Microchemical Journal 136 (2018) 56–60

Contents lists available at ScienceDirect

Microchemical Journal

journal homepage: www.elsevier.com/locate/microc

Combining passive sampling and tandem mass spectrometry for the

determination of pharmaceuticals and other emerging pollutants in
drinking water
Emanuele Magi ⁎, Marina Di Carro, Cristiana Mirasole, Barbara Benedetti
Department of Chemistry and Industrial Chemistry, University of Genoa, Via Dodecaneso 31, 16146 Genoa, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Passive sampling and liquid chromatography-tandem mass spectrometry can be profitably employed to detect
Received 29 July 2016 emerging contaminants in waters at very low concentration levels. In this work, two types of Polar Organic
Received in revised form 26 October 2016 Chemical Integrative Sampler (POCIS) were subjected to calibration at two different temperatures to calculate
Accepted 28 October 2016
the sampling rates of eight emerging pollutants (five pharmaceuticals, two perfluorinated compounds and caf-
Available online 29 October 2016
feine). Results obtained changing the temperature from +5 to +25 °C showed a limited influence on the sam-
Keywords:
pling rate values for all the selected analytes. Preliminary evaluations on storage life-time of POCIS devices
Passive sampling were also taken into account.
HPLC-MS/MS After calibration, samplers were deployed in the inlet and the outlet of two drinking water treatment plants in
POCIS calibration Northwestern Italy, for two and four weeks; the extracts were then analyzed by means of LC-MS/MS in multiple
Emerging contaminants reaction monitoring mode, which provided high sensitivity and allowed the detection of the selected compounds
Pharmaceuticals at the low ng L−1 level. Three analytes were measured in both treatment plants: the two perfluorinated com-
Water analysis pounds, in the range 2.93–13.42 ng L−1, and caffeine, in the range 0.07–0.93 ng L−1.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction introduced in 2004 [6], contains a sorbent phase sandwiched between

two membranes and, once exposed in water, is able to sample and con-
In recent years, environmental studies have partially shifted their at- centrate hydrophilic contaminants. Since POCIS are usually deployed for
tention from classical pollutants to the so-called “emerging contami- periods of time up to several weeks, they are able to pre-concentrate
nants”. These substances have been detected in the environment, and analytes from a large volume of water. Moreover, the continuous sam-
in particular in waters, at very low concentrations (ng L− 1 to μg L− 1 pling allows to take into account episodic events that are not easily
levels) [1–2]; nonetheless their presence should not be neglected, due identified by classical spot sampling [7].
to the unknown effects of long-term exposition. In fact, even though As a result of prolonged exposition, POCIS provides the time-
their low environmental concentrations are usually considered non- weighted average (TWA) concentration of a compound, derived from
harmful for humans, the contemporaneous presence of many pollutants the fluctuations of contamination levels [6]. In fact, the amount of
could cause unpredicted synergistic effects [3]. chemicals found in the sorbent phase after deployment is correlated
Very sensitive methodologies are therefore required to detect with their concentration in water, mediated over time, and it depends
emerging contaminants: the choice of the appropriate preconcentration on the sampling rate (Rs), i.e. the volume of water that the POCIS is
methods and analytical technique is crucial to obtain satisfactory results able to “clear” from a specific compound in a time unit, according to
[4–5]. A successful approach can be represented by the combination of the following equation [8]:
passive sampling with sensitive analytical techniques such as HPLC-
MS. Passive sampling presents the remarkable advantage of combining Cs ¼ Cw Rs t=Ms ð1Þ
sampling and preconcentration in one step; in fact, samplers are usually
deployed for a certain period of time in order to accumulate contami- where Cs and Cw are the concentrations of the compound in the POCIS
nants, whose levels would be otherwise difficult to detect by spot sam- sorbent (ng g−1) and in water (ng L−1), respectively, Rs is the specific
pling. The Polar Organic Chemical Integrative Sampler (POCIS), first sampling kinetic constant, known as sampling rate (L day−1), t is the
exposure time (days) and Ms. is the mass of the sorbent in the POCIS (g).
⁎ Corresponding author. To calculate TWA concentrations, the sampling rates of the analytes
E-mail address: emanuele.magi@unige.it (E. Magi). must be determined through calibration experiments; this is a crucial

http://dx.doi.org/10.1016/j.microc.2016.10.029
0026-265X/© 2016 Elsevier B.V. All rights reserved.
 E. Magi et al. / Microchemical Journal 136 (2018) 56–60 57

step and can be performed in situ or in the laboratory [9–10]. In situ cal- Quantitation of the analytes was achieved using polarity switching
ibration enables to obtain specific Rs for a determined site, and takes MS in multiple reaction monitoring mode (MRM) to maximize sensitiv-
into account the physico-chemical conditions of the site itself (water ity. Two different transitions were chosen for each compound on the
temperature and flow, ionic strength, biofouling, pH…), but is costly basis of literature data [15–16]: the first and more abundant was used
and time consuming [11–12]. Laboratory calibration is the most com- for the quantitation and the second for confirmation of the results. In
mon method because of its simplicity; experiments can be performed Table 1 chemical formulas, retention times and MRM transitions are re-
using either a static approach or a recirculating flow system [13–14]. ported for the considered compounds.
TWA concentrations obtained can only provide semi-quantitative Quantitative analysis was performed by means of the internal stan-
values of water contamination, nonetheless passive samplers are pow- dard method. The internal standard concentration (Ketoprofen-d3)
erful tools to perform screening studies and to monitor sites for long pe- was maintained constant at 25 ng mL−1, while the analyte concentra-
riods in a much easier and more economical way than spot sampling. tions were 2, 5, 10, 15 and 30 ng mL−1. Each point of the respective cal-
The aim of this work was to use POCIS and tandem mass spectrom- ibration curves was the mean of three replicates. All analytes showed
etry for the determination of some emerging contaminants in drinking good linearity (R2 between 0.992 and 0.999).
water, performing the calibration of passive samplers in different condi- Limits of detection and quantitation (calculated as a signal to noise
tions. In fact, calibration was carried out comparing POCIS assembled in ratio of 3 and 10, respectively) were slightly different for each analyte;
our laboratory to commercial samplers, at different working tempera- in particular perfluorinated compounds provided the highest analytical
tures. The obtained Rs values were then employed to calculate the response with a limit of detection of 0.15 ng mL−1 and a limit of quan-
TWA concentrations in two case studies. The chemicals chosen were titation of 0.50 ng mL−1. Nonetheless, for practical reasons, the lowest
five anti-inflammatory drugs (ibuprofen, naproxen, diclofenac, point of the calibration curve (2 ng mL−1) was considered as the limit
ketoprofen and mefenamic acid), two perfluoroalkyl compounds of quantitation for all of the compounds.
(perfluorooctanoic acid and perfluorooctane sulfonate) and caffeine,
considered a tracer of anthropic contamination. 2.3. Passive samplers

2. Experimental Commercial POCIS with HLB (hydrophilic-lipophilic balance) phase

were supplied by Environmental Sampling Technologies (St. Joseph,
2.1. Standard and reagents USA). Home-made POCIS were assembled using HLB sorbent phase
(60 μm particle size), purchased from Sigma-Aldrich (Milan, Italy),
Diclofenac (DIC), ketoprofen (KET), mefenamic acid (MEF), and 0.1 μm pore size polyethersulfone (PES) membranes (Pall Italia,
naproxen (NAP), ibuprofen (IBU), ketoprofen-d3 (KET-d3), Buccinasco, Italy), with the same characteristics of the commercial
perfluorooctanoic acid (PFOA), perfluorooctanesulfonate (PFOS) and ones (200 mg as mass of the sorbent phase and 45.8 cm2 as sampler sur-
Caffeine (CAF) were obtained from Sigma-Aldrich (Milan, Italy). All face area). PES membranes were washed before assembling in a H2O/
standards were of high purity grade (N97%). Methanol (MeOH), aceto- CH3OH solution (80:20 v/v) for 24 h and then with CH3OH for 24 h.
nitrile (ACN) and acetic acid were obtained from Merck (Darmstadt, After drying in a laminar hood, the membrane-sorbent-membrane
Germany). All solvents were of analytical or chromatographic grade. layers were compressed between two stainless-steel support rings
Water was purified by Milli-Q system (Millipore, Watford, Hertford- held together with three thumbscrews and stored frozen at −20 °C.
shire, UK). Both commercial and assembled POCIS samplers were used for static
Stock solutions of nonsteroidal anti-inflammatory drugs (NSAIDs) calibration at two different temperatures: room temperature (RT,
and caffeine were prepared by dissolving each compound in CH3OH at + 25 °C) and +5 °C. The samplers were deployed in triplicate, as de-
a concentration of 2000 ng mL− 1. Working solutions of NSAIDs, scribed in the section “POCIS calibration”; upon retrieval, the samplers
perfluorinated compounds and caffeine were prepared at a concentra- were rinsed with Milli-Q water, wrapped in aluminum foil and stored
tions of 200 ng mL−1 by subsequent dilution of the stock solution in frozen at −20 °C.
MQ water. Both stock and working solutions were stored at − 20 °C. Prior to processing, the samplers were thawed and rinsed with Milli-
The working solutions at different concentration levels were prepared Q water. Each POCIS was dismantled and the sorbent was transferred by
by dilution using Milli-Q water. means of Milli-Q water into a 1 cm i.d. glass syringe cartridge fitted with
Teflon frit and glass wool. The sorbent was dried for 30 min under vac-
2.2. LC-MS/MS analysis uum. Prior to extraction, 50 μL of a 1000 ng mL− 1 solution of
ketoprofen-d3 were added into the sequestering phase, which was sub-
Analyses were performed on an Agilent Liquid Chromatograph Se- sequently eluted with 50 mL of acetone. The eluate was collected in a
ries 1200SL (consisting of a binary HPLC pump, an online vacuum flask, reduced to dryness in a rotary evaporator and redissolved in
degasser, an automatic sampler ALS and a thermostatted column com- 1 mL of methanol; this solution was diluted 1:1 with Milli-Q for the
partment and a DAD detector) coupled to an Agilent 6430 MSD triple- LC-MS/MS analyses of real samples, while, during calibration experi-
quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, ments, it was diluted 1:100 with a water - methanol mixture 50:50.
USA) with an Electrospray source. Separation of the analytes was car-
ried out by means of a Hypersil Gold Aq column (3 × 30 mm, particle 2.4. POCIS calibration
size 1.9 μm), purchased by Thermo Scientific (San Jose, CA, USA), at
60 °C. An isocratic elution with 50% Milli-Q water containing 0.1% of The Rs calibration experiments were performed for both commercial
acetic acid and 50% acetonitrile was performed with a flow rate of and home-made POCIS at two different temperatures, + 5 °C and
0.2 mL min−1 and injection volume 10 μL, allowing the separation of +25 °C. A 1.8 mL mixture containing 22.5 μg of each analyte was spiked
compounds within 10 min. Negative ionization was used to analyze into a clean 5 L-capacity beaker containing 4.5 L of tap water. The mix-
NSAIDs, PFOA and PFOS compounds, while positive ionization was ture was allowed to equilibrate for about 30 min at a stirring rate of
employed for caffeine investigation. ESI conditions for both kinds of ion- 1300 rpm using a F30 magnetic stirrer (Falc Instruments s.r.l., Italy),
ization were: drying gas flow (N2) 10 L min−1, capillary potential ± resulting in a nominal concentration of 5 μg L−1 of each compound.
3000 V, nebulizer pressure 35 psi and drying gas temperature 350 °C. For each experiment, three POCIS were suspended in the spiked solu-
Mass calibration for MS experiments was performed by infusion of ESI tion and the beaker was covered with aluminum foil. The calibration
Tuning Mixture (Agilent Technologies, Palo Alto, CA, USA). at +5 °C was performed in a fridge. Both experiments were conducted
MassHunter software was used for data acquisition and processing. for 72 h under stirring at 1300 rpm.
58 E. Magi et al. / Microchemical Journal 136 (2018) 56–60

Table 1
Chemical formulas, retention times and MRM transitions of the eight emerging pollutants.

Compounds Structure Retention time Precursor ion (m/z) Product ion (m/z)

Caffeine 1.0 195 138–110

PFOS 1.1 499 99

PFOA 1.1 413 369

Naproxen 1.8 229 169–185

Ketoprofen 1.9 253 209–197

Diclofenac 2.9 294 250–214

Ibuprofen 3.4 205 161–159

Mefenamic acid 4.3 240 196–192

Blank POCIS have been deployed in parallel with calibration, show- In our previous studies the real experimental conditions were simi-
ing no contamination by targeted compounds during the experiment, lar to a laminar flow and Rs were calculated by means of a home-made
except for caffeine, whose signal was anyway negligible. recirculating system, simulating this kind of flow [18–19].
In the two case studies described in this work, experimental condi-
2.5. Sampling rate calculation tions were different; in fact POCIS have been deployed in two drinking
water treatment plants, where the flow was rather turbulent. For this
Accumulation of contaminants by passive samplers typically follows reason the calibration experiments were performed in a stirred beaker
first-order kinetics, which includes an initial integrative phase, followed (Fig. 1), as described in Section 2.4. Besides, it is reported that this cali-
by curvilinear and equilibrium-partitioning phases. POCIS requires a rel- bration method, called “static calibration” (closed system, with analyte
atively long sampling time before reaching equilibrium and thus accu- spiking at the beginning of the experiment), is suitable when the mole-
mulation remains in the integrative phase for a long period after cules studied are not quickly degraded or adsorbed, or when the calibra-
deployment. In this period of the linear uptake, the concentration Cs of tion duration is short, i.e., less than one week [20].
a chemical accumulated in the sampler is described by the Eq. (1), al- Experiments of this work were performed in January, when the
ready reported in the Introduction: water temperature was close to +5 °C; as well known, also temperature
can influence the sampling rate values [17].
Cs ¼ Cw Rs t=Ms Although POCIS samplers can be purchased ready to use, in the last
years there is a growing tendence by research groups working with pas-
where Cs is the concentration of the compound in the POCIS sorbent sive sampling to self-assemble the samplers, with the same characteris-
(ng g−1), Cw is the concentration of the compound in water (ng L−1), tics as the original ones [21–23]. To the best of our knowledge, in the
Rs is the specific sampling kinetic constant, known as sampling rate literature the storage life-time of a POCIS sampler before use is not
(L day−1), t is the sampling period (days) and Ms. is the mass of the sor- taken into account; the only indication provided by the manufacturer
bent in the POCIS (g). The experimental data obtained from the labora- is to store the samplers at −20 °C. In our laboratories a suitable protocol
tory calibration tests were used to calculate the sampling rates (Rs) of for POCIS assembly was developed [18]; nevertheless some commercial
the target chemical compounds according to Eq. (1). For each experi- samplers were purchased in 2009 and stored at −20 °C. Therefore we
mental condition Rs were calculated as the mean of the three values ob- decided to perform the same calibration experiments using both these
tained by POCIS exposed in triplicate. “old” commercial POCIS and home-made POCIS assembled just before
their use.
3. Results and discussion Results are reported in Table 2, comparing Rs values obtained expos-
ing both types of samplers in tap water, spiked with 5 μg L−1 of each
3.1. Experimental sampling rates compound for 72 h under stirring at RT and +5 °C.
Results obtained in both temperature conditions highlight that Rs
A calibration of a passive sampler is necessary for each compound values of commercial samplers are decidedly lower than those of assem-
since there are no “standard” sampling rates [11]. Various calibration bled POCIS, except for caffeine. For some analytes, for example
methods have been described in the last years, leading to different sam- diclofenac, the difference is about one order of magnitude. Values ob-
pling rates values for the same analyte and making comparison between tained in this work for the assembled POCIS are similar to the sampling
them difficult [9]. rates summarized in the review of Morin et al. [10] reported in the last
In the choice of the calibration method it is important to consider the column of Table 2, and to Rs observed in our previous works [18–19]. Al-
environmental conditions in which samplers will be deployed, in partic- though a rather wide range of Rs is shown for each analyte, the sampling
ular the presence of a laminar or turbulent flow [17]. rates obtained in this work for commercial POCIS seem excessively low,
 E. Magi et al. / Microchemical Journal 136 (2018) 56–60 59

indicating a possible alteration of the sorptive performances of the sam-

plers, likely due to the long storage time. Therefore we decided to aban-
don the “old” commercial POCIS and use the home-made POCIS for the
case studies, assembling them just before the deployment.
As regards the influence of temperature on the sampling rates of the
assembled POCIS (Table 2), both increases and decreases in Rs values
can be observed changing the temperature from +5 to +25 °C; the dif-
ferences are in general limited and can be considered not significant
taking into account the standard deviation. Anyway, in the subsequent
experiments in the drinking water treatment plants, the sampling
rates used for the calculation of the TWA concentrations were those ob-
tained at +5 °C (Table 2).

3.2. Application of POCIS to drinking water

The home-made POCIS were deployed in the inlet and the outlet of
two drinking water treatment plants, located in Northwestern Italy in
January 2016; average water temperature in both sites was relatively
stable (+5 ± 2 °C) over the entire period. Four samplers were exposed
in the influent and four in the effluent of Site 1; two of these were re-
trieved after for 14 days, while the other two after 28 days. In Site 2,
four POCIS were deployed only in the effluent with the same timing.
The inlet water of both plants consists of surface water which is sub-
sequently subjected to mild chemical-physical treatments. Laboratory-
derived Rs from the calibration experiment at +5 °C were used to calcu-
late the TWA concentrations (Table 3) from the amount of the analytes
sequestered by POCIS and measured by LC-MS/MS.
Only three of the eight analytes - namely PFOS, PFOA and caffeine -
were measured, while pharmaceuticals were below the detection limit
in both sites. Levels of concentration were very low, with caffeine in
the sub-ng L−1 range, suggesting a negligible anthropic contamination
of the raw surface waters. Higher concentrations were measured for
PFOS and particularly for PFOA, with a maximum value of
13.42 ng L−1. These perfluorinated compounds are involved in industri-
al processes for the manufacturing of grease- and waterproof products;
they are highly stable in the environment and have been inserted in the
European priority list for future regulation of drinking water (Annex X
of the Water Framework Directive 2000/60/EC).
In Site 1 the level of contamination is similar in the inlet and in the
Fig. 1. Stirred beaker calibration experiment: three POCIS in a spiked solution with target outlet, indicating that analyte concentration is not reduced during the
analytes. The mixture was stirred at 1300 rpm with the beaker covered with aluminum treatment applied in the drinking water plant, in accordance with liter-
foil. ature data [24]. Besides, the TWA concentrations for PFOA after 2 and

Table 2
Experimental Rs values obtained exposing assembled and commercial POCIS for 72 h in tap water spiked with 5 μg L−1 of each compound, at +5 and +25 °C.

Rs (L day−1) (+5 °C) Rs (L day−1) (+25 °C) Literature Rs [10]

Assembled POCIS Commercial POCIS Assembled POCIS Commercial POCIS

Caffeine 0.033 ± 0.001 0.141 ± 0.011 0.035 ± 0.007 0.102 ± 0.009 0.127 (±0.021)
0.151 (±0.018)
0.096 (±0.008)
PFOS 0.076 ± 0.001 0.021 ± 0.004 0.079 ± 0.007 0.048 ± 0.018 n.r.
PFOA 0.262 ± 0.008 0.063 ± 0.006 0.209 ± 0.052 0.052 ± 0.019 n.r.
Naproxen 0.075 ± 0.006 0.032 ± 0.014 0.056 ± 0.015 0.018 ± 0.005 0.392 (±0.024)
0.298 (±0.016)
0.239 (±0.009)
0.200 (±0.037)
0.116 (±0.053)
0.083 (±0.055)
Ketoprofen 0.068 ± 0.019 0.008 ± 0.024 0.087 ± 0.012 0.031 ± 0.013 0.135 (±0.035)
0.083 (±0.078)
Diclofenac 0.133 ± 0.004 0.013 ± 0.004 0.117 ± 0.015 0.008 ± 0.001 0.166 (±0.052)
0.092 (±0.055)
Ibuprofen 0.121 ± 0.011 0.030 ± 0.011 0.091 ± 0.018 0.022 ± 0.007 0.348 (±0.052)
0.254 (±0.019)
0.204 (±0.004)
0.197 (±0.013)
Mefenamic acid 0.055 ± 0.007 0.008 ± 0.002 0.089 ± 0.020 0.006 ± 0.001 n.r.

n.r.: not reported in ref. [10].

60 E. Magi et al. / Microchemical Journal 136 (2018) 56–60

Table 3
Estimated mean concentrations (TWA ± standard deviation) of the emerging pollutants detected in drinking water treatment plants (ng L−1) in two sites of Northwestern Italy.

Site 1 Site 2

2 weeks 4 weeks 2 weeks 4 weeks

In Out In Out Out Out

PFOS 2.93 ± 0.49 3.99 ± 0.43 7.98 ± 2.14 9.39 ± 1.68 7.74 ± 1.20 8.97 ± 0.52
PFOA 13.42 ± 3.32 11.48 ± 2.03 11.76 ± 2.23 8.56 ± 2.80 9.63 ± 0.25 9.06 ± 0.83
Caffeine 0.86 ± 0.05 0.93 ± 0.24 0.19 ± 0.04 0.15 ± 0.03 n.c. 0.07 ± 0.06

n.c.: not calculated because of an accidental contamination.

4 weeks of deployment are comparable, indicating that the contamina- [7] N. Morin, J. Camilleri, C. Cren-Olivé, M. Coquery, C. Miège, Determination of uptake
kinetics and sampling rates for 56 organic micropollutants using “pharmaceutical”
tion level is rather constant. PFOS exhibits a different behavior, showing POCIS, Talanta 109 (2013) 61–73.
an increase in the TWA concentration measured after 4 weeks, both at [8] R. Greenwood, G. Mills, B. Vrana, Passive Sampling Techniques in Environmental
Monitoring, Elsevier Science, 2007.
the inlet and the outlet. Nonetheless, considering also the relatively [9] C. Harman, I.J. Allan, E.L.M. Vermeirssen, Calibration and use of the polar organic
high standard deviation observed for data regarding the 4 week deploy- chemical integrative sampler-a critical review, Environ. Toxicol. Chem. 31 (2012)
ment, further experiments would be necessary to confirm and interpret 2724–2738.
[10] N. Morin, C. Miège, M. Coquery, J. Randon, Chemical calibration, performance, vali-
this trend as well as the behavior of caffeine. dation and applications of the polar organic chemical integrative sampler (POCIS)
in aquatic environments, TrAC Trends Anal. Chem. 36 (2012) 144–175.
[11] C. Harman, M. Reid, K.V. Thomas, In situ calibration of a passive sampling device for
4. Conclusions
selected illicit drugs and their metabolites in wastewater, and subsequent year-long
assessment of community drug usage, Environ. Sci. Technol. 45 (2011) 5676–5682.
In this work, a combination of passive sampling and tandem mass [12] R. Jacquet, C. Miège, P. Bados, S. Schiavone, M. Coquery, Evaluating the polar organic
chemical integrative sampler for the monitoring of beta-blockers and hormones in
spectrometry was employed for the determination of eight emerging wastewater treatment plant effluents and receiving surface waters, Environ. Toxicol.
pollutants in the influent and the effluent of two drinking water treat- Chem. 31 (2012) 279–288.
ment plants. Two types of POCIS samplers were subjected to calibration [13] A. Arditsoglou, D. Voutsa, Passive sampling of selected endocrine disrupting com-
pounds using polar organic chemical integrative samplers, Environ. Pollut. 156
experiments at two different temperatures; results showed that Rs (2008) 316–324.
values of commercial samplers stored for several years were excessively [14] C. Harman, O. Boyum, K.V. Thomas, M. Grung, Small but different effect of fouling on
the uptake rates of semipermeable membrane devices and polar organic chemical
low, suggesting an alteration of the sorptive performances of the integrative samplers, Environ. Toxicol. Chem. 28 (2009) 2324–2332.
samplers. [15] M. Farre, M. Petrovic, D. Barcelo, Recently developed GC/MS and LC/MS methods for
Sampling rate values for home-made POCIS were in accordance to determining NSAIDs in water samples, Anal. Bioanal. Chem. 387 (2007) 1203–1214.
[16] R. Loos, J. Wollgast, T. Huber, G. Hanke, Polar herbicides, pharmaceutical products,
those reported in the literature; the effect of the temperature was not perfluorooctanesulfonate (PFOS), perfluorooctanoate (PFOA), and nonylphenol
particularly relevant. and its carboxylates and ethoxylates in surface and tap waters around Lake
Rs obtained for assembled samplers at +5 °C were used for the cal- Maggiore in Northern Italy, Anal. Bioanal. Chem. 387 (2007) 1469–1478.
[17] B. Vrana, I.J. Allan, R. Greenwood, G.A. Mills, E. Dominiak, K. Svensson, J. Knutsson, G.
culation of the TWA concentration of the analytes in two drinking water Morrison, Passive sampling techniques for monitoring pollutants in water, TrAC,
treatment plants, where only PFOS, PFOA and caffeine were measured at Trends Anal. Chem. 24 (2005) 845–868.
[18] M. Di Carro, L. Bono, E. Magi, A simple recirculating flow system for the calibration of
trace and ultra-trace levels. polar organic chemical integrative samplers (POCIS): effect of flow rate on different
water pollutants, Talanta 120 (2014) 30–33.
Appendix A. Supplementary data [19] S. Tanwar, M. Di Carro, E. Magi, Innovative sampling and extraction methods for the
determination of nonsteroidal anti-inflammatory drugs in water, J. Pharm. Biomed.
Anal. 106 (2015) 100–106.
Supplementary data to this article can be found online at doi:10. [20] H.X. Li, P.A. Helm, C.D. Metcalfe, Sampling in the Great Lakes for pharmaceuticals,
personal care products, and endocrine-disrupting substances using the passive
1016/j.microc.2016.10.029.
polar organic chemical integrative sampler, Environ. Toxicol. Chem. 29 (2010)
751–762.
References [21] J.C. Carlson, J.K. Challis, M.L. Hanson, C.S. Wong, Stability of pharmaceuticals and
other polar organic compounds stored on polar organic chemical integrative sam-
[1] D. Fatta-Kassinos, S. Meric, A. Nikolaou, Pharmaceutical residues in environmental plers and solid-phase extraction cartridges, Environ. Toxicol. Chem. 32 (2013)
waters and wastewater: current state of knowledge and future research, Anal. 337–344.
Bioanal. Chem. 399 (2011) 251–275. [22] G. Poulier, S. Lissalde, A. Charriau, R. Buzier, F. Delmas, K. Gery, A. Moreira, G.
[2] S. Tanwar, M. Di Carro, C. Ianni, E. Magi, Occurrence of PCPs in Natural Waters From Guibaud, N. Mazzella, Can POCIS be used in water framework directive (2000/60/
Europe, Springer, Berlin Heidelberg, 2014 1–35. EC) monitoring networks? A study focusing on pesticides in a French agricultural
[3] A. Nikolaou, S. Meric, D. Fatta, Occurrence patterns of pharmaceuticals in water and watershed, Sci. Total Environ. 497–498 (2014) 282–292.
wastewater environments, Anal. Bioanal. Chem. 387 (2007) 1225–1234. [23] A. Zenker, H. Schmutz, K. Fent, Simultaneous trace determination of nine organic
[4] S.D. Richardson, T.A. Ternes, Water analysis: emerging contaminants and current is- UV-absorbing compounds (UV filters) in environmental samples, J. Chromatogr. A
sues, Anal. Chem. 86 (2014) 2813–2848. 1202 (2008) 64–74.
[5] W.W. Buchberger, Current approaches to trace analysis of pharmaceuticals and per- [24] C. Flores, F. Ventura, J. Martin-Alonso, J. Caixach, Occurrence of perfluorooctane sul-
sonal care products in the environment, J. Chromatog. A 1218 (2011) 603–618. fonate (PFOS) and perfluorooctanoate (PFOA) in N.E. Spanish surface waters and
[6] D.A. Alvarez, J.D. Petty, J.N. Huckins, T.L. Jones-Lepp, D.T. Getting, J.P. Goddard, S.E. their removal in a drinking water treatment plant that combines conventional and
Manahan, Development of a passive, in situ, integrative sampler for hydrophilic or- advanced treatments in parallel lines, Sci. Total Environ. 461–462 (2013) 618–626.
ganic contaminants in aquatic environments, Environ. Toxicol. Chem. 23 (2004)
1640–1648.