4068
J. Agric. Food Chem. 2009, 57, 4068–4078 DOI:10.1021/jf900358r
Multiresidue Determination and Uncertainty Analysis of 87 Pesticides in Mango by Liquid Chromatography-Tandem Mass Spectrometry KAUSHIK BANERJEE,* DASHARATH P. OULKAR, SHUBHANGI B. PATIL, MANJUSHA R. JADHAV, SOMA DASGUPTA, SANGRAM H. PATIL, SUNITA BAL, AND PANDURANG G. ADSULE National Research Centre for Grapes, P.O. Manjri Farm, Pune 412 307, India
A liquid chromatography-tandem mass spectrometry (LC-MS/MS)-based method was optimized and validated for the multiresidue analysis of 87 pesticides in mango at the e10 ng g-1 level. The method involves extraction of 10 g of homogenized mango samples (+10 mL of water + 1 g of sodium acetate + 10 g of sodium sulfate) with 10 mL of ethyl acetate; cleanup by dispersive solidphase extraction with a combination of primary secondary amine (PSA, 50 mg), graphitized carbon black (GCB, 25 mg), and anhydrous sodium sulfate (150 mg); and final estimation by LC-MS/MS with multiple reaction monitoring. Direct analysis (no clean up) resulted in significant suppression in ionization of the majority of the test compounds over the electrospray ionization probe. However, clean up with the above combination of PSA + GCB reduced the matrix-induced signal suppressions significantly, and the signals in the cleaned extracts were nearly equivalent to the corresponding solvent standards. Substitution of PSA with florisil also gave equivalent clean up effects. The method was quite rugged as evident from a low Horwitz ratio (mostly 25 mg for 4 mL of ethyl acetate extract) affected the recovery of several pesticides viz. carbendazim, phosalone, paraoxon-methyl, diflubenzuron, forchlorfenuron, imazalil, emamectin benzoate, and spinosad due to surface adsorption on GCB. The comparative effect of different cleanup strategies is presented in Figure 1 for selected pesticides. For oxydemeton-methyl, cleanup with PSA + C18 (50 mg each) increased the peak area and S/N by more than two times as compared to the uncleaned extract, but the peak area decreased
Article
J. Agric. Food Chem., Vol. 57, No. 10, 2009
4073
Figure 1. Effect of different cleanup strategies on the peak area of selected pesticides at 10 ng g-1. Table 2. Mean Recovery ((RSD), HorRat, and Matrix Effect of the Test Pesticides recovery % (mean ( RSD) sr. no.
pesticide name
Ia
IIa
IIIa
IVa
HorRat (10 ng g-1)
ME (%)b
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
acephate acetamiprid alachlor atrazine azinphos-methyl azoxystrobin benalaxyl bitertanol buprofezin butachlor carbaryl carbendazim carbofuran carbofuran-3-OH clothianidin cymoxanil demeton-S-methyl demeton-S-methyl sulfone diazinon dichlofluanid dichlorvos difenconazole diflubenzuron dimethoate dimethomorph diniconazole DMSA emamectin benzoate ethion etrimphos famoxadone fenamidone fenarimol fenobucarb fenpyroximate fenthion flufenoxuron flusilazole forchlorfenuron hexaconazole imazalil imidacloprid indoxacarb
NDc 86 ((9) 88 ((18) 80 ((13) 83 ((29) 93 ((26) 91 ((5) 78 ((17) 73 ((12) 69 ((23) 96 ((11) 71 ((16) 95 ((12) 104 ((12) 81 ((32) 94 ((20) 128 ((58) 90 ((7) 81 ((13) 81 ((43) 42 ((84) 72 ((15) 78 ((20) 89 ((11) 81 ((9) 91 ((19) 81 ((13) 75 ((22) 60 ((24) 91 ((32) ND 86 ((10) 69 ((32) 94 ((10) 65 ((15) 71 ((26) ND 99 ((27) 92 ((11) 79 ((14) 90 ((9) 99 ((18) 60 ((33)
82 ((24) 91 ((7) 98 ((11) 96 ((6) 87 ((21) 85 ((8) 84 ((7) 76 ((19) 70 ((6) 63 ((20) 97 ((9) 85 ((10) 117 ((17) 98 ((9) 95 ((17) 105 ((13) 118 ((35) 89 ((7) 81 ((9) 92 ((20) 45 ((51) 74 ((11) 85 ((15) 85 ((9) 86 ((7) 84 ((16) 80 ((8) 73 ((12) 67 ((12) 75 ((28) 85 ((22) 84 ((9) 89 ((24) 92 ((9) 53 ((24) 75 ((14) 83 ((25) 88 ((21) 82 ((13) 79 ((13) 90 ((11) 89 ((8) 59 ((21)
78 ((10) 87 ((7) 94 ((5) 89 ((7) 84 ((12) 80 ((5) 79 ((6) 71 ((9) 78 ((11) 73 ((12) 93 ((5) 81 ((6) 90 ((6) 89 ((6) 85 ((9) 102 ((8) 116 ((18) 87 ((7) 83 ((8) 88 ((11) 53 ((28) 71 ((6) 92 ((14) 86 ((8) 77 ((7) 76 ((12) 85 ((7) 75 ((10) 66 ((7) 77 ((16) 96 ((13) 79 ((8) 75 ((16) 81 ((8) 69 ((17) 77 ((16) 84 ((17) 77 ((12) 80 ((8) 75 ((6) 84 ((7) 88 ((9) 58 ((12)
78 (9) 85 ((4) 98 ((7) 91 ((6) 88 ((9) 75 ((7) 73 ((6) 77 ((10) 75 ((9) 72 ((10) 93 ((4) 83 ((5) 91 ((5) 87 ((8) 81 ((8) 91 ((7) 96 ((12) 86 ((8) 88 ((5) 89 ((10) 52 (25) 76 ((7) 98 ((10) 90 ((4) 83 ((6) 79 ((9) 80 ((7) 76 ((8) 64 ((7) 79 ((10) 92 ((12) 80 ((9) 71 ((11) 74 ((6) 70 ((11) 69 ((10) 89 ((15) 77 ((7) 77 ((6) 71 ((7) 80 ((6) 82 ((6) 62 ((12)
0.32 0.22 0.16 0.22 0.38 0.16 0.19 0.28 0.35 0.38 0.16 0.19 0.19 0.19 0.28 0.25 0.57 0.22 0.25 0.35 0.88 0.19 0.44 0.25 0.22 0.38 0.22 0.32 0.22 0.50 0.41 0.25 0.50 0.25 0.54 0.50 0.54 0.38 0.25 0.19 0.22 0.28 0.38
-76.8 -4.3 -38 -38 -15.1 -12.6 -32.4 -18.5 -7.8 -51 -61.1 -14.7 -9.6 -8.9 -30 -21.3 -33.7 -25.4 -28.5 -36.4 -29.1 -28.9 -45 -10.3 -13.9 -34.6 -26.4 -25.1 -56.4 -29.5 -33.4 -15.9 -18.1 -27 -53.3 -42.8 -43.3 -31.9 -24.6 -31.6 -17.7 -26.8 -41.9
4074
J. Agric. Food Chem., Vol. 57, No. 10, 2009
Banerjee et al.
Table 2. Continued recovery % (mean ( RSD) sr. no. 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87
pesticide name iprovalicarb isoprothiolane isoproturon iprobenfos/kitazin kresoxim methyl malaoxon malathion mandipropamid metalaxyl methamidophos methidathion methomyl metribuzin mevinphos monocrotophos myclobutanil omethoate oxydemeton methyl paraoxon methyl penconazole pendimethalin phenthoate phosalone phosmet phosphamidon profenophos propargite propiconazole pyraclostrobin quinalphos simazine spinosyn A spinosyn D tebuconazole temephos tetraconazole thiacloprid thiamethoxam thiodicarb thiometon triadimefon triadimenol triazophos trifloxystrobin
Ia
IIa
IIIa
IVa
HorRat (10 ng g-1)
ME (%)b
93 ((7) 84 ((9) 95 ((6) 85 ((20) 114 ((30) 85 ((9) 95 ((11) 95 ((18) 89 ((13) 95 ((23) 103 ((13) 101 ((6) 79 ((33) 103 ((12) 98 (12) 98 ((15) 90 ((16) 87 ((8) 43 ((87) 81 ((14) ND 74 ((12) 70 ((22) 92 ((26) 88 ((12) 76 ((36) 60 ((23) 88 ((15) 70 ((8) 71 ((29) 65 ((30) 70 ((13) 86 ((15) 78 ((19) 40 ((41) 82 ((29) 94 ((13) 96 ((17) 55 ((6) 118 ((42) 79 ((21) 80 ((24) 83 ((9) 80 ((11)
89 ((5) 93 ((16) 94 ((6) 89 ((15) 89 ((17) 89 ((7) 93 ((10) 99 ((13) 90 ((6) 73 ((14) 106 ((9) 98 ((7) 98 ((15) 92 ((7) 92 (11) 94 ((15) 83 ((10) 83 ((7) 57 ((42) 86 ((9) 75 ((18) 74 ((10) 68 ((14) 94 ((17) 94 ((15) 79 ((18) 61 ((13) 82 ((14) 70 ((6) 82 ((15) 61 ((18) 75 ((19) 82 ((11) 80 ((13) 44 ((32) 82 ((17) 95 ((11) 95 ((15) 60 ((4) 124 ((30) 78 ((17) 81 ((10) 85 ((5) 74 ((7)
79 ((8) 82 ((5) 86 ((4) 87 ((16) 119 ((16) 85 ((9) 85 ((6) 97 ((14) 84 ((6) 73 ((15) 99 ((5) 89 ((6) 95 ((14) 87 ((5) 86 (8) 82 ((6) 78 ((11) 81 ((8) 75 ((15) 82 ((7) 74 ((12) 76 ((8) 72 ((8) 88 ((19) 89 ((12) 73 ((12) 61 ((14) 81 ((11) 76 ((8) 77 ((9) 79 ((18) 71 ((11) 78 ((10) 77 ((8) 42 ((27) 80 ((11) 85 ((4) 90 ((8) 56 ((7) 98 ((22) 76 ((8) 81 ((8) 80 ((7) 82 ((9)
73 ((6) 75 ((7) 78 ((5) 92 ((10) 100 ((19) 80 ((6) 78 ((10) 98 ((8) 77 ((3) 87 ((13) 85 ((8) 84 ((5) 85 ((10) 85 ((7) 89 (6) 86 ((6) 74 ((5) 74 (v7) 80 ((10) 73 ((5) 77 ((10) 78 ((8) 66 ((8) 92 ((19) 83 ((9) 75 ((8) 62 ((12) 72 ((9) 81 ((7) 77 ((7) 78 ((16) 73 ((10) 71 ((6) 72 ((9) 44 ((15) 77 ((7) 88 ((3) 83 ((6) 60 ((4) 93 ((20) 80 ((8) 88 ((6) 84 ((9) 87 ((6)
0.25 0.16 0.13 0.50 0.50 0.28 0.19 0.44 0.19 0.47 0.16 0.19 0.44 0.16 0.25 0.19 0.35 0.25 0.47 0.22 0.38 0.25 0.25 0.60 0.38 0.38 0.44 0.35 0.25 0.28 0.57 0.35 0.32 0.25 0.85 0.35 0.13 0.25 0.22 0.69 0.25 0.25 0.22 0.28
-22.8 -21.7 -15.8 -21.8 -18.2 -13.8 -13.8 -24.6 -9.1 -87.6 -72.5 -37.9 -21.5 -13.6 -13.9 -21.3 -62.4 -3.8 -33.3 -27.4 -35 -50.4 -38.8 -23.8 -11.2 -33.3 -63.3 -19.6 -27.4 -25.1 -29.5 -36.3 -42.7 -29.7 -69.7 -22.4 -26.5 -51.8 -15.2 -55.6 -28.3 -14.4 -19.4 -32.3
a I, 2.5 ng/g; II, 5 ng/g; III, 10 ng/g; and IV, 25 ng/g. b ME (%) pertains to matrix-induced signal suppressions in LC-(ESI)-MS/MS when the extracts were directly analyzed after solvent exchange without any cleanup. The “-” sign indicates signal suppressions. c ND, not detected.
by 15% when C18 was substituted with 25 mg of GCB, indicating adsorption of this chemical on the surface of GCB. In the case of spinosyn A and D, the peak area remained unchanged on cleanup with PSA + GCB, but the S/N reduced by nearly 60% when the extract was cleaned with PSA + C18, which might have occurred as a result of adsorption of this pesticide on C18 sorbent. For some azole derivatives like myclobutanil, tebuconazole, triadimefon, and tetraconazole and organophosphorus pesticides like acephate, ethion, methidathion, profenophos, temephos, and phosmet, the recoveries increased by nearly 20% when DSPE cleanup was performed with 50 mg of PSA + 25 mg of GCB. In general, the recoveries of all of the test pesticides were within the range of 70-120% at all four levels of fortifications except for selected compounds like acephate, famoxadone, flufenoxuron, and pendimethalin (Table 2), which were not detectable at 2.5 ng g-1 level. The addition of sodium acetate during extraction significantly improved the recoveries, which could be due to
its buffering as well as salting out effects during phase separation. The recoveries of the nonpolar pesticide temephos were low (4044%) at all of the fortification levels on account of its limited extraction in ethyl acetate, which is in agreement with our results reported earlier for grapes (4 ). The poor recovery and high RSD for dichlorvos could be due to its volatile and unstable nature. For some compounds, the RSD of recovery was >30% (Table 2) at the fortification levels of 2.5 and 5 ng g-1, but at 10 and 25 ng g-1 levels, their RSDs were within 20% except for dichlorvos, temephos, and thiometon, where RSDs were within 20-30%. The HorRat of most of the compounds was within 0.5 (Table 2), indicating satisfactory intralaboratory precision. The highest HorRat was for dichlorvos (0.88), which could be attributed to its relatively less repeatable signal because of its volatile nature. The results received from the six participating laboratories, which validated the current method as per our instructions, are similar to our results. The small-scale PT results were satisfactory
Article
J. Agric. Food Chem., Vol. 57, No. 10, 2009
Table 3. Results of Individual and Global Uncertainties for Each Pesticide uncertainty components (expressed as relative measures, calculated at 10.0 ng g-1) precision
sr. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
name of pesticides (group designationa in Roman numerals) acephate (I) acetamiprid (I) alachlor (I) atrazine (I) azinphos-methyl (IV) azoxystrobin (I) benalaxyl (I) bitertanol (II) buprofezin (III) butachlor (IV) carbaryl (I) carbendazim (I) carbofuran (I) carbofuran-3-OH (I) clothianidin (I) cymoxanil (I) demeton-S-methyl (III) demeton-S-methyl sulfone (I) diazinon (I) dichlofluanid (II) dichlorvos (III) difenoconazole (II) diflubenzuron (II) dimethoate (I) dimethomorph (I) diniconazole (II) DMSA (II) emamectin benzoate (II) ethion (II) etrimfos (II) famoxadone (II) fenamidone (I) fenarimol (II) fenobucarb (I) fenpyroximate (IV) fenthion (III) flufenoxuron (IV) flusilazole (II) forchlorfenuron (I) hexaconazole (I) imazalil (I) imidacloprid (II) indoxacarb (II) iprobenfos (I) iprovalicarb (I) isoprothiolane (I) isoproturon (I) kresoxim methyl (IV) malaoxon (I) malathion (I) mandipropamid (II) metalaxyl (I) methamidophos (II) methidathion (I) methomyl (I) metribuzin (II) mevinphos (I) monocrotophos (I) myclobutanil (I) omethoate (I) oxydemeton methyl (I) paraoxon methyl (II) penconazole (II)
accuracy/bias
calibration curve (U1)
U2
U3
U4
U5
global uncertainty (U)
expanded uncertainty (2U)
0.0182 0.0173 0.0190 0.0165 0.0164 0.0170 0.0183 0.0180 0.0177 0.0233 0.0173 0.0165 0.0193 0.0202 0.0203 0.0171 0.0256 0.0193 0.0203 0.0216 0.0212 0.0175 0.0156 0.0179 0.0186 0.0171 0.0171 0.0238 0.0183 0.0176 0.0210 0.0183 0.0215 0.0183 0.0244 0.0159 0.0268 0.0165 0.0178 0.0153 0.0166 0.0218 0.0158 0.0155 0.0171 0.0166 0.0170 0.0253 0.0189 0.0198 0.0145 0.0206 0.0184 0.0185 0.0203 0.0216 0.0228 0.0207 0.0216 0.0216 0.0240 0.0212 0.0228
0.0022 0.0013 0.0022 0.0015 0.0087 0.0016 0.0019 0.0025 0.0020 0.0052 0.0013 0.0016 0.0015 0.0023 0.0020 0.0023 0.0049 0.0016 0.0020 0.0028 0.0063 0.0022 0.0032 0.0014 0.0019 0.0031 0.0031 0.0031 0.0041 0.0038 0.0036 0.0013 0.0029 0.0015 0.0049 0.0058 0.0065 0.0027 0.0020 0.0017 0.0018 0.0027 0.0040 0.0013 0.0015 0.0013 0.0015 0.0063 0.0018 0.0014 0.0020 0.0015 0.0031 0.0017 0.0014 0.0024 0.0021 0.0013 0.0016 0.0016 0.0031 0.0031 0.0017
0.0026 0.0013 0.0024 0.0017 0.0019 0.0016 0.0022 0.0030 0.0065 0.0079 0.0011 0.0015 0.0015 0.0020 0.0017 0.0019 0.0042 0.0013 0.0017 0.0034 0.0040 0.0029 0.0029 0.0019 0.0020 0.0029 0.0028 0.0028 0.0032 0.0033 0.0039 0.0019 0.0029 0.0013 0.0063 0.0041 0.0049 0.0029 0.0009 0.0023 0.0020 0.0029 0.0035 0.0019 0.0013 0.0020 0.0013 0.0052 0.0013 0.0027 0.0034 0.0013 0.0031 0.0022 0.0014 0.0023 0.0019 0.0015 0.0016 0.0016 0.0013 0.0026 0.0028
0.0214 0.0127 0.0201 0.0126 0.0869 0.0164 0.0189 0.0254 0.0198 0.0522 0.0130 0.0161 0.0151 0.0228 0.0200 0.0229 0.0495 0.0157 0.0200 0.0282 0.0630 0.0224 0.0326 0.0175 0.0187 0.0308 0.0308 0.0305 0.0410 0.0375 0.0365 0.0134 0.0293 0.0150 0.0497 0.0586 0.0648 0.0276 0.0202 0.0178 0.0179 0.0270 0.0404 0.0136 0.0154 0.0128 0.0152 0.0628 0.0177 0.0141 0.0205 0.0150 0.0310 0.0173 0.0143 0.0242 0.0214 0.0131 0.0164 0.0164 0.0310 0.0313 0.0167
0.0262 0.0132 0.0210 0.0155 0.0190 0.0159 0.0225 0.0292 0.0647 0.0785 0.0109 0.0151 0.0148 0.0212 0.0169 0.0185 0.0416 0.0134 0.0170 0.0335 0.0396 0.0290 0.0303 0.0191 0.0196 0.0198 0.0198 0.0305 0.0323 0.0333 0.0405 0.0193 0.0284 0.0130 0.0635 0.0411 0.0476 0.0292 0.0097 0.0238 0.0120 0.0274 0.0354 0.0194 0.0134 0.0203 0.0132 0.0522 0.0131 0.0276 0.0350 0.0132 0.0319 0.0223 0.0143 0.0232 0.0196 0.0149 0.0162 0.0161 0.0013 0.0263 0.0278
0.039 0.025 0.035 0.026 0.091 0.029 0.035 0.043 0.070 0.098 0.024 0.028 0.029 0.037 0.033 0.034 0.070 0.028 0.033 0.049 0.078 0.041 0.047 0.032 0.033 0.041 0.041 0.049 0.056 0.053 0.059 0.030 0.046 0.027 0.085 0.074 0.085 0.044 0.029 0.034 0.027 0.044 0.056 0.028 0.027 0.029 0.026 0.086 0.029 0.037 0.043 0.029 0.048 0.034 0.029 0.040 0.037 0.029 0.032 0.032 0.039 0.046 0.040
0.077 0.051 0.070 0.052 0.182 0.057 0.069 0.086 0.141 0.195 0.049 0.055 0.057 0.074 0.066 0.068 0.140 0.057 0.067 0.098 0.155 0.082 0.095 0.063 0.066 0.081 0.081 0.099 0.111 0.107 0.117 0.060 0.093 0.054 0.169 0.147 0.170 0.087 0.057 0.067 0.055 0.089 0.112 0.057 0.053 0.059 0.053 0.172 0.058 0.074 0.087 0.058 0.097 0.068 0.057 0.080 0.074 0.057 0.063 0.063 0.079 0.092 0.080
4075
4076
J. Agric. Food Chem., Vol. 57, No. 10, 2009
Banerjee et al.
Table 3. Continued uncertainty components (expressed as relative measures, calculated at 10.0 ng g-1) precision
sr. no. 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 a
name of pesticides (group designationa in Roman numerals) pendimethalin (III) phenthoate (II) phosalone (II) phosmet (IV) phosphamidon (I) profenophos (II) propargite (III) propiconazole (I) pyraclostrobin (II) quinalphos (II) simazine (III) spinosyn A (III) spinosyn D (II) tebuconazole (III) temephos (IV) tetraconazole (II) thiacloprid (I) thiamethoxam (I) thiodicarb (II) thiometon (IV) triadimefon (I) triadimenol (I) triazophos (I) trifloxystrobin (II)
accuracy/bias
calibration curve (U1)
U2
U3
U4
U5
global uncertainty (U)
expanded uncertainty (2U)
0.0268 0.0238 0.0199 0.0164 0.0208 0.0219 0.0212 0.0224 0.0162 0.0218 0.0232 0.0193 0.0181 0.0240 0.0244 0.0183 0.0158 0.0242 0.0158 0.0540 0.0168 0.0205 0.0184 0.0193
0.0045 0.0022 0.0026 0.0087 0.0016 0.0028 0.0051 0.0016 0.0122 0.0025 0.0041 0.0042 0.0042 0.0044 0.0064 0.0026 0.0016 0.0016 0.0036 0.0097 0.0025 0.0023 0.0012 0.0028
0.0060 0.0027 0.0031 0.0019 0.0015 0.0034 0.0036 0.0027 0.0028 0.0036 0.0039 0.0050 0.0036 0.0043 0.0046 0.0028 0.0011 0.0015 0.0010 0.0130 0.0024 0.0016 0.0020 0.0033
0.0449 0.0221 0.0258 0.0869 0.0161 0.0276 0.0512 0.0156 0.0220 0.0223 0.0409 0.0426 0.0424 0.0437 0.0638 0.0263 0.0164 0.0158 0.0362 0.1000 0.0249 0.0233 0.0119 0.0276
0.0589 0.0272 0.0308 0.0190 0.0153 0.0344 0.0365 0.0266 0.0277 0.0362 0.0391 0.0502 0.0362 0.0434 0.0456 0.0286 0.0114 0.0154 0.0100 0.1302 0.0237 0.0159 0.0203 0.0333
0.079 0.043 0.045 0.091 0.031 0.049 0.067 0.038 0.041 0.048 0.061 0.069 0.059 0.066 0.083 0.043 0.026 0.033 0.041 0.174 0.038 0.035 0.030 0.048
0.158 0.085 0.090 0.182 0.061 0.099 0.133 0.076 0.082 0.096 0.123 0.138 0.118 0.133 0.165 0.086 0.051 0.066 0.082 0.347 0.077 0.070 0.060 0.095
Refer to the text in the Results and Discussion.
Figure 2. Z scores of the test pesticides in the PT sample for six participating laboratories.
Figure 3. Comparison of different sample preparation methods for selected pesticides at 10 ng g-1.
with the Z score of all of the participating laboratories being within +2 and -2 (Figure 2). Each laboratory could identify all of the target compounds, and their results were close to the true values. None of the participating laboratories reported any inconveniences in adopting this method or criticized its performance. We plan to do large-scale interlaboratory validation in our future endeavor. The cleanup effect could be attributed to the removal of fatty acids and sugars by PSA, whereas GCB was effective in removing carotenoids and any other plant pigments. Because β-carotene is the chief carotenoid compound in mango, its concentration in uncleaned and cleaned extracts was compared by HPLC to assess the cleanup effect. PSA alone could not remove any β-carotene as observed by HPLC analysis. However, DSPE with 25 mg of GCB
could remove more than 90% β-carotene from the ethyl acetate extract. An increase in GCB to 50 mg could completely remove the carotenoids, but it affected the recovery of several compounds. The addition of C18 sorbent did not result in any significant improvement in recoveries and hence was not considered. Substitution of PSA with florisil gave a nearly equivalent cleanup effect for the majority of the pesticides. This indicates that along with the carotenoids, the fatty acids might be the major coextractives. The cleanup effect rendered by PSA or florisil was evaluated by exploring into the coextracted matrix peaks of mono- and polyunsaturated fatty acids by gas chromatography-time-of-flight mass spectrometry (GCTOFMS), and the results showed significant removal of these interfering compounds upon cleanup. In general, the matrix
Article
J. Agric. Food Chem., Vol. 57, No. 10, 2009
4077
Figure 4. LC-MS/MS chromatogram of 87 pesticides in mango at 10 ng g-1.
influence was less for ethyl acetate when compared to acetonitrile extraction, which might be on account of less solubility of sugars and other polar matrix components in ethyl acetate. The Japanese method (2 ) involved multistep sample preparation, which caused losses in recoveries and, in turn, poor repeatability. The validation data generated at the NRL as well as the participating nominated laboratories indicate a high RSD in the range of 60-70% or even higher at 10 and 25 ng g-1 levels for a large proportion of the test compounds. In comparison, the corresponding RSDs in the current method were in general less than 10%. A comparative assessment for selected compounds is presented in Figure 3, which clearly establishes superiority of the current method over the Japanese method. The performance of the method was, however, statistically similar to the method of Mol et al. (3 ) or Lehotay (8 ) as per the Student’s t test. The use of ethyl acetate was also economically cheaper and toxicologically safer than acetonitrile used in the Japanese or Lehotay’s method and thus found to be more appropriate for extraction of a matrix like mangoes, which contains high sugar and less fat. Method Performance/Fitness for Purpose. All of the 87 pesticides could be analyzed by a single chromatographic run within 20 min (Figure 4). The dwell time of 10 ms was found to be
optimum for all of the compounds to achieve good peak shape having at least 15 data points across a peak at the 10 ng g-1 level. The linearity of the calibration curve was established for all of the pesticides. The correlation coefficient (R2) of the calibration curve was >0.99 (Table 1). For matrix-matched calibration, too, the R2 was >0.99 for most of the compounds. The LODs and LOQs for all of the compounds are presented in Table 1. Measurement Uncertainty of Analyses. The global uncertainty of the test pesticides varied until 10%, with the exception of thiometon (17.4%). On the basis of the global uncertainty values, the test pesticides could be classified into four groups: group I (U < 4%), group II (U ∼ 4-6%), group III (U ∼ 6-8%), and group IV (U > 10%). Forty-two compounds could be graded as group I, 28 as group II, 9 as group III, and 8 as group IV. The compounds classified under group I had lower uncertainties associated with precision (
