Model fuel cell contamination

The effects of mixtures of contaminants have not been extensively studied and little experimental data is available. There is a need for experimental studies and modeling in this area. Chapter 7 closes with a discussion on critical needs in fuel cell contamination modeling. 

One of the difficulties in fuel cell contamination modeling is estimating the unknown ORR parameters. In the case of no toluene being present, we needed to know the forward and backward reaction rates or their ratios for reaction (6.36) to reaction (6.38). To do this, we simulated experimental baseline data free of toluene contamination. The parameters used for the simulation are listed in Table 6.1, and the modeling results and experimental polarization curves are shown in Figure 6.3. 

Fuel cell contamination is a significant concern for PEM fuel cell operation and applications. Understanding the effects and mechanisms, and providing predictable tools to account for the influence of impurities on fuel cell performance, are the objectives of contamination modeling. 

Shi, Z., Song, D. T., Li, H. et al. 2009. A general model for air-side proton exchange membrane fuel cell contamination. Journal of Power Sources 186 435-445. 

St-Pierre, J. and Jia, N. (2002) Successful demonstration of Ballard PEMFCs for space shuttle applications. J. New Mater. Electrochem. Syst. 5, 263-271 St-Pierre, J., Jia, N. and Rahmani, R. (2008) Proton exchange membrane fuel cell contamination model - Competitive adsorption demonstrated with NO J. Electrochem. Soc. 155, B315-B320 St-Pierre, J., WrUdnson, D. P., Knights, S. and Bos, M. (2(XX)) Relationships between water management, contamination and Ufetime degradation in PEFC. J. New Mater. Electrochem. Syst. 3,99-106... 

Zhang J, Wang H, Wilkinson DP, Song D, Shen J, Liu ZS (2005) Model for the contamination of fuel cell anode catalyst in the presence of fuel stream impurities. J Power Sources 147 58-71... 

Figure 6.5. Modelled data and experimental results. Top current density = 0.5 A cm bottom current density = 1.0 A cm [33]. (Reproduced by permission of ECS—The Electrochemical Society, from Shi Z, Song DT, Zhang JJ, Liu ZS, Knights S, Vohra R, et al. Transient analysis of hydrogen sulfide contamination on the performance of a PEM fuel cell.)...

Based on the above general model, Shi et al. [33] has proposed a transient kinetic model to describe the contamination of the PEM anode catalyst layer by H2S present in the fuel feed stream. Figure 6.5 shows flie model-predicted cell voltages in comparison with experimental results. It can be seen that their model provides an excellent fit with the experimental results. 

Dr. Zheng Shi obtained her Ph.D. in Physical Chemistry from Dalhousie University in 1990. Dr. Shi has over fifteen years of theoretical modeling experience in the fields of electronic structure, reaction mechanisms, drug development, catalysis development, and structure activity relationships. She has worked at several pharmaceutical companies as well as universities, including Simon Fraser University and the University of British Columbia, prior to joining the National Research Council of Canada Institute for Fuel Cell Innovation. Currently, Dr. Shi is working in the area of fuel cell catalysis development and catalyst contamination studies. Dr. Shi s research focuses on the fundamental understanding of electrocatalysis reaetion meehanisms, and the development of structure activity relationships and fuel eelt eontamination kinetic models. 

The class of chemicals designated as VOCs includes a wide range of carbon-based molecules of sufficient vapor pressure to be present in the air, such as aldehydes and ketones. The most common VOC is methane, the primary component of nafural gas. There are various sources, both natural and human, of VOCs. The response of the fuel cell to VOCs will vary significantly, depending on the molecules in question, but can be significant. For example, benzene and toluene at the ppm level have both been found to significantly affect performance, with the dominant effect believed to be due to adsorption on the catalyst surface resulting in kinetic losses. A semiem-pirical model for toluene contamination based on kinetic losses is described in chapter 3, section 3.8. 

Contamination modeling is an important aspect of fuel cell development. It is required to interpolate and extrapolate experimental results to expected conditions in real-world operation, as it is impractical to test all combinations of reactant concentrations and fuel cell operating conditions. Modeling also assists in the development and validation of hypothesized contamination mechanisms. Model development for the anode is more extensive than that for the cathode contamination. The majority of the modeling deals with the kinetic effects associated with adsorption of contaminant species on the cathode and anode catalysts. 

Chapter 7 provides a comprehensive literature review and discussion on fuel cell anode confamination modeling. An introduction to fuel contamination is provided, including a hydrogen fuel quality specification for transportation hydrogen. 

The model results are compared to experimental data based on cesium contamination of a fuel cell operated in hydrogen pump mode. 

Like CO, H2S is a major contaminant in the fuel streams of PEM fuel cells. Modeling and experimental studies [65-74] have indicated that trace amounts of H2S can cause dramatic degradation in fuel cell performance. For example. 

Semiempirical Model for Fuel Cell Performance in the Absence of Contaminants... 

Semiempirical Model for Fuel Cell Performance in the Presence of Toluene In the presence of toluene in the air stream, the fuel cell performance degraded. Figure 3.15 illustrates two sets of representative results of toluene contamination tests, conducted with various levels of toluene concentration at current densities of 0.75 and 1.0 A cm , respectively. The cell voltage experienced a transient period (nonsteady state) immediately after the introduction of toluene, then reached a plateau (steady state). The duration of the transient period and the magnitude of the cell voltage drop to the plateau were strongly dependent on toluene concentration and current density. 

Contamination in a PEM fuel cell directly affects the kinetics, conductivity, and mass transport properties of the cell. In particular, the blocking of electrocatalysts by impurity adsorption can drastically reduce the effective surface area of the catalysts and, thus, slow down the kinetics and hinder cell performance. This chapter is devoted to cathode contamination modeling. 

We have developed a general cathode contamination model [20] for PEM fuel cells, which is presented in this section. 

When contaminant presents in the fuel cell, its concentration at the catalyst layer varies with both the inlet contaminant concentration and the current density, as discussed in Shi et al. [18]. Furthermore, the contaminant adsorption (desorption) rate constant is also related to the electrode potential. This variation of the contaminant concentration can be obtained by introducing the CGDL and cathode flow field into the model, which definitely increases its complexity. For simplicity here, we considered the product of the contaminant adsorption (desorption) rate constant and the contaminant concentration at the CCL, as a fimction of current density and contaminant inlet concentration (kCp-- f Cpr J))/ where Cp is the contaminant inlet concentration in the cathode charmel. Based on the experimental data at current densities of 0.2, 0.5, 0.75, and 1 A/cnP, and contaminant inlet concentrations of... 

Besides the kinetic models we discussed above, the other cathode contamination models available in the literature are the empirical model and the competitive adsorption model [4,57]. Empirical models have been successfully used to describe FEM fuel cell performance at different temperatures and pressures [55,56]. Equation (6.73) was proposed by Kim et al. [56] ... 

Empirical models are a valid approach for providing estimates of contamination effects on fuel cell performance. The major drawback of the empirical model approach is the lack of physical interpretation of various parameters. 

For nonempirical models, NO2 contamination has been reported in St-Pierre et al. [57]. This model is based on the competitive adsorption kinetics of O2 and NO2 on Ft catalysts. The species considered on the Pt surface are adsorbed 02and NO2, and free Pt sites. With this simple model, the authors discussed NO2 threshold concentrations. With the assumptions that the cell voltage was 0.8 V and the allowable performance loss for a fuel cell was 10%, the predicted allowable NO2 concentration was 0.39 ppm. This is a preliminary model only. Considerable work is still anticipated to extend the model s capability of handling contamination at different current densities, and extend it to other contaminants, which involves electrochemical reactions as well. 

---