Multivariate analysis of reagent profiles for slow reactions using PLS analysis

Abstract

A methodology is proposed in order to use the reagent profiles obtained in coulometric or volumetric experiments when the titration reaction is slow. The proposal consists of the use of the whole signal–time profile for the reagent to characterize the sample, by using multivariate calibration methods. The reagent profiles have been simulated for a slow second-order reaction and their possibilities for quantitative purposes evaluated. Experimental verification has been performed using two chemical systems: hexacyanoferrate(III)–fructose in alkaline medium (coulometric approach, amperometric monitoring) and hexacyanoferrate(III)–cysteine at pH 4 (volumetric approach, photometric monitoring). Partial least squares regression (PLS) has been applied as calibration algorithm.