材料科学
光电子学
晶体管
突触重量
信道长度调制
排水诱导屏障降低
电压
电气工程
阈值电压
计算机科学
人工神经网络
机器学习
工程类
作者
Keonwon Beom,Paul Yang,Daehoon Park,Hyung Jun Kim,Hyun Ho Lee,Chi Jung Kang,Tae‐Sik Yoon
出处
期刊:Nanotechnology
[IOP Publishing]
日期:2018-10-16
卷期号:30 (2): 025203-025203
被引量:28
标识
DOI:10.1088/1361-6528/aae8d2
摘要
We demonstrate single- and double-gate synaptic operations of a thin-film transistor (TFT) with double-gate stack consisting of an Al-top-gate/SiO x /TaO x /n-IGZO on a SiO2/n+-Si-bottom-gate substrate. This synaptic TFT exhibits a tunable drain current, mimicking synaptic weight modulation in the biological synapse, upon repeatedly applying gate and drain voltages. The drain current modulation features are analog, voltage-polarity dependently reversible, and strong with a dynamic range of multiple orders of magnitude (∼104). These features occur as a consequence of the changes in mobility of the IGZO channel, gate insulator capacitance, and threshold voltage. The drain current modulation responsive to the timing of the voltage application emulates synaptic potentiation, depression, paired-pulse facilitation, and memory transition behaviors depending on the voltage pulse amplitude, width, repetition number, and interval between pulses. The synaptic motions can be realized also by a double-gate operation that separately tunes the channel conductance by top-gate biasing and senses it by bottom-gate biasing. It provides the modulated synaptic weight with a wide level of synaptic weight through the read condition using a bottom-gate stack without read-disturbance. These results verify the potential application of TaO x /IGZO TFT with single- and double-gate operations to artificial synaptic devices.
科研通智能强力驱动
Strongly Powered by AbleSci AI