Table of contents

The author evaluates the possibility that stable isotopes under certain conditions can form ordered distributions. It is argued that all the basic ingredients that lead to the spontaneous ordering of impurities or voids are present in the isotopic case as well. Therefore, the search for positional isotopic correlations and 'isotopic superlattices' is a meaningful task.

The density of states and the band structure of the cubic phase of non-magnetic CeOs₂ is determined. The splitting between the different 4f structures is analysed and the experimental results related to the large peaks that appear in the DOS at the Fermi level are interpreted.

Electrons (or holes) added to a commensurate charge-density-wave insulator can bind to form bi-polaron states. Bose condensation of these bi-polarons may occur at relatively high temperatures. This mechanism may be relevant for the recently discovered high-T_c superconductors La_2-xSr_xCuO₄.

The authors study a spin-one-half tight-binding Hamiltonian that includes both a random potential (independent of spin) and spin-orbit couplings. They find numerical evidence for one-parameter scaling and an Anderson transition at a critical value of the disorder. The results agree with perturbation theory for weak disorder and recent rigorous results for strong disorder. In the absence of potential scattering a divergent localisation length of the band centre is found but no apparent divergence in the density of states.

Dyre's work (1986) on the Meyer-Neldel rule in disordered semiconductors leads to the AC conductivity sigma varies as omega ^s. Although his expression s=1-T/T₀, with T₀ the 'compensation' temperature, is characteristic of his model, it is pointed out that the power-law form of sigma is a consequence of non-analytic behaviour of the relaxation spectrum, examples being cited from ion migration in a dielectric solvent, defect-diffusion-assisted relaxation, variable-range hopping and the tunnelling model of a glass.

The lucky drift theory developed by B.K. Ridley (1983) is modified to include the effect of field variation over the 'dead space'. The new theory is applicable to small-geometry logic devices for which the total energy available from the supply voltage is comparable to the ionisation threshold energy. It is shown that the new theory predicts a substantially lower level of multiplication than do conventional theories based on an ionisation coefficient.

The authors have investigated the weak-localisation corrections to the resistance of AuPd wires of rectangular cross section in which the effective width is shorter than the inelastic scattering length. The transverse magnetoresistance of the wires can be described equally well by an inelastic scattering rate tau ₁ varies as T^2/3 or T^1/2. This conclusion can also be applied to measurements of pseudo-one-dimensional electron scattering made by other authors and casts doubt on the identification of Nyquist scattering as the dominant inelastic mechanism. Universal conductance fluctuations observed in the same wires have periods that are consistent with the observed measurements of the inelastic scattering rates.

An argument is presented to explain the fact that the same mean-field results apply to several spin glass models, such as the 'random-energy' model and the large-p Potts glass model. The dilute simplest spin glass model is solved, yielding the same mean-field results. The argument is based on the fact that the symmetry of the discrete-spin system can be determined by the discrete-spin type or by the form of the interactions. The mapping between the models can be extended to all forms of discrete-spin systems.

Diffuse neutron scattering measurements on a single crystal of alpha -YD_0.17 indicate an order-disorder transition between 170 K and 280 K. Above 280 K, a broad ridge of scattering intensity in a plane normal to the (00l) direction at l approximately ⁴/₃ indicates that the deuterons have formed pairs on next-nearest tetrahedral sites along the c direction of the yttrium host lattice. As the temperature is lowered through the transition region, the feature becomes narrower and more intense and shifts somewhat closer to the l=⁴/₃ position. This behaviour can be largely explained by a simple model that assumes a repulsive hard-core plus long-range attractive potential between deuteron pairs along lines in the c direction. As the temperature is reduced these interactions cause the deuterons to arrange themselves in lines along this direction, each line being uncorrelated with all other such lines. The origin of the interaction is presumably electronic and may lie in the nature of the Fermi surface of yttrium.

The selection rules for electron transfer between impurities with respect to spin, orbital symmetry and number of tunnelling electrons have been formulated, and a method of fractional parentage coefficients has been applied.

The speed of a second-sound wave in liquid helium II is known to depend on amplitude, and thermal shocks may form. Prediction of the shock velocity depends on which conservation laws hold across the discontinuity. When dissipation is taken into account a smooth but rapidly varying wave profile is predicted in place of a discontinuity. It is shown how waves of this kind may be used to resolve an uncertainty surrounding one of the conserved quantities across a shock. A number of wave profiles are also exhibited in the vicinity of a critical equilibrium temperature.

A new recursion algorithm which is free from symmetry difficulties in the resultant Green function associated with the axis rotation is developed as a straightforward matrix extension of the conventional recursion method of Haydock and co-workers (1975, 1980). The present method is particularly useful when one needs to calculate the symmetry preserving inter-site Green function from a relatively small number of the recursion coefficients, or whenever the symmetry properties of the Green function is important.

The T(X)(e+t₂) Jahn-Teller system is studied in the strong-coupling approximation on a cluster model by transformation methods. An energy-minimisation procedure is used to show that the system will be localised in potential energy minima having either tetragonal, trigonal or orthorhombic symmetry depending upon the strengths of the e and t₂ coupling constants. Detailed results are given for the orthorhombic case to underline the physical processes involved in the Jahn-Teller effect and provide a detailed mathematical framework for the calculation of other properties of the system (e.g. first- and second-order reduction factors, random-strain effects).

A calculation is made of the phonon-drag contribution S_g to the thermopower of a quasi-2D electron gas coupled to 3D phonons. It is based on coupled Boltzmann equations for the electrons and the phonons. The equations are linearised and solved in a relaxation time approximation. A final formula for S_g is developed for weak electron-phonon coupling which is the case of interest in the low-dimensionality systems currently under experimental investigation.

For pt.I see ibid., vol.20, no.13, p.1985 (1987). A general formula for the phonon-drag contribution to the thermopower of quasi-2D electrons coupled to 3D phonons is evaluated for low temperatures when the electrons are in the electric quantum limit and the phonons are in the boundary scattering regime. The formula is most simply applied to GaAs/AlGaAs heterojunctions. An extension of the theory is required before it can be applied to Si MOSFETs due to the anisotropic nature of Si. In both cases good agreement is found between theoretical predictions and recent experimental results.

The authors present results concerning the magnetic susceptibility of the recently synthesised diluted magnetic (semi-magnetic) semiconductors Hg_1-xFe_xSe, Cd_1-xFe_xSe and Hg_1-x-yCd_yFe_xSe. The high-temperature data (temperature T>or=77 K) obey the Curie-Weiss law. The orbital moment is almost quenched in the system Hg_1-xFe_xSe. The compounds Cd_1-xFe_xSe for x<0.10 exhibit the Van Vleck type of paramagnetism at low temperatures. For x>or=0.10 the susceptibility has a maximum for all the systems studied which is interpreted as the onset of the spin glass phase. The temperature dependence of the specific heat for Hg_1-xFe_xSe shows a broad maximum which is also present in the samples without the corresponding maximum in the susceptibility; thus it cannot be associated solely with the spin glass transition. The dominant exchange integral determined from the Curie-Weiss temperature is of antiferromagnetic character for all the systems studied and diminishes with the increasing band gap. The results are discussed qualitatively in terms of the virtual bound state of the 3d electrons in Hg_1-xFe_xSe which transforms gradually into a localised state with growing content of Cd in Hg_1-x-yCd_yFe_xSe. The authors also discuss the relative roles of the super-exchange and the Bloembergen-Rowland interactions in these systems.

An efficient red-to-green up-conversion arising from Er³⁺ single sites of C_2v symmetry has been observed in CdF₂:0.1 mol% ErF₃, 0.1 mol% NaF at 77 K, under CW dye-laser excitation in a phonon sideband. It is shown that the mechanism responsible for this anti-Stokes fluorescence is a sequential two-photon absorption process via the lowest excited state of the Er³⁺ ion. This up-conversion phenomenon is used to study the excited-state absorptions of two major sites of Er³⁺ in CdF₂:0.1 mol% ErF₃. The excited-state absorption processes cannot be neglected in the interpretation of the excitation spectra of the laser-induced anti-Stokes fluorescences. In addition, the authors emphasise the importance of the two-step absorption in site-selective spectroscopy.

The decay of GR1 cathodoluminescence associated with the neutral vacancy in diamond has been studied in the temperature range 10 to 77 K. Exponential decay of the luminescence is only obtained when short (5 ns) widely spaced (10 mu s) pulses are used, yielding a decay time of 3.1 ns in the temperature range studied. If much longer excitation pulses are used (10 mu s at 10 kHz), or a higher duty cycle (100 ns at 1 MHz), additional structure is observed in the zero-phonon region of the luminescence spectrum identical with that obtained using continuous excitation. Under these conditions the luminescence decay is slower and non-exponential. The value of 3.1 ns, obtained using short widely spaced pulses, is therefore expected to represent the upper limit for the radiative decay time of the centre, and compares with a decay time of 2.55 ns obtained using photo-excitation. Differences between the cathodoluminescence and photoluminescence spectra, and the complex decay of the former, are attributed to the indirect excitation mechanism, rather than to saturation effects.

Hydrogenated amorphous thin films of Si_1-xAg_x have been prepared by a radiofrequency sputtering technique. The absorption coefficient and band gap of the film were determined from the optical measurements within the 0.5 to 3.0 eV photon energy region. For the alloys considered-with up to 6% of Ag-the square root of the product of the absorption coefficient and the photon energy is found to be proportional to the photon energy in the higher-energy region and the optical band gap is found to decrease with the increase in the concentration of Ag. The band gap obtained for a-Si is about 1.5 eV, which is in good agreement with the values reported by previous workers.

Angle-resolved photo-emission spectra for Cu(001)c(2*2)-Cl were obtained by synchrotron radiation between 13 and 45 eV, for two angles of incidence, in order to identify the symmetry types of the initial states. To interpret the spectra in detail, a layer KKR formalism was used to calculate, on the footing of the same effective potential, the photo-emission, layer-projected density of states and bulk band structure. From the good overall agreement between theory and experiment, the authors infer the existence, symmetry and localisation of four adsorbate-induced p/d hybridisation surface states. Furthermore, bulk inter-band transitions of two kinds are identified: (i) those already found on clean Cu(001) and (ii) adsorbate-induced 'surface Umklapp' in the initial state. The c(2*2) adsorbate geometry is impressed on the electron density of the second substrate layer.