Abstract
Copper will probably replace aluminum alloys as the interconnect metallurgy of choice for high performance semiconductor devices. This transition will challenge the suitability of established practices in focused ion beam (FIB) chip repair. A fundamental rethink in methodology, techniques, and process gases will be required to deal with the new metal films. This paper discusses the results of recent experiments in the areas of FIB exposure, cuts and connections to buried copper lines. While copper tends to mill faster than aluminum, etch rate variations due to grain structure tend to make reliable isolation cuts more difficult. The films also have been shown to suffer regrowth and surface reactions during long term storage following FIB exposure. Attempts at halogen gas assisted etch (GAE) mills result in undesirable removal characteristics, and in the case of bromine, the spontaneous destruction of all exposed copper in the immediate area. Resistance measurements and reliability of deposited tungsten connections to copper lines are also presented. In addition, the latest techniques developed for aluminum wiring isolation and device characterization are shown. These include 'cleanup' methods for achieving good circuit isolation without the extensive use of local oxide deposition, and the latest multilevel version of the FIB ‘wagon wheel’ for SRAM cell characterization. Also included is preliminary data from a custom built FIB chamber four manipulator prober module.
