DOE OSTI2011
The main objectives of this project were to investigate the effect of a series of potential impurities on fuel cell operation and on the particular components of the fuel cell MEA, to propose (where possible) mechanism(s) by which these impurities affected fuel cell performance, and to suggest strategies for minimizing these impurity effects. The negative effect on Pt/C was to decrease hydrogen surface coverage and hydrogen activation at fuel cell conditions. The negative effect on Nafion components was to decrease proton conductivity, primarily by replacing/reacting with the protons on the Bronsted acid sites of the Nafion. Even though already well known as fuel cell poisons, the effects of CO and NH3 were studied in great detail early on in the project in order to develop methodology for evaluating poisoning effects in general, to help establish reproducibility of results among a number of laboratories in the U.S. investigating impurity effects, and to help establish lower limit standards for impurities during hydrogen production for fuel cell utilization. New methodologies developed included (1) a means to measure hydrogen surface concentration on the Pt catalyst (HDSAP) before and after exposure to impurities, (2) a way to predict conductivity of a Nafion membranes exposed to impurities using a characteristic acid catalyzed reaction (methanol esterification of acetic acid), and, more importantly, (3) application of the latter technique to predict conductivity on Nafion in the catalyst layer of the MEA. H2-D2 exchange was found to be suitable for predicting hydrogen activation of Pt catalysts. The Nafion (ca. 30 wt%) on the Pt/C catalyst resides primarily on the external surface of the C support where it blocks significant numbers of micropores, but only partially blocks the pore openings of the meso- and macro-pores wherein lie the small Pt particles (crystallites). For this reason, even with 30 wt% Nafion on the Pt/C, few Pt sites are blocked and, hence, are accessible for hydrogen activation. Of the impurities studied, CO, NH3, perchloroethylene (also known as tetrachloroethylene), tetrahydrofuran, diborane, and metal cations had significant negative effects on the components in a fuel cell. While CO has no effect on the Nafion, it significantly poisons the Pt catalyst by adsorbing and blocking hydrogen activation. The effect can be reversed with time once the flow of CO is stopped. NH3 has no effect on the Pt catalyst at fuel cell conditions; it poisons the proton sites on Nafion (by forming NH4+ cations), decreasing drastically the proton conductivity of Nafion. This poisoning can slowly be reversed once the flow of NH3 is stopped. Perchloroethylene has a major effect on fuel cell performance. Since it has little/no effect on Nafion conductivity, its poisoning effect is on the Pt catalyst. However, this effect takes place primarily for the Pt catalyst at the cathode, since the presence of oxygen is very important for this poisoning effect. Tetrahydrofuran was shown not to impact Nafion conductivity; however, it does affect fuel cell performance. Therefore, its primary effect is on the Pt catalyst. The effect of THF on fuel cell performance is reversible. Diborane also can significant affect fuel cell performance. This effect is reversible once diborane is removed from the inlet streams. H2O2 is not an impurity usually present in the hydrogen or oxygen streams to a fuel cell. However, it is generated during fuel cell operation. The presence of Fe cations in the Nafion due to system corrosion and/or arising from MEA production act to catalyze the severe degradation of the Nafion by H2O2. Finally, the presence of metal cation impurities (Na+, Ca 2+, Fe3+) in Nafion from MEA preparation or from corrosion significantly impacts its proton conductivity due to replacement of proton sites. This effect is not reversible. Hydrocarbons, such as ethylene, might be expected to affect Pt or Nafion but do not at a typical fuel cell temperature of 80oC. In the presence of large quantities of hydrogen on the anode side, ethylene is converted to ethane which is very nonreactive. More surprisingly, even more reactive hydrocarbons such as formic acid and acetaldehyde do not appear to react enough with the strong Bronsted acid sites on Nafion at such low temperatures to affect Nafion conductivity properties. These results clearly identify a number of impurities which can have a detrimental impact on fuel cell performance, although some are reversible. Obviously, fuel cells exposed to impurities/poisons which are reversible can recover their original performance capabilities once the impurity flow is stopped. Impurities with irreversible effects should be either minimized in the feed streams, if possible, or new catalytic materials or ion conductors will need to be used to minimize their impact.