Preparing Allicin Nanocapsules and Determining the Factors Controlling Their Particle Size through Artificial Intelligence

Abstract

Abstract Allicin nanocapsules were prepared via ionotropic pre-gelation. The wall materials were alginate-chitosan biopolymers. Nanocapsules were characterized using Fourier transform infrared spectroscopy (FT-IR) and field emission scanning electron microscopy (FE-SEM). We tried to simulate the effects of three different variables on particle size through artificial intelligence approaches. Feedforward neural networks (FFNN) and adaptive neuro-fuzzy inference system (ANFIS) were employed to model the size of allicin nanocapsules, and the latter was found to be relatively more successful in this regard. Finally, genetic algorithms were employed to determine the optimal values for the variables at which the smallest particles were formed.