Implementation of Chemometric Tools To Improve Data Mining and Prioritization in LC-HRMS for Nontarget Screening of Organic Micropollutants in Complex Water Matrixes.

Abstract

One of the most critical steps in nontarget screening of organic micropollutants (OMP) in complex environmental samples is handling of massive data obtained from liquid chromatography coupled with high-resolution mass spectrometry (LC-HRMS). Multivariate chemometric methods have brought about great progress in processing big data obtained from high-dimensional chromatographic systems. This work aimed at a comprehensive evaluation of two LC-Q-Orbitrap mass spectrometry full-scan data sets for target and nontarget screening of OMPs in drinking and wastewater samples, respectively. For each data set, following segmentation in the chromatographic dimension, at first multivariate curve resolution alternating least-squares (MCR-ALS) was employed for simultaneous resolution of global matrices. The chromatographic peaks and the corresponding mass spectra of OMP were fully resolved in the presence of highly co-eluting irrelevant and interfering peaks. Then partial least-squares-discriminant analysis was conducted to investigate the behavior of MCR-ALS components in different water classes and selection of most relevant components. Further prioritization of features in wastewater before and after ozonation and their reduction to 24 micropollutants were then obtained by univariate statistics. Two-way information retrieved from MCR-ALS of LC-MS1 data was also used to choose common precursor ions between recoveredand measured data through data-dependent acquisition. MS1 and MS2 spectral features were used for tentative identification of prioritized OMPs. This study indicates that the described strategy can be used as a promising tool to facilitate both feature selection through a reliable classification and interference-free identification of micropollutants in nontargeted and class-wise environmental studies.