13C cross-polarization (CP) and direct-polarization (DP) spectra of an 83 mg sample of a chemical vapor deposited (CVD) diamond film (combined from 12 separate depositions) have been obtained via dynamic nuclear polarization (DNP) combined with magic-angle spinning (MAS). With DNP, the presence of unpaired electron spins in the sample, measured to be 2 × 1018 spins/g, provides a way to enhance the 13C or the “residual” 1H signal by irradiating the sample with microwaves at or near the electron spin resonance (ESR) Larmor frequency; the interactions between the unpaired electrons and protons or 13C spins lead to a transfer of polarization from the electron spin system to the 1H and/or 13C spin systems. No signal for sp2 hybridized carbons could be observed. The DNP-CP-MAS spectrum, obtained in an experiment in which the DNP-enhanced proton polarization is in turn transferred via CP to the 13C spin system, differs significantly from the DNP-DP-MAS spectrum, in which the 13C spins are directly enhanced.

We describe two complementary studies of diamond growth by chemical vapor deposition. In the first, the early stages of growth of randomly distributed nuclei on silicon are studied by scanning tunneling microscopy. For growth times from 1 to 30 min nearly all crystallites are three dimensional, and increase in volume as t1.5. Although this result could be interpreted in terms of diffusion limited growth, the conditions for diamond CVD are more consistent with rate limited growth where the crystals are expected to gain volume as t3. This anomaly can be explained in terms of a two-species growth mechanism in which the rate constant for carbon addition is proportional to the diffusion limited flux of atomic hydrogen. Other mechanisms giving rise to the observed t1.5 dependence are also considered. The second study uses both scanning electron and tunneling microscopies to examine the morphology of a boron-doped film homoepitaxial to the {100} surface of natural type 2a diamond. In regions distant from gross defects, this film is very smooth. However, gross defects appear to initiate growth of new epitaxial layers at a rate much higher than in defect-free regions. This observation suggests that diamond growth is promoted by “enabling defects” and that without such defects nucleation of new layers is a slow process and permits layer-by-layer growth at a much lower rate.

Recent alloy development efforts have shown that Fe3Al-based alloys can have room temperature tensile ductilities of 10–20% and yield strengths of 500 MPa at temperatures to 600 °C. These property improvements are important for enabling the use of iron-aluminides for structural applications that require their excellent corrosion resistance. New data are presented here from creep-rupture studies on Fe3Al and on Fe3Al-based alloys containing molybdenum or niobium plus zirconium. Binary Fe3Al alloys have low creep resistance, but the addition of 2 at. % Mo or 1% Nb plus 0.1% Zr increases the creep life and reduces the minimum creep rate, with the niobium-containing alloy being the strongest. The improvement in creep life is the result of a combination of factors which include grain boundary strengthening, resistance to dynamic recrystallization during stressing, precipitation strengthening, and changes in the formation and mobility of the dislocation network. Correlation of optical, scanning electron, and transmission electron microscopy data suggests that the intergranular creep failure found in Fe3Al after creep testing at 550–650 °C is related to weak high-angle grain boundaries and to formation of subgrain boundary arrays, which reduce the ability of dislocations to glide or multiply to produce matrix plasticity. The addition of niobium/zirconium results in solid solution strengthening effects, as well as the formation of fine MC precipitates (a small amount of carbon is present as a contaminant from the casting process) which strengthen both the matrix and grain boundaries. The result relative to the binary alloy is increased creep-rupture strength and life coupled with a change to a ductile-dimple transgranular failure mode. This suggests that the mechanisms that cause failure during creep can be controlled by macro- and microalloying effects.

Two equations for determining the hardness of thin films from depth-sensing indentation data are examined. The first equation is based on an empirical fit of hardness versus indenter displacement data obtained from finite element calculations on a variety of hypothetical films. The second equation is based on a model which assumes that measured hardness is determined by the weighted average of the volume of plastically deformed material in the coating and that in the substrate. The equations are evaluated by fitting the predicted hardness versus contact depth to data obtained from titanium coatings on a sapphire substrate. Only the volume fractions model allows the data to be fitted with a single adjustable parameter, the film hardness; the finite element equation requires two thickness-dependent parameters to obtain acceptable fits. It is argued that the difficulty in applying the finite element model lies in the use of an unrealistic area function for the indenter. For real indenters, which have finite radii, the area function must appear explicitly in the final equation. This is difficult to do with the finite element approach, but is naturally incorporated into the volume fractions equation. Finally, using the volume fractions approach the hardnesses of the titanium films are found to be relatively insensitive to film thickness. Thus, the apparent increase in hardness with decreasing film thickness for the titanium films is most likely due to increased interactions between the film and substrate for the thinner films rather than to a change in the basic structure of the titanium films.

The preferred orientation of pitch precursor fibers and carbon fibers from an isotropic pitch was studied using x-ray diffraction techniques. The φ scan profile of a pitch precursor fiber was divided into two parts: φ scan-dependent (anisotropic) and φ scan-independent (isotropic) profiles. The half width at half maximum (HWHM) intensity of the anisotropic part became smaller as the pitch viscosity increased during spinning and as the filament diameter decreased. The area ratio of the anisotropic part to the total area increased when spinning at higher pitch viscosities and when making the diameter thinner; no preferred orientation existed for pitch precursor fibers with very large diameters. A thermodynamic model is proposed for the preferred orientation formation when spinning an isotropic pitch. The carbon fiber exhibited no skin-core difference in preferred orientation. A crystallite with a larger crystallite size Lc(002) in the carbon fiber was shown to be better aligned along the fiber axis than that with a smaller Lc(002). Furthermore, a well-aligned crystallite possessed a larger Lc(002) than one that was misaligned. The pitch precursor fiber also exhibited such a correlation between Lc(002) and the preferred orientation, but the correlation was weak.

The room-temperature and 700 °C tie lines of the early transition metal–Si–Cu ternary systems have been calculated for the metals Cr, Mo, Nb, Ta, V, and W. The tie lines are determined by considering only binary and elemental phases in observance of strict stoichiometry and zero solid solubility. We find that copper is stable with the majority of transition metal silicides under consideration at 700 °C. The result is relevant to the design of stable copper metallizations for silicon.

The role of product evaporation during lead monoxide (PbO) powder generation by aerosol decomposition (spray pyrolysis) was investigated at various temperatures in a flow reactor. Particles consisting of phase pure litharge and a mixture of litharge and massicot were formed with the dominant phase changing from litharge to massicot as the pyrolysis temperature was increased. Scanning electron microscopy showed particles produced at lower temperatures had a lumpy surface morphology and at higher temperatures appeared to be solid, indicated by the faceted surfaces and a plate-like morphology. Evaporative losses of PbOx, to the reactor walls were observed due to the substantial vapor pressure of PbOx. A simple model was developed that accounts for particle evaporation and mass transfer of lead oxide vapor to the reactor walls. This model suggested that the loss of lead oxide to the reactor walls was limited by diffusional transport of lead oxide vapor to the reactor walls.

Using elemental powders, combustion reaction was carried out to form intermetallic-ceramic composites in the Ti–Al–N system. Ti and Al powders reacted exothermically in gaseous nitrogen and formed a mixture product which had a fine distribution of the Ti2AlN particles in the matrix TiAl with a small amount of Ti3Al. Subsequently, these reacted products were arc-melted to obtain fully dense button ingots. The resulting composites had about 30 vol. % Ti2AlN, and the Ti2AlN particles were ellipsoidal or columnar in shape with sizes of 2–10 μm and appeared to be homogeneously distributed and well bonded to the matrix TiAl. It was found that such composite materials have a high strength at both room and elevated temperatures and some intrinsic compressive ductility at room temperature. Therefore, the processing technique in the present investigation is of interest as a new combustion reaction process to make intermetallic-based composite materials.

Epitaxial TiN/NbN superlattices with wavelengths/ranging from 1.6 to 450 nm have been grown by reactive magnetron sputtering on MgO(100). Cross-sectional transmission electron microscopy (XTEM) studies showed well-defined superlattice layers. Voided low-angle boundaries, aligned perpendicular to the film plane, were also present. High-resolution images showed misfit dislocations for Λ = 9.4 nm, but not Λ = 4.6 nm. Up to ninth-order superlattice reflections were observed in diffraction, indicating that the interfaces were relatively sharp. Analysis of the first-order x-ray superlattice reflection intensities indicated that the composition modulation amplitude increased and the coherency strains decreased for Λ increased from 2 to 10 nm. Vickers microhardness H was found to increase rapidly with increasing Λ, from 1700 kg/mm2 for a TiN–NbN alloy (i.e., Λ = 0) to a maximum of 4900 kg/mm2 at Λ = 4.6 nm. H decreased gradually for further increases in Λ above 4.6 nm, to H = 2500 kg/mm2 at Λ = 450 nm. The hardness results are compared with theories for strengthening of multilayers.

Phase transformation and preferred orientation in ZrO2 thin films, deposited on Si(111) and Si(100) substrates, and prepared by heat treatment from carboxylate solution precursors were investigated. The deposited films were amorphous below 450 °C, transforming gradually to the tetragonal and monoclinic phases on heating. The monoclinic phase developed from the tetragonal phase displacively, and exhibited a strong (111) preferred orientation at temperature as low as 550 °C. The degree of preferred orientation and the tetragonal-to-monoclinic phase transformation were controlled by heating rate, soak temperature, and time. Interfacial diffusion into the film from the Si substrate was negligible at 700 °C and became significant only at 900 °C, but for films thicker than 0.5 μm, overall preferred orientation exceeded 90%.

Diamond crystals have been selectively grown on the apex of anisotropically chemically etched silicon pyramids. A novel process sequence is developed which exposes a patterned sharp apex of silicon pyramids surrounded by thermally grown silicon dioxide to a high pressure microwave plasma-assisted chemical vapor deposition (HPMACVD) process where the reactant feed gases are methane and hydrogen. The nucleation rate of diamond is very high on the sharp edge of a silicon mesa structure or an apex of a silicon pyramid, as anticipated. Selective growth of diamond particles on the apex of silicon pyramids fabricated using various approaches was analyzed by scanning electron microscopy.

The effects of scanning speed and constitutional supercooling during zone-melting recrystallization (ZMR) of silicon-on-insulator (SOI) structures were studied numerically. Effects of strip motion and constitutional supercooling on the temperature profiles and the size and structure of the molten region were investigated. The temperature distribution and melt zone width does not alter for processing speeds lower than 400 μm/s. For higher speeds, motion causes the temperature profiles to lag behind the strip. The size of the molten zone decreases for increasing scanning speeds and increases for increasing degrees of supercooling. The combined effects of strip motion and supercooling on the size of the molten zone were determined. Supercooling reduced the size of the solidification region where liquid and solid phase silicon coexist.

The nucleation and growth of diamond crystals on single crystal copper surfaces has been studied. Microwave plasma enhanced chemical vapor deposition (MPECVD) was used for diamond nucleation and growth. Prior to diamond nucleation, the single crystal copper surface is modified by carbon ion implantation at an elevated temperature (∊820 °C). This procedure leads to the formation of a graphite film on the copper surface, resulting in an enhancement of diamond crystallite nucleation. A simple lattice model has been constructed to describe the mechanism of diamond nucleation on graphite as 〈111〉diamond parallel to 〈0001〉graphite and 〈110〉diamond parallel to 〈11$\overline 1$0〉graphite. This leads to a good understanding of diamond growth on carbon-implanted copper surfaces.

Isothermal and nonisothermal sintering behaviors of glass-filled ceramics are compared in this paper. Equations developed to describe the kinetics of shrinkage for both techniques can be readily used to interpret the experimental data. It is shown that the kinetics of linear shrinkage of powder compacts during isothermal and nonisothermal constant rate of heating are consistent and can be described by the viscous flow of glass.

It now appears generally accepted that the process of stress-induced orientation and graphitization of a thermoset-resin-derived matrix in a carbon-fiber/carbon-matrix (C/C) composite is principally a result of molecular orientation induced during the pyrolysis process as a consequence of the restraint of pyrolysis shrinkage at the fiber/matrix interface by attractive or frictional forces between fiber and matrix. We hypothesize that the critical factor for the formation of lamellar graphite (by subsequent high-temperature heat treatment), instead of fibrillar or isotropic glassy carbon, is a state of multiaxial deformation in the pyrolysis step. Finite-element stress analyses of the relative stresses in the regions of interfilament matrix, as the matrix pyrolyzes from polymer to carbon, reveal patterns of biaxial and triaxial stress consistent with experimental observations of lamellar graphite formation in the matrix by the techniques of optical microscopy, scanning electron microscopy, and transmission electron microscopy. The implications of localized matrix orientation and graphitization for C/C composite properties are discussed in terms of a “duplex” system of composite reinforcement. An example is presented showing crack deflection and blunting at the matrix/matrix-sheath interface produced as a result of such orientation and graphitization.

A new processing route for intermetallics starting from elemental metal powders and using laser beam heating has been investigated, by which three Al3Ti-based ternary intermetallic systems alloyed with Fe, Ni, or Cu, respectively, have been produced. Structures and compositions of five phases found in the three systems have been analyzed using optical metallography, transmission electron diffraction and microscopy, and thin foil energy dispersive x-ray spectrometry. The structure of three matrix phases in the three systems was identified as L12 type, which is consistent with previous work. The space group of the second phases in the two systems alloyed with Fe or Ni was reconfirmed or newly determined as Fm3m, using convergent beam electron diffraction. The lattice parameter measurements of these five found phases indicated atomic size-effect on the formation of intermetallics.

Interface studies of CVD Si3N4 grown on (11$\overline 1$) SiC single crystal substrates have been made by transmission electron microscopy (TEM). It is found that there are two orientation relationships both of which involve the same (10$\overline 1$0)Si3N4//(11$\overline 1$)SiC planar relationship. However, the orientation relationships are not perfect and rotations of 2–6° are commonly seen between both directions and planes involved. High resolution electron microscopy (HREM) of the interfaces shows that the SiC and Si3N4 are continuous up to the interface and that no intermediate phases are formed. However, due to the small rotations, the HREM images are difficult to interpret directly in terms of atomic positions. Nevertheless, possible atomic models of the interface are proposed based upon the experimental findings. These models exploit the similarities between the [SiN4] and [SiC4] tetrahedra in Si3N4 and SiC, respectively. The observed orientation relationships appear to be due to matching of tetrahedra across the interface along with adequate lattice matches.

Metallic iron and iron oxide particles were produced by injecting ferrocene into the afterglow region of a low pressure, low power, plasma generated using a microwave power source. This was done as part of an effort to explore the feasibility of using flow type microwave plasmas for the production of metal nanoparticles. It was found that two parameters had the largest impact on the particles: injection point and plasma composition. Analysis done using Mössbauer effect spectroscopy, transmission electron microscopy, and x-ray diffraction indicated that low yields of small particles (ca. 10 nm) resulted from injection into the afterglow region. Much higher yields of large particles (ca. 50 nm) formed if the ferrocene was injected through the coupler. In hydrogen plasmas the particles that were produced were metallic iron, whereas in oxygen and argon plasmas the particles were iron oxide. In all cases significant amounts of graphitic carbon formed around the metal particles.

Experiments have been conducted to examine the relative importance of different gas feed geometries and molecular species in the hot filament assisted vapor phase deposition of diamond. Remote and local (relative to the substrate) gas feeds of methane and acetylene were tested and it was found that although dramatic effects on uniformity and rate of deposition can be observed using a gas feed of methane local to the substrate surface, little or no variation in growth rate or deposition uniformity was produced by feeding acetylene locally. The growth rate observed on a defined area of the substrate using a local gas feed of methane was 0.50 mgms/h versus 0.17 mgms/h for a local feed of acetylene. These results, as well as the patterns of deposition observed, suggest that the major species contributing to growth in the hot filament assisted CVD of diamond is the methyl radical. This species is readily formed in the gas phase from methane and atomic hydrogen by hydrogen abstraction. It is, however, further suggested that numerous different hydrocarbon species may be of importance, with the nature of the species contributing most to the growth dependent on the method used.

The formation and characterization of zirconia coatings on graphite/carbon filaments were investigated. The objective was to eliminate or minimize degradative chemical reactions and improve bonding at the metal/carbon-fiber interface when the coated fiber is used as reinforcement in metal matrix composites. Thin, homogeneous films of zirconium oxide, ZrO2, less than 1μm thick, were formed on carbon monofilaments (35 μm diameter) from a zirconium oxychloride solution in water (less than 1.0 wt. % ZrO2) by dip coating and heating. Chemical changes during thermal decomposition and polycondensation were examined by FTIR spectroscopy. Through dynamic x-ray diffraction tests, the zirconia coating was found to transform first to a metastable tetragonal phase on heating to 330 °C, and then upon cooling, to a stable monoclinic structure. Organotitanate coatings were formed by electrodeposition of the ionizable organometallic complex, titanium di(dioctylpyrophosphate) oxyacetate (TDPA). Single fiber tests revealed a slight reduction in the strength of fibers with thicker coatings, probably due to crack initiation by brittle fracture of the ZrO2 coating. Thin coatings applied from 0.25% ZrOCl2 did not cause such strength reduction. Effective bonding of ZrO2 coatings to the filaments was revealed in single filament composite tests which showed the interfacial shear strength (IFSS) of the coated filaments in an epoxy matrix to be higher than that of the uncoated filaments. The IFSS of monofilaments electrocoated by the titanate was also higher. Examination of the fracture surfaces showed fiber pull-out associated with poor bonding in the case of specimens prepared from uncoated monofilaments. The coated monofilaments showed no fiber pull-out, suggesting that maximum fiber strength was achieved and transferred through the interface to the matrix. Finally, the chemical and thermal stability of the interfacial region of coated and uncoated graphite rods embedded in an aluminum matrix were evaluated. The uncoated rod showed little or no interfacial bonding to the metal matrix, suggesting poor wetting, while the dip-coated and electrocoated rods showed good wetting and compatibility with the metal matrix.