The three-dimensional quantum crystal-melting Hamiltonian of Dijkgraaf, Orlando and Reffert acts on Hilbert-space states labeled by plane partitions; its one- and two-dimensional reductions are exactly solvable (the 2D model is the XXZ spin chain in disguise), and on this evidence the 3D model was conjectured to be integrable and to share the mass gap of its lower-dimensional cousins. Araujo (2021) reformulated the model as a particle-occupation problem on the Kagome lattice and then as a classical hexagon vertex model with 18 local Boltzmann weights, proving that the natural alternating row-to-row transfer matrices of the original connectivity fail to commute, while two modified even/odd descendant subsystems are Yang–Baxter integrable. The open problem is to determine the 18 weights as explicit functions of the couplings and a spectral parameter, to construct a commuting transfer-matrix scheme for the original, unmodified Kagome rules or prove an obstruction, and to determine whether the integrable descendants compose to reconstitute the full quantum Hamiltonian on the plane-partition sector — the intended route to the excited-state spectrum and the mass-gap conjecture. No resolution is known in the literature to date. The problem is open.

Background

A plane partition is a two-dimensional array of non-increasing non-negative integers, visualized as a stack of unit cubes piled against the corner of the positive octant of R3\mathbb{R}^3 (a "3D Young diagram"). The statistical mechanics of a melting crystal corner is the weighting and enumeration of such stacks, with answer encoded in the MacMahon function M(q)=∏n≥1(1−qn)−nM(q)=\prod_{n\geq1}(1-q^{n})^{-n}; Okounkov, Reshetikhin and Vafa connected this enumeration to topological-string amplitudes, which is one reason the model sits at the interface of combinatorics, statistical mechanics and string theory.

Dijkgraaf, Orlando and Reffert introduced the quantum version: a Hamiltonian whose basis states are plane partitions and whose elementary moves create or remove a single box,

H=−J∑(∣■⟩⟨□∣+∣□⟩⟨■∣)+V∑(q ∣□⟩⟨□∣+1q∣■⟩⟨■∣),H=-J\sum\left(|\blacksquare\rangle\langle\square|+|\square\rangle\langle\blacksquare|\right)+V\sum\left(\sqrt{q}\,|\square\rangle\langle\square|+\frac{1}{\sqrt{q}}|\blacksquare\rangle\langle\blacksquare|\right),

where ∣□⟩⟨□∣|\square\rangle\langle\square| counts positions where a box can consistently be added to the stack (addable corners), ∣■⟩⟨■∣|\blacksquare\rangle\langle\blacksquare| counts removable boxes, and JJ, VV, qq are real couplings (Hermiticity). Restricted to growth along one axis (the 1D problem) the model reduces to free fermions; in a quadrant (the 2D problem) it maps exactly onto the spin-1/2 XXZ Heisenberg chain, the paradigmatic spin chain solvable by the Bethe ansatz (an exact algebraic method for diagonalizing one-dimensional quantum Hamiltonians), whose spectrum in the domain-wall setting relevant here (fixed opposite spin orientations at the two chain ends, the boundary condition under which boxed plane partitions arise) was derived by Orlando, Reffert and Reshetikhin. The 1D and 2D models turn out to have the same mass gap, i.e. the same ground-state-to-first-excited-state energy difference. On this evidence Dijkgraaf, Orlando and Reffert conjectured that the full 3D model is also integrable and shares the same mass gap; the conjecture is supported by numerics but unproven. The ground state is the q#Boxes/2q^{\#\text{Boxes}/2}-weighted sum over all plane partitions, with norm squared given by the MacMahon function.

Araujo, Orlando and Reffert recast the 3D problem on a two-dimensional lattice: through a two-step map — boxes of the stack to dimers (occupied links) on a hexagonal lattice, and each dimer to a point particle at a dual site — box growth becomes a particle-occupation problem on the Kagome lattice (the lattice of corner-sharing triangles), described by two families of free fermions organized in alternating rows ("X-rows" and "Y-rows"); the bosonic plane-partition states were related to the MacMahon representation of the affine Yangian Y[gl^(1)]\mathcal{Y}[\widehat{\mathfrak{gl}}(1)] — an infinite-dimensional quantum algebra whose representation theory organizes plane partitions, an algebraic hint of integrability for which no constructive link to the Hamiltonian is known.

Araujo [1] pushed this program toward Yang–Baxter integrability. In a two-dimensional vertex model, each local configuration of edges meeting at a site carries a Boltzmann weight; integrability follows if the weights can be organized into a matrix (an RR-matrix or Lax operator) satisfying the Yang–Baxter equation, which guarantees that the row-to-row transfer matrices T(u)T(u) — traces of ordered (monodromy) products of Lax operators along a row, depending on a spectral parameter uu — commute for all uu; commutation of transfer matrices is the algebraic engine of exact solvability (via the RTT or fundamental commutation relation between the monodromy matrix entries). Araujo showed that the plane-partition states of the Kagome system decompose into exactly 18 allowed local hexagon configurations with position-independent weights W(Oi)\mathcal{W}(\mathcal{O}_i), so that the quantum problem becomes a classical hexagon vertex model: once the weights are known as functions of (J,V,q)(J,V,q) and a spectral parameter uu, the quantum eigenenergies follow. A Lax operator was defined as the scattering matrix of the three-particle processes at a hexagon; it is a singular operator acting non-trivially only on a 5-dimensional physical subspace. However, the natural alternating row-to-row transfer matrices t(u)\mathbf{t}(u), t~(v)\tilde{\mathbf{t}}(v) built with the original Kagome connection rules provably fail to commute, [t(u),t~(v)]≠0[\mathbf{t}(u),\tilde{\mathbf{t}}(v)]\neq0, which blocks the standard route: the Yang–Baxter equation cannot even be imposed on the weights. Under modified rules connecting the hexagons, each family does commute with itself, [t(u),t(v)]=0[\mathbf{t}(u),\mathbf{t}(v)]=0 and [t~(u),t~(v)]=0[\tilde{\mathbf{t}}(u),\tilde{\mathbf{t}}(v)]=0, defining two Yang–Baxter integrable subsystems, called descendants of the 6-vertex model (the archetypal integrable vertex model; "descendant" means the auxiliary and quantum spaces of the Lax operator are not isomorphic — here the X-row states carry the fundamental representation of su(2)su(2) while the Y-row pairs carry the adjoint of sl(2,C)sl(2,\mathbb{C})). Crucially, these integrable subsystems belong to a modified connectivity, not to the original Kagome model of the crystal-melting Hamiltonian.

Problem Statement

Does the three-dimensional quantum crystal-melting Hamiltonian admit a global Yang–Baxter integrable structure, and can it be established through the classical hexagon reformulation? The source (Araujo 2021) poses the following jointly linked questions, whose common objective is the exact solvability of HH; a complete answer addresses them together:

(a) Determine the 18 local Boltzmann weights W(Oi)\mathcal{W}(\mathcal{O}_i) — equivalently the local energies εi\varepsilon_i — as explicit functions of the quantum couplings (J,V,q)(J,V,q) and a spectral parameter uu, such that the classical hexagon model reproduces the eigenvalues of the quantum Hamiltonian without diagonalizing the full HH.

(b) For the original Kagome connectivity (the diagonally linked tiling on which HH acts, with no modification of the local connection rules), construct a row-to-row or rows-to-rows transfer-matrix scheme — a different monodromy ordering, a higher-rank RR-matrix, or an equivalent device — whose transfer matrices commute for all spectral-parameter values; or prove that an obstruction exists, i.e. that no commuting transfer-matrix object can be correctly defined from the original rules at all. (The naive alternating scheme already fails, [t(u),t~(v)]≠0[\mathbf{t}(u),\tilde{\mathbf{t}}(v)]\neq0; as the source emphasizes, the absence of solutions for a particular choice of scheme does not by itself imply non-integrability, so the failure of any single proposed family — including this one — does not discharge the negative branch.)

(c) Determine whether the two integrable even/odd descendant vertex models can be composed or embedded so as to reconstitute the original Kagome connectivity and recover the full quantum crystal-melting Hamiltonian on the plane-partition sector — in particular whether solutions of the descendant models' Yang–Baxter equations can be used as building blocks for plane partitions — and develop the projection that separates the plane-partition macrostates from the classical configurations that have no plane-partition analogue.

The structure (a)–(c) is the intended route to the explicitly stated downstream goal: the determination of the spectrum beyond the ground state — in particular whether the first excited state admits an expansion ∣Ψ1⟩=∑ΛfΛ(q)∣Λ⟩|\Psi_1\rangle=\sum_\Lambda f_\Lambda(q)|\Lambda\rangle over plane partitions Λ\Lambda with computable coefficients fΛ(q)f_\Lambda(q) — and thereby a test of the Dijkgraaf–Orlando–Reffert conjectures that the 3D model is integrable and shares the mass gap of its 1D and 2D reductions. An answer may be constructive (explicit weights, a commuting transfer-matrix family, and recovery of HH on the plane-partition sector, with whatever spectral consequences it enables) or negative (a proof that no commuting transfer-matrix structure can be defined for the original rules — an obstruction to the correct definition of such objects, not merely the failure of a solver-selected subclass of schemes); a construction that only works for the modified descendant connectivity, or only for a proper subsystem, does not complete the problem as stated.

The source's open list also contains complementary directions — a concrete algebraic embedding of the model into the affine Yangian Y[gl^(1)]\mathcal{Y}[\widehat{\mathfrak{gl}}(1)], and the relations to N=2\mathcal{N}=2 supersymmetric theories, topological strings, the AGT correspondence (which relates two-dimensional conformal field theories to four-dimensional supersymmetric gauge theories) and AdS3_3/higher-spin duality. These are alternative routes and surrounding context rather than additional required deliverables: work along them resolves this problem exactly insofar as it settles (a)–(c).

References

  1. T. Araujo, "Towards Yang-Baxter integrability of quantum crystal melting: From Kagome lattice to vertex models", Nucl. Phys. B 963 (2021) 115287; DOI: 10.1016/j.nuclphysb.2020.115287; arXiv:2009.06812; https://arxiv.org/abs/2009.06812
  2. R. Dijkgraaf, D. Orlando, S. Reffert, "Quantum Crystals and Spin Chains", Nucl. Phys. B 811 (2009) 463–490; DOI: 10.1016/j.nuclphysb.2008.11.027; arXiv:0803.1927; https://arxiv.org/abs/0803.1927
  3. T. Araujo, D. Orlando, S. Reffert, "Quantum crystals, Kagome lattice and plane partitions fermion-boson duality", Phys. Rev. D 103 (2021) 026020; DOI: 10.1103/PhysRevD.103.026020; arXiv:2005.09103; https://arxiv.org/abs/2005.09103
  4. D. Orlando, S. Reffert, N. Reshetikhin, "On domain wall boundary conditions for the XXZ spin Hamiltonian", arXiv:0912.0348; https://arxiv.org/abs/0912.0348
  5. A. Okounkov, N. Reshetikhin, C. Vafa, "Quantum Calabi-Yau and classical crystals", in: The Unity of Mathematics, Prog. Math. 244 (2006) 597–618; DOI: 10.1007/0-8176-4467-9_16; arXiv:hep-th/0309208; https://arxiv.org/abs/hep-th/0309208
  6. N. Wang, "3D fermions and affine Yangian", Nucl. Phys. B 969 (2021) 115461; DOI: 10.1016/j.nuclphysb.2021.115461; https://doi.org/10.1016/j.nuclphysb.2021.115461
  7. A. Morozov, N. Tselousov, "Hunt for 3-Schur polynomials", Phys. Lett. B 840 (2023) 137887; DOI: 10.1016/j.physletb.2023.137887; arXiv:2211.14956; https://arxiv.org/abs/2211.14956
  8. F. Balducci, A. Gambassi, A. Lerose, A. Scardicchio, C. Vanoni, "Localization and melting of interfaces in the two-dimensional quantum Ising model", Phys. Rev. Lett. 129 (2022) 120601; DOI: 10.1103/PhysRevLett.129.120601; arXiv:2203.09495; https://arxiv.org/abs/2203.09495
  9. A. Prakash, J. S. Rao, S. A. Parameswaran, A. Lerose, "Interface phases and dynamics in two-dimensional quantum magnets: A 'holographic' approach from universality to quantum simulation", arXiv:2608.12312; https://arxiv.org/abs/2608.12312
  10. D. Galakhov, W. Li, "Charging solid partitions", J. High Energy Phys. 2024 (2024) 043; DOI: 10.1007/JHEP01(2024)043; arXiv:2311.02751; https://arxiv.org/abs/2311.02751
  11. J.-M. Stéphan, "Exact time evolution formulae in the XXZ spin chain with domain wall initial state", J. Phys. A: Math. Theor. 55 (2022) 204003; DOI: 10.1088/1751-8121/ac5fe8; arXiv:2112.12092; https://arxiv.org/abs/2112.12092
  12. T. Araujo, "Comments on Slavnov products, Temperley-Lieb open spin chains, and KP tau functions", Nucl. Phys. B 972 (2021) 115566; DOI: 10.1016/j.nuclphysb.2021.115566; arXiv:2107.13060; https://arxiv.org/abs/2107.13060
  13. T. Araujo, "Q-Boson model and relations with integrable hierarchies", Nucl. Phys. B 1006 (2024) 116640; DOI: 10.1016/j.nuclphysb.2024.116640; arXiv:2405.01213; https://arxiv.org/abs/2405.01213