Coble creep, which is controlled by mass transport along grain boundaries, has been identified in a powder-metallurgically prepared nickel aluminide with the nominal composition Ni–22.8Al–0.6Hf–0.1B (at. %). Diffusional creep rates ∊ as a function of temperature T, stress σ, and grain size L are well described by ∊ = 33δb Db Ωσ (kTL3), where δb Db = 3 × 10−6 m3 s−1 × exp [– (313 kJ/mol)/(RT) (δb is the diffusional grain boundary width, Db is the grain boundary diffusivity, R is the gas constant, and T is the absolute temperature). The activation energy of 313 kJ/mol is unusually high as compared to that for volume diffusion.

Gax In1 − x Pepilayers grown under a range of growth conditions by organometallic vapor phase epitaxy (OMVPE) on GaAs substrates have been studied in the electron microscope. The results show the presence of an ordering of the group III sublattice parallel to some of the {111} planes. Dark-field images directly reveal ordered domains of different orientations that appear not to be perfect, but contain many planar defects parallel to the growth surface.

Metallasiloxanes of the type, M (OSiR3)4, which contain zirconium and hafnium can be prepared by the facile silanol-induced cleavage of organometallic compounds. The metallasiloxanes vary from mobile liquids to insoluble nonmelting powders depending upon the constituents bonded to the silicon. The compounds can be fired into inorganic refractory oxides at temperatures varying from 350–600 °C. The resulting materials have the general formula M Si4O10. Electron spectroscopy for chemical analysis (ESCA) data suggest that the M-O–Si linkage remains intact during this process. Thus the M Si4O10 material can be considered a mixed oxide alloy rather than a simple mixture of the oxides M O2 and SiO2.

Measurements are presented of the electrical double layer and van der Waals forces between the (0001) surfaces of two single-crystal sapphire platelets immersed in an aqueous solution of NaCl at pH values from 6.7 to 11. The results fit the standard Deryaguin-Landau-Verwey-Overbeek (DLVO) theory, with a Hamaker constant of 6.7 × 10−20 J. These are the first measurements made using the Israelachvili surface forces apparatus without mica as a substrate material, and they demonstrate the possibility of using this technique to explore the surface chemistry of a wider range of materials.

The oxide superconductor Y–Ba–Cu–O is studied using Auger scanning microscopy. The chemical depth profiles of the samples were obtained. It is concluded that two phases are present in the sample, one corresponding to the standard composition and another that is Ba enriched. The first shows a platelet shape and the second a granular appearence that covers the surface of the sample.

Systematic rapid quenching experiments on the icosahedral phase-forming system Al1 − xRux were performed for × < 20 at. %. The solidified alloys have been studied by electron microscopy and x-ray diffraction to determine their composition, constituent phases, and phase morphology. It has been determined that the icosahedral phase must form directly from the liquid and, at these quench rates, is always found in the presence of second or third phases. The results have been summarized in a metastable phase diagram appropriate for rapid solidification of Al1 − xRux at piston and anvil quench rates. This metastable phase diagram describes the results for Ru concentrations less than 14 at. %.

The present article first describes some characteristics of structure, chemistry, and electronic (bond) nature for grain boundaries in the A3B Li2-type intermetallic compounds. Next, the phenomenological aspects for the grain boundary brittleness of the Li2-type intermetallic compounds are reviewed with respect to the combination of the constituent atoms, the alloying effect, the stoichiometry effect, and a role of impurity or gaseous atoms. It is emphasized that the brittleness of grain boundaries in the intermetallic compounds is directly controlled by the atomistic and electronic structures at grain boundary regions. Based on these systematic investigations, it is suggested that the brittleness of the Li2-type intermetallic compounds can be manipulated by appropriate control of composition and the corresponding electrochemical bond environment at grain boundary planes and by control of test environment. Furthermore, some examples of the materials development are described.

Individual particles of argon-atomized Fe-30Ni powder are electrodynamically levitated and remelted by a CO2 laser pulse. The thermal history of each droplet during remelting and solidification is monitored by single-color radiation pyrometry at each of three wavelengths (850, 750, and 550 nm). Experiments are done in an atmosphere of either air or nitrogen. The average supercooling of six experiments performed in nitrogen is 298 K with a standard deviation of 14 K. This value is of the same order as several others reported in the literature using bulk levitation and emulsification techniques. The average supercooling of seven experiments performed in air is 163 K with a standard deviation of 20 K. The difference suggests that oxides are forming in the air-remelting experiments and catalyzing nucleation at relatively low supercoolings. The average cooling rate of the liquid droplets prior to solidification in nitrogen is 1.5 × 105 K/s. This measured cooling rate is somewhat higher than that predicted by Newtonian heat flow modeling, and the difference is attributed to radiative losses not considered in the Newtonian model. The measured cooling rate is used to estimate the total heat transfer coefficient characterizing cooling of a small metal droplet in a quiescent gas atmosphere. A lower bound of 1.5 × 106 K/s on the droplet heating rate during recalescence and a minimum average liquid/solid interfacial velocity during recalescence of 0.1 m/s are estimated.

The crystalline-to-amorphous transformation in U6Fe produced by thermal fission fragment damage was studied using resistivity and differential scanning calorimetry. The results are described in terms of a model of radiation-produced defect buildup in the crystalline matrix followed by transformation of small regions to an amorphous phase when a critical local defect concentration is reached. This can occur directly in a single cascade or from cascade overlap. The total resistivity is modeled assuming an inhomogeneous media consisting of a crystalline matrix containing a dose-dependent concentration of defects and amorphous zones. The crystallization behavior is initially, starting at Tc = 388 K, a kinetically limited process of shrinkage of amorphous zones that gradually transforms to nucleation and growth in fully amorphous material at Tc = 555 K.

Thin-film reactions of Co with Zr have been studied in the temperature range between 473 and 523 K by electrical conductance measurements and cross-sectional transmission electron microscopy (CS-TEM). The reduction of the electrical conductance during the solid state reaction is explained by formation and growth of an amorphous phase at every Zr/Co interface. For long reaction times the growth of the layer thickness follows a shifted $\sqrt t$ law. For short reaction times the measurements show a linear time law, which is expected for an interface limited reaction.

The oxidation behavior of a Pb-2.9 at.% Sn alloy was studied using Auger electron spectroscopy (AES) combined with sputter depth profiling. Samples were oxidized in air at temperatures less than 90 °C and at low oxygen pressures at room temperature. The oxide formed on the alloy after oxidation in air has an “oxide inversion layer,” where the amount of preferentially oxidized tin increases with depth into the sample, reaches a maximum, and then decreases to its bulk value. Internal oxidation of tin was also observed in this alloy, probably occurring along grain boundaries. Low pressure oxidation studies indicate that grain boundaries play an important role in the initial development of the oxide.

A procedure based on the embedded atom method (EAM) is presented for developing atomistic models for use in computer simulation calculations, with an emphasis on simple but general schemes for matching experimental data with fitting parameters. Both the electron density function and the two-body potential are taken as exponentially decreasing functions and the model is derived for any choice of cutoff distance. The model has been applied successfully to seven fcc and three hcp metals, but the extension to bcc metals was unsuccessful because of difficulty in matching the shear anisotropy ratio.

A stochastic model is presented for staging transitions in graphite intercalation compounds. Three types of kinetic processes are introduced into the domain model of Daumas and Herold, and a one-dimensional Monte Carlo simulation is carried out within the framework of a single-column model to study the time evolution of the system for a variety of cases of staging transitions. Results of Monte Carlo simulations from stage 3 to stage 2 after a sudden change of chemical potential show that the staging transition depends sensitively on the final value of the chemical potential μf, temperature of the system, and the kinetic coefficients. When μf is taken in a certain range in the phase diagram, the time evolution of the structure factor demonstrates the coexistence and no significant broadening of peaks corresponding to the initial and final stage states. For other values of μf, it is observed that the staging transitions proceed via disordered states and in some cases the final states are also disordered. Similar results are obtained for staging transitions from stage 4 to stage 3. Simulations are also made for staging transitions in which the start is from stage 4 and suddenly μ changes to a value in the stage-2 stable region. It is found that the system transforms either directly into a stage-2 state or into a metastable stage-3 state without showing evidence of further transition into a final stage-2 state, depending upon the values of μf, and kinetic coefficients. A discussion is presented on the possibility of observing an intermediate metastable stage-3 state during the transition from stage 4 to stage 2.

A series of unistructural materials was used to investigate dependence of fracture energy upon crystal bond ionicity. Fracture energy values obtained from indentation measurements were found to agree with those obtained from a simplistic sine function force law calculation. In addition, an indentation procedure to determine whether environments enhance fracture is discussed.

Nanosecond time-resolved reflectivity and ellipsometry experiments have been performed on (100) Si wafers encapsulated by 5.5–76.2 nm thick thermal oxides, using pulsed KrF (248 nm) laser energy densities sufficient to melt the Si beneath the oxide. Post-irradiation nulling ellipsometry, optical microphotography, and surface profiling measurements were carried out. It was found that the threshold energy density required to melt the Si varies with oxide thickness; this is explained primarily by the reflective properties of the oxide overlayer. The time-resolved reflectivity and ellipsometry measurements show that rippling of the SiO2 layer occurs on the 20–40 ns timescale and results in a decrease in specular reflectivity of the rippled silicon surface beneath. Optical model calculations suggest that pulsed laser annealing through a thick oxide layer results in a damaged near-surface silicon layer (∼ 30 nm thick); this layer contains defects that are probably responsible for the degraded performance of devices.

Angle-resolved x-ray photoelectron spectroscopy (ARXPS) results in a very detailed analysis of minor amounts (≥0.3 nm) of interface compounds and their spatial distribution. First experimental results on the Si–SiO2 interface used in microelectronics, are presented. The ARXPS results on plane, single crystalline (100) Si, oxidized to about 5 nm SiO2, indicate a planar Si surface connected by about one monolayer Si∗ to a compressed SiO∗2 layer coated by SiO2. Separated from this interface region, Si clusters stabilized by a compressing SiO∗2 coating have been found in SiO2. Dehydrogenation is showing up in binding energy changes in SiO2 indicating that H or OH is not only saturating Si∗ but is also bonded to SiO2. The O deficiency of amorphous SiO2,x is increasing toward the outer SiO2 surface.

Recently it was reported that high quality, 10–100 μm thick, single-crystalline Si films were formed on oxidized single-crystalline Si wafers by the lateral epitaxial growth over oxide (LEGO) process. Although this recrystallization process is reliable and reproducible, periodic regions of dislocations in the otherwise relatively dislocation-free Si film were not well understood. In this paper, therefore, the film stress and defect properties are investigated in detail, and devices made in recrystallized wafers are compared with devices in conventional wafer structures. Stress levels were found to be too low to cause defects, with TEM data suggesting an impurity mechanism (SiO2 precipitation) for small dislocation loops and slight crystalline misorientation for long dislocation lines in the periodic, defective areas. Device results confirmed that LEGO is a viable alternative to the dielectric isolation (DI) technology.

Antiphase domain-free, monocrystalline β-SiC thin films have been epitaxially grown on off-axis Si (100) substrates. The effects of degree of misorientation and substrate preannealing on the antiphase domain boundaries (APB's) have been investigated. Wet oxidation, optical microscopy, and transmission electron microscopy were used to characterize the defect structures in both β-SiC thin films grown on exact Si (100) and on off-axis Si (100) substrates. The results revealed that many dislocations were included in APB's and that APB's were eliminated in the β-SiC thin films grown on Si (100) substrates that were oriented 2°–4° from [100] toward the [011] direction. Some APB's were observed near the edge of the β-SiC film on the 2° off-axis Si (100) substrates; however, they were eliminated by substrate preannealing. The carrier concentration and electron mobility of these β-SiC films were similar to those of β-SiC films grown on exact Si (100) substrates as determined by differential capacitance-voltage and Hall effect measurements. Au-β-SiC Schottky diodes were also fabricated on the β-SiC thin films on 4° off-axis substrates and had an ideality factor of 1.4. Differential capacitance-voltage measurements and current-voltage characteristics of Au-β-SiC Schottky diodes indicated that APB's caused significant leakage current in the β-SiC thin films.

The effect on dislocation density of elastic anisotropy (relation of growth direction to elastic stiffness tensor of cubic symmetry) and of plastic anisotropy (relation of growth direction to crystallographic slip systems of {111} ≪110≫ type) is investigated for the growth from the melt of shaped, single-crystal semiconductors of III–V compounds (typically GaAs, InP) by the Czochralski method. The thermal stresses are determined numerically by using three-dimensional finite element techniques for growth along the ≪100≫, ≪111≫ or ≪110≫ directions. The equivalent shear stress is calculated and it is assumed that the dislocation density is proportional to the sum over all slip systems of the excess of the resolved thermal shear stress over the crystal's yield stress in shear. It is shown that proper account of both elastic and plastic anisotropy is necessary in order to correlate numerical estimates of dislocation density to experimentally determined patterns. In particular, it is shown that the effect of elastic anisotropy is to produce significantly higher dislocation densities, especially for the case of ≪111≫ and ≪110≫ growth. For moderate levels of elastic anisotropy, the dislocation density levels during ≪111≫ and ≪110≫ growth become significantly larger than density, during ≪100≫ growth, at specific locations within the crystal. The correlation between dislocation density and equivalent shear stress is presented, and the effect of plastic anisotropy is thus discussed. It is shown that the assumption of elastic isotropy leads to dislocation density levels that considerably underestimate the levels but, for a given growth direction, are qualitatively similar to the dislocation density patterns corresponding to the three-dimensional numerical solutions in which elastic anisotropy is accounted in full.

To produce a homogeneous composite of silicon carbide whiskers in a silicon nitride matrix, the whiskers were formed in situ. The resulting bulk density was higher than that obtained by a physical mixing of the components using a dispersion procedure. Bending strength at room temperature of the in situ formed composite was about the same as that for the physically mixed composite when the small amount of carbon remaining in the in situ formed composite was removed.