RTN and BTI in nanoscale MOSFETs: A comprehensive statistical simulation study

• Published in: Solid-state electronics 2013-06, Vol.84, p.120-126
• Main Authors: Amoroso, Salvatore Maria, Gerrer, Louis, Markov, Stanislav, Adamu-Lema, Fikru, Asenov, Asen
• Format: Article
• Language: English
• Subjects: Applied sciences; BTI; Circuit design; Device Simulation; Electronics; Exact sciences and technology; Molecular electronics, nanoelectronics; Nanoscale MOSFET devices; Reliability; RTN; Semiconductor electronics. Microelectronics. Optoelectronics. Solid state devices; Statistical Variability; Transistors
• ISSN: 0038-1101; 1879-2405
• DOI: 10.1016/j.sse.2013.02.016

► Random telegraph noise and bias temperature instabilities in nanoscale MOSFETs have been investigated using 3D TCAD ‘atomistic’ simulations. ► The statistical distributions of charge-trapping time constants and noise amplitudes are evaluated. ► The effects of 3D electrostatics, random dopant fluctuations and many-traps interactions on reliability figures of merit are studied. ► The absence of correlation between random telegraph noise and bias temperature instabilities is highlighted. This paper presents a thorough statistical investigation of random telegraph noise (RTN) and bias temperature instabilities (BTIs) in nanoscale MOSFETs. By means of 3D TCAD ‘atomistic’ simulations, we evaluate the statistical distribution in capture/emission time constants and in threshold voltage shift (ΔVT) amplitudes due to single trapped charge, comparing its impact on RTN and BTI. Our analysis shows that the individual BTI ΔVT steps are distributed identically as the RTN ΔVT steps. However, the individual traps in a device cannot be considered as uncorrelated sources of noise because their mutual interaction is fundamental in determining the dispersion of capture/emission time constants in BTI simulation. Further, we show that devices strongly affected by RTN are not necessarily strongly affected by BTI (and vice versa), revealing the uncorrelated nature of these two reliability issues. The presented results are of utmost importance for profoundly understanding the differences and similarities in the statistical behavior of RTN and BTI phenomena and assisting a reliability-aware circuit design.