High Energy Density Proton Exchange Membrane Fuel Cell With Dry Reactant Gases

Supramaniam Srinivasan, Serguey Gamburzev, Omouriag A. Velcv, Frank Simoncaux, and A. John Appleby

Center for Electrochemical Systems and Hydrogen Research Texas Engineering Experimental Station

Texas A&M University System College Station, TX 77843-3402, USA

Introduction Proton exchange membrane fuel cells (PEMFC) require careful control of humidity levels in the cell slack to achieve a high and stable level of performance. External humidification of the rcactant gases, as in the state-of-the-art PEMFCs, increases the complexity, the weight, and the volume of the fuel ccll power plant. A method for the operation of PEMFCs without external humidification (i.e., self-humidified PEMFCs) was first developed and tested by Dhar at BCS Technology (1)'. A project is underway in our Center to develop a PEMFC cell stack, which can work without external humidification and attain a performance level of a current density of 0.7 A/cm2 at a cell potential of 0.7 V, with hydrogen/air as reactants at 1 atm pressure. In this paper, the results of our efforts to design and develop a PEMFC stack requiring no external humidification will be presented. This paper focuses on determining the effects of type of elcctrodcs, the methods of their preparation, as well as that of the membrane and electrode assembly (MEA), platinum loading and types of electrocatalyst on the performance of the PEMFC will be illustrated.

Optimization of Structures of Electrodes The traditional method of electrode preparation was significantly modified in order to retain the maximum amount of water in the active layer and hence to maintain the high ionic conductivity in the proton exchange membrane and the active layer in the clcctrode containing only the supported electrocatalyst and Nafion. The electrodes were fabricated by spreading an emulsion of platinum supported catalyst and Nafion 950EW onto teflonizcd substrate (carbon cloth)/diffusion layer (2, 3, 4). By this method of preparation pcrfluorosulfonic acid polymer was evenly distributed in the active layer of the electrode and thus provided the largest possible-area of the elcctrocatalyst/electrolyte interface. Attaining this high interfacial area was confirmed by the very high electrochcmical utilization of the platinum electrocatalyst, as determined using cyclic voltammctry. For the 20% Pt supported on Vulcan XC-72, this value was 48% (vs 10 - 20% obtained with the standard impregnation method). Another significant difference in the composition of the active layer was the elimination of Teflon which is a strong hydrophobic agent. In this way a better water balance was possible, as required for preventing drying-out of the membrane.

Experimental Procedure Experiments were carried out in 50 cm2 PEMFC single cells at different temperatures and atmospheric pressure with H/air as reactants. The fuel cell electrodes were prepared in-house with Pt supported on Vulcan XC-72 carbon clectrocatalysts (E-TEK Inc.). The Pt loading in the cathode was between 1.2 and 1.5 mg Pt/cm2, and that in the anode between 0.3 and 0.4 mg Pt/cm2. The proton exchange membranes, Nafion® 112 and GORE-SELECT™, were used as received. The rcactant utilization was 95 % for hydrogen and 50 % for oxygen at all currcnt densities.

PEMFC Performance with Dry Reactant Gases The data presented in Figure 1 illustrates the PEMFC (with a Nation 112 membrane) performance with dry reactant gases. For comparison the performance of the PEMFC with 100% humidification of air and hydrogen is also presented. At low current densities there is hardly any difference in performance but at high current densities a lower hydration of the membrane is reflected in the ohmic controlled region of the cell potential vs. current density plot. In the PEMFC with the GORE-SELECT™ membrane, the difference in the performances for operation without and with external humidifications was considerably less than in the PEMFC with the Nafion membrane.

Figure 1 Cell potential vs. current density plots for cells with different membranes. 50 cm2 ccll, hydrogen/air rcactants at atmospheric pressure, 50 °C. Anode elcctrocatalyst loading 0.4 mg/cm\ cathode elcctrocatalyst loading 1.4 mg/cm2. Reactant gas utilization - 95% for hydrogen and 50% for oxygen.

Current density, mA/cm2

Figure 1 Cell potential vs. current density plots for cells with different membranes. 50 cm2 ccll, hydrogen/air rcactants at atmospheric pressure, 50 °C. Anode elcctrocatalyst loading 0.4 mg/cm\ cathode elcctrocatalyst loading 1.4 mg/cm2. Reactant gas utilization - 95% for hydrogen and 50% for oxygen.

The dependence of the time on current density at a constant potential for the PEMFC operating with dry gases is shown in Fig. 2. Its performance was stable over a 200 h period. The current density at this potential obtained from the dynamically measured cell potential vs current density plot (Fig. 1) shows good agreement with its value, as obtained from the time study (Fig. 2). Such a result was obtained only when the operating temperature of the PEMFC was below 60* C. When the temperature was raised above this value, the PEMFC performance dropped drastically due to dehydration of the proton exchange membrane.

PEMFC Performance with only Hi Gas Humidification If the restrictions on the operating conditions imposed by the low relative humidity of the reactant gases arc not acceptably a compromise may be rcachcd by providing extra water only to the hydrogen side of the PEMFC. The data in Figure 3 shows that the PEMFC performance in this case is practically the same as in the PEMFC when both hydrogen and oxygen arc humidified. Another advantage of this type of humidification is that it bccomcs easier to remove the product water as a vapor, which thereby reduces the amount of external cooling required to remove the heat generated in the PEMFC.

0 0

Post a comment