Convenient two step synthesis of29Si labelled tetraalkoxysilanes

Article abstract of DOI:10.1039/d0cc05525e

29Si enrichment would be advantageous for many NMR studies of the structural properties of sol-gel derived materials. The required starting materials (29Si enriched alkoxysilanes), however, are expensive and difficult to provide. Here we present a scalable, reliable synthesis of 29Si enriched tetraethoxysilane starting from silicon dioxide or elemental silicon with a total yield up to 75%.

Full text of DOI:10.1039/d0cc05525e

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
2
9
Convenient two step synthesis of Si labelled
tetraalkoxysilanes†
Cite this: Chem. Commun., 2020,
6, 13631
5
a
a
a
a
Robert Gericke, ‡ Christoph Gondek, Andre Stapf,
Jorg Wagler,
¨
´
a
b
Received 13th August 2020,
Accepted 22nd September 2020
Edwin Kroke
and Erica Brendler
*
DOI: 10.1039/d0cc05525e
rsc.li/chemcomm
2
9
Si enrichment would be advantageous for many NMR studies of overcome this problem are Dynamic Nuclear Polarization
the structural properties of sol–gel derived materials. The required (DNP) methods. But these are limited to high surface systems
29
starting materials ( Si enriched alkoxysilanes), however, are expen- or materials which allow to incorporate radicals without chan-
sive and difficult to provide. Here we present a scalable, reliable ging the structure, and of course to the availability of the
synthesis of Si enriched tetraethoxysilane starting from silicon technique. Another approach uses the reduction of the silicon
29
3
2
+
dioxide or elemental silicon with a total yield up to 75%.
relaxation times by adding paramagnetic cations such as Mn
2
+ 4
and Ni . The incorporation of these cations is not always
Tetraalkoxysilanes, tetraethoxysilane (TEOS) in particular, are possible from a synthetic point of view and may also have an
essential starting compounds for a vast range of syntheses in unfavourable influence on the material properties.
1
,2
29
sol–gel chemistry.
As long as the resulting materials are
Therefore, Si enrichment is an alternative way to signifi-
crystalline, X-ray diffraction methods can be applied for char- cantly improve the sensitivity of the experiments without alter-
2
9
acterization of their structures. As a matter of fact of sol–gel ing of other properties of the products. Si enriched materials
chemistry, a great share of the materials obtained by these can be purchased as silica (SiO ) or elemental silicon, with the
routes are completely (or at least in great parts) amorphous. advantage of the former which can be used as obtained for
2
5
,6
Thus, NMR spectroscopy is the method of choice to investigate some glass or zeolite syntheses.
For other applications,
2
9
the structure of these products. Especially the Si nucleus is compounds synthesized therefrom and used to label the mate-
suited to gain insights into the structural units formed and rials under investigation are silicon tetrachloride SiCl , tetra-
their interconnections. With a spin of 1/2, resulting in long phenoxysilane Si(OPh) , or sodium metasilicate Na SiO3
2
9
4
2
9
29
4
2
7
,8
relaxation times, and only 4.7% natural abundance of the NMR and its solution.
2
9
accessible isotope Si, investigations are very time consuming. For access to TEOS, several synthetic routes have been
Broad signals (caused by amorphicity) and the need of using proposed in the literature starting from SiO by reaction with
2
direct pulse techniques (rather than cross-polarization) for dialkylcarbonates or alcohols and potassium hydroxide in an
9
quantitative evaluation of individual Si coordination spheres autoclave reaction. Yields up to 80% were reported, depending
2
already pose the first hurdles. Even more problematic, when on the SiO quality. Small particle, high surface oxides like rice
2
9
29
it comes to Si– Si-correlation experiments for further eluci- hull ash resulted in high yield, whereas for other qualities the
dation of the wider Si site environment, NMR experiments on reaction output can drop dramatically.
2
9
samples with natural abundance Si are, in most cases, not
Elemental silicon as a starting material is reacted with alcohols
10,11
practicable within a reasonable time frame. One way to with and without (mainly copper based) catalysts.
Either
trialkoxysilane/tetraalkoxysilane mixtures are obtained, or the
conversion rate for silicon, especially for the ethoxy silanes, was
a
b
Technische Universit a¨ t Bergakademie Freiberg, Department of Chemistry and
Physics, Institute of Inorganic Chemistry, Leipziger Straße 29, 09599 Freiberg,
Germany
2
9
too low to be tolerable for the price level of Si enriched
material. In addition, the reaction started with removal of
Si–O-groups from the surface by cleaning with HF, which would
reduce the yield and thus makes this route less desirable when
working with enriched material. Only a reaction using an
autoclave and a copper based catalyst resulted in yields of up
to 80% in TEOS, but again required fine powdered silicon in
Technische Universit a¨ t Bergakademie Freiberg, Department of Chemistry and
Physics, Institute of Analytical Chemistry, Leipziger Straße 29, 09599 Freiberg,
Germany. E-mail: Erica.Brendler@chemie.tu-freiberg.de
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0cc05525e
Current address: School of Chemistry, Trinity College Dublin, The University of
‡
1
2
Dublin, College Green, Dublin 2, Ireland.
100 nm range.
This journal is ©The Royal Society of Chemistry 2020
Chem. Commun., 2020, 56, 13631--13633 | 13631

View Article Online
Communication
ChemComm
Scheme 1 Reaction pathway of the TEOS synthesis.
When working with the very expensive isotopic enriched
material, a reliable and reproducible synthetic route, which
does not so highly depend on the morphology of the starting Fig. 1 Sections of the H NMR spectrum of Si enriched TEOS with
1
29
J(Si,H).
material, is needed. Here, we present a convenient way to
2
9
obtain Si enriched TEOS in high yield (475%) independent
of the SiO or silicon particle size, which in addition can be yield contaminated with 25% SiF(OEt) , respectively. In reflux-
2
3
2
9
scaled to the amount of the available Si source.
ing tetrahydrofuran (THF) TEOS formation was not observed.
The overall synthesis route is sketched in Scheme 1. Our key
The NMR data for the obtained TEOS are given in Table S1
reaction is the synthesis of TEOS from Na
which can be accessed from both SiO and Si in a similar The enriched material was successfully used for the synthesis and
manner. The starting materials (SiO or Si) are dissolved in characterization of phosphosilicates.
concentrated HF or a HF/HNO mixture, respectively. The
A further advantage of the method is, that the silicon can be
2 6
SiF , the latter of (ESI†), the spectra are provided in Fig. 1 and the ESI† (Fig. S1–S3).
2
19
2
3
dissolution is essentially quantitative, irrespective of the grain recovered from the materials as SiO2 after performing all
size of the silicon dioxide or silicon used. Of course, smaller investigations of interest and can thus be converted again to
2
9
6
particles are favourable regarding the reaction time, but are not
essential for the yield and can be accounted for by slightly laboratory scale amounts but can also be used to produce larger
Si enriched TEOS. The synthesis can be applied to small
2
9
prolonged reaction time. The dissolution of silicon in HF/HNO
3
badges of TEOS, limited only by the amount of Si enriched
allows the use of the silicon irrespective of its surface condition. starting material available. According to our tests the yield may
Using silicon as a starting material we face an exothermic even further improve when the synthesis is scaled up. The
reaction which brings about the risk of stripping SiF
losing silicon via exhaust gas flow. Thus, damping the procedure should be adaptable to other calcium alkoxides (e.g.
reaction by adding the silicon in small portions after the Ca(OMe) ) to afford the respective Si enriched alkoxysilanes
4
and experiments in this work aimed at obtaining TEOS, but the
1
3,14
2
9
2
2
9
dissolution of the previous portion improved the yield signifi- (e.g. Si(OMe) ). However, adjustment of the solvent used for
4
cantly. Na
This pair of salts was chosen because of the good solubility of ent boiling point) in order to avoid contamination of the final
NaCl in HF solution, and there is no hydrate of Na SiF known tetraalkoxysilane with traces of the solvent used.
to date as it would be the case for hexafluorosilicates of other
2 6
SiF is then precipitated by salting out with NaCl. calcium alcoholate preparation might be necessary (to a differ-
2
6
1
5
The financial support from the German Research Founda-
tion (DFG; KR 1739/32-1; BR 1540/5-1) is gratefully acknowl-
1
6
cations, e.g., Mg, Ca, Pb, and first row transition metals.
Furthermore, related attempts of alkoxylation starting from edged. The authors thank Claudia Sieber for assistence during
K SiF did not lead to TEOS formation. In a second step dry literature research and Lukas Jahn for synthesis support.
2
6
Na
2
SiF
6
is mixed and heated with freshly prepared calcium
Further details of the synthesis are given in the Experi-
mental footnote§, and the ESI†.
2
9
ethanolate Ca(OEt)
For the conversion of Na
carefully dried solvents and Na
inert conditions. Ca(OEt) was chosen because of its easy
2
to give Si enriched TEOS.
SiF into TEOS it is essential to use
SiF and to work under dry and
2
6
2
6
Conflicts of interest
2
accessibility from ethanol and elemental calcium as well as
because of the formation of CaF2 as a driving force for the
There are no conflicts to declare.
reaction. The fluoroethoxysilanes SiF (OEt) (b.p.: À7 1C), Notes and references
3
1
7
SiF
2 2 3
(OEt) (b.p.: 84 1C) and SiF(OEt) (b.p.: 135 1C), despite
§
Experimental: All manipulations and reactions involving ethanolates
their lower boiling points in comparison to TEOS, have not
been observed in the final product. We address this observation standard Schlenk techniques. Si enriched SiO was purchased from
2
were carried out under inert conditions (dry Ar atmosphere) using
29
2
+
CortecNet, France. Calcium and diethyl ether were purchased from
ABCR, ethanol and n-hexadecane from VWR, 48% HF solution and 67%
HNO solution from Honeywell, and sodium chloride from ESCO.
3
All synthesis steps were trained and optimized with natural abun-
dance materials in a scale planned for the final synthesis. Yields for
SiO
of Na
to the presence of a fluoride trap (Ca ) for efficient F/OEt
substitution, inherent fluorosilicate formation through the first
stages of substitution (thus keeping the fluorosilanes trapped
as solids) and, in case of release of gaseous fluoroethoxy-
dissolution to obtain Na
silanes their tendency to undergo substituent redistribution.
SiF with Ca(OEt) resulted in 75–85% TEOS. After obtaining
6
2
Heating a mixture of powders of Ca(OEt) and Na SiF gave reproducible high yields the SiO route was performed for two batches
2 6
SiF were between 70–95%. The reaction
1
8
2
2
6
2
2
2
with enriched material, details of the larger one are given in the
Supporting Information. It was observed that the yield was comparable
to the training synthesis and slightly improved for the larger SiO
the best results. Performing the reaction in the presence of a
high boiling solvent, i.e., tetraethylene glycol dimethyl ether
29
2
(Tetraglyme) or n-hexadecane, did not succeed or gave only 13% batch compared to the smaller.
13632 | Chem. Commun., 2020, 56, 13631--13633
This journal is © The Royal Society of Chemistry 2020

View Article Online
ChemComm
Communication
For initial experiments with natural abundance silicon starting
materials chromatography grade Silica gel (SiO , 60 Å, 60–100 mesh,
2
5 (a) I. I. Ivanova, Y. G. Kolyagin, I. A. Kasyanov, A. V. Yakimov,
T. O. Bok and D. N. Zarubin, Angew. Chem., Int. Ed., 2017, 56,
15344; (b) S. Cadars, A. Lesage, N. Hedin, B. F. Chmelka and
L. Emsley, J. Phys. Chem. B, 2006, 110, 16982; (c) A. R. Brough,
C. M. Dobson, I. G. Richardsen and G. W. Groves, J. Am. Ceram. Soc.,
1994, 77, 593.
Davisil grade 635) purchased from Aldrich and silicon powder o200 mm
kindly provided by SolarWorld were used.
Organic solvents were dried and purified according to standard
methods and stored under Ar atmosphere over molecular sieve 3 Å
(
ethanol) or sodium (n-hexadecane). Solution NMR spectra were recorded
6 E. Ebrecht, PhD thesis, TU Bergakademie Freiberg, 2003.
7 D. Wisser, S. I. Br u¨ ckner, F. M. Wisser, G. Althoff-Ospelt,
J. Getzschmann, S. Kaskel and E. Brunner, Solid State Nucl. Magn.
Reson., 2015, 66–67, 33.
8 (a) A. McCutcheon, J. Hu, G. S. Kamali Kannangara, M. A. Wilson
and N. Reddy, J. Non-Cryst. Solids, 2005, 351, 24; (b) K. Spinde,
K. Pachis, I. Antonakaki, S. Paasch, E. Brunner and K. D. Demadis,
Chem. Mater., 2011, 23, 4676; (c) S. I. Br u¨ ckner, S. Donets, A. Dianat,
M. Bobeth, R. Guti ´e rrez, G. Cuniberti and E. Brunner, Langmuir,
2016, 32, 11698; (d) F. Kolbe, F. Daus, A. Geyer and E. Brunner,
Langmuir, 2020, 36, 4332.
1
13
on a Bruker Nanobay 400 MHz spectrometer ( H, 400.13 MHz; C,
1
29
00.61 MHz, Si, 79.49 MHz). The chemical shifts were referenced to SiMe
Synthesis of calcium ethanolate Ca(OEt) : See ESI†, p. 4.
Synthesis of sodium hexafluorosilicate (Na SiF ): Caution! Experi-
4
.
2
2
6
ments with hydrofluoric acid must be performed in a special equipped
HF-laboratory in a HF-approved fume hood with HF-resistant labora-
tory and personal protective equipment.
29
29
Starting from SiO
2 2
: To a slurry of fine powdered SiO (1.800 g,
2
9.51 mmol) and water (3 mL) in a beaker hydrofluoric acid (5 mL, 48%
HF) was added dropwise, and the suspension was stirred at 35 1C. After
approx. 40 min further 48% HF (2 mL) was added, and stirring was
9 (a) M. Akiyama, E. Suzuki and Y. Ono, Inorg. Chim. Acta, 1993, 207, 259;
(b) Y. Ono, M. Akiyama and E. Suzuki, Chem. Mater., 1993, 5, 442;
(c) N. Fukaya, S. J. Choi, T. Horikoshi, S. Kataoka, A. Endo, H. Kumai,
M. Hasegawa, K. Sato and J.-C. Choi, New J. Chem., 2017, 41, 2224.
continued until all SiO
0.51 mmol) was added, and the solution was stirred for 1 h. Then, the
mixture was allowed to attain room temperature and was stored undisturbed
for crystallization overnight. The precipitate of Na SiF
6
was filtered, washed 10 M. N. Temnikov, A. S. Zhiltsov, V. M. Kotov, I. V. Krylova,
2
has been dissolved. Thereafter, NaCl (3.536 g,
6
2
with a 2 : 3 water–ethanol mixture (3 Â 3 mL), then air-dried for at least one
day and finally dried in vacuo until constant weight. Yield Z = 88%.
Starting from silicon: See ESI†, p. 4.
M. P. Egorov and A. M. Muzafarov, Silicon, 2015, 7, 69.
11 I. V. Krylova, M. P. Egorov and O. M. Nefedov, Russ. Chem. Bull.,
2017, 66, 260.
Synthesis of tetraethoxysilane (TEOS): Under an Ar atmosphere, a 12 O. Y. Alhussein, A. M. Elgorashi and H. Bayahia, Asian J. Chem.,
29
2
flask was charged with Na
ethanolate (10.0 g, 76.0 mmol). The compounds were thoroughly mixed 13 M. Lippold, U. B ¨o hme, C. Gondek, M. Kronstein, S. Patzig-Klein,
and vacuum-dried for 5 h. Thereafter, the contents were set under Ar
M. Weser and E. Kroke, Eur. J. Inorg. Chem., 2012, 5714.
SiF
6
(4.881 g, 25.96 mmol) and calcium
2016, 28(12), 2741.
atmosphere, and the flask with the white mixture was connected to a 14 A. Stapf, C. Gondek, E. Kroke and G. Roewer, Wafer Cleaning,
shortcut still. Using a metal alloy bath (Wood’s metal: 50% Bi, 25% Pb,
2.5% Cd, 12.5% Sn) the dry mixture was slowly heated to 280 1C,
Etching, and Texturization, in Handbook of Photovoltaic Silicon,
ed. D. Yang, Springer-Verlag, Heidelberg, 2019.
1
whereupon TEOS was distilled from the mixture (at 280–290 1C (metal 15 F. I. Z. Karlsruhe, Inorganic Cambridge Structure Database - ICSD,
bath temperature), and the entire fraction in the range 130–168 1C
version 4.3.0, 2020.
2
0
(
head temperature) was collected within approx. 20 min. m = 4.586 g, Zdist
=
16 (a) T. G. Cherkasova, E. S. Tatarinova and E. V. Cherkasova, Zh.
Neorg. Khim., 2004, 49, 1161; (b) S. Frisoni, S. Brenna and
N. Masciocchi, Powder Diffr., 2011, 26, 308; (c) A. M. Golubev,
A. M. Ilinets, A. N. Tsvigunov and L. G. Makarevich, Kristallografiya,
29
1
3
8
5%, Ztotal = 75%, d Si [ppm]: À81.7; d H [ppm]: 1.17 (dt, CH
3
, J(H,H) 7 Hz,
4
3
3
13
2
J(Si,H) 0.44 Hz), 3.78 (dq, CH , J(H,H) 7 Hz, J(Si,H) 2.8 Hz); d C [ppm]: 18.1
2
3
(d, CH
3
, J(Si,C) 2.4 Hz), 59.2 (d, CH
2
, J(Si,C) 0.95 Hz).
1991, 36, 214; (d) F. A. Cotton, L. R. Falvello, C. A. Murillo and
1
2
3
C. J. Brinker and G. W. Scherer, Sol–gel science, the physics and
J. F. Quesada, J. Solid State Chem., 1992, 96, 192; (e) A. Torn,
K. Ogawa, H. Tamura and K. Osaki, Acta Crystallogr., 1997,
C53, 833; ( f ) P. G. Chevrier, A. Hardy and G. J ´e hanno, Acta Crystal-
logr., 1981, A37, 578; (g) S. Ray, A. Zalkin and D. H. Templeton, Acta
Crystallogr., 1973, B29, 2741; (h) S. Ray, A. Zalkin and
D. H. Templeton, Acta Crystallogr., 1973, B29, 2748.
chemistry of sol–gel processing, Academic Press, Boston, 1990.
U. Schubert and N. H u¨ sing, Synthesis of inorganic materials, Wiley-
VCH, Weinheim, 2019.
(a) M. Lelli, D. Gajan, A. Lesage, M. A. Caporini, V. Vitzthum,
P. Mieville, F. H ´e roguel, F. Rascon, A. Roussey, C. Thieuleux,
M. Boualleg, L. Veyre, G. Bodenhausen, C. Coperet and L. Emsley, 17 J. Emel ´e us and H. G. Heal, J. Chem. Soc., 1949, 1696.
J. Am. Chem. Soc., 2011, 133, 2104; (b) D. Lee, G. Monin, N. T. Duong, 18 M. G. Voronkov, E. V. Boyarkina, I. A. Gebel, A. I. Albanov and
I. Z. Lopez, M. Bardet, V. Mareau, L. Gonon and G. De Pa ¨e pe, J. Am.
Chem. Soc., 2014, 136, 13781.
(a) K. Scholz and B. Thomas, Solid State Nucl. Magn. Reson., 1995,
S. V. Basenko, Russ. J. Gen. Chem., 2005, 75, 1927.
19 C. Viehweger, J. Kowalke, S. Schwarzer, C. Vogt, E. Kroke and
E. Brendler, New J. Chem., 2020, 44, 4613.
4
4
, 309; (b) C. Coelho, T. Azais, L. Bonhomme-Coury, J. Maquet, 20 L. Gmelin, Gmelin Handbuch der Anorganischen Chemie, Verlag
D. Massiot and C. Bonhomme, J. Magn. Reson., 2006, 179, 114.
Chemie, Weinheim, Silizium, volume C/15, 1958, p. 333.
This journal is ©The Royal Society of Chemistry 2020
Chem. Commun., 2020, 56, 13631--13633 | 13633