Quiet Breakout: The Strongest Bond in Organic Chemistry Has Six New Ways to Break

In a parking lot in Grand Rapids, Michigan, a shipping-container-sized reactor takes in roughly 160,000 gallons a day of brown landfill leachate and shoots it past 374°C and 221 bar into the region of the phase diagram where water stops being a liquid or a gas and starts behaving like a solvent that can dissolve oxygen the way it normally dissolves salt. Inside that supercritical fluid, the strongest single bond in organic chemistry comes apart. The trace per- and polyfluoroalkyl substances (PFAS, the “forever chemicals”) get torn down to fluoride, carbon dioxide, and water. The output is clean enough that it goes back to the municipal treatment plant. Heat from the reactor is recovered and fed back into the loop.

The reactor is called the PFAS Annihilator, and the company running it, Revive Environmental, was carved out of the nonprofit research organization Battelle Memorial Institute in 2023, capitalized by Viking Global Investors. That site is the first permitted commercial PFAS destruction job in the United States. It is also a quiet milestone in a much larger reordering of an industry: for the first time, the patent office is filling up with reactor designs that don’t capture forever chemicals — they mineralize them.

Until very recently, the entire PFAS economy was built around the assumption you couldn’t do that. The carbon-fluorine bond is the strongest single covalent bond in organic chemistry, with a dissociation energy near 485 kilojoules per mole. PFAS owe their thermal stability, their water and oil repellency, and their resistance to enzymes, microbes, and household incinerators to that one chemical fact. Activated-carbon filters, ion-exchange resins, foam fractionators, nanofiltration membranes: every commercial method shipped to a wastewater plant or fab line for the last two decades does the same thing: it concentrates the molecules onto a sorbent, which is then trucked somewhere else. Usually a hazardous-waste incinerator. Often a landfill. The problem is always relocated, never solved.

That assumption broke on April 10, 2024, when the EPA finalized its first enforceable drinking-water limits for PFOA and PFOS, at four parts per trillion each, and added separate limits for three more PFAS at ten parts per trillion. Public water systems are on the clock for monitoring by 2027 and for compliance by 2029 (with a proposed extension to 2031 for the first two compounds). Concurrently, the agency listed PFOA and PFOS as CERCLA hazardous substances, which means whoever sent the molecules to a landfill or wastewater plant can be pulled back in years later to pay for cleanup. Suddenly the architecture of “concentrate and ship” carried a liability that traveled with the molecule forever.

The patent office has been watching this in real time. Since January 2023, US grants mentioning PFAS in their title or abstract have moved from 27 a year to 47 in 2025, with 17 more issued in the first five months of 2026. What’s interesting is not the headline number, which is a vocabulary-frequency observation, but the subset that does something new. At least seventeen grants since 2023 describe reactors that actually break the C-F bond rather than concentrating it, spread across six distinct destruction chemistries:

Supercritical water oxidation. Five grants, all to Revive Environmental (US 11,780,753; US 11,970,409; US 12,091,331; US 11,891,323; US 12,371,358). The last two pair SCWO with reverse-osmosis pre-concentration so the salt does not foul the reactor. The newest, granted September 2024, adds a foam-fractionation step in front, which lets the system process AFFF firefighting foam without choking on its surfactants.

222-nanometer UV photolysis. Two grants to Claros Technologies of Minneapolis (US 12,304,850 and US 12,351,498). The chemistry uses persulfate plus a sulfite salt at pH 10 to generate hydrated electrons, which attack the perfluorinated tail. In July 2025 Claros ran a field pilot at a Daikin America fluoropolymer plant in Decatur, Alabama, that the company says destroyed more than 99.99% of the targeted PFAS in 50,000 gallons of wastewater.

Thermomechanochemical milling. Two grants assigned to the United States government on behalf of the EPA (US 12,006,238 and US 12,384,710). The reactor is a ball mill with a heater. PFAS-laden sludge or activated carbon is co-milled with a reactive solid (quartz, calcium oxide, or, in a follow-on paper in Nature last year by EPA chemist Endalkachew Sahle-Demessie’s group, sodium phosphate), and the mechanical impact plus heat liberates fluoride that can be reused. It is genuinely strange to find the federal government as the named assignee on a destruction patent. It is also a tell that the agency writing the regulation has been writing the chemistry too.

Sonochemical pyrolysis. A reactor full of transducers that drives ultrasonic cavitation bubbles to collapse temperatures locally above 4,000 K, hot enough to pyrolyze the C-F bond inside the bubble (US 12,576,297, granted March 2026; US 12,522,521).

Pulsed plasma in liquid. Onvector LLC’s plasma-discharge reactor (US 12,295,090, granted May 2025) generates non-thermal plasma directly inside a microbubble suspended in water; the radicals do the defluorination. The same plasma platform is referenced in Waste Management’s leachate patents through AECOM’s DE-FLUORO brand.

Hydrated electrons from “structurally altered” water. A more exotic claim from H2Plus (US 11,993,683): generating reductive electron pairs by destabilizing water structure with a small molecule, then infusing the resulting gas into PFAS-bearing waste. Battelle has a parallel patent (US 11,774,414) using sodium metal dispersions in alcohol to do reductive defluorination on the analytical bench.

What lifts this past being a list of cool reactors is who is filing alongside the chemistry companies. Waste Management, the largest US trash hauler, has four PFAS grants since 2023 (US 11,623,884; US 11,795,090; US 12,209,043; US 12,410,071). One of them is a pyrolysis process for sewage sludge that turns the solids into biochar while removing the PFAS. Another wires AECOM’s plasma reactor directly into landfill-leachate treatment. Waste Management is, in essence, patenting the chemistry to defuse its own long tail of CERCLA liability before someone hands it the bill. Battelle is doing the same for its government and defense customers. The Department of Defense, which owns the largest catalog of PFAS-contaminated firefighting-foam sites in the country, is co-funding the next phase of Revive’s reactor.

Two consequences worth thinking about.

The first is that PFAS economics flip. For twenty years the rational play for an industrial site holding a PFAS plume was to capture, ship, and let the lawyers price the long tail. Destruction at point of capture changes the math: a granular-activated-carbon filter that gets regenerated on-site rather than landfilled becomes a recurring service contract instead of a stranded liability. Revive’s GAC Renew™ system is exactly that play; so, more quietly, is Battelle’s silica-assisted destruction patent (US 12,012,343), which converts the same etching reaction that ruined laboratory glassware for decades into a way to mineralize PFAS at relatively mild temperatures.

The second is that “forever” stops being a marketing word and becomes a regulatory one. The EPA’s drinking-water rule sets a number, 4 ppt, that nobody can hit by dilution and nobody can credibly hit by relocation. The cheapest way to comply on a long enough timeline is to break the bond. Six different reactor designs is the patent office saying out loud that it expects multiple chemistries to coexist, sized for different waste streams: SCWO at landfills, 222-nm UV at fluoropolymer plants, ball mills at military bases, sonochemistry inside concentrate-handling skids. Whichever wins, the implicit promise of the C-F bond, that whatever you make with it will outlast everything else you make, no longer holds.

Method note. Counts of US utility patent grants come from a local index of 9.3 million US grants sourced from the USPTO bulk grant feed, filtered to grants issued between January 2018 and early June 2026 whose title or abstract mention PFAS, perfluoroalkyl, or polyfluoroalkyl substances. The six-chemistry decomposition was done by reading the abstracts and representative claims of grants since January 2023 and grouping them by the disclosed destruction mechanism, not by keyword. Literature trend numbers reflect counts from OpenAlex on the same query terms. Specific deployment facts (Revive’s Grand Rapids site, Claros’s Daikin pilot, EPA’s TMC research line, the April 10, 2024 MCL) are sourced from Battelle, Revive Environmental, Claros Technologies, the EPA, and Federal Register filings. Patent numbers are written as granted; assignees reflect the primary assignment at issue and may have changed hands since.