Table of contents

Neutron diffraction experiments were carried out on an inorganic quasi-one-dimensional magnet with spin 1/2 , CuGeO₃, and showed satellite peaks arising due to the dimerization of Cu atoms. Displaying a spin-Peierls transition, Cu ions with spin 1/2 were found to reach a singlet ground state having alternating displacements between Cu atoms. Taking the shift of a Cu atom as an order parameter, the critical exponent beta becomes 0.5 and the shift resulting from dimerization along the c axis is about 1% of the lattice constant c at 2.7 K.

The equation of motion of the auto-correlation function has been solved analytically using a hyperbolic secant form of the memory function. The analytical result obtained for long-time expansion together with short-time expansion provides a good description over the whole time domain as judged by a comparison with the numerical solution of the Mori equation of motion. We also find that the time evolution of the auto-correlation function is determined by a single parameter tau which is related to frequency sum rules up to fourth order. The autocorrelation function has been found to show simple decaying or oscillatory behaviour depending on whether the parameter tau is greater than or less than some critical value. Similarities as well as differences in the time evolution of the auto-correlation have been discussed for exponential, hyperbolic secant and Gaussian approaches of the memory function.

The influence of small additions of iron on the electrical resistivity of liquid gold has been investigated for an iron content x_Fe<or=0.24. The drastic increase in the resistivity due to small amounts of iron as well as the negative temperature coefficient apparent in the more concentrated alloys are not understandable on the basis of the Faber-Ziman formalism. The localization of large magnetic moments on the impurity atoms is assumed to be responsible for this exceptional behaviour. A tentative description is presented in terms of spin-disorder scattering including some particular aspects of the Kondo effect.

The structure of the stable decagonal quasicrystal Al_70.5Mn_16.5Pd₁₃ was determined by single-crystal X-ray methods. Higher-dimensional structure-solution techniques were employed, such as five-dimensional (5D) Patterson and Fourier syntheses as well as 5D least-squares structure refinements. A 3D physical-space maximum-entropy method (MEM) was used in the course of structure-factor phasing and for the calculation of a high-quality electron-density distribution function. Geometrically, the structure may be described as a periodic stacking of two different quasiperiodic layers A (puckered +or-0.3 AA) and B (planar) with sequence ABAaba (a and b denote the layers A and B, respectively, rotated by 2 pi /10), consistent with the centrosymmetric 5D superspace group P10₅/mmc. Basic structure-building elements, however, are columnar clusters (OE approximately 20 AA) with cross-sections corresponding to parts of a decorated Penrose tiling. The clusters are packed in a way that results in a disordered Robinson-triangle tiling. The title compound is isotypic to metastable decagonal Al₇₈Mn₂₂ and locally shows a strong resemblance to icosahedral Al_68.7Mn_9.6Pd_21.7, orthorhombic Al₃Mn and hexagonal mu -Al_4.12Mn.

We performed temperature-dependent extended X-ray absorption fine structure (EXAFS) measurements on a series of nanocrystalline zirconia powders with different crystallite sizes. The results show that there is a tetragonal-to-cubic phase transition with reduction of crystallite size. The temperature dependence of the EXAFS Debye-Waller factor is also discussed.

Neutron Compton scattering (NCS) measurements have been used to determine the kinetic energies of atoms in samples of amorphous hydrogenated carbon (a-C:H), graphite and diamond at momentum transfers between 40 and 300 AA^-1. We find that the kinetic energy of individual carbon atoms is the same within statistical error in a-C:H and graphite but significantly higher in diamond. The kinetic energy of the hydrogen in a-C:H is lower than expected from previous spectroscopic measurements and we infer that the sample contained molecular hydrogen. Observed deviations from the impulse approximation are consistent with theoretical calculations. We discuss future prospects for NCS measurements on non-crystalline materials.

Using the spectral moments method, we have studied the inelastic neutron scattering by phonons in an Al-Mn quasi-crystal. The structural model is derived from the work of Duneau and Oguey. For the dynamical model we assume an effective potential with force constants depending upon equilibrium distances. Calculations are performed with the first seven rational approximants of the quasi-periodic structure. Pseudo-acoustic dispersion curves associated with intense Bragg peaks are reported. Broadening of inelastic neutron scattering peaks with increasing phonon frequency is found in high-order approximants. We compare our results with experimental spectra obtained on an Al-Mn-Pd quasi-crystal where such a behaviour has been observed. The relation of the peak broadening to the critical character of the modes is discussed.

Mossbauer spectra of Fe-8.0 at.%C, Fe-10.9 at.%Al-8.2 at.%C, Fe-1.4 at.%Mn-7.9 at.%C, Fe-2.4 at.%Mn-7.8 at%C and Fe-14.6 at.%Ni-6.2 at.%C FCC gamma -irons have been measured in order to investigate the distribution and the local interactions of C in the lattice. The local interaction energies in nearest and second-nearest C-C and M-C (M: Al, Mn, Ni) atomic pairs have been determined by a Monte Carlo method simulating the distribution obtained. The interaction energies in nearest and second-nearest C-C pairs are strongly repulsive, in contrast to the weak interaction in second-nearest N-N pairs. The results are compared with the interaction energies derived from activity data. The interaction in nearest Al-C pairs is repulsive but that in second-nearest pairs is strongly attractive, which can lead to the formation of the perovskite Fe₃AlC-type ordering. The interaction between nearest Mn and C atoms is strongly attractive, while between Ni and C the interaction is very weak.

The absorption of longitudinal and transverse ultrasound in plastically deformed single crystals of tungsten was measured as a function of temperature. The maximum in the temperature dependence of the electron absorption was found to be more pronounced in deformed specimens. The electron absorption of longitudinal waves decreases in a magnetic field as the degree of plastic deformation is raised. The results are interpreted, allowing for intersheet scattering of electrons in hot spots of the Fermi surface.

The influence of hydrostatic pressure on the temperature behaviour of acoustic properties in the vicinity of phase transitions of CsD₂PO₄ and RbD₂PO₄ crystals is studied. The pressure-temperature phase diagrams are obtained. The results are considered within the framework of phenomenological theory.

It is well known that the value of room-temperature conductivity sigma _RT of boron-doped silicon films is one order lower than that of phosphorus-doped silicon films, when they are deposited in an identical plasma-enhanced chemical vapour deposition system. We use surface acoustic wave and secondary-ion mass spectrometry techniques to measure the concentration of total and electrically active boron atoms. It is shown that only 0.7% of the total amount of incorporated boron is electrically active. This is evidence that hydrogen atoms can passivate substitutional B-Si bonds by forming the neutral B-H-Si complex. By irradiating the boron-doped samples with a low-energy electron beam, the neutral B-H-Si complex converts into electrically active B-Si bonds and the conductivity can be increased by about one order of magnitude, up to the same level as that of phosphorus-doped samples.

A new bond auxiliary particle method for the Hubbard, t-J and Heisenberg Hamiltonians is presented. First, the one-dimensional (1D) Heisenberg Hamiltonian is studied as a test case. The exact ground state is approached rapidly via perturbation theory. At the level of second-order perturbation theory, the ground state energy is accurate within a few per cent, while in the 15th order it is accurate to five figures. Without any renormalization, the spin velocity is given with 10% accuracy, and the spectrum is correctly gapless. The degeneracy is also correct, giving a Wilson ratio R=2. The method is used to calculate analytically the full Green's function for a single hole in a half-filled (one electron per site) t-J model.

We study the dynamics of the spin-boson model using a simple mean-field-type approximation which treats phonons classically. This simple approximation allows us to cover the whole range of system parameters and to determine boundaries between regimes with different behaviours. The validity of the model is tested by its ability to recover most of the previous results. The results are also compared with the results of some more sophisticated variational approaches.

We present the results of tight-binding molecular dynamics calculations for studying the equilibrium structures and the bonding properties of Si_n clusters for n up to 18. To prevent unphysically large charge transfer between different atoms in clusters, we employ the atomic charge neutrality constraint that each atom has approximately four valence electrons. With a limited number of parameters in the tight-binding scheme, the structures of minimum energy are well reproduced, as compared with results from previous ab initio quantum mechanical calculations. We find the abundant cluster sizes n=4, 7 and 10, which are in good agreement with other theoretical and experimental results. For n>or=7, surface-like compact structures with a pentagon or a hexagon base are found to be energetically favourable, resulting in the metallic nature of the cluster bonding, while a core-based structure appears first for Si₁₅.

Calculations are made of the energies of the dominant transition (1s to 2p₊₁-like) of a hydrogenic donor impurity in a multi-quantum-well (MQW) system subject to a magnetic field applied at an angle theta to the direction of growth of the MQW layers. The model used involves constructing suitable basis states, and then performing a matrix diagonalization procedure. The results are compared to the available experimental data, including the extreme cases of fields perpendicular ( theta =0 degrees ) and parallel ( theta =90 degrees ) to the layers. The results are also found to compare well with the results of existing variational calculations.

We present a novel approach to the calculation of shallow impurity states in single- and multi-quantum-well systems. The Schrodinger equation is directly integrated numerically using an iterative technique. Thus, by avoiding a variational ansatz for the wave function and including a proper treatment of the coupling between the wells, the whole range from low to high magnetic fields as well as from the two-dimensional to the three-dimensional limit is treated on the same level. For the case of a single quantum well we discuss the dependence of the binding energies on the magnetic field as well as on the position of the impurity. For the case of a multi-quantum-well system we analyse the influence of the barrier width on the energies and the localization of the wave functions and we show that for barriers thicker than about 10 nm a single-well analysis is sufficient. A comparison with experimental results shows good agreement.

The Vogel-Fulcher-Tammann (VFT) ( sigma varies as exp(-E_a/k(T-T₀))) temperature dependence of ionic conductivity has been found for the Na_3-2xCa_xPO₄ orientational glass at x=0.5. The VFT behaviour of sigma is accounted for within a two-level phenomenological model including attractive interactions between the mobile ions. In this case, an increase in the E_a-value for the phase transition of Na₂Ca_0.5PO₄ to a modulated state (q_1,2=0.167a* and q₃=0.263a*, where a* is the reciprocal-lattice cell constant of the cubic modification of the crystal) at about 518 K is due to a deepening of the low-energy level.

The electronic properties of a series of nearly stoichiometric sputtered a-GaP films have been investigated as a function of increasing deposition and annealing temperatures up to 270 degrees C. The optical absorption coefficient (10 cm^-1< alpha <10⁵ cm^-1) as a function of photon energy, deduced both from transmission (T) and reflection (R) measurements and from photothermal deflection spectroscopy (PDS), shows a very large shift of the edge towards lower energies relative to that of c-GaP. The electrical conductivity data have an activation energy that varies continuously with temperature. The results are related to the atomic structure of films which was reported earlier by Elgun, Gurman and Davis and also to the microstructure which is shown to contain voids. In addition, a density-of-states picture based on theoretical calculations by O'Reilly and Robertson is used for interpretation of the results.

Magnetic phase transitions as a function of concentration and temperature for ordered and disordered (Mn_1-xFe_x)Pt₃ alloys are investigated by neutron diffraction and magnetic measurements. On the basis of these results the magnetic phase diagrams are drawn. The diagram for the ordered alloys is characterized by a tricritical point in the region of the ferromagnetic-to-antiferromagnetic transition at X_c=²/₃. A canted noncollinear magnetic structure and a region of two-phase mixtures are shown. The diagram for the disordered alloys is characterized by two regions of heterogeneous ferromagnets, both showing re-entrant spin-glass states at low temperatures, but differing in the character of exchange interaction between magnetic atoms. The diagrams are discussed within the framework of the phenomenological theory of the phase transitions with related order parameters.

W11, W12, W13 and W14 are four paramagnetic defects with S=³/₂ observed in electron or neutron irradiated type Ib diamonds. Although spin Hamiltonian parameters have been determined that describe the ESR results very well, it will be shown that the normal spin Hamiltonian with S=³/₂ is not entirely sufficient to describe the angular dependence of the ESR spectra. Possible tentative models for the defects will be discussed.