Indexed by:Journal paper First Author:Haowen Luo Correspondence Author:Xingsheng Wang Co-author:Ruihan Li,Xiangshui Miao Journal:SCIENCE CHINA Information Sciences Included Journals:SCI、EI Affiliation of Author(s):Huazhong University of Science and Technology Place of Publication:China Discipline:Engineering First-Level Discipline:Electronic Science And Technology Document Type:J Volume:66 Issue:2 Page Number:129402 ISSN No.:1674-733X Key Words:nanosheet transistor, TCAD simulation, interplay, quantum confinement variation, statistical variability DOI number:10.1007/s11432-021-3399-3 Date of Publication:2023-02-01 Impact Factor:7.275 Abstract:In this paper, subject to the strong quantum confinement effect, the variability of 3D sub-5nm nanosheet transistors (NSTs) is comprehensively studied by a TCAD simulation platform, which is strictly calibrated against the experimental data and 2D Poisson-Schrodinger (PS) quantum simulations. Firstly, following the process-induced global variation, the NST channel geometry dependence of the key figures of merit including subthreshold slope, drain-induced barrier lowering, on-state current, off-state current and on-state current to off-state current ratio is presented and analyzed. It is found that the device performance of the nominal transistor can be further improved. Finally, the statistical variability (SV) due to random discrete dopants (RDD), metal gate granularity (MGG), and nanosheet edge roughness (SER) considering the interplay with quantum confinement variation effect, is explored. RDD mainly generated in extension region results in the considerable threshold voltage and large on-state current variations. MGG is still the dominant statistical variability source in NSTs but strong quantum confinement contributes slightly. SER can also cause the large threshold voltage fluctuation for thin NSTs boosted by the quantum confinement variation