A pipe full of water, a hole two centimeters wide, the pressure that pushes the leak out and a soft material that closes it almost immediately. The scene seems to come out of a laboratory designed to convince even the most sceptical, those who raise an eyebrow at every new “revolutionary material” and wait for the real proof. This time the proof is there: a underwater adhesive gel developed by researchers at Hokkaido University has achieved an adhesion strength superior to 1 megapascala value approximately ten times higher than many underwater hydrogels described so far in the literature.
The result comes from very concrete work, even if behind it there are heavy words such as data mining, adhesive proteins and machine learning. The group analyzed thousands of proteins present in nature, searching in their sequences for those small patterns that allow some organisms to remain attached to surfaces even in wet, saline, unstable environments. From that biological catalog, synthetic hydrogels were designed, then tested underwater, corrected, rethought and optimized with predictive models.
From natural proteins to 180 hydrogels tested underwater
The researchers started from approximately 24,700 natural adhesive proteinscoming from bacteria, eukaryotes, archaea and viruses. Organisms very distant from each other, with different evolutionary histories, united by a precious ability: to adhere in environments where water tends to infiltrate everywhere and make almost every bond fragile. Within those sequences, recurring motifs emerged, combinations of amino acids capable of suggesting how to build polymer networks more suitable for adhesion in difficult conditions.
From there the work fell onto the laboratory bench. They have been synthesized 180 hydrogels with different polymer structures and each was measured to understand how well it adhered underwater. The data collected was used to train machine learning models, which indicated new combinations to try. After three cycles of prediction and experimental verification, the best material exceeded the 1 MPa threshold, a level that marks an important leap for soft adhesives designed for humid or submerged environments.
A hydrogel, said without making it sound like something from a university textbook left open halfway, is a soft material composed of water and polymeric networks. We already find it in many areas, from contact lenses to biomedical research. The tricky part comes when you ask it to behave like glue underwater. Water gets between material and surface, disturbs contact, weakens adhesion. Here the new gel manages to push away that barrier and attach itself with surprising force.
From the stamp that an adult can support to the duck left on the rock
To give an idea, Hokkaido University uses a simple image: a piece of this gel the size of a postage stamp, about 2.5 centimeters per side, could theoretically support the weight of an adult of about 63 kilos. It is a comparison to be taken for what it is, linked to specific test conditions, but it helps to understand the scale of the result.
The most curious demonstration concerns a rubber duck. The gel kept her glued to a rock, exposed to waves and tides, for over a year. It seems like an almost comical detail, and in fact it works precisely for this reason: the yellow duck, a bath and play object, becomes a small visual test of resistance. The material remains there, while the sea continues to do its work.
Then there are the less photogenic and more useful tests. The gel quickly sealed a hole from 2 centimeters in a tube filled with water under pressure. In other tests it showed adhesion on different surfaces, from glass to titanium, from ceramic to plastic, and maintained its performance even in conditions of variable salinity, from simple water to environments similar to sea water.
The advantage of the method also lies here: the machine helped to narrow down a search which, proceeding only by trial and error, would have become long and dispersive. There are many possible combinations between chemical components, gel structure and underwater behavior. Machine learning indicated the most promising formulations, then the laboratory did the rest. The digital part suggested, the physical proof decided.
The possible applications are wide. In medicine, an adhesive capable of working on wet fabrics it could become useful in surgery, in advanced wound dressings or in the repair of tissues that are difficult to treat with traditional systems. In the sea, such a material could help maintain submerged infrastructure, sensors, deep exploration tools and soft robotics devices.
The research also points to another practical aspect: the components used are commercially available. This makes the work more interesting, because it brings the material closer to possible future production. Other tests will be needed: durability, safety, dirty surfaces, clinical and industrial use. For now the gel is holding. Even underwater.
