Catalyst layer surface property changes

The result of these degradation mechanisms is that that the GDL and the MPL both lose their hydrophobic character [133, 153, 154], and that the pore structure of the materials changes. The relation between microstructure and surface properties and mass transport properties has been the subject of several recent experimental studies [155,156], which indicate that indeed mass transport can be seriously affected by the hydrophobicity of the GDL and MPL as well as by the pore size. This will contribute to the gradual decay of the performance, though it is hard to distinguish the effects of changes in the GDL/MPL to those of changes in the catalyst layer. 

Electrochemical corrosion of carbon supports was widely studied in the context of phosphoric acid fuel cell development (Antonucci et al. 1988 Kinoshita 1988), but recently also the low-temperature fuel cell community paid more attention to this process (Kangasniemi et al. 2004 Roen et al. 2004). Carbon corrosion in fuel cell cathodes in the form of surface oxidation leads to functionalization of the carbon surface (e.g., quinones, lactones, carboxylic acids, etc.), with a concomitant change in the surface properties, which clearly results in changes of the hydrophobicity of the catalyst layer. Additionally, and even more severe, total oxidation of the carbon with the overall reaction... 

However, in the case of multimetallic catalysts, the problem of the stability of the surface layer is cmcial. Preferential dissolution of one metal is possible, leading to a modification of the nature and therefore the properties of the electrocatalyst. Changes in the size and crystal structure of nanoparticles are also possible, and should be checked. All these problems of ageing are crucial for applications in fuel cells. 

The variations of acidic properties in the surface layers and in the bulk solid catalysts after calcination, reduction, or coking were examined by pyridine Nls XPS [4,7] and by the pyridine infrared adsorption techniques, respectively. This provides a means to compare the changes in the characteristic BrBnsted and Lewis acidity functions after those treatment conditions. First of all, TPD of ammonia revealed that both coked and regenerated samples exhibited much decreased acidity as compared with either calcined or reduced samples before the reaction of n-heptane conversion in either N2 or H2 stream [7]. The adsorption of pyridine may cause further perturbation to the Pt4+ or Pt 2+ species in the zeolite as indicated by the increase in binding energies of Pt3d5/2 electrons, as shown in Table 3 and Figure 4,... 

From these observations it can be concluded that the changes in catalyst properties in water under reductive circumstances especially at high pH are the result of two phenomena 1) Growth of the platinum particles by a process in which the crystallites become mobile 2) Coverage of the platinum surface. The latter can be caused either by coverage with carbonaceous products originating from the support or by disappearance of the mobile platinum particles in between the graphite layers. 

At low temperatures where the surface ionic mobility is restricted the catalytic activity of a divided oxide for oxidation or reduction processes is determined primarily by the nature and the concentration of lattice defects in the surface layer and by the strength of the bond between oxygen and these defects. The nature and concentration of the defects depend upon the chemical nature of the catalyst, its previous history, and on the course of the catal diic reaction itself. In some instances, a small modification in the preparation procedure or in the pretreatment may result in an important change of cataljrtic activity. Such abrupt changes of activity may be caused by the occurrence of different reaction paths on apparently similar catalysts. Since the catalytic act is localized on particular surface structures, the energy spectrum of the active surface is of paramount importance and correlations between catal3rtic activities and collective or average properties of the catalyst are crudely approximate. 

Although such effects are well known for hydrocarbon reactions over platinum and other metals,it is unlikely to apply to the present situation where considerable surface reconstruction, corresponding to the formation of an oxide film, is likely. The properties of such a layer would change with particle size simply because of the increase in surface energy with decreasing size. The smaller particles would tend more towards the composition Ag20 than the larger ones. Since it is known that silver(i) oxide is not a selective catalyst for ethylene oxidation such a model could explain these size effects. 

---