The biological brain stores and processes information with exceptional energy efficiency, motivating the search for alternatives to traditional von Neumann computing systems that separate memory from processing. Nanofluidic memristors have emerged as promising building blocks for non-conventional neuromorphic computing, in which ion transport under nanoscale confinement generates history-dependent conductance for analog or digital switching. Their performance is governed by surface chemistry, device geometry and electrolyte properties, which together give rise to diverse memristive mechanisms and hysteresis loop types. Exploring new materials not only improves functionality but also deepens our understanding of the diverse physical mechanisms driving memristive behaviour. Despite rapid progress, challenges remain in mechanisms identification, fabrication scalability, variability control, fine tuning of memristive properties, and integration into higher-order neuromorphic architectures. This Perspective summarizes recent advances in materials and mechanisms, outlines characterization protocols and highlights emerging applications from in-sensor computing to visual processing and reservoir computing, while discussing how tailored nanochannel materials may enable next-generation, brain-inspired iontronic neuromorphic systems. In the brain, memory involves release of neurotransmitters and transport of ions through nanoconfined channels. This Perspective discusses how nanofluidic memristors emulate this confined ion transport, highlighting the materials, design strategies and challenges involved in developing brain-inspired computing technologies.
原文链接
分享
引用
