Analytical Gradient Theory for Resolvent-Fitted Second-Order Extended Multiconfiguration Perturbation Theory...

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Extended multiconfiguration quasidegenerate perturbation theory (XMCQDPT2) is crucial for studying complex electronic systems.
• Accurate optimization of molecular configurations and evaluation of nuclear gradients are essential for understanding chemical processes.

Purpose of the Study:

• To formulate and implement an analytical gradient algorithm for XMCQDPT2 using Granovsky's resolvent-fitting approximation.
• To demonstrate the algorithm's power and accuracy for optimizing molecular geometries and analyzing electronic states.

Main Methods:

• Development of an analytical gradient algorithm for XMCQDPT2 with resolvent-fitting.
• Comparison of resolvent-fitting approximations with canonical XMCQDPT2.
• Application to frequency analyses and optimization of minimum energy conical intersection geometries.
• Parallelization using OpenMP/MPI hybrid approach.

Main Results:

• The resolvent-fitting approximation shows comparable accuracy to canonical XMCQDPT2.
• The algorithm effectively optimizes molecular configurations and evaluates nuclear gradients for systems with many electronic states.
• The parallelized program demonstrates efficient computational performance.

Conclusions:

• The developed analytical gradient algorithm for XMCQDPT2 with resolvent-fitting is a powerful tool for computational chemistry.
• This method facilitates the study of complex molecular systems, including reaction pathways and excited states.
• The parallel implementation enhances computational efficiency for large-scale calculations.