Impact Induced Metal-Crush Failures

Abstract

The authors show for the first time that metal-crush failures can also be caused by mechanical impact which occurs during handling of the package and pick and place processes. So, not only thermo-mechanical effects might cause these failures. The paper also show a clear correlation between the observed failures on sample IC that returned from the field, and the failures observed on samples which were studied with the impact test. Both the position and the failure signature are the same. The author also showed that the sensitivity of packages for this failure cause highly depends on the chip coverage and that it is layout dependent. No die coating and polyimide coating show the best resistance against impact damage. Single and dual drop coatings are very sensitive. As such, die coat needs to be avoided as much as possible. However, since some designs count on its stress relieving ability to keep parameters within specification, the effectiveness of polyimide needs to be characterized. Moreover, it was found that this kind of impact can cause passivation cracks without causing an electrical failure of the chip, or even no failure after ATE test and after additional bake and solder reflow tests. The Charpy system is a representative tool for the evaluation of the robustness of packages within the logistic flow. The most effective activation stress test to combine with mini-Charpy tests still needs to be defined. This possible failure cause was up to now never investigated and never taken into account in the qualification process of packaging material. It is clear that the sensitivity of different packaging types and moulding components to such impact should be tested in advance and added to the qualification process. Finite element modelling clearly indicates that particle indentation is causing high tensile stresses in the passivation layer. These stresses are in the order of the ultimate stress values for these materials.