Table of contents

In situ high-pressure X-ray diffraction data for polycrystalline beryllium to 28.3 GPA at ambient temperature show that beryllium is transformed from the HCP phase (I) into a slightly distorted HCP phase (II) at pressures between 8.6 and 14.5 GPa. The volume change for the transition is extremely small ( approximately 0.4%); the effect of pressure on the c/a ratio for both the Be(I) and Be(II) phases is also very small.

Edwards' cancellation theorem is re-examined by taking the proper weak-scattering limit of his self-consistent second-order perturbation theory for the electrical conductivity. It is shown that the theory can be rewritten in an equivalent but fully renormalised form, and two important contributions have turned out to be missed out in previous studies. This analysis leads to several correction terms to the famous Ziman formula; in particular, the free-electron mass in the formula is shown to be replaced by the effective mass defined in a natural way.

The observed structure factors of expanded liquid rubidium (for temperatures up to 1900K and pressures up to 122 bar near saturation conditions) can be explained using (a) nearly-free electron theory to describe the interatomic forces and then (b) standard techniques of classical liquid theory to deduce the corresponding structures. The high-K observations correspond to close binary collisions involving atomic cores of fairly hard-sphere character, while the low-K behaviour can be described by long-wavelength density fluctuations, the description of which requires, in addition, the use of the interatomic potential tails.

Results of electromotive force measurements of Na activity in dilute Na-Bi and Na-Sn alloys in the temperature range 573-723K are reported. The limiting slopes are determined for the logarithm of the activity coefficient of Na as a function of its concentration in the solvent. Using the equations derived by Kirkwood and Buff (1951), the sign and magnitude of the slopes are shown to have structural implications. Data from this study and published literature data on some Na and Li alloys are consistent with a net repulsion between alkali atoms in solution in electronegative metals.

Structural relaxation of a metal-metal-type amorphous alloy has been simulated using the method of molecular dynamics for a model Cu₅₇Zr₄₃ alloy constructed by central force interatomic potentials. The sharpening of the peaks in the total radial distribution function (RDF) after annealing agrees well with that of experiment. The decrease of the internal energy Delta E is related to the density change Delta rho by the equation - Delta E/ mod E mod =1.6 Delta rho / rho , which means that the structural relaxation is achieved by the densification which increases the average coordination number. The densification is found to occur effectively by developing compositional short-range order (CSRO), i.e. by increasing the number of unlike pairs of the small atom (Cu) and the large atom (Zr). The formation of CSRO causes the shift of the first peak of the partial RDF (PRDF), outward for Cu-Cu pairs and inward for Zr-Zr pairs, resulting in the sharpening of the total RDF. The atomic diffusion has been found to be quite heterogeneous with a dimension of about five atomic spacings.

Specimens of the intermetallic compound Ni₃Al were irradiated by 2 MeV electrons at room temperature or quenched from 900 degrees C to about 1300 degrees C and subsequently annealed isochronally. The behaviour of the vacancies was studied using the positron lifetime. The vacancies annealed out around 215 degrees C. The formation energy of a vacancy was determined to be E_v^F=1.6+or-0.2 eV. The activation energy for the motion of the vacancy was estimated to be E_v^M=1.2+or-0.2 eV. The annealing of the vacancies was found to be a second-order reaction.

The elastic constants of BCC transition metals (Fe, Nb, Mo and W) are calculated by using the tight-binding d band and the Born-Mayer repulsive potential. Introducing a small distortion characteristic to C₄₄ (or C') elastic deformation and calculating the energy change up to second order in the atomic displacement, the authors determine the shear elastic constants C₄₄ and C'. The elastic constants C₁₁ and C₁₂ are then calculated by using the relations B=¹/₃(C₁₁+2C₁₂) and C'=¹/₂(C₁₁-C₁₂), where B is the bulk modulus. In general, the agreement between the present results and the experimental values is satisfactory. The characteristic elasticity behaviour, i.e. the strong N_d (number of d electrons) dependence of the observed anisotropy factor A=C₄₄/C', is also discussed.

The evolution of the internal friction under irradiation is measured over three orders of magnitude of the electron flux. The peaking effect is shown to be flux dependent at some temperatures. The results are explained qualitatively in terms of a model for the dislocation dynamics where a long-range dislocation-point defect interaction is included.

The relationships between the elastic constants and the force constants for the modified models of Krebs (1965) and Cheveau (1968) have been re-evaluated. This evaluation shows that the relationships currently used in the literature for these models neglect two contributions. One is due to the volume dependence of the dielectric function which is contained in the screened electron-ion term. The other arises from the Umklapp contribution contained in the screened electron-ion expressions of Krebs and Cheveau. These contributions have been calculated for the modified model of Krebs and are found to be significant. Furthermore, it was found that, to satisfy equilibrium, electron exchange and correlation must be included.

Precision total neutron cross section measurements of ZrH_1.58 were performed using a highly monochromatic neutron beam for neutron energies between 0.035 and 1.1 eV. The experimental values of the cross section in the region of the one-phonon optical peak (0.100 eV<E_n<0.250 eV) compare well with calculations if one assumes that the phonon spectrum g( epsilon ), is represented by three Gaussians centred at 0.132, 0.137 and 0.151 eV, with associated widths of 0.05, 0.02 and 0.01 eV, respectively. The phonon spectrum g( epsilon ) departs from the calculations as exemplified by the central force (CF) model of ZrH₂. Such departures may arise due to non-central forces and tetragonality. The structure of g( epsilon ) in some respects is similar to that observed in inelastic neutron scattering cross section measurements. The results from these two independent and different types of measurements provide sufficient basis for further attempts toward a more realistic theoretical model of ZrH₂.

A cluster theory method based on a continued fraction expansion of the Green function has been used to calculate the density of states of an interstitial metal-hydrogen alloy for a simple single-band model. Disorder has been introduced through both diagonal and off-diagonal matrix elements. Partial densities of states seen by electrons on hydrogen and metal atoms have been calculated for a whole range of hydrogen concentrations. A cluster of four coordination shells with 91 configurations for the hydrogen and 63 configurations for the metal atom at the central site has been considered. Numerical calculations have been performed for a number of parameter sets. The formation of a hydrogen-metal bonding band has been observed. The application of the simplified single-band model to the Pd/H system has led to a density of states consistent with an earlier photoemission experiment by Eastman et al. (1971).

Simultaneous measurements of quantum oscillations in magnetostriction and torque in a single crystal of ruthenium have been used to determine the derivative with respect to tensile stress along the hexad crystal axis of the central cross section, normal to that axis, of the hole sheet of the Fermi surface in zone 10. The uniaxial stress derivative delta lnA/ delta sigma ₃=-(3.6+or-1.0)*10^-12 cm² dyn^-1, is found to be an order of magnitude larger than the corresponding hydrostatic pressure derivative. The Fermi surface is therefore strongly influenced by distortion of the hexagonal crystal axial ratio.

The authors present results of relativistic band structure calculations for the noble metals Cu, Ag and Au. Particular effort was focused on good-quality self-consistent charge densities and on well converged excited bands (up to 30 eV above the Fermi level) to serve present needs of photoemission analysis within that energy regime. The calculations rest on Kohn-Sham-type relativistic one-particle equations with a local exchange-correlation potential either of the form derived by Hedin and Lundqvist (1971) or to the X_alpha form suggested by Slater. In using the latter potential best agreement with the experiments could be obtained by assigning alpha the values 0.70, 0.82 and 0.85 for Cu, Ag and Au, respectively. Self-consistency effects caused by relaxation of the core states were accounted for fully. The authors' findings do not support results of recent (differently performed) calculations by other authors, using the same (or similar) one-particle equations, who generally obtain d bands which are bound too weakly and whose widths come out too large. In contrast, they even find good agreement with spin-orbit effects observed in the photoemission spectra of Cu.

Bulk samples of the intermetallic compounds Mg₃Bi₂ and Mg₃Sb₂ have been investigated experimentally using soft X-ray emission and X-ray photoemission spectroscopy. The electronic band structure has been calculated using an extended Huckel technique. The results are discussed and critically compared with existing data on electrical conductivity and soft X-ray absorption.

The phasing of the charge-density wave/periodic structural distortion relative to the 3a₀ commensurate unit cell of low-temperature 2H-TaSe₂ is investigated in the light of recent experimental and theoretical work. In trying to accommodate all the data it becomes apparent that the phasing selected alters substantially with applied magnetic field; interpretation of the NMR and Hall coefficient measurements is at stake, and to a lesser extent the electron microscopy. The phase diagram develops striking resemblance to that modelled by Littlewood and Rice (1982) in connection with the pressure variable. The low-field/low-pressure structure is identified Consequences for the discommensuration array are discussed.

The authors have calculated the resistivity, caused by the scattering of free electrons on an impurity described by a localised potential with cubic symmetry in terms of phase shifts (l<or=2), using an exact solution of the linearised Boltzmann equation and an approximation to the solution consisting of four cubic harmonics with t₁z symmetry. The latter approach is based on the variational principle, whereas the former one is developed along the lines indicated by Mertig and co-workers (1982). The resistivity calculated with the exact solution still contains three numerical integrals, which have to be computed for every set of scattering phase shifts. The resistivity calculated with the approximation implies an inversion of a symmetric 4*4 matrix for each set of phase shifts. The resistivity calculated with the approximation deviates markedly only for some sets of phase shifts from the resistivity calculated with the exact solution. The phase shifts, deduced from recent UPS measurements on AuNi, cannot account for the experimental resistivity within the model.

A simple technique for calculating the phonon-limited resistivity ( rho _ph) of metals is described which is analogous to the Gaspari-Gyorffy method (1972) for calculating the electron-phonon coupling constant ( lambda ), rho _ph and lambda are calculated for all 4d metals and silver. The HCP metals have been treated as cubic, as is common in calculations of lambda . For rho _ph, however, this approach turns out to be too rough because it depends strongly on the average Fermi velocity which is very sensitive to crystal structure. In all transition metals investigated the superconducting and transport electron-phonon coupling constants are nearly equal, lambda approximately= lambda _tr. Also nearly equal are their electronic parts, eta approximately= eta _tr. The only exception is Pd, where eta _tr is nearly 1.5 times larger than eta ; this is because the superconducting and transport properties of Pd are defined by different parts of the Fermi surface. Palladium is the only 4d metal where the Mott model of conductivity (s-d scattering) is valid.

This study examines two parameters describing the electromagnetic properties of superconductors: the upper critical field H_c2 and the low-field penetration depth lambda . The current state of the theory for H_c2 is summarised and some new results for lambda are derived to focus on the dominant role of the Fermi surface topology and resultant anisotropies. These properties are evaluated for Nb and Nb₃Sn as specific cases. Global parameters are found to be in good agreement with experimental data but other properties exhibit some scatter. The anisotropic deviations of the electromagnetic response are given to about 10%. An estimate is obtained for the dependence of the penetration depth on the direction of the shielding current.

The authors study the effect of small anisotropic interactions on the stability of the ferromagnetic spin component in a semi-spin glass. They conclude that the ground state consists of large ferromagnetic domains separated by narrow walls.

The influence of magnetism on the short-range order (SRO) and its temperature dependence is clearly pointed out in a Fe_0.9V_0.1 alloy. Emphasis is placed on the metallurgical aspects of the study carried out by means of residual resistivity and nuclear magnetic resonance (NMR) measurements at low temperature. The ordering kinetics were measured in order to determine the heat treatments required to obtain the thermodynamic equilibrium state for SRO and the appropriate conditions necessary to retain it at low temperature by quenching. The measured ordering rates are in good agreement with previous anelasticity results. The temperature dependence of the SRO below the Curie temperature T_CM exhibits a strong departure from the behaviour expected in a classical mean-field approximation. This dependence is in qualitative agreement with that given by a magnetic mean-field model (Ising-Heisenberg). Preliminary results obtained at temperatures above T_CM (1<T/T_CM<1.4), in an Fe_0.96V_0.04 alloy, exhibit the expected behaviour for the paramagnetic phase.

Polarisation analysis of neutron diffuse scattering has been used to characterise atomic short-range order in disordered FCC MnNi alloys. The diffuse scattering observed at 4.2K is shown to arise from fluctuations of the average atomic unit cell structure. The data have yielded the concentration dependence of the Cowley short-range order parameters which express the interaction of a given atom with the atoms in its immediate local environment and show that Ni atoms prefer a Mn nearest-neighbour environment. The result of a linearised mean-field model of the pair interactions which relates the pair potentials to the diffuse scattering has also been applied to the experimental data and shows the characteristic Friedel oscillations on the pair interactions which, in general, reflect some features of the disordered alloy electron spectrum.

A model which includes the effects of magnetic as well as atomic environment on the spatial magnetic moment fluctuations in binary alloys has been applied to measurements of the diffuse magnetic scattering from antiferromagnetic gamma -MnNi alloys. The application of this model has shown that the effects of magnetic environment are particularly important at concentrations approaching the critical concentration for the onset of long-range antiferromagnetic order and signifies some further success of the formalism of a model which has previously accounted for the moment defects in nickel-based ferromagnetic alloys.

The crystallographic and magnetic properties of Tm₆Fe₂₃ before and after hydrogen absorption were studied by means of X-ray diffraction, bulk magnetic measurements and ¹⁶⁹Tm and ⁵⁷Fe Mossbauer spectroscopy. The data were analysed in terms of a mean-field model. Coupling constants describing the magnetic interaction between the Fe moments and between the Tm and Fe moments were derived for Tm₆Fe₂₃ as well as for its hydride.

The authors show that, because of the existence of a distribution of relaxation times in the percolation model of the spin-glass transition, magnetisation relaxes logarithmically in time.

By using the screening charge model of Alfred and Van Ostenburg (1969) for the valence effect and the continuum model of lattice strain for the size effect, electric field gradients (EFG) in Cu-Ni and Cu-vacancy are calculated at the first four nearest-neighbour host sites. The strain-field gradient coupling parameter lambda has been fixed by fitting the calculated EFG at the fourth nearest neighbour with experiment. The results show that except for the discrepancy in the principal Z direction of the EFG tensor at the first nearest neighbour in Cu-Ni and in the magnitudes of the EFG at the second nearest neighbours, there is good agreement between theory and experiment in both the systems. Possible reasons for the existing discrepancy are discussed.

The electronic structure of ferromagnetic FCC Co_0.92Fe_0.08 has been studied by angle-resolved UV photoemission from single-crystal faces in the (111), (110) and (100) orientations. The spin-polarised energy bands have been computed in the average t matrix as well as in the virtual crystal approximation using the KKR method. A Lambda ₁ symmetry surface state has been found at Gamma (zone centre) at an energy of E_i=-0.35 eV relative to the Fermi level E_F. The remaining structure can be interpreted in terms of emission from bulk states and is compared with the calculated band structure. Owing to a strong lifetime broadening only the transitions within a few eV of E_F can be used for the analysis. They support the calculation. The experimental value of the exchange splitting Delta E_ex has been found to be 1.3 eV, which essentially confirms our theoretical value of 1.5 eV.

The authors present angle-resolved photoemission spectra from (100) faces of Cu₉₆Ge₄ and Cu₈₉Ge₁₁ single crystals using 16.85 eV radiation, together with the corresponding Korringa-Kohn-Rostoker coherent potential (KKR CPA) complex band computations. The results are compared and contrasted with an earlier photoemission study of Cu₈₇Al₁₃ single crystals. Among the features of the alloy spectrum discussed are the characteristic shifts in the Cu d bands and in the s-p-type valence bands, changes in the Fermi surface dimensions and the disorder smearing of states. The computations provide a good understanding of the various observed effects of alloying on the bulk electronic structure of Cu.

High-resolution photoemission spectra are presented of Ag-1.90 at.% Pt and Ag-10 at.% Pt alloys, together with low-temperature thermopower measurements on Ag-1.05 at.% Pt and Ag-1.28 at.% Pt samples. The photoemission data clearly show a splitting of the virtual bound Pt 5d states due to spin-orbit and crystal-field effects. A simple atomic theory, including these effects, is presented and shown to fit the photoemission data. The spin-orbit coupling is found to be only slightly reduced from the spectroscopic free-atom value. The authors show that the virtual bound state parameters extracted from the photoemission spectra can describe the experimental transport data within a free-electron phase-shift model incorporating only spin-orbit coupling. It is found to be necessary to include s and p phase shifts.