Samsung Advanced Institute of Technology

We demonstrate a computational workflow based on quantum chemical calculations and application of generative neural networks for the discovery of novel materials. We apply the developed workflow for search of molecules suitable for fusion of triplet-triplet (TTF) excitations in blue OLED devices. Application of generative neural network allowed us to screen only the most promising regions of the chemical space. Another neural network based on graph convolutions for prediction of excitation energies made it possible to estimate excitation energies and filter molecules before running time consuming quantum chemical calculations. The described approach can be useful for the computer-aided design of materials with energy level alignment favorable for efficient energy transfer, triplet harvesting, exciton fusion processes, which are crucial for the development of the next generation of OLED materials.