熵权TOPSIS法结合多指标成分综合评价金钱草药材的质量 严玉晶 崔婷 丁青 洪婉敏 李秀枝 孙冬梅摘 要 目的：建立金錢草药材的超高效液相色谱（UPLC）特征指纹图谱，同时测定其中3种有效成分的含量，并综合评价不同产地金钱草的整体质量方法：采用UPLC法建立10批金钱草全草、茎及叶的特征指纹图谱并测定山柰酚-3-O-芸香糖苷、槲皮素、山柰素的含量色谱柱为Waters CORTECS UPLC T3，流动相为乙腈-0.2%磷酸水溶液（梯度洗脱），检测波长为364 nm，柱温为30 ℃，流速为0.2 mL/min，进样量为1 ?L采用《中药色谱特征图谱相似度评价系统（2012版）》进行相似度评价，确定共有峰;以山柰酚-3-O-芸香糖苷、槲皮素、山柰素、总灰分、酸不溶性灰分、二氧化硫残留量及醇溶性浸出物为指标，采用熵权优劣解距离法（TOPSIS）对金钱草药材的整体质量进行综合评价结果：10批金钱草全草、茎及叶共有7个共有峰，指认了其中3个峰分别为山柰酚-3-O-芸香糖苷、槲皮素、山柰素同一部位不同批次金钱草全草的相似度均不低于0.830;同一批次金钱草茎与叶的相似度为0.504～0.859，全草与茎的相似度为0.593～0.904，全草与叶的相似度为0.885～0.995。
山柰酚-3-O-芸香糖苷、槲皮素、山柰素检测质量浓度的线性范围分别为0.392 0～39.197 0、0.397 0～39.703 4、0.380 9～38.093 0 μg/mL（r均大于0.999 0）;精密度、稳定性、重复性试验的RSD均小于2%;加样回收率分别为96.43%（RSD=0.63%，n=9）、100.32%（RSD=0.46%，n=9）、101.80%（RSD=0.32%，n=9）金钱草中上述3种成分的含量分别为0.006 3%～0.041 1%、0.002 9%～0.008 6%、0.004 4%～0.017 5%（茎）， 0.024 8%～0.290 5%、0.000 9%～0.009 0%、0.001 3%～0.012 4%（叶），0.007 9%～0.118 0%、0.001 5%～0.008 8%、0.002 8%～0.012 5%（全草）不同产地样品间3种成分含量比较，差异均无统计学意义（P>0.05）;金钱草不同部位中山柰酚-3-O-芸香糖的含量大小顺序依次为叶>全草>茎，槲皮素与山柰素的含量大多以茎中含量较高熵权TOPSIS法结果显示，四川省中江县、双流县以及重庆市石柱县产金钱草的最优解欧氏贴近度均值分别为0.446、0.512、0.287。
结论：所建特征指纹图谱和含量测定方法稳定、可行，结合熵权TOPSIS法所建的多指标评价模型可用于金钱草药材质量的综合评价;四川省产金钱草药材质量较优关键词 金钱草;不同部位;超高效液相色谱法;含量测定;特征指纹图谱;熵权优劣解距离法ABSTRACT OBJECTIVE： To establish UPLC characteristics fingerprints of Lysimachia christinae， and to simultaneously determine 3 effective components and to comprehensively evaluate the quality of L. christinae from different production areas. METHODS： UPLC method was adopted to establish characteristics fingerprint of the whole plant， stem and leaves of 10 batches of L. christinae， and determine the contents of kaemperfol-3-O-rutinoside， quercetin， kaemperfol. The determination was performed on Waters CORTECS UPLC T3 column with mobile phase consisted of acetonitrile-0.2% phosphoric acid aqueous solution （gradient elution） at the flow rate of 0.2 mL/min. The detection wavelength was set at 364 nm， and column temperature was 30 ℃. The sample size was 1 ?L. Similarity Evaluation System for TCM Chromatographic Fingerprint（2012 edition） was adopted to evaluate its similarity， and common peaks were confirmed. Using the contents of kaemperfol-3-O-rutinoside， quercetin， kaemperfol， total ash， acid-insoluble ash and sulfur dioxide residue， the ethanol-soluble extract as index， entropy weight TOPSIS was used to evaluate the overall quality of L. christinae comprehensively. RESULTS： There were 7 common peaks in the whole plant， stem and leaves of 10 batches of L. christinae， among which 3 peaks were identified as kaemperfol-3-O-rutinoside， quercetin and kaemperfol. The similarity of same part in the whole plant of L. christinae from different batches were not lower than 0.830. The similarity between stem and leaves of L. christinae in same batch was 0.504-0.859; the similarity between whole plant and stem was 0.593-0.904; the similarity between whole plant and leaves was 0.885-0.995. The linear ranges were 0.392 0-39.197 0 μg/mL for kaempferol-3-O-rutinoside， 0.397 0- 39.703 4 μg/mL for quercetin， 0.380 9-38.093 0 μg/mL for kaempferol （r>0.999 0）. RSDs of precision， stability and repeatability tests were all lower than 2%. The recoveries were 96.43% （RSD=0.63%， n=9）， 100.32% （RSD=0.46%， n=9）， 101.80% （RSD=0.32%， n=9） ，respectively. The content range of above components in L. christinae were 0.006 3%-0.041 1%， 0.002 9%-0.008 6%， 0.004 4%-0.017 5%（stem）; 0.024 8%-0.290 5%， 0.000 9%-0.009 0%， 0.001 3%-0.012 4%（leaves）; 0.007 9%-0.118 0%， 0.001 5%-0.008 8%， 0.002 8%-0.012 5%（whole plant）. There was no significant difference in the contents of 3 components in L. christinae among different producing areas （P>0.05）. The order of the contents of kaempferol-3-O- rutinoside in different parts of L. christinae was leaves>whole plant>stem. The contents of quercetin and kaempferol were high relatively in the stem. Results of entropy weight TOPSIS method showed that mean values of Ci for L. christinae from Zhongjiang county and Shuangliu county of Sichuan province， Shizhu county of Chongqing city were 0.446，0.512，0.287. CONCLUSIONS： Established fingerprint and content determination method are stable and feasible， and multi-index evaluation model constructed by characteristic chromatogram combined with entropy weight TOPSIS analysis method can be used for comprehensive quality evaluation of L. christinae. The quality of L. christinae from Sichuan province is better.2.1.5 稳定性试验 取“2.1.3”项下供试品溶液（编号：S1）适量，分别于室温下放置0、2、4、8、12、24 h时按“2.1.1”项下色谱条件进样测定，以槲皮素为参照，计算各共有峰的相对保留时间和相对峰面积。
结果，7个共有峰相对保留时间的RSD≤0.48%（n=6），相对峰面积的RSD≤1.14%（n=6），表明供试品溶液于室温下放置24 h内稳定性良好2.1.6 重复性试验 取金钱草样品粉末（编号：S1）约1.0 g，精密称定，共6份，按“2.1.3”项下方法制备供试品溶液，再按“2.1.1”项下色谱条件进样测定，以槲皮素为参照，计算各共有峰的相对保留时间和相对峰面积结果，7个共有峰相对保留时间的RSD≤0.49%（n=6），相对峰面积的RSD≤1.98%（n=6），表明方法重复性良好2.1.7 特征指纹图谱的建立 取金钱草全草及不同部位样品各10批，按“2.1.3”项下方法制备供试品溶液，再按“2.1.1”项下色谱条件进样测定，记录色谱图，采用《中药色谱指纹图谱相似度评价系统（2012版）》建立特征指纹图谱分别以S1、J1、Y1样品色谱图为参照图谱，设置时间窗宽度为0.1 min，通过多点校正、全谱峰匹配分别生成共有特征指纹图谱，同时采用中位数法分别生成对。
