Confirmation of a predictive process dependent model of oxide charging

Abstract

Summary form only given. The concept of building in reliability (BIR) has been gaining attention in the semiconductor industry. Full realization of BIR, in our view, requires the development of physics-based models of the effects of process parameter variations on reliability mechanisms and subsequent incorporation of these models into predictive semiconductor TCAD tools. Previously, a physics-based model of charge trapping in "intrinsic" SiO/sub 2/ was introduced (Conley et al, IEEE Integrated Reliability Workshop Final Report, p. 134-141, 1996) and its potential predictive power demonstrated on a limited array of oxides (by "intrinsic", it is meant that trapping due to extrinsic contaminants is insignificant). Here, work is presented that confirms the validity of the model's parameters and shows that the model is predictive for a variety of oxides. In addition, very preliminary results are presented that address the equilibrium kinetics assumptions that were made in order to calibrate this model.