材料科学
光电子学
铁电性
晶体管
神经形态工程学
非易失性存储器
极化(电化学)
阈值电压
氮化镓
电压
氧化物
场效应晶体管
纳米技术
铟
硅
纤锌矿晶体结构
电子工程
泄漏(经济)
半导体
接口(物质)
氮化物
逻辑门
随机存取
作者
Kai Wang,Tingting Gao,胡凯生,Suyuan Cai,Huibo Liu,Haotian Liang,Jianye Zhang,Jiayao Hao,Chenchen Ye,Zhiqiang Wang,Yang Chai,Xuefei Li
标识
DOI:10.1038/s41467-026-79253-7
摘要
Wurtzite-structured aluminum scandium nitride (AlScN) ferroelectric materials are potential candidates for non-volatile memory (NVM) applications because of their high remanent polarization (Pr > 100 μC·cm−2) and compatibility with back-end-of-line (BEOL) processes. However, achieving high-performance wurtzite ferroelectric field-effect transistors (FeFETs) remains a significant challenge because of their high interface trap density (Dit) and large threshold voltage (VTH) variation. Here we show the distinct polarization switching mechanisms of AlScN FeFETs based on an ultrathin indium gallium oxide (IGO) channel with three different structures. FeFETs with a metal–ferroelectric–metal–insulator–semiconductor (MFMIS) structure show high interface quality with low Dit and a stable VTH, achieving a large memory window (MW) of up to 20.6 V and a high endurance of 106. The stable 6-bit operation of our devices with 64 distinct states shows the feasibility of multilevel storage applications. Furthermore, we demonstrate the applicability of the FeFET in neuromorphic networks, with a high recognition accuracy of 93.02%. AlScN field-effect transistors for non-volatile memory are limited by high interface trap density and threshold voltage instability. Wang et al. demonstrate annealing-free AlScN ferroelectric transistors with low interface trap density and stable threshold voltage, enabling a large memory window and robust 6-bit operation.
科研通智能强力驱动
Strongly Powered by AbleSci AI