Reconstruction and modeling of pure and Cd-doped In (111) surface: Ab Initio study of structural, electronic, and hyperfine properties at and near the surface

DARRIBA G. N.; R. J. FACCIO; M. RENTERÍA

Goa

Conferencia; HYPERFINE 2019 - International Conference on Hyperfine Interactions and their Applications; 2019

The reconstruction of pure surfaces has great importance since its most stable configuration is the structural starting point for the study of the doped surface. We present here a complete ab initio/DFT study of structural, electronic, and hyperfine properties of pure and Cd-doped (111) In surface, with Cd at different depths from the surface. For the pure system we performed a complete optimization of all atomic positions in the supercell. For the doped systems, a Cd atom replaced an In atom at the different inequivalent In sites of the pure optimized surface, followed by a new complete optimization. We studied the structural and electronic changes produced by the generation of the surface itself and by the inclusion of the Cd impurity. The magnitude, symmetry, and orientation of the electric-field gradient (EFG) tensor was studied as a function of the depth from the surface. For the doped systems, we obtained basically 2 EFG tensors: one when Cd is localized at the two surface sites, at different depths but both with only vacumm above the impurity, and another one for all the other Cd positions at different depths from the surface. These two predicted EFG tensors are in perfect agreement with the ones observed in PAC experiments on layer-by-layer 111In-doped (111) In surface, assigned to 111Cd probes localized at the surface (HFIS) and at the bulk (HFIB) of these In samples [1]. Our combined approach enabled to confirm that HFIS is related with probes localized at the surface, with a V33 magnitude four times larger that the bulk value. The precision of the EFG ab initio calculation could have resolved among the 3 different Cd positions existing at the surface: the adatom, and the two surface sites mentioned above. But the combination of these predicted EFGs and the EFG distribution observed in the experiments would indicate that the Cd atom is localized at these 3 positions in the real samples. [1] W. Körner, et al., 1982, Phys. Rev. Lett., 49, 1735