The hydrogen fuel cell is a promising technology that supports the development of sustainable energy systems and zero emission vehicles. One of the key technical challenges for the use of fuel cells in the transportation sector is the high durability requirements 1–3. One of the key components that control the overall life time of a hydrogen fuel cell is the ionomer membrane that conducts the protons and allows the separation between the anode and the cathode.
During fuel cell operation, the membrane is subjected to two categories of degradation: mechanical and chemical. These degradations lead to reduction in the performance, crossover of reactants between anode and cathode and ultimately total failure of the fuel cell. The mechanical degradation occurs when the membrane swells and shrinks under the variation of the local hydration level. This leads to fatigue of the ionomer structure and ultimately irreversible damage. However, under pure mechanical degradation the damage takes a very long time to occur 4,5. Sadeghi et al. 5 observed failure of the membrane after 20,000 of accelerated mechanical stress testing. This translates into a longer lifetime in comparison to what is observed in field operation 6.
The chemical degradation on the other hand is caused by the presence of harmful chemicals such as OH radicals that attack the side chains and the main chains of the ionomer 7,8. Such attacks weaken the structural integrity of the membrane and make it prone to severe mechanical damage. Hence understanding the effect of combining both categories of membrane degradation is the key to accurate prediction of the time to failure of the fuel cell.
In this work we propose a novel model that represents accurately the structural properties of the membrane and couples the chemical and the mechanical degradations to estimate when the ultimate failure is initiated. The model is based on a network of agglomerated fibrils corresponding to the basic building block of the membrane structure 9–11. The mechanical and chemical properties are defined for each fibril and probability functions are used to evaluate the likelihood of a fibril to break under certain operating conditions. The description of the fundamentals behind the approach will be presented. Two set of simulations will be presented and discussed. The first one corresponding to standard testing scenarios that were used to validate the model. The second set of results will highlight the impact of coupling both degradation mechanisms on the estimation of the failure initiation time. The main strengths of the model and the future development will be discussed as well.
T. Sinigaglia, F. Lewiski, M. E. Santos Martins, and J. C. Mairesse Siluk, Int. J. Hydrogen Energy, 42, 24597–24611 (2017).
T. Jahnke et al., J. Power Sources, 304, 207–233 (2016).
P. Ahmadi and E. Kjeang, Int. J. Energy Res., 714–727 (2016).
X. Huang et al., J. Polym. Sci. Part B Polym. Phys., 44, 2346–2357 (2006).
A. Sadeghi Alavijeh et al., J. Electrochem. Soc., 162, F1461–F1469 (2015).
N. Macauley et al., J. Power Sources, 336, 240–250 (2016).
K. H. Wong and E. Kjeang, J. Electrochem. Soc., 161, F823–F832 (2014).
K. H. Wong and E. Kjeang, ChemSusChem, 8, 1072–1082 (2015).
P.-É. A. Melchy and M. H. Eikerling, J. Phys. Condens. Matter, 27, 325103–6 (2015).
J. A. Elliott et al., Soft Matter, 7, 6820 (2011).
L. Rubatat, G. Gebel, and O. Diat, Macromolecules, 37, 7772–7783 (2004).
Electrochemical Performance Improvement of the Catalyst of the Methanol Micro-Fuel Cell Using Carbon Nanotubes