Sticking probabilities

In all experiments designed to determine aj-, ay, and the energy contents of the atoms and molecules are scrutinized before and after scattering. Ultimately, we would like to study the scattering species while it is on the surface during the collision process to learn about the details of bonding and local geometries on the 

TABLE 4.4. Structure Sensitivity of H2 D2 Exchange at Low Pressures ( 10 torr) 

The Sticking probability may decrease with increasing temperature or remain unchanged. When molecules must dissociate in order to adsorb on a surface, under certain experimental conditions S may increase with increasing temperature, indicating that there is an activation energy for adsorption. 

---

The rate of physical adsorption may be determined by the gas kinetic surface collision frequency as modified by the variation of sticking probability with surface coverage—as in the kinetic derivation of the Langmuir equation (Section XVII-3A)—and should then be very large unless the gas pressure is small. Alternatively, the rate may be governed by boundary layer diffusion, a slower process in general. Such aspects are mentioned in Ref. 146. 

As with any system, there are complications in the details. The CO sticking probability is high and constant until a 0 of about 0.5, but then drops rapidly [306a]. Practical catalysts often consist of nanometer size particles supported on an oxide such as alumina or silica. Different crystal facets behave differently and RAIRS spectroscopy reveals that CO may adsorb with various kinds of bonding and on various kinds of sites (three-fold hollow, bridging, linear) [307]. See Ref 309 for a discussion of some debates on the matter. In the case of Pd crystallites on a-Al203, it is proposed that CO impinging on the support... 

Figure Al.7.8. Sticking probability as a fimction of surface coverage for tliree different adsorption models.

If adsorption occurs via a physisorbed precursor, then the sticking probability at low coverages will be enhanced due to the ability of the precursor to diflfiise and find a lattice site [30]. The details depend on parameters such as strength of the lateral interactions between the adsorbates and the relative rates of desorption and reaction of the precursor. In figure Al.7,8 an example of a plot of S versus 0 for precursor mediated adsorption is presented. 

It shows that sticking is proportional to the availability of empty sites (because there are no lateral interactions in the adsorbate), and the sticking probabilities, Sy and S, are weighted by the fraction of the adsorbate-free surface that is reconstructed or not. This can obviously introduce a substantial temperature dependence in the sticking coefficient. 

If we now assume that this surface at temperature T is in equilibrium with a gas then the adsorption rate equals the desorption rate. Since the atoms/molecules are physisorbed in a weak adsorption potential there are no barriers and the sticking coefficient (the probability that a molecule adsorbs) is unity. This is not entirely consistent since there is an entropic barrier to direct adsorption on a specific site from the gas phase. Nevertheless, a lower sticking probability does not change the overall characteristics of the model. Hence, at equilibrium we have... 

Methane is a stable molecule and therefore hard to activate. As a result the sticking probability for dissociative chemisorption is small, of the order of 10 only, and ruthenium is more reactive than nickel. However, a stretched overlayer of nickel is significantly more active than nickel in its common form, in agreement with expectation. 

Frequently, adsorption proceeds via a mobile precursor, in which the adsorbate diffuses over the surface in a physisorbed state before finding a free site. In such cases the rate of adsorption and the sticking coefficient are constant until a relatively high coverage is reached, after which the sticking probability declines rapidly. If the precursor resides only on empty surface sites it is called an intrinsic precursor, while if it exits on already occupied sites it is called extrinsic. Here we simply note such effects, without further discussion. 

Figure 4. Variation of the relative sticking probability, s/s, with 6(H), for H on Ni(110). The dashed line refers to ex-...

An linportant aspect of the present work is the dependence of the sticking probability on the surface coverage of an adsorbate. This will be the topic of the second part of the discussion. 

Figure 3b. The sticking probability of methanol on a Cu(llO) surface predosed with half a monolayer of oxygen. There is an induction period to adsorption taking place and formaldehyde is evolved coincidentally with the sticking. Adsorption temperature of 353 K.

The plasma-wall interaction of the neutral particles is described by a so-called sticking model [136, 137]. In this model only the radicals react with the surface, while nonradical neutrals (H2, SiHa, and Si H2 +2) are reflected into the discharge. The surface reaction and sticking probability of each radical must be specified. The nature (material, roughness) and the temperature of the surface will influence the surface reaction probabilities. Perrin et al. [136] and Matsuda et al. [137] have shown that the surface reaction coefficient of SiH3 is temperature-independent at a value of = 0.26 0.05 at a growing a-Si H surface in a... 

Nienhuis [189] has used a fitting procedure for the seven most sensitive elementary parameters (reactions SiH4 -t- SiH2 and Si2H6 -I- SiHi, dissociation branching ratio of SiH4, surface reaction coefficient and sticking probability of SiHa, and diffusion coefficients of SiH and H). In order to reduce the discrep-... 

Further improvement of the previously described model [74] allowed the inclusion of a sticking probability a superimposed to the aggregation of both carbon and nitrogen (C and N species in the model). This was done to represent the block-... 

---