Among chemically bonded ceramics (i.e., those not utilizing thermally activated diffusion for bonding) the French synthetic opal gilsonite provides an excellent existence theorem. By using optical, scanning, and electron microscopy techniques and x-ray, chemical, and differential thermal analyses, it is shown for the first time that the synthetic opal is composed of two separate phases: noncrystalline silica and crystalline (tetragonal) zirconia balls. The zirconia balls with sizes ranging from 7-50 nm appear to be present in an extraordinary regular “lattice” in the void spaces of the silica “balls” of mean size 200 nm. A comparison of the fracture toughness, Kic, data for the gilsonite and natural opal shows that the former is significantly tougher than the latter. The KIC values for gilsonite fall between those of the Corning 0337 Glass Ceramic and Wesgo Al-500 alumina, showing that surprisingly tough ceramics can be made near room temperature by resorting to chemical bonding.

The glass-forming ability of the three alloy systems Co–Zr, Cu–Zr, and Ni–Zr has been analyzed for three distinct production routes: (1) liquid quenching, (2) vapor deposition, and (3) solidstate reaction. Using the free energy and heats of formation curves obtained from the thermodynamic characterization of the respective alloy systems, a satisfactory rationale can be obtained for amorphous phase formation by all three routes. The analysis shows that while amorphous phase formation by quenching from the high-temperature liquid is clearly dependent on factors such as quench rate and the value TG/TM, it is the low-temperature stability of the amorphous phase relative to the other crystalline structures that enables amorphous phases to be formed by both vapor deposition and solid-state reaction. The underlying free energy curves indicate the interesting possibility of a supersaturation sequence in the nucleation of an amorphous phase by solid-state reaction. The principles underlying thermodynamic characterizations are briefly discussed, and a characterization for Co–Zr is presented.

A number of grain boudary structures prepared from evaporated Au thin films have been investigated using very high-resolution electron microscopy. Particular emphasis has been placed on analyzing [110] tilt boundaries with both low- and high-angle misorientations. Observations of the atomic positions at dislocation cores and close-packed polyhedral shapes at the interface have been made. Symmetric boundaries have been observed as well as interfacial regions where the grain orientations deviate from perfect coincidence and hence the boundary structure exhibits perturbations to the regular polyhedron stacking. The special case of “plane coalescence” has also been observed, as well as the boundary splitting near a hole. The fabrication and observation of a generalized boundary, which cannot be classified as a tilt structure, will also be demonstrated.

Radiation effects in a natural, metamict zirconolite from Sri Lanka that has received an alpha-decay dose greater than 1026 alpha decays/m3 have been studied using x-ray diffraction (XRD), high-resolution transmission electron microscopy (HRTEM), electron paramagnetic resonance spectroscopy (EPR), extended x-ray absorption fine structure spectroscopy (EXAFS), and x-ray absorption near edge structure spectroscopy (XANES). The same techniques were applied to the sample annealed between 1000°and 1100°C. The heat of recrystallization was measured by differential thermal analysis (DTA) with values of 40 to 50J/g. In contrast to previous work [A. E. Ringwood, Am. Scientist 70, 201 (1982); Mineralog. Mag. 49, 159 (1985)], these results demonstrate that there are fundamental differences at the atomic level between the annealed, crystalline, and the natural, fully damaged zirconolite. We suggest that the most likely structure for the fully damaged state is that of a random, three-dimensional network with no atomic periodicity extending beyond the first coordination sphere. Even within the first coordination sphere, there is a reduction in coordination number and an associated decrease in bond length for principal cations (Ca and Ti). Despite this structural modification and the great age of the specimen (∼550 million years), the natural zirconolite shows only minor signs of geochemical alteration. The metamict structure is readily annealed, initially to a disordered, flourite-type structure, and finally, at higher temperatures, to a highly twinned, monoclinic zirconolite structure.

Electron microscopic observations, electrical resistivity, and lattice parameter measurements of annealed Pd-12.5 at. % Y alloy indicate the presence of long-range order in the form of a Pt7Cu superlattice. This is similar to the order found in Pd7Ce, which is also isomorphous with Pt7Cu. The order-disorder transformation temperature was found to be Tc = 783 ± 5 K. Evidence was found from resistivity measurements for the existence of an ordered structure in the hypostoichiometric alloys: Pd-10.0 at. % Y and Pd-8.0 at. % Y. The order-disorder transformation temperatures were lower than that for the stoichiometric alloy Pd7Y.

Self-propagating high-temperature synthesis (SHS), also called combustion synthesis, is useful for fabricating numerous ceramics. In the case of SiC, heat released from the exothermic reaction is not sufficient to completely convert the mixed reactants of constituent elements into SiC in the usual nonadiabatic experimental system. This disadvantage could be overcome by a new ignition process called, the “direct passing method of electric current.” By using this method, stoichiometric fine SiC powder could be obtained rapidly and efficiently with low electric power. This paper also involves the effect of particle size of Si and C initial reactant powders on conversion efficiency into SiC and also on particle size of SiC powder fabricated by this method.

The effects of substituting divalent metal ions (Mg, Ca, Sr, Ba) for Y in YCrO3 were investigated by electrical conductivity, Seebeck coefficient, and thermal conductivity measurements. The electrical conductivity results were consistent with the hopping-type conduction of a temperature-independent concentration of small polarons, with measured activation energies of 0.18-0.26 eV. The Seebeck coefficient increased nearly linearly with temperature and indicated p-type conductivity. Both electrical conductivity and Seebeck coefficient results show a strong dependence on dopant size (ionic radius) and indicate that the highest carrier concentrations were associated with Ca as the dopant, which is attributed to the similar ionic radii of Ca2+ and Y3+. The thermal conductivity decreased slightly with temperature and dopant concentration.

The reliability of microfocus x radiography for detecting internal voids in structural ceramic test specimens was statistically evaluated. The microfocus system was operated in the projection mode using low x-ray photon energies (<20 keV) and a 10 μm focal spot. The statistics were developed for implanted internal voids in green and sintered silicon carbide and silicon nitride test specimens. These statistics were compared with previously obtained statistics for implanted surface voids in similar specimens. Problems associated with void implantation and characterization are discussed. Statistical results are given as probability-of-detection curves at a 95% confidence level for voids ranging in size from 20–528 μm in diameter.

The effect of carbon on the mechanical properties of ordered, face-center-cubic Ni3Al has been studied. It has been found that carbon provides no ductihzation to the intermetallic compound, but exerts a large solid solution strengthening effect. The strengthening rate measured is Δσy/ΔC∼0.5G per atom percent carbon, where G is the Ni3Al shear modulus. Auger analysis and lattice parameter measurements were also carried out. The results are discussed with respect to the nature of carbon in grain boundary regions and in the bulk.

Several novel oxides have been prepared by the decomposition of carbonate precursors of calcite structure of the general formulas Mn1−xMxCO3 (M = Mg,Co,Cd), Ca1−xMx'CO3, and Ca1−x−yMxMy”CO3

Implanted, polymeric films have been produced by accelerator-ion-beam irradiation of metallic substrates immersed in hydrocarbon gases. Typical substrates include silver, aluminum, and steel; hydrocarbon gases include 1,3 butadiene and ethylene at 6.6 Pa pressure; ion beams employed include singly ionized H, He, and Ar at 30 keV. Experimental procedures and corrosion-resistance properties of the films are reviewed (each discussed elsewhere). A theory of the film-formation process is presented. It is concluded that the films form as the result of a two-stage process: glow-discharge adhesion and polymerization followed by radiation-induced implantation resulting from collisional recoil and substrate sputtering.

Using a two-zone method, the possible formation of an intercalation compound of hexagonal boron nitride (BN) with Cs and Br2 was investigated. Only a few percent weight increase was observed by doping BN with Cs and Br2. The electron paramagnetic resonance (EPR) signal was significantly modified by Cs doping, which is attributed to the reaction between the Cs atoms and spin resonance centers (N vacancies) in BN; no change in the EPR spectra was observed with Br2 doping. However, the deep blue colored Cs-BN complex reported by Mugiya and co-workers was not obtained with the two-zone method. Though no evidence of systematic intercalation reaction in BN was observed in contrast to graphite host materials, intercalation islands induced by the introduction of Cs atoms were suggested by the transmission electron microscopy (TEM) observations.

The dominant diffusing species in VA13, Co2Al9, CrAl7, and MoAl12 formation was determined. These are the initial phases to form in metal-Al reactions and the most Al-rich phases on the phase diagrams (except for VA13). It was found for the growth of VA13, Co2Al9, CrAl7, and MoAl12 that Al is the dominant moving species. This is very different from results with metal-Si marker experiments where either the metal or Si can be the dominant moving species in the initial phase. In aluminides the dominant moving species is related to the initial phase formed. For the growth of the Ni-rich phase, Ni3Al, Ni was the dominant moving species.

Raman spectroscopy (RS) and low-angle x-ray diffraction (LAXRD) have been used to characterize semiconductor multilayer interfaces. In the present study a model for Raman spectra of multilayers is developed and applied to the specific case of the interfaces of a-Si/a-Ge multilayers. Quantification of the “blurring” of interfaces is possible because peak heights in the Raman spectra of thin films are proportional to the number of scatterers, thus RS is capable of directly “counting” the total number of chemical bonds of a given type in the film. Multilayers, prepared by various deposition techniques, are compared. The relative roles of LAXRD and RS in investigating interfaces are contrasted. Several a-Si/a-Ge multilayers deposited by ultra-high vacuum (UHV) evaporation (MBD) are found to exhibit very regular periodicities and exceptionally sharp interfaces (<1.0 Å intermixing).

The pH of the solution that zirconia is precipitated from defines the crystal phase formed after calcination of the material at 400°to 600°C. A metastable tetragonal form is obtained for either low (less than about 5) or high (greater than about 13) pH. The tetragonal phase formed at high pH is much more stable at the calcination temperature than the material obtained at a low pH is. For a material obtained by precipitation at a low pH, monoclinic crystallites, determined by x-ray line broadening, were obtained that were smaller than the tetragonal crystals that produced them. A crystallite size effect, based on x-ray line broadening, is not responsible for the formation or stabilization of the tetragonal phase.

Determination of the amount of cubic and tetragonal phase in yttria–zirconia using high-angle XRD (x-ray diffraction) has been complicated by problems of resolution and interpretation. The evidence, from electron diffraction studies, for a c→t shear transformation also needs to be taken into account. Two compositions, a 3 and a 5.7 mol % Y2O3−ZrO2, were sintered and thermally treated at different temperatures between 1450°and 1700°C. X-ray diffraction traces revealed the anticipated tetragonal (400) and (004) reflections. The region of the diffraction pattern that was thought to arise from the cubic phase could be best interpreted as a second tetragonal phase t'. The original amount of cubic phase computed from the t' reflections showed good agreement with the phase diagram of Scott. The lattice parameters of the t' phase were determined, and the volume of the tetragonal (t') unit cell was calculated.

The effectiveness of oxidation-inhibiting nitride layers formed in silicon by ion implantation has been studied by in situ thermogravimetric analysis. The oxidation time delay and the oxidation rate after nitride breakdown were measured for various fluences of implanted nitrogen. The results generally confirm previous findings for the oxidation delay; however, a detailed analysis indicates that the onset of oxidation is not as abrupt as had been previously assumed. After the delay period, the oxidation of an implanted Si/SiO2 structure follows a linear time dependence at oxide thicknesses where unimplanted Si/SiO2 structures follow a parabolic dependence.

The creep behavior of a poly crystalline nickel aluminide with the composition Ni-23.5 at.% Al-0.5 at. % Hf-0.2 at. % B has been measured as a function of stress, temperature, and grain size. At high stresses, of the order of 100 MPa, the strain rate is nonlinear in the stress, with a stress exponent greater than two. Below approximately 10 MPa, at 1033 K, the steady-state strain rate is almost proportional to the stress, indicating that diffusional creep is rate controlling. Calculations of expected Nabarro—Herring and Coble creep rates did not answer whether diffusive mass transport through the grains, or along the grain boundaries, is rate controlling. The grain-size dependence of the strain rate, however, indicates predominance of volume diffusion control, i.e., Nabarro—Herring creep, for our experimental conditions.

High concentrations of self-interstitials are trapped by dopant atoms during ion implantation into Si. For group V dopants, these complexes are sufficiently stable to survive solid-phase-epitaxial (SPE) growth but break up on subsequent thermal processing and cause a transientenhanced diffusion. Dopant diffusion coefficients are enhanced by up to five orders of magnitude over tracer values and are characterized by an activation energy of approximately one half of the tracer values. In the case of group III dopants, any complexes formed during implantation do not survive SPE growth but a second source of self-interstitials becomes significant and leads to similar transient effects. This is the damaged layer underlying the original amorphous/crystalline interface. These observations provide direct evidence for longrange self-interstitial migration in Si, and we believe these are the first observations of the interstitialcy diffusion mechanism with no vacancy contribution. We propose that the complexes are simply interstitial dopant atoms (in a split <100> interstitialcy configuration) that are particularly stable in the case of group V dopants. As they decay self-interstitials are released and cause the transient-enhanced diffusion.

As-received P/P + (100) epitaxial silicon wafers were heat treated using the one-, two-, and three-step internal gettering heat treatment cycles in wet oxygen, dry oxygen, and nitrogen ambients. The results indicate that ambients have an effect on the growth of bulk defects and denuded zone formation in the epitaxial silicon wafers.