Multiresidue Method for the Rapid Determination of Pesticide Residues in Tea Using Ultra Performance Liquid Chromatography Orbitrap High Resolution Mass Spectrometry and In-Syringe Dispersive Solid Phase Extraction.

A method based on in-syringe dispersive solid phase extraction (IS-D-SPE) and ultra performance liquid chromatography Orbitrap high resolution mass spectrometry for the multiresidue analysis of 117 pesticides in tea was developed. Full scan mode was acquired over an m/z range of 100-800 with Orbitrap resolution at 70000, followed by full scan/dd-MS2 mode for confirmation. The identification criteria of retention time and mass accuracy tolerance was ±0.20 min and ±5.0 ppm, respectively. MS/MS fragment ions obtained dd-MS2 were necessary to identify the pesticides with the same molecular mass weight. The IS-D-SPE technique involved a mixture of 200 mg PSA, 100 mg C18, and 15 mg multiwalled carbon nanotubes for the cleanup of tea matrix. Good linearity (R2 > 0.99) for 117 pesticides was obtained. Satisfactory recoveries in the range of 70-120% were obtained for 105 pesticides, while intraday and interday precisions were below 20%. Limits of quantification were generally 10 μg kg-1. Finally, this method was employed to analyze 117 pesticides in 70 tea samples.

Introduction

Tea is the most widely consumed beverage next to water in the world. By drinking a cup of tea, one can take in plenty of special nutrient substances, such as catechins, theanine, polysaccharide, vitamins, and essential minerals. These substances are beneficial for human health due to their effective medicinal and therapeutic potentialities, such as cancer prevention, antioxidants, anticardiovascular diseases, correcting skin disorder, and body weight reduction.[1] Besides beneficial healthy function, the fascinating aroma and comfortable taste attract people to enjoy tea, especially young people.

Tea (Camellia sinensis L) is a perennial woody plant and is grown in regions such as tropical, subtropical, and temperature zones. These growth conditions are favorable for pests, diseases, and competing grasses. In China, more than 800 pests and diseases have been found in tea gardens, which results in serious damage of tea yield and quality.[2] Currently, application of pesticide formulations is the most effective measure to prevent pests and disease, and improve tea yield and quality. Regarding the negative effects of the application of pesticides in tea, there are potentially harmful residues. The presence of pesticide is an important issue in tea safety and tea trade.[3] Many countries and regions, such as Japan, European Union (EU), China, and Codex Alimentarius Commission (CAC), have formulated maximum residue levels (MRLs) in tea. The strictest MRLs are set by the EU, where 515 pesticide MRLs are regulated, and MRLs for unregulated pesticides are set at 0.010 mg kg–1.[4] Japan also has set up 264 pesticide MRLs and the unregulated pesticide MRLs are at 0.01 mg kg–1.[5] Chinese national standard has formulated 387 pesticide MRLs in agricultural products, where 48 pesticide MRLs in tea have been regulated in the range of 0.1–20 mg kg–1.[6] Therefore, high throughput pesticide screening at trace levels is necessary to ensure tea safety and reduce the tea economic loss due to exceeding the pesticide MRLs.

Chromatography coupled with mass spectrometry is widely used for the screening and determination of pesticide residues in tea.[7,8] Gas chromatography coupled with single mass spectrometry (GC-MS) or tandem mass spectrometry (GC-MS/MS) is employed for the analysis of volatile and thermostable pesticides, while liquid chromatography tandem mass spectrometry (HPLC-MS/MS) is alternatively used to detect the polar, nonvolatile, and thermally unstable pesticides.[9] A high throughput GC-MS and HPLC-MS/MS method for the multiresidue determination of 490 and 448 pesticides, respectively, has been developed and validated by Pang et al.[10] For GC-MS/MS and HPLC-MS/MS, several parameters, such as the precursor and two MS/MS transitions, retention time, and the area ratio between two MS/MS transitions, are required to be monitored. However, false positive or negative findings for the MS/MS technique may occur due to the use of nonspecific transitions, characterized by a poorly resolved chromatographic peak.[11] Other drawbacks and limitations for the MS/MS technique which are inherent to targeted acquisition are as follows: time-consuming optimization of MS/MS parameters, limited number of compounds in one instrumental method, time-consuming and constant definition of acquisition-time window, and nonretrospective data analysis.[12] Although LC-MS/MS and GC-MS/MS employing triple-quadrupole MS systems are traditionally used for multipesticide analysis, the use of high resolution mass spectrometry (HRMS) with full-scan technique has become a promising analytical technique for the high throughput screening and determination of organic pollutants in foods, the environment, and other matrices.[13,14] Compared with MS/MS, the HRMS technique has several advantages that are important for screening and identification of unknown compounds: higher sensitivity, higher accurate mass for calculating elemental composition, and higher accurate mass of product ions (MSn).[15]

The Orbitrap high resolution mass analyzer was first described in 2000 and commercially introduced in 2005.[16,17] In Orbitrap HRMS, the option for an internal mass calibration can be chosen to correct all the scans in a LC-Orbitrap MS experiment.[18] A mass accuracy greater than 5 ppm at resolving power 60,000 fwhm (full-width at half-maximum peak height at m/z 400) can be achieved due to the use of the Orbitrap mass analyzer preceded by an external injection device based on trapping ions RF-only gas-filled curved quadrupole (named C-trap).[19] Due to accurate mass and high sensitivity, the utility of HPLC-Orbitrap-MS is sufficient for the measurement of a wide range of pesticides at trace residue in various matrices, such as fruit and vegetables,[20] honey,[21] fish,[22] and baby foods.[23] Currently, there are few literature references on this analytical technique for high throughput screening pesticide residues in tea using HPLC Orbitrap MS. Until now, only two studies related to high throughput screening pesticide residues in green tea and its nutraceuticals employing HPLC Orbitrap MS have been reported. One previous report focused on screening pesticide residues in the nutraceutical products, such as tea powder and extracts.[24] The tea nutraceuticals are generated from raw tea leaves or manufactured tea and would have undergone complex processing. The presence of matrix differences may result in different matrix interference and method performance when HPLC Orbitrap MS is applied. Another previous study focused on the UPLC-Q-Orbitrap MS identification of pesticide residues in agricultural products, such as vegetables, fruits, and green tea.[25] Therefore, it is necessary to establish a novel methodology for screening and determination of multipesticide residues in tea using UPLC-Q-Orbitrap MS.

The purpose of this study was to develop a high throughput screening and multi residue analysis of pesticides in various tea samples (green tea, black tea, and oolong tea) using UPLC-Q-Orbitrap-MS. A rapid and simple sample preparation employing in-syringe dispersive solid phase extraction (IS-D-SPE) cleanup was optimized. Full scan/dd-MS2 monitoring mode was used for screening and identification of targeted pesticides.

Experimental Section

Chemicals and Reagents

All pesticides were of certified quality and purchased from Dr. Ehrenstorfer (Augsburg, Germany), Sigma–Aldrich (Steinheim, Germany), J&K Chemical (Beijing, China). Full name, chemical formula, and accurate mass are listed in Supporting Information Table-S1. HPLC-grade acetonitrile, methanol, and acetone were provided from Sigma–Aldrich (Merck, Germany). HPLC-MS grade formic acid and ammonium acetate were purchased from Sigma–Aldrich. Deionized water was obtained using a Milli-Q plus ultrapure water system from Millipore (Miford, USA). The sorbents of primary second amine (PSA), multiple walled carbon nanotubes (MWCNTs), and octadecylsilane (C18) were provided from Agela (Tianjin, China). Analytical-grade sodium chloride and anhydrous magnesium were provided from Zhejiang Medicine (Hangzhou, China), and baked for 3 h at 650 °C prior to use.

Samples

All tea samples (green tea, black tea, and oolong tea) were obtained from Laboratory of Tea Safety and Risk Assessment, Ministry of Agriculture, P.R. China. Organic tea samples free of pesticides were used for matrix standard solution and recovery test. Ten positive tea samples, which were previously detected and confirmed using UPLC-MS/MS, were employed for the optimization of the extraction procedure. All tea samples were stored in a dark room at 1–4 °C.

Standard Solution Preparation

Individual stock standard solutions (generally at 1000 μg mL–1) were prepared in methanol or dissolved in a few volumes of acetone and then added to the scale of volumetric flask using methanol. Of these stock solutions, standard mixture solutions were diluted at 5 μg mL–1 using methanol. Working standard solutions were prepared by the serial dilution of standard mixture solutions using methanol. Matrix matched standard solutions were prepared by adding less than or equal to 100 μL of working standard solution into about 900 μL of the blank extractions (total volume of standard solution and blank extraction was equal to 1 mL), which were obtained from organic tea samples and cleaned up by IS-D-SPE.

Sample Preparation

The proposed sample preparation method was performed using acetonitrile extraction by vortex and IS-D-SPE cleanup. Homogenized tea powder (2.5 g) was weighed into a 50 mL centrifuged tube, and then 5 mL of deionized water was added and mixed by a vortex at 1200 rpm for 1 min. Ten mL of acetonitrile was added, followed by extraction using a vortex at 1200 rpm for 2 min. Afterward, 2 g of NaCl and 4 g of MgSO4 were added and mixed by a vortex at 1200 rpm for 1 min, and followed by centrifugation at 5000 rpm for 5 min. Two mL of upper layer extraction was transferred into a syringe containing 200 mg PSA, 100 mg C18, and 15 mg MWCNTs. After plugging the bottom, the syringe was blended using a vortex at 1200 rpm for 1 min. Finally, the plug was changed with a 0.22 μm filter membrane. The extraction was transferred into sample vial for UPLC-Q-Orbitrap-MS analysis.

To determine the effect of water immersion on extraction efficiency, the incurred samples were extracted with acetonitrile with and without water immersion. Comparison of different extraction methods, e.g., stand by for overnight, homogenizer, vortex, and shanking device, were carried out for shortening extraction time and streamlining sample preparation.

UPLC-Q-Orbitrap-MS Parameters

UltiMate 3000 HPLC system (Thermo Fisher Scientific, Waltham,MA, USA) was used for the separation of the analytes on the reverse phase Hypersil GOLD C18 (100 mm × 2.1 mm i.d., 3 μm particle size, (Thermo Fisher Scientific, Waltham,MA, USA). The mobile phase consisted of 0.1% (v/v) formic acid and ammonium formate 4 mM in water (A) and methanol (B). A gradient elution was applied as follows: 10% B, 0–1 min; 10–75% B, 1–3 min; 75–100% B, 3–4 min; 100% B, 4–10 min; 100–10% B, 10–11 min; 10% B, 11–14 min. The sample tray temperature, column oven temperature, injection volume, and flow rate set at 20 °C, 30 °C, 10 μL, and 0.3 mL min–1, respectively.

The UPLC system was coupled with a Q-Orbitrap MS, a high-performance benchtop quadrupole Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham,MA, USA). Ionization was achieved by operating a heated electrospray ionization (H-ESI) probe in the positive mode. The Orbitrap MS parameters were as follows: sheath gas (N2, >99%), 35 mL; auxiliary gas (N2, >99%), 10 mL; heater temperature, 350 °C; capillary temperature, 300 °C; spray voltage, 4 kV; skimmer voltage, 18 V; capillary voltage, 35 V; tube lens voltage, 95 V. Quantification and screening data were acquired using full scan and full scan/dd-MS2 mode, respectively. The full scan mode was acquired over an m/z range of 100–800 with Orbitrap resolution at 70000 full width at half-maximum (fwhm) at m/z = 200. Automatic gain control (AGC) target ion set at 1.0e[6] for a maximum injection time of 100 ms. Stepped normalized collision energies (NCEs) were set at 30, 50, 70 eV. The dd-MS2 resolution was set at 17500 fwhm (m/z = 200). The precursor ions with a signal threshold >1.0e[4] were automatically performed for MS2 fragmentation activation with 40% collision energy.

Data Processing

Data acquisition and processing were performed using Tracefinder 3.3 software package. Data calculations and statistical analysis were performed using Microsoft Excel and SPSS 12.0, respectively. Accurate-mass database of protonated molecular ion [M + H]+ or ammoniated molecular ion [M+NH4]+ and MS/MS fragment ions, and retention time were obtained by injection of individual standard solutions at 500 ng mL–1. The defined criteria for identification analysis was as follows: retention time ±0.20 min, mass accuracy tolerance ±5.0 ppm, identical molecular ions, and MS/MS fragment ions. The most abundant ion, typically the protonated molecular or ammoniated molecular ion, was used for quantification.

Validation Procedure

Matrix effect was evaluated by comparing the slopes of the analytical curves obtained using matrix matched standards and solvent. Each analytical curve was established at six different concentrations in the range of 1–200 μg L–1 (equal to 8–800 μg kg–1 for tea samples). Matrix effect is calculated as follows:The method was validated according to the international SANTE guidelines SANTE/11945/2015.[26] Representative matrices, including green tea (unfermented), black tea (fermented), and oolong tea (semifermented), were used for validation. Linearity was evaluated using matrix matched standard solution in the range of 1–200 μg L–1 (equal to 4–800 μg kg–1 for tea sample). Accuracy was measured in terms of recoveries obtained by five blank samples fortified at 10 and 50 μg kg–1 of the respective analytes. The lowest fortification level was chosen based on the lowest maximum residue levels (MRLs) which were generally set up at 10 μg kg–1. Precision was evaluated as intraday and interday repeatability, expressed as relative standard deviation (RSD). Limits of detection (LODs) were obtained by injecting fortified samples at 0.5, 1.0, 2.0, 5.0, and 10.0 μg kg–1. Limits of quantification (LOQs) were determined by analyzing minimum fortified samples that provided suitable recovery in the range of 70–130% and RSDs within 20%.

Results and Discussion

Establishment of Compound Database for Qualification and Quantification

To develop a database of target compounds, individual standard solutions for 117 pesticides at 50 μg L–1 were injected into the UPLC-Q-Orbitrap-MS, and the MS acquisition in positive was performed with and without fragmentation in the HCD collision cell. Table S-1 shows the accurate mass, retention time, and fragmented ions obtained from full scan/dd-MS2 mode. Mutual interference was not observed for 113 pesticides based on chromatographic retention time window ±0.2 min and mass error ±5 ppm. For full scan mode, the compounds with the same mass (isomers) are hardly distinguished because they have the same molecular mass weight. As shown in Figure , quinalphos and phoxim have the same chemical formula C12H15N2O3PS with the theoretical mass m/z at 299.0613 Da (experimental mass at m/z 299.0607 Da), and the retention time deviation was less than 0.1 min. In this case, dd-MS2 monitoring mode is necessary for obtaining fragmented ions to identify quinalphos and phoxim. In the HCD collision cell, the protonated molecular ion [M + H]+ of quinalphos was fragmented into the product ions at m/z 147.0550 and 163.0326, while the fragmented product ions at m/z 129.0448 and 114.9609 were obtained from the phoxim protonated molecular ion. Besides quinalphos and phoxim, the isomers pretilachlor and metolachlor have the same accurate mass and similar retention time and different fragmented ions of protonated molecular ion, which is shown in Figure S-1. Therefore, UPLC-Q-Orbitrap MS was performed with full scan monitoring mode for identification and quantification, and full scan/dd-MS2 was complementary for identification of the compounds with the same chemical formula and molecular weight.

Figure 1

Extracted ion chromatograms (m/z 299.0607 with exact mass error 5 ppm) (A1, B1), full scan mass spectrogram (A2, B2), and dd-MS2 fragmented ions (A3,B3) for quinalphos (A) and phoxim (B) in standard solution at 50 μg L–1.

In this study, interference from the tea matrix was investigated. No interference from the matrices was found for any of the pesticides, except allethrin, when retention time window set at 0.5 min. As shown in Figure , three chromatographic peaks (extracted ion m/z 303.1952) obtained from blank tea sample could result in the interference with allethrin (mass error at 5 ppm). To evaluate retention time tolerance, 20 successive injections of a mixture of green tea matrix matched calibration solution at 50 μg L–1 were carried out. The results showed that retention time deviations for all pesticides were less than 0.15 min. Therefore, the retention time window set at 0.2 min in this study, and none of the 117 pesticides interfered with the tea matrix.

Figure 2

Extracted ion chromatograms (mass error 5 ppm) for allethrin (m/z 303.1952, RT 6.99 min) at 1 μg L–1 in methanol (A), green tea matrix matched calibrated solution (B), and blank green tea matrix (C).

Determination of Extraction Efficiency of Pesticide Residues in Incurred Tea Samples Using Different Extraction Techniques

In order to evaluate the extraction efficiency of multi pesticide residues, we choose different types of pesticides, which have a wide range of water solubility (0.48–39800 mg/L, 20 °C) and octanol–water ratio (Log Kow 0.8–6.37). The effect of extraction technique on detected concentrations of 11 pesticide residues in 10 incurred tea samples was investigated. First, the effect of the additional water on extraction efficiency was investigated. Figure illustrates that all detected concentrations of 11 incurred pesticides increased, ranging from 1.1 to 18.6 times, when tea samples were allowed to be immersed in boiling water prior to acetonitrile extraction, in comparison with the use of acetonitrile as extraction solvent without soaking in boiling water. The enhancement ratio seems negatively correlated with octanol–water partition coefficient (Log Kow). Second, the effect of soaking by boiling water and room-temperature water on extraction efficiency was investigated. As shown in Table S-2, the detected concentrations obtained from soaking in water (at 20–25 °C) (E5) were less than 3.6–14.9% of the average values obtained from six extraction procedures. The results illustrated that boiling water was helpful for acetonitrile to permeate tea leaves and improve extraction efficiency of pesticides. Finally, different extraction techniques were compared and the results were shown in Table S-2. Statistical analysis based on SPSS LSD test showed that there was no significant difference of detected concentrations of 11 pesticides in 10 incurred tea samples using five extraction techniques, e.g., standing overnight (about 12 h) (E1), vortexed for 2 min (E2), rotational oscillation at 300 rpm for 10 min (E3) or 20 min (E4), and homogenization with blend probe (E6). The results obtained from E6 were slightly higher than from other extraction techniques. The extraction procedure is one of the key points for high throughput sample preparation and it requires simultaneous processing of multiple samples. For homogenization with one blend probe, it is inconvenient, slow, laborious, and potentially hazardous due to its extensive and problematic cleaning steps.[27] For high throughput sample analysis, extraction procedures E1 (stand by for overnight) and E3 (rotational oscillation for 10 min) were proposed, while extraction procedures E2 and E3 were suitable for quick sample preparation.

Figure 3

Enhancement of extraction efficiency of 11 representative pesticides (values of Log Kow are given in brackets) in incurred tea samples with or without water soaking prior acetonitrile extraction (n = 5).

Optimization of IS-D-SPE Cleanup Procedure

IS-D-SPE is a dispersive solid extraction (D-SPE) sample preparation technique using a syringe containing adsorbents for cleanup, which does not need centrifugation and therefore takes less time for sample preparation.[28−31] Several dispersive adsorbents (PSA, C18, GCB, SAX, florisil, carbon nanotubes) have been previously reported for removal of tea matrix as D-SPE adsorbents.[8,32−38] In this study, PSA, C18, and MWCNTs were prepared for cleanup in the mixture of IS-D-SPE adsorbents, and each mass of adsorbents was optimized. To evaluate the recovery and cleanup effects, blank extracts spiked with 117 pesticides at 25 μg L–1 were investigated using different adsorbents. All pesticides could obtain good recovery in the range of 90–110%, and the color density of extracts had little change when PSA and C18 were individually used as IS-D-SPE adsorbents with the amount of 0–200 mg, which indicated that both adsorbents had little adsorption capacity of 117 pesticides, as well as pigments or other color substances. MWCNTs, as a carbon-based nanomaterial, has been reported as an excellent adsorbent for the removal of agricultural sample matrix owing to the large surface area and structural characteristics.[39−41] Color density of tea extracts became lighter when the mass of MWCNTs increased from 0 to 50 mg. In consideration of the adsorptivity of MWCNTs for pesticides, mixture adsorbents of 200 mg PSA, 100 mg C18, and MWCNTs with the mass of 0–50 mg were optimized. As shown in Figure , satisfactory recoveries in the range of 80–120% were obtained for 109 pesticides when 50 mg MWCNTs was used. Eight pesticides (e.g., chlorbenzuron, clofentezine, diflufenican, pendimethalin, phosmet, profenofos, rotenone, and spirodiclofen) seriously lost their recovery with the MWCNTs cleanup treatment. By comparison, 15 mg of MWCNTs was proposed because higher recovery of chlorbenzuron, clofentezinethe, and phosmet weas observed. Finally, a mixture of 200 mg PSA, 100 mg C18, and 15 mg MWCNTs was used as IS-D-SPE adsorbent.

Figure 4

Recovery of 117 pesticides with 200 mg PSA, 100 mg C18, and different amounts of MWCNTs for cleanup of the tea matrix.

Method Validation

The obtained data of matrix effects are shown in Figure . Ion suppression with matrix effect value below 0 was observed for most analytical pesticides in black tea, green tea, and oolong tea. Pesticides with significant matrix effect below −20% from black tea, green tea, and oolong tea accounted for the percentage of 11%, 21%, and 26%, respectively. Generally, matrix suppression was stronger for unfermented tea (green tea) and semifermented tea (oolong tea) than fermented tea (black tea). Eleven pesticides, e.g., carbendazim (51–70%), dichlorvos (22–90%), dimethoate (20–30%), flonicamid (81–85%), methacrifos (22–36%), monocrotophos (49–51%), sebuthylazine (27%), simetryn (54–55%), spirotetramat (31–35%), thionphanate-methyl (43–66%), and trichlorfon (41–52%), showed strong ion suppression below −20%. Therefore, matrix matched calibration solutions were employed to establish the calibrated curves for each pesticide. Good linearity was achieved for all 117 pesticides with correlation coefficients (R2) higher than 0.991 (see Table S-3).

Figure 5

Evaluation of matrix effect in black tea (BT), green tea (GT), and oolong tea (OT).

Accuracy was evaluated in terms of recovery at spiked levels of 10 μg kg–1 (20 μg kg–1 for dimetachlone) and 50 μg kg–1, shown in Table . Except for 12 pesticides, e.g., chlorbenzuron (21–31%), chlorfluazuron (61–71%), chlorotoluron (46–52%), clofentezine (22–34%), diflubenzuron (40–57%), diflufenican (43–64%), fenazaquin (49–61%), imazalil (49–68%), pendimethalin (33–48%), spirodiclofen (50–58%), spirotetramat (52–61%), and tricyclazole (42–53%), recoveries were in the range of 70–110%. For different kinds of tea samples, there is little difference in recovery among black tea, green tea, and oolong tea. Physicochemical properties of analytical pesticides were responsible for the loss of recovery, especially due to the adsorption of MWCNTs during IS-D-SPE cleanup.

Table 1

Recoveries % (n = 5) and Precisiona of 117 Pesticides from Black Tea, Green Tea, and Oolong Tea

		level of fortification, μg kg–1
		black tea		green tea		oolong tea
number	pesticide	10	50	10	50	10	50	interday RSD,%
1	Acetamiprid	83 (9.5)	88 (5.6)	75 (11.5)	78 (3.2)	83 (6.3)	86 (5.9)	9.6
2	Acetochlor	94 (8.7)	96 (3.3)	92 (7.9)	88 (3.1)	87 (10.6)	91 (5.7)	6.7
3	Allethrin	94 (14.7)	92 (9.7)	81 (8.9)	88 (6.3)	85 (11.6)	93 (8.5)	10.5
4	Ametryn	89 (6.3)	94 (4.7)	78 (3.2)	83 (5.1)	87 (7.6)	84 (5.2)	4.8
5	Atraton	85 (5.5)	87 (4.9)	81 (7.1)	82 (6.8)	78 (8.8)	86 (7.3)	6.2
6	Atrazine	92 (3.6)	96 (5.1)	82 (5.6)	90 (7.3)	85 (2.5)	93 (4.7)	4.9
7	Azoxystrobin	95 (4.6)	91 (2.7)	92 (1.8)	94 (3.2)	85 (3.2)	95 (1.9)	3.7
8	Buprofezin	87 (2.1)	90 (1.7)	89 (5.4)	87 (7.3)	93 (3.4)	94 (4.1)	1.8
9	Butachlor	87 (1.8)	91 (0.7)	86 (3.2)	76 (5.6)	82 (1.7)	81 (3.6)	2.8
10	Carbaryl	83 (14.6)	87 (7.5)	94 (16.3)	83 (10.5)	90 (13.7)	79 (7.6)	9.7
11	Carbendazim	68 (6.2)	73 (4.8)	66 (6.9)	69 (7.5)	68 (8.1)	73 (7.6)	5.9
12	Carbofuran	81 (2.3)	93 (3.1)	85 (1.7)	90 (1.9)	78 (4.3)	84 (3.8)	3.8
13	Chlorantraniliprole	84 (10.7)	88 (5.6)	76 (9.8)	73 (7.5)	80 (6.9)	84 (5.7)	8.2
14	Chlorbenzuron	21 (15.8)	26 (8.8)	28 (16.3)	30 (7.5)	31 (13.9)	27 (6.8)	8.6
15	Chlorfluazuron	61 (12.7)	65 (6.4)	61 (11.6)	71 (7.0)	65 (9.9)	63 (6.3)	9.1
16	Chlorotoluron	46 (7.9)	49 (8.5)	52 (7.5)	47 (6.2)	48 (7.2)	49 (6.7)	8.1
17	Chlorpyrifos	85 (8.6)	82 (5.1)	80 (2.8)	84 (2.6)	77 (1.9)	86 (1.1)	3.0
18	Chlorpyrifos-methyl	88 (8.5)	91 (4.3)	89 (7.6)	96 (4.1)	80 (3.6)	85 (5.3)	6.7
19	Chromafenozide	92 (9.6)	97 (3.4)	83 (7.5)	92 (8.6)	92 (9.7)	94 (4.2)	8.4
20	Clofentezine	22 (8.9)	27 (6.8)	32 (9.7)	34 (6.1)	25 (11.6)	23 (6.6)	9.3
21	Coumphos	82 (1.7)	84 (2.5)	90 (3.0)	96 (2.1)	85 (0.8)	89 (3.1)	2.7
22	Cyanazine	80 (9.4)	83 (6.2)	84 (8.5)	87 (7.7)	78 (6.9)	80 (7.8)	7.5
23	Diazinon	89 (0.8)	93 (1.5)	92 (2.7)	95 (1.5)	79 (1.7)	86 (2.4)	2.4
24	Dichlorvos	83 (8.5)	88 (4.9)	86 (7.6)	95 (6.4)	81 (6.9)	89 (7.8)	8.6
25	Difenoconazole	95 (2.3)	92 (2.7)	93 (3.9)	91 (2.1)	88 (4.2)	90 (5.9)	4.3
26	Diflubenzuron	40 (6.7)	42 (5.3)	54 (8.1)	57 (9.4)	41 (7.6)	53 (7.8)	8.1
27	Diflufenican	48 (7.9)	59 (4.7)	64 (6.3)	61 (5.1)	43 (8.3)	49 (3.7)	8.9
28	Dimetachlone	86b (9.7)	92 (7.6)	90b (11.6)	95 (7.9)	87b (14.3)	95 (9.8)	7.7
29	Dimethachlor	93 (3.7)	91 (2.6)	79 (4.8)	90 (1.5)	91 (5.0)	88 (1.7)	2.9
30	Dimethenamid	90 (1.1)	95 (2.0)	87 (2.5)	92 (1.4)	90 (2.4)	93 (1.7)	3.2
31	Dimethoate	82 (5.7)	85 (4.6)	79 (6.2)	80 (4.1)	80 (6.7)	82 (3.9)	5.9
32	Dimethomorph	100 (8.5)	97 (3.5)	101 (6.1)	99 (4.9)	99 (5.8)	100 (4.2)	7.1
33	Diuron	87 (8.6)	85 (3.5)	72 (9.6)	85 (3.7)	83 (8.3)	78 (5.2)	6.8
34	Epoxiconazole	89 (1.1)	91 (3.2)	92 (4.6)	95 (2.7)	89 (5.7)	87 (3.2)	4.6
35	Ethiofencarb	87 (16.4)	90 (5.3)	79 (9.8)	72 (6.8)	82 (15.7)	87 (9.6)	8.7
36	Ethion	95 (3.2)	95 (4.7)	93 (1.8)	86 (2.5)	82 (4.6)	84 (1.9)	3.3
37	Ethoprop	96 (1.7)	95 (0.8)	93 (2.3)	96 (3.1)	92 (2.8)	96 (1.8)	2.9
38	Fenazaquin	57 (5.3)	53 (2.1)	58 (2.2)	61 (4.3)	49 (2.7)	57 (1.7)	4.1
39	Fenobucarb	89 (1.8)	95 (1.1)	93 (2.4)	98 (1.6)	94 (3.2)	93 (1.8)	3.7
40	Flonicamid	72 (4.3)	82 (5.1)	79 (2.9)	77 (7.4)	80 (6.3)	76 (2.1)	6.5
41	Flufenacet	93 (8.6)	96 (4.3)	92 (9.6)	90 (3.9)	91 (9.6)	89 (7.9)	8.2
42	Flusilazole	91 (2.3)	97 (5.4)	97 (4.8)	100 (1.9)	94 (5.3)	96 (4.9)	4.7
43	Fonofos	88 (2.7)	104 (3.2)	91 (1.8)	98 (2.5)	96 (2.5)	91 (1.6)	3.7
44	Furalaxyl	86 (1.3)	94 (2.7)	84 (0.9)	88 (1.5)	90 (3.2)	88 (1.6)	3.2
45	Hexaconazole	83 (2.8)	90 (3.1)	90 (5.4)	92 (2.9)	91 (6.4)	89 (1.8)	4.8
46	Hexazinone	79 (7.5)	90 (7.1)	82 (3.7)	87 (6.8)	78 (5.3)	85 (4.9)	7.6
47	Imazalil	52 (1.8)	68 (1.0)	62 (4.3)	65 (1.7)	49 (3.2)	64 (0.8)	3.8
48	Indoxacarb	97 (3.5)	101 (4.7)	105 (5.1)	103 (1.4)	95 (2.6)	103 (3.2)	4.9
49	Imidacloprid	80 (10.6)	88 (8.9)	73 (13.2)	76 (9.7)	72 (11.7)	81 (10.9)	12.1
50	Iprobenphos	84 (11.2)	94 (3.6)	91 (10.8)	86 (4.8)	90 (9.7)	83 (6.1)	9.0
51	Isazophos	94 (2.1)	98 (4.3)	89 (0.9)	93 (4.3)	92 (2.1)	95 (3.7)	4.8
52	Isoproturon	90 (1.5)	95 (0.5)	82 (2.5)	89 (1.5)	83 (3.6)	82 (2.4)	3.7
53	Kadethrin	95 (1.1)	95 (4.2)	90 (5.3)	103 (2.6)	94 (4.8)	91 (7.9)	8.5
54	Linuron	93 (16.7)	96 (11.2)	99 (7.9)	104 (9.4)	94 (10.5)	93 (7.3)	13.4
55	Malathion	100 (8.5)	99 (4.7)	97 (8.6)	102 (7.3)	97 (9.5)	99 (3.7)	7.8
56	Mefenacet	90 (3.7)	95 (5.8)	91 (3.6)	88 (2.1)	92 (6.9)	91 (9.4)	6.6
57	Metaflumizone	73 (7.8)	79 (8.1)	83 (5.9)	89 (10.5)	73 (8.6)	83 (4.8)	7.9
58	Metalaxyl	98 (2.1)	96 (3.2)	87 (2.8)	84 (3.7)	80 (3.8)	85 (2.6)	4.3
59	Metazachlor	91 (8.5)	97 (6.3)	83 (5.8)	80 (4.7)	94 (5.5)	96 (6.3)	5.9
60	Methabenzthiazuron	94 (6.8)	97 (5.9)	80 (4.7)	89 (5.1)	92 (8.5)	97 (4.3)	8.9
61	Methacrifos	91 (7.8)	97 (8.3)	94 (10.5)	81 (8.4)	90 (9.7)	97 (9.5)	10.8
62	Methidathion	90 (8.5)	95 (4.7)	82 (5.9)	79 (6.8)	92 (9.6)	89 (5.7)	9.6
63	Methoxyfenozide	84 (7.9)	97 (9.5)	89 (8.3)	80 (6.4)	83 (7.9)	95 (7.3)	9.5
64	Metobromuron	90 (6.4)	94 (7.8)	88 (10.5)	93 (4.6)	90 (9.5)	95 (3.6)	8.9
65	Metolachlor	95 (2.5)	95 (1.9)	92 (5.3)	102 (5.8)	92 (2.5)	97 (3.1)	6.3
66	Metolcarb	72 (8.5)	81 (6.3)	74 (9.0)	71 (6.6)	69 (6.3)	74 (6.9)	9.1
67	Metoxuron	73 (8.6)	76 (4.3)	70 (6.5)	77 (6.8)	72 (8.4)	70 (3.2)	7.7
68	Monalide	89 (2.1)	95 (7.6)	96 (4.6)	99 (5.7)	94 (2.6)	92 (6.7)	8.3
69	Monocrotophos	88 (11.5)	85 (6.9)	87 (13.6)	77 (10.5)	87 (13.5)	89 (11.3)	13.2
70	Monolinuron	83 (5.3)	91 (4.7)	85 (6.8)	89 (5.8)	83 (3.1)	89 (6.3)	8.6
71	Myclobutanil	99 (3.6)	101 (2.5)	89 (6.8)	99 (6.4)	95 (5.7)	98 (5.0)	7.3
72	Napropamide	90 (14.6)	96 (12.5)	93 (15.0)	91 (10.6)	83 (11.7)	90 (16.7)	11.6
73	Nitenpyram	86 (6.4)	82 (5.8)	75 (8.4)	83 (6.0)	74 (9.1)	72 (6.6)	8.6
74	Penconazole	87 (2.5)	92 (3.8)	91 (4.6)	93 (1.5)	99 (2.0)	90 (4.8)	5.8
75	Pendimethalin	44 (0.8)	46 (4.3)	40 (3.5)	48 (6.2)	33 (4.4)	37 (5.0)	7.3
76	Pethoxamid	87 (1.1)	95 (3.2)	92 (2.8)	89 (3.6)	84 (3.5)	91 (2.1)	4.5
77	Phorate sulfone	102 (6.3)	99 (7.2)	97 (5.8)	100 (4.9)	105 (5.0)	94 (6.3)	7.9
78	Phosalone	88 (3.5)	91 (6.1)	85 (6.8)	93 (9.3)	80 (3.6)	89 (3.9)	8.0
79	Phosmet	78 (5.6)	74 (3.6)	72 (6.2)	71 (5.8)	70 (6.4)	73 (7.6)	6.9
80	Phosphamidon	84 (5.8)	82 (8.7)	80 (5.3)	85 (2.0)	82 (8.3)	80 (6.9)	7.3
81	Phoxim	91 (3.7)	94 (0.8)	85 (4.6)	94 (6.1)	90 (0.6)	97 (4.5)	5.5
82	Pirimicarb	83 (0.6)	84 (3.5)	82 (6.3)	79 (3.5)	75 (2.4)	77 (5.1)	6.0
83	Pirimiphos-ethyl	92 (1.3)	90 (2.7)	88 (3.6)	91 (5.2)	81 (4.2)	85 (4.0)	5.4
84	Pirimiphos-methyl	93 (3.2)	91 (6.3)	92 (4.6)	95 (5.3)	82 (5.7)	83 (6.6)	7.1
85	Pretilachlor	96 (8.5)	95 (7.5)	100 (9.6)	102 (4.3)	97 (10.5)	92 (10.9)	9.4
86	Prochloraz	89 (7.9)	93 (6.3)	88 (9.7)	82 (10.0)	83 (6.6)	81 (7.4)	8.2
87	Profenofos	87 (10.5)	89 (3.6)	89 (5.3)	93 (5.7)	82 (7.3)	82 (8.1)	7.2
88	Prometryn	92 (2.0)	95 (1.7)	90 (3.6)	93 (5.8)	89 (3.5)	90 (3.9)	4.8
89	Propachlor	89 (3.6)	94 (6.8)	81 (9.5)	86 (4.2)	83 (6.9)	86 (11.1)	9.8
90	Propanil	88 (8.7)	85 (10.3)	88 (5.7)	98 (8.5)	87 (6.8)	89 (3.6)	8.6
91	Propargite	79 (4.4)	85 (3.7)	72 (5.1)	83 (6.5)	78 (5.3)	80 (4.5)	5.4
92	Propazine	89 (0.4)	95 (4.6)	93 (2.6)	99 (5.3)	90 (6.3)	98 (2.7)	3.6
93	Propoxur	83 (5.6)	91 (7.1)	84 (8.5)	88 (8.0)	84 (6.4)	81 (3.4)	6.1
94	Pyraclostrobine	79 (1.3)	83 (5.6)	88 (6.7)	86 (3.5)	79 (8.3)	81 (2.5)	7.1
95	Pyridaben	70 (3.6)	73 (3.1)	73 (7.8)	66 (4.4)	65 (2.0)	69 (4.7)	5.6
96	Pyrimethanil	71 (1.6)	75 (2.6)	76 (3.1)	78 (5.2)	70 (4.3)	73 (4.8)	6.3
97	Quinalphos	90 (0.7)	94 (3.6)	85 (4.3)	95 (5.2)	88 (4.3)	94 (4.8)	3.7
98	Rotenone	80 (6.7)	73 (3.6)	78 (6.2)	82 (5.4)	80 (7.1)	82 (7.7)	8.6
99	Secbumeton	91 (3.2)	94 (3.0)	82 (5.3)	87 (2.7)	81 (2.8)	83 (6.1)	6.7
100	Simazine	89 (4.1)	88 (6.2)	87 (5.8)	92 (5.4)	83 (3.1)	85 (3.9)	5.7
101	Simetryn	83 (6.2)	80 (2.9)	77 (5.5)	72 (6.0)	75 (2.1)	74 (4.2)	6.7
102	Spirodiclofen	53 (8.6)	50 (4.6)	55 (10.5)	58 (7.8)	53 (11.6)	55 (9.5)	9.9
103	Spirotetramat	53 (7.9)	58 (6.3)	53 (11.6)	61 (8.4)	55 (6.4)	52 (7.7)	8.5
104	Sulfotepp	91 (3.2)	96 (2.2)	94 (1.9)	92 (4.7)	90 (3.2)	93 (4.4)	7.1
105	Sumithrin	72 (9.5)	74 (6.7)	70 (5.6)	74 (7.3)	66 (6.9)	73 (7.4)	9.1
106	Tebufenozide	94 (13.6)	96 (8.4)	91 (9.6)	90 (12.6)	95 (11.9)	92 (8.9)	9.3
107	Temephos	91 (6.7)	94 (7.1)	95 (6.9)	107 (4.3)	92 (6.8)	89 (5.6)	7.5
108	Tetrachlorvinphose	96 (8.9)	98 (7.4)	98 (9.3)	104 (6.7)	97 (5.5)	95 (6.8)	8.5
109	Tetramethrin	83 (6.7)	92 (8.3)	95 (7.3)	88 (5.6)	91 (8.3)	94 (5.7)	9.2
110	Thiacloprid	71 (9.5)	78 (7.8)	70 (6.9)	73 (6.3)	67 (8.1)	69 (5.2)	9.3
111	Thiodicarb	92 (5.7)	89 (7.5)	80 (6.3)	78 (6.9)	83 (10.6)	82 (8.5)	8.7
112	Thiophanate	90 (8.7)	95 (7.4)	90 (9.1)	87 (11.6)	95 (5.8)	92 (7.3)	8.1
113	Thiophanate-methyl	81 (9.8)	84 (6.7)	84 (7.5)	82 (11.3)	76 (8.9)	78 (9.3)	10.8
114	Triadimefon	98 (6.5)	101 (7.5)	97 (8.2)	101 (3.8)	100 (6.7)	98 (5.8)	9.0
115	Triazophos	93 (1.7)	97 (1.8)	97 (4.3)	96 (2.5)	96 (4.8)	95 (5.8)	6.3
116	Trichlorfon	75 (15.6)	78 (11.8)	82 (13.4)	88 (14.1)	83 (9.5)	74 (11.8)	12.7
117	Tricyclazole	53 (3.7)	53 (2.7)	43 (6.3)	48 (3.4)	45 (3.0)	42 (5.7)	7.5

Intraday precision, expressed as RSD (%), is given in brackets (n = 5) and interday precision (RSD %) was evaluated using spiked green tea samples at 50 μg kg–1 during 5 consecutive days.

Fortification level at 20 μg kg–1.

Precision was measured employing intraday repeatability and interday reproducibility testing. Intraday repeatability was evaluated at the fortified levels of 10 μg kg–1 and 50 μg kg–1 during the same day. In this case, the repeatability, expressed as RSDs, was less than 17% (see Table ) for all analytical pesticides. Interday reproducibility was evaluated analyzing five blank green tea samples spiked with 50 μg kg–1 of 117 pesticides during 5 consecutive days. Interday RSDs for all analytical pesticides were below 18%.

LODs were estimated analyzing three blank samples spiked at 0.5, 1.0, 2.0, 5.0, and 10.0 μg kg–1. They are defined as the minimum concentration at which the accurate mass error was ±5 ppm,[42] and LODs were less than or equal to 5 μg kg–1 for all pesticides except dimetachlone at 10 μg kg–1 (see Table S-3). According to accuracy and precision measurement, LOQs were 10 μg kg–1 for 104 pesticides which provided recoveries between 70% and 130%, and RSDs below 20%. LOQs of dimetachlone were 20 μg kg–1 due to poor sensitivity. Although low recoveries for 9 pesticides, including chlorfluazuron, chlorotoluron, diflubenzuron, diflufenican, fenazaquin, imazalil, spirodiclofen, spirotetramat, and tricyclazole, were in the range of 40–70%, these pesticides could be quantified using this method according to European Council No SANTE/11945/2015,[26] where the document shows that a mean recovery below 70% may be acceptable when the method is quite consistent (with a good precision). Due to the low recovery below 40%, this method was not suitable for quantification of chlorbenzuron, clofentezine, and pendimethalin in tea.

Real Sample Analysis

A total of 70 tea samples (30 green tea samples, 20 black tea samples, and 20 oolong tea samples) purchased from local markets in Zhejiang province were analyzed using this proposed UPLC-Q-Orbitrap-MS method. For each batch of less than 20 tea samples, quality control samples spiked with 117 pesticides at 10 and 50 μg kg–1 were analyzed to evaluate the characteristics of the analytical method. Target analyte identification was based on retention time window ±0.2 min and mass accuracy ±5 ppm. To confirm positive samples, characteristic fragments and isotopic pattern were monitored using UPLC-Q-Orbitrap-MS in full scan/dd-MS2 mode. Table shows the frequency and concentrations of 11 detected pesticides, including acetamiprid, buprofezin, carbaryl, carbendazim, chlorpyrifos, difenoconazole, indoxacarb, imidacloprid, myclobutail, pyridaben, and triazophos. Higher frequencies above 10% occurred for two neonicotinoid pesticides (acetamiprid and imidacloprid), buprofezin, carbedazim, and pyredaben. Multi residues containing over 3 pesticides in the same sample were observed for green tea (9.2%), black tea (2.8%), and oolong tea (15.9%). Figure shows an oolong tea containing actamiprid (5 μg kg–1), chlorpyrifos (0.3 μg kg–1), imidacloprid (79 μg kg–1), and triazophos (0.1 μg kg–1). None of the 117 pesticides was higher than EU maximum residue levels (MRLs), CAC MRLs and China MRLs.

Table 2

Frequency (F, %) and Residue Ranges (R, μg kg–1) of 11 Pesticides in Green Tea, Black Tea, and Oolong Tea Samples

	Acetamiprid		Buprofezin		Carbaryl		Carbendazim		Chlorpyrifos		Difenoconazole		Imdoxacxarb		Imidacloprid		Myclobutanil		Pyridaben		Triazophos
samples	F	R	F	R	F	R	F	R	F	R	F	R	F	R	F	R	F	R	F	R	F	R
Green tea (30)	13	12–159	6.7	24–47	0	0	20	11–219	13	10–49	0	0	7	11–27	17	11–69	0	0	20	11–36	10	11–23
Black tea(20)	10	10–16	10	13–27	10	42–274	0	0	0	0	5	121	0	0	0	0	5	38	0	0	0	0
Oolong tea(20)	70	12–90	40	18–93	5	11	20	24–106	10	11–12	0	0	15	10–16	25	13–40	5	25	45	17–106	5	26
Total (70)	29	10–159	13	13–93	4	11–274	14	11–219	9	10–49	1	121	9	10–27	14	11–69	3	25–38	21	11–106	6	11–26

Figure 6

Extracted ion chromatograms of imidacloprid (m/z 256.0596) at 79 μg kg–1, acetamiprid (m/z 223.0745) at 5 μg kg–1, triazophos (m/z 314.0725) at 0.1 μg kg–1, and chlorpyrifos (m/z 349.9333) at 0.3 μg kg–1 detected in an oolong tea sample. (note: Quantification by single point calibration using matrix matched calibration solution at 1 μg L–1.)

Conclusions

A high throughput method based on in-syringe dispersive solid phase extraction and UPLC-Q-Orbitrap MS was developed for rapid analysis of multiple pesticide residues in green tea, black tea, and oolong tea. A mixture of adsorbents containing PSA, C18, and MWCNTs was used as IS-D-SPE adsorbents to reduce tea matrices. UPLC-Q-Orbitrap-MS at full scan monitoring mode provided high method selectivity, accuracy, and precision. Full mass/dd-MS2 monitoring mode is necessary to identify the compounds with the same chemical formula. Matrix suppression was more serious for unfermented tea and semifermented tea than fermented tea. This newly developed method allows high sample preparation thanks to its simplicity and effectiveness. Compared with previous analytical methods, the advantages of this proposed method are less time-consuming, solvent-consuming, lower cost, and higher throughout analysis. Therefore, this method is suitable for routine analysis and identification of multiple pesticides in various types of tea samples. This method was applied to analyze 70 tea samples. High frequency of two neonicotinoid pesticides acetamiprid and imidacloprid, buprofezin, carbendazim, and pyredaben was found.