Characterization of Random Telegraph Noise in Scaled High-κ/Metal-Gate MOSFETs with SiO2/HfO2 Gate Dielectrics

摘要

In the paper, random telegraph noise (RTN) in high-κ/metal-gate MOSFETs is investigated. The RTN in high-κ MOSFETs is found different compared to that in SiON MOSFETs, and faces challenges in characterization. Therefore, the characterization method is improved based on clustering and Hidden Markov Model, which greatly enhances the ability to extract RTN with non-negligible “ghost noise” in high-κ MOSFETs. The RTN signal and “ghost noise” in devices fabricated by two SiO2/HfO2 stack processes with two different formation methods are compared. It is found that the real RTN signal in SiO2/HfO2 MOSFETs originates from the oxide defects in the HfO2 layer, while the “ghost noise” originates from the SiO2 interfacial layer and has strong dependence on the quality and formation process of interfacial layer.

出版物
In ECS Transactions
李萌
李萌
助理教授、研究员、博雅青年学者

李萌,北京大学人工智能研究院和集成电路双聘助理教授、研究员、博雅青年学者。他的研究兴趣集中于高效、安全的多模态人工智能加速算法和芯片,旨在通过算法到芯片的跨层次协同设计和优化,为人工智能构建高能效、高可靠、高安全的算力基础。

var dimensionValue = 'SOME_DIMENSION_VALUE'; ga('set', 'dimension1', dimensionValue);