Table of contents

On the basis of the Overhauser (1971) and Axe (1980) treatment of Debye-Waller factors in incommensurate systems, a calculation allows the authors to determine the temperature range Delta T on which the phase is floating in modulated structures. Their estimations are in good agreement with experimental results obtained by Blinc et al. (1983) in Rb₂ZnBr₄ from NMR measurements ( Delta T approximately=10K) and by Emery et al. (1983) in ThBr₄ from EPR measurements ( Delta T approximately=2.6K).

The authors have investigated the two-dimensional (2D) electron hopping energy integral in order to calculate the impurity density of states of doped semiconductors. A cluster model is outlined for the two-dimensional disordered system. It is shown that the hopping matrix is very sensitive to a change in the dimensionality of the system, i.e. from 3D to 2D system. The impurity band is symmetric and has a considerable bandwidth for high concentration, while for low concentration it is drastically reduced by the cut-off of the long-range hopping energy. The results of other models are discussed.

The authors show that it is possible to suppress the metal-insulator transition in (TMTSF)₂ReO₄ by fast quenching at moderate pressures. Fast cooling down to the temperature where the resistivity reaches its maximum value, and a slow cooling below, increases the electrical conductivity below the transition temperature by almost three orders of magnitude and maximises the critical temperature to the superconducting ground state at 1.52K under 10.5 kbar. In addition, evidence for a spin density wave state is reported for the first time in (TMTSF)₂ReO₄ at 10 kbar below 12K under appropriate cooling conditions.

The author explains why Kimball's considerations in two special models are not appropriate for showing essential features of localisation.

RuS₂ single crystals have been grown by chemical transport reaction using ICI₃ as transport agent. As confirmed by X-ray investigations, the specimens crystallise with the correct pyrite structure. The electrical resistivity and Hall effect studies showed n-type semiconducting behaviour. At room temperature, the electrical resistivity was 6.2*10^-2 Omega cm, the carrier concentration was 3.2*10¹⁷ cm^-3 and the Hall mobility was 310 cm² V^-1 s^-1. The optical gap was determined at about 1.3 eV.

The deviations from one-parameter scaling recently reported in Si inversion layers (see ibid., vol.16, p.L285, 1983) can be understood in terms of a weak spin-orbit scattering.

The problem of the Ising spin glass is reformulated in terms of new spins, naturally related to the Edwards-Anderson order parameter. It is shown that in terms of the new spins the system is ferromagnetic, providing a simple explanation of possible spin glass ordering. This result is used to obtain upper bounds on spin glass correlations as well as on the transition temperature.

Optical detection of the electron spin resonance of the (AlO₄)⁰ centre in crystalline smoky quartz at 1.4K has allowed identification of bands in the visible spectrum associated with this centre, at 1.96(5) and 2.85(5) eV.

A new formula is derived for elastic constants using a rigorous quantum statistical treatment based on obtaining the derivatives of the Helmholtz free energy with respect to strain elements. The stress tensor and the second-order elastic constants are formulated exactly for any homogeneous stress state by using a perturbation method applied to the Schrodinger equation. Third- and higher-order elastic constants may also be obtained in the same way. The formula for the elastic constants contains four terms, all of which are thermal averages. Two of these terms are the momentum term and the quantum stress fluctuation term. Another term involves the thermal average of the potential energy of a lattice and it reduces, in the case of two-body central forces, to the well known formula for the elastic constants of Born and Huang (1954) without making any of the following three assumptions: zero stress, sublattice equilibrium, every lattice point at a centre of inversion. The remaining term is a second-order perturbation term with energy difference denominators, and is the contribution of internal strain to the elastic constants.

Quantum rotational excitation energies in phase II of solid CH₄ consisting of only A-spin species are calculated using the extended James-Keenan model. It is shown that the first excited rotational levels of the D_2d and O_h molecules split into six levels at wavevector k=0. This result is compared with classical lattice dynamical calculations.

A renormalised response theory proposed recently by Falter and Selmke (1980-1) is applied to the lattice dynamics of covalent materials. Using a model suggested earlier by Falter and for the lattice dynamics of insulators the phonon dispersion of Si has been calculated. Within this framework the influence of the higher excited states on the phonon spectrum and the dependence of the phonon frequencies on the dimension of the density response function are studied. Furthermore, the role of exchange and correlation on the spectrum is discussed.

The correlated-effective-theory described by Lines in 1974 is applied to the description of phase transitions in Ising systems in which the interactions involve homogeneous strains. The critical properties of such systems are expected to be classical, as Colwey predicted (1976), and there is an anomaly at q=0 in the effective Ising interaction J_q. Depending on the relative strength of short-range and long-range contributions to J_q, the phase transition may be of either first or second order. It is of first order if Sigma _q(J₀-J_q)^-2>(3/2N)(Z_q(J₀-J_q)^-1)². A simple parametric description of J_q is introduced, and the relevance of this theory to the description of the first-order Jahn-Teller phase transition in DyAsO₄ is discussed.

Precise measurements of the order parameter and the c axis electric susceptibility have been made near the Jahn-Teller phase transitions in DyAsO₄ and DyVO₄. Both compounds show large departures from simple mean-field (MF) theory. The phase transition in DyAsO₄ is of first order while that of DyVO₄ is continuous. The data are analysed using a correlated-effective-field (CEF) model in which a simple approximation, consistent with classical critical behaviour, is made for the dependence on q of the effective exchange constant J(q). The CEF model gives an excellent description of the data over a wide range of temperatures and magnetic fields, both below and above T_D. This agreement establishes that the competing effects of short- and long-range interactions are responsible for the deviations from MF behaviour; the model accounts for the first- and second-order nature of the phase transitions in these compounds. The large c axis dielectric anomalies imply the existence of appreciable electric dipole moments and are interpreted in terms of coupling to odd-symmetry q=0 modes. From the measurements they find that the odd-symmetry (electric dipolar) couplings are in order of magnitude larger in DyVO₄ than in DyAsO₄. These interactions of relatively long range appear to be strong enough in DyVO₄ to ensure a continuous phase transition, but are too weak to do so in DyAsO₄.

Measurements of the electric susceptibility (which is orthogonal to the magnetic susceptibility) in LiHoF₄ show a sharp cusp at T_c that cannot be explained by simple mean-field theory. Correlated-effective-field (CEF) theory is applied to this system and is found to provide an excellent fit to the data over a wide range of temperatures. In addition the critical behaviour of LiHoF₄ is re-examined by the application of the authors' CEF model to the precise order-parameter data of Griffin and co-workers (1980). Although the authors' CEF model ultimately shows classical critical behaviour, over the temperature range of the data it yields a fit of quality equivalent to that obtained by Griffin and co-workers using renormalisation-group theory. Hence the experiments of Griffin and co-workers do not unequivocally establish a marginal dimensionality of three for LiHoF₄.

The authors study the electronic structure of the various charge states of the ideal vacancies in Ge, GaAs and ZnSe crystals in the tight-binding approximation. They show that the vacancy states are strongly localised on the orbitals of the neighbouring atoms but that they are localised more on the s or on the p orbitals depending upon the chemical nature of the neighbours of the vacancy. The authors study also the splitting between the A₁ and T₂ vacancy levels as a function of the hopping integrals between two of their nearest neighbours; they show that the absolute and relative importances of these hopping integrals depend also upon the chemical nature of the neighbours of the vacancy. They study also the charge states of the vacancies and give estimates for the effective Coulomb interaction U. The results are compared with available calculations and experiments.

Anderson localisation in two dimensions is modelled numerically using the Langreth-Abrahams generalisation of the Landauer formalism. This approach allows an investigation of systems with disorders as small as W/V=0.1. The increasing effect of the finite system size as seen in the surface resistance is studied and taken into account. Results agree with the scaling theory of Abrahams et al. (1979) in the small disorder region.

The bond disordered two-dimensional Ising model is studied for a small concentration of randomly substituted defects. The fermion mapping is used and the development is in terms of alloy scattering theory. Three approximations are used. At the lowest level, the ATA reproduces the results of the superlattice model. The CPA is equivalent to the annealed model. An adequate treatment of the quenched model requires the leading contributions from all cross-line diagrams to be included. This is accomplished by parquet theory, and the scattering from three channels is evaluated. Cancellation between two of the channels leaves the third channel as the only contributor. Three conflicting estimates of the specific heat behaviour of the quenched model have appeared in the literature recently. The results of this paper agree with those of Dotsenko and Dotsenko (1982), namely ln(-ln mod T-T_c mod ).

AC susceptibility and ⁵⁷Fe Mossbauer spectroscopy measurements are reported for Fe²⁺ aluminosilicate glasses. The frequency dependence of T_f (over a range from 10 to 10⁸ Hz) is described by a Fulcher law. This indicates a progressive spin freezing at T_f, comparable with results reported in Co and Mn aluminosilicate glasses. At T<<T_f the Fe²⁺ glasses order speromagnetically, without angular correlation between Fe²⁺ moments and local-field-gradient axes. A linear correlation between isomer shift and quadrupole splitting distributions is established. This is tentatively attributed to a changing covalency of bonding of six-coordinated Fe²⁺ ions. The wide distribution of quadrupole splitting cannot be interpreted unequivocally but indicates a large distribution of crystal-field parameters.

A simplified version of the exact curved-wave theory of EXAFS is described. The simplifications arise from the angle-averaging process which is necessary when analysing data from polycrystalline or amorphous samples. Such an angle-averaging process is here performed analytically, resulting in a much simpler theoretical expression for the EXAFS function, but one which still treats the outgoing spherical wave exactly. Test runs show that such a theory is about a factor of forty faster in computation time than the original form of the curved-wave theory. Such rapid computation allows the authors routinely to use the curved-wave theory in structural analyses, rather than the plane-wave approximation which is limited to high photoelectron energies.

The static base approach (SBA) and adiabatic base approach (ABA) are analysed for an electronic two-level system coupled to a single vibrational mode. In the ABA due account is taken of the non-adiabatic contribution to the oscillatory potential. This permits the introduction of a suitable model potential whose eigensolutions can be calculated exactly. It is proven that the calculation of the non-adiabatic transition matrix elements by means of these model eigenfunctions yields the correct first-order result in the promoting-mode coupling constant Delta and thus provides the same accuracy as the SBA calculation. The analysis evinces the importance of the diagonal non-adiabaticity operator, which has been neglected in earlier work. The transition matrix elements of the two approaches are compared. It is found that the absolute value of the SBA result is always larger than that of the ABA result and rises up to about a factor of two times the latter for higher Huang-Rhys factors. The authors finding resolves an old controversy.

The surface structure of Al(110) is determined by means of r-factor comparisons of new measurements at 100K of LEED intensity-energy spectra with spectra calculated by multiple-scattering theory. The effect of correlations between structural and non-structural variables on the process of structure determination is analysed in detail. The surface structure is found to exhibit a damped, oscillatory relaxation of the interlayer spacings. The relaxations of the first three interlayer spacings, expressed as deviations from the bulk value, are determined to be Delta d₁=-8.6%, Delta d₂=+5.0% and Delta d₃=-1.6%. These results are discussed in the context of previous LEED studies of Al(110) and in the light of recent model calculations of surface structure.