Monopole Operators and Quantum Coulomb Branches

A three-dimensional monopole operator is a local disorder insertion: it is defined by fixing magnetic flux through a small two-sphere linking the insertion. In a supersymmetric theory the corresponding flux sector can contain chiral operators that serve as quantum Coulomb-branch coordinates. Determining whether such an operator exists requires four separate steps—choose the magnetic lattice, impose the BPS boundary condition, quantize fermion zero modes, and satisfy the Chern–Simons Gauss law.

Required background. We use the supersymmetric gauge actions and FI convention, the parity-anomaly and contact-term shifts, and the general definition of disorder operators by singular boundary conditions. Helpful background. Zero-mode selection rules parallel those for four-dimensional instantons.

Magnetic boundary conditions and the GNO lattice

Let the compact connected gauge group be $G$ with maximal torus $T$. A monopole at the origin is labeled by a cocharacter

$$m\in\operatorname{Hom}(U(1),T)/W_G,$$

where $W_G$ is the Weyl group. On a small sphere, one can gauge-transform the leading field strength into the Cartan and write

$$F\sim\frac{m}{2}\sin\theta\,d\theta\wedge d\varphi, \qquad \frac{1}{2\pi}\int_{S^2}F=m.$$

For every electric weight $\rho$ of a genuine field, Dirac quantization requires $\rho(m)\in\mathbb Z$. This is why the global gauge group matters: groups with the same Lie algebra but different genuine Wilson lines have different cocharacter lattices and therefore different local monopole sectors.

Equivalently, $m$ is a weight of the GNO or Langlands dual group. Passing from a gauge group to a quotient changes this lattice even when the local Lagrangian is unchanged: a flux allowed for the quotient may fail to lift to the covering group, while gauging a discrete subgroup of topological symmetry projects the magnetic sum in the opposite direction. The transition-function condition $\exp(2\pi i m)=1_G$ is the reliable test; informal statements such as “$m$ is an element of the Cartan” do not fix the lattice Cremonesi, Hanany, and Zaffaroni 2014, §2, Eqs. (2.1)–(2.2).

For a disconnected gauge group, this connected-component cocharacter record is not the whole classification: nontrivial component-group bundles and twisted sectors must be added explicitly.

An $\mathcal N=2$ BPS insertion also turns on the vector scalar with a correlated singularity, schematically $\sigma\sim m/(2r)$, with sign tied to whether the operator is chiral or antichiral. The exact auxiliary-field boundary condition follows from setting the gaugino variation to zero. Flux alone defines a disorder sector; flux plus this supersymmetric completion selects a BPS operator.

Under radial quantization, the insertion becomes a state on $S^2$ with magnetic flux $m$. Fermion fields of gauge weight $\rho$ see $|\rho(m)|$ zero modes. Quantizing these modes fixes the quantum numbers of the bare monopole vacuum Borokhov, Kapustin, and Wu 2002a, §§2–3.

Zero-mode charges without double counting

Let $V_m$ denote the bare monopole. Choose Cartan gauge fields $a^a$ and a background field $A_F$ for an ordinary Abelian symmetry $F$. In the symmetric zero-mode vacuum convention, and at a point where no extra fermion becomes massless, its global charge is

$$Q_F(V_m) =+k^{\rm bare}_{Fa}m_a -\frac12\sum_{\psi}Q_F(\psi) \sum_{\rho\in R_\psi}|\rho(m)|,$$

where the first sum runs over complex two-component fermions, including gauginos, and repeated Cartan indices are summed. The first term is the charge induced by a bare mixed Chern–Simons coupling $A_Fda^a/(2\pi)$. The absolute value in the second term counts zero modes; it is not the sign of an ordinary real mass.

Gauge charge is computed by the same normal ordering, replacing $Q_F(\psi)$ by the Cartan weight $\rho_a$:

$$q_a(V_m)= +k^{\rm bare}_{ab}m_b -\frac12\sum_{\psi}\sum_{\rho\in R_\psi} \rho_a\,|\rho(m)|.$$

There are two equivalent bookkeeping descriptions. One may quantize the zero modes and add the bare Chern–Simons charges as above, or, in a Coulomb chamber where the relevant fermions are massive, integrate them out and use the chamber-dependent $k^{\rm eff}$. Adding the zero-mode normal-ordering charge to a level that already contains the same fermion determinant counts that fermion twice. A wall where $\rho(\sigma)+m_F=0$ must instead retain the light field explicitly.

If chiral multiplet $i$ has scalar trial $R$-charge $r_i$, its fermion has charge $r_i-1$, while the gaugino has $R=1$. Hence

$$R(V_m)= \frac12\sum_i(1-r_i) \sum_{\rho_i\in R_i}|\rho_i(m)| -\frac12\sum_{\alpha\in\mathrm{roots}(G)}|\alpha(m)|.$$

At a superconformal fixed point, a gauge-invariant chiral monopole obeys $\Delta=R$ for the exact infrared $R$ symmetry. The formula becomes a scaling dimension only after $R$-mixing and accidental symmetries are resolved. If a candidate hits the unitarity bound and becomes free, extremization must be redone with that decoupled sector treated correctly.

This also identifies the semiclassical boundary of the formula. The charge calculation is protected, but the exponential photon coordinate and the interpretation of a smooth branch require a region with $|\sigma|$ large compared with the strong scale. Near the origin, $\Delta=R$ can remain exact even though the semiclassical metric and the elementary description of the operator fail.

For a $U(N)$ factor, the topological charge is

$$Q_J(V_m)=\operatorname{Tr}m$$

in the fundamental trace convention. A monopole superpotential can break $U(1)_J$ to a subgroup or entirely.

Chern–Simons Gauss law and dressing

Magnetic flux in a Chern–Simons theory carries electric charge. With $D=\partial-iq a$, positive flux $m=\frac{1}{2\pi}\int F$, and the term $+\frac{k}{4\pi}\int a\wedge da$, our convention gives

$$Q_{\rm gauge}^{\rm classical}(V_m)=+km.$$

Mixed gauge–background levels analogously give global charges with the same sign. This convention matches the level-one bosonization fixture, where a positive-flux monopole has gauge charge $+1$ and is dressed by a charge-$-1$ scalar Seiberg et al. 2016, §2.1, Eq. (2.2). One-loop shifts from matter and massive W-bosons must be evaluated in the same Coulomb chamber as the monopole. In a non-Abelian background, $m$ breaks $G$ to its stabilizer $G_m$; gauge neutrality is required under every unbroken factor.

A bare $V_m$ that is electrically charged is not a gauge-invariant local operator. More generally, in a non-Abelian flux it can transform in a representation of the stabilizer $G_m$. It may be dressed by matter fields and residual-vector scalars so that the tensor product contains a $G_m$ singlet and every Abelian gauge charge cancels. The dressing must also survive F-term relations and the zero-mode quantization. If no such dressing exists, the magnetic sector can exist in the Hilbert space without providing the proposed chiral-ring generator.

This observation explains the structural difference between level-zero Aharony duality, where minimal monopoles can appear as singlets, and nonzero-level Giveon–Kutasov duality, where the Chern–Simons Gauss law obstructs those same bare coordinates.

Example: minimal monopoles in U(Nc) SQCD

For the standard global group $U(N)=(SU(N)\times U(1))/\mathbb Z_N$ with a genuine fundamental representation, a cocharacter can be diagonalized as $m=\operatorname{diag}(m_1,\ldots,m_N)$ with $m_i\in\mathbb Z$, modulo permutations. This concrete lattice statement is part of the example; a different quotient or line spectrum must be recomputed.

Consider $U(N_c)_0$ with $N_c\ge1$ and $N_f$ pairs $(Q_i,\widetilde Q^i)$ in the fundamental and antifundamental, no superpotential, common trial scalar $R$-charge $r$, and axial charge $+1$ for both. Choose the two minimal cocharacters

$$m_+=(1,0,\ldots,0), \qquad m_-=(-1,0,\ldots,0).$$

For either flux, every fundamental or antifundamental supplies one unit to the absolute-weight sum. The $2N_f$ matter fermions therefore give axial charge $-N_f$ and matter contribution $N_f(1-r)$ to $R$. The roots connecting the fluxed color to the remaining $N_c-1$ colors give the vector contribution $-(N_c-1)$. Thus

$$\begin{array}{c|ccc} &U(1)_J&U(1)_A&U(1)_R\\ \hline V_+&+1&-N_f&N_f(1-r)-(N_c-1)\\ V_-&-1&-N_f&N_f(1-r)-(N_c-1) \end{array}$$

These quantum numbers are exactly what the singlet monopoles on the magnetic side of Aharony duality must reproduce. They also show a limitation: the ultraviolet trial value $r$ cannot simply be inserted to assert an infrared dimension. Mixing with $U(1)_A$, superpotential constraints, and possible free monopoles decide the exact value.

For $N_c=1$, the root term vanishes. With one flavor pair, $R(V_\pm)=1-r$ and $Q_A=-1$. This is the basic $\mathcal N=2$ SQED monopole calculation used in Abelian mirror symmetry Borokhov, Kapustin, and Wu 2002b, §§3–5.

Exact N=4 SQED benchmark

An $\mathcal N=4$ $U(1)$ theory with $N\ge1$ charge-one hypermultiplets has two $\mathcal N=2$ chirals per hyper, each with scalar $R=1/2$. For flux $m$, the formula above gives

$$\Delta(V_m)=\frac{N}{2}|m|.$$

Let $\varphi$ be the complex scalar in the $\mathcal N=4$ vector multiplet. In one complex structure, the full quantum Coulomb-branch ring is

$$\mathbb C[\mathcal M_C] =\frac{\mathbb C[\varphi,V_+,V_-]} {(V_+V_- - \varphi^N)}.$$

The relation passes both grading checks: the two sides have topological charge zero and dimension $N$. It exhibits something that the dimension formula alone cannot supply—the operator product that glues positive- and negative-flux patches. This exact benchmark becomes the simplest coordinate-ring square in $\mathcal N=4$ mirror symmetry Cremonesi, Hanany, and Zaffaroni 2014, §3.1, Eqs. (3.2)–(3.3).

From semiclassical photons to a quantum Coulomb branch

At a generic point where $G$ is broken to its maximal torus, dualizing an Abelian photon gives a periodic scalar $\gamma$. Semiclassically, exponential combinations of $\sigma$ and $\gamma$, of the form

$$Y\sim\exp\!\left(\frac{2\pi\sigma}{e^2}+i\gamma\right),$$

provide local coordinates in an appropriate chamber. This expression is only heuristic near strongly coupled loci: the true holomorphic coordinates are monopole operators, and different Weyl chambers can require different patches.

Quantum effects can alter the classical picture in several ways.

• Fermion zero modes determine whether a fundamental monopole can generate a superpotential. Precisely two unlifted gaugino zero modes are the standard $\mathcal N=2$ superpotential measure; additional matter zero modes forbid the term unless interactions absorb them.
• A generated monopole superpotential lifts directions and can impose chiral-ring relations.
• Chern–Simons couplings make the dual photon massive and charge the flux operator, often removing an undressed Coulomb coordinate.
• At singular loci, additional monopoles or matter fields become massless, so the semiclassical Abelian description fails.

Accordingly, “the Coulomb branch” can mean a semiclassical vacuum region, a quantum moduli space, or the subring generated by monopole chirals. These coincide only after the lifting, dressing, and patching questions have been answered Aharony et al. 1997, §§2–4.

Independent checks

A proposed monopole dictionary should pass at least three distinct tests.

1. Magnetic lattice: the flux must be allowed by the global gauge group and genuine line spectrum.
2. Quantum numbers: zero-mode and Chern–Simons calculations must match all gauge, flavor, topological, and exact $R$ charges.
3. Protected observables: the same operator should appear with the predicted fugacities in a supersymmetric index, Hilbert series, or localized partition function when that observable is defined.

A match of trial $R$-charge alone is weak: distinct magnetic sectors can share it, and an electrically charged bare monopole is not a local gauge invariant.

The chapter’s monopole–contact–duality map places these operator checks beside the regulator and infrared-duality data they constrain.

Common pitfalls

Using the Lie algebra as the magnetic lattice. A magnetic charge is a cocharacter of the global group. Quotients change allowed monopoles even when all perturbative fields transform under the same algebra.

Forgetting the gaugino roots. The second term in $R(V_m)$ is essential for non-Abelian theories. Omitting it changes the rank dependence and spoils Aharony dictionaries.

Calling every flux insertion a chiral operator. The BPS completion, gauge neutrality, zero-mode vacuum, dressing, and chiral-ring relations must all be checked.

Exercises

1. Reproduce the $R$-charge of $V_+$ in $U(N_c)$ SQCD by counting the fundamental, antifundamental, and root weights with nonzero pairing with $m_+=(1,0,\ldots,0)$.

Solution

Each of the $N_f$ fundamentals and $N_f$ antifundamentals has one weight with absolute pairing one, so matter contributes $\frac12(2N_f)(1-r)=N_f(1-r)$. There are $2(N_c-1)$ roots $e_1-e_j$ and $e_j-e_1$ with absolute pairing one, so the vector contribution is $-\frac12\,2(N_c-1)=-(N_c-1)$. Their sum is $N_f(1-r)-(N_c-1)$.

2. In compact $U(1)_k$ with a scalar field of charge $+1$, a bare monopole of flux $m=1$ has classical electric charge $+k$. What polynomial dressing is required when $k\ge0$, ignoring other quantum corrections?

Solution

Multiplying by $(\phi^\dagger)^k$ adds electric charge $-k$, so $V_1(\phi^\dagger)^k$ is classically neutral. Whether it is an actual chiral operator still depends on fermion zero modes, real-mass chambers, F-term relations, and the availability of a scalar zero mode in that monopole background.

3. In $\mathcal N=4$ $U(1)$ SQED with $N$ hypermultiplets, check that $V_+V_-=\varphi^N$ is homogeneous under scaling and $U(1)_J$. What happens for $N=1$?

Solution

$V_+$ and $V_-$ have dimensions $N/2$ and topological charges $+1$ and $-1$, so their product has dimension $N$ and charge zero. The vector scalar $\varphi$ has dimension one and no topological charge, so $\varphi^N$ has the same gradings. For $N=1$, the relation eliminates $\varphi=V_+V_-$ and leaves a polynomial ring $\mathbb C[V_+,V_-]$: the Coulomb branch is $\mathbb C^2$, as required for the free twisted-hypermultiplet mirror.

References

• Aharony, O., Hanany, A., Intriligator, K., Seiberg, N., and Strassler, M. J. (1997), “Aspects of $N=2$ Supersymmetric Gauge Theories in Three Dimensions,” Nuclear Physics B 499, 67–99. doi:10.1016/S0550-3213(97)00323-4. Open PDF
• Borokhov, V., Kapustin, A., and Wu, X. (2002a), “Topological Disorder Operators in Three-Dimensional Conformal Field Theory,” Journal of High Energy Physics 2002(11), 049. doi:10.1088/1126-6708/2002/11/049. Open PDF
• Borokhov, V., Kapustin, A., and Wu, X. (2002b), “Monopole Operators and Mirror Symmetry in Three Dimensions,” Journal of High Energy Physics 2002(12), 044. doi:10.1088/1126-6708/2002/12/044. Open PDF
• Cremonesi, S., Hanany, A., and Zaffaroni, A. (2014), “Monopole Operators and Hilbert Series of Coulomb Branches of 3d $\mathcal N=4$ Gauge Theories,” Journal of High Energy Physics 2014(01), 005. doi:10.1007/JHEP01(2014)005. Open PDF
• Seiberg, N., Senthil, T., Wang, C., and Witten, E. (2016), “A Duality Web in 2+1 Dimensions and Condensed Matter Physics,” Annals of Physics 374, 395–433. doi:10.1016/j.aop.2016.08.007. Open PDF

Next steps

Mirror symmetry exchanges these quantum Coulomb coordinates with classical Higgs operators in $\mathcal N=4$ theories. In four-supercharge theories they enter the singlet and superpotential dictionaries of Aharony duality.