Researchers at Arizona State University have tested a water-treatment membrane that can do more than capture “forever chemicals”: it can break some of their exceptionally strong carbon–fluorine bonds. Their findings are promising, but the small laboratory reactor is not yet a proven solution for public drinking-water systems.
In research published October 19, 2026, in the peer-reviewed Proceedings of the National Academy of Sciences, a team led by environmental engineer Bruce Rittmann reported a continuous 58-day experiment involving six PFAS compounds. The system removed an average of about 96% of the chemicals from the flowing water. Separate measurements of fluoride release indicated approximately 83% overall defluorination — evidence of chemical breakdown rather than separation alone.
The distinction is important. Getting PFAS out of the water is not the same as destroying them. Traditional treatment methods such as activated carbon, ion-exchange resins and reverse osmosis may leave PFAS concentrated in spent filters or waste streams that still require management. The peer-reviewed study and Arizona State University’s report describe a possible way to address that waste.
How the membrane attacks durable chemicals
PFAS, or per- and polyfluoroalkyl substances, appear in certain industrial processes, firefighting foams and products designed to resist water, grease or heat. The carbon–fluorine bonds responsible for this performance are unusually difficult to break, allowing some compounds to persist in the environment and living organisms.
The prototype contains hollow polymer fibers coated with palladium nanoparticles. Hydrogen gas travels inside the fibers to the catalytic surface, where the metal activates it. The reactive hydrogen can replace fluorine atoms in PFAS molecules in a process called hydrodefluorination.
The researchers identified complementary roles for hydrogen: hydrogen on the surface helps position PFAS near the palladium, while hydrogen stored below the catalyst surface helps drive the bond-breaking chemistry. Controlled comparisons of hydrogen delivery supported this mechanism. The process operates near room temperature, although it still uses hydrogen at modest pressure and carefully controlled conditions.
What the 58-day test found
The continuous experiment treated a mixture of PFOA, PFOS, PFHxS, PFNA, PFBS and a GenX-related compound. Researchers recorded individual removal efficiencies of roughly 93% to 99%, averaging approximately 96% across the six. They separately measured an overall defluorination efficiency of 83% by analyzing released fluoride.
That 83% result strengthens the evidence that the membrane actually breaks carbon–fluorine bonds. It does not establish that 96% of all molecules were fully destroyed, that every intermediate product was harmless or that the treated water met drinking-water limits. The researchers observed differences among the chemicals; GenX was comparatively difficult to break down.
Performance remained consistently high throughout the 58-day run, but the paper also reports that the palladium coating lost about 10% of its mass. That material loss is important when assessing maintenance, long-term stability and cost.
What still separates the prototype from a water plant
In the study’s controlled continuous-flow test, the reactor held approximately 40 milliliters of water. Flow was only 0.05 milliliters per minute, with a roughly 12-hour hydraulic residence time. Each of the six PFAS entered at 50 micrograms per liter, and the operating conditions included acidic water at pH 4.
Such conditions are valuable for investigating chemistry but do not demonstrate performance in a municipal drinking-water system, where volumes, pH, natural constituents and contaminant concentrations can differ substantially. The team checked selected potential interferents, including nitrate and glucose, but full-scale validation would need to address water chemistry, transformation products, hydrogen management, catalyst recovery, replacement costs and energy demand.
Palladium itself is expensive. Rittmann told the university that reducing the amount of metal required is a key research priority. A plausible future role is to treat PFAS-rich waste produced by conventional filters rather than to send every gallon of municipal water through the new reactor. That application still needs field testing and economic evidence.
Why the research matters for Americans
The CDC’s Agency for Toxic Substances and Disease Registry reports that nearly all people in the United States have measurable PFAS in their blood, although levels of several older compounds have declined over time. The widely repeated figure that “97% of Americans” carry PFAS comes from older monitoring summaries and should not be mistaken for a newly measured 2026 national rate. CDC/ATSDR population monitoring.
Risk depends on which PFAS are involved and on the level and duration of exposure. Federal health authorities describe associations between exposure to certain PFAS and elevated cholesterol, reduced antibody responses, pregnancy complications and some cancers. These findings do not mean that every exposure causes illness. CDC/ATSDR health-effects review.
In April 2024, the U.S. Environmental Protection Agency finalized the first national drinking-water limits for six PFAS, including enforceable limits of four parts per trillion for PFOA and PFOS. In May 2026, EPA proposed retaining those two limits while rescinding several others and offering possible compliance extensions. These were proposals, not regulations that had already taken effect. EPA’s drinking-water standards and proposal summary.
Destroying PFAS instead of moving it elsewhere
EPA’s updated 2026 guidance reviews options for the destruction and disposal of PFAS-containing waste, including underground injection, hazardous waste landfills and properly managed thermal treatment. Not all disposal options break PFAS apart, and the agency acknowledges uncertainties in available and emerging technologies. The guidance itself does not require every facility to destroy PFAS. EPA’s 2026 guidance.
Arizona State University’s membrane could eventually help close the gap between removing contamination and eliminating persistent waste. Before public-health benefits can be claimed, researchers must test real-world waste streams, characterize the resulting chemicals and show that the process works at practical scale and acceptable cost.
Video: An independent explanation of the study
This independent video commentary describes the same research, its experimental results and scale-up questions. It is not a presentation from the study authors. The peer-reviewed paper remains the primary scientific source.
Sources
- Long and colleagues, Proceedings of the National Academy of Sciences, October 19, 2026.
- Arizona State University research announcement, October 19, 2026.
- EPA drinking-water PFAS standards and 2026 proposals.
- EPA 2026 destruction and disposal guidance.
- CDC/ATSDR population exposure and health effects.
