References & Citation¶

Citation¶

If you use this code, data, or methodology in your research, please cite:

D. A. Garzón, L. Himanen, L. Andrade, S. Sadewasser, J. A. Márquez, "ML-guided screening of chalcogenide perovskites as solar energy materials", arXiv:2602.21812 (2026). https://arxiv.org/abs/2602.21812

@misc{garzon2026mlguided,
  title   = {{ML-guided screening of chalcogenide perovskites
             as solar energy materials}},
  author  = {Garz{\'o}n, Diego A. and Himanen, Lauri and Andrade, Luisa
             and Sadewasser, Sascha and M{\'a}rquez, Jos{\'e} A.},
  year    = {2026},
  eprint  = {2602.21812},
  archivePrefix = {arXiv},
  primaryClass  = {cond-mat.mtrl-sci},
  doi     = {10.48550/arXiv.2602.21812},
  url     = {https://arxiv.org/abs/2602.21812},
}

Key References¶

The methods and data sources used in this pipeline:

Step	Reference	DOI
SISSO	Ouyang, R. et al. Phys. Rev. Materials 2, 083802 (2018)	10.1103/PhysRevMaterials.2.083802
Tolerance factor (τ)	Bartel, C. J. et al. Science Advances 5, eaav0693 (2019)	10.1126/sciadv.aav0693
Goldschmidt factor	Goldschmidt, V. M. Naturwissenschaften 14, 477–485 (1926)	10.1007/BF01507527
CrystaLLM	Antunes, L. M. et al. Nature Communications 15, 10570 (2024)	10.1038/s41467-024-54639-7
CrabNet	Wang, A. Y.-T. et al. npj Computational Materials 7, 77 (2021)	10.1038/s41524-021-00545-1
HHI / mineral data	U.S. Geological Survey. Mineral Commodity Summaries 2025	10.5066/P13XCP3R
ESG data	World Bank. Environment, Social and Governance Data (2023)	Data Catalog
Synthesizability (GCNN)	Gu, G. H. et al. npj Computational Materials 8, 71 (2022)	10.1038/s41524-022-00757-z
Synthesizability	Jang, J. et al. J. Am. Chem. Soc. 142, 18836–18843 (2020)	10.1021/jacs.0c07384
UMA MLIP (structure relaxation)	Meta FAIR. UMA: A Family of Universal Models for Atoms arXiv:2506.23971 (2025)	10.48550/arXiv.2506.23971

Raw Data Sources¶

Full citations for every file in data/raw/ — including ionic radii, band gap databases, device data, and PCE limits — are listed on the Raw Data Sources page.

Sustainability Data Sources¶

Full citations for every file in data/sustainability_data/ — including USGS mineral commodity data and World Bank ESG indicators — are listed on the Sustainability Data Sources page.

License¶

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgements¶

DAG acknowledges the support by FCT — Fundação para a Ciência e Tecnologia, I.P. (project ref. 2023.00258.BD). Authors acknowledge the COST Action "Emerging Inorganic Chalcogenides for Photovoltaics (RENEW-PV)", CA21148.