Key Engineering Materials Vol. 391 (2009) pp 45-78 online at http://www.scientific.net © (2009) Trans Tech Publications, Switzerland NanoStructured Sonogels Nicolás de la Rosa-Foxa)*, Víctor Morales-Flóreza), Manuel Piñero b) and Luis Esquivias c). a) Departamento de Física Materia Condensada. Facultad de Ciencias b) Departamento de Física Aplicada. CASEM. Universidad de Cádiz. 11510 P uerto Real (Cádiz). SPAIN. c) Departamento de Física de la Materia Condensada. Facultad de Físicas. Instituto de Ciencias de los Materiales de Sevilla (CSIC). Universidad de Sevilla. 41012 Sevilla. SPAIN a b c d nicolas.rosafox@uca.es, victor.morales@uca.es, manolo.piniero@uca.es, luisesquivias@us.es Keywords: Solution, Sol, Gel, Precipitate, Xerogel, Aerogel. Abstract Acoustic cavitation effects in sol- gel liquid processing permits to obtain nanostructured materials, with size-dependent properties. The so-called “hot spots” produce very high temperatures and pressures which act as nanoreactors. Ultrasounds force the dissolution and the reaction stars. The products (alcohol, water and silanol) help to continue the dissolution, being catalyst content, temperature bath and alkyl group length dependent. Popular choices used in the preparation of silica-based gels are tetramethoxysilane (TMOS), Si(OCH3 )4 , and tetraethoxysilane (TEOS), Si(OC2 H5 )4 . The resultant “sonogels” are denser gels with finer and homogeneous porosity than those of classic ones. They have a high surface/volume ratio and are built by small particles (1 nm radius) and a high cross- linked network with low –OH surface coverage radicals. In this way a cluster model is presented based on randomly-packed spheres in several hierarchical levels that represent the real sonoaerogel. Organic modified silicates (ORMOSIL) were obtained by supercritical drying in ethanol of the corresponding alcogel producing a hybrid organic/inorganic aerogel. The new material takes the advantages of the organic polymers as flexibility, low density, toughness and formability whereas the inorganic part contributes with surface hardness, modulus strength, transparency and high refractive index. The sonocatalytic method has proven to be adequate to prepare silica matrices for fine and uniform dispersion of CdS and PbS quantum dots (QDs), which show exciton quantum confinement. We present results of characterization of these materials, such as nitrogen physisorption, small angle X-ray/neutrons scattering, electron microscopy, uniaxial compression and nanoindentation. Finally these materials find application as biomaterials for tissue engineering and for CO 2 sequestration by means the carbonation reaction. 1.Sol-gel nanomaterials processing. In many fields of the natural sciences the boundary between two subjects is difficult to establish, and the frontier between extended solids and nanostructured materials is no exception. The difference between the macro and nano scales is both quantitative and qualitative. It is well established that a material can be considered “nano” when the grain size scale extends from the molecular level (0.1 nm) up to around 100 nm; that is, from a physico-chemical point of view, it extends from the scale of the chemical bond, where its behaviour is governed by quantum mechanics, to the mesoscopic systems (100 nm), at which specific models still are needed, before passing to the macroscopic level. * To who m all correspondence should be addressed: nicolas.rosafox@uca.es All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of the publisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 150.214.245.142-07/07/08,18:04:03) 46 Progress in Sol-Gel Production Nanostructured materials take advantage of properties that derive from confinement effects, larger interface-to-volume ratios, links between wave phenomena and structural features, and where new atomic and macromolecular structures can be generated. The criteria currently used to identify a nanostructured material is the existence of a size-dependent physico-chemical property in such materials, but it is not safe to extrapolate from this up or down in scale. From current work on the development of nanostructured materials the question arises of how the various different properties change as the microstructural scale descends to nanometre dimensions. A nanostructured material is thus recognized because it is formed from the assemb ly of either nanoparticles, nanocrystals or nanolayers, which can be adapted to specific requirements in a dispersion or a coating, functional nanostructures, compacted materials, biological systems, etc. Manufacturing processes have been designed to take advantage of the following effects: a) New physical, chemical or biological properties derived from the grain size scaling. b) New phenomena due to the reduced grain size, where interaction length scales become comparable to the size of the particle, crystal or grain microstructure. c) The generation of new atomic, molecular and macromolecular structures in materials. d) The significant increase of the degree of complexity and speed of processes in particulate systems. All of these characteristics or effects give rise to novel mechanical, optical, electric, magnetic, thermal, chemical and biological properties, but only a small proportion of these properties have been fully identified and quantified [1]. In this chapter we describe one of the strategies for building nanomaterials, known as the sol-gel method. This process is widely used for the preparation of multicomponent nanostructured materials by the hydrolysis and polycondensation of metal alkoxides. From a structural point of view a gel can be considered, in the simplest picture, as a giant molecule which has been formed as a consequence of growth by condensation of polymers or aggregation of particles, but no la tent heat is evolved. This giant molecule extends to the walls of the vessel that contains it, and the coherent solid 3-D network inside the fluid phase is known as a gel. The primary purpose of this method is to produce very specific composite materials, and among the most important of these are nanomaterials. The main processes involved in this method begin at the molecular level; hence it is possible to act on the structure at the very initial stages. Applications of the method include the tailoring of mechanical, optical, electronic, and chemical properties for sensors, bio-composites, structural and other materials, which are used in all kinds of manufacturing industry [2]. The new features in coatings and films include ferroelectric, magnetic and photo chromic properties, electrical conductors and materials with large third-order optical nonlinear coefficients [3, 4]. Another active field is the use of organic- inorganic hybrid materials (OIHM) developed from the pioneering work of H. Schmidt in 1985 [5]. He called some of these materials ORMOSILs (ORganic MOdified SILicate) and others ORMOCERs (ORganic MOdified CERamic). The mechanical properties are dependent on the organic content and rubber- like elasticity can replace brittleness [6]. Advantage is taken of the inorganic component that governs the degree of hardness, brittleness and transparency whereas the density, porosity and thermal stability are governed by the organic polymer. Using the classification given by Mackenzie [7], we are now in the second generation of gels, including new aerogels with improved mechanical properties, termed “aerormosils” [8]. These include hard coatings for plastic ophthalmic lenses in which the transparency is maintained by the small size of the colloidal particles, and the Ormosil coating reduces the permeation by water, oxygen and numerous other molecules [9]. Considerable progress has been made in the field of nano-building blocks by the co-polymerization of different oxide species [10, 11]. Key Engineering Materials Vol. 391 47 2. Ultrasonically-induced nanomaterials. Some of the most important recent developments in sonochemistry have been its application in the synthesis and modification of both organic and inorganic materials. High-power ultrasound can induce a wide range of chemical and physical conseq uences. The chemical effects of ultrasound can be grouped in three main areas: homogeneous sonochemistry of liquids, heterogeneous sonochemistry of liquid- liquid or liquid-solid systems, and sonocatalysis (which overlaps the other two areas). Applications of ultrasound to materials chemistry are found in all of these categories [12]. Physical effects of high- intensity ultrasound, which often have chemical consequences, include enhanced mass transport, emulsification, bulk thermal heating, and a variety of e ffects on solids. In all of these cases the process is conducted by means of acoustic cavitation effects. When the sound wave propagates in a medium it travels via a series of compression and rarefaction stages induced in the molecules through which it passes. At sufficiently high intensity the rarefaction cycle may exceed the attractive forces of the molecules and cavitation bubbles will form. Such generated bubbles grow by a process known as rectified diffusion i.e. small amounts of vapour (or gas) from the medium enter the bubble during its expansion stage but this vapour is not fully expelled during the compression step. The bubbles continue to grow over a few cycles until reaching an equilibrium size for the ultrasound frequency applied. It is the fate of these bubbles to collapse in succeeding compression cycles, at which point they act as a “hot spot” [13], as shown in Figure 1. 150 Bubble Radius (m) Collapse 100 Growth Cooling 50 "hot spot" Cavitation seeds 0 0 100 200 300 400 500 600 Time (ms) Figure 1: Ultrasonic cavitation can produce bubbles in liquids. Along this process the bubbles oscillate induced by the pressure cycle of the sound wave. Eventually they undergo a violent collapse, which generates hot -spots of extremely high pressures (1000 At m) and temperatures (5000K), in lifet imes of less than 2 s. Thus the chemical consequences of the use of high-power ultrasound do not arise from an interaction of acoustic waves and matter at a molecular or atomic level. Instead, in liquids submitted to high-power ultrasound, acoustic cavitation (a process which takes place in three steps: the formation, growth and collapse of bubbles) provides the primary mechanism for sonochemical effects [14, 15]. During cavitation, bubble collapse produces intense local heating, high pressures, and very short lifetimes; these transient, short-lived localized hot spots provide enough high-energy to generate the chemical reactions. As described in detail by Suslick et al. [16], these hot spots can reach temperatures as high as 5000K, pressures of about 1000 atm, and heating and cooling rates above 1010 K/s, all in a time scale of the order of picoseconds [17]. 48 Progress in Sol-Gel Production Thus, cavitation hot spots can be used as chemical nanoreactors because they serve as a means of concentrating the diffuse energy of sound into a unique set of conditions to produce unusual materials from dissolved (and generally volatile) precursors. 3. Sonogels As already stated, the sol- gel process is based on the possibility of forming the disordered oxide network by various reactions in a liquid phase, followed by solvent elimination. One of the approaches to forming the initial gel is the “alkoxide method”. The precursors used in this case are metal alkoxides M(OR)n , where M is a metal and R an alkyl group. These compounds react with water, in the presence of any acid or basic catalyst, and undergo hydrolysis and polycondensation reactions, which lead to the formation of the metal oxide. The polymerization leads first to a colloidal dispersion of particles, the aggregation of which will form the gel when the percolation level is accomplished, forming a coherent solid structure immersed in a fluid medium. To obtain nanostructured metal oxides using the action of high power ultrasounds to promote the chemical reactions, the “sonogel” method can be used. This procedure avoids the need to use a common solvent for the alkoxide and water mixture [18, 19]. The acoustic wave effect on the liquid mixture acts in two ways: firstly, the alkoxide/water mixture is emulsified and diffusion between the two phases takes place; secondly, chemical reactions take place in the small bubbles (hot spots) generated by cavitation. The two acoustic processes take place inside the bubbles dispersed throughout the liquid, and the rapid compression/decompression of the ultrasonic wave (20 kHz) causes the bubble radius to oscillate around some equilibrium size and finally to collapse. When high-power ultrasound is used (100 W output power), extremely hot spots are formed in the liquid in its three discrete stages: nucleation, growth, and implosive collapse [20, 21]. The extreme conditions generated on bubble collapse lead to the production of excited states, breakage of chemical bond, and formation of free radicals. Thus, three areas can be identified in a cavitation system. The center of the hot spot is where the primary chemistry involved in atomic and radical recombination takes place [33]. The surrounding heterogeneous liquid is relatively unaffected, although active molecular species generated inside the bubble may diffuse out and react with reagents in the liquid. At this point, the interfacial region has very large gradients of temperature, pressure, surface tension, electrical field, shear stresses and rapid motion of molecules, leading to efficient mixing [22, 23]. It is interesting to note that the presence of foreign pa rticles enhances the cavitation because they act as nucleation sites. The temperature inside the hot-spot increases suddenly as a consequence of the adiabatic bubble collapse, and has been estimated at several thousand degrees Kelvin [24, 25]. The alkoxide-water mixture involved in sonogels can be classified, under Luche‟s rule [26], as a heterogeneous system in which ionic reactions are stimulated inside the cavitation bubbles; therefore the product of the reaction will be the same as it would be in the absence of ultrasound (as by classic or conventional processes). The application of ultrasound to the synthesis of organosilane and organosiloxane precursors has been studied by Boudjouk [27] and Price et al. [28]. 3.1. Pure Silica Sonogels This section accounts for the systematic studies carried out on silicon alkoxide+water mixtures. The aim of such work was to elucidate the differences between sonogels and gels obtained by the conventional (classic) method. As the first parameter involved is the absence of alcohol, one can expect to obtain denser sono-sols due to a lower dilution and, in consequence, a fast polymerization process. Key Engineering Materials Vol. 391 49 Tarasevich first reported observation of the TEOS/water mixture reaction under the action of high-power ultrasound in 1984 [29]. He reported that the reaction shows an intense increase in temperature with the release of alcoholic vapours. He obtained a homogeneous and transparent solution in a few minutes. Later, Zarzycki‟s group in France and Esquivias‟ group in Spain began intensive collaborative work to establish the consequence of ultrasound influence on the textural characteristics of “sonogels” [30, 31, 32] and derived materials [27, 28]. The combined effect of sonocatalysis and organic- inorganic hybrids has opened new routes to nanostructured materials [15]. Pure silica sonogels have also been studied in depth by Donatti, Vollet and Ibañez Ruiz, in Brazil. They report a systematic study on the hydrolysis of TEOS/water and TMOS/water liquid mixtures under the action of ultrasounds, proposing a dissolution reaction model based on dynamic calorimetric measurements [33, 34, 35]. The results are thermally activated hydrolysis with apparent activation energy of 36 kJ/mol. From this, the action of ultrasound can be explained as follows: ultrasound forces the dissolution and the reaction begins; the products (alcohol, water and silanol) help to continue the dissolution, with the reaction being dependent on catalyst content, bath temperature and alkyl group length [36, 37]. 3.1.1. Sonocatalysis. The corresponding mixture of the silicon alkoxide and acidified water (pH < 1) was placed in a double-volume beaker. No reaction was observed when a neutral or basic catalyst was used. A twophase system, as shown schematically in Figure 2, could be discerned in the beaker. At this point the tip of the ultrasonic device was immersed some millimetres into the liquid. The ultrasonic waves (20 kHz, 60-100 W) are then produced by an electrostrictive device equipped with a stainless steel horn terminated with a titanium tip Popular choices used in the preparation of silica-based gels are tetramethoxysilane, Si(OCH3 )4 , and tetraethoxysilane, Si(OC 2 H5 )4 , known as TMOS and TEOS, respectively. Since these compounds and water are immiscible, a common solvent (usually methyl or ethyl alcohol) has to be added to obtain an initially- homogeneous liquid. However, no solvent is necessary when high-power ultrasound is applied to the liquid mixture. The chemical reactions involved can be summarised as follows: 50 Progress in Sol-Gel Production PDMS TEOS H 2O Inmiscible starting precursors Ultrasounds exposure hydrolysis Si (OR) 4  4 H 2 O )))us Transparent hybrid organic-inorganic sonosol Si (OH ) 4  4 ROH polycondensation a) Si (OH ) 4  Si (OR) 4 )))us 2SiO2  4 ROH b) Si (OH ) 4  Si (OH ) 4 )))us 2SiO2  4 H 2 O Figure 2: Step procedure for h igh-power u ltrasounds treatment of an alko xide -PDMS-water mixture. where R is –CH3 in the case of TMOS, and –CH2 -CH3 for TEOS. Hydrolysis and polycondensation can be accelerated or slowed down by employing an appropriate acid catalyst. These reactions proceed simultaneously and are generally incomplete. Depending on the amount of water present, hydrolysis may go to completion or stop while the metal is only partially hydrolyzed. On the other hand, the polymerization can be made reversible because silica solubility increases for small particles or small negative curvature (spheres or “necks”). This process gives a strengthened network because the small particles dissolve and are deposited on the larger ones and in the “necks” between two larger particles. When several different cations are used to form mixed-oxide networks, an initial complexing step may be required. When the alkoxide precursors have different hydrolysis rates (e.g. Al, Zr, Y or Ti, compared with Si), prehydrolysis of the alkoxysilane is preferred [27]. After a complex sequence of polymerization, sol formation and gelation, a microporous gel with high surface area is formed, constituted by small particles; its formula is approximated by: (MO)x (M’O)x’(OH)y (OR)z The radicals -OH and -OR account for reaction by-products that can, reasonably easily, result in a system that yields a complex three-dimensional oxide -M-O-M‟-O-M- polymer which preconfigures the network of corresponding oxide glasses. Key Engineering Materials Vol. 391 51 In the case of copolymerization between (organic) PDMS and (inorganic) TEOS, the reaction proceeds as follows, in order to form the hybrid organic- inorganic network: The n value determines the polymer chain length and leads to the desired the physical properties by the interconnecting of silica particles through the polymer chain cross- links. Me Me 2 Si(OH)4 + HO OH HO Si OH O Si O Si Me Me Me Me Si Me O OH n OH Si Me O n Si OH + 2H2 O OH Continuing with the effects of ultrasound, the ultrasonic energy delivered to the solution can be estimated by the temperature increase of a fixed volume (V) of water during the sonic treatment, and consequently can be calculated by the relationship: dQ dT W   mc dt dt The mass of water and its specific heat are m and c respectively. Then the energy delivered by the ultrasound will be: W t Us   K  t ( J  cm  3 ) V where t is the sonic treatment time in minutes and K is a constant characteristic of the apparatus. Many factors can affect these K values, such as room temperature, thermostatic bath temperature, catalyst content, molar ratios of water/alkoxide or solvent/alkoxide, beaker diameter and volume, depth of horn tip immersion in the liquid solution and tip diameter, among others. Once the liquid mixture reacts and a homogeneous solution is observed, the energy dose is taken as the threshold of the hydrolysis reaction. Also there is an energy dose at which the liquid gels “in situ”; that is, gelation takes place in the beaker during the ultrasonic treatment. At this point, a transparent soft solid gel, with the tip hole on top, can be observed. Between these two limits one can tune the gelation at any particular dose within the energy range. As a reference, some of these energy doses, Us (min) and Us (max), are shown in Table 1 for different precursors and contents. In this way, if the minimum dose is related to the hydrolysis reaction, as can be seen in Table 1, the expected decrease of the hydrolysis rate due to the dilution is compensated by the action of ultrasound. At the other limit, the maximum dose that can be re lated to the polycondensation rate produces greater reactivity of the TMOS. However, this effect is reduced by the presence of the PDMS organic polymer. This fact could be a consequence of the different functionality of the precursors, f=4 for TEOS and TMOS and only f=2 for PDMS. Once the sonosol has been exposed to the corresponding ultrasonic energy dose, the liquid sonosol is kept in a hermetic container and allowed to gel at the chosen temperature. Since the 52 Progress in Sol-Gel Production sonosol is a low- viscosity liquid, it can be cast in a mould with a selected shape. In all cases, for the silicon alkoxides, a homogeneous and transparent solution is obtained. This property is due to the fine dispersion of the colloidal particles formed, which begin to aggregate to produce the so-called nanostructured sonogel. Table 1. Th reshold and limiting of the ultrasonic energy dose. 3.1.2. Sonogel Gelation The point at which the sol passes from a viscous fluid to an elastic solid is taken as the gelation point. At this point the viscosity increases abruptly by several orders of magnitude. The gelation time is taken as the time interval between the end of the ultrasonic treatment and the abovementioned transition. This transition can be estimated visually when the solution surface is no longer horizontal when the container is tilted. Figure 3 shows the results of the gelation time tG as a function of Us for a molar ratio of H2 O/TEOS=4, at several mixture temperatures. In fact, the temperature, as well as the increase in the ultrasonic energy dose, both activate the gelation process. As can be observed, two regimes are present, separated by a level of Us = 600 J cm-3 . This point indicates the rapid increase in the polycondensation rate. As a reference for classic gels, the gelation time is of the order of several days for similar compositions, and is dependent on solvent content. From the experimental values of Figure 3 it is possible to establish the thermal behaviour of the gelation process. The plot of tG vs. 1/T is depicted in the inset of Figure 3 for different ultrasonic energy doses. Assuming an Arrhenius behaviour, the corresponding linear fit gives an activation energy of 50–70 kJ mol-1 TEOS, which is of the same order of magnitude as the results reported by Tiller et al. [38], 45 kJ mol-1 TEOS. The pH dependence of the activation energy, that is, the activation energy obtained for sonogels, agrees with the results reported by Coudurier et al. [39], of 61 kJ mol-1 for the condensation process, and 63 kJ mol-1 for the aggregation process. This indicates that the polycondensation in pure sonogels is achieved mostly in the early stages, producing a more reticulated structure than in classic processes. Key Engineering Materials Vol. 391 53 US Figure 3: Evolution of the gelation time fo r water/TEOS = 4 pure sonogels as a function of the ultrasonic energy dose evaluated at different temperatures. Inset evolution of the gelation time as a function of 1/T for d ifferent ultrasonic energy doses, fro m top to the bottom [32]. By using a simple model of the polycondensation ratio (the prod uct of a second order consecutive reaction [40, 41]), the concentration of polycondensed species results in a sigmoidallike growth. The shape for every particular case depends not only on the standard parameters for gel preparation (pH, water ratio, precursor, etc.) but also on the ultrasound power of supplied by the device and the time applied, this is to say energy. Figure 4 are plot of the polycondensation ratio evolution c(t) during the whole process for three different energies furnished. During insonation the concentration of polycondensed species follows an evolution c(t) Ultrasound assisted reactions c(t) Silent reactions tG1 Ultrasound assisted reactions Silent reactions c(t) Ultrasound assisted reactions tG2 cc cc t cc t Silent reactions tG3 = 0 t Figure 4: Polycondensation ratio evolution c(t) during the whole process for three different ultrasound energies supplied. Similar results are obtained for a sol of 90% wt TEOS + 10% wt PDMS and a molar ratio H2 O/TEOS = 2, in which the two regimes can also be observed. However in this case hybrid materials are separated by a level of Us = 450 J cm-3 , accounted for by the influence of the PDMS organic polymer. This effect may be due to the chain breaking as a consequence of the ultrasound cavitation, preventing the formation of rings. In fact, in the first regime, pure sonogels are obtained twice as fast, and in the second, three times as fast as the hybrid gels. Since the reactions in the PDMS/TEOS hybrid system are more complex than those in pure silicon alkoxides, the activation energy is difficult to calculate. Mackenzie and Hoshino report a value of 60 kJ mol-1 for a 20% wt PDMS Ormosil [42], which is similar to that of pure silica sonogels, revealing the relatively minor influence of the temperature in this process, compared with that of the ultrasound applied. 54 Progress in Sol-Gel Production 3.1.3 Sonogel drying Two methods are used to dry the wet sonogel: 1) A process of slow evaporation or the use of a c hemical additive (DCCA), which give a very shrunk solid called a sono-xerogel, in which the syneresis and permeability collapse the gel microstructure. 2) Venting off a supercritical solvent, which produces a solid known as a sono-aerogel, in which the original microstructure of the gel is preserved. There are several strategies for preventing theifferential stresses between adjacent pores that provoke fracture of the solid if the tensile strength of the material is exceeded : one is to add a chemical additive (DCCA) to the precursor liquid mixture, before gelling, to control the drying [43]. The additive most commonly used as a DCCA is the protic solvent formamide (HCONH 2 ) [44]; the bonding of hydrogen to hydroxyl ions reduces the catalytic activity and influences both the hydrolysis and condensation reactions. The influence of formamide as the DCCA is a consequence of its low vapour pressure and surface tension. The reduction in the capillary pressure is accomplished by the formation of a film on the pore wall, reducing the contact angle and the solvent evaporates very slowly, providing a plasticizing effect. The success of this method can be attributed to the coarsening of the microstructure and the strengthening of the gel network. Other solvents that are effective in this application include dimethyl formamide (DMF) [45], oxalic acid and glycerol [46], among others. The other strategy for maintaining the original microstructure of the gel as it exists at the gelation point is by supercritical drying. When the critical point of one liquid is surpassed, there are no distinctions between liquid and vapour phases and their densities become equal (as a supercritical fluid); then capillary pressure stops and the permeability decreases. To achieve this, the wet gel is placed in an autoclave and the temperature and pressure are raised up to the critical point of the corresponding solvent, taking care not to cross the liquid-vapour boundary; then the solvent is vented isothermally. The critical points of various solvents are shown in Table 2. The gel obtained is a solid network with the pores filled by air, termed an aerogel. Table 2. Crit ical points of some common solvents The experimental process is crucial for maintaining the solid in one monolithic piece, so the ramp or gradient applied to raise the temperature must be as slow as possible (< 1ºC/min) because of the differences in the thermal expansion coefficients of the liquid and solid. This difference gives rise to pressure gradients and can cause the solid gel to crack. It is also common practice to apply the extra pressure of an inert gas from the beginning, in order to avoid crossing the liquid-vapour boundary. In the case of a volatile solvent, one disadvantage is the high temperature necessary (Table 2); the use of CO 2 permits the process to be performed at near ambient temperature. This process was first applied by T. Woignier [47] and Tewari et al. [48]. Supercritical drying needs the pore liquid (alcohol+water) to be replaced by liquid CO 2 ; a complete solvent exchange is necessary because capillary compression is produced by the immiscible boundary between water and CO 2 . Key Engineering Materials Vol. 391 55 The main application found this method is in the preparation of organic- inorganic hybrid aerogels because it prevents degradation of the organic polymer by temperature. Another approach is to freeze the pore liquid and sublime the resulting solid under vacuum, which is termed freezing-drying [49]; though this is widely used in the preparation of foods, it does not permit the preparation of monolithic pieces. 3.2. NanoStructure of dry silica sonogels. The special characteristics of sonogels after drying is that present a particulate structure 50, contrary to gels obtained by hydrolysis of metallorganic compounds under acid catalys t without applying ultrasound, which are filamentous. We have shown that can be represented of a hierarchic arrangement of agglomerated of elementary particles of 1- 2 nm size forming agglomerates 4- 6 nm size and, in some cases, aggregates of these aggregates level [51, 52]. Sonogels, in the form of xerogels and aerogels, are examples of solids with both microporous and mesoporous structure, respectively. For our samples these features can be seen in Figure 8; that is, a type I isotherm for the xerogel sample and type IV isotherm for the aerogel. There is adsorption at low pressure in both cases, but not further in the case of the xerogel, and capillary condensation with hysteresis in the desorption branch in the case of the aerogel. In line with the previously reported results, the xerogel is an entanglement of elementary particles of 2.8 nm radius forming a microporous network of 1.1 nm pore radius, as indicated in the t-plot in the inset of Figure 8. The aerogel, however, is formed by particles of 1.7 nm radius forming a mesoporous network of 2.3 nm pore radius. The pore size distribution of these samples shows a narrow peak in the micro- and the mesopore regions, respectively. The results are explained by the differences between the collapsed microstructure of the xerogel and the original nanostructure sonogel in the case of the aerogel. 800 3 Vads (cm /g) 200 3 -1 Vads (cm g ) 600 1500 150 1000 100 500 400 50 0 4 200 6 8 10 12 14 16 18 0 20 t (Å) 0 0,0 0,2 0,4 0,6 0,8 P/P0 Figure 8: N2 -physisorption isotherms from xerogel (type I-squares) and aerogel (type IV-ciscles) made of silica, full symbols is the adsorption branch and open symbols to the desorption one. Inset shows the t-plot of these gels and the continous line corresponds to a nonporous silica sample. 3.3. Cluster Model: random-packing of spheres. Given the above results, some attempts have been made to depict such hierarchies on several levels, using models constructed with the Monte-Carlo technique [59, 60, 61]. The approach is based on comparing the pore size distribution of sonogels with that of a random close-packed hard sphere model studied by Bernal and Mason [53], Scott [54] and Finney [55], and developed for this application by Zarzycki [56] and Rodríguez-Ortega and Esquivias [59, 60, 57]. Essentially, the structural approach consists in building models of the solid phase of a gel by depicting its structure 56 Progress in Sol-Gel Production as a collection of packed spherical particles. Then, we created a catalogue of pore size distributions [60] according to different conditions of particle coordination number and compaction according to the features obtained from its pore space. The pore size distribution (PSD) of the models are represented as a function of the variable K = r/R, where r is the radius of the largest inscribed sphere in the interstice and R the radius of the particles. Finally, pore distributions can be compared to the experimental meas urement. A logarithmic scale for K is used to make easier to fit the experimental data, by simply sliding it along the K axis until the position considered to give the best fit is reached. Given the maximum, the fit allows the particle size of a mono-disperse system to be calculated from the maxima of the pore size distributions. From the distribution of each of these contributions to the experimental distribution, successive sizes and local densities of hierarchic distribution may be deduced [66. Data on the pore volumes associated with different hierarchical levels, size of aggregates, the local density of the i-th aggregation level, and packing of the successive levels can be obtained. Some of these models and their pore size distributions (PSD) are shown in Fig. 9. Figure 9: Mesopore fractions of SG2(sono-aerogel TEOS) and STM S (sono-aerogel TMOS) fro m Hg porosimetry fitted with specific built models (continuous line). The insets correspond to the N2 physisorption which fit with the models. Bi-d imensional representations of such models (SG2 in the middle and STM S in the right) are also included [52]. It is generally accepted that the Cluster-Cluster Aggregation regimes (DLCA – DiffusionLimited Cluster Aggregation, RLCA – Reaction- Limited Cluster Aggregation) describe quite well the typical structures obtained via sol-gel. One of the goals pursued most by researchers is to reproduce the formation and growth processes of aerogels, using the RLCA or DLCA algorithms or some modifications of these [58, 59]. Scherer and col. [60] used structures generated with modified DLCA algorithms, characterising them by their fractal dimension, to achieve the power law exponent, and they have presented several models to explain the relationship between structure a nd mechanical properties [61, 62]. Since then, Woignier and col. have introduced a new technique for characterising these porous systems [63, 64], and conclude that pore size distribution and hydroxyl content are relevant for understanding the mechanical p roperties of these materials [65]. In a previous study, Woignier and Phallipou proposed one approach starting from a cubic structural Key Engineering Materials Vol. 391 57 model [66] and for a rigid assembly of cohesive spheres [67]. The Cluster Model that is introduced here has also been applied as an initial approach to the study of mechanical properties [68]. The diagram shown as Figure 10 explains the process of construction. The algorithm works as follows: first we place one elementary sphere in the centre of our system. Then we randomly p lace as many other spheres as are needed to cover fully the surface of the first one; this produces the first random shell. Every sphere has to be in contact with at least one other; that is, the sphere centres must be at a previously defined distance from each other. One can build as many shells of randomly-placed spheres as are required. This shell then constitutes the basic aggregate of the first hierarchical level, and its diameter is measured. This diameter will be taken as the diameter of a secondary sphere. Figure 10: a) Diagram of the Cluster Model algorith m. b) Sketch of a Cluster Model, corresponding to the simulat ion box cropped fro m system 2 of table 5 [70]. The next level in the hierarchy is constructed in the same way, treating the ba sic aggregate as if it were an elementary sphere to construct the second level with secondary spheres. After constructing this new aggregate, each secondary sphere forms a new basic aggregate, to obtain a two- level hierarchically-ordered assembly of randomly-packed spheres. This process can be repeated as many times as required. Typical values of our models are 1000 to 60,000 particles organised in 2 to 4 shells of randomly-packed spheres and 2 or 3 hierarchical levels; their contact distance L usually is found in the interval 0.85d < L < 1.0d, d being the particle diameter. A sketch of a cluster model can be seen in Figure 10-b rendered using the free software POVRay [69]. The system illustrated corresponds to the simulation box cropped from model #2, and has around 3000 particles [70]. We have built several cluster models that represent the microstructure of various real systems. Table 5 gives the characterization parameters of some real systems and their corresponding models. Systems #1 and #2 correspond to two different silica aerogels prepared from TEOS, and were characterized by the analysis of the N 2 adsorption/desorption isotherms [71]. These results confirm the good agreement between the Cluster Model and the actual nanostructured sonogel. 58 Progress in Sol-Gel Production System 1 (pure silica aerogel) Apparent density: 0.83 g/cm 3 Specific surface: 387-407 m2/g Specific porous volume: 0.73-0.74 cm3/g Model 1 Apparent density: 0.80 g/cm 3 Specific surface: 384 m 2/g Specific porous volume: 0.72 cm 3/g System 2 (pure silica aerogel) Elemental Sphere radius: 1.2 nm First aggregate radius: 4.5 nm Specific surface: 640 m 2/g Model 2 Elemental Sphere radius: 1.1 nm First aggregate radius: 4.5 nm Specific surface: 612 m 2/g Table 5: Structural parameters of several real systems and the g eometric and structural parameters of their corresponding counterpart cluster models. 4. Nanocomposites from sonogels. Nanocomposites are materials in which the constituents are mixed on a nanometer scale to give properties that are superior to conventional microscale composites. They can be synthesized by a variety of techniques including the sol- gel process and, more specifically, by using ultrasound. This produces cavitation in the liquid sol, which in general leads to the matrix phase, as an additional parameter for controlling the mixing at the nanometer length scale with the secondary phase. Considerable research has been devoted to this type of material since 1989, covering different phases, designs, morphologies and technological uses, to produce ceramic [72, 73] and optoelectronic [74, 75] materials and polymers [76]. Hybrid aerogel nanocomposites made from sonogels, with the incorporation of a second and active phase, have found applications as bioactive materials [77, 78] and for CO 2 sequestration [79, 80]. 4.1. Nanocrystals embedded in a sonogel matrix. Adopting the sonosol liquid stage at the early stages of the process also favours the formation of nanocrystals with uniformity in composition, shape, size, internal structure and surface che mistry, which are essential characteristics for designing and controlling the behaviour of materials. In this field, materials containing semiconductor nanocrystallites immersed in a dielectric matrix have attracted great interest because they display third-order optical non- linearities and dramatic changes in the optical absorption spectra with respect to the conventional bulk semiconductor [81]. A II-VI semiconductor such as CdS has been studied in detail because it precipitates easily in a silica gel matrix to give good optical transmission. The sonocatalytic method has been proven suitable for preparing silica matrices for fine and uniform dispersion of CdS quantum dots (QDs) [82]. The silica sono- xerogels were prepared by ultrasound-assisted hydrolysis of a mixture of tetramethoxysilane (TMOS):acid water:Formamide, in the molar ratio 1:10:3, using an ultrasound dose of 64 J·cm-3 . Different amounts of Cd(NO 3 )2 (1, 3 and 10 % weight related to the total silica content) were added under mechanical agitation to the sono-solution before gelation. Once the resulting gels had been left to age at room temperature, H2 S gas was diffused by thermal decomposition of thioacetamide (TAA). Small CdS crystals are then produced by precipitation inside the silica matrix. The quantum confinement behaviour is manifested by the characteristic blue-shift in the UV-Vis spectrum (Figure 11-a left axis) in relation to the bulk optical absorption band of CdS with a bandgap energy of 2.53 eV (490 nm wavelength). This behaviour is consistent with the Éfros-Éfros model [3] which gives the absorption threshold energy E dependent on the inverse square of particle mean radius R as follows: Key Engineering Materials Vol. 391 E  Eg  59 2 2 2R2 5 3 6x10 CdS1 CdS3 CdS10 CdO 5 5x10 5 4x10 2 5 -d 3x10 5 2x10 1 Photoluminiscence (cps) where ħ is the Planck constant,  is the electron effective reduced mass and Eg is the bandgap energy for the bulk crystal. These energy values can be obtained from the position of the optical absorption band of the different samples. 5 1x10 0 0 300 400 500 600 700 800 900  (nm) Figure 11: Optical density (left axis) of the outlined samples together the CdO one included as transparent reference (bottom). Photoluminescence (PL) signal (right axis) of the same samples [82bis]. Figure 11 also includes the photoluminescence (PL) yield (right axis) in which two bands can be observed. The higher-energy band is related to the intrinsic recombination mechanism. There may also be weak transitions for dots with RaB=3.2 nm in CdS (aB being the bulk exciton Bohr radius), because kinetic energy terms are dominant over Coulomb effects, and this produces the observed line broadening. On the other hand, the lower-energy band is attributed to the recombination of trapped carriers. A red-shift of the PL emission relative to the absorption band can also be observed, which is crystal size-dependent: the red-shift increases as the particle size decreases. The origin of this effect seems to be in a distortion of the crystal lattice after the transformation of the CdO (cubic) into CdS (hexagonal) at 150ºC, which provokes compressive strains. This redshift/distortion influences the intrinsic band due to the sulfur ion vacancies acting as potential hole traps. The observable differences in the PL spectra also inform about the crystal surface traps; the CdS3 sample seems to have lower trap states than the CdS1. However, the CdS10 sample shows an overlapping of the two bands indicating a significant distortion of the lattice. In the light of these size-dependent effects (blue- and red-shifts), the lack of correlation between crystal size and CdS content must be attributed to the slight difference of the silica network pore structure which affects diffusion of the H2 S gas. The PL yields were measured by using the third harmonic (3.49 eV) of a pulsed Nd:YAG laser (6 ns pulse at 10 Hz repetition rate). Using a cylindrical lens, the laser beam is focused on the edge of the sample to form a narrow rectangular strip (50 m to 2 mm long and 20 m wide) [82, 83]. The amplified luminescence (AL) at right angles to the direction of the strip is measured for different strip lengths. In all cases the PL yield measurements were taken at atmospheric pressure in air at room temperature. Use of the variable strip length (VSL) method is reported in [84]; in this 60 Progress in Sol-Gel Production method the PL yield along the focus axis is related to the optical gain by the relationship: I I AL  SP e gL  1 , g   where IAL and I SP are the amplified and spontaneous emission, respectively. The net gain coefficient is g and L is the stripe length. The PL yield increases in a superlinear fashion for the intrinsic band, accounting for a net gain coefficient as can be seen in Figure 12. The gain is spectrally broad with a steeper decrease on the high-energy side and a long tail stretching to lower energies; in all cases the optical gain reveals the stimulated emission from the CdS nanocrystals as QDs. CdS1 CdS3 -1 gain coeff. (cm ) 100 50 0 400 450 500 550 600 650 700  (nm) Figure 12: Optical gain spectra for the filtered intrinsic band (Schott BG-14) of the outlined samples. The TEM micrograph in Figure 13 confirms the fine and homogeneous distribution of the CdS nanocrystals immersed in the porous silica matrix, with an average crystal size of 4.6 nm, as indicated by their size distribution in the inset of Figure 13. Figure 13: TEM micrograph of CdS/SiO2 co mposite corresponding to a 10 wt % Cd S doped SiO2 gel. The corresponding particle size distribution is shown in the inset [74]. Key Engineering Materials Vol. 391 61 The electronic and optical properties of IV-VI semiconductors such as PbS have also been topics for extensive research in the last three decades. Their narrow band-gap permits strongly quantum-confined excitons (electron- hole pairs) to be created when microcrystallites are smaller than the bulk exciton Bohr radius (aB=18 nm for PbS) [4,85, 86]. This is the case for PbS crystallites presenting a particle size smaller than 18 nm, with a threshold of the optical absorption band at 0.41 eV for the bulk material. Many approaches have been explored for the formation of small PbS nanocrystals and their incorporation into solid and transparent matrices [87, 88, 89], and one approach is to use the sol- gel process combined with colloidal chemistry [90, 91, 92]. This technique prevents agglomeration and allows control of the particle size and size distribution; although the small precipitates produced are unstable, surface-capping methods [93] may help to overcome this drawback. Gel matrices and films have been used as hosts for semiconductor quantum dots since more than ten years ago, mainly because low temperatures are required to create the network. However, they present a high porosity which makes them especially active for humid condensation and oxidation, thus favouring subsequent crystal growth. In spite of these inconveniences, this topic arouses interest and recently several studies have been published [94, 95, 96, 97, 98] . Oxidation could be limited by sealing pores and thus reducing the spec ific surface area. One of the methods for achieving this is by using organic polymers in the gel formation; this allows mechanically- improved silica gel matrices to be synthesised via copolymerisation with silicon precursors, named ormosils. These OIHM present properties that make them especially attractive for optical applications [99, 100]. Using this procedure PbS quantum dots stabilised with surface capping agents (SCA), incorporated into rigid transparent SiO 2 sono-ormosils, have been successfully synthesized for nonlinear optical applications. Table 6 gives the textural parameters evaluated from N 2 isotherms. Sample SB ET (m2 ·g -1 ) Vp (cm-3 ·g -1 ) * ρa (g·cm-3 ) **ρ Geom. (g·cm-3 ) Undoped Matrix 616 0.3237 1.28 1.29 ± 0.05 677 0.4830 1.07 1.03 ± 0.05 PbS 542 0.4676 1.09 1.08 ± 0.05 Content 414 0.2101 1.56 1.55 ± 0.05 * 1/pa = Vp + 1/ps ** parallelepiped casting Table 6.- Specific surface (SBET), Pore volume (Vp ), and apparent density (ρa) calculated from N2 physisorption data. The last column shows the apparent density (ρ) values evaluated from geometrical measurement. A decrease in the specific surface area with an increase in the PbS content is observed. This behaviour could be due to differences in pore sizes as a consequence of the steric effects derived from the non-bonded 3- mercaptopropyl groups, since they are electrostatically opposed to the silica network. This assumption is confirmed by the good agreement existing between the apparent densities calculated from the porous volume and from the geometrical dimensions. Bulk densities (s ), were calculated taking into account the corresponding percentages by weight in the composite (SiO 2 )1-x-(PbS)x , by using 2.2 gcm-3 for vitreous silica and 7.5 gcm-3 for PbS. Finally, the average pore size values in the last column corroborate the fine porosity of the host silica matrix. An HRTEM micrograph from PbS/SiO 2 sample is shown as Figure 14-a. Spherical and/or cubic morphology of particles, and lattice planes in some cases, can be clearly seen. A mean diameter of around 10.5 nm with a standard deviation of 1.2 nm was obtained from the distribution plotted in 62 Progress in Sol-Gel Production Figure 14-b. In such a situation, the SCA molecules passivate the Pb surface sites on the subcritical PbS crystals and prevent their oxidation. Surplus uncoordinated S sites act as potential surface traps which produce important steric effects on the fine silica network. The final size of the PbS particle is then controlled by adjusting the relative concentrations of SC A and sulphide [96, 101]. Moreover, the ormosil matrix surrounding the PbS crystals prevents their subsequent growth and oxidation. Figure 14: (a) HRTEM micrograph of the PbS/SiO2 sample (b) Corresponding particle size d iameter d istribution [76]. In the structural analysis an additional measurement was made by means of Small Angle Neutron Scattering (SANS), at room temperature, using a neutron wavelength of 6.02 Å, at the Berlin Neutron Scattering Center (HMI) in Berlin. The corrected SANS intensity c urves are shown in Fig. 20 [102, 103]. In all cases, two regions can be discerned, separated by the well-defined crossover point around q=0.1 nm-1. For high q- values a wide plateau is apparent, which is characteristic of fine well-shaped scatters homogeneously distributed, typical of a material with fine porosity. In contrast, the increased signal at low q- values is characteristic of larger polydisperse particles, as could be produced by the polymer cross- links or/and the PbS nanocrystals (6.5 nm size). The scattering from the undoped matrix is also included, inset of Fig. 20. In the absence of crystals, a smooth broad peak develops at q = 0.5 nm-1 , which accounts for the short-range order of the homogeneous pore distribution. It is well-known that crystals in gels grow by a diffusion-controlled process. In such a situation, the concentration gradient around the crystal depletes a surrounding volume larger than the crystal in which no other crystals are present [104]. In accordance with this understanding, the experimental intensities were fitted by the two-correlation model proposed by Debye and Bueche [105, 106] where the scattering cross-section behaves as:  q2 a 22  A1 d  (q)   A 2 exp     d 4 1  q2 a12 2   (1) where the correlation lengths a1 and a2 represent the medium-range and short-range fluctuations of the scattering length density, respectively. The parameters A1 and A2 are related to the corresponding volume fraction of each phase. Eq.(1) can be considered as the form factor P(q) of the heterogeneities, then the total scattering cross-section becomes I(q)=P(q)S(q),  and S(q) being the particle number density and the structure factor, respectively. The structure factor S(q) was evaluated assuming the Percus-Yevick hard-sphere model, and using the expression of Ashcroft and Lekner [107]. This interpretation of the SANS patterns indicates the existence of a depleted region around each PbS crystal, where the growth of another crystal is inhibited [111].   Key Engineering Materials Vol. 391 63 -1 -1 d/d (cm str ) 10 1 -1 d/d (cm str ) 10 -1 undoped matrix 0,1q(nm-11) x3 1 x1.7 x1 PbS content 0,1 -1 q(nm ) 1 Figure 15. SANS scattered intensities fro m increasing PbS content samples. The solid lines correspond to the nonlinear leat-square fit fro m I(q). Inset shows the undoped matrix samp le. Sample -detector distance was set at 1, 4 and 16 m with a λ = 0.6 n m as neutron wavelength. Scattering vector modulus is q = (4 π sin θ/2)/ λ, θ being the scattering angle. So me contributions of the incoherent scattering comong fro m the ormosil matrix can be noticed at the end of the curves (q >2 n m-1 ), this fact mask probably the Porod behavior in this q-reg ion (I α q -4 ) [111] The structure factor accounts for the interparticle interference terms which could be produced either by crystals or pores. The results of such a fitting are shown in Table 7 and the fitted curves from I(q) are drawn in Figure 15. The undoped matrix does not fit (inset Figure 15) this twofold correlated structure, as can be deduced from the noisy signal at low-q. Thus the observed increase of intensity must be due to the local heterogeneities produced by the polymer cross- links between chains [108] with the silica network at the boundary of the SCA molecules surrounding the PbS nanocrystals. From Table 7, it can be stated that the crossover point at q=0.1 nm-1 resolves the twofold structure for the doped samples. Thus, A1 and a1 grow with the PbS content. These values are surprisingly high compared with those of A2 and a2 , in spite of the higher scattering length density of the matrix/pores (short-range) relative to that of the PbS/matrix (medium-range). These results indicate a dot density higher than those for QDs in Schott filters [111] due to the smaller depleted region. We attribute the underestimation of a 2 values with regard to the pore size calculated from physisorption (see Table 6), to the incoherent scattering contribution for q>1 nm-1 , in the absence of which the “knee” of the fitted curves at q=2 nm-1 would shift to lower values, thus indicating larger pores. 64 Progress in Sol-Gel Production PbS-ormosil A1 (cm-1) a 1(nm) TEM (nm ) PbS content 9.8 81.1 104.2 22.5 35.6 44.0 6.5±0.5 10.5±0.5 13.5±0.5 depleted region (nm) 8 12 15 A2 (cm-1) 0.52 0.49 0.56 a2(nm) 0.50 0.38 0.85 Table 7.- Structural parameters fro m the fitted experimental intensities by using the two-correlation function to the Debye-Bueche model accord ing to Eq. (14). These results are promising to obtain strong quantum confinement (R<a B ) of the PbS quantum dots from which large non- linear optical behavior may be expected. 4.2. Hybrid organic/inorganic sonogels. Organic modified silicates (ORMOSIL) can be obtained by supercritical drying in ethanol of the corresponding alcogel, producing a hybrid organic/inorganic aerogel. The new material obtained takes advantage of the properties of organic polymers, including flexibility, low density, toughness and formability, whereas the inorganic component contributes properties such as surface hardness, modulus strength, transparency and high refractive index. These hybrid aerogels can be classified as Type C Ormosils, according to the Mackenzie classification; that is, the organic and the inorganic parts are chemically bonded, via covalent or iono-covalent bonds. The stronger nature of the covalent bond gives considerable improvement of the mechanical properties of the composites. One of the best methods for obtaining organicinorganic hybrid materials (OIHM) by this approach is the combined reaction of a silanolterminated polymer (PDMS), as the organic component, with a silicon alkoxide (namely, TEOS or TMOS) as the inorganic component. In the silanol-terminated polymers, the non-bridging oxygen Si-CH3 groups are dominant, depending on the average polymer molecular weight. The end chain groups –OH or –OR radicals link the polymer with the silica network via the copolymerisation between PDMS and TEOS, as stated in paragraph 3.1.1. Type C ormosils comprising TEOS and PDMS can be also promoted by high power ultrasound, when they are known as “sono-ormosils”; these are denser and have enhanced mechanical properties. As a comparison, in Table 8, the Vickers hardness value of some representative materials is shown for comparison with this new sono-ormosil. This new material falls between the softest glasses and the hardest transparent plastics. With higher PDMS content the ormosil behaves as a more rubbery material. Key Engineering Materials Vol. 391 65 N2 PDMS10 pure silica 0 200 400 600 T (ºC) 800 O2 DTG (ºC/min) DTG (ºC/min) This proposed route for obtaining these materials incorporates an organic phase in the inorganic precursor sol, in combination with the assistance of high-power ultrasound [34, 109, 110,111]. Given that in the sol- gel process little or no heating is required, organic molecules with low thermal stability can be incorporated into an inorganic ceramic or glassy host. Thus the “sonogel” route is another approach to modification of the sol- gel process. Ormosils can also be excellent matrices for non- linear optical materials due to their high transparency, inertness, mechanical strength and ease of preparation [112]. Similarly semiconductor quantum dots have been dispersed in ormosils [157] and dye lasers are enhanced [113]. A new route is now being exploited in this field of OIHMs. The process consists of drying the wet sono-ormosil at the supercritical conditions of pressure and temperature of the added liquid, to obtain a sono-ormosil aerogel, also known as an aer-ormosil. Aerogels have been successfully obtained from a TEOS/PDMS sono-ormosil. Results seem to confirm the hybrid characteristics of these aerogels; that is, the organic groups are retained in the sample that can still be considered as an OIHM. We have studied the pyrolysis process in producing these aerogels from sonogels (550 J cm-3 ); Figure 16 shows the DTG results for the sample aerogel in air and in inert atmospheres. As can be seen, the organics decompose at 230ºC (DTG in air), and then there is a further gradual weight loss between 500-600ºC, explained by the remaining hydroxyl groups producing water as by-product. In contrast, the pyrolysis under an inert atmosphere (DTG in inert gas) shows a different behaviour; up to 400ºC the aerogel is therma lly stable, and between 400-600ºC carbonisation produces a redistribution between Si-O and Si-C bonds, but no destruction of the organics groups, as reported by Zang and Pantano [114]. The transition to an inorganic material occurs between 700-800ºC, with decomposition of the methyl groups. Above 1000ºC Babonneau et al. [115] report, from 29 Si MAS-NMR experiments, the presence of an oxycarbide phase and free carbon in a DMDES/TEOS ormosil. 1000 PDMS10 pure silica 100 200 300 400 500 T (ºC) Figure 16: DTG diferential weight loss curves for a sono-aerogel made of TEOS (green line) and a hybrid TEOS/PDMS (red line), in N2 and O2 atmospheres. 66 Progress in Sol-Gel Production These new solids also present variation in several physical properties; as an example, the methyl group on the polymer chain makes these aerogels hydrophobic as is apparent from the photograph of Figure 17, from which a contact angle greater than 120º can be estimated. Figure 17: Water drop in the top of a hybrid T EOS/PDMS aerogel p iece, contact angle125º [115bis]. 4.2.1. Mechanical behaviour. Mechanical test were performed in an AG-I Autograph from Shimadzu. For the experiments an uniaxial compression device was used, equipped with a load cell of 5kN. In all the experiments the strain rate was fixed at 0.5 mm/s at room temperature. For this mechanical test samples were cast as cylinders of 20 mm length and 10 mm diameter. 30 1 2 3 1,84 1,83 1,82 log [Er(t) (MPa)]  (MPa) 20 10 1,94 1,81 1,80 1,79 1,78 1,77 0 0,0 0,1 0,2  (%) 0,3 0,4 1,76 1,93 1 2 3 log [t(s)] Figure 18: Stress-strain curves obtained on pure silica aerogel (circles) and on organic/inorganic aerogel (t riangles), last point correspond to the fracture. Inset shows the log-log plot of the relaxation modulus for the same samples [115b is]. The results of the stress-strain experiments are shown in Figure 18; the incorporation of organic polymer chains into the inorganic structure of silica gel dramatically changes the mechanical behaviour. First, the PDMS aerogel (PDMS 40 % by wt.: bottom curve) presents an elastomeric behaviour indicating a softer solid network than that of the pure silica gel (top curve); in other words, the polymeric chains are acting as springs. As the stress increases, the PDMS aerogel begins to deform less and less, indicating that the solid network is becoming stiffer; then the action of the polymer stops (at 25% strain) and the inorganic silica network begins to act, and ends with a similar slope (elastic modulus) as the pure silica aerogel (top curve). The upward curvature Key Engineering Materials Vol. 391 67 indicates a continuous increase in its elastic modulus, which is characteristic of elastomers that loose their stiffness as the polymer chains deform. First the polymer chain deforms like a spring by reducing the Si-O-Si bond angle, giving large strain with low stress. In a second state, the chain penetrates into the micropores of the inorganic silica nodules behaving like a dashpot to dampen the small strains induced as the stress increases, and finally the failure of the network provokes the fracture. Thus the solid presents first an elastic behaviour and then, in a second step, becomes like a viscous liquid, giving a total mechanical behaviour of a visco-elastic solid. A similar elastomer signature can be observed in the other samples as the polymer content is increased. The experimental values from the stress-strain curves are given in the Table 9. PDMS (wt. %) 0 10 20 40 50 Max. Load (N) 391.10 444.60 1159.90 1508.20 1252.50 Compressive strength (Nmm-2) 8.00 9.60 23.10 28.60 23.70 Strain at fracture (%) 8.40 13.40 34.70 36.80 42.90 Young´s Modulus 50% (Nmm-2) 113.80 105.70 55.10 41.50 39.40 Young´sModulus 90% (Nmm-2) 101.00 89.90 65.60 78.75 55.92 Table 9. Mechanical parameters results fro m uniaxial co mp ression test. As can be seen, this polymeric aerogel exhibits a complex behaviour on deformation, flowing like a liquid or responding elastically like a solid, depending on the time scale of the dynamics of the molecules. This mechanical duality is called visco-elasticity [116, 117] and can be characterized by the relaxation modulus, the stress time evolution at fixed strain. Such relaxation behaviour is illustrated in the inset of Figure 18 for a TEOS/PDMS aerogel (PDMS 40 % by wt); under small deformation (=15%), a biexponential decay reveals two mechanisms in the relaxation process: a fast mechanism corresponding to the elastic polymer chains (like springs) and another slow mechanism due to the stiff porous silica matrix (like a dashpot). The technique known as nanoindentation enables the mechanics of nanomaterials to be probed [118] and permits direct measurements of physical property of heterogeneous materials with close spatial resolution. This technique has traditionally been used to characterize elasto-plastic hard materials, although many groups have used the technique on soft tissues, such as demineralized dentin [119] and vascular tissues [120]. The problem of suitable analytical techniques comes from the influence of interfacial adhesive forces, which can affect the indentation contact area and has not been validated for very soft materials, with an elastic modulus below 5 MPa. Some authors have demonstrated the validity of nanoindentation in measuring the elastic modulus of pure PDMS with different degrees of crosslinking [121]. In this context several studies have concentrated on the sol-gel method and specifically on aerogels [122]. This technique seems to be very useful for the study of such complex structures as hybrid organic/inorganic aerogels, which are composed of an inorganic phase covalently bonded to an organic polymer chain. In a standard load-depth curve, the initial slope on the unloading branch (stiffness-S) is related to the reduced elastic modulus (Er) by the following relationship: 2  dP  S Er A     dh  m ax 68 Progress in Sol-Gel Production in which A is the area of the imprint on the sample, in this case a Berkovich pyramidal tip (A=24.5h2 , h being the depth). The slope S is calculated after least square fitting by the power law function proposed by Oliver and Pharr [123, 124] in the form: P  h m , where  contains geometric constants, the elastic modulus of both sample and indenter, and Poisson‟s ratio, and m is the power law exponent related to the geometry of the indenter (for a cone m=2). The hardness is calculated simply by H  P A and the elastic recovery parameter as ERP  h m ax  h f , hf being the h m ax non-recovered depth of the indenter inside the aerogel sample. Figure 19 shows a typical load-depth hysteresis curve, in which the dramatic difference between the two samples (pure silica and PDMS hybrid aerogel) is apparent; as a first indication, the load necessary to provoke a displacement of 1 m is 33% greater for the pure silica sample compared with the hybrid aerogel with a polymer content of 10 % by wt. Thus the influence of the polymer chains is mainly on the stiffness of the silica; there is a decrease in the hardness from 226 MPa for the pure silica to 24 MPa for the hybrid aerogel sample, and consequently a better elastic recovery indicated by a greater depth interval (hf, hmax ). 2,0 Young's modulus (MPa) 200 Load (mN) 1,5 loading 1,0 unloading 0,5 180 E soft E stiff E medium E uniaxial E average 160 140 120 100 80 60 40 20 0 10 20 30 40 50 PDMS content (wt%) 0,0 0 500 1000 1500 2000 2500 3000 Depth (nm) Figure 19: Depth-load hysteresis curves for the pure silica aerogel (red) and hybrid TEOS/PDMS 10 wt.% samp les. Maximu m load 1.5 mN with a penetration rate of 5 μN/s. Inset average Young‟s modulus as a function of the PDMS content for different sites and macroscopic (black rho mb) [115bis] The inset of Figure 19 shows a plot of the Young‟s modulus versus polymer content. The macroscopic Young‟s modulus from uniaxial compression (solid rhombus) is also included; this value comes close to that of the softer sites indicating that the polymer phase controls the mechanical behaviour at the macroscopic level. The silica clusters, bonded by the spring- like elastomeric polymer chains, act like a dashpot to give the final elastic modulus of a soft material. Key Engineering Materials Vol. 391 69 Grid maps were recorded of 10X10 indentations, each 20 m apart. The total load-depth curves show a fan feature indicating stiffest sites against softest ones. The former are dominated by the inorganic silica clusters and the latter by the organic polymer. This picture is apparent in the map of the elastic recovery parameter depicted in Figure 20a, in which elastic sites (white) correspond to those with low reduced modulus, and plastic sites (black) correspond to the hard regions of the aerogel. 1450 a) b) 2300 1400 h(nm) 1350 h(nm) 2250 1300 2200 1250 Pure silica aerogel - Load 1mN Hybrid aerogel 50% wt PDMS - Load 0.8mN 2150 1200 0 1000 2000 3000 t (s) Figure 20: 10x10 grid maps for the PDMS 50 wt.% sample: a) for the elastic recovery parameter, each indentation is separated 20 μm and b) creep curves compared with the pure silica aerogel sample [115bis]. In order to study the mechanical properties of these organic- inorganic aerogels in greater depth, we have performed creep tests on hybrid aerogels, as the time dependence of the strain at fixed stress [125]. Assays of this type have been used widely on different kinds of materials and, more specifically, on polymers [126]. From these tests the creep curve shows an instantaneous elastic deformation followed by a transitory state of retarded deformation, known as primary creep. During this period, the material can be understood to be undergoing a reinforcement process by deformation [127]. Then, the system reaches the stage of secondary creep, a steady state where equilibrium is found within the different mechanisms of strain and recovery. Creep curves from pure silica and hybrid PDMS aerogels are shown in Figure 20b. They depict the typical saturation shape with an instant deformation, and then the primary regime of retarded strain and secondary creep of steady-state linear regime. It can be seen that the depth of the retarded strain is much greater in the hybrid than in the pure silica aerogel. 4.2.2. Structural aspects. SANS measurements were carried out at room temperature on the V4 workstation at the Berlin Neutron Scattering Center of the HMI (Berlin). A neutron wavelength of 0.602 nm, at three sampledetector distances 1, 4 and 16 m, was used to cover a q-range from 0.036 to 3.6 nm-1 . Data were corrected using the software package available at HMI [128]. Figure 21 shows the SANS experimental intensities of hybrid aerogels with different polymer content. Several features can be discerned; first, an increase of the intensity is apparent towards the low q-side. Then wide plateaus appear with Guinier radii ranging from 5 to 8 nm as the polymer content is increased from 10 to 50 % by wt, as displayed in the inset of Figure 21. The plateaus are followed by power law decay, with a slope getting steeper in line with the polymer content, up to a maximum value of -2.5. The final parts of the curves are dominated by the incoherent scattering due to the 1 H atoms of the organic polymer in the methyl radicals. 70 Progress in Sol-Gel Production 10% Onuki+RC fit of ajustes_B 20% Onuki+RC fit of ajustes_D 30% Onuki+RC fit of ajustes_F 40% Onuki+RC fit of ajustes_H 50% Onuki+RC fit of ajustes_J 10 1 0,1 Gyration radius (nm) -1 d(q)/d (cm ) factor 10% 1 20% 0.9 30% 0.6 40% 0.3 50% 0.2 8 Rg(Gui) Rg(RC) 7 6 5 10 20 30 40 50 PDMS wt.% 0,01 -1 q0,1 (nm ) 1 Figure 21: Plot of experimental SANS intensities versus the transfer mo mentum q, for different PDMS aerogel hybrid samples (Curves are shifted). Continuous line is the least-squares fit to the Gaussian random coil model. The corresponding Gu inier radius (red square) and random coil gyration rad ius (green square) as a function of the polymer content is included in the inset [128b is] In Figure 21 a first shoulder at low-q (0.04 nm-1 ) is produced by the long range characteristic length governed by the “frozen-in” elastic constraints, which act through topologically-connected polymer cross- links. A shorter range correlation length at q=0.3 nm-1 in the plateau corresponds to the size of the so called “blobs”, which can be considered to be composed of a nucleus of the silica cluster surrounded by the polymer distributed as a random coil (RC); the average size of the “blob” can be identified with the Guinier radius. This supposition is supported by the agreement between the results of non- linear fitting of the experimental curves and a mixed Gaussian random coil form factor (continuous line in Figure 21) [129, 130]. The wide Guinier region indicates the homogeneous distribution of these “blobs” as scatterers. As the polymer content is increased, the “blobs” grow as a consequence of the entanglement of the polymer chains. The aggregates of the entangled polymer chains surrounding the silica clusters, the “blobs”, form mass fractal units, their dimension D=2.5 being characteristic of a cross- linked random coil structure. Although the aggregate does not strictly fulfil the conditions of self-similarity on different length scales to be considered a fractal object, it can be viewed as an object that does not fill the space totally. Moreover, this result suggests a kinetic growth mechanism based on the diffusion limited aggregation (DLA) of the monomer-cluster that has a fractal dimension of 2.5. 4.3. Silica ae rogel nanocomposites. Finally we present here some applications of the nanostructured sonogels described, in which a second component added in the liquid will act as the active phase. 4.3.1 Bioactivity Hybrid organic/inorganic materials (OIHMs) are being used for implants since they are tolerated by the human organism, which creates a fibrous tissue when they are embedded in the body. However, they do not become bonded to the bone unless they are bioactive. In such case, a layer of Key Engineering Materials Vol. 391 71 hydroxycarbonate of apatite (HCA) grows and wraps around the material when it is immersed in blood plasma. An HCA layer is also formed when bioactive materials are soaked in solutions mimicking the features of plasma. To perform fully as an implant, these materials need to present mechanical behaviour very similar to that of human bone. Recently, several mixtures obtained by ultrasonic agitation of colloidal silica with a sol solution containing tetraethoxysilane (TEOS) have been used to form crack- free monoliths. We have used the combination of colloid-polymer as precursor of the inorganic part of the OIHM with the aim of tailoring the porosity to control the density and mechanical strength in the range of those of human bone. Finally, in vitro bioactivity of this material has been promoted by adding colloidal silica particles to the initial sol. In ormosils, it has been shown that Ca ++ cations, together with unreacted silanols, form the bioactive hydroxyapatite layer in a simulated body fluid [131]. Si 0d O 3d Ca 0.04% 10 m Si O Ca P 0.14% 0.14% 10 m 1d Si 7d Si O 10 m P Ca 0.12% 0.10% O 10 m P Ca 0.75% 1.40% Figure 22: SEM micrographs of aerogel sample and the correspondent EDX spectra before and after be soaked for different times (0, 1, 3 and 7 days) in SBF. The ato mic weight percentages of Ca and P in samples after each immersion time are also included [77] In general terms, those materials coated with a layer of amorphous calcium phosphate that crystallises into HCA after being soaked in SBF (Simulated Body Fluid) are categorized as bioactive. The biologically active HCA layer facilitates interaction between the material and biological entities, and it has been proposed as a first phase in the sequence of reactions that result in the creation of a mechanically strong bond between the bioactive materials and the living tissues [132]. To confirm the formation of calcium phosphate on the hybrids after the in vitro assays, the variations in their surfaces with the time of soaking in SBF were also studied by SEM and EDX. Figure 22 presents the SEM images and the correspo nding EDX spectra of an aerogel composite after 0, 1, 3 and 7 days of immersion in SBF. After 1 day, the SEM micrograph shows no significant differences with respect to the initial sample (t=0). However, after 3 days in SBF, the surface appears cracked, with deep channels forming. After 7 days, some kind of material appears to have formed inside the channels. The corresponding EDX spectra show that the proportion of Ca and P on the material surface increases slowly during the first 3 days of assay, up to 0.14 mol-% for the two elements. However, the EDX spectrum of the sample immersed for 7 days shows the presence of significant amounts of P (0.75 mol-%) and Ca (1.4 mol- %). This result demonstrates the formation of calcium phosphate on the surface of the hybrid [84, 85]. In another approach using synthetic wollastonite powders as the active phase and modifying the inorganic TEOS gel matrix with MTES (methyl triethoxysilane), the bioactive behaviour was improved. The formation of a layer of HA crystals on the aerogel surface is apparent in the micrograph of Figure 23; roughly spherical shaped particles, formed by fine lamellar crystals, can be seen. A quantitative yield analysis on the EDX spectrum indicates a Ca/P ratio of 1.7, almost the same composition as that of the HA. 72 Progress in Sol-Gel Production Figure 23: SEM micrograph and EDX pattern of SiO2 /CaO co mposite aerogel after soaking in SBF during 25 days [132b is]. 4.3.2. CO2 sequestration Among the gases contributing to the „„green house effect‟‟, the influence of CO 2 has been estimated at 40–50% of the total. Currently, the policy proposed for dealing with this problem is the reduction of gas emissions from power generation processes, transport and especiallycontaminating industries (such as cement factories). Any viable strategy with this objective requires technologies that incorporate CO 2 elimination processes, i.e., separation, capture (sequestration), disposal or storage and, finally, elimination. Everything must be carried out at costs that are feasible for the gas-emitting industries, organisations and countries to bear. Figure 24: a) Diffract ion profiles of the TEOS/wo llastonite composite before and after exposure to 30 min of CO2 flow b) SEM of the byproducts of the same composite after exposure to 30 min of CO2 flow. There can be observed the individual monocrystals (rho mbohedric-like) o f carbonates together with unreacted silica gel [80] It is known that some natural silicates form carbonate when in contact with CO 2 . In particular, wollastonite (CaSiO 3 ) reacts to produce calcite and silica. This is a natural reaction that has operated over millions of years to reduce the CO 2 in the Earth‟s atmosphere. Wu et al. (2001) [133] and Tai et al. (2006) [134] have published results on the rate of conversion of wollasto nite (CaSiO 3 ) into calcite (CaCO 3 ) for different experimental protocols. These values must be interpreted to determine the efficiency of CO 2 fixation by the minerals. Thus, starting from a powdered wollastonite sample in a reactor at atmospheric pressure a nd room temperature, Wu et al. obtained a Key Engineering Materials Vol. 391 73 conversion rate of 14% of wollastonite into calcite after 22 days. Later, a comparative study has been carried out [135] analysing different types of sample and experimental conditions, and has given more encouraging results. In Figure 24 the carbonation reaction is illustrated by the X-Ray diffraction peak and by the SEM micrograph. In this case, the wollastonite is almost entirely dissolved and the calcite peak can be observed standing out from the other peaks. The reaction efficiency can be estimated as relatively high: their analysis reveals that almost 85% of the wollastonite is dissolved. The SEM micrograph shows the by-products of the aerogel composite after exposure to 30 min of CO 2 flow, where the individual monocrystals (rhombohedric- like) of carbonates together with unreacted silica gel can be observed. These results indicate advantages in terms of fast reaction time and efficiency, which should open up new uses for composites of this type [86, 87]. 74 Progress in Sol-Gel Production ACKNOWLEDGEMENTS The authors thank to the Spanish Ministry of Science and Technology for financial support under project MAT2005-1583. The authors are members of the PAI TEP-115 Research Group of the Junta de Andalucía (Spain). REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] M.C. Rocco, J. Nanoparticle Res., 1, 1 (1999). “Sol-gel technology for thin films, fibers, preforms, electronics and specialty shapes” Ed. L. Klein, Noyes publications, New Jersey, (1988). Y.H. Xu, CH. Cheng and J.D. Mackenzie, J. Non-Cryst. Solids 176, 1 (1994). CH. Cheng, Y.H. Xu and J.D. Mackenzie, Ferroelectrics 152, 31 (1994). H. Schmidt, J. Non-Cryst. Solids 73, 681 (1985) J.D. Mackenzie, Q. Huang and T. Iwamoto, J. Sol-Gel Sci. and Tech., 7,151 (1996) J.D. Mackenzie, J. Sol-Gel Sci. and Tech. 26, 23 (2003). S.J. Kramer, F. Rubio-Alonso and J.D. Mackenzie, in “Better Ceramics through Chemistry VII: Organic-Inorganic Hybrid Materials”, Eds. by B.K. Coltrain, C. Sanchez, D.W. Schaefer and G.L. Wilkes, MRS, Pittsburgh (1996) p. 295. S. Amberg-Schwab, M. Hoffmann, H. Bader and M. Gessler, J. Sol-Gel Sci. and Tech. 13, 141 (1998). C. Sánchez, G.J.de A.A. Soler-Illia, F. Ribot, T. Lalot, C.R. Mayer and V. Cabuil, Chem. Mater. 13, 3061 (2001) P. Gómez-Romero and C. Sánchez, “Functional Hybrid Materials”, Willey-VCH, Germany, (2004). K.S. Suslick and G.J. Price, Annu. Rev. Mater. Sci. 29, 295 (1999). K.S. Suslick, Science 247, 1439 (1990). Lord Rayleigh, Phil. Mag. 34, 94 (1917). T.G. Leighton, in “The Acoustic Bubble”, London Academic (1994). E.B. Flint and K.S. Suslick, Science 253, 1397 (1991). K.S. Suslick, Y. Didenko, M.M. Fang, T. Hyeon, K.J. Kolbeck, W.B. McNamara III, M.M. Mdleleni and M. Wong, Phil. Trans. R. Soc. Lond. A 357, 335 (1999). E. Blanco, L. Esquivias, R. Litran, M. Piñero, M. Ramírez-del-Solar, and N. de la Rosa-Fox, Appl. Organomet. Chem. 13, 399 (1999). J. Zarzycki, Heterogeneous Chem. Rev. 1, 243 (1994). J. P. Lorimer and T. J. Mason, Chem. Soc. Rev. 16, 239 (1987). E. A. Neppiras, Ultrasonics 22-1, 25 (1984). J.L. Luche, in “Current TRENDS I Sonochemistry” (G.J. Price, Ed.) p.34 RSC, Cambridge 1192. T. J. Mason, “Practical Sonochemistry.” Ellis Horwood, Chichester, 1991. P. Kruus, Ultrasonics 21, 201 (1983). K. Morita, Y. Hu, and J. D. Mackenzie, J. Sol–Gel Sci. Technol. 3, 109 (1994). J. L. Luche, in “Current Trends in Sonochemistry” (G. J. Price, Ed.), p. 34. RSC, Cambridge, 1992 P. Boudjouk, in “Science of Ceramics Chemical Processing”, L. L. Hench and D. R. Ulrich, Eds., Wiley, New York (1986). p. 363. Key Engineering Materials Vol. 391 [28] [29] [30] [31] 75 G. J. Price, M. P. Hearn, E. Wallace, and A. M. Patel, Polymer 37, 2303 (1996). M. Tarasevich, Ceram. Bull. 63, 500 (1984)(Abstract only). L. Esquivias and J. Zarzycki, in “Current Topics on Non Crystalline Solids” (M. D. Baró and N. Clavaguera, Eds.), p. 409. World Scientific, Singapore, 1986. L. Esquivias and J. Zarzycki, in “Ultrastructure Processing of Ceramics, Glasses and Composites” (J. D. Mackenzie and D. R. Ulrich, Eds.), p. 255. Wiley, New York, 1988. N. de la Rosa-Fox, L. Esquivias, and J. Zarzycki, Diffusion Defect Data 53/54, 363 (1987). D.A. Donatti and D.R. Vollet, J. Sol–Gel Sci. Technol. 4, 99 (1995). D.A. Donatti and D.R. Vollet, J. Non-Cryst. Solids. 208, 99 (1996). D.A. Donatti, A. Ibañez Ruiz and D.R. Vollet, Ultrasonics Sonochem. 9, 133 (2002). D.A. Donatti, D.R. Vollet and A. Ibañez Ruiz, J. Sol-Gel Sci. Technol. 18, 5 (2000). H.-J. Tiller, R. Gobel, and U. Hartung, J. Non-Cryst. Solids 105, 162 (1988). H. Coudurier, R. Baudru, and J. B. Donnet, Bull. Soc. Chim. Fr. 9, 3147, 3154, 3161 (1971). Pouxviel J C , Boilot J P, Beloeil J C . and Lallemand Y , J. Non-Cryst. Solids,1987, 89 : [32] [33] [34] [35] [36] [38] [39] [40] 345. [41] Orcel G and Hench L , J. Non-Cryst. Solids, 1986, 79 : 177. [42] Y. Hoshino and J. D. Mackenzie, J. Sol–Gel Sci. Technol. 5, 83 (1995). [43] L.L. Hench and J.K. West, Chem. Rev. 90, 33 (1990) [44] G. Orcel and L.L. Hench, J. Non-Cryst. Solids 79, 177 (1986). [45] T. Adachi and S. Sakka, J. Mater. Sci. 22, 4407 (1987) [46] L.L. Hench, in Science of Ceramic Chemical Processing, Eds. L.L. Hench and D.R. Ulrich, Wiley, NY, (1986) p. 52 [47] T. Woigner, PhD Thesis, University of Montpellier, France (1984). [48] P.H. Tewari, A.J. Hunt and K.D. Loffus, Mater. Lett., 3, 363 (1985). [49] S. Bruin and K.Ch.A.M. Luyben in Advances in Drying, vol. 1, ed. A.S. Mujumdar, Hemisphere, NY, (1980) p. 155. [50] J. Rodriguez-Ortega and L. Esquivias, J. Sol-Gel Sci. Tech., 8, 117 (1997) [51] J. Rodriguez-Ortega, PhD thesis, Universidad de Cadiz, Spain (1996) [52] L. Esquivias, J. Rodríguez-Ortega, C. Barrera-Solano and N. de la Rosa-Fox, J. Non-Cryst. Solids, 225,239 (1998) [53] J.D. Bernal and J. Mason, Nature, 188, 910 (1960) [54] G.D. Scott, Nature, 188,908 (1960) [55] J.L. Finney, Proc. R. Soc (London) A319, 479 (1970) [56] J. Zarzycki, J. Non-Cryst. Solids, 147&148, 176 (1992) [57] L. Esquivias, N. de la Rosa-Fox, M. Bejarano and M. J. Mosquera (2004) Langmuir 20:3416. [58] A. Emmerling, J. Fricke. J. Sol-Gel Sci. and Tech. 8 (1997) 781-788. [59] M. Grzegorczyk, M. Rybaczuk, K. Maruszewski. Chaos, Solitons & Fractals 19 (2004) 1003-1011. [60] H. Ma, J. H. Prevost, G. W. Scherer. International Journal of Solids and Structures 39 (2002) 4605-4614 [61] H. Ma, A. P. Roberts, J. H. Prevost, R. Jullien, G. W. Scherer. J. Non-Cryst. Solids 277 (2000) 127-141 [62] H. Ma, J. H. Prevost, R. Jullien, G. W. Scherer, J. Non-Cryst. Solids 285 (2001) 216-221. [63] T. Woignier, J. Reynes, A. H. Alaoui, I. Beurroies, J. Phallipou, J. Non-Cryst. Solids 241 (1998) 45-52. [64] J. Primera, A. Hasmy, T. Woignier, J. Sol-Gel Sci. and Tech. 26 (2003) 671-675. 76 [65] Progress in Sol-Gel Production T. Woignier, F. Despetis, A. Alaoui, P. Etienne, J. Phalippou, J.Sol-Gel Sci. and Tech. 19 (2000) 163-169. [66] T. Woignier, J. Phalippou, J. Non-Cryst. Solids 100 (1988) 404-408. [67] A. Rumpf, Chem. Ing. Tech. 30 (1958) 144. [68] V. Morales-Florez, N. de la Rosa-Fox, M. Pin˜ero, L. Esquivias, J. Sol–Gel Sci. Technol. 35 203 (2005). [69] Persistence Of Vision, www.povray.org. Last visit, 23/nov/2006. [70] V. Morales-Flórez, M. Piñero, N. de la Rosa-Fox, L. Esquivias, J.A. Anta, J. Primera, J. Non-Cryst. Solids, 354, 193-198 (2008) [71] M.C. Barrera-Solano, N. de la Rosa-Fox and L. Esquivias. J. Non-Cryst. Solids 147/148 (1992) 194-200. [72] Piñero M., Atik M. , Zarzycky J., J. Non-Cryst. Sol.. 147&148, 523 (1992) [73] Piñero M., Zarzycki J., J. Sol-Gel Sci.Tech., 1, 275 (1994) [74] M. Piñero, R. Litran, C. Fernández-Lorenzo, E. Blanco, M. Ramírez-del-Solar, N. De la Rosa-Fox, L. Esquivias, A. Craievich, J. Zarzycki, J. Sol-Gel Sci.Tech., 2, 689 (1994) [75] A. Craievich, N. De la Rosa-Fox, E. Blanco, M. Piñero, M. Ramírez del Solar, L. Esquivias, Nanostruc. Mater., 5 (3) 363 (1995). [76] R. Erce-Montilla, M. Piñero, N. de la Rosa-Fox, A. Santos, and L. Esquivias, J. Mater. Res. 16, 2572 (2001). [77] L. Esquivias, V. Morales-Flórez, M. Piñero, N. de la Rosa- fox, J. Ramírez, J. GonzálezCalbet, A. Salinas, M. Vallet-Regí, Mater. Res. Soc. Symp. Proc. Vol.847 EE12.1.1 (2005) [78] J. A. Toledo-Fernández, R. Mendoza-Serna, V.Morales-Flórez, N. de la Rosa-Fox, A. Santos, M. Piñero, L. Esquivias, Bol. Soc. Exp. Ceram. V., 46 [3] 138-144 (2007) [79] A. Santos, J. A. Toledo-Fernández, R. Mendoza-Serna, L. Gago-Duport, N. de la Rosa-Fox, M.Piñero, L. Esquivias, Ind. Eng. Chem. Res, 46, 103-107 (2007) [80] A. Santos, M. Ajbary, A. Kherbeche, M. Piñero, N. de la Rosa-Fox, L. Esquivias, J. Sol-Gel Sci. Tech. DOI: 10.1007/s10971-007-1672-1 (2008) [81] Jain, R.K., Lind, R.C., J. Opt.Soc. Am. 73, 647 (1983) [82] J. Butty,Y.Z. Hu, N. Peyghambarian, Y.H. Kao and J.D. Mackenzie, Appl. Phys. Lett. 67, 2672 (1996). [82bis] N. de la Rosa-Fox, M. Piñero, R. Litrán, L. Esquivias. J. Sol-Gel Science and Technology 26, 947-951, (2003) [83] J. Butty, N. Peyghambarian, Y.H. Kao and J.D. Mackenzie, Appl. Phys. Lett. 69, 3224 (1996). [84] K.L. Shaklee, R.E. Nahory and R.F.Leheny, J. of Lumin. 7, 284 (1973). [85] N. Peyghambarian, S. Koch and A. Mysyrowicz, in “Introduction to Semiconductor Optics”. (Prentice-Hall, New Jersey, 1993). Pp. 245-253. [86] A. Lipovskii, E. Kolobkova, V. Petrikov, I.Kang, A. Olkhovets, T. Krauss, M. Thomas, J. Silcox, F. Wise, Q. Shen, S. Kycia, Appl. Phys. Lett. 71(23), 3406,(1997). [87] N. F. Borrelli, D. W. Smith, J. Non-Cryst. Solids., 180, 25 (1994). [88] M. Mukherjee, A. Datta, D. Chakravorty, J. Mater. Res., 12(10), 2507, (1997) [89] F del Monte, Y. Xu, J. D. Mckenzie, J. Sol-Gel Sci.&Tech., 17, 37 (2000) [90] S. Gallardo, M. Gutiérrez, A. Henglein, E. Janata, Ver. Bunsenges. Phys. Chem. 93,1080, (1989) [91] M. Guglielmi, A. Martucci, J. Fick, G. Vitrant, J. Sol-Gel Sci.&Tech., 11, 229 (1998) [92] A. Martucci, P. Innocenzi, J. Fick, J.D. Mackenzie, J. Non-Cryst. Solids., 244, 55, (1999) [93] L. Spanhel, E. Arpac, and H. Schmidt, J. Non-Cryst. Solids., 147&148, 657, (1992) [94] N.N. Parvathy, G.M. Pajonk, and A. Venkateswara, J. Cryst. Growth, 179, 249, (1997). Key Engineering Materials Vol. 391 77 [95] N.N. Parvathy, G.M. Pajonk, and A. Venkateswara, Mat. Res. Bull., 32(4), 397, (1997) [96] A. Martucci, M. Guglielmi, and K. Urabe, J. Sol-Gel Sci. &Tech., 11, 105 (1998). [97] A. Martucci, J. Fick, J. Scell, G. Battaglin and M. Guglielmi, J. Appl. Phys. 86(1), 79 (1999). [98] M. Guglielmi, A. Martucci, E. Menegazzo, G.C. Righini, S. Pelli, J. Fick, G. Vitrant, J. SolGel Sci. & Tech., 8, 1017, (1997) [99] H. Lin, E. Bescher, J.D. Mackenzie, H.Dai, and O.M. Stafsudd, J. Mat. Sci., 27, 5523, (1992). [100] P. T. Guerreiro, S. Ten, N. F. Borrelli, J. Butty, G. E. Jabbour, N. Peyghambariam, Appl. Phys. Lett. 71(12), 1595, (1997) [101] A. Martucci, J. Fick, J. Scell, G. Battaglin and M. Guglielmi, J. Appl. Phys. 86(1), 79 (1999). [102] M. Piñero, N. de la Rosa-Fox, R. Erce-Montilla and L. Esquivias, J. Sol-Gel Sci. Tech., 26, 527 (2003) [103] N. de la Rosa-Fox, R. Erce-Montilla, M. Piñero and L. Esquivias, Opt. Mater., 22, 1 (2003) [104] G.Banfi, V. Degiorgio and B. Sepeit, J. Appl. Phys. 74, 6925 (1993). [105] P. Debye and M. Bueche, J. Appl. Phys. 20, 518 (1949). [106] P. Debye, H.R. Anderson Jr. and H. Brumberger, J. Appl. Phys. 28, 179 (1957). [107] N. Ashcroft, J. Lekner, Phys. Rev, 145, 83 (1966). [108] E. Geissler, F. Horkay, A.M. Hecht and C. Rochas, Polymer 38 (1), 15 (1997). [109] K. Morita, Y. Hu and J.D. Mackenzie, Mat. Res. Soc. Symp. Proc., Vol 271, 693 (1992). [110] E. Blanco, L. Esquivias, R. Litran, M. Piñero, M. Ramírez-del-Solar and N. de la Rosa-Fox, Appl. Organomet. Chem. 13, 399 (1999). [111] N. de la Rosa-Fox, L. Esquivias and M. Piñero in Handbook Of Organic-Inorganic Hybrid Materials and Nanocomposites, Vol. 1, Ed. H.S. Nalwa, American Scientific Publishers, USA (2003) Ch. X (in press) [112] J.D. Mackenzie, J. Ceram. Soc. Japan 101, 1 (1993). [113] B. Dunn, J.D. Mackenzie, J.I. Zink and O.M. Stafsudd, SPIE Proc. 1328, 174 (1990). [114] H. Zang and C. Pantano, J. Am. Ceram. Soc. 73, 958 (1990). [115] F. Babonneau, L. Bois and J. Livage, J. Non-Cryst. Solids 147&148, 280 (1992). [115bis] N. de la Rosa-Fox, V. Morales-Flórez, J.A Toledo-Fernandez, M. Piñero, R. MendozaSerna, L. Esquivias. “Journal of the European Ceramic Society” 27, 3311-3316 (2007) [116] J.D. Ferry, “Viscoelastic properties of Polymers”, 3rd Ed. (Wiley, NY, 1980). [117] I.M. Ward, J. Sweeney. “An introduction to the mechanical properties of solid polymers”. (2nd Ed.) John Wiley & Sons, (2004). [118] Schuh, C.A., Mater. Today, 2006, 9, 32-40. [119] Marshall, G.W., Habelitz, S., Gallagher, R., Balooch, M., Balooch, G. and Marshall, S.J., J. Dent. Res. 2000, 80, 1768. [120] Ebenstein, D.M. and Pruitt, L.A., J. Biomed. Mater. Res. A, 2004, 69A, 222. [121] Carrillo, F., Gupta, S., Balooch, M., Marshall, S. J., Marshall, G.W., Pruitt, L. and Puttlitz, C.M., J. Mater. Res., 2005, 20, 2820-2830. [122] Moner-Girona, M., Roig, A., Molins, E., Martínez, E. and Esteve, J. Appl. Phys. Lett., 1999, 75, 653-655 [123] Oliver, W. C. and Pharr, G.M., J. Mater. Res., 1992, 7, 1564-1583. [124] Pharr, G.M., Oliver, W.C. and Brotzen, F.R., J. Mater. Res., 1992, 7, 613-617. [125] Piñero, M., Morales-Flórez, V., de la Rosa-Fox, N. and Esquivias, L., Mechanical properties of hybrid aerogels, Bol. Soc. Esp. Ceram. V., 2005, 44(5), 291-293. [126] Ferry, J.D., “Viscoelastic Properties of Polymers”, Wiley, N. Y., 3rd. ed., 1980. 78 Progress in Sol-Gel Production [127] Callister Jr., W. D. “Introducción a la Ciencia e Ingeniería de los Materiales”, Ed. Reverté, Barcelona, 1995. [128] U. Keiderling, Physica B, 234–236, (1997), 1111. [128bis] N. de Rosa-Fox, V. Morales-Flórez , J. A. Toledo-Fernández M. Piñero, L. Esquivias y U. Keiderling. J Sol-Gel Sci. Technol , 45, 245-250.(2008) [129] Kirste, R.G. and Oberthur, R.C., in “Small Angle X-ray Scattering”, Ed. Glatter, O. and Kratky, O., Academic Press, (1982), Chap. 12, 387-431. [130] Pedersen, J.S., J. Appl. Cryst., 33, (2000), 637-640. [131] M. Kamitahara, M. Kawashita, N. Miyata, T. Kokubo and T. Nakamura, J. Mat. Sci.: Mat. In Med. 13 (2002) 1015-1020 [132] Hench L. L. J. Am. Ceram. Soc. 81 (1998) 1705. [133] Wu JC-S, Sheen J-D, Chen S-Y, Fan Y-Ch. Ind Eng Chem Res 40, 3902 (2001) [134] Tai CY, Chen W-R, Shih S-M AIChE J 52, 292 (2006) [135] Zevenhoven R, Eloneva S, Teir S Catal Today 115, 73 (2006) View publication stats