Modeling of the enzymatic synthesis of fructo-oligosaccharides

DRA. ANDREA GOMEZ-ZAVAGLIA; DRA. EMMA ELIZABETH TYMCZYSZYN; DRA. MARIA MICAELA URETA; ING. FAROUX, JUAN MANUEL

Buenos Aires

Congreso; WWCE11 ? 11th Worl Congress of Chemical Engineering; 2023

AAIQ (Asociación Argentina de Ingenieros Químicos. Centro para la seguridad de los Procesos Químicos)

Fructo- and galacto-oligosaccharides (FOS and GOS) are oligosaccharides widely known as prebiotic compounds. In addition, they have relevant nutritional characteristics; they are low caloric sweeteners, with low glycemic index and not cariogenic. For this reason they are incorporated in many products, such as dairy products, beverages, bakery, and infant formula. The production of FOS is a complex process that involves transfructosylation mechanisms; synthesis and hydrolysis occur simultaneously. As a result mixtures of short chain FOS (represented by 1-kestose-DP3, nystose-DP4 and 1-fructofuranosylnystose-DP5), glucose (byproduct) and sucrose (remaining substrate), are obtained. In previous experimental work, the conditions for the enzymatic synthesis of FOS were studied, using sucrose as raw material and a commercial enzyme (Vizcozyme L) as the catalyst. The influence of the initial concentration of sucrose solution (40% and 50% w/vol), temperature, and reaction time (4 to 24 h) was investigated. The experiments were performed in a lab scale fermenter (1 l) and also at pilot plant scale (50 l). The composition of the reaction product was analyzed by high pressure liquid chromatography (HPLC).The aim of this work is the simulation of the whole process of FOS synthesis, that allows performing the scale up from the pilot plant to an industrial scale. As a first step of the simulation, a model of the coupled set of reactions is needed. Based on previous models available in literature [1], a Michaelis-type mechanism of four enzymatic reactions and ten kinetic parameters is proposed, with sucrose (S) as the main reactive, and glucose (G), DP3, DP4 and DP5 as the reaction products. Different approaches, including (or not) substrate inhibition and competitive glucose inhibition were considered in the model. The kinetic parameters of the enzymatic reactions were determined as an inverse problem solution, with the experimental composition of FOS as model input. Since the reactions are described by a set of ordinary differential equations (ODEs), the non-linear least square fitting procedure was coupled to a 4th order Runge-Kutta method solved with a MATLAB routine. Good agreement was found in the fitting procedure which indicates the adequacy of the proposed mechanism. Once the reactions are characterized, the next step is the simulation of the reactor. In this sense, these results are promising, given that the model can be coupled to a simulation software allowing to represent the reactor unit operation with accuracy.