A model for corrosion induced failure in wire bond devices made with either Cu or Au wire was developed. The model is based on detailed analysis of the chemical composition, crystallography, and microstructure of the corrosion induced failure sites. The detailed analysis was enabled by both a scanning electron microscope (SEM) and a transmission electron microscope (TEM) equipped with the appropriate analytical detectors.The combined characterization results were used to develop a detailed failure mechanism model which explains not only the overall chemical reactions but also the resultant two phase microstructures of the corrosion product observed in both metal systems. In addition the model helps to explain why Cu wire bonded devices are more susceptible to corrosion than Au wire bonded devices. In both systems corrosion occurs in the IMC not the end components pure metals. It is proposed this is due to the oxide on the surface of the IMCs being less resistant to pitting corrosion than that for Al. Once IMC passivity is broken down, corrosion of the IMC proceeds via selective oxidation of Al. This leads to the formation of a corroded region which is composed of a two phase microstructure, crystalline -Al2O3 with embedded crystalline Au and Cu metal particles. The resulting oxidized interface is highly susceptible to fracture, which is the ultimate reason for device failure.Although similar there are differences in the Cu/Al and Au/Al systems. There is a notable difference in the size and distribution of Cu particles in the aluminum oxide corrosion product of the Cu-Al system as compared to the Au particles in the aluminum oxide corrosion product of the Au-Al system. In particular, the Cu particles appear to be more uniformly distributed as compared to their Au counterparts. This difference is likely related to the crystal structure of the IMC from which the corrosion product was formed. It is proposed that this difference is related to the presence and probability of Al to Al bonding in the IMC phase/s. Furthermore, the IMC structures are also suspected to be responsible for the better immunity to contamination (specifically Halide) induced corrosion of Au relative to Cu.

A series of erosion–corrosion (EC) tests was carried out on HVOF Cr3C2–NiCr cermet coatings with different fractions of NiCr metal binder phase. Testswere conducted using a nozzle type elevated temperature erosion tester at test temperatures up to 450°C. Bed and fly ash retrieved from operating fluidised bed combustor boilers as well as SiO2 angular quartz were used as the erodent materials for tests at both shallow and steep impact angles. The fly ash contained chemically active compounds and a corrosive environment was involved. The morphology of the specimens was examined by scanning electron microscopy. It was found that the cermet coatings were eroded by SiO2 quartz and fly ash at a low velocity with peak wastage occurring at an intermediate value of carbide content at both the shallow and the steep impact angles. Erosion caused by the bed ash at a high velocity resulted in no peak wastage at either the shallow or the steep angles. Instead, the EC thickness loss increased with increasing metal binder content. The EC behaviour of the coatings is related to the morphology of the coatings and the characteristics of the erodent particles.