Organo-modified silica aerogels and implications for materialhydrophobicity and mechanical properties{

Laura Martın,ab J. Oriol Osso,ac Susagna Ricart,a Anna Roig,*a Olga Garcıad and Roberto Sastred

Received 15th August 2007, Accepted 1st November 2007

First published as an Advance Article on the web 14th November 2007

DOI: 10.1039/b712553d

Two types of organo-modified silica gels, with the organic groups trimethoxymethylsilane

(TRIMOS) and 3-(trimethoxysilyl)propyl methacrylate (TMSPMA), have been synthesized. The

wet gels were dried by supercritical evacuation of the solvent. The materials with enhanced

performance were those hybridized with TRIMOS and dried above the critical conditions of

methanol. Such aerogels were found to be hydrophobic with mechanically improved properties

whilst still maintaining the characteristic transparency of pure silica aerogels. The condensation

degree and the number of superficial hydroxyl groups were determined from solid-state 29Si-NMR

spectroscopy. Mechanical properties were measured using nanoindentation.

Introduction

Aerogels are low density highly porous solids (¢90% of

accessible mesoporosity); they can be fabricated by sol–gel

processes and supercritical drying of the wet gels. The potential

of aerogels for technological applications has often been

limited by their extreme sensitivity to moisture and poor

mechanical properties. Large surface area materials

(.200 m2 g21) usually adsorb important amounts of molecular

gases, i.e. silica aerogels can contain up to 15 wt% of adsorbed

air and water vapour. Wettability of a material depends on

both the physical and the chemical nature of the surface, i.e.

surface roughness and the chemical functional groups of the

surface. A recent comprehensive review on the subject has been

written by Feng and Jiang.1 The fabrication of superhydro-

phobic solids, with contact angles (CA) larger than 150u, has

recently attracted extensive interest both from the fundamental

as well as from the practical points of view since the repellence

of water can confer practical applications to the material,

such as self cleaning surfaces, microfluidic devices, prevention

of the adhesion of moisture to antennae and windows,

avoidance of corrosion, reduction of friction drag and

functional low k-dielectric materials.2–4

In the case of silica aerogels, the MSi–OH groups at the

surfaces of the silica clusters are the main source of

hydrophilicity promoting condensation reactions. In order to

decrease the pore affinity for moisture adsorption and make

aerogels less hydrophilic a number of strategies have been

used, i.e. avoiding the presence of terminal hydroxyl groups

by co-gelling with silicon precursors containing at least one

non-polar chemical group,5–8 incorporation of fluorinated

chains covalently bound onto the surface9–11 and reactions of

aerogels with gaseous reagents.12

Regarding their mechanical properties, typical Young’s

modulus values for aerogels are a factor of 102–104 smaller

than that of silica glass13 so they can be easily compressed. We

have previously shown that nanoindentation is a suitable

method to assess their mechanical properties.14 It is a major

challenge to improve the mechanical properties of such

materials without adversely affecting their other properties.

There have been a few studies reporting attempts to improve

the mechanical properties of inorganic aerogels, for silica15–17

and more recently for alumina aerogels.18 An interesting

overview describing the mechanical properties of hybrid

organic–inorganic films has also recently been published.19

In this work, we revisit the synthesis of organo-modified

silica aerogels with the aim of fabricating large pieces of

mechanically improved hydrophobic materials while still

maintaining the characteristic transparency of base catalysed

silica aerogels. Two organic groups (trimethoxymethylsilane

and 3-(trimethoxysilyl)propyl methacrylate) were used to

hybridize the inorganic silica network. The wet gels were dried

either at the supercritical conditions of methanol or by first

exchanging the methanol for liquid CO2 and subsequently

drying the gels at the supercritical conditions of CO2. It was

found that 3-(trimethoxysilyl)propyl methacrylate did not

withstand high temperature supercritical drying. Moreover,

the materials dried at the supercritical conditions of CO2 were

of lower quality in terms of transparency. In this scenario, we

have focused on the materials hybridized with trimethoxy-

methylsilane in which the effects of the organic concentration

on the degree of hydrophobicity, the transparency and the

mechanical properties have been investigated. Solid-state29Si-NMR spectroscopy was used to determine the condensa-

tion degree and the number of superficial hydroxyl groups,

and the resulting data were correlated with the water surface

contact angle. It was found that some of the materials present

contact angles larger than 160u. Nanoindentation was used to

assess their mechanical properties.

aInstitut de Ciencia de Materials de Barcelona (ICMAB-CSIC),Campus de la UAB, Cerdanyola, 08193, Spain. E-mail: roig@icmab.es;Fax: +34 935805729; Tel: +34 935801853bDepartment of Chemistry, Universitat Autonoma de Barcelona,Campus UAB, Cerdanyola, 08193, SpaincMATGAS 2000 A.I.E., Campus de la UAB, Cerdanyola, 08193, SpaindInstituto de Ciencia y Tecnologıa de Polımeros, CSIC, Juan de laCierva 3, Madrid, 28006, Spain{ Electronic supplementary information (ESI) available: IR and 13CNMR spectra; hardness and Young’s modulus values. See DOI:10.1039/b712553d

PAPER www.rsc.org/materials | Journal of Materials Chemistry

This journal is � The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18, 207–213 | 207

Experimental

Synthesis

Tetramethoxysilane (TMOS), trimethoxymethylsilane

(TRIMOS) and 3-(trimethoxysilyl)propyl methacrylate

(TMSPMA) were purchased from Sigma-Aldrich (purity of

98%) and were used as received. The molecular structures of

these alkoxides are shown in Scheme 1.

The organo-modified silica aerogels were produced in a one-

step base catalysed sol–gel synthesis by hydrolysis and poly-

condensation of TMOS and TRIMOS (series 1) or TMOS and

TMSPMA (series 2), diluted in methanol and using aqueous

ammonia (0.5 mol L21 NH3 aqueous solution) as a catalyst.

The following procedure was used for all syntheses: over a

stirred mixture of TMOS and TRIMOS or TMOS and

TMSPMA, methanol was carefully added. In some cases, an

ice-water bath was used to control the exothermic reaction.

Afterwards, the distilled water and the catalyst were added.

The solution was poured into cylindrical moulds that were

then sealed until the gel was formed. Gelation times ranged

from 3 min to 2 h. Next, the gels were washed with methanol

for 24 h, this procedure was repeated three times and the gels

were subsequently dried under supercritical conditions.

The molar ratios of the reactants were: (TMOS : TRIMOS

or TMSPMA) : CH3OH : H2O : NH3 aqueous solution =

(1.0) : 12.25 : 2.0 : 0.065. The amounts of TMOS, TRIMOS

or TMSPMA, expressed as molar ratios, were varied as

follows: from 0.90TMOS : 0.10TRIMOS to 0.40TMOS :

0.60TRIMOS in series 1 and from 0.95TMOS : 0.05TMSPMA

to 0.75TMOS : 0.25TMSPMA in series 2. A sub-stoichiometric

quantity of water was used to slow the gelation and extend

the range of transparent materials. The concentration of the

organic part was increased up to a concentration where the

gels become milky.

High temperature supercritical extraction of the methanol

was carried out in a fully automated autoclave which allowed

the precise control of temperature and pressure. The pressure

was initially raised to 80 bar using CO2. Once the final pressure

was reached, the temperature was slowly increased to 260 uC

and maintained at this value for 4 h. Then, supercritical

methanol was removed by slow depressurization, and finally,

the temperature was lowered to room temperature. Some gels

were also dried by first exchanging, inside the autoclave, the

methanol for liquid CO2 (at room temperature and a pressure

of 100 bar, over 6 h). After that, the liquid CO2 was trans-

formed to supercritical CO2 by increasing the temperature to

42 uC. This step was done at 120 bar of pressure over 2 h. Next,

supercritical CO2 was removed by slow depressurization. This

process will be referred to as low temperature drying.

Characterization techniques

The materials were systematically characterized using comple-

mentary techniques. The bulk densities of the organo-modified

silica aerogels were estimated by measuring the dimensions and

mass of each monolithic sample. Surface area, SBET, mean

pore size and pore volume determinations were performed by

the Brunauer–Emmett–Teller (BET) and Barrett–Joyner–

Halenda (BJH) methods using an ASAP 2000 surface area

analyzer (Micromeritics Instruments Corp.). Samples of

approximately 0.03 g were heated to 180 uC under vacuum

(1025 Torr) for at least 22 h to remove all adsorbed species.

SBET was obtained from the adsorption isotherm of N2 at 77 K.

IR spectra were recorded on a PerkinElmer Spectrum One

Fourier transform infrared spectrophotometer in the 4000–

450 cm21 range with the conventional KBr method.

The experiments of solid-state 13C and 29Si cross-polarity/

magic-angle spinning (CP-MAS) NMR spectroscopy

(13C-NMR and 29Si-NMR, respectively) were performed in a

Bruker AvanceTM 400 spectrometer (Bruker Analytic GmbH,

Karlsruhe, Germany) equipped with a Bruker UltraShieldTM

9.4 T (13C and 29Si resonance frequencies of 100.62 and

79.49 MHz, respectively), 8.9 cm vertical-bore superconduct-

ing magnet. In both cases, spectra were acquired at ambient

temperature using a standard Bruker broad-band MAS probe.

Representative samples were grounded and packed in 4 mm

zirconia rotors, sealed with Kel-FTM caps and spun at 5 kHz.

The 90u pulse width was 3.5–4.5 ms and, in all cases, high-

power proton decoupling was used. All free-induction decays

were subjected to standard Fourier transformation and phasing.

The chemical shifts were externally referenced to TMS.

The 13C-NMR spectra were acquired with 1 ms CP contact

time and 5 s recycle decay. Each spectrum was obtained with

800 averages and 5 Hz line broadening.

The 29Si-NMR spectra were obtained with 4 ms CP contact

time, 5 s recycle delay, 6000 averages and 75 Hz line broadening.

The spectra were deconvoluted by using Gaussian/Lorentzian

fits. The NMR spectra were evaluated with the software

package XWIN-NMRTM provided by Bruker.

The contact angles measurements were performed in a

DSA100 instrument (Drop Shape Analysis System) from

Kruss. A droplet of water is placed on a piece of aerogel

located on a moveable table. The droplet is illuminated from

one side and a camera at the opposite side records the image.

Optical measurements were performed with a Cary|5|

Varian Ultraviolet-Visible-NIR spectrophotometer with an

ultraviolet-visible range from 200–800 nm. Aerogels were

carefully cut into 1 cm thickness and with parallel sides.Scheme 1 Molecular structures of alkoxides TMOS, TRIMOS and

TMSPMA.

208 | J. Mater. Chem., 2008, 18, 207–213 This journal is � The Royal Society of Chemistry 2008

For the mechanical characterization, an MTS NanoXP

indenter system with a Berkovich tip was used. Hardness and

modulus were calculated from the load–displacement curves

following Oliver and Pharr methodology.20 Maximum loads

between 1.5 and 3 mN were used leading to tip penetrations in

the range of 6 to 7 mm. The reported data were derived from at

least 20 individual indentation tests on each sample.

Results and discussion

The materials were obtained as monolithic cylinders of

approximate 1 cm diameter and 6–7 cm length. The samples

dried by supercritical evacuation of the methanol were

transparent and free of cracks, the materials become more

translucent as the amount of the organic part increases. Fig. 1

shows some of the materials obtained. Samples dried using

supercritical CO2 were less transparent and presented a

significant number of cracks. FT-IR spectroscopy of the solid

samples was carried out to monitor the presence of organic

groups after the high temperature supercritical drying. In the

case of series 1, IR characterization was sufficient to verify the

presence of the CH3 groups. For series 2, IR and 13C-NMR

analysis confirmed the absence of the methacrylate group on

the materials dried by supercritical evacuation of methanol

(see ESI{). This is in agreement with the previously reported

results that indicated the degradation of the methacrylate

chain with the high temperature drying process.21

Table 1 summarises the information on the main structural

aspects of the materials of series 1 dried at high temperature. It

can be seen that the density values are characteristic of those

reported for silica aerogels and, within the experimental error,

do not change significantly as the amount of the organic part is

increased. Large values of the specific surface area were also

obtained for all samples. As the amount of organic part is

increased, the specific surface area values increase substantially

(from 399 to 837 m2 g21) and the average pore size decreases

(from 403 to 162 A) whilst the total pore volume remains

almost constant (from 4.7 to 3.4 cm3 g21). This indicates an

increase in the number of pores and a more branched

microstructure.

A preliminary check on the hydrophobic/hydrophilic nature

of the materials was done by placing a water droplet on the

surface of the aerogels. It was observed that for series 1 the

water droplet remained on the top of the surface, with large

contact angle, for both low and high temperature dried

materials indicating that the pore surface is deficient in

hydroxyl terminations (see Fig. 2a)). In the case of series 2,

Fig. 1 Images of the materials obtained with high temperature

drying: a) series 1: TMOS : TRIMOS; b) series 2: TMOS : TMSPMA.

Table 1 Density, surface area, total pore volume and average pores size for series 1 of organo-modified silica aerogels

TMOS : TRIMOS(% molar ratio) Density ¡ 0.01/g cm23 BET Surface area ¡ 40/m2 g21 Pore volume ¡ 0.3/cm3 g21

Average pore size ¡ 10/A(= 4V/SBET)

90 : 10 0.16 399 4.4 40385 : 15 0.13 487 4.2 33875 : 25 0.12 709 4.7 26560 : 40 0.11 719 4.4 23950 : 50 0.13 864 3.9 16645 : 55 0.13 880 3.4 14040 : 60 0.13 837 3.6 162

Fig. 2 a) Pieces of aerogels of series 1 with a water droplet on the top,

when dried under low (i) and high (ii) temperature. b) Pieces of

aerogels of series 2 with a water droplet on the top, when dried under

low (i) and high (ii) temperature.

This journal is � The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18, 207–213 | 209

the water did not damage the low temperature dried materials

but it did damage the high temperature dried ones (see

Fig. 2b)). The latter procedure was used as a quick test to

asses the presence of the methacrylate groups. The same test

was done using other organic solvents, namely cyclohexane,

dimethylformamide, chloroform, ethanol and acetonitrile.

All of them were absorbed by the aerogel surfaces of

both series.

Fig. 3 shows the dependence of the surface contact angle and

the percentage of silanols on the surface of series 1 aerogels.

It can be noticed that as the organic fraction is increased,

the contact angle increases and the percentage of MSi–OH

bonds decreases. Moreover, up to a percentage of 40% of

TRIMOS the contact angle increases with a small slope while a

significant increase in the slope can be observed for molar

fractions above 40%. A contact angle of 160u was measured for

aerogels with a 60% TRIMOS molar fraction.

The percentage of MSi–OH bonds at the aerogel surface was

calculated considering the data obtained by NMR (see below)

and according to the equation: % MSi–OH = [3.0 (% area T0) +

2.0 (% area T1) + 1.0 (% area T2) + 0.0 (% area T3) + 4.0

(% area Q0) + 3.0 (% area Q1) + 2.0 (% area Q2) + 1.0 (% area

Q3) + 0.0 (% area Q4)]/(3 + 4). The % MSi–OH bonds varies

from 5.87% for TMOS : TRIMOS (90 : 10) to 2.71% for

TMOS : TRIMOS (40 : 60); such a decrease of the % MSi–OH

groups at the surface of the aerogel directly relates to the

hydrophobicity of the material and is revealed by the increase

of the contact angle.29Si-NMR of solids was carried out to investigate the

coordination of the Si atoms depending on the proportions of

the organic group. Fig. 4 shows the 29Si-NMR spectra of

selected samples. Qn (n = 0, 1, 2, 3, 4) and Tm (m = 0, 1, 2, 3)

correspond to the number of siloxane bridges bonded to the

silicon atom of TMOS and hybrid system respectively; in all

cases presented only Q3, Q4 and T2, T3 are detected. Fig. 5

presents the results in a more quantitative form. The degree of

condensation was calculated according to equation previously

reported by Ji et al.22 The degree of condensation of

hybrid aerogels = [1.0 (% area T1) + 2.0 (% area T2) + 3.0

(% area T3) + 1.0 (% area Q1) + 2.0 (% area Q2) + 3.0 (% area

Q3) + 4.0 (% area Q4)]/(3 + 4).

Introducing TRIMOS into a TMOS system results in fewer

condensation bonds and hence the degree of condensation

decreases. It can be observed (see Fig. 5) that the condensation

degree radically decreases between TMOS : TRIMOS (100 : 0)

and TMOS : TRIMOS (90 : 10). However, from 90 : 10 to

40 : 60, the degree of condensation remains almost constant

even when the amount of the organic part is increased; the Q3

and Q4 species have been replaced by T2 and T3 ones. The

results previously reported by Husing and Schubert23 indicate

that under basic conditions the hydrolysis and condensation

rates of TMOS are much faster than those of TRIMOS. They

Fig. 3 Left: contact angle of a water droplet with the aerogel surface

for series 1, right: % Si–OH bonds for series 1.

Fig. 4 a) 29Si-NMR spectrum of sample TMOS : TRIMOS (40 : 60); b) 29Si-NMR spectrum of TRIMOS; c) 29Si-NMR spectrum of TMOS.

210 | J. Mater. Chem., 2008, 18, 207–213 This journal is � The Royal Society of Chemistry 2008

explained that the gel network forms quasi-totally from

TMOS, and then TRIMOS condenses with the silanol surface

groups of the gel giving a Si–O–Si–CH3 surface. Therefore,

hydrophobicity results due to the attachment of hydrolytically

stable MSi–CH3 groups on the surfaces of SiO2.24 From our

results we can say that the difference in reactivity of TMOS

and TRIMOS is not reflected in a regular decrease of the

condensation degree as the amount of organic part increases

since a sharp decrease in the condensation rate from TMOS :

TRIMOS (100 : 0) to TMOS : TRIMOS (90 : 10) is clearly

observed but a less prominent decrease is seen as the TRIMOS

concentration increases.

Our materials show the characteristic transparency of pure

base catalysed silica for up to a molar ratio of 50% of

TRIMOS. The transparency range has been increased from

previously reported materials,24 whilst the materials with 55%

and 60% of TRIMOS are translucent. Fig. 6 shows the

absorbance versus wavelength in the UV-VIS range from 350–

800 nm for aerogels of series 1 carefully cut into 1 cm slabs

with parallel sides. All materials are very transparent in the

visible range. The materials absorbed more at shorter

wavelengths and the absorbance increases as the amount of

the organic part increases. The inset of Fig. 6 shows the optical

transmission at 500 nm versus the molar ratio of TRIMOS. In

the visible range, as the amount of the organic part increases,

the transmission decreases implying more efficient light

scattering and a decrease in the transparency of the materials.

This may be due to less homogeneity in the sizes of the

scattering centers (pores and/or particles) or to a phase separa-

tion of two gel networks from TMOS and TRIMOS and

Fig. 5 Analysis of the results obtained by 29Si-NMR of series 1.

Fig. 6 Absorbance versus wavelength in the UV-VIS range from

350–800 nm for 1 cm slabs of aerogels of series 1. Inset: optical

transmission (at 500 nm) versus the percentage for the same series.

This journal is � The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18, 207–213 | 211

differences in the refractive index between the two phases as

was previously pointed out in ref. 24. More in-depth investiga-

tions are needed to discern between the two hypotheses.

The manner in which aerogels’ mechanical properties are

affected by the organic modification of the materials was

investigated in series 1. Fig. 7 shows the curves of load versus

penetration depth for a few selected samples. The shape of

the curves demonstrates the elasto-plastic character of the

materials, the materials becoming more elastic as the organic

part is increased, this is inferred from the decrease of the area

between the loading and downloading branches. From the

macroscopic point of view the materials range from brittle

to rubber-like solids. Fig. 8 presents the dependence of the

Young’s modulus and hardness on the concentration of

organic part and the numerical values are gathered in ESI{(SM Table 1). It can be observed that the modulus decreases

by a factor of two when increasing the concentration from

TMOS (100) to TMOS : TRIMOS (40 : 60), as has been

recently observed also in other studies;25 while the hardness,

more directly related to the yield strain behaviour of the

material, is not affected. Thus, the solid networks for the

materials with higher concentration of organic part will be

more deformable, most likely due to a decrease of the network

degree of crosslinking (as a consequence of a decrease of the

condensation degree), whilst at the same time maintaining

almost constant the number of contact points, necking

between silica particles, which would account for a constant

value of the hardness along the series.

Conclusions

In summary, we have synthesised trimethoxymethylsilane and

3-(trimethoxysilyl)propyl methacrylate modified silica aerogels

with the aim of fabricating large pieces of mechanically

improved hydrophobic aerogels while still maintaining the

characteristic transparency of base catalysed pure silica ones.

The wet gels have been dried either at supercritical conditions

of methanol or by prior exchange of methanol with liquid CO2

and subsequently drying the gels at supercritical conditions of

CO2. IR and 13C-NMR analysis confirmed that the high

temperature drying causes degradation of the methacrylate

chain, although the methacrylate is not affected by the low

temperature drying. The materials with improved performance

were those hybridized with trimethoxymethylsilane and dried

above the critical conditions of methanol. For such series,

it was found that when increasing the concentration of

TRIMOS, the surface area increases, the average pore size

decreases and the water surface tension of the material

diminishes, reaching a maximum contact angle over 160u for

60% TRIMOS, even though it is only up to 40% of TRIMOS

that the transparency (at 500 nm) is unaffected. Concerning

the mechanical properties, as the organic part increases, the

hybrid aerogels can be more easily deformed nevertheless

the hardness is not affected. Summarizing, we have shown

that it is possible to fabricate custom-made aerogels with

functional properties.

Acknowledgements

This work has been partially financed by the Ministerio de

Educacion y Ciencia (MAT2006-13572-C02-01) and by the

Generalitat de Catalunya (2005SGR00452). We are grateful to

S. Villar, N. Roma, A. Bernabe and R. Solanas for technical

assistance. MATGAS is also acknowledged for the use of

the supercritical facilities. O. G. also wants to thank to

Comunidad Autonoma de Madrid (CAM) for financial

support (CAM-200660M025).

References

1 X. Feng and L. Jiang, Adv. Mater., 2006, 18, 3063–3078.2 Y. Li, W. Cai, B. Cao, G. Duan, F. Sun, C. Li and L. Jia,

Nanotechnology, 2006, 17, 238–243.3 D. A. Doshi, P. B. Shah, S. Singh, E. D. Branson, A. P. Malanoski,

E. B. Watkins, J. Majewski, F. van Swol and C. J. Brinker,Langmuir, 2005, 21, 7805–7811.

4 J. T. Han, X. Xu and K. Cho, Langmuir, 2005, 21, 6662–6665.5 F. Schwertfeger, W. Glaubitt and U. Schubert, J. Non-Cryst.

Solids, 1992, 145, 85.6 F. Schwertfeger, N. Husing and U. Schubert, J. Sol–Gel Sci.

Technol., 1994, 2, 103.7 N. Husing, U. Schubert, K. Misof and P. Fratzl, Chem. Mater.,

1998, 10, 3024–3032.8 A. V. Rao and G. M. Pajonk, J. Non-Cryst. Solids, 2001, 285,

202–209.9 A. Roig, E. Molins, E. Rodrıguez, S. Martınez, M. Moreno-Manas

and A. Vallribera, Chem. Commun., 2004, 2316–2317.10 P. R. Coronado, J. F. Poco and L. W. Hrubesh, US Pat. Appl.

20030032681 A1, Feb. 13, 2003.

Fig. 7 Displacement versus load curves for aerogels of series 1.

Fig. 8 Hardness and Young’s modulus values for aerogels of series 1.

212 | J. Mater. Chem., 2008, 18, 207–213 This journal is � The Royal Society of Chemistry 2008

11 L. W. Hrubesh, P. R. Coronado and J. H. Satcher, Jr., J. Non-Cryst. Solids, 2001, 285, 328.

12 S. S. Prakash, C. J. Brinker, A. J. Hurd and S. M. Rao, Nature,1995, 374, 439.

13 M. Gronauer and J. Fricke, Acustica, 1986, 59, 177.14 M. Moner-Girona, A. Roig, E. Molins, E. Martınez and J. Esteve,

Appl. Phys. Lett., 1995, 75(5), 653–655.15 N. Leventis, C. Sotiriou-Leventis, G. Zhang and A. M. M.

Rawashdeh, Nano Lett., 2002, 2, 957.16 M. Moner-Girona, E. Martınez, J. Esteve, A. Roig, R. Solanas and

E. Molins, Appl. Phys. A:. Matter. Sci. Process., 2002, 74, 119.17 E. M. Lucas, M. S. Doescher, D. M. Ebenstein, K. J. Wahl and

D. R. Rolison, J. Non-Cryst. Solids, 2004, 350, 244.18 S. O. Kucheyev, T. F. Baumann, C. A. Cox, Y. M. Wang,

J. H. Satcher, Jr, A. V. Hamza and J. E. Bradby, Appl. Phys. Lett.,2006, 89, 041911.

19 F. Mammeri, E. Le Bourhis, L. Rozes and C. Sanchez, J. Mater.Chem., 2005, 12, 3787–3811.

20 W. C. Oliver and G. M. Pharr, J. Mater. Res., 1992, 7,1564.

21 N. Husing, F. Schwertfeger, W. Tappert and U. Schubert, J. Non-Cryst. Solids, 1995, 186, 37–43.

22 X. Ji, S. Jiang, X. Qiu, D. Dong, D. Yu and B. Jiang, J. Appl.Polym. Sci., 2003, 88, 3168–3175.

23 N. Husing and U. Schubert, Chem. Mater., 1998, 10,3024–3032.

24 A. Venkateswara Rao and G. M. Pajonk, J. Non-Cryst. Solids,2001, 285, 202–209.

25 Published after submission of this article: N. de la Rosa-Fox,V. Morales-Florez, J. A. Toledo-Fernandez, M. Pineiro,R. Mendoza-Serna and L. Esuivas, J. Eur. Ceram. Soc., 2007,27, 3311–3316.

This journal is � The Royal Society of Chemistry 2008 J. Mater. Chem., 2008, 18, 207–213 | 213