Premium
Atomic Structure of Hexagonal SiC Surfaces
Author(s) -
Starke U.
Publication year - 1997
Publication title -
physica status solidi (b)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.51
H-Index - 109
eISSN - 1521-3951
pISSN - 0370-1972
DOI - 10.1002/1521-3951(199707)202:1<475::aid-pssb475>3.0.co;2-e
Subject(s) - annealing (glass) , materials science , dangling bond , bilayer , silicon , silicon carbide , crystallography , hydrofluoric acid , wafer , vacancy defect , phase (matter) , stacking , etching (microfabrication) , stoichiometry , layer (electronics) , nanotechnology , chemistry , composite material , optoelectronics , metallurgy , biochemistry , organic chemistry , membrane
A review of the atomic structures found on hexagonal SiC surfaces is presented. Ex situ preparation methods such as oxidation and subsequent etching with hydrofluoric acid or hydrogen etching generate nearly bulk truncated surfaces whose dangling bonds are saturated by atomic or molecular adspecies. These surfaces exhibit the two‐dimensional periodicity of a SiC bulk bilayer. The topmost surface layer arrangement and thus the surface morphology seem to be dependent on polytype and orientation of the sample, and on the particular preparation treatment. In several cases of nominally Si rich orientated samples such as 6H‐SiC(0001) step bunching can be found. Then a linear surface layer stacking is preferred. On the nominally carbon rich orientation, for example 6H‐SiC(0001‐), mostly single steps are present. Heating the samples in vacuum in combination with Si evaporation leads to different reconstructed phases with a stoichiometry depending on annealing temperature and Si flux. The atomic structure of these phases is still under debate. On the Si rich orientation a (3×3) phase is generated upon large Si exposure. Its surface presumably consists of a Si bilayer and additional Si adatoms. Lower silicon flux during annealing or heating the (3×3) phase leads to the development of a (√3×√3)‐R30° phase. This phase can also be obtained starting on an ex situ prepared sample by heating alone. Adatom, adcluster and vacancy models have been proposed in the literature for the (√3×√3)‐R30° phase. On the opposite sample orientation the development of a (1×1) phase was reported. Further annealing causes silicon depletion for both surface polarities. On the Si side a phase of apparent (6√3×6√3)‐R30° periodicity develops while on the carbon side a carbon rich (3×3) phase was found. Eventually, heating leads to the development of graphitic overlayers.