High throughput first-principles calculations of bixbyite oxides for TCO applications

Abstract

We present a high-throughput computing scheme based on density functional theory (DFT) to generate a class of oxides and screen them with the aim of identifying those that might be electronically appropriate for transparent conducting oxide (TCO) applications. The screening criteria used are a minimum band gap to ensure sufficient transparency, a band edge alignment consistent with easy n- or p-type dopability, and a minimum thermodynamic phase stability to be experimentally synthesizable. Following this scheme we screened 23 binary and 1518 ternary bixbyite oxides in order to identify promising candidates, which can then be a subject of an in-depth study. The results for the known TCOs are in good agreement with the reported data in the literature. We suggest a list of several new potential TCOs, including both n- and p-type compounds.

4f fine-structure levels as the dominant error in the electronic structures of binary lanthanide oxides

“…The origin of this difference may be the incomplete counteraction of self‐interaction induced by the localized hole states, as mentioned in previous work . Advanced methods, such as GW [13,14] or hybrid functionals based on fixed experimental lattice constants, substantially improve band structure calculations and reproduce the variation behavior found by Prokofiev et al, as reported in our previous work and that of Jiang et al Additionally, our previous work found that the use of such methods as hybrid functionals in conjunction with pseudopotential methods can also introduce errors in the DFT‐predicted lattice parameters and defect levels …”

Comparison of approximations in density functional theory calculations: Energetics and structure of binary oxides

“…In particular, high-throughput first-principles calculations based on density functional theory (DFT) [12,13] are powerful when generating large data for both known and hypothetical materials [14][15][16][17][18]. With the aid of methods for high-throughput computations [17,[19][20][21][22][23][24][25], such data have been used in many studies for understanding the tendency of physical and chemical properties and exploring novel materials [26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43], some of which effectively combine machine learning techniques [33][34][35][36]. High-throughput first-principles calculations have also been used for studying metal oxide systems, for instance to explore transparent conducting oxides [37,38], photocatalysts [39][40][41], and high-κ dielectrics [42].…”

Easily doped p-type, low hole effective mass, transparent oxides

“…High-throughput ab initio computations, on the other hand, can be used to screen large classes of materials, searching for those that exhibit a predetermined basic set of properties, qualifying them as potential candidates for a specific application 19 20 . This approach has already been used in the search for novel organic p -type semiconductors (not transparent) 21 and candidate TCO materials 16 22 , as well as new thermoelectric 23 , piezoelectric 24 and scintillator materials 25 .…”