Spatially decoupled design and AI-driven optimization for thermal uniformity and nitridation resistance in NH3-fueled micro-tubular SOFCs

Ammonia-fueled micro-tubular solid oxide fuel cells (MT-SOFCs) offer a compact route for carbon-free power generation, but the endothermic NH 3 decomposition can cause thermal non-uniformity, and local NH 3 enrichment can increase the thermodynamic risk of Ni nitridation in the anode. To address this issue, we propose a spatially decoupled MT-SOFC configuration that integrates an internal catalytic layer (ICL) into the fuel inlet tube. NH 3 decomposition kinetic experiments, multiphysics simulations and AI-assisted optimization are combined to evaluate flow resistance, electrochemical performance, thermal distribution and anode nitridation risk. The ICL shifts most NH 3 decomposition from the anode to the inlet tube, forming an H 2 -rich fuel environment while avoiding the severe pressure-drop penalty associated with packed-catalyst designs. At 100 sccm, the ICL design increases power density and electrical efficiency by 23.6% relative to the Base design and reduces the predicted Ni nitridation risk in the anode. Co-optimization of ICL geometry and operating conditions yields Pareto-optimal solutions constrained by maximum temperature gradient ( TG cell ) and anode nitridation-risk volume fraction ( V nit ). Compared with the optimized Base design* under the TG cell ≤ 10 K cm −1 constraint, the representative dual-constrained design increases power density from 962.6 to 1251.8 W m −2 and electrical efficiency from 7.8% to 19.6%. Meanwhile, it reduces the predicted V nit to 0% under the adopted thermodynamic criterion. These results indicate that spatially relocating NH 3 decomposition through safety-constrained optimization provides a potential route toward durable and efficient NH 3 -fueled MT-SOFCs.