目的 建立UPLC-MS/MS法同时测定荆防颗粒中15种活性成分的含量，并综合评价其质量。方法 采用Poroshell 120 SB-C₁₈色谱柱（100 mm×2.1 mm，2.7 μm）；流动相为水（含0.1%甲酸和5 mmol/L乙酸铵）-乙腈，梯度洗脱：正离子模式，0～2.0 min，12%～25%乙腈；2.0～5.0 min，25%～50%乙腈；5.0～7.0 min，50%～90%乙腈；7.0～9.0 min，90%乙腈；9.0～9.1 min，90%～5%乙腈；9.1～11.0 min，5%乙腈；负离子模式，0～2.0 min，10%～15%乙腈；2.0～6.0 min，15%～30%乙腈；6.0～8.0 min，30%～90%乙腈；8.0～10.0 min，90%乙腈；10.0～10.2 min，90%～10%乙腈；10.2～13.0 min，10%乙腈；体积流量0.3 mL/min；柱温25 ℃；进样量3 μL。电喷雾离子源（ESI）；正、负离子扫描；多反应监测（multiple reaction monitoring，MRM）模式；雾化器压力103.42 kPa（15.0 psi）；毛细管电压-3 500 V（负离子）、4 000 V（正离子）；干燥气温度300 ℃；体积流量11 L/min。采用SPSS 27.0与SIMCA 14.1软件对含量测定结果进行化学模式识别分析，并利用熵权TOPSIS法计算8个厂家共24批荆防颗粒的综合质量排名。结果 15种成分在各自范围内具有良好的线性关系（r≥0.999 0），平均加样回收率在99.30%～100.97%，RSD在1.53%～2.71%。层次聚类分析（hierarchical cluster analysis，HCA）和主成分分析（principal component analysis，PCA）结果表明，荆防颗粒样品聚为3类，4个主成分累积方差贡献率达到82.180%。变量重要性投影（variable importance projection，VIP）结果表明，柴胡皂苷A、橙皮苷、紫花前胡苷、柚皮苷、甘草苷、佛手柑内酯及蛇床子素是影响荆防颗粒质量差异的主要差异成分。熵权TOPSIS法可用于不同批次荆防颗粒质量优劣的评价，其中Q4厂家与Q5厂家整体表现较优；Q6～Q8厂家综合得分多处于中游位置；Q1～Q3厂家综合得分普遍较低。结论 该方法准确、灵敏，结合化学计量学及熵权TOPSIS法可用于荆防颗粒的质量综合评价。

Objective To establish a UPLC-MS/MS method for simultaneously determining the contents of 15 active components in Jingfang Granules (JG, 荆防颗粒), and conduct a comprehensive evaluation of their quality. Methods A Poroshell 120 SB-C₁₈ column (100 mm×2.1 mm, 2.7 μm) was used; the mobile phase was water (0.1% formic acid-5 mmol/L ammonium acetate aqueous solution)-acetonitrile, with gradient elution as follows: in positive ion mode, 0—2.0 min, 12%—25% acetonitrile; 2.0—5.0 min, 25%—50% acetonitrile; 5.0—7.0 min, 50%—90% acetonitrile; 7.0—9.0 min, 90% acetonitrile; 9.0—9.1 min, 90%—5% acetonitrile; 9.1—11.0 min, 5% acetonitrile; negative ion mode, 0—2.0 min, 10%—15% acetonitrile; 2.0—6.0 min, 15%—30% acetonitrile; 6.0—8.0 min, 30%—90% acetonitrile; 8.0—10.0 min, 90% acetonitrile; 10.0—10.2 min, 90%—10% acetonitrile; 10.2—13.0 min, 10% acetonitrile; flow rate 0.3 mL/min; column temperature 25 ℃; injection volume 3 μL. Electrospray ionisation (ESI); positive and negative ion scanning; multiple reaction monitoring (MRM) mode; nebuliser pressure 103.42 kPa (15.0 psi); capillary voltage -3 500 V (negativeions), 4 000 V (positiveions); dry gas temperature 300 ℃; flow rate 11 L/min. Chemical pattern recognition analysis of the content determination results was performed using SPSS 27.0 and SIMCA 14.1 software, and the entropy-weighted TOPSIS method was employed to calculate the comprehensive quality rankings of 24 batches of JG from eight manufacturers. Results All 15 components show a good linear relationship within their ranges (r ≥ 0.999 0), with average sample recoveries between 99.30% and 100.97%, and RSD from 1.53% to 2.71%. The results of hierarchical cluster analysis (HCA) and principal component analysis (PCA) showed that the JG samples were clustered into three categories, with the four principal components accounting for 82.180% of the cumulative variance. Variable importance projection (VIP) results indicated that saikosaponin A, hesperidin, nodakenin, naringin, liquiritin, bergapten, and osthol were the primary components influencing quality variations in JG. The entropy-weighted TOPSIS method can be used to evaluate the quality of JG across different batches, with manufacturers Q4 and Q5 demonstrating superior overall performance; manufacturers Q6 to Q8 generally achieving mid-range composite scores; and manufacturers Q1 to Q3 generally scoring lower. Conclusion This method is accurate and sensitive, and combined with chemometrics and the entropy weight TOPSIS method, it can be used for the comprehensive quality evaluation of JG.

R283.6

国家药品监督管理局中药材及饮片质量控制重点实验室项目资助（2024GSMPA-KL/5）；2021年度甘肃高等学校产业支撑计划项目资助（2021CYZC-40）；兰州市人才创新创业项目资助（2023-RC-9）；兰州市科技计划项目（2025-5-255）