Tiny Metal Particles Use Light to Shatter the Forever Chemicals in Water

Per- and polyfluoroalkyl substances, the notorious family of synthetic chemicals known as PFAS, have earned the nickname forever chemicals for good reason. Their carbon-fluorine bonds are among the strongest in organic chemistry, requiring roughly 485 kilojoules per mole to break, and roughly 15,000 distinct PFAS structures are known to contaminate water supplies worldwide. Conventional treatment plants can capture these molecules on activated carbon or membranes, but destroying them typically demands extreme heat, harsh reagents or energy-hungry plasma reactors. Now a research team led by scientists at Yale University reports a dramatically milder route: they have shown that tiny palladium and platinum nanoparticles, when excited by ordinary ultraviolet light at room temperature, can strip nearly every fluorine atom from two of the most infamous PFAS compounds in water.

The study, published in Nature Water, demonstrates that ultraviolet irradiation of quantum-sized palladium and platinum catalysts anchored to amine-functionalized mesoporous silica nanoparticles triggers localized surface plasmon resonance, a collective oscillation of free electrons on the metal surface. When the incoming photons match the plasmon resonance frequency, the particles generate highly energetic hot electrons, local electromagnetic fields and, crucially, hydrated electrons in solution. These hydrated electrons turn out to be the workhorses of the reaction, delivering electrons directly into the fluorinated carbon chains of perfluorooctanoic acid, or PFOA, and perfluorooctane sulfonic acid, or PFOS, and prying off fluorine atoms one by one.

The experimental results are striking. Under low-intensity UVC light at 254 nanometres, with a light intensity of just 2.6 milliwatts per square centimetre, both catalyst systems destroyed more than 99 percent of PFOA and PFOS within 24 hours at room temperature. More importantly, ion chromatography and time-resolved fluorine-19 nuclear magnetic resonance spectroscopy confirmed near-stoichiometric fluoride release, meaning essentially all of the fluorine locked inside the parent molecules was liberated as inorganic fluoride within 72 hours. No-catalyst controls showed no measurable defluorination, and catalyst suspensions kept in the dark were equally inert, proving that the plasmonic interaction between light and metal is what drives the chemistry.

The design of the catalyst itself is central to the achievement. The team synthesized spherical mesoporous silica particles with an average diameter of 77.7 nanometres, an exceptionally high surface area of 496.3 square metres per gram, and pores roughly 3.6 nanometres wide. These pores were lined with primary amine groups, which served as anchoring sites for the subsequent solvothermal growth of palladium and platinum nanoparticles. The resulting metal clusters were astonishingly small, averaging 0.7 nanometres for palladium and 1.5 nanometres for platinum, well within the quantum size regime where plasmon resonance can be excited by ultraviolet photons. Diffuse reflectance spectroscopy confirmed plasmon absorption bands between 370 and 250 nanometres, and transmission electron microscopy together with inductively coupled plasma mass spectrometry showed the particles remained stable throughout 96-hour reactions, with no aggregation and no measurable metal leaching.

Mechanistic probing through a matrix of scavenger and control experiments revealed which reactive species matter. Adding silver nitrate, which scavenges hydrated electrons, slashed PFOA degradation from 99.1 percent to just 25 percent, while scavengers of superoxide radicals and photogenerated holes had little effect. This points decisively to hydrated electrons as the dominant reductive intermediate. The reactions also performed best under mildly acidic conditions, maintained by carbon dioxide sparging to a pH between 4.2 and 4.5, consistent with a model in which positively charged catalyst surfaces attract the anionic PFAS molecules and concentrate them near the reactive metal sites. Sorption isotherms confirmed favourable PFAS uptake under these conditions, suggesting that interfacial concentration of contaminants plays a synergistic role alongside plasmon-driven electron generation.

One of the most elegant features of the system is that it generates its own reductant supply. Conventional reductive defluorination strategies typically require exogenous hydrogen gas and metal catalysts to activate the stubborn carbon-fluorine bonds. In contrast, gas chromatography confirmed that the plasmonic nanoparticles produce hydrogen gas in situ upon ultraviolet irradiation, both with and without PFAS present. Control experiments with added external hydrogen showed nearly identical reaction kinetics, demonstrating that the system is self-sustaining and that hydrogen availability does not limit the reaction. Significantly, when external hydrogen was sparged into the reactor without ultraviolet light, no meaningful PFAS degradation occurred, further confirming that plasmon excitation, not merely the presence of hydrogen, is the essential driver.

High-resolution tandem mass spectrometry allowed the researchers to map the transformation products in unprecedented detail. The dominant pathway was stepwise hydrodefluorination: each carbon-fluorine bond is replaced by a carbon-hydrogen bond, preserving the carboxylate or sulfonate head group while the fluorinated chain is progressively stripped. The first intermediates, in which a single fluorine is swapped for hydrogen, appeared early and were themselves consumed within 12 hours, giving way to ever more hydrogenated products, ultimately reaching species with 14 and 17 hydrogen atoms in place of fluorine. Mass spectrometry fragmentation analysis suggested that fluorine loss could occur from many positions along the carbon backbone, generating a rich diversity of positional isomers, in contrast to the common assumption that the carbon adjacent to the head group is the first site of attack.

Equally notable is what the researchers did not observe. Oxidative PFAS treatment routes frequently leave behind shorter-chain perfluorinated carboxylic acids such as PFHpA and PFHxA, compounds that are more mobile in water, poorly understood toxicologically and increasingly targeted by regulators. In this plasmonic system, quantitative liquid chromatography-mass spectrometry detected no measurable accumulation of these chain-shortened by-products. By suppressing hydroxyl radicals with methanol, the team effectively shut down the oxidative decarboxylation-hydroxylation-elimination-hydrolysis pathway and channelled the reaction exclusively through the reductive route, which is also theoretically the lower-energy option and therefore potentially more cost-effective and scalable.

The researchers also compared the two ultraviolet regimes. Under longer-wavelength UVA light at 368 nanometres, palladium catalysts degraded 60.7 percent of PFOA and platinum catalysts 69.8 percent in 24 hours, with roughly 80 percent fluoride evolution over 96 hours. When reaction rates were normalized to photon flux, UVC proved roughly 10 to 75 times more efficient per photon, reflecting both the higher photon energy and more favourable metal absorption in that range. Yet the UVA result carries practical weight: longer-wavelength light is compatible with solar irradiation and near-visible LEDs, which are cheaper, longer-lived and safer to deploy than traditional UVC lamps, suggesting a credible path toward field-scale reactors as the catalysts are further optimized.

The authors emphasize that their system operates catalytically rather than stoichiometrically, with modest noble-metal loadings of 4.0 weight percent palladium and 3.7 weight percent platinum on the silica support, and an estimated raw material cost of about 3.80 US dollars per gram of catalyst, excluding operating and irradiation costs. Future work will focus on reducing or replacing the noble metals with earth-abundant alternatives, developing visible-light-responsive variants, resolving the fine details of how hydrated electrons and surface-bound hydrogen atoms cooperate within the confined pores, and integrating the catalyst into continuous-flow treatment systems. For now, the demonstration stands as a landmark: at room temperature, in ordinary water, and with nothing more than ultraviolet light, quantum-sized metal particles can dismantle the very bonds that make forever chemicals forever.

Subject of Research: Plasmon-enhanced photocatalytic reductive defluorination of PFAS in water

Article Title: Plasmon-enhanced reductive defluorination of PFAS in water under ambient conditions

Article References: Durak, O., Lee, S. S. S., Watt, L. M., Maisto, S. K., Menges, F., Kim, C., Mayer, J. M., & Fortner, J. D. (2026). Plasmon-enhanced reductive defluorination of PFAS in water under ambient conditions. Nature Water, 4(9), 1166-1176. https://doi.org/10.1038/s44221-026-00690-1

Image Credits: AI Generated

DOI: 10.1038/s44221-026-00690-1

Keywords: PFAS, PFOA, PFOS, plasmonic photocatalysis, hydrated electrons, defluorination, water treatment, mesoporous silica, palladium nanoparticles, platinum nanoparticles, UVC irradiation, environmental remediation