The inverse materials design is a key topic of materials science nowadays. The proposed software solutions are useful tools for decision support at a pre-synthetic stage. Though, the existing methods are restricted by predefined elemental composition and can search for new materials only in a small part of an entire chemical space. Here we would like to present the machine-learning approach i.e. free from the mentioned restriction and able to propose novel materials with different elemental compositions and crystal structures. The method was tested on generating super-hard materials and proved and ability to generate well-known oxides or carbides, as well as novel compounds with three or four elements inside.
Please, cite if using: Vadim Korolev, Artem Mitrofanov, Artem Eliseev, and Valery Tkachenko. Machine-learning-assisted search for functional materials over extended chemical space. Materials Horizons, 2020, 7, 2710-2718, DOI: 10.1039/D0MH00881H