Particle Physics
The Chiral Anomaly: How Quantum Loops Break a Classical Symmetry
Write down the Lagrangian for a massless charged fermion coupled to electromagnetism and you find two conserved currents: the ordinary vector current (charge) and the axial current (the difference between left- and right-handed particles). Noether's theorem guarantees both — at the classical level. But quantize the theory and one of them dies: the axial current acquires a divergence ∂μ jμ5 = (e²/16π²) εμναβFμνFαβ, sourced by parallel electric and magnetic fields. This is the chiral (Adler–Bell–Jackiw) anomaly: a symmetry that is exact in the classical action is unavoidably broken by the quantum measure of the path integral.
Far from a pathology, the anomaly is a hard, measurable prediction. It fixes the rate of π⁰ → 2γ decay to within a percent — Γ ≈ 7.75 eV, matching a neutral-pion lifetime of ~8.5×10⁻¹⁷ s — and it drives negative longitudinal magnetoresistance in Weyl semimetals like TaAs, turning a subtle feature of relativistic quantum field theory into a benchtop transport signature.
- RegimeRelativistic quantum field theory; massless (chiral) fermions
- Key relation∂_μ j^μ5 = (e²/16π²) ε^{μναβ} F_μν F_αβ
- DiscoveredAdler; Bell & Jackiw, 1969 (roots in Steinberger 1949)
- Characteristic scaleΓ(π⁰→γγ) ≈ 7.75 eV; τ ≈ 8.5×10⁻¹⁷ s
- Realized inπ⁰→2γ decay; Weyl/Dirac semimetals (TaAs, Na₃Bi)
- Matters forQCD, Standard-Model consistency, topological transport
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
A conservation law that quantum mechanics refuses to honor
Symmetries and conservation laws are supposed to be inseparable — that is Noether's theorem. For a massless Dirac fermion, the classical action is invariant under two independent U(1) phase rotations: an ordinary one (ψ → eiαψ) giving the conserved vector current jμ = ψ̄γμψ, and an axial one (ψ → eiβγ⁵ψ) rotating left- and right-handed components oppositely, giving the axial current jμ5 = ψ̄γμγ⁵ψ. Classically ∂μjμ5 = 0, so the difference NR − NL (net chirality) is conserved.
The chiral anomaly is the discovery that this second law is false in the quantum theory. No matter how you regularize, the loop corrections generate a nonzero divergence. It is not an approximation artifact or a broken assumption — it is exact, one-loop-complete, and cannot be removed by any legitimate field redefinition. This makes it one of the sharpest examples of a quantum effect with no classical shadow.
The mechanism: the triangle diagram and the path-integral measure
Two complementary derivations expose the same truth. (1) The triangle diagram. Compute the correlator of one axial current with two vector currents ⟨jμ5 jν jρ⟩ — a fermion loop with three vertices. You cannot simultaneously enforce conservation of the vector current (needed for charge/gauge invariance) and the axial current: regularizing the linearly divergent loop forces the anomaly onto whichever current you sacrifice. Physics keeps jμ conserved, so the anomaly lands on jμ5. Crucially the coefficient is exact at one loop (Adler–Bardeen theorem) — higher loops add nothing.
(2) Fujikawa's measure. An axial rotation is a change of variables in the path integral. Its Jacobian is not 1: the fermion measure Dψ Dψ̄ is not invariant, and the noninvariance equals the anomaly. Its integral, ∫ε F F̃ d⁴x, counts the difference in zero-mode number between chiralities — the Atiyah–Singer index of the Dirac operator. The anomaly is thus topological: it counts winding of the gauge field, tying particle physics to index theory and instantons.
The equation, the coefficient, and the numbers
For a single fermion of charge e in QED the anomalous divergence is
∂μ jμ5 = (e²/16π²) εμναβ FμνFαβ = (e²/2π²) E·B.
The right side is nonzero only when E and B are non-orthogonal — the pseudoscalar F·F̃ ∝ E·B. The coefficient 1/16π² is fixed, dimensionless, and universal; it depends only on the charges and multiplicities of the looping fermions, never on coupling strength or energy scale. In QCD, gluon fields replace F and the anomaly of the flavor-singlet axial current carries a factor Nc = 3 (three quark colors) and a Tr over generators.
The most quantitative payoff is the π⁰ → 2γ width. The anomaly predicts Γ = (α² mπ³ Nc²)/(576 π³ fπ²) ≈ 7.75 eV with no free parameters, corresponding to a lifetime τ ≈ 8.5×10⁻¹⁷ s. The Nc² dependence is why π⁰ decay was an early confirmation that quarks come in three colors.
How it is measured: from a 7.75 eV width to a resistor that drops
Particle physics. The cleanest test is the neutral-pion lifetime. Naive PCAC (partial conservation of the axial current) would make π⁰ → 2γ nearly vanish; the observed, large rate is entirely the anomaly. The PrimEx experiments at Jefferson Lab used the Primakoff effect (π⁰ photoproduction off a nuclear Coulomb field) to measure Γ(π⁰→γγ) = 7.80 ± 0.12 eV, in striking agreement with the anomaly prediction of ~7.75 eV. Percent-level agreement validates both the anomaly coefficient and Nc = 3.
Condensed matter. In a Weyl semimetal the low-energy quasiparticles are 3D chiral (Weyl) fermions living at nodes of opposite chirality. Apply E ∥ B and the anomaly pumps charge from one node to the other — the same E·B term — creating a chirality imbalance that relaxes slowly. The result is negative longitudinal magnetoresistance: resistance drops as field aligns with current. This was reported in TaAs (Phys. Rev. X, 2015) and Na₃Bi, a direct solid-state readout of a QFT anomaly.
Where it operates — and what it is not
The anomaly is generic to any theory with chiral fermions coupled to gauge fields: QED, QCD, the electroweak sector, and lattice/emergent Weyl systems. It is essential that at least one current be axial (γ⁵-carrying) — a purely vector theory has no anomaly to worry about. It requires massless or nearly massless fermions for chirality to be a good symmetry; a mass term explicitly breaks chirality and adds a separate 2m ψ̄γ⁵ψ piece, but the anomaly term survives independently.
Distinguish it carefully from relatives. Spontaneous symmetry breaking hides a symmetry in the vacuum but keeps the current conserved; the anomaly instead breaks the conservation law itself. The scale (trace) anomaly breaks classical scale invariance and generates the running of couplings — different current, same 'quantum kills a classical symmetry' theme. And the gauge anomaly is fatal: if the anomaly afflicted a gauged current the theory would be inconsistent, which is why Standard-Model hypercharges must cancel per generation.
Why it matters: from anomaly cancellation to topological transport
The chiral anomaly is a load-bearing beam of modern physics. Its cancellation is a consistency requirement: the sum of gauge anomalies over each Standard-Model generation must vanish, which quantizes and correlates the quark and lepton hypercharges — a nontrivial constraint the observed particle content precisely satisfies. The related global-anomaly counting underlies why π⁰ decays and, via the Wess–Zumino–Witten term, organizes low-energy meson interactions.
The QCD version resolves the U(1)A problem: the anomaly explains why the η′ meson is heavy (~958 MeV) rather than a light pseudo-Goldstone boson. But it also creates the strong CP problem (the θ-term ∝ E·B is anomaly-related), motivating axions. In cosmology, anomalous B+L violation via sphalerons enables electroweak baryogenesis. And in materials, the chiral magnetic effect and anomaly-driven transport now define an active frontier of topological matter. Open questions span axion detection, anomaly-induced transport in quark–gluon plasma (the chiral magnetic effect at RHIC), and using anomalies as nonperturbative constraints on strongly coupled theories.
| Quantity / setting | Classical (tree level) | Quantum (with loop) | Physical consequence |
|---|---|---|---|
| Vector current j^μ | Conserved (∂·j=0) | Conserved — anomaly kept off it by regularization | Electric charge is exactly conserved |
| Axial current j^μ5 | Conserved for m=0 | ∂_μ j^μ5 = (e²/16π²)ε F F̃ ≠ 0 | Chirality not conserved; L↔R interconvert |
| π⁰ → 2γ (QCD+QED) | Suppressed (~0 by PCAC) | N_c=3 anomaly triangle sets amplitude | Γ ≈ 7.75 eV, confirms three colors |
| Weyl semimetal in E∥B | Separate node charges fixed | Charge pumped between Weyl nodes | Negative longitudinal magnetoresistance |
| Global U(1)_A symmetry | Exact symmetry of action | Explicitly broken by measure | Solves the η′ mass ('U(1) problem') |
Frequently asked questions
Why does the anomaly break the axial current but not the electric current?
The triangle diagram cannot conserve both currents at once — regularization forces the anomaly onto one of them. Electric charge conservation is tied to gauge invariance, which is non-negotiable for a consistent theory, so we impose conservation of the vector current j^μ. The anomaly then unavoidably shows up in the axial current j^μ5. It is a choice dictated by consistency, not by the diagram, which is symmetric until you regularize.
Is the anomaly coefficient a one-loop approximation that gets corrected at higher orders?
No. The Adler–Bardeen theorem states the anomaly coefficient is exact at one loop — higher-order diagrams do not renormalize it. This is because the anomaly is topological: its integral counts the Atiyah–Singer index, an integer that cannot shift continuously. That exactness is what makes the π⁰→2γ rate a parameter-free, percent-accurate prediction rather than a rough estimate.
What exactly does E·B have to do with it?
The anomaly source ε^{μναβ}F_μν F_αβ equals −8 E·B (magnitude 8 E·B), a pseudoscalar. It is nonzero only when the electric and magnetic fields have a component along each other. Parallel E and B pump net chirality into the fermion sea. In Weyl semimetals this same E·B term transfers charge between Weyl nodes, producing negative longitudinal magnetoresistance.
How does π⁰ → 2γ prove there are three colors of quarks?
The anomaly-fixed decay width scales as N_c². With N_c=1 the predicted rate is nine times too small; with N_c=3 it is Γ ≈ 7.75 eV, matching the measured 7.80 ± 0.12 eV from Jefferson Lab's PrimEx. So the neutral-pion lifetime is a direct, quantitative count of quark colors — an early and clean piece of evidence for N_c=3 in QCD.
How is a high-energy field-theory anomaly seen in a solid?
Weyl semimetals host emergent 3D chiral fermions at pairs of band-touching nodes with opposite chirality. Applying E parallel to B triggers the identical (e²/2π²)E·B anomaly, pumping electrons from one node to the other. The resulting chirality imbalance boosts conductivity along the field, so resistance falls with increasing aligned field — the negative longitudinal magnetoresistance observed in TaAs and Na₃Bi.
What is the difference between the chiral anomaly and spontaneous symmetry breaking?
Spontaneous symmetry breaking keeps the symmetry (and its conserved current) intact in the equations but chooses a non-symmetric ground state, producing Goldstone bosons. The chiral anomaly instead destroys the conservation law itself at the quantum level — the current has a genuine divergence. In QCD both act on chiral symmetry: quark condensation breaks it spontaneously (giving pions as pseudo-Goldstones), while the anomaly explicitly breaks the singlet U(1)_A (making the η′ heavy).