Machine Learning Optimized FeCoMn-Trimetallic MOF-Decorated Nanofibers for Enhanced OER Catalysis

Despite significant advancements in noble metal-free trimetallic MOF-based electrocatalysts for efficient oxygen evolution reaction (OER), limited attention is given to identify which metal will play most significant role in controlling OER performance. Thus, to address this gap, herein ternary metallic (FeCoMn) squarate-based MOF via a solvothermal approach is synthesized. Additionally, machine learning (ML) algorithms are employed on experimental datasets during synthesis strategy to optimize metal concentrations more swiftly and efficiently to design highly efficient ternary metallic (FeCoMn) squarate MOF-based electrocatalysts. Interestingly, ML optimization has identified Fe as a key element significantly influencing OER efficacy. To further boost OER efficacy, ML-optimized FeCoMn MOF is drop-casted onto highly conductive electrospun polycaprolactone (PC) nanofibers, facilitating smooth, uniform flow of ions and electrons across the entire surface, maximizing exposed active sites, all anchored on a sponge-like conductive Ni foam (NF) substrate. Results reveal that ML-optimized FeCoMn/PC displays high electrocatalytic activity with lower overpotential (170 mV at a current density of 10 mA cm⁻²), Tafel slope of 66.6.8 mV dec⁻¹, as compared to FeCoMn (overpotential 180 mV, Tafel slope 89.3 mV dec⁻¹). To the best of knowledge, the first time ML optimized FeCoMn/PC-based electrocatalyst for OER is reported.