Characterisation of intermetallic growth in copper and gold ball bonds on aluminium metallization

Abstract

While the characterisation of intermetallic coverage and intermetallic phase (IP) growth in gold ball bonding on aluminium is quite well understood, there is relatively little literature concerning the morphology and growth of 1P's between Cu halls bonded on aluminium pad metallisation.The difference between Cu-AI IP growth compared with the well known Au-AI IP's has been studied mainly of larger wire diameter (35-50pm) in the early 1980's. Cu wire ball bonding has been established for many years mainly for high power devices at wire diameters >= 38pm and fine wire for discrete device applications. However, there is now interest in fine pitch Cu wire ball bonding at smaller wire diameters of 25pm and smaller for high pin count applications, driven mainly by cost reduction. Development and optimisation of robust copper wire bonding processes for such applications requires an assessment of intermetallic coverage and Cu-AI intermetallic growth after isothermal aging. This paper describes the problems associated with coverage determination, some characteristics of Cu-AI and Au-AI intermetallic compounds and characterises the difference in the IP growth between Au-AI and Cu-AI. The relative merits of gold and copper ballhonding are also briefly discussed.

What is the future of bonding wire? Will copper entirely replace gold?

“…Intermetallic growth in copper ball bonds was much faster because of the higher temperature. In 2005 Wulff et al [32] published a study on intermetallic growth in copper and gold ball bonds. Cross-sections of copper ball bonds aged at 175°C in Fig.…”

“…At 300°C, shear strength increased significantly and at long times failed in the copper ball, a mode C failure, which was attributed to annealing and softening of the ball. A paper by Wulff et al in 2004 [19] studied bonded ball hardness and high temperature storage of copper ball bonds at 250°C. Intermetallic growth in copper ball bonds was much faster because of the higher temperature.…”

“…However, as every metallurgist knows, materials like copper that work harden rapidly, will do so more when they are softer and a softer FAB might work harden more than a harder FAB during bonding. Wulff et al [8] showed by measuring Vickers hardness of bonded balls ( Fig. 8) that gold and copper ball bonds can workharden and that copper does so more than gold.…”

“…A further complication with copper ball bonds, but not necessarily a great handicap, is the difficulty in measuring intermetallic coverage because the phase created during thermosonic bonding that bonds ball to bond pad is simply to thin to be seen with routine methods such as optical and scanning electron microscopy [19]. It is necessary to use transmission electron microscopy [20].…”

Some thoughts on bondability and strength of gold wire bonding

“…Finite element analysis (FEA) has been used extensively by several researchers to analyse the plastic deformation that occurs during the wire bonding process [23][24][25][26]. Saiki et al [23] investigated the effect of capillary tip shape on ball bonding using finite element analysis.…”