Computational Chemistry · SandboxAQ
aqcat25-dataset
13.5M DFT calculation trajectories for heterogeneous catalysis and machine learning potential training.
What aqcat25-dataset is
aqcat25-dataset is a computational chemistry dataset published by SandboxAQ. It comprises 13.5 million density functional theory (DFT) calculation trajectories targeted at heterogeneous catalysis and machine learning potential training.
The dataset addresses data requirements in chemistry, materials science, and engineering, supplying large-scale quantum mechanical trajectories for fitting data-driven atomic potentials and studying catalytic interfaces.
How it was made
The dataset consists of 13.5 million DFT calculation trajectories formatted as tabular and text data. Records are stored in Parquet files and can be accessed with data processing libraries such as pandas, Polars, and Hugging Face datasets.
Key facts
- Size
- 13.5M trajectories
- Data format
- Parquet
- Modalities
- Tabular, text
- Primary method
- Density functional theory (DFT)
- Primary fields
- Chemistry, materials science, engineering
- Associated preprint
- arXiv:2510.22938
What you can use it for
- Machine learning potential training
- Fitting and benchmarking machine learning interatomic potentials using quantum mechanical energy and trajectory calculations.
- Heterogeneous catalysis investigation
- Analysing catalytic reaction mechanisms and surface interactions using DFT simulation trajectories.
- Materials science simulation
- Informing computational chemistry and engineering workflows that require reference trajectory data for surface and material systems.
Explore the data
This dataset is gated
Its authors require their terms to be accepted before the data can be read. Accept them on the original dataset page. The platform does not accept terms on your behalf, so the rows are not shown here.Strengths and limits
Good at
- Provides a collection of 13.5 million DFT calculation trajectories for catalysis and potential training.
- Stored in standard Parquet format compatible with common tabular processing tools such as Polars and pandas.
Questions
Can I look inside aqcat25-dataset without downloading it?
Not here: its authors gate it behind terms each user must accept on the original page. Its description and statistics are still shown.
What licence is aqcat25-dataset under?
Its authors declare cc-by-nc-sa-4.0. Read the terms on the original page before reusing the data.
What data does aqcat25-dataset contain?
It contains 13.5 million density functional theory calculation trajectories.
What are the primary applications of this dataset?
The dataset is designed for heterogeneous catalysis research and the training of machine learning potentials.
What formats and libraries can be used to read the dataset?
The dataset is provided in Parquet format and is compatible with libraries including pandas, Polars, and datasets.
How to cite it
arXiv:2510.22938
Written from aqcat25-dataset's own dataset card. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.