Advanced atomistic models to accurately simulate graphitic nanostructures, bio-interfaces, and aromatic molecules

Graphene and carbon nanomaterials are used in biosensors, batteries, renewable composites, and functional materials. We introduce new atomistic models for the simulation of graphitic nanostructures, bio-interfaces, and aromatic molecules using common energy expressions (CHARMM, AMBER, CVFF, OPLS-AA, PCFF). The model contain virtual π-electrons in graphene/benzene ring that significantly increase the reliability of computed properties. Benchmarking of the new model shows improvements in the reproduction of cation-pi interactions, pi-pi stacking, and electrolyte interfacial properties by ~10 fold. Two dummy electrons are attached to carbon atom to add the multipoles due to π-electrons in the aromatic rings. This leads to columbic interaction in accordance with experiment, which is missing in the old models. The models can accurately predict binding free energies, adsorption sites and biomolecular recognition, binding orientations, surface diffusion, and competitive adsorption of molecules. Details of the validation include agreement with experimental lattice parameters (<0.5% deviation), surface energy (<5% deviation), elastic constants (<15% deviation), hydration energy (10% deviation), water contact angles (<10% deviation) and cation-pi interaction (<15% deviation).