Computational study of molecular properties with dual basis sets

Area of Science:

• Computational chemistry
• Quantum chemistry
• Theoretical chemistry

Background:

• Accurate calculation of molecular properties is crucial in chemistry.
• Second-order Møller-Plesset perturbation theory (MP2) is a widely used method for post-Hartree-Fock calculations.
• Basis set size significantly impacts computational cost.

Purpose of the Study:

• To evaluate the performance of dual basis (DB) sets for molecular property calculations using MP2.
• To investigate the impact of virtual subspace truncation on computational efficiency and accuracy.
• To assess the accuracy of DB sets compared to full basis calculations.

Main Methods:

• Utilized dual basis (DB) sets for Hartree-Fock (HF) and MP2 calculations.
• Implemented systematic truncation of the virtual subspace.
• Calculated total energies and static electrical response properties (dipole moments, polarizabilities).
• Used aug-cc-pV5Z as the parent basis set.

Main Results:

• DB sets with full virtual space showed excellent agreement with full basis calculations.
• Mean absolute errors for DB-HF and DB-MP2 total energies were on the order of 10^-5 au.
• Mean absolute relative errors for electrical response properties were generally below 1%.
• DB coupled with virtual space truncation at the MP2 level resulted in errors within 2% of full basis values.

Conclusions:

• Dual basis sets offer a computationally efficient approach for MP2 calculations.
• The DB method provides accurate molecular properties with reduced computational overhead.
• Virtual space truncation can further enhance efficiency, albeit with slightly increased errors.