Proposal of a new empirical model with flow velocity to improve time-weighted average concentration estimates from the Polar Organic Chemical Integrative Samplers.

It is now widely recognized that the sampling rate of Polar Organic Chemical Integrative Samplers (POCIS) is significantly affected by flow velocity, which can cause a consequent bias when determining time-weighted average concentrations (TWAC). We already observed the desorption of deisopropylatrazine (DIA) over time when added to the receiving phase of a POCIS. This desorption rate was particularly influenced by flow velocity, in an agitated water environment in situ. In the method presented here, we calibrated 30 pesticides under controlled laboratory conditions, varying the flow velocity over four levels. We simultaneously studied the desorption rate of DIA-d5 (a deuterated form of DIA) over time. An empirical model based on a power law involving flow velocity was used to process the information from the accumulation kinetics of the compounds of interest and elimination of DIA-d5. This type of model makes it possible to consider the effect of this crucial factor on exchange kinetics, and then to obtain more accurate TWACs with reduced bias and more acceptable dispersion of results.

Adaptation of the o-DGT for the sampling of 12 hormones: calibration, performance evaluation, and recommendation.

This work highlights the methodology for the development of diffusive gradients in thin films (o-DGT) through its adaptation for 12 natural and synthetic hormones belonging to three different families (estrogens, progestins, and androgens). A reliable strategy must be applied during o-DGT lab adaptation to avoid issues related to the analysis (i.e., presence of matrix effects in grab or passive samples) but also to the o-DGT configuration (i.e., undesirable sorption or desorption, lack of performance with insufficient elution or unreliable diffusion coefficient). To avoid analytical issues due to the presence of salts in grab samples, CaCl2 exposure solutions must be used on a lab-scale development to monitor the hormone concentration. The selected o-DGT was composed of an Oasis® HLB binding gel and a diffusive gel in agarose because they provided better performance than polyacrylamide gels (i.e., higher elution factors and more repeatable diffusion coefficients). The elution factors of the binding gel were then from 0.79 ± 0.13 to 1.04 ± 0.13 (RSD < 15%) and the diffusion coefficients at 25 °C were from 4.07 ± 0.24 to 5.49 ± 0.28 × 10-6 cm2 s-1 (RSD < 9%). A laboratory exposure to a synthetic solution was performed to check the consistency with the DGT quantification model validating the calibration parameters for all hormones (except 17α-ethinylestradiol with a bias of 40%). Therefore, the o-DGT configuration is suitable for sampling hormones in the natural environment with LOQDGT ranging from 0.3 to 6.6 ng L-1.

Interest of a new large diffusive gradients in thin films (L-DGT) for organic compounds monitoring: On-field comparison with conventional passive samplers.

In this work, the performances of a Large Diffusive Gradients in Thin films (L-DGT, i.e., a DGT based on a Chemcatcher® holder with a 5-fold larger sampling area) were compared on-field with the conventional DGT and the Polar Organic Chemical Integrative Sampler (POCIS) for the monitoring of a wide range of organic contaminants (i.e., 65 pesticides and metabolites, 53 pharmaceuticals and 12 hormones). These three passive samplers were simultaneously deployed in four rivers during 14 days. Their performances were then evaluated according to their detection and quantification capacities and their physical robustness. The results obtained confirm the advantages of the L-DGT over the conventional DGT regarding its sensitivity but also its robustness during field deployment. The POCIS provides the higher sensitivity, allowing the detection of more organic compounds compared to the DGT and, to a lesser extent, the L-DGT. However, both L-DGT and DGT reduces the uncertainty on the determination of the time-weighted average concentrations (CW), mainly due to the narrow range of variation of their calibration parameters. Indeed, for a given compound, CW can vary up to only a 3-fold factor with DGT and L-DGT compared to a 2 to 10-fold factor (up to 50) with POCIS. Thus, the L-DGT appears to be more suitable than DGT in low-contaminated contexts, which require higher sensitivity, or than POCIS when a CW determination is needed. For a qualitative evaluation however, the POCIS remains the most suitable passive sampler.

Multivariate Tiered Approach To Highlight the Link between Large-Scale Integrated Pesticide Concentrations from Polar Organic Chemical Integrative Samplers and Watershed Land Uses.

This paper presents a multi-step methodology to identify relationships between integrative pesticide quantifications and land uses on a given watershed of the Adour-Garonne Basin (Southwestern France). In fact, a large amount of pesticide concentration data was collected from 51 sites located in the Adour-Garonne Basin for a 1 year monitoring period in 2016. The sampling devices used here were polar organic chemical integrative samplers (POCIS), which provided time-weighted average concentration estimates. For each study site, its associated watershed and land cover distribution were determined using Corine Land Cover 2012 (CLC 2012) and Geographic Information System (GIS). The large-scale data were analyzed using multivariate statistical analyses, such as hierarchical cluster analysis (HCA) and principal component analysis (PCA). HCA grouped the 51 sites into five clusters with similar primary land uses. Next, the integrated pesticide concentration and land use distribution data sets were analyzed in a PCA. The key variables responsible for discriminating the sample sites showed distribution patterns consistent with specific land uses. To confirm these observations, pesticide fingerprints from sites with contrasting land uses were compared using a waffle method. The overall multivariate approach allowed for the identification of contamination sources related to their likely initial use, at the watershed level, that could be useful for preventing or containing pesticide pollution beyond simply acting on areas at risk.

Development of a multi-hormone analysis method by LC-MS/MS for environmental water application using diffusive gradient in thin films.

An analysis method for four families of hormones (estrogens, progestins, androgens and prostaglandins), dedicated to an efficient water monitoring with passive sampling, was developed using a liquid chromatography tandem mass spectrometry with triple quadrupole coupling and universal electrospray ionisation. Thirteen natural and synthetic hormones in ultra-pure water could be analysed in a single run according to the French Standard NF T90-210: calibration range of 0.1 (except for 17ß-Estradiol, Estriol, Estrone and Diethylstilbestrol, from 0.5 µg/L; and Ethinylestradiol, from 1 µg/L) to 20 µg/L with linear regressions (R2 ≥ 0.96), maximum accuracy deviations of 30% at intermediate fidelity for three concentration references (1, 10 and 20 µg/L) and instrumental LOQs from 0.05 to 1 µg/L. The stability of 11 hormones (10 µg/L) was studied under several storage conditions and sample evaporation. All selected hormones were stable for 60 days at -18 °C, 7 days at 4 °C and 7 days at 20 °C but continued drying flow after evaporation should be avoided, especially for 17α-Estradiol, Estrone and Diethylstilbestrol. Observed matrix effects using o-DGT extracts (diffusive gradient in thin-film sampler for polar organics) containing an environmental matrix varied from 24 to 92% but all matrix effects were corrected with IS use. Therefore, the developed method, coupled with o-DGT, was tested with the o-DGT deployment in rivers. Using diffusion coefficients from the literature or lab determined, the concentrations in the rivers varied for Estrone from 1.8 ng/L to 2.5 ng/L, and for Androstenedione from 0.4 to 1.1 ng/L.

Adaptation of diffusive gradients in thin films technique to sample organic pollutants in the environment: An overview of o-DGT passive samplers.

The adaptation of the diffusive gradients in thin films technique (DGT) to sample organic pollutants in the environment, called o-DGT has been performed since 2011 for various types of organic compounds (e.g. pesticides, pharmaceuticals, hormones, endocrine disrupting chemicals, household and personal care products). To sample these different compounds, configuration of the samplers (mainly receiving phase and diffusive gel) has to be adapted. Up-to-date, sampling of 142 organic compounds by this passive sampler have been tested. This review provides the state-of-art of o-DGT passive sampler development, describing theory and modelling, calibration, configuration of the devices, and field applications. The most used configurations were agarose-XAD-18 and agarose-HLB configuration. o-DGT can be used to sample soils and most of natural waters (range of pH 4-9 and ionic strength 0.001-0.1 M). This review discusses current limitation of o-DGT in light of the feedback of DGT use to sample inorganic contaminants. It mainly concern the low sampling rates currently obtained by o-DGT compared to other passive samplers. This weakness could be compensated in the future with new sampler's design allowing an increase in exposure area.

Limitation of flow effect on passive sampling accuracy using POCIS with the PRC approach or o-DGT: A pilot-scale evaluation for pharmaceutical compounds.

Flow velocity is known to alter passive sampling accuracy. We investigated the POCIS (Polar Organic Chemical Integrative Sampler) with PRC (Performance Reference Compounds) approach and Diffusive Gradients in Thin Films samplers (o-DGT) to limit the effect of flow on the quantification accuracy of ten model pharmaceuticals compounds (0.16 ≤ log KOW ≤ 4.51). POCIS and o-DGT samplers were exposed for seven days in controlled pilot-scale (hundreds of liters) experiments under quiescent or flowing (2 < V < 18 cm s-1) conditions. Under flowing conditions, both POCIS-PRC and o-DGT efficiently limited the flow effect and led, in most cases, to biases within analytical uncertainty (20%). Under quiescent conditions, o-DGT performed accurately (bias < 30% for most compounds) whereas the PRC approach was unsuitable to improve upon the accuracy of POCIS (PRC was unable to desorb). Therefore, both approaches are helpful in limiting the effects of flow on accuracy, but only o-DGT is efficient in quiescent conditions. However, o-DGT currently suffers from poorer sensitivity compared to POCIS, but the future development of o-DGT devices with wider windows could overcome this limitation.

Semi-continuous pharmaceutical and human tracer monitoring by POCIS sampling at the watershed-scale in an agricultural rural headwater river.

Pharmaceutical monitoring (37 pharmaceuticals and 3 human tracers) was conducted in a headwater streams in southwest France, an area characterized by a low population density with an elderly population (30% > 60 years old) and extensive agriculture (cow cattle breeding). Polar Organic Chemical Integrative Sampler (POCIS) were exposed for 14-day consecutive periods in 2016 at three sampling points. Three human wastewater tracers and 20 pharmaceuticals commonly used for human and/or cattle were quantified in headwaters. Succession of small Wastewater Treatment Plant (WWTP), non-collective sanitation, discharges of untreated effluents as well as the river ability to dilute discharged wastewater, mainly explain the pharmaceuticals and human tracers concentrations. Pharmaceutical loads were time-dependent and were higher during cold season due to increase of pharmaceutical consumption. In contrast, better degradation and/or sorption onto river biofilms in warm season induced the decrease of headwater pharmaceutical content. The headwaters streams were contaminated by compounds found in other type of watershed, but ß-blocker were the compounds quantified in higher concentration with frequencies of 100%, which was consistent with the elderly population living in the watershed. Specific compounds (sulfamerazine and sulfamethoxazole) used to cattle medical care were detected in waters, but at a low content.

Two sampling strategies for an overview of pesticide contamination in an agriculture-extensive headwater stream.

Two headwaters located in southwest France were monitored for 3 and 2 years (Auvézère and Aixette watershed, respectively) with two sampling strategies: grab and passive sampling with polar organic chemical integrative sampler (POCIS). These watersheds are rural and characterized by agricultural areas with similar breeding practices, except that the Auvézère watershed contains apple production for agricultural diversification and the downstream portion of the Aixette watershed is in a peri-urban area. The agricultural activities of both are extensive, i.e., with limited supply of fertilizer and pesticides. The sampling strategies used here give specific information: grab samples for higher pesticide content and POCIS for contamination background noise and number of compounds found. Agricultural catchments in small headwater streams are characterized by a background noise of pesticide contamination in the range of 20-70 ng/L, but there may also be transient and high-peak pesticide contamination (2000-3000 ng/L) caused by rain events, poor use of pesticides, and/or the small size of the water body. This study demonstrates that between two specific runoff events, contamination was low; hence the importance of passive sampler use. While the peak pesticide concentrations seen here are a toxicity risk for aquatic life, the pesticide background noise of single compounds do not pose obvious acute nor chronic risks; however, this study did not consider the risk from synergistic "cocktail" effects. Proper tools and sampling strategies may link watershed activities (agricultural, non-agricultural) to pesticides detected in the water, and data from both grab and passive samples can contribute to discussions on environmental effects in headwaters, an area of great importance for biodiversity.

Passive sampling of anionic pesticides using the Diffusive Gradients in Thin films technique (DGT).

DGT passive samplers using Oasis® HLB or Oasis® MAX sorbent were developed for anionic pesticides sampling. They were tested using four model compounds (i.e. bentazon, chlorsulfuron, ioxynil and mecoprop). Polyacrylamide diffusive gel was found to be more suitable than agarose gel for most anionic pesticides sampling. An elution procedure was optimized and diffusion coefficients were determined for quantitative use of the samplers. Depending on the DGT configuration used (HLB or MAX), accuracies better than 30% were demonstrated in laboratory for pH from 3 to 8 and ionic strengths from 10-2 to 1 M. Combined with the effective binding capacities of samplers (≥9 µg for each pesticide) and limits of quantification of the method (≤13 ng.L-1 using Q-TOF detector) monitoring of numerous aquatic systems can be expected. Except for ioxynil, accurate quantifications were demonstrated in laboratory using a spiked natural water for HLB-DGT whereas MAX-DGT did not give satisfactory results. A further in situ validation was performed in two rivers and showed identical detection frequency between HLB-DGT and POCIS of anionic pesticides (bentazon and mesotrione) whereas calculated concentrations, although within the same order of magnitude, could differ (<70%). HLB-DGT could therefore constitute an interesting alternative to other passive samplers for the monitoring of several anionic pesticides in aquatic systems but more work is required for quantification of molecules from hydroxybenzonitrile chemical group (ioxynil).

Improvement of POCIS ability to quantify pesticides in natural water by reducing polyethylene glycol matrix effects from polyethersulfone membranes.

The presence of polyethylene glycol compounds (PEG) in extracts from polar organic chemical integrative samplers (POCIS) was shown by high resolution time-of-flight mass spectrometry. PEG compounds, which are released by polyethersulfone (PES) membranes used to build POCIS, can induce matrix effects during quantification of performance reference compounds (PRC, DIA-d5) and target pesticides by mass detection, even after chromatographic separation. Dilution of POCIS extracts can reduce this matrix effect, but dilution may induce a decrease in POCIS performance, primarily for quantification limits. To reduce PEG interference during chromatographic analysis, a simple non-damaging washing protocol for PES membranes is proposed. The method consists of 2 successive baths of washing solution (140 mL per membrane) of ultrapure water (UPW) and methanol (50/50), stirred at 300 rotations per minute (rpm), followed by a final membrane rinse with UPW (140 mL). The signal from PEG compounds was significantly decreased for washed membranes (between 4 and 6 fold lower). After field deployment, total ion current chromatograms of extracts from POCIS built with washed PES membranes did not display a significant PEG fingerprint. This led to improved quantification accuracy for compounds co-eluting with PEG, i.e. PRC (performance and reference compound, DIA-d5) and some pesticides and metabolites. With washed membranes, an accurate quantification of PRC and pesticides sampled by POCIS was indeed possible without a large extract dilution; 10 times instead of the 25 times needed in unwashed conditions. Assuming that the PRC approach corrects for environmental conditions and sampling rates (Rs), a proper PRC (DIA-d5) quantification significantly improved pesticide time weighted average concentration (TWAC) determination in natural water after field deployment.

Coupling passive sampling and time of flight mass spectrometry for a better estimation of polar pesticide freshwater contamination: Simultaneous target quantification and screening analysis.

The aim of this study was first to develop and validate an analytical method for the quantification of 35 polar pesticides and 9 metabolites by ultra-high-performance-liquid chromatography combined with a high resolution time-of-flight mass spectrometer detector (UHPLC-(Q)-TOF). Various analytical conditions were investigated (eluent composition and mass parameters) to optimize analyte responses. Analytical performance (linearity, limit of quantification, and accuracy) was then evaluated and interference in the extract of a passive sampler exposed in freshwater (POCIS: Polar Organic Chemical Integrative Sampler) was studied. The proposed quantification method was validated for 43 compounds with variation of calibration slopes below 10% in environmental matrix. For the unvalidated compound DIA (atrazine-desisopropyl: an atrazine metabolite), interference increased the error of concentration determination (50%). The limits of quantification obtained by combining POCIS and UHPLC-(Q)-TOF for 43 target compounds were between 0.1 (terbuthylazine) and 10.7 ng/L (acetochlor). Secondly, the method was successfully applied during a 14-day POCIS river exposure, and gave concentration values similar to a more commonly used triple quadrupole detector regarding concentration, but allowed for the detection of more compounds. Additionally with the targeted compound quantification, the (Q)-TOF mass spectrometer was also used for screening non-target compounds (other pesticides and pharmaceuticals) in POCIS extracts. Moreover, the acquisition of full scan MS data allowed the identification of the polyethylene glycol (PEG) compounds which gave unresolvable interference to DIA, and thus questions the ability of DIA to be used as performance reference compound (PRC) to determine sampling rates in situ. This study therefore illustrates the potential, and proposes a pathway, of UHPLC-(Q)-TOF combined with POCIS in situ pre-concentration for both quantitative and screening analyses of organic contaminants in water.