A comprehensive model for oxide degradation

https://doi.org/10.1016/j.microrel.2004.11.032Get rights and content

Abstract

In this work we present two analytical (and physically supported) models to describe trap kinetics in both thin and ultra-thin SiO2 films. The models are based on the different mechanism controlling the carrier transport through the oxides and on the assumption of a two step process for creating stable traps, through defect precursors. Experimental data of stress induced leakage current confirm the validity of models predictions. Furthermore, a systematic study of the transient of trap kinetics experimentally demonstrates the existence of defect precursors as well as a reduction of oxide damage under pulsed stress condition respect to DC case.

Introduction

Scaling of CMOS devices, as specified by the ITRS, makes understanding of process involved in oxide damage of crucial importance for getting better performances.

For this reason, a lot of studies have been devoted toward the physics of oxide degradation. However, the debate on which is the main damage responsible (the electric field [1] or the carrier energy [2]) is still open.

Here, we propose two analytical models, physically based, able to predict stress induced leakage current (SILC) in thin and ultra-thin (tOX < 4 nm approximately) oxides.

Section snippets

Degradation models

In this section, two analytical models describing trap kinetics in thin and ultra-thin SiO2 films under DC electrical stress are presented. In both cases, defect creation is thought to be ruled by hot carrier energy release, which causes SiH bond breaking [2]. However, the mechanism of energy release changes with oxide thickness. This difference is taken into account writing two different rate equations for thin and ultra-thin oxides degradation kinetics.

Transient of trap kinetics

Degradation models presented in Section 2 are valid under steady state stress conditions and cannot take into account the transient behaviour of trap kinetics. This transient can be related to a phenomenological picture based on the hypothesis that the defect creation is a two-step process [6]. During an electrical stress, at first defect precursors are formed in the oxide, which represent unstable energetic states of the network. They are converted in stable traps if the external perturbation

Conclusions

In this work we have presented a comprehensive, physically based, analytical model for both thin and ultra-thin oxides degradation kinetics, which have been experimentally validated. Furthermore, starting from the microscopic theory of defect precursors, we have demonstrated that the pulsed stress with an appropriate waveform duty cycle can strongly reduce oxide damage respect to DC case.

References (8)

  • J.W. McPherson et al.

    Underlying physics of thermochemical E model in describing low field time dependent dielectric breakdown in SiO2 thin films

    J Appl Phys

    (1998)
  • D.J. DiMaria

    Electron energy dependence of metal–oxide–semiconductor degradation

    Appl Phys Lett

    (1999)
  • F. Irrera

    Degradation kinetics of stressed oxides

    Appl Phys Lett

    (2001)
  • B. Riccò et al.

    Modeling and simulation of stress induced leakage current in ultrathin SiO2 films

    IEEE Trans Electron Devices

    (1998)
There are more references available in the full text version of this article.

Cited by (2)

  • 3D KMC simulations of crater growth during the reduction of oxide nanoislands on metal surfaces

    2011, Surface Science
    Citation Excerpt :

    Oxide reduction is also frequently involved in hydrometallurgical processes for recovering pure metals from their native oxides [8]. Other processes of oxide reduction include fabrication of electronic devices, magnetic memory components and active/passive solar materials systems [9–14]. Traditionally, the reduction process of metal oxides has been described using phenomenological kinetic models.

  • Reduction of CuO nanowires confined by a nano test tube

    2014, RSC Advances
View full text
Copyright © 2004 Elsevier Ltd. All rights reserved.