In extreme-ultraviolet (EUV) lithography sources, laser pulses create hot, dense tin plasma whose highly charged ions emit light near the 13.5-nanometer wavelength used for patterning. After the pulse, the plasma expands and cools, passing through conditions that are harder to characterize but important for understanding its interaction with the surrounding environment. Time- and space-resolved tin plasma spectroscopy offers a way to track this afterglow using optical emission from lower charge states.
A study of laser-produced microdroplet-tin plasma used optical spectra from neutral tin and ions Sn II through Sn IV to examine both plasma evolution and the Stark effect. Its results illustrate the diagnostic value of line widths and shifts—and the need to validate theoretical Stark parameters before relying on them to infer operating conditions.
Optical emission as a plasma diagnostic
When a laser strikes a molten tin droplet, it creates plasma with a broad range of charge states. The EUV-producing phase is hot and dense, but as the plasma expands, lower charge states become visible in the optical range. Their spectral lines carry information about the local conditions: line intensity can help estimate temperature, while line broadening and displacement can provide information about electron density and atomic interactions.
The reported measurements focused on 11 Sn I–IV emission lines between 315 and 425 nanometers. Spectra were collected at different times and positions, from 20 to 40 nanoseconds after the end of a 15-nanosecond laser pulse. This combination of temporal, spatial and wavelength information helps distinguish changes in the plasma over time from differences between regions.
What Stark broadening and shifts reveal
The Stark effect describes changes to spectral lines caused by charged particles in a plasma. Interactions with electrons can broaden a line and shift its position from its low-density wavelength. In a diagnostic application, the width can be related to electron density, provided the relevant Stark broadening parameter and plasma temperature are known.
Temperature is needed because the relationship between width and density is not entirely independent of plasma conditions. In this study, temperature was estimated from the relative strengths of Sn I and Sn II lines using Saha-Boltzmann fits. Stark widths were then used alongside that temperature estimate to determine how electron density evolved.
The researchers also compared each line’s Stark shift with its width. This shift-to-width ratio is largely independent of density and temperature under the conditions examined, making it useful for testing predictions from atomic physics models. Across several lines, experimental ratios and available theoretical predictions agreed poorly. For some transitions, experiment and theory even differed on the direction of the shift. These discrepancies matter because the same atomic data used to interpret spectra can influence derived plasma parameters.
Cooling and density decline in the afterglow
The measurements show a clear cooling trend during the observed afterglow. The local plasma temperature declined from approximately 2 electronvolts to 1 electronvolt in the measured time window. This followed a much hotter EUV-emitting phase, reported at roughly 30 electronvolts during the laser pulse.
Electron density also fell rapidly. Using Stark widths from Sn II lines, the study found an approximately exponential reduction from around 1018 to 1017 electrons per cubic centimeter over about 20 nanoseconds. Density estimates from several charge states were broadly similar at early times, though the authors noted a tentative trend in which lower charge states appeared to probe higher densities over time. Uncertainties in the Stark parameters and the line-of-sight integration of emission complicate interpretation of that difference.
Measurements at nearby viewing positions showed no significant change in the Sn II density estimate across the tested displacement. This suggests a locally rather flat density profile in the observed region; it does not establish that the entire expanding plasma is uniform.
Implications for plasma diagnostics
For engineers and technical teams working with EUV source plasmas, the study highlights several practical points:
- Use multiple diagnostics together. Spectral line strengths can constrain temperature, while Stark widths provide a route to estimating electron density.
- Check the atomic data. Disagreement between measured and predicted shift-to-width ratios can signal that a theoretical Stark model needs further validation.
- Account for measurement geometry. A line of sight integrates light from different plasma regions, so a spectrum may not represent a single, uniform condition.
- Interpret charge states carefully. Different ionization states may sample different parts or conditions of an evolving plasma, particularly as it cools and expands.
The work examined the afterglow rather than directly measuring the densest, hottest EUV-emitting region. Its optical measurements therefore complement, rather than replace, diagnostics aimed at the initial source conditions. The results also point to the value of extending spectroscopy to higher charge states where suitable transitions can be observed.
Conclusion
Time-resolved tin plasma spectroscopy can map important changes as laser-produced plasma moves from its EUV-emitting phase into a cooler, less dense afterglow. In this case, optical line measurements indicated cooling from about 2 to 1 electronvolt and a rapid decline in electron density. Just as importantly, measured Stark shift-to-width ratios exposed substantial disagreements with some theoretical predictions. For plasma diagnostics, the practical lesson is to combine line intensities and widths with careful attention to atomic data, spatial integration and the changing charge-state distribution.