What Controls the Quality of Photodynamical Simulations? Electronic Structure Versus Nonadiabatic Algorithm

Abstract

The field of nonadiabatic dynamics has matured over the last decade with a range of algorithms and electronic structure methods available at the moment. While the community currently focuses more on developing and benchmarking new nonadiabatic dynamics algorithms, the underlying electronic structure controls the outcome of nonadiabatic simulations. Yet, the electronic-structure sensitivity analysis is typically neglected. In this work, we present a sensitivity analysis of the nonadiabatic dynamics of cyclopropanone to electronic structure methods and nonadiabatic dynamics algorithms. In particular, we compare wave function-based CASSCF, FOMO-CASCI, MS- and XMS-CASPT2, density-functional REKS, and semiempirical MRCI-OM3 electronic structure methods with the Landau–Zener surface hopping, fewest switches surface hopping, and ab initio multiple spawning with informed stochastic selection algorithms. The results clearly demonstrate that the electronic structure choice significantly influences the accuracy of nonadiabatic dynamics for cyclopropanone even when the potential energy surfaces exhibit qualitative and quantitative similarities. Thus, selecting the electronic structure solely on the basis of the mapping of potential energy surfaces can be misleading. Conversely, we observe no discernible differences in the performance of the nonadiabatic dynamics algorithms across the various methods. Based on the above results, we discuss the present-day practice in computational photodynamics.

Simulating Attochemistry: Which Dynamics Method to Use?

“…We analyzed the nuclear motion in the branching space of the nearby CI to investigate the charge-directed reactivity. Mixed quantum-classical methods can reproduce quantitatively the quantum nuclear motion induced by pure electronic adiabatic states (or an incoherent superposition of them) as in standard photochemistry, consistently with literature. − In this case, the resulting nuclear motion is mostly along the gradient difference coordinate, in the positive or negative direction, depending on the excited electronic state.…”

“…Recent works have evaluated the effect of the choice of the dynamics methods and algorithms on femtochemistry with initial excitation to a single electronic state with little observed difference (in the absence of quantum nuclear tunneling). − In the context of attochemistry with initial coherent excitation to a superposition of electronic states, mixed quantum-classical methods, like Ehrenfest and TSH, are commonly employed in the literature. − ,− However, the impact of the approximations underlying such dynamics methods on the simulations of attochemical reactions remains unknown. The goal of the present paper is to compare the suitability of several dynamics methods for charge-directed reactivity: DD-vMCG, TSH, and Ehrenfest (see Figure for a schematic representation).…”

Which Electronic Structure Method to Choose in Trajectory Surface Hopping Dynamics Simulations? Azomethane as a Case Study

“…Interestingly, the suitability of the different electronic structure methods could hardly be predicted from a static analysis of the potential energy surfaces (see SI for energetics at the Franck–Condon geometry and scan along the CNNC dihedral angle coordinate). This suggests that the common approach to benchmark and validate electronic structure methods before simulating dynamics may not be sufficient, as concluded in a recent work . Of course, the present conclusions were obtained in the specific case of the cis-to-trans photoisomerization reaction of azomethane upon S 1 excitation and thus cannot be assumed to be general for all molecules, all electronic states, nor all types of photoinduced processes.…”

“…A common approach in the literature to benchmark electronic structure methods is to compare energies and geometries of key structures or along specific cuts of the potential energy surfaces. , But, globally, the impact of the choice of the electronic structure methods on the nonadiabatic dynamics has been little studied, − since simulating and comparing dynamics with several electronic structure approaches is often computationally beyond reach. Very recently, the nonadiabatic dynamics of cyclopropanone was extensively studied with different electronic structure methods and nonadiabatic dynamics algorithms, showing a much larger sensitivity to the former than the latter . The goal of the present Letter is to benchmark the suitability for nonadiabatic dynamics simulations, in particular internal conversion, of a series of different electronic structure methods: (extended) multistate complete active space second-order perturbation theory (MS-CASTP2 − and XMS-CASPT2), state-averaged complete active space self-consistent field (SA-CASSCF), , second-order approximate coupled cluster singles and doubles (CC2), ADC(2), , configuration interaction singles (CIS), , and TD-DFT , with several functionalsPBE, PBE0, and CAM-B3LYP, as respective representatives of local, global hybrid, and range-separated hybrid modelswith and without TDA .…”

“…A common approach in the literature to benchmark electronic structure methods is to compare energies and geometries of key structures or along specific cuts of the potential energy surfaces. , But, globally, the impact of the choice of the electronic structure methods on the nonadiabatic dynamics has been little studied, − since simulating and comparing dynamics with several electronic structure approaches is often computationally beyond reach. Very recently, the nonadiabatic dynamics of cyclopropanone was extensively studied with different electronic structure methods and nonadiabatic dynamics algorithms, showing a much larger sensitivity to the former than the latter .…”