IBM Supercomputers Discover New Type of Plastic

IBM researchers have used a novel computational chemistry approach to discover the first new class of polymer in over 20 years, creating an ultra strong, lightweight and recyclable type of plastic ideal for electronics.

Chemists at IBM Research used supercomputers to accelerate the discovery process of the synthetic polymer, combining lab experimentation with high-performance computing to develop a material that is stronger than bone, capable of self-healing and 100% recyclable.

The research - undertaken in collaboration with UC Berkeley, Eindhoven University of Technology and King Abdulaziz City for Science and Technology (KACST) - is published in the journal Science.

"By joining forces with IBM Research and bringing together the minds of KACST and IBM scientists, we have managed to merge the strengths of both sides, making it possible to bring forth novel green materials that exhibit excellent properties while being completely recyclable," said Turki bin Saud, vice president of research at KACST.

"We believe that this work can have a significant impact on multiple industries and hope to see more great things come from our collaboration."

The polymer's ability to reform to its original shape, coupled with its recyclable properties, mean that it could replace materials currently used to make high value microchips and manufactured parts in the semiconductor industry that are unusable once defective.

The material could also find use in the transportation and aerospace industries in which environmental factors cause stress fractures to parts that are are difficult to repair or recycle.

Accelerating discovery

Material science has for decades been considered a mature field of study, with most research experimenting with the combination of different types of existing polymers to create desired properties.

By using advanced computing, IBM scientists were able to model new polymer forming reactions to identify several previously undiscovered classes of polymer.

It is hoped that this type of computational chemistry will eventually lead to scientists inserting a list of desired properties into a computer to design a material that perfectly matches those specifications.

"Although there has been significant work in high-performance materials, today's engineered polymers still lack several fundamental attributes," said James Hedrick, an advanced organic materials scientist at IBM Research. "New materials innovation is critical to addressing major global challenges, developing new products and emerging disruptive technologies.

"We're now able to predict how molecules will respond to chemical reactions and build new polymer structures with significant guidance from computation that facilitates accelerated materials discovery."