A Numerical Study of the Superconducting Proximity Effect in Topological Surface States

TL;DR

This study numerically investigates the superconducting proximity effect in topological insulator surface states, revealing the limitations of the Fu-Kane model at higher energies and the conditions for inducing a significant gap.

Contribution

It provides a detailed numerical analysis of the proximity effect, highlighting the energy-dependent validity of the Fu-Kane model and the optimal coupling for inducing a surface state gap.

Findings

The Fu-Kane model is valid at energies much lower than the superconducting gap.

Strong modifications occur in the interface-state dispersion near the gap energy.

An intermediate coupling strength optimizes the proximity effect and gap induction.

Abstract

We study the superconducting proximity effect induced in the surface states of the 3-d topological insulator Bi$_2$Se$_3$ by a singlet, s-wave superconductor deposited on its surface. To this effect, the $\mathbf{k}\cdot\mathbf{p}$-Hamiltonian of Bi$_2$Se$_3$ and the BCS-Hamiltonian are mapped onto tight-binding chains which we couple through a transfer-Hamiltonian at the interface. We then employ the Recursive Green's Function technique to obtain the local spectral function and infer the dispersion of the interface-states from it. In agreement with earlier microscopic studies of this problem, we find that the Fu-Kane model is a reasonable approximation at energies $\epsilon\ll \Delta_{\rm SC}$. However, for energies close to the SC bulk gap, the Fu-Kane model is expected to break down. Indeed, our numerical calculations show strong modifications of the interface-state dispersion for $\epsilon \gtrsim \Delta_{\rm SC} $. We find that the proximity effect can be strong enough to induce a gap in the surface state that is comparable to the superconducting gap. An analysis of the spatial profile of the states shows that their weight shifts towards the SC as the coupling strength increases. We conclude that an intermediate coupling is ideal for realising the Fu-Kane scenario.