Complexation of cress seed mucilage and β-lactoglobulin; optimization through response surface methodology and adaptive neuro-fuzzy inference system (ANFIS)

Abstract

Complex coacervation is a common technique that enables the encapsulation of nutraceuticals/bioactive compounds and the modification of biopolymeric properties. Here, the complex coacervation occurred between cress seed mucilage (CSM) and β-Lactoglobulin (Blg) and the results were systematically evaluated under the influence of polymer concentrations (CSM & Blg), pH and Blg: CSM mixing ratio. In this regard, thirty-six experiments were performed, including particle size and zeta potential measurements to find the optimum reaction conditions. Then, Response Surface Methodology (RSM) and Adaptive Neuro Fuzzy Inference System (ANFIS) were applied to predict model-effects and to reach optimization in experimental variables on the size and zeta potential of CSM-Blg complexes. The correlation coefficient of RSM (size: 0.984, zeta potential: 0.975) and ANFIS models (size: 0.998, zeta potential: 0.997) were acceptable in the case of both responses. They also indicated a higher correlation of the predicted values with the experimental data by ANFIS. According to RSM results, CSM concentration and pH were the most important variables in determining both size and zeta potential of complexes. In addition, the best conditions (CSM ​= ​0.1 %w/v, Blg ​= ​0.066 %w/v, pH ​= ​5.0 and mixing ratio ​= ​1.25) were calculated by ANFIS for the formation of the smallest and most stable CSM-Blg complexes. The complexity and nonlinearity of the relationships between the input and output variables further revealed the ability of intelligent models such as ANFIS in predicting and controlling the complex properties. This ability of the model was consistent with experimental inferences.