Infrared fate of the noncompact CP$^1$ gauge theory: is the Coulomb-to-crystal transition of the three-dimensional classical dimer model continuous or weakly first order?

The close-packed classical dimer model on the cubic lattice undergoes a direct transition between a high-temperature Coulomb phase — a disordered phase with algebraic correlations, an emergent noncompact U(1) gauge field, and deconfined monomers — and a low-temperature columnar dimer crystal. A transfer-matrix mapping to quantum bosons, followed by a vortex duality, shows that the long-wavelength theory of this transition is the noncompact CP$^1$ model (NCCP$^1$): a two-component complex field minimally coupled to a noncompact U(1) gauge field in three space-time dimensions. The central open question, posed by Powell and Chalker (2009) and still unresolved, is whether this theory possesses a nontrivial renormalization-group fixed point — so that the transition is genuinely continuous with an emergent symmetry, the SU(2) manifest in $\mathcal{L}_0$ being its minimal form and the strictly larger SO(5) (the rotation group of a five-component order parameter) its sharpest form, or whether the renormalization-group flow instead produces an anomalously weak first-order transition with only quasiuniversal 'pseudocritical' behavior and approximate symmetry. Satellite open questions concern the microscopic dependence of the gauge stiffness $\kappa$, the irrelevance of symmetry-breaking operators, possible tricriticality of the nearest-neighbor model, and rigorous locality of the effective quantum Hamiltonian generated by the transfer matrix. No resolution is known in the literature through 2026.

Background

A classical dimer model is a statistical-mechanical ensemble of close-packed dimer coverings: each elementary bond of a lattice either carries a dimer or not, with the constraint that every site is touched by exactly one dimer. On the cubic lattice, the unbiased ensemble of coverings forms a Coulomb phase: a disordered phase whose long-distance correlations are described by an emergent noncompact U(1) gauge field, with correlation functions decaying algebraically and having nontrivial directional dependence, analogous to the power-law correlations of magnetic-field fluctuations in magnetostatics. The divergence-free part of the dimer occupation pattern plays the role of the magnetic field, and defects violating close packing (empty or doubly occupied sites, called monomers) carry the gauge charge; in the Coulomb phase monomers are deconfined, meaning that separating a monomer–antimonomer pair to infinity costs only a finite energy.

Interactions between dimers can drive ordering. With a classical energy $-v_2 N_2$, where $N_2$ counts flippable plaquettes (plaquettes carrying two parallel dimers), the cubic dimer model develops, below a critical temperature, one of six columnar dimer crystals that break the translation and rotation symmetries of the lattice. Monte Carlo simulations by Alet et al. (2006) found a direct transition between the Coulomb phase and the crystal that appears continuous: no latent heat or discontinuity is resolved up to the largest sizes simulated, which implies a correlation length (the distance over which critical fluctuations extend) of at least hundreds of lattice spacings. The measured critical exponents and order-parameter distributions (Misguich et al. 2008) are, however, unlike those of any conventional universality class.

The transition is unconventional because the two adjacent phases are characterized by unrelated diagnostics: the crystal has a conventional order parameter (the dimer magnetization $m_\mu$, the imbalance between the fraction of dimers oriented along lattice axis $\mu$ and the fractions oriented along the other axes), while the defining property of the Coulomb phase is topological — algebraic correlations and deconfined monomers — with no local order parameter. A Landau–Ginzburg–Wilson theory of the crystal order parameter alone cannot describe a continuous transition between such phases without fine tuning. This is the classical-statistical counterpart of deconfined quantum criticality, proposed by Senthil et al. (2004) for transitions between antiferromagnetic phases and valence-bond-solid phases (nonmagnetic states in which spins pair into localized singlet bonds in a pattern that breaks lattice translation symmetry) of two-dimensional quantum magnets.

Powell and Chalker (2009) made the correspondence precise through a two-stage mapping. First, the transfer matrix of the classical model (the operator whose trace gives the partition function), taken along the [111] direction, maps the three-dimensional classical dimer problem exactly onto a quantum model of hard-core bosons on the kagome lattice (a two-dimensional lattice of corner-sharing triangles) at fractional filling (1/6 per site): the Coulomb phase is the boson superfluid, and the dimer crystal is a Mott insulator (a commensurate, incompressible insulating phase) with density-wave order. Second, a duality transformation to the vortices (points around which the boson phase winds by $2\pi$ and the superfluid density vanishes) of this boson model yields the continuum theory

$\mathcal{L}_0 = |(\nabla - iA)\varphi|^2 + s|\varphi|^2 + u(|\varphi|^2)^2 + \kappa|\nabla\times A|^2,$

the noncompact CP$^1$ model (NCCP$^1$): a two-component complex field $\varphi$ (the vortex, or matter, field) minimally coupled to a noncompact U(1) gauge field $A$. 'Noncompact' means that monopole configurations of $A$ are excluded: they map onto monomers, which the close-packing constraint forbids. The same continuum theory was obtained independently by direct three-dimensional duality mappings (Charrier–Alet–Pujol 2008; Chen et al. 2009). The transition corresponds to condensation of $\varphi$, which gaps the gauge field via the Anderson–Higgs mechanism (the condensate screens the gauge field; the crystal is this screened, or Higgs, phase), as the tuning parameter $s$ crosses its critical value. After tuning $s$, the theory retains one free parameter, the gauge stiffness $\kappa$, whose dependence on the microscopic interactions is unknown.

Remarkably, $\mathcal{L}_0$ has a manifest SU(2) symmetry acting on the two-component field $\varphi$, although the microscopic dimer model has only cubic lattice symmetry. The lowest-order symmetry-allowed term that breaks SU(2) is eighth order in the fields, $\mathcal{L}_1 = v\sum_\mu m_\mu^4$, where $m_\mu$ is the dimer magnetization (a bilinear in $\varphi$). If this term and all higher-order symmetry-breaking terms are irrelevant under renormalization-group (RG) flow — meaning they shrink in strength as the length scale grows, leaving the long-distance physics invariant — a continuous transition would exhibit emergent SU(2) symmetry: observables would become symmetric under SU(2) on long length scales even though no microscopic symmetry requires it. Whether this happens, and whether the RG flow of $\mathcal{L}_0$ has any nontrivial fixed point in three space-time dimensions at all — as opposed to running away toward a (weakly) first-order transition — is not settled: an expansion in $\epsilon = 4 - d$ (a perturbative RG technique in the dimensionality deficit from the upper critical dimension $d = 4$, at which the theory becomes Gaussian) has no weak-coupling fixed point to expand about, and the pre-2009 numerical evidence was contradictory (simulations of the J-Q model, a square-lattice quantum spin Hamiltonian mixing Heisenberg exchange $J$ with singlet-projection terms $Q$ and designed to realize the antiferromagnet–valence-bond-solid transition, supported continuity; Kuklov et al. argued for a generic first-order transition). The measured critical exponents of the nearest-neighbor dimer model were additionally noted to resemble those of a tricritical point — the meeting point of continuous and first-order transition lines — and the locality of the effective quantum Hamiltonian generated by the transfer matrix was argued only heuristically, within the low-energy sector.

Problem Statement

Determine the infrared fate of the NCCP$^1$ field theory in three space-time dimensions — equivalently, the order and universality class of the Coulomb-to-columnar-crystal transition of the close-packed classical dimer model on the cubic lattice. Powell and Chalker (2009) leave this as a cluster of linked open questions whose common objective is to establish whether the continuum gauge theory $\mathcal{L}_0$ actually describes a continuous transition:

(i) Does $\mathcal{L}_0$ have a nontrivial renormalization-group fixed point, so that the dimer-model transition is genuinely continuous, or does the flow instead produce a (weakly) first-order transition? The $\epsilon = 4 - d$ expansion provides no controlled determination, and the numerical evidence was inconclusive at the time and remains contested.

(ii) How does the gauge stiffness $\kappa$ depend on microscopic details of the classical configuration energy (for instance, whether it increases or decreases when further-neighbor interactions are added), and are there regimes of $\kappa$ — or microscopic models realizing them — with qualitatively different behavior, such as continuous versus first-order transitions?

(iii) Are the SU(2)-breaking operator $\mathcal{L}_1 = v\sum_\mu m_\mu^4$ and all higher-order microscopic symmetry-breaking terms irrelevant at the critical point, so that emergent SU(2) symmetry holds asymptotically? In the sharpest current form of this question (established by later work), the putative emergent symmetry is the larger SO(5) rotating the dimer magnetization together with the monopole operators (topological disorder operators that insert monopoles of the gauge field $A$ at fixed probe positions — equivalently, in the dimer language, monomers held at those positions; the close-packing constraint excludes such defects from the equilibrium ensemble, but their correlation functions are well defined, and their deconfinement — separating the pair to infinity at finite energy cost — is the defining property of the Coulomb phase), and the issue is whether it is exact at a fixed point or only approximate.

(iv) Does the unperturbed nearest-neighbor cubic dimer model lie close to a tricritical point, as suggested by comparisons of critical exponents, and does a continuous line of transitions connect tricritical and generic NCCP$^1$ regimes as microscopic parameters are tuned?

(v) Can the locality of the effective quantum Hamiltonian $\mathcal{H}$ obtained from the transfer matrix — argued to hold within the low-energy sector — be established rigorously, or can $\mathcal{H}$ be constructed explicitly, on arbitrarily large lattices?

The central open problem is (i): settling the order of the transition and the structure of the renormalization-group flow — a nontrivial fixed point; a pseudocritical flow that approaches but never reaches a fixed point and ends in an anomalously weak first-order transition; or a direct runaway to a strongly first-order transition. Items (ii)–(v) are the satellite open questions introduced above: each is posed in Powell and Chalker (2009) as a related remark of independent interest, not as a component that must additionally be settled before a resolution of (i) counts as complete.

References

1. Stephen Powell and J. T. Chalker, 'Classical to quantum mapping for an unconventional phase transition in a three-dimensional classical dimer model', Phys. Rev. B 80, 134413 (2009). DOI: 10.1103/PhysRevB.80.134413; https://arxiv.org/abs/0907.1564
2. Stephen Powell and J. T. Chalker, 'SU(2)-invariant continuum theory for an unconventional phase transition in a three-dimensional classical dimer model', Phys. Rev. Lett. 101, 155702 (2008). DOI: 10.1103/PhysRevLett.101.155702; https://doi.org/10.1103/PhysRevLett.101.155702
3. David A. Huse, Werner Krauth, R. Moessner, and S. L. Sondhi, 'Coulomb and liquid dimer models in three dimensions', Phys. Rev. Lett. 91, 167004 (2003). DOI: 10.1103/PhysRevLett.91.167004; https://doi.org/10.1103/PhysRevLett.91.167004
4. Fabien Alet, Grégoire Misguich, Vincent Pasquier, Roderich Moessner, and Jesper Lykke Jacobsen, 'Unconventional continuous phase transition in a three-dimensional dimer model', Phys. Rev. Lett. 97, 030403 (2006). DOI: 10.1103/PhysRevLett.97.030403; https://doi.org/10.1103/PhysRevLett.97.030403
5. Grégoire Misguich, Vincent Pasquier, and Fabien Alet, 'Correlations and order parameter at a Coulomb-crystal phase transition in a three-dimensional dimer model', Phys. Rev. B 78, 100402(R) (2008). DOI: 10.1103/PhysRevB.78.100402; https://doi.org/10.1103/PhysRevB.78.100402
6. D. Charrier, F. Alet, and P. Pujol, 'Gauge theory picture of an ordering transition in a dimer model', Phys. Rev. Lett. 101, 167205 (2008). DOI: 10.1103/PhysRevLett.101.167205; https://doi.org/10.1103/PhysRevLett.101.167205
7. Gang Chen, Jan Gukelberger, Simon Trebst, Fabien Alet, and Leon Balents, 'Coulomb gas transitions in three-dimensional classical dimer models', Phys. Rev. B 80, 045112 (2009). DOI: 10.1103/PhysRevB.80.045112; https://doi.org/10.1103/PhysRevB.80.045112
8. D. Charrier and F. Alet, 'Phase diagram of an extended classical dimer model', Phys. Rev. B 82, 014429 (2010). DOI: 10.1103/PhysRevB.82.014429; https://doi.org/10.1103/PhysRevB.82.014429
9. G. J. Sreejith and Stephen Powell, 'Critical behavior in the cubic dimer model at nonzero monomer density', Phys. Rev. B 89, 014404 (2014). DOI: 10.1103/PhysRevB.89.014404; https://doi.org/10.1103/PhysRevB.89.014404
10. G. J. Sreejith, Stephen Powell, and Adam Nahum, 'Emergent SO(5) symmetry at the columnar ordering transition in the classical cubic dimer model', Phys. Rev. Lett. 122, 080601 (2019). DOI: 10.1103/PhysRevLett.122.080601; https://arxiv.org/abs/1803.11218
11. T. Senthil, Ashvin Vishwanath, Leon Balents, Subir Sachdev, and Matthew P. A. Fisher, 'Deconfined quantum critical points', Science 303, 1490 (2004). DOI: 10.1126/science.1091806; https://doi.org/10.1126/science.1091806
12. A. B. Kuklov, M. Matsumoto, N. V. Prokof'ev, B. V. Svistunov, and M. Troyer, 'Deconfined criticality: generic first-order transition in the SU(2) symmetry case', Phys. Rev. Lett. 101, 050405 (2008). DOI: 10.1103/PhysRevLett.101.050405; https://doi.org/10.1103/PhysRevLett.101.050405
13. Adam Nahum, J. T. Chalker, P. Serna, M. Ortuño, and A. M. Somoza, 'Deconfined quantum criticality, scaling violations, and classical loop models', Phys. Rev. X 5, 041048 (2015). DOI: 10.1103/PhysRevX.5.041048; https://arxiv.org/abs/1506.06798
14. Adam Nahum, P. Serna, J. T. Chalker, M. Ortuño, and A. M. Somoza, 'Emergent SO(5) symmetry at the Néel to valence-bond-solid transition', Phys. Rev. Lett. 115, 267203 (2015). DOI: 10.1103/PhysRevLett.115.267203; https://doi.org/10.1103/PhysRevLett.115.267203
15. Chong Wang, Adam Nahum, Max A. Metlitski, Cenke Xu, and T. Senthil, 'Deconfined quantum critical points: symmetries and dualities', Phys. Rev. X 7, 031051 (2017). DOI: 10.1103/PhysRevX.7.031051; https://doi.org/10.1103/PhysRevX.7.031051
16. Hui Shao, Wenan Guo, and Anders W. Sandvik, 'Quantum criticality with two length scales', Science 352, 213 (2016). DOI: 10.1126/science.aad5007; https://doi.org/10.1126/science.aad5007
17. Bowen Zhao, Phillip Weinberg, and Anders W. Sandvik, 'Symmetry-enhanced discontinuous phase transition in a two-dimensional quantum magnet', Nature Physics 15, 678 (2019). DOI: 10.1038/s41567-019-0484-x; https://doi.org/10.1038/s41567-019-0484-x
18. Pablo Serna and Adam Nahum, 'Emergence and spontaneous breaking of approximate O(4) symmetry at a weakly first-order deconfined phase transition', Phys. Rev. B 99, 195110 (2019). DOI: 10.1103/PhysRevB.99.195110; https://arxiv.org/abs/1805.03759
19. Ruochen Ma and Chong Wang, 'Theory of deconfined pseudocriticality', Phys. Rev. B 102, 020407(R) (2020). DOI: 10.1103/PhysRevB.102.020407; https://doi.org/10.1103/PhysRevB.102.020407
20. Yu Nakayama and Tomoki Ohtsuki, 'Necessary condition for emergent symmetry from the conformal bootstrap', Phys. Rev. Lett. 117, 131601 (2016). DOI: 10.1103/PhysRevLett.117.131601; https://doi.org/10.1103/PhysRevLett.117.131601
21. Anders W. Sandvik and Bowen Zhao, 'Consistent scaling exponents at the deconfined quantum-critical point', Chin. Phys. Lett. 37, 057502 (2020). DOI: 10.1088/0256-307X/37/5/057502; https://doi.org/10.1088/0256-307X/37/5/057502
22. Bilal Hawashin, Astrid Eichhorn, Lukas Janssen, Michael M. Scherer, and Shourya Ray, 'The Nordic-walking mechanism and its explanation of deconfined pseudocriticality from Wess-Zumino-Witten theory', Nature Communications 16, 20 (2025). DOI: 10.1038/s41467-024-54884-w; https://doi.org/10.1038/s41467-024-54884-w
23. Yi Cui, Rong Yu, and Weiqiang Yu, 'Deconfined quantum critical point: a review of progress', Chin. Phys. Lett. 42, 047503 (2025). DOI: 10.1088/0256-307X/42/4/047503; https://arxiv.org/abs/2504.10027
24. Hu-Xiao Peng, Zheng Yan, and Shuai Yin, 'Short-time critical dynamics in the classical cubic dimer model', Commun. Theor. Phys. 78, 075602 (2026). DOI: 10.1088/1572-9494/ae5941; https://arxiv.org/abs/2605.15730
25. Ryan Flynn and Anders W. Sandvik, 'SU(N) quantum spin model with weak and strong first-order Néel to valence-bond solid transitions' (2026). arXiv:2603.16106; https://arxiv.org/abs/2603.16106
26. Chunhao Guo, Zhe Wang, Danhe Wang, Zenan Liu, Haiyuan Zou, and Zheng Yan, 'Lee-Yang zeros and pseudocritical drift in J-Q Néel-VBS transitions' (2026). arXiv:2603.15116; https://arxiv.org/abs/2603.15116
27. Shai M. Chester, Alessandro Piazza, Marten Reehorst, and Ning Su, 'Bootstrapping the simplest deconfined quantum critical point' (2025). arXiv:2507.06283; https://arxiv.org/abs/2507.06283