Skip to Main Content
Your organization might have access to this article on the publisher's site. To check, click on this link:http://dx.doi.org/+10.1063/1.1459756
We use modeling and simulation tools to determine the beneficial additives or dopants to Cu interconnect. We have designed a virtual simulation procedure to cover several important aspects in screening a potential dopant to Cu with the assumption that grain-boundary (GB) diffusion is dominant for Cu electromigration performance. The procedure investigates dopant segregation to GB, bulk diffusion, dopant and Cu self-diffusion at the GB, and the effect of the dopant’s presence on Cu diffusion at the GB. Defect formation and migration energies as well as activation energies were calculated using the state of the art ab initio method. Two primary mechanisms for a dopant to be effective were identified, namely, dopant blocking and dopant dragging mechanisms. For dopant blocking mechanism the desired dopants occupy the GB interstitial sites and block the fast diffusion pathway for Cu. In the case where Cu atoms occupy the GB interstitial sites, the desired dopants segregate to the nearby substitutional sites and drag the fast diffusing Cu. Early experimental results have confirmed model prediction for several dopants identified so far. The mean time to failure has increased more than 60% with a dopant concentration as low as 0.01 at. % in Cu and the resistivity increase can be controlled below 15% compared to undoped Cu. We demonstrate that modeling and simulation have become valuable alternatives to experiment for design of advanced materials systems for technology research and development. © 2002 American Institute of Physics.