DFT vs experiment: formation enthalpies
Dataset ORCHID-Parity. Formation enthalpies of binary compounds read from published papers (literature data), plotted against the lowest-energy Materials Project and OQMD entry at the same composition. Both axes in kJ/mol-atom. The two DFT sources use different energy references and are shown side by side, never averaged. Hover or tap a point for its source, quote and page.
Materials Project
OQMD
x = literature value, y = DFT; dashed line y = x; shaded band ±10 kJ/mol-atom. Residual = DFT − literature.
Statistics for the current filters
| Source | n | MAE | RMSE | mean signed | median |res.| | Pearson r | same sign | within ±10 |
|---|
Errors in kJ/mol-atom. RMSE is dominated by a few gross outliers (see below); the median absolute residual is more robust.
Residuals, Materials Project
Residuals, OQMD
Largest residuals
| Source | Compound | literature | DFT | residual | DFT entry | Paper | Record |
|---|
Typical causes: a per-formula value read as per atom, a table printing magnitudes with the sign given by the column heading, a high-temperature phase or polymorph missing from DFT, or a strongly correlated oxide where the DFT functional and corrections matter.
- DFT formation energies are 0 K values without zero-point or thermal terms; experimental enthalpies are usually at 298 K (some at high temperature).
- The Materials Project applies fitted energy corrections (notably to anions such as O, halides, S, N, H and to some transition-metal oxides; GGA/GGA+U mixing). OQMD's "standard" fit uses elemental reference energies fitted to experimental formation enthalpies. Neither is independent of experiment, and the two must not be averaged.
- The lowest-energy DFT entry at the composition is used; the measured phase may be a different polymorph (the number of DFT entries at that composition is shown), or a disordered or off-stoichiometric phase.
- Literature values include values cited from older work (often a compilation), model values when selected, and machine-extracted values: each point keeps its quote and page so it can be checked. Values printed without a sign or flagged magnitude-suspect are excluded by default.
- Only binary compounds whose phase label gives a formula are matched; see dft_parity.json for the matching rules and exclusion counts.
Sources and credit
Please cite
- R. Arróyave, J. Saal, D. Shin, ORCHID-Values: thermodynamic and phase-equilibrium values read from the literature (version 2026.10; Version 1, machine-extracted), Zenodo (2026). doi:10.5281/zenodo.23735569
- R. Arróyave, J. Saal, D. Shin, ORCHID-Parity: DFT-vs-experiment formation-enthalpy benchmark (version 2026.10; Version 1, machine-extracted), Zenodo (2026). doi:10.5281/zenodo.23735943
- R. Arróyave, J. Saal, D. Shin, ORCHID data: umbrella record of the ORCHID datasets (version 2026.10; Version 1, machine-extracted), Zenodo (2026). doi:10.5281/zenodo.23736065
Also cite the original source of every value you use (paper_citation / paper_doi, and cited_dois for values the paper cites from earlier work); A. Jain et al., The Materials Project: A materials genome approach to accelerating materials innovation, APL Materials 1 (2013) 011002, https://doi.org/10.1063/1.4812323 (CC BY 4.0); J. E. Saal et al., JOM 65 (2013) 1501, https://doi.org/10.1007/s11837-013-0755-4 and S. Kirklin et al., npj Comput. Mater. 1 (2015) 15010, https://doi.org/10.1038/npjcompumats.2015.10 (CC BY 4.0); G. Kim, S. V. Meschel, P. Nash, W. Chen, Experimental formation enthalpies for intermetallic phases and other inorganic compounds, Scientific Data 4 (2017) 170162, https://doi.org/10.1038/sdata.2017.162; B. Hallstedt, The SGTE collection of binary datasets, Calphad 89 (2025) 102833, https://doi.org/10.1016/j.calphad.2025.102833 (CC BY 4.0), and the original assessment named in each file.
@misc{orchid_literature_2026,
author = {Arr{\'o}yave, Raymundo and Saal, James and Shin, Dongwon},
title = {{ORCHID-Values: thermodynamic and phase-equilibrium values read from the literature}},
version = {2026.10},
publisher = {Zenodo},
year = {2026},
doi = {10.5281/zenodo.23735569},
url = {https://doi.org/10.5281/zenodo.23735569},
note = {Version 1, machine-extracted}
}
@misc{orchid_dft_benchmark_2026,
author = {Arr{\'o}yave, Raymundo and Saal, James and Shin, Dongwon},
title = {{ORCHID-Parity: DFT-vs-experiment formation-enthalpy benchmark}},
version = {2026.10},
publisher = {Zenodo},
year = {2026},
doi = {10.5281/zenodo.23735943},
url = {https://doi.org/10.5281/zenodo.23735943},
note = {Version 1, machine-extracted}
}
@misc{orchid_data_2026,
author = {Arr{\'o}yave, Raymundo and Saal, James and Shin, Dongwon},
title = {{ORCHID data: umbrella record of the ORCHID datasets}},
version = {2026.10},
publisher = {Zenodo},
year = {2026},
doi = {10.5281/zenodo.23736065},
url = {https://doi.org/10.5281/zenodo.23736065},
note = {Version 1, machine-extracted}
}
Literature data: CC BY 4.0 (cite ORCHID and the original papers); quotes are short excerpts under the publishers’ copyright, kept for verification. Materials Project and OQMD data: CC BY 4.0, see
NOTICE. Data: dft_parity.json (built by tools/build_dft_parity.py).