Crystalline SnS2 accommodates Na ions through intercalation-conversion-alloying (ICA) reactions, exhibiting a natural potential for high energy storage, while its layered structure facilitates rapid charging. However, these intrinsic advantages are not fully realized in practical battery applications. Herein, utilizing an innovative integration of machine-learning-based thermodynamics, artificial-neural-network-assisted molecular dynamics, and density functional theory, specific solvents are demonstrated to effectively tailor the reaction pathways. This strategy not only steers phase transition pathways but also significantly reduces the formation of the solid electrolyte interphase (SEI), which is a common issue in recent battery research. These characteristics of solvents enable reversible ICA reactions and also aid the transformation of microsized SnS2 particles into 3D porous nanostructures with minimal SEI formation. The performance of our Na-SnS2 half-cells achieve 1100 mAh g-1 (97% of the theoretical capacity) at 0.5 C, placing them among the top performers for Na storage. By moving beyond the traditional view of electrolyte solvents as a simple medium for ion transport, this work highlights the critical impact of solvent selection on enabling reversible reactions and morphological transformation of SnS2 anodes with minimal SEI formation and setting benchmarks for anode performance in energy storage systems based on ICA reactions.
Keywords: Na-ion battery; SnS2 anodes; intercalation−conversion−alloying reactions; machine-learning-assisted atomic simulations; porous nanostructure.