Exact Solution for Bulk-Edge Coupling in the Non-Abelian $\nu=5/2$ Quantum Hall Interferometer

TL;DR

This paper provides an exact solution for how bulk-edge coupling affects interference patterns in a non-abelian /2 quantum Hall interferometer, revealing universal phase shifts and the conditions under which localized quasiparticles influence the edge.

Contribution

It derives an exact solution for bulk-edge tunneling effects on interference in the /2 quantum Hall state, clarifying how coupling strength modifies the interference pattern.

Findings

Strong coupling effectively incorporates localized QPs into the edge.

Universal phase shift occurs as a function of coupling strength.

Interference magnitude varies with bulk-edge coupling.

Abstract

It has been predicted that the phase sensitive part of the current through a non-abelian $\nu = 5/2$ quantum Hall Fabry-Perot interferometer will depend on the number of localized charged $e/4$ quasiparticles (QPs) inside the interferometer cell. In the limit where all QPs are far from the edge, the leading contribution to the interference current is predicted to be absent if the number of enclosed QPs is odd and present otherwise, as a consequence of the non-abelian QP statistics. The situation is more complicated, however, if a localized QP is close enough to the boundary so that it can exchange a Majorana fermion with the edge via a tunneling process. Here, we derive an exact solution for the dependence of the interference current on the coupling strength for this tunneling process, and confirm a previous prediction that for sufficiently strong coupling, the localized QP is effectively incorporated in the edge and no longer affects the interference pattern. We confirm that the dimensionless coupling strength can be tuned by the source-drain voltage, and we find that not only does the magnitude of the even-odd effect change with the strength of bulk-edge coupling, but in addition, there is a universal shift in the interference phase as a function of coupling strength. Some implications for experiments are discussed at the end.