Table of contents

A two-dimensional molecular dynamics model is introduced in order to obtain a better understanding of interface amorphization. We start from a monodisperse system of particles arranged on a perfect two-dimensional triangular lattice, which is divided into two separate regions. In one of these regions the particle size is uniformly reduced, while in the other it is enlarged, preserving the total system 'volume'. Using a whole wealth of diagnostic tools, and visualizations of the particle distributions, a clear distinction between the bulk and the interface disordering transition emerges. Several interesting phenomena, such as the formation of defects, hysteresis of the transition, single-layer displacement and grain boundary development in the vicinity of the interface, show up in our simulations.

CVV Auger electrons emitted from K, Rb and Cs layers are studied using spin-resolved spectroscopy. Oriented 3p, 4p and 5p hole states are excited by circularly polarized radiation in normal incidence. The photon energies range from 12 to 24 eV. With all three materials, the degree and sign of the Auger electron spin polarization vary with the photon energy. As an atomic model of the Auger process predicts, and as a comparison of measurements with the calculated densities of states shows, the spin polarization is essentially determined by the symmetry of the final states reached in the primary (photo)excitation. Just above the excitation threshold, the preferential spin direction of the Auger electrons is measured to be parallel to the spin of the exciting photons corresponding to a predominantly s-like symmetry of the unoccupied final states reached by the excitation. At higher photon energies the preferential spin direction changes to be antiparallel to the photon spin, corresponding to the mainly d-like symmetry of unoccupied states reached by the excitation.

We discuss retarded electromagnetic modes of a film containing one or two species of carriers in the presence of an in-plane magnetic field. It is assumed that the film is characterized by a gyrodielectric tensor containing cyclotron and plasma response. We deal with the Voigt geometry (propagation transverse to the field) for both a symmetric film and a film on a substrate. A detailed discussion is given of the guided-wave modes that occupy the bulk continuum region. For the symmetric film, dispersion curves of both surface-type and guided-wave modes are reciprocal, omega (-q)= omega (q), but field profiles are different for +q and -q. For the film on a substrate both the dispersion curves and the field profiles are non-reciprocal. Numerical illustrations are given for the parameters of CMT, Cd_1-xHg_xTe.

The electronic structures of Pd and Pt atoms adsorbed on the (110) surface of silver and of the related dilute Ag-(3 at.% Pd) and Ag-(3 at.% Pt) alloys have been studied at room temperature using photoelectron spectroscopy. The observed difference between the lineshape of the virtual bound palladium 4d (platinum 5d) state on the respective isolated atoms supported on the (110) surface and that of an impurity atom embedded in the bulk is well accounted for by a many-body calculation performed within the Anderson single-impurity model. This analysis clearly indicates stronger s-d hybridization in the bulk than at the surface, consistent with the change in coordination number.

We studied the charge and potential distribution near the (100) GaP surfaces which bear the characteristic features common to polar surfaces of semiconductors. A self-consistent periodic quantum-chemical technique and a slab model are employed and the atomic and electronic structures for several surface stoichiometries are calculated. We compare these results with the predictions of the model approaches in order to check the limits of applicability of simplified models such as the electron-hole counting model or the model suggested by Nosker, Mark and Levine. We have found that zero dipole moment of the slab is an important criterion of the surface stability. The charges of the surface ions are flexible in our model and decrease strongly at the surface with respect to the bulk values. Therefore this criterion does not impose such severe restrictions on the surface stoichiometry, composition, and structure as are suggested in other models. We conclude that structure and stability of polar surfaces are determined by a balance of stoichiometric factors and electron density redistribution which should both be taken into account in a self-consistent manner.

The electronic structure of Ti₈C₁₂ and Zr₈C₁₂, in both D_2d and T_h symmetry, are investigated by the first-principles DV-X alpha SCC method. The calculated binding energies of the D_2d structure are stronger than those of the T_h structure. It reveals that the D_2d cluster is more stable than the T_h cluster. Strong interactions between C-C and Ti-C, Zr-C covalent bonding are observed. When the transition-state procedure is used, ionization potentials of 7.04 eV and 6.58 eV are obtained for D_2d Ti₈C₁₂ and Zr₈C₁₂ molecules, respectively.

The dynamics and thermodynamic properties of the sine-Gordon system with the Kac-Baker long-range interaction are considered. The dispersion relations are derived and a closed-form kink solution is obtained. It appears that the kink width and energy increase as the range of interaction increases. Using a suitable conversion of the functional integral, the transfer matrix techniques are used to evaluate the classical thermodynamic properties of the model in the low-temperature regime. They are seen to depend strongly on the range of interaction.

Magneto-quantum oscillations of the Hall constant have been observed in metallically doped three-dimensional semiconductors in a magnetic field far below the magnetic-field-induced metal-insulator (MI) transition. Especially around the magnetic field where all the electrons enter the lowest Landau level, a strong increase of the Hall constant is observed. This phenomenon is explained in terms of a field-induced MI transition in the tails of the spin-split Landau levels which can be understood in analogy to the well known MI transition in metallic semiconductors.

The interaction energy of two Li impurities in a KTaO₃ host lattice is calculated within a non-linear polarizable shell model as a function of Li-Li distance, dipole orientation, and bond direction. Our results show a complex dependence of the interaction on these parameters, suggesting that a large amount of information is lost when representing K_1-xLi_xTaO₃ by means of simplified spin-glass-like models. At low impurity concentrations, where two-body interactions are dominant, it is observed that the tendency towards ferroelectric and ferroelastic ordering increases with Li concentration. However, the complexity of the interactions together with the random spatial distribution of the impurities leads to a disordered-frustrated-ground state.

We have measured the energies and the Zeeman splitting of excitations of the non-ionized Li acceptor in CdTe by means of far-infrared absorption spectroscopy in order to determine the valence band parameters of the host material (0<or=B<or=7T, B//(100) and B//(110), 1.5K<or=T<or=4.2K). We have studied in detail the splitting of the G line (1S₃/2( Gamma ₈) to 2P₃2/( Gamma ₈)) using polarized light; g factors for initial and final states are given. Shifts of the transition energies caused by interactions with phonons are identified and separated by a critical comparison of the excitation energies measured for Li with results published for other acceptors in CdTe. Using the recently developed theory of Schmitt and co-workers, a set of valence band parameters of the host material is obtained, which reproduces very well both the excited acceptor states at B=OT and the Zeeman patterns. With a value of the static dielectric constant epsilon =10.31 taken from the literature the Baldereschi-Lipari set amounts to mu =0.68, delta =0.13, R₀=30.9 meV. The Luttinger parameters are determined as gamma ₁=4.14, gamma ₂=1.09, gamma ₃=1.62 and kappa =0.46. We compare these numbers with results obtained by different methods and published previously in the literature.

We consider a model of a weakly disordered wire connected to a clean sample. The disordered wire acts as an imperfect contact, coupled in varying degrees to different out going and in coming states in the sample. We consider here a strong magnetic field where the states are edge states, and scattering between the different edge states is weak. We derive the distribution between edge states of electrons emerging from an infinitely long, disordered current lead, and the sensitivity of an infinitely long voltage lead to electrons in different edge states, in terms of the ratio of the backscattering rate to the scattering rate between the inner and outer edge states on one side of the lead. Corrections for finite lead lengths are exponentially small so long as the lead is long compared to the backscattering length. We discuss how the results relate to recent experiments.

We present an analysis of the behaviour of a strongly disordered lead in a magnetic field connected to a less disordered sample. The disordered lead acts as an imperfect contact, coupled in varying degrees to different outgoing and incoming states in the sample: in a magnetic field these states are the different edge states. For strong disorder what would be the inner edge state in the absence of disorder does not propagate through the disordered lead, and we discuss what happens in this case. We show how an activated occupation of the inner edge state results from the different energies at which the inner Landau level starts to propagate for different disorders. We present numerical results for some particular disordered leads, quantifying the difference in energies.

Anomalies in the transport properties on RNi₅ single crystals are presented. These anomalies are characterized by the appearance of a peak, in some cases, and hysteresis in the electrical resistivity just below the Curie temperature. A qualitative analysis suggests that these features could be correlated with spin fluctuations in these compounds. The presence of two crystalline-electric-field levels close to each other as the lowest-energy states and the nature of these levels could play a crucial role in these phenomena.

The change in the electrical resistivity of Y₂O₃-doped ZrO₂ ceramics with microstructural modification was investigated. Two kinds of ZrO₂ powder containing 8 mol.% Y₂O₃ (8Y) and 3 mol.% Y₂O₃ (3Y) were used as the raw powders. Six kinds of microstructure were designed: two single-phase specimens (8YS and 3YS) and four mixed-phase specimens (Par, Ser, GG and CC). Using an impedance analyser, the contributions of the structural components, i.e. Bulk and boundary, to the electrical resistivity were identified in air up to 750 degrees C. Microstructures that contained a large phase boundary area revealed a high electrical resistivity. In the low-temperature region, the grain boundaries and/or phase boundaries played important roles in the electrical resistivity while the contribution of the bulk component increased as the temperature increased.

We have studied the permittivity of a thin (planar) percolation system (PPS) spread over a plane which is in turn sunk into the host medium. We present the numerical simulation of the PPS as well as RG speculations. It appears that u and s indices ( epsilon (L, p_c, omega =0) approximately L^sv/, epsilon (L= infinity , p_c, omega ) approximately omega ^1-u) of the planar and 2D systems are different: Delta (s/v) approximately=0.25 and u_pps approximately=0.42<u_2D=0.5. We find out that the Shklovskii-Efros scaling relation is true in spite of the fact that the PPS is a non-universal system. The indices depend upon the permittivity of the host medium.

The exact solution for the Peierls-Heisenberg antiferromagnetic (AF) model in a finite cluster is presented. The spin-spin coupling is retained up to second neighbours. The quantum mechanical character of the phonon variables is maintained in order to properly describe the fluctuations, which have a substantial role in one dimension. The problem of the frustrated AF state is briefly addressed. For the purpose of comparison, Born-Oppenheimer (BO) calculations are also provided. A noticeable result is the fact that the (semiclassical) BO approximation overestimates quantum fluctuations.

A Monte Carlo sampling technique based on 2*2 rotations of the Heisenberg spin-1 Hamiltonian containing both anisotropic and next-nearest-neighbour (NNN) terms has been attempted. Closed rings of six, eight, 10 and 12 spin-1 sites have been studied and the ground and the first excited states have been obtained. The validity of the procedure has also been tested using the spin correlation function for the ground state.

Ti-doped CaYAlO₄ crystals were grown in a reducing atmosphere. The absorption spectrum with peaks at 408 nm and 540 nm observed at room temperature was strongly polarized along the a and b axes. EPR measurements in the temperature range of 4.2-300 K have found that Ti³⁺ ions in CaYAlO₄ show orthorhombic symmetry. In order to explain the optical and EPR results, we have proposed that the Ti³⁺ ions are perturbed by an O^2- ligand ion vacancy on the c axis and additionally by either an Al³⁺ ion vacancy or a substituted Ti⁴⁺ ion at neighbouring Al³⁺ ion sites in the c plane containing the Ti³⁺ ions. This model can explain the polarization of the absorption spectrum and, also, enable us to deduce that the absorption bands with peaks at 408 nm and 540 nm are due to the perturbed Ti⁴⁺ and Ti³⁺ octahedra, respectively.

We present an ab initio multiple-scattering (MS) analysis of the Fe K-edge x-ray absorption fine structure (XAFS) in FeF₃ having a rhombohedrally distorted ReO₃-like structure. Good agreement between theoretical and experimental data was found in the energy range 2.5-14.0 Å^-1. It is shown that the MS contribution influences strongly the XAFS signals from atoms located beyond the first coordination shell. In particular (i) the XAFS signal from iron atoms Fe₂ in the second shell is amplified due to the focusing effect within the Fe₀-F₁-Fe₂ chains, caused by fluorine atoms F₁ of the first shell (Fe₀ denotes the absorber, and the number is related to the coordination shell); (ii) the contribution from iron atoms Fe₂₁, located as far as at 7.5 Å, is observed due to the 'superfocusing' effect within the Fe₀-F₁-Fe₂-F₁₀-Fe₂₁ chains with FeFFe approximately=153 degrees and FFeF approximately=180 degrees . It was also found that, despite general similarities between FeF₃ and ReO₃ crystallographic structures, the Fourier transforms of the XAFS signals at the Fe K and Re L₃ edges differ significantly beyond the first-shell peak, mainly due to differences in iron and rhenium scattering amplitudes and phases (the rhombohedral distortion becomes important only for outer shells located beyond 4 Å).