Analytic gradients for density cumulant functional theory: the DCFT-06 model

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Theoretical Chemistry

Background:

• Electron correlation methods are crucial for accurate molecular property prediction.
• Wavefunction-based methods can be computationally expensive.
• Density cumulant functional theory (DCFT) offers an alternative, size-extensive approach derived from reduced density matrices.

Purpose of the Study:

• To derive and implement analytic gradients for the DCFT-06 variant.
• To assess the accuracy and computational efficiency of DCFT-06 gradients.
• To benchmark DCFT-06 against established methods like coupled cluster theory.

Main Methods:

• Derivation of analytic gradient expressions for DCFT-06.
• Solution of coupled, perturbation-independent orbital and cumulant response equations.
• Iterative solution with rapid convergence for response equations.
• Implementation and benchmarking against CCSD, CCSD(T), and experimental data.

Main Results:

• DCFT-06 analytic gradients were successfully derived and implemented.
• The method requires solving coupled orbital and cumulant response equations iteratively.
• Benchmarking showed DCFT-06 results closer to CCSD(T) and empirical data than CCSD for most cases.
• The computational cost of the gradient calculation is significantly lower than expected.

Conclusions:

• DCFT-06 provides an accurate and computationally efficient method for calculating analytic gradients.
• The developed gradients show promise for future applications in computational chemistry.
• DCFT represents a viable and advancing alternative to traditional wavefunction-based methods.