Determination of paraquat and diquat: LC-MS method optimization and validation

This study describes the optimization and single-laboratory validation of a single residue method for determination of two bipyridylium herbicides, paraquat and diquat, in cowpeas by UPLC-MS/MS in a total run time of 9.3min. The method is based on extraction with an acidified methanol-water mixture. Different extraction parameters (extraction solvent composition, temperature, sample extract filtration, and pre-treatment of the laboratory sample) were evaluated in order to optimize the extraction method efficiency. Isotopically labeled internal standards, Paraquat-D6 and Diquat-D4, were used and added to the test portions prior to extraction. The method validation was performed by analyzing spiked samples at three concentrations (10, 20 and 50μgkg(-1)), with seven replicates (n=7) for each concentration. Linearity (r(2)) of analytical curves, accuracy (trueness as recovery % and precision as RSD%), instrument and method limits of detection and quantification (LOD and LOQ) and matrix effects were determined. Average recoveries obtained for diquat were between 77 and 85% with RSD values ⩽20%, for all spike levels studied. On the other hand, paraquat showed average recoveries between 68 and 103% with RSDs in the range 14.4-25.4%. The method LOQ was 10 and 20μgkg(-1) for diquat and paraquat, respectively. The matrix effect was significant for both pesticides. Consequently, matrix-matched calibration standards and using isotopically labeled (IL) analogues as internal standards for the target analytes are required for application in routine analysis. The validated method was successfully applied for cowpea samples obtained from various field studies.