Zinc oxide thin film transistor with high UV photoelectric sensitivity for artificial neuro networks

Abstract

Photosynaptic transistors based on amorphous oxide semiconductors are a potential device to break von Neumann bottleneck due to their low consumption and integration of sensing, storage, and computing. Till now, there has been a lack of studies on the photosynaptic transistors based on zinc oxide (ZnO) under two dimensional optoelectronic controls. In this work, through size modulation, high-performance and back-end of line compatible 16-nm-thick ZnO thin film transistors (TFTs) by atomic layer deposition is fabricated with channel length and width of 10 and 30 μm, respectively. The device possesses outstanding electrical and photoelectric properties with a subthreshold swing of 273 mV/dec, a mobility of 28.0 cm² (V s) ¹, an on/off current ratio of 3.7 × 10⁸, high responsivity of 2.9 × 10⁵ A/W, and photo-to-dark-current ratio of 2.3 × 10¹⁰%. Moreover, three different trends of paired-pulse ratios under different V_g are exhibited with illustrations. This work demonstrates the potential of scaled down ZnO TFTs for the artificial neural networks.