Table of contents

The recent calculation of Devlin (1982) predicting that the energy of rigid (001) twist boundaries in a central force pairwise interaction model with a long-ranged potential increases linearly with Sigma (for 5<or= Sigma <or=37) is considered. By use of a general argument based on the 'geometry of numbers' it is shown that this linear increase of the energy is too strong to be physically reasonable. Instead, the energy approaches a constant asymptotic value in this range, and any variations can be represented by a higher order term which varies as Sigma ^-34/.

X-ray diffraction experiments on Y₄Co₃ single crystals suggest that the positions of the Co atoms in the channels between the Y prism columns ('c axis sites') are only weakly correlated with the rest of the structure. The Y₄Co₃ phase can be stabilised by silicon such that peritectic rather than peritectoid formation occurs. The substitution of about 2 at.% Co by Si destroys both superconductivity and magnetism.

A field-theoretical renormalisation group method is used to investigate the possible existence of multicritical points in the phase diagram of a ferromagnetic superconductor, which is described by a free energy functional with electromagnetic coupling between the magnetic and superconducting order parameters. The lack of stable fixed points indicates no multicritical behaviour in contrast to the predictions by Horneich and Schuster (1979).

The Ginzburg-Landau equation, previously used for bulk Pd-Ni alloys, is applied to discuss the appearance of magnetic moments at T=0 in a Pd-Ni-Pd sandwich. Using parameters deduced from the analysis of the bulk alloys the authors predict that the critical concentration of Ni atoms is 14.1% for a (100) monolayer and 11.1% for a (111) monolayer.

Small-angle neutron scattering has been used in a study of the precipitation mechanisms of an Al-5 wt.% Cu alloy doped with small additions of Ag and Mg. The experiments were carried out using the D11a instrument at ILL which was arranged to provide an observation range of 0.01<K<0.1 AA^-1 with incident wavelengths between 8 and 12 AA. The high neutron flux available on this instrument permitted the collection of adequate statistics from each sample in a period of approximately 10 minutes. Three reaction temperatures of 120, 145 and 170 degrees C were selected for the experiment which was carried out in two parts: (i) with samples aged 'in situ' for periods of up to 19 hours; (ii) with samples pre-aged for greater periods up to completion of the reaction. The results indicated time- and temperature-dependent features which reflect detail concerning inter-zone/particle separations and particle size distributions throughout the reaction process. It is apparent from the scattering data that the mechanisms associated with GP zone formation in the doped alloy are markedly different from those operative in the classical binary Al-Cu system and the results were found to correlate well with other physical property measurements carried out on the system.

Small-angle neutron scattering measurements have been performed on the amorphous metallic alloy TbCu_3.54 in its as-prepared state and after annealing for 5 h at 240 degrees C. The sample impurity atoms such as hydrogen, argon and oxygen were identified and their concentrations determined. At very small momentum transfer values a very strong scattering is observed which decreases as q^-3 in an oscillatory fashion. A well defined weak ring is observed at about 0.2 AA^-1. Annealing doubles the intensity of the whole pattern. In a model description of the observed small-angle scattering, the alloy is composed of domains several thousand of angstroms wide, in which the impurities tend to accumulate near the walls. The impurities might be diluted in the alloy or concentrated such as in small bubbles. In any case, their distribution within the domains is characteristic. The annealing purifies the centres of the domains and repels the impurities to the periphery.

Using the computer-simulated model structures of two typical metallic glasses (Ni₈₀P₂₀ and Cu₃₃Zr₆₇) reported earlier, the authors present an extension of these calculations which consists of a quantitative description of short-range order effects. Making use of the Bhatia-Thornton number-concentration formalism, partial structure factors and distribution functions are first determined and discussed. The radial concentration-correlation functions are then interpreted in terms of a generalised Warren chemical short-range order parameter; the values obtained for this parameter (-0.13 to -0.14 for Ni₈₀P₂₀ and -0.05 to -0.06 for Cu₃₃Zr₆₇) are found to be in good agreement with the corresponding experimental data.

For pt.II see ibid., vol.11, p.2231 (1981). The stable configurations and binding energies of interstitial-solute complexes in irradiated dilute Al-Cu, Al-Ag and Al-Au alloys were calculated using interatomic potentials derived from first principles. The interactions between the host lattice atoms were governed by the Al interatomic potential of Dagens et al. (1975), while the host-solute interactions were described by pair potentials determined via the recent resonant-model-potential theory of Dagens (1976, 1977). It was found that all three noble-metal solutes-Cu, Ag, and Au-interacted strongly with self-interstitials in Al, forming stable complexes with large binding energies, greater than 0.5 eV. Despite the size differences of these solutes in Al, the stable complexes had the same symmetries; in each alloy, the complexes were bound in three configurations: type b first-neighbour, and type a and type b second-neighbour configurations.

The lattice deformation in a simple metal host due to an interstitial hydrogen-like impurity is studied by extending the jellium density approximation of Popovic et al. (1976) to calculate displacing forces and the method of lattice statics to find equilibrium. The validity of the approach is discussed in the framework of the perturbed electron liquid formalism. It is shown, in particular, that the method is incomplete for the long-wavelength components of the strain field, a generalisation of the earlier result concerning the special case of linear screening. In determining the volume dilatation, the 'slope theorem' connecting density distribution with the derivative of the impurity energy in jellium, recently derived by Stott and Zaremba (1980), proves to be essential. The energy of lattice relaxation around a tetrahedral interstitial is found to be -0.18 eV, the heat of solution is 0.71 eV, near to the experimental value (0.66 eV). The energy minimum is at the tetrahedral site, but the barrier is extremely low so that the energetics seem to allow an almost free motion of the proton along the cubic diagonal. The individual displacements of near neighbours are in reasonable agreement with the available experimental data.

A computer simulation of the static lattice distortions in substitutional potassium-rubidium alloys is presented. Interatomic forces and volume energies are derived from a non-local first-principles pseudopotential; a novel variant of the cluster relaxation technique is used to calculate the equilibrium atomic configuration of finite clusters. It is shown that substantial lattice distortions occur in K-Rb alloys and that they have a non-negligible influence on the heat of formation. The authors find that the distribution of the interatomic spacings and the relaxation energy are well described by the linear harmonic response theory of Froyen and Herring (1981).

The applicability of the tight-binding (TB) type of electronic theory to lattice defect problems of transition metals has been investigated. For this purpose, three kinds of TB schemes, the second moment approximation (SMA), the fourth-order moment approximation (FMA) and the usual TB recursion method (TBRM), were considered. To represent the repulsive interaction between atomic sites i and j, the Born-Mayer potential was used and the parameters were chosen so as to reproduce the correct equilibrium lattice constant and the bulk modulus of the perfect crystal. It was shown that the FMA provides reliable results for the problems of lattice defects of transition metals.

Positron lifetime and annihilation lineshape measurements are used to study defect recovery in pure and H-doped Nb after low-temperature electron and neutron irradiation. In neutron-irradiated specimens correlated vacancy migration and the concomitant vacancy clustering within the collision cascades are observed around 160K. However, in electron-irradiated Nb, no sign of vacancy annealing is seen until above 380K where the vacancies agglomerate strongly. According to earlier ideas this is interpreted as being due to hydrogen impurities bound at vacancies. Vacancy migration and clustering are possible only after the dissociation of the vacancy-hydrogen pairs at 380K. The results also indicate that hydrogen impurities play an important role in the stability of small three-dimensional vacancy clusters in Nb.

Palladium hydride formed in the Pd-C alloy is less stable than in carbon-free palladium, while the hydrides in FCC Fe-Ni-C and Ni-C alloys are more stable than in carbon-free metals. To investigate the effect of carbon on the stability of hydrides, the outgasing process of the hydrogenated alloys is studied by differential scanning calorimetry and Mossbauer spectroscopy. The former measurement gives the values of the heat of reaction and the latter clarifies the carbon-metal, hydrogen-metal and carbon-hydrogen interactions in the alloys. The observed effect of carbon on hydride formation is interpreted by combining lattice distortion theory and electron theory.

The diffusion of ⁶⁸Ge in single crystals of high-purity (over 99.99% pure) Ni has been measured using the sectioning technique. The plot of log D versus 1/T is straight from 939 to 1675K, over 6¹/₂ orders of magnitude in D, with parameters D₀=2.10+or-0.47 cm² s^-1, Q=2.74+or-0.03 eV (264+or-3 kJ mol^-1). The data suggest a low Ge-vacancy binding enthalpy for Ni.

The activation energy of hydrogen in Al and the noble metals is calculated as the sum of the strain field, the configurational and local-mode energies. The prescription of Flynn and Stoneham (1970) is used to estimate the strain-field energy. The self-consistent non-linear screening of the proton and the Blatt correction are not incorporated to calculate the configurational energy. The bare ions of the host lattice are represented by the quasi-non-local model potential which accounts for the d band effects in noble metals. The local-mode energies are taken from experimental measurements and from earlier calculations. The estimated activation energies are found to be in agreement with the experimental values.

The authors have calculated the energy band structure of BCC europium self-consistently by using the KKR method. The density of states curve has been computed by using Jepsen and Andersen's (1971) tetrahedron method. It has been confirmed that the Fermi energy E_F lies near the dip of the density of states curve, from which the electronic specific heat coefficient gamma and the paramagnetic susceptibility chi are obtained, and their temperature variations are examined. Also they have made a quantitative calculation of the electronic density at the nucleus, and obtained the contribution of the conduction electrons to the isomer shift of the ¹⁵¹Eu nucleus: 2 Sigma mod Phi _k(0) mod ²=0.44*10²⁶ cm^-3 atom^-1.

The nearly free-electron self-consistent-field pseudopotential theory is used to analytically evaluate the dielectric function epsilon (k, omega ) for the alkali metals Li, Na and K, the effects of the crystal potential being included up to second order in the U_G pseudopotential form factor, in the high-frequency limit. The plasmon dispersion omega _p(k) is then obtained from the strong peak position of the electron energy loss function. The results are compared with experiment for Li, Na and K, and for increasing mod k mod the discrepancies between the theoretical and experimental plasmon dispersion become noticeable because the calculated dispersion curve is too steep when compared with experiment. The anisotropy of the dispersion is also calculated and found to be rather small for Na and K, while for Li, which has a larger pseudopotential coefficient, the plasmon anisotropy can be quite considerable for large values of mod k mod and should be noticeable experimentally.

The electronic structure of the ordered Zintl (B32) phase of LiAl has been studied in a molecular cluster model within the framework of the Hartree-Fock-Slater theory. Knight shifts and electronic field gradients induced by vacancies are calculated and compared with recent NMR studies. The effects of lattice relaxation about a vacancy are studied and combined with recent neutron diffraction analyses to predict relaxation-induced field gradients.

The low-field (0.005-2T) induced torque in potassium has been examined using a new modulation method of induced torque. An oscillatory field is applied perpendicular to the main field and the direction of torque measurement. An expression for the detected oscillatory torque is presented. The amplitude of the oscillatory torque for a free-electron metal increases linearly with magnetic field at low fields, reaches a maximum at omega _c tau =2 and decreases with increasing field at high fields. The predictions of the properties of induced torque presented by Bishop and Overhauser (1977, 1978) (for the case when the residual resistivity of potassium at zero field is higher in one direction) and by Lass (1972, 1976) (for a sample departing from spherical shape) are compared. It is concluded that the models are experimentally distinguishable in the low-field region omega _c tau <1. A twofold torque anisotropy with peaks separated by approximately 180 degrees was observed at low fields and a fourfold pattern with peaks separated by approximately 90 degrees was dominant at higher fields. The ratio of the torques in the directions of the maximum and minimum of the twofold pattern was measured as a function of this field. This ratio was found to increase with magnetic field as predicted by the Lass theory but contrary to that predicted by Bishop and Overhauser. The zero-field limit of the ratio was in the range 1.05-1.25. The presence of the fourfold pattern at high fields depends on the precise direction of the modulation field. This is explained in terms of the theory of Lass. A reason is given why Hunt and Werner (1974) did not observe the fourfold pattern using an inductive technique. Although the Lass model explains properties of the induced torque, the samples were sufficiently spherical to rule out the direct applicability of the model.

For pt.I see ibid., vol.73, p.839 (1983). Lattice distortion (LD) effects on Dingle temperatures x* and cross-sectional area changes delta A are studied in dilute Cu alloys. An impurity cluster formalism is applied to a cluster consisting of 13 atoms embedded in Cu. First model calculations are reported for a pure Cu crystal with artificial relaxation. The results mainly lead to the conclusion that summing up to l_max=3 is necessary if LD is accounted for. Applying the theory to Si, P, Z, Al and Ge as impurities in Cu it is found that the effect of LD on x* and delta A depends linearly on the shift of the first shell around the impurity. Further it appears that x* is much more sensitive to LD than is delta A. Depending on the orbit the corrections to x* from LD and charge transfer may be of nearly equal importance. Because these corrections sum up rather independently, neglecting both effects may cause errors up to an order of 20%.

The fitted crystal-field parameters for lanthanide atoms in metals are shown to be consistent with the usual assumption that the crystal field is a one-electron interaction. Also, the relationship between crystal-field parameters for different host crystals can be understood in terms of the superposition model. Some consequences of these results in relation to ab initio calculations are discussed.

The electronic and magnetic specific heat of CuNi alloys has been determined in the temperature range 1.5-240K for Ni concentrations between 25 and 70 at.%. For concentrations up to 58 at.% Ni the specific heat at 240K is close to the unenhanced electronic value deduced from a recent CPA band-structure calculation for paramagnetic CuNi. At lower temperatures the electronic specific heat is strongly enhanced in the critical concentration region. Arguments are presented against a superparamagnetic cluster explanation, and it is concluded that the results may best be understood in terms of an electron effective mass enhancement due to their coupling to spin fluctuations localised in regions of high Ni density.

Low-temperature specific heats for amorphous Co_1-xB_x alloys (x=0.16, 0.18, 0.20, 0.22, 0.26 and 0.30) have been measured in the temperature range between 1.5 and 6K. The results are found to fit well to the equation C= gamma T+ alpha T³+ sigma _sT³2/+ sigma _nT^-2. The electronic specific heat coefficient gamma shows only a weak composition dependence and is close to the value for pure HCP Co. This suggests that the density of states near the Fermi level does not change appreciably with varying B content. The Debye temperature always exceeds 300K and increases monotonically with increasing B content. The observed nuclear specific heat coefficient sigma _n allows the effective magnetic field H_eff at the Co nuclear site to be determined. The value of 114 kOe mu _B^-1 is in good agreement with the reported valued derived from the NMR experiment on the amorphous Co₇₅B₂₅ alloy.

¹¹⁹Sn Mossbauer spectra of the Heusler alloy Pd₂MnSn have been studied in the temperature range 4.2K<or=T<or=298K. Although the temperature dependence of the Sn magnetic field is found to scale with the host magnetisation, the recoil-free fraction and the isomer shift display an anomalous temperature variation. As the temperature is reduced from above the Curie temperature T_C, the recoil-free fraction is found to increase while the contact electron density mod n(0) mod ² is found to decrease anomalously. The former result provides evidence for magnon-phonon coupling while the latter provides evidence for electron-magnon coupling which is mediated by phonons.

Measurements of the temperature dependence of the elastic constants of single Q-domain chromium in the cubic (T>T_N), orthorhombic (T_sf<T<T_N) and tetragonal (T<T_sf) phases as well as in the cubic multi-Q phase, are reported. The behaviour of the shear constants near T_sf differs significantly for the single-Q and multi-Q phases. The authors found no magnetic contribution, except near T_sf, to c₄₄ and c₅₅ for T<T_N. In the case of c₆₆, on the other hand, a significant magnetic contribution exists below T_N. The constants c₁₁, c₂₂, c₃₃, c₁₂, c₁₃ and c₂₃ show large lambda -type anomalies at T_N with smaller anomalies at T_sf. The results are discussed in terms of the phenomenological theories of Cowan (1978) and Walker (1980).

The complex multiplet structure of the partially filled 5f shell in the actinides is calculated in intermediate coupling. The eigenfunctions obtained are used for a calculation of photoemission intensities in the fractional parentage scheme including all final states. The calculated spectra are significantly different from the equivalent results for the 4f states in the rare-earth metals presented recently.