Samsung Advanced Institute of Technology

Technologies relieving the burden of arduous manual synthesis are essential to accelerate the development of organic compounds, and their efficacy will be further increased by arming autonomous functions beyond automation. To address this issue, we developed an autonomous synthesis-robot in which artificial intelligence (AI) and robots collaborate to find optimal synthetic recipe. Given a target molecule, the AI initially plans synthetic pathways and reaction conditions, receives experimental feedback from the robots to update its models, and progressively reach an optimal recipe. The performance is successfully validated on three organic electronic materials by finding synthetic recipes whose conversion yields are better than known references. The autonomous system is based on batch-reactors compatible with conventional laboratories, making it highly pragmatic and valuable for chemists to assist researches.