Low-K wire bonding

When we encountered the bond pad metal peeling defects with low-K devices during wire bonding, we have a tendency to optimize ´bonding power´, ´bonding force´ and ´bond time´ so far. But the effectiveness of improvement by evaluations was not so satisfactory as much as we have got with normal Si devices. So this brought us to assume existence of another important factors against the defects and initial force and bonding force were revealed to play an important role in the defects through DOE (design of experiments) implementation. As the results of DOE and some more evaluations, the defects could be reduced by squashing initial balls (free air balls) using the initial force and bonding force before bonding starts or at the very beginning of bonding because the stress onto the bond pads through balls during wire bonding can be distributed and lessened as much as the squashed area. And when the initial force is larger than the bonding force, it was more effective to reduce the defects