Comprehensive Basis-Set Testing of Extended Symmetry Adapted Perturbation Theory

Hybrid or "extended" symmetry-adapted perturbation theory (XSAPT) replaces traditional SAPT's treatment of dispersion with better-performing alternatives, while at the same time extending two-body (dimer) SAPT to a many-body treatment of polarization, via a self-consistent charge-embedding procedure. The present work presents a systematic study of how total interaction energies and also energy components converge with respect to the choice of Gaussian basis set in XSAPT. Pople-style basis sets, while very efficient, consistently afford errors > 1 kcal/mol with respect to benchmark interaction energies for standard data sets of noncovalent complexes, whereas Dunning's correlation-consistent basis sets and Karlsruhe basis sets perform much better. Hybrid treatments of dispersion afford must faster basis-set convergence as compared to traditional SAPT, and benchmark-quality results can be obtained using basis sets of triple-zeta quality, although the use of diffuse functions proves to be essential. The use of dimer Hartree-Fock calculations as a correction for higher-order induction can be performed in even smaller basis sets with negligible error, leading to a composite approach that offers speedups of 100x even for small systems.