Researchers at the University of Basel and ETH Zurich switched the polarity of a ferromagnet using a laser beam. The magnet's north and south poles reversed without applying any external magnetic field.
Conventional magnetic switching requires a magnetic field stronger than the material's coercivity — the threshold field that overcomes the material's preference for its current orientation. This is how hard drives write data: an electromagnet applies a field to flip each bit. The process is slow relative to optical timescales and requires physical proximity between the write head and the medium.
Reader note (2026-11-16): The Basel and ETH Zurich release (Smoleński, Imamoğlu) describes a different mechanism from the one this essay gives. The material is two atomically thin layers of molybdenum ditelluride stacked with a slight twist, the laser pulse changes the collective orientation of the spins, and the release says the switching is done using only light, without raising the temperature; it says heating above the critical temperature is what is normally needed. The essay's femtosecond heating past the Curie temperature, its two-sublattice ferrimagnet account and its picosecond timescale are not in the release (it gives no switching time); they describe a different, known family of all-optical switching. The release gives the paper as Nature, 3 March 2026. Checked against the release only, not the paper.
Laser switching bypasses the field entirely. The mechanism uses ultrafast heating — a femtosecond laser pulse heats the material past its Curie temperature, momentarily destroying the magnetic order. As the material cools, angular momentum transfer between sublattices in the ferrimagnet drives the magnetization into the opposite orientation. The switching occurs in picoseconds.
The key constraint is the material. This works in special ferrimagnets where two sublattices with different magnetic elements contribute opposing magnetic moments. The sublattices demagnetize at different rates when heated. The faster-demagnetizing sublattice loses its moment first, and the angular momentum it releases drives the other sublattice past the equilibrium point, landing on the opposite pole.
The reversal happens not because the laser tells the magnet which way to point, but because the two competing internal structures respond to heat at different speeds. Remove the equilibrium, and the asymmetry in recovery determines the outcome. The laser is not the instruction — it is the disruption. The material's internal dynamics are the instruction.