Dynamics of the chemisorption of N2 on W(100): Precursor-mediated and activated dissociation

Publisher Website
Google Scholar
Abstract
The dissociativechemisorption probability of N2 on W(100) is found to proceed by way of two dynamically distinct channels. At low kinetic energies E i , dissociation proceeds primarily by way of a precursor‐mediated process, where the dissociation probability is found to fall with increasing E i , reflecting the energy dependence of the trapping probability into this state. Dissociation at low energies is also strongly dependent on surface temperature T s which effects the fraction of trapped species that desorb. For energies above about 0.45 eV, the dissociation probability is found to rise from a minimum of about 0.14 at T s =800 K to over 0.45 at E i =5 eV. Over this range we believe that kinetic energy enables the incident molecules to directly overcome a barrier in the reaction coordinate. Throughout the entire range of energies we observe only slight variations of the dissociation probability with the angle of incidence, with no discernible sensitivity for energies below ∼0.5 eV. For energies between 1 and 4 eV, associated with the ‘‘activated’’ channel, we observe a slight preference for non‐normal incidence, with a clear preference for normal incidence only for E i >5 eV. While the ‘‘precursor‐mediated’’ channel displays a considerable sensitively to surface temperature, results at high energy are found to be essentially independent of this parameter. Moreover, dissociation by way of the precursor‐mediated channel is found to be insensitive to surface coverage, in contrast to a roughly linear decrease in the dissociation probability with surface coverage observed for dissociation via the activated process. In this latter case, we find that the saturation coverage remains approximately constant at about 0.6 atomic monolayers for all conditions, up to the highest incidence energies. This differs from previous observations for the dissociation of O2 and N2 on W(110), where the saturation coverage was found to rise with increasing E i . Finally we find that the dissociation probability vs kinetic energy curve for the ‘‘direct’’ dissociation case is qualitatively similar to that for the N2/W(110) system, but with a threshold that is ∼0.4 eV lower. We argue that the ‘‘precursor‐mediated’’ mechanism does not contribute significantly to dissociativechemisorption in the W(110) case and conclude that the primary difference between N2dissociation on the W(110) and W(100) surfaces is that the barrier to dissociation is slightly higher in the W(110) case.