Second-order chromatographic photochemically-induced fluorescence emission data coupled to chemometric analysis for the simultaneous determination of urea herbicides in the presence of matrix co-eluting compounds

J. ARANCIBIA, G. ESCANDAR

Analytical Methods

Royal Society of Chemistry

Lugar: Londres; Año: 2014 vol. 6 p. 5503 - 5511

1759-9660

This paper presents a novel approach for the simultaneous determination of isoproturon, rimsulfuron and monuron, three widely used urea-derivative herbicides, in interfering environmental samples, combining the advantages of photoinduced fluorescence (PIF) emission, liquid chromatography and second-order chemometric algorithms. Chromatographic detection is made with a fast-scanning spectrofluorimeter, which allowed the efficient collection of PIF through a post-column photoreactor. Thus, second-order elution time-PIF emission data matrices are rapidly obtained (in less than 4 min) with a chromatographic system operating in isocratic regime using a minimal amount of organic solvent. The goal of the present study was the successful resolution of a system in the presence of foreign compounds which can be present in real samples. The study was employed for the discussion of the scopes of the applied second-order algorithms selected for data processing, namely parallel factor analysis (PARAFAC), multivariate curve resolution- alternating least-squares (MCR-ALS), and multidimensional and unfolded partial least-squares coupled to residual bilinearization (N- and U-PLS/RBL). U-PLS/RBL showed the best performance to quantify the herbicides, even when the foreign compounds showed very similar spectral and time profiles to the analytes. The quality of the proposed technique was assessed on the basis of the analytical recoveries from different types of water samples spiked with analytes and other selected agrochemicals. After a solid-phase extraction, reaching a preconcentration factor of 50, detection limits of 2.9, 2.4, and 1.7 ng mL?1 for isoproturon, rimsulfuron, and monuron, respectively, were obtained in those interfering matrices, with relative prediction errors lower than 5 %.