Nanofluidic Memristive Transition and Synaptic Emulation in Atomically Thin Pores.

Abstract

Ionic transport across nanochannels is the basis of communications in living organisms, enlightening neuromorphic nanofluidic iontronics. Comparing to the angstrom-scale long biological ionic pathways, it remains a great challenge to achieve nanofluidic memristors at such thinnest limit due to the ambiguous electrical model and interaction process. Here, we report atomically thin memristive nanopores in two-dimensional materials by designing optimized ionic conductance to decouple the memristive, ohmic, and capacitive effects. By conducting different charged iontronics, we realize the reconfigurable memristive transition between nonvolatile-bipolar and volatile-unipolar characteristics, which arises from distinct transport processes governed by energy barriers. Notably, we emulate synaptic functions with ultralow energy consumption of ∼0.546 pJ per spike and reproduce biological learning behaviors. The memristive nanopores are similar to the biosystems in angstrom structure, rich iontronic responses, and millisecond-level operating pulse width, matching the biological potential width. This work provides a new paradigm for boosting brain-inspired nanofluidic devices.