Kinetic Modeling of a Consecutive Enzyme-Catalyzed Enantioselective Reaction in Supercritical Media

Abstract

Based on experimental data of both batch and continuous enzyme-catalyzed kinetic resolutions of (±)- trans -1,2-cyclohexanediol in supercritical carbon dioxide, kinetic models of increasing complexity were developed to explore the strengths and drawbacks of various modeling approaches. The simplest, first-order model proved to be a good fit for the batch experimental data in regions of high reagent concentrations but failed elsewhere. A more complex system that closely follows the true mechanism was able to fit the full range of experimental data, find constant reaction rate coefficients, and was successfully used to predict the results of the same reaction run continuously in a packed bed reactor. Care must be taken when working with such models, however, to avoid problems of overfitting; a more complex model is not always more accurate. This work may serve as an example for more rigorous reaction modeling and reactor design in the future.

Kinetic modelling: Regression and validation stages, a compulsory tandem for kinetic model assessment

“…[23] Based on these multiple observables (experimental concentrations), kinetic modellers could be tempted to develop complex kinetic models, requiring the estimation of several parameters and increasing the risk of overfitting. [24][25][26][27] A validation stage can determine the presence of overfitting.…”