High-Throughput ab Initio Screening for Two-Dimensional Electride Materials

High-throughput ab initio screening of approximately 34000 materials in the Materials Project was conducted to identify two-dimensional (2D) electride materials, which are composed of cationic layers and anionic electrons confined in a 2D empty space. The screening was based on three indicators: (1) a positive total formal charge per formula unit; (2) layered structures for two-dimensionality; (3) empty spaces between the layer units. Three nitrides, Ca2N, Sr2N, and Ba2N, and the carbide Y2C were identified as 2D electrides, where Ca2N is the only experimentally confirmed 2D electride (Lee, K.; et al. Nature 2013, 494, 336-341). Electron density analysis using ionic radii revealed a smaller number of anionic electrons in Y2C than those in the three nitrides as a result of the partial occupation of the anionic electrons in the d orbitals of Y. In addition, no candidates were identified from the p-block elements, and thus the ab initio screening indicates that the s-block elements (i.e., alkali or alkaline-earth metals) are highly preferable as cation elements. To go beyond the database screening, a tailored modeling was conducted to determine unexplored compounds including the s-block elements that are suitable for 2D electrides. The tailored modeling found that (1) K2Cl, K2Br, Rb2Cl, and Rb2Br dialkali halides are highly plausible candidates, (2) Li2F and Na2Cl dialkali halides are highly challenging candidates, and (3) the Cs2O1-xFx halogen-doped dialkali oxide is a promising candidate.