Multilayer development for the generation beyond EUV: 6.x nm

Radiation of 6.x nm wavelength is a possible candidate for the next generation of optical lithography aiming at sub 8 nm resolution in single exposure. Construction of multilayer optics based on La and B remains a challenge due to a stronger requirement for layer thickness and interface roughness. The main challenge is to prevent intermixing of the multilayer materials, particularly since individual layer thicknesses (~1.5 nm) are in the same range as the naturally formed compounded interfaces. In recent research we have shown that the formation of La-nitride by N-ion treatment of the La layers reduces compound formation, providing a better optical contrast, thus leading to an increase of the multilayer mirror reflectivity. We also argue that LaN is optically preferred over pure La. Mirrors for 6.x nm require more than 150 periods for a maximum theoretical peak reflectance of 78% for an ideal structure. To deposit such an amount of periods the deposition technique is being optimized. A first optimization of the deposition process has enabled full stack (150 periods) LaN/B4C multilayer mirror samples yielding 47.2% reflectivity at l=6.63 nm (Fig. 1) at near normal incidence. Further optimization of the deposition and nitridation technique is planned aiming to further reduce interface compound formation and roughness value. First of all the magnetron plasma should be optimized in order maximize plasma ion polishing effects and minimize layers intermixture induced by high energy ions and neutrals. Further interface roughness reduction can be achieved by tuning the energy of the atoms being deposited and by low energy ion treatment of the deposited layers