Science of the Total Environment 687 (2019) 355–368

Contents lists available at ScienceDirect

Science of the Total Environment

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

Suspect, non-target and target screening of emerging pollutants using

data independent acquisition: Assessment of a Mediterranean
River basin
Alexander Ccanccapa-Cartagena a,c,⁎, Yolanda Pico a, Xavier Ortiz b, Eric J. Reiner b
a
Environmental and Food Safety Research Group of the University of Valencia (SAMA-UV), Desertification Research Centre CIDE (CSIC-UV-GV), Moncada-Naquera Road km 4.5, 46113 Moncada,
Valencia, Spain
b
Ontario Ministry of the Environment and Climate Change, 125 Resources Road, Toronto, ON M9P 3V6, Canada
c
Escuela Profesional de Antropología, Universidad Nacional de San Agustín de Arequipa¡, Av. Venezuela s/n, 04000 Cercado, Arequipa, Peru

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Suspected, non-target and target

screening to establish River contamina-
tion profile
• Identification 68 contaminants against a
database of 2000 compounds
• Non-target screening identifies addi-
tional compounds as Eprosartan
• New findings enrich target screening up
to 171 pesticides and 33 pharmaceuticals

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

Article history: A single workflow based on three approaches (target, suspected and non-target screening) using liquid chromatog-
Received 31 January 2019 raphy coupled to quadrupole-time-of-flight mass spectrometry (LC-QTOF-MS) in data independent acquisition
Received in revised form 3 June 2019 mode (DIA) was developed to assess the presence of emerging pollutants (EPs) in water and sediments from a Med-
Accepted 4 June 2019
iterranean River Basin. Identification of potential contaminants was based on mass accuracy, isotopic ratio pattern,
Available online 6 June 2019
theoretical fragmentation, and retention time using Waters UNIFI software. In the suspect screening against a library
Editor: Damia Barcelo containing 2200 components, 68 contaminants were tentatively identified, 6 of which were confirmed and quanti-
fied with analytical standards. Non-target screening (NTS) required additional manual processing and the aid of an
Keywords: on-line database (ChemSpider) to tentatively identify compounds. Eprosartan, an antihypertensive drug not in-
Turia river cluded in the library used for suspected screening, was confirmed and semi-quantified. The identification of
Target screening Eprosartan proved the workflow to be functional for NTS. Target screening of 171 pesticides and 33 pharmaceuticals
Non-target screening and personal care products (PPCPs) including the compounds confirmed using suspect (6) and non target
Suspected screening (1) screening achieved monitoring of the most abundant contaminants from the head to the mouth of the Turia
Contamination profiling
High-resolution-mass-spectrometry

⁎ Corresponding author at: Environmental and Food Safety Research Group (SAMA-UV), Desertification Research Centre CIDE (CSIC-UV-GV), Faculty of Pharmacy, University of
Valencia, Burjassot, Valencia, Spain and Escuela Profesional de Antropología, Universidad Nacional de San Agustín de Arequipa, Av. Venezuela s/n, 04000 Cercado, Arequipa
E-mail addresses: accanccapa@unsa.edu.pe, alexander.ccanccapa@uv.es (A. Ccanccapa-Cartagena).

https://doi.org/10.1016/j.scitotenv.2019.06.057
0048-9697/© 2019 Elsevier B.V. All rights reserved.
356 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368
basin to establish their spatial distribution. QTOF-MS screening versatility with its high-resolution capability allows
for a comprehensive assessment of EPs in the aquatic environment.
1. Introduction
Priority and emerging pollutants (EPs) potentially harmful to wild-
life, such as pesticides, pharmaceuticals and personal care products
(PPCPs), hormones, phthalates or artificial sweeteners, have been de-
tected in the aquatic environment (Andrés-Costa et al., 2014; Bolong
et al., 2009; Campo et al., 2014; Carmona et al., 2014; Ccanccapa et al.,
2016a; Daughton and Ternes, 1999; Ibáñez et al., 2005; Masiá et al.,
2014b; Silva et al., 2011). The current legislation do not yet regulate
the presence of these EPs in the environment (Bletsou et al., 2015;
Mailler et al., 2016; Petrović et al., 2003; Sanchez-Prado et al., 2010).
Once EPs are released into the environment, they are subject to biotic
and abiotic transformation through hydrolysis, photolysis, oxidation,
and microbial metabolism; thereby, generating transformation prod-
ucts that may be more persistent and/or toxic than their parent com-
pounds and possibly present at relatively higher concentrations
(Bletsou et al., 2015; Gago-Ferrero et al., 2016; Picó and Barceló,
2015). EPs and their transformation products (TPs) typically migrate
to surface and groundwater via run-off, sub-soil tile drains and waste-
water treatment plants (WWTPs) (Ccanccapa et al., 2016a; Farré et al.,
2008; Gonzalez-Rey et al., 2015; Hug et al., 2014).
The most common technique to analysis EPs is liquid
chromatography–mass spectrometry (LC–MS) using triple quadrupole
(QqQ), quadrupole linear ion trap (QTRAP), quadrupole time-of-flight
(QTOF) or Orbitrap (Andrés-Costa et al., 2016b; Andrés-Costa et al.,
2014; Campo et al., 2014; Gonzalez-Rey et al., 2015; Hollender et al.,
2018; Ibáñez et al., 2017; Ieda et al., 2019; Martínez-Domínguez et al.,
2016; Paíga et al., 2016; Pedrouzo et al., 2009; Silva et al., 2011; Storck
et al., 2016; Yao et al., 2016). Target screening methods, where the
chemicals are selected in advance, cover a relatively small proportion
of these organic contaminants. This can result in bias (due to the pre-
selection), as potential chemical stressors may not be detected (Gago-
Ferrero et al., 2015). In recent years, there is a growing trend to analyze
environmental samples beyond the intended target compounds using
non-target or unknown screening (Arsand et al., 2018; Chiaia-
Hernandez et al., 2014; Masiá et al., 2014b; Masiá et al., 2013; Müller
et al., 2011; Newton et al., 2018; Park et al., 2018; Ruff et al., 2015;
Singer et al., 2016; Zedda and Zwiener, 2012a). High-resolution MS
(HRMS) represented by QTOF and Orbitrap combines sensitive full-
spectrum data with high mass resolution (N25,000 FHWM) and mass
accuracy (mass error b 5 ppm). In theory, the presence of an unlimited
number of compounds can be investigated at the required sensitivity,
without pre-selection of analytes or even without having reference
standards available (Bletsou et al., 2015; Gago-Ferrero et al., 2016;
Masiá et al., 2014a). In fact, the ultimate suspect and non-target ap-
proaches reported for environmental matrices (water and sediments)
are based on semi-quantitative analysis (using peak intensities) if stan-
dard substances are not available (Arsand et al., 2018; Park et al., 2018)
or identification as last step (using seven golden rules) (Kind and Fiehn,
2007; Schymanski et al., 2014b; Sibiya et al., 2019). Although suspect
and non-target analysis has grown rapidly, they are still uncommon
for most environmental monitoring agencies and environmental scien-
tists due to they are not well-resolved yet and represent a challenge
since inability to provide quantitative information about identified
compounds.
HRMS can be used (Baz-Lomba et al., 2016; Gago-Ferrero et al.,
2016; Glauner et al., 2016; Rosnack et al., 2016) in the data-
dependent acquisition (DDA) or data independent analysis (DIA). The
later is commonly preferred because it avoids specific selection during
© 2019 Elsevier B.V. All rights reserved.
LC-MS analysis by co-selection and co-fragmentation (Masiá et al.,
2016; Moran et al., 2014; Zhou et al., 2017). The most common way to
perform DIA is by simultaneously obtaining the mass spectra at high
and low collision energy (CE) to alternately acquire spectra of all detect-
able precursors and their products, which are then, time aligned and
available for exploration. One of the challenges of HRMS is how to pro-
cess and rationalize the high amount of information obtained in a reli-
able way. The precursor ion accurate mass and its isotopical pattern
attain elucidation of the most plausible empirical formula, which is
the starting point of complicated searching to propose a tentative struc-
ture. The information provided by the retention time (tR) and the frag-
ments of the product ion mass spectrum is an important milestone in
structure elucidation. The software algorithms to process and combined
chromatographic and spectral information are still under development,
mainly due to limited comparability of MS/MS libraries and consequent
difficulty to develop models that predict fragmentation or retention
time. The suspected screening is carried out against a database that con-
tains, at least, information on empirical formula and accurate mass of a
more or less long list of compounds and additionally, can also contain
information on their retention time in a defined LC system and the “in
silico” or experimental MS/MS fragmentation compiled in libraries.
Non-target screening is more difficult and requires the help of software
capable of going beyond the instrument's own database by searching
online public internet databases such as ChemSpider or PubChem. In ad-
dition, several additional tools as “in silico” fragmentation simulations
can be used to decrease the number of potential chemical structures
(Bletsou et al., 2015; Chiaia-Hernandez et al., 2014; Ibáñez et al., 2017;
Masiá et al., 2014b; Masiá et al., 2013; Picó and Barceló, 2015; Ruff
et al., 2015; Schymanski et al., 2015).
The aim of this study was to apply HRMS technology for establishing
EPs contamination profile in a River Basin by suspected HRMS-DIA
screening against a commercial database (Waters corporation) with in-
formation on 2200 compounds (pesticides, pharmaceuticals, personal
care products and toxins) and non-target screening that combines
searching in the ChemSpider online database both using UNIFI platform
that helps to simplify the workflow. Target screening with 171 pesti-
cides and 33 PCPP including the compounds confirmed (with purchased
standards) in the suspect and non-target steps was the last step. Previ-
ous reports showed EPs as the greatest threats to freshwater ecosys-
tems, especially in Mediterranean watersheds, which are
characterized by periodical low flows that may exacerbate chemical ex-
posure (Andrés-Costa et al., 2016a; Andrés-Costa et al., 2016b;
Ccanccapa et al., 2016b; Masiá et al., 2014b). The selected study case is
the Turia River (East of Spain; sampled in 2016) a typical Mediterranean
River but the proposed workflow has a global application to any other
environmental matrix (ground water, lake water, reservoir water, efflu-
ents from WWTP, etc.). This study also shows the potential and versatil-
ity of a HRMS workflow as a routine tool for the comprehensive analysis
of pollution in aquatic environments, which so far has not been fully
exploited. Furthermore, it also includes as an interesting aspect, not pre-
viously reported to our knowledge: the analysis of water and sediment
samples taken in the same location by HRMS and the comparison of the
data obtained.
2. Materials and methods
A list of materials used including solvents and standards is provided
in the Supporting information (SI-1 and Table S-1).
 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368
2.1. Site description and sampling
The area of study is located in the Central-Eastern part of Spain. The
Turia River is 243 km long and drains an area of approximately
6250 km2, making it one of the most important rivers in the Mediterra-
nean basin (Carmona et al., 2014; Ccanccapa et al., 2016a; Navalon et al.,
2010). 22 water and 29 sediment samples were taken as homogenously
as possible through the course of the river from its source to its mouth
on June of 2016 (Fig. S-1). Additionally, two effluent samples collected
from the Waste Water Treatment Plants (WWTP) of Pinedo I and
Pinedo II in Valencia (sampling points are georeferenced in Table S-2,
Supplementary material) were included (SI-2 Sampling details).
2.2. Extraction procedure
2.2.1. Water samples
Analyte isolation and pre-concentration from water samples is car-
ried out by off-line Solid Phase Extraction (SPE) as described previously
(Masiá et al., 2013). Water samples (200 mL) were passed through an
Oasis HLB (200 mg sorbent/6 mL) cartridge (Waters Corporations, Mil-
ford, MA) at flow rate ca. 10 mL min−1 previously conditioned with
methanol-dichloromethane (50:50, v/v) and water using vacuum. The
analytes were eluted with 10 mL of methanol-dichloromethane mix-
ture, evaporated to dryness and re-dissolved in 1 mL of mobile phase
to achieve an enrichment factor of 200 (SI-3 Sample extraction details).
2.2.2. Sediment samples
Stepwise, 5 mL of distilled water, 5 mL of methanol with 5 mL of
Macllvaine-EDTA (100 mL of 0.1 M of citric acid, 62.5 mL 0.2 M
Na2HPO4 and 6.05 g of Na2-EDTA) buffer were added to 1 g of sediment,
previously lyophilized. This mix was sonicated for 30 min in an S 120H
Elma Ultrasonic (Singen, Germany) and centrifuged for 6 min at
3000 rpm with a 5810 R Centrifuge from Eppendorf (Hamburg,
Germany). The supernatant was collected; 200 mL of distilled water
was added, and clean up by SPE as water samples.
2.3. Ultra-high pressure liquid chromatography
The chromatography column and mobile phases were selected
based on a pre-established screening method (Masiá et al., 2013). The
chromatograph was a Waters Acquity UPLC system (Milford, MA,
USA). Chromatographic separation was carried out using a column
Luna C18 (15.0 cm × 0.21 cm) with a 3 μm particle size (Phenomenex,
Torrance, USA). A binary mobile phase of A (10 mM formic acid in
water) and B (10 mM formic acid in methanol) was applied with follow-
ing program: 0–15 min, 10% B; 15–18.5 min, 95% B; 18.5–19 min, 95% B;
19–23 min, 10% B. The analytical column and the sample manager were
kept at 35 °C and 7 °C, respectively. An aliquot of 10 μL was injected into
UPLC-QTOF-MS with a flow rate of 0.4 mL min−1.
2.4. Quadrupole time-of-flight mass spectrometry
A Xevo G2-S Q-TOF mass spectrometer (Waters, Milford, MA USA)
was used in positive ESI (majority of EPs could be detected in the posi-
tive ion mode) (Schymanski et al., 2014a) for acquisition in two modes:
MS for target analysis and DIA (MSE) for suspect and non-target analy-
sis. The MS mode was set as full scan mode between 100 and 1200 m/z;
scan speed at 4 scan/s; collision energy off; mass accuracy calibrated to
b2 ppm; mass resolution N24 K and XIC tolerance of 20 mDa. The DIA
mode allows both precursor and fragmentation data to be simulta-
neously acquired during a single run. The DIA method consists of
three functions, the first (low energy, LE) applied a collision energy of
4 eV, the second function (high energy, HE) acquired spectra through
a collision energy ramp of 10–45 eV and the third function acquired
the lock mass data for internal on-the-fly mass calibration. The MS
mass range was 50–1200 m/z with a scan time of 0.25 s in continuum
357
mode, preserving the peak shape of the exact-mass precursor and prod-
uct ions. The source conditions to achieve maximum intensities were:
capillary voltage 2 kV, sample cone, 80 V, source offset 80 V, source tem-
perature 125 °C, desolvation temperature 250 °C, cone gas flow rate
50 L h−1, desolvation gas (N2) flow rate 600 L h−1.
During data acquisition, the mass to charge ratio was corrected using
an external reference (Lock-Spray™) consisting of a 0.2 mg mL−1 solu-
tion of leucine-enkephalin (Waters, Milford, MA, USA) infused continu-
ously at 10 mL min−1 via a lock spray interface. This generated a
reference ion in positive mode at m/z 556.2771 that was used for real-
time mass corrections in order to maintain the mass accuracy and
reproducibility.
2.5. Data processing
Data acquisition and processing were performed using the UNIFI
screening platform (Waters Corporation, Milford MA, USA) on suspect
and non-target screening (DIA) and Masslynx™ v4.1 (Waters Corpora-
tion, Milford MA, USA) on target screening (MS) (Fig. S-2).
The workflow (Fig. 1) and identification confidence criteria used on
suspect and non-target screening were based on those described by
Krauss et al. (2010), Schymanski et al. (2015) and Pedersen et al.
(2013). The analysis method for DIA was performed as follows: Firstly,
all the continuum data was peak-detected using a 3D peak algorithm
and can be considered as two mountain ranges; one for the low-
energy data, and one for the high-energy data, respectively. The 3D
peak detection algorithm within UNIFI locates the summits of the
peaks in a given mountain range. The peak summits, or apexes, detected
for the data shown in Fig. 2-A correspond to the dots overlaid on the
data in Fig. 2-B. Each dot represents a single ion characterized by a m/
z value and a retention time (RT), together with a measure of intensity
which is derived from the volume of the peak. Thus, a unique m/z-RT
pair is derived for each ion. Secondly, the identity of the compounds
was established by setting target mass tolerance at 5.0 ppm and frag-
ment match tolerance at 2.0 mDa. The mass defect setting was enabled
by the selecting H+ adduct. Halogen match settings included four Cl and
three Br atoms.
2.5.1. Suspect screening workflow
The library used in the suspected method (Waters, Milford, MA USA)
consisted of 2200 components including pesticides, pharmaceuticals,
personal care products among other toxins. Each component included
mass spectra, retention time, molecular formula and structure.
Suspected screening was performed matching the compounds detected
in the samples with this library, which included the structure (as a.mol
file) for theoretical fragmentation prediction. The tentative identifica-
tion was based on retention time (tolerance 0.3 min), accurate mass
(b 5 ppm), isotope fit (N90%) and coincidence of the fragments obtained
with the theoretical fragmentation (5 or more fragments in common, 2
mDa accuracy), chromatographic ratio (b 1), mass peak resolution (N
7.000) and response (N25.000) (F1). Fig. S-2 shows the interface of
UNIFI platform where the data are processed and matched as follow
as: (5A) template of workflow, (5B) component identification list,
(5C) selected ion chromatogram of a single component corresponding
to 5B, and finally (5D) the respective mass spectrum and fragmentation
(example of cucurmenol identified by the library).
2.5.2. Non-target screening workflow
The UNIFI software is able to process the peak list and organize it in
components. Each component is the group of data which relates to the
same chemical entity (mass, retention time, isotopic pattern, fragments,
adducts, drift time, etc.…). Components not identified using suspect
screening were candidates for non-target screening- The non-target
screening was performed in six steps.
358 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368
Fig. 1. Workflow of suspect and non-targeted screening.
1) Selecting: The components of each sample obtained by suspected
screening are reported in a list of identified and non-identified com-
ponents exportable to Microsoft Excel (including observed m/z, tR,
response, and identification status) in order to be checked by the op-
erator. The eight samples (water and sediment) with more identi-
fied components were selected for further non-target analysis of
the unidentified ones. Initially, there were around 15,000–20,000
components per selected sample.
2) Filtering: in order to reduce the number of components to a more af-
fordable amount. The filter 2 (F2) designed for this step was based
on chromatographic peak width (b 1), mass resolution (N 7000)
and peak intensity (N 25,000 detector counts); it reduced the com-
ponents to 500–1000 per sample.
3) Elucidation tool set: The first step is to determine the empirical for-
mula of the unknown components that have pass the filter (the soft-
ware uses the exact mass of the intact precursor component, exact
mass and abundance ratio of the isotopic peaks, as well as the
exact mass of the fragments corresponding to the precursor ions).
The UNIFI system uses an algorithm, i-FIT™; a similar concept to
the Seven Golden Rules by (Kind et al., 2007). Potential molecular
formulas were displayed with an i-FIT confidence percentage
(representing the relative certainty of the assigned molecular for-
mulas). All the potential empirical formulas with an i-FIT confidence
N90% were considered. In addition to the elemental composition
containing C, H, N, O, and S, other elements commonly present in en-
vironmental contaminants such as F, Cl and Br can be selected as a
filter in order to reduce the number of compounds. Although the
tentative identification was performed without this filter, it was
also applied to focus specially on those contaminants more toxic
for the environment. Then, the parameter halogen (Kaufmann
et al., 2017) in match (containing Cl, Br, F) was added to the filter
resulting in a reduction of 50–100.
4) Matching online database: The next step in this identification pro-
cess is to search for possible names and structures associated with
the empirical formula or exact mass through online libraries. The
UNIFI software can be connected directly to on-line databases such
as ChemSpider Library \\including PubChem (over 10,000,000
structures) and Thomson Pharma libraries (over 2,000,000)\\,
which allows the possible structures to be obtained quickly and
automatically.
5) In silico ranking: the UNIFI program has a software that helps to con-
firm which of the possible structures is the most likely candidate by
comparing their theoretically predicted fragments using a combina-
torial fragmentation approach to the experimental spectrum of the
unidentified compound obtained at high energy (Gago-Ferrero
et al., 2015; Hollender et al., 2017; Kaufmann et al., 2017; Zedda
and Zwiener, 2012b). The combinatorial fragmentation approach at-
tempts to match observed product ions to the candidate structures
by disconnecting covalent bonds in the candidate structure.
Disconnecting a bond gives a score, with the lowest score being
the most probable. The candidate structures are then ranked based
on predicted intensity (%-matched intensity of total intensity) and
a number of matched product ions, the most probable structure
being the one with highest predicted intensity and largest number
of matched product ions. When several structures matched equally
well, the most probable candidate structure was picked based on
the product ion scoring. Fragment match was set N5 with b2 mDa
of mass error. Elucidation was based on accurate mass (b 5 ppm),
i-FIT (N 90%), matching fragments (b 2 mDa, N N 5) and the relative
intensity of the fragments. 6). Tentative identification: All those
compounds that followed the filter application were ranked based
on number of citations (N10), number of match fragmentation (N5)
and intensity (N25,000). If a relevant compound was identified,
(HR)MS spectra were sought in the literature for confirmation. The
comparison was performed manually for thirteen compounds iden-
tified in this step. Reference standards were required for the final
identity confirmation.
2.5.3. Target screening
Target screening was focused on all 171 pesticides and 33 PPCPs, in-
cluding the compounds identified on suspected and non-target
 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368 359

Fig. 2. 3D LC/MS plot of a significant sample (WWTP-II).

screening. Quantification was by external standard calibration (pesti-
cides 0.5, 1, 5, 10, 20, 40, 80 ng mL−1 and pharmaceuticals 20, 200 and
2000 ng mL−1 in the vials injected equivalent to pesticides 2.5, 5, 25,
50, 100, 200 and 400 ng L−1 and pharmaceuticals 100, 1000 and
10,000 ng L−1 in water).
The extraction methods have been widely reported using different
detection systems (Carmona et al., 2014; Masiá et al 2014; Ccanccapa
et al., 2016a; Ccanccapa et al., 2016b). Taking into account that recovery
depends mostly on the extraction procedure, the values reported in
those studies are appropriate.
The low reported levels (LRLs) were b10 ng L−1 for water and
b10 ng g−1 for sediments.
The matrix effects were not studied. These constitute an additional
uncertainty at established concentrations that may be under- or
overestimated. However, the results obtained can be compared be-
tween the different matrices.
360 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368
2.6. Sediment-water distribution of EPs (Kp)
Using the data of pesticides and pharmaceutical concentrations in
surface sediment (Cs) and in water samples (Cw), the partition coeffi-
cients (Kp) between sediment and water were calculated:
Kp ¼ Cs=Cw
3. Results
3.1. Suspect screening
Suspect screening was performed on 10 water and 10 sediment
samples of 32 water and 29 sediment samples collected. Criteria to se-
lect the samples were based on the distribution of the contamination
in previous studies about the occurrence of EPs in the Turia basin
(Andrés-Costa et al., 2016a; Carmona et al., 2014; Ccanccapa et al.,
2016a). From these studies, it was concluded that highest pollution
levels were mostly in the mouth of the river basin, specifically at points
in the headwater (Alfambra tributary) and areas close to the WWTP
(Fig. S-1).
Structure assignments based on exact mass, isotope pattern and
fragmentation are an important step to identify non-target contami-
nants. However, its theoretical nature sometimes matched false posi-
tives (e.g. dexamethasone Fig. 3) in terms of mass accuracy, tR, or
fragmentation. Finally, tentative structural assignments were confirmed
with their authentic standards. Standards used in the present study are
described in Supplemental information (Table S-1). Standards were
injected and compared with the sampled candidates based on retention
time, exact mass and fragmentation. Dexamethasone, isoprocarb,
bupirimate and penoxsulam did not match with the samples candidates
in term of exact mass (mass error N 5 ppm) and retention time
(highlighted in red with clear blue background in Table S-4 and S-3).
However, imazalil, tebuconazole, nytempriran, metalaxyl,
Fig. 3. False positive peak identified as Dexamethasone in the suspected screening by the Library of Waters (Waters Corporation) A) sediment sample (observed average accurate mass
393.207183) and B) standard injected (observed average accurate mass 393.207185).
thiabendazole and oxytetracicycline were an excellent match with the
sample and could be monitored in the target screening in all water
and sediment samples.
Following this approach, 68 compounds were identified (Table S-3).
Forty-five pollutants were identified in water samples, whereas 42 pol-
lutants were identified in sediment. With this step, all the identified
ð1Þ compounds were at level 2 (probable structure). A total of 51 pesticides,
15 pharmaceuticals and 2 mycotoxins were detected. All the identified
compounds had b5 ppm mass error and the tR were from 3.4 min (tri-
methoprim) to 17.6 min (betamethasone).
In water samples, the pollutants identified were pesticides and phar-
maceuticals (Table S-4A). Each water sample showed at least 7 different
compounds. The samples with highest number of contaminants were
WWTP-II (17), WWTP-I (10) and ALF-5 (13). Most of the compounds
detected in water samples were also detected at higher concentrations
in wastewater. However, 15 compounds (names highlighted in bold
and grey rows) were detected only in wastewater (Table S-4).
Pesticides, pharmaceuticals and mycotoxins were also identified in
sediment samples (Table S\\4B). The most polluted were ALF1 (23),
ALF3 (10) y ALF5 (11) and TUR4 (12). The sediments of the tributary
Alfambra (ALF) in the Turia River identified additional EPs; ALF5
reaching critical levels (more EPs, especially for ALF5, which was one
of the most polluted samples).
3.2. Non-target screening
Based on the suspect screening results, the eight highest contami-
nated samples (4 water samples (ALF5, GUA2, GUA6, TUR8), 2 effluents
(WWTP-I and WWTP-II) and 2 sediment samples (ALF5 and TUR4)
were selected and screened for non-target EPs. This approach was ap-
plied to those candidate masses that were identified as a peak but that
appear in the suspected screening with an “no identified” status. Ini-
tially, there were around 15,000–20,000 features per sample. The filter
2 (F2) designed for this step was based on chromatographic peak
width (b 1 rel. chrom. resolution), mass resolution (N 7000) and peak
 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368 361

intensity (N 25,000 detector counts); it reduced the features to
500–1000 per sample. Then, the parameter halogen match (containing
Cl, Br, F) was added to the filter resulting in a reduction to 50–100 fea-
tures per sample. Finally, candidate structures were selected via the elu-
cidation of the molecular formula and then tentatively identified via in
silico fragmentation. Elucidation was based on accurate mass (b
5 ppm), i-FIT (N 90%), matching fragments (b 2 mDa, N N 5) and frag-
mentation relative intensity. ChemSpider was used for the identifica-
tion; the results were ranked according the fragment matches (N5)
and citation (N10).
Table 1summarizes the compounds identified as main candidates in
each sample. These candidates are mostly analgesics, anti-
inflammatories, anaesthetics, antidepressants and antihypertensive
agents. Several tentative candidates not covered in the suspected
screening library were identified in sediment and water samples with
the exception of GUA2 and TUR8. WWTP-I, WWTP-II and ALF5 (both,
water and sediment samples) showed the highest number of tentative
candidates.
Alprenolol, tanacaine and 5-carboxamidotryptamine (5-CT) were
tentatively identified in the effluents of both WWTP, ampyrone and
tepantadol only in WWTP-I and eprosartan only in WWTP-2. In the
MS obtained at high energy, N5 experimental fragments overlaps with
those of the theoretical fragmentation. Furthermore, many scientific ar-
ticles reported the occurrence of these compounds in the environment
(N 29 citations found). The main criteria to rank the compounds and se-
lect candidates were the number of experimental fragments that match
the theoretical ones and the number of papers that reports their occur-
rence. Rolipram, paraxypropine, ibuverine and cyclopent were identi-
fied in sample ALF5. These compounds have log P values b 3
indicating that they can be soluble in water.
The last step was selecting the candidates (potential hits) to confirm
their presence in the samples using an authentic standard. Eprosartan
and 5-CT analytical standards were purchased. 5-CT was selected be-
cause its presence in two important samples (WWTP-I and WWTP-II)
and the number of citation reporting its environmental occurrence
(40) and common fragments of the theoretical fragmentation (8) in
ChemSpider data base. These clues were important in deciding to pur-
chase a real standard for confirmation. Finally, both standards were
injected. 5-CT did not match with its corresponding sample in terms
of time tR and product ions. However, eprosartan was an excellent
match with the sample and could be quantified in the target screening
in all water and sediment samples (Fig. 4).
3.3. Target screening
Under this approach, data evaluation on target screening is based on
the high resolution UPLC–QTOF MS full scan. The total run time was
20 min, with pesticides eluting between 1.27 and 19.14 min and phar-
maceuticals between 1.84 and 17.34 min. Thirty-two and twenty-nine
water and sediment samples, respectively, were analyzed. This method
includes 171 pesticides and 33 PPCPs, and includes some of the com-
pounds that after being identified in suspected and non-target screen-
ing tests were added to the target analysis. Thirty-three pesticides and
7 pharmaceuticals were detected in water samples; whereas 34 pesti-
cides and 6 pharmaceuticals were detected in sediment samples.
Results obtained for water and sediment samples in Turia River 2016
expressed as average and frequency of detection are summarized in
Table 2. In water samples, pesticides with the highest concentrations
were 3-hydroxycarbofuran (778 ng L−1), dimethoate (203 ng L−1)
and thiabendazole (533 ng L−1); whereas pharmaceutical compounds
with the highest concentrations were eprosartan (1970 ng L−1), carba-
mazepine (1200 ng L−1), trimethropim (2570 ng L−1) and lidocaine
(2400 ng L−1).
In sediment samples, pesticides with the highest concentrations
were: ethofumesate (483 ng g−1) and 3-hydroxycarbofuran
(83 ng g−1); whereas pharmaceutical compounds with the highest con-
centrations were oxytetracycline (2030 ng g−1), diazepam
(1090 ng g−1) and doxycycline (845 ng g−1). Confirmed compounds
in suspect and non-target screening were included in the quantification.
The compounds confirmed in suspect screening as imazalil (24%) was
the most frequent and thiabendazole (533 ng L−1) had the highest con-
centration in water samples. In sediment samples the most frequent
were tebuconazole (31%) and imazalil (17%) and the highest concentra-
tion was represented for tebuconazole (13 ng L−1).
Eprosartan, a pharmaceutical confirmed through non-target
workflow WWTP-II, was present in 18% of the samples and the maxi-
mum concentration was 1970 ng L−1 localized is the mouth of the basin.
Samples can contain several EPs. The monitoring of the river showed
that 81% of water samples and 86% of sediment samples contained at
least 5 pesticides and 21% of the water samples contained N10 pesti-
cides. The pharmaceuticals showed less co-occurrence than pesticides
(at least 12% of the samples contained N2 compounds).
4. Discussion
Different studies conducted on water and sediment samples aimed
to develop target, suspect and non target screening approaches to iden-
tified emerging pollutants (Table 3). Most of them have adopted
(Schymanski et al., 2014a) criteria, establishing different levels of con-
firmation, such as exact mass, unequivocal molecular formula, tentative
candidates, probable structure and confirmed structure (with authentic
standard) and some of them reached to a semi-quantitative stage (peak
area). (Park et al., 2018) focused on PPCPs in water samples from
Yeongsan river (Korea), (Newton et al., 2018) in drinking water in

Table 1
List of compounds identified by ChemSpider on line database and UNIFI platform by the non-target screening.

Compound m/z Fragments # Frag tR No. Cites Water Sediment

1 Alprenolol 250.1814 102–84 14 2.7 29 WWTP-I –

WWTP-II
2 Ampyrone 204.113 171–125 7 5.06 737 WWTP-I
3 Tapentadol 222.1855 107–120 13 5.45 41 WWTP-I
4 Safingol 302.3054 268–284 6 10.92 57 ALF5
5 Rolipram 276.1602 260–233 8 14.69 1176 ALF5
6 Paroxypropione 151.0751 137–119 12 14.12 380 ALF5
7 Tinabinol 375.2353 331–345 4 8.6 18 ALF5-GUA6 TUR4
8 Cyclopent 320.2228 206–163 9 6.23 25 GUA6 ALF5
9 Ibuverine 291.1966 163–209-233 17 14.69 12 GUA6-ALF5
10 Tanacaine 235.181 174–86 5 3.62 562 WWTP-I –
WWTP-II
11 5-CT 204.1132 159–161-146 8 5.06 40 WWTP-I –
WWTP-II
12 Eprosartan 425.1539 273–207 18 7.54 319 WWTP-II
13 Crotetamide 227.1762 102–116–124-142 5 4.33 33 TUR4
 362
A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368
Fig. 4. Chromatogram (tR) and high energy (collision energy ramp of 10–45 eV) fragment spectra of Eprosartan identified by non-target screening in (A) water sample and (B) standard.
 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368 363

Table 2
Maximum and Mean concentrations and frequency of detection (Freq) of EPs in water and sediments using the target method.

Water Sediment

Concentration (ng L−1) Freq (%) Concentration (ng g−1) Freq (%)

Maximum Mean Maximum Mean

Pesticides
3-Hydroxycarbofuran 778.00 321.51 32 (97) 83.80 31.60 28 (97)
Azoxystrobin 22.50 0.70 1 (3) n.d n.d n.d
Bromuconazole 35.50 1.11 1 (3) 6.80 6.80 1 (3)
Buprofezin n.d n.d n.d 0.90 0.90 2 (6)
Butoxycarboxim 60.50 1.89 1 (3) n.d n.d n.d
Chloroxuron n.d n.d n.d 10.70 7.64 13 (44)
Cyproconazole n.d n.d n.d 4.80 4.80 1 (3)
Carbendazim 46.50 9.64 9 (27) n.d n.d n.d
Dimethoate 203.00 6.34 1 (3) n.d n.d n.d
Dimethomorph 41.50 3.67 4 (12) 6.10 6.10 4 (13)
Emamectin-benzoate b1a n.d n.d n.d 6.40 6.40 1 (3)
Epoxiconazole 32.50 1.02 1 (3) 6.40 6.40 1 (3)
Etaconazole 40.50 23.91 21 (63) 6.80 6.62 19 (65)
Ethofumesate n.d n.d n.d 483.70 137.44 20 (69)
Ethiprole 31.50 0.98 1 (3) n.d n.d n.d
Etoxazole 23.00 1.56 3 (9) n.d n.d n.d
Fenarimol 37.00 21.31 21 (63) 5.90 5.78 18 (62)
Fenobucarb n.d n.d n.d 6.80 6.80 1 (3)
Fenpropimorph n.d n.d n.d 2.20 2.08 4 (13)
Fenuron n.d n.d n.d 14.90 7.65 20 (68)
Fludioxinil n.d n.d n.d 18.60 18.60 1 (3)
Fenoxycarb 48.00 3.68 7 (21) n.d n.d n.d
Flusilazole 21.00 0.66 1 (3) 4.30 4.30 1 (3)
Furalaxyl 48.00 3.68 7 (21) n.d n.d n.d
Flutolanil n.d n.d n.d 2.00 1.95 2 (6)
Hexaflumuron n.d n.d n.d 6.30 6.30 1 (3)
Hexaconazole 28.50 0.89 1 (3) n.d n.d n.d
Imazalil 103.40 9.48 8 (24) 7.90 6.54 5 (17)
Imidacloprid 81.00 12.37 9 (27) n.d n.d n.d
Isoproturon 40.50 4.08 4 (12) n.d n.d n.d
Mefenacet 24.50 14.66 20 (60) 4.80 4.48 11 (38)
Metalaxyl 96.00 6.94 5 (15) 5.30 5.30 1 (3)
Metconazole n.d n.d n.d 5.60 5.60 1 (3)
Neburon 30.50 27.34 30 (90) 7.40 6.06 29 (100)
Nitenpyram n.d n.d n.d 5.00 5.00 1 (3)
Oxadixyl 76.50 2.39 1 (3) n.d n.d n.d
Penconazole 29.50 0.92 1 (3) 5.90 5.90 1 (3)
Prochloraz n.d n.d n.d 10.50 7.52 5 (17)
Propamocarb n.d n.d n.d 5.20 5.20 2 (6)
Propiconazole 42.50 29.82 28 (84) 8.80 6.84 15 (51)
Pyraclostrobin 33.00 16.14 18 (54) 5.60 5.30 4 (14)
Pyrimethanil 16.00 2.58 6 (18) 2.90 2.90 1 (3)
Sulfentrazone n.d n.d n.d 8.00 8.00 1 (3)
Tebuconazole 73.50 7.75 7 (21) 13.50 6.73 9 (31)
Tebuthiuron 29.50 1.07 2 (6) n.d n.d n.d
Terbutryn 5.00 0.79 6 (18) 0.90 0.90 2 (7)
Thiabendazole 533.50 17.77 2 (6) n.d n.d n.d
Tricyclazole 51.00 1.59 1 (3) 6.40 6.40 1 (3)

Pharmaceuticals
Atorbastatin 349.50 19.05 2 (6) n.d n.d n.d
Caffeine 161.50 4.89 1 (3) n.d n.d n.d
Carbamezapine 1196.00 121.32 9 (27) 97.3 68.15 10 (34)
Chlorotetracycline n.d n.d n.d 81.5 81.50 1 (3)
Diazepam n.d n.d n.d 1085.2 78.39 2 (6)
Doxycycline n.d n.d n.d 845.4 480.65 2 (6)
Oxytetracycline n.d n.d n.d 2033.3 639.43 4 (13)
Ketoprofen 686.50 22.06 2 (6) n.d n.d n.d
Lidocaine 2407.00 209.52 6 (18) n.d n.d n.d
Trimethoprim 2569.50 145.36 3 (9) 12.7 6.35 1 (3)
Eprosartan 1967.00 161.55 6 (18) n.d n.d n.d

North Carolina (USA), (Arsand et al., 2018; Singer et al., 2016) in pesti-
cides and pharmaceuticals in effluents of WWTPs. These authors devel-
oped strategies that priorize the features of the suspect screening that
confirmed the potential candidates and achieved a semi-quantitative
analysis of EPs. These steps confirm the presence of some EPs in the ma-
trix selected. The monitoring of these compounds is a fundamental step
to stablish its spatial distribution, contamination profile and distribution
in water-sediment phase (Fairbairn et al., 2015; Zhou and Broodbank,
2014). The distribution of the compounds between water and sediment
mostly depend on their hydrophobic characters that determine low
water solubility. Commonly, the log P or log Kow is used as a measure
of their hydrophobicity, the higher the log P, the greater affinity for sed-
iments. Most of the studies established a limit value for log P of 3, that
means that compounds with log P values N3 tend to be mainly in
364 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368

Table 3
Overview of the current studies assessing EPs in environmental matrices based on target, suspected and non-target analysis.

Ref. Study scope Environmental matrix Approach Extraction procedure Analytical

instrumentation
Target Suspect Non-Target

(Park et al., 2018) PPCPs identification River water ✔ ✔ SPE-multi-layer LC-HRMS (Orbitrap)
cartridge
(Newton et al., 2018) Prioritize unknowns Drinking water ✔ ✔ Soxhlet LC-HRMS (qTOF)
(Arsand et al., 2018) Pharmaceuticals and pesticides Superficial water, effluent ✔ ✔ SPE LC-qTOF-MS/MS
WWTP
(Singer et al., 2016) Pharmaceutical Effluents WWTP ✔ SPE multi-layer LC-ESI-HRMS
cartridge
(Andrés-Costa et al., Emerging drugs of abuse Superficial water, influent ✔ ✔ SPE UHPLC–QqTOF–MS/MS
2016a) and effluent
WWTP
(Gago-Ferrero et al., Pharmaceuticals Influent WWTP ✔ ✔ SPE LC-qTOF-MS
2015)
(Schymanski et al., Organic contaminants and Effluents WWTP ✔ ✔ ✔ SPE-multi-layer HPLC-HR- MS/MS
2014b) transformation products cartridge
(Hug et al., 2014) Micropollutants Effluents WWTP ✔ ✔ SPE LC-HRMS
(Masiá et al., 2013) Pesticides and organic contaminants River water ✔ ✔ SPE LC–qTOF-MS and
(LC–QqQ-
MS/MS
(Müller et al., 2011) Organic trace substances Landfill leachates ✔ SPE HPLC-MS
(Sibiya et al., 2019) Organic pollutants and Leachate and sediment ✔ ✔ Solid Liquid LC-qTOF-MS
organophosphorus flame retardants Extraction (SLE)
(Ieda et al., 2019) PCBs, PBDEs, PCDD/Fs Sediment samples ✔ ✔ Soxhlet (GC × GC-HRToFMS
PAHs and organochlorine pesticides
(Hollender et al., Micropollutants Drinking water, ✔ ✔ SPE LC-HRMS/MS
2018) groundwater
(Storck et al., 2016) Pesticides and transformation Sediment ✔ Solvent extraction UHPLC-qTOF-MS
products (Acetone)
(Chiaia-Hernandez Organic contaminant Lake sediments ✔ ✔ Pressurized Liquid LC-HRMS (Orbitrap)
et al., 2014) Extraction

sediments but log P values b3 tend to be balanced between both phases
(Table S-5). Additionally, the sediment (particulate) water partition co-
efficient, Kp, is used as the ratio of the concentration of pollutant in the
sediment (or particulate) phase (Cp) to the concentration of pollutant
in the water (Cw) (Table S-6). These two parameters could provide a
clue of EPs behavior in both phases, since previous studies of
sediment-water interactions are typically based on laboratory experi-
ments. Such experiments provide important insight on the process
characteristics, and their controlling parameters. However, due to in-
herent limitations such as controlled conditions, and unrealistic reflec-
tion of field environment, there is a need to conduct field-based
partition experiments.
Imazalil, tebuconazole, nytenpiram, metalxyl, thiabendazole and
oxytetracycline were found in water and sediment samples using
suspected as well as target screening (last step). The log P values of
these compounds (between 2 and 3) are consistent with the existence
of a balance between water and sediments and the presence of these
compounds in both compartments with the exception of oxytetracy-
cline (Table S-5). The log P of this compound amphoteric molecule
with several ionic/polar groups is −0.99 then, do not explain its pres-
ence in sediments. However, the presence of tetracyclines in sediments
has been widely reported and their possible mechanisms of adsorption
include the ion exchange, electrostatic forces, and surface complexation
between its hydroxyl groups and sediment metal oxides (González-
Gaya et al., 2018; Li and Zhang, 2016; Siedlewicz et al., 2016).
Two of them −3,4,5-Trimethacarb and japothrin, both non-
authorized insecticides in the EU- were widely found in sediments.
Japothrin (log P = 5.66) is non-water soluble then, it rapidly disappears
from water and is adsorb to suspended solids on wastewater and accu-
mulate in the sediments of the rivers. In the case of, 3,4,5-trimethacarb
(log P = 2.47), some percentage could also adsorb to suspended solids
and sediment (this will explain its presence in this matrix) whereas it
will disappear from water by hydrolysis (Honing et al., 1996). The
other compounds are characterized by short half-lives (in the range of
hr), then their concentration should decline rapidly once are release to
surface waters.
The Kp values were calculated for those EPs detected simultaneously
in water and sediment samples (Table S-6). The results were highly var-
iable, for example, this varied from 26 to 289 mL g−1 for carbamezapine
and from 36 to 289 mL g−1 for 3-hydroxycarbofuran. If released into
water, carbamazepine is expected to adsorb to suspended solids and
sediment based upon the estimated Koc. (Wijekoon et al., 2013) found
that carbamazepine could significantly accumulate in sludges since
this compound is recalcitrant and has moderate hydrophobicity due to
the presence of an amide functional group. 3-hydroxycarbofuran, a
carbofuran-derivative with a half-life time (30–120 d) depending on
environmental conditions (Morales et al., 2012), was detected in 98%
in sediment samples. This insecticide is moderately persistent in sedi-
ments. (Bermúdez-Couso et al., 2011) carried out a study about
adsorption-desorption kinetics of carbofuran in acid soils. This study
found carbofuran adsorption capacity of the soils to be low and strongly
dependent on their clay and organic carbon contents. Such difference in
Kp values found in this study are likely due to several factors, including
the difference in sediment composition (e.g. organic carbon content).
Eprosartan is an angiotensin II receptor antagonist, which is used in
the treatment of hypertension and been detected recently in wastewa-
ter at moderately to high concentrations (Gurke et al., 2015; Letzel et al.,
2015; Stankiewicz et al., 2015). (Shah et al., 2011) has reported
Eprosartan through TOF-MS/MS spectrum with sixteen fragments, one
of them was m/z 207. Based on this study, the possibilities of proton-
ation in the drug structure showed four assigned routes for its fragmen-
tation. The first two were inter-related, and proceeded with subsequent
losses of H2O or CO to generate fragments of m/z 407, m/z 389, m/z 379
and m/z 361. The third and major route was initiated by the loss of thio-
phene, yielding a fragment of m/z 341 that was further reduced to ions
of m/z 297 and m/z 207 through neutral loss of CO2 and C8H6O2, respec-
tively. Both these ions further produced a fragment of m/z 163 by re-
spective neutral losses of C8H6O2 and CO2. There are other studies
 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368 365

Fig. 5. Spatial distribution as mean concentration of pesticides and pharmaceuticals in water (A) and sediment (B) samples.
366 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368
based on target screening that have also reported the same product ions
identified with this approach (207, 245. 407 and 379) (Grabic et al.,
2012; Gurke et al., 2015).
These results can be explained by the higher environmental stability
of pesticides (designed to be sprayed in the environment reaching toxic
concentrations). Instead, pharmaceuticals are less stable than
pesticides.
These results are coherent considering the hydrophobic interactions
(Log P values) since tebuconazole has log P N 3 favoring their accumula-
tion in sediments. Imazalil is moderately polar (log P = 2.56). Then,
their high presence in water together with a log P near to 3 justify
their presence in sediments. Regarding pharmaceutical compounds,
oxytetracycline was present in 13% of the sediment samples and had
one of the highest concentrations (2030 ng g−1).
Regarding the spatial distribution of EPs, target screening strat-
egy confirmed the pattern of pollution of previous monitoring in
the Turia River (Fig. 5) (Andrés-Costa et al., 2016a; Aznar et al.,
2016a; Aznar et al., 2016b; Carmona et al., 2014; Ccanccapa et al.,
2016a; Lorenzo et al., 2015; Masiá et al., 2014b). The mouth and spe-
cific points located in the head of the basin are the most polluted
sites. WWTP-I and WWTP-II effluents demonstrated to be a source
of EPs. From TUR12 to TUR15, pesticides and pharmaceutical were
detected at higher frequency. In TUR15, 20 pesticides were detected
in water samples and 14 in sediment samples (Fig. S-5). Another area
impacted by EPs was the head of the river, specifically the Alfambra
tributary, where the stretch from ALF1 to ALF6 showed high fre-
quency of pollutants in water and sediment samples, pharmaceuti-
cals as oxytetracycline, doxycycline, diazepam and carbamazepine
showed high concentrations (N500 ng L−1 and ng g−1). (Carmona
et al., 2014; Ccanccapa et al., 2016a) have previously reported the
presence of pharmaceutical and pesticides pollutants in water and
sediment samples from the Turia River Basin. The concentration
level and spatial distribution in the present study were consistent
with those findings in the main in the head of the basin (Alfambra
tributary) and effluents from WWTP. (Carmona et al., 2014) ana-
lyzed influent and effluent samples from WWTP and detected up to
20 PPCPs indicating that conventional treatment processes do not
completely remove PPCPs. Among them, triclocarban, gemfibrozil
and diclofenac were predominant in effluents. Spatial and temporal
patterns of pesticide residues were studied by (Ccanccapa et al.
2016a). This report showed highest concentration of pesticides in
the mouth of the river (comprise the regions of L'horta and Valencia
city that make up the coastal area), which could be related to the ag-
ricultural practices and rainfall. Oppositely, the headwater had a spe-
cific point of contamination (Alfambra tributary) and the sum of the
compounds detected surpassed the tolerance threshold of 0.5 μg L−1
for the sum of pesticide residues established by the European Union
Commission in order to guarantee water quality.
5. Conclusions
HRMS showed versatility and potential to analysis EPs through three
approaches (suspected, non-target and target screening) and could be
used as a routine tool to perform a comprehensive assessment of the
pollution in the River basins.
A comprehensive screening combining suspect, non-target and
semi-quantitative target screening was performed in a typical Mediter-
ranean river basin as result a total of 51 pesticides, 15 pharmaceuticals
and 2 mycotoxins were identified with the library. Of these 68 com-
pounds, 6 were confirmed and quantified with the real standard in
water and sediment samples (imazalil, tebuconazole, nytenpiram,
metalaxyl, thiabendazole and oxytetracycline).
Declaration of Competing Interest
The authors declare no competing financial interest.
Acknowledgments
This work has been supported by the Spanish Ministry of Science, In-
novation and Universities and the ERDF (European Regional Develop-
ment Funds) through the project GCL2015-64454-C2-1-R (ECO2risk-
dds) and the project RTI2018-097158-B-C31(Wetanpack) and through
the Generalitat Valenciana (ANTROPOCEN@, PROMETEO/2018/155). A.
Ccanccapa gratefully acknowledges the Conselleria D'Educació, Cultura
i Sport de la Generalitat Valenciana for the financial support through
“Santiago Grisolía” Scholarship Program. We gratefully acknowledge
the contributions of Irene Gimeno, Claudia Keller and Giuly Lazzaretti.
Appendix A. Supplementary data
Additional information as noted in the text (chemicals; sample prep-
aration; sampling points; compounds identified by library; frequency of
compounds identified by library; non-target results, interface of UNIFI
and more results). Supplementary data to this article can be found on-
line at https://doi.org/10.1016/j.scitotenv.2019.06.057.
References
Andrés-Costa, M.J., Rubio-López, N., Morales Suárez-Varela, M., Pico, Y., 2014. Occurrence
and removal of drugs of abuse in wastewater treatment plants of Valencia (Spain).
Environ. Pollut. 194, 152–162.
Andrés-Costa, M.J., Andreu, V., Picó, Y., 2016a. Analysis of psychoactive substances in
water by information dependent acquisition on a hybrid quadrupole time-of-flight
mass spectrometer. J. Chromatogr. A 1461, 98–106.
Andrés-Costa, M.J., Carmona, E., Picó, Y., 2016b. Universal method to determine acidic licit
and illicit drugs and personal care products in water by liquid chromatography quad-
rupole time-of-flight. MethodsX 3, 307–314.
Arsand, J.B., Hoff, R.B., Jank, L., Dallegrave, A., Galeazzi, C., Barreto, F., et al., 2018. Wide-
scope determination of pharmaceuticals and pesticides in water samples: qualitative
and confirmatory screening method using LC-qTOF-MS. Water Air and Soil Pollution
229.
Aznar, R., Moreno-Ramón, H., Albero, B., Sánchez-Brunete, C., Tadeo, J.L., 2016a. Spatio-
temporal distribution of pyrethroids in soil in Mediterranean paddy fields. J. Soils
Sediments 1–11.
Aznar, R., Sánchez-Brunete, C., Albero, B., Moreno-Ramón, H., Tadeo, J.L., 2016b. Pyre-
throids levels in paddy field water under Mediterranean conditions: measurements
and distribution modelling. Paddy Water Environ. 1–10.
Baz-Lomba, J.A., Reid, M.J., Thomas, K.V., 2016. Target and suspect screening of psychoac-
tive substances in sewage-based samples by UHPLC-QTOF. Anal. Chim. Acta 914,
81–90.
Bermúdez-Couso, A., Fernández-Calviño, D., Pateiro-Moure, M., Nóvoa-Muñoz, J.C., Simal-
Gándara, J., Arias-Estévez, M., 2011. Adsorption and desorption kinetics of carbofuran
in acid soils. J. Hazard. Mater. 190, 159–167.
Silva, BFd, Jelic, A., López-Serna, R., Mozeto, A.A., Petrovic, M., Barceló, D., 2011. Occur-
rence and distribution of pharmaceuticals in surface water, suspended solids and sed-
iments of the Ebro river basin, Spain. Chemosphere 85, 1331–1339.
Bletsou, A.A., Jeon, J., Hollender, J., Archontaki, E., Thomaidis, N.S., 2015. Targeted and non-
targeted liquid chromatography-mass spectrometric workflows for identification of
transformation products of emerging pollutants in the aquatic environment. TrAC-
Trends in Analytical Chemistry 66, 32–44.
Bolong, N., Ismail, A.F., Salim, M.R., Matsuura, T., 2009. A review of the effects of emerging
contaminants in wastewater and options for their removal. Desalination 239,
229–246.
Campo, J., Masiá, A., Picó, Y., Farré, M., Barceló, D., 2014. Distribution and fate of
perfluoroalkyl substances in Mediterranean Spanish sewage treatment plants. Sci.
Total Environ. 472, 912–922.
Carmona, E., Andreu, V., Picó, Y., 2014. Occurrence of acidic pharmaceuticals and personal
care products in Turia River Basin: from waste to drinking water. Sci. Total Environ.
484, 53–63.
Ccanccapa, A., Masiá, A., Andreu, V., Picó, Y., 2016a. Spatio-temporal patterns of pesticide
residues in the Turia and Júcar Rivers (Spain). Sci. Total Environ. 540, 200–210.
Ccanccapa, A., Masiá, A., Navarro-Ortega, A., Picó, Y., Barceló, D., 2016b. Pesticides in the
Ebro River basin: occurrence and risk assessment. Environ. Pollut. 211, 414–424.
Chiaia-Hernandez, A.C., Schymanski, E.L., Kumar, P., Singer, H.P., Hollender, J., 2014. Sus-
pect and nontarget screening approaches to identify organic contaminant records
in lake sediments. Anal. Bioanal. Chem. 406, 7323–7335.
Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal care products in the en-
vironment: agents of subtle change? Environ. Health Perspect. 107, 907–938.
Fairbairn, D.J., Karpuzcu, M.E., Arnold, W.A., Barber, B.L., Kaufenberg, E.F., Koskinen, W.C.,
et al., 2015. Sediment–water distribution of contaminants of emerging concern in a
mixed use watershed. Sci. Total Environ. 505, 896–904.
Farré, Ml, Pérez, S., Kantiani, L., Barceló, D., 2008. Fate and toxicity of emerging pollutants,
their metabolites and transformation products in the aquatic environment. TrAC
Trends Anal. Chem. 27, 991–1007.
Gago-Ferrero, P., Schymanski, E.L., Bletsou, A.A., Aalizadeh, R., Hollender, J., Thomaidis,
N.S., 2015. Extended suspect and non-target strategies to characterize emerging
 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368
polar organic contaminants in raw wastewater with LC-HRMS/MS. Environmental
Science & Technology 49, 12333–12341.
Gago-Ferrero, P., Schymanski, E.L., Hollender, J., Thomaidis, N.S., 2016. Nontarget analysis
of environmental samples based on liquid chromatography coupled to high resolu-
tion mass spectrometry (LC-HRMS). In: Sandra Pérez, P.E., Damià, B. (Eds.), Compre-
hensive Analytical Chemistry. Vol. 71. Elsevier, pp. 381–403 (Chapter 13).
Glauner, T., Wüst, B., Faye, T., 2016. A comprehensive workflow for target, suspect, and
non-target screening by LC/MS demonstrated for the identification of CECs in efflu-
ents from waste water treatment plants. Assessing Transformation Products of
Chemicals by Non-Target and Suspect Screening − Strategies and Workflows. Vol.
2.1242. American Chemical Society, pp. 113–130.
González-Gaya, B., Cherta, L., Nozal, L., Rico, A., 2018. An optimized sample treatment
method for the determination of antibiotics in seawater, marine sediments and bio-
logical samples using LC-TOF/MS. Sci. Total Environ. 643, 994–1004.
Gonzalez-Rey, M., Tapie, N., Le Menach, K., Dévier, M.-H., Budzinski, H., Bebianno, M.J.,
2015. Occurrence of pharmaceutical compounds and pesticides in aquatic systems.
Mar. Pollut. Bull. 96, 384–400.
Grabic, R., Fick, J., Lindberg, R.H., Fedorova, G., Tysklind, M., 2012. Multi-residue method
for trace level determination of pharmaceuticals in environmental samples using liq-
uid chromatography coupled to triple quadrupole mass spectrometry. Talanta 100,
183–195.
Gurke, R., Rößler, M., Marx, C., Diamond, S., Schubert, S., Oertel, R., et al., 2015. Occur-
rence and removal of frequently prescribed pharmaceuticals and corresponding
metabolites in wastewater of a sewage treatment plant. Sci. Total Environ. 532,
762–770.
Hollender, J., Schymanski, E.L., Singer, H.P., Ferguson, P.L., 2017. Nontarget screening with
high resolution mass spectrometry in the environment: ready to go? Environmental
Science & Technology 51, 11505–11512.
Hollender, J., Rothardt, J., Radny, D., Loos, M., Epting, J., Huggenberger, P., et al., 2018. Com-
prehensive micropollutant screening using LC-HRMS/MS at three riverbank filtration
sites to assess natural attenuation and potential implications for human health.
Water Research X 1, 100007.
Honing, M., Riu, J., Barceló, D., van Baar, B.L.M., Brinkman, U.A.T., 1996. Determination of
ten carbamate pesticides in aquatic and sediment samples by liquid
chromatography-ionspray and thermospray mass spectrometry. J. Chromatogr. A
733, 283–294.
Hug, C., Ulrich, N., Schulze, T., Brack, W., Krauss, M., 2014. Identification of novel
micropollutants in wastewater by a combination of suspect and nontarget screening.
Environ. Pollut. 184, 25–32.
Ibáñez, M., Sancho, J.V., Pozo, Ó.J., Niessen, W., Hernández, F., 2005. Use of quadrupole
time-of-flight mass spectrometry in the elucidation of unknown compounds present
in environmental water. Rapid Commun. Mass Spectrom. 19, 169–178.
Ibáñez, M., Borova, V., Boix, C., Aalizadeh, R., Bade, R., Thomaidis, N.S., et al., 2017. UHPLC-
QTOF MS screening of pharmaceuticals and their metabolites in treated wastewater
samples from Athens. J. Hazard. Mater. 323, 26–35.
Ieda, T., Hashimoto, S., Isobe, T., Kunisue, T., Tanabe, S., 2019. Evaluation of a data-
processing method for target and non-target screening using comprehensive two-
dimensional gas chromatography coupled with high-resolution time-of-flight mass
spectrometry for environmental samples. Talanta 194, 461–468.
Kaufmann, A., Butcher, P., Maden, K., Walker, S., Widmer, M., 2017. Using in silico frag-
mentation to improve routine residue screening in complex matrices. J. Am. Soc.
Mass Spectrom. 28, 2705–2715.
Kind, T., Fiehn, O., 2007. Seven Golden Rules for heuristic filtering of molecular formulas
obtained by accurate mass spectrometry. BMC Bioinformatics 8, 105.
Krauss, M., Singer, H., Hollender, J., 2010. LC–high resolution MS in environmental analy-
sis: from target screening to the identification of unknowns. Anal. Bioanal. Chem. 397,
943–951.
Letzel, T., Bayer, A., Schulz, W., Heermann, A., Lucke, T., Greco, G., et al., 2015. LC-MS
screening techniques for wastewater analysis and analytical data handling strategies:
Sartans and their transformation products as an example. Chemosphere 137,
198–206.
Li, J., Zhang, H., 2016. Adsorption-desorption of oxytetracycline on marine sediments: ki-
netics and influencing factors. Chemosphere 164, 156–163.
Lorenzo, M., Campo, J., Pico, Y., 2015. Optimization and comparison of several extraction
methods for determining perfluoroalkyl substances in abiotic environmental solid
matrices using liquid chromatography-mass spectrometry. Anal. Bioanal. Chem.
407, 5767–5781.
Mailler, R., Gasperi, J., Coquet, Y., Buleté, A., Vulliet, E., Deshayes, S., et al., 2016. Removal of
a wide range of emerging pollutants from wastewater treatment plant discharges by
micro-grain activated carbon in fluidized bed as tertiary treatment at large pilot scale.
Sci. Total Environ. 542 (Part A), 983–996.
Martínez-Domínguez, G., Romero-González, R., Garrido Frenich, A., 2016. Multi-class
methodology to determine pesticides and mycotoxins in green tea and royal jelly
supplements by liquid chromatography coupled to Orbitrap high resolution mass
spectrometry. Food Chem. 197 (Part A), 907–915.
Masiá, A., Ibáñez, M., Blasco, C., Sancho, J.V., Picó, Y., Hernández, F., 2013. Combined
use of liquid chromatography triple quadrupole mass spectrometry and liquid
chromatography quadrupole time-of-flight mass spectrometry in systematic
screening of pesticides and other contaminants in water samples. Anal. Chim.
Acta 761, 117–127.
Masiá, A., Blasco, C., Picó, Y., 2014a. Last trends in pesticide residue determination by liq-
uid chromatography–mass spectrometry. Trends in Environmental Analytical Chem-
istry 2, 11–24.
Masiá, A., Campo, J., Blasco, C., Picó, Y., 2014b. Ultra-high performance liquid
chromatography–quadrupole time-of-flight mass spectrometry to identify contami-
nants in water: an insight on environmental forensics. J. Chromatogr. A 1345, 86–97.
367
Masiá, A., Suarez-Varela, M.M., Llopis-Gonzalez, A., Picó, Y., 2016. Determination of pesti-
cides and veterinary drug residues in food by liquid chromatography-mass spectrom-
etry: a review. Anal. Chim. Acta 936, 40–61.
Morales, J., Manso, J.A., Cid, A., Mejuto, J.C., 2012. Degradation of carbofuran and
carbofuran-derivatives in presence of humic substances under basic conditions.
Chemosphere 89, 1267–1271.
Moran, D., Cross, T., Brown, L.M., Colligan, R.M., Dunbar, D., 2014. Data-independent ac-
quisition (MSE) with ion mobility provides a systematic method for analysis of a bac-
teriophage structural proteome. J. Virol. Methods 195, 9–17.
Müller, A., Schulz, W., Ruck, W.K.L., Weber, W.H., 2011. A new approach to data evalua-
tion in the non-target screening of organic trace substances in water analysis.
Chemosphere 85, 1211–1219.
Navalon, S., Alvaro, M., Alcaina, I., Garcia, H., 2010. Multi-method characterization of DOM
from the Turia river (Spain). Appl. Geochem. 25, 1632–1643.
Newton, S.R., McMahen, R.L., Sobus, J.R., Mansouri, K., Williams, A.J., McEachran, A.D., et
al., 2018. Suspect screening and non-targeted analysis of drinking water using
point-of-use filters. Environ. Pollut. 234, 297–306.
Paíga, P., Santos, L.H.M.L.M., Ramos, S., Jorge, S., Silva, J.G., Delerue-Matos, C., 2016. Pres-
ence of pharmaceuticals in the Lis river (Portugal): sources, fate and seasonal varia-
tion. Sci. Total Environ. 573, 164–177.
Park, N., Choi, Y., Kim, D., Kim, K., Jeon, J., 2018. Prioritization of highly exposable pharma-
ceuticals via a suspect/non-target screening approach: a case study for Yeongsan
River, Korea. Sci. Total Environ. 639, 570–579.
Pedersen, A.J., Dalsgaard, P.W., Rode, A.J., Rasmussen, B.S., Muller, I.B., Johansen, S.S., et al.,
2013. Screening for illicit and medicinal drugs in whole blood using fully automated
SPE and ultra-high-performance liquid chromatography with TOF-MS with data-
independent acquisition. J. Sep. Sci. 36, 2081–2089.
Pedrouzo, M., Borrull, F., Marcé, R.M., Pocurull, E., 2009. Ultra-high-performance liquid
chromatography–tandem mass spectrometry for determining the presence of eleven
personal care products in surface and wastewaters. J. Chromatogr. A 1216,
6994–7000.
Petrović, M., Gonzalez, S., Barceló, D., 2003. Analysis and removal of emerging contami-
nants in wastewater and drinking water. TrAC-Trends in Analytical Chemistry 22,
685–696.
Picó, Y., Barceló, D., 2015. Transformation products of emerging contaminants in the en-
vironment and high-resolution mass spectrometry: a new horizon. Anal. Bioanal.
Chem. 407, 6257–6273.
Rosnack, K.J., Reid, M.J., Ladak, A., Cleland, G., 2016. Screening solution using the software
platform UNIFI: an integrated workflow by waters. ACS Symp. Ser. 1242, 155–172.
Ruff, M., Mueller, M.S., Loos, M., Singer, H.P., 2015. Quantitative target and systematic
non-target analysis of polar organic micro-pollutants along the river Rhine using
high-resolution mass-spectrometry–identification of unknown sources and com-
pounds. Water Res. 87, 145–154.
Sanchez-Prado, L., Garcia-Jares, C., Llompart, M., 2010. Microwave-assisted extraction: ap-
plication to the determination of emerging pollutants in solid samples. J. Chromatogr.
A 1217, 2390–2414.
Schymanski, E.L., Jeon, J., Gulde, R., Fenner, K., Ruff, M., Singer, H.P., et al., 2014a. Identify-
ing small molecules via high resolution mass spectrometry: communicating confi-
dence. Environmental Science & Technology 48, 2097–2098.
Schymanski, E.L., Singer, H.P., Longrée, P., Loos, M., Ruff, M., Stravs, M.A., et al., 2014b.
Strategies to characterize polar organic contamination in wastewater: exploring the
capability of high resolution mass spectrometry. Environmental Science & Technol-
ogy 48, 1811–1818.
Schymanski, E.L., Singer, H.P., Slobodnik, J., Ipolyi, I.M., Oswald, P., Krauss, M., et al., 2015.
Non-target screening with high-resolution mass spectrometry: critical review using a
collaborative trial on water analysis. Anal. Bioanal. Chem. 407, 6237–6255.
Shah, R.P., Sahu, A., Singh, S., 2011. Identification and characterization of geometrical iso-
meric photo degradation product of eprosartan using LC-MS and LC-NMR. Eur.
J. Chem. 2, 152–157.
Sibiya, I., Poma, G., Cuykx, M., Covaci, A., Daso Adegbenro, P., Okonkwo, J., 2019. Targeted
and non-target screening of persistent organic pollutants and organophosphorus
flame retardants in leachate and sediment from landfill sites in Gauteng Province,
South Africa. Sci. Total Environ. 653, 1231–1239.
Siedlewicz, G., Borecka, M., Białk-Bielińska, A., Sikora, K., Stepnowski, P., Pazdro, K., 2016.
Determination of antibiotic residues in southern Baltic Sea sediments using tandem
solid-phase extraction and liquid chromatography coupled with tandem mass spec-
trometry. Oceanologia 58, 221–234.
Singer, H.P., Wössner, A.E., McArdell, C.S., Fenner, K., 2016. Rapid screening for exposure
to “non-target” pharmaceuticals from wastewater effluents by combining HRMS-
based suspect screening and exposure modeling. Environmental Science & Technol-
ogy 50, 6698–6707.
Stankiewicz, A., Giebułtowicz, J., Stankiewicz, U., Wroczyński, P., Nałecz-Jawecki, G., 2015.
Determination of selected cardiovascular active compounds in environmental aquatic
samples - methods and results, a review of global publications from the last 10 years.
Chemosphere 138, 642–656.
Storck, V., Lucini, L., Mamy, L., Ferrari, F., Papadopoulou, E.S., Nikolaki, S., et al., 2016. Iden-
tification and characterization of tebuconazole transformation products in soil by
combining suspect screening and molecular typology. Environ. Pollut. 208, 537–545.
Wijekoon, K.C., Hai, F.I., Kang, J., Price, W.E., Guo, W., Ngo, H.H., et al., 2013. The fate of
pharmaceuticals, steroid hormones, phytoestrogens, UV-filters and pesticides during
MBR treatment. Bioresour. Technol. 144, 247–254.
Yao, L., Zhao, J.-L., Liu, Y.-S., Yang, Y.-Y., Liu, W.-R., Ying, G.-G., 2016. Simultaneous deter-
mination of 24 personal care products in fish muscle and liver tissues using
QuEChERS extraction coupled with ultra pressure liquid chromatography-tandem
mass spectrometry and gas chromatography-mass spectrometer analyses. Anal.
Bioanal. Chem. 408, 8177–8193.
368 A. Ccanccapa-Cartagena et al. / Science of the Total Environment 687 (2019) 355–368

Zedda, M., Zwiener, C., 2012a. Is nontarget screening of emerging contaminants by LC-
HRMS successful? A plea for compound libraries and computer tools. Anal. Bioanal.
Chem. 403, 2493–2502.
Zedda, M., Zwiener, C., 2012b. Is nontarget screening of emerging contaminants by LC-
HRMS successful? A plea for compound libraries and computer tools. Anal. Bioanal.
Chem. 403, 2493–2502.
Zhou, J., Broodbank, N., 2014. Sediment-water interactions of pharmaceutical residues in
the river environment. Water Res. 48, 61–70.
Zhou, J., Li, Y., Chen, X., Zhong, L., Yin, Y., 2017. Development of data-independent acqui-
sition workflows for metabolomic analysis on a quadrupole-orbitrap platform.
Talanta 164, 128–136.