材料科学：AI設計による生物から着想を得た水中用接着剤

Super-sticky glues designed with the help of an artificial intelligence (AI) model and inspired by naturally occurring adhesive proteins are described in Nature this week. The glues are shown to fix a hole in a water pipe and stick items under water, demonstrating a wide range of potential applications.

Substances that can adhere in wet environments can be challenging to design. While AI-driven methods have shown success in designing hard materials, soft materials are more complex; super-sticky compounds are particularly tricky as the properties that make materials soft are often the opposite of those that promote adhesion. However, organisms such as bacteria and molluscs produce natural sticky proteins, offering potential inspiration for the design of new super-sticky compounds called hydrogels.

Using a dataset of 24,707 adhesive proteins, Jian Ping Gong and colleagues first develop a protein data-mining tool to guide the design and synthesis of 180 new underwater adhesives. They then use the measured strengths of these adhesives as a data set to train a machine-learning tool, and used that outcome to inform another round of machine-learning-driven design, resulting in superior underwater adhesives. One of the hydrogels, named R1-max was used to glue a rubber duck to a rock in the ocean, which remained stuck despite wave impacts and withstood the tides. Another hydrogel, R2-max, was used as a patch to seal a 20-millimetre-diameter hole in a pipe filled with water. The patch prevented leaking for more than five months.

“Super-adhesive hydrogels such as these that stick strongly to irregular and wet surfaces could be transformational for many biomedical applications, including prosthesis coatings and wearable biosensors,” notes Laura Russo in an accompanying News & Views. She adds that the design approach is versatile, and “could be adapted for other types of functional soft materials”.