Table of contents

The determinations of rho _F (resistivity contribution per Frenkel defect) by damage-rate measurements is discussed critically. An integral equation is derived from which rho _F can be evaluated using the threshold energy surface concept. The inherent approximations in the numerical procedure of King et al. (1981) can be understood within the scope of this analysis.

The magnetic susceptibility, NMR Knight shift and spin-lattice relaxation time of sodium have been measured up to 1000 degrees C. Strong deviations from simple interacting electron-gas behaviour are found above 400 degrees C. The paramagnetism is still further enhanced and the NMR data show that changes in the wavevector dependence of the susceptibility consistent with fluctuations towards a spin density wave occur. Such behaviour is expected for an itinerant antiferromagnet above its ordering temperature as predicted by some recent theories of the metal-nonmetal transition in an expanded fluid metal.

The electronic band structure of Ni₃Al has been calculated using the ASW method, and a high-precision density of states has been derived. Comparison with existing specific-heat measurements data suggests that Ni₃Al remains in the fluctuation regime even in the ferromagnetic state. The authors present calculations of the extremal cross-sectional areas of the Fermi surface and the corresponding orbital masses to stimulate further experiments.

The stability of static martensitic phase boundaries (i.e. interphase interfaces or domain boundaries) in an unbounded crystal is examined in the framework of a one-dimensional Ginzburg-Landau theory. It is shown that at the three-phase coexistence temperature T=0 the kink-shaped boundaries interpolating the shear-strain field between the austenitic and one of the martensitic phases, as well as the phase boundaries interpolating between the martensitic domains in the two-phase sector of the model, are energetically stable. Both the pulse-shaped solitary-wave excitations of the metastable martensitic states in the one-phase sector, as well as the pulse-like excitations of the metastable austenitic state in the two-phase sector are energetically unstable.

The internal friction and the torsion modulus have been measured in an inverted torsion pendulum using quenched beta Cu-Zn and beta Cu-Zn-Al alloys. Peaks observed at -50 degrees C for the B2 ordered alloys and at -15 degrees C for the DO₃ ordered alloys have been associated with (111) nearest neighbour vacancy-Cu atom pairs. Peaks found at 100 degrees C and 200 degrees C have been interpreted as being due to next nearest neighbour Cu pairs on the Zn rich sublattice. From the results a vacancy formation energy of 52 kJ mol^-1 is deduced. Comparisons with existing diffusion data are presented.

The concentrations of hydrogen and deuterium dissolved in Fe_0.5Ti_0.5 in equilibrium with the gas (pressure range, 0-120 bar; temperature range, 400-800K) were determined by electrical resistivity measurements. For both isotopes the electrical resistivity increases by 820 mu Omega cm per unit concentration of x in Fe_0.5Ti_0.5H_x/D_x. Hydrogen and deuterium dissolve endothermically, and Sieverts' law is satisfied up to the highest obtainable concentration (x approximately 0.03). For Fe_0.5Ti_0.5H_x the x-p-T relation (with p in bar, T in K) is log₁₀x=¹/₂log₁₀p-1.86-556/T. The partial molar enthalpies and excess entropies of solution at infinite dilution exhibit only a small isotope dependence. The solvus concentration of the solid solution Fe_0.5Ti_0.5H_x is log₁₀x=1.1-1340/T.

New measurements of the hydrogen chemical potential and the partial enthalpy of hydrogen in Pd-H alloys are reported. These results are combined with information available in the literature to obtain a consistent set of values for the partial excess enthalpy and entropy of hydrogen as a function of composition. The partial excess entropy of hydrogen varies from negative values at low hydrogen concentrations to positive values at high hydrogen concentrations and it is demonstrated that the principal contribution to this excess entropy is configurational. Problems associated with trying to interpret these properties in terms of pairwise interaction models are discussed.

The calculation of the electronic states lifetime and of the enhancement factor of the electronic specific heat related to the electron-electron interaction is presented in the case of the 3d metals V, Cr, Mn and Fe with b.c.c. structure. The average lifetime of the electronic states has been found to be simply related to the interstitial electron density while no simple behaviour has been found for the enhancement factor.

The (100), (110) and (111) directional Compton profiles and their reciprocal form factors (B functions) have been deduced from Compton scattering measurements on single crystal slices of iron with ¹⁹⁸Au 412 keV gamma radiation. The accuracy of the individual profiles is sufficient not only to detect the small asphericity in the electron momentum density but also to discriminate between APW and LCAO band-structure models. In general terms the LCAO calculations with a Gaussian basis describe the momentum density better, although, in common with studies of other transition metals, the asphericity is still overestimated. Inspection of the reciprocal form factor reveals aspects of the density distribution not adequately by either model.

The electronic structures of dilute niobium-based molybdenum alloys have been studied by using the de Haas-van Alphen effect. The rate of increase of the Dingle temperature ( Delta T_D*) on alloying is found to be fairly small compared with that of noble-metal alloys and to be anisotropic over the Fermi surfaces that is, 3K (at.% Mo)^-1 for nu oscillations arising from the ellipsoids, and 7K (at.% Mo)^-1 for alpha and eta oscillations arising from the jungle gym. The change in frequency of the nu ₃₄₅₆ oscillation is about ²/₃ that predicted by the rigid band model. Applying the partial-wave analysis to these experimental facts, the following results are obtained. (i) The d wave phase shift dominates over s, p and f wave phase shifts in this alloy system and the anisotropy in Delta T_D* results from the anisotropy in the amount of d component of the wavefunction over the Fermi surfaces. (ii) The non-rigid band-like behaviour of the nu ₃₄₅₆ oscillation arises from the same origin as Delta T_D*. (iii) The significantly small Delta T_D* found in the present experiment is related to a small electron group velocity on the Fermi surfaces in niobium.

The powerful functional formulation used in quantum field theory and statistical physics is extended to deal with electronic properties of liquid metals and disordered semiconductors. The averaged one-electron Green operator is extended to a functional G mod eta mod of a position-dependent ionic number density eta _R. Functional equations are derived for the purely random case and for the correlated-ion case, within a generalised Kirkwood superposition approximation (GKSA), which treats exactly the ionic correlations of n ions up to n= nu ( nu =2 is the ordinary KSA; nu =3 involves the exact two-ion and three-ion distribution functions, etc). The functional equation for the site-site scattering operator is reduced formally to ordinary integral equations by introducing vertex operators. An approximation is derived which extends the effective medium approximation (EMA) beyond the KSA. The functional formulation provides various exact identities which are used to derive internal consistency relations for approximations such as the EMA and to formulate a general identity related to the non-overlapping condition.

The authors present a first-principles description of the electronic structure in metallic spin-glass alloys. The method is illustrated by explicit calculations for the Ag_cMn_1-c system. In this case it consists of a Korringa-Kohn-Rostoker coherent potential approximation (KKR CPA) calculation for a three-component alloy. Namely, the lattice sites are taken to be occupied by Ag atoms with probability c, Mn atoms with local moments pointing 'up', Mn up arrow , with probability ¹/₂(1-c) and Mn atoms with the opposite orientation, Mn down arrow , with the same probability ¹/₂(1-c). They find that the Mn impurity band is exchange split, as in spin-polarised band theory with aligned spins, and the resulting gain in energy supports the moments. In this sense their theory is a generalisation of the Anderson model of impurity moment formation to finite concentrations. They study the density of states, Bloch spectral functions and the Fermi surface.

The authors analyse in detail the KKR CPA calculations of Pindor et al. (1983) on the electronic structure of a series of non-stoichiometric Pd-rich Pd-Ag hydrides. The authors find the hydrogen-metal bonding band split off the bottom of the metal d bands for all H concentrations. They study the electronic structure at high symmetry points and along the Gamma L direction, and in particular they find two L_2' states. They comment on the relationship of their results to the stability of hydride formation.

A nearly free electron description of the parallel bands responsible for the 1.6 eV absorption edge in aluminium gives rise to an inverse square-root singularity in the joint density of states. The authors have explored the possibility of excitons associated with the corresponding optical transitions. A calculation based on the Bethe-Salpeter equation with a statically screened Coulomb interaction between carriers shows that the singularity in the joint density of states leads to a singlet exciton with a binding energy of only 5 meV. For the case of a triplet exciton they estimate a binding energy of 20 meV. These results are discussed in relation to experiments.

Starting from the realistic wavefunctions of Faust et al. (1976) the authors have calculated the electron self-energy Sigma and corresponding corrections to the band energies of potassium in Hedin's GW approximation. The inverse dielectric function (DEF) was evaluated in the plasmon pole approximation (PPA) using a biquadratic dispersion law. The dynamical effects, i.e. the energy dependence of Sigma in the quasiparticle equation, were taken into account fully self-consistently, showing significant shifts with respect to the usual first-order calculations. They found the expected reduction of self-energy contributions as compared with the static approximation from which they started. Moreover, the core polarisation could be included in a consistent way, yielding better agreement with experimental data.

A previously developed theory of itinerant-electron magnetism at finite temperatures is generalised to include the effect of degenerate multi-bands with t_2g and e_g symmetry by adopting the static functional integral method and the coherent potential approximation (CPA). Numerical calculations of the Curie temperature, the magnetisation curve, the amplitude of local moments and the susceptibility are made for iron by using the detailed density of states generated by the recursion method for the tight-binding model. Calculated results are shown to account for finite-temperature properties of iron. In particular, the calculation explains the observed very weak temperature dependence in the spin-density asphericity, which cannot be explained by the Stoner theory. The authors' results are compared with those obtained by alternative approaches in which the five sub-bands are taken to be equivalent.

A theoretical study of the distribution of directions of magnetisation in ferromagnets with sixfold hexagonal anisotropy is presented. The use of various approximate treatments for ferromagnets with a large axial anisotropy, such as terbium and dysprosium, is discussed; the calculated results are compared with magnetisation and nuclear orientation experiments on polycrystalline and single-crystal samples.

Using a mean-field effective medium approach, the authors study the phase diagram of a random Heisenberg model. In particular they study the phases and magnetisation beyond the percolation limit, and compare these with the experimental observations in AuFe alloys with high concentrations of Fe.

New measurements of the magnetisation of a CuMn spin glass (0.25 at.%) are presented in order to test the hypothesis that the field-cooled magnetisation M(H,T) of a spin glass can be treated by the relationships of equilibrium thermodynamics. Comparisons are made between direct measurements of ( delta M/ delta T)_H and ( delta ²M/ delta T²)_H and those deduced from recent heat capacity and magnetocaloric experiments performed on similar specimens. The results show that above 0.05 T all three experiments agree closely. At low fields, however, there are small but significant differences that most probably arise from the use of different specimens for the three experiments but might have different causes. It is concluded that the equilibrium hypothesis is consistent with the experimental data at present available but that further experiments at low fields are required to test it thoroughly. The hypothesis has the remarkable implication that just below the glass temperature the entropy of a field-cooled spin glass increases with field in low fields and goes through a maximum as the field is increased (at 0.05 T for our specimen). An analogy is drawn between this behaviour and that of an ordered antiferromagnet for which the entropy is also an increasing function of field.

The authors report Mossbauer and magnetisation studies of the alloy Fe_0.69Pd_0.31 with special emphasis on the FCC-FCT martensitic transformation. The transition temperature of 184K is determined from the T^3/2 dependence of the quadrupole splitting which undergoes a change in slope after the transformation. The hyperfine field distribution in the FCC phase at various temperatures shows a steady increase in the distribution width with temperature, which indicates a short-range order in the material. The recoilless fraction of the material is determined using the method of Trooster et al. (1978) and the value of f_a obtained at room temperature is 0.568 in this alloy.