A study of twisted graphene reports a set of current-dependent electrical features that the authors say are consistent with a ferro-Josephson-like form of domain-wall dynamics. A domain wall is a boundary between magnetic regions. The evidence comes from indirect electrical transport signatures, so the proposed magnetic motion is being assessed through its effects on the signal rather than directly observed.
The clearest signal was a narrow peak in longitudinal resistance at zero in-plane magnetic field. It appeared only in the gate-region sliver where the t1+3 state was valley polarized. The peak depended strongly on AC current and disappeared in the reported dataset below 0.4 nA.
The primary sample was a monolayer graphene sheet twisted above trilayer graphene, forming a t1+3 stack with a twist angle of 1.29 degrees. Four-terminal lock-in transport used a 1.4 nA AC bias for standard longitudinal and Hall resistance measurements. For differential resistance, the experiment added a 10 to 80 pA AC modulation to a DC bias, tracking how the voltage response changed as current was varied.
A pattern that shifts with magnetic field
The differential-resistance data showed multiple peaks. At zero in-plane field, the peaks approached zero bias. Increasing the in-plane field moved the first peak to higher bias and eventually removed it. That shifting pattern is central to the study’s test of whether the transport features are consistent with ferro-Josephson-like domain-wall dynamics.
The response also depended on field direction. The spectra were roughly, though not exactly, isotropic within the plane of the graphene, while much larger out-of-plane fields had little effect. Changing the out-of-plane field from 18 to 30 mT altered the observed fingerprints only weakly.
Field history mattered as well. After the out-of-plane field was ramped past about -130 mT and then brought back toward zero, magnetoresistance was absent. It reappeared immediately when the field crossed zero. The result indicates that the transport response depended on how the magnetic state had been prepared, although the domain configuration was inferred from transport rather than directly imaged.
Where the resonances dispersed cleanly, the reported threshold-current slope was a few nanoamps per millitesla. The authors interpreted that scale as consistent with a domain wall containing about 200 precessing moments. The slope varied across datasets and in-plane directions, and no formal uncertainty interval was reported.
An intriguing signal, with an unfinished case
The proposed picture also has to account for the temperature at which the signal survives. Magnetoresistance was observed at 700 millikelvin, while the thermal energy exceeded the stated tens-of-nanoelectronvolt Zeeman scale by more than three orders of magnitude. The authors argue against simple thermal or voltage activation over an energy barrier, while acknowledging that this is a mechanism argument rather than a direct exclusion of every alternative.
The central in-plane-field-dependent features were not observed in every scan, even at gate settings where they usually appeared. The supplied analysis centers on a primary t1+3 device and shows a t2+2 device for comparison, but it does not report a total device count or an independent replicate count. That history dependence and incomplete scan-to-scan reproducibility limit how broadly the result can be generalized.
The authors interpret the pattern through a model in which polarized current could supply spin-transfer torque, potentially overcoming an in-plane restoring torque and being associated with precession of moments in a domain wall. The model links that motion to a Josephson-like voltage, but the transport data do not establish this mechanism as the unique cause of the observed features. The paper also says the model applies to small out-of-plane fields and does not capture the suppression seen at large out-of-plane fields.
On the evidence supplied, the result is a suggestive transport signature rather than a settled identification of the mechanism. The document is an arXiv preprint identified as version 3, and the supplied metadata does not identify a journal or peer-review status.
Paper data and sources
Original title: Signatures of a ferro-Josephson effect in twisted graphene
Authors: Ruiheng Su, Zhenxiang Gao, Christopher Coleman et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-26
DOI: Not available
Original paper · Full text