Table of contents

A positron annihilation of Doppler broadening in crystalline, quasicrystalline and amorphous Al-Mn-Si alloys is presented. The density of trapping sites is found to be similar in the crystalline and quasicrystalline alloys but much greater in the amorphous alloy.

EXAFS measurements have been made through the melting point up to 291 degrees C using a Lytle type cell with a 25- mu m thick tin sample. The best agreement with the liquid state results is found with a model that retains the 4-2 coordination of the white tin structure.

X-ray and neutron diffraction measurements have been made on three NiTa metallic glasses, Ni₅₅Ta₄₅, Ni₆₂Ta₃₈ and Ni₆₇Ta₃₃. The total structure factors have been derived and the curves obtained with neutrons found to be particularly well defined. The total radial distribution functions (RDF) obtained are discussed in relation to models of close packing and to the stable crystalline phases in the NiTa alloy system. A derivation of the partial structure factors has been made for the Ni₅₅Ta₄₅ glass by an approximate method. The results of this suggest there is a progressively broadened distribution through the Ni-Ni, Ni-Ta and Ta-Ta first-neighbour distances which allows the form of the first peak in the total RDF to be understood. The broadening and expansion of particularly the Ta-Ta first-neighbour distances in the glassy state is consistent with the kind of local order which exists in the Ni₂Ta and mu -NiTa crystalline phases. The possible identification of glassy NiTa alloys as disordered Frank-Kaspar phases, or alternatively as structurally relaxed 'soft sphere' assemblies, is discussed.

Neutron scattering determination of the total structure factors of the amorphous Tb_xSi_1-x system has been performed for three compositions in order to determine the atomic arrangement and the evolution of the local order as a function of the silicon content. Comparison has been made with a conventional structural model consisting of a binary mixture of hard spheres of different diameters. Total pair correlation functions have been determined. The authors propose a model for the local order around both the terbium and the silicon atoms which is in reasonable agreement with the physical properties of the system.

H.-J. Rockosch (1983) has recently obtained a set of vacancy-solute binding energies for seven Cu-based and five Ag-based alloys from an analysis of diffusion data. The author investigates the extent to which these binding energies correlate with (i) the difference between solute and host valences Delta Z, (ii) the slope of the solidus curve, dT_m/dC_s, and (iii) the lattice constant change da/dC_s. D. Lazarus (1954) and A.D. LeClaire (1978) have previously developed a model relating binding energies to Delta Z; in this work, the author presents new models that predict correlations (ii) and (iii). Specifically, the author derives the approximate relations E_vs=-(E_1v/z)(dT_m/dC_s) (where E_1v is the monovacancy formation energy in the pure solvent and z is the coordination number), and E_vs=-(9K Omega /z)(d ln n_at(R₁)/d ln R₁)^-1 d ln a/dC_s (where K is the bulk modulus, n_at is the atomic charge density of host atoms, R₁ is the nearest-neighbour separation and Omega is the atomic volume). The correlation between the Rockosch binding energies and either Delta Z or dT_m/dC_s is only fair; the correlation with da/dC_s, however, is quite close.

Two aspects of point defect-dislocation interactions, peaking effect and kinetics following pulsed electron and neutron irradiations have been studied in high-purity polycrystalline copper. In all experiments, an initial very rapid rise in the internal friction and Young's modulus was observed, followed by a slower continuous change. The rise of the internal friction shows the presence of the peaking effect. The magnitude of the rapid change in the modulus was proportional to the fluence. The initial rise was too fast for detailed kinetics analysis; however, the slower process was analysed. Results are interpreted in terms of a model wherein interstitials are rapidly deleted from the lattice and responsible for the initial rise, while the slower process is due to vacancy diffusion to dislocations. These data also show that interstitials are responsible for the peaking effect. Models proposed to explain the peaking effect are discussed in the light of these results.

The authors report non-equilibrium molecular dynamics simulations of liquid rubidium using the isokinetic sllod algorithm for the dynamics and the Price potential for the interactions. The three state points simulated are the same three chosen by Bansal and Bruns (1984) who used a quite different non-equilibrium method. However, the authors' results differ substantially from theirs. Specifically, they find that the Price potential consistently predicts better results for the shear viscosity. A more fundamental disagreement between the authors' results and theirs is that the authors' results underestimate the experimental data where theirs overestimate and vice versa. This suggests that the method used by Bansal and Bruns has systematic errors.

The consistent treatment of the static and dynamic properties of metallic systems using interatomic potentials which are based on a one-body potential plus pair interactions containing volume or mean density dependence is examined. The focus is on the discrepancy between the compressibility of metals calculated in two alternative and legitimate ways, i.e. from elastic constants deduced from the long-wavelength limit of phonon spectra and from a homogeneous deformation. A simple, physically motivated resolution of this long-standing 'compressibility problem' is proposed which for simple metals provides an interpretation of earlier treatments involving third- and fourth-order perturbation terms in the total energy. The resolution hinges on taking reasonable account of inhomogeneities in the electron density which modulate the one-body and pair potentials. Subtleties in the interplay between these two terms in the particular case of interatomic potentials based on weak pseudopotential theory are exposed and pitfalls are pointed out in the haphazard application of empirical interatomic potentials. The consequences of the work for the phonon spectra are investigated and a first-principles method for treating this and related situations is outlined.

The solution of the helicon problem in a sphere held in a mounting of arbitrary rigidity and damping is calculated. The results show that measurement of R_H may be significantly affected by the sample mounting, but measurement of the magnetoresistance is little affected.

New experiments are reported which investigate the field dependence and anisotropy of the Hall coefficient and magnetoresistivity in single crystals of potassium. These experiments shed light on whether the electronic ground state of potassium is a charge density wave (CDW). They comprise the first independent test of the high field structure in the orientation dependence of the magnetoresistance observed by Coulter and Datars (1980) using an induced torque experiment. At similar levels of sensitivity the authors find no evidence of open orbits in potassium. This result casts doubt on the CDW hypothesis.

It is shown that photon energy dependent photoemission, using a synchrotron radiation source, is a useful tool in studying the electronic structure of alloys. In AuNi a satellite structure at 1.9 eV is identified using a combination of cross-section effects and resonant photoemission. A model calculation is used to show that the authors' observations are consistent with the magnitude of parameters describing the electronic behaviour in terms of correlation and hybridisation, obtained from other techniques.

Low-temperature specific heats and electrical resistivities were measured on 12 samples in two series of Ca_0.7Mg_0.3-xAl_x and Ca_0.6Mg_0.4-xAl_x metallic glasses over the temperature ranges 1.5-6 and 2-300 K respectively. The Hall coefficient was also measured for all Ca_0.7Mg_0.3-xAl_x metallic glasses in the range 77-300 K. A substantial departure from the free electron behaviour is deduced from the composition dependence of the electronic specific heat coefficient. The electrical resistivity is found to vary widely from 50 to 400 mu Omega cm as Mg is replaced by Al. The temperature dependence of resistivity in the low resistivity regime ( rho <150 mu Omega cm) can be explained in the generalized Faber-Ziman theory. The electron transport in the high-resistivity regime (200-400 mu Omega cm) is found to be still strongly composition- and temperature-dependent. These unique features are, the authors believe, far beyond the scope of the generalised Faber-Ziman theory.

Magnetic and one-electron properties of Ni(100) at finite temperatures are studied by the author's recently developed calculation method in which effects of spin fluctuations on the first ten layers of the semi-infinite FCC lattice, described by the realistic canonical d-band model, are included by means of the functional-integral method within the static and single-site coherent potential approximations. It is shown that the surface moment, whose ground-state value is slightly enhanced compared with the bulk one, decreases more rapidly than the bulk upon heating. The calculation of spectral densities indicates that their exchange-split peaks at all the investigated wave-vectors in the surface Brillouin zone approach each other with broadened linewidth when the temperature is raised. The recent ARPE experiments probing the bulk band structure are discussed on the basis of the calculated surface and bulk spectral densities.

The theory of the electron paramagnetic resonance (EPR) on magnetic impurities in type-II superconductors is presented. The kinetic equations describing the coupled motion of the magnetisation of localised spins and that of conduction electrons' spins are derived and solved. Analytical expressions for the relaxation rates, being the real parts of kinetic coefficients entering the equations, are obtained. These expressions are valid at arbitrary temperatures and frequencies of resonance within the approximation of the homogeneous superconductor. The detailed numerical analysis, allowing the authors to plot the temperature dependence of the EPR linewidth, is carried out. It is shown that the 'electronic bottleneck effect' becomes stronger in the superconducting state and, as a result, the linewidth becomes a monotonic function of temperature. The EPR experiments on superconducting systems are discussed.

The theory of EPR in type-II superconductors, developed previously (see ibid., vol.17, p.695, 1987), is extended to include the inhomogeneous broadening and hyperfine structure effects. The appropriate kinetic equations are obtained and solved describing the coupled motion of transverse magnetisation of conduction electrons and transverse magnetisation of magnetic impurities, subject to the action of static spatially random local fields. The solution is analysed analytically and numerically for specific local-field distribution functions. It is shown that even under very soft 'bottleneck' conditions the EPR line exhibits a sharp narrowing just below the superconducting transition temperature T_c. This narrowing occurs primarily because of the decreasing of the inhomogeneous linewidth due to 'superconducting' enhancement of the Korringa relaxation rate, which governs the degree of the narrowing under the bottleneck conditions. It is demonstrated how the exponential decrease of the relaxation rates at temperatures T<<T_c can be used to resolve the hyperfine structure and extract the splitting constant. The connection with experiment is discussed.

The magnetic phase diagram of Cr-Mn alloys determined from measurements of the thermal expansion and bulk moduli is reported for Mn concentrations close to the triple point, which was found to occur at c=0.20+or-0.02 at.% Mn and T=330+or-15 K. The concentration dependence of the boundary lines separating the various phases agrees well with theory.

The authors have measured the AC susceptibility of nine PdGd alloys, varying in concentration from 2 to 10 at.% Gd, between 1.5 and 10K in 15 or more static biasing fields up to 0.1 T. These data were analysed in terms of the predictions of a scaling law equation of state in an attempt to characterise the nature of the ordered ground state of this system.

Al-Mn quasicrystals of several kinds of compositions were prepared by melt-quenching in order to investigate the magnetic properties. The curves of magnetic susceptibility against temperature were obtained and the effective magnetic moment P_eff was deduced from the Curie constant. With increasing Mn content from 14.3 to 22.5%, P_eff increases from 0.56 to 1.53 mu _B. The increment of P_eff becomes remarkable with the increase of Mn content. The spin glass behaviour, namely the cusp in the AC susceptibility curve, was observed above 20% Mn. The spin glass freezing temperature T_f increases from 3.3 to 8.5 K with increasing Mn content from 20.0 to 22.5% Mn. The magnetic moment is drastically reduced and the spin glass behaviour is eliminated, in the crystalline state, by annealing.

Thermomagnetic studies and X-ray analysis made on the amorphous Fe_95-xW₅B_x alloys with x=15, 20 and 25 show the presence of stable alpha -Fe, Fe₂B and FeWB phases. Low-temperature (4.2-300 K) magnetisation measurements indicate the existence of spin waves and Stoner or single-particle excitations. Isochronal and isothermal annealing treatments reveal two stages of relaxation, one of which is attributed to compositional or chemical short-range ordering (CSRO). The value of the critical exponent beta for all the three compositions is determined to be 0.43 by means of scaling law analysis and it is found to be independent of annealing treatments.

The relaxations of structurally sensitive magnetic parameters, in particular those of magnetic permeability, were investigated on a pre-annealed nearly zero-magnetostrictive amorphous (Fe_0.1Ni_0.35Co_0.55)₇₈Si₈B₁₄ alloy at temperatures up to 300 degrees C. The ordinary disaccommodation (DA) and the continuous permeability decay in the total permeability relaxation exhibit a characteristic spectrum with a single relaxation peak. The distributions of the activation energies in the relaxation processes were evaluated. The most probable activation energies for DA and the decay are approximately 1.12 and 1.76 eV, respectively. It was confirmed that the permeability decay by aging is a process of reversible structural rearrangements. The crossover effect in the decay was observed and quantitative estimation made for the magnetic stability at lower temperatures.

The equation of motion approach is employed to derive the two-body Green function, relevant for the calculation of Auger spectra, in the case of systems well described by the Hubbard Hamiltonian with many bands. The results have been used to calculate the M₁VV (M₁ valence-valence) Auger rate of copper, which has been shown to be well described by the present theory, as distinct contributions belonging to different bands are clearly present.