We have used ultraviolet photoelectron spectroscopy to study the strain dependence of the valence-band offset in situ for InAs/GaAs(100) heterojunctions grown by molecular-beam epitaxy. The Ga 3${\mathit{d}}_{5/2}$ and In 4${\mathit{d}}_{5/2}$ core-level to valence-band-maximum binding-energy separations and the Ga 3${\mathit{d}}_{5/2}$ to In 4${\mathit{d}}_{5/2}$ core-level energy separations were measured as functions of strain. This requires the growth of fully relaxed ${\mathrm{In}}_{\mathit{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathit{x}}$As buffer layers of varying composition as virtual substrates. We account for the true shape of the density of states near the valence-band edge by using results of k⋅p theory. Thus a full empirical pseudopotential calculation can be avoided. Our measurements yield valence-band offset values of 0.04±0.10 eV for InAs strained to GaAs(100), 0.57±0.10 eV for GaAs strained to InAs, and 0.36±0.06 eV for both strained to InP(100), respectively, the valence-band maximum being lower at the InAs side of the junction.

We report on detailed and systematic investigations of the structural properties of thin (t<32 nm) granular ${\mathrm{Pd}}_{\mathit{x}}$${\mathrm{C}}_{1\mathrm{-}\mathit{x}}$ films with 0.1<x<0.5, where x is the metal volume fraction. The films are prepared by coevaporation of palladium (Pd) and carbon (C) onto quartz- and NaCl-crystal substrates at room temperature in an uhv system. Since carbon is known to be insoluble in palladium within the whole composition region, we obtain granular ${\mathrm{Pd}}_{\mathit{x}}$${\mathrm{C}}_{1\mathrm{-}\mathit{x}}$ films retaining their typical granular structure upon large variations of the metal volume fraction x. As revealed from transmission electron microscopic (TEM) investigations, granular ${\mathrm{Pd}}_{\mathit{x}}$${\mathrm{C}}_{1\mathrm{-}\mathit{x}}$ films with x<0.3 consist of mainly isolated small Pd clusters with mean diameters Φ of a few nanometers embedded in an amorphous carbon matrix. With increasing x clusters progressively coagulate, until at a certain metal volume fraction ${\mathit{x}}_{\mathit{p}}$=0.3—which is the so-called percolation threshold—an infinite percolative network exists throughout the entire sample. From the analysis of TEM micrographs, we obtain size distributions for both cluster diameter Φ and cluster separation s for various films with different x, from which we obtain mean values as well as typical values for Φ and s with respect to the metal volume fraction x. This offers a quantitative comparison of the structural properties of a real granular system with that of a simple cubic model, often proposed being appropriate to describe the structural properties of granular systems.

We have studied photoemission from an adsorbate-covered Ag film in an attenuated-total-reflection configuration and with direct illumination of the film. From the peaks in the photoemission as a function of angle of incidence of the light, P(ϑ), we have determined the dispersion relation for the excitation responsible for the peaks. There are three branches. The dispersion curve for the photoemission below 3.6 eV is the same as that for surface plasmon-polariton excitation. In the range between the surface and bulk plasmon limits, 3.6–3.8 eV, the dispersion curve bends backwards, i.e., with increased energy the angle for peak photoyield decreases. The third branch to the dispersion curve is above 3.8 eV and the peak photoyield occurs at increasing angle with increasing energy E. Large photoyield in both the second and third regimes requires a large imaginary part of the dielectric constant. The dispersion curve E(ϑ) and the shape of P(ϑ,E) agree with a calculation based on a model of Pepper [J. Opt. Soc. Am. 60, 805 (1970)]. We have found the temperature coefficient of the minimum in P(E) at ${\mathrm{\varepsilon }}_2$(ω)=0 to be -2.5×${10}^{\mathrm{-}4}$ eV/K.

The structure of Ni(100)-c(2×2)-Na has been investigated by analysis of extensive new low-energy electron-diffraction (LEED) data. The structure is found to contain Na atoms adsorbed in fourfold hollow sites on an unreconstructed and essentially unrelaxed substrate. The Na-Ni layer spacing is 2.38±0.04 Å and the first Ni-Ni layer spacing is 1.75±0.01 Å. These results are in good agreement with the conclusions of an early study by Demuth et al. [J. Phys. C 8, L25 (1975)]. Good agreement between experimental and calculated LEED intensities is obtained using the dynamic theory of LEED, with a conventional, muffin-tin potential for the adsorbed Na atoms and a step potential at the surface. The good agreement pertains not only to the energy positions and relative intensities of peaks in intensity-energy spectra for the diffracted beams, but also to the relative beam intensities, which span a range of nearly two orders of magnitude.

The interaction of ${\mathrm{O}}^{\mathrm{-}}$ ions with an Al surface is studied in the coupled-angular-mode method (CAM) to yield all the quantities relevant to study the O/${\mathrm{O}}^{\mathrm{-}}$-Al surface resonant-charge-transfer process. The present work takes into account the existence of a few equivalent electrons in the free O and ${\mathrm{O}}^{\mathrm{-}}$; this results in the fact that the O/${\mathrm{O}}^{\mathrm{-}}$ surface resonant-charge-transfer process does not only imply the O ground state but also the two metastable singlet states ${(}^1$D and ${}_{}{}^1{}_{}{}^{}$S). The CAM results are used to evaluate the importance of a quenching process in low-energy oxygen collisions with a metal surface. This process involves the formation of the ${\mathrm{O}}^{\mathrm{-}}$ ion as a transient intermediate.

Monolayer Cu stripes attached to the steps of W(110) are distinguished from W terraces by scanning tunneling microscopy using a Cu-induced empty state, which is characterized by inverse photoemission spectroscopy. While the contrast between Cu and W atoms is dominated by the difference of the atomic sizes at most biases, it can be reversed with a bias such that tunneling into the Cu state overwhelms the size difference. This effect provides an unambiguous and well-understood mechanism for identifying different metal atoms.

Single-scattering calculations of extended x-ray-absorption fine-structure (EXAFS) amplitude and phase were compared with the experimental first-shell data for copper, rhodium, and platinum metal. Theoretical standards used were the tables of Teo and Lee, the tables of McKale et al., and the codes excurv90, mufpot, and feff. The quality of the experimental data was shown to be very high. The experimental first-shell data were obtained by Fourier filtering. The errors introduced in the separation of the first shell from the complete EXAFS spectrum by Fourier filtering were negligible as shown in a model study. The comparison of experimental and theoretical data shows that the accuracy of theoretical standards depends mainly on the treatment of the exchange potential and the energy-dependent losses. The most accurate description of the exchange potential is the energy-dependent Hedin-Lundqvist potential with an energy-dependent self-energy, as used in feff. The use of ground state Xα or energy-independent exchange, as in the tables of McKale et al. or the codes excurv90 and mufpot, is found to be inadequate and leads to large phase and amplitude errors. Addition of an energy-dependent mean free path to the tables of McKale et al. improved the accuracy by 15–25 %. The physical reasons for the differences in accuracy of the theoretical methods examined are discussed.

We present high-resolution transmission electron-microscopic observations of the ${\mathit{sp}}^2$-bonded material that remains with the ${\mathit{sp}}^3$-bonded cubic boron nitride (cBN) in films grown by ion-assisted deposition. These observations show regions of ${\mathit{sp}}^2$-bonded material that are in a three-layer stacking configuration rather than the two-layer configuration of hexagonal boron nitride. Measurement of the lattice fringe angles shows that the observed three-layer stacking is consistent with the metastable, rhombohedral structure (rBN). Significantly, rBN allows for a diffusionless pathway for cBN synthesis under high pressure, unlike the high-activation-energy route that is required to directly convert the hexagonal phase to cBN. This low-energy pathway is considered in relation to recent work in the literature indicating that ion-induced compressive stress plays a critical role in the synthesis of thin-film cBN.

We present measurements on two serially coupled, small-capacitance tunnel junctions formed using a scanning tunneling microscope and a film composed of isolated metallic islands on an oxidized conducting substrate. Grains near the current-carrying island serve as mesoscopic charge traps. This geometry provides a model system in order to understand the effect of charge traps as noise sources in single-electron devices. We measure discontinuous shifts in tunnel current caused by changes in the local electrostatic environment. These shifts depend on the tip height above the sample. The tip is thus a driving force for charge motion near the device; its electrostatic influence extends over an area larger than the region of tunneling contact.

Recently it was demonstrated that the manipulation of atoms with the tip apex of a scanning probe microscope is a promising tool for the fabrication of nanometer-scale structures. We report time-dependent calculations of the motion of gold atoms on the NaCl(100) surface occurring after the collision with the tip of an atomic-force microscope (AFM). The interaction between the adsorbates and the tip-sample junction is calculated by considering a summation of pairwise potentials. The time-dependent trajectories of the adsorbed species are then derived from a classical dynamical dissipative method. Within this framework, the possibility of moving a single gold atom is investigated. We have also studied the formation of gold dimers and trimers from such manipulations. The stability of such clusters is discussed, in particular, its dependence on the initial position of the AFM tip. The range of tip-surface distances, which allows a controlled manipulation, has been determined.

An efficient Green’s-function method is developed to calculate electronic and ground-state properties of heavy-elements containing random alloys and their surfaces. Based on the local-density approximation, the all-electron fully relativistic linear-muffin-tin-orbital method in the tight-binding representation is used to describe disorder within the coherent-potential approximation and the semi-infinite geometry of surfaces. As a first application the electronic and ground-state properties of a random fcc-${\mathrm{Cu}}_{75}$${\mathrm{Au}}_{25}$ alloy and its (001) surface are evaluated.

The structural, electronic, and magnetic properties of small ${\mathrm{Rh}}_{\mathit{N}}$ clusters (N=2–8,10,12,13, and 19) are studied using the discrete-variational local-spin-density-functional method. The ground-state structures of ${\mathrm{Rh}}_2$, ${\mathrm{Rh}}_3$, and ${\mathrm{Rh}}_4$ are obtained and found to exhibit a tendency toward higher-dimensional geometries with longer bond lengths and more nearest-neighbor bonds. The equilibrium bond lengths in the chosen geometry for the ${\mathrm{Rh}}_5$–${\mathrm{Rh}}_8$, ${\mathrm{Rh}}_{10}$, ${\mathrm{Rh}}_{12}$, and ${\mathrm{Rh}}_{13}$ clusters are determined and show 4–6 % bond length contractions compared with the bulk interatomic spacing. A complex size dependence of the magnetic properties of the ${\mathrm{Rh}}_{\mathit{N}}$ clusters is found to be consistent with a recent experiment. The average magnetic moment per atom of the ${\mathrm{Rh}}_{\mathit{N}}$ clusters varies from 0${\mathrm{\mu }}_{\mathit{B}}$ to 2${\mathrm{\mu }}_{\mathit{B}}$. The clusters with magnetic ground states have ferromagnetic interactions, while the local moments of the 5s and 5p often align antiferromagnetically with that of the 4d. The calculated magnetic moments of the ${\mathrm{Rh}}_{10}$, ${\mathrm{Rh}}_{12}$, ${\mathrm{Rh}}_{13}$, and ${\mathrm{Rh}}_{19}$ clusters are compared and discussed with the experimental ones. Icosahedral growth is suggested for Rh clusters. The reactivity of the ${\mathrm{Rh}}_{\mathit{N}}$ clusters toward ${\mathrm{H}}_2$, ${\mathrm{N}}_2$, and CO molecules is predicted. The densities of states, exchange splittings, valence band widths, and ground-state electronic configurations are presented for all the clusters. Finally, an energy difference is identified which may be used as the criterion for the existence of multiple magnetic solutions in local-spin-density-functional calculations.

The effect of substrate steps on the adsorption of particles is considered. The problem is formulated as a lattice-gas model with nearest neighbor interactions and it is studied by a numerical transfer-matrix method. In particular, the influence of the substrate-induced row potential on adsorbed monolayers is discussed. It is found that strong row-transition-like features appear in the presence of a row potential and it is suggested that these may be seen in adsorption on vicinal faces.

The multilayer relaxations of the Zr(0001) surface and the surface electronic structure are studied using first-principles total-energy and force calculations. The distance between the outermost two layers is 4.4–4.7 % smaller than the ideal interlayer spacing, while inner layers show oscillatory relaxations. We examined two possible termination sequences for the surface and found that the hcp surface termination is more stable by about 0.06 eV per surface atom as compared to the fcc terminated structure. We observed an increase in the local density of states (LDOS) for the surface atoms around the Fermi level relative to the bulk atoms. The increase in the LDOS is found to be closely related to the existence of surface states and resonances around the Fermi level.

The electronic band-gap properties of high-density, highly tetrahedrally bonded, hydrogenated amorphous carbon are related to the size and overlap distribution of small π-bonded clusters embedded within a strained rigid ${\mathit{sp}}^3$ matrix of atomic-scale models generated by semiempirical density-functional molecular dynamics. Compared to the hydrogen-free analogues of similar density and chemical composition, the residual strain in the network is reduced. As a consequence the overlap of p orbitals between undercoordinated sites is enforced in favor of forming strong π bonds, which is consistent with a further band-gap opening relative to the hydrogen-free structures.

It is experimentally observed that adsorbate atoms and vacancies on (111) surfaces of fcc metals cluster into islands which are approximately hexagonal, but which on closer inspection turn out to have equilibrium facets that alternate in length ABABAB around the six sides of the island. By contrast, previous theoretical models for island faceting predict a rotating sequence of three lengths ABCABC around the island. We propose a model for the observed shapes, whose physical basis is the variation of the local arrangements of substrate atoms seen by an adsorbate atom. We map our model onto a generalized form of the two-dimensional Ising model having three- as well as two-spin interactions, and estimate using atom-embedding calculations the strengths of these interactions for Cu adsorbed on a Cu(111) surface. We then describe a highly efficient Monte Carlo technique for calculating the equilibrium crystal shapes of general Ising-type models in two or three dimensions, and apply it to the model in hand. Our results do indeed show alternating facet lengths very similar to those seen in experiments.

We investigate growth on vicinal surfaces by molecular-beam epitaxy making use of a generalized Burton-Cabrera-Frank model. Our primary aim is to propose and implement an analytical program based on a perturbative solution of the nonlinear equations describing the coupled adatom and dimer kinetics. These equations are considered as originating from a fully microscopic description that allows the step boundary conditions to be directly formulated in terms of the sticking coefficients at each step. As an example, we study the importance of diffusion barriers for adatoms hopping down descending steps (Schwoebel effect) during growth and post-growth equilibration of the surface.

We present a nonrelativistic multiple-scattering theory of a spin-polarized electron including the spin-orbit interaction. It is generalized to the N-scatterer case and the scattering by an ordered layer. A proper design of the potential form and the radius of a quantum dot may enhance the electron spin polarization very effectively utilizing the spin-orbit interaction.

Nuclear-magnetic-resonance spectroscopy has been used to probe the electronic structure of the M cationic sites in ${\mathit{M}}_2$${\mathrm{Mo}}_6$${\mathrm{Se}}_6$, M=(Li,Rb,Cs,In), a family of low-dimensional electronic conductors. Measured values of the quadrupole coupling parameter constant are compared to calculated values of the contribution from an ionic lattice. The measured values deviate significantly from calculation only for M=In, where the observed value substantially exceeds the calculated value. This deviation is attributed to an In p-orbital contribution to the conduction band. Such a contribution would provide a mechanism for the interchain electronic coupling observed in ${\mathit{M}}_2$${\mathrm{Mo}}_6$${\mathrm{Se}}_6$ chain systems when M=In or Tl.

Dimensionality crossovers of plasmons in carbon nanotubes, modeled as curved layered electron-gas superstructures, are investigated. For small wave vectors, the σ plasmons exhibit a transition from a one- to a three-dimensional character as the number of graphitic shells increases, in correspondence with recent experiments. For large wave vectors, plasmons of two-dimensional character are predicted.