Machine learning coarse-grained models of dissolutive wetting: a droplet on soluble surfaces

Abstract

Based on neural networks, coarse-grained glucose–water dissolution pair models with all-atomic system precision are developed, which can be used to explore the dynamics of dissolutive wetting on larger scales.

Shape optimization of a meniscus-adherent nanotip

“…The solid–liquid interaction ε SL is fixed to make the solid hydrophilic with the liquid (the equilibrium contact angle is approximately 70°). 38 Simulations for various nanotip widths and dissolubility ( w = 59.16, 79.56, 99.96, 120.36, and 140.76 Å, ε = 0.32, 0.4, 0.48, 0.56, 0.64, and 0.72) were conducted; 39 w and ε = ε SS / ε SL represent the nanotip width and the ratio of solid–solid and solid–liquid interactions, respectively. The dissolubility of the solid is controlled by ε : a smaller ε generates a higher dissolubility.…”