Defatted wheat germ as source of polyphenols-Optimization of microwave-assisted extraction by RSM and ANN approach

Wheat germ was firstly defatted using supercritical CO2 extraction and raffinate was further utilized for polyphenols isolation. For that purpose solid-liquid, ultrasound-assisted and microwave-assisted extraction (MAE) techniques were applied. MAEs were performed at different levels of ethanol concentrations, extraction time, liquid-solid ratio and irradiation power, while ethanol concentration was only variable for other extraction techniques. Obtained extracts were evaluated in terms of total extraction yield, total phenols yield and in vitro antioxidant activity, whereas under certain conditions MAE exhibited the highest response values. MAE was analyzed by response surface methodology (RSM) and artificial neural network (ANN) techniques which were compared in terms of fitting quality, influence analysis and optimization. Results indicated that RSM and ANN could be successfully used to describe experimental data. Both methods exhibited certain advantages, i.e. ANN showed minor advantage of fitting quality, while RSM provided further insight in influence analysis. Optimized MAE conditions determined by RSM were ethanol concentration of 58.03%, extraction time of 14.48 min, liquidsolid ratio of 25 mL/g and irradiation power of 800 W. MAE conditions optimized by ANN were different in terms of ethanol concentration (50%), extraction time (15 min) and irradiation power (400 W), and same in terms of liquid-solid ratio.