... is a compilation of surface hopping datasets that offers high-quality quantum chemical data (e.g. energies, forces, nonadiabatic couplings)
of photochemical reactions; ideal for benchmarking and machine learning applications.
... is an integrated Python software package for the streamlined collection, managment, processing, analysis and visualization of surface hopping data.
SHNITSEL is a centralized platform compiling surface hopping datasets relevant for photochemical and photophysical studies. The database currently provides information on key photochemical processes, including photodissociation, roaming mechanisms, E/Z-isomerization, and electrocyclic reactions. These processes are fundamental in various applications, from synthetic organic chemistry to the development of new theoretical and machine learning (ML)-based methods.
The dataset consists of extensive ab initio calculations covering ground and excited-state properties, including energies, forces, dipole moments, nonadiabatic couplings, transition dipoles, and spin-orbit couplings. These high-accuracy quantum chemical properties serve as a valuable benchmark for validating computational approaches and training ML models for excited-state dynamics.
SHNITSEL is intended as a growing resource for the community. Researchers are encouraged to contribute by sharing links to their own datasets, helping expand the database’s scope. If you would like to add your dataset, feel free to reach out!