The first-principles transformation of the linear muffin-tin orbitals (LMTO) into a tight-binding (TB) basis provides a reliable and very efficient method of obtaining the so-called Slater-Koster (SK) parameters, which can be used to gain insight into the electronic structure of complex systems, where conventional ab initio methods become too computer intensive. In this work we study the usefulness and the reliability of the SK parameters obtained from the TB-LMTO method, by examining their (i) transferability with changes in structural and chemical environments and (ii) applicability over a wide range of interatomic separations through explicit charge self-consistent LMTO calculations using Ni and Ni-Al systems as examples. We further test these parameters by comparing the densities of states and the effective pair interactions of Ni-Al system with the results of the more accurate Korringa-Kohn-Rostoker–coherent potential approximation method. We also compare our results with the SK parameters obtained empirically.

We discuss the transition from a metal to charge- or spin-insulating phases characterized by the opening of a gap in the charge- or spin-excitation spectra, respectively. These transitions are addressed within the context of two exactly solvable Hubbard and t-J chains with long-range, 1/r hopping. We discuss the specific heat, compressibility, and magnetic susceptibility of these models as a function of temperature, the band filling, and the interaction strength. We then use conformal-field-theory techniques to extract ground-state correlation functions. Finally, by employing the g-ology analysis we show that the charge-insulator transition is accompanied by an infinite discontinuity in the Drude weight of the electrical conductivity. While the magnetic properties of these models reflect the genuine features of strongly correlated electron systems, the charge-transport properties, especially near the Mott-Hubbard transition, display a nongeneric behavior.

First-principles electronic structure and positron-state calculations for transition-metal carbides and nitrides are performed. Perfect NaCl structures as well as structures with metal or carbon/nitrogen vacancies are considered. The positron affinities and lifetimes are determined. The trends are discussed and the results are compared with recent positron lifetime measurements for group-IV and -V refractory metal carbides. The present analysis suggests, contradictory to an earlier interpretation, that positrons are trapped and annihilated at both carbon and metal vacancies. The concentration of metal vacancies detected by positron annihilation methods is probably very low, below the sensitivity limit of other experimental methods.

Based on the Sterne-Inkson model for the self-energy correction to the single-particle energy in the local-density approximation (LDA), we have implemented an approximate energy-dependent and k-dependent GW correction scheme to the orthogonalized linear combination of atomic orbital-based local-density calculation for insulators. In contrast to the approach of Jenkins, Srivastava, and Inkson, we evaluate the on-site exchange integrals using the LDA Bloch functions throughout the Brillouin zone. By using a k-weighted band gap ${\mathit{E}}_{\mathit{g}}$ and a plasmon frequency ${\mathrm{\omega }}_{\mathit{p}}$ determined by valence-electron density for the estimation of the dielectric constant, our approach retains the first-principles nature for the single-particle energy correction. Test calculations on semiconductors such as diamond, Si, Ge, GaAs, GaP, and ZnSe show good results with the GW-corrected gap values generally within 10% of the experimental values. It is shown that an accurate and well-converged LDA result is very important for the correct self-energy correction, and its convergence with respect to the number of k points needed in the computation is much slower than that in the LDA calculation.

We present the polarized electric field modulation spectra of dilute blends of the conjugated polymer poly[2-methoxy,5-(2’-ethyl-hexoxy)-p-phenylenevinylene] (MEH-PPV) oriented in ultrahigh-molecular-weight polyethylene (PE). The photoluminescence and electric-field-modulation spectra of tensile drawn MEH-PPV/PE blends are highly anisotropic, with preferred polarization parallel to the draw axis; the field-induced absorption and photoluminescence demonstrate polarization anisotropies in excess of 150:1 and 60:1, respectively. Because of the orientation, the enhanced order and the dilution of the MEH-PPV conjugated polymer in the polyethylene matrix, the MEH-PPV/PE blend is an ideal system in which to investigate the intrinsic linear and nonlinear optical properties of oriented conjugated polymer chains. The absorption is directly proportional to the imaginary part of ${\mathrm{\chi }}^{\mathrm{}\left(1\right)\mathrm{}}$(ω), and the electroabsorption is directly proportional to the imaginary part of ${\mathrm{\chi }}^{\mathrm{}\left(3\right)\mathrm{}}$(ω,0,0) of the ordered MEH-PPV π-electron system. The optical data obtained from the oriented blends are therefore directly relevant to existing theoretical models which do not account for disorder. Using a model for the off-axis transition dipole moment of the conjugated polymer, which is based on the measured off-axis transition dipole moment of trans-stilbene and the geometric structure of MEH-PPV, we demonstrate that the observed anisotropy of 150:1 in the field-induced absorption requires that the instantaneous excited-state wave functions are delocalized over a minimum of 50 unit cells (400 Å).

Heavily carbon-doped GaAs samples with doping levels as high as 4.1×${10}^{20}$ ${\mathrm{cm}}^{\mathrm{-}3}$ were studied by photoluminescence and luminescence excitation spectroscopies. Luminescence and absorption band gaps were determined, from which the positions of the Fermi levels were deduced. A variety of factors which affect the observed emission spectra, including diffraction and refraction of the luminescence, substrate effects, and lattice contraction in the carbon-doped epilayer, were examined. A first-principles calculation was performed to calculate the density of states, the band gap, the Fermi energy, and the emission spectrum. Comparison of the calculation results with the experiments shows that the theoretical model we use is a good approximation for describing band structure for doping levels higher than ${10}^{20}$ ${\mathrm{cm}}^{\mathrm{-}3}$. However, the discrepancy between the measured and the calculated emission spectra indicates that the usual assumptions such as constant momentum matrix elements and quasiequilibrium distribution of photoexcited electrons in the conduction band may not be appropriate in treating the transition processes in highly degenerate semiconductors.

The saturated metastable defect density under AM1 (100 mW ${\mathrm{cm}}^{\mathrm{-}2}$) illumination as well as their formation and annealing kinetics have been studied in a large number of hydrogenated, and deuterated amorphous silicon films, prepared by plasma enhanced chemical-vapor deposition of silane at substrate temperatures of 250 °C–450 °C at different deposition rates, having low stable defect densities. The hydrogenated films exhibit a characteristic saturated defect density of (2–5)×${10}^{16}$ ${\mathrm{cm}}^{\mathrm{-}3}$ independent of deposition conditions. There is no variation in the formation or annealing kinetics for samples with initial defect densities <3×${10}^{15}$ ${\mathrm{cm}}^{\mathrm{-}3}$, but samples with higher initial defect densities have an order-of-magnitude longer time constants. On the other hand, there is a marked reduction of the saturated defect density in the deuterated films when the deposition temperature is increased even though the kinetics are similar to their hydrogenated counterparts. This shows that the saturated metastable defect density can be varied but is not a simple consequence of the average sample properties or the defect kinetics.

The embedded-cluster linear-combination-of-atomic-orbitals–self-consistent-field method, within the local-density framework, is used to study the substitutional impurities Li, Be, B, C, N, and O in InP. The pure semiconductor gap is found to be 0.8 eV, in fair agreement with experiments. Impurities (Im) are substituted at either In or P sites in succession forming (Im)${\mathrm{P}}_4$${\mathrm{In}}_{12}$ and (Im)${\mathrm{In}}_4$${\mathrm{P}}_{12}$ variational clusters. Li, Be, and O are found to be acceptors whereas B and C are donors. N is ambivalent: a donor at the In site, and an acceptor at the P site. Binding-energy calculations indicate that Li, Be, and B prefer to occupy the In site while C, N, and O settle at P sites. Our calculated results are compared with experiments and other calculations wherever available.

In this paper Ga ligand electron-nuclear double resonance measurements are reported on the substitutional ${\mathrm{Mn}}^{2+}$ center in GaAs. On the basis of these experiments it is concluded that the ${\mathrm{Mn}}^{2+}$ center is substitutional on a Ga site. From the electron paramagnetic resonance linewidth and the crystal-field parameter an estimate is made of the hyperfine interaction with the nearest-neighbor As atoms. The data are compared to those obtained for ${\mathrm{Fe}}^{3+}$ in III-V semiconductors. It is found that ${\mathrm{Mn}}^{2+}$ has a more localized character than the ${\mathrm{Fe}}^{3+}$ center. Finally, the quadrupole interaction of the Ga atoms is discussed.

The temperature dependences of the luminescence intensity, lifetime, and decay profiles for porous Si are studied from 5 to 271 K. The radiative decay rates were determined from the tails of the decay curves and found to have an activation-type temperature dependence above 10 K. To describe the nonradiative process we propose a model in which we assume a tunneling and a thermally activated escape of the photoexcited carriers through barriers with a Gaussian distribution in height. The temperature dependence of intensity, lifetime, and nonexponential decay profiles are successfully interpreted in terms of this model.

Several related defect centers have been observed in synthetic and type-Ib diamond following electron or neutron irradiation. Such centers, termed W centers, have similar characteristic features in that all have a spin of S=3/2 and exhibit an extremely large zero-field splitting. Spin-Hamiltonian parameters have been determined which describe the EPR results and characterize the defects. The origin of the zero-field splitting at such centers is discussed within a distorted-negative-vacancy model in which the zero-field splitting is associated with mixing of negative-vacancy molecular states through a spin-orbit interaction. It is concluded that W centers are associated with negative vacancies that have been extensively distorted.

Magnetopolaron features in the spectra of resonant Raman scattering from bulk semiconductors are studied theoretically in the scattering configuration which allows for Fröhlich interaction with the Raman phonon. We show that the theoretical treatment of scattering efficiency should include the vertex corrections to the electron-phonon and hole-phonon interaction in order to preserve the well-known dipole-forbidden nature of the scattering process. The fine balance between the electron and hole contributions to the scattering amplitude is grossly upset if the vertex corrections are neglected, resulting in a tremendous but unphysical increase of the scattering efficiency. The resonant profile for the scattering intensity is calculated as a function of the magnetic field. It shows the splitting characteristic of the magnetopolaron range.

Two-photon absorption is used to measure the shift of the longitudinal ${\mathit{Z}}_3$ and ${\mathit{Z}}_{1,2}$ excitons and the 2p exciton in CuCl under hydrostatic pressure. Hydrostatic deformation potentials for excitons and band states are determined with high accuracy. In CuBr the shift of the longitudinal ${\mathit{Z}}_{1,2}$ exciton is measured, showing a pronounced superlinear pressure dependence. In the high-pressure phase (above 4 GPa) the pressure dependence of luminescence is observed up to 9 GPa in both compounds.

In this work we use spin-dependent photoconductivity (SDPC) to study the recombination process of photoexcited carriers in hydrogenated amorphous germanium (a-Ge:H) and silicon-germanium alloys (a-${\mathrm{Si}}_{\mathit{x}}$${\mathrm{Ge}}_{1\mathrm{-}\mathit{x}}$:H). The a-Ge:H SDPC signal is found to be strongly affected by the larger spin-orbit coupling λ, when compared to a-Si:H (${\lambda }_{\mathrm{Ge}}$≊7${\lambda }_{\mathrm{Si}}$), which results in a reduced spin-lattice relaxation time. The decrease in the spin-lattice relaxation time gives the following characteristics to the a-Ge:H SDPC signal: (i) small amplitudes (-Δσ/σ≤${10}^{\mathrm{-}6}$); (ii) a linear dependence on microwave power, and strong temperature dependence. In a-${\mathrm{Si}}_{\mathit{x}}$${\mathrm{Ge}}_{1\mathrm{-}\mathit{x}}$:H alloys, the incorporation of Ge is marked by a sudden change in the SDPC signal from Si-like to Ge-like, for x<0.9. The origin of the spin-dependent recombination in a-Ge:H and a-${\mathrm{Si}}_{\mathit{x}}$${\mathrm{Ge}}_{1\mathrm{-}\mathit{x}}$:H is discussed.

Using a variational technique, the change of ground-state energy and binding energy of an impurity atom with impurity position and well width in a GaAs-${\mathrm{Ga}}_{0.7}$${\mathrm{Al}}_{0.3}$As quantum well is calculated. In the calculation the electron–surface-optical-phonon interactions are taken into account. The result is discussed. It is found that the electron-phonon interaction energy depends not only on the width of the well but also on the position of the impurity atom in the well.

The infinite-U Anderson model is applied to transport through a quantum dot. The current and density of states are obtained via the noncrossing approximation for two spin-degenerate levels weakly coupled to two leads. At low temperatures, the Kondo peak in the equilibrium density of states strongly enhances the linear-response conductance. Application of a finite voltage bias reduces the conductance and splits the peak in the density of states. The split peaks, one at each chemical potential, are suppressed in amplitude by a finite dissipative lifetime. We estimate this lifetime perturbatively as the time to transfer an electron from the higher-chemical-potential lead to the lower-chemical-potential one. At zero magnetic field, the clearest signatures of the Kondo effect in transport through a quantum dot are the broadening, shift, and enhancement of the linear-response conductance peaks at low temperatures, and a peak in the nonlinear differential conductance around zero bias.

We have investigated by means of in situ reflection high-energy electron diffraction the surface topography of various ${\mathrm{III}}_{\mathit{A}}$${\mathrm{III}}_{\mathit{B}\mathrm{-}}$V and III-${\mathrm{V}}_{\mathit{A}}$${\mathrm{V}}_{\mathit{B}}$ layers grown by solid-source molecular-beam epitaxy on (311) InP and GaAs substrates. In a wide temperature range, ${\mathrm{III}}_{\mathit{A}}$${\mathrm{III}}_{\mathit{B}\mathrm{-}}$V (311) surfaces are corrugated on a nanometer scale in a way similar to that previously reported for (311) GaAs surfaces [R. Nötzel et al., Phys. Rev. Lett. 67, 3812 (1991)]. The geometry of the corrugations can be tuned in a well defined manner by adjusting the overlayer strain. The higher the strain is, the lower the lateral periodicity and the step height are. In addition, the surface of relaxed layers exhibits the structure of the underlying substrate. Finally, on (311)-oriented substrates the relaxation of overlayer strain at least up to 4% occurs without transition toward a three-dimensional agglomerated morphology. This indicates that the surface structure modifies the relaxation paths and mechanisms. On the other hand, the III-${\mathrm{V}}_{\mathit{A}}$${\mathrm{V}}_{\mathit{B}}$ (311) surfaces are not corrugated. These results are interpreted in terms of the interplay between surface free energy, strain energy, and local-strain fields. We demonstrate that the overlayer strain is the key parameter to control the surface topography. Our work thus provides the means to directly select the interface morphology of semiconductor heterostructures.

Transport, reflectivity, and magneto-optical measurements have been performed on a p-type zero-band-gap ${\mathrm{Hg}}_{\mathit{x}}$${\mathrm{Zn}}_{1\mathrm{-}\mathit{x}}$Te/CdTe superlattice for systematic investigation of the anomalous properties of holes. The observed magnetotransport, intrasubband, and intersubband optical-transition data are fully correlated with a k⋅p band calculation and show consistent behavior resulting from two holes with different mobilities. The distinctive features of the holes are accurately explained by the extremely nonparabolicity and anisotropic nature of the valence-band structure.

We show that the Chern-Simons-Ginzburg-Landau theory of the quantum Hall effect needs a modification, because the order parameter in that theory carries an intrinsic orbital angular momentum. This quantum number contains additional information about the topological order of the Hall liquids. We propose to measure this angular momentum by circular-polarized Raman scattering.

The dissociative chemisorption of ${\mathrm{Br}}_2$ on Si(100)-2×1 at 300 K and the initial stages of layer-by-layer etching at high temperature have been studied with scanning tunneling microscopy. The Si surface maintains the 2×1 reconstruction upon Br exposure with low-yield adsorption-induced etching at room temperature. Annealing Br-covered surfaces at temperatures up to ∼800 K results in thermally activated etching, but the yield is low. Higher-yield etching of the steps and terraces takes place at ∼900 K. Healing of this etched surface begins at ∼1000 K, and there is complete Br desorption and restoration of the pre-etch morphology by ∼1100 K. Etching under continuous bombardment by ${\mathrm{Br}}_2$ at 900 K was studied by varying the halogen flux while keeping the fluence constant. Quantitative analyses of images obtained during the early stages of continuous etching showed differences in the morphologies that offered insight into etching mechanisms.