The Wigner crystal of electrons in a couple of two-dimensional layers in semiconductor structures is shown to form a two-component lattice and to possess a specific excitation, namely, an optical phonon mode. The calculated frequencies of optical magnetophonons in the magnetically frozen double-layer Wigner crystal demonstrate exponential dependence on the interlayer spacing and electronic density, and we guess that their observation can give a direct indication of the crystalline ordering in the magnetic freeze-out regime.

The resistance of a wire is calculated from the ballistic up to the diffusive transport regime through a semiclassical transmission approach. A formula is derived which describes the transition from the Shravin resistance to the Drude resistance when the mean free path becomes comparable to the wire length. The exact expression differs less than 2.5% from an interpolation formula which simply adds the two resistances. Good agreement with a recent experiment by Tarucha et al. is obtained.

The excitation of a long-range surface mode supported by gold, copper, aluminum, and iron metal island films has been investigated. These metal island films have an imaginary part of the effective dielectric constant larger than the real part at the wavelength 632.8 nm, which is distinctly different from those of the bulk metals. This indicates that the surface mode supported by these metal island films is not the free-electron surface-plasmon mode but another type of long-range surface mode. This type of surface mode can exist in any metal island film as long as the film has the proper thickness and appropriate boundary layers.

Charge transfer in stage-1 and stage-2 Rb graphite intercalation compounds was analyzed in terms of total electron-density distributions. The net charge transfer was decomposed into geometrical and ‘‘physical’’ components, obtained by the integration of reference and difference charge densities [${\mathit{n}}_{\mathrm{ref}}$(r) and δn(r)=n(r)-${\mathit{n}}_{\mathrm{ref}}$(r)], respectively, over regions associated with the host or the intercalant. The reference charge density is the superposition of charge densities for the host and intercalant sublattices treated separately, and the host-intercalant partitioning is defined by the condition that the normal component of the density gradient vanish at the interface. Employing self-consistent local-density-functional-theory calculations based on separable pseudopotentials and a plane-wave basis, we obtained a physical charge transfer of 0.2 (0.35)e and a geometrical charge transfer of 0.55e, which yield a net charge transfer of 0.75 (0.90)e for stage 1 (2).

The molecular structure factor of fullerene ${\mathrm{C}}_{60}$ has recently been determined with a high accuracy using neutron scattering over a large momentum-transfer range. The two bond lengths ${\mathit{d}}_{\mathrm{C}–\mathrm{C}}$ and ${\mathit{d}}_{\mathrm{C}=\mathrm{C}}$ have been measured with a precision of ${10}^{\mathrm{-}3}$ Å. As the temperature is lowered from 295 K down to 4 K, the single bond is elongated and the double bond shortened, but the radius of the sphere remains almost unchanged. Unfortunately, the covariance matrix obtained from the fit shows that the two lengths are strongly correlated; this means that any error made on the determination of one bond length is compensated by an opposite error on the other bond length without significantly increasing the ${\mathrm{\chi }}^2$ and, thus, the exactness of the reported temperature effect could be argued against. We present in this paper an alternative way of processing the data which allows us to remove almost completely the coupling, using an appropriate change of variables. The temperature effect is confirmed and resolved with a better accuracy.

Using scanning tunneling microscopy, a long-wavelength pattern was observed by Müller et al. [Phys. Rev. B 46, 12 899 (1992)] in Pb monolayers on Ag(111) formed by underpotential electrochemical deposition. We point out that the monolayer is incommensurate and so it should be described as having substrate-induced spatial modulations, and we also demonstrate that one can accurately calculate the long-wavelength spacing and rotation angle, which are shown to depend on electrode potential.

In our original article [Phys. Rev. B 46, 12 899 (1992)] we observed a superstructure in lead underpotential deposits on Ag(111). In their paper, Toney et al. point out that one should see a potential dependent variation in period and angle of the long-wavelength pattern. Using the dimensionless ratio κ of the long-wavelength periodicity of the moiré pattern and the atomic spacing of the lead atoms we compare the calculated dependence of κ based on the data by earlier grazing incidence x-ray-scattering experiments with data from scanning-tunneling-microscopy investigations. We find a potential dependence of the lead nearest-neighbor distance close to the expected one. Horizontal displacements proposed by Toney et al. as an explanation of the superstructure could not be detected by scanning tunneling microscopy.

The electronic structure of the ground state associated with a resonant Sc donor in CdSe and ${\mathrm{Cd}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Mn}}_{\mathit{x}}$Se was investigated by far-infrared reflectivity and millikelvin magnetoresistance measurements. Similarities in the results for samples doped either with Sc or with In, particularly in the spectral region of the 1s-2p excitations and near the metal-insulator transition, demonstrate that the ground state of the Sc impurity is hydrogenic. Thus, our findings prove that the Coulombic part of the potential of a resonant donor impurity can create hydrogenic states in the forbidden energy gap. Furthermore, the data provide support for the two-fluid model of electronic states in the vicinity of the metal-insulator transition.

We report on two hydrogen-gold-related deep levels (G2 and G3) in gold-doped p-type silicon. The levels are formed after hydrogenation by wet chemical etching. Using deep-level-transient-spectroscopy depth profiling and capacitance-voltage analysis we demonstrate that the levels are caused by injection of hydrogen into gold-doped silicon. The etching treatment results in a decrease of the gold-donor concentration, which can be fully explained by a corresponding increase of the G2 trap concentration. The results indicate that the G2 level belongs to an electrically active gold-hydrogen (Au-H) complex that changes into an electrically inactive Au-H complex upon heat treatment between 150 and 200 °C. Prolonged heat treatments within this temperature range result in dissociation of these electrically inactive Au-H complexes and the gold-donor concentration approaches its initial value.

Theoretical calculations for the resonant Raman spectra of spin-density transitions elucidating the influence of hole states participating in the resonance scattering process as well as a continuum of intersubband transitions are reported for periodically δ-doped GaAs. Raman cross sections calculated in conditions of extreme resonance with the spin-split edge show remarkable coincidence with the experimental line shapes we have measured for the resonance spectra of these transitions.

For a two-dimensional electron gas in the quantum Hall regime the time dependences of the induced charge and the current through the sample are investigated. The fastest electrodynamical process in the system is the propagation of a wave packet along the edge with the corresponding charge being held by the edge channel structure. The width of the charge strip is determined and the dispersion relation is obtained.

Conductance asymmetry of quantum point contacts at near-zero applied biases is reported. This nonlinearity in the current-voltage characteristics depends in sign and magnitude on the quantization condition of the point contact (i.e., the gate voltage). The variation in the conductance is of the order of tens of ohms against a background of several kΩ, and occurs in a very small bias region, ‖V‖≪(${\mathit{E}}_{\mathit{n}}$-${\mathit{E}}_{\mathit{n}\mathrm{-}1}$)/e around zero bias, where ${\mathit{E}}_{\mathit{n}}$ is the nth one-dimensional subband energy in the quantum point contacts. The non-Ohmic and asymmetric behavior causes a rectified dc signal as the response to an applied ac current. Second-harmonic-generation measurements also confirm the observation. The phenomenon is discussed in light of recent mechanisms for nonlinearities in point contacts due to the influence of an electric field on a two-level scatterer or on an intermode interference between mode-mixing points.

Excitonic nonlinearities in the weak-excitation limit condition are examined with time-integrated and time-resolved four-wave mixing (FWM). Below an excitation of 3×${10}^9$ excitons/${\mathrm{cm}}^2$, coherent third-order processes prevail. Echo-type signals are observed in parallel and perpendicular polarization configurations. The coherent process induced by the two-photon coherence of the two-exciton state is attributable to signals in the perpendicular configuration. The polarization dependences of a temporal profile and the intensity of FWM signals are well explained with third-order perturbational analysis that takes into account the two-exciton states and inhomogeneity.

A blue-photoluminescence (PL) band, centered at 2.6 eV, was observed in thermally and chemically oxidized light-emitting porous silicon layers with an efficiency that can exceed 0.1%. The PL decay is found to be independent of excitation intensity and detection wavelength, and to be nonexponential with a characteristic time of ∼1 nsec. A correlation between the intensity of the blue PL and the intensity of the infrared absorption related to bonded oxygen has been established. These results are examined in relation to the possible mechanisms for the blue-PL band.

We present a theoretical study of the electronic excitations in fullerene crystals by calculating the density-density correlation function in a fully nonlocal linear response theory. Our results indicate that the collective features associates with the π→${\mathrm{\pi }}^{\mathrm{*}}$ transitions show strong anisotropic properties, with peaks changing by as much as 0.7 eV in different directions. Meanwhile, the calculated mode dispersion exhibits rather weak wave-number dependence along a given direction, in general agreement with experimental results. The oscillator strength also shows anisotropic behavior, with significant weight redistribution for different directions. We also analyze this system in terms of a classical point-dipole array model, and show that this simple model approximates well the quantum results.

The angular distribution of inelastically scattered electrons has been measured with a LEED system that provides high resolution in energy and momentum simultaneously. The dipole broadening has been measured with the central dip in the forward direction, as predicted by theory. The difference between a thick dipole-active film [e.g., ZnO(10 1¯ 0)] and a thin film (${\mathrm{Al}}_2$${\mathrm{O}}_3$) on Al(111) is clearly seen in the experiment.

Inner-shell-electron energy-loss spectroscopy studies have been done for the systems in which a shallow p or d core hole exists in the final state of a cation. Optically allowed and forbidden transitions have been distinguished from spectral variations due to the breakdown of dipole selection rules. The Ti and V ${\mathit{M}}_{2,3}$ spectra and the Zr, Nb, and Mo ${\mathit{M}}_{4,5}$ spectra in layered transition-metal disulfides and related misfit-layer compounds are well explained in terms of the energy-band structures. The overall structures of the Pb and Bi ${\mathit{O}}_{4,5}$ spectra are understood within the atomic model containing j-j coupling. The Sn ${\mathit{N}}_{4,5}$ spectra in SnS and ${\mathrm{SnS}}_2$ are intermediate between them. A reasonable explanation is given by both the atomic model and the band-structure model.

We study the spherical version of a model of localized particles in a random potential which are subject to a power-law interaction ${\mathit{U}}_{\mathit{i}\mathit{j}}$∼${\mathit{r}}_{\mathit{i}\mathit{j}}^{\mathrm{-}\mathrm{\sigma }}$. In the case of a repulsive interaction with σ=1 the model is identical to the spherical version of the Coulomb-glass model of disordered localized electrons. The use of continuous variables instead of discrete occupation numbers of the sites renders the model exactly solvable. Analytic results are obtained for the free energy and for the single-particle density of states (DOS) as an example for single-particle properties. The zero-temperature DOS shows a hard gap close to the chemical potential.