Graduate Student Award of the Energy Technology Division - Electrochemical Award

In both automotive and stationary applications, proton-exchange-membrane fuel cells (PEMFCs) must permit rapid startup with minimal energy from subfreezing temperatures, known as cold-start. Under subfreezing conditions, water solidifies to form ice in the membrane-electrode assembly (MEA), severely inhibiting cell performance and often causing cell failure. By way of example¹, Figure 1 displays typical evolutions of MEA cell voltage (filled symbols) during isothermal cold-start from temperatures of –6 and –20 °C at a current density of 20 mA/cm² (open symbols). Following an initial current ramp of 0.4 mA/cm²/s, cell voltage remains constant until failure (i.e., when cell voltage rapidly declines to 0 mV).  Note the short MEA cell-failure time,  t_fail, at –20 °C (i.e., 10 min). Clearly, elucidation of the mechanisms and kinetics of ice formation within PEMFC-porous media is critical to improving PEMFC cold-start capability.

          We measure and predict ice-crystallization kinetics within PEMFC gas-diffusion layers (GDLs) and catalyst layers (CLs).^1-3 To validate ice-crystallization kinetics within PEMFCs, we further measure and predict cell-failure time during isothermal galvanostatic cold-start. Using a simplified PEMFC isothermal cold-start continuum model, cell-failure times are predicted using the newly obtained rate expression and compared to those using a traditional thermodynamic-based approach.¹ From this comparison, we identify conditions under which including ice-crystallization kinetics is critical and elucidate the impact of freezing kinetics on low-temperature PEMFC operation. 

          As an example, Figure 2 plots isothermal t_fail at a current density of 20 mA/cm² as a function subcooling, ΔT, where ΔT is defined as the magnitude of the difference in temperature and 273 K. As ΔT increases,   t_fail decreases substantially. In Figure 2, t_fail decreases from 15.5 and 33 h to 0.19 and 0.2 h for an increase in ΔT  from 5 to 10 K, respectively. Solid lines in Figure 2 represent ice-crystallization kinetics for two cCL carbon-support materials with considerably different ice-crystallization kinetics, whereas the dashed line corresponds to a traditional thermodynamic-based approach.¹ In all cases, t_fail decreases substantially with increasing ΔT, in good agreement with experiment. As subcooling extends beyond 11 K, ice-nucleation times in both the cCL and cGDL are negligible, and ice-crystallization kinetics is well approximated by the thermodynamic-based approach. Therefore, we conclude that including ice-crystallization kinetics is critical in the “nucleation-limited” regime (see Figure 14 of           Dursch et al.²) where ice-nucleation times are long. However, the particular ΔT that establishes the “nucleation-limited” regime relies heavily on all heat transfer and kinetic parameters.¹ Consequently, these controlling parameters can be adjusted to lengthen ice-nucleation times, significantly delaying or even preventing ice formation.^2,3

Acknowledgement

This work was funded by the Assistant Secretary for Energy Efficiency and Renewable Energy, Office of Fuel Cell Technologies, of the U. S. Department of Energy under contract number DE-AC02-05CH11231.