“…Indeed, the typical distance between two nearest-neighbor C 8 H 8 molecules in fullerene-cubane is about 10 Å, well beyond the optimal 6.5 Å intermolecular distance. 43 We therefore considered the parameters bb , bb , and as tunable, for our purposes, rather than to be taken from a fit to cubane-cubane interactions. A priori requirements for our potential constants are reproducing ͑i͒ a fcc phase with a lattice constant close to the experimental value a = 14.74 Å at room temperature, ͑ii͒ a structural phase transition at about T Ϸ 140 K, and ͑iii͒ a transition from freely rotating to frozen fullerene molecules at about the same temperature as the structural phase transition.…”