Table of contents

The authors calculate a structure factor for the liquid alkali metals assuming that the soft core of the potential plays the dominant structure-determining role. The structure factor is calculated in an analytic form suitable for fitting to experimental data. They obtain good agreement with existing measurements on Na, K and Rb.

The existence of charge density waves in alkali metals can cause smearing at the Fermi momentum of the positron annihilation angular correlation profile. It is suggested that a significant part of the apparent non-thermalisation of positrons could be a reflection of the charge density wave ground state.

The electronic structure of vacancies and hydrogen-trapped vacancies in aluminium are investigated using the self-consistent scattering X alpha method. The energy levels, electron density and charge distribution are obtained. The results are used to compute the lifetime of a positron in a vacancy and in a hydrogen-vacancy complex.

Internal friction and modulus measurements have been performed on high-purity cold-worked magnesium which was bombarded by electrons at low temperature during successive anneals up to 250K. The variation of Bordoni-type peaks is discussed in terms of pinning and diffusion of defects along dislocations lying along the Peierls' valleys. Some very low-temperature peaks are associated with an interaction between radiation-induced defects and geometrical kinks.

If the static lattice Green function of a crystal is known then the distortions produced in the crystal by a point defect may be obtained simply from the Green function and the forces acting on the crystal due to the defect. The numerical evaluation of the Green function for a given model of the crystal force constants involves Brillouin zone summation of a function which is singular at the origin and this singularity can cause considerable numerical convergence problems. It is shown that a useful calculational procedure involves evaluating the Brillouin zone summation by using the Chadi and Cohen sets (1973) of special points in the Brillouin zone together with an efficient extrapolation procedure. Green functions are calculated for some FCC and BCC metals using the force constants obtained from Born-von Karman fits to experimental lattice vibration spectra.

The fine structure of low-energy phonon dispersion branches in Mg has been investigated by inelastic neutron scattering. The irregularities observed in the photon group velocity are a kind of Kohn anomaly. They are compared with those determined in Cd. Theoretical calculations of the fine structure in Mg were made in a new way by taking band gaps into account. The theory explains the difference between Mg and Cd quite well, but fails to describe the experimental data quantitatively. In particular, the varying strength of different anomalies are not reproduced.

The low-temperature lattice, specific heat of dilute copper-gold alloys has been studied on the basis of the Green function theory. The effect of lattice expansion on the specific heat has been emphasised. An impurity model including central and non-central force constant change is employed. The change in specific heat is dominated by mass-induced resonance modes appearing in the F_1u irreducible representation. The calculated results are in reasonable agreement with the experimental data.

The charge transfer in liquid alloys discussed previously in terms of the partial structure factors in the low-wavelength limit is modified using the Thomas-Fermi screening of a point charge. This may enable us to apply it to compound-forming liquid alloys which have a considerable volume contraction. Numerical calculations are carried out for liquid Na-K and Na-Hg alloys.

High-resolution quasi-elastic neutron scattering has been used to study the hydrogen diffusion in the multi-component Laves phase hydride Ti_0.8Zr_0.2CrMnH₃, one of the most promising hydrogen storage materials for technical applications. The data analysis is complicated by multiple scattering which, even with only moderately scattering samples, leads to a fictitious line broadening at small momentum transfers and considerably hinders the determination of diffusion coefficients. An evaluation procedure is presented which allows for the necessary multiple scattering corrections independent of details of the single diffusive steps. In the temperature range of the study, 230-360K, the hydrogen self-diffusion coefficient in Ti_0.8Zr_0.2CrMnH₃ obeys the Arrhenius law: D=(3+or-1)*10^-4 cm² s^-1 exp(-(220+or-20) meV/k_BI). This comparatively fast hydrogen diffusion is not the rate determining step in the absorption and desorption kinetics.

A Green function formalism is developed in the muffin-tin approximation to calculate the local electronic densities in structurally disordered metals. The local atomic arrangement and also the long-range distribution of the atomic scattering centres are treated on an equal footing by dividing the coordinate space into three regions. In the first region, in the immediate neighbourhood of the central space into three regions. In the first region, in the immediate neighbourhood of the central atom, the further atoms are treated individually, as in a cluster. Outside the cluster, the second region, the atomic distribution is represented by the pair distribution function. The nearly constant tail of this function defines the third region. The local density of states is calculated for different atomic arrangements inside the cluster, and the results are then averaged for different clusters. By applying the method to iron, a single-peaked averaged density of states curve was obtained which is somewhat shifted to lower energies.

Multiple scattering theory in the small scattering atom approximation is used to evaluate the Green function at high energies. A rearrangement of the series expansion in terms of the individual t matrices is used where the leading term is the Green function of the free atom. The energy density matrix well above the Fermi energy can then be calculated from the wavefunctions of the central atom as well as from the position and scattering amplitudes of the other atoms. An angular momentum expansion appears only for the central atom. The sum over all closed loops can be restricted to the neighbouring atoms inside a cluster around the central atom due to the imaginary part of the optical potential. The EXAFS amplitude is calculated using this approach for an arbitrary angular momentum of the initial state. The result contains all multiple scattering contributions and reduces in the single scattering approximation to the well known results for the K and L edges.

The radial distribution functions of Cu_xZr_1-x (x=0.35, 0.50 and 0.65) have been calculated by the means of the relaxed dense random packing model. The calculated results agree well with the experimental data obtained by both X-ray and neutron diffraction. Based on the present model structures and the pair potentials, the vibrational densities of states have been calculated; from these the sound velocities and elastic constants were estimated. Agreement between theory and experiment is fairly good.

The effect of spin-orbit coupling upon the energy band structure of the light actinide metals has been examined. A measure of the importance of spin-orbit coupling is the ratio of the number of f electrons in the j=⁵/₂ and j=⁷/₂ bands, which approaches 6/8 when spin-orbit coupling is small. The computed ratio from ab initio self-consistent relativistic calculations for the metals Ac-Am varies from 0.75 for Ac to 2.4 for Am. The change in atomic volume due to the preferential occupation of the j=⁵/₂ band has also been computed. It is small for the elements Ac-U but the volumes of Np-Am increase progressively. There is a corresponding decrease in the change of volume for the expected phase transition in americium under pressure.

The electrical resistivity of liquid Sn-Te alloys has been measured as a function of temperature and concentration. The resistivity isotherm has a remarkable maximum around the equiatomic composition, suggesting that a kind of compound formation exists at this composition. The compound fraction is estimated as a function of concentration using thermodynamic analysis. The electrical resistivities and the magnetic susceptibilities are interpreted consistently in terms of the compound formation of SnTe.

Experimental results are presented for the Hall coefficients of a number of amorphous Cu-Zr, Cu-Ti and Cu-Hf alloys in the temperature range 4.2-270K. With the exception of Cu₉₂Zr₈, all these alloys are found to have positive Hall coefficients and none show any significant temperature dependence. The fact that R_H is temperature independent can probably be understood as a direct consequence of the isotropic nature of the scattering. The positive values cannot be understood on the basis of nearly free electron ideas nor within the random phase model if hybridisation is neglected. It is therefore argued that the positive values are probably due to the s-d hybridisation leading to negative group velocities at the Fermi level. Using this approach a qualitative account can be made of the compositional dependence of R_H in the alloys used here and in a number of related systems.

The electrical resistivity and the absolute thermoelectric power (TEP) of Pd₈₀Ge₂₀ and Pd_77.5Ge_22.5 alloys, in both glassy and crystalline states, have been measured over a wide temperature range. In the glassy alloys, the room-temperature resistivity is high (about 100 mu Omega cm) and has a positive temperature coefficient. The room-temperature TEP for the alloys in both states is negative. The magnitude is, however, a factor of four larger in the corresponding crystalline state. In Pd_77.5Ge_22.5 glassy alloys, the TEP also changes sign at about 230K as the temperature is lowered. The overall transport behaviour of the glassy alloys can be understood in terms of the Evans-Greenwood-Lloyd diffraction model (1971) for electron concentration in the alloys. This involves electron localisation at the Ge sites and filling of the transition-metal d band. Since there is a delicate interplay between the structure factor and the d phase shift terms of the TEP, its sign is decided by the relative magnitudes of these two terms.

The authors evaluate the magnetoelastic contributions to the coefficients in the free energy which determine whether the ground state of neodymium is single Q, double Q or triplet Q. They also develop a model free energy which leads to a second-order polarisation and wavevector reorientation phase transition at a temperature somewhat below the Neel transition.

The electronic structure of paramagnetic iron is calculated self-consistently within the local spin density functional formalism by applying the coherent potential approximation with the muffin-tin potential model. The effective exchange parameters J_n are calculated as the interaction energy between two magnetic moments in the paramagnetic medium obtained. A discussion on the magnetism of iron above the Curie temperature is presented using the Heisenberg model with the calculated J_n.

The optical absorption spectra epsilon ₂(0, omega ) of Ni₂MnSn and Pd₂MnSn are evaluated on the basis of the realistic band structure obtained by the tight-binding orthogonalised plane wave (TB OPW) method. The transitions between OPW in the hybridised states contribute dominantly to the main structure of the spectra, as was the case in Cu₂MnAl. The common properties in epsilon ₂(0, omega ) of the ferromagnetic Heusler alloys X₂MnY (X-Ni, Cu, Pd: Y=Al, Sn) are examined using the present results and those obtained previously for Cu₂MnAl. Three peaks below 4 eV and a peak above 7 eV beyond which epsilon ₂(0, omega ) decreases rapidly are observed as the common structures of the three alloys. These structures are related to the d states of Mn in the three alloys.

High pressure and high magnetic field experiments have been performed on the ferrimagnetic Er₆Fe₂₃ compound. The erbium-iron interaction parameter is found to be independent of pressure, magnetic field and temperature in the range of the experimental facilities (p<or=5 kbar, mu ₀H<or=35 T and 4.2<or=T<or=300K). The relative volume dependence of the iron moment in Er₆Fe₂₃ at 4.2K is almost equal to the value in pure iron.

The concentration dependence of the average hyperfine field and the hyperfine field distribution in Fe_100-xB_x metallic glasses (0<or=x<or=25) is explained in terms of a local environment model based on dense radiation packing of hard spheres' model structures. It is assumed that the hyperfine field at iron sites without boron neighbours is sensitive to local compression or dilation in analogy to crystalline close-packed iron. In this way even the strong decrease of the average hyperfine field at low boron content can be explained.

Pulsed NMR results on the ²⁷Al, ¹⁵⁵Gd and ¹⁵⁷Gd nuclei in the ordered phase of GdAl₂ are presented and discussed. It is shown that the quadrupole interaction at the Gd sites is too small to be measured ( mod P mod <or approximately=1 kHz). This result is used, in conjunction with some theoretical estimates of Gehring and Walker (1981), to infer that the pseudoquadrupole interaction in GdAl₂ is too small to account for the temperature quadrupole interaction at the ²⁷Al site. Some new results are presented for the ²⁷Al resonance. It is shown, both on experimental and theoretical grounds, that the 'a' and 'b' peaks are generated primarily by nuclei within domains and domain walls respectively. In particular, it is argued that the 'b' peak is generated by nuclei whose local axis of magnetisation within a domain wall lies close to a (110) hard axis.

High-field NMR Knight-shift measurements of ¹⁵⁷Gd and ²⁷Al in a ferromagnetically ordered single crystal of GdAl₂ are described. The values of K(¹⁵⁷Gd)=+52*10^-4 and K(²⁷Al)=-25*10^-4 are derived at T=1.8 and 1.5K, respectively. The results are analysed in comparison with the corresponding values in non-magnetic LaAl₂ and in combination with known susceptibility data. A predominance of non-s conduction electrons in the ferromagnetic phase of GdAl₂ is concluded, indicating a substantial change in the conduction band character at the Al sites upon magnetic ordering.

Measurements of the low-energy satellites of the core X-ray photoemission lines of silver and gold are presented. The results are compared with electron energy loss spectra and dielectric measurements. Interpretation of the results on the basis of plasmon excitation is given and compared with earlier interpretations.

The author have measured the angle-resolved energy distributions of electrons photoemitted from the (100) faces of single crystals of pure Cu and Cu-5 at.% Ni. Cu-23 at.% Ni and Cu-40 at.% Ni random substitutional alloys, using various photon energies and incident directions. They interpret the richly structured data using first-principles calculations of the configurationally averaged photocurrent within the framework of the Korringa-Kohn-Rostoker coherent potential approximation (KKR CPA). The quantitative agreement found between theory and experiment shows that the KKR CPA can describe electron states in these alloys in k point by k point detail. They conclude that the electronic structure of Cu-Ni alloys consists, for the most part, of bands whose smearing in energy and wavevector is less than their separation.