Atomic geometry of semiconductor surfaces

Abstract While considerations of the space‐group symmetries of semiconductor surfaces as inferred from elastic low energy electron diffraction (ELEED) spot patterns have been given for decades,1 only during the past 2 years have quantitative analyses of ELEED intensities been performed to determine the atomic geometry of semiconductor surfaces. Specific systems which have been studied using ELEED intensity analyses include the {100} faces of rocksalt structure the low index faces of the wurtzite structure semiconductor Zn06–9 the cleavage faces of several zincblende structure semiconductors,9–14 the {100}15–17 and {11 l}18.19 faces of silicon, and the {OOOl} faces of MoSZ and NbSel.zo.zL The important finding emerging from these analyses is the common occurrence of substantial deviations of the actual surface structures from those characteristic of a truncated but otherwise “bulk” solid: a phenomenon which is referred as “reconstruction” (or “rearrangement”) of the surface. Moreover, these reconstructions both are pervasive (i.e., occurring for all but possibly the most ionic materials)3–5 and involve large (Δa∼0.5 Å) measurements of atomic species in the surface layer and possibly in the layers beneath as well. These results, which are in marked contrast to those obtained for metals,22 are manifestations of the differences between metallic, covalent, ionic and Van der Wads' bonding. Indeed, it seems possible22 that a covalent‐ionic surface structural phase transition occurs within the tetrahedrally coordinated compound semiconductors, all of which exhibit covalent bonding from the perspective of their bulk structure.23