We examine transport in single and double barrier tunneling structures with neutral impurities. For this purpose, we have developed a three-dimensional, tight-binding supercell model of quantum transport capable of simulating potential variations both along the growth direction and in the lateral directions. We find that an isolated attractive impurity in a single barrier can produce a transmission resonance whose position and strength are sensitive to the location of the impurity within the barrier. We also study transmission in the presence of two closely spaced impurities as a function of their separation and orientation relative to the incident plane wave. Multiple impurities can lead to a complex resonance structure that fluctuates widely with impurity configuration. In addition, impurity resonances can give rise to negative differential resistance.

We provide numerical evidence that the ground state of a short-range-interaction model at ν=1/2 is incompressible and spin singlet for a wide range of repulsive interactions. Furthermore it is accurately described by a trial wave function studied earlier. For the Coulomb interaction we find that this wave function provides a good description of the lowest-lying spin-singlet state, and propose that fractional quantum Hall effect would occur at ν=1/2 if this state became the global ground state. Some conclusions of an earlier paper are invalidated by the present numerical study of a larger system; these will be indicated.

We describe photon-emission scanning-tunneling-microscopy experiments on polycrystalline silver surfaces. By performing a method of simultaneous photon emission and tunneling spectroscopy measurements, we are able to show that the photon emission process and, in particular, the intergranular contrast in the photon maps depend upon interactions between individual grains in the silver film. Factors known to affect photon emission, such as grain shape and surface morphology, are shown to be inconsistent with the detailed structure of the photon maps. We observe that certain grains behave as isolated entities while others radiate in a more collective manner. This is indicative that for these materials the photon emission behavior is strongly affected by grain-boundary effects and we discuss a recent model which takes these into account.

We derive general expressions for the self-energy of a charged particle placed in the gap between two metal surfaces and near a metallic slab. Treating the electron gas in the metal within the hydrodynamic approximation, the self-energy of a charged particle is obtained, taking into account the dispersion of the surface plasmons and the effect of the recoil of the charged particle. We show that the self-energy saturates to a finite negative value at the metal surface and inclusion of recoil and dispersion reduces the magnitude of the self-energy.

The mechanisms responsible for the core-level binding-energy (BE) shifts due to alkali-metal and halogen adsorption on metal surfaces are identified and characterized through theoretical analyses of the surface electronic structure. By means of cluster model calculations of the adsorption of K and F atoms on the Cu(100) surface, we show that ionic adsorbates, both cationic and anionic, lead to small BE shifts, typically <200 meV, of the substrate metal atoms. These small shifts arise from the cancellation of two large initial-state effects, the electric field created by the ions at the surface and the consequent polarization of the metal conduction-band electrons. These two mechanisms induce rather large shifts of opposite sign and similar magnitude in the substrate core-level BE’s, with resulting small final shifts. This is true for all electronic states, clusters, adsorption sites, and substrate-adsorbate distances. Thus, substrate BE shifts do not provide information about the bonding nature and the adsorption site. On the other hand, ionic and covalent bonding between the substrate and the adsorbate lead to significantly different shifts in the core-level BE’s of the adsorbate. The BE shifts of alkali-metal atoms adsorbed on metal surfaces as functions of the coverage provide an indication of the transition from an ionic bond at low coverage to a covalent bond at high coverage.

The mechanisms responsible for the core-level binding-energy (BE) shifts that occur for surface and bulk atoms of the Al(100) surface are identified and characterized through theoretical analyses of the surface electronic structure. Two initial-state mechanisms are shown to be responsible for the shifts of the surface atom BE’s from bulk values; final-state effects make only very minor contributions to these shifts. We are able to directly relate the important initial-state effects to the chemical environment and the chemical bonding of atoms at the surface and in the bulk. These mechanisms affect the different levels differently and we predict that the Al surface core-level shifts will be larger in magnitude for the Al 2p level than for the 2s level. The difference predicted, ∼50–100 meV, can, in principle, be resolved by present-day high-resolution photoemission measurements.

The annealing properties of the 0.5-monolayer Pd/Cu(100) surface alloy at 353 K have been investigated using reemitted-positron spectroscopy, low-energy electron diffraction, Auger electron spectroscopy, electron work function, and thermal-desorption spectroscopy. During the thermal treatment, Pd dissolution into the substrate is observed, with a time constant of ∼33 min. In addition, near-surface defects (possibly steps) are observed to anneal out with a time constant of 82 min.

We have studied the electronic properties of ultrathin layers of Sm on Ta and Cu substrates using valence-band and resonant photoemission. The average formal chemical valence of Sm on Ta increases from below 2.4 at low coverage to the value measured from bulk Sm at higher coverage, while the valence of Sm on Cu is 3 at low coverage and decreases abruptly to the same value measured for Ta substrates at higher coverage. These changes in electronic properties reflect interlayer and intralayer interactions in these films, particularly the weak Sm-Ta and strong Sm-Cu interactions.

An alternative approach to the ionization potential of transition-metal clusters has been presented. The mean effective coordination number of surface atoms is taken as a parameter to reflect the d-band width of the cluster and a tight-binding theory is used to establish this relation. We have constructed a simple expression for the ionization potential and calculated the ionization potentials of the clusters ${\mathrm{Ni}}_{\mathit{n}}$ and ${\mathrm{Nb}}_{\mathit{n}}$, and have found that they agree quite well with the experimental results.

Recent experimental work on the (001) surface of tantalum carbide has shown that this material is among the class of compounds which exhibit a marked rippling relaxation. We report here on a total-energy calculation of the relaxation of TaC(001), using a full-potential, linear-combination-of-muffin-tin-orbitals electronic-structure method. We obtain good agreement between calculated and experimental values for the degree of relaxation, and a detailed analysis of the physical mechanisms involved leads us to conclude that the nature of the rippled relaxation is determined primarily by the electrostatics associated with the bulk charge density. We also discuss other properties of the TaC(001) surface, including surface core-level shifts.

The transition-metal carbides are commonly formed with a substantial density of carbon vacancies. Here we examine the effect of carbon vacancies on the surface properties, primarily the work functions, of HfC and TaC. The method used is a full-potential linear-muffin-tin-orbital formulation, with a supercell approach for obtaining variations in the vacancy distribution. We find that the differing signs of the vacancy-induced changes in the work functions of HfC and TaC are essentially determined by the differing effect of vacancies on the relative position of the Fermi energies in the two materials.

The resistivity and Hall effect of the stage-1 ${\mathrm{CdCl}}_2$ graphite intercalation compound were investigated in the temperature range between 4 and 300 K with the current parallel or perpendicular to the c axis. The resistivity for both directions is of the form ${\mathrm{\rho }}_0$+AT+${\mathit{BT}}^2$ that results from electron-phonon scattering. The anisotropy of the resistivity is of the order of ${10}^3$ and is explained by the undulating cylindrical Fermi surface of the compound. The Hall coefficient is independent of temperature to within the experimental uncertainty and has a value of 7.0×${10}^{\mathrm{-}9}$ ${\mathrm{m}}^3$ ${\mathrm{C}}^{\mathrm{-}1}$ from a net carrier density of 8.9×${10}^{26}$ ${\mathrm{m}}^{\mathrm{-}3}$.

The specular reflectivity from the (110) surface of a Cu single crystal has been measured for temperatures between 300 and 1100 K. At lower temperatures, the reflectivity is well described by a model which yields a contraction of the spacing between the first and the second surface layers by 8% relative to the bulk layer separation. The fits also show large surface-normal vibrational amplitudes within the first few atomic layers. The line shape of the scattering transverse to the specular direction has two components below 1000 K, and evolves with increasing temperature. However, our data do not preclude the existence of a surface-roughening transition.

Picosecond time dynamics of the isotropic as well as the anisotropic transient absorption saturation have been measured in an α-sexithyenil (${\mathit{T}}_6$) thin film at room temperature by photoinduced absorption and resonant Kerr-gate experiments. Strongly damped photoacoustical oscillations have been detected superimposed on the photobleaching signal, providing a way to evaluate the sound speed in ${\mathit{T}}_6$ films along the normal to the substrate; ${\mathit{V}}_{\mathit{S}}$≃16 Å/ps. A stretched exponential function has been used to describe the time recovery of the total photoexcited population yielding a time constant τ=3.5 ns with an exponent α=0.5. This is indicative of dispersive diffusion towards recombination centers via intermolecular hopping; the Kerr-gate transmission decay provides an estimate of the photoexcitation diffusion constant, namely, D=6×${10}^{\mathrm{-}3}$ ${\mathrm{cm}}^2$/s.

Thin gold films (t<10 nm) are prepared in an UHV system simultaneously onto quartz substrates for resistance measurements and onto NaCl crystals for electron-microscopic investigation. Both samples are irradiated by ${\mathrm{Ar}}^{+}$ ions with various fluences. The temperature dependence of the resistance and the magnetoresistance (B<5 T) is measured for the nonirradiated samples as well as for the irradiated samples. Accordingly, each pair of samples is investigated by TEM to examine the influence of irradiation on the film topography. From this it is possible to explain the variation of the conductivity behavior for the irradiated and nonirradiated gold films. By magnetoresistance measurements we find a gradual shift from normal to anomalous electron diffusion due to the topographical changes in the Au films.

Electrodeposition experiments were carried out using silver nitrate or copper sulfate solutions (${\mathrm{AgNO}}_3$ and ${\mathrm{CuSO}}_4$), with organized molecular films at the interface of the solutions. We found that the two-dimensional silver films have a fractal transition boundary which grows outwards when a constant current is applied while for copper films a space-filling fractal structure grows within the film’s current boundary when a constant voltage is applied. The kinetics of the film growth is described and a model is given to explain the fractal growth of silver films. Some results for two-dimensional fractal deposition of copper in the absence of monolayers are also reported.

We have studied the growth of thin cerium films on Pt(111) with x-ray photoemission spectroscopy (XPS) and low-energy electron diffraction. No ordered structures were observed for the Ce films immediately after room-temperature deposition. After heat treatment to 770 K, ordered Ce-Pt compounds were formed. For films with initial coverages greater than 3.5 ML, a hexagonal pattern with periodicity nearly twice that of Pt(111) appeared after annealing. For this 1.96×1.96 structure, the periodicity is close to that of the compound ${\mathrm{CePt}}_2$, the XPS data show a stoichiometry of ${\mathrm{CePt}}_{2.23}$, and the observed Ce valence is 3.07. For films having initial coverages from 0.9 to 1.8 ML, a 1.94×1.94 pattern with small satellites around the main spots appeared; the Ce valence is 3.11 and the stoichiometry is ${\mathrm{CePt}}_3$. Between 2.1 and 3.2 ML, a 1.96×1.96 pattern with a superimposed 1.96×1.96 net rotated by 30° was observed with valence around 3.09. Studies of the take-off angle dependence of the emission indicate that the valence has the same value near the surface as in the bulk. We have also performed CO and ${\mathrm{O}}_2$ adsorption studies and found extremely low sticking coefficients for these molecules on the 1.96×1.96 ${\mathrm{CePt}}_{2.23}$ surface. Since these well-characterized single-crystal Ce compound films are highly suitable for photoemission experiments and adsorption studies, we have attempted to formulate empirical rules for predicting which compound species will be generated by similar means.

Quantitative low-energy electron-diffraction (LEED) IV spectra analyses of Ag, Li, and metastable clean (√3 × √3 )R30 ° Si(111) reveal that striking similarities in the LEED spectra are due to common scattering blocks of substrate atoms while the metal atoms are adsorbed at open sites on the Si lattice. This leads to the use of LEED IV spectra as a characterization method. For the (√3 × √3 )R30 ° phases of Li and Ag on Si(111), it is found that both structures can be described by a modified honeycomb-chained-trimer model. The main structural element consists of Si trimers and Si atoms in the deeper reconstructed layers. A counterexample is provided by Si(111)-(√3 × √3 )R30 °-Au for which the IV spectra are very different from those of the other three systems because the common scattering block is absent.

The electronic structure and magnetism of ${\mathrm{V}}_9$, ${\mathrm{V}}_{15}$, ${\mathrm{Cr}}_9$, and ${\mathrm{Cr}}_{15}$ clusters having body centered cubic (bcc) geometry have been studied using a Gaussian orbital basis and the local spin density approximation. For ${\mathrm{V}}_9$ and ${\mathrm{Cr}}_9$, calculations were made for several different lattice spacings. Our calculated moment of 0.33${\mathrm{\mu }}_{\mathit{B}}$ of ${\mathrm{V}}_9$ is much smaller than the previously calculated 2.89${\mathrm{\mu }}_{\mathit{B}}$ and is in very good agreement with a recent experiment. Our calculated moments of 0.67${\mathrm{\mu }}_{\mathit{B}}$, 0.07${\mathrm{\mu }}_{\mathit{B}}$, and 0.40${\mathrm{\mu }}_{\mathit{B}}$, respectively for ${\mathrm{Cr}}_9$, ${\mathrm{V}}_{15}$, and ${\mathrm{Cr}}_{15}$ at the 0 K bulk lattice spacing are also in good agreement with experiment. But ${\mathrm{Cr}}_9$ was found to have a transition near the bulk lattice spacing such that the moment becomes very large above the room temperature lattice spacing. Calculations of ${\mathrm{V}}_9$ and ${\mathrm{Cr}}_9$ at various interatomic distances show that they go through small magnetic moment to large moment transitions at certain distances, similar to the nonmagnetic to magnetic transition in bulk phase. The density of states profiles of ${\mathrm{V}}_{15}$ and ${\mathrm{Cr}}_{15}$ show the features characteristic of the bulk as had been the case in our previous calculations of Fe, Ni, and Cu.

A method for solving wave propagation problems in bounded media that operates entirely in wave-vector space is presented. No boundary conditions are used in the solution in contrast to the real-space method. The transmission/reflection problem of an electromagnetic wave on a dielectric half-space is solved as an illustration of the method. The dispersion relations of vacuum and medium, Snell’s law, the reflection angle law, and the Fresnel reflection/transmission formulas are obtained from these procedures. Applications to wave-vector dispersion problems, where the method appears to be essential, are pointed out.