Mathematical modeling of enzymatic syntheses of biosurfactants catalyzed by immobilized lipases

• Published in: Reaction kinetics, mechanisms and catalysis Vol. 130; no. 2; pp. 699 - 712
• Main Authors: de Carvalho Lima Torres, Alice, de Lima, Lionete Nunes, Tardioli, Paulo Waldir, de Sousa Júnior, Ruy
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.08.2020
• Subjects: Catalysis; Chemistry; Chemistry and Materials Science; Industrial Chemistry/Chemical Engineering; Physical Chemistry; Sugar fatty esters syntheses; Kinetic model; Ping Pong Bi Bi mechanism; Enzymatic route
• ISSN: 1878-5190; 1878-5204
• DOI: 10.1007/s11144-020-01812-w

This work aimed at the kinetic modeling of enzymatic reactions for the production of sugar fatty esters type biosurfactants. The production of biosurfactants by esterification (72 h, 250 rpm stirring and 1.5 g of molecular sieve) of oleic and lauric acids with fructose and lactose, in the presence of tert -butanol and 2-methyl-2-butanol, has already been evaluated experimentally using Candida antarctica lipase B immobilized (CALB-IM-T2-350) and Pseudomonas fluorescens lipase (PFL) immobilized on octyl-silica. Acid conversion against time assays were carried out changing the following parameters: acid:sugar molar ratio (1:1 and 1:2), sugar concentration (25 and 50 mM), temperature (45 and 55 °C) and mass of biocatalyst (0.25 and 0.5 g). The Ping Pong Bi Bi kinetic model was fitted to data selected from the above-cited assays, assuming the production of only monoesters (relating the consumption of sugar and acid by an equimolar relation). The values for V max , K sugar and K acid were estimated. When CALB-IM-T2-350 was used, at 55 °C, V max  = 1133 ± 221 mmol/Lh, K sugar  = 1378 ± 1696 mmol/L and K acid  = 3298 ± 1015 mmol/L. In the syntheses with PFL-octyl-silica, the results indicated lower conversions, with V max varying between 124 ± 13 and 221 ± 34 mmol/Lh (depending on reaction conditions) and K  = 10,241 ± 1526 mmol/L (in average). A very good fit of the proposed model to the experimental data was obtained. For validation purpose, a different set of experimental data was used. The validation concentration profile showed an excellent prediction capability of the model.