Abstract and research overview
Paper record · Original source linked above
This paper combines DeePMD-based molecular-dynamics sampling and DeePTB electronic-structure models for finite-temperature optical calculations. Electronic information from smaller hybrid-functional (HSE) calculations is used to train a model that evaluates larger atomic configurations, including supercells up to 4,096 atoms. The authors compare phonon-assisted optical absorption and temperature-dependent bandgap renormalization for Si, GaAs and GaP with experimental findings, while discussing the limitations of the finite-temperature sampling and local approximations.
Background on the underlying DeePTB method.
Citation: Yang, Shengguo, et al. “Deep neural network for phonon-assisted optical spectra of semiconductors at finite temperatures.” npj Computational Materials 12, 268 (2026).
Note: This page is an independently written overview, not a verbatim reproduction of the publisher’s abstract.
This page is part of the author's maintained research bibliography. The text may be an author-written overview or an abstract from an earlier manuscript version; consult the linked publisher or preprint for the version of record.
Citation
Shengguo Yang, Shishir Kumar Pandey, Zhanghao Zhouyin, and Qiangqiang Gu. "Deep neural network for phonon-assisted optical spectra of semiconductors at finite temperatures." npj Computational Materials 12, 268 (2026).