A Simple View of a Complex Phenomena
Muhammad A. Alam
Department of Electrical and Computer Engineering
Purdue University, West Lafayette, IN
Negative Bias Temperature Instability (NBTI) is defined as the shift (instability) in threshold voltage of PMOS transistors as a function of stress time. Since NBTI is a 40-year old problem, known from the early days of PMOS-only integrated circuits, one might assume that a thorough review of the classic papers (circa 1970) would be helpful in understanding the phenomena. Indeed early studies established that NBTI increases with temperature and effective fields and requires the presence of holes at the Si/SiO2 interface (hence no NBTI concerns for buried channel PMOS). The degradation appears to be correlated to depassivation of Si-H bonds at the Si/SiO2 interface and is well described by a power-law with time-exponent close to 0.25-0.35 and an activation energy of 0.1-0.2 eV. The studies in 1970s also found that once the stress is turned off, a fraction of the degradation appears to recover. However over the last five years since 2000, as the industry refocused on the NBTI issues and examined the degradation more thoroughly, we gradually discovered the classic conclusions are not completely correct. It is true that NBTI has a power-exponent close to 0.25-0.35, but the exponent is not robust, but rather gradually saturates over time. It is true that NBTI relaxes with time, but the relaxation is frequency independent! And indeed NBTI temperature activation is 0.1-0.2 eV, but on closer inspection the activations appears dispersive, not Arhennius-like as was originally assumed. And finally, although everyone agrees that NBTI reflects depassivation of Si-H bonds, there is no agreement regarding the nature of Hydrogen specie that diffuses through the oxide (e.g. H, H2, or H+) and determines the time exponent. Given all these confusing modifications of the classic observations, i.e. quasi-saturation of time-degradation, dispersion of temperature activation, the frequency-dependence of NBTI relaxation, it is hardly surprising that physicists and engineers working on NBTI issue can no longer agree on a common protocol to extrapolate NBTI lifetime to operating conditions.
Over the last several years, we have developed a simple reformulation of the classical Reaction-Diffusion model (requiring no more than 3-5 lines of algebra!) to show that all the newly discovered features are not isolated physical phenomena at all, but can be consistently interpreted within the Reaction-diffusion (R-D) framework. In fact, the saturation characteristics, dispersive diffusion, frequency-independent degradation are all different facets of the same phenomena. This allows us to address the problem of NBTI lifetime extrapolation to operating conditions in a straightforward manner. In this tutorial, I will discuss the effectiveness of our analytical reformulation of the R-D model in interpreting the non-classical aspects of NBTI degradation and suggest a consistent protocol for NBTI lifetime extrapolation.
Muhammad Ashraful Alam
Muhammad Ashraful Alam is a Professor of Electrical and Computer Engineering at Purdue University where his research and teaching focus on physics, simulation, characterization and technology of classical and novel semiconductor devices. He received a B.S.E.E. degree from Bangladesh University of Engineering and Technology in 1988, the M.S. degree from Clarkson University, Potsdam, NY, in 1991, and the Ph.D. degree from Purdue University, Lafayette, IN, in 1994, all in electrical engineering. From 1995 to 2000, he was with Bell Laboratories, Lucent Technologies, Murray Hill, NJ, as a Member of Technical Staff in the Silicon ULSI Research Department. From 2001 to 2003, he was the Technical Manager of the IC Reliability Group at Agere Systems, Murray Hill, NJ. He joined Purdue University in 2004 and his current research includes theory of oxide reliability, transport in nanocomposite thin film transistors, and nano-bio sensors. Dr. Alam has published over 70 papers in international journals and has presented many invited and contributed talks at international conferences. He received IRPS Best paper award in 2003 and Outstanding paper award in 2001, both for his work on gate oxide reliability. Most recently, he was elected an IEEE Fellow for contribution to physics of CMOS reliability and received IEEE Kiyo Tomiyasu Award for contributions to device technology for communication systems. .