Abstract
Indeed, structural and electronic behavior of organic semiconductors control their performance for organic solar cells. To attain the higher performance a deeper understanding of materials is required. Here, multi-scale computational modeling is used to study the effect of structural variation on the halogen substitution. Quantum chemical calculations, molecular dynamics simulations and machine learning are used. Halogens are introduced at the terminal position of electron-acceptors and their electronic properties are further examined. Quantum chemical analysis has shown that fluorinated and chlorinated acceptors have lower exciton binding energy, higher transfer integral, and lower reorganization energy; suggesting that these acceptors are better than others. Moreover, the power conversion efficiency of newly designed acceptor materials is also predicted through already trained machine learning models. Fluorinated and chlorinated acceptors showed higher PCE, but the difference is not very large as compared with other acceptors. Further, the mixing behavior of the designed acceptors with the polymer donor PBDB-T is investigated using the Florgy-Huggins parameter. The molecular packing of donor and acceptor molecules is studied using radial distribution function. Fluorinated and chlorinated acceptors showed lower Florgy-Huggins parameter and free energy of mixing. We believe that multiscale modeling has the potential to explore various electronic and photovoltaic aspects of organic semiconductors even before synthesis.