Anyons: Braiding Statistics Between Bosons and Fermions

What anyons are and why the plane is special

In three dimensions, quantum statistics is dictated by the permutation group: two paths that swap identical particles can be continuously deformed into one another, so exchanging particles twice is equivalent to doing nothing. This forces the exchange operator to square to the identity, allowing only its two eigenvalues +1 (bosons) and −1 (fermions) — the spin-statistics dichotomy.

In two dimensions the argument fails. A clockwise exchange cannot be smoothly deformed into a counterclockwise one, because the worldlines braid around each other and the plane keeps track of the winding. The relevant symmetry is the braid group Bₙ, whose generators need not square to the identity. Its one-dimensional representations are pure phases e^iθ for arbitrary θ — Abelian anyons — and its higher-dimensional representations give non-Abelian anyons. This is why Jon Leinaas and Jan Myrheim (Oslo, 1977) found that 2D particles can carry any statistical angle, and why Frank Wilczek (1982) coined "anyons." Anyons are not fundamental particles but emergent quasiparticles of strongly correlated 2D matter.

The mechanism: charge–flux composites and the Aharonov–Bohm phase

The cleanest microscopic picture is the charge–flux composite. Attach to each particle a thin magnetic flux tube of strength Φ. When one composite (charge q) encircles another, it acquires an Aharonov–Bohm phase e^iqΦ/ℏ. An exchange is half a full loop, so the exchange phase is θ = ½·(qΦ/ℏ). Tuning the bound flux tunes θ continuously between bosonic and fermionic values — a concrete model realizing fractional statistics.

In the fractional quantum Hall (FQH) effect this happens naturally. Robert Laughlin's 1983 wavefunction for the ν = 1/m state describes an incompressible electron liquid whose elementary excitations (quasiholes) carry fractional charge e/m. Each quasihole is effectively a bound state of charge and the depleted correlation hole (an emergent flux). Daniel Arovas, John Robert Schrieffer, and Frank Wilczek (1984) computed the Berry phase for adiabatically dragging one quasihole around another and found a statistical phase θ = π/m — precisely the anyonic result. Braiding is topological: the phase depends only on how many times worldlines wind, not on the path's geometry, energy, or timing.

The key relations, phases, and characteristic scales

For Abelian anyons, exchanging two identical anyons multiplies the many-body state by a single phase:

ψ(1,2) → e^iθ ψ(2,1),  θ = πα,  α ∈ [0,2] (α=0 boson, α=1 fermion).

A full braid (one anyon encircling another, = two exchanges) yields the monodromy phase 2θ. For the Laughlin ν = 1/m state, θ = π/m and the braid phase is 2π/m; at ν = 1/3 this is a braid phase of 2π/3 with quasiparticle charge e/3.

For non-Abelian anyons, braiding acts as a unitary matrix on a degenerate ground-state manifold: n Ising anyons span a Hilbert space of dimension ~2^n/2, and different braids generally do not commute — the outcome depends on the order. Energy scales set the stage: FQH gaps are ~1–5 K (of order 0.1–0.5 meV), demanding temperatures of ~10 mK, magnetic fields B ≈ 5–15 T, and ultra-clean GaAs with mobilities exceeding 10⁷ cm²/V·s so that disorder does not localize the quasiparticles.

How anyons are measured: interferometry and collisions

Because the braiding phase is topological, it is read out by making anyon worldlines wind and detecting interference. In 2020, James Nakamura and colleagues (Purdue) built an electronic Fabry–Pérot interferometer in a GaAs 2D electron gas at ν = 1/3. As anyons circulated the loop, the Aharonov–Bohm interference of the edge current showed discrete phase slips of 2π/3 each time an additional e/3 quasiparticle entered the interference region — the smoking-gun signature of a braiding phase θ_anyon = 2π/3. The key was operating where the device charging energy is small compared to the quasiparticle formation energy, isolating statistics from Coulomb effects.

Independently and nearly simultaneously, Gwendal Fève's group (Bartolomei et al., Paris, 2020) collided two dilute beams of e/3 anyons at a beamsplitter and measured cross-correlations of current fluctuations. Anyonic statistics produces a bunching tendency absent for fermions; fitting the partition noise yielded an exchange phase ϕ = π/3, matching θ = π/m for m = 3. Graphene interferometers (2023–2024) have since reproduced anyonic phase slips with telegraph noise as a control.

Where anyons live, and how they differ from related effects

Anyons require topological order — long-range entanglement and a ground-state degeneracy that depends on the manifold's genus, not on any broken symmetry. Confirmed and candidate hosts include the Laughlin FQH states ν = 1/3, 1/5, 2/5 (Abelian, well established); the ν = 5/2 Moore–Read Pfaffian state (candidate non-Abelian Ising anyons, charge e/4); the proposed ν = 12/5 Read–Rezayi state (Fibonacci anyons); Kitaev's toric-code and honeycomb models; and Majorana zero modes in topological superconductors and semiconductor–superconductor nanowires, which realize Ising anyons at defects.

Distinctions matter. Anyons are not merely fractionally charged particles — fractional charge (e/3) and fractional statistics (θ = π/3) are logically independent, though they coincide in Laughlin states. Anyonic statistics is a genuine many-body topological property, unlike the single-particle Berry phase of a Bloch electron (though both are geometric phases). And unlike spin-statistics in 3D, the anyon connection between spin and statistics is continuous: the topological spin e^2πi·s equals the self-braiding phase.

Applications, open questions, and significance

The headline application is topological quantum computation (Kitaev 1997; Freedman, Nayak et al.). Because braiding is topological, a qubit encoded in the fusion space of non-Abelian anyons is immune to local noise — errors would require a physical worldline to wind incorrectly, which local perturbations cannot cause. Gates are implemented simply by braiding anyons around one another. Ising anyons (ν = 5/2, Majorana modes) yield only the Clifford gates — powerful for error correction but not universal, requiring a supplementary non-topological T-gate (magic-state distillation). Fibonacci anyons (candidate ν = 12/5) are computationally universal by braiding alone.

Open frontiers: unambiguously confirming non-Abelian statistics at ν = 5/2 (interferometry there is hard and results remain contested); scaling Majorana-based platforms into demonstrable braiding operations; and finding anyons in lattice spin liquids and Kitaev materials. Conceptually, anyons show that quantum statistics is richer than the boson/fermion binary and tie together topology, entanglement, and fault tolerance — a rare case where an abstract mathematical possibility became a measured number and a computing paradigm.

Exchange statistics of identical particles: bosons, fermions, Abelian anyons, and non-Abelian anyons

Property	Bosons	Fermions	Abelian anyons	Non-Abelian anyons
Exchange phase	e^{i0} = +1	e^{iπ} = −1	e^{iθ}, θ arbitrary	matrix U (unitary)
Governing group	Permutation Sₙ	Permutation Sₙ	Braid group Bₙ (1D reps)	Braid group Bₙ (higher-D reps)
Ground-state degeneracy	None required	None required	None (Abelian phase only)	Degenerate fusion Hilbert space
Spatial dimension	any D	any D	D = 2 only	D = 2 only
Example / realization	photons, He-4	electrons, He-3	ν=1/3 FQH quasiholes (charge e/3)	ν=5/2 Moore–Read (charge e/4)
Order of operations	commutes	commutes	commutes (phases)	does NOT commute (non-Abelian)