Version History
Changes made to the QMOF Database: https://doi.org/10.6084/m9.figshare.13147324
Changelog
v14: New single-point calculations at the HLE17, HSE06* (i.e. 10% HF ex.), and HSE06 (25% HF ex.) levels of theory. 12/09/21.
v13: Locked-in version to match the PBE files uploaded to NOMAD. 09/15/21.
v12: Several MOFs taken from the Genomic MOF Database with over/underbonded atoms were removed, as the original authors of the Genomic MOF Database uploaded a fairly large fraction of structures with missing H atoms. Supplemental results from new non-self-consistent (NSCF) calculations with a higher k-point density are now provided (note: this was reverted in v13, as it was discovered that LMAXMIX was not set high enough, effecting the NSCF results for a subset of structures). Removed raw VASP files from Figshare to instead host them on NOMAD. Gave each MOF a unique hash-based identifier, which will match the identifiers on the forthcoming Materials Project MOF Explorer app. 09/14/21.
v11: Same changes as in v12, but the bandgaps.csv file was not made backwards-compatible here. 09/13/21.
v10: Removed irrelevant data from the JSON, reducing the filesize. 09/01/21.
v9: Added ~3000 new DFT-optimized MOFs from the CoRE MOF Database (based on the clean subset identified by Chen and Manz), the Genomic MOF Database, and the CSD MOF Subset. Deprecated 13 structures. Added spacegroup info. Added "synthesized?" flag. Added missing PLDs and LCDs. Fixed 186 structures that had EDIFF = 1e-4 instead of EDIFF = 1e-6 in the INCAR. Removed structures that were duplicates according to Pymatgen's StructureMatcher to avoid confusion. The user no longer needs to run the StructureMatcher as a result. 09/01/21.
v8: Added 1243 new DFT-optimized MOFs. 623 were taken from the Boyd & Woo dataset, 485 were taken directly from the 2019 CoRE MOF FSR Database, 92 were Cu triangle MOFs taken from ToBaCCo, and 44 were Hf MOFs obtained by exchanging the Zr metals of ToBaCCo MOFs by Anderson and coworkers. For the CoRE MOFs, only those found in this pre-curated list were included to maximize structural fidelity. For the hypothetical MOFs, some new ones were introduced using the Boyd & Woo structures as a starting point (e.g. by exchanging metal cations). 3 MOFs were deprecated. Added MOFids, DOIs, spin-dependent CBM/VBM, and initial CIFs for the hypothetical MOFs. 07/12/21.
v7: Deprecated 12 MOFs. Added more properties to JSON file and made it easier to parse. 06/08/21.
v6: Added 2620 DFT-optimized MOFs. 1217 were taken from the CSD using the usual protocol. 1188 were hypothetical MOFs obtained from the Boyd & Woo dataset. 148 were hypothetical MOF-74 and MOF-5 analogues obtained from Haranczyk's nanoporousmaterials.org. 48 were hypothetical Zr MOFs made with ToBaCCo and obtained from Anderson and coworkers. 19 were experimental pyrene MOFs from Smit and coworkers. The maximum number of atoms per unit cell was raised to 500. 5/7/2021.
v5: Release corresponding to the published Matter paper. No changes to the database compared to v3. Fixes a bug in
get_subset_data.pythat did not correctly write out the updated.jsonfile. 2/12/21.v4: Includes a few minor typo fixes and better
.xlsxreader. 1/12/21.v3: Added CM5 partial charges for every structure and 3000+ Bader charges (and spin densities). Patched some minor bugfixes with the unrelaxed properties for a few MOF structures, deprecated a few structures, and flagged more duplicates. Continued restructuring of main QMOF database for increased useability. 12/23/20.
v2: ~1500 new structures with pore-limiting diameter greater than 2.4 Å, computed using Zeo++ prior to structure relaxation, were added to the QMOF database along with their DFT-computed properties. The cap on the maximum number of atoms per primitive cell was raised from 150 to 300. 12/05/20.
Last updated