Carbon based catalysts containing N, Fe or Co have shown activity for the oxygen reduction reaction (ORR) comparable to platinum in acid electrolytes at low current densities.(1–3) While promising, these catalysts require high loading to achieve satisfactory activity resulting in a thick catalyst layer and significant mass transport limitations at the high current densities desired in fuel cells. Further improvements in volumetric activity are therefore needed. Furthermore, long-term stability and suppression of H₂O₂ selectivity need to be addressed. While the active site of Fe and N containing graphene is believed to be an FeN₄C₁₂ site,(4) a nitrogen-free active site was recently proposed for graphite encapsulated Fe₃C, which showed promising durability activity and durability in acid electrolyte.(2)

Here we use atomic-scale density functional theory to investigate whether it is possible for Fe₃C to enhance the activity of nitrogen-free graphite for the 4e^- reduction of O₂ to H₂O. Activity and selectivity of various extended model surfaces are investigated within the framework of the computational hydrogen electrode model,(5) which previously has been shown to capture trends in 4e^- and 2e^- reduction of O₂.(6, 7)

We find significant activity enhancement of graphite zigzag edges and a single graphene layer on Fe₃C(001), while activity enhancement of multiple layers of graphite is not observed. Comparing the activity of graphite zigzag edges anchored on Fe₃C(001) to Pt(111), we predict the edge sites are less active for the 4e^- ORR and significantly more selective for the 2e^-ORR than Pt. This is caused by weaker binding of ORR intermediates on the graphite step edge compared to Pt, as show in the figure below.

It is, however, possible to increase activity and selectivity for the 4e^- ORR through expansive strain. Although Fe₃C significantly facilitate O₂ activation on isolated graphene sheets, we find the activation is in sufficient to catalyze the 4e^- reduction. Instead H₂O₂ formation is predicted. These results alludes to the necessity of adding dopants such as nitrogen or boron to further enhance reactivity and suppress H₂O₂ production on graphite-Fe₃C based catalysts for the ORR. 

Acknowledgements

This work is funded by Innovation Fund Denmark through 4106-00012A (NonPrecious) Initiative Towards Non-Precious Metal Polymer Fuel Cells.

 References

1.            G. Wu, K. L. More, C. M. Johnston, P. Zelenay, Science. 332, 443–7 (2011).

2.            Y. Hu et al., Angew. Chem. Int. Ed. Engl. 53, 3675–9 (2014).

3.            M. Lefevre, E. Proietti, F. Jaouen, J.-P. Dodelet, Science. 324, 71–74 (2009).

4.            A. Zitolo et al., Nat. Mater. 14, 937–942 (2015).

5.            J. K. Nørskov et al., J. Phys. Chem. B. 108, 17886–17892 (2004).

6.            J. Greeley et al., Nat. Chem. 1, 552–556 (2009).

7.            V. Viswanathan, H. A. Hansen, J. Rossmeisl, J. K. Nørskov, J. Phys. Chem. Lett. 3, 2948–2951 (2012).

Figure (a) Free energy diagram at 0.9 V vs RHE for the 4e^- ORR on Pt(111) and graphite zigzag edges anchored to Fe₃C(001). (b) Free energy diagram for H₂O₂ production at 0.69 V vs RHE.