Parameter optimization of extreme ultraviolet radiation from lithium targets driven by short-wavelength solid-state lasers

Short-wavelength solid-state lasers are considered promising drivers for next-generation extreme ultraviolet (EUV) lithography sources owing to their compact architecture and high electro-optical conversion effciency. Using a one-dimensional radiation-hydrodynamic model, the EUV emission characteristics of laser-driven Li plasmas were investigated, and the dependence of the in-band conversion effciency (CE) within a 2% bandwidth centered at 13.5 nm on laser wavelength and intensity was analyzed. The results show that the optimal laser intensity increases with decreasing wavelength, while the CE exhibits a nonmonotonic dependence on wavelength. A maximum CE of 2.5% was obtained at a wavelength of 0.5 μm and an intensity of 6 × 10¹⁰ W/cm², accompanied by a relatively high spectral purity (SP) of 22.4%. Further analysis reveals that shorter wavelengths shift laser energy deposition toward denser plasma regions, thereby increasing the radiation effciency. However, the resulting increase in plasma optical depth enhances radiation self-absorption and re-emission processes, leading to spectral broadening and a reduction in SP. The evolution of the CE is therefore governed by the competition between radiation enhancement and spectral broadening. To boost optimization effciency, the enhanced Bayesian optimization framework identified the reference global optimum within 25 total objective-function evaluations in a representative run, compared with 272 simulations required for exhaustive grid scanning. Additional comparisons with available Li-target experiments show good agreement in the CE-intensity trends and major EUV spectral features. These results provide physical insight into short-wavelength laser-driven Li-plasma EUV emission and demonstrate the potential of Bayesian optimization for effcient parameter optimization of laser-driven EUV sources.