Simulation-Assisted Deep Learning Techniques for Commercially Applicable OLED Phosphorescent Materials

Abstract

Phosphorescent light-emitting materials play a central role in organic light-emitting diode (OLED) devices. Due to their synthesis difficulties, unsystematic trial-and-error synthesis is prohibitively challenging. For this reason, deep learning (DL), which has shown success in various fields, is being actively applied to materials discovery. However, one challenge in applying DL to phosphorescent materials is the limited amount of experimental data set. One way to circumvent this issue is to apply powerful DL techniques that have been successfully implemented in several domains. Another solution would be to use a large amount of data set for pretraining DL models with simulated properties highly relevant to target properties. In this work, phosphorescent materials are represented as strings, molecular graphs, and point clouds, which are employed by language models, two-dimensional graph, and three-dimensional graph neural networks. In addition, more than 200 000 molecules with simulated properties highly relevant to experimental properties are used for pretraining the DL models. Our work shows high performance in the prediction of five experimental properties that are importantly considered when commercializing OLED devices. This means that faster material discovery for OLEDs can be achieved through DL models that are trained with simulation information that is highly correlated with experimental properties.