Machine-Learning-Assisted Synthesis of Bimetallic Metal–Organic Frameworks for the Optimized Oxygen Evolution Reaction

Although a wide range of bimetallic metal-organic frameworks (MOFs) have been reported as electrocatalysts for the oxygen evolution reaction (OER), there is still a need for precise tuning of the metal precursors and composite ratios to minimize the overpotential and design highly efficient electrocatalysts. To achieve this, we applied machine learning (ML) algorithms to optimize the metal precursor and composite ratios, identifying the key factors that govern the OER performance. We first synthesized a bimetallic FeCo squarate-based MOF (FeCo-Sq MOF) using a solvothermal method and then optimized the metal precursor ratios using ML algorithms to achieve a low overpotential. To further enhance the OER efficacy, the ML-optimized FeCo-Sq MOF was coated with S-doped graphitic carbon nitride (SCN) and wrapped with polydopamine (PDA). The PDA wrapping not only increased the number of binding/adsorption sites for -OH but also enhanced the stability, charge/electron transfer kinetics, and effective anchoring of SCN on the MOF surface. To obtain optimal OER catalysts, the SCN loading was further fine-tuned through ML. The ML-optimized PDA-SCN@FeCo-Sq MOF exhibited high electrocatalytic performance, achieving a low overpotential of 310 mV and a Tafel slope of 56 mV/dec at a current density of 10 mA cm^-2 in 1 M KOH. This study presents a promising ML-assisted strategy for designing high-performance PDA-SCN@FeCo-Sq MOF electrocatalysts for efficient water splitting.