Session: Physical Chemistry Poster:

The use of materials that undergo singlet fission (SF) is a successful strategy for improvement of the efficiency of solar cells. In order to be a singlet fission chromophore applicable for organic solar cell material, the molecules should absorb intensely at about 2 eV and also to be chemically stable. The latter is, however, discordant with the recent findings that molecules with low to intermediate diradical character are promising candidates for singlet fission chromophores. Moreover, specific restrictions are defined for the ratios of the excitation energies to the first and second singlet and triplet excited states. Taking into account all these requirements, it is not surprising that the amount of known singlet fission chromophores is relatively small. Another impediment in the quest for new SF chromophores represents the non-trivial spectroscopic measurement and theoretical description of excited states for open-shell molecular systems. Such molecules should be investigated using highly accurate multiconfigurational (MC) methods, which predict correctly the diradical character of a SF chromophore, as well as all excitation energies, whereof the feasibility conditions can be estimated. However, these methods are also very resource demanding and limited to small size molecules. Naturally, arises the question whether a fast and accurate large scale computational screening for new singlet fission chromophores can be developed. Possible solution of the problem is to train a machine learning model that can predict feasibility conditions or at least give qualitatively correct trends at the MC level. For the purpose, a database is needed. Therefore, we have initiated and present here the first (to our knowledge) ab intio database of MC data for singlet fission chromophores.

Dr. Julia Romanova

Sofia University "St. Kliment Ohridski"

Appearances