EUV Lithography
# EUV Collector Mirror and Hydrogen Radical Debris Cleaning
The EUV collector mirror sits directly in the line of fire of the tin plasma that generates the light it's supposed to focus — a nearly spherical mirror wrapped almost entirely around the plasma to capture as much of the emitted 13.5nm light as possible, which also means it's the single surface in the entire scanner most exposed to the debris that plasma throws off. Every tin-droplet vaporization event that produces usable EUV light also ejects tin ions, neutral atoms, and droplet fragments outward in all directions. The collector mirror's geometry — large, curved, positioned close to the plasma to capture the widest possible solid angle of emitted light — makes it unavoidably the primary target for that debris. Over time, tin accumulates on the mirror's reflective coating, and even a thin tin layer measurably degrades its EUV reflectivity, directly cutting the usable light reaching the rest of the scanner.
Hydrogen radical cleaning solves the contamination problem by converting deposited tin into a volatile compound that pumps away, rather than trying to prevent the tin from landing on the mirror in the first place. Hydrogen gas, dissociated into reactive hydrogen radicals near the mirror surface, reacts with deposited tin to form stannane, a volatile tin-hydrogen compound that evaporates off the mirror and gets pumped out of the chamber. This is a continuous, in-situ cleaning process running throughout scanner operation — it doesn't stop tin from depositing, it keeps removing it fast enough that the mirror's reflectivity stays within an acceptable range over the scanner's operating life, rather than requiring the mirror to be physically removed and cleaned or replaced on a short cycle.
Collector mirror reflectivity directly multiplies into overall EUV source efficiency, so even a few percentage points of reflectivity loss from accumulated tin propagates straight through to scanner throughput — which is why hydrogen radical cleaning rate, not just droplet generator output, is one of the parameters EUV source engineers tune hardest. A collector mirror that degrades faster than its cleaning system can keep up with doesn't just need maintenance sooner, it silently reduces the light delivered to the wafer during every exposure in between, making the cleaning system's steady-state balance point — deposition rate versus removal rate — as consequential to scanner economics as the droplet generator's raw output.
The tin droplet source generates the light; the collector mirror is the very next thing that light touches, and keeping it clean turns out to be exactly as continuous and chemistry-dependent a problem as generating the light in the first place — proof that EUV's practical limits run through a chain of material surfaces, not a single component.