Automated Workflows for PFAS Screening and Unknown Identification

Emerging Environmental Applications for High Resolution Mass Spectrometry
Oral Presentation

Prepared by R. Cochran1, S. Choyke2, C. Meyers2, R. Tautenhahn3
1 - Thermo Fisher Scientific, 3000 Lakeside Drive, Suite 116N, Bannockburn, IL, 60015, United States
2 - Eurofins Environment Testing, 4955 Yarrow Street, Arvada, CO, 80002, United States
3 - Thermo Fisher Scientific, 355 River Oaks Pkwy, San Jose, CA, 95134, United States


Contact Information: [email protected]; 310-633-1338

ABSTRACT

PFAS are contaminants of concern that are monitored by various regulatory agencies. However, regulatory methods using targeted quantitative analysis are based on reference standards that do not address the thousands of novel PFAS for which there are no commercially available standards. Non-targeted methods that detect and identify compounds in complex mixtures with little to no prior knowledge about the PFAS present are therefore needed to understand the true extent of contamination and to facilitate source fingerprinting for remediation. In this work we show how ultra-high mass resolution mass spectrometry and a comprehensive suite of PFAS-specific libraries, databases and in-silico tools are fundamental and critical requirements to effectively annotating unknown PFAS with the highest levels of confidence. We also show how automation workflows within powerful software further facilitate the process of characterizing the overall PFAS composition of even the most complex sample matrices.
In this work we use a unified non-targeted data acquisition and processing workflow that combines the Orbitrap Exploris 240 mass spectrometer with Compound Discoverer 3.5 software. An analysis aqueous fire fighting foam (AFFF) and AFFF-impacted soil samples obtained through the NIST PFAS Non-targeted Analysis Interlaboratory Study (NTAILS) are used to explain how leveraging a full suite of PFAS-specific databases, reference and in-silico spectral libraries, fragment databases and neutral-loss detection leads to confident annotation of unknown PFAS. Software features that enable complete and consistent implementation of the Schymanski annotation confidence scale, including definition of the priority of spectral libraries and databases used for PFAS annotation, are presented. Data comparison and visualization tools that facilitate determination of PFAS sample composition are also covered, including principal component analysis (PCA) and differential analysis plots, mass defect plots for identification of homologous series of PFAS, and molecular networks to view structurally related PFAS.