Table of contents

Crystal optical and X-ray diffraction studies show that the high-temperature phase I of the perovskite-type layer-compounds (C_nH_2n+1NH₃)₂CuCl₄ with n=1, 2, 3 is Jahn-Teller distorted orthorhombic with space group Bbcm (D_2h¹⁸). As a consequence, the order of the C_nH_2n+1NH₃ group leads to two configurations, realised in a monoclinic phase II with space group P2₁/c (C_2h⁵) and an orthorhombic phase III with Pbca (D_2h¹⁵). The transition sequences found above 300 K depend on n: I from T_cl=348 K to II for n=1; I from T_cl=364 K from II from T_c2^h=356 K, T_c2^c=338 K to III for n=2; and I from T_cl=434 K to III for n=3. The domain phenomena at T_cl and T_c2 point to a subtle balance between the tilt and order configurations of phases II and III. Birefringence and lattice constraints indicate a very loose inter-layer contact in phase I.

The authors describe a high-resolution power diffraction experiment performed on PLZT (8.7/65/35) using the synchrotron radiation source at Daresbury. The experiment confirms the existence of previously reported structural changes at 290 and 360 degrees C; the results discussed here are more detailed and allow some quantitative analysis. In particular, the synchrotron source allowed the use of a geometry that enabled very small changes in Bragg angle to be detected; these changes were not seen when using a conventional diffractometer.

Inelastic neutron scattering and isotope separation factor measurements are reported for the potential hydrogen storage system Ti_0.4Mn_0.6H_n. Inelastic spectra were measured for six hydrogen concentrations ranging from 0.001 to 1.0 hydrogens/metal atom, the first five in the alpha and alpha + beta phases and the last in the beta phase. The spectra consisted of two groups of peaks, the first from 70 meV to 110 meV and the second from 110 to 180 meV. The intensity of the former was independent of the hydrogen content and could be associated with a precursor region in the isotherm while that of the latter rose linearly with hydrogen content. Peak shape analysis of the high energy peaks show them to be consistent with occupation of first Ti₃Mn and then Ti₂Mn₂ sites and with changes between the alpha - and beta -phases. The identification of the precursor sites is not easy if based entirely on the IINS spectra as both-OH and near surface interstitial sites could be involved. However, the measurement of the H/T isotope separation factor unambiguously points to the existence of TiOH because only this can explain the dramatic fall of alpha near n=0.

The authors use local density functional theory to calculate the structure and vibrational properties of small molecules containing phosphorus and selenium atoms in varying concentrations in an attempt to shed light on the composition of zero-dimensional P-Se glasses. They examine the structure and modes of P₄Se_n n=0, 1, 2, 3, 4, 5, 6, 10, P₈Se₈ and P₁₀Se₆. Their results enable them to exclude some molecules from being present. For the Se-rich glasses they find that the singly bonded molecules P₄Se_n, n=3, 4, 5, 6 are stable and are likely to be present. For the P-rich glasses P₄Se₃, P₄Se₂ and possibly P₁₀Se₆ are present.

Short-range shell-model potentials between substitutional Ag⁺ ions and host-lattice halide ions in RbCl, KCl, NaCl, and NaF are derived using the ICECAP methodology. In this method the impurity and its nearest neighbours are treated as an embedded quantum cluster in unrestricted Hartree-Fock self-consistent field approximation. The embedding lattice is described by the shell model. The consistency of the derivation is tested in various ways. The potentials are then applied to calculate activation energies for diffusion by vacancy and interstitial mechanisms for Ag⁺. Previously derived potentials for Cu⁺ are used to calculate interstitial diffusion activation energies in KCl and NaCl. The potentials are tested for transferability from one host crystal to another. Deficiencies of the method are discussed and improvements are identified.

Renormalisation-group recursion relations are used to construct equations of state for the critical region of the three-state Potts model with symmetry-breaking perturbations, to first-order in epsilon =4-d. Explicit scaling forms are obtained for the free energy and the longitudinal susceptibility above and below the critical point, as well as for the order-parameter discontinuity on the first-order phase boundary. Application to the trigonal-to-pseudotetragonal phase transitions in SrTiO₃ under stress yields reasonable agreement with the experimental phase boundary.

Exposure of a caesiated Si(100)2*1 surface to O₂ and subsequent heating to about 800 degrees C caused the formation of SiO₄ and SiO₂. For a saturation Cs coverage and the same O₂ exposure, the growth of oxides is more pronounced on a stepped than on a non-stepped Si(100)2*1 surface. This is due to the higher coverage of Cs on the stepped surface. When the above oxidation cycle is repeated, the SiO₂ predominates. After six oxidation cycles the average thickness of SiO₂ is about 11 AA and 16 AA on Si(100)2*1 surfaces without and with steps, respectively.

The relevant quantity in the comparison of the absolute positron energy levels in different materials is the sum of the internal electron and positron chemical potentials, i.e. the sum of the Fermi level and the bottom of the lowest positron band relative to a common, well-defined reference energy. This sum is defined as the positron affinity. The positron affinity reflects the preference of the positron for different components in heterostructures made of different materials and the preference between the host matrix and precipitates in alloys. Moreover, the affinity is closely related to the positron work function and positronium formation potential which are important parameters in the slow-positron-beam experiments. The authors have determined the positron affinity for the alkaline and alkaline-earth metals, 3d-, 4d-, and 5d-transition metal series, and for some metals on the right in the Periodic Table. The diamond structure semiconductors are also considered. The determination is based on the self-consistent electron structure calculations and the subsequent calculation of the positron band structure within the local-density approximation. The trends are studied and interpreted along the different columns and rows of the Periodic Table. The results are also compared with available experiments.

DC and AC electrical conductivity measurements in single crystals of diammonium hydrogen phosphate along the c axis show anomalous variations at 174, 246 and 416 K. The low-frequency dielectric constant also exhibits peaks exactly at these temperatures with a thermal hysteresis of 13 degrees C for the peak at 416 K. These specific features of the electrical properties are in agreement with earlier NMR second-moment data and can be identified with three distinct phase transitions that occur in the crystal. The electrical conductivity values have been found to increase linearly with impurity concentration in specimens doped with a specific amount of SO₄^2- ions. The mechanisms of the phase transition and of the electrical conduction process are discussed in detail.

The magnetic ordering of a single crystal of the cubic polymorph of FeGe has been studied by small-angle neutron scattering. The compound orders magnetically at T_N=278.7 K into a long-range spiral (period approximately 683-700 AA) propagating along equivalent <100> directions at high temperatures and along equivalent <111> directions at low temperatures. The length of the spiral wavevector is nearly independent of temperature. The transition at T_N is first order with very little hysteresis. The transition at which the direction of the spiral turns is rather sluggish. It takes place in a temperature interval of approximately 40 K and shows pronounced temperature hysteresis (T_{2 down arrow} =211 K, T_{2 up arrow} =245 K). Applied magnetic fields of 20-40 mT, depending on the temperature and the field direction, cause the spiral axis to turn into the direction of the applied field. As the field is further increased, the amplitude of the antiferromagnetic spiral decreases and the ferromagnetic component increases until at fields above approximately 200-300 mT cubic FeGe becomes magnetically saturated. The magnetic ordering in cubic FeGe is a Dzyaloshinskii spiral similar to the structure observed in the isostructural compound MnSi. However, in MnSi the spiral propagates along equivalent <111> directions at all temperatures below T_N=29.5 K.

The field dependence of the magnetic excitations in the re-entrant spin glass (RSG) Fe₂₈Cr₇₂ is investigated over a wide range of temperatures. The zero-field spin dynamics are anomalous not only in the spin-glass phase but also in the paramagnetic phase. In an applied field (H=8.8 kOe) spin waves are induced at low temperatures with an energy gap g mu _BH approximately=0.1 meV. This observation is in marked contrast with results from Fe_1-xAl_x, where no gap is induced. The results suggest that the ground state of RSG is not unique.

It is known that quantum fluctuations significantly modify the energy required to produce a mobile domain wall, or soliton, on a chain of exchange-coupled magnetic atoms subject to an anisotropy field or magnetic field. The authors show here that spin-wave theory applied about the classical soliton to the sine-Gordon equation can be adapted to calculate these zero-point energy effects by matrix diagonalisation. The results for solitons on ferromagnetically coupled chains are qualitatively similar to those obtained by other techniques. In antiferromagnets, however, the quantum fluctuations appear to produce, under certain conditions, a negative soliton creation energy, i.e. a soliton-containing ground state. This situation is a result of the non-linear response of the zero-point energy to small changes in the effective anisotropy of the one-dimensional antiferromagnet.

Samples taken from the growth band region of a KDP:Cr crystal have been studied by EPR spectroscopy at room temperature. The observed lines are consistent with a centre in which Cr²⁺ substitutes for K⁺ with two proton vacancies in nearest positions. From the powder spectrum and the angular variation of the resonance fields the defect principal axes and the spin-Hamiltonian parameters have been calculated. The experimental spectra, the angular variation and the parameters obtained resemble those from samples taken from regions out of the growth bands. The differences between Cr³⁺ spectra in both types of sample are attributed to variations in the crystal-field strength produced by changes in the lattice parameters.

⁵⁵Mn and ¹²¹Sb spin-echo NMR measurements have been carried out at liquid-He temperature on MnSb/Sb multi-layered films. The MnSb layers were shown to grow epitaxially with the c axis perpendicular to the film plane. The hyperfine tensors and the anisotropy fields were deduced from the variations in the resonance frequency and linewidth under a DC magnetic field along the c axis.

The theory of temperature dependence of photoemission spectra is reviewed in a systematic manner and an attempt to identify the essential components is made. Using synchrotron light the authors have explored the temperature dependence of the photoemission from the valence band of Cu(111) over a broad energy range (h nu =20 to 85 eV). Remarkable differences in the thermal responses were observed as the photon energy was varied. Particularly unusual was the observation of a direct transition peak which increased in intensity as the temperature was raised. The experimental data support the idea that, under certain circumstances, phonon-assisted transitions from k-space regions where the initial-state bands are flat, e.g. at band edges, may result in spectral peaks, which compensates for the reduced intensities of direct transitions as the temperature is raised. The authors have also performed full numerical calculations in a time-reversed LEED formalism, and the results of these are compared with their experimental data.

Experimentally determined coherent dynamic scattering functions for liquid bismuth and lead are used to derive the van Hove correlation functions G(r,t) and G_s(r,t). It is demonstrated that it is possible to obtain direct and physically adequate information on the atomic self-motion from a neutron inelastic scattering measurement on a completely coherent scatterer.

Expressions for zeroth-, second- and fourth-frequency sum rules of energy current density correlation function have been derived. These exact expressions involve static correlation functions of up to five particles. Because of the non-availability of any information about triplet and higher-particle correlation functions, the authors use a low-order decoupling approximation for these. The resulting expressions for the frequency sum rules, which now contain only a static pair correlation function and an inner-atomic potential, are suitable for applications. Numerical results for these frequency sum rules have been obtained for Lennard-Jones fluids over a wide range of densities and temperatures. The density and temperature dependence and the relative importance of higher-particle correlation contributions to the frequency sum rules have been discussed.

For pt.I see ibid., vol.1, p.6181 (1989). Expressions for the energy current density correlation function and the thermal conductivity have been derived using a continued-fraction representation of correlation function. A secant hyperbolic, a Gaussian and an exponential form of the memory function have been used. The parameters appearing in the memory function are determined from the zeroth-, second- and fourth-frequency sum rules of the energy current density correlation function. For the time dependence of the energy current density correlation function and the thermal conductivity, the results obtained for the Lennard-Jones fluids, have been compared with recent computer simulation data. Overall good agreement has been obtained over a wide range of densities and a wide range of temperatures.

The transverse current correlation function, C_T(k,t), is re-examined and some of the limitations of the familiar viscoelastic model, particularly for wavelengths less than a few interatomic spacings, emphasised. An alternative theory, which stresses the role of the velocity autocorrelation components in C_T(k,t), is proposed, with momentum transfer effects to the surrounding atomic shells of a given atom calculated separately. When applied to a liquid rubidium model, the results are in good agreement with computer simulation data. The theory is also used to investigate the generalised wavevector-dependent fluidity, eta ^-1(k), and gives new insight into the detailed structure which has been revealed by computer simulation data. A model for eta ^-1(k) is proposed, and the relationship of the self-diffusion coefficient of the liquid to the fluidity investigated.

The percolation cluster characteristics of 2D Lennard-Jones (LJ) fluid and solid phases have been determined by molecular dynamics computer simulation. The variation of the percolation threshold with temperature from the soft-core to hard-core limits is reported and interpreted in terms of changes in the short-range structure. The distribution of clusters is analysed using the cluster number distribution function, n_s. Based in the small systems considered, the authors find that at the percolation threshold, the measured percolation exponent tau departs from the universal 2D value of 2.05, ranging from 1.6+or-0.1 in the soft-core limit to 2.0+or-0.05 in the hard-core limit. However, the fractal dimension of the clusters, D_f, at the percolation threshold is indistinguishable from the 2D universal value of 1.9, having the range 1.9+or-0.1 for all state points considered. The same trends are observed in 3D LJ fluids. They examine the time dependence of the percolating clusters and show them to be ephemeral on a molecular timescale.

The authors present a detailed report of the observation of a strong, thermally induced optical non-linearity in a system comprised of a silver film and a nematic liquid crystal, in which the surface plasmon-polariton (SPP) mode may be excited. Various measurements of the reflected intensity from the experimental system as a function of the incident angle, including a novel 'AC-on-DC' probe technique, reveal that the observed sharp optical switching is due to warming of liquid crystal, close to the metal/liquid crystal interface, into the isotropic phase through loss of energy of SPPS in Joule heating of the silver film. The many factors contributing to the observed reflectances are investigated and lead to a detailed model of the non-linearity for which a computer model is constructed. Conclusions drawn from the results of this work have serious implications for the use of SPPS in non-linear optics.

It is suggested that a generalisation of a variational relation between the equilibrium free energy and the pair potential also holds for changes of the system along paths of non-equilibrium states. Combined with a parametrised model for the radial distribution function, this relation leads to an expression for the configurational entropy of an amorphous phase in terms of empirically determined radial distributions and a given pair potential. It is applied in combination with the 'structural diffusion' model to computer simulated amorphous aluminium, using recently determined model parameters from the normal liquid down to the amorphous solid at 273 K. The entropy difference between the liquid at melting point and the amorphous solid is estimated to be of the order of the entropy of melting, implying a negligible residual entropy. The temperature dependence of the structural diffusion coefficient, as determined from structural data, exhibits a glass transition of the amorphous system.

The residual resistivity rho ₀ as a function of composition has been measured on the pseudo-binary glassy alloy series (Fe_100-xCo_x)₉₀Zr₁₀ and (Fe_100-xCo_x)₈₀B₂₀, respectively, in the concentration range 0<or=x<or=100. Absolute values for rho ₀ are found to exceed those of corresponding crystalline disordered Fe-Co alloys by at least a factor of ten. The observed composition dependence of rho ₀ in the glassy alloys finds a straightforward explanation in terms of the two-current conduction model, adapted to the peculiarities of the amorphous structure, and substantiates the earlier finding of significant magnetic contributions to the total resistivity in such amorphous systems.

The behaviour of the antiferromagnetic nearest-neighbour spin-¹/₂ Heisenberg chain is studied in a space of interactions that include exchange anisotropy ( alpha ), bond alternation ( lambda ), and spin-lattice interaction ( gamma ). Using the coherent state method for spin, the authors analyse magnetically driven lattice instabilities and find that phase transitions occur under the condition of large bond alteration or strong spin-lattice coupling, which means that more staggered interactions make it easy for phase transitions to take place. The three-phase diagram and the tri-critical point are obtained.

The authors provide evidence for both voltage- and current-controlled negative differential resistance instabilities in the one-dimensional ballistic resistor driven into the hot-electron regime. Quantum reflections are used to explain the former effect, while extra electron subbands provide the current 'filaments' in the latter case.

The authors have investigated magneto-transport through a lateral quantum box defined by a patterned Schottky barrier on a GaAs-AlGaAs heterojunction. As the size of the box, and the reflecting barrier potential, are varied, edge states are successively reflected with a consequent change in the value of the plateaux of resistance. Superimposed on a plateau are modulations of resistance having both electrical and magnetic origin. Resonant transport, with an increase in transmission ratio to unity, (corresponding to oscillations in resistance at, or above, the plateaux values), occurs when an integer number of wavelengths fit into the box perimeter and are rapidly modulated by flux quantisation of the Aharonov-Bohm type. In addition, resonance can arise when an incident wave tunnels through the barriers via a circulating, trapped, edge current. This behaviour brings in an extra mode of current and causes the resistance to drop below the value of a plateau in contrast to the previously mentioned resonance arising from transport above the barrier.

Spin assignments of fractional QHE states in the N=0 Landau level are determined from finite-size calculations of the Coulomb energy. There is a spin-unpolarised ground state at ²/₃ filling, with partial polarisation at ³/₅ filling and full polarisation at 1/q filling with q odd. This explained in terms of competition between Coulomb repulsion, exchange and interaction of unlike spins. The experimental consequences are briefly discussed.

The variation of the superconducting critical temperature, T_c, with pressure in ⁶⁷Zn has been studied making use of a recent experimental study on the pressure-dependent Lamb-Mossbauer factor in the range 0-5.8 GPa at 4.2 K. It is found that the critical temperature decreases with pressure. At 5.8 GPa the value of T_c ( approximately=0.123 K) is about 1/6.8 of that at ambient pressure. The variation of T_c with pressure can easily be measured.

The spin system with two-axis anisotropy is considered. The temperature dependence of the probability of transition from the metastable state due to quantum and thermal fluctuations is determined. The method of calculation is based on exact correspondence between the spin system and a particle moving in a potential field.

Positron diffusion in an electric field has been experimentally investigated by implanting positrons in the depleted zone of an Au/Si Schottky diode. The results show that fields with an intensity of about 10⁴ V cm^-1 have a strong effect on positron diffusion. Before annihilation, thermalised positrons can drift distances of about 2 mu m.

The momentum dependence of an effective enhancement factor is investigated both for itinerant and core electrons. The local density approach is applied to describe electron-positron interaction. The possibility of over-enhancement of HMC is predicted for the first time. The results obtained for core electrons are consistent with experimental data. It is shown that for some momenta electron-positron attraction can give not only enhancement, but also damping of the probability annihilation.