The hydrolysis of organic silicates ( mainly Si(OEt)4: [TEOS] and Si(OMe)4: [TMOS]) is the key reaction in sol gel process for preparation of silica. The mechanism of this reaction is not well known because of the competition taking place between hydrolysis itself and condensation reactions (homo or heterocondensations) but also because of redistribution reactions which take place competitively. Furthermore the competition between these steps depends also on the catalysis (acid,basic or nucleophilic) [1, 3].

High pressure Raman spectroscopy is used to monitor the hydrolysis reaction of TMOS in solutions with methanol, acetonitrile, acetone, dioxane and formamide, and of TEOS as a function of pH and the catalyst used. The rate constants for various solvents, temperatures and pressures are experimentally determined from the time dependence of Raman band intensities. It is shown that the reaction is slow in dioxane and fast in methanol or formamide. The volume of activation is found from the pressure dependence of the rate constant. The volume of activation, the dielectric constant, dipole moments and hydrogen bonding properties and their role in the hydrolysis reaction are discussed. It is shown that solvents which can form hydrogen bonding with Si-OR groups can increase the rate of the reaction.

This paper surveys a few of the current issues in sol-gel reaction kinetics. Many times seemingly modest changes in reactants or reaction conditions can lead to substantial differences in the overall reaction rates and pathways. For example, qualitative features of the reaction kinetics can depend on catalyst concentration. At very high acid-catalyst concentrations, reverse reactions are significant for TMOS solgels, while for moderate acid-catalyst concentrations, reverse reactions are substantially reduced. The reaction kinetics of two similar tetraalkoxysilanes: tetramethoxysilane (TMOS) and tetraethoxysilane (TEOS), can be markedly different under identical reaction conditions. Under acid-catalyzed reaction conditions, a TMOS sol-gel undergoes both water- and alcoholproducing condensation reactions while a TEOS sol-gel undergoes only water-producing condensation. The early time hydrolysis and condensation reactions of a TMOS sol-gel are statistical in nature and can be quantitatively described by a few simple reaction rate constants while the reaction behavior of a TEOS sol-gel is markedly nonstatistical. A comprehensive theory of sol-gel kinetics must address these diverse experimental findings.

Sol-gel polymerization of [Si8O12](OCH3)8 in CH3CN under neutral conditions yields very high surface area (SBET > 900 m2/g) xerogels. This property is seen to result from the structure of the gel on the molecular level. According to N2 adsorption studies, model studies, and TEM studies, the large size and rigidity of the cubic [Si8O12] core structure leads to polymers whose rigidity inhibits extensive crosslinking of the type observed in orthosilicate derived xerogels.

The reaction of Bi[OC(CH3)3]3 with [Cu(μ-O2CCH3)2]2 has been investigated as a possible route to mixed-metal materials containing bismuth and copper. The product distribution obtained from this reaction depends not only upon the stoichiometry, but also upon the method of product isolation. Products containing bismuth and copper are obtainable as well as an unusual, soluble hexametallic copper(II) alkoxide acetate complex, Cu6[OC(CH3)3]6(O2CCH3)6. The mixed-metal products were examined by differential scanning calorimetry, and thermolyzed and hydrolyzed samples of these products were compared by scanning electron microscopy.

Oxo-alkoxides and allied derivatives are the organic counter part of polyanions and polycations. Most of them do not lead to further polymerization and are therefore dead endswhich are grnerally not directly involved in the formation of macromolecular oxide networks. However their well defined structure make them good models for the basic understanding of the first steps of hydrolysis- condensation reactions. The structural characterization of a new metal oxo species Ce6(μ3-O)4(μ3-OH)4(acac)12 obtained via the hydrolysis of acetylacetone modified cerium isopropoxide is presented. Possible mechanisms of formation of this compound as well as for some other titanium Ti6(μ2-O)2(μ3-O)2(μ2-OAc)4(μ2-OPri)6(OPri)6 and niobium Nb8O10(OEt)20 based oxo-alkoxides will be discussed.

A general route to low temperature synthesis of ternary metal tin oxides with controlled stoichiometries from single component molecular metal alkoxide precursors is described. The solid state and solution structures of homoleptic tin(IV) alkoxide compounds have been investigated to establish criteria for the determination of their solution structure. Tin alkoxide compounds suitable for metathesis reactions have been synthesized and used to prepare the mixed metal alkoxide compounds [ZnSn(OEt)6] and [((COD)Rh)2Sn(OEt)6] (where Et = ethyl and COD = 1,5-cyclooctadiene). Hydrolysis of [ZnSn(OEt)6] at neutral pH results in the formation of a high surface area, mainly amorphous hydrous oxide powder which forms Zn2SnO4 and SnO2 on heating to 600°C and ZnSnO3 on heating to 1000°C.Thermolysis of [((COD)Rh}2Sn(OEt)6] results in formation of rhodium and tin(IV) oxide (cassiterite phase).

Yttrium is involved as oxide in a variety of advanced materials. Hydrolysis of its most common alkoxide, yttrium oxoisopropoxide Y5O(OiPr)13 , is difficult to control as a result of the high electropositivity of the metal. Functional alcohols such as 2-methoxyethanol and acetylacetone have thus been used as modifiers. These tailored precursors have been fully characterized (single crystal X-Ray diffraction, IR, NMR. ). The tendency of methoxyethanol to act as a bridging ligand and to stabilize high oligomers - Y(OC2H4OMe)3 is a decamer - gives some insight into its behavior as “network builder”. The unexpected cleavage of acetylacetone offers a route to an interesting acetatoacetylacetonate derivative, Y2(OAc)2(acac)4(H2O)2 , but also suggest that this classical “modifier” could be affected by the hydrolysis-polycondensation reactions of the metallic species during the sol-gel process. Comparison of the hydrolysis behavior of the different precursors was performed.

Lead titanate gels were prepared by sol-gel processing from lead acetate and titanium isopropoxide in methoxyethanol. In the present study, 1H, and 13C FTNMR spectroscopic and mass spectrometric techniques were used to identify the structures of lead and titanium precursors formed by refluxing lead acetate and titanium isopropoxide in methoxyethanol, respectively. Lead acetate was found to be partially alkoxylated, whereas, the titanium precursor was identified as a dimer with complete alkoxy group exchange. Similar procedures were applied to identify the lead titanium complex formed from lead and titanium precursors.

The word “hydrolysis,” literally “water splitting,” refers here to those reactions of metallic ions with water that liberate protons and produce hydroxy or oxy complexes in solution and precipitate hydroxide or oxide solids. Over the wide range of pH available in water, the strong tendency of hydroxide to bond to and bridge among cations leads to a multitude of species for many of the metals. Properties of the metal ion such as “hardness” and, especially, charge and size, determine the strength of these interactions and the species formed by each. Over 200 mononuclear species have been identified that predominate in dilute solutions across the pH scale. It is the stability of these species that largely determines the pH dependence of the solubility of the oxide or hydroxide solid formed. Both the tendency of the variety of polynuclear species (those containing more than one metal ion) to form and the solubility of the hydrolytic solid correlate with the stability of the first mononuclear species. Other correlations allow one to predict the enthalpy and entropy of a variety of hydrolysis reactions, and therefore the dependence of the stability of hydrolysis products on temperature.

The nuclearity and structure of aluminum cations generated by the hydrolysis of aluminum salt solutions depends markedly on the method of preparation. This speciation affects not only the thermochemistry of processes leading to ceramics, but the microstructures of the ceramics themselves. A brief review of aluminum ion hydrolysis is presented along with a study of the thermal evolution of gels derived from aluminum salt solutions via several different hydrolysis methods. The observation of 5 - coordinate aluminum in a bulk transition alumina by MAS-NMR is reported. A new, high defect precursor of η-alumina dubbed “high-5 alumina” is described and the far reaching implications of this discovery are discussed.

We have used small angle neutron scattering, static light scattering and 27Al NMR to examine the structure and composition of alumina sol-gels formed by the hydrolysis of aluminum alkoxides. For LT sols at low acid concentrations, and HT gels over a wide range of acid concentration, 27Al solution NMR suggests, by the dearth of spectral resonances, that high molecular weight species are being formed. Analysis of the small angle neutron scattering data in the Porod regime indicates these sol-gels exhibit a power-law dependence consistent with mass fractal dimensions ranging from 1.45 to 1.8. These fractal dimensions are consistent with models based on diffusion limited cluster aggregation. The fractal dimensions do not differ significantly between LT and HT sols at the same acid concentration. However, for both temperature regimes, the fractal dimension increases with increasing acid concentration, suggesting a progression to a more compact network. Static light scattering measurements indicate the Guiner radii of the cluster aggregates vary from 600 to 2000Å.

Small angle X-ray scattering techniques have been used to investigate the shape and size distribution of Zr(IV) species in aqueous solution. This study has shown that when zirconyl chloride solutions, containing the zirconyl tetramer, are subjected to various treatments polymerisation occurs. While ageing and addition of base produces an increase in particle size the shape remains globular. Refluxing the solutions produces “rod” like particles of varying length but constant cross sectional radius.

The conventional mechanism developed by LaMer (1) is often considered as the most relevant model for describing the precipitation of uniform particles. In this model, the concentration of a species is slowly increased above its equilibrium value until a critical concentration is reached and nucleation occurs. The resulting particles consume soluble species and the supersaturation level is reduced until there is a balance between particle growth and the generation of reactive species. At this point nucleation stops. Particle growth then continues by molecular addition of soluble species to the growing particles. Uniformity is achieved through a short nucleation time and a particle growth mechanism where the small particles grow more rapidly than the large particles.

Most conventional and experimental methods for the formation of ceramic objects involve the manipulation of small particles and/or structures comprised of these particles. Consequently, a better understanding of the physics and chemistry of small particles could contribute to the development of better ceramics. Here the results of some recent computer simulation studies of processes such as particle aggregation, deposition and segregation are surveyed. More realistic models are needed to represent accurately the behavior of most real systems.

Control of the preparation of monodisperse particles for green body formation can be achieved by understanding the growth mechanism. Growth processes for colloidal silica and titania were followed by cryo-TEM, which allows direct observation of the particles in the liquid state. Structural development was tracked by quenching samples at successive reaction times. The preparation of silica particles involved the formation of ramified species which subsequently collapsed after reaching a certain size. In the titania system, initially small dense nuclei are observed which eventually form uniformly textured homogeneous particles.

The microstructure or interparticle ordering in concentrated dispersions of colloidal pmma (polymethylnethacrylate) particles in a mixture of tetralin and decalin have been monitored using light scattering techniques. These sterically stabilized, uniformly sized, nearly hard, colloidal spheres are observed to exhibit an equilibrium phase transition from a liquid-like ordering to a crystal-like ordering of suspended particles as the volume fraction of solids increases. The crystals have a close packed - random stacked structure. At the largest volume fractions a nonequilibrium glassy phase results.

Samples at different volume fractions are subjected to steady and oscillatory shear flow. Four basic structures are observed to exist: liquid or distorted liquid-like, string-like, sliding or randomly stacked layers, and face centered cubic (FCC) structures. Oscillatory shear studies will be reported here and are made as a function of strain amplitude and shear history, in addition to volume fraction. Generally, oscillatory shear is effective in ordering samples. For example, an unstable FCC ordering can be induced in an equilibrium liquid-like sample.

A key step in the processing of ceramics is the consolidation of powders into engineered shapes. Colloidal processing uses solvents (usually water) and dispersants to break up powder agglomerates in suspension and thereby reduce the pore size in a consolidatd compact. However, agglomeration and particle rearrangement leading to pore enlargement can still occur during drying. Therefore, it is beneficial to consolidate the compact asdensely as possible during the suspension stage. The consolidation techniques of pressure filtration and centrifugation were studied and the results are reported in this paper. In particular, the steady-state pressure-density relationship was studied, and information was obtained regarding the consolidation process, the microstructure, and the average density profile of consolidated cakes. We found that the compaction processes in these two consolidation methods are quite different. In general, a consolidated cake is a particulate network made up of many structural units which are fractal objects formed during aggregation in the suspension. In pressure filtration, compaction is a process of breaking up the fractal structural units in the particulate network by applied pressure; the resulting partie rearrangement is produced by overcoming energetic barriers which are related to tee packing density of the compact. Recently, we performed Monte Carlo simulations on a cluster-cluster aggregation model with restructuring,1 and found the exponential relationship between pressure and density is indeed the result of the breaking up of the fractal structural units.2 On the other hand, in centrifugation, compaction involves the rearrangement of the fractal structural units without breaking them so that the self-similar nature of the aggregates is preserved. Furthermore, we calculated density profiles fom the bottom to the top of the cakes by solving the local static force balance equation n the continuum particulate network. In pressure filtration of alumina and boehmite, the cakes are predicted have uniform density. The results of γ-ray densitometry3 on a pressure-filtrated alumina cake confirmed this prediction. In contrast, in centrifugation, the density profiles are predicted to show significant variation for cakes on the order of one centimeter high. Moreover, the pressure-filtered boehmite cakes showed no cracking during drying. This indicated that pressure filtration is a good consolidation technique for nanometer-sized particles such as boehmite. The improved drying property is probably a result of a minimal shrinkage due to the formation of higher packing densities during filtration.

We give an overview of some of the experiments currently underway to study the coupling of the microstructure and rheology of concentrated suspensions. Nuclear magnetic resonance imaging, real-time x-ray radiography, and refractive index matching allow the viewing of particles in concentrated suspensions. Both shear flow experiments and falling ball rheometry are reviewed. In the slow flow of these suspensions of large, hard, particles in a viscous Newtonian fluid, colloidal forces are negligible and hydrodynamic forces dominate.

Large local concentration changes are shown to occur rapidly in suspensions of uniform spheres subjected to flow between concentric rotating cylinders. Suspensions of spheres with a bimodal size distribution not only show similar phenomena, but also exhibit particle separation according to size. In addition, the large particles in the bimodal suspension migrate into ordered, concentric, cylindrical sheets, parallel to the axis of the cylinders. These sheets of particles rotate relative to each other. The particle migration and structure formation induced by this inhomogeneous shear flow is believed to be responsible for torque reductions and other anomalous behavior witnessed during the rheological testing of concentrated suspensions reported in the literature. Thus, suspensions may not always be characterized by a viscosity that is a scalar material property.

Suspensions of fibers also show markedly different rheological properties when the particles are aligned by flow. Falling ball rheometry is shown to be an effective tool to determine the bulk viscosity of a suspension while only slightly influencing the microstructure. This is illustrated by showing that falling ball rheometry can isolate the effect of orientation on the viscosity of a suspension of fibers.

Shear flow in α-A12O3 suspensions having a volume fraction of solids (φ) in the range between 0.17 and 0.50 was investigated between pH 4 and 12. It is Newtonian if the magnitude of the zeta potential exceeds a critical value which depends on φ;its value is 39 mV if φ= 0.40 and 74 mV if φ= 0.50. If this potential is less than the critical value, shear flow is pseudoplastic; its yield value markedly changes (e.g., 0 to >100 Pa) in a slightly φ-dependent, narrow pH range (<0.5 units). If a second oxide, t-ZrO2 , is present, its pH-dependent colloidal behavior governs the overall rheology, though its concentration may be only 11 % that of α-A12O3. Scanning electron microscopy of composite pieces indicates that a detrimental, rheologically detectable interaction between α-Al2O3 and t-ZrO2 can be avoided and the distribution of t-ZrO2 can be optimized during pressure casting by control of pH. Best conditions correspond to either Newtonian flow or pseudoplastic flow with a very low yield value.