Simultaneous Determination of Neonicotinoid Insecticides and Metabolites Residues in Milk and Infant Formula Milk Powder by EMR-Liquid Chromatography-Tandem Mass Spectrometry

An analytical method based on enhanced matrix removal–lipid liquid chromatography–tandem mass spectrometry (EMR-LC–MS/MS) was developed for the determination of neonicotinoid insecticides and metabolites residues (imidacloprid (IMI) and its metabolites imidacloprid-urea (IMI-U), imidacloprid-olefin (IMI-O), acetamiprid (ACE) and its metabolite N-desmethyl acetamiprid (IM 2–1), dinotefuran (DIN) and its metabolite [1-methy1-3(tetrahydro-3-furylmethy1) urea (DIN-UF), thiacloprid (THIA), thiamethoxam (TMX), clothianidin (CLO, metabolite of thiamethoxam), and flupyradifurone (FLU)) in milk and infant formula milk powder. The residual of neonicotinoid insecticides and their metabolites in samples were exacted by acetonitrile and extraction kits. The quantitative detection was performed by LC–MS/MS with multiple reaction monitoring (MRM) modes under positive ion electrospray ionization (ESI+). The isotope dilution internal standard or external standard method was used for quantitation. The limits of quantification (LOQs, S/N = 10) were 2 μg/kg (IMI, IMI-U, ACE, IM 2–1, DIN-UF, THIA, and TMX) and 5 μg/kg (IMI-O, DIN, CLO, and FLU) for milk; 2 μg/kg (ACE), 15 μg/kg (THIA, IM 2–1, DIN-UF, THIA, and TMX), and 40 μg/kg (IMI-U, IMI-O, DIN, CLO, and FLU) for infant formula milk powder. At three spiked levels of 5 μg/kg, 10 μg/kg, 50 μg/kg (milk), or 40 μg/kg, 80 μg/kg, 400 μg/kg (infant formula milk powder), the recoveries were in the range of 71.7–108.7% and 71.9–107.1%; the relative standard deviations were below 12.6% and 13.9%, respectively. This method was simple, rapid, and accurate to determine the neonicotinoids and their metabolites residues in milk and infant formula milk powder.


Introduction
Milk is an indispensable food in the human diet for its rich nutrition, easy digestion and absorption, good quality, low price, and convenient consumption, and it is also the main raw material for infant formula milk powder. Therefore, its quality and safety (e.g., pesticides or veterinary drugs residues) directly affects the health of consumers and infants. Neonicotinoids (NEOs) have been widely used for controlling pests in crops. They worked by blocking the normal conduction of the central nervous system in insects, leading to paralysis or death (Matsuda et al. 2001;Chagnon et al. 2015). Studies have shown that NEOs can affect animals through the food chain, such as by reducing the population of bees (Lu et al. 2012;Laurino et al. 2013). NEOs have also been detected in the bodies of honeybees and in honey (Kaczynski et al. 2017;Mitchell et al. 2017). Along with the transmission of the food chain, NEOs and their metabolites have been found not only in mammals (Ozsahin et al. 2014;Berheim et al. 2019) but also in human body fluids (e.g., urine, breast milk) (Osaka et al. 2016;Ueyama et al. 2020;Chen et al. 2020). To protect the quality of milk and the health of consumers, the maximum residue limits (MRLs) for some neonicotinoids and their metabolites in milk have been clearly defined in the United States of America (USA), China, European Union (EU), and Japan, as shown in Table 1

3
Due to the insecticidal properties of NEOs, the determination of NEOs residues is currently centered on plant-origin products such as vegetables, fruits, and grains (Xie et al. 2011;Wang et al. 2012;Vichapong et al. 2013;Pastor-Belda et al. 2016), while animal origin foods are mainly honey, royal jelly, and other bee products (Tanner and Czerwenka 2011;Giroud et al. 2013;Yá˜nez et al. 2013;Hou et al. 2019). The detection methods include high-performance liquid chromatography (HPLC) (Jovanov et al. 2015;Vichapong et al. 2015Vichapong et al. , 2016, liquid chromatography-tandem mass spectrometry (LC-MS/MS) (Anand et al. 2018;Zhang et al. 2018a, b;Cui et al. 2021), liquid chromatography-high-resolution mass spectrometry (LC-HRMS) Zhao et al. 2020), electrochemical detection (ECD) (Papp et al. 2010;Guzsvány et al. 2011;Brycht et al. 2012), capillary liquid chromatography (CLC) (Carbonell-Rozas et al. 2020a, and micellar electrokinetic chromatography (MEC) (Carbonell-Rozas et al. 2020b). It should be noted that the determination of NEOs residues in other foods of animal origin (e.g., milk, meat, and aquatic products) has been less studied (Alaa et al. 2010;Xiao et al. 2011;Xiao et al. 2013;Craddock et al. 2019). Enhanced matrix removal-lipid (EMR) is a technique that is based on hydrophobic interaction and size exclusion. It can be used to remove fatty acids, phospholipids, triglycerides, and other compounds with long-chain aliphatic functional groups from extracts. However, it does not retain analytes of interest (DeAtley et al. 2015). EMR has been used to determine the presence of NEOs in honeybee and wild boar (Sus scrofa L) matrix (Kaczynski et al. 2017(Kaczynski et al. , 2021, as well as for the detection of drug or pollutant residues in milk, infant formula powder, and chicken eggs (Zhang et al. 2018a, b, c;Luo et al. 2020;Zhang et al. 2021).

Sample Preparation
5.00 g sample (12.5 g infant formula milk powder was diluted with 87.5 g water in advance) into a 50 mL centrifuge tube, adding isotope internal standard (50 μL, 1000 ng/mL), 15 mL of ACN and QuEChERS extraction kit to the tube. The mixture was shaken by vortex for 10 min at room temperature, followed by centrifugation for 5 min at 8500 rpm. Transfer the supernatant to a new 50 mL centrifuge tube, and add ACN to bring the volume up to 20 mL, mixing the solution. Take 4 mL of the extract solution and transfer it to a 15 mL centrifuge tube, adding 1 mL of water and mixing. Transfer the solution to an EMR purification cartridge and collect the eluate, then elute the cartridge with 2 mL of ACN:H2O (20:80, vol/vol) and 3 mL of ACN, collecting all of the eluates. Evaporated the eluate to dryness using a rotary evaporator with a water bath at 40 ℃. The dried extract was reconstituted with 2 mL of MeOH:0.15% formic acid solution (10:90, vol/vol), vortex mixed for 60 s. Filter the solution through a 0.2 μm nylon membrane, take 0.1 mL of the filtered solution, and add 0.9 mL of MeOH:0.15% formic acid solution (10:90, vol/vol), mixing using a vortex for 60 s, and then were used for LC-MS/MS analysis.

Analytical Conditions of LC-MS/MS
The LC-MS/MS was carried out on liquid chromatography-tandem mass spectrometer of 1260-6495 (Agilent Technologies, Germany) with an ESI source. A Zorbax Eclipse XDB-C 18 column (150 mm × 4.6 mm i.d., 5 μm particle, Agilent) was used for chromatographic separation at a flow rate of 0.4 mL/min. The column temperature was held at 40 °C, and the injection volume was 10 μL. Mobile phase A was 0.15% formic acid (with 5 mM ammonium acetate), and mobile phase B was MeOH. The gradient program of mobile phase B was set as follows: 0 min, 10% B; 0-6.0 min, 10-70% B; 6.0  ionization (ESI + ). The operational conditions were as follows: capillary voltage, 3000 V; ion source temperature, 150 ℃; drying gas (nitrogen) flow rate, 14 L/min; sheath temperature, 350 ℃; sheath gas (nitrogen) flow rate, 10 L/min. The collision energy of each compound was optimized by flow injection analysis. The transition information and optimized parameters for each compound were listed in Table 2.

Optimization of LC-MS/MS Conditions
As shown in Fig. 2, the precursor ions and the major fragment information were monitored in positive by using flow injection for the standard solution of the compound to be measured at a concentration of 1.0 μg/mL. According to the European Union (EU) Directive 2021/808 (European Union 2021), the quantitative confirmation by LC-MS/MS must meet 5 identification points (1 point for chromatographic separation, 1 point for the precursor ion, and 1.5 points for one production). We selected two characteristic ion pairs, in which the ion pair with a high signal-to-noise ratio, good peak shape, and low interference was used as the quantitative ion pair. The ion information as well as the optimized collision energy parameters were shown in Table 2.

Optimization of Liquid Chromatography Conditions
According to references (Hou et al. 2019), the amino column and C 18 column (4.6 × 100 mm, 5 μm, of both sizes) were selected and compared in four separation systems: MeOH -0.15% formic acid solution (containing 5 mM ammonium acetate), MeOH -0.15% formic acid solution, ACN -0.15% formic acid solution (containing 5 mM ammonium acetate), and ACN -0.15% formic acid solution. The result showed that the separation effect and the signal intensity of the target compounds were better  IMI, IMI-U,  IMI-O, ACE, IM 2-1, DIN,  DIN-UF, THIA, TMX, CLO,  and FLU on the C 18 column than on the amino column, the signal response of THIA being more than 10 times higher, and IMI-U, ACE, IM 2-1, and DIN-UF being 3 to 6 times higher. For the C 18 column under different conditions, the result showed that the signal response of each compound was significantly higher when MeOH was in the organic phase, and the mobile phase contained ammonium acetate; therefore, MeOH -0.15% formic acid solution (containing 5 mM ammonium acetate) was finally used as the mobile phase for separation experimenting. Under this separation system, the MRM chromatograms of LC-MS/MS of milk spiked with each compound (spiked level: 5 μg/kg) were shown in Fig. 3.

Optimization of Extraction
The extraction process parameters were optimized to obtain the optimal extraction efficiency and remove proteins and lipids from milk or infant formula milk powder to reduce the matrix effects. The sample was extracted with several different solvents, such as MeOH, ACN, and trichloroacetic acid solution (50%); among the MeOH and ACN experiments, extraction kits (containing sodium chloride, anhydrous magnesium sulfate, sodium citrate, and sodium hydrogencitrate sesquinydrat) were added to improve the efficiency of extraction and cleanup. The results were shown in Fig. 4; when the trichloroacetic acid solution was added, the recoveries of IMI (61.1%) and IMI-U (63.4%) were lower than 65%; the recovery of DIN (157.8%, MS/ MS transition of 203.1/129.1) would have significant matrix effects in the test. When MeOH was used as an extracted solution and cleanup with EMR, the recoveries of IMI-U (62.2%), ACE (46.9%), IM 2-1 (39.4%), DIN UF (54.2%), THIA (40.5%), and CLO (58.1%) were all lower than 60%, and when ACE was used as an extracted solution, the recoveries of each compound were higher than 80%. Therefore, ACN was used to extract and precipitate protein directly.

Optimization of Cleanup Method
The cleanup procedure was optimized by a spiked standard solution. Both C 18 , HLB, and EMR were compared; the results were shown in Fig. 5

Matrix Effect Evaluation
Matrix effects were evaluated by a modified version of the equation described by Stahnke et al. (2012). Matrix effects (%) = 100% × [peak area (postextracted spiked sample) − peak area (solvent standard)]/peak area (solvent standard), where peak area (postextracted spiked sample) is the analyte spiked into extracted matrix after the extraction and cleanup procedure. The peak area of the solvent standard is the same concentration of analyte in the solvent solution, and this solvent solution was the reconstitution solution used for the postextracted spiked sample. So, the negative result indicates suppression, and the positive result indicates enhancement of the analyte signal in the postextracted spiked sample. The results were shown in Table 3; IMI-O has a significant matrix enhancement effect in milk or infant formula milk powder with matrix effects higher than 50%, while the matrix effects of other compounds were lower than 10%. In order to reduce the matrix effects, the experiment finally used the addition of isotope-labeled internal standards and matrix-matched solution calibration curves for quantitative determination.

Assay Specificity
The specificity was evaluated by the analysis of 20 blank milks and 20 infant formula milk powers. No interfering peaks from endogenous compounds were found in the retention time of the target analyte for samples.

Accuracy and Precision
The method accuracy was evaluated by analyzing a series of spiked samples following the developed method. Three different concentrations (high, medium, and low) of the standard target compounds were spiked into the "blank" samples and then treated following the optimized experimental procedure, analyzed by LC-MS/MS. The recoveries were 71.7-108.7% (milk) and 71.9-107.1% (infant formula milk powder). To evaluate the precision of this method, both intra-day and inter-day repeatability were examined by determination of milk and infant formula milk powder sample at three different concentrations. Good stability and satisfactory repeatability were achieved, the relative standard deviations (RSDs) values of milk were below 11.9% and 12.6%, infant formula milk powder was below 12.2% and 13.9% for intra-day and inter-day analyses, respectively.   . 6 The LC-MS/MS MRM chromatograms of ACE in spiked milk (5 μg/kg) and real sample was at a dilution ratio of 1:8 with warm water and then cooled to room temperature). The results were shown in Fig. 6; one milk sample contained residues of ACE (45.4 µg/ kg), and no neonicotinoids and their metabolites were detected in any of the infant formula milk powder samples.

Conclusions
In the present study, the method of simultaneous determination of seven neonicotinoids and four metabolites in milk and infant formula milk powder samples by EMR-LC-MS/ MS was established. Both isotope-labeled internal standards and matrix-matched calibration standards were used to alleviate the matrix effects. Good recoveries (71.7-108.7%) and precision (RSDs below 13.9%) were obtained; the results indicate that the developed method was simple and rapid and the LOQ meets the current requirements of the maximum residue limits of relevant compounds in the USA, China, EU, and Japan, which can be applied for the simultaneous determination of neonicotinoids and their metabolites in milk or infant formula milk powder. Data Availability All data generated or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations
Ethical Approval This article does not contain any studies with human participants or animals performed by any of the authors.

Consent to Participate
All authors consented to participate.

Consent for Publication
All authors reviewed the manuscript and consented to publication.
Conflict of Interest Jianbo Hou declares that he has no conflict of interest. Wen Xie declares that she has no conflict of interest. Yan Qian declares that she has no conflict of interest. Wenhua Zhang declares that he has no conflict of interest. Yingzhu Shi declares that she has no conflict of interest. Wei Song declares that he has no conflict of interest. Chengjie Lou declares that he has no conflict of interest.