Solution-Processed Multiterminal Artificial Synapses Based on Ion-Doped Solid Electrolytes

Abstract

Brain-inspired neuromorphic computing is a revolutionary technology to create an intelligent and energy-efficient computing system. However, neuromorphic devices fabricated with complementary metal-oxide-semiconductor technology have a few drawbacks such as high energy consumption and circuit complexity. Also, such devices have different mechanisms compared to biological synapses, which transmit signals using ion dynamics. Here, we propose a three-terminal artificial synapse based on ion movement. The synapse uses solution-processed reduced graphene oxide (rGO) and ion-doped solid polymer electrolyte. The rGO-based artificial synapse shows synaptic behaviors including excitatory postsynaptic current and paired-pulse facilitation as short-term plasticity. The artificial synapse emulates the transition characteristics of a biological synapse from short-term to long-term plasticity depending on the amplitudes of the applied bias pulses. These results demonstrate the feasibility of using rGO and ion-doped solid electrolyte in construction of neuromorphic computing systems.