Table of contents

Amorphous Al₉₀Fe₅Ce₅, prepared by rapid solidification from the melt, shows unusual tensile strength. The authors have investigated various physical properties of this alloy. X-ray diffraction measurements indicate a mean interatomic spacing comparable to that in FCC aluminium. Thermal analysis studies show an onset of crystallisation at 543 K, corresponding to the precipitation of FCC Al in an amorphous matrix. A second exothermic reaction at 593 K corresponds to the crystallisation of the remaining amorphous component into FCC Al and an intermetallic phase, possible Al(Ce, Fe)₃. The electrical resistivity is well described by the Ziman-Faber theory with a Debye temperature of approximately 200 K. Magnetic susceptibility shows the alloy to be paramagnetic and to exhibit combined Curie and Pauli behaviour. ⁵⁷Fe Mossbauer effect measurements show the short range ordering of this alloy to be similar to that found in transition metal-metalloid glasses.

In undoped paratellurite ( alpha -TeO₂) electron-irradiated near room temperature, in addition to the V_O^. centre, two further intrinsic, strongly anisotropic radiation defects have been characterised using ESR and ENDOR spectroscopy. The first new centre is assumed to be an aggregate of two oxygen vacancies with a net positive charge and denoted as (V_O)₂^.. According to this model the unpaired spin is mainly distributed along with a quasilinear Te-O-Te-O-Te chain with the vacancies in short-bond positions relative to the central tellurium. Based on apparent analogies a similar delocalisation is suggested for the V_O^. centre as well. For the second new centre the 'small peroxy radical' model originally proposed for SiO₂ by Edwards and Fowler (1982) has been adapted, where an oxygen interstitial is trapped next to a lattice oxygen, the latter also maintaining its bonds to both Te neighbours (Te-O₂-Te).

The off-centre displacement delta of a Li ion in KTaO₃ and the surrounding lattice distortion is calculated using a nonlinear shell model developed for the pure lattice. The method of the static lattice Green function is applied and the range of the Li dipole forces is limited to a few lattice constants. A short range Li-O interaction potential is used which leads to a lithium displacement in reasonable agreement with estimations made from NMR measurements. A lattice polarisation is found to develop in a region around the Li, whose total dipole moment exceeds by a factor of 4.5 that arising from the Li displacement itself. This is in satisfactory agreement with the observed dielectric properties. The shape of the polarised region is strongly anisotropic, the longer axis being coincident with the (incipient ferroelectric) polar axis.

The stability of SiC polytypes has been discussed elsewhere in terms of successive, pairwise, interplanar interactions J_n between Ising-like pseudo-spins. Here the model is first critically re-examined to resolve some doubts about the generality of the energy expression and the meaning of the J_n. The original expression is validated as the correct most general form, with the four-spin coupling term K re-interpreted as a modification of the third-neighbour interaction depending on the local geometry. Secondly the authors note the moderately long range of the interaction to third-neighbour atomic double layers found in density functional calculations on Si and SiC. They argue that this can be interpreted as the remnant in the semiconductor of what would be Friedel oscillations in a metal.

The authors report a strict comparison of long-wavelength optical lattice vibrations in a superlattice obtained with the macroscopic continuum model and a closely parallel microscopic model. It is shown that the proper envelopes of the atomic displacement patterns can be found if the uncertainties inherent in any dispersionless theory such as the continuum model are appropriately treated. Apart from small discrepancies due to the discrete nature of the matter and effective material layer thicknesses the main failure of the continuum theory is the neglect of bulk phonon dispersion.

Cohesive energies, bulk moduli and equilibrium lattice constants have been calculated for the 4d and 5d transition metals with face-centred cubic crystal lattices (Rh, Pd, Ag and Ir, Pt, Au). For the total energy calculations according to the density functional theory on the local density approximation the authors have used an ab initio pseudopotential method. Two calculations have been performed for each element using either nonrelativistic or scalar-relativistic ionic pseudopotentials. The pseudo-wavefunctions and charge densities of the valence electrons have been represented by a mixed basis of plane waves and localised orbitals derived from the atomic d pseudo-wavefunctions. For the 5d metals they find a significant improvement of the results by the relativistic treatment, as expected because of their heavy atomic nuclei. In the case of the 4d metals the relativistic results are of similar quality as for the 5d metals, but now the nonrelativistic values are slightly closer to the experiment, possibly due to an error cancellation effect.

A means of correcting total energy pseudopotential calculations performed using a fixed cut-off energy for the plane waves in the basis set is presented. The use of a finite set of special k-points in such a calculation will introduce errors in the total energies which decrease only slowly with increasing cut-off energy. In particular, total energy differences are not accurate unless the cut-off energy used is sufficiently large that the total energies are themselves converged. This would not be the case if a truly constant cut-off energy could be used. Unfortunately this can only be achieved by using an infinite k-point set. The authors have derived a correction which will explicitly eliminate these errors to give total energies which can correspond to a strictly constant cut-off energy. In this way, total energy differences and hence many physical properties can be accurately calculated using cut-off energies significantly lower than otherwise possible, with substantial savings in computational time.

The authors investigate the character of the native defects (antisites, interstitials, vacancies and complexes of these or with impurities) in undoped and n- or p-type modulation-doped AlAs/GaAs superlattices by means of the developed recursion method. The presented electronic structures, and local and partial density of states give a detailed identification of these defects in various types of AlAs/GaAs superlattice. They have calculated the energies of highly localised electronic states and compared the results with the existing theoretical and experimental data. The stable states of the EL2 defect (As_Ga-I_As pair) are studied under the assumption that it has the same atomic configuration as that in the bulk GaAs. With respect to the energy distribution, the electronic occupancy of various defects and of their atomic orbitals are determined; these show the occupied charge states directly. On the basis of these calculations, they find that the ionicity of a normal group III or V atom in AlAs/GaAs superlattices will always be positive when it becomes an interstitial and will be changed with a different type of doping, when it becomes an antisite. By considering the Coulombic nature of a Frenkel pair, some specific complexes between the native defects themselves or with the impurity Si are suggested. Finally, they describe the self-diffusion and the impurity diffusion mechanisms which occur in AlAs/GaAs superlattices with and without doping. Results are compared and contrasted with those of the bulk GaAs materials.

While the evolution of the nonlinear quantum dimer can be solved exactly in the adiabatic limit, only numerical methods and general analytical arguments have been used in the nonadiabatic case. The authors present approximate analytical expressions for the dependence of the complete density matrix of the system through an application of an averaging method. The method consists of selecting a constant of the motion for the adiabatic case, studying its time evolution, and using the results to obtain the evolution of the entire system.

The authors have examined some aspects of the ground state of a quantum two-dimensional electron gas interacting with ln(r) potential. It is shown that the divergence appearing in the Hartree-Fock contribution to the ground-state energy exactly cancels the divergence in the electrostatic contribution. The next leading contribution to the ground-state energy is explicitly calculated and is also found to be finite. Further, the exact asymptotic expression for the structure factor at large wavevectors is obtained using the exact asymptotic solution of the two-electron Schrodinger equation for small separation r.

The authors have studied the electrical conductivity of n⁺-n^--n⁺ GaAs structures in which the thickness of the n^- layer is comparable with the mean donor separation. At low temperatures, electrical conduction is dominated by a process in which electrons tunnel across the sample via shallow donor impurities close to the centre of the n^- layer. By studying the magnetoresistance of such samples they have investigated the effect of a magnetic field on the donor wavefunction in GaAs. Measurements are compared with the approximate analytical expression for the donor wavefunction and with numerical solutions of the ground-state hydrogenic wavefunctions in megatesla fields developed by astrophysicists.

A phase diagram of the anisotropic Heisenberg model with competition between the nearest and the next-to-nearest neighbour exchange interactions is investigated by using the bosonisation scheme. An instability leading to a spontaneous dimerisation is found. The new 'superfluid' phases with both nonzero, though nonnominal, Z-magnetisation and broken symmetry in the XY-plane are detected in the anisotropic case.

The Ho³⁺ ion (4f¹⁰, ⁵I₈) in holmium trifluoride has a singlet ground state and an excited state, also a singlet, at 6.59 cm^-1. In HoF₃ the nuclear moments (¹⁶⁵Ho, I=⁷/₂) are strongly enhanced through the hyperfine interaction, and previous results concluded that at T_N=0.53 K the crystal undergoes a transition to an ordered antiferromagnetic state with both nuclear and electronic moments parallel and antiparallel to the orthorhombic a axis. Neutron diffraction studies at temperatures well below T_N have revealed that the ordered magnetic state is at variance with earlier conclusions and is more complex. A refinement of the neutron diffraction data shows that the ordered state may be described as Pnm'a' (F_xC_z) with Ho³⁺ moments of 5.7(2) mu _B at 66 degrees to (001). The temperature dependence of the intensities of the (100), (140) and (420) reflections have been determined for 0.07<T<0.6 K and are compared with the predictions of molecular field theory.

An S=³/₂ Heisenberg antiferromagnet on a triangular lattice CuCrO₂, in which stacking of the triangular lattice of magnetic Cr atoms forms a layered rhombohedral antiferromagnet, is studied by neutron powder diffraction. In the paramagnetic phase the powder diffraction pattern shows asymmetry, which proves a two-dimensional character. In the ordered phase, magnetic Bragg scattering has large width, indicating that the scattering is distributed on a line (¹/₃¹/₃ zeta ) with peaks where zeta takes integer values. Although the magnetic long-range order is established in the c plane, correlation in the c direction is finite or the modulation vector is distributed on the line. Intensity of magnetic reflections is consistent with the 120 degrees structure in the a-c plane with moment (3.1+or-0.2) mu _B.

First-order phase transitions are induced by an axial magnetic field in the disordered antiferromagnet Fe_0.5Ni_0.5Cl₂ (T_N=40.9 K) below the tricritical point (H=2.97 T, T=29.3 K). The thermodynamic first-order transition line separates the antiferromagnetic (AF) from the paramagnetic (PM) phase down to a critical end point (H=3.40 T, T=19.3 K), below which AF-to-spin-flop (SF) transitions are found. Second-order SF-to-PM transitions occur at slightly higher fields. This is the first experimental evidence of the unusual multi-critical behaviour of Heisenberg antiferromagnets with intermediate uniaxial anisotropy as predicted by Vilfan and Galam (1986). Magnetic phase diagrams calculated using realistic exchange and anisotropy model parameters in the virtual-crystal approximation are in qualitative agreement with the experimental data obtained from Faraday rotation measurements.

High pressures and temperatures were used to synthesise the previously unknown compound EuNi. Magnetic measurements, and Mossbauer effect measurements for ¹⁵¹Eu and ¹¹⁹Sn impurity atoms are reported. Ferromagnetic ordering below 139 K and a mean valence value of 2.3+or-0.1 for Eu ions were found.

The shift with concentration variation of the optical absorption F_A band in KCl-KBr mixed crystals, with Li impurities is studied by means of the method of Bartram et al. (1968). The distortion around the F_A centre is calculated considering the change in the repulsive and polarisation energies due to the impurity. The energy due to this distortion is obtained by a perturbative method. This correction gives good agreement with the experiment values for one of the two absorption bands of the F_A centre, but not for the other. Calculations using several types of wavefunction are in approximate agreement with the experimental observations of Asami and Ishiguro (1986).

The authors report on the absorption and luminescence spectra of a laser nonlinear self-frequency-doubling Nd_xY_1-xAl₃(BO₃)₄(NYAB) crystal at room temperature. Following the Judd-Ofelt theory, they have obtained the oscillator strength for seven bands and three intensity parameters Omega ₂, Omega ₄, Omega ₆. The radiative lifetime tau of the metastable state ⁴F_3/2 and a series of laser coefficients of ⁴F_3/2 to ⁴I_j/2 transitions are also calculated. The primary results on the Cr-Nd sensitisation effect on Nd:Cr:YAB crystals are also reported.

The authors study the effects of Brownian coagulation driven by the thermal motion of droplets on the competitive droplet growth in a quenched binary fluid. To consider such an effect, they extend the classical theory of the Ostwald ripening by using a systematic method that was originally proposed by Lifshitz and Slyozov (1961), for the case of the encounter mechanism. They discuss the effect of Brownian coagulation on the asymptotic form of the droplet size distribution function and the growth law of the average droplet size.

The authors report the observation of a fluctuation-induced first order phase transition between two states of adsorption at a wall-liquid interface. WDA density functional theory modified to enable input of the measured bulk equation of state is used to confirm the critical nature of the mean-field transition. MD computer simulation results show a first-order phase transition associated with dramatic collective fluctuations in the amount of adsorbed fluid. The qualitative appearance of the large-scale fluctuation events is described.