Table of contents

Different models for relaxation times under the presence of random energy barriers E with a distribution P(E) are compared. Anomalous temperature dependences (such as the Vogel-Fulcher law) and anomalous relaxation (such as power laws) can be classified in this context.

Experiments on equilibrium metallic and helium crystals show an 'anomalous' mean-field behaviour of crystal shape near facet edges, in contradiction with existing microscopic theories. The possible finite-size, or 1/R, region of the observed behaviour is analysed. It is shown that these 1/R effects are negligible, even for very small crystals ( approximately 1 mu m). The author conclude that, neither finite-size effects nor classical long-range step-step interaction can account for the observed experimental behaviour.

A new infrared luminescence feature in Al_xGa_1-xAs is found to correlate with Si doping. Magnetic resonance signals are detected on this luminescence and on the visible L-bound effective-mass donor-to-acceptor recombination luminescence. An observed resonance line with g=1.95 and the infrared luminescence are explained by a simple model involving the bistable character to the DX centre.

The author presents the general formula for Lyapunov exponents of quasi-one-dimensional systems, suitable for any energy E of an electron. Some applications and discussions of singularities arising in weak disorder expansion are given.

Micrometer-sized loops of two-dimensional electron gas have been made on GaAs-AlGaAs heterostructures by electrostatic confinement. A split gate is used to define the loop, allowing the width of the conducting channels to be varied by changing the gate voltage. The magnetoresistance has been measured at low temperatures (T<100 mK) and shows strong Aharonov-Bohm oscillations with amplitudes of up to 7% of the total resistance in the narrowest devices. The oscillations are strong out to B approximately=0.5 T and then die out as B increases to approximately=1 T, with a possible dependence on the channel width. Magnetic depopulation of the ID-sub-bands is also seen.

The authors derive a variationally accurate expression for the centre of mass z₀( sigma ) of the induced charge density at a semi-infinite jellium metal in the presence of a static electric field. This expression generalises the 'displaced-profile change-in-self-consistent-field' (DP Delta SCF) expression of the work function and is shown to give good results with sample two-parameter trial profile for a wide range of electron densities and surface charges.

Monte Carlo calculations show that liquid KI near the melting maximum, at 17 kbar, has a larger coordination number (n approximately=7) that the NaCl structure solid (n=6). At this pressure the liquid solidifies into the CsCl structure (n=8). The properties of the liquid at high pressures may be calculated with good accuracy using a new modification of the hypernetted-chain equation.

A theory about the vibrations of semiconductor alloys is presented using the framework of the coherent potential approximation (CPA). Very simple self-consistent equations are obtained. A Born-type Hamiltonian was used in order to calculate the phonon density of states at a given site in real space. This CPA theory is applied to ultra-heavily doped n-type Ge in which the impurity density exceeds the normal solid solubility limits, giving rise to electrically active, as well as inactive, impurity ions. These calculations are in good agreement with Raman experimental data.

The ground-state energy of an atomic monolayer system in a periodic substrate potential is shown to be a two-dimensional fractal function of the structural parameters. The fractal nature of the ground-state energy originates from the hierarchical distribution of commensurate structures in the plane of the structural parameters. When the chemical potential is changed continuously, the system is locked to various commensurate structures in the fashion of a Devil's Staircase.

The authors present electrochemical and in situ X-ray diffraction results on the pseudoternary system Li_xCu_yMo₆Se₈ for 0<x<4 and 0<y<2.5. When prepared above 1250 degrees C, the compounds Cu_yMo₆Se₈ are single phase except for Cu_0.5Mo₆Se₈ which separates into two similar rhombohedral phases. The electrochemical behaviour and structural changes remain reversible over 0<x<4 for y<or=1. When y>1 and X+y approaches 4, irreversible phase separation is observed. An unidentified Cu-rich phase then coexists with the remaining Cu-deficient Chevrel phase, even after subsequent removal of Li. They prepared this Cu-rich phase by intercalating Cu into a Cu/Cu_yMo₆Se₈ cell but could not index the Bragg peaks. Because of the irreversible nature of this system, they present only a partial pseudo-ternary phase diagram.

The application of image intensification techniques to the study of the spatial, temporal and spectral characteristics of the light emitted during the crystallisation of sodium chloride is described. The crystallisation is induced by adding concentrated hydrochloric acid to a saturated aqueous solution of sodium chloride. It is found that during an early stage of its growth a homogeneously nucleated crystal emits a burst of some 10⁵ photons (measured in the range 390 to 570 nm). Assuming that this light emission accompanies an amorphous-to-crystalline transition in the growing 'crystal', simple thermodynamic arguments indicate that it occurs when the crystal diameter is of order 10^-7 m. Comparisons of the crystalloluminescent spectrum with the photoluminescent spectra of the crystalline and dissolved salt suggest that the light emission occurs from crystalline sodium chloride. Temporal studies revealed some long-lived crystalloluminescent emission (lasting up to several seconds) that are consistent with local high levels of supersaturation. Overall, the results broadly support the conclusions reached in earlier photomultiplier-based work by Garten and Head (1963). Unlike photomultipliers, image intensifier allow the spatial behaviour of crystalloluminescence to be studied and provide a sensitive and accurate way of recording the spectra of these transient and low-level light emissions.

A position-space renormalisation group method for calculating the electronic densities of states of disordered chains, which was first suggested by Langlois and co-workers (1983), is described. It is shown that the method enables one to calculate densities of states of random binary chains accurately. A new approach of configuration averaging is also developed in an effort to save computational time. Results are compared with numerical simulations.

A Green function diagrammatic approach is utilised to calculate the density of states of a Kondo lattice described by a periodic Anderson Hamiltonian. The hybridisation between f and conduction electrons is included in the atomic term of the Hamiltonian, which can be solved exactly. The hopping terms are considered through an approximate Dyson equation, equivalent to a generalised Hubbard I scheme. The results at T=0 K indicate the existence of a 'Kondo peak' at the Fermi level. Metallic or insulating phases are obtained for different occupations of the f and conduction states.

The author is concerned with the polaronic states of a tightly bound particle interacting with the elastic displacements of a linear host crystal. A new variational approach to Holstein's molecular-crystal model is introduced, which includes the 'adiabatic' and the 'dynamic' approximate solutions treated so far separately in the literature. It turns out that the transition from one type of solution to the other does not coincide with the transition from small to large polarons. Instead, there is a continuous transition from adiabatic to dynamic solutions in the small-polaron region while there is a discontinuous crossover in the region polarons, and six different types of polarons have to be distinguished with respect to their size and degree of adiabaticity.

The effect of neutral and charged soliton defects on the molecular geometry of long polyyne chains has been investigated using self-consistent field (SCF) molecular orbital theory. This lattice kink was found to be relatively compact for neutral and singly charged solitons but considerably more extended for doubly charged solitons in closed-shell ion polyyne chains. The spin-density wave for neutral solitons was slightly more extended than the corresponding lattice kink but its width became identical with that of the lattice kink in ion radical polyyne chains. For charged solitons, the charge-density wave was found to be considerably more extended than either the spin-density wave or the lattice kink. The present SCF results are in good agreement with the general analytical form predicted for kink-soliton solutions by field theory models. The formation energy for isolated soliton-anti-soliton pair production in long even-membered polyyne chains has also been calculated using a multi-electron configuration interaction procedure.

The authors study both the pure anisotropic and the anisotropically diluted transverse Ising model at zero temperature on the square lattice. In the former the coupling along each Cartesian direction are allowed to be different, whereas in the latter it is the bond concentration which may be different along the Cartesian directions. By means of a position space renormalisation group scheme, they find that in both cases the isotropic systems are stable with respect to anisotropy perturbations. They also discuss their results in view of a recently proposed criterion for the stability of pure systems of fixed points with respect to correlated dilution.

The bond-diluted Ising antiferromagnet in a uniform field (DAFF) on a square lattice is studied and comparison with the random-field problem is made. Position-space renormalisation group calculations for the DAFF give rise to a staircase structure in the zero-temperature phase boundary H (critical field) versus p (bond concentration), separating ordered and disordered phases. Since there is no stable long-range order in the equivalent random-field system, the authors discuss ways in which a consistent picture can be obtained, incorporating both our results and the correspondence between the two problems.

Cobalt orthophosphate, Co₃(PO₄)₂, is known to crystallise in the monoclinic system with space group P2₁/c, unit-cell dimensions a=5.063, b=8.361 and c=8.788 AA, beta =121.0 degrees and Z=2. Susceptibility measurements show the onset of antiferromagnetism at 30.0(3) K and the magnetic structure at 4.2 K has been determined from neutron diffraction measurements on synthetic powder and single-crystal samples. The magnetic structure is commensurate with the chemical unit cell with the magnetic cell doubled along a. The chemical cell contains one twofold and one fourfold site for Co²⁺ ions, whose moments lie close to b and in a more general direction within the unit cell, respectively. The refined values for the ionic moments are Co₁=3.77(10) and Co₂=3.49(7) mu _B, where Co₁ refers to the twofold site. The orientation of the magnetic moments is related to the spin-orbit coupling of the Co²⁺ ions and the exchange paths in the structure.

The dynamic properties of a system constituted by small interacting particles (iron grains dispersed in an amorphous alumina matrix) have been studied by AC susceptibility measurements at low frequency ( nu =2*10^-2 and 2*10^-3 Hz). Maxima have been found for the in-phase susceptibility chi ' (at T_B) and for the out-of-phase susceptibility chi ". The frequency dependence of T_B is analysed over a large frequency range, including previous results from AC susceptibility measurements (17<or= nu <or=10⁴ Hz) and Mossbauer spectroscopy. The results are compared with (i) the predictions from a superparamagnetic model, which takes into account the effect of magnetic interactions on the relaxation time, and (ii) the dynamic scaling laws proposed for spin glasses (the Fulcher law, the generalised Arrhenius law (transition at T_c=0) and the power law (transition at T_c not=0)). The results are not compatible with a transition at a finite temperature. They are satisfactorily explained by the authors superparamagnetic model, including inter-particle interactions, which implies a transition at a temperature of 0 K.

The probability f' of recoilless absorption in the Mossbauer effect was determined for different tin compounds, by recording the scattered radiation. The tin compounds used are CaSnO₃ SnO₂, ¹¹⁹SnO₂ and SnO₂.2H₂O. The experimental results showed that the probability f" of recoilless re-emission is larger than the probability f' of recoilless absorption: f">f'.

The behaviour of guanidine aluminium sulphate hexahydrate (GASH) at low temperatures is similar to that of triglycine sulphate (TGS). At 80 K, all GASH samples with thickness 20-8500 mu m appear to be homogeneous from careful far-infrared transmission measurements. At 4 K, for thickness below 200 mu m, the GASH samples are still homogeneous and keep their 80 K absorption coefficient values. For the thickest samples, at 4 K, a surface layer model has to be considered in a first approximation, with a quite transparent bulk and a surface layer that also maintains the 80 K absorption coefficient values. An explanation is suggested for both TGS and GASH in terms of reduction to one formula unit per primitive cell at some temperature T₀. The transition should only occur if the sample dimensions are large enough, and it should start from the centre of the sample. From extensive measurements, they have found far-infrared spectroscopic evidence at temperatures T below the transition temperature T₀ that both GASH and TGS single crystals have a superstructure where the electric dipole associated with one formula unit is slightly rotated when the primitive cell is translated from the centre to the surface of the sample.

Integrated intensity and lifetime measurements of the Eu²⁺ and Mn²⁺ emissions were carried out in the temperature range 10-300 K in order to obtain information on the mechanism of the Eu to Mn energy transfer in monocrystalline NaCl. Calculations based on the Forster-Dexter model of energy transfer indicate that the fast rise time of the Mn fluorescence after pulse excitation of Eu²⁺ can only be explained if a short-range interaction such as an electric dipole-quadrupole of exchange-type interaction is active. These calculations also substantiate the model proposed by Rubio and co-workers (1987) in the sense that energy transfer from Eu²⁺ to Mn²⁺ in the lattice of NaCl takes place in the Eu-Mn pairs which are preferentially formed in this material.