Electronic structure of the Si-rich3C−SiC(001)3×2surface

Abstract

The electronic structure of the Si-rich 3C-SiC͑001͒ surface with a single-domain 3ϫ2 long-range order has been studied by angle-resolved photoemission using synchrotron radiation. By identifying the topmost bulk valence band the Fermi level position at the surface is accurately determined to be located at 2.1 eV above the valence-band maximum. Four different surface state bands are clearly identified within the bulk band gap at 1.4Ϯ0.1, 2.3Ϯ0.1, 2.9Ϯ0.1, and 3.8Ϯ0.1 eV below the Fermi level, respectively. These states show no dispersion with photon energy in normal emission and both hydrogen adsorption and Si sublimation make them disappear. Dispersions and symmetry properties of the surface states were determined in detail. All four surface state bands have unusually small dispersions throughout the surface Brillouin zone suggesting strong localization of the electron orbitals within a unit cell and a unique surface bond configuration. The origins of the surface states observed in this study are ascribed to the dangling bonds of both the Si ad-dimers and the second layer Si atoms in the 3ϫ2 surface reconstruction.

Atomic structure determination of the Si-rich β-SiC(001)3×2surface by grazing-incidence x-ray diffraction: A stress-driven reconstruction

“…Many different opposite models have been proposed based on structural and nonstructural experimental techniques such as low-energy electron diffraction ͑LEED͒, reflection highenergy electron diffraction ͑RHEED͒, medium-energy ion scattering ͑MEIS͒, real-space atom-resolved scanning tunneling microscopy ͑STM͒, Auger and photoemission spectroscopies, 16 -24 and also a few ab initio total-energy theoretical calculations. [25][26][27] These include ͑i͒ the double dimer row model ͑DDRM͒ with a surface terminated by a 2 3 monolayer ͑ML͒ Si, 16,12,20,[22][23][24] ͑ii͒ the single dimer row model ͑SDRM͒ terminated by a 1 3 -ML Si, 19 ͑iii͒ the alternate dimer row model ͑ADRM͒ predicted theoretically and having 2 ϫ3 reconstruction with a 1 3 Si ML coverage and asymmetric dimers, 25,26 ͑iv͒ another ADRM having a 3ϫ2 surface array and asymmetric dimers as established by atom-resolved scanning tunneling microscopy ͑STM͒, 21 and finally, ͑v͒ a two adlayer asymmetric dimer model ͑TAADM͒ predicted by ab initio pseudopotential total-energy and grand canonical potential calculations. 27 The TAADM is basically the ADRM 3ϫ2 model built on top of the DDRM.…”

Optical Anisotropy of theSiC(001)-(3×2)Surface: Evidence for the Two-Adlayer Asymmetric-Dimer Model

“…The b-SiC(001) surface, exhibiting three major surface phases, c(2 3 2), (2 3 1), and (3 3 2), serves as a prototype in that respect. It has been intensively investigated by both experiment [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16] and theory [17][18][19][20][21][22]. Of particular importance are Si-rich conditions, because of the high Si vapor pressure in most methods of growth.…”

Series of(n×2)Si-rich reconstructions of β-SiC(001): A prospective atomic wire

“…It was found that the 1 ML TAADM model was the preferred model for the (3×2) reconstruction over the entire range of allowable chemical potential. Our calculated dispersion values for the TAADM contradict experiment however, we find that the only model that matches the photoemission data 44 is the ADRM. The mapping between observed surface states and one electron Kohn-Sham eigenvalues is not well defined however.…”

“…We attribute this to our modeling of a perfectly ordered periodic (3×2) surface, whereas the (3×2) surface has been observed to be almost perfectly disordered 45,46 , in the sense that there is no observable correlation between the addimer tilt in one (3×2) unit cell, and the next (in either the ×3 or ×2 direction). The measured dispersion of the surface state bands is ∼0.2 eV along Γ − J ′ , with no measured dispersion along J 44 . We find that the only model which reflects this anisotropy of dispersion is the ADRM, with the DDRM and the TAADM both possessing larger dispersions along Γ − J and smaller dispersions along Γ − J ′ , contrary to experiment.…”

“…We can calculate the differences (the 'dispersion') between the Kohn-Sham eigenvalues calculated at different points of the Brillouin zone, and compare these with recent photoemission experiments performed on the (3×2) surface [41][42][43][44] . Our results, and equivalent results in the literature are presented in Table 2.…”