In this work, home-made Polar Organic Chemical Integrative Samplers (POCIS) were studied for passive sampling of 15 endocrine disrupting compounds (4 alkylphenols and steroid hormones) in influent and effluent samples of wastewater treatment plants (WWTPs) as well as up- and downstream of the receiving river water.

POCIS calibration at laboratory conditions was carried out using a continuous-flow calibration system. The influence of the exposure position of the POCIS within the calibration device, horizontal or vertical, to the water flow direction was evaluated. While the sampling rates of most of the target substances were not affected by the sampler position, for cis-ADT, E1, E2 and E3, the vertical position provided the highest analyte accumulation. Hence, the POCIS samplers were preferably exposed vertical to the water flow in overall experiments. Using the continuous-flow calibration device, lab-based sampling rates were determined for all the target compounds (R_SBPA = 0.0326 L/d; R_ScisADT = 0.0800 L/d, R_SE1 = 0.0398 L/d, R_SEQ = 0.0516 L/d, R_STT = 0.0745 L/d, R_SE2 = 0.0585 L/d, R_SEE2 = 0.0406 L/d, R_SNT = 0.0846 L/d, R_SPG = 0.0478 L/d and R_SE3 = 0.1468 L/d), except for DES, MeEE2, 4tOP, 4OP, 4NPs, where the uptake after 14 days POCIS exposure was found to be insignificant or indicated a no linear behaviour. Recoveries from POCIS extractions were in the range between 71 and 152% for most of the target analytes except for DES and E3 with around 59%. Good precision of the sampling procedure up till 20% was observed and limits of detection were at ng/L level. Two deuterated compounds ([²H₃]-E2 and [²H₄]-EQ) were successfully tested as performance reference compounds (PRC, ke[H32]−E2 = 0.0507 L/d and ke[H42]−EQ = 0.0543 L/d)).

Finally, the POCIS samplers were tested for monitoring EDCs at two wastewater treatment plants, in Halle and Leipzig (Germany). BPA, E1, EQ, E2, MeEE2, NT, EE2, PG and E3 were quantified and their time-weighted average concentrations calculated on the basis of the lab-derived sampling rates were compared with the results based on conventional grab samples. While the influent concentration of BPA, cisADT, E1, TT, PG, EE2 reached the μg/L level, the rest of the target analytes were determined at ng/L. The analyte concentrations in the effluent never exceed ng/L level except for BPA. The concentration determined by spot sampling was partially lower (BPA, E1, TT) or comparable (EQ, E2, EE2, PG, E3) to the concentration obtained by POCIS using performance reference compounds (PRC).