a-Si-24

A dataset of synthetic amorphous Silicon structures, taken from Signatures of paracrystallinity in amorphous silicon. Each structure is the final snapshot from a unique melt-quench MD trajectory. The combined dataset covers a wide range of quench rates and densities, and includes the labels from the MTP \(M_{16}^{''}\) potential used to generate the structures.

>>> from load_atoms import load_dataset
>>> load_dataset("a-Si-24")
a-Si-24:
    structures: 3,069
    atoms: 1,317,240
    species:
        Si: 100.00%
    properties:
        per atom: (forces)
        per structure: (energy, label)

Citation

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

@misc{Rosset-24-07,
    title = {Signatures of Paracrystallinity in Amorphous Silicon},
    author = {
        Rosset, Louise A. M. and Drabold, David A.
        and Deringer, Volker L.
    },
    year = {2024},
    doi = {10.48550/arXiv.2407.16681},
}

Properties

Per-atom:

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

Per-structure:

Property	Units	Type	Description
energy	eV	float64	energy of the structure (as labelled by MTP)
label		str	a unique identifier for each structure

Miscellaneous information

a-Si-24 is imported as an InMemoryAtomsDataset:

Importer script for a-Si-24

from __future__ import annotations

from pathlib import Path
from typing import Iterator

import ase
import ase.io
from ase import Atoms
from load_atoms.database.backend import BaseImporter
from load_atoms.database.internet import FileDownload
from load_atoms.progress import Progress

_HASHES = {
    64: "25627b8c50d9",
    216: "da49808517c3",
    512: "654e2e1d1349",
    1000: "ae52f05f2231",
}


class Importer(BaseImporter):
    @classmethod
    def files_to_download(cls) -> list[FileDownload]:
        _base_url = (
            "https://github.com/lamr18/aSi-data/raw/refs/heads/main/data/xyz/"
        )

        return [
            FileDownload(
                url=f"{_base_url}{n}-atoms.xyz",
                expected_hash=hash,
            )
            for n, hash in _HASHES.items()
        ]

    @classmethod
    def get_structures(
        cls, tmp_dir: Path, progress: Progress
    ) -> Iterator[Atoms]:
        with progress.new_task("Parsing files", total=len(_HASHES)) as task:
            for n in _HASHES:
                path = tmp_dir / f"{n}-atoms.xyz"
                for atoms in ase.io.iread(path, index=":"):
                    del atoms.info["cell_origin"], atoms.info["config_type"]
                    yield atoms

                task.update(advance=1)

DatabaseEntry for a-Si-24

name: a-Si-24
year: 2024
category: Synthetic Data
minimum_load_atoms_version: 0.3
description: |
    A dataset of synthetic amorphous Silicon structures, taken from
    `Signatures of paracrystallinity in amorphous silicon <https://arxiv.org/abs/2407.16681>`__.
    Each structure is the final snapshot from a unique melt-quench MD trajectory.
    The combined dataset covers a wide range of quench rates and densities, and includes the labels
    from the MTP :math:`M_{16}^{''}` potential used to generate the structures.
citation: |
    @misc{Rosset-24-07,
        title = {Signatures of Paracrystallinity in Amorphous Silicon},
        author = {
            Rosset, Louise A. M. and Drabold, David A.
            and Deringer, Volker L.
        },
        year = {2024},
        doi = {10.48550/arXiv.2407.16681},
    }

per_atom_properties:
    forces:
        desc: force vectors (as labelled by MTP)
        units: eV/Å
per_structure_properties:
    energy:
        desc: energy of the structure (as labelled by MTP)
        units: eV
    label:
        desc: a unique identifier for each structure

representative_structure: 10