Effect of Heat Treatment on Activity of Fe/N/C Catalyst Prepared by Hemin for Oxygen Reduction Reaction

Introduction

Polymer electrolyte membrane fuel cells (PEMFCs) are one of the promising candidates for next generation clean power sources. As a low-cost alternative to the Pt based catalysts used in PEMFCs for oxygen reduction reaction (ORR), metal/nitrogen/carbon (M/N/C) catalysts have been paid much attention. Recently, highly active M/N/C catalysts have been reported [1-6].

Carbon-supported Fe/N/C catalysts synthesized from Hemin have recently shown high fuel cell performances [3]. Hemin, an iron-containing porphyrin, contains catalytically active structure, the iron coordinated to nitrogen, for the ORR. The Fe/N/C catalysts were synthesized by the calcination of Hemin at 600ºC for 2-3 hours [3, 4]. The heat treatment process is usually applied for Fe/N/C catalysts because the pyrolysis and graphitization of organic carbons as the source of iron and nitrogen are considered necessary for the activity and stability of catalysts [5]. At the same time, however, the high temperature calcination in a long period of time accompanies with the aggregation of iron nanoparticles as a result of desorption of Fe/N/C catalytic centers, which would cause the decrease of ORR activity [5, 6]. On the other hand, Kamiya et al. reported Fe/N/C catalyst produced by instantaneous calcination (45s) of pentaethylenehexamine coordinated with iron (Fe-PEHA) and graphenoxide [6]. They found that the instantaneous calcination would preferentially form atomic iron coordinated to nitrogen without producing aggregated iron nanoparticles.

Here, we investigated the effect of brief calcination on ORR activity with different precursors, Hemin and Fe-PEHA. If it is possible to selectively place iron atoms into catalytically active sites, which is an advantage of the method [6], it would be useful to maximize the performance of Fe/N/C catalysts.

Experimental

Hemin and Fe-PEHA were mixed with carbon black (CB) in dimethylformamide and ethanol, respectively. After ultrasonicating the solution, dried mixture was obtained by a rotary evaporator, and then heated at 900 ºC for 100s and 70s for Hemin and Fe-PEHA, respectively. ORR activity of the catalysts was evaluated in 0.5M H₂SO₄(aq) by rotating disk electrode(RDE) experiments.

Single cell tests were performed using a 36cm²MEA which used the Fe/N/C catalyst and Pt catalyst as cathode and anode, respectively. Current-voltage (I-V) curves were obtained at a cell temperature of 65 ºC by feeding fully humidified hydrogen and air or pure oxygen. The gases were supplied with a restricted stoichiometry which resembles the realistic operating conditions. Under these conditions, one can reasonably evaluate a performance difference between the Fe/N/C and Pt catalysts.

­­Results and Discussion

Fig. 1 shows the result of RDE for the synthesized Fe/N/C catalysts. Our Hemin based catalyst (red curve) showed the onset potential of about 0.91V vs. RHE. The onset potential was defined as the potential realizing a current density of 50µA in this study. This onset potential is comparable to those of other Hemin-based catalysts supported on graphene nanoplatelets or CB by long calcination of 2-3hours [3,4]. It is difficult to precisely compare ORR activity of these catalysts at this point because they may be mounted on RDE in a different manner. However, it is still interesting that the two different time scales of calcination, a couple of hours and 100s, can realize the comparable onset potential. The detailed analysis for comparing the structure and ORR mechanism in the two types of catalysts is in progress. We also note that our TEM and XPS analysis showed the coordination of iron to nitrogen without any sign of aggregated iron nanoparticles.

The purple line in Fig. 1 shows the result for the Fe/N/C catalyst by 70s-calcination of Fe-PEHA. Although the Hemin- and Fe-PEHA based catalysts seem to possess similar Fe-N-modified graphite structure, the latter showed lower onset potential (about 0.8 V vs. RHE). Since it is not clear what critically causes the difference in ORR activity between them, we are unable to conclude the superiority of Hemin as the precursor over Fe-PEHA. However, it seems worth investigating how the structures of Fe/N/C active centers are modulated by choosing different precursors.

We also obtained I-V curves for the Hemin-based Fe/N/C catalysts with different loadings of 0.2-4mg/cm² by single cell tests. It was found that the cell performance improves with increasing catalyst loading. For the case of 4mg/cm², OCV was 0.86V and a cell voltage of 0.4V was obtained at a current density of 0.5A/cm². Further analysis will be discussed in the session.

References

[1] G. Wu et al., Science 332, 443 (2011).

[2] E. Proietti  et al.,  Nature Commun. 2:416 (2011).

[3] R. Jang et al., Electrochem. Commun. 19, 73 (2012) .

[4] Z. Liao et al., J. Phys. Chem. C 115, 2604 (2011).

[5] Bezerra et al., Electrochim. Acta 53, 4937 (2008).

[6] K. Kamiya et al., Chem. Commun. 48, 10213 (2012).