Table of contents

The authors have used the classical molecular dynamics procedure to study the static structure factors of the equimolar liquid Na-K alloy at room temperature. A particularly striking feature of the results is the fact that the main peaks in S_NaNa(k) and S_KK(k) turn out to be decidedly asymmetric about the position of the peak height in contrast to the structure factors of the corresponding pure systems. This effect is qualitatively similar to that resulting from a hard-sphere calculation by Enderby and North (1968) and should be observable experimentally.

Boundary conditions are derived using the known electron density for the perfect crystal to perform the matching calculations on a vacancy cell in order to calculate the electron density within the vacancy cell by the Thomas-Fermi method. The electron density outside the vacancy cell is assumed to be given by n_II(r)=f²(r)n_III(r), where n_III(r) is the electron density for the perfect crystal (e.g. for a jellium model n_III(r)= rho ₀) and f(r)=C sin (r square root epsilon + alpha )/r. C, epsilon and alpha are parameters to be determined numerically such that some boundary conditions are satisfied at the same time. One should note that the authors' numerical results are not self-consistent. The electron densities for Na and Al have been calculated by using the boundary conditions. In spite of the rough approximations used, the results inside the vacancy cell were in good agreement with ones obtained from the self-consistent calculations by Manninen et al. (1975).

The angular force models of de Launay and Clark, Gazis and Wallis are bound to give different eigenvalues for FCC metals. A search for their equivalence should not be encouraged.

A recent paper by Ovadyahu (see ibid., vol.8, p.403 (1978)) purports to have demonstrated the existence of a double-hexagonal close-packed phase of lead. The authors suggest that this interpretation is not correct on the basis of the evidence presented, and that the structure observed is consistent with orthorhombic PbO.

An expression for ideal phonon limited resistivity is proposed. In this formula the electron-phonon coupling is factorised with a temperature dependent phonon part similar to that in the theory of superconductivity after McMillan. The calculation of transition metals resistivity using the Debye model for the phonon spectrum is presented.

The spin wave dispersion relation has been measured in the B2 structure of Fe₃Al up to the frequency nu =9 THz. The dispersion relation is found to be isotropic with a stiffness constant D=168+or-5 meV AA². This result is compared with the value D=139+or-6 meV AA² in the ordered DO₃ structure and with theoretical values for both structures.

In nuclear orientation experiments performed on 3d radioactive ions of ⁵⁴Mn, ⁵¹Cr and ⁶⁰Co dissolved in FCC ytterbium, the occurrence of the 3d magnetism is exactly that found for divalent matrices. At 4.2K the magnetisation of pure ytterbium varies linearly up to 130 kG. These measurements are in agreement with the existence, in a divalent matrix, of localised Yb³⁺ condensed at low temperature in a Kondo state. The authors discuss the possible origin of these Yb³⁺.

A discussion is made on some general characteristics of phonons in topologically disordered systems with strong short-range order. Attention is paid to the dispersion relation of longitudinal phonons. With some characteristic assumptions the numerical calculations of the phonon dispersion relations are performed mainly on liquid alkaline metals and it is pointed out that the softness of the core of the effective ion-ion pair interaction predicted by Hiwatari and Matsuda (1972) seems to play an important role in the lifetime of the phonons and that some general properties, which seem to be caused by the configurational short-range order, are likely to appear, compared with other classical simple liquids, in the liquid alkaline metals with soft cores.

The interaction of migrating Al interstitials with solutes has been investigated in dilute aluminium alloys with Cr, V and Ti. The effective capture radii of impurities for migrating interstitials have been determined from stage I recovery suppression and above stage I from damage rate measurements up to 200K by applying the usual diffusion theory to the data. The capture radii deduced from the recovery data seem to be appreciably smaller than the values extrapolated from the damage rate measurements to temperatures below 50K. Above 50K the capture radii decrease with increasing irradiation temperature. The impurity Ti loses its trapping capability between 86K and 123K, whereas the other impurities do trap interstitials even at 200K. In the case of Al(Cr) the complex formed by an interstitial trapped at a chromium atom seems to change its configuration at 58K.

Elastic after-effect measurements were conducted in a fast electron flux to monitor, at every instant of the irradiation, the atomic mobility, as derived from the kinetics of directional ordering in a Ag-30 at.% Zn alloy. The evolution profiles on application or suppression of flux agree well with the predictions of the model generally used to estimate the defect supersaturations under irradiation. Flux and temperature dependence of every one of the parameters which describe the various phases of the increase of the atomic mobility, as the irradiation proceeds, i.e. initial enhancement rate, characteristic time for the build-up of a quasistationary regime, a quasistationary and stationary enhancement levels, were determined. Again full agreement with theory was observed. These results appear as the first systematic experimental check for the phenomenological model.

The attenuation of 10-90 MHz longitudinal sound waves has been measured from 1.2K upwards in the superconducting mixed state of niobium near B_c2. The attenuation was determined as a function of the directly measured average induction, B, within single crystal specimens which had resistance ratios ranging from 83-3380. The specimens tended to the 'clean' limit (electron mean free path, l>> xi ₀, the superconducting coherence length) in which there is a strong purity dependence of the relative attenuation. For the authors' purest crystals close to B_c2, the results are in reasonable agreement with the purity and field dependence predicted by microscopic theories, and agree over a much wider field range with a phenomenological modification of these theories. Measurements of attenuation and other transport properties in clean type II superconductors by other authors are discussed, and it is shown that many of the evident discrepancies can be attributed to neglect of the effects of magnetic irreversibility and crystalline anisotropy.

The dispersion equation for helicon-phonon interaction for wave propagation parallel to the magnetic field has been solved numerically and sections of the inverse velocity surface by the principal planes have been plotted for four different metals, namely lithium, potassium, sodium and Nb₃Sn. The variation of the velocities of the four different modes with the magnetic field had been studied and it is found that, barring the longitudinal elastic mode, the remaining three hybrid modes show interesting variation with the magnetic field. For Nb₃Sn at low temperatures, the helicon propagating along a direction close to the diagonal direction has a low value for its velocity, though it cannot be called a soft mode.

A simple description of the forces in a homovalent disordered alloy allows a prediction of the excess entropy of mixing in terms of the excess volume and the volumes, entropies and Gruneisen parameters of the pure metal components. To within the usually rather large experimental errors, the theory is quite successful. Some implications for more general systems emerge.

The formalism of lattice statics is applied to the calculation of the lattice relaxation at a crystal surface. Two alternative methods of solution are demonstrated, one based on lattice Green functions and the other on standard difference equation methods. The techniques are illustrated in calculations for the (100) and (111) surfaces of alpha -Fe, in which a simple bond-breaking model of the surface is employed. The methods described are also applicable to certain types of internal surfaces, and a calculation for a twin boundary in an FCC crystal is outlined.

Simultaneous measurements of quantum oscillations in the ultrasonic absorption and dispersion in In and dilute In/Pb alloys have been carried out using a sensitive continuous wave technique. A comparison of the field dependences and the harmonic contents of the quantum oscillations in the absorption, dispersion and magnetisation has shown that the dispersion and magnetisation are in agreement with the existing theories. The absorption oscillations, however, do not follow the well known theories. A modified expression for the absorption lines has been derived by calculating the giant quantum oscillation lineshape in the presence of significant phase smearing. The Dingle temperatures and harmonic content calculated from this modified expression are in excellent agreement with the results obtained from the de Haas-van Alphen effect and the velocity oscillations.

A group-theoretical treatment of the symmetry properties of electron momentum distributions in cubic crystals is given. The contribution to the momentum density from a band not belonging to the totally symmetric representation possesses nodes in certain directions. A selection rule allows the location of these nodes. Tables are presented which summarise the consequences of the selection rule for simple cubic, FCC and BCC lattices, and some examples are discussed.

The scattering of electrons by the long-range elastic strain field of an edge dislocation in a metal is treated in the Born approximation by explicitly incorporating the concept of a limiting scattering angle. Simple expressions are obtained for the field-dependent cross section, reciprocal relaxation time and Dingle temperature. The results are in satisfactory agreement with currently available experimental data. The relation of the theory to other theoretical approaches is discussed.

Cyclotron effective masses of two extremal orbits on the third zone electron arm in lead have been studied for the symmetry direction (110) using the de Haas-van Alphen effect. The temperature dependence of the amplitudes of both the node and antinode for the well known 42.5 cycle beat pattern has been analysed to evaluate each of the two individual cyclotron effective masses, i.e. m_c* and m_n* of the gamma _c and gamma _n oscillations, respectively. Here gamma _c and gamma _n denote the dHvA oscillations arising from the larger and smaller extremal orbits on the electron arm, respectively. The cyclotron effective mass ratios thus determined are m_c*/m₀=0.539+or-0.006 and m_n*/m₀=0.571+or-0.007, where m₀ is the rest mass of the free electron. These mass ratios are favourably compared with the ones recently determined by Onuki et al. (1977) using the Azbel-Kaner cyclotron resonance technique.

Measurements of positron lifetime changes from 100K to the onset of trapping at vacancies in Al, Cu, Ag and Au are presented. The extracted temperature coefficients of positron lifetimes increase with temperature so that they become greater than theoretical predictions based on expansion of a static lattice. In addition, a simple representation of a one-dimensional momentum distribution is used to establish a relationship between changes in positron annihilation rates with temperature and the corresponding changes in shape parameters of the momentum distributions observed in angular correlation or Doppler broadening measurements, permitting a comparison of theoretical predictions of the temperature dependence of annihilation rates with various experimental determinations of the temperature coefficients of the momentum shape parameters.

The resistivity of an amorphous ferromagnet is calculated in the multiple scattering approximation. The scattering at each magnetic site has a static component represented by a real parameter V which characterises the amount of disorder in the system and a dynamic component whose dominant contribution is the inelastic scattering of conduction electrons by diffusive modes associated with the longitudinal component of the magnetisation. For sufficiently large values of V the resistivity has a minimum as observed.

A quantitative path integral calculation of the temperature dependence of the Hall coefficient R_H in potassium has been made for the temperature range approximately 2-60K using a model Fermi surface and an anisotropic relaxation time. The previous disagreement between low-field calculations of R_H(T) and experiment is shown to arise from the high-field contribution present in the experimental results.

The electrical resistance of Al containing Cu, Ag and Ge has been measured for impurity concentrations from about 0.002 to 0.2 at.%, and at temperatures 723-905K, 293K and 4.2K. An experimental technique is employed in which the impurity material is vacuum deposited onto a pure Al foil at high temperatures, and the resistance increment during the subsequent diffusion process is measured. The method allows accurate determination of impurity resistivity at high temperatures by eliminating possible errors due to difference in size factors between pure and alloyed specimens. Results show that, above room temperature, the ratio of the impurity resistivity to the residual resistivity tends to decrease with increasing temperature and impurity concentration. A general feature of the deviation from Matthiessen's rule is established for these impurities in Al by combining the present high-temperature data with low-temperature ones reported by other investigators.

The magnetic susceptibility, chi , of AgMn alloys with concentrations, c, between 10 at ppm and 4 at.% has been measured in the temperature and field ranges 1.4-300K and 0.9-8.5 kOe. The impurity susceptibility has been analysed according to a modified Curie-Weiss law chi = chi ₀+(C/(T+ theta )). The effective magnetic Mn moment is p=(5.1+or-0.1) mu _B and theta =(0.05+or-0.04)K for c<0.1 at.%. For c>or approximately=0.5 at.%, theta is dependent on the concentration and p=(5.4+or-0.1) mu _B. The susceptibility is independent of concentration over the whole temperature range for c<or approximately=100 at ppm and a characteristic temperature T_K approximately=20-30 mK has been estimated from a comparison with the calculations of the temperature dependence of chi by Krishna-Murthy et al. (1975) in which a spin value S=⁵/₂ was used as a scaling parameter. Special attention has been given to the sample control and a reliable determination of the magnetic solute concentration.

A derivation of an effective ion Hamiltonian is presented from an exact starting Hamiltonian describing a substitutional magnetic 3d impurity in an otherwise nonmagnetic metal. The impurity is assumed to have a well defined spin and orbital angular momentum in its groundstate throughout. The starting Hamiltonian contains spin-orbit coupling, the kinetic energy of electrons and nuclei, and the electron-electron and electron-nuclei potential energies. Zeeman energies have been omitted. It is found that the absolute magnitude of the effective spin-orbit coupling is increased over the free-ion value and 'crystal field' energies become considerably modified by the large exchange and Coulomb couplings between localised magnetic 3d and itinerant electrons, in agreement with recent experiments on Au:Co and Zn:Cr. The exchange interaction also gives rise to new zero-field contributions and in axial symmetries, evidence for which has been found in Zn:Cr and Zn:Mn.

For pt.III see ibid., vol.7, p.1051 (1977). The author studies the conditions for the existence of a local moment on a transitional impurity in a transitional alloy. The influence of the host concentration of the local environment of the impurity and of the characteristics of the impurity itself on the amplitude and on the stability of the moment were investigated. Applications to the Rh-Pd (Co), Rh-Pd (Fe) and Nb-Mo (Fe) systems are given.

The magnetisation and Mossbauer effect have been measured for the Invar type FCC Fe-Pt alloys. The atomic moment of the alloys varies linearly with concentration, following the Slater-Pauling curve, which indicates that the iron atom possesses a full moment of 2.7 mu _B even in the Invar region. The hyperfine field acting on the Fe nucleus in these alloys is independent of concentration. These results imply that the Fe-Pt Invar alloys are strong ferromagnets, i.e., weak itinerant ferromagnetism is not responsible for the Invar properties of Fe-Pt alloys, in contrast to the Fe-Ni Invar alloys.

Measurements of the specific magnetisation, sigma , of high-purity samples of the weak ferromagnetic ZrZn₂ in the temperature range 2.5-270K and in pulsed magnetic fields of up to 17 T are presented. The data indicate that ferromagnetism is an intrinsic property of ZrZn₂ and are consistent, below the ordering temperature with the Stoner-Wohlfarth model of very weak itinerant ferromagnetism. The Curie temperature, T_c, is higher than that previously measured with the same samples (27K compared with 21K). It is estimated that T_c increases significantly with thermal expansion in the paramagnetic region but is accompanied by a decrease in the 0K susceptibility chi ₀; thus the prediction of the Stoner-Wohlfarth model of a linear dependence of chi ^-1 on T² is retained. The observed Curie-Weiss temperature dependence in conjunction with the parallel nature of the sigma ² against H/ sigma isotherms is thus inconsistent with this model.

The changes in the characteristic antiferromagnetic properties of chromium produced by allowing with small quantities of platinum (less than 3 at.%) have been studied by neutron diffraction intensity measurements in the vicinity of the (100) magnetic peak as a function of temperature for a series of polycrystalline powders. The results indicate that dilute additions of Pt increase the Neel temperature T_N and decrease T_co and T_sf. At concentrations beyond about 0.6 at.% Pt a state is produced which is commensurate at all temperatures below T_N. This behaviour can be understood in terms of current theories if it is assumed that Pt contributes its excess d electrons to a common band.

Studies of magnetic viscosity in Au-Fe alloys containing 2 at.% and 7 at.% Fe are presented. The time decays of the remanences IRM and TRM are given for a range of temperatures up to and including the spin glass temperature, T_f. All the decays are best described by the relation M=M₀-Snlnt. S_TRM(T), S_IRM(T) both have maxima at 0.6 T_f. The IRM data suggests a linear relation between S and M₀. Plots of S/M₀=V against T suggest the existence of two regimes of temperature below T_f: (1)T<or=0.8 T_f, V_IRM nearly constant, independent of T, and much larger than V_TRM. (2) 0.8<or=T<or=T_f, both V_IRM and V_TRM strongly dependent on T and of nearly the same magnitude. These results at low temperatures can be qualitatively reproduced using a very simple activation model for noninteracting fine particles.

Reports the measurement of the linewidths of the ¹⁹⁹Hg and ²⁰¹Hg nuclear magnetic resonances in liquid mercury from 240K to 420K. From these data the authors have extracted the longitudinal relaxation rates, R₁. For ¹⁹⁹Hg, which has a nuclear spin of ¹/₂, this rate is determined solely by the magnetic dipolar interaction with the conduction electron spins. However ²⁰¹Hg has a nuclear spin of ³/₂ and of large quadrupole moment. Its relaxation rate is therefore a composite sum of the dipolar rate, R_1m, together with an additional quadrupolar contribution , R_1q. This is reflected in the results for ²⁰¹Hg where R₁, which is approximately 3.5*10⁴ s^-1 at 300K, diminishes slowly with increasing temperature. Both R₁ for ¹⁹⁹Hg and the extracted R_1m for ²⁰¹Hg may be explained by the hyperfine interaction with the s-like part of the conduction electron density together with a small contribution from other sources.

The effect of ten species of ion-implanted impurities on the thermal oxidation of copper has been studied. The majority of these increased the oxidation resistance of the copper, the greatest effect (85% decrease in oxidation) being caused by Ti. Only Xe and Cs caused significantly increased oxidation in dry oxygen, but in the presence of water vapour the effect of Ti⁺ implants was reversed and increases in oxidation rate up to 100% were observed. Electrical measurements on devices formed by applying a gold probe to the oxide surface showed a correlation between switching voltage and the effect of the implant on the oxidation rate. A simple model is proposed whereby the observed behaviour is ascribed to the creation of an n type region in the predominantly p type oxide such that inward hole diffusion is inhibited during oxide growth.