Mammeri, F., Le Bourhis, E., Rozes, L., and Sanchez, C. Journal of Materials Chemistry 15 (35-36), 3787 (2005).
Sanchez, C., Julian, B., Belleville, P., and Popall, M. Journal of Materials Chemistry 15 (35-36), 3559 (2005).
CHAPTER II
Sol-Gel Background.
The inorganic polymerization, from molecular precursors (usually metal alkoxides M(OR)n, M = Si, Ti, Zr,
Al, etc. and OR = OCnH2n+1) in organic solvents at low temperature associated with the versatility of processing
of colloidal states allows an easy elaboration of inorganic frameworks, generally metal oxo–polymers. 1,2
The inorganic polymerization proceeds by hydrolysis of the alkoxide precursors to introduce a reactive
hydroxy group on the metal. This step is followed by the formation of metal oxo-bridges or metal hydroxobridges by condensation or addition reactions respectively. The formation of hydroxobridges occurs when the
coordination number of the metal is higher than the valence state.
Hydrolysis:
Condensation:
M–OR + H2O → M–OH + R–OH
M–OH + XO–M → M–O–M + X–OH, X = H or R
formation of oxo-bridges by oxolation reactions
M–OH + XO–M → M–OH–M–OX, X = H or R
formation of hydroxo-bridges by olation reactions
Alkoxides are not miscible with water so that a common solvent, usually the parent alcohol ROH, is often
used. The oxo metallic network progressively grows from the solution, leading to the formation of oligomers,
oxopolymers, colloids (sols or gels), and a solid phase. These reactions can be described as SN2 nucleophilic
substitutions and the chemical reactivity of metal alkoxides toward hydrolysis and condensation depends mainly
on the electronegativity of the metal ion and the ability to increase its coordination number.1,2
Silicon has a low electrophilicity and remains four-coordinated in the monomeric Si(OR)4 alkoxide
precursors as well as in silica. Thus, silicon alkoxides are not very reactive. Hydrolysis–condensation reaction
rates must be increased using catalysts (acidic, basic or nucleophilic activation). On the other hand, non-silicate
metal alkoxides, including transition metals, lanthanides and aluminium are much more reactive than silicon
towards nucleophilic reactions. This high chemical reactivity is due to the lower electronegativity of the metal as
compared to silicon, and the metal atom ability to exhibit several coordination states.3 The coordination
expansion spontaneously occurs when the metal alkoxide reacts with water and direct addition of water to
transition metal alkoxide leads to uncontrolled precipitation of polydispersed oxide powders. An appropriate
choice of the alkoxide, especially the steric hindrance of the alkoxy groups, and of the solvent allows the control
of the reactivity towards hydrolysis–condensation reactions of M(OR)4 alkoxides. Indeed, the coordination
expansion of the metal occurs by solvation or alkoxy bridging leading to less reactive oligomeric species when
the steric hindrance effects are limited and when the solvent such as the parent alcohol can solvate the metal.3,4
Precipitation can be prevented by using inhibitors which leads the system toward the formation of other final
states including sols and gels.2,3,5 The most commonly used inhibitors are complexing ligands (β-diketones and
allied derivatives, polyhydroxyacid ligands such as polyols, and also α- or β-hydroxyacids or carboxylic acid),
which reduce the reactivity of the precursor by increasing its coordination number and decreasing the number of
easily hydrolysable groups. The strong complexing ligands (HL) react readily with transition metal as follows,
leading to new reactive precursors:
M(OR)n + xHL → M(OR)n-x(L)x + xROH.
Inhibition can also be performed by the use of inorganic acids, which inhibit the condensation reactions,
favour the reesterification of the hydrolysed precursors and enhance redissolution–precipitation processes.6 The
growth of transition metal oxo–polymers can be achieved to produce gels and sols by adjusting the hydrolysis
ratio and the inhibitor ratio. Increasing the hydrolysis ratio, or decreasing the inhibitor ratio, mainly results in an
increase in the size and in the fractal dimension of the oxo–polymers.7
The sol–gel route allows the powderless processing of glasses and ceramics. Moreover, the easy control of
the rheology of the colloid states and gels allows the formation of monoliths and also of fibres and films with
tuneable thickness produced directly from the solution by techniques like dip-coating, spin-coating and spindrawing.8
Mammeri, F., Le Bourhis, E., Rozes, L., and Sanchez, C. Journal of Materials Chemistry 15 (35-36), 3787 (2005).
Sanchez, C., Julian, B., Belleville, P., and Popall, M. Journal of Materials Chemistry 15 (35-36), 3559 (2005).
Sol-Gel Route and Hybrid Precursors: ORMOCER®S.
The strategy to construct class II hybrid materials consists of making intentionally strong bonds (covalent or
iono-covalent) between the organic and inorganic components. Organically modified metal alkoxides are hybrid
molecular precursors that can be used for this purpose.9 The chemistry of hybrid organic–inorganic networks is
mainly developed around silicon containing materials. Currently, the most common way to introduce an organic
group into an inorganic silica network is to use organo-alkoxysilane molecular precursors or oligomers of
general formula R’nSi(OR)4-n or (OR)4-nSi-R”-Si(OR)4-n with n = 1,2,3.
In most sol–gel conditions the Si-C bond remains stable towards hydrolysis and the R’ group introduces
focused new properties to the inorganic network (flexibility, hydrophobicity, refractive index modification,
optical response, etc.).
Organic groups R’ can be introduced into an inorganic network in two different ways: as network modifiers
or network formers. Both functions have been achieved in the so-called ORMOCER®s (registered trademark of
Fraunhöfer- Gesellschaft zur Förderung der angewandten Forschung e. V. in Germany). Since the eighties, they
have been extensively studied and developed by the Fraunhöfer Institut für Silicatforschung, Würzburg.9-11 The
organic group R’ can be any organofunctional group. If R’ is a simple nonhydrolyzable organic group (Si-CH3,
Si-phenyl, etc.) it will have a network modifying effect. Moreover if R’ can react with itself (R’ contains a vinyl,
a methacryl or an epoxy group for example) or additional polymerisable monomers, it acts as a network former.
Some examples of network formers and network modifiers were gathered by Sanchez et al.12. Polymeric
components can also be introduced in hybrid nanocomposites by using functionalised macromonomers (R”) of
general formula (OR)4-nSi-R”-Si(OR)4-n with n = 1,2,3. Some examples of these trialkoxysilyl functionalised
polymers are gathered in the following table.12
Table: Structure of some functionalized macromonomers used in sol-gel synthesis of hybrid materials: (a) Hydroxyl-terminated
polydimethylsiloxane, (b) Triethoxysilyl-terminated polydimethylsiloxane, (c) Hydroxyl-terminated phenylsilsesquioxane, (d)
Trimethoxysilyl-functionalized, (e) Triethoxysilyl-terminated poly(arylene ether) ketone, (f) Triethoxysilyl-terminated poly(tetramethylene
oxide), (g) Triethoxysilyl-terminated poly(oxipropylene), (h) Polyethyleneglycol, (i) Triethoxysilyl-terminated polyoxazoline, (j)
Ethoxysilyl-functionalized polyimide.
CH3
CH3
HO
Si
CH3
CH3
O
Si
CH3
OH
(EtO) 3Si
Si
CH3
n
O
CH3
(a)
CH2
A
Si(OEt) 3
HO
Si
O
n
Si
OH
C
CO
n
O
(b)
O
O
C
NH
C
O
O
C
O
(CH 2)3
O
C
C
(CH2)3
O
(CH2)4
O
(CH2)4
O
(CH2)3
NH
C
NH
(CH2)3
n
(e)
Si(OEt) 3
(EtO)3Si
Si(OMe) 3
(d)
(c)
HN
A
n
Si(OEt) 3
Si(OEt)3
n
(EtO) 3Si
(f)
H3C
(CH2)3
n
O Si(OEt) 3
HO
(g)
O
CH2 CH2 O
n
CH2OH
(h)
Si(OEt) 3
N
(EtO) 3Si
O
N
n
(i)
O
Si(OEt) 3
O
O
NH
HO
O
O
(j)
O
50
NH
NH
HO
OH
O
O
NH
O
OH
H3C Si CH3
OEt
50
Mammeri, F., Le Bourhis, E., Rozes, L., and Sanchez, C. Journal of Materials Chemistry 15 (35-36), 3787 (2005).
Sanchez, C., Julian, B., Belleville, P., and Popall, M. Journal of Materials Chemistry 15 (35-36), 3559 (2005).
The sol-gel synthesis of siloxane based hybrid organic-inorganic compounds usually involves di- or
trifunctional organosilanes co-condensed with metal alkoxides, mainly Si(OR)4, Ti(OR)4, Zr(OR)4 or Al(OR)3.
Each of these components has specific roles that have been reviewed and discussed extensively.13
Trifunctional alkoxysilanes and metal alkoxides are efficient crosslinkers that usually lead after cocondensation to hybrid materials having high Young’s modulus (E > 10 GPa) and either high glass transition
temperature (Tg > RT) or no Tg. Difunctional alkoxysilane co-condensed metal alkoxides generate linear
polymers and cyclic oligomers that behave as hybrid elastomers and exhibit low Young’s modulus (E = 5 - 100
MPa) and low glass transition temperature (Tg < RT). Therefore, mixing these different molecular precursors (triand di-functional alkoxysilanes, metal alkoxides, functional macromonomers, etc.) allow mechanical and
functional properties of hybrid materials to be tuned between those of polymers and glasses.
For example the microstructure of the resulting hybrids, ORMOCER®s, can be strongly modified by using
trifunctional silicon alkoxide R’Si(OR)3 in which R’ is a polymerisable organic group such as vinyl, epoxy or
methacrylate. Patternable coatings have been prepared from sols obtained by co-condensing TEOS, metal
alkoxides and organosilane precursors such as γ-glycidyloxypropyltrimethoxysilane, γ-methacryloxypropyl
trimethoxysilane or vinyl trimethoxysilane. Organic polymerisation is then induced by photochemical or thermal
curing. Patterning can be realized by photolithographic methods or even by direct laser writing.13 The synthesis
and processing of hybrid contact lenses provides a nice example of the versatile tailoring permitted by hybrid
materials chemistry. A ‘‘silicon’’-like unit, having a -Si-CH2-CH2- group was chosen to favour oxygen diffusion,
while a CH2OH group was considered to be suitable for hydrophobicity. However the combination of these
functional silanes in a sol–gel process leads to back esterification reactions of the –Si(CH2)nOH to the Si–R
group by destroying the hydrophilic –CH2OH groups. Therefore a -Si(CH2)n-epoxy group was chosen and the
ring opening was performed after hydrolysis and condensation of the alkoxy groups. The resulting hybrid
material exhibits good oxygen permeation and hydrophobicity but a porous texture and poor mechanical
properties. Ti(OR)4 was then added as a condensation catalyst leading to dense but brittle materials which were
not suitable for lenses. Consequently, -Si(CH2)n-O-CO-CH(CH3)=CH2 was introduced and copolymerized with
MMA (methyl methacrylate). Mechanical properties were sufficient but wettability was drastically decreased.
The wettability came back when MMA was substituted by HEMA (hydroxyl methyl methacrylate). The resulting
material exhibited all required properties but the unexpected part was that it showed a better scratch resistance.11
Another interesting discovery was that hybrids made from PDMS and
TEOS could be rubbery even when the inorganic component weight was in
excess of 70%. These rubbery hybrids materials14,15 exhibited properties
comparable to those of organic rubbers. The more rubbery hybrid was
synthesized with a much higher acid concentration (H+/Si = 0.3). The strong
acidity of the medium increases the hydrolysis rates of Si-OR species
rapidly providing silanol groups but slows down condensation reactions
providing small silica polymers and it also modifies the distribution of the
species contained in the commercial PDMS polymers. The removal by
evaporation of cyclic species may lead to an increase of the matrix porosity.
The resulting material consist of a matrix of medium chain length and small
silica particles with a porous structure. This porous texture produces free
volume needed for chain motion. As a consequence, the PDMS chains can
curl and uncurl in the presence of external stress and the material exhibits
rubbery elasticity. Furthermore, these rubbery hybrids can be specifically
shaped, giving objects such as those illustrated in the accompanying figure
(Courtesy of E. Bescher and J. D. Mackenzie) 15,16 Extensive investigations
have been also performed about organically derived silica-based materials
obtained through surface grafting reactions for other key applications such as gas chromatography17 and
membranes.18
References:
1
2
3
4
5
6
Brinker, C. J. and Scherer, G. W., Sol-Gel Science, The Physics and Chemistry of Sol-Gel Processing. (Academic Press,
San diego, 1990).
Livage, J., Henry, M., and Sanchez, C. Progress in Solid State Chemistry 18 (4), 259 (1988).
Sanchez, C. and Ribot, F. New Journal of Chemistry 18 (10), 1007 (1994).
Nabavi, M., Doeuff, S., Sanchez, C., and Livage, J. Journal of Non-Crystalline Solids 121 (1-3), 31 (1990).
Sanchez, C., Livage, J., Henry, M., and Babonneau, F. Journal of Non-Crystalline Solids 100 (1-3), 65 (1988).
Kallala, M., Sanchez, C., and Cabane, B. Physical Review E 48 (5), 3692 (1993).
Mammeri, F., Le Bourhis, E., Rozes, L., and Sanchez, C. Journal of Materials Chemistry 15 (35-36), 3787 (2005).
Sanchez, C., Julian, B., Belleville, P., and Popall, M. Journal of Materials Chemistry 15 (35-36), 3559 (2005).
7
8
9
10
11
12
13
14
15
16
17
18
Blanchard, J., Ribot, F., Sanchez, C., Bellot, P. V., and Trokiner, A. Journal of Non-Crystalline Solids 265 (1-2), 83
(2000).
Klein, L. C., Sol-Gel Optics, Processing and Applications. (Kluwer Publications, Berlin, 1994).
Schmidt, H. and Seiferling, B. Materials Research Society Symposium Proceedings 73, 739 (1986).
Schmidt, H., Kaiser, A., Patzelt, H., and Scholze, H. Journal de Physique 12 (C9), 275 (1982); Schmidt, H., Scholze, H.,
and Kaiser, A. Journal of Non-Crystalline Solids 63 (1-2), 1 (1984); Schmidt, H., in Ultrastructure Processing of
Advanced Materials: Ceramics, Glasses and Composites, edited by D.R. Ulhmann and D.R. Ulrich (John Wiley, New
York, 1992); Schmidt, H. Materials Research Society Symposium Proceedings 180, 961 (1990); Schmidt, H. Journal of
Non-Crystalline Solids 73 (1-3), 681 (1985).
Philipp, G. and Schmidt, H. Journal of Non-Crystalline Solids 63 (1-2), 283 (1984).
Sanchez, C., Julian, B., Belleville, P., and Popall, M. Journal of Materials Chemistry 15 (35-36), 3559 (2005).
Schmidt, H. and Popall, M., in Sol-Gel Optics I, edited by J.D. Mackenzie and D.R. Ulrich (Proc. SPIE 1328,
Washington, 1990), pp. 249.
Hu, Y. and Mackenzie, J. D. Materials Research Society Symposium Proceedings 271, 681 (1992); Chung, Y. J., Ting,
S., and Mackenzie, J. D., presented at the Better Ceramics Through Chemistry IV, Pittsburg, 1990 (unpublished).
Mackenzie, J. D., Chung, Y. J., and Hu, Y. Journal of Non-Crystalline Solids 147, 271 (1992).
Bescher, E., Pique, F., Stulik, D., and Mackenzie, J. D. Journal of Sol-Gel Science and Technology 19 (1-3), 215 (2000).
Balard, H., Papirer, E., Khalfi, A., and Barthel, H. Composite Interfaces 6 (1), 19 (1999); Barthel, H., Rosch, L., Weis,
J., Khalfi, A., Balard, H., and Papirer, E. Composite Interfaces 6 (1), 27 (1999); Hommel, H., Legrand, A. P., Benouada,
H., Bouchriha, H., Balard, H., and Papirer, E. Polymer 33 (1), 181 (1992).
Galvan, J. C., Aranda, P., Amarilla, J. M., Casal, B., and Ruiz-Hitzky, E. Journal of Materials Chemistry 3 (6), 687
(1993).