Researchers have used a robotic platform to produce - with no physical intervention - three pharmaceutical compounds with yields and purities comparable to those achieved by manual efforts, they say. Their results represent a new approach to automation of chemical synthesis. Historically, most work to synthesize chemicals must be done manually because automation is limited to single classes of reactions or work-flows. Synthetic chemists who perform this manual work use senses and skills that they may take for granted. Finding a way to automate these steps in machine-driven approaches has been challenging. Here, seeking to overcome this barrier, Sebastian Steiner and colleagues created a way to automate a decision-based procedure that trained chemists typically accomplish visually - a task related to the assembly of molecules. They achieved this automation using a conductivity sensor - one that could potentially outperform human vision, the authors say. The steps in their automated process were implemented in a modular robotic platform, instructed by a chemical programming language that formalizes and controls the assembly of the relevant molecules. Critically, the instructions the platform relies on can be coded by users with little programming knowledge, the authors say. The researchers demonstrated their platform's ability by instructing it to make three pharmaceutical compounds (Nytol, Rufinamide, and Sildenafil), which it did without any physical intervention, and at high quality. The synthesis paths of all three compounds are captured as digital code that can be transferred flexibly between platforms, the authors say, greatly enhancing reproducibility and reliable access to complex molecules.
###
Journal
Science