Small angle x-ray scattering (SAXS) is the technique of choice for the determination of structure on the 10–1000Å scale. We have used this technique to study the growth and topology of the macromolecules which precede gelation in several chemical systems used in sol-gel glass technology. The results show that branched polymers, as opposed to colloids, are formed. The alcoholic silica system is akin to organic systems where gelation occurs through growth and crosslinking of chain molecules. Data are reported from both the Porod and Guinier regions of the SAXS curve and these data are interpreted in terms of geometrical structures predicted by various disorderly growth processes. The results indicate that the degree of crosslinking can be controlled by catalytic conditions. The degree of crosslinking may, in turn, control phase separation and processability to a dense glass.

Small angle scattering experiments have demonstrated that the structure of the silicate species produced by the hydrolysis of silicon alkoxides in non-aqueous solvents ranges from extended, weakly cross-linked polymers to highly condensed, colloidal particles. In contrast, inorganic, aqueous silicate solutions yield primarily colloidal particles because the silicate species have a number of different silanol sites available and the preferred condensation reaction is that of weakly condensed species with highly cross-linked branch sites, such as those on an amorphous silica surface. It is proposed that in the alkoxide systems, however, the hydrolysis reaction may control the number and type of silanol sites available for condensation. In acid catalyzed reactions, the rate of hydrolysis of a silicate tetrahedron tends to decrease as alkoxide groups are removed. This favors the production of silanol sites on the end of chains, thus generating linear polymers. In base catalyzed reactions, it is argued that each subsequent hydrolysis of a tetrahedron should proceed more rapidly than the previous one, producing numerous branch points which are the preferred sites for condensation.

Silica gels were prepared by three methods in which the original silicate species varied from extended linear or randomly branched polymers to more highly crosslinked clusters to colloidal particles of anhydrous silica. During isothermal sintering experiments, the viscosities of the two polymer gels increased significantly (up to 3 orders of magnitude) while the isothermal viscosity of the colloidal gel was constant. Viscosity increases were explained by crosslinking and structural relaxation of the polymeric gels.

Two series of gels were prepared by hydrolyzing TEOS with HCl or NH4OH catalyst. In the first series, the molar ratio of water to TEOS was 4:1 and the catalyst addition was (1 × 10−3), (4 × 10−3), (1 × 10−2), (4 × 10−2), and (1 × 10−1). In the second series, the catalyst addition was (1 × 10−3) moles and the molar ratio of water to TEOS was 2:1, 4:1, 8:1, 16:1, and 32:1. Dried samples were characterized for surface area, porosity and C and H content. Samples were heated to 800°C and thermogravimetric weight loss was recorded. Both HCl and NH4OH are catalysts for hydrolysis, but during polymerization, the relative rates of hydrolysis and condensation determine resulting dried microstructures.

Transmission electron microscopy of leached NaOċ3SiO2 and K2Oċ3SiO2 glasses showed that the hydrateg glass surface is a silica rich gel which phase separates during leaching. Phase separation is initiated when pockets of an aqueous phase nucleate and grow in the hydrated silica matrix. The aqueous pockets eventually become a network of interconnected pores with a pore diameter of ∼30 nm. In advanced stages of leaching, the remaining interconnected silica rich phase resembles an aggregation of colloidal silica particles. The observed phase separation is consistent with literature models concerning the chemistry of silica in aqueous solutions and in leached glass.

The production of small silica gel monoliths by the hydrolysis and polycondensation of tetramethyl orthosilicate (TMS) in a methanol solution is now a common procedure. Drying is generally done slowly in a methanolic atmosphere [1] or under hypercritical conditions [2]. The potential application of sol-gel technology in the production of materials for large structures requires a rapid processing time, which causes a problem due to cracking. For large scale space structures the materials produced also require a low molecular weight, a low densification temperature, control of devitrification and a wide range of physical properties and gel densities. These requirements are potentially met by multicomponent silica based gels [3]. To decrease the drying time, which is the longest part of gel processing, the gel strength needs to be increased to resist cracking and the drying stress reduced. The gel strength can be improved by optimizing the variables in the gel manufacturing process.

Strengths of as drawn gel-derived SiO2 fibers are found to depend on H2O and alcohol content of the sol, as well as drawing viscosity. Additional factors, including fiber diameter and drying and consolidation schedules are found to dramatically influence strengths of the consolidated glass fibers.

The use of sol-gel techniques to prepare glasses and crystalline ceramics offers outstanding opportunity for breakthroughs in technology. The areas of particular promise include novel glasses; crystallineceramics with exceptional microstructures; coatings for modification of electrical, optical, mechanical and chemical properties; porous media with high surface area and tailored chemistry; ceramic powders with high chemical homogeneity and narrow distributions of particle size; matrix materials in ceramicceramic composites; and a wide spectrum of specialty ceramic materials, ranging from abrasives and fibers to glass ceramics and films. Opportunities in each of these areas will be discussed and related to the advances in understanding and process technology required for their achievement. The theses will be advanced that creative chemistry provides the key to many of these advances, that ceramists simply MUST learn more chemistry, but that we dare not rest from our labors when the chemistry is done.

A number of investigators [1–6] have studied the sol-gel-glass transformation and the aging, drying and densification processes involved. However, monolithic, amorphous, large scale xerogels are still difficult to produce because of insufficient understanding of basic changes in microstructure during the sol-gel-glass transformation and the chemical reactions of the precursors, solvents and catalysts used. Some papers sugqest ways of obtaining monolithic gels by (1) increasing the mechanical strength of gel by aging [7]; (2) diminishing the magnitude of capillary forces by enlarging the pore size and/or decreasing the surface energy by using surfactants [8] (3) reducino the rate of evaporation of the solvent from the pores by using a semipermeable membrane during drying [9] (4); and eliminating the pore liquid-solid interface by hypercritical evacuation [10].

The addition of a dryinq control chemical additive (DCCA) to the solvent allows short drying times for monolithic gel samples. The variables involved in the use of a DCCA in preparing gels from the Na2O-SiO2 -B2O3 system are described. Due to its higher reactivity, a DCCA influences the chemical behavior of the material at each step of the processing. The effects of formamide as a DCCA on the qelation process, surface charge, and surface area are presented. The variations of the physical properties of the gel, due to the addition of the DCCA, are related to the reactions of the formamide with the sol. Although the DCCA helps in producing monolithic dried gels, the problem of densification is not solved as yet.

The densification behavior of a gel-derived borosilicate glass was studied. Surface area, thermal gravimetric and infrared measurements were used to calculate the surface hydroxyl coverage as a function of time at several temperatures. Application of a viscous sintering model to the isothermal shrinkage data showed viscosity to increase isothermally by almost two orders of magnitude at the lowest investigated temperature. Although most of this increase is attributable to increased crosslinking accompanying dehydroxylation, structural relaxation was also shown to significantly affect the densification kinetics.

A review has been made of our works on two types of sol-gel methods for obtaining oxide fibers through low temperature process: (1) the alkoxide method and (2) the freezing-of-gel method. In the alkoxide method, metal alkoxides in an alcoholic solution are hydrolyzed and polycondensed into chain-like polymers, drawn into gel fibers near room temperature and heated to several hundred °C. Various problems encountered in the process are discussed with SiO2-based fibers. In the freezing-of-gel method, a hydrogel is unidirectionally frozen by lowering the gel cylinder into a cold bath. Thawing of unidirectionally grown ice crystals leaves a bundle of hydroxide fibers. Factors affecting the formation of fibers are discussed with titanium and zirconium hydroxide fibers as examples.

Various methods have been developed for producing crystals, powders, coatings and monoliths from gels [1]. The scientific basis for understanding compositional effects, gelation, aging, drying and densification is also advancing rapidly [2]. However, there is as yet relatively little information on the durability, weathering, or corrosion resistance of glasses, glass-ceramics, ceramics, or composites made via the gel route. Data is also sparse on the effects of vacuum, thermal exposure, or mechanical stress on the stability of gel-derived solids. Relationships between sol-gel processing variables and environmental stability are especially lacking at the present time. Since many end-use applications of gel derived materials involve exposure to severe environments, it is essential that the durability of these materials be established during their development.

Gels and gel-monoliths in the SiO2 -GeO2 system were prepared by different procedures using metal alkoxides as starting materials. Several gel compositions with increasing GeO2 were prepared. Crystallization behavior of gels in relation to the preparation procedures, homogeneity, and molecular structure of gels is discussed. Supercritically dried porous gel-monoliths of high silica compositions were densified into glass by sintering at 1280°C.

The structures of alumina gels made from aluminum sec butoxide have been studied by X-ray diffraction and scanning electron microscopy as a function of pH as well as of the temperature of peptization and gelling. The structures of these gels appear to be very sensitive to the conditions of preparation, and lead to a variety of gelled products and transformation behaviors on firing. Specific attention is directed to preferential orientation and to structures which are X-ray super-amorphous.

Transition metal oxide gels can be obtained, through a polycondensation process, by acidification of aqueous solutions. Thin layers can be easily deposited onto a substrate. Their electronic and ionic properties could lead to new developments of the sol-gel process. The semiconducting properties of V2O gels can be used for antistatic coatings or electrical switching devices. These gels exhibit a lamellar structure and can be described as particle hydrates. They are fast proton conductors and could behave as a host lattice for intercalation.

Catalyst supports are required to meet a wide range of objectives, including chemical purity and composition, optimised surface area/porosity and thermal stability. Gel processing offers means by which many of these requirements can be achieved. In addition, the use of sols of controlled rheology allows the support material to be prepared in a form which facilitates its application to substrates (including metals) as a coating, thereby greatly improving its versatility for producing novel catalysts. The preparation of typical support materials will be discussed, together with examples of their use in catalysis.

Gels with compositions Na2O-(1-x)SiO2 were prepared from tetraethylorthosilicate (TEOS) with sodium acetate and from tetramethylorthosilicate (TMOS) with sodium methoxide or TEOS with sodium ethoxide. After gelation (10 min.) the samples were slowly dried for about one month at room temperature. Thermogravimetric analysis (up to 500°C) in air and O2 atmosphere was performed. The electrical properties were studied in air by impedance spectroscopy during successive temperature cycles starting from room temperature dried gels of composition 0.10 Na2O 0.90 SiO2. Protonic conductivity appears to be eliminated after reaching the 2nd stage of the drying process (200°C). The activation energies for conduction and the conductivities show small discrepancies from known values for homogeneous glasses of equivalent compositions. Considering the effect of water on the conductivity of glasses prepared by the melting procedure, one can estimate the residual water content in the so-called dried gels to be close to 0.5 wt %.

The formation and characterization of thin ceramic coatings is currently an active area in materials science. Specifically, there has been increased interest in these materials as dielectric and optical layers for advanced microelectronic circuitry [1]. This is in addition to their continued use and evaluation as protective coatings in a variety of applications [2]. One method of forming these coatings is via the sol-gel technique. However, only a few papers have been published in this area [3–7]. Dense silicon dioxide, aluminium oxide and various glass compositions have been deposited in thin film form on various substrates. Further, silicon dioxide films, formed by the sol-gel method, have been nitrided in ammonia at high temperatures and partially converted to an oxynitride [8–11]. These films were shown to be highly non-homogeneous, with a predominance of nitrogen at the surface. Despite the desirability of dense uniform well bonded coatings of the nitrided ceramics, i.e. Si3N4 and the oxynitrides, no method yet exists for the formation of nitrides by the sol-gel method.

Diketonate aluminum alkoxides were prepared and allowed to react with acetoxyalkylsilanes. The conversion of the aluminosiloxanes thus obtained to aluminosilicates at 450°C was demonstrated. These aluminosiloxane ceramic precursors are glassy, oligomeric materials which readily dissolve in organic solvents to give viscous solutions ideal for casting films. Films of the metal-organic compounds prepared in this manner yield monolithic, crack-free aluminosilicate films directly on thermal curing, with a thickness limit of 3000Å for single crystal silicon wafer substrates. These aluminosilicate films have been found to be effective anticorrosion barriers for various metal substrates. The chemistry of this organoaluminosilane system and the nature of the aluminosilicate films obtained has been investigated using a variety of chemical and physical methods. The results of this investigation will be described and the possible advantages of this direct conversion process over the sol-gel method as a means of obtaining aluminosilicate films will be discussed.