Chlorinated paraffin analysis by gas chromatography Orbitrap high-resolution mass spectrometry: Method performance, investigation of possible interferences and analysis of fish samples

摘要

For decades, high quantities of short-chain chlorinated paraffins (SCCP) and medium-chain chlorinated paraffins (MCCP) have been widely used, for instance as plasticizers or flame retardants, leading to global pollution due to unintentional emissions from products or waste. Due to the high complexity of chlorinated paraffins with several thousand congeners there is no consensus on an analytical procedure for SCCPs and MCCPs in food samples. Amongst the multitude of methods currently in use, high-resolution mass spectrometry is particularly valuable for in-depth studies of homologue patterns. Here we analyse SCCPs and MCCPs with gas chromatography coupled to high-resolution Orbitrap mass spectrometry (GC-Orbitrap-HRMS) operated in full-scan acquisition in electron capture negative ion (ECNI) mode at 60,000 and 120,000 resolution (FWHM, m/z 200, equals roughly 30,000 and 60,000 at 5% peak height). Linear dynamic range, selectivity and sensitivity tests confirmed an excellent linearity in a concentration range of 25-15,000 pg/mu L with very low limits of detection (LODs) in the low pg/mu L. range. Spiking experiments with high levels of native mono- and di-ortho-polychlorinated biphenyls (PCBs) and mixtures of MCCP and SCCP standards did not have a negative impact on isotope ratios of the examined homologues. Besides the [M-CI](-) fragment ions used for quantification, the mass spectra of homologues also featured [M-HCl](-) ions whose abundance increased with decreasing chlorination degree. In addition, [M-HCI-Cl](-) ions were detected with a relative abundance of 5-10%. Three salmon (Salmo salar) samples farmed in Norway showed a consistent CP homologue pattern which differed both from the CP pattern in a sample from Scottish aquaculture and a wild salmon sample. These measurements produce evidence that discretely different CP patterns may exist in different areas of origin. Our results demon strate that GC/ECNI-Orbitrap-HRMS is well-suited for the analysis of CPs by