The Electroweak Sphaleron: A Saddle That Erases Baryon Number

What the sphaleron is and why it matters

The electroweak sector, SU(2)ₗ × U(1)ᵧ with the Higgs field, has infinitely many degenerate classical vacua. They look identical locally but differ by an integer topological label — the Chern–Simons number N_CS — that counts how many times the gauge field winds around the group manifold. To pass from one vacuum to an adjacent one, the fields must climb over an energy barrier. The sphaleron is the configuration sitting exactly at the top of the lowest such barrier: a static, spherically-arranged, but unstable saddle point of the energy functional with exactly one negative (downhill) direction.

Why care? Because of the Adler–Bell–Jackiw chiral anomaly, a change ΔN_CS = 1 is inescapably tied to a change in fermion number: ΔB = ΔL = 3. The Standard Model, entirely on its own, violates baryon number. This is the essential ingredient — and the essential constraint — for any explanation of why the universe contains matter but almost no antimatter.

The mechanism, step by step

Start with the vacuum structure. Pure SU(2) gauge configurations that are pure gauge at spatial infinity fall into homotopy classes indexed by an integer winding number; the energy plotted against N_CS looks like a tilted washboard of degenerate minima separated by barriers. This is the periodic vacuum first appreciated by 't Hooft (1976) in the instanton context.

Now interpolate between two neighbouring vacua along the lowest path. The energy rises, peaks, and falls. Manton (1983) proved on topological (non-contractible loop) grounds that such a barrier maximum must exist; Klinkhamer and Manton (1984) constructed the explicit solution in the full Weinberg–Salam theory, giving it the name sphaleron. At the peak, N_CS = ½ (half-integer, midway between vacua), and the gauge and Higgs fields form a localized, magnetically-charged-looking knot.

The anomaly does the bookkeeping. As the fields sweep from N_CS = 0 to 1, the Dirac sea levels cross zero energy — one level per fermion doublet per colour. Twelve fermionic levels cross, producing the 12-fermion operator (uude)-like structure: nine quarks plus three leptons, ΔB = ΔL = +3, with B−L untouched.

The key equation, characteristic numbers and scales

The zero-temperature barrier height is set by the weak scale and the SU(2) coupling g:

E_sph = (4π v / g) · B(λ/g²) ≈ (2 M_W / α_W) · B

where v ≈ 246 GeV is the Higgs vacuum expectation value, M_W ≈ 80.4 GeV, α_W = g²/4π ≈ 1/30, and B is a dimensionless shape factor that runs from about 1.56 to 2.72 as the Higgs self-coupling λ/g² varies. Plugging in the physical Higgs mass (125 GeV) gives E_sph ≈ 9 TeV.

The transition rate has two faces. In the broken (cold) phase it is Boltzmann-suppressed, Γ/V ∝ e^(−E_sph(T)/T); a Higgs VEV keeps the barrier tall. In the symmetric (hot) phase the barrier vanishes and dimensional analysis plus lattice measurement give an unsuppressed rate Γ/V = (25.4 ± 2.0) α_W⁵ T⁴ (Moore and collaborators). By comparison the T = 0 tunnelling amplitude carries a factor e^(−2π/α_W) ≈ e^(−170) ≈ 10⁻⁷⁴ — deader than dead.

How it is realized, measured, or observed

The sphaleron has never been produced in a laboratory, and probably cannot be: at a collider the ~9 TeV energy is in principle available at the LHC, but the relevant amplitude for a two-particle initial state to fake a coherent multi-particle sphaleron transition is exponentially suppressed (the "unitarity" or Ringwald–Espinosa problem). Dedicated LHC searches for anomalous multi-jet-plus-lepton, B+L-violating final states (ATLAS, CMS) have found nothing and set limits, consistent with expectations.

Its quantitative properties are instead established two ways. First, by explicitly solving the classical Euler–Lagrange equations for the static saddle — a boundary-value problem in a single radial coordinate with the sphaleron ansatz — which yields the energy, profile, and the single unstable negative mode (recently re-examined in 2025 lattice-and-numerical studies). Second, by lattice thermal field theory: real-time classical-statistical simulations of hot SU(2) gauge–Higgs theory directly measure the diffusion of N_CS and hence the rate Γ/V through the electroweak crossover, pinning the symmetric-phase coefficient near 25 α_W⁵.

Where it operates, and how it differs from related effects

The sphaleron's arena is the hot early universe, at temperatures above the electroweak crossover (T_c ≈ 160 GeV, roughly 10⁻¹¹ s after the Big Bang). There, N_CS diffuses freely and B+L is driven to zero over a Hubble time; below T_c the rate freezes out and whatever asymmetry survives is locked in.

Contrast with the instanton: same underlying vacuum topology, but the instanton tunnels quantum-mechanically through the barrier at T = 0 (amplitude ~10⁻⁷⁴), whereas the sphaleron is a thermal, over-the-barrier process. Contrast too with QCD sphalerons, which change chirality but not baryon number (SU(3) has no anomaly linking to B). And note what the sphaleron does not do: it exactly conserves B−L, because that combination is anomaly-free in the Standard Model. So a sphaleron can never create net B−L — a crucial selection rule that shapes every viable baryogenesis scenario.

Applications, open questions, and significance

The sphaleron is the linchpin of the third Sakharov condition applied to the electroweak scale. In electroweak baryogenesis, a strongly first-order phase transition provides departure from equilibrium; sphalerons acting in front of the expanding bubble walls, combined with new CP violation, generate a baryon asymmetry that the broken-phase suppression then preserves. The Standard Model fails here — with a 125 GeV Higgs the transition is a smooth crossover, not first-order, and CP violation is too weak — which is a leading motivation for physics beyond the Standard Model.

In leptogenesis, heavy right-handed neutrino decays make a lepton asymmetry; sphalerons then reprocess a definite fraction (the factor 28/79 in the Standard Model) into baryon number, tying the observed η_B ≈ 6×10⁻¹⁰ to neutrino physics. Open questions abound: what makes the transition first-order (new scalars? the Higgs potential's true shape?), how large the true sphaleron rate is near a real first-order wall, and whether any collider signature of B+L violation is reachable at all.

Sphaleron versus the related instanton, and the two thermal regimes of the electroweak transition rate

Object / regime	Mechanism	Rate or amplitude	B+L violation
Instanton (T = 0)	Quantum tunneling under the barrier	∝ exp(−2π/α_W) ≈ exp(−170) ≈ 10⁻⁷⁴	Utterly negligible today
Sphaleron (T > 0)	Thermal hop over the barrier	∝ exp(−E_sph(T)/T)	Boltzmann-suppressed, but non-zero
Broken phase (T ≲ T_c ≈ 160 GeV)	Higgs VEV ≠ 0, barrier large	Γ/V ≈ (M_W/α_W T)⁴ e^(−E_sph/T)·T⁴	Freezes out; asymmetry preserved
Symmetric phase (T ≳ T_c)	No barrier, unsuppressed	Γ/V ≈ (25 ± 2)·α_W⁵ T⁴	Rapid; equilibrates B+L to zero