Unraveling In-Source Fragmentation Patterns of Organophosphate Esters by LC-ESI-HRMS: Implications for False Positive Identification in Nontarget Analysis of Environmental Matrices

In-source fragmentation (ISF) is a challenge in liquid chromatography-electrospray ionization high-resolution mass spectrometry (LC-ESI-HRMS) analysis of organophosphate esters (OPEs), leading to lower abundances of protonated OPEs and severe false positive annotations in environmental nontarget analysis. Herein, we investigated ISF behaviors of 75 OPEs and assessed their application for environmental nontarget analysis. Screening of mass spectrometry (MS) acquisition parameters indicated that capillary temperature (CT) and S-lens radio frequency (RF) level strongly influenced ISF, whereas auxiliary gas heater temperature (AGHT) mainly affected the protonated OPE response. Utilizing spectral quality index (SQI), the optimized acquisition parameters (CT = 200 °C, AGHT = 450 °C and RF = 55%) were established to maximize protonated OPEs responses while suppressing ISF. All 75 OPEs exhibited detectable ISF, seven ISF pathways were identified among OPEs, with the McLafferty rearrangement and the π-electron-mediated intramolecular displacement (π-EMID) reaction as the dominant proposed pathways. Non-aryl and mixed aryl-alkyl OPEs bearing γ-hydrogen atoms predominantly undergo ISF via McLafferty rearrangement, while OPEs with unsaturated side chains like benzyl, allyl and aryl proceed ISF via the unique π-EMID reaction. Nontarget analysis of atmospheric fine particulate matter (PM2.5) samples identified 107 candidate features, including 47 OPE features, 14 OPE features accompanied by ISF-derived signals at different retention times, 28 ISF-derived false positive OPE annotations, 13 hydrocarbon carbocations and 5 protonated oxygenated hydrocarbon ions derived from OPE ISF. These findings provide a comprehensive characterization of the ISF pathways of OPEs and emphasize the importance of accounting for ISF-induced false positives in LC-ESI-HRMS nontarget analysis.