Authors: Horie Adabi, Pietro Giovanni Santori, Abolfazl Shakouri, Xiong Peng, Karam Yassin, Igal G. Rasin, Simon Brandon, Dario R. Dekel, Noor Ul Hassan, Moulay-Tahar Sougrati, Andrea Zitolo, John R. Varcoe, John R. Regalbuto, Frederic Jaouen, William E. Mustain
One of the most important needs for the future of low-cost fuel cells is the development of highly active platinum group metal (PGM)-free catalysts. For the oxygen reduction reaction, Fe–N–C materials have been widely studied in both acid and alkaline media. However, reported catalysts in the literature show quite different intrinsic activity and in-cell performance, despite similar synthesis routes and precursors. Here, two types of Fe–N–C are prepared from the same precursor and procedure – the main difference is how the precursor was handled prior to use. It is shown that in one case Fe overwhelmingly existed as highly active single-metal atoms in FeN4 coordination (preferred), while in the other case large Fe particles coexisting with few single metal atoms were obtained. As a result, there were drastic differences in the catalyst structure, activity, and especially in their performance in an operating anion exchange membrane fuel cell (AEMFC). Additionally, it is shown that catalyst layers created from single-atom-dominated Fe–N–C can have excellent performance and durability in an AEMFC using H2/O2 reacting gases, achieving a peak power density of 1.8 W cm−2 – comparable to similar AEMFCs with a Pt/C cathode – and being able to operate stably for more than 100 h. Finally, the Fe–N–C cathode was paired with a low-loading PtRu/C anode electrode to create AEMFCs (on H2/O2) with a total PGM loading of only 0.135 mg cm−2 (0.090 mgPt cm−2) that was able to achieve a very high specific power of 8.4 W mgPGM−1 (12.6 W mgPt−1).