Predicting radiation-induced carbon contamination of EUV optics

Predictions are made for the radiation-induced carbon contamination threat to ruthenium-coated extreme ultraviolet (EUV) optics for a range of incident EUV intensities, exposure pressures and types of hydrocarbon. A calculational philosophy is developed that acknowledges the ruthenium capping layer may have adsorbed oxygen on it and that the carbon contamination film is partially hydrogenated. The calculations incorporate the Nitta Multisite Adsorption framework, which accounts for the configurational adsorption difficulty encountered by the adsorption of large molecules on surfaces. Contributions from “out-of-band” radiation are included, both in the direct photon-induced dissociation of hydrocarbon molecules and in the out-of-band production of secondary electrons. For the hydrocarbon molecules, n-tetradecane, n-dodecane, n-decane, and benzene, for a range of EUV powers and hydrocarbon pressures, predictions are made for carbon thicknesses, the overall carbon deposition rates, and the relative amounts of contamination produced by primary photon excitation, secondary electrons, and out-of-band radiation. The comparison is made to relevant prior experiments. The model, with no adjustable parameters, provides a good account of prior experiments on n-tetradecane, n-decane, and benzene over the pressure ranges examined by the experiments (∼1 × 10−10 to ∼1 × 10−7 Torr) and over the EUV intensity range 0.001–100 mW/mm2. The level of agreement is within a factor of ∼4 or better, which is consistent with expectations based on the experimental uncertainties. Comparison with prior data for n-decane indicates that the carbon deposit produced by the EUV-induced dissociation of hydrocarbons is substantially hydrogenated. Out-of-band radiation accounts for ∼9%–12% of the overall optic contamination. Secondary electrons account for ∼2% of the overall optic contamination. The results show that the dominant mechanistic cause of the EUV carbon contamination is primary photon absorption by the adsorbed hydrocarbon molecule. The removal of carbon or hydrogen by electron stimulated desorption due to secondary electrons or photon stimulated desorption by primary EUV absorption can be safely ignored as negligible compared to the EUV-induced carbon deposition rate. The results allow comparison with past experiments, provide a framework for conducting future experiments, and predict contamination threats relevant for practical EUV lithography tool operation. The calculations also clarify the underlying physical phenomena at work in the EUV carbon contamination problem.