Scientific Deep Machine Learning Concepts for the Prediction of Concentration Profiles and Chemical Reaction Kinetics: Consideration of Reaction Conditions

Emerging concepts from scientific deep machine learning such as physics-informed neural networks (PINNs) enable a data-driven approach for the study of complex kinetic problems. We present an extended framework that combines the advantages of PINNs with the detailed consideration of experimental parameter variations for the simulation and prediction of chemical reaction kinetics. The approach is based on truncated Taylor series expansions for the underlying fundamental equations, whereby the external variations can be interpreted as perturbations of the kinetic parameters. Accordingly, our method allows for an efficient consideration of experimental parameter settings and their influence on the concentration profiles and reaction kinetics. A particular advantage of our approach, in addition to the consideration of univariate and multivariate parameter variations, is the robust model-based exploration of the parameter space to determine optimal reaction conditions in combination with advanced reaction insights. The benefits of this concept are demonstrated for higher-order chemical reactions including catalytic and oscillatory systems in combination with small amounts of training data. All predicted values show a high level of accuracy, demonstrating the broad applicability and flexibility of our approach.

Segmented Flow Processes to Overcome Hurdles of Whole-Cell Biocatalysis in the Presence of Organic Solvents

In modern process development, it is imperative to consider biocatalysis, and whole‐cell catalysts often represent a favored form of such catalysts. However, the application of whole‐cell catalysis in typical organic batch two‐phase synthesis often struggles due to mass transfer limitations, emulsion formation, tedious work‐up and, thus, low yields. Herein, we demonstrate that utilizing segmented flow tools enables the conduction of whole‐cell biocatalysis efficiently in biphasic media. Exemplified for three different biotransformations, the power of such segmented flow processes is shown. For example, a 3‐fold increase of conversion from 34 % to >99 % and a dramatic simplified work‐up leading to a 1.5‐fold higher yield from 44 % to 65 % compared to the analogous batch process was achieved in such a flow process.

Concepts for flow chemistry with whole-cell biocatalysts

By combining continuous flow processing and biocatalysis, efficient, stable and cost-effective processes can be realised. In this review, an overview about different concepts for continuous flow processes based on the use of whole-cells as catalysts is given.

Flow Process for Ketone Reduction Using a Superabsorber-Immobilized Alcohol Dehydrogenase from Lactobacillus brevis in a Packed-Bed Reactor

Flow processes and enzyme immobilization have gained much attention over the past few years in the field of biocatalytic process design. Downstream processes and enzyme stability can be immensely simplified and improved. In this work, we report the utilization of polymer network-entrapped enzymes and their applicability in flow processes. We focused on the superabsorber-based immobilization of an alcohol dehydrogenase (ADH) from Lactobacillus brevis and its application for a reduction of acetophenone. The applicability of this immobilization technique for a biotransformation running in a packed bed reactor was then demonstrated. Towards this end, the immobilized system was intensively studied, first in a batch mode, leading to >90% conversion within 24 h under optimized conditions. A subsequent transfer of this method into a flow process was conducted, resulting in very high initial conversions of up to 67% in such a continuously running process.

An Asymmetric Organocatalytic Aldol Reaction of a Hydrophobic Aldehyde in Aqueous Medium Running in Flow Mode

Reaction conditions have been identified to conduct a one-pot asymmetric organocatalytic aldol reaction with a hydrophobic substrate in aqueous medium via a process running in flow mode. By employing a mixture of water and 2-propanol, a hydrophobic aldehyde and 3.6 mol% of an organocatalyst, this microreactor process affords the desired aldol adduct with a conversion of 74% and an enantioselectivity of 89% after a reaction time of 60 minutes.