Showing papers by "Robert Evans published in 2012"

Phase behavior and structure of a fluid confined between competing (solvophobic and solvophilic) walls.

Abstract: We consider a model fluid with long-range r(-6) (dispersion) interparticle potentials confined between competing parallel walls. One wall is solvophilic and would be completely wet at bulk liquid-gas coexistence μ(co)(-), whereas the other is solvophobic and would be completely dry at μ=μ(co)(+). When the wall separation L is large and the system is below the bulk critical temperature T(C) and close to bulk liquid-gas coexistence, a delocalized interface or soft-mode phase forms with a liquid-gas interface near the center of the slit; this interacts with the walls via the power-law tails of the interparticle potentials. We use a coarse-grained effective Hamiltonian approach to derive explicit scaling expressions for the Gibbs adsorption Γ, the surface tension γ, the solvation force f(s), and the total susceptibility χ. These quantities depend on the dimensionless scaling variable (L/σ)(3)βδμ, where β=(k(B)T)(-1), σ is the diameter of the fluid particles and δμ=μ-μ(co) is the chemical potential deviation from bulk coexistence. Using a nonlocal density functional theory, we calculate density profiles for the asymmetrically confined fluid at different chemical potentials and for sufficiently large L confirm the scaling predictions for the four thermodynamic quantities. Since the upper critical dimension for complete wetting with power-law potentials is less than 3, we argue that our (mean-field) scaling predictions should remain valid in treatments that incorporate the effects of interfacial fluctuations. As the wall separation L is decreased at μ(co), we predict a capillary evaporation transition from the delocalized interface phase to a dilute gas state with just a thin adsorbed film of liquidlike density next to the solvophilic wall. This transition is closely connected to the first-order prewetting transition that occurs at the solvophilic wall in the semi-infinite system. We compare the phase diagram for the competing walls system with the phase diagrams for the fluid confined between identical solvophilic and identical solvophobic walls. Comparisons are also made with earlier studies of asymmetric confinement for systems with short-range forces.

Relationship between Local Molecular Field Theory and Density Functional Theory for non-uniform liquids

Abstract: The Local Molecular Field Theory (LMF) developed by Weeks and co-workers has proved successful for treating the structure and thermodynamics of a variety of non-uniform liquids. By reformulating LMF in terms of one-body direct correlation functions we recast the theory in the framework of classical Density Functional Theory (DFT). We show that the general LMF equation for the effective reference potential phi_R follows directly from the standard mean-field DFT treatment of attractive interatomic forces. Using an accurate (Fundamental Measures) DFT for the non-uniform hard-sphere reference fluid we determine phi_R for a hard-core Yukawa liquid adsorbed at a planar hard wall. In the approach to bulk liquid-gas coexistence we find the effective potentials exhibit rich structure that can include damped oscillations at large distances from the wall as well as the repulsive hump near the wall required to generate the low density 'gas' layer characteristic of complete drying. We argue that it would be difficult to obtain the same level of detail from other (non DFT based) implementations of LMF. LMF emphasizes the importance of making an intelligent division of the interatomic pair potential of the full system into a reference part and a remainder that can be treated in mean-field approximation. We investigate different divisions for an exactly solvable one- dimensional model where the pair potential has a hard-core plus a linear attractive tail. Results for the structure factor and the equation of state of the uniform fluid show that including a significant portion of the attraction in the reference system can be much more accurate than treating the full attractive tail in mean-field approximation. We discuss further aspects of the relationship between LMF and DFT.