A process optimization study on ultrasonic extraction of paclitaxel from Taxus cuspidata

This study aimed to improve the extraction rate of paclitaxel from Taxus cuspidata in order to determine the most effective combination of ultrasonic extraction and thin-layer chromatography–ultraviolet (TLC-UV) rapid separation method. The study was performed using the Box–Behnken test design to conduct single-factor experiments using ultrasonic extraction of paclitaxel from Taxus cuspidata. The study showed ethanol to be the best extraction solvent. When mixed with dichloromethane (1:1), the ratio of material to liquid was 1:50 when using an ultrasonic time of 1 hr at a power of 200 W. The correction coefficient K for the separation and detection of paclitaxel using the TLC-UV spectrophotometric method was 0.009152. Multifactor experiments determined the effect of the rate of liquid to material (X1), ultrasonic time (X2), and ultrasonic power (X3) on extraction using extraction volume as the dependent variable. Response surface analysis allowed a regression equation to be obtained, with the optimal conditions for extraction when the rate of liquid to material was 53.23 mL/g as an ultrasonic time of 1.11 hr and an ultrasonic power of 207.88 W. Using these parameters, the average amount of extracted paclitaxel was about 130.576 µg/g, which was significantly better than for other extraction methods.