Understanding the desalination mechanism of a two-dimensional graphene-like membrane using data-driven design

Abstract

Global freshwater resources are not inexhaustible, and the deterioration in water quality and the shortage of drinkable water are serious challenges for human beings. To date, the development of two-dimensional (2D) nanomaterials provides an essential guarantee for seawater desalination technology. However, 2D nanomaterials are still lacking in desalination efficiency. Herein, we used a data-driven design to analyze the desalination mechanism of a 2D graphene-like membrane. A 2D graphene-like desalination membrane (Desa-C) with a multiporous structure has been screened first through machine learning (ML). Then, the desalination efficiency and other comprehensive properties of mechanics, adsorption, diffusion, and electron are analyzed based on molecular dynamics simulations and first-principles density functional theory. The results indicate that the low free energy barrier of water molecules, abundant salt ions adsorption sites, and periodically distributed 0.53 nm nanopores endow Desa-C with ideal permeability and selectivity; In particular, the salt ions rejection efficiency could constantly be ensured at 100 % within 125 MPa pressure, and this value holds 97 % even if the pressure increases to 145 MPa. Moreover, the Desa-C membrane featuring metallicity could achieve a self-cleaning effect with reverse voltage conditions, which prolongs the lifetime of the desalination membrane. These findings have potential application values in desalination technology and provide a positive theoretical contribution to the innovation of desalination membranes.