Table of contents

The structure factor of liquid aluminium is calculated in the optimised random-phase approximation. The effective pair potential is obtained with Shaw's model-potential form factor including the contribution from the effective masses and the exchange and correlation corrections. The choice of the latter corrections is found to be particularly important at low q.

The metal beryllium (Be) has been studied under high pressure using an electrical resistivity technique with in situ pressure calibration. A sluggish phase transition at 12 GPa is observed which the authors identify as the HCP to distorted HCP phase transition reported in a recent high-pressure X-ray diffraction study. However, a reanalysis of the X-ray data on the high-pressure phase Be(II) indicates that a smaller orthorhombic cell with four atoms fits the observed interplanar spacings. The observed intensity ratios of the diffraction lines can be explained reasonably well on the basis of the space group P2₁2₁2 with four Be atoms occupying 4c positions.

Field-cooled and zero-field-cooled magnetic susceptibilities of a Pt-Mn spin glass containing 10 at.% Mn have been measured between 1.2 and 40K in fields of 1, 10 and 100 mT. The difference between the field-cooled and zero-field-cooled susceptibilities is expressed as Delta _chi =A exp(-E_a/k mod T-T_i mod ), T_i being the temperature at which the irreversibility sets in. A and E_a are constants and E_a is interpreted as the activation energy between the low-lying metastable states in the Pt-Mn spin glass. A and E_a have been found to have the values of 1.12*10^-4 EMU g^-1 Oe^-1 and 1.42 meV respectively. The latter agrees well with the value obtained from effective-field results.

⁵⁷Fe Mossbauer spectra of an amorphous Fe₂₄Zr₇₆ alloy obtained both in the paramagnetic state and in applied magnetic fields show that the quadrupole splitting (QS) is almost exclusively positive. The spectra are consistent with a normally distributed QS of positive mean and a mean asymmetry parameter eta =0.67+or-0.07. The results are in contradiction with the predictions of dense random packing models and indicate the existence of strong short-range order.

The Huang diffuse scattering (HDS) of fully annealed and quenched Pb(Au) single crystals has been measured and compared with lattice parameter measurements in order to determine the defect state of Au in the quenched crystals. The Au concentration was 730 PPM as determined by neutron activation analysis. The double force tensor as determined from the lattice parameter change is Tr P_if=-10.5 eV and from X-ray HDS is mod TrP_ij mod =11.9 eV. Combining both methods the authors conclude that the Au atoms in the quenched state form single substitutional defects. The proposed Au₄ clusters exist in a negligible fraction, if at all, whereas a small fraction of Au₂ dimers (<or=20%) cannot be excluded.

For pt.I see ibid., vol.14, p.1073 (1984). A double-crystal backscattering neutron spectrometer has been used to perform a careful measurement of the difference in lattice parameter between pure Pb and Pb+640 at.PPM Au. The results in the single-phase region at temperatures above 200 degrees C show that the Pb(Au) lattice contracts with respect to pure lead with increasing temperature. This contraction is explained by the conversion of Au₂ dimers to single substitutional defects. A quantitative analysis shows that the substitutional gold atom contracts the lattice by about -1 AA per Au atom, while the Au₂ dimer expands the lattice only slightly by 0.3 AA per dimer. At 215 degrees C the concentration of substitutional gold is 380-500 PPM and the concentration of dimers is 70-130 PPM.

The hydrogen-aluminium atomic force constants and the interionic force constants between aluminium ions are determined by means of the pseudopotential. The inclusion of a proton within a homogeneous electron gas produces a large change in the electron density, so that the electron density around hydrogen cannot be described correctly by the linear screening theory. In obtaining the hydrogen-aluminium atomic force constants, the authors have taken account of the non-linear screening effect by means of the Kohn-Sham formalism. The coupling between interstitial hydrogen and aluminium atoms is found to be fairly weak, and the hydrogen behaves as a free particle in the small region around its site.

Long-range concentration fluctuation (spinodal-type decomposition) with a wavelength of 45 AA has been reported by Wagner, Poerschke and Wollenberger in a concentrated Cu-Ni alloy during electron irradiation. The differences from thermodynamic models in the kinetics of this decomposition have prompted the present investigation of an irradiation-induced process. The system is treated as a dissipative structure in which a small bias in the vacancy-interstitial recombination reaction and irradiation mixing effect in enriched zones of the alloy produces fluxes of interstitials leading to growth and shrinkage of such zones. The system is modelled in a mean-field treatment as a homogeneous lossy medium with a non-linear coupling of the defect concentrations and an averaged parameter characterising these zones. A linear stability analysis shows that an instability can develop in such a solid solution leading to concentration fluctuations. The model yields a temperature- and dose-independent wavelength which is not inconsistent with experimental observations.

Experiments on specific heat (1.3-40K), susceptibility (4-300K) and neutron inelastic scattering (6 and 215K) have been performed on Cu-Zn-Al samples for concentrations close to the onset of the martensitic transformation at 0K. The electronic specific heat term is small and insensitive to the proximity of the transition: the electron-phonon enhancement of the electronic specific heat must be unimportant. For nearly martensitic samples, the phonon specific heat is anomalous: in the range 1.3-4K, the T³ term reaches 0.089 mJ K^-4 mol^-1; at higher temperatures, the results are well described by a Debye spectrum with, essentially, an Einstein peak added, containing about 25% of the modes, and with a characteristic temperature of 78K. The latter figure is in agreement with the frequency corresponding to the Brillouin zone boundary for the softened TA₂ branch as measured by neutrons on one of the authors samples. However, the high mode content of the Einstein specific heat term shows that the softening of this branch extends well outside the direction explored. The neutron results also show that the TA₂ branch is not altered appreciably when the temperature varies from 6 to 215K. This tends to show that the transition is first order. The same indication can be drawn from the strong hysteresis shown by the magnetic susceptibility of the samples exhibiting a transformation, and from the fact that anharmonic processes can be accounted for by mode renormalisation.

High-resolution capacitance dilatometric measurements of the thermal expansion of the a, b and c axes of alpha uranium are reported, with emphasis on the phase transitions at 23, 37 and 43K. Comparisons with earlier X-ray and strain-gauge measurements are made and the relationship of the phases involved to the behaviour of the superconducting transition temperature as a function of pressure is discussed.

The authors present a simple model for the structure and thermodynamics of liquid alloys with strong chemical interactions. Starting from the fact that these systems are characterised by the existence of a non-vanishing ordering potential phi _CC=c₁c₂( phi ₁₁+ phi ₂₂-2 phi ₁₂), the authors model the interatomic potentials by a hard-sphere potential for the mean pair interaction phi _NN=c₁² phi ₁₁+c₂² phi ₂₂+2c₁c₂ phi ₁₂ and a hard-sphere potential of the same diameter plus a Yukawa tail for phi _CC. The volume forces are calculated in an electron-gas model. Based on the solution of this model in the mean spherical approximation, the authors find analytic expressions for the thermodynamic excess functions and the partial structure factors. They have applied their model to liquid Li-Pb, Li-Ag and Li-Mg alloys and find very good results for the heat and the entropy of formation and good agreement between their structure data and neutron scattering experiments.

The cohesive energies of the light actinide metals have been calculated from first principles using the density functional formalism and self-consistent RLMTO energy band calculations. The band contribution to cohesion increases across the series but the contribution from free-atom preparation energies increases more rapidly, resulting in cohesive energies that decrease across the series. The modification of the band contribution to cohesion by spin-orbit splitting is significant in the middle of the series.

The parallel resistor formula describing the saturation phenomenon of electrical resistivity in systems with strong electron-phonon interactions is derived. The higher-order terms in the electron-phonon interaction are described by a self-energy which is determined self-consistently.

The glass-forming region in liquid-quenched Ag-Cu-X (X=Mg, Al, In, Si, Sn and Sb) alloy systems was investigated by utilising the melt-spinning technique. The electrical resistivities were measured in the range 2-700K and the low-temperature specific heats in the range 1.5-6K for these Hume-Rothery-type simple metallic glasses. A distinct difference is found to exist in the low-temperature electronic properties between the low-resistivity Ag-Cu-Mg and Ag-Cu-Sn alloys and high-resistivity Ag-Cu-metalloid alloys. The temperature dependence of the electrical resistivity can be discussed within the framework of the extended Ziman theory. The sign of the TCR can be well correlated with Ziman's parameter 2k_F/K_p only for the low-resistivity alloys. The saturation effect develops in the high-resistivity alloys. This can be easily identified from the unique temperature dependence of the resistivity at low temperatures below about 100K. The electronic specific heat coefficient data for the low-resistivity Ag-Cu-Mg alloys can be well understood in terms of the free-electron model. Furthermore, it turns out that the 2k_F=K_p condition does not noticeably affect the density of states at the Fermi level. However, some depression of the density of states at the Fermi level is apparently present in the case of Ag-Cu-metalloid metallic glasses.

Measurements of the Hall coefficient, the angular correlation of the positron annihilation, the low-temperature specific heat, and the electrical resistivity are reported for the simple metallic glasses Mg_0.7Zn_0.3-xGa_x in the composition range 0<or=x<or=0.2. The Hall coefficient is negative and temperature independent between 78 and 300K. It becomes more negative with increasing Ga concentration, indicating the decrease in the carrier density. This is in contradiction to the free-electron prediction. The one-dimensional Fourier transform of the angular correlation curve for the Mg_0.7Zn_0.24Ga_0.06 alloy also indicates a deviation from the free-electron curve. The electronic specific heat coefficient decreases slightly with increasing Ga content, again being opposite to the trend expected from the free-electron model. All these results are consistent with the departure from the free-electron behaviour when Ga is added to the free-electron-like Mg_0.7Zn_0.3. The electrical resistivity at 300K, rho _300K, increases from 54 to 94 mu Omega cm when Ga is substituted for Zn up to x=0.2. It is also found that the detailed temperature dependence of the electrical resistivity at low temperatures depends strongly on the magnitude of rho _300K. The electrical resistivity data above 300K, in combination with X-ray diffraction studies, revealed that the addition of Ga changes the crystallisation process drastically and contributes to suppress the formation of the metastable Mg₅₁Zn₂₀ compound.

The transfer matrix of a solid described by the stacking of principal layers is obtained by an iterative procedure which takes into account 2" layers after n iterations, in contrast to usual schemes where each iteration includes just one more layer. The Green function and density of states at the surface of the corresponding semi-infinite crystal are then given by well known formulae in terms of the transfer matrix. This method, especially convenient near singularities, is applied to the calculation of the spectral as well as the total densities of states for the (100) face of molybdenum. The Slater-Koster algorithm for the calculation of tight-binding parameters is used with a basis of nine orbitals per atom (4d, 5s, 5p). Surface states and resonances are first identified and then analysed into orbital components to find their dominant symmetry. Their evolution along the main symmetry lines of the two-dimensional Brillouin zone is given explicitly. The surface-state peak just below the Fermi level (Swanson hump) is not obtained. This is traced to the difficulty in placing an appropriate boundary condition at the surface with the tight-binding parameterisation scheme.

The authors have studied the temperature, magnetic and electric field dependence of the electrical resistance of thin Cu/Pb contacts. The results demonstrate the coexistence of superconducting fluctuations and weak localisation effects well above the critical temperature. The magnetoresistance data are explained on the basis of the theories of Hikami et al. (1980) and Larkin (1980) for two-dimensional disordered systems. Values of the inelastic scattering time tau _i, the spin-orbit scattering time tau _so and the scattering time from magnetic impurities tau _s were obtained for this proximity system.

The charge and spin form factors have been analysed away from the reciprocal lattice points in crystalline solids. The relevant formulae leading to the form factors have been discussed for a static as well as a thermally vibrating lattice.

A previously developed microscopic theory for the wavevector-dependent spin susceptibility chi (q) of paramagnetic iron based on the Hubbard-Hasegawa approach has been applied to nickel. The equal-time spin-spin correlation function evaluated with the use of the calculated chi (q) and the classical fluctuation-dissipation theorem accounts well for the observed feature in the integrated neutron scattering intensity. The magnitude of local magnetic moments and the spatial correlation function are also calculated by summing chi (q) over the FCC first Brillouin zone, and they are shown to be consistent with experimental data.

Magnetic and ⁵⁷Fe Mossbauer measurements are reported for Y(Fe_xAl_1-x)₂, 0<x<or=0.8. The mean effective moment per Fe atom is only mildly dependent on concentration. In contrast, the mean spontaneous moment decreases rapidly and the long-range ferromagnetic order disappears at about x=0.78. For lower iron content short-range order and freezing effects are observed, typical for mictomagnetism. The low-temperature data were interpreted as the sum of a susceptibility and of a field-dependent magnetisation caused by magnetic clusters originating from the statistical replacement of Al for Fe. These clusters exist throughout the whole concentration range. From both magnetic and Mossbauer measurements indications for the presence of Fe moments were obtained even at high dilution.

The stress dependence of the Neel temperature T_N of chromium, which is detected by means of the ultrasonic attenuation peak at the first-order transition, is measured for compressive stress sigma along a (100) and a (110) axis. It is found to be isotropic within the experimental accuracy, having a value of dT_ZN/d sigma =-1.5+or-0.4K kbar^-1. Both ultrasonic and neutron scattering measurements fail to reveal the expected approach to a tricritical point for sigma along (110) because of irreversible broadening of the transition at values of sigma of the order of 100 bar.

New measurements are reported which show substantial agreement with published values below 1K and above 15K, but very large differences at intermediate temperatures, the new values being smaller. Analysis of the results shows that (i) the hyperfine specific heat is consistent with the results of nuclear magnetic resonance experiments, (ii) the electronic specific heat coefficient gamma =10+or-0.2 mJ mol^-1 K^-2, and (iii) the ferromagnetic-antiferromagnetic transition temperature is 19.2K. The temperature dependence of the magnetic specific heat below 7K appears to be anomalously rapid.

The authors present a detailed comparison of the data on magnetic susceptibility and resistivity as a function of temperature with the predictions of the theory of weak itinerant ferromagnetism which takes spin fluctuations into account. The agreement is good provided the authors decompose the paramagnetic susceptibility into the Curie-Weiss and the temperature-independent parts. They also compare the results obtained for Y₄Co₃ with those for other weak ferromagnets. The Stoner enhancement factor and the molecular field at T=0 are determined. The authors also draw some conclusions about the electronic structure of Y₄Co₃ (or Y₉Co₇) above the superconducting transition temperature.

⁹³Nb and ¹H Knight shifts and nuclear relaxation rates in NbH₂ have been measured in the temperature range 120-300K. The separation of (T₁T)_Nb^-1 into the s-spin, d-spin and orbital contributions is reported and the resulting densities of states N_s(E_F) and N_d(E_F) are discussed.

A number of applications of the calculational scheme developed by Puska and Nieminen (1982-3) are reported and the predictive power of the scheme is substantiated. Effects on positron parameters of relaxation and of N or H impurities in vacancies in Mo are calculated and employed to analyse recent experiments. Predictions pertaining to H decoration of vacancies in Al and Ni suggest the use of positron lifetime studies of these systems. Positron responses to submicroscopic vacancy clusters decorated with Kr and to large Kr bubbles in Cu are calculated and used to analyse recent experiments. To accomplish this the scheme is generalised to incorporate crystals of inert gas. In turn this makes it possible to estimate the average enhancement factors of the inert-gas crystals. The cases considered in this paper indicate that the potential of the positron techniques can be enhanced further by computational analysis.

Experimental angle-resolved photoemission spectra from Cr (001) are compared with the relevant intensities calculated in a 'one-step model' theory. The calculations reproduce most features observed in the spectra and relate them to the bulk band structure. This confirms the interpretation of the experimental band structure deduced previously from the photoemission data by means of the usual 'three-step model' assuming a free-electron-like final band dispersion. Yet the measured d bandwidth is about 20% smaller than predicted by standard one-electron band-structure calculations and a satellite is seen at about 7 eV below the Fermi level, suggesting that correlation effects are essential to the understanding of photoemission from Cr. Using perturbation theory (limited to second order in U/W where U is the Coulomb integral and W the d bandwidth) within the Hubbard model, the authors calculate the corrections induced by these electronic correlations: good agreement with the 'experimental' band structure, the observed d-band narrowing, the hole lifetimes and the satellite position is obtained for U=1.5 eV.