Entangled electron current through finite size normal-superconductor tunneling structures

TL;DR

This paper provides a theoretical analysis of entangled electron tunneling from a superconductor into a normal metal, emphasizing the importance of momentum dependence in tunneling models for accurate predictions of current distribution and entanglement decay.

Contribution

It introduces a local tunneling Hamiltonian accounting for arbitrary interface geometries and demonstrates the significance of momentum dependence for correct physical behavior.

Findings

Momentum dependence is crucial for unitarity and correct thermodynamic limits.

Entangled electron current decays as r^{-4} with distance, not r^{-2}.

Neglecting momentum dependence leads to incorrect predictions in tunneling models.

Abstract

We investigate theoretically the simultaneous tunneling of two electrons from a superconductor into a normal metal at low temperatures and voltages. Such an emission process is shown to be equivalent to the Andreev reflection of an incident hole. We obtain a local tunneling Hamiltonian that permits to investigate transport through interfaces of arbitrary geometry and potential barrier shapes. We prove that the bilinear momentum dependence of the low-energy tunneling matrix element translates into a real space Hamiltonian involving the normal derivatives of the electron fields in each electrode. The angular distribution of the electron current as it is emitted into the normal metal is analyzed for various experimental setups. We show that, in a full three-dimensional problem, the neglect of the momentum dependence of tunneling causes a violation of unitarity and leads to the wrong thermodynamic (broad interface) limit. More importantly for current research on quantum information devices, in the case of an interface made of two narrow tunneling contacts separated by a distance $r$, the assumption of momentum-independent hopping yields a nonlocally entangled electron current that decays with a prefactor proportional to $r^{-2}$ instead of the correct $r^{-4}$.