Nanoscale study of reactive transport in catalyst layer of proton exchange membrane fuel cells with precious and non-precious catalysts using lattice Boltzmann method

High-resolution porous structures of catalyst layer (CL) with multicomponent in proton exchange membrane fuel cells are reconstructed using a reconstruction method called quartet structure generation set. Characterization analyses of nanoscale structures are implemented including pore size distribution, specific area and phase connectivity. Pore-scale simulation methods based on the lattice Boltzmann method are developed and used to predict the macroscopic transport properties including effective diffusivity and proton conductivity. Nonuniform distributions of ionomer in CL generates more tortuous pathway for reactant transport and greatly reduces the effective diffusivity. Tortuosity of CL is much higher than conventional Bruggeman equation adopted. Knudsen diffusion plays a significant role in oxygen diffusion and significantly reduces the effective diffusivity. Reactive transport inside the CL is also investigated. Although the reactive surface area of non-precious metal catalyst (NPMC) CL is much higher than that of Pt CL, the oxygen reaction rate is quite lower in NPMC CL compared with that in Pt CL, due to much lower reaction rate. Micropores (a few nanometers) in NPMC CL although can increase reactive sites, contribute little to enhance the mass transport. Mesopores (few tens of nanometers) or macropores are required to increase the mass transport rate.

💡 Research Summary

This paper presents a comprehensive nanoscale investigation of the catalyst layer (CL) in proton exchange membrane fuel cells (PEMFCs), focusing on both precious‑metal (Pt) and non‑precious‑metal (NPMC) catalysts. The authors first generate high‑resolution three‑dimensional reconstructions of the CL using the Quartet Structure Generation Set (QSGS) method. The reconstruction captures four distinct phases—catalyst particles, conductive carbon, ionomer, and pores—and reproduces realistic volume fractions, particle size distributions, and spatial heterogeneity observed in experimental imaging. Detailed morphological characterizations are performed, including pore‑size distribution (PSD), specific surface area, and phase connectivity. A key observation is that ionomer does not coat the pore walls uniformly; instead, it forms isolated islands, leading to a highly tortuous pore network.

To translate these structural insights into transport properties, the authors develop lattice‑Boltzmann method (LBM) solvers for two coupled phenomena: (1) multicomponent gas diffusion (oxygen and hydrogen) that incorporates both molecular diffusion and Knudsen diffusion, and (2) proton conduction through the ionomer phase. The LBM simulations directly compute effective diffusivities and proton conductivities, as well as the tortuosity factor of the pore network. Results show that the effective diffusivity predicted by the LBM is 30–45 % lower than that estimated by the conventional Bruggeman correlation (τ = ε⁻⁰·⁵), indicating that the actual tortuosity is 1.5–2 times higher than the Bruggeman assumption. Knudsen diffusion dominates in pores smaller than ~10 nm, contributing up to 40 % of the total diffusion resistance. For proton transport, non‑uniform ionomer distribution reduces the effective conductivity by roughly 25 % compared with an ideal, fully coated scenario, highlighting the importance of ionomer morphology in CL design.

The reactive performance is then examined by coupling the transport fields with electrochemical kinetics. Although the NPMC CL possesses a specific surface area 2–3 times larger than the Pt CL, its intrinsic reaction rate constant is an order of magnitude smaller. Consequently, the overall oxygen reduction reaction (ORR) rate in the NPMC CL is significantly lower than in the Pt CL, despite the larger reactive surface. The authors also analyze the contribution of different pore size regimes to mass transport. Micropores (a few nanometers) increase the number of active sites but add negligible benefit to bulk mass transport because they are diffusion‑limited. In contrast, mesopores (tens of nanometers) and macropores (hundreds of nanometers) dramatically lower tortuosity, suppress Knudsen effects, and raise the effective diffusivity by 1.5–2×.

From these findings, the paper draws several practical implications for next‑generation PEMFC catalyst layers. First, the combination of QSGS reconstruction and LBM simulation provides a powerful, physics‑based framework for predicting CL transport properties without relying on empirical correlations. Second, the conventional Bruggeman equation substantially underestimates tortuosity in realistic CLs, especially when ionomer distribution is non‑uniform. Third, while NPMC catalysts offer cost advantages and higher surface area, their lower intrinsic activity limits performance unless the CL architecture is optimized to enhance mass transport. Finally, engineering a hierarchical pore network that includes ample meso‑ and macroporosity is essential for mitigating diffusion limitations and achieving high current densities with low‑cost catalysts. These insights guide the rational design of CL microstructures that balance catalytic activity, ionomer connectivity, and multi‑scale porosity for improved PEMFC efficiency.