The van der Waals induced supramolecular organization of hydrophobic tetrahedral units in the course of hydrolytic polycondensation

Article abstract of DOI:10.1039/b512450f

This paper describes the hydrolytic polymerization of a tetrahedral precursor in the four directions. The hydrolytic polycondensation occurring at silicon led to highly polycondensed solids (around 100%). The solids obtained exhibited an organization at the nanometric scale evidenced by X-ray powder diffraction and another at the micrometric scale (birefringence experiments). The optical axis determination which corresponds to the orientation of the cross-linked colloids exhibits an inverse micelle type geometry. Thus the self-organization of the tetrahedral units is controlled by the poor hydrophilicity of the medium. the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006.

Full text of DOI:10.1039/b512450f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
The van der Waals induced supramolecular organization of hydrophobic
tetrahedral units in the course of hydrolytic polycondensation
´ ´
Frederic Lerouge, Genevieve Cerveau and Robert J. P. Corriu*
Received 5th September 2005, Accepted 22nd November 2005
First published as an Advance Article on the web 7th December 2005
DOI: 10.1039/b512450f
This paper describes the hydrolytic polymerization of a tetrahedral precursor in the four
directions. The hydrolytic polycondensation occurring at silicon led to highly polycondensed
solids (around 100%). The solids obtained exhibited an organization at the nanometric scale
evidenced by X-ray powder diffraction and another at the micrometric scale (birefringence
experiments). The optical axis determination which corresponds to the orientation of the cross-
linked colloids exhibits an inverse micelle type geometry. Thus the self-organization of the
tetrahedral units is controlled by the poor hydrophilicity of the medium.
phenyl]germane and tetrakis[4-tris(isopropoxy)silylphenyl]-
Introduction
stannane [Scheme 2(a)].
The hydrolytic polycondensation at silicon (sol-gel process1)
of hybrid organic–inorganic monophasic materials leads to
short range organized solids (Scheme 1).
Both precursors present twelve directions for solid forma-
tion, oriented regularly on the four directions of the tetra-
hedron. Their sol-gel hydrolysis–polycondensation leads to
hybrid solids organized from the nanometric scale (X-ray
diffraction) to the micrometric (birefringence) and even the
millimetric one.9 This organization occurs during the forma-
tion of the Si–O–Si bonds which transforms intermolecular
interactions into intramolecular ones corresponding to the
weak van der Waals interactions existing between the organic
units. These results are noteworthy since it has been recently
described that rigid molecules containing Ar4C and Ar4Sn as a
core, explored as possible mesophases, are not mesogenic in
spite of the presence of eight lipophilic long chains [Scheme
2(b) and (c)].10
This organization, induced by supramolecular2,3 van der
Waals type interactions between hydrophobic units occurring
during polycondensation at silicon, represents an attractive
field of investigations. A large variety of precursors containing
organic units has been investigated, leading to non-crystalline
materials in which the organic and inorganic parts are cova-
lently bound maintaining their physical and chemical proper-
ties.4–6 van der Waals-induced supramolecular organization of
these solids is currently under study in our group. In contrast
to silica, which is always amorphous when prepared by this
route, self-organization in nanostructured hybrid organic–
inorganic materials has been evidenced in many examples.7
Organic units included in the framework, whatever their
geometry, exhibit an organization at the nanometric scale, as
evidenced by X-ray powder diffraction. They exhibit also an
organization at the micrometric scale as shown by birefrin-
gence experiments in cross-polarized light microscopy.7 These
results have been observed in the case of rigid-rod linear,
twisted, planar and mesogenic organic spacers.7 Only flexible
alkyl organic groups lead to non-birefringent solids.8 More-
over it has been shown that the hydrophilic conditions are an
important factor.
In this paper we describe the case of the formation of
polysiloxane polymers obtained from monomers presenting a
tetrahedral geometry. The precursors investigated present a
germanium or tin tetrahedral central atom and only the
four directions of monofunctional silicon atoms for solid
formation: tetrakis(4-dimethylisopropoxysilylphenyl)germane
1 and tetrakis(4-dimethylisopropoxysilylphenyl)stannane 2
(Scheme 3).
Furthermore the precursors 1 and 2 were selected because
they present the same number of directions and the same
tetrahedral geometry as Si(OR)4 which never exhibits any
organization when prepared in mild conditions. Indeed,
crystals are obtained only after thermal treatment (thermo-
dynamic control). The main difference is the complete
hydrophilicity of Si(OR)4 in the hydrolytic polycondensation.
In contrast 1 and 2 exhibit a hydrophobic central bulky
organic group. Thus in the hydrolytic conditions used for
In order to determine the limits of the organization of such
systems, we are investigating new precursors with different
geometry and connectivity at silicon. We have focused our
interest in the case of tetrahedral geometry, expected not to be
very favorable to self-organization. In a recent paper9 we
described the surprising results obtained from two precursors
presenting such a geometry, tetrakis[4-tris(isopropoxy)silyl-
´
Laboratoire de Chimie Moleculaire et Organisation du Solide, UMR
5637 CNRS, Universite Montpellier II, Sciences et Techniques du
´
Languedoc, Place E. Bataillon, F-34095 Montpellier Cedex 5, France.
E-mail: corriu@univ-montp2.fr
Scheme 1 Sol-gel process of nanostructured hybrid organic–inor-
ganic materials.
ꢀc
272 | New J. Chem., 2006, 30, 272–276 This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006

View Article Online
Scheme 2
polycondensation at silicon, the two precursors behave
differently.
tion diagrams of the solids (Fig. 1) exhibited three broad
signals and the shape of the curves was similar whatever the
precursor and the catalyst employed.
The signals were centred at 1.61 AÀ1 (3.9 A), 1.06 AÀ1 (5.9
A) and 0.49 AÀ1 (12.8 A) in the case of germanium, and were
shifted to slightly longer distances for the tin compound, at
1.49 AÀ1 (4.2 A), 1.01 AÀ1 (6.2 A) and 0.48 AÀ1 (12.9 A). The
two signals at 3.9 and 4.2 A were attributed to the contribution
of the Si–O–Si units.12 The signals at 5.9 and 6.2 A could
correspond to the distance between planes of central metallic
(germanium or tin) on one hand and silicon atoms on the
other hand. The values obtained by molecular modelling,
using Chem 3D pro (MM2 method), were, respectively 6.7 A
and 6.9 A. The broad signals at 12.8–12.9 A might correspond
to an average organization between the tetrahedrons. How-
ever it is difficult to make further comparisons with distances
deduced by molecular modelling since we do not know how
the tetrahedrons are arranged one to the other. Nevertheless
the presence of these X-ray diffraction signals means that there
is an organization at the nanometric scale between the tetra-
hedrons.13 The irreversible formation of the Si–O–Si bonds
induces the auto-organization at the nanometric scale by
forcing the molecules to be bonded to each other: the poly-
condensation at silicon changes irreversibly the intermolecular
interactions into intramolecular ones. This process eliminates
the main part of the entropy and permits us to understand why
very weak van der Waals interactions of tetrahedral units are
sufficient for inducing the organization of the solid.
Results and discussion
1 and 2 were prepared by reaction of the Grignard reagent of
4-(bromophenyl)dimethylisopropoxysilane11 with germanium
and tin tetrachloride, respectively. They were isolated as oils in
good yields. The hydrolysis was performed in THF (concen-
tration 0.5 M) with a stoichiometric amount of water (2 eq.) in
the presence of a nucleophilic (TBAF, tetrabutylammonium
fluoride) or acid (HCl) catalyst (5 mol%) (Scheme 3). Opaque
monolithic gels formed after a few hours (Table 1). The gels
were aged for 6 days, then treated as usual to give insoluble
non-porous powders, (SBET o 10 m2 gÀ1).
These solids presented similar 29Si CP MAS NMR spectra:
only one thin signal at d B À0.9 ppm was observed and
attributed to C–Si(Me)2–O– units. This was indicative of
almost total polycondensation since no signal appeared at
d B 5 ppm (precursor) whatever the catalyst and the pre-
cursor, taking into account the accuracy of NMR.
As for all nanostructured hybrid materials reported until
now,12 the X-ray diffraction diagrams did not exhibit any
sharp Bragg signal. However broad and intense signals were
observed. It is well established that such broad signals corre-
spond unambiguously to the existence of a nanometric scale
order in the solid.12 This point is very important since it means
that this tetrahedral precursor is able to induce self-organiza-
tion at the nanometric scale. As a first approximation, assum-
ing Bragg’s law a priori, the distances associated with the q
values were evaluated (d = 2p/q). The X-ray powder diffrac-
For the birefringence measurements a fraction of the homo-
geneous solution, containing the precursor, THF, the catalyst
and H2O, was introduced by capillarity, a few minutes before
sol-gel transition, in a thin Teflon-coated glass cell, which was
then sealed. After gel formation (2 hours), the cell was
analyzed by microscopy in polarized light. All the solids
Table 1 Some data for solids 1A, 1B, 2A, 2B
Gel time
(hour)
29Si NMR
d (ppm)
Birefringence
Dn (Â103)
Solid
Catalyst
1A
1B
2A
2B
TBAF
HCl
TBAF
HCl
9
3
13
3.5
À0.91
À0.93
À0.89
À0.90
4.9
5.1
6.0
6.1
Scheme
1 and 2.
3 Hydrolysis–polycondensation reaction of precursors
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 272–276 | 273

View Article Online
Fig. 1 X-Ray diffraction diagrams of solids 1A (plain), 1B (dotted), 2A (plain) and 2B (dotted).
presented birefringence properties after 24 hours. The higher
values were found for the tin compounds 2A and 2B. The
birefringence intensities are given in Table 1. They correspond
to an average of 10 measurements recorded at different loca-
tions of the cell. The morphologies of the solids in the cells
were very similar whatever the precursor and the type of
catalyst. Many circular nodules of various size were observed.
These isotropic zones corresponded to bubbles formed during
the polycondensation. They were empty of material since only
Teflon coating is observed in their center, as confirmed by
Raman experiments.14 The other parts of the solid around the
bubbles were anisotropic, the higher birefringence values being
observed in the near environment of the bubbles (Fig. 2).
The orientation of the optical axis was determined using a
Berek compensator (Fig. 2). It corresponds to the orientation of
anisotropy at the micrometric scale. Interestingly we observed
the formation of empty bubbles instead of the cracks that are
usually formed under equivalent conditions with rigid-rod line-
ar, planar and even tetrahedral spacers.7,9 We have previously
evidenced that reorganization within hybrid solids occurs during
the ageing step, that is when densification of the Si–O–Si bridges
and solvent evaporation generate stress and shrinkage, forming
empty spaces in the cell as shown in Scheme 4(a).
In other words, the organization of aggregates of matter at
the micrometric scale occurs with an anisotropic orientation
around the empty bubbles in the case of polycondensation of
precursors 1 and 2 containing monodimensional Me2Si(OR)
groups. This situation is completely different from the orienta-
tion observed in the case of the hydrolysis–polycondensation
of tetrahedral precursors containing tridimensional Si(OR)3
units (Scheme 2a).9 In this case, regular parallel rectangular
chunks are always observed and the orientation of the optical
axis was always found perpendicular to the edges of the
chunks (Scheme 4b). This difference of birefringence figures
and the orientation of the optical axes towards the centre of
the bubbles suggested the geometry of an inverse micelle: the
cross-linked hydrophobic polymer is oriented towards the
hydrophilic centre.
Fig. 2 Birefringence pictures of: (a) 1A, (b) 1B, (c) 2A, (d) 2B and schematic representation of a gel picture.
ꢀc
274 | New J. Chem., 2006, 30, 272–276 This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006

View Article Online
Scheme 4 (a) Schematic representation of stress and shrinkage during ageing step, and orientation of optical axis in a chunk of solid 1A, 1B, 2A or
2B; (b) cracks and orientation of optical axis in a chunk of solid obtained from tetrahedral precursor with Si(OR)3 groups cf. Scheme 2(a).
In fact the work-up always employed involves the use of a
solution corresponding to the stoichiometric amount of H2O
necessary for the hydrolysis–polycondensation. In the present
case, the amount of H2O is 1 mole of H2O for 13.4 moles of
hydrophobic species (precursor and THF) whereas in the
general cases and particularly in the case of the tetrahedral
precursors shown in Scheme 2(a), this molar ratio was B1/4.
The polycondensation goes on with the elimination of hydro-
philic isopropanol molecules in a geometry similar to that of
an inverse micelle. It is possible to explain this organization
considering that the polycondensation begins to induce the
organization of the oligomers at the nanometric scale by van
der Waals interactions. The growth of the oligomers is then
controlled by the structure of the inverse micelle existing in the
medium owing to the presence of a low amount of hydrophilic
entities (H2O, isopropanol).
recorded on a Micromeritics Gemini III 2375 apparatus. The
specific surface area was determined using the BET equation.
The X-ray experiments were performed on powders of solids
in a Lindeman tube with an imaging plate two-dimensional
detector (Marresearch 2D ‘‘Image-Plate’’) with a rotating
anode apparatus (Rigaku RU 200). The radiation used was
Cu Ka (l = 1.5418 A). Optical properties of the materials
were observed with a Laborlux12POLS polarizing microscope.
Photographs were taken using a Leica wild MPS28 camera.
The birefringence Dn of the gels was obtained from the
expression Dl = (Dn)d, where Dl is the optical path difference
and d is the cell thickness which is evaluated by UV-Vis
spectroscopy (B15 mm). Dl was measured by a Berek com-
pensator. Elemental analyses were carried out by the ‘‘Service
Central de Micro-Analyse du CNRS’’.
Syntheses
These observations show that it will certainly be possible to
obtain different orientations by varying the amount of H2O
employed. Work is in progress in this direction.
Tetrakis(4-dimethylisopropoxysilylphenyl)germane 1. The
Grignard reagent of 4-bromophenyldimethylisopropoxysilane
was prepared in THF from 7.3 g (26.72 mmol) bromo com-
pound and 0.78 g (32.10 mmol) of magnesium according to a
literature procedure.11 The greenish mixture was stirred during
3 hours at room temperature, then 0.71 mL (6.07 mmol) of
tetrachlorogermane were added. After 12 additional hours of
stirring, THF was evaporated and the solid residue was taken
up in hexane. After filtration it was evaporated under vacuum
to give 1 as a translucent thick oil: 3.85 g (4.55 mmol), yield
75%. 1H NMR (CDCl3) d: 0.42 (s, 24H), 1.17 (d, J = 6.0 Hz,
24H), 4.02 (hept, J = 6.0 Hz, 4H), 7.53 (dd, J = 7.6 Hz, 16H);
13C NMR (CDCl3) d: À0.7 (CH3), 26.1 (CH3), 65.9 (CH),
124.7 (Ar), 131.4 (Ar), 135.5 (Ar), 137.8 (Ar); 29Si NMR
(CDCl3) d: 4.94; elemental analysis: calcd for C44H68O4GeSi4:
C 62.47, H 8.10, Ge 8.58, Si 13.28. Found: C 62.14, H 8.06, Ge
8.21, Si 13.42%.
Conclusions
In conclusion tetrahedral precursors led to the formation of
long range ordered solids. This particular example shows the
degree of generality of the auto-organization phenomenon
since systems presenting four directions for polycondensation
regularly distributed in space led to highly polycondensed
ordered solids. The results reported here clearly illustrate the
fact that the micrometric organization takes place during
Oswald ageing and is governed by completely different para-
meters than the nanometric one.
Experimental
All reactions were carried out under argon using a vacuum line
and Schlenk techniques. Solvents were dried and distilled just
before use. The 1H and 13C NMR spectra were recorded on a
Bruker DPX-200 spectrometer and the 29Si NMR spectra were
recorded on a Bruker WP-200 SY spectrometer. The 29Si CP
MAS NMR spectra were recorded on a Bruker Avance 300
spectrometer operating at 59.6 MHz using a recycling delay of
10 s and a contact time of 5 ms. The spinning rate was 5 kHz.
Chemical shifts are given relative to tetramethylsilane. The
nitrogen adsorption–desorption isotherms at 77.35 K were
Tetrakis(4-dimethylisopropoxysilylphenyl)stannane 2. 2 was
prepared similarly from 10.3 g (37.69 mmol) of 4-bromophe-
nyldimethylisopropoxysilane, 1.10 g (45.23 mmol) of magne-
sium and 1.00 mL of tin tetrachloride. 2 was obtained as a
translucent thick oil: 5.37 g (6.02 mmol), yield 72%. 1H NMR
(CDCl3) d: 0.45 (s, 24H), 1.20 (d, J = 6.0 Hz, 24H), 4.08 (hept,
J = 6.0 Hz, 4H), 7.68 (dd, J = 7.6 Hz, 16H); 13C NMR
(CDCl3) d: À0.7 (CH3), 26.2 (CH3), 65.8 (CH), 124.7 (Ar),
133.9 (Ar), 137.1 (Ar), 139.8 (Ar); 29Si NMR (CDCl3) d: 5.18;
elemental analysis: calcd for C44H68O4Si4Sn: C 59.24, H 7.68,
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 272–276 | 275

View Article Online
Si 12.59, Sn 13.30. Found: C 59.12, H 7.96, Si 12.14, Sn
12.94%.
Acknowledgements
The authors thank Dr Philippe Dieudonne
´
(GDPC-UMII,
Montpellier, France) for X-ray measurements and Dr Jean-
Francois Bardeau (Laboratoire de Physique de l’Etat Con-
Preparation of the solids 1A, 1B, 2A and 2B. The preparation
of the gels was carried out according to the following general
procedure. The preparation of 1A is given as an example: in a
Schlenk tube, 946 mL of a solution containing 47 mL (47 mmol)
of TBAF (1 mol LÀ1 in THF), 34 mL (1.89 mmol) of H2O and
865 mL of dried THF, were added to a solution of 0.800 g
(0.946 mmol) of 1 in 946 mL of dried THF. After homogeniza-
tion the mixture was kept in the Schlenk tube. A few minutes
before the sol-gel transition, a part was introduced by capil-
larity into a Teflon-coated cell for birefringence measure-
ments. The gel obtained after 9 h in the Schlenk tube was
aged for 6 days, then crushed and washed twice with acetone,
ethanol and diethyl ether, and the resulting solid was dried at
120 1C in vacuo for 3 hours yielding a white powder.
dense, Universite du Maine, Le Mans, France) for Raman
´ ´
experiments.
References
1 C. J. Brinker and G. W. Scherer, Sol-Gel Science: The Physics and
Chemistry of Sol-Gel Processing, Academic Press, San Diego, 1990.
2 J. M. Lehn, Supramolecular Chemistry, VCH, Weinheim, 1st edn,
1995.
3 M. W. Hosseini, Coord. Chem. Rev., 2003, 240, 157.
4 D. A. Loy and K. J. Shea, Chem. Rev., 1995, 95, 1431.
5 R. J. P. Corriu, Angew. Chem., Int. Ed., 2000, 39, 1376, and
references cited.
6 G. J. de Soler-Illia, C. Sanchez, B. Lebeau and J. Patarin, Chem.
Rev., 2002, 102, 4093.
7 B. Boury and R. J. P. Corriu, Chem. Rec., 2003, 3, 120, and
references cited.
8 B. Boury, F. Ben, R. J. P. Corriu, P. Delord and M. Nobili, Chem.
Mater., 2002, 14, 730.
9 F. Lerouge, G. Cerveau and R. J. P. Corriu, J. Mater. Chem., 2006,
DOI: 10.1039/b509017b.
10 A. Pegenau, T. Hegmann, C. Tschierske and S. Diele, Chem.–Eur.
J., 1999, 5, 1643.
11 J.-P. Bezombes, C. Chuit, R. J. P. Corriu and C. Reye, J. Mater.
Chem., 1998, 8, 1749.
1A: 0.587 g (0.915 mmol); yield 97%. 29Si CP MAS NMR
(d ppm): À0.91. SBET o 10 m2 gÀ1
.
1B: 0.800 g (0.946 mmol) of 1 in 1.892 mL of THF contain-
ing 47 mL (47 mmol) of HCl and 34 mL (1.89 mmol) of H2O
gave a gel after 3 h. 1B: 0.565 g (0.880 mmol); yield 93%. 29Si
CP MAS NMR (d ppm): À0.93. SBET o 10 m2 gÀ1
.
2A: 0.500 g (0.560 mmol) of 2 in 1.120 mL of THF contain-
ing 28 mL (28 mmol) of TBAF and 20 mL (1.12 mmol) of H2O
gave a gel after 13 h. 2A: 0.378 g (0.550 mmol); yield 98%. 29Si
12 B. Boury and R. J. P. Corriu, in Supplement Si: The Chemistry of
Organic Silicon Compounds, ed. Z. Rappoport and Y. Apeloig,
Wiley & Sons, Chichester, 2001, ch. 10, p. 565, and references cited.
13 B. Boury, R. J. P. Corriu, P. Delord and V. Le Strat, J. Non-Cryst.
Solids, 2000, 265, 41.
CP MAS NMR (d ppm): À0.89. SBET o 10 m2 gÀ1
.
2B: 0.500 g (0.560 mmol) of 2 in 1.120 mL of THF contain-
ing 28 mL (28 mmol) of HCl and 20 mL (1.12 mmol) of H2O
gave a gel after 3.5 h. 2B: 0.365 g (0.531 mmol); yield 95%. 29Si
14 Raman experiments were performed by Dr Jean-Francois Bardeau
from University of Maine, Le Mans.
CP MAS NMR (d ppm): À0.90. SBET o 10 m2 gÀ1
.
ꢀc
276 | New J. Chem., 2006, 30, 272–276 This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006