Modeling the Encapsulation of Thymus Essential Oil (Thymus vulgaris) in Sodium Caseinate, Maltodextrin and Modified Starch Using Response Surface (RSM) and Artificial Neural Network (ANN)

Microencapsulation process is down in protecting volatile and sensitive compounds to chemical reactions. Process modeling can be effective in assessing and predicting the impact of conditions affecting the qualitative properties of the product. In order to investigated the effect of wall material concentration, protein/polysaccharide ratio and ultrasound waves time on the microencapsulation efficiency, total phenolics, antioxidant capacity and volatile compounds, nano-emulsion of Thymus vulgaris essential oil in aqueous phase containing protein (sodium caseinate) and polysaccharide (maltodextrin and modified starch) was prepared by ultrasound waves. Modeling of dependent variables was done by response surface methodology and artificial neural network. The results showed that increasing the wall concentration and protein/polysaccharide ratio improved the retention of volatile compounds and antioxidant capacity of the encapsulated essential oil, and at higher values of variables the same trend was observed for the phenolic content. Increasing the protein/polysaccharide ratio and ultrasound waves resulted in increased the microencapsulation efficiency. Models obtained from the interaction effects of neural network and Response Surface Method at higher values resulted in better preservation of total phenolic, antioxidant activity and increased of Microencapsulation efficiency of microcapsules. Microencapsulation produced containing of 20% wall material, 31% protein/polysaccharide ratio and 97s ultrasound time were the best treatment in terms of the microencapsulation efficiency and volatile compounds.Experimental results of the experiments indicative appropriate and uniform function of both response surface and neural network methods and the superiority of the artificial neural network in predicting the desired variables.