Advertisement

Analytical and Bioanalytical Chemistry

, Volume 410, Issue 22, pp 5521–5528 | Cite as

Method development and validation for total haloxyfop analysis in infant formulas and related ingredient matrices using liquid chromatography-tandem mass spectrometry

  • Urairat Koesukwiwat
  • Lukas Vaclavik
  • Katerina Mastovska
Research Paper
Part of the following topical collections:
  1. Food Safety Analysis

Abstract

According to the European Commission directive 2006/141/EC, haloxyfop residue levels should not exceed 0.003 mg/kg in ready-to-feed infant formula, and the residue definition includes sum of haloxyfop, its esters, salts, and conjugates expressed as haloxyfop. A simple method for total haloxyfop analysis in infant formula and related ingredient matrices was developed and validated using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The sample preparation consisted of an alkaline hydrolysis with methanolic sodium hydroxide to release haloxyfop (parent acid) from its bound forms prior to the extraction with acetonitrile. A mixture of magnesium sulfate (MgSO4) and sodium chloride (NaCl) (4:1, w/w) was added to the extract to induce phase separation and force the analyte into the upper acetonitrile-methanol layer and then a 1-mL aliquot was subsequently cleaned up by dispersive solid phase extraction with 150 mg of MgSO4 and 50 mg of octadecyl (C18) sorbent. The analytical procedure was developed and carefully optimized to enable low-level, total haloxyfop analysis in a variety of challenging matrices, including infant formulas and their important high-carbohydrate, high-protein, high-fat, and emulsifier ingredients. The final method was validated in two different laboratories by fortifying samples with haloxyfop and haloxyfop-methyl, which was used as a model compound simulating bound forms of the analyte. Mean recoveries of haloxyfop across all fortification levels and evaluated matrices ranged between 92.2 and 114% with repeatability, within-lab reproducibility, and reproducibility RSDs ≤ 14%. Based on the validation results, this method was capable to convert the haloxyfop ester into the parent acid in a wide range of sample types and to reliably identify and quantify total haloxyfop at the target 0.003 mg/kg level in infant formulas (both powdered and ready-to-feed liquid forms).

Graphical abstract

LC-MS/MS-based workflow for the determination of the total haloxyfop in infant formula and related ingredients

Keywords

Pesticides Phenoxy acid herbicides Alkaline hydrolysis QuEChERS Infant formula Liquid chromatography-tandem mass spectrometry 

Introduction

Haloxyfop is an acidic compound belonging to the group of aryloxyphenoxypropionic acid herbicides. It is used for the control of grass weeds in broad-leaf crops that might be fed to livestock. Acidic pesticides are applied in crop production either as parent acids or as esters linked to a variety of alcohol groups. Most esters degrade quickly in soil or plants to their parent acids. Acids covalently bind to matrix components via ester-, glycoside-, or other bonds to form secondary conjugated residues in crops [1, 2]. Based on metabolism studies, the European Food Safety Authority (EFSA) reported the occurrence of haloxyfop residues in goat milk and animal tissues. In milk, haloxyfop was mainly present as haloxyfop esters incorporated into lipids (triglycerides) [2, 3] and found to concentrate into fatty fractions of milk fat, body fat, and egg yolk in livestock [2].

According to the European Commission (EC) directive 2006/141/EC [4], haloxyfop and its esters should not be used in agricultural production intended for the production of infant formulae, and the haloxyfop residue level should not exceed 0.003 mg/kg in ready-to-feed infant formula for the control purposes. The residue definition for monitoring of haloxyfop is “sum of haloxyfop, its salts and esters including conjugates, expressed as haloxyfop” [4]. To be able to measure all components included in the residue definition, ester cleavage or deconjugation must be part of the sample preparation step in order to release haloxyfop (parent acid) from these bound forms [2, 3]. Considering the above, the total haloxyfop analysis cannot be performed within typical multiclass, multiresidue analysis methods because hydrolysis step affects determination of many other pesticides. A European Union Reference Laboratory (EURL) previously reported a single class method for the analysis of acidic pesticides without a hydrolysis step [5, 6, 7, 8]. Several methods demonstrated the analysis of acidic pesticides using hydrolysis under acidic [9, 10], alkaline [11, 12, 13, 14, 15, 16], or enzymatic [1, 17] conditions. However, these methods were intended for the analysis in food crops (mostly vegetables and cereals), for which the lowest maximum residue limits (MRLs) of acidic pesticides were not below 0.01 mg/kg. To the best of our knowledge, no validated methods suitable for the total haloxyfop analysis in infant formula (milk- and soy-based) and relevant raw materials, such as soy protein isolate, maltodextrin, soybean oil, or soy lecithin, have been published.

The aim of this study was to develop and validate a reliable method based on alkaline hydrolysis incorporated with the QuEChERS (Quick, Easy, Cheap, Effective, Rugged, and Safe) sample preparation workflow [18] for determination of total haloxyfop in infant formula and related raw (ingredient) materials using ultra high-performance liquid chromatography combined with tandem mass spectrometry (UHPLC-MS/MS). The method development was mainly focused on the optimization of sample preparation conditions to enable effective analysis of total haloxyfop in various highly challenging matrices, such as infant formulas and their important high-protein, high-fat, high-carbohydrate, and emulsifier ingredients. Haloxyfop-methyl (ester form) was employed as a model compound to monitor the completeness of hydrolysis during the method development and validation. The method was successfully validated and implemented in two different laboratories and later transferred to a third pesticide testing laboratory.

Materials and methods

Chemicals and materials

Haloxyfop (CAS no. 69806-34-4) and haloxyfop-methyl (CAS no. 69806-40-2), all ≥ 95% purity, were obtained from Sigma-Aldrich (St. Louis, MO, USA). Isotopically labeled internal standard, (±)haloxyfop-d4, was obtained from Toronto Research Chemical Inc. (Toronto, Ontario, Canada).

Acetonitrile (MeCN), methanol (MeOH), and hexane used for sample extraction were HPLC grade and were obtained from Fisher Scientific (Geel, Belgium). Water and MeOH were optima LCMS grade (Fisher Scientific) used for the preparation the LC mobile phases. Ammonium acetate (LCMS grade) was obtained from Sigma-Aldrich. Sodium hydroxide (NaOH) and sulfuric acid (H2SO4) were purchased from J.T. baker (Phillipsburg, NJ, USA). Prepacked salt mixture of 4 g anhydrous magnesium sulfate (MgSO4) and 1 g sodium chloride (NaCl) was purchased from UCT (Bristol, PA, USA) and Agilent Technologies (Santa Clara, CA, USA). Minicentrifuge tubes (2 mL) containing 150 mg anhydrous MgSO4 and 50 mg C18 were purchased from UCT. Mini-UniPrep syringeless PVDF (0.45 μm) filter vials were obtained from Whatman (Florham Park, NJ, USA).

All standard solutions were prepared in MeCN. Individual standard stock solutions of haloxyfop, haloxyfop-methyl, and haloxyfop-d4 were prepared at about 2000 μg/mL. Individual intermediate standard and internal standard (ISTD) solutions were prepared at 100 μg/mL. For recovery experiments, spiking solutions of haloxyfop-methyl and haloxyfop were prepared separately at 0.3 and 0.6 μg/mL and ISTD working solution was prepared at 10 μg/mL. Mixed standard solutions containing haloxyfop and haloxyfop-methyl were prepared at 0.1 and 10 μg/mL. Calibration standard stock solutions were prepared at 2.5, 5, 10, 25, and 50 ng/mL (ISTD at 100 ng/mL) by diluting appropriate volumes of the mixed standard solutions and the ISTD working solution. All prepared standard solutions were stored at – 20 °C. Calibration standard working solutions were prepared at 0.25, 5, 1, 2.5, and 5 ng/mL (ISTD at 10 ng/mL) by adding 50 μL of each calibration standard stock solutions into vial body of a Mini-UniPrep syringeless filter vials containing 150 μL of MeCN and 300 μL of the mobile phase A.

Samples of infant formula (ready-to-feed and powdered milk- and soy-based) and soybean oil were purchased from local stores. Maltodextrin was obtained from Sigma-Aldrich, soy protein isolate (natural unflavored and non-GMO) was purchased from Now Sports (Bloomingdale, IL, USA), and soy lecithin (emulsifier) was from Modernist Pantry (online purchase).

Sample preparation procedure

The optimized sample preparation procedure entailed the following steps: (1) weigh 2.5 g of a thoroughly homogenized sample or 5 g of a ready-to-feed (liquid) infant formula into a 50-mL disposable polypropylene centrifuge tube; (2) add 30 μL of the ISTD working solution (10 μg/mL); (3) add 5 mL of MeOH and vortex to mix; (4) add 1 mL of 5 M NaOH; (5) shake the tube vigorously and place it on a platform shaker for 2 h; (6) add 1.5 mL of 2.5 M H2SO4 and vortex; (7) add 5 mL of deionized water and vortex (pH of the sample mixture ~ 2.5); (8) for high-fat samples (e.g., oils and emulsifiers), add 3 mL of hexane and vortex the tube to solubilize/disperse the sample in hexane; (9) add 10 mL of MeCN using a solvent dispenser; (10) vortex the tube to fully disperse the sample in the solvent for 1 min; (11) add 4 g of anhydrous MgSO4 and 1 g of NaCl; (12) immediately seal the tube and shake vigorously by hand (or vortex) for 1 min to prevent formation of crystalline agglomerates during MgSO4 hydration; (13) centrifuge the tube at > 2000 rcf for 5 min; (14) transfer 1 mL of the supernatant (mixed MeCN-MeOH layer) to a 2-mL minicentrifuge tube containing 150 mg of anhydrous MgSO4 and 50 mg of C18 (avoid transferring the upper hexane layer added to high-fat sample); (15) mix/vortex the extract with the sorbent for 30 s; (16) centrifuge the tube at > 10,000 rcf for 5 min; (17) transfer 160 μL of the supernatant into the vial body of a Mini-UniPrep syringeless filter vial and add 240 μL of the mobile phase A; (18) partially compress the Mini-Uniprep plunger to keep the liquid in the vial body and vortex/mix thoroughly; and (19) fully compress the filter plunger to filter the extract for the UHPLC-MS/MS analysis.

For samples that are prone to precipitation in the alkaline hydrolysis step (e.g., protein isolates), 2-fold increased volumes of MeOH, 5 M NaOH, and 2.5 M H2SO4 should be added to the sample in steps (3), (4), and (6), respectively.

UHPLC-MS/MS analysis

UHPLC-MS/MS analysis was performed using an Agilent 1290 Infinity UHPLC system with a binary pump, autosampler, and thermostated column compartment, coupled to an Agilent 6490 triple-quadrupole mass spectrometer equipped with Jet Stream electrospray ionization (ESI) source (Agilent Technologies, Santa Clara, CA, USA). The MassHunter software (version B.06) was used for the instrument control, data acquisition, and processing. The sample injection volume was 10 μL. The autosampler was kept at 4 °C during analysis. A Zorbax Eclipse Plus C18 RRHD column (2.1 × 50 mm; 1.8-μm particle size) maintained at 35 °C was employed for the LC separation. The mobile phase was composed of (A) 10 mM ammonium acetate in water-MeOH (95:5, v/v) and (B) 10 mM ammonium acetate in water-MeOH (99:1, v/v). Gradient elution was performed as follows: 0–5 min: 0–100% B; 5–8 min: 100% B; 8–8.5 min: 100–0% B; and 8.5–9 min: 0% B. The flow rate of the mobile phase was 0.5 mL/min. A divert valve was placed between the column outlet and MS source to eliminate the introduction of co-extracted matrix components into the MS instrument before and after the elution of haloxyfop and haloxyfop-methyl, respectively. The divert valve positions were set as follows: 0–3 min to waste; 3–5 min to MS; and 5–9 min to waste.

The MS detection was performed in positive and negative ESI modes by monitoring at least two most selective and sensitive multiple reaction monitoring (MRM) transitions. Table 1 shows analyte-specific MS/MS conditions and LC retention times. The MS ion source conditions were as follows: drying gas, N2 (220 °C, 14 L min−1); nebulizer gas, N2 (40 psi); sheath gas, N2 (325 °C, 11 L min−1); capillary voltage, 3000 V; nozzle voltage, 500 V (ESI+)/0 V (ESI-); high pressure RF, 150; low pressure RF, 60; and cell accelerator voltage, 4 V.
Table 1

Analyte-specific LC-MS/MS conditions including retention time (RT)

Analyte

Polarity

RT (min)

Precursor ion (m/z)

Product ion (m/z)

Collision energy (V)

Dwell time (ms)

Haloxyfop

ESI−

3.98

360.0

288.0

13

5

360.0

251.9

28

5

360.0

195.9

48

5

360.0

34.9

30

5

Haloxyfop-d4 (ISTD)

ESI−

3.98

364.0

292.0

12

5

364.0

255.0

28

5

364.0

196.0

40

5

Haloxyfop

ESI+

3.98

362.0

316.0

20

5

362.0

288.1

35

5

362.0

271.9

40

5

362.0

91.0

36

5

Haloxyfop-d4 (ISTD)

ESI+

3.98

366.0

320.0

20

5

366.0

319.0

20

5

366.0

291.0

28

5

Haloxyfop-methyl

ESI+

4.67

376.0

315.9

16

50

376.0

288.0

28

50

376.0

271.8

40

50

376.0

91.0

32

50

Method validation

Validation parameters investigated in this study included linearity and matrix effects, trueness (recovery) and precision (repeatability, within-laboratory reproducibility, and reproducibility), limit of quantitation (LOQ), and robustness of the method. The method validation was conducted according to the SANTE method validation guidelines for pesticide residue analysis [19].

Recovery and repeatability were validated analyzing at least five spike replicates at the following concentration levels: 0.003 and 0.03 mg/kg for infant formulas, 0.003 and 0.01 mg/kg for maltodextrin, and 0.01 and 0.03 mg/kg for other ingredient matrices (soy protein isolate, soybean oil, and soy lecithin). Within-laboratory reproducibility was evaluated by analyzing another set of five spike replicates by different analysts (1–3 additional analysts) on different days using different sets of calibration standard preparations. Reproducibility of the method was evaluated for milk-based powdered infant formula matrix based on data generated by different analysts on different days in two different laboratories. Identical instrumentation and instrument parameter settings were used in both laboratories that participated in the method validation.

The selected samples were fortified using relevant volumes of the spiking solutions containing either haloxyfop-methyl or haloxyfop at 0.3 and 0.6 μg/mL in MeCN. The purpose of using haloxyfop-methyl to spike the evaluated matrices (in addition to the separate spiking experiments with haloxyfop-free acid) was to monitor the completeness of the alkaline hydrolysis step. After fortification, the spiked samples were vortexed to mix and left at room temperature for 15 min prior to the addition of the ISTD (prior step 2 in the sample preparation procedure). Solvent-based standard solutions were used to construct calibration curves (normalized to ISTD) and calculate the spike recoveries. Matrix-matched standards were also analyzed to assess matrix effect.

Results and discussion

Sample preparation method development and optimization

To analyze haloxyfop according to its residue definition “sum of haloxyfop, its salts and esters including conjugates, expressed as haloxyfop”, alkaline hydrolysis has been demonstrated to be suitable to release the parent acid compound from its bound forms [1, 2]. Some currently available methods for the analysis of total haloxyfop and other acidic herbicides are based on the QuEChERS extraction procedure [18] with the hydrolysis step performed in either aqueous [13, 15] or aqueous-organic solutions [16] with NaOH and H2SO4 used for alkalization and subsequent pH adjustment, respectively. Considering the advantages of the QuEChERS method, which is widely used for the analysis of pesticide residues in foods, we also employed this approach to extract haloxyfop after the hydrolysis. The hydrolysis and clean-up steps had to be carefully optimized for the range of the target matrices.

Results of the initial experiments focused on optimization of alkaline hydrolysis step conditions in milk-based powdered infant formula samples spiked with haloxyfop-methyl have shown that full conversion to haloxyfop could not be achieved in aqueous solutions of these samples. Neither increase of 5 M NaOH volume added to the sample (from 0.2 up to 1 mL) nor prolongation of the incubation/shaking (hydrolysis) time from 0.5 to 2 h allowed for complete cleavage of the ester bond. Increased temperature of the reaction mixture could provide a higher conversion yield, but this would require heating of the samples, thus would complicate the sample preparation workflow and was therefore avoided. To deal with this problem, we evaluated the conversion yield of the alkaline hydrolysis performed in aqueous-methanolic sample solutions. In the optimized procedure, the infant formula samples were dissolved in MeOH followed by addition of 5 M NaOH (1 mL) and shaken for 2 h at ambient temperature. Then, the pH value of the mixture was adjusted to approximately 2.5 by addition of 1.5 mL 2.5 M H2SO4 and 5 mL water in order to keep haloxyfop in its undissociated free acid form which was subsequently extracted and partitioned into acetonitrile. As demonstrated in the method validation, the above adjustments to the sample preparation protocol led to complete conversion of haloxyfop-methyl to free haloxyfop in infant formula samples and most of other matrices evaluated in this study. Soy protein isolate was found to be prone to excessive matrix precipitation after alkalization of the extraction mixture. The sample-to-solvent ratio (2.5 g of soy protein isolate sample and 5 mL of MeOH) employed for other samples was too high in this case and resulted in formation of a thick mixture, which was difficult to shake. To eliminate this issue and allow for effective hydrolysis and extraction, we increased 2-fold the volumes of MeOH, 5 M NaOH, and 2.5 M H2SO4 as recommended in the final sample preparation procedure.

For matrices with high content of fat, it was necessary to add 3 mL of hexane to the sample prior to the QuEChERS extraction in order to reduce co-extraction of non-polar matrix components. After centrifugation, the fat and other less polar compounds were in the hexane upper layer, while the analyte, haloxyfop (parent acid), partitioned into the middle MeCN-MeOH layer, and more polar matrix components remained in the aqueous (bottom) layer.

To further purify crude acetonitrile extracts, freeze-out clean-up was used in some of the published methods [13, 15]. This approach is rather time-consuming and limits the sample throughput. Therefore, we decided to employ an alternative approach based on dispersive SPE. Dispersive SPE clean-up in the original QuEChERS includes 150 mg MgSO4 and 25 mg primary-secondary amine (PSA). MgSO4 helps to remove remaining water from the MeCN extract, further reducing/eliminating more polar matrix co-extractives. The PSA sorbent is very effective in removing fatty acids and some other matrix components that may cause analyte signal suppression in the LC-MS/MS determination. Unfortunately, PSA could not be used in this method because haloxyfop contains a carboxylic group that is retained by the PSA sorbent, resulting in decreased analyte recoveries. Therefore, instead of PSA, we included 50 mg of C18 in the clean-up step because it is beneficial for clean-up of high-fat matrices [20]. Thus, a combination of 150 mg of MgSO4 and 50 mg of C18 was employed in the final dispersive SPE clean-up protocol.

UHPLC-MS/MS method development and optimization

The LC-MS/MS analysis employed the UHPLC technique using a short narrow bore C18 column with 1.8-μm particles for fast and efficient separation. The method development mainly involved tuning for analyte-specific MS/MS conditions (shown in Table 1) and optimization of UHPLC conditions. Ammonium acetate was chosen as an additive in the mobile phase because it provided adequate sensitivity for haloxyfop in both negative and positive ESI modes. Optimization of the MS/MS conditions for haloxyfop, haloxyfop-methyl, and (±)haloxyfop-d4 was performed in both positive and negative ESI modes and included selection of precursor and production ions and optimization of collision energy using Agilent MassHunter Optimizer tool [21]. The optimization was conducted via sample injection with LC column using LC conditions similar to that of the final method.

Haloxyfop contains a carboxylic group that is easily ionized in the negative ESI mode; however, it also yielded abundant precursor and product ions in the positive ESI mode. Initially, we attempted to analyze haloxyfop using only negative ESI mode because of expected higher selectivity with less signal interferences and suppression as compared to ESI positive. Recently, LC-MS/MS technology has significantly improved in fast polarity switching which enables simultaneous data recording in both positive and negative polarity settings in a single-analysis run. Therefore, we chose to include also positive ESI transitions in the final method to maximize useful identification information that serves as additional evidence in case if haloxyfop is found in the sample. Haloxyfop-methyl showed predominant transitions in positive ESI. At least two most intensive precursor-product ion transitions were chosen for each analyte. Dwell time of each MRM transition was optimized for the final method to give at least 10 data points across peak. The preferred ESI polarity mode for haloxyfop quantification may be matrix dependent. Also, the choice of quantitative transitions maybe modified due to potential matrix interferences. Negative ESI was the preferred ionization mode with the quantification transition m/z 360.0 > 288.0 for all matrices validated in this study, except for soybean oil for which positive ESI mode was preferred with quantification transition m/z 362.0 > 316.0.

ESI source conditions were fine-tuned using the final LC gradient program and MRM transitions to achieve maximum analyte detectability. The injection volumes (5 and 10 μL) were evaluated and 10 μL was selected to provide sufficient sensitivity and precision for haloxyfop and haloxyfop-methyl at the limit of quantitation (LOQ) of 0.003 mg/kg. Figure 1 shows an example of extracted ion chromatograms of a milk-based powdered infant formula sample spiked at 0.003 mg/kg with haloxyfop-methyl, which were obtained using the optimized sample preparation procedure.
Fig. 1

Extracted ion UHPLC-MS/MS chromatograms of quantitation MS/MS transitions for haloxyfop (ESI+ and ESI−), haloxyfop-d4 (ESI+ and ESI−), and haloxyfop-methyl (ESI+) in (A) a calibration working solvent standard at 5 ng/mL (ISTD, haloxyfop-d4, at 10 ng/mL) and (B) a milk-based powdered infant formula sample spiked at 0.003 mg/kg with haloxyfop-methyl (demonstrating effective conversion of haloxyfop-methyl to haloxyfop free acid with no quantifiable signal left for haloxyfop-methyl and acceptable recoveries for haloxyfop as documented in Table 2)

To prevent carry-over between injections, we included a short system-cleaning program (in-run solvent blank injections) in each analytical run. This program started at 5 min (after the elution of haloxyfop-methyl peak) and involved triplicate 20-μL injections of MeCN while the composition of the mobile phase was at 100% B. This helped to clean the autoinjection system (needle, needle seat, injection port, and valve) from the potential analyte and matrix residues during the actual run without increasing the analytical cycle time. Subsequently, the mobile phase composition changed to the initial condition and the column was equilibrated for the next injection.

Interlaboratory method validation

The method validation was conducted in infant formulas and related raw materials in two different routine testing laboratories using the SANTE method validation guidelines and criteria [19]. The following parameters were evaluated: linearity and matrix effects, trueness (recovery) and precision (repeatability, within-lab reproducibility, and reproducibility), limit of quantitation (LOQ), and robustness of the method.

Linearity and matrix effects

Matrix-matched and solvent-based calibration curves of 0.25, 0.5, 1, 2.5, and 5 ng/mL were prepared, using (±)haloxyfop-d4 (at 10 ng/mL) as the internal standard. Calibration curves were constructed by plotting analyte concentration versus the analyte-to-internal standard area response factor. The data was fit with either linear or quadratic fit using 1/x weighting. The coefficient of determination (r2) values obtained during the validation on different analysis days were all > 0.995 with back-calculated standard concentrations with no more than 20% relative difference from the theoretical concentrations at all calibration levels.

The matrix effect was calculated as the difference between the linear best-fit slopes of the solvent-based and matrix-matched calibration curves divided by the slope of the solvent-based calibration curve. Matrix effects were evaluated in milk-based infant formula (both powder and ready-to-feed) and soy-based infant formula. For haloxyfop, the matrix effect was negligible, ranging between 5.7 and 8.9% for haloxyfop in both negative and positive ESI modes, in all cases. Therefore, solvent-based calibration curves were used for the evaluation of haloxyfop recoveries during the method validation and have been employed in the routine analysis.

Trueness and precision

Trueness (recovery) and precision (repeatability, within-lab reproducibility, and reproducibility) were evaluated for total haloxyfop using blank matrices fortified with haloxyfop-methyl at two spiking levels in at least five replicates. Haloxyfop-methyl (methylester of haloxyfop) was used as a model compound to demonstrate acceptable conversion (hydrolysis) of esters and other bound forms of haloxyfop to the free acid form and its recovery during the extraction and clean-up. Additionally, the presence of haloxyfop-methyl was monitored in sample extracts. No detectable amounts of haloxyfop-methyl were present in any of the infant formula or ingredient-spiked samples, indicating the hydrolysis completeness. Limited experiments employing samples fortified with haloxyfop were also performed to demonstrate the performance of the extraction and clean-up steps alone.

For the calculation of haloxyfop recoveries in samples spiked with haloxyfop-methyl, obtained concentrations were multiplied by a factor of 1.04 (haloxyfop-methyl-to-haloxyfop molecular weight ratio) to compensate for the difference between molecular weight of both compounds. As shown in Table 2, mean recoveries of released and spiked haloxyfop across all fortification levels and evaluated matrices ranged between 92.2 and 114% with repeatability, within-lab reproducibility, and reproducibility RSDs ≤ 14%, meeting the SANTE validation criteria and demonstrating acceptable method performance.
Table 2

Mean recovery and precision (repeatability for n = 5, within-lab reproducibility for n = 10–30, and reproducibility for n = 35) obtained for blank samples spiked with (A) haloxyfop-methyl or (B) haloxyfop (free acid)

Matrix

Spiking level (mg/kg)

Lab

A. Spike with haloxyfop-Me

B. Spike with haloxyfop

n

Recovery (%)

RSD (%)

n

Recovery (%)

RSD (%)

Milk-based powdered infant formula

0.003a

1

5b

114

3.7

N/A

N/A

N/A

2

30c

106

11

30c

104

14

0.030

2

30c

92.3

7.6

30c

91.6

5.5

Milk-based ready-to-feed infant formula

0.003

2

30c

92.2

12

30c

94.9

9.7

0.030

30c

95.2

9.0

30c

97.0

9.4

Soy-based ready-to-feed infant formula

0.003

2

30c

94.4

10

30c

96.0

7.3

0.030

30c

99.9

12

30c

99.6

7.4

Maltodextrin

0.003

1

10d

108

4.7

5b

107

1.3

0.010

10d

101

4.0

5b

95.2

4.4

Soybean oil

0.010

1

5b

109

6.4

N/A

N/A

N/A

0.030

5b

108

5.9

N/A

N/A

N/A

Soy protein isolate

0.010

1

5b

115

6.8

N/A

N/A

N/A

0.030

5b

104

6.2

N/A

N/A

N/A

Soy lecithin

0.010

1

5b

111

4.7

N/A

N/A

N/A

0.030

5b

108

3.7

N/A

N/A

N/A

N/A value not available, experiment not performed

aOverall recovery and RSD calculated based on pooled data generated by both labs were 107 and 11%, respectively

bExperiments performed by one analyst on one day

cExperiments performed by two analysts on three different days

dExperiments performed by two analysts on two different days

Limit of quantitation

The LOQ was determined as the lowest spiking level that met the trueness (recovery) and precision acceptance criteria in the given matrices: 0.003 mg/kg for powdered and ready-to-feed (liquid) infant formulas and high-carbohydrate ingredients (maltodextrin); and 0.01 mg/kg for all other matrix types (see Table 2).

Robustness

Robustness of the method was evaluated by altering the parameters of the sample preparation (hydrolysis) procedure: volume of 5 M NaOH added to the samples (original volume 1 mL; altered volume 0.5 mL) and hydrolysis time (original time 2 h; altered time 1.5 h). The robustness testing was performed with the use of milk-based powdered infant formula spiked with either haloxyfop-methyl or haloxyfop at 0.03 mg/kg (n = 6).

The original and altered parameter settings provided acceptable individual recoveries (81.0–110%) and RSDs (4.2–8.4%) for both released and spiked haloxyfop. The spiking with haloxyfop-methyl was more critical for the evaluation of the two parameters related to the hydrolysis step. The method was shown to be robust towards modification of hydrolysis time. Average recoveries of released haloxyfop with 2- vs. 1.5-h hydrolysis times were 90.8% (6.5% RSD) comparing to 92.1% (4.2% RSD), respectively. In contrast, a considerable difference was observed for mean recoveries of released haloxyfop obtained with the original (1 mL) and altered (0.5 mL) volume of 5 M NaOH: 84.6% (4.4% RSD) vs. 97.7% (7.2% RSD), respectively. This was obviously due to the reduced conversion of haloxyfop-methyl to the free acid. Therefore, the volume of 5 M NaOH is an important parameter, which should not be altered and needs to be highlighted as critical in the method SOP for the routine analysis.

Conclusions

The method validation results demonstrate that the developed UHPLC-MS/MS-based method meets the data quality requirements and is suitable for the analysis of total haloxyfop in infant formulas and their important ingredients. It can reliably identify and quantify total haloxyfop at the target 0.003 mg/kg level in infant formulas (both powdered and ready-to-feed liquid forms) to meet the 2006/141/EC directive [4]. In addition, the method was also successfully validated in various baby foods (to meet the 2006/125/EC directive [22]), cereals, nutritional bars, and multivitamin tablets (data not presented or discussed in this article). The sample preparation procedure was optimized for effective conversion of esters and other bound forms of haloxyfop to the free acid (target analyte) form in a wide range of challenging matrix types. The method has been interlaboratory validated and is currently being used in three routine pesticide testing laboratories in the USA, EU, and Asia.

Notes

Acknowledgements

The authors wish to thank Zdenka Veprikova, Hana Novotna, and Lucie Drabova from the University of Chemistry and Technology in Prague, Czech Republic, for their contribution to the method development. Jean-Francois Halbardier, Erika Deal, and Shi Ting Ong from Covance Food Solutions are acknowledged for their contribution and support to the method validation.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

References

  1. 1.
    Anastassiades M, Roux D, Mack D, Sigalov I. Advance in the analysis of acidic pesticides. EPRW 2010 Poster presentation; 2010. (http://cvuas.xn--untersuchungsmter-bw-nzb.de/pdf/AcidicPA134.pdf). Accessed Sep 2015.
  2. 2.
    European Food Safety Authority (EFSA). Reasoned opinion on the review of the existing maximum residue levels (MRLs) for haloxyfop-P according the Article 12 of Regulation (EC) No 396/2005. EFSA J. 2014;12:8–60.Google Scholar
  3. 3.
    FAO Manual on the submission and evaluation of pesticide residues data for the estimation of maximum residue levels in food and feed, 1st Edition 2009. (http://www.fao.org/fileadmin/templates/agphome/documents/Pests_Pesticides/JMPR/FAO_manual2nded_Oct07.pdf). Accessed Oct 2017.
  4. 4.
    Commision Regulation No. 2006/141/EC of 22 December 2006 on infant formulae and follow-on formulae and amending Directive 1999/21/EC. Off J Eur Union, 30.12.2006, L 401/1.Google Scholar
  5. 5.
    EU Reference Laboratory for Residues of Pesticides. Analysis of acidic pesticides using QuEChERS (EN15662) and acidified QuEChERS method (reported by EURL-SRM, version 1, last update). 2015;20:05.Google Scholar
  6. 6.
    EU Reference Laboratory for Residues of Pesticides. Phenoxyalkanoic acids in milk using modified QuEChERS method (version 1, last update). 2014;10:04.Google Scholar
  7. 7.
    EU Reference Laboratory for Residues of Pesticides. Analysis of phenoxyalkanoic acids in milk using QuEChERS method and LC-MS/MS (version 2, last update) 05.05.2014.Google Scholar
  8. 8.
    Santilio A, Stefanelli P, Dommarco R. Fast determination of phenoxy acid herbicides in carrots and apples using liquid chromatography coupled triple quadrupole mass spectrometry. J Environ Sci Health B. 2009;44:584–90.CrossRefPubMedGoogle Scholar
  9. 9.
    Crosby DG. Metabolism of herbicides, metabolites of 2,4-dichlorophenoxyacetic acid (2,4-D) in bean plants. J Agric Food Chem. 1964;12:3–6.CrossRefGoogle Scholar
  10. 10.
    Bristol DW, Cook LW, Koterba MT, Nelson DC. Determination of free and hydrolyzable residues of 2, 4-dichlorophenoxyacetic acid and 2, 4-dichlorophenol in potatoes. J Agric Food Chem. 1982;30:137–44.CrossRefGoogle Scholar
  11. 11.
    Steinwandter H. Contributions to the analysis of chlorophenoxy acids – I. A simple method for the determination of five chlorophenoxyalkane carboxylic acids in cereals. Fresenius J Anal Chem. 1989;334:133–5.CrossRefGoogle Scholar
  12. 12.
    Schaner A, Konecny J, Luckey L, Hickes H. Determination of chlorinated acid herbicides in vegetation and soil by liquid chromatography/electrospray-tandem mass spectrometry. J AOAC Int. 2007;90:1402–10.PubMedGoogle Scholar
  13. 13.
    EU Reference Laboratory for Residues of Pesticides. Analysis of acidic pesticides in wheat flour samples by LC-MS/MS using the QuEChERS method, incl. optional alkaline hydrolysis to release covalently bound compounds (last update) 07.05.2007).Google Scholar
  14. 14.
    Santilio A, Stefanelli P, Girolimetti S, Dommarco R. Determination of acidic herbicides in cereals by QuEChERS extraction and LC/MS/MS. J Environ Sci Health B. 2011;46:535–43.PubMedGoogle Scholar
  15. 15.
    Sack C, Vonderbrink J, Smoker M, Smith RE. Determination of acid herbicides using modified QuEChERS with fast switching ESI+/ESI LC-MS/MS. J Agric Food Chem. 2015;63:9657–65.CrossRefPubMedGoogle Scholar
  16. 16.
    Steinborn A, Alder L, Spitzke M, Dork D, Anastassiades M. Development of a QuEChERS-based method for the simultaneous determination of acidic pesticides, their esters, and conjugates following alkaline hydrolysis. J Agric Food Chem. 2017;65:1296–305.CrossRefPubMedGoogle Scholar
  17. 17.
    Løkke H. Analysis of free and bound chlorophenoxy acids in cereals. Bull Environ Contam Toxicol. 1975;13:730–6.CrossRefPubMedGoogle Scholar
  18. 18.
    Anastassiades M, Lehotay SJ, Stajnbaher D, Schenck FJ. Fast and easy multiresidue method employing acetonitrile extraction/partitioning and “dispersive solid-phase extraction” for the determination of pesticide residues in produce. J AOAC Int. 2003;86:412–31.PubMedGoogle Scholar
  19. 19.
    Guidance document on analytical quality control and method validation procedures for pesticides residues analysis in food and feed. SANTE/11945/2015 Supersedes SANCO/12571/2013 Implemented by 01/01/2016.Google Scholar
  20. 20.
    Lehotay SJ, Mastovska K, Yun SJ. Evaluation of two fast and easy methods for pesticide residue analysis in fatty food matrixes. J AOAC Int. 2005;88:630–8.PubMedGoogle Scholar
  21. 21.
    Mastovska K, Zulkoski J, Zweigenbaum J. Triggered MRM LC/MS/MS method development-practical considerations for MRM optimization using Agilent MassHunter Optimizer software. In: Agilent Technologies publication number 5991-7195EN; 2017.Google Scholar
  22. 22.
    Commission Regulation No. 2006/125/EC of 5 December 2006 on processed cereal-based foods and baby foods for infants and young children. Off J Eur Union, 6.12.2006, L 339/16.Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Urairat Koesukwiwat
    • 1
  • Lukas Vaclavik
    • 2
  • Katerina Mastovska
    • 3
  1. 1.Covance (Asia) Pte. Ltd., Covance Food SolutionsSingaporeSingapore
  2. 2.Covance Laboratories Inc., Covance Food SolutionsNorth YorkshireUK
  3. 3.Covance Laboratories Inc.Covance Food SolutionsMadisonUSA

Personalised recommendations