同时测定合成样中铜钴镍钒含量的分光光度法——主成分回归法和偏最小二乘法比较

Simultaneous Determination of Copper, Cobalt, Nickel and Vanadium inSynthetic Samples by Spectrophotometry——Comparison betweenPrincipal Component Regression and Partial Least _ Squares

研究了主成分回归和偏最小二乘方法在多组分分光光度法分析中的应用 ,以5_Br_PADAP(2[(5_溴_2_吡啶 )_偶氮]5_二乙氨基苯酚 )为显色剂 ,OP(聚乙二醇辛基苯基醚 )作为增溶增稳剂 ,在pH=3.6的条件下 ,用主成分回归及偏最小二乘分光光度法同时测定了合成样中的铜、钴、镍、钒4组分含量 ,测定相对误差在 -6.00 %～4.00 %之间。实验证明 ,对于加和性不好的体系偏最小二乘分光光度法要优于主成分回归法。

The application of principal component regression(PCR) and partial least _ squares(PLS) method for the simultaneous determination of components in multicomponent system by spectrophotometry is reported. Contents of copper, cobalt, nickel and vanadium in synthetic samples were determined by using 5 _ Br _ PADAP as a color agent, OP as sensitizing and stabilizing agent. The relative error of the determined results for the two methods ranges from -6.00% to 4.00%. It can be found that for the system with poor additivity PLS method is superior to PCR.