化学计量学-荧光光谱法同时测定四种农药

应用化学计量学方法对荧光光谱严重重叠的灭螨猛、三唑磷、蝇毒磷和多菌灵四组分混合体系进行光谱解析,建立了四种农药同时测定的新方法。在pH=7.87 Britton-Robison缓冲介质中,以λ=228 nm为共激发波长,扫描四种农药混合物在260~440 nm波长范围荧光光谱,然后用不同化学计量学模型对重叠光谱数据进行解析并同时测定四组分。研究表明,偏最小二乘和主成分回归算法对四组分混合物合成样预报结果要优于径向基人工神经网络模型和经典最小二乘算法,偏最小二乘和主成分回归算法的预报浓度与实际浓度的总相对预报误差均为7.5%。用偏最小二乘法结合荧光光谱法对实际样品进行测定,结果满意。

Simultaneous Fluoprescence Spectrophotometry Determination of Four Pesticides By Chemometrics

A new fluorescence method by chemometrics for simultaneous determination of chinomethionat,triazophos,coumaphos and carbendazim was developed.The experimental conditions were investigated and the spectra from 260 to 440 nm were scanned.The different chemometric methods were applied to resolve the overlapping fluorescence spectrum and determine the four compounds simultaneously in a B-R buffer solution(pH 7.87) and at excitation wavelength of 228 nm.It was found that the partial least squares and principal components regression gave better results for the synthetic mixtures of four compounds than radial basis function artificial neural network and classical least squares.Their relative prediction error for all components(RPET) obtained by the predicted and actual concentrations were both 7.5%.The partial least squares was applied to the analysis of practical samples with satisfied results.