Electrons in a Crystal Kept Oscillating at 60 Tesla, Past the Point Where Theory Says They Stop

Put a metal in a strong magnetic field and cool it far enough, and its electrical resistance starts to wobble up and down in a regular rhythm as the field increases. The effect, called Shubnikov–de Haas oscillation, is one of the oldest and most trusted tools in condensed matter physics: the spacing of the wobbles tells you the shape of the electron orbits inside the material. Push the field high enough and the oscillations are supposed to stop entirely, because every electron has been squeezed into the lowest available energy level and there is nothing left to oscillate.

In zirconium pentatelluride, they do not stop. A team led by researchers at the University of São Paulo, working with Los Alamos National Laboratory and the University of Washington, took crystals of ZrTe5 to magnetic fields as strong as 60 tesla — more than a million times Earth's field — at a temperature of about 0.7 kelvin, roughly minus 272.45 degrees Celsius. The oscillations persisted past the quantum limit, and they were not periodic in the way the textbook effect requires. The work appeared in Nature Communications on August 17.

The explanation the team proposes is that the electron's spin, not just its orbital motion, is doing part of the bookkeeping. "When we apply strong magnetic fields, the interaction between spin and the magnetic field profoundly alters energy levels," said Julio Larrea Jiménez of the USP Physics Institute, who advised the study. "Landau levels that would normally move away can 'return' and cross it again." Levels that should have been swept out of reach come back into play, and each crossing produces another feature in the resistance.

The first author is Cauê Kaufmann Ribeiro, with Johanna Palmstrom and Sean Thomas at Los Alamos as co-advisors; the 60-tesla pulsed fields came from the National High Magnetic Field Laboratory facility there. The measurements were paired with theoretical modelling that reproduces the non-periodic pattern from a topological origin rather than from the electron-electron interactions that are the usual suspects when oscillations misbehave.

ZrTe5 has been an argument in the field for years. Different laboratories growing nominally identical crystals have reported it as a weak topological insulator, a strong one, and a Dirac semimetal, depending on sample and temperature. A mechanism that explains why the same material gives different answers in different hands is worth more than another data point, and the paper's account of spin-driven level crossings does that work.

The broader claim is that a topological insulator carries more than charge. If the spin degree of freedom is shaping transport this strongly at high field, these materials become candidates for spintronic devices, where information moves as spin orientation rather than as current, and the energy cost of a switching operation drops accordingly. That is a long way from a device, but it is the reason a resistance curve that refuses to flatten is interesting to anyone outside the low-temperature community.