Energy and analytic gradients for the orbital-optimized coupled-cluster doubles method with the density-fitting...

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Conventional orbital-optimized coupled-cluster doubles (OCCD) methods face computational limitations for large systems due to expensive integral transformations.
• Density-fitting (DF) and Cholesky decomposition (CD) approaches offer potential for reducing the computational burden of coupled-cluster methods.

Purpose of the Study:

• To present efficient implementations of the density-fitting orbital-optimized coupled-cluster doubles (DF-OCCD) method and its analytic energy gradients.
• To compare the computational cost and performance of DF-OCCD with conventional OCCD and other related methods.

Main Methods:

• Implementation of DF-OCCD and its analytic gradients.
• Implementation of a Cholesky-decomposed variant (CD-OCCD) for energy computations.
• Comparison of computational costs using the Q-CHEM and PSI4 packages for various molecular systems and basis sets.

Main Results:

• DF-OCCD demonstrates dramatically lower computational costs compared to conventional OCCD, with significant reductions in computational time (e.g., eightfold for C6H14).
• DF-OCCD shows substantial improvements over DF-CCD for open-shell geometries, interaction energies, and hydrogen transfer reactions.
• The method provides accurate harmonic vibrational frequencies, especially in symmetry-breaking scenarios, and offers advantages over CCSD in several aspects.

Conclusions:

• DF-OCCD provides a computationally efficient and accurate approach for electronic structure calculations.
• The method's advantages, including no need for orbital relaxation in gradients and compatibility with active spaces, make it highly attractive for various chemical applications.
• DF-OCCD and its variants offer superior performance and accuracy, particularly for challenging problems like symmetry breaking and near-degeneracy regions.