Oxygen sticking coefficient

More recently, a kinetic model for describing the occurrence of temporal oscillations on Pt( 110) was developed, which is more simplified than that originally derived for Pt(100), but nevertheless reproduces the qualitative features well and enables, in addition, analysis of further effects, such as forced oscillations, etc. (69). It starts from Eqs. (4) and (5) outlined previously, which were the basis for successful description of the overall kinetics. The system was then extended in order to include the CO-induced structural transformation of the surface as well as the different oxygen sticking coefficients on the two modifications. 

Equation (5) was modified in order to take into account the different oxygen sticking coefficients on the 1 x 1 and 1 x 2 patches. If their ratio is denoted by a [which was taken to be 1.5 from the experimental estimates (66)], the temporal variation of the oxygen coverage now reads... 

Both limits will increase with increasing p0, however, with different slopes, so that an intersection occurs that marks the lower limit of pQ, and pco, below which oscillations are no longer possible. The difference between these slopes is primarily due to the difference in oxygen sticking coefficient (rate of removal of adsorbed CO) for the reconstructed and nonreconstructed surface phases. 

The mechanism of the kinetic oscillations occurring with the CO + 07 reaction on clean Pt( 100) and Pt( 110) surfaces was based on the reversible transformation of the surface structure by the presence of adsorbed CO and by an associated variation of the oxygen sticking coefficient that increased upon CO-induced lifting of the reconstruction of the clean surface. The most densely packed Pt(lll) surface is not reconstructed and its structure is also not affected by CO adsorption. Accordingly, kinetic oscillations with a clean Pt(lll) surface (i.e., for partial pressure <10 3 torr) could never be observed (13, 26, 27, 38). Again no reconstruction... 

Fig. 23. Oxygen sticking coefficient on [lll 2 x 1 silicon surfaces as a function of step density (expressed as tana). The data also show the influence of a hot filament on the sticking coefficient (after Ibach et al. [231]). , Auger spectroscopy ot surface vibrations A, ellipsometry Q, on 7 X 7 surface.

No kinetic oscillations, rather just bistability, could be detected with the Pt(l 11) surface, whereas the other orientations of the crystal showed bistability and oscillations [8]. The reason for this is that the lifting of the reconstruction enhances the oxygen sticking coefficient. Consequently, oscillatory behavior occurs when the partial pressures pco and po are adjusted in such a way that the CO adsorption rate is faster (slower) than the O2 uptake on the reconstructed (1x1) surface. The reconstructed surface will take up CO and the 1 X 1 will form, which adsorbs oxygen faster, which, in turn, reduces the CO coverage by reactive removal. (Note that CO adsorption is not noticeably structure sensitive.)... 

Equation (4d) provides a negative feedback loop, since the oxygen sticking coefficient on the reconstructed surface Sr is smaller than on the 1 x 1 phase (si). The function f u) takes account of the fact that the lifting of the reconstruction does not start until a certain threshold value of u is reached (0.2 for Pt(l 10)). This delay in the inhibitor production can cause pronounced changes in the spatial structures compared to standard models (see Section 3.4). 

I ig. 25. Simulation of solitary pulses. A defect with slightly higher oxygen sticking coefficient on the reconstructed surface is located around x = 40 /j,m. The same defect can explain wave splitting, soliton-like behavior and partial annihilation, (c) differs from (b) only in a stronger asymmetry of the initial conditions after [115]. 

---