All‐Optical‐Controlled Excitatory and Inhibitory Synaptic Signaling through Bipolar Photoresponse of an Oxide‐Based Phototransistor

• Published in: Advanced optical materials Vol. 11; no. 14
• Main Authors: Mi, Yen‐Cheng, Yang, Ching‐Hsiang, Shih, Li‐Chung, Chen, Jen‐Sue
• Format: Journal Article
• Language: English
• Published: Weinheim Wiley Subscription Services, Inc 01.07.2023
• Subjects: bipolar photoresponse; Broadband; broadband sensing; Computer vision; Coupling (molecular); Defects; Indium gallium zinc oxide; Materials science; Metal bonding; negative photoresponse; Optics; photosynaptic transistors; phototransistors; Transistors
• ISSN: 2195-1071; 2195-1071
• DOI: 10.1002/adom.202300089

Using light signals for computation and communication is a vital approach for advanced neuromorphic designs. In this study, an all‐optical‐controlled IGZO/ZrOx phototransistor is demonstrated to emulate synaptic functions via both positive and negative photoresponse arisen from the ionization of neutral oxygen vacancies (VO) and metalmetal bonding (MM) defects in IGZO at an illumination with visible light (405 and 520 nm) and near‐infrared light (750, 890, and 980 nm), respectively. With the coupling effect of photogenerated electrons and the charged MM++ defect scattering, the IGZO/ZrOx photosynaptic transistor not only shows broadband photosensing performance but also emulates the excitatory/inhibitory contrasting synaptic functions, such as learning‐ and regulating‐experience behavior of human brain, via applying 405 and 890 nm light pulses, respectively. The all‐optical‐controlled IGZO/ZrOx photosynaptic transistor therefore may convey optical information effectually for the streaming sensor processing in biologically inspired computer vision application. An all‐optical‐controlled IGZO/ZrOx photosynaptic transistor with bipolar photoresponse is demonstrated to emulate the excitatory and inhibitory synaptic functions via the application of 405 and 890 nm light, respectively, including the learning/regulating‐experience behavior and the long‐term potentiation/depression cycles. This work provides insight to develop advanced photonic neuromorphic computing architecture without complicated electronic circuit design.