Performance Degradation of PEMFC by Sea Salt Aerosol Contamination

Thursday, 5 October 2017: 17:40
National Harbor 14 (Gaylord National Resort and Convention Center)
S. Uemura, M. Yamazki, T. Yoshida, T. C. Jao, and S. Hirai (Tokyo Institute of Technology)
In this study, performance degradation of proton exchange membrane fuel cells (PEMFCs) by 1 μm scale sea salt aerosol was investigated.

The regulations of NOx emission control on ships are strengthened year by year. Installation of PEMFCs as electrical power source is one of important choice. In order to investigate the influence of seawater contamination, several studies focus on the effects of sodium chloride (NaCl) [1, 2]. The fuel cell performance degradation by NaCl was investigated by injecting NaCl solution into the cathode side gas inlet, however, fuel cell has a filter and there is no such a situation. On the other hand, conventional filter cannot remove extreme small sea salt particles (1 μm scale sea salt aerosol). Effects of the invaded sea salt aerosol for the fuel cell performance were not elucidated. In this study, artificial sea salt aerosol generator was developed and the effects of sea salt aerosol contamination in the cathode side were investigated.

Figure 1 shows experimental setup. Sea salt aerosol generator was consisted by ultrasonic vibrator, NaCl solution tank, and heated pipe. Ultrasonic vibration generates fine droplets of NaCl solution. Water evaporates from the droplet while flowing through the heated pipe and finally, fine NaCl crystals remain in the gas as aerosol. The diameter of sea salt aerosol can be controlled by varying concentration of NaCl solution. Figure 2 shows SEM image of collected sea salt aerosols at the end of cathode gas supply line. With using 3 mol/L NaCl solution, approximately 1 μm scale sea salt particles were stably produced.

The cell has an active area of 4 cm2 (2 × 2 cm) with serpentine channels. The channel width and depth were 1.0 and 1.0 mm, respectively, and the rib-to-channel ratio was 1. Fuel cell temperature and relative humidity of anode/cathode gas were set 60°C and 70%, relatively. Flow rate of sea salt aerosol was 1.2 mg/h. Fuel cell was operated with constant current density condition (0.5 A/cm2).

Figure 3 shows time variation of cell voltage. When the sea salt aerosol injection was started, the cell voltage clearly dropped, and it suggested cell performance degradation.

The cell was disassembled after completing the power generation test. NaCl deposition was not observed in the channel. On the other hand, it was found that NaCl particles existed in the GDL by polarizing microscope imaging. A large number of NaCl particles were present under the channel, whereas almost no particles presented under the rib. The NaCl particles not only adhered to the channel side of the GDL but also invaded into the region close to the catalyst layer and the particle diameter has increased compare to initial one (Figure 4). Similar distribution was observed from inlet to outlet in the serpentine channels. However, NaCl particles did not block the pore in the GDL, and there were still enough gas flow path.

Figure 5 shows result of cyclic voltammetry (CV) measurement. Comparing the effective surface area of the platinum catalyst before and after adding the sea salt aerosol, decrease of 22.6% has been occurred. It is considered that poisoning of platinum catalyst was caused by adsorption of chloride ion (Cl-) [1].

These results suggest that 1 μm scale fine sea salt aerosol reaches catalyst layer in the cathode with passing the GDL easily, and finally, cell performance degradation will be caused. Present experiment has been performed under accelerated condition, investigation of degradation process with different flow rate of sea salt aerosol is future task. Investigation of performance recovery by generated liquid water is also important issues. There is a possibility that high humidity cell operation prevent the poisoning.


[1] M.S. Mikkaola et al., Fuel cells, 2007, 2, 153-158.

[2] T.J. Schmidt et al., Journal of Electroanalytical Chemistry, 2001, 508, 41-47.