Advertisement

Analytical and Bioanalytical Chemistry

, Volume 410, Issue 22, pp 5491–5506 | Cite as

Further improvements in pesticide residue analysis in food by applying gas chromatography triple quadrupole mass spectrometry (GC-QqQ-MS/MS) technologies

  • Elena Hakme
  • Ana Lozano
  • Samanta Uclés
  • Amadeo R. Fernández-Alba
Research Paper
Part of the following topical collections:
  1. Food Safety Analysis

Abstract

Nowadays, the control of pesticide residues in food is well established. The capacity of triple quadrupole technology to satisfy the current food regulations has been demonstrated. However, the permanent high demand of consumers for more sensitive and faster testing is driving the development of improved analytical methodologies that increase the performances of sensitivity and robustness and reduce the analysis time. In this work, the feasibility of decreasing the run time to 12.4 min by modifying the oven temperature program, for a multiresidue method covering 203 pesticides, was evaluated. Satisfactory sensitivity results were achieved by reaching a limit of quantitation of 2 μg kg−1 for a great variety of fruits and vegetables. The validated method based on updated GC-QqQ-MS/MS has confirmed the abovementioned challenges with adequate robustness by its application to routine analyses for 69 real samples. The proposed method can represent great benefit for laboratories as it allows increasing samples throughput. It is also very useful for risk assessment studies, where the needs of low reporting limits and very wide analytical scope are necessary.

Keywords

Fast GC-MS/MS analysis Low reporting limits Multiresidue method Method validation 

Introduction

The interest in the development of a fast gas chromatography (GC) method continues to increase. Today’s laboratories are looking for ways to shorten analytical time in order to increase sample throughput and reduce analysis costs, without compromising results. Using standard runtime GC-MS analysis of 18 up to 45 min [1, 2, 3, 4, 5, 6] or even up to 61 min [7] is usual. Regarding sensitivity, reaching LOQ values of 10 μg kg−1 on GC-MS/MS has been largely reported [1, 2, 3]. According to the runtime classification, the analysis time of fast GC is in minute range, very fast GC is in seconds, and ultra-fast GC is in sub-second range [8]. The challenge in the implementation of a fast chromatography-mass spectrometry analysis can be very different depending on the type of samples, the number of components to be analyzed, and the chosen analytical conditions [9]. Reducing analysis runtime without sacrificing the analytical performance can improve laboratory productivity allowing running more samples in a reduced time. This will also allow for quicker results. This approach depends on commercialization of new techniques and adequate instrumentations.

As for pesticide applications, even before the emergence of mass spectrometry, investigations yielded for the development of fast GC-ECD using narrow-bore columns and increasing the make-up gas flow [10]. A fast GC-ECD method of 3.6 min runtime method was developed for 10 multiclass pesticides in a standard mixture [11]. Fast GC analysis of 5 min has been developed and proven effective as well as by Tanaka et al. [12] for the determination of 15 pesticides in a standard mixture.

The application of fast GC-MS for the analysis of pesticide residues in water without the losses of sensitivity and analytical quality has been demonstrated to be successful with a total runtime of less than 9 min [13, 14] for a group of 15 to 17 multiclass pesticides. More recently, multiresidue analysis in water with single quadrupole analyzer with analysis time about 10 min [15, 16] was developed. However, when it comes to a large multiresidue pesticide method, and in food matrices, shortening runtime method can lead to ambiguous results. For the analysis of pesticide residues in food matrices by MS analyzers, reduction of GC-MS method for less than 20 min and up to 6 min was developed but for a limited number of pesticides, between 13 and 19 multiclass pesticides, and for a specific food category as spice and apple [17, 18, 19]. LOQs reached were in a range of 5 to 20 μg kg−1 [17, 18, 19]. Kirchner et al. [20] and Húšková et al. [21] developed a GC-MS method with a runtime less than 20 min for 61 pesticides in apple matrix. Fast GC analyses were also developed using ToF analyzers with a total runtime less than 24 min, reaching LOQ values lower than 10 μg kg−1 [22] and even up to 16 min and reaching LOQ values between 2.5 and 500 μg kg−1 for a limited number of pesticides [23].

Many parameters can be manipulated in order to have a short runtime as the use of short columns, fast oven temperature ramp rates, or higher carrier gas linear velocities [24, 25, 26, 27, 28, 29]. Changing any of these parameters is unlikely to cause a loss in chromatographic resolution. Column length influences three parameters of major concern: efficiency, retention (analysis time), and carrier gas pressure. In this term, these parameters must be optimized when modified. When changing the column, the ratio of length to internal diameter should be kept the same. Short columns (≤ 20 m) with narrow ID (≤ 0.2 mm) with a higher carrier gas liner velocity could be used. By opting to use narrow bore columns, reduced sample injection volume is required to prevent column overload. Peak broadening, tailing, overlapping, or bad peak shapes can result in higher LOQs. Signal loss can occur when increasing the flow rate [30]. Faster temperature programming is an easier and attractive way to reduce the analysis time. Another approach for a fast runtime method is the use of low-pressure gas chromatography-mass spectrometry (LP-GC-MS) [31, 32, 33]. LP-GC was applied in a study for the analysis of 243 pesticides in a short runtime method (10.25 min) [32]. LP-GC has some limitations: due to the use of short column, there is a loss in the number of theoretical plates [31]. Moreover, it is not a rigorous method; it exhibits low repeatability, when running long sequences, caused by peak diminishment and retention time shifting in matrix [32].

This work is accomplished to demonstrate the effectiveness and sensitivity of a short runtime method on the new Intuvo 9000 GC system. The new MS system allows enhanced sensitivity due to the higher efficiency electron impact source. Different runtime methods were developed by modifying the oven temperature program. The optimum method, in terms of peak shapes and chromatographic resolution, was chosen for method validation. The 12.4-min fast GC runtime method was developed and validated at 2, 10, and 50 μg kg−1 for 203 multiclass pesticides in three food matrices (tomato, apple, and orange) in terms of linearity, recoveries, limits of identification and quantitation, matrix effects, intra-day and inter-day precisions, and uncertainty.

Experimental

GC-QqQ-MS/MS systems

Analyses were performed on an Agilent Intuvo 9000 GC system equipped with an Agilent 7693 autosampler and an Agilent 7010 GC-MS/MS triple quadrupole. Table 1 shows the specifications of this system. For sensitivity comparison purposes, this new system was compared to an Agilent 7890 GC equipped with an Agilent 7693B autosampler and an Agilent 7000 series GC–MS/MS triple quadrupole system. The instrument specifications including GC parameters [1] are represented in Table 1. Table 2 shows the list of compounds covered in the study with the two transitions for each analyte, the collision energy and the retention time.
Table 1

GC-MS/MS systems with corresponding specific parameters

GC-MS/MS system specifications

System 1

System 2

GC

Agilent Intuvo 9000 GC

Agilent 7890 GC

Autosampler

Agilent 7693 autosampler

Agilent 7693 B autosampler

MS

Agilent 7010 GC-MS/MS

Agilent 7000 GC-MS/MS

Column

Planar column HP-5MS UI (15 m long × 0.25 mm i.d. × 0.25 μm film thickness)

HP-5MS UI (15 m long × 0.25 mm i.d. × 0.25 m film thickness)

Pre-column

Guard ship

No pre-column

Flow/pressure

Constant flow = 1.611 mL/min column 1, 1811 mL/min column 2

Constant pressure = 14.1 psi

Carrier gas

Helium

Helium

Collision gas

Nitrogen

Nitrogen

Injection mode

Splitless

Splitless

Injection volume

2 μL

2 μL

Liner

Ultra-inert inlet liner with a glass wool frit from Agilent

Ultra-inert inlet liner with a glass wool frit from Agilent

Inlet temperature

80 °C hold for 0.1 min, then up to 300 °C at 600 °C/min and up to 250 at 100 °C/min

80 °C hold for 0.1 min, then up to 300 °C at 600 °C/min

Oven

Contact heating

Air bath oven

Oven temperature program

60 °C for 0.5 min, up to 170 °C at 40 °C/min, and up to 310 °C at 10 °C/min

70 °C for 1 min, up to 150 °C at 50 °C/min, then up to 200 °C at 6 °C/min and finally up to 280 °C at 16 °C/min, and then maintained for 4.07 min

Transfer line temperature

280 °C

280 °C

Ion source temperature

280 °C

280 °C

Ion source

EI, High efficient ion source (HES)

EI

Connections

Flow ships, gaskets

Ferrules

Total runtime

19.3 min

20 min

Acquisition method

2 transitions per analyte, dynamic MRM (time window 0.2 min)

2 transitions per analyte, time segment MRM

Table 2

List of compounds analyzed with their corresponding acquisition parameters (retention times, transitions, and collision energies)

 

Compound

Retention time (min)

SRM 1

CE1 (V)

SRM 2

CE2 (V)

1

2,4′-DDE

7.083

246 > 211

20

246 > 176

30

2

2-Phenylphenol

4.412

170 > 141

30

170 > 115

40

3

4,4′-DDD

7.758

235 > 199

15

235 > 165

20

4

4,4′-DDE

7.364

246 > 211

20

246 > 176

30

5

4,4′-DDT

8.110

235 > 199

20

235 > 165

20

6

Acrinathrin

9.076

289 > 93

5

208 > 181

5

7

Alachlor

6.103

188 > 160

10

188 > 130

40

8

Aldrin

6.506

293 > 257

8

293 > 186

40

9

Ametryn

6.090

227 > 212

8

227 > 185

5

10

Atrazine

5.349

215 > 173

5

215 > 58

10

11

Azoxystrobin

11.449

344 > 329

10

344 > 156

40

12

Benalaxyl

8.021

204 > 176

2

148 > 105

20

13

Bifenox

8.707

311 > 279

14

311 > 216

25

14

Bifenthrin

8.527

181 > 166

10

181 > 115

50

15

Biphenyl

3.894

154 > 126

40

154 > 102

40

16

Bixafen

10.301

413 > 159

12

159 > 139

15

17

Boscalid

10.069

140 > 112

10

140 > 76

25

18

Bromopropylate

8.564

341 > 185

20

341 > 155

20

19

Bupirimate

7.441

273 > 193

5

273 > 108

15

20

Buprofezin

7.440

305 > 172

5

119 > 91

5

21

Butralin

6.613

266 > 190

12

266 > 174

20

22

Butylate

4.013

174 > 146

3

156 > 57

5

23

Cadusafos

5.090

213 > 73

10

158 > 97

15

24

Carbofuran

5.319

164 > 149

12

164 > 122

12

25

Carbophenothion

7.997

342 > 157

10

199 > 143

10

26

Chinomethionate

7.079

234 > 206

10

206 > 148

15

27

Chlorbromuron

3.940

233 > 205

12

233 > 124

25

28

Chlordane

7.160

373 > 301

10

373 > 266

20

29

Chlorfenapyr

7.563

247 > 227

15

247 > 200

25

30

Chlorfenvinphos

6.848

295 > 267

5

267 > 81

40

31

Chlorobenzilate

7.649

139 > 111

15

139 > 75

30

32

Chlorothalonil

5.755

266 > 231

20

266 > 133

40

33

Chlorpropham

4.902

213 > 171

5

213 > 127

5

34

Chlorpyrifos

6.474

314 > 286

5

314 > 258

5

35

Chlorpyrifos-methyl

6.045

288 > 93

26

286 > 271

26

36

Chlorthal-dimethyl

6.535

330 > 299

12

330 > 221

12

37

Chlozolinate

6.791

331 > 216

5

259 > 188

5

38

Coumaphos

9.606

362 > 109

15

210 > 182

15

39

Cyfluthrin

9.813

226 > 206

10

263 > 127

10

40

Cypermethrin

10.022

209 > 141

20

163 > 127

20

41

Cyproconazole

7.595

222 > 125

18

139 > 111

18

42

Cyprodinil

6.719

224 > 208

20

224 > 197

20

43

Deltamethrin

11.160

253 > 172

5

253 > 93

5

44

Desmethyl-pirimicarb

5.886

224 > 152

8

152 > 96

8

45

Diazinon

5.554

304 > 179

15

137 > 84

15

46

Dichlofluanid

6.384

224 > 123

8

167 > 124

8

47

Dichloran

5.318

206 > 176

5

206 > 148

5

48

4,4′-Dichlorobenzophenone

6.531

250 > 139

8

139 > 111

8

49

Dichlorvos

3.385

185 > 109

15

185 > 93

15

50

Diclobutrazol

7.497

270 > 201

8

270 > 159

8

51

o,p´-dicofol and p,p´-dicofol

6.531/9.197

251 > 139

15

139 > 111

15

52

Dieldrin

7.454

345 > 263

8

279 > 243

8

53

Diethofencarb

6.353

207 > 151

10

151 > 123

10

54

Dimethenamid

5.984

230 > 154

10

154 > 111

10

55

Dimethipin

5.385

124 > 76

5

118 > 58

10

56

Diphenylamine

4.832

169 > 77

35

168 > 140

40

57

Disulfoton

5.651

142 > 109

5

142 > 81

12

58

Disulfoton-sulfoxide

3.667

212 > 153

15

125 > 97

3

59

Dodemorph

6.730

154 > 97

10

154 > 82

20

60

Endosulfan sulfate

8.134

387 > 289

5

272 > 237

15

61

Endosulfan-alpha

7.211

239 > 204

15

195 > 160

5

62

Endosulfan-beta

7.745

207 > 172

15

195 > 159

10

63

Endrin

7.660

263 > 193

35

245 > 173

30

64

EPN

8.580

157 > 110

15

157 > 77

25

65

Epoxiconazole

8.389

192 > 138

10

192 > 111

35

66

Ethion

7.772

231 > 175

5

231 > 129

25

67

Ethofenprox

10.165

163 > 135

5

163 > 107

15

68

Ethofumesate

6.288

207 > 161

5

207 > 137

10

69

Ethoprophos

5.090

158 > 114

5

158 > 97

15

70

Ethoxyquin

5.300

202 > 174

15

202 > 145

30

71

Etrimfos

5.700

292 > 181

5

292 > 153

20

72

Fenamidone

8.679

268 > 180

20

238 > 103

20

73

Fenarimol

9.197

219 > 107

10

139 > 111

15

74

Fenazaquin

8.712

160 > 145

5

160 > 117

20

75

Fenbuconazole

9.828

198 > 129

5

129 > 102

15

76

Fenchlorphos

6.181

285 > 270

15

285 > 240

30

77

Fenhexamid

8.125

177 > 113

10

177 > 78

20

78

Fenitrothion

6.282

277 > 260

5

277 > 109

20

79

Fenpropathrin

8.604

265 > 210

10

181 > 152

25

80

Fenpropidin

6.219

273 > 98

3

98 > 55

12

81

Fenpropimorph

6.437

128 > 110

10

128 > 70

12

82

Fenthion

6.455

278 > 169

20

278 > 109

20

83

Fenvalerate

10.624/10.772

167 > 125

12

125 > 89

20

84

Fipronil

6.824

213 > 178

10

213 > 143

20

85

Fipronil sulfone

7.399

452 > 383

8

383 > 255

20

86

Fipronil-desulfinil

6.055

388 > 333

20

333 > 281

15

87

Flamprop-isopropyl

7.695

276 > 105

5

276 > 77

40

88

Flamprop-methyl

7.403

276 > 105

8

230 > 170

15

89

Fluacrypyrim

7.813

145 > 115

15

145 > 102

30

90

Fluazifop-p-butyl

7.512

282 > 238

20

282 > 91

15

91

Flucythrinate

10.160

199 > 157

5

157 > 107

15

92

Fludioxonil

7.308

248 > 154

25

248 > 127

30

93

Fluopicolide

8.115

209 > 182

20

173 > 109

25

94

Fluopyram

6.821

223 > 196

15

173 > 145

15

95

Fluquinconazole

9.608

340 > 298

20

340 > 286

30

96

Flusilazole

7.435

233 > 165

20

233 > 152

20

97

Flutolanil

7.217

323 > 281

5

323 > 173

15

98

Flutriafol

7.212

219 > 123

12

219 > 95

20

99

Fluvalinate-tau

10.750

250 > 200

20

250 > 55

15

100

Fonofos

5.566

246 > 137

5

137 > 109

5

101

Formothion

5.841

224 > 125

20

170 > 93

5

102

Fosthiazate

6.670

195 > 139

5

195 > 103

5

103

HCB

5.315

284 > 249

25

284 > 214

40

104

HCH-alpha

5.235

219 > 183

5

219 > 145

25

105

HCH-beta

5.449

219 > 183

5

219 > 145

25

106

Heptachlor

6.174

272 > 237

10

272 > 143

40

107

Heptachlor endo-epoxide

6.904

183 > 155

15

183 > 119

30

108

Heptachlor exo-epoxide

6.865

217 > 182

22

183 > 119

25

109

Heptenophos

4.614

215 > 200

10

124 > 89

15

110

Hexaconazole

7.281

214 > 172

20

214 > 159

20

111

Indoxacarb

11.105

264 > 148

25

203 > 134

10

112

Iprodione

8.140

314 > 245

10

314 > 56

20

113

Iprovalicarb

7.380

158 > 116

5

158 > 98

10

114

Isazofos

5.696

257 > 162

5

161 > 119

5

115

Isocarbophos

6.541

230 > 212

8

136 > 108

8

116

Isofenphos

6.842

213 > 185

3

213 > 121

3

117

Isofenphos-methyl

6.709

199 > 167

10

199 > 121

10

118

Isoprothiolane

7.290

162 > 134

5

162 > 85

5

119

Isopyrazam

9.303

359 > 303

8

159 > 139

8

120

Kresoxim-methyl

7.432

206 > 131

10

206 > 116

10

121

Lambda-Cyhalothrin

9.023

197 > 161

5

197 > 141

5

122

Lindane

5.520

219 > 183

5

219 > 145

5

123

Malaoxon

5.986

195 > 125

15

127 > 99

15

124

Malathion

6.343

173 > 99

15

158 > 125

15

125

Mecarbam

6.837

329 > 160

3

131 > 74

3

126

Mepanipyrim

7.120

222 > 207

30

222 > 158

30

127

Merphos

7.337

169 > 113

3

169 > 57

3

128

Metalaxyl

6.141

206 > 162

8

206 > 132

20

129

Metazachlor

6.789

209 > 133

10

133 > 117

25

130

Metconazole

8.744

125 > 99

20

125 > 89

20

131

Methidathion

7.042

145 > 85

5

145 > 58

15

132

Methiocarb

6.273

168 > 153

10

153 > 109

10

133

o,p´-methoxychlor and p,p´-methoxychlor

8.213/8.610

227 > 169

25

227 > 115

40

134

Metolachlor

6.452

238 > 162

8

162 > 133

10

135

Mevinphos

3.988

127 > 109

10

127 > 95

15

136

Molinate

4.492

187 > 126

3

126 > 55

12

137

Myclobutanil

7.415

179 > 152

5

179 > 125

10

138

Napropamide

7.257

271 > 128

3

128 > 72

3

139

Nuarimol

8.253

235 > 139

12

203 > 107

10

140

Ofurace

7.989

232 > 186

5

232 > 158

20

141

Oxadixyl

7.795

163 > 132

15

163 > 117

25

142

Paclobutrazol

7.099

236 > 167

20

236 > 125

10

143

Paraoxon-methyl

5.645

230 > 200

5

109 > 79

5

144

Parathion

6.485

291 > 109

10

139 > 109

10

145

Parathion-methyl

6.046

263 > 109

10

233 > 124

10

146

Penconazole

6.798

248 > 192

15

248 > 157

25

147

Pendimethalin

6.775

252 > 191

10

252 > 162

10

148

Pentachloroaniline

5.945

263 > 227

15

263 > 192

25

149

Permethrin

9.478

183 > 153

15

163 > 127

5

150

Phenothrin

8.750

183 > 153

15

123 > 81

8

151

Phenthoate

6.887

274 > 246

5

274 > 121

10

152

Phorate

5.120

231 > 175

20

231 > 129

20

153

Phorate sulfone

6.440

199 < 143

8

153 > 97

10

154

Phosmet

8.573

160 > 133

15

160 > 77

30

155

Phthalimide

4.161

147 > 103

5

147 > 76

30

156

Picolinafen

8.560

376 > 238

25

238 > 145

25

157

Picoxystrobin

7.156

335 > 173

10

303 > 157

15

158

Pirimicarb

5.804

238 > 166

10

166 > 96

20

159

Pirimiphos-methyl

6.267

305 > 180

5

290 > 151

15

160

Procymidone

6.951

283 > 255

8

283 > 96

8

161

Profenofos

7.320

337 > 309

5

337 > 267

15

162

Prometon

5.290

225 > 183

3

225 > 168

10

163

Prometryn

6.113

241 > 226

8

241 > 184

12

164

Propaphos

6.997

220 > 140

12

220 > 125

25

165

Propazine

5.377

229 > 187

3

214 > 172

8

166

Propiconazole

8.080

259 > 191

8

259 > 173

10

167

Propyzamide

5.507

173 > 145

15

173 > 109

30

168

Prosulfocarb

6.177

251 > 128

5

128 > 86

3

169

Prothiofos

7.295

309 > 239

15

309 > 221

25

170

Pyraclostrobin

10.666

164 > 132

10

132 > 77

20

171

Pyrazophos

9.175

221 > 193

10

221 > 149

15

172

Pyridaben

9.553

147 > 132

10

147 > 117

20

173

Pyrifenox

6.820/7.068

262 > 227

10

262 > 200

20

174

Pyrimethanil

5.575

198 > 156

25

198 > 118

25

175

Pyriproxyfen

8.913

136 > 96

10

136 > 78

20

176

Quinalphos

6.890

157 > 129

15

146 > 91

30

177

Quinoxyfen

8.056

307 > 272

5

307 > 237

25

178

Quintozene

5.558

295 > 265

10

295 > 237

15

179

Secbumeton

5.643

225 > 196

5

225 > 169

5

180

Spirodiclofen

9.496

312 > 259

10

312 > 109

20

181

Spiromesifen

8.451

272 > 254

3

272 > 209

12

182

Sulfotep

5.053

238 > 146

10

202 > 146

10

183

Sulprofos

7.894

322 > 156

10

156 > 141

15

184

Tebuconazole

8.233

250 > 153

12

250 > 125

20

185

Tebufenpyrad

8.638

333 > 276

5

333 > 171

20

186

Tecnazene

4.796

215 > 179

10

203 > 143

20

187

Tefluthrin

5.617

177 > 137

15

177 > 127

15

188

Terbufos

5.505

231 > 175

10

231 > 129

25

189

Terbumeton

5.391

225 > 169

3

169 > 154

5

190

Terbutryn

6.244

241 > 185

3

241 > 170

10

191

Tetrachlorvinphos

7.108

329 > 109

25

329 > 79

35

192

Tetraconazole

6.517

336 > 218

30

336 > 204

30

193

Tetradifon

8.831

356 > 229

10

356 > 159

10

194

Tetrahydrophthalimide

4.237

151 > 122

8

151 > 80

5

195

Tetramethrin

8.500

164 > 107

15

164 > 77

30

196

Thiobencarb

6.389

125 > 89

15

100 > 72

3

197

Tolclofos-methyl

6.096

265 > 250

15

265 > 220

25

198

Tolylfluanid

6.844

238 > 137

10

137 > 91

20

199

Triadimefon

6.503

208 > 181

5

208 > 127

15

200

Triazophos

7.887

161 > 134

5

161 > 106

10

201

Trifloxystrobin

8.003

222 > 190

3

222 > 130

15

202

Trifluralin

4.966

306 > 264

10

264 > 160

15

203

Vinclozolin

6.020

212 > 172

15

212 > 109

40

Comparison of the sensitivity of conventional and improved GC system

System evaluation was made by comparing the new GC-QqQ-MS/MS to the conventional system. A calibration curve prepared in cucumber matrix and including 203 pesticide standards has been injected at eight calibration levels (1, 2, 5, 10, 100, 200, 500, and 1000 μg kg−1) on both GC-MS systems using the same conditions.

Shortening runtime method

Classic conventional ovens used in most GC systems first heat the air inside the oven, then transfer the heat to the GC column. The new GC with a planar column design enables efficient direct contact conduction heating. Direct heating enables faster analysis. This system uses less than half of the power of a conventional convection air bath oven and can be heated and cooled off much faster, improving analysis throughput.

Once the instrumental capabilities of the Intuvo 9000 GC system had been evaluated, further work was conducted. In this study, column, flow, and carrier gas velocity were kept constant. The same columns were used to develop shorter runtime methods. Only oven temperature program was modified. In this case, with different runtime methods, peaks are expected to relatively shift from each other, maintaining the same retention time’s order. An optimal temperature program for fast runtime method is the one that renders the best separation in the least time.

The starting point of the development of different runtime GC methods was a 19.3-min standard GC method. The parameters optimized in the oven temperature program are the initial hold time, ramp rates, final temperature, and hold time. To create a method of 17.6 min, initial temperature of 60 °C was kept the same, but holding time was reduced from 1 °C/min for 1 min to 0.5 °C/min for 0.5 min. An increase in the first oven temperature ramp from 40 to 70 °C reduced the runtime from 19.3 to 17.6 min. The run could be made even faster by doubling both temperature program ramps, from 40 °C/min up to 170 °C and from 10 °C/min to 310 °C to 80 °C/min and 20 °C/min, respectively, to develop a 12.4-min method based on the 19.3-min runtime method. In these cases, peak elution order is maintained the same. The method with a runtime of 15.3 min was built by decreasing ramps temperature of the 12.4-min runtime method to 70 and 13 °C/min, respectively. The 9.4-min method runtime was based on the 12.4-min method as well as on slightly increasing the ramps temperature to 90 and 30 °C/min, respectively. Figure 1 shows temperature program for each runtime method as well as the distribution of concurrent MRMs regarding time, with a total of 428 transitions, 2 for each compound. In a time segment of 1 min, e.g., from 5.0 to 5.9 min, the number of transition set was 152, 70, 30, 18, and 18, respectively, with the 9.3-, 12.4-, 15.3-, 17.6-, and 19.3-min runtime method. Retention times were set for 203 pesticides with the five discussed methods.
Fig. 1

GC temperature program and the distribution of concurrent MRMs with the five different runtime methods (Intuvo 9000 GC and 7010 MS/MS QQQ)

Runtime method selection

During the development of these methods, several factors were taken into consideration such as efficiency, retention, selectivity, and chromatographic resolution. The challenge is in developing a fast method without affecting the resolution. Optimizing chromatographic method requires compromising between speed, selectivity, and sensitivity. The detector must be able to obtain sufficient data points per peak to ensure proper peak quantitation. For the runtime methods of 12.4, 15.3, 17.6, and 19.3 min, the number of cycles per second was set to 5 data points for 1 s per peak. As for 9.4 min, data point per second was set to the maximum possible value which is 4 data points for 1 s per peak.

Once the chromatographic parameters were optimized, and to evaluate the analytical time and the separation efficiency, a mix of 203 compounds was injected at 2 μg kg−1 in five different matrices: tomato, apple, green pepper, leek, and orange. The choice of matrices was made via their representative categories and difficulty of the matrix. Tomato, apple, leek, and pepper are commodities of high water content. Tomato is considered an easy matrix [34]. Pepper and leek are considered difficult matrices due to the high amount of co-extractives remaining in the final extracts. Orange contains high water content and high acid content [35]. Fifty microliters of matrix extract previously prepared by the conventional citrate QuEChERS [36] without cleanup was evaporated and reconstituted with 50 μL of 2 μg kg−1 mix of pesticide standards prepared in ethyl acetate. Every sample was injected three times with the five methods in order to study the sensitivity and repeatability of each analytical method with the different matrices.

Method validation

The optimal runtime method was chosen and validated. The validation of the method was performed according to the EU quality control procedures [35]. The analytical parameters evaluated are selectivity, sensitivity, linearity, recovery, method repeatability, and matrix effect. Intra- and inter-day precision as well as uncertainty were also estimated.

Precision of the instrument was evaluated with the 12.4-min runtime method. Two concentration levels were studied, 2 and 5 μg kg−1, in tomato, apple, and orange. Five replicate injections were carried out for each sample. The linearity of the method was evaluated by establishing three matrix-matched calibration curves with three different matrices of tomato, apple, and orange. Seven calibration levels of 1, 2, 5, 10, 50, 100, and 200 μg kg−1 were prepared by spiking the corresponding blanks extracts of tomato, apple, and orange, previously prepared by citrate QuEChERS extraction. The limit of identification for all compounds was also studied by checking the lowest calibration level of 1 μg kg−1, with correct accuracy, by checking the two transitions and the ion ratio (< 30%). To determine the effect of co-extracted matrix components on the sensitivity, matrix effect was studied by the comparison of the same calibration curve in pure solvent (ethyl acetate) to the calibration curves obtained during matrix-matched calibration [37]. Matrix effect was also assessed by the comparison of the calibration curves of apple and orange to the one of tomato.

In order to check the trueness and precision of the overall method, recoveries and repeatability of the method were studied at three calibration levels, of 2, 10, and 50 μg kg−1, by spiking blank matrix of tomato, apple, and orange. Reaching a LOQ of 2 μg kg−1 is important and challenging despite the fact that MRLs are usually set in a range of 0.01–10 mg kg−1 or lower, as for baby food [38]. Intra-day precision was studied by injecting spiked samples for five different days. Inter-day precision was studied by calculating the RSDs obtained from checking recoveries of five spiked samples in the same day. Uncertainty of the overall procedure was also estimated using the validation data.

Spiking procedure

Thirty-five grams of a homogenized blank of tomato was spiked with a mix of 203 pesticide standards at 2 μg kg−1. The sample was stirred for 30 min to ensure homogenization and was then allowed to stand for 30 min more at room temperature, prior to extraction. The portion was split in three samples of 10 g each for extraction. The spiking procedure is performed at two more concentration levels of 10 and 50 μg kg−1. These fortification levels were selected after checking the sensitivity of the system that has been proved at 2 μg kg−1. The same procedure was repeated in two other matrices, apple and orange. The subsequent steps of sample preparation were applied for each concentration level.

Sample preparation

When considering the merits of developing and validating a routine GC method, the time for sample preparation should be considered as well; otherwise, the benefits derived from speeding up the GC separation become less significant. For this purpose, citrate QuEChERS [39] was adopted and was simplified excluding the cleanup step. Accordingly, 10-g portion of sample was weighted in a 50-mL PTFE centrifuge tube. Ten milliliters of acetonitrile was added. Ten microliters of 10 mg kg−1 of a mix of three procedural standards (dichlorvos-D6, malathion-D10, and triphenyl phosphate) was added as well to verify the extraction efficiency, and the tubes were shaken in an automatic axial agitator (AGYTAX, Cirta lab. S.L., Spain) for 4 min. Afterwards, 4 g of magnesium sulfate, 1 g of sodium chloride, 1 g of trisodium citrate dihydrate, and 0.5 g of disodium hydrogen citrate sesquihydrate were added and the samples were shaken again in the automatic axial agitator for 4 min. The extract was then centrifuged at 3500 rpm for 5 min. Prior to injection into GC, 50 μL of the extract was evaporated and reconstituted with 50 μL of ethyl acetate. Of the injection standard of Lindane-D6, 50 μg kg−1 was also added to control the performance of the analytical system.

Real samples

After the validation study was completed, the method was applied for the analysis of pesticide residues in 69 real samples purchased from local market in Andalusia, Spain. Some of these samples are local; others are imported from different countries. Among those samples were tomato, apple, orange, leek, pineapple, melon, watermelon, potato, cabbage, grapes, pear, carrot, onion, okra, nectarine, eggplant, banana, broccoli, garlic, peach, lettuce, cauliflower, kiwi, plum, cucumber, celery, grapefruit, and raspberry.

Results and discussions

Comparison of the sensitivity of conventional and improved GC system

On the basis of sensitivity, results showed that the new GC system is more sensitive than the conventional GC system. All compounds showed a limit of identification of 1 μg kg−1 on the new GC while only 43% of compounds typically reached 1 μg kg−1 on the conventional GC system. This increase in sensitivity is based on a more intense electron beam source (using same voltage, 70 eV) together with a longer interaction path length allowing the increase of number of ions produced and so the sensitivity by a factor from 2 to 10 depending on the compounds.

Shortening runtime method

Optimizing runtime temperature on GC was carefully studied. At higher temperatures, compounds interact less with stationary phase, spend more time in the mobile phase, and thus are eluted faster. However, by pressing peaks all together, they may not be resolved. Oven program should result in a reasonable separation of peaks. Any decrease in runtime analysis should not affect the resolution. Decreasing runtime and maintaining resolution is an important aspect that has to be considered. The resolution describes the system’s ability to separate compounds from each other. In the first measurement, separation and detection steps were optimized. The performance of the five runtime methods and the separation were satisfactory. Careful consideration of peak separation for compounds having same transition is necessary. The only case is the one of cyproconazole and chlorobenzilate that were not well separated in a runtime method of 9.4 min. Quantitation is difficult using the common transition 139 > 111 (Fig. 2). However, other transitions can be used for the quantitation of both compounds. In general, there was no loss in resolution when shortening the method from 19.3 to 9.4 min. Two hundred three pesticides were properly separated with the five discussed GC temperature program methods.
Fig. 2

a Cyproconazole and chlorobenzilate peak separation, corresponding to the same transition 139 > 111 with 9.4- and 12.4-min runtime methods. b Peak resolution for thiobencarb from matrix interference in tomato at 19.3- and 12.4-min runtime methods

The five runtime methods provided good sensitivity with the five matrices analyzed. In total, 99% of compounds showed good sensitivity at 2 μg kg−1 with all the runtime methods. By comparing the signals (peak areas) of the 203 compounds with the different runtime methods, no loss of sensitivity was observed; the relative standard deviations of areas were less than 20% for 99% of the compounds. Regarding the high number of coinciding transitions in 1 min at shorter GC methods (e.g., from 5.0 to 5.9 min), sulfotep, one of the compounds eluted in the same time segment with the three different runtime methods, 12.4, 15.3, and 17.6 min, was checked. No difference in sensitivity was observed for this compound. Shortening runtime did not have a cost regarding sensitivity even for the most challenging compounds such as disulfoton, disulfoton-sulfoxide, dicofol, and tolylfluanid. Good precision was obtained as well in all cases (< 5%).

Pesticides showed reproducible responses with different runtime methods. Contrarily, pesticides had different responses in different matrices due to matrix effects. In general, leek showed different results due to the difficulty of this matrix. All compounds were detected at 2 μg kg−1 in tomato and apple. Ninety-nine percent of compounds were detected in orange and pepper matrices. Ninety-six percent of compounds showed good sensitivity in leek. Chlordane was not detected in orange, pepper, and leek with the five runtime methods. Diethofencarb was discriminated in orange and leek. Additionally, chlorothalonil, dichlofluanid, ethoxyquin, isoprothiolane, and mecarbam were not detected in leek. Leek is a difficult matrix as it contains large amount of pigments and sulfur-containing compounds [40]. Moreover, the instability of chlorinated compounds (dichlofulanid) and compounds that contain sulfur (isoporthiolane) and phosphorus (mecarbam) in GC explains the results obtained in this specific matrix.

Even though the number of coincident transitions per 1 min time segment did not affect peaks resolution, or sensitivity, taking this aspect into consideration is important for the choice of the adequate runtime method. The highest number of transitions (152) in 1-min time segment (5 to 6 min) is observed with the 9.4-min runtime method. The highest numbers of transitions observed with the four larger runtime methods were, respectively, 119, 93, 71, and 66 transitions per 1 min.

Regarding peak shapes, a general beneficial effect is that fast analysis produces rapid and narrow peaks. Peak shapes improved considerably at 9.4 and 12.4 min. Peaks are more narrowed and symmetric. Some compounds showed better results regarding peak shape when moving to the 9.4- and 12.4-min methods, case of napropamide and thiobencarb. These compounds resulted in high matrix interference at 19.3 min. With the 12.4-min runtime method, reason being of more narrowed peaks, peaks corresponding to these compounds were well shaped and separated from any peak corresponding to matrix interference. Figure 2 shows the case of thiobencarb.

Fast GC-MS requires an analyzer able to provide an adequate scan speed to give sufficient data points per peak. Data points per peak were checked. For compounds that eluted first such as 3-chloroaniline and atrazine, no significant change was observed regarding the number of data points per peak with the different runtime methods. A more important decrease in data point was observed at the end of analysis, where azoxystrobin, the last eluted compound, showed 16 data points per peak at 9.4 min runtime method, 38 data points per peak at 12.4 and 15.3 min runtime methods, and 42 data points per peak at 17.6 and 19.3 min runtime methods. An overview of the modification of data points per peak shows that in all the methods, the data points per peak were maintained approximately the same, except for the 9.4-min method: the reason why the 12.4-min method was preferred.

Therefore, taking into consideration, the data point per peak, number of transitions, peaks resolution, as well as the system sensitivity, the 12.4-min method was preferred over the 9.4-min method. The 12.4-min method did not affect the analytical quality and proved its reliability for quick and correct identification.

Method validation

Recoveries and repeatability

The 12.4-min method was validated. Recovery experiments were established in order to evaluate the trueness of the method. Generated data showed that 97% of compounds had good recovery rates, in the acceptance range, between 70 and 120%, at the concentration levels of 2, 5, and 10 μg kg−1 in tomato, apple, and orange. Recovery and repeatability values obtained are presented in Fig. 3.
Fig. 3

Average recovery values and RSDs obtained with the 12.4-min runtime method with tomato, apple, and orange

The analytes that were demonstrated to be non-sensitive at low concentration, which are mainly acid-sensitive analytes prone to hydrolyzation in acidic matrices as orange, showed low recoveries. Citrate QuEChERS allows adjusting the pH. However, in orange, the pH is acid, and with the elimination of the cleanup step with PSA, some basic analytes have been lost. On the other hand, by avoiding the cleanup step, all acidic analytes were dissolved in extractive solvent and recoveries with acceptable range were obtained. Some compounds were lost in the evaporation step in sample preparation. The problems experienced for these analytes are discussed below.

In general, same behavior with different concentration levels was observed for most of the compounds. Some compounds were demonstrated to have low sensitivity at 2 μg kg−1 in the above sensitivity study, case of diethofencarb (pK a = 12.7, strong base) that is as well prone to hydrolyzation in acid matrices as orange and chlordane that is only recovered at high concentration levels.

As for ethoxyquin, only satisfactory recoveries were obtained with orange. The low pH in orange was suitable for the stability of ethoxyquin which is a weak acid (pK a = 4.56) and is only recovered at low pH [41]. The discrimination of ethoxyquin in apple and tomato can be explained as follows: ethoxyquin is a quinoline-based antioxidant and its decomposition is faster in extracts of commodities with poor antioxidative potential [42]. In apple, chlozolinate showed low recoveries at 2 μg kg−1. Pyrifenox (pK a = 4.61, weak acid) showed low recoveries in apple at 10 and 50 μg kg−1. It could be prone to hydrolyzation. At 2 and 10 μg kg−1, HCB showed low recoveries in tomato. HCB is very stable in acid and base matrices but can be not well extracted with acetonitrile (log P = 3.93) [43].

Low recovery rates were observed for biphenyl and butylate in tomato and apple at the three concentration levels of 2, 10, and 50 μg kg−1. There is a potential that the loss of these compounds occurred in the evaporation step. Biphenyl and butylate have high volatility, 1238 and 1730 mPa at 25 °C, respectively. Tecnazene as well showed low recovery rates in tomato at 2 and 10 μg kg−1 due also to a probable loss during evaporation (25 mPa at 25 °C) [44].

Repeatability experiments were established to evaluate the method precision. Satisfactory repeatability results (RSDs < 10%) were obtained for all compounds at different concentration levels and in all matrices.

Limits of identification and limits of quantitation

All compounds could be identified at 1 μg kg−1 in cucumber. Identification is demonstrated with two transitions, good peak shapes and right accuracy values. This instrumental limit was as well checked when studying the linearity of the method with the concerned matrix studied.

According to EU analytical quality control procedures, LOQ is the lowest concentration tested for which recovery and repeatability values were satisfactory [35]. Therefore, the LOQ for 97% of the compounds was 2 μg kg−1, the lowest validated level with acceptable trueness and precision.

Linearity

Good linearity was achieved in all cases with residuals lower than 20% and correlation coefficients R 2 higher than 0.99. All compounds were linear up to a concentration level of 200 μg kg−1, the highest concentration level examined. However, linearity concentration ranges were different for some pesticides. For 99% of compounds analyzed in tomato, the linearity range was between 1 and 200 μg kg−1. Phenothrin and buprofezin had a narrow range of linearity between 2 and 200 μg kg−1. Propaphos had a linear range of 1 to 100 μg kg−1. As for linearity study with apple, 98% compounds had a linearity range between 1 and 200 μg kg−1. Quinalphos, metconazole, fipronil, and fenhexamid showed a linear range from 2 to 200 μg kg−1 because of low accuracy at 1 μg kg−1. Chlordane showed linearity results between 5 and 200 μg kg−1. The image is quite different for orange due to the difficulty of the concerned acidic matrix. Ninety-four percent of compounds showed good linearity between 1 and 200 μg kg−1. Sulfotep, paraoxon methyl, merphos, mercabam, isazofos, chlordane, disulfoton, diethofencarb, malaoxon, quinalphos, and secbumeton showed narrowed linear range due to compound sensitivity. Ethoxyquin was the only compound that showed a linear range between 1 and 10 μg kg−1.

Matrix effect

Matrix effect is defined by the effect of co-eluting residual of matrix components resulting in either signal suppression or enhancement. All compounds showed signal enhancement on GC except for fenpropidin and chlorobenzilate that showed signal suppression in all matrices, but the strongest effect (< −50%) was observed in orange. This is due to the fact that the matrix component of orange is an H+ donor, which can explain the signal enhancement. In general, significant matrix effect expressed as signal enhancement is observed on GC [45, 46, 47]; however, this aspect is resolved by building matrix-matched calibration for quantitation purposes [35]. Consequently, the assessment of matrix effect of apple and orange was studied in reliance to the easiest matrix, tomato. In apple, no matrix effect was observed for 91% of the compounds. In orange, no matrix effect was observed for 85% of the compounds (see Electronic Supplementary Material (ESM) Table S1). Therefore, it is possible to quantify the majority of pesticides in different matrices by building one matrix-matched calibration curve.

Inter- and intra-day precision

For inter-day precision (5 days), 97% of the compounds studied showed satisfactory results (RSD < 20%). Some compounds showed higher RSD values as biphenyl and butylate. These compounds are discussed above; they are susceptible of loss during the evaporation step, what explains the high RSDs obtained.

For intra-day precision, 97% of pesticides in all matrices showed a RSD below 20% except for biphenyl, butylate, chlozolinate, and pyrifenox. Chlozolinate and pyrifenox showed different behavior with apple. The intra-day precision study includes different matrices, which explains the high RSDs obtained for these two compounds.

Uncertainty

Uncertainty was evaluated for the analytical methodology developed. Recoveries and repeatability obtained from spiked tomatoes at 10 μg kg−1 with a mix of 203 pesticides and for 5 consecutive days were integrated in the calculation.

Expanded uncertainty was calculated with the following formula: \( {U}^{\hbox{'}}=2{u}^{\hbox{'}}=2\ \sqrt{\left({u}^{\hbox{'}}{\mathrm{Rw}}^2\right)+{u}^{\hbox{'}}{\mathrm{bias}}^2} \)where \( {u}^{\hbox{'}}\mathrm{bias}=\sqrt{\frac{{\mathrm{RMS}}^{\hbox{'}}{\mathrm{bias}}^2}{n}+{u}^{\hbox{'}}C{\mathrm{ref}}^2} \), RMS′bias is the root mean square of bias, n is the number of days,\( {u}^{\hbox{'}}C\mathrm{ref}=\frac{100-\mathrm{standard}\ \mathrm{purity}}{2} \), and u′Rw is the uncertainty of recovery values associated with the whole procedure over 5 days.

Uncertainties associated with purity of pesticide standards did not exceed 3% for 96% of the compounds. Uncertainty of bias did not exceed 20% for 92% of the compounds. Recovery uncertainty values were as well less than 20% for 92% of the compounds. Expanded uncertainty values obtained were lower than 50% for all compounds except for the problematic compounds discussed previously. Figure 4 shows the RSDs obtained with inter-day precision, intra-day precisions, and uncertainty.
Fig. 4

RSD values generated from inter-day precision, intra-day precision, and expanded uncertainty. Note: The default value of uncertainty is 50% and the criteria of precision is ≤ 20% according to SANTE document

Real samples

After the validation study was completed and in order to prove the effectiveness of the 12.4-min runtime method and its suitability in routine analysis, it was applied first to EUPT-FV18 (spinach) and EUPT-FV19 (lemon) samples 48]. Pesticides included in the EUPT-FV18 and suitable for GC analysis were fluopyram, indoxacarb, and metalaxyl. Pesticides included in the EUPT-FV18 were fluopyram, boscalid, chlorpyrifos, diazinon, fipronil, iprodione, and pyraclostrobin. Z scores have been calculated for each compound. Z score is a parameter assessment of laboratory’s performance. Z score is calculated from the assigned value of an analyte and the fit-for-purpose relative standard deviation. The results were acceptable (ǀzǀ ≤ 2). Z scores obtained were between − 0.9 and 0.1 for EUPT-FV18 and between − 1.49 and 0.84 for EUPT-FV19.

The validated method was later on applied to 69 real samples. No pesticide residues were observed in only 12% of the samples (eight samples). In 23% of the samples (16 samples), only one detected pesticide was observed. Thirty-two percent of samples showed two to three detections. Twelve percent of samples showed five to six pesticide detections. In 6% of samples, higher number of pesticides was detected, up to 12 detections in 1 sample, case of plum, pear, nectarine, and peach. These latter are the most contaminated samples considering the number of pesticide detected in each sample.

The total number of pesticides identified was 58. Boscalid, myclobutanil, and deltamethrin showed high detection frequency. They were identified in 21, 19, and 19% of the samples, respectively. Cypermethrin, chlorpyrifos, and azoxystrobin were detected in 14% of the samples. Other pesticides are detected in 7 to 12% of samples: chlorpyrifos-methyl, fludioxonil, fluopyram, iprodione, lambda-cyalothrin, metalaxy, etofenprox, pyriproxyfen, phthalimide, pyraclostrobin, pyrimethanil, tetrahydrophtalimide, tetramethrin, and tebuconazole. Some pesticides are only detected in one, two, or three samples: 2-phenylphenol, 4,4′-DDE, benalaxyl, bupirimate, carbofuran, chlorothalonil, chlorpropham, cyproconazole, cyprodinil, diazinon, diphenylamine, fenazaquin, fenbuconazole, fenhexamid, fenpropidin, fenpropimorph, fipronil sulfone, fluopicolide, flutolanil, flutriafol, fluvalinate-tau, indoxacarb, malathion, oxadixyl, paclobutrazol, paraoxon-methyl, penconazole, pendimethalin, permethrin, phenothrin, phenthoate, pirimicarb, prometon, propiconazole, spiromesifen, tetraconazole, and trifloxystrobin. This large number of pesticides reported is due to the detection of pesticides at concentrations levels below 10 μg kg−1.

In seven cases (4% of total detections), the measured residue concentrations, with corresponding uncertainty taken into consideration, exceeded the MRLs (6% of samples). The pesticides found violating EU MRLs are carbofuran, diphenylamine, phthalimide, chlorpyrifos, and chlorpyrifos-methyl. Carbofuran was detected in two pear samples at concentrations of 0.062 and 0.032 mg kg−1 (MRL = 0.001 mg kg−1). Diphenylamine was detected in one pear sample at a concentration of 1.23 mg kg−1 (MRL = 0.1 mg kg−1). Phthalimide was detected as well in two pear samples at concentrations of 0.19 and 0.54 mg kg−1 (MRL = 0.03 mg kg−1). Chlorpyrifos was detected at a concentration of 1.31 mg kg−1 in asparagus, exceeding the MRL fixed by the EU of 0.05 mg kg−1. Chlorpyrifos-methyl was detected in one pear sample at 0.14 mg kg−1, exceeding the MRL of 0.03 mg kg−1.

Figure 5 shows the percentage of samples with no pesticide residues (12%), those with detections higher than 10 μg kg−1 (52%), and the percentage of samples with detections between 2 and 10 μg kg−1 (36%). These data show that lower concentration reporting limits can increase the number of detection rate. Therefore, reaching lower concentrations can be a valuable tool for monitoring of specific compounds or production types (e.g., organic production) and for risk assessment.
Fig. 5

Pesticide residue detections in 69 real samples of fruits and vegetables

Conclusions

The implementation of a fast GC program temperature allows the reduction of the total GC analysis time by a factor of 1.5 without compromising the quality of results and the sensitivity of the method. The main benefit of a fast GC method is the increase of laboratory throughput. The minor modifications to the method allowed reducing analysis time without the loss of chromatographic resolution. Careful consideration of peak resolution and method validation parameters allowed the development of a fast method without compromising the results. Good quality results include separation capabilities and satisfactory method validation parameters (recoveries, repeatability, linearity, matrix effect). In this term, peaks are properly separated. An improvement in the peak shapes was observed due to narrow peak widths. The validated method is demonstrated to be effective and also sensitive as it showed good recovery rates and reproducibility at 2 μg kg−1 in most of the cases. The proper quantitation of the method was evaluated by the analysis of 2 EUPT-FV samples and 69 real samples. It is noticeable that the increasing rate of positive detections when lowering the concentration limits from 10 to 2 μg kg−1 is around 36%.

Notes

Acknowledgments

The authors acknowledge funding support from the European Commission, DG SANTE (Grant decision SI2.726352) and thank Joerg Riener from Agilent Technologies for his technical support.

Compliance with ethical standards

Conflict of interest

The authors declare no conflict of interest with any of the instruments or materials referred to in this work.

Supplementary material

216_2017_723_MOESM1_ESM.pdf (158 kb)
ESM 1 (PDF 158 kb)

References

  1. 1.
    Uclés S, et al. Validation of a multiclass multiresidue method and monitoring results for 210 pesticides in fruits and vegetables by gas chromatography-triple quadrupole mass spectrometry. J Environ Sci Health. 2014;49(8):557–68.CrossRefGoogle Scholar
  2. 2.
    Lee J, et al. Rapid and simultaneous analysis of 360 pesticides in brown rice, spinach, orange, and potato using micropore GC-MS/MS. J Agric Food Chem. 2017;65(16):3387–95.CrossRefPubMedGoogle Scholar
  3. 3.
    Walorczyk S, Kopeć I, Szpyrka E. Pesticide residue determination by gas chromatography-tandem mass spectrometry as applied to food safety assessment on the example of some fruiting vegetables. Food Anal Methods. 2016;9(5):1155–72.CrossRefGoogle Scholar
  4. 4.
    Machado I, et al. Determination of pesticide residues in globe artichoke leaves and fruits by GC–MS and LC–MS/MS using the same QuEChERS procedure. Food Chem. 2017;227:227–36.CrossRefPubMedGoogle Scholar
  5. 5.
    Sawant D, et al. Fast gas chromatography with tandem mass spectrometry analysis of selected persistent organic pollutants and organophosphorus and synthetic pyrethroid pesticides in Indian prawn (Fenneropenaeus indicus) in compliance with the EU-MRLs. J AOAC Int. 2017;100(3):610–5.CrossRefPubMedGoogle Scholar
  6. 6.
    Silva AS, et al. A multiclass analytical method for pesticides determination in water using DLLME and GC-MS. Rev Virtual Quim. 2017;9(2):550–62.Google Scholar
  7. 7.
    Rai S, et al. A rapid method for the quantitative determination of 34 pesticides in nonalcoholic carbonated beverages using liquid-liquid extraction coupled to dispersive solid-phase cleanup followed by gas chromatography with tandem mass spectrometry. J AOAC Int. 2017;100(3):624–30.CrossRefPubMedGoogle Scholar
  8. 8.
    Van Deursen MM, et al. Evaluation of time-of-flight mass spectrometric detection for fast gas chromatography. J Chromatogr A. 2000;878(2):205–13.CrossRefGoogle Scholar
  9. 9.
    Matisová E, Hrouzková S. Analysis of endocrine disrupting pesticides by capillary GC with mass spectrometric detection. Int J Environ Res Public Health. 2012;9(9):3166–96.CrossRefPubMedPubMedCentralGoogle Scholar
  10. 10.
    van Ysacker PG, et al. Electron capture detection in high speed narrow bore capillary gas chromatography: fast and sensitive analysis of PCBs and pesticides. J High Resolut Chromatgr. 1995;18(7):397–402.CrossRefGoogle Scholar
  11. 11.
    Dömötörová M, et al. Fast gas chromatography of polar and non-polar compounds with the on-column injection. Chem Anal. 2005;50(2):365–75.Google Scholar
  12. 12.
    Tanaka KM, Miyagawa H, Nakagawa K. Analysis time reduction using fast gas chromatography-mass spectrometry. J Anal Sci. 2001;17:885–6.CrossRefGoogle Scholar
  13. 13.
    Hada M, et al. Trace analysis of pesticide residues in water by high-speed narrow-bore capillary gas chromatography-mass spectrometry with programmable temperature vaporizer. J Chromatogr A. 2000;874(1):81–90.CrossRefPubMedGoogle Scholar
  14. 14.
    Korenková E, Matisová E, Slobodník J. Study on the feasibility of coupling large-volume injection to fast gas chromatography with mass spectrometric detection for analysis of organochlorine pesticides. J Sep Sci. 2003;26(12–13):1193–7.CrossRefGoogle Scholar
  15. 15.
    Purdešová A, et al. Evaluation of calibration approaches for quantification of pesticide residues in surface water by SPE with small-size cartridges followed by fast GC-MS. J Anal Methods. 2013;5(13):3403–9.CrossRefGoogle Scholar
  16. 16.
    Cherta L, et al. Multiclass determination of 66 organic micropollutants in environmental water samples by fast gas chromatography-mass spectrometry. ABC. 2012;402(7):2301–14.Google Scholar
  17. 17.
    Kochman M, et al. Fast, high-sensitivity, multipesticide analysis of complex mixtures with supersonic gas chromatography-mass spectrometry. J Chromatogr A. 2002;974(1–2):185–212.CrossRefPubMedGoogle Scholar
  18. 18.
    Hercegová A, et al. Fast gas chromatography with solid phase extraction clean-up for ultratrace analysis of pesticide residues in baby food. J Chromatogr A. 2005;1084(1–2):46–53.CrossRefPubMedGoogle Scholar
  19. 19.
    Hercegová A, et al. Study on pesticide residues in apples, apple-based baby food, and their behaviour during processing using fast GC-MS multiresidue analysis. Int J Environ Anal Chem. 2007;87(13–14):957–69.CrossRefGoogle Scholar
  20. 20.
    Kirchner M, et al. Fast gas chromatography for pesticide residues analysis using analyte protectants. J Chromatogr A. 2008;1186(1–2):271–80.CrossRefPubMedGoogle Scholar
  21. 21.
    Húšková R, Matisová E, Kirchner M. Fast GC-MS pesticide multiresidue analysis of apples. Chromatographia. 2008;68(SUPPL. 1):S49–55.CrossRefGoogle Scholar
  22. 22.
    Čajka T, Hajšlová J. Gas chromatography-high-resolution time-of-flight mass spectrometry in pesticide residue analysis: advantages and limitations. J Chromatogr A. 2004;1058(1–2):251–61.CrossRefGoogle Scholar
  23. 23.
    Patel K, et al. Analysis of pesticide residues in lettuce by large volume-difficult matrix introduction-gas chromatography-time of flight-mass spectrometry (LV-DMI-GC-TOF-MS). Analyst. 2003;128(10):1228–31.CrossRefPubMedGoogle Scholar
  24. 24.
    Maštovská K, Lehotay SJ. Practical approaches to fast gas chromatography-mass spectrometry. J Chromatogr A. 2003;1000(1–2):153–80.CrossRefPubMedGoogle Scholar
  25. 25.
    Klee MS, Blumberg LM. Theoretical and practical aspects of fast gas chromatography and method translation. J Chromatogr Sci. 2002;40(5):234–47.CrossRefPubMedGoogle Scholar
  26. 26.
    Blumberg LM, Klee MS. Optimal heating rate in gas chromatography. JMS. 2000;12(9):508–14.Google Scholar
  27. 27.
    Vaclavik L, et al. Survey of mass spectrometry-based high-throughput methods in food analysis, in high-throughput analysis for food safety. 2014; 15–72.Google Scholar
  28. 28.
    Woolt L, Decker D. Practical fast gas chromatography for contract laboratory program pesticide analyses. J Chromatogr Sci. 2002;40(8):434–40.CrossRefGoogle Scholar
  29. 29.
    Mondello L, et al. Ultra-fast essential oil characterization by capillary GC on 50 μm ID column. J Sep Sci. 2004;27(9):699–702.CrossRefPubMedGoogle Scholar
  30. 30.
    McNair HM, Reed GL. Fast gas chromatography: the effect of fast temperature programming. JMS. 2000;12(6):351–5.Google Scholar
  31. 31.
    Martínez Vidal JL, et al. Trace determination of organotin compounds in water, sediment and mussel samples by low-pressure gas chromatography coupled to tandem mass spectrometry. RCM. 2003;17(18):2099–106.Google Scholar
  32. 32.
    Maštovská K, Hajšlová J, Lehotay SJ. Ruggedness and other performance characteristics of low-pressure gas chromatography-mass spectrometry for the fast analysis of multiple pesticide residues in food crops. J Chromatogr A. 2004;1054(1–2):335–49.PubMedGoogle Scholar
  33. 33.
    Han L, Sapozhnikova Y, Lehotay SJ. Method validation for 243 pesticides and environmental contaminants in meats and poultry by tandem mass spectrometry coupled to low-pressure gas chromatography and ultrahigh-performance liquid chromatography. Food Control. 2016;66:270–82.CrossRefGoogle Scholar
  34. 34.
    Malato O, et al. Benefits and pitfalls of the application of screening methods for the analysis of pesticide residues in fruits and vegetables. J Chromatogr A. 2011;1218(42):7615–26.CrossRefPubMedGoogle Scholar
  35. 35.
    European Commission DG-SANTE. Guidance document on analytical quality control and method validation procedures for pesticide residues analysis in food and feed, N.S. 2015.Google Scholar
  36. 36.
    European Standard. Foods of plant origin—determination of pesticide residues using GC-MS and/or LC-MS/MS following acetonitrile extraction/partitioning and clean-up by dispersive SPE-QuEChERS-method. CSN EN 15662 2008.Google Scholar
  37. 37.
    Furey A, et al. Ion suppression; a critical review on causes, evaluation, prevention and applications. Talanta. 2013;115:104–22.CrossRefPubMedGoogle Scholar
  38. 38.
    EU pesticides database available at: http://www.ec.europa.eu.
  39. 39.
    Anastassiades M, et al. Fast and easy multiresidue method employing acetonitrile extraction/partitioning and “dispersive solid-phase extraction” for the determination of pesticide residues in produce. J AOAC Int. 2003;86(2):412–31.PubMedGoogle Scholar
  40. 40.
    Shen CY, et al. Determination of 17 pyrethroid residues in troublesome matrices by gas chromatography/mass spectrometry with negative chemical ionization. Talanta. 2011;84(1):141–7.CrossRefPubMedGoogle Scholar
  41. 41.
    Uclés A, et al. Application of zirconium dioxide nanoparticle sorbent for the clean-up step in post-harvest pesticide residue analysis. Talanta. 2015;144:51–61.CrossRefPubMedGoogle Scholar
  42. 42.
    EURL-SRM, improvement of ethoxyquin recoveries by QuEChERS through the addition of ascorbic acid initiator/Institution: EURL-SRM version 1 (last update: 22 July 2014).Google Scholar
  43. 43.
    PubChem database available at: https://www.ncbi.nlm.nih.gov/pccompound.
  44. 44.
    Pesticides properties database PPDB available at: http://sitem.herts.ac.uk/aeru/ppdb/en/.
  45. 45.
    Lozowicka B, Rutkowska E, Hrynko I. Simultaneous determination of 223 pesticides in tobacco by GC with simultaneous electron capture and nitrogen-phosphorous detection and mass spectrometric confirmation. Open Chem. 2015;13(1):1137–49.CrossRefGoogle Scholar
  46. 46.
    Uclés S, et al. Matrix interference evaluation employing GC and LC coupled to triple quadrupole tandem mass spectrometry. Talanta. 2017;174:72–81.CrossRefPubMedGoogle Scholar
  47. 47.
    Lozano A, et al. Pesticide analysis in teas and chamomile by liquid chromatography and gas chromatography tandem mass spectrometry using a modified QuEChERS method: validation and pilot survey in real samples. J Chromatogr A. 2012;1268:109–22.CrossRefPubMedGoogle Scholar
  48. 48.
    Ferrer C, et al. European Union proficiency tests for pesticide residues in fruit and vegetables from 2009 to 2016: overview of the results and main achievements. Food Control. 2017;82:101–13.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  • Elena Hakme
    • 1
  • Ana Lozano
    • 1
  • Samanta Uclés
    • 1
  • Amadeo R. Fernández-Alba
    • 1
  1. 1.Agrifood Campus of International Excellence (CeiA3), European Union Reference Laboratory for Pesticide Residues in Fruit and Vegetables, Department of Chemistry and PhysicsUniversity of AlmeríaAlmeríaSpain

Personalised recommendations