Elliptic eight-vertex model with domain wall boundary conditions: normalization-level identification of the domain wall partition function at generic parameters

The eight-vertex model with domain wall boundary conditions (DWBC) is the most general exactly solved nearest-neighbour lattice model of its class, and its domain wall partition function (DWPF) with elliptic (theta-function) weights was obtained in 2008-2009 by several independent routes: the algebraic Bethe ansatz (ABA) of Yang and Zhang, giving a permutation-sum closed form that reduces to the Izergin determinant in the trigonometric limit; elliptic weight-function formulations of Pakuliak, Rubtsov and Silantyev and of Rosengren; and functional-equation derivations of Razumov and Stroganov. The 2009 source paper explicitly posed one follow-up problem — whether the DWPF admits a single determinant, the authors having obtained only a sum of N determinants — and announced an extension to open (reflecting-end) boundary conditions; the six-vertex determinant underlies norms of Bethe states, correlation functions, thermodynamic properties, and alternating-sign-matrix-type enumeration, from which analogous applications of the elliptic formula are implied as the natural continuation rather than posed. The single-determinant question was resolved by Galleas (2016) for the vertex-face-equivalent elliptic solid-on-solid model, and the open-boundary extension was resolved by Yang et al. (2011) and Filali (2011). What remains open, at the original generality, is the residual of the single-determinant question: an explicit normalization-level identification of the algebraic-Bethe-ansatz expression with the independently derived closed forms, since carrying the single-determinant form over to the eight-vertex ABA normalization requires exactly this identification. The source-implied downstream directions — correlation functions (local height probabilities) of the domain wall ensemble, and the thermodynamic limit of the DWPF, at generic elliptic parameters — also remain open.

Background

The eight-vertex model is a statistical-mechanical model on a square lattice whose edges carry arrow variables, with the constraint that an even number of arrows point into each site; all eight allowed local arrow configurations are assigned Boltzmann weights $a,b,c,d$. In the most general ('elliptic') regime these weights are parameterized by Jacobi theta functions — doubly periodic entire functions of a spectral parameter, with a modulus $p$ — and the trigonometric limit $p \to 0$ gives the six-vertex model, whose weights are ordinary trigonometric functions.

Domain wall boundary conditions (DWBC) prescribe that all arrows on the left and right outer boundaries of an $N \times N$ lattice point inward, and all arrows on the top and bottom point outward. This is the boundary condition natural to the transfer-matrix eigenproblem rather than to bulk statistical mechanics, and the corresponding partition function (the domain wall partition function, DWPF) is the central object of the theory: for the six-vertex model, Korepin characterized it by symmetry, recursion relations, and initial conditions, and Izergin solved that characterization as a single determinant. The Izergin determinant underlies the enumeration of alternating sign matrices (ASMs — square matrices with entries 0, ±1 whose partial row and column sums are all 1), boundary free energies, and asymptotics in the KPZ class (the Kardar–Parisi–Zhang universality class governing fluctuating interfaces), which is why the elliptic (eight-vertex) generalization matters.

Yang and Zhang (2009) derived the elliptic DWPF by the algebraic Bethe ansatz (ABA) — the construction of transfer-matrix eigenstates from creation operators acting on a reference state — extended to the elliptic setting, where the Yang–Baxter equation (the condition making row-to-row transfer matrices commute) becomes 'dynamical': it acquires an extra shift parameter, and the relevant algebra is an elliptic or dynamical quantum group. Their answer is a sum over the symmetric group $S_N$ (a permutation sum of products of theta functions), reducing in the trigonometric limit to the Izergin determinant, together with a representation as a sum of $N$ determinants. Independently, three groups obtained the partition function of the elliptic solid-on-solid (SOS) model with DWBC — the face model equivalent to the eight-vertex model under the vertex–face correspondence, in which heights on adjacent dual-lattice sites differ by fixed amounts and local face weights replace vertex weights — via elliptic weight functions — projections of products of total elliptic currents in the elliptic quantum group, whose integral-transform kernel reproduces the partition function (Pakuliak–Rubtsov–Silantyev; Rosengren) — and functional equations (Razumov–Stroganov).

The posed single-determinant question was resolved for the vertex-face-equivalent SOS object, but the surviving core is an explicit normalization-level identification between the ABA expression and the independent formulations, as made precise in the Problem Statement: the competing closed forms were derived in different normalizations and gauges (boundary theta-function factors, dynamical-parameter conventions), and no published proof equates the ABA expression with the elliptic weight-function or determinantal formulations at the level of explicit normalization factors. Two source-implied downstream directions remain open as context. First, correlation functions — for face models conventionally expressed as local height probabilities (LHPs), the probabilities of observing prescribed heights at chosen sites, the analogue of spin correlators — are available for the domain wall ensemble at special parameter points (cyclic, i.e. roots of unity; and rational limits) but not at generic elliptic parameters. Second, the thermodynamic limit (the behaviour of the DWPF as $N \to \infty$, including the boundary or surface free energy characteristic of the domain wall geometry) is known for the six-vertex model in detail, but no corresponding result at generic elliptic parameters is known.

Problem Statement

Resolve the surviving open residual of the source problem for the elliptic eight-vertex model with domain wall boundary conditions, at the original generality (arbitrary finite lattice size $N$, generic elliptic modulus $p$, generic inhomogeneity spectral parameters, no restriction to roots of unity or to trigonometric or rational limits):

Normalization-level identification: prove an explicit identity, including all boundary, gauge, and normalization prefactors, between the algebraic-Bethe-ansatz permutation-sum expression of the domain wall partition function and at least one independently derived formulation (the elliptic weight-function expressions, or one of the determinant representations, including the single-determinant form available for the vertex-face-equivalent solid-on-solid model). The identity must be valid for arbitrary $N$ and generic parameter values, and must degenerate correctly to the Izergin determinant in the trigonometric limit.

This identification is the surviving residual of the source paper's posed single-determinant question: the single determinant obtained by Galleas is established for the elliptic SOS object, and carrying it over to the normalization of the eight-vertex ABA expression requires exactly such an identity. The source-implied downstream directions described above — local height probabilities of the domain wall ensemble at generic elliptic parameters, where existing results are confined to cyclic (root-of-unity) and rational limits, and the $N \to \infty$ asymptotics of the domain wall partition function, including its bulk and surface free energies, at generic elliptic parameters — remain open; results on them are genuine related progress, and the identification removes the normalization ambiguity that currently obstructs transferring correlator and asymptotic technology between formulations, but these directions were not posed and form no part of the completion criterion for this problem.

References

1. W.-L. Yang and Y.-Z. Zhang, Partition function of the eight-vertex model with domain wall boundary condition, Journal of Mathematical Physics 50, 083518 (2009). DOI 10.1063/1.3205448. https://doi.org/10.1063/1.3205448 (arXiv:0903.3089, https://arxiv.org/abs/0903.3089)
2. A. V. Razumov and Yu. G. Stroganov, Three-coloring statistical model with domain wall boundary conditions: Functional equations, Theoretical and Mathematical Physics 161, 1325-1339 (2009). DOI 10.1007/s11232-009-0119-y. https://doi.org/10.1007/s11232-009-0119-y (arXiv:0805.0669, https://arxiv.org/abs/0805.0669)
3. S. Pakuliak, V. Rubtsov, and A. Silantyev, The SOS model partition function and the elliptic weight functions, Journal of Physics A: Mathematical and Theoretical 41, 295204 (2008). DOI 10.1088/1751-8113/41/29/295204. https://doi.org/10.1088/1751-8113/41/29/295204 (arXiv:0802.0195, https://arxiv.org/abs/0802.0195)
4. H. Rosengren, An Izergin–Korepin-type identity for the 8VSOS model, with applications to alternating sign matrices, Advances in Applied Mathematics 43, 137-155 (2009). DOI 10.1016/j.aam.2009.01.003. https://doi.org/10.1016/j.aam.2009.01.003 (arXiv:0801.1229, https://arxiv.org/abs/0801.1229)
5. W. Galleas, Refined functional relations for the elliptic SOS model, Nuclear Physics B 867, 855-871 (2013). DOI 10.1016/j.nuclphysb.2012.10.014. https://doi.org/10.1016/j.nuclphysb.2012.10.014 (arXiv:1207.5283, https://arxiv.org/abs/1207.5283)
6. W. Galleas, Partition function of the elliptic solid-on-solid model as a single determinant, Physical Review E 94, 010102(R) (2016). DOI 10.1103/PhysRevE.94.010102. https://doi.org/10.1103/PhysRevE.94.010102 (arXiv:1604.01223, https://arxiv.org/abs/1604.01223)
7. W. Galleas, On the elliptic gl2 solid-on-solid model: Functional relations and determinants, Journal of Mathematical Physics 60, 023503 (2019). DOI 10.1063/1.5039148. https://doi.org/10.1063/1.5039148 (arXiv:1606.06144, https://arxiv.org/abs/1606.06144)
8. W. Galleas, New determinants in the 8VSOS model with domain-wall boundaries, arXiv:1812.10781 (2018). https://arxiv.org/abs/1812.10781
9. W.-L. Yang, X. Chen, J. Feng, K. Hao, K.-J. Shi, C.-Y. Sun, Z.-Y. Yang, and Y.-Z. Zhang, Domain wall partition function of the eight-vertex model with a non-diagonal reflecting end, Nuclear Physics B 847, 367-386 (2011). DOI 10.1016/j.nuclphysb.2011.01.029. https://doi.org/10.1016/j.nuclphysb.2011.01.029 (arXiv:1108.0961, https://arxiv.org/abs/1108.0961)
10. G. Filali, Elliptic dynamical reflection algebra and partition function of SOS model with reflecting end, Journal of Geometry and Physics 61, 1789-1796 (2011). DOI 10.1016/j.geomphys.2011.01.002. https://doi.org/10.1016/j.geomphys.2011.01.002 (arXiv:1012.0516, https://arxiv.org/abs/1012.0516)
11. J. Lamers, Integral formula for elliptic SOS models with domain walls and a reflecting end, Nuclear Physics B 901, 556-583 (2015). DOI 10.1016/j.nuclphysb.2015.11.006. https://doi.org/10.1016/j.nuclphysb.2015.11.006 (arXiv:1510.00342, https://arxiv.org/abs/1510.00342)
12. K. Motegi, A duality formula between elliptic determinants, arXiv:1901.01352 (2019). https://arxiv.org/abs/1901.01352
13. D. Levy-Bencheton and V. Terras, Algebraic Bethe ansatz approach to form factors and correlation functions of the cyclic eight-vertex solid-on-solid model, Journal of Statistical Mechanics: Theory and Experiment (2013) P04015. DOI 10.1088/1742-5468/2013/04/P04015. https://doi.org/10.1088/1742-5468/2013/04/P04015 (arXiv:1212.0246, https://arxiv.org/abs/1212.0246)
14. D. Levy-Bencheton and V. Terras, Multi-point local height probabilities of the CSOS model within the algebraic Bethe ansatz framework, Journal of Statistical Mechanics: Theory and Experiment (2014) P04014. DOI 10.1088/1742-5468/2014/04/P04014. https://doi.org/10.1088/1742-5468/2014/04/P04014 (arXiv:1311.7692, https://arxiv.org/abs/1311.7692)
15. H. Rosengren, Elliptic pfaffians and solvable lattice models, Journal of Statistical Mechanics: Theory and Experiment (2016) 083106. DOI 10.1088/1742-5468/2016/08/083106. https://doi.org/10.1088/1742-5468/2016/08/083106 (arXiv:1605.02915, https://arxiv.org/abs/1605.02915)
16. N. Slavnov, A. Zabrodin, and A. Zotov, Scalar products of Bethe vectors in the 8-vertex model, Journal of High Energy Physics 2020, 123 (2020). DOI 10.1007/JHEP06(2020)123. https://doi.org/10.1007/JHEP06(2020)123 (arXiv:2005.11224, https://arxiv.org/abs/2005.11224)
17. A. J. Gerrard, K. Motegi, and K. Sakai, Higher rank elliptic partition functions and multisymmetric elliptic functions, Nuclear Physics B 1011, 116805 (2025). DOI 10.1016/j.nuclphysb.2025.116805. https://doi.org/10.1016/j.nuclphysb.2025.116805 (arXiv:2412.13561, https://arxiv.org/abs/2412.13561)
18. J. Lamers, The Functional Method for the Domain-Wall Partition Function, Symmetry, Integrability and Geometry: Methods and Applications (SIGMA) 14, 064 (2018). DOI 10.3842/SIGMA.2018.064. https://doi.org/10.3842/SIGMA.2018.064 (arXiv:1801.09635, https://arxiv.org/abs/1801.09635)
19. V. E. Korepin and P. Zinn-Justin, Thermodynamic limit of the six-vertex model with domain wall boundary conditions, Journal of Physics A: Mathematical and General 33, 7053-7066 (2000). DOI 10.1088/0305-4470/33/40/304. https://doi.org/10.1088/0305-4470/33/40/304