Organic Synaptic Perovskite Transistors Based on Patternable Fluorinated Block Copolymers for Reconfigurable Neuromorphic Computing and Logic Operations

ABSTRACT Artificial organic synaptic transistors integrating sensing, memory, and learning capabilities have great potential for application in emerging neuromorphic intelligent vision systems. Herein, we propose an organic synaptic perovskite transistor (OSPT) based on fluorinated block copolymer, i.e. poly(heptafluorobutyl methacrylate)‐ block ‐poly(2‐vinylpyridine) (PHFBMA‐ b ‐P2VP), that exhibits light‐electric dual modal modulation for reconfigurable neuromorphic computing and logic integration. Rapid self‐assembly of PHFBMA‐ b ‐P2VP at 80°C endows MAPbBr 3 quantum dots with the advantage of being templated in P2VP domains to form ordered nanocomposite floating gate at low temperature, which not only prevents the loss of captured charges, but also avoids the damage to device performance caused by perovskite quenching due to high temperature. The OSPT exhibits multilevel memory, broadband response from 280 nm to 740 nm, and extremely low power consumption of 1.92 fJ. Furthermore, the device can execute optoelectronic information memory and simulate a series of synaptic functions. Notably, the MNIST pattern recognition based on the OSPT reveals a high recognition accuracy of 94.8%. Meanwhile, this device is also capable of performing manipulated optoelectronic logic operation functions of “AND” and “OR”. This work advances the development of versatile integrated neuromorphic intelligent vision systems.