In this study the influence of diffusion limitation on enzymatic kinetically controlled cephalexin synthesis from phenylglycine amide and 7-aminodeacetoxycephalosporinic acid (7-ADCA) was investigated systematically. It was found that if diffusion limitation occurred, both the synthesis/hydrolysis ratio (S/H ratio) and the yield decreased, resulting in lower product and higher by-product concentrations. The effect of pH, enzyme loading, and temperature was investigated, their influence on the course of the reaction was evaluated, and eventually diffusion limitation was minimised. It was found that at pH >or=7 the effect of diffusion limitation was eminent; the difference in S/H ratio and yield between free and immobilised enzyme was considerable. At lower pH, the influence of diffusion limitation was minimal. At low temperature, high yields and S/H ratios were found for all enzymes tested because the hydrolysis reactions were suppressed and the synthesis reaction was hardly influenced by temperature. The enzyme loading influenced the S/H ratio and yield, as expected for diffusion-limited particles. For Assemblase 3750 (the number refers to the degree of enzyme loading), it was proven that both cephalexin synthesis and hydrolysis were diffusion limited. For Assemblase 7500, which carries double the enzyme load of Assemblase 3750, these reactions were also proven to be diffusion limited, together with the binding-step of the substrate phenylglycine amide to the enzyme. For an actual process, the effects of diffusion limitation should preferably be minimised. This can be achieved at low temperature, low pH, and high substrate concentrations. An optimum in S/H ratio and yield was found at pH 7.5 and low temperature, where a relatively low reaction pH can be combined with a relatively high solubility of 7-ADCA. When comparing the different enzymes at these conditions, the free enzyme gave slightly better results than both immobilised biocatalysts, but the effect of diffusion limitation was minimal.

A method is proposed for identification of kinetic parameters when diffusion of substrates is limiting in reactions catalyzed by immobilized enzymes. This method overcomes conventional sequential procedures, which assume immobilization does not affect the conformation of the enzyme and, thus, consider intrinsic and inherent kinetics to be the same. The coupled equations describing intraparticle mass transport are solved simultaneously using numerical methods and are used for direct estimation of kinetic parameters by fitting modeling results to time-course measurements in a stirred tank reactor. While most traditional procedures were based on Michaelis-Menten kinetics, the method presented here is applicable to more complex kinetic mechanisms involving multiple state variables, such as ping-pong bi-bi. The method is applied to the kinetic resolution of (R/S)-1-methoxy-2-propanol with vinyl acetate catalyzed by Candida antarctica lipase B. A mathematical model is developed consisting of irreversible ping-pong bi-bi kinetics, including competitive inhibition of both enantiomers. The kinetic model, which fits to experimental data over a wide range of both substrates (5-95%) and temperatures (5-56 degrees C), is used for simulations to study typical behavior of immobilized enzyme systems.

No abstract

Lipase-catalyzed kinetic resolution of racemates is a popular method for synthesis of chiral synthons. Most of these resolutions are reversible equilibrium limited reactions. For the first time, an extensive kinetic model is proposed for kinetic resolution reactions, which takes into account the full reversibility of the reaction, substrate inhibition by an acyl donor and an acyl acceptor as well as alternative substrate inhibition by each enantiomer. For this purpose, the reversible enantioselective transesterification of (R/S)-1-methoxy-2-propanol with ethyl acetate catalyzed by Candida antarctica lipase B (CAL-B) is investigated. The detailed model presented here is valid for a wide range of substrate and product concentrations. Following model discrimination and the application of Haldane equations to reduce the degree of freedom in parameter estimation, the 11 free parameters are successfully identified. All parameters are fitted to the complete data set simultaneously. Six types of independent initial rate studies provide a solid data basis for the model. The effect of changes in substrate and product concentration on reaction kinetics is discussed. The developed model is used for simulations to study the behavior of reaction kinetics in a fixed bed reactor. The typical plot of enantiomeric excess versus conversion of substrate and product is evaluated at various initial substrate mixtures. The model is validated by comparison with experimental results obtained with a fixed bed reactor, which is part of a fully automated state-of-the-art miniplant.