Desarrollo de un modelo predictivo para la obtención de Maltodextrinas del Almidón de Yuca.

The objective of this work was to obtain a predictive model of the production of maltodextrins, measured as dextrose equivalents (ED), in terms of four factors: enzyme concentration, substrate concentration, temperature and time of digestion whose minimum and maximum levels were 20 and 30% m/m, 0,01 and 0,03%, 40 and 60°C, 15 and 30 minutes respectively. The substrate was obtained from the pulp of cassava tubers (Manihot esculenta Crantz), which was characterized and used as raw material to produce starch solutions at 10, 20 and 30% m/m, which were hydrolysed with commercial alpha amylase (300 NKU/g) to produce the maltodextrins. A factorial design 24 was used, and by means of ANOVA and the software Minitab 17 it was determined that all the factors and their interactions, except the interaction temperature-concentration of substrate, were statistically significant for a α = 0,05. Mathematical expression was simplified by considering the factors and their most effective interactions; these being the Temperature, Enzyme Concentration, Time, Substrate Concentration, and the Interaction Temperature-Concentration of enzyme; thus the mathematical model remained as: ED=0,454375+0,128781 x temperatura-0,060186 x [sustrato]-10,249998 x [enzima] +0,065625 x tiempo +5,753121 x temperatura x [enzima] with a deviation s = 0,197, an R² = 99,78% and an average error rate of 1,18%. The model was validated through a residual analysis; as well as using published works, finding good reproducibility in two cases, even when the substrate was not yucca but yam. [ABSTRACT FROM AUTHOR]