Removing the Deadwood from DFT/MRCI Wave Functions: The p-DFT/MRCI Method

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Theoretical Chemistry

Background:

• The combined density functional theory and multireference configuration interaction (DFT/MRCI) method is crucial for studying the excited electronic states of large molecules.
• A significant challenge in DFT/MRCI calculations is the presence of numerous superfluous configurations within the wave function, increasing computational cost.

Purpose of the Study:

• To develop an efficient method for reducing the computational burden of DFT/MRCI calculations.
• To improve the performance of DFT/MRCI by removing unnecessary configurations without sacrificing accuracy.

Main Methods:

• A configuration pruning algorithm based on second-order Epstein-Nesbet perturbation theory was developed.
• This algorithm effectively identifies and removes superfluous configurations from the DFT/MRCI wave function.
• The performance of the pruned DFT/MRCI (p-DFT/MRCI) method was evaluated.

Main Results:

• The p-DFT/MRCI method demonstrated orders of magnitude savings in computational timings compared to the standard DFT/MRCI approach.
• The accuracy of the excited electronic state calculations was preserved using the pruned method.
• The efficiency of calculating excited states for large molecules was substantially enhanced.

Conclusions:

• The developed configuration pruning algorithm offers a computationally efficient alternative for DFT/MRCI calculations.
• The p-DFT/MRCI method provides a practical solution for studying excited states of large molecular systems.
• This approach significantly reduces computational resources required for accurate electronic structure calculations.