From virtual to reality: A practical route to design new materials

Abstract

Modern computational techniques that use a combination of electronic structure calculations, adaptive genetic algorithms, and machine learning data analysis have been recently predicting many new unknown structures that may exhibit desired physical properties. Yet, most of these theoretically discovered structures belong to the realm of virtual phase space, and the great challenge to experimentally observe them still remains. Based on the example of the C36 Laves phase in a Co-Fe-Ta system, we demonstrate a practical route to design and produce a material with desired properties.