Correlation between Local Temperature and Degradations in Polymer Electrolyte Membrane Fuel Cells

In Proton Exchange Membrane Fuel Cells, local temperature is a driving force for many degradation mechanisms such as hydrothermal deformation, creep, platinum dissolution, and bipolar plates corrosion [1, 2]. In order to investigate and quantify those effects, durability testing in automotive-related operating conditions is conducted in this work. In particular, two ageing tests have been conducted: a stationary test over 2000 h under nominal operating conditions and a load/thermal cycling test (1500 cycles) which corresponds to peaks in power demand.

During the ageing tests, punctual characterizations like polarization curves, Cyclic Voltammetry (CV), Linear Sweep Voltammetry (LSV), Electrochemical Impedance Spectroscopy (EIS) and Fluoride Release Rate (FRR) are performed in order to investigate respectively the overpotentials  (and global performance), the electrochemical surface area, the hydrogen crossover though the membrane, the transfer resistances and the membrane chemical degradation. Continuous in situ measurements of current density distribution and local temperature are made with a printed circuit board (Current Scan Lin S++ device [3]) inserted in the middle of 30 cells stacks.

At the end of life, post-mortem analysis of the aged components via optical microscopy, Scanning Electron Microscopy (SEM), Energy Dispersive X-ray Spectroscopy (EDX) and X-ray Photoelectron Spectroscopy (XPS) are conducted in order to investigate their microstructure and chemical composition. Infrared Imaging (IR) is used to study the local permeation in the aged membrane.

In order to study the correlation between the observed degradation and the local conditions inside the PEMFC (temperature, relative humidity, molar fraction of hydrogen and oxygen, water fluxes…), a pseudo-3D physic based model has been developed at the cell scale (220 cm²) and validated against the in situ experimental data of the printed circuit board. Given the aspect ratio of the different cell components, the idea of the pseudo-3D approach is to consider each component as a plane layer, which is coupled to the other components through appropriate exchange conditions. With this approach, it is thus possible to compute the 2D (in-plane) distribution of the different physical parameters (temperature, species concentration…) in each cell component (active layers, gas diffusion layers, membrane, bipolar plates, cooling circuit…) at a reasonable cost. The model takes into account the thermal effects, including the exchanges with the cooling circuit, species transport and the electrochemical behavior of the electrodes. The model is implemented in the software Comsol Multiphysics.

The simulation results highlight different zones where temperature and linked parameters, such as humidity, have extreme values causing notably lower local performance and increasing risks of water flooding. The post-mortem analysis mainly point out heterogeneous bipolar plates and membrane degradations that are different in the anode and cathode sides both at the cell and at the channel/rib scale but quite reproducible along the stack (See Figure 1). The correlation of the simulations with the experimental results highlight the effects of local temperature and gases humidity on in-situ performance losses and ex-situ observed degradations from the cell scale to the channel/rib scale. Furthermore, the impact of dynamic thermal cycles on degradation acceleration is quantified.

[1]   P. Pei et al. Main factors affecting the lifetime of Proton Exchange Membrane fuel cells in vehicle applications: A review, Applied Energy 125 (2014) 60–75

[2]   S.F. Burlatsky et al. A mathematical model for predicting the life of polymer electrolyte fuel cell membranes subjected to hydration cycling, Journal of Power Sources 215 (2012) 135 - 144

[3]   S++ Simulation Services, Waldstraße 5, 82418 Murnau-Westried, Germany.  www.splusplus.com