research30 min studyLesson 755 in Physics

Euclidean instanton actions and collective coordinates

Derive the Yang–Mills instanton bound with explicit trace conventions and identify the measure and boundary assumptions behind a semiclassical sector.

On this page

What you’ll understand

  • Distinguish Euclidean action saddles from static Lorentzian solitons.
  • Derive the topological action bound in one declared gauge normalization.
  • Explain collective coordinates, gauge quotient conventions, and instanton-size limitations.

An instanton is localized in Euclidean spacetime rather than being a static particle localized only in space. It contributes to an appropriately defined path integral through a saddle weighted exponentially by its Euclidean action. The exponent is only part of that contribution: fluctuation determinants, collective-coordinate measures, boundary states, and the integration contour matter too. We will derive the four-dimensional Yang–Mills action bound, identify the one-instanton parameters in a framed SU two problem, and avoid equating every finite-action configuration with a physical decay event.

Specify signature and normalization

Use positive Euclidean metric, orientation epsilon one two three four equal to plus one, and natural units with hbar equal to one. Gauge generators obey trace T a T b equal to one half delta a b. Put the coupling outside the action: F equals derivative terms minus i times the gauge commutator, with no coupling inside F. This differs from the preceding lesson’s canonical component convention and is declared anew to prevent importing an extra factor of g.

A Wick rotation is a prescription for relating suitable Lorentzian observables to Euclidean integrals, including boundary conditions and convergence information. The Euclidean action is not simply the real-time action evaluated on an arbitrary imaginary trajectory. For smooth finite-action Yang–Mills configurations on Euclidean four-space, assume the field approaches a pure gauge sufficiently rapidly at infinity. The resulting map on the three-sphere at infinity gives an integer topological sector under the usual regularity and bundle assumptions.

SE=14g2∫d4x FμνaFμνa,F~μνa=12ϵμνρσFρσa,Q=132π2∫d4x FμνaF~μνaS_E=\frac1{4g^2}\int d^4x\,F^a_{\mu\nu}F^a_{\mu\nu},\qquad \widetilde F^a_{\mu\nu}=\frac12\epsilon_{\mu\nu\rho\sigma}F^a_{\rho\sigma},\qquad Q=\frac1{32\pi^2}\int d^4x\,F^a_{\mu\nu}\widetilde F^a_{\mu\nu}
F has mass dimension two, so S E and Q are dimensionless in the chosen units. Integer Q relies on the stated finite-action asymptotics and regularity, not on an arbitrary field on a bounded region with unspecified boundary data.↔ Scroll wide equations horizontally.

Self-duality saturates the action bound

In four Euclidean dimensions the dual of a two-form can be applied twice to recover it. Completing the action into the square of F minus or plus its dual isolates the integral of F times its dual. The square is nonnegative, so its topological part gives a lower bound. Positive charge is saturated by a self-dual field and negative charge by an anti-self-dual field under the declared orientation. Such fields also satisfy the second-order equations through the Bianchi identity.

The path-integral sector weight contains exponential minus action and, with a theta angle, a phase proportional to Q. The sign of that phase is a convention fixed here by writing the weight explicitly. Different charge sectors can interfere in an amplitude or partition function; one must not square each factor independently and then call that the same observable. A tunneling splitting, a vacuum-energy theta dependence, and a metastable decay rate involve related methods but different boundary and contour problems.

SE=18g2∫d4x (Fμνa∓F~μνa)2±8π2Qg2≥8π2∣Q∣g2,weight=e−SE+iθQS_E=\frac1{8g^2}\int d^4x\,(F^a_{\mu\nu}\mp\widetilde F^a_{\mu\nu})^2\pm\frac{8\pi^2Q}{g^2}\ge\frac{8\pi^2|Q|}{g^2},\qquad \text{weight}=e^{-S_E+i\theta Q}
The upper square sign is chosen to match positive charge. The real action bound controls exponential suppression; theta contributes a phase with the explicitly chosen sign. Prefactors and sector measures are not included in this schematic weight.↔ Scroll wide equations horizontally.

A family of saddles needs a measure

A charge-one SU two instanton on four-space has four center coordinates, one size, and three gauge-orientation parameters in a framed convention where gauge transformations are fixed at infinity. Quotienting additional global gauge rotations changes the moduli description and its volume factors. Those directions generate zero modes rather than ordinary Gaussian fluctuations, so they must be replaced by collective-coordinate integrations with Jacobians. Counting parameters without stating the gauge quotient can give an apparent disagreement that is only conventional.

Quantum fluctuations renormalize the coupling and instanton measure, making size integration a physical issue rather than a harmless normalization. Small sizes probe short distances, while large sizes can leave the weakly coupled semiclassical regime. A formal dilute-instanton expression is not automatically a controlled description of uncompactified strongly coupled Yang–Mills theory. The saddle and its action are exact classical statements under their assumptions; dominance of that saddle and convergence of its collective-coordinate integral are additional questions.

Compare topological action sectors

Under the declared Euclidean normalization let g2=π2/10g^2=\pi^2/10. Find the charge-one self-dual action, the minimum action in charge two, and the ratio of their bare exponential suppressions if charge two also saturates its bound. Set theta = 0 and omit prefactors only for this comparison.

  1. For charge one, S1=8π2/(π2/10)=80S_1=8\pi^2/(\pi^2/10)=80. Its bare exponential weight is therefore e−80e^{-80}.

  2. The charge-two bound is twice as large: S2≥160S_2\ge160. A self-dual charge-two field attains 160 under the same boundary and normalization assumptions.

  3. The ratio of bare weights at saturation is e−160/e−80=e−80e^{-160}/e^{-80}=e^{-80}. This does not include the different moduli integrations, determinants, or interactions needed for a full sector contribution.

The result

The actions are 80 and at least 160; saturated charge-two bare suppression is e⁻⁸⁰ relative to charge one. These are action comparisons, not a computed physical transition rate.

Explore the model

Interactive lab

Interactive lab · quantum

A sector phase is not a probability change

Use the basis states only as abstract sector labels. Reset, set Mixing angle θ = 90° and Relative phase φ = 0°, and predict equal basis probabilities. For a fresh comparison Reset, set θ = 90° and φ = 90°, and predict unchanged probabilities but a rotated Bloch direction. Inspect the initial paused states before playback adds phase.

A sector phase is not a probability changeA normalized two-level pure state, not an interacting field simulation. Playback adds one radian per second to the relative phase; basis probabilities stay fixed. Current readings: Probability |0⟩: 71.1%; Probability |1⟩: 28.9%; Bloch (x, y, z): 0.78, 0.45, 0.42. All parameters are adjustable with the labeled controls below.xyz|0⟩|1⟩Basis probabilitiesP(0)71.1%P(1)28.9%
Bloch vector / P(0)P(1)Phase latitude
Probability |0⟩
71.1%
Probability |1⟩
28.9%
Bloch (x, y, z)
0.78, 0.45, 0.42
t = 0 s

A normalized two-level pure state, not an interacting field simulation. Playback adds one radian per second to the relative phase; basis probabilities stay fixed. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forThe lab is a chosen pure two-state geometry. It does not perform Wick rotation, evaluate an instanton action or moduli integral, or determine the theta dependence of a gauge theory.

Check your understanding

Why does the classical instanton action not by itself give a complete semiclassical contribution?

Try it yourself

Suppose charge +1 and −1 sectors have equal real prefactor C and the same real action S. With the stated weight convention, sum their contributions at theta = pi/2. Is this cancellation a statement that each sector’s probability or every tunneling rate vanishes?

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

Add the complex amplitudes before assigning an observable interpretation.

Compare with the solution

The sum is Ce−S(eiπ/2+e−iπ/2)=Ce−S(i−i)=0Ce^{-S}(e^{i\pi/2}+e^{-i\pi/2})=Ce^{-S}(i-i)=0. This is a cancellation of the specified equal-prefactor sector contributions to an amplitude or partition-function term. It does not say that each separate magnitude squared is zero, nor that every tunneling or decay observable has this structure. Different prefactors, sectors, boundary states, or rate definitions require a new calculation.

Take it with you

  • Euclidean signature, gauge normalization, and finite-action boundaries fix the instanton bound.
  • Collective coordinates replace symmetry zero modes and depend on a declared gauge quotient.
  • Classical action suppression, phase interference, and a controlled physical nonperturbative result are separate statements.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.