A State-Averaged Formulation for Variational Multiconfigurational Pair-Density Functional Theory.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Accurate prediction of excited states is crucial but challenging for systems with static and dynamic correlation.
• Existing methods often struggle with complex electronic structures, necessitating new approaches.

Purpose of the Study:

• To develop and assess a novel state-averaged formulation for variational multiconfiguration pair-density functional theory (MC-PDFT).
• To evaluate the performance of MC-PDFT for calculating excitation energies in open-shell radicals and closed-shell organic molecules.

Main Methods:

• Implementation of a state-averaged MC-PDFT formulation.
• Comparison of translated pure, global hybrid, and range-separated PDFT functionals against spin-TD-DFT and wave function methods.
• Assessment on diverse sets of open-shell radicals and closed-shell organic molecules.

Main Results:

• MC-PDFT with global hybrid functionals significantly improves results for small open-shell radicals compared to standard DFT and EOM-CCSD.
• Range-separated MC-PDFT achieves mean absolute errors as low as 0.06 eV for excitation energies.
• Translated pure GGAs struggle with π → π* transitions, but range separation corrects these deviations effectively.

Conclusions:

• The developed MC-PDFT offers a robust and computationally efficient strategy for modeling complex chemical systems.
• Range-separated functionals, like sr-ctBLYP, provide a balanced description of different electronic transitions and are competitive with high-level wave function methods.