Atomic-Scale Rational Design of Hard Carbon Electrodes for Sodium Behaviors: An Image-Guided Modeling and ReaxFF-MD Study

Hard carbon is a leading anode candidate for sodium-ion batteries (SIBs), yet its complex microstructure has prevented consensus on sodium storage mechanisms, impeding rational electrode design. Herein, we present a computational framework that integrates image-guided atomistic modeling with reactive force field molecular dynamics (ReaxFF-MD) to systematically investigate sodium storage behaviors in hard carbon. Quantitative HRTEM analysis—employing semantic segmentation and Gabor filtering with a data throughput reaching 106—enables the decomposition of carbon structures into curved (amorphous-like) and stacked (crystalline-like) basic structural units, which are then reassembled into three-dimensional atomic models. ReaxFF-MD simulations reveal three sodium storage states: intercalation, surface adsorption, and pore clustering. Crucially, this framework achieves decoupled analysis of three critical structural parameters—crystalline volume fraction (CVF), intracrystallite d-spacing (d002), and crystallite size—thereby overcoming the inherent limitation of trial-and-error approaches in which multiple parameters co-vary with processing conditions. The results demonstrate that (i) higher CVF increases intercalation sites and enhances binding energy; (ii) within the investigated parameter ranges, a favorable d-spacing window exists at 0.37–0.38 nm; and (iii) crystallite size below approximately 20 Å exerts minimal influence. This research technology chain—spanning HRTEM analysis, automated modeling, molecular simulation, and experimental calibration—provides quantitative structure–property relationships that support the rational design of hard carbon anodes for SIBs within the investigated parameter ranges.