Quick monitoring of routine samples of foods in the regulatory analysis of pesticide residues

doi:10.21203/rs.3.rs-5217790/v1

Download PDF

Research Article

Quick monitoring of routine samples of foods in the regulatory analysis of pesticide residues

https://doi.org/10.21203/rs.3.rs-5217790/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 14 Jun, 2025

Read the published version in Archives of Environmental Contamination and Toxicology →

You are reading this latest preprint version

In this study, a rapid visualization method was developed to simultaneously evaluate the on-going performance of routine analysis and ensure that the concentration of multiple pesticides in food samples with high water content complies with Maximum Residue Limits (MRLs). In order to accomplish this, we used a bubble chart known as Fast Risk Estimation and Analysis (FREA). In this chart, each pesticide is represented by a bubble. By looking at its color, position on the graph, and the size of the bubble, you can quickly determine whether it meets the requirements for relative standard deviation (RSD), recovery, Index of Quality for Residues, and matrix effect. Onion and pepper were chosen as commodity group with high water content. A single matrix-matched calibration using pepper was performed to analyse all these products. The risk visualisation allows simultaneous checking of on-going validation and quality control parameters as recovery between 60–140% of the spike samples analysed at the same time, such as the historical RSD (value and alarm if is upper at 20%). At the same time, the bubble chart monitoring other parameters such as the exceeding of the MRL in the analyzed samples, the complex definition of the residue in different pesticides or inconsistencies such as limits of quantification higher than the MRL could be quickly identified.

Control Chart

On-going validation

Pesticides

Quality control

Pesticides are extensively utilized to safeguard livestock and crops in the agricultural sector. For example, in 2021, approximately 333,000 tonnes of pesticides were used, along with the introduction of over 400 new pesticides to the market (Eurostat, 2023).

When these pesticides are used on crops, it is important to ensure that only a minimal amount of them ends up in the food supply. To address this, Maximum Residue Limits (MRLs) have been established by regulations worldwide to prevent health issues linked to the overuse of pesticides (Veiga-del-Baño, Cuenca-Martínez, Andreo-Martínez, Cámara, Oliva, & Motas, 2023). In the European Union (EU), for instance, the European Commission's (EC) consolidated version of Regulation 396/2005 sets MRLs for a range of foodstuffs (EC, 2005; Kuchheuser & Birringer, 2022).

The development of appropriate methods for assessing food safety in accordance with the requirements of international quality standards is one of the main objectives in pesticide analysis (SANTE, 2021). To determine as many pesticides as possible in the most cost-effective way with the least effort, Multi-Residue Methods are needed. For this reason, pesticides are determined by gas chromatography (GC) and liquid chromatography (LC) coupled with mass spectrometry (MS), and the most widely used pesticide residue extraction method in routine laboratories today is the method with the acronym "Quick, Easy, Cheap, Effective, Rugged, and Safe" (QuEChERS). The QuEChERS method involves two steps, the first being an initial extraction with different salt formulations to drive the separation of the organic extraction solvent and water. The second, where an aliquot of the organic phase goes through a clean-up process in dispersive solid phase extraction (d-SPE) using different sorbents to remove the matrix component prior to injection by chromatographic analysis (Anastassiades, Lehotay, & štajnbaher, 2002).

The extraction method in the QuEChERS has been subject to certain variations and modifications to cover a broad range of products, such as the addition of water (for spices, flour and other dry matrices) or the addition of a buffer method when analysing pH-sensitive pesticides as suggested by EN 15662 (ECS, 2008).

The cleanup approaches are aimed to reducing the matrix effect (ME) in the variability of foods that can be analysed, such as foods with chlorophyll and other natural pigments (e.g. spices), fat or lipid content (e.g. nuts), essential oils and flavonoids (e.g. herbs), etc. (Rutkowska, Łozowicka, & Kaczyński, 2019). These ME lead to analytical problems that affect the accuracy of the results and can be seen in the recovery of pesticides by reducing or increasing the acceptable value for the purpose of validation from 70–120% (Damale et al., 2023; SANTE, 2021).

One of the most widely approaches to reduce ME methods is the matrix-matched calibration (Cuadros-Rodrı́guez et al., 2003). Some of its advantages include the ability to utilize a wide range of matrices (Damale et al., 2023; Fu, Zhang, Qin, Dou, Luo, & Yang, 2022; Kardani et al., 2023; Rutkowska et al., 2019; UNE-CEN/TS-17061, 2019; Zhao et al., 2021). Annex A of the SANTE 11312/2021 (SANTE, 2021) document, which describes commodity groups such as high-water content (e.g. vegetables), high acid and high water content (e.g. citrus fruits), high starch and/or protein and low water and fat content (e.g. cereals) or unique commodities such as tea, spices or coffee, can be used to extrapolate some recommendations for the preparation of matrix calibrations.

For a laboratory analysing a wide range of products and a large number of samples within the same product group, it is most effective to use a single matrix calibration for different matrices within the same product, which may result in reduced accuracy for some pesticides. In this sense, the design of an analytical Quality Assurance (AQA) system is very important, and the most efficient system for AQA is on-going validation or performance verification with the spike matrix to demonstrate applicability to other commodities in the same commodity group with the same matrix calibration.

Any commercially available chromatography instrument has both acquisition and processing software to analyze AQA requirements, including retention times, sample recovery, blank, calibration, or MS/MS identification. However, these softwares do not allow an assessment of whether the recovery achieved is within the range of the average recovery and the relative estandar desviation (RSD) from on-going recovery results [within laboratory reproduciblity (RSD_wR)]. Another major drawback when working with samples subject to MRLs, which may change over time, is knowing whether the result obtained for a particular pesticide exceeds the MRL, because although this is not defined as a specific AQA criterion, it is a criterion that affects the quality of the product and other criteria such as those defined in section D15 (SANTE, 2021) relating to confirmation of higher values exceeding the MRL.

The great development of information technology in recent years has made it possible to provide commercial AQA software that allows, externally to the instrument software, many statistical and evaluation tools that allow easy and intuitive evaluation of the range of mean recovery and RSD_wR through, for example, Shewhart's charts, which are commonly used for quick graphical visualisation of historical data (Agüera, López, Fernández-Alba, Contreras, Crespo, & Piedra, 2004; ISO-7870-2, 2023). However, this tool and software also have some disadvantages when used in a multiresidue method to analyze pesticides. For instance, there is a different interpretation of the classic Shewhart's chart when used in analytical chemistry, as shown in ISO/TS 13530 (ISO/TS-13530, 2009). Two plots are required, one for recovery and the other for precision for each pesticide, and it can be complex to study a multiresidue method with many pesticides (between GC and LC). In addition, other intrinsic characteristics of pesticide analysis, such as the MRLs values or the risk associated with the results of the pesticide versus MRLs, are not considered in this type of chart (Caldas, 2023; IEC, 31010; Li et al., 2023; Su et al., 2024).

Therefore, the aim of this study is to create and to evaluate an alternative graphical tool to Shewart´s charts, called Fast Risk Estimation and Analysis (FREA), for multiresidue analysis of pesticides by chromatography and mass detector in a routine laboratory.

The purpose of this graph is to simultaneously display information on the recovery, the RSD_WR of the method, and ME for each pesticide and sample analyzed. This will provide insight into the on-going method performance verification and whether the product analyzed exceeds the MRL in the samples through the Index of Quality for Residues (IqR) parameter. This type of graph would enable a quick and visual assessment of the pesticides that have been analyzed with values above the limit of quantification (LOQ), as well as their associated risk. This would be reported in terms of AQA and also by their value in relation to the MRL.

2.1. Sample selection

The commodity group used was High Water Content (G.HW) because of the wide variability of the different products it contains and the diversity of pesticides and MRLs. The products analyzed were onions and tomatoes, which are shown in Table 1.

A single matrix-matched calibration using pepper was performed to analyse all these products. Pepper was used because it was previously validated according to the SANTE 11312/2021 (SANTE, 2021). The pepper matrix was chosen based on internal validation studies previously carried out, and because it was considered very different from the other matrices to be analyzed (tomato and onion), so that the possible ME could be evaluated (SANTE, 2021).

Table 1 also shows the samples used as blank and spike, and the samples used for analyzed as blind-incurred. The blank samples were supplied by an accredited laboratory for pesticide analysis in food (ISO/IEC-17025, 2017) located in the Region of Murcia (Spain), which are usually used by the laboratory as matrix calibration in its routine analyses.

The blind-incurred samples were chosen based in the different classification in the EURL MRL database (EU, 2023b) within the commodity group G.HW. These samples were obtained from local companies prior to sale in supermarkets. The subsampling and preparation of each G.HW sample were carried out according to R.D. 290/2003 (RD, 2003). The homogeneous products obtained were kept in the freezer at -20°C until extraction and analysis.

Table 1

Samples analyzed by commodity group.
Blank-spike samples G.HW	Blind samples G.HW	Commodity EURL database (Blind samples)	Product EURL database (Blind samples)
Onions	Onions	0220000: Bulb vegetables	0220020: Onions
Tomatoes	Tomatoes	0230000: Fruiting vegetables	0231010: Tomatoes

G.HW: High water content group; Commodity: Commodity group and code assigned in EURL database; Product: code assigned to the product inside the commodity EURL database.

All the samples described in Table 1 were analyzed only once like any routine laboratory.

2.2. Chemical and reagents

All pesticides had certified reference standards of at least 95% purity and were purchased from LGC Standards (Teddington, UK).

Gas chromatography and liquid cromatography solvents [(ethyl acetate, methanol, water (resistivity > 18 MΩ) and residual formic acid] were supplied by J.T. Baker (Center Valley, PA, USA). The ammonium formate, formic acid and acetonitrile were supplied by Merck (Darmstadt, Germany). Sodium chloride (1 g), disodium hydrogen citrate sesquihydrate (0.5 g), trisodium citrate dihydrate (1 g), sodium sulfate (4 g), primary secondary amine (PSA) for dispersive solid phase extraction (dSPE) were supplied in an extraction kit by Agilent Technologies (Santa Clara, USA). All reactives were supplied with analytical quality.

2.3. Preparation of solvent, matrix calibration and spiked samples

The stock standard solution was prepared for individual pesticides at a concentration of about 1000 µg/ml. The solvent used for LC was acetone and for GC n-hexane/acetone (9:1, v/v).

The standard working matrix calibration of multiple compounds was performed by serially dissolving the appropriate amounts of each stock solution. The matrix calibration used for G.HW was a blank pepper. The concentrations of the matrix calibration were in the range of 2 to 50 ng/ml. The fortified samples for the ME study were a white matrix of onion and another of tomato.

2.4. Pesticide analysis

The QuEChERS extraction, based on EN 15662 (ISO, 2019), used 10 g into a 50 ml polypropylene tube without water addition. After, 10 ml of acetonitrile was added [E1 EN 15662 extraction (ISO, 2019)] and the mixture was shaken for 1 min. After shaking, the salts extraction kit was added, and it was centrifuged for 5 min at 3000 rpm. A clean-up process using PSA [based in C2 EN 15662 clean-up (ISO, 2019)] on the organic aliquot obtained and finally it was centrifuged again at 3000 rpm for 5 min to get the final aliquot for GC-MS/MS and LC-MS/MS analysis.

A total of 23 pesticides were analyzed by GC-MS/MS using an Agilent (Santa Clara, USA) 7890A GC. A Gas Chromatography system coupled with a 7000A quadrupole tandem mass spectrometer. Chromatographic separation was performed on a RTx-5MS column (30 m, 0.25 mm, i.d., 0.5 µm) RESTEK. Helium was used as the carrier gas at a constant flow rate of 1.2 mL/min, while argon was used as the collision gas. The oven temperature was adjusted as follows: the initial temperature was set at 70°C for 2 min, then increased to 150°C at a rate of 25°C min^− 1, to 200°C at 3°C min^− 1 with a hold time of 1 min, and finally to 280°C at 10°C min^− 1 with a hold time of 10 min. The total run time was 45 min. The temperatures of the transfer line and ion source were set at 280 ºC and 250 ºC, respectively. The mass spectrometer was operated in multiple reaction monitoring (MRM) mode with three mass transitions.

A total of 18 pesticides were analyzed by LC-MS/MS using an Agilent liquid chromatography system (Santa Clara, USA) coupled with a 6470 triple quadrupole tandem mass spectrometer. The chromatographic separation was performed on a Poroshell C18 column (150 mm, 2.1 mm i.d., 2.7 µm) (Agilent, USA) with a flow rate of 0.1 ml at 40°C. The elution solvent used was water 5 mM ammonium formate with 0.01% formic acid (A) and methanol 5 mM ammonium formate with 0.01% formic acid (B). The gradient elution was carried out as follows: 40% solvent B for 0–5 min, changing to 60% solvent B for 6–12 min and finishing with 100% solvent B for 17–20 min. Pesticides were analyzed using programmed MRM in positive and negative modes simultaneously. Ion source parameters include optimized drying gas temperature, drying gas flow, nebulizer pressure, sheath gas temperature and flow, capillary voltage, nozzle voltage and high and low radio frequency voltage.

Limit of quantification (LOQ), validation information, MRM transitions and the retention times for tested pesticides can be found in Supplementary Table S1.

The number of pesticides studied using each technique, and the technique used, were chosen based on the history of positive results for the matrices analyzed in this study.

2.5. Performance of the method and quality assurance

The method for pepper was previously validated according to the SANTE 11312/2021 (SANTE, 2021) requirements. The blind-incurred samples of G.HW were analyzed in the same sequence or batch according to the quality assurance section titled “on-going method performance verification during the routine analysis” from SANTE 11312/2021 (SANTE, 2021) guide. Therefore, together with the sequence of blind-incurred samples, the blank samples spiked at a LOQ of 0.005 mg/kg validated with all pesticides were analyzed.

2.6. Database creation and data analysis

The Structured Query Language (SQL) database is one of the most widely used because it is a free tool and easy to implement in different programming languages (Strassemeyer, Daehmlow, Dominic, Lorenz, & Golla, 2017; Xia, Stinner, Brinkman, & Bennett, 2003). A database of MRLs in the EU was created by downloading information in Extensible Markup Language (XML) from the EU website (EU, 2023a).

In the same database, another table was created with information of the data with recoveries for each pesticide in the commodity group studied for automatic RSD_wR calculations. Both tables are relational tables linked by the pesticide identifier (ID) present in the XML file from EU web page and they are periodically updated automatically, allowing both the MRL and the definition of the pesticide to be updated.

The results of the samples analyzed (blind and spike) were exported through the Agilent MassHunter software (Anonymous, 2008) in an XML format to a processing computer code in Python (3.11.4) or PHP 7.0 language using open libraries for data analysis.

The Phyton and PHP code generate a hypertext markup language (HTML) with a graphical through the open code Google chart (Google, 2023b). The graphical option selected to generate a FREA chart was the bubble chart (Google, 2023a).

The Figures generated in HTML code that appear in this manuscript, as well as the source code used in PHP, together with the SQL databases, can be obtained on request from the authors and the author´s GitHub repository.

2.7. AQA data analysis

The quality control based on the check of the routine recovery (Rec) of the spiked samples was calculated using Eq. 1.

$\:\%Rec=\frac{Measured\:concentration}{Spiked\:concentration}\times\:100$ Eq. 1

Where measured concentration is the concentration for each blind-incurred sample and each pesticide. Spiked concentration is the theorical concentration spiked, 0.005 mg/Kg in this case.

The quality results, based on the reproducibility on-going method, are given by the Relative Standard Deviation (RSD_wR) of all the running verification samples (8 historical data) together the spiked samples analysed in the same time and batch, by the Eq. 2.

$\:\%RSD=\frac{Standard\:deviation}{Average\:}\times\:100$ Eq. 2

Where the standard deviation and mean are the results of calculating the standard deviation and mean for each historical routine recovery data for each pesticide together with the current recovery in the batch analysed.

The limits for both calculations are given in section C43 of the SANTE Guide (SANTE, 2021). For routine analysis, a practical default of 60–140% can be used for individual recoveries, but with a maximum RSD of 20%.

To assess the quality of the food in the samples analyzed, the index of quality for residues (IqR) could be used (Bibi, Rafique, Khalid, Samad, Ahad, & Mehboob, 2022; Mac Loughlin et al., 2018) as shown in the Eq. 3.

$\:IqR=\frac{PRC}{MRL\:}$ Eq. 3

Where PRC is the pesticide residue concentration (mg/kg) in the blind sample. The results thus obtained for IqR could be evaluated as good (0-0.6), adequate (0.6-1.0) and inadequate (> 1).

The combination of the three equations gives information about the analytical compliance. These three variables were combined in a bubble chart. Figure 1 shows a bubble chart with data of different pesticides (each black bubble), recovery values (axis x), and RSD_wR values (axis y). The diameter of the bubble indicates the value of IqR.

The validation of a method for the analysis of multiple residues of pesticides must take into account that the limit of quantification is appropriate to the MRL of the pesticide and that there are pesticides with complex definitions that include different compounds to be validated, such as carbofuran (not analyzed in this study), where the MRL is set for various compounds and metabolites expressed as the sum of carbofuran (including any carbofuran formed from carbosulfan, benfuracarb or furathiocarb, and 3-OH carbofuran expressed as carbofuran). However, these variables may change rapidly depending on the changes introduced by Regulation 396/2005 (EC, 2005), such as the introduced value of IqR (IqRm) to add additional evaluations about the analytical compliance in the pesticide analysis. The Eq. 4 shows the modifications to identify two different cases:

$\:{IqR}_{m}=\frac{{PRC}_{m}}{MRL\:}\times\:100\:\text{I}\text{q}\text{R}\text{m}=\left\{\begin{array}{c}Case\:1:\:\:MRL<Spiked\:concentration\\\:Case\:2:MRL\:complex\:definitions\\\:\end{array}\right.$ Eq. 4

Case 1

Specific pesticides in diverse products present MRLs lower than the LOQ, e.g. carbofuran in wine grapes have a MRL of 0.002 mg/kg. Also, the MRLs values may be modified in future changes to the Regulation 396/2005 (EC, 2005). Updating XML files in the database allows detect MRLs < spiked concentration and show a 150% result of IqR_m.

Case 2

Some of pesticide are expressed and quantified through a single component (e.g. oxyfluorfen), but other pesticides have a complex residue definition because the MRL is based in the sum of more of one component that is analyzed separately (e.g endosulfan MRL is based in the analysis of alpha endosulfan, beta endosulfan and endosulfan sulphate). In these cases, only to evaluate the risk with the MRL value, it is necessary to consider that each component must comply with the MRL or use the residue definition in a specific manner taking into account the molecular weight conversion factor (EC, 2015). In the case of a complex definition, IqRm use the MRL of the sum for each component.

All these cases were represented on different colored bubbles, as shown in the following scheme:

Bubble chart =

$$\:\left\{\begin{array}{c}Green:\:\:IqRm<60\%\:and\:RSD<20\:\%\:and\:recovery\:between\:60\%-140\%\:\\\:Yellow:\:IqRm\ge\:60\%\:and<100\%\:\:and\:\:RSD<20\:\%\:and\:recovery\:between\:60\%-140\%\\\:Red:IqRm>100\%\:or\:RSD>20\%\:or\:recovery<60\%\:or>140\%\\\:Grey:LMR<spike\:concentration\:or\:(ME>20\%\:\:and\:at\:least\:two\:other\:non-compliance)\end{array}\right.$$

The scheme shows the different criteria used to plot the bubble chart. Note that for grey bubble chart could be for a non-compliance for at least two non-compliance about the IqRm, RSD_wR or recovery.

The ME could be estimated by comparing the slopes of matrix-matched calibration curves with the solvent calibration curves (Damale et al., 2023; de Sousa, Guido Costa, de Queiroz, Teófilo, Neves, & de Pinho, 2012). However, this presents a big problem for the analytical effort (cost and time) involved for each product in the same commodity group. For these reasons, an alternative calculation (Eq. 5) to evaluate ME was proposed.

$\:ME=\frac{{Absolute(Rec}_{1}-{Rec}_{2})}{\:{(Rec}_{1}+{Rec}_{2})/2}\times\:100$ Eq. 5

The ME proposed calculates the relative percentage difference (RPD) (Anonymous, 2017) between two different recoveries of the two different spike samples in the routine batch analyzed. In this study, Rec1 is the recovery of the tomato spike sample and Rec 2 is the recovery obtained from the onion spike. A maximum value for RPD of 20% was fixed, by similarity, to the value of RSD_wR fixed in SANTE 11312/2021 (SANTE, 2021).

Table 2 shows the results obtained for the pesticides found in the blind samples analized of tomato and onions in a routine analysis by GC and LC. The concentrations are expressed with two significant figures according to SANTE 11312/2021 (SANTE, 2021) for all results over of the LOQ of 0.005 mg/kg. Table 2 also shows the MRL for both samples.

Table 2

Pesticide results and MRL values.
Pesticide	Analysis	mg/kg (Onion)	mg/kg (Tomato)	MRL (Onion)	MRL (Tomato)
Chlorpropham	GC	0.053	< 0.005	0.01	--
Fenpropathrin	GC	0.095	< 0.005	0.01	--
Fluopyram	LC	0.079	0.090	0.07	0.5
Mandipropamid	LC	0.0083	< 0.005	0.1	--
Spirotetramat (sum)	LC	0.078	0.072	0.4	1
Spirotetramat enol	LC	0.097	0.089	--	--
Pendimethalin	GC	0.071	< 0.005	0.05	--
Acetamiprid	LC	< 0.005	0.028	--	0.5
Azoxystrobine	LC	< 0.005	0.0082	--	3
Boscalid	LC	< 0.005	0.043	--	3
Chlorantraniliprole	LC	< 0.005	0.012	--	0.6
Cypermethrin	GC	< 0.005	0.185	--	0.5
Cyproconazole	GC	< 0.005	0.024	--	0.05
Cyprodinil	GC	< 0.005	0.107	--	1.5
Dimethomorph	GC	< 0.005	0.031	--	1
Fenhexamid	LC	< 0.005	0.049	--	2
Fludioxonil	GC	< 0.005	0.025	--	3
Iprodione	GC	< 0.005	0.036	--	0.01
Metaflumizone	LC	< 0.005	0.039	--	0.7
Pyraclostrobin	LC	< 0.005	0.013	--	0.3
Pyriproxyfen	GC	< 0.005	0.279	--	1
Spinosad	LC	< 0.005	0.045	--	0.7
Spirodiclofen	LC	< 0.005	0.026	--	0.5
Spiromesifen	LC	< 0.005	0.055	--	1
Thiacloprid	LC	< 0.005	0.049	--	0.5

Spirotetramat (sum): Spirotetramat and spirotetramat-enol (sum of), expressed as spirotetramat: Spinosad: Spinosad (spinosad, sum of spinosyn A and spinosyn D); Cypermethrin: Cypermethrin [cypermethrin including other mixtures of constituent isomers (sum of isomers)]; Conc: Concentration in mg kg^− 1; MRL: Maximum Residue Level from EURL database; --: MRL value not shown due to the current lack of MRL values in the EURL database, therefore the default value of 0.01 mg/kg can still be used.

The number of pesticides with values above the LOQ or reported as positive were 7 in the onion sample and 21 in the tomato sample. The number of positive pesticides in tomato and onions were similar in number to other pesticide studies in these matrices (Jirata, Asere, Balcha, & Gure, 2024; Ouakhssase & Ait Addi, 2023). The number of pesticides with values in above the LOQ in onions is much lower than in tomatoes, could be due to the inhibitory effect of onion organosulphur compounds (propylpropane thiosulphinate and propylpropane thiosulphonate) on many pests (Falcón-Piñeiro et al., 2023; Skovgaard, Encinas, Jensen, Andersen, Condarco, & Jørs, 2017) and mainly because they are bulbs (grown under the soil).

The three highest values for onion correspond to the pesticides fluopyram (0.079 mg/kg), fenpropathrin (0.095 mg/kg) and spirotetramat (0.097 mg/kg). In the case of spirotetramat, the value is obtained exclusively from the result of spirotetramat enol. As can be observed, all pesticides are below the MRL due to spirotetramat, even though it is a pesticide belonging to the derivatives of tetranic acid, it has a low persistence (Mandal, Joshi, Bansal, Sharma, & Kang, 2022). However, in the case of fluopyram it is above the MRL due to a greater persistence of the compound (Patel et al., 2016).

In the case of tomato, the three pesticides found were cyprodinil (0.279 mg/kg), cypermethrin (0.185 mg/kg) and pyriproxyfen (0.107 mg/kg). These pesticides are commonly used in tomatoes, but they have low persistence due to factors such as photodegradation (Lin, Gerrard, & Shaw, 2005). In the case of cypermethrin, the results are the sum of different chromatographic peaks automatically via software in GC analysis (Khazri et al., 2016).

Table 3 shows the data calculated of the spiked samples of onion and tomato (blank samples) such as % of recovery (%Rec), % RPD (as estimation of ME) described by Eq. 5, historical data for %RSD in the commodity group studied (G.HW), or the results for the IqRm for all the pesticides analyzed by GC and LC. Of note, pesticide concentrations in the blank matrix were not included in Table 3 because they were below the LOQ (< 0.005 mg/kg) in all cases.

Table 3

Calculations obtained for the generation of FREA chart
Pesticide	%Rec (Onion)	%Rec (Tomato)	%RPD (ME)	%RSD (G.HW)	IqR_m (Onion)	IqR_m (Tomato)
Chlorpropham	64	86	29	23	530	0
Fenpropathrin	106	70	41	17	950	0
Fluopyram	89	104	16	22	113	0
Mandipropamid	109	75	37	19	8	0
Spirotetramat (sum)	115	91	23	20	20	0
Spirotetramat enol	115	91	23	20	24	0
Pendimethalin	116	103	12	18	142	0
Acetamiprid	106	87	20	17	0	6
Azoxystrobine	102	85	18	13	0	0
Boscalid	77	85	10	15	0	1
Chlorantraniliprole	111	80	32	22	0	2
Cypermethrin	87	108	22	17	0	37
Cyproconazole	113	90	23	13	0	48
Cyprodinil	105	89	16	9	0	7
Dimethomorph	92	69	29	16	0	3
Fenhexamid	81	107	28	18	0	2
Fludioxonil	96	109	13	21	0	1
Iprodione	99	115	15	10	0	360
Metaflumizone	99	83	18	15	0	6
Pyraclostrobin	78	91	15	16	0	4
Pyriproxyfen	94	84	11	14	0	28
Spinosad	118	115	3	15	0	6
Spirodiclofen	88	67	27	19	0	5
Spiromesifen	71	119	51	13	0	6
Thiacloprid	86	62	32	24	0	10

As can be seen in Table 3, there are four pesticides in the onion with not adequate IqR_m, but only the chlorpropham and fluopyram present a non-compliance in the RSD, according to the requirements fixed in the Eq. 5. However, in the tomato sample, only iprodione presented a not adequate IqRm but all other requirements are fulfilled.

The ME effect, calculated with the RPD using the Eq. 5, shows that there are 12 (48%) pesticide with a non-compliance value upper of the LOQ. This could indicate that there is a ME to be considered when using pepper as matrix calibration in tomato or onion since the effects of signal suppression and co-extracted compounds can have a significant effect on the results (Gómez-Ramos, Rajski, Lozano, & Fernández-Alba, 2016; Wu & Ding, 2023). However, even taking into account the possible matrix effects, only the pesticides chlorpropham and fenpropathrin represent a high risk in onions due to the concentration of the pesticide with respect to the MRL that using Eq. 4 gives a very high value of the IqRm.

As mentioned earlier, the FREA chart can be generated in two different ways depending on the objective pursued. Figure 2 illustrates how a graph can display information related to pesticides with values above the limit of quantification in a specific sample. These values must be evaluated from the perspective of both the AQA data analysis and the associated risk of that value against the MRL of the pesticide in that sample.

In Fig. 2, it is evident that the grey bubbles, representing chlorpropham, pose a high risk due to non-compliance in RSD, ME, and a larger radius indicating a high IqRm value. On the other hand, the position of fluopyram, the red bubble in the top right-hand corner of Fig. 2, indicates a historical RSD > 20%, but it does not exhibit EM and has a > 100 value of the IqRm. However, its lower radius suggests that the quantified value of the pesticide compared to the MRL is lower in this pesticide than in chlorpropham.

The FREA chart also allows numerical information to be displayed by hovering over each of the bubbles, as shown for the pesticide chlorpropham.

An alternative to displaying positive pesticides as shown in Fig. 2 is to display pesticides with values below the LOQ as shown in Fig. 3. In this way, the AQA criteria can be evaluated only and simultaneously. In Fig. 3, those pesticides with adequate recovery values (60–140%) and RSD (< 20%) appear in green and in yellow when there is a non-compliance, as in the case of penconazole that does not comply with the RSD value by having a value of 21%.

The results of this study showed that the bubble chart, called FREA chart, allows a rapid visualization in a larger number of pesticides of the classical on-going parameters RSD_wR and recovery simultaneously.

Using the FREA chart, it was easy to see that only 4 pesticides in the onion exceeded the limit of quantification analyzed by GC-MS/MS and LC-MS/MS. These pesticides are chlorpropham (IqRm of 530), fenpropathrin (IqRm of 950), fluopyram (IqRm of 113), and pendimetalin (IqRm of 142). This poses a risk that needs to be reviewed as the values in the sample exceeded the MRL. Hovering over the graph, it was observed that chlorpropham had an RSD value of 23% and an ME of 29%, which violates the AQA requirements. For fenpropathrin, only the ME (41%) was found to be non-compliant. In the case of fluopyram, non-compliance was shown for an RSD value over 20% (22%), while pendimetalin did not have any non-compliance issues.

The FREA chart also showed its usefulness by indicating those pesticides that, although they were not positive in the samples, presented some type of non-compliance in the AQA as was the case with chlorpropham and penconazole with RSD values > 20%.

Using the information from the MRL and the periodic updates of the existing databases, the IqRm index allows graphical visualisation both to assess the quality of the sample analyzed and to detect changes in the MRL or the definition of the pesticide that may affect the LOQ. This graph also allows to estimate the ME by RPD calculation between two recoveries of different samples analyzed in the batch. However, it has the limitation that it is only evaluated in a single concentration and sample, and a validation in a specific product or alternative solutions could be necessary (Kwon, Lehotay, & Geis-Asteggiante, 2012).

This type of graph can be very helpful in making decisions about the analysis of pesticides, as well as determining the risk of a pesticide being present at a concentration above the LOQ, based on the concentration obtained and the MRL in place at the time for the specific type of sample being analyzed.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding

This research received no external funding.

Agüera, A., López, S., Fernández-Alba, A. R., Contreras, M., Crespo, J., & Piedra, L. (2004). One-year routine application of a new method based on liquid chromatography–tandem mass spectrometry to the analysis of 16 multiclass pesticides in vegetable samples. Journal of Chromatography A, 1045(1), 125-135. https://doi.org/https://doi.org/10.1016/j.chroma.2004.06.039.
Anastassiades, M., Lehotay, S. J., & štajnbaher, D. (2002). Quick, Easy, Cheap, Effective, Rugged, and Safe (QuEChERS) approach for the determination of pesticide residues. 4th European Pesticide Residues Workshop (EWPR). Book of Abstracts, p. 51. Rome.
Anonymous. (2008). Agilent Technologies. Materials Today, 11(7), 55. https://doi.org/https://doi.org/10.1016/S1369-7021(08)70153-6.
Anonymous. (2017). 1020 QUALITY ASSURANCE. In Standard Methods For the Examination of Water and Wastewater.
Bibi, A., Rafique, N., Khalid, S., Samad, A., Ahad, K., & Mehboob, F. (2022). Method optimization and validation for the routine analysis of multi-class pesticide residues in Kinnow Mandarin and fruit quality evaluation. Food Chemistry, 369, 130914. https://doi.org/https://doi.org/10.1016/j.foodchem.2021.130914.
Caldas, E. D. (2023). Approaches for cumulative dietary risk assessment of pesticides. Current Opinion in Food Science, 53, 101079. https://doi.org/https://doi.org/10.1016/j.cofs.2023.101079.
Cuadros-Rodrı́guez, L., Garcı́a-Campaña, A. M., Almansa-López, E., Egea-González, F. J., Lourdes Castro Cano, M., Garrido Frenich, A., & Martı́nez-Vidal, J. L. (2003). Correction function on biased results due to matrix effects: Application to the routine analysis of pesticide residues. Analytica Chimica Acta, 478(2), 281-301. https://doi.org/https://doi.org/10.1016/S0003-2670(02)01508-8.
Damale, R. D., Dutta, A., Shaikh, N., Pardeshi, A., Shinde, R., Babu, K. D., . . . Banerjee, K. (2023). Multiresidue analysis of pesticides in four different pomegranate cultivars: Investigating matrix effect variability by GC-MS/MS and LC-MS/MS. Food Chemistry, 407, 135179. https://doi.org/https://doi.org/10.1016/j.foodchem.2022.135179.
de Sousa, F. A., Guido Costa, A. I., de Queiroz, M. E. L. R., Teófilo, R. F., Neves, A. A., & de Pinho, G. P. (2012). Evaluation of matrix effect on the GC response of eleven pesticides by PCA. Food Chemistry, 135(1), 179-185. https://doi.org/https://doi.org/10.1016/j.foodchem.2012.04.063.
EC. (2005). Regulation (EC) No 396/2005 of the European Parliament and of the Council of 23 February 2005 on maximum residue levels of pesticides in or on food and feed of plant and animal origin and amending Council Directive 91/414/EEC Text with EEA relevance. Available online: https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=celex%3A32005R0396.
EC. (2015). SANCO/12574/2014 30/11-01/12 2015 rev. 5(1). Working document on the summing up of LOQs in case of complex residue definitions. Available on: https://food.ec.europa.eu/system/files/2016-10/pesticides_mrl_guidelines_wrkdoc_12574.pdf.
ECS. (2008). European Standard EN 15662. Foods of plant origin-Determination of pesticide residues using GC-MS and/or LC-MS/MS following acetonitrile extraction/partitioning and clean-up by dispersive SPE-QuEChERS-method. European Committee for Standardization. Available on: http://www.chromnet.net/Taiwan/QuEChERS_Dispersive_SPE/QuEChERS_%E6%AD%90%E7%9B%9F%E6%96%B9%E6%B3%95_EN156622008_E.pdf.
EU. (2023a). EU Pesticides Database (v3.1)-Download MRLs Data. Available on: https://ec.europa.eu/food/plant/pesticides/eu-pesticides-database/start/screen/mrls/download.
EU. (2023b). EU Pesticides Database (v3.1)-Search Pesticide Residues. Available on: https://ec.europa.eu/food/plant/pesticides/eu-pesticides-database/start/screen/mrls.
Eurostat. (2023). Pesticide sales. Available on: https://ec.europa.eu/eurostat/databrowser/view/aei_fm_salpest09/default/table?lang=en.
Falcón-Piñeiro, A., García-López, D., Gil-Martínez, L., de la Torre, J. M., Carmona-Yañez, M. D., Katalayi-Muleli, A., . . . Baños, A. (2023). PTS and PTSO, two organosulfur compounds from onion by-products as a novel solution for plant disease and pest management. Chemical and Biological Technologies in Agriculture, 10(1), 76. https://doi.org/https://doi.org/10.1186/s40538-023-00452-1.
Fu, Y., Zhang, J., Qin, J., Dou, X., Luo, J., & Yang, M. (2022). Representative matrices for use in matrix-matched calibration in gas chromatography-mass spectrometry for the analysis of pesticide residues in different types of food-medicine plants. Journal of Food Composition and Analysis, 111, 104617. https://doi.org/https://doi.org/10.1016/j.jfca.2022.104617.
Gómez-Ramos, M. d. M., Rajski, Ł., Lozano, A., & Fernández-Alba, A. R. (2016). The evaluation of matrix effects in pesticide multi-residue methods via matrix fingerprinting using liquid chromatography electrospray high-resolution mass spectrometry. Analytical Methods, 8(23), 4664-4673. https://doi.org/https://doi.org/10.1039/C6AY00436A.
Google. (2023a). Google Charts-Visualization: Bubble Chart. Availabe on: https://developers.google.com/chart/interactive/docs/gallery/bubblechart.
Google. (2023b). Google Charts. Available on: https://developers.google.com/chart.
IEC. (31010). IEC 31010:2019. Risk management. Risk assessment techniques. Available on: https://www.iso.org/standard/72140.html.
ISO-7870-2. (2023). Control charts — Part 2: Shewhart control charts. Available on: https://www.iso.org/standard/78859.html.
ISO. (2019). Foods of plant origin - Multimethod for the determination of pesticide residues using GC- and LC-based analysis following acetonitrile extraction/partitioning and clean-up by dispersive SPE - Modular QuEChERS-method. Available on: https://www.une.org/encuentra-tu-norma/busca-tu-norma/norma?c=N0061576.
ISO/IEC-17025. (2017). General requirements for the competence of testing and calibration laboratories. Available on: https://www.iso.org/obp/ui#iso:std:iso-iec:17025:ed-3:v2:es.
ISO/TS-13530. (2009). Water quality — Guidance on analytical quality control for chemical and physicochemical water analysis. Available on: https://www.iso.org/standard/52910.html.
Jirata, U., Asere, T. G., Balcha, Y. B., & Gure, A. (2024). Levels of organochlorine pesticides in onion and tomato samples from selected towns of Jimma Zone, Ethiopia. Heliyon, 10(15), e35033. https://doi.org/https://doi.org/10.1016/j.heliyon.2024.e35033.
Kardani, F., Jelyani, A. Z., Hashemi, M., Rashedinia, M., Shariati, S., Mirzaei, R., . . . Noori, S. M. A. (2023). Determination of 323 pesticide residues in Iran’s cereal by GC-MS and HPLC-UV combined with QuEChERS extraction and mixed-mode SPE clean-up method. Journal of Food Composition and Analysis, 124, 105670. https://doi.org/https://doi.org/10.1016/j.jfca.2023.105670.
Khazri, A., Sellami, B., Dellali, M., Corcellas, C., Eljarrat, E., Barceló, D., . . . Mahmoudi, E. (2016). Diastereomeric and enantiomeric selective accumulation of cypermethrin in the freshwater mussel Unio gibbus and its effects on biochemical parameters. Pesticide Biochemistry and Physiology, 129, 83-88. https://doi.org/https://doi.org/10.1016/j.pestbp.2015.11.001.
Kuchheuser, P., & Birringer, M. (2022). Pesticide residues in food in the European Union: Analysis of notifications in the European Rapid Alert System for Food and Feed from 2002 to 2020. Food Control, 133, 108575. https://doi.org/https://doi.org/10.1016/j.foodcont.2021.108575.
Kwon, H., Lehotay, S. J., & Geis-Asteggiante, L. (2012). Variability of matrix effects in liquid and gas chromatography–mass spectrometry analysis of pesticide residues after QuEChERS sample preparation of different food crops. Journal of Chromatography A, 1270, 235-245. https://doi.org/https://doi.org/10.1016/j.chroma.2012.10.059.
Li, Q., Zhang, J., Lin, T., Fan, C., Li, Y., Zhang, Z., & Li, J. (2023). Migration behavior and dietary exposure risk assessment of pesticides residues in honeysuckle (Lonicera japonica Thunb.) based on modified QuEChERS method coupled with tandem mass spectrometry. Food Research International, 166, 112572. https://doi.org/https://doi.org/10.1016/j.foodres.2023.112572.
Lin, H. M., Gerrard, J. A., & Shaw, I. C. (2005). Stability of the insecticide cypermethrin during tomato processing and implications for endocrine activity. Food Additives & Contaminants, 22(1), 15-22. https://doi.org/https://doi.org/10.1080/02652030400027938.
Mac Loughlin, T. M., Peluso, M. L., Etchegoyen, M. A., Alonso, L. L., de Castro, M. C., Percudani, M. C., & Marino, D. J. G. (2018). Pesticide residues in fruits and vegetables of the argentine domestic market: Occurrence and quality. Food Control, 93, 129-138. https://doi.org/https://doi.org/10.1016/j.foodcont.2018.05.041.
Mandal, K., Joshi, S., Bansal, T., Sharma, S., & Kang, B. K. (2022). Residual determination of spirotetrament and imidacloprid in onion and tomato crop by mass spectrometery. International Journal of Environmental Analytical Chemistry, 1-12. https://doi.org/https://doi.org/10.1080/03067319.2022.2103690.
Ouakhssase, A., & Ait Addi, E. (2023). Monitoring 432 potential pesticides in tomatoes produced and commercialized in Souss Massa region-Morocco, using LC-MS/MS and GC-MS/MS. Environmental Pollution, 337, 122611. https://doi.org/https://doi.org/10.1016/j.envpol.2023.122611.
Patel, B. V., Chawla, S., Gor, H., Upadhyay, P., Parmar, K. D., Patel, A. R., & Shah, P. G. (2016). Residue decline and risk assessment of fluopyram + tebuconazole (400SC) in/on onion (Allium cepa). Environmental Science and Pollution Research, 23(20), 20871-20881. https://doi.org/https://doi.org/10.1007/s11356-016-7331-8.
RD. (2003). Real Decreto 290/2003, de 7 de marzo, por el que se establecen los métodos de muestreo para el control de residuos de plaguicidas en los productos de origen vegetal y animal. Available on: https://www.boe.es/buscar/doc.php?id=BOE-A-2003-4786.
Rutkowska, E., Łozowicka, B., & Kaczyński, P. (2019). Three approaches to minimize matrix effects in residue analysis of multiclass pesticides in dried complex matrices using gas chromatography tandem mass spectrometry. Food Chemistry, 279, 20-29. https://doi.org/https://doi.org/10.1016/j.foodchem.2018.11.130.
SANTE. (2021). Analytical quality control and method validation procedures for pesticide residues analysis in food and feed SANTE 11312/2021 v2. Available on: https://food.ec.europa.eu/system/files/2023-11/pesticides_mrl_guidelines_wrkdoc_2021-11312.pdf.
Skovgaard, M., Encinas, S. R., Jensen, O. C., Andersen, J. H., Condarco, G., & Jørs, E. (2017). Pesticide Residues in Commercial Lettuce, Onion, and Potato Samples From Bolivia-A Threat to Public Health? Environ Health Insights, 11, 1178630217704194. https://doi.org/https://doi.org/10.1177/1178630217704194.
Strassemeyer, J., Daehmlow, D., Dominic, A. R., Lorenz, S., & Golla, B. (2017). SYNOPS-WEB, an online tool for environmental risk assessment to evaluate pesticide strategies on field level. Crop Protection, 97, 28-44. https://doi.org/https://doi.org/10.1016/j.cropro.2016.11.036.
Su, Y., Lu, J., Liu, J., Li, F., Wang, N., Lei, H., & Shen, X. (2024). Optimization of a QuEChERS–LC–MS/MS method for 51 pesticide residues followed by determination of the residue levels and dietary intake risk assessment in foodstuffs. Food Chemistry, 434, 137467. https://doi.org/https://doi.org/10.1016/j.foodchem.2023.137467.
UNE-CEN/TS-17061. (2019). Foodstuffs - Guidelines for the calibration and quantitative determination of pesticide residues and organic contaminants using chromatographic methods (Endorsed by Asociación Española de Normalización in October of 2019.). Available on: https://www.normadoc.com/spanish/une-cen-ts-17061-2019.html.
Veiga-del-Baño, J. M., Cuenca-Martínez, J. J., Andreo-Martínez, P., Cámara, M. Á., Oliva, J., & Motas, M. (2023). Uncertainty and associated risks in the analysis of pesticides in homogeneous paprika samples. Food Chemistry, 429, 136963. https://doi.org/https://doi.org/10.1016/j.foodchem.2023.136963.
Wu, X., & Ding, Z. (2023). Evaluation of matrix effects for pesticide residue analysis by QuEChERs coupled with UHPLC-MS/MS in complex herbal matrix. Food Chemistry, 405, 134755. https://doi.org/https://doi.org/10.1016/j.foodchem.2022.134755.
Xia, Y., Stinner, R. E., Brinkman, D., & Bennett, N. (2003). Agricultural chemicals use data access using coldfusion markup language and a relational database. Computers and Electronics in Agriculture, 38(3), 217-225. https://doi.org/https://doi.org/10.1016/S0168-1699(03)00003-6.
Zhao, J., Pu, J., Wu, X., Chen, B., He, Y., Zhang, Y., & Han, B. (2021). Evaluation of the matrix effect of pH value and sugar content on the analysis of pesticides in tropical fruits by UPLC-MS/MS. Microchemical Journal, 168, 106375. https://doi.org/https://doi.org/10.1016/j.microc.2021.106375.

Supplementary.docx

Download PDF

Journal Publication

published 14 Jun, 2025

Read the published version in Archives of Environmental Contamination and Toxicology →

Editorial decision: Accept after minor revision
11 Mar, 2025
Reviewers agreed at journal
02 Dec, 2024
Reviewers invited by journal
02 Dec, 2024
Editor assigned by journal
07 Oct, 2024
First submitted to journal
07 Oct, 2024

You are reading this latest preprint version

Quick monitoring of routine samples of foods in the regulatory analysis of pesticide residues

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Materials and Methods

2.1. Sample selection

2.2. Chemical and reagents

2.3. Preparation of solvent, matrix calibration and spiked samples

2.4. Pesticide analysis

2.5. Performance of the method and quality assurance

2.6. Database creation and data analysis

2.7. AQA data analysis

3. Results and discussion

4. Conclusions

Declarations

Conflicts of Interest

Funding

References

Supplementary Files

Status:

Journal Publication

Version 1