Synergistic Enhancement of Water-Splitting Performance Using MOF-Derived Ceria-Modified g-C3N4 Nanocomposites: Synthesis, Performance Evaluation, and Stability Prediction with Machine Learning

Diminishing the charge recombination rate by improving the photoelectrochemical (PEC) performance of graphitic carbon nitride (g-C₃N₄) is essential for better water oxidation. In this concern, this research explores the competent approach to enhance the PEC performance of g-C₃N₄ nanosheets (NSs), creating their nanocomposites (NCs) with metal-organic framework (MOF)-derived porous CeO₂ nanobars (NBs) along with ZnO nanorods (NRs) and TiO₂ nanoparticles (NPs). The synthesis involved preparing CeO₂ NBs and g-C₃N₄ NSs through the calcination of respective precursors, while the sol-gel method is employed for ZnO NRs and TiO₂ NPs. Following the subsequent analysis of the physicochemical properties of the materials, the binder-free brush-coating method is deployed to fabricate NC-based photoanodes, followed by an evaluation of the PEC performance through various electrochemical techniques. Remarkably, the binary g-C₃N₄/CeO₂ NCs with 20 wt % CeO₂ NBs (gC20 NCs) exhibited a significantly enhanced current density of 0.460 mA/cm² at 1.23 V vs reversible hydrogen electrode, which is 2.3 times greater than that of bare g-C₃N₄ NSs (0.195 mA/cm²). Further improvements are observed with ternary gC20/TiO₂ (gCT50) and gC20/ZnO (gCZ50) NCs, achieving current densities of 1.810 and 1.440 mA/cm², respectively. These enhanced current densities are attributed to increased donor densities, reduced charge transfer resistances, and efficient charge transport within the NCs. In addition, higher surface areas with beneficial instinctive defects are perceived for gCT50 and gCZ50 NCs, as revealed by Brunauer-Emmett-Teller and electron spin resonance analysis. Finally, the stability of gCZ50 and gCT50 NC-based photoanodes is predicted and forecasted with the help of the recurrent neural network-based long short-term memory technique. Overall, this study demonstrates the efficacy of organic-inorganic hybrids for efficient photoanodes, facilitating advancements in water-splitting studies.