Asymmetric nanowire SQUID: linear CPR, stochastic switching, and symmetries

TL;DR

This paper investigates ultrathin superconducting nanowire SQUIDs, revealing their unique linear current-phase relationship, asymmetric critical current behavior, and developing a model that accurately explains these phenomena and vorticity stability regions.

Contribution

It introduces a novel model assuming linear CPR and sample-specific critical phases, accurately fitting experimental data and explaining complex vorticity and switching current behaviors.

Findings

Critical current versus magnetic field is multivalued and asymmetric.

The model accurately predicts vorticity stability regions.

A technique for detecting hidden phase-slips was developed.

Abstract

We study nanodevices based on ultrathin superconducting nanowires connected in parallel to form nanowire SQUIDs. The function of the critical current versus magnetic field, $I_{C}(B)$, is multivalued, asymmetric and its maxima and minima are shifted from the usual integer and half integer flux quantum points. The nanowire interference device is qualitatively distinct from conventional SQUIDs because nanowires do not obey the same current-phase relationship (CPR) as Josephson junctions. We demonstrate that the results can be explained assuming that (i) the CPR is linear and (ii) that each wire is characterized by a sample-specific critical phase, which is usually much larger than $\pi/2$. Our proposed model offers accurate fits to $I_{C}(B)$. It explains the single-valuedness regions where only one vorticity (i.e., the order parameter winding number) is stable as well as regions where multiple vorticity values are allowed for the SQUIDs. We also observe and explain regions in which the standard deviation of the switching current is independent of the magnetic field. We develop a technique that allows a reliable detection of hidden phase-slips. Using this technique we find that our model correctly predicts the boundaries of vorticity regions, even at low currents where $I_C(B)$ is not directly measurable.