Abstract:

Ethanol aqueous solution was used to extract total flavonoids from Artemisia lavandulaefolia under ultrasonic assistance. To optimize extraction yield of total flavonoids, 30 experiments resulting from a five-variable and four-level central composite design were first carried out; subsequently, experimental data obtained were subjected to multiple variable regression fitting to establish a mathematical model for extraction yield of total flavonoids as a response to 5 variables and a significance test was performed for the model; and finally response surface analysis was conducted for investigating the interaction effects of 5 variables on extraction yield of total flavonoids. Results showed that the developed model could describe the relationship between extraction yield of total flavonoids and 5 variables such as ultrasonic treatment time, extraction temperature, ethanol concentration in aqueous solution and solid/liquid ratio very well. The optimal values of the variables were as follows: ultrasonic treatment time 36.7 min, extraction temperature 74.99 ℃, ethanol concentration in aqueous solution 71.78%, and solid/liquid ratio 1:65.68 (g/mL). Under such conditions, the predicted value of extraction yield of total flavonoids was 2.8%, which was very consistent with the experimental value with a relative error of 1.1%.

Key words: Artemisia lavandulaefolia, total flavonoids, response surface analysis methodology, ultrasonic-assisted extraction