Gradient porous carbon has become a potential electrode material for energy storage devices, including the aqueous zinc-ion hybrid capacitor (ZIHC). Compared with the sufficient studies on the fabrication of ZIHCs with high electrochemical performance, there is still lack of in-depth understanding of the underlying mechanisms of gradient porous structure for energy storage, especially the synergistic effect of ultramicropores (<1 nm) and micropores (1-2 nm). Here, we report a design principle for the gradient porous carbon structure used for ZIHC based on the data-mining machine learning (ML) method. It is clarified that the combination of 0.6-0.9 nm ultramicropore and 1.6 nm micropore achieves the highest specific capacity. Molecular dynamic simulation was further employed to investigate the electric double-layer structures in several kinds of electrified gradient porous carbon electrode/electrolyte interface. It is found that the Zn2+ ions in the 1.6 nm micropore balance the most charges of the electrode surface as the counterion with the modification of the solvation structure. Furthermore, the ML-based force field is trained and employed in the simulation of the ion charging dynamic in the gradient porous carbon electrode. Based on the free energy profile result, the remarkable benefit of the step-by-step desolvation process is found in the 0.86 and 1.6 nm gradient porous structure, which could be the origin of the enhanced ion charging dynamic and better capacity retention performance.
Keywords: electrochemical interface; machine learning; molecular dynamic; porous carbon; zinc-ion hybrid capacitor.