Direct machine-learning prediction of solar-cell efficiency limits is cheaper and more accurate than predicting absorption spectra first, and the resulting errors are comparable to the spread between different DFT methods.
High-throughput optical absorption spectra for inorganic semiconductors
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
An optical absorption spectrum constitutes one of the most fundamental material characteristics, with relevant applications ranging from material identification to energy harvesting and optoelectronics. However, the database of both experimental and computational spectra is currently lacking. In this study, we designed a computational workflow for the optical absorption spectrum and integrated the simulated spectra into the Materials Project. Using density-functional theory, we computed the frequency-dependent dielectric function and the corresponding absorption coefficient for more than 1000 solid compounds of varying crystal structure and chemistry. The computed spectra show excellent agreement, as quantified by a high value of the Pearson correlation, with experimental results when applying the band gap correction from the HSE functional. The demonstrated calculated accuracy in the spectra suggests that the workflow can be applied in screening studies for materials with specific optical properties.
citation-role summary
dataset 1
citation-polarity summary
fields
cond-mat.mtrl-sci 1years
2025 1verdicts
CONDITIONAL 1roles
dataset 1polarities
use dataset 1representative citing papers
citing papers explorer
-
The carbon cost of materials discovery: Can machine learning really accelerate the discovery of new photovoltaics?
Direct machine-learning prediction of solar-cell efficiency limits is cheaper and more accurate than predicting absorption spectra first, and the resulting errors are comparable to the spread between different DFT methods.