“…[48] assume however that all gold aluminide compounds are formed during ballbonding, which appears inconsistent with the majority of reactive diffusion literature. Such a failure mechanism is not consistent with the experiments reported in this paper that show degradation within Au 4 Al and between The effect of isothermal ageing temperature on the proposed de-bonding mechanism remains an open question except for some limited work in a recent study on 40 lm pitch ballbonding [49], which shows the same HSBL and LSBL failure modes occur after baking at 150°C, 175°C and 200°C and with increased bake temperature both types of failure occur at shorter times. Interestingly however, the same study also shows that for each ageing temperature, the use of a different capillary geometry results in the de-bonding of gold balls at approximately the same time regardless of squash height.…”