Table of contents

Responding to the potential of a new neutron-beam method for studying the time-dependent properties of magnetic fluctuations, well suited for quasi-two-dimensional magnets, we predict the form of the signal for scattering by paramagnetic materials. The model used is expected to describe and related quasi-two-dimensional magnets, with an Ising-like character. Calculations are made using the coupled-mode theory of spin fluctuations. Applied to an isotropic Heisenberg model, results from the theory are not the same as those stemming from Fick's Law for spin diffusion . For we predict out-of-plane fluctuations to be the same as for the Heisenberg model, and in-plane fluctuations to be different and an exponential function of time.

The drainage of liquid in a foam may be described in terms of a nonlinear partial differential equation for the foam density as a function of time and vertical position. We review the history and recent development of this theory, analysing various exact and approximate solutions and relating them to each other.

The scattering from binary molten salt mixtures of nominally two thirds alkali halide concentration, where x = `natural', `zero-scattering' and `62' enrichments of nickel, has been measured by time-of-flight neutron diffraction on ISIS LAD. A complementary experiment was performed on molten samples, also of nominally two thirds alkali halide concentration, using isotopic substitution of chlorine. In the case of - KCl, the results indicate a large first sharp diffraction peak (FSDP) in and but not in . Reverse Monte Carlo (RMC) modelling indicates a well-ordered, almost regularly tetrahedral local structure for the ion, the preponderance of large ions as next-nearest neighbours to the metal cation and the apparent association of this with enhanced intermediate-range order (IRO) in the mixtures. The results confirm the findings of the initial composition study of this system and provide further support for the suggestion that enhanced IRO arises from ordering between tetrahedral units and alkali counter-ions. Modelling of the data suggests a strong similarity with molten even though the pure and molten salt structures are confirmed to be dissimilar.

Recent studies of the dynamic theory of liquids have been much influenced by Mori's formulation of the generalized Langevin equation. An essential component of this formalism is the memory function. For realistic models like a Lennard-Jones liquid it is too difficult to calculate the memory function. Recent approaches have been to propose various forms for it including the hyperbolic secant. In perhaps the most serious attempt yet, Tankeshwar and Pathak have been said to derive hyperbolic secant memory directly from the generalized Langevin equation. Here the validity and justification of this work is closely examined. Other approximate studies which have yielded hyperbolic secant memory are also examined and comparisons are made.

We have simulated a liquid layer consisting of quasi-Stockmayer molecules with the molecular dynamics method. Graphical illustration of the ensemble shows that the molecules form chains. The calculation of the chain size distribution for the bulk part and for the surface of the layer shows that in the bulk part of the liquid longer chains are more likely to occur. We reason that this is due to the higher mobility of the molecules in the surface region which lowers the probability of bond formation.

In addition to this we have analysed the topological correlation between chains. To do so a distance between chains has been defined to obtain a criterion for whether two chains are neighbours or not. We have examined the angles between those molecules of two neighbouring chains which are closest or between those with the lowest pair energy. It seems that there is a slight preference that the two closest molecules of the neighbouring chains align their dipoles. If the chain sizes of the two participating chains exceed three an increase in the probability that the two closest molecules antialign their dipoles can also be observed. A rise in the dipole moment intensifies this behaviour. Lowering the moment of inertia yields a decrease in the probability to antialign the neighbouring dipoles. A possible reason is the loss of stability of the antiparallel arrangement due to an increased amplitude of the rotational vibration caused by the smaller moment of inertia. However, the parallel arrangement is still a preferred orientation between the correlated dipoles.

The electrical conductivity , thermopower S and magnetic susceptibility of liquid have been measured as a function of c. It is found that decreases monotonically on addition of AgCl to S and reaches a value typical of a molten salt when at which point a semiconductor to molten salt transition takes place. The thermopower is negative for c > 0.3 at which composition it changes sign. The magnetic susceptibility measurements show a steady increase with c, consistent with the changes in the ionic diamagnetism and a significant Pauli contribution from the electrons for the S rich compositions. The behaviour of the system is analysed and described in terms of a simple ionic band picture using a Kubo - Greenwood formalism in which the conductivity gap increases with increasing concentration of AgCl.

The reorientation process of a molecular probe dissolved in the fragile glass former o-terphenyl is studied via electron spin-resonance spectroscopy. Owing to the good resolution of the ESR lineshape in the so-called slow-motion regime, we are able to discriminate between diffusive and jump rotations. In the region of the glass transition ( the results are at variance with the diffusion models and are fairly well reproduced by assuming that the probe jumps with steps . The temperature dependence of the correlation time is well described, in the temperature range investigated, by a double Arrhenius law broken at . The findings are compared with previous NMR, dielectric, fluorescence and molecular dynamics studies.

Composition-dependent mechanical properties and free volumes are compared for miscible, amorphous blends of bisphenol-A polycarbonate (PC) with (a) polyaryloxysiloxane (PAS), (b) a copolyester of 1,4-cyclohexanedimethanol and a mixture of isophthalic and terephthalic acids (EASTAR) and (c) an experimental polyester of 1,4-cyclohexane dicarboxylic acid and 1,4-cyclohexanedimethanol (CDACD). The free volumes were measured by the positron annihilation lifetime spectroscopy (PALS) technique. The strength of specific interactions, as indicated by data, is relatively weak in all of the blends. However, the fractional free volume quantity measured by PALS is less than additive in the polyester blends and is additive, or greater than additive, in the PC - PAS blends. The mechanical behaviour of the blends can be rationalized in terms of the free volume behaviour. The polyester blends which lose free volume (contract) on mixing exhibit higher than averaged yield strengths and brittle impact responses. The PC - PAS blends which retain or gain free volume on mixing exhibit averaged yield strengths and averaged ductile impact responses of the constituent polymers.

The volume dependence of local electronic and magnetic properties of has been investigated using the self-consistent spin-polarized linear muffin-tin orbital (LMTO) calculation. The results obtained indicate that the Fe magnetic moment at the face-centred sites decreases more rapidly than that at the corner sites with decreasing unit-cell volume. The decrease of Fermi-contact hyperfine field at Fe sites with the decrease of the unit-cell volume not only results from the corresponding decrease of the Fe magnetic moment, but also is affected by the abnormal change of the transferred at these sites. The discussion on at Fe sites partly reveals the unit-cell volume dependence of hyperfine interactions among atoms. Moreover the change of the calculated Mössbauer isomer shift (IS) at Fe sites with decreasing unit-cell volume is in agreement with the experimental results.

The near-edge absorption fine structure of carbon in graphite has been calculated using the multiple-scattering approach. The effects of both a neutral carbon absorber and taking into account the effect of the core hole by using a nitrogen (Z + 1) absorber atom are considered. The best agreement with experiment was obtained for the case of the nitrogen absorber and a cluster of 145 atoms, with the calculated near-edge structure (NES) containing the main features observed experimentally. A splitting near the onset of the -peak, similar to that seen experimentally, was found to occur in the calculated NES for one of the two inequivalent carbon atoms in the graphite structure. This splitting was not observed in the calculated NES for a single sheet of graphite, supporting the proposition that it only occurs in well ordered graphite. This result supports the proposition that the splitting originates from a sharp -peak and a dipole-allowed bandlike contribution associated to the branch of the free-electron-like interlayer states.

thin films, with x = 0.14 and 0.23, were prepared from alternate Al and Fe layers by irradiation with a Xe beam at two different temperatures in order to reach either the amorphous or the quasicrystalline state. Through thermal treatment, crystalline alloys are obtained. Rutherford back-scattering spectrometry of -particles and transmission electron microscopy have been used to characterize the atomic distribution and structural state of the samples. X-ray emission spectroscopy and conversion electron Mössbauer spectroscopy provided information about the electronic distributions whereas electrical resistivity measurements allowed us to determine the metallic character of the compounds. The existence of a pseudo-gap and dehybridization of the s and d electronic states of Fe are suggested to interpret the high-resistivity behaviour and the isomer shift values, measured in the Fe concentration range studied here.

The projection of the eigenfunctions obtained in standard plane-wave first-principles calculations is used for analysing atomic orbital basis sets. The `spillage' defining the error in such a projection allows the evaluation of the quality of an atomic orbital basis set for a given system and its systematic variational optimization. The spillage is shown to correlate with the mean square error in the energy bands obtained from the projected Hamiltonian matrix. The method is applied to the characterization of finite-range pseudo-atomic orbitals (Sankey O F and Niklewski D J 1989 Phys. Rev. B 40 3979) in comparison to infinite-range pseudo-atomic and Slater-type orbitals. The bases are evaluated and optimized for several zinc-blende semiconductors and for aluminium; the finite-range orbitals display high quality in spite of the limited range. A simple scheme is proposed to systematically enlarge the basis without increasing its range.

Bayesian probability theory along with the maximum entropy concept is widely used for inferential problems, particularly to infer dynamic properties of strongly correlated fermion systems from quantum Monte Carlo (QMC) imaginary time data. In current QMC applications, however, the error covariance of the QMC data is not treated consistently. Here we present a consistent Bayesian analysis of all the information provided by the QMC simulation. This approach allows us to infer reliable results with the least amount of computer time.

Neutron diffraction measurements on the title compounds have been carried out over a range of temperature from 10 to 300 K. The compounds have a rhombohedrally distorted perovskite structure. A paramagnetic to ferromagnetic transition is observed for samples with . The ordered magnetic moments obtained for x = 0.2 and 0.3 are found to be 1.5 and respectively at 10 K and are in agreement with those reported from other macroscopic measurements. A possible magnetic structure is discussed.

For technical applications the commonly used thermal processing of the amorphous alloy consists of one-hour anneals at about . This procedure leads to very soft magnetic properties due to the formation of the nanocrystalline Fe(Si) phase in the remaining amorphous matrix. In order to study the onset and development of the nanocrystalline phase the as-quenched amorphous samples were annealed in an Ar atmosphere at temperatures of and for different annealing times: 15 s, 30 s, 60 s, 90 s, 120 s, 180 s and 300 s. For sample characterization we used the conventional Mössbauer effect technique and additionally Mössbauer measurements in radiofrequency magnetic fields, which are very sensitive to magnetostriction and magnetic anisotropies. The measurements were supplemented with electron transmission microscopy and electron diffraction studies. Our short-time annealing experiments clearly show that in order to form good magnetically soft properties the annealing times of 1 - 2 min at and of only 15 - 30 s at are sufficient.

Pulsed nuclear quadrupole resonance (NQR) and Raman spectroscopy studies were performed to get information about the structural and dynamic thermal behaviour of the dichlorobiphenyl. NQR measurements of the line-shape and the spin - lattice relaxation time were obtained in the temperature range 80 K - 320 K. Raman spectra from powder samples were recorded for temperatures between 4.5 K and 300 K. Although the morphology of the dichlorobiphenyl molecule is very close to the biphenyl one, there is no evidence of any structural instability, such as the incommensurate phase transitions in biphenyl. Up to the lowest temperatures scanned, the compound seems to remain in the same crystalline ordered phase reported by x-ray diffraction at room temperature. The and Raman measurements suggest that there are no significant dynamical changes (for example, the softening of a mode) in the scanned temperature range. Some considerations are made on the size of the para-substituents of biphenyl, the crystal packing and the stability of the periodic phases.

On the basis of the superposition principle for the optical anisotropy parameters and the Jones calculus methods a phenomenological model is developed to interpret the optical properties of absorbing, spatially modulated dielectric materials. The character of the normal electromagnetic waves propagated in a uniaxial, weakly dichroic crystal for which the complex gyration tensor is modulated with the simple square wave form is ascertained. The structure with a perfect modulation wave and the unipolar structure are considered and should describe the multidomain ferroelectric and incommensurately modulated phases. Specific crystal optical effects are found which have no analogues in uniform materials. Proceeding from the normal wave polarization the modulated crystal is shown to be equivalent to a low-symmetry biaxial crystal. The origin of a non-zero eigenwave ellipticity is discussed in connection with the problem of optical activity in macroscopically centrosymmetric incommensurate crystals of group.

Results of optical absorption, photoluminescence, fluorescence lifetime and fluorescence-line-narrowing studies on several -doped crystals with the gallogermanate crystal structure are presented. Such crystals are unusual in that most fluoresce via both the broad-band transition and the sharper transition (R line), both of which are considerably broadened by the effects of substitutional disorder. An established adiabatic model is used to reproduce the selectively excited broad-band photoluminescence spectra and this shows that the substitutional disorder creates a distribution in the strength of both the octahedral crystal field and the electron - phonon coupling. The variations in electron - phonon coupling are attributed to variations in the strength of the non-octahedral crystal fields that determine the magnitudes of the coupling to non-symmetric phonon modes. This effect is shown to dominate the broadening of the band. The analysis of results of fluorescence-line-narrowing measurements on the R lines shows that these lines are also broadened by a distribution of both octahedral and non-octahedral crystal fields. In the case of crystals that contain a substitutionally disordered site in the close vicinity of the ion, the strong-field sites that give rise to R-line emission are shown to comprise an independent distribution and the broadening is dominated by the effects of the non-octahedral component of the crystal field. When the substitutionally disordered sites are further from the ion, the broad-band and R-line fluorescence derive from a single distribution of crystal fields. In this case the effect of the variation of the crystal field is again shown to be dominated by the non-octahedral contribution though the R-line fluorescence shows little evidence of this due to the strong-field sites being the extreme of the overall distribution.

Photoluminescence spectra from quantum well structures grown at high temperatures are investigated. The luminescence properties are found to be very different from those of free excitons. To describe correctly the spectral lineshape, the radiative recombination from excitons trapped on the local potential fluctuations in quantum wells must be considered. At low sample temperatures, the luminescence is mainly from the trapped excitons. With increasing temperature, trapped excitons are thermally activated into free excitons and luminescence peaks shift to higher energies. By comparing measured spectra with the calculated spectra, the trap density and the trap energy are derived. The origin of the trap and its relation with the crystal growth are discussed.

Photoemission measurements have been made of the intermetallic compound utilizing synchrotron radiation in the energy range 40 - 180 eV. The valence band extends to a binding energy of 7 eV. Measurements made at the Cooper minimum for Pd 4d indicate that the highest density of states for Ti 3d is at the Fermi edge and that it decreases towards higher energy. Using this density of states and correcting for the photoemission cross-sections with energy, it has been possible to determine the density of states for Pd 4d which also contributes to the Fermi level. Resonance photoemission measurements are consistent with this conclusion and also indicate that the Pd 4d band is not completely full.

The following corrections should be made to the above paper:

1. In the first line of the abstract, `Polycrystalline pyrolytic graphite was' should read `Polycrystalline and pyrolytic graphite were'.

2. In the third line from the bottom of page L637, `in a sample volume' should read `in the sample volume'.

3. In the first line of page L640, `3 mu m' should read `3 µm'.

4. In the tenth line from the bottom of page L641, the word `sufficiently' should be replaced with `substantially'.