Predicting the Topological and Transport Properties in Porous Transport Layers for Water Electrolyzers

The porous transport layer (PTL) in polymer electrolyte membrane (PEM) electrolyzers governs the overall efficiency. Its structural, thermal, and electronic properties determine how effortlessly the gases can be produced and can exit the PEM electrolyzer. In this study, we apply a stochastic reconstruction method for titanium felt-based PTLs to generate PTLs with different porosity, fiber radii, and anisotropy parameters. The morphology and topology of these PTLs are numerically characterized, and transport properties, such as gas diffusion coefficients and electrical and thermal conductivity, are computed via pore-scale modeling. Customized graded PTLs are proposed, exhibiting the optimal topology and bulk structure for the removal of gases, the conductance of electrons, and the transport of heat. The results indicate that the surface and transport properties of PTLs can be tailored by certain morphology parameters: PTLs with lower porosity and smaller fiber radii feature a more sufficient interfacial contact and superior electrical and thermal conductivity. Lowering the anisotropy parameters of PTLs results in a slight loss of interfacial contact but a substantial increase in the electrical and thermal conductivity in the through-plane direction. We outline that the design of PTLs should be differentiated depending on the operating conditions of electrolyzers. For nonstarvation conditions, PTLs should feature low porosity and small fiber radii, whereas for starvation conditions, PTLs should feature high porosity, low anisotropy parameters, and small fiber radii. Furthermore, graded PTLs with enhanced structural and transport properties can be developed by customizing the porosity, fiber radius, and fiber orientation.

Tailored Porous Transport Layers for Optimal Oxygen Transport in Water Electrolyzers: Combined Stochastic Reconstruction and Lattice Boltzmann Method

The porous transport layer (PTL) plays an integral role for the mass transport in polymer electrolyte membrane (PEM) electrolyzers. In this work, a stochastic reconstruction method of titanium felt-based PTLs is applied and combined with the Lattice Boltzmann method (LBM). The aim is to parametrically investigate the impact of different PTL structures on the transport of oxygen. The structural characteristics of a reconstructed PTL agree well with experimental investigations. Moreover, the impact of PTL porosity, fiber radius, and anisotropy parameter on the structural characteristics of PTLs are analyzed, and their impact on oxygen transport are elucidated by LBM. Eventually, a customized graded PTL is reconstructed, exhibiting almost optimal mass transport performance for the removal of oxygen. The results show that a higher porosity, larger fiber radius, and smaller anisotropy parameter facilitate the formation of oxygen propagation pathways. By tailoring the fiber characteristics and thus optimizing the PTLs, guidelines for the optimal design and manufacturing can be obtained for large-scale PTLs for electrolyzers.

Pore Network Simulation of Gas-Liquid Distribution in Porous Transport Layers

Pore network models are powerful tools to simulate invasion and transport processes in porous media. They are widely applied in the field of geology and the drying of porous media, and have recently also received attention in fuel cell applications. Here we want to describe and discuss how pore network models can be used as a prescriptive tool for future water electrolysis technologies. In detail, we suggest in a first approach a pore network model of drainage for the prediction of the oxygen and water invasion process inside the anodic porous transport layer at high current densities. We neglect wetting liquid films and show that, in this situation, numerous isolated liquid clusters develop when oxygen invades the pore network. In the simulation with narrow pore size distribution, the volumetric ratio of the liquid transporting clusters connected between the catalyst layer and the water supply channel is only around 3% of the total liquid volume contained inside the pore network at the moment when the water supply route through the pore network is interrupted; whereas around 40% of the volume is occupied by the continuous gas phase. The majority of liquid clusters are disconnected from the water supply routes through the pore network if liquid films along the walls of the porous transport layer are disregarded. Moreover, these clusters hinder the countercurrent oxygen transport. A higher ratio of liquid transporting clusters was obtained for greater pore size distribution. Based on the results of pore network drainage simulations, we sketch a new route for the extraction of transport parameters from Monte Carlo simulations, incorporating pore scale flow computations and Darcy flow.