Introduction

Metazachlor [2-chloro-N-(pyrazol-1-ylmethyl)acet-2′,6′-xylidide] (Fig. 1), a chloroacetamide herbicide developed by BASF, was produced by introducing 2,6-dimethylaniline (2,6-DMA; C8H11N) into the structure of chloroacetamide (C2H4ClNO) [1,2,3]. Metazachlor is mainly used to control annual weeds and broadleaf weeds before or after germination, and is absorbed through the roots or stems. Absorbed metazachlor affects elongase activity during lipid biosynthesis, thereby inhibiting the synthesis of very-long-chain fatty acids, as well as interfering with cell division and tissue differentiation during growth, resulting in the production of deformed seeds [1,2,3,4,5,6]. Metazachlor was completely metabolized in the plant and was not detected at harvest [7]. Furthermore, metazachlor residues were not detected in kohlrabi after pesticide application [6], as it is quickly metabolized in plants [8]. Therefore, to monitor metazachlor in plants, analytical methods are required for determination of its metabolites Pesticide metabolism can be divided into two phases. By these processes, nonpolar pesticides are generally converted into polar pesticides [9]. In plants, the Cl atom in metazachlor is replaced by glutathione, which can be decomposed by oxidation and hydroxylation [2] to form ethanesulfonic acid (ESA) and oxanillic acid (OA) in the environment [10]. The major metabolites of metazachlor are 479M04 [N-(2,6-dimethylphenyl)-N-(1H-pyrazol-1-ylmethyl)oxalamide], 479M08 [N-(2,6-dimethylphenyl)-N-(1H-pyrazol-1-ylmethyl)aminocarbonylmethylsulfonic acid], and 479M16 [3-[N-(2,6-dimethylphenyl)-N-(1H-pyrazol-1-ylmethyl)aminocarbonylmethylsulfinyl]-2-hydroxypropanoic acid], the structures of which are shown in Fig. 1 [8]. In the European Community, maximum residue limits (MRLs) of 0.01–0.4 mg/kg have been established for metazachlor in more than 378 products, including oranges and eggs. The residue definition for foods of plant origin was proposed as the sum of metabolites 479M04, 479M08, and 479M16, expressed as metazachlor, whereas for foods of animal origin, the sum of metazachlor and its metabolites containing a 2,6-DMA moiety, expressed as metazachlor, was used [8, 11]. Therefore, similar definitions should be used to establish MRLs in Korea.

Fig. 1
figure 1

Structures of metazachlor and its metabolites

Full size image

Although no prior analytical method for metazachlor determination has been reported in Korea, various chloroacetamide herbicides (alachlor, metolachlor, metazachlor, etc.) are included in an official multiresidue analytical method (Ministry of Food and Drug Safety (MFDS) code 7.1.2) [12]. In the case of acetochlor, metabolites extracted from agricultural products using acetonitrile and then purified by dispersive-solid phase extraction (MFDS code 7.1.4.216) are also included [13]. Outside of Korea, chloroacetamide herbicides and their acidic metabolites (–ESA, –OA) in water have been purified using C18 cartridges and analyzed simultaneously using liquid chromatography–tandem mass spectrometry (LC–MS/MS) [10, 14, 15]. Recently, sample preparation using a quick, easy, cheap, effective, rugged, and safe (QuEChERS) method has been reported for studying metazachlor in soil, fruit, and kohlrabi [6, 16, 17]. In addition, Valls-Cantenys et al. [18] and Karier et al. [5] developed an analytical method for the determination of metazachlor and its metabolites (479M04 and 479M08), which was applied to monitor river water samples. Notably, 479M04 was detected in all samples, whereas 479M08 was detected in a drinking water reservoir in Luxembourg. Owing to this serious contamination, there is a necessity for safety management in consideration of the hazards to consumers, even though the metabolites have lower toxicities than the parent compound [5]. Therefore, the purpose of this study was to develop an analytical method for the determination of metazachlor metabolites in crops.

Materials and methods

Reagents, standard compounds and crop samples

479M04 (metazachlor-OA, 96.5%), 479M08 (metazachlor-ESA, 99.1%), and 479M16 (94.2%) standards were supplied from Kyung Nong Co. Ltd. (Seoul, Korea). Acetonitrile and methanol were HPLC grade (Merck, Darmstadt, Germany). Formic acid and ammonium formate were ACS grade (Sigma-Aldrich, Buchs, Switzerland). Sodium chloride (NaCl) was obtained from Junsei Chemical Co. Ltd. (Tokyo, Japan). Hydrophilic–lipophilic balance (HLB) cartridges (6 cm3, 500 mg) were purchased from Waters (Milford, MA, USA) and syringe filters (PVDF, 0.2 μm × 13 mm) were obtained from Teknokroma (Barcelona, Spain). Hulled rice, potato, soybean, mandarin, and green pepper of “residue-free grade” were purchased from a local market. The agricultural samples were homogenized and kept at a temperature below −50 °C in sealed containers until use.

Stock and working solution

The stock solutions of 479M04, 479M08, and 479M16 at a concentration of 1000 μg/mL were prepared dissolving the appropriate amount of each compound in methanol. The working solutions were prepared by appropriate dilution of the stock solutions with methanol. For the matrix-matched calibration curves, the concentrations of 0.002–0.2 μg/mL were obtained by mixing with blank extracts and standard solutions were mixed at a ratio of 9:1. All the standard solutions were stored at 4 °C in amber vials and the matrix-matched standards were newly prepared before each analysis.

Sample preparation

For homogenization of grains and soybeans, approximately 1 kg of sample pulverized to pass through a standard sieve of 420 μm. For fruits and vegetables, approximately 1 kg of sample were ground. Then, 5 g of each homogenized sample was accurately weighed and placed in a 50 mL centrifuge tube. For grains and soybeans, 5 mL of distilled water was added and the mixture was left to stand for 10 min, whereas this step was not necessary for the other samples. Following the addition of 25 mL of acetonitrile, the mixture was shaken for 1 min and then the pH was adjusted to 3 using formic acid (for grains and soybeans, 4 g of NaCl was also added). After shaking for 10 min, the extract was centrifuged for 5 min (4000g, 4 °C), the supernatant was collected, and 25 mL of acetonitrile was added. Subsequently, the above procedure was repeated twice and the extracts were combined. The extract was evaporated to dryness at 40 °C and redissolved in 12.5 mL of methanol. After loading 5 mL of the extract on an HLB cartridge, which was conditioned with 3 mL of methanol and 3 mL of distilled water, 5 mL of the eluate was collected. Subsequently, the cartridge was eluted with 5 mL of methanol, which was then mixed with the previous eluate and the total volume was adjusted to 10 mL. This solution was then filtered through a syringe filter (PVDF, 0.2 μm × 13 mm).

LC–MS/MS conditions

LC–MS/MS analysis was performed using an Acquity UPLC/Xevo TQ-S system (Waters, Milford, MA, USA) equipped with a Capcell Core ADME column (100 mm × 2.1 mm i.d., 2.7 μm, Shiseido, Tokyo, Japan) at a temperature of 40 °C. The injection volume and flow rate were 5.0 μL and 0.3 mL/min, respectively. Separation was performed by gradient elution using methanol (mobile phase A) and 5 mM ammonium formate in water (mobile phase B). The initial mobile phase of 10/90 (A/B, v/v) was retained for 3 min, changed to 90/10 (A/B, v/v) over 5 min and then held for 2 min, and finally changed to 10/90 (A/B, v/v) over 0.1 min and then retained until 10 min. MS/MS was conducted using positive electrospray ionization (ESI) and multiple reaction monitoring (MRM). The ion source temperature and the desolvation temperature were 150 °C and 500 °C, respectively, and the capillary voltage was 1.0 kV. The MRM conditions are summarized in Table 1.

Table 1 LC-MS/MS parameters for the analysis of metazachlor metabolites
Full size table

Method validation

To validate the developed analytical method, the recovery (accuracy and repeatability), linearity, selectivity, limit of detection (LOD), and limit of quantification (LOQ) in agricultural samples were assessed based on the MFDS guidelines on standard procedures for preparing analysis methods and the CODEX guidelines [19, 20]. The selectivity was compared using chromatograms of blank samples and blank samples spiked with metazachlor metabolites to confirm whether interfering peaks occurred at the retention times of the standards. The linearity of the matrix-matched calibration curves at concentrations of 0.002–0.2 μg/mL were assessed by the coefficients of determination (r2). The LOD and LOQ values of the instrument were determined as 3 and 10 times, respectively, the signal to noise (s/n) ratio of the chromatogram. Subsequently, the LOQ of the analytical method was calculated by considering the LOQ of the instrument, the weight of the sample, and the volume of solvent. Recovery tests were conducted at three fortification levels (0.01 (LOQ), 0.1 (10 LOQ), and 0.5 mg/kg (50 LOQ), n = 5) in the five representative crops. The accuracy and repeatability of the analytical method were assessed by calculating the average and relative standard deviation (RSD) values. Finally, inter-laboratory testing was conducted at Chonnam National University to confirm the validity of the developed method.

Results and discussion

Optimization of LC–MS/MS conditions

Currently, there are no analytical methods for the determination of metazachlor and its metabolites in the MFDS food code, although metazachlor has been included in a multiresidue analytical method using gas chromatography–tandem mass spectrometry (GC–MS/MS) [12]. However, metazachlor metabolites are difficult to analyze with GC, as these compounds are relatively polar (for 479M04 and 479M16, log Pow = − 0.54). Therefore, owing to the introduction of the positive list system, analysis methods are required for these analytes, and LC–MS/MS has been chosen as an appropriate analysis technique owing to its high selectivity and low LOQ (≤ 0.01 mg/kg) [5, 18]. ESI has a lower noise level and better sensitivity than atmospheric pressure chemical ionization (APCI) for pesticide analysis [21]. In particular, for metazachlor, ESI has been shown to have higher sensitivity than APCI [18]. When a mobile phase of 0.1% formic acid in water and 0.1% formic acid in acetonitrile was used, peak splitting was observed for the metazachlor metabolites, likely owing to the presence of diastereomers, as observed for the acidic metabolites of chloroacetamide [10, 22, 23]. Shortening the column length [10] and increasing the analysis temperature [22, 23] have been reported to provide improved performance for the analysis of metazachlor. In this study, the analysis conditions were further improved by using ammonium formate and changing from acetonitrile to methanol (Fig. 2). As peak tailing occurred when formic acid or acetic acid was added to the mobile phase, no acid was added. A comparison of the peak shapes with ammonium formate or ammonium acetate added to water (mobile phase B) showed that the tailing issue was resolved when formate was used. This result is consistent with the findings of Valls-Cantenys et al. [18], who found that the addition of formate improved the peak shape and increased the s/n ratio. For 479M04 (exact mass: 273.11), 479M08 (exact mass: 323.09), and 479M16 (exact mass: 379.12), the direct injection of standard solutions (0.1 μg/mL) into the mass spectrometer at a constant rate (10 μL/min) gave [M + H]+ peaks at m/z 274, 324, and 380, respectively. The product ion with the highest intensity was used as the quantification ion and the product ion with the second highest intensity was used as the qualification ion (Table 1). When operating LC–MS/MS, the extent of ionization suppression or enhancement for each compound due to the matrices was accounted for by matrix-matched calibration.

Fig. 2
figure 2

Optimization of peak shape according to mobile phases

Full size image

Optimization of sample extraction

Acetonitrile and methanol, which have been used in previous studies [3, 23,24,25], were first considered as extraction solvents. The QuEChERS method, which has the advantages of requiring relatively low amounts of solvent and short extraction times, was evaluated using three different procedures (original, AOAC 2007.01, and CEN15662) that differ according to the reagents used in the extraction step [26,27,28]. Using the three QuEChERS methods, extraction with acetonitrile and methanol from mandarin and soybean gave poor recoveries (22.3%–74.0%). Therefore, the pHs of the acetonitrile and methanol extracts were adjusted to 1–7 using formic acid (Fig. 3). While methanol still exhibited low extraction efficiencies, all the metabolites exhibited excellent recovery with acetonitrile at pH 3. Although this extraction method was found to be inefficient for soybeans, this issue was resolved by repeating the extraction process twice (89.0% → 95.2%). Furthermore, when dried samples (hulled rice and soybean) were extracted, amount of sodium chloride was increased and then recovery results were improved (479M04 77.2% → 87.4%, 479M08 69.2% → 81.0%, 479M16 78.8% → 82.4%). Because it increases ionic strength and thus the distribution efficiency from the water layer to the organic solvent layer via the salting out effect [29].

Fig. 3
figure 3

pH dependence of the extraction efficiencies of metazachlor metabolites from the mandarin matrix using a acetonitrile and b methanol

Full size image

Optimization of cartridge clean-up

Various purification methods, such as using a C18 cartridge [24, 30] or liquid–liquid partitioning [25], have been widely studied. However, as polar compounds, the metazachlor metabolites elute without adsorption to C18 and are difficult to collect following partitioning, as they move into the water layer. Recently, in recent studies on metazachlor monitoring in water sample, an HLB cartridge was applied for sample clean-up [5, 18]. Thus, an HLB cartridge was adopted for sample clean-up in this study. The HLB cartridge gave good recoveries of the metazachlor metabolites from a mixed standard solution at 0.05 μg/mL (479M04: 100.7%, 479M08: 109.8%, and 479M16: 128.5%).

Method validation

To validate the developed analytical method, the recovery (accuracy and repeatability), linearity, selectivity, LOD, and LOQ were assessed for the five agricultural samples. When analyzing the blank (nonspiked) samples, no interfering peaks appeared at the same retention times and mass-to-charge ratios (m/z) as those of the metabolites. Therefore, this analytical method provided high separation ability and selectivity for the metazachlor metabolites (Fig. 4). The standard solutions were diluted with a blank extract to obtain matrix-matched standards at concentrations of 0.002–0.2 μg/mL, and the linearity of each matrix-matched calibration curve was assessed based on the coefficient of determination (r2). From the analyses at various concentrations, LOQ values of ≤ 0.01 mg/kg were determined for all five products. Recovery tests for the five agricultural products were performed in triplicate at three fortification levels (0.01, 0.1, and 0.5 mg/kg), and the accuracy and repeatability of the analytical method were assessed based on the calculated average and RSD. The results of interlaboratory test satisfied the criteria of both the CODEX and MFDS guidelines [19, 20] (Table 2).

Fig. 4
figure 4

Representative MRM (quantification ion) recovery chromatograms (fortification concentration: 0.1 mg/kg) of a 479M04, b 479M08 and c 479M16 in (1) mandarin control, (2) matrix-matched standard, and (3) mandarin test sample

Full size image
Table 2 Validation results for the determination of 479M04, 479M08, and 479M16 in agricultural samples
Full size table

In conclusion, the analytical methods were established for the determination of the residues of metazachlor metabolites in five agricultural commodities using LC–MS/MS. Through method validation, the method achieved good accuracy and repeatability, and the results were confirmed with chromatogram of LC–MS/MS (MRM). Finally, this study was can be used as an analytical method for the determination of metazachlor metabolites in agricultural products.