ALBERT

Notizen: Studies of nonlinear acoustic interactions in superfluid helium at temperatures below 0.2 K have culminated in the construction of an all-acoustic parametric amplifier at gigahertz frequencies. This device represents the shortest wavelength parametric amplifier ever made, with signal wavelengths shorter than 1000 A(ring) and pump wavelengths shorter than 600 A(ring). In the experiment, plane waves are mixed at a predefined angle in the superfluid helium. Two gain regimes are observed. The first regime is a noncollinear phase-matched process, in which the signal phonon stimulates decay of the pump phonon to create gain at the signal frequency. The second regime is a four-phonon collinear process, in which gain on the signal is created by the second harmonic of the pump. This four-phonon process is unusual and is shown to be a combination of three-phonon processes wherein the lack of conservation of energy and momentum in one process is mostly compensated by the other.

Notizen: The ferromagnetic Heusler alloy Pd2MnSn shows large reduction in magnetization by cold working, without any appreciable change in the Curie temperature. In order to clarify the relationship between the reduction and the defect structure in Pd2MnSn, transmission electron microscopic observation, high-resolution neutron powder diffraction, and magnetization studies have been performed. In the deformed Pd2MnSn sample, dislocations are distributed inhomogeneously and form a cell structure. In the cell walls, a large amount of antiphase domain boundaries exist. A method of Rietveld refinement has been proposed for analyzing the powder neutron diffraction patterns of the cold-worked sample, where all the contributions of defects (size, strain, and antiphase domain boundary) and the instrumental resolution to each reflection line are expressed in terms of Fourier coefficients, and inverse Fourier transforms of those product represent the calculated profile which is used for least-squares refinement. Analyzing the experimental data with the method proposed, it is concluded that the reduction in magnetization in the deformed Pd2MnSn is due to an increase in the number of small antiphase domains which couple antiferromagnetically to each other.

Notizen: A Raman study has been performed, under resonant conditions, on a GaAs bevelled-edge layer grown on a Si substrate to characterize the optical and crystalline properties of the epilayer near the interface. According to the geometrical characteristics of the sample, a theoretical expression for the Raman intensities profile has been established and compared to the experimental data. This fitting procedure enables us to investigate the absorption coefficient of the GaAs layer due to the disorder-induced softening of the E1 edge. A quantitative analysis of the lattice disorder has been carried out on both longitudinal and transverse optical modes by studying the Raman line-shape evolution versus the laser spot position on the bevel edge. From this study, we have followed the recovery of the crystalline quality of the epilayer while going away from the interface, and evaluated the "Raman thickness'' of the dislocated layer. Using the spatial correlation model as a relationship between the disorder amount and the outcoming effects on the Raman line peaks, we have estimated the dislocation density at the heterostructure interface.

Notizen: Examination with x-ray photoelectron spectroscopy (XPS) of CF4 plasma etched GaAs(100) wafers unveils a residual surface reaction layer composed of Ga-fluoride and Ga-oxide. Efficient removal of this contamination layer by a brief immersion in a dilute NH4OH wet etch is demonstrated. The formation of a thin native oxide upon exposure to atmosphere of the clean substrate surface cannot be avoided, however. Prospective replacement of this wet etch processing by in situ thermal annealing in hydrogen was investigated. The recorded XPS spectra show almost complete desorption of fluorine after annealing at 200 °C, whereas a temperature of ∼600 °C is required for entire removal of residual surface Ga-oxide. Heat treatment in H2 also compares favorably with vacuum annealing, for which a noticeable reduction of the surface contamination layer was found only after annealing at 600 °C. The cleaning efficiency of hydrogen processing may be attributed to the reactive nature of this ambient.

Notizen: Time-dependent production of higher-silane gases and a-Si:H film are measured relative to decomposed silane in rf and dc, hot and cold cathode, static-gas discharges. From the absence of higher-silane production in very low silane partial-pressure discharges, it is inferred that most higher silanes are produced by gas-phase SiH2-initiated reactions. The higher silanes are thus tracers of SiH2, while the film production traces the fraction of H, SiH, and SiH3 in the initial decomposition. From the measured stable product yields, we deduce that SiH4→SiH2+2H is the dominant electron-collisional dissociation channel.

Notizen: Three methods of evaluating the second-order contributions to the mechanical velocity shift and reflection coefficient for isotropic groove gratings are presented. The second-order terms contain the effect of evanescent and bulk modes generated through the interaction of a surface acoustic wave with a periodic overlay. All three methods will be shown to be numerically equivalent when calculated self-consistently, verifying the validity of all three within their respective limitations. The extension of these methods to arbitrary anisotropic substrates and overlays is discussed.

Notizen: Characteristics of phototriggered discharges in pure argon and neon have been experimentally and theoretically investigated at pressures between 0.5 and 5 bars and reduced electric fields E/N in the range 5 to 50 Td. Emphasis is laid on the breakdown delay time, defined as the time lag between the application of a very short preionization pulse and the occurrence of the discharge. The model, developed in the frame of the local field approximation, gives breakdown delay time values in excellent agreement with the experimental results. However, in pure neon, the experiments have revealed a large influence of very small amounts of easily ionizable impurities. The calculations point out that the discharge behavior is mainly determined by the electron multiplication rate due to direct ionization of ground state atoms. Other ionizing processes, such as two-step ionization or Penning effects, or the electrical parameters of the circuit (inductance, capacitance, and intensity of the preionization) have only a secondary effect on the breakdown delay time.

Notizen: Fully developed, viscous liquid-metal velocity profiles and induced magnetic field contours were studied for Hartmann numbers of M=2 and 10 and for different load currents for a particular rectangular channel configuration (two-dimensional Couette flow). The rectangular channel was assumed to have a homogeneous external (axial) magnetic field parallel to the moving, perfectly conducting top wall and the stationary, perfectly conducting bottom wall. The two stationary side walls were also perfect conductors. The small gap between the moving wall and each side wall was an insulating, free surface. The method of weighted residuals was used to obtain truncated series solutions for the variables of interest. The heavy load currents across the channel were obtained by simulating an external potential to the conducting moving wall. The load currents in each case were opposed by the induced electric field. Since there is no pressure gradient, the flow along the channel is driven by the viscous effects of the moving wall and the Lorentz body force and is retarded by the stationary walls. In the case where no load current is applied across the channel, the current circulates in the channel.The circulation is driven by the generator that is due to the axial variation of velocity in an axial magnetic field. The numerical results presented show that the radial gap and the free surface region represent electrical resistances in parallel between the perfectly conducting stationary wall and the perfectly conducting moving wall. The numerical results also show that the resistance of the radial gap increases as M2 while that of the free surface increases by M or M1/2. Thus, as M increases, the division of current shifts to the free surface region and the current density in the radial gap decreases as M−1. The theoretical magnetohydrodynamic model presented here was developed to provide numerical parameters to help in the design of liquid-metal current collectors. Numerical results were computed for one-dimensional Couette flow with no pressure gradient in an external, homogeneous axial magnetic field. One-dimensional Couette flow has no end effects, and thus the numerical results were compared with corresponding numerical results for the two-dimensional Couette flow case to determine end effects.

Notizen: The heat capacities of liquid and crystalline Au-Pb-Sb alloys in the glass-forming composition range were measured with droplet emulsion and bulk samples. Based on the measured Cp data, the entropy, enthalpy, and Gibbs free-energy differences between the eutectic solid mixture and undercooled liquid were determined as a function of temperature over ∼60% of the undercooling range below the liquidus temperature and compared with theoretical predictions. The results indicate an isentropic temperature at 313 (±5) K, which agrees well with experimental data for the glass transition. The thermodynamic evaluation was applied further to develop a kinetics analysis of the nucleation undercooling response during cooling. Use of different approximations for the Gibbs free energy leads to a variation of the prefactor terms of six orders of magnitude for classical nucleation theory and, consequently, large variation in calculated transformation diagrams which is more pronounced with increasing undercooling. Extrapolations into the glass-forming temperature range and the effects of viscosity, transient nucleation, and estimated Kauzmann temperatures on the crystallization kinetics at high undercooling have been evaluated. This analysis reveals the importance of using measured values of thermophysical properties, even if they represent a limited temperature range at modest undercooling, rather than model approximations in order to obtain reliable evaluations of crystallization kinetics at high undercooling in the glass-forming temperature range.

Notizen: Shock waves of order 100 kbar were launched into 50-μm-thick single crystals of silicon (111) by irradiation with nanosecond pulses of 1.05-μm laser light at irradiances in the region of 2×1010 W cm−2. A separate laser beam, synchronous but delayed with respect to the shock-driving beam, and containing approximately 25 J of 0.53-μm laser light in a pulse of 1 ns (FWHM), was focused to a tight (〈100 μm) spot on a separate titanium target to produce a plasma which was a prolific source of He-like Ti x rays. The x rays were Bragg diffracted from the rear surface of the shocked crystal and the spectrum recorded on an x-ray streak camera. Changes in interatomic spacings in a region within several microns of the surface were thus deduced from the resultant shift in Bragg angle with a temporal resolution of 50 ps. Shock waves with compressions of order 6% were observed. We observed the crystal in a state of dynamic tension as the two rarefaction waves met. The results are in good agreement with hydrocode simulations in conjunction with x-ray diffraction calculations.

Notizen: Amorphous Si/Ge artificial multilayers with a repeat length around 60 A(ring) have been partially mixed with 1.5-MeV Ar+ ions at temperatures in the range 77–673 K. The diffusive component of the square of the mixing length, obtained by subtracting out the ballistic contribution, does not depend on the dose rate at a given dose, and shows an Arrhenius-type temperature dependence with activation enthalpies between 0.13 and 0.22 eV. Possible mechanisms for migration and annihilation processes of defects are discussed to understand these low activation enthalpies.

Notizen: ZnTe-ZnS short period (∼30 A(ring)) strained layer superlattices were grown by molecular beam epitaxy and the growth process was observed in situ by using reflection high energy electron diffraction (RHEED). Surface lattice constant (a(parallel)) behavior shows a transition from one state in which a(parallel) is equal to that of the substrate to the other in which a(parallel) depends on the depositing constituents. RHEED patterns show the same streaks for both states except the difference in streak separation. Transmission electron micrograph reveals that the resultant multilayer structures of the two states have the same quality. The experimental data may suggest a possible assumption that there exists multiple states for two-dimensional growth of a strained system with substantial lattice mismatch.

Notizen: A detailed optical study of the metastable III/V semiconductor alloy InP1−xSbx is presented. InP1−xSbx layers are grown throughout the entire compositional range by atmospheric pressure organometallic vapor phase epitaxy on InP, InAs, and InSb substrates. Composition and strain are measured by combined electron microprobe analysis and x-ray diffractometry. The dependence of band gap on composition is experimentally established for the first time from absorption spectra measured at 10 and 300 K. The resultant value of the band-gap bowing parameter is 1.52±0.08 eV, independent of temperature. The absorption spectra show the InP1−xSbx layers to have long band tails, which extend further into the gap as the Sb concentration is increased. The band tails are induced by compositional clustering. Photoluminescence (PL) spectra are measured between 10 and 300 K. The PL peaks are assigned to recombination between carriers occupying band-tail states or to recombination via deep centers in the gap.

Notizen: Hole trapping phenomena in SiO2 were examined using an optically assisted hot carrier injection technique on p-channel insulated gate field effect transistors. It was found that only a single, field-dependent, capture-cross-section hole trap is present. The capture cross section of these hole traps at a field of 4 mV/cm across the gate insulator, corresponding to a gate voltage just above the threshold voltage, was found to be 8.5×10−14 cm2. Injected holes were found to trap with an initial efficiency of approximately 60% at this gate field. Depopulation of trapped holes at room temperature was also examined, and found to be significant. The neutral hole trap density in unirradiated device gate insulators after post-metal annealing was found to be approximately 7.0×1012 cm−2. Based on a study of the threshold voltage shift as a function of gate insulator thickness, coupled with the model recently proposed by Walters and Reisman for determining charge centroid, it appears that for oxides with thicknesses greater than 10 nm, the hole traps lie in a band of finite thickness with a charge centroid 5 nm from the substrate-SiO2 interface. In addition, there exists a layer approximately 3.7 nm thick at each interface that appears void of trapped charge. Therefore, oxides less than 7.4 nm thick should not trap charge, which was found to be the case experimentally. This implies that as devices are scaled down, hole trapping will disappear, which is of particular significance in oxides subjected to ionizing irradiation, either during processing or during use.

Notizen: The problem of calculating the valence-band structure of strained-layer quantum wells in the effective-mass approximation is reviewed. Using the spherical approximation and exploiting the simplicity of the infinitely deep well model we show that the in-plane effective mass is determined by two factors—a splitting contribution which is dominant at large strains, and a quantum confinement contribution. A model for finite-depth wells is presented which gives analytic expressions for the zone-center in-plane mass and associated nonparabolicity factor, and it is applied to the system InxGa1−xAs/GaAs. The model allows the computation of valence-band structure using no more than a pocket calculator. It is shown to give results in reasonable agreement with experiment.

Notizen: Silicon dioxide layers of 250 A(ring) thick were grown on Si at 1000 °C in a dry O2/TCA ambient. Thermal nitridation of the samples was performed in a pure ammonia ambient at temperatures from 900 to 1100 °C with one hour time intervals up to a maximum of 4 h. The fixed charge state densities at the interface of the samples were determined from high frequency C-V measurements, and the breakdown fields from I-V curves. Secondary ion mass spectroscopy depth profiles show low levels of contaminants, and high levels of nitrogen at the interface for samples annealed at temperature greater than or equal to 1000 °C and for periods longer than 2 h. Post metal annealing of the nitrided samples appears to help in reducing the trapped charges. Better quality films with lower Qf and VFB shifts, and higher breakdown fields were achieved for samples annealed at 1100 °C. Metal-oxide-semiconductor device quality nitrided films with a Qf of 1010/cm2 were achieved by optimizing the process conditions at 1100 °C. The fixed charge build up for lower nitridation temperatures (〈1000 °C) and times (〈2 h)is due to the dissociation of Si—O bonds in the presence of hydrogen, and is in accordance with the earlier results in the literature. However, the reduction in the fixed charge buildup at 1100 °C, we believe, is due to the increased levels of nitrogen.

Notizen: Josephson voltage standards utilize microwave-induced constant voltage steps in the dc characteristic of Josephson tunnel junctions. This paper describes the design and operation of array circuits with 108 and 2000 junctions connected in series. In contrast with similar realizations, simple Q-band equipment is used for the microwave supply. The microwave attenuation of 1000 junctions was about 1 dB. The version with 2000 junctions generated Josephson voltages up to 1.2 V when operated at 35 GHz. The stability times of the quantized levels were, under normal laboratory conditions (unshielded room), better than 10 min.

Notizen: The results obtained in part I are interpreted in terms of the viscosity field arising from independent processes of directional ordering for magnetic defects dispersed in the amorphous structure and interacting with the magnetization vector. A specific model is developed in order to take into account the changes in the ordering kinetics induced by the periodic magnetization rotations described in part I. This model, however, requires that the magnetic induction remain constant during the whole measurement; as a consequence, the model's predictions cannot be directly compared with the experimental results, obtained instead at constant applied field. This difficulty is overcome by deriving a general relationship between the magnetic-induction decay and the viscosity field kinetics for an arbitrary number of half-periods of the square-wave field. The agreement of our theory with the experimental results turns out to be quite satisfactory. As consequence, the ordering processes responsible for the magnetic aftereffect in amorphous ferromagnets may be described as essentially uncorrelated.

Notizen: We have observed a very interesting feature in the mid-infrared region of copper-doped semi-insulating liquid encapsulated Czochralski grown GaAs at liquid helium temperature. After gettering copper using backside mechanical damage, the transmittance of this particular feature became much weaker. The first absorption feature occurs at 1182 cm−1 (0.147 eV), the second at 1460 cm−1 (0.181 eV), and the third at 1750 cm−1 (0.217 eV), which is a transition of an electron from the valence of three copper levels. This electronic mode absorption allows us to obtain information about the activation energies of deep-level impurities like Cu and their relative concentration after each process. This electronic mode absorption at multilevels of Cu in GaAs is proposed for the first time using the Fourier transform infrared technique to detect deep-level metal impurities in GaAs.

Notizen: The authors report time-resolved measurements of the emission of positive and negative charge from Si and Ge surfaces irradiated with 248-nm KrF excimer laser pulses. With pulse energies both below and above the melting threshold, the time evolution of the emission currents is complex and strikingly different for Si and Ge. The positive ion emission signal from Ge persists only for the duration of the laser pulse (〈60 ns), but in sharp contrast, the signal from Si continues for several microseconds. A tentative suggestion is made that the positive ions encounter a Knudsen layer created just above the surface of the Si target. More refined experiments, coupled with a theoretical effort, are proposed.

Notizen: Ultraviolet laser irradiation of surfaces, in the course of photoemission or surface photochemical studies, often produce copious electron emission, up to 1000's of A/cm2. The time-dependent fields produced by these electrons accelerate some of the electrons up to 5.4 times their initial energies. The steady-state fields return most of the emitted electrons to the surface. We discuss and illustrate both phenomena with theoretical simulations and experiment, and discuss possible implications.

Notizen: The composition of the reaction gases in a hot-filament reactor for chemical vapor deposition of diamond films was analyzed using a gas chromatograph coupled with a quartz microprobe. Concentrations of several hydrocarbons were determined as functions of filament temperature (FT) and the position of the probe relative to the filament for two feed gases, methane/hydrogen and acetylene/hydrogen. The diamond growth rate was measured as a function of FT in both feed gases. The major chemical process in these reaction systems is found to be conversion between methane and acetylene with ethane and ethylene as reaction intermediates. For FT≤1800 °C, the chemical reactivity is low, and no diamond deposition is observed. For FT≥1900 °C, nearly identical chemical composition near the filament is obtained from both feed gases (indicating possible attainment of thermodynamic equilibrium in the gas mixtures), and the measured diamond growth rates are similar. A substantial depletion of carbon in the reaction gases near the filament is observed and is shown to be due to thermal diffusion. Downstream of the filament, a nonequilibrium state is observed, possibly caused by slow methane/acetylene conversion and/or fast diffusion. The trend of the diamond growth rate, which increases with increasing FT from 1900 to 2200 °C, correlates well with the concentration profile of acetylene but not that of methane.

Notizen: Using (100) Cu(1000 A(ring)) /Ni/Cu(1000 A(ring)) structures with a different Ni thickness, a complete reversal in magnetic anisotropy is observed between the structures with 1000 and 50 A(ring) Ni, respectively. The structures were deposited on (100) Si, with the first Cu layer in the (100) orientation as the seed, followed by epitaxial Ni and another Cu layer. Comparing the magnetization for the field applied parallel to the film plane, M(parallel), with that perpendicular to it, M⊥, M⊥ is found to increase and M(parallel), to decrease with decreasing Ni thickness. At a Ni thickness of 50 A(ring), a squared hysteresis loop is observed for M⊥. The M(parallel) curve, on the other hand, changes from a normal one at 1000 A(ring) Ni to one with little hysteresis at 50 A(ring) Ni. These are accompanied by an increasing deformation of the (100)Ni cubic lattice into a tetragonal one with decreasing Ni thickness, due to its mismatch with Cu. The deformation is discussed to assess the roles of stress in the observed reversal in magnetic anisotropy.

Notizen: The emission spectra resulting from ion beam sputtering a Y-Ba-Cu-O target were observed as a function of beam voltage and beam current. The spectra were relatively clean with several peaks readily attributed to each of Y, Ba, and Ar. Monitoring of copper and oxygen was more difficult with a single CuO peak and one O peak evident. The intensities of the cation peaks were linear with respect to beam voltage above 400 V. Since target current was found not to be directly proportional to beam current, target power was defined as the product of beam voltage and target current. The response of cation peak height to changes in target power was linear and similar for variations of either beam voltage or target current.

Notizen: Optically transparent, free-standing single crystals of the high-temperature superconductor Bi2Sr2CaCu2O8 have been prepared with a typical size of a×b×c=1 mm×1 mm×1500 A(ring). The thickness of the samples has been determined by x-ray transmission measurements and was found to be uniform to within a lattice spacing. Transmission electron microscopy and diffraction revealed a pseudotetragonal lattice with a twinning free, incommensurate superstructure. X-ray microanalysis shows that the composition of the sample is homogeneous. Electrical resistivity and magnetic susceptibility measurements indicate that the crystals are good quality superconductors.

Notizen: Spatially resolved measurements of the change in resistance with temperature of bulk high-temperature superconductors are generated by heating a localized region of a sample with a modulated laser beam. The amplitude of the thermally induced resistance change is related to the heat capacity and derivative of resistance with temperature whereas the phase of the signal is related to the thermal diffusivity of the sample. By moving the sample with respect to the focused laser beam, spatially resolved measurements may be made. Images with ∼50-μm resolution have been used to locate defects in polycrystalline Bi-Sr-Ca-Cu-O sample and to estimate the thermal diffusivity of the bulk material.

Notizen: An efficient numerical algorithm was developed to solve the continuity equations describing charged particle transport and potential distribution in a low-pressure glow discharge in an argonlike gas. A parametric investigation of the effect of gas pressure and electrode spacing on the discharge properties was conducted.

Notizen: We have measured the epitaxial growth rate of Si in the disilane/hydrogen system for different experimental conditions in a rapid thermal processing–low-pressure chemical vapor deposition reactor. The measured activation energy is about 43 kcal/mol for temperatures below 700 °C. The growth rate dependence on disilane or hydrogen partial pressures is measured in order to verify the reliability of a proposed model for the decomposition of the disilane molecule in the torr pressure regime. The obtained crystal quality is comparable to that obtained with silane gas except for higher growth rate in the disilane system.

Notizen: Low-temperature photoluminescence was used to study the Si+- and Si++P+-implanted InP. A broadband at ∼ 1.26 eV appears in photoluminescence spectra for Si+-implanted InP after annealing. The intensity of this broadband decreases with increasing the coimplanted P+ dose and increases with increasing the implanted Si+ dose. The temperature dependence of the photoluminescence data shows that the change in the half-width of the broadband can fit the configuration coordinate model. This band is believed to be due to VP-SiP complex. The results indicate that silicon is an amphoteric species in InP.

Notizen: Stable and highly reproducible current-limiting characteristics are observed for polycrystalline ceramics prepared by sintering mixtures of coarse-grained, donor-doped BaTiO3 (tetragonal) as the major phase and ultrafine, undoped cubic perovskite such as BaSnO3, BaZrO 3, SrTiO3, or BaTiO3 (cubic). The linear current-voltage (I-V) relation changes over to current limiting as the field strength increases, when thermal equilibrium is attained. The grain-boundary layers with low donor and high Sn, Zr, or Sr have depleted charge carrier density as compared to that in the grain bulk. The voltage drop at the grain-boundary layers diminishes the temperature gradient between the interior and surface regions.

Notizen: A new self-aligned subtractive gate process is proposed by alternating several masking and polycrystalline silicon gate etching steps from a conventional process to build high-voltage (up to 100 V) and complementary thin-film transistors (TFT) on insulating substrates. The new process is compatible with conventional TFT mask sets and standard processing techniques. The advantage of using this new process is to eliminate the difficulty in stripping photoresist on insulating substrates after high-dose phosphorus implant and to improve the off-state breakdown voltage in high-voltage transistors. Complementary metal-oxide-semiconductor and high-voltage TFT devices were successfully fabricated by this new process.

Notizen: Nonlinear refractive index changes for the wavelengths near the fundamental edge in an ultrapure epitaxial layer of GaAs were obtained directly by the photoreflectance method. The nonlinearities are due to the electric field effect for pump intensities lower than 0.06 W/cm2 and the plasma screening and the band filling effect for pump intensities higher than 0.06 W/cm2.

Notizen: An improved version of a prior model is given for specific resistance of ohmic contacts to n-GaAs. Instead of the two-band model used in the prior work for the transmission coefficient, a more rigorous three-band model due to E. O. Kane [J. Phys. Chem. Solids 1, 249 (1957)] is used in the improved version. In the doping range of interest, the theoretical contact resistance can differ by more than an order of magnitude, particularly at high intrinsic barrier heights. The contact resistance is given as a function of doping and the intrinsic barrier height. An ultimate limit to the contact resistance is also given.

Notizen: Time-dependent intensities of the spectral lines emitted by laser-induced plasmas generated in several gases are presented. The time-resolved and spatially varying intensities of two once-ionized nitrogen lines were used to calculate radial temperature distribution of temperature within the plasma. A modified blast wave theory, in which ionization was included through the Saha equation and the equation of conservation of charge, was used to calculate time-dependent intensities of several spectral lines of C, N, He, and Ar. The temporal profiles of the spectral lines appear to be dependent on the ionization potentials of the species in the plasma.

Notizen: The general approach to cylindrically symmetric force-free magnetic fields first introduced by Lüst and Schlüter [Z. Astrophys. 34, 263 (1954)], is restricted to fields of the form H=[0,Hφ(r),Hz(r)], and subsequently used to determine a set of solutions to the force-free field equations with nonconstant α. The first element of the set is the well-known constant α solution of Lundquist [Ark. Fys. 2, 361 (1951)]. These solutions may have practical applications with respect to high-temperature superconductors.

Notizen: The performance characteristics of ridge waveguide In0.2Ga0.8As/GaAs multiquantum-well lasers are reported. The lasers emit more than 50 mW/facet in the temperature range 20–100 °C in the fundamental transverse mode. The external differential quantum efficiency of 250-μm-long lasers is 0.4 mW/mA/facet. The internal optical loss is 14 cm−1. The optical gain is found to vary linearly with current. The increase in threshold current with increasing temperature in these lasers is primarily due to decreasing carrier lifetime (increased carrier loss) at high temperature. Increasing the heterobarrier height may further improve the high-temperature performance of these devices.

Notizen: Third-harmonic generation (THG) was used to monitor ground-state atomic hydrogen H(1s 2S1/2) in a dc plasma system. A 364.6 nm laser beam focused through H2 or CH4/H2 plasmas induced THG at 121.5 nm, near the atomic hydrogen 2p 2PoJ→1s 2S1/2 Lyman-α transition. Both the intensity and frequency shift of the excitation spectra exhibited dependence on the plasma power. Absolute H atom concentration was estimated by comparing the frequency shift to that obtained in a calibrated microwave discharge flow system. The sensitivity was ∼4×1013 cm−3 (100 ppm). The measured atomic hydrogen densities were substantially less than in other diamond chemical vapor deposition methods and may explain the lower diamond deposition rates obtained with dc plasma systems of this type.

Notizen: A low-loss, high-intensity color modulator based on the dichroic-mirror-induced separation of the light into three color channels, modulated by polymer dispersed liquid-crystal shutters, is described. The maximum contrast obtained with the light-scattering-based shutter is estimated and compared with experimental values. The quality of color reproduction of the modulator is presented in a standard 2D CIE diagram while the luminance is included in the 3D version with an incandescent lamp used as the light source. The white color is reconstructed by the modulator with an efficiency of 50%.

Notizen: Tin-diffused GaAs layer samples, with mean concentrations larger than 1019 cm−3, were obtained by irradiating with a ruby laser GaAs substrates covered with thin tin layers. Mobilities between 100 and 200 cm2/V s were obtained. Both carrier concentration and mobility decrease with increasing energy density. Strong segregation effects were observed in the samples processed with large energy densities.

Notizen: The results of an experimental investigation to improve the performance of a discharge-pumped iodine monofluoride laser are reported. Lasing was observed at 478.7, 484.7, 490.7, and 496.5 nm. Electrical measurements of the discharge characteristics permitted the energy flow in the circuit to be followed and laser efficiencies to be calculated. Parametric studies of gas mixtures were carried out. By optimizing several parameters, single-pulse lasing energies greater than 50 mJ were obtained.

Notizen: The temporal variation of the electron and current densities as well as the discharge voltage in a XeCl excimer laser discharge is studied as function of gas pressure and HCl+Xe concentration. The results show that independent of pressure and gas mix composition the electron drift velocity is a constant of vd=(1.2±0.2)×106 cm s−1. While in discharges containing only helium the current and electron density are independent of pressure P, both quantities vary as P0.6 once Xe and HCl have been added. The results are examined considering the most important atomic reaction rates.

Notizen: The optical and magneto-optical properties of weakly guiding single-mode buried channel waveguides in substituted yttrium iron garnet have been investigated experimentally. The propagation constant difference Δβ of the two coupled TE and TM modes, which is of particular interest for waveguide isolator application, is found to be strongly dependent on waveguide geometry. Measured results can be explained by the specific growth of the top cladding layer, which occurs into different crystallographic directions around the etched waveguide core. This leads to a local variation of the refractive index, which effects the guidance of light, and to a local variation of birefringence, which effects Δβ. The measured variation of Δβ is shown to be primarily caused by growth induced birefringence. It is demonstrated that Δβ can be controlled to small values at selectable waveguide widths as required for waveguide isolator application simply by an appropriate setting of fabrication parameters.

Notizen: Be ions are implanted into Te-doped Ga0.96Al0.04Sb layers grown by liquid-phase epitaxy. Be distribution is analyzed from secondary-ion mass spectrometry profiles and is found to be in good agreement with a computed simulation one. Hall-effect measurements show a complete electrical activity of the Be-implanted ions. Mesa devices are realized on the Ga0.96Al0.04Sb p+/Ga0.96Al0.04Sb n−/GaSb+ system: Be implantation leads to good quality junctions exhibiting homogeneous multiplication.

Notizen: Ionization coefficients of holes and electrons (kp and kn) in Ga0.96Al0.04Sb have been derived from the variations of the photocurrent multiplication as a function of the reverse bias in Be-implanted p+Ga0.96Al0.04Sb/n Ga0.96Al0.04Sb/n+ GaSb mesa diodes. Three wavelength values have been used for the illumination in order to obtain pure hole injection, nearly pure electron injection, and mixed injection. The kp and kn values thus determined for electric fields varying from 18 to 32 V/μm exhibit a high ratio value showing the interest of this material for avalanche photodiodes devices.

Notizen: The nonlinear piezoelectric photoacoustic (PA) effect of silicon wafers in the applied dc electric field is studied experimentally and theoretically. The apparent nonlinear dependencies of PA amplitude and phase on the laser power are observed in experiments with different dc electric fields. The nonlinearity is explained theoretically by considering the nonlinear surface and bulk Auger recombinations. We present a "subsurface approximation'' method to solve the nonlinear transport equation for photogenerated carriers in semi-infinite samples. A good agreement between the experimental results and theoretical predictions is obtained as the incident laser beam with and/or less than medium power.

Notizen: The Laplace equation can be solved in any two- and three-dimensional porous medium by means of a vectorized numerical code. It is applied to several structures such as random media derived from site percolation; close to the percolation threshold, the critical exponents are found to be very close to the ones corresponding to networks; the results are usefully compared to previous variational upper bounds and to the prediction of an approximate space renormalization. Media with double porosity such as catalyst pellets are also addressed. Finally the conductivity of most fractals is shown to follow an Archie's law in the limit of large generation numbers; the exponents of the power laws can be retrieved by various renormalization arguments.

Notizen: A two-dimensional model is developed for material damage caused by melting and vaporization during pulsed laser irradiation. The problem is formulated by using the energy conservation equation (the Stefan condition) at various points on the solid-liquid and liquid-vapor interfaces. The effect of curvature of the solid-liquid and liquid-vapor interfaces are taken into account and the problem is solved numerically by using the Runge–Kutta method. For determining the maximum damage that can occur during laser irradiation, the laser energy is considered to be utilized only to melt and vaporize the material. The effect of various laser parameters, such as the laser power, laser beam diameter, pulse-on time, and the number of pulses per second on the depth and the radius of the crater is presented. Also, the tapering angle of the crater and the formation of recast layer in the crater during laser irradiation are examined in this study. Finally, a linear relationship between the maximum crater depth and the "gross'' laser intensity is derived phenomenologically and verified by using the numerical results of this study.

Notizen: A new Brownian motion simulation technique developed by Torquato and Kim [Appl. Phys. Lett. 55, 1847 (1989)] is applied and further developed to compute "exactly'' the effective conductivity σe of n-phase heterogeneous media having phase conductivities σ1, σ2, ..., σn and volume fractions φ1, φ2, ..., φn. The appropriate first passage time equations are derived for the first time to treat d-dimensional media (d=1, 2, or 3) having arbitrary microgeometries. For purposes of illustration, the simulation procedure is employed to compute the transverse effective conductivity σe of a two-phase composite composed of a random distribution of infinitely long, oriented, hard cylinders of conductivity σ2 in a matrix of conductivity σ1 for virtually all volume fractions and for several values of the conductivity ratio α=σ2/σ1, including perfectly conducting cylinders (α=∞). The method is shown to yield σe accurately with a comparatively fast execution time.

Notizen: A continuum model was used to analyze charged particle transport and potential distribution in low-pressure radio frequency (rf) glow discharges. The method of lines with orthogonal collocation on finite elements for the spatial discretization was found to be an effective numerical technique for solving the model equations. An argonlike (electropositive) discharge was compared to a pure chlorine (electronegative) discharge. The electronegative discharge was found to have much thinner sheaths, much greater potential drop and electric field strength in the bulk plasma, and severe modulation by the applied rf (10 MHz frequency) of the electron temperature, ionization, and excitation rate, even in the bulk. The effect of varying excitation frequency was also examined. The results showed that continuum models can capture the essential features of both kinds of discharges. Integration of these models with neutral species transport and reaction can result in powerful tools for the modeling and design of plasma reactors.

Notizen: Ion bombardment energy and angle distributions have been measured in an argon plasma. The measured ion angle distribution at 10 mTorr shows that 30% of the ions have incident angles greater than 10° from the surface normal. However, ions with large incident angles have much lower energies than those incident perpendicular to the surface. At 500 mTorr a very large fraction of the ions have large incident angles, and the average energies of these ions are relatively independent of incident angle. Monte Carlo simulations of the sheath kinetics predict the trends shown in the experimental data for ion energy and angle distributions. Fine structure in the ion energy distribution was observed below 50 mTorr and is shown to be caused by charge-exchange collisions in the sheath. The average ion energy in a symmetric parallel plate system is linearly related to the voltage applied across the electrodes for measured plasma pressures up to 500 mTorr.

Notizen: The energy density of the electron flux in magnetized 30-Torr He and He:N2:CO2 dc glow discharges was obtained by Monte Carlo simulation of the cathode region. The application of transverse magnetic fields to the cathode region caused an increase in collisions. This effect became increasingly important in the second half of the cathode region, where magnetic trapping of high-energy electrons occurred. Ionization was enhanced in the cathode region as well as the negative glow, with maximal electron multiplication observed for magnetic field strengths of approximately 0.1 T. This suggests both that a lower cathode fall may be possible in a magnetized discharge and that the electric field reversal in the negative glow will be accentuated. Thus, magnetization of the cathode region may improve the electrical efficiency of CO2 lasers.

Notizen: We discuss extreme ultraviolet and x-ray spectral lines emitted by ions of the Cu i and Na i isoelectronic sequences and show that certain intensity ratios of these lines are sensitive to electron density in the range from about 1018 to 1022 cm−3. These lines therefore provide density diagnostics for high-temperature plasmas. We illustrate the density dependence for four ions of the Cu i sequence (Z=60, 71, 82, 92) and seven ions of the Na i sequence (Z=32, 42, 52, 62, 72, 82, 92). We also show that the dependence of the ratios on electron temperature is not strong, and demonstrate that opacity can be kept small by employing the spot spectroscopy technique.

Notizen: We present a calculation of the characteristic curves of the silver/chalcogenide glass inorganic photoresists which is based on a theory of photoinduced diffusion of silver into the chalcogenide glasses. The silver concentration profiles in the glasses are calculated for the cases of semi-infinite as well as finite slabs. We also consider the effect on the silver concentration profiles of the intrinsic absorption of photons, which results when illumination by photons whose energies are above the band-gap energies of the glass is employed. It is found that very high values of the contrast parameter are attainable for prolonged development.

Notizen: The effect of group-V donor impurities on the behavior of indium atoms implanted into silicon single crystals has been studied by Rutherford backscattering spectrometry/channeling spectrometry, and differential Hall-effect/resistivity measurements. Flat arsenic or phosphorus profiles of concentrations between ∼1.4×1020 cm−3 and 3×1020 cm−3 were produced by As+ or P+ implantation followed by rapid thermal annealing (RTA) (20 s, 1100 °C). Subsequently In was implanted to peak concentrations of 1.7×1020 cm−3 or 3.5×1020 cm−3 and annealed by RTA (15 s, 700–1050 °C).Compared to single indium implants, the presence of the preimplanted group-V impurities is found to reduce the redistribution of the implanted indium atoms during RTA and to increase the concentration of indium atoms incorporated on (or close to) lattice sites (up to ∼2×1020 cm−3). The value of the indium substitutional fraction is found to be dependent on anneal temperature and type of donor impurity. A reduction in the free-electron concentration is observed in both the phosphorus and arsenic predoped samples at the same depths as that of the indium atoms.

Notizen: Double crystal x-ray analysis has been carried out on GaAs wafers implanted with 1.15 MeV sulfur ions at a dose of 5×1014/cm2, followed by rapid thermal anneal for 10 s at temperatures between 300 and 1100 °C. A systematic reduction of strain with increased annealing temperature has been observed, as measured from the separation between the peak of the unimplanted substrate and the major peak of the strained region. Calculations of strain distribution based on existing numerical models are correlated with implantation parameters using Pearson's type-VI distribution functions. Strain reduction after a silicon nitride encapsulation process is found to be equivalent to that after a 300 °C, 10 s rapid thermal anneal (RTA). It is also found that strain relaxation by RTA is strongly dependent on both sample size and the annealing geometry.

Notizen: We have investigated the effect of CF4/O2 plasma treatment before ion implantation on the activation efficiency for Si-implanted GaAs by rapid thermal annealing. WSiN-capped, SiO2-capped, and capless rapid thermal annealings at 900 °C for 10 s in an Ar atmosphere were compared. The annealed samples were characterized by Hall-effect, capacitance-voltage, secondary-ion mass spectrometry, and 77-K photoluminescence measurements. The large decrease in activation efficiency by CF4/O2 plasma treatment was observed in SiO2-capped rapid thermal annealing. On the other hand, the decrease in activation efficiency by this plasma treatment was small in both WSiN-capped and capless rapid thermal annealings.〈squeeze1.6p〉

Notizen: Mercury ions were implanted in amorphized potassium titanyl phosphate (KTiOPO4) in the range of 50–400 keV. The implanted depth distribution was measured by Rutherford backscattering of 2.1 MeV He ions. The mean projected ranges and range stragglings obtained are compared with our calculation based on Biersack's angular diffusion model and Monte Carlo simulation. The results show that the experimental mean projected range is in good agreement with the calculated value within the experimental errors but the range straggling is systematically higher than the calculated value. However, significant deviations from the transport of ions in matter (TRIM'89) prediction for mean projected range and range straggling are found.

Notizen: Development of solid-state laser drivers of higher operating fluence is dependent upon the damage resistance of the frequency conversion crystals. We have demonstrated the correlation between the purity of the crystal growth solution and the laser damage threshold of single-crystal potassium dihydrogen phosphate (KDP). Impurities introduced in the growth process can be atomic species, inorganic or organic compounds (dissolved or particulate), or bacteria. We have developed a purification process for KDP that minimizes contamination: initial recrystallization of bulk KDP followed by ozonation to remove oxidizable material, ultrafiltration to remove nonoxidized particles, and UV-light exposure to suppress bacterial growth. We have also developed a crystal growth method that excludes load-bearing surfaces, which are a potential source of particulate contamination, from the growth environment. The method, which incorporates continuous filtration and continuous flow of the growth solution has yielded crystals with damage thresholds of 28.6 J/cm2 at 355 nm and greater than 64 J/cm2 at 1064 nm (10-ns pulse length).

Notizen: A computational method is developed for analyzing interactions between near-surface lateral cracks and residual stresses due to localized plastic zones. The latter are approximated as either prismatic or shear dislocation dipole pairs. Two-dimensional plane-strain conditions are assumed. A complete linear elastic solution is obtained using half-space dislocations for the dipole pairs and for the dislocation distribution function that accounts for the stress-free crack surface boundary condition. The crack-tip stress intensity factors show that mode-II driving forces are important for both prismatic and shear zones. The tendency for crack deflection to the free surface derived from energy release rates is complex. The dipole representation results do not provide a simple rationalization for lateral crack chipping.

Notizen: Crystals formed upon crystallization of amorphous solids are often plate shaped and contain a high density of twins. The objective of the present work was to further the understanding of the microstructure of such crystals. It is suggested that the elastic energy associated with the crystallization-induced dilatational strains can be reduced significantly by the formation of a macroscopically invariant plane, lying parallel to the basal plane of plate-shaped crystals. The invariant plane is obtained by means of a double-twinning mechanism. The model accounts for continuity being maintained across the crystal-amorphous matrix interface in spite of the volume change involved in the crystallization. The situation bears close resemblance to that dealt with by the crystallographic theory of the martensitic transformation.

Notizen: The internal friction of YBa2Cu3O7−δ is measured at 7.8, 40, and 144 kHz as a function of temperature (from 300 to 1100 K) and oxygen partial pressure. At 40 kHz a marked damping peak occurs at ∼800 K which is attributable to oxygen hopping in the basal plane. The oxygen diffusion coefficient calculated from the hopping relaxation time is in good agreement with other estimates yielding an activation enthalpy, hm=1.07±0.05 eV and D0=1.8×10−4 cm2/s. At a critical value of δ∼0.25 the diffusion coefficient begins to rise rapidly similar to the occurrence of a fast-ion transition.

Notizen: The growth kinetics of an amorphous (a-)interlayer in polycrystalline Zr and Hf thin films on (111)Si have been investigated by cross-sectional transmission electron microscopy. The growth of the a-interlayer in group-IVb metals and silicon systems was found to exhibit similar behaviors. The growth was found to follow a linear growth law initially. The growth rate then slows down and deviates from a linear growth law as a critical thickness of the a-interlayer was reached. Crystalline silicide (ZrSi or HfSi) was found to nucleate at the a-interlayer/Si interface in samples after prolonged and/or high-temperature annealing. Silicon atoms were found to be the dominant diffusing species during the formation of amorphous alloys. The activation energy of the linear growth and maximum thickness of the a-interlayer were measured to be 1.4 eV, 17 nm and 1.2 eV, 27 nm in Zr/Si and Hf/Si systems, respectively. The correlations among the differences in atomic size between metal and Si atoms, growth rate and activation energy of the linear growth, critical and maximum a-interlayer thickness, the largest heat of formation energy for crystalline silicides, the calculated free-energy difference in forming amorphous phase, as well as the atomic mobility in Ti/Si, Zr/Si, and Hf/Si systems are discussed.

Notizen: The diffusion of Si in GaAs was investigated using diffusion sources from different tie triangle regions on the Si-Ga-As ternary phase diagram which provide constant chemical potentials for the three components under isothermal conditions. The Si profiles are measured with secondary ion mass spectrometry and differ significantly for the different sources. The Ga-Si-GaAs tie triangle source produces p-type Si doping with concentration independent diffusion coefficients, and a neutral As or Ga vacancy is proposed as the dominant mobile defect for these conditions. Arsenic-rich sources from two tie triangle regions and also a Si-GaAs tie line source produce Si donor diffusion with concentration-dependent diffusion behavior. The concentration-dependent diffusion coefficients of donor Si for As-rich source diffusion are related to the net ionized donor concentration showing three different regimes: intrinsic regime, intermediate regime, and saturation regime. Ga vacancies are proposed to be responsible for donor Si diffusion in GaAs: a V0Ga and/or V−Ga mechanism for the intrinsic regime; and a V−Ga related mechanism for the extrinsic and saturation regimes. The Si-GaAs tie line source resulted in two branch type profiles, intermediate between the As-rich and the Ga-rich source diffusion cases.

Notizen: Microlithographic patterning has been used to elucidate the mechanisms controlling diamond film nucleation and grain growth. The approach is capable of establishing a degree of control over diamond nucleation on the substrate, which can be used to improve film uniformity and enhance fine grained microstructure. The observed microstructures in the patterned films are consistent with an intrinsic growth mechanism based upon defect-initiated renucleation.

Notizen: This paper compares Hall-effect measurements combined with rapid thermal annealing (RTA) and low-temperature photoluminescence (4-K PL) as characterization techniques for the optimization of the growth of pseudomorphic InGaAs channel modulation-doped field-effect-transistor structures. The Hall-effect measurements with the RTA were used to determine the optimized growth temperature. 4-K PL was used to determine the quality of the InGaAs quantum well and the In mole fraction.

Notizen: The first-order Raman spectra of individual 8-μm-diam PAN-derived carbon fibers which had been annealed at five temperatures from 1700 to 2800 °C are measured as a function of incident laser power from 1 to 140 mW. In all fibers studied, the Raman frequencies of the graphite G band at ca. 1580 cm−1 and the disordered-induced D band at ca. 1360 cm−1 shift to lower frequency with increasing laser power. The largest shifts observed before the fiber is physically damaged are about 13 cm−1 at a laser power of 30 mW. The band positions decrease further at higher laser power, up to a maximum of about 20 cm−1 at 40 mW, and at powers above 30 mW the linewidths and the ID/IG intensity ratio change irreversibly as the fiber begins to erode. The irreversible changes extend several hundred μm away from the 2-μm spot illuminated by the laser. The effects are attributed to laser heating. To quantify the degree of laser heating the temperature dependence of the Raman shift (G band) in pure bulk, highly oriented pyrolytic graphite (HOPG) is measured. To study the heating effect in more detail, a Raman imaging experiment is carried out in which a 0.1-mm spot on a fiber is heated by a focused laser while the fiber is illuminated for Raman characterization along a 1.6 mm length by a low power probe laser. The spatially resolved Raman shifts obtained in this way are combined with the HOPG frequency-temperature calibration to obtain the in situ temperature profile of the laser heated fiber. The measured temperature profile along the fiber is in excellent agreement with a simple convective heat-transfer model. Assuming that the temperature dependence of the G-band position is the same in the fibers as in bulk graphite, the present experiments show that a laser power of 30 mW heats an 8 μm fiber to 330 °C and that above 330 °C irreversible changes are produced by erosion of the fiber. Unperturbed room-temperature Raman frequencies for the five groups of fibers are obtained by extrapolating to zero laser power.

Notizen: Electron-beam (EB)-induced pattern etching of AlxGa1−xAs (0≤x≤0.7) is described. An ultra-thin GaAs oxide at the surface of a GaAs/AlGaAs heterostructure wafer is used as a resist film. The GaAs oxide resist can be selectively removed by EB irradiation in a Cl2 ambient, which results in pattern etching of GaAs/AlGaAs. The etch rate of AlGaAs is examined as functions of substrate temperature, AlAs mole fraction, and EB flux. The results indicate that pattern etching is realized in the AlAs mole fraction range of 0≤x≤0.7

Notizen: Buried oxynitride layers were formed in silica glass by N+ only or both N+- and Si+-ion implantation. These samples were characterized by secondary-ion mass spectroscopy, Rutherford backscattering spectrometry, x-ray photoelectron spectroscopy, and scanning electron microscope. The depth profile of implanted N without Si implantation is a trapezoidal shape at a high-dose condition (more than about 3 × 1016 ions/cm2). In the N+ implantation only, a part of implanted N reacts with Si, and forms a SiON layer which is thermally unstable. In both Si+ (100 keV, 1 × 1017 ions/cm2) and N+ (50 keV, 11× 1017 cm2) implantation, the depth profile of N is a Gaussian distribution, and the thermal stability of a SiON layer is dramatically improved. However, bubbles were generated near the projected range of Si and N at a high-dose condition (the total dose was 4 × 1017 ion/cm2 and the dose ratio Si/N=3/4). The generation of bubbles was suppressed by two-step N+ implantation (40 and 60 keV).

Notizen: The adsorption and decomposition of PH3 on Si(111)-(7×7) was investigated in ultrahigh vacuum by means of temperature programmed desorption, low-energy electron diffraction, Auger electron spectroscopy (AES), and electron stimulated desorption (ESD) methods. Phosphine adsorbs on Si(111)-(7×7) at T=120 K with an initial sticking coefficient of S0(approximately-equal-to)1 through a mobile (extrinsic) precursor state. Some PH3 dissociative adsorption at 120 K is observed. Thermal activation of the adsorbed species results in desorption of a molecular PH3 species up to 550 K. Further heating produces H2(g) desorption at T(approximately-equal-to)740 K and P2(g) desorption at T(approximately-equal-to)1010 K, thus indicating that PH3 decomposition has occurred. AES and ESD studies of the adsorbed species reveal that decomposition takes place by the breaking of PH bonds in PHx(a) to form SiH species on the surface for 120 K〈T〈700 K.

Notizen: The high-performance AuCr/SiO2/InSb metal-oxide-semiconductor capacitor was fabricated successfully using photoenhanced chemical vapor deposition. The 1200-A(ring)-thick SiO2 layer was deposited on the InSb substrate at a temperature below 200 °C. Their electrical and structural properties were analyzed by capacitance-voltage and Auger electron spectroscopy, respectively. The capacitance-voltage results show that the optimal growth temperature of SiO2 is 150 °C at which the flat-band voltage of the capacitor is close to ideal and the slow interface state density is less than 5 × 1010 cm−2. For SiO2 deposited at a lower temperature, although the flat-band voltage and interface state are poorer, the subsequent thermal annealing at 150 °C for 12 h improves both quantities to the level as the optimal condition. However, for SiO2 deposited at a higher temperature (190 °C), the flat-band voltage shifts to −4 V and the slow state density increases to 1.1 × 1011 cm−2. It is found from the Auger depth profile that whatever the deposition temperature was a Si-rich region followed by an oxygen-rich region was formed at the SiO2/InSb interface. These observations are shown to be consistent with the electrical characteristics of the capacitor.

Notizen: We describe reverse bias diode leakage and physical analysis (secondary ion mass spectroscopy, Rutherford backscattering and transmission electron microscopy) data from shallow n+/p junctions made by implanting various doses of arsenic into polycrystalline Si and out diffusing at various temperature/time conditions, into the underlying (100) Si substrate. The polycrystalline Si is cobalt disilicided to provide the first level of metallization. The minimum process specifications giving good junction quality (reverse bias diode leakage current density ≤10 nA/cm2 at 10 V) are 7×1015 cm−2 As out diffused at 950 °C, 30 min with a junction depth 1200 A(ring) below the polycrystalline Si. At these conditions, most of the As ((approximately-greater-than)90%) uniformly redistributes in the polycrystalline Si. There is a segregation to the polycrystalline Si/Si interface which is characterized by a porous, 〈50 A(ring) thin silicon oxide interface. The polycrystalline Si/Si interface is also characterized by a laterally non-uniform large grain(∼2000 A(ring)) or epitaxial regrowth at the 950 °C process temperature. These interface conditions may be used to explain the ∼20% high leakage population which is independent of out-diffusion condition at high As dose. Such a high sport population is not observed when P is implanted into polycrystalline Si and out diffused at 900 °C, 30 min to obtain ∼2500-A(ring)-deep junctions below the polycrystalline Si/Si interface.

Notizen: The light-induced decrease of the photoconductivity in deuterated amorphous silicon is a factor of 3 less even though the defect density increase is greater than in hydrogenated material having equivalent as-deposited properties. Consequent changes in the average recombination cross section of the defects is illustrated. Since the differences in the light soaking behavior upon isotopic substitution has been found to disappear in films deposited at low temperatures, the changes are thought to arise from differences in the silicon network occurring during growth.

Notizen: The current-voltage (I-V) behavior of a GaSb(p)/AlSb/InAs/AlSb/GaSb(p) resonant interband tunneling (RIT) heterostructure is analyzed experimentally and theoretically. The structure has been successfully grown on a (100)-oriented GaAs substrate by molecular-beam epitaxy, demonstrating that more exotic lattice-matched substrates (such as InAs or GaSb) are not required for RIT devices. Theoretical simulations of I-V behavior are developed, employing a two-band tight-binding model. Experimental I-V curves show pronounced negative differential resistance, with a peak-to-valley current ratio of 8.3 at 300 K. Good agreement is observed between measured and calculated peak current densities, consistent with light-hole tunneling through the confined InAs conduction-band state.

Notizen: Coupling-loss-arrested misaligned semiconductor laser arrays are designed and fabricated by liquid-phase epitaxy. In these devices, the losses in the normal misaligned arrays is collected by a gain-induced positive waveguide, and then the collected light couples with light in the adjacent waveguides evanescently, therefore avoiding the coupling loss and strengthening the coupling. The constructed co-cavity element should have better characteristics with respect to transverse-mode selection. A clean single-lobe far field as narrow as 3° and a relatively low threshold of 130 mA for a seven-element array are obtained.

Notizen: P/Be and Ar/Be dual implantations were performed into In0.53Ga0.47As. Significantly higher Be dopant activation was obtained for P/Be dual implantation compared to Be implantation. Lower dopant activation was obtained for Ar/Be dual implantation. Be in-diffusion during annealing is reduced for both P/Be and Ar/Be dual implantations.

Notizen: Measured length variations of laser written marks in rare earth–transition metal (RE-TM) optical data storage media were found to be a strong function of test conditions. In particular, the mark length standard deviation was found to be inversely proportional to the difference between the threshold and bias power levels. Differences in mark length variabilities of various RE-TM compositions written with the same bias power were found to be attributable to their differing marking thresholds.

Notizen: Thin films of high-temperature superconducting YBaCuO deposited by dc-sputtering on SrTiO3 substrates are structured by electron beam lithography on a submicron scale. Details of the technology processes involved are presented. The sudden transition of structures below 1 μm into the semiconducting state is discussed.

Notizen: The room-temperature cathodoluminescence of diamondlike films produced by glow discharge is reported. The material was deposited onto dc biased substrates maintained at relatively low temperature (〈100 °C). Two visible peaks around 2.3 eV (green) and 2.7 eV (blue-violet) were identified which are commonly found in natural and synthetic diamond, indicating the presence of crystalline particles in the films. Moreover, x-ray diffraction spectra of the samples before cathodoluminescence studies are identical to the ones reported for natural powder diamond.

Notizen: We analyzed the Raman spectra and x-ray diffraction of two GaAs/Al0.3Ga0.7As superlattices grown by molecular-beam epitaxy on a 4 ° misoriented (001) GaAs substrate and an exactly (001)-oriented GaAs substrate respectively. From the frequency shifts of the longitudinal-optical- (LO) confined phonons in the Raman spectra and the variation in linewidths of both the LO-confined phonons and the satellite peaks in x-ray diffraction, we found that the 4 ° misorientation of the GaAs substrate from (001) toward [110] direction improves the interface quality of the superlattices.

Notizen: The physics of the growth mechanisms, characterization of epitaxial structures and device properties of GaAs and other compound semiconductors on Si are reviewed in this paper. The nontrivial problems associated with the heteroepitaxial growth schemes and methods that are generally applied in the growth of lattice mismatched and polar on nonpolar material systems are described in detail. The properties of devices fabricated in GaAs and other compound semiconductors grown on Si substrates are discussed in comparison with those grown on GaAs substrates. The advantages of GaAs and other compound semiconductors on Si, namely, the low cost, superior mechanical strength, and thermal conductivity, increased wafer area, and the possibility of monolithic integration of electronic and optical devices are also discussed.

Notizen: We present the first experimental results of a uniformly insulated extraction applied-B ion diode. The diode typically attained beam generating efficiencies of 70% at the time of peak beam power. This is a significant improvement over previous extraction applied-B ion diodes, demonstrating the importance of uniform insulation as suggested by theory [J. Appl. Phys. 59, 2685 (1986)]. Furthermore, we found that the beam current is nearly independent of the ion-emitting area down to a minimum area. At the minimum area, the beam current density was enhanced by 150 times the monopolar Child–Langmuir value. This data supports a recently proposed theory [Phys. Rev. Lett. 59, 2295 (1987)] that predicts large enhancements due to the electron diamagnetic effect on the virtual cathode.

Notizen: The HAWK facility has been developed as a tool to investigate the kinetics of electron-beam-pumped lasers at a pressure of 1 atm for pump rates of 0.040–1 kW/cm3 and pump times of 0.05–1 ms. The highly collimated (1 cm FWHM within the lasing medium) relativistic electron beam propagates through fast valves that separate the accelerator vacuum from a 2-m-long vacuum isolation section, a 0.4-m-long gas-filled buffer section, and 0.7 m along a lambda-geometry laser cell. A 6.5 kG magnetic field confines the 1-MeV electron beam laterally and guides it around a 30° bend into the 3.3-cm-diam laser cell. A magnetic mirror at the far end of the laser cell reflects a large portion of the transmitted electron beam, thereby protecting the laser optics from the electron beam and making the axial deposition more uniform. In this paper we describe modeling of the electron beam energy deposition in HAWK using the three-dimensional Monte Carlo electron/photon transport code acceptm and compare our results with the measured energy deposition. These numerical simulations were begun during the design phase to define the operational range of the facility. There is good agreement between the calculated energy deposition in the laser cell and the deposition inferred from measurements with a segmented, totally stopping calorimeter at different axial locations in cases where hydrodynamic effects are unimportant. ln cases where these effects are important, however, the calculations can predict too large an amount of energy deposited in the laser gas. The numerical simulations are being used to infer the radial and axial deposition profiles of the electron beam energy as an aid in determining the electron-beam-pumped laser performance.

Notizen: Detailed measurements of the time-dependent electron density, xenon excited-state densities, and total HCl depletion have been recently made in electron-beam- (e-beam-) pumped XeCl. This paper presents the results of extensive computer modeling of these experiments and detailed comparisons with the measured results. The model used includes updated HCl reaction cross sections and an enlarged xenon excited-state manifold. A reduced Boltzmann equation is used to calculate the high-energy electron-energy distribution function, and the low-energy distribution function assumes a quasi-Maxwellian distribution. These changes are upgrades to prior models. The model accurately predicts the total HCl depletion, the time-dependent electron density for initial HCl concentrations ≥0.16% (4.8 Torr), and the time-dependent xenon excited-state densities for lean ((approximately-equal-to)0.04%) initial HCl concentration cases. The model tends to underpredict the rate of electron-density growth after the electron density begins to run away. Since depletion of HCl is a key factor in understanding the kinetics data, possible mechanisms that can contribute to this process are also discussed.

Notizen: A theoretical investigation of the effect of induction frequency shows that deviations from local thermodynamic equilibrium (LTE) are strongly related to this parameter. Computations are carried out for an argon plasma at atmospheric pressure over the frequency range 3–40 MHz. Higher frequencies result in lower-temperature levels in the torch, and also the difference between the electron and the atom/ion temperatures is increased. This is in agreement with the observations of other investigators. The results of the proposed model also show good agreement with the measured temperature profiles. Similar calculations, which are based on the LTE assumption, are reported and compared with the present non-LTE model.

Notizen: A low-pressure mercury-argon discharge is probed with a new technique of laser spectroscopy to determine the influence of the Hg(63P2) population on discharge emission. The discharge is excited with inductively coupled rf power. Variations in the intensity of emission lines in the discharge were examined as λ=546.1 nm light from a cw laser excited the Hg 6 3P2-to-7 3S1 transition. The spectrum of the discharge viewed in the region of laser irradiation showed increased emission in λ=546.1, 435.8, 404.7, 253.7, and 194.2 nm lines. Other lines in Hg i exhibited a decrease in emission. When the discharge was viewed outside the region of laser irradiation, all lines exhibited an increased emission. Based on these results we conclude that the dominant mechanism for the excitation of higher lying levels of mercury is the two-step electron impact excitation via the 3P2 level. The depopulation of the metastable level is also responsible for the observed increase in the electron temperature when the laser irradiates the discharge. We also conclude that the 3P2 metastable level of mercury does not play a significant role in the excitation of the 2P1/2 level of mercury ion.

Notizen: The control of minority-carrier lifetimes in n-type silicon by means of low-dose, high-energy proton irradiation at room temperature has been investigated. The irradiation-induced defects, their electrical properties, their annealing behavior, and their relevance for carrier lifetimes have been studied by deep level transient spectroscopy (DLTS) and the open-circuit voltage decay technique. Emphasis was placed on the complete electrical characterization of the dominant recombination centers induced by the proton bombardment of silicon. The energy level position within the forbidden band, the capture cross sections, and emission rates for majority and minority carriers, as well as the defect concentration were determined for each center. By means of a special recombination selective DLTS measurement mode (double-double DLTS) it was even possible to localize in the spectra the deep centers which are relevant to minority-carrier lifetimes. The results are quantitatively discussed in the framework of Shockley–Read–Hall recombination statistics with respect to the reduction of carrier lifetimes for high- and low-injection conditions in semiconductor devices. The results have been corroborated qualitatively by the correlation of the minority-carrier lifetimes in proton-irradiated devices with the concentrations of the induced recombination centers which were influenced by different temperature treatments.

Notizen: The volume dependence of the x-ray Debye temperature in aluminum at high pressure up to 6 GPa was measured with high-energy synchrotron radiation at the National Laboratory for High Energy Physics. High-pressure high-temperature diffraction experiments were carried out by use of a multianvil press system with the energy-dispersive diffraction method. The Debye temperature was determined from the integrated intensity ratio at different temperatures. In order to avoid the intrinsic anharmonic effects and the influence of the energy shift for the Bragg diffraction due to the thermal expansion, constant volume was maintained in a sample by controlling the pressure at elevated temperatures. The interpolated values of the Debye temperature and the Grüneisen parameter of the atmospheric volume were θD0 =367 K and γ0 =3.0, respectively. Excellent agreement was found upon comparison with the published results of conventional experiments.

Notizen: A hot wall technique was applied to grow very thin BiI3 single crystals about 100–1000 A(ring) thick by the achievement of low-temperature growth. The so-called van der Waals epitaxy became possible by employing layered materials such as PbI2 and CdI2 as substrate. BiI3 thin films grown with optimized conditions exhibited a sharp band-edge direct exciton absorption line of 30-meV half-width, which is much sharper than that of a single crystal grown by a conventional method reported so far. The transmission electron diffraction measurements were carried out on the films. It was revealed that highly perfect single crystals with a 4-mm diameter size were obtained.

Notizen: A detailed in situ spectroellipsometric analysis of the nucleation and growth of hydrogenated amorphous silicon (a:Si:H) is presented. Photoelectronic quality a-Si:H films are deposited by plasma-enhanced chemical vapor deposition on smooth metal (NiCr alloy) and crystalline silicon (c-Si) substrates. The deposition of a-Si:H is analyzed from the first monolayer up to a final thickness of 1.2 μm. In order to perform an improved analysis, real time ellipsometric trajectories are recorded, using fixed preparation conditions, at various photon energies ranging from 2.2 to 3.6 eV. The advantage of using such a spectroscopic experimental procedure is underlined. New insights into the nucleation and growth mechanisms of a-Si:H are obtained. The nucleation mechanism on metal and c-Si substrates is very accurately described assuming a columnar microstructural development during the early stage of the growth. Then, as a consequence of the incomplete coalescence of the initial nuclei, a surface roughness at the 10–15 A(ring) scale is identified during the further growth of a-Si:H on both substrates. The bulk a-Si:H grows homogeneously beneath the surface roughness. Finally, an increase of the surface roughness is evidenced during the long term growth of a-Si:H. However, the nature of the substrate influenced the film growth. In particular, the film thickness involved in the nucleation-coalescence phase is found lower in the case of c-Si (67±8 A(ring)) as compared to NiCr (118±22 A(ring)). Likewise films deposited on c-Si present a smaller surface roughness even if thick samples are considered (〉1 μm). More generally, the present study illustrates the capability of in situ spectroellipsometry to precisely analyze fundamental processes in thin-film growth, but also to monitor the preparation of complex structures on a few monolayers scale.

Notizen: We investigated subboundary generation in silicon-on-insulator (SOI) structures which have square-shaped SOI regions and surrounding seed stripes. The SOI films were melted using a pseudoline electron beam, which was produced by oscillating a spot beam faster than the thermal response time of the substrate, and recrystallized through laterally seeded epitaxial growth from the seed stripes. It was found that a molten zone with a concave trailing edge was not effective to form a subboundary-free SOI region wider than several tens of micrometers, since the curved liquid-solid (L-S) interface easily generated subboundaries along 〈100〉 directions. It was also found that the temperature fluctuation due to the oscillation of the spot beam was not directly related to subboundary generation, although it produced some crystalline defects in the films. Concerning the size of SOI patterns, a two-dimensional array of small SOI patterns, each of which is surrounded by seed stripes, was found much more effective to increase the total subboundary-free area than that of larger SOI patterns. From these results, an optimum oscillation waveform was synthesized by combining the triangular and sinusoidal waveforms, so that the trailing edge of the molten zone became straight. Then, in order to minimize generation of {111} facets at the L-S interface which is known to be the origin of subboundaries, the pseudoline beam was swept in the 〈110〉 direction as a whole. As a result, a 150-μm-square SOI region or 5×7 array of 50-μm-square SOI regions was recrystallized by a single sweep without generation of subboundaries.

Notizen: The stability of W, Mo, and Ta in contact with single-crystal β-SiC at elevated temperatures has been investigated using Auger sputter profiling. All three metals were found to form a thin-mixed layer of metal carbide and silicide upon metal deposition at room temperature. This layer is thought to be the result of surface defects which weaken the Si—C bonds and allow a low-temperature reaction to occur. Upon heating, the Ta readily reacts with the SiC substrate and forms a mixed layer of Ta carbide and silicide at annealing temperatures as low as 400 °C, however, the W/SiC and Mo/SiC systems are stable and change very little after annealing at 850 and 800 °C, respectively.

Notizen: Photoluminescence and photoluminescence excitation spectra have been performed on GaAs/AlGaAs quantum well structures in the temperature range 4–300 K. Sharp exciton resonances are present up to room temperature and can be ascribed to localized excitons for T≤50–70 K and to free excitons at higher values of T. Nevertheless, a line-shape analysis of the PL spectra clearly shows the presence of band-to-band recombination. A fit based on a simple statistical model reproduces with high accuracy the photoluminescence spectrum line shape and allows to evaluate the relative densities of excitons and free carriers generated by the exciton dissociation. We find that the ratios of the relative densities can be interpreted on the basis of the law of mass action for describing the thermal equilibrium between excitons, electrons, and holes.

Notizen: Superconducting YBa2Cu3O7−δ thin films with surface roughness Ra∼20 A(ring) have been successfully prepared by either rf or dc magnetron sputtering. The substrate temperature was kept at 600–670 °C during deposition and a subsequent in situ plasma oxidation treatment was performed at 480–520 °C. The films deposited on single-crystal SrTiO3 (100) exhibited zero resistance at 91 K and had the critical current density of 3×106 A/cm2. In this paper we describe the influences of fabrication conditions on film superconductivity.

Notizen: A study of grain alignment and its effect on the dc transport critical current in fine-grained bulk Y1Ba2Cu3O7−δ is reported in magnetic fields from 10−4 T to 26 T. Two features distinguish the critical current density Jc of aligned bulk Y1Ba2Cu3O7−δ from unaligned material. First, the effective critical field where the intergranular Jc approaches zero is about four times higher (30 T) for aligned samples with field parallel to the a, b planes, than it is for polycrystalline unaligned samples (7 T). Second, the nearly field independent plateau value of Jc between 10 mT and 1 T is one to two orders of magnitude higher than typical plateau values of Jc in unaligned bulk-sintered Y1Ba2Cu3O7−δ, for field parallel to the a, b planes. A low-field (〈10 mT) weak-link decrease in Jc with magnetic field is still observed, but it is much smaller than for unaligned material. These data clearly demonstrate that alignment alone significantly reduces the weak-link problem in fine-grained polycrystalline samples with low-aspect-ratio (4:1) grains (unlike melt-grown samples where there has been some ambiguity as to the relative importance of alignment versus large grain growth). Furthermore, the results provide strong evidence that there are two parallel components of intergranular current conduction, one consisting of weak-linked material, the other behaving like intrinsic intragranular material that is not weak-linked. A comparison with unaligned Y1Ba2Cu3O7−δ indicates that the volume fraction of such nonweak-linked material is significantly enhanced by grain alignment, but still only 0.01%–0.1% of the grain boundary area. Field-cooled and force-free Jc data are also presented, along with detailed measurements of the shapes of the voltage-current characteristics.

Notizen: Transmission and reflection measurements in the energy range 0.5–1 eV have been performed on thin CoSi2 films grown by molecular-beam epitaxy on Si(111). For film thicknesses above 200 A(ring), the transmission factor decreases exponentially with film thickness, with an optical attenuation length of 180 A(ring) nearly independent of photon energy. Deviations from this law for film thicknesses below 200 A(ring) are explained by reflection effects. These data, supplemented by reflection measurements, can be fitted by theoretical calculations of the transmission and reflection factors, thus leading to the determination of the CoSi2 optical indexes. The energy dependence of the real and imaginary indexes is found to be consistent with the Drude model. Some interesting aspects of this intraband absorption either in ultrathin films or at low temperature are presented.

Notizen: Raman spectroscopy has been used to monitor the changes induced in glassy carbon as a result of irradiation with H, He, C, N, Si, and Xe ions. The Raman spectrum of unirradiated glassy carbon consists of an intrinsic graphite peak at 1590 cm−1 (the a peak) and a disorder-induced D peak at 1350 cm−1. The G peak position and FWHM and the ratio of the D peak intensity to that of the G peak [I(D)/I(G)] are plotted as functions of the calculated damage density induced by the ion beam. The results show that at very low damage levels [ 〈 0.008 displacements per atom (DPA)], considerable disorder is being introduced into the system, and the average crystallite size has been reduced from 35 A(ring) for the unirradiated material to about 25 A(ring). At damage levels of ∼0.21 DPA, the material starts to undergo an ion-beam-induced modification which saturates at about 5 DPA. The Raman spectra for these heavily irradiated samples are very similar to those obtained from amorphous carbons. At yet higher doses there is some evidence suggesting that ion-beam-induced partial graphitization of the irradiated glassy carbon has occurred. The results are discussed in the context of previously reported Raman studies of amorphous carbons and highly oriented pyrolytic graphite.

Notizen: Damage in GaAs induced by Ga+ focused-ion-beam-assisted Cl2 etching is studied by photoluminescence (PL) intensity measurements as functions of ion energy, ion dose, and substrate temperature. By decreasing the ion energy from 10 to 1 keV, the damage depth decrease to 1/5, where damage depth is taken as the thickness at which the PL intensity recovers by wet etching. The damage depth is shallower when the etching yield is larger with the same ion energy. By increasing the ion dose, the normalized PL intensity decreases, but damage depth is nearly constant. Over 1015 ion dose, the normalized PL intensity shows a constant value. By increasing the sample temperature, the damage depth becomes shallower. At 150 °C with ion energy of 1 keV, the damage depth is less than 0.5 μm, which is the detection limit of the PL measurement in GaAs substrate.

Notizen: Chlorine reactive ion etching of silicon requires the initial removal of the native oxide prior to rapid etching of silicon. The threshold energy for sputter removal of the native oxide on silicon was measured from the apparent oxide thickness on the silicon surface as determined by x-ray photoemission spectroscopy of the Si 2p core level. Using model computations, the threshold energy to sputter the modified native oxide was determined to be 72±5 eV. The surface film chemistry during etching is different above and below the sputter threshold energy. Above the threshold, the silicon surface is contaminated with 1–2 monolayers of SiClx (x=1,2,3) and residual SiClOy. Below the threshold, the oxide is not etched and SiClOy forms on the native oxide surface. The film thickness is observed to increase with ion energy to the threshold energy indicating that ion induced chemisorption phenomena control the film thickness.

Notizen: (Ho,Tm,Er):YAG single crystals of high optical quality suitable for 2.09-μm laser operation have been grown by the Czochralski method. Accurate crystal composition was obtained by new analytical technique. This technique enables the determination of the effective distribution coefficients for the yttrium as well as the other rare-earth ions incorporated into the garnet lattice. Lattice parameter and absorption spectra of the grown crystals are reported together, along with results of cw laser performances.