Robotic Modules for the Programmable Chemputation of Molecules and Materials

摘要

Before big data methods like machine learning and artificial intelligence methods can be used in chemistry, the digitization of chemistry and materials requires the development of a universal standard that is both affordable and broadly applicable. This has parallels with the foundations of the digital revolution which required standard architectures with a well-defined specification. Recently, we have developed automated platforms for the chemical artificial intelligence driven discovery, synthesis of molecules, materials, nanomaterials, and formulations. To avoid the use of highly expensive systems we focussed on the design and construction of standard hardware and software modules creating a road map for the digitization of chemistry across different fields. Our platforms can be divided into four different categories depending on their application: i) discovery systems for the search of chemical space and new reactivity, ii) synthesis and manufacture of fine chemicals, iii) formulation discovery and exploration, and iv) materials discovery and synthesis. We describe the evolution and convergence of these platforms in terms of, common hardware, firmware, and software along with the development of a programming language for chemical and material systems. This programming approach is not only useful for reliable synthesis, but for design of experiments, discovery, optimisation and providing new standards for collaboration. This approach is also vital for the verification of findings published in the literature, databases, to increase the reliability of experimental outcomes, and to allow collaboration across different research laboratory settings as well as the sharing of failed experiments.