Reconfigurable graded adaptive asymmetry-Schottky-barrier phototransistor for artificial visual system with zJ-energy record

Visual perception, memory, and adaptation processes are critical functions in biological systems that enhance responsiveness, improve survival fitness, and reduce information redundancy in complex environments. Therefore, the development of adaptive bionic vision systems with high efficiency, low complexity, and minimal energy consumption has become a key objective. However, most adaptive devices suffer from either complex structures or non-reconfigurable functionalities, hindering the further application for bionic vision systems. Here, for the first time, an asymmetry-Schottky-barrier MoS2 phototransistor is demonstrated for reconfigurable visual system with visual selective memory and graded adaptation functions. More importantly, the device exhibits a new record with the ultra-low energy consumption of ∼90 zJ per synaptic event. Several important adaptive behaviors, such as the sensitivity, desensitization, accuracy, and self-recovery, are successfully realized and adjusted by asymmetry-Schottky-barriers. These results pave a new way toward the efficient, low-energy, and reconfigurable bionic visual systems for applications of machine vision, bionic robotics, and human-machine interfaces.