Anyons and Fractional Statistics

Anyons crossed from theory into measured fact in 2020. A Paris group (Bartolomei et al.) built an "anyon collider" — a beam-splitter for $\nu=1/3$ quasiparticles — and read the fractional exchange statistics directly from cross-correlated current noise, while a Purdue group (Nakamura et al.) resolved the anyonic braiding phase $\theta=2\pi/3$ as flux-periodic conductance oscillations in a Fabry–Pérot interferometer. Both confirmed what Arovas, Schrieffer, and Wilczek had computed 36 years earlier. On the non-abelian frontier, the $\nu=5/2$ quantum Hall plateau is the leading candidate for Ising anyons: a 2018 thermal-Hall measurement found a half-integer quantized thermal conductance consistent with a chiral Majorana edge mode, though its interpretation is still debated. The engineered route runs through semiconductor–superconductor nanowires and, more recently, planar Josephson-junction and quantum-dot arrays aimed at demonstrating Majorana braiding on a chip; in 2023 a Google-led collaboration braided non-abelian anyons in a surface-code processor as a digital simulation, exhibiting the predicted noncommuting statistics on qubits rather than in a material. The prize at the end — a topological qubit whose error rate is suppressed exponentially by an energy gap rather than by active error correction — is why major industrial efforts continue to chase Majorana and Fibonacci platforms.

Start with the topology, because everything else is a corollary of it. Track two identical particles as their worldlines sweep upward through time. To exchange them and return the configuration to where it began is to draw a closed loop in the space of two-particle configurations; the operator that acts on the wavefunction depends only on which homotopy class that loop belongs to. In three dimensions there is always a spare direction into which a double-exchange loop can be tilted and reeled in, so it is contractible and $\sigma^2=\mathbb{1}$. In the plane there is no spare direction: a double exchange threads one strand through the region swept by the other, and the winding cannot be undone. The first figure is that contrast, side by side.

Once fractional statistics is allowed, the cleanest mental model of an abelian anyon is a charge–flux composite: bind an electric charge $q$ to a thin solenoid carrying magnetic flux $\Phi$. Carry one composite all the way around another and its charge sees the partner's flux, acquiring an Aharonov–Bohm phase $q\Phi/\hbar c$ that depends only on whether the loop encloses the partner — a topological invariant. A single exchange is half of that loop, so the statistical angle is $\theta = q\Phi/2\hbar c$. This makes the fractional phase concrete and shows why it cannot be gauged away locally. The second figure places the real systems on the statistical-angle line and records the two numbers that matter for non-abelian anyons.

Everything begins with the fundamental group of the configuration space of $n$ indistinguishable points. Its homotopy classes are the exchanges, and the group they form decides what statistics are possible.

where: $C_n(\mathbb{R}^d)$ — configuration space of $n$ indistinguishable points in $d$ dimensions (identical points removed and identified); $\pi_1$ — fundamental group, the homotopy classes of closed loops; $B_n$ — braid group on $n$ strands; $S_n$ — symmetric (permutation) group. Reading: in the plane the loop space is infinite (braids never repeat), in higher dimensions it collapses to finite permutations. The braid group has a presentation by generators $\sigma_i$ that exchange neighbours. Two relations define it, and — crucially — the relation $\sigma_i^2=1$ of the symmetric group is absent.

where: $\sigma_i$ — the generator that carries strand $i$ counterclockwise over strand $i+1$. The first identity is the Yang–Baxter (braid) relation; the second says distant exchanges commute. Because no power of $\sigma_i$ is the identity, $\sigma_i$ has infinite order — the algebraic face of "a double exchange winds." For a one-dimensional (abelian) representation of $B_n$, every generator maps to the same scalar phase, and the many-body wavefunction transforms under exchange by that phase.

where: $\Psi$ — many-anyon wavefunction; $\theta$ — statistical angle, continuous in 2D; $\mathbf{r}_i$ — position of the $i$-th anyon. Setting $\theta=0$ recovers bosons, $\theta=\pi$ recovers fermions, everything else is an anyon. The charge–flux model gives $\theta$ a physical origin as half an Aharonov–Bohm phase.

where: $q$ — charge bound to the composite; $\Phi$ — flux bound to it; $\hbar c$ — in Gaussian units; $\gamma_{\text{loop}}$ — phase for a full encirclement, twice the exchange angle. The phase is topological: it counts enclosures, not distances. Beyond phases, anyons are organized by their fusion rules — the superselection algebra of how two anyons combine.

where: $a,b,c$ — anyon charge labels; $N_{ab}^{\,c}$ — nonnegative integer fusion multiplicities; $d_a$ — quantum dimension of $a$, so the Hilbert space of $n$ copies grows as $d_a^{\,n}$. If any $\sum_c N_{ab}^{\,c}>1$ the anyons are non-abelian and $d_a>1$. The braiding and associativity data are packaged in the $R$- and $F$-matrices, constrained by two coherence equations that make the whole theory consistent.

where: $R^{ab}_c$ — the unitary from braiding $a$ past $b$ inside fusion channel $c$; $F^{abc}_d$ — the basis change between the fusion orderings $(ab)c$ and $a(bc)$. The pentagon and hexagon identities are the anyonic Yang–Baxter and associativity constraints; their solutions are the modular tensor categories. Two model theories fix the anchor numbers of the field.

where: $\varphi$ — golden ratio, from $\tau\times\tau=1+\tau\Rightarrow d^2=1+d$; $\sqrt2$ — from $\sigma\times\sigma=1+\psi\Rightarrow d^2=2$. Fibonacci braiding is universal; Ising braiding is Clifford-only. The Ising theory is realized by Majorana zero modes, whose braid is an explicit exponential of the Majorana algebra.

where: $\gamma_i$ — Majorana operator, Hermitian with $\gamma_i^2=1$; $U_{ij}$ — the unitary implementing a counterclockwise exchange of adjacent modes. The identity uses $(\gamma_i\gamma_j)^2=-1$, so the exponential collapses to a $\pi/4$ rotation.

The statistical angle $\theta$ and the quantum dimension $d$ together classify every planar particle type; the familiar bosons and fermions are two corners of a much larger space.

1 · Charge and statistics of a ν = 1/3 Laughlin quasiparticle (fully worked)

The $\nu=1/3$ fractional quantum Hall plateau hosts Laughlin quasiholes. Find (a) the electric charge, (b) the full-braid phase when one quasihole encircles another, and (c) the single-exchange statistical angle. Then say what a full braid does to the two-quasihole wavefunction.

Step 1. Read off $m$. The filling factor is $\nu=1/m$ with $m=3$ (odd, as required for a fermionic Laughlin state). Every fractional quantity below is a fixed function of this one integer.

Step 2. Charge. From Derivation 2 the quasihole carries $q^*=e/m=\mathbf{e/3}$. This is the sharpest signature of fractionalization: a localized excitation with one-third of an electron's charge, and it is what fractional shot-noise experiments actually measure (Fano factor $1/3$).

Step 3. Full-braid phase. Encircling one quasihole completely around another contributes the statistical phase $\gamma_{\text{loop}}=2\pi/m=2\pi/3$, so the wavefunction is multiplied by $e^{2\pi i/3}$ — a nontrivial cube root of unity, $\mathbf{e^{2\pi i/3}}$, neither $+1$ (boson) nor a sign flip alone.

Step 4. Single exchange. A single swap is half a braid, so $\theta=\tfrac12\cdot 2\pi/3=\pi/3\approx\mathbf{1.047\ rad}$, i.e. $60^\circ$. Since $0<\pi/3<\pi$, this quasiparticle sits strictly between boson and fermion on the statistical line of the second figure.

Step 5. Interpretation. Braiding two $e/3$ quasiholes once multiplies the state by $e^{2\pi i/3}$; doing it three times returns to the identity ($e^{2\pi i}=1$). Charge e/3, full-braid phase e^{2πi/3}, exchange angle θ = π/3. The abelian anyon is fully specified by these numbers — there is no hidden degeneracy, because $d=1$.

2 · Counting the Fibonacci fusion space (why d = φ)

A collection of $n$ Fibonacci anyons $\tau$ (with $\tau\times\tau=1+\tau$) is prepared with total charge $\tau$. How many internal fusion states does it have as $n$ grows, and what is the quantum dimension?

Step 1. Build the fusion tree left to right, fusing one $\tau$ at a time and tracking the running total charge, which can be $1$ (vacuum) or $\tau$. Let $a_n$ be the number of trees ending in total charge $\tau$ and $b_n$ the number ending in $1$. The rule $\tau\times\tau=1+\tau$ and $\tau\times1=\tau$ give the recursion $a_{n+1}=a_n+b_n$, $b_{n+1}=a_n$.

Step 2. Eliminate $b$: $a_{n+1}=a_n+a_{n-1}$ — the Fibonacci recursion. Starting from a single anyon ($a_1=1,\ b_1=0$) the counts are $a_1,a_2,\dots=1,1,2,3,5,8,13,21,\dots$, the Fibonacci numbers $F_n$. So $n$ anyons fusing to total charge $\tau$ have $F_n$ internal states.

Step 3. Quantum dimension. By definition $d_\tau$ is the growth rate of the space, $\dim\sim d_\tau^{\,n}$, so $d_\tau=\lim_{n\to\infty}F_{n+1}/F_n$. The Fibonacci ratios $1,2,1.5,1.667,1.6,1.625,\dots$ converge to $\mathbf{\varphi=(1+\sqrt5)/2\approx 1.618}$.

Step 4. Consistency with fusion. The quantum dimension must satisfy $d^2=1+d$ directly from $\tau\times\tau=1+\tau$ (take dimensions of both sides). Solving, $d=\varphi$ — the same golden ratio, from algebra rather than counting. n Fibonacci anyons carry F_n fusion states; d_τ = φ = 1.618, irrational, so braiding is a resource.

Step 5. Why it matters. Because $d_\tau>1$ is irrational, the degeneracy cannot be reproduced by any local tensor-product structure; the information is genuinely nonlocal. Four $\tau$ anyons with total charge $1$ span a two-dimensional space — one logical qubit — and their braids generate a dense subgroup of $SU(2)$, so any single-qubit gate is approximable to arbitrary accuracy by braiding alone.

3 · The Ising braid gate

Two Majorana modes $\gamma_1,\gamma_2$ are exchanged counterclockwise. Write the braid unitary explicitly and identify the gate.

Step 1. From equation (8), $U_{12}=\exp(\tfrac{\pi}{4}\gamma_1\gamma_2)$. Use $(\gamma_1\gamma_2)^2=-1$ so the series splits into cosine and sine: $U_{12}=\cos\tfrac{\pi}{4}\,\mathbb{1}+\sin\tfrac{\pi}{4}\,\gamma_1\gamma_2$.

Step 2. Evaluate: $\cos\tfrac{\pi}{4}=\sin\tfrac{\pi}{4}=1/\sqrt2$, giving $U_{12}=\tfrac{1}{\sqrt2}(\mathbb{1}+\gamma_1\gamma_2)$. In the fermion basis $c=(\gamma_1+i\gamma_2)/2$, the operator $i\gamma_1\gamma_2=2c^\dagger c-1$ is the parity $Z$, so $U_{12}=\tfrac{1}{\sqrt2}(\mathbb{1}-iZ)=e^{-i\pi Z/4}$.

Step 3. Identify the gate. $e^{-i\pi Z/4}$ is the phase gate $S$ up to a global phase — a Clifford operation, the single-qubit $\sqrt{Z}$. The Ising braid is U₁₂ = (1 + γ₁γ₂)/√2 = e^{-iπZ/4}, a Clifford (phase-gate) rotation — protected, but needing a magic-state T gate for universality.

Fractional statistics was a theorist's discovery before it was an experimentalist's — imagined from the topology of the plane a decade before anyone measured a fractional charge, and named with a joke that stuck.

Texts. Preskill, Lecture Notes on Quantum Computation, Ch. 9 (Topological Quantum Computation) — the canonical pedagogical account · Nielsen & Chuang, Quantum Computation and Quantum Information, §10.6 · Nayak, Simon, Stern, Freedman, Das Sarma, "Non-Abelian Anyons and Topological Quantum Computation," Rev. Mod. Phys. 80 (2008) 1083 — the definitive review · Kitaev, "Fault-tolerant quantum computation by anyons," Annals of Physics 303 (2003) 2 · Kitaev, "Anyons in an exactly solved model and beyond," Annals of Physics 321 (2006) 2 · Wilczek, Fractional Statistics and Anyon Superconductivity (World Scientific, 1990) · Pachos, Introduction to Topological Quantum Computation (Cambridge, 2012). The key results collected in this article are: (1) $\pi_1(C_n(\mathbb{R}^2))=B_n$ and $\pi_1(C_n(\mathbb{R}^d))=S_n$ for $d\geq3$, with $\sigma_i$ having infinite order; (2) the Laughlin quasihole charge $q^*=e/m$ and statistics $\theta=\pi/m$ — for $m=3$, $q^*=e/3$ and $\theta=\pi/3$, consistent with the 2020 anyon-collider and Fabry–Pérot experiments; (3) the Majorana braid $U_{ij}=\exp(\tfrac{\pi}{4}\gamma_i\gamma_j)=\tfrac{1}{\sqrt2}(1+\gamma_i\gamma_j)$, derived using $(\gamma_i\gamma_j)^2=-1$, equal to $e^{-i\pi Z/4}$ (a Clifford $S$-gate); (4) the quantum dimensions $d_{\text{Fib}}=\varphi=(1+\sqrt5)/2$ from $\tau\times\tau=1+\tau$ and $d_\sigma=\sqrt2$ from $\sigma\times\sigma=1+\psi$, with $n$ Fibonacci anyons carrying $F_n$ fusion states.

Prerequisites. Topological Quantum Computation · Toric Code and Topological Order · Non-Abelian Braiding and Fusion Rules. Leads to. Fibonacci and Ising Anyons · Majorana Zero Modes · Magic State Distillation.

Ask the tutor: Ask me to walk through the Berry-phase derivation of the $e/3$ quasiparticle, or to show explicitly how a Fibonacci braid approximates a Hadamard gate.