Table of contents

In this work, magnetization studies on quasi bi-dimensional nanoparticles of Pd encapsulated in a graphite host as a function of magnetic field and temperature are presented. We found that an important magnetic signal can be attributed to the Pd nanoparticles. The obtained magnetic behaviour is interpreted in view of theoretical studies which predict ferromagnetism in slabs of Pd with some monolayers of thickness.

The elastic constant tensors of thin layers of In_xGa_1-xAs and In_xGa_1-xP with compositions around x = 0.5 have been determined using surface acoustic waves measured by Brillouin light scattering spectroscopy. Contrary to the case for the usual estimates, the experimentally obtained elastic constants are different from the ones calculated by simple linear combination of those of the corresponding constituent binary compounds. An averaging model combining a parallel-serial disposition of springs with the elastic constants corresponding to the binary compounds has been used to explain the elastic constants of the ternary alloys. This model explains perfectly the elastic properties of the In_xGa_1-xAs system, while those of the In_xGa_1-xP system are not reproduced.

The results of time-dependent remanent magnetization (M_IRM) behaviour (magnetic relaxation studies) in polycrystalline Gd_2-xY_xPdSi₃ alloys in a magnetically ordered state are reported. We observed that M_IRM undergoes slow relaxation with time. We have also measured ac susceptibility as a function of frequency; however, we did not find any upward shift of the peak-temperature with an increase of frequency, which suggests that these alloys are not spin-glasses. Measurements on single-crystalline Gd₂PdSi₃ show that the magnetic relaxation is an intrinsic property of the materials.

Scanning second-harmonic microscopy was used to study the lateral distribution of the second-order optical nonlinearity of a - copolymer [poly(disilanylene-2,2'-bipyridine-5,5'-diyl) ruthenium complex (PDSBpy-Ru)] thin film on a glass substrate. The orientation of the dipolar chromophores and establishment of the noncentrosymmetry required by the macroscopic second-order optical nonlinearity were resulted from natural coordination of the PDSBpy-Ru molecules to the surface Si-O-H groups on the glass substrate. A homogeneous lateral distribution of the coordination-induced poling was observed.

Colloidal spheres interacting via a purely repulsive potential provide an excellent model system to study the kinetics of solidification from the melt. Such systems are readily accessible by comparably simple yet powerful optical methods like time resolved static light scattering and microscopy, reviewed in the first part of this paper. We then present results from our own recent studies within a framework of data available from literature. In particular we investigated nucleation and growth of hard sphere crystals using combined Bragg and small angle light scattering and of charged sphere crystals using Bragg microscopy. Special attention is given to the range of validity of classical nucleation theory and of Wilson Frenkel growth and a discussion of kinetic prefactors for both processes.

We have carried out ab initio molecular-dynamics simulations for the compound-forming liquid K-Pb alloys to investigate the composition dependence of their ionic structure and electronic states. It is shown that the tetrahedral Pb₄^4- clusters, which are the so-called `Zintl ions' seen in the crystalline compound KPb, appear for the liquid K-rich alloy, while the Zintl ions are connected to each other and a complex structure is formed for the liquid equiatomic alloy. The observed structure factor characterized by the first sharp diffraction peak can be reproduced and it is found that the prepeak arises from the intermediate-range ordering of the Zintl ions. The composition dependence of the resistivity is well reproduced by means of the Kubo-Greenwood formula and its characteristic features are understood on the basis of the calculated electronic states and the ionic structure obtained.

The sound velocity, v, and the sound attenuation coefficient, , of fluid mercury have been measured at 20 MHz in the temperature and pressure range up to 1600 °C and 200 MPa. To obtain the precise sound attenuation data under high temperature and pressure, we have derived the formula for estimating by taking into account the sound absorption in the buffer rods and the acoustic impedance mismatches between the buffer rods and the sample Hg. The measurements have been carried out with four different sample lengths, and the agreements among these measurements are fairly good in the common density range. Beside the critical attenuation of sound propagation, we have observed the secondary maximum in the density dependence of at a density near 9 g cm^-3, where the metal-nonmetal (M-NM) transition occurs. In contrast to the critical attenuation, the height of the secondary maximum is almost independent of temperature. Assuming a Debye-type relaxation for the frequency-dependent adiabatic compressibility, we have estimated the relaxation time from the anomalous attenuation. We conclude that in expanded liquid Hg slow dynamics is generated by the sound pressure in the M-NM transition range.

A semi-empirical `universal' corresponding-states relationship, for the dimensionless transport coefficients of dense fluids as functions of the reduced configurational entropy, was proposed more than twenty years ago and established by many simulations. Here it is shown analytically, by appealing to Enskog's original results for the inverse-power potentials, that the quasi-universal entropy scaling can be extended also to dilute gases. The analytic form and the possible origin for the entropy scaling for dense fluids are discussed in view of this unexpected result. On the basis of the entropy scaling we predict a minimum in the shear viscosity as a function of temperature for all soft inverse-power potentials, in quantitative agreement with the available simulations.

Dynamical properties of ions are studied in SiO₂ melt by using the molecular dynamics method. The diffusion constant, ionic conductivity and velocity autocorrelation function are calculated at various pressures and temperatures. It is found that the simulated ionic conductivities are close to experimental values, and show an increase with temperature. Diffusion constants become maximum around 10 GPa, in close relation with a marked shift in the coordination number of the Si ion. The velocity autocorrelation function and its spectra are calculated by using the memory function method. These compare well with the molecular dynamics results. Discussion is given on the pressure dependence of dynamical quantities.

This paper investigates the evolution of defects in two-dimensional (2D) wet foams by using the dynamic bubble simulation approach. Two defect types are considered: a single large bubble, and a cluster of small bubbles inserted in an otherwise monodisperse hexagonal lattice. In the case of a single large defect bubble, the disorder of the cluster associated with the defect is seen to increase and peak before returning to a state having a degree of order different from that of the disordered foam scaling state. This long-timescale behaviour agrees with recent 2D wet-foam experiments yet disagrees with the majority of existing simulations on dry foam. The inclusion of a finite liquid content in the present simulations is identified as a possible reason for the improved predictions. In the case of a defect of small bubbles, coarsening trends observed in experiments are reproduced. Unlike the case for a single large bubble defect, no peak was observed in the disorder of the bubble cluster.

The phonon dispersion curves of molybdenum at zero temperature as well as at elevated temperatures are studied, using our recently developed semi-empirical tight-binding model. At zero temperature the phonon dispersion curves are obtained by frozen-phonon calculations as well as by a dynamical matrix approach. The phonon frequency shifts due to an increase in temperature are calculated by molecular dynamics (MD) simulations and the quasiharmonic contribution to the total frequency shift is determined. An alternative approach for calculating the frequency shifts without performing MD simulations is also presented.

A model is proposed for the growth of crystals with molecular orientation from a melt. The proposed free-energy functional depends on a local order parameter , which may be interpreted as a density, and the orientation of a unit vector , which is interpreted as a molecular orientation. Initial studies of the relation between the growth conditions and the growth rate and uniformity of the molecular orientation are reported.

This paper presents an analytical and numerical investigation of the energy spectrum of two-dimensional Bloch electrons subject to a periodic potential of square and hexagonal symmetry and a perpendicular sine-modulated magnetic field, applying the tight-binding model. The energy spectrum is found using an effective Hamiltonian obtained by employing the Peierls substitution in the ground-state energy band function. The resulting Schrödinger equation is solved by applying a matrix method. The energy spectrum found exhibits recursive properties similar to those discussed by Hofstadter for the case of a uniform perpendicular magnetic field. It is the objective of this paper to show that this technique can be extended successfully in the presence of a modulated magnetic field of arbitrary strength. We successfully demonstrate a Hofstadter-type spectrum in the presence of both a uniform and a modulated magnetic field.

Above 30% deuteration, ammonium hexachloropalladate undergoes the phase transition Fm3mP2₁/n, with T_c increasing to 30.2 K for pure (ND₄)₂PdCl₆. Rotational tunnelling of NH₄ and NH₃D was used to study this phase transition. The transition energies are related to the tunnelling matrix elements h_i.NH₃D can be described as a one-dimensional rotor, with the one-dimensional matrix elements being almost exactly the averages of the three-dimensional ones. In the high-temperature phase the intensities of the respective tunnelling transitions confirm a statistical occurrence probability of the isomers. In the low-temperature phase the strength of the rotational potential increases by about 30%. A broadening of the NH₃D tunnelling line with decreasing temperature above the crystallographic phase transition is interpreted as an intermediate glass phase which evolves due to long-range dipolar coupling combined with positional disorder.

We present the results of mean-field calculations on a symmetry based semimicroscopic model for betaine calcium chloride dihydrate. The model takes into account the three degrees of freedom associated with the lowest energy phonon branches as relevant variables and experimental inelastic neutron scattering data and observed transition temperatures are used to fit the coupling constants of the model potential. Although the method does not allow us to predict incommensurate phases, the results show the failure of previous models based on continuous variables to reproduce the observed phase diagram. The insertion of a new term in the model potential corresponding to anharmonic interactions between nearest neighbours is shown to be crucial to stabilize some experimentally observed commensurate phases.

In the familiar Korringa-Kohn-Rostoker (KKR) or Green function method, wave propagation between scattering sites is described by the so-called structure constants of the lattice (KKR structure constants), where these quantities are treated as having infinite spatial extent. In a recent development, it has been shown that the KKR method can be formulated in terms of screened structure constants that are of finite range (the screened KKR method). Here, we present an alternative formulation of the screened KKR method that is derived from simple manipulations of the multiple-scattering equations. We carry out density-of-states and total-energy calculations for spin-polarized and non-spin-polarized materials using the method. We point out possible unphysical features of the method.

On the basis of a method used earlier to predict the off-centre relaxed triplet self-trapped excitons in alkali halides, similar systems in caesium halides are studied. This work confirms the expected off-centre relaxation along the [100] cubic axes with increasing magnitude in the order CsI, CsBr and CsCl. The calculated emission energies are in reasonable agreement with observed values. The well-known 2.95 eV emission band of CsI:Na has been studied as a tunnelling recombination between a close pair consisting of a V_K centre and a Na atom. For a number of close-pair geometries the emission energies are close to 3 eV. The strong emission bands of CsI:Tl at 2.25 eV and 2.55 eV have been interpreted as arising from tunnelling recombination of close pairs each consisting of a Tl⁰ and a V_K centre. The calculated emission energies and polarizations are discussed in conjunction with the experimental data.

We give an exact expression for the extension of the Kane function for periodic potentials with an inversion centre. The variance of the Kane function is the sum of the semiclassical expression and the variance of the Wannier function of Kohn type.

The structure and magnetic properties of Ho₂(Co_1-xFe_x)₁₅Al₂ compounds with x = 0.0, 0.1, 0.2, 0.3, 0.4, 0.6, 0.8, 1.0 have been investigated by x-ray diffraction (XRD) and magnetization measurements. XRD patterns demonstrate that all samples are single phase with the hexagonal Th₂Ni₁₇-type structure. The substitution of Fe for Co leads to a nearly linear increase in the unit cell volume and a monotonic decrease of Curie temperature. Saturation magnetization increases with x and then decreases, attaining a maximum at x = 0.2. The magnetic anisotropy phase diagram of Ho₂(Co_1-xFe_x)₁₅Al₂ compounds has been determined from the temperature dependence of magnetization and the XRD patterns of magnetically oriented powder samples. The spin-reorientation temperature of Ho₂(Co_1-xFe_x)₁₅Al₂ compounds was found to decrease firstly with Fe content increasing for x<0.1 and then increase for 0.1<x<0.4, attaining a lowest value of 381 K at x = 0.1.