Estimation of selectivities in the lipase-catalyzed esterification of trimethylolpropane with fatty acids

Abstract

Interest has increased in using enzyme-catalyzed esterification to produce fatty acid esters of trimethylolpropane for use as biolubricants. Mathematical models of this process can be used to guide the design and scale-up of production processes. The selectivities of the enzyme for the various trimethylolpropane species (unesterified, mono-esterified, di-esterified and tri-esterified) are important parameters of such models. Previous attempts to estimate selectivities in the esterification of trimethylolpropane have not produced accurate estimates. In the current work, we show that the fingerprinting method can be used to obtain accurate estimates. We apply the method to various data sets from the literature, using Bayesian techniques for model fitting. We show that, depending on the reaction conditions, different versions of the model are necessary: (i) A model that treats the reactions as irreversible; (ii) an “irreversible model” that recognizes mass transfer limitations, but treats them using a pseudoprocessive model; and (iii) a model that treats the first reaction as reversible and is appropriate for systems in which water is added at the beginning of the reaction.