Insights into the Pore Structure Effect on the Mass Transfer of Fuel Cell Catalyst Layer Via Combining Machine Learning and Multiphysics Simulation
Laiming Luo,Xinrui Liu,Jujia Zhang,Jian Yao,Baohua Liu,Jin Zhang,Haining Wang,Shanfu Lu,Yan Xiang
DOI: https://doi.org/10.1016/j.ces.2024.120830
IF: 4.7
2025-01-01
Chemical Engineering Science
Abstract:Analysis of the catalytic layer (CL) pore structure and three-phase (electron phase, proton phase, and reactant phase) mass transfer in high-temperature polymer electrolyte membrane fuel cell (HT-PEMFC) is critical for improving concentration polarization and optimizing performance. However, there is a lack of studies on the pore structure's impact on the catalytic layer three-phase mass transfer. In this work, the influence of pore structure on catalytic layer three-phase mass transfer and multiphysics distribution was investigated by combining machine learning and multiphysics simulation. Among many key factors investigated by machine learning that impact the performance of catalytic layer, porosity emerges as the primary influencing factor. Porosity accounts for 32.7 % of the electric potential drop, with macro/micro porosity contributing 27.6 % and 5.1 %, and 32.8 % of the current density, with macro/micro porosity contributing 15.6 % and 17.2 %. Furthermore, the macropore structure has a much greater influence on the performance of the catalytic layer compared to the micropore structure. Moreover, the simulation results demonstrated that porosity has a great influence on three-phase mass transfer, the oxygen molarity increases with increasing porosity. The electric potential difference increases exponentially as porosity increases. The relationship between current density and porosity is a volcanic relationship, which can be expressed as a quadratic polynomial function. The mesoscopic pore-scale model (PSM) with an epsilon eff of 0.502 (epsilon mic of 0.2 and epsilon mac of 0.378) exhibits the highest mean current density at 619.526 mA/cm2@0.5 V, attributed to a more uniform and efficient three-phase transfer path. The accuracy of the mesoscopic pore-scale model and the polynomial equation is further confirmed through corresponding experiments. The experiment results show that the current density decreased as the epsilon eff increased from 0.682 to 0.702, which aligns with the theoretical predictions of the PSM model and polynomial equation. These studies are valuable for comprehending the internal three-phase mass transfer behavior and optimizing the catalytic layer pore structure to enhance HT-PEMFC performance.