Simultaneous determination of 25 pesticides in Zizania latifolia by dispersive solid-phase extraction and liquid chromatography-tandem mass spectrometry

An improved quick, easy, cheap, effective, rugged, and safe (QuEChERS) method combined with ultrapressure liquid chromatography tandem mass spectrometric method (UPLC-MS/MS) was developed to simultaneously determine 25 pesticides in Zizania latifolia. The samples were extracted with methanol(MeOH) and 0.1% formic acid (80:20, v/v) and cleaned with C18 absorbent and primary-secondary amine (PSA). LC separation was performed on a BEH C18 UPLC column under the condition of gradient elution with the mobile phase consisted of 0.5% formic acid (10 mM ammonium acetate)/MeOH. External standard calibration method with matrix-matched was used for quantification, and good linearity was obtained over a concentration range of 0.5–100 μg/l, with correlation coefficients greater than 0.9901. The limit of detection (LOD) and the limit of quantitation (LOQ) of the 25 pesticides were in the range of 0.2–1.0 µg/kg and 0.5–3.3 µg/kg, respectively. The recoveries ranged from 72% to 118%, and the relative standard deviations (RSDs) were less than 20%. Thus, the proposed method is suitable for the simultaneous determination of 25 pesticides in Z. latifolia.

Chromatographic conditions. A Waters Acquity UPLC instrument (Milford, MA, USA) was used for analysis, and an Acquity BEH C 18 column (2.1 mm × 100 mm, 1.7 μm) was utilized for separation while maintained at 35 °C. The mobile phase consisted of solvent A (0.5% formic acid containing 10 mM ammonium acetate) and solvent B (MeOH). The initial gradient conditions were set at 20% B and held for 1.1 min. Then, the gradient was increased linearly to 90% B at 3.5 min and maintained for 4.5 min. then the gradient was programmed to return to the initial conditions at 8.1 min to re-equilibrate the column for 1.9 min. The flow rate was 0.30 ml/min. Total run time of one injected sample was 10 minutes with the injection volume of10 µl in full-loop injection mode. www.nature.com/scientificreports www.nature.com/scientificreports/ Mass spectrometry conditions. MS/MS detection was performed on a Waters Xevo TQ triple-quadrupole MS system equipped with an electrospray ionization (ESI) source operated in positive mode. The ion source and desolvation temperatures were optimized at 150 °C and 500 °C, respectively. The capillary voltage and the flow rate of the desolvation gas (N 2 ) were set at 2.2 kV and 1000 L/h, respectively. The collision cell pressure was 3.0 mbar sustained by the collision gas argon. Detection was carried out in a multiple-reaction monitoring (MRM) mode. Other parameters are shown in Table 1.
Sample preparation. A homogeneous sample (5 g) was weighed, placed in a 50 ml polypropylene centrifuge tube, and 10 ml of 0.1% formic acid/MeOH (20:80, v/v) was added to sample. The mixture was homogenized for 1 min using a high-speed dispersing device (Ultra-Turrax T 25; IKA, Germany) and vortexed for 1 min. The tube was subsequently centrifuged at 9500 rpm for 5 min, and a 1 ml aliquot was transferred to a tube containing PSA solid-phase extraction (SPE) sorbent (75 mg) and ODS C 18 sorbent (75 mg). Next, the tube was vortexed for 30 s Method validation. Analytical performance was examined in terms of the selectivity, linearity, mean recovery, repeatability, LOD and LOQ of the method in accordance with the SANCO document 26 .
To confirm the absence of interfering substances around the retention times of the 25 pesticides, 20 blank samples were analysed.
The recoveries and repeatability (intra-day and inter-day) of the method were determined with spiked blank samples at three concentrations (0.05, 0.1 and 0.25 mg/kg for fenaminosulf, procymidone, and hexaconazole; 0.01, 0.05 and 0.1 mg/kg for the other compounds). The intra-day repeatability was determined with five replicates at each calibration level on the same day, and the inter-day repeatability was calculated from five replicates at 0.05 mg/kg per day over 3 consecutive days. The intra-day and inter-day repeatability values were expressed as the relative standard deviation (RSD).
The LOD and LOQ were calculated from the signal-to-noise ratio (S/N) of a chromatographic peak, where LOD = 3 S/N and LOQ = 10 S/N.

Results and Discussion
LC-MS/MS optimization. In this study, positive mode produced higher precursor ion signal intensities than the negative mode for all pesticides. Therefore, the analysis of target compounds.is carried out in the [M + H] + ion mode as the precursor ion. One parent ion and two transitions were chosen. The most intense transition was used for quantitation 27 , while the other transition was employed for qualitative. The optimal parameters for each compound are shown in Table 1.
After optimizing the MS cinditions, the mobile phase composition was explored by the chromatographic column. It is well known that the [M + H] + ion forms easily under acidic conditions. Therefore, 0.1% formic acid/ACN and 0.1% formic acid/MeOH solutions were first investigated. Satisfactory separation was difficult to achieve for the 9 triazole pesticides when 0.1% formic acid/ACN solution was used, while the peak shape of pymetrozine was poor when 0.1% formic acid/MeOH was used (Fig. 2a). To achieve both these goals simultaneously, the 0.1% formic acid solution was replaced with 0.1% formic acid containing 10 mM ammonium acetate. With this solvent system, the separation of the 9 triazole pesticides did not improve significantly, but the peak shape of pymetrozine improved (Fig. 2b). Thus, 0.1% formic acid (10 mM ammonium acetate)/MeOH was chosen initially. However, the ionization of procymidone and hexaconazole was suppressed in this mobile phase (Fig. 2b). The responses of procymidone and hexaconazole obviously increased when the concentration of formic acid was changed from 0.1% to 0.5% (Fig. 2c). Thus, 0.5% formic acid (10 mM ammonium acetate)/MeOH was finally chosen as the mobile phase in the current study. Moreover, in order to improve the sensitivity of all compounds, the chromatogram was divided into five regions.
Optimization of sample preparation. Salting-out assisted water-acetonitrile extraction is an convenient sample preparation technique when pesticide residue analytical method development. Compared with traditional liquid-liquid extraction and SPE, this method is more environmentally friendly, more cost-efficient and faster. www.nature.com/scientificreports www.nature.com/scientificreports/ Hence over the past decade, it has obtained growing interest in QuEChERS sample preparation [28][29][30] . However, salting out was not used in the present study because fenaminosulf and pymetrozine are highly soluble in water. To achieve satisfactory recoveries for all target compounds from the Z. latifolia samples, three extraction solvents (0.1% formic acid/ACN (20:80, v/v), 0.1% formic acid/MeOH (20:80, v/v) and ACN) were evaluated at a fortification level of 50 µg/kg. The best recoveries for most of the compounds were obtained with 0.1% formic acid/MeOH (20:80, v/v), which was selected as the optimal extraction solvent.
Z. latifolia mainly contains carbohydrates, proteins and fats. Pesticides with high polarity are highly susceptible to interferences from impurities. To reduce the level of the co-extracted matrix, and obtain good purification efficiency, a simple and effective clean-up procedure with dispersive SPE (dSPE) is often used. The original QuEChERS method involves cleaning up with PSA sorbent 31 . PSA can effectively adsorb organic acids, fatty acids, sugar, and other interferences in the matrix. However, compounds with carboxyl groups are easily retained    www.nature.com/scientificreports www.nature.com/scientificreports/ by PSA. The QuEChERS method has been modified to enable the use of C 18 for clean-up, and strong adsorption of low-polarity matrix interferences such as fatty acids, olefins, and large molecules, such as sterols and pigments has been achieved 32,33 . In the present study, a mixed PSA-C 18 (1:1) sorbent was used for clean-up. The effects of the amount of PSA-C 18 (1:1) sorbent (50-300 mg) on the matrix effect and recoveries were examined in detail (Fig. 3). The matrix effect was counted by the following formula: matrix effect = (external calibration slope for matrix-matched standards/external calibration slope standard in solvent) 34,35 . For most compounds, the recoveries were above 90%, and the matrix effect did not obviously change when the amount of PSA-C 18 (1:1) was varied from 50 to 300 mg. However, there were significant differences in the recoveries and matrix effects of fenaminosulf and iprodione when the amount of PSA-C 18 (1:1) was increased from 50 to 300 mg (Fig. 3). According to the data in Fig. 3, 150 mg of PSA-C 18 was selected as the optimal sorbent amount.
Matrix effect. The detector response of pesticides may be influenced by co-extracted materials from the sample. To evaluate the matrix effect, the ratio of the external calibration slope for the matrix-matched standard and the external calibration slope for the standard in solvent was compared for each target compound ( Table 2). According to the study of Frenich et al., when the value is between 0.8 and 1.2, signal suppression or enhancement