Related Experiment Video
Updated: Aug 6, 2026
08:59
Concurrent Quantitative Conductivity and Mechanical Properties Measurements of Organic Photovoltaic Materials using AFM
Published on: January 23, 2013
Machine Learning Driven Analysis Yields Insight into the Processing Robustness of Organic Photovoltaics
Stephen Wong1, Ankush Kumar Mishra2, Baskar Ganapathysubramanian2,3
1Department of Chemical Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802, United States.
ACS Macro Letters
|July 23, 2026
Summary
Optimizing organic photovoltaics (OPVs) processing is key for efficiency. Acetone as an additive creates a robust processing window, improving reproducibility for high-performance devices.
Area of Science:
- Materials Science
- Chemical Engineering
- Renewable Energy
Background:
- Reproducible manufacturing of high-efficiency organic photovoltaics (OPVs) is challenging due to sensitivity to processing variations.
- Morphological changes during film formation significantly impact device performance and yield.
Purpose of the Study:
- To develop and apply a machine learning (ML)-enabled framework for quantifying and optimizing the processing robustness of OPV films.
- To compare the robustness offered by different processing additives, specifically acetone and 1-chloronaphthalene.
Main Methods:
- Combined Taguchi and space-filling design-of-experiments (DoE) with ML surrogates to map the processing parameter space.
- Quantified robustness using parameter-space fraction, persistence curves, and fill fraction relative to a power conversion efficiency (PCE) threshold.
- Investigated PM6:Y6 blend films using acetone and 1-chloronaphthalene as processing additives.
Main Results:
- Acetone as a solvent additive resulted in a broad, robust processing regime, with approximately 73% of conditions exceeding a 9% PCE threshold.
- Faster evaporation rates, associated with acetone, were hypothesized to yield morphology more tolerant to variability.
- The ML framework successfully disentangled manufacturing robustness from peak power conversion efficiency (PCE).
Conclusions:
- The developed ML-enabled framework provides a data-efficient method for assessing and comparing OPV processing window robustness.
- Acetone demonstrates significant potential for enhancing OPV manufacturing reproducibility.
- This approach integrates with automated laboratories to discover new processing strategies for OPVs.
