Magnetic bion condensation: A new mechanism of confinement and mass gap in four dimensions

TL;DR

The paper presents a microscopic mechanism for confinement in four-dimensional QCD-like theories by identifying magnetic bions—topologically neutral, magnetically charged bound states—as the drivers of a mass gap and area-law confinement in SU(N) QCD(adj) on S^1 × R^3. By leveraging abelian duality, index theorems, and a careful treatment of monopole composites, the authors derive a 3D dual Lagrangian where bions generate a bosonic potential for the dual photon, yielding confinement and a calculable mass gap. The analysis extends to SU(N) and reveals a hidden integrable structure via a prepotential, with sine-law scaling for string tensions and a detailed portrait of domain walls and chiral symmetry realizations; in contrast, noncompact adjoint-Higgs setups and N=2 theories lack confinement due to the absence of magnetically neutral bound states. The work offers a controllable, analytic handle on confinement in vector-like gauge theories, suggests potential integrability behind QCD(adj), and points to lattice tests and extensions toward QCD on R^4 and beyond.

Abstract

In recent work, we derived the long-distance confining dynamics of certain QCD-like gauge theories formulated on small $S^1 \times \R^3$ based on symmetries, an index theorem, and Abelian duality. Here, we give the microscopic derivation. The solution reveals a new mechanism of confinement in QCD(adj) in the regime where we have control over both perturbative and nonperturbative aspects. In particular, consider SU(2) QCD(adj) theory with $1 \leq n_f \leq 4$ Majorana fermions, a theory which undergoes gauge symmetry breaking at small $S^1$. If the magnetic charge of the BPS monopole is normalized to unity, we show that confinement occurs due to condensation of objects with magnetic charge 2, not 1. Because of index theorems, we know that such an object cannot be a two identical monopole configuration. Its net topological charge must vanish, and hence it must be topologically indistinguishable from the perturbative vacuum. We construct such non-self-dual topological excitations, the magnetically charged, topologically null molecules of a BPS monopole and ${\bar{\rm KK}}$ antimonopole, which we refer to as magnetic bions. An immediate puzzle with this proposal is the apparent Coulomb repulsion between the BPS-${\bar{\rm KK}}$ pair. An attraction which overcomes the Coulomb repulsion between the two is induced by $2n_f$-fermion exchange. Bion condensation is also the mechanism of confinement in $\N=1$ SYM on the same four-manifold. The SU(N) generalization hints a possible hidden integrability behind nonsupersymmetric QCD of affine Toda type, and allows us to analytically compute the mass gap in the gauge sector. We currently do not know the extension to $\R^4$.

Figures (3)

• Figure 1: Summary of perturbative analysis: Solid line indicates the running of the gauge coupling in QCD(adj) compactified on a small circle $S^1$ with circumference $L$, and dashed line is the usual running on ${\mathbb R}^4$. In the regime $1/L \gg \Lambda$ perturbative Coleman-Weinberg analysis is reliable, and leads to a radiatively induced gauge symmetry breaking $G \rightarrow H$ where $G=SU(2)$ and $H= U(1)$. To all orders in perturbation theory, the long-distance theory described by $H$ is free due to absence of charged massless excitations. This is reminiscent of the ${\cal N}=2$ SYM theory on ${\mathbb R}^4$, for which gauge symmetry breaking takes place on the semi-classical domain of the moduli space.
• Figure 2: (Left)Magnetically and topologically charged monopoles carry compulsory fermion zero-modes. Consequently, they cannot induce a bosonic potential for the dual photon. (Right) Topologically null, magnetically charged bions have no external fermionic legs. Hence, they induce the leading bosonic potential, which implies mass for the dual photon and confinement. The figure is for $SU(2)$ with $n_f=2$. The combination of the BPS- KK monopoles (which is not depicted) is an instanton (or caloron). It is present in confined phase, but is not the source of the dual photon mass term.
• Figure 3: The cartoon of the behavior of the center, discrete and continuous chiral symmetry realization in QCD(adj), for $SU(N)$ where $N=$ few, $n_f=2$ and $n_f=1$ (${\cal N}=1$ SYM). The spatial center symmetry is unbroken at any $L$ in both cases $\langle {\rm tr} U \rangle =0$. In $n_f=2$, the continuous chiral symmetry is unbroken at small $S^1$ and broken at large $S^1$, and discrete chiral symmetry is always broken. The red (dotted) line is the chiral condensate in ${\cal N}=1$ SYM, and the discrete chiral symmetry is always broken. In the small $S^1$ regime, the string tensions and thicknesses (the inverse mass gap in gauge sector) are calculable, and $n_f=2$ theory exhibits confinement without continuous chiral symmetry breaking. The lines slightly on top of the horizontal axis are all zero and are split to guide the eye.