Many essential structural materials used in applications including electronics, civil infrastructure and aerospace must be especially strong and durable, but the same chemical properties that give these materials their ideal traits make them difficult to recycle. Wind turbine blades, for example, endure incredibly strong forces and may last for 20 to 30 years. After they wear out, though, the massive structures – which can be up to 400 feet long – end up in landfills.

In a newly published paper, Beckman and Massachusetts Institute of Technology researchers describe a novel method for creating tough, but recyclable materials with properties that persist across generations.

“This is a major advance in a series of work we’ve been doing as part of our Department of Energy center, geared at designing an end-of-life strategy for thermosetting plastics that we used to think of as not recyclable,” said Nancy Sottos, head of the Department of Material Science and Engineering and a Swanlund Endowed Chair and Center for Advanced Study Professor. “This paper lays the groundwork for rethinking the lifecycle of thermoset polymers and opens up a new set of building blocks for designing them.”

On a microscopic scale, many materials are made of polymers: long chains of repeating molecules. For certain materials, called thermosets, these polymers are connected through a process called crosslinking. In crosslinking, other molecules form bridges between the polymers, like rungs on a rope ladder. On a larger scale, when thousands of crosslinks and polymer strands are involved, the structure may look more like a 3D spiderweb. 

Unfortunately, crosslinks are also the reason why thermosets are hard to recycle. In order to melt a material down and change it into something else, the crosslinks must be removed to get the original, disconnected polymer chains – otherwise, the material will be stuck in its previous shape. However, it is difficult to reverse crosslinking, and the removal process often alters the polymers and deteriorates the thermoset’s quality. Other materials made of polymers — for example, single-use plastics like water bottles — do not have crosslinks and can be recycled much more easily, but lack the properties required for industrial applications. 

However, crosslinking is not the only way to make tough materials. Polymer chains experience entanglement, which is exactly as it sounds: long strands interlacing to form a dense, tangled network. As anyone who has ever had to untangle a box of cables can attest, intermeshed strands can be very difficult to pull apart, even if they are not physically connected to each other.

Zhenchuang Xu, lead author on this paper and previous postdoctoral researcher at Beckman, explained that this work began by investigating whether tough materials require crosslinking, or if entanglement can do the job just as well. The group made samples using two similar polymers – one was able to crosslink, while the other could not. Surprisingly, both materials had comparable properties. 

The researchers still determined that a small number of crosslinks would be important for maintaining these properties over time, as without any connections at all, the polymers could wriggle out of their entangled states. This would weaken the material. They created polymers that would only crosslink at a few sites, and specifically chose molecules that would crosslink in a controlled, reversible way.

More work was conducted to better understand how changing certain variables, such as polymer length, altered the properties of the material. The thermosets were also broken down and reconstructed multiple times, and their properties – including appearance, strength and stiffness – remained consistent.

“One of the innovations of this paper is how we inferred all these design rules and network properties from things we can’t directly measure or see, even with a microscope,” said Sameh Tawfick, a professor of mechanical science and engineering. “In our case specifically, I think this comes because of all the groups in the Autonomous Materials Systems Group – from chemistry, material science, mechanical engineering and so on – who can exchange different languages and help each other infer these rules.”

For example, the researchers performed mechanical tests in which they pulled on their sample materials to see how they behaved and when they broke. They also inferred chemical traits based on how material properties changed at different temperatures. 

In the future, the group intends to incorporate new experimental tools, including high-throughput pipelines, different models and artificial intelligence to understand how the fundamentals described in this study could be used with other molecules to make materials with specific properties.

“Another direction, going forward, is learning how to reprogram these materials so that you can steer the performance in another direction — regenerating them not with the same properties, but with a targeted set of properties that are different from the original generation,” said Jeff Moore, a professor of chemistry and Stanley O. Ikenberry Endowed Chair Emeritus. “Maybe you want a material that is tougher than the previous generation, or maybe you want to improve its stiffness. I think we are opening up a new field of generational materials systems.”

The paper, “Chain entanglements enable regeneration of high-performance thermosets”, can be accessed here: https://doi.org/10.1038/s41563-026-02646-

This work was a collaborative effort with the Massachusetts Institute of Technology, led by Jeremiah Johnson, the A. Thomas Guertin Professor of Chemistry, with graduate student Shuyi Zhang. This paper’s authors also include Beckman-affiliated researchers Edgar Mejia, Tyler Price, Ignacio Arretche, Ruishi Lei, Valerie Chen, Hannah Liu, Shaofeng Huang and Boran Chen.

This research was supported as part of the Regenerative Energy-Efficient Manufacturing of Thermoset Polymeric Materials, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences at the University of Illinois Urbana-Champaign and the Massachusetts Institute of Technology under award DE-SC0023457.