Ohmic assisted hydrodistillation extraction of musk willow (Salix aegyptiaca L.) essential oil and artificial neural network modeling of extraction kinetic

In the current research, the effects of different ohmic assisted hydrodistillation (OAHD) parameters including voltages gradient (5, 15 and 25 v/cm), extraction time (30, 75 and 120 min) and NaCl concentrations (0.5, 1 and 1.5%) on the extraction yield, energy consumption and total phenol content (TPC) of Salix aegyptiaca L.essential oil were investigated, and then compared with conventional hydrodistillation (HD). Finally, artificial neural network (ANN) modeling is utilized to predict kinetics of essential oil extraction. Result revealed that extraction time and voltage gradient had significant effect on extraction yield, energy consumption and TPC (p<0.05). Extraction yields of essential oil obtained by OAHD and HD were 0.119 ± 0.012 and 0.081 ± 0.01, respectively. There was no significant difference (p>0.05) between specific gravity, refractive index and TPC of essential oil of OAHD and HD methods, however IC50 of essential oil extracted by OAHD was significantly higher than essential oil obtained with HD method (p<0.05). To design the ANN model, voltages gradient, extraction time and salt concentrations and their interactions were considered as input vectors while the extraction yield of essential oil was considered as the model output. The results showed that the best prediction performance belonged to 3-9-8-1 ANN architecture (RMSE=0.036 and R2=0.99). Therefore, it can be concluded that the OAHD method is applicable for S. aegyptiaca L. essential oil extraction and ANN model is an efficient quantitative tool to predict the kinetics of essential oil extraction.