High-Pressure Phase Diagram of the Ti–O System

Abstract

Titanium oxides are technologically important compounds. The chemistry of the Ti–O system is quite rich, largely because of the multiple oxidation states that titanium atoms can take. In this work, using a combination of variable-composition evolutionary crystal structure prediction (USPEX code) and data mining (Materials Project), we predicted all of the stable titanium oxides in the pressure range 0–200 GPa and found that 27 compounds can be stable at different pressures. We resolved contradictions between previous works and predicted four hitherto-unknown stable phases: P21/c-TiO3, I4/mmm-Ti3O2, Imm2-Ti5O2, and R3̅-Ti12O5. We also showed that the high-pressure P6̅m2-TiO phase is an electride.

High-throughput computational materials screening of transition metal peroxides

“…Compared with the P2 1 /c and Pm% 3n TiO 3 predicted by previous work, the total energy of the Pmmn-TiO 3 respectively decreases by 1.41 and 2.75 eV per f.u. 18,19 The detailed crystal information for both the P2 1 /c-ScO 2 and Pmmn-TiO 3 structures can be found in the ESI, † Table S1.…”

“…[11][12][13][14] In order to predict novel structures, researchers commonly employ structure search software such as RG 2 , 15 USPEX, 16 and CALYPSO. 17 Novel structures such as TiO 3 , 18,19 YO 3 , 20 VO 4 , 21 CaO 3 , 22 LaS 2 , 23 have been predicted to be stable at high pressure, but they have not yet been included in crystal structure databases, such as the Material Project Database, 24 the AFLOWLIB Database, 25 and the Open Quantum Materials Database. 26 In the global structure optimization, the computing of a sufficient number of structures is required to establish an accurate potential energy surface, and the diversity of structures will significantly increase the difficulty of predicting multi-component systems.…”

Machine Learning-Accelerated Discovery of A2BC2 Ternary Electrides with Diverse Anionic Electron Densities

“…Although this screening method can highly accelerate the pace of electride identification, new electrides are limited to materials already existing in the databases. The second approach is to design electrides with completely new structures using structure prediction algorithms. − In contrast to computational screening, structure prediction may yield unexpected discoveries beyond the content of the material databases. However, it would be too demanding to use the characteristics of electrides as the target for a global structure search without any restrictions on possible compositions.…”

Electride Formation in Ba–P System and the Unexpected Structure Transition of Electrides under Pressure

“…The identification of typical electrides, such as C12A7:e – and Ca 2 N:e – , was done by experimental screening. , Recently, high-throughput structure search algorithm combined with first-principles calculations has played an increasingly important role in discovering new materials including electrides. ,,,− For example, 1D electride Sr 5 P 3 was predicted using the evolutionary structure search method USPEX − in the strontium–phosphorus (Sr–P) system and then was experimentally synthesized . It is known that alkaline earth (AE) metals, calcium (Ca) and barium (Ba), have similar outer electron structures and electronegativities with Sr.…”