Monitoring organic pollutants in rainwater is important to understand relations between air pollutants and water safety. Non-targeted liquid chromatography–mass spectrometry (LC–MS) is a powerful tool but because of big differences of pollutants in polarity, ionization efficiency, and low concentrations, it is challenging to detect all pollutants in a single analysis. Chemical derivatization is a widely adapted strategy to fractionate complex samples with enhanced sensitivity and selectivity. Herein, we propose the usage of dansylation as a chemical derivatization method to improve both LC retention and MS ionization of organic pollutants containing amine, hydroxyl, and carboxyl for non-targeted LC–MS analysis. We first evaluated the labeling coverage of dansylation to organic pollutants in water matrix. Using dansyl chloride (DnsCl) and dansyl hydrazine (DnsHz) to label 100 amine- and hydroxyl-containing compounds and 100 carboxyl compounds, respectively, we found DnsCl and DnsHz had over 60% labeling coverage for 8 categories of compounds. Then dansylation was applied to label the rainwater, source water, disinfected rainwater, and drinking water samples. To facilitate the annotation of dansylated compounds, we also established a web-based tool termed DansylFinder. Using DansylFinder, 3889, 5813, 6077, and 4050 tentative annotations were found in rainwater, source water, disinfected rainwater, and drinking water samples by dansylation-based non-targeted LC-HRMS analysis. Four hundred fifty four were persistently detected in all water samples, suggesting significant organic overlaps of the four water samples. In addition, two degradation pathways reported in drinking water disinfection process were also detected in disinfected rainwater, suggesting rainwater is a potential path for air pollutants to infiltrate drinking water system.
Read full abstract