An efficient method by combining different basis sets and SAPT levels.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Molecular Interactions

Background:

• Accurate calculation of non-covalent interactions (NCI) for large molecular systems (>50 atoms) using symmetry-adapted perturbation theory (SAPT) is computationally expensive with large basis sets.
• Smaller basis sets often lead to inaccurate predictions of dispersion forces and total interaction energies in NCI calculations.

Purpose of the Study:

• To develop computationally efficient yet accurate methods for calculating NCIs in large molecular complexes.
• To improve the prediction of dispersion energies and overall interaction energies compared to standard SAPT approaches.

Main Methods:

• Proposed two composite methods combining SAPT2 calculations with different basis sets (aug-cc-pVDZ and aug-cc-pVTZ) as baselines.
• Incorporated dispersion term corrections at the SAPT2+ level using bond functions and smaller basis sets with MP2 corrections.

Main Results:

• Evaluated methods on benchmark NCI datasets (S22, S66).
• Demonstrated significant improvements in accuracy compared to the original SAPT Silver standard.
• Achieved accuracy comparable to the SAPT Gold standard in certain cases.

Conclusions:

• The proposed composite methods offer a substantial improvement in the accuracy of NCI calculations for large systems.
• These methods provide a computationally feasible alternative to traditional approaches, maintaining high accuracy with reduced cost.