FTIR-ATR结合膜富集技术测定中药材地龙中微量铜的含量

Quantitative analysis on Cu content in pheretimae by FTIR-ATR spectra combining with membrane enrichment technology

目的利用傅里叶变换-衰减全反射红外光谱（FTIR-ATR）结合膜富集技术,建立中药材地龙中微量铜质量分数的快速、经济的测定方法。方法首先,对30批校正集样品进行湿法消解,利用铜与金属络合剂1-（2-吡啶偶氮）-2-萘酚（PAN）发生络合反应,优化各种反应条件,包括p H、PAN用量、反应时间。再将络合物抽滤富集在微孔滤膜上,采集其FTIR-ATR光谱,经9点平滑、一阶导数9点平滑等预处理后,利用随机森林回归算法建立其铜质量分数的定量校正模型。校正集按3∶1随机分配为训练集和测试集,对模型进行参数优化和评价。利用优化后模型预测广东、广西和福建3个产地地龙样品中铜的质量分数,并与ICP-MS测定结果进行比较。结果最优模型预测3个产地地龙样品的质量分数在5.85~6.98 mg/kg之间,与ICP-MS分析结果相比,相对误差均小于20%。结论本方法利用膜富集技术提高了FTIR-ATR的检测灵敏度,为中药材重金属元素的定量分析提供了一种新方法。

Objective To establish a rapid, economic method for the determination of Cu content in the TCM pheretimae by FTIR-ATR spectra combined with membrane enrichment technology. Methods 30 calibration samples were processed by wet digestion, and complexed by PAN, a metal complexing agent for Cu. The complexing conditions, including pH, PAN quantity and reaction time were optimized with characteristic absorption of complex compounds as index. Then, complex compounds were filtered in microporous membranes, and the FTIR-ATR spectra of which were collected, processed by 9-point smoothing and 9-point smoothing coupled with first differential etc. for the calibration. A calibration model of Cu content in pheretimae was established using random forest algorithm, in which the calibration set was randomly assigned as training set and test set in 3 ： 1 ratio, and the parameters were optimized as well. Cu contentsof pheretimae in Guangdong, Guangxi and Fujian were predicted by optimized model, and compared with ICP-MS results. Resluts The prediction concentrations in pheretima samples of three different roots using optimized model were between 5.85-6.98 mg/kg, with relative error less than 20% compared with ICP-MS results. Conclusion A new method for the quantitative analysis of heavy metals in TCM was established, and the sensitivity of FTIR-ATR was improved combined with membrane enrichment technology. The content of microscale Cu could be predicted in pheretimae properly.