Multivariate chemical reaction optimization involving catalytic systems is a non-trivial task due to the high number of tuneable parameters and discrete choices.
Multivariate chemical reaction optimization involving catalytic systems is a non-trivial task due to the high number of tuneable parameters and discrete choices. Closed-loop optimization featuring active Machine Learning (ML) represents a powerful strategy for automating reaction optimization. However, the translation of chemical reaction conditions into a machine-readable format comes with the challenge of finding highly informative features which accurately capture the factors for reaction success and allow the model to learn efficiently. Herein, we compare the efficacy of different calculated chemical descriptors for a high throughput generated dataset to determine the impact on a supervised ML model when predicting reaction yield. Then, the effect of featurization and size of the initial dataset within a closed-loop reaction optimization was examined. Finally, the balance between descriptor complexity and dataset size was considered. Ultimately, tailored descriptors did not outperform simple generic representations, however, a larger initial dataset accelerated reaction optimization.
Multivariate chemical reaction optimization involving catalytic systems is a non-trivial task due to the high number of tuneable parameters and discrete choices. Closed-loop optimization featuring active Machine Learning (ML) represents a powerful strategy for automating reaction optimization. However, the translation of chemical reaction conditions into a machine-readable format comes with the challenge of finding highly informative features which accurately capture the factors for reaction success and allow the model to learn efficiently. Herein, we compare the efficacy of different calculated chemical descriptors for a high throughput generated dataset to determine the impact on a supervised ML model when predicting reaction yield. Then, the effect of featurization and size of the initial dataset within a closed-loop reaction optimization was examined. Finally, the balance between descriptor complexity and dataset size was considered. Ultimately, tailored descriptors did not outperform simple generic representations, however, a larger initial dataset accelerated reaction optimization.
There has been an increasing interest in the use of automated self-optimising continuous flow platforms for the development and manufacture in synthesis in recent years. Such processes include multiple reactive and work-up steps, which need to be efficiently optimised. Here, we report the combination of multi-objective optimisation based on machine learning methods (TSEMO algorithm) with self-optimising platforms for the optimisation of multi-step continuous reaction processes. This is demonstrated for a pharmaceutically relevant Sonogashira reaction. We demonstrate how optimum reaction conditions are re-evaluated with the changing downstream work-up specifications in the active learning process. Furthermore, a Claisen-Schmidt condensation reaction with subsequent liquid-liquid separation was optimised with respect to three-objectives. This approach provides the ability to simultaneously optimise multi-step processes with respect to multiple objectives, and thus has the potential to make substantial savings in time and resources.
We herein report a novel kinetic modelling methodology whereby identification of the correct reaction model and kinetic parameters is conducted by an autonomous framework combined with transient flow measurements to enable comprehensive process understanding with minimal user input. An automated flow chemistry platform was employed to initially conduct linear flow-ramp experiments to rapidly map the reaction profile of three processes using transient flow data. Following experimental data acquisition, a computational approach was utilised to discriminate between all possible reaction models as well as identify the correct kinetic parameters for each process. Species that are known to participate in the process (starting materials, intermediates, products) are initially inputted by the user prior to flow ramp experiments, then all possible model candidates are compiled into a model library based on their potential to occur after mass balance assessment. Parallel computational optimisation then evaluates each model by algorithmically altering the kinetic parameters of the model to allow convergence of a simulated kinetic curve to the experimental data provided. Statistical analysis then determines the most likely reaction model based on model simplicity and agreement with experimental data. This automated approach to gaining full process understanding, whereby a small number of data-rich experiments are conducted, and the kinetics are evaluated autonomously, shows significant improvements on current industrial optimisation techniques in terms of labour, time and overall cost. The computational approach herein described can be employed using data from any set of experiments and the code is open-source.
Self-optimising chemical systems have experienced a growing momentum in recent years. Herein, we review algorithms used for the self-optimisation of chemical reactions in an accessible way for the general chemist.
Abstract Despite growing applications being reported both in academia and industry, continuous flow chemistry remains a relatively untaught field across most chemistry undergraduate courses. This is particularly true in laboratory practical classes, where it is often deemed simpler to carry out synthetic reactions in traditional batch mode using round-bottomed flasks. Herein, we report the development of an undergraduate project that utilises cheap and readily available materials to construct continuous flow reactors. The students compare the performance of different types of reactors and conditions in a biphasic selective acetylation of a symmetrical diamine. Throughout the investigation, the students can vary multiple parameters as they optimise the reaction, thus actively learning and readjusting them based on their improved understanding. The experiments give the students an appreciation of continuous flow techniques in comparison to batch.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
'/%3e%3cuse xlink:href='%23a' transform='translate(288.889 -214.211)'/%3e%3cuse xlink:href='%23a' transform='translate(108 342.749)'/%3e%3cuse xlink:href='%23a' transform='translate(469.189 342.749)'/%3e%3c/svg%3e)
