A

brcgagaara©

K

1

rtpTp

cglbD[Tv

0d

Available online at www.sciencedirect.com

Journal of Chromatography A, 1187 (2008) 25–33

Comparative study of pesticide multi-residue extraction in tobaccofor gas chromatography–triple quadrupole mass spectrometry

Jeong-Min Lee ∗, Jin-Won Park, Gi-Chul Jang, Keon-Joong HwangKT&G Central Research Institute, 302 Shinseong-dong, Yuseong-gu, Daejeon 305-805, Republic of Korea

Received 4 December 2007; received in revised form 1 February 2008; accepted 4 February 2008Available online 14 February 2008

bstract

This study was performed to apply the more rapid and accurate sample preparation, and the high selectivity and sensitivity of the analyte detectiony gas chromatography–triple quadrupole mass spectrometry (GC–MS/MS). We compared with the three-sample preparation methods of pesticideesidues in tobacco such as the conventional liquid–liquid extraction (LLE), pressurized liquid extraction (PLE) and QuEChERS (quick, easy,heap, effective, rugged and safe) methods. These methods were validated for 49 pesticides in the CORESTA Agrochemical Advisory Committeeuide and amenable to GC–MS/MS determination. LLE with acetonitrile, and PLE with acetone were followed by solid-phase extraction (SPE)nd analyzed by GC–MS/MS. In QuEChERS method, the effects of sorbents, such as primary secondary amine (PSA), octadecylsilane (C18) andraphitized carbon black (GCB), and matrix of the analytes in tobacco types, such as flue-cured, burley and oriental were investigated. MS/MS

cquisition provided high specificity and selectivity for pesticides and low limit of detection and quantification. QuEChERS by using PSA alonend the matrix-matched standards gave good recoveries and RSD values in three types of tobaccos. The method yielded higher and more consistentecoveries of these analytes than LLE and PLE. QuEChERS method was no needed to be complex clean-up procedure and would be used to rapidnd sensitive analysis of pesticides as an alternative to LLE and PLE methods. 2008 Elsevier B.V. All rights reserved.

ss spe

cbp(

nsrnodtqd

eywords: Tobacco; Pesticides; LLE; PLE; QuEChERS; Triple quadrupole ma

. Introduction

Nowadays, public interest in pesticide residues in foods andelated commodities has been increased. This situation has leto regulation setting maximum residue limits (MRLs) [1] ofesticide residues in agricultural products including tobacco.hus, the rapid multi-residue determination of wide range ofesticides in many samples has been required.

The multi-residue determination of pesticides in samplesonsists of extraction, clean-up and instrumental analysis. Theeneral sample preparation methods based on the conventionaliquid–liquid extraction (LLE) [2] and the automated instrumentased extraction technique, pressurized liquid extraction (PLE;ionex trade name ASE for accelerated solvent extraction)

3], followed by clean-up such as solid-phase extraction (SPE).hese methods are complicated and require large amount of sol-ents. In order to cover these problems, QuEChERS (quick, easy,

∗ Corresponding author. Tel.: +82 42 866 5389.E-mail address: jminlee@ktng.com (J.M. Lee).

mtmu

aa

021-9673/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.chroma.2008.02.035

ctrometry

heap, effective, rugged and safe) method was firstly establishedy using the dispersive-SPE [4]. After these sample preparationrocedures, analytes have been analyzed by gas chromatographyGC) or liquid chromatography (LC).

For GC separations, selective detectors (electron-capture,itrogen-phosphorus and flame photometric detection) or masspectrometry [5–8] are popular for detection of pesticideesidues. But, recently more sensitive analytical technique iseeded to qualify and quantify pesticide residues because manyf modern pesticides are either not amenable to GC or notetectable at sufficiently low levels. Instead of these detec-ors and single mass spectrometry, GC coupled with tripleuadrupole mass spectrometry is used for the rapid and sensitiveetermination of trace-level analytes [3,9]. Triple quadrupoleass spectrometry can increase the number of co-eluted pes-

icides to be analyzed if it is operated in selected reactiononitoring (SRM) mode that only monitors a few selected prod-

ct ions per analyte.Therefore, this study was performed to select the more rapid

nd accurate sample preparation of pesticide residues in tobaccos compared with the three sample preparations such as LLE,

2 atogr

Pwt

2

2

opffUDfolc1sdgM

(s

2

cpwscdr

2

2ati

cttp

2

a2a

wacd(

ltpTd

2

isv

TIftapeueodstr

ltpTd

22waiv4rwoAto2

6 J.M. Lee et al. / J. Chrom

LE and QuEChERS methods. Also the purpose of this studyas to apply triple quadrupole mass spectrometry (MS/MS) for

he high selectivity and sensitivity detection of analytes.

. Experimental

.1. Reagents and materials

Acetone, n-hexane and acetonitrile (pesticide grade) werebtained Mallinkrodt Baker (Phillipsburg, NJ, USA). Selectedesticides and the internal standard (I.S., mirex) were obtainedrom Riedel-de Haen (Seelze, Germany), Supelco (Belle-onte, PA, USA) and Chem Service (West Chester, PA,SA). Hydromatrix (diatomaceous earth) was purchased fromionex (Sunnyvale, CA, USA). Sorbents (40 �m particle size)

or dispersive-SPE included primary secondary amine (PSA)btained from Varian (Lake Forest, CA, USA), octadecylsi-ane (C18) from Merck (Darmstadt, Germany) and graphitizedarbon black (GCB) from Supelco. SPE cartridges containingg of florisil were purchased from Varian. Anhydrous magne-

ium sulfate, sodium chloride, trisodium citrate dihydrate andisodium hydrogencitrate sesquihydrate, which were all ACS-rade or better, were obtained from Sigma–Aldrich (St. Louis,O, USA).Three types of tobacco, flue-cured (B1O, Korea), burley

B1T, Korea) and oriental (Basma, Greece) were used in thistudy.

.2. Standard preparation

Stock standard solutions of individual pesticides with con-entration 100 �g/mL were prepared by exact weighing of theowder or liquid and dissolution in 100 mL of acetone, whichere then stored in a freezer at −19 ◦C. The working standard

olutions of multi compounds (approx. 0.01–0.4 �g/mL con-entration of each compound) were prepared by appropriateilutions of the stock solutions with acetone and stored in aefrigerator at 4 ◦C.

.3. Sample preparation

Three types of tobacco are dried in the oven set at 50 ◦C forh to obtain a mass fraction of moisture of approximately 5%fter drying. The tobacco is grounded through a 2 mm mesh,aking care to avoid heating above 50 ◦C. The grounded tobaccos stored in sealed containers and is excluded from light.

Recovery tests carried out three spiking levels and three repli-ates. A solution of analytes in acetone was spiked to tobaccoo achieve a concentration of 0.05, 0.5 and 1 �g/g for each pes-icide. Extraction procedures were performed as three samplereparation methods.

.3.1. Liquid–liquid extraction (LLE)

Ground tobacco (5 g) was exactly weighed into a glass beaker

nd 100 �L of I.S. solution (20 �g/mL of mirex in acetone) and5 mL of water was added, then 50 mL of acetonitrile was addednd homogenized by Ultra Turrax for 1 min. Sodium chloride

pofw

. A 1187 (2008) 25–33

ere added to homogenized sample. The extract was shakennd filtered. An aliquot of 5 mL was transferred to the coni-al tube and concentrated to dryness with a stream of N2. Theried residue was re-dissolved with 1.5 mL of acetone/n-hexane20:80).

After that, 1 mL aliquot of the concentrated extract wasoaded onto a florisil cartridge (1 g, 6 mL) previously condi-ioned with 5 mL of n-hexane and acetone/n-hexane (20:80). Theesticides were eluted with 5 mL of acetone/n-hexane (20:80).he eluates were concentrated to dryness with N2 and re-issolved with 1 mL of acetone.

.3.2. Pressurized liquid extraction (PLE)Extraction was carried out using an ASE 200 (Dionex). This

nstrument can extract automatically up to 24 samples. Stainlessteel extraction cells and glass collecting vials with 22 and 40 mLolumes, respectively, were used.

Ground tobacco (2.5 g) was mixed with 3 g of hydromatrix.he mixture was filled into the extraction cell. Then, 0.1 mL of

.S. solution (20 �g/mL of mirex in acetone) was added. Theree cell space was filled up with hydromatrix. The cell washen closed and placed on the PLE vial tray. Acetone was useds extraction solvent under the following PLE conditions: tem-erature: 100 ◦C; pressure: 1500 psi; heating time: 5 min; staticxtraction time: 3 min; flush volume: 60% of extraction cell vol-me; purge: N2, 60 s; number of cycles: 3. The total volume ofxtract obtained under those conditions was 50 mL and it showednly very little variations for analysis of different samples. Theseifferences were compensated with the presence of the internaltandard. An aliquot of 10 mL was transferred to the conicalube and concentrated to dryness with a stream of N2. The driedesidue was re-dissolved with 1 mL of acetone/n-hexane (20:80).

After that, 1 mL aliquot of the concentrated extract wasoaded onto a florisil cartridge (1 g, 6 mL) previously condi-ioned with 5 mL of n-hexane and acetone/n-hexane (20:80). Theesticides were eluted with 5 mL of acetone/n-hexane (20:80).he eluates were concentrated to dryness with N2 and re-issolved with 1 mL of acetone.

.3.3. QuEChERS preparation

.3.3.1. Extraction and clean-up. Ground tobacco (2 g) waseighed into a 50 mL centrifuge tube, 10 mL of water and

cetonitrile and 100 �L of I.S. solution (20 �g/mL of mirexn acetone) were added and the tube was closed and shakenigorously by shaker for 1 min. After shaking a mixture ofg magnesium sulfate, 1 g sodium chloride, 1 g trisodium cit-

ate dihydrate and 0.5 g disodium hydrogencitrate sesquihydrateere added. The tube was closed and immediately shaken vigor-usly by shaker for 1 min and centrifuge for 5 min at 6000 rpm.n aliquot of the extract was transferred into a PP-single use cen-

rifugation tube which contains following sorbents: (1) 25 mgf PSA and 150 mg of magnesium sulfate per mL extract; (2)5 mg of PSA, 25 mg of C18 and 150 mg of magnesium sulfate

er mL extract; (3) 25 mg of PSA, 2.5 mg of GCB and 150 mgf magnesium sulfate per mL extract. Tube is shaken vigorouslyor 30 s and centrifuged. After centrifugation the cleaned extractas transferred into a vial.

matog

2btoasi

2

geslosa

cnaua1t32ar

3

3

t

inwpritGs

wa(ognat

3

ticswGa

(sv

J.M. Lee et al. / J. Chro

.3.3.2. Preparation of calibration solutions. To prepare cali-ration solutions, a blank matrix is necessary. The blank wasreated as any other sample, without I.S. solution. An aliquotf the blank extract was fortified with the desired amount ofn pesticide mixture stock solution and a known amount of I.S.olution was added at approximately the same concentration asn the sample extracts.

.4. GC–MS/MS analysis

GC–MS/MS analysis was carried out on a Varian 3800as chromatograph/1200L triple quadrupole mass spectrom-ter (Varian, Walnut Creek, CA, USA) equipped with aplit/splitless injector and CombiPAL autosampler (CTC Ana-ytics, Zwingen, Switzerland). The data system contains allf the software required for calibration, collection of GC/MSpectra and data processing for qualitative and quantitativenalysis.

The chromatographic separation was performed by a VF-5msapillary column (30 m × 0.25 mm i.d., 0.25 �m film thick-ess, Varian, Middelburg, The Netherlands). Helium was useds the carrier gas at flow rate of 1.0 mL/min. Argon wassed as the collision gas. The injector temperature was sett 250 ◦C and 1 �L was injected in undividable mode formin. Samples were analyzed as following oven tempera-

ure program was used: initial temperature 50 ◦C (hold formin), ramp to 150 ◦C at 20 ◦C/min and finally ramp to80 ◦C at 2 ◦C/min (hold for 5 min). The manifold, sourcend transfer line temperature were set at 40, 230 and 250 ◦C,espectively.

. Results and discussion

.1. Optimization of GC–MS/MS conditions

To optimize MS/MS conditions, triple quadrupole mass spec-rometer was performed for the fragmentation of pesticides. MS,

se

m

Fig. 1. GC–MS/MS chromatogram f

r. A 1187 (2008) 25–33 27

n the selected ion monitoring and full scanning mode, couldot avoid matrix interference. Fragmentation of each analyteas caused by collisions of precursor ions. Product ions of eachesticide were chosen for the purposes of quantification andeduction of interferences. The product ions chosen were theons (m/z) with high intensity. The optimized MS/MS acquisi-ion was carried out with the transitions reported in Table 1, andC–MS/MS chromatogram of pesticides standard solution was

hown in Fig. 1.The limits of detections (LODs) of the MS/MS acquisition

ere evaluated by using solution of the different compounds incetone. LODs of pesticides ranged from 0.004 to 0.039 �g/mLTable 1). MS/MS acquisition could detect the low levelf pesticides without interference peak. The gas chromato-raphic separation of the target analytes in SRM mode wasot very important. Because this MS/MS acquisition wasble to determine several co-eluated analytes at the sameime.

.2. Matrix effect and recovery on sorbent types

Three types of sorbent, such as PSA, C18 and GCB were usedo investigate the influences on recovery rate and matrix effectn flue-cured tobacco (Table 2, Fig. 2). The matrix effects werealculated by comparing matrix-matched with solvent-matchedtandards. C18 was used to remove co-extracted fat and waxesith PSA and magnesium sulfate in the dispersive-SPE step.CB was used to remove carotenoids or chlorophyll with PSA

nd magnesium sulfate [4].As shown in Fig. 2, PSA alone and additional use of C18

PSA + C18) showed 35.4% of matrix effects (less than 20%ignal enhancement and suppression). In the distribution ofalidated pesticides, the additional use of GCB (PSA + GCB)

howed the 43.8% of matrix effect (less than 20% signalnhancement and suppression).

In the dispersive-SPE clean step, PSA sorbent and anhydrousagnesium sulfate alone were found to remove many matrix

or pesticide standard solution.

28 J.M. Lee et al. / J. Chromatogr. A 1187 (2008) 25–33

Table 1SRM transition for the different pesticides

No. Compound Typea Precursor ion (m/z) Product ion (m/z) Collision energy (eV) Retention time (min) LOD (�g/mL)

1 Heptenophos OP 250 124 −10 13.85 0.0102 Ethoporphos OP 200 97 −30 15.26 0.0093 Trifluralin DN 306 264 −15 15.87 0.0084 Benfluralin DN 292 160 −25 16.00 0.0185 Phorate OP 260 75 −10 16.64 0.0066 Quintozene OC 237 143 −30 18.22 0.0307 Terbufos OP 231 129 −30 18.69 0.0108 Diazinon OP 179 121 −25 19.09 0.0149 Disulfoton OP 274 88 −10 19.60 0.014

10 Pirimicarb CB 166 96 −25 20.41 0.02011 Phosphamidon OP 264 127 −25 21.12 0.02512 Chlorpyrifos methyl OP 286 93 −25 21.53 0.01013 Tolclofos methyl OP 265 93 −30 21.93 0.00814 Alachlor AM 188 130 −30 21.94 0.01015 Metalaxyl AM 206 132 −30 22.37 0.00516 Fenchlorphos OP 285 270 −30 22.46 0.01917 Pirimiphos methyl OP 305 180 −30 23.20 0.02118 Fenitrothion OP 277 109 −25 23.32 0.02019 Malathion OP 173 99 −30 23.90 0.03420 Chlorpyrifos OP 314 258 −30 24.16 0.00821 Fenthion OP 278 109 −20 24.45 0.02022 Dicofol OC 139 111 −15 25.05 0.00723 Butralin DN 266 174 −30 25.23 0.00824 Diphenamid AM 167 165 −30 25.48 0.02225 Isopropalin OP 280 238 −10 25.79 0.02026 Pendimethalin DN 252 162 −25 26.06 0.00427 Chlorfenvinphos OP 267 159 −25 26.70 0.01028 Phenthoate OP 274 121 −15 26.99 0.00629 Quinalphos OP 146 118 −10 27.02 0.00930 Methidathion OP 145 85 −10 27.70 0.00731 Flumetralin DB 143 107 −25 28.46 0.00432 Prothiofos OP 309 239 −15 29.46 0.01933 Profenofos OP 337 267 −15 29.75 0.02034 Oxadixyl AM 163 132 −10 32.38 0.01135 Ethion OP 231 129 −25 32.49 0.00636 Benalaxyl AM 148 105 −20 33.78 0.00837 Tebuconazole HC 250 125 −25 35.41 0.00638 cis-Permethrin PY 183 153 −15 43.35 0.01039 trans-Permethrin PY 183 153 −20 43.79 0.01140 Cyfluthrin 1 PY 226 206 −15 45.18 0.01041 Cyfluthrin 2 PY 226 199 −10 45.53 0.01042 Cyfluthrin 3 PY 226 206 −10 45.69 0.00843 Cyfluthrin 4 PY 226 206 −10 45.84 0.00844 Flucythrinate 1 PY 199 107 −15 46.73 0.00545 Flucythrinate 2 PY 199 107 −30 47.40 0.00446 Deltamethrin 1 PY 253 174 −30 50.77 0.01047 Deltamethrin 2 PY 253 174 −15 51.36 0.00648 Dimethormorph 1 HC 301 301 −5 52.29 0.01049 Dimethormorph 2 HC 301 301 −5 53.40 0.004

– Mirex I.S. 272 237 −25 40.39 –

ne; Ds

ctpGntpo

c[

3

a Abbreviations for pesticide type: OP, organophosphorus; OC, organochloritandard.

o-extractives without affecting pesticide recoveries. The addi-ional use of C18 did not affect recoveries, but neither did itrovide additional clean-up over PSA alone in tobacco samples.CB did a good job of removing additional matrix compo-

ents from the QuEChERS extracts, but unfortunately, it alsoended to remove certain pesticides, such as terbufos, disulfoton,irimicarb, metalaxyl and other analytes. Similar results werebtained from the past study that the removal of certain pesti-

ea

N, dinitramine; AM, amides; PY, pyrethroids; HC, heterocyclic; I.S., internal

ides by using GCB was shown in non-fatty fruits and vegetables11].

.3. Matrix effect and recovery on tobacco types

Three types of tobacco such as flue-cured, burley and ori-ntal were used to investigate the influences on recovery ratend matrix effect by using PSA sorbent in QuEChERS method

J.M. Lee et al. / J. Chromatogr. A 1187 (2008) 25–33 29

Table 2Recoveries and relative standard deviation of pesticides on sorbent types in flue-cured tobacco by QuEChERS method

Compound PSA PSA + C18 PSA + GCB

Recovery (%) RSD (%) Recovery (%) RSD (%) Recovery (%) RSD (%)

Heptenophos 81.4 7.5 118.7 8.7 122.1 12.1Ethoporphos 118.1 9.6 89.5 6.1 124.7 1.1Trifluralin 100.5 7.0 104.6 11.5 121.7 13.3Benfluralin 81.5 8.1 91.8 7.3 102.3 10.3Phorate 70.9 9.9 100.7 4.8 90.4 11.7Quintozene 82.6 10.5 70.0 15.6 118.6 11.8Terbufos 87.3 8.7 122.2 4.3 65.9 13.6Diazinon 71.9 11.6 113.9 9.7 90.1 8.3Disulfoton 85.6 7.4 111.3 9.0 62.7 9.4Pirimicarb 81.6 7.6 91.2 8.8 74.0 12.3Phosphamidon 104.1 9.5 92.1 8.9 87.9 7.5Chlorpyrifos methyl 83.3 7.4 111.9 11.1 73.3 15.2Tolclofos methyl 94.6 2.9 108.7 12.1 111.6 7.1Alachlor 76.8 5.8 102.3 10.7 122.7 13.6Metalaxyl 106.5 9.7 100.5 9.7 73.3 13.7Fenchlorphos 104.3 3.7 105.5 9.1 83.3 9.5Pirimiphos methyl 111.6 0.7 102.6 12.3 89.8 12.5Fenitrothion 94.8 8.8 117.8 10.4 122.9 10.1Malathion 88.0 11.2 93.1 5.3 86.5 9.4Fenthion 105.1 6.0 104.9 9.4 116.3 8.9Chlorpyrifos 99.2 1.2 57.6 10.0 120.1 7.2Dicofol 107.0 3.2 84.3 1.6 104.9 12.7Butralin 107.9 3.9 92.3 10.2 115.0 7.8Diphenamid 120.1 3.8 111.0 6.8 110.1 9.5Pendimethalin 81.7 2.9 86.2 14.5 89.6 13.4Isopropalin 105.1 6.0 106.2 12.9 129.0 8.8Chlorfenvinphos 102.3 6.9 121.8 3.4 91.6 7.4Phenthoate 98.7 10.9 94.8 11.8 109.6 9.9Quinalphos 98.4 3.5 109.7 10.5 111.3 1.6Methidathion 116.0 7.1 108.9 13.4 101.5 10.8Flumetralin 110.5 5.8 111.6 8.0 120.2 9.0Prothiofos 104.0 6.0 94.5 7.2 113.3 14.7Profenofos 101.7 7.6 113.6 2.5 91.5 8.1Oxadixyl 119.5 9.4 105.5 9.9 101.6 11.5Ethion 117.1 3.0 110.7 7.4 98.0 7.3Benalaxyl 98.9 6.9 97.7 2.0 93.2 8.9Tebuconazole 116.7 6.4 102.8 7.4 78.4 8.7cis-Permethrin 114.5 5.8 71.5 10.9 81.7 13.7trans-Permethrin 117.8 1.5 87.6 12.3 80.8 7.6Cyfluthrin 1 105.6 11.0 94.0 10.7 76.6 15.8Cyfluthrin 2 110.7 7.6 93.9 8.9 77.1 15.4Cyfluthrin 3 100.0 13.1 88.2 1.7 74.7 13.5Cyfluthrin 4 97.2 6.3 109.8 10.8 102.2 19.2Flucythrinate 1 98.0 5.4 95.8 15.7 107.3 14.8Flucythrinate 2 114.2 11.0 113.2 17.4 87.8 9.4Deltamethrin 1 97.4 8.0 109.7 15.5 84.3 14.7Deltamethrin 2 92.4 7.1 139.7 16.5 84.3 11.7Dimethormorph 1 119.5 13.9 119.7 7.9 108.0 10.1D 114.

(iiwtbtc

iLcd

imethormorph 2 113.5 4.3

Fig. 2, Table 3). The matrix effects were calculated by compar-ng matrix-matched with solvent-matched standards. As shownn Fig. 2, the matrix effects caused by different types of tobaccoere flue-cured 35.4%, burley 33.3% and oriental 41.7% (less

han 20% signal enhancement and suppression) in the distri-ution of validated pesticides, respectively. Tobacco samples ofhree types, which prepared QuEChERS by using PSA alone andalculated the matrix-matched standards, showed good recover-

toa

6 4.7 80.5 8.3

es and relative standard deviation (RSD) values. Anspach andinkerhagner [10] also reported that the matrix effects wereaused by different types of tobacco, and matrix-matched stan-ards were required to quantify.

As a result, one type of matrix-matched might be representa-ive for the other types. Nevertheless, for a proper quantificationf the analytes it is necessary to know the type of tobacco indvance.

30 J.M. Lee et al. / J. Chromatogr. A 1187 (2008) 25–33

nd to

3p

iLPm

oG

uF

EmA1

FT

Fig. 2. Influences of sorbent a

.4. Comparison of LLE, PLE and QuEChERS samplereparation methods

QuEChERS has been performed to extract pesticide residuesn tobacco, and the technique compared with conventionalLE and the automated instrument based extraction technique,LE for selection of the rapid and accurate sample preparation

ethod.Recoveries, RSD values and limits of quantitation (LOQ)

f these methods in flue-cured tobacco showed in Table 4.C–MS/MS chromatograms of flue-cured tobacco spiked

7tor

ig. 3. GC–MS/MS chromatograms of flue-cured tobacco spiked using (a) LLE, (bable 1 (I.S. refers to the internal standard).

bacco types on matrix effect.

sing LLE, PLE and QuEChERS methods were shown inig. 3.

LOQs of analytes by analysis of LLE, PLE and QuECh-RS methods were similar. The LOQs determined by theethod were below the guidance residue levels of CORESTACAC. Recoveries and RSD values were 42.9–123.4% and.1–21.1% for LLE, 39.0–136.8% and 1.8–21.1% for PLE, and

0.9–120.1% and 0.7–18.9% for QuEChERS, respectively. Ashe Table 4, results of QuEChERS method for the wide rangef 49 pesticides generally fell within the commonly acceptedange of 70–120% and ≤20% RSD for quantitative pesticide

) PLE, and (c) QuEChERS methods. Peak identification numbers are detailed

J.M. Lee et al. / J. Chromatogr. A 1187 (2008) 25–33 31

Table 3Recoveries and RSD of pesticides on tobacco types by QuEChERS method

Compound Flue-cured Burley Oriental

Recovery (%) RSD (%) Recovery (%) RSD (%) Recovery (%) RSD (%)

Heptenophos 81.4 7.5 106.4 11.5 92.2 7.7Ethoporphos 118.1 9.6 111.8 9.7 104.3 10.0Trifluralin 100.5 7.0 75.1 7.5 87.5 5.7Benfluralin 81.5 8.1 111.9 5.1 91.1 10.4Phorate 70.9 9.9 114.0 10.0 118.9 7.1Quintozene 82.6 10.5 92.0 13.2 71.8 12.1Terbufos 87.3 8.7 87.7 7.4 104.1 7.6Diazinon 71.9 11.6 110.7 2.7 84.1 12.1Disulfoton 85.6 7.4 98.2 9.9 81.8 10.9Pirimicarb 81.6 7.6 99.3 8.4 76.1 7.3Phosphamidon 104.1 9.5 108.4 8.4 89.2 8.9Chlorpyrifos methyl 83.3 7.4 100.6 12.5 117.3 13.1Tolclofos methyl 94.6 2.9 82.1 6.9 111.2 5.9Alachlor 76.8 5.8 94.0 8.1 89.9 2.1Metalaxyl 106.5 9.7 99.5 10.5 116.7 9.1Fenchlorphos 104.3 3.7 93.5 8.4 111.5 11.9Pirimiphos methyl 111.6 0.7 91.6 11.2 92.8 7.3Fenitrothion 94.8 8.8 85.2 10.9 106.8 9.5Malathion 88.0 11.2 101.6 6.5 114.0 10.3Fenthion 105.1 6.0 94.9 8.8 94.4 6.0Chlorpyrifos 99.2 1.2 75.0 5.5 70.5 10.9Dicofol 107.0 3.2 113.0 7.0 94.6 5.0Butralin 107.9 3.9 71.8 9.0 75.6 11.1Diphenamid 120.1 3.8 117.7 6.8 89.4 7.6Pendimethalin 81.7 2.9 73.3 5.9 74.0 8.8Isopropalin 105.1 6.0 98.8 9.6 107.8 7.6Chlorfenvinphos 102.3 6.9 127.1 11.9 77.0 15.0Phenthoate 98.7 10.9 119.3 5.4 95.5 9.7Quinalphos 98.4 3.5 119.0 12.3 82.4 14.8Methidathion 116.0 7.1 120.2 10.6 98.2 3.6Flumetralin 110.5 5.8 72.6 5.5 88.2 3.9Prothiofos 104.0 6.0 101.0 7.4 79.0 8.3Profenofos 101.7 7.6 102.0 6.6 90.5 8.3Oxadixyl 119.5 9.4 108.5 7.7 86.0 8.9Ethion 117.1 3.0 82.2 6.1 87.6 4.8Benalaxyl 98.9 6.9 102.3 9.3 95.5 7.9Tebuconazole 116.7 6.4 89.0 10.5 72.9 15.5cis-Permethrin 114.5 5.8 99.7 9.7 79.4 2.8trans-Permethrin 117.8 1.5 99.0 5.0 73.7 4.6Cyfluthrin 1 105.6 11.0 85.8 15.8 95.6 13.2Cyfluthrin 2 110.7 7.6 126.0 5.1 74.6 10.4Cyfluthrin 3 100.0 13.1 116.5 11.2 104.2 15.9Cyfluthrin 4 97.2 6.3 94.5 1.5 85.5 10.3Flucythrinate 1 98.0 5.4 79.4 6.6 93.3 8.3Flucythrinate 2 114.2 11.0 108.0 9.5 86.9 8.5Deltamethrin 1 97.4 8.0 80.9 11.3 91.2 8.5Deltamethrin 2 92.4 7.1 99.5 16.5 72.0 1.6D 96.D 93.

rsmTTmS

p

mrbfS

imethormorph 1 119.5 13.9imethormorph 2 113.5 4.3

esidue methods [10]. But, results of LLE and PLE methodshowed the poor values on 9 and 10 analytes such as terbufos,alathion and other analytes that were not in acceptance criteria.his was previously observed in case of the food matrices [11].he dispersive-SPE procedure, QuEChERS yielded higher and

ore consistent recoveries of these analytes than the traditionalPE procedure, LLE and PLE.

Compared with LLE and PLE, QuEChERS requires less sam-le weight and less organic solvent amount. Also, QuEChERS

pto

4 10.9 96.9 4.68 10.0 99.9 8.4

ethod was easier, cheaper and faster, and requires less appa-atus than the LLE and PLE methods using SPE. The slightlyetter clean provided by the SPE technique does not compensateor the better recoveries and practical advantages of dispersive-PE.

As a result, QuEChERS method does not require complicatedrocedure, and it could be a rapid, accurate and effective alterna-ive to LLE and PLE methods using SPE for the determinationf pesticide residues in tobacco.

32 J.M. Lee et al. / J. Chromatogr. A 1187 (2008) 25–33

Table 4Comparison of LLE, PLE and QuEChERS methods

Compound LLE PLE QuEChERS

Recovery (%) RSD (%) LOQ (�g/g) Recovery (%) RSD (%) LOQ (�g/g) Recovery (%) RSD (%) LOQ (�g/g)

Heptenophos 102.7 9.1 0.030 111.0 11.2 0.020 81.4 7.5 0.040Ethoporphos 110.8 13.7 0.027 103.8 7.0 0.018 118.1 9.6 0.036Trifluralin 102.3 8.8 0.024 136.8 10.2 0.016 100.5 7.0 0.032Benfluralin 86.0 8.0 0.054 111.9 8.8 0.036 81.5 8.1 0.072Phorate 72.7 11.4 0.018 43.1 8.6 0.012 70.9 9.9 0.024Quintozene 82.1 5.7 0.090 76.5 13.2 0.060 82.6 10.5 0.120Terbufos 42.9 15.3 0.029 64.2 5.9 0.019 87.3 8.7 0.039Diazinon 73.0 3.8 0.041 76.4 12.7 0.027 71.9 11.6 0.054Disulfoton 87.5 6.3 0.042 70.2 5.6 0.028 85.6 7.4 0.056Pirimicarb 50.2 9.3 0.060 80.0 10.9 0.040 81.6 7.6 0.080Phosphamidon 77.9 7.9 0.075 81.1 9.2 0.050 104.1 9.5 0.100Chlorpyrifos methyl 70.7 1.1 0.030 96.1 5.9 0.020 83.3 7.4 0.040Tolclofos methyl 68.9 5.0 0.024 91.1 5.7 0.016 94.6 2.9 0.032Alachlor 86.0 4.4 0.029 88.9 7.5 0.019 76.8 5.8 0.039Metalaxyl 92.7 8.7 0.014 70.1 4.9 0.010 106.5 9.7 0.019Fenchlorphos 54.3 7.1 0.057 90.9 12.9 0.038 104.3 3.7 0.076Pirimiphos methyl 92.8 2.7 0.063 78.6 10.9 0.042 111.6 0.7 0.084Fenitrothion 74.2 6.6 0.060 106.9 13.3 0.040 94.8 8.8 0.080Malathion 44.9 8.4 0.102 39.0 9.0 0.068 88.0 11.2 0.136Fenthion 82.1 7.3 0.025 52.0 7.1 0.017 105.1 6.0 0.033Chlorpyrifos 98.0 3.4 0.060 81.3 6.7 0.040 99.2 1.2 0.080Dicofol 73.5 1.5 0.022 95.7 1.8 0.015 107.0 3.2 0.030Butralin 81.1 6.8 0.025 104.8 7.5 0.016 107.9 3.9 0.033Diphenamid 77.1 1.2 0.065 52.4 8.6 0.043 120.1 3.8 0.086Pendimethalin 84.4 5.0 0.059 72.6 8.2 0.039 81.7 2.9 0.157Isopropalin 71.5 3.2 0.011 75.5 4.0 0.007 105.1 6.0 0.015Chlorfenvinphos 61.0 9.2 0.031 59.7 18.4 0.020 102.3 6.9 0.041Phenthoate 82.6 2.3 0.019 53.6 8.8 0.013 98.7 10.9 0.026Quinalphos 81.1 5.8 0.027 77.8 9.5 0.018 98.4 3.5 0.036Methidathion 79.0 7.9 0.021 76.0 12.9 0.014 116.0 7.1 0.028Flumetralin 85.1 6.1 0.012 85.3 7.0 0.010 110.5 5.8 0.016Prothiofos 93.5 8.3 0.057 94.2 5.7 0.038 104.0 6.0 0.100Profenofos 90.1 8.1 0.060 70.1 6.5 0.040 101.7 7.6 0.080Oxadixyl 65.2 7.7 0.034 85.2 9.3 0.023 119.5 9.4 0.045Ethion 60.4 14.1 0.017 48.8 10.4 0.011 117.1 3.0 0.022Benalaxyl 67.9 10.3 0.024 54.8 6.7 0.016 98.9 6.9 0.032Tebuconazole 77.4 11.8 0.018 49.0 7.7 0.012 116.7 6.4 0.024cis-Permethrin 81.7 7.0 0.030 89.8 6.4 0.020 114.5 5.8 0.040trans-Permethrin 89.5 4.6 0.032 95.2 8.8 0.021 117.8 1.5 0.043Cyfluthrin 1 105.5 6.8 0.030 70.7 7.9 0.020 105.6 11.0 0.040Cyfluthrin 2 108.2 7.8 0.030 112.9 11.0 0.020 110.7 7.6 0.040Cyfluthrin 3 107.2 8.7 0.024 89.8 7.8 0.016 100.0 13.1 0.032Cyfluthrin 4 101.2 4.9 0.024 79.9 6.6 0.016 97.2 6.3 0.032Flucythrinate 1 56.0 13.7 0.021 61.5 7.1 0.014 98.0 5.4 0.019Flucythrinate 2 75.4 18.4 0.017 72.9 12.2 0.011 114.2 11.0 0.015Deltamethrin 1 95.4 7.2 0.030 103.5 9.3 0.020 97.4 8.0 0.040DDD

4

apipftm

flbly

eltamethrin 2 105.4 6.4 0.018 92.4imethormorph 1 113.4 10.7 0.030 101.2imethormorph 2 123.4 16.7 0.022 94.2

. Conclusions

Dispersive-SPE method (QuEChERS) and SPE method (LLEnd PLE) carried out to select the rapid and accurate samplereparation method for the determination of pesticide residuesn tobacco by GC–MS/MS. MS/MS technique was sensitive for

esticide analysis in tobacco. QuEChERS method showed dif-erent recoveries and matrix effects on sorbent types and tobaccoypes. QuEChERS method by using PSA alone and the matrix-atched standards gave good recoveries and RSD values in

R

5.9 0.012 92.4 7.1 0.02410.2 0.020 119.5 13.9 0.04015.2 0.018 113.5 4.3 0.016

ue-cured, burley and oriental tobaccos. The LOQs determinedy the QuEChERS method were below the guidance residueevel of CORESTA ACAC. Conventional LLE and PLE methodsielded lower recoveries than QuEChERS.

eferences

[1] L. Mueller, M.R. Ward, in: D.L. Davis, M.T. Nielsen (Eds.), Tobacco,Production, Chemistry and Technology, Blackwell Science, London, 1999,p. 250.

matog

J.M. Lee et al. / J. Chro

[2] Y.R. Tahboub, M.F. Zaater, Z.A. Al-Talla, J. Chromatogr. A 1098 (2005)150.

[3] J. Haib, I. Hofer, J.M. Renaud, J. Chromatogr. A 1020 (2003) 173.

[4] M. Anastassiades, S.J. Lehotay, D. Stajnbaher, F.J. Schenck, J. AOAC Int.

86 (2003) 412.[5] International Standard Organization, ISO 4389:2000(E), Geneva, 2000.[6] C.S. Barrio, J.S. Asensio, J.G. Bernal, Chromatographia 39 (1994) 320.[7] G. Dugo, G.D. Bella, L.L. Torre, M. Saitta, Food Control 16 (2005) 435.

[

[

r. A 1187 (2008) 25–33 33

[8] B. Albero, C. Sanchez-Brunete, J.L. Tadeo, J. Agric. Food Chem. 52 (2004)5828.

[9] A.G. Frenich, M.J. Gonzalez-Rodriguez, F.J. Arrebola, J.L. Martinez Vidal,

Anal. Chem. 74 (2005) 4640.

10] T. Anspach, M. Linkerhagner, Proceedings of the Paper Presented at theCooperation Centre for Scientific Research Relative to Tobacco Congress,Paris, France, October 20, 2006, p. 134.

11] S.J. Lehotay, K. Mastovska, S.J. Yun, J. AOAC Int. 88 (2005) 630.