Electronic Transport and Machine-Learned Modeling of Defect Dynamics in MoS₂ Monolayers for Resistive Switching Devices

Memristors based on 2D materials hold great promise for next-generation neuromorphic computing systems; however, their switching behavior remains poorly understood. We investigate resistive switching in monolayer MoS₂ planar channels where sulfur-vacancy concentration and spatial distribution control conductance. Using density-functional theory combined with nonequilibrium Green’s function (DFT+NEGF), we quantify how vacancy density and configuration modulate electronic transport. Furthermore, we train an equivariant machine-learning interatomic potential with DFT accuracy and employ molecular dynamics (MD) simulations under an applied electric field to simulate vacancy kinetics, field-driven defect reconfiguration, and the resulting evolution of channel conductance and device hysteresis. Our framework links atomistic defect dynamics to electronic transport properties, clarifying the microscopic origins of switching in MoS₂ memristors and providing insight into design rules for stable 2D memristive devices.