Three independent teams have produced metre-scale polymer and covalent-organic-framework (COF) membranes that separate hydrocarbons without boiling crude oil, with one group fabricating 50-metre rolls for refinery trials. The work targets a process—fractional distillation—that accounts for 1% of global primary energy and emits more than 160 million tonnes of CO₂ each year, according to figures cited by Chemistry World.

"We're slowly getting to the point where these membranes can be long lasting and work with actual crude oil streams rather than idealised trials," science correspondent Mason Wakley told the Chemical Breakdown podcast. "A lot of these were produced within the lab. Whether there is a way that you can have that production line on a commercial scale, I'm not too sure about."

What does fractional distillation actually cost?

Heating crude to sort hydrocarbon fractions by boiling point consumes 1% of global primary energy and releases over 160 million tonnes of CO₂ per year. Refineries also burn roughly 10% of the crude they receive to generate that heat. Membrane separation sorts molecules by size, shape, or chemical affinity instead—potentially reclaiming much of that energy input while keeping the carbon in the barrel.

How close are the three prototypes to industrial use?

• Cross-linked polymer membranes: researchers fixed pore size by linking polymer chains, preventing the swelling that destroyed selectivity when crude contacted the membrane.
• Covalent organic frameworks (COFs): another team built 3D porous networks from tunable organic linkers, adjusting pore shape and chemistry with alkyl chains.
• Fabrication scale: one group produced 50-metre membrane sheets, spiral-wound them into modules, and tested them on Arabian Light crude—a low-sulphur, high-fluidity grade—for up to one month.

Heavier crudes remain untested. "Whilst that is a major step… there is still more that needs to be done," Wakley said.

What did Seveso leave for today's researchers?

July 2026 marks 50 years since the 10 July 1976 rupture of a safety valve at Icmesa's trichlorophenol reactor, 20 kilometres north of Milan. Operators skipped a routine water quench; an exothermic reaction ran unchecked for six hours; a dioxin-laced plume drifted south across 18 square kilometres, exposing roughly 35,000 residents—the largest such episode on record.

Clinical chemist Paolo Mocarelli drew and froze blood samples within days. Methods did not yet exist to detect dioxin at those concentrations. It took another decade, at roughly £1,000 per sample, before measurements became possible. Epidemiologist Brenda Eskenazi later estimated that analysing 1,000 samples cost about $1 million; the CDC ultimately funded the work. Nearly 90% of original participants still return for assessments today.

Senior science correspondent Jamie Durrani said Mocarelli's experience taught that "scientists need to remain humble… we need to be as transparent as possible if you want to be able to reassure people." Long-term follow-up has also revised the risk model: "Dioxin is known to be carcinogenic… [but] the dioxin exposure was not as harmful as was initially feared," Durrani reported. Seveso also reshaped European regulation—the EU's "Seveso directives" trace directly to the event.

What else moved research portfolios this week?

• US doctoral admissions: Research universities admitted 15% fewer PhD applicants for 2026 than for 2025; the Association of American Universities attributes the drop to Trump-administration actions on federal research.
• UK facilities: The Science and Technology Facilities Council will trim £160 million over four years; the three major sites stay open, smaller national labs face substantial cuts.
• Germany: The Deutsche Forschungsgemeinschaft warned of rising political interference and urged institutions to harden academic freedom.
• Image integrity: Researchers flagged hundreds of doctored Western blots in Thermo Fisher Scientific's online antibody catalogue. The company calls the manipulation "mostly cosmetic" and is investigating.

A historical footnote on miscredited research

The episode closed with Alice Augusta Ball, born 24 July 1892, who at 23 became the University of Hawaii's first African-American and female professor. She converted chaulmoogra oil's fatty acids into water-soluble ethyl esters, producing the first injectable leprosy treatment. She died in 1916 before publishing; colleagues Arthur Dean and Richard Wrenshall rebranded it the "Dean Method." University of Hawaii professors restored her name in the 1970s, and Hawaii proclaimed 28 February Alice Augusta Ball Day in February 2022.

If the membrane prototypes clear heavier crudes and commercial fabrication, refineries built on 30- to 40-year capex cycles will face a choice: retrofit at the next turnaround, or wait for the next generation of separation chemistry.