Cao–Nissim–Sheffield (2025) proved Wilson's area law for lattice Yang–Mills with gauge group U(N), SU(N), or SO(2N) in the 't Hooft strong-coupling regime by verifying the Durhuus–Fröhlich σ-model mass-gap condition through the Bakry–Émery curvature criterion, and explicitly left open three directions: whether Chatterjee's boundary-uniform mass-gap condition follows from Bakry–Émery, what happens for gauge groups with trivial center such as SO(2N+1), and whether U(N) admits a direct Bakry–Émery treatment. Later work has reshaped these questions. Nissim (2025) established the mass gap, infinite-volume limit, and large-N limit for U(N) via a U(1)×SU(N) random-environment decomposition, so the direct-Bakry–Émery question for U(N) is now mostly methodological. Nissim (2026) proved that SO(3) lattice Yang–Mills, whose center is trivial, fails Wilson's confinement criterion at strong coupling — deconfinement, the opposite of an area law — so the correct surviving question for centerless groups is to prove failure of confinement, not to extend the area law. The problem posed here is the surviving open core: prove that the remaining trivial-center orthogonal groups SO(2N+1), N ≥ 2, likewise fail Wilson's criterion in the strong-coupling regime (or refute the screening prediction by proving an area law), and determine whether the Bakry–Émery condition implies Chatterjee's mass-gap criterion under arbitrary boundary conditions; within the settled SO(3) case, removing the 'almost every β' restriction and establishing a uniform perimeter-law lower bound remain open strengthenings of the same component.

Background

Lattice Yang–Mills theory is a Gibbsian discretization of Euclidean quantum gauge theory introduced by Wilson (1974). Group elements of a compact Lie group GG (the gauge group) are assigned to the oriented edges of the lattice Zd\mathbb{Z}^d; a plaquette is a unit square, and the Wilson action in the 't Hooft normalization weights a configuration QQ by exp⁡(Nβ∑pReTr(Qp))\exp(N\beta\sum_p \mathrm{ReTr}(Q_p)), where QpQ_p is the ordered product of edge variables around a plaquette and β\beta is the inverse coupling strength. The regime β\beta below a small dimensional constant (with the NβN\beta scaling above) is called the 't Hooft regime; it is the regime in which the large-NN limit is nontrivial and in which all recent rigorous results cited here are proved. The central observable is the Wilson loop Wℓ=tr(Qℓ)W_\ell = \mathrm{tr}(Q_\ell), where tr=1NTr\mathrm{tr} = \frac{1}{N}\mathrm{Tr} is the trace normalized by the 't Hooft parameter NN (the dimension of the defining matrix representation for SU(N)\mathrm{SU}(N) and U(N)\mathrm{U}(N); for the orthogonal groups in scope the choice of normalization constant is immaterial, since rescaling tr\mathrm{tr} by a fixed factor changes V(R)V(R) only by a term vanishing as T→∞T\to\infty), applied to the holonomy (ordered product of edge variables) around a closed loop ℓ\ell. For a rectangular loop ℓR,T\ell_{R,T} with side lengths RR and TT, Wilson's criterion for confinement of static quarks is that the quark–antiquark potential V(R)=−lim⁡T→∞1Tlog⁡⟨WℓR,T⟩V(R) = -\lim_{T\to\infty}\frac{1}{T}\log \langle W_{\ell_{R,T}} \rangle diverges as R→∞R \to \infty. A sufficient condition is the area law, ∣⟨Wℓ⟩∣≤c1e−c2 area(ℓ)|\langle W_\ell \rangle| \le c_1 e^{-c_2\,\mathrm{area}(\ell)} for all rectangular loops, which yields a linearly growing potential (a nonzero string tension); decay at perimeter-law rate — ∣⟨Wℓ⟩∣|\langle W_\ell \rangle| decaying like e−c ∣ℓ∣e^{-c\,|\ell|} with ∣ℓ∣|\ell| the length of the loop — instead indicates screening of the fundamental charge by dynamical degrees of freedom.

Rigorous area-law results have historically required strong coupling. Osterwalder and Seiler (1978) proved area law at very strong coupling (β\beta below a threshold of order 1/N1/N in the normalization above) via a cluster expansion (a convergent expansion of expectations in powers of the coupling). Chatterjee (2021) identified a probabilistic mechanism: he rigorously defined center symmetry for lattice gauge theory, proved that unbroken center symmetry implies the area law for representations acting nontrivially on the center, and showed that a sufficient condition for unbroken center symmetry is exponential decay of correlations under arbitrary boundary conditions (his Definition 2.3, recalled in the problem statement). The center of GG is the subgroup of elements commuting with all of GG; e.g. {e2πik/NI}\{e^{2\pi i k/N} I\} for SU(N)\mathrm{SU}(N) and {±I}\{\pm I\} for SO(2N)\mathrm{SO}(2N), while SO(2N+1)\mathrm{SO}(2N+1) and SO(3)\mathrm{SO}(3) have trivial center. Shen–Zhu–Zhu (2023) introduced a dynamical approach: the Langevin dynamics of the lattice field (a diffusion on the configuration space having the Gibbs measure as its invariant measure) is proved to mix rapidly once a Bakry–Émery curvature condition holds (a criterion from Riemannian geometry: Ricci curvature of the configuration space minus the Hessian of the action is bounded below, which forces exponential decay of correlations, log-Sobolev and Poincaré inequalities), giving a mass gap for SU(N)\mathrm{SU}(N) and SO(N)\mathrm{SO}(N) in the 't Hooft regime. Cao–Nissim–Sheffield (2025) combined these threads: they verified, via Bakry–Émery, the mass-gap condition of Durhuus and Fröhlich (1980) — exponential decay for the (d−1)(d-1)-dimensional nonlinear σ\sigma-model (a spin system on vertices obtained by conditioning a height-1 slab of the lattice on its top and bottom boundary fields, with action coupling each spin to its neighbors through the fixed fields), uniformly in the boundary fields — and concluded the area law in the whole 't Hooft regime for G∈{U(N),SU(N),SO(2N)}G \in \{\mathrm{U}(N), \mathrm{SU}(N), \mathrm{SO}(2N)\}. For U(N)\mathrm{U}(N), whose Ricci curvature is not uniformly positive, they condition on the U(1)\mathrm{U}(1) factor and reduce to the SU(N)\mathrm{SU}(N) case.

Every group covered by this argument has a nontrivial center, and this is essential rather than accidental: in both the Durhuus–Fröhlich and Chatterjee arguments, a central element zIzI, z≠1z \ne 1, forces the one-point functions of the σ\sigma-model to vanish, which is what converts covariance decay into decay of the loop expectation itself. The physics literature has long predicted the opposite behavior for gauge groups with trivial center, such as SO(2N+1)\mathrm{SO}(2N+1) or G2G_2: dynamical gluons screen the fundamental charge, so Wilson loops should show perimeter-law rather than area-law decay and Wilson's criterion should fail. Nissim (2026) made this rigorous for the first time: for SO(3)\mathrm{SO}(3) lattice Yang–Mills in dimension d≥3d \ge 3, for almost every β\beta in a strong-coupling interval, the quark–antiquark potential is uniformly bounded, so Wilson's criterion fails. In parallel, Nissim (2025) established the mass gap, the unique infinite-volume limit, and large-NN factorization for U(N)\mathrm{U}(N) in the 't Hooft regime by recasting U(N)\mathrm{U}(N) as a random-environment U(1)×SU(N)\mathrm{U}(1)\times\mathrm{SU}(N) model, precisely because, as that paper notes, the Bakry–Émery condition cannot be easily verified for U(N)\mathrm{U}(N). The open questions that remain are set out in the problem statement.

Problem Statement

The problem is to complete the confinement and mass-gap picture for lattice Yang–Mills theory in the strong-coupling ('t Hooft) regime for the cases left open by the dynamical approach of Cao–Nissim–Sheffield (2025). The components share one scientific objective: to determine, by rigorous proof, which lattice Yang–Mills theories confine and which mass-gap criteria capture confinement, beyond the nontrivial-center cases already settled. Component (1) is the central question: it is resolved by proving failure of Wilson's criterion for the trivial-center groups SO(2N+1)\mathrm{SO}(2N+1), N≥2N\ge 2, or by proving an area law for some such group. The Bakry–Émery question of component (2), the two SO(3)\mathrm{SO}(3) strengthenings, and the analogous questions for other centerless compact groups named within component (1) are further open questions.

(1) Trivial-center groups (primary component). For gauge group G=SO(2N+1)G = \mathrm{SO}(2N+1) with N≥2N \ge 2 (trivial center), dimension d≥3d \ge 3, and the Wilson action in the 't Hooft normalization, prove that there exists β∗(d,N)>0\beta^*(d,N) > 0 such that for all β∈(0,β∗)\beta \in (0, \beta^*) — or, matching the current state of the art, for almost every such β\beta — the theory fails Wilson's confinement criterion: the quark–antiquark potential V(R)=−lim⁡T→∞1Tlog⁡⟨WℓR,T⟩V(R) = -\lim_{T\to\infty}\frac{1}{T}\log \langle W_{\ell_{R,T}} \rangle is uniformly bounded in RR. A rigorous proof that some SO(2N+1)\mathrm{SO}(2N+1) instead satisfies the area law in this regime, refuting the screening prediction, is an equally valid resolution of this component. Within the settled case SO(3)\mathrm{SO}(3), two strengthenings remain open: removing the 'almost every β\beta' restriction, and establishing the uniform perimeter-law lower bound. The analogous questions for other centerless compact groups (e.g. SU(N)/ZN\mathrm{SU}(N)/\mathbb{Z}_N, G2G_2) are natural generalizations of the same component.

(2) Mass-gap criteria from Bakry–Émery (secondary component). Determine whether the Bakry–Émery curvature condition — verified in the 't Hooft regime for SU(N)\mathrm{SU}(N) and SO(N)\mathrm{SO}(N) — implies Chatterjee's mass-gap criterion (Definition 2.3 of Chatterjee 2021): there exist constants K1,K2>0K_1, K_2 > 0 depending only on GG, β\beta, dd such that for every cube, every boundary condition on its boundary edges, and all local [−1,1][-1,1]-valued functions f,gf, g of the edge variables supported inside the cube, ∣⟨fg⟩−⟨f⟩⟨g⟩∣≤K1e−K2 dist(f,g)|\langle fg \rangle - \langle f \rangle \langle g \rangle| \le K_1 e^{-K_2\,\mathrm{dist}(f,g)}, where dist(f,g)\mathrm{dist}(f,g) is the Euclidean distance between the midpoints of the edges on which ff and gg are supported. An affirmative proof gives the alternative route to area law envisioned in Remark 1.1 of the source paper; an explicit counterexample (Bakry–Émery holds while boundary-uniform decay fails) resolves it negatively. As a further methodological sub-question, determine whether a direct Bakry–Émery argument, or a suitably modified curvature criterion, can treat U(N)\mathrm{U}(N) without conditioning on its U(1)\mathrm{U}(1) factor — noting that the principal consequences for U(N)\mathrm{U}(N) (mass gap, infinite-volume limit, large-N limit, area law) are already established by other means, so this sub-question is purely about the method.

References

  1. Sky Cao, Ron Nissim, Scott Sheffield. Dynamical approach to area law for lattice Yang-Mills. arXiv:2509.04688 (2025). DOI: 10.48550/arXiv.2509.04688. https://arxiv.org/abs/2509.04688
  2. Ron Nissim. U(N) lattice Yang-Mills in the 't Hooft regime. arXiv:2510.22788 (2025). DOI: 10.48550/arXiv.2510.22788. https://arxiv.org/abs/2510.22788
  3. Ron Nissim. Deconfinement for SO(3) lattice Yang-Mills at strong coupling. arXiv:2605.16162 (2026). DOI: 10.48550/arXiv.2605.16162. https://arxiv.org/abs/2605.16162
  4. Sourav Chatterjee. A probabilistic mechanism for quark confinement. Communications in Mathematical Physics 385:1007-1039 (2021). DOI: 10.1007/s00220-021-04086-y. arXiv:2006.16229. https://arxiv.org/abs/2006.16229
  5. Sky Cao, Ron Nissim, Scott Sheffield. Expanded regimes of area law for lattice Yang-Mills theories. arXiv:2505.16585 (2025). DOI: 10.48550/arXiv.2505.16585. https://arxiv.org/abs/2505.16585
  6. B. Durhuus, J. Fröhlich. A connection between ν-dimensional Yang-Mills theory and (ν-1)-dimensional, non-linear σ-models. Communications in Mathematical Physics 75:103-151 (1980). DOI: 10.1007/bf01222514. https://doi.org/10.1007/bf01222514
  7. Hao Shen, Rongchan Zhu, Xiangchan Zhu. A stochastic analysis approach to lattice Yang-Mills at strong coupling. Communications in Mathematical Physics 400:805-851 (2023). DOI: 10.1007/s00220-022-04609-1. https://doi.org/10.1007/s00220-022-04609-1
  8. K. Osterwalder, E. Seiler. Gauge field theories on a lattice. Annals of Physics 110:440-471 (1978). DOI: 10.1016/0003-4916(78)90039-8. https://doi.org/10.1016/0003-4916(78)90039-8
  9. Kenneth G. Wilson. Confinement of quarks. Physical Review D 10:2445-2459 (1974). DOI: 10.1103/PhysRevD.10.2445. https://doi.org/10.1103/PhysRevD.10.2445