Table of contents

Volume 10

Number 3, March 1980

Previous issue Next issue

LETTERS TO THE EDITOR

L95

and

The temperature dependence of the thermal conductivity of indium has been measured in the high-temperature regime for both solid and liquid phases to an accuracy of +or-1%. These data lead to a Wiedemann-Franz ratio that is (i) temperature independent for the solid, (ii) temperature independent for the liquid, and (iii) continuous across the melting point. The value of the measured Wiedemann-Franz ratio is 9% higher than the theoretical Lorenz number.

L101

, and

Starting from the three-dimensional model for grain boundary conductivity theoretical expressions for the conductivity, its TCR and the Hall coefficient are derived. In the limit of a small magnetic field the relative grain boundary Hall coefficient is approximately equal to the product of the relative grain boundary conductivity and the relative TCR.

L105

, , and

Spin glass behaviour of slowly cooled beta -phase CoxGa1-x, compounds is reported for compositions 0.51<or=x<or=0.55. This is indicated by a sharp cusp in the temperature dependence of the AC magnetic susceptibility. Below the spin glass transition temperature the magnetic susceptibility obeys a compositionally independent universal function of the temperature.

PAPERS

337

The internal energy of the normal twins: (1100), (1101), (1102) (1120); (1121), (1122) (1100) in HCP normal metals (Li, Be, Mg, Zn, Cd) has been calculated in the framework of the pseudopotential theory. For each twin misorientation, there appear several relative minima that require translations of adjacent grains from the coincidence position, as was found for grain boundaries in aluminium. The lowest twin energy values are obtained when atoms in the core of the boundary are able to approach a close-packed configuration, i.e. when their coordination number is maximum in a 'nearly' hard-sphere model.

347

, , , and

Measurements are reported of the elastic constants of single-crystal yttrium in the range of temperature 4-300K. Single-crystal samples, with a range of impurity contents, were obtained from a rod that had been subjected to solid state electrotransport. This allowed a study of the dependence of elastic constants on concentration of impurity and permitted a more meaningful comparison of the data with the earlier work by Smith and Gjevre (1960). In particular, the points of inflection that they reported were not observed in the present work.

353

and

The authors have made thermal expansion measurements on Pd-H and Pd-D alloys. The measurements were performed between 9 and 270K, and with H(D) concentrations ranging from 0.69 to 0.82 H(D)/Pd atom ratios. Above 80K (in pure beta phase), the thermal expansion of Pd-H and Pd-D is increased compared with that of pure Pd, for comparable concentrations, D-doped samples have a larger expansion than those doped with H. For PdH0.7(PdD0.7), the authors find an essentially isotope-independent Gruneisen constant gamma H=3.0+or-0.7 for the optical phonons and a Gruneisen constant gamma ac=3.0+or-0.5 for the acoustic phonons.

367

The results of a study, using electrical resistivity, of the solubility of Au in Pb in PbAu alloys in the range of concentration 22 to 650 at PPM Au are reported. In the range of temperature below 493K, the temperature dependence of the solubility of Au in Pb was found to be CAue=1.42*103 exp(-0.61/(kT)). The contribution to the resistivity arising from gold in the solute state depends on the temperature at which the resistivity is measured. The data are compared with previous measurements and the significance of the results is discussed.

377

The variation of the electronic Gruneisen parameter ( gamma e) is evaluated for the high-atomic-number elements thorium and lead by use of the relativistic augmented-plane-wave method and the pseudopotential method respectively. The variation of gamma e in thorium is found to be in qualitative agreement with that obtained from the Thomas-Fermi atom model, but the variation in lead is found to be in the opposite direction. The thermal electronic energy for thorium is also found to depart from the T2 behaviour often assumed in shock-wave calculations.

383

, and

The simultaneous lattice diffusion of 59Fe, 51Cr and 57Ni tracers has been measured in the alloys Fe-15Cr-20Ni, Fe-15Cr-45Ni, Fe-22Cr-45Ni and Fe-15Cr-20Ni-1.4Si between 960 and 1400 degrees C. In all four alloys at all temperatures, DCr>DFe>DNi, with DCr/DNi approximately=2.5 and DFe/DNi approximately=1.8. The activation enthalpies for diffusion of the major constituents in any one alloy are quite close to each other. No correlation is found between swelling under irradiation and the diffusion behaviour of the major constituents. The results are discussed on the basis of the random-alloy model. Grain-boundary diffusion is found in these alloys at surprisingly high temperatures.

399

The classical Friedel sum rule has been extended to the high-concentration region and the self-consistency condition so obtained has been applied to the coherent potential approximation. It is shown that the present approach is useful in realistic calculations for obtaining a condition for the self-consistent potential, which cannot under any conditions be independent of composition.

407

, , and

Results of a calculation of cyclotron masses in Cu and Au for five extremal orbits and for the orientation dependence in the planes (110) and (100) are presented. The numerical integration along the orbit over the Fermi surface was performed using a modification of the well known tetrahedron method. The electronic bandstructure of Cu was obtained from a H-NFE-TB calculation. For Au a relativistic H-NFE-TB scheme was used. The theoretical results are compared with experimental cyclotron masses obtained from de Haas-van Alphen investigations.

417

and

For pt.I see ibid., vol.10, no.3, p.407 (1980). Calculations of Dingle temperatures for 3d transition metal impurities in Cu are presented for five extremal orbits. Also the orientation dependence of the Dingle temperatures in the planes (110) and (100) for Ni and Fe impurities is calculated. The scattering properties of the impurities are described in terms of effective scattering phaseshifts obtained from an impurity model potential of muffin-tin type. The bandstructure of the Cu host was calculated using the H-NFE-TB scheme; the numerical integration along the orbits over the Fermi surface was performed using the tetrahedron method.

427

and

A method developed recently by Williams and co-workers (1979) for the calculation of electronic structure and cohesive properties is applied to Ni3Al, NiAl and Al3Ni and is generalised to make possible the performance of self-consistent spin-polarised calculations which are applied to Ni3Al. Results are given for the densities of states, the magnetic moment and its pressure derivative, lattice constants, bulk modulus and the heats of formation, and are compared with experimental data. In an attempt to isolate dominating binding forces, the heats of formation are decomposed into site- and angular-momentum contributions.

441

and

Structural energies of the three heavy alkali metals, namely K, Rb and Cs, have been calculated in the energy-independent model pseudopotential formalism. For K and Cs, BCC structure is found to have the lowest energy. For Rb, although the BCC phase is favoured over the FCC phase, it is seen to have an energy higher than the HCP phase, but with a relatively small difference. The results of this work have been compared with those from other pseudopotential calculations. An attempt has also been made to compute the structure with the lowest energy for different polymorphs of Cs.

445

, and

The energy-independent model pseudopotential theory, developed and used previously for simple metals, is extended to the binary alloys of these simple metals and a self-consistent pseudopotential, which contains a detailed concentration dependence, is derived for the calculation of various properties of these alloys. This pseudopotential is applied within a low-order perturbation theory to calculate the form factors and transport coefficients for the K-Rb, Na-K and Na-Cs alloys in the liquid state. The calculated results show that a very significant fraction of the valence electrons is localised on the electronegative component in the liquid Na-Cs alloy, as compared with the other alloys considered, and, as a result, the electrical resistivity is very much greater for the liquid Na-Cs alloy than for the liquid Na-K and K-Rb alloys, as demonstrated in experiment. Finally, the applicability of the pseudopotential perturbation theory to the liquid alloys of simple metals is discussed.

461

, and

Measurements of the residual resistivities of the dilute Au-4d alloys AuY, AuZr, AuNb, AuRh and AuPd have been made. These are in agreement with the essential features of the Friedel d resonance scattering picture unlike previous measurements on this alloy series. In one Au-4d alloy, AuPd, close estimates of the s, p and d effective Friedel phaseshifts have been obtained from the resistivity and the Friedel sum rule. The effects of the host lattice on the resistivity of the Au-4d alloys are considered within the approach developed by Coleridge and co-workers (1974) and found to be relatively small.

471

and

Extending the free-electron theory of electronic transport in simple amorphous metals, the Eliashberg function alpha 2F( omega ) and the electron-phonon coupling constant lambda are calculated for such materials. A formula relating lambda to the temperature coefficient of the resistivity at higher temperatures (T>or approximately= theta D/2) is derived. The validity of this formula is tested for amorphous Ga.

477

Measurements have been made of the Hall coefficient of 'pure' potassium over the range of temperature 4.2-16k for fields between 0.01 and 0.5 T. The predicted peak in the magnitude of the low-field Hall coefficient near 6K is visible, but it is an order of magnitude smaller than expected. A new feature is that the low-field coefficient also exhibits a minimum in its magnitude near 12K, and this is replaced by a maximum at fields above 0.25 T. The experiments show that the currently available theories of the Hall coefficient are inadequate.

489

and

The magnetic scattering from the paramagnetic moment aligned by a 4.8 T applied field in technetium has been measured using the polarised neutron technique. The measurements were made at 100K and were scaled to bulk magnetisation measurements made on the same crystals at the same temperature and in the same field. The results yield a paramagnetic form factor for technetium which drops less rapidly with increasing scattering angle than that calculated for spin scattering from the 4d electrons. The shape of the form factor is consistent with a model in which the paramagnetic moment has 70% orbital character.

497

and

The intermetallic compound Fe0.5Co0.5Ti and its hydride phases Fe0.5Co0.5TiHx were investigated through transmission Mossbauer spectroscopy in the composition range 0<or=x<or=1 and a temperature range of 4.2 to 294K. The Mossbauer spectral data and X-ray diffraction data imply that the hydride consists of a mixture of two phases, alpha and beta . The spectra for both Fe0.5Co0.5Ti and the alpha -phase hydride are single lines with identical widths and isomer shifts, and the beta -phase spectrum is a quadrupole doublet. Any spectrum for a mixture of phases can be resolved as a sum of a single line and a quadrupole doublet; the relative spectral area of each component is proportional to the concentration of the phase to which the component corresponds. The Mossbauer parameters in the beta -phase can be explained in terms of the expansion and distortion of the lattice without having to introduce a direct effect arising from the hydrogen.

511

A detailed study is presented on self-absorption of the K-M23, Mg-L23, and Al-L23 edges as well as new data on the anomalous Be-K edge which has already been presented by the author. The self-absorption mechanism is examined in some detail and it is concluded that under favourable circumstances, the absorption spectrum at the edge can be evaluated and its separation from the emission edge measured with a fair degree of accuracy. Broadening and other parameters have been determined for both emission and absorption and are discussed with reference to similar determinations by other workers.