A retrospective on the automation of laboratory synthetic chemistry

Abstract The quixotic pursuit to automate laboratory synthetic chemistry has a long history. The rationale for attempting automated synthesis is to increase productivity, to improve quality, to increase precision, to liberate the scientist from monotony, and to provide capabilities for exhaustive experimentation. Early research efforts were consumed in building homemade hardware components and overcoming inadequate computers; these limitations have diminished with recent technology. Diverse sample-handling problems, particularly unexpected phase separations, remain the primary obstacle to the development of general-purpose automated synthesis machines. Though the automation of total synthesis has figured prominently as a goal, there are many roles for automation in support of a research program in synthetic chemistry. The key stages of research encompass exploratory work, systematic physical and synthetic studies, total synthesis, and pre-pilot plant characterizations. No general-purpose synthesizers have been constructed. Five special-purpose automation designs have been investigated for various tasks in the research cycle, including continuous flow reactors, single-batch reactors, single-robotic synthesizers, dual-robotic synthesizers, and integrated workstations. The system design depends both on sample-handling issues and on the scientific objective. Continuous flow reactors have been applied to the characterization of oscillating reactions and enzymatically catalyzed reactions. Single-batch reactors have been applied to peptide and DNA synthesis, rapid synthesis of radiopharmaceuticals, and process optimization studies. Single-robot systems have been constructed to engender greater flexibility, and dual-robotic synthesizers have been constructed with each robot dedicated to a particular set of tasks. Integrated workstations comprised of multiple reactors, hardware modules, and analytical instruments have been applied to total synthesis, fundamental studies of self-assembly processes, and technically complex sample preparations in molecular biology. Primitive machines for studying evolution are harbingers of powerful automation systems for fundamental research in molecular biology. Among these machines there exist numerous solutions for difficult sample-handling problems. Several automation systems have capabilities for automated decision-making based on experimental data, a key feature for autonomous experimentation. Automation of synthetic chemistry is a science-driven effort that must draw electically from the disciplines of analytical chemistry, chemical engineering, robotics, and computer science. Accelerated scientific research can occur now if scientists have access to the full diapason of special-purpose automation machines for synthetic chemistry. Future automation systems will likely include enhanced general-purpose capabilities.