Determine accuracy and practicality of SS-CASSCF for photochemical systems

Determine the accuracy and practicality of state-specific complete active space self-consistent field (SS-CASSCF) theory for predicting electronically excited energy surfaces in chemical systems of photochemical interest, assessing whether SS-CASSCF can reliably deliver continuous and physically meaningful potential energy surfaces across relevant nuclear coordinates.

Background

State-specific CASSCF has been proposed as a route to predict excited-state energy surfaces with smaller active spaces and bespoke orbital relaxation, addressing limitations of state-averaged CASSCF. While the present work evaluates SS-CASSCF on ethylene torsion, the broader question of its overall accuracy and practicality across photochemical systems remains unsettled.

The paper demonstrates both promising aspects (e.g., improved diffusivity of the valence state and comparable excitation energies to larger SA-CASSCF active spaces) and pitfalls (e.g., solution coalescence and disappearance due to unbalanced dynamic correlation), underscoring the need for a systematic determination of SS-CASSCF’s reliability in general photochemical applications.