Recombinative thermal desorption of hydrogen from Si (100) surfaces has been extensively studied from a kinetics as well as dynamics point of view. The measured translational energy distribution was found to be well related to the adsorption dynamics, strongly supporting the validity of detailed balance [1, 2]. The kinetics experiments so far done prefer the 2H, 3H, and 4H interdimer mechanism to the 2H* intradimer mechanism. Yet, the temperature-programmed-desorption (TPD) spectra of hydrogen molecules from monohydride Si(100) surface have not been fully understood because of the problem of cross-over coverage.
We measured D2 TPD spectra under various experimental conditions changing initial D coverage and desorption angle. As a result, the TPD spectra were found to be hardly affected by those parameters within the experimental error bars. This result seems to imply only single desorption kinetic mechanism governs the whole TPD process. Indeed, the desorption dynamics experiments showed up a feature that the 4H path is quite minor compared to the 2H (+3H) path. We invoked the diffusion-mediated desorption (DMD) mechanism by which H adatoms diffuse over the surface via hopping to dangling bonds. Such diffusion was considered to render Si-H bonds in the vibrationally excited states, thereby effectively reducing the desorption energy. Hence the DMD occurs at such sites where unoccupied Si dimers are present closely to occupied Si dimers. Taking into the clustering effect of occupied Si dimers, we counted up configurations of unoccupied Si dimers packed with occupied dimers as a function of D coverage. Thence, the experimental TPD spectrum was fitted with simulation based on a rate equation after choosing suitable values for the desorption energies and prefactors for various kinetic channels. The TPD curve from the 1 ML D/Si(100) surface could be explained reasonably, but we observed considerable TPD peak shift to a higher temperature with decreasing coverage, deviating from the expected first-order rate law.
[1] M. Durr and U. Hofer, Surf. Sci. Rep. 61(2006)465.
[2] A. Namiki, Prog. Surf. Sci. 81(2006)337.
|