Analytic gradients for the single-reference driven similarity renormalization group second-order perturbation theory.

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Method Development

Background:

• Accurate calculation of molecular geometries is crucial for understanding chemical properties and reactions.
• Existing methods like Møller-Plesset perturbation theory (MP2) have limitations in scaling and accuracy for certain systems.
• The driven similarity renormalization group (DSRG) method offers a potential avenue for improved accuracy in electronic structure calculations.

Purpose of the Study:

• To derive and implement analytic energy gradients for the single-reference driven similarity renormalization group second-order perturbation theory (DSRG-PT2).
• To assess the computational efficiency and accuracy of the DSRG-PT2 method for geometry optimization.
• To compare the performance of DSRG-PT2 with established methods like MP2 and coupled cluster theory.

Main Methods:

• Derivation of analytic energy gradients for DSRG-PT2.
• Implementation of the DSRG-PT2 gradient calculations in a computational chemistry framework.
• Geometry optimization of 15 small molecules using the developed DSRG-PT2 method.

Main Results:

• The derived DSRG-PT2 equations exhibit favorable asymptotic scaling, similar to MP2.
• Geometry optimizations using DSRG-PT2 yielded equilibrium bond lengths comparable to MP2.
• The DSRG-PT2 method demonstrated a mean absolute error of 0.0033 Å and a standard deviation of 0.0045 Å against coupled cluster with singles, doubles, and perturbative triples (CCSD(T)).

Conclusions:

• The developed analytic energy gradients for DSRG-PT2 are computationally viable and accurate for molecular geometry optimization.
• DSRG-PT2 presents a promising alternative to MP2, offering similar computational efficiency with potentially improved accuracy.
• The findings support the broader applicability of DSRG methods in quantum chemistry for electronic structure and property predictions.