It seems to be general agreement in relating the dielectric breakdown of thin S i 0 2 films to the previous degradation of its network by external stress. In a recently presented model we developped the idea of relating the breakdown with the generation of a critical density of electron traps [1,2]. These idea has also been proposed by the group of the Hebrew University of Jerusalem after extensive studies of the dynamics of generation and occupation of neutral trapping sites (NTS) in S i 0 2 films subjected to high-field injection [3,4]. For the moment, the relation between the generation of NTS and the breakdown has been treated from a deterministic point of view since the breakdown condition was related to an average density of generated defects ( 1 . 2 1 . In this paper, we study the breakdown statistics in the same theoretical framework. The obtained distribution of failures in constant-current and constant-voltage stresses is compared with the widely used Weibull distribution [5]. fi FIRST THEORETICAL APPROACH TO THE BREAKWWN STAT1 STICSThere is experimental evidence to support the idea that the breakdown is a local phenomenon [6]. In fact, the self-healing technique would have been a nonsense if the breakdown hadn't been local. The main hypothesis of the presented model consists in assuming that the breakdown takes place when a critical number of defects , "bd, have been generated in a local zone defined by a minimum area, S O .If the total area of the MOS structure is divided in N cells of area S o , the distribution of defects in those cells follows a Poisson law if the local degradation rate does not depend on the already generated defects. Under these hypothesis, straightforward calculations lead to the following distribution of failures :
The energetic behaviour of hot electrons in the SiO2 of MOS capacitors subjected to constant-current stresses has been studied by means of a Monte Carlo simulation. The authors have observed that, for thick enough oxides, the average electron energy increases near the Si-SiO2 interface up to its steady-state value. These theoretical results are in accordance with the experimental evidence that this interface is the most degraded region of the oxide during electrical injection stresses. From the energy distributions of hot electrons, the fraction of electrons with energy below a threshold value (2.2 eV) has been registered for different electric fields and oxide thicknesses. Since this fraction is proportional to the time-to-breakdown, the authors explain the experimental thickness dependence of the time-to-breakdown in the cases of high and low values of oxide current density.
A model for the degradation and breakdown of thin gate oxide films is presented. During electrical stresses, a small fraction of the energy of the tunnel electrons that is dissipated in the oxide is converted into the creation of electron traps. When a critical density of traps is achieved, a fast runaway process leads the oxide to break down and its insulating properties are irreversively lost. It is demonstrated that the total charge injected to breakdown depends on the applied current in accordance with recently published results. The quasi‐linear log (time‐to‐breakdown) versus log (current density) plot experimentally obtained for VLSI oxides (tox ≈ 100 Å) is correctly predicted.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2025 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.