The solid-solution phase of hydrogen in hexagonal close-packed yttrium (α-${\mathrm{YH}}_{\mathit{x}}$) is studied using the pseudopotential method within the local-density-functional approximation with a plane-wave basis. The binding energies associated with different interstitial sites are evaluated for several ordered structures: ${\mathrm{YH}}_{0.5}$, ${\mathrm{YH}}_{0.25}$, and ${\mathrm{YH}}_{0.167}$. It is found that the occupation of the tetrahedral site is always energetically favorable. The hydrogen potential-energy curves around the tetrahedral sites along the c axis and along the path connecting the adjacent octahedral sites are also calculated for ${\mathrm{YH}}_{0.25}$. In particular, the local vibrational mode along the c axis is estimated to be 100 meV, in excellent agreement with that measured in neutron-scattering experiments. Finally, the intriguing pairing phenomenon is investigated by calculating the total energy for various pairing configurations. The possibility of pairing between nearest-neighbor tetrahedral sites is excluded due to the high energy. It is found that the pairing of hydrogen across a metal atom is indeed energetically favorable compared with other kinds of pairs considered and also with isolated tetrahedral hydrogen atoms. The connection with the electronic structure of the system is also examined.

The electronic and structural properties of random Cu-Au alloys have been determined using the (all-electron) fully relativistic Korringa-Kohn-Rostoker coherent potential approximation. The concentration was varied in steps of 10% (5% for the ${\mathit{A}}_3$B and ${\mathit{AB}}_3$ compositions) and total energy calculations were carried out to determine the corresponding equilibrium lattice constants. Just like the experimental data from which they differ at most by 1.5%, the calculated equilibrium lattice constants show a positive deviation from Vegard’s law. The calculated heat of mixing agrees well with the existing experimental data. The equilibrium spectral properties of random Cu-Au alloys are discussed in terms of theoretical photoemission spectra and are compared to experimental data for ${\mathrm{Cu}}_3$Au and pure Au. The structural properties of the compositional series are presented using lattice-constant- and concentration-dependent effective pair (and triplet) interactions up to fourth (tenth) -nearest neighbors and related quantities such as ordering energies and antiphase boundary energies. The ordering processes that govern the ${\mathrm{Cu}}_3$Au and CuAu structure are discussed.

We present two techniques that make feasible the application of the adaptive Riemannian metric technique to all-electron local-density-functional calculations. The first overcomes both the real- (r=0) and Fourier- (G=0) space divergences of the nuclear Coulomb potential by computing the electron-ion energy as the smooth periodic electrostatic potential due to the electrons measured at the positions of the ions. The second overcomes the problem of slow convergence of the extreme metrics which the r=0 Coulomb divergence necessitates by giving an explicit prescription for a suitable metric for arbitrary ionic configurations. All-electron-diamond calculations then serve as a proving ground for these ideas and demonstrate the viability of adaptive Riemannian methods for bypassing the pseudopotential approximation in solid-state calculations.

Density of states calculations for the tight-binding model with diagonal disorder are presented. An instanton approach is used to calculate the tails of the spectrum, including all prefactors. It is shown that a Hartree resummation improves the predictions. Furthermore, an effective-medium approximation is used to calculate the density of states in the band. Combining both approaches, a prediction is obtained for all energies. Large numerical simulations have been performed to check the validity of the approach for the Anderson model with Gaussian and with binary disorder on square and simple cubic lattices. The theory agrees well with the simulations.

The Poisson equation for the electrostatic potential in a solid is solved using three different cellular techniques. The relative merits of these different approaches are discussed for two test charge densities for which an analytic solution to the Poisson equation is known. The first approach uses full-cell multiple-scattering theory and results in the famililar structure constant and multipole moment expansion. This solution is shown to be valid everywhere inside the cell, although for points outside the muffin-tin sphere but inside the cell the sums must be performed in the correct order to yield meaningful results. A modification of the multiple-scattering-theory approach yields a second method, a Green-function cellular method, which only requires the solution of a nearest-neighbor linear system of equations. A third approach, a related variational cellular method, is also derived. The variational cellular approach is shown to be the most accurate and reliable, and to have the best convergence in angular momentum of the three methods. Coulomb energies accurate to within ${10}^{\mathrm{-}6}$ hartree are easily achieved with the variational cellular approach, demonstrating the practicality of the approach in electronic structure calculations.

We report on trapping effects observed in small-signal photoconductivity and steady-state photocarrier-grating (SSPG) experiments. In the former case trapping effects appear since the small-signal photocurrent decays exponentially with time after the excitation is turned off, but with a decay time which depends on the ratio of free to trapped carriers and is determined by the intensity of the cw background illumination. In the case of (steady-state) SSPG measurements, carried out as a function of electric field, an electron drift mobility is obtained which is proportional to the ratio of free to trapped carriers and is therefore again determined by the background illumination. Both experiments show that around 0.4 eV below the electron mobility edge the density of states falls off much more slowly with energy towards midgap than near the mobility edge and may even exhibit a minimum.

We present Raman-scattering spectra of n-doped trans-(CH${)}_{\mathit{x}}$, trans-${(}^{13}$CH${)}_{\mathit{x}}$, and trans-(CD${)}_{\mathit{x}}$ systems. Experimental data have been obtained by using different excitation wavelengths (${\lambda }_{\mathit{L}}$) from the near uv (351.1 nm) to the near ir (1064 nm). For ${\lambda }_{\mathit{L}}$=457.9 nm, doping-induced bands are localized at ≊1270 and 1602 ${\mathrm{cm}}^{\mathrm{-}1}$ in trans-(CH${)}_{\mathit{x}}$,≊1240 and 1550 ${\mathrm{cm}}^{\mathrm{-}1}$ in trans-${(}^{13}$CH${)}_{\mathit{x}}$, and 1564 ${\mathrm{cm}}^{\mathrm{-}1}$ in trans-(CD${)}_{\mathit{x}}$. The peak position and the shape of these bands change slightly when ${\lambda }_{\mathit{L}}$ is moved to the red and near-ir range. Also, additional bands appear in the frequency range 1000–1120 ${\mathrm{cm}}^{\mathrm{-}1}$ in trans-(CH${)}_{\mathit{x}}$ and around 880 ${\mathrm{cm}}^{\mathrm{-}1}$ in trans-(CD${)}_{\mathit{x}}$. The interpretation of these features is given by using the perturbed-Green-function method. In the framework of this model, we have been able to calculate the Raman-active doping-induced vibrational modes by considering the perturbation due to the trapped charges on the lattice dynamics of the conjugated segments in all the samples studied. The calculated modes are found in excellent agreement with experimental data.

We report an ensemble Monte Carlo simulation of impact-ionization-induced impurity breakdown in p-type germanium at liquid-helium temperatures. In our Monte Carlo simulation, the impurities are treated as two-level systems (ground state and first excited state) exchanging particles with the continuum of free holes by capture and thermal generation from the excited level and by impact ionization from both levels. From the simulation we directly obtain the experimentally observed negative differential mobility and demonstrate the cooling effect of the impact-ionization process. For a detailed analysis of the nonequilibrium phase transition between low and high conducting states we extract the characteristic relaxation times of fluctuations exhibiting ‘‘critical slowing down’’ and a strong increase of the current-noise spectral density in the phase-transition regime. Finally, we demonstrate that our Monte Carlo simulations can be used for a quantitative investigation of the complex nonlinear dynamics of current filaments in semiconductors.

The intensity of the dark electron-spin-resonance (ESR) signal, which is commonly observed in nitrogen-rich hydrogenated amorphous silicon nitride (a-${\mathrm{SiN}}_{1.6}$:H), is shown to depend on temperature. This temperature dependence is compared to that observed in amorphous arsenic (a-As), which has a similar thermally generated ESR signal, and to the temperature-independent ESR signal observed in a-Si:H. Comparisons of optically induced absorption spectra for a-${\mathrm{SiN}}_{1.6}$:H and a-As suggest that the defects contributing to the subband-gap absorption may not exhibit the strong electron-lattice interactions that characterize those in a-As but, rather, result most probably from large potential fluctuations.

Two similar defects in silicon, resulting from doping with platinum in an atmosphere containing water vapor, were studied by means of electron paramagnetic resonance. Both spectra have effective electron spin S=1/2, and exhibit platinum- and hydrogen-related hyperfine structure and remarkable behavior under band-gap illumination. One of the centers has been identified with the recently reported Pt-${\mathrm{H}}_2$ complex. Ligand hyperfine interaction and values of electron localization on two silicon neighbors and two protons have been measured and found to support the model of the Pt-${\mathrm{H}}_2$ center. The second spectrum, labeled Si-NL53, is also reported here. This spectrum has trigonal symmetry and displays hyperfine interactions with three equivalent platinum nuclei. Spin-Hamiltonian parameters of the spectrum were determined and the microscopic model of the defect is proposed. It has been concluded that annealing at high temperature in water-vapor atmosphere provides a high concentration of hydrogen in the bulk silicon and that it can be trapped there forming stable complexes with platinum.

A theoretical model for second-order resonant Raman scattering is presented. The effect of Coulomb interaction between electrons and holes is fully taken into account in the framework of the effective-mass approximation. By introducing discrete and continuous excitonic intermediate states in the Raman process, an explicit expression for the Raman scattering efficiency is given for long-range Fröhlich electron-phonon interaction. The model developed can be used to evaluate Raman profiles around the resonant region. A closed-form expression for all matrix elements of the exciton-phonon interaction is obtained once the Coulomb problem for the relative electron-hole motion is separated in spherical coordinates. For the first time, to our knowledge, transitions between states from the excitonic ionization continuum for nonzero phonon wave vectors are exactly included in the calculations. The dependence of the Raman scattering efficiency on electron and hole masses is analyzed. The contribution of the different excitonic transitions to the scattering process is also studied. Finally, the model is compared to available experimental data for GaP, InP, GaAs, and GaSb. The overall agreement with the measured resonance profiles and their absolute scattering efficiencies confirms that excitonic effects are required for a satisfactory interpretation of these phenomena.

We have measured valence-band photoemission spectra and dark conductivity of a-${\mathrm{SiN}}_{\mathit{x}}$:H compounds for compositions between x=0 and x=1.35. The photoemission spectra have been measured with Zr Mζ and Al Kα radiation of 151.4 and 1486.6 eV, respectively. At hν=151.4 eV the spectra resemble directly the total density-of-states (DOS) of the system; the most important change with x is the shift of spectral weight from near the valence-band maximum (VBM) toward the center of the band, indicating the change from a band of Si-Si bonding states to a band of Si-N bonding states. At hν=1486.6 eV the spectra are dominated by the contribution of the Si-3s partial DOS; this contribution is located at the bottom of the band and shifts toward higher binding energies with increasing x. We compare our results at x=0.36 and x=1.35 with those of two recent calculations. Combining results of the dark-conductivity measurements and the photoemission spectra with a previous determination of the optical gaps we make a plot of the VBM, Fermi-level position, and conduction-band minimum (CBM) versus x. It is shown that the sudden opening of the gap at x∼1 is due mainly to the recession of the CBM.

The effect of a static uniform electric field on the photoionization process of an isolated impurity in a semiconductor is studied. The impurity potential is treated with the quantum-defect approach. The limiting hydrogenic and Lucovsky models are also considered. The photoionization process is analyzed with the Fermi’s golden rule in the dipolar approximation. We show that the main effects of the static electric field are similar to the Franz-Keldysh effects well known for the photon-induced interband transition: the appearance of a redshift of the lower absorption edge and of an oscillation pattern superimposed on the zero-field photoionization cross section for higher energies. These oscillations are much stronger when the electric-field polarization of the photons is parallel to the static electric field than when it is perpendicular.

The efficiency of resonant Raman scattering by TO and LO phonons in molecular-beam epitaxially grown HgTe is studied for (001), (110), and (111) surfaces with incident photon energies near the ${\mathit{E}}_1$ and ${\mathit{E}}_1$+${\mathrm{\Delta }}_1$ band gaps. Absolute values of the Raman scattering efficiency are obtained by means of a sample substitution method and by correcting the experimental data with respect to absorption and reflection losses of the incident and the scattered radiation as well as the different angles of collection inside and outside the sample. We compare our results with calculated resonance profiles assuming uncorrelated electron-hole pairs as intermediate states. Interference effects between dipole-allowed (deformation-potential and electro-optic coupling) and dipole-forbidden scattering by LO phonons in HgTe(001) are used to separate extrinsic (electron-impurity interaction) from intrinsic contributions (Frölich intraband coupling) to dipole-forbidden scattering. The optical deformation potentials near the ${\mathit{E}}_1$ and ${\mathit{E}}_1$+${\mathrm{\Delta }}_1$ gaps are obtained from TO scattering in HgTe (111) [HgTe (110)] to be ${\mathit{d}}_{1,0}^5$=-19.9 eV (-21.9 eV) and ${\mathit{d}}_{3,0}^5$=14.7 eV (13.0 eV). The Faust-Henry coefficient of HgTe is determined.

Data on donor magneto-optical 1s→2${\mathit{p}}_{+1\mathrm{}}$ transitions in four center-doped GaAs/${\mathrm{Al}}_{\mathit{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathit{x}}$As quantum wells ranging in width from 10 to 52 nm is presented and analyzed for magnetic fields between 0 and 6 T. It is found that both band nonparabolicity and central-cell corrections must be considered in the narrower wells in order to obtain satisfactory agreement with experiment. A theory of band nonparabolicity is introduced which includes important terms arising from the donor Coulomb potential, terms previously ignored in the analysis of quantum-well donor spectra. In the narrower wells relatively large central-cell corrections are calculated for Si donors from bulk data. A central-cell correction of ∼0.75 ${\mathrm{cm}}^{\mathrm{-}1}$ is estimated for bulk Si donors at zero magnetic field from analysis of published data on Si-doped GaAs epitaxial films.

We have carried out a quantitative low-energy electron-diffraction intensity analysis of the Ge(111) √3 × √3 R30°-Ag structure. It is found that the Ag atoms arrange according to a honeycomb chained trimer model (HCT) similar to the situation of Si(111) √3 × √3 R30°-Ag. This result shows that the Ge-Ge bond is stronger than the Ag-Ag bond on the surface, thus causing the HCT structure to form.

We present the results of first-principles-derived Monte Carlo simulations of H adatom diffusion on the Si(100)-2×1 surface. We developed an analytical Si/H potential which was fit to the results of first-principles electronic-structure calculations of H adatom adsorption and diffusion on embedded silicon clusters designed to model Si(100)-2×1. With this interaction potential, we calculated the rate constants for a H adatom hopping from one site to another, both parallel and perpendicular to the silicon dimer rows, using Monte Carlo simulations to extract exact classical transition-state-theory rate constants. The diffusion constants for H adatoms moving parallel and perpendicular to the surface dimer rows both were found to obey an Arrhenius temperature dependence (over the temperature range T=700–900 K) with preexponential factors and activation energies of ${\mathit{D}}_0^{\mathrm{\parallel }}$=4.0×${10}^{\pm 0.5}$ ${\mathrm{cm}}^2$ ${\mathrm{s}}^{\mathrm{-}1}$, ${\mathit{E}}_{\mathit{a}}^{\mathrm{\parallel }}$=38.1±1.7 kcal/mol, and ${\mathit{D}}_0^{\mathrm{\perp }}$=4.8×${10}^{\mathrm{-}1.0\pm 1.8}$ ${\mathrm{cm}}^2$ ${\mathrm{s}}^{\mathrm{-}1}$, ${\mathit{E}}_{\mathit{a}}^{\mathrm{\perp }}$=62.8±6.4 kcal/mol, respectively. These results confirm our previous suggestion that anisotropic diffusion of H adatoms on the Si(100)-2×1 surface will occur preferentially along the edges of silicon dimer rows. However, these predicted H adatom diffusion rates are orders of magnitude faster (along the dimer rows) or slower (across the dimer rows) than measured values for the rates of ${\mathrm{H}}_2$ desorption from Si(100)-2×1-H. Thus these results suggest that diffusion of hydrogen atoms may not be involved in the rate-limiting step for hydrogen desorption from Si(100).

The kinetics and mechanism of the polymorphic transformation in titanium disilicide thin films from a base-centered orthorhombic structure (C49-${\mathrm{TiSi}}_2$) to a face-centered orthorhombic structure (C54-${\mathrm{TiSi}}_2$) were investigated with the emphasis on the effect of film thickness scaling on the phase transition using in situ resistance measurements, x-ray diffraction, and transmission electron microscopy (TEM). The C49 disilicide films of different thicknesses were prepared by annealing Ti/polycrystalline Si thin film couples of varying Ti thicknesses (250, 550, and 1000 Å). The transformation rate was found to be a strong function of the film thickness and temperature, and can be described by τ∝exp(${\mathit{E}}_{\mathit{a}}$/kT) with an activation energy of ${\mathit{E}}_{\mathit{a}}$=3.73, 4.44, and 5.08 (±0.07) eV for 1000, 550, and 250 Å Ti, respectively. An unusual transition between nucleation sites was observed as a result of the scaling: for samples with 250 Å Ti, the transformation proceeded by nucleating the C54 polymorph at grain edges (three-grain junctions) of the C49-${\mathrm{TiSi}}_2$ films while the C54 nuclei were predominantly formed at the grain boundaries (two-grain junctions) in thicker films, which are in good agreement with the predictions of the nucleation mode from kinetics analyses. It has been suggested that the nucleation of the C54-${\mathrm{TiSi}}_2$ is likely to be the rate-limiting step in the overall transformation. Based upon the microstructural evidences provided by TEM analyses, the significance of these observations due to the reduction in film thickness was discussed by considering energetics of nucleation at different geometrical sites, nucleation site density, and effects of surface and stress. It was demonstrated that the surface contribution and nucleation site density become important for nucleation in thinner films with largely increased surface-to-volume ratio. The nucleation-mode transition is driven by the morphological and microstructural changes associated with thickness scaling.

Observations of the magnetic-field splitting of the circularly polarized photoluminescence excitation spectrum of light-hole excitons in the well region and of the excitons in the magnetic barrier layers of various CdTe/${\mathrm{Cd}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Mn}}_{\mathit{x}}$Te multiple quantum wells are described. These results indicate that the magnetic response of the first few monolayers of the barrier layer is different from that of equivalent bulk material, and can be understood in terms of appreciable short-range potentials situated at the well-barrier interface region.

Within the sequential tunneling approach we calculate the zero-frequency shot noise for electrons tunneling through a mesoscopic system with internal scattering. This scattering is included by coupling the resonant level to another set of ‘‘internal’’ states. Our results show that the internal scattering has no effect on the noise provided that the internal states are not coupled to the collector. On the other hand, when electrons can hop from the internal states to the collector, the noise is affected. In some cases the zero-frequency noise can be suppressed below one-half of full shot noise, 2eI. All limiting cases for the ratio of noise to the current, S/2eI, are presented and discussed.