Table of contents

Bragg reflections of NbD_x single crystals in the alpha phase were studied under hydrostatic pressure using gamma ray diffraction. The integrated intensity of the (211) reflection decreases by -1.5+or-0.2%/kbar and -0.7+or-0.1%/kbar in NbD_0.04 and NbD_0.02 respectively. The (200), (110) and (222) reflections show no change of intensity under hydrostatic pressure. In a pure Nb crystal no decrease of the intensity of the (211) reflection was found under hydrostatic pressure.

Examination of defect production rates during electron irradiation of dilute copper or aluminium alloys in the stage II temperature range indicates a correlation between the size-effect of foreign atoms and their capture radii for migrating interstitials. In copper the capture radius increases with size effect, suggesting a mainly elastic interaction. The trend is reversed in aluminium so that other effects probably dominate the long-range interaction.

The submicrosecond resistive pulse heating of wire-shaped metallic samples in a highly incompressible medium leads to a thermodynamic state very close to the critical point of the liquid metal. The additional application of a static pressure may result in a critical or supercritical transition. First results on the critical data of molybdenum are reported: T_c=(11150+or-550)K, p_c=(5460+or-1160) bar, nu _c=(36.5+or-3.5) cm³ mol^-1.

Self-consistent relativistic bandstructure calculations for face centred cubic lanthanum, cerium and thorium have been performed using the linear muffin-tin orbital (LMTO) method and the local spin density (LSD) approximation for exchange and correlation. Results are given for equilibrium distance, bulk modulus, spin susceptibility and the electron-phonon matrix element in each case. A significant contribution of f partial waves to the crystal bonding for all three metals has been found. At the atomic volume corresponding to the gamma phase, cerium exhibits a ferromagnetic instability.

Reports the observation of a plasmon-gain satellite of the KL_2,3L_2,3(¹D) Auger line in Mg and Al metal. For both metals the integrated intensity of this satellite is approximately 1.2% of the parent line; this is in good agreement with an estimate based on a theory presented by Almbladh (1974) and supports the importance of intrinsic plasmon losses as derived from intensity measurements of the plasmon-loss lines.

The author applies the fluctuation formalism through the functional-integral method to obtain an expression for the conductivity of a single-impurity Kondo system. The response to an external field, when evaluated by summing the second-order diagrams in the fluctuation field, gives the electrical conductivity in the 'transition' region. This expression contains the well known high-temperature results as a limiting case.

A rapid energy-dispersive X-ray diffraction (EDXRD) method has been applied to the determination of the ferrite content in austenitic stainless steel (weld metal). No external calibration or standard reference sample is necessary. The analysing times were some minutes. The method was adapted to the measurements of thermal expansion and phase fractions of austenitic stainless steel at temperatures from 17 degrees C to 700 degrees C.

The field dependence of the field evaporation rate of surface layers has been measured for several transition metals. From the data, the polarisability and the effective distance to the metal surface of kink-site atoms are derived.

Calculations of single-vacancy loss to dislocations in high quality single crystals indicate that negligible loss should occur even in air quenches from temperatures approaching the melting point. Experimental deviations from these predictions can be used to determine the importance of divacancies and higher clusters to the loss mechanism.

The elastic constants c₁₁, c₃₃, c₄₄ and c₆₆ of single-crystal neodymium have been measured at 4.2K as a function of applied magnetic field of up to 8 T. The results are in good agreement with the neutron diffraction and magnetisation data for fields in the (1120) and (0001) directions. There is no comparable neutron work to help understand the effects on the elastic constants of a field in the (1100) direction. The magnetic field produces marked anisotropy in the basal plane of Nd, which is strongly reflected in c₄₄ in particular.

A method for the calculation of the formation entropy, per vacancy, in metals is proposed. An application in FCC and BCC metals gave values in close agreement with the experimental ones. It is shown that the entropy increases with the temperature, due to anharmonic effects. This increase must be responsible for the curvature in the Arrhenius plots, observed in some metals. A new method for the analysis of the experimental data is proposed. The introduction of divacancies, trivacancies etc. should be avoided in a number of metals.

For pt.I see ibid., vol.7, p.1651 (1977). The temperature dependence of monovacancy formation parameters such as the formation free energy, energy, entropy and fractional formation volume has been studied for simple metals using pseudopotential theory and the Gibbs-Bogoliubov inequality on the Helmholtz free energy. Application is made to Na and Al. Work presented in a previous paper is extended so that the relaxation in mean position and the shift in vibrational frequency in response to the vacancy are calculated variationally for every ion. This amounts to an extension to finite temperatures of the Kanzaki lattice statics method. The additional flexibility in the model leads to a doubling of the calculated formation entropy for Na over the previous result. Furthermore, the results support the earlier conclusion that, at least for Na and Al, the vacancy formation parameters at high temperatures are insensitive to temperature. This is also found to be the case for the lattice relaxations.

Thermophysical data of solid and liquid Mo and Pd up to the normal boiling temperature obtained by a submicrosecond pulse heating method are presented. These data include the heats of fusion, the thermal expansion and the electrical resistivities. Good agreement is reached with other dynamic and with steady-state data except for the Mo heat of fusion, which comes out to (48.5+or-2.3) kJ mol^-1. Comparative values are 15-33% lower.

The author applies the pseudo-Green function formalism of Moriarty (1972) to the problem of dilute transitional impurities in transition and noble metal hosts. The result is an expression for the electronic structure of these dilute transitional impurity systems which is an appropriate generalisation of Friedel's simple theory and in which the transition metal pseudopotentials appear directly. This theory is shown to predict a virtual bound state in dilute CuNi alloys which arises from the s-d hybridisation and whose position and width are in good agreement with the available experimental data and earlier theoretical calculations.

The authors have found twelve de Haas-van Alphen effect frequencies, in the range 3*10⁶ G to 10⁸ G, for a single crystal of yttrium. The specimen, of resistance ratio 700, was purified by solid state electrotransport at the Centre for Materials Science, University of Birmingham. The frequencies have been identified with orbits on the Fermi surface proposed by Loucks (1966) from an APW bandstructure calculation. Agreement is generally good. In particular the 'webbing' feature which is thought to give rise to spiral magnetic structures in rare earth metals was found to exist, although the thickness is about 17% less than that given by Loucks. The results were all consistent with a spin-orbit coupling large enough to inhibit magnetic breakdown across the AHL face between the third and fourth Brillouin zones.

The de Haas-van Alphen effect has been measured in the ordered alloy beta '-AgMg. The relativistic LMTO bandstructure has been calculated, and predicts a Fermi surface in good agreement with experiment. Both bandstructure and Fermi surface are similar to those of other beta brasses with the same number of conduction electrons per unit cell.

The generalised perturbation method previously introduced for discussing ordering effects in transition metal alloys is extended to include charge transfer effects. A detailed numerical discussion of the validity of this method and of the role of charge transfers is given by using simple models for the electronic structure of the pure metals.

Experimental data describing the equilibrium temperature dependence of positron annihilation in Cu, Ag, Au, Zn, Cd, Al and Pb are considered in terms of the simple trapping model. The results show that there is a correlation between a temperature, T_v, associated with the onset of measurable positron trapping at vacancies and the absolute temperature of melting, T_m, of the metal; T_v approximately 0.6T_m. The correlation may be shown to imply that the product of the specific trapping rate of positrons at vacancies, mu , and the mean positron lifetime, tau , characteristic of the bulk metal, is similar for each of these elements. Values of mu are then extracted from the data using vacancy concentrations based on a present assessment of parameters describing the temperature dependence of the equilibrium vacancy populations of these metals. It is found that values of the product, mu tau , are indeed similar, as indicated by the T_v/T_m correlation. The result can be qualitatively understood in terms of a theory of positron-vacancy interactions which suggests that changes in mu tend to be cancelled by opposite changes in tau .

The nature of electronic bound states on dislocations in crystals is discussed and it is argued that the existence of scattering resonances very close to the Fermi energy may be understood as a many-body effect arising from the screening of bound charges on dislocation lines. The effect of this bound charge upon de Haas-van Alphen frequencies is considered.

In previous calculations of the resistivity the iterative series of the total T matrix of the system has been truncated so that multiple scattering effects have been neglected. In an attempt to include such effects the T matrix is obtained here within the quasicrystalline approximation, resulting in a modified Ziman formula in which the scattering cross section now includes the effects of multiple scattering. Application is made to the liquid metals V, Cr, Mn, Fe, Co, Ni, Cu, Pt, Pd, Ag and Sn, showing good agreement with experiment.

The Boltzmann transport equation for the electrical and thermal resistivities of copper and silver and their dilute alloys is solved numerically for the energy dependence of the electron distribution function, assuming just the lowest order angular dependence. The pseudopotential derived by Dagens (1976) is used to calculate the electron-phonon matrix elements. This pseudopotential gives reasonable values for the transport coefficients at high temperature. At low temperature, the detailed energy dependence of the electron distribution function leads to large decreases in both the electrical and thermal resistivities relative to the values obtained with the simplest form. It also causes significant deviations from Mathiessen's rule (DMR) not only in the thermal resistivity but also in the electrical resistivity. For the latter, the DMR are comparable to those caused by the angular dependence in the distribution function.

A general expression for all elements of the orbital magnetic susceptibility tensor of conduction electrons in a disordered system is derived in the one-electron approximation using the Green-function method. The result is used to calculate the orbital magnetic susceptibility of alloys within the coherent potential approximation by a single-band model and that of dilute impurities and simple liquid metals by the pseudopotential method. Compact expressions for the orbital magnetic susceptibility of these systems are obtained and a numerical result is given for the case of dilute impurities which have a bound state.

Simple expressions for the transverse dynamical spin susceptibility are obtained within the local exchange approximation by neglecting local field effects. The nature and validity of this further approximation is discussed with particular reference to long-wavelength spin waves in a ferromagnet. This paper provides the first practical procedure for calculating the susceptibility solely from the results of a band calculation for the magnetically ordered groundstate.

Nuclear spin-lattice and spin-spin relaxation measurements of ⁶¹Ni metal in the paramagnetic solid phase, and Knight shift measurements extended into the liquid phase are reported. It is shown that the spin-lattice relaxation rate is the sum of an electronic contribution of the form (T,₁T)^-1=(4.6+or-0.2)(sK)^-1, which is mainly orbital in origin, and a spin fluctuation contribution in the vicinity of the Curie temperature. From 820K to 1000K, the Knight shift varies linearly with the susceptibility, yielding a d spin hyperfine field of -(113+or-5)kG mu _B^-1. From 1000K to the melting point (1726K) a departure from this linearity is observed and analysed in terms of the change in the orbital susceptibility due to volume thermal expansion in the solid. Upon melting a further change Delta K=(0.26+or-0.03)% is observed. It is also attributed to the effect of a volume change on the orbital susceptibility.

The influence of the Ho substitution on the Co moments in (Ho_xY_1-x)Co₂ has been investigated by means of magnetic, resistivity and lattice parameter measurements. A first-order phase transition and a field-induced metamagnetism can be detected for x>or=0.7. At the magnetic transition temperature anomalies of the lattice parameters were measured for x>or=0.3. In connection with the first-order transition a discontinuity in the rho (T) curves appears for x>or=0.7. For 0.3<or=x<or=0.7 these curves are characterised by the occurrence of minima above T_c. At high temperatures (T>150K) the rho (T) curves show a tendency to saturation, which is the same for all x values. Considering the substitution of Ho into YCo₂ the existence of a critical RE concentration (x>0.1) for the appearance of a long-range magnetic order is evidenced. The Co moments mu _Co are compared with neutron diffraction results. The overall agreement is very good and supports the assumption of induced Co moments. For the whole pseudobinary system, a Fermi energy situated in a steep flank of the density of states function was assumed.

The spin density functional theory is used to derive a variational principle (lower bound) for the wavevector dependent static transverse ferromagnetic spin susceptibility chi _+-(q) for a general exchange-correlation energy functional. This variational principle is a direct consequence of the Hohenberg and Kohn minimum energy theorem for the ground state. The importance of spin rotational invariance is stressed in choosing the trial function so that the variational expression for chi _+-(q) has the correct behaviour for small q. Noting the relationship between the spin wave stiffness coefficient D and chi _+-(q), a rigorous upper bound for D is derived for general ferromagnetic systems. The relationships between the present general theory for D and recent theories based on the local exchange model are discussed.

The hyperfine fields, determined by NMR, at all the three nuclear sites in gold Heusler alloys, Au₂MnAl and Au₂MlnIn, are reported. These values are compared with those in electronically isostructural copper-based Heusler alloys and discussed on the basis of various theoretical models.

The hyperfine field distributions in solid solutions of FeRh, FeIr and FeMo have been measured at both solute and solvent sites by the use of spin echo nuclear magnetic resonance. The composition range is 0.5-10 at.% of the solute in iron. The distributions have been analysed into constituent lines, each of which is characteristic of a particular nuclear environment.

The nuclear quadrupole interaction at substitutional Ta atoms in the alloy system Hf_xZr_1-x has been investigated over the entire range of composition by time differential perturbed angular correlations (TDPAC). Source configurations investigated include oriented and randomised single crystals as well as textured, cold-worked and random polycrystals, all at room temperature. For both pure Zr and pure Hf, a pure electric quadrupole interaction is found of magnitude nu _Q=eQV_zz/h=303.9(15) MHz and 311.01(11) MHz respectively, which exhibits no electric field gradient (EFG) asymmetry and negligible damping due to finite distribution of the EFG strength. As increasing amounts of impurity are added at either end of the phase diagram, a two-component TDPAC pattern becomes evident: in addition to the undamped pure-metal spin rotation, a second component, exhibiting both damping and asymmetry, is seen at a somewhat higher frequency than that of either pure metal. The fraction of this second component, its frequency, the degree of asymmetry, and the distribution width of the EFG all increase monotonically with increasing impurity concentration.

The Fe site hyperfine field has been measured in the Heusler alloy Ni₂Mn_0.9Fe_0.1Sn using the Mossbauer effect. A value of 193+or-1 kOe was obtained at 110K. The magnetic moment per molecule was measured using a vibrating sample magnetometer. A value of 4.10+or-0.04 mu _B was obtained at 4K. Assuming a value of 4 mu _B per Mn atom, this implies that the magnetic moment associated with the Fe site is 5.0+or-0.4 mu _B. This result has been interpreted using a model for the magnetic coupling in which the dominant interaction aligning the Mn spins is by way of the Sn atoms.

Describes a new method of determination of the complex dielectric constant in the infrared for highly reflecting metals, which takes advantage of the excitation of surface plasmon waves at the metal/air interface. The authors present and discuss the results obtained by this method for Pd, on carefully characterised evaporated films, and the new features which show up in the absorption spectrum.

Presents XPS and UPS photoemission data for polycrystalline RbAu and CsAu. The spin-orbit splittings of the valence bands is employed to interpret the spectra in terms of flat bands. The experimental density of states distribution is compared with new relativistic KKR band calculations. In these calculations a significant charge transfer from the alkali metal to the gold site has been found and, consequently, a calculation starting from an ionic crystal potential has been employed. The resulting bands compare well with the photoemission spectra. The unexpected small photoionisation cross section from the Au d bands and a significant deviation of the surface composition from the stoichiometric bulk composition are discussed.