Removal and transformations of diclofenac and sulfamethoxazole in a two-stage constructed wetland system

[Display omitted] •Two-stage CW offered 96% removal of SMX and 78% removal of DCF.•TPs of DCF&SMX indicated multiple transformation routes including abiotic reduction.•Some TPs detected in the first stage were not detected in the second stage.•Forced aeration may hinder the removal of SMX in subsurface-flow CWs.•DCF&SMX displayed phytotoxic effect in the surface-flow CW. The aim of this study was to evaluate the removal of sulfamethoxazole (SMX) and diclofenac (DCF) in a mesocosm two-stage constructed wetland system consisting of a subsurface-flow bed (first stage; CW-1) and a surface-flow system (second stage; CW-2) with an emphasis on the effect of artificial aeration as a potential performance intensification option. Furthermore, the additional aim was to identify the transformation products of DCF and SMX along the flow path of the treatment system. This two-stage system was fed for 442 days with an artificial municipal wastewater containing DCF and SMX at a concentration of 2 mg/L each. The removal of SMX in the CW-1 was found to be significantly higher under non-aerated conditions (>77%) as compared to aerated conditions (26%), but the aeration of the CW-2 was insignificant for the SMX removal. In contrast, the removal of DCF was not affected by aeration in the CW-1 (removal efficiency of approximately 50%), but was significantly positively influenced by the combined effect of plants and aeration in the CW-2. The overall removal of SMX and DCF in the two-stage system was 95.8% and 77.6%, respectively. The transformation products of SMX and DCF detected in the studied system indicated the potential contribution of both oxic and anoxic transformation pathways, potentially including also abiotic reaction between SMX and Fe(II) leading to the cleavage of the isoxazole ring in this compound. Some of the transformation products found in the CW-1 were not further detected in the CW-2. Based on the literature data the antimicrobial activity and potential activity of the identified transformation products cannot be ruled out. Giant miscanthus growing in the CW-1 was unaffected by the high concentrations of DCF and SMX in contrast to the emergent macrophytes (on the floating island) or the free-floating plants present in the CW-2, which failed to adapt to the presence of the pharmaceuticals. As a result, the CW-2 was colonized by an unidentified species of a water moss and algae.