Comparison of LC/MS/MS and GC/MSD Analyses of Pesticide Residues in Food Products when Using the QuEChERS Sample Preparation Technique

Transcription

1 Comparison of LC/MS/MS and GC/MSD Analyses of Pesticide Residues in Food Products when Using the QuEChERS Sample Preparation Technique Monika Kansal, A. Carl Sanchez, Michael Rummel, and Art Dixon Phenomenex, Inc., 411 Madrid Ave., Torrance, CA USA PO _W_1

2 Introduction QuEChERS, an acronym for Quick, Easy, Cheap, Effective, Rugged, and Safe, is a sample preparation procedure commonly used for the extraction and cleanup of multiple pesticide residues from food and agricultural products. This relatively simple procedure provides extracts suitable for both LC/MS and GC/MS analyses. In this study the results obtained when analyzing the same extracts from multiple food products using both GC/MSD and LC/MS/MS are presented and compared. The relative strengths of each instrument platform are highlighted and considerations for choosing the most appropriate platform are presented.

3 Results and Discussion Iceberg Lettuce Extraction The total number of possible failures is % passing is difficult to achieve The most problematic analytes are discussed in this study Analytes LC/MS LC/MS LC/MS GC/MS GC/MS GC/MS Recovery % RSD Linearity Recovery %RSD Linearity Atrazine % Azoxystrobin % Bifenthrin N/A N/A N/A 108 6% Carbaryl % Chlorothalonil N/A N/A N/A 103 7% Chlorpyrifos % Chlorpyrifos-methyl % Cyprodnil % Dichlorvos % Endosulfan sulfate N/A N/A N/A 110 5% Ethion % Imazalil * % Imidacloprid N/A N/A N/A Kresoxim-methyl % Linuron N/A N/A N/A L-Cyhalothrin N/A N/A N/A 110 8% Methamidophos * % Methomyl N/A N/A N/A o,p-ddd N/A N/A N/A 100 5% o-phenolphenyl N/A N/A N/A 103 5% Permethrins N/A N/A N/A 112 8% Procymidone N/A N/A N/A 107 5% Pymetrozine N/A N/A N/A Tebuconazole % Thiabendazole * % Tolyfluanid % Trifluralin N/A N/A N/A 101 5% Recovery LC/MS- All analytes (18) passed Association of Analytical Chemists AOAC criteria except Dichlorvos All calculations based on External Standardization (ESTD) Pymetrozine is low as expected by AOAC method GC/MS- All analytes (23) showed good recovery All calculations based on Internal Standardization (ISTD) % RSD for sample replicates LC/MS- All analytes passed AOAC criteria except Dichlorvos Based on 5 sample replicates GC/MS- All analytes passed AOAC criteria Based on 5 sample replicates Correlation coefficient (linearity) LC/MS - All analytes passed AOAC criteria except Dichlorvos All calculations based on ESTD GC/MS - Linearity criteria was low for a few analytes Sensitivity for methamidophos, imazalil and thiabendazole was inadequate at low levels

4 Spinach Extraction Pigmented matrices require graphitized carbon black (GCB) GCB is known to adsorb planar molecules* Procymidone suffered loss in GC/MS due to matrix co-elution** Optimized analysis conditions as determined in this study Analytes LC/MS LC/MS LC/MS GC/MS GC/MS GC/MS Recovery %RSD Linearity Recovery %RSD Linearity Atrazine % Azoxystrobin % Bifenthrin N/A N/A N/A 92 4% Carbaryl % Chlorothalonil * N/A N/A N/A 26* 9% Chlorpyrifos % Chlorpyrifos-methyl % Cyprodnil * 14* * 6% Dichlorvos N/A N/A N/A 81 8% Endosulfan sulfate N/A N/A N/A 105 6% Ethion % Imazalil N/A N/A N/A Imidacloprid N/A N/A N/A Kresoxim-methyl % Linuron N/A N/A N/A L-Cyhalothrin N/A N/A N/A 108 9% Methamidophos N/A N/A N/A Methomyl N/A N/A N/A o,p-ddd N/A N/A N/A 102 4% o-phenolphenyl N/A N/A N/A 94 4% Permethrins N/A N/A N/A 95 4% Procymidone** N/A N/A N/A 46** 33% Pymetrozine * 10* N/A N/A N/A Tebuconazole % Thiabendazole * 12* N/A N/A N/A Tolyfluanid % Trifluralin N/A N/A N/A 93 3% Thiabendazole* Pymetrozine* Cyprodnil*

5 LC/MS - Challenges and Solutions Problematic Analytes Injection solvent mismatch AOAC injection solvent 25 % Acetonitrile (ACN) peak distortion Early eluters gave split peaks (pymetrozine, methamidophos, thiabendazole) Reduced solvent strength - sample precipitation Filter membranes absorb analytes to various degrees Mobile Phase Limited flexibility Increase initial % organic - early eluters not retained Early eluters unretained in some published papers Decrease initial % organic reconstitution solvent mismatch Matrix effects Each analyte responds differently to matrix interferences One internal standard for all analytes does not give reliable results

6 LC/MS Injection Solvent - Peak Distortion Challenge Early eluting peaks distorted with 25 % acetonitrile injection solvent (as suggested by AOAC) Initial mobile phase: 5 mm Formic Acid (FA) in 25 % methanol (MeOH) Intensity, cps Methamidophos Sample dissolved in 25 % ACN 5mM FA As per AOAC min App ID Solution Intensity, cps 1.04e4 1.00e Pymetrozine, split peaks Reduced solvent strength 25 % MeOH Peak splitting issue resolved Precipitation 3.82 Intensity, cps Sample dissolved in 25 % ACN 5mM FA As per AOAC min Methamidophos Sample dissolved in 25 % MeOH 5mM FA min App ID App ID Intensity, cps 4.2e4 4.0e4 3.8e4 3.6e4 3.4e4 3.2e4 3.0e4 2.8e4 2.6e4 2.4e4 2.2e4 2.0e4 1.8e4 1.6e4 1.4e4 1.2e4 1.0e Pymetrozine Sample dissolved in 25 % MeOH 5mM FA min App ID 20674

7 LC/MS Injection Solvent Sample Precipitation (Filtration) Challenge Matrix components precipitate when 25 % methanol is used 5 ng/g Cal Standard prepared in 25 % Methanol 5 mm Formic acid All filter membranes absorb analytes different for each filter Below is the % loss with filtration for the most problematic analytes Solution Samples injected without filtration (centrifuged) No problems with column fouling % Loss % Loss on Filtration-5 ng/g cal standard Chlorpyrifos Ethion Imazalil 1 Methamidophos 2 Thiabendazole 1 D10-Parathion % Loss with GHP filter % Loss with Nylon filter % Loss with RC filter % Loss with PTFE filter % Loss with PTVDF filter

8 LC/MS Limited Flexibility of Mobile Phase Different mobile phase conditions (JCA Article (2010)) Column: Prodigy ODS-3 5 µm Dimensions: 150 x 3.0 mm Mobile Phase A: Water with 0.1 % Formic acid B: Acetonitrile with 0.1 % Formic acid Gradient: 30 to 100 % B in 8 min, 5.5 min final hold at 100 % B Flow Rate: 0.3 ml/min Instrument: MS: API4000 (ESI+), LC: Agilent 1200 Sample solvent: 40 % ACN in water with 0.4 % Acetic acid Intensity, cps 1. 20e e e e e Pymetrozine unretained (t o =1.8 mins ), RT=1.79 mins min App ID Mobile phase/sample solvent mismatch (AOAC 2007) Column: Luna C18(2) 3 µm, Dimensions: 150 x 3.0 mm Mobile Phase A: Water with 5 mm Formic acid B: Methanol with 5mM Formic acid Gradient: 25 to 90 % B in 15 mins, 15mins final hold at 100 % B Flow Rate: 0.3 ml/min Instrument: MS: API4000 (ESI+), LC: Agilent Sample Solvent: 25 % ACN in water with 0.25 % Acetic acid Intensity, cps Pymetrozine retained RT=3.04 mins, but splitting min App ID 20675

9 LC/MS Matrix Effects Matrix interferences affect the internal standard (d10-parathion as per AOAC) differently than analytes % RSD ISTD >> % RSD analyte -> matrix effects with ISTD Stable label ISTD for all analytes not generally used in multi-residue methods (GC/MS) Standard practice for LC/MS due to matrix effects Effect of Matrix Interference in QC Sample LC/MS Analysis for Lettuce Carbaryl % RSD=2 % Cyprodnil % RSD=2 % Ratio IS/TPP % RSD=17 % D10-IS % RSD=17 % TPP% RSD=2 %

10 GC/MS - Challenges and Solutions Problematic Analytes Overall fewer problems than LC/MS Matrix interferences Co-eluting matrix components Some matrices more problematic than others Chromatographic degradation (inlet fouling, losses) Accumulation of matrix components Poor peak shape Reduced sensitivity

11 GC/MS Interferences - Methamidophos Early eluter co-elutes with interference Analyze by LC/MS for optimum results 50 ng/g Cal standard (apple extract) Blank (apple extract) Interference Methamidophos min Interference min App ID App ID 20671

12 GC/MS Inlet Losses - Imazalil in Different Matrices Accumulation of matrix components in inlet liner leads to poor peak shape and reduced sensitivity Analyze by LC/MS for optimum results Lettuce extract 50 ng/g Cal standard Imazalil min Spinach extract 50 ng/g Cal standard Imazalil min

13 Advantages of LC/MS Over GC/MS Analysis Methamidophos, imazalil and thiabendazole show better performance with LC/MS than GC/MS Better peak shape in LC/MS Higher signal intensity in LC/MS Good linearity and % RSD with LC/MS

14 Advantages of LC/MS GC/MS Analysis Analyze by LC/MS for optimum results Methamidophos using LC/MS and GC/MS LC/MS 100 ng/g (lettuce extract) Intensity, cps 6.2e4 6.0e4 5.5e4 5.0e4 4.5e4 4.0e4 3.5e4 3.0e4 2.5e4 2.0e4 1.5e4 1.0e Methamidophos min GC/MS 100 ng/g (lettuce extract) Interference Methamidophos min Thiabendazole using LC/MS and GC/MS LC/MS 5 ng/g (lettuce extract) GC/MS 5 ng/g (lettuce extract) 3.8e4 3.6e4 3.4e4 3.2e4 3.0e4 2.8e4 2.6e4 2.4e4 2.2e4 2.0e4 1.8e4 1.6e4 1.4e4 1.2e4 1.0e Thiabendazole min Thiabendazole min

15 Advantages of GC/MS Analysis Over LC/MS Analysis Internal standardization possible In general, GC/MS does not suffer from compound specific matrix effects Solubility No solubility problems, all analytes are soluble in toluene and acetonitrile Dramatically improved sensitivity for procymidone vs. LC/MS (no fragmentation with CID) Dichlorvos shows reproducible results in GC/MS LC/MS results very poor

16 Internal Standardization Plot of D10-Parathion(internal standard) GC/MS and LC/MS/MS In GC/MS matrix generally does not affect ionization In LC/MS, co-eluting matrix components affect ionization of ISTD differently than analytes % RSD ISTD >> % RSD analyte -> matrix effects with ISTD GC/MS Analysis LC/MS Analysis Effect of matrix interferences in lettuce QC samples Cyprodnil % RSD=6 % Carbaryl % RSD=7 % TPP% RSD=7 % Carbaryl % RSD=2 % Cyprodnil % RSD=2 % Ratio IS/TPP % RSD=17 % D10-IS % RSD=17 % Ratio IS/TPP % RSD=14 % TPP% RSD=2 % D10-IS % RSD=8 % Dichlorvos improved performance Dichlorvos shows significant matrix effects in LC/MS It failed all criteria in LC/MS but passed all in GC/MS Analyze by GC/MS for optimum results Analyte results using both GC/MS and LC/MS techniques Analytes LC/MS-Lettuce LC/MS-Lettuce LC/MS-Lettuce GC/MS-Lettuce GC/MS-Lettuce GC/MS-Lettuce Recovery % RSD Linearity Recovery % RSD Linearity Atrazine % Azoxystrobin % Carbaryl % Chlorpyrifos % Chlorpyrifos-methyl % Cyprodnil % Dichlorvos % Ethion % Imazalil % Kresoxim-methyl % Methamidophos % Tebuconazole % Thiabendazole % Tolyfluanid % 0.992

17 Conclusions Results for all analytes met AOAC performance expectations by carefully optimizing analysis conditions Most appropriate platform selected for each analyte. Best technique, GC or LC, is analyte specific Methamidophos, imazalil and thiabendazole - best results with LC/MS Dichlorvos and procymidone - best results with GC/MS Injection solvent and solubility LC/MS -reconstitute samples with 25 % MeOH without filtering GC/MS -no solubility problems Matrix interferences External standardization for LC/MS analysis (if no stable label ISTDs) Internal standardization for GC/MS analysis

18 Trademarks Luna is a registered trademark, and Prodigy is a trademark of Phenomenex. Agilent is a registered trademark of Agilent Technologies. API 4000 is a trademark of AB SCIEX pte Ltd. AB SCIEX is being used under license. Disclaimer Phenomenex is not affiliated with Agilent Technologies. Comparative separations may not be representative of all applications Phenomenex, Inc. All rights reserved.