New perspectives in defect centric model for NBTI reliability. (July 2019)

• Record Type: Journal Article
• Authors: Nouguier, D.; Ghibaudo, G.; Federspiel, X.; Rafik, M.; Roy, D.
• Is Part Of: Microelectronics and reliability. Volume 98(2019)
• Page Start: 119
• Page End: 123
• Publication Date: 2019-07
• Subjects: Reliability -- Statistics -- Bias temperature instability -- Kinetics -- Defect centric model; Electronic apparatus and appliances -- Reliability -- Periodicals; Miniature electronic equipment -- Periodicals; Appareils électroniques -- Fiabilité -- Périodiques; Équipement électronique miniaturisé -- Périodiques; Electronic apparatus and appliances -- Reliability; Miniature electronic equipment; Periodicals; 621.3815
• DOI: 10.1016/j.microrel.2019.05.010
• Languages: English
• ISSNs: 0026-2714

Abstract

Abstract: Inconsistency in the single-family defect centric model for the prediction of the cumulative distribution function of NBTI Vth shifts after stress and relaxation with the same average degradation is outlined. Then, for relaxation, a two-component ΔVth model is used to extract the slow and fast contributions of the cumulative distributions during the relaxation phase. The bi-layer DCM model based on two-defect independent families for the prediction of the cumulative distribution is applied but without success. Indeed, the fast and slow components during relaxation phase were found to be significantly correlated, justifying the inadequacy of the bi-layer DCM analysis on our data. Subsequently, this model is modified to take into account the correlation between fast and slow components of relaxation and a new correlated dual defect centric model (c-DDCM) able to predict the cumulative distribution functions in the presence of correlated two-defect families is proposed. Highlights: Inconsistency in DCM based on one defect family after stress and relaxation with the same average degradation is outlined. Fast and slow components during relaxation phase are found to be significantly correlated. A new correlated dual DCM able to predict the CDF in the presence of correlated two-defect families is proposed.