In a Grand Rapids wastewater plant, a pressurized tube of water is being held above its critical point (374°C and 218 atmospheres, the regime where liquid and gas stop being different things) and pumped with oxygen until the molecules dissolved in it stop existing. The molecules in question were, until very recently, considered indestructible. Teflon’s chemical ancestors. Scotchgard’s working ingredient. The contaminant the EPA has just told 11,000 public water systems they have until 2031 to get below four parts per trillion in tap water.

The carbon-fluorine bond that defines per- and polyfluoroalkyl substances is the strongest single bond in organic chemistry. It is the reason “forever chemicals” got their nickname. It is also, increasingly, the reason a small industry’s worth of patent attorneys are billing extra hours.

US grants that mention PFAS by name jumped to 31 last year, up from a flat baseline of about ten a year through the back half of the 2010s. Eleven more issued in the first four months of 2026, putting the run-rate above 30 again. That alone is a vocabulary trend. The interesting subset is the one that doesn’t just talk about capturing the molecules. It talks about destroying them. Twenty-four US grants since 2022 describe how to break the C-F bond, versus essentially zero in the decade before. The forever is starting to come off.

Three different ways to crack a bond nobody could crack

Read the actual claims and the variety is striking. They are not the same invention with different keywords. They are several different physics experiments converging on the same target.

The Battelle Memorial Institute, the Columbus, Ohio nonprofit lab that runs Oak Ridge and several other federal labs, owns a 2024 grant (US12012343) on what may be the most quietly elegant approach: heat PFAS in water in the presence of plain silica at 100°C and at least 90% of it falls apart. The byproduct is hydrofluoric acid, which then etches the glass. The patent specifically claims borosilicate as the reaction vessel, meaning the destruction container slowly consumes itself as it works. A laboratory beaker is, on a long enough timescale, a reagent.

Battelle has also patented a sodium-metal version (US11774414) that throws PFAS into a sodium dispersion in ether and gets nearly complete defluorination in twenty-five minutes. The point is partly analytical for “total organofluorine,” because the regulatory targets are now so low that you have to count the fluorine atoms themselves to know if you’ve cleaned the sample.

Then there is Revive Environmental Technology, a company Battelle spun out in 2022 specifically to commercialize what it calls the PFAS Annihilator. Four grants in two years (US11780753, US11891323, US11970409, US12371358) describe variations on supercritical water oxidation: concentrate the PFAS via reverse osmosis, strip the salt, drive the brine into a reactor running above water’s critical point, inject an aqueous oxidant, and out the other end you get carbon dioxide, inert salts, and water with PFAS measured below five parts per trillion. The residence time the patent claims is twenty seconds. As of mid-2023, a permitted full-scale unit has been chewing through landfill leachate at a Heritage-Crystal Clean facility in Grand Rapids, Michigan, the partnership branded “4never,” processing 300 to 500 gallons of concentrated leachate per day, with about 160,000 gallons of raw landfill liquid behind it. Revive took home a 2025 Water Environment Federation technology award for the system; the Department of Defense has been deploying it against aqueous film-forming foam at military sites where decades of fire-training drills left a plume.

A third approach skips the supercritical reactor and skips the silica entirely. In 2022, Brittany Trang and William Dichtel at Northwestern published a paper in Science showing that perfluorocarboxylic acids — the PFOA family — can be mineralized at 80 to 120°C in a mixture of dimethyl sulfoxide and sodium hydroxide at ambient pressure, with the molecules falling apart not one carbon at a time but two or three at a time in a kind of unzipping cascade. The paper has now been cited 340 times. Dichtel was a MacArthur fellow; Trang was a graduate student. The lab cost was, by the standards of forever-chemical destruction, almost nothing.

These are not the same engineering. One is a bomb calorimeter run in reverse; one is a hot beaker; one is a sodium fire. They are coherent because they all attack the same molecular target, the C-F bond, and because they all answer the same operational question: what do you do with the rinsate from the activated-carbon filter once it’s full?

The unusual companies showing up in the assignee field

The thing to notice in the patent record is who is filing.

Waste Management’s intellectual-property holding company has four PFAS grants since 2023. One of them (US11795090) is a method for slow-pyrolyzing sewage sludge at 580°C in a continuous screw conveyor, no catalysts, no inert gas, until the resulting biochar is, in the patent’s language, “essentially free of PFAS.” Another (US12410071) covers a leachate-treatment train designed to feed a destruction unit downstream. The trash company has stopped pretending it can land-store the contamination it is paid to take.

WM’s Crossroads Landfill in Maine is now building the first biosolids dryer in the state, due online in the second quarter of 2026, capable of handling roughly 73,000 tons a year, or about 83% of the biosolids the state generates, according to Waste Dive. Maine banned land-spreading of PFAS-contaminated biosolids in 2022 after dairy farms turned up with contamination from decades-old applications. The dryer, paired with leachate treatment, is the practical answer to a question that didn’t have one three years ago.

Revive’s deployments and Battelle’s R&D are funded in large part by the Department of Defense, which had identified 723 installations requiring PFAS assessment as of March 2025 and a Congressional mandate to clean them. The 3M Company, which made the original PFAS chemistries, agreed in 2023 to a $10.3 billion settlement with public water utilities and committed to exiting all PFAS production by the end of 2025. The math on the supply side has flipped: the people who used to be paid to store the contamination are now patenting the destruction process, and the people who used to make the molecules are now paying for the cleanup.

What’s still missing

PFOS, the older sulfonate-based forever chemical that covered the inside of waterproof jackets for forty years, is harder to crack than PFOA. Northwestern’s NaOH/DMSO method explicitly does not work on it. Revive’s supercritical reactor does — the 2024 ACS ES&T Water paper from a multi-institution team validates destruction in alcohol-resistant firefighting foams — but supercritical pressure vessels are not cheap, and the regional model implied by the patents (concentrate at the source, ship the brine to a destruction hub) requires a logistics network that does not yet exist. The EPA’s four-parts-per-trillion rule survived an early-2025 challenge from the new administration but had its compliance deadline pushed from 2029 to 2031, giving water utilities two extra years to either find a destruction vendor or hope someone else does. White & Case noted in a May 2025 client alert that the “any combination” hazard-index limit on four other PFAS variants was rescinded entirely, narrowing what utilities have to remove.

For an R&D director or a corp-dev scout, the read-through is straightforward. The forever-chemical destruction stack has gone from zero patents to roughly two dozen in five years, with three commercially distinct mechanisms, two operational sites, one MacArthur fellow, and one waste hauler quietly reinventing what it sells. There is now an answer to the question regulators have been asking since 2024. The scarce input is the engineering capacity to deploy it.

Method note

Counts are taken from the full corpus of US utility patent grants from 1976 through April 28, 2026 (about 9.3 million documents sourced from USPTO bulk-grant XML), filtered for grants whose title mentions PFAS or its full chemical names alongside destruction-related verbs (destroy, degrade, defluorinate, mineralize). Assignee tallies combine variant spellings and subsidiary filings. The Northwestern paper is Science 376:1107–1111 (2022); citations come from OpenAlex (357 million scientific works as of this writing). Commercial-deployment details, settlement figures, and EPA rule history come from Crain’s Grand Rapids Business, Waste Dive, Chemical & Engineering News, the EPA’s April 2024 final rule, and the May 2025 White & Case client alert. The headline number — twenty-four destruction grants since 2022 versus essentially none in the prior decade — counts only grants whose titles describe both PFAS and a destruction or degradation method; it excludes sorption, capture, and analytics, which together account for the bulk of the broader PFAS patent record.