Formatted Title
Untargeted PFAS Analysis and Forensics: Promise and Bottlenecks to Widespread Adoption from a Commercial Lab Perspective
Background/Objectives
As PFAS measurement needs expand from targeted analysis of a select group of PFAS to cover the thousands of uncharacterized/partially characterized entities, the use of high-resolution accurate mass (HRAM) analysis has increased significantly. From the discovery of previously uncharacterized factory emissions in North Carolina and New Jersey, to their use in source characterization, source attribution and forensics, the vast majority if this work has been carried out in academic, government and non-profit institutions. While this work continues to provide valuable information on uncharacterized PFAS and suspect screening, scale on adoption of these techniques will need development of capabilities at commercial and contract laboratories. Our objectives with this study were to understand and elucidate the major challenges with untargeted PFAS workflows from a commercial lab perspective such as 1) technical instrument capabilities, sensitivity of the instruments, speed, resolution and capabilities to generate actionable information on suspect screening workflows and untargeted workflows, 2) data analysis and reporting workflows, 3) standardization and benchmarking for untargeted PFAS analysis, and 4) alignment of technical needs of workflow with analyst skillset available in a commercial lab.
Approach/Activities
We selected three different time-of-flight and ion trap instruments as part of this evaluation. We used calibration standards from EPA 1633 draft to assess capabilities of instrument to detect the 40 PFAS that are part of 1633 and to understand the sensitivity of the instrument as assessed by calibration standards. We also provided biosolids extracts which had already been characterized using EPA 1633 and known to contain high levels of uncharacterized PFAS in order to assess the targeted, suspect screening and untargeted workflows. All instrument and data analytical work was done by the instrument vendor technical groups at their facilities and data analysis reports were prepared by the instrument teams. We then assessed reports provided to benchmark instrument capabilities, the data analysis workflows for target analysis, suspect screening and untargeted analysis. We also performed a literature review to assess the status of benchmarking and quality control of non-target analysis workflows.
Results/Lessons Learned
The preliminary results are available at the time of writing the abstract. Preliminary results indicate that all instruments were in general able to detect the 40 target PFAS in 1633 at the default low calibration level with a few exceptions. There were instrument-specific differences in relative sensitivity by PFAS structure type. With suspect screening workflows, the number of available PFAS with previously compiled information, integration with publicly available databases and general ease of use of workflow varied significantly between the options tested. Results for the untargeted workflow are being compared using Kendrick Mass Defect results, number of features detected, Schymanski scale evaluation for confidence of detection, ease of use of instrument-specific software, integration with publicly available software such as MZmine and Fluoromatch and will be presented.