Sticking Probability of Ammonia Molecules on Tungsten and 316L Stainless Steel Surfaces

We present measurements of the sticking probability of ammonia on two metals, tungsten and 316L stainless steel, covered with natural surface impurities as they will be used for the international nuclear fusion experimental reactor ITER Using a collimated supersonic molecular beam at two different kinetic energies (55 and 255 meV), varying the sample temperature in the 130-425 K range, and characterizing the surface composition with Auger electron spectroscopy, we observe similar sticking features on both surfaces, consistent with a nondissociative adsorption mediated by two precursors having different trapping probabilities. First, the initial sticking probability decreases with increase in the surface temperature. Second, the sticking probability increases with the surface coverage up to near-saturation coverage, where it declines. Both features cannot be described together with the Kisliuk model (intrinsic + extrinsic precursors with identical trapping probabilities) or the modified Kisliuk model (direct adsorption + extrinsic precursor). Thus, we derive a generalized and separable Kisliuk (GSK) model that is able to reproduce quantitatively these two experimental observations thanks to intrinsic and extrinsic precursors having different trapping probabilities. The GSK model assumes a negligible transfer from the intrinsic precursor to the extrinsic precursor, which allows one to extract precursors kinetic parameters in a two-step analysis. The GSK analysis indicates that the ammonia trapping probability is lower on the bare surface (intrinsic precursor) than on the NH3-covered surface (extrinsic precursor). Furthermore, the barriers between the two precursor wells and the deep adsorption well are found below the vacuum level. Finally, we measure that the sticking probability does not decline to zero, i.e., steady-state sticking is observed with a probability up to 0.15 at a beam energy of 55 meV and a surface temperature of 220 K. This observation is consistent with NH3 multilayer adsorption. These experimental results and their fitting with the GSK model offer the starting point to a predictive determination of the fusion fuel inventory related to ammonia in the international experimental reactor ITER.