High-throughput systematic topological generation of low-energy carbon allotropes

• Main Authors: Vladislav A. Blatov, Changhao Yang, Dingyi Tang, Qingfeng Zeng, Andrey A. Golov, Artem A. Kabanov
• Format: Article
• Language: English
• Published: Nature Publishing Group 2021-01-01
• Series: npj Computational Materials
• Online Access: https://doi.org/10.1038/s41524-021-00491-y

Abstract The search for new materials requires effective methods for scanning the space of atomic configurations, in which the number is infinite. Here we present an extensive application of a topological network model of solid-state transformations, which enables one to reduce this infinite number to a countable number of the regions corresponding to topologically different crystalline phases. We have used this model to successfully generate carbon allotropes starting from a very restricted set of initial structures; the generation procedure has required only three steps to scan the configuration space around the parents. As a result, we have obtained all known carbon structures within the specified set of restrictions and discovered 224 allotropes with lattice energy ranging in 0.16–1.76 eV atom−1 above diamond including a phase, which is denser and probably harder than diamond. We have shown that this phase has a quite different topological structure compared to the hard allotropes from the diamond polytypic series. We have applied the tiling approach to explore the topology of the generated phases in more detail and found that many phases possessing high hardness are built from the tiles confined by six-membered rings. We have computed the mechanical properties for the generated allotropes and found simple dependences between their density, bulk, and shear moduli.