Table of contents

The conduction-band density of states of amorphous (a-) GeSe₂ photo-dissolved by Ag and Cu metals has been examined by means of inverse-photo-emission spectroscopy (IPES). The characteristic feature of the IPES spectrum of a-GeSe₂ rapidly smears out with the photo-dissolution of Ag or Cu metal, and each spectrum of the fully photo-doped samples is quite similar to that of a-GeSe_1.5. This result is consistent with a structural model in which the dopant atoms are surrounded by four Se atoms and many Ge-Ge bonds are induced by photo-dissolution. We also measured the change of the IPES spectra in the process of Ag photo-doping.

The hyperfine-coupled electronic and nuclear moments on the Ho ions in HoF₃ order in a ferrimagnetic structure at T_c=0.53 K. The heat capacity has been measured from above T_c down to 0.09 K. The electronic contributions are vanishingly small below T_c, whereas the nuclear heat capacity increases rapidly with decreasing temperature in the ordered phase, reaching a maximum of about 9 J K^-1 mol^-1 near 0.2 K. The polarization of the nuclear spins and the heat capacity have been calculated. The theory, which includes the modification of the hyperfine field due to the single-site fluctuations of the electronic moments, leads to a very satisfactory account of the heat capacity measurements.

The phase equilibria of ternary mixtures of water, oil and surfactants and binary mixtures of water and surfactants are studied using a triangular lattice-gas model that is an extension of the spin-1 Blume-Capel model with the addition of the orientational degrees of freedom for surfactants. The phase diagram of the model is investigated via Monte Carlo simulations in conjunction with the extrapolation technique of Ferrenberg and Swendsen combined with the finite-size scaling method of Lee and Kosterlitz. Five phases are found for ternary mixtures in the case of equal water and oil concentrations: a water-oil coexistence region, a lamellar phase, a liquid crystalline phase, which has the symmetry of a rhombic lattice, and an extended disordered phase, which is divided by a Lifshitz line into disordered fluid and microemulsion regions. In the limiting case of a binary mixture of water and surfactants, we found a rich phase diagram corresponding to a lamellar phase, a hexagonal phase and a disordered phase. which is again divided by a Lifshitz line into disordered fluid and structured fluid regions. The latter can be considered as the analogue of the microemulsion found for the ternary systems.

For a very short (t_p<0.25 mu s) Joule heating pulse in a metal film heater submerged in liquid ⁴He at low temperatures (T<0.1 K), the phonons created in the liquid separate into distinct low-frequency and high-frequency distributions. Phonons of intermediate frequencies are absent from the signal because they spontaneously decay to lower frequencies. High-frequency phonons above the decay cut-off ( omega > omega _c^{( infinity} )) are stable as T to 0. We find that for a heater power of 10 mW mm^-2, these high-frequency phonons are not distributed over the available frequency range up to the maxon, but are concentrated in a narrow band around omega _c^{( infinity} ). Times of flight show that the distribution of high-frequency phonons depends upon pressure P such that the peak is always at the pressure dependent decay cut-off omega _c^{( infinity} )(P). We suggest that the majority of these detected high-frequency phonons are produced in the liquid ⁴He by frequency up scattering processes amongst the injected phonons.

For part I see ibid, vol.6, p.2813 (1994). We have measured the angular distribution of phonons emitted by an Au film heater in liquid ⁴He. For long heater pulses (10 mu s), the distribution is almost Lambertian ( alpha cos theta ), but for short pulses (0.1 mu s), the distribution is very anisotropic. The HWHM Of the beam of high-frequency (h(cross) omega /k_B>10 K) phonons is approximately 3.5 degrees , and the HWHM Of the low-frequency (h(cross) omega /k_B<1 K) phonon beam is approximately 10.5 degrees . We suggest that the angular distribution of the low-frequency phonons is due initially to classical phonon transmission at the heater/⁴He interface, but is then broadened by decays and interactions in the liquid. These interactions in the liquid create the high-frequency phonons. The high-frequency phonon beam is narrower than the low-frequency phonon beam due to momentum conservation in the four-phonon process (4PP) interactions that create them. Narrow beams of phonon emission from cleaved crystal surfaces have been observed before, but we believe that this is the first time that similarly narrow beams from evaporated metal surfaces have been observed.

Using the combined results of extensive MD simulation and neutron scattering studies of liquid lead at 623 and 1170 K, details of the self-diffusion process within the atomic domain are studied. From analysis of F_s(q,t) it is found that there is an element of structural relaxation in the self-diffusion process which may lead to a jump diffusion component. Also from the derivation of memory functions Gamma _s(Q,t), particularly for the velocity autocorrelation Gamma _s(t), it is found that the MD results are not well described by present mode-coupling theories. It is believed that a more sophisticated theory which does not separate the characteristic parts of Gamma _s(t) and takes many-body effects into account more properly is needed to explain the observations.

A noticeable change in the n=1 exciton reflection band of KI and RbI crystals exposed to air was revealed after cleavage in liquid helium. This change is explained by disturbance of the near-surface layer by charged surface dislocations and other defects, which appeared after cleavage at a low temperature. A moderate irradiation by photons in the exciton region and/or heating leads to the removal of the disturbance and, correspondingly, to relaxation of the crystal surface.

We present an atomistic simulation of the structure of liquid Al_1-xMn_x alloys based on interatomic forces derived from quantum theory. Using the intra-atomic potentials calculated within the tight-binding bond approach, we study the structural properties of liquid Al₈₀Mn₂₀ and Al₆₀Mn₄₀ alloys using molecular dynamics simulations. The results are in good agreement with the most accurate diffraction data. The correlation of the structural trend with the characteristic variation in the electronic structure is established.

The electronic structure of Ni₂Ti₄O is studied using first-principles calculations based on the linear-muffin-tin-orbital method. The role of oxygen in stabilizing the metallic NiTi₂-type structure is analysed by site-projected energy calculations which indicate that the Ti(f) sites with nonicosahedral symmetry determine the structural stability of Ni₂Ti₄O. The origin of this behaviour, contrary to that found for NiTi₂, is due to the effects of 2p(O)-3d(Ti(f)) hybridization.

The source of random static electrical fields, which is determined by both the piezoelectric coupling and the static elastic deformations induced by substitutional disorder of ions with different ionic radii has been considered for mixed crystals of the KH₂PO₄ family. It is shown that the random electrical fields act mainly on the lattice polarization along the c axis and up-down proton configurations. The short-range interactions of four acid protons surrounding a PO₄ group are represented as the random local uniaxial anisotropy for the mixed crystals provided the ice rule condition is fulfilled. It is also shown that the distinction of symmetry between the up-down and lateral proton configuration, and the peculiarities of proton-lattice interaction can lead to the coexistence of long-range order (for up-down/lateral configuration) with the glass or paraelectric (for lateral/up-down configuration) state.

The motivation of this work is to search for an elementary common physical mechanism for the non-exponential decay observed in a large number of disparate phenomena. Probably the most elementary assumption is an initial Poissonian distribution of activation energies E with each fraction decaying independently according to an Arrhenius rate r(E) approximately exp(-E/T). Despite this drastic oversimplification, this elementary model will account for a large variety of observed shapes of decay, ranging from nearly exponential to power law and nearly logarithmic functions. The fundamental feature that governs the shape is the ratio E/T, denoted 1/b, of the mean E at temperature T. Analysing limited experimental data sets is also easy due to the closed form of the resulting normalized decay curves g(b, tau )=b tau ^-b gamma (b,tau/), where tau =r₀t is the time t normalized by the rate r₀. The incomplete gamma function gamma (b, tau )= integral ₀^tau s^b-1e^-sds is an important correction of the power law tau ^-b In an alternative interpretation, this decay form g(b, tau ) could also be considered as generated by a b-dependent form of an effective barrier energy U_eff(g) increasing as a function of the decaying normalized observable g. In this paper the forms g(b, tau ) and U_eff(g) are evaluated, displayed and compared with In a separate paper the widespread applicability of the form g(b,tau/) will be demonstrated by interpreting various types of measured data using individual established temperature dependencies and by treating the 'ageing' or 'memory' effect.

For part I see ibid, vol.6, p.2893 (1994). The 'elementary decay model' (EDM), which allows an interpretation of most nonexponential decay functions, is applied to representative published decay curves measured in superconductors and in a metallic spin glass. Features attributed previously to different regimes connected by crossover regions can be interpreted by monotonically changing parameters. For high-T_c superconductors I-V curves are also analysed with the adequately transformed EDM. In addition, the excess conductivity in the paraconductive region above T_c is well modelled by an Arrhenius excitation, which also extends below T_c in multilayers of YBCO. Measurements previously interpreted by U_eff varies as J^{- mu} can also be interpreted by the EDM, which is close to U_eff varies as -ln J evaluated by Blatter et al. (1990). The rapid drop in resistivity close to a transition into a vortex-glass state measured in the pV range by Gammel et al. (1991) in YBCO could also be fitted by the EDM. Further, the 'ageing' or 'memory' effect in a metallic spin glass is well interpreted by the EDM. In particular, the 'memory' effect is shown to be a consequence of plotting nonexponential decay data against the logarithm of a retarded time and is therefore not a unique indication for a spin glass.

Magnetotransport measurements have been carried out on the organic superconductor kappa -(BEDT-TTF)₂Cu(NCS)₂ at temperatures down to 500 mK and in hydrostatic pressures up to 16.3 kbar. The observation of Shubnikov-de Haas and magnetic breakdown oscillations has allowed the pressure dependences of the area of the closed pocket of the Fermi surface and the carrier effective masses to be deduced and compared with simultaneous measurements of the superconducting critical temperature T_c. The effective mass measured by the temperature dependence of the Shubnikov-de Haas oscillations is found to fall rapidly with increasing pressure up to a critical pressure P_c approximately=5 kbar. Above P_c a much weaker pressure dependence is observed; T_c also falls rapidly with pressure from 10.4 K at ambient pressure to zero at around P_c. This strongly suggests that the enhanced effective mass and the superconducting behaviour are directly connected in this organic superconductor. A simplified model of the band structure of kappa -(BEDT-TTF)₂Cu(NCS)₂ has been used to derive the bare band masses of the electrons from optical data. Comparisons of these parameters with cyclotron resonance data and the effective masses measured in the present experiments indicate that the greater part of the enhancement of the effective mass necessary for superconductivity in this material is due to quasiparticle interactions, with the electron-phonon interactions playing a secondary role. The dependence of T_c on the effective mass may be fitted satisfactorily to a suitably parametrized weak-coupling BCS expression, although this cannot be taken as a definitive proof of the nature of superconductivity in organic conductors.

The quantum and classical transitions of the Neel vector in antiferromagnets are considered. Expressions for the tunnelling and thermal activation rates of the transitions are given for low temperatures, near the cross-over temperature and for high temperatures, so that they could be checked experimentally. Quantum and thermally activated nucleation in bulk solids and two-dimensional films of antiferromagnets is also presented.

Hydrogen diffusion in ZrH_x (1.58<or=x<or=1.98) in the temperature range 600 K to 970 K has been measured by means of pulsed-field-gradient nuclear magnetic resonance. The activation enthalpy for hydrogen diffusion, H_a, obtained by fitting an Arrhenius expression D=D₀exp(-H_a/k_BT) to the diffusivities, increases sharply as x approaches the limiting value of two in good agreement with results deduced from the proton spin-lattice relaxation rate, Gamma ₁, measured on the same samples. Hydrogen atoms jump predominantly between nearest-neighbour tetrahedral sites in ZrH_x. The observed concentration dependence of both the effective value of H,and the pre-exponential factor D₀ suggests, however, that at high hydrogen concentrations and high temperatures another interstitial site is occupied in addition. At x to 2 a small fraction of hydrogen atoms located on an interstitial site other than the tetrahedral site appears to contribute significantly to the diffusivity. The temperature and concentration dependence of the diffusion data can quantitatively be described by such a model. The activation enthalpies for all possible jumps in this system with two different kinds of site are shown to be independent of the hydrogen concentration x. The corresponding attempt frequencies nu _a approximately=10¹³ s^-1 are compatible with the picture of a classical diffusion mechanism.

The two complementary techniques of positron lifetime spectroscopy and x-ray diffraction have been applied to the structural characterization of commercial titanium nitride hard coatings grown by hollow cathode arc evaporation. Use of a pulsed positron beam allows the layers to be analysed more or less independently of the substrate. The layers are found to consist of a single delta -phase TiN, with pronounced texture and residual stress. The dominant defects appear to be grain boundaries with a positron lifetime of 225 ps. Values for the bulk positron lifetime and diffusion coefficient are also obtained.

The multiple-scattering cluster method was first employed to calculate the theoretical near-edge X-ray absorption fine structure (NEXAFS) spectrum of graphite. Studies on the polarization dependence on the incident vector E classify spectral features into pi and sigma resonances. Characterization of the carbon K-shell NEXAFS spectra of the species modelled by both a single layer and a multilayer has been performed. It reveals that a pi * resonance originally located at around 989 eV splits with the accumulation of the number of atoms included in the cluster. Multilayer calculation has little influence on the NEXAFS spectrum. Our calculation coincides with the experimental results fairly well. In addition, it has been found to be in good agreement with the final density of states derived from band-structure calculations.

Based on the two-conduction-band periodic Anderson model, including the effects of Kondo holes and conduction band impurities, the residual magnetoresistivities (MRS) of heavy-fermion (HF) alloys have been calculated in the single-site coherent-potential approximation. The strong correlation of HF systems is treated by the slave-boson technique. Our results show that only negative residual MRS can be obtained in 'pure' Kondo-hole alloys and only positive residual MRS can be obtained in 'pure' conduction band impurity alloys. In realistic alloy systems, competition exists between the above two kinds of disorder, which may give rise to a transition in the residual MR from positive to negative with increase in the Kondo-hole concentration.

We report a revision of the rigid-ion potential for UO₂ introduced by Sindzingre and Gillan (see ibid., vol. 21, p. 4017, 1988). In its original form, the potential is unstable at the reported lattice parameter. We have re-fitted the potential to remove this problem, whilst keeping as close as possible to the original parametrization. The calculated static quantities are little changed, but there is a small increase in the O Frenkel pair formation energy. In dynamics both the superionic transition temperature and the rate of O diffusion in the superionic state are altered, but there is no effect on the calculated thermal conductivity. The conclusions of previous work with the original potential are not affected by the changes we propose.

A methodology for investigating dynamics of various spin probes in mixed/glassy systems of the KH₂PO₄ type is developed. Using this methodology, electron spin resonance (ESR) features of the AsO₄^4- centre in the paraelectric phase of Rb_1-x(NH₄)_xH₂PO₄ are shown to arise from strain-broadening and motional-broadening effects. These conclusions are in disagreement with those of a recent study, which implicated the various features observed in the existence of ferroelectric-antiferroelectric fluctuations in Rb_1-x(NH₄)_xH₂PO₄. It is additionally shown that hyperfine couplings from Rb isotopes make the dominant contribution to the ESR spectra in Rb-containing compounds.

By taking ⁷⁵As hyperfine coupling constants for AsO₄^4- in hydrogen bonded crystals of RbH₂PO₄ and (NH₄)H₂PO₄ as fingerprints of ferroelectric (FE) and antiferroelectric (AFE) phases respectively, it is shown that the variation of the coupling constant as a weighted average of Rb⁺ and NH⁴+ concentrations in Rb_1-x(NH₄)_xH₂PO₄ clearly proves the existence of FE and AFE fluctuations in the mixed crystal.