Electromigration threshold in damascene versus plasma-etched interconnects
Joris Proost, Karen Maex, L. Delaey
Abstract
Joris Proost, Karen Maex, L. Delaey
Abstract
Electromigration threshold has been measured for unpassivated damascene Al(Cu) interconnects by drift experiments on Blech-type test structures, and results are compared to equivalent plasma-etched lines. When using a damascene implementation, the critical product of current density and stripe length increases significantly. As a result, the incubation period, needed for Cu depletion beyond a critical length, is increased over a wide temperature range (155–230 °C). This is shown to be an intrinsic effect, directly controlling the relative electromigration performance of both metallization schemes.
OpenAlex reports 13 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Electromigration threshold has been measured for unpassivated damascene Al(Cu) interconnects by drift experiments on Blech-type test structures, and results are compared to equivalent plasma-etched lines. When using a damascene implementation, the critical product of current density and stripe length increases significantly. As a result, the incubation period, needed for Cu depletion beyond a critical length, is increased over a wide temperature range (155–230 °C). This is shown to be an intrinsic effect, directly controlling the relative electromigration performance of both metallization schemes.
Key concepts: Electromigration, Copper interconnect, Plasma, Materials science, Current density, Optoelectronics, Copper, Metallurgy