Table of contents

A theory of Raman scattering by superlattice LO phonons is outlined. The formalism is based on the dispersive continuum model obtained by the extension of the Born and Huang theory (1968) to include dispersion. This model allows a straightforward derivation of the response functions entering the light scattering cross section. Although the theory can be formulated so as to apply to any scattering geometry, only the familiar back-scattering configuration is considered here. The predictions of the theory are examined for this case and compared with the corresponding experimental data from a GaAs/Ga_1-xAl_xAs superlattice.

The author argues that in fact there is no difference between Holstein's result (1986) and the author's previous result (1985) in the case of the double-well potential. He emphasises that it is the exact standard WKB result.

The authors measure the segregation of platinum at the (111) surface of the Pt_0.1Ni_0.9 single-crystal alloy by means of low-energy electron diffraction (LEED) and find that platinum is enriched to 30+or-4% in the first layer and depleted to 5+or-3% in the second; the corresponding relaxations of the interlayer spacings are 0+or-0.2% and -0.8+or-0.9%. They let the averaged t-matrix approximation mimic LEED and substitutional disorder, and use metric distances for fitting a structural model to the data.

The theory of elastic constants of crystalline solids, which relates the elastic constants to the derivatives of the interatomic potential is generalized for a noncrystalline structure. The nearest-neighbour (NN) distance is assumed to have a distribution about its mean. Analytic expressions for the elastic constants C₁₁, C₁₂ and C₄₄ are obtained when only NN interactions are taken into account. The current theory predicts the isotropy (i.e. C₁₁-C₁₂=2C₄) of a glassy system. The theory is applied to a two-component colloidal glass where the average interaction potential is obtained by a mean-field approach. For a number of simple probability distributions, closed-form expressions are derived for bulk and shear moduli. The calculated shear moduli agree well with the reported values for colloidal glasses.

The authors calculate the high-pressure behaviour of magnesium and cobalt oxides using the best available potentials within the shell model. They obtain the Hugoniot equation of state to a pressure of 250 GPa and show good agreement with the available experiments. Static volume-temperature relations are also obtained and the transition pressures for the B1-B2 (rocksalt-caesium-chloride) structural transition predicted. The results suggest that this transition is unlikely to be important in the lower mantle.

The authors have recorded inelastic neutron scattering spectra of GaAs due to LO and LA modes at the high symmetry points X and L of the Brillouin zone. Accurate measurement of the relative intensity of the two modes results in the eigenvectors e(Ga mod LO)=e(As mod LA)=1, e(Ga mod La)=e(As mod LO)=0 for X and e(Ga mod LO)=e(As mod LA)=0.58+or-0.07, e(Ga mod LA)=-e(As mod LO)=0.81+or-0.05 for L. These numbers are compared with a variety of lattice-dynamical model predictions and quantum mechanical results.

Thermodynamic functions of the binary chalcogenide systems, S-Se, S-Te and Se-Te, have been analyzed to provide evidence for structural phase transitions in the liquid state of tellurium-rich alloys. A cusp-like minimum in the enthalpy of mixing as a function of concentration observed for the last two systems is in consequence of the steep increase in the enthalpies of tellurium-rich alloys in the narrow temperature range. Apparent complications in the thermodynamic functions are indicative of structural phase transitions rather than some local order in the liquid state. The inhomogeneous structure model of Tsuchiya and Seymour (1982, 1985) can consistently explain these characteristic features of thermodynamic functions together with other thermodynamic properties.

The adequacy of the Callaway integral (1959, 1974) and the Leibfried-Schlomann formula (1979) for describing the pressure dependence of the thermal conductivity of insulators is investigated. It is concluded that, overall, the Leibfried-Schlomann formula and the Callaway integral offer roughly the same agreement between calculated and measured results.

Adsorption of sodium on the W(100) surface is studied at room temperature by angular resolved ultraviolet photoemission spectroscopy (ARUPS) and LEED. A reconstruction of the W(100) surface induced by sodium adsorption is observed at least at low coverage. The high-lying tungsten surface states situated at 0.4 eV below the Fermi level at normal emission ( Gamma ) suffer a continuous shift towards higher binding energies when sodium coverage increases. With a monolayer of sodium, the shift attains 2 instead of 1 eV for caesium. The low-lying surface state located at 4.2 eV below the Fermi level at Gamma shifts towards higher binding energies in contrast with caesium adsorption. These effects reveal a stronger sodium atom-substrate interaction than that which occurs in caesium. The description of the Cs/W(100) system may be extended to Na/W(100): the d(z²)-like surface states of W(100) hybridise with the sodium 3s valence states to form a polarised covalent bond.

The general features of the repulsive potential introduced by oxygen atoms successively agglomerating around 450 degrees C thermal donors is determined by fitting the ladders of energy levels for both charge states observed by infrared spectroscopy and accounting for electron-screening effects in the neutral charge state. Recent evidence suggesting that the thermal-donor ground-state wavefunction is constructed from a pair of (100) valleys has been used to determine qualitative features of the perturbing potential. Directional agglomeration is seen to reproduce the experimental energy shifts, a result in accord with the view that thermal donors are nucleation sites for the subsequent formation of coesite.

The electronic structure of ferroelectric LiNbO₃ at room temperature has been investigated using the X alpha molecular orbital method applied to a NbO₆ cluster of C_3v symmetry. The total and partial densities of states associated to the O 2p valence band built up from the X alpha eigenvalues are compared with XPS and UPS data as well as the calculated binding energies of the inner states. The band gaps E/sub /// and E_{perpendicular to} relevant to light polarised along or parallel to the spontaneous polarisation P_s, are obtained from the allowed O 2p-Nb 4d transitions. Moreover, the imaginary part of the dielectric constants epsilon ₂// and epsilon _{2 perpendicular to} are estimated from these transitions and used to explain the strong anisotropy observed in the reflectivity data. A population analysis, performed from the electronic distributions in the cluster, leads to the cluster ionic charges. These latter are compared with the effective charges obtained from IR measurements and from point-charge model calculations. They set up the covalency of Nb-O bonds and allow the authors to estimate P_s.

The Feynman path integral formalism is used to calculate the energy and the effective mass of one-dimensional optical and acoustic polarons. It is shown that the optical polaron yields a continuous change as the coupling goes from the weak to the strong limit. On the other hand an abrupt change of the acoustic polaron state at alpha _ac=0.5 is observed.

Light scattering spectra in the energy-transfer range 4< nu <500 cm^-1 have been measured for molten MgBr₂, CaBr₂, SrBr₂ and BaBr₂. The spectra are similar to those found for the molten alkaline-earth chlorides. First-order Raman scattering gives a central peak, whose width is determined by dissipative effects such as electrical conduction, and a noncentral peak or shoulder whose frequency is of the order expected for TO modes in the corresponding crystals. In the case of MgBr₂ the peak is well resolved and in CaBr₂ it is almost resolved. Second-order Raman scattering gives an exponential tail at high frequencies. The spectra have been fitted with the semi-empirical model of Bunten and co-workers (1984) in order to derive the relaxation times associated with these contributions to the scattering. In Mg₂₊ and Ca²⁺ salts, where the noncentral contribution is completely polarised, its frequency scales linearly with reduced mass.

The relationships among the structure parameter of the Cr³⁺-vacancy centre, the EPR parameter D and the splitting Delta E(²E) of the ²E state in MgO:Cr³⁺ are established according to third-order perturbation formulae and the SCF d orbit of a Cr³⁺ ion. By using these, the structure parameter of the Cr³⁺-V_c centre is determined quantitatively from the EPR and optical spectra data, and the results show that the displacement Delta R of the anion O^2- along the (001) axis is Delta R=0.056+or-0.004 AA. This method would be effective in determining the structure parameter of other Mⁿ⁺-V_c centres.

For the origin of the high-temperature superconductivity and its correlation with the structural phase transition in A15 compounds, an electron mechanism based on the Geilikman model (1965) is considered. The electronic band structure near the Fermi level epsilon _F in these compounds consists of the very narrow, twofold degenerate Gamma ₁₂ subbands referred to as the 'd bands' and other wide bands referred to as the 's bands'. The Coulomb interaction between the s and d electrons causes an indirect attractive interaction via the particle-hole pair excitations in the d-electron system. The s-electron system with the attractive interactions due to both the electron and phonon mechanisms is assumed to be responsible for the main part of the superconductivity, while the d-electron system is for the structural phase transition. The superconductivity of the s-electron system forces d electrons to form Cooper pairs. The two phase transitions interfere with each other, since the structural phase transition causes a decrease in the density of states near epsilon _F of the d bands, which is unfavourable for superconductivity. The temperature dependences of the crystal distortion and of the elastic constants are calculated. This model can explain successfully the experimental results on V₃Si and Nb₃Sn.

An anisotropic exchange Heisenberg chain in the classical continuum approximation is studied. Single-soliton solutions and their corresponding energies for the case of small anisotropy are derived analytically. Semiclassical quantisation of the soliton energy is performed and the result compared with the energy of the bound magnon states.

Magnetic inhomogeneity in a magnetic medium can be demonstrated with the use of Landau-Ginzburg theory. One-dimensional solutions for the spatial variation of the magnetisation M(X) have been found in terms of the Landau-Ginzberg coefficients under various conditions. The solutions show two types of variations: (i) oscillatory between positive and negative values, and between two positive values of M; (ii) discontinuous in M(X). The discontinuities in the latter lead to M(X)=constant as the only solution. An analytical expression has been obtained that helps to predict values at peaks or troughs in M(X).

Two effective-field treatments are presented for the diluted transverse Ising model. The methods include the effects of correlation. The critical temperature, the longitudinal and transverse magnetisations, the initial transverse susceptibility and the two-site spin correlation function are calculated for a square lattice. It was found that, in particular, the transverse magnetisation and the initial transverse susceptibility exhibit behaviour which is interesting in the diluted case.

Monte Carlo simulations are used to investigate a spin system with randomness defined by a Hamiltonian. The phase diagram is investigated for a range of concentrations of the single-ion anisotropy term for the set of parameters (J/sub ///, J_{perpendicular to} , D)=(¹/₃, ¹/₆, 1). The phase boundaries are determined from the positions of the maxima in the transverse and longitudinal susceptibilities. The ordered state is a longitudinal ferromagnet for small values of p and a transverse ferromagnet for large values of p. For intermediate p-values the ground state is a mixed phase characterised by the presence of both transverse and longitudinal order parameters. It is shown that the two order parameters are coupled in the mixed phase, in agreement with the experimental findings for the material Fe_pCo_1-pCl₂. Isotherms of the difference between the transverse and longitudinal order parameters are smooth functions of p in the mixed phase. The model seems to exhibit a tetra-critical point close to (p,T)=(0.75, 0.30 J/sub ////k_B).

The electronic properties of the ferrous ion in (Fe(H₂O)₆)(NH₄)₂(SO₄)₂ are investigated using far-infrared (FIR) Fourier transform spectra and FIR EPR measurements at low temperatures in the energy range from 8 to 100 cm^-1. Several absorption lines can be identified to originate from electronic excitations of the ferrous ion.

The interaction of the non-Kramers Tb³⁺ ions at tetragonal sites in CaF₂ with the second and third fluorine shells has been studied by ENDOR. Displacements of the Tb³⁺ and F^- ions near a paramagnetic centre and interstitial F^- have been evaluated.

The resonant Raman spectra due to the symmetric Cl-Pt-Cl stretching mode in (Pt(en)₂)(Pt(en)₂Cl₂)(ClO₄)₄ and (Pt(en)₂)(Pt(en)₂Br₂)(ClO₄)₄ single crystals have been observed to be resolved into fine structures with three main components at 2K. It is found that both the Raman frequencies and the relative intensities of these components depend considerably on the incident photon energy and on the order of the overtone. The origin of the fine structures is discussed from the detailed measurements of this dependence.

Polarised luminescence spectra of Cr³⁺-doped beta -alumina crystals were taken at temperatures between 78 and 295K. From luminescence excitation spectra the energy level scheme of the Cr³⁺ ion is obtained. This turns out to be very similar to that of ruby, with the Cr³⁺ ion substitutional for the Al³⁺ ion in the octahedral sites of the spinel block structure in beta -alumina. Time-resolved spectroscopy allows for the separation of the contributions from non-equivalent centres, characterised by lifetimes ranging from 0.7 ms to 50 ms. Three different sets of centres were identified, corresponding to Cr³⁺ ions in centrosymmetric and noncentrosymmetric sites. The partially disordered structure of the host crystal is reflected in the shape of the emission lines, whose characteristics are intermediate between those typical of ordered compounds and amorphous ones.

Experimental normal photoemission spectra from Cu(100) were taken with linearly polarised synchrotron radiation at photon energies between 13 and 48 eV and at two different angles of light incidence. They are compared with intensity profiles calculated by nonrelativistic 'one-step-model' theory including a classical macroscopic radiation field inside the crystal. All direct transition features of the spectra are reproduced quite well, while agreement in detail is not yet perfect. Several possibilities for improving the agreement are discussed.