A framework for stochastic reconstruction of gas diffusion layers with enhanced fiber morphology modulation and additive incorporation
Yang, Danan LU ; Garg, Himani LU
and Andersson, Martin
LU
(2026)
In Materials Today Communications
53.
- Abstract
Gas diffusion layers (GDLs) are critical fibrous porous components utilized in proton exchange membrane fuel cells and water electrolysis systems. Stochastic reconstruction of GDL microstructures facilitates the shift from empirical trial-and-error to performance-driven material design. However, current reconstruction methods often ignore intricate morphological nuances inherent to commercial GDLs, such as fiber curvature, binder/PTFE integration, and localized porosity distribution. Besides, the precise computation of porosity during the generation process remains challenging, typically constrained by the trade-off between simplistic analytical models that neglect fiber intersections and computationally intensive voxel-based... (More)
Gas diffusion layers (GDLs) are critical fibrous porous components utilized in proton exchange membrane fuel cells and water electrolysis systems. Stochastic reconstruction of GDL microstructures facilitates the shift from empirical trial-and-error to performance-driven material design. However, current reconstruction methods often ignore intricate morphological nuances inherent to commercial GDLs, such as fiber curvature, binder/PTFE integration, and localized porosity distribution. Besides, the precise computation of porosity during the generation process remains challenging, typically constrained by the trade-off between simplistic analytical models that neglect fiber intersections and computationally intensive voxel-based approaches. To overcome these limitations, this study proposes a novel, unified GDL reconstruction framework, featuring three key innovations: (1) a generalized fiber generation scheme utilizing diverse curves to construct versatile geometries ranging from straight to highly curved carbon fibers; (2) a hybrid stacking strategy that reduces local porosity oscillations and precisely controls through-plane anisotropy; and (3) a dual-stage porosity calculation algorithm that improves computation efficiency while maintaining accuracy. Furthermore, additive components are integrated via image morphology techniques to mimic realistic carbonized binder and hydrophobic coating distributions. Finally, simulations of fluid transport properties confirm the fidelity of the reconstructed GDLs against prior experimental data and analytical models, establishing this framework as a powerful tool for the targeted design of next-generation proton exchange membrane fuel cells and water electrolysis systems.
(Less)
- author
- Yang, Danan
LU
; Garg, Himani
LU
and Andersson, Martin
LU
- organization
- publishing date
- 2026-04
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Binder and PTFE, Diffusivity, Fiber curvature, Fibrous porous structure, Permeability, Stochastic reconstruction
- in
- Materials Today Communications
- volume
- 53
- article number
- 115303
- pages
- 21 pages
- publisher
- Elsevier
- external identifiers
-
- scopus:105039552562
- ISSN
- 2352-4928
- DOI
- 10.1016/j.mtcomm.2026.115303
- language
- English
- LU publication?
- yes
- id
- 1b86f554-3264-4a80-bffa-c1cba0d6ce1b
- date added to LUP
- 2026-06-04 19:02:52
- date last changed
- 2026-08-12 11:00:16
@article{1b86f554-3264-4a80-bffa-c1cba0d6ce1b,
abstract = {{<p>Gas diffusion layers (GDLs) are critical fibrous porous components utilized in proton exchange membrane fuel cells and water electrolysis systems. Stochastic reconstruction of GDL microstructures facilitates the shift from empirical trial-and-error to performance-driven material design. However, current reconstruction methods often ignore intricate morphological nuances inherent to commercial GDLs, such as fiber curvature, binder/PTFE integration, and localized porosity distribution. Besides, the precise computation of porosity during the generation process remains challenging, typically constrained by the trade-off between simplistic analytical models that neglect fiber intersections and computationally intensive voxel-based approaches. To overcome these limitations, this study proposes a novel, unified GDL reconstruction framework, featuring three key innovations: (1) a generalized fiber generation scheme utilizing diverse curves to construct versatile geometries ranging from straight to highly curved carbon fibers; (2) a hybrid stacking strategy that reduces local porosity oscillations and precisely controls through-plane anisotropy; and (3) a dual-stage porosity calculation algorithm that improves computation efficiency while maintaining accuracy. Furthermore, additive components are integrated via image morphology techniques to mimic realistic carbonized binder and hydrophobic coating distributions. Finally, simulations of fluid transport properties confirm the fidelity of the reconstructed GDLs against prior experimental data and analytical models, establishing this framework as a powerful tool for the targeted design of next-generation proton exchange membrane fuel cells and water electrolysis systems.</p>}},
author = {{Yang, Danan and Garg, Himani and Andersson, Martin}},
issn = {{2352-4928}},
keywords = {{Binder and PTFE; Diffusivity; Fiber curvature; Fibrous porous structure; Permeability; Stochastic reconstruction}},
language = {{eng}},
publisher = {{Elsevier}},
series = {{Materials Today Communications}},
title = {{A framework for stochastic reconstruction of gas diffusion layers with enhanced fiber morphology modulation and additive incorporation}},
url = {{http://dx.doi.org/10.1016/j.mtcomm.2026.115303}},
doi = {{10.1016/j.mtcomm.2026.115303}},
volume = {{53}},
year = {{2026}},
}