Table of contents

Be is a small atom (V=4.89 cm³ mol^-1) compared with Fe and Sn (V=7.09 and 16.30 cm³ mol^-1 respectively). As a result, size mismatch theories predict an attractive component of the potential between Be and Sn when dissolved in iron. The author presents spectra of ¹¹⁹Sn in a FeBe alloy, from which he deduces repulsive interactions at first- and second-neighbour distances. The first-neighbour interaction is in fair agreement with the semi-empirical theory of Miedema and Krolas. The repulsive interaction at the second-neighbour distance contradicts the expectation developed in earlier papers. He also makes some comments on the influence of solute-solute interactions on the deduction of probe-solute interaction potentials.

A metastable BCC Si-substituted Pd-rich Pd_100-xSi_x solid solution Pd(Si) of the A2 type has been obtained in the concentration range 8<x<13 by rapidly quenching the alloys from the liquid state. The solid solution decomposes into the equilibrium phases Pd and Pd₅Si depending on the composition and temperature. Once the Pd(Si) solid solution is formed, it remains stable up to 620 K before it decomposes under heat treatment.

Measurements of the specific heat in the temperature range 2-32 K have been made on a high-purity single crystal of holmium and two polycrystalline samples of samarium. Holmium shows a discontinuous transition at 19.46 K with an entropy of 0.14+or-0.02 J mol^-1 K^-1 and a latent heat of 2.7+or-0.3 J mol^-1. There is a small peak at 17.3 K. Samarium has a large peak at 13.7 K, a smaller peak at 20.4 K and a very small anomaly at 9.6 K. The size and shape of the peaks depends upon the purity of the sample.

Rabin and Klick showed empirically that low-temperature F-centre formation in alkali halides is easy or hard depending on whether the value of a geometric ratio S/D is greater or less than 0.45. The authors show that a similar correlation holds between the ease of F-centre formation at low temperature and the value of a parameter Delta E₂/( Delta E₁+ Delta E₂), where Delta E₁ and Delta E₂ are the calculated energies of odd-parity relaxation of the self-trapped exciton (STE) in modes that do not involve halogen pair translation and do involve halogen pair translation, respectively. These are the local relaxation modes used previously in calculation of the potential curves for on-centre-off-centre motion of the STE, and the values of Delta E₁ and Delta E₂ are obtained directly from such calculations for 12 alkali halides in the present work. The energy barrier that inhibits low-temperature F-centre formation is found to be comparable to the barrier already associated with thermally activated defect formation. That is, the barrier height against defect production does not change greatly from one side of the Rabin-Klick diagram to the other, in contrast to what was held to be the case in previous interpretations of the S/D parameter. What changes is the division of available STE relaxation energy between the productive and non-productive modes. This picture provides a unifying account of both thermally activated and low-temperature defect production in the alkali halides. Furthermore, this description of the Rabin-Klick diagram is based on a model in which low-temperature as well as high-temperature F-centre formation occurs on the lowest adiabatic potential surface connecting the STE pi -luminescent state and the F-H pair.

This paper addresses the problem of calculating the vacancy formation energy in simple metals and presents results for aluminium. The calculations are based on the pseudopotential method, the local-density approximation for exchange and correlation, and periodically repeating geometry. The approach used is similar to that proposed by Car and Parrinello, and allows the simultaneous relaxation of the electrons and the ionic positions. Problems peculiar to metals in this approach are discussed and a way of overcoming them is presented. The calculated vacancy energy (0.56 eV) in aluminium is in quite good agreement with experiment (0.66 eV). Comparisons with perturbation theory show that the calculated value is subject to a technical error of only approximately 0.03 eV and that corrections due to periodic boundary conditions are also of this order. The contribution to the vacancy energy from non-linear effects is similar to the jellium estimate of Evans and Finnis.

The authors report a study of transverse magnetoresistance of thin, rolled potassium films. The measurements in the Sondheimer geometry, when the magnetic field is along the outwards normal to the plane of the film, yield no oscillatory pattern in the transverse magnetoresistance as a function of increasing field. For the high-field regime, with omega tau =10-30, there is a non-saturating linear increase in the magnetoresistance. The Kohler slope of this increase is dependent on the ratio of the film thickness to the electronic mean free path and is up to 20 times the appropriate slope for the bulk metal. From a comparison of these results with the data available for the MacDonald geometry (with the field lying in the film plane), they deduce that the origin of the magnetoresistance increase for the Sondheimer geometry is related to size effects. They discuss several possible explanations for the absence of Sondheimer oscillations, such as a large Hall field, open orbits due to the preferred orientation of the Q-domains associated with charge-density waves and inherent bulk magnetoresistance. None of them seems to provide a sufficient argument for the absence of the oscillatory pattern in the field dependence of the transverse magnetoresistance in the potassium thin films under study.

Soliton excitations in the anisotropic Heisenberg chain are studied in the classical limit by means of the Holstein-Primakoff boson representation, taking into account the complete series. The results obtained are equivalent to those obtained by the classical treatment of spins. The connection with different calculations using limited numbers of terms is established, as well as the relation to the generalised coherent-states method.

Elastic neutron scattering experiments were performed on single crystals of the solid solution RbFeCl_3-xBr_x (x=0.03, 0.06, 0.15, 1.0, 2.56, 2.76, 2.85 and 2.94) to study the magnetic ordering behaviour and the magnetic phase diagram of this mixed pseudo-one-dimensional ferromagnetic-antiferromagnetic system. Small amounts of dopant destroy the magnetic long-range order of the pure parent compounds, RbFeCl₃ and RbFeBr₃. On the RbFeCl₃ side of the phase diagram the magnetic correlations appear finite for x as small as 0.03. The periodicities of these correlations are similar to those of the incommensurate and commensurate magnetic phases of pure RbFeCl₃. On the RbFeBr₃ side of the phase diagram the ordering vector remains at the H point (Q=(¹/₃¹/₃1)_N). At an intermediate composition (x=1.0) there are no detectable one-dimensional ferro- or antiferromagnetic correlations down to a temperature of 1.3 K.

Magnetic properties of the series of ThMn₁₂-structure intermetallic compounds R(Fe₁₁Ti) have been determined for rare earths from Nd to Lu plus Y. The highest Curie temperature (607 K) is for R=Gd, and R-Fe exchange interactions are much stronger for light rare earths than for heavy ones. The temperature dependence of the iron sublattice magnetisation and anisotropy are determined for the Y and Lu compounds. Spin reorientation transitions are found as a function of temperature for the rare earths with a negative second-order Stevens coefficient alpha _J(Nd, Tb, Dy), and a set of crystal-field parameters is derived to account for the transitions in a consistent way. A sharp increase in magnetisation observed for Sm(Fe₁₁Ti) below 130 K in a field of about 10 T applied perpendicular to the easy direction indicates that J-mixing may be important for Sm³⁺. Compared with R₂Fe₁₄B, the iron anisotropy in R(Fe₁₁Ti) is greater, and the rare-earth anisotropy is much weaker at low temperature, with the opposite sign for the rare-earth crystal-field coefficient A₂₀. The average iron moment is 1.7 mu _B in R(Fe₁₁Ti) at 4.2 K; Mossbauer spectra are analysed to yield the average moments on each site. Limits set by the intrinsic magnetic properties on the performance of magnets made from these families of alloys are discussed.

Electron paramagnetic resonance studies on a single crystal of Cu²⁺-doped monopyrazine zinc sulphate trihydrate (PZST) have been made over an extended temperature range (4.2-375 K). Three physically equivalent, but magnetically inequivalent, Cu²⁺ complexes have been observed. The spin-Hamiltonian parameters are rigorously evaluated at 334, 295, 77 and 4.2 K, by the method of least-squares fitting, utilising numerical diagonalisation of the spin-Hamiltonian matrix on a digital computer. The principal values of the g- and A-tensors indicate that, in PZST, the Cu²⁺ ion experiences an octahedral symmetry with orthorhombic distortion. The data are interpreted to conclude the occurrence of both static and dynamic Jahn-Teller effects over the temperature range of investigation, the transition from static to dynamic Jahn-Teller effect occurring at 334+or-1 K.

Brillouin spectroscopy was used to study the elastic properties of LiNH₄SO₄ in the vicinity of the ferroelastic first-order transition occurring at approximately 285 K. Eleven acoustic modes were observed in the temperature range from 273 to 293 K. Their sound velocities and the temperature dependences of the corresponding elastic constants were calculated from the measured frequency shifts. The measured elastic constants exhibited step-like discontinuities at the transition temperature and were interpreted using a symmetry-adapted Landau free-energy expansion. The order parameter for the ferroelectric phase transition at 460 K was chosen to be an infrared-active soft mode; while for the ferroelastic phase transition at 285 K, a two-component soft mode from the X point of the Brillouin zone was chosen. Both of these order parameters are coupled to polarisation and strains.

The electron energy loss and X-ray absorption spectra of two phases of titanium dioxide-rutile and anatase-are reported. Although the nearest-neighbour environments of titanium and oxygen in both these structures are very similar, noticeable differences are observed between the electron energy loss and the X-ray absorption spectra. Attempts to model these differences using real-space multiple-scattering calculations are reasonably successful at the Ti K and O K edges. The authors propose an interpretation of the Ti L_2.3 edge, notably for the presence of a previously unobserved splitting on the L₃ edge.

The author shows that the timescale, tau _g, for rapid growth onset following the nucleation stage, should in general be non-monotonic as a function of quench depth. This behaviour is controlled by the corresponding non-monotonic variation of the critical germ size.

Diffraction model calculations of electrical resistance for strong electron-phonon coupling in two- and three-dimensional alloys, incorporating Pippard-Ziman phonon ineffectiveness, are reported. It is shown that in an electron-phonon scattering based theory, although strong electron-phonon coupling is necessary, it is not a sufficient condition for the production of strong saturation in the normal-state resistivity. Thus, absence of saturation does not preclude strong electron-phonon coupling.

Long-wavelength magnetic excitations in aperiodic layered structures consisting of alternating magnetic and non-magnetic layers are investigated. Real and periodic boundary conditions are used. Spin-wave spectra for Fibonacci superlattices with copper mean and golden mean are compared.