Effects of radiation-induced carbon contamination on the printing performance of extreme ultraviolet masks.

摘要

This dissertation investigates one of the remaining issues for extreme ultraviolet (EUV) lithography, the effects of radiation induced carbon contamination on the printing performance of patterned EUV masks. The impact of carbon contamination on EUV masks is significant due to the throughput loss and potential effects on imaging performance, and occurs when multilayer surfaces are exposed to EUV radiation with residual carbonaceous species present. Current carbon contamination research is primarily focused on the lifetime of the multilayer surfaces, determined by reflectivity loss and reduced throughput in EUV exposure tools. However, contamination on patterned EUV masks can cause additional effects on absorbing features and affect the printed images.;In this work, various carbon contamination experiments were performed to study the impacts of carbon contamination on mask features and its effects on imaging. Selected fields on patterned EUV masks were contaminated with a series of exposures, and then analyzed to determine the effects of carbon contamination. Using several techniques such as top-down inspection with a reticle scanning electron microscope (SEM), mechanical surface measurement using critical dimension -- atomic force microscopy (CD-AFM), and aerial image analysis with EUV microscope, we observed that the CD of contaminated features was increased, as well as the printing results showed an increased dose to print the target CD from contaminated features.;In order to understand the effects of carbon contamination topography on mask absorbing features, direct measurements of transmission electron microscopy (TEM) cross-section images were used to characterize the contaminated features. Non-uniform and asymmetric carbon contamination topography was observed at various feature sizes and pitches on different masks. With the knowledge of real contamination topography, we can then calculate the effect of contamination topography on the printing performance.;The baseline simulation was developed with 40-nm 1:1 lines and spaces features, and the simulation parameters and contamination film properties were determined based on current tool status and our experimental results. The simulation was then compared to the actual printing results to ensure the accuracy, as well as the printing results from other mask feature designs. With the confidence of the simulation, various feature size, duty cycles, and mask absorber heights were modeled to see the effects of carbon contamination with different conditions. For high volume manufacturing (HVM) tools and future mask designs, one should be able to reproduce this model and predict the effects of carbon contamination on the printing performance of any patterned EUV masks.