TBHubbard: tight-binding and extended Hubbard model dataset for metal-organic frameworks

Area of Science:

• Materials Science
• Computational Chemistry
• Condensed Matter Physics

Background:

• Metal-organic frameworks (MOFs) offer diverse applications in chemical separations due to their tunable structures.
• Computational screening of MOFs is hindered by the complexity of existing crystallographic data.
• Machine learning (ML) and quantum computing (QC) require extensive, curated datasets for effective application.

Purpose of the Study:

• To augment the QMOF dataset with simulated electronic structure data for metal-organic frameworks (MOFs).
• To enable advanced computational screening and design of MOFs for specific applications.
• To provide data supporting machine learning and quantum computing tasks.

Main Methods:

• Applied tight-binding (TB) and density functional theory (DFT) for electronic structure calculations on MOFs.
• Generated TB representations for 10,000 MOFs.
• Developed Extended Hubbard Model (EHM) representations for 240 transition metal-containing MOFs, including self-consistent U and V parameters.

Main Results:

• Provided extensive simulated electronic structure data for 10,000 MOFs within the QMOF dataset.
• Generated self-consistent Hubbard parameters for a subset of MOFs, crucial for EHM.
• Established computational workflows for identifying structure-property correlations.

Conclusions:

• The augmented QMOF dataset facilitates ML-driven discovery and design of MOFs.
• The provided data supports quantum computing algorithms leveraging TB Hamiltonians and Hubbard parameters.
• Public availability of the data promotes reproducibility and further research in MOF computational science.