Anal. Chem. 2007, 79, 1492-1501
Liquid Chromatography Quadrupole Time-of-Flight Mass Spectrometry Analysis of Carbosulfan, Carbofuran, 3-Hydroxycarbofuran, and Other Metabolites in Food Carla Soler,*,† Brett Hamilton,‡ Ambrose Furey,‡ Kevin J. James,‡ Jordi Man˜es,† and Yolanda Pico´†
Laboratori de Bromatologia i Toxicologia, Facultat de Farma` cia, Universitat de Vale` ncia, Avenue Vicent Andre´ s Estelle´ s s/n, 46100 Burjassot, Spain, and PROTEOBIO, Mass Spectrometry Centre for Proteomics and Biotoxin Research, Department of Chemistry, Cork Institute of Technology, Bishopstown, Cork, Ireland
The potential of liquid chromatography quadrupole timeof-flight mass spectrometry (LC-QqTOF-MS) to identify and confirm carbosulfan and seven of its main metabolites (carbofuran, 3-hydroxycarbofuran, 3-ketocarbofuran, 3-hydroxy-7-phenol carbofuran, 3-keto-7-phenolcarbofuran, 7-phenolcarbofuran, dibutylamine) at trace levels from food is explored for the first time. The analytical method developed consists of pressurized liquid extraction (PLE) and LC-QqTOF-MS in positive ion mode, which attains unequivocal identification and quantification of the studied compounds in food, at levels well below of those of concern (0.05 mg/kg for the sum of carbosulfan, carbofuran, and 3-hydroxycarbofuran). PLE recoveries ranged from 55 to 94% with limits of quantification from 10 (for carbosulfan, carbofuran, 3-hydroxycarbofuran, and dibutylamine) to 70 µg/kg (3-keto-7-phenolcarbofuran). The method is precise, with relative standard deviations varying between 5 and 11% for the repeatability (withinday) and 8-13% for the reproducibility (interday). This method was used to monitor the presence and fate of the target compounds in orange, potato, and rice crops treated with a commercial product containing carbosulfan. Field degradation studies show that carbofuran, 3-hydroxycarbofuran, and dibutylamine are the main degradation products formed in the environmental disappearance of carbosulfan.
reason, the analysis of pesticides in food is a hot issue within the field of food safety, and it is of great importance to individuals and health organizations around the world.3 The ensuing problem of pesticide application can be aggravated by the formation of highly toxic degradation products. In this way, both the United States and the European Union, have set different documents that remark the necessity to assess the environmental impact of relevant degradation products and their identification and routine analysis, since some of those degradation products can be rather persistent and as hazardous as their parent compounds.4,5 A typical example is carbosulfan, which is a broad-spectrum insecticide, extensively used in the United States, Europe, and Asia for pest control in a wide range of crops, mainly citrus, potato, and rice.6 In the environment, the metabolism of carbosulfan involves hydroxylation or oxidation reactions, or both, to be metabolized first to carbofuran, which is actually more toxic than the carbosulfan itself,7,8 and then to 3-hydroxycarbofuran and 3-ketocarbofuran.9 Other metabolites of carbosulfan, 3-hydroxy7-phenolcarbofuran, 3-keto-7-phenolcarbofuran, 7-phenolcarbofuran, and dibutylamine, have also been reported.10 Despite that these metabolites are not abundant in the degradation process, their presence is also important as a consequence of their persistence in the environment. In recent years, liquid chromatography-mass spectrometry (LC-MS) techniques have almost replaced gas chromatography-
Extensive use of agrochemicals at various stages of cultivation for improving agricultural productivity has increased the concern of the consumers about the possible contamination of food.1 Many of these pesticides are harmful for humans and the environment owing to their toxic effects at short or long term.2 The toxicity level of a pesticide depends on the deadliness of the chemical, dose, length of exposure, and absorption by the body. For this
(3) Irace-Guigand, S.; Aaron, J. J.; Scribe, P.; Barcelo´, D. Chemosphere 2004, 55, 973-981. (4) U.S. Environmental Protection Agency. Pesticide Tolerances. Dermal Exposure Assessment: Principles and Applications; EPA/600/8-91./011B. Office of Research and Development, Washington, DC, 1992. (5) EU. Council Directive 91/414/EEC of 15 July 1991 concerning the placing of plant protection products on the market. (6) Carbosulfan 203. Carbosulfan (145) explanation. Food and Agricultura Organization (hptt:/www.fao.org/ag/AGP/AGPP/Pesticid/JMPR/Download/ 93_eva/carbosu.pdf). (7) de Melo, L. P.; Costa, L.; Lonardoni, L.; Pimentel, L. R. Chemosphere 2005, 60, 149-156. (8) Tevisan, M. J.; Casadei de Baptista, G.; Pimentel Trevizan, L. R.; Papa, G. Rev. Bras. Frut. 2004, 26, 230-236. (9) Detomaso, A.; Mascolo, G.; Lopez, A. Rapid Commun. Mass Spectrom. 2005, 19, 2193-2202. (10) Rasul Chaudhry, G.; Matten, A.; Kaskar, B.; Sardessai, M.; Bloda, M.; Bhatti, A. R.; Walia, S. K. FEMS Microbiol. Lett. 2002, 214, 171-176.
* To whom correspondence should be addressed. E-mail:
[email protected]. † Universitat de Vale`ncia. ‡ Cork Institute of Technology. (1) Pussemier, L.; Larondelle, Y.; Van Peteghem, C.; Huyghebaert, A. Food Control 2006, 17, 14-21. (2) Herna´ndez, F.; Sancho, J. V.; Pozo, O. J. Anal. Biochem. 2005, 382, 934946.
1492 Analytical Chemistry, Vol. 79, No. 4, February 15, 2007
10.1021/ac060709+ CCC: $37.00
© 2007 American Chemical Society Published on Web 01/23/2007
Table 1. Structures and Origin/Use of the Selected Pesticides and Metabolites
mass spectrometry for determining pesticides. LC/MS has advantages, such as the capability to determine transformation products (TPs) because most of them are more polar compared with the parent compounds, less volatile, and often thermolabile.11,12 The huge interest in the application of LC-MS techniques has significantly stimulated developments and improvements in mass analyzer technology.13 A particularly important and innovative LC-MS/MS configuration involves the combination of quadrupole (Q) and time-of-flight (TOF) mass spectrometers.14,15 LC-TOF-MS has already demonstrated its capability for the analysis of (unknown) compounds in food because of its high mass accuracy and inherent sensitivity in full scan mode that allows recording the full spectrum at all times, which is not possible with standard monitoring practices that use single-ion monitoring or multiple reaction monitoring.16 In this field, Ferrer et al.17 reported a LC-TOF-MS method for the identification of 15 pesticides in fruits and vegetables. Thurman et al.18 used the same technique for the nontarget searching and identification of postharvest fungicides and their degradation products in different citrus fruit extracts. However, the application of TOF is limited due to its incapability of providing structural information from fragmentation, which sometimes makes difficult the identification of compounds at trace levels in complex matrixes.18 QqTOF has been introduced as an (11) Thurman, E. M.; Ferrer, I.; Ferna´ndez-Alba, A. Comp. Anal. Chem. 2005, 43, 369-401. (12) Iba´n ˜ez, M.; Sancho, J. V.; Pozo, O. J.; Niessen, W.; Herna´ndez, F. Rapid Commun. Mass Spectrom. 2005, 19, 169-178. (13) Niessen, W. M. A. J. Chromatogr., A 2003, 1000, 413-436. (14) Ferrer, I.; Thurman, E. M. TrAC, Trends Anal. Chem. 2003, 22, 750(15) Blasco, C.; Font, G.; Pico´, Y. Mass Spectrom. Rev. 2004, 23, 45. (16) Ferrer, I.; Garcı´a-Reyes, J. F.; Ferna´ndez-Alba, A. R. Trends Anal. Chem. 2005, 24, 671-682. (17) Ferrer, I.; Garcı´a-Reyes, J. F.; Mezcua, M.; Thurman, E. M.; Ferna´ndezAlba, A. R. J. Chromatogr., A 2005, 1082, 81-90.
alternative strategy for more accurate identification of trace-level analytes,19 because of its ability to fragment and to provide exact mass measurements. Its applications are still very scarce due to several disadvantages attributed to the QqTOF that, in addition to its high cost, are the same reported for the TOF: low efficiency in obtaining quantitative information at trace levels, narrow dynamic ranges, little robustness, and lack of accuracy for quantitative purposes.20 The last few published studies21-23 are not in complete conformity with those “established disadvantages” and dealt with the confirmation and quantification of (unknown) compounds and their metabolites by LC-QqTOF-MS showing appropriate results. However, these studies were carried out only in matrixes such as water,12,24,25 urine, and plasma.26 To our knowledge, there is no work reporting the application of these techniques for the quantification and identification at trace level of pesticides in such complex matrixes as food. (18) Thurman, E. M.; Ferrer, I.; Zweigenbaum, J. A.; Garcı´a-Reyes, J. F.; Woodman, M.; Ferna´ndez-Alba, A. R. J. Chromatogr., A 2005, 1082, 71-80. (19) Chernushevich, I. V.; Loboda, A. V.; Thomson, B. A. J. Mass Spectrom. 2001, 36, 849-865. (20) Herna´ndez, F.; Iba´n ˜ez, M.; Sancho, J. V.; Pozo, O. J. Anal. Chem. 2004, 76, 4349-4357. (21) Herna´ndez, F.; Pozo, O. J.; Sancho, J. V.; lo´pez, F. J.; Marı´n, J. M.; Iba´n ˜ez, M. Trends Anal. Chem. 2005, 24, 596-612. (22) Hopfgartner, G.; Chernushevich, I. V.; Covey, T.; Plomley, J. B.; Bonner, R. J. Am. Soc. Mass Spectrom. 1999, 10, 1305-1314. (23) Barcelo´-Barrachina, E.; Moyano, E.; Galcera´n, M. T. J. Chromatogr., A 2004, 1054, 409-418. (24) Iba´n ˜ez, M.; Sancho, J. V.; Pozo, O. J.; Herna´ndez, F. Anal. Chem. 2004, 76, 1328-1335. (25) Bodeldijk, I.; Vissers, J. P. C.; Kearney, G.; Major, H.; van Leerdam, J. A. J. Chromatogr., A 2001, 929, 63-74. (26) Stolker, A. A. M.; Niesing, W.; Vreeken, R. J.; Niessen, W. M. A.; Brinkman, U. A. Th. Anal. Bioanal. Chem. 2004, 378, 1754-1761.
Analytical Chemistry, Vol. 79, No. 4, February 15, 2007
1493
Table 2. Exact Mass Measurements and Fragment Mass Errors for Studied Pesticide and Metabolites in a Potato Matched Standarda optimized parameters
observed mass (m/z)
CE ) 20 V DP ) 60 V FP ) 250 V DP2 ) 15 V
381.2212
CE ) 10 V DP ) 35 V FP ) 100 V DP2 ) 15 V
222.1137
CE ) 10 V DP ) 30 V FP ) 150 V DP2 ) 15 V
238.1067
CE ) 20 V DP ) 25 V FP ) 130 V DP2 ) 15 V
236.0916
CE ) 10 V DP ) 20 V FP ) 100 V DP2 ) 15 V
a
181.0850
CE ) 20 V DP ) 45 V FP ) 200 V DP2 ) 20 V
179.0715
CE ) 20 V DP ) 20 V FP ) 100 V DP2 ) 20 V
165.0925
CE ) 30 V DP ) 30 V FP ) 150 V DP2 ) 20 V
130.1585
major ions (m/z)
calcd mass (m/z)
elemental composition
error (ppm)
error (mDa)
Compound: Carbosulfan 381.2212 (40%) 381.2212 160.1152 (100%) 160.1160 118.0678 (70%) 118.0690 128.1423 (50%) 128.1439 165.0907 (20%) 165.0915
C20H33N2O3S+ C8H18NS+ C5H12NS+ C8H18N+ C10H13O2+
0 -4.99 -10.16 -12.49 -4.84
0 -0.8 -1.2 -1.6 -0.8
Compound: 222.1137 (51%) 165.0908 (100%) 123.0428 (90%) 137.0576 (20%) 147.0769 (15%)
C12H16NO3+ C10H13O2+ C7H7O2+ C8H9O2+ C10H11O+
3.15 -4.24 -14.63 -19.69 -27.19
0.7 -0.7 -1.8 -2.7 -4
Compound: 3-Hydroxycarbofuran 238.1067 (35%) 238.1079 220.0957 (100%) 220.0973 181.0850 (80%) 181.0864 163.0732 (40%) 163.0759
C12H16NO4+ C12H14NO3+ C10H13O3+ C10H11O2+
-5.04 -7.27 -7.73 -16.55
-1.2 -1.6 -1.4 -2.7
Compound: 3-Ketocarbofuran 236.0916 (30%) 236.0922 179.0696 (80%) 179.0708 151.0652 (70%) 151.0633 161.0589 (100%) 161.0602 208.0961 (20%) 208.0973
C12H14NO4+ C10H11O3+ C8H9NO2+ C10H9O2+ C11H14NO3+
-2.50 -6.70 12.58 -8.07 -5.77
-0.6 -1.2 1.9 -1.3 -1.2
Compound: 3-Hydroxy-7-phenolcarbofuran 181.0850 (25%) 181.0864 C10H13O3+ 163.0752 (100%) 163.0759 C10H11O2+ 135.0806 (80%) 135.0809 C9H11O+ 107.0508 (60%) 107.0497 C7H7O+
-7.7 -4.29 -2.22 -10.27
-1.4 -0.7 -0.3 -1.1
Compound: 3-Keto-7-phenolcarbofuran 179.0715 (25%) 179.0708 C10H11O3+ 161.0594 (100%) 161.0602 C10H9O2+ 151.0652 (40%) 151.0633 C8H9NO2+ 123.0431 (30%) 123.0446 C7H7O2+
-3.91 -4.97 -12.58 -12.19
-0.7 -0.8 -1.9 -1.5
Compound: 7-Phenolcarbofuran 165.0925 (20%) 165.0915 123.0452 (100%) 123.0446 137.0616 (40%) 137.0602 91.0530 (20%) 91.0547
C10H13O2+ C7H7O2+ C8H9O2+ C7H7+
6.06 4.87 1.02 -18.67
-1 -0.6 -1.4 -1.7
Compound: Dibutylamine 130.1585 (1005) 130.1595 74.0952 (80%) 74.0969 57.0685 (90%) 57.0704
C8H20N+ C4H12N+ C4H9+
7.68 22.94 33.29
-1 -1.7 -1.9
Carbofuran 222.1130 165.0915 123.0446 137.0603 147.0809
In italics, the compounds with abundance >40%
The objective of this paper is to demonstrate the potential of LC-QqTOF-MS in the identification and confirmation of metabolites of the pesticide carbosulfan, at trace levels, in food. QqTOF can offer the same perspectives of TOF to identify unknown compounds and, in addition, the capability to fragment analyte molecules. Specifically, this work is based on the unambiguous identification of carbosulfan metabolites in oranges, potatoes, and rice by QqTOF that will allow studying its metabolism under field conditions. It is aimed to highlight the possibilities of QqTOF to quantify and identify target compounds. EXPERIMENTAL SECTION Reagents and Chemicals. HPLC-grade Suprasolv methanol and organic trace analysis dichloromethane and ethyl acetate were purchased from Merck (Darmstat, Germany). Deionized water 1494
Analytical Chemistry, Vol. 79, No. 4, February 15, 2007
(>18 MΩ cm resistivity) was purified using a Milli-Q SP Reagent Water System (Millipore, Bedford, MA). All solvents were passed through a 0.45-µm cellulose filter purchased from Scharlau (Barcelona, Spain). Analytical grade reagent anhydrous sodium sulfate was also obtained from Scharlau. Standards of carbosulfan (2,3-dihydro-2,2-dimethylbenzofuran7-yl(dibutylaminothio) methyl carbamate), carbofuran (2,3-dihydro2,2-dimethylbenzofurany-7-yl methyl carbamate), 3-hydroxycarbofuran (2,3-dihydro-3-hydroxy-2,2-dimethylbenzofuran-7-yl methyl carbamate), and 3-ketocarbofuran (2,3-dihydro-3-oxy-2,2-dimethylbenzofuran-7-yl methyl carbamate), were provided by Riedel de Haen (Seelze, Germany). Metabolite standards of dibutylamine and 7-phenol (2,3-dihydro-2,2-dimethylbenzofuran-7-ol) were purchased from Aldrich (Madrid, Spain) and those of 3-keto-7-phenol
Figure 1. Full scan product ion spectrum chromatograms obtained by LC-QqTOF- MS/MS of (A) standard solution at LOQ level and (B) orange-matched standard solution fortified at LOQ level. Insets correspond to extracted ion chromatograms. Identification peaks: (1) dibutylamine, (2) 3-hydroxy-7-phenolcarbofuran, (3) 3-hydroxycarbofuran, (4) 3-keto-7-phenolcarbofuran, (5) 3-ketocarbofuran, (6) carbofuran, (7) 7- phenolcarbofuran, and (8) carbosulfan.
(2,3-dihydro-2,2-dimethyl-3-oxobenzofuran-7-ol) and 3-hydroxy-7phenol (2,3-dihydro-2,2-dimethylbenzofuran-3,7-diol) were synthesized by FMC Corp. (Philadelphia, PA). Table 1 lists the structures of carbosulfan and its metabolites selected for this study, their use, and their molecular weight. Individual pesticide stock solutions (1 mg/mL) were prepared in methanol and stored in stained glass-stopped bottles at 4 °C. All solutions, with the exception of carbosulfan, were found to be stable for at least 3 months. Carbosulfan was prepared once a week, and its precise concentration was verified by UV spectrophotometry using the molar extinction coefficients, after proper dilution with water.27 Aliquots of stock solutions were used to daily prepare standard mixtures of these compounds, at different concentrations, in methanol because carbosulfan photolyzes mainly to carbofuran and dibutylamine. The standard mixtures were used as spiking solutions for preparation of the matrix matched standards and in the recovery study. Study of Carbosulfan Persistence in Field Experiments. The study of carbosulfan metabolism was performed on orange, potato, and rice samples from different fields located in the Valencia community (Spain) area that were treated with carbosulfan. Carbosulfan was applied using a commercial product, Cenit (27) Soler, C.; Man ˜es, J.; Pico´, Y. J. Chromatogr., A 2006, 1109, 228-241.
(Afrasa, Valencia, Spain), that is an emulsifier concentrate of carbosulfan (25% p/v) at the level specified for each crop. In all cases, pesticide-free samples, before carbosulfan treatment, were used as the control blank, to spike sample aliquots at different concentrations for recovery studies and to prepare matrix matched standards for calibration. (a) Orange Samples. The experiments were carried out on Valencia Late cultivar oranges. The orchard consisted of four trees covering an area of 50 m2 with a plant spacing of 6 m × 6 m (10-year-old trees; 2.0-m height, 1.5-m diameter) in a citrus grove located in central western Valencia, receiving standard horticultural practices. The application of carbosulfan was made on May 20, 2005 at 1 kg/ha, according to the recommendations of the manufacturer. The insecticide solution was applied by using a pressurized hand gun sprayer at high volume to runoff, in order to obtain a good uniformity in insecticide distribution on the experimental plot. There was no rain during this period. Valencia Late is a late-season orange that ripens in the late spring through the early summer. This slow maturation in the final stages of development is in marked contrast to many other fruits, in which final changes occur in a few days. Furthermore, all fruits on a given tree are not at the same stage of maturity at one time. Analytical Chemistry, Vol. 79, No. 4, February 15, 2007
1495
Figure 2. (A) Full scan LC-QqTOF-MS/MS of carbosulfan with CE of 25V in (a) methanol, (b) orange-matched standard solution fortified at 10 µg/kg, and (c) treated orange sample (concentration ∼10 µg/kg). (B) Full scan LC-QqTOF-MS/MS of carbofuran with CE of 20 V in (a) methanol, (b) orange-matched standard solution fortified at 20 µg/kg, and (c) treated orange sample (concentration ∼20 µg/kg). (C) Full scan LC-QqTOF-MS/MS of 3-hydroxycarbofuran with CE of 20V in (a) methanol, (b) orange-matched standard solution fortified at 20 µg/kg, and (c) treated orange sample (concentration ∼20 µg/kg). (D) Full scan LC-QqTOF-MS/MS of dibutylamine with CE of 30 V in (a) methanol, (b) orangematched standard solution fortified at 10 µg/kg, and (c) treated orange sample (concentration ∼10 µg/kg).
(b) Potato Samples. The crops of Duquesa were grown in an experimental field located 30 km to the south of the Valencia city. This crop consisted of three 5-m2 plots, each with three rows 1496
Analytical Chemistry, Vol. 79, No. 4, February 15, 2007
of 10 potato plants. The crop was planted on July 20, 2005. The insecticide carbosulfan was sprayed onto the potato crop on September 13, 2005, 55 days after planting. The aqueous solution
Table 3. LC-QqTOF-MS/MS Ion Ratios (Aqualifying Ion/Aquantifying Ion) for Pesticides Fortified into Solvent and Matrix Samplea observed ratioc (RSD %)
pesticide
spiked level (µg/kg)
carbosulfan
10
carbofuran
10
3-hydroxycarbofuran
10
3-ketocarbofuran
50
3-hydroxy-7-phenolcarbofuran
20
3-keto-7-phenolcarbofuran
70
7-phenolcarbofuran
40
dibutylamine
10
ratio of product ions monitored 160.1152 118.0678 165.0908 123.0428 220.0957 181.0850 161.0589 179.0696 163.0752 135.0806 161.0594 151.0652 123.0452 137.0616 57.0685 74.0952
expected ratiob (RSD %)
orange
potato
rice
0.71 (10)
0.68 (7)
0.74 (9)
0.62 (10)
0.89 (7)
0.93 (8)
0.80 (6)
0.86 (8)
0.77 (5)
0.76 (6)
0.83 (5)
0.78 (12)
0.82 (6)
0.87 (8)
0.82 (9)
0.77 (10)
0.76 (5)
0.79 (7)
0.80 (3)
0.82 (7)
0.43 (8)
0.36 (8)
0.45 (6)
0.38 (9)
0.37 (12)
0.38 (8)
0.42 (10)
0.39 (7)
0.80 (4)
0.82 (11)
0.77 (6)
0.85 (10)
a The EU guidelines sets criteria for the observed ratio as follows; expected ratio >0.5, observed ratio should be within (20%, expected ratio 0.2-0.5, observed ratio should be within (25%; expected ratio 0.1-0.2, observed ratio should be within (30%; expected ratio 20 ppm. The poor accuracy in the exact masses of dibutylamine can be explained because its m/z is smaller than that of the other compounds, needing more collision energy for the fragmentation, and obtaining smaller m/z species. In any case, the error values for dibutylamine are always 70%). Only two metabolites dibutylamine, which is very soluble in water, and 7-phenolcarbofuran, which can be bond to some material components, are recovered in lower percentage. Considering the characteristics of the analytes and the aspects of the extraction procedure, recoveries could with difficulty be improved. In this sense, there are data in the literature that support that dichloromethane is the best solvent to simultaneously determine carbosulfan and their metabolites.7,8,27 Other solvents, such as hexane, acetone, acetonitrile, ethyl acetate, and different mixtures, were checked. Some of the analytes were not extracted and, for the other recoveries, are in general lower. Analytical Performance and Matrix Effects. Table 3 lists the extracted fragment ions that were monitored to quantify and identify, and the calculated ratio of their abundances. These ion ratios, obtained in a standard in methanol, were compared with those calculated for fortified orange, potato, and rice samples at the same concentration. Confirmatory analysis was found to be successful in all the cases. The quantification ion was the most abundant. The linearity of the method was calculated on the basis of the linear response to the limit of the quantitative measurements. Calibration graphs of standard solutions and spiked blank abstract areas versus concentration were constructed by the use of leastsquares linear regression analyses. The developed LC-QqTOFMS/MS method shows good linearity (r g 0.99, Table 4). Matrix effects, which occur during electrospray ionization, may alter the
signal intensity of an analyte in ES-MS analysis of real sample. These effects are reliably detected by comparing the slopes of the calibration obtained from the standard addition of the analytes into a sample extract and from the standards in methanol. Table 4 shows the slope ratios (matrix/solvent) in the three different matrixes. The insecticide and its seven metabolites presented slope ratios that indicate response reduction of (50% in orange and (70% in potato and rice. Therefore, a reliable quantification of carbosulfan and its metabolites from food samples using LCQqTOF-MS requires matrix-matched standards. The reproducibility and repeatability of the results obtained were evaluated on matrix-matched solutions at different concentration levels. The repeatability study was carried out by injection of the same standard solution five consecutive times in the same day (intraday values). The reproducibility study was carried out for five successive days using the same solution (interday values). The relative standard deviation (RSD) values obtained from runto-run and day-to-day precision are summarized in Table 5 at three different concentrations levels. From the results obtained, the developed method was found to be precise (with run-to-run instrumental RSD values between 3 and 9% and day-to-day RSD values between 5 and 13%). The LODs and LOQs of the method were experimentally estimated from the analysis of spiked orange, rice, and potato samples as the minimum concentration of an analyte giving signalto-noise ratios of 3 or 10, respectively. The results obtained in the three different matrixes are included in Table 6. All the tested matrixes presented similar LOQs for the compounds, ranging between 10 and 70 µg/kg. The accuracy of the method, determined as recoveries of each compound, was calculated by adding a standard mixture at LOQ and 10 times the LOQ concentration levels to noncontaminated sample. In the three studied matrixes, the recoveries for each compound were very similar. Good recoveries were observed (ranging between 60.3% for dibutylamine and 93.6%for carbofuran) except in the case of 7-phenolcarbofuran (
