An improved Chemcatcher-based method for the integrative passive sampling of 44 hydrophilic micropollutants in surface water – Part B: Field implementation and comparison with automated active sampling

Abstract

Integrative passive sampling is particularly useful in the monitoring of hydrophilic contaminants in surface water, but the impact of hydrodynamics on contaminant uptake still needs to be better considered. In part A (Glanzmann et al., 2023), Chemcatcher-like hydrophilic samplers (i.e., SDB-RPS extraction disks covered by PES microporous membranes) were calibrated to determine the sampling rates RS of 44 hydrophilic contaminants (pesticides, pharmaceuticals, industrial products) taking into account the hydrodynamic conditions. In this study, Chemcatcher-like passive sampling devices that allowed co-deploying hydrophilic samplers and performance reference compounds (PRC)-spiked silicone disks were tested in a Swiss river with intermediate water velocities (5–50 cm s−1, 23 cm s−1 on average) during 11 consecutive 14-day periods. The PRC dissipation from silicone disks – combined with the calibration data from part A – allowed to determine in-situ RS that took into account hydrodynamic conditions. The obtained aqueous time-weighted average (TWA) concentrations were found to be robust with good concordance between duplicates (mean quotient of 1.16 between the duplicates). For most measurements (76 %), TWA concentrations showed no major difference (<factor 2) from concentrations obtained with automated sampling (14-day composite samples). This observation was also valid for TWA concentrations calculated with extrapolated RS at infinite water velocity (RS,MAX), revealing that the added value of using in-situ RS compared to RS,MAX is limited above intermediate water velocities (>20 cm s−1). RS from the literature (RS,LIT) – obtained at water velocities between 8 and 37 cm s−1 – were also shown to provide comparable TWA concentrations in the studied hydrodynamic conditions (average water velocity of 24 cm s−1). The estimated errors due to the use of RS,MAX or RS,LIT rather than in-situ RS are given as a function of the water velocity to determine in which conditions the developed method is required (or not) in monitoring programs.