It was recalled that a model for this phenomenon had recently been proposed, according to which interconnect failure was associated with the build-up of stresses to a critical level. It was shown here that the analytical solution for a semi-infinite line with a blocking boundary was a good approximation only when the stress build-up was small. This was not usually the case for narrow encapsulated interconnect lines, where the electromigration-induced stress could be quite high before failure. A complete model description, and a more accurate analytical solution to the differential equations which described electromigration-induced stress build-up at a blocking boundary, were then presented.

J.J.Clement, C.V.Thompson: Journal of Applied Physics, 1995, 78[2], 900-4