The next storage revolution, up close and personal
Machine-learning helps identify weaknesses in next-gen materials.
This is an extreme close-up metal-organic framework, or MOF, a self-assembling three-dimensional compound made of connected metallic and organic atoms.
A hot topic for materials science research, MOFs are potentially suitable for hundreds of applications, including water extraction, storing hazardous chemicals, or functioning as fuel cells for hydrogen-powered cars. Pundits predict they will be as important in this century as plastics were in the last one.
The wonder-substances, however, have one major drawback: they break.
“That MOFs are so porous makes them highly adaptable for all kinds of different applications, but at the same time their porous nature makes them highly fragile,” says chemical engineer David Fairen-Jimenez from Cambridge University in the UK.
To better understand this aspect, he and colleagues have developed a machine-learning algorithm in order to better predict the vulnerabilities and stresses of the materials. In a study published in the journal Matter, they detail results of tests conducted on 3000 existing MOFs, and a number yet to be synthesised.
The research, the scientists say, forms an important data base that will save time for other people in the field.
“We are now able to explain the landscape for all the materials at the same time,” says Fairen-Jimenez. “This way, we can predict what the best material would be for a given task.”