Small-Circle Abelianization, Bions, and Adiabatic Continuity

On $\mathbb R^3\times S^1_L$, a center-symmetric holonomy can separate non-Abelian and Abelian scales so that monopole-instantons and correlated bions become calculable. Which event generates the bosonic dual-photon potential depends on the matter zero modes: fundamental monopoles do so in deformed Yang–Mills, whereas magnetic bions lead in massless QCD(adj). Neutral bions have a different role. Extending either result to large $L$ requires a separate continuity argument.

Required background. Fractional events and caloron constituents supplies the holonomy-resolved monopoles and their fractional topological charges. Compact U(1) in 2+1 dimensions supplies the dual-photon and kink mechanism.

Helpful background. Multi-saddle sums and dilute ensembles supplies the correlated-event expansion and its quasi-zero-mode checks.

Shared comparison. The claim–evidence comparison marks the small-circle result as a controlled mechanism under explicit hypotheses.

Center symmetry creates a calculable hierarchy

Take $SU(N)$ on a spatial circle with periodic boundary conditions for adjoint fermions when present. Let

$$\Omega=\mathrm P\exp\!\left(i\oint_{S^1}A_4\,\mathrm d x^4\right)$$

have center-symmetric eigenvalues. Then the adjoint Higgs effect breaks

$$SU(N)\longrightarrow U(1)^{N-1}, \qquad m_W=\frac{2\pi}{NL},$$

up to convention-dependent factors for non-nearest roots. At fixed $N$, $L\Lambda\ll1$ makes $g^2(m_W)$ small; more uniformly in $N$, the useful separation is $NL\Lambda\ll1$. A center-breaking holonomy destroys this Abelian regime, so “small circle” alone is not enough.

There are $N$ fundamental monopole-instantons, associated with the simple and affine roots $\alpha_i$. At center symmetry each has

$$Q_i=\frac1N, \qquad S_0=\frac{8\pi^2}{g^2(m_W)N},$$

and the product of one event of every type has unit instanton charge. These data, including the dilute monopole ensemble and the hierarchy $m_\gamma\ll m_W$, are derived for center-stabilized Yang–Mills in Ünsal and Yaffe 2008, §§2–3, especially Eqs. (3.11)–(3.23).

Controlled compactification mechanism chain

The figure separates three superficially similar event chains. Inspect the event inventory and zero-mode test before following an arrow to a bosonic potential.

Controlled confinement mechanisms in a compact Abelian theory and in center-symmetric small-circle theories. The chains are schematic but their hypotheses and event orders are explicit; the final continuity test is not part of the semiclassical derivation and does not prove undeformed four-dimensional Yang–Mills confinement.

Deformed Yang–Mills: monopoles contribute directly

In center-stabilized pure Yang–Mills, no fermion zero modes obstruct the fundamental monopole vertices. On one chosen semiclassical branch, the leading dual-photon potential has the affine-Toda form

$$V_{\mathrm{dYM}}(\boldsymbol\sigma) =-2\zeta\sum_{i=1}^{N} \cos\!\left(\alpha_i\mathbin{\cdot}\boldsymbol\sigma+\frac{\theta}{N}\right), \qquad \zeta\sim e^{-S_0}.$$

Away from special destructive-interference points, its Hessian gives masses to the $N-1$ dual photons, and a Wilson loop imposes a weight-lattice discontinuity whose minimum-action interpolation has nonzero string tension; at a special theta value, subleading composite events may be required for a lifted mode. The visible $\theta/N$ does not give the microscopic theory a $2\pi N$ period: shifting $\theta$ by $2\pi$ relabels the semiclassical branches, and the full branch sum or minimum is $2\pi$-periodic. The mechanism is the locally four-dimensional analogue of the compact-$U(1)$ monopole plasma, but it relies on the double-trace deformation maintaining center symmetry Ünsal and Yaffe 2008, §§2.1 and 3.1–3.3.

Massless QCD(adj): zero modes force a composite event

Now include $n_f$ massless adjoint Weyl fermions with periodic spatial boundary conditions. The index theorem gives $2n_f$ fermion zero modes on each fundamental monopole. Schematically,

$$\mathcal M_i \sim e^{-S_0}e^{i\alpha_i\cdot\boldsymbol\sigma} (\alpha_i\mathbin{\cdot}\psi)^{2n_f}.$$

An isolated monopole therefore cannot appear as a purely bosonic potential term. The leading magnetically charged bosonic object is a magnetic bion, a correlated monopole–anti-monopole pair of different types,

$$\mathcal B_i=\mathcal M_i\overline{\mathcal M}_{i+1}, \qquad Q_{\mathcal B}=0, \qquad \mathbf m_{\mathcal B}=\alpha_i-\alpha_{i+1}.$$

Fermion exchange can bind the magnetically repelling constituents. The resulting operators occur at order $e^{-2S_0}$ and generate

$$V_{\mathrm{bion}}(\boldsymbol\sigma) \sim-e^{-2S_0}\sum_i \cos[(\alpha_i-\alpha_{i+1})\mathbin{\cdot}\boldsymbol\sigma],$$

producing a dual-photon mass and spatial string tension. The microscopic binding and bion-plasma mapping are worked through in Anber and Poppitz 2011, §§4.1–4.4.

A neutral bion $\mathcal M_i\overline{\mathcal M}_i$ instead has zero magnetic and topological charge. It affects the holonomy potential and, after a properly defined quasi-zero-mode integral, participates in semiclassical ambiguity cancellation. Because it carries no magnetic charge, it does not directly supply the dual-photon cosine responsible for confinement. Calling both composites simply “bions” erases the physical distinction.

Poppitz gives a convention-complete comparison of deformed Yang–Mills, QCD(adj), and supersymmetric Yang–Mills on the circle in Poppitz 2022, §§3–6.

Continuity is a test, not an arrow in the derivation

The semiclassical calculation is controlled while

$$m_\gamma\ll m_W, \qquad e^{-S_0}\ll1, \qquad \text{center symmetry is intact}.$$

As $L$ grows toward $\Lambda^{-1}$, the event gas ceases to be parametrically dilute and the Abelian/non-Abelian scale separation closes. To claim adiabatic continuity to a large-circle regime one must separately check:

1. the same exact symmetries and anomalies on both sides;
2. no intervening center, chiral, or other phase transition;
3. no level crossing or new massless sector that changes the infrared realization;
4. which deformation or fermion mass is held fixed and whether it is eventually removed.

For deformed Yang–Mills, absence of an order parameter separating small and large $L$ within the deformed theory motivates continuity, and large-$N$ volume independence adds a distinct statement when center symmetry remains unbroken. Neither statement removes the deformation at finite $N$ or constitutes a proof for ordinary $\mathbb R^4$ Yang–Mills. For QCD(adj), the center may remain intact while chiral symmetry changes, so center symmetry alone cannot establish continuity.

The bounded conclusion is strong but specific: small-$L$ center-symmetric theories provide analytic continuum examples of a mass gap and linear confinement generated by identified topological events. The extrapolation beyond the controlled hierarchy is a hypothesis to be tested, not another term in the semiclassical series.

Exercises

1. Identify the leading event. For $n_f=2$ massless adjoint Weyl fermions, why can a fundamental monopole not generate a dual-photon mass, and at what exponential order does the magnetic-bion potential appear?

Solution

Each monopole has $2n_f=4$ fermion zero modes, so its vertex contains four fermion fields and is not a bosonic potential. A correlated magnetic bion can saturate the exchanged fermions internally and has action approximately $2S_0$, so its bosonic potential appears at order $e^{-2S_0}$.

2. Separate the bions. State one observable role of a magnetic bion and one of a neutral bion.

Solution

A magnetic bion carries net magnetic charge and generates a periodic dual-photon potential, hence a dual-photon mass and spatial string tension. A neutral bion carries no magnetic charge; it contributes to the holonomy potential and to the cancellation structure of semiclassical ambiguities, but not directly to the confining dual-photon cosine.

References

• Anber, Mohamed M., and Erich Poppitz. “Microscopic Structure of Magnetic Bions.” Journal of High Energy Physics 2011, no. 6 (2011): 136. DOI. Open PDF.
• Poppitz, Erich. “Notes on Confinement on $\mathbb R^3\times S^1$: From Yang–Mills, Super-Yang–Mills, and QCD(adj) to QCD(F).” Symmetry 14 (2022): 180. DOI. Open PDF.
• Ünsal, Mithat, and Laurence G. Yaffe. “Center-Stabilized Yang–Mills Theory: Confinement and Large-$N$ Volume Independence.” Physical Review D 78 (2008): 065035. DOI. Open PDF.