Fibrous Organic Electrochemical Transistors for Neuromorphic Computing with Enhanced Cyclic Stability
발표자
()
초록
내용
Up to date, synaptic functionalities and simple simulation for learning have been demonstrated with organic electrochemical transistors (OECTs). However, the cyclability of the OECTs such as cyclic endurance which is requirement for iterative neuromorphic computing has not been studied. The representative OECT system based on polythiophene (PT) cannot be reversibly recovered to the initial state after repeated erased/programmed process due to residual of anions in the film which cause unstable cyclability. Here, we suggested carboxylic acid-fucntionalized-PT (P3CT) which has deprotonating group to balance charge for reversible redox reactions in the film, leading to the enhanced cyclability of the device. We also demonstrated a new structure of fiber-shaped OECT (neurofiber) with multiple junctions of synapses for artificial neural network. Finally, we successfully exploited neurofiber to the speech recognition as an iterative learning simulation with 88.9 % classification accuracy.