Phase diagrams based on numerical studies are presented for a Frenkel-Kontorova model in which the periodic potential V is either a cosine or a scalloped function formed from identical parabolas, while the interatomic potential W is the minimum of two offset parabolas, and thus is not convex. The phase diagrams exhibit Υ points, which also occur in some models in which W is convex, and superdegenerate points, which are only possible when W is not convex. Explicit conditions are given for the existence of superdegenerate points. In contrast with other nonconvex cases, all phase transitions in the neighborhood of a superdegenerate point are of the singular-continuous (‘‘devil’s staircase’’) type. This, and also the presence of Υ points, can be understood through the fact that W is locally convex, with a positive second derivative wherever it exists, so that defect and interface interaction energies are positive.

With a pump-probe laser-induced transparency technique the ground-state recovery of the F center in hydrogenated NaBr is measured. For a random distribution of F and substitutional hydrogen (${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$) centers a nonexponential F-center relaxation is observed, faster than the one in undoped crystals. The aggregation of F and ${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$ centers by subsequent optical bleachings is monitored by means of resonant Raman scattering. The corresponding relaxation measurements reveal three decay components, whch we associate with the formation of ${\mathit{F}}_{\mathit{H}}$(${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$) and ${\mathit{F}}_{\mathit{H}2}$(${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$) centers and to the remaining F centers. The relaxation time of the ${\mathit{F}}_{\mathit{H}}$(${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$) center is established to be ∼290 ps, which is 20 times faster than the one of the pure F center. An estimate of the decay rate assuming dipole-dipole coupling between the F-center electron and the ${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$ vibration shows that this interaction may play an important role in reducing the lifetime of the excited F center. The frequency shift of the hydrogen vibration of the ${\mathit{F}}_{\mathit{H}}$(${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$) center with respect to the one of the ${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$ centers cannot be explained by diple-dipole coupling only. This is possibly related to a displacement of the ${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$ equilibrium position of the ${\mathit{F}}_{\mathit{H}}$(${\mathrm{H}}_{\mathit{s}}^{\mathrm{-}}$) center, away from the lattice site.

${\mathrm{Cr}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Fe}}_{\mathit{x}}$ alloy films (≃3000 Å) were prepared by thermal coevaporation at room temperature at a deposition rate of 25 Å/min. It was found that a pure amorphous phase can be obtained in the composition range 0.25<x<0.60, while out of this range the bcc crystalline phase appears. The structural characterization and the magnetic properties were studied by x-ray diffraction and Mössbauer spectroscopy, respectively. The crystallization temperature for an amorphous ${\mathrm{Cr}}_{70}$${\mathrm{Fe}}_{30}$ film was found to be about 630 K as determined by in situ Mössbauer measurements. Since Cr-Fe is not a ready-glass-forming system, the amorphization process was discussed through the thermodynamics of the system and the kinetics of the condensation process.

The formalism for the generalized Kanzaki lattice static method is developed for a metallic crystal with bcc symmetry. The interactions up to second-nearest neighbors are included in the derivation of the dynamical matrix, impurity-induced force, and the atomic displacements. The formalism is applied to calculate the strain field due to 3d, 4d, and 5d substitutional transition-metal impurities (Ti, Cr, Mn, Fe, Nb, Mo, Ta, and W) in the vanadium metal, the only bcc systems in which the study of electric field gradients has been performed. The effective ion-ion interaction potential for the transition metals, proposed by Wills and Harrison, which includes properly the d-band effects, is used in the numerical calculations. In all the systems the atomic displacements are calculated up to 21 nearest neighbors and these show oscillatory behavior. The maximum displacement (or strain field) is caused by a Ta impurity and is about 3.5% of the first-nearest-neighbor distance. The strain field is found to depend, both in strength and range, on the excess ion-ion interaction potential due to the impurity. The atomic displacements exhibit the same trend, as shown by x-ray studies of the fractional change in the lattice parameter.

Using both numerical and analytical techniques, we investigate various ways to enhance the cubic nonlinear susceptibility ${\mathrm{\chi }}_{\mathit{e}}$ of a composite material. We start from the exact relation ${\mathrm{\chi }}_{\mathit{e}}$ =${\mathit{tsum}}_{\mathit{i}}$${\mathit{p}}_{\mathit{i}}$${\mathrm{\chi }}_{\mathit{i}}$〈(E⋅E${)}^2$${\mathrm{〉}}_{\mathit{i},\mathrm{l}\mathrm{i}\mathrm{n}}$/ ${\mathit{E}}_0^4$, where ${\mathrm{\chi }}_{\mathit{i}}$ and ${\mathit{p}}_{\mathit{i}}$ are the cubic nonlinear susceptibility and volume fraction of the ith component, ${\mathit{E}}_0$ is the applied electric field, and 〈${\mathit{E}}^4$${\mathrm{〉}}_{\mathit{i},\mathrm{l}\mathrm{i}\mathrm{n}}$ is the expectation value of the electric field in the ith component, calculated in the linear limit where ${\mathrm{\chi }}_{\mathit{i}}$=0. In our numerical work, we represent the composite by a random resistor or impedance network, calculating the electric-field distributions by a generalized transfer-matrix algorithm. Under certain conditions, we find that ${\mathrm{\chi }}_{\mathit{e}}$ is greatly enhanced near the percolation threshold. We also find a large enhancement for a linear fractal in a nonlinear host. In a random Drude metal-insulator composite ${\mathrm{\chi }}_{\mathit{e}}$ is hugely enhanced especially near frequencies which correspond to the surface-plasmon resonance spectrum of the composite. At zero frequency, the random composite results are reasonably well described by a nonlinear effective-medium approximation. The finite-frequency enhancement shows very strong reproducible structure which is nearly undetectable in the linear response of the composite, and which may possibly be described by a generalized nonlinear effective-medium approximation. The fractal results agree qualitatively with a nonlinear differential effective-medium approximation. Finally, we consider a suspension of coated spheres embedded in a host. If the coating is nonlinear, we show that ${\mathrm{\chi }}_{\mathit{e}}$/${\mathrm{\chi }}_{\mathrm{coat}}$≫1 near the surface-plasmon resonance frequency of the core particle.

Recently, the non-Newtonian viscosity of silicate liquids at high stress levels has attracted a fair amount of attention. Different explanations for this interesting phenomenon have been proposed. The purpose of the present paper is to demonstrate that an extension of the configurational entropy theory for the occurrence of the glass transition by Adam and Gibbs [J. Chem. Phys. 43, 139 (1965)] suffices to explain the development of non-Newtonian viscosity in liquids under high stress before rupture occurs. Attention is drawn to the fact that under steady-state conditions the logarithmic values of reduced viscosities of silicate liquids under high stress show a linear dependence on the square of the applied stress. This relationship suggests that the elastic work done by the stress on the liquid is the cause of the generation of configurational entropy. Adding this contribution to the configurational entropy term in the Adam-Gibbs equation for the ensemble average of the configurational transition probability is all that is needed to explain the heretofore mentioned non-Newtonian viscosity in liquid silicates and a molecular-dynamics Lennard-Jones liquid.

The cluster expansion technique is used in combination with first-principles calculations of the total energy of ordered compounds to study the energetics of the disordered phase for a number of alloy systems. The effect of short-range order, as seen in the energy differences between an alloy with the configuration corresponding to finite temperature and the perfectly random one is studied. The role of relaxation due to large size mismatch is discussed in terms of an effective cluster volume approximation. Very good agreement with measured energies of formation is obtained when short-range order and relaxation effects are taken into account.

Nonequilibrium AlSi solid solutions have been prepared by means of thermal annealing (T∼1000 K) at high pressure (P∼4 GPa). This synthesis is possible due to the metallic character of silicon under high pressure which greatly enhances the solubility of silicon in aluminum. The study of these supersaturated AlSi solid solutions has enabled us to investigate the chemical destabilization of a crystalline lattice. We experimentally show by means of inelastic neutron scattering and low-temperature specific heat that, as the silicon concentration in the aluminum lattice is enhanced, transverse acoustic phonon modes are becoming softer. This corresponds to a dramatic decrease of the shear modulus. Using differential scanning calorimetry, the positive heat of silicon dissolution in aluminum is measured under normal pressure: ΔH=38 kJ/mole. We interpret it as being mainly due to the energy difference between a metallic bonding of silicon in aluminum compared to its usual covalent bonding after demixing. This large heat of dissolution appears as the major reason for the observed lattice instability. Comparison of our experimental results with corresponding effects observed in usual amorphous metals suggests that further increase of the silicon concentration would induce a structural transformation of this crystal if the silicon segregation could be prevented.

Interstitial solid solutions of H in random fcc metal alloys have been examined in detail as far as the nature of interstitial sites and their spatial correlations are concerned. Both octahedral and tetrahedral sites have been considered and distinguished on the basis of the atomic composition and atomic arrangements within their first shell of neighbors. The conditional probability ${\mathrm{q}}_{\mathrm{uu}\prime }$ of finding a site of type u′ as nearest neighbor of a given site of type u were calculated. Expressions for selective blocking factors and occupancy probabilities have been derived for the approximation of the hard-core model, assuming blocking of sites at distances smaller than about 0.21 nm from the occupied site. The application of the model to the ${\mathrm{Pd}}_{1\mathrm{-}\mathrm{y}}$${\mathrm{Pt}}_{\mathrm{y}}$ system shows that H solubility data can be interpreted on the basis of a site-filling process. Finally, transition probability factors have been estimated to predict the occurrence of anelastic relaxations.

In the present work, the electronic structure of polythiophene at several doping levels is investigated by the use of the Hückel Hamiltonian with σ-bond compressibility. Excess charges are assumed to be stored in conformational defects of the bipolaron type. The Hamiltonian matrix elements representative of a bipolaron are obtained from a previous thiophene oligomer calculation, and then transferred to very long chains. Negative factor counting and inverse iteration techniques have been used to evaluate densities of states and wave functions, respectively. Several types of defect distributions were analyzed. Our results are consistent with the following: (i) the bipolaron lattice does not present a finite density of states at the Fermi energy at any doping level; (ii) bipolaron clusters show an insulator-to-metal transition at 8 mol % doping level; (iii) segregation disorder shows an insulator-to-metal transition for doping levels in the range 20–30 mol %.

The evolution of a one-dimensional Ising model with Glauber dynamics and an extrinsic energy barrier, submitted to continuous cooling and heating processes, is studied. The model is shown to exhibit many of the properties observed in real glasses. Under slow cooling, the dependence of the residual energy on the cooling law and rate is analyzed. The width of the glass transition is proportional to the inverse of the logarithm of the cooling rate. Heating processes can be studied in terms of a special solution of the master equation. Using the properties of this normal solution, the origin of the hysteresis effects when the system is cooled and reheated is discussed. The heat capacity under continuous heating presents a maximum, whose position and height can be used to characterize the thermal cycle described by the system.

We show that a previous failure to describe the frequency shift Δ and linewidth Γ of the optical zone-center mode in silicon with an anharmonic shell model arises from the neglect of the electronic polarizability effects in the perturbative treatment of the anharmonic interactions. We consider cubic anharmonic shell-shell interactions between nearest-neighbor atoms in a shell model used in the past for silicon. We apply an anharmonic perturbation theory which properly takes into account the adiabatic constraint for the shell’s polarization. We obtain Δ and Γ as functions of temperature in good agreement with experimental data.

We have studied the electronic structure of ${\mathrm{BaPb}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Bi}}_{\mathit{x}}$${\mathrm{O}}_3$ and its changes across the composition-dependent metal-semiconductor transition using photoemission and O 1s x-ray-absorption spectroscopy. For the parent insulator ${\mathrm{BaBiO}}_3$, the peak-to-peak splitting of the Bi 6s band is found to be large, whereas, the minimum gap is much smaller, qualitatively consistent with the large (∼2 eV) optical gap of the direct type and the smaller (∼0.5 eV) transport gap of indirect type. Pb substitution for Bi induces new states of Pb 6s character outside the band gap of ${\mathrm{BaBiO}}_3$ and does not induce in-gap spectral weight unlike in cuprate superconductors; the splitting of the Bi 6s band is not significantly reduced by Pb substitution throughout the semiconducting region. Substituted Pb remains tetravalent in the semiconducting phase and does not supply the Bi-O network with extra holes, which explains the stability of the semiconducting phase up to the Pb content of ∼65%. Nevertheless, the band gap collapses in the metallic region (x<0.35), where a clear Fermi edge is observed. The shifts of the core-level and valence-band peaks with x suggest that a small (∼0.3 eV) rigid-band shift of the Fermi level occurs in the metallic region. Comparison of the photoemission, optical absorption, and transport data on ${\mathrm{BaBiO}}_3$ strongly suggests that polaronic levels are created within the band gap of ${\mathrm{BaBiO}}_3$ and accommodate thermally excited charge carriers.

The electrical resistivity, Hall effect, and magnetic susceptibility of single-crystal ${\mathrm{UPd}}_2$${\mathrm{Si}}_2$ have been studied between 4.2 and 300 K. A large anisotropy was observed in both the magnetic and transport properties. There is a quadratic temperature dependence of the resistivity for a range of temperatures between 4.2 and 80 K. At higher temperatures, the resistivity indicates a Kondo-type behavior. The behavior of these quantities is accounted for by the magnetic phase transitions at 108 and 136 K reported from neutron-scattering studies. At high temperatures, the magnetic susceptibility of ${\mathrm{UPd}}_2$${\mathrm{Si}}_2$ is Curie-Wiess-like along the c axis. The temperature dependence of the Hall coefficient above 108 K is accounted for by a theoretical model invoking skew scattering of conduction electrons by localized magnetic moments.

The generalization of the local-density-approximation method for the systems with strong Coulomb correlations is proposed, which restores the discontinuity in the one-electron potential as in the exact density functional. The method is based on the model-Hamiltonian approach and allows us to take into account the nonsphericity of the Coulomb and exchange interactions. The calculation scheme could be regarded as a first-principle method due to the absence of adjustable parameters. The method was applied to the calculation of the photoemission (x-ray photoemission spectroscopy) and bremsstrahlung isochromat spectra of NiO.

Five signals which emerge after transmission through potassium slabs (∼0.1 mm thick) have been studied by Dunifer, Sambles, and Mace at 1.3 K with 79-GHz microwaves and large magnetic fields, Hz^, perpendicular to the slabs. They are (i) conduction-electron-spin resonance (together with spin-wave sidebands); (ii) Gantmakher-Kaner (GK) oscillations; (iii) cyclotron resonance; (iv) cyclotron-resonance subharmonics; and (v) high-frequency oscillations. Spin-resonance transmission has been understood for many years, but not the other four. The isotropic-Fermi-surface model is used here to calculate the transmitted power versus H by solving Maxwell’s equations self-consistently with the Boltzmann transport equation. GK oscillations do emerge, but the (theoretical) transmitted power is too large by a factor ∼10 000. In contrast, the remaining three signals should not even exist. It has been shown that a charge-density-wave broken symmetry creates a family of (higher-order) minigaps which cuts through the Fermi surface near its extremities along the slab normal. These gaps diminish the effectiveness of conduction electrons having ${\mathit{v}}_{\mathit{z}}$ near the Fermi velocity. Calculation confirms that this ineffectiveness dramatically reduces the predicted amplitude of the GK oscillation, and thereby provides an interpretation of the huge discrepancy. The minigaps also give rise to a small Fermi-surface cylinder which (as recently shown) leads to Landau-level oscillations that quantitatively explain signal (v). Signals (iii) and (iv) can also be attributed to the small Fermi-surface cylinder, but only because its axis is tilted ∼45° relative to z^. A nonlocal theory for such a scenario is yet to be formulated.

In order to reveal the microscopic origin of the metal-nonmetal transition in NiS, we have studied the Ni 2p x-ray-absorption spectra of substituted systems ${\mathrm{Ni}}_{1\mathrm{-}\mathit{x}}$${\mathit{M}}_{\mathit{x}}$S (M=Fe,Co), whose transition temperature ${\mathit{T}}_{\mathit{t}}$ varies from ∼500 K (${\mathrm{Ni}}_{0.5}$${\mathrm{Fe}}_{0.5}$S) to below 4 K (${\mathrm{Ni}}_{0.94}$${\mathrm{Co}}_{0.06}$S). Spectra taken at room temperature have been analyzed using a configuration-interaction cluster model to deduce changes in the electronic structure parameters, namely, the S-3p-to-Ni-3d charge-transfer energy Δ and the Ni-3d–S-3p transfer integral T. We find that Fe substitution does not increase the local Ni-S distance as would be expected from the lattice parameters but affects energy levels as is reflected in Δ. The changes in Δ and T derived from the cluster-model analysis explain the increase in the charge-transfer-type band gap in ${\mathrm{Ni}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Fe}}_{\mathit{x}}$S and hence the increase in ${\mathit{T}}_{\mathit{t}}$ within the cluster model, whereas they do not explain the decrease in ${\mathit{T}}_{\mathit{t}}$ in ${\mathrm{Ni}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Co}}_{\mathit{x}}$S. In order to explain the dramatic decrease in ${\mathit{T}}_{\mathit{t}}$ with Co substitution, we propose that the Ni-3d–Co-3d hybridization increases the bandwidths leading to the closure of the band gap.

We report a self-consistent linear-combination-of-Gaussian-orbitals study of the electronic states and ground-state geometry of an undoped and doped single, infinite chain of trans-polyacetylene using the density-functional theory in the local-density approximation. We find a dimerized ground state for an undoped chain with a dimerization amplitude of about 0.01 Å, which is lower than the experimental value of 0.023–0.03 Å. A pure Hartree calculation neglecting all exchange and correlation gives a much smaller dimerization amplitude of less than 0.005 Å. The local exchange-correlation energy thus significantly favors the dimerization although its effect is not strong enough. In the calculations of the doped chains, the dopant ions were approximated by a uniform background charge. We find that the undimerized state becomes energetically more favorable than any uniformly dimerized state at a critical doping level of about 0.04 (0.03) extra holes (electrons) per CH unit. The band structures and total energies of polaron and soliton lattices at a higher doping level of 0.2 holes per CH unit are calculated and compared with those of the uniformly dimerized and undimerized lattices, and possible models of the metallic state of trans-polyacetylene are discussed. According to our study, the bonds become increasingly similar with increasing doping. The undimerized chain model seems to be a good approximation for the metallic state of trans-polyacetylene at high doping levels although the possibility for a marginal soliton lattice cannot be fully excluded.

The angular correlation between Compton scattered photons and the recoil electrons has been measured in thin self-supporting carbon and aluminum foils using 100-keV synchrotron radiation from a wiggler at the DORIS III storage ring at Deutsches Elektronen-Synchrotron (DESY). The coincidence count rate is proportional to the three-dimensional electron momentum density (3D EMD) of the target, after integration over the energy distribution of the scattered photons and electrons it is proportional to the 2D EMD. The effect of multiple scattering of the recoil electrons within the target on the measured angular correlation is studied both by Monte Carlo calculations and by experiment.

We present calculations of interband optical conductivity of the compound ${\mathrm{Ni}}_3$Al using the linear-muffin-tin-orbital method. Calculations are performed using constant matrix elements and calculated matrix elements. We find that matrix elements play a significant role in influencing the magnitude of optical conductivity σ(ω) and the position of peaks. Since ${\mathrm{Ni}}_3$Al is weakly ferromagnetic (magnetic moment equal to 0.31${\mathrm{\mu }}_{\mathit{B}}$ per unit cell), we have also calculated σ(ω) using the ferromagnetic band structure but without spin-orbit coupling. A comparison of our results with experimental data is presented.