Leveraging Metaheuristics to Uncover Water Confinement in Multilayer Graphynes

Abstract

Global optimization is an effective approach to study the geometries and energetics of atomic or molecular confinement within nanostructures. The high computational cost associated with modeling such complex chemical systems calls for the adoption of stochastic global optimization techniques. Herein, we employ a swarm intelligence-based technique, namely, particle swarm optimization (PSO), to study the confinement of water clusters in monolayer and multilayer graphynes (GYs), including γ-GY-2, γ-GY-3, and γ-GY-4. The water molecules are described using the TIP4P model. The non-electrostatic part of GY-water and GY-GY interactions are modeled using the optimally fitted improved Lennard-Jones potential and the anisotropic Hod’s interlayer potential, while the Coulombic potential is employed to account for the electrostatic interactions between GYs and water. Our PSO results reveal that the pore size of GYs is vital to the confinement of water clusters in multilayer γ-GYs. The γ-GY-2 multilayer tends to accommodate water as a monolayer between its two layers for large cluster sizes. A single-file confinement of water molecules is observed when water clusters were confined within the γ-GY-3 trilayer. In contrast, γ-GY-4, with the largest pore size, allowed clustering of water molecules within the triangular channels. Our findings established the importance of incorporating the twist features of GYs in the modeling formulation, as well as the accurate description of empirical formulations that can enable large-scale simulations. Our findings hold promise for extended research on water transport through twisted multilayer GYs.