PBTI under dynamic stress: From a single defect point of view

In this paper, fundamental aspects of the Bias Temperature Instability (BTI) in FETs with metal gate/high-k (HKMG) gate stacks are discussed from a single defect point of view. First, Random Telegraph Noise (RTN) measurements are used to show that the capture/emission processes of individual defects...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhao, K, Stathis, J H, Linder, B P, Cartier, E, Kerber, A
Format: Conference Proceeding
Language:English
Subjects:
Bias Temperature Instability
Charge carrier processes
duty cycle dependence
FETs
frequency dependence
Frequency measurement
HKMG
Logic gates
NBTI
PBTI
RTN
single defects
Stress
Temperature
Temperature dependence
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, fundamental aspects of the Bias Temperature Instability (BTI) in FETs with metal gate/high-k (HKMG) gate stacks are discussed from a single defect point of view. First, Random Telegraph Noise (RTN) measurements are used to show that the capture/emission processes of individual defects in highly scaled HKMG FETs exhibit very similar Poisson statistics and can be fully characterized by a characteristic electron/hole capture, τ c , and emission time, τ e , in NFET/PFET. In all cases, capture and emission are found to be thermally activated. These observations suggest that NBTI and PBTI share similar microscopic trapping/de-trapping mechanism, for holes and electrons, respectively. Based on these findings, a simple physical model is introduced which describes the behavior of a distribution of identical defects (characterized by τ c and τ e ) but provides deep insights into the BTI dynamics under AC stress in general. The occupancy level of identical defects at equilibrium is found to becomes frequency, f, independent for f ≫ [1/τ c , 1/τ e ], such that the BTI behavior at operation conditions (~GHz) can be measured at relatively low frequencies (in the kHz range). The single defect model was then expanded to predict the macroscopic BTI behaviors in NMOS devices for arbitrary stress conditions. Excellent agreement between model prediction and experimental data is demonstrated, confirming that PBTI in HKMG gate stacks can be understood as a superposition of trapping/de-trapping events from individual defects in the gate stack. The overall dynamics of PBTI is thus largely governed by the distribution of electron capture and emission times of the defects in the gate stack. The challenges for using a capture and emission time based model for product lifetime predictions are addressed.
ISSN:1541-7026
1938-1891
DOI:10.1109/IRPS.2011.5784502