A two-shell jellium-on-jellium model has been used to study the collective electronic excitations of ${\mathrm{C}}_{60}$ coated by an increasing number N of Na atoms. We predict a transition from the π-electron polarization of the fullerene to the metallic sodium polarization as a function of increasing N. For low coverages, the Na layer produces only a broadening and fragmentation of the π plasmon of ${\mathrm{C}}_{60}$. However, if the coverage is large enough, a Na surface plasmon appears. This occurs only after the electron density of the cluster clearly shows two distinct regions, the outer one having an ‘‘average’’ density similar to that in pure Na metal or Na clusters. The induced density at the collective-mode frequency shows structure corresponding to two interfaces, ${\mathrm{C}}_{60\mathrm{-}}$Na and Na vacuum, the first peak becoming less pronounced as the coverage increases due to a transfer of oscillator strength to the metallic counterpart. The static polarizability per electron also shows this trend and tends slowly to the Na value from below after an initial sharp rise in the region N∼13–21.

Core-level and valence-band photoemission studies of films of La@${\mathrm{xaC}}_{82}$ and ${\mathrm{C}}_{82}$ demonstrate charge transfer from La to the fullerene cage and the formation of a nonmetallic solid with the uppermost band centered 0.64 eV below the Fermi level. Comparison to results for ${\mathrm{C}}_{82}$ demonstrates that there are modifications to the electronic-state distribution throughout the valence band. Bonding of the molecular radicals in the pure solid is found to be significantly stronger than in empty fullerene solids which sublime at much lower temperatures. Detailed calculations for isolated La@${\mathrm{xaC}}_{82}$ molecules allow comparison with the experimental results. They show the tendency of the La ions to move from the center of the molecule toward the cage. Comparisons of calculations for La@${\mathrm{xaC}}_{82}$ based on ${\mathit{C}}_{3\mathit{v}}$ and ${\mathit{C}}_2$ structures show that much of the energy difference between the two structures is lost upon La incorporation. Bond-length analyses allow comparison with extended x-ray-absorption fine-structure data for the metal-carbon coordination which has been measured for Y@${\mathrm{xaC}}_{82}$. They show sizable La-induced effects on the structure of the carbon cage in the proximity of the La ion.

We design a ${\mathrm{C}}_{60}$-intercalated all-carbon graphite intercalation compound, a ${\mathrm{C}}_{60}$-graphite cointercalation compound, and a K-doped ${\mathrm{C}}_{60}$-graphite cointercalation compound. Their electronic structure is studied within the density-functional theory. The geometrical structures assumed are the stage-1 cointercalation compound ${\mathrm{C}}_{60}$${\mathrm{C}}_{32}$ with alternating stacking of the ${\mathrm{C}}_{60}$ monolayer and the graphite sheet, and ${\mathrm{K}}_4$${\mathrm{C}}_{60}$${\mathrm{C}}_{32}$ where all the interstitial sites of ${\mathrm{C}}_{60}$${\mathrm{C}}_{32}$ are occupied by K atoms. In ${\mathrm{C}}_{60}$${\mathrm{C}}_{32}$, a small but finite charge transfer from the graphite layer to the ${\mathrm{C}}_{60}$ layer, as well as the redistribution of electrons within the graphite layer, is found to take place. This is a consequence of the round shape of ${\mathrm{C}}_{60}$ and the flat graphite sheet. Around the Fermi level, nearly degenerate graphite-layer valence and conduction bands exist as in the case of the graphite crystal. On the other hand, ${\mathrm{K}}_4$${\mathrm{C}}_{60}$${\mathrm{C}}_{32}$ is found to have conduction electrons on both ${\mathrm{C}}_{60}$ layers and graphite layers. The system has a high Fermi-level density of states due to the narrow ${\mathrm{C}}_{60}$ ${\mathit{t}}_{1\mathit{u}}$ conduction band but still the screening for Coulomb repulsion between conduction electrons is expected to be effective due to the coexistence of carriers in a wide graphite conduction band. Hence, the K-doped ${\mathrm{C}}_{60}$-graphite cointercalation compound is a promising candidate to exhibit superconductivity. Energetical considerations suggest that undoped as well as metal-doped ${\mathrm{C}}_{60}$-graphite cointercalation compounds should be stable and can be synthesized.

The density-functional pseudopotential method has been used to study the equilibrium lattice parameter, elastic constants, TO frequency, band structure, and valence charge distribution of AgCl. The Ag 4d electrons are fully included using an optimized pseudopotential. Results for the lattice parameter and the elastic and vibrational quantities are in satisfactory agreement with experiment, as is the calculated valence-band width. The relation of the present results with recent Hartree-Fock calculations is discussed.

Results from a projector-basis implementation of the linearized-augmented-plane-wave (LAPW) method for calculating electronic structure are presented. The implementation employs projector-basis functions to couple the plane-wave and augmentation components of the wave function. The resulting method has the same favorable scaling with system size as plane-wave-based methods such as the Car-Parrinello method, but it retains the flexibility of the LAPW basis set. Total-energy calculations for Cu, O, and Mo are used to demonstrate the accuracy of the projector-basis scheme. In addition, timings for supercells containing up to 64 Mo atoms show that the computational efficiency of the method exceeds that of the conventional LAPW method for large systems.

In silver-doped silicon, a photoluminescence band is observed with zero-phonon lines at 778.92, 779.85, and 784.31 meV, which is concluded to be associated with the spin-triplet and spin-singlet states of a bound exciton. The photoluminescence excitation spectrum of this defect, with the transitions to s-like electronic excited states dominating, is measured using Fourier-transform techniques. The electronic structure of this defect, which is known in the literature as the deep ${\mathit{E}}_{\mathit{v}}$+0.34 eV silver donor, is well described by the pseudodonor model.

Tight-binding calculations of the electronic structure of the M${\mathrm{PS}}_4$ (M=B, Al, Ga, and In) thiophosphate family, are performed. The theoretical density of states (DOS) is compared with the x-ray photoemission spectroscopy valence-band spectrum for ${\mathrm{InPS}}_4$, from which we are able to identify the main character of the experimental peaks. Differences in the density of states due to changes in the group-III element are also analyzed, since there are marked effects in the DOS due to the differences in the atomic radii of the M element. We predict these compounds to be insulators and from the calculated charges we are able to account for the oxidation number associated to each atom in these systems.

We perform first-principles theoretical studies for the doping efficiency and the diffusion mechanism of carbon dopants in GaAs. The C acceptor occupying an As site is found to be the most stable and is responsible for the high doping efficiency; however, the hole concentration is saturated to about ${10}^{20}$ ${\mathrm{cm}}^{\mathrm{-}3}$ due to the compensation by the donors such as the [100]-split interstitial (CC${)}_{\left[100\right]}$ complexes. We propose a mechanism for C diffusion accompanied with the formation and dissociation of the (CC${)}_{\left[100\right]}$ complex, with the activation energy lower than that for atomistic diffusion.

A self-consistent calculation of the traversal time in a resonant tunneling device has been presented. The model is then used to study the effect of a transverse magnetic field on the traversal time. The calculation shows the presence of bistability in the traversal time. It is also observed that the traversal time becomes progressively longer in the presence of increasing magnetic field. A quantity called the effective traversal time has been defined which can be compared to an important time scale, namely, the carrier thermalization time, and may be experimentally measurable.

In this paper we present a study of the electroreflectance spectra of several quantum wells. Using polarized light we are able to separate the contributions related to the linear electro-optical tensor (Pockels effect) and the quadratic electro-optical tensor (Kerr effect). We have observed an enhancement of the linear electro-optical contribution to the electroreflectance spectra in the light-hole region of the fundamental transition. Such enhancement is not observed in the fundamental transition of the III-V compounds and is due to the mixing of the atomiclike orbitals induced in a quantum well.

Most of the techniques used to determine the structure of reconstructed surfaces are relatively insensitive to displacements perpendicular to the surface, especially those of subsurface atoms. Reflection high-energy electron diffraction (RHEED) is sensitive to these displacements. We have performed a dynamical simulation of RHEED intensities and show that for the Si(111)-7×7 reconstruction, the dimer-adatom stacking-fault model optimized by ab initio parallel computation has given more accurate atomic positions than previous determinations by other approaches. This indicates that the relative intensities of reflections in a RHEED pattern contain important, sensitive surface structural information.

A theoretical study is presented of the lateral confinement potential (CP) in the very narrow mesa channels fabricated in the conventional two-dimensional (2D) electron gas in GaAs-${\mathrm{Al}}_{\mathit{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathit{x}}$As heterostructures. The 1D electronic structures are calculated in the framework of the confinement potential: V(x)=${\mathit{m}}^{\mathrm{*}}$${\mathrm{\omega }}_0^2$${\mathit{x}}^2$/2 for ‖x‖<${\mathit{d}}_0$/2 and V(x)=∞ at ‖x‖=${\mathit{d}}_0$/2 where ${\mathit{d}}_0$ and ${\mathrm{\omega }}_0$, the parameters characterizing such a composite potential well, are the effective width and the characteristic frequency, respectively. Calculations suggest that, for wires having structural widths ${\mathit{W}}_{\mathrm{str}}$=1500, 550, and 500 nm, the CP models consisting of flat bottoms and soft walls are more realistic than this complex potential. The experimental result for a wire of ${\mathit{W}}_{\mathrm{str}}$=250 nm cannot be explained within the composite-well model including these two cases. Thus how to explain the experiment for this wire remains an open question.

We have investigated experimentally an open semiconductor system in which electron confinement around an obstacle is obtained using a magnetic field. The magnetic field gives rise to Landau levels, and each associated edge state circulates around the obstacle, forming a set of quantized states. Tunable constrictions are fabricated by using a technique which enables us to control transport in and out of these states, producing Aharonov-Bohm oscillations as the magnetic field is swept. Surprisingly, a strong extra oscillation with the same h/e frequency develops, phase shifted by π so that the frequency appears to have doubled. We explain these results in terms of charging of isolated circulating edge states.

A high-resolution photoemission study of the Be (101¯0) surface, at a sample temperature of 100 K, is reported. Three surface-shifted Be 1s components are revealed having shifts of -700, -500, and -220 meV. These are interpreted as originating from the first, second, and third plus fourth atomic layers while the bulk signal originates from the fifth and deeper layers. The surface layer shift for this more open Be(101¯0) surface is found to be smaller than for the close-packed Be(0001) surface. Reasons for this unusual behavior are suggested.

An all-electron study on neutral, singly, and doubly charged ${\mathrm{Al}}_{\mathit{n}}$ clusters (n=2–6) has been carried out within the Kohn-Sham formalism incorporating gradient corrections to the exchange and correlation energy contributions self-consistently. We present results on their equilibrium geometries, adiabatic and vertical ionization potentials, electron affinities, atomization energies, global hardness, and possible fragmentation channels. It is shown that global hardness decreases with cluster size and that it is useful in selecting the energetically more feasible fragmentation route.

We extend the analytic solution of ${\mathrm{C}}_{60}$ molecular orbitals to include nonzero next-nearest-neighbor hopping. This next-nearest-neighbor hopping term does not destroy the icosahedral symmetry in the ground state as was claimed, and plays a role in alternating electronic structures.

The collapse of the 4f wave function in various Cs systems was studied by means of inverse photoemission (IPE). At about 10 eV above the Fermi level, the ionic compounds CsI and ${\mathrm{CsO}}_{\mathit{x}}$ display sharp IPE spectral structures that are attributed to Cs 4${\mathit{f}}^1$ electron-addition states, but no 4f signal is found in the case of Cs metal. These observations provide strong evidence for a collapse of the 4f wave function in ionic compounds; this collapse is caused by charge transfer from Cs to the anions.

Semiconducting chalcogenide glasses become optically anisotropic solids with the optical axis parallel to the direction of the inducing light beam after exposure to unpolarized light. This photoinduced anisotropy (PA) can be detected in bulk samples by means of a probing beam directed perpendicularly to an unpolarized inducing beam. In chalcogenide films, PA induced by an unpolarized glancing-angle beam can be detected by means of a probing beam that is incident normal to the films.

We have studied the photoresponses of poly(para-phenylene vinylene) (PPV) light-emitting diodes (LED’s) with PPV derivatives sandwiched between tin oxide (ITO) and metals including calcium, aluminum, and copper. Under illumination all diodes exhibit relatively large photoconductive I(V) responses which cross the dark I(V) curve at a forward-bias voltage ${\mathrm{V}}_0$ that scales with the difference in work functions between the ITO and metal electrodes, the open-circuit voltage saturates at ${\mathrm{V}}_0$ and is temperature independent, and the enhanced electroluminescence intensity of the illuminated LED’s correlates with the photocurrent. (AIP)

Positron-lifetime measurements have been performed on anodized porous silicon as a function of temperature by using a monoenergetic pulsed positron beam. In the lifetime spectra, a long-lived component, which can be attributed to orthopositronium formed at the surface of pores, has been observed. The intensity and the lifetime of the orthopositronium component were found to be strongly influenced by the sample temperature and the annealing temperature. A dramatic increase in the orthopositronium intensity after annealing at 300–400 °C suggests a reduction of positron traps near the surface due to a microstructural change. A temperature-reversible change of the orthopositronium lifetime after annealing at 400 °C suggests that there exist thermally activated dangling bonds at the pore surface at high temperatures. The present study revealed that positron lifetime is highly sensitive to changes in microstructure and pore surface conditions in porous Si.