Silicide formation due to thermal treatment of thin (5–10 nm) molybdenum films on single-crystal, polycrystalline, and hydrogenated amorphous silicon substrates in the temperature range of 100 to 1000 °C was studied, with an emphasis on the initial interactions. The molybdenum deposition, annealing, and characterization using Raman scattering and Auger electron spectroscopy was carried out in UHV in order to minimize the effects of contaminants. Raman spectroscopy is used to distinguish between tetragonal (t-MoSi2) and hexagonal MoSi2 (h-MoSi2). The Raman spectrum of bulk tetragonal MoSi2 exhibits two prominent lines which are associated with the A1g (325 cm−1) and Eg (440 cm−1) modes. The only silicide phases detected in the thin film experiments were t-MoSi2 and h-MoSi2. While hexagonal MoSi2 does not appear in the bulk phase diagram, it is the first silicide phase formed in thin film reactions at a temperature between 300 and 400 °C. The nucleation temperature of h-MoSi2 was the same for Si〈100〉, Si〈111〉, and amorphous Si. Indirect evidence for disordered intermixing of silicon and molybdenum before nucleation of h-MoSi2 is found. Annealing at approximately 800 °C causes the silicide to transform from the hexagonal phase to the tetragonal phase for all substrates. Contaminants interfere with the formation of h-MoSi2 and also retard the transformation of h-MoSi2 to t-MoSi2. For the thin films considered here, the transformation to t-MoSi2 is accompanied by islanding of the silicide film. A lower interfacial energy between the silicon and silicide for h-MoSi2 has been proposed to explain the nucleation of h-MoSi2 before t-MoSi2.

In undoped liquid-encapsulated Czochralski-grown semi-insulating gallium arsenide (LEC SI GaAs), the carbon acceptor, CAS, is responsible for the ∼1.49 eV emission due to donor-acceptor (D–A) pair radiative recombination. We have studied the effect of different carbon concentrations on the line shape and decay kinetics at and near the D–A emission peak for temperatures ranging from 4.2 to 34 K. The photoluminescence decay has been found to follow a double exponential law at the early stage of the decay but follows a power law of the form L(t)αT−p at a later stage. Analysis of the temperature dependence of the exponential components gives two activation energies of ≃5.7 and ≃15 meV. We find that the 5.7 meV is most likely associated with SiGa shallow donor, while the 15 meV is due to a more indirect process involving an EL2 related donor cluster near 20 meV. The exponent, p, of the power-law decay is found to be both temperature (T) and concentration dependent of the form p = βT −1, where β is a linear function of the concentration of the acceptors, CAs, in the temperature range of 18  T < 30 K. A promising, simplified model based on a continuous set of trapping levels is presented and compared to the experiments. Some aspects of the observed exponential and power law decays are predicted by the theory.

An analysis of nucleation from solid nuclei in Polix photovoltaic polycrystalline silicon is reported. The bulk texture is studied by neutron diffraction using a transmission method. The crystallographic orientation is determined for three adjacent samples along the growth axis in the lower central part of the ingot. The size and orientation of all crystallites are obtained in the domains showing different morphologies: a nucleation zone with equiaxed crystallites, an upper domain with columnar structure, and an intermediate part where both types of crystallization are present. The crystallites are counted in terms of their volume in the three parts. The mean grain size increases in the columnar domain, but a pronounced elimination of particular crystalline orientations is not observed as the morphology is changed.

In this paper we report on the growth of pseudomorphically strained Si1−xGex alloys on 〈001〉 Si by solid phase epitaxy. One set of amorphous alloys was formed by high dose ion implantation 74Gc implanted at an energy of 200 kcV to a fluence of 9.6 ⊠ 1020/m2). Our TEM observations show that regrowth of these Si1−xGex(xmax = 0.14) films at ≍590°C results in a high density of planar defects and that these defects are associated with faceting of the amorphous/crystalline interface during annealing. These results were compared with the solid phase regrowth of MBE-grown Si0.7Ge0.3 amorphized with 170 keV 28Si ions which exhibited identical defects and faceting during regrowth. Attendant with this faceting was a decrease in the regrowth velocity, a result of a change from a planar {001} growth morphology to a multi-faceted growth surface containing many <50 nm deep pyramidal impressions. The regrowth rate was quantified, at a particular temperature, by the use of Si homoepitaxy for calibration of in situ TEM experiments. It was shown that the regrowth rate at 594°C in pure Si was 51 nm/min, whereas in the Si0.7Ge0.3 the regrowth rate decreased, as a result of {111} faceting, to 21 nm/min. RBS was used to characterize Ge concentrations and lattice resolution TEM was used to study the development of the faceted interface and associated planar defects during regrowth.

The microstructures of carbon-rich regions in chemically vapor deposited SCS-6 SiC fibers have been studied by electron diffraction and high resolution transmission electron microscopy. The microstructures of the carbon in the substrate filament and the outer coating layers of the fiber are consistent with different distributions of Oberlin's model of Basic Structural Units (BSUs)1 while that of the inner substrate coating is consistent with turbostratic carbon (TC).

The finding of Jang and Koch that the melting enthalpy/mass, ΔHm, of a Sn matrix containing a fine dispersion of Ge particles sharply decreases with increasing Ge volume fraction, νGe, >0.5 and vanishes at νGe = ν°Ge ≍ 0.81, is accounted for by supposing that the Sn is distributed between an interfacial and bulk state. The interfacial statc is one in which the Sn is assumed to be in a disordered, possibly amorphous, structure coating the Ge particles uniformly to a constant thickness, δ. The remaining “bulk” Sn is assumed to exhibit the normal enthalpy of fusion, ΔH°m. The model accounts for the dependence of ΔHm on νGe within the experimental uncertainty. With the average width of Ge particles −10 nm, δ is estimated to be −0.23 nm; i.e., of the order of the thickness of one Sn monolayer.

Aluminum nitride (AlN) is currently under investigation as a potential candidate for replacing aluminum oxide (Al2O3) as a substrate material for electronic circuit packaging. The requirements for such a material are that it can be metallized and joined to produce hermetic enclosures for semiconductor devices. A technique for brazing AlN using a nonactive metal braze has been investigated. The process involves the in situ decomposition of an active metal hydride. This process improves the wetting of the AlN and led to the development of strong bonding between braze and ceramic. The ceramic-braze interface was studied using scanning electron microscopy (SEM). The nature of the interfacial reactions and the reaction products have been identified using x-ray diffraction (XRD). The progress of the reaction has been followed using differential thermal analysis (DTA). The experimental results have been correlated with thermodynamic predictions of the reaction processes. In addition to joining ceramic to ceramic, braze joints of AlN to copper and to a low expansion iron-nickel lead frame alloy were made. Residual stress resulting from a mismatch of thermal expansion coefficients between AlN and copper caused cracking in the ceramic upon cooldown from the brazing temperature. No cracking occurred in the ceramic when joined to the iron-nickel alloy. The results obtained are important for the realization of AlN as a ceramic packaging material for semiconductor devices.

A new parameter, hardness/modulus2 (H/E2), has been derived from the equations used to calculate the hardness and elastic modulus from data taken during continuous depth-sensing microindentation tests. This paper discusses the use of this parameter to treat the data obtained from a sample whose surface roughness was of the same scale as the size of the indents. The resulting data were widely scattered. This scatter was reduced when the data were plotted in terms of H/E2 versus stiffness. The effect of surface roughness on the hardness and elastic modulus results is removed via stiffness measurements, provided single contacts are made between the indenter and the specimen. The function relating the cross-sectional area of the indenter versus the distance from its point is not required for calculation of H/E2, but the hardness and modulus cannot be determined separately. The parameter H/E2 indicates resistance to plastic penetration in this case.

Silicone gels are becoming more accepted as protective coatings for Very Large Scale Integrated circuits (VLSI) against severe environments due to their excellent electrical, thermal, and mechanical properties. Recent studies indicate that high performance silicone gels in low-cost, non-hermetic plastic packaging may replace conventional hermetic ceramic packaging. This paper describes the use of the soft silicone gels as coatings on Integrated Circuit (IC) devices, and the correlation between the material's cure temperature and cure time versus their adhesion and electrical reliability during 85°C, 85% RH and bias accelerating testing. In addition, the reaction kinetics of the silicone gel based on the Differential Scanning Calorimetry (DSC) study of the uncured sample will be reported.

Mixed carbon-12/carbon-13 diamond films were synthesized by hot-filament chemical vapor deposition, using mixtures of 13CH4 and 12CH4 or 12C2H2 in H2. The first-order Raman peak at 1332 cm−1 for 12C-diamond was found to shift by 50 cm−1 to 1282 cm−1 for pure 13C-diamond. For mixed-isotope films, the Raman peak frequency shifts linearly between these values as a function of the 13C mole fraction. The mechanism of diamond film growth by hot-filament CVD has been investigated by growth from mixtures of 13CH4 and 12C2H2, using the shifted Raman frequency to determine the relative incorporation rates of 13C and 12C into the film. The 13C mole fraction in the film agrees closely with the 13C mole fraction inferred for the methyl radical but differs substantially from that of acetylene, indicating that the methyl radical is the dominant growth precursor under the conditions studied.

In this paper we report the influence of process variables, viz., substrate temperature, oxygen partial pressure, and external electric field bias, on phase precipitation and microstructure of tin oxide films as revealed by small-angle x-ray diffraction and conversion electron Mössbauer spectroscopy.

Tin-lead (40 wt. % Pb) solder-matrix composites containing 8–54 vol.% continuous unidirectional copper plated carbon fibers were fabricated by squeeze casting for use as low thermal expansion solder preforms. The low thermal expansion greatly increased the thermal fatigue life for solder joints between materials with low thermal expansion coefficients. For example, for 29 vol.% fibers, the thermal expansion coefficient was 8 ⊠ 10−6/°C (25–105°C) in the direction parallel to the fibers compared to a corresponding value of 24 ⊠ 10−6/°C for plain solder. The thermal fatigue life for cycling 2 cm long alumina-to-alumina solder joints between 25 and 100°C was increased from 98 to 183 cycles by using 29 vol.% carbon fibers in the composite solder. The fibers also increased the tensile modulus and tensile strength of the solder, but the ductility was decreased. The copper coating on the carbon fibers increased the tensile strength and ductility of the composite.

We have studied the structural and electrical properties of ultrathin cobalt silicide films deposited on Si(211) and (311) and annealed at a variety of temperatures. Transmission electron microscopy reveals two regimes of silicide growth. If more than 10 Å of Co is deposited, textured CoSi forms at low temperatures, which transforms to two epitaxial orientations of CoSi2 at about 500°C. If less than 10 Å of Co is deposited, only one epitaxial CoSi2 orientation is observed, which occurs in discontinuous islands or lines. The resistivities of the thinnest films reflect the degree of continuity of the silicide layers. In thicker films, where disorder is the major factor, the resistivity for intermediate annealing temperatures (∼500°C) depends dramatically on the orientation of the Si substrate, being an order of magnitude lower for Si(311) and (211) than for (111). The differences between these observations and those made on films grown under the same conditions on Si(111) can be understood by noting both the greater tendency of the film to form intermediate phases which are epitaxial on the high symmetry Si(111) surface and the faceting of the CoSi2-Si(311) and (211) interfaces.

Improved structural quality Hg1−xCdxTe epitaxial films have been grown by metal organic chemical vapor deposition (MOCVD) using the interdiffused multilayer process (IMP) on lattice-matched CdZnTe substrates at temperatures above 400°C with diethyltelluride. Double-crystal rocking curve data with values as low as 25 arcsec correlated with cross-sectional transmission electron micrographs. The process of growing alternate layers of HgTe/CdTe for improved compositional uniformity, and the inherent mismatch, was not detrimental to the structural quality, which was limited by that of the substrate.

The surface morphology of SIMOX wafers implanted at 180 keV with doses of 0.4–2.2 ⊠ 101816O+ cm−2 in a temperature range of 400–700 °C has been investigated using transmission electron microscopy (TEM) replica and cross-sectional TEM (XTEM) techniques. Wafer temperature during oxygen implantation strongly affects the morphology. A number of dents are formed on the surface of wafers implanted at temperatures higher than 510 °C with a dose of 1.8 ⊠ 1018 cm−2. Increasing the wafer temperature causes the dents to grow. The dents disappear by a high-temperature anneal of 1260 °C after the implantation. It is found that oxygen implantation through a 50-nm-thick screen oxide film prevents dent formation. A model explaining the dent formation and dent growth is also proposed.

This article proposes a new scheme for fabricating homogeneous Al2O3–ZrO2 composite, in which a thermodynamic theory of interfacial electrochemical phenomena is applied. The theory predicts that a heterogeneous Al2O3 interface in colloidal dispersion can induce an enhanced concentration of the ionic species needed for a selective formation of the ZrO2 precursor at the Al2O3/aqueous solution interface. Based on this proposition, a homogeneous Al2O3–ZrO2 composite powder was fabricated by a surface-induced coating of the fine ZrO2 precursor on the kinetically stable colloid particles of Al2O3. The composite prepared by the surface-induced coating was characterized by a uniform spatial distribution of the dispersed ZrO2 phase and by the absence of large ZrO2 grains formed from hard ZrO2 agglomerates. The composite also showed highly uniform grain size distribution of both the dispersed ZrO2 and the matrix Al2O3 phases. The uniform grain size distribution of the matrix phase indicates that the homogeneous coating of the fine ZrO2 particles is effectively pinning the Al2O3 grain boundaries.

Shallow n-p GaAs solar cells have been made by implantation of Si into Zn-doped (p-type) GaAs substrates followed by rapid thermal annealing. The structure of the GaAs crystal has been determined by the DSL photoetching method (Diluted Sirtl-like etchants used with Light). It was found that implantation-induced-damage (revealed by DSL as microroughness and craters) was not removed after annealing for energies exceeding 60 keV. This leads to substrates that contain many precipitates, which appears to be disastrous for the fabrication of good solar cells. In addition, good cell performance is hampered by compensation effects in the n-p transition region and in the n-type layer itself.

The ICECAP methodology is used to derive interionic potentials of some cation and anion impurities in MgO, namely, Li+, Na+, K+, Be2+, H−, S2−, and O2−. Analysis is given of the defect energies obtained by using the derived impurity potentials. Based on the available experimental data, comparison is made to justify the reliability of the derived impurity potential for Be2+. The calculated activation energy for Be2+ diffusion comes out to be 1.54 eV as compared to the experimental value of 1.60 eV, which is considered to be very satisfactory.

The reduction of propagation losses in ZnO thin film waveguides for integrated optics after CO2 laser treatment is studied. RHEED patterns show that the measured decrease of optical losses is accompanied by an increase of ordering after the laser treatment. The reordering depends on the initial film state and the laser irradiation conditions.

We have used transmission electron microscopy techniques to study the nature of dislocations, stacking faults, twins, and grain boundaries in CVD (chemical-vapor-deposition) diamond thin films. Perfect a/2〈110〉 and partial a/6〈112〉 and a/3〈111〉 type dislocations are observed; the partial dislocations are also associated with twins and stacking faults. The most common defect in diamond thin films, particularly in 〈110〉 textured films, is Σ = 3 grain boundary or the primary twin. These twins in 〈110〉 textured films can lead to formation of fivefold microcrystallites. We have also investigated the splitting of Σ = 9 grain boundary (second order twin) into two Σ = 3 boundaries or primary twins via reaction Σ9 = 2Σ3. A rapid thermal annealing treatment has been shown to result in annealing of Σ = 3 boundaries and produce “defect-free” regions in thin films. A mechanism of annealing (removal) of Σ = 3 boundaries is discussed. Atomic structure and energetics of dislocations, twins, and grain boundaries are calculated using Tersoff potentials. The calculated atomic structure for Σ = 3 boundary is compared with high-resolution TEM images and a good agreement is obtained. These boundaries consist of periodic units of 5–7 rings which are similar to the core structure of 90° a/2〈110〉{001} dislocations. The energy of the 5–7 rings in the grain boundaries is considerably lower, due to overlapping and strain cancellation effects, than that associated with single dislocations. The 5–7 ring energy and consequently the boundary energy increases as the overlapping effects decrease. An interesting analogy between the diamond and silicon results is drawn.