红外光谱对苏云金杆菌制剂中违禁添加溴虫腈的定量检测  

Quantitative Study on Prohibited Addition of Chlorfenapyr in Bacillus Thuringiensis Formulations by Infrared Spectroscopy

作　　者：韦芸 夏静静 徐惟馨 陈玥瑶 毛欣然 熊艳梅[1] 闵顺耕[1] WEI Yun;XIA Jing-Jing;XU Wei-Xin;CHEN Yue-Yao;MAO Xin-Ran;XIONG Yan-Mei;MIN Shun-Geng(College of Science,China Agricultural University,Beijing 100193,China)

机构地区：[1]中国农业大学理学院,北京100193

出　　处：《分析化学》2022年第3期482-490,共9页Chinese Journal of Analytical Chemistry

摘　　要：建立了利用衰减全反射傅里叶变换红外光谱(ATR-FTIR)结合偏最小二乘法(PLS)对苏云金杆菌制剂中违禁添加的溴虫腈含量进行快速测定的方法。在3种不同来源的苏云金杆菌制剂中加入不同质量的97.00%(m/m)溴虫腈原药,配制成溴虫腈浓度为0.00%~5.00%的153个混合样品。选择3种预处理方法:Savitzky-Golay卷积平滑法(S-G)、标准正态变换法(SNV)和多元散射校正法(MSC)以及5种变量选择算法:区间偏最小二乘法(iPLS)、移动窗口偏最小二乘法(MWPLS)、无信息变量消除法(UVE)、竞争性自适应重加权采样(CARS)、自助软收缩法(BOSS),考察了不同预处理方法及变量选择方法对模型结果的影响。其中,MSC预处理结合BOSS算法获得最优的模型效果:交叉验证均方根误差(RMSECV)为0.0017,交叉验证集决定系数(R^(2)_(cv))为0.9859,预测均方根误差(RMSEP)为0.0016,预测集决定系数(R^(2)_(pre))为0.9868。对于溴虫腈含量范围在0.50%~5.00%的样品,外部检验样品预测平均相对误差为0.0540;并且BOSS算法选取的变量主要集中在溴虫腈特征峰的吸收区域。本方法不仅具有优异的模型效果,而且具有很好的化学可解释能力,可用于苏云金杆菌制剂中违禁添加溴虫腈的快速检测。A method for rapid determination of chlorfenapyr in Bacillus thuringiensis formulations with prohibited additions using attenuated total reflection Fourier transform infrared spectroscopy(ATR-FTIR)combined with partial least squares(PLS)analysis was developed.Three Bacillus thuringiensis formulations from different sources were added with different masses of 97.00%(m/m)chlorfenapyr prodrug to prepare 153 mixed samples with chlorfenapyr concentrations ranging from 0.00%to 5.00%.Three pretreatment methods(Savitzky-Golay smoothing(S-G),standard normal variation(SNV)and multivariate scattering correction(MSC))and five variable selection algorithms(interval partial least-squares(iPLS),moving window partial least-squares(MWPLS),elimination of uninformative variables(UVE),competitive adaptive reweighted sampling(CARS),and the bootstrapping soft shrinkage(BOSS))were employed to investigate the effects of different pretreatment methods and variable selection methods on the model results.Among them,the MSC pretreatment method combined with the BOSS algorithm obtained the optimal model results,and with this method,RMSECV=0.0017,R^(2)_(cv)=0.9859,RMSEP=0.0016,and R^(2)_(pre)=0.9868.For samples with chlorfenapyr concentration ranging from 0.50%to 5.00%,the average relative error of external test sample prediction was 0.0540,and the variables selected by the BOSS algorithm were mainly concentrated in the absorption region of the characteristic peak of chlorfenapyr.This method not only had excellent modeling effect,but also showed good chemical interpretability,and could be applied to the rapid detection of chlorfenapyr in Bacillus thuringiensis formulations with prohibited additions.

关 键 词：红外光谱 偏最小二乘法 变量选择 溴虫腈 苏云金杆菌 

分 类 号：O657.33[理学—分析化学] TQ450.7[理学—化学]