Journal of Food Composition and Analysis 103 (2021) 104090

Contents lists available at ScienceDirect

Journal of Food Composition and Analysis

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

Original Research Article

Optimization of straightforward methods for the monitoring of organic and

inorganic contaminants in strawberries
Ignacio Machado *, Isabel Dol
BIOESP Group, Analytical Chemistry, Faculty of Chemistry, DEC, Universidad de la República (UdelaR), Av. Gral. Flores 2124, P.O. Box 1157, 11800, Montevideo,
Uruguay

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

Keywords: Straightforward methods for the monitoring of chlorothalonil, cyprodinil, fludioxonil, metalaxyl-m, mancozeb
Strawberries and ethylenethiourea in strawberries, based on modified QuEChERS protocols followed by HPLC-DAD, were
Fungicides validated according to DG-SANTE Guidelines. Compounds showed satisfactory recoveries in the range
Cadmium
85.4–100.3% at spiking levels of 0.01, 0.02 and 0.05 mg kg− 1, with RSDs below 15 %. LOQs were in the range
Lead
0.01 – 0.02 mg kg− 1, being below the corresponding maximum residue levels (MRLs). Also, a simple method for
QuEChERS
HPLC-DAD the determination of Cd and Pb, based on a microwave-assisted digestion followed by electrothermal atomic
ETAAS absorption spectrometry (ETAAS), was validated according to Eurachem Guide. Trueness and precision were
evaluated by analysis of a certified reference material (CRM). Results were statistically equivalent to the certified
values at the 95 % significance level, while RSDs were below 5%. Methods were finally applied to the analysis of
commercial samples. Residues of mancozeb and ethylenethiourea were detected in some of the samples but
below the LOQs. Inorganic contaminants were in the range 0.0031 – 0.0088 mg kg− 1 (Cd) and 0.0057 – 0.0212
mg kg− 1 (Pb) (fresh weight), being below the corresponding maximum limits established by national regulation.
The proposed methods can be easily implemented as routine analysis for food safety monitoring.

1. Introduction statistics. Also, commercial products containing chlorothalonil, cypro­

dinil, and fludioxonil as active ingredients are allowed (MGAP, 2021).
Strawberries (Fragaria x ananassa Duch.) cultivation in Uruguay oc­ However, being a product mostly consumed at a national level in
cupies around 120 ha and is produced by approximately 250 producers. Uruguay, it is not often included in the pesticide residues assays, as other
The national average yield is 32 tons/ha occupying the first places in products destined for export.
Latin America. Its production is around 5000 tons per year (INIA, 2021). The evaluation of mancozeb is of utmost importance, since as a
It is considered as one of the most consumed fruits at a national scale, member of ethylenebis(dithiocarbamates) (EBDCs), its main product of
used in many homemade and industrial preparations. This crop can degradation is ethylenethiourea (ETU), which is considered a risk to
reach the harvest period in 6 months, fact that makes it available for sale human health according to the Food and Agriculture Organization
as fresh fruit for long periods during the year (DIGEGRA, 2021). How­ (FAO) of the United Nations (FAO, 2021). Mancozeb is characterized by
ever, the way of cultivation used, allows the presence of fungal diseases, a broad spectrum of activity against various plant pathogens, low acute
which involves the use of fungicides for its control. Within the fungicides toxicity, and low production costs (Crnogorac and Schwack, 2009).
authorized by the Ministry of Agriculture (MGAP) for strawberries, there Since it is a protective fungicide, it needs to be sprayed on the surfaces of
are some commercial products containing metalaxyl-m and mancozeb as leaves and crops. Its oral median lethal dose is 11,200 mg kg− 1 in rats
active ingredients (MGAP, 2021). These products are recommended in (Xu et al., 2017). The degradation of mancozeb far from being benefi­
the presence of phytophthora species (Agrios, 2005; Reglinski et al., cial, increases the risk to human health. Mancozeb can be partially
2009). When infected plants are cut open, a brown discoloration can be eliminated with vigorous washing (up to 30 and 50 % in some fruits) but
seen in the crown vascular tissue or throughout the crown tissue. the rest can be transformed into ETU if submitted, for example, to a
Infection of the roots causes a brown to black root rot. Approximately 6 heating process (FAO, 2021). The degradation of mancozeb into ETU
tons of these products are imported each year, according to the MGAP may occur through several mechanisms, such as hydrolysis, oxidation,

* Corresponding author.
E-mail address: [email protected] (I. Machado).

https://doi.org/10.1016/j.jfca.2021.104090
Received 11 March 2021; Received in revised form 6 July 2021; Accepted 21 July 2021
Available online 24 July 2021
0889-1575/© 2021 Elsevier Inc. All rights reserved.
I. Machado and I. Dol Journal of Food Composition and Analysis 103 (2021) 104090

and microbial degradation (López-Fernández et al., 2014; Rossi Lemes mancozeb (96.8 %), metalaxyl-m (99.4 %) and ethylenethiourea (99.7
et al., 2014). Commercial products containing mancozeb as an active %) were purchased from Fluka (Steinheim, Germany).
ingredient may present ETU as an impurity or as degradation product. Acetonitrile for liquid chromatography was purchased from Merck
The concentration of ETU may increase during storage, decreasing that (Darmstadt, Germany). Magnesium sulphate (MgSO4) (99.0 %), sodium
of the fungicide. This change occurs to a greater extent if temperature chloride (NaCl) (99.1 %), trisodium citrate dihydrate
and humidity are high (IPCS-INCHEM, 2021). (C6H5Na3O7.2H2O) (99.5 %), disodium hydrogencitrate sesquihydrate
In addition to the above-mentioned, scarce information is available (C6H6Na2O7.1.5H2O) (99.0 %), sodium phosphate monobasic
on the scientific literature related to these fungicides’ levels in Uru­ (NaH2PO4) (99.0 %) and disodium hydrogen phosphate (Na2HPO4)
guayan strawberry crops. These facts inspired this work, considering the (99.0 %) were purchased from J.T. Mallinckrodt Baker, Inc. (Phillips­
risk posed over human health, especially the vulnerability of children burg, NJ, USA). Primary secondary amine (PSA) was provided by
that find strawberries as extremely attractive fruits. SUPELCO (Bellefonte, PA, USA). Ethylenediaminetetraacetic acid diso­
Pesticide residues must comply with the Maximum Residue Levels dium salt 2-hydrate (EDTA-2Na) (99 %) was provided by Sigma-Aldrich
(MRLs) established by national and international regulations (Machado (Saint Louis, MO, USA).
et al., 2017). In this contextual framework, there is a need of simple, Individual pesticide stock solutions of 1000 mg L− 1 were prepared in
accurate, fast, and sensitive analytical methods for pesticide residues acetonitrile and stored in the dark at − 18 ◦ C. Mixed solutions used for
analysis in food. Modern tendencies seek for miniaturized and calibration and for spiking were prepared from the stock standards at
cost-effective sample preparation procedures, such as the Quick, Easy, appropriate dilutions using acetonitrile.
Cheap, Effective, Rugged and Safe (QuEChERS) protocol, which is For inorganic contaminants determinations, calibration curves were
suitable for large scale residue analysis in a great variety of matrices prepared by serial dilution of commercial 1000 mg L− 1 stock solutions of
(Anastassiades et al., 2003; Lehotay et al., 2005; Machado et al., 2017). Cd and Pb (Merck, Darmstadt, Germany) in 0.01 mol L− 1 nitric acid
QuEChERS is one of the most popular sample preparation methods for (HNO3), prepared from concentrated HNO3 (Merck, Darmstadt, Ger­
pesticide residues analysis. The general procedure involves a many). A commercial solution of palladium nitrate (Pd(NO3)2) (Merck,
single-phase extraction of analytes from the sample with acetonitrile Darmstadt, Germany) containing 10.0 g L− 1 and a commercial solution
followed by a liquid–liquid partitioning with magnesium sulphate of magnesium nitrate (Mg(NO3)2) (Sigma-Aldrich, St. Louis, MO, USA)
(MgSO4) and NaCl. A simple dehydration and cleanup step, using a containing 20.0 g L− 1, were used to prepare the chemical matrix mod­
dispersive solid-phase extraction, is performed with MgSO4 and primary ifier for Cd and Pb determinations. All glassware was soaked overnight
secondary amine (PSA). The simplicity of the procedure and the easy in 1.4 mol L− 1 HNO3 and then rinsed exhaustively with ultrapure water,
development make it extremely attractive for routine analysis (Anas­ before its use.
tassiades et al., 2003; Zhou et al., 2013). Ultrapure water, ASTM Type I of 18.2 MΩ cm resistivity, was ob­
On the other hand, strawberries may also accumulate potentially tained from a Millipore Direct-Q 5 purifier (Sao Paulo, SP, Brazil).
toxic elements such as cadmium (Cd) and lead (Pb). The elemental
composition of a fruit can be affected by genetic factors, soil charac­ 2.2. Samples
teristics and environmental conditions (Shaltout et al., 2013). The
geographical origin, the use of agrochemicals, the harvesting time and Strawberries samples (approximately 2 kg) used as matrix blank
the type of soil represent the main sources of inorganic contaminants in during optimization and validation of the procedure were collected at a
plants (Machado et al., 2016). In this regard, the MERCOSUR regulation family farm dedicated to grow these crops in Montevideo-Uruguay.
establishes maximum limits (MLs) for small fruits of 0.05 mg kg− 1 for Cd Blank samples were checked showing non-detectable levels of the
and 0.20 mg kg− 1 for Pb, respectively (MERCOSUR, 2011). studied pesticides and heavy metals. Ten different samples (approxi­
The trend of analytical chemistry in the field of metal analysis in mately 2 kg) were also purchased in local markets in Montevideo-
food, is to avoid drastic treatments and to look for efficient quantitative Uruguay, harvested in different farms. All samples were chopped and
multi-extraction procedures under mild conditions, more aligned with homogenized in a commercial blender and stored in the dark at − 18 ◦ C
the principles of green chemistry (Machado et al., 2019). In this regard, until analysis. A national cultivar called ‘Mayte’ was used throughout
microwave-assisted extraction is a very efficient strategy of sample this study.
preparation. This method has the advantage of working with closed Portions destined for inorganic contaminants determinations were
vessels, thus reducing the risk of contamination (Machado et al., 2016). dried in an oven with forced air circulation at 70 ◦ C and stored at 20 ◦ C.
The aim of this work was the optimization of simple methods, based A certified reference material (CRM) consisting of strawberry leaves
on different variations of the QuEChERS protocol, for the analysis of (LGC 7162) was used for trueness and precision evaluation during the
chlorothalonil, cyprodinil, fludioxonil, metalaxyl-m, mancozeb and validation process.
ethylenethiourea in strawberries, using liquid chromatography coupled
to diode array detection (HPLC-DAD). Also, the development and vali­ 2.3. Sample preparation procedures
dation of a microwave-assisted extraction method using dilute acid,
followed by electrothermal atomic absorption spectrometry (ETAAS), For chlorothalonil, cyprodinil, fludioxonil and metalaxyl-m extrac­
for Cd and Pb assessment. This group of pesticides was selected since tion, a modification of QuEChERS citrate protocol was employed (CEN
they are some of the most extensively used fungicides in Uruguayan Standard Method EN 15662, 2008). A 10.0 g portion of blended sample
strawberry crops. In addition, this two heavy metals were chosen, since was weighed in a polypropylene conical centrifuge tube. Next, 10.0 mL
they are included in the MERCOSUR regulation, and are also the two of acetonitrile were added, and the sample was shaken by hand for 1
inorganic contaminants mostly found in fruits in Uruguay (MGAP, min. Afterwards, 4 g of MgSO4 anhydrous, 1 g of NaCl, 1 g of
2021). Therefore, this work will contribute with simple analytical tools C6H5Na3O7.2H2O and 0.5 g of C6H6Na2O7.1,5H2O were added and the
for the surveillance of both organic and inorganic contaminants, which sample was shaken in a vortex for 1 min.
are usually reported separately. For mancozeb extraction, a 10.0 g portion of blended sample was
homogenized with 5.0 mL of a pH 7.8 buffer solution (0.2 M NaH2PO4 +
2. Materials and methods 0.2 M Na2HPO4), 0.1 g of L-cysteine and 0.5 g of EDTA-2Na. Next, 10.0
mL of 0.05 M dimethyl sulphate solution in acetonitrile were added, and
2.1. Reagents the sample was shaken by hand for 5 min. Then, 4 g of MgSO4 anhydrous
and 1 g NaCl were added, and the mixture was shaken by hand again for
Chlorothalonil (99.3 %), cyprodinil (99.8 %), fludioxonil (99.9 %), 5 min.

2
I. Machado and I. Dol Journal of Food Composition and Analysis 103 (2021) 104090

For ETU extraction, a 10.0 g portion of blended sample was weighed Table 1
into a 50 mL PP conical centrifuge tube. After addition of 10.0 mL of 1 % Temperature programs for the determination of Cd and Pb by ETAAS.
v/v NH3.H2O in acetonitrile, the sample was shaken by hand for 5 min. Stage Temperature (ºC) Ramp rate (ºC s− 1) Hold time (s)
Then, 4 g MgSO4 anhydrous and 1 g NaCl were added into the combi­
Drying 120 10 30
nation of supernatant to induce phase separation and partitioning, and Pyrolysis 800 (Cd)/1100 (Pb) 150 20
the mixture was shaken by hand again for 5 min (Zhou et al., 2013). Atomization 1800 (Cd)
/1900 (Pb) 0 3
In all three cases, the extract was then centrifuged (14,000 g) for 5 Cleaning 2600 0 3
min and 3.0 mL of the supernatant were transferred to a 15 mL PP
centrifuge tube containing 150 mg of PSA and 450 mg of MgSO4
determinations consisted of 10 μL of a solution containing 5 μg of Pd
anhydrous. The extract was vortexed for 30 s and centrifuged again
(NO3)2 and 3 μg of Mg(NO3)2 (Machado et al., 2016). The temperatures
(14,000 g) for 5 min. An aliquot of 2.0 mL was evaporated under a gentle
programs are shown in Table 1.
stream of N2 and reconstituted to 1.0 mL with acetonitrile for HPLC-DAD
analysis. Prior to injection, the sample was filtered through a Millex-FG
0.45 μm PVDF filter provided by Millipore (Mildford, CT, USA). 2.5. Validation procedures
For Cd and Pb determinations, a microwave-assisted acid digestion
was carried out using a CEM Mars 6 microwave oven (CEM, Charlotte, Method validation for pesticide residues was performed according to
NC, USA) provided with 12 Easy Prep Plus® vessels. Portions of 0.5 g of the Directorate-General for Health and Food Safety (DG-SANTE)
sample were accurately weighted into each vessel and 10.0 mL of 4.2 Guidelines (European Commission DG-SANTE, 2019). The figures of
mol L− 1 HNO3 were added. The program consisted of a 15-minute ramp merit evaluated were: linearity (calibration curves - residuals < ± 20 %),
time until 200 ◦ C and then holding for 10 min, with power varying trueness (average recovery for spike levels tested - recoveries in the
between 400 and 1800 W, and a maximum pressure of 500 psi. The range 70–120 %), precision (repeatability for spiked levels tested – RSDr
obtained solutions were used for analytical determinations. Reagent < 20 %), detection limit (LOD) (3 s criteria), quantification limit (LOQ)
blanks were also run. (lowest spike level meeting the method performance criteria for trueness
and precision - LOQ ≤ MRL), matrix effect (comparison of solvent
2.4. Analytical determinations standards and matrix-matched standards - assessment of matrix effect
percentage) and uncertainty (estimated expanded measurement uncer­
Analytical determinations by HPLC-DAD were performed with a tainty < 50 %).
Prominence LC-20AT system coupled to a SPD-M20A diode array de­ Method validation for inorganic contaminants was performed ac­
tector provided by Shimadzu (Kyoto, Japan). Chromatographic separa­ cording to Eurachem Guide (Eurachem Guide, 2014). The figures of
tions were performed on a C18 Luna column (150 mm × 4.6 mm, 5 μm) merit evaluated were linearity, trueness, precision, LOD (3 s criteria)
from Phenomenex (Torrance, CA, USA) and the column oven tempera­ and LOQ (10 s criteria).
ture was set at 40 ◦ C.
For chlorothalonil, cyprodinil, fludioxonil and metalaxyl-m de­ 2.6. Statistical analysis
terminations, the mobile phase consisted of (A) water and (B) acetoni­
trile. The following elution program was used: started at 100 % (A) and A Student’s t-test was performed to compare the obtained experi­
increased to 100 % (B) over 10 min where it was held for 15 min (25 min mental means with the certified values of the CRM. Analysis was per­
total run). The percentage of (A) was increased to 100 % again and held formed using MS Excel®. Differences among mean concentrations at a
during 10 min before starting a new run. Injection volume was 50 μL and 5% significance level (p < 0.05) were considered statistically significant
flow rate was 1.2 mL min− 1, for data processing and acquisition. LC (Miller and Miller, 2010).
Solution software version 1.24 from Shimadzu (Kyoto, Japan) was used
to acquire and to process data obtained. Eluents were analyzed at 224 3. Results and discussion
nm (Abd Al-Rahman et al., 2012).
For mancozeb determinations, an extract aliquot of 50 μL was 3.1. Method selection and validation: pesticide residues
injected into the column thermostated at 40 ◦ C with a constant flow-rate
of 1.2 mL min− 1 at the following gradient conditions: water/acetonitrile For this reason, it was chosen in this work, aiming to develop
95:5 (v/v) was changed for 9 min to 10:90 (v/v), before recovering straightforward methods for the determination of the selected fungicides
initial conditions in just 1 min and keeping them for 5 min for column by HPLC-DAD.
stabilization. Eluents were analyzed at 270 nm (López-Fernández et al., With the use of citrate buffer, the pH of the extract gets in the range
2012). 5.0–5.5, facilitating the extraction of low-pH susceptible compounds
For ETU determinations, the mobile phase was composed of water/ and those that present stability problems (Rizzetti et al., 2016). Straw­
acetonitrile 95:5 (v/v) in isocratic mode. The flow rate was 1.0 mL berry is a matrix with a pH value around 3.5, which makes it difficult to
min− 1 and the injection volume was 50 μL. Eluents were analyzed at 240 extract some pesticides without pH adjustment. For this reason, this
nm (Biswas et al., 2003). modification was selected for chlorothalonil, cyprodinil, fludioxonil,
Total determinations of Cd and Pb were performed by ETAAS using a and metalaxyl-m determinations.
Thermo Scientific iCE 3500 spectrometer equipped with auto-sampler Mancozeb is the most problematic among all pesticides since it is
and employing Zeeman correction (Thermo Scientific, Cambridge, almost insoluble in common organic solvents, being difficult to extract
United Kingdom). A transversely heated graphite tube furnace module directly from fruits, so multiresidue extraction procedures with other
from Thermo Fisher Scientific was used (Thermo Scientific, Cambridge, groups of pesticides cannot be applied (Xu et al., 2017). Traditional
United Kingdom). The analytical lines employed were: 228.8 nm (Cd) methods for the analysis of mancozeb consist of acid hydrolysis of the
and 283.3 nm (Pb) and the signal used for quantification was integrated compound to carbon disulfide (CS2) and its subsequent determination by
absorbance (peak-area). Pyrolytically coated graphite tubes from gas chromatography or UV spectrophotometry. But these methods are
Thermo Scientific were used and the purge and protective gas was argon time-consuming and have poor stability and repetitiveness due to the
99.998 % (Air Liquide, Montevideo, Uruguay). The graphite furnace volatility of CS2 (López-Fernández et al., 2012). The QuEChERS method
heating programs used for the analytical determinations and the is rarely used in this case. However, Chen et al. (2011) has applied
experimental conditions are reported in Table 1. Injection volumes were dimethyl sulfate as a derivation reagent and successfully extracted
of 30 μL. The chemical matrix modifier used for Cd and Pb mancozeb using the QuEChERS protocol. Despite of the solubility

3
I. Machado and I. Dol Journal of Food Composition and Analysis 103 (2021) 104090

problem of mancozeb, it can be readily transformed into a soluble salt in matrix effects were quantified. Percentage matrix effect (ME%) for each
water by the addition of an alkaline EDTA solution. Then, mancozeb can compound was calculated as: ME% = [(matrix matched calibration
be transformed into mancozeb-dimethyl from its sodium salt by curve slope – solvent calibration curve slope)/ solvent calibration curve
S-methylation with methyl iodide through the ion-pair alkylation. slope] × 100. Matrix matched calibration was performed using stan­
However, a higher methyl derivative yield has been reported with the dards prepared from extracts of the same blank matrix. The analyzed
use of dimethyl sulphate (Hayama and Takada, 2008). Based on this, pesticides in QuEChERS extracts presented strong matrix effect, being
dimethyl sulphate was selected as methylation reagent in this work. greater than 50 % (Machado et al., 2017).
Linearity was evaluated, considering peak areas by constructing six- The strong matrix effects observed (>50 %) may be due to the
point calibration curves (each level in triplicate) with a wide concen­ presence of phenolic compounds from the matrix, represented by fla­
tration range (0.5, 1.0, 2.0, 3.0, 4.0 and 5.0 mg kg− 1). Good linearity vonoids (mainly anthocyanins, flavonols and flavanols), followed by
was observed for all compounds at the concentrations within the tested hydrolysable tannins (ellagitannins and gallotannins) and phenolic acids
interval, up to 5 mg kg− 1, with determination coefficients (R2) greater (hydroxybenzoic acids and hydroxycinnamic acids) (Giampieri et al.,
than 0.99. Individual residuals were studied for each compound, and 2012). The findings of strong matrix effects are also supported by several
deviations from the calibration curve in the relevant region proved to be researchers (Prodhan et al., 2015, 2016). The quantifications were al­
less than ± 20 % in all cases, as stated by DG-SANTE Guidelines. For this ways performed using matrix-matched calibration curves as recom­
task, the deviation of the back-calculated concentrations of the cali­ mended by the guidelines. The high matrix effects can lead to false
bration standards from the true concentrations using the calibration positives or negatives findings when the residue levels are closer to the
curve in the relevant region was evaluated. Results are shown in Table 2. MRL. In such cases, the guidelines suggest confirming the presence of
Recoveries were performed at three concentration levels: 0.01, 0.02 the pesticide by standard additions (European Commission DG-SANTE,
and 0.05 mg kg− 1. Five replicates were analyzed for each spiking level. 2019).
Recovery values obtained at the highest concentration levels assayed, Uncertainty was estimated using the approach based on intra labo­
were in the range 70–120 %, as stated by DG-SANTE Guidelines. Pre­ ratory validation data as stated by DG-SANTE Guidelines. The estimated
cision (repeatability) expressed as RSDr was below 20 % in all cases. expanded measurement uncertainty, calculated at the 0.02 mg kg− 1
Regarding the low recoveries obtained at 0.01 mg kg− 1 for chlor­ spiking level, varied between 27.6 and 41.0 % (assuming a coverage
othalonil, cyprodinil, fludioxonil, and metalaxyl-M, it could be factor of 2) as shown in Table 2, not exceeding the 50 % default value.
explained by the fact that strawberry is a complex matrix that contains Thus, regulatory authorities may use the 50 % default value for
both inorganic and organic components characterized by diverse enforcement decisions.
chemical structures and physicochemical properties. Therefore, pesti­ The good results obtained with QuEChERS method are probably due
cide residues analysis at low concentration levels can be a difficult and to its relatively high selectivity that can be associated to the properties of
challenging task, mainly due to interfering substances that are co- acetonitrile, as well as the sorbents applied in the clean-up step.
extracted with the analytes and can adversely affect recoveries. Also, Acetonitrile co-extracts relatively small amounts of non-polar sub­
the adsorption of this sort of fungicides at low concentrations onto stances such as waxes and fatty acids that can be removed by PSA in the
different sample matrixes has been previously described by other au­ clean-up step (Machado et al., 2017). Strawberries are known to be a
thors (Fernandes et al., 2003; Hou et al., 2016). source of essential unsaturated fatty acids that come from the seeds’ oil.
Estimation of LOD values was performed according to the Eurachem Additionally, the good performance of PSA was also related to its ca­
Guide as 3 s/b (being “s” the standard deviation of 10 replicates of a low pacity to remove sugars, such as sucrose, glucose and fructose, well
concentrations standard and “b” the slope of the calibration curve) and known components of the matrix. Besides, more than 25 different
expressed in the sample (Eurachem Guide, 2014). Analytes presented anthocyanin pigments have been described in strawberries from
LOQ values between 0.01 mg kg− 1 and 0.02 mg kg− 1, which were far different varieties. These compounds are the best-known polyphenolic
below the corresponding MRLs established by the European Union for compounds and quantitatively the most important in strawberries,
strawberries (European Union Pesticides Database, 2021), thus agreeing ranging from 150 to 800 mg kg− 1 of fresh weight (Giampieri et al.,
with DG-SANTE Guidelines. The LOQ values were also below the cor­ 2012). Anthocyanins are derived from anthocyanidins by adding sugars,
responding MRLs established by the Codex Alimentarius, shown in and so can be successfully removed by PSA.
Table 2, which correspond to the Uruguayan national regulation (Codex In this context, Fig. 1 was included as an example, showing relatively
Alimentarius, 2021). clean extracts, with no signals neighboring the studied analytes (1b),
Solvent and matrix matched calibration curves were compared, and demonstrating once again the effectiveness of the applied QuEChERS

Table 2
Figures of merit obtained for the determination of pesticide residues.
Parameter Chlorothalonil Cyprodinil Fludioxonil Mancozeb Metalaxyl-M ETU

Linearity (mg kg− 1) 0.020 to 5.0 0.020 to 5.0 0.020 to 5.0 0.010 to 5.0 0.020 to 5.0 0.010 to 5.0
Residuals (%) 7.0/4.4/− 1.2/− 7.8/ 6.2/4.6/− 2.1/ − 8.5/ − 2.2/6.7/ − 9.9/− 3.0/ − 3.0/2.9/3.2–5.5/ − 3.2/8.9/9.5/ 6.0/ − 3.9/3.7/3.4–6.2/
9.4 9.7 3.1 6.1 − 10.1 7.9
LOD (mg kg− 1) (fresh 0.011 0.009 0.010 0.005 0.009 0.004
weight)
LOQ (mg kg− 1) (fresh 0.020 0.020 0.020 0.010 0.020 0.010
weight)
a b c a b c a b c a b c a b c
Precision, RSD (%) 16.5 /10.8 / 10.3 15.4 /12.4 /10.6 10.3 /9.9 / 8.7 9.8 /9.2 / 8.3 12.5 /11.9 / 10.5 9.5 /8.3b/ 8.0c
a

Trueness, Recovery (%) 62.3a/90.2b/ 95,4c 59.4a/87.8b/92.2c 54.8a/85.4b/ 99.2c 81.3a/97.5b/97.9c 63.9a/92.3b/ 98.7c 86.0a/98.1b/100.3c
Matrix effect (%) 57.0 51.3 55.9 61.6 122.7 64.9
Uncertainty (%) 36.4 41,0 36.7 28.9 35.7 27.6
MRL (mg kg− 1) 4.0d/5.0e 5.0d/10.0e 4.0d/3.0e 5.0d/5.0e 0.5d/0.2e –
a
Level 0.01 mg kg− 1.
b
Level 0.02 mg kg− 1.
c
Level 0.05 mg kg− 1.
d
MRLs from European Commission.
e
MRLs from Codex Alimentarius. Five replicates were analyzed for each spiking level.

4
I. Machado and I. Dol Journal of Food Composition and Analysis 103 (2021) 104090

◦
C at 200 ◦ C intervals. Once the optimum temperature ranges were
found, the fine adjustment was made, obtaining the conditions described
in Table 1. The pyrolysis and atomization curves, between the aqueous
standard solutions and the digested samples, showed no significant
differences in the maximum applicable pyrolysis temperatures without
analyte losses. The lowest temperatures for quantitative atomization
were practically identical as well, suggesting no matrix effects. The
highest pyrolysis temperatures without analyte loss were finally chosen
to remove as much as possible of the residual matrix.
The microwave-assisted method was validated for Cd and Pb de­
terminations based on Eurachem Guide (Eurachem Guide, 2014). For
trueness evaluation, a Student’s t-test was performed to compare the
obtained values with the certified values of the CRM. All experimental
t-values were below the theoretical t (0.05, 5) 2.57 indicating that, at the
95 % confidence level, the concentrations did not differ significantly
from the certified values (Miller and Miller, 2010). Recoveries were in
the range 97–102%. Repeatability expressed as RSD (%) for the analysis
of CRM (n = 6) was better than 10 % for all the studied elements. So, the
accuracy of the method was ensured. Obtained LODs and LOQs as well
as linearity ranges for each element, are summarized in Table 3. Line­
arity was evaluated considering peak areas by constructing five-point
calibration curves (each level in triplicate). Good linearity was
observed for both elements at the concentrations within the tested in­
terval, up to 4 mg kg− 1 (Cd) and 50 mg kg− 1 (Pb), respectively, with
determination coefficients (R2) greater than 0.99. Individual residuals
were studied for each analyte, using back-calculated concentrations.
Results are shown in Table 3.
Fig. 1. Chromatograms obtained for mixed solutions containing: (1) Metalaxyl-
M, (2) Fludioxonil, (3) Chlorothalonil and (4) Cyprodinil. Letter (a) corresponds
to the mix in solvent at 1 mg kg− 1, while (b) corresponds to the mix of recovery 3.3. Application to commercial strawberry samples
test at 0.05 mg kg− 1 using strawberry matrix blank.
After the validation process, 10 strawberry samples from different
method for this mix of fungicides. markets of Uruguay were tested for pesticide residues. Analytical de­
terminations of pesticides were performed using matrix matched cali­
3.2. Method selection and validation: inorganic contaminants bration and finally expressed as mg of analyte per kg of sample (fresh
weight). The obtained concentrations for all the analyzed pesticides
Since Cd and Pb levels were expected to be low, ETAAS seemed to be were below the corresponding MRL. Even though mancozeb was
an adequate technique for its determination. The optimization was detected in 4 samples, their values were below the corresponding LOQ
carried out using the strawberry leaves CRM. For this technique, the (Table 4).
need of chemical modifiers is well established, to decrease volatilization Mancozeb is widely used in agriculture especially in combination
and gas-phase interferences. These reagents allow the thermal stabili­ with modern systemic fungicides to control resistances and to expand
zation of volatile analytes, so that higher pyrolysis temperatures can be the spectrum of activity. Therefore, it is one of the most frequently
applied, as well as the conversion of matrix interferences into more detected pesticides (López-Fernández et al., 2012). Since it is not sys­
volatile forms (Welz and Sperling, 1999). Different modifiers have been temic, the residues of which from foliar application always stay on the
described for Cd and Pb determinations in the literature (Machado et al., surface of plants. Accordingly, ETU is also present on the surface of the
2016; Shaltout et al., 2013; Tissot and Machado, 2020). crop (FAO, 1993). Mancozeb can be partially eliminated with a vigorous
The greatest difficulty for determining Cd by ETAAS has been asso­ washing (up to 30 and 50 % in some fruits) but the remainder can be
ciated to its high volatility. Ammonium salts, such as ammonium transformed into ETU if it is subjected, for example, to a heating process,
phosphate, have been used as modifiers to thermally stabilize Cd up to as in the production of jam (FAO, 2021).
about 900 ◦ C. However, for the determination of trace Cd concentrations The determination of ETU residue in strawberries was successfully
in the presence of high chloride concentrations, spectral interferences achieved by the modified version of QuEChERS-HPLC-DAD method.
may appear. The Pd-Mg(NO3)2 modifier has proven to be especially
good in these cases, being the maximum pyrolysis temperature about Table 3
800 ◦ C (Welz and Sperling, 1999). When it comes to Pb determinations, Figures of merit obtained for the determination of inorganic contaminants.
the Pd-Mg(NO3)2 modifier brings excellent results. It allows pyrolysis Parameter Cd Pb
temperature of 1100− 1400 ◦ C, which enables the separation of most − 1
Linearity (mg kg ) 0.0027 to 4.0 0.0050 to 50.0
interfering concomitants. The stabilizing effect of this modifier also Residuals (%) 1.1/− 0.5/− 0.9/ 1.9/− 0.8/− 1.3/
causes a relatively high atomization temperature of about 2000 ◦ C. The 1.7 2.2
ammonium phosphate modifier, that is also frequently used, allows a LOD (mg kg− 1) (3 s, n = 10) (fresh 0.0008 0.0015
weight)
lower atomization temperature of about 1600 ◦ C, however, the lower
LOQ (mg kg− 1) (10 s, n = 10) (fresh 0.0027 0.0050
maximum pyrolysis temperature may interfere with the separation of weight)
concomitants. Precision, RSD (%) 3.5a 4.4a
Herein, the mixture Pd-Mg(NO3)2 was evaluated for all analytes, Trueness, Recovery (%) 99.0a 102.0a
using 5 μg of Pd and 3 μg of Mg(NO3)2. The optimization of pyrolysis and ML (mg kg− 1) 0.05b 0.2b

atomization temperatures was carried out exhaustively, by performing a

CRM LGC 7162.
the corresponding pyrolysis - atomization curves, from 600 ◦ C to 3000 b
ML from national regulation.

5
I. Machado and I. Dol Journal of Food Composition and Analysis 103 (2021) 104090

Table 4
1
Levels of contaminants found in the analyzed strawberry samples expressed in mg kg− (fresh wight).
Sample Chlorothalonil Cyprodinil Fludioxonil Mancozeb Metalaxyl-M ETU Cd Pb

1 ND ND ND ND ND ND 0.0031 0.0064
2 ND ND ND ND ND ND 0.0077 0.0099
3 ND ND ND < LOQ ND < LOQ 0.0088 0.0212
4 ND ND ND ND ND < LOQ 0.0069 0.0103
5 ND ND ND ND ND ND 0.0039 0.0057
6 ND ND ND < LOQ ND < LOQ 0.0081 0.0202
7 ND ND ND ND ND ND 0.0042 0.0087
8 ND ND ND ND ND ND 0.0044 0.0075
9 ND ND ND < LOQ ND < LOQ 0.0056 0.0154
10 ND ND ND < LOQ ND < LOQ 0.0063 0.0178
1
ND: Non detected (<LOD), LOQ: 0.010 mg kg− (for Mancozeb and ETU).

Compared to previous methods, this method turned out to be simple, of pesticide residues in strawberries. The methods were good in terms of
rapid, and sensitive with the advantage of a relatively low consumption figures of merit, allowing the analysis of fungicides from different
of organic solvents. Setting a MRL for ETU was considered not beneficial families by HPLC-DAD, providing good selectivity, trueness, and preci­
to ensure that Good Agriculture Practices (GAP) were being followed in sion. The LOQs obtained were lower than the corresponding MRLs
the use of EBDCs, so the Joint Meeting on Pesticide Residues (JMPR) established by the European Union and the Codex Alimentarius for each
agreed to recommend the withdrawal of all MRLs for ETU in 1993 (Zhou compound. Mancozeb and ETU were detected in some of the analyzed
et al., 2013). However, the LOD (0.004 mg kg− 1) and LOQ (0.010 mg samples, but below the corresponding LOQs. This information is of
kg− 1) values obtained in this work, are considered suitable for fast ETU utmost importance at the national scale, since only few reports about
monitoring. this topic are readily available.
As stated before, EBDCs are generally unstable in the presence of On the other hand, a simple method for the extraction and subse­
moisture, oxygen and in biological systems (Garcinuno et al., 2004), quent determination of Cd and Pb was also validated, proving to be
being ETU its main degradation product by hydrolysis. This compound is accurate for this application. Commercial samples showed quantifiable
a possible human carcinogen so its presence at any level is undesirable, levels for both elements, but below the corresponding MLs established
thus, monitoring studies are mandatory. In this regard, 5 samples by national and international regulations.
showed detectable levels but below the corresponding LOQ, which In general terms, the validated methods turned out to be appropriate
agrees with previous results from Brazilian samples (Rossi Lemes et al., for the surveillance of these contaminants, providing a holistic approach
2014). for the monitoring of both organic and inorganic residues.
Once the validation for inorganic contaminants was completed, the
same commercial samples were analyzed. Analytical determinations Author statement
were performed using external calibration and finally expressed as mg of
analyte per kg of sample (fresh weight). The average values for Cd were Ignacio Machado: Conceptualization; Data curation; Formal analysis;
in the range 0.0031 ± 0.0002 to 0.0088 ± 0.0005 mg kg− 1 (fresh basis), Investigation; Methodology; Resources; Software; Supervision; Valida­
while average values for Pb were in the range 0.0057 ± 0.0004 to tion; Visualization; Roles/Writing - original draft; Writing - review &
0.0212 ± 0.0012 mg kg− 1 (fresh basis) (Table 4). The obtained results editing.
comply with the requirements of MERCOSUR regulation related to these Isabel Dol: Conceptualization; Investigation; Methodology; Re­
inorganic contaminants in small fruits, being Cd levels below 0.05 mg sources; Software; Supervision; Validation; Visualization; Roles/Writing
kg− 1 and Pb levels below 0.20 mg kg− 1 (MERCOSUR, 2011). Further­ - original draft; Writing - review & editing.
more, Pb levels agree with the limit of 0.02 mg kg− 1 recommended by
Joint FAO/WHO Expert Committee on Food Additives (JECFA) for small
Declaration of Competing Interest
fruits (Codex Alimentarius Commission, 2011).
Vascular plants take up metals from their roots, transpiration
The authors declare that they have no known competing financial
through stomata on the leaf surface and deposition on the leaf surface.
interests or personal relationships that could have appeared to influence
The uptake, translocation, and accumulation of metallic species by
the work reported in this paper.
edible crops are extremely important processes, strongly related to
human health. The partitioning of inorganic elements in different parts
of the plant, due to binding with specific ligands, is a common strategy Acknowledgements
to prevent toxicity of the aboveground parts (Machado et al., 2016).
Pb is a potentially toxic element, that can enter the body by inges­ The authors would like to thank Agencia Nacional de Investigación e
tion, inhalation, or skin resorption. The continuous assimilation of small Innovación (ANII) and PEDECIBA-Química for its support.
quantities is the main cause for acute Pb poisoning. This element is
initially bound to the erythrocytes and only a small part is excreted References
through the urine, being most part accumulated in the bones. It is one of
Abd Al-Rahman, S.H., Almaz, M.M., Ahmed, N.S., 2012. Dissipation of fungicides,
the most widely distributed contaminants in the nature. It is a ubiquitous insecticides, and acaricide in tomato using HPLC-DAD and QuEChERS methodology.
element that is present in all areas of the Biosphere. Deposition on edible Food Anal. Methods 5, 564–570.
crops is its main route of entry in the food chain. Therefore, the evalu­ Agrios, G.N., 2005. Control of plant diseases. Plant Pathology, 5th ed. Elsevier Academic
Press, New York, NY, USA, pp. 293–353.
ation of these potentially toxic elements in fruits is of utmost importance
Anastassiades, M., Lehotay, S.J., Stajnbaher, D., Schenck, F.J., 2003. Fast and easy
from the toxicological point of view. multiresidue method employing acetonitrile extraction/partitioning and ‘dispersive
solid phase extraction’ for the determination of pesticide residues in produce.
J. AOAC Int. 86 (2), 412–431.
4. Conclusions
Biswas, S.K., Banerjee, K., Handa, S.K., 2003. Metabolic fate of mancozeb in tomato
(Lycopersicon esculentum). Toxicol. Environ. Chem. 85 (1–3), 33–38.
Three different QuEChERS protocols were validated for the analysis Chen, W.Y., Dong, F.S., Liu, X.G., Qin, D.M., Liao, X.L., Cheng, L., Wang, C.R., Zheng, Y.
Q., 2011. Simplified method for determination of mancozeb residues in apple using

6
I. Machado and I. Dol Journal of Food Composition and Analysis 103 (2021) 104090

ultra performance liquid chromatography-mass spectrometry. Chinese J. Anal. López-Fernández, O., Rial-Otero, R., González-Barreiro, C., Simal-Gándara, J., 2012.
Chem. 38 (4), 508–512. Surveillance of fungicidal dithiocarbamate residues in fruits and vegetables. Food
Codex Alimentarius, 2021. International Food Standards. Retrieved February 2021. http Chem. 134, 366–374.
://www.fao.org/fao-who-codexalimentarius/codex-texts/dbs/pestres/pesticides López-Fernández, O., Rial-Otero, R., Cid, A., Simal-Gándara, J., 2014. Combined
/en/. determination and confirmation of ethylenethiourea and propylenethiourea residues
Codex Alimentarius Commission, 2011. Joint FAO/WHO Food Standards Programme in fruits at low levels of detection. Food Chem. 145, 1002–1010.
Codex Committee on Contaminants in Foods. Retrieved February 2021 from. http:// Machado, I., Dol, I., Rodríguez-Arce, E., Cesio, M.V., Pistón, M., 2016. Comparison of
www.fao.org/tempref/codex/Meetings/CCCF/CCCF5/cf05_INF.pdf. different sample treatments for the determination of As, Cd, Cu, Ni, Pb and Zn in
Crnogorac, G., Schwack, W., 2009. Residue analysis of dithiocarbamates fungicides. globe artichoke (Cynara cardunculus L.sUbsp. Cardunculus). Microchem. J. 128,
Trends Analyt. Chem. 28 (1), 40–50. 128–133.
Dirección General de la Granja (DIGEGRA) (2021). Retrieved February 2021 from: https Machado, I., Gérez, N., Pistón, M., Heinzen, H., Cesio, M.V., 2017. Determination of
://www.gub.uy/ministerio-ganaderia-agricultura-pesca/granja. pesticide residues in globe artichoke leaves and fruits by GC–MS and LC–MS/MS
Eurachem Guide: the fitness for purpose of analytical methods, 2014. In: Magnusson, B., using the same QuEChERS procedure. Food Chem. 227, 227–236.
Örnemark, U. (Eds.), A Laboratory Guide to Method Validation and Related Topics, Machado, I., Faccio, R., Pistón, M., 2019. Characterization of the effects involved in
2nd ed. Retrieved February 2021 from. http://www.eurachem.org ultrasound-assisted extraction of trace elements from artichoke leaves and soybean
European Commission – DG-SANTE, 2019. Document SANTE/12682/2019. Analytical seeds. Ultrason. Sonochem. 59, 104752.
quality control and method validation procedures for pesticide residues analysis in MERCOSUR, 2011. Technical Regulation. Retrieved March 2021 from. http://www.
food and feed. Supersedes Guidance Document SANTE/11813/2017 From 01 puntofocal.gov.ar/doc/r_gmc_12-11.pdf.
January 2020. Miller, J.N., Miller, J.C., 2010. Statistics and Chemometrics for Analytical Chemistry, 6th
European Commission – European Union Pesticides Database (2021). Retrieved February ed. Prentice Hall, Pearson Education Limited, Harlow, England.
2021 from: https://ec.europa.eu/food/plant/pesticides/eu-pesticides-db_en. Ministerio de Ganadería, Agricultura y Pesa (MGAP) (2021). Retrieved February 2021
European Committee for Standardization (CEN), 2008. Standard method EN 15662: from: https://www.gub.uy/ministerio-ganaderia-agricultura-pesca/.
2008. Foods of Plant Origin – Determination of Pesticide Residues Using GC-MS and/ Prodhan, M.D.H., Papadakis, E.N., Papadopoulou-Mourkidou, E., 2015. Determination of
or LC-MS/MS Following Acetonitrile extraction/partitioning and Cleanup by multiple pesticide residues in eggplant with liquid chromatography-mass
Dispersive SPE – QuEChERS-method. spectrometry. Food Anal. Methods 8, 229–235.
Fernandes, M.C., Cox, L., Hermosín, M.C., Cornejo, J., 2003. Adsorption–desorption of Prodhan, M.D.H., Papadakis, E.N., Papadopoulou-Mourkidou, E., 2016. Analysis of
metalaxyl as affecting dissipation and leaching in soils: role of mineral and organic pesticide residues and their variability in cabbage using QuEChERS extraction in
components. Pest Manag. Sci. 59, 545–552. combination with LC-MS/MS. Food Anal. Methods 9, 3470–3478.
Food and Agriculture Organization of the United Nations (FAO), 1993. Pesticide Residue Reglinski, T., Spiers, T.M., Dick, M.A., Taylor, J.T., Gardner, J., 2009. Management of
in Food, Report of the Joint Meeting of the FAO Panel of Experts on Pesticide phytophthora root rot in radiata pine seedlings. Plant Pathol. 58, 723–730.
Residues in Food and the Environment and the WHO Expert Group on Pesticide Rizzetti, T.M., Kemmerich, M., Martins, M.L., Prestes, O.D., Adaime, M.B., Zanella, R.,
Residues. Geneva, 20–29 September. 2016. Optimization of a QuEChERS based method by means of central composite
Food and Agriculture Organization of the United Nations (FAO) (2021). Retrieved design for pesticide multiresidue determination in orange juice by UHPLC–MS/MS.
February 2021 from: http://www.fao.org/home/en/. Food Chem. 196, 25–33.
Garcinuno, R.M., Ramos, L., Fernandez-Hernando, P., Camara, C., 2004. Optimization of Rossi Lemes, V.R., Martins-Júnior, H.A., Vitorino Carvalho de Souza, S., Colacioppo, S.,
a matrix solid-phase dispersion method with subsequent clean-up for the 2014. Ethylenethiourea in fruits: optimization and in-house validation of a method
determination of ethylene bisdithiocarbamate residues in almond samples. by liquid chromatography tandem mass spectrometry, occurrence, and dietary
J. Chromatogr. A 1041, 35–41. exposure assessment. Food Control 42, 321–328.
Giampieri, F., Tulipani, S., Alvarez-Suarez, J.M., Quiles, J.L., Mezzetti, B., Battino, M., Shaltout, A.A., Abdel-Aal, M.S., Welz, B., Castilho, I.N.B., 2013. Determination of Cd, Cu,
2012. The strawberry: composition, nutritional quality, and impact on human Ni, and Pb in black tea from Saudi Arabia using graphite furnace atomic absorption
health. Nutrition 28 (1), 9–19. spectrometry after microwave-assisted acid digestion. Anal. Lett. 46, 2089–2100.
Hayama, T., Takada, M., 2008. Simple and rapid method for the determination of Tissot, F., Machado, I., 2020. Evaluation of essential and potentially toxic elements in
ethylenebisdithiocarbamate fungicides in fruits and vegetables using liquid popcorn: study of cooking effect and in vitro bioaccessibility. J. Food Meas. Charact.
chromatography with tandem mass spectrometry. Anal. Bioanal. Chem. 392 (5), 14 (4), 1842–1849.
969–976. Welz, B., Sperling, M., 1999. Atomic Absorption Spectrometry, 3rd ed. WILEY-VCII
Hou, F., Zhao, L., Liu, F., 2016. Determination of chlorothalonil residue in cabbage by a Verlag GmbH, Weinheim, Germany.
modified QuEChERS-based extraction and gas chromatography–mass spectrometry. Xu, G., Nie1, J., Li, H., Yan, Z., Cheng, Y., 2017. Determination of mancozeb residue in
Food Anal. Methods 9, 656–663. fruit by derivatization and a modified QuEChERS method using ultraperformance
Instituto Nacional de Investigación Agropecuaria (INIA) (2021). Retrieved February liquid chromatography–tandem mass spectrometry. Anal. Bioanal. Chem. 409,
2021 from: http://www.inia.uy. 5057–5063.
International Programme on Chemical Safety (IPCS - INCHEM) (2021). Retrieved Zhou, L., Liu, X., Kang, S., Zhang, F., Pan, C., 2013. A rapid determination method for
February 2021 from: http://www.inchem.org/. ethylenethiourea in potato and cucumber by modified QuEChERS – high
Lehotay, S.J., Mastovská, K., Lightfield, A.R., 2005. Use of buffering and other means to performance liquid chromatography–tandem mass spectrometry. Food Chem. 138,
improve results of problematic pesticides in a fast and easy method for residue 1355–1359.
analysis of fruits and vegetables. J. AOAC Int. 88 (2), 615–629.