单纯形线性规划多组份同时分光光度分析5-Br-PADAP光度法测定八种常见离子

The Application of Simplex Optimization to Multicomponent Simultaneous Spectrophotmetry

本文将分光光度法无机分析中多组分同时分析的问题归结成线性规划问题。用单纯形法求解。编成FORTRAN和BASIC程序,在长城0520或IBM微机上应用均获得成功。实验体系为常见八种金属离子Cd、Co、Ni、cu、Pb、Zn、Mn、Fe与5-Br-PADAP-OP的络合物。当pH为9,取10个波长数可任选2-5种组分组合同时测定,其相对偏差为1.0-8.0%。随组份数增加偏差增大。方法用于标准水样,铝合金标样和铬铝钴钢标样中多组分分析,其平均误差小于11.0%,可满足例行分析要求。

The simultaneous spectrophotometry was treated as a problem of Linear-programming. FORTRAN and BASIC programs were worked out by using simplex optimization. The formation of complexes of Cd, Co(Ⅱ), Cu, Fe(Ⅲ), Mn(Ⅱ), Ni, Pb(Ⅱ), and Zn with 5-Br-PADAP and emulsifying agent OP was studied. All cations form stable complexes at pH 9. A series of Co-Ni-Cu-Zn mixtures were studied and it was found that as few as ten evenly spaced wavelengths can give satisfactory calculations in the simultaneous spectrophotometry. The Linear-programming technique gave better results than the Least-squares. Diverse combinations of two to five components of the Cd, Co, Cu, Fe, Mn, Ni, Pb, and Zn complexes with 5-Br-PADAP and OP were studied. The average relative errors ranged 1.0-8.0%, the more complex mixtures gave less accurate results. The method has been applied to the analyses of two water reference samples (for Pb, Zn, Cu and Cd), two aluminum alloy reference samples (for Fe, Ni, Mn, Zn nad Cu), and a Cr-Al-Co steel reference sample (for Co and Ni). The maximum average error was less than 11.0%.