Multifunctional Optoelectronic Logic Gates and Artificial Synapses Based on Coexistent Positive and Negative Photoconductivity SnSe2 Photodetectors.

To address the computational bottlenecks of the von Neumann architecture in the era of big data, the development of an optoelectronic computing platform, which combines high speed and low power consumption, has emerged as a viable solution for processing high-density information. In this study, the logic computation and synaptic sensing integrated photodetector is fabricated through a mechanically exfoliated SnSe2 film. The coexistent positive and negative photoconductive behavior of the device is exhibited over a wide spectral range under an ultralow bias voltage of 1 mV. The six fundamental optoelectronic logic gates (YES, AND, OR, NOT, NAND, and NOR) are successfully demonstrated, where the output current of the logic gates is modulated under different light intensities. The logic gate for a 1 × 4 array is tested, where the logic state stability is up to 100%, and logic imaging of the device can be performed. Due to the excellent retention time (>30 s) of the nonvolatile photoconductivity phenomenon, the synaptic behavior is effectively emulated, such as the transition from short-term memory to long-term memory, and thereby, memory learning and associative learning capabilities are validated. In the work, the SnSe2 device is successfully implemented in a highly integrated way with optoelectronic logic functions and provides a physical foundation for bionic intelligent visual systems by integrating optical pulse computation with brain-like visual perception.