Compactification, Semiclassical Continuity, and Order of Limits

Compactifying pure $\mathcal N=1$ super-Yang–Mills on a small spatial circle can turn its strong dynamics into a controlled dilute gas of monopole-instanton constituents. The calculation reproduces the discrete vacua and gaugino condensate and gives a semiclassical abelian mass gap in that small-circle regime. Only the protected pieces can be transported to four-dimensional strong coupling without an extra continuity assumption.

Required background. Instantons, zero modes, and condensates supplies the Euclidean index calculation, and pure SYM vacua and domain walls supplies the four-dimensional vacuum structure. Helpful background. Fractional events and caloron constituents develops the topology of the monopole constituents.

The compactified theory and its control parameter

Take simply connected $SU(N)$ pure SYM on

$$\mathbb R^3\times S_L^1$$

with circumference $L$ and periodic boundary conditions for the gaugino. This is not a thermal partition function: the periodic spin structure preserves four supercharges, which appear as three-dimensional $\mathcal N=2$ supersymmetry. The three-dimensional coupling is

$$g_3^2=\frac{g_4^2(1/L)}{L}.$$

For periodic boundary conditions, perturbative bosonic and gaugino contributions to the holonomy potential cancel. The nonperturbative monopole superpotential then selects supersymmetric center-symmetric vacua in which the Wilson-line eigenvalues are evenly spaced and

$$SU(N)\longrightarrow U(1)^{N-1}.$$

The lightest off-diagonal vector multiplet has mass

$$m_W=\frac{2\pi}{NL}.$$

Abelian semiclassics requires $m_W\gg\Lambda$, or parametrically

$$\boxed{NL\Lambda\ll1.}$$

The factor $N$ is physical. Merely requiring $L\Lambda\ll1$ is insufficient at large rank because adjacent holonomy eigenvalues approach one another.

In Euclidean signature, $A_4$ becomes a compact adjoint scalar and each three-dimensional photon may be dualized to a periodic scalar $\sigma$. They combine into complex Coulomb-branch coordinates. The barred fermion is an independent integration variable during the saddle calculation; Lorentzian Majorana reality is restored only after analytic continuation.

Fractional monopole events

At center-symmetric holonomy there are $N-1$ monopoles associated with the simple roots and one Kaluza–Klein monopole associated with the affine root. In the displayed classical action take $g_4\equiv g_4(1/L)$; running down to the constituent scale $m_W$ is combined with the fluctuation determinant and the prefactor below. Each event has

$$S_0=\frac{8\pi^2}{g_4^2N}, \qquad Q_{\mathrm{top}}=\frac1N,$$

and exactly two adjoint-gaugino zero modes. One event can therefore contribute to the holomorphic superpotential. A collection containing one monopole of every type has action $8\pi^2/g_4^2$, topological charge one, and $2N$ zero modes: it reconstructs the four-dimensional instanton.

Let $Y_i$ denote the $i$th monopole operator, including its holonomy and dual-photon exponential. Their product is fixed by the four-dimensional instanton factor,

$$\prod_{i=1}^{N}Y_i=\eta, \qquad \eta=e^{-8\pi^2/g_4^2+i\theta},$$

up to the declared renormalization convention. The monopoles generate the affine-Toda superpotential

$$W_{\mathbb R^3\times S^1}=\kappa(L,g_4)\sum_{i=1}^{N}Y_i,$$

where the common prefactor depends on normalization but the root structure does not. This controlled construction and its zero-mode measure were derived in Davies, Hollowood, Khoze, and Mattis 1999, §§ 3–5, pp. 128–139.

General simple groups: co-roots and co-marks

The equality “number of monopole species $=$ number of vacua” is special to $SU(N)$. For a simple rank-$r$ group, the fundamental events correspond to the $r$ simple co-roots and one affine co-root. Define the dual Kac labels, or co-marks, by

$$\alpha_0^\vee=-\sum_{i=1}^{r}k_i^\vee\alpha_i^\vee, \qquad k_0^\vee=1, \qquad \sum_{i=0}^{r}k_i^\vee=h^\vee.$$

There are only $r+1$ monopole species $\mathcal M_i$, but a unit four-dimensional instanton contains them with multiplicities

$$\mathcal I_{4d}\sim \prod_{i=0}^{r}\mathcal M_i^{\,k_i^\vee}.$$

The distinction is visible in a two-row comparison. The co-marks are shown as an unordered multiset, so no Dynkin-node numbering convention is hidden.

Simply connected group	Fundamental species	Co-mark multiset	$h^\vee$	Unit-instanton content
$SU(N)$	$N$	$N$ entries equal to $1$	$N$	one event of every species
$G_2$	$3$	$\{1,1,2\}$	$4$	one of the three species occurs twice

With a Coulomb-branch coordinate $X$ defined so that a fundamental operator contains $e^{\alpha_i^\vee\cdot X}$, the general superpotential has the structural form

$$\begin{aligned} W={}&\kappa\sum_{i=1}^{r} \frac{2}{\alpha_i^2}e^{\alpha_i^\vee\cdot X}\\ &+\kappa\frac{2}{\alpha_0^2} e^{2\pi i\tau+\alpha_0^\vee\cdot X}. \end{aligned}$$

The root-length factors matter for non-simply-laced groups, whose construction is associated with the appropriate twisted affine algebra. At the supersymmetric holonomy each fundamental event has topological charge $1/h^\vee$; the co-mark multiplicities then give total charge one and $2h^\vee$ gaugino zero modes. The stationary points number $h^\vee$, although there are only $r+1$ species. These statements and the group tables are derived in Davies, Hollowood, and Khoze 2003, §§ 3–5, especially Eqs. (4.16), (5.3), and (5.7)–(5.14), and Appendix A, arXiv PDF.

Return to SU(N): N vacua from one constrained extremization

Extremize $\sum_iY_i$ subject to $\prod_iY_i=\eta$. A Lagrange multiplier gives

$$Y_1=Y_2=\cdots=Y_N,$$

and hence

$$\boxed{ Y_i=\eta^{1/N}e^{2\pi i k/N}, \qquad k=0,1,\ldots,N-1.}$$

These are the $N$ vacua required by $\mathbb Z_{2N}\to\mathbb Z_2$ chiral-symmetry breaking. Differentiating the holomorphic vacuum functional with respect to the gauge coupling produces

$$\langle\operatorname{Tr}\lambda\lambda\rangle_k \propto\left(\Lambda^{3N}\right)^{1/N}_k =\lvert\Lambda\rvert^3e^{i(\theta+2\pi k)/N},$$

where the last expression declares $\Lambda^{3N}=\lvert\Lambda\rvert^{3N}e^{i\theta}$. The proportionality is fixed only after the trace and scale conventions are fixed. The small-circle result agrees with the weak-coupling determination of the four-dimensional condensate; see Davies, Hollowood, and Khoze 2003, §§ 4–5, especially Eqs. (4.16), (5.3), and (5.7)–(5.14).

The scalar potential obtained from this superpotential and the Coulomb-branch Kähler metric gives masses to the holonomy fluctuations and dual photons. In the controlled regime this produces an abelian mass gap and confinement of electric probes charged under the low-energy photons. Correlated monopole–antimonopole events are visible in the component potential generated by the monopole superpotential. Numerical values of masses and string tensions are nonholomorphic and depend on the Kähler normalization.

What continuity does and does not transport

Periodic compactification preserves supersymmetry and leaves the discrete chiral and center transformations as symmetries of the action. Their realization is dynamical: in the small-circle supersymmetric vacua the center is unbroken. Provided no phase transition occurs and no vacuum escapes to infinity as $L$ changes, the supersymmetric index, number of vacua, holomorphic superpotential, and condensate can be continued from $NL\Lambda\ll1$ toward $L\Lambda\gg1$. This is the controlled route around the ambiguous strong-coupling four-dimensional instanton calculation.

The same argument does not compute the large-$L$ mass gap, string tension, Kähler metric, or the microscopic mechanism of confinement. Abelianization disappears once $m_W$ is comparable to $\Lambda$; the four-dimensional theory may remain in the same phase while its useful quasiparticles change. “No phase transition” is weaker than “the same semiclassical mechanism remains dilute.”

Lattice simulations of $SU(2)$ SYM with one adjoint Majorana fermion and periodic boundary conditions find center stability and no intervening transition in the explored light-fermion regime Bergner, Piemonte, and Ünsal 2018, §§ 4–5. This is numerical evidence, not a proof beyond $SU(2)$, outside the explored lattice masses and radii, or after an uncontrolled lattice-to-continuum extrapolation. The analytic mass–circumference phase diagram of Poppitz, Schäfer, and Ünsal 2012, §§ 2–4 is likewise an $SU(2)$ analysis.

For a regime-by-regime separation of exact statements, controlled calculations, continuity assumptions, and numerical evidence, consult SQCD regimes: exact results and evidence limits.

Four limits that do not commute

Spatial versus thermal circle

Antiperiodic fermions define a thermal ensemble, explicitly break supersymmetry, and change the holonomy potential. A thermal center transition is compatible with smooth periodic-circle behavior. Results from one spin structure cannot be cited as evidence for the other without a separate continuation.

Large N versus small L

The abelian window is $NL\Lambda\ll1$. At fixed $L\Lambda$, taking $N\to\infty$ first eventually violates it and makes the $W$ bosons light. Large-$N$ volume independence, when applicable, is a different mechanism and must not be conflated with the dilute abelian expansion.

Finite small circle versus the strict three-dimensional limit

At finite $L$, the Kaluza–Klein monopole and the four-dimensional instanton factor $\eta$ remain present. A strict three-dimensional limit instead takes $L\to0$ while holding $g_3^2=g_4^2/L$ fixed. Then $g_4^2\to0$, so

$$\eta=e^{-8\pi^2/g_4^2+i\theta}\longrightarrow0.$$

The affine monopole term disappears. The remaining non-affine Toda superpotential has no stationary point, and the Coulomb-branch coordinate runs away. This strict limit is therefore not the small but finite spatial-circle regime used to recover the four-dimensional vacua Davies, Hollowood, and Khoze 2003, § 4, especially Eq. (4.18), arXiv PDF.

Decompactification versus semiclassical expansion

Taking $L\to\infty$ at fixed $N$ eliminates the small parameter. Protected holomorphic answers may remain constant, but term-by-term monopole-gas control does not survive that limit. One should continue the answer, not the validity of the dilute expansion.

The gauge-group global form supplies another discrete choice. Replacing $SU(N)$ by $SU(N)/\mathbb Z_p$ changes genuine line operators, allowed magnetic sectors, and discrete theta angles. Local Lie-algebra formulas for $S_0$ do not determine those global identifications or the vacuum counting after a one-form symmetry is gauged.

Common pitfalls

Calling the periodic circle finite temperature. Thermal fermions are antiperiodic and break supersymmetry. The periodic calculation is a spatial compactification.

Transporting an unprotected number. The condensate and vacuum count are protected; a string tension or glueball mass is not. Their small-circle values are not four-dimensional predictions.

Taking large N inside the abelian formula. Since $m_W=2\pi/(NL)$, the scale separation collapses at fixed $L$ as $N$ grows.

Exercises

1. Reassemble an instanton

Show that one monopole event of each of the $N$ types has the action, topological charge, and gaugino zero modes of a four-dimensional instanton.

Solution

The actions add to $NS_0=8\pi^2/g_4^2$, the charges add to $N(1/N)=1$, and the two zero modes per monopole add to $2N$, the adjoint index for a unit $SU(N)$ instanton.

2. Find the vacua

Extremize $W=\kappa\sum_iY_i$ with $\prod_iY_i=\eta$ and explain the origin of the branch label.

Solution

For $\mathcal W=\kappa\sum_iY_i+u(\log\eta-\sum_i\log Y_i)$, the $Y_i$ equation gives $\kappa Y_i=u$, so all $Y_i$ are equal. Their common value is an $N$th root of $\eta$, producing $Y_i=\eta^{1/N}e^{2\pi ik/N}$ with $k\in\mathbb Z_N$. The branches are permuted by a $2\pi$ shift of $\theta$ and correspond to the broken discrete chiral symmetry.

3. Distinguish species from constituents

Use the $G_2$ row of the co-mark table to determine the number of fundamental monopole species, the number of constituent events in a unit instanton, its total topological charge, and its total number of gaugino zero modes.

Solution

$G_2$ has rank two, so there are three species. The co-marks $\{1,1,2\}$ require four constituent events in total. At the supersymmetric holonomy each carries charge $1/h^\vee=1/4$ and two gaugino zero modes, giving total charge one and eight zero modes. This equals $2h^\vee$ because $h^\vee(G_2)=4$.

References

• Bergner, Georg, Stefano Piemonte, and Mithat Ünsal. “Adiabatic Continuity and Confinement in Supersymmetric Yang–Mills Theory on the Lattice.” Journal of High Energy Physics 11 (2018): 092. DOI; arXiv.
• Davies, N. Michael, Timothy J. Hollowood, Valentin V. Khoze, and Michael P. Mattis. “Gluino Condensate and Magnetic Monopoles in Supersymmetric Gluodynamics.” Nuclear Physics B 559 (1999): 123–142. DOI; arXiv.
• Davies, N. Michael, Timothy J. Hollowood, and Valentin V. Khoze. “Monopoles, Affine Algebras and the Gluino Condensate.” Journal of Mathematical Physics 44 (2003): 3640–3656. DOI; arXiv.
• Poppitz, Erich, Thomas Schäfer, and Mithat Ünsal. “Continuity, Deconfinement, and (Super) Yang–Mills Theory.” Journal of High Energy Physics 10 (2012): 115. DOI; arXiv.