We compare several electronic structure models (based on density-functional theory, or pseudopotential theory with linear or nonlinear screening) and liquid structure theories (variational approach or integral equation methods) applied to the calculation of thermodynamic functions in liquid simple metals. The numerical consequences of these choices are shown and discussed. Comparison with other work and experiment is made. A prescription for choosing the bridge function used in the modified hypernetted-chain equation is proposed. The central role of the ionic excess entropy in the study of liquid structure and its relation with the ion-ion interaction is emphasized.

Measurements of the electric conductivity and the relaxation times for reorientation in magnetic fields are presented for the cesium perfluorooctanoate (CsPFO)–${\mathrm{D}}_2$O micellar system, as a function of temperature and concentration. Five sample concentrations are presented, namely 41%, 48%, 50%, 55%, and 60% of CsPFO in ${\mathrm{D}}_2$O by weight. At these concentrations, the system shows a nematic liquid crystal, between a lower-temperature lamellar smectic and a higher-temperature isotropic micellar phase. At the transition from the nematic to lamellar phase, the principal conductivities change rather continuously, and do not show any marked features. The measurements show that at the same reduced temperature, the structure of the nematic phase becomes more anisotropic and the relaxation times become longer with decreasing CsPFO concentration in the system, suggesting that at the same reduced temperature, the size of the nematic micelles increases with decreasing CsPFO concentration. Comparison of the conductivity anisotropy data with values calculated from structural models allows estimates of the aspect ratio of the micelles and the fraction of the area of the smectic planes that is covered by water. The behavior of the relaxation times at the nematic-to-lamellar transition shows dependence on the concentration. The relaxation times at the nematic-to-lamellar transition for the four lower concentrations diverge; the critical exponents in order of increasing CsPFO concentration are 0.71, 0.76, 0.77, and 0.85. The 60% sample does not show any pretransitional behavior at this transition.

We report on a high-resolution x-ray study of the smectic-A phase near the nematic–smectic-A transition of a side-group polymeric liquid crystal, polyacrylate. Except at the very near vicinity of the phase transition, the first- and the second-harmonic x-ray structure factors were found to be consistent with the harmonic theory of de Gennes and Caille. However, very close to the phase transition, the attempts to fit the two harmonics completely failed. We believe that this is evidence for the importance of anharmonic corrections near the phase transition. The fits to the experimental data yield the compressibility constant B and the splay elastic constant ${\mathit{K}}_{\mathit{s}}$. We found that B can be described by a power law B∼${\mathit{t}}^{\mathrm{\phi }}$, where t=(${\mathit{T}}_{\mathit{c}}$-T)/${\mathit{T}}_{\mathit{c}}$,${\mathit{T}}_{\mathit{c}}$ being the phase-transition temperature. φ was found to be 0.82±0.08, in good agreement with theoretical predictions using exponents from the literature, but in disagreement with previous experimental results on monomeric liquid crystals. The splay elastic constant ${\mathit{K}}_{\mathit{s}}$ has roughly the same magnitude as in monomeric liquid crystals but tends to decrease by approximately 50% upon approaching the transition from below. This temperature dependence is further evidence for the importance of anharmonicity in the system.

The saddle-splay modulus K¯ of a surfactant monolayer determines the stability of a spherical droplet microemulsion phase. Negative values of K¯ increase the droplet-size polydispersity, but inhibit shape fluctuations. Large values of the observed polydispersity may therefore not be indicative of small values of the bending modulus K, but of a ratio of K¯/K∼-2.

As a primitive model for structural breakdown in elastic media, we analyze the failure of random resistor-fuse networks with various distributions of properties. We show that variations in breakdown voltage have a more significant effect than variations in resistance values. This is analogous to the fluid-displacement problem [D.Y.C. Chan, B. D. Hughes, L. Paterson, and C. Sirakoff, Phys. Rev. A 38, 4106 (1988)], in which variations in fluid capacity have a greater effect on displacement efficiencies than variations in permeability. An exponential distribution of breakdown voltages creates much more disorder than any uniform distribution, but power-law distributions that emphasize weak bonds can create even greater disorder, up to the percolation limit, in which bonds are broken independently at random.

We consider the statistical mechanics of a system of semiflexible chains, which can represent polymer liquid crystals, long-chain biomolecules, stiff wormlike micelles, or columns of discotic liquid crystals. Semiflexible chains can order in a hexagonal lattice (also called a columnar or canonic phase), and they can form nematic and isotropic fluids. We calculate the structure factor and the director fluctuation spectrum in the hexagonal and nematic phases using continuum elastic theory. The structure factor and the director fluctuation spectrum of the hexagonal phase are found to be intimately related to those of smectic-A liquid crystals through a ‘‘line-plane’’ duality. The elastic constants of the hexagonal phase are computed for the case of pure steric repulsion between chains. These elastic constants show ‘‘universal’’ behavior in the close-packing limit where steric repulsion dominates.

We develop a model for the phase transitions of a system of infinitely long, semiflexible chains that interact through a steric, excluded-volume repulsion. This model combines density-functional theory for freezing with the Onsager theory of orientational ordering. We derive phase diagrams in terms of pressure or density versus persistence length or temperature. If the temperature is low or the persistence length is long, the phase diagram shows an isotropic fluid, a nematic liquid-crystal phase, and a hexagonal lattice (also known as a columnar or canonic phase) as the pressure or density is increased. If the temperature is high or the persistence length is short, the phase diagram shows a direct isotropic-hexagonal transition, with no intermediate nematic phase. The theoretical phase diagram agrees qualitatively with several experiments on lyotropic discotic liquid crystals, and it should also apply to polymer liquid crystals, long-chain biomolecules, and wormlike micelles.

This paper reports on studies of wetting phase behavior and interfacial structure of square-well fluid adsorbed at square-well walls. Choosing a particular wetting isotherm at bulk liquid-vapor coexistence, we present our final and complete comparison between molecular-dynamics (MD) simulation and weighted-density-approximation (WDA) density-functional theory. The properties of wall-liquid, wall-vapor, and liquid-vapor interfaces are measured and used to determine contact angles and to locate the positions and order of interfacial phase transitions (wetting and drying). Due to the presence of a suitable collective mode, it was possible to directly observe the collective dynamics of the fluctuation-induced first-order drying transition in MD simulation. A practical implementation of contact-angle measurement by generalized WDA density-functional theory is detailed, enabling one to input the bulk equation of state as a boundary condition. An attempt is made to gauge the generality of our results by comparison with simulation data from other systems and with alternative versions of WDA theory.

Possible phases in a nematic liquid crystal confined to a spherical submicrometer droplet embedded in a solid polymer are analyzed in terms of a Landau–de Gennes theory. For a droplet with a radial structure we show that the strength of the nematic-polymer interfacial interaction affects the nematic-paranematic (partially ordered isotropic phase) phase transition and may in addition induce a boundary-layer nematic phase. This boundary layer phase exists only in a narrow (∼0.1 K) temperature interval above the nematic phase for a restricted range of interfacial interactions. Also in the radial structure the degree of ordering is suppressed close to the center of the droplet where a defect is located. As the size of the droplet decreases, the relative size of this region of suppressed ordering increases. Below a critical radius ${\mathit{R}}_{\mathit{c}}$ (0.22 μm for 4-n-pentyl-4’-cyanobiphenyl), if the surface interaction is above a critical value (${\mathit{q}}_{\max }$=1.85×${10}^{\mathrm{-}3}$), the transition between the nematic phase and the paranematic phase no longer occurs. A three-dimensional phase diagram is presented to demonstrate the effect of the surface interaction strength, droplet radius, and sample temperature on the stability of phases within a droplet.

The electrostatic interaction between two planar dipolar lattices in dielectric continua is investigated. If there are no dielectric images, the pressure is equally likely to be attractive or repulsive, depending upon the relative lateral displacement of the two lattices. In addition to perfect dipolar lattices, we consider the effects of real dipoles, of disorder, and of finite-sized domains. When the surfaces constitute the boundary between dielectric media, and when there are no correlations between the two surfaces, then there occurs a strong exponentially decaying repulsion due to the interaction between a lattice and its dielectric images. The relevance of this repulsion to the hydration force is discussed, and an experimental test that should discriminate between different theories for that force is proposed.

Viscous fingering at a nonzero viscosity ratio on percolating clusters is considered to study morphological changes of patterns formed by the injected fluid in porous media. A fraction P of bonds is filled by the displaced fluid, while the others (1-P) are blocked, where P is the usual percolation probability. Fluid with a low viscosity is injected into the percolating cluster filled by the displaced fluid with high viscosity. Morphological changes of patterns of the injected fluid are described in terms of crossover phenomena by making use of a four-parameter position-space renormalization-group method. It is found that when ${\mathrm{\mu }}_{\mathit{I}}$/${\mathrm{\mu }}_{\mathit{D}}$≪(P-${\mathit{P}}_{\mathit{c}}$)≪1 the double crossover occurs from the diffusion-limited aggregation (DLA) on an incipient percolation cluster through the DLA on the perfect lattice to the dense structure, and when 1≫${\mathrm{\mu }}_{\mathit{I}}$/${\mathrm{\mu }}_{\mathit{D}}$≫(P-${\mathit{P}}_{\mathit{c}}$) the other double crossover appears from the DLA on an incipient percolation cluster through the invasion percolation to the dense structure, where ${\mathrm{\mu }}_{\mathit{I}}$/${\mathrm{\mu }}_{\mathit{D}}$ is the viscosity ratio and ${\mathit{P}}_{\mathit{c}}$ the critical percolation probability.

The effect of hydrodynamic flow upon diffusion-limited deposition on a line is investigated using a Monte Carlo model. The growth process is governed by the convection and diffusion field. The convective diffusion field is simulated by the biased-random walker resulting from a superimposed drift that represents the convective flow. The development of distinct morphologies is found with varying direction and strength of drift. By introducing a horizontal drift parallel to the deposition plate, the diffusion-limited deposit changes into a single needle inclined to the plate. The width of the needle decreases with increasing strength of drift. The angle between the needle and the plate is about 45° at high flow rate. In the presence of an inclined drift to the plate, the convection-diffusion-limited deposit leads to the formation of a characteristic columnar morphology. In the limiting case where the convection dominates, the deposition process is equivalent to ballistic deposition onto an inclined surface.

We have expanded our earlier Monte Carlo model [Phys. Rev. A 38, 2447 (1988); J. Crystal Growth 100, 313 (1990)] to three dimensions and included reevaporation after accommodation and growth on dislocation-induced steps. We found again that, for a given set of growth parameters, the critical size, beyond which a crystal cannot retain its macroscopically faceted shape, scales linearly with the mean free path in the vapor. However, the three-dimensional (3D) the systems show increased shape stability compared to corresponding 2D cases. Extrapolation of the model results to mean-free-path conditions used in morphological stability experiments leads to order-of-magnitude agreement of the predicted critical size with experimental findings. The stability region for macroscopically smooth (faceted) surfaces in the parameter space of temperature and supersaturation depends on both the surface and bulk diffusion. While surface diffusion is seen to smooth the growth morphology on the scale of the surface diffusion length, bulk diffusion is always destabilizing.

The atomic surface roughness increases with increase in growth temperature and supersaturation. That is, the tendency of surface kinetics anisotropies to stabilize the growth shape is reduced through thermal and kinetic roughening. It is also found that the solid-on-solid assumption, which can be advantageously used at low temperatures and supersaturations, is insufficient to describe the growth dynamics of atomically rough interfaces where bulk diffusion governs the process. For surfaces with an emerging screw dislocation, we find that the spiral growth mechanism dominates at low temperatures and supersaturations. The polygonization of a growth spiral decreases with increasing temperature or supersaturation. When the mean free path in the nutrient is comparable to the lattice constant, the combined effect of bulk and surface diffusion reduces the terrace width of a growth spiral in its center region. At elevated temperatures and supersaturations, 2D nucleation-controlled growth can dominate in corner and edge regions of a facet, while the spiral growth mode prevails in its center. Thus, in addition to confirming the experimental observation that the critical size of a growing crystal depends on the prevailing growth mechanism, we are able to obtain detailed insight into the processes leading to the loss of face and facet stability.

A method of simulating the response of an irregular interface is investigated. It is based on an exact mapping between the Laplace equation and the steady-state diffusion equation with mixed boundary conditions. Simulations in two dimensions show that diffusion-limited aggregation (DLA) and other self-similar fractal electrodes exhibit the so-called constant-phase-angle (CPA) behavior. In the case of DLA electrodes the CPA exponent is found by this method to be close to the inverse of the fractal dimension. We show that this is directly related to the fact that in d=2 the admittance is proportional to the overall size of the self-similar electrode.

The present investigations consider the case of a low-density laser-produced plasma expanding into a vacuum in the presence of an axial magnetic field. The time-integrated line intensities of neutral and singly ionized barium have been measured for magnetic fields up to 300 G. These measurements reveal three prominent changes in the intensities of individual lines as a function of increasing magnetic field: extinction, growth, and severe attenuation followed by enhancement. Measurements support a model that predicts the quenching of higher-lying transitions and the enhancement of lower-lying transitions for increasing magnetic fields.

Magnetic-field effects are known to play an important role in determining current-collection characteristics of electrodes in magnetoplasmas. In the Parker-Murphy model of current collection, for example, the magnetic field excludes electrons that originate from outside a characteristic collection radius from reaching the anode. In this paper we show that these excluded electrons can dominate current collection through ionization effects, even at pressures well below the Paschen threshold for ionization. We demonstrate this by releasing a controlled current near a spherical anode in a magnetic field, and measuring the resultant anode current. The use of a spatially localized electron source reveals aspects of current collection that are not evident when a background plasma constitutes the electron source. We find that the anode current can exceed the source current by more than two orders of magnitude. The conditions are such that we can neglect the contribution to the anode current from secondary emission at the chamber wall. We therefore attribute the additional current to the flow of electrons released by ionization of the ambient neutral gas and show how the ionization arises from electron trapping.

In laser-fusion experiments, the degree of preheating of the thermonuclear fuel and of the imploding shell that compresses the fuel is an important factor determining the efficiency of fuel compression. Results are presented here of fast-electron-, radiation-, and shock-wave-induced preheating in the laser-driven thermonuclear fusion experiments with high-aspect-ratio shells at the laser intensity ${10}^{14}$ W/${\mathrm{cm}}^2$. An alternative technique is proposed for the measurement of fast-electron-induced preheating. The technique is based on analysis of superthermal plasma x-ray radiation. The roles of radiation and shock-wave-induced preheating are also evaluated.

In hot plasmas the theoretically predicted photoabsorption cross sections rest upon the electronic states predicted by the plasma model. We compare the theoretical photoabsorption based on two models, the ion-sphere model and the ion-correlation model. We discuss the underlying physics and present calculations for iron, aluminum, and bromine plasmas, and also for a mixture of 14 elements under conditions occurring below the convective region of the sun.

Measurements of transient and steady-state light emission and of current transients from Ar at very high electric field to gas-density ratios E/n and low gas densities show that the excitation of the 811-nm lines of Ar is primarily by fast neutral Ar atoms. A drift tube with parallel-plane electrodes produces a spatially uniform electric field. Pulsed laser irradiation at 266 nm of the semitransparent cathode produces current pulses of ≊10 mA and ≊10 ns width. The time and spatial dependences of the line emission at 811±1 and 750±1 nm were measured for E/n in the range 280 townsend (Td) to 84 kTd at gas densities from 1.3×${10}^{23}$ to 9×${10}^{20}$ ${\mathrm{m}}^{\mathrm{-}3}$, where 1 Td=${10}^{\mathrm{-}21}$ V ${\mathrm{m}}^2$. At E/n<4 kTd the time dependence of the 811-nm emission is consistent with direct and cascade excitation due to electron collisions with Ar. At E/n>18 kTd the emission transient consists of an early peak caused by the initial electron avalanche followed by a delayed peak that moves toward the cathode with an apparent velocity of about twice that of the ions. Steady-state 811-nm emission is also observed in a region of very low electric field near the cathode where the ${\mathrm{Ar}}^{+}$ ions have insufficient energy for excitation and the fast atoms are injected from a high-field region. These observations of 811-nm emission, plus previous measurements of steady-state spatial distributions of emission, are interpreted as demonstrating that excitation of the 811-nm lines is primarily by fast Ar atoms formed by charge transfer in ${\mathrm{Ar}}^{+\mathrm{-}}$Ar collisions. For 750-nm emission the prompt portion is much larger and the delayed portion is much weaker than for 811-nm emission showing the importance of excitation by electrons even at these high E/n values. Steady-state observations of 811-nm emission when slow ${\mathrm{Ar}}^{+}$ ions are injected into a region of high electric field further confirm the model.

By adopting the modified-convolution approximation for the electron-electron correlation, the density-functional theory for the electron-ion correlation, and the hypernetted-chain approximation for the ion-ion correlation, an integral-equation formalism is developed for strongly coupled plasmas. With this scheme, the interparticle correlations are analyzed for hydrogen plasmas near the metal-insulator transition boundary. Several results of existing theories are reproduced with this approach. The boundaries of the phase diagram containing both the fully-to-partially ionized plasma transition and a metal-insulator transition are studied through investigation of the behaviors of the degree of ionization and the compressibility.