Atomically dispersed high-loading Pt-Fe/C metal-atom foam catalyst for oxygen reduction in fuel cells

Designing efficient and stable single-atom catalysts (SACs) for the oxygen reduction reaction (ORR) is essential to promote fuel cell commercialization. In this study, the Bayesian Regularization Neural Network (BRNN) has been used to determine the carbothermal shock (CTS) conditions for preparing high-loading Pt-Fe/C SACs. The BRNN model effectively identified a low reaction temperature of 283 degrees C for achieving high-loading (15 wt %) Pt-Fe/C SACs. Based on this model, we successfully synthesized Pt-Fe/C SACs with a metal mass content of 11.31 wt % using the CTS method. These ultrahigh content metal Pt-Fe/C SACs consisted of dispersed single atoms and foamlike atomic constructions, namely Pt-Fe/C atom foam catalysts (AFCs), rendered them promising candidates for electrocatalysis applications. The Pt-Fe/C AFCs exhibited good catalytic activity and durability toward ORR in alkaline conditions. The Pt-Fe/C AFCs were employed in a direct borohydride fuel cell (DBFC) as cathode catalyst, which resulted in a maximum power density of 214 mW cm-2 at 60 degrees C and demonstrated stable discharge for 40 h at room temperature. The electrocatalytic performance of the Pt-Fe/C AFCs surpassed that of low-loading Pt-Fe/C SACs and commercial Pt/C catalysts. The rapid ramp-up rate and the rapid cooling rate of the CTS process enabled the incorporation of high-loading Pt and Fe metal atoms into the graphitized carbon layer, resulting in an increased number of active sites of the Pt-Fe/C AFCs. This work provides an effective strategy for manufacturing efficient and stable SACs for practical applications in fuel cells.