Visual growth of nano-HOFs for low‐power memristive spiking neuromorphic system

Abstract

Memristors showcasing progressive conductance modulation are now recognized as one of the most promising candidates to act like electronic synapses inside computer circuits, mimicking the behavior of neurons in the human brain. Recently, the concept of biomimic synapse devices based on organic materials has attracted enormous interest, nevertheless, largely limited by their poor electrical conductivity and charge transfer capacity. Here, 2D ribbon-structured hydrogen-bonded organic frameworks (Nano-HOFs) embedded with the transition metal nanoparticles are firstly grown under visual surveillance to serve as highly reliable memristive materials, where the localized surface plasmon resonance effect and enhanced charge-carrier transport are strongly confirmed. The HOFs@Au-based memristor presents gradient electrical conductances under continuous voltage sweep or pulse algorithms, thereby closely simulating the synaptic behaviors in biological neurons. The newly designed artificial synapses owns an ultra-low areal switching energy density (35.8 fJ µm−2) and maintains extremely stable synaptic functions even after being exposed to ambient conditions over 6 months without any encapsulation. Further brain-inspired cognitive systems are established to realize color coding and image cognition in a secure way. This work creates a precedent for developing highly viable Nano-HOFs materials to manipulate artificial synapse mimicking, prospective neuromorphic computing and intelligent cognition applications.