Development of a Rapid Method for Simultaneous Determination of Pesticides in Plant Oil Using GC-MS/MS

Abstract

Multi-residue methodologies necessitate a tailored and precise approach across all areas of analysis. Analyte extraction must be closely correlated with the sample matrix to ensure the optimal recovery for the maximal array of analytes, thereby fulfilling all legal requirements concerning analytical determination. Although the QuEChERS method offers undeniable advantages, it proves unsuitable for pesticide residue determination in matrices with high oil contents. A pivotal component of the method involves employing n-hexane as the extraction solvent and utilizing solvent exchange in obtaining the final extract. The analytical method developed by our laboratory, as presented here, enabled the determination of all requisite pesticides in accordance with European Union (EU) Regulation 396/2005. The linearity, limits of detection and quantification, matrix effects, accuracy, and precision of the method were evaluated in line with the aforementioned regulation. Of the analyzed pesticide residues, 273 exhibited expanded uncertainty with an RSDr ≤ 20%, with recoveries falling within the range of 70–120%, meeting all the requirements of document SANTE/11312/2021 V2. For 9 pesticides, the recovery was below 30%, and the precision of the obtained content for 21 compounds surpassed 20%, necessitating the removal of these compounds from the analytical method.

1. Introduction

Olive oil, renowned for its rich flavor and numerous health benefits, stands out as a cornerstone of the Mediterranean diet, widely acknowledged for its positive effects on cardiovascular health and overall well-being. Beyond olive oil, various plant-based oils, such as those derived from nuts, seeds, and avocados, offer a diverse array of nutritional advantages. Polyunsaturated fatty oils, including omega-3 and omega-9, play pivotal roles in supporting heart health, reducing inflammation, and maintaining optimal cognitive function. Omega-3 fatty acids, commonly found in fish oil and flaxseed oil, are particularly lauded for their ability to lower triglyceride levels, alleviate symptoms of depression, and promote healthy brain development. Similarly, omega-9 fatty acids, prevalent in olive oil and certain nuts, contribute to reducing LDL cholesterol levels, thereby aiding in the prevention of cardiovascular diseases. Moreover, these oils boast potent antioxidant properties, effectively combating oxidative stress and mitigating the risk of chronic diseases such as cancer and diabetes [1].

Incorporating a variety of plant-based oils into one’s diet can thus provide a holistic approach to nutrition, offering an abundance of essential fatty acids and antioxidants crucial for maintaining overall health and vitality [2]. Pesticides have been integral to agricultural production for over a century, becoming an indispensable component thereof used in food production. Plant protection agents are employed in cultivation to increase crop yields, minimize potential losses due to pests, extend the product shelf life, enhance the suitability for consumption, and maximize organoleptic quality attributes. According to the European Union database, approximately 717 pesticide residues and about 1483 active substances have been found in products available on the market [3,4].

However, the introduction of pesticides into food production has had various health-related consequences. Due to the physicochemical properties and metabolism of a wide range of pesticides, many substances have the capacity to bioaccumulate in animal tissues, owing to their non-polar nature, lipophilicity, and affinity for lipid fractions [5]. Growing public concern about potential health hazards stemming from the presence of pesticide residues in the human diet underscores the need for stringent regulations. To safeguard health and ensure food safety to the fullest extent possible, the European Commission has established maximum residue levels (MRLs) for each agricultural product [6].

In that context, the development of multi-residue analytical methodologies for the quantification of hydrophobic pesticides in edible oils at trace levels remains a formidable challenge, primarily due to the complex nature of these matrices and the low concentrations at which these contaminants are typically found. The chemical diversity of hydrophobic pesticides, which includes various classes such as organochlorines, organophosphates, and pyrethroids, further complicates the analytical process. Each class possesses unique physicochemical properties that can its their extraction, separation, and detection, thereby necessitating tailored approaches for accurate quantification. The diverse spectrum of commercially available pesticide formulations, encompassing multiple active compounds with varied chemical properties, necessitates the continual advancement of cutting-edge analytical methodologies. These methodologies must demonstrate robustness and versatility to use, capable of accommodating a diverse spectrum of analytes, thus ensuring the precise and accurate detection and quantification of residues, even at trace concentrations. The ongoing development and refinement of multi-residue analytical methodologies are essential to address the challenges posed by the quantification of hydrophobic pesticides in edible oils. Advanced instrumental equipment such as GC-MS/MS and LC-MS/MS is indispensable for laboratories to meet the challenges posed, allowing for precise, accurate, selective, and simultaneous determination of substances in diverse food matrices. Methods of implementing such analysis include optimizing extraction procedures to efficiently recover pesticides from oil matrices, fine-tuning the chromatographic conditions for optimal separation, and calibrating the mass spectrometric detection parameters to achieve the desired sensitivity and selectivity [7,8,9].

The analysis of pesticide residues is an extremely difficult task in view of the low concentrations of these compounds in plant matrices, which hinder their detection. Determining pesticide residues in matrices with high oil contents and low to moderate water contents poses a challenge due to signal suppression in chromatographic analysis, often resulting in poor recovery and inadequate precision for certain pesticide residues. The QuEChERS method successfully enables analyte extraction from products with high water contents, but it encounters limitations when applied to items with high oil contents [10].

To enhance detection and extraction methods, various pre-treatment techniques have been employed to concentrate analytes, significantly reducing solvent consumption. Among the innovative methods for the quantitative determination of pesticide residues, solid-phase extraction (SPE) [11], liquid–liquid extraction, dispersive liquid–liquid microextraction (DLLME) [12], solid-phase microextraction (SPME) [13], and single-drop microextraction (SDME) have been utilized [14].

In the case of sample preparation for determination using LC-MS/MS, the normative analytical method QuOil was employed [14]. Moreover, many research works utilized the Quick, Easy, Cheap, Effective, Rugged, and Safe (QuEChERS) technique for extracting pesticide residues from plant products with high oil contents [15].

The objective of this study was to develop an analytical method successfully enabling the determination of 273 pesticides, to meet the requirements outlined in document SANTE/11312/2021 V2, and with the requisite analytical scope in accordance with 396/2005. The developed method aligns with the latest European guidelines for pesticide residue analysis, as outlined in document SANTE/11312/2021 v2 [16].

2. Materials and Methods

2.1. Chemicals, Reagents, and Standards

Pesticide-grade acetonitrile (ACN) (purity ≥ 99.9%) and n-hexane (purity ≥ 99.9%), and primary–secondary amine (PSA) sorbent were purchased from JT Baker Chemicals (Phillipsburg, NJ, USA). The selection of pesticides was performed according to the EU Pesticide Database and the literature. Pesticide analytical quality standards were purchased from LGC Standards, CPAChem, and Tusnovic (purity > 99%). All analytical standards were compliant with ISO 17025 [17] and 17034 norms [18].

2.2. Instrument and Experimental Conditions

GC-MS/MS analyses were carried out with a Shimadzu Nexis GC-2030 (Shimadzu Corporation, Kyoto, Japan), equipped with an AOC-6000 Plus auto-sampler (Shimadzu Corporation, Kyoto, Japan) and coupled with a triple quadrupole mass spectrometer equipped with an electronic ionization source (70 eV, 280 °C) (GCMS-TQ8040NX (Shimadzu Corporation, Kyoto, Japan)). The GC-MS/MS analyses were performed according to the method and parameters presented below. Sample extracts were injected using an AOC-6000 Plus auto-sampler (Shimadzu Corporation, Kyoto, Japan), with a 10 μL syringe, into a multi-mode injector equipped with an ultra-inert dimpled inlet liner (2 mm ID) from Shimadzu.

The injector temperature was set at 250 °C. Splitless-mode injection for 1 min was applied with an injection volume of 1 μL. Separation was performed on an SH-I-5MS (Shimadzu Corporation, Kyoto, Japan) fused silica capillary column of 30 m × 0.25 mm, at 0.25 µm of the film thickness (Shimadzu). Helium (purity 99.9995%) was used as the carrier gas at a constant-pressure flow rate of 1.4 mL/min. The column oven temperature program was as follows: the initial temperature was set at 105 °C and maintained for 3 min; then, it was increased to 130 °C at a rate of 10 °C/min, increased to 200 °C at a rate of 4 °C/min, increased to 280 °C at a rate of 8 °C/min, and held for 10 min. The total program time was 43 min. The mass spectrometer was operated with an electron impact (EI) source using the multiple reaction monitoring (MRM) mode. In Table S1, a characterization of the ions used for qualitative analysis of the pesticides is presented. The ion source and transfer line temperatures were set at 225 °C and 280 °C, respectively. Each target compound was monitored in the multiple reaction monitoring (MRM) mode used for both quantification and identification purposes. The software LabSolutions GCMS (version 4.20) was used for instrument control, and data acquisition and processing. The identification of target compounds relied on comparing the retention times (tR) of chromatographic peaks for the estimation and qualification of ions with the peaks of reference standards. Each standard was injected twice at the beginning and end of the validation sequence. Table S1 presents the operational parameters of the tandem mass spectrometer MS/MS, with a specific emphasis on the most intense transitions in multiple reaction monitoring (MRM), as well as collision energies. Additionally, the Kovats retention index (IR) and retention time (tR) are provided for each of the analyzed compounds.

2.3. Preparation of Standard Solutions

Preparation of standard solutions involved precise procedures wherein each analyte was dissolved in an acetone mixture to achieve a concentration of approximately 1000 µg/mL. These solutions were then stored at −20 °C for a period not exceeding twenty-four months, secured in 10 mL dark glass vials. Furthermore, formulated working solutions, encompassing all analytes at calibrated concentrations ranging from 2 to 10 µg/mL in toluene, were devised. These working solutions were preserved at 20 °C and, prior to utilization, allowed for equilibrating to room temperature to ensure accuracy in subsequent analyses. Subsequently, these calibrated working standards played a pivotal role in enriching samples earmarked for the method’s validation, thereby ensuring accuracy and reliability. Additionally, they were utilized in constructing a calibration curve spanning the linearity range from 0.002 mg/kg to 0.5 mg/kg and used to enrich the samples used for the method’s validation.

2.4. Sample Preparation and Clean-Up Procedure

The analytical procedure was developed for the effective analysis of plant-derived foodstuffs containing a high oil content and low to moderate water content. The scheme of the method is presented in Figure 1. To do so, 3 g of oil sample was weighed into a 50 mL centrifuge tube. The sample was enriched with a working solution. Subsequently, 1.5 mL of n-hexane was added followed by 6 mL of ACN. The sample was vortexed for 15 min using a vortex mixer. The sample was then left for 20 min to allow for the separation of immiscible solvent phases. The maximum volume of solvent (15 mL) was transferred to a tube. The extraction step was repeated, and then 6 mL of ACN was added, vortexed for 15 min, and left for 20 min for phase separation. The maximum volume of solvent was again transferred to the tube (15 mL). The obtained extract was vortexed for 2 min. Then, 4 mL of the extract was transferred to a new tube, and 200 mg of PSA was added to the new tube with a volume of 15 mL. PSA is commonly employed for the removal of organic acids, fatty acids, sugars, and pigments. The tube was vortexed and centrifuged at 3000 rpm for 5 min. The maximum volume from the above sediment was transferred to a pre-washed vial and evaporated to dryness with a drop of n-decane. The dried residue was reconstituted with a mixture of toluene: acetonitrile 9:1 in a volume of 1 mL.

2.5. Method’s Validation

The analytical method was verified against the assumptions outlined in document SANTE/11312/2021v2. The evaluated parameters included the linearity of the calibration curve based on the matrix samples. The linearity of the calibration curve was determined by calculating the coefficient of determination (R2) and BCC ≤ ±20% at 5 concentration levels performed for each compound in the linearity range of 0.002 mg/mL–0.5 mg/mL. The matrix effect was examined in comparison with the calibration curve for chromatographic standards in the solvent. The matrix effect verification aimed to quantify the analytes’ suppression or enhancement of the response by the detection system. The mean percentage recovery (RE%) was set at 70–120%, with an RSD ≤ 20%. In exceptional cases, an average recovery in the range of 70–120% may be accepted if the analytes are consistent (RSD ≤ 20%), but the average absolute recovery should not be below 30% or above 140%. According to the assumptions of the document, the lowest concentration should be ≥LOQ, while the next validation level should fall within the range of 2–10 times the LOQ value. In the examined validation experiment, the second enrichment level was set at 5 × LOQ. The method’s precision should be within ≤20%, and the expanded uncertainty of the method should be within ≤50%. Quality parameters of the chromatographic resolution should be ±30% (relative value) of the average ratio of ions of calibration standards in the same sequence, and the retention time should be ±0.1 min. Additionally, limits of detection (LODs) and limits of quantification (LOQs) were investigated, where the LOQ value should be ≤MRL. The values of LOQ and LOD were calculated using the signal-to-noise (S/N) ratio comparison method. The determination of LOQs and LODs involved comparing the concentration of the chromatographic peak with the lowest signal to the chromatographic signal obtained from the measurement of a pesticide-free sample. The LOD value represents a signal-to-noise ratio of 3, while the LOQ is set at 10 compared to the noise level of the pesticide-free sample signal. The test material subjected to examination consisted of olive oil samples devoid of pesticide residues. The oil matrix was organic olive oil purchased for validation purposes.

3. Results

In summary, this is a novel sample preparation method utilizing the addition of n-hexane and sample evaporation, coupled with GC-QqQ-MS/MS, for the multi-class determination of pesticide residues in edible oils, providing the possibility of effectively removing matrix components.

The analytical method was validated according to a plan based on the criteria of document SANTE/11312/2021 V2. Calibration curves were prepared in a matrix with a high oil content and low to moderate water content (oil). Recovery studies were conducted at two levels of concentrations, LOQ and 5 × LOQ, with concentration ranges consistent with Table S2. Based on the obtained results, the following validation parameters were determined: LOQ, recovery, repeatability, linearity, matrix effect, and expanded uncertainty (Table S1).

The utilization of three-stage solvent extraction results in the dilution of the extract, which may lead to a significant reduction in signal intensity during direct analysis. Consequently, the application of final extract evaporation becomes necessary. Substituting the final solvent with a toluene–acetone mixture in a ratio of 9:1 enables excellent dissolution of all analytes. The selection of the necessary pesticide analytes was based on the guidelines provided by the European Commission, Regulation 396/2005, taking into account the maximum complexity and chemical diversity, as well as laboratory availability.

The compounds were identified through injection of a standard mixture, taking into account the retention time with a permissible difference of 0.1 min and MRM ion transition ratios within ±30%. The limit of quantification (LOQ) for each pesticide was established at the lowest calibration level corresponding to the concentration of a single pesticide yielding a signal-to-noise ratio greater than 10. Meanwhile, the limit of detection (LOD) value was consistently higher than 3. Linearity was evaluated in the range of 0.002 to 0.5 mg kg⁻¹, with coefficients of determination ranging from 0.995 to 0.9999.

Recovery was assessed by analyzing enriched samples at the LOQ and 5 × LOQ concentration levels, according to Table S2. Recoveries ranged from 35% to 116%, as shown in Table S2. Recoveries below 30% were obtained for nine compounds: chlorotoluron, dichlorobenzophenone, 4,4′-dichloran, nitralin, pentachlorophenol, piperophos, propachizafop, terbacil, tetrachlorvinphos, and thiometon. Meanwhile, the criterion of RSD ≤ 20% was exceeded for 21 compounds: bifenazate, chinomethionat, chlortoluron, cypermethrin, disulfoton sulfoxide, dodemorph, ethofumesate, fenamiphos sulfone, fenhexamid, fenmedifam, fenoxaprop-ethyl, fenpropidin, fenpropimorph, fenthion oxon sulfone, furathiocarb, furilazole, pendimethalin, phosmet, pyridalyl, spiroxamine, and trinexapac ethyl. Furthermore, the recovery of aldrin, biphenyl, bromociclen, carbofuran, chlorothalonil, endrin, folpet, heptachlor, hexachlorobenzene, pentachloroaniline, proquinazid, quinocetone, and quinoxyfen analytes was found to be below the 70–120% range. Nevertheless, these compounds were positively assessed in the validation experiment due to their relative standard deviation (RSD) being ≤ 20%, and thus they were included in the analysis.

The matrix effect was examined for all analyzed compounds, representing the ratio between the slope of the calibration curve in the matrix and the slope of the calibration curve in the solvent. For 14 pesticides, the matrix effect was lower than 20%; in the range of 20–50%, this effect was observed for 73 pesticides, and a value above 50% was noted for 207 pesticide residues. The assessment of method robustness revealed that employing a two-step extraction of the analytical sample increased the analyte recovery by 15%. A prolonged 20 min mixing duration was necessary to achieve satisfactory results. Furthermore, it was demonstrated that changing the organic solvent from acetonitrile to a mixture of toluene–acetone (1:1), due to evaporation to dryness, improved the symmetry of the chromatographic peak and reduced the interference levels, resulting in decreased detection levels for the majority of the analytes.

4. Discussion

The formulation of multi-residue methodologies for the quantification of hydrophobic pesticides within oil matrices poses a rigorous analytical hurdle. To uphold optimal chromatographic system efficacy, meticulous sample decontamination from interfering agents is paramount [19].

This study assessed the extraction efficiency of 294 pesticides from a high-fat matrix-unrefined olive oil. A key element of the extraction is the addition of n-hexane. Previously, several extraction methodologies utilizing this solvent were developed; however, the range of analyzed compounds was not as extensive [20,21].

The main analytical challenge during chromatographic analysis of pesticide residues is the matrix effect. To minimize the matrix influence, Soltani et al. employed a eutectic solvent consisting of thymol and vanillin (1:1) deep eutectic solvent (DES), and utilized a mixture of acetonitrile and DES. The initial extraction included a 4 mL addition of n-hexane, similar to the presented procedure. The addition of n-hexane significantly improves the extraction of non-polar pesticides. The researchers removed the upper layer of n-hexane, and further analysis was based on utilizing the potential of the eutectic mixture. The use of an acetonitrile: DES (1:1) mixture significantly increased the peak height in chromatography and reduced the intensity of noise. The recovery for 16 pesticides ranged from 63.1% to 119.4% [22].

In another study, the QuEChERS method was implemented, incorporating enhancements such as the use of d-SPE with EMR-Lipid^TM as the second purification step after freezing at −76 °C. Recoveries of 70–120% and a mean RSD of 4.0% were achieved for 177 pesticides, which constituted 83% of the total number of pesticides investigated [23].

Similar results were obtained by Dias et al. using a specific EMR-Lipid purification sorbent. Recoveries in the range of 70–120% were achieved for 91% of the compounds out of 165 pesticides tested. The use of other purification salts did not yield such significant results [24].

The EMR-Lipid dSPE cleaning vial was introduced to the market in 2015, containing specially developed sorbents that interact specifically with matrix components. The effectiveness of cleaning oily matrices was also confirmed by Zhao et al. (2019) using the EMR-Lipid dSPE cleaning cartridge. Forty-six pesticides were tested, of which >95% achieved an average recovery of >70% and RSD < 15% [19].

In another study focusing on the determination of polar pesticide residues using HILIC-TOFMS, an attempt was made to add n-hexane and saturated naphtha ether with acetonitrile during the extraction procedure to avoid interferences from non-polar compounds. However, this approach was rejected due to the lack of a significant effect in a subsequent analysis using HILIC-TOFMS [25].

Extraction using the Quick, Easy, Cheap, Effective, Rugged, and Safe (QuEChERS) method and its modifications is chosen by many researchers for the preparation of analytical sample extracts. Sample extraction in most proposed analytical methods occurs via a liquid–liquid extraction (LLE) technique using a solvent mixture of n-hexane, acetonitrile, and saturated naphtha ether. However, a drawback of this method is the significant co-extraction of interfering substances. It is common practice to freeze the sample to remove lipid components; however, this may decrease the solubility of analytes at low temperatures. Extraction using dSPE salts is also not a suitable choice for matrices with high fat contents due to the lack of selectivity in lipid removal [26,27]. Drakopoulou et al. (2024) used the QuEChERS extraction method, and 65 pesticides met the validation requirements, but the analytical scope of the method is too narrow to meet EU requirements [28].

Although the matrix effect was reduced in the presented method, this issue was not completely eliminated. Wang et al. (2022) presented an interesting proposal for a novel clean-up method employing emulsification and demulsification (ED) for edible oil sample preparation. Spontaneous emulsification generates micron-sized droplets in a water/co-solvent/oil system. These droplets are demulsified using a hydrophilic polyamide membrane with submicron pores, where they aggregate and are retained, permitting the acetonitrile/water phase containing pesticides to pass through efficiently [29]. However, the analysis of analytes in oils is challenging due to the co-extraction of lipophilic matrices with the target analytes. Despite the advantages of SFO-DLLME, the complexity of edible oil samples and the solidification of oil at low temperatures complicate its application, necessitating an additional sample preparation step [11].

The selection of cleaning sorbents can be crucial for the efficiency of extraction. Primary secondary amine (PSA) is frequently used, among others, for the removal of fatty components. In analytical chemistry, the choice of appropriate sorbents for the clean-up step significantly impacts the overall performance and accuracy of the extraction process. PSA, in particular, is widely employed due to its effectiveness in eliminating the lipid content, which can otherwise interfere with the detection and quantification of target analytes. Analytical methods based on the QuEChERS method use this substance to effectively remove lipid components [15,30,31].

An important trend in analytical research is the minimization of organic solvent usage. A notable approach was proposed by Zhang et al. in 2022, where they employed an innovative technique using miniaturized kapok-fiber-supported liquid extraction (mini-KF-SLE). This method exemplifies advancements in environmentally friendly analytical methodologies, focusing on minimizing the ecological footprint of solvent-based extraction while maintaining or enhancing analytical performance. The mini-KF-SLE technique utilizes the unique properties of kapok fibers to support the extraction process, providing an efficient and sustainable alternative to conventional extraction methods [32]. If following current trends, it is essential to first develop an effective extraction procedure and then aim for maximum minimization of organic solvent usage, while ensuring compliance with all EU regulations for the analytical method to be implementable. Therefore, the next stage of developing this method will proceed accordingly. In the analytical method’s development, the primary focus should be on establishing a robust and efficient extraction process. Following this, efforts should be directed towards improving the method to minimize its organic solvent usage and thus enhance its sustainability and reduce resource consumption. Compliance with stringent EU regulatory standards is crucial to ensure that the method is not only scientifically sound but also legally viable for practical implementation. Future advancements in this method will prioritize these objectives to align with contemporary analytical research trends [33].

Zayats et al. (2013) observed a clear dependence. A hexane/acetonitrile extraction solution in a ratio of 1:4 resulted in logarithms with partition LgP values of around 3.2 for all pesticides in the extracted food samples. Recovery for a wide range of analyzed pesticide residues could be achieved with over 99% efficiency, except for HCH and fenpropimorph, through three-stage extraction using n-hexane with acetonitrile. Additionally, n-hexane extraction allowed for the removal of triglycerides, the main matrix component causing significant signal suppression in chromatographic analysis [30].

Various studies conducted to develop analytical methods are bringing us closer to meeting the requirements imposed by document SANTE/11312/2021 V2. Nonetheless, continuous validation attempts must be made to meet the criteria for all analyzed pesticide residues according to 365/2005 [33,34].

5. Conclusions

The identification of pesticides in food samples is crucial for food safety monitoring purposes. This has study presented detailed data obtained from the analysis of 294 different pesticides, of which 273 met the requirements specified in document SANTE/11312/2021 V2. The methodology underwent validation for linearity, accuracy, the matrix effect, and precision, with the detection and quantification limits of this method also estimated. The application of a chromatographic technique coupled with triple quadrupole mass spectrometry (GC-QqQ) enables the identification of both the most volatile and stable pesticide residues in olive oil. The extraction method is characterized by its speed, simplicity, and high sensitivity, allowing for the detection of most compounds in a single analysis. The presented method meets the requirements for high-throughput detection of pesticide residues in olive oil and serves as a reference point for the analysis of pesticide residues in other oils, as well as for the development of automated sample preparation processes for complex matrices.