The continually shrinking dimensions of today’s semiconductor technology occasionally allow for novel approaches in imaging defects. It has become desirable to image subsurface voids prior to cross sectioning and some efforts have been made to address this need including the construction of specialized instrumentation [1]. The thickness of the metallization levels at the 65 nm technology node and smaller now allow for the use of the electron beam in a scanning electron microscope (SEM) as a material sensor. At high accelerating voltages (between 20-30 kV) in backscatter imaging mode the numerical gray level values at each pixel location can correlate to the amount of material directly under the electron beam at that location. This is particularly evident when dealing with defined geometries and material sets offering high contrast changes between materials such as those found in semiconductor technology like copper (Cu) metal and conventional dielectric materials. As a result, subsurface voids can be mapped to a reasonable representation prior to cross sectioning and precise pinpointing of the defect location in test structures can occur. This paper discusses this methodology on 65 nm technology with Cu metal lines in a low-k dielectric material for a two level metal test structure. To some extent this work represents a natural extension of a paper presented previously by the author [2].

This content is only available as a PDF.
You do not currently have access to this content.