Table 4 Selected application-specific datasets for solid-state systems.
From: Graph neural networks for materials science and chemistry
Dataset | Size |
|---|---|
AFLOW239 - calculated properties of materials | >3,400,000 |
Inorganic Crystal Structure Database (ICSD)299 - extensive and well curated experimental database | ≈210,000 |
Pure carbon and C-H-N-O structures at different pressures300 | ≈200,000 |
Materials Project100 - calculated properties of materials | ≈145,000 |
Hypothetical MOF database236 | 137,953 |
NREL Materials Database (NRELMatDB)301 - computational materials database focused on renewable energy applications | ≈60,000 |
CO and H surface binding energy dataset302 | ≈40,000 |
Inorganic materials synthesis recipes303 | 19,488 |
Perovskite structures and energies304 | 18,928 |
CoRE MOF database305 - Experimental MOF database | >14,000 |
bcc iron structures with energies and various kinds of defects306 | 12,193 |
MOF methane adsorption volume of CoRE MOFs (from GCMC)234 | 10,102 |
Elemental boron structures with energies307 | 5038 |
≈4000 | |
DDEC MOF point charges310 | 2932 |
