Effects of systematic and stochastic errors on estimated failure probability of anisotropic conductive film

This paper analyzes the effects of systematic and stochastic errors on the failure probability of anisotropic-conductive-film (ACF) assemblies estimated using the V-shaped-curve method. It is shown that the effect of systematic errors varies as a function of the volume fraction and the volume-fraction bias. The effects of stochastic errors are investigated by using an in-house software program to generate random conductive-particle distributions in the pad and interpad regions of the ACF package for the given volume fractions and package geometries. The dependences of the coefficient of variation (CV), essentially the degree of uniformity of the particle distribution, and the failure probability on the volume fraction are examined, and the corresponding results are used to derive the correlation between the stochastic error and the CV for a given volume fraction. In general, the current results indicate that the effects of systematic errors on the accuracy of the estimated failure probability can be controlled by improving the accuracy with which the resin and conductive-particle components of the ACF-compound material are weighed during the ACF fabrication process. However, the effects of stochastic errors cannot be controlled and vary as a function of the volume fraction and the degree of nonuniformity of the particle distribution. Nevertheless, the results indicate that the effects of both systematic and stochastic errors can be suppressed by specifying the volume fraction as the value corresponding to the tip of the V-shaped curve when designing the ACF compound.