The name for a family of commercial chemicals usually provides a fairly clear idea of the related compounds within it. An example is organophosphate pesticides. Another is brominated flame retardants.
But the moniker per- and polyfluoroalkyl substances (PFAS) is less distinct. What constitutes this clan of thousands of environmentally persistent and widely used synthetic compounds, including some toxic molecules, depends on a detailed definition.
There’s more than a single definition of PFAS. The one that policy makers use will ultimately determine which fluorinated chemicals are scrutinized for possible regulation.
In the US, environmental and health activists are worried about a PFAS definition that the Environmental Protection Agency rolled out a year ago. It includes far fewer molecules than one crafted by an international panel that included EPA scientists. Meanwhile, the US chemical industry’s largest trade association says both the international panel’s definition and the EPA’s are too broad. Ultimately, Congress may decide which definition federal agencies follow.
The term PFAS was popularized in a 2011 paper by a team based in Europe and North America led by Robert C. Buck, a chemist who then worked at E. I. du Pont de Nemours and Company. This groundbreaking work described these substances as highly fluorinated aliphatic compounds with one or more carbon atoms on which all hydrogen substituents are replaced by fluorine atoms and “contain the perfluoroalkyl moiety CnF2n+1–” (Integr. Environ. Assess. Manage. 2011, DOI: 10.1002/ieam.258).
In their formula-based definition, Buck and his colleagues contrasted PFAS with molecules that have undergone polyfluorination and contain scattered, multiple fluoride atoms—compounds such as CH2FCHFCHFCH2OH. “We consider that only those polyfluorinated substances having at least one perfluoroalkyl moiety CnF2n+1– belong to the PFAS family,” the team wrote.
Definitions of PFAS have since evolved.
Notably, an international panel of scientists—led by Zhanyun Wang of the Swiss Federal Institute of Technology (ETH), Zurich, and including Buck—recently expanded on the 2011 paper’s definition. Convened by the Organisation for Economic Co-operation and Development (OECD), which is a coalition of 38 of the world’s highest-income countries, and the United Nations Environment Programme, the panel worked for 3 years before aligning on a PFAS definition based on molecular structure.
In a 2021 report, the panel states that, with a few exceptions, “any chemical with at least a perfluorinated methyl group (–CF3) or a perfluorinated methylene group (–CF2–) is a PFAS.” This broadens the PFAS universe from the Buck team’s definition to include chemicals that lack fluorines at either end of a carbon chain and instead have a hydrogen or a functional group at both ends. It also includes aromatic compounds that have at least one aliphatic side chain containing at least one fully fluorinated, saturated-carbon moiety.
“The report openly and clearly acknowledges that all the PFAS substances are not the same,” Buck, who recently retired from Chemours, tells C&EN. “You’re talking about things that are gases, liquids, solids, water soluble, not water soluble.” The report states that its definition of PFAS is broad and doesn’t indicate whether a chemical is harmful, he points out.
Although the US is a member of the OECD, the EPA didn’t incorporate the panel’s definition into the agency’s 2021 strategy for selecting PFAS for safety testing under the Toxic Substances Control Act (TSCA), the federal law that regulates commercial chemicals. Instead, the EPA followed a novel and narrower definition.
The agency’s strategy states that PFAS are “chemicals with at least two adjacent carbon atoms, where one carbon is fully fluorinated and the other is at least partially fluorinated.” In addition, the EPA made this definition public in a 2021 action under the Safe Drinking Water Act, when the agency listed unregulated contaminants in drinking water, including 29 PFAS, that may require legal limits in the future.
Michal Freedhoff, the head of the EPA’s chemical safety office, told a Senate hearing on June 23 that the agency developed this definition in 2006. It has been used since by the part of the EPA that reviews new chemicals under TSCA before they are marketed. She pointed out that the agency released its rule on PFAS testing weeks before the OECD definition was unveiled.
The EPA has not publicly explained the rationale underlying its definition.
“The definition is the fundamental building block for regulation and EPA has an obligation to explain the basis for its PFAS definition,” says Kyla Bennett, science policy director for the advocacy group Public Employees for Environmental Responsibility (PEER). “EPA’s current approach will leave many toxic and persistent PFAS beyond federal control—a result that makes no sense at all.” PEER is suing the EPA to obtain documents explaining how the agency came to its definition.
A team of EPA scientists published a paper in April that suggests why the agency opted for a narrower definition than the OECD. Led by Antony J. Williams of the agency’s Center for Computational Toxicology and Exposure, the team notes that the international definition “includes almost all structures that could possibly be considered a PFAS.” The EPA attempted to trim the list of compounds considered PFAS to focus on those of greatest concern to regulators. The paper acknowledges, however, that this may create loopholes that omit compounds that regulators are interested in (Front. Environ. Sci. 2022, DOI: 10.3389/fenvs.2022.850019).
“EPA has developed specific definitions for some efforts, such as the National PFAS Testing Strategy,” an agency spokesperson tells C&EN in an email. The EPA “is considering options for defining the scope of PFAS the agency can best address under other programs, such as Toxic Substances Control Act reporting, the Safe Drinking Water Act Contaminant Candidate List, and the Comprehensive Environmental Response, Compensation, and Liability Act,” the federal statute commonly known as the Superfund law for cleaning up contaminated land.
The American Chemistry Council (ACC), a chemical industry trade group that includes PFAS makers, says it appreciates the efforts of the EPA and OECD to define PFAS but prefers the Buck team’s 2011 definition.
The OECD and EPA definitions “are overly broad in scope and consequently include compounds not relevant to chemical management, such as important pharmaceuticals that would be restricted if policymakers were to use these definitions,” the ACC tells C&EN in an email. “Policymakers need to recognize that all PFAS are not the same and they should not all be regulated the same way.”
Meanwhile, the definition of PFAS continues to evolve. In May, a chemical informatics group published a definition that is neither as broad as the OECD’s nor as narrow as the EPA’s. The researchers, based in Europe and the US, were led by Shadrack J. Barnabas of OntoChem, which conducts data mining for clients in material and life sciences. They identified PFAS on the basis of chemical structures in Core, an aggregator of open-access research papers from repositories and journals, and in the Google Patents database.
The definition this group came up with: “Each compound that contains a (AH)(AH)(F)C–C(AH)F2 group is considered a PFAS, where the AH groups could be hydrogen or any other atom and the bond between both aliphatic carbon atoms is a single bond” (Digital Discovery 2022, DOI: 10.1039/D2DD00019A). The paper calls this “a straightforward structural definition as a compromise” between the OECD and EPA definitions.
The scientific community had discussed this third definition a few years ago, but it appears to have gone undocumented until now, corresponding author Emma L. Schymanski, head of environmental cheminformatics at the Luxembourg Centre for Systems Biomedicine, tells C&EN.
Scientists’ work aside, Congress might have the last word on defining PFAS for the EPA. Lawmakers have already done so for another slice of the federal government.
A part of the 2020 military spending act (Public Law 116–92) defines a polyfluoroalkyl substance as “a man-made chemical containing a mix of fully fluorinated carbon atoms, partially fluorinated carbon atoms, and nonfluorinated carbon atoms.” It also says a perfluoroalkyl substance is one in which all the carbon atoms are fully fluorinated.
This definition applies to sections of the statute affecting the Department of Defense’s disposal of firefighting foams containing PFAS and barring its procurement of ready-to-eat meals packaged in materials that include them. But it does not specifically apply to provisions that direct the EPA to gather data on the production and use of PFAS under TSCA, says Daniel Rosenberg, director of federal toxics policy at the Natural Resources Defense Council.
This leaves the military adhering to a congressionally mandated PFAS definition and the EPA relying on its own to direct the testing of new chemicals and select drinking-water contaminants. But this situation could change, Rosenberg tells C&EN.
A bipartisan bill (H.R. 5987) pending in the House of Representatives would amend TSCA by defining PFAS consistently with the 2020 military spending law. Sponsored by Reps. Deborah Ross (D-NC) and Nancy Mace (R-SC), it would define PFAS as chemicals containing at least one fully fluorinated carbon atom.
The Ross-Mace bill likely won’t get legislative action on its own. But Congress could add it as an amendment to the National Defense Spending Act for Fiscal 2023 (H.R. 7900).
That bill “is the most sensible vehicle” for the change, Rosenberg says. He points out that the part of the 2020 military spending law requiring the EPA’s collection of production and use data originated as a freestanding bill cosponsored by Sens. Shelley Moore Capito (R-WV) and Thomas Carper (D-DE).
It’s possible that federal legislators will allow the EPA’s definition to stand. Yet a number of lawmakers are attracted to a consistent PFAS definition across federal programs. At the hearing where EPA chemical chief Freedhoff testified, Capito said of a PFAS definition, “I think uniformity obviously is easier.”