Surface reconstructions of InSb(100) observed by scanning tunneling microscopy

Abstract

A number of surface reconstructions formed on InSb(100) have been observed with atomic resolution using scanning tunneling microscopy (STM). Cycles of low-energy ion bombardment and annealing result in the formation of an In-rich surface with a c(8 X2) diffraction pattern. Filled-states images of this indicate a (4X1) periodicity arising from imaging occupied lone-pair orbitals on exposed second-layer Sb atoms. The controlled deposition of Sb4 from a Knudsen effusion source onto the c(8 X2) surface at elevated temperatures led to the formation of a (1X1), c(4X4), and an asymmetric (1X3)structure, in order of decreasing Sb/In ratio. Filledand empty-states STM images of all these Sb-rich reconstructions were generated and the coexistence of the different structures was also observed. Both the (1X1) and asymmetric (1 X 3) surfaces showed a high degree of disorder while the c(4 X 4) surface was highly ordered. The c (4 X4) structure, which involves the chemisorption of Sb onto an already Sb-terminated surface, is characterized by blocks of three pairs of well-resolved Sb dimers. Structural models are proposed for the ordered structures observed.

Bismuth-stabilizedc(2×6)reconstruction on a InSb(100) substrate: Violation of the electron counting model

“…It describes well the previous experimental findings of the ͑4 ϫ 3͒ and c͑8 ϫ 2͒ reconstructions on the Sb/InSb͑100͒. 5 It is also worth noting that the c͑8 ϫ 2͒ phase is much more stable on the InSb͑100͒ than on the GaSb͑100͒ one ͑Fig. 4͒.…”

“…To our best knowledge, there is no experimental report on the GaSb͑100͒c͑8 ϫ 2͒ formation. The Sb/ InSb͑100͒c͑4 ϫ 4͒ reconstruction appears in experiments, 5 although it is not energetically stable. This can be understood due to finite temperature and configurational entropy, as we show below.…”

“…1,2 However, there are indications that the detailed structures of the ͑1 ϫ 3͒ and c͑2 ϫ 6͒ surfaces have remained unresolved. [3][4][5][6] Indeed, structures, shown in Figs. 2͑a͒-2͑e͒, have been recently proposed for them on the basis of high-resolution STM findings and theoretical calculations.…”

Evidence from scanning tunneling microscopy in support of a structural model for the InSb(001)-c(8×2) surface

“…Comparison with empty states images from chemisorbed Sb-Sb dimers on the c(4ϫ4)-InSb(001) surface clearly indicate the presence of these states. 4 It should also be noted that the empty state image shows the A-C-A features aligned along the ͓110͔ direction, whereas the structural model in Fig. 3 shows the C features lying halfway between two A features in a given row.…”

“…1,2 This has clearly been demonstrated by the controversy surrounding the InSb͑001͒-c(8ϫ2) surface reconstruction. [3][4][5] The initial model for the InSb(001)-c(8ϫ2) surface was proposed by John et al, 6 based on high energy electron diffraction, soft x-ray photoemission data and comparison with the same surface reconstruction previously observed by Biegleson et al for GaAs͑001͒ using scanning tunneling microscopy ͑STM͒. 7 This model had the surface terminated with 0.75 monolayer ͑ML͒ of indium ͓where 1 ML is the In concentration in a (1ϫ1) unreconstructed In-terminated surface͔, made up of blocks of three In dimers and a missing dimer forming a (4ϫ2) block with the dimer bonds parallel to the ͓110͔ direction, on top of a complete Sb layer.…”