Microstructure-Based Modeling of Aging Mechanisms in Catalyst

ARTICLE pubs.acs.org/JPCB

Microstructure-Based Modeling of Aging Mechanisms in Catalyst Layers of Polymer Electrolyte Fuel Cells Kourosh Malek*,†,‡ and Alejandro A. Franco*,§ †

Institute for Fuel Cell Innovation, National Research Council of Canada, 4250 Wesbrook Mall, Vancouver, BC, Canada V6T 1W5 Department of Chemistry, Simon Fraser University, Burnaby, BC, Canada V5A 1S6 § DRT/LITEN/DEHT/Laboratory of Components for Fuel Cells and Electrolysers, and of Modeling (LCPEM), Atomic and Alternative Energies Commission of France (CEA), 17 rue des Martyrs, 38054 Grenoble Cedex 9, France ‡

ABSTRACT: This work is comprised of a versatile multiscale modeling of carbon corrosion processes in catalyst layers (CLs) of polymer electrolyte fuel cells (PEFCs). Slow rates of electrocatalytic processes in CLs and materials aging are the main sources of voltage loss in PEFCs under realistic operating conditions. We combined microstructure data obtained from coarse-grained molecular dynamics (CGMD) simulations with a detailed description of the nanoscale elementary kinetic processes and electrochemical double-layer effects at the catalyst/electrolyte and carbon/electrolyte interfaces. We exclusively focused on morphology and microstructure changes in the catalyst layer of PEFCs as a result of carbon corrosion. By employing extensive CGMD simulations, we analyzed the microstructure of CLs as a function of carbon loss and in view of ionomer and water morphology, water and ionomer coverage, and overall changes in carbon surface. These ingredients are integrated into a kinetic model, which allows capture of the impact of the structural changes on the PEFC performance decay. In principle, such multiscale simulation studies allow a relation of the aging of CLs to the selection of carbon particles (sizes and wettability), the catalyst loading, and the level of ionomer structural changes during the CL degradation process.

I. INTRODUCTION The performance of polymer electrolyte fuel cells (PEFCs) is strongly dependent on the complex processes that occur in the membrane electrodes assembly (MEA). Durability of state-ofthe-art MEA nanomaterials is one of the main shortcomings limiting the large-scale development and commercialization of PEFCs. For automotive-operating conditions, MEA durability rarely exceeds 1000 h, but at least 3000
5000 h are demanded from industry for widespread commercialization.1 During the MEA operation, the nanoproperties of the Pt or Pt alloy catalyst, the CL microstructure, and the physicochemical properties of the polymer electrolyte membrane (PEM) evolve. These spatiotemporal nanostructural changes translate into long-term cell degradation. Among the components of MEA, the catalyst layers (CLs), in particular, cathode CLs (CCLs) are the most complex and critical ones. Conventional CCLs are random heterogeneous media that consist of a solid phase comprised of carbon particles or agglomerates decorated with catalyst nanoparticles (typically Pt-based) for conducting electrons and catalyzing reactions, a proton-conducting network of Nafion ionomer, and a particular water-filled porous network for gas transport.2,3 During preparation in an organic solvent, Pt/C particles and ionomer molecules self-organize into agglomerated structures with internal porosities that depend on the composition of the ink.4,5 Published 2011 by the American Chemical Society

Despite significant improvement in microstructure characterization techniques such as high-resolution transmission electron microscopy (HRTEM), the interplay between phase-segregated morphology and transport properties of CLs is not well understood.6 In this context, a significant number of computational approaches have been employed to complement experimental observations. An appropriate model should be able to predict microstructure formation processes at the nano-to-meso level while still being able to capture the morphology at long time and length scales.7
9 Recently, Malek et al. applied the coarsegrained molecular dynamics (CGMD) technique to consolidate the main features of microstructure formation in CLs of PEFCs.10
12 For a highly hydrophobic model of the carbon surface, they showed that the final microstructure depends on carbon particle choices and ionomer
carbon interactions. While ionomer side chains are contained in hydrophilic domains with a weak contact to carbon domains, the ionomer backbones are attached to the surface of carbon agglomerates. The correlation between hydrophilic species and the ionomer is significantly stronger than that between those species and carbon particles. A layer of ionomer of ∼4
10 nm thickness with well-packed Received: December 1, 2010 Revised: April 3, 2011 Published: June 07, 2011 8088

dx.doi.org/10.1021/jp111400k | J. Phys. Chem. B 2011, 115, 8088–8101

The Journal of Physical Chemistry B morphology around Pt/C clusters was evident in the presence of polar solvents. The compactness of carbon agglomerates was found to strongly depend on the polarity of the solvent.13 Electrochemical degradation processes of the CCLs include Pt catalyst oxidation, Pt dissolution and ripening, ionomer aging, and catalyst
support corrosion. Carbon black (CB) catalyst
support is thermodynamically unstable at typical cathode operating conditions (potential > 0.2 V versus NHE at 25 °C).14,15 The carbon corrosion reaction is quite slow at common PEFC operating temperatures (250 ns after equilibrating the microstructures of the structure at a given carbon loss percentage. The color code is same as that in Figure 2.

Figure 8. Simulated long-time behavior of the anode, cathode, and cell potentials for a static nominal current density at 0.1 A 3 cm
2.

Figure 6. Calculated water coverage (including hydronium ions) on Pt as a function of carbon losses.

attract most of the water, while a relatively constant amount of water is adsorbed at the ionomer surface at different carbon losses. By increasing the carbon loss, more Pt particles are exposed to water, and the water coverage increases as a function of carbon loss, as depicted in Figure 6. Ionomer coverage, however, is less affected by carbon loss (Figure 7), as a result of strong Pt
ionomer interactions in our simulations and more preserved structure of the ionomer network on the Pt surface. After reaching the optimum point at 40% carbon loss, most of the carbon surface is covered by ionomer as a result of ionomer rearrangement, and therefore, the overall water coverage decreases. The situations depicted in Figure 5b correspond to rather dominant scenarios of wetting properties and water distribution, which will have a significant impact on electrochemical and transport properties. In the case of a predominantly hydrophilic Pt/C surface, a water film is expected to form between Pt/C and the ionomer film. This water film will facilitate uniform wetting of Pt; it will accommodate a high concentration and mobility of protons in the confined water layer, which will enable high proton

conductivities and increase the turnover rate of the strongly pHsensitive carbon corrosion kinetics. On the other hand, for a predominantly hydrophobic Pt/C surface, hydrophobic ionomer backbones will be preferably oriented toward the Pt/C surface, resulting in a small active fraction of Pt due to poor wetting and in low rates of proton transport and ORR due to the inhibited access of protons to the region between Pt/C and the ionomer. C. Kinetic Results. Figure 8 shows the simulated long-term behavior of the anode, cathode, and cell potentials within MEMEPhys for a static nominal current. First, we can note that the cathode and cell potential decrease with time because of the carbon corrosion-driven degradation mechanism. When the cathode carbon is degraded, the cathode carbon electronic conductivity and platinum active surface area decrease as a function of time. The latter is mainly due to the catalyst nanoparticle reorganization as a result of carbon support corrosion. In fact, carbon corrosion induces the detachment of the nanoparticles from the carbon support. Notice that the CGMD simulations do not capture the nanoparticles coarsening, although Pt coarsening process has already been observed experimentally.1 Figure 8 also highlights that at a given simulated operation time (∼1000 h), the simulated cathode potential decreases dramatically. This potential “failure” or “collapse” occurs when the cathode platinum active surface area becomes too small to ensure the externally demanded current Iext. This phenomenon has already been observed experimentally, albeit without augmentation of PEM crossover.36,38,47 8096

dx.doi.org/10.1021/jp111400k |J. Phys. Chem. B 2011, 115, 8088–8101

The Journal of Physical Chemistry B

Figure 9. The impact of the nominal current on the cumulative carbon mass loss [%ΔmC = mC loss (%)] at different various static nominal current densities. The inset illustrates the short-term behavior.

Figure 10. Calculated electronic conductivity loss (g/g0) at (a) long and (b) short times at various static nominal currents.

Figure 9 shows the impact of the nominal current on the cumulative carbon mass loss. In agreement with the experimental data, the corrosion rate increases as the current decreases (where the Butler
Volmer potential increases).20 This behavior induces the trends in Figures 10 and 11 for the electronic conductivity loss and the catalyst active surface area loss, which are responsible for the potential decay in Figure 8. An interesting feature can also be pointed out from Figure 11; at short-term operation of less than 20 h, the catalyst surface area increases over time instead of decreasing. This is indeed a fingerprint based on the fact that at the beginning of the carbon corrosion process, Pt catalyst in smaller pores becomes exposed to water and ionomer. In this manner, access to protons becomes easier, and therefore, performance is increased at very

ARTICLE

Figure 11. Calculated catalyst active surface area loss at various static nominal currents. The color code is identical to that in Figure 10.

short term. This effect is only slightly visible in Figure 8. This interesting result can be validated by ultrasensitive timeresolved experiments. Figure 12 shows the associated Pt ad-species (OH, OOH, H2O) and evolution of adsorbed ionomer coverage during the carbon corrosion process for two current densities, 0 (OCV condition) and 0.75 A 3 cm
2. This coverage evolution is a consequence of the loss of catalyst surface area induced by carbon corrosion. The main feature is that the OH coverage increases with time; that is, the catalyst activity diminishes. The ionomer coverage is higher at the higher current densities, which agrees with the recent experimental data reported by Markovic et al.13 Figure 13 shows the corresponding coverage of carbon corrosion ad-species (OH, OOH, H2O) and ionomer coverage evolution. At long-term operation, the ionomer coverage becomes higher than the water coverage, and the difference between water and ionomer coverage decreases as the current increases (as does the degradation rate and electrode reorganization). Furthermore, the carbon oxide ad-layer increases as the current increases, implicating an increase in stability of the carbon at high currents (“passivation” effect). In Figure 14a, we report the calculated MEA durability as a function of the nominal current density for the full model (where ionomer coverage from CGMD simulations is included) and for C a model where θPt ionomer = 0 and θionomer = 0. The durability is expressed in terms of time when the potential collapse occurs. We emphasize that the ionomer adsorption on both Pt and C enhances MEA durability at a given current density. This can be understood from the associated cumulative carbon mass loss and Pt active surface area loss at 0.5 A 3 cm
2 provided in Figure 14b; the adsorbed ionomer shields the carbon from being exposed to water and thus mitigates the carbon corrosion reaction and the associated catalytic activity loss. This mitigation effect, captured here for Nafion, strongly depends on the chemical nature of the ionomer. In Figure 15, we compare the cumulative carbon mass loss over time for the full model at open circuit voltage (OCV) conditions with estimated ionomer coverage from CGMD simulations on both Pt and carbon and without ionomer coverage on Pt (θPt ionomer = 0). It is evident that the ionomer adsorption on Pt has an impact on the carbon corrosion kinetics. In fact, ionomer adsorption on Pt slows down the rate of carbon corrosion. The latter effect is explained by the electron pumping mechanism previously proposed by Franco et al.20 The latter is in agreement with experimental data, where Pt catalyzes the carbon 8097

dx.doi.org/10.1021/jp111400k |J. Phys. Chem. B 2011, 115, 8088–8101

The Journal of Physical Chemistry B

ARTICLE

Figure 12. Coverage of Pt ad-species (OH, OOH, H2O) and evolution of adsorbed ionomer coverage during the carbon corrosion process for two current densities, [0 (OCV) and 0.75 A 3 cm
2] at (a) long, (b) intermediate, and (c) short time frames.

degradation with an effectiveness increasing as the ORR kinetics (consuming electrons) increases on Pt.45,47

IV. CONCLUSIONS Here, we focused on morphology and microstructure changes in the CL of PEFCs as a result of carbon corrosion. By employing extensive CGMD simulations, we initially

analyze the microstructure of CLs as a function of carbon loss and in view of ionomer and water morphology, water and ionomer coverage, and overall changes in carbon surface area. By estimating water and ionomer coverage on Pt/C during the carbon corrosion process, our simulations unravel the effects of ionomer phase evolution as a major consequence of carbon loss in CLs. Carbon corrosion and ionomer reconfiguration influence Pt degradation and Pt
C and Pt
ionomer 8098

dx.doi.org/10.1021/jp111400k |J. Phys. Chem. B 2011, 115, 8088–8101

The Journal of Physical Chemistry B

ARTICLE

Figure 14. (a) Calculated MEA durability as a function of the nominal current density for the full model (where the ionomer coverage from CGMD simulations is included) and for a model where θPt ionomer = 0 and θCionomer = 0. (b) Cumulative carbon mass loss and Pt active surface area loss at 0.5 A 3 cm
2. [%ΔmC = mC loss (%)].

Figure 13. Coverage of carbon ad-species (OH, OOH) and evolution of adsorbed ionomer coverage during the carbon corrosion process for two current densities, [0 (OCV) and 0.75 A 3 cm
2] at (a) long, (b) intermediate, and (c) short time frames.

interactions at various I/C ratios. These structural data have been used to simulate performance decay based on the multiscale kinetic model. This multiscale/multiphysics model allowed description of the nanomaterials' degradation in in situ PEMFC conditions, based on an ab-initio elementary kinetics framework and through analysis of competitive aging mechanisms as well as MEA durability prediction. We initially focused on catalyst
support corrosion under anode fuel-starvation conditions. On the basis of these multiscale simulations, the impact of reorganization of Nafion, Pt nanoparticles, and the local water content on carbon corrosion rate was studied. In particular, we found that Nafion adsorbs on both Pt and C materials, reducing the specific ORR activity of Pt (even if OH and O2H adsorbate coverages appear to be reduced compared to a Nafion-free Pt surface) and decreasing the carbon corrosion kinetics (decrease of quinone and lactone products). Furthermore, the decrease of the ORR activity contributes to mitigating the carbon corrosion process (Pt “catalyzes” less the carbon oxidation process).

Figure 15. The cumulative carbon mass loss over time for the full model at OCV conditions with an estimated ionomer coverage from CGMD simulations on both Pt and carbon and without ionomer coverage on Pt (θPt ionomer = 0).

This work demonstrates that the ionomer structure has a major effect on the C corrosion process and thus on performance degradation. This modeling approach opens interesting perspectives in engineering practice to predict the MEA degradation and durability as a function of its initial nanomaterial composition and nano-to-micro structure and identifies new operation strategies to mitigate CL degradation.68 8099

dx.doi.org/10.1021/jp111400k |J. Phys. Chem. B 2011, 115, 8088–8101

The Journal of Physical Chemistry B Finally, such mesoscale simulation studies allow relating the aging of CLs to the selection of carbon particles (sizes and wettability), catalyst loading, and the level of ionomer structural changes during CL degradation

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected] (K.M.); alejandro.franco@ cea.fr (A.A.F.).

’ ACKNOWLEDGMENT The NRC
CEA collaboration program and the French Ministry of Foreign Affaires through the program EGIDE are gratefully acknowledged. We thank Dr. Michael Eikerling, Dr. Atsushi Ohma, and Mr. Tetsuya Mashio for useful discussions. ’ REFERENCES (1) Borup, R.; et al. Chem. Rev 2007, 107, 3904. (2) Vielstich, W., Lamm, A., Gasteiger, H., Eds. Handbook of fuel cells: fundamentals, technology, applications; VCH-Wiley: Weinheim, 2003. (3) Eikerling, M.; Kornyshev, A. A.; Kucernak, A. R. Phys.Today 2006, 59, 38. (4) Wilson, M. S.; Gottesfeld, S. J. Electrochem. Soc. 1992, 139, L28. (5) Uchida, M.; Aoyama, Y.; Eda, E.; Ohta, A. J. Electrochem. Soc. 1995, 142, 463. (6) Zhang, J.; Wu, J.; Shi, Z.; Holdcroft, S. IEEE Trans. NanoTechnol. 2005, 4, 616. (7) Eikerling, M.; Kornyshev, A. A.; Kulikovsky, A. A. Physical modeling of fuel cells and their components. In Encyclopedia of electrochemistry; Bard, A. J., Stratmann, M, Macdonald, D, Schmuki, P, Eds.; VCH-Wiley: Weinheim, 2007; Vol, 5, electrochemical engineering, pp 447
543. (8) Gottesfeld, S.; Zawodzinski, T. A. Polymer electrolyte fuel cells. In Advances in electrochemical science and engineering Alkire, R. C., Gerischer, H, Kolb, D. M., Tobias, C. W., Eds.; Wiley-VCH: Weinheim, 1997; Vol. 5, pp 195. (9) Eikerling, M.; Malek, K.; Wang, Q. Catalyst Layer Modeling: Structure, Properties, and Performance. In PEM Fuel Cell Electrocatalysts and Catalyst Layers; Zhang, J.J., Ed.; Springer: New York, 2008; p 381. (10) Malek, K.; Eikerling, M.; Wang, Q.; Navessin, T.; Liu, Z. J. Phys. Chem. C 2007, 111, 13627. (11) Malek, K.; Eikerling, M.; Wang, Q.; Liu, Z.; Otsuka, S.; Akizuki, K.; Abe, M. J. Chem. Phys. 2008, 129, 204702. (12) Malek, K.; Mashio, T.; Eikerling, M. Electrocatal 2011, 21, 41. (13) Subbaraman, R.; Strmcnik, D.; Stamenkovic, V.; Markovic, N. M. J. Phys. Chem. C 2010, 114 (18), 8414. (14) Ball, S. C.; Hudson, S. L.; Thompsett, D; Theobald, B. J. Power Sources 2007, 171 (1), 18. (15) Gallagher, K.; Fuller, T. Phys. Chem. Chem. Phys. 2009, 11, 11557. (16) Teranishi, K.; Kawata, K.; Tsushima, S.; Hirai, S. Electrochem. Solid-State Lett. 2006, 9 (10), A475. (17) Tang, H.; Qi, Z.; Ramani, M.; Elter, J. F. J. Power Sources 2006, 158, 1306. (18) Perry, M. L.; Patterson, T. W.; Reiser, C. ECS Trans. 2006, 3 (1), 783. (19) Li, J.; He, P.; Wang, K.; Davis, M.; Ye, S. ECS Trans. 2006, 3 (1), 743. (20) Franco, A. A.; Gerard, M. J. Electrochem. Soc. 2008, 155 (4), B367. (21) Shao, Y. Y.; Yin, G. P.; Gao, Y. Z.; Shi, P. F. J. Electrochem. Soc. 2006, 153, A1093.

ARTICLE

(22) Coloma, F.; Sepulveda Escribano, A.; Rodriguez Reinoso, F. J. Catal. 1995, 154, 299. (23) Reiser, C. A.; Bregoli, L.; Patterson, T. W.; Yi, J. S.; Deliang Yang, J.; Perry, M. L.; Jarvi, T. D. Electrochem. Solid-State Lett. 2005, 8, A273. (24) de Bruijn, F. A.; Dam, V. A. T.; Janssen, G. J. M. Fuel Cells 2008, 8, 3. (25) Fujii, Y.; Tsushima, S.; Teranishi, K.; Kawata, K.; Nanjo, T; Hirai, S. ECS Trans. 2006, 3, 735. (26) Baumgartner, W. R.; Walln€ofer, E.; Schaffer, T.; Besenhard, J. O.; Hacker, V.; Peinecke, V.; Prenninger, P. ECS Trans. 2006, 3, 811. (27) Maass, S.; Finsterwalder, F.; Frank, G.; Hartmann, R. J. Power Sources 2008, 176, 444. (28) Avasarala, B; Moore, R.; Haldar, P. Electrochim. Acta 2010, 55 (16), 4765. (29) Gallagher, K.; Fuller, T. Phys. Chem. Chem. Phys. 2009, 11, 11557. (30) Fujii, Y.; Tsushima, S.; Teranishi, K.; Kawata, K.; Nanjo, T; Hirai, S. ECS Trans. 2006, 3, 735. (31) Fuller, T. F.; Gray, G. ECS Trans. 2006, 1, 345. (32) Franco, A. A.; Schott, P.; Jallut, C.; Maschke, B. Fuel Cells 2007, 7 (2), 99. (33) Franco, A. A.; Tembely, M. J. Electrochem. Soc. 2007, 154 (7), B712. (34) Franco, A. A.; Schott, P.; Jallut, C.; Maschke, B. J. Electrochem. Soc. 2006, 153 (6), A1053. (35) Franco, A.A. PhD Thesis, Universite Claude Bernard Lyon-1, France; No. 2005LYO10239, 2005. (36) Franco, A.A. Habilitation Thesis, Universite Claude Bernard Lyon-1, France, 2010. (37) Cheah, S. K.; Lemaire, O.; Gelin, P.; Franco, A. A. ECS Trans. 2010, 25 (35), 275. (38) Franco, A. A.; Coulon, R.; de Morais, R. F.; Cheah, S.
K.; Kachmar, A.; Gabriel, M. A. ECS Trans 2009, 25, 65. (39) Ferreira de Morais, R.; Loffreda, D.; Sautet, P.; Franco, A. A. ECS Trans. 2010, 25 (23), 167. (40) Ferreira de Morais, R.; Loffreda, D.; Sautet, P.; Franco, A. A. Electrochim. Acta 2010Submitted. (41) Franco, A. A.; Passot, S.; Fugier, P.; Anglade, C.; Billy, E.; Guetaz, L.; Guillet, N.; De Vito, E.; Mailley, S. J. Electrochem. Soc. 2009, 156 (3), B410. (42) Gabriel, M.; Genovese, L.; Krosnicki, G.; Lemaire, O.; Deustch, T.; Franco, A. A. Phys. Chem. Chem. Phys. 2010, 12, 9406. (43) Franco, A. A.; Guinard, M.; Barthe, B.; Lemaire, O. Electrochim. Acta 2009, 54, 5267. (44) Coulon, R.; Bessler, W.; Franco, A. A. ECS Trans. 2010, 25 (35), 259. (45) Franco, A. A.; Passot, S.; Fugier, P.; Anglade, C.; Billy, E.; Guetaz, L.; Guillet, N.; De Vito, E.; Mailley, S. ECS Trans. 2008, 13 (17), 29. (46) Cheah, S.-K.; Lemaire, O.; Gelin, P.; Franco, A.A. ECS Trans 2010, 25 (35), 275. (47) Franco, A. A.; Gerard, M.; Guinard, M.; Barthe, B.; Lemaire, O. ECS Trans. 2008, 13 (15), 35. (48) Eikerling, M.; Malek, K. Electrochemical Materials for PEM Fuel Cells: Insights from Physical Theory and Simulation. In Modern Aspects of Electrochemistry, No. 43 ed. Schlesinger, M. Ed.; Springer: New York, 2009. (49) Wescott, J. T.; Qi, Y.; Subramanian, L.; Capehart, T. W. J. Chem. Phys. 2006, 124, 134702. (50) Lee, S. J.; Mukerjee, S.; McBreen, J.; Rho, Y. W.; Kho, Y. T.; Lee, T. H. Electrochim. Acta 1998, 43, 693. (51) Antolini, E.; Cardellini, F.; Giacometti, E.; Squadrito, G. J. Mater. Sci. 2002, 37, 133. (52) Antolini, E. J. Mater. Sci. 2003, 38, 2995. (53) Ferreira, P. J.; la O’, G. J.; Shao-Horn, Y.; Morgan, D.; Makharia, R.; Kocha, S.; Gasteiger, H. J. Electrochem. Soc. 2005, 152, A2256. (54) Izvekov, S.; Violi, A. J. Chem. Theory Comput. 2006, 2, 504. (55) Izvekov, S.; Violi, A. J. Phys. Chem. B 2005, 109, 2469. (56) Izvekov, S.; Violi, A.; Voth, G. A. J. Phys. Chem. B 2005, 109, 17019. 8100

dx.doi.org/10.1021/jp111400k |J. Phys. Chem. B 2011, 115, 8088–8101

The Journal of Physical Chemistry B

ARTICLE

(57) Markvoort, A. J. Coarse-Grained Molecular Dynamics. In Computational Methods in Catalysis and Materials Science; van Santan, R. A., Sautet, P., Eds.; Wiley-VCH: New York, 2010. (58) Lindahl, E.; Hess, B.; van der Spoel, D. J. Mol. Model. 2001, 7, 306. (59) Berendsen, H. J. C.; van der Spoel, D.; van Drunen, R. Comput. Phys. Commun. 1995, 9, 43. (60) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graph. 1996, 14, 33. (61) Franco, A.A. Habilitation (H.D.R.) Manuscript, Universite Claude Bernard Lyon-1, France, 2010. (62) Wang, L.; Roudgar, A.; Eikerling, M. J. Phys. Chem. C 2009, 113, 17989. (63) Franco, A. A.; Passot, S.; Fugier, P.; Anglade, C.; Billy, E.; Guetaz, L.; Guillet, N.; De Vito, E.; Mailley, S. ECS Trans. 2008, 13 (17), 29. (64) Franco, A.A.; Schott, P.; Jallut, C.; Maschke, B. In Proc. of the 5th MATHMOD Conference, Vienna, February 8
10, 2006; ISBN 3-90160825-7; p 93. (65) Takeuchi, N.; Fuller, T. J. Electrochem. Soc. 2010, 157 (1), B135. (66) Meyers, J.; Darling, R. J. Electrochem. Soc. 2006, 153, A1432. (67) Sahimi, M. Heterogeneous Materials, Part I and Part II; Springer: Heidelberg, Germany, 2003. (68) Franco, A.A. PEMFC degradation modeling and analysis. In Polymer electrolyte membrane and direct methanol fuel cell technology (PEMFCs and DMFCs) Vol. 1: Fundamentals and performance; Hartnig, C., Roth, C., Eds.; Woodhead: Cambridge, U.K., 2010.

8101

dx.doi.org/10.1021/jp111400k |J. Phys. Chem. B 2011, 115, 8088–8101