This Glue Sets Underwater in 10 Seconds, Held 2 Kilograms for Three Years, and Dissolves Back Into Liquid

Every surface underwater is covered by a hydration layer — a molecularly thin film of water that clings to it and refuses to be displaced. That film is why adhesives that work fine in air fail when submerged: the glue never actually touches the thing it is supposed to stick to.

Chinese researchers have built an adhesive that removes the film itself. The material, which they call BP16TPB, is a supramolecular ionic liquid — a salt-like substance whose molecules are held together by reversible interactions rather than permanent chemical bonds — mixed with dimethyl sulfoxide, or DMSO.

The mechanism is a surface-tension effect. When the mixture meets water, the DMSO begins diffusing outward, and the resulting difference in surface tension across the boundary drives a rapid motion known as Marangoni flow. That flow strips the hydration layer off the surface while the remaining molecules reassemble into a water-resistant network. "Marangoni flow not only speeds up supramolecular aggregation but also effectively penetrates hydration layers," the researchers wrote.

The numbers are the reason the work is getting attention. On ceramic, the adhesive reached 1.1 megapascals of bonding strength in 10 seconds and 1.3 megapascals within five minutes. A test joint held a 2-kilogram weight underwater for more than three years without failing. The team also bonded copper, epoxy resin and polyamide, and ran the tests in acidic, alkaline and salty water — the last of which is the condition that matters for anything intended to be used in an ocean.

Then there is the property most structural adhesives do not have: it comes apart on purpose. Dissolve the bond back into DMSO and the material returns to liquid form, ready to be applied again. Across eight cycles of reuse, the researchers measured no meaningful loss of performance. For a class of materials whose entire design philosophy is permanence, a glue that can be recovered and reapplied changes the economics of repair.

The limitation is heat. Performance falls off sharply above 70 degrees Celsius, roughly 158 Fahrenheit, which rules out engine compartments, exhaust systems and anything else that runs hot. Cold seawater, by contrast, is well inside the working range.

The applications the team points to are all ones where the alternative is currently expensive and slow: patching ship hulls and offshore platforms without dry-docking them, sealing underwater pipelines in place, and attaching environmental sensors to submerged structures that would otherwise need mechanical fasteners drilled into them. None of those jobs is impossible today. All of them currently involve divers, epoxy that cures for hours, and a fair amount of hoping.