ALBERT

Notes: Posthydrogenation of a-Si produced by low-pressure chemical-vapor deposition was investigated using an internal lamp which can dissociate hydrogen molecules directly. After hydrogenation, the photo–to–dark-current ratio was increased to greater than 4.5×104. The hydrogen in the hydrogenated films was mainly bonded as Si-H, and the penetration depth was about 300 nm. The posthydrogenated thin-film transistor had about 7 orders on/off current ratio and its electron and hole mobilities were 0.1 and 0.01 cm2/V s, respectively.

Notes: The thermal decomposition of decaborane (B10H14) and its doping effects on Si(111)-(7×7) has been investigated by surface spectroscopies. Upon adsorption between 100 and 300 K, molecular decaborane was identified on the surface by high-resolution electron-energy-loss spectroscopy (HREELS) by the absence of Si-H surface species production. The thermal decomposition of adsorbed decaborane molecules at higher temperatures involves a preferential removal of hydrogen from the weaker B—H—B linkage. H2 thermal desorption was observed to cover a wide temperature range between 300 and 900 K. Clean boron deposition on the surface was achieved at ∼900 K. Upon heating to ∼1275 K, extensive boron diffusion into bulk silicon produced a highly B-doped region below the surface (∼103 A(ring)) with a carrier hole concentration on the order of ∼1019 cm−3 depending upon the initial surface boron coverage and annealing conditions. The surface adopted a ((square root of)3×(square root of)3)R30° reconstruction with a nominal 1/3 ML boron occupying subsurface substitutional sites. Both the localized B-Si vibration and carrier surface plasmon excitation were observed by HREELS at 100 K.

Notes: This article investigates the epitaxial regrowth of n-type and p-type polycrystalline silicon (polysilicon) layers deposited on silicon, which are subjected to either a single emitter diffusion or consecutive base and emitter diffusions. A wide range of diffusion conditions is considered, covering both rapid thermal and furnace diffusion in the temperature range 950–1200 °C. The sheet resistances of single-diffused n-type polysilicon layers are significantly higher than those of double diffused layers for rapid-thermal emitter diffusions in the temperature range 1025–1125 °C. This is explained by the epitaxial regrowth of the polysilicon during the emitter diffusion, caused by the partial break-up of the interfacial oxide during the base diffusion. In contrast the sheet resistances of single- and double-diffused p-type polysilicon layers are found to be similar. Rutherford backscattering spectra are presented which show that the structures of the single- and double-diffused polysilicon layers are similar. This is explained by the effect which fluorine, incorporated into the polysilicon during the BF2 emitter implant, has in accelerating the break-up of the interfacial oxide during the early part of the emitter diffusion. Estimates are made of the time to break up the interfacial oxide layer and the time to vertically epitaxially align the polysilicon at different temperatures, and activation energies of 4.9 and 4.7 eV, respectively obtained. In n-type polysilicon, the epitaxial regrowth is dominated by the time to break up the interfacial oxide layer, whereas in BF2 implanted p-type polysilicon it is dominated by the time to vertically epitaxially align the polysilicon. A vertical epitaxial alignment rate of 1000 A(ring)/s is obtained for n-type polysilicon at 1050 °C, compared with 240 A(ring)/s for p-type polysilicon at 1100 °C.

Notes: Results of measurements of the hole recombination process at DX centers related to group-IV (silicon) and group-VI (tellurium) donor elements in AlxGa1−xAs are presented. For the DX(Si) and DX(Te) defects the values of hole capture cross section σh=2×10−13 and σh=10−13 cm2 were found, respectively. In both cases no systematic change of σh with temperature was observed. The results show that the DX center in the ground state is negatively charged as a consequence of its negative-U two-electron character.

Notes: Detailed transmission and scanning electron microscopic images of electromigration-induced open circuit failures are presented for fine line aluminum alloy thin film interconnects. A characteristic slit open circuit, similar to stress migration open circuits in narrow interconnects, is shown for various film compositions, processing, and deposition conditions. It is suggested that slit failure morphology is more generally observed for low (≈1) ratios of conductor line width to film grain size. The slit failures observed often occur near copper rich precipitates. The morphology of several slit voids suggests that they are transgranular across the linewidth, consistent with other recent reports of electromigration induced damage in single crystal interconnects.

Notes: We have studied the Wannier–Stark effect in GaAs/GaAlAs short-period superlattices under applied electric field perpendicular to the layers by room- and low-temperature photocurrent measurements. The changes in the transition intensities with biasing are well fitted to a theoretical calculation based on the finite Kronig–Penney model on which the potential of an applied electric field is superposed. With increasing electric field, the 0h peak grows to a maximum while the −1h and +1h peaks monotonousely decrease. By a comparison of the spectra measured at different temperatures, the two peaks in the room temperature photocurrent spectra at relatively low electric field (1.0×104 V/cm) are identified to be caused by the Wannier localization effect instead of saddle-point excitons.

Notes: An empirical formula for the binding energy of an electron-heavy-hole exciton is proposed which modifies the known formula for the Wannier exciton binding energy in the case when the well width is smaller than the bulk diameter of the exciton. This formula is utilized in calculations of the peak wavelengths in photoluminescence spectra which correspond to the ground-state electron-heavy-hole exciton recombinations in Al0.3Ga0.7As/GaAs single quantum wells of various widths.

Notes: High energy B implantations were performed into n-type GaAs and InP at room temperature in the range of energies from 1 to 5 MeV and fluences from 1011 to 1016 cm−2. The material did not become amorphous for any of the fluences used. Buried layers with resistivities as high as 108 Ω cm and 106 Ω cm were obtained in GaAs and InP, respectively, after heat treatments. The breakdown voltages corresponding to the highest resistivities are 80 and 35 V, respectively, in GaAs and InP. In GaAs, the Rutherford backscattering analysis on the annealed samples showed an aligned yield close to that of a virgin sample, whereas, the yield in InP is more than that of the as-implanted sample.

Notes: Formation of SiO2 precipitates in Czochralski Si is associated with a volume expansion of more than 100%. The needed extra volume for precipitate growth to occur is primarily supplied by emission of Si self-interstitials (I) into the Si matrix, in balance with a compressive growth residual strain. During cooling after the anneal, an additional compressive cooling strain component also develops because of the different thermal expansion coefficients of SiO2 and Si. For precipitates grown to a sufficiently large size, the growth residual strain and/or the cooling strain can be further relieved by punching interstitial type prismatic dislocation loops into the Si matrix. Otherwise, only I emission can occur. Up to now, there have been no quantitatively determined strain values, which constitute in a given experiment a measure of the I emission efficiency on the one hand, and a basis for determining whether prismatic punching can also occur on the other. In this study, we have calculated the strain values and obtained a quantitative criterion for prismatic punching to occur. In the order of ∼10−3–10−2, the growth residual strain component values indicate that I emission has attained an efficiency of relieving the precipitate growth strain by ∼90%–99%. Available experimental data on the precipitate size dependence of prismatic dislocation loop punching have been satisfactorily fitted using the obtained strain values and the punching criterion, indicating that these calculated values are in acceptable accuracy ranges.

Notes: Bismuth titanate thin films were deposited by the excimer laser ablation technique directly onto bare silicon substrates. Current-voltage studies suggested the formation of a heterojunction at the Bi4Ti3O12/Si interface. The current characteristics at voltages ≥1 V were ascribed to trap-limited space-charge conduction. Hysteretic capacitance-voltage data indicated that charge injection was dominant, masking any polarization mode necessary for nonvolatile memory operation. Ferroelectric hysteretic studies, however, indicated that the films were ferroelectric with Pr(approximately-equal-to)10 μC/cm2 and Ec(approximately-equal-to)150 kV/cm.

Notes: Heteroepitaxial Y1Ba2Cu3O7−x films were grown on MgO(100) substrates by using the reactive evaporation method. The in situ streak reflection high energy electron diffraction pattern of films gradually faded with an increasing thickness when the hydrogen partial pressure during growth was higher than 1×10−6 Torr. Transmission electron microscope observation revealed that the film showed heteroepitaxial growth in the early stage of coevaporation. However, as with the increase of film thickness, the quality of the film became worse and finally reached polycrystalline. These results are attributed to the accumulation of hydrogen induced defects during heteroepitaxial growth of Y1Ba2Cu3O7−x films.

Notes: The influence of heat treatment on conventional magnetic properties and Barkhausen noise were studied in ferromagnetic amorphous Fe40Ni38Mo4B18 during a half magnetization cycle for both amorphous and crystallized states. The coercivity, Hc, was low in the amorphous specimens and maximum as the crystallization starts, and decreased with further annealing. The differential permeability near to the Hc, μc, was high in the amorphous specimens and minimum at the onset of crystallization, then increased with annealing. The single peak behavior of the Barkhausen noise (BN) envelope was a common feature in both amorphous and crystallized specimens, and the variation of peak height with heat treatment was similar to that of μc. The magnetic anisotropy energy constant, Ku, was lowest when the crystallite started to nucleate in the amorphous specimen, and increased as the crystallite size grows. The difference between ribbon and transverse directions of BN characteristics such as BN energy, peak height of BN envelope, Hp, and corresponding field strength to Hp, respectively, was the lowest at the onset of crystallization, and increased with Ku. A model was proposed to predict the linear relationship between the magnetic anisotropy energy constant detected by torque measurement and the one evaluated by BN measurement.

Notes: The results of paramagnetic defect and positive fixed oxide charge creation in x-irradiated buried SiO2 formed by ion implantation and annealing are presented. Charged oxygen-vacancy centers are argued to be the source of spin active defects but not the primary source of fixed oxide charge. Hydrogenation or fluorination of the oxide enhances the radiation sensitivity for creation of spin active defects but not of trapped positive charge. Annealing of the spin active defects may proceed by a mechanism similar to that involved in charged defect annealing or by the trapping of thermally detrapped, diffusing electrons. Annealing does not involve the trapping of diffusing molecular species as is usual in bulk oxides.

Notes: The influence of near-contact electrical field nonuniformities on the injection of minority carriers into semi-insulating GaAs has been investigated experimentally. Variable length p+-v-n+ (200–10 μm) diodes are used to show that near-contact regions do affect the electrical measurements in high-resistivity trap-dominated semiconductors even though linear I-V characteristics are observed. As the diode length is decreased from 200 to 10 μm, the current density through the device, at a constant average field, decreases by a factor of 10. The experimental results are compared, with good agreement, to a previously proposed theoretical model in which semi-insulating GaAs is treated as a trap-dominated relaxation semiconductor.

Notes: Using a multiparticle Monte Carlo method, a theoretical analysis of the spectral density of velocity fluctuations in semiconductors has been performed, under both stationary and transient conditions (when the electric field applied to a semiconductor changes). In the case of the transient analysis a general method has been developed and applied to N-type GaAs and InP. The results obtained are interpreted in terms of the microscopic processes occurring during the transient. Significant differences between these materials have been observed. The following results were found: (i) The main source of noise is the presence of carriers in the Γ valley subject to the action of high fields, and the velocity-disorienting effect of the intervalley mechanisms; (ii) the maximum in the spectral density is essentially due to the presence of intervalley mechanisms; (iii) the dominant frequencies in the spectral density are strongly affected by the duration of the free flights; (iv) for long times, the transient spectral density converges on the steady-state one of the final field.

Notes: Schottky barrier height enhancement on n-InGaAs is studied on structures with thin surface layers of different compositions. Counter-doped p+-InGaAs layers, as well as layers of n- and p-InP, n-GaAs, and n-InGaP of different thicknesses and dopant densities, respectively, were used to enhance the barrier. Titanium was used as a barrier metal to prepare Schottky diodes of different areas and the barrier height is analyzed by current-voltage measurements. It is observed that the barrier height enhancement by p+-InGaAs layers increases with the layer thickness and dopant density, respectively, and effective barrier heights up to 0.63–0.68 eV, i.e., higher values than previously reported, have been measured. The barrier height enhancement by counter-doped p+-InGaAs layers on n-InGaAs can be described by the two-carrier model. Schottky diodes with extremely low reverse current densities have been prepared, JR(1 V) =4.5×10−6 A/cm2. It is shown that lattice-matched InP surface layers can be used as an alternative to enhance the barrier height on n-InGaAs. The barrier height increases with the layer thickness up to φB=0.53–0.55 eV, i.e., up to values previously reported as barrier heights on thick n-InP. Additional barrier enhancement can be achieved by counter doping of the InP surface layer and barrier heights of 0.66 eV have been obtained by p-InP surface layers on n-InGaAs. On structures with barrier-enhanced n-GaAs layers, a remarkable decrease of the reverse current density is observed if the layer thickness is reduced to the critical layer thickness, but the barrier height is very low due to the small n-GaAs thickness. For structures with slightly lattice-mismatched n-InGaP layers (xGaP=0.11) measured barrier heights are similar to those for n-InP enhancement layers of the same thicknesses.

Notes: We have performed photoreflectance measurements of the exciton transitions associated with the first (n=1) subbands in GaAs(dG A(ring))/InxAl1−xAs(dI A(ring)) strained-layer superlattices with (dG, dI, x)=(100,100,0), (100,100,0.1), (100,100,0.2), and (40,40,0.2) grown on (001) GaAs substrates. The photoreflectance-intensity profiles of the n=1 excitons, which reflect the oscillator strength, clearly demonstrate that the order of the n=1 ||J=3/2, mJ=±3/2≥ (heavy-hole) subband and the n=1 ||3/2,±1/2≥ (light-hole) subband is changed by the In concentration (lattice-mismatch strain) and the layer thickness (quantum-size effect). We have analyzed the experimental results of the hole-subband order as a function of the In concentration and the layer thickness on the basis of an effective-mass approximation taking into account strain effects.

Notes: The stability of the Bi2−xPbxSr2Ca2Cu3O10 (Pb-2223) phase contained in silver-sheathed oxide-powder-in-tube specimens has been investigated by x-ray diffraction, transmission electron microscopy, and energy dispersive x-ray analysis. Silver tubes loaded with Pb-2223 precursor powders were processed into tapes using established metallurgical techniques. The tapes were heat-treated in a specially designed equilibration apparatus at selected temperatures (800–845 °C) for a range of times (10–5500 min) and quenched in liquid gallium held at ∼40 °C. The results showed that the Pb-2223 phase is stable in a limited temperature interval between 810 and 830 °C in 7.5% oxygen. At 800 °C, this phase decomposes to Bi2Sr2CaCu2O8 (2212), Ca2PbO4, and CuO; while at temperatures ≥840 °C it partially melts with precipitation of Bi2Sr2CuO6 (2201) and Ca2CuO3. The effects of the silver cladding on the Pb-2223 phase stability and microstructure are also discussed.

Notes: Demagnetizing curves after field cooling were measured in the melt processed YBa2Cu3Ox superconductors using a vibrating sample magnetometer and a SQUID magnetometer. The curves showed that the samples worked as strong bulk permanent magnets. This is attributed to the strong flux pinning of the bulk superconductors and is well explained in terms of the Bean–London model. The magnetic properties of the permanent magnet can be continuously changed using the field-cooling technique. The magnet properties of the melt-powder–melt-growth processed YBa2Cu3Ox with 1.5 mm thickness obtained at 5 K after field cooling in 10 kOe are as follows: the residual flux densities Br are 9.6 kG (H(parallel) c) and 8.5 kG (H⊥ c); the coercive forces bHc are 33 kOe (H(parallel) c) and 17 kOe (H⊥ c); and the maximum energy product, (BH)max, are 110 MGOe (H(parallel) c) and 49 MGOe (H⊥ c).

Notes: Electrical properties of gold-containing plasma-polymerized thin films have been studied in the dielectric regime (isolated conducting clusters dispersed in a polymeric matrix). dc measurements over a wide temperature range provide evidence for a transport process involving the matrix itself as opposed to tunneling directly across the insulating barrier between metallic clusters. The films display space-charge-limited conduction which is due to the existence of trap states in the polymeric phase. An exponential distribution of traps with a peak value of the order of 1017 cm−3 eV−1 has been deduced from the voltage-current data. The ac behavior is dominated by conduction losses at low frequency with a dissipation peak due to interfacial polarization between metal and matrix in the kHz range. Another relaxation is found for gold-rich films. The full interpretation requires more details on the polymeric phase which composition and electrical properties change gradually with an increasing gold concentration.

Notes: The values of nonlinear constants of piezoelectric crystals are frequently calculated from overdetermined linear systems using the least-squares fit. In such systems, all nonlinear constants, whose values are known from earlier publications, are replaced with these values. The results of the least-squares fit, executed after the substitutions, are interpreted only mechanically, with no regard for the substitutions made. It is shown that this procedure frequently leads to a loss of valuable information and/or to unnoticed distortions and misinterpretations of the estimates of the nonlinear constants and of their standard errors. The reasons for these undesired phenomena and how to avoid them are explained. The calculation process should start with the least-squares fit with the substitutions to be made later, when their effect cannot be misunderstood. In a number of cases, however, the substitutions serve no useful purpose and should be avoided. A practical illustration is made for the nonlinear constants of quartz determined by the resonator method. The ideas and results of this paper apply generally to the determination of any material constants of crystals where the least-squares method is used.

Notes: High-fluence KrF-laser-irradiated silicon wafers exhibit a wealth of surface structures, which are different depending on the medium present above the surface during treatment with ultraviolet light. A special surface feature, appearing after treatment under water and in air, is analyzed with a Raman microprobe spectrometer. It is found to consist of leaflike formed silicon structures located above but still attached to the treated surface. During Raman scattering measurements the temperature of these morphological peculiarity rises considerably even at low cw probe laser powers because of geometrically limited thermal conductivity. It is shown that the temperatures, which were derived from the Stokes–to–anti-Stokes scattering intensity ratio are more reliable than those extracted from line shifts since the stress developing inside the Raman laser-heated leaflike microstructures is not known. A Raman microprobe scan of a damage spot, generated by UV treatment under water, reveals in the most cases an enhanced Raman intensity (factor of 10) in comparison with untreated areas. This enhancement is even larger at surface spots of smaller (≤1 μm) sizes and is attributed to the field enhancement of the incident and scattered radiation field caused by geometrical resonances.

Notes: Room temperature measurements are reported of the mid-infrared local vibrational mode (LVM) absorption caused by CAs acceptors in GaAs. A 30:1 range of carbon content was found among a group of 33 samples of melt-grown semi-insulating GaAs. The LVM absorption band area is compared with the dip ΔT this band makes in a transmittance spectral trace. The ratio (ΔT/t) of transmittance dip to sample thickness shows a useful correlation with the integrated LVM band area.

Notes: Diamond crystallites have been nucleated and grown by hot filament chemical vapor deposition at 600 °C on the untreated fracture surface of a cleaved Si(111) sample. The flat surface of the cleaved crystal was inactive towards diamond nucleation while, on the terraced surface formed by the propagating crack, a high density of nuclei was found. The crystallites were nucleated in correspondence of edges between (111) terrace planes and step planes. The occurrence of edges, as determined by scanning electron microscopy (SEM) observation, is a necessary but not sufficient condition for the nucleation and this fact suggests that particular atomic arrangements are required for the diamond nucleus formation.

Notes: Electroluminescence (EL) efficiency increase due to recombination-enhanced annealing of recombination centers has been observed. The EL of the band-to-band emission from GaAs tunnel diodes has been investigated. The EL efficiency is found to increase as the diodes are degraded due to the recombination-enhanced process. The EL efficiency increase is attributed to the recombination-enhanced annealing of the recombination centers.

Notes: ac breakdown properties of alumina, aluminum nitride, and glass ceramic (SiO2-Al2O3-MgO-TiO2) materials have been investigated. The breakdown strength of glass ceramics (∼65 kV/mm) was found to be higher than that of alumina (∼31 kV/mm) and aluminum nitride (∼16 kV/mm). The specimens were characterized for their density, porosity, and microstructural parameters, such as grain size and distribution, as they are believed to influence the breakdown properties significantly. In all the cases, the prebreakdown and breakdown conduction were found to be associated with visible-light emission at the electrode-specimen interface. The samples that underwent breakdown were found to exhibit irregular puncture channel, terminated with craters at both ends. Several types of crystallization structures were observed at the rim of the craters and in some cases at the edges of the breakdown channels. It is suggested that the breakdown process of the ceramic materials is a combination of electronic, electromechanical, and thermal processes.

Notes: Photothermal laser ablation is studied using poly-(methyl methacrylate) films doped with a dye, IR-165, which functions as a molecular heater and thermometer. Direct optical measurements of temperature are performed on samples heated by 100 ns near-IR pulses at 1.064 μm, at rates dT/dt≈5×109 deg/s. Below ablation threshold, the heat capacity measured by optical calorimetry is precisely the value obtained by conventional calorimetry. At ablation threshold, the peak surface temperature is Tabl=600 °C. The increase in heat capacity observed above threshold, together with the results of a conventional thermal analysis, is used to determine the weight fraction of material decomposed at ablation time χth=0.02. With increasing pulse energy, the fraction decomposed increases and a more forceful ablation is observed, but the surface temperature does not continue to increase past Tlim=715 °C, which is determined to be the limiting temperature for thermal decomposition.

Notes: Plasma-polymerized fluorocarbon (PPFC) films were analyzed by thermogravimetric analysis (TGA), direct pyrolysis/mass spectrometry, and laser-ablation/electron-impact mass spectrometry. Fourier transform mass spectrometry was used to detect products. The films were made by plasma polymerizing tetrafluoroethylene in an argon plasma. Two types of films were studied: with and without fine gold particles incorporated in the PPFC films. TGA showed that gold-containing films decompose more rapidly and at lower temperature with increasing gold content. Pyrolysis products were determined as a function of temperature. The predominant positive product ions, using 20 eV electron-impact ionization, were C2F4+, CF3+ and a distribution of higher-mass unsaturated fluorocarbon species, CnFm+, up to at least n=14 and m≥n+1. The predominant negative ions, formed by electron attachment, were also unsaturated fluorocarbon ions which extended up to 1145 u in mass. These species are different from those observed from polytetrafluoroethylene pyrolysis. No dependence of the positive-ion mass spectral distributions of pyrolysis products on gold content of the films was observed. However, the negative-ion pyrograms had a dramatically different temperature dependence for films with and without gold. The results imply that the gold-containing films have regions of lower molecular weight, and a possible formation mechanism is suggested. Laser ablation formed low-mass neutral fluorocarbon species. Ablation of the gold-containing film under direct laser ionization conditions (higher laser power), produced a spectrum resembling the pyrolysis spectrum, suggesting that the gold particles affect the mechanism of the ablation.

Notes: The Ti–Si system is chosen as a model system to study the mechanical alloying of elemental powder blends in which one of the components is brittle. In addition, the stability of intermetallic compounds upon milling is investigated, in order to interpret the metastable phase formation during the milling process. Detailed x-ray investigations show that during mechanical alloying, a solution of Si in Ti preceeds the formation of amorphous or intermetallic phases. The selection of phases formed, as well as the final state, are influenced by the overall composition of the powder blends, and by the milling procedure applied. However, predictions can be made by considering the free energy curves of the stable and metastable phases in the Ti–Si system, which have been calculated by the CALPHAD method. Milling of the intermetallic compounds generally leads to the same final state as the mechanical alloying of elemental powder blends with corresponding compositions. These results demonstrate that the intermetallic compounds are energetically destabilized upon milling, allowing for phase transformations into metastable phases, such as amorphous or metastable crystalline alloys. The energetic destabilization of intermetallic compounds mainly originates from the chemical disorder caused by the milling process.

Notes: The lifetime of a surface plasma wave (SPW) on a semi-infinite air/metal interface is calculated. The calculation is based on the assumption that SPW has a real frequency and a complex propagation constant. Thus the lifetime of the SPW can be calculated in a wide spectrum range from 0.1 to 10 μm. Two different metals (aluminum and silver) have been used separately as the media for generating the SPW. Within the spectrum of visible light, the lifetime of the SPW on the air/metal interface is in subpicoseconds.

Notes: We have studied the magnetoresistance (MR) properties of sputtered Si/Re 50A(ring)[Co 20 A(ring)/Re x A(ring)]40 superlattices with xRe=3–32 A(ring). Transverse and longitudinal MR exhibit oscillation of the interlayer coupling as a function of Re layer thickness with a period of (approximately-equal-to)12 A(ring) and a maximum transverse MR of (approximately-equal-to)1.2% for xRe=5 A(ring) at 4.2 K. The saturation magnetic field Bs and MR magnitude ΔR/R for samples with xRe=5 and 7 A(ring) are nearly constant up to 40 K, above which they decrease approximately linearly with increasing temperature. In addition, the perpendicular MR in magnetic fields up to 1 T changes sign from negative for samples with xRe≤7 A(ring) (strong antiferromagnetic coupling) to positive for those with xRe≥10 A(ring).

Notes: Electrical resistance of Al-1 wt % Si thin-film conductors has been measured as a function of time t, temperature, and current polarity in order to investigate both generation and recovery of (microstructural) damage caused by electromigration. The fractional change of electrical resistance ΔR/R is characterized by three distinct stages: (i) undetectable ΔR/R during an incubation period τ; (ii) linear increase of ΔR/R with t−τ; and (iii) abrupt decrease of ΔR/R when polarity is reversed, followed by gradual resumption of the previous linear increase. Examination of the conductor surface during these three stages by scanning electron microscopy reveals: (i) undetectable microstructural damage; (ii) generation of (first) holes and (then) hillocks; and (iii) recovery followed by further generation of microstructural damage. Results are interpreted by (i) generation of stress σ in grain boundaries; (ii) formation of holes when σ exceeds a critical tensile stress σ+c and hillocks when σ exceeds a critical compressive stress σ−c (||σ+c|| 〈 ||σ−c||), and (iii) interchange of tensile and compressive stress by polarity reversal. The last stage, in fact, represents superposition of a continuation of the linear increase (degradation) of ΔR/R due to the applied current and an exponential decrease (healing) of ΔR/R, characterized by τ, due to stress relaxation. In general, damage and subsequent healing by electromigration involve a delicate balance between applied current, time, and spatial distribution of (elastic) tensile and compressive stress, (anelastic) formation of holes, and (plastic) formation of hillocks, as dictated by the concomitant microstructure.

Notes: The results of a scanning tunneling microscope (STM) study of the morphology of Au films thermally evaporated onto heated glass substrates are presented. Au films of thickness 20–80 nm were evaporated onto Corning glass cover slips at temperatures of 20–465 °C. Before the evaporation, the glass substrates were prebaked at 300–400 °C for 12 h to remove surface contamination. Grain size and surface roughness of the films have been measured, and the best results were obtained with Au films 80 nm thick evaporated onto 300 °C substrates. These films have 250-nm-diam grains with large, atomically flat tops exhibiting step-free terraces as large as 200×200 nm2. These results are comparable to the best reported results for Au grown on heated mica and show that epitaxial growth is not required for the production of films with large atomically flat regions. This study of Au on glass was motivated by failed attempts to produce flat films of gold on mica using recipes in the literature. Possible reasons for the poor results with mica are discussed.

Notes: Low-frequency (10–60 Hz) electrical noise has been measured for the first time in carbon fibers while they were subjected to tensile stress. The measurements were performed at room temperature. In all the fibers, a minimum in the noise was observed with applied stress. Low-modulus ex-PAN (polyacrylonitrile) fibers showed three times the increase in noise at the highest stress compared to the higher modulus fiber. Measurements were also made on vapor grown carbon fibers (VGCFs) both as-grown and annealed. The highly graphitized VGCF showed the smallest increase in the noise with stress among the fibers studied. Our noise results can be related to the microstructure of the fibers. The fiber microstructure becomes better aligned with applied stress. Therefore, the noise would be expected to decrease. Our results, however, indicate that stress dependent noise in these fibers is caused by a combination of parameters such as fibril alignment and electronic characteristics.

Notes: Raman scattering measurements on cubic boron nitride up to 1600 K are reported. The results indicate that at 1600 K the interatomic force constant is only ∼10% weaker than at room temperature and that cubic BN is therefore a good candidate for applications where high-temperature mechanical strength is required. However, contrary to the behavior found for diamond, where the frequency changes with temperature were very nonlinear, it is found that in boron nitride both the TO and LO shift linearly with temperature at a rate of −0.036 cm−1/K.

Notes: The analysis of transient photoluminescence measurements and extraction of carrier recombination lifetimes in GaAs and AlGaAs double heterostructures is discussed. In contrast to recently reported claims, it is demonstrated that even in regions where the measured decay curves show single exponential behavior, the slopes do not, in general, correspond to any single physical carrier lifetime such as the minority-carrier lifetime. A series of measurements over a range of incident optical intensities is required to extract such lifetimes.

Notes: Thermomagnetic measurements were used to determine the oxidation kinetics of bulk and powdered Nd-Fe-B magnets under a pressure of 1 Pa in the 360–600 °C temperature range. Those conditions are roughly similar to those undergone by sintered or rapidly quenched magnets during heat treatment. The first step of the oxidation kinetics obeys a parabolic law for bulk magnets and a cubic one for powders, with activation energies in the range 65–78 and 35–50 kJ/mol, respectively. Some anomalies in these patterns related to time or temperature are discussed.

Notes: The selectivity of copper deposition from copper (II) bis-hexafluoroacetylacetonate on SiO2 patterned with a platinum seeding layer is studied as a function of the reagent gas mixture. On platinum, the copper film growth rate increases with the amount of water vapor in the gas flow, and is independent of the chemical nature of the carrier gas used (H2 or He). The selectivity of the copper deposition is significantly improved when using He rather than H2 as carrier gas, especially at high water vapor concentrations where rapid film growth can be obtained.

Notes: The effects of high temperature rapid thermal annealing processes on carrier concentration and mobility of bulk AlInAs and AlInAs/GaInAs high electron mobility transistor structures with planar Si doping are studied. At annealing temperatures of 700 °C and 800 °C, slight reduction in mobilities and carrier concentration are observed in samples annealed with a Si3N4 cap or GaAs pieces in close proximity. The reduction in mobility is thought to be due to enhanced diffusion of the donor Si atoms towards the two-dimensional electron gas channel. Preferential vacancy enhanced diffusion of Si atoms towards the surface is projected to be responsible for the loss in carrier concentration. At these annealing temperatures, the reduction in mobility in the samples annealed with SiO2 capping is more pronounced, and is as high as 80% at the measurement temperature of 15 K. This behavior is attributed to the outdiffusion of Ga and In atoms into the oxide thereby creating vacancies and resulting in interface mixing. Reduction in mobility and carrier concentration are much more substantial in the 900 °C anneals done with Si3N4 cap and GaAs pieces in close proximity. This indicates the destruction of the heterostructure integrity of the AlInAs/GaInAs interface. For the particular anneal with a SiO2 cap at this temperature, the carrier concentration increases above its reference value due to effective doping of the ternary material by the back-diffusing Si atoms from the SiO2 cap.

Notes: The photoluminescence (PL) of porous Si that has been hydrogenated in boiling water has been investigated. The PL intensity is observed to increase with a concurrent shift of the spectral peak to shorter wavelengths. These effects can be explained in terms of size effects in the microstructures of porous Si. The spectral changes after annealing and after rehydrogenation are similar to the behavior of a-Si:H. We conclude that hydrogen plays an important role in the luminescence of porous Si.

Notes: The possibility of replacing Pt in the Ti/Pt/Au base and traditionally used metallurgical structure by Ni, while bonding InP laser chip to a submount with AuSn (80% Au) solder, has been investigated. Various Ni-based metal alloys have been prepared by evaporation. Reflow experiments were conducted in a chamber under forming gas-controlled ambient. The Ti/Ni/AuSn system provided much longer surface local freezing duration compared to the Ti/Pt/AuSn system. Scanning electron microscopy analysis revealed a smoother surface morphology for the Ti/Ni/AuSn system after the metal refroze. Auger electron spectroscopy depth profiles indicated the formation of a Ni-Sn-Au interacted layer. The interaction took place in two steps: the first stage was the dissolution of Ni into the Au-Sn liquid followed by precipitation of a Ni-Sn-Au intermetallic compound; the second stage was a solid-state interdiffusion of Sn, Au, and Ni which occured in the interacted layer and in the original Ni layer. The latter step was a diffusion-controlled process, resulting in a very slow growth rate. Both Au and Sn reacted to form Ni alloy layers of almost equal thickness, regardless of the reaction duration (up to about 5 min). This intensive reaction, however, did not lead to full consumption of the Ti interfacial layer, which provided an excellent adhesion layer between the submount and the metallurgical structure.

Notes: Epitaxial regrowth and dopant activation of Sb-implanted Si and Si0.93Ge0.07 were investigated with ion channeling techniques and transmission electron microscopy. The presence of Sb greatly enhanced the regrowth rate of both Si and Si0.93Ge0.07. The initial crystallinity of Si0.93Ge0.07 was fully recovered after furnace annealing at 500–550 °C. Approximately 95% of Sb atoms were found on substitutional sites and most of them were electrically active.

Notes: The phenomenon of simultaneous compositional disordering and the formation of electrical resistive layers induced by oxygen implantation in InGaAs/InP superlattices has been investigated. The disordering characteristics have been studied as a function of implantation temperature and ion dose. It was found that implantation at elevated temperatures (referred to as the IM or ion mixing process) usually leads to much more efficient disordering compared to implantation at room temperature followed by annealing at the same elevated temperature (referred to as the implantation plus annealing process). Of particular interest is the observation that ion mixing at 550 °C with 1×1013 O+/cm2 leads to significantly more disordering than implantation with the same dose at room temperature followed by annealing at 550 °C for the same period of ion mixing time. In addition, the electrical resistance of the ion-mixed layer at 550 °C increases 2600 times for the p-type InGaAs/InP superlattice structure, whereas the sample implanted at room temperature and annealed at 550 °C showed only a 20 times increase in electrical resistance. These results indicate a distinct advantage for the IM process in achieving simultaneous compositional disordering and electrical isolation for optoelectronic applications.

Notes: We investigate the behavior of paramagnetic AsGa+-related defects under 1.2 eV illumination at 4.2 K in plastically deformed bulk semi-insulating GaAs. We find them identical to the grown-in antisites with a similar photoquenchability and analogous parameters. Further, there is effective generation of additional AsGa centers after deformation. The photobehavior can be interpreted by a charge transfer model, where the metastability involved with the AsGa+ signal decay is not a defect inherent feature.

Notes: Phase formation kinetics in Nb/Al multilayered thin films having overall compositions of 25, 33, 50, and 75 at. % Al have been investigated using scanning calorimetry, x-ray diffraction, and cross-sectional transmission electron microscopy. The first phase to form upon annealing the Nb/Al layered structure of all samples is the NbAl3 intermetallic. Calorimetry clearly identifies the NbAl3 formation to be a two-stage process. The first stage is the formation of a planar layer by nucleation and growth to coalescence while the second stage is the thickening of the planar layer. The large amount of heat released (and hence large volume fraction of NbAl3 formed) during the first reaction stage is consistent with heterogeneous nucleation at well-isolated sites in the Nb/Al interface. This is surprising in light of the large thermodynamic driving force expected for nucleation and suggests that the local nonequilibrium nature of the Nb/Al interface greatly reduces the driving force for nucleation. The next phase observed in samples of 25 and 33 at. % Al is the A15 superconducting phase, Nb3Al. The Nb3Al growth completes a first reaction stage similar to the NbAl3, but the subsequent thickening reaction stage is not observed without simultaneous Nb2Al growth. The high interface velocities derived from the calorimetry for formation of both NbAl3 and the A15 Nb3Al indicate that atomic transport must be by grain boundary diffusion.

Notes: In this study, we report a high-performance ultrathin oxide (≈80 A(ring)) prepared by a low-temperature wafer loading and N2 preannealing before oxidation. This recipe can reduce native oxide thickness and thermal stress compared to the conventional oxidation recipe. The high-resolution transmission electron microscopy reveals that the SiO2/Si interface is atomically flat, and a thin crystalline-like oxide layer about 7 A(ring) exists at the interface. Oxides prepared by the proposed recipe show a very high dielectric breakdown field (≥16 MV/cm) and a very low interface state density (Nit ≈ 3 × 109 eV−1 cm−2 at midgap). The effective barrier height at cathode derived from the slopes of log(Jg/E2ox) vs 1/Eox and tbd vs 1/Eox plots is about 3.9 eV, instead of 3.2 eV for the control sample. It also shows a better immunity to the charge trapping and interface state generation under high-field stressing, and superior time-dependent dielectric breakdown characteristics.

Notes: The dependence of the excitonic lifetime on the well width in InGaAs/InP (unstrained), InGaAs/GaAs, and InGaAsP/InP strained quantum well heterostructures is investigated by time-resolved photoluminescence spectroscopy. We show that the main contribution to the lifetime variation comes from the change of the exciton binding energy. In the strained InGaAs/GaAs wells, the widest wells have a shorter lifetime than expected. We attribute this to the onset of plastic relaxation.

Notes: Magnetic levitation and stiffness have been measured in several systems composed of a permanent magnet elastically suspended above a stationary melt-textured sample of Y-Ba-Cu-O. The levitation force and vertical stiffness have been calculated on the basis of magnetization measurements of the same system, and the calculated results showed excellent agreement with the experimental measurements. Based on the force and magnetization measurements, it is predicted that the same Y-Ba-Cu-O material configured in a geometry suitable for magnetic bearings could produce a levitation pressure of 100–400 kPa at 20 K.

Notes: For many excimer laser applications for example, in medicine and semiconductor processing, the long-term stability of the output pulse energy and the laser-beam profile are very important. The spatial intensity distribution of an excimer laser with a Fabry–Perot cavity depends on many discharge conditions (e.g., the loading voltage, the total gas pressure, the concentrations of the gas components, the repetition rate, the electrode profiles, and the spatial distribution of the preionization intensity). A one-dimensional model of a discharge pumped excimer laser is described. It can be shown that the laser output power of a KrF laser depends on the spatial variations of the discharge parameters to a higher degree than does a XeCl laser. The double structure of the transverse intensity distribution for long-term KrF laser operation can be explained as a consequence of the depletion of F2 concentration within the gas mixture after many discharges.

Notes: Comparisons between experimentally measured time-dependent electron energy distribution functions and optical emission intensities are reported for low-frequency (100 and 400 kHz) radio-frequency driven discharges in argon. The electron energy distribution functions were measured with a time-resolved Langmuir probe system. Time-resolved optical emissions of argon resonance lines at 687.1 and 750.4 nm were determined by photon-counting methods. Known ground-state and metastable-state excitation cross sections were used along with the measured electron energy distribution functions to calculate the time dependence of the optical emission intensity. It was found that a calculation using only the ground-state cross sections gave the best agreement with the time dependence of the measured optical emission. Time-dependent electron density, electron temperature, and plasma potential measurements are also reported.

Notes: Heteroepitaxy immediate to thermodynamical equilibrium is investigated for liquid phase epitaxy of Ge0.85Si0.15 on Si(111). Below and beyond the critical thickness for strain relaxation, the GeSi layer grows in a two-dimensional mode. The critical thickness agrees well with mechanical equilibrium considerations. Strain relaxation is found to take place by the confined formation of misfit dislocations directly at the Ge0.85Si0.15/Si interface. We do not observe half-loop formation at the layer surface. The misfit dislocations (b=1/6〈112〉) form regular networks, which are buried under a structurally perfect relaxed GeSi layer virtually free of threading dislocations.

Notes: It is demonstrated that a closed delayed equation for the phase of the electric field can be used to describe accurately the transient switch on of a laser with external feedback when it can be described by field evolution equations. In contrast, it is shown that several more severe approximations to the dynamics, including an adiabatically evolving potential for one-dimensional relaxation dynamics (which had previously been used to accurately predict the laser linewidths for the steady-state solutions) fail to reliably describe the final state selection during transient switch on of the laser, the delay-induced oscillatory approach to the chosen state, and features of the amplitude and frequency spectra near multiples of the external cavity mode spacing which may be important for the stability and switching dynamics of the steady states that correspond to excitation of different external cavity resonances.

Notes: Recent theoretical work on atomic level sliding friction is summarized. Some previous analytic results are verified by numerical simulations, and finite-size scaling arguments based on several time scales appropriate to finite crystals undergoing shear motion are used to interpret the results of the simulations and to give insight into the methods by which energy is dissipated in such a shearing process. One conclusion is that the existence of a lifetime for the lattice vibrations plays an important role in determining the velocity dependence of the force of friction, for any finite-size crystal. Finally the force of friction found in recent microbalance experiments between a solid rare-gas-element film on a metallic substrate is calculated by the present methods and compared to experiment. The main conclusion is that the observed friction is probably due to atomic level defects such as substitutional impurities. Larger-scale defects on the surface contribute a much smaller value.

Notes: The reaction kinetics of tungsten deposition from a SiH4/WF6 gas mixture was investigated for SiH4/WF6 feed-gas flow ratios lower than unity. By performing selective depositions, care was taken to ensure that the kinetic data was obtained in the surface-kinetics controlled regime. It was found that the reaction kinetics of the deposition process depends on the SiH4/WF6 ratio. For low SiH4/WF6 ratios (smaller than ≈0.7) the growth rate is slightly negative in [WF6] and almost quadratic in [SiH4]. The growth rate is almost independent of temperature between 200 and 350 °C. On the basis of our measurements it is argued that in this regime the surface is covered mainly with W, WFx, and F, and that the rate-limiting step is the dissociative adsorption of SiH4 on this surface. For higher SiH4/WF6 ratios (0.7≤SiH4/WF6≤1) the growth rate increases sharply as the SiH4/WF6 ratio increases. In this regime the growth rate cannot be expressed in the simple form: rate=A⋅[WF6]α⋅[SiH4]β since the growth rate is more sensitive to the SiH4/WF6 partial pressure ratio than to the absolute values of the SiH4 and WF6 partial pressure. For SiH4/WF6 ratios as high as 1, the growth rate is limited only by the flux of the species hitting the surface.

Notes: The Maxwell displacement current generated from single fatty acid monolayers containing azobenzene by photoisomerization was investigated in the context of the phase diagram of monolayers on a water surface. Transient displacement-current pulses were found to be produced when successive isomerizations were induced in monolayers by irradiation with ultraviolet and visible light. In contrast, the displacement current was never generated in the so-called solid phase by the application of surface pressure or photoirradiation, probably because of the lack of free volume for the isomerizing chromophore in single monolayers and as well as the aggregation of molecules in the monolayer. The displacement current generated during monolayer compression was also investigated in connection with phase transitions.

Notes: We report the optical study of the interband transitions in InAsxP1−x/InP strained-layer multiple quantum wells grown by gas-source molecular beam epitaxy. Low-temperature photoluminescence, photoluminescence excitation, and room temperature photomodulated transmission measurements were performed to investigate optical interband transitions. In addition to transitions associated with the heavy-hole and the light-hole bands, a transition involved with the spin-orbit split-off band was observed. We also observed spectral linewidth broadening due to compositional inhomogeneity and layer-thickness fluctuations from the sample using short-period superlattices as the well materials. Calculations based on the envelope-function approximation and phenomenological deformation potential theory, including both band nonparabolicity and strain-induced valence-band mixing, were compared with experimental data to identify the optical transitions between quantized states in the wells. We found good agreement between theory and experiment.

Notes: Two Curie temperature Tc peaks in differential scanning calorimetry (DSC) are observed in a well-crystallized, nonpoled film of a ferroelectric copolymer of vinylidene fluoride and trifluoroethylene [P(VDF-TrFE)] when it is reannealed at temperatures near Tc. The DSC peak at higher Tc shifts to higher temperature with increasing reannealing temperature, the phenomenon being very similar to that observed in thermally depolarized P(VDF-TrFE) film. This result can be explained by the formation of large 60° or 180° domains during reannealing. The free energy of the simple 180° domain model is used to discuss the temperature dependence of Tc on the domain width and thickness. It is concluded that large domains are generally formed in a P(VDF-TrFE) film by reannealing at temperatures near Tc irrespective of its macroscopic polarization.

Notes: Low-temperature photoluminescence experiments were performed on high-purity InP substrates as a function of heat treatment, flat-profile ion implantation and subsequent annealing procedures. Whereas the well resolved near band edge luminescence proved the superior substrate quality, up to 6 sharp luminescence lines (some of them not reported so far) and one shallow broad-band emission were observed after Si3N4 deposition and various annealing procedures. From the temperature dependence of the luminescence peak intensities the activation energies of the novel lines were determined and discussed in terms of excitons bound to processing-induced complex defects. These assignments are supported by the observation of a strong enhancement of the luminescence after low dose Ar+ ion implantation.

Notes: Room-temperature synchrotron radiation photoemission studies of rare-earth/Hg1−xCdxTe(110) junctions were performed as a function of metal coverage for the rare-earth metals Sm, and Gd. These new results are compared to our previous results for Yb/Hg1−xCdxTe junctions. At low rare-earth metal coverages, we observe in all cases a rare-earth Te reaction which removes Hg and Cd from the interface region, and the three-dimensional island growth of a rare-earth telluride layer 4.0–4.5 monolayer thick. The most abrupt interface region is observed for Gd/Hg1−xCdxTe, in agreement with the more reactive character of this interface as compared to Yb/Hg1−xCdxTe and Sm/Hg1−xCdxTe. For higher rare-earth metal coverages, we observe the formation of a metallic rare-earth rich layer at the surface, which effectively traps Hg atoms diffusing across the interface through the formation of a rare-earth–Hg alloy.

Notes: A rare gas positive column plasma theory that takes into account the effect of various chemical processes on the production, loss, and interconversion of metastable and ion species is developed and compared with experimental results. The experimental measurements have been made in a 4 cm i.d. radio frequency (rf) discharge tube for He, Ne, and Ar. An electrostatic triple probe was used to monitor the electron temperature and plasma density, and a mass spectrometer was used to analyze the ion species. The results show that the molecular ion becomes important when gas pressure is above 1 Torr in the Ar and Ne plasmas. For a He plasma, the molecular ion always plays an important role in the plasma process. Good agreement has been observed between present model and experiments.

Notes: The exact mass, momentum, and energy conservation equations for electron transport in a dc glow are derived from the Boltzmann equation. A Monte Carlo particle simulation is used to explicitly calculate the individual terms of the moment equations, and to gain insight into the behavior of the electron distribution function (EDF) moments such as density and average velocity. Pure forward scattering and isotropic scattering are considered as two limiting scattering mechanisms. When forward scattered, the electron fluid shows the maximum change in properties and in transport mechanisms at the field transition point between the cathode fall (CF) and the negative glow. Isotropic scattering, however, results in property changes a short distance inside the sheath. Diffusion of the low-energy, high-density, bulk plasma electrons into the CF causes dilution of the low-density, high-energy beam from the CF before the beam actually arrives at the low-field region. The applicability of commonly used closure relations which yield a fluid description of the system is evaluated. Use of fluid equations to characterize this system with no a priori knowledge of the EDF is limited by kinetic effects, such as heat flow against the temperature gradient, especially in the forward-scattered case where the EDF is very anisotropic. The description of inelastic rates by Arrhenius kinetics is found to be surprisingly accurate with both scattering mechanisms. However, while temperature is an adequate gauge of the characteristic energy under isotropic scattering, the energy of the bulk electron motion must be included under forward scattering. Also, Arrhenius kinetics sometimes produce a spurious double peak in the inelastic rate profile which is not reproduced by the Monte Carlo simulation. The anisotropy of the EDF under the forward-scatter assumption makes it difficult to justify the use of the mobility and heat conduction closure relations. Under isotropic scattering, however, electron inertia is negligible. In that case, under the discharge conditions used here, the drift-diffusion approximation to the flux is good to within a factor of 2. Classical heat conduction theory overestimates the heat flux by a factor of 4 at the sheath edge.

Notes: Absolute concentrations of ground state atomic hydrogen are determined by two-photon absorption laser-induced fluorescence in a pure hydrogen rf discharge. Axial concentration profiles obtained across the interelectrode space show a relatively constant concentration (of the order of 1014 atoms/cm3) in the region around the center of the discharge until it decreases sharply within 3–4 mm of each metal electrode surface. In addition, the temporal evolution of the atomic concentration is monitored by means of a pulsed discharge from short times to the establishment of a steady-state concentration. We present a model according to which the formation and decay of the distribution of atomic hydrogen is dominated by diffusion from the production regions towards both the center of the discharge and the metallic electrode surfaces where H atoms are absorbed.

Notes: A detailed study of tungsten low-pressure etching in a helicon source reactor was performed. In correlation with surface analyses (transmission electronic microscope, Rutherford backscattering, nuclear reaction analysis), a complete parametric study of the plasma and etching parameters versus the macroscopic parameters [gas pressure, radio frequency (rf) power, substrate bias voltage) has been carried out. Using a model developed by Hoffman and Heinrich for silicon etching (Proceedings of the 9th ISPC, Pugnochiuso, Italy, 1989, p. 1003), and taking into account the experimental results, it has been shown that tungsten etching mechanisms can be expressed by the formation and destruction of a low-density reactive top layer. Sputtering of this layer under argon-ion bombardment has been studied by optical emission spectroscopy. Consequently, the tungsten etch rate can be expressed as an analytical function of the macroscopic parameters such as gas pressure, rf power, and substrate bias voltage.

Notes: For the numerical simulation of electron cyclotron resonance plasma production, a two-dimensional model that describes wave propagation and plasma transport is developed. The modeling code calculates profiles of electromagnetic wave fields, power absorption of electrons, and temporal evolution of plasma densities in a bounded, inhomogeneous, cylindrical system. The calculation of the plasma production in a mirror magnetic field shows that the plasma production property is very sensitive to the antenna location.

Notes: We have systematically investigated the use of high-temperature sputtering in conjunction with excimer laser planarization for planarizing Al alloy surfaces and filling submicrometer, high aspect ratio (∼1–3) contact vias. In general, the process window for the complete contact filling increased significantly with the increasing sputtering temperature, and decreased with shrinking contact/via geometry. Better as-deposited metal step coverage and profile obtained at higher sputtering temperatures were identified as the key parameters contributing to this significant improvement in the process window. The process window achieved by this combined process is significantly wider than that obtained by high-temperature sputtering alone. In addition, the poor surface morphology generally observed for the Al alloy films sputtered at elevated temperatures is significantly improved due to recrystallization during laser reflow.

Notes: The transport-simulation computer program amps has been used to examine the role of contact barrier heights in determining the performance of a-Si:H p-i-n homojunction detector and solar-cell devices. Current-voltage performance, with and without illumination present, is considered. It is determined that for p and n layers that are sufficiently thick and sufficiently doped, the reverse bias currents, both in the dark and under illumination, do not depend or depend only weakly on the barrier heights of the contacts to the doped layers. However, these contact barrier heights can strongly influence both the dark and illuminated currents in forward bias. Thus, it is determined that the quality of the contacts to the doped layers does not affect the performance of a-Si:H p-i-n detector structures, but it can strongly influence the performance of a-Si:H p-i-n solar-cell structures.

Notes: Quantum theory is applied to calculate the free carrier absorption coefficients in n-ZnS. This treatment includes all major scattering mechanisms and screening. We present the results of calculations of 77 K free carrier absorption and derivative logarithmic absorption coefficients, as a function of the free carrier concentration and compensation ratio, at 10 μm. We discuss the unambiguous determination of both free carrier concentration and compensation ratio for unknown samples.

Notes: Generation of a microsecond ribbon electron beam with a strongly elongated cross section 4 cm×140 cm in a magnetically insulated diode at 1 MV voltage and its transport at the distance 2 m in a slit vacuum channel with a guiding magnetic field about 1 T are described. High efficiency of the ribbon beam generation and transport at the total energy about 140 kJ is experimentally proven.

Notes: Spatial distributions of persistent magnetic field trapped by sintered and melt-textured ceramic-type high-temperature superconductor (HTS) samples have been studied. The trapped field can be reproduced by a model of the current consisting of two components: (i) a surface current Js and (ii) a uniform volume current Jv. This Js + Jv model gives a satisfactory account of the spatial distribution of the magnetic field trapped by different types of HTS samples. The magnetic moment can be calculated, based on the Js + Jv model, and the result agrees well with that measured by standard vibrating sample magnetometer (VSM). As a consequence, Jc predicted by VSM methods agrees with Jc predicted from the Js + Jv model. The field mapping method described is also useful to reveal the granular structure of large HTS samples and regions of weak links.

Notes: Light-ion inertial confinement fusion requires beam transport over distances of a few meters for isolation of the diode hardware from the target explosion and for power compression by time-of-flight bunching. This paper evaluates ballistic transport of light-ion beams focused by a solenoidal lens. The ion beam is produced by an annular magnetically insulated diode and is extracted parallel to the axis by appropriate shaping of the anode surface. The beam propagates from the diode to the solenoidal lens in a field-free drift region. The lens alters the ion trajectories such that the beam ballistically focuses onto a target while propagating in a second field-free region between the lens and the target. Ion orbits are studied to determine the transport efficiency ηt (i.e., the fraction of the beam emitted from the diode which hits the target) under various conditions relevant to light-ion inertial confinement fusion. Analytic results are given for a sharp boundary, finite thickness solenoidal lens configuration, and numerical results are presented for a more realistic lens configuration. From the analytic results, it is found that ηt can be in the range of 75%–100% for parameter values that appear to be achievable. Numerical results show that using a more realistic magnetic-field profile for the lens yields similar values of ηt for small radius diodes but significantly reduced values of ηt for large radius diodes. This reduction results from the radial gradient in the focusing field at larger radius.

Notes: Shock- and release-wave measurements are reported for 6061-T6 aluminum [J. R. Asay and L. C. Chhabildas, in Shock Waves and High-Strain-Rate Phenomena in Metals, edited by M. A. Meyers and L. E. Murr (Plenum, New York, 1981), pp. 417–431], oxygen-free-electronic copper, and a Si-bronze alloy. Significant departure from ideal elastic-plastic response is observed in all three materials. Experimentally determined release-wave profiles show evidence for the onset of reverse plastic flow immediately upon release from the shocked state. This phenomenon is analyzed in terms of internal stresses acting on straight dislocation pileups and pinned dislocation loops created by the shock-compression process. Following shock compression and prior to release, the internal stresses are opposed by the applied shear stress; that is, they exactly balance each other and no plastic flow occurs. As the applied stress is reduced in the unloading wave, reverse plastic flow occurs immediately due to internal reverse stresses acting on these pileups and pinned loops. This effect reduces the longitudinal modulus, and hence, the release-wave speed in what we normally think of as the "elastic-wave'' regime. Interpretations of quasielastic release-wave data and calculations are expressed in terms of micromechanical concepts.

Notes: The effective diffusivity of substitutional impurities in silicon is reconsidered by involving the point defects. Four basic reactions, including the interstitials and vacancies contributions, are necessary to derive a general formulation of the effective diffusion coefficient. The model has been fitted to data on arsenic diffusion in silicon under nitridation conditions, with a very good agreement. From these experimental results, it has been deduced that arsenic diffuses mostly by a vacancy mechanism with a partial interstitial contribution, which cannot be neglected. It has been also observed that, under these experimental conditions, the point-defects concentrations are a complex function of the observed phosphorus and antimony diffusivities, in contrast with previous approaches. The general behavior of the dopants diffusivities as a function of point-defects concentrations is also presented. This allows the importance of each reaction involved in this analysis to be shown clearly. Finally, the recombination of point defects mediated by dopants pairs has been emphasized over a wide range of conditions for point-defects concentrations.

Notes: The growth kinetics and inhibition of growth of chemical vapor deposited thin W films on Si(100) from WF6 was studied with in situ growth stress and reflectivity measurements and ex situ weight gain measurements. A systematic series of experiments at varying WF6 flow, total pressure, and temperature show that the thickening kinetics and inhibition of the growth are controlled by two processes: WF6 diffusion through the gas phase and Si diffusion through the thickening columnar film. The steady state growth kinetics are controlled by WF6 diffusion in the gas phase whereas inhibition of the growth occurs at the transition from WF6 gas diffusion limited to Si solid state diffusion limited growth. A simple model based on WF6 gas phase diffusion and Si solid state diffusion is presented which gives a quantitative description of the experimental results.

Notes: Deep-level transient spectroscopy (DLTS) and constant-capacitance DLTS (CC-DLTS) techniques have been used to investigate selenium-related DX centers in AlGaAs alloys. The value of the thermal activation energy obtained by both techniques was the same (0.21 eV); however, experimental curves show some important differences. While CC-DLTS curves show only one peak, which reveals that there exists only one DX center in Se-doped AlGaAs, in DLTS curves it is possible to resolve up to two peaks lying at a lower temperature than the one observed by CC-DLTS. This disagreement may be due to the fact that DLTS measurements are strongly affected by refilling effects which occur in the edge zone of the space-charge region during capacitance transients performed at constant voltage. These effects accelerate the capacitance transients and can lead to too high thermal-emission rates. In contrast, these effects do not affect CC-DLTS curves, because in constant-capacitance voltage transients the edge of the space-charge region remains unchanged and refilling effects do not take place. These effects are rather important on DX levels because they exhibit thermally activated capture cross sections and very low ionization factors at the experimental temperatures and, therefore, capture processes are slow and their time constants can be similar to those of the emission processes.

Notes: We have investigated the effects due to the capture of tunneling electrons by interface traps to the measured capacitance, Cm, and equivalent series resistance, Rm, of insulator-semiconductor interfaces in metal-insulator-semiconductor (MIS) capacitors in accumulation. A new circuit model taking into account the capture of tunneling electrons by interface traps is derived. Theoretical and experimental results of Si3N4/Si/GaAs, Si3N4/epi-Si, and SiO2/epi-Si MIS capacitors are compared. The Si, Si3N4, and SiO2 layers investigated were deposited in situ by electron cyclotron resonance generated plasma in an ultrahigh vacuum chemical vapor deposition chamber. Frequency dispersion of both Cm and Rm can be adequately explained by the new circuit model.

Notes: Carbon-doped GaAs with carbon concentrations ranging from 2×1017 cm−3 to 2.6×1020 cm−3 has been characterized by variable temperature Hall effect measurements, secondary ion mass spectrometry (SIMS), and double-crystal x-ray diffraction (DCXD). The samples studied were grown by metalorganic chemical vapor deposition (MOCVD) and by metalorganic molecular beam epitaxy (MOMBE). The hole mobility is dominated by degenerate conduction for hole concentrations ≥1×1019 cm−3, and the 77 K resistivity is typically 30%–35% lower than at 300 K in these samples. The mobilities of C-doped p+-GaAs are found to be significantly higher than for Zn- or Be-doped p+-GaAs for doping concentrations in excess of 2×1018 cm−3. The maximum achievable hole mobilities for C-doped material grown by the two techniques are nearly identical, indicating that neither MOCVD nor MOMBE has an inherent advantage over the other for producing low-resistivity p-type GaAs. SIMS analysis and Hall effect measurements reveal that the total carbon concentration, [C], is higher than the as-grown hole concentration, p, in the most heavily doped samples. DCXD measurements show general agreement with the lattice mismatch predicted by Vegard's law. However, for [C](approximately-greater-than)1020 cm−3 a discrepancy between the predicted and measured mismatch suggests that partial lattice relaxation or the presence of interstitial carbon may need to be considered in order to adequately describe the lattice contraction.

Notes: The relation between short-wavelength optical loss increases and defect concentrations has been examined for a single-mode germanosilicate fibers exposed to H2 at 150–250 °C. The hydrogen-induced loss increases under these conditions appear to be due to the reaction of H2 at Ge–E' centers, resulting in the formation of —OH and "GeH'' defects, the latter of which is responsible for a 650-nm fluorescence. "GeO'' defect levels, characterized by fluorescence and optical loss measurements, remain unchanged in these experiments. This indicates that "GeO'' defects are not part of the reaction mechanism that lead to short-wavelength loss increase at these temperatures, although they may cause other types of short-wavelength loss increase at higher temperatures.

Notes: In recent years a number of workers have observed, using laser light-scattering methods, contamination particles suspended in rf process plasmas. Some recent studies show that the regions occupied by the particles appear finite and well defined, e.g., a ring. In this paper, a tuned Langmuir probe has been used to measure the plasma potential of these regions. Five cases are considered, where a case is distinguished by a disk of material A placed upon a larger diameter disk of material B, which in turn is placed upon an aluminum electrode driven by the rf power. It is consistently found that the regions occupied by particles are at a larger potential than the surrounding or ambient plasma. These regions are electrostatic traps. It is found that the incremental increase in potential as well as the geometry of the trap is dependent on materials A and B. For example, the trap configurations found are: case 1, silicon on graphite: ring and disk traps; case 2, silicon on silicon: ring trap; case 3, 304 stainless steel on silicon: disk trap; case 4, aluminum on silicon: bowl-shaped trap; case 5, silicon on aluminum: T-shaped or mushroom-shaped trap. Case 5 also contains a ring-shaped antitrap, i.e., a region where the potential is less than the ambient potential. The scenario proposed is that traps are system generated and depend on the design of the tool and the electrode. Particles subsequently flow into the traps. Particles may distort the trap boundaries as the density of trapped particles increases. However, particles do not determine the basic configuration (e.g., disk) of a trap, nor do they determine the plasma potentials within the trap or the surrounding ambient plasma region. In nearly all cases the trap and antitrap regions appear to be bounded by a double layer. Measurements of the plasma-sheath interface show that it generally follows the topography of the driven electrode. Explanations are presented for many of the observed phenomena.

Notes: The interdiffusivity within multilayered electrodeposited thin films of NiPx/NiPy has been measured using an x-ray technique. The NiPx, which can be deposited in the range 0≤x≤25 at. %, is crystalline at low phosphorus content but amorphous at high phosphorus content, so that films with a range of crystalline and amorphous structures could be prepared. Measurements on fully amorphous multilayers, where x,y(approximately-greater-than)18 at. %, give an interdiffusivity of order 10−22 m2 s−1 at 120 °C, which decreases with time, probably due to structural relaxation. The diffusivity decreases with decreasing repeat length of the films, with a critical repeat length of 40 A(ring). Crystalline/amorphous films show qualitatively different behavior, with the low-phosphorus layer governing the extent of a more limited diffusion phenomenon. This has been interpreted as the diffusion of phosphorus into a limited number of sites within the grain boundaries of the crystalline layers.

Notes: The microhardness, H, and Young's modulus, E, of a polycrystalline diamond film and several amorphous diamondlike carbon (DLC) films were determined from force-displacement curves obtained using an ultralow-load microhardness instrument (UMIS-2000). Measurements were made at a constant loading rate of 3 mN/s, to a maximum applied force of 67 and 100 mN with contact force of 0.06 and 1.07 mN, respectively. The diamond film had a surface morphology typical of microwave plasma chemical vapor deposition films (crystallite size 0.5–3 μm), and the force-displacement curves showed nearly complete elastic behavior. The average values of hardness (80–100 GPa) and modulus (500–533 GPa) are comparable to those of natural (001) diamond reference standards (H=56–102 GPa, E=1050 GPa). The DLC films were prepared by low-energy ion-assisted unbalanced magnetron sputtering. By varying the bombarding ion energy, five films were prepared having different sp3/sp2 bonding ratios (3–6), optical gaps (1.2–1.6 eV), and hydrogen concentrations (4–20 at %). The force-displacement measurements are characterized by substantial elastic recovery, and individual films show a very narrow range of hardness and modulus values. It is found that high hardness and improved modulus in DLC films correlate with increasing ion energy, sp3/sp2 bonding ratio, and energy gap. Individual films have mean values of hardness and elastic modulus in the range 12–30 GPa and 62–213 GPa, respectively.

Notes: Ion beam assisted evaporation was used to deposit cubic and hexagonal boron nitride thin films. Boron was evaporated and bombardment was by argon and nitrogen ions. The effect of preparation conditions on the resulting phase was studied, and the relationship between the phase and the energy and momentum transferred into the film through ion bombardment was examined. It is shown that for a given temperature, the controlling factor in the resulting thin film phase is the momentum transferred into the film per depositing boron atom. At 300–400 °C a sharp threshold value of momentum-per-atom exists below which films are hexagonal and above which they are cubic. For 400 °C this threshold occurred at 200 (eV×amu)1/2 which is equal to 3.3×10−21 m kg s−1. Depositions performed using krypton and xenon instead of argon as the second bombarding gas confirmed this momentum-per-atom value. A second threshold was also observed, which was bombarding species dependent, above which either complete resputtering of the deposited material or reversion to the hexagonal phase occurred. Cubic boron nitride deposition was seen to occur in a window of momentum-per-atom values between these two thresholds. Using this information it was possible to grow cubic boron nitride using only nitrogen bombardment, although the window of momentum-per-atom values for nitrogen is very narrow. The effect of substrate temperature was studied, and it was found to be difficult to grow predominantly cubic phase films below 300–400 °C. The relationship between intrinsic stress and phase of the films is also discussed. A diagram is presented showing film phase as a function of bombardment, substrate temperature, and system chemistry. The parameter of momentum-per-atom is shown to combine into a single value the variables of ion beam assisted deposition: deposition rate, ion energy, ion flux, and ion species. It is suggested that, in general, for properties affected by ion bombardment the momentum-per-atom transferred into the film is the controlling factor. The results are shown to support momentum transfer as the dominant process in cubic boron nitride thin film formation.

Notes: We have modeled the lateral quantum confinement in a GaAs quantum well due to partial strain release on the side walls of an InGaAs stressor located on the top of the well. We used the finite element method to solve both the continuum elasticity equation and the Luttinger four-band Hamiltonian with strain. This model was used to systematically study the wire width dependence of lateral band edge modulation, subband splitting, and relative transition strength in the strain-induced quantum well wire structures.

Notes: Subband effective mass and mobility of a two-dimensional electron gas in uniformly Si-doped Ga0.47In0.53As/Al0.48In0.52As multiple quantum wells are estimated. In three samples with well widths of 130, 200, and 300 A(ring) and a fixed barrier width of 68 A(ring), up to three two-dimensional subbands are found. Maximum enhancement of electron effective mass is 25% over the bulk value at a Fermi energy of 108 meV. These estimated effective masses are slightly smaller than previously reported values, and this is explained by recent theories which assume only nonparabolic correction. When the temperature ratio, (Dingle temperature)/(mobility temperature), is assumed to be identical in different subbands in each sample, the electron mobility of the first excited subband is at most twice that of the ground subband in the sample with the widest well. The effects of intersubband scattering and of screening on the subband mobilities are also discussed.

Notes: The chemical composition of the surfaces and interfaces of NbN-MgO-NbN trilayers has been studied by x-ray photoelectron spectroscopy during the fabrication of the trilayer without breaking the vacuum. The NbN and MgO layers were prepared by rf magnetron sputtering. The results of the chemical analysis have been correlated to the electrical characteristics of the completed NbN-MgO-NbN tunnel junctions. During the deposition of the MgO barrier layer the presence of a high amount of energetic oxygen ions and atoms in the sputtering plasma results in a strong plasma oxidation of the NbN base electrode and, hence, in mixed Nb2O5-MgO barriers. The oxygen ions and atoms are generated by the dissociation of the target material and the water of the background pressure. Their amount was found to increase with increasing argon pressure during the MgO sputtering process. Also, adsorption layers of hydroxides on the MgO-target result in the formation of an uncontrollable amount of niobium oxide components at the interface between the NbN base electrode and the MgO barrier. The current-voltage characteristics of tunnel junctions with such barriers show large subgap leakage currents. Pure MgO barriers can be prepared by reducing the oxygen bombardment of the NbN films during the MgO deposition. Pure MgO barriers are oxygen deficient and easily adsorb hydroxides. These hydroxides react with the first layers of the NbN top electrode to NbO2 and NbO thereby reducing and broadening the sumgap value of the tunnel junctions. Tunnel junctions with pure MgO barriers of a nominal thickness of less than 2 nm usually have current-voltage characteristics indicative for microshorts. A special annealing procedure of the NbN-MgO bilayers prior to the deposition of the top electrode desorbs the hydroxides, transforms Mg(OH)2 to MgO without forming metallic magnesium and prevents the formation of intermediate layers of niobium suboxides and metallic shorts.

Notes: Thin films of potassium tantalum niobate, KTa0.6Nb0.4O3, with a Curie temperature of 20 °C were deposited on a variety of substrates by metalorganic deposition. These films had peak relative permittivities of 16 000 at 20 °C. Hysteresis plots of electric displacement as a function of electric field, taken at 0 °C, revealed a coercive field of 800 V/cm, a spontaneous polarization of 3.9 μC/cm2, and a remnant polarization of 0.5 μC/cm2. The hysteresis loops did not change significantly as the temperature was varied down to −100 °C.

Notes: All 17 infrared lattice modes for yttrium aluminum, yttrium gallium, and yttrium iron garnets Y3R2(RO4)3 are obtained from single-crystal reflection and absorption spectroscopy, and the powder technique. Assignments for 12 modes are firmly establish based on chemical substitution among these garnets and within related series. Frequencies of the modes associated with translation of yttrium are nearly constant across the series. Frequencies of all other modes decrease quasilinearly as cell volume increases. Values the high and low frequency dielectric constants are in good agreement with independent measurements; also, the Lyddane–Sachs–Teller relationship was found to hold. Calculations of bulk moduli from vibrational spectra using a modification of Brout's formula linearly correlate with experimental values for KT.

Notes: First phase formation has been determined in Cu binary thin film systems with Ti, Zr, Mg, Sb, Pd, and Pt using transmission electron microscopy and Rutherford backscattering spectrometry. CuTi, CuZr, CuMg2, Cu2Sb, Cu3Pd, and Cu3Pt are the first phases to form upon annealing the Cu/metal bilayers. The effective heat of formation model is used to predict first phase formation in 14 Cu/metal systems.

Notes: A modification of the optical technique of impulsive stimulated thermal scattering that allows real-time data acquisition is introduced. The method is used to characterize the pseudo-Rayleigh acoustic modes of thin (∼1 μm) polyimide films attached to silicon substrates and its potential use as a means for extracting the elastic constants of thin-film coatings is demonstrated.

Notes: Studied here is the temperature rise in thin films during heating produced by ac magnetic fields. Eddy currents induced by ac magnetic fields in tantalum foils and silicon thin films are calculated for previously reported recrystallization equipment. Two kinds of transient temperature rise, produced by the power loss resulting from the eddy currents, are calculated on the basis of a one-dimensional variation of Fourier's thermal equation. The first calculation, along the r axis of a cylindrical coordinate system centered at the equipment coil center, shows that temperature is strongly dependent on distance from the coil center. The calculation further suggests that tantalum can be successfully heated to over 1000 °C in a short time with the equipment, and that undoped silicon can be heated only when the effective initial temperature is over 500 °C. The second calculation along the z axis perpendicular to a SOG (silicon on glass) wafer surface, shows that undoped silicon thin film can be heated to over 500 °C in a short time when the Ta-foil/SOG-wafer gap is kept within a 0.1–1 mm range. Calculated temperature distributions agree well with those measured experimentally on Ta-foil surfaces. The results of these calculations contribute to an understanding of the recrystallization process and provide important suggestions for improving the SOG recrystallization produced by ac magnetic fields.

Notes: The effects of heavy Si doping in GaAs layers grown by molecular beam epitaxy as a function of varying arsenic to gallium (group V/III) flux ratio have been studied. By using a slow growth rate of 1500 A(ring)/h, a high electron concentration of ≈5.5×1018 cm−3 was achieved, which decreases sharply following a decrease in the group V/III flux ratio. The low-temperature photoluminescence spectra also show a broad low-energy peak (≈1.284 eV) as the flux ratio decreases. Secondary ion mass spectroscopy measurements show that the effect of Si accumulation near the GaAs surface is increased at lower flux ratios.

Notes: Epitaxial growth of BaTiO3 and SrTiO3 films by the reactive evaporation method was investigated using reflection high-energy electron diffraction (RHEED). The investigations were carried out using two growth methods: coevaporation and alternate evaporation of the metal elements in an oxygen atmosphere. Atomic layer growth was achieved by the alternate supply of Ba or Sr and Ti on the growing surface. In the case of coevaporation, epitaxial growth occurred in a two-dimensional unit-cell-by-unit-cell mode. The surface of each unit cell is terminated by a (TiO2) layer. Artificial superlattices of BaTiO3/SrTiO3 were fabricated by monitoring the film thickness with the RHEED oscillations.

Notes: The stoichiometry, crystallinity, defect concentration, and the excess As lattice location in GaAs layers grown by molecular beam epitaxy at low growth temperatures (≤300 °C) were studied using ion beam techniques. The excess As concentration in the layers was measured by particle induced x-ray emission and was found to increase as the growth temperature was lowered. Excess As concentrations up to 1.5 at. % were measured in layers grown at 190 °C. After annealing at temperatures higher than 400 °C under As overpressure, the excess As atoms coalesce to form As precipitates as revealed by transmission electron microscopy. Ion channeling on the unannealed layers grown at 200 °C revealed that they have good crystalline quality with a large fraction of the excess As atoms sitting at interstitial sites close to the normal As sites in the lattice. The rest of the excess As atoms are believed to be in an AsGa antisite position.

Notes: The density and distribution of dangling-bond defects in a given sample of hydrogenated amorphous silicon (a-Si:H) depends on thermal history and on electron and hole densities. As a consequence, the defect density of a-Si:H devices changes during their operation. Optoelectronic device modeling therefore must be coupled with the modeling of the density and distribution of dangling-bond defects. In this paper such a defect model is presented. The model accounts for the effects of the temperature of film growth, the rate of film growth, the film thickness, light-soaking intensity, time, and temperature, and the temperature and duration of thermal annealing. The model is based on the assumption of a limited pool of defects with a Gaussian distribution of thermal annealing energies. This is the most simple model that agrees reasonably well with a variety of experimental data. The applicability of the model is demonstrated on many examples drawn from measurements of dark and photoconductivity, transmission spectroscopy, and subgap optical absorption.

Notes: The thermal diffusivity, electrical resistivity, density, and heat capacity have been measured for Cu6Sn5, Cu3Sn, and Ni3Sn4. These properties were used to evaluate the apparent Lorenz number for comparison with the theoretical value. Acceptable agreement was found.

Notes: An algorithm for separating the bulk and surface components of recombination lifetime obtained via a contactless single laser excitation/microwave reflection decay measurement is presented. The surface recombination component of lifetime is determined by extrapolating the tail portion of the carrier decay curve to the carrier axis. Although the slope of this curve depends on both surface and bulk properties, it is shown that the y intercept depends only on the surface component of lifetime. A wide range of surface lifetimes, corresponding to surface recombination velocities from 102 to 105 cm/s, and bulk lifetimes from a few microseconds to several hundred microseconds can be measured. An experimental verification of the analysis is presented using microwave absorption/reflection measurements on silicon wafers representing a wide variety of bulk and surface lifetime components.

Notes: The static field electrical conductivities of several discontinuous fiber composites are reported. Novel fabrication techniques are employed to produce systems of highly aligned and well-dispersed carbon fiber and coated carbon fibers. The thermal and breakdown properties of the matrix resin and breakdown behavior of the composites are examined. The distribution of orientations of the included phase is determined and recorded as an average orientation descriptor. Degree of fiber aggregation is found to be the principal determinant of bulk conductivity, eclipsing the effects of orientation and fiber fraction for the dilute concentrations investigated. A conductivity scale dependence is found for materials with tightly constrained microstructures.

Notes: The frequency-dependent conductivity of sputtered (with and without hydrogen) and glow-discharge amorphous silicon-nitrogen alloy films has been measured in the temperature range 15–330 K. The ac loss decreases with the introduction of hydrogen. The origin of the ac loss is interpreted in terms of electron transfer between dangling-bond states of silicon and nitrogen-related two-level systems.

Notes: Silicon samples doped with tin and/or phosphorus have been irradiated at room temperature by 2.0 MeV electrons in order to study the generation of divacancy centers (V2). The samples were analyzed using low-temperature infrared absorption measurements and deep level transient spectroscopy. The production of V2 is found to be very small in the tin-doped samples, which suggests that the generation process hinges strongly on the existence of mobile monovacancies. Substitutional Sn atoms act as efficient traps for migrating vacancies and form stable vacancy-tin centers with a corresponding reduction in the formation of "ordinary'' vacancy-related defects. Indeed, at temperatures above ∼150 °C where the vacancy-tin centers dissociate, a pronounced increase in the V2 concentration occurs. On the other hand, measurements of the V2 concentration in the non tin-doped samples as a function of electron dose using different dose rates reveal that the steady-state concentration of monovacancies during irradiation has no influence on the generation of V2. Thus, pairing of two single vacancies created by different bombarding electrons can be excluded as a major formation mechanism for V2. Evidence is presented for a new model where the formation of V2 is predominantly due to migration and subsequent agglomeration of two single vacancies created directly (electron-atom collision) and indirectly (recoiling atom-atom collision), respectively, by the same impinging electron. Finally, the presence of substitutional phosphorus atoms in high-purity float zone Si samples is shown to enhance the production of V2. The experimental data exhibit excellent relative agreement with computer simulations assuming a defect reaction model where annihilation of vacancy-phosphorus centers by Si self-interstitials is taken into account, i.e., the role of phosphorus is to provide an alternative recombination center for the Si self-interstitials with a corresponding growth of surviving V2 centers.

Notes: High current discharge plasma produced by the Ta-LaB6 cathode was used for ion plating of TiN films. The dependence of the deposition rate of TiN on the substrate dc bias potential was measured using the Rutherford backscattering spectrometry (RBS) method. The results indicate that the particles impinging on the substrate, i.e., vaporized particles (Ti) and reactive gases (N2), could be ionized at high efficiency of 70%. Epitaxial TiN films were grown at temperatures between 400 and 700 °C on cleaved MgO(100) substrate supplied with negative dc anf rf potential. Structures of the epitaxial film were investigated by RBS and ion channeling, and scanning tunneling microscope (STM). The minimum channeling yields for (100) and (110) axes obtained at a substrate temperature of 550 °C with the growth rate of 40 A(ring)/s were found to be 1.4% and 2.0%, respectively, which were almost the same as that of the MgO substrate. Atomically resolved tunneling images of (100) surface of the epitaxial TiN film were observed by STM.