A Novel Numerical Multiphysics Framework for the Modeling of Cu-Al Wire Bond Corrosion under HAST Conditions

Abstract

Copper wire bond interconnect has been used widely as the first-level interconnect in semiconductor packaging industry over gold wire bond interconnect. It has several advantages over gold wire bond interconnect such as lower material cost, advanced electrical, mechanical and thermal properties as well as lower interdiffusion rate when bonded with aluminum pad under high temperature operational conditions. However, the use of copper wire for first level interconnects also come with some tradeoffs, such as narrow process window, higher hardness that cause cratering and poor corrosion resistance. Compared to Au wire, Cu wire is more likely to suffer galvanic corrosion when the operational environment is humid and when electric bias is involved during the aging. A few regression-based models have been purposed to address the reliability performance of Cu-Al wire bonds over the last several years. In this paper, a novel numerical multiphysics framework for modeling of Cu-Al wire bond corrosion under HAST conditions is proposed. The model is characterized by Butler-Volmer equation and Nernst Planck equation. It features corrosion front tracking, a moving meshing-based approach to keep track of the corrosion progression at the Cu-Al bond pad interfacial area. It captures the effect of IMC growth, contaminant transport during the high humidity aging process as well as the electric bias. The model has been used to predict the lifespan of Cu-Al wire bond interconnects in packages molded with different type of molding compounds under different aging conditions.