Abstract
Automation is ubiquitous in modern life, but the chemical synthesis research laboratory remains largely manual, using methods from the early 20th century. Automation and digitisation stand to improve the throughput of chemistry synthesis laboratories, both by increasing the pace of work and automating data generation and capture. The Chemputer is a platform for digital synthesis developed by the Cronin group in Glasgow, offering a wide range of capabilities for batch chemistry. Unfortunately, the Chemputer costs between R100 000 to R700 000 build. This project aimed to develop an inexpensive and open-source alternative to the Chemputer using the design science research methodology. The project also investigated 3D printing for the manufacture of low-cost fluid handling parts. A syringe pump, commercial selector valve actuator, and reactor module were designed to form the base of a low-cost, automated, chemical synthesis system. A 3D printed, low-cost selector valve and non-return valve were designed and tested but were unsuccessful. The modules were automated using an Arduino Mega and RepRap Arduino Mega Pololu (RAMPS) connected to a computer running a Python software package. A webserver running on the Flask framework was developed to coordinate multiple robots and provide a database to store reaction data securely. The cost of a system comprising two syringe pumps, two valves, and one reactor module, driven by a Raspberry Pi, was R34 145.33.
The synthesis of the pharmaceuticals aspirin and paracetamol were performed using the system. The syntheses showed that the system was capable of performing simple chemical syntheses. Three robots were used to collaboratively search an azo dye space, communicating with the server to receive reactions and send images of the reaction products. Together the robots performed 117 reactions over a week with minimal operator intervention. The colours of the dye solutions were determined automatically using the OpenCV library. The collaborative search showed that the robots were reliable and could be used for large-scale parallel syntheses, such as library generation. The automated system presented here is capable of simple liquid-phase chemistry and could be expanded upon to provide more capabilities in the future. All designs, source code, and instructions for operation were made freely available on Github.
Keywords:
Automated chemical synthesis, automation, IoT, 3D printing, open-source, Python.