Analysis of the Si(111)5*5-2*1 phase boundary

Abstract

Expxhentd scanning hmnelling microscopy data for the Si(II1)S x 5-2 x 1 phase boundary are analysed with reference to lhe structure of the 2 x I surface. The height change data are subjed to significant correction facton, making interpretation difficult due to the small theoretical differences involved. The theoxtical lateral differences are much larger. being 3.3 A, and these are analysed.

Surface structures on cleaved silicon by scanning tunnelling microscopy

“…Therefore, the large value of the observed [112]-direction corrugation is very difficult to reconcile with the MP model, although it arises naturally from the TBS structure. In this connection we note that STM height scans performed across a 5 × 5 to 2 × 1 phase boundary [35] were consistent with the observed lateral space between the structures using the TBS model, but there was a 0.33 nm lateral anomaly if the MP model was used [21].…”

“…STM senses contours of equal charge density outside the atom cores. Estimates of small height differences in STM studies are subject to many factors, discussed in some detail for the case of a 5 × 5 to 2 × 1 boundary [21], and the quantitative values observed can be affected by tip voltage and scanning conditions as seen from data in [3]. However, atomic steps of theoretical height difference 0.313 nm are readily measured and have been studied in detail [3].…”

“…This is certainly true in the present case, because there are believed to be bonded p-type orbitals on the top (chain) atoms, and sp mixtures, perhaps sp 3 on the MP model, on the atoms between the chains. The height at which a given charge density occurs above an atomic nucleus for a p orbital and for an sp hybrid could differ by as much as about 0.05 nm ( [21] and references therein). However, the required extra 0.12 nm for the MP model seems to be too large to be attainable by this argument.…”

“…This involves single-bond (shuffle plane) (111) cleavage. However, it has been argued that this structure is not in accord with crack healing data [19,20], or the lateral onset of a heat-induced 5 × 5 structure adjoining the 2 × 1 [21], as well as other criticisms [13]. Strong evidence in favour of the MP model was adduced from considerable dispersion observed in the -J direction in angle-resolved photoemission measurements [22,23].…”