One of the most fundamental concepts in modern physics is spontaneous symmetry breaking. When a system undergoes a phase transition into an ordered state, it must choose, apparently at random, a specific configuration from an infinite set of equivalent possibilities. A ferromagnet picks a direction of magnetization. A crystal picks a position for its lattice. And a Bose-Einstein condensate, the state of matter in which a macroscopic number of particles occupy the same quantum state, must pick a phase.
This moment of choice, the emergence of a global phase from nothing, is central to BEC theory. It was predicted by Fritz London in 1938 and has been a pillar of condensed matter physics ever since. But it has never been directly observed in the time domain.
Now, physicists at RPTU University Kaiserslautern-Landau in Germany, together with a collaborator at the University of Colorado Colorado Springs, have filmed it happening inside a millimeter-scale crystal of yttrium-iron-garnet (YIG). Their results were published in Nature Physics.
“For the first time, we are able to directly measure the spontaneous emergence of coherence in a magnon condensate,” said Malte Koster, the study’s first author. “We can show that the phase of the condensate is independent of any external source, proving the formation of a true BEC.”
Magnons as a BEC platform
Magnons are the quantum quasi-particles of spin waves, collective excitations of the magnetic order in a material. They are bosons, and under the right conditions they can condense into a BEC, just as atoms can in a gas laser-cooled to nanokelvin temperatures. The difference is that magnon BECs operate at room temperature and inside a solid crystal, making them far more accessible for experiments.
The researchers used a 2.1-micrometer-thick film of YIG, a synthetic ferrimagnetic garnet with extraordinarily low magnetic damping, the lowest known for any magnetic material. They pumped the film with 1-microsecond microwave pulses at 7.8 gigahertz through a microstrip antenna, applying a magnetic field of 281 millitesla. After each pulse, the magnons in the film thermalized through four-magnon scattering processes and, when the pump power exceeded a threshold of approximately 21 dBm, condensed into a coherent state at the bottom of the magnon spectrum.
The critical innovation was the detection method: an IQ mixer that measures the instantaneous phase of the precessing magnetization in a single shot, without averaging over cycles. This preserves the phase information of each individual condensation event.
The phase appears
Three observations confirm spontaneous symmetry breaking. First, the phase of the condensate is uniformly distributed between 0 and 2π across 1,000 independent experimental runs. The pump phase is fixed, the same every time, but the magnon phase is random, proving it is not imposed externally.
Second, the onset is abrupt. Below approximately 21 dBm pump power, no coherence appears. Above this threshold, the coherence metric jumps sharply to approximately 0.9, a classic phase transition signature.
Third, once formed, the condensate maintains its phase until the magnon density decays below the noise floor. There is no dephasing; the state is stable for its entire lifetime.
“This is the smoking-gun confirmation of U(1) symmetry breaking in a quasiparticle BEC,” said Georg von Freymann, one of the senior authors. “The phase is not determined by the pump, by the geometry, or by the crystal. It is chosen spontaneously, each time anew.”
Why it matters
The experiment closes a long-standing gap in BEC physics. Spatial phase differences had been observed in interference experiments, and second-order coherence had been measured indirectly. But the direct, time-domain observation of the emergence of a global phase from an incoherent state had not been achieved for any BEC system, atomic, exciton-polariton, or magnon.
The result also validates that quasiparticle BECs obey the same fundamental coherence physics as atomic BECs, despite being non-equilibrium, dissipative systems. This has practical implications: magnon BECs operate at room temperature and microwave frequencies, making them potentially useful platforms for phase-based information processing and magnonic supercurrent devices.
Several caveats apply. Magnon BECs are non-equilibrium condensates, they exist only under continuous pumping, which differs from equilibrium atomic BECs. The measurement is inductive, not a direct quantum measurement of the wavefunction, and the antenna acts as a spatial filter that averages over the film. Nevertheless, the observation is unambiguous: the phase emerges from nothing, chosen by the system itself.
Source: Koster, M., Schweizer, M.R., Noack, T. et al. “Emergence of phase coherence in a magnon Bose-Einstein condensate.” Nature Physics (2026). DOI: 10.1038/s41567-026-03373-6