摘要
提出了一种简化数据处理过程的多波长自动融合指纹图谱策略,并应用于蜂胶乙醇提取物的质量评价。运用超高效液相色谱(UPLC)技术对5个厂家的15批蜂胶乙醇提取物进行220、245、270、320 nm的多波长检测。样品采用Agilent Poroshell 120 EC-C_(18)(3 mm×100 mm,2.7μm)色谱柱进行分离;以甲醇-0.2%磷酸水溶液为流动相进行梯度洗脱。利用“中药色谱指纹图谱相似度评价系统”进行全峰匹配,实现了每个样品多个波长下各色谱峰数据的自动融合。以270 nm单波长为对照,比较其与多波长融合在共有色谱峰、相似度评价、聚类分析、主成分分析方面的差异。蜂胶多波长融合指纹图谱包含33个特征峰,其中14个特征峰可定性,分别为咖啡酸、ρ-香豆酸、阿魏酸、异阿魏酸、3,4-二甲氧基肉桂酸、肉桂酸、槲皮素、山奈素、芹菜素、异鼠李素、乔松素、白杨素、高良姜素、咖啡酸苯乙酯。各批次特征峰保留时间的相对标准偏差(RSD)均小于1.00%,峰面积RSD为14.39%~112.02%,平均值为37.94%。15批蜂胶融合指纹图谱的相似度为0.902~0.994,主成分分析、聚类分析均可将样品分为4类。蜂胶270 nm单波长的指纹图谱包含32个特征峰,12个特征峰可定性。15批蜂胶单波长指纹图谱的相似度为0.928~0.997,主成分分析、聚类分析均可将样品分为3类。相比于270 nm单波长,多波长融合的相似度评价更敏感,主成分分析和聚类分析更准确。该多波长自动融合指纹图谱策略可对多波长数据进行快速、自动处理,丰富不同组分在各波长下的色谱信息,提高蜂胶指纹图谱差异评价的准确性。
A multi-wavelength automatic fusion fingerprint strategy was proposed to simplify the data processing process and applied to the quality evaluation of propolis ethanol extracts.15 batches of propolis ethanol extracts from 5 manufacturers were tested at multiple wavelengths of 220,245,270,320 nm through UPLC technology.The sample was separated on an Agilent Poroshell 120 ECC_(18)(3 mm×100 mm,2.7μm)column by gradient elution,with methanol and 0.2%phosphoric acid aqueous solution as mobile phases.The multi-wavelength fusion fingerprint for each batch was gener⁃ated by using the“Similarity Evaluation System for Traditional Chinese Medicine Chromatographic Fingerprint”to achieve automatic fusion of various chromatographic peaks under multiple wave⁃lengths.The differences between multi-wavelength fusion and 270 nm single wavelength were com⁃pared in common chromatographic peaks,similarity evaluation,cluster analysis,and principal component analysis.The multi-wavelength fusion fingerprint of propolis contained 33 characteristic peaks,in which 14 characteristic peaks have been identified to be caffeic acid,ρ-coumaric acid,fe⁃rulic acid,isoferulic acid,3,4-dimethoxy group cinnamic acid,quercetin,kaempferin,apigen⁃in,isorhamnetin,chocine,chrysin,galangin,caffeic acid,phenethyl ester.The retention time RSDs of characteristic peaks in each batch were less than 1.00%and the peak area RSDs were ranged from 14.39%to 112.02%,with an average value of 37.94%.The similarity of 15 batches of propolis fusion fingerprints ranged from 0.902 to 0.994.The samples could be divided into four categories through the principal component analysis and the clustering analysis.The 270 nm single wavelength fingerprint of propolis contained 32 characteristic peaks,of which 12 have been identified.The simi⁃larity of 270 nm single wavelength fingerprint of 15 batches of propolis ranged from 0.928 to 0.997.The samples were divided into three categories by the principal component analysis and cluster analy⁃sis.Compared to the singl
作者
姜慧洁
章越
慎凯峰
姜艳
周丹英
赵洪静
JIANG Hui-jie;ZHANG Yue;SHEN Kai-fen;JIANG Yan;ZHOU Dan-ying;ZHAO Hong-jing(Zhejiang Traditional Chinese Medicine Research Institute Co.,Ltd.,Hangzhou 310023,China;Center for Food Evaluation,State Administration for Market Regulation,Beijing 100070,China)
出处
《分析测试学报》
CAS
CSCD
北大核心
2023年第12期1588-1597,共10页
Journal of Instrumental Analysis
基金
浙江省属科研院所扶持专项(2022F03355)。
关键词
多波长
自动融合
指纹图谱
蜂胶
质量评价
化学模式识别
multi-wavelength
automatic fusion
fingerprint
propolis
quality evaluation
chemical pattern recognition
