银黄口服液近红外光谱测定和相关性模型建立  

Determination of Yinhuang Oral Liquid by near-infrared spectroscopy and establishment of dependency model

作　　者：张叶霞[1] 潘金火[1] 蔡宝昌[2,3] 秦昆明 刘晓[1,2] 王靥润 严国俊[1] 

机构地区：[1]南京中医药大学药学院,江苏南京210023 [2]南京中医药大学,国家教育部中药炮制规范化及标准化工程研究中心,江苏南京210023 [3]南京海昌中药集团有限公司,江苏南京210061

出　　处：《中成药》2016年第3期565-569,共5页Chinese Traditional Patent Medicine

基　　金：国家自然科学基金青年基金项目(81202922);江苏省科技支撑计划工业项目(BE2012011);江苏省自然科学基金青年基金项目(BK2012089)

摘　　要：目的近红外光谱法测定银黄口服液（金银花、黄芩）中绿原酸和黄芩苷的含有量,并建立相关性模型用于定量分析。方法采集30批银黄口服液的近红外光谱数据,HPLC和偏最小二乘法建立绿原酸和黄芩苷的相关性模型。结果绿原酸和黄芩苷的最佳波段均为9 401.6-7 826 cm^-1,R2分别为0.985 2和0.9881,交叉验证均方差（RMSECV）分别为0.002 96和0.018,相对分析误差（RPD）分别为8.23和9.27。将该模型应用于检测两者的含有量时,其预测值的马氏距离（Mash distance）均小于1,相对误差分别为2.12%和1.67%。结论该模型预测性良好,可用于银黄口服液的实时监测和质量控制。AIM To determine the contents of chlorogenic acid and baicalin in Yinhuang Oral Liquid（ Lonicera japonica and Scutellaria baicalensis） by near-infrared spectroscopy and to establish a dependency model for quantitative analysis. METHODS Based on the near-infrared spectroscopy data of thirty batches of Yinhuang Oral Liquid,the dependency model for chlorogenic acid and baiclin was established by HPLC and partial least squares.RESULTS 9 401. 6- 7 826 cm^- 1served the best waveband for the detection of both chlorogenic acid and baicalin,0. 985 2 and 0. 988 1 were for R2,0. 002 96 and 0. 018 were for root-mean-squares error of cross-validation（ RMSECV）,and 8. 23 and 9. 27 were for regression point displacement（ RPD）,respectively. When this model was used for the determination of these two constituents,the predicted Mash distance values of both constituentswere less than 1,whose relative errors were 2. 12% and 1. 67%,respectively. CONCLUSION With good predictability,this model can be applied to the real time monitoring and quality control of Yinhuang Oral Liquid.

关 键 词：银黄口服液 绿原酸 黄芩苷 相关性模型 近红外光谱 HPLC 偏最小二乘 

分 类 号：R927.2[医药卫生—药学]