Experimental Realization of Topological Insulator

Presenter	Pu Xiao
Date	June 23, 2026
Topic	Experimental Realization of Topological Insulator
Notes	Slides & PDF below

A survey of how topological insulators are actually made and measured: from the band-inversion picture and the 2D/3D material platforms that realize it, to the experimental probes — ARPES, molecular beam epitaxy, and STM/STS — that reveal and characterize the protected surface states.

Topological insulators: the band-inversion picture

Starting from an ordinary band insulator, turning on strong spin–orbit coupling can invert the conduction and valence bands. Restoring the gap in the presence of this inversion forces gapless, spin-momentum-locked states to appear at the boundary — the hallmark edge/surface states of a topological insulator.

Material platforms

• 2D TI (quantum spin Hall insulator): HgTe quantum wells (Bernevig et al., Science 314, 1757 (2006); König et al., Science 318, 766 (2007)); InAs/GaSb quantum wells; monolayer 1T′-WTe$_2$.
• 3D TI: Bi$_x$Sb$_{1-x}$ alloys (Fu & Kane, Phys. Rev. B 76, 045302 (2007); Hsieh et al., Nature 452, 970 (2008)) and the Bi$_2$Se$_3$ family (Zhang et al., Nat. Phys. 5, 438 (2009); Xia et al., Nat. Phys. 5, 398 (2009); Chen et al., Science 325, 178 (2009)) — a single Dirac cone, large bulk gap, and simple surface chemistry made this family the workhorse system for the experiments below.

Probe 1: ARPES

Angle-resolved photoemission spectroscopy extends the photoelectric effect ($E_{kin} = h\nu - \phi - |E_B|$) to map the occupied band structure directly: measuring the photoelectron’s kinetic energy and emission angle recovers its in-plane crystal momentum,

$$ \mathbf{p}_\parallel = \hbar \mathbf{k}_\parallel = \sqrt{2mE_{kin}}\,\sin\vartheta . $$

ARPES directly imaged the single Dirac surface state of the Bi$_2$Se$_3$ family — the linear, spin-momentum-locked cone connecting valence and conduction bands at the Kramers point (Hsieh et al., Nature 460, 1101 (2009); Chen et al., Science 325, 178 (2009)).

Probe 2: molecular beam epitaxy

Thin films grown by MBE (slow growth rate $\sim 0.2\,\text{nm/min}$, ultra-high vacuum $\sim10^{-10}\,\text{mbar}$, monitored by RHEED) give atomic-layer control over thickness and interfaces. Tracking the ARPES spectrum quantum-layer by quantum-layer shows the 2D-to-3D crossover in Bi$_2$Se$_3$: the surface Dirac cone only emerges once the film is thick enough that top and bottom surface states decouple (Zhang, He, Chang et al., Nat. Phys. 6, 584–588 (2010)).

Probe 3: STM / STS

Scanning tunneling microscopy resolves real-space surface topography from the tip–sample tunneling current, $I \sim e^{-2\kappa d}$. Scanning tunneling spectroscopy extends this to energy-resolved measurements — the differential conductance $dI/dV$ at fixed tip position is proportional to the local density of states,

$$ \left.\frac{dI}{dV}\right|_{V_B} \propto \rho_S(eV)\,T(z,eV) + \int_0^{eV_B} \rho_S(E)\, \frac{\partial T(eV,E)}{\partial V}\,dE . $$

Discussion points

• How robust is the Dirac surface state to non-magnetic vs. magnetic disorder, and how is that tested in STS?
• MBE thickness control as a knob for opening a hybridization gap — connection to proposals for gapping the surface state deliberately (e.g., for axion electrodynamics).
• What ARPES and STM each see that the other cannot (occupied-only vs. real-space/LDOS).

References

1. Hasan & Kane, Rev. Mod. Phys. 82, 3045 (2010).
2. Qi & Zhang, Rev. Mod. Phys. 83, 1057 (2011).
3. Bernevig, Hughes & Zhang, Science 314, 1757 (2006).
4. König et al., Science 318, 766 (2007).
5. Hsieh et al., Nature 460, 1101 (2009).
6. Chen et al., Science 325, 178 (2009).
7. Zhang, He, Chang et al., Nat. Phys. 6, 584–588 (2010).