RSM and ANN modeling-based optimization approach for the development of alginate-chitosan particles containing Rosmarinus officinalis essential oil
摘要
Alginate is a multifunctional biopolymer employed in diverse applications, particularly in delivery systems that safeguard bioactives in acidic environments. Chitosan demonstrates various functionalities, including enhanced mucoadhesion and the facilitation of controlled release in alkaline conditions. Current research focuses on encapsulating essential oils in alginate and chitosan. The study aimed to encapsulate Rosmarinus officinalis essential oil within the alginate-chitosan sphere to enhance its physical properties and investigate the protective effects of alginate-chitosan Rosmarinus essential oil particles during sperm cryopreservation. The effects of factors influencing cell motility were optimized and modeled using response surface methodology and artificial neural networks. Results indicated a high goodness of fit for the artificial neural network model with an R2 of 0.999, compared to 0.991 for the response surface methodology. The optimal solution involved 2 g of alginate, 1.05 g of chitosan, and 80 μl of Rosmarinus essential oil. The optimal formulation was characterized by assessing drug entrapment efficiency, drug-excipient interaction, swelling index, and lipid peroxidation state. At pH 6.8, alginate-chitosan-Rosmarinus essential oil particles exhibited a significant swelling index. These findings suggest that the formation of alginate-chitosan through cross-linking holds promise for solubilizing and delivering Rosmarinus essential oil to sperm under cryopreservation conditions. Adding alginate-chitosan-Rosmarinus essential oil complex to the extender enhances its ability to remove free radicals and thus protects sperm during the cryopreservation process. A methodological approach utilizing either response surface methodology or artificial neural networks can establish alginate-chitosan encapsulation effectively. Importantly, the artificial neural network model outperformed response surface methodology in prediction and optimization accuracy.