Role of boron and (√3 × √3)‐B surface defects on the growth mode of Si on Si(111): A photoemission and electron diffraction study

Abstract

The influence of boron on Si molecular beam epitaxy was investigated as function of coverage and temperature by reflection high‐energy electron diffraction (RHEED). The development of the boron‐covered Si surface was studied additionally by ultraviolet photo electron spectroscopy (UPS) as function of boron coverage (cB) and annealing condition. We found a direct correlation between the appearance of surface states in UPS and the transient growth behaviour observed in RHEED. For cB > 0.4 monolayer (ML) regular RHEED oscillations occur with a period typically for two bilayers (BLs), whereas for lower cB a transient behaviour with irregular intensity oscillations was observed in the initial growth stages. The appearance of this transient behaviour is discussed in terms of an initial surface defect‐induced nucleation of BL‐high Si islands and the formation of two BLs‐high Si islands on top of the van der Waals‐like perfect boron‐covered surface, respectively. The occurrence of surface defects for cB < 0.6 ML is clearly established in UPS measurements, where an intensity peak slightly below the Fermi level (−0.4 eV) was visible. We suggest that this surface defect level is related to Si adatom dangling bonds acting also as preferential nucleation centres in epitaxy. Furthermore, surface defects results in a significant Fermi level pinning.

Towards controlled molecular beam epitaxial growth of artificially stacked Si: Study of boron adsorption and surface segregation on Si(111)

“…1(c), the shift of the back-bond state indicates a shift of the Fermi level towards the valence band maximum (VBM). The shift is more distinct seen for the Si bulk states, as shown recently [8]. The shift of the Fermi level for the B-covered surfaces could be attributed by the upwards band bending at the surface [8].…”

“…The shift is more distinct seen for the Si bulk states, as shown recently [8]. The shift of the Fermi level for the B-covered surfaces could be attributed by the upwards band bending at the surface [8]. Since the shift of the Fermi level is constant for 0.3 rc B r0.43 ML, it can be further suggested that the Fermi level is pinned at the surface defects (dangling bond states).…”

“…The topmost curve shows an UPS spectrum of the initial Si(1 1 1)(7 Â 7) surface structure. was observed, which indicates the formation of surface defects [8]. Annealing at 1130 K for 10 min already leads to a significant shift of about 0.3 eV towards lower energy (Fig.…”

“…This is accompanied by the development of a well-pronounced surface state at around À 1.45 eV, indicating a better arrangement of the surface structure, where B at the surface is only within the subsurface sites. However, a peak related to surface defects is visible at around À0.4 eV indicating that there is not sufficient B accumulated at the surface during annealing for 10 min at T up to 1040 K. This is accompanied by a shift of the spectra by 0.3 eV to the low-energy side resulting from the Fermi level pinning due to Si dangling bonds, as explained recently [8].…”

“…B was evaporated from an effusion cell at T42000 K, corresponding to a deposition rate of around 0.02 ML/min. RHEED intensity measurements during the Si(1 1 1)(O3 Â O3)R301-B surface superstructure formation were used to determine c B , as described elsewhere [6,8]. UPS measurements were conducted in the same multichamber vacuum system at a working pressure of 2 Â 10 À 8 Torr using a He discharge source (Specs 10/35), with unpolarized light of 21.2 eV photon energy (He I a ).…”