杰纳斯
单层
非易失性存储器
材料科学
光电子学
铁电性
超短脉冲
极化(电化学)
电介质
晶体管
神经形态工程学
铁电电容器
纳米技术
实现(概率)
功率消耗
栅极电介质
电压
逻辑门
切换时间
光子学
高-κ电介质
作者
Ngoc Thanh Duong,Swathi Krishna,Minh Chien Nguyen,Hang Sik Kim,Vu Tien Lam,Hamza Zad Gul,Soo Ho Choi,Minjeong Kim,D.U Lee,Thi Uyen Tran,Vu Khac Dat,Van Dam Do,Van Huy Hoang,Heesuk Rho,Yong Wook Kim,Woo Jong Yu,Ki Kang Kim,Seong Chu Lim
标识
DOI:10.1038/s41467-026-73279-7
摘要
The discovery of two-dimensional (2D) semiconducting ferroelectric (SFe) materials marks a pivotal step toward the realization of ferroelectric random-access memory. Despite their potential as ideal platforms for ultrafast switching and ultralow power consumption, SFe monolayers remain largely unexplored. In this work, we demonstrate field-effect transistor employing monolayer Janus MoSeS, which exhibits switchable polarization when integrated on an amorphous ZrO2 layer with a high dielectric constant. The non-centrosymmetric structure of MoSeS enables gate-switchable spontaneous out-of-plane polarization, attributed to atomic displacements. The device demonstrates counterclockwise (CCW) memory windows of ~5 V, an on/off ratio exceeding 106 at zero gate bias, and a retention time much longer than 10,000 s. Furthermore, it achieves fatigue-free endurance over 107 cycles and ultrafast responses to 16.5 ns voltage spikes, resulting in ultralow power consumption of 113.9 fJ for programming and 89.8 fJ for erasing in each cycle. These performances establish Janus monolayers as promising candidates for next-generation nonvolatile memory and neuromorphic computing applications. Further work will be necessary to establish the switching mechanisms of the devices. The authors report the realization of switchable polarization field-effect transistors based on monolayer Janus MoSeS, showing on/off ratios exceeding 106 at zero gate bias, retention times up to 10,000 seconds, endurance of over 107 cycles, programming response times down to 16.5 ns and power consumption down to 113.9/89.8 fJ per program/erase cycle.
科研通智能强力驱动
Strongly Powered by AbleSci AI