The reaction mechanism at the fuel cell cathode

Using density functional theory (DFT) with the B3LYP gradient-corrected exchange–correlation functional, we systematically studied the cathode reaction of polymer electrolyte membrane (PEM) fuel cells in gas-phase:
 
As catalyst material we assumed Pt, which was modeled with a cluster of 35 atoms.
Figure 1: Motivation for choosing Pt14.13.8.


We first calculated binding structures and energetics for each intermediate on the Pt(111) surface plane separately: O, H, O2, H2, OH, OOH, H2O. Atomic oxygen binds most strongly at the µ3-fcc position (77.71 kcal/mol), while molecular O2 prefers the bridge site (11.30 kcal/mol). Thus, OOH prefers the same geometry with one O covalently bound on top of a Pt atom (23.85 kcal/mol). Including zero-point energy (ZPE) a single H atom prefers the µ3-fcc over an on top site by ˜3.2 kcal/mol, whereas molecular H2 undergoes dissociation to two on top bound H atoms wile adsorbing on Pt. OH and water show comparable binding structures (on top bound), but a different type of binding. The hydroxyl radical binds covalently to one Pt atom (47.45 kcal/mol), and water uses the remaining lone pair orbital of its oxygen to attach to the surface atom (13.90 kcal/mol).

In order to study whole reaction pathways we calculated all possible dissociation processes of the various intermediates on the Pt cluster. Using all energetics we calculated heats of formation (?Hf) and combined these with the dissociation barriers. Since on the cathode oxygen and hydrogen catalytically reacts to water, we studied possible reaction pathways starting with gas-phase H2 and O2. We distinguished between two main reaction pathways:




Along the O2-Dissociation pathway oxygen adsorbs on the surface, dissociates, and finally reacts with hydrogen to form water. With a barrier of 31.66 kcal/mol the rate-determining step for this mechanism is the Oad + Had --> OHad reaction and not the dissociation of O2. Since O2 changes its adsorption structure during dissociation the dissociation barrier is lowered to only 15.02 kcal/mol. Along the OOH-Formation pathway adsorbed O2 first forms OOH with a surface hydrogen, and the generates OH via O–OH dissociation, which finally reacts with another hydrogen to water. For this mechanism the OOHad --> OHad + Oadfcc dissociation has the highest barrier with 17.13 kcal/mol. Thus, we propose the OOH-Formation mechanism to be the most likely pathway for the cathode reaction. This pathway may additionally be supported by recombination of two adsorbed surface oxygens.

Further investigations will address the effects of solvation and the presence of an external electric field. In this context, we will also present first simulations using the reactive ForceField (reaxFF).


Personnel: Dr. Timo Jacob

This project is co-directed by Dr. Boris Merinov and Dr. Adri van Duin.