Table of contents

The authors bring together tables of current defect nomenclature and a summary of the rules actually practised (rather than idealised schemes) in choosing such labels for signals obtained with a range of spectroscopies. As well as providing a source of reference for the user lost in a maze of labels, the compilation also indicates parallels between similar defect species in very different systems (e.g. ice and quartz), even though the relationships may be far from obvious from the labels. The systems considered are all non-metals, namely ionic crystals (including oxides), silica, semiconductors (e.g. III-V and tetrahedrally coordinated II-VI), valence crystals (e.g. diamond, c-Si, a-Si) and other special hosts like ice and conducting polymers.

Positron annihilation measurements made on Mg-Cd alloys are considered. Some pronounced effects at the order-disorder transition were found when the alloy composition was close to stoichiometry. No apparent effects of the order-disorder transformation were exhibited when the alloy was far from the stoichiometric composition. The measured mono-vacancy energies of the system at the stoichiometric composition showed a pronounced decrease when the alloy passed from an ordered to a disordered state.

A detailed temperature dependence of the energy gap E₀ in the PbSe and PbTe was obtained from the transmittance measurements. A non-linear change of E₀ in the vicinity of the Debye temperature was noticed. An attempt to find an empirical formula describing the experimental dependence E₀(T) led to the conclusion that the non-linear change can be attributed to the electron interaction with the optical phonons.

It is shown that ionisation waves, familiar in connection with the positive column of a gaseous plasma, can occur in solids. This is demonstrated with a simple model in which a single type of carrier (electrons) impact ionise a deep-level trap in an insulator. Impact ionisation is described by lucky-drift theory taking into account the non-local nature of the process. An analytic theory is presented for linear waves and conditions for the existence and stability of ionisation waves are derived. It is shown that stationary, forward-travelling and backward-travelling waves are all possible, and they occur as a consequence of the non-local nature of the impact ionisation process. It is also shown that the presence of stationary waves in a finite length of sample leads to a current-controlled negative differential resistance (NDR). The end point of the growth of stationary waves in time is shown to be saw tooth waves of unchanged wavelength with an amplitude determined by the occupation of the trap at electrical neutrality; being zero for completely full or completely empty traps and being a maximum when the traps are half filled. The theory is applicable to semiconductors as well as insulators.

Discusses the common origin of 1/f noise in the power spectrum of random walks in a random environment and the 1/E-like spectral anomaly for tight-binding electronic systems with off-diagonal disorder. In one dimension the existence of the Dyson singularity, ( rho (E)) varies as mod ln^-3 mod E mod mod / mod E mod , for the density of states of the random chain as mod E mod to 0 is inevitably linked to Sinai ultraslow diffusion with the law (x²(t)) varies as ln⁴t as t to 0 and the presence of S(f) varies as ln⁴f/f power spectral density for small frequencies f. Different forms of power-law singularities are expected instead for a correlated disorder model appropriate to describe random symmetric diffusion and magnon or phonon excitations. The two problems are discussed in terms of the averaged moments of the electronic wavefunction. The 1/E behaviour is shown to rely on the underlying very general validity of the log-normal distribution for strongly disordered electronic systems. Analytical arguments are given and numerical evidence reported for the asymptotic presence of these singularities in the dimensions d=2, 3 when the disorder is sufficiently strong.

The strong-coupling behaviour of a single oxygen hole in the copper-oxide planes of the high-T_c superconductors is determined by exact diagonalisation on a finite cluster. The only energy scales are those determined by the hybridisation of the hole via Cu⁺ or Cu³⁺ excitations. As is already known for hole motion via virtual Cu³⁺ excitations the relevant two-band model ('d-p' model) may be mapped onto an effective one-band model, and the constant phase Nagaoka 'ferromagnetism' (S=(N-1)/2, where N is the number of lattice sites) is the groundstate. However, hole motion via virtual Cu⁺ excitations is more interesting and it is shown that the one-band model is not a useful approximation: the magnetic phase coherence preferred by the propagating hole is a total spin singlet. The stability of the total spin singlet state with respect to the ferromagnetic' state is discussed as the oxygen energy level is raised from the lower Hubbard copper band to the upper Hubbard band. Finally, the consequences of this calculation if super-exchange between the copper spins were included are considered.

The authors consider the variational Monte Carlo method of studying fermion systems in condensed matter physics. This method is a powerful tool for studying the properties of interacting electrons which are responsible for phenomena like the new high-T_c superconductors. Using a parallel algorithm for updating the determinants of large complex matrices on transputer arrays they study the energy and spin correlations for a 12*12 and a 22*22 lattice. Using the Hubbard model and the Gutzwiller projected wavefunction they have evaluated these quantities as functions of band filling and coupling constant. They find that although the Gutzwiller state is stable for densities near half filling it is not at low densities. For the stable states they find antiferromagnetic correlations. Statistically the thermodynamic limit is already attained with a 12*12 lattice.

Rare-earth-based magnetic alloys are formed with local random anisotropy, which is weak or strong depending on the type of rare earth used. A continuum model is used to study the correlations, susceptibility and magnetisation as functions of the anisotropy and magnetic field.

The crystallographic structure, atomic site occupancy and magnetic properties of a series of Fe_3-xV_xGe alloys in the range 0<or=x<or=1.0 have been determined. For x<or=0.4 the HCP DO₁₉-type structure of the parent Fe₃Ge is retained but beyond this composition a transformation to the doubly ordered L2₁ Heusler structure occurs. Pulsed neutron diffraction measurements suggest that the V atoms enter one Fe site preferentially for the range 0.6<or=x<or=1.0. The magnetisation, magnetic moment per Fe atom and Curie temperature all decrease with increasing x. In conjunction with magnetic susceptibility measurements carried out up to 900 K the results suggest that the L2₁ alloys have a largely itinerant character.

A complete analysis of the linear magnetoelectric (ME) effect is carried out which exploits the PT symmetry of the ME susceptibility tensor alpha . The existing (or 'old') theory in effect considers only the time-reversal antisymmetry of alpha , with the result that it restricts the effect to magnetic crystals. It is shown that this presents only half the 'picture', and that consideration of the space-inversion antisymmetry of the tensor reveals, in accord with relativistic considerations, that the effect should also occur in certain paramagnetic electric (ferroelectric and antiferroelectric) crystals, which constitute the other half of the picture. A new classification of the eligible ME crystals is proposed. Among these crystals are BaCoF₄ and Ti₂O₃, which, although prohibited by the old theory, have nonetheless been found to exhibit the effect. Other predicted candidates for the effect are presented.

ODMR investigations of recombination processes in AgCl have been extended in an effort to reconcile the apparently contradictory interpretations of other workers. The experiments described indicate that for AgCl no single recombination mechanism dominates for low-temperature excitations; the energy can relax either by a donor-acceptor pair recombination, or by an excitonic recombination. Background impurity levels could determine the relative intensities of ODMR transitions associated with these two recombination processes. Only shallowly trapped (isotropic) electrons have been identified from analysis of these spectra; no tetragonally distorted electron traps have been observed. By way of contrast, only deeply self-trapped holes (STH) are seen. Yamaga's models for the two self-trapped excitons (STE) are shown to describe adequately the ODMR data for all transitions associated with the excitons in AgCl. Observation of the STE by EPR is shown to be possible because of the excessive excited-state lifetime of this species.

The dielectric constant along the crystallographic axes was measured. The dielectric anomalies at T_I along the a and b axes show critical exponent beta =0.50+or-0.01 in the incommensurate phase II. At the second phase transition T_II no anomaly in dielectric constant was observed.

Photoacoustic (PA) measurements of F₂⁺ centres in additively coloured NaCl:OH^- crystals have been carried out. The 1.06 mu m line of a pulsed Nd:YAG laser was used for excitation and the PA signal was detected with a piezoelectric detector. The F₂⁺ centres were produced by photoaggregating samples in situ, allowing measurements of the PA signal as a function of the centre concentration to be obtained without altering the experimental geometry. In agreement with previous observations of temperature independence for both lifetime and luminescence intensity, a quantum efficiency of almost 100% was measured. The relative quantum efficiency of (F₂⁺)_H centres was determined by a comparison of the luminescent emissions of both centres at liquid-nitrogen temperature.

Some special effects in a 2D or 3D packing appear when this packing is made with particles belonging to an initially dense system. In order to achieve a better understanding of 'arching effects', a computer simulation of the collective random packing of disks in a plane under the influence of a weak central force has been set up. This work can be regarded as the first approach to the modelling of actual granular stackings resulting from mixing, pouring or fluidising particles. The dependence of the final packing fraction and mean coordination number on the initial density and on the step length has been studied on stackings of uniform disks. These functions can reach values as low as 0.76 and 3.40.

The long-wavelength-limit structure factor of liquid alkali metals is investigated using the mean-density approximation. The one-component classical plasma model is used as the unperturbed reference system. It is shown that a consistent description of long-wavelength-limit structure factor and the overall structure factor as well as the thermodynamics can be achieved to some degree using the reduced plasma parameter.

The elastic stability of a discotic liquid crystal in the theoretically predicted hexatic N+6 phase against fluctuations of orientational order is tested. For this purpose, the author derives the elastic stability conditions for Frank constants of a discotic liquid crystal, and finds that the coupling constant gamma ₃, which couples the director distortions to local rotations of the hexagonal two-dimensional lattice, is forced to be weak with respect to the other Frank constants, in order to preserve the long-range sixfold orientational order. He then shows that critical enhancements of Frank elastic constants, in the hexatic N+6 phase near the supposed continuous transition to the hexagonal discotic phase, fulfil the elastic stability conditions previously derived. Critical fluctuations of orientational order, therefore, are not able either to decorrelate the hexatic phase or to drive the transition to first order. Such a fluctuation-induced first-order transition, on the contrary, is known to play a role in smectic liquid crystals. We conclude that the N+6 phase remains orientationally correlated in the critical region, which proves the self-consistency of the model assuming such an intermediate phase, as yet experimentally undiscovered, between nematic and hexagonal discotic phases.

The authors report work-function measurements and Auger spectroscopy of liquid gallium-aluminium alloys which indicate that the surfaces of liquid dilute alloys can have an instability in vacuum less than ultra-high that is quite distinct from normal surface oxidation. The solute (aluminium) continuously diffuses to the surface where it oxidises to form an insoluble oxide. Even in very dilute liquid alloys the surface eventually becomes covered in a layer of the oxidised solute, so the surface properties are unlike those of either the solute metal or its oxide. This may explain some anomalies that have been obtained in the surface properties of liquid alloys.

The authors present theoretical and computer simulation results for the structure of a hard-sphere fluid adsorbed onto a lattice of attractive sites placed on a planar hard wall. It is shown that packing effects in the dense fluid may induce a structural phase transition in the adsorbed layer. This transition is predicted by a theory which relates the properties of the interface to those of a 2D lattice gas with an effective interaction potential.

The authors study the process of diffusion-limited colloid aggregation (DLCA) using both static and dynamic light scattering. Static light scattering is used to measure the fractal dimension of the clusters as well as their structure factor, which is found to be in good agreement with that obtained from calculation using computer-generated clusters. Dynamic light scattering is used to probe both translational and rotational diffusion motion of the clusters. A method to separate their respective contributions is developed, allowing a quantitative determination of the average hydrodynamic radius. In addition, they determine the ratio of the hydrodynamic radius to the radius of gyration for individual aggregates, and find beta =0.93. A method is developed to scale all the dynamic light scattering data onto a single master curve, whose shape is sensitive to key features of the DLCA process. Good agreement is found between the prediction of the shape of the master curve and that obtained from experiments. Using several completely different colloids, they find that the shape of their master curves are identical, their fractal dimensions are identical and their aggregation kinetics are identical. This provides strong evidence of the universality of the DLCA regime of colloid aggregation.

A relationship between the static structure factor and the 'density-density' response function in the long-wave limit is established for the case of arbitrary parameters of a system. The structure factor of the degenerate ideal Fermi gas is also examined.

A displaced squeezed state is used as the variational ground state of a spin-boson Hamiltonian describing the interaction of a two-state system with a phonon bath. The phase diagram for localisation in the ohmic dissipation is obtained by a method based on the self-consistent calculation of the effective tunnelling splitting. A comparison with the results of the usual displaced state approximation and the renormalisation group theory is also given.

Electric field heating of electrons in thin metal films deposited on silicon substrates has been used to determine the phonon transmission across the metal-silicon interface. A simple rate equation model is used to fit the electron temperature as a function of applied field. From the fit the authors have established that the phonon transmission deviates from the usual acoustic mismatch model. The authors find the phonon escape to comprise two components: ballistic escape of phonons from the film and strong scattering close to the interface.

The authors suggest that uncoupled ferromagnetic films in the monolayer regime provide a physical realisation of the two-dimensional anisotropic Heisenberg ferromagnet. Using a real space renormalisation group approach they calculate the spontaneous magnetisation of the model system. The calculations account for the observed slow decrease of this magnetisation at low temperatures, with reasonable values of the anisotropy and also for its sharp drop close to T_C.

A polarised Raman study of YBa₂Cu₃O_{7- delta} single crystals revealed significant changes of light scattering anisotropy with reducing oxygen content. Strongly (y,y) and (x,x) polarised scattering from free carriers and its coupling with the phonon system (Fano interference) are characteristic for the orthorhombic crystal ( delta approximately=0). In contrast, in the tetragonal crystal ( delta approximately=0.8) enhanced continuum scattering and asymmetry lines were observed only in the tetragonal (z,z) polarisation. These findings are attributed to the reduction of free carriers in the xy CuO₂ planes due to chains-to-planes charge redistribution initiated by the decreased oxygen concentration.