We present molecular-dynamics and energy-minimization calculations for glassy ${\mathrm{Cu}}_{33}$${\mathrm{Y}}_{67}$ and for crystalline Cu-Y alloys with nonadditive Lennard-Jones interactions. The results reproduce satisfactorily the x-ray and neutron data on the glass and the structure of the crystalline alloys. The calculations suggest that atomic compressibility or energetic effects concur with atomic-packing constraints to determine the glass structure.

NMR spectra have been measured for the quadrupole perturbed 1/2→-1/2 transition of ${}_{}{}^{93}{}_{}{}^{}\mathrm{Nb}$ in barium sodium niobate (BSN) between 300 and 650 K. The spectra are inhomogeneously broadened in the incommensurate (I) phase between ${\mathit{T}}_{\mathit{I}}$=580 K and ${\mathit{T}}_{\mathit{c}}$=540 K as well as in the quasicommensurate (QC) phase below ${\mathit{T}}_{\mathit{c}}$. However, the observed line shapes do not exhibit the characteristics of the lines for other 1q and 2q modulated incommensurate materials. The temperature dependence of the observed line shape can be explained using a model in which phase fluctuations induced by random, weak pinning of the modulation wave by defects produce Gaussian line broadening. Random pinning of the modulation wave also creates a large phason gap which makes spin-lattice relaxation via phasons inefficient. These results suggest that the presence of defects due to nonstoichiometry causes the I and QC phases of BSN to be ‘‘chaotic.’’

The Bragg diffraction of waves in one-dimensional doubly periodic media is analyzed by means of Kogelnik’s coupled-waves technique. The spectrum problem and the problem of reflection from a half-space and from a layer are considered. It is shown that a devil’s-staircase type of spectrum causes characteristic peaks and valleys in the frequency dependence of the reflection coefficient.

The interface thickness in two triblock copolymers were determined using small-angle x-ray scattering in the context of the theory proposed by Ruland. The thickness was found to be nonexistent for the samples at three different temperatures. By viewing the systems as ideal two-phase systems with a sharp boundary, the roughness of the interface was described by fractal dimension ${\mathit{D}}_{\mathit{s}}$ which slightly increased with temperature. Length scales in which surface fractals are proved to be correct exist for each sample at every temperature, and become wider as the temperature increases. The interface in block copolymers was treated as fractal, which reasonably explained the differences between Porod’s law and experimental data, and gained an insight into the interaction between different segments on the interface.

We report phason defects in the stable Al-Cu-Co-Si decagonal quasicrystal obtained by an Ammann-line analysis of a tiling derived from a high-resolution electron microscope image, covering an area of about 25×34 ${\mathrm{nm}}^2$. Ammann-line jags, corresponding to matching rule violations, are about 4% of the total tiling edges. This supports the quasiperiodic-tiling model of quasicrystals.

Electron-spin-resonance and dc conductivity data show a thermal-history-dependent transition at 240 K in pressed pellets of ${\mathrm{ClO}}_4^{\mathrm{-}}$-doped poly(3-methylthiophene) (P3MT). We discuss the possibility of this transition to be a Peierls transition from a room-temperature-metallic to a charge-density-wave state driven by anions ordering at this temperature. Below 100 K, dc conductivity shows a change from linear to exponential decay. Nonlinear conductivity has also been observed in this system for very low electric fields.

A scaling theory of the weakly nonlinear conductivity and flicker noise near the percolation threshold of a two-component composite is formulated and some of its physical consequences are examined. Both components are assumed to have finite Ohmic conductivities, but with a ratio that is very different from 1. We predict that under certain conditions a nonmonotonic dependence of the weakly nonlinear conductivity on the volume fraction of the good conductor can be observed, with a maximum very near to the threshold and a local minimum somewhat above it. We also predict that, depending on the values of the important physical parameters of the system, either the good conductor or the poor conductor may dominate the critical behavior both above and below the threshold. Some of these predictions are compared with results of numerical simulations on two-dimensional random resistor networks and with an effective-medium approximation (EMA). The numerical results are shown to reproduce some of the predictions of the scaling theory. The EMA is shown to follow the same qualitative behavior but to reproduce poorly the quantitative results.

We analyze the possibility of the appearance of long-range correlations of states of quenched defect systems in disordered solids. For randomly distributed defects with a finite number of degenerate internal degrees of freedom such correlations are shown to appear near the defect percolation threshold if they obey certain nearest-neighbor correlation rules. The corresponding effective Hamiltonian can be viewed as that of a generalized m-component long-range correlated Weinrib-Halperin model. Our renormalization-group investigation shows, however, that it asymptotically decomposes into a set of m noninteracting one-component Weinrib-Halperin models. The coordinates of the stable fixed point of this model are determined to O(${\mathrm{\epsilon }}^{1/2}$).

We introduce an operator algebra for the description of the low-energy physics of one-dimensional, integrable, multicomponent quantum liquids. Considering the particular case of the Hubbard chain in a magnetic field and chemical potential, we show that at low energy its Bethe-ansatz solution can be interpreted in terms of a pseudoparticle-operator algebra. Our algebraic approach provides a concise interpretation of, and justification for, several recent studies of low-energy excitations and trasnport which have been based on detailed analyses of specific Bethe-ansatz eigenfunctions and eigenenergies. A central point is that the exact ground state of the interacting many-electron problem is the noninteracting pseudoparticle ground state. Furthermore, in the pseudoparticle basis, the quantum problem becomes perturbative, i.e., the two-pseudoparticle forward-scattering vertices and amplitudes do not diverge, and one can define a many-pseudoparticle perturbation theory. We write the general quantum-liquid Hamiltonian in the pseudoparticle basis and show that the pseudoparticle-perturbation theory leads, in a natural way, to the generalized Landau-liquid approach.

We show that the recently developed pseudoparticle-operator algebra which generates the low-energy Hamiltonian eigenstates of multicomponent integrable systems with contact interactions also provides a natural operator representation for the Virasoro algebras associated with the conformal-invariant character of the low-energy spectrum of these models. Studying explicitly the Hubbard chain in a nonzero chemical potential and external magnetic field, we establish that the pseudoparticle-perturbation theory provides a correct starting point for the construction of a suitable critical-point Hamiltonian. We derive explicit expressions in terms of pseudoparticle operators for the generators of the Virasoro algebras and the energy-momentum tensor, describe the conformal-invariant character of the critical point from the point of view of the response to curvature of the two-dimensional space time, and discuss the relation to Kac-Moody algebras and dynamical separation.

The vibrational spectroscopy of hydrogen isotopes in α-phase solid solution with Lu was probed by incoherent-inelastic-neutron-scattering methods. Low-resolution spectra of H and D located at tetrahedral interstices are qualitatively similar to those of other rare-earth/hydrogen α phases, indicating a vibrational mode along the c direction that is ∼30% softer and significantly broader than the doubly degenerate vibrational modes in the basal plane. High-resolution spectra of these broad c-axis vibrations reveal a temperature- and concentration-dependent line shape believed to consist of local acoustic and optic bands associated with the dynamical coupling of hydrogen atoms paired on either side of a metal atom along the c direction. Previous diffuse-elastic-neutron-scattering studies indicate the existence of short-range ordering of these hydrogen pairs into c-axis-oriented chains possessing interchain correlations. The shape and width of these bands at low temperature near the α/β-phase boundary compared to those of the analogous Sc and Y α-phase systems suggest that the extent of ordering in α-${\mathrm{LuH}}_{\mathit{x}}$ is intermediate with respect to that in the (less ordered) α-${\mathrm{ScH}}_{\mathit{x}}$ and (more ordered) α-${\mathrm{YH}}_{\mathit{x}}$ solid solutions.

The in-plane thermal conductivity ${\mathrm{\kappa }}_{\mathrm{\parallel }}$ has been measured over the temperature range 5–400 K for samples of chemical-vapor-deposited (CVD) diamond made by both the microwave and hot-filament processes. The samples span a range of defect level, grain size, and degree of thinning. Comparison with a model of heat transport suggests that ${\mathrm{\kappa }}_{\mathrm{\parallel }}$ is limited by scattering of phonons from point defects, extended defects of ∼1.5 nm diameter, dislocations, grain boundaries, and microcracks, as well as by phonon-phonon scattering at high temperatures. The Callaway model of thermal conductivity is used to include the effects of normal three-phonon scattering processes. In the higher-conductivity samples, scattering of long-wavelength phonons is very weak even at grain boundaries, indicating relatively smooth boundaries. The value of ${\mathrm{\kappa }}_{\mathrm{\parallel }}$=20 W ${\mathrm{cm}}^{\mathrm{-}1}$ ${\mathrm{K}}^{\mathrm{-}1}$ at room temperature for some of the microwave CVD samples is the highest reported to date for CVD diamond. Measurements of the anisotropy in conductivity obtained from the measured perpendicular conductivity ${\mathrm{\kappa }}_{\mathrm{\perp }}$ consistently show a higher conductivity along the (columnar) grains. The hot-filament-CVD sample measured exhibits a room-temperature conductivity approaching that of the best microwave-plasma samples, indicating that the thermal conductivity is determined more by the specific conditions of growth than by the type of CVD growth (microwave or hot filament).

The presence of a magnetic field in an optically active medium produces a rotation of the polarization of light: it is the well-known Faraday effect, which breaks the time-reversal symmetry. The averaged light intensity in the multiple scattering of light by disordered systems is described by the weak-localization theory based on the direct and reverse sequences of scatterings, which are founded on the time-reversal symmetry. The multiple scattering of electromagnetic (vectorial) waves by spherical particles is considered in the presence of a magnetic field. We have shown that the electric field of the reversed path can be obtained from the direct one by a simple matrix transposition. In systems of reduced dimensionality (1 and 2), we have shown that for the same polarization channel, the peak of the backscattering cone is not affected by the Faraday effect even though the time-reversal symmetry is broken. The intensity correlation function is obtained for a one-dimensional system. This simple model furnishes two results: (i) even though the wave vector is randomized, there is no decorrelation of the polarization for paths of the same length and (ii) the correlation function has an oscillatory behavior as a function of the magnetic field. In three dimensions, we have calculated analytically the attenuation of the backscattering cone as well as the decorrelation length for the multiple Rayleigh scattering. Mie scattering has been considered by Monte Carlo simulations. In the diffusion regime (thick slabs) our results are in accord with previous results and with experiments. Nevertheless, for the intermediate regime in transmission, we have found oscillations of the intensity correlation as a function of the magnetic field. For reflection and strong magnetic field, we have observed the convergence of the enhancement factor to nontrivial asymptotic values.

The polaron path integral introduced by Feynman is generalized to the case of an electron moving in d space dimensions. A scheme of systematic calculations of the obtained path integral is developed to estimate the ground-state energy of the polaron in the same way for different values of the electron-phonon coupling constant α. The leading-order term in this approach yields an upper bound to the polaron self-energy and improves Feynman’s variational estimates for d=2 and d=3. A scaling relation between polaron self-energies for different space dimensions is obtained for this term. The next correction to this estimation is calculated by numerical integration for d=2 and d=3.

We present an ab initio calculation of lattice-dynamical properties of the alkaline-earth oxides MgO, CaO, and SrO in the rocksalt structure. We employ the density-functional perturbation theory within the local-density approximation. We show phonon dispersion curves along several symmetry directions in the Brillouin zone. The calculated frequencies agree well with the available experimental data. The investigation of the phonon-induced charge-density fluctuations of MgO and CaO at the L point of the Brillouin zone partially supports the breathing-shell model of lattice dynamics and rules out the charge-transfer model for this class of materials. Moreover, our calculations show that the breathinglike charge-density response is more pronounced for oxygen than for the cations in these compounds.

The magnetocrystalline anisotropy (MA) of ferromagnetic fcc Ni, fcc Co, and bcc Fe (001) slabs is calculated to second order in the spin-orbit coupling constant using realistic tight-binding models with parameters fitted to available first-principles electronic-structure calculations. Results for slabs with 1–17 atomic layers show that MA favors magnetization perpendicular to the slab for Fe but in-plane for Ni and Co, in agreement with experiment. Fluctuations in the MA as the thickness is varied indicate that surface MA is not purely of local origin in the surface atomic layer.

The structure of epitaxial Fe/Pt(001) multilayers deposited by e-beam evaporation was characterized by a four-circle x-ray diffractometer and the magnetic anisotropy was investigated using a torque magnetometer. For epitaxial Fe/Pt(001) multilayers, a structural transition of the Fe layer is found to occur with increasing Fe layer thickness. The critical thickness is around 8 Å. Magnetic anisotropy and coercive force change with the transition. The out-of-plane volume anisotropy decreases for the Fe layer thickness less than 8 Å. The fourfold in-plane anisotropy is lost for the Fe layer thickness less than 8 Å. For the Fe layer thickness beyond 8 Å, it is found that the in-plane interface anisotropy is negative and that the in-plane volume anisotropy is identical to the magnetocrystalline anisotropy of bulk bcc Fe.

Moriya’s perturbation theory is applied to the distorted ${\mathrm{NiO}}_2$ and ${\mathrm{CuO}}_2$ planes, and the effective spin Hamiltonians are derived. Spin-wave theory is explored for the Hamiltonians, and the spin-wave excitations are studied for nickelates and cuprates with low-temperature orthorhombic and low-temperature tetragonal (LTT) structures. It is shown that in the cuprates the origin of the anisotropy responsible for the spin-wave gaps is different for in-plane and out-of-plane modes: i.e., anisotropic superexchange for the in-plane mode, and anisotropic direct exchange for the out-of-plane mode. This difference explains the observed structural dependence of the spin-wave gaps in ${\mathrm{La}}_{1.65}$${\mathrm{Nd}}_{0.35}$${\mathrm{CuO}}_4$. In nickelates, the single-ion anisotropy is solely responsible for the spin-wave gaps, which are observed in ${\mathrm{La}}_2$${\mathrm{NiO}}_4$. We discuss the mechanism of the weak ferromagnetism; in the nickelates the Dzyaloshinski-Moriya (DM) interaction causes the weak ferromagnetism in the LTT phase where an appropriate spin configuration is prepared by the single-ion anisotropy. In the cuprates, where the single-ion anisotropy for preparing the spin configuration is absent, the DM and pseudodipolar interactions themselves determine the spin configuration: the weak ferromagnetism originates only from the competition between these two interactions. It is proposed that a multiorbital effect is essential to explain the observed weak ferromagnetism in the LTT phase of ${\mathrm{La}}_{1.65}$${\mathrm{Nd}}_{0.35}$${\mathrm{CuO}}_4$.

In this paper we discuss the specific features of K-edge magnetic circular x-ray-dichroism (MCXD) spectra of Fe and Co in bcc Fe-Co alloys and Fe/Co multilayers. In the alloys, the systematic variations of shape and amplitude of the near-edge MCXD signals as a function of concentration can be correlated with the changes in local atomic environment. The difference between the Co K-edge and Fe K-edge spectra of the same alloy shows that the signal is central atom dependent. Well above the absorption edge (E-${\mathit{E}}_0$>80 eV) the extended x-ray-absorption fine structure (EXAFS)-like oscillations in the MCXD spectra clearly reflect the local bcc structure and appear not to depend on the absorbing atom. In both alloys and multilayers, the magnetic EXAFS intensity can be correlated to the average magnetization in the neighborhood of the absorbing atom.

Finite cluster models and a variety of ab initio wave functions have been used to study the electronic structure of bulk ${\mathrm{KNiF}}_3$. Several electronic states, including the ground state and some charge-transfer excited states, have been considered. The study of the cluster-model wave functions has permitted an understanding of the nature of the chemical bond in the electronic ground state. This is found to be highly ionic and the different ionic and covalent contributions to the bonding have been identified and quantified. Finally, we have studied the charge-transfer excited states leading to the optical gap and have found that calculated and experimental values are in good agreement. The wave functions corresponding to these excited states have also been analyzed and show that although ${\mathrm{KNiF}}_3$ may be described as a ligand-to-metal charge-transfer insulator there is a strong configuration mixing with the metal-to-metal charge-transfer states.