학술논문

Wafer-Scale, Efficient In2S3-Based Optical Memory Devices for Neuromorphic Computing
Document Type
Periodical
Author
Source
IEEE Transactions on Electron Devices IEEE Trans. Electron Devices Electron Devices, IEEE Transactions on. 71(4):2760-2765 Apr, 2024
Subject
Components, Circuits, Devices and Systems
Engineered Materials, Dielectrics and Plasmas
In₂S₃
microfabrication
optical memory devices
PPF index
Language
ISSN
0018-9383
1557-9646
Abstract
This article demonstrates fabrication and testing of novel memory devices using indium sulfide (In2S3) as an active material. A wafer scalable process is used to develop nanostructured In2S3 for the mass production of optical memory devices. The developed devices have shown outstanding light-sensing capability as well as memory functionality. The In2S3 devices are highly responsive toward the 880 nm with higher responsivity and detectivity of 15.88 A/W and $2.84\times 10^{{11}}$ Jones, respectively. In addition, these optical memory devices were able to mimic cognitive functionalities like high paired-pulse facilitation (PPF) index, long-term plasticity (LTP), and short-term plasticity (STP) compared to the state of the art. The optical memory devices have shown a higher PPF index of 227% and long-term memory (LTM) retention capability of $2\times 10^{{3}}$ s. Devices were optically potentiated for broadband spectrum and reset using electrical stimulation. Moreover, In2S3-based devices have highly repeatable, scalable, stable, and reproducible characteristics at room temperature with a rise/fall time of 0.4/0.01 s. The fabricated optical memory devices can be deployed for future neuromorphic computing applications.