Topological interface states in the natural heterostructure (PbSe) 5 ( Bi2Se3 ) 6

Abstract

We study theoretically the electronic band structure of (PbSe) 5 (Bi 2 Se 3 ) 6 , which consists of an ordinary insulator PbSe and a topological insulator Bi 2 Se 3 . The first-principles calculations show that this material has a gapped Dirac-cone energy dispersion inside the bulk, which originates from the topological states of Bi 2 Se 3 layers encapsulated by PbSe layers. Furthermore, we calculate the band structures of (Bi x Pb 1−x Se) 5 (Bi 2 Se 3 ) 6 with Bi Pb antisite defects included in the PbSe layers. The result shows that a high density of Bi Pb defects can exist in real materials, consistent with the experimentally estimated x of more than 30%. The Bi Pb defects strongly modify the band alignment between Bi 2 Se 3 and PbSe layers, while the topological interface states of Bi 2 Se 3 are kept as a gapped Dirac-cone-like dispersion.

Evolution of nontrivial Fermi surface features in the band structures of the homologous members Pb5Bi6Se14 and Koumoulis1, Fang2, Chung3 et al. 2020Phys. Rev. B3000View full textAdd to dashboardCite

“…Furthermore, the observed temperature dependence of 1/T 1 T for m=2 rules out the possibility that the material can be described as a gapped semiconductor (similar to m=1), since there is a temperature independence of 1/T 1 T (Figure 2b). Hence, the dominance of a conducting-electron-caused relaxation mechanism for 77 Se nucleus of m=2 across the experimental temperature range is apparent in Figure 2B, and aligns with both ARPES [4] and ab initio results [8].…”

“…As a result, there is an inverted band symmetry related to the former (e.g. Pb x Sn 1-x Te [6,7], (PbSe) 5 (Bi 2 Se 3 ) 3m [4,8] etc.). Scalar relativistic, electronic and structural terms are essential contributions to the BI mechanism.…”

“…An angle-resolved photoemission spectroscopy (ARPES) study on m=1 and m=2 phases by Nakayama et al [4] and first-principles calculations by Momida et al [8] have shown a marked difference on the band dispersions near the Fermi level (E F ) in these two samples. Namely, the band structure for m=1 is a single parabolic band, whereas for m=2 the band structure consists of multiple bands associated with a shallow dip at the top due to the presence of additional bands that cross E F .…”

“…In particular, ARPES experiments on m=2 [4] revealed six dispersive features around the near-E F region where two of them cross E F and overlap with the Dirac-cone surface states, which mainly originate from the Bi 2 Se 3 layers. Recent ab initio calculations [8] for the m=2 system have shown that its electronic structure near the Fermi energy consists of a mixture of Bi 2 Se 3 -and PbSe-layer bands.…”

Nanomosaic of Topological Dirac States on the Surface of Pb5Bi24Se41 Observed by Nano-ARPES

“…31 Moreover, recent first-principles calculations for PSBS (m = 2) with realistic lattice parameters, in which the strain effect is naturally taken into account, show a substantially small hybridization gap of ∼20 meV (c.f., 100-200 meV for unstrained 2-QL Bi 2 Se 3 ). 32 It is noted that, while strain effect may also arise from crystal imperfections, the observed sample insensitivity of the hybridization gap size ( Figure S2) rules out such a possibility because the crystal imperfection, if it exists, would be sample-dependent and hence the gap size should change from sample to sample. In addition, surface reconstruction that may also lead to lattice strain at the surface is also ruled out because our band-structure mapping over a wide momentum space does not show a signature of band-folding indicative of surface reconstruction.…”