Predicting Intersystem Crossing Rates with AIMS-DFT Molecular Dynamics

Accurate prediction of the intersystem crossing rates is important for many different applications in chemistry, physics, and biology. Recently, we implemented the ab initio multiple spawning (AIMS) molecular dynamics method to describe the intersystem crossing processes, where nonradiative transitions between electronic states with different spin multiplicities are mediated by spin orbit coupling. Our original implementation of the direct AIMS dynamics used the complete active space self-consistent field (CASSCF) method to describe multiple coupled electronic states on which multidimensional Gaussian wave packets were propagated. In this work, we improve the computational efficiency and versatility of the AIMS dynamics by interfacing it with the density functional theory (DFT). The new AIMS-DFT and the earlier AIMS-CASSCF implementations are used to investigate the effects of electronic structure methods on the predicted intersystem crossing rate constants and the lowest triplet state lifetime in the GeH2 molecule. We also compare the rates and lifetimes obtained from the AIMS simulations with those predicted by the statistical nonadiabatic transition state theory (NA-TST). In NA-TST, the probabilities of spin transitions are calculated using the Landau-Zener, weak coupling, and Zhu-Nakamura formulas. Convergence of the AIMS rate constants with respect to the simulation time and the number of initial trajectories (Gaussian wave packets) is analyzed. An excellent agreement between AIMS-DFT and AIMS-CASSCF can be explained by cancelation of two effects: higher energy barriers and a stronger spin orbit coupling in DFT relative to CASSCF. The rate constants obtained with the AIMS-DFT dynamics are about a factor of 2 larger than those predicted by the statistical NA-TST. This is likely due to the importance of the nonlocal interstate transitions missing from the NA-TST description.