Water and oxygen unmake plastic, Canada keeps medicine unpatentable, and hydrogen crumbles old magnets

The stakes are concrete. Most chemical recycling leans on metal catalysts that are expensive, sometimes toxic, and easily poisoned by the dyes, stabilizers, and mixed materials found in real waste. Stripping the catalyst out removes a barrier that has kept such methods largely confined to the laboratory.[3]

How it works01

When the plastic is melted and stirred in water under oxygen pressure, it disperses into microdroplets, tiny water-and-oil droplets whose surfaces carry intense electric fields.[1] Those interfaces spontaneously generate hydroxyl radicals, highly reactive fragments of a water molecule that act as chemical scissors, pulling hydrogen atoms off the polymer chains and cutting them apart.[1][3] The end products are diacids, small molecules with an acid group at each end that industry uses to make nylons, plasticizers, food additives, and medicines.[3]

Running polyethylene, the plastic in bags and bottles, at 125 degrees Celsius under oxygen for 18 hours converted it completely, with 69 percent of the carbon ending up as diacids and succinic acid making up 58.4 percent of that mixture.[1] An isotope test that swapped ordinary water for heavy water slowed the reaction sevenfold, pinning the water-derived radical as the rate-limiting step.[1]

Robustness and scale02

The reaction tolerated the additives that trip up conventional catalysts, including light stabilizers and antioxidants dosed at up to 4.8 percent by weight, with no measurable loss of conversion.[1] It also ran on tap water and even seawater, shrugging off the halide salts that would deactivate many metal systems.[1] Commercial gloves, caps, and mixed packaging all cleared 60 percent diacid yields.[1]

The authors pushed the process from a 0.2 gram bench test to 300 grams of polyethylene in a five-liter reactor, holding 89 percent conversion with more than half the carbon recovered as saturated diacids.[1] A techno-economic model of a plant handling 60,000 tonnes a year estimated 144 million dollars in capital cost against 72.1 million in annual after-tax profit, a payback of roughly 3.3 years, and a break-even capacity near 9,000 tonnes a year.[1] A life-cycle estimate put the route at minus 0.30 kilograms of carbon-dioxide-equivalent per kilogram of plastic, a small net sink, against 2.11 for incineration and 0.15 for landfill.[1]

What it does not yet settle03

The demonstration remains a laboratory and intermediate-scale result, and the reactions run for 18 to 48 hours, far slower than industrial throughput would need.[1] The products skew toward short-chain acids in the four-to-eight-carbon range, and the economic and emissions figures are model projections rather than plant data.[1] "We are awash with plastic waste, and we need viable solutions for its effective recycle," said Graham Hutchings of Cardiff, framing the appeal as much economic as chemical.[3] "By eliminating the need for expensive or toxic catalysts entirely, we have removed one of the major economic and environmental barriers," Wang said.[3] Whether the chemistry survives the jump to industrial tonnages will decide if that promise holds outside the reactor.

On 17 July 2026 the Supreme Court of Canada ruled that Janssen Inc. may keep its patent on a dosing schedule for a long-acting schizophrenia injection, while reaffirming that "methods of medical treatment", meaning the clinical judgment a physician exercises in caring for a patient, cannot be patented.[1] The court dismissed an appeal by the generic drugmaker Pharmascience Inc., leaving Canadian law where it stood.[2]

At issue was Canadian Patent No. 2,655,335, which claims a dosing regimen for paliperidone palmitate, an injectable antipsychotic that Janssen sells as Invega Sustenna.[1] Pharmascience, which makes generic copies that reach the market once a brand's patents expire or are struck down, argued the claims were really an unpatentable method of medical treatment; had that argument won, cheaper competition could have arrived sooner.[4]

Canadian courts have long refused patents on methods of medical treatment, a bar the judges built themselves rather than one Parliament wrote in plain words.[6] The concern is that a physician should not face a claim of patent infringement for deciding how to treat a patient.[2]

The line the court drew01

Writing for seven members of the court, Justice Mahmud Jamal rejected the neat distinction lower courts had flirted with, between a fixed dose and a variable one.[1] The majority framed the inquiry instead around an "ultimate question": whether a claim fences in an area of medical practice by monopolizing a professional's skill and judgment, a question it assessed through three non-exhaustive factors.[5]

Applied to the '335 patent, the claims survived. Once a physician decides to use the regimen, the majority reasoned, carrying it out demands no further skill or judgment because the patent itself sets out the doses and their timing.[1] The dosing schedule was therefore a patentable invention rather than the practice of medicine.[4]

A rule two judges would scrap02

Two justices, Michelle O'Bonsawin and Mary Moreau, agreed the appeal should fail but wrote separately to say the prohibition on methods of medical treatment should be abolished altogether.[6] They noted that the statutory provision once thought to anchor the rule, former subsection 41(1) of the Patent Act, has since been repealed, and would instead police such claims through the ordinary requirement that an invention be useful.[1]

The decision closes a case that had run through two lower courts. The Federal Court dismissed Pharmascience's challenge, and the Federal Court of Appeal affirmed in 2024, asking whether using the invention requires the exercise of skill and judgment.[3] The Supreme Court's dismissal leaves that reasoning intact.[5]

For drug companies, the ruling confirms that dosing-regimen patents remain a workable way to protect a medicine's later refinements. For generic manufacturers and the public payers who buy most prescription drugs, the window for cheaper copies of Invega Sustenna stays where the lower courts had put it.[7] Because the court's test is narrow and fact-specific, future disputes will turn on how a claim is written rather than on any bright line.[4]

The latest sign came on 27 July, when HyProMag USA laid out a phased plan for a recycling and manufacturing hub in the Dallas-Fort Worth area, putting magnet finishing ahead of full production.[1] Under the schedule, finished neodymium-iron-boron magnets would begin coming off the line in the first half of 2027, with an integrated hydrogen-processing section following in the second quarter of 2028.[1] At full run the plant is designed to take in roughly 400 tonnes of recycled sintered magnets a year, ramping toward about 1,526 tonnes of magnetic product.[1]

The chemistry01

The reaction it relies on is old and unfussy. Scrap devices holding magnets go into a rotating porous drum inside a pressure vessel, which is then filled with hydrogen gas.[5] Over about thirty to forty minutes the hydrogen works into the neodymium-iron-boron alloy, which swells and shatters into a friable powder while losing its magnetism.[5] Because the same reaction loosens coatings and glue, the demagnetised grains fall away as the drum turns, sparing the labour of prying each magnet out by hand.[5]

The method, called hydrogen decrepitation, was worked out at the University of Birmingham's Magnetic Materials Group, where Rex Harris first demonstrated it and where the group has studied rare-earth alloys for more than forty years.[3] The recovered powder can be re-melted into master alloys, spun into bonded-magnet feedstock, or, in the lowest-impact route, pressed and sintered straight back into new magnets.[5]

One dominant node02

The appeal is less the chemistry than where it sits in a chain that runs through a single country. China refines roughly nine in ten of the world's rare earths and makes an even larger share of the finished sintered magnets, a position built up over two decades.[2]

Beijing has begun to use that position. Starting in April 2025 it placed seven heavy rare earths and finished magnets under export licensing, then widened the rules through the year to cover parts and assemblies containing Chinese-sourced material and, by December, foreign-made goods that used Chinese rare earths or processing technology.[2] That is the established backdrop against which the recycling plants are being financed.

The short loop03

Recycling reaches into the chain at a different point. Because old magnets already hold the rare earths in roughly the right proportions, the direct re-sintering route can crush and reform them without returning to a mine or a separation plant.[5] A lifecycle assessment put the energy needed for that short loop at about 88 percent below primary production from ore.[5]

The approach is no longer confined to a bench. A facility at Tyseley Energy Park in Birmingham opened in January and can recover more than 400 kilograms of alloy per batch, with a design capacity of 100 tonnes a year on a single shift.[3] In April a HyProMag plant in Pforzheim, Germany, began commercial runs, aiming to scale from about 100 tonnes a year toward a permitted ceiling of 750.[4] The Texas project would add a third node, on a continent where magnets for electric drives, wind turbines and defence systems are treated as a strategic gap.[6]

What the recyclers cannot yet claim is scale. The plants now running or planned are each measured in hundreds of tonnes a year,[1][4] a fraction of an output that China still supplies almost in full.[2] Recovery also leans on a steady flow of scrap, from hard drives, motors and speakers, that has to be collected and sorted before any hydrogen is added.[5] The mechanism is proven and three plants are now real, yet on their stated designs each would have to multiply its throughput many times over, and lock in the scrap to feed it, before the totals begin to register against a Chinese supply counted in the tens of thousands of tonnes.