材料科学
光电子学
铁电性
脉冲激光沉积
晶体管
外延
薄膜
场效应晶体管
无定形固体
纳米技术
电压
图层(电子)
电气工程
电介质
工程类
化学
有机化学
作者
Yooyeon Jo,Ji Young Lee,Eunpyo Park,Hyun Soo Kim,Hyung‐Jin Choi,Seunguk Mun,Yunseok Kim,Sunghoon Hur,Jung Ho Yoon,Ji‐Soo Jang,Chong‐Yun Kang,Seung-Hyub Baek,Jeong Min Baik,Joon Young Kwak,Hyun‐Cheol Song
出处
期刊:ACS applied electronic materials
[American Chemical Society]
日期:2023-08-14
卷期号:5 (8): 4549-4555
被引量:4
标识
DOI:10.1021/acsaelm.3c00691
摘要
Neuromorphic computing systems that mimic the human brain have recently attracted substantial attention because they allow for the efficient processing of large amounts of data. These systems are composed of neurons and synapses to transfer information; synapses play a particularly important role in transmitting and integrating processed signals between the neurons. The ferroelectric field-effect transistors (FeFETs) can meet the required properties of artificial synaptic devices because the channel current can be controlled with changes in applied gate voltage due to two stable polarization states, meaning that the data can be memorized in various states. In this study, the epitaxial Pb(Zr0.20Ti0.80)O3 (PZT) film was grown on La0.67Sr0.33MnO3 (LSMO) buffered SrTiO3 (STO) single crystal substrate using pulsed laser deposition (PLD). As the channel layer, the amorphous indium gallium zinc oxide (a-IGZO) was employed due to its large carrier mobility and good uniformity. The epitaxially grown single-crystal PZT thin film has a residual polarization (Pr) value of 20.2 μC/cm2 and the a-IGZO thin film transistor has a carrier mobility of 10.23 cm2/V s. The biological synaptic behaviors were emulated using the fabricated FeFETs based on the PZT and a-IGZO thin film as a gate insulator and channel material, respectively. The synaptic plasticity was analyzed according to the applied voltage pulse condition. The calculated nonlinearity values were 0.00 and 5.41 with 16 pulse numbers and 0.51 and 7.05 with 32 pulse numbers for potentiation and depression, respectively.
科研通智能强力驱动
Strongly Powered by AbleSci AI