Competitive Multi-Ion Transport in Sub-Nanopores Enables Memristive Behaviors and Synaptic Plasticity

In fluidic environments, the diversity in ionic species and valence states provides charge carriers with high information density for emulating synaptic plasticity. Here, we reported a nanofluidic device based on a monolayer graphene sub-nanopore array. In a multi-ion system, tuning the pore diameter significantly modulates the free energy barrier of specific ions, enabling the modulation of dynamic ion competition and realizing multiple computational functions. Notably, this device originates from a new mechanism in which energy barriers or ion pairs temporarily block the pore, thereby modulating the ionic conductance. Specifically, at a pore diameter of 4 Å, the accumulation of Ca2+ ions at the pore entrance induced steric hindrance, resulting in voltage-gated switching and memristive hysteresis. Expanding the pore to 6 Å enabled voltage-driven dynamic ion sieving, which facilitated the construction of a fluidic ternary logic gate that effectively moves beyond binary limitations. At a diameter of 9 Å, the formation of ion pairs in the nanopore blocked the transport of other ions, thereby inducing the negative differential resistance (NDR) effect. Furthermore, based on the dynamic blockage and relaxation-induced dissociation of ion pairs under pulsed stimulation, the device mimicked the "inactivation-reset" kinetics of biological voltage-gated Ca2+ channels and emulates synaptic weight depression. Notably, the NDR effect allows the device to achieve self-sustained electrical oscillations, emulating the continuous spiking signals of biological neurons. This work establishes a theoretical foundation for the development of multifunctional fluidic neuromorphic computing systems based on competitive multi-ion transport.