Rational Placement of Catalysts for Oxygen Reduction and Evolution Reactions Based on the Reaction Sites in Porous Gas Diffusion Electrodes

Abstract

Metal-air secondary batteries using alkaline electrolyte solutions are promising candidates for next-generation large-scale energy storage systems because of their potential high capacity, low cost, and high safety standard. However, large overpotential of bifunctional air electrodes hindered their wide applications. Recently, increasing research on electrocatalysts for oxygen reduction reaction (ORR) and oxygen reduction reaction (OER) has been conducted to reduce the overpotential, and many active catalysts have been developed. On the other hand, formation of reaction sites is also important since the reaction sites for ORR and OER are basically different1. In this work, we investigated the reaction sites in porous gas diffusion electrodes (GDE) and constructed a GDE with a double-layered catalyst layer based on the difference in the reaction sites for ORR and OER2.La0.6Ca0.4CoO3 (LCCO, bifunctional catalyst), La0.4Sr0.6MnO3 (LSMO, ORR catalyst) and Ca2FeCoO5 (CFCO, OER catalyst) were synthesized by the polymerized complex method. Porous gas diffusion electrodes are constructed by the hot-press method. Catalyst layers were composed of the prepared catalyst (50 wt%), graphitized carbon black (35 wt%), and poly(tetrafluoroethylene) (15 wt%). Gas diffusion layers (GDL) were commercial carbon papers (TGP-H-030H, Chemix). IR-corrected polarization curves were measured by the constant current and AC impedance measurements. Oxygen diffusion resistances for ORR and OER were estimated from the difference in the steady-state potential between pure oxygen and air atmosphere (⊿E = |E O2 – Eair|). Larger increase in ⊿E indicates larger oxygen diffusion resistances in the oxygen reaction processes3.Figure 1 (a, b) shows ⊿E-I curves of LCCO-GDE for ORR and OER in 4.0 and 8.0 dm–3 KOHaq. The oxygen permeability(D O2×C O2) in 4.0 dm–3 KOHaq is 16 times as large as that in 8.0 dm–3 KOHaq while the ion conductivities of these electrolyte solutions are at the same extent. The increase in ⊿E for ORR was larger in 8.0 dm–3 KOHaq, indicating dissolved oxygen involved in the ORR process. In contrast, the ⊿E for OER was nearly negligible in these electrolyte solutions, suggesting that oxygen was mainly transported as bubbles. We also measured ⊿E-I curves of the LCCO-GDE in 8.0 dm–3 KOHaq having different catalyst layer thicknesses. The increases in ⊿E were almost independent of the catalyst layer thicknesses for ORR and OER, suggesting that the reaction sites were concentrated in the catalyst layer. Finally, we compared polarization curves of the GDEs with single-layered catalyst layer (LSMO+CFCO|DGL-GDE) and double-layered catalyst layers (LSMO|CFCO|GDL-GDE and CFCO|LSMO|GDL-GDE) to identify the reaction sites for ORR and OER (Fig. 1 (c, d)). The smallest overpotential of CFCO|LSMO|GDL-GDE for both ORR and OER showed that the reaction sites for ORR and OER were concentrated at the electrolyte-side and gas-side, respectively. In addition, CFCO|LSMO|GDL-GDE showed higher durability than LSMO+CFCO|DGL-GDE. From these results, it can be concluded that the placement of the ORR and OER catalysts on the electrolyte-side and gas-side of the CL can improve both activity and durability of the bifunctional air electrode.