Perovskite Properties · foundry-ml
dataset_perovskite_asr
Perovskite oxide area-specific resistance data across four temperatures for solid-oxide fuel cell machine learning.
What dataset_perovskite_asr is
dataset_perovskite_asr is a curated tabular materials science dataset designed for predicting the catalytic and electrochemical behaviour of perovskite electrodes. It provides area-specific resistance (ASR) measurements across 289 distinct perovskite oxide compositions, addressing the challenge of identifying efficient electrode materials for solid-oxide fuel cells.
Assembled by Ryan Jacobs, Jian Liu, Harry Abernathy, and Dane Morgan, the collection links structural and chemical descriptions of perovskite oxides with their experimental and modelled transport characteristics. The dataset serves researchers developing regression and classification models to screen candidate electrode formulations without relying solely on slow experimental syntheses.
How it was made
The dataset tabulates material descriptors alongside performance outcomes. Input features include material composition, site-specific elemental groupings, an abbreviated material identifier, the electrolyte class used in electrochemical testing, an indicator showing whether the measurement averages four or more independent studies, and a machine learning-predicted activation barrier for ASR in electronvolts.
The target variables record logarithmic area-specific resistance—measured in Ohm-cm²—at four specific operating temperatures: 773 K, 873 K, 973 K, and 1073 K. The data is structured as a single training split formatted for tabular machine learning workflows.
Key facts
- Number of materials
- 289
- Data format
- Tabular Parquet
- Primary target
- Log ASR (Ohm-cm²) at 773 K, 873 K, 973 K, and 1073 K
- Input features
- Material composition, site groupings, electrolyte class, study count indicator, and ML-predicted ASR barrier
- Machine learning tasks
- Tabular regression and tabular classification
- Release year
- 2023
- Digital Object Identifier (DOI)
- 10.18126/aw1n-3z96
What you can use it for
- Temperature-Dependent ASR Regression
- Train tabular regression models to predict logarithmic area-specific resistance of perovskite electrodes at operational temperatures between 773 K and 1073 K.
- Electrode Screening for Solid-Oxide Fuel Cells
- Evaluate how different chemical compositions and site substitutions influence resistive losses across compatible electrolyte classes.
- Descriptor Assessment
- Benchmark machine learning algorithms against electronic and composition-based descriptors to determine predictive efficacy for perovskite catalytic activity.
Explore the data
The schema, real rows, full-text search and a SQL filter — read in your browser, nothing downloaded.
Explore
Read the real rows without downloading anything
Reading rows…
LiveRead from foundry-ml/dataset_perovskite_asr at the moment you asked. Nothing is cached or stored — every row above came from that request.
Strengths and limits
Good at
- Provides standardized logarithmic resistance targets across four consistent operational temperatures.
- Includes compositional site groupings and electrolyte pairings relevant to solid-oxide fuel cell setups.
- Distinguishes well-studied perovskites by flagging entries averaged over four or more independent materials sources.
Watch out for
- Contains a small sample size restricted to 289 materials.
- Provides only a predefined single training split, requiring users to partition their own test and validation splits.
- Target values are constrained to four discrete temperatures rather than a continuous operational spectrum.
Questions
Can I look inside dataset_perovskite_asr without downloading it?
Yes. This page reads its schema, real rows and per-column statistics in the browser, with full-text search and a SQL filter. Nothing is downloaded to your machine.
What licence is dataset_perovskite_asr under?
Its authors declare other. Read the terms on the original page before reusing the data.
What is area-specific resistance (ASR) in this context?
Area-specific resistance quantifies the electrical and electrochemical resistance of a perovskite electrode normalized by surface area, expressed in Ohm-cm².
At which temperatures is ASR reported?
Logarithmic ASR targets are provided at 773 K, 873 K, 973 K, and 1073 K.
What does the 'Is well studied?' feature denote?
It is an input field indicating whether the reported data for that material composition is averaged over four or more independent measurements.
Which machine learning tasks does the dataset support?
It is formatted for tabular regression to predict continuous ASR values and tabular classification tasks.
How to cite it
Jacobs, Ryan and Liu, Jian and Abernathy, Harry and Morgan, Dane. Critical Assessment of Electronic Structure Descriptors for Predicting Perovskite Catalytic Properties. Materials Data Facility, 2023. https://doi.org/10.18126/aw1n-3z96
Written from dataset_perovskite_asr'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.