Automated Search for Minimum Energy Conical Intersection Geometries between the Lowest Two Singlet States S0/S1-MECIs by the Spin-Flip TDDFT Method

Automated search for minimum energy conical intersection geometries between the lowest two singlet states (S-0/S-1-MECIs) was performed by combining the anharmonic downward distortion following (ADDF) method, the seam model function (SMF) approach, and the spin-flip (SF) TDDFT method. SMF/ADDF has been employed previously in automated searches for MECIs on potential energy surfaces (PESs) with expensive multireference methods. In this work, we adopt the SF-TDDFT method that enables efficient optimization of S-0/S-1-MECIs in the framework of TDDFT. To evaluate the performance of the present approach, it was applied to ethylene and 1,3-butadiene. The present method automatically gave unknown S-0/S-1-MECIs as well as all previously reported ones. Therefore, the present hybrid method of SMF/ADDF and SF-TDDFT is shown to be a promising approach to locate S-0/S-1-MECIs of large systems automatically with modest computational costs.