Database-driven screening of South African surface water and the targeted detection of pharmaceuticals using liquid chromatography - High resolution mass spectrometry

Pharmaceuticals and personal care products are released into aquatic environments, largely as a result of ineffectual removal during wastewater treatment. Here we present a screening strategy based on the use of three commercially available mass spectral databases, combined into a single searchable entity and parallelized by cluster computing. In addition to this, a targeted solid phase extraction method with Ultra High Pressure Liquid Chromatography coupled to quadrupole time of flight mass spectrometry (UHPLC-QTOF) was used to quantify 99 pharmaceuticals in South African surface water on a national level. Limits of quantification were in the low ng/L range for the majority of the compounds and it was found that nationally both Lamotrigine and Nevirapine occurred most often. Prednisolone and Ritonavir were present at the highest average concentration; 623 and 489 ng/L respectively. It is however shown that more than 50% of the targets chosen for analysis are not detectable in any of the samples, which highlights the utility of untargeted, database driven screening; prior to the use of costly analytical standards. Untargeted screening detected 45% of the compounds detected in targeted mode, and furthermore tentatively identified a total of 4273 unique compounds across the samples. Automatically triggered MS/MS analyses yielded 92 unique hits with greater than 95% confidence. It is therefore suggested that untargeted screening should precede the targeted approach as a matter of economy and to guide the selection of targets for quantification. There is however great room for improvement in current commercial database search methodologies as a large bottleneck exists due to processing time. [Display omitted] •Quantitation of 96 pharmaceuticals by UHPLC-QTOF in South African surface water.•Lamotrigine and Nevirapine occurred most often.•Prednisolone and Ritonavir were present at the highest average concentrations.•Developed a procedure to screen samples prior to costly targeted analysis.•Standard analytical software was parallelized on a cluster reducing processing time. In this work, we describe for the first time the extent of pharmaceutical pollution in South African surface water; using database-driven mass spectral analyses in a user friendly cluster computation format.