Machine learning for predicting product distributions in catalytic regioselective reactions

Abstract

Gaining predictable control over various forms of selectivities, such as enantio- and/or regio-selectivities, has been a long-standing goal in chemical catalysis. Although a number of factors such as the molecular features of the reactants and catalysts, as well as the reaction conditions, can influence the outcome of a reaction, it is not quite conspicuous as to what combinations of these parameters would offer a desired form of selectivity. We use machine learning tools, such as the neural network (NN), decision tree (DT), logistic regression (LR) and Random forest algorithms, to (a) analyze the outcome of an important catalytic regio-selective difluorination reaction of alkenes, and (b) decipher the complex interplay of various molecular parameters and their non-linear dependencies. The connection between what features of alkenes will yield 1,1-difluorination and how subtle changes would steer the reaction to 1,2-difluorination under identical conditions is enunciated. The NN was able to accurately predict whether a given alkene would yield a 1,1- or 1,2-difluorinated product. A combination of DT and the random forest classifier offered important chemical insights, which could be used in making a more rational choice of the reactant alkene for the desired regioisomeric product. The results could have far reaching implications in predicting which regioisomer is likely to be formed under a given set of conditions, and thus this technique is capable of expediting the development of catalytic transformations.

Machine learning dihydrogen activation in the chemical space surrounding Vaska's complex

“…[62][63][64][65][66][67][68] In other areas, including organic synthesis, [69][70][71][72][73] and theoretical [74][75][76][77][78][79][80][81] and inorganic 82,83 chemistry, the use of ML is rapidly growing. In catalysis, 84,85 several examples have been reported for both heterogeneous [86][87][88][89][90][91][92][93] and homogeneous [94][95][96][97][98] systems.…”

Quantum-mechanical transition-state model combined with machine learning provides catalyst design features for selective Cr olefin oligomerization

“…92 Sunoj used a combination of a neural network and random forest model to identify the regioselectivity of catalytic diuorination of alkenes. 93 Brgoch screened over 100 000 compounds using a support vector machine regression to identify novel highly compressible metal materials, 94 and Buehler used convolutional neural networks to search for new composite metal materials. 95 Xin used articial neural networks to identify heterogeneous metallic catalysts for CO capture and reduction.…”

Autonomous Discovery in the Chemical Sciences Part I: Progress

“…[218] Using informative electronic (e.g., Fukui functions [219,220] )a nd steric (e.g., Sterimol [221,222] )d escriptors can help model generalization and performance,e specially in low data environments.G iven suitable descriptors and holding other process parameters constant, complex properties have been described with linear or nearly linear models,f or example,c atalyst performance and enantioselectivity ( Figure 11). [223][224][225][226][227] Descriptors tailored to aspecific reaction class can be effective representations for predicting regioselectivity [228] and yield [229] among other performance metrics,a lthough they may not be broadly applicable across reaction and substrate classes.I np rinciple, these descriptors could be calculated with greater universality than expert-selected ones already known to be relevant. [230] Similarly,s electivity in complex synthetic steps can be explained by expert-defined DFT calculations [231] that could, in principle,bea utomated.…”

“…Given suitable descriptors and holding other process parameters constant, complex properties have been described with linear or nearly linear models, for example, catalyst performance and enantioselectivity (Figure 11). [223–227] Descriptors tailored to a specific reaction class can be effective representations for predicting regioselectivity [228] and yield [229] among other performance metrics, although they may not be broadly applicable across reaction and substrate classes. In principle, these descriptors could be calculated with greater universality than expert‐selected ones already known to be relevant [230] .…”