In situ transformation and cleaning of tin-drop contamination on mirrors for extreme ultraviolet light

Abstract

Tin-drop contamination was cleaned from multilayer-coated mirrors by induction of phase transformation. The β→α phase transition of tin was induced to initiate material embrittlement and enable facile removal of thick tin deposits. The necessary steps were performed under high-vacuum conditions for an in-situ demonstration of the removal of severe tin contamination from optics used for reflection of extreme ultraviolet light.Molten tin of high purity was dripped onto mirror samples, inoculated with small seed particles of gray tin and then cooled to temperatures in the range of -25 °C to -40 °C. As recorded by photographic imaging, the drops were converted in an evacuated chamber to gray tin by induction of tin pest leading to their disintegration within a few hours. They could then be easily cleaned or fell off from the surface without causing any damage of the multilayer coating. Cleaning of tin contamination from the mirrors with almost complete structural transformation of the tin drops and subsequent removal by puffs of dry gas could be achieved within a day. The fraction of area coverage of untransformed tin remaining on the samples after cleaning was evaluated from the images and generally found to be well below 1%. After tin dripping, phase transition and cleaning, analysis of the reflectance of a Mo/Si-coated mirror with measurements at wavelengths of 13.6 nm and 13.5 nm showed a reduction by only 0.5%, with an upper limit of 1%.

Sticking behavior and transformation of tin droplets on silicon wafers and multilayer-coated mirrors

“…Subsequently, the loose pieces of transformed tin can easily be removed from the surface. We have already verified that multilayer Mo/Si-coated mirror samples do not undergo any substantial degradation in EUV reflectance at 13.5 nm by this process [20]. Using a refined transformation procedure, we now investigate samples with capped EUV ML-coating layers directly relevant for collector mirrors in plasma light sources and also examine if there is any influence of the substrate coating composition.…”

“…The base pressure of the unbaked vacuum chamber was 10 -5 Pa. Tin dripping experiments were typically conducted at vacuum pressures of around 10 -4 Pa. Before pump-down several pieces of tin (mass each between ~100 mg and ~170 mg) were placed on an insertible Mo trip tray with two beveled slots (2 mm width) for dripping. The pieces melted in vacuum during slow heating of the tray to temperatures above the melting point of Sn and generally contracted while on the tray into elongated or round balls at tray temperatures of about 240 -250 °C before dripping through the slots at slightly higher temperatures [20]. For spheres of liquid tin this resulted in droplet diameters in the range of 3.0 -3.6 mm.…”

“…Previously, an embrittlement process for solidified tin drops resulting from induction of a phase transformation was described by one of us [19]. Based on this scheme, we have recently reported first tests of an alternative in-situ optics cleaning scheme carried out in a vacuum chamber where massive tin drop contamination was examined on mirror samples coated with uncapped Mo/Si multilayers [20]. Complete disintegration of tin drops could be initiated by means of induction of tin pest when cooling the substrates to negative Celsius temperatures [21,22].…”

“…In this process, tin in its normal metallic phase, β-Sn, undergoes an allotropic phase change to a lower-density semiconducting phase, α-Sn, which is stable at temperatures below 13 °C [21]. The transformation occurs by a process of nucleation and growth that is induced by bringing seed crystals of α-Sn in contact with metallic tin [19][20][21][22][23]. Tin deposits on ML optics crack and detach during the transition, since the new phase takes up considerably more volume.…”

“…To our previously described apparatus [20] we have added the option to heat the sample holder during tin dripping in order to mimic the heating of the collector surface by the hot plasma in a commercial EUV source. On the other hand, by cooling the substrates to 6 negative Celsius temperatures we now explore a parameter range that has not been investigated before in the context of tin dripping.…”