Recently, a family of bismuth-based materials, in particular bismuth chalcogenides (Bi_{2}Se_{3}, Bi_{2}Te_{3}), have been identified as three-dimensional (3D) topological insulators (TIs), i.e. materials characterized by a non-trivial bulk insulating gap and topologically protected surface states with linear dispersion and helical spin texture, traversing the gap [1]. A question of great fundamental and practical importance is how electronic and spin properties of topological surface states are modified in the presence of external perturbations, in particular time-reversal-breaking ones, such as magnetic dopants [2]. Experimentally, this question can be addressed by using advanced experimental probes, such as spin-sensitive angle-resolved photoemission spectroscopy (ARPES) and scanning tunnelling microscopy (STM). On the theoretical side, there is a need for atomistic modelling of TIs that enables quantitative analysis and comparison with experiments, while keeping the computational overhead to a minimum. Microscopic tight-binding models, combined with input from ab initio calculations, provide a convenient platform to study surface states in TIs [3].
We present results of realistic tight-binding modelling of 3D TIs, with particular focus on Bi_{2}Se_{3}. Our implementation is based on the sp3 tight-binding model for Bi_{2}Se_{3} by Kobayashi [4], with parameters calculated from density functional theory. We start with a thorough analysis of the calculated band structure of a slab of Bi_{2}Se_{3} of varying thickness, with surface states identified using quantitative criteria according to their actual spatial distribution. We investigate the thickness-dependent energy gap for thin slabs, attributed to inter-surface interaction, and the emergence of gapless surface states for slabs above a critical thickness. A quantitative description of the transition to infinite-thickness limit is provided by calculating explicitly the associated modifications in the spatial distribution and spin character of the wave function. We find that the system must be at least forty quintuple-layers thick to displace surface states that are essentially localized on either surface. In addition, we discuss the effect of an external magnetic field on the electronic structure of Bi_{2}Se_{3}. The peculiar structure of the Landau levels, found experimentally in this system [5], is a characteristic signature of the presence of Dirac surface states and can be used to extract the dispersion of the surface band. Furthermore, building upon previous work on GaAs [6], we develop a finite-cluster tight-binding approach, where an infinite slab is represented by a large but finite cluster with the same thickness. We find that the electronic structure for a finite cluster, with periodic boundary conditions along the length and width of the cluster, is in excellent agreement with that of an infinite slab. On the other hand, the finite-cluster approach allows us to directly probe mesoscopic effects, associated with real boundaries of a finite system, especially when the topological surface states are present. The approach enables also the investigation of individual impurities and defects. As a first application we present a case study of substitutional Bi defect at Se site near the surface of a slab of Bi_{2}Se_{3} of varying thickness, focusing in particular on the interplay between defect-induced states and gapless surface states. The analysis of spatial features of the calculated local density of states around the defect reveals similarities with STM studies [7]. Finally, we discuss strategies for incorporating single transition-metal dopants (Mn, Fe) in the current model and draw connections to recent STM experiments [8].
[1] M. Z. Hasan and C. L. Kane, Rev. Mod. Phys. 82, 3045 (2010); X.-L. Qi and S.-C. Zhang, Rev. Mod. Phys. 83, 1057 (2011).
[2] H. Beidenkopf et al., Nature Physics 7, 939 (2011); L. A. Wray et al., Nature Physics 7, 21 (2011).
[3] W. Zhang, R. Yu, H.-J. Zhang, X. Dai, and Z. Fang, New. J. Phys. 12, 065013 (2010); M. S. Bahramy et al., Nature Communications 3, 1159 (2012).
[4] K. Kobayashi, Phys. Rev. B 84, 205424 (2011).
[5] T. Hanaguri et al., Phys. Rev. B 82, 081305(R) (2010).
[6] T. O. Strandberg et al., Phys. Rev. B 80, 024425 (2009).
[7] S. Urazhdin et al., Phys. Rev. B 66, 161306(R) (2002); S. Urazhdin et al., Phys. Rev. B 69, 085313 (2004).
[8] Y. S. Hor et al., Phys. Rev. B 81, 195203 (2010); T. Schlenk et al., Phys. Rev. Lett. 110, 126804 (2013).