Solid oxide fuel cells (SOFCs) are high-efficient solid-state energy conversion devices. However, all-ceramic self-supported SOFCs face several challenges such as high brittleness, difficulty in mechanical processing, poor thermal shock resistance, and limited weldability, which result in high manufacturing costs and restrict the applications in mobile power systems. In contrast, metal-supported SOFCs (MS-SOFCs) with metal materials as the external structural support, exhibit remarkable mechanical strength, low cost, and rapid start-up capability, making it highly promising for mobile applications. The anode is a critical component of MS-SOFCs, serving as the site where fuel oxidation occurs to generate electrons. Its microstructure significantly influences the density and effectiveness of the triple-phase boundaries (TPB), where the gas phase, the ionic phase, and the electronic phase intersect. The TPB density largely determines the polarization resistance, with its low-frequency component being inversely related to anode gas diffusion. A common strategy to enhance gas transport is the incorporation of pore-formers, such as graphite, into the anode raw materials. Most studies focus on the type, particle size, and content of pore-formers, which directly affect the pores number, size, and distribution. In this work, atmospheric plasma spraying (APS) was employed to fabricate three types of anodes and corresponding cells. APS reduces thermal input to the metal substrate, effectively preventing oxidation, deformation, and elemental interdiffusion between the metal support and the anode at high temperatures. This study systematically investigates the influence of graphite incorporation methods on the microstructural evolution of the anode and the resulting cell performance, providing important theoretical insights into the operational mechanisms of SOFC anodes.

Porous 430L stainless steel substrates were used as supports. Three different NiO-GDC (Gd_0.2Ce_0.8O_1.9) anode powders were prepared. C1 is the baseline without a pore-former. C2 contains 40% (in volume fraction) graphite, which is mixed by spray granulation process to produce composite particles. C3 is made by mechanically mixing 40% of the same graphite with C1 powder. All powders are spherical with good fluidity. Anode layers were deposited via APS. Subsequently, the C2 and C3 anodes were heat-treated in air to remove the graphite pore-former at 750 ℃ for 2 h. A ScSZ (Sc₂O₃-ZrO₂) electrolyte and an LSCF (La_0.6Sr_0.4Co_0.2Fe_0.8O_3-δ) cathode were subsequently deposited by APS to build single cells. The microstructure and porosity of the anodes were characterized using scanning electron microscopy (SEM) and image analysis software. The surface roughness was measured by profilometer. The electrochemical performance, including the open-circuit voltage (OCV), current-voltage-power (I-V-P) and electrochemical impedance spectroscopy (EIS), were evaluated in the range of 600-750 ℃ using humidified H₂ as fuel and air as oxidant. The equivalent circuit fitting of EIS data is carried out to quantitatively analyze the contribution of charge transfer, surface adsorption/dissociation and gas diffusion in polarization impedance.

The incorporation of graphite pore-former significantly modified the pore size distribution and total porosity within the anodes. The measured porosity of C1, C2, and C3 was 26%±2%, 37%±3.1%, and 42%±2.3%, respectively. The C1 anode featured a relatively dense structure with uniformly distributed pores, which primarily consisted of submicron cracks and fine pores originating from the thermal stress inherent to the APS process. In contrast, the C2 anode showed a notable increase in both the number and size of pores, which were homogeneously dispersed without significant agglomeration. The C3 anode, however, contained a substantial amount of large pores, mostly ~5 μm in diameter, attributed to the agglomeration of graphite particles during mechanical mixing, resulting in coarse and irregular pore structures after heat treatment. Furthermore, the addition of graphite modified the thermal response characteristics of the agglomerated powder during the spraying process, promoting the formation of a more uniformly melted microstructure in the anode layer. The average surface roughness (R_a) values for C1, C2, and C3 were 6.64, 7.06 nm, and 7.66 μm, respectively, indicating that graphite addition increased anode surface roughness. This phenomenon is due to the thermal decomposition of graphite during the plasma spraying process, where high temperatures cause partial oxidation of graphite in the open atmosphere, thereby generating CO₂ gas. The release of this gas from the incompletely solidified anode surface etches irregular pits and protrusions, ultimately leading to increased surface roughness. The cell without graphite pore-former (C1) consistently demonstrated the highest OCV and maximum power density, reaching 1.0 V and 957 mW·cm^-2 at 750 ℃, respectively. EIS analysis revealed that the anodes with graphite pore-former (C2 and C3) exhibited improved charge transfer capability, thereby reducing the high-frequency polarization resistance. Despite this, the overall output performance of C2 and C3 did not show effective enhancement, which is attributed to their increased ohmic resistance (R_o) and lower OCV. The elevated surface roughness and inherent porosity in the C2 and C3 anodes adversely affected the quality of the subsequently sprayed electrolyte layer, introducing microcracks and gas permeation pathways. This resulted in increased R_o and reduced OCV, ultimately weakening the benefits gained from the reduced polarization resistance.

The method of graphite pore-former addition significantly affects the anode microstructure and overall cell performance. Compared with mechanical mixing, spray granulation produces a superior and uniform pore structure. However, contrary to conventional expectation, the introduction of graphite pore-forming agent into the APS anode reduces overall cell performance due to induced electrolyte defects, which elevated R_o and lowered OCV. Future optimization should focus on strategies to reduce anode surface roughness and refine pore structure without affecting the quality of electrolyte deposition. It is anticipated that this will further enhance the output performance of MS-SOFCs.