MOFFieldBench: A Machine Learning Benchmark for Continuous Electronic Fields in Metal-Organic Frameworks

Abstract

Machine-learning models for metal-organic frameworks (MOFs) predominantly rely on atomic structures, whereas electron-density (ED) and electrostatic-potential (ESP) fields remain underexplored as scalable learning representations. These continuous fields encode complementary spatial and electronic information relevant to structure-property relationships, particularly pore geometry and guest-framework interactions, but costly first-principles calculations, storage requirements and the scarcity of systematically aligned datasets limit their broader use. Here we introduce MOFFieldBench, a multimodal benchmark comprising approximately 46,000 MOFs with one-to-one aligned crystallographic structures, ED fields and ESP fields. A unified workflow integrates field generation, error-controlled compression and sparse point-cloud representation, supporting field reconstruction, prediction of 13 material properties, multimodal property prediction and pairwise cross-modal retrieval. ED and ESP provide competitive representations for pore geometry, adsorption and separation properties. Multimodal fusion improves several prediction tasks relative to the best unimodal reference, supporting complementary predictive roles for atomic structures and electronic fields. Learned ED reconstruction closely approximates first-principles reference densities, while predicted-density initialization reduces self-consistent-field calculation time by 13.59\% on average in paired tests. These results support electronic fields as practical, informative learning modalities and establish MOFFieldBench as a resource for studying representation choice, multimodal learning and electronic-field-driven MOF modelling.

uMOF: A Universal Database, Benchmark, and Machine Learning Interatomic Potentials for Metal-Organic Frameworks
Théo Jaffrelot Inizan, Prathami Divakar Kamath, Alin Marin Elena +1
arXiv8 Sep 2026PreprintCC BY 4.0Full textCondensed matterarXiv
Multimode continuous-variable quantum key distribution with CMOS-compatible photonic-electronic integration
Yiming Bian, Sarah Bastiaens, Saeed Hakimi +11
arXiv26 Oct 2026PreprintarXiv licenceFull textQuantum physics & quantum informationarXiv
Hybrid Analytical Numerical and Machine Learning Frameworks for Solving Deterministic and Stochastic Differential Equations with Stability, Convergence, and Uncertainty Quantification
Suresh Kumar Sahani
Journal of Intelligent Decision Making and Information Science23 Jul 2026Peer-reviewedCC BYFull text13 citedStatistical & nonlinear physicsOpenAlex
MLIP Detective: Active Failure Mode Discovery Beyond Benchmark Scores for Machine-Learning Interatomic Potentials
Ryuhei Okuno, Nontawat Charoenphakdee, Kaoru Hisama +1
arXiv8 Sep 2026PreprintCC BY 4.0Full textPhysical chemistry Condensed matterarXiv
Physics-Constrained Learning of Dose-Dependent Spectral Degradation in Metal--Organic Frameworks from In Situ Low-Loss EELS
Gabriel T. dos Santos, Roberto dos Reis, Vinayak P. Dravid
arXiv30 Sep 2026PreprintCC BY 4.0Full textCondensed matterarXiv

Cite

BibTeX RIS

BibTeX

@misc{hu2026moffieldbench,
  title = {{MOFFieldBench: A Machine Learning Benchmark for Continuous Electronic Fields in Metal-Organic Frameworks}},
  author = {Zhaolin Hu and Hehe Fan and Chenhao Rao and Wangyihan Guo and Meng Xu and Qiwei Yang and Yi Yang},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.27036},
  eprint = {2610.27036},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.27036},
  note = {CC BY-NC-SA 4.0},
}

RIS

TY  - UNPB
TI  - MOFFieldBench: A Machine Learning Benchmark for Continuous Electronic Fields in Metal-Organic Frameworks
AU  - Zhaolin Hu
AU  - Hehe Fan
AU  - Chenhao Rao
AU  - Wangyihan Guo
AU  - Meng Xu
AU  - Qiwei Yang
AU  - Yi Yang
PY  - 2026
DA  - 2026/10/19
PB  - arXiv
DO  - 10.48550/arxiv.2610.27036
UR  - https://arxiv.org/abs/2610.27036
AB  - Machine-learning models for metal-organic frameworks (MOFs) predominantly rely on atomic structures, whereas electron-density (ED) and electrostatic-potential (ESP) fields remain underexplored as scalable learning representations. These continuous fields encode complementary spatial and electronic information relevant to structure-property relationships, particularly pore geometry and guest-framework interactions, but costly first-principles calculations, storage requirements and the scarcity of systematically aligned datasets limit their broader use. Here we introduce MOFFieldBench, a multimodal benchmark comprising approximately 46,000 MOFs with one-to-one aligned crystallographic structures, ED fields and ESP fields. A unified workflow integrates field generation, error-controlled compression and sparse point-cloud representation, supporting field reconstruction, prediction of 13 material properties, multimodal property prediction and pairwise cross-modal retrieval. ED and ESP provide competitive representations for pore geometry, adsorption and separation properties. Multimodal fusion improves several prediction tasks relative to the best unimodal reference, supporting complementary predictive roles for atomic structures and electronic fields. Learned ED reconstruction closely approximates first-principles reference densities, while predicted-density initialization reduces self-consistent-field calculation time by 13.59\% on average in paired tests. These results support electronic fields as practical, informative learning modalities and establish MOFFieldBench as a resource for studying representation choice, multimodal learning and electronic-field-driven MOF modelling.
ER  -

Abstract

Related

Cite