Statistical characterization of fatigue lifetime of polysilicon thin films

Abstract

This paper discusses, from a statistical point of view, whether the fatigue behavior of polysilicon thin films can be formulated uniquely on the basis of initial defects which determine their strength distribution. The fatigue process determining the lifetime is modeled with a power law similar to the well-known Paris law as the fatigue crack extension process from the initial defects whose equivalent crack lengths were evaluated from the initial strength distribution. The parameters in the power law were here fit to the results of tensile fatigue tests performed on three groups of specimens patterned using three different conditions leading to different etching damage. The results showed that the obtained power law parameters of polysilicon are spread over a range where the calculated behavior changes insignificantly in spite of the different etching conditions leading to different initial strength distributions. They support a possibility that the fatigue fracture is caused by the accumulation of defects and that therefore the fatigue lifetimes can be practically predicted from their strength distributions by using the same values of the parameters in the power law.