Indexed on: 14 Mar '16Published on: 27 Nov '15Published in: Polyhedron
We evaluate the numerical accuracy of the potential energy surface (PES) that describes the interaction between Buckminster fullerene C60 and a range of graphene flakes of increasing size as computed by the M06-2X, ωB97-xD and B2PLYP-D3BJ density functionals, all of which have been suggested for accurate descriptions of systems that are dominated by van-der-Waals interactions that mainly originate from π–π stacking. We apply a methodology that has been recently established by us as being of good quality to describe the weak non-bonding interactions of buckyball–buckybowl complexes at reasonable computational cost. In the first step we use five different coordination modes of C60 on coronene to establish the minimum energy configuration the buckyball will adopt once adsorbed on the graphene flake. We then investigate the PES with respect to dissociation of the intermolecular complex and the rotation of the buckyball with respect to the surface of the graphene flake. Finally, we attempt to extrapolate our results to the Buckminster fullerene on graphene limit by applying the same methodology to larger and larger analogues of circumcoronene. Our results show that while the chosen density functional methods themselves describe the system well, care must be taken to ensure that the accuracy of the numerical integration (i.e. the grid size) is high enough to preserve the smallest energetic features (≈0.1 kcal/mol) of the PES. This is of particular importance when the rotation of the buckyball with respect to a graphene surface is described.