The liquid electrolytes dataset contains the electronic properties of $LiPF_6$ and seven common carbonate solvents, including ethylene carbonate (EC), propylene carbonate (PC), vinylene carbonate (VC), fluoroethylene carbonate (FEC), dimethyl carbonate (DMC), diethyl carbonate (DEC) and ethyl methyl carbonate (EMC), used in lithium-ion batteries. The dataset contains 362,382 entries with molecular properties, involving the total energy, gradient, dipole moment, and atomization energy, calculated at the $\omega$B97X-D3BJ/def2-TZVPD level of theory using Psi4 (version 1.3) quantum chemistry package.

Dataset Structure

Data Instances

An example of a data instance is as follows:

{
  "id": 0,
  "positions": [
    [-0.9714277386665344, 0.56951904296875, 1.1032522916793823],
    [-0.2717927396297455, 1.0743870735168457, -0.03295067325234413],
    [0.4464282691478729, 0.045661065727472305, -0.56782066822052],
    [0.22332225739955902, -1.0471138954162598, 0.04433632642030716],
    [-0.5465647578239441, -0.7467289566993713, 1.2349183559417725],
    [0.985609233379364, 0.09795906394720078, -1.576427698135376],
    [-0.6657967567443848, -1.599708914756775, 1.8853603601455688],
    [-1.4483437538146973, 1.1785120964050293, 1.8737382888793945]
  ],
  "atomicNumbers": [6, 8, 6, 8, 6, 8, 1, 1],
  "elements": ["C", "O", "C", "O", "C", "O", "H", "H"],
  "charge": 0,
  "multiplicity": 1,
  "totalEnergy": -341.3654864261416,
  "gradient": [
    [-0.10185397416353226, 0.1174161434173584, -0.0037562192883342505],
    [0.05161228030920029, -0.021420234814286232, -0.0743783488869667],
    [0.09550163149833679, -0.06375114619731903, -0.015527772717177868],
    [0.0476410873234272, 0.10103901475667953, -0.17489637434482574],
    [0.01629248820245266, -0.07748205959796906, 0.08027751743793488],
    [-0.11646807938814163, -0.03765171766281128, 0.16160888969898224],
    [0.010279906913638115, -0.01157551072537899, -0.004658835940063],
    [-0.0031323935836553574, -0.006509041413664818, 0.03129011392593384]
  ],
  "dipoleMoment": [
    -0.48487669229507446, -0.1025281623005867, 0.9352725744247437
  ],
  "atomizationEnergy": 0.021487715909586314
}

Data Fields

Field	Description
id	Unique identifier for the data instance
positions	3D Cartesian coordinates of the atoms in Angstroem
atomicNumbers	Atomic numbers
elements	Chemical symbols
charge	Total charge of the system
multiplicity	Spin multiplicity of the system
totalEnergy	Total energy of the system in Hartree
gradient	Gradient of the total energy with respect to atomic positions in Hartree/Angstroem
dipoleMoment	Dipole moment
atomizationEnergy	Atomization energy of the system

Data Splits and Configurations

The dataset has only one train split and one configuration/subset:

battery_round7 (default)

Dataset Creation

Curation Rationale

The present version of liquid electrolytes dataset has been extracted from its original repository, flattened and converted to the parquet format.

Source Data

The original liquid electrolytes dataset blob can be downloaded from Google Drive or ChemRxiv.

Initial Data Collection and Normalization

Other than the changes detailed in Sec. Curation Rationale, no data modification has been performed on the liquid electrolytes dataset, conformant to the terms of its original license (cc-by-nc-nd-4.0).

Personal and Sensitive Information

The liquid electrolytes dataset does not involve any personal or sensitive information.

Considerations for Using the Data

Social Impact of Dataset

The liquid electrolytes dataset enables research in materials science, in particular, training data-driven models for predicting the properties of Li-ion batteries, among others.

Additional Information

Dataset Curators

Steven Dajnowicz, Schrödinger, Inc., New York, New York 10036, United States
Garvit Agarwal, Schrödinger, Inc., New York, New York 10036, United States
James M. Stevenson, Schrödinger, Inc., New York, New York 10036, United States
Leif D. Jacobson, Schrödinger, Inc., Portland, Oregon 97204, United States
Farhad Ramezanghorbani, Schrödinger, Inc., New York, New York 10036, United States
Karl Leswing, Schrödinger, Inc., New York, New York 10036, United States
Richard A. Friesner, Schrödinger, Inc., New York, New York 10036, United States
Mathew D. Halls, Schrödinger, Inc., San Diego, California 92121, United States
Robert Abel, Schrödinger, Inc., New York, New York 10036, United States

Licensing Information

Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Citation Information

@article{Dajnowicz:2022:6271,
  author = {Steven Dajnowicz and Garvit Agarwal and James M. Stevenson and Leif D. 
  Jacobson and Farhad Ramezanghorbani and Karl Leswing and Richard A. Friesner and Mathew 
  D. Halls and Robert Abel},
  journal = {Journal of Physical Chemistry B},
  pages = {6271-6280},
  publisher = {American Chemical Society},
  title = {High-Dimensional Neural Network Potential for Liquid Electrolyte Simulations},
  volume = {126},
  url = {https://pubs.acs.org/doi/abs/10.1021/acs.jpcb.2c03746},
  year = {2022}
}