We have studied the optical second-harmonic (SH) response of vicinal Si(111) interfaces, H-terminated or covered with a thermally grown oxide film and subjected to different annealing temperatures. We observed that the rotational anisotropy of the SH signal of a thermally grown oxide film changes after an annealing procedure. We explain this result from a modification of atomic bonding at the steps of the surface. The change in amplitude and phase of the nonlinear-optical response can be well described with an anharmonic oscillator model. We found that the nonlinear-optical signal can be correlated with the electrical quality (density of interface traps) of metal-oxide semiconductor devices produced from these Si/${\mathrm{SiO}}_2$ structures.

We present in detail a quantitatibe description of Zeeman splittings of exciton states in a nonmagnetic quantum well with diluted-magnetic-semiconductor barriers which takes interface mixing into account. For structures with high Mn concentration in the barrier we show that this effect leads to a dramatic increase of the Zeeman splitting even to a value exceeding the barrier splitting. We describe a method of interface characterization based on Zeeman-effect measurements. It enables one to study the shape of the interface profile, the influence of the growth conditions on the interface, and provides a perspective to decouple the influence of interface mixing from truly two-dimensional magnetic effects in studies of ultrathin magnetic layers. The influence of the parameters of the model on the results obtained and the practical applicability of the method are discussed. In particular, we show that the quantum well profile is strongly asymmetric in relation to the growth direction and we demonstrate the small influence of the intrinsic effect in relation to interface magnetism.

We consider a mesoscopic region coupled to two leads under the influence of external timedependent voltages. The time dependence is coupled to source and drain contacts, the gates controlling the tunnel-barrier heights, or to the gates that define the mesoscopic region. We derive, with the Keldysh nonequilibrium-Green-function technique, a formal expression for the fully nonlinear, time-dependent current through the system. The analysis admits arbitary interactions in the mesoscopic region, but the leads are treated as noninteracting. For proportionate coupling to the leads, the time-averaged current is simply the integral between the chemical potentials of the time-averaged density of states, weighted by the coupling to the leads, in close analogy to the time-independent result of Meir and Wingreen [Phys. Rev. Lett. 68, 2512 (1992)]. Analytical and numerical results for the exactly solvable noninteracting resonant-tunneling system are presented. Due to the coherence between the leads and the resonant site, the current does not follow the driving signal adiabatically: a "ringing" current is found as a response to a voltage pulse, and a complex time dependence results in the case of harmonic driving voltages. We also establish a connection to recent linear-response calculations, and to earlier studies of electron-phonon scattering effects in resonant tunneling.

First-principles atomic-cluster calculations have been carried out in the local-density approximation (LDA) for small clusters (six-atom octahedra) for each metallic element of the Periodic Table through Cu. Trends in ground-state properties of these clusters have been found to bear a significant relationship to corresponding trends in the crystalline solids. Binding energies and bond lengths for these clusters, which form fragments of the crystal lattice, display trends with atomic number which are very similar to the trends in experimental cohesive energies and lattice constants for the solid. An even closer correspondence is found when comparing with results from LDA band-structure calculations. The accuracy of small clusters in representing these trends in solids indicates that local interactions are dominant in determining the variations in cohesive and structural properties that distinguish one metal from another. Energy shifts that result from coupling the cluster to the rest of the solid are rather unform within different families of metals.

On the basis of the Su-Schrieffer-Heeger model and the analysis of the physical pictures of the so-called endohedral complexes Na@${\mathrm{xaC}}_{{}_{60}{}^{}{}_{}{}^{}}$ and Na@${\mathrm{xaC}}_{{}_{70}{}^{}{}_{}{}^{}}$, which consist of ${\mathrm{C}}_{60}$ and ${\mathrm{C}}_{70}$ molecules and a sodium ion or atom trapped inside, we have introduced two reasonable terms in the Hamiltonian to describe the interaction between the metal ion and the fullerene cage. From this simple Hamiltonian, we can reproduce the results of other first-principles calculations. Using the simplex optimization method, we have calculated the lattice distortions when all carbon atoms relax, and their effect on the potential-energy curve. Our results show that this effect is large and important. We discuss the physical meaning of our results.

Electronic shell and supershell structures in spherical multiple-shell and liquidlike clusters, taking into account the electron-ion interaction through usual empty-core pseudopotentials, have been investigated for trivalent and alkali species and compared with the conventional jellium model. In the case of multiple-shell structures we find that the jelliumlike electronic shell structure is clearly developed, in spite of the strong inhomogeneities in the ion density. The supershell pattern is, however, strongly dependent on the geometrical structure and pseudopotential parametrization. In the case of a homogeneous ionic distribution, the overall effect of the non-Coulombic behavior of the pseudopotential at short range is reflected through a softness increase of the effective electronic potential, acting as a diffuse jellium. Results about the shell and supershell structures are in better agreement with experiments, as compared to the standard jellium-model predictions. This effective surface softness, arising only from the finite size of the ionic cores, could also partially correct the present disagreement between experiment and jellium calculations with regard to the cluster polarizabilities and surface plasmon frequencies.

Cluster diffusion on a surface has been simulated with a triangular-lattice-gas model and the kinetic Monte Carlo method. The dependence with temperature and cluster size and the detailed mechanisms of cluster motion have been studied, as well as the role of impurities. Results show that cluster motion occurs through a mechanism in which random atomic displacements along the cluster perimeter make it deform and crawl. In contrast, evaporation-condensation has a negligible role at low temperatures. If impurities with a lower binding energy are added, they concentrate at the cluster perimeter and greatly facilitate cluster motion by lowering activation energies, provided that atom-impurity exchanges are possible.

Within the framework of the time-dependent Anderson-Newns model, charge exchange between an atomic particle and a surface with arbitrary band spectrum during scattering is examined. The charge state of the scattered particle is shown to be, in general, an oscillating function of its energy. Such behavior is a consequence of interference effects arising from band finiteness and from the complex structure of electronic states, which develops when an atomic particle interacts with a surface. The possibility of a substantially nonexponential law of atomic-state decay is demonstrated. The results obtained are found to give the account for known experimental dependencies.

Diamondlike carbon films, prepared by the dual-ion-beam sputtering method on a glass substrate, were thermally annealed in vacuum for 30 min at 300 °C, 400 °C, 500 °C and 600 °C, respectively. Electrical resistivity, infrared transmittance, and Raman spectroscopy were used to characterize the films before and after annealing. The results showed that the vacuum thermal annealing caused a decrease and even the removal of bond-angle disorder, and an increase of the ${\mathit{sp}}^2$ C-C-bond-dominated crystallite size and/or number in the films. This then led to decreases in both film resistivity and optical transmittance of the specimens. Meanwhile, we found that the most substantial changes in structure and in electrical and optical properties occurred at a temperature higher than 400 °C. It was concluded that the annealing effects of diamondlike carbon films in vacuum are nearly the same as those in hydrogen or in nitrogen atmospheres.

When a fluid is confined between two molecularly smooth solid surfaces (e.g., mica surfaces) it tends to form layers of molecular thickness parallel to the surfaces. We study the process of the squeezing out of a monolayer as a function of an applied normal pressure. We estimate the pressure necessary to ‘‘nucleate’’ the thickness transition (n→n-1, where n is the number of monolayers) and study the ‘‘growth’’ or ‘‘spreading’’ of the removal process after the nucleation has occurred.

Neutron-diffraction investigations of powder and consolidated ultrafine-grain-sized chromium samples indicate that antiferromagnetic order in the body-centered-cubic phase of this material can be suppressed to well below the Néel temperature of coarse-grained and single-crystal chromium. The suppression is correlated strongly with decreasing grain size. Antiferromagnetic order was not observed in powder or in consolidated samples with grain sizes less than 16 nm, indicating that free surfaces and grain boundaries play the same role in preventing antiferromagnetic order with the structure of bulk chromium. Antiferromagnetic order was observed in nanocrystalline samples with grain sizes greater than 19 nm at 20 K. No correlation is seen between the Néel temperature and the degree of long-range microstrain, or with contents of light-element impurities in the samples. Even in cases where antiferromagnetic order is detected, the transversely polarized ${\mathit{AF}}_1$ spin-density-wave magnetic phase is never seen. While this may suggest that spin-density-wave phases do not occur in nanocystalline chromium, the presence of the longitudinally polarized ${\mathit{AF}}_2$ spin-density-wave phase cannot be ruled out unambiguously.

The time-resolved spontaneous emission from a disordered molecular monolayer is calculated using Green-function techniques. In addition to a fast-decaying transmitted and reflected coherent component in the forward direction, and a slowly decaying, incoherent, spatially homogeneous background, we predict a backscattering peak related to weak localization of excitons. Analytical and numerical calculations of the energy- and disorder-dependent cooperative radiative decay rate, the quantum yield of the coherent and incoherent components, and the exciton diffusion coefficient are presented.

Photoemission spectra and scanning tunneling microscopy (STM) images of the clean ${\mathrm{Pt}}_3$Ti(111) surface are presented. Grazing-emission core-level spectra show that the topmost layer is pure platinum, modified compared with the Pt(111) surface. The Pt 4f levels at the surface are shifted 0.4 eV toward the Fermi level relative to bulk ${\mathrm{Pt}}_3$Ti while the Pt 4f and Ti 2p levels in the bulk are shifted 0.4 and 1.3 eV to higher binding energy relative to pure bulk platinum and titanium, respectively, Tunneling measurements show a surface with only metallic atoms and a small p(2×2) buckling. Our observations of molecularly adsorbed CO are not compatible with metallic titanium atoms at the surface and the STM data thus indirectly confirm that only platinum atoms are present in the topmost layer. Linear muffin-tin-orbital calculations of the bulk band structure and valence-band photoemission spectra reveal highly hybridized electron states between the Pt d and Ti d levels. The calculations give a minor charge transfer from Ti to Pt, 0.37 electrons per Ti atom, but the large core-level shifts reflect the stability of the alloy and the response to the excitation rather than the amount of charge transfer. The observed segregation of platinum to the surface and the altered electronic structure of the topmost layer, due to interaction with the underlying alloy, are in full agreement with earlier conclusions based on low-energy electron-diffraction measurements and on the chemical properties of the surface.

Photon emission by means of a scanning-tunneling-microscope junction involves surface-dependent phenomena. Experiments may be performed in ultrahigh vacuum or in air. In this paper, the correlation between the tunnel image and the photonic map of a gold surface in air is presented and the manner in which the emission can be modified to create nonemissive features on an emissive area and emissive features on a nonemissive zone is also reviewed. Emission may be either enhanced or suppressed as required.

We report the results of (e,2e) studies of crystalline graphite films. A noncoplanar asymmetric geometry was used with energies of the incident, scattered, and ejected electrons of 20, 18.8, and 1.2 keV, respectively. Thin (≃150 Å) films of highly oriented pyrolitic graphite were obtained by cleaving followed by plasma etching in an Ar-${\mathrm{O}}_2$ mixture. The (e,2e) spectra were sensitive to oxygen contamination of the film due to the plasma etching. Clean, oxygen-free graphite films were obtained by annealing in vacuum. Spectra of these films show distinct peaks in their momentum and energy distributions, but in between the peaks in the momentum distributions there is an additional contribution of unknown origin. The sharp peaks can be associated with the σ band electrons and modeled as an average of the basal plane momentum density of crystalline graphite. Good agreement in the electron dispersion was obtained. Quantitative tests of the momentum density were difficult because of the lack of a full understanding of the additional contributions, but qualitatively the calculated and measured momentum densities agree. A comparison is made with similar measurements of amorphous carbon films. An asssessment of the potential of the (e,2e) technique as applied to solids is made.

We present an improvement over the nonorthogonal tight-binding molecular-dynamics scheme recently proposed by Menon and Subbaswamy [Phys. Rev. B 47, 12 754 (1993)]. The proper treatment of the nonorthogonality and its effect on the Hamiltonian matrix elements has been found to obviate the need for a bond-counting term, leaving only two adjustable parameters in the formalism. With the improved parametrization we obtain values of the energies and bonding distances which are in better agreement with the available ab initio results for clusters of size up to N=10. Additionally, we have identified a lowest energy structure for the ${\mathrm{Si}}_9$ cluster, which to our knowledge has not been considered to date. We show that this structure (a distorted tricapped trigonal prism with ${\mathit{C}}_{2\mathit{v}}$ symmetry) is also a minimum at the Hartree-Fock level and in approximate density-functional theory, and should therefore be seriously considered as a candidate for the ground-state structure of the ${\mathrm{Si}}_9$ cluster.

We present a formalism to calculate the change of linear and nonlinear response properties induced by single atoms adsorbed on simple metal surfaces in the adiabatic limit. The atom plus jellium model of Lang and Williams for adsorbate chemisorption was extended to incorporate the presence of an external static field. On the basis of these first-principles calculations in the frame of density-functional theory the mechanisms for the increase or decrease of second-harmonic generation by a number of representative adsorbate atoms on free-electron-like jellium substrates are discussed and correlated with the various orders of the adatom and substrate response. The dipole field of the adsorbed atom causes an important long-range contributon to the response. These dipole fields also give external-field screening and depolarization and lead to a quantitative model for adatom-adatom interactions in the low-coverage regime. The results of recent SHG experiments are interpreted with our approach.

The effect of a surface on electronic fluctuations in anisotropic metals is studied theoretically. The Coulomb interacton of the electron-hole excitations is taken into account self-consistently. A system consisting of the Boltzmann equation for electronic fluctuations and the Maxwell equations for the interaction field is solved with appropriate boundary conditions. The cross section of inelastic light scattering is calculated. The scattered light observed out of the metal is a result of the collective electron interaction with radiation inside the metal. The scatteirng cross section consists of the bulk contribution of electron-hole pairs and plasmons, which differs from the scattering by an infinite metal only in the effect of the penetration of radiation in the skin layer. The surface contribution to the cross section is attributed to the electromagnetic excitations of the nonradiative into-vacuum surface plasmons (the sharp peak) and the radiative surface plasmons (the continuum with a peak).

The dynamic Jahn-Teller splitting of the six equivalent ${\mathit{D}}_{5\mathit{d}}$ polarons due to quantum fluctuations is studied in the framework of the Bogoliubov–de Gennes formalism. The tunneling-induced level splittings are determined to be ${}_{}{}^2{}_{}{}^{}$${\mathit{T}}_{1\mathit{u}}$${\oplus }^2$${\mathit{T}}_{2\mathit{u}}$ and ${}_{}{}^1{}_{}{}^{}$${\mathit{A}}_{\mathit{g}}$${\oplus }^1$${\mathit{H}}_{\mathit{g}}$ for ${\mathrm{C}}_{60}^{1\mathrm{-}}$ and ${\mathrm{C}}_{60}^{2\mathrm{-}}$, respectively, which should give rise to observable effects in experiments.

Hindered rotation is combined with the vibrations of a ${\mathrm{C}}_{60}$ molecule and is discussed in the context of icosahedral symmetry. A large difference is predicted for the occurrence of vibrations and rotations between a pure ${}_{}{}^{12}{}_{}{}^{}\mathrm{C}_{60}^{}{}_{}{}^{}$ molecular solid and a solid containing molecules with the ${}_{}{}^{13}{}_{}{}^{}\mathrm{C}$ isotope.