III-V compound semiconductor (001) surfaces

There has been renewed interest in the structure of III-V compound semiconductor (001) surfaces caused by recent experimental and theoretical findings, which indicate that geometries different from the seemingly well-established dimer models describe the surface ground state for specific preparation conditions. I review briefly the structure information available on the (001) surfaces of GaP, InP, GaAs and InAs. These data are complemented with first-principles total-energy calculations. The calculated surface phase diagrams are used to explain the experimental data and reveal that the stability of specific surface structures depends largely on the relative size of the surface constituents. Several structural models for the Ga-rich GaAs (001)(4×6) surface are discussed, but dismissed on energetic grounds. I discuss in some detail the electronic properties of the recently proposed cation-rich GaAs (001)ζ(4×2) geometry. Received: 18 May 2001 / Revised version: 23 July 2001 / Published online: 3 April 2002     

Ad-atom interactions with III-V semiconductor surfaces

Photoemission techniques (core level, valence band and partial yield spectroscopies) using synchrotron radiation as the excitation source have been applied to study the changes in the surface electronic structure of the (110) cleavage face of III-V semiconductor surfaces as a function of different ad-atom coverages. In this paper we concentrate on Au overlayers on GaSb and in particular address the problem of the mechanism for Fermi level pinning and the formation of Schottky barrier heights. It appears that the Fermi level pinning is fully established at a small fraction of a monolayer coverage. Core level spectroscopy gives evidence for a strong metal-semiconductor interaction leading to decomposition of GaSb at the interface. The experimental results form the basis for proposing a new model for the Schottky barrier based on defect states at the interface.     

Subsurface strain in the Ge(001) and Ge(111) surfaces and comparison to silicon

The subsurface strain associated with surface reconstruction was measured for the Ge(001)-c(4×2) and Ge(111)-c(2×8) surfaces using high energy ion scattering. In the case of the Ge(001) surface we find the equivalent of ∼3 monolayers displaced by more than 0.12 Å, in accord with dimer models of the surface reconstruction. For the Ge(111) surface displacements are observed in off-normal incidence, indicating large displacements perpendicular to the surface or other reconstructions, such as a stacking fault configuration. The relationship between subsurface strain and stacking fault models is also discussed. The subsurface strain in these two Ge surfaces is remarkably similar to that of the corresponding Si surfaces, even though the details of the surface reconstruction are different. Measurements at low temperature indicate that the strain is essentially temperature independent, as expected. Measurements of the hydrogen covered surfaces show little change is strain, a surprising result when compared to the behavior of Si(001).     

Surface vibrations of Na(001) and K(001) surfaces

We present the results of a theoretical study of the dynamics of the atom motion of Na(001) and K(001) surfaces. The total electronic energy is calculated using a pseudopotential approach with a confined electron gas as unperturbed system. With this theory the dynamical matrix can he derived without resorting to empirical parametrizations. Surface phonon dispersion curves are reported for the high symmetry directions of the two-dimensional Brillouin zone for ideal and relaxed configurations. The calculated spectra are compared with the results of semi-empirical force constant calculations. The effects of single and multilayer relaxations on the location and the nature of the main surface bands are examined.     

Structural phase transition on Si(001) and Ge(001) surfaces

Among a variety of solid surfaces, Si(001) and Ge(001) have been most extensively studied. Although they seem to be rather simple systems, there have been many conflicting arguments about the atomic structure on these surfaces. We first present experimental evidence indicating that the buckled dimer is the basic building block and that the structural phase transition between the low-temperature c(4x2) structure and the high-temperature (2x1) structure is of the order-disorder type. We then review recent theoretical work on this phase transition. The real system is mapped onto a model Ising-spin system and the interaction parameters are derived from total-energy calculations for different arrangements of buckled dimers. The calculated critical temperature agrees reasonably well with the experimental one. It is pointed out that the nature of the phase transition is crucially affected by a small amount of defects on the real surfaces.     

Subsurface As-layer formation in Au films deposited on GaAs(001)

Au deposited at room temperature on MBE grown GaAs (001) has been studied by UPS, XPS and AES. At Au coverages below ～2Å the adsorption interaction spreads the Au (6s) states density over the width of the GaAs valence band, while at higher coverages the Au overlayer has metallic properties. For the thick Au overlayers a subsurface As-rich region is established. It is suggested that this has a stabilizing influence on the continuous Au overlayer.     

Asymmetric versus symmetric dimerization on the Si(001) and As/Si(001)2 × 1 reconstructed surfaces as observed by grazing incidence X-ray diffraction

The atomic structures of the clean Si(001) and As/Si(001)2 × 1, 1 × 2 reconstructed surfaces prepared in situ have been obtained by grazing incidence X-ray diffraction under ultra high vacuum. In the former case an asymmetric dimer inclined by 7.4° on the surface plane with a slightly contracted dimer bond length (−1.5% compared to the bulk value) has been indentified as the structural basis, whereas a symmetric dimer is found after adsorption of one monolayer of As at 350 ° C. The induced displacements in the first subsurface silicon layer have been derived in both cases. These results are compatible with the models proposed on the basis of spectroscopic and direct imaging methods.     

Vibrations at 3C-SiC(001)-(32) surfaces

Si-terminated 3C-SiC(001) surfaces with () and ()reconstructions were investigated by high-resolution electron energy-loss spectroscopy (HREELS), low-energy electron diffraction (LEED) and Auger electron spectroscopy. The surfaces were prepared by subsequent annealing steps after cleaning by heating in a Si flux. At ()-reconstructed surfaces, the HREELS intensity increases while the widths of the loss lines decrease with proceeding preparation. Eventually, weak loss structures at 380 and 700 cm^-1 are detected besides the strong Fuchs-Kliewer phonon loss peaks. They are attributed to surface-localized vibrations, i.e., to stretching modes of on-top Si dimers and of C-Si-C groups, respectively. The weak features vanish after exposure to atomic deuterium, but reappear after subsequent annealing. At () reconstructed surfaces the HREELS lines are broadened and no surface-localized modes were resolved. Received 18 January 1999     

Instability of one-dimensional dangling-bond wires on H-passivated C(001), Si(001), and Ge(001) surfaces

We investigate the instability of one-dimensional dangling-bond (DB) wires fabricated on the H-terminated C(001), Si(001), and Ge(001) surfaces by using density-functional theory calculations. The three DB wires are found to show drastically different couplings between charge, spin, and lattice degrees of freedom, resulting in an insulating ground state. The C DB wire has an antiferromagnetic spin coupling between unpaired DB electrons, caused by strong electron–electron interactions, whereas the Ge DB wire has a strong charge-lattice coupling, yielding a Peierls-like lattice distortion. For the Si DB wire, the antiferromagnetic spin ordering and the Peierls instability are highly competing with each other. The physical origin of such disparate features in the three DB wires can be traced to the different degree of localization of 2p, 3p, and 4p DB orbitals.