GO-MACE-23

The complete dataset ("iter-12-filtered") used for training and testing the GO-MACE-23 interatomic potential. This dataset covers a wide range of graphene-oxide relevant configurations, several different stoichiometries, and many different functional groups. DFT labels were generated using CASTEP with the PBE functional. Original data were obtained from Zenodo. Reference energies for isolated atoms are {"C": -148.6811580026, "H": -12.53432584235, "O": -431.6357255604}.

>>> from load_atoms import load_dataset
>>> load_dataset("GO-MACE-23")
GO-MACE-23:
    structures: 3,816
    atoms: 768,256
    species:
        C: 69.86%
        O: 19.54%
        H: 10.60%
    properties:
        per atom: (forces)
        per structure: (energy, split)

License

This dataset is licensed under the CC BY 4.0 license.

Citation

If you use this dataset in your work, please cite the following:

@article{El-Machachi-24-11,
    title = {
        Accelerated First-Principles Exploration
        of Structure and Reactivity in Graphene Oxide
    },
    author = {
        {El-Machachi}, Zakariya and Frantzov, Damyan
        and Nijamudheen, A. and Zarrouk, Tigany
        and Caro, Miguel A. and Deringer, Volker L.
    },
    journal = {Angewandte Chemie International Edition},
    volume = {n/a},
    number = {n/a},
    pages = {e202410088},
    doi = {10.1002/anie.202410088},
}

Properties

Per-atom:

Property	Units	Type	Description
forces	eV/Å	ndarray(N, 3)	force vectors

Per-structure:

Property	Units	Type	Description
energy	eV	float64	total structure energy
split		str	whether the structure was used for training ("train") or testing ("test")

Miscellaneous information

GO-MACE-23 is imported as an InMemoryAtomsDataset:

Importer script for GO-MACE-23

from load_atoms.database.backend import BASE_GITHUB_URL, SingleFileImporter
from load_atoms.database.internet import FileDownload


class Importer(SingleFileImporter):
    @classmethod
    def file_to_download(cls) -> FileDownload:
        return FileDownload(
            url=f"{BASE_GITHUB_URL}/GO-MACE-23/GO-MACE-23.extxyz",
            expected_hash="aeb61b35a44b",
        )

DatabaseEntry for GO-MACE-23

name: GO-MACE-23
year: 2023
description: |
    The complete dataset (``"iter-12-filtered"``) used for training and testing the
    `GO-MACE-23 <https://doi.org/10.1002/anie.202410088>`__ interatomic potential.
    This dataset covers a wide range of graphene-oxide relevant configurations,
    several different stoichiometries, and many different functional groups.
    DFT labels were generated using `CASTEP <https://www.castep.org/>`__ with the PBE functional.
    Original data were obtained from `Zenodo <https://zenodo.org/records/14066557>`__.
    Reference energies for isolated atoms are ``{"C": -148.6811580026, "H": -12.53432584235, "O": -431.6357255604}``.
category: Potential Fitting
minimum_load_atoms_version: 0.2
citation: |
    @article{El-Machachi-24-11,
        title = {
            Accelerated First-Principles Exploration
            of Structure and Reactivity in Graphene Oxide
        },
        author = {
            {El-Machachi}, Zakariya and Frantzov, Damyan
            and Nijamudheen, A. and Zarrouk, Tigany
            and Caro, Miguel A. and Deringer, Volker L.
        },
        journal = {Angewandte Chemie International Edition},
        volume = {n/a},
        number = {n/a},
        pages = {e202410088},
        doi = {10.1002/anie.202410088},
    }
license: CC BY 4.0
per_atom_properties:
    forces:
        desc: force vectors
        units: eV/Å
per_structure_properties:
    energy:
        desc: total structure energy
        units: eV
    split:
        desc: |
            | whether the structure was used
            | for training (``"train"``)
            | or testing (``"test"``)
representative_structure: 1894