Table of contents

The authors have measured the optical absorption spectrum of YBa₂Cu₃O_{7- delta} crystalline thin films and powders in the range 2000 to 10000 cm^-1. A broad absorption band near 5500 cm^-1 (0.7 eV) is observed, which has the same absorption profile in both cases. Their work on thin-film samples shows that the excitation causing this absorption is polarized in the ab plane. Further work on doped powder samples provides evidence that the excitation is localized in the CuO₂ planes of the structure, which are known to be the principal carriers of the superconductivity. Studies of the temperature dependence of the absorption have revealed that the peak intensity undergoes a transition at the superconducting transition temperature, providing firm evidence that the excitation is coupled to the superconducting order parameter.

The results of measurements of the internal friction and Young's modulus in yttrium 1-2-4 superconducting ceramics are reported. The internal friction maxima found in the temperature range from 30 K to 100 K are of relaxation nature. The activation energy and relaxation time of these phenomena have been estimated. The association of the Cu-O chain relaxation with one of the internal friction phenomena is briefly discussed.

The product obtained by the pyrolysis of a mixture of triphenoxy-triazine(TPTA) and benzoguanamine(BG) at 400-1000 degrees C has exhibited novel ferromagnetic behaviour. The product with the highest spin concentration (1.4*10²⁰ spins g^-1) was obtained when a TPTA/BG mixture of molar ratio 1.1:0.9 was pyrolysed at 700 degrees C. The possible compound structures of the products prepared at 350 degrees C and 600 degrees C are proposed from the results of elemental analysis. The ferromagnetic behaviour is considered to be caused by parallel spins of the nitrogen and/or oxygen radicals in those structures.

Many materials, including soap froths, polycrystalline alloys, ceramics, lipid monolayers and garnet films, have structures composed of either two- or three-dimensional polygonal domains separated by well defined boundaries. Usually, the surface energy of these boundaries makes the pattern unstable, causing certain grains to shrink and eventually to disappear. Thus the pattern coarsens continuously unless other factors arrest the motion of the boundaries. The authors review recent theoretical, computational and experimental progress in their understanding of the asymptotic scaling laws that describe coarsening. In most cases the elementary expectation, that the mean grain radius scales with the square root of time, is confirmed. They pay particular attention to the history of the field, to understand why this elementary result has remained in doubt until now.

A theory of elastic strain relaxation is developed for ternary semiconductor alloys A_1-xB_xC with bond length mismatch. The theory is applied to the problem of local structure in Zn_1-xMn_xSe using two models that separately consider random and ordered cation-site occupation. Comparison of results from the models with experimental data for the local structure suggests that the alloy is random and that both anion and cation sublattices relax.

With the aid of the defect-specified supercells used within the framework of the linear-muffin-tin-orbital (LMTO) method, the familiar difficulties in calculating the anti-structure (AS) defects in FeAl can be overcome. The calculations are based on the spin-dependent version of the Hohenberg-Kohn-Sham scheme. It is hence possible to obtain the electronic structure of the alloy along with the integral properties, e.g. local magnetic moments of defects. The occurrence and magnitude of the local moment associated with an AS-Fe defect is in agreement with the experimental findings.

The dynamics of Davydov solitons within the so-called mod D₁> state which allows quantum effects in the lattice are studied at physiological temperatures using Davydov's averaged Hamiltonian method. For this purpose the Euler-Lagrange method is used to obtain approximate equations of motion from a thermally averaged Hamiltonian. Within mod D₁> dynamics at T=0K and for parameter values appropriate for proteins, no solitons are found. It is demonstrated that temperature effects at 300 K shift the stability window for travelling solitons into regions of the parameter space which might be realistic for proteins.

The ionic distribution in the high-temperature rotor phase of lithium sulphate, Li₂SO₄, has been reinvestigated by single-crystal neutron diffraction at 923 K. Thermal vibrational tensor terms up to the fourth rank are refined to give a better description of the space-time-averaged Li⁺ and SO₄^2- distributions. The picture of Li⁺ mobility that emerges supports well the earlier proposed notion of rapidly reorienting SO₄^2- ions which effectively gate the movement of Li⁺ ions, i.e. a 'paddle-wheel mechanism'. The mode of transport for the Li⁺ ions would appear to be jumps of 3.5-3.7 AA between approximately=90% occupied, tetrahedrally distorted 8c sites (at ¹/₄, ¹/₄, ¹/₄), into a spherical shell of radius 2.8 AA surrounding each of four adjacent tetrahedrally coordinated SO₄^2- ions. Each shell contains, in total, approximately=0.2 Li⁺ ions. These Li⁺ ions can then jump further into one of the six adjacent 8c sites. The Li⁺ transport thus occurs in one of the six directions (110), (110), (101), etc., and gives some support to the notion (corroborated by earlier MD simulation) that the pathways are slightly curved, and hence somewhat in excess of 3.5 AA. This agrees well with the value of 3.7+or-0.4 AA deduced from Raman spectroscopy, and also with diffuse neutron scattering studies which suggest Li⁺-Li⁺ pair correlation distances of 3.6-3,7 AA.

Deposition of Na on 1*1 and 7*7 structures of Si(111) at room temperature (RT) forms a uniform layer with a rather weak metallic character. The coverage of Na can be determined at any time of the Na deposition by a flux meter. The binding energy of Na is greater on a reconstructed 7*7 than on the 1*1 structure of a Si(111) surface. The initial dipole moment and the maximum Delta phi are greater for Na on the 1*1 structure. Near saturation Na coverage, the oxygen goes initially under the Na layer forming a Na-O-Si complex and subsequently resides on top of it. The presence of Na on Si(111) causes a drastic increase of the initial sticking coefficient of oxygen by about ten times. When oxygen is adsorbed on Na-covered Si(111) surfaces at RT and the system is heated to 700 degrees C, Na is removed from the surface while oxygen forms SiO₄ and SiO₂. By repeating this cycle several times (seven) SO₂ predominates and its average thickness is about 9 AA. However, this thickness is smaller than when Cs is used instead of Na. This indicates that Cs is more effective as a promoter of the oxidation. However, the oxide formation occurs at a lower temperature than with Cs additives. Finally, the effect of Na on the oxidation of Si was almost the same on the 7*7 and 1*1 structures of Si(111) surfaces.

It is shown how information about the pseudoatom in germanium can be derived from knowledge of the charge density gained from total-energy calculations. The charge densities of the equilibrium crystal and of a crystal where the ions in each cell have been moved by small, equal amounts delta R in opposite directions (the LTO mode at Gamma ) are used to derive the Fourier components of the pseudoatoms at reciprocal lattice vectors. The total-energy calculations are carried out by solving the Kohn-Sham equations within the local density approximation. The harmonic contribution to the Fourier components of the change in charge density is derived from total-energy calculations at two different values of delta R. By using symmetry, linear equations involving the real and imaginary parts of the rigid ion and the deformation can be obtained from a total-energy calculation with just one distortion of the crystal. The parts of these linear equations involving the unknowns are linearly dependent, and the calculated results have approximately the same dependence. It is shown that any discrepancies are due to anharmonic effects. The results show that the rigid ion is atomic-like with only a small deviation from spherical symmetry, and that the deformation is significant and acts mainly on the bond charges. Contour plots of the various contributions, including the anharmonic one, in the (110)-(001) plane are drawn.

Results of numerical investigations of the Anderson model of localization are reported. Using the transfer matrix method and finite size scaling, the localization lengths of electronic states in 2D systems with energetic disorder are determined on honeycomb, square, and triangular lattices. While the actual values of the localization length at a particular disorder differ for the different coordination numbers, complete localization is found in all cases. This result is in agreement with the scaling hypothesis, but in contrast to recent claims.

The authors present a theoretical analysis of positron trapping into large voids in metals. They show that, even if the trapping rate vanishes as the positron energy goes to zero, a significant amount of trapping of non-thermalized positrons ensures a substantial trapping probability even at very low temperatures.

A correlation between collections of the experimental values of the crystal-field parameters (P₂⁰, b₂⁰, A₂⁰) characterizing the nuclear quadrupole interaction, the ground-state and excited-state splitting for ¹⁵⁷Gd³⁺ in various ionic crystals with the same type of anions has been established. The correlation consists of a tendency towards a decrease in P₂⁰ and an increase in A₂⁰ through a series of these collections provided that they are arranged according to descending values of b₂⁰. The great variety of the centres involved and the considerable magnitudes of the intervals within which the parameters of the collections obeying the correlation fall suggest that this correlation is a highly general characteristic of the collections (P₂⁰, b₂⁰, A₂⁰) and its explanation may be regarded as an important criterion of validity for microtheory crystal-field models. The authors have developed a model to interpret the correlation mentioned and some cases of deviation from it. The phenomenological values of the Sternheimer antiscreening coefficient gamma _x have been determined in different ways. These values agree with each other and with the value of gamma _x for Pr³⁺ calculated by Ahmad and Newman (1980) taking the interelectron correlation into account in addition to the interactions considered by Sternheimer originally.

The authors present measurements of the anisotropy between the longitudinal and transverse magnetoresistance in a number of Fe_1-xZr_x (0.08<or=x<or=0.11) metallic glasses in fields ( mu ₀ H_a) up to 1 T at a number of fixed temperatures between 1.5 and 300 K, and in a single, low field ( approximately 5 mT) as a function of temperature. The influence of hydrogen loading was also studied. A low-field (remanent-field) resistive anisotropy (RFRA) first appears at a temperature very close to the Curie temperature (T_C) (found from previous AC susceptibility measurements on the same specimens), confirming the emergence of a substantial, non-local exchange field at this temperature. A linear increase of the RFRA with decreasing temperature immediately below T_C (i.e. as A(T_C-T)/T_C) is reported for the first time, and a semiquantitative explanation for such an increase proposed. Structure in the RFRA has also been observed at temperatures well below T_C, although no definite correlation between it and the proposed re-entrant transition in this system could be established.

A self-consistent calculation has been made, using non-local norm-conserving pseudopotentials and the local-density approximation, to study the electronic states for LEED determined structure of an ordered monolayer deposition of Sb on the InP(110) surface. A number of chemisorbed induced states have been identified in the various band gaps of InP. In particular it is found that there are three occupied and two unoccupied states in the fundamental band gap. In agreement with the inverse photoemission study of Drube and Himpsel (1988) the authors find that for the monolayer coverage of Sb there is no unoccupied electronic state in the fundamental band gap of InP. Charge density analysis reveals that while states localized on the second-substrate-layer atoms retain their identity, states localized on the top substrate layer atoms are changed into Sb-In and Sb-P bonding and antibonding states. In addition they identify states originating from the adsorbate layer containing the Sb-Sb chain.

A calculation is presented of the size dependence of the normal electron-electron scattering and the Umklapp electron-electron scattering contributions to the electrical resistivity of copper whiskers. The authors find that the normal scattering term is larger then the Umklapp scattering term when the diameter of the copper whisker is smaller then about 10 mu m. These results are in good agreement with recent measurements.

Monte Carlo simulations are presented for Ising dipoles on body centred cubic and tetragonal lattices. A finite size scaling form that includes logarithmic corrections is proposed and found to significantly improve the data collapse. With lattice parameters appropriate to LiHoF₄ the authors obtain a ferromagnetic transition temperature T_c=1.51 K in good agreement with experiment.

Expressions for the dipolar and hyperfine contributions to the relaxation rate of muons implanted in a ferromagnet are presented, and analysed using the Heisenberg model of spin waves including dipolar and Zeeman energies. Calculations for EuO indicate that the temperature dependence of the hyperfine and dipolar contributions to the relaxation rate are similar, so the latter contribution will dominate if the ratio of the hyperfine and dipolar coupling constants is indeed very small. The hyperfine mechanism is sensitive to the dipolar energy of the atomic spins, whereas the dipolar mechanism depends essentially on the exchange energy. For both mechanisms there is an almost quadratic dependence on temperature, throughout much of the ordered magnetic phase, which reflects two-spin-wave difference events from the Raman-type relaxation processes.

For pt.I see ibid., vol.4, p.2043 (1992). Mechanisms for the relaxation of muon spin relaxation signals in paramagnets are analysed in terms of spin fluctuations in a Heisenberg magnet with a ferromagnetic exchange interaction. Explicit predictions provided for EuO are based on the assumption that the implanted muon occupies an interstitial site of high symmetry. Because the average dipolar field at the muon site is zero, and the material is ferromagnetic, the dipolar mechanism of relaxation is not enhanced by critical spin fluctuations. In consequence, the hyperfine mechanism dominates relaxation in the critical region, even though the ratio of the hyperfine and dipolar coupling constants might be small. Calculations of relaxation rates for both mechanisms, made on the basis of coupled-mode theory, are provided for critical and paramagnetic states.

The Mossbauer spectra of a sample of synthetic ferrihydrite have been investigated in the 26-220 K temperature range in order to identify the mechanism giving rise to the relaxation times. The results show that the temperature dependence of the relaxation times is in accordance with the Vogel-Fulcher law. This indicates that the magnetic interaction between the crystallites is significant. From the temperature dependence of the mean magnetic field under T_B the authors also argue for a strong correlation between the spatial orientations of neighbouring clusters.

A modified expression for the axial pressure is derived and implemented within an LMTO-ASA framework. The axial pressures of non-collinear antiferromagnets nets in FCC manganese and iron are calculated and bed to make predictions on the axial distortion and relative stability of the phases.

Electronic absorption, luminescence and laser excitation spectra of the complex ion Eu(N₃)₈^5- in single crystals of Cs₅Eu(N₃)₈ have been measured at temperatures down to 4 K. Although the crystal structure shows the presence of only one type of Eu(H₃)₈^5- complex ion, in which nearest N atoms form an approximate D_4d (square anti-prismatic) coordination polyhedron about the Eu atom, the spectra clearly show the presence of two distinct, but closely similar, types of Eu atoms. For one of these sites, emission from the ⁵D₀ state to all the expected crystal field states of the ⁷F_J(J=0-6) manifold and from the ⁵D₁ state to all of the ⁷F_J(J=0, 1, 2, 3) states is observed and, together with the absorption and the excitation spectra, permit the location of 56 of the 57 crystal field states expected below 23000 cm^-1.

The authors present the results from constant-pressure constant-temperature molecular dynamics simulations on the fluoride-perovskites: KMnF₃, KZnF₃ and KCaF₃. These simulations lead to the predictions that KMnF₃, and KZnF₃ are not superionic conductors while KCaF₃ shows limited superionic behaviour with a diffusion coefficient of 5.24*10^-6 cm² s^-1 ( sigma =0.145 Omega ^-1 cm^-1) and T_c=0.93 T_m. However, these results are only qualitative since the use of the rigid-ion model resulted in simulated temperatures above the experimental melting points.