Table of contents

For a long time, the large discrepancy between the various measurements of the thermal conductivity, k, of liquid UO₂ has represented an insoluble difficulty. The recent experimental determination of the heat capacity of this material, at temperatures from 2500 K to above the melting point, makes it possible to evaluate the heat diffusivity, alpha , in this temperature range. A discussion of the dependence of alpha on temperature indicates that the lowest measured value of k is the right one.

On the basis of semi-empirical INDO calculations of the atomistic and electronic structure of F⁺ and F centres in corundum crystals, a novel explanation of the old puzzle of F⁺ to F photostimulated conversion is suggested. Literature data are presented supporting a quite general character of this phenomenon in ionic solids with defects.

The crystal structures and molecular dynamics of solution-grown and melt-grown polycrystalline samples of n-hexatriacontane have been investigated by employing NMR, DSC and X-ray powder diffraction techniques. All solution-grown samples have the same orthorhombic structure, while the melt-grown samples are mixtures of a monoclinic and two orthorhombic structures. Defect motions of chain ends between all-trans and defect orientations are present in all crystal modifications at room temperature, but defect orientations have significantly higher populations in the orthorhombic structure which occurs only in the melt-grown sample.

The first-principles method for calculating the elastic stiffness tensor of composite crystals has been given. Applying the present method to the wurtzite and rocksalt structures of AlN under normal and elevated pressure up to 25.8 GPa, we have determined the elastic stiffness tensor completely from first principles. The theoretical values for the wurtzite structure at normal pressure are in good agreement with the experimental ones given by Tsubouchi and Mikoshiba. The elastic stiffness constants have been evaluated from the total energies of the deformed static lattices. These energies have been calculated by using a calculation method based on the variational principle.

Self-diffusion in a lattice-gas model with two-vacancy assisted hopping on the triangular lattice is investigated, by both Monte Carlo simulation and analytical calculation. A very rapid decrease of the tracer-correlation factor and marked size effects in finite lattices give evidence for strong dynamic correlations in both space and time at high particle concentration. Although the decrease of the self-diffusion coefficient over 3.5 decades for concentrations up to c=0.77 is best fitted by a power law (0.835-c)^3.54, it is argued that the model does not have a sharp dynamical phase transition with a critical concentration lower than one. The argument is based on a proof of absence of permanently blocked particles in infinite lattices at all concentrations below one. The self-diffusion coefficient is calculated analytically within a pair approximation which gives good results for lower concentrations, but fails at the higher concentrations. The approximation is in qualitative agreement with the Monte Carlo data for the tracer-correlation factor at all concentrations for a variant of the model with one-vacancy assisted hopping, in which the dynamic correlations are less pronounced.

The density-density correlation function and the site-occupation autocorrelation function are calculated for the lattice-gas model with two-vacancy assisted hopping on the triangular lattice, by both Monte Carlo simulation and analytical approximations. The non-exponential time dependence of the density-density correlation function for short wavelengths and high concentrations is explained by the strong spatial inhomogeneity of the collective diffusion process. The coefficient of collective diffusion decreases very rapidly at the highest concentrations, similarly to the self-diffusion coefficient. The analytical calculations for a two-variable approximation and a mode-coupling approximation give good results for lower concentrations. At higher concentrations the mode-coupling approximation is unstable, leading to a diverging rather than decaying correlation function for concentrations higher than 0.34.

The computational scheme of MacGillivray and Sholl (1983) for the calculation of the lattice Green's function is extended to crystalline compounds with more than one atom in the unit cell. The lattice Green's function is calculated for the high-temperature gamma -phase of Fe and for the intermetallic compound FeTi using force constants obtained from Born-von Karman fits to experimental lattice vibration spectra.

Ab initio pseudopotential calculations for H in Pd are used in order to construct an auxiliary short-ranged host-lattice-particle potential which well reproduces the large number of available ab initio data. This potential (instead of an empirical potential as employed in previous treatments) is used in a quantum mechanical calculation of hydrogen as well as of mu ⁺ and pi ⁺ states in Pd, which takes into account the dependence of the lattice relaxation on particle mass and particle state. These calculations allow us to determine Jahn-Teller coupling constants of excited states and the dependence of lattice displacements on the isotopic mass. It is found that coupling to lattice modes with E-symmetry is dominant. The effect of lattice relaxation on the total energy difference if hydrogen is localized at tetrahedral or octahedral sites and on the excitation energies as measured by inelastic neutron scattering experiments is also discussed.

For pt.I. see ibid. vol.6, 7679-7704, (1994). An auxiliary short-ranged Fe-H pair potential is constructed on the basis of ab initio pseudopotential calculations for hydrogen in alpha -Fe. This pair potential is used in self-consistent quantum mechanical calculations of hydrogen states and lattice displacements which have been performed in order to investigate the relative stability of self-trapped states formed at tetrahedral (T) or octahedral (O) sites as a function of the isotopic mass. It is found that for the light hydrogen isotopes mu ⁺ and pi ⁺ the total energy is lower at the O site whereas it is lower at the T site for the heavy isotopes p, d and t. The site with higher energy is found to be unstable. The physical origin of the stabilization of O sites for the lighter particles is the lower vibrational energy at the O site whereas the isotope effect in the lattice displacements is of minor importance in this respect.

The extended-ion method for the study of excited states in ionic crystals is modified for such two-electron defect systems as the F' centre and positronium in alkali halides. Exclusive use of 1s floating Gaussian functions as the basis is the main feature of the method. Some degree of electron correlation is built-in in this method. Applied to F' centres, it is found that in NaI the spin-singlet ground state is deeply bound, and there is also a bound triplet state. Comparison with experimental data regarding the optical and thermal dissociation energies in several alkali halides gives reasonable agreement.

We set out the effective Hamiltonian associated with the application of perturbations such as stress to ions in surroundings of cubic or tetrahedral symmetry that are subject to a T(X) tau ₂ Jahn-Teller interaction. The first- and second-order Ham factors that appear as parameters in this effective Hamiltonian are calculated by a numerical method, and checked by approximate analytical methods in the limits of strong and weak coupling. The results are compared with those produced by an approximate analytical method by Jamila and co-workers. Some discrepancies are found.

We study the transmission problem for electrons in an alternately linear-non-linear medium using a quasiperiodic Kronig-Penney model whose variance in spatial locations obeys the Fibonacci rules. We find that for waves with low intensity and long wavelength, non-linearity enhances the lattice transparency.

Convergent renormalized perturbation series in powers of the electron-electron interaction are proposed for calculating the energy of a quantum dot. The method is illustrated by calculating the ground and several excited states of a quantum dot consisting of two electrons. Comparison between the present results and those of an exact numerical integration shows the great accuracy of the proposed method over the whole range of the electron-electron coupling constant values.

Two-magnon Raman scattering in triangular Heisenberg antiferromagnets on a two-dimensional plane is discussed on the basis of the 120 degrees structure, where spins are confined on the plane by the single-ion-type anisotropy, by the use of the exchange-scattering mechanism. Spectra obtained by calculation reflect the symmetric properties of the present non-collinear spin structure. In addition to the normal structure in spectra of the alpha mode which is symmetric with respect to space inversion, we can find anomalous structure of beta and gamma modes which lacks symmetry with respect to space inversion. The former appears at twice the excitation energy for the single mode, which is also found in collinear structures for two-sublattice Neel states. In contrast, the latter is caused by the following anomaly characterizing the present 120 degrees structure in the triangular lattice antiferromagnets: the magnon for the beta mode with wave-number vector +k is effectively coupled with that for the gamma mode with +k.

We consider a two-site Anderson Hamiltonian which includes hybridization both inside the same and between different sites, as well as Coulomb and exchange interactions between them. An exact diagonalization of this Hamiltonian is performed and the competition between the Kondo and RKKY interactions is studied as a function of the exchange, J, and the hybridization parameters. The local, uniform and staggered magnetic susceptibilities are calculated. From the correlation functions it is possible to see that the existence of the Kondo effect depends strongly on the ratio between the on-site and inter-site hybridizations. No divergence of the susceptibility was found at the critical J value for finite temperatures.

We investigate the scaling behaviour of hysteresis in three-dimensional continuum models in a linear driving mode. A non-conserving N-component vector Phi is the order parameter and its dynamics are specified by the time-dependent Ginzburg-Landau theory. The ( Phi ²)² and ( Phi ²)³ models embody spatial fluctuations of Phi and their free-energy functionals have a ( Phi ²)² and a ( Phi ²)³ interaction, respectively, with O(N) symmetry. On the other hand, the mean-field model ignores all spatial fluctuations of Phi . The area A of hysteresis loop, the energy dissipation per cycle, is taken as functions of the scanning rate R and temperature. For the ferromagnetic first-order phase transition (FOPT), a scaling relation A=aRⁿ is determined where R is the field-sweeping rate, a is the coefficient and the exponent n approximately= 1/2 . We also discuss experimental evidence for real systems compatible with this result, which shows that the driving rate dependence of energy dissipation is important in non-equilibrium FOPT.

Locations and assignments of crystal-field levels in polarized absorption spectra (77 and 298 K) are reported for Eu³⁺ in the GdAl₃(BO₃)₄ huntite matrix. The levels are analysed in terms of 20 free-ion and six crystal-field parameters (D₃ symmetry). J mixing is taken into account. The influence of a parameter change on the calculated energy levels is discussed and it is shown that some parameters affect only particular levels. Spectral intensities of the 4f-4f transitions are given and parametrized using a set of B_{lambda kq} intensity parameters.

Absorption, emission and excitation spectra of LiF single crystals doped with LiOH have been measured in the UV and VUV regions of the spectrum. An absorption band at 137 nm has been identified as the UV absorption of OH/sup /ons, while bands at 172, 200 and 280 nm are assigned to defects caused by other impurities accidentally contained in the crystals. Only one emission band at 409 nm, assigned to O^2- centres, has been observed at all temperatures under excitation in the various absorptions belonging to different defects. As a consequence, an energy transfer mechanism between different centres is suggested. A weak luminescence at 255 nm, excited only in the fundamental absorption, is proposed as an intrinsic emission of LiF.

The two-dimensional angular correlation of electron-positron annihilation radiation (2D ACAR) from the heavy-fermion dense Kondo system CeB₆ was measured. The anisotropies of the measured momentum spectra were in reasonable agreement with those of model spectra based on the Fermi-surface topology suggested by band-structure calculations and de Haas-van Alphen measurements. However, the related k-space densities, obtained by an LCW analysis, were far removed from those expected from the Fermi-surface topology. The use of a recently developed filter procedure, designed to remove the characteristic long-range distortions that are frequently encountered in such LCW spectra, provided a much improved image of the Fermi-surface features.

We use molecular dynamics to study the phenomena occurring after ion impact. The simulations are two dimensional, the target contains approximately 5000 nickel atoms and the impacting ions have an energy in the range 100-1000 keV. The phenomena observed include focused collision sequences, sputtering and a thermal spike. The focused collision sequences transport much of the ion's energy away from the impact site and significantly reduce the size of the thermal spike. In an analogous process to Cerenkov radiation, it is shown that the focused collision sequences travel at supersonic speeds and lose energy via the production of sound waves. When an electron-phonon interaction is included the phenomena are slightly damped, but their essential features are unchanged.