Transport properties of band engineered $p\text{−}n$ heterostructures of epitaxial ${\mathrm{Bi}}_2{\mathrm{Se}}_3/{\left({\mathrm{Bi}}_{1-x}{\mathrm{Sb}}_x\right)}_2{\left({\mathrm{Te}}_{1-y}{\mathrm{Se}}_y\right)}_3$ topological insulators

The challenge of parasitic bulk doping in Bi-based 3D topological insulator materials is still omnipresent, especially when preparing samples by molecular beam epitaxy. Here, we present a heterostructure approach for epitaxial ${\left({\mathrm{Bi}}_{1-x}{\mathrm{Sb}}_x\right)}_2{\left({\mathrm{Te}}_{1-y}{\mathrm{Se}}_y\right)}_3$ (BSTS) growth. A thin $n$-type ${\mathrm{Bi}}_2{\mathrm{Se}}_3$ (BS) layer is used as an epitaxial and electrostatic seed which drastically improves the crystalline and electronic quality and reproducibility of the sample properties. In heterostructures of BS with $p$-type BSTS we demonstrate intrinsic band bending effects to tune the electronic properties solely by adjusting the thickness of the respective layer. The analysis of weak antilocalization features in the magnetoconductance indicates a separation of top and bottom conduction layers with increasing BSTS thickness. By temperature- and gate-dependent transport measurements, we show that the thin BS seed layer can be completely depleted within the heterostructure and demonstrate electrostatic tuning of the bands via a back gate throughout the whole sample thickness.

Figure 1

RHEED patterns of BSTS directly grown on STO (a), with BST seed layer (b) and BS seed layer (c). (d)–(f) RHEED patterns of BSTS (right) with BS seed layer on different substrates (middle).

Figure 2

Schematic band structure and density of states (DOS) of dopant levels in BS (a) and BSTS (b). (c) Heterostructure concept: depending on respective BS and BSTS thicknesses, band bending is introduced within the bilayer, leading to different sizes of metal-like (m-bulk) and semiconductor-like (semiC) bulk contribution additional to conduction of top (t-TSS) and bottom (b-TSS) topological surface states. (f) ARPES at $77\mathrm{K}$ imaging the band bending evolution for samples of 1 QL BS and 3 (i), 6 (ii), and 12 QL (iii) BSTS. The horizontal black dashed line represents the Fermi level. The red circles roughly mark the position of the Dirac point, visualizing the shift of the band structure with increasing BSTS thickness.

Figure 3

Sheet resistance as a function of temperature normalized to room temperature of $1+x$ (a), $2+x$ (b), and $4+x$ (c) series. (d) Conductivity at $4.2\mathrm{K}$ of all three series versus total sample thickness. The inset shows the $1+x$ series versus $1/t_{\mathrm{tot}}$ and a linear fit (light blue dashed line). The $y$ intercept yields the asymptote (black dashed line) in the main figure.

Figure 4

Magnetoresistance at $4.2\mathrm{K}$ for the $1+x$ (a) and $4+x$ (b) series. (c) HLN fits (white dotted lines) to $\mathrm{\Delta }G\left(B\right)$ for the $1+x$ series. (d) $\alpha$ values from HLN fits for all three series versus BSTS thickness at $4.2\mathrm{K}$. The insets show band structure sketches for different BSTS thicknesses corresponding to the evolution of $\alpha$.

Figure 5

Normalized sheet resistance against back-gate voltage at $4.2\mathrm{K}$ for the $4+x$ series (a) with a zoom-in for the thickest samples (b) and the $1+x$ series (c). (d) Dual-gated measurement of sheet resistance for sample $1+40$. The white dashed line is a guide to the eye along the maximum of $R_{\mathrm{S}}^{\mathrm{norm}}$.