Optimizing conical intersections by spin-flip density functional theory: application to ethylene.

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Theoretical Chemistry

Background:

• Conical intersections (CIs) are crucial for understanding photochemical reactions.
• Previous methods like linear response time-dependent density functional theory (LR-TDDFT) struggled to accurately describe the geometry of CIs in twisted ethylene.
• Accurate characterization of CIs is essential for predicting reaction pathways and dynamics.

Purpose of the Study:

• To accurately determine conical intersections (CIs) of ethylene using a novel computational approach.
• To present and analyze the specific structures of three key CIs: twisted-pyramidalized, hydrogen-migrated, and ethylidene.
• To validate the findings against high-level ab initio methods.

Main Methods:

• Utilizing spin-flip density functional theory (SFDFT) for electronic structure calculations.
• Employing a penalty-constrained optimization method to locate and optimize CI structures.
• Comparing results with multireference configuration interaction (MR-CI) and multistate second-order multireference perturbation theory (MS-CASPT2).

Main Results:

• SFDFT successfully optimized three distinct conical intersection structures in ethylene.
• SFDFT identified a pyramidalized structure as the S(1) global minimum, unlike LR-TDDFT.
• Calculated energies and geometries show excellent agreement with established high-level computational methods (MR-CI and MS-CASPT2).

Conclusions:

• Spin-flip density functional theory (SFDFT) provides an accurate and reliable method for optimizing conical intersections in twisted ethylene.
• This study presents the first successful DFT optimization of CI points for twisted ethylene, including pyramidalized structures.
• The findings validate SFDFT as a powerful tool for studying photochemical processes involving conical intersections.