Study on the Prevention of Optics Contamination in Photo-Lithography

Abstract

In semiconductor manufacturing technology, photolitho exposure tools have advanced scaling of integrated circuit (IC) chips for decades. The tools enable illumination optics to expose the patterns of photo-mask onto wafers coated with photo-resist (PR). As the line width of the chip has been scaled down below 20 nm, the contamination of optics degrades the uniformity of light so that it reduces the accuracy of optical patterns. Thus, contamination in the optic system is crucial to enhance wafer pattern scaling.The level of contamination in optics can be identified with light scattering (i.e., Straylight index). By tracing the Straylight index for years, we determined REMA (Reticle Masking) lens, which is one of the lens element, was contaminated by the chemical ions of FAB air. We implemented destructive inspection with the contaminated lenses to identify degradation mechanisms and detected plenty of positive ions (i.e., Si+) (Table.1) and negative ions (i.e., Silicon Oxide) (Table.2) on REMA Lens. (Fig.1). The humidity in FAB air which is attach to the polar surface of optics, light exposure transforms it to hydroxide ions or highly reactive radical (OH*, H*, O*) by the DUV light. And then the highly reactive radical is easy to combine with non-cleanable contamination (e.g., silicon, sulfur, and nitrogen). Although Photolithographic tools have a purging system using inert N2, XCDA (Extreme clean dry air), CA (Clean Air) to prevent contamination. Each optic has a different purge system and environment, so contamination has been strongly concentrated on the last lens element of REMA.In this paper, we performed three main tests to slow the rate of lens contamination. First of all, literally hundreds of tools in the factory have measured with the efficiency of the chemical filters for pollutions. Then, we exchanged types of chemical filters based on the different contamination speed. This test helps us to identify contamination sources. Secondly, we changed the air mixing ratio (i.e., Dilution factor). Despite the similar structure of the tool, the air mixing ratio is different respectively as the time of each tools is released. After investigating the three different values, we found an optimized value to slow down the contamination speed. Lastly, we reform the structure of tools to avoid mixing the pollution air. Overall, these test results tell a remedy of how to manage the lens contamination in optics. Figure 1