Simultaneous Determination of Paraquat and Diquat in Environmental Water Samples by HPLC-MS/MS

Oral Presentation

Prepared by R. Jack, L. Wang, X. Liu, C. Pohl
Thermo Scientific, 1214 Oakmead Parkway, Sunnyvale, CA, 94085

Contact Information: richard.jack@thermofisher.com; 408-481-4227


ABSTRACT

Paraquat (1,1’-dimethyl-4,4’-bipyridylium ion) and Diquat (1,1’-ethylene-2,2’
-bipyridyliumion) are quaternary amines widely used as herbicides for both terrestrial and aquatic plants. Due to their wide usage and moderate toxicities, their presence in runoff from application areas and in agricultural consumer products has been a major concern for aquatic life and human health. Diquat is currently regulated by the United States Environmental Protection Agency (US EPA) at 20 μg/L in drinking water. Paraquat is not currently regulated. Commonly used methods for Paraquat and Diquat analysis include ion pairing liquid chromatography (IP-LC)2,3, capillary electrophoresis4 (CE) with various detectors such as ultraviolet (UV) and mass spectrometry (MS).

This study describes an HPLC-MS/MS method for high throughput, sensitive and selective quantitation of Paraquat and Diquat in environmental waters. Without using an ion pairing reagent, sensitivity was significantly improved and detection limit scan be extended to low ng/L levels. Environmental water samples were prepared offline via anion exchange cleanup or injected directly for online solid phase extraction (SPE) using a weak anion exchange cartridge. Sufficient retention and improved chromatographic resolution were achieved with isocratic elution on a mixed mode column featuring reversed-phase, anion/cation exchange retention mechanisms. Analytical time was within 5 minutes per sample ensuring routine high throughput. SRM was used for quantitation with isotope labeled internal standards to achieve quantitation accuracy.

In this method, Paraquat and Diquat were separated within 5 minutes with chromatographic resolution greater than 7 to avoid SRM interferences and thus enhance the detection selectivity and sensitivity. Two SRM transitions for each analyte were used for quantitation and confirmation. The lower limit of quantitation (LLOQ) was determined to be 100 ng/L for both analytes. Quantitation precision was within 10% at 500 ng/L and 4% at 50 μg/L. Good linearity was achieved from 100 ng/L to 100 μg/L with a coefficient of determination (r2) greater than 0.999 for both analytes with using 1/x weighting factor. Other performance parameters such as specificity, carry over, accuracy, and matrix effect were also evaluated and are presented below. This method has been applied for various samples including drinking water and ground water.