Table of contents

The author proposes a new type of cluster in addition to those of Reynolds, Klein and Stanley for the square lattice and another in addition to those of Chao for the triangular lattice for a real-space renormalisation group approach for studying critical phenomena. Thus each lattice is associated with two types of clusters-the longitudinal and the diagonal. The results for the bond percolation and Ising fixed points obtained from the diagonal square and honeycomb clusters agree with the exact values to within 6-11% error; whereas those obtained from the longitudinal triangular and honeycomb clusters agree with the exact values to within 2-4.5% error. The critical exponents corresponding to the percolation and the Ising problems also show reasonably good results: in particular those calculated by using the longitudinal triangular cluster agree with the exact values to within 6.7% and 3.2% error respectively.

The Miller-Abrahams equivalent circuit is used to motivate a simple modification of the pair approximation to AC hopping conductivity which yields transparent analytical formulae showing the correct behaviour at all frequencies including zero.

X-ray and angle-resolved ultraviolet photoemission data on the layered compound NaCrS₂ are presented and related to the electronic structure. The data indicate that the Cr 3d electrons are essentially localised and give rise to a multiplet splitting of the Cr 3s emission line; the magnitude of the splitting is discussed in terms of the magnetic properties of the material and the net spin moment of the Cr ions.

Discommensuration arrays have been seen directly in the electron microscope for the first time in a charge-density-wave-bearing material. The material, 2H-TaSe₂, is also discovered to be orthorhombic in its commensurate phase (projected space group Cmcm) and not hexagonal. This leads to a natural description of the occurrence of the incommensurate stripe phase, observed only on warming. It likewise is orthorhombic. A new double-honeycomb discommensurate model is supported for the symmetric incommensurate phase. The remarkable dynamics and inhomogeneity in this wealth of microstructure is of much relevance to the new generation of fluctuation and transition theory papers on these and related incommensurate systems.

The authors study a two-parameter one-dimensional lattice model embedded in three-dimensional space at T=0, that has indirect oscillatory interactions between neighbours. It can describe succession of modulated structures separated by first-order phase transitions. Qualitative agreement with experimental data obtained by neutron diffraction on thiourea is found.

Liner response formulae for the phenomenological coefficients L_AA, L_AB, and L_BB for matter transport by vacancies are calculated exactly for the five-frequency model of a very dilute FCC alloy of solute B in solvent A. The method reduces each formula to a set of random walk problems in the homogeneous system, and these are solved by Green function methods. In this paper the results for L_BB are shown to be the same as that obtained by Manning in 1964 using the method of equivalent vacancy distributions. The key steps are noted and used in the following paper to calculate L_AB and L_AA.

For pt.I see ibid., vol.14, p.5453 (1981). Linear response formulae for the phenomenological coefficients L_AB and L_AA for matter transport by vacancies are calculated exactly for the five-frequency model of a very dilute FCC alloy of solute B in solvent A. The Green function method used is a straightforward application of the method developed in the preceding paper for L_BB. In the latter case the results are the same as those obtained by Manning in 1964 through the method of equivalent vacancy distributions.

Presents a theoretical investigation of the electronic properties of bulk silicides of near-noble metals (NiSi₂, NiSi, Ni₂,Si, PdSi, Pd₂Si, PtSi and Pt₂Si). The theoretical approach is provided by the iterative extended Huckel method. The densities of states of various compounds, differing in silicon content and in the crystal structure, are presented and the main features of the chemical bond are discussed. A detailed comparison with the available experimental data and with other theoretical work is presented. The relevance of the present results for the interpretation of the silicide/silicon interfaces is also discussed.

Exact formulae and sufficient conditions for a one-dimensional multi-channel localisation scattering are presented. The possibility of strong localisation in a very clean, perfectly cylindrical and rather thick wire is demonstrated.

The influence of non-parabolicity on the velocity operator is derived, and it is argued that the expression used previously by Arora and Jaafarian (1976) is not correct. With the correct form of the velocity operator it is shown that in the limit of vanishing scattering the Hall effect is not influenced by non-parabolicity.

The thermodynamic expressions for the long-wavelength limits of the partial structure factors given by Kirkwood and Buff are supplemented by expressions for temperature derivatives of the partial chemical potentials in terms of long-wavelength limits of energy density-number density correlation functions. These expressions must be modified when the particle numbers of each species cannot be varied independently and the author considers the modifications necessary for electron-ion liquids in terms of the response to hypothetical external fields, both generally and in the adiabatic approximation. In this approximation the ion-ion interactions are density dependent because of the influence of the conduction electrons and he relates first-order thermodynamic changes in the electron-ion liquid to those in a reference liquid, whose ion-ion interactions are identical at the original density but are independent of density. He uses his results to discuss the thermal diffusion resulting from the driving forces occasioned by electron-ion collisions as the electrons conduct heat along the temperature gradient, reducing the response to that of the reference liquid to effective driving forces and the temperature gradient. Returning to the equilibrium liquid, he evaluates his expression for the temperature derivative of the chemical potential of a pure fluid with density-independent forces, showing how it must be modified to apply to a liquid metal.

The authors report high-temperature expansions on spin systems with a combined dipole-dipole and Ruderman-Kittel interaction, the latter component varying in strength and sign. They keep track of both positive and negative transition temperatures to ordered states. The experimentally interesting positive temperature ordering of Cu nuclei should take place above 150 nK.

Small electric field gradients in cubic ferromagnets may be observed using nuclear resonance techniques. The hyperfine coupling generates an interaction of the same symmetry known as the pseudoquadrupole effect. The authors show that this effect is related to the difference in transverse and longitudinal magnetic susceptibilities. Model calculations are performed for a Heisenberg ferromagnet using molecular-field theory near T_c, spin-wave theory at low temperatures, and the itinerant model at absolute zero and T_c. They find that the pseudoquadrupole effect in iron, nickel and GdAl₂ appears to be too small to be observed using current techniques. They therefore infer that the observed quadrupole effects in cubic magnets are due to real electric field gradients.

The green emission at 530 nm in zinc selenide has been studied by means of optically-detected magnetic resonance (ODMR). Observation of hyperfine structure shows unequivocally that copper ions are involved with the emission. The copper ions lie in C_2v symmetry with g_X=1.951+or-0.004, g_Y=2.214+or-0.004, g_Z=2.042+or-0.004, mod A_X mod <10*10^-4 cm^-1, mod A_Y mod =(155+or-5)*10^-4 cm^-1 and mod A_Z mod <10*10^-4 cm^-1, with X, Y and Z respectively parallel to (110), (110) and (001) or equivalent directions. The spectrum is attributed to Cu²⁺ either at the zinc substitutional site or at the interstitial site that is surrounded by four selenium ions. The lowering of the symmetry from T_d to C_2v is attributed to the presence of another ion in an (001) type direction relative to the copper. The results of previous workers suggest that this second ion is itself copper, so that the centre would be a copper-substitutional-copper-interstitial pair. The ODMR results show surprisingly that the emission is due to donor-acceptor recombination, the copper centre acting as a deep acceptor.

The reflectivity spectra of CaAs₃, SrAs₃, BaAs₃, and alpha -EuP₃ in the far infrared region have been measured. The parameters of the long-wavelength optical phonons and the free carriers have been determined using a model of damped Lorentzian oscillators including a Drude term. The results are compared with Raman data.

The K_{beta 5} and L_alpha X-ray emission bands of titanium carbide and the compound TiC_0.65 are considered. Splittings of the bands have been found for the non-stoichiometric compound. Based upon the correspondence between experimental spectra and the results obtained in MO-LCAO calculations on both defect clusters and complete clusters, the authors conclude that the deviation from stoichiometry causes a change of valency for the transition metal. That valency change leads to an increase in the electronic density of states near the Fermi energy mainly associated with metal d character. No evidence for significant electron localisation at vacancy sites has been found either in theoretical or experimental results.

The discrete variational (DV)-X alpha cluster method is applied to elucidate the elementary process of chemisorption of the underlayer type forming a nitride layer at the Ti(0001) metal surface. The electronic structures of a Ti metal cluster (Ti₆) and the chemisorbed clusters (Ti₆N) are calculated. It is found that a characteristic peak of the density of states consisting of odd-parity symmetry orbitals appears in Ti₆ at the Fermi level. The peak may correspond to the surface band at the Fermi surface found in the metal film. A specific energy level of Ti₆ 1.6 eV below the Fermi level is found to be very sensitive to the chemisorption in agreement with photoemission data. This is confirmed by comparing the electronic structure of the Ti₆ system with that of the chemisorbed one in which a nitrogen atom sits at the body-centre site of the Ti₆ cluster. It is pointed out that the elevation of the -1.6 eV level far above the Fermi level in the chemisorption and the subsequent filling electrons in the surface states of Ti₆ would induce a relaxation of the Ti₆ framework helping the penetration of a N atom through the first layer. The change of the electronic structure of the Ti₆N system when the nitrogen atom moves from the outside to the inside of the Ti₆ cluster is also discussed in connection with the deformation process of the Ti metal surface.

The authors present a model calculation of the non-radiative damping of a dipole outside a metal surface. The model treats simultaneously several mechanisms of electronic excitation in the metal and the net damping rate is not simply additive. The dipole frequency is a free parameter so one can examine the decay of both electronic and vibrational excitations. In the low-frequency limit, a friction parameter can also be deduced from their results. Its evaluation is compared with a recent alternative calculation.

Brillouin scattering using surface plasmons in a Kretschmann geometry is analysed theoretically. Numerical estimates for scattering by both bulk and surface acoustic modes indicates that the Brillouin signals are enhanced by orders of magnitude when surface plasmon fields are established at the vacuum-metal boundary for the prism-metal film-vacuum sample geometry.