Nanocrystalline materials have a grain size of only a few nanometers and are expected to possess very high hardness and strength values. Even though the hardness/strength is expected to increase with a decrease in grain size, recent observations have indicated that the hardness increases in some cases and decreases in other cases. A careful analysis of the available results on the basis of existing models suggests that there is a critical grain size below which the triple junction volume fraction increases considerably over the grain boundary volume fraction and this is suggested to be responsible for the observed softening at small grain sizes.

The ternary site occupancy of two alloys, vanadium in NiAl + V and hafnium in nickel-rich, boron doped Ni3Al + Hf, was determined by ALCHEMI, or atom location by channeling enhanced microanalysis. Vanadium exhibited a preference for the aluminum sublattice in NiAl, and hafnium preferentially occupied the aluminum sites in Ni3Al. Spectra were acquired over a range of accelerating voltages from 80 kV to 200 kV. Delocalization effects were observed to increase as the accelerating voltage increased, which thus reduces the accuracy of the ALCHEMI data. For NiAl + V, both planar and axial channeling were performed, and delocalization effects were greater for axial channeling, further reducing the accuracy of the ALCHEMI data.

The change in the surface morphology of metallic glasses Fe78B13Si9 during the early stage of the crystallization process has been studied mainly by atomic force microscopy (AFM). Specimens of 50-μm thickness have been heated up to 323 K, 373 K, 423 K, and 523 K with a heating rate of 30 K/min in vacuum, and then the surface morphology of each specimen has been observed by AFM in air. In the surface image of the specimen heated to 323 K, many holes are observed and inside the holes single or plural protrusions can be observed. Clusters composed of aggregated protrusions are also found like islands in the amorphous sea. The specimen heated to 323 K has also been observed by a high-resolution transmission electron microscope and crystalline structures among the amorphous matrix have been detected. From the selected area diffraction study, the crystalline structure is found to be α–Fe crystallites. In contrast, in the AFM image of the surface of the specimen heated to 423 K, no holes are seen and many protrusions are found to extend above the surface and form several parallel lines. The spatial density of protrusions above the surface becomes much higher in the specimen heated to 523 K. We propose that this change in the surface morphology during heat treatments indicates the process of nucleation and growth of α–Fe crystallites in the surface of metallic glasses. It is also found that the stage of many protrusions extending above the surface corresponds with the beginning of exoelectron emission from the surface. This result suggests that exoelectron emission and surface crystallization connect with each other.

A fine-grained ($\overline \iota$m) laminate containing 91.0 wt.% magnesium and 9.0 wt. % lithium was prepared by a foil metallurgy technique involving rolling and pressing at low homologous temperature (0.39–0.49 Tm). The processed material exhibits superplastic characteristics above 70 °C (0.40 Tm). The strain-rate-sensitivity exponent is about 0.5 and an elongation-to-failure of 450% was obtained at 100 °C (0.43 Tm). The activation energy for plastic flow in the superplastic region is 65 kJ/mole. This value of the activation energy is related to the expected activation energy for grain boundary diffusion.

Molecular species were found to diffuse readily into the octahedral interstitial sites of the fcc lattice of C60. The 13C NMR spectrum of C60 under magic angle spinning (MAS) conditions consisted of a primary resonance at 143.7 ppm and a minor peak shifted 0.7 ppm downfield. The downfield shift obeys Curie's law and is attributed to the Fermi-contact interaction between paramagnetic oxygen molecules and all 60 carbon atoms of rapidly rotating adjacent C60 molecules. Exposure of C60 to 1 kbar oxygen for 1.75 h at room temperature resulted in a spectrum of seven evenly spaced resonances corresponding to the filling of 0 to 6 of the adjacent octahedral interstitial sites with oxygen molecules. The distribution of site occupancies about a C60 molecule provided evidence that the intercalation process is controlled by diffusion kinetics. Exposure to 0.14 kbar hydrogen gas at room temperature for 16 h filled a substantial fraction of the interstitial sites of C60 and C70 with hydrogen molecules.

The role of SiO2 in nucleation of diamond has been investigated in an oxy-acetylene flame. It was found that growth methods that minimize SiO2 formation enhance diamond nucleation. A short pretreatment of a scratched Si surface in a low oxygen-to-acetylene ratio flame, at a distance 1.5 cm from the flame core, significantly improved uniformity of subsequent diamond growth. When scratched surfaces were intentionally oxidized, nucleation of diamond was completely inhibited. By using a mask to controllably deposit SiO2 on a scratched Si surface, highly selective deposition of diamond was achieved with resolution below 5 μm. These results are discussed with reference to competing oxidation and carbon formation processes that take place during the nucleation of diamond. During the nucleation stage, carbon may be deposited on the scratched Si via a route in which the Si surface catalyzes carbon formation reactions that are otherwise kinetically unfavorable. The formation of an oxide layer, on the other hand, would act to passivate the surface, and thus inhibit carbon formation via a catalytic route. The decomposition of CO to C and CO2 is given as an example of a reaction that is favored at temperatures below 1000 K, but requires surface catalysis to proceed because it remains frozen out in the gas phase due to a very slow reaction rate.

A powerful micro SIMS technique coupled to a computer driven acquisition system has allowed the simultaneous recording of C−, MoO−, and Si− images of the sample surfaces, or of the transverse cross sections of the Mo-diamond interface. Diamond deposition has been shown to take place on a Mo2C layer, and the influence on the nucleation process of Si contamination, coming from the quartz tube etched by H atoms, has been investigated. Contamination can in fact occur during the shutdown procedures or during the whole experiment. This last contamination can be avoided by using suitable pressure ranges or gas combinations. Moreover, the deposition time necessary to obtain well-crystallized diamond films and the nucleation density could be optimized by an in situ pretreatment stage. This treatment reduces the delay observed before nucleation (which would correspond to the carbide formation), and increases the carbon activity at the sample surface.

PECVD reactors are increasingly used for the manufacturing of electronic components. This paper presents a reactor model for the deposition of amorphous hydrogenated silicon in a dc glow discharge of Ar–SiH4 The parallel-plate configuration is used in this study. Electron and positive ion densities have been calculated in a self-consistent way. A macroscopic description that is based on the Boltzmann equation with forwardscattering is used to calculate the ionization rate. The dissociation rate constant of SiH4 requires knowledge about the electron energy distribution function. Maxwell and Druyvesteyn distributions are compared and the numerical results show that the deposition rate is lower for the Druyvesteyn distribution. The plasma chemistry model includes silane, silyl, silylene, disilane, hydrogen, and atomic hydrogen. The sensitivity of the deposition rate toward the branching ratios SiH3 and SiH2 as well as H2 and H during silyl dissociation is examined. Further parameters that are considered in the sensitivity analysis include anode/cathode temperatures, pressure, applied voltage, gap distance, gap length, molar fraction of SiH4, and flow speed. This work offers insight into the effects of all design and control variables.

The existence and imperfection of several types of epitaxial layers and films on the substrate crystals were successfully detected and characterized nondestructively using a new interference fringes method. It was found that the interference fringes generated by amorphous thin films showed many very tiny bright spots that became discontinuous, which was quite different from the case of fringes generated by a crystalline epitaxially grown layer (smooth fringes). Multiple layers on a substrate were also observed by the fringes that were not concentric, but interlacing.

A study of the cross doping of GaAs layers grown by a two-step metalorganic chemical vapor deposition on Si substrates is reported. All as-grown, unintentionally doped layers of GaAs were n-type, and the carrier profiles tracked the Si atomic profiles. Furnace annealing at 850 °C for 30 min in an arsine overpressure, which is used to improve the crystalline quality of the GaAs near the heterointerface, caused additional Si to diffuse into the GaAs layer. Comparison of the Si concentration at the interface with the carrier concentration suggested the presence of compensating acceptors. Resonance Raman scattering by the SiAs local vibrational mode near the interface shows that a fraction of the Si atoms are localized at the As sites. The furnace annealing increased the Si concentration in the 1.7–1.8 μm thick initially grown GaAs layer. This, in turn, influenced the electrical profiles created with Si or Be implantation on a 2.3 μm thick GaAs layer.

The MBE growth of InxGa1−xAs (x ∼ 0.53) on silicon substrates has been investigated emphasizing the effects of substrate orientation and buffer layers between In0.53Ga0.47As and Si. It is shown that growth on silicon substrates misoriented from (001) toward a [110] direction eliminates the presence of antiphase domains. The best In0.53Ga0.47As surface morphology was obtained when a 0.9 μm epitaxial Si buffer was initially grown, followed by a pre-exposure of the silicon surface to As4 at 350 °C, followed by the growth of In0.53Ga0.47As. Threading dislocations, stacking faults, low-angle grain boundaries, and spinodal decomposition were observed by TEM in the InGaAs layers. The spinodal contrast scale was shown to depend on the buffer type and the total InGaAs thickness. Thick buffers consisted of GaAs or graded InxGa1−xAs layers, and large In0.53Ga0.47As thicknesses favor the development of a coarse-scale spinodal decomposition with periodicity around 0.1 μm. Thin GaAs buffers or direct In0.53Ga0.47As growth on Si may result in a fine-scale decomposition of periodicity ∼10 nm. The principal strain direction of the spinodal decomposition appeared along the [1$\overline 1$0] direction, parallel to the vicinal Si surface step edges. InGaAs immiscibility affects the InGaAs growth process, favoring a 3-D growth mode. X-ray diffraction measurements and photoreflectance spectra indicated that the sample quality was improved for samples exhibiting a fine-scale spinodal decomposition contrast even if they contained a higher dislocation density. Threading dislocations run almost parallel to the [001] growth axis and are not affected by strained layers and short period (InAs)3/(GaAs)3 superlattices. The lowest double crystal diffractometry FWHM for the (004) InGaAs reflection was 720 arc sec and has been obtained growing InGaAs directly on Si, while the lowest dislocation density was 3 × 109 cm−2 and was obtained using a 1.5 μm GaAs buffer before the In0.53Ga0.47As deposition.

The direct energy gaps, Eg, and the indirect gaps at the X point, E(X), of GaAs and AlxGa1−xAs alloys are essentially linear functions of hydrostatic pressure, P. Recent photoluminescence measurements of Tozer et al. for InP under high pressures, however, found that Eg(P) is not quite linear, but bends down slightly at high pressures. Using the first-principles pseudofunction method, we have calculated Eg and E(X) as functions of pressure, as well as the zero-temperature equation of state P(V). Our calculated gap curve for InP, Eg(P), bends down slightly, as found in photoluminescence studies. The slope dEg/dP is 8.8 meV/kbar for small pressures P, and is in good agreement with the experimental value, 8.32 meV/kbar. The observed nonlinearity in the dependence of Eg on pressure for InP is attributed to a large derivative of the bulk modulus with respect to pressure. The calculated bond length, bulk modulus, and critical pressure for a phase transition from the zinc blende to a rocksalt structure, and the unit cell volume change at this phase transition are all in good agreement with the data.

We have calculated the mobility of electrons in n-type Mercury Cadmium Telluride (MCT) and compared it to a calculation of the mobility of electrons in n-type Mercury Zinc Telluride (MZT) with nearly the same energy gap and with the same number of donors and acceptors. We also compared the results of the MZT calculation with experiment. We found for equivalent energy gaps that the mobilities in the two compounds (MCT, MZT) were nearly the same. The calculations for both MCT and MZT were based on the best set of material parameters that we could compile from the literature. Using these parameters, the comparison with experiment for MZT yielded good results. Since MZT is harder and structurally more stable with respect to Hg retention than MCT, the possibility of equivalent mobility for MCT and MZT is significant. This calculation is one of the first extensive calculations of the mobility of MZT, and we compared it with another which appeared to be less extensive. Our calculation involves scattering of the electrons by longitudinal optic phonons, acoustic phonons, ionized impurities, holes, and compositional disorder. Since not all of these interactions can be approximated by elastic scattering, the corresponding Boltzmann equation was solved by a variational principle. We also discuss directions for future work.

Optical, electron microprobe, and x-ray diffraction analysis of 88 samples of various compositions between Ag2S and Ag2Se synthesized at high temperature in sealed quartz tubing indicates the presence of two solid-solution series in this system at ambient (room) conditions. One series extends from Ag2S to approximately Ag2S0.4Se0.7 and has the Ag2S-III-type structure (monoclinic). The second series ranges from Ag2S0.3Se0.7 to Ag2Se and is characterized by the Ag2Se-II-type structure (orthorhombic). Members of both series, in appropriate proportions, characterize the apparent compositional gap between the two solid solutions. Gradual shifts in the locations of the x-ray diffraction peaks along the compositional gradient of each solid solution revealed an expansion of the d-spacing as the larger Se ion was substituted for S in the Ag2S-III-type structure and a contraction as S was substituted for Se in the Ag2Se-II-type structure. The reported discrete phase, Ag4SSe (aguilarite, orthorhombic), appears to be simply a member of the monoclinic Ag2S-III-type solid solution.

The microstructure of LaB6-base thick film resistors was investigated by cross-sectional transmission electron microscopy. The specimens were prepared by a technique that polished them to a thin wedge, thus avoiding ion-milling and permitting imaging over a distance of tens of microns. The resistor microstructure contained a finely divided electrically conductive phase of TaB2 and nonconducting crystals of CaTa4O11, formed during high temperature processing of glass and LaB6 ingredients of the thick film ink. Using higher surface area ingredients virtually suppressed the formation of CaTa4O11 crystals, and the microstructure became more uniform. Resistors made with higher surface area intermediates also had better voltage withstanding properties.

Sol-gel solutions prepared from tetraethylorthosilicate (TEOS) hydrolyzed in acidic conditions were used to make thin films on Si wafers via spin coating. Solutions with H2O/TEOS ratios of 2, 5, and 10 were prepared and aged at room temperature in sealed containers for times ranging from <2.5 h to 1 month. Transmission Fourier transform infrared spectra acquired around 1 min after spinning showed peaks characteristic of both SiO2 and SiOH. Additional features at ≍580, 1093, and 1130 cm−1 were observed and possible origins are discussed. Peaks at ≍580 and 1130 cm−1, which increased in intensity with solution age time, are attributed to cyclic Si–O, e.g., 4-member rings. When films were exposed to room air, ring concentration decreased significantly within 30 min. Rings were stable for hours, however, when films were kept in N2. The feature at 1093 cm−1, most sensitive to H2O/TEOS, is attributed to residual Si–O–C.

Acoustic emission (AE) behavior at scratch test has been studied for TiN coatings that deposited in an ion plating system. The AE intensity of the coatings measured by the scratch test depended not only on the adhesion of a coating-substrate interface but also on the intrinsic properties of a deposited film. In this work, a careful separation of coating parameters was investigated to clarify the relationship between the AE intensity and the intrinsic properties of coated films. For identical coating conditions, the AE intensity increased with coating thickness since the elastic energy released during the scratching is proportional to film thickness. It has been found that the coatings prepared at higher substrate temperature and/or at higher substrate bias voltage had a larger crystallite size and lower internal stress. With increasing bias voltage and substrate temperature, the AE intensity of TiN coatings decreased. An explanation for this behavior has been proposed in terms of the internal stress and the toughness of the deposited films.

Using the measured elastic constants of TiN and NbN single crystals with cubic symmetry, the effective elastic constants of single-crystal TiN/NbN superlattice films with tetragonal symmetry, namely c11, c12, c13, c33, c44, and c66 have been calculated for various thickness ratios of the layers. Using a line-focus acoustic microscope, measurements of surface acoustic waves (SAWs) have been carried out on single-crystal TiN/NbN superlattice films grown on the (001) plane of cubic crystal MgO substrates. The phase velocities measured as functions of the angle of propagation display the expected anisotropic nature of cubic crystals. Dispersion curves of SAWs propagating along the symmetry axes have been obtained by measuring wave velocities for various film thicknesses and frequencies. The SAW dispersion curves calculated from the effectiveelastic constants and the effective mass density of the superlattice films show very good agreement with experimental results. The results of this paper exhibit no anomalous dependence of the elastic constants on the superlattice period of TiN/NbN superlattices.

Ultrafine amorphous Si/C/N ternary powders have been prepared in a CO2 laser assisted process. We demonstrate the possibility to drive the C/N ratio in the powder by properly choosing the experimental conditions and the gaseous reactant ratio in the initial mixture containing silane, dimethylamine, and ammonia. A kinetic model that accounts for reaching the equilibrium between the gaseous reaction intermediates and the solid products is proposed.

The reaction kinetics involved in the continuous precipitation of monodispersed spherical yttrium basic carbonate particles was studied. The amount of powder produced was a linear function of H2CO3 generated by the decomposition of urea in mildly acidic solutions. A model equation, containing an empirical constant F, was derived to describe the conversion yield as a function of reactant concentrations, reaction time, and temperature. The calculated and experimental data agree well for both the batch and continuous processes.