Speaker
Description
The discovery of novel materials increasingly relies on first-principles electronic-structure theory, but the value of these predictions depends on their reproducibility and connection to measurable observables. A recent community effort demonstrates that standardized, automated workflows (such as AiiDA) can establish high-precision reference datasets. By cross-validating all-electron calculations for nearly 1,000 equations of state across the periodic table, this effort provides a shared test bed to verify pseudopotential-based approaches [1].
Building on this rigorous foundation, we apply high-throughput screening to resolve physics questions that remain elusive on a material-by-material basis. Utilizing the computational 2D materials database (C2DB), we screened hundreds of stable, nonmagnetic, direct-gap monolayers to determine if reduced dimensionality intrinsically enhances the radiative transition strength. Our analysis identified multiple 2D candidates with transition strengths that rival or exceed those of molybdenum disulfide (MoS2), while establishing the performance limits of monolayers relative to traditional III-V semiconductors like GaAs and GaN [2].
We further anchor these results to experimental benchmarks by examining the accuracy of band structure predictions beyond simple energy gaps. Using a curated dataset of 60 experimental effective masses from 18 semiconductors, we quantify the performance of different exchange-correlation approximations. Our findings highlight that achieving accurate band gaps does not guarantee reliable predictions of band curvature (effective mass) and demonstrate the critical role of nonlocal exchange in capturing the physics of charge carrier transport [3].
These results provide a rigorous framework for high-throughput discovery, ensuring that the next generation of data-driven materials design is anchored in both computational precision and physical reality.
[1] E. Bosoni, …, O. Rubel et al., Nat. Rev. Phys. 6, 45-58 (2024).
[2] A. F. Gómez-Bastidas, K. Sriram, A. C. Garcia-Castro, and O. Rubel, Phys. Rev. B 111, 195426 (2025).
[3] M. Laurien and O. Rubel, Phys. Rev. B 106, 045204 (2022).
| Keyword-1 | reproducible workflows |
|---|---|
| Keyword-2 | ab initio benchmarking |
| Keyword-3 | high-throughput screening |