Fidelity of Robotic Chemical Operations of Homogenous and Heterogeneous Reactions

Chemical synthesis always involves laborious and time-consuming operations. Autonomous chemical synthesis systems can liberate experimentalists from those routine tasks, but the design of automatic chemical reaction machines is of great challenge. We describe herein the development of a real-life automatic chemical operation platform by using a coordinated multi-robot system with artificial intelligence and machine learning methods. The capacities of this platform were demonstrated by its application in condensation reaction of 2,4-dinitrophenyldrazine with formaldehyde, and automatic catalyst evaluation of a heterogeneous aza Diels-Alder reaction, suggesting the designed machine is applicable to both homogenous and heterogeneous chemical reactions. Besides, the excellent repeatability and reproducibility of robotic tests on these two model reactions indicate the great promise of massive transfer of manual to automatic operation.