Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Scaling and microstructure effects on electromigration reliability for Cu interconnects

 
: Hu, C.-K.; Hübner, R.; Zhang, L.; Hauschildt, M.; Ho, P.S.

:

Baklanov, M.R.; Ho, P.S.; Zschech, E.:
Advanced interconnects for ULSI technology
Chichester: Wiley, 2012
ISBN: 978-0-470-66254-0 (Print)
DOI: 10.1002/9781119963677
pp.291-337
English
Book Article
Fraunhofer IZFP, Institutsteil Dresden ( IKTS-MD) ()

Abstract
Prior to the 65 nm technology node, the Cu interconnects commonly had a bamboo-like grain structure where the mass transport under electromigration was dominated by interface diffusion. In this case, the EM lifetime was found to degrade by about half for every new generation, even for the same current density. At the 65 nm node with line width reduced to 90 nm, small grains agglomerated to mix with large bamboo grains in Cu damascene lines. This further degraded the EM lifetime due to the contribution of the grain boundary diffusion to mass transport. These observations raise serious reliability concerns about the effect of scaling on grain structure and EM reliability for future technology. This chapter reviews the current understanding of the scaling and microstructure effects on EM reliability for Cu interconnects. It is organized into five parts. First we review the basic physics of the EM phenomenon, focusing on EM mass flow, lifetime scaling rule, test structures and the relationship between Cu EM lifetime and current density. This follows with a discussion on the techniques developed recently for Cu microstructure analysis and the correlation of Cu microstructure to EM-induced interconnect damage. In the third part, we present the techniques developed for improving EM lifetime, including the use of alloys, upper-level dummy vias, Cu surface treatments and surface metal coating. The fourth part describes the results from a recent study on the effects of grain size and cap layer on EM lifetime and statistics. Finally, in the last part, we discuss the methods developed for massive-scale statistical studies using the Wheatstone bridge method and conclude with a summary.

: http://publica.fraunhofer.de/documents/N-350479.html