Synthesis of Prebiotic Caramels Catalyzed by Ion-Exchange Resin Particles: Kinetic Model for the Formation of Di-d-fructose Dianhydrides.

Caramel enriched in di-d-fructose dianhydrides (DFAs, a family of prebiotic cyclic fructodisaccharides) is a functional food with beneficial properties for health. The aim of this work was to study the conversion of fructose into DFAs catalyzed by acid ion-exchange resin, in order to establish a simplified mechanism of the caramelization reaction and a kinetic model for DFA formation. Batch reactor experiments were carried out in a 250 mL spherical glass flask and afforded up to 50% DFA yields. The mechanism proposed entails order 2 reactions that describe fructose conversion on DFAs or formation of byproducts such as HMF or melanoidines. A third order 1 reaction defines DFA transformation into fructosyl-DFAs or fructo-oligosaccharides. The influence of fructose concentration, resin loading and temperature was studied to calculate the kinetic parameters necessary to scale up the process.

Purification of sugar beet vinasse - adsorption of polyphenolic and dark colored compounds on different commercial activated carbons.

The adsorption on activated carbons of dark colored compounds contained in sugar beet vinasse was studied. Four commercial activated carbons with different properties (particle size, residual acidity and microporous properties) were respectively checked for efficiency at two temperature levels (25 degrees C and 40 degrees C) and at four pH levels (2,3.5,7,10). The adsorption of organic molecules was determined by quantifying the amounts of total polyphenolic compounds and total organic carbon. The results showed that the adsorption capacity of dark colored compounds was enhanced by the decrease in both temperature and pH values of the solution. In this study, it is shown that this capacity depends on activated carbon characteristics which can be classified in the following order: particle size>residual acidity>microporous volume. Three models (Langmuir, Freundlich and Dubinin-Radushkevich) were tested from experimental data and compared. The Langmuir model provided the best correlation on all the activated carbons studied.