A catalyst that loosens the strongest bond in forever chemicals

They are in rain jackets and frying pans — but also in drinking water, food, and human blood. PFAS have become so widespread that in many places, avoiding them is no longer a realistic option.

The reason is not poor regulation or lack of effort. It is chemistry.

PFAS were deliberately designed to resist heat, water and oil — properties that made them immensely useful, but also allow them to persist for years in the environment and the human body. At the heart of that persistence lies one of the strongest bonds in organic chemistry: the carbon–fluorine bond — a link so stable that most treatment technologies simply bounce off it.

“The carbon–fluorine bond is the strongest chemical bond in organic chemistry,” says senior author Zongsu Wei, tenured Associate Professor at the Department of Biological and Chemical Engineering at Aarhus University. “It was designed to be stable — and that stability is exactly what makes PFAS so difficult to deal with once they are released.”

Instead of trying to overpower that bond with harsher conditions, the researchers took a different approach: make the molecule vulnerable first. In simple terms, the catalyst forces PFAS to bind in an uncomfortable position — stretching and weakening its strongest bond — before light-driven breakdown can occur.

The result is not just a new material, but a shift in strategy: whether so-called “forever chemicals” can be broken down at all may depend less on brute force and more on how precisely the chemistry is engineered at the surface.

Why PFAS are so hard to destroy

PFAS comprise thousands of synthetic chemicals that have been used for decades in products ranging from non-stick coatings and firefighting foams to food packaging and industrial surfactants. They were deliberately designed to be highly resistant to heat, water and oil – a chemical durability that also makes them persist for years in the environment and in the human body.

“So right now, the way we try to destroy PFAS compounds is a kind of brute-force approach,” says Zongsu Wei. “Because the carbon–fluorine bond is so difficult to degrade, we add more chemicals and more energy – but that still does not solve the problem.”

The result is a class of remediation strategies that may work in laboratory settings, but are costly, energy-intensive and difficult to scale – and often fail to destroy the molecules themselves.

“In many cases, we are not really eliminating PFAS,” Wei says. “We are transferring them from one place to another – from water to filters, sludge or solid waste. If fluorine stays bound in organic fragments, the problem is not solved.”

This distinction has shifted the focus of PFAS research toward strategies that can dismantle PFAS molecules themselves rather than merely relocating them.

Photocatalysis under ultraviolet (UV) light has therefore attracted growing attention, because it can in principle drive chemical reactions without adding reagents. “But light alone is not enough,” Wei says.

“The molecule has to be activated – meaning pushed out of its chemically comfortable, ultrastable state – and that depends on how it binds to the catalyst surface.”

From brute force to surface chemistry

This insight has directed researchers toward materials whose surfaces can be engineered at the molecular level. Coordination polymers and metal–organic frameworks offer precise control over surface chemistry, electronic structure and the presence of unsaturated metal sites.

“These unsaturated metal sites behave as electron-pulling docking points – what chemists call Lewis acid sites – metal atoms that attract electrons and create chemically sticky landing sites for PFAS molecules, redistributing charge and making the surface more reactive,” Wei says. “That principle guided the design of the catalyst’s local coordination environment.”

Bismuth-based photocatalysts have previously shown promise, thanks to their chemical stability and favourable electronic properties. But efficient cleavage of the carbon–fluorine bond has remained elusive.

“What was missing,” Wei says, “was a way to combine strong surface binding with effective charge transfer to the carbon–fluorine bond itself.”

That challenge set the stage for the new study’s design strategy: instead of trying to overpower the carbon–fluorine bond, the researchers set out to weaken it first.

Designing a catalyst that destabilises PFAS

To test whether surface design could actively promote carbon–fluorine bond activation, the researchers synthesised a bimetallic coordination polymer composed of bismuth with a small amount of titanium.

“We have a barrier to overcome,” Wei explains. “But if we can activate the PFAS molecule, that barrier becomes much lower.”

The two metals were linked through organic ligands, forming a porous, crystalline material with deliberately introduced unsaturated metal sites.

“The whole idea was to make the PFAS molecules less stable,” Wei says. That principle guided the design of the catalyst’s local coordination environment, in which unsaturated metal sites increase surface reactivity.

“Just imagine that you stretch a molecule,” Wei explains. “As the structure loosens, the bond energy drops – and degradation becomes easier.”

Why a trace of titanium makes a difference

Titanium was introduced only in trace amounts relative to bismuth. Rather than forming a separate phase, Ti atoms were incorporated into the coordination framework, subtly altering the local structure and electronic properties.

“Titanium does not dominate the structure,” Wei says. “Its role is to modulate the local electronic environment and increase the active sites.”

The resulting material was characterised using electron microscopy, X-ray diffraction, infrared spectroscopy and X-ray photoelectron spectroscopy to confirm its structure, composition and surface chemistry. Optical and electrochemical measurements probed light absorption, charge separation and electron transfer.

Photodegradation experiments were carried out under irradiation with UV light at 254 nm (UV₂₅₄) in aqueous PFOA solutions at defined laboratory concentrations, enabling direct kinetic comparison across catalyst variants and control conditions.

Crucially, the catalyst was used at an ultralow concentration, equivalent to only a pinch of material in a bathtub of water.

“Using such a low catalyst loading enables us to see whether the effect comes from surface chemistry,” Wei says, “rather than from simply adding more material.”

Testing whether chemistry – not quantity – drives the reaction

Increasing the catalyst concentration actually reduced performance, likely because excess particles scatter UV light and block UV penetration.

To distinguish true chemical degradation from mere adsorption, control experiments were performed without light and without catalyst. Radical scavenging experiments and electron spin resonance spectroscopy identified the reactive species involved – revealing which short-lived chemical “attackers” actually do the work.

“We wanted to understand which species actually drive the reaction,” Wei says. “That is essential if you want to design better catalysts rather than rely on trial and error.”

Degradation products and fluoride release were analysed using high-resolution mass spectrometry and ion chromatography, enabling the researchers to track both molecular breakdown and defluorination.

“By combining experiments with theory, we can connect what we see in the lab with what actually happens at the molecular level,” Wei says.

Finally, the catalyst was tested against mixtures of different PFAS and in real groundwater samples to assess robustness under more realistic conditions, in which competing ions and organic matter can occupy active sites and suppress performance – effectively crowding the catalyst surface and leaving fewer binding sites for PFAS.

Complete breakdown at ultralow catalyst loading

When exposed to UV₂₅₄, the bismuth–titanium coordination polymer broke down PFOA at catalyst concentrations far below what is typically reported. Using this ultralow catalyst loading, the researchers observed complete disappearance of PFOA within five hours – an efficiency that exceeded their own expectations.

“We did not expect full degradation under these conditions,” Wei says.

But molecular disappearance is not the same as full mineralisation, which is why the team also tracked defluorination and intermediate products.

To determine whether the most persistent chemical bonds in PFAS were actually being broken, the researchers measured fluoride release – fluorine appearing as free fluoride in the water – as a direct indicator of carbon–fluorine bond cleavage.

After five hours, more than half the fluorine originally bound in PFOA had been released into solution. The remainder persisted in organic intermediates – underscoring that “PFAS gone” and “fluorine removed” are not the same endpoint.

The efficiency of this process depended strongly on pH. Performance peaked under mildly acidic conditions and dropped sharply at neutral to basic pH – an important practical constraint, because real water systems are often neutral.

“Defluorination is a much stricter metric than degradation,” Wei explains. “It proves that the strongest bonds are actually being broken – not just rearranged – which has been missing in many earlier approaches.”

Why two metals work better than one

Direct comparison with single-metal catalysts highlighted the importance of the bimetallic design. Under identical conditions, the titanium-doped material consistently outperformed its bismuth-only counterpart, achieving nearly double the defluorination efficiency and substantially faster degradation kinetics.

“Adding a small amount of titanium made a disproportionate difference,” Wei says. “That points to a genuine synergistic effect between the two metals.”

Kinetic analysis showed that the reaction followed pseudo-first-order behaviour, meaning that the breakdown rate was predictable and could be compared cleanly across conditions. The rate constant was several times higher than under UV irradiation alone.

Further insight came from scavenger experiments and spectroscopic measurements, which showed that superoxide radicals and electrons dominated the reaction and hydroxyl radicals played only a minor role – consistent with a surface electron-transfer mechanism in which electrons move directly between the catalyst and the PFAS molecule rather than reacting randomly in the surrounding water, as in an indiscriminate oxidation bath.

“This is not a classical hydroxyl-radical-driven system,” Wei notes. “The mechanism is more selective and tightly linked to electron transfer at the surface.”

A controlled breakdown – and how it holds up in real water

Analysis of degradation products over time revealed a stepwise shortening of the perfluorinated carbon chain. Short-chain perfluorocarboxylic acids accumulated transiently before breaking down further, while fluoride release continued to increase as the reaction progressed.

“We see a clear sequence of intermediates,” says Wei. “That gives us confidence that the molecule is being dismantled in a controlled manner.”

When tested against mixtures of different PFAS and in real groundwater samples, overall efficiency decreased compared with ideal laboratory conditions. Even so, the material remained active and structurally stable over multiple cycles.

“Real water is always more challenging,” Wei says. “But the fact that the catalyst remains active and reusable is encouraging.”

Interestingly, performance improved over repeated reaction cycles. This suggests that prolonged UV exposure may gradually activate the material by exposing additional unsaturated metal sites, although the underlying mechanism still needs to be clarified.

Taken together, the results show that carefully engineered surface chemistry can enable both efficient PFAS degradation and meaningful defluorination – even at ultralow catalyst dosages normally considered insufficient for such chemically resilient compounds.

A shift in strategy, not a silver bullet

Although the study demonstrates efficient degradation and partial defluorination of PFOA, Wei is careful not to frame the results as a ready-made solution but as a shift in strategy for PFAS remediation.

“This is not about one perfect catalyst,” Wei says. “It is about showing that surface design can fundamentally change how PFAS molecules behave.”

At the same time, the study makes clear where the approach still falls short. Performance decreases in complex water matrices, and certain PFAS – particularly branched compounds and sulfonates – remain more resistant to degradation than linear carboxylates.

“Those limitations are important,” Wei says. “They show us where the design needs to be improved rather than hidden.”

Designing catalysts for specific PFAS – and the limits ahead

One likely direction, Wei suggests, is to combine catalytic degradation with upstream or downstream treatment steps, such as selective adsorption or filtration, to reduce competition from the matrix for active sites at the catalyst surface. Another is to further refine the coordination environment to increase selectivity toward specific PFAS classes.

“Once you understand which interactions matter,” Wei notes, “you can start designing catalysts for particular molecular structures instead of treating PFAS as one uniform group.”

Equally important is the question of scalability. The ultralow catalyst dosage demonstrated in the study suggests that material efficiency may no longer be the dominant bottleneck.

Cost and energy demands matter as well. “Perhaps in Denmark it is acceptable,” Wei says, “but in low-income countries the problem may simply be left untreated, because the solutions are too expensive.”

A broader playbook for tackling persistent pollutants

Instead, other constraints – such as energy input, light delivery and reactor design – become decisive.

“The chemistry works,” Wei says. “The next question is how to make it work under realistic conditions, at relevant concentrations and energy costs.”

Beyond PFAS, the researchers see broader implications. Many persistent organic pollutants resist degradation for similar reasons: exceptionally strong chemical bonds, weak interactions with catalyst surfaces and inefficient electron-transfer pathways.

“What we show here is a general principle,” Wei concludes. “If you can control how a molecule binds and how electrons move at the interface, you can begin to tackle chemicals that were previously considered almost untouchable.”

Seen in that light, Wei describes the study less as identifying a single best catalyst and more as a design playbook: engineer binding and charge transfer first – then tailor the material to the pollutant, the water chemistry and the treatment context, whether for PFAS or other persistent pollutants.