Nanoconfinement effect holds significant research implications across multiple disciplines, with interfacial interactions—particularly at solid–water interfaces—playing a central role in the field of confined spaces. The semi-disordered nature of solid–liquid interfaces under confinement critically influences various dynamic processes, such as crystallization, electrochemistry, and catalysis. Building on these insights, we employed ab initio molecular dynamics simulations to systematically investigate interfacial reactions in graphene-confined environments, with a focus on the water–CaCl semi-disordered interface to elucidate the underlying atomic-scale mechanisms. In particular, the non-uniform electronic density distribution at the CaCl semi-disordered interface governs the spatial arrangement, structural ordering, and stability of interfacial water molecules. Furthermore, our simulations revealed that strong Coulomb interactions drive pronounced proton transfer and peculiar hydrogen evolution reactions at the interface. Finally, combining machine learning techniques, we developed a set of potential functions with density functional theory accuracy to describe Ca–Cl systems with unconventional stoichiometry. These functions will support in-depth research on Ca–Cl systems with unconventional stoichiometry. These findings provide fundamental insights into interfacial phenomena under confinement and offer critical implications for the design of energy storage systems, batteries, and iontronic devices.