SHNITSEL - Surface Hopping Nested Instances Training Set for Excited-state Learning

SHNITSEL-dynamic K. Höllring, T. E. Röhrkasten & C. Müller The Surface Hopping Nested Instances Training Set for Excited-State Learning (SHNITSEL) is a comprehensive data repository designed to support the development and benchmarking of excited-state dynamics methods. SHNITSEL-dynamic contains datasets with comprehensive surface hopping trajectory data of five organic molecules: Alkenes: ethene (A01), propene (A02), 2-butene (A03) Ring structures: 1,3-cyclohexadiene (R02) The data are stored in xarray format using xarray.DataTree objects for efficient handling of multidimensional trajectory ensembles. Metadata such as units, electronic-structure method, charge, and number of electronic states are stored as attributes of the individual datasets within the tree. The datasets contain key electronic-structure quantities for singlet and triplet states, including energies, nuclear forces, dipole moments, transition dipole moments, nonadiabatic couplings, and spin-orbit couplings, computed at the multireference ab initio level. Two data representations are provided: Stacked (#271,700 data points in total): for each molecule, the DataTree contains a single dataset in which all trajectories are stacked along a trajectory dimension. Unstacked (#271,700 data points in total): for each molecule, the DataTree contains one dataset per trajectory, stored as individual leaves. These complementary formats allow users to choose between trajectory-resolved and ensemble-level data representations depending on their analysis needs. ---------------------------------------- Additional Datasets for I01 (CH2NH2+) S. Mai & J. C. B. Dietschreit a) About Two additional datasets for compound I01 are provided, aiming to sample the free energy surface around the S1/S0 conical intersection (I01_dietschreit_sa2-cas22_196_S1traj.nc) and sample the photoinduced dynamics (I01_dietschreit_sa2-cas22_500_fssh.nc). b) Computational Details Dynamics: NAMD method: SHARC4 surface hopping was turned off and a Langevin thermostat (1000K) was activated to sample the free energy surface around the S1/S0 conical intersection (I01_dietschreit_sa2-cas22_196_S1traj.nc) surface hopping (gradients for all states, NACs computed with nacdr, I01_dietschreit_sa2-cas22_500_fssh.nc) timestep: 0.5 fs, substeps: 25 total time: 5000 fs (10000 time steps, 01_dietschreit_sa2-cas22_196_S1traj.nc) 100 fs (200 time steps, I01_dietschreit_sa2-cas22_500_fssh.nc) number of trajectories: 196 (I01_dietschreit_sa2-cas22_196_S1traj.nc) 4999 (I01_dietschreit_sa2-cas22_500_fssh.nc) Electronic structure theory: engine: OpenMOLCAS (v24.06) reference method: SA(2)-CAS(2,2) (state-average CASSCF) basis set: cc-pVDZ charge: +1 states: two lowest singlets (S0 and S1) as active states ---------------------------------------- Data of Aza-Diarylethenes (R04 & R06) M. Hartinger & C. Müller a) About Aza-diarylethenes (aDAEs) are emerging photoswitches that undergo reversible light-induced C–N bond formation. This dataset contains surface-hopping trajectories for two representative aDAEs, R04 (corresponding to D-open in 10.1039/D6CP01330A) and R06 (corresponding to K-open in 10.1039/D6CP01330A), generated to investigate the mechanistic origin of their distinct photochemical outcomes. This dataset contains surface-hopping trajectories classified according to their photochemical outcome. Trajectories are categorized as productive (leading to formation of the closed isomer), unproductive (returning to the open isomer or forming rare side products), and still excited (remaining in an excited electronic state at the end of the simulation). R04: 45% productive, 43% unproductive, 11% still excited R06: 29% productive, 66% unproductive, 5% still excited b) Computational Details Dynamics: NAMD method: trajectory surface hopping with Zhu-Nakamura algorithm in PyRAI2MD timestep: 0.5 fs total time: up to 1 ps for R06 and up to 2 ps for R04 number of trajectories: 97 trajectories each for R04 and R06 Electronic structure theory: engine: OpenQP reference method: MRSF-TD-DTCAM-AEE (mixed-reference spin-flip TD-DFT) basis set: 6-31G* charge: 0 states: three lowest singlets (S0, S1 and S2) as active states