Completing the screening picture of adjoint matter: the all-rank string tension law of 1+1-dimensional adjoint QCD and confinement versus screening in 2+1-dimensional N=1 supersymmetric Yang-Mills

Gross, Klebanov, Matytsin, and Smilga (1996) established that two-dimensional $SU(N)$ gauge theory with a massless adjoint Majorana fermion — adjoint QCD2, the dimensional reduction of 4d $\mathcal{N}=1$ supersymmetric Yang-Mills (SYM) — screens rather than confines fundamental test charges, and posed three follow-up questions: (i) a paradox in which fermion zero modes counted in instanton sectors appeared to forbid the bilinear condensate predicted by bosonization, together with the mass dependence of the string tension $\sigma(m)$ for $N\ge 3$; (ii) whether the massless theory has a continuous spectrum above some threshold or a discrete spectrum with exponentially small gaps; (iii) whether 2+1- and 3+1-dimensional $\mathcal{N}=1$ SYM might also be in a screening phase. Later literature resolved the mechanism of (i): the naive zero modes are not topologically robust (a mod-2 index governs the protected ones), and non-invertible symmetries enforce deconfinement of the undeformed massless theory for every $N$. It also established the small-mass law $\sigma_k\propto|m|\sin(\pi k/N)$, but rigorously only for $N\le 5$, with an explicitly stated unproven conjecture for all $N$, and it settled (ii) (continuum onset at the two-particle threshold at large $N$; infinitely dense planar spectrum) and the 3+1d case of (iii) (confinement, so the screening conjecture is refuted there). What survives as the open core: a rigorous all-$N$ derivation of the $k$-string tension law together with its overall normalization, and a dedicated determination of confinement versus screening in 2+1-dimensional $\mathcal{N}=1$ SYM, for which no first-principles study exists.

Background

In a gauge theory, whether an isolated external charge can be separated is diagnosed by the Wilson loop $W(C)$, the trace of the gauge-field holonomy (path-ordered exponential) around a closed curve $C$. If $\langle W(C)\rangle$ decays exponentially with the enclosed area (an area law), the potential between an external charge-anticharge pair grows linearly with separation, $V(r)\approx\sigma r$, with coefficient $\sigma$ called the string tension, and the theory confines; if the decay follows the perimeter (a perimeter law), dynamical charges screen the external one and the potential saturates. For gauge group $SU(N)$ the sharp question concerns probes of nonzero N-ality, meaning representations that transform nontrivially under the center subgroup $Z_N\subset SU(N)$ (the fundamental representation is the simplest case). Matter in the adjoint representation has zero N-ality and cannot, by representation theory alone, screen a nonzero-N-ality charge in three or four spacetime dimensions.

The theory at the center of this problem is adjoint QCD2 (also written QCD(adj)): $SU(N)$ Yang-Mills theory in 1+1 dimensions coupled to a single Majorana fermion (a fermion that is its own antiparticle) in the adjoint representation, of mass $m$. In two dimensions the gauge coupling $g$ has dimensions of mass, and the standard large-$N$ expansion parameter is the 't Hooft coupling $\lambda=g^2N$. The theory is the dimensional reduction of 4d $\mathcal{N}=1$ supersymmetric Yang-Mills (SYM), the minimal supersymmetric gauge theory, in which the gluon is accompanied by a single adjoint Weyl fermion (the gluino); at one special nonzero value of $m$ the spectrum of the two-dimensional theory becomes supersymmetric. Two-dimensional dynamics is dominated by topology: since $\pi_1(SU(N)/Z_N)=Z_N$, Euclidean field configurations split into instanton sectors labeled by $k\in Z_N$ (instantons are topologically nontrivial classical gauge fields), and a classical solution in the $k$-sector carries $k(N-k)$ doublets of fermion zero modes — exact zero eigenvalues of the Dirac operator forced by an index theorem — a count due to Smilga in this context.

In 1996, Gross, Klebanov, Matytsin, and Smilga analyzed this theory and found that it screens. Using bosonization — the exact correspondence in 1+1 dimensions between fermionic theories and bosonic ones; $n$ free Majorana fermions are equivalent to the $Spin(n)_1$ Wess-Zumino-Witten (WZW) model, a solvable conformal field theory — they argued that the massless theory flows in the infrared to the coset topological quantum field theory $Spin(N^2-1)_1/SU(N)_N$, which has a number of vacua growing like $2^N$, and that the fundamental Wilson loop then obeys a perimeter law. Giving the fermion a mass restores confinement; for $N=2$ they computed the string tension exactly, $\sigma=2\Sigma m$, where $\Sigma$ is the magnitude of the fermion bilinear condensate.

The same paper posed the three questions this problem addresses. (i) For $N\ge 3$, each nontrivial instanton sector carries $2(N-1)$ zero modes (the $k=1$ case of the count above), which naively forbids a bilinear fermion condensate, contradicting the bosonization prediction of one — a paradox the authors called unresolved — and for $N\ge 3$ the mass dependence of the string tension, in their words, "has not been sorted out". (ii) Since the string tension vanishes at $m=0$, it is natural to ask whether the massless theory develops a continuous spectrum above some threshold energy, or instead a discrete spectrum with exponentially small level spacings. (iii) They asked whether 2+1- and 3+1-dimensional $\mathcal{N}=1$ SYM might also be in a screening phase, being "bold enough to conjecture" this for 3+1 dimensions.

Three strands of later work bear on these questions.

Spectrum. Gross, Hashimoto, and Klebanov argued that at large $N$ the spectrum becomes continuous above twice the mass of the lightest bound state — the two-particle decay threshold — with support from discretized light-cone quantization (DLCQ, a numerical method that puts a 1+1d theory on a finite light-cone momentum grid). Armoni, Frishman, and Sonnenschein derived exact string tensions in massive two-dimensional QCD and screening in its two-dimensional supersymmetric variants. The modern numerical program — the analytic large-$N$ truncation of Katz, Tavares, and Xu (six single-particle bound states below the first multi-particle threshold), finite-$N$ DLCQ, the basis-function method of Trittmann, and Hamiltonian and Euclidean lattice studies — confirms a gapped, discrete, low-lying spectrum with continuum onset at multi-particle thresholds. In the theory with fundamental matter added, Dempsey, Klebanov, and Pufu demonstrated quantitatively, using exact affine (Kac-Moody) algebra identities, that the meson spectrum becomes continuous above a threshold, which is the precise mechanism of screening there. Question (ii) is therefore regarded as answered: discrete one-particle states below the lowest multi-particle threshold, a continuum above it, and an infinitely dense planar spectrum at $N=\infty$ (Komargodski, Ohmori, Roumpedakis, and Seifnashri showed that the Hagedorn temperature — the temperature at which the exponential growth of the density of states makes the canonical partition function diverge — vanishes in the planar massless theory).

Symmetry and the zero-mode paradox. Cherman, Jacobson, Tanizaki, and Ünsal showed that only $[k(N-k)]\bmod 2$ zero-mode doublets are topologically protected (a mod-2 index); the rest of the naive count is an artifact of symmetric backgrounds and disappears under generic deformations, as Smilga demonstrated explicitly. The premise of the 1996 paradox — that topology forbids the condensate — is therefore wrong. The fate of screening was settled after an initial disagreement in the literature: Cherman and collaborators argued from mixed 't Hooft anomaly matching that the theory confines for $N>2$, while Komargodski, Ohmori, Roumpedakis, and Seifnashri proved that the massless theory admits roughly $2^{2N}$ non-invertible symmetries — generalized symmetries implemented by topological fusion defects that do not multiply like group elements — and that these enforce deconfinement of the fundamental Wilson line for every $N$. The vacua organize into universes: superselection sectors, labeled here by $p\in Z_N$, that anomalies prevent from being connected by local operators. For small fermion mass $m\ll g$, the theory confines, the non-invertible symmetries are softly broken, and their analysis yields the small-mass $k$-string tension law, for a Wilson loop of N-ality $k$,

$$\sigma_k \;\propto\; |m|\,\sin\!\left(\frac{\pi k}{N}\right),$$

computed analytically for $N\le 5$ and conjectured for all $N$ (the proportionality constant is of order $\sqrt{\lambda}$; it is written $m\lambda^{1/2}\sin(\pi k/N)$ in the conventions of Bergner, Piemonte, and Ünsal, and $g_{\rm YM}|m|\sin(\pi k/N)$ in those of Komargodski, Ohmori, Roumpedakis, and Seifnashri). Screening was then confirmed numerically: Bergner, Piemonte, and Ünsal found on the lattice that the string tension is finite at $m\ne 0$ and tends to zero in the chiral limit, and Dempsey, Pufu, Søgaard, and Klebanov, by exact diagonalization of a lattice Hamiltonian formulation, found that the $p=0$/$p=1$ universe splitting of $SU(3)$ vanishes in the continuum limit at $m=0$ and computed the $SU(3)$ chiral condensates directly.

Higher dimensions. For 3+1d $\mathcal{N}=1$ SYM the question is settled in the negative: the theory confines. Ünsal's magnetic-bion mechanism — condensation of bound states of monopole-instanton events in the small-spatial-circle regime, argued to be continuously connected to infinite volume — yields confinement in four-dimensional adjoint QCD including $\mathcal{N}=1$ SYM; lattice simulations reproduce its expected low-lying supermultiplet spectrum; and since gluinos are adjoint (center-neutral) they cannot screen an N-ality-1 charge, while the two-dimensional screening mechanism — non-invertible symmetries available only through 2d bosonization — has no known analog in higher dimensions (Aharony, Cuomo, Komargodski, Mezei, and Raviv-Moshe classify Wilson-line screening in three and four dimensions and find it requires appropriately charged massless matter). For 2+1d $\mathcal{N}=1$ SYM, the intermediate case that the source also raised, no dedicated determination exists: the known 2+1d adjoint-matter results concern Chern-Simons-matter dualities (Armoni) and do not cover the pure Yang-Mills case.

After this literature, what remains open of the 1996 question is precisely: (a) the all-$N$ status and normalization of the small-mass string-tension law (rigorous only for $N\le 5$), and (b) confinement versus screening in 2+1d $\mathcal{N}=1$ SYM. These two clauses are the surviving core.

Problem Statement

This problem addresses the surviving open core of the 1996 screening question of Gross, Klebanov, Matytsin, and Smilga. Two linked deliverables complete their program; the parts already settled by later literature — the zero-mode paradox mechanism, the fact that massless 1+1d adjoint QCD screens for every $N$, the continuum-onset question at large $N$, and the 3+1d $\mathcal{N}=1$ SYM case, which confines — are background and are not part of the requested answer.

(A) All-rank string tension in 1+1d adjoint QCD. In $SU(N)$ gauge theory in 1+1 dimensions with one adjoint Majorana fermion of mass $m\ll\sqrt{\lambda}$ (where $\lambda=g^2N$ is the 't Hooft coupling), establish the leading small-mass behavior of the $k$-string tension $\sigma_k$ — the string tension of a Wilson loop of N-ality $k$ — for all $N\ge 3$ and all N-alities $1\le k\le N-1$: prove or refute the conjectured law $\sigma_k\propto|m|\sqrt{\lambda}\,\sin(\pi k/N)$ (the same law written as $\sigma_k\propto g_{\rm YM}|m|\sin(\pi k/N)$ in the conventions of Komargodski, Ohmori, Roumpedakis, and Seifnashri), and determine the overall proportionality constant in terms of the ultraviolet parameters. The definition of $\sigma_k$ used (which universe sector the string belongs to, and how string breaking by dynamical adjoint matter is treated) and the status of the assumption — load-bearing in the existing $N\le 5$ derivations — that the massless theory flows to the $Spin(N^2-1)_1/SU(N)_N$ coset topological field theory are themselves part of what the determination must settle. Current status: proven for $N\le 5$; conjectured, with an explicit statement that no proof exists, for $N\ge 6$.

(B) Confinement versus screening in 2+1d $\mathcal{N}=1$ SYM. For $SU(N)$ Yang-Mills theory in 2+1 dimensions with one massless adjoint Majorana fermion, for any $N\ge 2$, determine whether the fundamental (nonzero N-ality) Wilson loop obeys an area law (confinement) or a perimeter law (screening) in the continuum, spatially infinite theory. A controlled first-principles determination is required. No dedicated determination of either outcome exists in the literature.

The common objective of (A) and (B) is a complete quantitative theory of when massless adjoint matter screens fundamental test charges: across gauge-group rank in 1+1 dimensions, and across spacetime dimension in the supersymmetric Yang-Mills family, between the settled 1+1d screening endpoint and the settled 3+1d confinement endpoint. Answering (A) closes the last quantitative gap in the two-dimensional theory whose screening behavior the source established; answering (B) supplies the one missing data point between those endpoints.

References

1. D. J. Gross, I. R. Klebanov, A. V. Matytsin, and A. V. Smilga (1996), "Screening versus confinement in 1 + 1 dimensions", Nuclear Physics B 461, 109-130, arXiv: hep-th/9511104, https://arxiv.org/abs/hep-th/9511104, DOI: 10.1016/0550-3213(95)00655-9, https://doi.org/10.1016/0550-3213(95)00655-9
2. A. V. Smilga (1994), "Instantons and fermion condensate in adjoint two-dimensional QCD", Physical Review D 49, 6836-6848, DOI: 10.1103/PhysRevD.49.6836, https://doi.org/10.1103/PhysRevD.49.6836
3. A. Armoni, Y. Frishman, and J. Sonnenschein (1998), "The string tension in massive QCD2", Physical Review Letters 80, 430-433, arXiv: hep-th/9709097, https://arxiv.org/abs/hep-th/9709097, DOI: 10.1103/PhysRevLett.80.430, https://doi.org/10.1103/PhysRevLett.80.430
4. D. J. Gross, A. Hashimoto, and I. R. Klebanov (1998), "Spectrum of a large N gauge theory near the transition from confinement to screening", Physical Review D 57, 6420-6428, arXiv: hep-th/9710240, https://arxiv.org/abs/hep-th/9710240, DOI: 10.1103/PhysRevD.57.6420, https://doi.org/10.1103/PhysRevD.57.6420
5. A. Armoni, Y. Frishman, and J. Sonnenschein (1999), "Screening in supersymmetric gauge theories in two dimensions", Physics Letters B 449, 76-80, arXiv: hep-th/9807022, DOI: 10.1016/S0370-2693(99)00018-0, https://doi.org/10.1016/S0370-2693(99)00018-0, https://arxiv.org/abs/hep-th/9807022
6. E. Katz, G. Marques Tavares, and Y. Xu (2014), "Solving 2D QCD with an adjoint fermion analytically", Journal of High Energy Physics 05 (2014) 143, DOI: 10.1007/JHEP05(2014)143, https://doi.org/10.1007/JHEP05(2014)143
7. M. Ünsal (2009), "Magnetic bion condensation: A new mechanism of confinement and mass gap in four dimensions", Physical Review D 80, 065001, DOI: 10.1103/PhysRevD.80.065001, https://doi.org/10.1103/PhysRevD.80.065001
8. G. Bergner, P. Giudice, I. Montvay, G. Münster, and S. Piemonte (2015), "The low-lying spectrum of N=1 supersymmetric Yang-Mills theory", Proceedings of Science (EPS-HEP2015) 372, arXiv: 1510.04942, https://arxiv.org/abs/1510.04942
9. A. Cherman, T. Jacobson, Y. Tanizaki, and M. Ünsal (2020), "Anomalies, a mod 2 index, and dynamics of 2d adjoint QCD", SciPost Physics 8, 072, DOI: 10.21468/SciPostPhys.8.5.072, https://doi.org/10.21468/SciPostPhys.8.5.072
10. Z. Komargodski, K. Ohmori, K. Roumpedakis, and S. Seifnashri (2021), "Symmetries and strings of adjoint QCD2", Journal of High Energy Physics 03 (2021) 103, arXiv: 2008.07567, https://arxiv.org/abs/2008.07567, DOI: 10.1007/JHEP03(2021)103, https://doi.org/10.1007/JHEP03(2021)103
11. R. Dempsey, I. R. Klebanov, and S. S. Pufu (2021), "Exact symmetries and threshold states in two-dimensional models for QCD", Journal of High Energy Physics 10 (2021) 096, DOI: 10.1007/JHEP10(2021)096, https://doi.org/10.1007/JHEP10(2021)096
12. A. Smilga (2021), "A comment on instantons and their fermion zero modes in adjoint QCD_2", SciPost Physics 10, 152, DOI: 10.21468/SciPostPhys.10.6.152, https://doi.org/10.21468/SciPostPhys.10.6.152
13. R. Dempsey, I. R. Klebanov, L. L. Lin, and S. S. Pufu (2023), "Adjoint Majorana QCD2 at finite N", Journal of High Energy Physics 04 (2023) 107, arXiv: 2210.10895, https://arxiv.org/abs/2210.10895, DOI: 10.1007/JHEP04(2023)107, https://doi.org/10.1007/JHEP04(2023)107
14. O. Aharony, G. Cuomo, Z. Komargodski, M. Mezei, and A. Raviv-Moshe (2023), "Phases of Wilson lines: conformality and screening", Journal of High Energy Physics 12 (2023) 183, DOI: 10.1007/JHEP12(2023)183, https://doi.org/10.1007/JHEP12(2023)183
15. R. Dempsey, I. R. Klebanov, S. S. Pufu, and B. T. Søgaard (2024), "Lattice Hamiltonian for adjoint QCD2", Journal of High Energy Physics 08 (2024) 009, arXiv: 2311.09334, https://arxiv.org/abs/2311.09334, DOI: 10.1007/JHEP08(2024)009, https://doi.org/10.1007/JHEP08(2024)009
16. G. Bergner, S. Piemonte, and M. Ünsal (2024), "Investigating two-dimensional adjoint QCD on the lattice", Journal of High Energy Physics 07 (2024) 048, DOI: 10.1007/JHEP07(2024)048, https://doi.org/10.1007/JHEP07(2024)048
17. R. Dempsey, I. R. Klebanov, S. S. Pufu, and B. T. Søgaard (2024), "Small circle expansion for adjoint QCD2 with periodic boundary conditions", Journal of High Energy Physics 11 (2024) 128, arXiv: 2406.17079, https://arxiv.org/abs/2406.17079, DOI: 10.1007/JHEP11(2024)128, https://doi.org/10.1007/JHEP11(2024)128
18. U. Trittmann (2024), "Solving two-dimensional adjoint QCD with a basis-function approach", Physical Review D 109, 016004, DOI: 10.1103/PhysRevD.109.016004, https://doi.org/10.1103/PhysRevD.109.016004
19. A. Armoni (2022), "Dualities of adjoint QCD3 from branes", Journal of High Energy Physics 09 (2022) 073, arXiv: 2205.15706, https://arxiv.org/abs/2205.15706, DOI: 10.1007/JHEP09(2022)073, https://doi.org/10.1007/JHEP09(2022)073
20. R. Dempsey, S. S. Pufu, B. T. Søgaard, and I. R. Klebanov (2025), "More about the lattice Hamiltonian for Adjoint QCD2", Journal of High Energy Physics 06 (2025) 260, arXiv: 2409.19164, https://arxiv.org/abs/2409.19164, DOI: 10.1007/JHEP06(2025)260, https://doi.org/10.1007/JHEP06(2025)260
21. I. R. Klebanov, S. S. Pufu, B. Søgaard, and E. Witten (2025), "Supercurrents and (partial) supersymmetry in adjoint QCD_2 and its generalizations", arXiv: 2507.17138, https://arxiv.org/abs/2507.17138