Optimized microwave-assisted azadirachtin extraction using response surface methodology

The neem tree (Azadirachta indica A. Juss) is grown mainly for shade, fuel, and numerous nontimber forest products using its leaves, fruit, and bark. It produces an essential oil that is used as a source for obtaining bioinsecticides, with a broad spectrum of action in agricultural production. Its bioinsecticidal activity is due to the presence of triterpenes, such as azadirachtin, a product in continued growth of the global biopesticide market. Optimal conditions for neem oil extraction using response surface methodology (RSM) and microwave-assisted extraction (MAE) methods have been defined. However, the extraction conditions for these methods tend to consume high volumes of organic solvent and long extraction times. The aim of the present study is to determine the optimal conditions for the extraction of azadirachtin from neem seeds in a hydroalcoholic medium using MAE and RSM with a Box-Behnken design (BBD). A BBD was applied to evaluate the effects of the factors, magnetron voltage (X-1), extraction time (X-2), and pH of the extraction medium (X-3), on the yield of the azadirachtin extraction process. The effect of each variable on the extraction yield was studied independently, considering the pure coefficients (linear and quadratic) on the three levels that were studied in the experiments. Moreover, the study experiments were conducted in triplicate, data were presented as mean and standard deviation, homogeneity of variances was estimated using Levene's test, and a two-way ANOVA with Tukey's post hoc analysis was performed to identify the experimental conditions that allowed us to find the highest extraction yield and to analyze whether the response surface model adequately described our data. The most significant effects of the model correspond to quadratic and interaction effects (p < 0.0001); the quadratic terms voltage (X-1), extraction time (X-2), and pH (X-3); and the interaction effects between voltage-pH (X-1*X-3) and time-pH (X-2*X-3), which had a significant influence on the model. Moreover, a canonical analysis was performed. The optimal conditions were as follows: 69.22 V, 6.89 min, and a pH value of 4.35, coinciding with the zones shown in the contour plots. Furthermore, the response obtained at the optimal conditions was 37.5 mu g of azadirachtin per gram of pretreated seed.