By use of the plane-wave method, we calculate the photonic band structure for electromagnetic waves of E and H polarization propagating in a system consisting of an infinite array of identical, infinitely long, parallel, metal cylinders of circular cross section, embedded in vacuum, whose intersection with a perpendicular plane forms a simple square or triangular lattice. The dielectric function of the metal from which the cylinders are formed has the simple, free-electron form ε(ω)=1-(${\mathrm{\omega }}_{\mathit{p}}^2$/${\mathrm{\omega }}^2$), where ${\mathrm{\omega }}_{\mathit{p}}$ is the plasma frequency of the conduction electrons. For electromagnetic waves of both polarizations, the problem of obtaining the photonic band structure is reduced to the solution of a standard eigenvalue problem despite the frequency dependence of the dielectric constant of the two-dimensional, periodic system. For electromagnetic waves of E polarization the photonic band structure at low filling fractions of the metallic cylinders is a slightly perturbed version of the dispersion curves of electromagnetic waves in vacuum, except for the appearance of a band gap below the lowest frequency band, whose width increases with increasing filling fraction. A band gap between the first and second bands is present in the photonic band structure of the square lattice, but no band gap is found in the band structure of the triangular lattice. In the case of electromagnetic waves of H polarization the photonic band structure at low filling fractions of the metallic cylinders is also, for the most part, a slightly perturbed version of the dispersion curves of electromagnetic waves in vacuum, but possesses additional, nearly dispersionless, bands in the frequency range ω<${\mathrm{\omega }}_{\mathit{p}}$. A possible origin of these flat bands is described.

The temperature dependence of quantum corrections to conductivity, which are due to the weak localization (WL) of electrons and electron-electron interaction (EEI), has been studied for thin Bi films. A change of the total quantum correction has been determined as well as the components related to WL and EEI under the action of an electric field associated with a high-density current flowing through the film. The quantum corrections are found to decrease with an increase in the current during electron overheating. However, the change in the quantum correction related to EEI is caused not only by the heating but also by a change in the electron interaction. This effect is considered together with a similar effect that shows itself as a change in the contribution of quantum corrections to the Bi-film magnetoresistance under the action of an electric field.

The low-field Hall coefficient of a number of dilute α-phase CuAl alloys and CuAl containing 110-ppm Fe has been measured at 4.2 K as a function of the magnetic field. The difference in the values of the Hall coefficient between the two systems has been interpreted as arising from different anisotropies of the relaxation times of the host conduction electrons. By use of a two-group model for the conduction electrons of copper, it has been found that the relaxation time ratios ${\mathrm{\tau }}_{\mathit{B}}$/${\mathrm{\tau }}_{\mathit{N}}$ for the CuAl and CuAl 110-ppm Fe systems amount to 0.63 and 0.82, respectively. The difference in ${\mathrm{\tau }}_{\mathit{B}}$/${\mathrm{\tau }}_{\mathit{N}}$ between these two alloys has been attributed to the scattering of the conduction electrons by the localized magnetic moments of the Fe atoms.

Exciton optical absorption in self-similar aperiodic one-dimensional systems is considered, focusing our attention on Thue-Morse and Fibonacci lattices as canonical examples. The absorption line shape is evaluated by solving the microscopic equations of motion of the Frenkel-exciton problem on the lattice, in which on-site energies take on two values, according to the Thue-Morse or Fibonacci sequences. Results are compared to those obtained in random lattices with the same stoichiometry and size. We find that aperiodic order causes the occurrence of well-defined characteristic features in the absorption spectra, which clearly differ from the case of random systems, indicating a most peculiar exciton dynamics. The origin of all the absorption lines is assigned by considering the self-similar aperiodic lattices as composed of two-center blocks, within the same spirit of the renormalization group ideas.

We propose an energy functional for localized electron systems that corresponds to the constrained-local-density approximation (LDA) but includes some corrections for spin and orbital polarization to take Hund’s first and second rules into account. The discontinuity of the one-electron potential known for an exact density functional can be easily incorporated in LDA in the scope of our formalism. Applications of the method to the electronic structure and configurational stability of d impurities in Rb are presented.

The two-electron-self-consistent-field (SCF2) approach in which the wave function is constructed from spin-unrestricted geminals (UG’s) optimized variationally is developed. It combines restricted-geminal (RG)-SCF2 and unrestricted Hartree-Fock methods, thus accounting for both short-range (intrapair) and long-range electron correlations. On this basis ground-state properties of the extended half-filled Peierls-Hubbard chains are calculated in a wide range of site-energy modulation α and on-site electron-electron repulsion U. Treating a generalized-Peierls transition in slight-to-strong correlation regimes within the same framework, the electronic-Peierls and spin-Peierls transition areas can be conventionally differentiated by bifurcation points in which UG-SCF2 solutions separate from RG solutions. The corresponding phase diagram in the U-α plane is obtained. For the monatomic (α=0) chain, the predicted dimerization amplitude is in a good agreement with the earlier Peierls and spin-Peierls calculations. For the diatomic strongly correlated chain, dimerization sharply rises with α at the bifurcation point ${\mathrm{\alpha }}_0$ and then rapidly drops near α=U/2; a property that can be related to phase transitions in some donor-acceptor molecular crystals.

We investigate the effect of strong electron-electron repulsion on the electron-phonon interaction from a Fermi-liquid point of view. In particular, we show that the strong interaction is responsible for vertex corrections, which are strongly dependent on the ${\mathit{v}}_{\mathit{F}}$q/ω ratio, where ${\mathit{v}}_{\mathit{F}}$ is the Fermi velocity and q and ω are the transferred momentum and frequency, respectively. These corrections generically lead to a strong suppression of the effective coupling between quasiparticles mediated by a single phonon exchange in the ${\mathit{v}}_{\mathit{F}}$q/ω≫1 limit. However, such effect is not present when ${\mathit{v}}_{\mathit{F}}$q/ω≪1. Analyzing the stability criterion for the compressibility, which involves the effective interactions in the dynamical limit, we show that a sizable electron-phonon interaction can push the system towards a phase separation instability. A detailed analysis is then carried out using a slave-boson approach for the infinite-U three-band Hubbard model describing the basic structure of a ${\mathrm{CuO}}_2$ plane in copper oxides. In the presence of a coupling between the local hole density and a dispersionless optical phonon, we explicitly confirm the strong dependence of the hole-phonon coupling on the transferred momentum versus frequency ratio. We also find that the exchange of phonons leads to an unstable phase with negative compressibility already at small values of the bare hole-phonon coupling. Close to the unstable region, we detect Cooper instabilities both in s- and d-wave channels supporting a possible connection between phase separation and superconductivity in strongly correlated systems.

The pressure-dependent optical conductivity spectra of K-TCNQ (tetracyanoquinodimethane) single crystals, a well characterized one-dimensional electron-correlated system, are interpreted on the basis of a Holstein-Hubbard model. The effect of increasing pressure on both electronic and vibronic excitations is accounted for by considering an increasing average value of the transfer integral (t¯) and of the nearest-neighbor electron-electron interaction (V). It is also shown that the pressure induces a decrease of the one-dimensional dimerization of the radical stacks and that the electron-electron correlation energy is always of the same order as the bandwidth.

Local vibrational modes induced by defect complexes involving nitrogen and silicon impurities in different configurations are calculated using Keating potentials in a supercell model. Both the frequency and the vibrational amplitudes of atoms in the complexes are determined for each mode. It is found that defect complexes involving interstitial impurities produce high-frequency localized modes and substitutional impurities produce low-frequency quasilocalized modes. The bond-center Si interstitial is found to have higher lattice energy than the substitutional Si-vacancy pair complex, and hence is less stable. Suggestive atomic models are given based on the present calculations.

We have taken magnetization and calorimetric measurements on ${\mathrm{Zn}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Cr}}_{\mathit{x}}$Te (x=0.003). The heat-capacity measurements show a Schottky peak indicating an energy-level splitting of 3.1 K between the ground and first excited states. Above 1.5 K we observe additional heat capacity, which indicates the presence of additional low-energy vibronic excitations. The magnetization data reveal a small anisotropy (∼7%) with the (111) direction giving the largest value. The magnetization data were fit with a model including a static Jahn-Teller distortion proposed previously in these materials [J. T. Vallin, G. A. Slack, S. Roberts, and A. E. Hughes, Phys. Rev. B 2, 4313 (1970)]. Reasonable agreement was found with the data for a spin-orbit parameter of -59 ${\mathrm{cm}}^{\mathrm{-}1}$ and a Jahn-Teller energy of 320 ${\mathrm{cm}}^{\mathrm{-}1}$.

Electrical resistivities along the ${\mathit{C}}_2$ and ${\mathit{C}}_3$ axes, Hall effect, Shubnikov–de Haas effect, and pulsed-laser-induced transient thermoelectric effect (TTE) have been measured in the temperature range 4.2–300 K for solid solutions of narrow-gap p-type ${\mathrm{Bi}}_{2\mathrm{-}\mathit{x}}$${\mathrm{Sn}}_{\mathit{x}}$${\mathrm{Te}}_3$ (0≤x≤0.035) single crystals. We have found systematic variations of the hole concentration, Hall mobility, Dingle temperature, and cyclotron mass with increasing Sn content. By doping Sn atoms, the hole Fermi energy of the upper valence band (UVB) is increased and a Sn-induced impurity band is formed near the top of the lower valence band (LVB) lying by about 15 meV below the top of the UVB. The observed TTE voltages decay exponentially with time, showing a multiple relaxation process with characteristic relaxation times for thermal diffusions of photogenerated carriers. The TTE data show that there are at least four different relaxation times (${\mathrm{\tau }}_1$–${\mathrm{\tau }}_4$) due to holes, and two, ${\mathrm{\tau }}_5$ and ${\mathrm{\tau }}_6$, due to electrons. Combined with the transport data, as well as the existing band model, we have evaluated the effective masses ${\mathit{m}}_1^{\mathrm{*}}$≃0.033${\mathit{m}}_0$ and ${\mathit{m}}_2^{\mathrm{*}}$≃0.060${\mathit{m}}_0$ for the many-valley hole Fermi surfaces of the UVB, as well as two effective masses ${\mathit{m}}_3^{\mathrm{*}}$≃0.16${\mathit{m}}_0$ and ${\mathit{m}}_4^{\mathrm{*}}$≃0.31${\mathit{m}}_0$ for the LVB along the ${\mathit{C}}_2$ direction, corresponding to ${\mathrm{\tau }}_1$–${\mathrm{\tau }}_4$, respectively. Discussion will be given for the energy spectrum and anisotropy in the UVB and LVB, as well as the lower electron band, in this material system. The thermal conductivities for holes and lattices are also evaluated from the TTE data.

In order to explain experimentally observed magnetophotoluminescence line shapes, theoretical models of interband recombination in the case of k conservation as well as in case of violation of that selection rule have been developed for narrow-gap IV-VI semiconductors. The combined density of states and the individual densities of states of the conduction and the valence bands, respectively, as well as the occupation probabilities of these levels are taken into account. Modifications of the band structure due to many-particle effects (free carriers due to doping of the samples and photoexcited electrons and holes) are also included in the theoretical calculations of the luminescence line shapes. The spectral intensity distribution of the recombination radiation can be explained as spontaneous and stimulated emission of a homogeneous nonequilibrium electron-hole plasma. From least-squares fits to the experimentally observed photoluminescence line shapes, averaged plasma parameters can be deduced. Without an external magnetic field the unrenormalized single-particle band gap, its modifications due to exchange and correlation energy terms, the quasi-Fermi levels, and the effective temperature of the photoexcited carriers can be obtained. In addition, in the case of an external magnetic field one can determine the Dingle temperature and estimate the magnetic-field dependence of the renormalization of the band gap, if the band structure of the investigated material is known. Measurements of the photoluminescence have been performed at a lattice temperature of ${\mathit{T}}_{\mathit{L}}$=1.7 K in a magnetic-field range of B=0–7 T for a variety of IV-VI narrow-gap semiconductors. In the present paper we discuss the results obtained using ${\mathrm{Pb}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Eu}}_{\mathit{x}}$Se epitaxial films with Eu contents x of approximately 4%. The agreement between experimentally observed and calculated line shapes is excellent. However, due to the large number of plasma parameters entering the model, it seems difficult to determine the band structure of a semiconductor only using photoluminescence data.

Carrier-diffusion measurements in silicon using a newly developed Fourier-transient-grating technique is presented. The method uses a laser light pulse projected through a semitransparent grid pattern to excite a sinusoidal excess carrier grating within the sample. The interdiffusion of carriers is monitored by free-carrier absorption of a focused infrared probe beam. The grating is scanned across the probe beam and the spatial Fourier transform is calculated at each sampling time following the excitation pulse. The resulting Fourier spectrum shows a peak for a frequency corresponding to the grating period and the decay of the amplitude of this spatial-frequency component represents a characteristic grating erasure time, which is related to the carrier diffusivity. This Fourier-transient-grating method allows sensitive measurements of the carrier diffusivity over a broad range of injection levels, both in the minority carrier regime as well as for high-injection conditions including the transition between the two regimes. Here, measurement data are presented for silicon samples of various doping concentration and types for excess carrier injections in the range ∼${10}^{12}$–${10}^{17}$ ${\mathrm{cm}}^{\mathrm{-}3}$. At low densities of injected carriers, our measurement data are in agreement with generally accepted low-injection lattice-scattering mobility values showing the transition to the high-injection range according to ambipolar theory. However, at excess carrier concentrations exceeding ${10}^{15}$ ${\mathrm{cm}}^{\mathrm{-}3}$, the diffusion coefficient is clearly reduced with respect to the ambipolar diffusivity (using constant electron and hole diffusivities) due to carrier-carrier scattering effects. This reduction is stronger above an injected carrier density of ${10}^{16}$ ${\mathrm{cm}}^{\mathrm{-}3}$ than that predicted by many-body quantum theory [J. F. Young and H. M. van Driel, Phys. Rev. B 26, 2147 (1982)]. The diffusivity data, converted to mobilities using Einstein’s relation, have also been compared to recent semiempiric drift-mobility models used for semiconductor device simulation.

We study theoretically the possible band mixings induced by a biaxial strain perpendicular to the [001] axis, at the Γ point of zinc-blende semiconductors. Simple group-theory arguments reveal that the usually labeled light-hole state is mixed both with the split-off valence band (intravalence band mixing) and with the upper conduction band (upper-conduction–valence interband mixing). This latter coupling often has been previously neglected in former studies concerning strained compounds and heterostructures. These mixings are related to the interband deformation potentials. The competition of this upper-conduction–valence interband mixing with the intravalence band mixing is analyzed as a function of the spin-orbit energy and of the ionicity at the zone center. Effects of the upper conduction bands comprise at least 30% for usual semiconductors. We specify the conditions and the experimental situations for which it must be taken into account.

We have measured the Hall mobility and Hall electronic concentration as a function of temperature of three low-compensation n-type InP epilayers of similar thickness grown by chemical-beam epitaxy. The Hall mobility of the samples was found to be exceptionally high at 77 K and significantly depressed at 300 K. Above the freeze-out region, strong Hall electronic excitation was consistently observed in the temperature range of ≊35–300 K. These features are in excellent agreement with a model accounting for a broad band of localized deep-donor centers or complexes of unknown origin, centered at 160 meV below the conduction band. In addition, all samples showed an anomalous metal-like behavior at the lowest temperatures, instead of the expected strong localization regime. A broad luminescent band of ≊11-meV linewidth linked to the presence of the band of deep donors was observed below the acceptor bound-exciton transition. Time-resolved photoluminescence and photoluminescence-excitation measurements supported the bound-exciton nature of the luminescent band.

We present a theoretical and experimental analysis of photorefractive two-beam coupling in undoped GaAs as a function of temperature. The theoretical treatment includes the metastable state of the EL2 defect with its optical properties (optical generation and optical recovery) in the photorefractive model. Three major features are predicted by this model: First, a change of the sign of the photorefractive beam coupling gain around 150 K; second, an enhancement of the space-charge field by a factor of 2 compared to the diffusion field; and, finally, the appearance of a strong peak due to an absorption grating around 150 K. All these features are actually observed experimentally with a good correlation between experimental data and theoretical simulation.

The effect of trace carbon impurities on the density of states in amorphous silicon has been studied. The deep-defect densities and mobility-gap electronic structure were characterized with electron-spin resonance, drive-level capacitance profiling, and transient photocurrent measurements. Good quantitative agreement in the defect densities deduced from these three methods has been found. This implies a ratio between charged and neutral dangling bonds of at most 2 to 1. Light-induced changes in the mobility-gap electronic structure were also investigated in these films. A small but significant increase in the density of light-induced defects was observed for samples with carbon impurities at the 1 at. % level. The time to saturation of the light-induced degradation for the carbon containing samples was also significantly increased. We discuss the interpretation of these results in terms of two possible mechanisms: either from the presence of carbon-related precursor sites, or by widening of the band gap with carbon alloying.

Neutral-impurity scattering of electrons and holes at low temperatures has been studied in isotopically engineered Ge single crystals. Use of the neutron transmutation doping technique provides the necessary dopant uniformity and low compensation. We find excellent agreement between the low-temperature experimental mobility and phase-shift calculations for the hydrogen atom scaled to the impurity atoms in semiconductors.

The effects of doping Si clathrate structures with alkali metals have been investigated theoretically. We discuss (a) the electronic structure of Na and other alkali-metal-doped Si clathrates, and (b) the possibility of a Jahn-Teller effect. We also discuss band-gap engineering for Si materials, as well as Mott’s conjecture for the electrical and magnetic properties of ${\mathrm{Na}}_{\mathit{x}}$${\mathrm{Si}}_{136}$ clathrate compounds. We use the self-consistent plane-wave technique to study the electronic structure, and an ab initio tight-binding-like molecular dynamics is performed to determine geometries, energetics, and Jahn-Teller effects.

We have observed Fano resonances in linear absorption experiments on GaAs under a magnetic field. In a bulk semiconductor in a magnetic field, Fano interference is the result of the coupling of higher-order magnetoexcitons and energetically degenerate one-dimensional continuum states. We show that these experimental findings can be described by the model of a two-band semiconductor in the effective-mass approximation. Numerical calculations of the linear magnetoabsorption are presented that demonstrate that the coupling between magnetoexcitons and continuum states is due to Coulomb interaction.