Calculation of interfacial properties via free-energy-based molecular simulation: The influence of system size

Abstract

We examine several issues related to the calculation of interfacial properties via analysis of an interface potential obtained from grand canonical Monte Carlo simulation. Two model systems are examined. One includes a monatomic Lennard-Jones fluid that interacts with a structureless substrate via a long-ranged substrate potential. The second model contains a monatomic Lennard-Jones fluid that interacts with an atomistically detailed substrate via a short-ranged potential. Our results are presented within the context of locating the wetting point. Two methods are used to compute the wetting temperature. In both cases we examine the system size dependence of the key property used to deduce the wetting temperature as well as the robustness of the scaling relationship employed to describe the evolution of this property with temperature near the wetting point. In the first approach we identify the wetting transition as the point at which the prewetting and bulk saturation curves meet. In this case, the prewetting saturation chemical potential is the key quantity of interest. In the second approach we find the point at which the spreading coefficient evaluates to zero. We find that the effect of system size is adequately described by simple scaling functions. Moreover, estimates of the wetting temperature for finite-sized systems characterized by a linear dimension greater than 12 fluid diameters differ by less than 1% from an otherwise equivalent macroscopic system. Modification of the details regarding the use of simulation data to compute the wetting temperature can also produce a shift in this quantity of up to 1%. As part of this study, we also examine techniques for describing the shape of the interface potential at a relatively high surface density. This analysis is particularly relevant for systems with long-ranged substrate potentials for which the interface potential approaches a limiting value asymptotically.

Surface phase transitions of multiple-site associating fluids

“…Recently, Errington and co-workers studied various aspects of the prewetting transition for the same system using the GC-TMMC method [43][44][45]. We employed the same methodology to evaluate the prewetting transition for associating fluids, and our results are in agreement with those of Errington and co-workers for the non-associating system.…”

“…For a non-associating fluid, using the aforementioned scaling relationship for the temperature range T ¼ 0.65-0.725, we found T w ¼ 0.5836 (2), which is in good agreement with the value reported by Sellers and Errington [44]. However, recently, Grzelak and Errington [45] found that the locus of the prewetting line in the vicinity of the bulk saturation line is sensitive to the system size and hence can affect the estimation of the wetting temperature. In order to evaluate the system size effect in the current work, we considered a one-site model on a smooth surface as an example with " af ¼ 0 and 4.…”

Boundary of prewetting transition of Ar on a Li surface

“…The small discrepancy may be due to the influence of system size effects, which has been discussed by Grzelak and Errington recently. 44 In contrast, the DFT study by Sartarelli and Szybisz 20 obtained a higher wetting temperature with T w = 110 K. The quantitative difference can be attributed to the different treatment of attractive interactions. They used a mean-field approximation while we used an accurate direct correlation function 42 to account for the inhomogeneity.…”

A simple grand canonical approach to compute the vapor pressure of bulk and finite size systems

“…25, 26 Errington has recently published a series of papers summarizing many of these contributions to the calculation of interfacial and coexistence phenomena in different molecular systems. [27][28][29] However, most of these methods cannot be adapted in a simple way to capture the vapor pressure of an interface of arbitrary shape or size. In particular, the vapor pressure of nanodroplets or molecular aggregates is very difficult to access experimentally, though it is of a fundamental interest in areas like classical nucleation theory.…”