Machine learning methods for excited-state dynamics simulations in light-induced spin-crossover complexes

Spin crossover complexes are bistable transition-metal compounds in which light, or other external simuli, is used to induce a change in the magnetic state of the system, allowing them to be employed as molecular switches in future spintronics and photonics technologies.

Despite the rapid development of experimental techniques to characterise these complexes, computational material science is not yet able to provide a quantitative description of the light-induced spin crossover mechanism.

These limitations are mainly due to the need for enormous computational resources to perform accurate excited-state dynamics simulations in medium-sized transition-metal complexes.

The aim of the SCOML project is to provide the proof of concept for a new machine learning-based strategy that will enable efficient and accurate simulations of the light-induced spin-crossover mechanism, thus paving the way for a systematic design of new materials.

Achieving this goal has the potential to revolutionise the production of new technological solutions to guide Europe towards a digital and green transition.