Table of contents

Using the one-component plasma as a reference system a positive expression for the ion-ion structure factor S_ii(q) is obtained for all q.

Square-shouldered and Yukawa-shouldered hard-sphere model potentials have been used to find the best possible fits of these types to the experimental structure factors, S(q), of liquid polyvalent metals. The trends in the model potentials which best fit the S(q) are seen to follow the corresponding trends for the crystal structures studied by Hafner and Heine (1983).

The Fermi surface of Hg_{3- delta} AsF₆ is studied by means of the de Haas-van Alphen effect in magnetic fields up to 35T. The authors observe magnetic breakdown over the gaps resulting from incommensurability. The delta signal decreases in amplitude in fields larger than 10 T whereas a normal increase is seen for all other main orbits. Estimates for the effective field values belonging to magnetic breakdown and scattering are given. They set an upper bound of approximately 3 mRyd for the matrix elements of the incommensurate component of the lattice potential. A spin-orbit splitting is observed for the epsilon orbit in agreement with a self-consistent calculation including relativistic effects.

A SQUID voltmeter with a noise of 8*10^-15 V Hz^-1/2 at 1K has been used to measure the resistance and thermoelectric properties of a number of superconductor-normal-metal-superconductor (SNS) sandwiches. Both the resistance and thermoelectric EMF diverge near the lower transition temperature in a fashion which is well explained by the results of a new theory for Pb/Cu/Pb sandwiches but sandwiches of Pb/Cu/eutectic PbBi show more puzzling features.

Stoichiometric molybdenum nitride with sodium chloride-type structure has been prepared by reactive sputtering and by ion implantation of nitrogen into molybdenum. The lattice parameter of this phase is 0.4212 nm. The superconducting transition temperature is only around 3K in contrast to expectations raised by theoretical and empirical predictions.

The superconducting transition temperature of lanthanum (La) and actinium (Ac) are calculated at various high pressures from the relativistic band-structure results obtained using the relativistic augmented plane-wave (RAPW) method. From the calculations it is seen that T_c increases with increasing pressure or decreasing cell volume for both the metals. A new empirical relation for the volume dependence of T_c is obtained.

Rigorous expressions for the exchange parameters and the spin-wave stiffness constant of ferromagnetic metals are obtained using a local spin-density functional formalism and the muffin-tin potential model. The calculated results for ferromagnetic iron are given as an example of the approach.

The cubic Laves phase of YCo₂ is near a transition from a paramagnetic to a magnetically ordered state. Self-consistent energy-band calculations yield the total energy and the magnetic moment as functions of volume. The new fixed spin-moment (FSM) method, which allows one to calculate the total energy as an explicit function of the magnetisation, is introduced. At the theoretical equilibrium volume YCo₂ is found to be a strongly enhanced Pauli paramagnet, but at a slightly larger lattice constant a metamagnetic transition seems possible. Its occurrence can be understood on the basis of a double minimum in the total energy obtained from the present FSM calculations, which lead to an estimated critical field of about 350 T. In the magnetic state the Y moment is coupled antiferromagnetically to the Co moment.

A theory for the surface tension of liquid metals is presented and applied to the alkalis and the polyvalent metals, Mg, Zn and Al. The numerical results are in excellent agreement with experimental data. The theory also predicts realistic surface density profiles for the ions and the electrons.

Perturbed angular correlation (PAC) and ion channelling techniques were used to study the trapping of vacancies by In atoms in a Cu-0.13 at.% In crystal after H⁺ irradiation. Four major defect complexes were observed; the fractions f₁, f₂, f₇ and f_T of In atoms which were associated with these complexes were measured as a function of annealing temperature and irradiation fluence. The complexes f₁, f₂ and f₇ were observed only by the PAC technique; the coupling constants for these defects were v_Q1=116 MHz, v_Q2=181 MHz and v_Q7=106 MHz, all having eta =0. The first two of these had previously been shown to be due to trapping of one and two vacancies. The third may be due to He trapping. A tetrahedral complex f_T, consisting of an In atom in the tetrahedral interstitial site surrounded by four nearest-neighbour vacancies was observed by channelling but was invisible by the PAC method. Complexes, f₁, f₂, and f_T were formed during annealing from 180-250K. The fractions f₁ and f₂ decreased strongly during annealing from 250-230K, while the tetrahedral complex was stable up to approximately 500K. The complex f₇ appeared during annealing above 320K and continued to increase up to 600K. The fractions f₁, f₂ and f_T were all reduced during 70K irradiations by the absorption of self-interstitial Cu atoms.

The normal modes of vibration in a Mo-Ru alloy were measured at 296K using inelastic neutron scattering. The resulting frequency-wavevector dispersion relations are compared with the pure Mo results of Powell, Martel and Woods (1968). The primary deviation is at the zone-boundary point H where a dip in the frequency curve is accentuated with the addition of Ru. A Born-von Karman analysis is attempted but is unable to specify which force-constant changes are significant, suggesting the importance of longer-range electronic effects on the spectra.

Thermal desorption spectroscopy (TDS) of hydrogen from metal hydrides is a technique where the rate of hydrogen desorption is measured as a function of temperature while the temperature is increased linearly with time. The authors present the TDS of a powdered PdH_x (particle size approximately 2 mu m) for various initial concentrations and for various heating rates. The spectra develop from a single peak (A) and a small shoulder curve at low initial H concentrations to two peaks (A,B) and a shoulder at high initial concentration (x equivalent to 0.9). The peaks are correlated to the phase diagram-the peak A to the alpha + beta region, the peak B to the beta region and the shoulder to the alpha region. Activation energies for desorption and preexponential factors are derived from the Arrhenius-like parts that the spectra exhibit in their onset regions and in the region alpha + beta . The results are discussed in terms of a model in which the desorption rate is determined by the rate of the recombination process H+H to ₂ at the surface.

Presents experimental data obtained for the variation of the specific volume of the liquid metals Al, Cu, Mo, Ta and W with applied energy produced by pulsed electric heating of wires of the material. The liquid volume was measured by two methods; the first is based on the fact that the variation of the electrical resistance of the wire with applied energy for free expansion of the liquid differs from that when the volume of the liquid is limited by a surrounding capillary tube, and in the second the volume is measured by using the electrical contact between the expanding wire and fixed probes.

A detailed calculation of the cyclotron masses in indium is presented for various extremal cyclotron orbits on both the electron and hole surfaces. The anisotropic mass enhancement lambda _k is also calculated at a number of places on the Fermi surface using realistic pseudopotential models for the phonon frequencies and polarisation vectors and for the Fermi surface, many-OPW electronic wavefunctions, Fermi velocities, and the electron-phonon interaction. The authors have carried out the line integrals on the Fermi surface both without and including lambda _k for the central and non-central orbits in all directions in the (110), (010), (111) and (001) planes for the second-zone hole surface and in the (011), (110) and (111) planes for the third-zone electron surface. In general, the calculated cyclotron effective masses are in reasonable agreement with the experimental results for most of the orbits observed by Mina and Khaikin (1960) and by Hughes and Shepherd (1969). The authors find a 43% difference between the maximum and minimum values of lambda _k over the second-zone surface. Their calculations also show that the lambda _k on the third-zone electron surface are generally larger than those on the hole surface. The variation of the orbit-averaged mass enhancement over the second-zone surface or over the third-zone surface does not correlate well with experiment. They attribute this to inaccuracies in the electron-band velocity.

A simple method is presented to obtain the static and dynamic local field corrections of the dielectric function of an electron liquid. The method can be extended easily to inhomogeneous systems and lower-dimensional cases. The authors present calculations for the static field correction of a homogeneous electron liquid; their results are in good agreement with other calculations.

The electrical resistivity of amorphous Ca₇₀Mg₃₀ alloys has been measured over the temperature range 4-300K. The absolute value and temperature dependence of the electrical resistivity have been calculated on the basis of the generalised Faber-Ziman theory, without an adjustable parameter. Excellent agreement between theory and experiment is found for a plane-wave phonon model.

A new model for the influence of surface scattering upon the thermopower in thin metal foils and wires is developed. It is based upon Soffer's angularly dependent surface-scattering theory. For most other transport properties the predictions of Soffer's theory only differ substantially from the commonly accepted Fuchs theory in the limit of the bulk mean free path being much greater than the thickness. However, in the case of the thermopower the angularly dependent specularity introduces an entirely new term which arises from the dependence of the surface scattering on wavelength, and hence on energy. The new term is evaluated for a large range of values of both surface roughness and the ratio of sample thickness to bulk mean free path. In circumstances which are readily obtainable experimentally it is shown that this term may be of equal magnitude to the more commonly accepted terms. This almost certainly explains why much of the data on the thermopower developed in thin samples is irreproducible.

The dependence of critical temperature T_c on pressure P and iron concentration x was measured for Fe_xSnMo₆S₈. The iron concentration was determined by activation analysis. The variation of T_c(x) is similar to that predicted by Abrikosov and Gorkov (1960). The critical concentration is x_cr=0.061 (0.42. at.%). For x=0.04, delta T_c/ delta P is three times higher in absolute value than for x=0 and reaches -3.2*10^-4 kbar^-1. The variation of T_c(P) is not linear and may be explained by the pressure-enhanced exchange interaction of electrons with the magnetic moment of the impurity atoms.

The magnetisation density in vanadium metal induced by an applied field of 4.6 T at 100K has been measured using the polarised neutron technique. The authors have also measured the nuclear scattering length of vanadium in their crystal by making absolute measurements of the integrated intensity of reflection of unpolarised neutrons. These yield a value of -0.0446(9)*10^-12 cm for the scattering length of vanadium, and indicate the presence of interstitial impurities giving a scattering length of 0.0010(1)*10^-12 cm per interstitial site of the type (0, 0, ¹/₂). The magnetisation density has been fitted to a simple model based on the vanadium 3d³4s² free-atom form factor. Coefficients giving the amounts of spin and orbital moment and the degree of asphericity were determined. The density is nearly spherical with respect to the vanadium sites and appears to be almost completely orbital in character. This latter result, which is at variance with gyromagnetic measurements, probably indicates a contraction of the radial wavefunction of states near the Fermi surface. The fit also suggests that about 20% of the magnetic moment does not contribute to the scattering at finite angles.

Specific heat measurements have been carried out on a high-purity single crystal of terbium of mass 3.734 g between temperatures of 200 and 400K. The data around the Neel temperature T_N, have been analysed in terms of the equation C^+or-=(A^+or-/ alpha ^+or-) mod t mod ^{- alpha +or-}(1+E^+or- mod t mod ^x+or-)+B^+or-+D^+or-t where t=(T-T_N^+or-)/T_N^+or-. The parameters obtained from our best fit are alpha ⁺= alpha =^-0.20+or-0.03, A⁺/A^-=0.58+or-0.34, B⁺=B^-=21.68+or-13.93 J mol^-1 K^-1, E⁺/E^-=-2.06+or-4.12, D⁺=D^-=4.99 J mol^-1 K^-1 and T_N⁺=T_N^-=229.9503+or-0.0035K. This fit is valid in the critical regions -3.4<lg mod t mod <-1.5 below T_N and -3.9<lg mod t mod <-2.2 above T_N. The authors find that the relation T_N⁺=T_N^- is obeyed and that the scaling law alpha ⁺= alpha ^- is valid in this region. In order for the condition B⁺=B^-, predicted by theory, to be satisfied they have found it necessary to include the confluent singular term in E in the fit. The value of alpha (=+0.20) that the authors obtain is not close to the predictions of renormalisation group theory for magnetically collinear systems with short-range interactions. A cross-over from the critical region to a region with different behaviour is observed at lg mod t mod =-2.2 above T_N, but not below it.

Reports ellipsometry measurements on ZrZn₂ at room temperature under ultra-high vacuum conditions. The optical constants have been determined in the energy range 0.6-3.8 eV. The results are compared with calculated optical conductivity curves based on the recent band-structure calculation of de Groot et al. (1980). Peaks in the experimental optical conductivity at 0.92, 1.22 and 3.02 eV could be assigned to direct transitions, in agreement with the band model.

The validity of the Lock-Crisp-West (LCW) theorem for analysing two-dimensional (2D) angular correlation of positron annihilation radiation (ACPAR) in V and paramagnetic Cr has been examined theoretically. A band-structure calculation has been made of the two-photon momentum density (TPMD), rho _ep(p), and the occupation density, rho _e(k), for V and paramagnetic Cr. The TPMD, rho _ep(p), has been remapped into k space and the resulting distribution rho _ep(k) has been compared with rho _e(k). The 2D distributions N_ep(k_y,k_z) and N_e(k_y,k_z) have been obtained from rho _ep(p) and rho _e(k) by a suitable integration and folding procedure and they are compared with each other. The results show a qualitative agreement between the two distributions, the observed quantitative difference being ascribed to the effects of the positron wavefunction. The present calculations explain the success of the LCW theorem for analysing the ACPAR data for the V and Cr group of metals.

Calculated angle-resolved photoemission spectra from TiN(100) are presented for photon energies between 13 and 31 eV and at different electron emission angles. The theoretical spectra were calculated using an extended version of Pendry's model for photoemission. A comparison with earlier published experimental results is made and the theoretical spectra are found to reproduce the experimental spectra very well.