▎ 摘　　要

In a recent study [Angew. Chem. Int. Ed.,2014, 53, 9957-9960] a new concept of chiral separation has been suggested, which is based on functionalized, nanoporous sheets of graphene. In this follow-up article we discuss the underlying principle in greater detail and make suggestions for suitable pore functionalizations with respect to a selection of chiral prototype molecules. Considering drug molecules as future targets for a chiral separation via membranes, the necessary pore sizes represent a big challenge for standard methods of computational chemistry. Therefore, we test two common force fields (GAFF, CGenFF) as well as a semiempirical tight-binding approach recently developed by the Grimme group (GFN-xTB) against the computationally much more expensive density functional theory. We identify the GFN-xTB method as the most suitable approach for future simulations of functionalized pores for the given purpose, as it is able to produce reaction pathways in very good agreement with density functional theory, even in cases where force fields tend to an extreme overestimation of barrier heights.