Py-Conjugated: Quantitative Morphology Data for Organic Photovoltaics

Published: 22 March 2021| Version 1 | DOI: 10.17632/9j7wm22cgn.1
Wesley Tatum,


This dataset contains 3 sub-sets of data, all of which describe a series of P3HT:PCBM OPV cells that have received different thermal annealing treatments. By annealing to alter the morphology of the thin films in this well-studied system, a workflow for extracting, reporting, and applying machine learning to quantitative morphology data are validated. The subsets of data are: 1) A .csv containing the extracted and distilled quantitative morphology descriptors for each of the 3 phases in the P3HT:PCBM bulk heterojunction (polymer-rich, fullerene-rich, and mixed) 2) A directory that contains .npy files of the m2py (10.1021/acs.jcim.0c00308) labeled morphologies of the OPV active layers. These are used to generate subset #1 3) A directory that contains the cleaned .npy files containing fast force-distance microscopy images of the OPV active layers. These are used to generate subset #2 Also contained in this dataset are .csv files that list the train-test split used in machine learning model training/validation.



Machine Learning, Polymer Morphology, Organic Electronics