Functional equations and homogeneous limit of the determinant identity between the domain-wall six-vertex partition function and anti-periodic transfer-matrix eigenvalues

The partition function $Z$ of the six-vertex model with domain-wall boundary conditions can be written as a single determinant $Z=\kappa_0\det(\mathcal{H}_L)$ whose entries are built from one eigenvalue $\Lambda(\lambda)$ of the commuting family of anti-periodic six-vertex transfer matrices (the off-diagonal, spin-flip twist $G=\sigma^x$, for which $T(\lambda)=B(\lambda)+C(\lambda)$). Galleas, who proved this identity, explicitly left its consequences unexplored: the inhomogeneous functional equations for $\Lambda$ obtained by equating $\kappa_0\det(\mathcal{H}_L)$ with the Izergin-Korepin determinant of $Z$; the complete homogeneous limit in which the matrix entries develop simple poles; the recurrences and pairwise eigenvalue relations induced by the identity; and the transfer of exact results between the domain-wall and anti-periodic problems. The anti-periodic spectrum is independently characterized by separation of variables and Bethe-type $T-Q$ equations, but the analysis of the determinant bridge itself - whether its functional equations characterize $\Lambda$ on their own and what its regularized homogeneous limit is - remains open.

Background

The six-vertex model is a two-dimensional lattice model of ice type: arrows live on the edges of a square lattice, each vertex carries one of six allowed arrow configurations, and the partition function $Z$ is a sum over configurations of products of vertex weights. It is integrable because its weights form an $R$-matrix that satisfies the Yang-Baxter equation, a consistency condition which guarantees that the row-to-row transfer matrices $T(\lambda)=\mathrm{tr}_a[R_{aL}(\lambda-\mu_L)\cdots R_{a1}(\lambda-\mu_1)]$ commute for all values of the spectral parameter $\lambda$; the $\mu_j$ are inhomogeneity parameters and $a$ is an auxiliary two-dimensional space. With domain-wall boundary conditions (DWBC), where the external arrows on the left and right lattice boundaries point inward and those on the top and bottom point outward, the finite-volume partition function $Z(\lambda_1,\dots,\lambda_L)$ is given by the Izergin-Korepin determinant: Korepin characterized $Z$ by a recurrence relation and symmetry properties, and the determinant found by Izergin satisfies them (see Izergin, Coker and Korepin 1992 for the standard presentation). The thermodynamic limit of the DWBC model is well understood, including the arctic phenomenon of an ordered region separating from a disordered region (Korepin and Zinn-Justin 2000).

For toroidal (periodic or twisted) boundary conditions one may insert a twist matrix $G$ in the auxiliary trace, $T(\lambda)=\mathrm{tr}_a[G\,R_{aL}(\lambda-\mu_L)\cdots R_{a1}(\lambda-\mu_1)]$. de Vega (1984) classified the integrable twists for the six-vertex $R$-matrix: they are either diagonal or off-diagonal. Writing the monodromy matrix as a two-by-two matrix with operator entries $A,B,C,D$, the off-diagonal choice $G=\sigma^x$ gives $T(\lambda)=B(\lambda)+C(\lambda)$; the associated quantum spin chain is the XXZ Heisenberg chain with anti-periodic (spin-flip) boundary conditions, hence the name anti-periodic six-vertex model. Although the family $\{T(\lambda)\}$ still commutes, the off-diagonal twist breaks the $U(1)$ symmetry, so there is no pseudo-vacuum reference state and the algebraic Bethe ansatz does not apply, making the eigenvalue problem $T(\lambda)|\Psi\rangle=\Lambda(\lambda)|\Psi\rangle$ markedly harder than in the diagonal case.

Galleas (2014) discovered an unexpected exact relation between the two boundary problems: by the algebraic-functional (AF) method, which derives functional equations for symmetric functions directly from the Yang-Baxter algebra commutation relations, he expressed the DWBC partition function in terms of eigenvalues of the anti-periodic transfer matrix. Galleas (2019), the source of this problem, recast that relation as a single determinant

$$Z(\lambda_0,\dots,\lambda_{L-1}) = \kappa_0\,\det(\mathcal{H}_L),$$

where $\mathcal{H}_L$ is a tridiagonal block matrix whose size grows with $L$ (for $L\ge3$ it is built on a vector with $3\cdot 2^{L-2}-2$ components; e.g. the $L=3$ determinant is $4\times4$ and $\mathcal{H}_4$ is $10\times10$) and whose entries contain one single eigenvalue $\Lambda(\lambda)$ together with explicit coefficient functions $M_i^{(n)}$ and $N_{j,i}^{(n)}$ (rational functions of the spectral parameters), and the constant $\kappa_0$ is the ratio of two projections of the eigenvector and is fixed by the specialization $Z(\mu_1,\dots,\mu_L)=\kappa_0\prod_j\Lambda(\mu_j)$. The identity holds for every eigenvalue of $T$, so it yields $2^L$ equivalent determinant representations and is an invariant on the spectrum. The paper derived but did not exploit three consequences: an inhomogeneous functional equation for $\Lambda$ (obtained by equating $\kappa_0\det\mathcal{H}_L$ with the Izergin-Korepin determinant, with $Z$ itself as the inhomogeneity term - the anti-periodic counterpart of the periodic-model functional equations that Galleas (2018) reinterpreted as an auxiliary linear problem showing that the periodic six-vertex spectrum solves Riccati-type nonlinear differential equations); recurrence relations for $\Lambda$ induced by Korepin's recurrence for $Z$; and pairwise relations $\kappa_0\det\mathcal{H}_L=\bar\kappa_0\det\bar{\mathcal{H}}_L$ linking any two eigenvalues. It also left the complete homogeneous limit untreated: entries of $\mathcal{H}_L$ develop simple poles when the spectral parameters $\lambda_j$ coalesce, already for $L=3$ and $L=4$.

Independently of this bridge, the anti-periodic spectral problem itself is well characterized. Batchelor, Baxter, O'Rourke and Yung (1995) diagonalized the anti-periodic strip transfer matrix by the commuting-transfer-matrix method and computed the interfacial tension. Niccoli (2013) used Sklyanin's separation of variables (SoV), a method in which one constructs a basis in which a set of quantum separate variables is diagonal and the spectral problem reduces to a finite system, to obtain the complete anti-periodic spectrum (eigenvalues and eigenstates), prove its simplicity, and compute determinant form factors; Niccoli and Terras (2015) proved that this characterization is equivalent to Bethe-type $T-Q$ equations, where the $T-Q$ relation is Baxter's functional equation $\Lambda(\lambda)=a(\lambda)Q(\lambda-\gamma)/Q(\lambda)+d(\lambda)Q(\lambda+\gamma)/Q(\lambda)$ solved by an auxiliary function $Q$, here either with an inhomogeneous extra term (as in the off-diagonal Bethe ansatz of Cao, Yang, Shi and Wang 2013) or with double-period $Q$-functions. Xin et al. (2018), Le et al. (2021) and Lu et al. (2024) developed the thermodynamic-limit program for this chain (twisted boundary energy, root patterns, ground state and excitations, finite-temperature free energy at root-of-unity anisotropy). None of these works analyzes the determinant bridge, which therefore remains the open object.

Problem Statement

For the six-vertex model with anisotropy $\gamma$ and inhomogeneities $\{\mu_j\}_{j=1}^{L}$, analyze the determinantal identity $Z(\lambda_0,\dots,\lambda_{L-1})=\kappa_0\det(\mathcal{H}_L)$ relating the domain-wall partition function $Z$ to a single eigenvalue $\Lambda(\lambda)$ of the anti-periodic transfer matrix $T(\lambda)=B(\lambda)+C(\lambda)$, turning the identity from a formal result into a working analytic tool. Concretely, as posed and left open by the source paper:

1. Write out and solve, or rigorously characterize, the inhomogeneous functional equations for $\Lambda$ obtained by equating $\kappa_0\det(\mathcal{H}_L)$ with established determinant representations of $Z$ (the Izergin-Korepin determinant or the continuous determinantal representations): determine the full set of admissible pairs $(\Lambda,\kappa_0)$, the analytic structure of admissible $\Lambda$ (zeros, poles, degree and symmetry constraints), and the relation of these solutions to the Bethe-type parametrizations of the anti-periodic spectrum (inhomogeneous $T-Q$ relations and double-period $Q$-functions).

2. Regularize the complete homogeneous limit $\lambda_j\to\lambda$, $\mu_j\to\mu$ of the determinant representation: the entries of $\mathcal{H}_L$ develop simple poles when spectral parameters coalesce (already at $L=3,4$), so the naive limit of $\kappa_0\det(\mathcal{H}_L)$ is indeterminate; derive the well-defined limiting formula, the limiting form of the spectral constraints it encodes, and its consequences for $\Lambda$ in the homogeneous model.

3. Exploit the bridge-induced spectral constraints: the recurrence relations for $\Lambda$ following from Korepin's recurrence relation for $Z$, and the pairwise relations $\kappa_0\det(\mathcal{H}_L)=\bar\kappa_0\det(\bar{\mathcal{H}}_L)$ linking any two eigenvalues $\Lambda,\bar\Lambda$. Determine what these constraints imply about the anti-periodic spectrum (eigenvalue counting, pairing and degeneracy structure, including the even/odd-$L$ parity effects) and how this characterization is reconciled with the existing SoV and $T-Q$ characterizations of the same spectrum.

4. Use the bridge to transfer exact results between the two boundary-condition problems, for example: anti-periodic-model eigenvalue asymptotics and thermodynamics from the known thermodynamic limit of the domain-wall partition function, or new domain-wall observables from anti-periodic spectral data.

The common objective uniting these tasks is the analysis of one and the same determinant identity. Items 1-3 are each required and together constitute the resolution: they must make the identity's content explicit and reconcile it with the existing spectral characterizations of the anti-periodic six-vertex model. Item 4 is illustrative of the payoff rather than a required deliverable, and the examples listed there are not a mandatory checklist.

References

1. W. Galleas, "Domain-wall boundaries through non-diagonal twists in the six-vertex model", J. Stat. Mech. (2019) 023101, DOI 10.1088/1742-5468/aaf633, arXiv:1807.03659, https://arxiv.org/abs/1807.03659
2. W. Galleas, "Twisted Heisenberg chain and the six-vertex model with DWBC", J. Stat. Mech. (2014) P11028, DOI 10.1088/1742-5468/2014/11/P11028, arXiv:1312.6817, https://arxiv.org/abs/1312.6817
3. W. Galleas, "Functional relations from the Yang-Baxter algebra: Eigenvalues of the XXZ model with non-diagonal twisted and open boundary conditions", Nucl. Phys. B 790 (2008) 524-542, DOI 10.1016/j.nuclphysb.2007.09.011, arXiv:0708.0009, https://arxiv.org/abs/0708.0009
4. W. Galleas, "Six-vertex model and non-linear differential equations I. Spectral problem", Commun. Math. Phys. 363 (2018) 59-96, DOI 10.1007/s00220-018-3232-7, arXiv:1705.03408, https://arxiv.org/abs/1705.03408
5. A. Bossart and W. Galleas, "Functional relations in nineteen-vertex models with domain-wall boundaries", J. Math. Phys. 60 (2019) 103509, DOI 10.1063/1.5095588, arXiv:1810.06482, https://arxiv.org/abs/1810.06482
6. H. J. de Vega, "Families of commuting transfer matrices and integrable models with disorder", Nucl. Phys. B 240 (1984) 495-513, DOI 10.1016/0550-3213(84)90240-2, https://doi.org/10.1016/0550-3213(84)90240-2
7. V. E. Korepin, "Calculation of norms of Bethe wave functions", Commun. Math. Phys. 86 (1982) 391-418, DOI 10.1007/BF01212176, https://doi.org/10.1007/BF01212176
8. A. G. Izergin, D. A. Coker and V. E. Korepin, "Determinant formula for the six-vertex model", J. Phys. A 25 (1992) 4315-4334, DOI 10.1088/0305-4470/25/16/010, https://doi.org/10.1088/0305-4470/25/16/010
9. V. Korepin and P. Zinn-Justin, "Thermodynamic limit of the six-vertex model with domain wall boundary conditions", J. Phys. A 33 (2000) 7053-7066, DOI 10.1088/0305-4470/33/40/304, https://doi.org/10.1088/0305-4470/33/40/304
10. M. T. Batchelor, R. J. Baxter, M. J. O'Rourke and C. M. Yung, "Exact solution and interfacial tension of the six-vertex model with anti-periodic boundary conditions", J. Phys. A 28 (1995) 2759-2770, DOI 10.1088/0305-4470/28/10/009, arXiv:hep-th/9502040, https://arxiv.org/abs/hep-th/9502040
11. G. Niccoli, "Antiperiodic spin-1/2 XXZ quantum chains by separation of variables: Complete spectrum and form factors", Nucl. Phys. B 870 (2013) 397-420, DOI 10.1016/j.nuclphysb.2013.01.017, arXiv:1205.4537, https://arxiv.org/abs/1205.4537
12. G. Niccoli and V. Terras, "Antiperiodic XXZ chains with arbitrary spins: Complete eigenstate construction by functional equations in separation of variables", Lett. Math. Phys. 105 (2015) 989-1031, DOI 10.1007/s11005-015-0759-9, arXiv:1411.6488, https://arxiv.org/abs/1411.6488
13. H. Pei and V. Terras, "On scalar products and form factors by separation of variables: the antiperiodic XXZ model", J. Phys. A 55 (2022) 015205, DOI 10.1088/1751-8121/ac3b85, arXiv:2011.06109, https://arxiv.org/abs/2011.06109
14. J. Cao, W.-L. Yang, K. Shi and Y. Wang, "Off-diagonal Bethe ansatz and exact solution of a topological spin ring", Phys. Rev. Lett. 111 (2013) 137201, DOI 10.1103/PhysRevLett.111.137201, arXiv:1305.7328, https://arxiv.org/abs/1305.7328
15. Z. Xin, Y. Qiao, K. Hao, J. Cao, W.-L. Yang, K. Shi and Y. Wang, "Thermodynamic limit and twisted boundary energy of the XXZ spin chain with antiperiodic boundary condition", Nucl. Phys. B 936 (2018) 501-519, DOI 10.1016/j.nuclphysb.2018.10.002, arXiv:1804.06144, https://arxiv.org/abs/1804.06144
16. X. Le, Y. Qiao, J. Cao, W.-L. Yang, K. Shi and Y. Wang, "Root patterns and energy spectra of quantum integrable systems without U(1) symmetry: antiperiodic XXZ spin chain", J. High Energy Phys. 2021 (2021) 044, DOI 10.1007/JHEP11(2021)044, arXiv:2108.08060, https://arxiv.org/abs/2108.08060
17. P. Lu, J. Cao, W.-L. Yang, I. Marquette and Y.-Z. Zhang, "T-W relation and free energy of the antiperiodic XXZ chain with $\eta=i\pi/3$ at a finite temperature", J. Phys. A 57 (2024) 445205, DOI 10.1088/1751-8121/ad85b3, arXiv:2402.04849, https://arxiv.org/abs/2402.04849