Protected pristine CsV3Sb5 devices showed no detectable intrinsic field-free superconducting diode effect, while a response was observed after further oxidation and in a selectively etched device.
The measurements used electrical devices made from CsV3Sb5 flakes, including ten hBN-encapsulated pristine devices, altered pristine devices and Ti-doped controls.
The protected flakes set the baseline
The baseline comprised ten hBN-encapsulated pristine CsV3Sb5 devices with different thicknesses and fabrication histories. With the flakes protected and no external field, forward and backward switching into the resistive state could not be distinguished within the experiment’s resolution.
A superconducting diode effect here means that the current at which a device switches from superconducting to resistive differs between the two current directions. The researchers treated the forward and backward critical currents as the onset currents of that transition.
They measured differential resistance, the change in voltage response as current varied, using combined direct-current and alternating-current bias, then compared the two onset currents.
Responses in altered devices
A field-free diode response was observed after further oxidation. In Device 11’s second oxidation step, the reported diode efficiency reached 83.61%.
A field-free superconducting diode effect was also observed in selectively etched Device 13, consistent with extrinsic inversion-symmetry breaking.
The response was not fixed from sweep to sweep. In Device 12, reported diode-efficiency values ranged from −30.86% to 17.07% across repeated current sweeps, indicating that the nonreciprocal response could change polarity.
Field-conditioned measurements showed greater stability
In Device 13, an out-of-plane field of 0.1 T largely fixed the polarity, while the amplitude remained broadly distributed. At 0.2 T, both polarity and amplitude became stable; the same evolution was observed in Device 14.
The variability persisted in a separate field-cooling test. After out-of-plane cooling under 1 T and removal of the field, stochastic diode behavior remained.
Time-dependent fluctuation measurements did not support simple fixed-bias fluctuations as the explanation for the changes in polarity.
The authors favor a CDW-related explanation
The authors associate the relevant time-reversal-symmetry breaking with charge-density-wave, or CDW, order in the normal state. A CDW is an organized pattern in electronic charge.
They favor that interpretation over one based only on the standalone superconducting condensate or on surface processing.
The Ti-doped control, CsV3−xTixSb5 at x = 0.24, was reported to have suppressed long-range CDW order and no field-free diode effect after the same inversion-breaking protocol.
That comparison does not isolate CDW suppression from other consequences of Ti substitution, which changes the material’s composition as well as its CDW order.
The measurements leave the mechanism unresolved
These are device-level transport observations, not a direct microscopic identification of the proposed time-reversal-symmetry-broken domains. The evidence supports an electrical association among nonreciprocity, externally introduced inversion breaking and inferred time-reversal-symmetry-broken order.
The study does not establish whether time-reversal-symmetry breaking in the superconducting state is generated by superconductivity or inherited from CDW-related order. It also leaves open how fabrication disorder, local current paths and domain configurations shape the sweep-dependent efficiency.
The report gives no formal inferential tests or confidence intervals, and it does not provide a consolidated total for the altered-device or Ti-doped groups. That makes the size and reproducibility of the response harder to assess from the reported results.
The report remains a preprint
The work is presented as a 16-page preprint. The authors report no competing interests, and supporting data are available from the corresponding author upon reasonable request.
The study’s central result is that a zero-field diode response was not detected in protected pristine devices, was observed after oxidation and in a selectively etched device, and varied across repeated sweeps, while the microscopic order behind it remains unidentified.
Paper data and sources
Original title: Symmetry Origins of the Field-Free Superconducting Diode Effect in the Kagome Superconductor CsV$_3$Sb$_5$
Authors: Xin-Jie Liu, Shengbiao Sun, Ke-Fan Song et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-26
DOI: Not available
Original paper · Full text