Enhancing the Performance of Sr₂Fe_1.3Ni_0.2Mo_0.5O_6−δ as Methane-Fueled SOFC Anode via In-Situ Exsolution of Ni-Fe Nano-Catalyst

3.1. Phase Structure and Morphology of SFNMO Figure 1a presents comparative X-ray diffraction (XRD) patterns of SFMO and SFNMO. Analysis confirms that SFNMO perovskite synthesized via high-temperature calcination in air exhibits a phase-pure cubic structure, indicating successful incorporation of Ni²⁺ ions into the SFMO lattice. Local magnification of the 31–34° region reveals a slight high-angle shift of SFNMO’s primary peak relative to SFMO (Figure 1b). This shift is attributed to Ni²⁺ substitution for Fe²⁺, where the smaller ionic radius of Ni²⁺ (0.72 Å) versus Fe²⁺ (0.76 Å) induces minor lattice contraction without altering crystal symmetry [33]. Based on the Rietveld refinement method, the calculated lattice parameter of SFNMO perovskite is 7.79 Å, which is slightly smaller than that of SFMO perovskite (7.84 Å). Figure 1c examines SFNMO’s phase evolution under reducing conditions. While the cubic perovskite framework persists, the emergence of Ruddlesden-Popper (RP)-type phases indicates partial structural reorganization. Concurrently, a weak reflection at 43.8° corresponds to metallic Ni-Fe alloy formation, confirming nanoparticle exsolution accompanied by partial matrix decomposition. To evaluate operational stability, reduced SFNMO was exposed to dry or wet methane (700 °C, 2 h). XRD analysis shows dry methane exposure degrades structural integrity, generating SrCO₃ and SrMoO₄ impurities (Figure 1d). Conversely, humidified methane preserves phase purity, with no secondary phases detectable (Figure 1e), demonstrating steam’s stabilizing effect. Consequently, all subsequent electrochemical evaluations employed humidified methane fuel. SFNMO perovskite powders were analyzed using FE-SEM before and after reduction to investigate morphological evolution during exsolution. Figure 2a,b present the SEM images of the as-synthesized SFNMO at different magnifications, revealing relatively smooth surfaces without detectable impurities. Following reduction treatment at 800 °C for 2 h, distinct morphological evolution occurs as shown in Figure 2c. Numerous particles emerge preferentially at crystallographic discontinuities (e.g., grain boundaries or defect sites), while featureless planar regions remain largely particle-free. The exsolved particles exhibit distinctive pyramidal morphologies with dimensions typically exceeding 100 nm. This heterogeneous spatial distribution suggests localized reduction kinetics: regions with enhanced surface energy facilitate preferential nickel exsolution. Similar findings have been reported in previous studies [25,28]. This accelerated exsolution triggers localized perovskite decomposition, resulting in phase separation and formation of layered structural domains adjacent to the exsolved particles, which is consistent with XRD-identified RP-type phase evolution. 3.2. Compositional Analysis of Exsolved Particles Figure 3a shows TEM image of the SFNMO perovskite after reduction in a dilute hydrogen atmosphere at 800 °C for 2 h. The image reveals in-situ exsolved nanoalloy particles on the perovskite surface, displaying a relatively uniform distribution with an average particle size of approximately 27.6 nm and anchored within the matrix. This uniform particle size distribution suggests that the dilute hydrogen atmosphere facilitates the in-situ exsolution of small and well-dispersed nanoparticles. For a more detailed analysis of the in-situ exsolution behavior of the SFNMO perovskite, higher-magnification TEM images of the interface between in-situ exsolved particles and the matrix are presented in Figure 3b,c. Lattice fringe analysis of this region yielded a spacing of 0.2114 nm, corresponding to the (116) plane of the RP-phase perovskite. To investigate the compositional differences between the matrix and the in-situ exsolved particles, EDS mapping was performed on a localized region (Figure 3d). The results demonstrate a uniform distribution of strontium (Sr), iron (Fe), molybdenum (Mo), and oxygen (O) throughout the perovskite matrix. In contrast, nickel (Ni) is predominantly concentrated within the in-situ exsolved particles, with only a minor fraction remaining relatively uniformly distributed in the matrix. Further compositional analysis of the particles and their exsolution characteristics was conducted via EDS line scanning along a trajectory connecting the particles and the matrix (Figures 4a). The line scan profiles for Ni and Fe are presented in Figures 4b and Figures 4c, respectively. The data indicate that the Ni intensity is significantly elevated within the initial 19 nm segment of the scan path, corresponding to the alloy particles. Beyond this 19 nm point, the Ni intensity does not abruptly decrease but rather diminishes gradually with increasing scan distance. This gradual compositional transition suggests that the particles are not fully detached from the host structure but remain partially embedded within the matrix. Conversely, the Fe concentration within the 19 nm segment is relatively low, and beyond this range, it gradually increases with scan distance. This observation suggests that the in-situ exsolved alloy particles contain not only a substantial amount of Ni but also a minor quantity of Fe. This is corroborated by the EDS mapping of individual particles presented in Figure 4d, confirming that the in-situ exsolved particles are composed of a Ni-Fe alloy. XPS analysis was performed to probe the evolution of elemental composition and valence states in SFNMO perovskite before and after reduction. As shown in Figure 5a, the deconvoluted Fe 2p_3/2 spectrum of pristine SFNMO exhibits three characteristic peaks at binding energies of 709.6 eV, 710.6 eV, and 712.7 eV, corresponding to Fe²⁺, Fe³⁺, and Fe⁴⁺ species, respectively [34,35,36,37,38]. Quantitative analysis yields relative proportions of 30.53% Fe²⁺, 41.68% Fe³⁺, and 27.79% Fe⁴⁺, corresponding to an average iron valence state of +2.97. After reduction, an additional peak emerges at 706.5 eV, attributable to metallic Fe⁰ [39]. However, its low spectral weight (<2%) indicates limited reduction extent. Concurrently, the relative contents shift to 35.40% Fe²⁺, 34.69% Fe³⁺, and 27.19% Fe⁴⁺, resulting in a reduced average valence state of +2.84. The Mo 3d_5/2 spectra (Figure 5b) reveal two oxidation states for both samples: Mo⁵⁺ (231.6 eV) and Mo⁶⁺ (232.0 eV) [40,41,42]. The pristine material shows a Mo⁵⁺/Mo⁶⁺ ratio of 21.48%:78.52% with an average valence of +5.79. This ratio shifts to 29.01%:70.99% (+5.71) post-reduction, confirming partial reduction of molybdenum. Critical evidence is provided by Ni 2p_3/2 analysis (Figure 5c). While only signatures of Ni²⁺ (854.7 eV and 856.5 eV) are observed in the pristine sample, the reduced specimen displays new peaks at 852.6 eV (Ni⁰) and 858.6 eV (Ni⁰ satellite) [39], diagnostic of metallic nickel formation. Combined with the detection of Fe⁰, these results demonstrate in-situ exsolution of Fe-Ni alloy nanoparticles during reductive treatment, establishing SFNMO as a self-exsolving perovskite matrix capable of generating catalytically active heterostructured interfaces. 3.3. Cell Structure Figure 6 displays the cross-sectional morphology of the symmetrical SFNMO/GDC/SFNMO cell. The micrograph reveals a tri-layer structure comprising two porous electrodes sandwiching a dense electrolyte layer, with both electrodes exhibiting intimate interfacial contact with the electrolyte. The electrode thickness is approximately 30 μm, featuring a porous microstructure conducive to gas diffusion. Conversely, the electrolyte layer measures approximately 660 μm in thickness; its fully dense microstructure effectively prevents gas crossover between fuel and air compartments. These structural characteristics indicate successful fabrication of the LSGM electrolyte-supported cell. 3.4. Cell Performance and Stability Figure 7a compares the electrochemical performance of SFNMO/GDC/SFNMO and SFMO/GDC/SFMO symmetrical cells, operating under wet methane at 800 °C. Both cells exhibited open-circuit voltages (OCVs) between 0.7 and 0.8 V. The depressed OCV values primarily stem from the electronic conductivity of GDC electrolytes under intermediate-temperature reducing conditions, resulting from partial reduction of Ce⁴⁺ to Ce³⁺ in low oxygen partial environments [43]. Notably, the SFNMO/GDC/SFNMO cell achieved a peak power density of 82 mW cm⁻², surpassing the 61 mW cm⁻² attained by the SFMO/GDC/SFMO cell. As established in prior analyses, this enhancement is attributed to the in-situ exsolution of abundant Ni-Fe alloy nanoparticles on the SFNMO anode surface under reducing conditions. The performance enhancement mechanism of the in-situ exsolved nano-alloy particles on the SFNMO anode is illustrated in Figure 8. It is widely recognized that methane conversion on SOFC anodes involves complex chemical and electrochemical processes. The reaction typically begins with methane adsorption and activation on the anode surface, followed by reforming reactions (steam reforming or dry reforming) and subsequent electrochemical oxidation of the resulting syngas (primarily H₂ and CO). For SFMO anodes, methane conversion is relatively slow due to the limited adsorption and dissociation of methane on the surface (Figure 8a), which has been identified as the rate-limiting step [21]. In contrast, for SFNMO anodes, the in-situ exsolved Ni-Fe nano-alloy catalysts can significantly enhance both the adsorption and catalytic conversion of methane (Figure 8b). On one hand, the Ni-Fe nanoparticles efficiently promote methane cracking and reforming reactions, leading to increased production of H₂ and CO. On the other hand, they also serve as highly active sites for the electrochemical oxidation of H₂, CO, and intermediate carbon species (CH_x). Furthermore, the reduced SFNMO anode exhibits an increased oxygen vacancy concentration, providing enhanced pathways for oxygen ion transport. These combined effects synergistically improve the electrochemical performance. Figure 7b presents the electrochemical impedance spectra acquired under OCV conditions at 800 °C. The total cell resistance comprises two distinct components: the ohmic resistance (R_o), primarily determined by the ionic conductivity of the electrolyte and electrodes, and the polarization resistance (R_p), associated with electrode reaction kinetics and gas diffusion processes. The SFNMO cell exhibited an R_o of 0.88 Ω cm², marginally lower than the 0.94 Ω cm² observed for the SFMO cell, potentially due to the enhanced conductivity of SFNMO perovskite compared to the SFMO perovskite [33]. Crucially, the predominant performance difference stems from a significant reduction in R_p, which serves as the key determinant of the performance disparity. This reduction indicates lower activation energy requirements for electrochemical reactions during methane-fueled operation, thus confirming that the in-situ exsolution of Ni-Fe nanoalloy particles effectively enhances the catalytic conversion efficiency of methane on the perovskite anode. To evaluate the operational stability of the SFNMO-based cell under methane, a constant-current test was conducted at 0.1 A cm⁻² (Figure 7c). The cell voltage initially decreased rapidly, subsequently increased gradually, and stabilized at approximately 0.67 V without degradation, demonstrating robust stability in wet methane. Post-stability testing revealed the peak power density increased from 82 to 113 mW cm⁻² (Figures 7a). Impedance analysis indicated a slight reduction in R_o from 0.88 to 0.76 Ω cm², likely due to improved electrode-electrolyte interfacial contact during polarization. More significantly, a pronounced decrease in R_p was observed, suggesting electrochemical activation-induced microstructural optimization of the electrode. Although methane-fueled performance was enhanced, hydrogen-fed operation exhibited compromised performance. The temperature-dependent electrochemical performance of SFMO/GDC/SFMO and SFNMO/GDC/SFNMO symmetrical cells are displayed in Figure 9a and Figure 9b, respectively. Both cells exhibited OCVs between 0.80 and 0.94 V across all temperatures—significantly lower than cells employing YSZ or LSGM electrolytes. Notably, while OCV decreased with increasing temperature, power density increased monotonically. The SFMO anode cell achieved peak power densities of 43, 77, 124, 182, and 256 mW cm⁻² at 600, 650, 700, 750, and 800 °C, respectively. Under identical conditions, the SFNMO anode cell yielded corresponding values of 35, 66, 109, 164, and 223 mW cm⁻², demonstrating the consistently superior performance of SFMO anodes in wet hydrogen. This performance degradation in SFNMO originates from structural alterations during reduction: In-situ exsolution of Ni-Fe nano-alloy particles induces partial perovskite decomposition, forming secondary RP phase impurities. This phase segregation reduces available electrochemically active sites and compromises interfacial ion transport efficiency. Although the in-situ exsolved Ni-Fe alloys enhance hydrogen cracking kinetics, this catalytic benefit is offset by the degraded structural integrity of the SFNMO matrix. This structural compromise is further evidenced by EIS spectra, that the polarization resistance of the SFNMO/GDC/SFNMO symmetrical cell (Figures 9d) is substantially higher than that of the SFMO/GDC/SFMO symmetrical cell (Figures 9c) at each equivalent temperature. These findings underscore the critical importance of preserving matrix structural integrity during in-situ exsolution of metallic nanoparticles, which is potentially achievable through reduced reaction temperatures or optimized doping concentrations.