ALBERT

Notes: Positron annihilation lifetime spectroscopy is a powerful method for characterizing free volumes in a variety of materials. Correlations between positron annihilation rates and the size of free-volume regions in which o-Ps localizes are well described. Unfortunately, difficulties in the analysis of positron annihilation lifetime data have limited the approach to the determination of average lifetimes and average free volumes. Recent advances in the development of numerical integral transform methods now make it possible to extract continuous distributions of positron annihilation rates. The application of these methods to the determination of free-volume distributions is described. The variable transformations required to convert positron annihilation rate probability density functions (PDF) to the corresponding lifetime, radius, and free-volume PDFs are given and the approach is illustrated by application to positron annihilation lifetime data for amorphous polytetrafluoroethylene.

Notes: Photon-gated persistent spectral hole burning was observed for metal-free tetraphenylporphine with halogenated anthracene derivatives in poly(methylmethacrylate) or poly(ethylene) at liquid-helium temperatures. The hole formation yield was markedly dependent on the polymer matrices and the sample composition as well as on the gating wavelength. The irreversible broadening of holes in the systems measured by temperature cycling experiment was smaller than that by proton tautomerization of tetraphenylporphine. We concluded that the hole formation mechanism is donor-acceptor electron transfer on the basis of combined analysis of the matix and acceptor concentration dependence of hole formation yield, the action spectrum of the gating photon and the photoproduct spectrum.

Notes: ArF(B-X) laser emission has been observed under long pulse electron-beam excitation at relatively low pump rate. ArF lasing only occurred with Ne buffered gas mixtures. Optimum performance of 1.93 J/l at an intrinsic efficiency of 1.35% occurred with an Ar/F2 mixture of 1/0.075% using a Ne buffer to 4.0 amagats. Total energy of 290 mJ in a 2-cm2 beam with a 1.0-μs full width at half maximum pulse width was recorded.

Notes: It is shown that the resultant in-plane acceleration sensitivity of contoured quartz resonators symmetrically supported along rectangular edges always vanishes. This result is a consequence of symmetry and applies to any shape for which the support system is symmetric with respect to the center of the mode shape. When the active plate is supported on one side only, a state of flexure is induced in the resonator which degrades the in-plane sensitivity. This emphasizes the importance of a symmetric support system to achieve low acceleration sensitivities. The extensional and induced in-plane flexural biasing deformations are determined by means of our variational approximation procedure using the variational principles in which all conditions appear as natural conditions. The resulting biasing states are employed in an existing perturbation equation along with the mode shapes of the contoured resonators to calculate the nonvanishing acceleration sensitivities when the resonator is supported on one side only.

Notes: A low-resistance electrical contact is provided by a liquid metal in a small gap between the perimeter of a rotating disk (rotor) and a static surrounding surface (stator). The liquid metal extends radially inward on both sides of the rotor to free surfaces with an inert cover gas, and there is a strong axial magnetic field. This paper presents results for the shape of the free surface and for the liquid-metal velocity and pressure adjacent to the free surface. The results depend on the magnetic-field strength, the surface tension, the wetting angle at the free-surface–solid intersections, and the voltage difference between the stator and rotor. The copper stator and rotor are perfect conductors compared to the liquid metal. Two cases are considered, with and without electrically insulating coatings on the sides of the rotor.

Notes: The threading dislocation morphologies and characteristics in In-doped 3-μm-thick GaAs films grown by molecular-beam epitaxy on Si (100) substrates tilted toward the [110] orientation by 2° have been examined mainly using cross-sectional transmission electron microscopy. Most of the observed threading dislocations are 30°- and 60°-type dislocations along the 〈211〉 and 〈110〉 directions on inclined {111} planes. Also, screw-type dislocations running parallel to the [001] growth direction are frequently detected. No appreciable effect of homogeneous In doping throughout the films with a content between 2×1019 and 2×1020 atoms/cm3 on the reduction of threading dislocation generation is observed. However, an In content of 2×1020/cm3 (a misfit of 7.5×10−4) in GaAs films doped in the middle of growth is found to be effective for changing the dislocation direction on the {111} planes.

Notes: In most theoretical and experimental investigations into the shock response of underdense solid media, the influence of the medium's mesostructure on the resulting pressure and degree of compaction has not been taken into account. In typical cases examined, shock pressures are well in excess of 1 GPa and this approach is clearly justified. However, at low pressures, calculations show that the distribution of void sizes can affect the final state achieved upon shocking the medium from a given initial porosity. This paper analyzes the response of porous aluminum to low pressure shocking and demonstrates a dependence of the final shocked state on the distribution of void sizes.

Notes: Fracto-emission is the emission of particles and photons during and after fracture of materials. The observed emissions include electrons, ±ions, neutral species in both ground states and in excited states, and visible photons. Here we examine primarily photon and electron emission during failure of axially loaded stainless-steel fixtures (e.g., rods) embedded in epoxy. The observed signals provide time resolved information on the sequence of events involving interfacial fracture preceding pullout. In addition, we examine the emissions during frictional pullout following debonding. Over a large range of strain rates this pullout exhibits stick-slip behavior. These effects model the processes of fiber/matrix debonding and fiber pullout in a brittle matrix composite.

Notes: A quantitative determination of the contributions of the triply negatively charged Ga vacancies (V3−Ga) and of the doubly positively charged Ga self-interstitials (I2+Ga) to the Ga self-diffusion coefficient in GaAs has been carried out. Under thermal equilibrium and intrinsic conditions, the V3−Ga contribution is characterized by an activation enthalpy of 6 eV for As-rich crystals and of 7.52 eV for Ga-rich crystals, while the I2+Ga contribution is characterized by an activation enthalpy of 4.89 eV for As-rich crystals and of 3.37 eV for Ga-rich crystals.

Notes: The viscosity of Pd77Cu6.5Si16.5 was measured from 995 to 1183 K by capillary flow method. In the experiment, fluxing of the molten specimen by boron oxide was necessary in order to inhibit the formation of crystallites at temperatures near its liquidus. It was found that the viscosity can be described very well by the following Vogel–Fulcher equation: ln η=−3.764+938.543/(T−726), where T is in K. No discontinuity in viscosity was observed at the thermodynamic melting temperature. Recently Tsao and Spaepen [Amorphous Materials: Modeling of Structure and Properties, edited by V. Vitek (The Metallurgical Society of AIME, New York, 1983), p. 323] determined the equilibrium viscosity of the same system near the glass transition temperature. It turned out that the viscosity data from the two regimes cannot be joined together by a single Vogel–Fulcher equation of constant parameters. Finally the present results were compared with those earlier viscosity data of Pd40Ni40P20 and Au77Ge13.6Si9.4 [D. E. Polk and D. Turnbull, Acta Metall. 20, 493 (1972)] to test the validity of the principle of corresponding states for atomic transport properties.

Notes: Multilayers of Fe0.33Zr0.67, prepared by electron beam evaporation, have been characterized by conversion electron Mössbauer spectroscopy, Rutherford backscattering spectroscopy, and x-ray diffraction. Two phases, one amorphous and another crystalline (FeZr3), occur by solid-state reaction. For temperatures of 350 and 500 °C and annealing times ranging from 10 min to 72 h the growth rates of both phases had been obtained. From these results we suggest a model to describe the phase growth kinetics of the amorphous-crystalline Fe0.33Zr0.67 multilayer thin film.

Notes: Epitaxial thin films of ErAs were grown by molecular beam epitaxy on (100) GaAs. The lattice mismatch induced strain and subsequent lattice relaxation were studied using Rutherford backscattering spectrometry and ion channeling. The channeling results showed a strong thickness dependence for the backscattered minimum yield. Channeled angular scans showed the thin films to be fully strained up to 15 nm in thickness. Thicker films were found to relax with full relaxation not being reached until the 100-nm thickness range. The thick films were found to tilt up to 0.26° with respect to the crystal axes of the substrate. The tilt is apparently a result of the strain relaxation mechanism.

Notes: An unconventional analytical/empirical approach was used to determine the intrinsic stress in thin films of hydrogenated amorphous carbon (a-C:H) from measurements at room temperature of total stress on glass. The a-C:H was deposited through the decomposition of methane in an rf plasma over a range of conditions defined by two parameters; namely, the self-bias voltage on the substrate support electrode, Vb, and the pressure of methane in the deposition chamber, P. The intrinsic stress was found to differ from the total stress at room temperature by a thermal stress introduced by the deposition process. Over the range of deposition conditions investigated, 400 ≤ Vb ≤ 1600 V and 1≤P≤16 mTorr, the intrinsic stress was compressive, high in level (varying only slightly between 1–3×1010 dynes/cm2) and proportional to a function of the deposition parameters, Vb−1/4P1/8. Over the same range of conditions, the thermal stress (at ∼20 °C) was tensile and proportional to Vb3/2P1/4. At the highest values of Vb and P investigated, 1600 V and 16 mTorr, respectively, the level of the thermal stress (∼0.8×1010 dynes/cm2) exceeded the level of the total stress (∼0.3×1010 dynes/cm2) by a significant margin. Implications are discussed.

Notes: The changes in electrical properties of n-GaAs as a result of irradiations with fast neutrons have been studied. Epitaxial layers doped with Si at concentrations in the range 1.35×1015 to 1.599×1016 cm−3 were irradiated with reactor neutron fluences up to 1.31×1015 cm−2. When the changes in carrier concentration, Hall mobility, and resistivity were more than 25% of their initial values, nonlinear dependence on neutron fluence was apparent. New theory is proposed which explains the changes in electrical properties in terms of rates of trapping and release of charges. A theoretical relationship is derived for the change in carrier concentration as a function of neutron fluence and doping level. A linear relationship between neutron fluence and Fermi level shift was found to be consistent with the observed changes in carrier concentration.

Notes: We studied nonradiative recombination centers in undoped (AlxGa1−x)0.5In0.5Pgrown by metalorganic vapor phase epitaxy using transient capacitance spectroscopy. We found three deep energy levels, including a mid-gap level. We drew an equation to get a capture cross section for minority carriers, and obtained it using isothermal capacitance transient spectroscopy measurement. The mid-gap level had an electron capture cross section of 2 × 10−10 cm2 and a hole capture cross section of 1 × 10−15 cm2. The time constant of nonradiative recombination through the mid-gap level was found to be comparable to that of radiative recombination. We concluded that the mid-gap level is an effective nonradiative recombination center that reduces photoluminescence intensity.

Notes: Capacitance and conductance measurements are presented for dc-driven Au/ZnS:Mn/p-Si electroluminescence metal-insulator-semiconductor (MIS) devices, where the ZnS:Mn films are deposited by radio frequency sputtering. Stable dc operation is achieved by introducing oxygen into the film during deposition and subsequently annealing. The effect of the post-deposition annealing upon the density of states at the ZnS:Mn/p-Si interface is investigated. As deposited, the devices show unusual MIS C-V characteristics, that indicate a very high interface state density. Annealing at 700 °C, normal C-V characteristics are observed, indicating that the very high density of states is greatly reduced. For these films the conductance technique has been used to measure the density of states at the interface between the ZnS:Mn and p-Si. The statistical model is found to describe most accurately the interface state conductance response. The interface state density consists of a tail of states that varies between 3.7×1013 cm−2 eV−1 at the silicon Fermi level and 1.1×1013 cm−2 eV−1 at the silicon mid-gap. A small peak is superimposed upon this tail at (−0.16±0.01) eV below mid-gap. The tail of states is believed to be intrinsic to the ZnS:Mn/p-Si interface, but evidence suggests that the small peak is due to the presence of oxygen, which is shown by secondary-ion mass spectrometry analysis to accumulate at the interface after annealing at 700 °C. It seems likely that the very high density of interface states in as deposited devices is a consequence of a plasma damage to the silicon surface during growth, creating defects such as silicon dangling bonds. One possible explanation for the decrease in this density is that by annealing at 700 °C, oxygen in the bulk of the film diffuses to the interface, where it mops up these defects by forming compounds such as SiOx. A simpler model of interface recrystallization is also suggested. The doping density in the depletion region of the silicon is calculated as (7.5±0.5)×1014 cm−3, and the interface state capture cross section for holes is found to have mean value of approximately 10−15 cm−2.

Notes: The correlations between structural quality and superconducting behavior in 1000-A(ring)-thick Ba2YCu3O7−δ (BYCO) films grown on LaAlO3(100) from the coevaporation of BaF2, Y, and Cu, followed by an optimized ex situ annealing process are reported. Epitaxial films with smooth, laminar morphology and excellent crystallinity can be grown to have critical current density Jc values nearly identical to single crystals. This finding contrasts with the typical observation that Jc values in thin films of BYCO are very high compared to those of single crystals. This is attributed to a greater density of flux pinning sites due to structural defects within the films. The most crystalline films presented here have penetration length λ∼2000 A(ring) with temperature dependencies described well by the Bardeen–Cooper–Schrieffer (BCS) theory. Material disorder of two types can be controlled by the high-temperature stage Ta of the annealing process. The first type is point defects and dislocations the same size or smaller than the coherence length ξab, which Rutherford backscattering/channeling suggests decrease in number with increasing Ta. The second is crevices, pinholes, and microcracks, which are at least one to two orders of magnitude larger than ξab. At Ta 〈 850 °C, crevices, which create areas of nonuniform thickness, occur due to incomplete epitaxial growth and correlate with the presence of weak links. Hence film resistivity is high, Tc is low, and λ is large. As Ta is increased, the film morphology becomes smoother and all electrical properties improve, except for Jc in nonzero applied magnetic fields, since the improved epitaxy correlates with reduced flux pinning. By Ta= 900 °C, the BYCO films are similar to single crystals in both cation alignment and Jc behavior. Above this annealing temperature, pinholes and microcracks develop and increase in both size and density with increasing Ta. Although these relatively large defects do not act as weak links, they do affect magnetic screening (and hence λ), to result in an anomalous temperature dependence that masks the intrinsic BCS behavior.

Notes: The changes in the real index of refraction and the optical absorption for conduction intersubband transitions in AlxGa1−xAs/GaAs/AlxGa1−xAs quantum wells are examined as a function of the carrier density. Various values for the input optical field and quantum well width are considered in the calculations. The linear contribution due to χ(1) as well as the nonlinear contribution from χ(3) is included. The relationship of the results to device applications such as waveguides and optical modulators is discussed.

Notes: In0.32Ga0.68P epitaxial layers doped with Te and Zn were grown on (100) GaAs0.61P0.39 epitaxial substrates by liquid-phase epitaxy using a supercooling method. The electrical properties of doped layers were determined by C-V measurements at 300 K. Room-temperature carrier concentrations ranging from 9×1016 to 2×1018 cm−3 for n-type and from 3×1016 to 6×1018 cm−3 for p-type dopants are obtained reproducibly. The full width at half maximum value of the 300 K photoluminescent spectrum increases with carrier concentration for Te- and Zn-doped layers. The relative intensity of 300-K photoluminescent peak presents the maximum values at 1×1018 and 6×1017 cm−3 for electron and hole concentrations, respectively. The 100-K photoluminescent spectra show three distinctive peaks and their relative intensities change with hole concentrations. Finally, the relationship between the acceptor ionization energy and hole concentration is described.

Notes: The oxidation of alloys of Si1−xGex differs significantly from that of pure Si in both the thermodynamics of the process and in the kinetics of the oxidation reaction. In this paper these fundamental differences are explored and are used to explain experimental observations of Si1−xGex oxidation that are presented herein and elsewhere in the literature. Alloys of Si1−xGex (with x=5.4, 11.6, and 17 at. %) approximately 200 nm thick were oxidized using the following two processes: (i) dry oxygen at 680 atm at a temperature of 550 °C and (ii) conventional, 1-atm steam at 800 °C. The wet oxidation conditions were chosen to produce an oxide thickness comparable (≈100 nm for xGe=11.6 at. %) to that obtained during high-pressure oxidation at 550 °C. Auger sputter depth profiling, x-ray photoelectron spectroscopy (XPS), and cross-sectional transmission electron microscopy were used to characterize the as-grown oxides. XPS studies reveal that high-pressure oxides formed from alloys of Si1−xGex chemically incorporate Ge directly into the oxide. In contrast, atmospheric oxidation of the same alloy composition produces oxides with no Ge incorporation that is detectable using XPS. It is shown that the selective removal of Si from alloys of Si1−xGex is predicted from Si-Ge-O ternary-phase equilibrium considerations. High-pressure oxidation conditions, however, allow the growth of oxides at much lower temperatures which minimizes Si-Ge interdiffusion and, as a consequence, prevents the selective oxidation of Si from the alloy. These results have direct significance to future device fabrication in Si1−xGex and to the issues of oxidation of multicomponent systems in general.

Notes: The uniform deposition of the oxide nitride oxide (ONO) film on the tunnel-structured polysilicon electrodes was examined by measuring the electrical characteristics of the capacitor and by using transmission electron microscopy. The conventional low pressure chemical vapor deposition method was adopted for the deposition of the nitride film of the ONO. The tunnel length is varied from 1.6 to 15.6 μm. The ONO film deposited over the entire tunnel of 15.6 μm showed good electrical characteristics, and the thickness uniformity as well. In addition to the maximum tunnel length which can accommodate the uniform ONO deposition, the growth mechanism of the ONO film inside of the tunnel is discussed.

Notes: Superconducting YBa2Cu3O7−x films have been deposited in situ onto III-V superlattice substrates. The substrates were GaAs/AlAs superlattices grown by molecular beam epitaxy onto GaAs substrates. For 5000-A(ring) -thick YBCO films grown at 615 °C substrate temperature, we have obtained Tc of 73 K. For thinner films the Tc's are lower, indicating poor interfaces. However, the onset of the superconducting transition is 90 K in all cases. X-ray diffraction, scanning electron microscopy, and transmission electron microscopy (TEM) show these films to be polycrystalline. TEM shows an interaction layer of about 1000 A(ring) at the interface. Low-temperature cathodoluminescence measurements of the substrate show that atomic interdiffusion has intruded at least 5000 A(ring) below the interface.

Notes: The optical absorption of hydrogenated amorphous carbon films (a-C:H) was measured by spectroscopic ellipsometry. The a-C:H films were deposited at different substrate temperatures by rf-plasma of methane. A volume distribution of graphitic cluster size was assumed to reproduce the experimental spectra of the absorption coefficient. The changes in the absorption coefficient and the optical gap, induced by deposition temperature, have been interpreted in terms of changes in the graphitic cluster size of the network. The increase in the deposition temperature produces an increase in the size of the graphitic clusters.

Notes: We have measured the tensor components c1111 and c1221 of third-order nonlinear optical susceptibility of two nematic liquid crystals, MBBA (p-methoxy-benzylidene p-n-butylaniline) and EBBA (p-ethoxy-benzylidene p-n-butylaniline), in the picosecond regime. We observed that the molecule with ethoxy group (EBBA) has a larger third-order nonlinearity than the one with methoxy group (MBBA) and present a model to explain this effect.

Notes: The thermally stimulated discharge current (TSD) behavior of native (dark rested) and irradiated amorphous selenium films (∼100 μm) has been studied in the temperature range 295–385 K. The TSD spectra of these films shows, in addition to the already reported relaxation at 310 K, a new relaxation peak in the temperature range 370–380 K. The origin of this relaxation is attributed to the trapping of charge carriers i.e., electron and holes at deep intrinsic and irradiation induced photostructural defect levels lying at ∼1.6 and 1.7 eV, respectively.

Notes: MeV He ion implantation is used to produce optical waveguides in KTa1−xNbxO3 (x=0.37). The detailed refractive index profiles of the implanted planar waveguides are reconstructed from the measured mode spectra at the wavelength of 632.8 nm. The absorption and the tunneling losses of the lowest mode of the waveguides are determined. Ionic collision is proposed to be the mechanism which causes the observed variation of the refractive index.

Notes: Insulated gate Ga0.47In0.53As/Al0.48In0.52As heterojunction field-effect transistors have been prepared using SiOx and SrF2 insulation and dilute acid and O2 plasma surface cleaning. The leakage current-gate voltage characteristics at different temperatures are analyzed by thermionic and field emission theories to extract the effective barrier heights. The transistor pinch-off characteristics and effective barrier height results are compared with elemental depth profiles obtained from Auger electron spectroscopy. It is concluded that O2 plasma cleaning followed by SiOx gate insulation, is the most suitable technique for fabricating devices and leads to an effective barrier height of 0.3±0.1 V. With refinement this fabrication route should lead to satisfactory room-temperature operation.

Notes: This paper reports a current transport study on the P+/N AlxGa1−xAs homojunction diodes with the AlAs mole fraction, x, ranging from 0 to 0.66. From the 2-kT surface recombination current behavior, the surface state density is nearly independent of x as x is smaller than 0.2. On the other hand, as x is larger than 0.2, the surface state density decreases as x increases. A photocurrent-aided method is proposed to extract the embedded 1-kT current out of the series resistance effect. It is found that the 1-kT current follows Shockley diffusion theory, and the hole diffusion length has a drastic change in the transition region from direct to indirect band. A three-band diffusion model considering the Γ-band-electron-concentration-dependent hole lifetime has been proposed to explain the experimental results successfully.

Notes: Results of combined ac capacitance and conductance measurements on two-terminal metal-oxide-semiconductor-oxide-semiconductor capacitors fabricated in a silicon-on-insulator substrate formed by oxygen implantation are presented for the first time. We demonstrate the efficiency of the technique to investigate the presence of bulk traps presumably caused by metallic contaminants introduced during the high-energy oxygen implantation, as well as to determine the interface trap density at the buried oxide/Si substrate interface.

Notes: The electron mobility and the two-dimensional electron gas (2DEG) concentration in different indium compositions (0.1〈x〈0.6) δ-doped GaAs/InxGa1−xAs/GaAs pseudomorphic structures grown by low-pressure metalorganic chemical vapor deposition are studied. The electron mobilities of a δ-doped GaAs layer are comparable to those of previous reports. Furthermore, the maximum mobility (5500 and 33 000 cm2/V s at 300 and 77 K, respectively) of the proposed pseudomorphic structure appears at x=0.37. Taking into account of strain and quantum effects, the variation trends of calculated 2DEG concentrations are in good agreement with the experimental results.

Notes: General variational expressions for the calculation of the binding energies of the low-lying bound states of an anisotropic donor in a quantum well (QW) are presented. To demonstrate the utilization of these expressions the binding energies of anisotropic donor states in a stepped quantum well are studied theoretically. The variations of the binding energies of the anisotropic donor states as a function of impurity position are presented for different values of the anisotropy factor γ=m*⊥(large-closed-square)m*(parallel), where m*⊥ and m*(parallel) represent the effective electron mass perpendicular to and parallel to the z axis (perpendicular to the interfaces of the QW), respectively. The peak position of the binding energy curves shifts to the deep side of the stepped well. The smaller the γ, the larger the peak shift. The peak shift and the maximum in the binding energies are also dependent on the potential V2 of the step barrier. For the 2p0-like state the corresponding binding energy curve exhibits double peaks, different from the s- and 2p±-like states.

Notes: The critical parameters in the single-target magnetron sputtering of YBa2Cu3O7 have been identified and sufficiently optimized to allow the reproducible deposition of films with Tc's of (approximately-greater-than)90 K and Jc's of (very-much-greater-than) 106 A/cm2 at 77 K. It was found that during film growth the bombardment of the YBa2Cu3O7 by energetic particles must be minimized and also a stronger oxidizing agent than molecular oxygen must be present to obtain films with these properties. Otherwise, films are deposited that, by x-ray diffraction and energy dispersive x-ray spectroscopy analyses, are indistinguishable from the highest-Tc 1:2:3 stoichiometric material but which have critical temperatures of (very-much-less-than)90 K. Films need not have 1:2:3 overall stoichiometry to have optimum superconducting properties. In such cases the excess elements are present as second-phase particles.

Notes: The remanent magnetization of fine particle media of perpendicular anisotropy is known to exhibit a time decay of pseudologarithmic form. The viscous properties are to a large extent determined by the magnetostatic particle interaction and the particle size. A mean-field model of a perfectly aligned ensemble of Stoner–Wohlfarth particles is presented that gives a qualitative description of the hysteretic and time-dependent properties of particulate perpendicular media. The time decay of the remanent magnetization of Alumite media after initial magnetic saturation was measured. The decay was found to be logarithmic within the measuring period, and a coefficient of magnetic viscosity was obtained. Although the reversal mechanism for the particles in Alumite media is known to be incoherent, a good qualitative agreement between the theoretical model and the measurements was found by introducing a reduced effective volume acting as a scaling factor that accounts for the discrepancy in reversal mechanisms.

Notes: FeSi multilayered films (MLFs) with intermediate SiO2 or Ni80Fe20 layers have been found to have superior corrosion resistance, as well as excellent soft magnetic properties. Corrosion resistance for the MLFs depends on the intermediate layer material, intermediate layer thickness, and FeSi layer thickness. In the MLFs with appropriate intermediate layers, oxygen diffusion is restricted to only 20 nm depth from the surface, which is less than the FeSi layer thickness, and no rust is observed after 600 h exposure in 60 °C and 90% relative humidity conditions. In contrast, in FeSi single-layered films, oxygen spreads throughout the entire film, and a reddish brown rust of Fe2O3 was observed. Due to the high saturation magnetization and high corrosion resistance, the FeSi MLFs are forseen to be soft magnetic materials applicable for higher density magnetic recording heads.

Notes: We investigate the influence of interaction between dispersed particles to ac conductivity of suspension of charged particles immersed in electrolyte solution. Using the Rayleigh technique, we establish an identity for charged suspensions, and derive a formula for effective complex conductivity of the systems. A useful scheme for computation of effective constants of composite media is proposed.

Notes: A detailed photothermal electrostatic consideration of the Pd-pyroelectric junction H2 sensor is presented. Experimental evidence is in agreement with the fundamental features of the theory, which supports two possible mechanisms of ac-mode device operation: pyroelectric coefficient dependence on the hydrogenic dipole-induced charge density at the Pd-insulator polyvinylidene fluoride interface, and thermal-wave modulation of the hydrogen-concentration-dependent Pd work function. The dominant operating mechanism is found to depend on the experimental conditions. The concept of image dipole thermostatistical vibration and libration in the pyroelectric matrix is further successfully used to explain the temperature dependence of the photopyroelectric signal in support of the former above-mentioned mechanism.

Notes: Theoretical calculations show that inhomogeneous doping can increase the efficiency of thermoelectric generators and refrigerators.

Notes: Multiple-stripe semiconductor laser arrays are analyzed using a broad-area mode coupling approach. Rather than considering coupling between individual waveguide modes as in the conventional supermode theory, a basis of broad-area modes is chosen. These modes are coupled through a perturbation of refractive index and gain profiles caused by nonuniform carrier injection, thermal effects, and/or built-in weakly guiding or antiguiding profiles. Present theory reveals that earlier simplified analysis involving broad-area mode-coupling may lead to significant errors in modal gains of high-order array modes.

Notes: We have developed a high-purity, intense, lithium ion beam source which operates at 500 kV and 120 A/cm2 with pulse widths of 125 ns full width half maximum. The beams were generated using a lithium chloride anode in planar magnetically insulated geometry. We have found that the combination of vacuum baking of the anode at 250 °C followed by the application of 100 W of pure argon, steady-state, glow discharge cleaning reduced the impurity concentration in the beam to approximately 10% (components other than chlorine or lithium were considered impurities). Although the impurities were low, the concentration of chlorine in the 1+ and 2+ charge states was significant (∼25%). The remaining 65% of the beam consisted of Li+ ions. Without the special cleaning process, over half the beam particles were impurities. It was determined that these impurities entered the beam at the anode surface but came originally from material in the vacuum chamber. After the cleaning process, recontamination was observed to occur in approximately 6 min. This long recontamination time, which was much greater than the expected monolayer formation time, was attributed to the elevated temperature of the anode. We also compared the electrical characteristics of the beams produced by LiCl anodes to those generated by a standard polyethylene proton source. In contrast to the polyethylene anode, the LiCl source exhibited a higher impedance, produced beams of lower ion current efficiency and had longer turn on times.

Notes: Several Si/Si1−xGex multilayer structures, grown by molecular-beam epitaxy (MBE) or synthesized by high-dose Ge implantation into Si substrates, have been implanted with Co atoms at doses of typically 1.0×1015 and 1.0×1016 ions/cm2. These samples are annealed at elevated temperatures (typically 850 °C) to study the diffusion and precipitation of the Co atoms. Depth distributions of Co atoms are measured with secondary-ion mass spectrometry. X-ray diffraction is used to measure the strain evolution of theSi/Si1−xGex MBE multilayers during processing. Channeling Rutherford backscattering spectrometry is used to study the lattice position of the Co atoms after heat treatment. Upon annealing, Co atoms are observed to diffuse out of the Si1−xGex layers. The efficiency of this outdiffusion process strongly depends on the implantation dose. The released Co atoms are gettered by the Si layers adjacent to the Si1−xGex layers and, in case of the MBE samples, at the interface between the substrate and MBE-grown bufferlayer. A physical model to explain these observations is presented.

Notes: Density and resistivity of sputtered tungsten films are directly measured as a function of angle-of-vapor incidence. The film density is found to drop off quickly at deposition angles greater than 50°, and the resistivity rises sharply at angles greater than 70°. These deposition angle effects in films deposited over topography can be predicted using simulation by ballistic deposition (SIMBAD), a two-dimensional computer simulation of thin-film growth using ballistic deposition of hard disks. Simulated film structure and density is in good agreement with real tungsten films deposited over vias. To confirm density predictions, etch rate measurements made on planar films are presented which confirm the dependence of etch rate on film density. The results of the measurement are used to develop an etching algorithm using SIMBAD density predictions. Films deposited over vias are etched to reveal the presence of low-density regions of film covering the via sidewalls, showing excellent agreement with the simulation results.

Notes: The resistivity of an inhomogeneous medium is expanded in Fourier components and the effective resistivity is defined in terms of the rate of entropy production per unit current density. The effective resistivity lies below the volume average by an amount proportional to the mean square fluctuation in resistivity and containing an angular factor, as previously obtained by Herring. The angular factor is obtained in the absence of a preferred direction, and for the case of ellipsoidal symmetry. The effect of the neglect of higher order terms is discussed. The theory is applied to two-phase systems of low contrast and to cases of nonuniform solute concentration. Homogenization increases the resistivity. The time dependence of this increase is discussed in general and for solute clusters.

Notes: An optimized nearest-neighbor tight-binding description of valence bands in strained-layer III-V semiconductors is developed and applied to the calculation of valence-band offsets at strained heterojunctions. It is first shown that a single set of universal interatomic matrix elements can be found which, when appropriately scaled for bond length, simultaneously provide near-optimum tight-binding predictions of valence-band uniaxial deformation potentials, trends in photoelectric thresholds, and valence bandwidths for the common III-V compounds. Application of the optimized tight-binding model to the calculation of valence-band offsets at strained heterojunctions is then discussed, and one simple approach is described which combines a fully strain-dependent version of the optimized tight-binding model with Tersoff's quantum-dipole heterojunction model. Offsets calculated using this combined approach are shown to agree with experimental data better than either strain-dependent natural tight-binding offsets or offsets calculated directly from Tersoff's model. Finally, convenient quadratic expressions for the composition dependence of light-and heavy-hole valence-band offsets, as calculated using the combined approach, are tabulated for several strained and unstrained ternary-on-binary III-V heterojunctions. The balance between accuracy and simplicity offered by our approach should render it useful for exploratory heterojunction device modeling.

Notes: Thin films of boron nitride have been deposited in a plasma-enhanced chemical vapor deposition system, for use as a gate dielectric layer on InP. The InP substrates were pretreated by in situ HCl vapor etching prior to the film deposition. Borane-dimethylamine and ammonia were used as sources for boron and nitrogen. The depositions were carried out at low temperatures (320 °C), the samples being not directly exposed to the plasma which minimizes radiation damage to the substrates. The deposited films were characterized by ellipsometry, x-ray photoelectron spectroscopy, infrared and ultraviolet−visible spectroscopy. Metal-insulator-semiconductor structures were realized to evaluate the bulk dielectric and insulator−semiconductor interface properties. The growth rates are low (20 nm/h). X-ray photoelectron spectroscopy and infrared measurements showed that the layers are essentially boron nitride, with some boron excess (typically N/B(approximately-equal-to)0.8). The dielectric films were found to have an optical index n(approximately-equal-to)1.7, an optical band gap E0(approximately-equal-to) 5.8 eV, a resistivity of about 1013 Ω cm and a dielectric constant of 3.5. Frequency dispersion, between 100 Hz and 1 MHz, of capacitance in accumulation did not exceed 10%. The interface state density was in the 1011–1012 cm−2 eV−1 range.

Notes: The defects present in a sample of YBa2Cu3O7−δ, which has been prepared by the chemical decomposition of YBa2Cu4O8, are examined in detail in an attempt to ascertain with a high degree of probability which of them is primarily responsible for the enhanced flux pinning at low fields. The CuO precipitates are not significant in this regard as they are not numerous enough and produce little if any strain in the surrounding matrix. Imperfections in the twin boundaries are also found to have little if any contribution to pinning in this material. It appears that the partial dislocations bounding stacking faults in the material are the major contributor to the flux pinning as has been suggested previously in the literature.

Notes: Magnetostriction and magnetomechanical coupling coefficient have been measured on Ho0.85Tb0.15Fe2−xCox (x=0, 0.5, and 1.2) alloys. The alloys were prepared by arc melting as well as by arc melting followed by zoning using induction furnace. Saturation magnetostriction (λs) and magnetomechanical coupling coefficient are found to decrease with the increase of cobalt concentration. The λs values of the zoned samples of x=0 and 0.5 are seen to be more than those of arc melted samples whereas, it is otherwise in the case of x=1.2. Magnetomechanical coupling coefficient values for all the samples are seen to increase with zoning. Change in the easy direction of magnetization in the case of x=1.2 is reflected in the magnetostriction of zoned samples.

Notes: The influence of the tensile stress on the flux reversal phenomena in as-quenched state of Fe77.5Si7.5B15 is investigated. A large and stable Barkhausen jump is observed at a field H*, called the switching field. The switching field H* passes through a minimum as tensile stress increases and the jump in magnetization M* at H* grows with stress. It is observed that H* of the sample can be modified by flash annealing under stress. The results on magnetic properties like coercive field, squareness ratio, power loss of as-quenched and stress-annealed samples are presented. The stress annealing induces elastic and plastic components in anisotropy.

Notes: Electronic properties of various single crystals of CdIn2Te4 have been investigated, and more particularly the static dielectric constant εr. This material is usually known to have a very large dielectric constant, values as large as 200–265 are frequently claimed. Two methods were used to study εr. One is based on the direct measurement of the capacitance of a In/CdIn2Te4/In structure and a second one uses a Schottky diode. In this latter case, the εr value is deduced with the help of the Hall measurements. From our measurements the dielectric constant would be much smaller than 200–265, it is estimated to about 11. Finally our results are compared with those obtained by optical measurements.

Notes: Nd3+-doped CaF2 monocrystalline films have been grown by molecular-beam epitaxy using CaF2 and NdF3 evaporation on CaF2 (100) substrates. Auger electron spectrometry and high-energy electron diffraction have been used as in situ techniques to characterize the grown layers. No degradation of crystallinity was observed even for the highest levels of doping considered. The effect of Nd3+ concentration is studied by means of the photoluminescence signal of Nd3+ ions. The results can be compared favorably with those of bulk CaF2:Nd crystals. For a given Nd3+ concentration a lower quantity of Nd3+ aggregate centers is observed. Nd3+ ions can, therefore, be efficiently incorporated in isolated Nd3+-F− centers up to concentrations of 3.7 wt % Nd3+ without emission quenching of the associated line at 1.0457 μm. The results are related to the thermodynamic conditions imposed by the molecular-beam epitaxy technique.

Notes: We present experimental results of photoablated product and kinetic energy (KE) distributions from the ultraviolet laser ablation of a sintered YBa2Cu3O6+x wafer at the laser wavelengths 351, 248, and 193 nm. Data is presented which spans the laser fluence range beginning at the threshold for species ejection (50 mJ/cm2) to nearly that required for the formation of an above surface plasma (800 mJ/cm2). The goal of this experiment was to measure changes in the photophysical process as the incident laser fluence was increased above the threshold value. Our results show, that near the laser threshold fluence, the photoejected products consist of small atomic and oxide species. An unforeseen result for the UV laser wavelengths used, is the lack of CuO+/CuO and free oxygen (O+/O, O+2/O2) in the mass spectrum. In addition, the product distributions are dependent on the laser wavelength. Measured also, at laser threshold fluence, are the nascent photoejected cation kinetic energy distribution. Here our results show that the KE distribution is independent of both the laser fluence and the wavelength. The mean kinetic energy, 〈KE〉, exceeds 3 eV and cannot be explained by a thermal excitation process. With increasing laser fluence (50% above threshold), we detected the photodissociation of the ejected oxide species, and the appearance of the O− anion. Unlike the KE of the cation species, the O− kinetic energy is nearly thermal (〈1 eV). With additional increases in the laser fluence, we measure the photoejection of slow KE neutral species and the simultaneous KE enhancement of the laser ablated ions (KE(approximately-greater-than)30 eV). At the highest laser fluences used in this experiment, it was noted that atomic cluster formations were enhanced. These compounds are presumably formed in the expanding ablated plume. In summary, our results show that, at threshold laser fluences, the photoejection process is via non thermal excitation. With increasing laser fluence the ejected species mass spectrum includes products from the plume photolysis and the plume chemistry.

Notes: Stable compounds in microwave plasmas of various source gases for vapor-phase diamond growth have been measured by using in situ Fourier transform infrared spectroscopy. Each gas mixture of CH4+H2, C2H2+H2, C2H4+H2, CH3OH+H2, C2H5OH+H2, CO+H2, CCl4+H2, and CH4+O2+H2 was introduced into the plasma discharge region as a source gas. The detected carbonaceous compounds were CH4, C2H2, C2H4, and CO, and no other carbonaceous compounds were observed. CH4 and C2H2 were observed in plasmas of all source gases, whereas C2H4 was detected only at higher concentrations of carbonaceous compounds in source gases. CO was produced from source molecules containing oxygen atoms. The source molecules of CH3OH, C2H5OH, and CCl4 disappeared in the microwave plasma. Concentrations of products in the plasmas of CH4+H2, C2H2+H2, and C2H4+H2 were very similar. This result suggests that an equilibrium among CH4, C2H2, C2H4, H2, H, electrons, and various radicals is kept in the hydrogen excess plasma. Moreover, it was found that the addition of O2 to the CH4+H2 plasma reduced the concentrations of CH4 and C2H2 in the plasma.

Notes: The long-term compositional stability of hydrogenated and deuterated amorphous germanium was studied by infrared absorption measurements. The results reveal for substrate temperatures below about 200 °C the growth of a void-rich material which is oxidized rapidly by ambient oxygen and water.

Notes: A predictive formalism is developed that is applicable to the large class of activated physical systems described by a differential equation of the generic form: ∂n(φ,t)/∂t =−n(φ,t)F(t) exp(−(φ−R(t))/A(t)). Practical techniques to predict the behavior of activated physical systems for arbitrary time-dependent environments are both intuitively and mathematically developed. Useful techniques to experimentally determine the initial distribution of activation energies, utilizing arbitrary time-dependent laboratory environments, are presented. A number of fundamental results regarding the correct use and interpretation of common diagnostic techniques, such as Arrhenius plots, are derived. It is shown how the predictive results significantly enhance the ability to quantitatively evaluate the reliability of physical systems whose rate-limiting mechanisms are activated processes obeying the above differential equation. Specific issues regarding integrated circuit reliability are examined as potential applications of this predictive formalism, including time-dependent dielectric breakdown, metal electromigration, nonvolatile memory retention, annealing of radiation-induced trapped charge, and thin ferroelectric film switching properties.

Notes: We present results from a theoretical study of the influence of doping profile variations on the performance of delta-doped AlGaAs/GaAs high electron mobility transistors (HEMTs). An ensemble Monte Carlo simulation coupled with a self-consistent solution of the two-dimensional Poisson equation is used to investigate HEMTs which employ both single and double delta-doped profiles with varying doping configurations. The calculated results reveal that single delta-doped HEMTs designed with identical threshold voltages exhibit improved device behavior when thinner delta-doped layers with more heavily doped concentrations are utilized. For double delta-doped HEMTs with an identical total doping in the AlGaAs layer, improved threshold voltage control is obtained as the spacing between two delta-doped layers increases. However, this increase in spacing also causes a degradation in transconductance, cut-off frequency, and switching time. As gate bias increases, the dependence of device performance (or degradation) on the spacing between doping planes becomes less pronounced due to the upward shift in threshold or "onset'' of parallel conduction in the AlGaAs layer.

Notes: A numerical method is used to evaluate the surface photovoltage (SPV) in semiconductors as a function of surface recombination velocity, photon flux, substrate doping level, and oxide charge. It is found that SPV is reduced by the surface recombination. For a certain surface recombination velocity SPV increases approximately as the logarithm of photon flux. The effect of surface recombination on SPV is dependent on substrate doping level and oxide charge, to sum up, on the energy-band bend. SPV for depletion surface is more remarkably affected by surface recombination than that for strong inversion surface.

Notes: We used x-ray double-crystal diffractometry and MeV 4He channeling spectrometry to study quantitatively the damage produced in Si(100) at room temperature by 230-keV 19F, 230-keV 28Si, 250-keV 40Ar, or 570-keV 131Xe implantation. The measured defect concentration and the perpendicular strain have the same depth profile, and both are depleted near the surface compared to the Frenkel pair concentration calculated from computer simulation. The perpendicular strain is proportional to the defect concentration with a coefficient of B∼0.01 common to all implanted species. The maximum value of the perpendicular strain and of the defect concentration rises nonlinearly with the dose for all species. The damage produced by different implanted species depends on the dose in approximately the same way save for a scaling factor of the dose. In the regime of low damage, the strain and the defect concentration rise linearly with increasing dose. The slope of this rise with dose increases with the square of the Frenkel pairs produced per unit dose of incident ions, as calculated from computer simulations. This fact means that stable defects produced by room-temperature implantation in Si(100) cannot be predicted by a linear cascade model.

Notes: The characteristics of the formation and growth of buried oxide layers, formed by oxygen implantation into silicon at lower energies (50–140-keV 16O+), have been studied using secondary-ion mass spectrometry. Some results have been checked and compared with the results obtained by Rutherford backscattering and cross-sectional transmission electron microscopy. The critical doses, required to form a continuous buried stoichiometric oxide layer during implantation (ΦIc) and after annealing (ΦAc) have been estimated from experimental results. The thicknesses of the silicon overlayer (TASi) and buried silicon dioxide layer (TASiO2) for the annealed wafers have also been estimated. A set of semi-empirical formulas for ΦIc, ΦAc, TASi, and TASiO2 has been introduced. These formulas can be used to quickly calculate the critical doses and the layer thickness values.

Notes: Nonspecular x-ray (λ=0.154 nm) scattering from a 3.9-nm-period tungsten/carbon multilayer structure with scattering vectors near low-angle multilayer specular interference peaks is reported. Diffuse intensity results from kinematical small-angle scattering from in-plane structural inhomogeneities associated with the individual interfaces or layers having characteristic length scales much greater than the multilayer period. Modulations in this diffuse nonspecular intensity when the incident or observation angle is equal to the angle of the first-order multilayer Bragg peak result from standing-wave-enhanced scattering and other dynamical effects. The technique provides a sensitive measure of in-plane structural inhomogeneities of heterointerface systems.

Notes: In the attenuated total reflection method of excitation of the surface polariton, the plane-wave reflectivity has a singular behavior for angles of incidence which nearly lead to the phase-matching condition. A technique for the elimination of the singularity leads to the usual two canonical equations governing the excitation of the surface polariton. The characteristics of the antiresonance in the angular response of the reflectivity are treated, and the effect of a small loss in the metal on these characteristics is also discussed.

Notes: Titanium nitride (TiNx) thin films were deposited onto InP by means of the rapid-thermal-low-pressure-chemical-vapor-deposition (RT-LPMOCVD) technique, using the tetrakis (dimethylamido) titanium (Ti(NMe2)4 or DMATi) complex as the precursor. Depositions were successfully carried out at temperatures below 550 °C, pressure range of 5–20 Torr and duration of 50 to 90 s, to give layer thicknesses up to 200 nm and growth rates in the range of 0.8 to 4.5 nm/s. These films had a stoichiometric structure and contained nitrogen and titanium in a ratio close to unity, but also contained a significant amount of carbon and oxygen. The elements were spread uniformly through the films, the nitrogen was Ti bounded, and the carbon was partially titanium bonded and organic bonded as well. The film resistivity was in the range of 400–800 μΩ cm−2; the stress was always compressive, in the range of − 0.5 × 109 to − 2 × 1010 dyne cm−2, and the film had a good morphology. These layers performed as an ohmic contact while deposited onto p-In0.53Ga0.47As material, (Zn-doped 1.2 × 1018 cm−3), provided an excellent step coverage for high aspect ratio via holes and were deposited selectively onto the InP and based materials when using SiO2 mask. This represents the first report of TiNx films deposited in a commercial RT-LPMOCVD reactor using the DMATi precursor.

Notes: Reflectance anisotropy spectrometry is used to study the electronic properties of sulfide-passivated GaAs. The optical technique allows measurement of the Fermi-level displacement at the surface in situ, while the sample is immersed into a sulfide solution. The kinetics of the process was found to be very sensitive to conductivity type and doping level in GaAs.

Notes: Ion implantation is a possible technique to introduce electrochromically active dopants into fine-grained WO3 thin films. However, to be of use, we must be able to distinguish, and erase, effects due to collisional damage. Argon implants were performed in order to investigate damage effects. Pure damage was found to cause oxygen deficiency, leading to a broadband grey coloration. The formation of hydrogen bronzes (due to water vapor absorption), and of grain-grown material, are reported. Beam heating effects were found to be important due to the fine-grained nature of the material, and need to be suppressed in order for ion implantation to be a useful technique in this area.

Notes: Implantation of MeV erbium ions into micron-thick silica and phosphosilicate glass films and 1200-A(ring)-thick Si3N4 films is studied with the aim of incorporating the rare-earth dopant on an optically active site in the network. Implantation energies and fluences range from 500 keV to 3.5 MeV and 3.8×1015 to 9.0×1016 ions/cm2. After proper thermal annealing, all implanted films show an intense and sharply peaked photoluminescence spectrum centered around λ = 1.54 μm. The fluorescence lifetime ranges from 6 to 15 ms for the silica-based glasses, depending on annealing treatment and Er concentration. Silicon nitride films show lower lifetimes, in the range 〈0.2–7 ms. Annealing characteristics of all materials are interpreted in terms of annealing of ion-induced network defects. These defects are identified using photoluminescence spectroscopy at 4.2 K. Concentration quenching, diffusion and precipitation behavior of Er is also studied.

Notes: Infrared absorption experiments have been performed on hydrogenated and deuterated bulk boron- and aluminum-doped-Si and implanted P, As, and Sb donors in silicon. A first evidence of complex formation in bulk p-type Si is obtained and the spectra confirm the anomalous 3.3-cm−1 deuterium frequency shift with respect to boron isotopes. The ratio of the D-11B and D-10B peak areas is found to be the same as that of the two boron isotopes natural abundance. In donor-implanted silicon, a quantitative analysis of the obtained data has allowed a rough estimate of the passivating rate due to diffusing deuterium. While the frequencies of the various vibrational lines are found to be in agreement with those reported in the literature, the data on the broad line at 1660 cm−1 (H) or 1220 cm−1 (D) seem to suggest an assignment of this peak to a complex in the bulk involving some type of defect due to the implantation process.

Notes: Electroreflectance spectroscopy has been used to study a GaAs/Ga0.77Al0.23As symmetrically coupled double quantum well system as a function of externally applied electric field. Both intra- and inter-well exciton transitions were detected. The energy shifts of the coupled quantum confined levels are in good agreement with a theoretical calculation. Details of the lineshape fit yield important information about the modulation mechanism.

Notes: A mapping technique has been developed that uses differential imaging to characterize semiconductor optoelectronic devices. Although the technique was developed for examination of optical modulators, it has been extended to provide information on the uniformity and bandgap of any material through which one can transmit light. The sharpness of the bandedge, maximum absorption strength, and the film thickness determine the sensitivity of the technique for any particular application. For example, the bandedge sensitivity in 100 periods of InGaAs/GaAs multiple quantum wells was ±20 μeV and can be detected at each location in the field of view in under three minutes. This technique has been applied to modulators designed using strained layer materials for use at a 1-μm operating wavelength.

Notes: We report the results of a systematic investigation of radiative recombination processes of the free-carrier plasma confined in Ga0.47In0.53As/Al0.48In0.52As multiple-quantum-well structures on InP substrate, either undoped or modulation doped. Photoluminescence under low- and high-excitation intensity, luminescence excitation, and optical gain measurements have been used to study the electronic transitions and the optical amplification in the temperature range 10–300 K. Space-resolved luminescence has also been adopted to distinguish between the spontaneous and the stimulated spectral contributions to the observed luminescence collected along different directions with respect to the [001] growth axis of the heterostructures. Optical gain up to 300 K has been observed in the undoped samples under photoexcitation quasiresonant with the confined states in the quantum wells. The theoretical analysis of the optical gain spectra furnishes quantitative data on the electron-hole plasma ground level. In the n-type modulation-doped samples the application of intense optical pumping allows us to observe the transition from the one-component electron plasma to a two-component electron-hole plasma through the rising of a sharp stimulated emission in the optical spectra. Also in this case we observe optical gain up to room temperature.

Notes: Rapid isothermal processing (RIP) based on incoherent sources of light is emerging as a reduced thermal budget (product of processing time and temperature) processing technique. As compared to stand alone annealing unit, the in situ RIP unit is very attractive for the next generation of devices. A number of unwanted physical phenomena can be suppressed or completely eliminated in the in situ RIP case leading to improved quality of materials when compared to their ex situ rapid isothermal annealed and furnace annealed counterparts. We have used in situ rapid isothermal processor for the in situ rapid isothermal chemical cleaning of InP and GaAs substrates and in situ metallization of InP and GaAs Schottky diodes. As compared to ex situ annealing, the negligible oxygen content at the surface and interface of in situ RIP samples result in better current-voltage characteristics, lower and compressive stress values, as well as smooth and continuous morphologies of the ohmic contact. In this paper, we have highlighted the role of in situ RIP in the metallization of InP and GaAs devices.

Notes: We have studied Sb segregation at the moving crystal-amorphous Si(001) interface during preparation of Sb delta-doping layers. X-ray reflectivity measurements reveal a broadening of the delta-doping profile as a result of segregation during amorphous Si crystallization. From ion backscattering measurements, in conjunction with channeling and blocking, it is inferred that bulk diffusion constants are too low to account for the observed segregation behavior. The observed broadening is ascribed to enhanced diffusion at the c-Si/a-Si interface. This interfacial diffusion coefficient is determined to be at least two orders of magnitude larger than the diffusion coefficient in bulk a-Si. Possible causes of this enhancement are discussed.

Notes: The successful synthesis of silicon nitride directly on the surface of a single-crystalline silicon wafer by multipulse (up to n=2500) XeCl* excimer (λ=308 nm) laser irradiation in ambient ammonia or nitrogen atmospheres is reported. The amount and quality of the nitrided layers formed are shown to depend on both the number of subsequent laser pulses applied to the same irradiation site as well as on the nature of the ambient gas. Nuclear reactions, Auger electron spectroscopy, Rutherford backscattering spectroscopy, electrical measurements, and optical and electron microscopy were applied to characterize the synthesized surface layers.

Notes: Determining the composition of strained InGaAsP films through measurements of the lattice spacings and band-gap energies (Eg) requires converting the measured values to ones which would correspond to unstrained material. In strained layers the lattice constants perpendicular and parallel to the growth plane can vary significantly from the relaxed value, and the optically measured Eg is affected by a strain-induced splitting of the valence band and a shifting of the direct gap energy. By combining double-crystal x-ray data with room-temperature photoreflectance results, we determine the InGaAsP composition accurately using an iterative procedure. Film compositions calculated using strain adjusted values of Eg agree with those determined by energy dispersive spectroscopy to within 1–2 at. %, whereas if energy shifts are not considered, the error approaches 10 at. % for strain on the order of 0.4%.

Notes: The diffusion length of minority carriers in n-type floating-zone Si samples is obtained with the electron-beam-induced current technique in planar configuration. The charge collection current data as a function of the beam-junction distance are analyzed on the basis of the "moment method'' developed by Donolato [C. Donolato, Solid-State Electron. 28, 1143 (1985)], which is based on the calculation of the variance of the derivative of the current profile. With respect to other methods reported in literature, this has the advantage that it requires no assumptions on the surface recombination velocity and thus provides a diffusion length value free from its influence. The data are also analyzed with the asymptotic method, which requires conventional assumptions on the surface recombination velocity. The comparison between the results has allowed us to test the capabilities of the above-mentioned method. Particular attention is paid to the injection level and its influence on bulk and surface properties.

Notes: We present experimental results of Al/n- and Al/p-type GaAs Schottky barrier structures grown in situ by molecular-beam epitaxy with thin Si epitaxial interfacial layers. The barrier heights are measured by the I-V, thermal activation, C-V, and photoresponse methods. Barrier heights in the range of 0.3〈φbN〈1.04 eV for n-type GaAs and 0.28 〈 φbP〈1.01 eV for p-type GaAs were obtained for Si layer thicknesses between 6 and 100 A(ring). Annealing studies conducted on the samples indicate that the structures are thermally stable to temperatures up to 450 °C. These results imply that the GaAs surface Fermi level at the Si/GaAs interface is unpinned from its customary near-midgap value. A model which involves the energy-band discontinuities ΔEC and ΔEV between GaAs and Si, the thickness, and the doping of the Si layer is suggested to account for the different barrier-height values obtained.

Notes: Continuous tuning over the entire 8–12 μm wavelength range is demonstrated for the intersubband absorption resonance in n-doped GaAs/AlxGa1−xAs multiple quantum-well structures following partial interdiffusion of the well and barrier layers via rapid thermal annealing. The data indicate that redshifting of the intersubband absorption resonance arises both from interdiffusion-induced modification of the confining potential and from a decrease in the depolarization shift. The latter effect is due in part to a decrease in the free-carrier concentration within the Si-doped quantum wells following rapid thermal annealing. Significant diffusion of the localized Si dopant is also observed over the range of annealing temperatures investigated here. Calculated values of the Al-Ga interdiffusion coefficient, as a function of anneal temperature, indicate that Si diffusion through the heterointerfaces contributes substantially to layer intermixing.

Notes: The photocarrier grating (PCG) technique is used for the first time in the high electric field regime. The results are shown to confirm previously unproven theoretical predictions for this regime. It is demonstrated that, by application of the PCG technique in the high-field regime, accurate values for the ratio of the two carriers mobility-lifetime products can be deduced. This is in contrast with the fact that their sum cannot be derived accurately from the measurement of photoconductivity because one cannot determine accurately the carrier generation rate. Combining the ratio, determined by the high field PCG, with the low-field data yields accurate values for the mobility-lifetime products of both carriers.

Notes: Boron nitride films prepared by reactive ion plating from a boron evaporation source were characterized structurally using three independent electron optical techniques: energy filtered electron diffraction with radial distribution function analysis; electron energy-loss spectroscopy of the near-edge structure of the boron K edge; and spectroscopy of the plasmon region of the energy-loss spectrum. Both specimens had a graded BNx layer between the BN layer and the silicon substrate and in addition one specimen had a titanium bonding layer underneath the BNx layer. The presence of c-BN in both specimens was confirmed by all techniques. The specimen with the titanium bonding layer was examined in cross section and showed essentially pure c-BN on the surface. A model for the formation of c-BN assisted by the compressive stress generated during deposition is proposed.

Notes: The upper critical field of YBa2(Cu1−xFex)3O7−δ with x=0 and 0.05 is determined over the whole temperature range from 4.2 K to Tc. The data behave according to the theory of Werthamer, Helfand, and Hohenberg [N. R. Werthamer, E. Helfand, and P. C. Hohenberg, Phys. Rev. 147, 295 (1966)]. However, at low temperatures the upper critical field is larger than expected from the original expressions of Werthamer and co-workers. Generalized equations, which take into account the anisotropy of the Fermi surface, yield values for the upper critical field, which are in excellent agreement with the experimental data.

Notes: We present a phenomenological expression for the calculation of the optical constants, the refractive index, the extinction coefficient, and the absorption coefficient, of amorphous (a) semiconductors. Our proposed model requires four parameters, namely Eg (optical energy gap), Ec (high-energy cutoff), D (nondirect-transition strength), and Γ (damping). The damping parameter Γ is considered to reflect the distribution of electronic states rather than being the consequence of disorder broadening. Analyses are presented for a-Si, a-Ge, and a-GaAs, and results are in satisfactory agreement with the experimental data over the entire range of photon energy (1.5–6.0 eV).

Notes: We present a differential modulation technique which is effective in extracting the photoreflectance from GaAs/AlxGa1−xAs samples exhibiting excessive room-temperature photoluminescence. Using the technique, we obtain the photoreflectance from the triangular potential-well region of a high-electron-mobility transistor. Surprisingly, the signal from the potential well can be extracted from underneath two layers of heavily doped material, making differential photoreflectance useful in detection of two-dimensional electron gas in high-electron-mobility transistors cladded with doped protective GaAs caps which usually obliterate the PR from the buried interfaces. We isolate absorptionlike photoreflectance peaks at 1.447 and 1.472 eV and show that their energies and the ratio of their amplitudes agree with the 2DEG theory for asymmetric triangular potential well.

Notes: We have studied the dependence of breakdown field in nitrided oxides on nitridation conditions. Nitridation improves the breakdown field of the oxides. In an attempt to explore the cause of the improved breakdown, the breakdown field was related with electron-trap density and refractive index (which increases with nitrogen concentration in the oxide). Our results suggest that the enhanced breakdown field of nitrided oxides is mainly due to incorporation of nitrogen, which results in structural changes in the nitrided oxide.

Notes: The temperature dependence of photoluminescence from the Mg-doped In0.5Ga0.5P layers on (100) GaAs substrates grown by liquid-phase epitaxy has been studied. At low temperature, the spectra show only two major emission peaks involving intrinsic recombination and conduction–band-to-acceptor transition. The intrinsic recombination dominates in the doping concentration range studied (1.0×1017–7.0×1018 cm−3) above 60 K. Below 50 K, these two peaks merged with each other when the doping concentration is higher than 1×1018 cm−3. The temperature dependence of band gap in In0.5Ga0.5P layers determined from the photoluminescence peak energy varies as 1.976 − [7.5 ×10−4 T2/(T + 500)] eV. For the moderately doped concentration (p 〈 1.4 × 1018 cm−3), the Mg acceptor ionization energy obtained from 50-K photoluminescent spectra is in the range from 37 to 40 meV.

Notes: Several physical properties of thin plasma-polymerized films have been measured using a new fiber-optic-based technique. Films of plasma-polymerized tetrafluoroethylene (PPTFE) deposited on the end of an optical fiber form an optical cavity, the reflectivity of which is very sensitive to the film thickness. The fiber is used as an in situ monitor of the deposition rate in the plasma and, after removal from the plasma, the mechanical properties of the film can be measured. With this measurement technique the thermal expansion of the film normal to its surface as well as the swelling of the film when exposed to an array of organic solvents have been determined. A significantly smaller thermal-expansion coefficient and larger degree of swelling are observed relative to bulk PTFE. X-ray photoelectron spectroscopy measurements show that the fluorocarbon chains are highly branched and have a fluorine-to-carbon ratio of 1.45. These results suggest that the plasma-polymerized films are not crystalline and are heavily cross linked.

Notes: The profile of titanium (Ti) lines produced by KrF excimer laser direct writing on lithium niobate (LiNbO3) has been investigated in detail since it is critical in the fabrication of Ti in-diffused LiNbO3 optical waveguides. Lines written at speeds varying from 0.5 to 10 μm/s have typical thicknesses and linewidths varying from 10 to 185 nm and from 2 to 24 μm, respectively. At a low power density E, the maximum thickness t is proportional to the number of photons or the power density. Increasing E allows a sufficient heating, which leads to the diffusion of Ti into LiNbO3, resulting in a sharp decrease of the thickness in the middle of the deposited line.

Notes: InP crystals were etched by reactive ion etching (RIE) with gas mixture of ethane and hydrogen (C2H6/H2), and etching damages were characterized by photoluminescence (PL) measurements of near-edge and defect-related emissions. Near-edge PL emission intensities after RIE were equal to or larger than those before RIE, except for the samples etched for 50 min. The damage introduced by RIE was restricted to the very-near-surface region which can be removed by HF treatment. The peak energy of defect-related 1.1-eV deep emission bands shifted toward the lower-energy side for the crystals with etching damages at the surface. The peak shift is attributable to the increase of defect complexes such as P-vacancy–P-interstitial or P-vacancy–In-vacancy.

Notes: Temperature effects in plasma etching of polymers have been reinvestigated. A stepwise increase in the etch rate is observed when the substrate temperature exceeds the glass transition temperature Tg of the polymer. This phenomenon is correlated with changes in the physical properties of the polymer such as heat capacity. Furthermore, whereas temperature is usually assumed to enhance the etching process, the ion-induced etching component, below Tg, is shown to increase when the polymer temperature is decreased. This behavior can be explained by a simple model involving changes in the adsorption rate of reactive species with temperature.

Notes: Modeling the cutoff frequency (fT) of the advanced bipolar transistors at avalanche breakdown has been developed. The analytical equations developed account for temperature and impact ionization effects on fT and high-current effects at the collector-base junction. Process sensitivity of the collector charging time at the avalanche breakdown regime has been examined. The experimental data reported in the literature are compared in support of the model. Good agreement between the model prediction and measurement is obtained.

Notes: High resolution images of domains written on Co/Pt magneto-optical disks were obtained using scanning magnetic force microscopy. The key advantage of this technique compared to electron microscopy techniques is that regular disks can be used. We studied the influence of the multilayer composition and preparation conditions as well as the effect of the pregrooves and the incorporation of a Pt underlayer, on the structure of the domains on the disk. Marks were written both with laser and magnetic field modulation, and carrier and noise levels were determined. Regularly shaped and subdomain-free marks for both writing schemes were obtained for multilayers with about 4-A(ring)-thick Co layers and Pt layers thinner than 15 A(ring). The magnitude of the coercive field had only a minor influence on the domain structure. On the other hand, the preparation conditions and the structure of the base layers have a large impact on the recording conditions and domain structures.

Notes: The β-diketonate process for chemical vapor deposition Y1Ba2Cu3O6.5+x is extended to deposition on enlarged areas up to 20 cm2. The films are prepared on monocrystalline SrTiO3 and reveal a surface topography comparable to films by laser ablation. The inductively measured transition temperature Tc is up to 93 K and the transition width is smaller than 1 K; Tc is well reproduced from run to run. The critical current density jc at 77 K attains values as high as 6×106 A/cm2 in zero magnetic field; the current carrying capacity in magnetic fields up to 8 T is presented.

Notes: Electron microscopic studies and the electrical properties of Pt contacts on n-type Si annealed by conventional furnace and rapid thermal annealing (RTA) were investigated with scanning electron microscopy, transmission electron microscopy, x-ray analysis, capacitance-voltage, and deep-level transient spectroscopy measurements. An incomplete reaction between Pt and Si is observed after furnace annealing at 550 °C due to the presence of oxygen in the ambient. An almost complete reaction between Pt and Si is achieved after silicidation by RTA at 550 °C. An acceptorlike level located 0.11 eV below the conduction-band edge with a concentration of 2×1013 cm−3 is detected only after furnace annealing at 550 °C. Silicidation by RTA at 550 °C does not introduce traps in the Si band gap.

Notes: Detailed synchrotron radiation photoemission studies of Yb/Hg1−xCdxTe junctions as a function of Yb coverage were performed at room temperature. Photoemission from physisorbed xenon after cooling the sample to 35 K was also used to examine the local overlayer work function and the development of interface morphology. For Yb coverages less than 6 A(ring), the data provide evidence for the lateral growth of islands consisting of Yb-Te reaction products involving divalent Yb, and an associated Hg depletion within an 18-A(ring)-thick near-surface layer. Upon island coalescence at an Yb coverage of 6 A(ring), the formation of a metallic Yb-rich layer at the surface is observed. This layer traps Hg atoms diffusing across the interface through the formation of an Yb-Hg alloy, and is responsible for the nonmonotonic behavior of the Hg interface concentration as a function of Yb coverage.

Notes: A fully quantum-mechanical description of high-field carrier transport in real space representation in semiconductors is developed. The present approach treats a high electric field unperturbatively, while a realistic electron-phonon interaction is included through the perturbation scheme. Direct comparisons of transport properties at very short time in real space with the corresponding semiclassical (Monte Carlo) descriptions are presented. The results explicitly show the quantum effects on the spreading of electron wave packets due to the electron-phonon interaction and help to clarify the limitation of semiclassical descriptions based on the Boltzmann transport equation.

Notes: Heteroepitaxial Si/CoSi2/Si structures have been synthesized by implanting 170-keV Co+ with doses in the range 1–3×1017 Co+ions/cm2 into (100) and (111) Si substrates and subsequent annealing. The microstructure of both the as-implanted and annealed structures is investigated in great detail by transmission electron microscopy, high-resolution electron microscopy, and x-ray diffraction. In the as-implanted samples, the Co is present as CoSi2 precipitates, occurring both in aligned (A-type) and twinned (B-type) orientation. For the highest dose, a continuous layer of stoichiometric CoSi2 is already formed during implantation. It is found that the formation of a connected layer, already during implantation, is crucial for the formation of a buried CoSi2 layer upon subsequent annealing. Particular attention is given to the coordination of the interfacial Co atoms at the Si/CoSi2 (111) interfaces of both types of precipitates. We find that the interfacial Co atoms at the A-type interfaces are fully sevenfold coordinated, whereas at the B-type interfaces they appear to be eightfold coordinated.It is shown that these interface configurations introduce defects in the three-dimensional CoSi2 precipitates and Si matrix. As a result, the nuclei are subjected to compressive strain. It is argued that the combination of interface energy and strain results in a larger stability of small B-type nuclei as compared to A type. When the precipitates grow beyond a critical size of some 20–30 nm, A-type precipitates become more stable, finally resulting in a buried layer of aligned orientation if the layer thickness is larger than about 30 nm. If smaller, it is argued that upon prolonged annealing the layer will have a twinned orientation (B type). Annealed layers of aligned orientation in (100) Si are found to contain interfacial dislocations of edge type with Burgers vectors b=a/4〈111〉 and b=a/2〈100〉. These dislocations are associated with boundaries separating domains having different interface structures. For (111) Si, there exist edge-type dislocations with Burgers vector b=a/2〈110〉. The final state of strain can be attributed to the difference in thermal expansion between CoSi2 and Si. The strain at room temperature corresponds to a fully relaxed layer at about 700 °C. Below this temperature, dislocations become immobile.

Notes: We report a calculation on the current-voltage characteristics of double-barrier resonant tunneling diodes. The results can well-explain experimental observations, in particular, the temperature dependence of both the absolute current amplitude and the resonant voltage positions. Our calculation takes into account the self-consistent Poisson equation, quantum mechanical tunneling, and band-gap offset at the heterojunctions. In addition, the calculation includes a realistic scattering-induced broadening effect, with which a quantitative fit throughout the entire negative differential resistance (NDR) region is demonstrated. Finally, we show that the differential resistance in the NDR region can be tuned, when asymmetrical tunneling barriers are used.

Notes: The H− negative ion thermal energy measured using the two-laser-pulse photodetachment technique is reported to be in the range from 0.1 to 0.7 eV for various conditions of volume ion source operation (pressure−from 2 to 7 mTorr, discharge current−from 1.5 to 20 A). The hydrogen pressure has a significant effect in lowering the negative ion temperature, while the increase of the discharge current leads to a rise in T−. It is found that T− is a fraction of the electron temperature, Te. This fraction is strongly dependent on the gas pressure. T− scales linearly with the electron temperature and exceeds the highest values predicted by the theory of dissociative attachment. The possible mechanisms for H− ion heating are discussed.

Notes: A model is developed to investigate rail electrodynamics in a plasma armature railgun. This model describes the rail motion in response to the transitory Lorentz force and the compressive restoration force from the material which is between the rails and the containment structure. In this model the distance between the rails is found to oscillate with a frequency of β. The magnetic field and the dynamic behavior of the rails induce local electric fields. We investigate the significance of these electric fields in the laboratory frame and in the projectile frame. In the lab frame, rail electrodynamics induces local electric fields which have maximums spaced behind the projectile at locations where βtp is an odd multiple of π, where tp is the time since the projectile has passed an axial location on the rails. When the projectile is accelerating, rail dynamics induce electric fields in the projectile frame which have maximums where βtp is an even multiple of π. As the projectile velocity increases, the locations of the peak voltages move farther behind the projectile. For the CHECMATE railgun, calculations indicate that the rail displacement is on the order of 2 mm, the rail velocity is on the order of 50 m/s, and the voltages induced in the projectile frame are on the order of 20–40 V.

Notes: Defects produced in lightly doped crystalline silicon by 1-MeV electrons and 14-MeV neutrons are studied. Deep level transient spectroscopy (DLTS) and Hall voltage measurements have been used for investigating the energy depth, nature, and density of radiation-induced defects. The extent of damage caused by 14-MeV neutrons is found to be comparatively high as inferred from the continuous distribution of gap states in the DLTS measurement. Identical defects were observed in both cases where the A-center (Ec−0.17 eV), E-center (Ec−0.39 eV), and divacancy (Ec−0.23 eV) were prominent. An additional defect level (Ec−0.86 eV) was observed from DLTS for neutron-irradiated samples.

Notes: We present the first in situ measurement of the evolution of strain fields due to oxygen precipitation in silicon single crystals by means of high-resolution neutron backscattering. The integrated reflecting power R and the lattice parameter variations Δd/d which are directly related to the strain fields have been measured as a function of temperature and annealing time. In the temperature range from 300 to 1185 K, high purity float zone crystals maintain the R-values characteristic for perfect crystals. In contrast, Czochralski-grown crystals which contain on the order of 1018 oxygen atoms cm−3 (20 ppm), show a steep increase in reflectivity starting at 1160 K, and which goes through a maximum at 1350 K. At 1456 K, partial annealing occurs with a time constant of several hours.

Notes: Radiation effects on cubic silicon carbide (3C-SiC) metal-oxide-semiconductor (MOS) structures have been studied with high-frequency capacitance-voltage measurements. It was 〈m1;38p〉found that interface traps are generated at the 3C-SiC/SiO2 interface and oxide-trapped charges are built up in the oxide by 60Co gamma-ray irradiation. The generation of the interface traps and the oxide-trapped charges are affected by bias polarity applied to the gate electrode during the irradiation. Absorbed dose dependencies of their generation are discussed comparing with those of Si MOS structures.