PMOSFET NBTI (negative-bias temperature instability) measurement using ultra-fast switching method

Negative bias temperature instability was first discovered in 1966. It only became an important reliability issue in silicon integrated circuits because device scaling leads to an increase in gate electric field. Although numerous papers have been published on NBTI, there were many discrepancies bet...

Full description

Saved in:
Bibliographic Details
Main Author: Boo, Ann Ann.
Other Authors: Ang Diing Shenp
Format: Final Year Project
Language:English
Published: 2009
Subjects:
Online Access:http://hdl.handle.net/10356/17896
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Negative bias temperature instability was first discovered in 1966. It only became an important reliability issue in silicon integrated circuits because device scaling leads to an increase in gate electric field. Although numerous papers have been published on NBTI, there were many discrepancies between various models and the experimental results. Recently it has been understood that measurement methods with shorter delay characterize NBTI degradation and recovery better. Hence, researchers have begun to switch from conventional DC measurement to “On-The-Fly” in order to eliminate the recovery effect during the DC measurement. However, there were also several critical issues with the implementation of OTF, i.e. it measures the degradation at the stress level and the degradation of Idlin measured at gate voltages is much higher than at any operation condition. The recovery effect of NBTI could not be studied by employing OTF measurement. The most recent measurement method used to characterize NBTI degradation is the Ultra Fast Switching (UFS) method. UFS gives a better insight of the degradation as it measures the experimental data by injecting extremely short interruptions at the measurement voltage. By doing so, UFS is able to eliminate the recovery effect during measurement and it will not overestimate the shift in threshold voltage as measurement is not taken during stress level and thus, NBTI recovery could be understood better. In this project, NBTI degradation is characterized using UFS at various gate biases and stressing temperatures. The dominant factor which contributes to NBTI degradation at different stressing conditions is studied and analyzed. Besides, the effect of nitrogen which is incorporated in the gate oxide during fabrication to stop the diffusion of boron from the polygate is also studied in this project. As a result, it is found that hole trapping is the dominant contributing factor when nitrogen content is found in the oxide. Based on the Arrhenius plot, there is only one activation energy in the range of stressing temperatures for a pure gate oxide. This is in good agreement with the R-D Model. However, two activation energies are observed for a nitride oxide. Therefore, a conclusion is drawn that besides the dissociation of Si-H bonds, there might be other unknown mechanisms which cause NBTI degradation in a nitride gate oxide.