The two directions of current
Reversing the current through a superconductor can change the threshold at which resistance-free flow ends. In the peer-reviewed Nature Communications experiment by Han Zhong and colleagues, a twist of just 1 degree between two NbSe2 layers accentuated this directional difference. With a perpendicular magnetic field of −13 oersted, critical-current asymmetry reached 27.6 per cent. That percentage is the difference between forward and reverse critical-current magnitudes divided by their sum. The measurement describes the separation of the superconducting thresholds in the two directions, rather than the fraction of electrical energy conserved.[1]
The small angle does not select the entire physical explanation by itself. The interface between twisted layers can alter inversion symmetry and the relationship between electron spin and motion. The researchers interpret the perpendicular-field response through a mechanism that makes magnetic-vortex entry at the sample edges different in the two directions. Directional asymmetry weakened as thickness increased. The responses of the 16-, 30- and 60-nanometre devices therefore also belong among the observations a theory must accommodate; twist angle alone is an incomplete description.[1]
The open question at zero field
The two interpretations diverge where the applied field is nominally set to zero. The experiment measured a negative asymmetry of approximately 2.1 per cent there. The paper proposes an explanation involving spontaneous time-reversal symmetry breaking, while retaining a possible contribution from several oersted of residual field caused by trapped flux in the magnet. These possibilities lead to different physical accounts. My narrow inference is that the nominal-zero-field measurement cannot select spontaneous symmetry breaking on its own: the device’s magnetic order and a small remaining external field can still contribute to the same signal.[1]
For the in-plane field, a straightforward alternative is quantitatively narrowed. A sample-alignment error of 0.1 degree converts an in-plane field of 200 oersted into an unwanted perpendicular component of about 0.35 oersted. The strongest perpendicular response occurred around 10–15 oersted; mimicking a component of 13 oersted requires a deviation of approximately 3.7 degrees. This weakens an explanation in which misalignment alone produces the measured in-plane response. The calculation leaves the residual-flux question at nominal zero field open. The two comparisons test different physical assumptions.[1]
Levels retained by the potential
The subsequent quantum-circuit calculation reveals another constraint. At a Josephson-to-charging energy ratio of 20, the effective model produces three bound energy levels at 10 per cent rectification, two at 27.6 per cent and only one at 50 per cent. Stronger directionality consequently offers no automatic gain for a two-level qubit in this model: tilting the potential well also changes how many levels it retains. This result comes from numerical simulation. Connecting the measured current thresholds to the performance of a qubit made from the material still requires a physical test within a circuit.[1]
What makes this study valuable to me is the way it connects the same percentage to two physical questions: where current becomes resistive in each direction, and how many quantum levels an effective potential retains. The first is determined experimentally; the second depends on model parameters. A tighter bound on residual magnetic field offers a route to distinguishing the origin of the nominal-zero-field signal, while direct measurements of circuit energy levels can test the simulated landscape. The fundamental question left by a one-degree twist is how much directional asymmetry attributable to the material’s own symmetry breaking survives as the external-field contribution shrinks.[1]