Inverted Multivariate Calibrations as an Instrument for the Separate Determination of Similar Analytes from a Mixture Spectrum with Non-Additive Absorbance

Abstract

Multivariate calibrations are used in spectrophotometric analysis to determine multiple analytes in multicomponent solutions. These calibrations relate generalized signals measured at multiple wavelengths with concentrations of the analytes. The aim of this study was to assess the applicability of inverted multivariate calibrations (IMCs) for the separate determination of similar analytes under conditions of nonadditive absorbance. The test samples were model aqueous solutions simultaneously containing Cu(II), Co(II), Ni(II), Zn(II), and Pb(II) along with an excess of the photometric reagent 4-(2-pyridylazo)resorcinol. In these solutions, statistically significant deviations from additive absorbance were observed, likely due to a shift in the complexation equilibrium. The input data for constructing the IMC were the spectra of model mixtures from the training set. The number of analytical wavelengths (m) and the number of mixtures in the training set (n) were varied during the experiment. The metal concentrations in the mixtures of the test set were calculated individually by multiple linear regression, using different spectral regions and different IMCs. The best results were obtained with m = 16 and n = 30. The determination errors for Co, Ni, and Zn in single mixtures did not exceed 25 rel % (in modulo), while the generalized errors (RMSEP) were 10–15 rel %. The determination errors for copper and lead were significantly higher. The experiment demonstrated that IMCs allow to determine the separate components of mixtures with similar but nonadditive spectra. However, the amount of the input data required must be significantly larger than in assessing the total amount of the same analytes, the accuracy of the results will be lower, and the correct determination of all analytes cannot be guaranteed.