EPA 1633 - What Happens When Automated Solvent Extraction and Robust LC-MS/MS are Coupled for Soil Analysis

Per- and Polyfluoroalkyl Substances (PFAS) in the Environment
Oral Presentation

Prepared by R. Marfil-Vega1, B. Liu2, A. Stell3, O. Shrestha1, K. Luo1, M. Davis1, L. Wiest1
1 - Shimadzu Scientific Instruments, 7102 Riverwood Drive, Columbia, MD, 21046, United States
2 - CEM, 3100 Smith Farm Rd, Matthews, NC, 28104, United States
3 - CEM, 3100 Smith Farm Rd, Metthews, NC, 28104, United States


Contact Information: [email protected]; 410-910-0884

ABSTRACT

With the increasing interest in meeting regulatory requirements and understanding PFAS levels in various sample types, automated workflows are essential for improving lab productivity. In this work we evaluated the performance of an automated solvent extraction system for soils extraction, coupled with robust LC-MS/MS for PFAS analysis according to EPA 1633. The EDGE PFAS was used for the extraction of Ottawa sand samples in a series of experiments to assess background PFAS, extraction recovery, and method detection limits, among others. Samples were spiked with the Extracted Internal Standards before extraction, and extracts were subsequently cleaned-up as described in EPA 1633. Three extraction and instrumentation replicates were performed to evaluate robustness. The 40 target compounds were chromatographically separated on a C18 column by gradient elution with 2 mM ammonium acetate in water and 100% acetonitrile. A C18 delay column was used to decrease the interference from native PFAS from the system. After addition of the Non-Extracted Internal Standards, samples were then analyzed and data processed using the Shimadzu LCMS-8060NX and LabSolutions Insight. Since the MDLs obtained for aqueous samples with the same LCMS are up to 10x lower than those listed in EPA 1633, the calibration curve started at a lower concentration than the Cal 1 level in the final EPA method. Average %accuracy of calibration verification standards were injected every 10 samples and ranged from 79 – 128% for all targeted PFAS analytes throughout the analysis. The small %accuracy range demonstrates the robustness of the LC-MS/MS. No PFAS were measured in the blank solvent extraction, indicating that the automated solvent extraction system along with its consumables did not contribute to background PFAS concentration. In this presentation we will share the outcomes from the performance evaluation of the workflow described above and discuss critical steps to achieve optimal system performance.