We have examined the Ge(111)-c(2×8) and related surfaces in detail with the scanning tunneling microscope. A comparison of tunneling images in both filled and empty states for the Ge(111) surface with similar images for Si(111)-(5×5), -(7×7), and -(9×9) surfaces shows the surface structure to be inconsistent with recent models, such as the dimer chain proposed by Takayanagi and collaborators. Comparison with filled- or empty-state images for the laser-stabilized Si(111) surface shows that both this surface and the Ge(111)-c(2×8) are most consistent with a simple model of alternating rows of 2×2 and c(4×2) adatoms on sites on a 1×1 substrate. We find no evidence in the tunneling images of Ge(111)-c(2×8) to support the existence of a subsurface stacking fault. Electronic and geometric sources for the contrast in the tunneling images are considered and it is concluded that electronic structure forms the basis for the images of adatom-reconstructed semiconductor surfaces.
- Received 12 July 1988
- Published in the issue dated 15 January 1989
© 1989 The American Physical Society