*Article* **Determination of Fosetyl-Aluminum in Wheat Flour with Extract-Dilute-Shoot Procedure and Hydrophilic Interaction Liquid Chromatography Tandem Mass Spectrometry**

**Xianjiang Li \*, Sheng Wang, Zhen Guo, Xiuqin Li, Qinghe Zhang and Hongmei Li**

Food Safety Laboratory, Division of Metrology in Chemistry, National Institute of Metrology, Beijing 100029, China; wangsh@nim.ac.cn (S.W.); guozh@nim.ac.cn (Z.G.); lixq@nim.ac.cn (X.L.); zhangqh@nim.ac.cn (Q.Z.); lihm@nim.ac.cn (H.L.)

**\*** Correspondence: lixianjiang@nim.ac.cn; Tel.: +86-10-64524737

**Abstract:** Fosetyl-aluminum is a widely used ionic fungicide. This pesticide is not amenable to the common multi-residue sample preparation methods. Herein, this paper describes a novel method for the simple and sensitive determination of fosetyl-aluminum residue in wheat flour. The sample preparation method involved extraction with water under ultrasonication and subsequent dilution with six-fold acetonitrile. The fosetyl-aluminum concentration was determined by hydrophilic interaction liquid chromatography tandem mass spectrometry. The limit of detection and quantification were only 5 and 10 ng/g, respectively, which meet the requirement of the current European legislation. Matrix-matched linearity (r<sup>2</sup> = 0.9999) was established in the range of 10–2000 ng/g. Satisfactory recoveries were achieved in the range of 95.6% to 105.2% for three levels of spiked samples (10, 50, and 100 ng/g). Finally, the method was applied to analyzing 75 wheat flour samples produced in four provinces in China. Two samples were positive with concentrations over the limit of detection. This is the first method focusing on fosetyl-aluminum determination in wheat flour with an extractdilute-shoot strategy and is very promising for the routine quality control of fosetyl-aluminum in similar cereal matrices.

**Keywords:** extract-dilute-shoot; fosetyl-aluminum; hydrophilic interaction liquid chromatography; wheat flour

#### **1. Introduction**

Fosetyl-aluminum (fosetyl-Al) is a polar fungicide, and a replacement for the banned sodium arsenite [1]. This fungicide is widely used to control rot in plant roots. Due to its toxic effects, the Joint Meeting on Pesticide Residues recommended set the acceptable daily intake for fosetyl-Al to 0–1 mg/kg bw per day in 2017 [2]. The maximum residue level of fosetyl-Al in food is strictly controlled by Regulation 396/2005 of the European Commission [3], GB 2763-2021 of China [4], and 40CFR180.415 of the USA [5].

Fosetyl-Al is a highly polar compound with an ionic structure. This highly polar pesticide is not amenable to the common multi-residue sample preparation methods because it is difficult to partition into common organic solvents [6] and needs dedicated chromatographic conditions. In addition, it is difficult to retain fosetyl-Al in typical reverse-phase liquid chromatography and the co-eluted salts and polar matrix components seriously interfere with fosetyl-Al determination [7]. Therefore, there is an urgent demand for the development of a simple and general method that could be used to detect this "orphan pesticide". Until now, few solutions that address this problem have been reported, such as applying a specific column with a polar stationary phase or using an ion-pair reagent within the mobile phase [7]. Because it lacks UV absorption and fluorescence, fosetyl-Al is rather difficult to determine by conventional liquid chromatography detectors such as diode array detectors. In recent years, the use of liquid chromatography mass

**Citation:** Li, X.; Wang, S.; Guo, Z.; Li, X.; Zhang, Q.; Li, H. Determination of Fosetyl-Aluminum in Wheat Flour with Extract-Dilute-Shoot Procedure and Hydrophilic Interaction Liquid Chromatography Tandem Mass Spectrometry. *Separations* **2021**, *8*, 197. https://doi.org/10.3390/ separations8110197

Academic Editor: Fabio Gosetti

Received: 2 October 2021 Accepted: 21 October 2021 Published: 24 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

spectrometry (LC-MS) for analyte determination has aroused great attention due to its inherent selectivity and sensitivity [8]. Several recent works have described the use of LC-MS for fosetyl-Al determination in different food commodities, such as lettuce [7], tomato [9], grape [1], mango [10], olive oil [11], oat, and soy beans [12]. In these LC-MS methods, polar pesticides are separated with graphitized carbon (Hypercarb) columns [13] or hydrophilic interaction liquid chromatography (HILIC) columns [9,14,15]. To the best of our knowledge, no work has reported the determination of fosetyl-Al in wheat flour with the HILIC-MS/MS method.

Wheat flour is a staple food in many countries. It is a typical matrix that is high in carbohydrates and proteins and belongs in category 5 of the AOAC Food Triangle. During its long-term storage, fosetyl-Al may be illegally spiked into wheat flour to avoid fungus growth. Scarce literature is available for the analysis of this polar pesticide in wheat flour. Therefore, the State Administration for Market Regulation of China set up a project and supported us in developing a method for fosetyl-Al determination in wheat flour. Effective removal of the polar, soluble carbohydrates and proteins poses a great challenge in the sample preparation process. For the LC-MS method, sample preparation is necessary prior to analyte detection to eliminate interferences [16] and a lower matrix effect (ME) for MS determination [17–19], such as the hydrophilic-lipophilic balance cartridge [1] and the anion-exchange column [20]. One widely adopted sample preparation method was developed by the European Reference Laboratory to determine highly polar pesticides in plant-derived foods such as soy flour, named the Quick Polar Pesticides (QuPPe) method [21]. Although the method is capable of extracting various polar analytes, the extracts probably contain large amounts of matrix interferences that would contaminate the instruments. Furthermore, this QuPPe strategy is tedious and time-consuming. On the contrary, the extract-dilute-shoot procedure is a promising strategy for its simple and fast operation. As an improved method from dilute-and-shoot [22], this method is fit for solid samples. In detail, solid samples are extracted with a suitable extraction solvent to facilitate the migration of analytes into the liquid phase. Then, the liquid extract was diluted with a suitable dilution solvent before the shoot step. The dilution minimizes the ME and thus, reduces the need for an additional clean-up procedure. Recently, this procedure has gained increased attention in food analysis, including rice [23], tomato [24], fruit jam [25], and gingerbread [26]. However, the HILIC-MS/MS method with the extract-dilute-shoot procedure has not yet been applied for fosetyl-Al determination.

The aim of this work was to develop a simple and sensitive analytical method for fosetyl-Al determination in wheat flour. Therefore, we adopted an extract-dilute-shoot strategy for sample preparation, and combined it with HILIC separation and tandem mass spectrometry determination. To the best of our knowledge, this is the first report that focuses on fosetyl-Al determination in wheat flour with extract-dilute-shoot sample preparation and HILIC-MS/MS determination.

#### **2. Materials and Methods**

#### *2.1. Chemicals and Materials*

HPLC-grade acetonitrile and methanol were bought from Merck (Darmstadt, Germany). Water was supplied by Hangzhou Wahaha Group Co., Ltd. (Hangzhou, China). Ammonium formate and formic acid (FA) were provided by Sigma-Aldrich (St. Louis, MO, USA). Analyte-grade fosetyl-Al standard was purchased from Dr. Ehrenstorfer (Augsburg, Germany). All wheat flour samples were commercial products from the local market.

For the syringe filter (0.22 μm), GHP was from PALL Life Sciences (Ann Arbor, MI, USA), nylon was from UA Filter & Chrom (Taibei, Taiwan, China), PTFE PVDF and PP were from Jinteng Experimental Equipment (Tianjin, China). The syringe (2 mL) was from Jiangsu Yuzhi Medical instrument (Taixing, China).

Chromatography columns were Hypercarb columns (2.1 × 100 mm, 5 μm) from Thermo Fisher Scientific (Bellefonte, PA, USA); BEH C18 column (2.1 mm × 50 mm, 1.7 μm) and BEH amide column (2.1 × 50 mm, 1.7 μm) were obtained from Waters (Milford, CT, USA). Guard columns of the same stationary phase were connected in front of the separation columns. All the columns were preconditioned according to the manufacturer's instructions before use.

#### *2.2. Preparation of Calibration Solutions*

A stock solution of fosetyl-Al was prepared by dissolving an accurately weighed portion of the pure standard compound in acetonitrile at a concentration of 250 μg/g. Blank wheat flour samples were spiked with the fosetyl-Al stock solution to prepare matrixmatched external calibration solutions. The concentrations of fosetyl-Al varied between 5 and 2000 ng/g. Then, the spiked samples were extracted by water, and the extracts were diluted with 6-fold acetonitrile. After centrifugation and filtration, the matrix-matched standards were ready for further use. All solutions were stored at −20 ◦C in the dark.

#### *2.3. Instruments*

Samples were centrifuged by a Hettich universal 320R centrifuge to separate the supernatant (Tuttlingen, Germany). Samples and solutions were mixed by a Vortex Genie-2 from Scientific Industries Inc. (New York, NY, USA). Samples were sonicated with a Branson 8510 Ultrasonic Cleaner (Danbury, CT, USA).

HILIC-MS/MS experiments were performed using a Waters instrument. An AC-QUITY UPLC® system was used for LC separation. The LC system was connected to a triple quadrupole MS (TQ-S, Manchester, UK) with a Z-spray electrospray ionization interface. MassLynxTM 4.1 software (Milford, CT, USA) was used for instrument control and data acquisition. Nitrogen was used as the nebulizer gas and was supplied by the generator NM31LA of Peak Scientific (Scotland, UK).

#### *2.4. Optimization of Extract-Dilute-Shoot Parameters*

To accurately quantify fosetyl-Al concentration in wheat flour, it was important to extract fosetyl-Al and remove matrix interference by efficient sample preparation. To obtain optimal extraction efficiency for fosetyl-Al, several important parameters influencing the extraction efficiency were evaluated in this study, such as type and volume of extraction solvent, dilution factor of the extract, and type of filter. To obtain a reliable result, all the optimizations were carried out in fosetyl-Al-spiked wheat flour at a concentration of 50 ng/g in triplicate.

#### *2.5. Sample Pretreatment*

A 2.00 g portion of wheat flour was accurately weighed into a 50 mL polypropylene centrifuge tube, and mixed with 12 mL water using an automatic pipette. The tube was capped tightly and vortexed for 2 min to form a homogeneous paste. Then, the paste was extracted under sonication for 15 min at room temperature. Afterward, the mixture was vortexed again for 1 min. Then, the tube was centrifuged at 9000 rpm for 3 min, and 1 mL supernatant was collected and diluted with 6-fold acetonitrile. Finally, the extract was filtered with a 0.22 μm GHP syringe filter before analysis by HILIC-MS/MS.

#### *2.6. Hydrophilic Interaction Liquid Chromatography Tandem Mass Spectrometry*

Three different columns were tested for the retention of fosetyl-Al, including BEH C18, BEH amide, and Hypercarb. Afterward, the mobile phase and additive were also optimized to achieve better separation and peak shape in liquid chromatography.

The ion source parameters were optimized automatically by a TQ-S system (Waters, Manchester, UK) with the direct infusion of fosetyl-Al solution at a flow rate of 10 μL/min, source temperature of 150 ◦C, desolvation temperature of 500 ◦C, capillary voltage of −2.3 kV, corn voltage of 30 V, source offset voltage of 50 V, desolvation gas at 700 L/h, cone gas at 150 L/h, and collision gas at 0.13 mL/min. The detection of fosetyl-Al was performed in multiple reactions monitoring mode with a collision energy of 10. The precursor ion with *m/z* 109 corresponded to the fosetyl anion. The product ion with *m/z* 81

was the product of the McLaffery rearrangement with the loss of ethene, and the product ion with *m/z* 63 was PO2<sup>−</sup> [7]. Therefore, the most intense transition of *m/z* 109 > *m/z* 81 was selected for quantification and the transition of *m/z* 109 > *m/z* 63 was used for qualification. Two transitions were selected to qualitatively and quantitatively detect fosetyl-Al in the validation study.

#### *2.7. Method Validation*

To validate the applicability of the developed method for fosetyl-Al determination, the linearity, linear range, limit of detection (LOD), limit of quantification (LOQ), ME, recovery, and precision were investigated. A calibration curve for the quantitative analysis was established using the matrix-matched standard of spiked wheat flour samples. The standards were produced using the developed extract-dilute-shoot method. The spike concentrations were in the range of 5–2000 ng/g. The linearity of the method was evaluated by using a linear regression curve fit with the areas obtained for the matrix-matched standard. The LOD and LOQ were calculated from chromatograms of fortified samples according to SANTE/2020/12830, defined as the concentration of signal-to-noise ratios larger than 3 and 10, respectively. To quantitively evaluate the ME (suppression or enhancement), standard solutions of both solvent and matrix were shot into the HILIC-MS/MS. The ME was assessed by the slope of the calibration curve between the matrix and solvent standards by the below equation [27]. An ME of less than 100% indicates matrix suppression; an ME greater than 100% indicates matrix enhancement. To establish the reliability and validity of the analytical method, recovery and precision tests of fosetyl-Al were carried out in blank samples fortified with three different levels (10, 50, and 100 ng/g) with five replicates.

$$\text{ME} = \text{b}\_{\text{m}} / \text{b}\_{\text{s}} \times 100\%$$

where bm and bs are the angular coefficients of the curve in the matrix and in the solvent, respectively.

#### *2.8. Real Sample Analysis*

Seventy-five commercial wheat flour samples were collected from four provinces of China, including Shandong (25 samples), Henan (15 samples), Hebei (3 samples), and Jiangsu (32 samples). Two replicates were tested from each sample to ensure that reliable results were collected. Retention time and intensity of product ions were used to identify positive samples. Quantification was achieved using an external matrix-matched calibration curve that was produced from the peak area of fosetyl-Al versus the corresponding concentration of the spiked wheat flour samples.

#### **3. Results and Discussion**

#### *3.1. Extract-Dilute-Shoot Optimization*

#### 3.1.1. Type of Extraction Solvent

The choice of extraction solvent directly impacted the extraction efficiency of fosetyl-Al for further analysis. Initially, acidified methanol, described by the QuPPe method, was used for the extraction of the fosetyl-Al. However, this extraction method was not suitable to the wheat flour matrix, because a doughy mixture was inevitably generated when the acidified methanol was added. The doughy substance was difficult to remove from the solution via high-speed centrifugation or syringe filter. This finding is consistent with a previous study about the analysis of fosetyl-Al in soy nutraceuticals [28]. To achieve a satisfactory extraction efficiency, five kinds of polar extraction solvents were tested in this study. The extraction was evaluated by recoveries of spiked fosetyl-Al. From the results in Figure 1A, polarity played a dominated role in the extraction; therefore, recoveries were in the following order: water > water/methanol (50%) > methanol > water (0.5% FA) > isopropanol > acetonitrile. This indicated that water had excellent performance for the extraction of fosetyl-Al. In water, fosetyl-Al transforms into fosetyl anions. The addition of FA would turn the negative ion to neutral molecular and lower

extraction efficiency. Therefore, water was used in the following extractions, consistent with a previous report [29].

**Figure 1.** Optimization of the (**A**) type and (**B**) volume of extraction solvent; (**C**) effect of dilution fold; (**D**) type of filter (*n* = 3).

#### 3.1.2. Volume of Extraction Solvent

The volume of water can directly influence the extraction efficiency of wheat flour. A larger volume of extraction solvent would extract more fosetyl-Al in theory, while signal intensity may decrease remarkably because of the dilution effect. On the other hand, a smaller volume would not provide enough quantity of the sample for LC-MS analysis and the reproducibility would be poor. Herein, the volume of water was optimized from 4 L to 16 mL for 2.00 g wheat flour to find the optimal volume. From Figure 1B, the recovery increased with water volume from 4 to 12 mL, and no significant signal enhancement was observed with increasing water volume. Better recovery would provide better accuracy and that was very important for the real samples test. Moreover, larger volume led to less variation. As a result, 12 mL was chosen for following experiment.

#### 3.1.3. Effect of Dilution Fold

In wheat flour, the main compositions were 71.2% carbohydrate, 15.1% protein, 2.7% lipid, and 9.4% water [30]. During the water extraction, some soluble carbohydrates and proteins are extracted simultaneously. Therefore, acetonitrile was chosen to dilute the extract solution to precipitate these compounds and minimize the ME. Acetonitrile is widely used for protein precipitation [31] and it can lower the solubility of carbohydrates [32]. Once acetonitrile was added, the extract turned cloudy immediately. The employed dilution factors usually range from 2 to 50 according the matrix [22]. Thus, dilution fold was optimized from 2 to 10, as exhibited in Figure 1C. Recoveries increased with higher dilution and relative standard deviations (RSDs) decreased. This indicated that lower ME and better repeatability were achieved with acetonitrile dilution. Conversely, large dilution factors would subsequently reduce the sensitivity of the method in terms of LOD. In conclusion, a six-fold dilution factor was selected as a compromise between sensitivity and repeatability.

#### 3.1.4. Type of Filter

To remove the cloudy particles in solution and prolong the lifetime of the column, filtration is necessary before LC separation. Sometimes, pesticides can be adsorbed by the membrane filters and lead to analyte loss and low recovery [33]. As a result, filter

adsorption of fosetyl-Al was evaluated among five widely used filters (GHP, nylon, PTFE, PVDF, and PP). From the results in Figure 1D, there was no significant difference between the tested filters. So, any kind of the tested filter was feasible for further tests.

#### *3.2. Hydrophilic Interaction Liquid Chromatography Tandem Mass Spectrometry*

The column was selected on the basis of the polarity of fosetyl-Al. As shown in Figure 2, preliminary testing with the BEH C18 column demonstrated its low retention capability with fosetyl-Al, which appeared in the peak void with a retention time of 0.35 min. Due to the highly polar nature of fosetyl-Al, it was difficult to obtain enough retention on a C18 column. Conversely, the Hypercarb column exhibited such a strong retention for fosetyl-Al that it resulted in an increased elution time. The BEH amide column had better retention behavior toward fosetyl-Al, with 17 times higher sensitivity than the Hypercarb column. With regard to the optimization of the mobile phase, acetonitrile/water showed better elution performance than methanol/water on the BEH amide column. Moreover, to keep the pH value and the retention time of fosetyl anion constant throughout the run, the buffer concentration was further optimized with gradient elution. Referring to previous research on polar pesticides [34], mobile phase A was chosen as water with 5.0 mmol/L ammonium formate and B was acetonitrile. The gradient started at 90% of phase B to elute non-polar compounds while fosetyl-Al was still retained. Afterward, phase B decreased linearly to 50% over 3 min and the polar compounds eluted gradually. Then, the mobile phase composition returned to the initial condition in 0.1 min and was held for 1.9 min for re-equilibration. The separation was operated at a flow rate of 0.3 mL/min and the column temperature was kept at 35 ◦C. The total chromatographic run time was 5 min. The retention time of fosetyl-Al was 1.86 min under this condition.

**Figure 2.** Performance comparison between three types of columns.

#### *3.3. Method Validation*

As listed in Table 1, the linear range of the developed method covered from 10 to 2000 ng/g with eight concentration levels in the matrix-matched standards. Simultaneously, correlation coefficient (r2) of the matrix-matched calibration curve equaled 0.9999, which is very satisfactory for accurate quantification. The LOD and LOQ of the developed method were 5 and 10 ng/g, respectively, using the signal-to-noise ratio of the qualifier transition signal. This is sufficiently low to meet the maximum residue limits for many regulations, including the European Commission, China, and the USA. The proposed method showed satisfactory accuracy with recoveries of 95.6%, 105.2%, and 104.6%, respectively. Method precision was evaluated by the RSDs of five repetitions. The result was lower than 6.2%, indicating that this method is quite fit for routine analysis. The ME was calculated by comparing the slopes of standards prepared in wheat flour extract and water/acetonitrile

solvent. The ME evaluation (88.33%) showed that the residual matrix could suppress the fosetyl-Al signal. Therefore, a matrix-matched calibration solution was used for accurate quantitation during the real sample analysis.

**Linear Range (ng/g) Linearity (r2) LOD (ng/g) ME RSDs% (***n* **= 5) Recovery 10 ng/g 50 ng/g 100 ng/g**

**Table 1.** Validation data of the developed method for the detection of fosetyl-Al.

From the above, a simple and sensitive extract-dilute-shoot HILIC-MS/MS method was established. This proposed method is easy to operate and has excellent sensitivity (Table 2), which is promising for the analysis of fosetyl-Al in wheat flour samples.

10–2000 0.9999 5 88.33% 6.2 95.6% 105.2% 104.6%


**Table 2.** Method comparison of fosetyl-Al in different matrices.

<sup>1</sup> ng/mL.

#### *3.4. Real Sample Analysis*

In order to study the applicability of the proposed method, the developed method was applied to analyze the fosetyl-Al residual in commercial wheat flour samples. In total, 75 samples were collected from four provinces that covered China's main wheat flour production areas. For each kind of sample, the determination was repeated two times. Matrix-matched calibrations were injected in every sequence of samples in order to explore the carry-over effect and to ensure that a reagent blank was injected immediately after the highest standard. As a result, two samples were positive for fosetyl-Al, as shown in Figure 3. The signal intensity reflected that their concentrations were higher than the LOD and lower than the LOQ. One sample was produced in Shandong Province, the other was from Henan Province. All samples were below the maximum residue limits, suggesting that it is generally safe to consume wheat flour in China.

**Figure 3.** Chromatograms of the positive samples.

#### **4. Conclusions**

In this study, we developed and validated a method by combining an extract-diluteshoot strategy for sample preparation with HILIC-MS/MS for fosetyl-Al determination in wheat flour. This simple and fast dilution procedure effectively lowered the ME and improved the repeatability. Moreover, HILIC column provided both sufficient retention of fosetyl-Al and removal of the matrix, and MS/MS measurement demonstrated excellent sensitivity and selectivity. This approach offers simple operation, minimal consumption of chemicals, wide linear range, and high sensitivity. Satisfactory results were obtained as evidenced by matrix-matched standards. The observed LOQ was 10 ng/g in wheat flour. The satisfactory precisions and recoveries achieved with the spiked samples demonstrated the reliability and practicability of the developed method. The proposed method was applied to the quantitation of fosetyl-Al in 75 wheat flour samples, which overall showed low levels of contamination. These results showed that our proposed method is very promising for routine analysis and could find more applications in quality control of fosetyl-Al in high-carbohydrate food matrices.

**Author Contributions:** Conceptualization, methodology, validation, and writing—original draft preparation, X.L. (Xianjiang Li); data curation and writing—review and editing, S.W.; investigation, Z.G.; supervision, X.L. (Xiuqin Li); project administration, Q.Z.; funding acquisition, H.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Key R&D Program of China, grant number 2019YFC1604801.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Kaixuan Tong 1, Yujie Xie 1, Siqi Huang 2, Yongcheng Liu 2, Xingqiang Wu 1, Chunlin Fan 1, Hui Chen 1,\*, Meiling Lu <sup>3</sup> and Wenwen Wang <sup>3</sup>**


**Abstract:** Cottonseed hull is a livestock feed with large daily consumption. If pesticide residues exceed the standard, it is easy for them to be introduced into the human body through the food chain, with potential harm to consumer health. A method for multi-residue analysis of 237 pesticides and their metabolites in cottonseed hull was developed by gas-chromatography and liquidchromatography time-of-flight mass spectrometry (GC-QTOF/MS and LC-QTOF/MS). After being hydrated, a sample was extracted with 1% acetic acid in acetonitrile, then purified in a clean-up tube containing 400 mg MgSO4, 100 mg PSA, and 100 mg C18. The results showed that this method has a significant effect in removing co-extracts from the oily matrix. The screening detection limit (SDL) was in the range of 0.2–20 μg/kg, and the limit of quantification (LOQ) was in the range of 0.2–20 μg/kg. The recovery was verified at the spiked levels of 1-, 2-, and 10-times LOQ (n = 6), and the 237 pesticides were successfully verified. The percentages of pesticides with recovery in the range of 70–120% were 91.6%, 92.8%, and 94.5%, respectively, and the relative standard deviations (RSDs) of all pesticides were less than 20%. This method was successfully applied to the detection of real samples. Finally, this study effectively reduced the matrix effect of cottonseed hull, which provided necessary data support for the analysis of pesticide residues in oil crops.

**Keywords:** QuEChERS; gas-chromatography high resolution mass spectrometry; liquid-chromatography high resolution mass spectrometry; pesticide residues; cottonseed hull

#### **1. Introduction**

The composition of cottonseed hull is similar to that of soybean concentrate, with a high content of cellulose that can enhance the digestive systems of ruminants. Cottonseed hull has been widely used as an alternative feed for ruminants, due to its low price, easy availability, and excellent mixing performance [1–3]. The excessive and illegal use of pesticides during forage planting makes it easy for pesticides to enter the food chain and accumulate in animal adipose tissue [4], and human consumers may indirectly experience food safety problems through contact with livestock products. The composition of the oily matrix is complex: in addition to fat, it contains polysaccharides, proteins, pigments, and other substances. In the process of residue analysis, problems such as matrix enhancement, matrix inhibition, and retention-time shifts may occur in the detection of pesticides, which

**Citation:** Tong, K.; Xie, Y.; Huang, S.; Liu, Y.; Wu, X.; Fan, C.; Chen, H.; Lu, M.; Wang, W. QuEChERS Method Combined with Gas- and Liquid-Chromatography High Resolution Mass Spectrometry to Screen and Confirm 237 Pesticides and Metabolites in Cottonseed Hull. *Separations* **2022**, *9*, 91. https:// doi.org/10.3390/separations9040091

Academic Editor: Chiara Emilia Cordero

Received: 8 March 2022 Accepted: 31 March 2022 Published: 2 April 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

will hinder the detection of target compounds [5,6]. Therefore, it is urgent to develop a detection technique for the oily matrix to solve these problems.

The analysis of pesticide residue usually includes the following steps: (1) extraction of the target compound; (2) removal of interference from the extract; and (3) qualitative and quantitative detection of the target compound [4]. Lipophilic pesticides tend to be concentrated in fat. Improper pretreatment will affect the detection sensitivity, recovery, and sample throughput [7]. The current pretreatment methods for plant-derived oil substrates mainly include dispersion liquid-liquid micro-extraction (DLLME) [8], matrix solid phase dispersion (MSPD) [9,10], low temperature fat precipitation (LTFP) [11], solid phase extraction (SPE) [5], and QuEChERS [12–16]. The QuEChERS method requires fewer reagent consumables and short pretreatment time, so it is accepted by more and more experimenters [17]. Theurillat et al. established the QuEChERS method to determine the residues of various pesticides and verified the method for 176 pesticides in six oily matrices [12]. Rutkowska et al. investigated the matrix effect and recovery of four seed samples of cress, fennel, flax, and hemp. The final method verified 248 pesticides, and the LOQs reached 0.005 mg/kg [14]. Banerjee et al. used the QuEChERS method to analyze more than 220 pesticide residues in sesame seeds. This method can effectively reduce the interference of the matrix effect by freezing and degreasing at −80 ◦C and then purifying the oil.

The current trend of separation science is to develop new chromatographic mass spectrometry methods that can detect multiple compounds at the same time after a single injection, thereby reducing analysis time and cost [18]. The current detection technology for the detection of pesticide residues in oily matrices is mainly triple quadrupole mass spectrometry (MS/MS) [13,19–21]. The data was collected according to the specific nucleocytoplasmic ratio of the specified compound, but other compounds that were not in the list could not be identified. When analyzing a large number of compounds, the sensitivity and selectivity are limited. Due to their high resolution, precise mass accuracy, outstanding full-scan sensitivity, and complete mass spectrometry information, high-resolution mass spectrometry (HRMS), such as time-of-flight mass spectrometry (TOF/MS) and quadrupole Orbitrap mass spectrometry (Obitrap/MS), can be used without additional sample injection. Under retrospective analysis, with these advantages, HRMS has been widely used in the field of food analysis [22,23]. Lehotay et al. used GC-TOF to analyze 34 pesticides in flaxseed, dough, and peanuts [15]. Amadeo et al. used GC-QTOF to verify 166 pesticide residues in avocados and almonds [24].

To ensure the safety of livestock feed and to prevent pesticide residues from being introduced into the human body through the food chain, this work established a QuEChERS multi-residue analysis method, and used GC- and LC-QTOF/MS techniques to verify 237 pesticides in cottonseed hull. By optimizing the hydration volume, extraction solvent, salting-out agent, and clean-up sorbents, the influence of the matrix effect was reduced and the pesticide recovery was optimized. Finally, this method was successfully applied to the analysis of actual samples, providing data support for the risk of pesticide residues in oily substrate monitoring.

#### **2. Materials and Methods**

#### *2.1. Chemicals and Reagents*

Pesticide standards (purity ≥ 98%) were obtained from Tianjin Alta Scientific (Tianjin, China). Sodium chloride, magnesium sulfate, and sodium sulfate (analytical purity) were obtained from Tianjin Fuchen Chemical Reagent Ltd. (Tianjin, China). Primary secondary amine (PSA) and C18 were purchased from Agilent Technologies (Santa Clara, CA, USA). Methanol, acetonitrile, and toluene (chromatographic purity) were obtained from Anpel Laboratory Technology (Shanghai, China). Formic acid and ammonium acetate (mass spectrometry grade) were obtained from Honeywell (Muskegon, MI, USA).

#### *2.2. Apparatus*

HPLC-QTOF/MS Agilent 1290 and Agilent 6550 equipped with Agilent Dual Jet Stream ESI and GC-QTOF/MS Agilent 7890B and Agilent 7200 were obtained from Agilent Technologies (Santa Clara, CA, USA). A Milli-QTM Ultrapure Water System was obtained from Millipore (Milford, MA, USA). An N-EVAP112 Nitrogen Blowing Concentrator was obtained from Organomation Associates (Worcester, MA, USA). An AH-30 Automatic homogenizer was obtained from RayKol Group Corp., Ltd. (Xiamen, China). An MS204S Electronic Analytical Balance was obtained from Mettler Toledo (Shanghai, China).

#### *2.3. Standard Solution*

Ten mg of the standard substance was accurately weighed into a 10 mL brown volumetric flask. a suitable reagent was selected according to the solubility of the compound in the organic reagent. It was dissolved by ultrasound and diluted to the mark to a standard solution of 1 mg/L. The standard solution was stored at −18 ◦C in the dark. As needed, a pipette with an appropriate amount of the standard stock solution was diluted with methanol to prepare a working solution of appropriate concentration, and stored at 4 ◦C in the dark.

#### *2.4. Sample Preparation Method*

Based on other oily matrix sample preparation methods [12,16], a modified QuEChERS method was used for the detection of cottonseed hull. Two g (accurate to ±0.01 g) of sample were transferred into a 50 mL centrifuge tube; 2 mL of ultrapure water were added for hydration and then extracted with 10 mL of 1% acetic acid in acetonitrile. The homogenizer was used to homogenize the sample for 1 min at 13,500× *g*; then, 4 g MgSO4, 1 g NaCl and a ceramic homoproton were added. The mixture was shaken for 10 min and centrifuged at 3155× *g* for 5 min; then, 3 mL of supernatant was transferred to a clean-up tube containing 400 mg MgSO4, 100 mg PSA, and 100 mg C18. After shaking for 10 min and being centrifuged at 3155× *g* for 5 min, 1 mL of supernatant was dried under nitrogen, then ultrasonically redissolved with ethyl acetate containing internal heptachlor-exo-epoxide for GC-QTOF/MS analysis, and ultrasonically redissolved with acetonitrile aqueous solution (2:3, *v*/*v*) containing internal standard atrazine D5 for LC-QTOF/MS analysis.

#### *2.5. Instrument Parameters*

The instrument parameters of LC-QTOF/MS and GC-QTOF/MS were configurated according to a previous paper published by our laboratory [25].

An LC-QTOF/MS: ZORBAX SB-C18 column (100 mm × 2.1 mm, 3.5 μm, Agilent Technologies) was used for separation at 40 ◦C; 5 mmol/L ammonium acetate with 0.1% (*v*/*v*) formic acid aqueous solution and acetonitrile were applied as phase A and phase B. The flow rate was set at 0.4 mL/min. The gradient program was set as follows: 0 min, 1% B; 3 min, 30% B; 6 min, 40% B; 9 min, 40% B; 15 min, 60% B; 19 min, 90% B; 23 min, 90% B; 23.01 min, 1% B. The equilibrium time was 4 min. The injection volume was 5 μL.

The Agilent Dual Jet Stream (AJS) ESI source (Agilent Technologies) was set in positive full scan (m/z 50–1000) mode; the capillary voltage was 4 kV; nitrogen was used as the nebulizer gas at 0.14 MPa; the sheath gas temperature was set at 375 ◦C with 11.0 L/min; the drying gas flow rate was 12.0 L/min; the drying gas temperature was 225 ◦C; the fragmentation voltage was 345 V. In all ions Mass/Mass mode, the collision energy was 0 V at 0 min, and 0, 15, and 35 V at 0.5 min, respectively. The total program duration was 27.01 min.

GC-QTOF/MS: HP-5 MS UI (30 m × 0.25 mm, 0.25 μm, Agilent Technologies) was used for separation at 40 ◦C. The oven temperature gradient was started at 40 ◦C for 1 min, increased at 30 ◦C/min to 130 ◦C, heated at 5 ◦C/min to 250 ◦C, ramped to 300 ◦C at 10 ◦C/min, and maintained for 7 min. Helium (purity > 99.999%) was used as the carrier gas with a constant flow rate of 1.2 mL/min. The injection temperature was set to 270 ◦C

and the injection volume was 1 μL. The injection mode was not split injection, and the purge valve was opened after 1 min.

The ion source was an electronic ionization source (70 eV, 280 ◦C), and the temperatures of the transfer line and the quadrupole were 250 ◦C and 180 ◦C, respectively. Solvent delay was set to 3 min; the ion monitoring mode was full scan; scanning ranged (m/z) from 45 to 550; the scan rate was 5 Hz. The total program duration was 42 min.

Mass calibration was required before sample acquisition, and the instrument was tuned at intervals to ensure stability.

#### *2.6. Method Validation*

The screening method of high-resolution mass spectrometry can be validated through screening detection limits (SDL), and the quantitative method can be validated through limit of quantitation (LOQ). The SDL, LOQ, linearity, recovery, and precision of this experiment were verified by SANTE/12682/2019 guidelines. SDL is the minimum concentration at which more than 95% of a series of concentration levels meets the detection requirements (20 additional experiments were conducted in parallel for each concentration). When the SDL and recovery were validated, all the target pesticides were spiked to the sample and the spiked samples were placed at room temperature for 30 min, then treated according to the above method. After the 10-point matrix matching calibration was constructed, its linearity was evaluated with the coefficient of determination (R2). The recovery and precision were investigated in three different levels of spiked blank samples with 1-, 2-, and 10-times LOQ.

The matrix effect (ME) is the interference of other components in the matrix with the target compounds. The formula is:

$$\text{ME} \left( \% \right) = \left( \text{bm} - \text{bs} \right) / \text{bs} \times 100\% \tag{1}$$

where bm is the slope of the matrix standard curve and bs is the slope of the solvent standard curve.

Based on previous studies, we established several hundred kinds of pesticide databases on gas and liquid high resolution mass spectrometry, respectively [25]. According to the recovery and precision, 237 pesticides were divided into pesticides suitable for GC or LC detection.

#### **3. Results**

#### *3.1. Optimization of Hydration Volume*

For the oily matrix, adding an appropriate amount of water for hydration during sample pretreatment was conducive to the softening of the matrix epidermis, making it easier for residual pesticides in the matrix to be extracted. This experiment explored the effect of different hydration volumes on the recovery of multiple pesticides. The experiment results show that the proportion of pesticides that met the recovery requirements (70–120%) under a non-hydration condition was 74.9%, which was less than under the conditions with water additions of 2 mL and 5 mL. Under the condition of a 2 mL water addition, the number of pesticides meeting the recovery requirements was the most numerous, accounting for 83.5%. As shown in Figure 1, the average recovery under the 2 mL condition was 88.3%, which was higher than that under the other two conditions. The results were in line with our expectations. The oil-water partition coefficient (logP) is an important parameter for the solubility of compounds, which is a simulated value based on the soil sorption coefficient normalized to organic carbon content (log Koc) [26]. The smaller the logP value, the better the water solubility of the compound. The effect of hydration volume on recovery with different logP was investigated, showing that hydration had a great impact on recovery with a low logP. The overall recovery of 54 pesticides with hydrophilic compounds (logP < 2.0) was low under a non-hydration condition, with the pesticides meeting the requirements accounting for 42.6%. When the hydration volume was 5 mL, the pores were opened due to the increase in the hydration volume, and multiple interferents

in the matrix could be extracted together. The matrix promotion effect was enhanced, so that the overall recovery of pesticides with logP < 2.0 was higher than the recovery under the other two conditions. When the hydration volume was 2 mL, the pesticides that met the requirements of recovery were most numerous, accounting for 70.4%; therefore, 2 mL was finally selected as the optimal hydration volume.

**Figure 1.** Effects of hydration volumes on pesticide recovery.

#### *3.2. Optimization of Extraction Solvent Volume*

The extraction of target compounds is a critical step in pesticide residue analysis. Mol et al. [27] tested a series of solvents for extraction and found that methanol usually extracts too many compounds in the matrix, and further matrix removal steps were required. Acetonitrile has low solubility in fat and a low matrix effect when extracting from complex matrices. Therefore, acetonitrile was selected as the extraction solvent of cottonseed hull in this experiment. Three different extraction volumes of 10 mL, 16 mL, and 20 mL (i.e., a hydration volume and extraction volume ratio of 1:5, 1:8, and 1:10) were compared to explore the effect of different extraction volumes on the recovery of pesticide residues. The results are shown in Figure 2. It was found that when the extract volume was 10 mL, 16 mL, and 20 mL, the proportion of pesticides meeting the recovery requirements was similar, at 81.0%, 80.7% and 81.3% respectively. However, at the spiked level, the volume of the extraction solution decreased, the pesticide concentration per unit volume increased, and more pesticide compounds had better peak shapes. In addition, a lower organic reagent amount was recommended from the perspective of green environmental protection, so the final extraction volume was 10 mL.

#### *3.3. Optimization of Salting-Out Agent*

The salting-out agents commonly used in pesticide residue screening were EN buffer salt (4 g MgSO4, 1 g NaCl, 0.5 g disodium hydrogen citrate, and 1 g sodium citrate), the QuEChERS method for fruits and vegetables (4 g MgSO4 and 1 g NaCl), and AOAC buffer salt (6 g MgSO4 and 1.5 g NaAc). In this work, the effects of the above three salting-out agents on the recovery of pesticides were compared. As shown in Figure 3, although EN or AOAC salt forms a buffer system in the solution state, the results showed that the recovery using an MgSO4 + NaCl combination best met the requirements, accounting for 78%. The reason for this result was that the volume of the extract from the QuEChERS method was relatively small. If the amount of extraction salt was too large, the heat emitted during water absorption destroys the structure of thermally unstable pesticides and affects their recovery. Therefore, 4 g MgSO4 and 1 g NaCl with less salt consumption were finally selected as the salting-out agents.

**Figure 2.** Effect of extraction solvent volume on pesticide recovery.

**Figure 3.** Effect of salting-out agents on pesticide recovery.

#### *3.4. Optimization of Types and Amounts of Clean-Up Sorbents*

A clean-up procedure was a key step in the pretreatment of the oily matrix. Its purpose was to effectively purify the analyzed matrix, and most of target pesticides had acceptable recovery, precision, and matrix effect [14]. Although acetonitrile had low liposolubility, which can slightly reduce the interference of a fat-soluble matrix on target compounds [15], in order to effectively reduce the influence of high-fat matrix co-extraction on the detection sensitivity of pesticides, as well as instrument loss, the clean-up procedure was necessary. Theurillat established a d-SPE clean-up method containing 150 mg C18 and 150 mg PSA to determine 176 pesticide residues in fatty foods [12]. Therefore, this study was optimized on this basis.

In this work, the ability of MgSO4 + PSA + C18 + Z-sep and MgSO4 + PSA + C18 sorbents were compared. The structure of PSA had -NH2, which can form a strong hydrogen bond with -COOH, so it was often used to adsorb polar compounds, such as fatty acids, lipids, and carbohydrates. C18 was often used to adsorb non-polar compounds, such as long-chain aliphatic compounds and sterols [8,25]. Z-sep was a new adsorbent, based on zirconia, which can be used for the adsorption of hydrophobic compounds in the fat matrix [28]. It was seen that the bottom of the purification tube after Z-sep purification was dark yellow, while the sample without Z-sep purification was light yellow, indicating that Z-sep had an obvious effect on degreasing.

In order to further verify the ability of sorbents, the spiked experiments were carried out. As shown in Figure 4, A was the sorbent combination of MgSO4 + PSA + C18 + Z-sep, and B was the sorbent combination of MgSO4 + PSA + C18. As a result, the sorbent combination without Z-sep accounted for more pesticides that meet the requirements, reaching 81.04%. The reason for this result was that Z-sep adsorbs some target pesticides while removing lipids. According to the Lewis theory, the affinities of Z-sep on the analyte containing different substituent characteristics can be sorted in the following order: chloride < formate < acetate < sulphate< citrate < fluoride < phosphate < hydroxide [25]. In this work, a variety of pesticides, such as trinexapac-ethyl, abamectin containing -OH, fenamiphos sulfoxide containing phosphate, and sulfoxaflor containing sulphate, had substituents with a strong affinity to Z-sep. Therefore, the recovery of sorbent combinations with Z-sep was significantly lower than that without Z-sep. Although Z-sep was more efficient in removing lipid compounds, the sorbent combination of MgSO4 + PSA + C18 was finally selected as the purification filler in this work, from the perspective of method versatility.

**Figure 4.** Effect of clean-up sorbents on pesticide recovery. (**A**) MgSO4 + PSA + C18 + Z-sep; (**B**) MgSO4 + PSA + C18.

The amount of PSA and C18 was also optimized. The effects of PSA (50–150 mg) and C18 (100–300 mg) on the recovery of various pesticides were optimized by controlling other variables. The results showed that when the amount of PSA was 100 mg, the greatest number of pesticides with satisfactory recovery was obtained, accounting for 73.7%. With the increase in PSA amount, the recovery of organic nitrogen pesticides, such as propanil and fenbuconazole, and carbamate pesticides, such as aldicarb-sulfone and thiophanatemethyl, gradually decreased. When the amount of C18 was 100 mg, the proportion of pesticides that met satisfactory recovery was 82.0%. With an increase in the C18 amount, the recovery of various organic nitrogen pesticides obviously decreased, especially the chlorides with a benzene ring structure, such as monolinuron, novaluron, propanil, and pretilachlor. Therefore, 100 mg PSA and 100 mg C18 were finally selected as the optimal amounts of clean-up sorbents.

#### *3.5. Evaluation of Matrix Effect*

Analysis of pesticide residues in the oil matrix may be adversely affected by the matrix effect. The main result of the matrix effect is to increase or decrease the analyte signal when the same analyte exists in the solvent [29]. The methods for eliminating or reducing the matrix effect include: (1) optimizing the sample preparation method and reducing co-extraction; (2) changing the chromatographic mass spectrometry conditions; (3) diluting the samples; and (4) using matrix-matched standards or an additional standard method [30]. In this work, the purifying agent was optimized, and the matrix-matched standard was used to reduce the interference of the matrix effect on target compounds. The matrix effect distribution of 237 pesticides is shown in Figure 5. Among the 237 pesticides investigated in cottonseed hull samples, the proportion of pesticides with a negative matrix effect accounted for 81.4%, indicating that the substrate had a suppression effect on the tested pesticides as a whole. The matrix effect can be divided into three categories: no matrix effect (|ME| ≤ 20%); a weak matrix effect (20% < |ME| < 50%); and a strong matrix effect (|ME| ≥ 50%). In this work, only 8% of the pesticides in the cottonseed hull matrix showed a strong matrix effect; the weak matrix effect and no matrix effect accounted for 13.1% and 78.9%, respectively, indicating that this research method had a strong anti-matrix interference ability.

**Figure 5.** Matrix effect distribution of 237 pesticides.

#### *3.6. Method Validation and Method Performance* 3.6.1. SDL, LOQ, and Standard Curve

The method validation was carried out under the optimal sample preparation procedure, and the results are shown in Table 1. The typical extraction ion chromatograms of GC-Q TOF/MS and LC-Q TOF/MS are shown in Figures 6 and 7, respectively. The SDLs were in the range of 0.2–20 μg/kg, of which 224 pesticides (accounting for 94.5%) were in the range of 0.2–5 μg/kg. The LOQs were in the range of 0.2–20 μg/kg; 215 pesticides (accounting for 90.7%) had an LOQ range of 0.2–5 μg/kg. Shinde developed and verified 222 and 220 multi-pesticides residue analysis methods in sesame seeds, using LC-MS/MS and GC-MS/MS, respectively, and most pesticides offered an LOQ of 10 μg/kg for most compounds [16]. Kuzukiran et al. developed an SPE sample preparation method, combined with GC-MS, GC-MS/MS and LC-MS/MS, to analyze the residues of 322 organic pollutants in bats [31]. The LOQ of the method was in the range of 0.27–19.26 μg/kg, which was similar to that in our work; however, they paid more attention to environmental pollutants. This indicated that this method had high sensitivity in the detection of pesticide residues in cottonseed hull matrix. It is noteworthy that due to the large number of pesticides spiked, the retention time of some pesticides may overlap or be very close; for example, the RTs of Chloridazon and Mevinphos were 3.62 min. However, the excellent resolution of high-resolution mass spectrometry was sufficient to separate compounds that had a similar RT but a different mass (the quantitative ion mass of Chloridazon and that of Mevinphos were 222.04287 and 225.05230, respectively).

**Figure 6.** Overlay extraction ion chromatograms of GC-Q TOF/MS of cottonseed hull sample at spiking level of 200 μg/kg.

*Separations* **2022**, *9*, 91




**Table 1.** *Cont*.


**Table 1.** *Cont*.












**Table 1.** *Cont*.


**Figure 7.** Overlay extraction ion chromatograms of LC-Q TOF/MS of cottonseed hull sample at spiking level of 200 μg/kg.

The calibration curve was plotted using the matrix matching calibration method and the target analytes at 10 spiked levels (0.2, 0.5, 1, 2, 5, 10, 20, 50, 100, and 200 μg/kg) were spiked to the blank cottonseed hull sample. The linear ranges of 237 pesticide analytes were 1–200 μg/L. All target pesticides showed good linearity in the concentration range, and R2 was greater than 0.99, indicating that this method could meet the requirements of quantitative analysis.

#### 3.6.2. Recovery and Precision

The recovery and precision of the method was evaluated by spiked standard solutions at the levels of 1-, 2-, and 10-times LOQ for the cottonseed hull samples with six parallels at each spiked level. The results are shown in Figure 8. At the levels of 1-, 2-, and 10-times LOQ, the recoveries of the 237 pesticides in the range of 70–120% were 91.6%, 92.8%, and 94.5%, respectively, and the RSD of all the pesticides was less than 20%, indicating that the method had satisfactory recovery and precision.

Among the 237 pesticides, 60 pesticides were detected by two detection techniques, and most of them showed similar performance; however, individual pesticides were different in the two techniques. For example, the average recovery (81.2%) of clodinafoppropargyl detected by GC-QTOF/MS was lower than that (95.7%) detected by LC-QTOF/MS. In terms of precision, the RSD (10.8%) of the compound detected by GC-QTOF/MS was higher than that (4.8%) detected by LC-QTOF/MS. For Propiconazole, the average recovery and RSD of GC-QTOF/MS (89.0%, 5.5%) were better than those of LC-QTOF/MS (80.0%, 6.4%). Therefore, appropriate detection techniques should be selected in pesticide residue analysis, especially when compounds are suitable for these two detection techniques.

**Figure 8.** The recovery and RSD of the target pesticides at three spiked levels.

#### *3.7. Analysis of Real Samples*

The established method was applied to the analysis of 11 real cottonseed hull samples collected from several domestic pastures. The results showed that three pesticide residues were found in 11 cottonseed hull samples (butylate (three times), fenbuconazole (three times), and Diuron (two times)), with concentrations ranging from 10 to 28 μg/kg and above the LOQ. The determined three pesticides were slightly hazardous, according to WHO [32]. This method can be used for high-throughput trace detection of pesticide residues in cottonseed hull samples and improve the ability of risk-screening.

#### **4. Conclusions**

In this work, GC-QTOF/MS and LC-QTOF/MS were used to develop a high throughput method for qualitative screening and quantitative analysis of 237 pesticides in the cottonseed hull matrix. The modified QuEChERS extraction process seems to effectively eliminate the interference caused by the oily matrix, and the SDL, LOQ, recovery, and precision of the analysis method were verified under optimal conditions. In addition, compared with other methods for the oily matrix, this method has the advantages of being fast and simple, with high throughput and low solvent consumption. The results showed that the developed method could be applied to the screening of pesticide residues in the cottonseed hull matrix, effectively and generally.

**Author Contributions:** Conceptualization, H.C. and C.F.; methodology, H.C.; validation, K.T., Y.X. and X.W.; investigation, S.H. and Y.L.; resources, K.T.; data curation, Y.X.; writing—original draft preparation, K.T.; writing—review and editing, H.C., X.W. and C.F.; supervision, M.L. and W.W.; project administration, H.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financially supported by the Science and Technology Project of the State Administration for Market Regulation (2021MK165).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


#### *Article*
