Basis Set Incompleteness Errors in Fixed-Node Diffusion Monte Carlo Calculations on Noncovalent Interactions

Area of Science:

• Quantum Chemistry
• Computational Physics

Background:

• Basis set incompleteness error (BSIE) is a known issue in quantum chemistry.
• Fixed-node Diffusion Monte Carlo (FN-DMC) is often assumed to be less sensitive to BSIE.
• The impact of BSIE on FN-DMC binding energy calculations is not well-established.

Purpose of the Study:

• To systematically investigate BSIE in FN-DMC binding energy calculations.
• To evaluate the performance of different basis sets for FN-DMC.
• To identify strategies for mitigating BSIE in FN-DMC.

Main Methods:

• Utilized the A24 data set of 24 noncovalently bound dimers.
• Performed FN-DMC calculations with various basis sets (cc-pVDZ, cc-pVTZ, aug-cc-pVTZ).
• Assessed the impact of diffuse functions and counterpoise correction on BSIE.

Main Results:

• BSIE can be significant in FN-DMC binding energy calculations, especially with small basis sets like cc-pVDZ and cc-pVTZ.
• The aug-cc-pVTZ basis set offers a good balance between accuracy and computational cost.
• Augmenting basis sets with diffuse functions and using counterpoise correction effectively reduces BSIE.

Conclusions:

• The assumption of minimal BSIE in FN-DMC is not universally valid for binding energies.
• Augmented basis sets, diffuse functions, and counterpoise correction are crucial for accurate FN-DMC binding energy calculations.
• Strategies exist to mitigate BSIE, enabling the use of smaller basis sets like aug-cc-pVDZ.