AVS 50th International Symposium
    Surface Science Monday Sessions
       Session SS3-MoM

Paper SS3-MoM6
Sb-mediated Ge and Si Growth on Ge(001)

Monday, November 3, 2003, 10:00 am, Room 328

Session: Surface Diffusion and Wetting
Presenter: J. Wang, Yale University
Authors: J. Wang, Yale University
M. Li, Yale University
E.I. Altman, Yale University
Correspondent: Click to Email

Antimony-mediated Ge and Si growth on Ge(001) were studied using scanning tunneling microscopy and other surface characterization techniques. For the homoepitaxial growth on intermixed SbGe(001) with submonolayer Sb incorporated into the surface and below the Sb-Ge exchange temperature (470 K), Sb decreased the Ge island size with isolated ad-dimers seen on intermixed surfaces but not on pure Ge. The islands were found on top of Ge not Sb indicating that diffusion was not significantly inhibited on the Sb-passivated surface. The islands were always seen near Sb-Ge substrate domain boundaries. STM movies recorded at elevated temperatures showed that Ge dimers rapidly moved along Ge substrate dimer rows but could not pass or move onto Sb dimers embedded in the surface. These results can be explained by a repulsive barrier at Sb-Ge domain boundaries that inhibits Ge dimer diffusion from Ge to Sb domains but not from Sb to Ge domains. For Si heteroepitaxial growth at 520 K where Sb-Si and Sb-Ge exchange happens, a high density of small clusters were observed when Si was pre-deposited at 520 K and then 0.3 ML Sb was deposited at room temperature. The observed enhanced nucleation was attributed to zero-order nucleation due to both surface defects and adatom-surfactant exchange. The surface becomes smoother than Si deposition on bare Ge(001), with a much lower density of deep trenches on the terraces. Further annealing up to 590 K caused the small clusters to disappear producing a surface with square-shaped islands with typical sizes of 6 x 6 nm2. The isotropic island shape suggests that the anisotropic surface diffusion was disrupted due to the existence of domain boundary barriers between Sb and Si or Ge. Growth with different Sb coverages will also be discussed.