Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW

Dorothea Golze; Markus Hirvensalo; Patricia Hernández-León; Anja Aarva; Jarkko Etula; Toma Susi; Patrick Rinke; Tomi Laurila; Miguel A. Caro

doi:10.1021/acs.chemmater.1c04279

Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW

Dorothea Golze^*, Markus Hirvensalo, Patricia Hernández-León, Anja Aarva, Jarkko Etula, Toma Susi, Patrick Rinke, Tomi Laurila, Miguel A. Caro^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › Scientific › peer-review

36 Citations (Scopus)

167 Downloads (Pure)

Abstract

We present a quantitatively accurate machine-learning (ML) model for the computational prediction of core-electron binding energies, from which X-ray photoelectron spectroscopy (XPS) spectra can be readily obtained. Our model combines density functional theory (DFT) with GW and uses kernel ridge regression for the ML predictions. We apply the new approach to disordered materials and small molecules containing carbon, hydrogen, and oxygen and obtain qualitative and quantitative agreement with experiment, resolving spectral features within 0.1 eV of reference experimental spectra. The method only requires the user to provide a structural model for the material under study to obtain an XPS prediction within seconds. Our new tool is freely available online through the XPS Prediction Server.

Original language	English
Pages (from-to)	6240−6254
Journal	Chemistry of Materials
Volume	34
Issue number	14
DOIs	https://doi.org/10.1021/acs.chemmater.1c04279
Publication status	Published - 13 Jul 2022
MoE publication type	A1 Journal article-refereed

Access to Document

10.1021/acs.chemmater.1c04279Licence: CC BY

Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based MaterialsFinal published version, 5.59 MBLicence: CC BY

Cite this

Golze, D., Hirvensalo, M., Hernández-León, P., Aarva, A., Etula, J., Susi, T., Rinke, P., Laurila, T., & Caro, M. A. (2022). Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW. Chemistry of Materials, 34(14), 6240−6254. https://doi.org/10.1021/acs.chemmater.1c04279

@article{539299cff5e346a4853de7a0c2c83472,

title = "Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW",

abstract = "We present a quantitatively accurate machine-learning (ML) model for the computational prediction of core-electron binding energies, from which X-ray photoelectron spectroscopy (XPS) spectra can be readily obtained. Our model combines density functional theory (DFT) with GW and uses kernel ridge regression for the ML predictions. We apply the new approach to disordered materials and small molecules containing carbon, hydrogen, and oxygen and obtain qualitative and quantitative agreement with experiment, resolving spectral features within 0.1 eV of reference experimental spectra. The method only requires the user to provide a structural model for the material under study to obtain an XPS prediction within seconds. Our new tool is freely available online through the XPS Prediction Server. ",

author = "Dorothea Golze and Markus Hirvensalo and Patricia Hern{\'a}ndez-Le{\'o}n and Anja Aarva and Jarkko Etula and Toma Susi and Patrick Rinke and Tomi Laurila and Caro, {Miguel A.}",

note = "| openaire: EC/H2020/756277/EU//ATMEN Funding Information: The authors acknowledge funding from the Academy of Finland under Projects 316168 (D.G.), 334532 (P.R.), 310574, 329483, 330488 (M.A.C.), and 321713 (M.A.C. and P.H.-L.) and from their flagship program Finnish Center for Artificial Intelligence (FCAI), from the Emmy Noether Programme of the German Research Foundation under Project Number 453275048 (D.G.), from COST action CA18234, and from the European Research Council (ERC) under the European Union{\textquoteright}s Horizon 2020 Research and Innovation Programme, under Grant Agreement No. 756277-ATMEN (T.S.). Computing time from CSC–IT Center for Science, allocated for the Grand Challenge Project XPEC, is gratefully acknowledged. Part of this work was carried out during a HPC-Europa3 mobility exchange (Horizon 2020 Program under Grant Agreement 730897). We thank V.L. Deringer for providing the a-CO structural models used in this study. x Publisher Copyright: {\textcopyright} ",

year = "2022",

month = jul,

day = "13",

doi = "10.1021/acs.chemmater.1c04279",

language = "English",

volume = "34",

pages = "6240−6254",

journal = "Chemistry of Materials",

issn = "0897-4756",

publisher = "American Chemical Society",

number = "14",

}

Golze, D, Hirvensalo, M, Hernández-León, P, Aarva, A, Etula, J, Susi, T, Rinke, P , Laurila, T & Caro, MA 2022, 'Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW', Chemistry of Materials, vol. 34, no. 14, pp. 6240−6254. https://doi.org/10.1021/acs.chemmater.1c04279

Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW. / Golze, Dorothea; Hirvensalo, Markus; Hernández-León, Patricia et al.
In: Chemistry of Materials, Vol. 34, No. 14, 13.07.2022, p. 6240−6254.

Research output: Contribution to journal › Article › Scientific › peer-review

TY - JOUR

T1 - Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW

AU - Golze, Dorothea

AU - Hirvensalo, Markus

AU - Hernández-León, Patricia

AU - Aarva, Anja

AU - Etula, Jarkko

AU - Susi, Toma

AU - Rinke, Patrick

AU - Laurila, Tomi

AU - Caro, Miguel A.

N1 - | openaire: EC/H2020/756277/EU//ATMEN Funding Information: The authors acknowledge funding from the Academy of Finland under Projects 316168 (D.G.), 334532 (P.R.), 310574, 329483, 330488 (M.A.C.), and 321713 (M.A.C. and P.H.-L.) and from their flagship program Finnish Center for Artificial Intelligence (FCAI), from the Emmy Noether Programme of the German Research Foundation under Project Number 453275048 (D.G.), from COST action CA18234, and from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation Programme, under Grant Agreement No. 756277-ATMEN (T.S.). Computing time from CSC–IT Center for Science, allocated for the Grand Challenge Project XPEC, is gratefully acknowledged. Part of this work was carried out during a HPC-Europa3 mobility exchange (Horizon 2020 Program under Grant Agreement 730897). We thank V.L. Deringer for providing the a-CO structural models used in this study. x Publisher Copyright: ©

PY - 2022/7/13

Y1 - 2022/7/13

N2 - We present a quantitatively accurate machine-learning (ML) model for the computational prediction of core-electron binding energies, from which X-ray photoelectron spectroscopy (XPS) spectra can be readily obtained. Our model combines density functional theory (DFT) with GW and uses kernel ridge regression for the ML predictions. We apply the new approach to disordered materials and small molecules containing carbon, hydrogen, and oxygen and obtain qualitative and quantitative agreement with experiment, resolving spectral features within 0.1 eV of reference experimental spectra. The method only requires the user to provide a structural model for the material under study to obtain an XPS prediction within seconds. Our new tool is freely available online through the XPS Prediction Server.

AB - We present a quantitatively accurate machine-learning (ML) model for the computational prediction of core-electron binding energies, from which X-ray photoelectron spectroscopy (XPS) spectra can be readily obtained. Our model combines density functional theory (DFT) with GW and uses kernel ridge regression for the ML predictions. We apply the new approach to disordered materials and small molecules containing carbon, hydrogen, and oxygen and obtain qualitative and quantitative agreement with experiment, resolving spectral features within 0.1 eV of reference experimental spectra. The method only requires the user to provide a structural model for the material under study to obtain an XPS prediction within seconds. Our new tool is freely available online through the XPS Prediction Server.

UR - http://www.scopus.com/inward/record.url?scp=85135248837&partnerID=8YFLogxK

U2 - 10.1021/acs.chemmater.1c04279

DO - 10.1021/acs.chemmater.1c04279

M3 - Article

AN - SCOPUS:85135248837

SN - 0897-4756

VL - 34

SP - 6240−6254

JO - Chemistry of Materials

JF - Chemistry of Materials

IS - 14

ER -

Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW

Abstract

Access to Document

Other files and links

Fingerprint

NEXTCELL: Next generation interatomic potentials to simulate new cellulose based materials

LEARNSOLAR: Rinke-LearnSolar

-: Formation of CO, CH4 and CH3OH by electrochemical reduction of CO2

Science-IT

Aalto University: Machine Learning Gives Material Science Researchers a Peek at the Answer Key

A model trained to predict spectroscopic profiles helps to decipher the structure of materials

Cite this

Accurate Computational Prediction of Core-Electron Binding Energies in Carbon-Based Materials: A Machine-Learning Model Combining Density-Functional Theory and GW

Abstract

Access to Document

Other files and links

Fingerprint

Projects

NEXTCELL: Next generation interatomic potentials to simulate new cellulose based materials

LEARNSOLAR: Rinke-LearnSolar

-: Formation of CO, CH4 and CH3OH by electrochemical reduction of CO2

Equipment

Science-IT

Press/Media

Aalto University: Machine Learning Gives Material Science Researchers a Peek at the Answer Key

A model trained to predict spectroscopic profiles helps to decipher the structure of materials

Cite this