Charge transfer and asymmetry on Ge(111)-c(2×8) studied by scanning tunneling microscopy

Abstract

Scanning-tunneling-microscopy images of annealed Ge(111) show large c(2&8) domains with protrusions that match closely the arrangement of adatom and rest-atom dangling bonds in the simple adatom model. The unit cell, however, appears asymmetric. or "buckled"; namely, the two adatoms or the two rest atoms within the unit cell show significant diA'erences in apparent height. Both the adatom and rest-atom dangling bonds can be simultaneously observed with either bias polarity, indicating that both are partially filled. These results will be discussed in relation to the proposed charge transfer between the adatoms and rest atoms.The Si(111)-(7X7)surface appears to be nearly fully understood, yet there remain unresolved issues for the closely related Ge(111)-c(2X8) surface. The geometrical structure of the Ge(111)-c(2X8) surface is generally considered to be optimally described by a simple adatom model in which the (1 x 1) substrate is decorated with germanium adatoms in alternating (2 x 2) and c (2 x 4) subunits. In this model, each c(2X8) primitive unit cell of eight (1 xi) sites contains two adatoms and two rest atoms, totaling four dangling bonds. Verification of this model has been provided by studies using x-ray diffraction, ' scanning tunneling microscopy (STM), and medium-energy ion scattering.In the STM study of Becker et al. , a pattern of protrusions was observed which matches the arrangement of either the rest atoms or the adatoms, depending of the sample-to-tip bias polarity, but not both simultaneously This o. bservation, combined with information available from a model calculation, ' lead to the following suggestions: (i) The simple adatom model is valid. (ii) There is a complete electron transfer from the adatoms to the rest atoms resulting in a fully occupied rest-atom dangling-bond state and a completely empty adatotn dangling-bond state. A surface electronic band at a binding energy of about 0.8 eV observed by ultraviolet photoemission was ascribed to the occupied restatom dangling-bond state in analogy to the Si(ill)-(7 x 7) case. Two questions remain concerning the above tnodel. The first question arose from a more recent photoemission study revealing the existence of another occupied surface-state band with a dangling-bond character, which could not fit into the model. The second question had to do with the high Coulomb energy needed for a complete charge transfer from the adatoms to the rest atoms. A fully ionic surface appears to be highly unlikely 6 In the work reported here, we examined the Ge(111)-c(2X8) surface with STM under various bias conditions and found features not reported before, which could provide an explanation in connection with the above-mentioned questions. Our results confirm that the simple adatom model is valid except that the c(2X8) unit cell appears "buckled" or asymmetric; namely, the two adatoms or the two rest atoms in a unit cell show significant differences in apparent height depending on the imaging conditions. This observation is somewhat similar to the familiar c...

Electronic structure ofGe(111)c(2×8): STM, angle-resolved photoemission, and theory

“…3͑a͒ shows only the adatoms. This is in agreement with previous STM studies 3,4 and theoretical results by Klitsner and Nelson 20 that the primary unoccupied surface states are strongly localized on the adatom sites. The c͑2 ϫ 8͒ unit cell is also indicated in the image.…”

“…In the 0.7 theo and 1.0 theo V images, to which both the R R and the R T states contribute, the two types of rest atoms have a similar brightness. The rest atom inequivalence was also observed in the STM study performed by Lee et al 25 and Hirschorn et al 4 on the Ge͑111͒c͑2 ϫ 8͒ surface at small negative sample voltages ͑−0.5-−0.7 V͒. Based on the comparison between the experimental and theoretical results we conclude that the upper A2 band corresponds to the R R rest atom dangling bonds and that the lower A2Ј band originates from the dangling bonds of the R T rest atoms.…”

“…This suggested that there is a complete electron transfer from the adatoms to the rest atoms. An STM study by Hirschorn et al 4 confirmed that the simple adatom model is valid. The rest atoms were the dominant features in the images taken with a negative sample bias.…”

“…Figure 1͑a͒ shows the accepted model of the c͑2 ϫ 8͒ structure consisting of onefourth monolayer ͑ML͒ of adatoms. 3,4 Each adatom bonds to three atoms in the first full layer, thus saturating three-fourths of the surface dangling bonds. The surface is further stabilized by a charge transfer from the one-fourth ML of adatom dangling bonds to the remaining one-fourth ML of unsaturated dangling bonds on the rest atoms ͑atoms of the first layer not bonding to adatoms͒.…”

Reconstructions and phase transitions at semiconductor surfaces: Ge(111)

“…Section 2 deals with the structure and electronic properties of the low temperature stable c(2 X 8) structure. Our calculations show that asymmetries are present in the surface unit cell, which can explain low energy electron diffraction (LEED) [8] and scanning tunneling microscopy (STM) [9] data. Results pertinent to the c(2 × 8) ~ (1 X 1) disordering transition at T ~ 300°C are presented in Section 3.…”