Suspect screening of emerging contaminants in granular activated carbon filter used for drinking water treatment

Abstract

Drinking water has been contaminated with widespread distribution of various organic compounds. Recently, high-resolution mass spectrometry (HR-MS) has been used increasingly for investigating emerging contaminants and transformation products in environmental samples. Universal screening and identification of novel micropollutants make it possible to overcome the limited information on occurrence and concentration of numerous micropollutants. Target, suspect and nontarget screening by using HR-MS have been developed and optimized continuously to identify unpredictable and unknown comtaminants at trace levels.

In this study, suspect screening method was applied on granular activated carbon filter samples installed in water purifiers collected in Korea. With those used GAC filter samples, unused GAC filter samples were analyzed as blank sample.

The first step, GAC sample extracted with methanol was screened for the 8 target perfluoroalkyl compounds (PFCs) using known retention times, accurate masses and fragment ions. All target compounds were detected and confirmed with reference standard. 6 out of 9 PFCs were detected in all GAC samples. Detected PFCs is Heptadecafluorooctane sulfonic acid(Perfluorooctane sulfonic acid), Perfluorodecanoic acid(Perfluorocapric acid), Perfluoroheptanoic acid, Perfluorohexanoic acid, Perfluorononanoic acid, Perfluorooctanesulfonic acid, Perfluorooctanoic acid, Perfluoropentanoic acid.

Next step was performed by suspect screening at full-scan mode for accurate mass measurement with HR-MS, a post-target screening approach was undertaken, where searching and identification of other interesting compounds could be done at any time without performing additional analyses or without using reference standards. By GAC extracted sample analysis, tentative contaminants were identified. Data acquisition filter is applied for condition according to response value, mass error(ppm), the number of total fragment fount then we build up final comtaminant candidate list which is obtained as meaningful peak shape.

In the first extraction for 14 GAC filter using Acetonitrile solvent, 15 components are selected containing plant growth regulator (Heptopargil) in negative mode. Next extraction for 6 GAC filter detected Perfluoroakyl compounds using Methanol solvent, 25 components are selected containing veterinary antibiotic (Danofloxacin) in positive mode and 17 components are selected containing disinfection byproduct (Monobromoacetic acid) in negative mode. Finally, last extraction for 27 GAC filter using Dichloromethane solvent, 16 components are selected containing pesticide (Dipropetryne), hormone (Estradiol) in positive mode and 28 components are selected containing insecticide(Sophamide) in negative mode.

When collecting for 19 tap water sample using solid phase extraction was performed, 11 components are selected containing acaricide(fluacrypyrim) in positive mode and 14 components are selected containing fungicide(piperalin) in negative mode.

In this study, suspect screening methods using HR-MS were shown to be a powerful tool for universal screening and identification of presence of micropollutants in drinking water. Its utility in this study is illustrated in our identification of several compounds that would not otherwise be monitored in drinking water. The need for a more comprehensive suspect screening approach is highlighted by the large number of features present in the samples, and the limited number of which that were confirmed or tentatively identified. We have demonstrated limited candidate contaminants list that prioritized frequency ranking by response data for which standards were not available on hand.

This results indicate that granular activated carbon used in drinking water filtration systems likely remove various micropollutants. As a result, extracted GAC filter enable to detect some micropolutants which is hardly detected in tap water with trace level. Also, this study could overcome limitness of mornitoring method for long-term and time-lapse contaminants having concentration fluctuation in final water distribution system.