Analytical and Bioanalytical Chemistry

, Volume 389, Issue 6, pp 1715–1754 | Cite as

Modification and re-validation of the ethyl acetate-based multi-residue method for pesticides in produce

  • Hans G. J. Mol
  • Astrid Rooseboom
  • Ruud van Dam
  • Marleen Roding
  • Karin Arondeus
  • Suryati Sunarto
Open Access
Original Paper

Abstract

The ethyl acetate-based multi-residue method for determination of pesticide residues in produce has been modified for gas chromatographic (GC) analysis by implementation of dispersive solid-phase extraction (using primary–secondary amine and graphitized carbon black) and large-volume (20 μL) injection. The same extract, before clean-up and after a change of solvent, was also analyzed by liquid chromatography with tandem mass spectrometry (LC–MS–MS). All aspects related to sample preparation were re-assessed with regard to ease and speed of the analysis. The principle of the extraction procedure (solvent, salt) was not changed, to avoid the possibility invalidating data acquired over past decades. The modifications were made with techniques currently commonly applied in routine laboratories, GC–MS and LC–MS–MS, in mind. The modified method enables processing (from homogenization until final extracts for both GC and LC) of 30 samples per eight hours per person. Limits of quantification (LOQs) of 0.01 mg kg−1 were achieved with both GC–MS (full-scan acquisition, 10 mg matrix equivalent injected) and LC–MS–MS (2 mg injected) for most of the pesticides. Validation data for 341 pesticides and degradation products are presented. A compilation of analytical quality-control data for pesticides routinely analyzed by GC–MS (135 compounds) and LC–MS–MS (136 compounds) in over 100 different matrices, obtained over a period of 15 months, are also presented and discussed. At the 0.05 mg kg−1 level acceptable recoveries were obtained for 93% (GC–MS) and 92% (LC–MS–MS) of pesticide–matrix combinations.

Keywords

Foods/Beverages Pesticides GC-MS LC-MS/MS Multi-residue analysis 

Introduction

For monitoring and control of pesticide residues, multi-residue methods are very cost-effective and are used in many laboratories. The pesticides are usually first extracted with an organic solvent of high or medium polarity. Typical solvents used for this purpose are acetone [1, 2, 3, 4], ethyl acetate [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26] (Table 1), and acetonitrile [26, 27, 28, 29, 30, 31]. With all three options, pesticides are partitioned between an aqueous phase and an organic phase. With acetone and acetonitrile this is done in two successive steps, with ethyl acetate in one step. With regard to extraction efficiency, ethyl acetate has been shown to be equivalent to the water-miscible solvents for both polar and non-polar pesticides in vegetables, fruit, and dry products (after addition of water) [6, 7, 26, 32]. It is also suitable for products with a high fat content—because of the solubility of fat in ethyl acetate, pesticides are released and extracted efficiently. The extract obtained is compatible with gel-permeation chromatography (GPC), the clean-up procedure most suitable for this type of sample. Ethyl acetate is very suitable for GC analysis. It has good wettability in GC (pre)columns; this is of benefit for solvent trapping of the most volatile analytes, which is required for refocusing after injection. Its vapor pressure and expansion volume during evaporation also favor large-volume injection. Finally, it is compatible with all GC detectors. The same extract can also be used for LC analysis, after a solvent change into, e.g., methanol [11, 15, 16, 17, 18, 26], as is done for acetone-based methods also [33].
Table 1

Examples from literature. Conditions typically used in ethyl acetate-based multi-residue analysis

Sample (g)

Addition

EtAc (mL)

Na2SO4 (g)

Extr.

Phase separation

Re-extr.

Evap./reconst. (aliquot/to mL)

Clean-up

Evaporation (from/to mL)

Final extr. g mL−1

Inj. (μL)

Analysis

 

50

100

50

B

 

5→1

GPC

None

0.19

10

GC–NPD/ECD

1987 [5]

75

200

40

T

F/Na2SO4

100→5

GPC

Eluate→5

1.5

?

GC–NPD, FPD

1991 [6]

(dilute)

0.3

?

GC–ECD

 

5

20

10

T

Let settle

10→1

  

2.5

1–5

GC–FPD/NPD

1992 [7]

1

5 mL H2O (wheat)

0.5

50

100

50

T

F/Na2SO4

   

0.5

2–8

GC–MS/FPD/ECD

1996 [4]

50

250

100

B

F/Na2SO4

All→100

GPC

Eluate→1

1

1

GC–NPD/ECD

1998 [8]

75

200

40

T

F/Na2SO4

100→5

SPE (ENV+)

3 mL

1.25

2

GC–ITD/ NPD/ECD

1999 [9]

20

100

 

T

See clean-up

 

Cartridge water abs. Polymer+ GCB/Na2SO4

50→dry→2 ace/hex

1

2

GC–MS, GC–NCI-MS

2001 [10]

GC–FPD LC–PCR-Flu

8

2 g NaHCO3

50

70

T

F

Yes

All→20 MeOH

  

0.4

5–10

LC–MS–MS

2002 [11]

25

100

75

T

F (vac)

All→25 +25 cyclohexane

GPC

Eluate→1

1

1

GC×GC–TOF-MS, GC–TOF-HRMS

2003 [12]

F/Na2SO4

Rinse

2004 [13]

30

5–6 g NaHCO3

60

30–40

T (30 °C)

F/cotton wool

1+0.1 IS→1

 

0.5

10

GC–TOF-MS (DMI)

2003 [14]

25

50

25

T

Let settle or centrifuge

1→1 H2O

 

0.5

20

LC–MS–MS

2003 [15]

75

NaOH if pH < 4.5

200

40

T

F/Na2SO4

100→5

5→MeOH

2.5

10

LC–MS–MS

2004 [16]

15

1 mL 6.5 mol L−1 NaOH

90

13

T

F/Na2SO4

Rinse

All→15 MeOH

 

1

10

LC–MS–MS

2004 [17]

15

1 mL 6.5 mol L−1 NaOH

90

 

T

F/Na2SO4

Yes 2×

All→15 MeOH

 

1

50

LC–TOF-MS

2005 [18]

10

50a

10

B

F

All→5

GPC

Eluate 35→2

2

10

GC–MS–MS

2006 [19]

6

50

3

B

Centr.

Yes

All→5

GPC

Eluate 84→1b

5

1

GC–MS

2006 [20]

20

80

50–100

T

F

Yes

All→ace/hex

SPE SAX/PSA

All→3

2.4

2

GC–ECD

2006 [21]

50

100

75

T

F/Na2SO4

Rinse

All→10

 

5

2

GC–NPD/MS

2006 [22]

5

10

 

T

F/Na2SO4

Rinse

All→1

 

5

10

GC–MS–MS

2006 [23]

2.5

5

2

T

F (syringe)

 

 

0.5

50

GC–NPD

2006 [24]

30

5–6 g NaHCO3

60

30–40

T (30 °C)

F

1→0.9 +0.1 IS

 

0.5

20

GC–FPD

2006 [25]

5

10 mL H2O (barley)

50

15

S

F/Na2SO4

25→1

GPC

Eluate→10 ACN

0.25

25

GC–TOF-MS, LC–MS–MS

2006 [26]

25

2 mL 4 mol L−1 phosphate buffer

40

25

T

Centrifuge

GC: -

GCB/PSA disp

 

0.5

20

GC– MS

This work

LC: 0.48→1.5 (MeOH/water)

0.2

10

LC–MS–MS

aEthyl acetate–cyclohexane, 1:1

bAdditional SPE clean-up step with Florisil EtAc/Hex 1:1 5 mL evap. to 1 mL

T, Turrax; B, blender; S, shaking; F, filtration; MeOH, methanol; ACN, acetonitrile; ace, acetone; hex, hexane

Although multi-residue methods based on ethyl acetate extraction have been used for more than 20 years, and continue to be used in many laboratories (they are, for example, the official methods in Sweden and Spain and are also commonly used in the Netherlands, UK, Czech Republic, Japan, and China), the methods described in the literature frequently include steps that make them, in our opinion, unnecessary laborious. Such steps include repeated extraction, filtration, clean-up steps involving GPC for non-fatty matrices, column chromatography or solid phase extraction (SPE) manifolds and evaporative concentration. Typical examples are given in Table 1. It will be shown in this paper that most of the laborious steps can be replaced by more efficient alternatives—repeated extraction is not required, an aliquot is taken after settling or centrifugation rather than filtration, use of GCB instead of GPC for removal of chlorophyll, use of dispersive SPE instead of classical SPE for clean-up (analogous to an acetonitrile-based method [29]), and injection of larger volumes into the GC instead of manual evaporative concentration.

The objective of the work discussed in this paper was to update and improve the ethyl acetate-based multi-residue method for pesticides in vegetables and fruit in respect of straightforwardness, robustness, and ease and speed of sample and extract handling. Aspects studied include dispersive clean-up using combined GCB/PSA, the possibility of preventing unacceptable adsorption of “planar” pesticides by GCB, by addition of toluene, and large-volume (20 μL) injection in GC. The method has been validated for 341 pesticides and degradation products which are analyzed by GC–MS or LC–MS–MS. For the latter the initial raw extract was used and injected after a solvent change to methanol–water. The suitability of the method as a multi-residue, multi-matrix method is evaluated by use of analytical quality-control data generated during 15 months for 271 pesticides and degradation products for over 100 different matrices, including less common and exotic crops. Results obtained for proficiency test samples during three years are also presented.

Experimental

Chemicals and reagents

Pesticide reference standards were obtained from C.N. Schmidt (Amsterdam, The Netherlands). For GC–MS a mixed stock solution containing 135 pesticides (Table 7; concentration 50 mg L−1 for each pesticide) was obtained from Alltech–Grace (Breda, The Netherlands). The full chemical names of the metabolites of phenmedipham and pyridate are methyl N-(3-hydroxyphenyl)carbamate and 3-phenyl-4-hydroxy-6-chloropyridazine, respectively. Solvents were from J.T. Baker (ethyl acetate, Resi-analysed; Deventer, The Netherlands), Labscan (toluene, Pestiscan), and Rathburn (methanol). Anhydrous sodium sulfate, ammonium formate, potassium dihydrogen phosphate, disodium hydrogen phosphate, acetic acid, and diethylene glycol (all p.A. quality) were from Merck. Water was purified by use of a MilliQ reagent-water system (Millipore).

Bondesil primary secondary amine (PSA, 40 μm) was obtained from Varian (Middelburg, The Netherlands) and GCB (graphitized carbon black) was purchased as Supelclean ENVI-carb (120–400 mesh, Supelco, Zwijndrecht, The Netherlands).

For GC–MS, in addition to the mixed stock solution, individual stock solutions of other pesticides were prepared in ethyl acetate. From these, additional mixed solutions were prepared in ethyl acetate. For LC–MS–MS analysis, individual stock solutions were prepared in methanol. Mixed solutions were prepared from the individual stock solutions and diluted with methanol. The mixed solutions were used for fortification of samples and for preparation of matrix-matched standards.

The extraction solvent was a solution of internal standard (0.05 mg L−1 antor (diethatyl-ethyl)) in ethyl acetate. Matrix-matched standards were prepared by addition of mixed solutions to control sample extracts. Dilution of the sample extract with mixed solution was never more than 10%.

Instrumentation

GC–MS analysis

GC–MS analysis was performed with a model 8000 Top GC equipped with a Best PTV (programmed temperature vaporizer) injector, an AS800 autosampler, and a Voyager mass spectrometer (Interscience, Breda, The Netherlands). The instrument was controlled by Masslab software. The injector was equipped with a 1 mm i.d. liner with porous sintered glass on the inner surface. The GC was equipped with a 30 m × 0.25 mm i.d., 0.25 μm film, HP-5-MS column and a 2.5 m precolumn (same as the analytical column, connected by means of a press-fit connector).

For PTV injection in solvent-vent mode 20 μL was injected at 5 μL s−1. The solvent was vented at 50°C in 0.67 min using a split flow of 100 mL min−1. The split valve was then closed and the analytes retained in the liner were transferred to the GC column by ramping the temperature at 10° s−1 to 300°C. Total transfer time was 2.5 min after which the split was re-opened.

Helium was used as carrier gas at constant flow (1.5 mL min−1). The oven temperature was maintained at 90°C for 2 min after injection then programmed at 10° min−1 to 300°C which was maintained for 10 min. The transfer line to the MS was maintained at 305°C.

Mass spectrometry was performed with electron-impact (EI) ionization (electron energy 70 eV) at a source temperature of 200°C. Data were acquired in full-scan mode (m/z 60–400), after a solvent delay of 5.5 min, until 30 min. Scan time and inter-scan delay were 0.3 and 0.1 s, respectively, resulting in 2.5 scans s−1. The detector potential was 450 V.

Masslab software (Interscience, The Netherlands) and an Excel macro developed in-house were used for data handling and quantitative data evaluation.

LC–MS–MS analysis

LC was performed with an Agilent, model 1100 instrument comprising degas-unit, pump, autosampler, and column oven. A 4 mm × 2 mm i.d. C18 guard column (Phenomenex) and a 150 mm × 3 mm i.d. LC column (Aqua, 5 μm C18, Phenomenex) were coupled to a triple-quadrupole mass spectrometer (model API2000 or API3000, Applied Biosystems, Nieuwerkerk a/d Yssel, The Netherlands). Analyst 1.2 and, later, 1.4 were used for instrument control and data handling. Additional data processing was performed using an Excel macro developed in-house.

Compounds were separated by elution with a gradient prepared from methanol–water–1 mol L−1 ammonium formate solution, 20:79.5:0.5 (component A) and methanol–water–1 mol L−1 ammonium formate solution, 90:9.5:0.5 (component B). The composition was changed from 100% A to 100% B in 8 min and was then isocratic until 24 min. The composition was then changed back to 100% A in 1 min and the column was re-equilibrated for 10 min before the next injection. The flow rate was 0.3 mL min−1 which was introduced into the MS without splitting. The injection volume was 20 μL and 10 μL for the API2000 and API3000, respectively.

Data were acquired in multiple-reaction-monitoring (MRM) mode. Electrospray ionization (ESI) (called turbo ion spray for the instruments used) mass spectrometry was performed in positive-ion mode. For the API2000 the nebulizer gas, turbo gas, and curtain gas were 20, 50, and 40 arbitrary units (a.u.), respectively. The ion-spray potential was 5000 V. Nitrogen was used as collision gas (4 psi). For the API3000 the nebulizer gas and curtain gas were 12 and 10 a.u. and the turbo gas was 7.5 L min−1. The ion spray potential was 2000 V. Nitrogen was used as collision gas (4 psi). For both instruments, the pause time was 5 ms. The dwell times for the pesticide transitions varied between 10 and 25 ms. The precursor and product ions and the collision energy (data for API3000) for each pesticide or degradation product are listed in Table 8. In the acquisition method one transition for each pesticide was measured. All transitions were acquired in one time window. The total cycle time was 2.24 s resulting in 8–10 data points across the peak. To measure the second transition a second method was created and run if confirmation was needed.

Sample preparation

Vegetable and fruit samples were taken from batches of samples as received from the food industry and trade for routine multi-residue analysis. After removal of stalks, caps, stems, etc., as prescribed by 90/642/EEC Annex I [34], an amount corresponding, at least, to the minimum size of laboratory samples (usually 1–2 kg [35]) was homogenized in a large-scale Stephan food cutter. A subsample (25 g) was weighed into a centrifuge tube. Fortification was performed at this stage. Phosphate buffer (pH 7, 4 mol L−1, 2 mL) and extraction solution (ethyl acetate with internal standard, 40 mL) were then added. Just before Turrax extraction anhydrous sodium sulfate (25 g) was added. After Turrax extraction (1 min) the tubes were centrifuged (sets of four).

For GC–MS analysis, Eppendorf cups were prefilled with 25 mg PSA and 25 mg GCB. To avoid a weighing step, scoops were made in-house for this purpose. Their accuracy was established to be 25 ± 2 mg (n = 10). For clean-up, 0.8 mL extract and 0.2 mL toluene were added to the cup with the SPE materials. The cups were then closed and the samples were vortex mixed for 30 s and centrifuged (up to 24 at one time). One aliquot was transferred to an autosampler vial with insert, and a second aliquot was transferred to an autosampler vial and stored under refrigeration as back-up extract. The calculated amount of initial sample in the final extract was 0.5 g mL−1.

For LC–MS–MS analysis the initial extract (3.2 mL for the API2000 and 0.48 mL for the API3000) was transferred to a disposable glass tube. After addition of a solution of diethylene glycol in methanol (10%, 200 μL) the extract was evaporated to “dryness” under a gentle flow of nitrogen gas at 35°C (up to 36 tubes in a heater block). The residue was reconstituted in methanol (1 mL and 0.75 mL for the API2000 and API3000, respectively), by use of vortex mixing and ultrasonication (5 min). The extract was then diluted 1:1 with component A. After centrifugation one aliquot was transferred to an autosampler vial with insert, and a second aliquot was transferred into an autosampler vial and stored under refrigeration as back-up extract. The final extract concentration was 1 g mL−1 and 0.2 g mL−1 for the API2000 and API3000, respectively.

For dry products (e.g. cereals) 5 g was weighed and 20 mL water was added. After soaking for 2 h samples were processed as described above. A larger amount of extract was taken for evaporation to compensate for the reduced amount of sample processed and to bring the final extract concentration to 0.2 g mL−1.

With the final method, one person can process 30 samples in eight hours. Here processing includes specific preparation before homogenization (i.e. removal of caps from strawberries, etc.), homogenization of the samples, extraction, cleaning the Turrax between samples, clean-up for GC–MS, and solvent switch for LC–MS–MS, i.e. from laboratory sample to ready-to-inject solutions in autosampler vials.

Quantification

GC–MS

For each pesticide the concentrations were calculated for two diagnostic ions. In previous validation work (not published) using the same software it was found that for most pesticides automatic integration and repeatability of response were better when peak height, rather than area, was used. Peak height was therefore used, with few exceptions (e.g. pesticides prone to tailing, for example 2-phenylphenol). All responses were normalized to the response of the internal standard (antor). One-point calibration was performed using a fixed matrix-matched standard (tomato, see Results and discussion section) at a level corresponding to five times the LOQ. The linearity of the plot of MS response against concentration was verified periodically over the range 0.01 to 1–5 mg kg−1. For most pesticides linearity was adequate (relative response within 20% of the calibration standard) up to at least 1 mg kg−1.

LC–MS–MS

The internal standard (antor) was evaluated qualitatively only to confirm injection of the sample extract. Because of unpredictable and varying matrix effects for several of the matrices included in this work, normalization against the internal standard was not considered feasible. For each sample matrix that was fortified, a matrix-matched standard was also prepared by spiking the final extract of the corresponding control sample. Peak area was used for quantification. One-point calibration was performed using the matrix-matched standard at a level corresponding to five times the LOQ. Linearity of the MS response against concentration was verified periodically over the range 0.01 to 1 mg kg−1. For most pesticides, the relationship was linear (relative response within 20% of the calibration standard) up to at least 0.5 mg kg−1.

Validation

Initial method validation was performed in accordance with EU guidelines [36, 37]. Two times five portions of the homogenized sample were spiked with a mixture of pesticides at a low level (0.01 mg kg−1 or lower) and at a level ten times higher. Together with two unfortified control portions of the sample, they were processed and analyzed as outlined above.

Additional method-performance data were acquired by analyzing fortified samples concurrently with each batch of samples. The spike level (0.05 mg kg−1 for most pesticides) was five times the LOQ. With each batch different products were selected as much as possible. In the compilation the emphasis was on products which are less frequently reported in the literature to challenge the applicability of the method as a “multi-matrix method”. For this purpose samples were not pre-screened for absence of pesticides and, consequently, occasionally recoveries could not be determined, because of the relatively high levels incurred. Such results were eliminated from the data set.

Spectrophotometric measurement of removal of chlorophyll

For evaluation of the removal of chlorophyll by GCB and comparison with GPC, a lettuce extract was prepared by extracting 25 g lettuce with 40 mL ethyl acetate after addition of 25 g anhydrous sodium sulfate. As a reference, 0.8 mL ethyl acetate was added to 3.2 mL of this extract to bring the extract concentration to 0.5 g mL−1. For dispersive SPE, 100 mg GCB was added to sets of duplicate tubes and 3.2 mL extract was added to all tubes. Solvent was then added to four sets of tubes: set one 0.8 mL ethyl acetate, set two 0.4 mL ethyl acetate and 0.4 mL toluene (i.e. 10% toluene), set three 0.8 mL toluene (20% toluene), and set four 0.8 mL xylene (20% xylene). The extracts were vortex mixed and centrifuged.

For GPC clean-up, 2.5 mL lettuce extract was injected on to a 40 cm × 28 mm i.d. Biobeads SX3 column with 1:1 ethyl acetate–cyclohexane as eluent. The fraction collected was such that at least 50% of the pyrethroids were recovered (fraction from 105–200 mL). The eluate was first concentrated, by rotary evaporation at 40°C, to approximately 5 mL, then transferred to a tube for further concentration, under nitrogen gas, to 2.5 mL.

Final extract concentration before and after clean-up was always 0.5 g mL−1. Aliquots of the extracts were transferred to a cuvet for spectrophotometric analysis at 450 nm. If required, the extracts were diluted with ethyl acetate to bring absorption within the linear range. The amount of chlorophyll in the uncleaned extract was defined as 100%. For calibration purposes the uncleaned extract was diluted 10, 20, 40, 50 and 100 times with ethyl acetate and a calibration plot was constructed. Chlorophyll remaining after clean-up was determined from the decrease in absorption at 450 nm compared with the absorption of the uncleaned lettuce extract.

Results and discussion

Monitoring of residues in fresh produce for the food industry, especially trade and retail, calls for rapid turnaround, preferably within one or two days. This means sample preparation must be rapid and straightforward. With regard to cost and waste, consumption of solvents and reagents should be low. At the same time, EU directives with regard to sample definition (90/642/EEC, [34]) and laboratory sample size (2002/63/EC [35]) for residue analysis should be respected. This means, for example, that that a total of 2 kg grapes (after removal of stalks), five whole melons, or 1 kg strawberries (after removal of caps) must be processed. The actual analysis is performed on a subsample of the laboratory sample, after appropriate comminution. The more thorough the comminution, the smaller the subsample can be and the lower the amount of solvent needed for extraction. It has, furthermore, been reported that for well homogenized samples extraction by vortex mixing or shaking, instead of high-speed blending (Turrax) suffices for effective extraction [29], although there is still some debate on this matter [38].

Homogenization

For homogenization there are several possibilities. Food choppers or kitchen blenders are often used. Very thorough homogenization can be achieved with the latter, but it is not possible to process the entire laboratory sample at once. For this reason, large-scale food choppers are more suited. With such devices, homogeneity is not always optimum, as can be observed with, e.g., tomatoes, for which small pieces of skin drift in the “soup” obtained after homogenization. Subsampling of very small amounts is, therefore, not acceptable after this procedure, because the subsample would be insufficiently representative of the original sample. More thorough homogenization can be achieved after addition of dry-ice or liquid nitrogen (cryogenic homogenization). This procedure is recommended when reducing the subsample for analysis to 10 g. This procedure is more laborious, however, because it involves cutting the sample into pieces, freezing the sample (usually overnight), cryogenic comminution, then dissipation of the dry-ice or liquid nitrogen before further processing or storage. It also puts higher demands on the cutter (blades) and requires additional precautions for the operators (protection against low temperatures and noise). Cryogenic comminution has been recommended for some pesticides because it reduces their degradation during this step [39].

In recent years the food trade and retail have been intensifying their residue-monitoring programs and require analytical data before harvest, before accepting an assignment, or before releasing their products from distribution centers to supermarkets. For fresh produce this means there is a much pressure on laboratories for rapid turnaround (24–48 h). This is difficult to achieve when the analysis involves overnight freezing for cryogenic comminution. Thus, for reasons of ease and speed, it was decided to retain the current procedure—ambient homogenization of the entire laboratory sample by use of a large scale food cutter (thus accepting the consequence that for a limited number of pesticides the concentration found might be an underestimate). Because of non-optimum homogenization with the food cutter, subsamples should not be too small, and further comminution is required for efficient extraction of systemic pesticides. This can be achieved during extraction by use of an Ultra Turrax. We have previously established the minimum size of subsample that did not negatively affect the repeatability of the analysis. This was done with samples which contained residues. For subsamples (n = 7) of 50 and 25 g, the relative standard deviation (RSD%) was below 8% for several pesticide–matrix combinations. For pear leaves (regarded as a difficult matrix to homogenize) containing bromopropylate, phosalone, and tolylfluanide it was observed that the RSD increased from <8% to 14–18% when the amount of subsample was reduced from 25 g to 12.5 g. From this it was concluded that, with our procedure, 25 g was the minimum required amount of subsample.

pH adjustment

In the ethyl acetate-extraction procedure analytes are extracted and partitioned between water (from the matrix itself, or added water for dry crops) and ethyl acetate in one step. For basic and acidic compounds the partitioning can be affected by pH, which can vary substantially with the matrix. Because the same extract is to be used not only for GC–MS but also for LC–MS–MS (after changing the solvent to methanol) which, preferably, should also include analysis of basic and acidic pesticides, control of pH was regarded as necessary. A pH of approximately 6 was chosen as compromise for efficient extraction of basic and acidic compounds. Although acidic pesticides were not included in this work, data in the literature (for barley without pH adjustment, i.e. non-acidic conditions [26]) indicate they are extracted into ethyl acetate.

For pH adjustment others have used sodium hydroxide [16, 17, 18] or sodium hydrogen carbonate [11, 14, 25] (Table 1). A disadvantage of this is that the amount of salt needed depends on the acidity of the sample. Addition of too much will result in a high pH and possible degradation of base-sensitive pesticides. To keep the method as straightforward as possible the pH was adjusted using a solution of concentrated phosphate buffer (4 mol L−1, 2 mL). A solution was preferred over addition of solid salts because this enabled use of a dispenser and eliminated additional weighing of the salts. The buffer resulted in appropriate pH adjustment for most matrices, although there were exceptions, for example lemon and lime.

Extraction

The two conditions most relevant to extraction efficiency are the sample-to-solvent ratio and addition of salt, which in ethyl acetate-based multi-residue methods has always been sodium sulfate.

The amount of ethyl acetate (in mL) relative to the amount of sample (in g) is, typically, at least 2:1. This ratio has been used for many years (Table 1). It results in good extraction efficiency and is practical with regard to achieving phase separation and avoidance of emulsions. To avoid sacrificing decades of method history no attempts were made to reduce the ratio; to do so might also adversely affect recovery and/or complicate phase separation. Larger amounts (as used by several other laboratories; Table 1) result in greater solvent consumption and more dilute extracts. In previous work [15] it has been shown that the efficiency of extraction of polar pesticides improves with the amount of salt added. When 50 mL ethyl acetate and 25 g sample were used, 25 g sodium sulfate was sufficient to obtain recoveries of 80% or better, even for very polar and highly water-soluble compounds, for example acephate and methamidophos. Because these recoveries were obtained with a single extraction it was found unnecessary to perform repeated extraction, as some laboratories are doing [11, 18, 20, 21]. For addition of the sodium sulfate an automatic salt-dispenser coupled to a balance, as is used in our laboratory, or a scoop, was found to be very convenient.

The extraction procedure involves successive addition of buffer, extraction solution (ethyl acetate with internal standard), and sodium sulfate to the centrifuge tube containing the sample, after which the pesticides are extracted and partitioned in one step using a Turrax. During this step the subsample is further comminuted for efficient extraction of the pesticides from the matrix. Vortex mixing, shaking or sonication were regarded as less efficient for subsamples that were homogenized in a large-scale food cutter under ambient conditions, but this was not investigated, partly because a variety of samples containing residues would be required to do so in an appropriate manner.

It was noted from the literature that filtration is often performed to separate the solid pellet from the liquid. Again, there is no real need for this step, which involves additional glassware and, occasionally, rinsing (diluting) of the extract. For many samples a clear ethyl acetate extract is obtained after settling; if not the tubes can be centrifuged. This is no more laborious than filtration and does not involve additional glassware.

Because the same Turrax is used for several samples, carry-over is an aspect to be considered. Between samples the Turrax is cleaned first by rinsing with water, by means of a flow-through beaker, then by brief immersing in two beakers containing ethyl acetate. Using this procedure, carry-over was tested by analyzing a blank after a sample that had been fortified at 5 mg kg−1. Carry-over was less then 0.1%, indicating that the straightforward cleaning procedure was sufficient to avoid cross-contamination up to 5 mg kg−1 when setting reporting limits not lower than 0.01 mg kg−1.

GC–MS analysis

Clean-up

In ethyl acetate-based multiresidue methods either no clean-up or GPC clean-up is performed. This has hardly changed over the years (Table 1). In contrast with acetone and acetonitrile-based methods, in which SPE is commonly employed, this has been reported only occasionally for ethyl acetate-based methods. Obana et al. [10] used a cartridge packed with layers of water-absorbing polymer and GCB. Sharif et al. [21] described a clean-up using SAX/PSA but the scope of the method was restricted to organochlorine and organophosphorus pesticides. Zhang et al. [20] used a clean-up based on Florisil and achieved adequate recovery of many pesticides but not the more polar organophosphorus pesticides. It has been stated that in GC analysis with use of highly selective detectors, for example MS–MS no clean-up is required, even when injecting 15 mg equivalent of matrix (green bean, tomato, pepper, cucumber, marrow, egg plant, and water melon [40]). Other laboratories experienced problems with contamination of the GC inlet and tried to solve this by automatic exchange of liner inserts [14, 41]. This is in agreement with our experience that injection of 10 mg matrix equivalent, especially for leafy vegetables, does result in rapid deterioration of system performance because of accumulation of non-volatile material in the inlet. This makes the system less robust, and frequent exchange of the liner (daily) and GC–pre column (weekly) is required. Another problem encountered with injection of the uncleaned extracts was a shift in the retention times of pesticides relative to that of the calibration standard for some sample extracts. This shift was insufficiently corrected by automatic adjustment of retention times relative to that of the internal standard. Typically, shifts were in the range 0.05–0.20 min and were most abundant for the “azole” pesticides. Such shifts can complicate automatic peak assignment during data-handling. When data acquisition is performed in a non-continuous mode (e.g. selected-ion monitoring or MS–MS) such shifts also increase the risk of pesticides shifting from their acquisition window. For injection of relatively large amounts of matrix (e.g. 10 mg) in GC analysis clean-up for removal of bulk co-extractants is therefore regarded as a prerequisite for robust analysis of a wide variety of vegetable and fruit matrices.

For vegetables and fruit matrices, chlorophyll (MW ∼900) and other pigments, for example carotenoids (e.g. β-carotene, MW 537) are typical bulk co-extractants. Most of these compounds are of low volatility and are not apparent as interferences in the chromatograms; they do, however, accumulate in the liner of the GC and eventually have an adverse effect on transfer of analytes to the column and/or on peak shape. Because of its high molecular weight, chlorophyll can be removed by GPC. A disadvantage is that the extract is strongly diluted and reconcentration by rotary evaporation is almost inevitable when LODs of 0.01 mg kg−1 are required. Such a step would contribute substantially to overall sample-preparation time. Although a very efficient on-line combination of GPC and GC–MS was described recently [42], avoiding GPC whenever possible would be even more straightforward. Solid-phase extraction is an alternative clean-up procedure which involves less dilution and is less laborious. Even more efficient is SPE in the so-called dispersive mode, as described by Anastassiades et al. [29]. Here the solid phase is simply added to the extract, thereby avoiding typical SPE procedures such as conditioning, sample transfer, elution, and evaporative reconcentration. The pesticides partition between the solid phase and the solvent and after vortex mixing and centrifugation the supernatant is ready for analysis.

Two stationary phases, graphitized carbon black (GCB) and phases with amino functionality, have been shown to be particularly effective for removing co-extracted material from the raw extract while not removing most of the pesticides; this makes them very suitable for wide-scope methods [28, 29, 31, 38, 43, 44, 45].

Initially, a method was envisaged using SPE column clean-up with GCB, because for leafy vegetables this was found to be the only sufficiently effective alternative to GPC. After the publication on dispersive SPE [29] it was decided to investigate this approach, thus sacrificing some clean-up potential (as has been reported in the literature [31]) for ease and speed.

GCB is well known to adsorb planar molecules, including chlorophyll and other pigments but also pesticides with planar functionality. In acetonitrile-based methods, toluene (typically 25%) is often added to the eluent to desorb these pesticides also from the SPE column [28, 38, 43, 45]. One of the objectives of this work was to investigate the possibility of using GCB in a dispersive clean-up step without unacceptable losses of planar pesticides. First we investigated which pesticides, dissolved in ethyl acetate, are adsorbed by GCB. A somewhat arbitrary, 25 mg mL−1 GCB phase was added to standard solutions. After vortex mixing and centrifugation the solution was analyzed by GC–MS (165 pesticides) and, after changing the solvent to methanol, by LC–MS–MS (another 70 pesticides), and the responses were compared with those from untreated standard solutions. For 35 pesticides (15%) adsorption was observed (Table 2). In addition to the pesticides included in this test, it is known from the literature [44] that chinomethionate, furametpyr, and pyraclofos are also adsorbed by GCB (from acetone–cyclohexane, 1:4).
Table 2

Pesticides adsorbed by GCBa

Strong adsorption (rec. 0–50%)

Medium adsorption (rec. 50–70%)

Not consistent

Measured by GC–MS

 Chlorothalonil

Azinphos-ethyl

Phosmet

 Cyprodinil

Azinphos-methyl

Prochloraz

 Fenazaquin

Chlorpyrifos-methyl

Pyrazophos

 Hexachlorobenzene

Dicloran

Trifluralin

 Mepanipyrim

EPN

 

 Pentachloroaniline

Fenamiphos

 

 Phosalone

Phorate

 

 Pyrimethanil

Quintozene

 

 Quinoxyfen

  

Measured by LC–MS–MS

 Carbendazim

Fenpyroximate

 

 Clofentezine

Flufenoxuron

 

 Desmedipham

Tricyclazole

 

 Diflubenzuron

Triflumuron

 

 Flucycloxuron

Thiophanate-methyl

 

 Hexaflumuron

  

 Phenmedipham

  

 Pymetrozine

  

 Thiabendazole

  

aPesticides in ethyl acetate, 25 mg GCB mL−1 solvent

rec., recovered

To investigate how much toluene is required to prevent adsorption of planar pesticides by GCB in dispersive SPE, the partitioning experiment was repeated with standard solutions of 10, 20, or 30% toluene in ethyl acetate. This was done for the GC–MS pesticide mixture only.

As is apparent from Fig. 1, even 10% toluene dramatically improved recovery. With 20% toluene recovery of all pesticides was higher than 65%. It should be noted that this experiment with standard solutions is the worst case. For real samples chlorophyll and carotenoids will also affect the distribution in favor of the pesticides in solution. Use of 30% of toluene further improved recovery only slightly. Twenty percent was regarded as optimum with regard to distribution and ease of solvent elimination in large-volume injection (see below). In addition to toluene, two alternative analogues, benzene and xylene, were also considered. Benzene, was not tested because it could not be used in routine practice because of its carcinogenic properties (although it would have been favorable with regard to solvent elimination). Xylene was tested in a similar way as toluene. Results obtained for hexachlorobenzene and chlorothalonil by use of the two solvents are compared in Fig. 2. Slight but consistently better recovery was obtained with xylene—>70% recovery could now be obtained for all pesticides. Because of its greater volatility, however, toluene was finally selected.
Fig. 1

Effect of the amount (%) of toluene in ethyl acetate on recovery of pesticides adsorbed by GCB (25 mg mL−1). hcb, hexachlorobenzene; pca, pentachloroaniline; ctn, chlorothalonil; mep, mepanipyrim; cypr, cyprodinil; pyri, pyrimethanil; fena, fenazaquin; quin, quinoxyfen; pyra, pyrazophos; epn, EPN

Fig. 2

Comparison of toluene and xylene as additives for preventing adsorption of planar pesticides by GCB in dispersive SPE

Obviously, toluene is also likely to affect adsorption of chlorophyll and/or carotenoids and might reduce the effectiveness of clean-up. To investigate this, a lettuce extract was prepared, the dispersive clean-up experiments were performed with different amounts of toluene, and removal of chlorophyll was verified. Visually it was clearly apparent that, despite addition of toluene, the intense green color turned light yellow, indicating that chlorophyll was removed to a large extent. To enable more quantitative evaluation, the extracts were also measured with a spectrophotometer at 450 nm. For comparison, the same extracts were also cleaned by GPC. The results are presented in Table 3. Without toluene, chlorophyll was very effectively removed. Absorption at 450 nm was reduced by 94%. Toluene, as expected, reduced adsorption of chlorophyll, but removal was still 87% or 78%, after addition of 10% or 20% toluene in ethyl acetate, respectively. Similar to observations with the planar pesticides, adsorption was reduced slightly more by use of xylene than by use of toluene. With GPC, chlorophyll removal was 60%. It should be noted here that the elution window was relatively wide, to include pyrethroids within the scope of the method. The elution windows for chlorophyll (and carotenoids) partially overlap those for pyrethroids, as has also been reported by others [44]. From these experiments it can be concluded that chlorophyll has more affinity than the planar pesticides for GCB. In dispersive SPE toluene effectively prevents unacceptable adsorption of planar pesticides while to a large extent maintaining its cleaning properties in respect of chlorophyll. Dispersive GCB not only enables much faster chlorophyll removal, it is also more effective when including pyrethroids in the scope of the method. For non-fatty vegetable and/or fruit matrices, therefore, GPC is not required and dispersive GCB clean-up is a much faster alternative without sacrificing scope.
Table 3

Removal of chlorophyll by dispersive SPE (GCB) and GPC

Clean-up procedure

Chlorophyll removal (%)

Dispersive SPE, 100% ethyl acetate

94

Dispersive SPE, 10% toluene in ethyl acetate

87

Dispersive SPE, 20% toluene in ethyl acetate

78

Dispersive SPE, 20% xylene in ethyl acetate

71

GPC (fraction incl. pyrethroids)

60

Sample extract: lettuce 0.5 g mL−1. Dispersive SPE: 25 mg GCB mL−1. GPC: wide scope elution window, i.e. including pyrethroids.

The GCB clean-up enabled continuous injection of extracts of leafy vegetables without rapid system deterioration. With some matrices, however (e.g. plums, grapefruit), retention time shifts were still observed. In addition, depending on the matrix, quite intensive interferences could be observed in the GC–MS TIC chromatograms. Further clean-up by PSA, complementing the GCB clean-up by removing compounds such as organic acids and sugars by hydrogen bonding, was therefore investigated. To keep sample clean-up as straightforward and rapid as possible focus was on a combined dispersive GCB/PSA clean-up.

After the outcome of the GCB experiments, partitioning of the pesticides and co-extractants will be between PSA and ethyl acetate–toluene, 8:2. Because no information was available about the distribution of pesticides between these two phases, this was obtained by analyzing pesticide standards in ethyl acetate–toluene, 8:2, with and without PSA. Preliminary experience with dispersive PSA clean-up revealed that with some matrices (e.g. cereals) 25 mg mL−1 did not result in complete elimination of interfering compounds (e.g. fatty acids) typically removed by PSA. Partitioning with a much larger amount of adsorbent (200 mg mL−1) was, therefore, also studied.

With 25 mg mL−1 losses of 30–40% were observed for sixteen pesticides, most probably as a result of adsorption, although the possibility of degradation induced by the basic nature of the PSA material could not be fully excluded. The findings were confirmed by the experiment with 200 mg PSA mL−1 (Table 4). The pesticides for which interaction with PSA was observed all had a C=O or P=O group in common (except for chlorothalonil). Our findings are not in full agreement with those of Anastassiades et al. [29] who did not observe losses as a result of using PSA. For this there can be two explanations. In our experiment adsorption was tested with standard solution rather than matrix. Co-extractants in matrix are likely to compete with the pesticides during adsorption. Second, with our method the organic phase (ethyl acetate–toluene, 8:2) is less polar than the acetonitrile phase; this could result in a stronger interaction between the polar functionality of the pesticides and amino functionality of PSA. From our results it became clear that with regard to the amount of PSA “the more, the better” does not apply. Another observation was that a hump appeared in the TIC chromatogram after a 20-μL injection of solvent mixed with 200 mg PSA mL−1. This hump, which eluted between 6 and 12 min, consisted of many peaks and a variety of masses. Cleaning of the PSA by washing with ethyl acetate (3 × 20 mL for 1 g), then drying by rotary evaporation, eliminated this contamination without affecting the clean-up properties. To keep the method straightforward, 25 mg PSA mL−1 was used as default, and the material was not cleaned before use.
Table 4

Adsorption of pesticides by PSA

Pesticide

Recovery (%)

Acephate

43a

Acrinathrin

41b

Asulam

0a

Carbaryl

56b

Chlorothalonil

17b

Cycloxidim

39a

Dichlorvos

33b

Dimethoate

62b

Hymexazol

0a

Mevinphos

62b

Phosmet

25b

Phosphamidon

63b

Profenofos

56b

Pyridate

40a

Pyridate-metabolite

7a

Sethoxydim

48a

aAfter partitioning with ethyl acetate, 25 mg mL−1 and LC–MS–MS analysis

bAfter partitioning with ethyl acetate–toluene, 8:2, 200 mg PSA mL−1 and GC–MS analysis

The clean-up proved effective at reducing retention time shifts. As an example, for a plum extract without clean-up, the retention times of 24 pesticides (out of 140) were shifted by more than 0.05 min compared with the calibration standard. After clean-up this occurred for three pesticides only. With other matrices also shifts were reduced, but for some matrices (herbs, e.g. parsley) deviations were still quite common.

As an illustration of the removal of co-extractants from the ethyl acetate extract (or, in fact, from the ethyl acetate–toluene, 8:2, extract) by dispersive GCB/PSA clean-up, GC–MS total ion current chromatograms of extracts obtained with and without clean-up are shown in Fig. 3. The most apparent differences are indicated. Several abundant matrix peaks are removed or strongly reduced. For lettuce, the overall background level between 15 and 25 min was also reduced. This clearly visible clean-up was mainly caused by the PSA material. With GCB alone differences between cleaned and uncleaned were much less apparent. The main benefit of GCB was prevention of rapid build up of non-volatile material (chlorophyll) in the liner, which enables prolonged use of the system without maintenance. Experience with method for more than three years and analysis of over 15,000 vegetable and fruit samples shows that, on average, the liner must typically be replaced weekly (after 150–200 injections; iprodion, dimethipin, and chlorfenapyr are the first for which response is lost). Further GC–MS maintenance consists in replacement of pre-column once of twice a month. The GC column is replaced approximately twice a year. The source of the MS is cleaned once a month.
Fig. 3

GC–MS chromatograms. Overlay total ion chromatograms (TICs) obtained after 20 μL injection of an extract of mandarin (top) and lettuce (bottom) without (higher peaks) and with clean-up

In a continuing search for even further simplification of sample preparation, the possibility of combined extraction and dispersive SPE clean-up in one step was investigated. For two matrices (lettuce and mandarin, fortified with 140 pesticides, triplicate experiments) the solid phase materials (GCB/PSA, relative amounts similar to previous experiments) were added directly to the centrifuge tube containing the sample, sodium sulfate, and the extraction solvent (to which 20% toluene had been added). After Turrax extraction and centrifugation, the extract was ready for injection into the GC. Recovery was compared with that obtained by use of dispersive clean-up after separation of the ethyl acetate extract from the sample mixture. As could be seen from the color of the extract (the lettuce extract was almost colorless) the GCB remained effective. Adsorption of chlorophyll is based on planarity (shape) rather than polarity and, therefore, this will occur from both the aqueous and the organic phases. As was to be expected, the same was not true for PSA. The presence of water prevented adsorption of co-extractants with a hydroxyl group, i.e. almost identical GC–MS total-ion chromatograms were obtained from extracts which were not cleaned and from those cleaned in the centrifuge tube. Pesticide recovery obtained after use of successive or simultaneous dispersive SPE clean-up was very similar, although recovery of some pesticides in the combined approach was too high, because of co-elution of interferences. The final method therefore used successive extraction and dispersive SPE clean-up.

Large-volume injection

GC–MS analysis of sample extracts was performed in full-scan mode. This enables detection of any GC–amenable pesticide. Because system LOQ for a quadrupole mass spectrometer in full-scan mode is limited, conservatively estimated at 100 pg, 10 mg matrix equivalent must be introduced into the GC to reach a target LOQ of 0.01 mg kg−1. With an extract concentration of 0.5 g mL−1, this means 20 μL must be introduced into the GC. Off-line tenfold evaporative concentration and then 2 μL injection could also be performed, but this would involve clean-up of larger volumes of extract, the risk of loss of the volatile pesticides (e.g. dichlorvos), and an additional step in sample preparation. Although large-volume injection in GC is a well established technique [47, 48], many routine laboratories are still reluctant to apply it; if they do, the volume is often restricted to 5–10 μL. Such volumes can be accommodated in liners with a frit or even in empty (baffled) liners when injection speed is carefully adjusted. For larger volumes there is a risk of flooding [46], i.e. that extract is lost as liquid through the split exit. To prevent this, liners can be packed with a variety of materials. Packing materials often have the disadvantage of a large surface area with active sites, however, resulting in degradation and/or adsorption of thermo labile and/or polar pesticides; problems can also be encountered with splitless transfer of higher boiling pesticides (e.g. deltamethrin) from the liner to the GC column. Other disadvantages can be a pressure drop over the liner (slows down solvent elimination) and liner-to-liner variability requiring re-optimization of the solvent-elimination process after liner replacement. A means of by-passing the disadvantages of packed liners while still achieving accommodation of 20–50 μL of liquid was described in 1993 by Staniewski and Rijks [49]. They developed a liner with a sintered porous glass bed on the inner surface wall of the liner. The liquid is retained in the porous glass bed. The potentially active glass surface area is relatively small compared with the materials in packed liners. The gas flow is not obstructed, because the centre of the liner is empty. This enables efficient solvent vapor removal during solvent elimination and efficient transfer of analytes to the analytical column during splitless injection after solvent elimination. Since the early 2000s such liners have been commercially available for PTV injectors from several suppliers, and since then our laboratory has implemented 20 μL as default injection volume for ethyl acetate.

After the development of the dispersive GCB clean-up, the solvent to be introduced into the GC contained 20% toluene, which might effect the processes involved in large-volume injection differently from 100% ethyl acetate. Because toluene does not evaporate azeotropically with ethyl acetate and is less volatile, it will be the main solvent left at the end of the evaporation process. Injection of 20 μL 20% toluene in ethyl acetate means that 4 μL toluene is introduced. The PTV used in this work was equipped with a 1 mm i.d. porous glass bed liner that could hold approximately 30 μL within the zone that is appropriately heated during splitless transfer. Up to this volume there is no need for optimization of injection speed. To obtain information about splitless transfer of the last few microliters of toluene after solvent elimination, cold splitless injections of 1, 2, and 3 μL of standards in 100% toluene were performed. Even with 2-μL volumes peak distortion (fronting peak shape) was observed for pesticides of medium volatility. With 1 μL injections peak shape was good and for several pesticides even better than for ethyl acetate. On injection of 20 μL standard in ethyl acetate–toluene, 8:2, in the solvent-vent mode, no peak distortion was observed, indicating that less then 2 μL toluene remained in the injector after the solvent-vent step. As observed earlier with large-volume injection of ethyl acetate, the vent time (here set at 40 s using an initial PTV temperature of 50°C) was not at all critical, even for the most volatile pesticide (dichlorvos). Venting for 35 or 50 s did not dramatically affect responses or peak shape of the pesticides. In our experience, this phenomenon is typical for porous glass bed liners and contributes to the robustness of the method.

Validation of GC–MS method

In the past a method based on simple ethyl acetate extraction followed by direct GC–MS analysis of the raw extract [4] had been validated for concentrations in the range 0.05–0.5 mg kg−1. The modified method described here involved a dispersive clean-up step, large-volume injection, and injection of ten times more matrix into the GC. Re-validation was therefore required, and focused on method performance at low concentrations. This was done using lettuce as matrix. The validation set consisted of two control samples, five fortifications at 0.001–0.05 mg kg−1 and five fortifications at a level ten times higher. Over 200 pesticides were included in the validation procedure. The results are presented in Table 5. For the 0.01–0.5 mg kg−1 concentration range the EU criteria (recovery 70–110%, RSD 30%, 20%, or 15% for ≤0.01, >0.01–0.1, and >0.1–1 mg kg−1, respectively [37]) were met for 184 of the 201 pesticides included in the validation. At a level a factor of ten lower (fortification in the 0.001–0.01 mg kg−1 range for most pesticides) 147 pesticides could still be detected and for most (78%) of these recovery and RSDs were acceptable. For many pesticides S/N ratios were surprisingly good and background-corrected mass spectra often contained sufficient diagnostic ions (or were even recognizable mass spectra) to enable identity confirmation, as is illustrated in Fig. 4. The limits of detection, defined as S/N = 3 for one favorable diagnostic ion for each pesticide, were determined on the basis of the signals from the low fortification levels and the average noise observed in duplicate control samples. The LOD was at or below 0.001 mg kg−1 for 78 pesticides, between 0.001 and 0.005 mg kg−1 for 73 pesticides, between 0.005 and 0.01 mg kg−1 for 29 pesticides, between 0.01 and 0.05 mg kg−1 for 16 pesticides, and higher for four pesticides.
Table 5

GC–MS re-validation data for pesticides in lettuce

 

Pesticide

tR (min)

m/z (quant)

Level (mg kg−1)

Rec. (%)

RSD (%)

Level (mg kg−1)

Rec. (%)

RSD (%)

LOD (mg kg−1)

1

Acephate

10.45

136

0.026

35

4

0.257

58

9

0.006

2

Acrinathrin

22.06

289

0.018

118

15

0.178

94

9

0.003

3

Aldrin

16.58

265

0.003

139

25

0.031

94

2

0.002

4

Atrazine

14.17

215

0.002

91

21

0.018

98

7

0.002

5

Azinphos-methyl

21.64

160

0.01

119

9

0.098

110

7

0.009

6

Azoxystrobin

25.80

344

0.01

82

8

0.099

92

5

0.003

7

Benalaxyl

19.82

148

0.005

85

9

0.047

90

8

0.002

8

Benzoylurea (deg)a

8.90

141

 

113

5

0.025

110

6

 

9

Bifenthrin

20.91

181

0.007

84

9

0.068

89

13

≤0.001

10

Biphenyl

9.81

154

0.006

97

10

0.063

101

5

≤0.001

11

Bitertanol

22.89

170

0.003

83

9

0.031

90

4

0.002

12

Bromophos

17.02

331

0.003

99

7

0.032

105

2

≤0.001

13

Bromopropylate

20.94

343

0.003

103

13

0.032

89

5

0.001

14

Bromuconazole

20.86

173

0.002

109

12

0.024

91

6

≤0.001

15

Bupirimate

18.72

273

0.003

61

8

0.032

91

5

0.001

16

Buprofezin

18.68

172

0.002

85

14

0.019

92

8

0.001

17

Cadusafos

13.46

158

0.002

117

18

0.021

92

11

0.001

18

Carbaryl

15.84

115

0.004

93

9

0.04

93

8

0.002

19

Carbofuran

14.10

164

0.003

88

7

0.033

93

3

0.002

20

Chlordane, alpha-

17.81

373

0.001

*

*

0.015

92

4

0.002

21

Chlordane, gamma-

18.12

373

0.002

84

7

0.015

96

4

0.001

22

Chlorfenvinphos

17.47

323

0.003

84

6

0.03

97

5

0.001

23

Chloroaniline, 3-

7.49

127

0.002

*

*

0.025

25

46

0.003

24

Chlorobenzilate

19.10

251

0.005

*

*

0.05

95

4

0.010

25

Chlorothalonil

15.05

264

0.004

146

15

0.042

136

9

≤0.001

26

Chlorpropham

13.08

171

0.006

*

*

0.059

95

6

0.015

27

Chlorpyrifos

16.67

314

0.003

102

16

0.034

102

5

0.002

28

Chlorpyrifos-methyl

15.70

286

0.001

105

5

0.015

102

6

≤0.001

29

Chlorthal-dimethyl

16.77

301

0.005

90

7

0.051

91

4

0.001

30

Cinerin-1

18.67

150

0.053

84

3

0.528

93

6

0.041

31

Clofentezine

22.45

304

0.014

*

*

0.14

101

14

0.050

32

Cyfluthrin I

23.33

226

0.041

91

7

0.407

93

6

0.023

33

Cyfluthrin II

23.60

226

0.041

100

8

0.407

88

8

0.016

34

Cyhalothrin-lambda

21.91

181

0.003

110

10

0.029

93

6

0.002

35

Cypermethrin-I

23.65

163

0.018

107

29

0.184

96

5

0.008

36

Cypermethrin-II

23.83

181

0.018

94

16

0.184

97

5

0.006

37

Cypermethrin-III

24.07

181

0.018

96

10

0.184

96

6

0.013

38

Cyproconazole

18.97

222

0.006

72

20

0.059

88

7

0.001

39

Cyprodinyl

17.19

224

0.005

105

25

0.051

85

10

≤0.001

40

Cyromazine

14.47

166

0.013

*

*

0.13

82

56

0.040

41

DDE, o,p′-

17.90

248

0.002

*

*

0.015

92

3

0.009

42

DDE, p,p′-

18.50

248

0.001

110

11

0.015

100

5

≤0.001

43

DDT, o,p′-

19.32

235

0.001

102

9

0.015

94

7

0.001

44

DDT, p,p′-

20.28

235

0.002

86

11

0.016

95

8

0.001

45

Deltamethrin

25.44

253

0.022

114

9

0.223

106

5

0.014

46

Demeton-S-methyl-sulfone

16.11

169

0.03

71

15

0.302

91

9

0.004

47

Desmethylpirimicarb

15.42

152

0.003

*

*

0.026

76

7

0.005

48

Diazinon

14.70

137

0.002

98

14

0.019

94

3

0.001

49

Dichlofluanid

16.41

224

0.004

79

9

0.044

98

8

≤0.001

50

Dichlorvos

8.00

185

0.002

107

6

0.018

92

7

≤0.001

51

Dicloran

13.96

206

0.003

96

16

0.029

106

2

0.003

52

Dicofol (as DCBP)

16.75

250

0.005

*

*

0.049

126

33

0.010

53

Dieldrin

18.56

263

0.004

*

*

0.041

95

6

0.005

54

Diethofencarb

16.53

267

0.005

98

5

0.046

96

6

0.001

55

Difenoconazole-I

25.12

323

0.029

94

10

0.288

95

3

0.006

56

Difenoconazole-II

25.36

323

0.029

91

9

0.288

99

3

0.003

57

Diflubenzuron (deg)

6.63

153

0.005

124

9

0.05

107

2

0.002

58

Dimethoate

13.97

125

0.009

*

*

0.091

91

4

0.017

59

Dimethomorph

25.88

301

0.021

95

7

0.207

87

5

0.002

60

Diniconazole

19.54

268

0.002

*

*

0.018

89

12

0.003

61

Diphenylamine

12.76

169

0.003

86

10

0.028

72

15

≤0.001

62

Disulfoton

14.81

88

0.005

101

5

0.05

96

3

0.002

63

DMSA

13.19

200

0.005

87

9

0.052

92

7

0.002

64

DMST

14.37

214

0.005

*

*

0.053

73

32

0.019

65

Dodemorph

16.95

154

0.005

67

26

0.046

91

7

0.002

66

Edifenfos

18.07

310

0.005

96

10

0.05

94

8

0.001

67

Endosulfan-alpha

18.08

239+197

0.005

*

*

0.047

93

5

0.010

68

Endosulfan-beta

19.19

195+241

0.005

*

*

0.046

87

1

0.020

69

Endosulfan-sulfate

19.98

274+237

0.005

82

10

0.047

97

4

0.004

70

Endrin

20.94

245

0.005

*

*

0.051

90

8

0.006

71

EPN

20.57

169

0.01

103

23

0.099

94

7

0.001

72

Epoxiconazole

20.55

194

0.007

*

*

0.066

92

1

0.010

73

Esfenvalerate

24.77

125

0.004

*

*

0.036

98

5

0.008

74

Ethion

19.36

231

0.003

*

*

0.03

97

3

0.007

75

Ethoprofos

12.86

158

0.003

88

17

0.026

93

5

0.001

76

Etofenprox

23.85

164

0.005

100

11

0.049

93

5

0.004

77

Etridiazole

10.74

211

0.014

95

8

0.138

98

4

0.001

78

Etrimfos

15.01

292

0.003

96

4

0.025

93

5

≤0.001

79

Famoxadone

25.90

330

0.01

97

9

0.1

96

5

0.003

80

Fenamiphos

18.23

303

0.015

97

6

0.154

91

11

≤0.001

81

Fenarimol

22.13

139

0.004

*

*

0.038

101

4

0.008

82

Fenazaquin

21.22

160

0.003

152

12

0.027

114

8

0.001

83

Fenbuconazole

23.30

129

0.003

*

*

0.03

92

3

0.006

84

Fenhexamid

20.10

177

0.003

*

*

0.026

90

7

0.004

85

Fenitrothion

16.25

260

0.001

*

*

0.015

95

8

0.003

86

Fenoxycarb

20.89

116

0.015

117

8

0.154

94

4

0.002

87

Fenpiclonil

20.78

238

0.007

88

5

0.071

92

8

0.003

88

Fenpropathrin

21.05

181

0.005

77

13

0.05

92

13

0.001

89

Fenpropimorph

16.63

128

0.001

*

*

0.01

93

2

0.002

90

Fenthion

16.63

278

0.002

99

7

0.023

99

5

≤0.001

91

Fenvalerate

24.54

167

0.004

*

*

0.036

103

8

0.006

92

Fipronil

17.57

367

0.002

81

6

0.024

94

9

≤0.001

93

Flucythrinate-I

23.77

199

0.017

93

11

0.174

92

1

0.004

94

Flucythrinate-II

18.51

199

0.017

94

6

0.174

93

4

0.004

95

Fludioxonil

19.05

248

0.003

113

13

0.027

97

3

0.001

96

Flufenoxuron (deg)

14.79

331

0.012

104

13

0.118

118

19

0.005

97

Flusilazole

18.70

233

0.006

68

8

0.055

87

6

≤0.001

98

Flutolanil

18.30

323

0.003

81

9

0.025

86

8

≤0.001

99

Fluvalinate, tau-

24.80

250

0.025

95

11

0.245

95

5

0.004

100

Folpet

17.65

147

0.016

96

16

0.159

91

15

0.009

101

Fonofos

14.55

246

0.005

94

6

0.049

92

7

0.001

102

Formetanate

15.27

122

0.05

*

*

0.498

102

62

0.188

103

Formothion

15.27

170

0.005

102

13

0.049

89

4

0.004

104

Fuberidazole

15.79

184

0.005

83

29

0.051

55

17

0.001

105

Furalaxyl

17.59

242

0.005

95

10

0.051

101

9

0.002

106

Heptachlor

12.19

272

0.001

*

*

0.014

92

5

0.003

107

Heptachlorepoxide-I

17.45

353

0.003

*

*

0.033

97

12

0.004

108

Heptachlorepoxide-II

17.36

353

0.001

96

13

0.015

94

8

≤0.001

109

Heptenophos

12.24

124

0.003

95

5

0.03

93

3

≤0.001

110

Hexachlorobenzene

18.33

284

0.005

75

28

0.049

96

15

0.001

111

Hexaconazole

18.32

216

0.002

*

*

0.02

87

7

0.003

112

Imazalil

18.37

215

0.005

79

50

0.05

77

14

0.002

113

Iprodione

20.75

316

0.012

108

7

0.12

95

4

0.004

114

Isofenphos

17.46

213

0.005

*

*

0.051

93

3

0.010

115

Jasmolin-I

19.36

123

0.053

*

*

0.528

77

5

0.100

116

Kresoxim-methyl

18.73

206

0.014

95

6

0.139

91

9

0.005

117

Lindane

14.41

183

0.002

86

18

0.02

99

6

0.001

118

Linuron

16.35

248

0.005

*

*

0.048

79

9

0.010

119

Lufenuron (deg)

11.48

176

0.011

123

20

0.114

76

34

0.004

120

Malathion

16.43

173

0.003

*

*

0.034

98

5

0.005

121

Mecarbam

17.49

329

0.003

*

*

0.029

93

5

0.004

122

Mepanipyrim

18.07

222

0.001

*

*

0.013

92

8

0.002

123

Mepronil

19.54

269

0.002

*

*

0.023

87

10

0.005

124

Metalaxyl

15.95

206

0.003

92

10

0.028

97

5

0.002

125

Metaldehyde

8.87

89

0.005

*

*

0.05

111

62

0.021

126

Methacrifos

11.28

180

0.003

97

17

0.029

85

4

≤0.001

127

Methamidophos

7.75

141

0.026

36

24

0.258

47

15

0.005

128

Methidathion

17.82

145

0.003

81

20

0.03

101

5

0.001

129

Methiocarb

16.26

168

0.002

109

59

0.02

77

46

0.001

130

Methoxychlor

21.03

228

0.002

*

*

0.025

90

10

0.003

131

Metoprene

17.56

73

0.01

104

5

0.103

93

3

0.003

132

Mevinphos

10.36

192

0.003

104

16

0.03

99

1

≤0.001

133

Monocrotophos

13.43

192

0.046

84

8

0.456

88

7

0.021

134

Myclobutanil

18.66

150

0.006

*

*

0.055

97

5

0.012

135

Nuarimol

20.28

314

0.005

*

*

0.049

89

7

0.008

136

Omethoate

12.39

156

0.005

57

19

0.054

53

14

0.002

137

Oxadixyl

19.38

163

0.012

*

*

0.124

92

4

0.038

138

Oxydemeton-methyl (deg)

6.63

110

0.005

*

*

0.052

79

7

0.010

139

Paclobutrazole

18.11

238

0.007

197

28

0.07

90

6

≤0.001

140

Parathion

16.69

291

0.011

106

26

0.106

91

6

0.004

141

Parathion-methyl

15.71

263

0.002

88

7

0.021

94

2

≤0.001

142

Penconazole

17.35

248

0.003

90

10

0.03

94

4

≤0.001

143

Permethrin-cis

22.65

183

0.005

101

7

0.049

98

7

0.003

144

Permethrin-trans

22.77

183

0.001

*

*

0.011

98

7

0.001

145

Phenothrin-I

21.40

183

0.005

97

8

0.05

92

9

0.001

146

Phenothrin-II

21.51

123

0.005

93

6

0.05

93

10

0.004

147

Phenthoate

17.53

274

0.005

103

8

0.048

91

5

0.001

148

Phenylphenol, 2-

11.56

170

0.005

96

6

0.052

95

4

0.001

149

Phorate

13.56

260

0.005

98

6

0.05

92

5

0.001

150

Phosalone

21.61

182

0.001

117

5

0.009

101

5

≤0.001

151

Phosmet

20.90

160

0.005

123

16

0.052

100

4

≤0.001

152

Phosphamidon-I

14.75

127

0.011

93

16

0.105

90

3

0.002

153

Phosphamidon-II

15.49

127

0.011

89

9

0.105

91

2

0.005

154

Piperonyl butoxide

20.36

176

0.004

*

*

0.037

89

10

0.010

155

Pirimicarb

15.25

166

0.002

101

9

0.02

95

5

≤0.001

156

Pirimiphos-methyl

16.26

233

0.002

*

*

0.016

87

2

0.004

157

Prochloraz

22.97

180

0.004

*

*

0.038

101

6

0.007

158

Procymidone

17.68

285

0.003

104

15

0.029

91

7

0.001

159

Profenofos

18.42

337

0.005

97

8

0.052

95

10

0.001

160

Propargite

20.31

350

0.01

*

*

0.102

96

7

0.020

161

Propham

10.73

179

0.005

97

5

0.049

94

5

0.001

162

Propiconazole-I

19.89

259

0.014

92

5

0.141

89

9

0.003

163

Propiconazole-II

20.02

259

0.014

90

5

0.141

87

9

0.002

164

Propoxur

12.62

110

0.002

96

6

0.02

92

7

≤0.001

165

Propyzamide

14.58

175

0.005

76

39

0.046

99

2

0.001

166

Prothiofos

18.37

267

0.003

85

19

0.032

101

9

0.001

167

Pyrazophos

22.17

221

0.003

137

11

0.03

145

4

≤0.001

168

Pyrethrins

19.62

123

0.053

*

*

0.528

99

13

0.087

169

Pyridaben

22.82

147

0.005

96

9

0.051

94

3

0.001

170

Pyridaphenthion

20.80

199

0.005

99

10

0.048

93

5

0.003

171

Pyrifenox-I

17.39

262

0.011

84

7

0.106

95

6

0.003

172

Pyrifenox-II

14.68

264

0.011

*

*

0.106

90

6

0.170

173

Pyrimethanil

14.65

198

0.002

135

14

0.02

123

4

≤0.001

174

Pyriproxyfen

21.65

136

0.002

119

18

0.024

91

6

≤0.001

175

Quinalphos

17.55

146

0.004

70

9

0.041

87

8

0.002

176

Quinoxyfen

19.90

272

0.001

113

13

0.014

105

13

≤0.001

177

Quintozene

14.50

237

0.005

106

10

0.046

108

2

0.003

178

Simazine

16.17

201

0.004

91

9

0.039

95

7

0.002

179

Spiroxamine

15.67

198

0.018

99

17

0.176

81

2

0.009

180

TDE, o,p′-

18.67

235

0.003

99

5

0.028

95

4

≤0.001

181

TDE, p,p′-

19.36

235

0.001

86

10

0.014

90

7

≤0.001

182

Tebuconazole

20.28

250

0.009

*

*

0.089

91

9

0.031

183

Tebufenpyrad

21.12

171

0.005

92

17

0.052

87

7

0.001

184

Tecnazene

12.56

203

0.005

108

6

0.048

99

6

0.002

185

Teflubenzuron (deg)

8.12

197

0.003

174

25

0.025

124

25

0.002

186

Tefluthrin

14.91

197

0.001

*

*

0.014

89

14

0.002

187

Terbufos

14.46

231

0.005

100

8

0.052

95

3

≤0.001

188

Tetraconazole

16.85

336

0.003

95

3

0.026

88

6

≤0.001

189

Tetradifon

21.44

356

0.003

*

*

0.03

94

8

0.010

190

Thiometon

13.78

88

0.005

93

5

0.055

100

3

≤0.001

191

Tolclofos-methyl

15.80

265

0.001

91

6

0.01

102

5

≤0.001

192

Tolylfluanid

17.42

238

0.003

85

17

0.031

96

2

0.002

193

Triadimefon

16.75

208

0.007

90

14

0.065

97

6

0.005

194

Triadimenol

17.85

168

0.005

*

*

0.053

85

2

0.029

195

Triazamate

17.95

242

0.003

*

*

0.028

90

10

0.010

196

Triazophos

19.62

257

0.005

109

37

0.054

89

20

0.001

197

Trifloxystrobin

19.92

116

0.006

91

13

0.055

88

11

0.002

198

Triflumizole

17.70

278

0.007

102

15

0.066

80

15

0.001

199

Trifluralin

13.33

306

0.002

92

19

0.019

94

8

≤0.001

200

Vamidothion

17.95

87

0.019

*

*

0.187

100

5

0.045

201

Vinclozolin

15.71

198

0.005

97

16

0.047

93

7

0.003

aBenzoylurea(deg) = 2,4-difluorobenzamide

LOD: Amount for which S/N = 3, or in the event of an interfering peak, the average peak height for fortified sample (n = 5) should be 3.3 times the average peak height for control sample (n = 2)

*Fortification level below LOD as defined above

Underlined values are outside EU criteria for method validation

Fig. 4

GC–MS extracted-ion chromatograms obtained from lettuce with (upper traces) and without fortification with pesticides, and the corresponding mass spectra (upper, reference spectra; lower, background-corrected spectra from the sample). a, b, 0.005 mg kg−1 disulfoton (m/z 88); c, d, 0.002 mg kg−1 fipronil (m/z 367); e, f, 0.006 mg kg−1 biphenyl (m/z 154)

This initial validation clearly showed it is possible to introduce 10 mg of matrix equivalent of generic extracts obtained after ethyl acetate extraction of leafy vegetables. Adequate quantitative data are obtained for most of the pesticides at levels of 0.01 mg kg−1 or even below. Detection limits were usually well below 0.01 mg kg−1 after full-scan acquisition with a single-quadrupole MS. This means that for most pesticides at the target LOQ of 0.01 mg kg−1 (i.e. the lowest maximum residue limit set in the EU for vegetables and fruit), the signal-to-noise ratio is adequate for reliable automatic integration of peaks and that confirmation of identity of the pesticide is possible from its mass spectrum or at least one or two other diagnostic ions.

Pesticides that did not meet the EU criteria for quantitative analysis, and/or for which relatively high LODs were obtained, included many compounds known to be troublesome in GC analysis because of to their high polarity or thermal lability. Typical examples are acephate, cyromazine, dicofol (screened for as its degradation product dichlorobenzophenone), dimethoate, imazalil, metaldehyde, methamidophos, methiocarb, omethoate, and the benzoylureas (measured as one common and one compound-specific degradation product). The relatively low recovery of the polar organophosphorus pesticides (acephate, methamidophos, and omethoate) can be attributed to the GC measurement and not to poor extraction efficiency, as was apparent from LC–MS–MS analysis of samples using the same extraction technique (see section LC–MS–MS analysis). For several other polar or labile pesticides adequate quantitative data were obtained during this initial validation, but from previous experience and the results obtained after implementation of the method it was clear that for such compounds LC–based analysis is more robust than GC–MS analysis. Typical examples include carbaryl, carbofuran, clofentezin, monocrotophos, and oxydemeton-methyl.

Analytical quality-control data from routine GC–MS analysis

The initial validation data are continuously being supplemented by performance data generated as part of the analytical quality-control during routine analysis of the samples, to gain insight into reproducibility, robustness, recovery, and selectivity with other matrices. For this, with each analytical batch, one of the samples submitted for routine analysis was spiked with 135 pesticides at five times the target LOQ level (i.e. samples were spiked with 0.05 mg kg−1 of most of the pesticides). A compilation was made of recovery data from a period of 15 months which included analysis of approximately 100 different vegetable and fruit commodities. Given the wide variety of commodities, matrix-matched calibration is quite tedious and would substantially increase the number of standard solutions to be analyzed in the GC sequence. It was therefore decided to select one relatively simple matrix (tomato) as default for matrix-matched calibration, i.e. recoveries for all commodities were calculated against the tomato-matrix standard. For each pesticide, calculations were performed for two diagnostic ions. All together this resulted in approximately 30,000 values.

According to the current EU guideline on quality control in pesticide residue analysis [37], the recovery obtained during routine analysis should be within 60–140%. An overview of the percentage of recovery values within or outside the 60–140% criterion for a wide variety of matrices is presented in Table 6. With such large number of pesticides (or, actually, diagnostic ions) and matrices, one failing combination or more occurred for most matrices. There are several causes for this. Main reasons for recovery below 60% could be poor extraction efficiency or incomplete transfer of the pesticides to the GC column (e.g. adsorption and/or degradation in a contaminated inlet). Higher recovery may occur when a compound from the matrix generates the same diagnostic ion as a pesticide and co-elutes with that pesticide (i.e. detection was not selective). Another reason could be that the matrix effect induced in the GC inlet [50] for a pesticide in a particular matrix is more pronounced than that in the tomato-based calibration standard.
Table 6

Overview of percentage of recovery valuesa within or outside the EU 60–140% criterion [37] after GC–MS analysis

 

Matrix

Percentage of all recovery valuesa

60–140%

<60%

>140%

1

Beetroot

100

0

0

2

Cucumber (1/2)

100

0

0

3

Mint (1/2)

100

0

0

4

Sharonfruit (1/2)

100

0

0

5

Witloof

100

0

0

6

Asparagus

99

1

0

7

Bean sprouts

99

0

1

8

Corn syrup

99

0

0

9

Fennel leaves

99

0

1

10

Grape

99

0

1

11

Kohlrabi (1/3)

99

1

0

12

Lima bean

99

0

1

13

Pak choi (1/2)

99

0

1

14

Pear concentrate

99

0

1

15

Pumpkin

99

0

1

16

Salsify

99

0

0

17

Sharonfruit (2/2)

99

0

1

18

Strawberry

99

0

1

19

Sugar pea

99

1

0

20

Taro

99

0

1

21

Bitter cucumber

98

0

2

22

Cucumber (2/2)

98

1

1

23

Egg plant

98

0

2

24

Kidney bean

98

1

1

25

Kohlrabi (2/3)

98

1

1

26

Mushroom

98

0

2

27

Pineapple

98

1

1

28

Sweet pepper

98

0

2

29

Tomato puree (processed)

98

0

2

30

Turnip

98

1

0

31

Turnip tops (1/2)

98

0

2

32

Alfalfa

97

1

2

33

Cauliflower

97

1

2

34

Cherry

97

0

3

35

Chestnut

97

2

1

36

Endive

97

0

3

37

Fig

97

0

3

38

Kangkung (1/2)

97

1

2

39

Kangkung (2/2)

97

2

1

40

Ladies’ fingers

97

0

3

41

Mango

97

0

3

42

Pear puree (processed)

97

0

3

43

Sorrel

97

3

0

44

Soybean sprouts

97

0

3

45

Asparagus bean

96

1

3

46

Orange

96

2

2

47

Potato leaves

96

2

2

48

Rhubarb

96

2

2

49

Artichoke

95

0

5

50

Tangelo

95

2

3

51

Tarrragon

95

3

2

52

Wine (red)

95

1

4

53

Apricot

94

0

6

54

Chives (1/3)

94

3

3

55

Chives (2/3)

94

4

2

56

Dill leaves

94

4

2

57

Melon puree (processed)

94

1

5

58

Mineola

94

1

6

59

Pak choi (2/2)

94

2

4

60

Sugar water

94

6

0

61

Broad bean

93

1

6

62

Celery leaves (1/4)

93

3

4

63

Chervil

93

5

2

64

Dates

93

7

0

65

Sweetcorn (1/3)

93

4

3

66

Carrot

92

1

7

67

Haricot bean

92

0

8

68

Oregano

92

5

3

69

Parsnip

92

2

6

70

Fennel

91

0

9

71

Green pea (1/2)

91

4

5

72

Passion fruit (1/2)

91

2

7

73

Celery leaves (2/4)

90

6

4

74

Green pea (2/2)

90

1

9

75

Lemon puree

90

8

2

76

Mint (2/2)

90

5

5

77

Pomegranate

90

1

9

78

Purslane

90

1

9

79

Water cress

90

2

8

80

Lettuce

89

7

4

81

Chili pepper (1/2)

88

6

6

82

Chinese cabbage

87

0

13

83

Passion fruit (2/2)

87

3

10

84

Bamboo shoots

86

0

14

85

Celery leaves (3/4)

86

7

7

86

Honey

86

14

0

87

Potato puree (processed)

86

14

0

88

Sugar pea

85

0

15

89

Turnip tops (2/2)

85

0

15

90

Lime

84

4

12

91

Blueberry

83

2

16

92

Potato

83

15

2

93

Celery leaves (4/4)

82

3

15

94

Green pea

82

1

17

95

Apple pulp (processed)

81

6

13

96

Cassava

81

9

10

97

Chives (3/3)

81

7

12

98

Kohlrabi (3/3)

78

0

22

99

Parsley (1/2)

78

6

16

100

Thyme (1/3)

78

2

20

101

Kale

77

6

17

102

Chili pepper (2/2)

76

15

9

103

Coriander leaves

76

18

6

104

Sweetcorn (2/3)

75

18

7

105

Sweetcorn (3/3)

74

9

17

106

Parsley (2/2)

73

20

7

107

Thyme (2/3)

73

3

24

108

Rocket

72

3

25

109

Thyme (3/3)

66

29

5

110

Golden berry (physalis)

65

1

34

aRecoveries at 0.05 mg kg−1 (0.10–0.30 mg kg−1 for 22 pesticides). Calculated for 135 pesticides, two diagnostic ions each, against a standard prepared in blank tomato extract. The pesticides included are listed in Table 7

Failing pesticide–matrix combinations were most abundant for herbs, kale, sweetcorn, and golden berry, for which up to 35% of recovery values (calculated using the two diagnostic ions for each pesticide) were outside the 60–140% range. These products contain larger amounts of co-extractants than most other vegetables and fruits, which may result in insufficient detection selectivity, enhanced response as a result of a matrix effect (more shielding of active sites in the inlet), and contamination of the inlet. For this type of product more selectivity, e.g. by use of MS–MS would be beneficial. Such detection is also more sensitive than single quadrupole full-scan detection and would enable reduction in the amount of matrix introduced, thus reducing build up of contamination. Overall, when data for all 110 QC samples were included, recovery was acceptable for 91% of the diagnostic ions measured.

On the basis of the same data, an overview by pesticide is presented in Table 7. For each pesticide two diagnostic ions from the full-scan data were integrated and concentrations were calculated. In routine practice, however, the most convenient way of reviewing the data is by using one and the same diagnostic ion for each pesticide, irrespective the matrix. On the basis of the data set obtained (nearly 14,700 pesticide–matrix combinations) the most favorable of the two diagnostic ions, i.e. the ion for which the highest number of recoveries within 60–140% was obtained, was assigned as the Quan ion (default quantification ion). By using this ion, acceptable recoveries were obtained for 93% of pesticides–matrix combinations. This also means that 7% or, in absolute figures, 1008 of the pesticide–matrix combinations did not meet the criterion. 40% of these failing combinations could be accepted after use of the alternative ion, for which calculations were also performed automatically during data processing. Low recoveries (<60%) for both diagnostic ions were obtained for 2.7% of pesticide–matrix combinations. High recoveries (>140%) were obtained for 2% of the combinations. For this latter group manual evaluation of other ions, if available and sufficiently abundant, could further increase the number of acceptable recoveries. Because this is a time-consuming process, it was not done routinely. In the event of deviating recovery, assessment of the results to be reported was based on visual evaluation of the extracted ion chromatograms of the two diagnostic ions at least. On the basis of on the findings it was then concluded the pesticide could not be determined in that specific matrix, or only at higher levels.
Table 7

Recoveries over all matrices (GC–MS analysis)

Pesticide

Quan. ion m/z

Qual. ion m/z

Fortification level (mg kg−1)

# QCs matrices (see Table 6)

Both diagn. ions 60–140%

One of diagn. ions 60–140%

Both diagn. ions >140%

Both diagn. ions <60%

Average recov. (%) Quan. ion

RSD (%)

Acrinathrin

208

289

0.10

110

107

107

3

0

97

16

Azaconazole

173

217

0.05

110

107

107

2

1

97

14

Azoxystrobin

388

344

0.05

108

97

102

0

8

96

15

Benalaxyl

206

148

0.05

110

108

109

0

1

100

13

Bifenthrin

181

166

0.05

109

109

110

0

0

102

13

Biphenyl

154

153

0.05

110

93

94

7

9

98

20

Boscalid

112

140

0.13

109

98

100

2

8

96

16

Bromopropylate

341

343

0.05

110

100

101

9

0

109

14

Bromuconazole

295

173

0.05

110

100

105

4

1

102

18

Bupirimate

273

208

0.02

110

108

109

0

1

96

15

Buprofezin

172

105

0.05

109

105

108

2

0

102

12

Cadusafos

158

159

0.05

110

105

107

1

2

104

13

Chlorfenapyr

364

328

0.04

110

103

106

2

2

102

16

Chlorfenvinphos

323

267

0.05

110

103

103

7

0

103

16

Chlorpropham

213

127

0.05

108

101

106

2

2

105

14

Chlorpyrifos

314

286

0.05

109

107

109

0

1

101

14

Chlorpyrifos-methyl

288

286

0.05

108

101

104

4

2

102

16

Chlorthal-dimethyl

332

301

0.05

110

110

110

0

0

101

14

Cinerin-1

123

150

0.11

110

104

105

4

1

101

15

Cyfluthrin

226

199

0.20

110

102

106

0

4

100

17

Cyhalothrin, lambda-

208

181

0.05

108

104

109

1

0

99

16

Cypermethrin

163

181

0.15

105

99

107

2

0

102

14

Cyproconazole

222

224

0.05

110

103

105

1

4

102

16

Cyprodinil

224

225

0.05

109

101

102

0

8

85

15

DDE, p,p′-

246

318

0.06

110

110

110

0

0

101

13

DDT, o,p′-

235

237

0.05

110

106

107

2

1

103

14

DDT, p,p′-

237

235

0.05

110

82

90

9

11

98

20

Deltamethrin

253

255

0.10

110

91

98

4

8

95

17

Diazinon

179

137

0.05

109

108

110

0

0

101

13

Dichlorvos

185

109

0.05

110

90

96

8

6

99

20

Dicloran

206

160

0.05

108

96

102

3

5

99

15

Dieldrin

263

79

0.05

110

109

109

0

1

104

14

Diethofencarb

168

267

0.05

110

107

108

1

1

100

15

Difenoconazole

323

265

0.10

107

101

106

0

4

96

16

Dimethipin

118

76

0.05

110

95

104

5

1

104

16

Dimethomorph

387

301

0.10

110

98

100

0

10

89

16

Dimoxystrobin

205

116

0.05

110

108

109

0

1

100

12

Diniconazole

270

268

0.15

64

58

62

1

1

97

17

Diphenylamine

169

167

0.05

110

107

107

0

3

101

16

Dodemorph

238

154

0.05

110

109

109

0

1

96

15

Endosulfan-alpha

195+241

239+197

0.50

110

95

100

10

0

107

12

Endosulfan-beta

195+241

237+160

0.10

110

107

107

3

0

102

14

Endosulfan-sulfate

272+229

274+237

0.05

109

102

107

2

1

104

16

EPN

157

323

0.05

110

103

106

3

1

103

17

Epoxiconazole

192

138

0.05

110

106

108

1

1

98

14

Esfenvalerate

167

125

0.15

110

102

103

4

3

106

15

Ethion

231

153

0.05

110

106

106

4

0

103

14

Ethoprophos

158

200

0.05

110

107

108

1

1

104

13

Etofenprox

376

164

0.05

110

102

104

2

4

97

15

Etridiazole

211

183

0.05

109

80

82

21

7

97

21

Fenarimol

219

139

0.05

110

106

108

1

1

103

16

Fenazaquin

160

145

0.05

110

105

105

1

4

88

16

Fenbuconazole

129

198

0.05

110

105

107

1

2

99

17

Fenitrothion

277

260

0.05

108

99

102

7

1

106

16

Fenoxycarb

186

116

0.05

110

89

101

8

1

105

17

Fenpiclonil

238

174

0.05

110

101

106

3

1

102

17

Fenpropathrin

181

141

0.05

109

101

104

6

0

103

13

Fenpropimorph

128

129

0.05

110

108

109

1

0

101

14

Fenvalerate

167

125

0.25

110

102

103

2

5

98

15

Fipronil

367

369

0.05

110

101

100

3

7

99

18

Flucythrinate

199

157

0.05

110

102

106

3

1

103

15

Fludioxonil

248

182

0.05

109

105

107

1

2

98

17

Flusilazole

233

206

0.05

110

104

107

1

2

97

15

Flutolanil

323

281

0.05

110

107

109

1

0

100

13

Flutriafol

219

123

0.04

110

102

104

5

1

103

14

Fluvalinate, tau-

250

252

0.15

110

97

99

5

6

99

15

Furalaxyl

242

95

0.05

110

106

107

3

0

101

13

Heptenophos

124

126

0.05

109

97

104

6

0

102

18

Hexaconazole

216

214

0.05

110

106

108

1

1

102

14

Iprodione

316

314

0.10

103

79

88

8

13

100

20

Jasmolin-1

164

123

0.04

110

92

104

4

2

97

15

Kresoxim-methyl

116

206

0.05

109

106

109

0

1

100

15

Lindane

183

219

0.05

110

107

110

0

0

99

15

Malathion

173

127

0.05

108

103

107

3

0

104

17

Mecarbam

329

131

0.05

110

109

110

0

0

101

15

Mepanipyrim

223

222

0.05

110

88

91

7

12

85

19

Mepronil

269

119

0.10

110

109

110

0

0

97

15

Metalaxyl

206

160

0.05

107

105

108

2

0

103

12

Methidathion

145

85

0.05

109

85

89

19

2

107

15

Metrafenone

395

393

0.05

110

104

106

2

2

94

14

Mevinphos

192

127

0.05

110

88

90

17

3

104

17

Myclobutanil

179

150

0.05

110

102

107

2

1

98

15

Nitrothal-isopropyl

236

254

0.05

110

108

108

1

1

99

13

Nuarimol

235

203

0.05

110

108

110

0

0

101

15

Oxadixyl

163

132

0.15

110

106

107

1

2

99

13

Parathion

291

109

0.05

110

105

109

1

0

105

15

Parathion-methyl

263

247

0.05

109

86

102

8

0

107

17

Penconazole

159

248

0.05

109

108

110

0

0

100

15

Pentachloroaniline

267

265

0.11

110

96

97

0

13

81

15

Pentachlorothioanisole

296

246

0.05

110

87

89

0

21

77

16

Permethrin-cis

183

163

0.05

110

108

110

0

0

101

14

Permethrin-trans

183

163

0.05

110

106

107

3

0

100

13

Phenylphenol, 2-

170

141

0.05

109

102

107

3

0

98

13

Phosalone

182

184

0.05

110

90

92

13

5

101

19

Phosmet

161

160

0.05

109

76

90

16

4

100

22

Phosphamidon

264

127

0.05

110

91

94

13

3

103

19

Picoxystrobin

335

145

0.05

110

105

109

1

0

103

12

Piperonyl-butoxide

176

177

0.05

107

106

109

1

0

100

13

Pirimiphos-methyl

276

305

0.05

110

109

109

1

0

102

13

Procymidone

283

285

0.05

108

106

108

1

1

100

14

Profenofos

337

206

0.05

108

93

102

8

0

104

17

Propargite

173

135

0.33

109

104

109

1

0

103

16

Propiconazole

259

261

0.05

109

106

107

2

1

99

14

Propyzamide

173

175

0.05

110

107

108

2

0

102

12

Prothiofos

309

267

0.05

110

108

109

1

0

99

13

Pyrazophos

221

232

0.05

110

99

99

3

8

91

18

Pyrethrins

123

160

0.36

110

87

103

7

0

105

18

Pyridaben

147

148

0.05

110

107

107

1

2

99

14

Pyridaphenthion

340

199

0.05

110

96

101

7

2

102

17

Pyrifenox

262

264

0.05

110

108

110

0

0

100

15

Pyrimethanil

199

198

0.05

110

107

106

1

3

90

14

Pyriproxyfen

226

136

0.05

110

104

107

2

1

103

16

Quinalphos

157

146

0.05

110

104

105

4

1

104

14

Quinoxyfen

307

272

0.05

110

106

106

0

4

92

14

Quintozene

237

142

0.05

110

107

107

1

2

93

16

Silafluofen

179

286

0.05

110

106

106

0

4

98

14

Spirodiclofen

312

314

0.25

110

95

96

6

8

96

19

Spiromesifen

272

254

0.05

110

105

108

1

1

96

16

Spiroxamine

100

198

0.10

110

107

109

0

1

96

13

TDE, p,p′-

235

237

0.05

110

97

100

5

5

103

14

Tebuconazole

250

252

0.15

67

66

67

0

1

97

15

Tebufenpyrad

171

318

0.05

110

107

108

1

1

100

13

Tebupirimfos

234

318

0.05

110

108

109

1

0

101

14

Tefluthrin

177

197

0.05

110

106

107

3

0

103

13

Tetraconazole

336

338

0.05

110

109

109

1

0

99

14

Tetradifon

356

229

0.15

109

109

110

0

0

99

14

Thiometon

88

125

0.05

110

108

110

0

0

104

15

Tolclofos-methyl

265

267

0.05

108

107

107

2

0

101

13

Tri-allate

268

270

0.05

110

104

105

4

1

104

13

Triazamate

242

227

0.05

110

107

107

3

0

102

14

Triazophos

285

257

0.05

109

95

100

8

2

104

18

Trifloxystrobin

131

116

0.05

110

108

109

1

0

103

14

Triflumizole

278

287

0.03

110

105

107

0

3

99

15

Trifluralin

264

306

0.05

110

107

107

2

1

101

14

Vinclozolin

212

198

0.05

107

106

109

1

0

103

11

Total

   

14696

13688

14057

402

300

  

% of # QCs

    

93.1

95.2

2.7

2.0

  

It should be noted that the above evaluation applies to a level five times the reporting level, which was set at 0.01 mg kg−1, or the LOQ if higher than 0.01 mg kg−1. At lower levels interferences may have a larger effect and, consequently, more frequent deviations from the 60–140% criterion (most probably >140%) may be observed. For higher levels, the opposite would be true.

Pesticides for which low recoveries (<60%) were frequently obtained (10–21 of 110 QC samples) included iprodione and p,p′-DDT (degradation in inlet), dimethomorph (polar, relatively non-volatile, could be troublesome in splitless transfer), pentachloroanisole, pentachloroaniline, and mepanipyrim (no clear explanation, but probably related to the dispersive SPE clean-up). There were no indications for poor extraction efficiency.

High recovery (>140%) frequently occurred for etridiazole, methidathion, mevinphos, phosmet, phosalone, phosphamidone, and endosulfan-alpha (10–21 times out of 110 QC samples, often in herbs and peas). This was attributed to matrix effects and interferences.

Overall, the pesticides that failed most frequently (11–28 times out of 110) during routine analytical quality control were (in descending order) etridiazole, iprodione, methidathion, pentachlorothioanisole, mevinphos, phosmet, p,p′-DDT, mepanipyrim, phosalone, phosphamidon, biphenyl, dichlorvos, spirodiclofen, pentachloroaniline, deltamethrin, tau-fluvalinate, and pyrazophos. These would be the most relevant for inclusion in alternative methods, for example GC–MS–MS or LC–MS–MS.

Average recovery and RSD were calculated for pesticide–matrix combinations that passed the acceptable recovery criterion. The results are included in Table 7. Average recovery was usually close to 100% and RSDs approximately 15%. For the pesticides known to be adsorbed by GCB systematically lower average recovery (77–90%) was obtained, which is in agreement with the results obtained during method development.

These comprehensive data show that with a relatively inexpensive single-quadrupole MS detector in full-scan mode it is possible to obtain reliable quantitative data down to the 0.01 mg kg−1 level, or even lower, for a wide range of pesticides in a wide variety of matrices after generic rapid sample preparation based on extraction with ethyl acetate. Unified calibration based on a tomato-matrix standard is, furthermore, a feasible approach. One should, however, be aware there are also limitations and that some pesticide–matrix combinations cannot be determined in the 0.01–0.1 mg kg−1 range, and that for other pesticides calibration against the corresponding matrix instead of tomato is required to bring quantitative results within the AQC criteria, especially for MRL violations, when more stringent criteria apply. The data also reveal that the only way to gain full insight into analyte recovery and method selectivity with a wide variety of matrices is by performing analytical quality control on all pesticides which are reported, rather than on a subset, as is suggested in the EU guideline [37]. A subset will suffice for demonstration of adequate sample preparation and injection but will not reveal limitations in the selectivity of GC–MS.

GC single-quadrupole MS remains an effective tool for routine GC analysis of pesticide residues. For many vegetable and fruit matrices there is no real need to change to more advanced (and expensive) MS techniques, for example MS–MS (which has limited scope) or accurate mass TOF-MS (which has a limited dynamic range). Use of such equipment would be justified for more complex matrices and when low μg kg−1 LOQs are required—for example analysis of some pesticides in baby food.

LC–MS–MS analysis

Clean-up

The ethyl acetate extraction procedure is also appropriate for many pesticides not amenable to GC analysis [11, 15, 16, 18, 26]. Typically no clean-up is performed (Table 1). One reason for this is that with regard to chromatographic performance LC columns tend to be more tolerant of injection of bulk matrix than GC columns. In our experience, continual injection of 20 mg equivalent of vegetable and fruit extracts does not result in deterioration of chromatographic performance or unacceptable contamination of the ion source (the system used here was an API2000). In LC–MS co-extracted matrix does have an effect on the response, however, by interfering with the ionization process. This results in suppression (sometimes enhancement) of the response to a pesticide in a matrix compared with that in a solvent standard [51] and complicates quantification of pesticides in the samples. The possibility of reducing matrix effects by use of dispersive SPE clean-up was investigated in a similar way as for GC. First, the effectiveness of the clean-up step was investigated by addition of 25 mg GCB and 25 mg PSA to 1 mL raw extract of a mixed spinach–grape–onion sample (1:1:1, 1 g ml-1). Seventy pesticides (the ones in Table 8 with API2000 in the MS-MS column) were added after clean-up and analyzed by LC–MS–MS. The response was compared with that of solutions of equal concentration in the raw extract and a solvent standard. Clean-up increased the number of pesticides for which no pronounced matrix effect (less than 20% suppression or enhancement) was observed from 38 to 84%. Several of the pesticides (Tables 2 and 4) were adsorbed by the SPE material, however. Although adsorption by the GCB could have been avoided or reduced by addition of toluene (although less practical when changing from extraction solvent to methanol/water), it was concluded that PSA was not compatible with a generic method for pesticides amenable to LC–MS–MS. It was therefore decided not to include a clean-up step for LC–MS–MS analysis and to use the initial raw ethyl acetate extract. Another reason for not further pursuing clean-up in LC–MS–MS analysis was that the sensitivity of current triple-quadrupole instruments enables injection of only small amounts of matrix into the LC–MS–MS system (e.g. 2 mg) while still achieving the desired limits of quantification. Experiments showed that tenfold dilution of 1 g mL−1 extracts increased the number of pesticides for which no pronounced matrix effect occurred from 65 to 82% and from 10 to 65% for cucumber and cabbage, respectively.
Table 8

LC–MS–MS settings and performance-validation characteristics

Pesticide

tr (min)

Precurs.

Prod. ion 1

DP

FP

CE

CXP

Prod. ion 2

CE

CXP

Vegetables

n

0.01 mg kg−1

0.1 mg kg−1

Fruits

n

0.01 mg kg−1

0.1 mg kg−1

MS–MS

Matrix

Rec. (%)

RSD (%)

Rec. (%)

RSD (%)

Matrix

Rec. (%)

RSD (%)

Rec. (%)

RSD (%)

Abamectinea,c

21.7

891

305

46

340

33

22

145

49

10

Cuc/lett

4

66

18

68

15

Apple/grape

4

159

39

155

46

API2000

Acephate

5.5

184

143

31

150

11

12

95

33

6

Cuc/lett

4

80

21

75

9

Apple/grape

4

80

8

76

10

API2000

Acetamiprid

10.5

223

126

91

270

29

10

177

11

14

Cuc/lett

4

99

3

96

7

Apple/grape

4

117

4

98

6

API2000

Aldicarba

11.7

208

116

16

110

11

8

89

21

6

Cuc/lett

4

103

20

91

12

Apple/grape

4

99

13

109

13

API2000

Aldicarbsulfon

7.9

223

86

32

200

21

12

148

13

6

Cuc/lett

4

104

9

83

4

Apple/grape

4

120

5

91

3

API2000

Aldicarbsulfoxide

7.2

207

132

46

300

9

10

89

19

6

Cuc/lett

4

109

12

89

4

Apple/grape

4

109

4

86

3

API2000

Asulam

3.6

231

156

41

260

15

12

92

33

6

Cuc/lett

4

35

28

29

38

Apple/grape

4

13

23

10

30

API2000

Azamethiphos

12.1

325

183

36

220

23

14

112

53

8

Cuc/lett

4

101

6

94

4

Apple/grape

4

106

11

91

10

API2000

Azinfos-methyl

13.5

318

132

41

60

23

6

160

15

6

Lettuce

5

93

16

88

4

Orange

5

69

11

79

9

API3000

Bendiocarb

12.2

224

167

16

100

13

10

109

25

18

Lettuce

5

102

10

96

6

Orange

5

86

8

108

6

API3000

Bifenazate

13.9

301

198

16

110

13

16

170

27

14

Lettuce

5

35

9

33

7

Orange

5

93

7

83

5

API3000

Bitertanol

15.2

338

269

21

120

13

20

99

21

8

Lettuce

5

96

10

81

7

Orange

5

93

10

81

8

API3000

Butocarboximb

11.6

213

75

41

300

21

4

156

17

12

Lettuce

5

101

23

93

12

Orange

5

72

16

89

19

API3000

Butoxycarboxim

7.7

223

106

36

250

13

8

166

11

10

Cuc/lett

4

116

9

104

15

Apple/grape

4

118

4

95

7

API2000

Carbaryl

12.5

202

145

101

370

13

12

127

37

10

Cuc/lett

4

100

4

95

4

Apple/grape

4

111

12

100

10

API2000

Carbendazim

11.4

192

160

46

230

23

12

132

43

10

Cuc/lett

4

104

2

102

10

Apple/grape

2

122

1

105

1

API2000

Carbofuran

13.3

222

165

46

290

17

12

123

29

10

Cuc/lett

4

124

12

111

13

Apple/grape

4

104

11

93

4

API2000

Carbofuran, 3-OH

10.4

238

220

31

210

9

16

163

19

12

Lettuce

5

91

8

94

4

Orange

5

100

7

91

6

API3000

Carboxin

12.6

236

143

11

350

21

2

93

51

2

Lettuce

5

87

9

80

2

Orange

5

89

9

84

6

API3000

Chlorbromuron

14.0

295

206

41

350

27

12

182

25

4

Lettuce

5

100

19

86

5

Orange

5

83

30

84

6

API3000

Chlorfluazuron

18.1

542

385

40

270

29

30

158

29

12

Lettuce

5

79

7

89

5

Orange

5

74

21

86

8

API3000

Clofentezin

15.5

303

138

51

280

21

10

102

61

8

Cuc/lett

4

93

17

76

10

Apple/grape

4

127

24

101

16

API2000

Clomazone

13.6

240

125

31

190

25

8

89

67

6

Lettuce

5

97

4

104

6

Orange

5

90

5

89

9

API3000

Clothianidin

10.2

250

132

36

70

23

10

169

17

10

Lettuce

5

99

11

100

2

Orange

5

110

4

100

3

API3000

Cycloxydim

14.9

326

280

46

260

19

22

180

29

14

Cuc/lett

4

18

118

82

9

Apple/grape

4

38

45

70

28

API2000

Cymoxanil

11.1

199

128

18

120

13

10

111

25

8

Cuc/lett

4

83

13

95

7

Apple/grape

4

90

8

99

2

API2000

Cyromazine

7.1

167

85

40

240

26

6

125

25

10

Cuc/lett

4

96

10

78

11

Apple/grape

4

96

7

81

3

API2000

Demeton

13.6

259

89

26

180

13

6

198

11

16

Lettuce

5

97

14

85

4

Orange

5

76

15

76

9

API3000

Demeton-S-methyl

12.5

231

89

31

50

21

4

61

37

4

Lettuce

5

93

5

86

4

Orange

5

81

6

81

8

API3000

Dem-S-meth-sulfone

8.8

263

169

41

350

23

6

109

41

4

Lettuce

5

104

12

92

6

Orange

5

100

2

97

4

API3000

Desmedipham

13.1

301

182

51

340

13

14

154

25

12

Cuc/lett

4

86

10

88

3

Apple/grape

4

95

22

83

15

API2000

Diafenthiuron

18.1

385

329

41

260

27

22

278

45

18

Lettuce

5

0

0

Orange

5

104

9

92

7

API3000

Dichlofluanidec

14.1

333

224

46

270

17

18

123

37

8

Cuc/lett

4

21

116

36

116

Apple/grape

4

33

82

54

68

API2000

Dicrotophos

9.5

238

112

41

270

17

8

193

13

16

Cuc/lett

4

110

5

99

3

Apple/grape

4

100

12

93

10

API2000

Diflubenzuron

14.5

311

158

46

270

19

12

141

47

10

Cuc/lett

4

79

15

84

1

Apple/grape

4

101

6

102

12

API2000

Dimethirimol

13.1

210

71

51

290

45

4

98

37

8

Lettuce

5

99

7

97

5

Orange

5

91

10

105

5

API3000

Dimethoate

10.6

230

199

11

350

13

4

125

29

2

Lettuce

5

98

7

96

4

Orange

5

109

17

95

6

API3000

Diniconazole

15.6

326

70

56

310

63

14

159

45

16

Lettuce

5

78

10

93

6

Orange

5

94

16

96

5

API3000

Disulfotonc

15.7

275

89

11

90

27

6

61

41

10

Lettuce

5

53

6

64

7

Orange

5

85

16

86

4

API3000

Disulfoton-sulfone

12.8

307

97

31

150

39

8

153

17

14

Lettuce

5

113

10

105

7

Orange

5

81

6

106

8

API3000

Disulfoton-sulfoxide

12.8

291

185

26

140

17

16

213

15

14

Lettuce

5

111

10

115

6

Orange

5

92

5

101

5

API3000

Diuron

13.3

233

72

36

210

37

4

46

35

6

Lettuce

5

111

6

101

7

Orange

5

94

7

94

6

API3000

DMSA

11.6

201

92

26

150

25

6

137

13

10

Lettuce

5

102

13

97

4

Orange

5

85

13

87

7

API3000

DMST

12.3

215

106

26

160

21

8

151

13

10

Lettuce

5

97

5

95

6

Orange

5

84

13

85

5

API3000

Ethiofencarb

12.8

226

107

36

220

21

8

169

9

14

Cuc/lett

4

81

30

94

5

Apple/grape

4

99

17

94

20

API2000

Ethiofencarbsulfon

9.7

258

107

36

240

21

6

201

11

16

Cuc/lett

4

120

10

105

5

Apple/grape

4

101

8

97

11

API2000

Ethiofencarbsulfoxide

9.9

242

107

31

180

23

8

185

13

14

Cuc/lett

4

114

13

97

2

Apple/grape

4

127

10

107

7

API2000

Ethirimol

13.3

210

140

51

370

31

12

98

37

6

Cuc/lett

4

96

3

88

6

Apple/grape

4

86

26

81

26

API2000

Famoxadonea

14.6

392

331

11

130

15

22

238

25

18

Lettuce

5

90

15

80

1

Orange

5

88

9

80

1

API3000

Fenamiphos

14.5

304

217

41

350

29

4

234

21

4

Lettuce

5

87

8

87

4

Orange

5

93

7

93

5

API3000

Fenamiphos-sulfone

12.2

336

308

81

360

23

22

266

29

20

Lettuce

5

102

8

94

5

Orange

5

81

16

86

8

API3000

Fenamiphos-sulfoxide

12.1

320

171

56

230

27

14

233

35

14

Lettuce

5

114

10

94

4

Orange

5

97

8

108

5

API3000

Fenhexamid

14.2

302

97

51

290

35

8

55

59

8

Lettuce

5

84

15

82

4

Orange

5

85

6

84

5

API3000

Fenpyroximate

19.3

422

366

61

360

21

26

135

43

10

Cuc/lett

4

98

8

95

9

Apple/grape

4

111

9

104

10

API2000

Fensulfothione

13.0

309

281

46

260

21

22

253

25

18

Lettuce

5

96

7

89

3

Orange

5

101

23

83

8

API3000

Fensulfothion-sulfone

13.0

325

269

36

120

21

18

191

33

12

Lettuce

5

103

10

98

8

Orange

5

85

6

100

6

API3000

Fenthion

13.9

279

231

26

130

21

16

   

Lettuce

5

111

31

81

8

Orange

5

38

22

74

8

API3000

Fenthion-sulfone

12.5

311

125

51

320

29

8

279

25

22

Lettuce

5

95

6

90

4

Orange

5

101

1

89

6

API3000

Fenthion-sulfoxide

12.4

295

280

46

230

25

20

109

45

8

Lettuce

5

93

2

94

6

Orange

5

94

8

87

6

API3000

Fipronil

14.1

437

368

66

370

23

26

290

37

16

Lettuce

5

70

24

88

11

Orange

5

92

28

90

12

API3000

Flucycloxuron

17.3

484

289

66

360

15

20

132

49

10

Cuc/lett

4

113

4

104

3

Apple/grape

4

163

38

121

26

API2000

Flufenoxuron

17.1

489

158

101

360

27

12

141

65

10

Cuc/lett

4

107

17

90

8

Apple/grape

4

172

50

102

8

API2000

Formetanate

12.2

222

165

36

190

19

14

120

37

8

Lettuce

5

100

14

103

6

Orange

5

95

6

95

7

API3000

Fosthiazate

12.7

284

104

31

200

23

6

228

15

22

Lettuce

5

99

8

102

6

Orange

5

84

2

98

6

API3000

Furathiocarb

16.5

383

195

76

370

25

16

252

19

18

Cuc/lett

4

55

32

55

38

Apple/grape

4

87

17

84

7

API2000

Hexaflumuronc

15.2

461

158

51

300

27

10

141

61

10

Cuc/lett

4

91

24

82

7

Apple/grape

4

171

15

114

16

API2000

Hexythiazox

17.4

353

168

41

270

35

12

228

21

18

Cuc/lett

4

99

19

84

15

Apple/grape

4

120

26

84

11

API2000

Hymexazolc

5.8

100

54

66

360

21

4

44

29

2

Cuc/lett

4

76

34

50

49

Apple/grape

4

45

15

22

20

API2000

Imazalil

15.0

297

159

46

290

33

12

201

29

16

Cuc/lett

4

90

4

76

12

Apple/grape

4

111

7

90

13

API2000

Imidacloprid

10.0

256

175

41

240

25

14

209

21

18

Cuc/lett

4

99

9

81

11

Apple/grape

4

121

12

89

7

API2000

Indoxacarb

15.1

528

249

41

240

23

18

150

35

10

Lettuce

5

60

32

73

5

Orange

5

84

6

78

6

API3000

Iprovalicarb

14.1

321

119

31

160

29

10

203

13

18

Lettuce

5

108

5

104

7

Orange

5

97

4

90

10

API3000

Isoxaflutole

12.9

360

251

46

270

19

22

220

55

22

Lettuce

5

76

18

90

15

Orange

5

86

18

98

5

API3000

Linuron

13.8

249

160

46

290

25

12

182

21

14

Cuc/lett

4

103

16

86

11

Apple/grape

4

90

26

101

6

API2000

Metamitron

10.7

203

175

51

290

23

14

104

31

6

Cuc/lett

4

80

11

87

17

Apple/grape

4

97

17

95

9

API2000

Methabenzthiazuron

13.3

222

165

31

200

21

12

150

45

12

Lettuce

5

106

4

98

6

Orange

5

84

8

107

9

API3000

Methamidofos

4.6

142

94

41

240

21

6

125

19

8

Cuc/lett

4

83

16

79

19

Apple/grape

4

86

11

81

5

API2000

Methiocarb

13.8

226

169

46

300

13

14

121

25

10

Cuc/lett

4

94

9

95

4

Apple/grape

4

101

5

94

1

API2000

Methiocarbsulfon

10.7

258

122

56

370

25

8

201

13

16

Cuc/lett

4

109

12

99

11

Apple/grape

4

94

9

87

6

API2000

Methiocarbsulfoxide

10.1

242

185

46

290

19

14

170

31

14

Cuc/lett

4

116

5

101

3

Apple/grape

4

126

8

104

2

API2000

Methomyl

8.8

163

88

21

130

13

6

106

13

8

Cuc/lett

4

153

22

136

19

Apple/grape

4

125

14

103

7

API2000

Methoxyfenozide

13.8

369

313

24

200

13

24

133

34

10

Lettuce

5

93

7

91

4

Orange

5

91

13

91

3

API3000

Metobromuron

13.1

259

170

46

280

25

12

148

21

12

Cuc/lett

4

112

19

99

6

Apple/grape

4

96

9

99

12

API2000

Metoxuron

11.6

229

72

31

190

37

4

46

35

2

Lettuce

5

104

8

100

4

Orange

5

95

8

102

4

API3000

Monocrotofos

9.2

224

127

41

240

21

10

193

11

16

Cuc/lett

4

108

5

90

4

Apple/grape

4

111

8

98

10

API2000

Monolinuron

12.8

215

126

41

260

23

8

148

19

12

Cuc/lett

4

104

7

98

6

Apple/grape

4

111

7

107

8

API2000

Omethoate

6.5

214

125

36

230

29

10

183

15

14

Cuc/lett

4

98

13

85

13

Apple/grape

4

102

5

86

2

API2000

Oxamyla

8.0

237

72

21

160

23

4

90

11

6

Cuc/lett

4

107

31

90

7

Apple/grape

4

128

14

97

9

API2000

Oxamyl-oxim

6.6

163

72

36

230

17

4

90

25

6

Cuc/lett

4

100

6

85

3

Apple/grape

4

118

3

101

9

API2000

Oxycarboxin

10.9

268

175

26

170

19

14

147

35

10

Lettuce

5

98

6

96

4

Orange

5

85

22

78

5

API3000

Oxydemeton-methyl

8.5

247

169

41

230

19

14

109

35

8

Cuc/lett

4

98

11

89

7

Apple/grape

4

104

5

96

4

API2000

Paclobutrazole

13.8

294

70

36

320

45

4

125

51

10

Lettuce

5

96

9

87

8

Orange

5

77

67

69

6

API3000

Pencycuron

15.4

329

125

56

340

35

10

218

23

18

Cuc/lett

4

100

5

77

3

Apple/grape

4

118

4

92

9

API2000

Phenmedipham

13.2

301

168

51

290

13

14

136

29

10

Cuc/lett

4

99

7

96

5

Apple/grape

4

108

11

84

11

API2000

Phenm.-metabolite

10.0

168

136

31

200

14

10

108

26

8

Cuc/lett

4

107

9

103

5

Apple/grape

4

101

14

96

17

API2000

Phorate

15.5

261

75

26

150

21

4

47

45

8

Lettuce

5

96

27

91

11

Orange

5

104

2

88

6

API3000

Phorate-sulfone

12.9

293

171

26

150

17

10

115

37

10

Lettuce

5

114

10

95

9

Orange

5

83

6

104

4

API3000

Phorate-sulfoxide

12.8

277

199

41

270

17

6

97

45

4

Lettuce

5

99

8

96

3

Orange

5

98

6

91

4

API3000

Phosphamidon

11.7

300

174

41

250

19

14

127

33

10

Lettuce

5

101

5

107

5

Orange

5

98

7

100

7

API3000

Picolinafen

16.4

377

238

56

220

41

14

256

29

20

Lettuce

5

81

8

96

6

Orange

5

103

8

99

5

API3000

Pirimicarb

13.0

239

72

26

360

31

4

182

23

12

Lettuce

5

99

6

96

4

Orange

5

89

9

92

3

API3000

Pirimicarb, desmethyl

11.6

225

72

21

360

33

4

168

21

6

Lettuce

5

103

4

98

3

Orange

5

31

14

42

15

API3000

Prochloraz

15.4

376

308

46

310

13

22

70

41

16

Cuc/lett

4

90

15

78

13

Apple/grape

4

84

38

94

64

API2000

Profoxydim

16.2

466

280

66

140

27

20

180

35

12

Lettuce

5

33

25

30

6

Orange

5

49

34

55

5

API3000

Propamocarb

8.5

189

102

31

190

25

6

144

19

12

Lettuce

5

75

4

72

4

Orange

5

22

14

18

8

API3000

Propoxur

12.2

210

111

31

210

19

8

168

11

14

Cuc/lett

4

114

3

100

5

Apple/grape

4

118

3

98

5

API2000

Prothiocarb

7.4

191

146

46

240

21

12

   

Cuc/lett

4

85

26

63

37

Apple/grape

4

106

5

83

10

API2000

Pymetrozine

9.0

218

105

56

370

27

8

201

9

16

Cuc/lett

4

65

26

85

8

Apple/grape

4

47

7

71

7

API2000

Pyraclostrobin

15.1

388

194

1

350

19

6

163

33

6

Lettuce

5

72

13

77

6

Orange

5

87

4

83

7

API3000

Pyridate metabolite

10.4

207

77

56

340

45

6

104

31

8

Cuc/lett

4

100

12

87

4

Apple/grape

4

89

9

75

5

API2000

Rotenone

14.7

395

213

101

370

31

16

192

33

14

Cuc/lett

4

93

13

93

8

Apple/grape

4

94

16

94

30

API2000

Sethoxydim

15.2

328

178

46

260

25

14

220

19

18

Cuc/lett

4

67

39

88

3

Apple/grape

4

59

34

96

28

API2000

Spinosyn A

22.0

733

142

96

280

43

12

98

83

6

Lettuce

5

95

9

93

6

Orange

5

97

4

92

2

API3000

Spinosyn D

24.1

747

142

96

110

47

12

98

89

4

Lettuce

5

86

3

93

6

Orange

5

99

7

92

5

API3000

Tebuconazole

14.8

308

70

61

140

51

6

125

53

8

Lettuce

5

80

6

93

3

Orange

5

95

8

96

4

API3000

Tebufenozide

14.5

353

133

26

180

23

10

297

13

22

Cuc/lett

4

103

16

86

11

Apple/grape

4

106

42

78

33

API2000

Temephos

16.3

467

125

71

320

39

10

419

35

32

Lettuce

5

62

27

81

6

Orange

5

92

7

95

9

API3000

Tepraloxydim

12.7

342

250

31

180

19

28

166

29

12

Lettuce

5

44

19

60

7

Orange

5

73

15

62

4

API3000

Terbufos

16.7

289

103

11

120

13

10

57

37

8

Lettuce

5

73

27

75

8

Orange

5

80

24

81

12

API3000

Terbufos-sulfone

13.5

321

171

21

130

19

12

115

39

6

Lettuce

5

108

4

101

11

Orange

5

99

6

93

10

API3000

Terbufos-sulfoxide

13.5

305

187

6

110

17

10

97

59

8

Lettuce

5

106

3

103

5

Orange

5

98

5

97

9

API3000

Thiabendazole

12.2

202

175

56

370

35

12

131

45

10

Cuc/lett

4

87

12

101

3

Apple/grape

4

98

2

92

7

API2000

Thiacloprid

11.0

253

126

41

210

27

8

90

53

16

Lettuce

5

97

9

102

3

Orange

5

102

6

116

7

API3000

Thiametoxam

9.0

292

211

46

270

19

24

132

33

10

Lettuce

5

94

4

97

4

Orange

5

101

9

99

6

API3000

Thiocyclamd

12.6

182

137

21

160

21

12

73

29

14

Lettuce

5

96

11

89

6

Orange

5

100

15

82

11

API3000

Thiodicarb

12.7

355

88

20

130

31

6

108

21

8

Cuc/lett

4

37

115

42

98

Apple/grape

4

83

4

79

4

API2000

Thiofanox

12.9

219

57

11

90

19

6

61

15

4

Lettuce

5

nd

81

93

21

Orange

5

nd

84

30

API3000

Thiofanox-sulfone

10.2

251

57

16

350

26

2

76

21

4

Lettuce

5

110

16

101

5

Orange

5

85

25

85

8

API3000

Thiofanox-sulfoxide

9.8

235

104

31

320

17

4

57

27

2

Lettuce

5

110

2

105

3

Orange

5

109

11

88

6

API3000

Thiometonc

13.0

247

89

16

110

23

6

61

45

8

Lettuce

5

96

17

100

9

Orange

5

87

11

100

2

API3000

Thiophanate-methyl

12.1

343

151

30

210

25

12

311

17

23

Cuc/lett

4

66

8

75

16

Apple/grape

4

41

59

37

98

API2000

Tolylfluanidea

14.7

364

238

31

210

19

18

137

41

10

Cuc/lett

4

31

116

42

115

Apple/grape

4

75

93

24

81

API2000

Triadimefon

14.0

294

197

31

180

23

12

225

19

18

Lettuce

5

92

10

86

6

Orange

5

89

7

78

7

API3000

Triadimenol

14.1

296

70

16

130

31

4

99

21

8

Lettuce

5

101

7

87

6

Orange

5

89

7

82

9

API3000

Triazoxide

13.5

248

68

56

320

47

4

95

37

6

Lettuce

5

99

102

76

19

Orange

5

43

107

69

10

API3000

Trichlorfon

10.6

257

109

46

260

27

8

221

15

18

Cuc/lett

4

116

16

104

22

Apple/grape

4

114

8

99

4

API2000

Tricyclazole

11.5

191

136

56

360

39

10

163

31

12

Cuc/lett

4

105

5

92

6

Apple/grape

4

96

11

83

3

API2000

Triflumuron

14.9

359

156

30

200

23

12

139

47

10

Cuc/lett

4

94

9

92

7

Apple/grape

4

118

12

109

8

API2000

Triforine

13.2

435

390

12

100

13

30

215

40

15

Cuc/lett

4

98

13

101

6

Apple/grape

4

97

10

93

9

API2000

Vamidothion

10.4

288

146

46

300

19

12

118

31

8

Cuc/lett

4

111

16

96

3

Apple/grape

4

119

11

104

7

API2000

Cuc, cucumber

Lett, lettuce

aNH4 adduct

bNa adduct

cLOQ level 0.05 mg kg−1

dLOQ level 0.02 mg kg−1

Routine experience with LC–MS–MS analysis for over four years, both with the API2000 (20 mg matrix) and the API3000 (2 mg matrix) has shown that injection of uncleaned extracts does not result in special maintenance requirements. The source is cleaned with a tissue daily. The LC column typically lasts for 6 months.

Changing the solvent

Because ethyl acetate is less suitable for direct injection in reversed phase LC, the solvent was changed. Because only small amounts of the raw extract need to be evaporated (less than 0.5 mL in the final method) and evaporation blocks enable simultaneous evaporation of many (typically 24–36) extracts, this step adds little to the overall sample-preparation time. Changing the solvent was even regarded as advantageous. It resulted in more freedom in selection of the final solvent to be injected into the LC, which can be critical for very polar compounds (e.g. in acetonitrile-based extraction methods, injection of 100% acetonitrile easily leads to band-broadening for methamidophos). It is also easier to compensate for the smaller amount of sample processed for dry crops (because of the need for addition of water) by evaporating a larger amount of the ethyl acetate extract.

In previous work [15] a small amount of a diethylene glycol (added as solution in methanol) was added, because this was found to facilitate reconstitution, thereby improving recovery for some pesticide–matrix combinations. It was also shown that the evaporation step did not require special attention and that continuing the process for another half hour after completion of evaporation of the solvent did not affect recovery. The same procedure was therefore used here without re-evaluating the real need for it. Reconstitution was performed by first dissolving in methanol (ultrasonication) and then dilution with LC mobile phase component A.

Validation of LC–MS–MS method

The LC–MS–MS method was validated in three separate studies, one using the API2000 with injection of 20 mg matrix equivalent and the other two using the API3000 with injection of 2 mg matrix equivalent. A total of 140 pesticides and degradation products were included. In contrast with the full-scan acquisition in GC–MS, in LC–MS–MS data were acquired for a fixed, limited, set of pesticides. Although many pesticides from the GC–MS method can also be analyzed by LC–MS–MS, emphasis was on pesticides that were not, or less, amenable to GC analysis.

Recovery, based on matrix-matched calibration, and repeatability were evaluated at the 0.01 and 0.1 mg kg−1 level for vegetable and fruit matrices; the results are listed in Table 8. Although acceptable performance data were obtained for most of the pesticides, low recovery and/or high variability were observed for some. Among these were compounds that were also reported as troublesome by other workers using alternative multi-residue methods, e.g. asulam [30]. Low recovery could be partly attributed to poor extraction efficiency (asulam, hymexazol, and, in orange, propamocarb) or degradation during sample preparation (cycloxydim, sethoxydim, profoxydim, tepraloxydim, dichlofluanide, tolylfluanide, thiodicarb, thiophanate-methyl, and, in lettuce, disulfoton and furathiocarb). The degradation seems to be related to the change of solvent, as is apparent from comparison of GC–MS and LC–MS–MS validation data for dichlofluanide, tolylfluanide, and disulfoton. Fortunately, for many of these the degradation products formed are also part of the residue definition and are included in the method. Indeed, elevated recovery was observed for the degradation products when determined in the same validation set as the parent compound. In the analysis, therefore, degradation is not necessarily a problem, because the results (expressed as defined in the residue definition) have to be summed. In routine analytical quality control (see below) the data were evaluated this way.

Analytical quality-control data from routine LC–MS–MS analysis

In the same way as for GC–MS analysis, the initial validation data are continually being supplemented by performance data generated as part of analytical quality control during routine analysis of samples. With each set of analytical samples at least one was fortified with the full quantitative suite (i.e. 136 pesticides and degradation products) at the 0.05 mg kg−1 level. A compilation was made from all the data generated over a period of 12 months, which included data for more than one hundred vegetable and fruit matrices. A limited number of dry matrices (flour, milk powder) were also included in the set. The data were evaluated for one transition for each pesticide, using the API3000 and injection of 2 mg equivalent of matrix (10 μL of a 0.2 g mL−1 extract). Examples of typical extracted ion chromatograms are shown in Fig. 5.
Fig. 5

Typical extracted ion chromatograms obtained by LC–MS–MS analysis of vegetable and fruit extracts (calibration standard in mango matrix, 10 pg μL, corresponding to 0.05 mg kg−1)

For all fortified samples the matrix effect was also established by analyzing the corresponding matrix-matched standard, at the same level as in the extract of the fortified sample, against a solvent standard. Suppression (or enhancement) of up to 20% was regarded as acceptable for quantification. The number of compounds for which the response in matrix relative to that in solvent was between 80 and 120% is given in Table 9 for each matrix. Whereas for beetroot, asparagus, and kangkung little or no matrix effects exceeding 20% were observed, such effects were much more common for herbs and citrus fruits.
Table 9

Overview of matrix effects and recoverya within or outside the EU 60–140% criterion [37] after LC–MS–MS analysis

 

N

Matrix effects

n*

Recovery

# Pesticides

# Pesticides

Rel. resp. 80–120%

>20% suppr.

>20% enhanc.

Calc. using solvent std

Calc. using matrix-matched std

60–140%

<60%

>140%

60–140%

<60%

>140%

Corn syrup (2/2)

135

134

0

1

104

97

4

3

99

4

1

Beetroot

135

133

1

1

104

101

3

0

101

3

0

Corn syrup (1/2)

135

132

2

1

104

98

4

2

100

2

2

Kangkung

135

132

2

1

104

91

11

2

94

8

2

Green pea

135

131

3

1

104

97

5

2

99

3

2

Asparagus

135

130

4

1

104

97

7

0

98

6

0

Coco nut

135

130

4

1

104

63

41

0

59

45

0

Papaya

135

130

3

2

104

96

4

4

98

4

2

Cauliflower

135

129

1

5

104

101

2

1

102

1

1

Fennel

135

129

4

2

104

100

3

1

101

3

0

Cherry (2/3)

135

128

7

0

104

100

4

0

100

4

0

Cherry (1/3)

135

127

7

1

104

92

8

4

98

2

4

Ladies’ fingers

135

127

8

0

104

97

7

0

97

7

0

Mango (1/2)

135

127

6

2

104

97

3

4

98

2

4

Cherry (3/3)

135

126

8

1

104

100

4

0

102

2

0

Mango juice

135

126

3

6

104

101

1

2

104

0

0

Mushroom

135

126

7

2

104

102

2

0

103

0

1

Taro

135

126

7

2

104

96

4

4

99

1

4

Plum (3/3)

135

125

8

2

104

95

7

2

100

4

0

Fennel leaves (2/2)

135

124

5

6

104

99

2

3

99

2

3

Milk powder

135

124

6

5

104

58

45

1

59

45

0

Grape

135

123

9

3

104

98

3

3

98

3

3

Spinach

135

123

12

0

104

94

8

2

96

5

3

Tamarind

135

123

8

4

104

67

37

0

79

25

0

Cassava

135

122

7

6

104

87

16

1

78

26

0

Raspberry (1/3)

135

122

12

1

104

84

20

0

92

12

0

Sweet pepper

134

122

10

2

103

100

1

2

100

1

2

Apple puree

135

121

5

9

104

99

5

0

97

7

0

Corn flour

135

121

1

13

104

95

6

3

95

7

2

Courgette

135

121

7

7

104

100

2

2

100

3

1

Tomato puree

135

121

10

4

104

101

3

0

103

1

0

Raspberry (2/3)

135

120

15

0

104

98

5

1

100

3

1

Broccoli

135

119

14

2

104

90

10

4

93

8

3

Flour (2/2)

135

119

2

14

104

95

2

7

97

3

4

Peach (1/2)

135

119

16

0

104

99

5

0

100

4

0

Mango (2/2)

134

117

12

5

103

96

6

1

100

3

0

Milk/flour mix

135

117

12

6

104

43

60

1

55

49

0

Bitter cucumber

135

116

17

2

104

99

2

3

99

1

4

Melon puree

135

116

18

1

104

99

5

0

103

1

0

Tomato

135

116

13

6

104

93

8

3

96

5

3

Lettuce, crinkley

134

114

19

1

103

97

3

3

97

2

4

Pear

134

114

14

6

103

97

6

0

99

4

0

Flour (1/2)

135

113

14

8

104

73

28

3

85

19

0

Plum (1/3)

135

113

13

9

104

93

6

5

98

2

4

Celery leaves (1/3)

135

112

22

1

104

90

12

2

97

3

4

Purselane

135

112

23

0

104

96

6

2

98

4

2

Apricots

135

111

23

1

104

90

13

1

97

6

1

Artichoke

135

111

17

7

104

91

12

1

95

8

1

Cucumber

135

110

15

10

104

99

5

0

101

3

0

Horseradish powder

135

110

15

10

104

88

11

5

97

5

2

Tarrragon (2/2)

135

110

8

17

104

96

4

4

94

6

4

Avocado (1/2)

135

109

22

4

104

81

21

2

90

13

1

Haricot bean

135

109

25

1

104

83

20

1

90

13

1

Kiwi

135

109

10

16

104

97

6

1

100

2

2

Peach (12/2)

135

108

24

3

104

88

14

2

93

9

2

Raspberry (3/3)

135

107

26

2

104

80

22

2

90

11

3

Blackberry

133

106

17

10

102

91

10

1

91

9

2

Diced pumpkins

135

106

27

2

104

95

8

1

100

3

1

Plum (2/3)

135

106

23

6

104

86

18

0

85

18

1

Yam

135

106

1

28

104

97

6

1

96

8

0

Avocado (2/2)

134

103

29

2

103

68

34

1

80

22

1

Dill leaves

135

103

15

17

104

94

7

3

93

9

2

Honey

106

103

3

0

82

82

0

0

82

0

0

Chervil

135

102

29

4

104

95

9

0

98

5

1

Parsley

135

102

29

4

104

95

4

5

99

1

4

Nectarine

134

101

29

4

103

92

8

3

98

4

1

Bean sprouts

106

100

5

1

82

76

6

0

78

4

0

Sweetcorn (1/2)

106

99

6

1

82

76

5

1

77

3

2

Beetroot leaves

135

98

32

5

104

85

19

0

99

5

0

Chestnuts

106

98

1

7

82

76

4

2

79

3

0

Pomegranate (1/2)

135

97

37

1

104

84

20

0

100

4

0

Pomegranate (2/2)

135

97

37

1

104

84

20

0

100

4

0

Pear syrup

106

95

3

8

82

79

3

0

80

2

0

Alfalfa

106

94

11

1

82

75

7

0

78

4

0

Fennel leaves (1/2)

106

92

8

6

82

74

5

3

76

2

4

Chili pepper

135

91

40

4

104

95

8

1

101

1

2

Turnip tops

106

90

15

1

82

76

2

4

78

0

4

Blueberry

135

89

43

3

102

66

36

0

91

11

0

Litchi

135

88

45

2

104

78

26

0

99

4

1

Salak

135

88

42

5

104

82

20

2

99

4

1

Pepper powder

106

87

16

3

82

54

27

1

70

11

1

Celery leaves (2/3)

135

85

41

9

104

93

10

1

100

2

2

Lemon

134

84

47

3

104

78

20

6

97

3

4

Physalis

135

83

48

4

104

71

33

0

99

5

0

Maize (feed)

135

81

53

1

104

95

6

3

93

3

8

Sweetcorn (2/2)

135

80

50

5

104

79

22

3

98

6

0

Coriander (1/2)

135

79

56

0

104

68

34

2

95

6

3

Mangostan

135

76

40

19

104

46

54

4

69

35

0

Celery leaves (3/3)

134

75

58

1

103

86

16

1

99

2

2

Laos

135

73

57

5

104

70

33

1

99

4

1

Chives

135

71

57

7

104

98

5

1

102

1

1

Coriander (2/2)

135

65

60

10

104

83

21

0

98

6

0

Tea (black)

136

65

69

2

104

60

43

1

87

14

3

Lemon puree

135

53

80

2

104

68

36

0

103

1

0

Ginger

135

46

86

3

104

68

34

2

98

3

3

Grapefruit (1/2)

133

46

87

0

102

43

59

0

98

1

3

Grapefruit (2/2)

135

46

88

1

103

61

41

1

97

3

3

Oregano

135

46

75

14

104

52

50

2

87

16

1

Kumquat

135

38

95

2

104

47

56

1

94

6

4

Lime

134

38

94

2

103

48

52

3

96

4

3

Tarrragon (1/2)

135

38

95

2

104

41

63

0

90

13

1

Italian herb mix

135

33

101

1

104

54

49

1

95

8

1

Total QC results

13497

10488

2566

443

10395

8618

1613

164

9533

708

154

Percentage of total results

 

78

19

3

 

83

16

2

92

7

1

aRecovery at 0.05 mg kg−1 (higher for seven pesticides). The pesticides included are listed in Table 10

N is the total number of individual compounds (pesticides and metabolites) added to the matrix

n* is the total number of pesticides added to the matrix. Compounds belonging to the same residue definition counted as one

In contrast with GC, for which matrix effects are mainly caused by shielding of active sites in the inlet and were, to some extent predictable (in relation to the matrix load injected and the lability and/or polarity of analyte), in LC–MS–MS matrix effects are much less predictable. Although they do depend on the amount of matrix introduced into the system, and also tend to be more abundant in complex (“aromatic”) matrices, it cannot be readily predicted for which pesticides the effects occur. For this reason use of one matrix-matched standard as representative calibrant for a whole range of commodities, which worked reasonably well in GC–MS analysis, was not feasible in LC–MS–MS analysis. Consequently, critical evaluation of the matrix effect was required; if unacceptable suppression occurred there was no alternative to quantification by use of the appropriate matrix-matched calibration standard or, when not available, by standard addition.

Recovery of the pesticides from the fortified samples was calculated relative to that from a solvent standard and a matrix-matched standard and tested against the 60–140% criterion for evaluation of routine analytical quality-control samples [37]. A total of more than 10,000 recovery values were evaluated. Without matrix-matched calibration, acceptable recovery was obtained for 83% of the pesticides. Deviating recoveries were usually too low, mainly because of ion suppression, as is apparent from the results obtained from determination of recovery using matrix-matched calibration, for which 92% met the criterion.

Concentrating on performance at the pesticide level (Table 10) enables easy identification of troublesome pesticides. All compounds belonging to the same residue definition were summed (according to the residue definition) and counted as one, thereby compensating for possible conversion during sample pretreatment. This way the low recovery of dichlofluanide and the corresponding high recovery of DMSA were acceptable for most matrices because recovery for the sum met the criterion. Pesticides for which multi-matrix analysis under fixed conditions was less favorable included asulam, bifenazate, cyromazine, furathiocarb, propamocarb, pymetrozine, and thiocyclam (low recovery because of varying extraction efficiency and/or degradation). As already observed during validation, the method was also less suitable for cycloxydim, profoxydim, sethoxydim, and tepraloxydim. For these compounds recovery was too high, possibly because of degradation in the calibration standard used for preparation of the matrix-matched standards.
Table 10

Recovery over all matrices (LC–MS–MS)

 

 

# ACQ samples

# Recov. 60–140%

# Recov. <60%

# Recov. >140%

Average recov. (%)a

RSD (%)a

1

Abamectin

102

100

2

0

86

17

2

Acephate

102

93

9

0

78

13

3

Acetamiprid

102

97

5

0

90

11

Aldicarb

102

101

0

1

91

13

Aldicarb-sulfone

102

102

0

0

92

12

Aldicarb-sulfoxide

102

96

6

0

84

13

4

Asulam

102

69

32

1

85

17

5

Azamethiphos

102

102

0

0

89

12

6

Azinfos-methyl

102

96

5

1

87

15

7

Bendiocarb

93

93

0

0

88

12

8

Bifenazate

98

60

37

1

85

18

9

Bitertanol

102

98

4

0

84

15

Butocarboxim

102

101

1

0

88

14

Butoxycarboxim

102

101

1

0

91

12

10

Carbaryl

102

100

1

1

87

13

Carbendazim

100

97

2

1

93

14

Carbofuran

102

100

1

1

92

12

Carbofuran,3-hydroxy-

102

102

0

0

93

11

11

Carboxin

102

97

5

0

84

13

12

Chlorbromuron

102

98

4

0

86

14

13

Chlorfluazuron

102

93

8

1

87

15

14

Clofentezine

102

89

13

0

80

15

15

Clomazone

93

89

3

1

85

12

16

Clothianidin

93

91

2

0

91

12

17

Cycloxydim

102

68

11

23

104

19

18

Cymoxanil

102

102

0

0

91

15

19

Cyromazine

102

49

53

0

74

12

20

Demeton

102

102

0

0

89

14

Demeton-S-methyl

102

100

2

0

87

14

Demeton-S-methylsulfone

102

101

1

0

91

12

21

Desmedipham

102

96

6

0

83

14

Dichlofluanid

102

36

66

0

80

19

22

Dicrotophos

102

100

2

0

89

14

23

Diflubenzuron

102

98

4

0

82

15

24

Dimethirimol

93

90

3

0

89

11

Dimethoate

102

101

1

0

90

12

25

Diniconazole

93

84

8

1

86

16

Disulfoton

93

67

25

1

75

13

Disulfoton-sulfone

93

93

0

0

88

12

Disulfoton-sulfoxide

93

89

0

4

96

16

26

Diuron

93

92

1

0

87

14

DMSA

102

41

0

61

109

17

DMST

102

96

1

5

104

16

Ethiofencarb

102

99

3

0

86

14

Ethiofencarb-sulfone

102

102

0

0

90

13

Ethiofencarb-sulfoxide

102

101

1

0

92

15

27

Ethirimol

102

98

4

0

88

12

28

Famoxadone

102

95

7

0

83

14

Fenamiphos

102

100

2

0

89

14

Fenamiphos-sulfone

102

102

0

0

91

12

Fenamiphos-sulfoxide

93

92

1

0

90

11

29

Fenhexamid

102

96

6

0

85

12

30

Fenpyroximate

102

92

10

0

87

13

Fensulfothion

102

102

0

0

88

11

Fensulfothion-sulfone

93

91

2

0

85

12

Fenthion

102

99

3

0

87

14

Fenthion-sulfone

102

99

2

1

88

15

Fenthion-sulfoxide

102

102

0

0

93

14

31

Flucycloxuron

102

94

8

0

88

15

32

Flufenoxuron

102

93

9

0

87

14

33

Fosthiazate

93

93

0

0

90

12

34

Furathiocarb

102

79

20

3

84

16

35

Hexaflumuron

102

90

10

2

85

18

36

Hexythiazox

102

91

11

0

85

15

37

Imazalil

101

92

9

0

83

14

38

Imidacloprid

102

99

3

0

90

14

39

Indoxacarb

101

96

5

0

86

16

40

Iprovalicarb

93

92

1

0

87

13

41

Isoxaflutole

93

83

10

0

82

14

42

Linuron

102

97

4

1

85

12

43

Metamitron

102

97

5

0

88

15

44

Methabenzthiazuron

93

93

0

0

88

13

45

Methamidophos

102

90

12

0

75

12

Methiocarb

102

100

2

0

85

13

Methiocarb-sulfone

102

84

18

0

78

15

Methiocarb-sulfoxide

102

99

2

1

88

12

46

Methomyl

102

89

0

13

101

14

47

Methoxyfenozide

102

101

1

0

85

14

48

Metobromuron

102

97

4

1

87

12

49

Metoxuron

93

93

0

0

89

12

50

Monocrotophos

102

101

1

0

90

12

51

Monolinuron

102

101

1

0

86

14

Omethoate

102

99

3

0

83

12

Oxamyl

102

100

2

0

89

12

Oxamyl-oxime

102

101

1

0

88

12

52

Oxycarboxin

102

102

0

0

91

12

Oxydemeton-methyl

102

97

5

0

86

13

53

Paclobutrazole

102

101

1

0

87

12

54

Pencycuron

102

96

6

0

81

14

Phenmedipham

102

94

7

1

83

14

Phenmedipham-metabolite

102

100

2

0

93

15

Phorate

102

68

34

0

74

19

Phorate-sulfone

93

93

0

0

88

12

Phorate-sulfoxide

102

101

1

0

90

12

55

Phosphamidon

93

93

0

0

89

10

56

Picolinafen

93

86

6

1

84

15

Pirimicarb

102

101

0

1

89

12

Pirimicarb, desmethyl-

102

100

1

1

90

12

57

Prochloraz

101

94

7

0

83

14

58

Profoxydim

99

54

32

13

99

21

59

Propamocarb

101

9

92

0

70

15

60

Propoxur

102

100

2

0

88

16

61

Pymetrozine

102

73

29

0

89

20

62

Pyraclostrobin

102

95

7

0

85

14

63

Pyridate-metabolite

102

92

9

1

86

15

64

Rotenone

102

93

9

0

81

15

65

Sethoxydim

102

72

3

27

106

19

66

Spinosyn-A

93

88

5

0

82

17

Spinosyn-D

93

82

11

0

83

15

67

Tebuconazole

93

90

3

0

86

16

68

Tebufenozide

102

99

3

0

86

14

69

Temephos

102

94

8

0

87

16

70

Tepraloxydim

102

62

0

40

114

14

Terbufos

93

62

30

1

77

15

Terbufos-sulfone

93

90

3

0

86

13

Terbufos-sulfoxide

93

92

1

0

88

12

71

Thiabendazole

98

92

5

1

86

13

72

Thiacloprid

93

90

3

0

88

12

73

Thiametoxam

93

91

2

0

89

13

74

Thiocyclam

93

64

29

0

78

16

Thiodicarb

102

62

40

0

82

16

Thiofanox

102

98

3

1

85

14

Thiofanox-sulfone

102

102

0

0

90

13

Thiofanox-sulfoxide

102

101

1

0

92

14

75

Thiometon

93

88

4

1

87

16

Thiophanate-methyl

102

83

19

0

77

12

Tolylfluanid

101

36

65

0

76

22

Triadimefon

102

99

3

0

85

13

Triadimenol

102

98

3

1

87

12

76

Triazoxide

102

90

9

3

84

16

77

Trichlorfon

102

101

0

1

87

12

78

Tricyclazole

102

96

6

0

87

12

79

Triflumuron

101

89

10

2

84

18

80

Triforine

102

97

3

2

87

15

81

Vamidothion

102

101

1

0

89

11

82

Sum aldicarb

102

101

1

0

88

11

83

Sum butocarboxim

102

101

1

0

90

11

84

Sum carbendazim

101

97

4

0

83

12

85

Sum carbofuran

102

102

0

0

92

10

86

Sum dimethoate

102

100

2

0

86

10

87

Sum dichlofluanid

102

89

1

12

107

17

88

Sum disulfoton

93

89

4

0

86

13

89

Sum ethiofencarb

102

102

0

0

89

11

90

Sum fenamiphos

102

101

1

0

90

11

91

Sum fensulfothion

102

102

0

0

86

11

92

Sum fenthion

102

102

0

0

89

12

93

Sum methiocarb

102

100

2

0

83

12

94

Sum methomyl

102

100

2

0

87

12

95

Sum oxamyl

102

101

1

0

88

10

96

Sum oxydemeton-methyl

102

101

1

0

88

11

97

Sum phenmedipham

102

101

1

0

88

13

98

Sum phorate

102

97

5

0

81

12

99

Sum pirimicarb

102

101

1

0

90

12

100

Sum terbufos

93

88

5

0

81

13

101

Sum thiofanox

102

102

0

0

89

11

102

Sum tolylfluanid

101

95

6

0

80

15

103

Sum triadimefon

102

99

3

0

86

13

aAverage and RSD for recoveries within 60–140% range

Matrix-matched calibration, API3000

Level = 0.05 mg kg−1 for most pesticides/metabolites

Bold indicates pesticides, including metabolites that are part of residue definition, if appropriate

Averaging acceptable recoveries reveals there is some bias, because the values are mostly approximately 87% (in contrast with the GC–MS data, for which the average was approximately 100%). It was noted that for dry crops relatively low recovery (typically between 60–70%) was obtained for all pesticides. The cause is not clear. This bias can also be seen in tables in other papers (barley [26], soya grain [33]).

Independent evaluation of method performance by proficiency testing

From results obtained over the years from participation in proficiency tests, an additional and independent verification of method performance could be made. The data are summarized in Table 11 and clearly show that good quantitative data were consistently obtained from both GC–MS and LC–MS–MS, with method performance good (Z-score<2) 54 times, doubtful (2 < Z < 3) three times, and never poor. It also shows that the calibration approach (one-point calibration, tomato-matrix standard for GC and matrix-matched standard for LC) is fit-for-purpose.
Table 11

Results from the analysis of Fapas (series 19) proficiency test samples (2003–2005)

Sample

Pesticide

MRM

Spike level added (μg kg−1)

Inter-lab. result (μg kg−1)

TNO result (μg kg−1)

Z-score TNO

#53 Apple

Fenpropathrin

GC–MS

500

405

528

1.7

Parathion-methyl

GC–MS

70

59

47

−0.9

Tetradifon

GC–MS

140

115

91

−0.9

Triazofos

GC–MS

140

119

74

−1.7

Vinchlozolin

GC–MS

60

53

53

0.0

#52 Cucumber

Iprodione

GC–MS

100

94

89

−0.3

Methomyl

LC–MS–MS

28

25

28

0.5

Thiabendazole

LC–MS–MS

50

128

113

−0.5

#51 Pear

Carbendazim

LC–MS–MS

150

116

60

−2.2

Dodine

not in MRM

60

59

*

*

Imazalil

LC–MS–MS

400

237

273

0.8

#49 Melon

Chlorpropham

GC–MS

10

9

11

1.0

Chlorpyrifos

GC–MS

8

8

7

−0.7

Dimethoate

LC–MS–MS

15

19

15

−0.9

Pirimicarb

LC–MS–MS

20

19

16

−0.7

#48 Tomato

Azoxystrobin

GC–MS

Not given

201

166

−0.9

Bifenthrin

GC–MS

Not given

83

99

0.9

Buprofezin

GC–MS

Not given

108

131

1

Chlorpyrifos-methyl

GC–MS

Not given

319

281

−0.6

Procymidone

GC–MS

Not given

712

668

−0.4

#47 Grapefruit

Diazinon

GC–MS

Not given

262

294

0.6

Heptenophos

GC–MS

Not given

168

234

1.9

Malathion

GC–MS

Not given

715

690

−0.2

Methidathion

GC–MS

Not given

567

540

−0.3

#46 Lettuce

Bromopropylate

GC–MS

80

67

51

−1.1

Dimethoate

LC–MS–MS

300

285

316

0.6

Oxadixyl

GC–MS

120

127

134

0.3

Penconazole

GC–MS

100

82

51

−1.7

Tolclofos-methyl

GC–MS

160

137

75

−2.1

#42 Apple

Chlorfenvinphos

GC–MS

90

71

50

−1.3

Chlorpyrifos

GC–MS

400

259

241

−0.3

Methamidophos

LC–MS–MS

60

44

31

−1.3

Monocrotophos

LC–MS–MS

80

58

56

−0.1

Omethoate

LC–MS–MS

150

108

103

−0.2

Trifluralin

GC–MS

100

59

62

0.2

#41 Basil

Kresoxim-methyl

GC–MS

150

94

86

−0.4

Procymidone

GC–MS

120

87

78

−0.5

Propyzamide

GC–MS

100

81

59

−1.2

Vinclozolin

GC–MS

60

47

44

−0.3

#38 Tomato

Azoxystrobin

GC–MS

150

137

132

−0.2

Bupirimate

GC–MS

100

83

62

−1.1

Chlorpyrifos-methyl

GC–MS

80

72

53

−1.2

Quinalphos

GC–MS

140

124

105

−0.7

#37 Lemon

Diazinon

GC–MS

80

42

42

0.0

Fenitrothion

GC–MS

100

78

80

0.1

Metalaxyl

GC–MS

120

94

93

0

Methidathion

GC–MS

150

109

154

1.9

#35 Lettuce

Carbendazim

LC–MS–MS

80

53

31

−1.9

lambda Cyhalothrin

GC–MS

80

66

54

−0.8

Metalaxyl

GC–MS

120

94

86

−0.4

#34 Apple

Diphenylamine

GC–MS

50

39

29

−1.2

Pirimiphos-methyl

GC–MS

50

41

42

0.1

Propargite

GC–MS

200

162

172

0.3

Tetradifon

GC–MS

100

83

38

−2.5

#29 Sweet pepper

Dichloran

GC–MS

200

179

200

0.6

Mecarbam

GC–MS

100

90

120

1.5

Methamidophos

LC–MS–MS

60

51

54

0.3

Conclusions

The ethyl acetate-based multi-residue method has been modified to meet today’s demands in respect of ease and speed of sample preparation. For GC–MS analysis, combined GCB/PSA dispersive clean-up enables prolonged injection of vegetable and fruit extracts (10 mg matrix equivalent) without maintenance. Retention time shifts induced by some matrices compared with the calibration standard are reduced by the clean-up procedure. Interferences are partially removed, resulting in cleaner (extracted ion) chromatograms. The last two benefits aid correct automatic peak assignment and confirmation. Addition of toluene during dispersive clean-up prevented unacceptable adsorption of planar pesticides by GCB yet removal of chlorophyll and other pigments was still sufficient. Use of liners with a sintered porous glass bed on the inner wall makes 20 μL injection non-critical and robust. In GC, use of a universal matrix-matched standard (tomato) is a feasible means of compensating for the matrix effects of many other vegetable and fruit samples. For most pesticides, LOQs of 0.01 mg kg−1 can be obtained by GC–MS with full-scan acquisition.

The same initial extract (i.e. without any clean-up) can be used for LC–MS–MS analysis, after changing the solvent to methanol–water. LC–MS–MS is relatively tolerant of injection of matrix—despite the absence of any clean-up no special maintenance was required. Matrix-induced suppression was observed for several matrices, however, especially herbs and citrus, and must be evaluated for all pesticide-matrix combinations. In contrast with the GC–based method, use of a universal matrix-matched standard to compensate for matrix effects was not feasible.

Evaluation of analytical quality control data for 271 pesticides and degradation products in over one hundred matrices showed that, at the 0.05 mg kg−1 level, recovery was acceptable for 92% (LC–MS–MS) and 93% (GC–MS) of all pesticide–matrix combinations. It also revealed that the method fails in the other 7–8% because of lack of specificity (mostly in GC–MS) or because of poor extraction efficiency and/or degradation (LC–MS–MS). The only way to identify these limitations is by thorough and continual evaluation of the quantitative performance of the method for all the pesticides (rather then a “representative subset”) in all the matrices.

Notes

Acknowledgements

Jan Quirijns is acknowledged for development of the initial ethyl acetate-based method at the TNO laboratory and for investigation of sample homogenization. Gert Stil, Corina van Ballegooien, Piet van Prattenburg, Petra Dam, Rob van Dinter, Maarten Nooteboom, and Hans Kooiman are acknowledged for generation of the extensive set of analytical quality-control data during routine analysis of the samples.

References

  1. 1.
    Luke M, Froberg JE, Masumoto HT (1975) J Assoc Off Anal Chem 58:1020–1026Google Scholar
  2. 2.
    Specht W, Pelz S, Gilsbach W (1995) Fresenius J Anal Chem 353:183–190CrossRefGoogle Scholar
  3. 3.
    Stan HJ (2000) J Chromatogr A 892:347–377CrossRefGoogle Scholar
  4. 4.
    General Inspectorate for Health Protection (1996) analytical methods for pesticide residues in foodstuffs, 6th edn., Part 1, Ministry of Health, Welfare and Sport. The Hague, The NetherlandsGoogle Scholar
  5. 5.
    Roos AH, van Munsteren AJ, Nabs FM, Tuinstra LGM Th (1987) Anal Chim Acta 196:95–102CrossRefGoogle Scholar
  6. 6.
    Andersson A, Palsheden H, Fresenius (1991) J Anal Chem 339:365–367CrossRefGoogle Scholar
  7. 7.
    Steinwandter H, Fresenius (1992) J Anal Chem 343:887–889CrossRefGoogle Scholar
  8. 8.
    Hajslova J, Holadova K, Kocourek V, Poustka J, Godula M, Cuhra P, Kempny M (1998) J Chromatogr A 800:283–295CrossRefGoogle Scholar
  9. 9.
    Pihlstrom T, Osterdahl BG (1999) J Agric Food Chem 47:2549–2552CrossRefGoogle Scholar
  10. 10.
    Obana H, Akutsu K, Okihashi M, Hori S (2001) Analyst 126:1529–1534CrossRefGoogle Scholar
  11. 11.
    Taylor MJ, Hunter K, Hunter KB, Lindsay D, le Bouhellec S (2002) J Chromatogr A 982:225–236CrossRefGoogle Scholar
  12. 12.
    Zrostlikova J, Hajslova J, Cajka T (2003) J Chromatogr A 1019:173–186CrossRefGoogle Scholar
  13. 13.
    Cajka T, Hajslova J (2004) J Chromatogr A 1058:251–261CrossRefGoogle Scholar
  14. 14.
    Patel K, Fussell RJ, Goodall DM, Keely BJ (2003) Analyst 128:1228–1231CrossRefGoogle Scholar
  15. 15.
    Mol HGJ, van Dam RCJ, Steijger OM (2003) J Chromatogr A 1015:119–127CrossRefGoogle Scholar
  16. 16.
    Jansson Ch, Pihlstrom T, Osterdahl BG, Markides KE (2004) J Chromatogr A 1023:93–104CrossRefGoogle Scholar
  17. 17.
    Agüera A, López S, Fernández-Alba AR, Contreras M, Crespo J, Piedra L (2004) J Chromatogr A 1045:125–135CrossRefGoogle Scholar
  18. 18.
    Ferrer I, Garcia-Reyes JF, Mezcua M, Thurman EM, Fernandez-Alba AR (2005) J Chromatogr A 1082:81–90CrossRefGoogle Scholar
  19. 19.
    Fernández Moreno JL, Arrebola Liébanas FJ, Garrido Frenich A, Martínez Vidal JL (2006) J Chromatogr A 1111:97–105CrossRefGoogle Scholar
  20. 20.
    Zhang W-G, Chu X-G, Cai H-X, An J, Li Ch-J (2006) Rapid Commun Mass Spectrom 20:609–617CrossRefGoogle Scholar
  21. 21.
    Sharif Z, Bin Che Man Y, Sheikh Abdul Hamid N, Cheow Keat Ch (2006) J Chromatogr A 1127:254–261CrossRefGoogle Scholar
  22. 22.
    Berrada H, Fernandez M, Ruiz MJ, Molto JC, J Manes (2006) Food Add Contam 23:674–682CrossRefGoogle Scholar
  23. 23.
    Martinez Vidal JL, Arrebola Liebanas FJ, Gonzalez Rodriguez MJ, Garrido Frenich A, Fernandez Moreno JL (2006) Rapid Commun Mass Spectrom 20:365–375CrossRefGoogle Scholar
  24. 24.
    Cortes JM, Sanchez R, Diaz plaza EM, Villen J, Vazquez A (2006) J Agric Food Chem 54:1997–2002CrossRefGoogle Scholar
  25. 25.
    Patel K, Fussell RJ, Goodall DM, Keely BJ (2006) J Sep Sci 29:90–95CrossRefGoogle Scholar
  26. 26.
    Diez C, Traag WA, Zomer P, Marinero P, Atienza J (2006) J Chromatogr A 1131:11–23CrossRefGoogle Scholar
  27. 27.
    Liao W, Joe T, Cusick WG (1991) J Assoc Off Anal Chem 74:554–565Google Scholar
  28. 28.
    Fillion J, Hindle R, Lacroix M, Selwyn J (1995) J Assoc Off Anal Chem 78:1252–1266Google Scholar
  29. 29.
    Anastassiades M, Lehotay SJ, Stajnbaher D, Schenk FJ (2003) J Assoc Off Anal Chem 86:412–431Google Scholar
  30. 30.
    Lehotay SJ, de Kok A, Hiemstra M, van Bodegraven P (2005) J Assoc Off Anal Chem 88:595–614Google Scholar
  31. 31.
    Hercegova A, Domotorova M, Kruzlicova D, Matisova E (2006) J Sep Sci 29:1102–1109CrossRefGoogle Scholar
  32. 32.
    Reynolds 1997–2001: S.L. Reynolds, R. Fussell, M. Caldow, R. James, S. Nawaz, C. Ebeen, D. Pendlington, T. Stijve, S. Lovell, H. Diserens (1997, 1998, 2000, 2001) “Intercomparison study of two multi-residue methods for the enforcement of EU MRLs for pesticides in fruits, vegetables and grain”, European Commission, LuxembourgGoogle Scholar
  33. 33.
    Pizzuti IR, de Kok A, Zanella R, Adaime MB, Hiemstra M, Wickert C, Prestes OD (2007) J Chromatogr A 1142:123–136Google Scholar
  34. 34.
    Annex I to Directive 90/642/EEC [Part of products to which maximum limits apply], as amended by Directive 93/58/EEC (OJ L 211, 23.8.1993, p. 6)Google Scholar
  35. 35.
    Commission directive 2002/63/EC of 11 July 2002 establishing Community methods of sampling for the official control of pesticide residues in and on products of plant and animal origin and repealing Directive 79/700/EEC, Official Journal of the European Communities L187, 16/07/2002, P. 30–43Google Scholar
  36. 36.
    SANCO/825/00 rev. 7, 17.03.2004, Guidance document on residue analytical methodsGoogle Scholar
  37. 37.
    SANCO/10232/2006, Quality control procedures for pesticide residue analysis 4th ednGoogle Scholar
  38. 38.
    Okihashi M, Kitagawa Y, Akutsu K, Obana H, Tanaka Y (2005) J Pestic Sci 30:368–377CrossRefGoogle Scholar
  39. 39.
    Fussell RJ, Addie KJ, Reynolds SL, Wilson MF (2002) J Agric Food Chem 50:441–448CrossRefGoogle Scholar
  40. 40.
    Martinez Vidal JL, Arrebola FJ, Mateu-Sanchez M (2002) Rapid Commun Mass Spec 16:1106–1115CrossRefGoogle Scholar
  41. 41.
    Lehotay SJ (2000) J Assoc Off Anal Chem 83:680–697Google Scholar
  42. 42.
    Kerkdijk H, Mol HGJ, van der Nagel B, submitted for publicationGoogle Scholar
  43. 43.
    Barwick VJ, Ellison SLR, Lacey SJ, Mussell CR, Lucking CL (1999) J Sci Food Agric 79:1190–1196CrossRefGoogle Scholar
  44. 44.
    Ueno E, Oshima H, Saito I, Matsumoto H (2004) J Assoc Off Anal Chem 87:1003–1015Google Scholar
  45. 45.
    Pang GF, Liu Y-M, Fan C-L, Zhang J-J, Cao Y-Z, Li X-M, Li Z-Y, Wu Y-P, Guo T-T (2006) Anal Bioanal Chem 384:1366-1408CrossRefGoogle Scholar
  46. 46.
    Mol HGJ, Janssen H-G, Cramers CA, Brinkman UA Th (1995) J High Resolut Chromatogr 18:19–27CrossRefGoogle Scholar
  47. 47.
    Teske J, Engewald W (2002) Trends Anal Chem 21:357CrossRefGoogle Scholar
  48. 48.
    Godula M, Hajslova J, Mastouska K, Krivankova J (2001) J Sep Sci 24:355–366CrossRefGoogle Scholar
  49. 49.
    Staniewski J, Rijks JA (1993) J High Resolut Chromatogr 16:182–187CrossRefGoogle Scholar
  50. 50.
    Poole CF (2007) J Chromatogr A DOI 10.1016/j.chroma.2007.01.018
  51. 51.
    Niessen WMA, Manini P, Andreoli R (2006) Mass Spectrom Rev 25:881–899CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2007

Authors and Affiliations

  • Hans G. J. Mol
    • 1
  • Astrid Rooseboom
    • 2
  • Ruud van Dam
    • 3
  • Marleen Roding
    • 3
  • Karin Arondeus
    • 3
  • Suryati Sunarto
    • 3
  1. 1.Rikilt Institute of Food SafetyPesticides and ContaminantsWageningenThe Netherlands
  2. 2.TNO-Blgg-AgriQWageningenThe Netherlands
  3. 3.Department Analytical ResearchTNO Quality of LifeZeistThe Netherlands

Personalised recommendations