Determination of 107 Pesticide Residues in Wolfberry with Acetate-buffered Salt Extraction and Sin-QuEChERS Nano Column Purification Coupled with Ultra Performance Liquid Chromatography Tandem Mass Spectrometry

A multi-residue method for the determination of 107 pesticide residues in wolfberry has been developed and validated. Similar pretreatment approaches were compared, and the linearity, matrix effect, analysis limits, precision, stability and accuracy were validated, which verifies the satisfactory performance of this new method. The LODs and LOQs were in the range of 0.14–1.91 µg/kg and 0.46–6.37 µg/kg, respectively. The recovery of analytes at three fortification levels (10 µg/kg, 50 µg/kg, 100 µg/kg) ranged from 63.3–123.0%, 72.0–118.6% and 67.0–118.3%, respectively, with relative standard deviations (RSDs) below 15.0%. The proposed method was applied to the analysis of fifty wolfberry samples collected from supermarkets, pharmacies and farmers’ markets in different cities of Shandong Province. One hundred percent of the samples analyzed included at least one pesticide, and a total of 26 pesticide residues was detected in fifty samples, which mainly were insecticides and bactericide. Several pesticides with higher detection rates were 96% for acetamiprid, 82% for imidacloprid, 54% for thiophanate-methyl, 50% for blasticidin-S, 42% for carbendazim, 42% for tebuconazole and 36% for difenoconazole in wolfberry samples. This study proved the adaptability of the developed method to the detection of multiple pesticide residues in wolfberry and provided basis for the research on the risks to wolfberry health.


Introduction
Lycium barbarum, belonging to the Solanaceous family, also called Gouqizi in Chinese or wolfberry in English, has been widely cultivated in the northwest of China and the Mediterranean region [1,2]. Wolfberry is a kind of traditional Chinese medicine (TCM) and food homologous plant with health-protection functions. It has been widely used for the prevention and treatment of various diseases for more than thousands of years [3,4]. In recent years, wolfberry, a functional dietary supplement with nutritive value, has become increasingly prevalent in the Western countries and regions [5]. Wolfberry is rich in minerals, proteins, polysaccharides, amino acids, carotenoids, flavonoids and so on [6]. This characteristic makes it vulnerable to diseases and insect pests during its growth. Rust mite, gall mite, psylla, mealybug, aphid, anthracnose and powdery mildew are often encountered at the same time or alternately. As chemical agents for diseases and insect pests control as well as plant growth liquid storage tank is discharged through the vent hole. The Sin-QuEChERS Nano purification column is reversely guided in the downward process of the centrifuge tube, and the purified extract is stored in the liquid storage tank, then the built-in water blocking filter is bonded to the strong hydrophobic water blocking group to ensure that the aqueous solution does not contact with the purification packing and reservoir. The seal is used for sealing between the cylinder and the centrifuge tube.
In this study, in order to monitor the quality of wolfberry in China, a method with acetate-buffered salt extraction and Sin-QuEChERS nanocolumn purification coupled with ultra performance liquid chromatography tandem mass spectrometry was developed for the determination of pesticide residues in wolfberry samples. Different versions of buffer salts for sample extraction and several SPE cartridges for clean-up were compared. The linearity, matrix effect, analysis limits, precision, stability and accuracy were validated in detail. Finally, the method was applied to determinate 50 wolfberry samples to validate the feasibility.

Optimization and Comparison of the Extraction Procedure
In this study, three versions of buffer salts were compared, and the extraction efficiency was evaluated using three levels (10, 25, 50 µg/kg) spiked recovery with five parallel samples. The results, were in accordance with previous studies [54], where most pesticides except for pH-sensitive ones gave excellent results when extracted with three different versions of buffer salts. As shown in Figure 1, the QuEChERS version using acetate buffering or citrate buffering more often gave higher recoveries compared with the unbuffered method for pH-dependent pesticides. For acephate, dimethoate, fenobucarb, methamidophos, omethoate and profenofos that are unstable in alkaline medium, it is easier to obtain higher recovery in buffer salt system with the pH of the matrix was maintained between 5.0 and 5.5 throughout the experiment. In addition to, as Figure 1 results demonstrate, the recovery rate of pesticides in acetate buffer version (AOAC) is slightly higher than that in citrate buffer version (CEN), but the difference is not obvious. Further, referring to previous research results [42,54], the ionization efficiency of acetonitrile and the ability of the matrix itself to interfere with acidity and alkalinity is enhanced in the acetate-buffered version, so subsequent experiments used acetate buffer version.
Molecules 2019, 24, x FOR PEER REVIEW 3 of 23 during the purification, the liquid level rises, and the air in the liquid storage tank is discharged through the vent hole. The Sin-QuEChERS Nano purification column is reversely guided in the downward process of the centrifuge tube, and the purified extract is stored in the liquid storage tank, then the built-in water blocking filter is bonded to the strong hydrophobic water blocking group to ensure that the aqueous solution does not contact with the purification packing and reservoir. The seal is used for sealing between the cylinder and the centrifuge tube.
In this study, in order to monitor the quality of wolfberry in China, a method with acetatebuffered salt extraction and Sin-QuEChERS nanocolumn purification coupled with ultra performance liquid chromatography tandem mass spectrometry was developed for the determination of pesticide residues in wolfberry samples. Different versions of buffer salts for sample extraction and several SPE cartridges for clean-up were compared. The linearity, matrix effect, analysis limits, precision, stability and accuracy were validated in detail. Finally, the method was applied to determinate 50 wolfberry samples to validate the feasibility.

Optimization and Comparison of the Extraction Procedure
In this study, three versions of buffer salts were compared, and the extraction efficiency was evaluated using three levels (10, 25, 50 µ g/kg) spiked recovery with five parallel samples. The results, were in accordance with previous studies [54], where most pesticides except for pH-sensitive ones gave excellent results when extracted with three different versions of buffer salts. As shown in Figure  1, the QuEChERS version using acetate buffering or citrate buffering more often gave higher recoveries compared with the unbuffered method for pH-dependent pesticides. For acephate, dimethoate, fenobucarb, methamidophos, omethoate and profenofos that are unstable in alkaline medium, it is easier to obtain higher recovery in buffer salt system with the pH of the matrix was maintained between 5.0 and 5.5 throughout the experiment. In addition to, as Figure 1 results demonstrate, the recovery rate of pesticides in acetate buffer version (AOAC) is slightly higher than that in citrate buffer version (CEN), but the difference is not obvious. Further, referring to previous research results [42,54], the ionization efficiency of acetonitrile and the ability of the matrix itself to interfere with acidity and alkalinity is enhanced in the acetate-buffered version, so subsequent experiments used acetate buffer version.

Optimization and Comparison of the Clean-up Procedure
Five kinds of SPE cartridges were selected for optimization and comparison of the purification procedure, and were activated by 4 mL methanol before use. Extraction solution obtained as described in Section 3.3 was passed through the SPE cartridges and 4 mL of eluent was collected for the next process which involved treatment the same as the filtrate in Section 3.3. As the Figure 2 results demonstrate vividly, compared with the other three SPE cartridges, the solution purified by the Sin-QuEChERS Nano and HyperSep NH2 catridges showed a colorless and transparent state. PSA was used for adsorbing pigments, organic acids, some carbohydrates and fatty acids while C18 and MWCNTs absorbed the pigments. The Sin-QuEChERS Nano cartridge is packed with a certain

Optimization and Comparison of the Clean-up Procedure
Five kinds of SPE cartridges were selected for optimization and comparison of the purification procedure, and were activated by 4 mL methanol before use. Extraction solution obtained as described in Section 3.3 was passed through the SPE cartridges and 4 mL of eluent was collected for the next process which involved treatment the same as the filtrate in Section 3.3. As the Figure 2 results demonstrate vividly, compared with the other three SPE cartridges, the solution purified by the Sin-QuEChERS Nano and HyperSep NH 2 catridges showed a colorless and transparent state. PSA was used for adsorbing pigments, organic acids, some carbohydrates and fatty acids while C 18 and MWCNTs absorbed the pigments. The Sin-QuEChERS Nano cartridge is packed with a certain amount of MWCNTs in addition to PSA, which can better remove carotene, lutein and other colored interferents in wolfberry with less co-extracted compounds. Recoveries with the different purification processes are displayed in Figure 3, where most pesticides displayed higher recoveries for the Sin-QuEChERS Nano cartridge in comparison with other SPE columns, while all pesticides also gave satisfactory recoveries when purified by HyperSep NH 2 columns.  Figure 3, where most pesticides displayed higher recoveries for the Sin-QuEChERS Nano cartridge in comparison with other SPE columns, while all pesticides also gave satisfactory recoveries when purified by HyperSep NH2 columns.  amount of MWCNTs in addition to PSA, which can better remove carotene, lutein and other colored interferents in wolfberry with less co-extracted compounds. Recoveries with the different purification processes are displayed in Figure 3, where most pesticides displayed higher recoveries for the Sin-QuEChERS Nano cartridge in comparison with other SPE columns, while all pesticides also gave satisfactory recoveries when purified by HyperSep NH2 columns.   Some pesticides gave unsatisfactory recoveries (>120% or <70%) with the other three SPE cartridges owing to matrix enhancement or inhibition effects caused by matrix co-extracted compounds. Therefore, considering the recoveries, the purification effect and the simplicity of operation, Sin-QuEChERS Nano cartridges were considered the optimal choice for purification of wolfberry.
Under the optimal experimental conditions, the lowest concentration or the lowest content of the target component can be detected is considered as the detection limit (LOD). The limit of quantitation (LOQ) is the lowest concentration or the lowest quantity of the components to be measured in the sample by analytical method. The LOD and LOQ for each analytes were determined at the lowest concentration at a signal-to-noise ratio (S/N) of 3 and 10, respectively. The results were shown in Table 1 that the LOD and LOQ were 0.14-1.91 µg/kg and 0.46-6.37 µg/kg, respectively.

Precision and Stability
In this study, the stability and precision of the established method were tested. The precision was evaluated by measuring the intra-day and inter-day variations in the relative standard deviation (RSD%) of the peak area of each analyte. In order to estimate intra-day precision, the blank spiked samples at concentration levels of 10 and 100 µg/kg were pretreated and analyzed in one day, with each spike level with five parallel samples. Inter-day precision was achieved by measuring the spiked samples at concentration levels of 10 and 100 µg/kg once a day within five consecutive days. As shown in Table 2, the relative standard deviation values (RSD) of intra-day were 0.6-9.4% (10 µg/kg) and 0.8-9.1% (100 µg/kg), inter-day were 0.8-12.6% (10 µg/kg) and 1.5-11.7% (100 µg/kg). The stability was evaluated by injecting the same volume of wolfberry spiked samples (100 µg/kg) into the UPLC/MS/MS system at 0, 2, 4, 8, 12, 18 and 24 h under the same condition in one day. The RSD values were less than 11.7 for all analytes (Table 2), which indicating that the sample solutions were stable and unchanged.

Accuracy
The accuracy of this method was tested by the blank spike recovery experiment of 107 pesticides. And the spiking levels were 10, 50, and 100 µg/kg, respectively, meanwhile the determination had five parallel samples for each spiking level. Average recoveries of 107 kinds of pesticides at three fortification levels using the analytical procedure, as presented in Table 2, were 63.3-123.0%, 72.0-118.6%, and 67.0-118.3%, respectively, which were within the range of the acceptable values. Under the premise of ensuring the accuracy and reproducibility of the results, a better recovery rate can be obtained, which is fully in line with the AOAC 2007.01 and EN 15,662 standards. All samples were analyzed on the same day and the RSD values below 15.0% at the three fortification levels indicated the accuracy was acceptable.

Matrix Effects
Matrix effects are caused by the co-elution of matrix constituents that play an important role in multi-residue analysis of pesticides, which can affect the ionization efficiency of target pesticides [55,56] and then influence the quantitative results by matrix enhancement and attenuation effects caused by quality of chromatographic separation, the ionization type, the amount and the type of the sample matrix, and sample preparation procedure [57,58]. For complex matrix like wolfberry is wealthy in mineral substance, proteins, polysaccharose, amino acids, carotinoid, flavonoids and so on [6], it is critical to eliminate or attenuate matrix effects by minimizing matrix co-extractives through the sample preparation procedure. In this research, the matrix effect (ME) was calculated by the following equation: where A Matrix is the peak area of matrix standard sample and A s is the peak area of pure solvent standard sample. The matrix effect of the individual pesticide was studied at the level of 10 µg/kg with five parallel samples, which was deemed to be ignored if the ME value is between 90% and 110%, while it was regarded to be matrix suppression or enhancement effect when the value was less than 90% or greater than 110%, respectively [55,58,59]. The results showed that 71.96% of the pesticides presented a negligible ME, whereas 27.10% of the analytes presented matrix suppression effect, and only one pesticide (carbofuran) showed a matrix enhancement effect. Table 2 shows the specific ME values of each pesticide. Most of the compounds showed negligible matrix effect or mild matrix inhibition effect, suggesting that this method was suitable for the determination of 107 pesticide residues in wolfberry. However, there are still several pesticides with strong matrix effect: ME < 80% (boscalid, hexythiazox, metconazole) or ME > 120% (carbofuran). Therefore, to overcome and compensate for these matrix effects, matrix-matched standard curves were used in the quantitative analysis.

Real Samples Analysis
The applicability of the developed method was evaluated by analyzing a total of fifty wolfberry samples collected from supermarkets, pharmacies and farmers' markets in different cities of Shandong Province and three replicates of each sample were analysed. One hundred percent of the samples analyzed included at least one pesticide, and a total of 26 pesticide residues was detected in fifty samples, which mainly were insecticides and bactericides. Ninety-six percent of the samples was found to be acetamiprid with concentration ranging from 19.54 to 63.15 µg/kg, and imidacloprid was detected in eighty-two percent of the samples with concentration ranging from 7.80 to 60.26 µg/kg. In addition, several pesticides with higher detection frequency in wolfberry were thiophanate-methyl (54%), blasticidin-S (50%), carbendazim (42%), tebuconazole (42%) and difenoconazole (36%). The detection frequency of cyromazine and metalaxyl in wolfberry is only 8%, but the detection concentration is as high as 46.78-64.56 µg/kg and 64.75-86.22 µg/kg, respectively. Moreover, carbendazim (44.36 µg/kg), myclobutanil (51.95 µg/kg) and thiamethoxam (69.21 µg/kg) were found higher concentration in some samples occasionally. All pesticide residues detected in wolfberry were below the MRLs specified by the EU and China, except for pesticides without MRLs. The residue level of imidacloprid with a detection frequency of up to 82% was lower than the MRL (1 mg/kg) set by China (GB 2763-2016).

Chemicals and Reagents
Pesticide reference standards of all analytes were of purity > 98%, purchased from Aldrich-Sigma (Shanghai, China) and Dr. Ehrenstorfer GmbH (Augsburg, Germany). Each kind of pesticide standard substances was accurately labeled as 10 mg in a 100 mL volumetric flask, then dissolved with methanol and fixed to the scale line. A composite sample working standard solution was prepared by combining aliquots of each stock solution and diluting in methanol to obtain a final concentration of 10 mg/L, a series of matrix-matched standard solutions was obtained by gradually diluting the stock solution with blank matrix solution. All solutions were stored at 4 °C before use.

Instrument
An ACQUITY Quattro Premier XE system (Waters Corp., Milford, MA, USA) was used for analysis. Chromatographic separation was performed at 35 °C with an ACQUITY UPLC HSS T3 (1.8 μm, 2.1 × 100 mm, Waters Corp., Milford, MA, USA). The mobile phase A was 0.1% formic acid solution and the mobile phase B was 0.1% formic acid acetonitrile. Initial composition of the mobile phase was 80% of solvent A and 20% of solvent B, reaching the 95% of solvent B at 6 min. The concentration of B remained at 95% for 2 min before returning to the initial state in 1 min. Reequilibration of the column was performed for 1 min before the next injection was conducted. The flow rate of the eluent was 0.4 mL/min and the injection volume was 1 μL. The mass spectrometer was equipped with an electrospray ionization (ESI) interface operating at positive (ESI+) or negative (ESI−) mode, and MRM mode was adopted for date collection. The electrospray voltage, the

Chemicals and Reagents
Pesticide reference standards of all analytes were of purity > 98%, purchased from Aldrich-Sigma (Shanghai, China) and Dr. Ehrenstorfer GmbH (Augsburg, Germany). Each kind of pesticide standard substances was accurately labeled as 10 mg in a 100 mL volumetric flask, then dissolved with methanol and fixed to the scale line. A composite sample working standard solution was prepared by combining aliquots of each stock solution and diluting in methanol to obtain a final concentration of 10 mg/L, a series of matrix-matched standard solutions was obtained by gradually diluting the stock solution with blank matrix solution. All solutions were stored at 4 • C before use.

Instrument
An ACQUITY Quattro Premier XE system (Waters Corp., Milford, MA, USA) was used for analysis. Chromatographic separation was performed at 35 • C with an ACQUITY UPLC HSS T3 (1.8 µm, 2.1 × 100 mm, Waters Corp., Milford, MA, USA). The mobile phase A was 0.1% formic acid solution and the mobile phase B was 0.1% formic acid acetonitrile. Initial composition of the mobile phase was 80% of solvent A and 20% of solvent B, reaching the 95% of solvent B at 6 min. The concentration of B remained at 95% for 2 min before returning to the initial state in 1 min. Re-equilibration of the column was performed for 1 min before the next injection was conducted. The flow rate of the eluent was 0.4 mL/min and the injection volume was 1 µL. The mass spectrometer was equipped with an electrospray ionization (ESI) interface operating at positive (ESI+) or negative (ESI−) mode, and MRM mode was adopted for date collection. The electrospray voltage, the atomizing gas pressure, the auxiliary airflow speed and the ion source temperature were set as 3000 V, 7.0 Bar, 150 L/h and 400 • C, respectively. The 107 analytical parameters of liquid chromatography tandem mass spectrometry are listed in Table 1.

Sample Preparation
Wolfberry samples were collected from supermarkets, pharmacies and farmers'markets in different cities of Shandong Province, then the samples were ground into powder, passed through mesh screen (0.42 mm). The prepared samples were stored at 4°C and analysed withinn 24 h following the procedure described below. Wolfberry (pesticide-free) obtained from an organic production base was used as blank matrix for preparing standard curve and the recovery studies.
Five grams of ground wolfberry sample were weighed exactly into a 50 mL polypropylene centrifuge tube and soaked with 10 mL water for 5 min. Subsequently, 15 mL acetonitrile was added, next vortexed for 2 min, after following AOAC 2007. 01 extraction salt pack was added, after that vortexed for 1 min, succeeding centrifuged at 5000 rpm for 5 min. The Sin-QuEChERS Nano purification tube was taken into the 50 mL centrifuge tube, which was filled with extraction liquid, eventually the purification tube was slowly pressed down by an automatic apparatus to making up about 4 mL supernatant in the storage tank of the purification pipe. The structure and usage of Sin-QuEChERS Nano catridges were shown in Figure 5. The supernatant of 4 mL was accurately transferred to 10 mL centrifugal tube, dried with nitrogen gas stream and resolved with 1 mL methanol. Before UPLC/MS/MS analysis, the solution was filtered through a 0.22 µm filter (Dikma, Technologies). The flow chart of sample preparation procedure is presented in Figure 6. atomizing gas pressure, the auxiliary airflow speed and the ion source temperature were set as 3000 V, 7.0 Bar, 150 L/h and 400 °C, respectively. The 107 analytical parameters of liquid chromatography tandem mass spectrometry are listed in Table 1.

Sample Preparation
Wolfberry samples were collected from supermarkets, pharmacies and farmers'markets in different cities of Shandong Province, then the samples were ground into powder, passed through mesh screen (0.42 mm). The prepared samples were stored at 4 ℃ and analysed withinn 24 h following the procedure described below. Wolfberry (pesticide-free) obtained from an organic production base was used as blank matrix for preparing standard curve and the recovery studies.
Five grams of ground wolfberry sample were weighed exactly into a 50 mL polypropylene centrifuge tube and soaked with 10 mL water for 5 min. Subsequently, 15 mL acetonitrile was added, next vortexed for 2 min, after following AOAC 2007. 01 extraction salt pack was added, after that vortexed for 1 min, succeeding centrifuged at 5000 rpm for 5 min. The Sin-QuEChERS Nano purification tube was taken into the 50 mL centrifuge tube, which was filled with extraction liquid, eventually the purification tube was slowly pressed down by an automatic apparatus to making up about 4 mL supernatant in the storage tank of the purification pipe. The structure and usage of Sin-QuEChERS Nano catridges were shown in Figure 5. The supernatant of 4 mL was accurately transferred to 10 mL centrifugal tube, dried with nitrogen gas stream and resolved with 1 mL methanol. Before UPLC/MS/MS analysis, the solution was filtered through a 0.22 µ m filter (Dikma, Technologies). The flow chart of sample preparation procedure is presented in Figure 6.   atomizing gas pressure, the auxiliary airflow speed and the ion source temperature were set as 3000 V, 7.0 Bar, 150 L/h and 400 °C, respectively. The 107 analytical parameters of liquid chromatography tandem mass spectrometry are listed in Table 1.

Sample Preparation
Wolfberry samples were collected from supermarkets, pharmacies and farmers'markets in different cities of Shandong Province, then the samples were ground into powder, passed through mesh screen (0.42 mm). The prepared samples were stored at 4 ℃ and analysed withinn 24 h following the procedure described below. Wolfberry (pesticide-free) obtained from an organic production base was used as blank matrix for preparing standard curve and the recovery studies.
Five grams of ground wolfberry sample were weighed exactly into a 50 mL polypropylene centrifuge tube and soaked with 10 mL water for 5 min. Subsequently, 15 mL acetonitrile was added, next vortexed for 2 min, after following AOAC 2007. 01 extraction salt pack was added, after that vortexed for 1 min, succeeding centrifuged at 5000 rpm for 5 min. The Sin-QuEChERS Nano purification tube was taken into the 50 mL centrifuge tube, which was filled with extraction liquid, eventually the purification tube was slowly pressed down by an automatic apparatus to making up about 4 mL supernatant in the storage tank of the purification pipe. The structure and usage of Sin-QuEChERS Nano catridges were shown in Figure 5. The supernatant of 4 mL was accurately transferred to 10 mL centrifugal tube, dried with nitrogen gas stream and resolved with 1 mL methanol. Before UPLC/MS/MS analysis, the solution was filtered through a 0.22 µ m filter (Dikma, Technologies). The flow chart of sample preparation procedure is presented in Figure 6.

Conclusions
In this work, a simple and rapid method has been developed for the determination of 107 pesticide residues in wolfberry samples using UPLC/MS/MS analysis. The qualitative and quantitative analysis of 107 pesticides can be accomplished in 10 min at a time with one injection. The Sin-QuEChERS Nano purification column can not only effectively remove pigments, organic acids, alkaline interferents, fat and water, but also save sample preparation time and avoid the losses caused by solvent transfers. It greatly simplifies the purification process of the samples, while improving the detection efficiency and accuracy. The linearity, matrix effect, analysis limits, precision, stability and accuracy were validated in detail. The LODs and LOQs of the target pesticides obtained by this method were in the range of 0.14-1.91 µg/kg and 0.46-6.37 µg/kg, respectively. Under the premise of ensuring the accuracy and reproducibility of the results, the average recoveries (80%-120%) and RSD (<15%) of most target pesticides at the three spike levels were acceptable, which met the requirements of conventional pesticide residues screening. Fifty commercial wolfberry samples were tested for pesticide residues by the developed method, all of which were positive, and 26 pesticides were detected. These results demonstrated that the proposed method was sensitive, fast, simple and reliable for the simultaneous determination of 107 pesticide residues in wolfberry, furthermore did not require complex purification process to allow routine analysis of a large quantity of samples. Furthermore, the developed approach can further expand the types of target pesticides and be applied to the detection of pesticide residues in more other traditional Chinese medicine. This will be the focus of our future work.