Simulation-ready graphene oxide structures with hierarchical complexity: a modular tiling strategy

Abstract Graphene oxide (GO) sheet structures are highly variable and depend on preparation conditions. The use of molecular simulation is a complementary strategy to explore how this complexity influences the ion transport properties of GO membranes. However, despite recent advances, computational models of GO typically lack the required complexity as suggested by experiment. The labor required to create such an ensemble of such structural models with the required complexity is impractical without recourse to automated approaches, but no such code currently can meet this challenge. Here, a modular tiling concept is introduced, along with the HierGO suite of code; an automated approach to producing highly complex hierarchically-structured models of GO with a high degree of control in terms of holes and topological defects, and oxygen-group placement, that can produce simulation-ready input files. The benefits of the code are exemplified by modeling and contrasting the properties of three types of GO membrane stack; the widely-modeled Lerf–Klinowski structure, and two types of highly heterogeneous GO sheet reflecting differing processing conditions. The outcomes of this work clearly demonstrate how the introduction of the complexity modeled here leads to new insights into the structure/property relationships of GO with respect to permeation pathways of water, ions and molecular agents that are inaccessible using previously-considered models.