About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Late News Poster Session
|
| Presentation Title |
High-Throughput Quantitative Characterization of H-Plasma-Induced Material Degradation of EUV Lithography Mirrors |
| Author(s) |
Marvin Josue Calderon, Tiffany Kaspar, Semanti Mukhopadhyay, Arun Devaraj, Hari Harilal |
| On-Site Speaker (Planned) |
Marvin Josue Calderon |
| Abstract Scope |
Extreme Ultraviolet (EUV) lithography is the most precise microchip fabrication method currently available. Multilayer Mirror (MLM)s consist of thin (3.5 nm) stacked layers of Si and Mo, which reflect and concentrate EUV radiation onto a wafer. During use, debris accumulates on MLMs, degrading their reflectivity. During routine cleaning, MLMs interact with H-plasma to flush off contaminants, but results in blister formation, damaging it irrevocably. Although existing works describe this H-plasma induced blistering, analysis of blister formation and evolution over time have not been explored. Leveraging a H Focused Ion Beam system, we observed growth of these blisters in situ. We aimed to develop high-throughput workflows to enable quantitative characterization of these blisters. To this end, we explored several traditional and machine learning-based segmentation strategies to measure and understand blister nucleation and growth in MLMs. These strategies contribute to future efforts into blister growth modeling, and the development of advanced MLMs. |
| Proceedings Inclusion? |
Undecided |
| Keywords |
Machine Learning, Thin Films and Interfaces, Characterization |