A Simple Yet Effective Descriptor for Predicting the Glass Transition Temperature of Y-Series Acceptors

Owing to the advances in Y-series acceptors, organic solar cells have achieved over 20% power conversion efficiency, and it becomes imperative to design Y-series acceptors with a high glass transition temperature (T_g) to obtain long-term device stability for practical applications. In order to build an accurate and efficient data-driven model to predict T_g, we propose a simple yet effective descriptor, namely, the self-diffusion coefficient at high temperature, which can be obtained only via a short-time MD simulation. Compared to the conventional method, the computational cost is dramatically reduced by 2-3 orders of magnitude. Remarkably, our method is applicable to both monomeric and dimeric Y-series acceptors and has achieved an unprecedented prediction accuracy with a high determination coefficient (R²) of 0.94 and a small root-mean-square error (RMSE) of 9.0 °C. Furthermore, our results point out that the inner and outer side chains have a distinct influence on the T_g of the Y-series acceptors. Upon removing the outer side chains of Y6, the T_g can be increased from ∼100 to 225 °C. This work paves the way toward precise and rapid screening of high-T_g Y-series acceptors.