Table of contents

Electron diffraction patterns obtained from coexisting decagonal and icosahedral phases in Al₄Mn have confirmed that the decagonal phase is related to the icosahedral phase by a set of mirror planes intersecting a fivefold axis. The tenfold axis is shortened by about 5.1%. The misfit between the tenfold and fivefold axis is accommodated by discommensurations across the grain boundary between the decagonal and icosahedral phases. The idea that the diffuse streaking is due to the packing of rods has been supported by high-resolution electron microscopy.

The decomposition of CuCo alloys with Co contents of 0.5, 0.8 and 2.0 at.% was investigated by small-angle neutron scattering with the aim of determining the critical nuclei size of precipitates of the Co-rich alpha phase. The measurements reveal finite incubation periods for the 0.5 and 0.8 at.% Co alloys, and the smallest precipitate radii between 0.48 nm (2 at.% Co) and 4 nm (0.5 at.% Co). For states with precipitate radii around 0.5 nm the scattering curves suggest the presence of disc- or rod-shaped precipitates as well as spherical precipitates.

The plasmon spectra of Ti and TiH₂ have been measured using high-energy electron energy loss spectroscopy. The measured ratio of the TiH₂ plasmon energy to the Ti plasmon energy is compared with the ratio calculated for anionic, atomic and protonic models within the limits of the free-electron approximation. The best agreement between theory and experiment is obtained using the protonic model. The remaining difference may be attributed to the simplicity of the models employed, although the instability of the sample under electron beam irradiation may also be significant.

The problems associated with assigning a contact temperature to the metallic point contacts, used in microcontact spectroscopy, under bias conditions are discussed. It is demonstrated that the Kohlrausch relation can be solved numerically to obtain an estimate of the maximum contact temperature if the electron diffusion length is smaller than the effective contact diameter. This procedure allows a more accurate estimate of the maximum contact temperature to be made when the contact is in the Maxwell limit and it eliminates the need to use the Wiedemann-Franz relationship. The measured differential resistance curves of high purity Gd point contacts are compared with theoretical predictions of the contact temperature.

An iterative solution of the magnetoelastic equation of magnetostriction in amorphous and polycrystalline ferromagnets is given. In zeroth order (no fluctuations delta C of the elastic tensor) magnetostriction is totally determined by the strain derivative of the local anisotropy ('conventional' mechanism), and the effective tensor is given by a simple geometrical average over all local values. Inclusion of spatial fluctuations delta C yields modifications of the effective magnetostriction tensor and an additional magnetostriction mechanism ('reorientation' mechanism) characteristic of amorphous and polycrystalline materials.

The properties of single and multiple defects in Ni were calculated by molecular dynamics simulation using an effective interaction potential derived with the aid of a new method developed by Dagens for transition metals. The formation and migration energies of single vacancies are 1.33 and 0.90 eV, respectively. Di-, tri- and tetravacancies are weakly bound, with respective binding energies of 0.004, 0.10 and 0.24 eV. The divacancies are more mobile than the single vacancies, having a migration energy of 0.47 eV. Self-interstitials are found to be stable in the (100) dumbbell configuration; their energies of formation and migration are 4.16 and 0.12 eV, respectively. Multiple interstitials are strongly bound; like the single dumbbell interstitials, di-interstitials are highly mobile. The respective binding and migration energies for di- and tri-interstitials are 1.13 and 0.17 eV, and 2.67 and approximately 0.50 eV. The theoretical results are compared with the available experimental measurements and with previous defect calculations based on a short-range empirical potential.

The molar volumes V_M and the thermal expansion coefficients alpha _p of liquid Hg-In and Hg-Sn alloys were measured as detailed functions of composition in the range of dilute solute (In or Sn) concentration (below 15 at.% In and 10 at.% Sn). The excess molar volume Delta V_M for each kind of alloy decreases monotonically with increasing concentration of solute (In or Sn) atoms. On the other hand, the concentration dependences of alpha _p differ; a slight hump exists at a few at.% In for liquid Hg-In alloys, while no anomalous behaviour of alpha _p is found for liquid Hg-Sn alloys. These results are closely related to the fact that there is a minimum in the curve of the thermoelectric power Q plotted against concentration for Hg-In but not for Hg-Sn, and appear to show the importance of higher-order correlation effects in liquid Hg-In alloys. It is also pointed out that similar anomalies can be found in the literature values of the partial molar excess entropy of solute component in liquid Hg-In and Hg-Tl systems.

Measurements of a set of third-order elastic constants of a BCC Cu-Zn-Al alloy as functions of temperature allow the study of the combined effect of temperature and strain on the elastic constants. The influence of a shear strain on the lattice stability is examined at different temperatures and the results are integrated in the localised soft-mode theory of martensite. The influence of a uniaxial compression on the shear constants C' and C₄₄ and on the elastic anisotropy of the BCC matrix is calculated and it is shown that at the critical conditions for stress-induced martensite formation, the resistance against a (011)(011)-type shear decreases and the anisotropy increases with increasing temperature. This change in anisotropy of the matrix can influence the local stability around defects explaining in a qualitative way the occurrence of stress-induced martensite.

The renormalisation of the hydrogen transfer integrals in niobium and tantalum due to the displacement clouds dressing the particles is investigated quantitatively. An auxiliary matrix tau (T) is introduced and values of its elements are presented, which permits evaluation of the renormalisation parameter S(T) once the self-trapping displacements around the hydrogen impurities are known. Estimates are given of S(T) for self-trapped H (and D) in Nb and Ta. A microscopic model is described for treating H trapped at O impurities in Nb and the renormalisation parameter is extracted for this case.

In the structural map for dialuminides based on Miedema parameters, the compound ThAl₂, which crystallises in the AlB₂ structure at ambient conditions, falls in the region of Laves phase (MgCu₂) compounds. The authors have shown that only moderate pressure is required to transform it to the structure predicted by the structural map.

The authors present preliminary studies of the thermodynamic properties of liquid transition metals at their melting point. These have been studied by using a semi-empirical tight-binding model to describe the interatomic potential energy and with the fluid aspects modelled through a system of hard spheres optimised via the Gibbs-Bogoliubov variational scheme. Results are presented for the free energies and the equilibrium densities which exhibit trends in accordance with experiment. They have also calculated the coefficients of thermal expansion which are found to agree well with the available experimental data. The work reveals shortcomings in the model adopted for the interatomic potentials that were not evident in earlier applications to solid metals.

The authors carry out an analytic investigation of the form of the interatomic potential Phi (R) for (s,p)-bonded metals. They use a simple Ashcroft pseudopotential concentrating both on the shape of short-range screened repulsion and on the form of the Friedel oscillations. In particular, they consider their combined roles around the nearest-neighbour distance. The interaction in q space is fitted by simple but specially chosen functions to include correctly the important physical effects. The short-range repulsion has roughly the form of a simple exponential screening with a screening radius determined by the compressibility sum rule. However, this approximation is not adequate for most purposes, the real interaction cutting off much more sharply with a small overshoot, determined by momentum transfers up to about k_F. The authors' form includes the logarithmic singularity in the susceptibility chi (q), at 2k_F precisely, giving the Friedel oscillations over the wide range of distances including various phase-shift effects. This now explains physically all the features found previously in the numerically computed interactions for different pseudopotentials, electron densities and local field corrections.

De Haas-van Alphen measurements have been made of the angular distribution of sub-grains produced in dislocated high-purity copper single crystals bent about a (111) axis. The results are compared with models in which misorientations are formed by dislocations produced in various (121) directions depending on the resolved shear stress due to bending. The method has been shown to be an accurate probe of the distribution of misorientations in bent crystals, and is capable of measuring the components of misorientations in any direction by suitable choice of the crystal zone in which the direction of the magnetic field is varied. However, some discrepancies exist between the predicted values of the misorientations and the experimental data which are thought to be due to the inherent unrepeatability of producing dislocations, and hence misorientations, by bending. Reasons for the unrepeatability are discussed.

The authors calculate the Compton profiles of Ti and TiH₂, due to their valence electrons, along the (100), (110) and (111) directions, based on self-consistent augmented plane-wave functions. They present the main theoretical points of the method and the Compton profiles of Ti (FCC), Ti (BCC) and TiH₂ (CaF₂ structure), as well as their directional anisotropies. The average Compton profile of Ti (FCC) is in good agreement with available experiments for Ti (HCP). The introduction of hydrogen in forming TiH₂ causes remarkable changes in the Compton profiles.

In the intermetallic compound Fe₃Si there are two non-equivalent Fe sites, one (C) with cubic symmetry and another (T) with tetrahedral symmetry. Impurities from the first transition row show a marked tendency at thermal equilibrium to occupy one or other of the two types of sites preferentially; impurities from the left of Fe such as Mn preferring the C site and impurities from the right of Fe such as Co preferring the T site. In this paper the authors calculate (i) the band structure Fe₃Si, (ii) the local densities of states of the pure system for various sites and spin orientations from Haydock's method and (iii) similar local densities of states of the systems with various impurities in both possible positions. One-electron total energies are compared and they find that the calculations give the correct site preferences in terms of them. A simple physical interpretation of the results is proposed.

The influence of metal vacancies on the electronic structure of zirconium nitride is studied by means of the KKR-CPA method. Besides lowering the Fermi energy, the vacancies do not change profoundly the density of states. The contribution of the vacancy subcell to the density of states is analysed in terms of the resonance condition for virtual bound states. In particular it is found that contrary to the case of vacancies on the nonmetal sublattice, where a sharp a_1g-like virtual bound state was found, for metal vacancies there is only a somewhat weak e_g-like virtual bound state, but no t_2g-like peak. In the case of vacancies on both sublattices, the a_1g-like virtual bound state is present, but the e_g-like peak vanishes almost completely.

The energy band structures for the four compounds CeRu₂, CeRh₂, CeCo₂ and LaRu₂ in the cubic Laves phase are calculated by the self-consistent APW method. Wide 4f band are obtained in the Ce compounds, and the numbers of f electrons in them are nearly one. The Fermi surfaces have complicated forms in these compounds. Although the average contribution of the 4f components over the Fermi surface in CeRu₂ is 28.8%, the character of the states is not uniform. The average weight of 4f components on various cyclotron orbits varies from 5 to 45%.

A wide range of behaviour is observed in uranium materials with stoichiometry UX₃. Many compounds of this type are described well by density functional theory, but several are not. The results of fully relativistic self-consistent density functional calculations using the LMTO method are presented for the materials USi₃, USn₃ and URu₃. In the light of these results the authors discuss the applicability of band theory to these materials and indicate how this correlates with the position of the non-uranium atom in the Periodic Table. The importance of the f band width and the details of the hybridisation between uranium f states and non-f states on neighbouring sites are emphasised.

High precision electrical resistivity measurements were made on potassium samples, including a dilute alloy at temperatures between 0.9 and 4.2K. These samples were annealed and deformed by twisting at or below 4.2K in stages, in order to generate controlled amounts of dislocations. The resulting enhancement of the electron-electron and electron-phonon terms is analysed in the light of recent theories based on the anisotropy of electron-dislocation scattering. Difficulties in interpretation in connection with recent measurements below 1K by another group are discussed. Some metallurgical aspects are considered.

The longitudinal magnetoresistance of three PtMn alloys, containing 0.05, 0.1 and 0.15 at.% Mn respectively, has been measured in fixed fields of approximately 12.5, 25, 50 and 75 kOe between 1.5 and 10K. These data have been compared with a first Born approximation calculation of the resistivity based on an s-d model, thus enabling estimates for the impurity spin (S) and g factor (g_eff) to be made. The authors find an unenhanced spin(S=⁵/₂) and a small positive g shift.

The authors calculate the depression of T_c caused by the effect of a short mean free path on the Coulomb repulsion, for the special case of a very anisotropic superconductor. Specifically, they consider the linear chain model for A15 superconductors with a large difference between the in-chain and inter-chain scattering rates. Previous work based on the McMillan equation for T_c gave a large reduction in T_c. In this work they find T_c numerically from the Eliashberg equations. They find that the reduction of T_c due to this effect is about 25% to 30%, which is considerably less than observed experimentally. The effective Coulomb term is retarded, and as a result is frequency dependent. It is affected by the disorder mainly at small frequencies ( approximately 10 meV).

The surface magnetism of Cr (100) at finite temperatures is discussed by using the single-site spin-fluctuation theory for the semi-infinite BCC lattice described by the realistic tight-binding canonical d-band model. The average and root-mean-square magnetic moments on the first ten layers as well as in the bulk are calculated as functions of the temperature. It is shown that the ferromagnetic surface order, which at T=0K has a magnetic moment of 2.2 mu _B, much larger than the bulk moment of 0.6 mu _B, persists above the bulk Neel temperature in accordance with recent angle-resolved photoemission (ARPE) experiments. Calculations of the spectral density demonstrate that the temperature-induced change in the ARPE spectra depends on the initial-state wavevector in the two-dimensional surface-Brillouin zone.

Mg-Zn metallic glasses, characterised by low crystallisation temperatures, are useful alloys for investigating the room temperature stability of the amorphous metallic state. In this paper the structural stability of Mg-Zn metallic glasses is discussed. As-quenched ribbons and the same ribbons after storage for six years at room temperature were investigated through differential scanning calorimetry (DSC), X-ray diffraction and optical microscopy. The fully amorphous structure was found only for Mg₆₅Zn₃₅ composition. An Mg₆₈Zn₃₂ alloy exhibited an amorphous structure with precipitated Mg on the ribbon surface. Mg₇₅Zn₂₅, Mg₇₂Zn₂₈ and Mg₇₀Zn₃₀ showed a mixture of crystalline phases with a decreasing amount of amorphous phase on the Mg-rich side of the peritectic. The structure of the crystalline phases and the quantity of the remaining amorphous phase was determined. Thermal analysis of aged samples indicated two values of activation energy for crystallisation-lower for the Mg-rich side of the peritectic (including the peritectic composition) and higher for Zn-rich alloys. Additionally, for Zn-rich alloys thermal effects related to the structural relaxation of the amorphous phase were observed.

The low-temperature specific heats in the temperature range 1.5-6K have been measured on four (Ni_1-xCu_x)₈₀P₂₀ (x=0, 0.18, 0.67 and 0.73) and two (Ni_1-xCu_x)₇₇B₁₃Si₁₀ (x=0 and 0.1) pseudobinary metallic glasses. The results can be well fitted to the equation C= gamma T+ alpha T³+A for all alloys studied except for (Ni_0.27Cu_0.73)₈₀P₂₀, for which the equation C= gamma T+ alpha T³+ delta T⁵ is better fitted. The band structure change associated with the progressive filling of the Ni d band is discussed on the basis of the plot of the gamma value against the electron concentration. The d band is found to be essentially filled at the Cu content midway between x=0.18 and 0.67 in the Ni-Cu-P alloy system.

The authors have studied magnetic fluctuations in the paramagnetic phase of rare-earth-aluminium intermetallics of the type (RE)Al₂ by the method of transverse-field muon spin rotation ( mu SR). In the case with strong 4f exchange they could observe a strong increase in the damping of the mu SR signal on approaching the Curie point T_c. One important contribution to the increased damping rate is the slowing down of magnetic fluctuations attributed to the formation of paramagnetic spin correlations which are observed to persist far above T_c. The crystallographic site for the mu ⁺ is discussed as well as the hyperfine field induced at this site due to interstitial electron polarisation.

The optical properties of LaB₆ and CeB₆ have been investigated using spectroscopic ellipsometry. The results were compared with band-structure calculations. The main features of the spectra of both materials originate from optical transitions starting at boron-derived levels of similar character. A sharp peak in the joint density of states for CeB₆ around 0.5 eV is observed.

The Hall coefficient of single-crystal indium was measured in two geometries: B//(100)(R_{H perpendicular to} ) and B//(001)(R_H//) at 4.2K. An anisotropy was observed in the Kohler plot: in intermediate and low fields, R_{H perpendicular to} is more negative than R_H//. Isothermal measurements of R_H at 3.5K and 1.9K are also reported. The effects of the Fermi surface geometry and the wavenumber-dependent electronic relaxation time on the anisotropy are discussed.