The problem originates in Remark 3.1 of A. Dynin's "Mathematical Quantum Yang–Mills Theory Revisited II: Mass without Mass" (Russ. J. Math. Phys. 27(1), 22–30, 2020), where a positive mass gap is derived for the second quantization of a massless Dirac field coupled to a fixed classical Yang–Mills background field a′a' on a spatial ball of radius RR; the gap scales as 1/R1/R and depends on the chosen a′a'. In that construction the gauge field is never a dynamical quantum variable, and no fully gauge-invariant, background- and cutoff-independent gap is obtained. Once the fixed background and cutoff are removed, the open core is precisely the Clay Millennium Problem "Yang–Mills existence and mass gap": prove that for every compact simple gauge group GG there exists a nontrivial quantum Yang–Mills theory on four-dimensional Minkowski spacetime, with the gauge connection as a fully dynamical quantum variable, possessing a positive mass gap Δ>0\Delta>0 defined without reference to any classical background field or finite cutoff.

Background

Yang–Mills theory is the nonabelian generalization of electromagnetism: the field is a connection AA taking values in the Lie algebra of a compact gauge group GG (physically SU(3)SU(3) for the strong interaction), and its self-interaction makes the classical theory nonlinear even in the absence of matter. Classically the gauge bosons (gluons) are massless — a mass term would break gauge invariance — yet the physical strong interaction is short-ranged: confinement binds quarks and gluons into massive hadrons, and lattice QCD (numerical simulation of Wilson's spacetime discretization of the theory; QCD, quantum chromodynamics, is the SU(3)SU(3) Yang–Mills theory of the strong interaction) reproduces the observed hadron spectrum from essentially massless constituents. A mass gap means, roughly, that the spectrum of the quantum Hamiltonian above the vacuum is separated from zero by some Δ>0\Delta>0, or equivalently that suitable gauge-invariant correlations decay exponentially; Wilson loops (traces of parallel transport around closed curves, which are gauge-invariant observables) provide the canonical such observables. The mathematical challenge is to construct the quantum theory itself: a mathematically rigorous, interacting four-dimensional quantum Yang–Mills theory satisfying the standard axioms of axiomatic quantum field theory (the Wightman axioms, adapted to gauge theories by requiring them only of gauge-invariant observables: Hilbert-space positivity, Poincaré covariance, locality, a spectral condition, and a unique vacuum). Constructive quantum field theory succeeded for superrenormalizable interacting models (models with only finitely many ultraviolet divergences, so that renormalization requires only finitely many counterterms) in spacetime dimension d≤3d\le 3 (the Glimm–Jaffe program), but in four dimensions no interacting theory satisfying the axioms is known: removing ultraviolet and volume cutoffs while keeping a nontrivial (interacting) limit, and handling gauge redundancy — gauge fixing admits no global section (the Gribov ambiguity, the obstruction that different gauge choices identify the same physical configuration multiple times) — have resisted all available techniques. Because of this, the Clay Mathematics Institute in 2000 designated "Yang–Mills existence and mass gap" one of its seven Millennium Prize Problems: prove that for any compact simple gauge group GG, a nontrivial quantum Yang–Mills theory exists on R4\mathbb{R}^4 and has a mass gap. The source paper, A. Dynin's "Mass without Mass" (2020), approaches the mass-generation side of this question from the physics observation (Wilczek's "QCD Lite") that most of the visible mass of the universe emerges from the quantum interaction of massless quarks and gluons, with no Higgs field. Dynin studies the classical Cauchy problem (the evolution of the fields from prescribed initial data) for the coupled Dirac–Yang–Mills (DYM) equations in temporal gauge (the gauge choice that sets the time component of the connection to zero, giving the equations Hamiltonian form), solves the Gauss constraint (the condition, arising from the gauge-field equations, that the covariant divergence of the electric field equals the charge density, which restricts admissible initial data) on the initial-data manifold via Nash–Moser-type implicit function theorems (an infinite-dimensional version of the implicit function theorem that tolerates loss of derivatives), and then fixes a classical transversal Yang–Mills field (a′,e′)(a', e') and applies Berezin–Fock second quantization (the standard construction of a quantum field theory on fermionic Fock space, the direct sum of antisymmetrized tensor powers of a one-particle Hilbert space) only to the massless Dirac field in that background. His Theorem 3.1 asserts the resulting second-quantized interaction Hamiltonian has discrete spectrum with an isolated vacuum eigenvalue, i.e. a positive mass gap; but his own Remark 3.1 records that this gap runs in inverse proportion to the ball radius RR (i.e. with the effective classical energy scale) and "incidentally it depends on the selected a′a'". Thus within that framework the gauge field is a classical external background, the gap is cutoff- and background-dependent, and no gauge-invariant, background-independent gap — the object the Clay problem demands — is produced. The paper's introduction further asserts that its predecessor (Dynin 2017) already "solved" the Clay problem for pure Yang–Mills fields in vacuum; that stronger claim is not established by the arguments in that paper.

Problem Statement

Prove that for every compact simple gauge group GG there exists a nontrivial quantum Yang–Mills theory on four-dimensional Minkowski spacetime R3,1\mathbb{R}^{3,1} in which the gauge connection is a fully dynamical quantum variable, and that this theory has a positive mass gap: a constant Δ>0\Delta>0 such that the spectrum of the theory above the vacuum contains no masses in (0,Δ)(0,\Delta), equivalently suitable gauge-invariant correlations (for example, of Wilson loops) decay exponentially in their separation at rates bounded below by Δ>0\Delta>0. Here "nontrivial" is given its standard Clay-problem reading: the theory must be the quantization of the Yang–Mills action for GG and an interacting (non-Gaussian) theory — equivalently, one whose gauge-invariant correlations, in particular its Wilson-loop expectation values, differ from those of the free (Gaussian) field theory — obtained as a nontrivial non-Gaussian scaling limit rather than a free or otherwise degenerate construction. The theory must satisfy the standard axioms of axiomatic quantum field theory for its gauge-invariant observables, and the theory, the gap, and its definition must be formulated without reference to any fixed classical background gauge field or finite spatial cutoff, unlike the source paper, where the gauge field is frozen to a classical solution (a′,e′)(a',e') and the derived gap scales as 1/R1/R and depends on a′a'. This is the Clay Millennium Problem "Yang–Mills existence and mass gap". The source paper's construction — second quantization of a massless Dirac field in a fixed classical Yang–Mills background — points to a related open direction, the simultaneous quantization of massless fermions and the gauge field; a background-independent gap for such a coupled Dirac–Yang–Mills system would not by itself establish the pure-gauge core stated above and is not a substitute for it. The renormalization-group flow of the gap and its connection to Standard-Model particle masses are downstream phenomenological questions that presuppose the core construction above and are not part of the problem posed here.

References

  1. A. Dynin, "Mathematical Quantum Yang–Mills Theory Revisited II: Mass without Mass", Russian Journal of Mathematical Physics 27(1), 22–30 (2020). DOI: 10.1134/S1061920820010033; arXiv:1909.12090; https://arxiv.org/abs/1909.12090
  2. A. Dynin, "Mathematical quantum Yang–Mills theory revisited", Russian Journal of Mathematical Physics 24(1), 19–36 (2017). DOI: 10.1134/S1061920817010022; arXiv:1308.6571; https://arxiv.org/abs/1308.6571
  3. Clay Mathematics Institute, "Yang-Mills & the Mass Gap", official Millennium Prize Problem page, marked "Unsolved", https://www.claymath.org/millennium/yang-mills-the-maths-gap/ (accessed 2026-08-28)
  4. M. Frasca, A. Ghoshal, S. Groote, "Confinement in QCD and generic Yang-Mills theories with matter representations", Physics Letters B 846, 138209 (2023). DOI: 10.1016/j.physletb.2023.138209; arXiv:2202.14023; https://arxiv.org/abs/2202.14023
  5. R. Vilela Mendes, "A consistent measure for lattice Yang-Mills", International Journal of Modern Physics A 32, 1750016 (2017). DOI: 10.1142/S0217751X17500166; arXiv:1711.03598; https://arxiv.org/abs/1711.03598
  6. R. Vilela Mendes, "Measure and mass gap for generalized connections on hypercubic lattices", arXiv:1504.07798; https://arxiv.org/abs/1504.07798
  7. S. Farinelli, "Renormalized Four Dimensional Quantum Yang-Mills Theory and Mass Gap", arXiv:1406.4177; https://arxiv.org/abs/1406.4177