Table of contents

From the results of X-ray, magnetic and nuclear scattering experiments, the partial structure factors, and hence the pair correlation functions, have been obtained for amorphous TbFe₂. The atoms are highly coordinated on a local level, but with practically no order extending beyond second neighbours. The structure is quite distinct from either crystal or liquid order, but certain gross features are similar to each.

The electric field gradients (EFGs) at the nearest neighbours to impurities in dilute copper alloys have been determined experimentally for a large number of impurities. The principal directions of these EFGs are fixed by symmetry, being parallel and perpendicular to the direction towards the impurity. It is proposed that the most interesting EFG quantity might be the difference Delta between the two perpendicular components. Furthermore it is pointed out that Delta experimentally turns out to be almost proportional to the effective impurity charge Z_eff with Delta /Z_eff=0.22 AA^-3/electron charge. This is in contrast to the other EFG parameters, which vary quite unsystematically with the impurity type.

The lattice dynamics of gold along high-symmetry directions has been studied in the framework of Kreb's model (1965) taking into account the repulsive ion-ion interactions between closed ion shells extending out to two neighbours. Reasonable agreement with the recently measured inelastic neutron scattering data is observed.

An investigation was made within the framework of the Harrison first principles fully nonlocal pseudopotential theory to determine that influence of effective mass on the calculation of phonon spectra. Results indicate that a choice of effective mass significantly influences the pseudopotential and the resulting phonon dispersion relations. These modifications are comparable to those resulting from the choice of alpha in the X alpha approximation for the conduction band/core exchange interaction.

The allowed electron states and electron density of states have been calculated for copper films one, two and four atomic layers in thickness. The results show a severe narrowing of the d-band for very thin films which has disappeared by the time the film thickness reaches four layers. The results of the calculation are supported by UPS data.

Recently it has been shown by Cox et al. (ibid., vol.4, p.807 (1975)) that the effective Coulomb correlation energy of two 3d electrons in an iron series transition metal is approximated by the change in self-consistent Hartree-Fock eigenenergy of a local 3d state that occurs when the occupation of the 3d shell is altered by one. This result has now been applied to the lanthanide series of rare earth metals, and a value of effective Coulomb correlation energy U_ff for the 4f electrons deduced. U_ff is found to be about 9 eV and nearly constant across the table.

The resistivity rho of palladium has been investigated on both sides of the melting point, and the thermoelectric power S has been measured in the liquid phase. The values for rho and S in the liquid are respectively 83+or-2 mu Omega cm and -41+or-3 mu V deg^-1, which are in general accord with the theoretical predictions of Brown (1973). An interesting feature is the comparatively large value of Delta rho / rho _solid across the melting point. Two possible contributions to this effect are identified.

From observations of impurity n.m.r. satellite structure in Fe-based binary and ternary alloys at about 1% concentration, it is shown that there are strong short-range ordering effects for certain impurities.

A peak observed in the soft X-ray emission intensity of an Fe 8 at.% Al alloy lies some 8 eV below the Fermi edge, near the bottom of the Fe 3d-4s bands. There is no such peak in the spectrum for pure Fe, nor has a similar feature been observed with Ni-Al alloys. Al therefore is a strong perturbation in Fe, giving rise to a (virtual) bound state, and it is concluded that the 3d Fe electrons are ineffective in screening an Al impurity, in sharp contrast to Ni.

The X-ray threshold exponents of free-electron metals are determined from impurity resistivity data in the optical approximation: an X-ray excited atom with its core hole is assumed to be identical to an atom 'optically transmuted' into one with an additional 'proton'. This leads to an approximate upper bound of alpha ₀ approximately=0.21 for Na, Mg and Al L_2,3 edges.

The K absorption discontinuity of cobalt has been studied in the pure metal and in six of its intermetallic compounds of the type RCo₅, where R stands for a rare earth element. It is observed that the cobalt K discontinuity shifts towards the low-energy side (negative chemical shift) in all the compounds studied. The negative chemical shifts for cobalt are reported for the first time. The magnitudes of the shifts seem to depend upon the number of electrons taking part in bonding.

The atomic relaxation R_D^a accompanying photoemission from each atomic energy level of cadmium has been deduced from Hartree-Fock calculations. Estimates of the binding energies of free Cd atoms are compared with experimental values deduced from optical and X-ray energies. A comparison of experimental free-atom binding energies with photoelectron measurements on solid Cd shows the extra-atomic relaxation R_d^ea to be 3.3+or-0.5 eV.

Using pseudopotential theory the authors have calculated interatomic potentials for the Li-Mg alloy system for six different concentrations. The Li and Mg pseudopotentials were taken from the work of Dagens et al. (1975) who adjusted the parameters to take into account the nonlinearities in the electron scattering at a single ion site. Using the interatomic potentials the authors have calculated elastic constants and heats of mixing, both of which agree reasonably well with experiment. It was also found that the potentials predict a much greater degree of short-range order than is observed experimentally, indicating that the details of the nearest neighbour Li-Mg interaction are not reproduced accurately. However, it seems likely that these potentials will be suitable for the study of the concentration dependence of the properties of defects such as dislocations, stacking faults and twins.

Evaporation rates of zinc from crystalline alpha -Ag-Zn solid solutions with various initial concentrations were measured using new evaporation equipment. The results, along with concentration profile data and SEM micrographs of the bulk material, confirmed the assumption of porous surface layers in the specimens with initial concentrations in the range 24.3-30.1 at.% Zn. Transport coefficients (including not only volume diffusion but also vapour transport in the pores) were calculated from the evaporation losses. Fairly good agreement was found between the theory of diffusion and the experiment for the alloys with concentration less than 18.3 at.% Zn. Corresponding values of diffusion coefficients were calculated from concentration profiles and evaporation losses.

The introduction of a phenomenological relaxation time to take collision into account in the longitudinal dielectric formulation raises some questions of principle. The author attempts to elucidate the method adopted to deal with these questions, presents a complete derivation of the RPA dielectric function by this method, and makes some observations pertinent to the method.

By a suitable combination inside the muffin-tin sphere of a radial solution to the Schrodinger equation and its energy derivative, the dependence of the APW matrix elements on the energy E used to construct the basis functions can be greatly reduced. This has a number of advantages. The authors present an application of the method, previously suggested by Marcus (1967) and analysed by Andersen to the Chodorow copper potential and examine its precision and its convergence properties. It was found the method converges about equally well or somewhat more slowly than the standard APW method in number of basis functions. The eigenvalue error of the method is proportional to (E-E₀)⁴. The error of the wavefunction is proportional to (E-E_o)². The d states limit the range of mod E-E₀ mod <0.1 Ryd for acceptable wavefunctions. The limit for non-d states is larger than 1 Ryd.

The problem of understanding AB₂ transition metal alloy structural trends within the three Laves phases is approached by considering a common-band model alloy. Tight-binding d band densities of states are calculated for the three phases by the recursion method of Haydock, Heine and Kelly (1972). The three total densities of states are very similar, highly peaked structures. The A site local densities of states show a division into bonding and antibonding states. Cohesive energies are computed and it is found that their differences undergo a cyclic variation across the band, which is related to the moments of the densities of states, and hence to the layer stacking of the structures.

For pt.II see ibid., vol.4, p.1351 (1974). A cluster theory method based on a continued fraction expansion of the Green function has been used to calculate the density of states of a model simple cubic alloy with both diagonal and off-diagonal disorder. For a particular form of off-diagonal disorder the calculation reduces to a simple modification of a cluster theory method presented earlier by Jacobs. For a more general form of off-diagonal disorder the calculation is harder to carry out but results are nevertheless presented. The results show a density of states with a great deal of structure which may be expected to reproduce many more moments correctly than the best previous theory which reproduces only four moments correctly.

The authors extend the method of energy-dependent potentials introduced by one of the authors in an earlier paper (Clark and Dawber, ibid., vol.2, p.930 (1972)) to study the properties of random binary alloys. Some of the approximations made in that paper are removed and the improved model is used to calculate not only the specific heat and resistivity, but also to study the distribution of charge density around the two types of atom in the alloy. Calculations are made on silver-gold and silver-palladium alloys.

The detailed energy band structures of hexagonal beryllium and magnesium have been computed by the composite-wave variational method in conjunction with the quantum defect method. From these energy band data, the density of states and Fermi surface of both these metals have been worked out and compared with the experimental and theoretical results of other workers. The results show a marked deviation of the energy bands of Be from the free electron behaviour, that for Mg being small. The Fermi surface of Be consists of a monster and two identical cigars, while that of Mg, following more or less the free electron nature, consists of several pieces in the form of a monster, cigars, lens, pockets and butterfly. These are found to agree favourably with experimental results.

The thermal and electrical conductivities of samples of as-rolled and annealed 35% Nb-65% Ti alloy have been measured between 4 and 20K and between the superconducting critical and 270K respectively. Analysis shows that the thermal conductivity is almost entirely due to phonon carriers, with the magnetic primarily limited by electron-phonon scattering by an unusually small number of electrons. The electrical resistivity data are analyzed to show that s-d interband scattering occurs most frequently. The enhancement of the electrical conductivity due to the existence of short-range superconducting order above the bulk critical temperature is discussed.

The electrical resistivities ( rho ) and temperature coefficients of resistivity ( delta rho / delta T) have been measured for binary amalgams containing up to 1 at.% of Li, Mg, Ca, Sr and Ba. Amalgams with Ca, Sr and Ba exhibit small but real increases in both rho and delta rho / delta T on initially alloying; both these properties go through maxima at approximately 0.2 at.% solute. Amalgams with Li or Mg give similar results for delta rho / delta T but rho decreases very slowly on alloying. These results are in marked contrast to those for amalgams with solutes from the B sub-groups of the periodic table.

The authors have re-examined earlier measurements of the concentration dependence of the Hall coefficient R_H (Barnard et al., Phys. Rev., vol.176, p.761 (1968)) at helium temperatures over the complete range of composition and have added new data at the gold-rich end. It is found that, apart from a rapid change in R_H with concentration at the extreme ends of the range, R_H varies linearly with composition across the complete series, with a total change of less than 20%. The rapid fall in mod R_H mod on adding 1% solute can be easily explained by the diminished influence of the scattering mechanisms present in the parent metals; for dilute alloys at the Au end the authors have expressed the variation of mod R_H mod with solute concentration in terms of a simple two-band model. For the more concentrated alloys the linear increase in mod R_H mod from gold to silver can then be accounted for by an interpolation of the Fermi surface properties of the pure metals, and no change in the relaxation-time anisotropy is required.

The frequency- and wavevector-dependent transverse magnetic spin susceptibility matrix chi _st( rho , omega ) for a ferromagnet is studied at T=0K within the local exchange (or X alpha ) approximation. General formulae are obtained for a system of Bloch electrons. It is shown that spin waves are obtained as elementary excitations. Applications are made to the model of a ferromagnetic electron gas.

The effect of aluminium addition on the magnetic moments of the constituent atoms of binary BCC Fe-Co, Fe-Cr and FCC Ni-Fe alloys was investigated by comparing the results of Mossbauer and average magnetization measurements. It was found that the Al neighbours do not perturb the moment of Fe atoms in the Fe-Co and Ni-Fe alloys whereas there is a decrease in the Co and Ni moments. This different behaviour of Fe, Co and Ni towards the Al diluent is explained in terms of a simple local model. The simple model does not, however, explain the effects produced by Al in the BCC Fe-Cr alloys.

The magnetization and high-field susceptibility of disordered cold-worked Fe_1-cAl_c alloys with 0<c<or=0.51 have been determined at 4K in fields up to 150 kOe. The well known initial magnetic behaviour is found to extend to about c approximately=0.25. At higher c values the mean atomic moment decreases faster than a dilution but still has not vanished at c=0.51, while an increasingly strong paramagnetic term simultaneously appears. The experimental results, magnetic moments and susceptibilities, are satisfactorily described in a local environment model, where iron atoms carry a moment only if surrounded by at least eight neighbours. This moment is equal to that of pure iron, 2.2 mu _B, for 0<c<or=0.35; it decreases beyond this concentration to about 1 mu _B at c approximately=0.50.

A model for magnetic clustering in spin glass alloys is presented. The susceptibility will show a discontinuous change of slope at the glass transition temperature T_g where an infinite cluster first forms, provided that cluster is superparamagnetically blocked. The statistics of magnetic clustering can be specified exactly on a suitably defined Bethe lattice. In the spin glass limit, i.e. infinite dilution of the magnetic species, it is found that (i) the susceptibility above T_g has the same Curie constant as for non-interacting spins and (ii) the change in slope at T_g is Delta K=2, where K=d(In_chi )/d(InT). The role of inter-cluster interactions and loose spins is briefly discussed.

Magnetic correlations in paramagnetic CuMn alloys containing up to 10 at.% Mn measured using neutron polarization analysis are compared with the predictions of the RKKY interaction and found to agree with the accuracy of the measurements. This implies a ferromagnetic first neighbour interaction. The susceptibility decreases and the atomic short-range order increases in a 1 at.% alloy on annealing at 100 degrees C. The low temperature magnetic order is described more closely by the cluster model of Beck (1972) than by the analysis of Klein and Briut (1963).

Measurements are reported of the Hall effect for thin films of the transition metals chromium, cobalt and nickel. The films were prepared by the condensation of the metal from its vapour phase on to liquid helium cooled, sapphire substrates under ultra-high vacuum conditions.

The variations in the hyperfine fields on the dilute non-magnetic impurities Cu, Ga, As, In, Sb and Au in Fe were measured from 4 to 550K. For FeSb, H_emp(T)/ mu _Fe(T) increases as the temperature is raised; for the other cases this ratio drops with increasing temperature. It is suggested that the mass of the impurity may play a role in the temperature dependence.

The ²⁷Al NMR in ferromagnetic iron was found to be a weaker function of temperature (at constant pressure) than the ⁵⁷Fe NMR. The temperature dependence of the hyperfine coupling constant cannot therefore be attributed to a conduction electron contribution which decreases more rapidly than the magnetization.

The temperature dependence of the susceptibility and the La Knight shift for LaAg and LaZn (CsCl structure) is reported (10-280K) and discussed. Resistance measurements performed on LaZn between 4 and 300K exclude a structural phase transition in this range. It is shown that a general picture of the electronic structures can be formed on the basis of published APW calculations. The relation of these results to properties of the corresponding magnetic compounds (GdAg and GdZn) is mentioned.

L₂ absorption spectra of ytterbium and gold atoms in the intermetallic compounds Yb₇Au₃, YbAu, YbAu₂ and YbAu₃ are recorded photographically using a 400 mm Cauchois spectrograph. The ytterbium edge shifts to higher energy in YbAu₂ and YbAu₃ and to lower energy in Yb₇ Au₃ and YbAu. The gold edge shifts to higher energy in Yb₇Au₃, YbAu and YbAu₃ and to lower energy in YbAu₂. These results are used together with structural information to infer the nature of the chemical bond in these compounds. A systematic variation in the nature of the chemical bond is observed with increasing gold content of the phases.

The author has studied the optical properties of crystalline and amorphous bismuth films. The spectrum of crystalline bismuth is discussed on the basis of a pseudopotential band structure calculation by Golin (1968). The shift of the main absorption peak with temperature is discussed in terms of a pseudopotential calculation. The spectrum of amorphous bismuth could be described well by the Drude equations and the plasmon frequency and electron lifetime thus obtained were similar to those found in liquid bismuth.