 160.
(5) (a) Barnum, D. W. Inorg. Chem. 1970, 9, 1942. (b) Barnum, D. W.
Inorg. Chem. 1972, 11, 1424.
(6) Weiss, A.; Reiff, G.; Weiss, A. Z. Anorg. Allg. Chem. 1961, 311,
142, 151.
(9) (a) Laine, R. M.; Blohowiak, K. Y.; Robinson, T. R.; Hoppe, M. L.;
Nardi, P.; Kampf, J.; Uhm, J. Nature 1991, 353, 642. (b) Laine, R. M.;
Mueller, B. L.; Hinklin, T. U.S. Patent No. 5,418,298, May 23, 1995. (c)
Laine, R. M. U.S. Patent No. 5,440,011, Aug 8, 1995. (d) Laine, R. M.;
Waldner, K.; Bickmore, C.; Treadwell, D. R. U.S. Patent 5,614,596, March,
1997.
(10) (a) Blohowiak, K. Y.; Treadwell, D. R.; Mueller, B. L.; Hoppe, M.
L.; Jouppi, S.; Kansal, P.; Chew, K. W.; Scotto, C. L.; Babonneau, F.;
Kampf, J.; Laine, R. M. Chem. Mater. 1994, 6, 2177. (b) Hoppe, M. L.;
Laine, R. M.; Kampf, J.; Gordon, M. S.; Burggraf, L. W. Angew. Chem.
1993, 105, 283; Angew. Chem., Int. Ed. 1993, 32, 287.
(7) (a) Boer, F. P.; Flynn, J. J.; Turley, J. W. J. Am. Chem. Soc. 1968,
90, 6973. (b) Glynn, J. J.; Boer, F. P. J. Am. Chem. Soc. 1969, 91, 5756.
(8) Kinrade, S. D.; Del Nin, J. W.; Schach, A. S.; Sloan, T. A.; Wilson,
K. L.; Knight, C. T. G. Science 1999, 285, 1542.
10.1021/ja001885h CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/03/2000

10064 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Cheng et al.
Scheme 1. Proposed Reaction Mechanism for Silica
Dissolution with 1 Equiv of KOH
Scheme 2. Formation of Spiro-Tetralkoxysilicate
Ammonium Salts
°C (120 °C) to give 50-80% conversion to the corresponding
alkoxysilanes.17 Suzuki et al. find that gaseous dimethyl and
diethyl carbonate will react with alkali treated silica at 225 to
325 °C to form the respective alkoxysilanes quantitatively.18
We describe here a new, simple, and low-cost synthetic route
to neutral alkoxy and polyalkoxysilanes directly from silica. This
work was prompted by Frye’s observation that spirosilicates
will react with methanol and amines (e.g. Et3N) to form
pentacoordinate, anionic silicates that revert to tetracoordinate
spirosilicates on heating.19 These results suggested that catalytic
quantities of amine bases might promote silica dissolution to
generate neutral alkoxysilanes.
We have now explored this approach and find that silica will
dissolve in ethylene glycol in the presence of catalytic amounts
of amine bases. Dissolution can also be catalyzed effectively
by Group I metal hydroxides under conditions much simpler
than used by Bailey et al.17 In the following, we present detailed
studies on this dissolution, including kinetic data which il-
luminate specific aspects of the dissolution mechanism, and
identify the reaction products.
Experimental Section
mechanism (Scheme 1) driven by constant removal of water,
in which ethylene glycol reacts with alkali base to form alkoxide
that attacks the silica surface to form pendant alkoxysilanes.
Then, through a series of “unzipping” steps, a neutral spirosi-
loxane is released generating SiO- groups on the silica surface,
which are subsequently protonated to give Si-OH groups. For
the reaction to continue, the surface Si-OH groups most likely
condense with ethylene glycol producing H2O as byproduct,
which we proposed to be the rate-limiting step. Compound 1
forms when the neutral spirosiloxane reacts with free alkoxide.
At this point, following a thorough review of the literature,
it occurred to us that it might be possible to catalyze silica
dissolution to produce neutral alkoxysilanes. In contrast to the
numerous reports on the syntheses of organic anionic silicate
species from silica,1-10 much less work has been done on the
syntheses of neutral alkoxysilanes directly from mineral sources.
Goodwin and Kinney have described acidification of mineral
silicates to form silicic acid which is then esterified with ethanol
or propanol to give good yields of alkoxysilanes.16 A patent by
Bailey describes heating silica in mixed solvents of alcohols
(ethanol or propanol) and aromatic hydrocarbons (benzene or
xylene) with 0.1 equiv of KOH (CsOH) for 16 h (120 h) at 200
General Procedures. All chemicals were reagent grade, purchased
from standard vendors, and used as received except ethylene glycol
(EGH2). EGH2 used in these experiments was (1) purchased from
standard vendors and used as received, (2) recycled from reaction
distillate by double distillation under N2, or (3) recovered from used
antifreeze by double distillation under N2. All reactions were carried
out under N2 to minimize the chances of exposing the solution to air
or moisture.
Analytical Methods. (a) 29Si Nuclear magnetic resonance (NMR)
studies were carried out on a Bruker 360 MHz spectrometer using 11°
pulse widths and a scanning width of 20000 Hz. The FID signals were
treated with a LB value of 5. Samples (200-500 mg estimated by TGA
data) in 10-15 mL of EGH2 were placed in 10-mm NMR tubes with
an inner 5-mm NMR tube containing acetone-d6 as lock solvent and
tetramethylsilane (TMS) as an external given reference. Samples were
scanned 1000-4000 times.
(b) Solid-state 29Si MAS NMR spectra were recorded (Laboratoire
de la Matie`re Condense´e) on a Bruker MSL 400 at 79.5 MHz, using
pulse widths of 2 µs (30°), delays between pulses of 30 s, and a spinning
rate of 5 kHz.
1
(c) Solid-state 13C and H MAS NMR spectra were recorded on
a Bruker MSL 300 (75.5 and 300.1 MHz). 13C NMR spectra were
recorded with cross polarization (CP) techniques with contact times of
(11) Chew, K. W.; Dunn, B.; Faltens, T. A.; Hoppe, M. L.; Laine, R.
M.; Nazar, L.; Wu, H. K. J. Am. Chem. Soc. Polym. Prprts. 1993, 34, 254.
(12) Ray, D. J.; Laine, R. M.; Robinson, T. R.; Viney, C. Mol. Cryst.
Liq. Cryst. Sci. Technol., Sect. A 1992, 225, 153.
(13) Youngdahl, K. A.; Nardi, P.; Robinson, T. R.; Laine, R. M. NATO
Asi Ser. E: Appl. Sci. 1991, 206, 99.
(14) (a) Kansal, P.; Laine, R. M. J. Am. Ceram. Soc. 1994, 77, 875. (b)
Kansal, P.; Laine, R. M. J. Am. Chem. Soc. 1995, 78, 529.
(15) (a) Bibby, D. M.; Dale, M. P. Nature, 1985, 317, 157. (b) van Erp,
W. A.; Kouwenhoven, H. W.; Nanne, J. M. Zeolites 1987, 7, 268. (c) Herron,
B.; Carr, S. W.; Klinowski, J. Science 1994, 263, 1585.
(16) (a) Kenny, M. E.; Goodwin, G. B. U. S. Patent, No. 4,717,773,
Jan, 1988. (b) Kenny, M. E.; Goodwin, G. B. In Inorganic and Organo-
metallic Polymers; ACS. Symp. Ser. Vol. 360; American Chemical
Society: Washington, DC, 1988; p 238. (c) Goodwin, G. B.; Kenney, M.
E. AdV. Chem. Ser. 1990, 224, 251.
(17) Bailey, D. L.; Snyder, A.; O’Connor, F. M. U.S. Patent No. 2,-
881,198, April 7, 1959.
(18) Suzuki, E.; Akiyama, M.; Ono, Y. J. Chem. Soc., Chem. Commun.
1992, 136.
(19) Fry, C. L. J. Am. Chem. Soc. 1979, 92, 1205. Frye, C. L. J. Am.
Chem. Soc. 1971, 93, 6805.

Neutral Alkoxysilanes from Silica
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10065
2 ms, 10 s delays between pulses, and a spinning rate of 5 kHz. Sample
sizes for all solid-state spectra were 300-500 mg.
ppm and one at -82 ppm. The height of each peak was measured and
the ratio of heights (-106 to -82 ppm) was calculated and plotted in
Figure 4.
(d) Thermal gravimetric analysis (TGA) was conducted using a
TA Instruments 2950 Thermal Analysis Instrument. Samples (30-60
mg) were loaded in platinum pans and heated under N2 balance (40
cm3/min) with an air purge (60 cm3/min), using a high-resolution (Hi-
Res 4) program with a maximum ramp rate of 50 °C/min to
temperatures of 950 °C.
(e) Mass spectroscopic (MS) analyses were carried out using a VG
70-70-E mass spectrometer (Micromass Corporation) by fast atom
bombardment ionization in the positive ion mode (FAB+). The
instrument was calibrated using CsI salt clusters. The FAB atom gun
was run with xenon gas, at settings of 1 mA and 8 KV voltage. The
mass range was scanned from m/z 2800 to 75 at 8 s/decade and run at
5 kV accelerating voltage. The samples in ethylene glycol were placed
directly on the sample target, and three to five scans were signal
averaged while being collected in a continuum or MCA mode.
Selected samples were also analyzed by chemical ionization (CI)
with ammonia on a VG 70-250-S mass spectrometer (Micromass
Corporation). The instrument was calibrated with perfluorokerosene
and scanned from m/z 1000 to 35 at 10 s/decade and 6 kV accelerating
voltage.
Standard Conditions for Kinetic Studies. To a 250-mL round-
bottom flask containing 100.0 mL of ethylene glycol (EGH2) were
added a certain amount of silica and various amounts of amine or alkali
bases. The flask, equipped with magnetic stirring and standard
distillation setup, was immersed in an oil bath preheated to 200 ( 2
°C for a pre-set reaction time. The flask was then cooled quickly in an
ice water bath. The undissolved silica was separated from the solution
by centrifuging for 15 min. This undissolved silica was then washed
with about 100 mL of distilled ethanol, followed by separation again
using centrifugation. The recovered silica was then heated in a box
furnace at 1 °C /min to 60 °C and held for 2 h, then heated at 1 °C/
min to 80 °C and held for 2 h, and then finally heated at 5 °C /min to
500 °C and held for 6 h. The heat treatment was used to oxidize any
organic residues. The weight of unreacted silica could be weighed then
and the amount of silica dissolved determined. All experiments were
repeated a minimum of twice. Blanks were also run for each set of
kinetic studies.
Comparison of Catalytic Efficiencies of the Organic Amine Bases.
Silica (1.50 g, 25.0 mmol) and 10 mol % of each above amine base
were added into 100.0 mL of EGH2 and standard reaction conditions
described above were applied. The reaction time was set to 3 h. A
blank test, in which only silica and EGH2 were mixed and no amine
base was added, was also heated for 3 h. Dissolution yields were then
determined by the method described above. The results are shown in
Figure 1.
Effect of Reaction Times for Amine-Catalyzed Silica Dissolution.
Silica (1.50 g, 25.0 mmol) and THAMH3 (0.300 g, 2.48 mmol) were
added to 100.0 mL of EGH2 and standard reaction conditions were
applied. A series of reaction times were used: 1.5, 3, 6, 9, and 12 h.
A blank in which silica and THAMH3 were mixed in EGH2 with no
heating was also examined. The results are shown in Figure 2. A similar
reaction was run using 2.48 mmol THAMSiegH. The results are also
plotted in Figure 2.
Effects of Initial Amine Concentration on Silica Dissolution. SiO2
(1.50 g, 25.0 mmol) and 5-150 mol % THAMH3 were added to 100.0
mL of EGH2 and standard reaction conditions were applied. Reaction
time was set to 3 h. A blank was run in which the mixture of silica
with ethylene glycol without any amine was heated for 3 h. The results
are shown in Figure 3.
29Si NMR Study of Equilibria between Tetra- and Pentacoor-
dinate Silanes. Tetraethoxysilane (TEOS; 2.07 g, 10.0 mmol) was
mixed with 36.0 mL of EGH2 and heated at 155 °C in a 100-mL round-
bottom flask equipped with magnetic stirring and a standard Pyrex
distillation setup under N2 to distill off ethanol. To the resulting solution
was added 1.00 equiv of base (amine or alkali hydroxide). The mixture
was placed in an oil bath preheated to 200 °C for pre-set times (10
min, 1, 3, or 12 h). At the end of each reaction time, the reaction flask
was immediately cooled in an ice bath and a 29Si NMR spectrum was
recorded. Two peaks were observed for each spectrum: one at -106
Effects of Reaction Temperature on Silica Dissolution. SiO2 (1.50
g, 25.0 mmol) and THAMH3 (0.75 g, 6.25 mmol) were added to 100.0
mL of EGH2 in a 250-mL round-bottom flask equipped with magnetic
stirring and standard distillation setup. The mixture was immersed into
an oil bath preheated to the desired temperature (170, 180, 190, and
200 °C) using an Omega CN76000 temperature controller. The mixture
was heated at that temperature for 3 h before it was immediately cooled
by an ice bath. The silica dissolution yield was then determined by the
method described in the section of Standard Reaction Conditions. The
results are shown in Figure 5.
Effects of Initial Concentration of Alkali Base on Silica Dissolu-
tion. Silica was mixed with various amounts of KOH (1-10 mol %)
in 100.0 mL of ethylene glycol. Standard conditions were applied to
this mixture and the reaction time was set to 1 h. A blank in which
silica was mixed with EGH2 without any base was also performed.
The results are shown in Figures 6 and 7.
Effects of Reaction Temperature on Silica Dissolution with Alkali
Base Catalysis. Silica mixed with 1 mol % of KOH in EGH2 was
heated for 1 h using standard reaction conditions with the reaction
temperature set at 170, 180, 190, and 200 °C. The dissolution yields
were determined as described above. The results are shown in Figure
8.
Mass Spectral Studies for Silica Dissolution Catalyzed by Alkali
Base. Silica (30.0 g, 0.500 mol) and different amounts of NaOH (from
0.05 to 0.1 equiv) were mixed with 600 mL of EGH2 in a 1-L round-
bottom flask equipped with a magnetic stirrer and distillation head.
This mixture was heated to >200 °C to distill off EGH2 and byproduct
water. As distillation continues, additional doubly distilled EGH2 was
added to retain the total reaction volume. Distillation was continued
for 30 h to allow a full study of the reactions. A series of samples
were taken at different reaction times. These samples were analyzed
by mass and 29Si NMR spectroscopy. The results are listed in Table 3.
Synthesis of Si(eg)x. Mixtures of silica (10.0 g, 166 mmol), 5-10
mol % of amine or alkali base, and EGH2 (500 mL) were heated under
N2 with magnetic stirring in a standard Pyrex distillation setup such
that EGH2 and byproduct H2O distill off slowly. After all the silica
dissolved, the solution volume was reduced to ∼50 mL and a clear
liquid was obtained. Further removal of the remaining EGH2 (10-2 Torr/
220 °C) gave a cross-linked polymer. This material was characterized
by TGA and solid-state MAS 29Si NMR. Characterization data are listed
in Table 1.
Synthesis of TEASiegH. Fumed silica (240.0 g, 4.00 mol) and
triethanolamine (TEAH3) (565.0 mL, 1.05 equiv) were mixed with ≈3
L of EGH2 in a 5-L round-bottom flask equipped with a mechanical
stirrer and still head. All silica dissolved in 10 h. The solution was
kept heating at >200 °C under normal pressure to distill off EGH2
until the solution was very viscous. Then the solution was heated at
150 °C under vacuum to obtain additional drying. The remaining residue
was washed with freshly distilled ethanol, and after filtration, 850 g of
white solid was collected (90% yield). Characterization data are listed
in Table 2.
Synthesis of THAMSieg. Fumed silica (30.0 g, 0.500 mol) was
mixed with 500 mL of EGH2 in a 1-L round-bottom flask equipped
with magnetic stirring and standard still head. Trishydroxymethylene-
aminomethane (60.5 g, 0.550 mol) was then added. This mixture was
heated to >200 °C to distill off EGH2 and water produced during the
reaction. After all the silica dissolved, the solution volume was reduced
initially by simple distillation, followed by distillation under reduced
pressure. The resulting solid was washed with dry THF. The resulting
pure white solid is isolated in quantitative yield. Characterization data
are listed in Table 3.
Results and Discussion
As mentioned above, prompted by Frye’s observations,19 we
decided to explore the possibility of using organic amines to
catalyze silica dissolution in ethylene glycol. For silica dissolu-
tion to proceed, the reaction must be run under conditions where

10066 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Cheng et al.
As Figure 1 shows, all three amines are modest catalysts for
silica dissolution with similar catalytic efficiencies as dissolution
yields all fall in the 16-20 mol % range (corrected for blank
as are all data presented). Although the yields are not high for
3 h reaction times, they clearly show that amines will catalyze
dissolution as more than stoichiometric amounts of silica
dissolve.
Ethylene glycol with a pKa of ∼15 is quite similar to
water.22-24 Thus, as Frye noted, simple amines are strong enough
to deprotonate simple alcohols to form alkoxides.19 Thus,
ethylene glycol should be easily deprotonated by the chosen
amine bases to form ammonium alkoxides (TETA pKb1
)
3.24,25 THAMH3 pKb ) 5.97,26 and TEAH3 pKb ) 6.3527),
which in turn can be expected to attack the silica surface,
initiating a series of steps leading to silica dissolution.
In principle, the differences in catalytic efficiencies among
these three amines should be related to their basicities. The
literature pKb values indicate 3 orders of magnitude difference
in basicities for these amines. These differences should affect
their abilities to deprotonate EGH2 to form the alkoxide, which
is the reactive intermediate for silica dissolution.10 However,
the dissolution yields do not reflect these differences. All three
amines have similar “apparent” basicities toward silica dissolu-
tion in EGH2.
A plausible explanation (see below) for this behavior was
sought through further kinetic studies, which were conducted
also with the goal of improving dissolution rates. Below we
first present the results from kinetic studies examining the effects
of reaction time, temperature, and amine concentrations on the
dissolution rates, followed by a discussion of possible dissolution
mechanisms.
Figure 1. Silica dissolution yields vs type of amine base.
byproduct water is removed continuously. This also obviates
the potential for forming zeolites.15,20,21 Thus, organic amines
used in the reaction must have high boiling points to minimized
co-distillation. With this in mind, three high-boiling amines were
chosen: triethanolamine (TEAH3), trishydroxymethyleneami-
nomethane (THAMH3), and triethylenetetramine (TETA).
Reaction Time. A second set of kinetic studies was carried
out to examine the effects of reaction time on silica dissolution.
THAMH3 was used for this set of studies. Data are plotted in
Figure 2. A blank test consisted of stirring a mixture of silica,
EGH2, and THAMH3 without heating for 3 h which results in
13 mol % silica dissolution. Figure 2 clearly shows the catalytic
effects of THAMH3 in silica dissolution as over 80 mol % of
the silica dissolves in 12 h with the aid of 5 mol % base.
An important additional observation seen in Figure 2 is that
dissolution is linear over the 12 h reaction period, implying a
constant dissolution rate. This suggests that the rate determining
step does not involve silica since the total silica surface area
should change significantly with reaction times. Furthermore,
reaction rates should diminish as observed in our previous
dissolution studies with equivalent amounts of alkali bases.10
The data from Figure 2 were quite confusing initially. As
discussed above, the first step in silica dissolution is most likely
deprotonation of EGH2 by amine to form an alkoxide as an
active intermediate. One might interpret the data from Figure 2
to suggest that the slow step is formation of alkoxide. However,
if this were true, we would have observed a much higher silica
dissolution rate for TETA with a pKb of 3 vs the other two
amines with pKb values of ∼6, which contradicts the Figure 1
data showing that all three amines give similar dissolution yields
despite the differences in their basicities.
Initial dissolution studies showed that silica would dissolve
in ethylene glycol when any one of the above amines was
present in catalytic amounts. This result was encouraging and
thorough kinetic studies were then carried out.
Standard Conditions for Kinetic Studies. The first step was
to establish standard conditions modeled on our earlier studies.10
A flask containing a mixture of silica and ethylene glycol with
various millimolar amounts of amines or alkali base was
attached to a distillation setup above an oil bath preheated to
200 ( 2 °C. The bath was raised to immerse the flask and the
reaction was run for a pre-set reaction time. The flask was then
cooled quickly in an ice water bath. The silica dissolution yields
were then determined as described in the Experimental Section.
Initial dissolution rates were then calculated. Blanks were also
run for each set of kinetic studies.
Catalytic Efficiencies of Organic Bases. The first set of
kinetic studies established the catalytic efficiencies of the organic
amines chosen. Using the standard experimental procedure, 10
mol % of TEAH3, THAMH3 and 2.4 mol % of TETA (10 mol
% nitrogen content) were tested with a reaction time of 3 h.
The dissolution yields are plotted in Figure 1. A blank, run
without amine, gave 20 mol % silica dissolution.
(22) Arnet, E. M.; Small, L. E. J. Am. Chem. Soc. 1977, 99, 808.
(23) Ballinger, P.; Long, F. A. J. Am. Chem. Soc. 1959, 81, 1050.
(24) Takahashi, S.; Cohen, L. A.; Miller, H. K.; Peake, E. G. J. Org.
Chem. 1971, 36, 1205.
(25) (a) Jonassen, H. B.; LeBlanc, R. B.; Meibohm, A. W.; Rogan, R.
M. J. Am. Chem. Soc. 1950, 72, 2430. (b) von Schwarzenbach, G. HelV.
Chim. Acta 1950, 33, 975.
(26) Glasstone, S.; Schram, A. F. J. Am. Chem. Soc. 1947, 69, 1214.
(27) (a) Alner, D. J.; Smeeth, A. G. J. Chem. Soc. 1958, 4207. (b) Bates,
R. G.; Schwarzenbach, G. HelV. Chim. Acta 1954, 37, 1437.
(20) Szostak, R.; Molecular SieVes. Principles of Synthesis and Identi-
ficaion; Van Nostrand Reinhold: New York, 1989; Catalysis Series.
(21) (a) Qisheng, H.; Shouhua, F.; Ruren, X. J. Chem. Soc., Chem.
Commun. 1988, 1486. (b) Qisheng, H.; Ruren, X. J. Chem. Soc., Chem.
Commun. 1992, 168.

Neutral Alkoxysilanes from Silica
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10067
Figure 2. Silica dissolution as a function of reaction time with 5 mol
% THAMH3 or THAMSiegH.
Figure 3. Silica dissolution as a function of initial [THAMH3].
compound 4 and Si(egH)4, it likely provides some evidence for
this assumption.
29Si NMR Evidence for Proposed Equilibrium. As men-
tioned above, an equilibrium between tetra- and pentacoordi-
nated silane species may exist. For THAMH3-catalyzed reac-
tions, the likely equilibrium is given in eq 5. To test this
Amine Concentration Studies. A third set of kinetic studies
was run to explore the effects of initial base concentrations on
dissolution rates by using different initial THAMH3 concentra-
tions under standard reaction conditions. Theoretically, if
THAMH3 functions only as catalyst, the reaction rate should
be linearly dependent on initial THAMH3 concentration. The
Figure 3 results do not show such a trend.
Although reaction rates increase with increases in initial
THAMH3 concentrations, the increase is not linear and reaction
rate increases fall behind increases in initial THAMH3 concen-
trations. This change in dissolution rates can be explained if
THAMH3 also acts as a reactant in addition to serving as a
catalyst.
This was not unexpected because from previous studies, we
know that TEAH3 reacts with silica to form silatrane glycol
when equivalent amounts are used, (3). Thus, we examined silica
possibility, tetraethoxysilane (TEOS) was mixed with excess
EGH2 and heated at 155 °C to form Si(egH)4 by distillative
removal of ethanol. The resulting Si(egH)4 was cooled to room
temperature before an equivalent amount of THAMH3 was
added. The mixture was then heated at 200 °C for a pre-set
time (10 min and 1, 3, and 12 h) and then cooled immediately
in ice water. 29Si NMRs were then recorded. Two signals were
observed for all samples: a major signal at -106 ppm and a
weak signal at -82 ppm, corresponding to penta- and tetra-
alkoxy silanes, respectively. This observation clearly supports
the existence of the above equilibrium.
A crude estimate of the equilibration process can be obtained
by measuring the peak heights of the two signals for each
spectrum and calculating the relative ratio of these two signals.
The peak ratios as a function of reaction time are plotted in
Figure 4, which indicates that the equilibrium favors formation
of the pentacoordinate species. Furthermore, Figure 4 suggests
that this equilibrium is established very quickly as the peak ratio
at 10 min is quite close to the remaining data points. The ratios
of the two peak heights are ∼8:1, which roughly indicates the
ratio of the two species in solution.
dissolution in EGH2 with an equivalent amount of THAMH3
and found reaction 4 (see product identification section). This
suggests that even when THAMH3 is used in catalytic amounts,
reaction 4 occurs. However, if compound 4 can also react with
EGH2 to generate tetraalkoxysilanes and release “free” amine,
then this may be the source of base needed to catalyze silica
dissolution. Because EGH2 is used in large excess, it is likely
that the tetraalkoxysilane formed would be Si(OCH2CH2OH)4
[Si(egH)4]. Thus, if we can observe an equilibrium between
Similar experiments were run with TETA. That is, the Si-
(egH)4 in EGH2 was mixed with an equivalent amount of TETA
following the same experimental procedure described above and
in the Experimental Section, and 29Si NMRs were taken. Again,
we observed two signals, one at -106 and one at -82 ppm for
reactions of 10 min, 1 h, and 3 h with the -106 ppm peak as
the major peak. The ratio of the two peaks is ∼11 (-106/-82

10068 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Cheng et al.
Figure 4. 29Si NMR peak height ratios -106:-82 ppm for reaction
of Si(egH)4 with equivalent amounts of THAMH3 as a function of
reaction time.
ppm) for these reaction times. This supports the assumption that
the equilibrium between the tetraalkoxysilanes and the pen-
taalkoxysilicates strongly favors the formation of pentaalkoxy-
silicates. Similar to the THAMH3 reaction, equilibration is very
quick as the 10-min sample has a peak ratio almost identical to
those at longer reaction times.
Figure 5. (a) Silica dissolution as a function of temperature with 25
mol % THAMH3. (b) Arrenius plot for part a.
and (6) are rapidly established such that most of the amine base
in solution reacts with tetraalkoxysilane to form the pen-
taalkoxysilicates. The overall effect is that the concentration of
“free” alkoxide/amine in solution, essential to the silica dis-
solution process, drops to such a low level that (1) the
differences in basicity between these amines are no longer
relevant and (2) the silica surface areas are large enough to be
treated as a constant, resulting in pseudo-zero order reaction
rates.
These equilibria strongly control the rate-limiting step for
amine-catalyzed silica dissolution. In our previous studies with
equivalent amounts of alkali base, water removal from the silica
surface was proposed to be the rate-limiting step.10 However,
for silica dissolution with catalytic amounts of amines, the
kinetic studies suggest that the above-mentioned equilibria limit
the “free” amine concentrations so much that the rate-limiting
step changes to the step of generating “free” alkoxide. Further
support for this comes from kinetic studies wherein isolated
THAMSiegH (see below and the Experimental Section) is used
in place of added free THAMH3 in the dissolution vs time
studies (Figure 2). The dissolution rate is essentially identical
and linear to that obtained with free THAMH3.
Temperature Effects. A fourth set of kinetic studies was
run to follow silica dissolution as a function of temperature at
a constant [THAMH3] of 25 mol %. The results, plotted in
Figure 5a, indicate that the dissolution process over the 170 to
200 °C range has an activation energy of 58 kJ/mol or 14 kcal/
mol (Figure 5b).
Silica Dissolution Mechanism with Catalytic Amounts of
Amine Bases. Kinetic studies show that the three amines,
TETA, TEAH3, and THAMH3, all catalyze silica dissolution
in EGH2. The catalytic efficiencies are quite similar despite 3
orders of magnitude differences in basicities. The dissolution
rates appear to be pseudo-zero order with respect to silica surface
area. 29Si NMR studies reveal the existence of equilibria between
the free amines and tetraalkoxysilanes and the ammonium
pentaalkoxysilicates that favors the pentaalkoxysilicates. With
the above facts in mind, we can propose a dissolution mecha-
nism similar to that shown in Scheme 1.
Dissolution most likely begins with amine deprotonation of
ethylene glycol to form an alkoxide and ends with the release
of spirosiloxane from the silica surface (Scheme 3). In addition
to the steps shown in Scheme 1, the equilibria shown in (5)

Neutral Alkoxysilanes from Silica
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10069
Scheme 3. Proposed Mechanism for Silica Dissolution with
Catalytic Amounts of Organic Amines
provide similar catalytic efficiencies, KOH was chosen for these
studies.
Base Concentration Studies. First, silica dissolution as a
function of initial [KOH] was examined. Standard reaction
conditions were applied with KOH at 1-10 mol % of silica
used. Given the fact that the above comparative tests showed
70% dissolution in 3 h which does not allow comparison of
initial rates, 1 h reaction times were used. A blank without KOH,
heating for 1 h, gave 13 mol % dissolution.
Results from this set of experiments, plotted in Figure 7, show
that silica dissolution increases linearly as [KOH] increases up
to 3 mol % (Figure 7b). This linear relationship is expected for
a purely catalytic reaction, unlike that found for THAM, Figure
3. At 3 mol % [KOH], dissolution is essentially quantitative,
thus the data beyond [KOH] of 3 mol % lose their meaning.
29Si NMR Study of Possible Equilibria. Similar to the
studies of amine reactions, Si(egH)4 was mixed with an
equivalent amount of KOCH2CH2OH and heated at 200 °C for
a set of time before it cooled and NMR spectra were taken. A
single peak at -105 ppm is observed for all spectra taken from
this series, even for sample heated just 15 min. However, the
S/N ratio is ∼15:1 (Bruker 360 MHz) after 3000 scans, which
sets the signal detection limit. Thus, if the equilibrium between
penta- and tetraalkoxysilanes is >15:1, it cannot be detected.
Alkali Base Catalysis. Our previous studies demonstrated
that stoichiometric amounts of alkali bases dissolve silica in
ethylene glycol to produce pentacoordinate anionic glycolato-
silicate complexes.10 However, their utility as catalysts was not
tested. Given that organic amines catalyze silica dissolution in
ethylene glycol, we decided to run a comparative experiment
wherein 10 mol % of KOH was used under the standard reaction
conditions. In 3 h, 70 mol % of silica dissolution was obtained,
much higher than found with the amine bases, which are all
∼20 mol % under the same conditions. Following this initial
result, other alkali bases were also tested under identical
conditions using 1 h instead of 3 h reaction times to obtain initial
dissolution rates. All these alkali bases give similar dissolution
yields of ∼40 mol %.
The primary difference between (7) and equilibria 5 and 6 is
that in (7), alkoxide is directly regenerated on equilibration,
unlike in (5) or (6) where the amine is first produced and
alkoxide must then be regenerated by deprotonation of EGH2.
Temperature Effects. The effect of temperature on the initial
reaction rates for KOH-catalyzed silica dissolution is plotted
in Figure 8a. The activation energy calculated from Figure 8b
gives 20.6 kJ/mol, or ∼5 kcal/mol. This activation energy is
lower than the 14 kcal/mol value found in our previous study
with equivalent amounts of KOH,10 which implies a different
rate-limiting step.
Dissolution Mechanism with Alkali Base Catalysis. Silica
dissolution catalyzed by alkali base most likely goes through
the same cycle as the amine-catalyzed one. However, the
important difference is the rate-controlling step. In the amine-
catalyzed reaction, generation of free amine or alkoxide
thereafter appears to be rate limiting. In the alkali base catalyzed
reaction, the alkoxide is directly regenerated during equilibration
per (7); however, the 29Si NMR attempts to detect the presence
of the tetraalkoxysilane in a 1:1 mixture of alkoxide and
tetraglycoxysilane after just a few minutes of reaction were
unsuccessful. This suggests that this equilibrium also strongly
favors the pentacoordinated species. Hence, direct generation
of the alkoxide may still be rate limiting in alkali base catalysis
but it is not complicated by a deprotonation step. However, one
crucial set of experiments proves the process quite different.
A study of dissolution rate vs time (analogous to the Figure
2 studies) reveals that the rate slows (data not shown) in a typical
first-order fashion. This indicates that the rate-determining step
occurs after glycoxide attack at the silica surface. The found Ea
of 5 kcal/mol is different from the 14 kcal/mol found for the
stoichiometric reaction, thus the rate-controlling step must be
different. It may be that the slow step involves one of the surface
Figure 6. Silica dissolution with alkali hydroxides (1 h, 1 mol %
MOH).
These much faster dissolution rates for the alkali base-
catalyzed reaction reflect the differences in base strengths
between the alkali and amine bases. To further understand the
catalytic effects of alkali bases, kinetic studies similar to those
done with the amines were conducted. Since all the alkali bases

10070 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Cheng et al.
Figure 7. (a) Silica dissolution as a function of KOH concentration.
(b) Expansion of the plot for [KOH] up to 3 mol %.
Figure 8. (a) Silica dissolution as a function of temperature with 1
mol % KOH. (b) Arrenius plot.
reorganization processes proposed in Scheme 1; however, further
studies are necessary to delineate the exact course of events.
Characterization of the Products from Silica Dissolution.
To better understand silica dissolution in EGH2 catalyzed by
either high-boiling amines or alkali bases, we attempted to
isolate and identify the dissolution products.
one singlet at -82 ppm is observed, whereas species with Si-
O-Si bonds appear at ca. -97 ppm (see below).
TETA-Catalyzed Dissolution. Silica dissolution in ethylene
glycol using catalytic amounts of TETA under continuous
distillation conditions gives a clear solution with a 29Si NMR
exhibiting a single peak at - 82 ppm, which is in the same
position as seen for neat Si(OEt)4. This suggests the formation
of neutral alkoxysilanes in solution. The simplest structure would
be a spirosiloxane, Si(OCH2CH2O)2, which would be expected
if the mechanism shown in Scheme 1 is correct. However, Frye’s
efforts to prepare this compound were unsuccessful.28 It is likely
that in the presence of excess ethylene glycol and some base,
this spirosiloxane reacts with ethylene glycol to form Si(OCH2-
CH2OH)4 [Si(egH4)] and/or oligomers thereof. No compounds
containing Si-O-Si bonds are present in solution since only
Continuous removal of solvent from the TETA-catalyzed
dissolution in vacuo (220 °C) gives a brittle, transparent cross-
linked polymer [Si(eg)2] (reactions 8 and 9). This solid
(28) Frye, C. L. J. Am. Chem. Soc. 1969, 34, 2496.

Neutral Alkoxysilanes from Silica
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10071
Scheme 4. Transformation of Si(eg)2
Table 1. Si(eg)2Characterization Data
Table 2. Silatrane Glycol Characterization Data
a Sample dissolved in EG-d6. Chemical shift of residual 1H (in
CD3OD) δ 5.21 (br), 3.34 (multiplet); 13C δ 62.31 (pentet, JCD ) 21.3
Hz). b Peak present in some samples only. c Species containing Si-
O-Si linkages. d Pentacoordinate silicon.
completely redissolves in hot EGH2. The solution 29Si NMR of
this redissolved Si(eg)2 in EG exhibits a major peak at -82.1
ppm along with minor peaks at -88.3, -96.3, and -104.2 ppm.
The solid-state 29Si NMR was also taken and exhibits the same
peaks. The NMR data are listed in Table 1.
As mentioned above, the peak at -82.1 ppm arises from
tetraalkoxysilane. The -104.2 ppm peak arises from the
ammonium analogue of compound 1 formed from the residual
TETA retained in the solid (see below). The two peaks at -88.3
and -96.3 ppm most likely result from formation of Si-O-Si
containing species during the solvent removal process. For
example, the polymeric species Si(eg)2 on prolonged heating
may undergo â-elimination reactions leading to formation of
Si-OCHdCH2 and Si-OH in the presence of amine bases.
The latter goes on to form Si-O-Si bonds. Indirect support
comes from the observation that the -88.3 and -96.3 ppm
peaks in 29Si solution and solid-state NMR spectra are more
significant for products obtained following prolonged heating
at 220 °C. Nevertheless, these peaks are small, which suggests
that the Si-O-Si content is low.
TGA analysis of the dry solid gives a ceramic yield of 39.8%.
If we assume that the solid is polymeric “Si(eg)2“ (FW ) 148.19
Da), and that Si(eg)2 decomposes only to SiO2, then the
calculated ceramic yield would be 40.5%, which is very close
but a little higher than the observed value. Chemical analysis
of the dry solid gives C ) 31.3, H ) 6.1, and N ) 0.69. If the
N content defines the residual TETA, one can recalculate the
expected ceramic yield of this solid to be 39.8%, which is
exactly that found by TGA, supporting the formulation of a
fully crosslinked polymer, Si(eg)2.
1
a Sample dissolved in EGH2-d6. Chemical shift of residual H (in
CD3OD) δ 5.21 (br), 3.34 (multiplet); 13C δ 62.31 (pentet, JCD ) 21.3
Hz). b Sample dissolved in EGH2.
Triethanolamine Reactions. Silatrane glycol (TEASiegH)
can be produced directly from silica in quantitative yield
(reaction 3). The compound has been characterized by high-
resolution MS, solution NMRs, chemical analysis, and TGA as
summarized in Table 2.29
Silatrane glycol can be used immediately in conjunction with
alumatrane [N(CH2CH2O)3Al] and related compounds to for-
mulate precursors to mullite (3Al2O3‚2SiO2), cordierite (2MgO‚
5SiO2‚2Al2O3),30 barium aluminosilicate (BAS, BaO‚2SiO2‚
Al2O3), strontium aluminosilicate (SAS, SrO‚2SiO2‚Al2O3),30
and porous silicon oxynitride particles.31 Additional studies show
that silatrane glycol is soluble and stable in cold, neutral water
for days.32 We have also successfully used silatrane glycol as a
starting material to synthesize other silatrane compounds
The TGA analysis also shows this Si(eg)2 polymer has good
thermal stability in dry, synthetic air to ∼390 °C, which suggests
potential utility as a fire retardant material. Si(eg)2 can also be
easily transformed to other useful silicon-containing compounds
(Scheme 4).
(29) Laine, R. M.; Treadwell, D. R.; Mueller, B. L.; Bickmore, C. R.;
Waldner, K. F.; Hinklin, T. R. J. Mater. Chem. 1996, 6, 1441.
(30) Laine, R. M.; Sutorik, A. C.; Neo, S. S.; Treadwell, D. R. J. Am.
Ceram. Soc. 1998, 81, 1477.
(31) Laine, R. M.; Bickmore, C. R. J. Am. Ceram. Soc. 1996, 79, 2865.
(32) Cheng, H.; Laine, R. M. New J. Chem. 1999, 23, 1181.

10072 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Table 3. THAMSieg Characterization
Cheng et al.
Figure 10. Proposed structures for species observed in MS spectra.
273 (M + H+). But the relative intensity of this peak is very
low in most cases. The peaks at m/z 149 and 255 have relatively
high intensities. These are likely fragment peaks of the molecular
ion: the m/z 255 results from the (M + H+) molecule losing
one molecule of H2O and the m/z 149 is the (M + H+) molecule
losing two molecules of EGH2, or the putative spirosiloxane
(Figure 10).
Thus, the NMR and MS data support the conclusion that the
tetraalkoxysilanes, Si(egH)4, and its oligomeric derivatives are
the main components of silica dissolution in ethylene glycol.
As ethylene glycol is removed, the degree of cross-linking
increases and finally completely cross-linked Si(eg)2 is formed.
Conclusions
New, low-cost syntheses of tetra- and pentacoordinated silicon
alkoxides have been developed. Silica is found to readily
dissolve in ethylene glycol in the presence of catalytic amounts
of TETA, or TEAH3, or THAMH3. The products obtained are
the neutral monomer Si(OCH2CH2OH)4 and oligomeric forms,
e.g. Si(eg)2. This low-cost synthesis route may impact the use
of chemical methods of processing oxide ceramic materials for
applications ranging from ceramic fibers for structural applica-
tions to nanosized ceramic powders for diverse applications.
Kinetic studies show that all the amines provide similar
catalytic efficiencies and the dissolution rate is pseudo-zero order
to silica surface areas. These observations can be explained by
equilibria between free amines and neutral tetraalkoxysilanes,
and amine complexed pentacoordinated silicates, as supported
by NMR studies. These equilibria greatly reduce the available
“free” amine concentrations needed to promote dissolution
thereby limiting dissolution rates.
Figure 9. MS peak intensities vs reaction time for a 5 mol % NaOH
catalyzed silica dissolution.
containing potentially polymerizable groups. These chemical
transformations have been reported separately.32
Trishydroxymethyleneaminomethane Reactions. As men-
tioned above, THAMH3 serves as a catalyst at low concentra-
tions but functions as a reactant at high concentrations. White
solid can be isolated from the reaction of silica with an
equivalent amount of THAMH3 in EGH2 by removing the
Pentacoordinated silanes can also be obtained in high yield
by reacting equivalent amounts of amine bases with silica in
ethylene glycol. These compounds also provide unique proper-
ties and are useful for diverse applications.
1
excess EGH2 under vacuum. 29Si, H, and 13C NMR spectra
and chemical analysis all agree with the proposed structure
(Table 3).
The alkali hydroxides are found to be more active catalysts
for silica dissolution in ethylene glycol than the organic amines.
The dissolution rates are orders of magnitude greater than those
found using organic amine bases. The products of silica
dissolution with the alkali bases consist mostly of monomeric
Si(OCH2CH2OH)4 in solution and fully cross-linked Si(eg)2 after
solvent removal.
Products from Alkali Bases Catalyzed Silica Dissolution.
The reaction solutions obtained using catalytic amounts of
NaOH were examined by NMR. All give similar NMR signals
(Table 3). Typically, three signals are observed: a major peak
at -82.1 ppm and two minor peaks at -89.0 and -105.9 ppm.
The -105.9 ppm peak corresponds to Na+[HOCH2CH2OSi-
(OCH2CH2O)2]-. The major peak at -82.1 ppm, identical to
the product from amine-catalyzed silica dissolution, suggests
the tetraalkoxysilane, Si(egH)4, and/or its oligomeric derivatives
are the major products in solution. Similar to amine-catalyzed
reactions, the peak at -89.0 ppm may arise from Si-O-Si
bonds.
Acknowledgment. We would like to thank FAA for their
generous support of this work through grant No. 95-G-026. We
also thank Tal Materials, Inc. of Ann Arbor for partial support
of this work through NSF SBIR sub-contract (grant No. DMI
9661236). Many of the compounds synthesized in this paper
are now available from talmaterials.com.
Mass spectra (MS) using fast atom bombardment in positive
ion mode (FAB+) were also obtained. A typical plot of ion
intensity as a function of reaction time is shown in Figure 9.
The molecular peak corresponding to Si(egH)4 should be at m/z
JA001885H