Characterization and modeling of dynamic variability induced by BTI in nano-scaled transistors

• Published in: Microelectronics and reliability Vol. 80; pp. 100 - 108
• Main Authors: Garros, Xavier, Laurent, Antoine, Subirats, Alexandre, Federspiel, X., Vincent, E., Reimbold, Gilles
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-01-2018
• Subjects: BTI; Defect Centric Model; Dynamic variability; Finfet; Monte Carlo simulation; Nanowire; Nanowire; Defect Centric Model; Finfet; BTI; Dynamic variability; Monte Carlo simulation
• ISSN: 0026-2714; 1872-941X
• DOI: 10.1016/j.microrel.2017.11.025

In this paper, dynamic variability (DV) induced by BTI is deeply investigated in nano-scaled devices by means of statistical measurements and modeling. The impact of a single charge q on Vt is first investigated through 3D electrostatic simulations. In planar devices, this MC modeling allows proving that the average Vt shift induced by a single q denoted ηt is inversely proportional to the device area. In trigate 3D transistors, BTI trapping not only occurs at the top surface (TS) oxide but also at the device sidewalls (SW). For Πfet Nanowire, this implies that ηt exhibits a complex variation with device scaling unlike in planar structures. In contrast, Finfet rather behaves as a vertical planar device for which SW plays now the role of TS. Finally the impact of device scaling on NBTI degradation is thoroughly studied in 3D technologies. Enhanced NBTI is measured on narrower devices. This phenomenon is well explained and reproduced by 3D MC simulations considering a poorer quality of the SW gate oxide with respect to its TS counterpart.
Impact of a single charge on Vt of Nanowire Πfet & Finfet from 3D electrostatic simulations. [Display omitted]
•Characterization and modeling of dynamic variability induced by BTI in both 2D planar and 3D transistors•3D electrostatics simulation to model the impact of single charge on threshold voltage•Establishing scaling affecting dynamic variability in 2D and 3D technologies•New insight on geometry dependence of NBTI from Monte Carlo modeling