Non-separative determination of isomeric polycyclic aromatic hydrocarbons by electrospray Ag(I) cationization mass spectrometry and multivariate calibration

• Published in: Microchemical journal Vol. 183; p. 108072
• Main Authors: María Casas-Ferreira, Ana, del Nogal Sánchez, Miguel, Rodríguez-Gonzalo, Encarnación, Pérez Pavón, José Luis
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2022
• Subjects: Ag (I) cationization; Experimental design; Flow injection analysis; Mass spectrometry; PLS models; Polycyclic aromatic hydrocarbons; Flow injection analysis; PLS models; Ag (I) cationization; Polycyclic aromatic hydrocarbons; Experimental design; Mass spectrometry
• ISSN: 0026-265X; 1095-9149
• DOI: 10.1016/j.microc.2022.108072

[Display omitted] •A non-separative method for the semiquantificaction of PAHs isomers is proposed.•The method is based on FIA-ESI-MS/MS and analysis time was 1.7 min.•The ionization of PAHs was achieved by forming complexes with Ag (I)•Experimental design for calibration and PLS multivariate calibration were used. A new approach for the determination of isomeric polycyclic aromatic hydrocarbons using a stand-alone mass spectrometry method is proposed. The aim of the work is to study quantitative possibilities of multivariate calibration and electrospray Ag(I) cationization mass spectrometry for the non-separative determination of polycyclic aromatic hydrocarbons isomers. The method is based on flow injection analysis, electrospray ionization and tandem mass spectrometry (FIA-ESI-MS/MS). No chromatographic column was included into the instrumental configuration and the analysis time was 1.7 min. Seven polycyclic aromatic hydrocarbons were selected as test compounds and the ionization was achieved by forming complexes with Ag (I). Individual quantification of all the isomers was carried out by using PLS multivariate calibration and experimental design for calibration modeling. The PLS models were used to predict the concentration of the analytes in a set of external validation samples and satisfactory results were obtained. RMSE, expressed as a relative value, were found to be between 23 and 34 %. Results obtained with multivariate analysis were compared with those corresponding to univariate calibration to show its high potential.