Analytica Chimica Acta 641 (2009) 117–123

Contents lists available at ScienceDirect

Analytica Chimica Acta

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

A combination of ultrasonic assisted extraction with LC–MS/MS for the

determination of organophosphorus pesticides in sludge
Ana Isabel García-Valcárcel, José Luis Tadeo ∗
Departamento de Medio Ambiente, INIA, Ctra. de La Coruña Km 7, 28040 Madrid, Spain

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

Article history: A method has been developed for the analysis of a wide polarity range of the currently used organophos-
Received 5 September 2008 phorus pesticides (OPs) and their metabolites in sewage sludge samples. Extraction was carried out using
Received in revised form 20 March 2009 ultrasonic assisted extraction (UAE) with different solvents. The levels of OPs in sludge were determined
Accepted 27 March 2009
by liquid chromatography-tandem mass spectrometry (LC–MS/MS) using electrospray ionization (ESI) in
Available online 5 April 2009
positive or negative ion mode. Extraction with acetonitrile containing 1% acetic acid gave the best results
with recoveries between 83.2% and 106.4% and RSD ≤ 8.7%. Evaluation of matrix effect showed high ion
Keywords:
suppression for OPs of intermediate polarity, which decreased to approximately 50% by matrix dilution
Sludge
Organophosphorus pesticides
to an equivalent of 0.5 g of sludge per millilitre; however, for polar OPs, the matrix effect was negligible
Ultrasonic assisted extraction at the same concentration. Therefore, matrix-matched standards were used for OPs quantification. The
Liquid chromatography–tandem mass limits of quantification were in the range of 1–14 ng g−1 . This method was successfully applied to real
spectrometry sludge samples collected from Madrid Province and chlorpyrifos along with its metabolite TCPY were the
OPs found in sludge at the highest levels, 113–344 ng g−1 and 33–307 ng g−1 , respectively.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction derivatization and positive analyte identification and quantification

Owing to the processes used in sewage treatment plants can-
not eliminate all contaminants, a wide range of toxic compounds
can find their ways to reach sludge. Therefore, sludge may con-
tain compounds including pesticides and other toxic chemicals that
might be transferred into soil when used as fertilizer. Among the
organic contaminants, organophosphorus pesticides (OPs) (with
various chemical structures and polarities) constitute a class of
widely used insecticides, that can be adsorbed to solid particles of
sewage sludge and cause a negative impact on soil organisms [1].
These pesticides are toxic for mammals due to the inhibition of the
acetylcholinesterase, an enzyme necessary for the normal function
of the nervous system [2]. OPs can enter the trophic chain through
the fertilization of soil with sludge and hence sludge can be consid-
ered as a source of contamination. Nevertheless, scarce information
exists on the levels of these organic contaminants in sludge.
OPs have been generally analysed in different matrices by gas
chromatography with mass spectrometry detection [3–8], however
some of highly polar and/or thermolabile compounds seem to be
difficult for analysis using GC, due to their low volatility and poor
repeatability. A liquid chromatography coupled with mass spec-
trometry (LC–MS) is a suitable technique since it does not require
∗ Corresponding author. Tel.: +34 91 3476821; fax: +34 91 3572293.
E-mail address: tadeo@inia.es (J.L. Tadeo).
can be performed in one step. In order to achieve good sensitiv-
ity and selectivity, LC coupled to tandem mass spectrometry via
an electrospray ionization interface (ESI) is becoming a frequent
technique for the determination of OPs in different matrices [9–12].
In general, traditional methods for the extraction of pesticides
from solid matrices, such as Soxhlet or mechanical shaking, alter-
natively have been replaced with modern extraction methods
including ultrasonic assisted extraction (UAE), supercritical fluid
extraction (SFE) and pressurized liquid extraction (PLE) [13–15].
Though, some of these techniques require expensive equipments;
they are not laborious and require less amount of solvent. The
QuEChERS method [16] (quick, easy, cheap, effective, rugged and
safe) is a procedure based on acetonitrile extraction that has been
used lately for the determination of OPs in fruits and vegetables.
Matrix solid phase dispersion (MSPD) is another modern extrac-
tion technique applied in our laboratory for the analysis of several
contaminants in sewage sludge for the first time [17,18]. However,
scarce methods for the determination of a few OPs in sewage sludge
have been reported in the literature [7,19]. The wide range of polar-
ities of OPs may difficult their adequate extraction and subsequent
determination in a complex matrix such as sewage sludge.
The aim of this work was to develop a simple and rapid method
for the determination of seven organophosphorus pesticides and
three of their metabolites in sewage sludge, based on ultrasonic
assisted extraction, followed by dispersive solid phase extraction
(DSPE) as a purification technique and LC–ESI-MS/MS. This method

0003-2670/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.aca.2009.03.046
118 A.I. García-Valcárcel, J.L. Tadeo / Analytica Chimica Acta 641 (2009) 117–123
was applied to the analysis of sludge samples collected from differ-
ent wastewater treatment plants from Madrid (Spain). To the best of
our knowledge, this is the first work reporting an analytical method
for the determination of a wide range of OPs and their metabolites
in sewage sludge. The tested OPs represent some of most commonly
used pesticides at present. Additionally, this work provides infor-
mation on the levels of OPs existed in sewage sludge, which is also
very scarce in the literature.
2. Experimental
2.1. Chemicals, reagents and materials
Acetonitrile, ethyl acetate and methanol were supplied by Pan-
reac (Barcelona, Spain) and ammonium formate and acetic acid
were obtained from Sigma (Steinheim, Germany). Silica Bondesil-
C18 , 40 ␮m particle diameter and Bondesil-PSA (primary secondary
amine), 40 ␮m particle diameter, were purchased from Varian (Palo
Alto, USA). Unless otherwise stated, all solvents used were of ana-
lytical grade.
Pesticide standards, chlorpyrifos, chlorpyrifos methyl and their
metabolite TCPY (3,5,6-trichloro-2-pyridinol), diazinon and its
metabolite IMPY (2-isopropyl-6-methyl pirimidin-4-ol), pirim-
iphos methyl and its metabolite DEAMPY (2-diethylamino-6methyl
pirimidin-4-ol), methamidophos, dichlorvos and azamethiphos,
were obtained from Sigma (Steinheim, Germany).
Sludge samples were collected during 2007 from wastewa-
ter treatment plants (WWTPs) located in the province of Madrid
(Spain). Samples were transported to our laboratory under cool con-
ditions (4–7 ◦ C), frozen at −80 ◦ C for 48 h, lyophilized, homogenized
with a mill, sieved to 0.5 mm and stored at −20 ◦ C until they were
analysed in the following weeks.
2.2. Preparation of calibration standards
Individual stock solutions, containing 1 mg mL−1 of OPs, were
prepared in analytical grade ethyl acetate and a mixture of each
organophosphorus pesticide at 1 ␮g mL−1 was prepared by dilut-
ing the stock solutions in acetonitrile, methanol/water (90/10, v/v)
and methanol/water–acetic acid 1% (95/5, v/v). These intermedi-
ate standards solutions were diluted to give calibration solutions of
5 ng mL−1 , 10 ng mL−1 , 25 ng mL−1 , 50 ng mL−1 and 100 ng mL−1 .
To determine the matrix effect, matrix-matched multi-level cal-
ibration standards (5 ng mL−1 , 10 ng mL−1 , 25 ng mL−1 , 50 ng mL−1
and 100 ng mL−1 ) were prepared with extracts of sewage sludge
samples and compared with standards in the best solvent. The
matrix concentration in the extract (1 g mL−1 and 0.5 g mL−1 ) was
also evaluated. The matrix effect was expressed as the pesticide
response in matrix-matched standards compared to the response in
pure solvent and was evaluated with sludge from different WWTP.
As matrix effect depends on the amount of matrix per millilitre
of extract and the sample preparation method used, these effects
were examined for each compound. Taken into account the possible
suppression or enhancement of ionization by the co-eluting com-
pounds, matrix-matched standards were used instead of standards
in pure solvent.
2.3. Sample preparation
2.3.1. Matrix solid phase dispersion (MSPD) assisted by sonication
Approximately 2 g finely ground lyophilized sludge and silica C18
(2 g) were placed in a glass mortar and the mixture was blended
with a glass pestle with circular motion to yield a homogeneous
material. Afterwards, it was transferred to a 25 mL glass column
(10 cm × 20 mm i.d., Rothe, Portugal) with a polyethylene frit (2 cm
diameter and 20 ␮m pore, Supelco, Spain) at the end. 2 mL of ace-
tonitrile, or acetonitrile containing 1% acetic acid, were used to rinse
the mortar and pestle, and the mixtures were transferred to the
column along with another 3 mL of the same solvent. The glass
column was placed in an ultrasonic bath (Raypa, Spain) at 40 ◦ C
for 15 min. After extraction, the column was placed on a multiport
vacuum manifold where the solvent was eluted from the column,
filtered through the column frit and collected in a graduated tube.
This extraction procedure was repeated again with 5 mL of solvent
and the yielded extracts were combined and concentrated under a
gentle stream of air at 30 ◦ C in a fume hood to 2 mL. To the concen-
trated extract, 0.1 g of PSA (primary–secondary amine) were added,
shaken for 1 min and phases were allowed to separate for 10 min.
Then the liquid phase was transferred to a graduated tube, con-
centrated near dryness and reconstituted in methanol/deionized
water–acetic acid 1% (95/5, v/v) to an adequate volume (2 mL).
2.3.2. Ultrasonic assisted extraction (UAE)
Approximately 2 g finely ground lyophilized sludge were placed
in a glass column with a frit at the end. Acetonitrile or a mix-
ture of acetonitrile–1% acetic acid (5 mL) was added and the sludge
sample was extracted in an ultrasonic bath as mentioned above.
The concentrated extracts (2 mL) were cleaned-up by DSPE with
PSA (0.1 g) following the above mentioned procedure. After con-
centration near dryness, the extract was reconstituted in 2 mL of
methanol/deionized water–acetic acid 1% (95/5, v/v).
2.4. LC–MS/MS
An Agilent 1200 (Waldbronn, Germany) liquid chromatograph
equipped with an autosampler, a quaternary pump and a ther-
mostated column compartment was used. Separations were carried
out using a Zorbax Eclipse XDB-C18 (150 mm × 4.6 mm i.d., 5 ␮m
particle size) analytical column with a C18 security guard car-
tridge (Agilent, Waldbronn, Germany). Eluent flow rate was set at
0.3 mL min−1 and the column was kept at 45 ◦ C. Two gradient elu-
tion programs were applied. For positive ionization (PI) analysis, an
organic solvent containing 5 mM ammonium formate in methanol
and an aqueous solvent containing 5 mM ammonium formate in
deionized water were used. The gradient program started in 75%
aqueous phase, decreased to 50% in 6 min, then to 5% in 11 min,
maintained at 5% until 32 min and returned to initial conditions the
following 1 min. A 5 min equilibration time was included between
runs resulting in a total analysis time of 38 min. For negative ion-
ization (NI) analysis, the organic phase was acetonitrile and the
aqueous phase 0.1% acetic acid in deionized water, starting in 75%
aqueous mobile phase, decreasing to 5% in 11 min, held at 5% for
7 min and returning to initial conditions in 2 min, with three addi-
tional min for equilibration before injection. A total analysis time
of 23 min was employed. Injection volumes of 5 ␮L and 10 ␮L were
used in PI and NI analyses, respectively.
An Agilent 6410 triple quadrupole tandem mass spectrometer
equipped with an electrospray ionization interface, operating in
both positive (PI) and negative (NI) ion modes according to the pref-
erential ionization of each analyte, was used. Drying and nebulising
gas for the LC–MS/MS was produced in situ by a nitrogen generator
fed by compressed air at 7 bar. The optimized ESI parameters were:
drying gas flow rate 12 L min−1 ; drying gas temperature 350 ◦ C;
nebulizer gas pressure 40 psi and capillary voltage 2000/−1000 V
(PI/NI).
Two different transitions (precursor/product ion pair) were
monitored per compound. The optimized settings for fragmentor
(cone voltage) and collision energy were tested for each com-
pound by flow injection analysis. Chromatographic data acquisition
in the positive ionization mode was divided into three different
MS segments to maximize sensitivity and a dwell time between
300 ms scan−1 and 100 ms scan−1 was chosen for all compounds.
 A.I. García-Valcárcel, J.L. Tadeo / Analytica Chimica Acta 641 (2009) 117–123 119

The time window for each segment was: segment I from 0 min to
14 min, segment II from 14.1 min to 19 min and segment III from
19.1 min to 23.5 min. OPs were confirmed by their retention times,
the identification of the most abundant transition ion used as quan-
tifier and other product ion used as qualifier.

3. Results and discussion

3.1. LC–ESI-MS/MS

Protonated or deprotonated molecular ions, in positive and

negative ionization modes were chosen for each compound as
precursors of product ions, which were optimized by varying the
collision energy. The most abundant product ion was selected as
quantifier and other ion as qualifier. Due to the stability of TCPY
[20], fragmentation of its [M − H]− precursor ion was not possible,
thus [M − H]− was used as both precursor and product ion and one
isotope of the ion cluster was chosen for confirmation. The opti-
mized parameters used to determine OPs by LC–ESI MS/MS are
summarized in Table 1.
In order to obtain the optimum response, different mobile
phase compositions were evaluated in the LC–MS/MS analysis.
Fig. 1. Effect of the injection solvent on the linear response of representative OP
The response of analytes, in positive mode, was good when compounds.
methanol/deionized water, instead of acetonitrile/deionized water,
was used as a mobile phase, whereas ionization efficiency in nega-
tive mode was better with acetonitrile, because it is a weaker proton Moreover, the injection solvent showed an important effect on
donor than methanol [21]. the response of the more polar OPs. Standard solutions in dif-
Ammonium formate, acetic acid and a mixture of formic acid ferent solvents including acetonitrile, methanol/deionized water
and ammonium formate were assayed as modifiers, and ammo- (90/10, v/v) and methanol/deionized water–acetic acid 1% (95/5,
nium formate was selected because it gave the overall best results v/v), were injected and the responses were compared. Fig. 1 shows
in respect to sensitivity, particularly for OPs of intermediate polari- the effect of the injection solvent on the response of methami-
ties with ionization in positive mode. Ionization of TCPY in negative dophos, polar OP, and pirimiphos methyl, OP of intermediate
mode gave better results when acetic acid was used as modifier polarity. The response of methamidophos was clearly improved
in the aqueous phase. Therefore, methanol/deionized water–5 mM when it was injected in MeOH/deionized water–1% acetic acid
ammonium formate, was chosen as mobile phase when ESI in pos- (95/5). A peak shape enhancement for this analyte was also
itive mode was employed and acetonitrile/deionized water–acetic observed when injected in this solvent. However, the effect of the
acid (0.1%) was used in the negative mode. injection solvent on the response and peak shape of a compound

Table 1
LC–ESI-MS/MS parameters.

Compound log Kow a Type Precursor (m/z) Cone voltage (V) Product ion (m/z) Collision energy (V) Rt (min)

Positive ionization
Methamidophos −0.8 Q 142 80 94 10 6.6
q 142 80 125 10

IMPY n.a. Q 153 130 84 20 13.1

q 153 130 70 20

DEAMPY n.a. Q 182 130 154 15 15.9

q 182 130 84 25

Azamethiphos 1.05 Q 325 60 183 10 17.2

q 325 60 139 25

Dichlorvos 1.42 Q 221 120 109 15 17.4

q 221 120 127 15

Diazinon 3.30 Q 305 80 169 25 20.7

q 305 80 153 20

Pirimiphos methyl 4.20 Q 306 105 164 20 21.2

q 306 105 108 55

Chlorpyrifos methyl 4.24 Q 322 110 125 20 21.5

q 322 110 290 10

Chlorpyrifos 4.70 Q 352 120 200 20 23.0

q 352 120 97 35

Negative ionization
TCPY n.a. Q 198 100 198 0 15.1
q 196 80 196 0 15.1

Q quantifier; q qualifier., n.a., not available.

a
From Tomlin [27].
120 A.I. García-Valcárcel, J.L. Tadeo / Analytica Chimica Acta 641 (2009) 117–123

Fig. 2. Selected reaction monitoring (SRM) chromatograms for chlorpyrifos methyl following purification using DSPE either with PSA or C18 .

of intermediate polarity such as pirimiphos methyl was negligi- Table 3

Matrix effect on the OP responsea .
ble.
In general, a better response was obtained for polar pesti- OP Relative response (%)
cides (methamidophos, TCPY, IMPY, DEAMPY) in MeOH/deionized Matrix concentration Type of sludge
water–1% acetic acid (95/5, v/v) without affecting the response of
1 g ml−1 0.5 g ml−1 3 4 5
intermediate polar pesticides. This is in agreement with a recently
published paper reporting the use of methanol–water to prevent Methamidophos 80 106 98 92 110
IMPY 93 103 116 103 103
band broadening of highly polar pesticides [22]. Therefore, recon-
DEAMPY 71 102 101 83 102
stitution of sewage sludge extracts in that solvent was carried out Azamethiphos 83 96 90 91 96
before sample injection in LC. Dichlorvos 72 97 87 88 97
Diazinon 35 57 60 43 57
Pirimiphos methyl 17 43 25 28 32
3.2. Sample preparation Chlorpyrifos methyl 32 56 41 38 56
Chlorpyrifos 35 52 52 51 52
The extraction efficiencies of both MSPD and UAE using differ- TCPY 73 100 114 110 83
ent solvents (as described above) were evaluated in sewage sludge a
Relative responses were obtained by comparing the slopes of the calibration
samples fortified with a mixture of OPs. To reduce the presence of curves of standards in neat solvent and in matrix extracts spiked with the OPs at
sludge co-extractives, the extracts were cleaned up by DSPE after 5 ng mL−1 , 10 ng mL−1 , 25 ng mL−1 , 50 ng mL−1 and 100 ng mL−1 .
both extraction procedures. In the DSPE step, the extraction solvent
served as sample disperser and PSA was chosen over silica C18 as
solid phase, because it yields somewhat clean extracts as shown in assisted extraction was used, acetonitrile–1% acetic acid gave good
Fig. 2 for chlorpyrifos methyl. recoveries, between 83.2% and 106.4% and precision (RSD ≤ 8.7%)
Table 2 shows the recoveries achieved with different extrac- for all analytes, whereas metabolites and the most polar OPs
tion procedures, using matrix-matched standards. When ultrasonic (methamidophos) showed lower recoveries when acetonitrile was

Table 2
Recoveries (%) and relative standard deviations (RSD) using different extraction methodsa .

Analyte UAE MSPD

Acn Acn–AcH Acn Acn–AcH

Mean RSD Mean RSD Mean RSD Mean RSD

Methamidophos 67.6 6.3 86.4 2.4 78.8 20.0 73.4 5.8

IMPY 70.6 10.0 86.3 1.8 67.3 32.8 80.6 11.4
DEAMPY 54.9 5.4 83.2 1.9 25.6 5.4 74.5 5.2
Azamethiphos 83.1 11.5 97.9 4.9 92.4 18.8 87.5 4.7
Dichlorvos 63.2 13.7 100.7 8.1 77.3 17.2 83.5 8.3
Diazinon 88.8 10.0 105.1 2.7 98.2 11.5 87.7 2.2
Pirimiphos methyl 88.2 6.9 103.8 4.0 100.0 10.5 85.6 2.7
Chlorpyrifos methyl 88.5 19.4 102.3 8.7 89.5 9.0 83.9 5.0
Chlorpyrifos 103.0 6.4 106.4 3.7 105.1 9.0 92.5 1.7
TCPY 85.8 2.4 99.5 2.9 93.6 11.7 85.6 17.5
a
Fortification level 20 ng/g (n = 3). UAE: ultrasonic assisted extraction; MSPD: matrix solid phase dispersion. Acn: acetonitrile, Acn–AcH: acetonitrile–1% acetic acid.
 A.I. García-Valcárcel, J.L. Tadeo / Analytica Chimica Acta 641 (2009) 117–123 121

Table 4
Calibration data, limits of detection and quantification (LOD, LOQ; ng g−1 ) and intra and inter-day repeatability, RSD (%).

OP Equation r2 LOD LOQ RSD (%)

Intra-day Inter-day

Methamidophos y = 384.97x + 0.0029 0.9966 0.3 1.0 2.3 12.5

IMPY y = 248.56x − 0.0033 0.9990 2.1 7.0 1.2 8.4
DEAMPY y = 433.40x − 0.0207 0.9990 1.5 5.0 1.7 5.5
Azamethiphos y = 341.04x + 0.003 0.9978 0.3 1.0 1.7 6.2
Dichlorvos y = 105.78x + 0.0005 0.9940 0.6 2.0 7.3 10.6
Diazinon y = 237.54 + 0.001 0.9966 0.4 1.2 2.5 12.1
Pirimiphos methyl y = 111.87x − 0.001 0.9986 1.2 4.0 7.2 12.5
Chlorpyrifos methyl y = 10.10x + 0.0004 0.9937 1.5 5.0 4.7 5.9
Chlorpyrifos y = 40.73x + 0.0481 0.9946 4.2 14.0 2.7 10.1
TCPY y = 19.15x − 0.044 0.9999 2.4 8.0 8.4 19.1

RSD of peak areas (n = 8).

Table 5 3.3. Matrix effect

Recovery (n = 4).

OP Fortification level The matrix effect was evaluated by comparing the slopes of
10 ng g−1 50 ng g−1 100 ng g−1 the calibration curves of standards in neat solvent and in matrix
extracts spiked with OPs at 5 ng mL−1 , 10 ng mL−1 , 25 ng mL−1 ,
Methamidophos 93.1 (3.5) 79.8 (1.8) 87.3 (8.9)
IMPY 91.0 (1.2) 81.6 (4.7) 84.0 (4.0)
50 ng mL−1 and 100 ng mL−1 . Unspiked matrix extracts were ana-
DEAMPY 82.7 (2.0) 83.6 (2.0) 89.4 (9.8) lyzed and OPs levels found were subtracted. Matrix extracts
Azamethiphos 103.9 (8.8) 93.5 (2.9) 101.1 (5.0) were obtained by ultrasonic assisted extraction with acetoni-
Dichlorvos 109.7 (14.0) 91.8 (1.8) 91.3 (2.7) trile containing 1% of acetic acid, followed by purification by
Diazinon 113.3 (6.3) 96.9 (3.0) 99.1 (2.2)
DSPE with PSA, evaporation near dryness and reconstitution in
Pirimiphos methyl 109.8 (9.6) 97.8 (3.8) 96.2 (2.2)
Chlorpyrifos methyl 106.2 (16.9) 98.4 (1.8) 97.3 (3.0) MeOH/deionized water–acetic acid 1% (95/5, v/v). The effect of
Chlorpyrifos >LOQ 106.4 (3.6) 97.3 (3.1) dilution of the final extract, containing 1 g or 0.5 g of matrix per
TCPY 105.6 (2.7) 97.2 (6.9) 87.9 (11.4) mL, using the same calibration curves was also tested and com-
RSD values are shown in parentheses (%). pared.
The obtained results are shown in Table 3. The importance of
matrix effect was related to the pesticide retention time, proba-
bly due to co-eluting matrix compounds which interacted in the
employed as extraction solvent, which could be explained by their ionization step. High ion suppression was observed for pesticides
low log Kow as well as their lesser stability in that solvent. of intermediate polarities, particularly for pirimiphos methyl that
MSPD gave also acceptable recoveries when acetonitrile–1% showed the highest matrix effect (17% of the response obtained in
acetic acid was used, though the results were somewhat lower pure solvent). The matrix effect on pesticides eluted at the begin-
(73.4–92.5%) than those obtained by the ultrasonic assisted extrac- ning of the chromatogram (polar OPs) was not important, although
tion method with the same solvent. In both extraction procedures, a small ion suppression was also observed. Dilution of the final
recoveries were improved by the addition of acetic acid to acetoni- matrix extract, to 0.5 g mL−1 instead of 1 g mL−1 , decreased the
trile. The acidification of the extraction medium could prevent the matrix effect for all OPs as shown in Table 3. In case of the more
pH dependant degradation and improve the stability of problematic polar analytes (methamidophos, IMPY, DEAMPY, azamethiphos,
pesticides [22–24]. dichlorvos and TCPY), the matrix effect was negligible, with relative
Therefore, acetonitrile with 1% acetic acid was the solvent response values ranging from 96% to 106% at a matrix concentration
selected for the simultaneous extraction of wide polarity range of 0.5 g mL−1 in the final extract. For OPs of intermediate polari-
of the studied organophosphorus pesticides from sewage sludge ties, though a decrease of matrix effect with dilution was observed,
samples and ultrasonic assisted extraction followed by DSPE was an ion suppression around 50% was still obtained. This could be
the procedure selected for this work because it is a very simple, explained by the co-elution of other ionisable compounds as pre-
easy and cheap method that yielded good recoveries and clean viously reported by Hiemstra and Kok [23]. Because matrix effect
extracts. is negligible, the dilution of final sludge extract to 0.5 g of matrix

Table 6
Concentration of OPs (ng g−1 ) in sewage sludge samplea collected from WWTPs in the province of Madrid.

OP Sludge sample

1 2 3 4 5

Methamidophos n.d. n.d. n.d. n.d. n.d.

IMPY <LOQ 12.4 ± 2.2 15.6 ± 2.3 <LOQ <LOQ
DEAMPY n.d. <LOQ <LOQ <LOQ <LOQ
Azamethiphos n.d. n.d. n.d. n.d. n.d.
Dichlorvos 0.9 ± 0.1 <LOQ <LOQ 1.2 ± 0.3 1.9 ± 0.5
Diazinon 7.3 ± 0.9 3.2 ± 0.4 4.9 ± 0.5 6.2 ± 0.7 4.4 ± 0.8
Pirimiphos <LOQ 12 ± 2.0 <LOQ <LOQ <LOQ
Chlorpyrifos methyl 41.8 ± 1.6 <LOQ 7.6 ± 0.5 13.2 ± 2.0 n.d.
Chlorpyrifos 344 ± 20 164 ± 23 113 ± 19 270 ± 39 173 ± 17
TCPY 99 ± 11 157 ± 21 33 ± 6 92 ± 12 307 ± 55

n.d. not detected, <LOQ detected but not quantified.

a
Values are the mean of three replicates ± SE.
122 A.I. García-Valcárcel, J.L. Tadeo / Analytica Chimica Acta 641 (2009) 117–123

Fig. 3. Total ion chromatogram of a sludge sample fortified at 100 ng/g in positive ion
mode. (1) Methamidophos, (2) IMPY, (3) DEAMPY, (4) azamethiphos, (5) dichlorvos,
(6) diazinon, (7) pirimiphos methyl, (8) chlorpyrifos methyl and (9) chlorpyrifos.

per mL is appropriate to determine the more polar OPs in sewage

sludge samples. However, this is not true for pesticides of interme-
diate polarity and, in order to take into account the matrix effect on
the determination of these pesticides, analysis of OPs was carried
out using matrix- matched standards.
On the other hand, to investigate the variation of the matrix
effect from sludge to sludge, as reported in other matrices such
as vegetables [24–26], the comparison of slopes of matrix-matched
calibration curves of different sludge, at 0.5 g matrix mL−1 , was car-
ried out. The results obtained are presented in Table 3. In most cases,
a similar relative response was obtained for a pesticide in different
sludge samples. In turn, differences in matrix effect from sludge to
sludge are low and standards could be prepared in any of the sludge
samples.

3.4. Method validation

Linearity was evaluated using matrix-matched standards in the

range of 5 ng g−1 to 100 ng g−1 . Good linearity was found for all
pesticides, with coefficients of determination higher than 0.99
(Table 4).
The limits of detection (LOD) and quantification (LOQ) of the
method were calculated as the minimum amount of target analyte
that produced a chromatogram peak with a signal-to-noise ratio
of 3 and 10, respectively. The LOD and LOQ obtained for the dif-
ferent OPs are shown in Table 4. The LOQs ranged from 1 ng g−1 to
14 ng g−1 and the LODs were in the range of 0.3–4.2 ng g−1 , showing
chlorpyrifos the highest values due to the high background noise
obtained at its retention time. In general, the limits of quantification
obtained were lower than those reported for OPs in sewage sludge
[7], except for TCPY, which was somewhat higher. This could be
explained by the analytical technique (GC–MS) used in the men-
tioned paper for the determination of OPs.
The repeatability of the chromatographic determination was Fig. 4. MS/MS chromatograms of a sludge sample (no. 5). See concentration values
evaluated by injecting eight times a matrix-matched standard in Table 6. Black line: quantifier, grey line qualifier.
solution of 20 ng mL−1 in the same day. Good relative standard devi-
ations (RSD) were obtained for peak areas, which ranged from 1.2 to
8.4. The inter-day repeatability was also evaluated by injecting the
same standard solution during three weeks and RSD values lower by Díaz-Cruz and Barceló for the determination of chlorpyrifos
than 20% were obtained (Table 4). and diazinon in sewage sludge by GC–MS after pressurized liquid
Recovery through the method was studied with sludge samples extraction [7]. In the above mentioned study, the recoveries of the
fortified at 10 ng g−1 , 50 ng g−1 and 100 ng g−1 . The results are sum- metabolites were found to be around 40%, however, in the present
marized in Table 5 and a representative chromatogram is shown in study it was higher than 80%. This difference could be explained
Fig. 3. Good recoveries ranging from 79.8 to 113.3 with RSD val- by the analytical method used in the mentioned study, where a
ues <17% were obtained for all the OPs studied. Although these derivatization of metabolites is needed before their determination
values are in the range of acceptable recoveries, the most polar by GC–MS. The good recoveries and repeatability of the proposed
compounds, methamidophos and the OP metabolites, showed a method make it suitable for the quantification of a wide polarity
somewhat lower recovery that could be explained by their lower range of organophosphorus pesticides in sewage sludge at trace
log Kow (Table 1), These recoveries are similar to those reported level.
 A.I. García-Valcárcel, J.L. Tadeo / Analytica Chimica Acta 641 (2009) 117–123
3.5. Levels of OPs in sewage sludge samples
The developed method was applied to the determination of OPs
in various sludge samples obtained from municipal wastewater
treatment plants in the province of Madrid. Only one published
paper on the concentration of some OPs (diazinon, chlorpyrifos
and their respective metabolites IMPY and TCPY) in sewage sludge
from WWTPs located in other Spanish region has been found in the
literature [7].
The levels of OPs found in the studied sludge samples are shown
in Table 6 and representative MS/MS chromatograms of a sludge
sample are depicted in Fig. 4. Methamidophos and azamethiphos
were not detected in any of the analyzed sludge samples and
very low or even not quantified levels of dichlorvos were found.
Chlorpyrifos methyl levels were variable between sludge samples,
with values below LOQ to 41.8 ng g−1 . Pirimiphos was only quan-
tified in one sludge sample, whereas its metabolite DEAMPY was
detected but not quantified in some samples. Diazinon was present
in all sludge samples at low levels (3.2–7.3 ng g−1 ) and its metabo-
lite IMPY was detected in all samples but only quantified in two
sludge samples. The highest concentrations of OPs were found for
chlorpyrifos and its metabolite TCPY. The presence of the metabo-
lites IMPY and TCPY at concentrations higher than those of their
parent compounds could be explained by the rapid metabolism
of these pesticides in mammals [28,29]. Chlorpyrifos levels, from
113 ng g−1 to 344 ng g−1 , are in the same range of the level reported
(210 ng g−1 ) in a previously published paper [7], whereas TCPY
levels found in our study (33–307 ng g−1 ) are higher than the con-
centration reported in that paper (5.42 ng g−1 ) and those of IMPY
are somewhat lower. These differences may be explained by the
different origin of sludge samples.
4. Conclusion
In this work, an analytical method for the determination of OPs
in sewage sludge by LC–ESI-MS/MS has been developed. A wide
polarity range of currently used OPs, including their main metabo-
lites, were extracted from sludge samples by ultrasonic assisted
extraction using different solvents and 1% acetic acid in acetonitrile
gave the best results. Although extracts were cleaned up by disper-
sive solid phase extraction, it was necessary to use matrix-matched
standards to take into account the matrix effect in the quantifica-
tion of OPs by LC–MS/MS with electrospray ionisation in positive
or negative mode.
The analytical method was validated and its suitability for the
determination of OPs in sludge at trace levels was demonstrated.
This method was successfully applied to the analysis of OPs in var-
123
ious sludge samples from waste treatment plants of the province
of Madrid, Spain. From our point of view, the developed method is
simple and rapid and it will facilitate the analysis of OPS and their
metabolites in complex samples such as sewage sludge.
Acknowledgments
We thank E. Molero for her help in the experimental work and
Spanish Ministry of Science and Innovation (RTA 2008-00040) for
financial support.
References
[1] S.A. Reinecke, A.J. Reinecke, Ecotox. Environ. Safe 66 (2007) 244.
[2] C. Timchalk, R.J. Notan, A.C. Mendrala, A.D. Dittember, K.A. Brzak, J.L. Mattsson,
Toxicol. Sci. 66 (2002) 34.
[3] C. Basheer, A.A. Alnedhary, B.S.M. Rao, H.K. Lee, Anal. Chim. Acta 605 (2007)
147.
[4] U. Uygun, B. Senoz, H. Koksel, Food Chem. 109 (2008) 355.
[5] H.J. Stam, J. Chromatogr. A 892 (2003) 47.
[6] J. Fillion, J. Sauve, J. Selwyn, J. AOAC Int. 83 (2000) 698.
[7] M.S. Díaz–Cruz, D. Barceló, J. Chromatogr. A 1132 (2006) 21.
[8] C. Sánchez-Brunete, B. Albero, J.L. Tadeo, J. Agric. Food Chem. 52 (2004)
1445.
[9] F. Hernandez, J.V. Sancho, O. Pozo, A. Lara, E. Pitarch, J. Chromatogr. A 939 (2001)
1.
[10] S.J. Lehotay, J. AOAC Inter. 90 (2007) 485.
[11] A.O. Olson, J.V. Nguyen, M.A. Sadowski, D.B. Barr, Anal. Bioanal. Chem. 376
(2003) 808.
[12] L.N. Williamson, M.G. Bartlett, J. Liq. Chromatogr. Relal. Technol. 30 (2007) 273.
[13] J.A. Mendiola, M. Herrero, A. Cifuentes, E. Ibáñez, J. Chromatogr. A 1152 (2007)
234.
[14] J. Martinez Vidal, I. Martinez Salvador, T. López-López, A. Garrido Frenich, Anal.
Bioanal. Chem. 383 (2005) 1106.
[15] C. Sánchez-Brunete, E. Miguel, J.L. Tadeo, Talanta 70 (2006) 1051.
[16] M. Anastassiades, S.J. Lehotay, D. Staijnbaher, F.J. Schenck, J. AOAC Int. 86 (2003)
412.
[17] C. Sánchez-Brunete, E. Miguel, J.L. Tadeo, Talanta 74 (2008) 1211.
[18] C. Sánchez-Brunete, E. Miguel, J.L. Tadeo, J. Chromatogr. A 1148 (2007) 219.
[19] L. Villarosa, M.J. McCormick, P.D. Carpenter, P.J. Marriott, Int. J. Environ. Anal.
Chem. 54 (1994) 93.
[20] J.V. Sancho, O.J. Pozo, F. Hernandez, Rapid Commun. Mass Spectrom. 14 (2000)
1485.
[21] B.A. Ingelse, R.C.J. Van Dam, R.J. Vreeken, H.G.J. Mol, O.M. Steijger, J. Chromatogr.
A 918 (2001) 67.
[22] K. Banerjee, D.P. Oulkar, S. Dasgupta, S.B. Patil, S.H. Patil, R. Savant, P.G. Adsule,
J. Chromatogr. A 1173 (2007) 98.
[23] M. Hiemstra, A. Kok, J. Chromatogr. A 1154 (2007) 3.
[24] K. Mastovská, S.J. Lehotay, J. Chromatogr. A 1040 (2004) 259.
[25] A. Kruve, A. Künnapas, K. Herodes, I. Leito, J. Chromatogr. A 1187 (2008) 58.
[26] H.G.J. Mol, R.C.J. Van Dam, O.M. Steijger, J. Chromatogr. A 1015 (2003) 119.
[27] C.D.S. Tomlin, The Pesticide Manual, 12th ed., The British Crop Council, UK,
2000.
[28] C. Timchalk, A. Busby, J.A. Campbell, L.L. Needham, D.B. Barr, Toxicology 237
(2007) 145.
[29] T. Rodriguez, L. Younglove, C.S. Lu, A. Funez, S. Weppner, D.B. Barr, R.A. Fenske,
Int. J. Occup. Environ. Health 12 (2006) 312.