Why a magnet floats over a superconductor
A superconductor is a material that, once cold enough, carries electric current with no resistance at all. It also does something stranger: it refuses to let a magnetic field inside. Hold a magnet over it and it floats.
Why this is worth a look
Zero resistance is invisible; you measure it with a meter. Field expulsion you can see: a small magnet hanging above a cold disc, touching nothing. It is also the deeper property. A merely perfect conductor would not do this if the magnet was already there when it was cooled, for a reason worked out below, so a magnet that rises as the disc cools signals something other than very good conduction. The same push is now being tried in the laboratory to hang tiny magnets in vacuum as motion sensors: touching nothing, with almost no gas to rub against, they swing for over ten thousand seconds before the motion dies.
What the field does
A magnet fills the space around it with a magnetic field. A field line follows the direction of that field; a bar magnet's lines loop out of one end and back into the other. Over an ordinary metal, the lines pass straight through.
Cool the metal below its transition temperature, where it becomes superconducting, and it answers the field. Currents flow in a thin layer at its surface, making a field of their own, arranged so that inside the material the two fields cancel exactly. The interior is field-free. This expulsion is the Meissner effect.
Interactive Press cool to take the disc below its transition temperature, then drag magnet and weight and watch where it settles.
Why that lifts the magnet
The field lines that would have gone through the disc now bend around it and bunch up in the gap between magnet and surface. Bunched field lines push apart like a compressed spring, so squeezing them into a smaller gap costs energy. The magnet is pushed toward the larger gap: upward.
Said another way: the surface currents produce, above the disc, exactly the field of a mirror-image magnet below the surface, like poles nearest. Like poles repel. In this simplest model the push grows sixteen-fold each time the magnet's distance from the surface halves, while its weight stays the same. So the magnet settles at the one height where push equals weight. A stronger magnet floats higher, a heavier one lower. That fixes the height only; over a flat surface nothing holds the magnet sideways, which is why classic demonstrations with pure metals use a dished superconductor, a lead bowl, whose sloping walls push the magnet back to the centre.
Why a perfect conductor is not enough
A conductor with no resistance would also carry surface currents, but only in response to a change in field, as any conductor does. Whatever field was already inside when it was cooled would stay frozen in. A superconductor does not care about history. Its field-free state is the lowest-energy state, so it expels a field it was already sitting in. That difference is what the floating magnet makes visible.
One caution. In liquid-nitrogen demonstrations a magnet can also hang beneath the superconductor, or stay locked in place when nudged. That needs something extra: in those materials some field threads through in thin tubes that get pinned in place. The pure case, with all the field pushed out, is the one drawn here.
In short
Below its transition temperature a superconductor drives surface currents that cancel any magnetic field inside it. A magnet's field lines must bend around the material and bunch up in the gap, and bunched field pushes back. The push grows as the gap shrinks, the weight does not, and the magnet floats where the two agree.
Where this comes from
- Magnetic levitation on a type-I superconductor as a practical demonstration experiment for students linked only, not reproduced
arxiv.org/abs/1207.5992 - A Classroom Demonstration of Levitation and Suspension of a Superconductor over a Magnetic Track linked only, not reproduced
arxiv.org/abs/0803.3090 - Ultralow mechanical damping with Meissner-levitated ferromagnetic microparticles linked only, not reproduced
arxiv.org/abs/1912.12252