Dataset for "From FTIR Spectra to Gas Permeability: Machine Learning for Polymer Membrane Screening"
This dataset contains gas permeability measurements and ATR-FTIR spectroscopy data for 204 pure dense polymer membranes, compiled to support machine learning prediction of membrane transport properties. The dataset is intended for researchers working on data-driven prediction of polymer membrane properties, cheminformatics, and structure-property relationships in membrane science.
Cite this dataset as:
Jafari, M.,
2026.
Dataset for "From FTIR Spectra to Gas Permeability: Machine Learning for Polymer Membrane Screening".
Bath: University of Bath Research Data Archive.
Available from: https://doi.org/10.15125/BATH-01709.
Export
Data
Dataset.csv
text/csv (24kB)
Creative Commons: Attribution 4.0
Polymer data
graphs.csv
text/csv (7kB)
Creative Commons: Attribution 4.0
Spectrum-to-polymer mapping
Code
Gas_Permeability … Pipeline.ipynb
application/json (38kB)
Creative Commons: Attribution 4.0
Jupyter notebook containing machine learning pipeline
Creators
Mehryar Jafari
University of Bath
Contributors
Bernardo Castro Dominguez
Supervisor
University of Bath
University of Bath
Rights Holder
Documentation
Data collection method:
Permeability values (Barrer) for six gases (He, H2, CO2, N2, O2, CH4) were sourced from peer-reviewed literature and are provided alongside polymer names, SMILES strings, elemental composition, glass transition temperature, melting temperature, morphological category (amorphous or semi-crystalline), and chemical family classification across 11 classes. A total of 860 ATR-FTIR absorbance spectra are included as individual CSV files, each containing wavenumber (cm-1) and absorbance (arbitrary units) columns covering approximately 400 to 4000 cm-1. A mapping file links each spectrum to its corresponding polymer entry. All polymers are pure homopolymers; blends, copolymers, and composites are not included. Not all polymers have permeability measurements for all six gases.
Documentation Files
README.pdf
application/pdf (101kB)
Creative Commons: Attribution 4.0
Funders
UK Research and Innovation
https://doi.org/10.13039/100014013
DTP 2022-2024 University of Bath
EP/W524712/1
Publication details
Publication date: 9 September 2026
by: University of Bath
Version: 1
DOI: https://doi.org/10.15125/BATH-01709
URL for this record: https://researchdata.bath.ac.uk/1709
Contact information
Please contact the Research Data Service in the first instance for all matters concerning this item.
Contact person: Mehryar Jafari
Faculty of Engineering & Design
Chemical Engineering
Research Centres & Institutes
Centre for Digital, Manufacturing & Design (The Foundry)