3D X-ray imaging and simulations uncover how structure affects fuel-cell performance

Proton exchange membrane fuel cells (PEMFCs) represent a key clean energy technology for applications ranging from electric vehicles to backup power systems, owing to their high efficiency and zero emissions.

Central to every PEMFC is the membrane electrode assembly (MEA), a layered composite containing the proton-conducting membrane, catalyst layers, and gas diffusion layers. The performance and durability of PEMFCs depend critically on the interplay between these components [1]. However, the inherent complexity of real MEAs – where even minor modifications to one component can substantially affect others – makes rational design highly challenging.

Although MEA components have traditionally been optimized individually, such approaches can overlook vital cross-component synergies or unintended detrimental side effects [2]. Furthermore, many existing models often use oversimplified MEA geometries, resulting in predictions that frequently fail to reflect actual device performance.

A central challenge, therefore, is to develop experimental and modelling techniques capable of decoupling the complex relationships between MEA materials and fuel cell behaviour, thereby guiding the design of robust, high-performance devices.

To address these challenges, a combination of high-resolution structural imaging (Figure 1a) and multiphase, multi-physics simulations (Figure 1b) was employed. Three-dimensional reconstructions of MEAs were obtained using micro X-ray computed tomography (micro-CT) at beamline ID19 (via rapid access mode with a sample-changer robot) and focused ion beam-scanning electron microscopy (FIB-SEM).

Click figure to enlarge

Fig. 1: Protocol for 3D modelling and multi-physics-coupled visualization simulation. a) Micro X-ray computed tomography-based segmentation, pore network generation, and permeability analysis. (i) Representative slice showing segmentation of membrane electrode assembly (MEA) components. (ii) Volume rendering and 3D segmentation of the MEA structure. (iii) Pore network extraction and permeability simulation of the gas diffusion layer (GDL). (iv) Pore network and permeability simulation of the catalyst layer (CL) and microporous layer based on focused ion beam-scanning electron microscopy (FIB-SEM) data. b) Electrochemistry-based multi-physics simulation. (i) 3D model of polyvinylpyrrolidone-polyethersulfone membrane-based MEA and meshing. (ii) Simulated electrode potential relative to ground, with arrows indicating electrode current density vector. (iii) Electrolyte potential distribution, with arrows representing electrolyte current density vector. (iv) Total flux streamline of O2. (v) Surface distribution of water mole fraction.

The high-energy X-rays at beamline ID19 enabled visualization of the bulk MEA structure, as well as clear differentiation of the phosphoric acid (PA)-doped membrane from the catalyst and support layers at sub-micrometre resolution. FIB-SEM further resolved nanometre-scale features within the catalyst and microporous layers.

These 3D datasets formed the basis for detailed computational models, enabling the simulation of coupled physical and electrochemical processes, including gas and proton transport, acid migration, and electrochemical reactions.

Crucially, this approach allowed specific structural features – such as membrane pores, catalyst cracks, and material migration – to be modified in silico, thereby facilitating the construction of decoupled models to isolate the influence of individual variables.

This methodology revealed several key insights into the relationship between MEA morphology and PEMFC performance. Firstly, morphological features commonly classified as defects – such as cracks, pores, and material migration – are not always detrimental. For example, small membrane pores, usually considered undesirable, can regulate local humidity and enhance proton transport, improving performance under certain operating conditions. Similarly, cracks in the catalyst layer can promote local turbulence and increase the electrochemically active interface, although excessive cracking impairs overall conductivity.

Secondly, multi-physics simulations based on realistic 3D structures enabled controlled variable studies (Figure 2), which demonstrated that the impact of a given defect is context-dependent. For example, cracks may be either beneficial or harmful, depending on membrane flatness, and the interplay among membrane thinning, PA leaching, and catalyst migration governs both short-term performance and long-term degradation.

Click figure to enlarge

Fig. 2: Protocol for decoupled and recoupled modelling with controlled variables via image-based segmentation. Schematic representation of the workflow used to construct decoupled and recoupled membrane electrode assembly (MEA) models by selectively modifying structural features derived from 3D imaging data. Image segmentation enables the isolation of individual components or morphological features, which are then computationally altered or reintroduced to investigate their independent and combined effects on performance through multi-physics simulation. CL: catalyst layer, GDL: gas diffusion layer.

Moreover, the selective coupling and subsequent recoupling of features in the model revealed that certain morphological characteristics only become detrimental when combined with others, highlighting the need to treat MEAs as integrated systems.

This study demonstrates that combining advanced synchrotron imaging with physics-based simulations is an effective way to understand how MEA morphology influences PEMFC performance.

By enabling specific structural variables to be isolated within realistic 3D microstructures, this method goes beyond traditional trial-and-error approaches and offers a more systematic path for designing durable, high-efficiency fuel cells. The insights gained lay the foundation for next-generation MEAs tailored to specific applications. Moreover, the methodology can be applied to other electrochemical energy systems.

Future work will extend this strategy to a broader range of materials and operating conditions, further accelerating the development of sustainable energy technologies.

Principal publication and authors
Decoupling Membrane Electrode Assembly Materials Complexity from Fuel Cell Performance through Image-Based Multiphase and Multiphysics Modelling, J. Chen (a,b), W. Du (c,d), Z. Guo (b,e), X. Lu (c,f), M.P. Tudball (a), X. Yang (b), Z. Zhou (b), S. Zhou (a), A. Rack (g), B. Lukic (g), P. R. Shearing (c,d), S. J. Haigh (h), S.M. Holmes (b), T.S. Miller (a), Adv. Energy Mater. 2405179 (2025); 
https://doi.org/10.1002/aenm.202405179
(a) Electrochemical Innovation Lab, Department of Chemical Engineering, University College London, London (UK)
(b) Department of Chemical Engineering, University of Manchester, Manchester (UK)
(c) The Faraday Institution, Harwell Science and Innovation Campus, Didcot (UK)
(d) Department of Engineering Science, University of Oxford, Oxford (UK)
(e) Department of Mechanical and Industrial Engineering, University of Toronto, Toronto (Canada)
(f) School of Engineering and Materials Science, Queen Mary University of London, London (UK)
(g) ESRF
(h) Department of Materials, University of Manchester, Manchester (UK)

References
[1] J. Chen et al., Chem. Eng. J. 487, 150670 (2024).
[2] J. Chen et al., Nat. Commun. 15, 1-18 (2024).