Effects of fusion-relevant gas atmospheres and pressures on hydrogen isotope characterization for fuel-retention diagnostics by LIBS

Fuel retention in plasma-facing components (PFCs) affects the operational efficiency and safety of fusion devices. However, detection of retained fuel by Laser-Induced Breakdown Spectroscopy (LIBS) remains challenging because the Hα and Dα Balmer lines commonly used for hydrogen-isotope identification are separated by only about 0.18 nm, making their discrimination highly sensitive to fusion-relevant ambient conditions. In this study, LIBS was employed to systematically investigate Hα/Dα discrimination on real EAST-exposed graphite tiles under various simulated ambient conditions. Depth-resolved measurements showed that the second laser pulse generally provided a more favorable Hα/Dα emission ratio and improved fitting quality, with the noise-normalized RMS residual decreasing from 3.6 to 1.26 compared with the first pulse. Meanwhile, time-resolved plume analysis helped explain the role of ambient environments in governing spectral broadening and isotope discrimination. At 100 Pa, the initial plume-height expansion rates within 200 ns were 15.6, 14.2, and 8.0 km/s in He, air, and Ar, respectively, revealing pronounced gas-dependent differences in plasma expansion. Combined spectral and plume analyses showed that vacuum provided high spectral resolution but suffered from a relatively short plasma lifetime. Within the investigated ambient parameters, a helium atmosphere at a moderate pressure of 100 Pa offered the most favorable balance between plasma lifetime and spectral resolution. In contrast, higher pressures or denser gases tended to enhance Stark broadening, leading to more severe peak overlap and a narrower effective discrimination window. The effective discrimination window was evaluated using the resolution factor Rs, with Rs ≈ 1.5 taken as a reference for baseline separation. These findings provide practical guidance for LIBS-based fuel retention diagnostics under EAST-relevant conditions.