[co-author: Kristin Robrock, Ph.D.]
Per- and polyfluoroalkyl substances (PFAS), often referred to as “forever chemicals,” are a very diverse set of organic chemical compounds containing fluorine that have historically been used in a wide range of consumer products and industrial settings. Due to governmental concerns about the potential impact of PFAS on human health and the environment, regulation of these compounds is continuing to increase very quickly at the state and federal levels, both with respect to the ongoing presence of these compounds in products and addressing historical releases to the environment. A critical issue for such regulation is how to identify PFAS compounds, as opposed to other compounds that contain fluorine but are not actually PFAS. Because there are thousands of PFAS compounds, and cost-effective testing for each of them is not currently available, one of the most common methods for making this determination currently involves testing for Total Organic Fluorine (TOF) and treating TOF as a proxy for PFAS. However, TOF is not always a reliable way of identifying the presence of PFAS. Moreover, even where TOF analysis is reliable, it leaves open significant questions that are pertinent to compliance with many regulatory regimes. For these reasons, the use of the TOF methodology has major legal ramifications with respect to both regulatory compliance and liability risk.
TOF Is a Problematic Proxy for PFAS: Key Technical Issues
TOF is often used as a surrogate measurement for the presence of PFAS, due to the fact that analytical methods do not exist for detecting most PFAS and because naturally occurring fluorinated compounds are rare (Ruyle 2023; Petkowski 2024). In fact, many state consumer product regulations for PFAS use TOF (e.g., California Assembly Bill 1817 establishes a threshold of 100 ppm TOF for unintentionally added PFAS).
However, there is no single standard analytical method for TOF; rather there are multiple options and methodologies involved. There are two types of fluorine potentially present in samples: organic fluorine in which fluorine is bound to a carbon atom (e.g., PFAS) and inorganic fluorine, in which fluorine is not bound to a carbon atom (e.g., fluoride, which occurs naturally and is added to drinking water and toothpaste). TOF testing methods may involve analyzing both types of fluorine:
- Measuring total fluorine (organic plus inorganic), via combustion ion chromatography, and subtracting out the inorganic fluorine fraction, measured via selective fluoride electrode, to calculate the organic fluorine concentration (EPA 2021; EPA Method 9214).
- Extractable organic fluorine (EOF) involves extracting organic fluorine from a sample, either using a solvent for solid samples or solid phase extraction for aqueous samples, and analysis via combustion ion chromatography (Ruyle 2023).
- Absorbable organic fluorine (AOF) involves sorbing organic fluorine present in an aqueous matrix onto granular activated carbon and analysis via combustion ion chromatography (EPA Method 1621)
There are benefits and drawbacks to each of these methods, as they differ in their abilities to detect certain PFAS as well as their detection limits (SWQCB 2024; Forster 2023). Currently, only one of these methods has been formalized into an EPA Method (AOF per EPA Method 1621), whereas other methods rely on ASTM standard testing methods (e.g., total fluorine per ASTM D7359). In addition, there are many technical challenges with obtaining accurate TOF data:
- TOF is not always equal to total fluorine minus inorganic fluoride. This is because inorganic fluoride extraction methods using water or a buffer solution may only capture inorganic fluoride on the outside of the sample (for example, like the salt on the outside of a pretzel) or may not be able to extract inorganic fluoride bound to the matrix (e.g., fluoride bound within the pretzel dough), whereas total fluorine digests and thus measures fluorine in the entire sample (for example, the entire pretzel, dough and all). Therefore, inorganic fluoride on the inside of the material may be over-accounted for as organic fluorine.
- Different TOF methods vary in the degree to which they measure different PFAS compounds, and thus, results may vary significantly depending on the specific method selected for analysis (SWQCB 2024).
- EOF methods may not be able to extract fluoropolymers such as polytetrafluoroethylene (PTFE or Teflon), based on the extraction solvent being used (ASTM 2024)
- The State of California found that AOF may be unable to or inefficient at measuring ultra-short-chain PFAS compounds (Han et al. 2021; SWQCB 2024)
- Laboratory equipment such as glassware frequently contains fluoride, and the absorption materials used in AOF methods may contain background organic fluorine, both of which may interfere with accurate quantification (EPA Method 1621; SWQCB 2024)
Therefore, with respect to using TOF methods on consumer products, it is important to select a TOF method best suited for the expected type and concentration of PFAS that may be present.
Critically, the various TOF methods do not identify the specific fluorinated compounds present in a sample, but rather can only detect a fluorine atom that had at one point been attached to a carbon atom. Therefore, the identity of the fluorinated organic(s) present in a sample remains completely unknown.
Furthermore, as a single definition of PFAS does not exist, many of the current state regulations for PFAS in consumer products have created their own definition for PFAS as an organic compound with a single fully fluorinated carbon. Because TOF methods do not identify the original organic fluorinated compound, they cannot distinguish whether the fluorine detected originated from a compound that has a single fully fluorinated carbon or does not. Fundamentally, what this means is that, while TOF is used as a surrogate for PFAS compounds, there is no guarantee that the fluorinated organic chemicals measured using TOF are indeed PFAS compounds.
Many of the current state regulations for PFAS in consumer products distinguish between intentionally and unintentionally added PFAS. TOF methods, however, are unable to support the distinction that many consumer product regulations make. There are many potential sources of unintentionally added PFAS, including source water, recycled materials, and PFAS used in the manufacturing equipment, for example, as lubricants. TOF methods simply identify the presence of organic fluorine, but not its origin. Identification of the source of a PFAS present on a consumer product requires a lengthier and detailed evaluation of the manufacturing raw materials and processes.
The EPA has also proposed using AOF on environmental aqueous samples via the development of Method 1621 for AOF. Recently, the States of California and Hawaii have also considered using TOF for environmental samples (Hawaii 2024; SWRCB 2024). However, the EPA has acknowledged that Method 1621 is fundamentally a screening tool and not as accurate or precise as PFAS-specific analytical methods (e.g., EPA Methods 537 or 1633).
There are numerous man-made and a few naturally occurring fluorinated organic compounds that may be present in environmental samples, but are not PFAS. These non-PFAS compounds could be detected using the AOF method, thus confounding interpretation of the results for PFAS (Shoemaker 2021). Specifically, there are approximately 360 approved fluorinated pharmaceuticals (Hammel 2022). Examples include fluoxetine (Prozac) and atorvastatin (Lipitor), both of which may meet PFAS definitions (Hammel 2022). There are over 100 EPA-approved fluorinated pesticides, 66 of which meet the PFAS definition of a compound with a single fully fluorinated carbon (Donley 2024). Fipronil is an example of a pesticide that could be considered a PFAS and is commonly detected in the environment (Donley 2024). A recent analysis determined that approximately 22% of the fluorinated mass in biosolids originated from fluorinated pharmaceuticals and pesticides (Spaan 2023). In addition, there are other fluorinated compounds, such as man-made Freon™, as well as naturally-occurring compounds, such as trifluoroacetate and dichlorodifluoromethane (which is also manufactured as Freon-12) (Solomon 2016; Gribble 2005). Freon and trifluoroacetate are commonly detected in the environment (Höhener 2003; Solomon 2016). These compounds could potentially result in measurable organic fluorine using TOF methods. Therefore, the usage of TOF methods on environmental samples—where confounding non-PFAS compounds may be present—should be evaluated with caution to correctly interpret organic fluorine results as actually representing the presence of PFAS.
Addressing Regulatory Compliance and Liability Risk Associated with TOF Testing for PFAS
The problematic nature of reliance on TOF as a proxy for PFAS creates significant compliance and liability risks for the regulated community, both with respect to current manufacturing processes/products and legacy manufacturing sites undergoing environmental investigation/cleanup.
Risk Management for Current Production Processes
TOF testing for PFAS poses a major regulatory compliance risk for manufacturers of products not believed to contain PFAS. Numerous state statutes regulate PFAS in products by using a specific level of TOF detection as the trigger for applicability (e.g., 100 ppm). To the extent that a manufacturer believes they are excluded from regulation by such a statute, simply because the manufacturer’s product materials/content list and the safety data sheets (SDSs) for those materials/contents do not include a specific, known PFAS compound, they may be making an unwarranted assumption that the product is not regulated. For example, the American National Standards Institute (ANSI) standard for SDSs only requires inclusion of hazardous chemicals comprising over one percent of a product. Given that PFAS and non-PFAS fluorinated organic chemicals are often added in minute quantities, it is possible that they will not appear on an SDS, even if present in the product, absent a mandatory regulatory disclosure requirement or an effort by the SDS preparer to voluntarily disclose PFAS at very low levels.
Depending on the TOF testing method utilized by a regulator, a product containing a non-PFAS fluorinated organic chemical could well be flagged as violative of the statute, resulting in a regulatory enforcement action potentially seeking monetary penalties and/or injunctive relief (such as an order barring the sale of the product as currently formulated). Even if a manufacturer eventually convinces the enforcing agency, or a court, that the TOF testing at issue was inappropriate or misleading, the manufacturer will likely have incurred substantial legal defense costs by the end of the enforcement process. Moreover, that effort could take months or years, during which sales of the product may need to be suspended. As such, if the manufacturer is directed by a regulatory agency to test a product for PFAS, close attention should be paid to the proposed methodology, whether one of the TOF methods or otherwise. Many state regulations do not require the use of a specific type of test. In such instances, and where the proposed methodology is problematic, the manufacturer may wish to push back on the agency request, and propose the use of a methodology that is more accurate.
Moreover, if a manufacturer knows that its product contains any fluorinated organic chemical, even if not believed to be a PFAS compound, it should consider proactively testing the product via one or more TOF methods, to determine whether TOF levels in the product could trigger regulation. If so, the manufacturer will need to determine whether compliance is feasible and cost-effective, via reformulation, or if instead some type of disclosure to regulatory agencies, along with a request for an exemption (if available), should be pursued.
Similarly, a manufacturer may believe that its product does not contain any PFAS or non-PFAS fluorinated organic chemical at all. Yet the product may still contain TOF at levels triggering regulation, due to the use of recycled materials or contamination during the manufacturing process (for example, in process water or machine lubricants). Even materials marketed and labelled as “PFAS free” can be problematic if used in the product or manufacturing process, due to suppliers of those materials negligently or willfully misrepresenting the PFAS content (for example, due to contamination issues in their own supply chains, or by testing with detection limits too high to identify the presence of minute levels of PFAS). Manufacturers will need to scrutinize their supply chains and manufacturing processes to address such risks, including conducting TOF testing, to ensure their products remain “PFAS” free, as defined by these statutes relying on TOF analysis. If such sources do trigger regulatory coverage, exemptions for unintentionally added PFAS may apply.
Because of the technical challenges with obtaining accurate TOF data, as well as the fact that product material sources and production processes change over time, manufacturers should not treat TOF testing as a “one and done” process. Routine and regular testing of a product should be conducted, with multiple product samples tested during each round. This will ensure consistency of testing results, prevent a single false positive (or false negative) result from skewing the regulatory compliance analysis, and allow the manufacturer to identify and understand the cause of any change in the product’s regulatory compliance over time.
Finally, manufacturers should be aware that if they disclose to a regulator – either voluntarily or pursuant to a mandatory statutory requirement – testing results that show the presence of TOF in a consumer product, there could be non-regulatory liability implications. Even if there is no evidence that this TOF content equates to the presence of actual PFAS, consumer advocacy and environmental groups may claim that the product contains PFAS. This creates liability risk both with respect to personal injury claims by persons believing they have a PFAS-related medical condition, as well as failure-to-warn claims under state laws requiring consumer disclosures, such as California’s Proposition 65.
Risk Management for Investigation/Remediation of Legacy Manufacturing Sites
Manufacturers and other parties with liability for sites with known or suspected environmental contamination will also need to address the impact of TOF testing. Many contaminated legacy industrial properties that have been undergoing long-term investigation and/or remediation for non-PFAS contamination (such as chlorinated solvents used in manufacturing processes) are now being required by regulators overseeing those sites to determine whether PFAS is also present as a contaminant of concern (COC). Moreover, facilities that are known or suspected to have used PFAS in their manufacturing processes or in fire suppression systems containing aqueous film-forming foam (AFFF) are increasingly coming under scrutiny by environmental agencies, regardless of the existence of any other known environmental issues. For example, between 2019 and 2021, the California State Water Resources Control Board issued investigation orders to over 800 facilities across the state—including chrome platers, airports, landfills, publicly owned treatment works, and refineries and bulk fuel terminals—requiring environmental testing for PFAS.
As discussed above, use of the new EPA Method 1621 or another TOF method when testing for PFAS in the environment poses a real risk that non-PFAS fluorinated organic compounds will be detected and then potentially flagged by the environmental regulator as PFAS. To the extent that regulators use such testing as a screening tool at legacy non-PFAS contaminated facilities or for generic testing at former/current manufacturing plants identified as a PFAS risk, this creates significant potential cost and liability issues for responsible parties at those sites. Parties asked to conduct such testing will need to work with their consultants to ensure that the regulator requesting the testing understands the limitations of EPA Method 1621 and alternative TOF testing methods. In addition, they should work to ensure that the agency will consider multiple lines of evidence—such as testing with different methodologies, and site history and chemical usage data—before concluding that PFAS is present and imposing costly investigation and remediation requirements.
Regardless of how environmental agencies use TOF testing data—including whether they agree that no further regulatory action related to PFAS is required at a given site—those data will still pose liability risk for responsible parties, including current and former owners and operators of the facility. To the extent that local public water systems or private drinking water wells are impacted by PFAS, positive test results under EPA Method 1621 or another TOF method may be viewed by owners of those drinking water sources as evidence of causation, even though (as discussed above) the presence of TOF is not a reliable indicator of the presence of PFAS in many instances. Despite that major evidentiary flaw, such TOF testing could result in responsible parties being named as defendants in cost recovery litigation, either under the federal Superfund/CERCLA statute (given that certain PFAS have been designated as hazardous substances under that law) or pursuant to state common law tort theories (such as negligence, strict liability, and nuisance). Similarly, unreliable and potentially misleading TOF test results could result in personal injury or wrongful death claims by persons who were exposed to the PFAS-contaminated water supply.
Conclusion
TOF testing poses significant challenges for manufacturers and other parties with regulatory responsibility and potential liability associated with PFAS in products and the environment. Given the rapidly evolving nature of both the science and the regulatory environment, a comprehensive risk assessment and mitigation strategy to address these challenges will necessarily involve close collaboration between internal EHS teams, legal counsel, and expert technical consultants.
References
ASTM. 2024. ASTM WK90492 New Test Method for Total Organic Fluorine (TOF) in Solid Matrices by Solvent Extraction followed by Combustion Ion Chromatography (CIC). Accessed November 13, 2024 at: https://www.astm.org/products-services/standards-and-publications/standards/workitem-wk90492
Donley N., Cox C., Bennett K., Temkin A.M., Andrews D.Q., and Naidenko O.V. 2024. Forever Pesticides: A Growing Source of PFAS Contamination in the Environment. Environmental Health Perspectives 132 (7).
EPA. 1996. Method 9214. Potentiometric Determination of Fluoride in Aqueous Samples with Ion-Selective Electrode. December.
EPA. 2021. Total PFAS Analysis for Public Health Protection: Science, Applications, Benefits and Challenges. Workshop Proceedings. October 27–28. Accessible at: https://cfpub.epa.gov/si/si_public_file_download.cfm?p_download_id=545197&Lab=OSAPE
EPA. 2024. Method 1621. Determination of Adsorbable Organic Fluorine (AOF) in Aqueous Matrices by Combustion Ion Chromatography (CIC). January.
Forster A.L.B., Zhang Y., Westerman D.C., Richardson S.D. 2023. Improved total organic fluorine methods for more comprehensive measurement of PFAS in industrial wastewater, river water, and air. Water Research 235 (119859).
Gribble. 2002. Naturally Occurring Organofluorines. In A. H. Neilson (Ed), Organofluorines (pp. 352). Springer Berlin, Heidelberg.
Hammel E., Webster T.F., Gurney R., Heiger-Bernays W. 2022. Implications of PFAS definitions using fluorinated pharmaceuticals. iScience 25 (4).
Han Y., Pulikkal V.F., Sun M. 2021. Comprehensive Validation of the Adsorbable Organic Fluorine Analysis and Performance Comparison of Current Methods for Total Per- and Polyfluoroalkyl Substances in Water Samples. ACS ES&T Water, 1 (6), 1474-1482.
Hawaii. 2024. Memorandum from Roger Brewer, Hazard Evaluation and Emergency Response, to Interested Parties re: Interim Soil and Water Environmental Action Levels (EALs) for Perfluoroalkyl and Polyfluoroalkyl Substances (PFASs). April 12.
Höhener P. 2003. Worldwide diffuse pollution of groundwater with anthropogenic chemicals: Lessons learned from freons. Proceedings of the 2nd International Workshop on Groundwater Risk, Assessment at Contaminated Sites and Integrated Soil and Water Protection.
Petkowski, J.J., Seager, S. & Bains, W. 2024. Reasons why life on Earth rarely makes fluorine-containing compounds and their implications for the search for life beyond Earth. Sci Rep 14, 15575.
Ruyle B.J., Pickard H.M., Schultes L., Fredriksson F., Heffernan A.L., Knappe D.R.U., Lord H.L., Meng P., Mills M.A., Ndungu K., Roesch P., Rundberget J.T., Tettenhorst D.R., Van Buren J., Vogel C., Westerman D.C., Yeung L.W.Y., and Sunderland E.M. 2023. Interlaboratory Comparison of Extractable Organofluorine Measurements in Groundwater and Eel (Anguilla rostrata): Recommendations for Methods Standardization. Environmental Science & Technology 57 (48), 20159-20168.
Solomon, K. R., Velders, G. J. M., Wilson, S. R., Madronich, S., Longstreth, J., Aucamp, P. J., & Bornman, J. F. 2016. Sources, fates, toxicity, and risks of trifluoroacetic acid and its salts: Relevance to substances regulated under the Montreal and Kyoto Protocols. Journal of Toxicology and Environmental Health, Part B, 19(7), 289–304.
Spaan K.M., Seilitz F., Plassmann M.M., de Wit C.A, and Benskin J.P. 2023. Pharmaceuticals Account for a Significant Proportion of the Extractable Organic Fluorine in Municipal Wastewater Treatment Plant Sludge. Environmental Science & Technology Letters 10 (4), 328-336.
SWRCB. 2024. Memorandum from Erica Valve, Wendy Linck, and Daniel Newton, to Darrin Polhemus re: Broad-Spectrum Per- and Polyfluoroalkyl Substances (PFAS) Method Comparison Study Results and Selection of Broad-Spectrum Methods to Support Statewide Monitoring for the Class of PFAS. April 24.