This chapter begins by looking at the first-wall erosion problem for EUV sources and proceeds in Sec. 15.2 to introduce the concept of a directed discharge, which allows the hot EUV-emitting plasma to be held at a sufficient distance from the first wall to yield industrially acceptable component life while maintaining the high efficiency and low cost of a DPP source.
In EUV sources of the plasma type, either LPP or DPP, the challenge is to achieve rather hot and dense plasma conditions repetitively for at least 1010 pulses without damaging the illumination or collector optics, chamber walls, and (in the DPP case) electrodes. As an example of typical plasma conditions, the Star Pinch, which is driven by multiple converging plasma channels, dissipates 10 J in less than 500 ns within a 0.7-mm3 volume, reaching an estimated electron density of 4 × 1019 e cm−3 and a temperature greater than 30 eV. At its hottest, the plasma radiates broadband EUV power at the rate of 20 MW, so it cannot be sustained continuously. In the case of xenon plasma discharges, approximately 20% of the dissipated energy is released promptly in a 50-ns burst of radiation, most of it in the 10–15-nm band. The remaining energy is carried by the expanding plasma toward the facing walls, where ion sputtering is the main erosion mechanism.
The typical plasma particle energy impacting the wall is roughly between 300 eV and 1 keV. We reach this number by considering the case of a plasma comprising Xe10+ ions (the ionization state that contributes most strongly to 13.5-nm radiation) that expands adiabatically from an initial temperature of 30 eV without recombination. At asymptotic distances the electron kinetic energy is low and Xe ions carry the energy formerly stored by the electrons, totaling 300 eV. In practice, di-electronic recombination occurs very rapidly through many of the xenon ionization stages before the plasma has even doubled in diameter. As that occurs, the electron density is reduced, but the electron temperature tends to be raised. Highly excited ions produced in recombination further heat electrons via superelastic collisions, but most of this electron energy is then imparted to the ions as electron pressure drives the expansion. The net effect is for much of the xenon ionization energy (totaling 807 eV) to be added into the radial kinetic energy and for the expected asymptotic ion energy range to lie roughly between 300 eV and 1 keV.
Online access to SPIE eBooks is limited to subscribing institutions.