Estudo da cinética de secagem de esferas de alginato considerando encolhimento

Abstract

Sodium alginate is a biopolymer that has being studied since the 2000’s, having lots of applications in the biomedical and food areas. Alginate can be used to fabric spherical microcapsules of different sizes which can contain oils or extracts. It’s important to know how the moisture of the material variates with time, thus it’s possible to plan stocking and transport procedures, in addition to have more application of materials whose diffuse themselves to outside of the capsules. In this work spheres of alginate were generated by the method of ionic gelation, forming spherical capsules whose were submitted to a drying process to study its drying kinetics. Experimental data of moisture and size of the material were collected. A phenomenological model native of a mass balance in transient state was proposed to describe the moisture decreasing and its shrinkage. In the model it was necessary to know how the radius variate with time. This variation was described by two different empirical equations, Page equation (Modelo RP) and the sigmoidal curve equation (Modelo RS). A calculus routine was developed in MATLAB to adjust the curves in accord to the experimental data, generating parameters that minimize the quadratic difference between the calculated and the experimental data (objective function). Both models were able to represent the phenomena, having a satisfactory performance in the challenge that was describe the variation of radius and moisture simultaneously. The parameters behavior was evaluated with temperature, revealing that the decreasing rate of radius and moisture grow exponentially with temperature. A statistical analysis was made by observing the histograms and graphics of the quantis of the residues of both models, followed by the Shapiro-Wilk test to verify the hypothesis that the residues have a normal distribution, with 95% of confidence, around zero. The hypothesis to both models was rejected 50% of the time, revealing that in half of the cases the residues presented the normal behavior and in the other 50% they didn’t. The Akaike’s test was also used in this study, as a tool of comparison between the models obtained. The test revealed that the Modelo RP was the model with the better fit and presented the highest probability of being the fittest model to describe the experimental data, revealing that the possession of an additional parameter, in this case, was profitable.