Synthesis of scanning electron microscopy images by high performance computing for the metrology of advanced CMOS processes

Scanning electron microscopy is still the technique of choice for imaging and for in-line measurement of critical dimensions and overlay accuracy in most of the core technology processes. In particular, critical dimension microscopy provides information about design template matching and edge placement errors through links with design having proven beneficial effects on process yield and product reliability. In this paper, the use of high performance computing is demonstrated to simulate linescans and 2D secondary electron images to be used in optical proximity error correction strategies for nanometer scale technologies.