Table of contents

Neutron powder diffraction studies on PdD_0.65 demonstrate that the 50 K anomaly is due to an order-disorder transition of deuterium within octahedral interstitial sites. The slow transition to the ordered phase involves diffusion of deuterium from the nearest-neighbour deuterium positions to the second-nearest-neighbour positions. The ordered crystal structure remains cubic, and is accurately described in space group Pm3n by doubling the cell constant relative to the disordered structure (space group Fm3m).

In this paper the annealing of Na colloids in heavily irradiated NaCl with damage levels up to 10 mol% is discussed. Recently the melting properties of Na colloids in heavily irradiated NaCl have been studied using differential scanning calorimetry. The observation of three endothermal latent heat peaks, numbered 1, 2 and 3, as a function of the dose, has been reported. Here we discuss the annealing behaviour of heavily irradiated crystals doped with KBF₄, which show a large melting peak 3. The changes in the appearance of the latent heat peak provide us with information about the properties of the sodium colloids. The annealing of the colloids in two different temperature ranges is described. At moderate annealing temperatures between 160 and 210 degrees C the peak shifts with annealing time to lower melting temperatures very rapidly. The shift of the peak is activated by an energy of 1-1.4 eV and strongly depends on the initial amount of latent heat. At high temperatures between 250 and 320 degrees C a decrease of the latent heat is measured, due to the back reaction between Na colloids and molecular Cl₂. After continued annealing at high temperatures the melting behaviour becomes very anomalous. The melting peak splits into two peaks, the main peak and a new, sharp one arising at 92 degrees C. After the splitting of the peak the back reaction slows down significantly, activated by an energy of 1.2+or-0.3 eV. Both the moderate- and high-temperature annealing behaviour indicate that the processes are very local, with a short length scale of the order of 1 nm. This implies that the morphology of this type of colloid is highly irregular, with a dense and fine nanostructure.

Collective diffusion of lattice gases of arbitrary concentrations is investigated in lattices of dimensions d>or=2 with randomly distributed site energies. The Gaussian, exponential and dichotomic distributions of site energies are employed. An effective medium theory of the coefficient of collective diffusion is given using symmetrized single-particle transition rates. A previous phenomenological theory is rederived in the limit d to infinity . The effective medium theory gives good agreement with simulation data at smaller particle concentrations, and rough agreement at large particle concentrations.

We investigate the electronic states in low-dimensional semiconductor systems created by delta doping of donors by solving Schrodinger and Poisson equations self-consistently. The increase of the density of states near the population threshold leads to the existence of two physical solutions to the same boundary-value problem due to the electron-donor dipole charge potential. An empty and a filled state can be self-consistent for a depletion potential in the two-dimensional system for heavy carriers as well as in the one-dimensional and zero-dimensional systems, The bisolution behaviour for electrons in the GaAs wire persists up to approximately 37 K if level broadening is negligible. We find that the energy levels of the lowest subband are well described using simple variational functions. The carrier density does not vanish gradually as the doping density that compensates the fixed depletion charge is decreased, indicating that true bistability does not take place in these systems.

A nanostructure with hard walls is divided into two regions: the terminals and the 'cavity' that connects them. In the absence of a magnetic field the modes in the terminals may be written down immediately. The modes in the cavity are calculated by extending each terminal mode continuously into the cavity and applying hard-wall boundary conditions elsewhere on the cavity boundary. Then a general wave function Psi continuous everywhere may be written down. The relationships between the coefficients of the modes in Psi are determined by minimizing the mean square discontinuity in the normal derivative of Psi averaged over all the interfaces between the terminals and the cavity. This method has been found to be robust and efficient in calculations of the conductance matrix and scattering wave functions for a 2D quantum wire with a hard-wall finger of various shapes pushed in through one side.

A spin-flop transition is observed when a field of 20 T is applied parallel to the b axis of a specimen of natural goethite at 4.2 K. The goethite orders in a four-sublattice antiferromagnetic structure with the sublattice magnetizations inclined at +or-13 degrees to b, and an iron moment of 3.90 mu _B. There is a small net ferromagnetic moment of 0.004 mu _B along b which has the effect of more than doubling the threshold field for spin flop. The anisotropy field is deduced to be 0.11 T, corresponding to a uniaxial anisotropy energy K=6*10⁴ J m^-3. The origin of the weak moment is discussed, and it is suggested that mode superparamagnetism may arise in non-collinear antiferromagnets when two transverse spin configurations are effectively degenerate.

The relaxation rate for depolarization of a positive muon implanted in an isotropic magnetic salt with ferromagnetic exchange interactions is studied theoretically, on the basis of the coupled-mode theory of critical and paramagnetic spin fluctuations and a full numerical evaluation of the dipole field experienced by the muon. The main findings from studies of realistic models of EuO and EuS are (a) a significant dependence of the relaxation rate, lambda , on the assumed position of the implanted muon and (b) a monotonic temperature dependence, with lambda approximately xi ^3/2 in the approach to the critical temperature at which the correlation length, xi diverges. In contrast, previous results for a model of an isotropic magnet with an antiferromagnetic exchange, RbMnF₃, show that lambda for this magnet is not a monotonic function of the temperature, and in the precursor region to T_c, lambda increases with decreasing temperature with a power law behaviour lambda approximately xi ^1/2. The calculated values of lambda for EuO are consistent with data from preliminary experiments on the same salt.

The spin crossover transition curve of the compound (Fe(BPTN)(NCS)₂) (BPTN=N,N'-bis(2-pyridylmethyl)-1,3-propanediamine) was measured by Mossbauer spectroscopy and susceptibility measurements. The compound was prepared by two different methods yielding a fine powder (a) and single crystals of approximately 0.5 mm in size (b). The transition temperatures T_1/2 of the gradual transitions are 182 K (a) and 186 K (b). Within experimental accuracy the transition curves derived from the fractional absorption areas of the Mossbauer spectra are the same as derived from susceptibility measurements so that there is no indication for different Lamb Mossbauer factors of the high-spin (HS) and low-spin (LS) states as reported in the literature. The observation of quite different Debye temperatures Theta _D at temperatures where the compounds are completely converted to the HS and LS state is discussed.

This paper presents the results of dielectric susceptibility investigations performed in the radiowave, microwave, submillimetre wave and infrared (IR) regions on the AgTaO₃-AgNbO₃ solid solution with Nb contents of 0.3, 0.4 and 0.6. An intense dipolar relaxation, which is responsible for the temperature dependence of the static permittivity and which freezes out at low temperatures, has been evidenced in the submillimetre wave region.

The absorption spectra of ZnS:Ni and ZnSe:Ni were measured at 100 K under hydrostatic pressure up to 10 GPa. Two kinds of optical transition were investigated: (i) the d-d transitions between crystal field levels of Ni²⁺ in the tetrahedral field of the host lattice and (ii) a charge transfer band that occurs when an electron is transferred from the valence band to the Ni²⁺ ion. The pressure dependence of the crystal field parameter D_q and of the Racah parameter B were determined and a revised assignment of the d-d transitions was obtained. The charge transfer can be considered as a deep-acceptor transition. In this case the Ni acceptor level may be used as an energy reference, which enabled us to determine the hydrostatic deformation potential of the valence band.

In transmission electron microscopy, Al core losses excited by currents localized about atomic cores show fringes similar to the thickness fringes observed in zero loss, The fringes observed in the inelastically scattered electrons result from intraband scattering in the dispersion of Bloch waves even when the losses are over 1 kV. The analysis uses dynamical scattering with inelastic scattering. The impact parameter method is used to determine transition rates from electron currents distributed across the unit cell of the crystal.

A Kronig-Penney model with a constant electric field F for a non-interacting election is used to study the transmission properties of the Anderson transition in one-dimensional systems with disordered delta -function potentials. We examine the cases where the potential varies uniformly from 0 to W (barriers) or from -W to 0 (wells) for a given disorder W. We observe jumps in the transmission coefficient at the points E+Fx=n² pi ² (where E is the electron energy and n an integer). These jumps are related to the small oscillations observed by Soukoulis et al. in the mixed case (potentials from -W/2 to W/2). However, an interesting feature is found in the wells in the range between two jumps. It is observed that in the presence of a small field the states become more localized and the localization length decreases up to a minimum for a critical value F_m instead of increasing. Finally, we have studied the effect of the disorder on the Anderson transition by means of the participation ratio and the localization length.