On real-scale LIDT testing for high power, high repetition rate CO2 lasers for EUV lithography

This study presents a methodology for laser-induced damage threshold (LIDT) measurements and analysis of high-reflective CO2 laser optics (wavelength ~10 µm) under high repetition rate (≥50 kHz), high power (≥30 kW), a pulse duration in the nanosecond range and real-scale beam diameters. These operation conditions are relevant for CO2 drive lasers used for laser-produced plasma (LPP) EUV lithography. A complex experimental LIDT setup is designed to handle these operation parameters and provide comprehensive metrology for data analysis (absolute laser power, beam profiles, temporal damage evolution and morphology). R-on-1 LIDT measurements are performed under conditions close to operation, i.e., repetition rate, power and beam diameter. The use of near-to-application beam diameters challenges LIDT data evaluation. First, large beam diameters limit the available statistics (only few sites per optic possible). Second, the application of real, non-Gaussian beam profiles requires a local fluence analysis well-overlapped with the damage morphology of the probed optics. The derived LIDT results are evaluated in terms of defect density. For system design and assessment of the laser damage behavior under operation conditions, defect density curves are combined with measured fluence distributions. This finally allows calculation of failure probabilities. The presented approach is suited for LIDT evaluation and analysis of large-scale lasers. For the studied operation conditions, a laser damage mechanism driven by local defects is identified.