The discovery of chemical reactions is an inherently unpredictable and time-consuming process. An attractive alternative is to predict reactivity, although relevant approaches, such as computer-aided reaction design, are still in their infancy. Reaction prediction based on high-level quantum chemical methods is complex, even for simple molecules. Although machine learning is powerful for data analysis, its applications in chemistry are still being developed. Inspired by strategies based on chemists' intuition, we propose that a reaction system controlled by a machine learning algorithm may be able to explore the space of chemical reactions quickly, especially if trained by an expert. Here we present an organic synthesis robot that can perform chemical reactions and analysis faster than they can be performed manually, as well as predict the reactivity of possible reagent combinations after conducting a small number of experiments, thus effectively navigating chemical reaction space. By using machine learning for decision making, enabled by binary encoding of the chemical inputs, the reactions can be assessed in real time using nuclear magnetic resonance and infrared spectroscopy. The machine learning system was able to predict the reactivity of about 1,000 reaction combinations with accuracy greater than 80 per cent after considering the outcomes of slightly over 10 per cent of the dataset. This approach was also used to calculate the reactivity of published datasets. Further, by using real-time data from our robot, these predictions were followed up manually by a chemist, leading to the discovery of four reactions.

We utilise 3D design and 3D printing techniques to fabricate a number of miniaturised fluidic 'reactionware' devices for chemical syntheses in just a few hours, using inexpensive materials producing reliable and robust reactors. Both two and three inlet reactors could be assembled, as well as one-inlet devices with reactant 'silos' allowing the introduction of reactants during the fabrication process of the device. To demonstrate the utility and versatility of these devices organic (reductive amination and alkylation reactions), inorganic (large polyoxometalate synthesis) and materials (gold nanoparticle synthesis) processes were efficiently carried out in the printed devices.

A ‘dial-a-molecule’ platform for algorithm driven organic synthesis using real-time feedback, via in-line flow NMR spectroscopy, is demonstrated.

SummaryWe present a study in which the versatility of 3D-printing is combined with the processing advantages of flow chemistry for the synthesis of organic compounds. Robust and inexpensive 3D-printed reactionware devices are easily connected using standard fittings resulting in complex, custom-made flow systems, including multiple reactors in a series with in-line, real-time analysis using an ATR-IR flow cell. As a proof of concept, we utilized two types of organic reactions, imine syntheses and imine reductions, to show how different reactor configurations and substrates give different products.

We use two 3D-printing platforms as solid-and liquid-handling fabricators, producing sealed reactionware for chemical synthesis with the reagents, catalysts and purification apparatus integrated into monolithic devices. Using this reactionware, a multi-step reaction sequence was performed by simply rotating the device so that the reaction mixture flowed through successive environments under gravity, without the need for any pumps or liquid-handling prior to product retrieval from the reactionware in a pure form.Scheme 1 Fabrication scheme for the integration of 3D-printing techniques with automated liquid handling to produce sealed reactionware for multi-step syntheses. Dotted line indicates the only process not automated in the current work. ; Web: http://www.croninlab.com † Electronic supplementary information (ESI) available: Details of synthetic protocols, analysis of NMR spectra of crude reaction mixtures containing (1a, b) and (2a, b), along with characterization of products (3a, b) and 1 H NMR spectra of compounds (3a) and (3b). Diagrams of, and notes on, the 3D printed reactionware used are also provided. See