Non-linear multivariate curve resolution applied to the spectrophotometric determination of cerium(iii) in aqueous nitric acid solutions for process control

Abstract

It is shown that the recently proposed Non-Linear Multivariate Curve Resolution (NL-MCR) method can be effectively employed to develop an accurate calibration of cerium(III) using spectrophotometry measurements of mixtures of rare earth elements in nitric acid. Spectroscopic techniques provide a unique opportunity for the in-line determination of critical concentrations rapidly and without serious risks to operating personnel and the environment. Cerium has no absorbance bands in the visual and near-infrared range. In the ultraviolet range cerium(III) has a unique large peak which overlaps with even larger peaks of nitric acid. In the case of in-line control, where conventional analytical means are limited, we encounter a peak flattening and, consequently, a Lambert–Beer law violation. Therefore, the conventional calibration methods, such as Partial Least Squares (PLS) and Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS), yield useless results. Our previous attempt to overcome Beer's law violation by introducing a non-linear constraint in the MCR-ALS procedure failed. Application of the NL-MCR method with specially selected transition function not only yields accurate cerium determination but also provides an opportunity to assess the unknown nitric acid concentration in new samples. It is shown that the established calibration models are stable to some extent for out-of-control cases.