Condition Optimization on Microwave-Assisted Extraction of Determination of 7 Chlorogenic Acids and Their Analogues in Herbal Tea Granules by High Performance Liquid Chromatography Based on Response Surface Methodology

When using high performance liquid chromatography for determination, the total content of 7 chlorogenic acids and their analogues in the herbal tea granule sample was used as the response values, and microwave-assisted extraction conditions, including the volume fraction of the methanol solution (extraction solvent, factor A), extraction time (factor B), extraction temperature (factor C), and solid-liquid ratio (ratio of sample amount to extraction solvent volume, g:mL, factor D) were optimized through single factor method and response surface methodology. The 20 mL of 70% (volume fraction) methanol solution was added into 1.00 g of herbal tea granule sample for dissolution for 1 min by vortex. The mixture was placed into a microwave extractor, and extraction process was made at 80 ℃ for 2.0 min. The extract was passed through a 0.22 μm filter membrane, and the filtrate was determined according to chromatographic conditions. In chromatographic analysis, ZORBAX Eclipse Plus C₁₈ chromatographic column was used as the stationary phase, and mixed solutions composed of 0.1% (volume fraction) phosphoric acid solution and methanol at different volume ratios were used as the mobile phase for gradient elution separation. A diode array detector was used for detection at the wavelength of 325 nm. It was shown that the quadratic polynomial regression model was established through four factors and three levels experiment with the Box-Behnken central combination design. A² and B² had the significant effect on the total content of 7 targets, while AD and D² had the significant effect. The effect of interaction factors was consistent with those obtained from contour lines and response surfaces. Under optimized microwave assisted extraction conditions, the total determined value of the 7 targets were 32.98 mg·kg^-1, which was 80.3% of the predicted value (41.10 mg·kg^-1), indicating that the regression model had high reliability. Linear relationships between values of the mass concentration and the peak area of 7 targets were kept in the range of 0.01-4.0 mg·L^-1, with detection limits (3S/N) in the range of 0.12-0.60 mg·kg^-1. Test for recovery was made according to the standard addition method, giving recoveries of 7 targets in the range of 97.5%～105%, and RSDs (n=6) of the determined values ranged from 1.4% to 6.9%.