Occurrence and fate of selected anticancer, antimicrobial, and psychotropic pharmaceuticals in an urban river in a subcatchment of the Yodo River basin, Japan.
A multi-residue method for the simultaneous determination of more than 90 pharmaceuticals in water samples was developed and validated. The developed method utilizes a single liquid chromatography-tandem mass spectrometry (LC-MS/MS) run after sample enrichment using solid-phase extraction (SPE). The pharmaceuticals included in this method were chosen based on their potency (effect/concentration ratio) and potential to bioaccumulate in fish. Because the selection was based on ecotoxicological criteria and not on ease of detection, the pharmaceuticals have a wide range of physico-chemical properties and represent 27 distinct classes. No method for surface, waste water or similar matrices was previously described for 52 of the 100 target analytes. Four chromatographic columns were tested to optimize the separation prior to detection by mass spectrometry (MS). The resulting method utilizes a Hypersil Gold aQ column. Three different water matrices were tested during method validation: Milli-Q water, surface water (river water from the Umea River) and effluent from the Umea waste water treatment plant (WWTP). Four of the selected pharmaceuticals exhibited poor method efficiency in all matrices. Amiodarone, Dihydroergotamine, Perphenazine and Terbutalin were omitted from the final analytical method. In addition, five compounds were excluded from the method for surface water (Atorvastatin, Chloropromazin, Dipyridamol, Furosemid and Ranitidin) and three other pharmaceuticals (Glibenclamid, Glimepirid and Meclozine) from waste water method respectively. Absolute recoveries were above 70% for Milli-Q water, surface water, and sewage effluent for most pharmaceuticals. The limits of quantification (LOQs) ranged from 0.05 to 50 ng L(-1) (median 5 ng L(-1)). The use of matrix-matched standards led to the elimination of ionization enhancement or suppression. The recoveries of the method for real matrices were in the range of 23-134% for surface water (only three compounds were outside of the range of 40-130%) and in the range of 47-162% for waste water (five compounds were outside of the range of 40-130% at lower validated concentration).