Table of contents

Presents the results of a band structure study of the electronic properties of beta FeTiH. The density of states is characterised by a structure at 9 eV below the Fermi energy, resulting from the metal-hydrogen interaction, and the Fermi energy falls in a rising portion to the density of states; the author finds 23.93 states of both spin per Ryd FeTiH at the Fermi level. A comparison with the electronic structure of pure FeTi is also given.

The effect of spin-orbit coupling on the Hall coefficient of simple liquid metals is discussed. Through use of the dynamic equations, an expression for the additional contribution due to spin-orbit coupling is obtained.

In alpha ' Pd-D at low temperatures the deuterons become ordered on the octahedral sites of the Pd lattice due to D-D interactions. As this process does not measurably disrupt the palladium lattice the system provides an excellent approximation to a lattice gas. The Monte Carlo method of predicting the behaviour of the lattice gas model for particular values of the D-D interactions is described and it is shown that for repulsive first nearest-neighbour (V₁) and second nearest-neighbour (V₂) interactions with V₂=0.25 V₁, the Monte Carlo method gives similar results to those obtained by a cluster variation method calculation. However, the latter method only predicts the transition temperature accurately near stoichiometric compositions. The above choice of interactions also gives a good description of the real system for D concentrations between 64 and 68% and it is suggested how agreement at higher concentrations could be achieved by introducing longer range interactions. A method of estimating the values of V_n from diffuse neutron scattering above the transition temperature is also proposed.

The accuracy with which the liquid theory approximate methods describe the structure factors S(k) in liquid metals is investigated. The results of the Barker-Henderson (BH) and Weeks-Chandler-Anderson (WCA) perturbation theories and the Percus-Yevick (PY) and hyper-netted-chain (HNC) equations are examined. Values of S(k)=S_i(k) calculated by these methods for liquid metals are compared with the available computer simulation data using the same inter-ionic potentials phi (r). S_i(k) for Na and Rb calculated with the previously described reliable phi (r), are also compared with experiment. It is concluded from the comparisons that the approximations S_BH(k), S_WCA(k) and S_HNC(k) are not very accurate in liquid metals, particularly near the melting point. S_HNC(k) underestimates the heights of the S(k) peaks while S_WCA(k) overestimates them. Possibilities of using the conclusions for the experimental verification of theoretical phi (r) and in thermodynamic calculations for liquid metals are discussed. An effective iterative method for the solution of the PY and HNC equations is proposed.

Measurements have been made for twelve Cs-Tl alloys at 450 degrees C over the range of 0.65-4.5 eV. The measured reflectances show metallic characteristics for alloys rich in either component. The 40 and 50 at.% Tl alloys have reflectances with distinct semiconducting character. The data are analysed in terms of a simple density of states. Some differences with less ionic alloys are noted.

Dilute copper alloys were irradiated at 80K with 3 MeV electrons up to radiation-induced defect concentrations four times larger than the concentration of the solutes. The resistivity damage rates were measured as a function of the residual resistivity increases and the effective trapping radii were evaluated as a function of the average number of trapped interstitials per solute. In the case of the undersized solute Be the effective trapping radius was found to increase with increasing defect concentration, which can be explained by the growth of solute self-interstitial complexes. The radius increment due to the first interstitial trapped by a Be atom can be estimated to be one half of the recombination radius. The oversized solutes Au, Pd, and Sb behaved differently. The effective trapping radius remains about constant up to defect concentrations comparable with those of the solutes and starts increasing only at higher concentrations.

The phonon spectrum of a liquid rubidium model is investigated in a number of different approximations. The random phase approximation is very accurate in predicting the speed of sound while the Percus-Yevick method is accurate only at the shortest physically realistic wavelengths. The simple Egelstaff formula gives the best results at short wavelengths and the Hubbard-Beeby method is best in the intermediate region.

The shock Hugoniot equation of state for thorium has been computed, using the pseudopotential methods. The nuclear Gruneisen parameters ( gamma ), used in evaluating the ionic vibration contributions have been derived from phonon frequencies. The electronic thermal contribution is obtained using Thomas-Fermi-Dirac (TFD) theory, as the electronic Gruneisen parameter varied with temperature in accordance with TFD theory. The computed P-V curves is in agreement with the data of McQueen and Marsh (1960) although the Hugoniot temperatures at high compressions are different from those derived by them.

The electronic structures of amorphous Fe, Fe_xP_1-x and Fe_xB_1-x (x=0.85 and 0.75) are calculated for realistic, structurally relaxed, models by using the linear muffin-tin orbital method in the atomic sphere approximation and the recursion method. The electron charge density is determined self-consistently and the orbitals of Fe 3d, 4s and P 3p or B 2p electrons are included in the calculation. The resulting density of states agrees fairly well with the observed XPS and UPS. Splitting of the bonding and antibonding states is seen in the metalloid p band. It is shown that the experimental values of the linearly temperature-dependent specific heat in Fe_xB_1-x cannot be explained by an electronic origin. The variation of the magnetic moment with the metalloid atom concentration is discussed.

Electronic energy bands of FCC Rh have been calculated by the KKR method for a wide range of energy and a large number of k points inside the Brillouin zone. The structure constants required for the band structure calculation have been evaluated for a large number of axes inside the Brillouin zone by the improved method of polynomial fitting proposed by Segall and Yang (1980). The authors have used a l-dependent muffin-tin constant to obtain the correct position of the d bands, and from the band structure so obtained, the density of states and epsilon ₂ spectra for Rh are also calculated. The agreement of such calculations with the latest experimental results of Weaver et al. (1977) is excellent.

Two-photon momentum densities and the two-dimensional angular correlation of positron annihilation radiation (ACPAR) in V is calculated by using the band structure method due to Hubbard and Mihnarends (1972). The two-dimensional ACPAR curves N(p_yp_z) calculated for four sets of momentum directions: (i) A: (p_z, p_x)=((001),(100)); (ii) B: (p_z,p_x)=((001), (110)); (iii) C: (p_z,p_x)=((112), (111)); and (iv) D: (p_z,p_x)=((110), (110)) all in the p_y=0 plane, are presented and the results are discussed in terms of the partial contributions from different bands and the Fermi surface topology. The results indicate which parts of the Fermi surface can be examined in each orientation (p_z,p_x). In particular it is shown that the presence of the Fermi surface necks (along Gamma N) in V can alter the two-dimensional ACPAR curves significantly for the orientations B and D. The changes in the two-dimensional ACPAR curves in going from V to VD_0.72 are calculated in the rigid band approximation and these results show interesting features. Long-slit ACPAR curves for V and VD_0.72 are calculated.

Resistivity differences between pure yttrium metal and a solid solution of H or D with yttrium in the atomic concentration range 0.05-0.24 alloy isotope effects and deviations from Matthiessen's rule to be investigated. A model of pair condensation is proposed to explain the transition encountered between 170 and 280K, to calculate the 'liquid-gas' transition energy (about 0.1 eV per atom unit) and also its concentration dependence. The added contribution to the resistivity in the 'gas' phase is attributed to short-range interstitial interactions.

The volume dependence of the electrical resistivity of sodium is calculated for four compressed lattices, whose volumes are 0.9, 0.8, 0.7 and 0.6 of the normal volume. The empirical pseudopotential of Kaveh and Wiser (1972) is used for the electron-phonon interaction and the force constants in the Born-von Karman force constant model are determined from observed values of the elastic constants and their pressure coefficients. The theoretical values of the resistivity are in good agreement with experiment.

Detailed electrical resistivity measurements have been made on nominally pure Ag and on Ag alloys with the aim of testing current understanding of low-temperature transport in metals. The pure samples had residual resistivities between 0.12 and 0.8 n Omega cm, and the alloys, containing up to 2 at.% of Au, Pd or Pt, residual resistivities up to 1 mu Omega cm. Measurements spanned the temperature range 1.2-9K, and were made with a resolution of order 1 in 10⁶. For all except the highest purity (HP) samples, the temperature-dependent part of the resistivity, rho (T)- rho (0), can be described by a simple T⁴ power law over a substantial temperature range, which extends to higher temperatures at the solute concentration increases. Furthermore, the magnitude of rho (T)- rho (0) also increases, with a factor of five to ten between the purest and most concentrated samples. Au and Pt appear to behave similarly as solutes, Pd differently.

The electrical resistivity of the Kondo lattice is studied within the model of Lacroix and Cyrot (79): in this model, the Kondo interaction is transformed into a fictitious s-f hybridisation Jx/2. In order to describe the transport properties, the authors must go beyond a mean field theory and introduce the thermal fluctuations of the hybridisation x. At high temperatures (above T_k), the system may be treated as a collection of incoherent Kondo impurities. The resistivity, evaluated within the phase shift method exhibits a logarithmic decrease as expected. At low temperature, coherence develops between impurities and the system is really a Kondo lattice. Spatial fluctuations (from site to site) of the s-f hybridisation are considered as being responsible for the electron scattering. The resistivity depends strongly on the conduction electron number n: the authors find for n<1 a metallic behaviour with a linear increase and a maximum at a temperature T₁ approximately (1-n)^1/3 T_K; for n=1 the resistivity is that of a semiconductor with p approximately exp(T_K/T).

The CPA theory modified to the first four moments is used to calculate the density of states and magnetic moments in YCo₅. The obtained values of the spin contribution to the magnetic moment of Co for I and II sites are different. Results of calculations are in agreement with recent polarised neutron experiments made by Schweizer and Tasset (1980).

Measurements of spin waves in Fe_0.796Al_0.204 and Fe_0.249Al_0.251 monocrystals using neutron triple-axis spectrometry are reported. The vanishing of spin-wave peaks over the range nu =2.8 to 4.5-5.0 THz for wavevectors in the (111) direction, and their reappearance at higher frequencies, is discussed in terms of a calculated density of states of Stoner excitations for two specified bands of Fe₃Al which show appreciable structure in this ( nu , q) range.

The energy and entropy of a superposition of several modulated spin patterns is calculated from first principles, and the structures which give stationary free energy are identified. The addition of an anisotropic exchange term, of similar form and magnitude to that observed in praseodymium, suggests the conclusion that in neodymium a double q structure should be the most stable. The experimental evidence on neodymium is reviewed and is shown to be rather naturally accounted for by the double q structure, though as yet no conclusive evidence can be offered.

The results of a ¹⁶⁹Tm and ⁵⁷Fe Mossbauer study of the cubic Laves compound TmFe₂ over the temperature range 1.3-550K are presented and discussed. The new results are used, in conjunction with existing NMR, Mossbauer and magnetic anisotropy data for TmFe₂ and Tm metal, to deduce a value of P_4f=536+or-14 MHz for the saturation value of the first excited state of the ¹⁶⁹Tm nucleus. Estimates are also given for the exchange field mu _BB_ex(T=0K)/k_B=153+or-3K acting on the Tm³⁺ ion in TmFe₂, the quadrupole moment of the I=3/2 state of the ¹⁶⁹Tm nucleus, Q=-1.36+or-0.11 b, and the lattice contribution to the nuclear quadrupole interaction in Tm metal, P_c=-54.8+or-5 MHz (for Q=-1.20+or-0.07 b) and P_c=-61+or-8 MHz (for Q=-1.36+or-0.11 b). In addition, estimates are given for the various transferred and parent hyperfine fields in TmFe₂.

Positron trapping at structural defects in lead, tin and some lead-tin alloys has been studied by measurements of the temperature dependence of the Doppler broadening of the annihilation photon line. Monovacancy formation enthalpies have been deduced from the data. In a Pb-8 at.% Sn alloy a sharp transitory peak in the peak height to area ratio or S parameter is observed some 20 degrees below the solvus temperature. An interpretation is offered in terms of enhanced positron trapping at grain boundary sites associated with the initial stages of tin precipitation.