Unifying Experimental and Computational Views over the Oxygen Reduction Reaction: Elucidating a Volcano Dilemma
Abstract: Currently, the biggest drawback on the road to commercialization of low-temperature fuel cells is the sluggish rate of the oxygen reduction reaction (ORR) at the cathode. Despite decades of intensive scrutiny, there is still no absolute consensus over the ORR mechanism. From the viewpoint of many computational electrochemists, the activity is controlled by the strength of the binding to the catalyst surface via a volcano-shaped plot. On the other hand, most experimental electrochemists attribute activity decrease to the buildup of spectator oxygenated species and acknowledge only the strong binding side of the volcano (activity is given by a straight line).
In order to elucidate the above point of contention, we have investigated in detail the thermodynamics and kinetics of the ORR on Pt(111) at stable coverages that develop in the relevant (0–1.0 V) potential range. Thus we depart from the conventional approach in which mechanistic details are deduced from reaction energetics on adsorbate-free metal surfaces. We find surface kinetics (the O-O bond breaking steps) to be relatively potential independent and thus only play a minor role in platinum deactivation at potentials above 0.8–0.9 V.
At the same time, we find a large dependence of the O2 adsorption energetics in the same potential range. In the mixed kinetic-diffusion controlled region (hydrophilic surface), the rate determining step (RDS) is given by a low reaction prefactor owing to a lengthy process of exchange between the water and oxygen molecules, whereas in the kinetically controlled region (hydrophobic surface) oxygen acts as a spectator species that effectively blocks O2 adsorption and dissociation in agreement with surface science experiments.
The intepretation of these results is that the RDS is governed by the potential determining step (PDS), which explains why the left leg of the conventional ORR volcano provides good activity predictions without explicitly taking kinetics into account. Subsequently, we extended the analysis to other active catalysts (palladium, silver, Pt3Ni, Pt3Cu, strained platinum overlayers etc.) and arrived at the conclusion that the volcano plot is not a good model for describing the ORR activity!
Instead, we propose that ORR activity is controlled by (1) the binding energy of spectator species and (2) the ability to adsorb O2, whereas the selectivity for water or hydrogen peroxide formation is given by the kinetics of the O2 bond splitting. The proposed concept resolves the point of contention by aligning more favorably with the experimental standpoint. Furthermore, together other works, it contests the general volcano concept for describing activities of electrochemical reactions.