Best Practices in Using Foam-Type Electrodes for Electrocatalytic Performance Benchmark

Abstract

Open-pore foam-type electrodes are three-dimensional (3D) materials (typically, Ni, Cu, Fe, and C; images are shown in Figure S1A) with a wide porosity range (70−98%) where their porous cellular structures are interconnected so fluid and gas can pass through. The unique properties of conductive foams, such as high specific surface area and structural rigidity, make them suitable and popular supports on which the active materials are coated or grown. For example, nearly 8% of the water-splitting electrocatalysts reported in 2019 (Figure S1B) involve foamtype electrodes, claiming outstanding electrocatalytic performance and superiority over the others (Figure S2). Electrocatalysts comparable to commercial noble metal catalysts are often found amid the headlines of various journals. Two indicators, overpotential at given current density (often 10 or 100 mA cm−2, denoted as η10 or η100) and Tafel slope (potential required to achieve every current density increment of 10 mA cm−2), are usually employed for catalytic activity evaluation and comparison. However, we believe some oversimplified and overlooked quantitative and experimental issues in using foam-type electrodes have made the electrocatalyst benchmarking physically groundless and often incommensurable, eventually obstructing the understanding and rational design of electrocatalysts. In this Viewpoint, several problems of using foam-type electrodes as electrocatalyst support are discussed, proving some conventional approaches, such as current density normalization, surface area evaluation, and experimental setup, are problematic. (Figure 1) Surface Area Matters. In most reports, the surface area value used for current density (denoted as J) calculation is not defined, which directly impacts overpotential determination. The ideal normalization of current should be performed with the real active surface (RA, denoted as Areal): the number or the area of the active sites that are responsible for Faradaic current transfer during electrocatalysis (or atomic sites that transfer electrons for electrocatalysis). In an ideal situation where non-Faradaic current can be ignored, the normalization of current at any given overpotential to Areal would reflect the charge (Q) input/output per site/unit area every second. However, the determination of the real active surface area is challenging for many systems because it often depends on the applied potential, electrolyte, and the nature of the reaction involved, making it somewhat too complicated to estimate. Instead, two alternatives, geometric area (GA, denoted as AGA) and electrochemically active surface area (ECSA, denoted as AECSA), are often used. The use of GA for current normalization (denoted as JGA) could be traced back to the period when smooth planar electrodes were widely used in the electrochemical community, and it could be calculated based on the dimension of the surface contacting the electrolyte: AGA = 2hl + wl + 2wh (Figure 2A). In most cases, the thickness of the planar electrode is much smaller than its length and height, thus AGA ≈ 2hl. ECSA is generally evaluated based on the double-layer capacitance value (CDL) of the electrode. All surface species that are capable of forming a double-layer structure would contribute to the ECSA. However, the surface sites involved in the non-Faradaic charging−discharging processes do not necessarily participate in the Faradaic processes. Specifically, for a noncatalytic foam electrode (such as carbon foam): AECSA ≫ AGA, Areal = 0 (Figure 2A). For a foam electrode fully coated with a partially catalytic rough layer (Figure 2B), AECSA= AECSA of noncatalytic sites+ AECSA of catalytic sites and AECSA of catalytic sites= Areal, AECSA ≫ AGA, Areal > AGA. Normalizing the Faradaic current with AECSA (denoted as JECSA) will usually produce a smaller value than the real current