Stress and Electromigration
- PDF / 762,505 Bytes
- 12 Pages / 414.72 x 648 pts Page_size
- 3 Downloads / 197 Views
Introduction The interdependence of mechanical stress and electromigration behavior has been recognized for quite some time, but has only begun to be fully appreciated and dealt with in a meaningful manner. There have been a number of recent advances in understanding that also may lead to a change in the way that electromigration failure is viewed in real applications.
History In 1972 Ainslee et. al. (1) first recognized that mechanical stress may play an important role in electromigration damage and failure processes. They were trying to explain results where electromigration lifetime was considerably enhanced by the presence of an overlying passivation layer. (2,3) The idea that a stress gradient will oppose electromigration flow and the effect that stress will have on the diffusion coefficient were both presented. However, it was concluded that the stresses required to have a meaningful effect on stripe lifetime would never be experienced in real systems. Subsequent experimental and theoretical analyses have, however, shown that this later conclusion was based on an underestimate of the stress levels that can be obtained in thin film conductors undergoing electromigration. (4-22) In a series of what can be called landmark papers a few years later, I.A. Blech and co-workers shed greater light on the importance of stress in understanding electromigration. (6-9) Performing experiments on "islands" of metal deposited onto higher resistance refractory underlays, Blech et. al. showed there was a critical product of current density and the length of the sample that had to be exceeded before electromigration drift would occur. This was interpreted as the effect of the stress gradient that is generated when electromigration induced diffusion is stopped by a diffusion barrier. (9), for which there is direct experimental evidence (10, 22). Since then it has become common practice to refer to the maximum length of a conductor that can sustain a stress gradient that will halt electromigration as a "Blech Length". Additional advances were made by the Cornell group, particularly by Korhonen et. al. (21), who solved an equation first proposed by Kinsbron et. al. (10a) with the use of several simplifying assumptions to demonstrate that the main role of a non-thermal equilibrium vacancy concentration would be to generate mechanical stress. (23)
The Origin of Electromigration-Induced Stress Electromigration is best described as a biased diffusion where the driving force is a result of a momentum exchange between conducting electrons and defects in the metal lattice. The most important of these defects are vacancies, since each vacancy can be the place where a diffusing metal atom can move into. The electrical current, therefore, produces a kind of pressure on the diffusing atoms in the direction of the electron flow for n-type conductors, and a corresponding opposite pressure on vacancies. Electromigration has been spoken of as a biased diffusion of
231
Mat. Res. Soc. Symp. Proc. Vol. 391 ©1995 Materials Research Society
vac
Data Loading...
