Article pubs.acs.org/JAFC
Sensitive Detection of Organophosphorus Pesticides in Medicinal Plants Using Ultrasound-Assisted Dispersive Liquid−Liquid Microextraction Combined with Sweeping Micellar Electrokinetic Chromatography Jin-Chao Wei, Ji Hu, Ji-Liang Cao, Jian-Bo Wan, Cheng-Wei He, Yuan-Jia Hu, Hao Hu, and Peng Li* State Key Laboratory of Quality Research in Chinese Medicine, Institute of Chinese Medical Sciences, University of Macau, Macau 999078, People’s Republic of China S Supporting Information *
ABSTRACT: A simple, rapid, and sensitive method using ultrasound-assisted dispersive liquid−liquid microextraction (UADLLME) combined with sweeping micellar electrokinetic chromatography (sweeping-MEKC) has been developed for the determination of nine organophosphorus pesticides (chlorfenvinphos, parathion, quinalphos, fenitrothion, azinphos-ethyl, parathion-methyl, fensulfothion, methidathion, and paraoxon). The important parameters that affect the UA-DLLME and sweeping efficiency were investigated. Under the optimized conditions, the proposed method provided 779.0−6203.5-fold enrichment of the nine pesticides compared to the normal MEKC method. The limits of detection ranged from 0.002 to 0.008 mg kg−1. The relative standard deviations of the peak area ranged from 1.2 to 6.5%, indicating the good repeatability of the method. Finally, the developed UA-DLLME−sweeping-MEKC method has been successfully applied to the analysis of the investigated pesticides in several medicinal plants, including Lycium chinense, Dioscorea opposite, Codonopsis pilosula, and Panax ginseng, indicating that this method is suitable for the determination of trace pesticide residues in real samples with complex matrices. KEYWORDS: capillary electrophoresis, sweeping, ultrasound-assisted dispersive liquid−liquid microextraction, organophosphorus
■
INTRODUCTION Medicinal plants (MPs) have been widely used as therapeutic agents and nutritional supplements for thousands of years in many countries as a result of their remarkable beneficial effects and negligible adverse effects.1−3 It has been estimated that as many as 80% of the population in developing countries relies on plant-based medicines for their primary healthcare, and nearly 51% of all drug preparations in industrialized countries are directly or indirectly derived from plants.4 However, as with any product from wild or cultivated sources, MPs can be contaminated by many potentially hazardous substances, such as microbial contaminants, heavy metals, and pesticide residues.5 In particular, nowadays, various kinds of pesticides are used on a large scale to kill or control unwanted insects, fungi, or other pests in the process of herbal production, which has become one of the concerns when assessing the safety of herbal products. Among various pesticides, organophosphorus pesticides (OPPs) are the most extensively used insecticides or herbicides as a result of their more intensive pest control effects and faster degradation in the environment compared to the organochlorine pesticides (OCPs), which were widely used in the mid-20th century. Acute and chronic exposure to OPPs could cause several neurotoxic disorders in humans, including the cholinergic syndrome, the intermediate syndrome, organophosphate-induced delayed polyneuropathy, and chronic organophosphate-induced neuropsychiatric.6 Therefore, the development of sensitive, selective, and inexpensive methods for monitoring of the toxic organophosphorus compounds in © XXXX American Chemical Society
plant-based medicines is currently a research area of great interest. The determination of pesticide residues has been accomplished mainly by chromatographic methods, including gas chromatography (GC),7 liquid chromatography (LC),8 or supercritical fluid chromatography (SFC),9 coupled with different detectors, such as electron capture detector (ECD), nitrogen phosphorus detector (NPD), flame photometric detector (FPD), photodiode array (PDA) detector, or mass spectrometry (MS) detector.10−12 Nowadays, as a result of the advantages of high resolution and separation efficiency, short analysis time, and low sample and reagent consumption compared to the LC method and the extraordinary power in dealing with the nonvolatile, thermolabile, or polar compounds, which are not suitable for the direct GC analysis, capillary electrophoresis (CE) has emerged as a very good separation alternative in pesticide analysis.13−16 Especially, with the rapid development of various online and offline sample preconcentration techniques,17−19 the major drawback of CE, i.e., the relatively low sensitivity as a consequence of the limited amount of sample volume injected, has been well overcome, which makes CE a promising technique for the analysis of pesticide residues. Received: November 11, 2015 Revised: January 6, 2016 Accepted: January 12, 2016
A
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry
Figure 1. Structures of the studied pesticides.
knowledge, it is the first time for the application of the UADLLME−sweeping-MEKC method for the analysis of OPPs in real samples.
As an important branch of CE, micellar electrokinetic chromatography (MEKC) has often been used in the recent decade for the analysis of polar or nonpolar OPPs, owing to its powerful separation ability for both neutral and ionic compounds.20−22 As online pre-concentration methods, stacking and sweeping can be individually or simultaneously combined with MEKC for the determination of trace pesticide residues. In sweeping mode, the analytes can be enriched into narrow bands within the capillary by a pseudo-stationary phase, and the efficiency mainly depends upon the interaction of the analytes with the pseudo-stationary phase.23,24 In addition to sample enrichment, the offline pre-concentration methods also play an important role in removing the matrix interferences from the analytes. In comparison to the conventional strategies, such as liquid−liquid extraction (LLE), solid-phase extraction (SPE), solid-phase microextraction (SPME), liquid-phase microextraction (LPME), and cloud point extraction (CPE), dispersive liquid−liquid microextraction (DLLME) has attracted more attention in the trace pesticide analysis since introduced in 2006 as a result of the advantages of the simple operation process, short extraction time, low solvent consumption, and high enrichment factor.25−27 Especially when assisted by ultrasonic activation, the extraction efficiency of DLLME could be significantly increased because the cavitation effect of ultrasound is conducive to the formation of a fine cloudy solution with less dispersive solvent and to the acceleration of the mass transfer of analytes from the aqueous phase to the extractant droplets.28−30 In this work, a procedure using ultrasound-assisted dispersive liquid−liquid microextraction (UA-DLLME) and sweeping as offline and online pre-concentration techniques coupled with MEKC has been proposed for the simultaneous determination of nine OPPs (chlorfenvinphos, parathion, quinalphos, fenitrothion, azinphos-ethyl, parathion-methyl, fensulfothion, methidathion, and paraoxon). Some important parameters that influence UA-DLLME and sweeping efficiency were optimized. The practical applicability of the proposed method was investigated in several MPs, such as Lycium chinense Miller, Dioscorea opposita Thunb., Codonopsis pilosula (Franch.) Nannf., and Panax ginseng C. A. Meyer. To the best of our
■
MATERIALS AND METHODS
Chemicals and Reagents. Certified reference standards, including chlorfenvinphos (1), parathion (2), quinalphos (3), fenitrothion (4), azinphos-ethyl (5), parathion-methyl (6), fensulfothion (7), methidathion (8), and paraoxon (9) (structures shown in Figure 1), were all purchased from Dr. Ehrensdorfer (Augsburg, Germany). Individual stock solutions of the pesticides at a concentration of 100.0 mg L−1 were prepared in acetonitrile and stored at 4 °C. The mixed standard solution containing 10.0 mg L−1 of each pesticide was prepared in acetonitrile and stored at 4 °C. A series of standard working solutions was prepared daily by an appropriate dilution of the stock solutions after dryness under a stream of nitrogen. Bulk sorbents, including primary secondary amine (PSA, 50 μm) and graphitized carbon black (GCB, 45 μm) were obtained from Sigma-Aldrich (St. Louis, MO). Sodium dihydrogen phosphate (NaH2PO4), sodium dodecyl sulfate (SDS), hydrochloric acid (HCl), and sodium chloride (NaCl) were analytical reagents from Sigma-Aldrich (St. Louis, MO). Dichloromethane (CH2Cl2), chloroform (CHCl3), and chlorobenzene (C6H5Cl) were high-performance liquid chromatography (HPLC)-grade solvents from Sigma-Aldrich (St. Louis, MO). Methanol, ethanol, acetone, and acetonitrile were analytical-grade solvents from Damao Chemical Reagent (Tianjin, China). Deionized water was purified with a Milli-Q purification system (Millipore, Bedford, MA). Apparatus and Software. All experiments were performed on a Beckman P/ACE MDQ capillary electrophoresis system (Beckman Coulter, Fullerton, CA), equipped with a PDA. An uncoated fusedsilica capillary (Ruifeng, Handan, China) of 75 μm inner diameter and 57 cm total length (50 cm effective length) was used for all of the experiments. The temperature of the capillary was set at 25 °C. Data were analyzed by Beckman P/ACE MDQ 32 Karat software. Ultrasonic equipment (SCIENTZ SB-300 DTY, Ningbo Scientz Biotechnology Co., Ltd., China) was used for sample treatment. Sample Preparation. Dried MP samples, including L. chinense, D. opposite, C. pilosula, and P. ginseng, were purchased from a medicine store in Guangzhou, China, which have been processed as decoction pieces (the herbal materials processed according to the specifications of China Pharmacopeia). The sample pieces were cut into small slices about 2 mm in diameter and then mixed thoroughly. Aliquots of 0.5 g B
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry
Figure 2. Effects of (A) methanol content and (B) SDS concentration in running buffer on the migration time. Pesticide assignment: (1) chlorfenvinphos, (2) parathion, (3) quinalphos, (4) fenitrothion, (5) azinphos-ethyl, (6) parathion-methyl, (7) fensulfothion, (8) methidathion, and (9) paraoxon. Data are presented as the mean ± standard deviation (SD) from three independent experiments. of each sample were randomly sampled and weighed into a 15 mL glass centrifuge tube and soaked with 5 mL of acetone, and then the tube was closed and immersed in an ultrasonic water bath for extraction (10 min and 60 kHz) at 25 °C and centrifuged for 5 min at 800g at 25 °C. A total of 2.0 mL of the upper layer of the extract was transferred into a 15 mL glass centrifuged tube containing 200 mg of PSA and 30 mg of GCB. The tube was shaken for 1 min and centrifuged at 3000g for 5 min. Then, 1.0 mL of acetone supernatants was mixed with 350 μL of chloroform, and the acetone−chloroform mixture was quickly injected into 5 mL of deionized water containing 6% NaCl to form cloudy solution. The mixture was shaking for 1.0 min using an ultrasonic water bath. The fine droplets were coalesced and deposited at the bottom of the tube after centrifugation at 3000g for 5 min. The sedimented phase was removed completely and evaporated to dryness under a stream of nitrogen. The residue was reconstituted in 100 μL of 30 mM sodium dihydrogen phosphate solution for CE analysis. Electrophoretic Procedure. A new capillary was conditioned by rinsing at 20 psi for 10 min with methanol, 20 psi for 10 min with deionized water, 20 psi for 10 min with 1 M HCl, 20 psi for 5 min with deionized water, 20 psi for 20 min with 1 M NaOH, and 20 psi for 5 min with deionized water. Between consecutive analyses, the capillary was rinsed with water (3 min), 0.1 M HCl (5 min), deionized water (3 min), and NaH2PO4 buffer (80 mM, pH 3.0) containing 40% methanol for 5 min at 30 psi. Then, the large volume sample solution was hydrodynamically injected at 2.0 psi for 90 s. After large volume sample injection, the inlet and outlet reservoirs were replaced with running buffer (80 mM NaH2PO4 containing 200 mM SDS and 5% methanol at pH 3.0). Finally, electrophoresis was performed at a constant reverse voltage (−15 kV) with diode array detection at 200 nm. Because the electrophoretic velocity of micelles is greater than the electroosmotic flow (EOF) velocity at low pH, micelles from the cathodic vial enter the capillary and sweep the analytes toward the detector.
MEKC procedure were investigated in this experiment, including sample solvent, buffer pH and concentration, organic solvent content, SDS concentration, and injection volume. An aqueous mixture solution of nine OPP standards (5 mg L−1) was used to study the analytical parameters under different conditions. The proper sample matrix normally required lower conductivity than that of the background electrolyte solution.31 The analytes with SDS micelles decelerated at the boundary between the sample zone and background electrolyte solution (BGS) and formed narrow bands when a high voltage was applied. Six different concentrations (0, 10, 20, 30, 40, and 50 mM) of phosphate buffer without SDS or organic solvent as the sample solvent were studied. At first, the peak shape and the separation efficiency of the targeted analytes were improved when the phosphate concentration was increasing from 0 to 30 mM. However, with the further increase of the sample matrix conductivity, the separation efficiency and sensitivity showed no obvious improvement. Consequently, 30 mM NaH2PO4 was selected as the ideal sample solvent. The EOF velocity of the buffer solution was easily influenced by the pH values. To suppress EOF, the buffers were studied in a low pH range of 1.5−4.0. The increasing pH value brought larger EOF, which led to a longer migration time. In consideration of separation efficiency and analytical time, pH 3.0 was chosen as the optimal pH condition for the BGS. Then, different concentrations of sodium dihydrogen phosphate (10, 20, 40, 60, 80, and 100 mM) were tested to evaluate the effect on the separation behavior of the target analytes. The result demonstrated that the peak shapes were improved when the concentration was higher than 40 mM, and no significant difference on separation efficiency could be obtained when the concentration was more than 80 mM. However, the migration time of the pesticides decreased with the increase of the phosphate concentration. Considering the migration time and the resolution, 80 mM was chosen as the optimum concentration for the running buffer. The addition of organic solvent in the BGS can influence the separation efficiency and migration time of the analytes because organic modifiers can change the polarity and viscosity of the buffer solution, the capacity of the micelle, and the affinity between micelles and analytes.32 In this study, the effect of the methanol content within the range of 0−50% was investigated.
■
RESULTS AND DISCUSSION Optimization of Sweeping-MEKC Conditions. In the sweeping procedure, electrophoresis was performed at a reverse voltage (−15 kV) after a large hydrodynamic injection of the sample solution (2.0 psi for 90 s). The electrophoretic velocity of micelles is greater than the electroosmotic flow velocity when the buffers are at low pH. The analytes in sample solution would be concentrated and swept from the cathode toward the detector with the help of micelles. For the purpose of increasing the online pre-concentration efficiency and obtaining the best separation performance, several parameters of the sweepingC
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry
Figure 3. Effects of (A) dispersive solvent type and (B) extraction solvent volume on the recoveries of the investigated pesticides in UA-DLLME: (A) dispersive solvent, 1.0 mL; extraction solvent, 200 μL of chloroform; and aqueous solution, 5 mL of 4% NaCl solution and (B) dispersive solvent, 1.0 mL of acetone; extraction solvent, chloroform; and aqueous solution, 5 mL of 4% NaCl solution. Pesticide assignment: (1) chlorfenvinphos, (2) parathion, (3) quinalphos, (4) fenitrothion, (5) azinphos-ethyl, (6) parathion-methyl, (7) fensulfothion, (8) methidathion, and (9) paraoxon. Data are presented as the mean ± SD from three independent experiments.
As shown in Figure 2A, with the addition of 5% methanol in running buffer, the migration times of the nine pesticides showed obvious reduction. When the ratios of methanol increased from 5 to 50%, however, an increasing trend of migration times was noticed. Within the investigated range of the methanol content, the separation efficiency showed no obvious difference. Therefore, 5% methanol was added in the separation buffer for further study. For the sweeping procedure, SDS was used for picking and accumulating analytes as the charged pseudo-stationary phase during application of a voltage. To determine the SDS effect on the sweeping efficiency, the separation was optimized using 80 mM NaH2PO4 containing different concentrations (50, 100, 200, 300, and 400 mM) of SDS. The result indicated that the migration times of all nine pesticides decreased along with the increased SDS concentration; particularly, a sharp decrease could be discovered when the concentration of SDS increased from 50 to 200 mM (as indicated in Figure 2B). Moreover, the peak areas of the analytes decreased with the increase of the SDS concentration. When the concentrations of SDS were higher than 300 mM, the peaks became deteriorated and the sensitivity was decreased. Therefore, 200 mM was chosen as the optimum concentration for SDS in consideration of both the migration time and sensitivity. In general, the period of sample injection time can affect the amount of sample introduced into the capillary, which then affects the peak sensitivity of the analytes. Various sample injection time periods (45, 90, 120, and 150 s) at 2.0 psi were investigated. With the increase of the sample injection time, the peak areas of all analytes increased. However, overloading may cause peak broadening without the improvement of sensitivity for analytes. In this study, the large volume of sample injection resulted in peak distortion when the injected time was more than 90 s. Therefore, an injection of sample solution at 2.0 psi for 90 s was selected. Optimization of the UA-DLLME Procedure. The extraction efficiency of the UA-DLLME procedure depends upon some different parameters. To obtain the optimal conditions, the influences of types of dispersive solvent and extraction solvent, volumes of dispersive solvent and extraction solvent, salt addition, and ultrasonic extraction time were investigated. The extraction recovery (ER, %) was used to evaluate the extraction efficiency according to the following equation:
ER (%) =
CrecVrec × 100 C0Vaq
(1)
where Crec and C0 are the mixed standard concentration in the final reconstituted solution and the initial mixed standard concentration in the acetone sample, respectively, while Vrec and Vaq are the volume of the final reconstituted solution and the volume of the acetone sample, respectively. For the UA-DLLME method, the importance for the dispersive solvent is its miscibility with both aqueous solution and extraction solvent, and then organic emulsions would be dispersed for interfacial analyte exchange. In this study, acetone, acetonitrile, methanol, and ethanol were selected as potential solvents, and their effects on the performance of UA-DLLME were evaluated. A series of sample solutions was tested using 1.0 mL of each dispersive solvent, while chloroform was kept at a constant volume of 200 μL as the extraction solvent. As indicated in Figure 3A and Table S1 of the Supporting Information, the extraction recoveries showed significant differences when using various dispersive solvents. Interestingly, the use of acetone could obtain the best extraction recoveries for parathion, quinalphos, fenitrothion, azinphos-ethyl, and parathion-methyl, although the recoveries for fensulfothion, methidathion, and paraoxon were lower than those using acetonitrile. Consequently, acetone was chosen as the optimal dispersive solvent for further study. The extraction solvent plays an important role for the extraction of selected pesticides. It has to be immiscible with aqueous solution to allow for emulsification, and its density needs to be higher than that of water to form sedimentation. For this purpose, CH2 Cl2 , CHCl 3, and C 6H 5Cl were investigated for their extraction capability of targeted analytes. CHCl3 gave the best extraction efficiency for all of the analytes with acetone as the dispersive solvent in this study. Therefore, CHCl3 was chosen as the extraction solvent for subsequent experiments. The amount of dispersive solvent has a direct effect on the solubility of the analytes because of its chemical property in the ternary solvent system. To minimize the volume of acetone while allowing for a sufficient amount for effective emulsification, the volumes of acetone were investigated. Different volumes (0.8, 1.0, 1.2, 1.5, and 1.8 mL) of acetone mixed with 200 μL of CHCl3 were added to 5 mL of deionized water with 4% sodium chloride. The results showed that the extraction D
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry
Figure 4. Effects of amounts of (A) PSA and (B) GCB on the recoveries of the investigated pesticides in the cleanup procedure. Pesticide assignment: (1) chlorfenvinphos, (2) parathion, (3) quinalphos, (4) fenitrothion, (5) azinphos-ethyl, (6) parathion-methyl, (7) fensulfothion, (8) methidathion, and (9) paraoxon. Data are presented as the mean ± SD from three independent experiments.
Table 1. Analytical Performance of the UA-DLLME−Sweeping-MEKC Method L. chinense
chlorfenvinphos parathion quinalphos fenitrothion azinphos-ethyl parathion-methyl fensulfothion methidathion paraoxon a
D. opposite a
b
range (mg kg−1)
linearity (r2)
LOD (mg kg−1)
% RSD (n = 5)
% RSD (n = 5 × 3)
range (mg kg−1)
linearity (r2)
LOD (mg kg−1)
% RSDa (n = 5)
% RSDb (n = 5 × 3)
0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5
0.9969 0.9974 0.9775 0.9971 0.9970 0.9414 0.9949 0.9863 0.9951
0.002 0.005 0.003 0.003 0.005 0.005 0.002 0.007 0.004
1.5 3.3 1.2 5.5 6.5 4.0 3.9 4.1 3.8
1.8 6.1 1.3 4.9 5.5 6.4 5.5 5.2 4.5
0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5 0.025−2.5
0.9989 0.9983 0.9974 0.9984 0.9969 0.9987 0.9981 0.9915 0.9980
0.003 0.007 0.003 0.006 0.006 0.007 0.003 0.008 0.003
1.3 2.7 1.5 3.3 1.7 5.5 3.9 6.3 3.4
5.3 2.3 3.7 4.6 1.6 5.5 4.8 6.2 5.1
Intraday RSD value (n = 5). bInterday RSD value (n = 5 × 3).
efficiencies for most analytes at 1.0 mL of acetone yielded optimal values of 75−95%. As a result, 1.0 mL of acetone was selected for the experiment. To evaluate the effect of the extraction solvent volume, different volumes of CHCl3 (150−400 μL in 50 μL intervals) were studied, while acetone was kept at a constant volume of 1.0 mL. As shown in Figure 3B and Table S2 of the Supporting Information, the extraction recoveries for most analytes increased by increasing the volume of chloroform from 150 to 350 μL. When the volume further increased to 400 μL, the extraction efficiency of some pesticides slightly decreased. Consequently, 350 μL was selected as the optimum volume for CHCl3. Generally, the addition of salt results in the increase of the ionic strength of the aqueous phase and then the decrease of the solubility of the analytes in it. Finally, the extraction efficiency of the analytes will increase as a result of the partitioning enhancement of the analytes into the organic dropets. According to the literature, sodium chloride was chosen as the most common salt for exploring the effect of the ionic strength. In this study, the effect of the ionic strength was evaluated by adding different concentrations of sodium chloride from 0 to 10% (w/v). The results showed that the extraction efficiencies for most analytes increased with the increase of the NaCl concentration up to 6% and slightly decreased at a higher content. On the basis of the results, 6% NaCl was used in further experiments. Although the ultrasound-assisted process can increase the extraction efficiency of DLLME, the extension of ultrasonic
time might decrease the volume of the sediment phase as a result of the volatile loss of analytes and extraction solvent.28 Therefore, to obtain the best extraction efficiency and ensure enough sediment phases for injection, a series of ultrasonic time (0.5, 1.0, and 2.0 min) was investigated in this work. Finally, 1.0 min (60 kHz and 25 °C) was decided as the optimum ultrasonic time, which produced the best extraction performance for each analyte. Optimization of the Sample Cleanup Procedure. Because some matrix-related compounds may be co-extracted during the ultrasonic extraction process, a selective cleanup procedure is essential to minimize the effect of these possible interferences, especially for the samples with complex matrices. In the present work, PSA and GCB sorbents have been used to cleanup the complex matrices in several MPs and to concentrate the target pesticides. Generally, PSA could remove various sugars, pigments, and polar organic acids as an anion exchanger, while GCB was used for adsorption of nonpolar and medium-polar compounds. First, the amounts of PSA have been compared to evaluate the effects on the extraction recoveries and purification efficiency. For the purification of most matrices, 200 mg of PSA was acceptable. As shown in Figure 4A and Table S3 of the Supporting Information, 200 mg of PSA also exhibited higher extraction efficiencies for most of the investigated pesticides. Second, to prevent the significant losses of low polar and planar pesticides, the effects of the amount of GCB on the purification and recoveries were investigated. It revealed that the higher amount of GCB provided more capacity to eliminate E
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry Table 2. EFs Obtained by Online and Offline Pre-concentration Methods pesticide
EF of sweeping-MEKC
EF of UA-DLLME−sweeping-MEKC
chlorfenvinphos parathion quinalphos fenitrothion azinphos-ethyl parathion-methyl fensulfothion methidathion paraoxon
219.3 708.5 750.1 448.4 873.7 226.7 99.0 206.3 92.8
2360.6 6203.5 5607.0 3698.9 4499.4 1683.4 873.7 1398.6 779.0
Figure 5. Electropherograms of the standard mixture obtained (A) without pre-concentration (sample at 20.0 μg mL−1 and injection at 0.5 psi for 5 s) and (B) with offline and online pre-concentration (sample at 2.0 μg mL−1 and injection at 2.0 psi for 90 s after the UA-DLLME procedure). Pesticide assignment: (1) chlorfenvinphos, (2) parathion, (3) quinalphos, (4) fenitrothion, (5) azinphos-ethyl, (6) parathion-methyl, (7) fensulfothion, (8) methidathion, and (9) paraoxon.
capability as well as linearity with correlation coefficients (r2) higher than 0.99 for all analytical curves, except for the compounds quinalphos, parathion-methyl, and methidathion, whose values were 0.9775, 0.9414, and 0.9863 for the matrixmatched calibration in L. chinense, respectively. LOD of the proposed method was calculated at the concentration that gave a signal-to-noise (S/N) ratio of 3 and varied between 0.002 and 0.008 mg kg−1. All of them are lower than the maximum residue limits (MRLs) regulated by the United States Food and Drug Administration (FDA) and the European Union (EU). The precisions were carried out by spiking the mixed standard at 0.5 mg kg−1 and expressed as the relative standard deviation (RSD) of the peak area on the same day (n = 5) and on 3 consecutive days. The results in Table 1 showed acceptable precisions, with RSD lower than 6.5% for intraday precision and 6.4% for interday precision. To evaluate the matrix effect, slope ratios of calibration in the selected matrix and in acetone were compared. The slope ratios of 1 indicates that the matrix does not significantly suppress or enhance the signal, otherwise denoting suppression (1). According to the results, most pesticides exhibited matrix enhancement effects more or less in all matrixes, while for chlorfenvinphos, parathion, quinalphos, and fenitrothion, strong suppression appeared in C. pilosula and P.
matrix effects because the cleanup step with GCB is designed to retain interfering substances and allows for the analytes of interest into the acetone phase. However, as indicated in Figure 4B and Table S4 of the Supporting Information, with the increase of the amount of GCB, lower recoveries for some planar pesticides (such as quinalphos and azinphos-ethyl) were also observed. As a result, when 30 mg of GCB was applied in the cleanup step, the acceptable recoveries from 84.7 to 98.9% for most analytes could be obtained. According to the above results, the combination of 200 mg of PSA and 30 mg of GCB was considered to be the most favorable sorbent for the cleanup procedure in this study. The optimized conditions for the procedures of UA-DLLME and sample cleanup have been summarized in Table S5 of the Supporting Information. Analytical Characteristics and Method Validation. To evaluate the overall analytical performance of the UADLLME−sweeping MEKC method, linear range, matrix effects, limit of detection (LOD), accuracy, and precision were studied. The linearity and matrix effects were studied in calibration standards prepared in neat solvent (acetone) and in four blank sample matrix extracts (L. chinense, D. opposite, C. pilosula, and P. ginseng). As shown in Table 1, linearities for all nine analytes in L. chinense and D. opposite were observed in the range of 0.025−2.5 mg kg−1. The pesticides exhibited a wide calibration F
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry Table 3. Recoveries Obtained from the Determination of OPPs in Spiked Samples L. chinense pesticide chlorfenvinphos
parathion
quinalphos
fenitrothion
azinphos-ethyl
parathion-methyl
fensulfothion
methidathion
paraoxon
a
spiked (mg kg−1) 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5
measured (mg kg−1)
D. opposite recovery (%)
a
ND 0.0803 0.4923 ND 0.0991 0.5586 0.1381 0.2525 0.5685 ND 0.0992 0.5208 ND 0.0837 0.6386 ND 0.0879 0.5884 ND 0.0851 0.4540 ND 0.0841 0.5540 ND 0.0969 0.5157
80.33 98.46 99.08 111.73 114.42 86.07 99.22 104.16 83.72 127.71 87.85 117.68 85.05 90.81 84.07 110.79 96.86 103.14
spiked (mg kg−1)
measured (mg kg−1)
0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5 0 0.1 0.5
ND 0.0983 0.5214 ND 0.0899 0.4325 0.0578 0.1715 0.4553 ND 0.0951 0.5100 ND 0.0979 0.5104 ND 0.0957 0.5214 ND 0.1024 0.4821 ND 0.0835 0.6099 ND 0.0900 0.4149
recovery (%) 98.34 104.28 89.92 86.49 113.68 79.49 95.11 102.00 97.93 102.07 95.73 104.27 102.36 96.42 83.50 121.97 89.96 82.99
ND = not detected.
MPs, i.e., L. chinense and D. opposite, respectively. The corresponding electropherograms can be seen in Figures S1 and S2 of the Supporting Information. These concentrations were higher than the MRLs (0.05 mg kg−1) set by the legislation of the U.S.A. and Europe but lower than the MRLs (0.5 mg kg−1) for this pesticide in fruit set by the Ministry of Agriculture of People’s Republic of China (GB 2763-014). The MRL for quinalphos in herbal drugs has not been regulated in China. The proposed method showed satisfactory recoveries for all of the analytes, except for azinphos-ethyl and methidathion, with 127.71 and 121.97% at 0.5 mg kg−1 spiked level, which indicated that this method was suitable for the trace analysis of these pesticides in complex real samples, such as MPs. In this work, a sensitive UA-DLLME−sweeping-MEKC method for the determination of pesticide residues has been developed. In comparison to the normal MEKC method, up to 6203.5-fold improvement in concentration sensitivity was achieved using this method as a result of the combination with offline and online pre-concentration techniques. Featuring great enrichment factor, satisfactory recovery, good repeatability, minimal sample size, and solvent consumption, the proposed method offers a sensitive and reliable method for routine pesticide detection in real samples with complex matrices.
ginseng. Therefore, the matrix-matched calibration curves were required for accurate quantification. Next, the sensitivity enhancement effect of the UA-DLLME− sweeping-MEKC method has been studied. The enrichment factor (EF) was calculated as a ratio of the peak area obtained using the developed method to that obtained by the normal MEKC method. As shown in Table 2, the signals of the nine pesticides produced by the sweeping-MEKC method could increase 92.8−873.7 times in comparison to those by the normal MEKC mode. After further combination with the offline pre-concentration procedure, the proposed method could provide a 779.0−6203.5-fold enrichment compared to the non-pre-concentration method. The above results have revealed that the offline and online pre-concentration procedures, i.e., UA-DLLME and sweeping, could greatly improve the sensitivity of the normal MEKC. A comparison of the main performance characteristics obtained in the proposed method with other analytical approaches was summarized in Table S6 of the Supporting Information. The electropherograms of the nine investigated pesticides obtained by MEKC without and with pre-concentration are shown in Figure 5. Application to Real Samples. To evaluate the applicability of the proposed method for the analysis of real samples with complex matrices, the developed UA-DLLME−sweeping MEKC method was applied to determine the nine organophosphorus pesticides in four MPs, including L. chinense, D. opposite, C. pilosula, and P. ginseng. Two known concentrations (0.1 and 0.5 mg kg−1) of the pesticide mixture were spiked in the sample. As shown in Table 3, 0.1381 and 0.0578 mg kg−1 of quinalphos have been detected in two commercially available
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jafc.5b05369. G
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry
■
(8) John, H.; Worek, F.; Thiermann, H. LC−MS-based procedures for monitoring of toxic organophosphorus compounds and verification of pesticide and nerve agent poisoning. Anal. Bioanal. Chem. 2008, 391, 97−116. (9) Ishibashi, M.; Izumi, Y.; Sakai, M.; Ando, T.; Fukusaki, E.; Bamba, T. High-throughput simultaneous analysis of pesticides by supercritical fluid chromatography coupled with high-resolution mass spectrometry. J. Agric. Food Chem. 2015, 63, 4457−4463. (10) Sharma, D.; Nagpal, A.; Pakade, Y. B.; Katnoria, J. K. Analytical methods for estimation of organophosphorus pesticide residues in fruits and vegetables: a review. Talanta 2010, 82, 1077−1089. (11) Tekel, J.; Hatrík, S. Pesticide residue analyses in plant material by chromatographic methods: Clean-up procedures and selective detectors. J. Chromatogr. A 1996, 754, 397−410. (12) Alder, L.; Greulich, K.; Kempe, G.; Vieth, B. Residue analysis of 500 high priority pesticides: better by GC-MS or LC-MS? Mass Spectrom. Rev. 2006, 25, 838−865. (13) Rojano-Delgado, A. M.; Luque de Castro, M. D. Capillary electrophoresis and herbicide analysis: Present and future perspectives. Electrophoresis 2014, 35, 2509−2519. (14) Ravelo-Pérez, L. M.; Hernández-Borges, J.; Rodríguez-Delgado, M. A. Pesticides analysis by liquid chromatography and capillary electrophoresis. J. Sep. Sci. 2006, 29, 2557−2577. (15) Hernandez-Borges, J.; Frias-Garcia, S.; Cifuentes, A.; RodriguezDelgado, M. A. Pesticide analysis by capillary electrophoresis. J. Sep. Sci. 2004, 27, 947−963. (16) Picó, Y.; Rodríguez, R.; Mañes, J. Capillary electrophoresis for the determination of pesticide residues. TrAC, Trends Anal. Chem. 2003, 22, 133−151. (17) Kitagawa, F.; Otsuka, K. Recent applications of on-line sample preconcentration techniques in capillary electrophoresis. J. Chromatogr. A 2014, 1335, 43−60. (18) Breadmore, M. C.; Shallan, A. I.; Rabanes, H. R.; Gstoettenmayr, D.; Abdul Keyon, A. S.; Gaspar, A.; Dawod, M.; Quirino, J. P. Recent advances in enhancing the sensitivity of electrophoresis and electrochromatography in capillaries and microchips (2010−2012). Electrophoresis 2013, 34, 29−54. (19) Wen, Y.; Li, J.; Ma, J.; Chen, L. Recent advances in enrichment techniques for trace analysis in capillary electrophoresis. Electrophoresis 2012, 33, 2933−2952. (20) Wei, J. C.; Cao, J. L.; Tian, K.; Hu, Y. J.; Su, H. X.; Wan, J. B.; Li, P. Trace determination of five organophosphorus pesticides by using QuEChERS coupled with dispersive liquid−liquid microextraction and stacking before micellar electrokinetic chromatography. Anal. Methods 2015, 7, 5801−5807. (21) Soisungnoen, P.; Burakham, R.; Srijaranai, S. Determination of organophosphorus pesticides using dispersive liquid−liquid microextraction combined with reversed electrode polarity stacking modemicellar electrokinetic chromatography. Talanta 2012, 98, 62−68. (22) Zhou, L.; Luo, Z.; Wang, S.; Hui, Y.; Hu, Z.; Chen, X. Incapillary derivatization and laser-induced fluorescence detection for the analysis of organophosphorus pesticides by micellar electrokinetic chromatography. J. Chromatogr. A 2007, 1149, 377−384. (23) Aranas, A. T.; Guidote, A. M.; Haddad, P. R.; Quirino, J. P. Sweeping-micellar electrokinetic chromatography for the simultaneous analysis of tricyclic antidepressant and beta-blocker drugs in wastewater. Talanta 2011, 85, 86−90. (24) Dawod, M.; Breadmore, M. C.; Guijt, R. M.; Haddad, P. R. Strategies for the on-line preconcentration and separation of hypolipidaemic drugs using micellar electrokinetic chromatography. J. Chromatogr. A 2010, 1217, 386−393. (25) Rezaee, M.; Assadi, Y.; Milani Hosseini, M. R.; Aghaee, E.; Ahmadi, F.; Berijani, S. Determination of organic compounds in water using dispersive liquid−liquid microextraction. J. Chromatogr. A 2006, 1116, 1−9. (26) Ho, Y. M.; Tsoi, Y. K.; Leung, K. S. Highly sensitive and selective organophosphate screening in twelve commodities of fruits, vegetables and herbal medicines by dispersive liquid−liquid microextraction. Anal. Chim. Acta 2013, 775, 58−66.
Effects of dispersive solvent type and extraction solvent volume on the extraction recoveries of the investigated pesticides in UA-DLLME (Tables S1 and S2), effects of amounts of PSA and GCB on the extraction recoveries of the investigated pesticides in the cleanup procedure (Tables S3 and S4), optimized conditions for the procedures of UA-DLLME and sample cleanup (Table S5), comparison of UA-DLLME−sweeping-MEKC with other analytical methods for determination of OPPs in real samples (Table S6), and electropherograms of the L. chinense and D. opposite samples (Figures S1 and S2) (PDF)
AUTHOR INFORMATION
Corresponding Author
*Telephone: +853-8822-4874. Fax: +853-2884-1358. E-mail:
[email protected]. Funding
This research was supported by the National Natural Science Foundation of China (31160065), the Macau Science and Technology Development Fund (052/2012/A2 and 007/ 2014/AMJ), and the Research Committee of the University of Macau (MYRG109-ICMS13-LP and MYRG2014-00089ICMS-QRCM). Notes
The authors declare no competing financial interest.
■
ABBREVIATIONS USED MP, medicinal plant; OPP, organophosphorus pesticide; OCP, organochlorine pesticide; ECD, electron capture detector; NPD, nitrogen phosphorus detector; FPD, flame photometric detector; PDA, photodiode array; CE, capillary electrophoresis; MEKC, micellar electrokinetic chromatography; LLE, liquid− liquid extraction; SPE, solid-phase extraction; SPME, solidphase microextraction; LPME, liquid-phase microextraction; CPE, cloud point extraction; DLLME, dispersive liquid−liquid microextraction; UA-DLLME, ultrasound-assisted dispersive liquid−liquid microextraction; PSA, primary secondary amine; GCB, graphitized carbon black; EOF, electroosmotic flow; SDS, sodium dodecyl sulfate; BGS, background electrolyte solution; LOD, limit of detection; S/N, signal-to-noise; MRL, maximum residue limit; FDA, Food and Drug Administration; EU, European Union; RSD, relative standard deviation
■
REFERENCES
(1) Fan, T. P.; Yeh, J. C.; Leung, K. W.; Yue, P. Y.; Wong, R. N. Angiogenesis: from plants to blood vessels. Trends Pharmacol. Sci. 2006, 27, 297−309. (2) Huang, W. Y.; Cai, Y. Z.; Zhang, Y. Natural phenolic compounds from medicinal herbs and dietary plants: Potential use for cancer prevention. Nutr. Cancer 2009, 62, 1−20. (3) Zhu, H. L.; Wan, J. B.; Wang, Y. T.; Li, B. C.; Xiang, C.; He, J.; Li, P. Medicinal compounds with antiepileptic/anticonvulsant activities. Epilepsia 2014, 55, 3−16. (4) Zuin, V. G.; Vilegas, J. H. Pesticide residues in medicinal plants and phytomedicines. Phytother. Res. 2000, 14, 73−88. (5) Kosalec, I.; Cvek, J.; Tomić, S. Contaminants of medicinal herbs and herbal products. Arh. Hig. Rada Toksikol. 2009, 60, 485−501. (6) Jokanović, M.; Kosanović, M. Neurotoxic effects in patients poisoned with organophosphorus pesticides. Environ. Toxicol. Pharmacol. 2010, 29, 195−201. (7) Dömötörová, M.; Matisová, E. Fast gas chromatography for pesticide residues analysis. J. Chromatogr. A 2008, 1207, 1−16. H
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX
Article
Journal of Agricultural and Food Chemistry (27) Alves, A. C.; Gonçalves, M. M.; Bernardo, M. M.; Mendes, B. S. Dispersive liquid−liquid microextraction of organophosphorous pesticides using nonhalogenated solvents. J. Sep. Sci. 2012, 35, 2653−2658. (28) Yan, H.; Liu, B.; Du, J.; Row, K. H. Simultaneous determination of four phthalate esters in bottled water using ultrasound-assisted dispersive liquid−liquid microextraction followed by GC−FID detection. Analyst 2010, 135, 2585−2590. (29) Wang, W. X.; Yang, T. J.; Li, Z. G.; Jong, T. T.; Lee, M. R. A novel method of ultrasound-assisted dispersive liquid−liquid microextraction coupled to liquid chromatography−mass spectrometry for the determination of trace organoarsenic compounds in edible oil. Anal. Chim. Acta 2011, 690, 221−227. (30) Wu, J. W.; Chen, H. C.; Ding, W. H. Ultrasound-assisted dispersive liquid−liquid microextraction plus simultaneous silylation for rapid determination of salicylate and benzophenone-type ultraviolet filters in aqueous samples. J. Chromatogr. A 2013, 1302, 20−27. (31) Quirino, J. P.; Kim, J. B.; Terabe, S. Sweeping: concentration mechanism and applications to high-sensitivity analysis in capillary electrophoresis. J. Chromatogr. A 2002, 965, 357−373. (32) Wang, C. C.; Chen, J. L.; Chen, Y. L.; Cheng, H. L.; Wu, S. M. A novel stacking method of repetitive large volume sample injection and sweeping MEKC for determination of androgenic steroids in urine. Anal. Chim. Acta 2012, 744, 99−104.
I
DOI: 10.1021/acs.jafc.5b05369 J. Agric. Food Chem. XXXX, XXX, XXX−XXX