Chemical Assignment of Symmetry-Adapted Perturbation Theory Interaction Energy Components: The Functional-Group SAPT Partition

Abstract

Recently, we introduced an effective atom-pairwise partition of the many-body symmetry-adapted perturbation theory (SAPT) interaction energy decomposition, producing a method known as atomic SAPT (A-SAPT) [Parrish, R. M.; Sherrill, C. D. J. Chem. Phys. 2014, 141, 044115]. A-SAPT provides ab initio atom-pair potentials for force field development and also automatic visualizations of the spatial contributions of noncovalent interactions, but often has difficulty producing chemically useful partitions of the electrostatic energy, due to the buildup of oscillating partial charges on adjacent functional groups. In this work, we substitute chemical functional groups in place of atoms as the relevant local quasiparticles in the partition, resulting in a functional-group-pairwise partition denoted as functional-group SAPT (F-SAPT). F-SAPT assigns integral sets of local occupied electronic orbitals and protons to chemical functional groups and linking σ bonds. Link-bond contributions can be further assigned to chemical functional groups to simplify the analysis. This approach yields a SAPT partition between pairs of functional groups with integral charge (usually neutral), preventing oscillations in the electrostatic partition. F-SAPT qualitatively matches chemical intuition and the cut-and-cap fragmentation technique but additionally yields the quantitative many-body SAPT interaction energy. The conceptual simplicity, chemical utility, and computational efficiency of F-SAPT is demonstrated in the context of phenol dimer, proflavine(+)-DNA intercalation, and a cucurbituril host-guest inclusion complex.

Spatial assignment of symmetry adapted perturbation theory interaction energy components: The atomic SAPT partition

“…119 Such a partition would combine the detailed information of an order-2 partition with roughly the level of data compression of an order-1 atomic visualization, and would provide an important complementary alternative to order-1 atomic analyses. Moreover, we emphasize that any use of A-SAPT in such applications as force-field parametrizations will necessarily use the full order-2 atom-pair data: the order-1 techniques developed here are solely for the purposes of qualitative chemical understanding.…”

Communication: Practical intramolecular symmetry adapted perturbation theory via Hartree-Fock embedding

“…9 The former developed an effective two-body partition of SAPT, allowing one to parti- tion SAPT to the level of atom-pairwise intermolecular interactions. The latter modified A-SAPT to provide a chemically sensible partition to the level of functional-group pairwise intermolecular interaction.…”

“…This scheme of fragment identification is identical to that used in F-SAPT. 9 At this point, the particle assignment to the fragments is complete (and, given a list of the atoms in A, B, and C, can be completely automated by examination of the sum of orbital atomic charges). However, all three fragments in the system are fully interacting at a HF level; in particular, the wavefunctions of fragments A and B are interacting and orthogonal.…”

Visualizing dispersion interactions through the use of local orbital spaces

“…As previously mentioned, extensions have been proposed to SAPT to distinguish between contributions of different fragments within a molecule. In this section, we compare the local partitioning scheme to the functional‐group partition of SAPT (F‐SAPT) of Sherrill and coworkers . We look into the substituent effect on π ‐stacking interactions, using the same model systems as Parrish and Sherrill in a recent work .…”

“…The approach used within the paper is closely related to F‐SAPT and A‐SAPT methods, with the extension to intramolecular interactions. Therefore, a direct comparison between the two is warranted: The use of local correlation based analysis only allows for a decomposition of correlation energy terms, while the SAPT ansatz covers all types of interactions (including electrostatics, exchange, and so forth). However, the decomposition used in this work is a direct by‐product of a wave function calculation.…”