Improved generalized energy-based fragmentation approach and its applications to the binding energies of supramolecular complexes

Accurate prediction for the binding energies of host-guest complexes is important in the field of supramolecular chemistry. In this article, we have implemented an improved generalized energy-based fragmentation (GEBF) approach for computing the binding energies of supramolecular complexes. In this approach, a new scheme is proposed to generate primitive GEBF subsystems with reasonable size. With this scheme, those subsystems in the complex and the corresponding host molecule are very similar (the guest does not exist in any subsystem of host). Our results for ten supramolecular complexes show that the improved GEBF approach could accurately reproduce the conventional density functional theory (DFT) binding energies in those complexes. The GEBF binding energy errors relative to the corresponding conventional values are usually less than 1 kcal mol(-1). Furthermore, for each pair of complexes with the same or similar hosts, the GEBF relative binding energy difference between the two complexes with different guests is even below 1 kJ mol(-1). In addition, the GEBF approach could predict the second-order Moller-Plesset perturbation theory (MP2) energy of a complex with 3382 basis functions within only 3.3 h by massive parallelization using 312 processors on 13 computer nodes. Therefore, the improved GEBF approach could be a practical tool for predicting the binding energies or relative binding energies of supramolecular complexes.