Food Chemistry 249 (2018) 77–83

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Pesticide determination in sweet peppers using QuEChERS and LC–MS/MS T

⁎
Elisa Helena da Costa Morais , Carol Hollingworth Collins, Isabel Cristina Sales Fontes Jardim
Analytical Chemistry Department, Institute of Chemistry, State University of Campinas, 13083-970 Campinas, SP, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: In this work, a rapid, effective, and safe method, generating only a small amount of waste, based on the citrate
Pesticides version of QuEChERS was optimized and validated for multiresidue determination of pesticides of different
Capsicum annuum L. classes in sweet green peppers, determined by liquid chromatography coupled with tandem mass spectrometry.
QuEChERS citrate The matrix components influenced the measurement of the pesticides by the developed analysis technique, so
Liquid chromatography–tandem mass
that, analytical curves were prepared using pesticide-free matrix extracts for quantification of the analytes. The
spectroscopy
method provides satisfactory accuracy verified by recoveries of 70–120%, and good precision (coefficients of
Validation method
variation ≤20%). It also showed selectivity, linearity of response, and lower limits of quantification than the
maximum limit of residue for each compound, as established by ANVISA and Codex Alimentarius.

1. Introduction Brazil were deemed inadequate, i.e., contained residues of pesticides

Due to population growth and the consequent demand for food,
agriculture has increased intensely in productivity, using lesser acreage.
However, due to the undesired incidence of diseases of bacterial,
fungal, nematological and viral origin, to arthropods that cause dis-
turbances and to weed seeds, with consequent environmental im-
balance, there has been an increase in the use of pesticides, often ap-
plied inappropriately and indiscriminately, which causes
contamination of crops and, consequently, adverse health effects for
human and animals. In addition, the use of pesticides can cause con-
tamination of surface water, groundwater and soil and lead to animal
mortality (Ahmed, Randhawa, Yusuf, & Khalid, 2011).
In order to control the use of pesticides and limit concentrations of
residues in foods, many agencies, such as the Brazilian Health
Surveillance Agency (ANVISA) and Codex Alimentarius, have estab-
lished maximum residue limits (MRL) for pesticides.
Sweet peppers (Capsicum annuum L.) (Buckler, Pearsall, & Holtsiord,
1998), whose world production is approximately 32 million tonnes,
according to the latest survey conducted by the Food and Agriculture
Organization of the United Nations (FAOSTAT, 2017), are consumed
due their taste and to the presence of compounds that prevent some
diseases (Collera-Zúñiga, Jiménez, & Gordillo, 2005; Nishino,
Murakoshi, Tokuda, & Satomi, 2009; UNICAMP, 2011) and deserve
attention due to possible irregularities in control of pesticide residues
(ANVISA, 2014; FDA, 2017). According to a report of activities released
by ANVISA in 2014 regarding the Program for the Analysis of Pesticide
Residues in Food (PARA), 89% of the sweet peppers samples grown in
not allowed or at levels above the MRL. The Pesticide Residue Mon-
itoring Program 2015, published in 2017 and conducted by the U.S.
Food and Drug Administration (FDA), revealed that 9.0% of samples
obtained from several Brazilian states contained irregularities.
Due to negligence and the adverse effects of pesticides, there is in-
creased interest in conducting research addressing the development and
validation of methods to monitor the presence of multiresidues of
pesticides in food matrices, such as sweet peppers. In this context, one
of the steps for the determination of pesticide residues in food is the
preparation of the sample for extraction and concentration of the
analytes, as well as for the clean-up of the samples. However, some
techniques have limitations and drawbacks, such as not providing high
recovery of the compounds of interest, efficient clean-up of the samples
and sufficient accuracy of results, and often because they are time
consuming, costly and difficult to apply.
The QuEChERS (quick, easy, cheap, effective, rugged, safe) tech-
nique of sample preparation was introduced in 2003 by Anastassiades,
Lehotay, Štajnbaher and Schenck, in order to overcome the limitations
and disadvantages of some traditional extraction techniques for multi-
residues of pesticides. The method has been widely used in the de-
termination of pesticide residues in various matrices, being a fast, easy,
economical, effective, rugged and safe method. Moreover, the appli-
cation of this method, including the acetate buffer (Lehotay, Maštovská,
& Lightfield, 2005) and citrate buffer (Anastassiades, Scherbaum,
Taşdelen, & Štajnbaher, 2007) versions, enables the extraction of acidic,
basic and neutral compounds, obtaining precise and accurate results
due to high recoveries of the analytes.

⁎
Corresponding author.
E-mail address: elisahelenabq@yahoo.com.br (E.H. da Costa Morais).

https://doi.org/10.1016/j.foodchem.2017.12.092
Received 4 November 2016; Received in revised form 7 December 2017; Accepted 31 December 2017
Available online 02 January 2018
0308-8146/ © 2018 Elsevier Ltd. All rights reserved.
E.H. da Costa Morais et al.
Among the modern analysis techniques, liquid chromatography
(LC) coupled to different detectors stands out, due to its facility in ef-
fecting separations, identifications and quantification the species pre-
sent in a sample (Braga et al., 2007; Collins, Braga, & Bonato, 2006). In
recent years, liquid chromatography coupled to tandem mass spectro-
metry (LC–MS/MS) has shown great progress in terms of technological
development and application (Kmellár, Pareja, Ferrer, Fodor, &
Fernández-Alba, 2011). Satisfactory results have been obtained when
this technique is used, since it combines the high selectivity and effi-
ciency of separation by liquid chromatography with obtaining identi-
fication information of the separated compounds, due to the high de-
tectability and increased selectivity of mass spectrometry, allowing the
determination of low concentrations of mixtures of pesticide residues
belonging to different chemical groups in complex matrices in a single
analysis (Vékey, 2001; Niessen, 2006).
Based on this, in the present study a method was optimized, vali-
dated and applied to samples of commercially available sweet green
peppers (Capsicum annuum L.) for the determination of residues of
pesticide that are permitted to be applied to the crop, as well as some
that are not allowed, but were found, according to the PARA report.
This work used the techniques of QuEChERS for sample preparation
and LC–MS/MS with electrospray ionization and triple quadrupole
detection in the multiple reaction monitoring (MRM) mode for the
development and validation of a method for multiresidue determina-
tion of pesticides in sweet green pepper.
2. Materials and methods
2.1. Reagents and solvents
The analytical standards of pesticides, with their respective purity
and suppliers, for: clomazone (98.1% w/w), difenoconazole (97.0% w/
w), ethion (97.8% w/w), methamidophos (98.5% w/w), methomyl
(99.9% w/w), pyraclostrobin (99.9% w/w), pyriproxyfen (99.1% w/w),
thiabendazole (99.8% w/w), tiacloprid (99.9% w/w) and thia-
methoxam (99.7% w/w) were from Fluka (Madrid, Spain); those for
acephate (97.2% w/w), azoxystrobin (99.9% w/w), carbofuran (99%
w/w), fenarimol (99.8% w/w), iprodione (99.3% w/w), metconazole
(99.5% w/w) and tebuconazole (99.8% w/w) were from Pestanal,
Riedel-de Häen (Seelze, Germany); carbendazim (99.1% w/w) and
methiocarb (98.5% w/w) were from Dr. Ehrenstorfer GmbH (Augsburg,
Germany); carbaryl (99.5% w/w) was from Chem Service (West
Chester, PA, U.S.A.); and imidacloprid (99.9% w/w) was from Riedel-
de Häen (Seelze, Germany). Polyethylene membranes of 0.45 µm por-
osity (Millipore – Milli-Q, Bedford, MA, U.S.A.), formic acid (p.a.,
Synth, São Paulo, SP, Brazil), water from a Millipore – Milli-Q (Bedford,
MA, U.S.A.) and methanol (chromatographic grade, J.T. Baker,
Phillipsburg, NJ, U.S.A.) were used to prepare the mobile phases, the
latter was also used in the preparation of standard solutions.
Acetonitrile and methanol (chromatographic grade, J.T. Baker,
Phillipsburg, NJ, U.S.A.), acetone and formic acid (p.a., Synth, São
Paulo, SP, Brazil), ethyl acetate (Mallinckrodt Chemicals, Saint Louis,
MO, U.S.A.), magnesium sulfate, (p.a., Vetec, Rio de Janeiro, RJ,
Brazil), sodium chloride, (p.a., ECIBRA, São Paulo, SP, Brazil), dis-
odium hydrogen citrate sesquihydrate, and trisodium citrate dihydrate
(Sigma Aldrich, Madrid, Spain), primary-secondary amine, PSA
(Varian, Harbor City, CA, U.S.A.) and graphitized carbon (Hexis, São
Paulo, SP, Brazil) were used in sample preparation.
2.2. Stock solutions of pesticides
The standard stock solutions of each pesticide in concentrations of
1000 mg L−1 were prepared by solubilizing each analytical standard in
methanol. The working solutions were prepared by diluting the stock
standard solutions with the same solvent. All were stored at a re-
frigerator temperature of approximately 4 °C.
78
Food Chemistry 249 (2018) 77–83
2.3. Equipment for sample preparation
A multiprocessor (Model Faciclic – Arno, São Paulo, SP, Brazil), an
analytical balance with accuracy of 5 decimal places (Model CP225 D –
Sartorius, Göttingen, Germany), micropipettes of 0.5–10 µL, 10–100 µL
and 100–1000 µL (Eppendorf Research, Hamburg, Germany), a glass
vacuum filtration system, with vacuum pump (Model WP6111560 –
Millipore, Billerica, MA, U.S.A.), a vortex (Model GENIUS 3 – IKA
Vortex ®, Staufen, Germany) and a centrifuge (Model Rotofix 32 –
Analytical, Hettich, Germany) were used.
2.4. Liquid chromatography–tandem mass spectrometry
For the chromatographic analysis an Alliance 2695 liquid chroma-
tograph (Waters, Milford, MA, U.S.A.) was used. The chromatographic
separations were carried out with a Nova-Pak C18 analytical chroma-
tographic column (150 mm × 3.9 mm i.d., 4 µm) (Waters, Milford, MA,
U.S.A.) and a Nova-Pak C18 guard column (20 mm × 3.9 mm i.d.,
4 µm) (Waters, Milford, MA, U.S.A.) with a flow rate of 0.3 mL min−1.
The column was kept at (25 ± 2) °C and the sample injection volume
was 17 µL. Before chromatographic analyses, all the samples were fil-
tered through 0.2 µm PTFE membranes.
The mobile phase used was 0.1% aqueous formic acid (A) and
methanol (B). Gradient elution was used and the amount of methanol
was changed as follows: 0 min – 50%, 12 min – 50%, 13 min – 75%,
30 min – 90%, 33 min – 90%; 35 min – 50%, 43 min – 50%.
A tandem mass spectrometer with triple quadrupole and Z-spray
interface for electrospray (ESI) (Micromass Quattro Micro™ API spec-
trometer, Waters, Milford, MA, U.S.A.), operating in the positive mode
with MRM acquisition was used. The parameters of the mass spectro-
meter for analysis were: capillary voltage – 2 kV, cone extractor voltage
– 3 V, RF lens voltage – 0.2 V, source temperature – 120 °C, desolvation
gas temperature – 400 °C, desolvation gas flow rate – 500 L h−1, cone
gas flow – 50 L h−1. Nitrogen was used as the cone and desolvation gas
and argon as the collision gas at a constant pressure of 2.45 × 10−3
mbar. The data acquisition and processing were performed using Mass
Lynx v. 4.1 software from Waters (Milford, MA, U.S.A.).
2.5. Technique for QuEChERS sample preparation
2.5.1. Samples preparation and fortification of sweet pepper
In the optimization of the QuEChERS sample preparation technique,
organic sweet green peppers obtained in the Campinas, SP, region
(Brazil) were used, from which extracts were prepared and analyzed by
LC–MS/MS to confirm the absence of pesticides studied. Sweet peppers,
free of the pesticides studied, were chopped to pieces and ground in a
household multiprocessor until full homogenization. To a falcon tube of
50 mL capacity were added 10.00000 g of sweet pepper pulp, which
was then fortified with 100 µL of a working solution (concentrations of
50 mg L−1) containing the pesticides studied. The sample was left
standing for about 30 min and then subjected to the QuEChERS method
in the citrate buffer version.
2.5.2. Extraction of pesticides from samples of sweet pepper
After 30 min, the falcon tube containing the crushed and fortified
sample sweet pepper, with added solvent was vortexed. After vortexing,
the partitioning salts were added. The mixture was stirred using the
vortex and then centrifuged. Then, 7 mL of extract were transferred to
another falcon tube containing the salts for the clean-up step. After
vortexing and centrifugation, 5 mL of supernatant was transferred to a
flask containing 50 µL of 5% formic acid (v/v) in acetonitrile, kept
under a flow of nitrogen gas to dryness, resuspended in 1.0 mL of me-
thanol and stored in a glass tube in the freezer until the time of analysis
by LC–MS/MS. The tests and injections were each performed in tripli-
cate. Table 1 shows the salts, their amounts, and the times of agitation
used in the steps of the citrate buffer QuEChERS method.
E.H. da Costa Morais et al. Food Chemistry 249 (2018) 77–83

Table 1 residues. A correlation coefficient of at least 0.990 was targeted to test

Amount of reagents and modes and times of shaking used in the steps of the QuEChERS linearity. The instrument quantification limit (IQL) was determined
method in the citrate buffer version.
based on the analytical curve parameters (Ribani, Bottoli, Collins,
QuEChERS – citrate buffer Jardim, & Melo, 2004) and the method quantification limit, based on
Step Amount of reagents Mode and time of shaking the visual method, was considered as the lowest concentration of the
analyte in the matrix that could be determined with accuracy (re-
Extraction 10 mL of acetonitrile vortex (1 min)
coveries in the range of 70–120%) and precision (RSD < 20%), in
Partition 4 g of MgSO4 vortex (1 min) defining the linear working range. For the evaluation of accuracy, re-
1 g of NaCl centrifugation (5 min; 5000 rpm) covery assays were performed in which sweet peppers samples were
1 g of C6H5Na3O7·2H2O
0.5 g of C6H6Na2O7·1.5H2O
fortified at two concentration levels (MQL and 6 × MQL) in five re-
plicates each, and the extracts analyzed. The accuracy was estimated in
Clean-up 1.05 g of MgSO40.175 g of PSA vortex (30 s)
terms of recovery of the analytes obtained using the analytical curves,
centrifugation (1 min; 6000 rpm)
using Eq. (1). The precision, in terms of repeatability and intermediate
*
rpm: revolutions per minute. precision, were evaluated by preparation, injection and analysis of five
samples of sweet pepper extracts fortified with pesticides at con-
2.5.3. Optimization of the QuEChERS sample preparation technique centrations equal to MQL and 6 x MQL on the same day and on different
Because the sweet peppers have constituents that could interfere days (1 and 14th days), respectively. For this evaluation the coefficients
with the determination of the analytes, the standard solutions were of variation (CV) for the recoveries of the analytes were determined.
prepared in pesticide-free (matrix-matched in extractions). Results below 20% are recommended by SANTE/11945/2015.
Factors which influence the recovery of analytes and which lead to
optimal assay conditions were evaluated: solvent (acetonitrile, acetone,
2.8. Application of the QuEChERS-LC–MS/MS method to sweet green
ethyl acetate and methanol), sorbent (use, or not, of graphitized carbon
pepper samples
mixed with PSA) and use of dry ice in the partition step (Lee et al.,
2011). In this context, the best conditions were those that provided
After validation, the method developed using QuEChERS the citrate
higher recoveries for most of the pesticides studied, with coefficients of
buffer version, with determinations by LC–MS/MS, was applied to
variation lower than 20%, values recommended in the literature
analyze randomly selected samples of sweet green peppers from fifteen
(SANTE/11945/2015), and minimization of the influence of the matrix
markets from four states of southeastern Brazil (Espírito Santo, Minas
components on the ionization of the analytes, i.e., reduction of the
Gerais, Rio de Janeiro and São Paulo), five of these markets being de-
matrix effect.
scribed as selling as organic produce. Five sweet peppers from each
For the calculation of recoveries and of matrix effect the Eqs. (1) and
establishment were used to obtain the mixture for extraction, which
(2) were considered, respectively.
was then analyzed.
concentration of analyte found in the extract ⎞
Recovery (%) = ⎜⎛ ⎟ × 100

⎝ initial concentration of analyte in the extract ⎠ 3. Results and discussion

(1)
Table 2 gives important information and the conditions of frag-
concentration of analyte in the extract ⎞
Matrix effect (%) = ⎜⎛ −1⎟ × 100 mentation and analysis of the studied compounds.
⎝ concentration of analyte in solvent ⎠ When using MS/MS in the MRM acquisition mode, it is possible to
(2) quantify substances that coelute by the monitoring of the transitions of
If the matrix effect < 0, there is suppression with respect to the the precursor – product ions of each substance (Rodrigues, Caldas,
analyte signal, if matrix effect > 0, there is increase in relation to the Furlong, & Primel, 2011). However, chromatographic separation is
signal thereof. important because it provides increased detectability of the analyte.

2.6. Evaluation of the matrix effect 3.1. Optimization of the QuEChERS sample preparation technique

The influence of the constituents of sweet green pepper on the de-
termination of analytes by LC–MS/MS was evaluated by comparing
curves constructed by analysis of standard solutions containing pesti-
cides at seven levels of concentration prepared in solvent to those
prepared in extracts free from sample analytes and by Eq. (2).
2.7. Validation of the analytical method
The method developed was validated evaluating selectivity, line-
arity, precision (repeatability and intermediate precision), accuracy and
limit of quantification (LOQ), according to the Document SANTE/
11945/2015.
Selectivity was evaluated by comparing the chromatograms ob-
tained after injection of extracts free of the pesticides studied and ex-
tracts fortified with the matrix array. The analytical curves constructed
from the adjustment of responses (peak areas corresponding to pesti-
cides) resulting from the analysis of extracts from spiked samples in
seven concentration levels of the analytes (0.1, 0.2, 0.4, 0.6, 0.8, 1 and
1.2 × MRL) to a linear model was used to assess the linearity through
the correlation coefficients (r) of the model and by analyzing the
3.1.1. Evaluation of the extractor solvent
With the intention of obtaining recoveries between 70 and 120%
with coefficients of variation less than 20%, values recommended in the
literature (SANTE/11945/2015), the influence of the solvent on the
extraction of pesticides was studied, applying the QuEChERS method in
the citrate buffer version.
The use of methanol was not feasible, since it solubilized salts used
in the methods, resulting in extracts that had cloudy appearances.
Ethyl acetate, a solvent less polar than acetonitrile and acetone,
caused a significant decrease in recoveries of the polar compounds
acephate and methamidophos, whose values were below 70%.
Moreover, the recovery of the pesticide methomyl was greater than
120%. Acetone resulted in an increase in recoveries of methamidophos
and acephate, compared to acetonitrile, but a decrease in the recoveries
of many other compounds. Furthermore, it was observed that the ex-
tracts obtained in extractions with acetone showed darker green colors
compared to other solvents due to the extraction of a greater amount of
pigment, thus increasing the matrix effect for most compounds, verified
by the intense signal suppression (matrix effect < 0), as shown in the
Fig. 1. Thus, it was decided to use acetonitrile as the extractor solvent.

79
E.H. da Costa Morais et al. Food Chemistry 249 (2018) 77–83

Table 2
Information and conditions of fragmentation and analysis of the studied pesticides.

Pesticide Class Chemical group log Kow pKa M (g mol−1) MRM transition Cone Collision tR (min) Window Dwell
(m/z)* voltage (V) energy (eV) (min) time (s)*

methamidophos* acaricide organophosphorus −0.79 − 141.1 141.9 > 93.1 20.0 12.0 4.6 0–7 0.5
insecticide 141.9 > 124.5 12.0
acephate acaricide organophosphorus −0.85 8.35 183.2 184.0 > 142.5 10.0 10.0 4.6
insecticide 184.0 > 124.5 18.0
thiamethoxam insecticide neonicotinoid −0.13 − 291.7 292.1 > 210.9 15.0 14.0 5.2
292.1 > 180.7 24.0
methomyl* acaricide carbamate 0.09 − 162.2 163.0 > 87.3 10.0 12.0 5.2
insecticide 163.0 > 105.5 12.0
imidacloprid insecticide neonicotinoid 0.57 − 255.7 256.2 > 174.8 20.0 18.0 5.9
256.2 > 209.0 12.0

thiacloprid insecticide neonicotinoid 1.26 − 252.7 253.1 > 125.4 30.0 24.0 7.4 6–16 0.5
253.1 > 89.4 40.0
carbendazim* fungicide benzimidazole 1.48 4.20 191.2 192.1 > 159.8 25.0 16.0 7.6
192.1 > 131.8 30.0
thiabendazole fungicide benzimidazole 2.39 4.73 201.3 202.1 > 174.8 40.0 24.0 8.5
12.00 202.1 > 130.6 36.0
carbofuran* acaricide carbamate 1.80 − 221.3 222.2 > 164.9 15.0 12.0 12.7
insecticide 222.2 > 122.6 22.0
nematicide
termiticide

carbaryl* insecticide carbamate 2.36 10.40 201.2 202.2 > 144.7 10.0 10.0 16.0 10–20 0.9
202.2 > 126.7 24.0

clomazone herbicide isoxazolidinone 2.54 − 239.7 240.2 > 124.5 25.0 20.0 21.3 20–24 0.5
240.2 > 88.5 48.0
azoxystrobin fungicide strobilurin 2.50 − 403.4 404.2 > 372.0 20.0 14.0 21.5
404.2 > 328.9 32.0
methiocarb insecticide carbamate 3.18 − 225.3 226.2 > 168.8 15.0 10.0 22.3
226.2 > 120.6 18.0

*
Pesticides found, but not allowed, according to the ANVISA.
log Kow: logarithm of the octanol-water partition coefficient (pH = 7, temperature = 20 °C); pKa: co-logarithm of the acidity constant (temperature = 25 °C); M: molar mass; MRM:
multiple reaction monitoring; tR: retention time.
The first transition was used for quantification and the second for confirmation.
Dwell time: 15 points per peak.
Source: ANVISA, 2011, 2014; IUPAC, 2017.

3.1.2. Evaluation of sorbent No significant difference was observed in chromatographic re-

Due to the presence of coextractives in extracts obtained from sweet
pepper samples, which were not removed with the use of PSA, it be-
came necessary to evaluate the effectiveness of the addition of a gra-
phitized carbon sorbent in the clean-up of these extracts (Anastassiades
et al., 2007). In this step, 0.99750 g of MgSO4, 0.17500 g of PSA and
0.05250 g of graphitized carbon were added to the extract.
sponses for most analytes. However, graphitized carbon positively in-
fluenced the cleaning of the extracts, which minimized matrix effects,
providing recoveries between 70 and 120%, with coefficients of var-
iation less than 20% for all the pesticides studied. Thus, it was decided
to add this sorbent in the clean-up step.

Fig. 1. Matrix effects in relation to pesticides when the solvent was evaluated.

80
E.H. da Costa Morais et al.
Fig. 2. Influence of the dry ice in the QuEChERS method for extraction of 21 pesticides from in sweet green peppers. *Error bars refer to standard deviations (N = 3).
3.1.3. Evaluation of dry ice
Dry ice was used, substituting magnesium sulfate, for the purpose of
simultaneously promoting the separation of the aqueous and organic
phases, and removing coextractives of the matrix with low solubility in
acetonitrile, based on work carried out by Anastassiades et al. (2007)
and Lee et al. (2011). In this context, the partition was performed from
solidification of the water containing the largest portion of the matrix
constituents. This caused an increase in the recovery of most of the
studied pesticides, especially the more hydrophobic (log Kow > 1) ones
(Fig. 2). On the other hand, it caused a significant decrease in the re-
coveries of more hydrophilic analytes like methamidophos, acephate,
imidacloprid, thiamethoxam and ethion, due to the freezing of the
aqueous phase (Fig. 2). As the recoveries of the pesticides methami-
dophos and acephate were lower than 70% it was decided not to use dry
ice in the partition step.
3.2. Evaluation of the matrix effect
Analyzing the analytical curves, matrix effects were observed for all
studied pesticides except for tebuconazole, since there are inclination
differences between the curves fortified in the solvent and in the matrix
extract. The chromatographic responses related to pesticides from the
injection of standard solutions prepared in pesticide-free matrix extract
were smaller than those from the injection of standard solutions pre-
pared in solvent (signal suppression), evidenced also by the matrix ef-
fects much smaller than zero. This demonstrates the influence of the
coextractives, with regard to the decrease in the analyte signal. The
matrix effect was more pronounced for the more polar compounds
methamidophos, acephate, thiamethoxam, methomyl and imidacloprid
(log Kow < 1). According to Bonfiglio, King, Olah, and Merkle (1999)
the ionization efficiency of the polar compounds is more influenced by
the presence of coeluates from the matrix when compared to the apolar
compounds, mainly because most of the coextractives of the matrix are
polar and elute at the start of the chromatographic run.
Therefore, due to these results, the analytical curves used in the
validation of the analytical method and for quantification of real sam-
ples should be constructed using pesticide-free matrix extracts.
3.3. Validation of the analytical method
Comparing the chromatograms of extracts free of the pesticides
under study, with the chromatograms of extracts fortified with these, it
81
Food Chemistry 249 (2018) 77–83
is observed that there are no chromatographic peaks at the same re-
tention times as the analytes, showing that the method is selective.
Table 3 presents the validation data of the analytical method de-
veloped.
The analytical curves constructed from the results of injection of
extracts obtained from spiked samples in seven concentration levels of
the analytes between 1 and 4800 µg kg−1 were used to evaluate line-
arity. Coefficients of correlation greater than 0.9917, featuring a good
fit of the model to the observed responses, and random distribution of
residues left by adjusting the linear model, with deviations in the range
of ± 20%, indicate linearity of the method for the pesticides studied in
the concentration ranges considered.
Satisfactory recoveries between 70 and 120%, and coefficients of
variation lower than 20% for all analytes, respectively, indicate the
accuracy and precision of the method.
The method developed can be applied to samples of sweet pepper
for the quantification of pesticides studied at concentrations lower than
the maximum residue limit established by ANVISA (2015) and the
Codex Alimentarius. (2016), since the limits of quantification con-
sidered were lower or equal to the MRL of each compound.
There are some studies reported in the literature related to the de-
termination of pesticides in sweet pepper, as shown in Table 4. How-
ever, they do not present a study of the evaluation of factors that in-
fluence the recovery of analytes, as well as in the analysis, in order to
minimize the matrix effect, which should be evaluated and considered
in the quantification of the analytes.
3.4. Application of the QuEChERS-LC–MS/MS method to sweet pepper
samples
The method was applied to samples of sweet green pepper obtained
in four states in the Southeastern region of Brazil (Espírito Santo: three
samples; Minas Gerais: three samples, Rio de Janeiro: three samples,
and São Paulo: six samples). In most of the sweet green pepper samples
analyzed at least one of the pesticides studied was detected, including
in samples indicated to be of organic origin. It is worth noting the
presence of the pesticide acephate at a concentration equal to
1237 µg kg−1, above of the MRL. However, most sweet pepper samples
contained pesticides at concentrations below the limits of quantification
or below the MRL.
E.H. da Costa Morais et al. Food Chemistry 249 (2018) 77–83

Table 3
Maximum Residue Limit (MRL) and merit figures of the analytical method developed.

Pesticides MRL MRL (µg kg−1) IQL g MQL Concentration Correlation Accuracy Repeatability Intermediate precision
(µg kg−1) Codex L−1)* (µg kg−1) range (µg kg−1) coefficient (r)
ANVISA Alimentarius Recovery (%)* Coefficient of variation (%)

MQL 6 x MQL MQL 6 x MQL MQL 6 x MQL

methamidophos − − 412 100 100–1200 0.9933 77 72 11 4 9 6

acephate 1000 − 516 100 100–1200 0.9927 79 70 10 10 10 9
thiamethoxam 200 − 132 20 20–240 0.9917 74 76 4 13 20 15
methomyl − − 45 10 10–120 0.9978 88 87 10 7 10 6
imidacloprid 500 − 300 50 50–600 0.9979 88 78 19 14 20 14
thiacloprid 200 1000 11 20 20–240 0.9941 95 89 7 6 8 5
carbendazim − − 6 2 2–24 0.9971 112 111 12 6 16 13
thiabendazole 2000 − 21 200 200–2400 0.9954 71 107 11 3 9 9
carbofuran − − 5 1 1–12 0.9995 99 99 5 4 8 4
carbaryl − 5000 42 10 10–120 0.9995 117 98 8 4 11 5
clomazone 50 − 4 5 5–60 0.9928 92 95 6 4 7 6
azoxystrobin 500 − 2 50 50–600 0.9946 95 106 9 4 8 3
methiocarb 50 2000 5 5 5–60 0.9929 100 89 6 7 12 10
fenarimol − 500 8 2 2–24 0.9981 99 95 6 3 8 6
iprodione 4000 − 245 400 400–4800 0.9963 117 80 7 6 13 19
tebuconazole 100 1000 0.6 10 10–120 0.9933 88 95 5 4 16 8
pyraclostrobin 1000 − 3 100 100–1200 0.9937 103 105 7 4 11 7
metconazole 100 − 0.6 10 10–120 0.9958 113 105 7 8 12 8
difenoconazole 500 − 2 50 50–600 0.9952 91 96 10 8 8 7
ethion 1000 − 9 100 100–1200 0.9982 100 106 8 13 11 12
pyriproxyfen 500 − 2 50 50–600 0.9982 85 102 6 3 11 5

*
MRL: maximum residue limit; IQL: instrument quantification limit; MQL: method quantification limit.
Recovery = 100 × (concentration of analyte found in the extract/concentration of the analyte in the extract).
Source: ANVISA, 2015; Codex Alimentarius (2016).

Table 4
Bibliographical survey of methods developed for determination of pesticides in sweet peppers.

Classes of Chemical groups Technique Evaluation Evaluation Matrix- Analysis MQL range Accuracy Precision Reference
pesticides of sample of factors of the matched technique (µg kg−1) range range
(Number of preparation correlated to matrix quantification Recovery Coefficient
pesticides the matrix effect (%) of variation
analyzed) effect (%)

acaricide organophosphorus ESL – yes yes GC–FPD not 73–116 ≤20 Patel, Fussell,
insecticide determined Macarthur,
nematicide Goodall, and
(37) Keely (2004)
acaricide benzimidazole PSI – yes no LC–MS not not not Evard, Kruve,
fungicide carbamate determined determined determined Lõhmus, and
insecticide imidazole Leito (2015)
nematicide
(5)
acaricide several QuEChERS – yes yes UHPLC–MS/ 10–25 70–120 < 20 Kemmerich
fungicide (citrate MS et al. (2015)
herbicide buffer
insecticide version)
(81)
insecticide organophosphorus MSPE – yes no GC-NPD not 91–106 < 20 Mahpishania-
acaricide determined n et al.
(6) (2015)
insecticide (1) pyrethroid QuEChERS – yes no LC-FLD 3 90–111 < 20 Watanabe
(original) and Baba
(2015)
acaricide benzimidazole QuEChERS solvent yes yes LC–MS/MS 2–400 70–117 < 20 This method
fungicide carbamate (citrate sorbent
herbicide dicarboximide buffer dry ice
insecticide isoxazolidinone version)
nematicide neonicotinoid
termiticide organophosphorus
(21) pyridyl oxypropyl
ether
strobilurin triazole
pyrimidine

*
LC–MS/MS: liquid chromatography–tandem mass spectrometry; ESL: extraction solid liquid; GC-FDP: gas chromatography with flame photometric detection; UHPLC–MS/MS: ultra-high-
performance liquid chromatography-tandem mass spectrometry; PSI: paper spray ionization; LC–MS: liquid chromatography–mass spectrometry; MSPE: magnetic solid phase extraction;
GC-NPD: gas chromatography-nitrogen phosphorus detection; LC-FLD: liquid chromatography-fluorescence detection.

82
E.H. da Costa Morais et al.
4. Conclusions
The technique of sample preparation using QuEChERS with citrate
buffer and of determination using liquid chromatography coupled to
tandem mass spectrometry for multiresidue analysis of pesticides in
sweet pepper (Capsicum annuum L.), resulted in a rapid (analysis time of
35 min), effective and safe procedure, with only a small amount of
waste generation. Another advantage of the method is related to the
high recoveries of the pesticides, 70–120%, and to the precision of the
results, verified by the coefficients of variation of ≤20%, recommended
literature values.
It was observed that the matrix components influenced the mea-
surement of the pesticides by this analysis technique; therefore, ana-
lytical curves were prepared using pesticide-free matrix extracts for
quantification of the analytes.
The method developed can be applied reliably to samples of sweet
pepper for the determination of the pesticides studied, since it is se-
lective, showed response linearity, precision and accuracy and quanti-
tation limits lower than the maximum residue limit of each compound.
Acknowledgments
The authors acknowledge financial support and fellowships from
the Conselho Nacional de Desenvolvimento Científico e Tecnológico
(CNPq), the Fundação de Amparo à Pesquisa do Estado de São Paulo
(FAPESP, Process No 2006/57897-4) and the Instituto Nacional de
Ciências e Tecnologias Analíticas Avançadas (INCTAA).
References
Ahmed, A., Randhawa, M. A., Yusuf, M. J., & Khalid, N. (2011). Effect of processing on
pesticide residues in food crops – A review. Journal of Agricultural Research, 49,
379–390.
Anastassiades, M., Lehotay, S. J., Štajnbaher, D., & Schenck, F. J. (2003). Fast and easy
multiresidue method employing acetonitrile extraction/partitioning and “dispersive
solid-phase extraction” for the determination of pesticide residues in produce. Journal
of AOAC International, 86, 412–431.
Anastassiades, M., Scherbaum, E., Taşdelen, B., & Štajnbaher, D. (2007). Recent devel-
opments in QuEChERS methodology for pesticide multiresidue analysis. In H.
Ohkawa, H. Miyagama, & P. W. Lee (Eds.). Pesticide chemistry: Crop protection, public
health, environmental safety (pp. 439–458). Weinheim: Wiley-VCH.
ANVISA – Agência Nacional de Vigilância Sanitária. (2011). http://portal.anvisa.gov.br/
programa-de-analise-de-registro-de-agrotoxicos-para Accessed 25.11.17.
ANVISA – Agência Nacional de Vigilância Sanitária. (2014). http://portal.anvisa.gov.br/
programa-de-analise-de-registro-de-agrotoxicos-para Accessed 25.11.17.
ANVISA – Agência Nacional de Vigilância Sanitária. (2015). http://portal.anvisa.gov.br/
registros-e-autorizacoes/agrotoxicos/produtos/monografia-de-agrotoxicos Accessed
25.11.17.
Bonfiglio, R., King, R. C., Olah, T. V., & Merkle, K. (1999). The effects of sample pre-
paration methods on the variability of the electrospray ionization response for model
drug compounds. Rapid Communications in Mass Spectrometry, 13, 1175–1185.
Braga, J. W. B., Bottoli, C. B. G., Jardim, I. C. S. F. J., Goicoechea, H. C., Olivieri, A. C., &
Poppi, R. J. (2007). Determination of pesticides and metabolites in wine by high
performance liquid chromatography and second-order calibration methods. Journal
83
Food Chemistry 249 (2018) 77–83
of Chromatography A, 1148, 200–210.
Buckler, E. S., IV, Pearsall, D. M., & Holtsiord, T. P. (1998). Climate, plant ecology, and
Central Mexican archaic subsistence. Current Anthropology, 39, 152–164.
Codex Alimentarius. (2016). http://www.fao.org/fao-who-codexalimentarius/standards/
pesticide-mrls/pt/ Accessed 25.11.17.
Collera-Zúñiga, O., Jiménez, F. G., & Gordillo, R. M. (2005). Comparative study of car-
otenoid composition in three Mexican varieties of Capsicum annuum L. Food
Chemistry, 90, 109–114.
Collins, C. H., Braga, G. L., & Bonato, P. S. (2006). Fundamentos de cromatografia.
Campinas: Editora Unicamp.
Evard, H., Kruve, A., Lõhmus, R., & Leito, I. (2015). Paper spray ionization mass spec-
trometry: Study of a method for 8 fast-screening analysis of pesticides in fruits and
vegetables. Journal of Food Composition and Analysis, 41, 221–225.
FAOSTAT – Food and Agriculture Organization of the United Nations. (2017). http://
www.fao.org/faostat/en/#home Accessed 25.11.17.
FDA – U.S. Food and Drug Administration. (2017). https://www.fda.gov/Food/
FoodborneIllnessContaminants/Pesticides/ucm2006797.htm Accessed 25.11.17.
IUPAC – The PPDB - Pesticide Properties Database International Union of Pure and
Applied Chemistry. (2017). http://sitem.herts.ac.uk/aeru/iupac/ Accessed 25.11.17.
Kemmerich, M., Rizzetti, T. M., Martins, M. L., Prestes, O. D., Adaime, M. B., & Zanella, R.
(2015). Optimization by central composite design of a modified QuEChERS method
for extraction of pesticide multiresidue in sweet pepper and analysis by ultra-high-
performance liquid chromatography–tandem mass spectrometry. Food Analytical
Methods, 8, 728–739.
Kmellár, B., Pareja, L., Ferrer, C., Fodor, P., & Fernández-Alba, A. R. (2011). Study of the
effects of operational parameters on multiresidue pesticide analysis by LC–MS/MS.
Talanta, 84, 262–273.
Lee, S. W., Choib, J. H., Cho, S. K., Yu, H. Y., El-Aty, A. M. A., & Shim, J. H. (2011).
Development of a new QuEChERS method based on dry ice for the determination of
168 pesticides in paprika using tandem mass spectrometry. Journal of Chromatography
A, 1218, 4366–4377.
Lehotay, S. J., Maštovská, K., & Lightfield, A. R. (2005). Use of buffering and other means
to improve results of problematic pesticides in a fast and easy method for residue
analysis of fruits and vegetables. Journal of AOAC International, 88, 615–629.
Mahpishanian, S., Sereshti, H., & Baghdai, M. (2015). Superparamagnetic core-shells
anchored onto graphene oxide grafted with phenylethyl amine as a nano-adsorbent
for extraction and enrichment of organophosphorus pesticides from fruit, vegetable
and water samples. Journal of Chromatography A, 1406, 48–58.
Niessen, W. M. A. (2006). Liquid chromatography–mass spectrometry (3th ed.). Leiden: CRC
Press.
Nishino, H., Murakoshi, M., Tokuda, H., & Satomi, Y. (2009). Cancer prevention by
carotenoids. Archives of Biochemistry and Biophysics, 483, 165–168.
Patel, K., Fussell, R. J., Macarthur, R., Goodall, D. M., & Keely, B. J. (2004). Method
validation of resistive heating-gas chromatography with flame photometric detection
for the rapid screening of organophosphorus pesticides in fruit and vegetables.
Journal of Chromatography A, 1046, 225–234.
Ribani, M., Bottoli, C. B. G., Collins, C. H., Jardim, I. C. S. F., & Melo, L. F. C. (2004).
Validação em métodos cromatográficos e eletroforéticos. Química Nova, 27, 771–780.
Rodrigues, S. A., Caldas, S. S., Furlong, E. B., & Primel, E. G. (2011). Otimização e
validação de método empregando QuEChERS modificado e LC-ESI-MS/MS para
determinação de agrotóxicos em cebola. Química Nova, 34, 780–786.
European Commission Directorate-General for Health and Food Safety, Guidance docu-
ment on analytical quality control and method validation procedures for pesticides residues
analysis in food and feed, SANTE/11945/2015, December 1, 2015. http://www.eurl-
pesticides.eu/docs/public/home.asp?LabID=100∗Lang=EN Accessed 25.11.17.
UNICAMP – Universidade Estadual de Campinas. (2011). http://www.unicamp.br/nepa/
taco/ Accessed 25.11.17.
Vékey, K. (2001). Mass spectrometry and mass-selective detection in chromatography.
Journal of Chromatography A, 921, 227–236.
Watanabe, E., & Baba, K. (2015). Highly sensitive quantification of pyrethroid insecticide
etofenprox in vegetables with high-performance liquid chromatography and fluor-
escence detection. Journal of Chromatography A, 1385, 35–41.