Nonlinear Regression Model for Prediction of Dispersive Liquid-Liquid Microextraction-Solidification of Floating Organic Drop (DLLME-SFO) of DecaBDE from River Sediments Based on Undecanol

The impacts of several factors (extractant volume, dispersant volume, extractant type, dispersant type, pH, salt effect, extraction time and grain size) on dispersive liquid-liquid microextraction-solidification of floating organic drop of DecaBDE from river sediments based on undecanol were investigated by an interactive orthogonal experimental array design, and the significant factors were obtained though the range analysis. A nonlinear regression model for prediction of dispersive liquid-liquid microextraction-solidification of floating organic drop of DecaBDE was established based on the results of the range analysis with four significant factors: grain size, dispersant type, pH, and salt effect. The determination coefficient of the model was: R² = 0.9978 (F>F_0.01), which indicated that the nonlinear regression model could be employed to predict the extraction recovery of DecaBDE. Several groups of extraction recovery of dispersive liquid-liquid microextraction-solidification of floating organic drop of DecaBDE were predicted by the nonlinear model, and one of the predicted results showed that the extraction recovery of DecaBDE could be up to 99.0% under the conditions (undecanol of extractant, 80 μL of extractant volume, acetone of dispersant, 1.0 mL of dispersant volume, pH = 6.0, 7.0% of salt effect, 10 min of extraction time, 0.075 mm of grain size). Moreover, the relative deviation of 3.52% was calculated between the predicted and verified values. On the contrary, the experimental conditions could be also optimized through the model at given values of extraction recoveries using Lingo software.