1102 A. Hercegovµ et al. J. Sep. Sci. 2006, 29, 1102 1109 Andrea Hercegovµ Milena Dömötörovµ DµÐa Kružlicovµ Eva Matisovµ Institute of Analytical Chemistry, Slovak University of Technology, Faculty of Chemical and Food Technology, RadlinskØho, Bratislava, Slovak Republic Original Paper Comparison of sample preparation methods combined with fast gas chromatography mass spectrometry for ultratrace analysis of pesticide residues in baby food Four sample preparation techniques were compared for the ultratrace analysis of pesticide residues in baby food: ( modified Schenck s method based on ACN extraction with SPE cleaning; (b) quick, easy, cheap, effective, rugged, and safe (QuEChERS) method based on ACN extraction and dispersive SPE; (c) modified QuEChERS method which utilizes column-based SPE instead of dispersive SPE; and (d) matrix solid phase dispersion (MSPD). The methods were combined with fast gas chromatographic-mass spectrometric analysis. The effectiveness of clean-up of the final extract was determined by comparison of the chromatograms obtained. Time consumption, laboriousness, demands on glassware and working place, and consumption of chemicals, especially solvents, increase in the following order QuEChERS a modified QuEChERS a MSPD a modified Schenck s method. All methods offer satisfactory analytical characteristics at the concentration levels of 5, 10, and 100 lg/kg in terms of recoveries and repeatability. Recoveries obtained for the modified QuE- ChERS method were lower than for the original QuEChERS. In general the best LOQs were obtained for the modified Schenck s method. Modified QuEChERS method provides 21 72% better LOQs than the original method. Keywords: Baby food / Fast GC-MS / Pesticide residues / Quadrupole mass selective detector / Sample preparation / Received: November 3, 2005; revised: January 10, 2006; accepted: January 31, 2006 DOI 10.1002/jssc.200500422 1 Introduction According to the status list of all active substances on the EU market (Status of active substances under EU review (doc. 3010), http://europa.eu.int/comm/food/plant/protection/evaluation/stat_active_subs_3010_en.xls), more than 1100 pesticides substances or products that kill harmful organisms (pests) are currently registered. There are significant economic benefits associated with the use of pesticides. Pesticides are used by farmers to improve or safeguard yields, to improve or protect quality of the produce, and to minimize labor input. Unfortu- Correspondence: Professor Eva Matisovµ, Institute of Analytical Chemistry, Slovak University of Technology, Faculty of Chemical and Food Technology, RadlinskØho 9, SK-812 37 Bratislava, Slovak Republic. E-mail: eva.matisova@stuba.sk. Fax: +421-2-52926043. Abbreviations: MSPD, matrix solid phase dispersion; PSA, primary secondary amine; PTV, programmed temperature vaporizer; QuEChERS, quick, easy, cheap, effective, rugged, and safe; R, recovery nately, these substances are often also harmful to nontarget organisms. Consumers can be subject to indirect exposure, due to the presence of pesticide residues in agricultural produce. Possible long-term health effects of pesticide residue exposure may include cancer, acute and chronic injury to the nervous system, neuromuscular and mental deficits, reproductive problems, and many others. Particular attention must be devoted to an especially sensitive population group children. The European Commission specified the maximum residual limit (MRL) of 10 lg/kg of pesticide residue content in cereal-based foods and baby foods for infants and young children [1]. In our recent publication [2] the modified Schenck s method connected with fast GC was used to analyze pesticide residues in baby food utilizing common routine instrumentation (with quadrupole mass spectrometric (MS) detection). Also matrix solid-phase dispersion (MSPD) was successfully combined with fast GC-MS on commercially available instrumentation analyzing
J. Sep. Sci. 2006, 29, 1102 1109 Sample preparation methods fast GC-MS for pesticide residues analysis 1103 apples (the most common raw material for baby food production) [3]. Nowadays the research in the field of sample preparation methods of plant matrix heads toward faster, less laborious, less harmful, and cost-effective procedures, preserving high recoveries and good precision, and are multiresidual and rugged. In 2003, Anastassiades et al. [4] introduced the quick, easy, cheap, effective, rugged, and safe (QuEChERS) method for the analysis of pesticide residues in fruit and vegetables. In a follow-up study, Lehotay et al. [5] conducted validation experiments of the QuEChERS method for more than 200 pesticides in several matrixes using GC-MS and LC-MS-MS. The results were satisfactory for nearly all pesticide residues in lettuce and orange. Leandro et al. [6] applied the QuEChERS method to the determination of 12 priority pesticides in baby food at a concentration level below 10 lg/kg using GC-MS-MS instrumentation. Connection of advantages of both modern approaches in sample preparation and fast GC techniques promises unquestionable benefits [7], such as high throughput, low GC operating cost, simplicity, higher sensitivity, and more reliable analytical answer. Our recent studies proved that the connection of fast GC utilizing microbore column (with id of 0.15 mm) with a common bench top quadrupole mass spectrometer provides efficient separation of 18 studied pesticides from the apple matrix components [2], sufficient number of data-points acquired for the proper peak shape reconstruction and good repeatability of peak areas in SIM mode, and the good overall system ruggedness after repeated matrix injections [8, 9]. The purpose of this study is to evaluate different sample preparation methods combined with fast GC-MS with a quadrupole bench top mass selective detector (MSD) for the determination of pesticide residues in baby food. Four sample preparation methods were compared: ( modified Schenck s method based on ACN extraction, followed by liquid liquid partitioning (LLP) of the crude extract and clean-up with SPE; (b) QuEChERS method based on ACN extraction and dispersive SPE; (c) modified QuEChERS method which utilizes SPE on columns instead of dispersive SPE; and (d) MSPD method, which involves cell disruption, homogenization, extraction into ethyl acetate, and a clean-up step in one step. Results of validation at the concentration levels of 5, 10, and 100 lg/kg; reached LOQs; as well as laboriousness, speed, and cleaning effectiveness are taken as criteria for method evaluation. The used chromatographic method in this paper was optimized with respect to the speed/ separation efficiency trade-off [10]. Analytical columns of inner diameter of 0.15 mm id were utilized due to a compromise between the time of analysis, transport phenomena, and column capacity concerning complex analytical samples. The 0.15 mm id columns allow more flexibility in loadability, sample introduction (flow rate), and operation when compared to 0.10 mm id columns [11]. Moreover, the ruggedness of 0.15 mm id columns was shown to be very good in the analysis of pesticide residues in plant matrix [9]. Several ways to reduction of matrix effects on the injector and column site were practiced [2, 9, 12, 13] (Document N SANCO/10476/2003, REV.4, http://europa.eu.int/comm/ food/plant/protection/evaluation/guidance/wrkdoc12_ en.pdf). 2 Experimental 2.1 Reagents and materials Pesticide reference standards of purity A 95% were obtained from various sources (Table 1). Stock solution of 0.5 mg/ml was prepared in toluene (Suprasolv, Merck, Darmstadt, Germany) and stored at 188C. Working standard pesticide mixtures prepared in acetone (Suprasolv, Merck) were used for preparation of spiked samples and matrix-matched standards. Standards were weighed on Sartorius Analytic MCI balances (Sartorius, Göttingen, Germany) with a precision of l10 lg. ACN and ethyl acetate used were of GC grade (Suprasolv, Merck). Magnesium sulfate (anhydrous powder) and NaCl p.a. was from Lachema (Neratovice, Czech Republic). The sorbent Florisil (60 100 mesh) was from Rotichrom, Roth, Karlsruhe, Germany. Bulk primary secondary amine (PSA) sorbent was obtained from Varian (Varian Incorporated, Harbor City, USA), the SPE columns used were 500 mg Bond-Elut- NH 2 (IST, Mid Glamorgan, UK). Quartz glass wool was Table 1. List of pesticides, chemical classes, and monitored ions utilizing GC-MS in SIM mode Pesticide Chemical class Retention time (min) Monitored ions in SIM mode, Target ion Dimethoate Organophosphorus 3.946 87, 125 Terbuthylazine Triazine 4.152 214, 229 Diazinon Organophosphorus 4.061 276, 304 Pyrimethanil Anilinopyrimidine 4.242 198, 199 Chlorpyrifos-methyl Organophosphorus 4.528 286, 288 Fenitrothion Organophosphorus 4.782 260, 277 Chlorpyrifos Organophosphorus 4.874 286, 314 Cyprodinil Anilinopyrimidine 5.219 224, 225 Penconazole Triazole 5.239 248, 250 Methidathion Organophosphorus 5.405 145, 302 Kresoxim-methyl Oximinoacetate 5.677 131, 132 Myclobutanil Triazole 5.770 179, 245 Tebuconazole Triazole 6.469 250, 252 Phosalone Organophosphorus 6.943 182, 367 Bitertanol 1 Triazole 7.409 168, 170 Bitertanol 2 7.475 Cypermethrin 1 Pyrethroid 7.811 163, 181 Cypermethrin 2 7.850 Cypermethrin 3 7.903 Etofenprox Nonester pyrethroid 7.972 163, 376 Pesticides are arranged according to retention times. Target ion used for quantification.
1104 A. Hercegovµ et al. J. Sep. Sci. 2006, 29, 1102 1109 from Agilent Technologies (Avondale, PA, USA). For filtration purposes glass fiber paper (Papírna PerÐtejn, Czech Republic) was used. For preconcentration purposes, nitrogen of purity A 99.99% (Linde, Technoplyn, Bratislava, Slovak Republic) was used. 2.2 Sample preparation The apples with peel were homogenously mixed with blender Braun MX 2050 (Kronberg, Germany). The apples used for validation purposes were checked by GC-MS for the studied pesticide residues and none of the selected ions were found at the corresponding retention times of the selected pesticides. 2.2.1 Modified Schenck s method Twenty-five grams of apple sample was extracted with 50 ml of ACN using sonication sonda of the pulsed ultrasonic cell disrupter VibraCell (Sonics and Materials, Danbury, CT, USA, CVX 400, frequency 20 khz). The ultrasonic pulses at 80% amplitude with duration of 3 s paused for 3 s were applied for 1 5 min. The extract was filtered through glass fiber paper (Papírna PerÐtejn), and the filtrate was transferred into an Erlenmayer flask with a tap. NaCl (2.5 g) was added and the mixture was shaken for 1 min. Phases were allowed to separate for 15 min. The upper ACN phase was transferred into an Erlenmayer flask, anhydrous magnesium sulfate (2 g) was added, and the mixture was shaken for 1 min. Twenty-five microliters of the dried extract was evaporated to less than 1 ml in a vacuum evaporator and transferred into an SPE-NH 2 column. Magnesium sulfate (1 cm layer) was always added to the top of the SPE column; the column was previously conditioned with acetone. The eluates were collected into 20 ml vials. Analytes were eluted with 15 ml of acetone and eluates were evaporated to dryness under a stream of nitrogen. The final volumes of the extracts were adjusted with toluene to 5 ml and analyzed by GC- MS. Preconcentration factor of the method is 2.5. 2.2.2 MSPD Five grams of a sample and 8 g of Florisil were blended in a glass mortar, transferred into a 25 cm615 mm id glass column plugged with glass wool and containing a layer of 2.5 g anhydrous magnesium sulfate. The column head was covered with a second 2 mm layer of anhydrous magnesium sulfate. The column was eluted with 60 ml of ethyl acetate by gravitational flow. The eluate was collected, concentrated to dryness using a rotary vacuum evaporator, then the dry rest was dissolved in 1 ml of toluene (preconcentration factor 5). 2.2.3 QuEChERS Ten grams of apple sample weighed into the 40 ml centrifuge tube was extracted with 10 ml of ACN using Ultra-Turrax (IKA, Germany) homogenizer at 19 000 rpm for 3 min. Then LLP followed: 1 g NaCl and 4 g MgSO 4 were added and the mixture was shaken by hand for 1 min. The mixture was then centrifuged at 3000 rpm for 5 min. Portion of the upper layer was transferred into a 10 ml centrifuge tube containing 25 mg PSA sorbent and 125 mg MgSO 4 per 1 ml of the cleaned extract. The mixture was shaken by hand for 1 min, and then centrifuged for 5 min at 3000 rpm to separate solids from solution. Extract of minimum 1 ml (Document N SANCO/ 10476/2003, REV.4, http://europa.eu.int/comm/food/ plant/protection/evaluation/guidance/wrkdoc12_en.pdf) was transferred into the vial and evaporated under N 2 to dryness to perform solvent exchange to toluene at preconcentration factor 1. 2.2.4 Modified QuEChERS After the first centrifugation of the original QuEChERS method, 5 ml of the upper layer was transferred onto an SPE column filled with acetone-conditioned 0.5 g of NH 2 sorbent covered with 1 cm layer of MgSO 4. The SPE column was eluted with 10 ml of acetone. The cleaned extract was evaporated under N 2 to dryness, and the solvent exchange to 2 ml of toluene was performed (preconcentration factor 2.5). 2.3 GC-MS GC MS measurements were performed on an Agilent 6890N GC coupled to 5973 MSD (Agilent Technologies, Avondale) equipped with a programmed temperature vaporizer (PTV) and autoinjector Agilent 7683. MS with electron impact ionization (EI) mode (70 ev) was operated in SIM mode; for each pesticide two specific ions were selected and sorted into groups (maximal number of ions in one group was eight); the used dwell time was 10 ms. PTV was operated in cold splitless mode. The injection volume was 2 ll. Helium with purity 5.0 (Linde Technoplyn) was used as a carrier gas. Microbore chromatographic column CP-Sil 8 CB (Varian, Middelburg, The Netherlands) with 5% diphenyl 95% dimethylsiloxane stationary phase 15 m60.15 mm id60.15 lm was utilized. It was connected to a nonpolar deactivated precolumn (1 m60.32 mm id) for focusation purposes and better ruggedness of the chromatographic system [9, 13]. Constant pressure mode 363.5 kpa was used until the elution of the last analyte (etofenprox, 7.90 min); additional pressure ramp (1000 kpa/min, 685 kp was used to speed up elution of higher boiling matrix coextractives. Chromatographic separation was performed under a temperature program, 1308C (1.13 min), 27.258C/min, 2908C (8 min). PTV conditions: temperature program, 1508C,
J. Sep. Sci. 2006, 29, 1102 1109 Sample preparation methods fast GC-MS for pesticide residues analysis 1105 4008C/min, 3008C (2 min), 4008C/min, 3508C (5 min); split vent open time 1.13 min. (The modified Schenck s method was validated at slightly different temperature and pressure conditions [2]: concerning oven temperature programme 1208C, 308C/min, 2908C (5 min), and constant carrier gas flow 0.5 ml/min.) 2.4 Real samples Apples were obtained from the suppliers of raw material to baby food production. Apples of cultivar Topaz were obtained from locality NovØ Zµmky. Apples of cultivar Jonagold were of origin from Dvory nad Žitavou and were used in the technological process of baby food production-apple pureø (Novofruct s.r.o., NovØ Zµmky, Slovak Republic). Sampling apple pureø after each step of the technological process was carried out by the following steps: (i) warming up to 608C for ca. 15 min; (ii) heating to 808C for 30 min; (iii) heating the mixture to 908C and (iv) sterilization at 988C for 40 min. 3 Results and discussion The studied pesticides were chosen according to the statistical evaluation of the consumption of pesticides applied to the apple trees in the south-west part of Slovakia. Farmers from this region have delivered the raw material to the producer of baby food. Pesticides also represent a wide range of polarity and other physicochemical properties. 3.1 Development of sample preparation methods Method development concerning the modified Schenck s method and MSPD are published in our recent publications [2, 3]. In the original QuEChERS procedure certain changes were made according to our needs and possibilities: Comminution with a chopping device with dry ice was replaced by mixing in a blender, and therefore the homogenization with Ultra Turrax was used at the extraction step instead of shaking to ensure good extraction efficiencies. Various extraction conditions were tested [14]. Five microliters of the extract after the first centrifugation was taken instead of 1 ml for the reason that more final extract was obtained. Solvent exchange (2 ml of ACN extract to 2 ml of final toluene extract) was included to permit better comparison of results and for the convenience to use the same GC method developed for the solvent toluene. Also the effectiveness of different amounts (25 75 mg) of PSA sorbent was assessed with regard to the quantity of coextractives remaining in the final extract. This was determined by the comparison of chromatograms of apple matrix obtained in the full scan mode. Twenty-five milligrams PSA sorbent/1 ml extract was found sufficient [14]. The main disadvantage of the original QuEChERS method compared to the other common methods is that the 1 g/ml final extract concentration is lower than the 2 5 g/ml concentrated extracts of the most traditional methods. If matrix is not the limiting source of noise in the analysis, this leads to higher LOQs for the same injection volume in the QuEChERS method. We investigated also the option of concentrating the final extracts to yield 2.5 g/ml (equally to modified Schenck s method). In order to reach cleaner extracts in the modified QuE- ChERS method, the dispersive SPE was replaced by cartridge-based SPE using NH 2 sorbent. Some interfering peaks were removed when moving from dispersive to column-based SPE. 3.1.1 Cleaning efficiency comparison A critical aspect of pesticide residue analysis is the purification, which is required to isolate the residues from matrix components to reduce undesirable matrix effects and is essential for the sufficient column capacity and the satisfactory long-term chromatographic performance during the analysis of a range of samples. The effectiveness of clean-up of the final extract (expressed by those coextractants which elute from the column under the given chromatographic conditions) was compared for all sample preparation methods. Similar chromatograms of unspiked apple samples in the full scan mode were obtained for the modified Schenck s and the modified QuEChERS method (both with concentration 2.5 g/ ml of the final extract). Two times concentrated MSPD final extract (5 g/ml) shows different coextractants, which, with their intensity, represent the worst burden on the chromatographic system. Figure 1 shows extracted ion chromatograms of selected pesticides in apple matrix in SIM mode. In all cases no significant interfering peaks were observed at the higher ion masses m/z. For lower, nonspecific m/z ions (pesticides methidathion m/z 145, kresoxim-methyl 131, and etofenprox 163) more impurities were detected in the areas of interest; however, the 15 m microbore column with 0.15 mm id provides satisfactory resolution in the case of original and modified QuEChERS and MSPD. Modified Schenck s method provides satisfactory cleaning also from interferences giving lower m/z ions. These observations relate to LOQs (Section 3.2.2). 3.1.2 Comparison on laboriousness and speed The differences among methods as well as the demands on chemicals, material, and the time were considered. The most time consuming step of the single sample preparation procedure is evaporation under N 2. If there is a possibility to evaporate more extracts at the same time (by simple arrangement of the device or using of a Turbovap), the time limiting factor of the preparation of more
1106 A. Hercegovµ et al. J. Sep. Sci. 2006, 29, 1102 1109 Figure 1. Extracted chromatograms of target ions of pesticides in the apple matrix-matched standard solution from fast GC-MS measurements in SIM mode. (A) Modified Schenck's method; (B) QuEChERS method; (C) modified QuEChERS method; (D) MSPD, concentration of pesticides 5 lg/kg. (1) Dimethoate; (2) terbuthylazine; (3) diazinon; (4) pyrimethanil; (5) chlorpyrifosmethyl; (6) fenitrothion; (7) chlorpyrifos; (8) cyprodinil; (9) penconazole; (10) methidathion; (11) kresoxim-methyl; (12) myclobutanil; (13) tebuconazole; (14) phosalone; (15) bitertanol; (16) cypermethrin; and (17) etofenprox.
J. Sep. Sci. 2006, 29, 1102 1109 Sample preparation methods fast GC-MS for pesticide residues analysis 1107 Figure 2. Results of the R experiments of pesticide residues from apples at spiking levels of 5, 10, and 100 lg/kg. samples becomes evaporation in a rotary evaporator. From this point of view the most time-consuming methods are the modified Schenck s method and MSPD. As in the modified Schenck s method it is necessary to carry out evaporation under N2 and the cleaning step on SPE columns, this method is unambiguously the most time consuming and the most laboriousness approach. The original QuEChERS method does not involve the solvent exchange step, and therefore it is the fastest and less laboriousness method. In terms of time, space, and financial demands, laboriousness, material, and solvents consumption we offer the following order: QuEChERS a modified QuEChERS a MSPD a Schenck s method. i 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 3.2 Validation results Experience tells us that if the method works well for the selected representative pesticides, then it should work equally well for nearly all of the others monitored routinely in multiclass, multiresidue methods [15]. 3.2.1 Recovery (R) studies R studies were carried out by spiking untreated apple samples with known volumes of the appropriate working mixtures of pesticides. The mixture was left to stand for 30 min. Blank samples were spiked with the pesticides at three concentration levels 5, 10, and 100 lg/kg.
1108 A. Hercegovµ et al. J. Sep. Sci. 2006, 29, 1102 1109 Except dimethoate in the case of QuEChERS and modified QuEChERS method and phosalone in the case of modified Schenck s method all pesticides analyzed by each sample preparation method tested met commonly accepted validation requirements of 70 110% recoveries [16] (Fig. 2). Vast majority of R results exceed 80%. Differences due to analyte concentration were minimal in terms of R for all methods and nearly all analytes. Recoveries obtained for the modified QuEChERS method were lower than those for the original QuEChERS for all pesticides except for pesticide cypermethrin at the level of 5 lg/kg. In terms of repeatability (expressed as RSD) the vast majority of pesticides gave a 10% RSD. Except four cases (three for MSPD and one for modified Schenck s method) all met the criterion of f20% RSD [16]. RSD increased to some extent as concentration decreased. Table 3. Pesticide etofenprox content in apple samples obtained from a supplier to baby food production analyzed by modified Schenck s, QuEhERS and modified QuEChERS sample preparation methods connected with fast GC-MS Method Average concentration (lg/kg) RSD PA (%) RSD GC (%) Modified Schenck s 14.2 1.28 0.59 QuEChERS 16.8 2.67 1.93 Modified QuEChERS 16.5 2.20 1.14 RSD PA RSD of analysis of two parallel sample extractions (two test portions), calculated by Eckschlager et al. [17]. RSD GC RSD of GC-MS measurements (the highest value is given), n = 2, calculated by Eckschlager et al. [17]. In all methods compared Ultra turrax was used in the first extraction step; 10 m 60.10 mm60.1 lm column was used instead of 15 m60.15 mm60.15 lm. 3.2.2 LOQs Calculation of LOQs was also performed to validate the method. The values are summarized in Table 2 for all tested pesticides and sample preparation methods. LOQ was determined as the concentration of a compound that gives a response which relates to the S/N = 10 (calculated by MS software). In general the best LOQs were obtained for the modified Schenck s method. Notably for nonspecific masses a 170 m/z (see Fig. 1) LOQs obtained for the modified Schenck s method were lower than those for the other three methods (except dimethoate with m/z 87 in the case of modified QuEChERS). Modified QuEChERS method provides 21 72% better LOQs than the original QuEChERS for all pesticides analyzed (except myclobutanil). LOQs of MSPD in some cases exceeds the LOQs obtained for the modified Schenck s method in more than one order. On the contrary, especially at high Table 2. LOQs (lg/kg) Pesticide Modified Schenck s method Original QuEChERS Modified QuEChERS method MSPD Dimethoate 0.38 0.54 0.30 2.67 Terbuthylazine 0.17 0.44 0.16 0.25 Diazinon 0.50 0.54 0.32 0.36 Pyrimethanil 0.07 0.15 0.07 0.07 Chlorpyrifos-methyl 0.18 0.19 0.08 0.06 Fenitrothion 0.33 0.40 0.22 0.16 Chlorpyrifos 0.46 0.69 0.34 0.30 Cyprodinyl 0.11 0.33 0.18 0.65 Penconazole 0.17 0.40 0.22 0.60 Methidathion 0.15 1.90 1.01 2.76 Kresoxim-methyl 0.22 1.05 0.65 0.68 Myclobutanil 0.14 1.15 1.44 2.00 Tebuconazole 0.29 1.52 0.43 0.59 Phosalone 0.73 1.52 1.20 0.57 Bitertanol 1 0.60 0.49 0.21 0.24 Bitertanol 2 1.44 2.31 1.02 2.25 Cypermethrin 1 0.08 3.22 2.33 1.42 Cypermethrin 2 0.38 2.04 1.31 2.64 Cypermethrin 3 0.17 3.97 3.43 5.03 Etofenprox 0.50 1.23 0.69 1.17 masses even lower LOQs compared with Schenck s method are obtained. 3.3 Application to real-life samples Apple samples (cultivar Topaz) obtained from suppliers from south-west Slovakia (raw material to the producer of baby food) were analyzed by three methods combined with fast GC-MS (Table 3). Pesticide etofenprox was found at the level exceeding MRL. The result obtained using the modified Schenck s method was lower compared to QuE- ChERS methods, but the maximal difference did not exceed 20%. Two test portions were taken and analyzed in three GC measurements to check the repeatability. Excellent repeatability (RSDs f 2.2%) was obtained. As another application the modified QuEChERS procedure was used for tracing the fate of pesticides during the technological process of baby food production. Also the raw apples (cultivar Jonatan) from a relevant consignment were sampled and analyzed for the comparison. Two pesticides, fenitrothion and pyrimethanil were detected and quantified after each step and also in raw apples. The results are presented in Table 4. The concentration of both pesticide residues decreases after the first steps. A probable explanation of this could be the degradation of the residues after the first heating step. Differences in the next steps are not significant, although a slight decrease of concentration was observed in general. 4 Concluding remarks Four tested sample preparation methods combined with fast GC-MS analysis reached the requirements of EU legislation on R, repeatability, [16] and LOQs for baby food [1]. Modified QuEChERS method utilizing column-based SPE clean-up instead of dispersive SPE provides lower recoveries, but slightly better LOQs. Modified Schenck s
J. Sep. Sci. 2006, 29, 1102 1109 Sample preparation methods fast GC-MS for pesticide residues analysis 1109 Table 4. Pyrimethanil and fenitrothion content determined by the modified QuEChERS method connected with fast GC-MS in the samples collected after each step of the technological process of baby food production and in raw apples (step 0) Traced pesticide Step of the technological process Average concentration (lg/kg) Pyrimethanil RSD PA % b) RSD GC % Average concentration (lg/kg) Fenitrothion RSD PA % RSD GC b) % 0 1.13 35.8 2.2 6.77 4.5 3.2 1 0.72 8.0 4.7 1.68 6.7 1.8 2 0.67 8.0 4.7 1.56 10.8 6.2 3 0.59 11.0 3.8 1.53 9.4 4.8 4 0.80 2.3 5.1 1.47 4.3 6.1 n =2. b) n =3. method with the best cleaning efficiency provides the lowest LOQs, while MSPD provides the worst cleaning related to the highest LOQs. QuEChERS method without any evaporation step offers extreme improvement in rapidity and simplicity in comparison with the other methods tested. It makes the QuEChERS method most attractive and useful in the ultratrace analysis of pesticides. The fast GC-MS technique using a quadrupole analyzer has been shown to be a valuable technique in pesticide residue analysis of complex mixtures and complex matrix, providing faster analytical answers with higher sensitivity compared to conventional capillary GC-MS. The authors gratefully acknowledge the support of a part of this research within the framework of the Slovak Grant Agency (VEGA, project no. 1/2463/05) and NATO project no. SfP 977 983. 5 References [1] Commission Directive 1999/39/EC amending Directive 95/5/EC, Off. J. Eur. Com. 1999, L 124/8. [2] Hercegovµ, A., Dömötörovµ, M., Matisovµ, E., Kirchner, M., et al., J. Chromatogr. A 2005, 1084, 46 53. [3] Dömötörovµ, M., Matisovµ, E., Kirchner, M., de Zeeuw, J., Acta Chim. Slov. 2005, 52, 422 428. [4] Anastassiades, M., Lehotay, S. J., ƒtajnbaher, D., Schenck, F. J., J. AOAC Int. 2003, 86, 412 430. [5] Lehotay, S. J., de Kok, A. D., Hiemstra, M., van Bodegraven, P., J. AOAC Int. 2005, 88, 595 614. [6] Leandro, C. C., Fussell, R. J., Keely, B. J., J. Chromatogr. A 2005, 1085, 207 212. [7] Matisovµ, E., Dömötörovµ, M., J. Chromatogr. A 2003, 1000, 199 221. [8] Kirchner, M., Matisovµ, E., Hrouzkovµ, S., de Zeeuw, J., J. Chromatogr. A 2005, 1090, 126 132. [9] Kirchner, M., Matisovµ, E., Otrekal, R., Hercegovµ, A., de Zeeuw, J., J. Chromatogr. A 2004, 1084, 63 70. [10] Klee, M. S., Blumberg, L. M., J. Chromatogr. Sci. 2002, 40, 234 247. [11] Dömötörovµ, M., Kirchner, M., Matisovµ, E., de Zeeuw, J., J. Sep. Sci., DOI: 10.1002/jssc.200500472. [12] HajÐlovµ, J., Zrostlíkovµ, J., J. Chromatogr. A 2003, 1000, 181 197. [13] Kirchner, M., Matisovµ, E., Dömötörovµ, M., de Zeeuw, J., J. Chromatogr. A 2004, 1055, 159 168. [14] Kružlicovµ, D., Thesis, Slovak University of Technology, Bratislava 2005. [15] Lehotay, S. J., MaÐtovskµ, K., Lightfield, A. R., J. AOAC Int. 2005, 88, 615 629. [16] Council Directives 94/43/EC, Off. J. Eur. Com., 1994, L 227 31. [17] Eckschlager, K., Horsµk, I., KodejÐ, Z., in: Hugovµ, E. (Eds.), Evaluation of Analytical Results and Methods, SNTL, Praha 1980, p. 25.