The quantum world is full of surprises: swap two electrons, and their combined wavefunction just changes sign; swap two photons, and nothing happens at all. But there are particles that behave very differently when exchanged, as if they remember being swapped. These particles are called anyons, and they can only exist in a two-dimensional world.
In the new work, physicists studied the fractional quantum Hall state at filling factor 5/2. In this state, non-Abelian anyons are believed to exist. Ordinary (Abelian) anyons, when exchanged, merely add a phase to the wavefunction, like two strands simply twisting when braiding a plait. Non-Abelian anyons, on the other hand, transform the entire state of the system, as if the pattern that emerges from braiding depends on the order in which the strands are interwoven.
The experimentalists used a clever trick: they measured current noise from weak partitioning of a very dilute one-dimensional edge channel. Such noise arises from the random transmission or reflection of quasiparticles. It turned out that the noise from the 'lower' charged channel corresponds to Abelian anyons, while the noise from the 'upper' neutral channel corresponds to non-Abelian anyons. Calculations for the topological order 'particle-hole Pfaffian' matched the measurements.
This is the first time non-Abelian braiding has been observed in an experiment. Previously, physicists had demonstrated Abelian braiding at filling factor 1/3, and now they have taken a step toward more exotic states. Such quasiparticles could become the basis for topological quantum computers, where information is stored in 'braiding patterns' and is therefore less susceptible to noise.
Anyons can only exist in a two-dimensional world. In our three-dimensional space, exchanging two particles always results in either a sign change (fermions) or nothing (bosons), but in a flat world, intermediate possibilities are possible.
The idea of using anyon braiding for quantum computing evokes science fiction tales of topological computers, where information is encoded in knots and braids rather than bits.