Effect of Pretreatment on Detection of 37 Pesticide Residues in Chrysanthemum indicum

This study aimed to develop a method, followed by gas chromatography-mass spectrometry, for detecting 37 pesticides in Chrysanthemum indicum (C. indicum) and investigating the decrease in the matrix-induced enhancement effect. The influence of QuEChERS extraction and matrix solid-phase dispersion (MSPD) on the recovery and matrix effect (ME) was compared. extraction and matrix solid-phase dispersion (MSPD) on the recovery and matrix effect (ME) was compared to decrease the ME. The cleanup sorbents, volume and type of solvent, and treatment time were optimized. The accuracy (as recovery), precision (as relative standard deviation, RSD), linearity, limit of quantitation, and limit of detection were determined. The recoveries at the three levels using mixed standard solution ranged between 76% and 120% with RSD ≤15%, and 76% and 120% with RSD ≤11% for MSPD and QuEChERS extraction, respectively. The results suggested that the ME for 21 pesticides was in the range of 80%–120% after MSPD and 15% after QuEChERS extraction. QuEChERS extraction was simpler and faster than MSPD. This methodology was applied in the analysis of 27 C. indicum samples; phorate was most frequently detected (63.0% of the sample).


Introduction
Chrysanthemum indicum (Ye Ju Hua) is a medicinal herb with anti-inflammatory, analgesic, antioxidant, antibiotic, and other pharmacological effects. It is widely distributed and planted in China. It is very popular in Asian countries, such as China, Korea, and Japan, as traditional medicine, herbal tea, and functional foods [1].
With the implementation of standardized cultivation of Chinese herbal medicines, many herbs, including C. indicum, have established planting bases. Pest control is very important to improve the yield and quality of herbs. Good agricultural practices (GAP) attach great importance to the protection of the ecological environment in the production of Chinese medicinal materials and have strict usage norms and safety standards for the control of pesticide residues, such as selecting the rational pesticide varieties, application period, and mix pesticides. During the growth not exceed the standard. erefore, some residues of pesticides should remain in herbal products [2]. e residues of pesticides are closely related to the soil environment.
Pesticide residues in herbs are pesticide precursors, degradants, toxic metabolites, and impurities that remain in the medicinal parts after the use of pesticides in the growing environment, planting, processing, and storage [3]. According to the chemical structure, pesticides are divided into organophosphorus, organochlorine, pyrethroid, and carbamate. Organochlorine pesticides use benzene or cyclopentadiene as raw materials, such as hexachlorocyclohexane (BHC) and dichlorodiphenyltrichloroethane (DDT). ese pesticides are chemically stable and easily accumulate in living organisms. e long-term accumulation of DDT in the human body can cause immune system dysfunction and genetic and developmental toxicity; DDT can even be carcinogenic and teratogenic. Organophosphorus pesticides are currently widely used types of pesticides, including dichlorvos, dimethoate, and phorate. Most organophosphorus pesticides have a short residual time. Hence, chronic poisoning is less. At present, the acute poisoning caused by the consumption of food containing organophosphorus pesticides ranks first in the pesticide poisoning incidents in China [4].
Pesticide residue analysis is a technique used to analyze trace components in complex mixtures. It requires high sensitivity, high recovery, good specificity and reproducibility, and easy operation. e difficulty lies in the preprocessing technology, which requires extracting the components to be tested while removing the interference components as much as possible to ensure the accuracy of the detection results and avoid instrument pollution. e pretreatment methods reported so far include the mechanical oscillation, ultrasonic extraction, solid-phase extraction (SPE) [5,6], supercritical fluid extraction, QuEChERS (quick, easy, cheap, effective, rugged, and safe) extraction, and matrix solid-phase dispersion (MSPD) [7]. e MSPD was first reported in 1989 for the simultaneous fragmentation and extraction of the same class of components in solid and semisolid samples. MSPD includes both matrix dispersion and chromatographic separation processes [8]. e general operation is to mix and grind the sample with an SPE material, fill the obtained semi-dry state mixture in the column, and then elute with a solvent of different polarity to obtain the target compound. e method simplifies the steps in the traditional pretreatment method, such as sample homogenization, tissue cell lysis, and purification, thereby avoiding the loss. QuEChERS is a pretreatment method based on solid-liquid extraction and dispersive solid-phase extraction technology developed by Anastassiades et al. [9] and validated by Lehotay et al. [10]. Because QuEChERS is friendly, convenient, stable, and reliable, it is widely adopted by the scientific community, which can be applied to the detection of pesticide residues and the extraction of different analytes from different fields and matrices such as foods [11,12], biological fluids [13], and environmental samples [14]. e steps of the QuECh-ERS method can be roughly summarized as follows: pulverizing the sample, adding the extraction solvent, adding the salt to remove the water, adding the adsorbent to remove the impurities, and detecting the supernatant. e QuEChERS method can recover alkaline components that MSPD cannot recover and can extract pesticide components with a wide range of polarities. [15,16] Many approaches are used to detect pesticide residues, including GC, LC, GC-MS, LC-MS, and so forth [17][18][19][20]. Gas chromatography-mass spectrometry (GC-MS) not only has a high separation efficiency of gas chromatography but also the ability of mass spectrometry to accurately identify the compound structure [21][22][23]. It can accurately measure a small amount of various pesticide residues and corresponding metabolites in the sample at the same time. It is suitable for detecting pesticide residues in medicinal materials. e accuracy of the trace analysis with gas chromatography (GC) is often seriously affected by ME [24]. However, in previous studies, the interactions of pesticides with the matrix were observed in the detection of pesticide residues in C. indicum using GC-MS. In this study, 37 types of pesticides were analyzed by GC-MS, and the MSPD and QuEChERS pretreatment methods were compared in terms of purification and ME. We aimed to establish a simple, rapid, and efficient method to analyze pesticide residues in C. indicum.

Pretreatment Methods.
e pretreatment methods were based on the previously reported methods, and the cleanup sorbent and cleanup solvent were optimized. e main steps of the MSPD method included mixing the sample with the cleanup sorbent, filling the column, extracting, concentrating, and reconstituting (Supplementary Figure 1). e main steps of QuEChERS included two parts, namely, extraction and cleanup.

Gas Chromatography-Mass Spectrometry Analysis.
e analysis was performed using gas chromatographymass spectrometry (GC-MS). An Agilent 7890B GC system was equipped with Agilent 5977A MSD. e chromatographic separation was performed using the Agilent DB-1701 column with a length of 30 m × ID 0.25 mm × and film thickness of 0.25 μm. e flow of carrier gas helium was 1.3 mL/min. e oven temperature was as follows: 50°C (1 min) and 30°C min −1 to 160°C, 4°C min −1 to 200°C, 3°C min −1 to 230°C (2 min), 2°C min −1 to 250°C, 20°C min −1 to 270°C, and 5°C min −1 to 300 (5-min hold), with a total run time of 48.5 min. e injector temperature was 230°C. Injection mode was splitless mode. e MS was operated in the electron ionization mode with a transfer line temperature of 250°C and an ion source temperature of 230°C.

Method Validation.
e method validation was performed using the following parameters, namely, accuracy (expressed as recovery), precision (expressed as RSD), linearity (expressed as R 2 ), limit of detection (LOD), and limit of quantification (LOQ). e accuracy was expressed as recovery. ree different levels were analyzed (0.4 mg/kg, 2 mg/kg, and 10 mg/kg), with three replicates for each level. e linearity was studied by analyzing the mixed standard solution at five concentration levels.
e range of analyzed concentrations was 0.02-10.0 mg/L.

Matrix Effect.
Initially, two pretreatment methods were evaluated in terms of matrix effect (ME) by comparison between the areas of the standard in the extract and the standard in the solvent, shown by the following equation: ME (%) � (area of the standard in the matrix/area of the standard in the solution) × 100. When the ME value is close to 100%, there are no influences by the matrix. When the ME value is out of the range 80%-120%, it means that the matrix effect is significant. [25].

Qualitative and Quantitative Detection.
e qualitative analysis of pesticide residues referred to the method for the determination of pesticide residues (Chinese Pharmacopoeia 2015, fourth edition). During sample testing, if the retention time of the detected peak was consistent with the reference, the qualitative ions appeared in the mass spectrum after subtracting the background. Moreover, the relative abundance of the sample was consistent with the reference (relative abundance >50%, deviation allowed up to ±20%; relative abundance 20%-50%, deviation allowed ±25%; relative abundance 10%-20%, deviation allowed ±30%; and relative abundance <50%, deviation allowed ±50%). en, the presence of the pesticide in the sample was determined. e internal standard method was used for the quantitative analysis, and the internal standard was heptachlor exoepoxide.

Pesticide Selection and Grouping.
A total of 37 representative pesticides were selected for analysis, including pesticides and their metabolites used in planting bases and pesticides restricted or prohibited in China and other countries, which are not easily metabolized or highly toxic. Based on the relevant literature and the actual planting situation of the base, this study summarized pesticides commonly used in the control of C. indicum diseases and pesticides with more domestic dosage forms; 37 pesticides were used as the detection indicators. ese pesticides were divided into two groups (groups A and B) based on the following principle: (1) retention time: pesticides were grouped with close retention times and overlaps to avoid mutual interference between compounds and (2) physical and chemical properties: group A mainly included organochlorine pesticides, while group B mainly included organophosphorus pesticides (Table 1).

Chromatographic Analysis.
e ions for each compound were obtained by GC-MS analysis in full-scan ( Figure 1) and selective ion monitoring (SIM) modes (Table 1). e standard solutions at a concentration of 10 mg/L of pesticides were prepared in acetone. e sensitivity of some pesticide reference materials was related to chromatographic conditions. e examination of the injector temperature, transfer line temperature, and flow revealed that the injector temperature had a great influence on endosulfan sulfate, quizalofop ethyl, profenofos, and famphur.

Optimum Method of MSPD.
e MSPD processing method was as follows: 0.5 g C. indicum sample was grounded for 3 min in the agate mortar with 1 g of single sorbent or 0.5 g of mixed sorbents. It was filled in the column (10 mL tube, 15.8 mm × 88 mm) with anhydrous sodium sulfate at a 2 cm height of the extraction column, which was prerinsed with 4 mL of the extraction solvent. Before the liquid level reached the top of anhydrous sodium sulfate, 25 mL of extraction solvents (n-hexane : acetone � 4 : 6) was added to the extract. e elution solvent was concentrated to  dryness, followed by the addition of 1 mL of n-hexane for dissolution, and filtered for GC-MS analysis. e results of the purification effect of a single sorbent on C. indicum indicated that GCB had a better effect on pigment removal; both GCB and C 18 had strong adsorption on some pesticides, such as heptachlor epoxide and endrin ( Figure 2). When using the NH 2 + C 18 mixed sorbent, the recovery of pesticides in group B was mostly in the range of 75%-125%, but the recovery of 9 of 17 pesticides in group A was less than 75%; the recovery of 2 pesticides was higher than 125% (Figure 2(c)). When using the PSA + NH 2 mixed sorbent, the recovery of only three pesticides was less than 75% (Figure 2(a)). erefore, PSA + NH 2 mixed sorbent was chosen as the adsorbent filler.
Furthermore, the ratio and amount of PSA and NH 2 and the time for mixing of the cleanup sorbent with the sample powder also had a significant influence on the purification effect. Based on the recovery, a two-factor and four-level orthogonal experiment was designed to evaluate the optimal ratio of PSA and NH 2 , and the results were verified. e amount of the absorbent had a great influence on the recovery of organochlorine pesticides (group A) but had little effect on organophosphorus pesticides (group B). According to the orthogonal test results, an appropriate increase in the amount of cleanup sorbents increased the recovery; when the sorbent combination was 200 mg PSA +200 mg NH 2 , the best purification effect was achieved and verified. e recoveries of 37 pesticides ranged from 76.24% to 118.76% with RSD <10.0% ( Figure 3). As shown in Figure 4(b), the repeated mixing of the adsorbent and the sample by grinding improved the extraction and purification effect. If the grinding time was too long, some pesticides, such as 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22  quizalofop ethyl, were decomposed to reduce the recovery. Overall, it was most suitable to mix the sample with the adsorbent for 2 min. Pretreatment of pesticide residues in herbs is a process of extracting and enriching trace components. e ideal extraction solvent should be able to extract pesticides as characteristically as possible. Different types of samples have different requirements for extraction solvents; the choice of extraction solvent is a very important step in the pretreatment method. erefore, in this study, four kinds of solvent combinations were used for extraction, which were acetonitrile (A), n-hexane : ethyl-acetate (9 : 1) (B), n-hexane : acetone (9 : 1) (C), acetone : acetonitrile (3 : 7) (D), acetone : ethyl acetate (1 : 1) (E), and n-hexane : acetone (4 : 6) (F). e extraction solvents A, B, and C had a large interference with c-BHC (recovery >140%), while the extraction effect of B and C on dimethoate was very poor (undetectable). When using E as the extraction solvent, the recovery of most pesticides was less than 75%. e recovery of 16 pesticides, including heptachlor epoxide, was less than 75% using solvent combination D, while the extraction effect of F on dieldrin and endrin was poor. In summary, n-hexane : acetone (4 : 6) (F) was a suitable extraction solvent for C. indicum samples. e amount of the extraction solvent was also one of the factors affecting the recovery. Increasing the amount of the extraction solvent increased the recovery to a certain extent. However, excessive use not only caused unnecessary waste but also led to an increase in the number of transfers during the recovery of the solvent. In this method, 15 mL of the extraction solvent was sufficient to elute completely, and the pesticide recovery was in the range of 75%-125% (Figure 4(a)). Journal of Analytical Methods in Chemistry

Optimum Method of QuEChERS.
Despite the simplicity of the experiment, the effectiveness of the QuEChERS method depends on the nature of the target analyte, matrix composition, equipment, and analytical techniques available in the laboratory. erefore, when developing the QuEChERS protocol, several parameters that affect the extraction efficiency need to be considered and optimized. [14]. e initial steps of QuEChERS extraction were as follows: 1.0 g of the sample was accurately weighed and placed in a 50 mL centrifuge tube. en, 10 mL of the extraction solvent was precisely added and shaken at 1,000 rpm for 4 min. Furthermore, the extraction package was added, shaken at 1,000 rpm for 4 min, and centrifuged at 5,000 rpm for 5 min. e supernatant was transferred to a purification tube, shaken at 1,000 rpm for 4 min, and centrifuged at 5,000 rpm. e supernatant was filtered for GC-MS analysis. e cleanup sorbents, volume and type of solvent, and treatment time were optimized. Six kinds of cleanup sorbents were explored (Supplementary Table 2), of which No. 1, No. 5, and No. 6 had higher adsorption to profenofos ( Figure 5(a)). e interference of No. 4 and No. 6 with β-BHC was relatively large (Figure 5(a)). When the No. 2 cleanup sorbent was used, the recovery of pesticides ranged between 82.07% and 119.95%, indicating that it had a good purification effect.
Next, the extraction effect of acetonitrile (AcN) and a solution with different concentrations of HAc (0.1%, 0.2%, and 0.4%) was investigated. us, 0.1% HAc was added to AcN to increase the recovery of pesticides (76.92%-115.41%). e recovery of a few pesticides, such as metalaxyl and disulfoton, reduced when the acid concentration was too high (Figure 5(b)). Different volumes (8, 10, and 15 mL) of the extraction solvent had a certain influence on the extraction effect; 10 mL was found to be sufficient. e increase in extraction time had a great influence on the recovery of some organophosphorus pesticides. When the extraction time was 2 min, the recovery of disulfoton and methyl parathion was 66.69% and 138.47%, respectively ( Figure 5(c)). e purification time had minimal effect on the recovery rate ( Figure 5(d)). After adding the extraction solvent to the sample, the shaking time was 1 min. After adding the adsorbent, the same shaking was performed for 1 min.

Method Validation.
e optimal conditions for the application of MSPD and QuEChERS to detect 37 pesticides in C. indicum were as follows. (1) e optimal conditions of MSPD were that 0.5 g C. indicum sample was taken in the agate mortar, 0.2 g PSA and 0.2 g NH 2 were added, and the mixture was grounded for 2 min. e column was filled, followed by the addition of anhydrous sodium sulfate to an extraction column of about 2 cm height , and prerinsed with 4 mL of n-hexane : acetone (4 : 6). Before the liquid level reached the top of anhydrous sodium sulfate, 15 mL of nhexane : acetone (4 : 6) was added to the extract, the elution solvent was concentrated to dryness, and 1 mL of n-hexane was added for dissolution. e filtrate was used for GC-MS analysis. (2) e optimal conditions of QuEChERS were to accurately weigh 1.0 g sample in a 50 mL centrifuge tube. en, 10 mL of 1% HAc-AcN was added and shaken at 1,000 rpm for 1 min. e extraction bag (4 g MgSO 4 and 1 g NaCl) was added, shaken at 1,000 rpm for 1 min, and centrifuged at 5,000 rpm for 5 min. e supernatant was taken, transferred to a purification tube, shaken at 1,000 rpm for 1 min, and centrifuged at 5,000 rpm. e supernatant was taken and filtered for GC-MS analysis.
Different parameters, such as accuracy, precision, linearity, LOD, and LOQ, were determined (Supplementary  Tables 3 and 4). e recovery and precision of the method for all pesticides at three levels (0.4 mg/kg, 2 mg/kg, and 10 mg/kg) in three replicates were determined. e recoveries at the three levels using mixed standard solution ranged between 76% and 120% with RSD ≤15%, and 76% and 120% with RSD ≤11% for MSPD and QuEChERS extraction, respectively (Supplementary Tables 3 and 4).

Matrix Effect.
In the analysis of pesticide residues, the type and content of the matrix affect the recovery. During the gas chromatography injection process, the matrix reduces the decomposition of thermally unstable pesticides and reduces the injection of polar pesticides at the injection port. at is to say, the sample matrix increases the amount of pesticide to be analyzed that enters the column from the inlet. e calibration using pure solvent standard solutions is that ME can cause deviations in pesticide residue analysis results and recovery calculations. Several methods are used to compensate for ME, such as matrix purification. e most effective method is calibration with matrix-matched standard solutions or calibration with analytical protection agents [26,27].
Blank matrices of two extraction methods were prepared to compare the effects of the two pretreatment methods on ME, and a mixed standard solution was added to the matrix. e peak area of the standard in the solvent and the peak area of the matrix mixed standard were measured separately to calculate ME. After MSPD treatment, the ME value of 21 pesticides exceeded the range of 80%-120%, while after QuEChERS treatment, the ME value of only 15 pesticides exceeded 80%-120% ( Figure 6).
In the C. indicum matrix, most of the pesticides exhibited matrix enhancement effects, especially organochlorine pesticides. MSPD and QuEChERS methods compensated the matrix enhancement effects to a certain extent using matrix-matched standard solutions, which was more obvious in the QuEChERS method. key parameters can show a relevant impact on the extraction efficiency [28]. e extraction package of the QuEChERS method was screened, revealing that the extraction package of PSA + GCB (ANPEL Laboratory Technologies) had a good purification effect. In the matrix, the most suitable cleanup sorbent for MSPD was PSA + NH 2 . is combination also had a better purification effect when used in the QuEChERS method, but the recovery of heptachlor epoxide and profenofos was less than 75%. Rejczak and Tuzimski use liquid-liquid partitioning with acetonitrile followed by dispersive solid-phase extraction cleanup using primary and secondary amine along with zirconia-coated silica particles for extraction and purification, so that the recoveries of the vast majority of the analytes were from 70 to 100% with relative standard deviations less than 20% being observed [29].
e QuEChERS method eliminated the process of drying and redissolving the solution, which not only simplified the operation but also reduced the error caused by the operation.

Application of C. indicum Sample.
e qualitative and quantitative detection of pesticide residues in C. indicum was carried out using QuEChERS as a pretreatment method. A total of 20 pesticides were detected in 27 C. indicum samples, six of which were organochlorine pesticides. e detection rates of phorate and profenofos exceeded 50% (Table 2). e phorate and profenofos in the two C. indicum samples exceeded 2 mg/kg, and the remaining samples met the limit requirements of the Chinese Pharmacopoeia. At present, no requirements exist for pesticide residues in flowers and fruits of medicinal plants, such as C. morifolium Ramat, in the National Food Safety Standard-Maximum Residue Limits for Pesticides in Food (GB 2763-2016). However, it sets maximum residue limits for 50 pesticides in tea, including phorate and BHC. e phorate is a highly toxic dithiophosphate insecticide and acaricide. It inhibits cholinesterase activity and causes neurophysiological dysfunction. Profenofos is a spectral insecticide often used to control pests of cotton, vegetables, and other crops, especially for the control of resistant cotton bollworms.

Conclusions
In this study, two sample pretreatment methods (MSPD and QuEChERS) were evaluated for detecting pesticide residues in C. indicum, reducing the influence of matrix to improve the accuracy of detection. e results of the comparison of the two sample processing methods (MSPD and QuECh-ERS) show that QuEChERS is better than MSPD, but the price is also more expensive than MSPD. Compared with the extraction solvents commonly used in laboratories, such as n-hexane, acetone, and acetonitrile, acetonitrile is less toxic. QuEChERS method selects acetonitrile as the extraction solvent. Adding acetic acid according to the nature of the sample is beneficial to the recovery of pesticides. Anhydrous magnesium sulfate has a strong dehydration capacity, which can meet the rapid dehydration needs of QuEChERS. PSA can absorb impurities such as sugars and fatty acids in the sample to achieve the purpose of purification. Both methods have their own advantages. If we want to apply the method to actual sample detection, we need a simpler and faster method. A GC-MS method for detecting pesticide residues in C. indicum was developed. e accuracy, precision, linearity, and LOQ results showed that the proposed method was feasible and applicable to detect 37 pesticide residues in C. indicum samples. e calibration using matrix-matched standard solutions effectively compensated for the ME of pesticide residue analysis in C. indicum. As an important food and dual-use product, C. indicum is widely used by consumers in their daily lives, especially in China. erefore, it is necessary to establish a faster and more effective method for analyzing pesticide residues in C. indicum, a matrix to ensure a healthy product to its consumers.

Data Availability
All the data used during the study appear in the submitted article.

Conflicts of Interest
e authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Authors' Contributions
LU Xiao-ying completed data analysis and drafted the manuscript. OUYANG Yan-qin designed and executed the research. ZENG WEI-ya, LIN Cui-qing, XIAO Lu-hua, and LUO Gui-hua participated in the experiment. ZHAN Ruoting and YAN Ping conceived the study, participated in its design and coordination, and helped to draft the manuscript. All the authors read and approved the final manuscript.