J O U R N A L O F C H E M I S T R Y Materials Influence of the nature of the organic precursor on the textural and chemical properties of silsesquioxane materials Genevie`ve Cerveau, Robert J. P. Corriu,* Ce�dric Lepeytre and P. Hubert Mutin Laboratoire de Chimie Mole�culaire et Organisation du Solide, UMR 5637, Universite� Montpellier II, Case Courrier 007–34095 Montpellier Ce�dex 5, France Received 24th July 1998, Accepted 8th September 1998 The hydrolytic sol–gel polymerization of molecular organosilicon precursors with a rigid geometry C6H4[Si(OMe)3]2-1,4 and C6H3[Si(OMe)3]3-1,3,5 2 was investigated and compared to the results obtained with precursors having a more flexible structure C6H4RR¾-1,4 [R=R¾=CH2Si(OMe)3 3; R=R¾=CH2CH2Si(OMe)3 4].Compounds 1–4 have been studied in the same conditions.They were hydrolyzed under nucleophilic catalysis (TBAF: tetrabutylammonium fluoride) in MeOH and in THF. The structure of the organic group was found to be a determining parameter for both the physical and chemical properties of the resulting silsesquioxane materials. The molecular precursors 1 and 2 containing a ‘rigid’ organic group led to hydrophilic solids with similar degrees of condensation.In all cases, high specific surface area (370–1018 m2 g-1) and poor chemical reactivity towards Cr(CO)6 (11–33%) were observed. By contrast, the precursors containing a ‘flexible’ organic group (3 and 4) led to hydrophobic solids; the texture, the degree of condensation and the reactivity towards Cr(CO)6 of these solids strongly depended on the solvent.For instance the solids prepared in MeOH had no significant specific surface area. The solids derived from the most flexible precursor (4) exhibited the highest chemical reactivity. The short range organization of the solid is a function of the geometry of the precursor and the experimental conditions. tural and chemical properties of diVerent aryl bridged Introduction silsesquioxane xerogels.Organic–inorganic hybrid materials are a wide field of growing interest.1,2 Sol–gel chemistry, which corresponds to an inor- Experimental ganic polymerization, oVers an access to a wide variety of silica-like structures.3 The mild reaction conditions allow the All the syntheses of precursors and reactions of complexation incorporation of organic moieties into inorganic oxide net- with Cr(CO)6 were carried out under argon using a vacuum works.The preparation of monocomponent hybrid materials line and Schlenk tube techniques.15 Solvents were dried and in which organic molecules are covalently bound to silica is distilled before use. IR spectra were recorded using a Perkin opening interesting perspectives for chemists.The incorpor- Elmer 1600 FTIR spectrophotometer using KBr pellets or by ation of an organic unit in the core of an inorganic matrix the DRIFT method. Solid state NMR spectra were obtained with a Bruker FTAM 300 spectrometer: 13C CP MAS NMR can be achieved when at least two covalent bonds are formed at 75.47 MHz, recycling delay 5 s, and contact time 5 ms; 29Si between the organic molecule and the solid.A large variety of CP MAS NMR at 59.62 MHz, recycling delay 10 s, and nanostructured hybrid materials has been reported4–13 and it contact time 2 ms. Chemical shifts are given relative to tetra- has been shown that changing the nature of the organic group methylsilane. To obtain quantitatively reliable 29Si data, single- induces changes in the macroscopic properties of the hybrid pulse MAS NMR experiments (SPE-MAS) have been per- solid.For instance, linear rigid rod structures with p-phenylene formed on a Bruker ASX 200 spectrometer at 39.74 MHz, groups4,5 and flexible structures due to methylene groups6,7 using a pulse angle of 30°, a recycling delay of 60 s and high- influence properties such as the microporosity. Microporous power proton decoupling during the acquisition.These experi- bridged polysilsesquioxanes have been used as a confinement ments have been done for two xerogels derived from precursors matrix for nanosized particles.9 Short range organization in 1 and 4 by hydrolysis in MeOH. The spinning rate was the amorphous solids arising from the molecular structure of 5000 Hz in all cases. the precursor has been detected using the chemical reactivity Specific surface areas, pore volumes and pore size of organic spacers in the case of hybrid materials containing distribution were determined using a Micromeritics Gemini buta-1,3-diyne or thiophene bridging fragments.10,11 III 2375 apparatus. Elemental analyses were carried out by Moreover, the inclusion of charge transfer complexes in gels11a the ‘Service Central de Micro-Analyse du CNRS’.Oxygen illustrates the importance of weak interactions between the percentages were deduced by diVerence. X-Ray powder organic units on the texture of the solid. In a preliminary diVraction measurements were performed using a Seifert MZ4 report, we have shown that the nature of the organic spacer apparatus. appears as a very important parameter in the control of the Compounds 1, 2 and 3 were prepared according to literature properties of the solid14a such as the specific surface area, the procedures4a,12b while 4 was commercial and purified before hydrophilicity and the chemical reactivity.use. In this context, we were interested in investigating to which Gels starting from 1,2 and 3 have been already described extent can the structure of the precursor be a determining under diVerent experimental conditions.4a,5a,13 parameter for the solid state properties of the resulting materials.We therefore examined the gel formation from precursors Preparation of silsesquioxane gels with diVerent structures, under the same experimental conditions. The preparation of the gels was carried out according to the We report here our studies concerning the relationships following general procedure.The preparation of 1M is given as an example. To 2.52 g (7.92×10-3 mol) of 1 in 10 ml of between the structure of the organic groups and the tex- J. Mater. Chem., 1998, 8, 2707–2713 2707Table 1 Experimental conditions and gelation time of xerogels Precursor conc./ Gelation Entry Precursor Solvent Xerogel mol L-1 time/min 1 1 MeOH 1M 0.5 160 2 1 THF 1T 0.5 135 3 2 MeOH 2M 0.5 7 4 2 THF 2T 0.5 4 5 3 MeOH 3M 0.5 45 6 3 THF 3T 0.5 <1 7 4 MeOH 4M 0.5 720 8 4 THF 4T 0.5 <5 methanol was added a solution of 79×10-3 ml of TBAF (solution 1 mol l-1 in THF) and 428×10-3 ml of water (3 mol. equiv.) in 5.33 ml of methanol.After 160 min a monolithic opaque gel formed. After ageing during 5 days at room temperature the solid was collected, then ground and washed with ethanol, acetone and diethyl ether.The resulting solid was dried at 120 °C in vacuo during 3 h yielding 1.23 g of a Scheme 1 Molecular precursors 1–4. white powder 1M. The experimental conditions and the gel times of the xerogels are reported in Table 1.These analyses revealed an excess of carbon and hydrogen indicating the presence of residual hydroxy and methoxy Reaction of complexation with Cr(CO)6 groups. This excess of carbon and hydogen was larger when the reaction was carried out in MeOH (Table 2, entries 1, 3) The reaction of complexation of the xerogel was carried out according to the following general procedure.14a than in THF (Table 2, entries 2, 4) as shown by the experimental formula.This can be attributed to a higher degree of The complexation of 1M is given as an example. The xerogel 1M (0.58 g, 3.22×10-3 mol ) and an excess of Cr(CO)6 (1 g, hydrolysis in THF than in MeOH. The X-ray powder diVraction pattern of the samples showed 4.8×10-3 mol) were introduced in a mixture of 10 ml of THF and 40 ml of Bun2O.The mixture was refluxed during 65 h. absence of crystallinity in all cases. After this, the yellow–green solid was filtered, washed with dichloromethane and diethyl ether, then dried in vacuo at NMR characterization of the xerogels room temperature during 3 h. A pale green solid (0.72 g) was The IR and NMR characteristics of the xerogels prepared obtaih was analysed by IR and NMR spectroscopy.from 1–4 have been determined and are in agreement with the The degree of complexation was determined by chemical and conservation of the organic units bonded to the silica matrix: EDS elemental analysis. The experimental ratio (Cr/Si)exp was 29Si CP MAS NMR (Table 2) clearly established that the Si-C determined from the elemental analysis and compared to the bond was retained within the gel in all cases:17 no 29Si theoritical value (Cr/Si)th assuming a degree of complexation resonances attributable to SiO4/2 units were detected 18 (only of 100%.For xerogel 1M, (Cr/Si)exp=0.09; degree of com- T0, T1, T2, T3 units were observed). Futhermore 13C CPMAS plexation=18%. The results obtained were confirmed by NMR spectra showed that the organic fragments were not elemental EDS analysis. modified.When the hydrolysis–polycondensation reaction was Results and discussion performed in methanol, in the case of gels prepared from 1 and 2, 29Si CP MAS NMR showed a major substructure T2: The molecular precursors 1–4 containing phenylene units with CSiO2(OX) (X=Me or H). Starting from 3 and 4 a major various geometries have been investigated4a,12b (Scheme 1).substructure T2 and signals corresponding to a substructure Compounds 1 and 2 have a ‘rigid’ spacer, whereas 3 and 4 of type T0: CSi(OX)3 (X=Me or H) were observed (Fig. 1). contain more ‘flexible’ organic groups. When the solvent was THF, the 29Si CP MAS NMR characteristics were very similar to those obtained in methanol Preparation and characterization of silsesquioxane gels in the case of 1T and 2T.However the solid 3T appeared more polycondensed (T2 and T3 major substructures). Solid The sol–gel polymerization of monomeric precursors 1–4 was performed in MeOH and in THF, in the presence of (TBAF) 4T showed a major substructure T3: CSiO3 indicative of a higher degree of polycondensation (Fig. 2).tetrabutylammonium fluoride (1 mol%) as a catalyst,16 for a concentration of 0.5 mol l-1 of the precursor at room tempera- Quantitatively reliable 29Si SPE MAS NMR spectra have been collected for two typical silsesquioxane xerogels, 1M and ture (reaction 1). The experimental procedure was strictly controlled, all the reactions were performed three times and 4M, and compared to the CP MAS NMR spectra of the same precursors (Fig. 3). The percentages of the diVerent Tx units, were rigorously reproducible. When the reaction was performed in MeOH, opaque gels obtained by deconvolution of the spectra are reported in Table 3. The SPE MAS and CP MAS percentages found for formed within a short period of time for 1 and 2 whereas white precipitates were obtained for 3 and 4.When the solvent the diVerent Tx units were comparable in both cases. Accordingly, the degrees of condensation derived from the was THF, transparent gels formed in all the cases. Gel times are reported in Table 1. The gels were allowed to stand at 29Si CP MAS spectra were close to those derived from the quantitative SPE MAS NMR spectra. It was not possible to room temperature for 5 days.After washing with ethanol, acetone and ether, the powders were dried in vacuo at 120 °C perform quantitative SPE MAS experiments for all the samples; thus the degrees of condensation reported in Table 4 during 3 h. Elemental analysis showed that in all cases the hybrid gels deviated from the ideal stoechiometry based on have been estimated from CP MAS spectra. All the samples derived from the rigid precursors 1 and 2 totally polycondensed silsesquioxane materials (Table 2). 2708 J. Mater. Chem., 1998, 8, 2707–2713Table 2 Elemental analyses, experimental formulas, and 29Si CP MAS NMR data of xerogels Elemental analysis (found %) 29Si CP MAS NMR (d)e Entry Xerogel C H Si Experimental formula T0 T1 T2 T3 1 1M 38.73 3.77 23.40 C7.76H9.01Si2O5.16 a — -61 -70 -78 2 1T 36.15 3.21 25.40 C6.66H7.06Si2O4.88 a — -61 -70 -78 3 2M 33.36 3.82 26.10 C8.93H12.22Si3O7.38 b — -62 -70 -78 4 2T 30.85 3.20 27.35 C7.89H9.78Si3O7.44 b — -61 -70 -78 5 3M 46.62 5.88 25.45 C8.62H12.94Si2O3.16 c -46 -55 -63 -71 6 3T -47 -55 -63 -70 7 4M -42 -50 -59 -67 8 4T 50.31 5.42 22.75 C10.32H13.24Si2O3.30 d — — -57 -66 Ideal formula: aC6H4Si2O3; bC6H3Si3O4.5; cC8H8Si2O3; dC10H12Si2O3; emajor resonances in bold type.Fig. 1 29Si CP MAS NMR spectra of xerogels; (a) 1M, (b) 2M, (c) 3M, (d) 4M. Fig. 2 29Si CP MAS NMR spectra of xerogels; (a) 1T, (b) 2T, (c) 3T, (d) 4T. had similar degrees of condensation, lying in the range 61–67%, whatever the solvent and concentration. These values are comparable to those reported by Shea and coworkers6 on than in MeOH, in good agreement with the elemental analysis other examples.reported in Table 2. On the other hand, the degree of condensation of the samples derived from the more flexible precursors 3 and 4 Hydrophilic character of the xerogels depended on the nature of both the solvent and the precursor. Thus, the samples prepared in THF showed significantly higher All the solids obtained from 1 and 2 exhibited significant hydrophilic character, whereas the gels derived from 3 and 4 degrees of condensation than those prepared in MeOH; for a given solvent, the samples prepared from precursor 4 showed always showed a low aYnity for water independent of the gelation solvent.Weight increases in a 60% humidity atmos- significantly higher degrees of condensation than those prepared from precursor 3.phere at 25 °C (E0.6%) are reported in Table 4 and are in agreement with the IR data of the xerogels: the n(OH) 13C CP MAS NMR spectra (Fig. 4) showed very weak signals attributable to residual methoxy groups when the absorption band centered at 3370 cm-1 due to the presence of silanol groups was stronger for the xerogels obtained from 1 solvent was THF and more intense signals in the case of methanol indicative of a higher degree of hydrolysis in THF and 2 than from 3 and 4.J. Mater. Chem., 1998, 8, 2707–2713 2709Fig. 3 29Si SPE MAS NMR and 29Si CP MAS NMR spectra of xerogels; (a) 1M, (b) 4M. Table 3 Comparison between SPE MAS and CP MAS 29Si NMR data for samples 1M and 4M %T0 %T1 %T2 %T3 %condensation Xerogel SPE/CP SPE/CP SPE/CP SPE/CP SPE/CP 1M 0/0 17.0/18.7 61.5/63.5 21.5/17.8 68.2/66.4 4M 4.5/6.4 26.0/27.1 46.2/44.3 23.4/22.3 62.9/60.9 Fig. 4 13C CP MAS NMR spectra of xerogels; (a) 4M, (b) 4T. Texture of the solids Nitrogen BET measurements19 gave specific surface areas very high for both samples, which is also indicative of microporosity. which were very diVerent depending on the structure of the precursor and the nature of the solvent.The solids obtained All the xerogels prepared from 2 exhibited isotherms intermediate between type I and IV (characteristics of meso- from 1 and 2 exhibited very high specific surface areas whatever the solvent used: 549–1018 m2 g-1 in MeOH and porous solids) and high specific surface areas (Table 5, entries 3 and 4). No narrow pore size distribution was observed in 370–766 m2 g-1 in THF (Table 4, entries 1–4).By contrast, the solvent employed for the hydrolysis– all cases. Sample 3T (Table 5, entry 5) was mesoporous and showed polycondensation reaction had a drastic influence on the texture of the solids in the case of 3 and 4. The xerogels 3M a substantial specific surface area (277 m2 g-1), whereas 3M presented no significant specific surface area.and 4M prepared in MeOH exhibited no significant surface areas (Table 4, entries 5, 7) while high specific surface areas For xerogels derived fom precursor 4, the nature of the solvent used during the hydrolysis–polycondensation had a were observed for samples 3T and 4T prepared in THF (Table 4, entries 6, 8). drastic influence on the texture of the resulting solids since sample 4T exhibited a high specific surface area (565 m2 g-1) Adsorption–desorption isotherms of 1M, 1T, 2M, 2T, 4T are shown in Fig. 5 and 6. The determination of the porous whereas sample 4M had no significant specific surface area. This diVerence may be correlated to the increase of the rigidity volume by the BJH method 20 and the evaluation of the microporous volume by the analysis of the t-plot diagram of the network of the sample prepared in THF owing to the higher degree of condensation (61% for 4M and 87% for 4T).have been performed in each case. The BET specific surface areas and porous volumes of the xerogels are reported in For 4T the solid was mainly mesoporous, with a low microporous contribution (estimated to be ca. 20% of the total porous Table 5. The xerogels 1M and 1T (Table 5, entries 1 and 2) showed volume). No narrow pore size distribution was observed. Despite its high specific surface area, gel 4T had a low aYnity type I isotherms,21 characteristic of microporous solids. Indeed, the microporous volume represented respectively 60 for water (5–6%) in accord with the low amount of hydroxy groups as shown by IR spectroscopy.and 79% of the total porous volume. The BET constant c was Table 4 Degree of condensation, hydrophilicity, specific surface area and degree of complexation of xerogels Degree of Surface area/ Degree of Entry Xerogel condensation (%)a E0.6 (%) m2 g-1 complexation (%) 1 1M 66 22 549 18 2 1T 67 18–22 370 11 3 2M 61 22–24 1018 27 4 2T 64 30–31 766 33 4 3M 55 2 <10 25 6 3T 72 3 277 42 7 4M 61 2 <10 84 8 4T 87 5–6 565 64 aEstimated from the 29Si CP MAS NMR spectra. 2710 J. Mater. Chem., 1998, 8, 2707–2713Fig. 5 N2 adsorption–desorption isotherms of xerogels; (a) 1M, (b) 2M. The most important feature for the silsesquioxanes reported here is the diVerence of the texture of the solids connected with the geometry of the organic unit: the materials with a ‘rigid’ bridge obtained from 1 and 2 exhibit in all cases a high hydrophilicity (18–31%) and a high surface area (370–1018 m2 g-1).The hydrophilic character (E0.6) is larger in THF than in MeOH for 2 (Table 4, entries 3, 4) although the surface area is lower, which suggests a larger amount of hydroxy groups. This observation is consistent with the IR and 13C NMR data.By contrast, with more ‘flexible’ spacers 3 and 4, the texture is dependent on the nature of the solvent Fig. 6 N2 adsorption–desorption isotherms of xerogels; (a) 1T, (b) used in the hydrolysis–polycondensation reaction. No signifi- 2T, (c) 4T. cant surface areas were observed using MeOH while high surface areas were obtained using THF. at ca. 1970 and 1880 cm-1 due to carbonyl groups, the Chemical reactivity of xerogels intensity of which depended on the degree of complexation with Cr(CO)3. 13C CP MAS NMR spectra clearly revealed We have previously shown that the chemical reactivity of hybrid solids can be used as a tool for studying the solid the presence of both uncomplexed aryl and aryl–Cr(CO)3 units: for example, for 4M, the spectrum showed signals arrangement as a function of molecular structure.10,14 The accessibility of the aromatic groups of the hybrid network has corresponding to carbonyl groups at d 234, complexed aromatic carbons aryl–Cr(CO)3 at d 93 and 117 and uncomplexed been studied, using the reaction of complexation of aryl fragments with Cr(CO)6 according to reaction 2.aromatic groups at d 128 and 145 (Fig. 7). These 13C CP MAS NMR data are reported in Table 6. 29Si CP MAS NMR Complexation reactions occur under heterogeneous conditions. The silsesquioxanes obtained from 1–4 were treated spectra before and after complexation with Cr(CO)6 were unchanged. The degree of complexation was determined by with Cr(CO)6. After 65 h of reflux in a 80/20 mixture of Bun2O–THF a yellow–green powder was isolated.The com- EDS and chemical elemental analysis. The results deduced from elemental analysis (see Experimental section) showed a plexation of arene groups was evidenced by FTIR and 13C CP MAS NMR spectroscopies as described in the case of ‘auth- greater reproducibility and are reported in Table 4. The xerogels obtained from ‘rigid’ precursors 1 and 2 led to entic’ materials.12b FTIR spectra showed new absorption bands J.Mater. Chem., 1998, 8, 2707–2713 2711Table 5 N2 adsorption desorption data of xerogels Entry Xerogel BET surface/m2 g-1 Vpore tot a/cm3 g-1 Vpore ads b/cm3 g-1 Vpore des b/cm3 g-1 Vmicropore c/cm3 g-1 c 1 1M 549 0.27 0.11 0.11 0.16 470 2 1T 370 0.19 0.04 0.03 0.15 814 3 2M 1018 0.80 0.74 0.74 0.10 106 4 2T 766 0.53 0.55 0.49 — 146 5 3T 277 0.20 0.17 — — 83 6 4T 565 0.37 0.29 0.27 0.08 122 aP/P0=0.99.bCumulative pore volume of pores between 17 and 3000 A° diameter. cEstimated by the t-plot method, using Harkins and Jura standard isotherm and thickness range between 4 and 6 A° . attributed to flexibility due to the methylene groups present in the organic moiety. When the gelation solvent was THF a decrease of the degree of complexation (64%) was observed for 4T (Table 4, entry 8).This decrease can be explained by the higher degree of polycondensation of the hybrid material which induces a lower swelling of the solid corresponding to a lower diVusion of Cr(CO)6. However the poor reactivity observed with the gels obtained from 3 appears unexpected, since the solids obtained from 3 and 4 in MeOH exhibited the same specific surface areas (<10 m2 g-1) and the same aYnity for water (2%).This can be explained by a less ‘flexible’ organic spacer for 3 and consequently more diYcult diVusion of Cr(CO)6 in the hybrid network. The precursor 3 appears to be intermediate between 1 and 4. Fig. 7 13C CP MAS NMR spectrum of the solid obtained by reaction of xerogel 4M with Cr(CO)6.Analysis of the surface of solids 1M and 4M by time of flight secondary ion mass spectrometry (TOF SIMS) was in agreement with the results presented above. For 1M, the main Table 6 13C CP MAS NMR data (d) of xerogels after reaction with ion detected is at m/z=45 (SiOH+), in accord with the high Cr(CO)6 hydrophilicity (OH groups at the surface) and poor chemical Complexed reactivity.22 By contrast, for 4M, ions at m/z=77 (phenyl ) Residual Unchanged and m/z=91 (tropylium) are observed23 in accord with an Xerogel CH2Si CH2Ar CH3O Ar Ar CO hydrophobic solid and a high chemical reactivity.For 1M, the OH groups are located at the surface in agreement with the 1M — — 50 134 —a —a high hydrophilicity, while for 4M the presence of ions at m/z= 1T — — 50 134 —a —a 2M — — 50 131, 143 88, 105 232 77 and 91 is in accord with the hydrophobicity observed.This 2T — — 51 130, 143 —a —a means that the arrangement of the organic groups in the two 3M — 21 50 129, 133 96, 109 235 solids is completely diVerent. 3T — 21 50 133, 129 106, 94 234 4M 14 29 51 145, 128 93, 117 234 4T 15 29 50 143, 128 114, 92 233 Conclusion aNot detected.The results presented here show that the kinetic parameters involved in the hydrolysis–polycondensation reactions have an influence on both the physical and chemical properties of poor levels of complexation (11–33%), despite their high specific surface areas (370–1020 m2 g-1). However, the surface the resulting silsesquioxane materials. The structure of the organic precursor appears to be a area and the degree of complexation of the xerogels derived from precursor 2 were about twice as high as those of the determining parameter for the solid state properties of the xerogels containing aromatic groups.Molecular precursors xerogels derived from precursor 1. These results suggest an organization of the solids in which only a few aromatic groups containing a ‘rigid’ organic group always led to hydrophilic solids with similar degrees of condensation.In all cases, these are accessible at the surface. Futhermore the precursors 1 and 2 have diVerent geometries: 1 is linear with six hydrolysable solids had high specific surface areas and no narrow pore size distribution whatever the solvent and the concentrations used groups, while 2 has a planar geometry with nine hydrolysable groups.during the hydrolysis–polycondensation process and poor chemical reactivity towards Cr(CO)6 was observed (11–33%). For xerogels obtained from ‘flexible’ precursors 3 and 4 the degrees of complexation were highly dependent on the nature By contrast when the precursor contained a ‘flexible’ organic spacer, the texture, the degree of condensation and the reactiv- of the organic spacer and the solvent and poor reactivity was observed when 3 was hydrolysed in MeOH (Table 4, entry 5).ity towards Cr(CO)6 of the resulting solids strongly depended on the solvent. The solids prepared in methanol had very low A moderate degree of complexation was observed for 3T which may be attributed to the higher specific surface area significant specific surface area, whereas in THF high specific surface area products were observed.The degree of com- observed in this case (277 m2 g-1). For 4, a high degree of complexation (84%) was obtained plexation with Cr(CO)6 depended on the flexibility of the organic group: when four methylene groups were present (high in the case of 4M (Table 4, entry 7), showing a high accessibility of the organic groups, although the specific surface area flexibility) a high reactivity was observed.By contrast, a low reactivity was observed when the flexibility was decreased (two was very low (<10 m2 g-1). This behaviour can be explained by the facile diVusion of Cr(CO)6 in the network due to methylene groups). In both cases, however, the solids were hydrophobic.swelling of the solid in the presence of solvent. This may be 2712 J. Mater. Chem., 1998, 8, 2707–271310 R. J. P. Corriu, J. J. E. Moreau, P. The�pot and M. Wong Chi The results are rigorously reproducible if the experimental Man, Chem. Mater., 1996, 8, 100. conditions are strictly controlled. 11 (a) R. J. P. Corriu, J. J. E. Moreau, P. The�pot, M.Wong ChiMan, Work is now in progress in order to determine the influence C.Chorro, J. P. Le`re-Porte and J. L. Sauvajol, Chem. Mater., of other kinetic parameters on the properties of the solids 1994, 6, 640; (b) J. L. Sauvajol, C. Chorro, J. P. Le` re-Porte, obtained by this sol–gel process. R. J. P. Corriu, J. J. E. Moreau, P. The�pot and M. Wong Chi Man, Synth. Met., 1994, 62, 233. 12 (a) G. Cerveau, R. J. P. Corriu and N. Costa, J. Non-Cryst. Solids, 1993, 163, 226; (b) G. Cerveau, R. J. P. Corriu and C. Lepeytre, References Chem.Mater., 1997, 9, 2561; (c) R. J. P. Corriu, P. Hesemann and 1 (a) H.K. Schmidt, Mater. Res. Soc. Symp. Proc., 1984, 32, 327; G. Lanneau, J. Chem. Soc. Chem. Commun., 1996, p. 1845. (b) H.K. Schmidt, Inorganic and Organometallic Polymers, ACS 13 (a) P.J. Barrie, S. W. Carr, D. Li Ou and A. C. Sullivan, Chem. Symp. Ser. 360, American Chemical Society, Washington DC, Mater., 1995, 7, 265; (b) S. W. Carr, M. Motevalli, D. Li Ou and 1988, p. 133; (c) H.K. Schmidt, Mater. Res. Soc. Symp. Proc., A. C. Sullivan, J. Mater. Chem., 1997, 7, 865. 1990, 180, 961 and references therein; (d) Hybrid 14 (a) G. Cerveau, R.J. P. Corriu and C. Lepeytre, J. Mater. Chem., 1995, 75, 793; (b) G. Cerveau, C. Chorro, R. J. P. Corriu, Organic–Inorganic Composites, ACS Symp. Ser. 585, American C. Lepeytre, J. P. Le`re-Porte, J. Moreau, P.The�pot and M. Wong Chemical Society, San Diego DC, 1994; (e) New J. Chem.,1994, Chi Man, Hybrid Organic–Inorganic Silica Materials, ACS Symp. 18, 989; ( f) J.D. Mackenzie, J.Sol–Gel Sci. Technol., 1994, 2, 81 Ser. 585, American Chemical Society, Washington DC, 1995, and references therein; ( g) U. Schubert, N. Hu�sing and A. Lorenz, p. 210 and references therein; (c) G.Cerveau, P. Chevalier, Chem. Mater., 1995, 7, 2010; (h) D. A. Loy and K.J. Shea, Chem. R. J. P. Corriu, C. Lepeytre, J. P. Le`re-Porte, J. Moreau, Rev., 1995, 95, 1431; (i) P. Judenstein and C.Sanchez, J. Mater. P. The�pot and M. Wong Chi Man, Tailor-maide Silicon Oxygen Chem., 1996, 6, 511. Compounds. From Molecules to Materials, ed. R. Corriu and 2 (a) R. J. P. Corriu and D. Leclercq, Angew. Chem., Int. Ed. Engl., P. Jutzi, Vieweg Publ., 1996, p. 273 and references therein. 1996, 35, 1420 and references therein; (b) R. J. P. Corriu, C.R. 15 D. F. Schriver, The manipulation of air-sensitive compounds, Acad.Sci. Paris, Se�r. II, 1998, 83 and references therein; MacGraw-Hill, New York, 1969. (c) R. J. P. Corriu, Polyhedron, 1998, 17, 925 and references 16 R. J. P. Corriu, D. Leclercq, A. Vioux, M. Pauthe and therein. J. Phalippou, Ultrastructure Processing of Advanced Ceramics, ed. 3 K. J. Shea, D. A. Loy and O. W. Webster, Polym. Mater.Eng. J. D. Mackenzie and D. R. Ulrich, Wiley, New York, 1988, p. 113. Sci., 1990, 63, 281. 17 E. A. Williams, NMR Spectroscopy of Organosilicon Compounds, 4 (a) R. J. P. Corriu, J. J. E. Moreau, P. The�pot and M. Wong Chi ed. S. Pataý� and Z. Rappoport, Wiley, New York, 1989, p. 511. Man, Chem. Mater., 1992, 4, 1217; (b) R. J. P. Corriu, 18 (a) M. Ma� gi, E. Lippmaa, A. Samoson, G. Engelhardt and J. J. E. Moreau and M. Wong Chi Man, J. Sol–Gel, Sci. Technol., A. R. Grimmer, J. Phys. Chem., 1984, 88, 1518; (b) H.Marsmann, 1994, 2, 87. 29Si NMR Spectroscopic Results, ed. P. Diehl, E. Fluck and 5 (a) K. J. Shea, D. A. Loy and O. Webster, J. Am. Chem. Soc., R. Kosfeld, Springer Verlag, Berlin. 1992, 114, 6700 and references therein; (b) K. M. Choi and 19 S. Brunauer, P. H. Emmet and E. Teller, J. Am. Chem. Soc., 1938, K. J. Shea, J. Am. Chem. Soc., 1994, 116, 9052; (c) D. A. Loy, 60, 309. G. M. Jamison, B. M. Baugher, S. A. Myers, R. A. Assink and 20 E. Barett, L. Joyner and P. Halenda, J. Am. Chem. Soc., 1951, K. J. Shea, Chem. Mater., 1996, 8, 656. 73, 373. 6 H. W. Oviatt Jr, K. J. Shea and J. H. Small, Chem. Mater., 1993, 21 S. Brunauer, L. S. Deming, W. S. Deming and E. Teller, J. Am. 5, 943. Chem. Soc., 1940, 62, 1723. 7 A. Antony, R. J. P. Corriu, J. J. E. Moreau and M. Wong Chi 22 G. Cerveau, R. J. P. Corriu, J. Dabosi, J. L. Aubagnac, Man, unpublished results. R. Combarieu and Y. de Puydt, J. Mater. Chem., 1998, 8,1761. 8 S. T. Hobson and K. J. Shea, Chem. Mater., 1997, 9, 616. 23 G. Cerveau, R. J. P. Corriu, C. Lepeytre and R. Combarieu, to 9 (a) K.M. Choi and K. J. Shea, Chem.Mater., 1993, 5, 1067; (b) K. be published. M. Choi and K. J. Shea, J. Phys. Chem., 1994, 98, 3207; (c) K. M. Choi, J. C. Hemminger and K. J. Shea, J. Phys. Chem., 1995, 99, 4720 and references therein. Paper 8/05794J J. Mater. Chem., 1998, 8,