The aim of this work was to develop procedures for the simultaneous determination of selected brominated flame retardants (BFRs) in river water and in river bed sediment. The target analytes were polybrominated diphenyl ethers (PBDEs) and tetrabromobisphenol A (TBBPA). To determine dissolved BFRs, a novel mixed-mode solid-phase extraction procedure was developed by combining a hydrophobic sorbent (C(18)) with a silica-based anion exchange sorbent, so as to overcome the negative artefact induced by dissolved organic carbon. Extraction recoveries exceeded 73% for most analytes, except for BDE-183 and BDE-209 (57%). As regards suspended sediment and river bed sediment, extraction was carried out by means of ultrasonication (recoveries: 73-94%). These procedures, combined to gas chromatography coupled to negative chemical ionisation mass spectrometry (GC-NCI-MS), enabled the determination of BFRs at trace level: 3-160 pg L(-1) in river water, 5-145 pg g(-1) in bed sediment. These methods were applied to the determination of PBDEs and TBBPA in a suburban river (near Paris, France). PBDEs were systematically detected in the water column (SigmaBDEs, 2,300-4,300 pg L(-1)); they partitioned between the dissolved and particulate phases and BDE-209 was the dominant congener, followed by BDE-99 and BDE-47. TBBPA was detected in the dissolved phase only (<35-68 pg L(-1)). All selected BFRs were ubiquitous in bed sediments and levels ranged from 3,100 to 15,100 pg g(-1) and from 70 to 280 pg g(-1) (dry weight), for SigmaBDEs and TBBPA, respectively.

Polybrominated diphenyl ethers (PBDEs), polychlorinated biphenyls (PCBs), and phthalates were investigated from July 2009 to April 2010 in three fish species from the Orge river, which flows in a densely populated area of Ile-de-France. In two Cyprinidae (roach and chub) and one Percidae (perch), muscle contents in increasing order ranged as follows: 12-18 ng g(-1) dw for PBDEs (Σtri-hepta), 120-170 ng g(-1) dw for PCBs (Σ7), and 2,250-5,125 ng g(-1) dw for phthalates (Σ7). No variation was observed between contaminant contents and lipid levels. No biomagnification was found according to the trophic level for PBDEs and PCBs, whereas for phthalates the highest contents were found in perch. Seasonal variations were observed with the lowest PBDE and PCB contents occurring in July after spawning in roach and perch (p < 0.001). PBDE content followed a decreasing trend-gonad > liver > muscle-whatever the period. For PCBs, gonad and liver contents remained greater than that of muscle (p < 0.05). Our results indicate a preferential accumulation of halogenated compounds in gonad and liver outside the reproduction period. Bioaccumulation factors for PCBs in muscle were significantly correlated to their chlorination degree in perch (p < 0.01) and roach (p < 0.01). In roach, that correlation slope was by decreasing importance order as follows: gonad > liver > muscle. The biota-sediment accumulation factors varied from 0.1 to 29.2, from 1.6 to 4.8, and from 1 to 123.5 for PBDEs, PCBs, and phthalates, respectively. These results contribute to document the use of freshwater fish as bioindicators of river quality.