Computational Framework for Hydrogen-Bond–Mediated Proton Transport in Confined Water Systems for Neuromorphic Materials

This record documents a computational study of hydrogen-bond–mediated proton transport in confined water systems, developed as part of the project CIS240528: Molecular Dynamics Modeling of Building Elements Based on Water Hydrogen Bonds for Molecular Artificial Intelligence. The work establishes a reproducible molecular-modeling framework for analyzing how nanoscale confinement geometry controls hydrogen-bond network topology and, consequently, proton-conduction pathways relevant to neuromorphic materials. Using a combination of classical molecular dynamics (MD) simulations and quantum-chemical calculations, the study investigates water confined in carbon-based nanostructures, including graphene sheets, nanoribbons, nanotubes, nanopores, and multilayer membrane assemblies. Water molecules are treated explicitly, while carbon frameworks serve as rigid or semi-rigid confinement scaffolds. Both periodic and finite boundary conditions are employed to model extended nanochannels, junctions, gated pores, and reservoir-like geometries. The analysis focuses on the topology of hydrogen-bond networks rather than on explicit charge-transport simulations. Structural motifs such as single-file water wires, parallel chains, branching points, junctions, constrictions, and gating elements are identified and used as proxies for Grotthuss-type proton transfer pathways. Systematic variations in channel width, connectivity, hydration level, and geometric perturbations reveal conditions under which proton-conduction pathways emerge, switch, compete, or break down. The accompanying figure set (Figures 1–55) provides a topology-based design map linking confinement geometry to emergent proton-transport functionality. These configurations illustrate key neuromorphic-relevant behaviors, including directional transport, gating, branching, competition between pathways, and reservoir-like dynamics arising from coupled nanochannel networks. This dataset and report are intended to support further research on ion–proton neuromorphic matter, hydrogen-bond networks, and molecular-scale artificial intelligence architectures, as well as to serve as a reference framework for future large-scale, high-throughput simulations and experimental comparison.