Self-heating-induced blocking in nanopores enables neuromorphic ionic computing.

Abstract

Inspired by the brain's efficient ionic processing, fluidic memristors using ions as charge carriers have emerged as a promising platform for future neuromorphic systems. To better replicate the brain's dynamic functions, it's essential to explore additional memristive materials and switching mechanisms. In this work, we present a self-heating-induced blocking memristor (SIBM), in which the memory effect arises from thermally triggered reversible precipitation. When subjected to periodic voltage stimulation, SIBM shows threshold-type unipolar resistive switching and features a clear negative differential resistance (NDR). We demonstrate various neuromorphic functions and experimentally validate a fluidic memristor array capable of repeated memory operations. These results provide insights for the design of next-generation neuromorphic fluidic devices.