Evaluation of Antioxidant Properties of Brown Algae (Sargassum Glaucescens) Extract and Optimization of Extraction of Its Antioxidant Compounds with Artificial Neural Network

Marine algae is one of the best sources of bioactive substances. In the present study, optimization of effective factors (time, temperature and solvent concentration) on the extraction of antioxidant compounds of extract of brown algae Sargassum glaucescens in Chabahar beaches with the artificial neural network was conducted.

Sargassum algae was extracted with methanol at three times (1, 3 and 5 hours), three temperatures (24, 47 and 70 ° C) and three concentrations (60, 80 and 100%). DPPH free radical scavenging activity method was used to evaluate its antioxidant properties. Total phenolic compounds were measured according to the standard method. Artificial neural networks (multilayer perceptron with hyperbolic tangent function) with 1 latent layer and 5 neurons were used to optimize the extraction of antioxidant compounds from sargassum algae extract.

The amount of total phenolic compounds in the extract of Sargasom algae was 3.2±0.56 mg Galic acid/g extract. The artificial neural network predicted the antioxidant properties of sargassum algae with a coefficient of determination of R2=0.9439 and Root mean square error (RMSE) of 1.465654. A positive and significant correlation was observed between observed and predicted antioxidant properties. The optimal extraction conditions of antioxidant compounds based on the artificial neural network method were 100% concentration, temperature 70 ° C and time 5 hours.

According to the present findings, the artificial neural network has good potential to predict the optimal extraction conditions of sargassum algae to determine the antioxidant properties using a multilayer perceptron with hyperbolic tangent. High concentration of total phenolic compounds in sargassum algae extract may be the reason for its antioxidant properties.