MIT researchers let AI identify materials’ best strain options

February 12, 2019 // By Julien Happich
In a paper titled "Deep elastic strain engineering of bandgap through machine learning", a team of researchers at MIT and in Russia and Singapore present how they were able to leverage artificial intelligence to predict the property changes of different semiconductor or crystalline materials based on specific strain levels and orientations.

Known as “strain engineering”, the use of localized strain to modify the properties of semiconductors is often used in the electronic industry. Already, based on earlier work at MIT, some degree of elastic strain has been incorporated in some silicon processor chips. Even a 1 percent change in the structure can in some cases improve the speed of a device by 50 percent, by allowing electrons to move through the material faster.

And unlike other ways of changing a material’s properties, like doping which produces a permanent, static change, strain engineering allows properties to be changed on the fly. “Strain is something you can turn on and off dynamically,” explains one of the authors Ju Li, MIT professor of nuclear science and engineering and of materials science and engineering. Very often, strain can be induced through localized heating.

But Strain can be applied in any of six different ways (in three different dimensions, each one of which can produce strain in-and-out or sideways), and with nearly infinite gradations of degree, so the full range of possibilities is impractical to explore simply by trial and error. “It quickly grows to 100 million calculations if we want to map out the entire elastic strain space,” Li says.


Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.