Non-targeted analysis of wastewater treatment plant effluents by high performance liquid chromatography–time slice-solid phase extraction-nuclear magnetic resonance/time-of-flight-mass spectrometry

Extracts of effluents from two different wastewater treatment plants (WWTP) in Switzerland taken during the application period of pesticides were examined by coupling an HPLC–MS system to a nuclear magnetic resonance spectrometer using a post column peak trapping device. By trapping 1 min portions of the chromatogram onto post column solid phase extraction cartridges (time slice-SPE-NMR) a comprehensive overview of proton carrying constituents could be achieved. Non-supervised statistical analysis of the NMR spectra obtained by this approach revealed NMR resonances pointing to contaminants present in decreasing proton concentration in the extracts. Comparison of exact mass data acquired during the trapping process to these NMR resonances enabled the identification of the pesticides Linuron, Metazachlor, Ethofumesate, Isoproturon, Metamitron, Propazine and Chloridazon. Desaminometamitron, a known transformation product of Metamitron could also be identified together with unexpected highly concentrated C8, C10 and C12 fatty acids and their glycerol mono- and di esters. Other compounds identified were a drug metabolite (3-Carboxymefenamic acid), a sun screen agent (Ensulizole: 2-Phenyl-1H-1,3-benzodiazole-6-sulfonic acid) and industrial chemicals (Benzotriazole, N-Benzyl-indole). In addition, a number of well-resolved proton spectra cannot be attributed to a mass response showing the need of further investigations using 2D-NMR and different ionization techniques.

► We describe an approach for the completely non-targeted analysis of complex mixtures based on the universal response of proton detected nuclear magnetic resonance.
► Separation of wastewater treatment plant effluents was done on an HPLC column.
► 1 min portions of the eluent are trapped to a post-column SPE device and eluted to an NMR spectrometer for detection.
► After non-supervised statistical analysis of time slice spectra and interpretation of the corresponding NMR and MS data contaminants were identified, e.g. pesticides, drug metabolites and industrial chemicals.