Related Experiment Video
Updated: Jun 19, 2026
12:12
On the Preparation and Testing of Fuel Cell Catalysts Using the Thin Film Rotating Disk Electrode Method
Published on: March 16, 2018
Probing Ink Evaporation-Deposition Behavior for Scalable Catalyst-Layer Fabrication in PEM Electrolysis: Systematic
Ziying Huang1, Xiangyu Fu1, Ronghui Qi1
1School of Chemistry and Chemical Engineering, South China University of Technology, Guangzhou 510640, China.
ACS Applied Materials & Interfaces
|June 18, 2026
Summary
This study uses machine learning to optimize catalyst ink deposition for proton exchange membrane electrolyzers, improving fabrication uniformity and performance. Key factors like solvent content and temperature were identified to control deposition patterns for scalable manufacturing.
Area of Science:
- Materials Science and Engineering
- Electrochemistry
- Chemical Engineering
Background:
- Nonuniform catalyst-ink deposition, like coffee-ring effects, hinders scalable fabrication of proton exchange membrane (PEM) membrane-electrode assemblies (MEAs).
- This nonuniformity negatively impacts catalyst layer uniformity, pore structure, and mass transport in PEM electrolysis, limiting device performance and manufacturability.
Purpose of the Study:
- To integrate single-droplet metrology with explainable machine learning for quantifying catalyst ink evaporation-deposition behavior.
- To establish quantitative morphology descriptors and identify key formulation/process parameters influencing deposition uniformity.
- To develop a framework for predicting optimal process windows for scalable PEM catalyst layer fabrication.
Main Methods:
- Systematic single-droplet metrology combined with optical microscopy and 3D profilometry to characterize ink morphology.
- Development of quantitative descriptors: normalized variance, thickness-partitioned area fractions, and edge-to-center grayscale ratio (ECCR).
- Training a gradient-boosting decision tree model with SHAP analysis on a dataset spanning ionomer properties, solid content, solvent fraction, and substrate temperature.
Main Results:
- The model accurately captured relationships between parameters and morphology metrics (R² = 0.74–0.88), identifying critical interaction terms like isopropanol fraction × equivalent weight.
- An empirical deposition uniformity index (DUI) was defined, enabling prediction of optimal process windows for uniform deposition.
- Validation confirmed improved deposition uniformity and electrochemical stability within the predicted optimal window (e.g., specific isopropanol fraction and substrate temperature ranges).
Conclusions:
- An integrated framework combining droplet metrology, explainable AI, and process-window prediction offers practical guidance for catalyst ink formulation.
- The developed approach facilitates scalable fabrication of uniform PEM catalyst layers, addressing a key bottleneck in electrolyzer manufacturing.
- Relative humidity and droplet volume showed limited impact on deposition patterns compared to formulation and substrate temperature.
