Machine-Learning-Guided Multiscale Design of Nitrogen-Doped Carbon for Enhanced Supercapacitor Performance

Hierarchical structural control of nanomaterials across multilength scales is critical for optimizing mass and charge transport in energy-related applications. In this study, we combine machine learning analysis and rational synthesis to design nitrogen-doped carbonaceous hybrid materials with multiscale architectures tailored for supercapacitor electrodes. Data-driven insights from the literature reveal that, in addition to surface area, the presence of pyrrolic nitrogen strongly correlates with enhanced capacitance. Guided by this finding, we synthesized nanohybrids through the heterogeneous self-assembly of dopamine and block copolymers on graphene oxide (GO) nanosheets, followed by polymerization and thermal carbonization. Dopamine serves as both the carbon and nitrogen sources and facilitates the formation of mesoporous structures. The resulting hybrids feature atomic-scale nitrogen doping (10–10 m), mesoporous carbon networks (10–8 m), and macroscopic sheet-like morphology (10–6 m) with tunable thickness. Uniform coating of mesoporous N-doped carbon on both sides of the reduced GO increases the accessible surface area while maintaining ion mobility. This work demonstrates a data-informed strategy for designing hierarchical carbon-based materials and highlights the synergy between machine learning and multiscale structural engineering for high-performance energy storage applications.