High current density operation is essential to cost reduction of polymer electrolyte fuel cell (PEFC) system for automotive application. Although extensive numerical^1,2) and experimental^3,4) approaches have been conducted in order to understand and to predict a cell performance under high current density condition, there was no comprehensive analysis with microscopic transport resistance on Pt surface, catalyst activity change　associate with Pt oxide formation, water and heat transport. In this study, polarization curves under high current density were validated with　quantification of both proper catalyst activity and mass transport effects. Obtained results suggested that catalyst activity with oxide formation is important for deeper understanding of over-potential under high current density condition.

Experimental and Numerical Modeling

Polarization curves were evaluated under extremely low oxygen concentration (0.4%) and normal condition (10%) in order to validate the effect of heat, water and proton transport respectively. Supplied gases were controlled at 200 kPa (Abs.) with high enough flow rate into low pressure drop straight flow channel to minimize oxygen and reaction distribution along flow direction. To evaluate the initial state dependence of adsorbed species on Pt surface, scanning rate of potential was controlled very low (0.5 mV· sec^-1) from low potential (0.3 V) to high potential (1.0 V) region. Pt loadings of test cell were 0.35 mg· cm^-2 and 0.10 mg· cm^-2to evaluate the effect of both mass transport resistance and activity change in catalyst associate with effective Pt surface area.

In the numerical analysis model, continuous equations, momentum, conservation formula of chemical species, and potential (electronic and electrolyte) formulas were combined, and the source and sink term associated with electrochemical reaction was given by the Butler Volmer equation. Transport characteristic values were measured by the evaluation method described in previous reports ^1,2) . Oxide formation on platinum surface was considered the active site change assuming Langmuir adsorption and adsorption energy change exerted by adsorbed species covering the site in the vicinity assuming Temkin adsorption^3,4), coupled analysis of electrochemical reactions and three-dimensional mass transport corresponding to activity changes due to potential changes. In addition, the oxide coverage ratio was experimentally approximated by the CV test. The numerical model also made it possible to simplify to one-dimensional analysis for studying three-dimensional mass transport effect on polarization curves.

Results and Discussions

Fig.1 shows results of experiment (dashed line) and numerical validation (solid line) with/without considering effect of Pt oxide formation and three-dimensional mass transport. One-dimensional models analysis indicated that Pt oxide formation doesn’t have effects on limiting current density but has effects on over-potential nonlinearly from the region below 0.75 V. It was roughly in agreement that oxide species on catalyst surface almost all desorbed based on the CV test result in previous report³⁾ and indicates that electrochemical reaction characteristics in low potential regions is also important for cell performance analysis. Furthermore, regarding the comparison of one-dimensional model and three-dimensional model considering Pt oxide, even if thin rib (0.2 mm) configuration is used under extremely low oxygen concentration operation, over-potential under high current density increased due to influences of three dimensional mass transport, electrochemical reaction distribution.

Conclusion

In this study, polarization curves under high current density condition were validated by the numerical analysis with both proper catalyst activity and three-dimensional mass transport effects. Validation results suggested over-potential under high current density operation is strongly affected by not only three-dimensional mass transport but also Pt oxide formation associated with potential change as well.

Reference

[1] Y. Tabuchi et al. / Electrochimica Acta., 56, 352 (2010)

[2] Y. Fukuyama et al. / Electrochimica Acta., 117, 367 (2014)

[3] S. Sugawara et al. / J. Eletrochemisty.,79, 404 (2011)

[4] N.M. Marković et al. / J. Eletroanal. Chem., 467, 157 (1999)