Journal article
Prediction of optoelectronic properties of Cu2O using neural network potential
Physical chemistry chemical physics : PCCP, Vol.22(26), pp.14910-14917
07/14/2020
DOI: 10.1039/d0cp01112f
PMID: 32584353
Abstract
Neural network potentials (NNPs) trained against density functional theory (DFT) are capable of reproducing the potential energy surface at a fraction of the computational cost. However, most NNP implementations focus on energy and forces. In this work, we modified the NNP model introduced by Behler and Parrinello to predict Fermi energy, band edges, and partial density of states of Cu2O. Our NNP can reproduce the DFT potential energy surface and properties at a fraction of the computational cost. We used our NNP to perform molecular dynamics (MD) simulations and validated the predicted properties against DFT calculations. Our model achieved a root mean squared error of 16 meV for the energy prediction. Furthermore, we show that the standard deviation of the energies predicted by the ensemble of training snapshots can be used to estimate the uncertainty in the predictions. This allows us to switch from the NNP to DFT on-the-fly during the MD simulation to evaluate the forces when the uncertainty is high.
Details
- Title: Subtitle
- Prediction of optoelectronic properties of Cu2O using neural network potential
- Creators
- Balaranjan Selvaratnam - University of South DakotaRanjit T. Koodali - University of South DakotaPere Miro - University of South Dakota
- Resource Type
- Journal article
- Publication Details
- Physical chemistry chemical physics : PCCP, Vol.22(26), pp.14910-14917
- Publisher
- Royal Soc Chemistry
- DOI
- 10.1039/d0cp01112f
- PMID
- 32584353
- ISSN
- 1463-9076
- eISSN
- 1463-9084
- Number of pages
- 8
- Grant note
- Department of Chemistry of the University of South Dakota (USD) DGE-1633213 / National Science Foundation; National Science Foundation (NSF) 1626516 / NSF; National Science Foundation (NSF)
- Language
- English
- Date published
- 07/14/2020
- Academic Unit
- Chemistry
- Record Identifier
- 9984618515902771
Metrics
3 Record Views