Intrinsically flexible multimode reconfigurable transistors for polymorphic circuits and neuromorphic devices

Abstract

Abstract Reconfigurable transistors hold considerable significance for both integrated circuits and neuromorphic electronics. At the same time, flexible electronics is developing rapidly, promoted by various emerging applications, such as wearable electronics and smart robots. Therefore, flexible reconfigurable transistors are highly expected to create emerging application scenarios while they are seldom reported. Here, intrinsically flexible multimode reconfigurable transistors (IFMRTs) with dual-gate structure are proposed and realized. Either top or bottom gate can serve as the modulation terminal to switch the transistor among p-type, n-type, and ambipolar modes. The threshold voltage of the reconfigurable transistor is modulable. Based on IFMRTs, we have demonstrated inverters and polymorphic logic circuits with key bit-selectable NAND or NOR logic functions, which provides solutions for hardware security. We have also demonstrated artificial heterosynapse and dendrite, achieving reconfigurable synaptic responses and dendrite integration. Taking the simulation of automatic obstacle avoidance and coordination control of hand movement by low-level and high-level neural centers as examples, we illustrate the potential intelligence of IFMRTs in robotic decision-making and arm control. IFMRTs maintain their reconfigurability even after 5000 bending cycles at 4 mm radius. The design and demonstration of IFMRTs open up possibilities for flexible wearable electronics and soft intelligent robots.