Scanning Tunneling Microscopy and Surface Simulation of Zinc-Blende GaN(001) Intrinsic4×Reconstruction: Linear Gallium Tetramers?

Scanning tunneling microscopy images confirm electron difraction studies that the zinc-blende GaN(001)-4× reconstruction consists of rows aligned along $[110]$ with a spacing along [11̄0] of $4a$. Dual-bias imaging shows a 180^∘{} shift of the corrugation maximum position between the profiles of empty and occupied states, in agreement with surface simulations based on the 4×1 linear tetramer model of Neugebauer et al. [Phys. Rev. Lett. 80, 3097 (1998)]. Electronic structure calculations predict a surface band gap of 1.1 eV, close to the measured value of 1.14 eV and the previously predicted value (1.2 eV). Despite the successes of this model, high-resolution images reveal an unexpected 3× periodicity (not seen in diffraction) along the $[110]$ row direction, indicating the need for a 4×3 model, and putting into question the existence of linear Ga tetramers.