Determination of Cationic Polar Pesticides and Plant Growth Regulators Using UPLC-MS/MS with the ACQUITY UPLC BEH Amide Column

The purpose of this work is to demonstrate a single extract LC-MS/MS method for the determination of highly polar cationic pesticide residues and plant growth regulators in several food commodities that meet or exceed the MRL detection requirements in the European Commission pesticides database. The method performance study was completed on an ACQUITY UPLC I-Class System with a Xevo TQ-S micro using an ACQUITY UPLC BEH Amide Column after extraction following the QuPPe method. A method validation study was carried out on 4 representative commodities; namely apple, cucumber, flour, and potatoes. The method performance was assessed using 2 spike levels, 0.01 and 0.05 mg/kg for all analytes except maleic hydrazide which was spiked at 0.5 and 1.5 mg/kg with 5 replicates at each level. Difenzoquat and aminocyclopyrachlor were the only analytes not internally standardized. Method performance for trueness was 92 to 108% across all commodities with the exception of difenzoquat in cucumber (60–67%) where it was identified that PVDF filters were not suitable to use for this analysis. RSDs were all at or below 12%. A FAPAS QC flour sample was extracted on two occasions in triplicate one month apart and all results were within 20% of the assigned value and within the range necessary to achieve an acceptable z-score. All calibration graphs had residuals below 20% and R² values of 0.99 or higher. Retention time stability across all the method validation study batches for all analytes was lower than 3%.

Benefits

• Provides a single extraction (QuPPe) and LC-MS/MS method suitable for the determination of various highly polar cationic pesticides and plant growth regulators in cereals, fruit, and vegetable commodities to facilitate monitoring of MRL/tolerance compliance
• Offers sufficient chromatographic retention, selectivity, peak shape, and stability to comply with SANTE guidelines
• Provides sufficient sensitivity to determine residues at concentrations as low as 0.01 mg/kg in crude extracts without cleanup

There are various methods available to analyze food for pesticide residues focusing on multi-residue methods such as QuEChERS. The extraction and determination of polar, however, still remain a considerable challenge. The QuPPe (Quick Polar Pesticides) method¹ has been developed by EURL-SRM (European Union Reference Laboratory – Single Residue Method) which allows for the simultaneous extraction of many highly polar pesticides, their metabolites, and plant growth regulators. The QuPPe method focuses on using LC-MS/MS instruments offering high sensitivity in part to deal with matrix effects as there is no current generic clean-up that effectively deals with all matrix types.

For the cationic polar pesticides in this study there are a variety of MRLs (see Table 1 for a selection) which range from default MRLs of 0.01 mg/kg to 7 mg/kg for chlormequat.² In addition, there are components that are not currently part of official EU MRL residue definitions such as melamine (metabolite of cyromazine), aminocyclopyrachlor, ETU, and PTU. The monitoring of their concentrations in food is still of interest for safety reasons.

In this application note, example performance data is provided from Waters ACQUITY UPLC I-Class System and Xevo TQ-S micro on four commodities which represent high water content and high starch, low water content sample types. Organic wheat flour, cucumber, apple, and potato were extracted following the QuPPe method, to assess various performance factors of the UPLC-MS/MS method such as calibration linearity, retention time stability, method precision, trueness, and analyte identification.

Sample Description

Organic apple, cucumber, and potato were purchased from a retail outlet and finely homogenized in the laboratory and stored at 4 °C in a fridge until analysis. Organic wheat flour was purchased and stored at room temperature. 

A certified QC sample (T09127QC) from FAPAS was purchased. The sample contained a mix of polar pesticides (including chlormequat and mepiquat) with assigned values and acceptance limits for the compounds included.

Method Conditions

Homogenized organic apple, cucumber, potato, and wheat flour were extracted using the QuPPe method as shown in Figure 1. For wheat flour additional steps were carried out, freezing for 2 hours at -20 °C and reduced sample mass was used (as outlined in the QuPPe method). The supernatant from the QuPPe extracts were then filtered using a 0.45 µm PTFE filter before analysis by LC-MS/MS.

Recovery spikes were carried out for all 4 commodities with 5 replicates at 0.01 mg/kg (0.5 mg/kg for maleic hydrazide) and 5 replicates at a higher level of 0.05 mg/kg (1.5 mg/kg for maleic hydrazide). Matrix matched standards were prepared in the respective blank extracts and were spiked after filtering. Calibration ranges for apples, cucumber, and potatoes were 0.002 to 0.2 mg/kg for all analytes except maleic hydrazide which had a calibration range of 0.1 to 2 mg/kg. For flour the calibration range values were 0.004 to 0.4 mg/kg for all analytes except for maleic hydrazide at 0.2 to 4 mg/kg. Quantification of spiked samples was by matrix matched bracketed calibration. The MRMs listed in Table 2 were used in this application for quantification and confirmation of residues.

The MRMs listed in Table 2 highlight the optimized transitions used for quantification and confirmation of the cationic polar pesticides in this application. Auto dwell was used to calculate the dwell times. Using this feature, at least 12 points across the peak were acquired at the bottom calibration standard. The transition used for quantification is denoted in Table 2 in bold. For certain analyte/commodity combinations this was not always the most abundant transition, but was used to ensure consistency across the batches.

Method performance was assessed over 4 validation batches which covered the following commodities with a range of different properties; apples, cucumbers, flour, and potatoes. Each of these batches contained 5 spiked recoveries at two levels, 0.01 and 0.05 mg/kg (except maleic hydrazide where the spike levels were 0.5 and 1.5 mg/kg). Figure 2 demonstrates a typical chromatogram at the 0.02 mg/kg (0.5 mg/kg for maleic hydrazide) in a flour matrix matched calibration standard that was routinely achieved using this method during the method validation work. This level of sensitivity was achieved using a low injection volume of 0.5 µL which helps to mitigate matrix effects.

As can be seen from Figure 2, ETU and PTU elute before 2 column void volumes, but upon investigation both showed a consistent Gaussian peak shape across all validation batches with good retention time stability. The effects of matrix on response of early eluting compounds was investigated for ETU and PTU. When matrix effects were calculated for PTU they ranged between 102–145% and for ETU 95–191% without internal standardization. For all analytes that labelled internal standards were available for, the use of these corrected for matrix effects and analyte recovery. For difenzoquat and aminocyclopyrachlor, no labelled internal standards were available and matrix effects ranged between 98 to 107% and 132 to 216% respectively. The results of the 4 validation batches are presented in Figure 3. Melamine was detected at significant levels in the cucumber blank material and the results have been excluded for this residue/commodity combination in the validation data.

The results of the validation batches indicate that ETU and PTU gave acceptable performance with the use of the labelled internal standards. Aminocyclopyrachlor exceeded validation requirements without the need for a labelled internal standard. The low recoveries of difenzoquat were investigated and found to be an issue depending on which type of filters were used prior to LC-MS/MS analysis. Where PTFE filters (0.45 µm) were used, the validation results exceeded the requirements in the SANTE⁴ method validation criteria. All other analytes met or exceeded method validation criteria.

Matrix matched calibration was used throughout the method validation study and examples of calibration graphs achieved are demonstrated in Figure 4. For all analytes, each calibration graph in the validation study displayed coefficients of determination of 0.99 or higher, with residuals all <20%. These values exceed the requirements for calibration as set out in SANTE guidelines.

The results were assessed in accordance with SANTE guidelines for identification and all analytes in the recovery samples confirmed by both ion ratio and retention time. Figure 5 displays the retention times for all analytes across the 4 method validation batches and demonstrates retention time stability during the validation studies. Additional retention time stability tests were conducted on the wheat flour sample where 200 injections of a 0.02 mg/kg matrix calibration standard were run without operator intervention. Figure 6 demonstrates the 1^st and 200^th injection of selected analytes. The retention times were found to be stable and there was no significant change in the observed peak shape.

The method was tested with a FAPAS QC sample (T09127, wheat flour) containing chlormequat and mepiquat where the sample was extracted on two occasions in triplicate one month apart. The results obtained from this work are shown in Table 3. All the replicates fall within the range for the QC material with the average concentration within 20% of the assigned value for both analytes and are within range for an acceptable z-score. The within laboratory repeatability (RSD) was <2% for the 6 replicates. Both analytes were confirmed by retention time and by ion ratio in all replicates as per SANTE guidelines.⁵

The method validation study results demonstrate a robust analytical method for the determination of cationic polar pesticides using the established QuPPe extraction protocol with analysis using an ACQUITY UPLC BEH Amide Column fitted in an ACQUITY UPLC I-Class PLUS System coupled to a Xevo TQ-S micro. The results for most compounds exceed the requirements for method validation in the SANTE guidelines. The trueness and precision of this LC-MS/MS method determined at two matrix QC levels with 5 replicate injections was found to be acceptable for all compounds. Retention time stability and robustness was proven over the course of the study with RSDs for all compounds under 2% across all method performance batches. In most cases, limits of quantification surpass the MRL requirements. The short method run time of 10 minutes and utilizing the QuPPe extraction method allows a high sample throughput for the analysis of the cationic polar pesticides in various food commodities investigated.

Scientists must validate the method in their own laboratories and demonstrate that the performance is fit for purpose and meets the needs of the relevant analytical control assurance system.

1. QuPPe Method V11. European Commission (2020). https://www.eurl-pesticides.eu/userfiles/file/EurlSRM/meth_QuPPe_PO_V11(1).pdf (Accessed online December 2020).
2. EU Pesticide Database. European Commission (2020) [Online] https://ec.europa.eu/food/plant/pesticides/eu-pesticides-db_en
3. International Experts Limit Melamine Levels in Food. World Health Organization (2010) http://www.who.int/mediacentre/news/releases/2010/melamine_food_20100706/en/ (Accessed online 06 January 2021).
4. Document No. SANTE 12682/2019. Guidance Document on Analytical Quality Control and Method Validation Procedures for Pesticides Residues Analysis in Food and Feed European Union (2019, Implemented by 01/01/2020).