Characterization of Surface Morphology in Epitaxial Growth
- PDF / 2,490,452 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 90 Downloads / 202 Views
95 Mat. Res. Soc. Symp. Proc. Vol. 399 01996 Materials Research Society
increased with coverage. In addition, the experimental RHEED profiles exhibited a peak at finite momentum transfer k1l corresponding to the mound separation. MODEL AND SIMULATIONS In order to try to reconcile these results we have carried out simulations of RHEED and HRLEED profiles using a model appropriate for Fe/Fe(100) growth at room temperature (see [11] and [12]) in which the correct bcc crystal geometry has been taken into account. This model has already been shown [11], [12] to reproduce the observed mound angle and coarsening behavior observed in Ref. 10 and is described briefly below. In our model, atoms are incorporated only at the four-fold hollow sites formed by the four nearest-neighbor atoms in the layer below. Atoms which land on a site which is not a four-fold hollow site, funnel downwards [13] until they reach the nearest four-fold hollow site below. Atoms with no lateral bonds may diffuse at a rate D = Doe-Ea,/kBT while for atoms which try to diffuse down a step to the layer below, there is an extra step-barrier energy EB. While atoms with two-or-more lateral bonds are assumed to be immobile at room temperature [14], atoms with one bond may diffuse along the edge of an island (edge-diffusion) at a rate D, = De-Ee/kBT. Values of the monomer diffusion rate (D _ 3100 sec-'), edge-diffusion barrier (E, -= 0.125 eV), and step barrier (EB = 0.07 eV) were selected which reproduce the observed submonolayer island density and submonolayer square island morphology [14], [15] as well as the measured r.m.s. surface width w in the first few layers of growth [16]. In our simulations, the antiphase kinematic HRLEED intensity IH(7r, k1l) was calculated by taking into account scattering from all surface atoms (including atoms just below a four-fold hollow) using the expression, IH(7r, kjl) =
0 reflects the realistic assumption that sites which lie "below" nearby sites will scatter with a reduced intensity amplitude while taking r = 1/4 implies that atoms which lie below the four-fold hollow sites of the (100) surface are completely shielded by the atoms above. We have carried out simulations with different values of r but find that our qualitative results are unaffected by the particular value used. In contrast to HRLEED, in low-angle RHEED shadowing may play an important role. For the case of the Fe/Fe(100) experiment the RHEED scattering takes place at an angle of 30 from the surface while the typical mound angle is of the order of 13' [10]. As a consequence, the dominant contribution comes from the peaks of the highest mounds and there is very little scattering from the valleys. In order to mimic this effect in our simulations the antiphase
Data Loading...
