Screening effect in low dimension : role of dielectrics and substrates on electronic strucuture of 1- and 2-dimensional materials

Abstract

For TIs, the surface states could be influenced due to the underlying substrates. Using density functional method and spin-orbit coupling, the Dirac cones and the bands around them are studied. Graphene has been found to be a suitable substrate for topological insulators (TIs) in recent experiments. In our study, six quintuple layers (QLs) as well as just one QL were considered for Bi2Se3. When six QLs of Bi2Se3 are supported by a graphene substrate on either one or both sides, the Dirac cone of the TI is shifted below the Fermi energy and a small band gap of graphene is opened. In addition, the influence of the graphene substrate on the real space density of the topological surfaces states (TSS) is found to be negligible. The Dirac point of Bi2Se3 splits to two Dirac points when graphene is deposited on only one side. For one QL Bi2Se3 the electronic structure near the band gap was strongly perturbed due to the interaction with graphene orbitals. These results will be compared with other works of TIs on substrates and their impact on device properties will be discussed in detail.

In summary, substrates are found to play active roles in nanoscale devices ranging from substrate induced giant renormalization of band gap in GNRs to strong surface state modulations in TIs due to underlying substrate.

Nanodevices based on low dimensional structures such as graphene nanoribbons (GNRs), topological insulator (TI) thin films are typically deposited on dielectric substrates. However, the effect of dielectric screening arising from the surrounding materials has not been investigated for 1 and 2- dimensional emerging structures. Using large-scale electronic structure calculations based on the Greens function (G) and screened Coulomb (W) interaction (GW approach), we show that when GNRs are deposited on dielectrics, the band gaps are strongly modulated even though the GNR-substrate interaction is weak. This effect, which is negligible in graphene (where the intrinsic screening is relatively complete), could renormalize the band gap of GNRs by as much as 1 eV when deposited on a substrate. In addition, the screening effect is dependent on the edge structure of GNRs and could be understood on the basis of charge localization. Such an understanding is critical to the band gap engineering of graphene based device components.