Mechanistic Insights into the Oxygen Reduction Reaction on Metal–N–C Electrocatalysts under Fuel Cell Conditions

Fuel cells can play a crucial role in a future sustainable world; however, these devices are currently based on platinum group metals (PGMs), which is a catalyst group of materials considered as critical raw materials. Furthermore, the oxygen reduction reaction (ORR), which is more than six orders of magnitude slower than the hydrogen oxidation reaction, requires the cathode of polymer electrolyte fuel cells (PEFCs) to utilise eight times the amount of platinum of the anode (i.e. 0.4 mgPt cm 2 cathode vs. 0.05 mgPt cm 2 anode). [2] As a sustainable alternative, non-precious oxygen reduction catalysts have been a main focus of research, and, in this regard, carbon-based catalysts are emerging as a viable alternative on a long-term scenario. 5] They are based on naturally abundant elements, for example, C ( 90%), N, Fe or other abundant transition metals, and we have recently shown that carbon catalysts can also be highly selective towards the oxygen reduction reaction, with tolerance to a wide range of poisons that adversely affect PGM-based catalysts. Even though extremely promising, some challenges remain regarding fundamental and applied developments of carbon catalysts, for instance: (1) understanding the kinetics of the oxygen reduction reaction on these materials; (2) translation of fundamentally (e.g. RRDE) observed performances into power output in practical devices; (3) dealing with degradation of these catalysts under operational conditions that limit their ultimate lifetime. There has been progress in addressing the oxygen reduction activity of carbon catalysts; 12–19] however, there is a gap in understanding the activity of the carbon catalyst cathodes, in part due to issues in appropriately determining ORR kinetics and mass transport overpotential under fuel cell conditions. One of the first attempts to establish an ideal performance for carbon-catalysed cathodes of PEFC was set by Gasteiger et al. , aimed at promoting high practical power outputs. This proposition was based on the assumption of equivalent Tafel kinetics of carbon catalyst to platinum (a Tafel slope of 70 mVdec 1 at 353 K) over the entire potential window, which, in fact, may not be appropriate for carbon-based catalysts or even platinum for that matter. Current carbon-based catalyst cathodes in a PEFC are unlikely to reproduce the performance of platinum catalyst cathodes, given that the location of the active sites on Pt/C are on the surface of an extended lattice, whereas those on the carbon catalysts are individual “molecular” sites mostly located at micropores. This will hinder species transport; that is, oxygen to and water from the active sites. Compared with previous work utilising rotating disc Three different transition metal-C-N catalysts are tested under a range of fuel cell conditions. It is found that common features of the polarisation curve can be explained by a change in electrocatalytic mechanism. Utilising a simple model to quantify the change in mechanisms, iR-free results of the fuel cell experiments are fit and found to be represented by a common set of parameters. The change in mechanism is assumed to be a switch from four-electron reduction of oxygen to water to a two-electron reduction to hydrogen peroxide followed by disproportionation of the hydrogen peroxide to water and oxygen. The data is used to estimate a mass specific exchange current density towards the oxygen reduction reaction (ORR) to water in the range 10 –10 13 Ag 1 depending on the catalyst. For the reduction of oxygen to hydrogen peroxide, the mass specific exchange current density is estimated to be in the range 10 –10 3 Ag . Utilising the electrokinetic model, it is shown how the mass transport losses can be extracted from the polarisation curve. For all three catalyst layers studied, these mass transport losses reach about 100 mV at a current density of 1 Acm . Finally a discussion of the performance and site density requirements of the non-precious metal catalysts are provided, and it is estimated that the activity towards the ORR needs to be increased by an order of magnitude, and the site density by two/three orders of magnitude to compete with platinum as an ORR electrocatalyst.