Rigorous completeness of the boundary-parameter Q-system for the open Heisenberg spin chain: the open-chain analogue of the Mukhin–Tarasov–Varchenko theorem

For the open spin-1/2 Heisenberg chain with diagonal boundary magnetic fields, Nepomechie (2020) constructed two inequivalent boundary-deformed rank-1 Q-systems and a boundary-deformed discrete Wronskian relation tying the fundamental Q-function Q(u) to its dual P(u), and conjectured — by analogy with the Mukhin–Tarasov–Varchenko theorem for periodic chains — that polynomiality of P(u) is equivalent to admissibility of the Bethe roots, i.e. to the Bethe ansatz being complete. The problem is to settle this conjecture rigorously, including the degenerate special manifold of boundary parameters alpha-beta=±1 where the Wronskian vanishes and two factorization conjectures (equations (A.4) and (A.9) of that paper) remain unproven, and to derive the boundary-deformed Wronskian relation and the two Q-system recursions at the operator level from explicit Q-operators.

Background

An open quantum spin chain is one whose ends carry boundary interactions instead of being joined periodically. For the spin-1/2 Heisenberg (XXX) chain, the integrable such boundaries are diagonal boundary magnetic fields, introduced by Sklyanin (1988): one builds a double-row transfer matrix — the trace over an auxiliary site of a product of $R$-matrices (the fundamental $2\times 2$ solution of the Yang–Baxter equation that encodes the model's integrability) together with two $2\times 2$ boundary reflection ($K$-)matrices, which for the diagonal case are parametrized by two complex boundary parameters $\alpha$ and $\beta$ — yielding a commuting family of conserved quantities that includes the Hamiltonian. The algebraic Bethe ansatz then reduces the spectral problem to algebraic Bethe equations for complex Bethe roots $u_1,\ldots,u_M$, where $M$ is the number of down spins relative to the ferromagnetic reference state and $M=0,\ldots,N$ for a chain of $N$ sites. The boundary parameters enter the Bethe equations only through the ratio of the two functions $f(u)=(u-i\alpha)(u+i\beta)$ and $g(u)=f(-u)$.

Not every solution of the Bethe equations corresponds to an actual eigenvector of the transfer matrix: solutions with repeated roots, or with roots at $0$ or $\pm i/2$, are spurious. Solutions that do correspond to genuine eigenstates are called admissible, and deciding exactly which solutions are admissible — the completeness problem for the Bethe ansatz — is a foundational rigor question of the field.

For the periodic chain this question has a complete answer in terms of functional relations. Baxter's TQ equation relates the transfer-matrix eigenvalue $T(u)$ to a polynomial Q-function $Q(u)=\prod_{k=1}^{M}(u-u_k)(u+u_k)$; the TQ equation admits a second (dual) solution $P(u)$, and the two are tied by a discrete Wronskian (Casoratian) identity, where $F^{\pm}(u)=F(u\pm i/2)$. Pronko and Stroganov (1999) observed that whether the dual solution $P(u)$ is a polynomial distinguishes physical from unphysical solutions, and Mukhin, Tarasov and Varchenko (2009), through their study of the Bethe algebra (the commutative algebra generated by the transfer matrix and the Q-operator) and its simple spectrum, together with Tarasov (2018), proved for the periodic isotropic chain that polynomiality of $P(u)$ is exactly equivalent to admissibility — the Bethe ansatz is complete; Chernyak, Leurent and Volin (2022) later extended this completeness of Wronskian Bethe equations to rational supersymmetric gl(m|n) periodic chains.

These functional relations were later organized into the rational Q-system, a tower of Q-functions $Q_{a,n}(u)$ obeying bilinear QQ-recursions with a built-in difference operator $D$ (see Bajnok, Granet, Jacobsen and Nepomechie (2020) for the generalized construction): all the Q-functions are polynomials precisely on physical solutions, which makes the Q-system an efficient complete-spectrum solver. Nepomechie (2020) extended this framework to the open chain with boundary parameters, and two new features arise. First, the TQ equation and its dual are no longer the same equation, and the Wronskian is deformed by the boundary factors $f$ and $g$:

$g\,P^{+}Q^{-}-f\,P^{-}Q^{+}=u\,Q_{0,0},\qquad Q_{0,0}(u)=u^{2N}.$

Second, there is not one Q-system but two inequivalent ones: the QQ-recursions may deform either the $Q_{1,n}$ tower (by $u\,Q_{1,n}\propto f^{[-(n-1)]}Q_{1,n-1}^{+}-g^{[n-1]}Q_{1,n-1}^{-}$, with $F^{[n]}(u)=F(u+n i/2)$) or the $Q_{0,n}$ tower by the same boundary factors; both are solved by $Q_{1,n}\propto D^{n}Q$, and the two systems deform the difference-operator recursions in complementary ways — the recursion defining $D^{n}Q$ is boundary-deformed in the first system, the one defining $D^{n}P$ in the second. Numerics for small $N$ support the claim that polynomiality of $P(u)$ — equivalently, of all the Q-functions of either system — holds exactly on admissible solutions, but no proof is known. The same paper leaves two further rigor gaps. On the special manifold of boundary parameters $\alpha-\beta=+1$ one has $f^{+}=g^{-}$: the Bethe equations degenerate, each Bethe root being either a special root $u=\pm i(\alpha-\tfrac{1}{2})$ or a root obeying the SU(2)-invariant chain's Bethe equations, and the Wronskian vanishes on class-I configurations — those containing such a special root (class-II configurations contain none). On $\alpha-\beta=-1$ one has $f^{-}=g^{+}$, and the dual Bethe equations degenerate in the analogous way, with dual special root $\pm i(\alpha+\tfrac{1}{2})$. The factorizations $Q(u)=f^{+}(u)P(u)$ and $P(u)=f^{-}(u)Q(u)$ are only conjectured (equations (A.4) and (A.9) of that paper), with no general proof and no complete classification of class-I versus class-II configurations for all $N$. Finally, at the operator level: Q-operators — operators on the chain's Hilbert space, constructed as traces of monodromy matrices over infinite-dimensional auxiliary spaces, whose eigenvalues are the Q-functions — were built for the open chain with diagonal boundaries by Frassek and Szécsényi (2015), but the boundary-deformed Wronskian relation and the two Q-system recursions (including the deformed $D^{n}$ recursion) have been established only at the level of eigenvalues, not as operator identities. The Q-systems and the Wronskian relation have trigonometric analogues for the open XXZ chain, treated in parallel in the source, where the Q-functions are Laurent polynomials in $t=e^{u}$ and the boundary parameters enter the corresponding q-deformed factors.

Problem Statement

Establish the rigorous completeness theory for the boundary-parameter Q-system of the open spin-1/2 Heisenberg chain with diagonal boundary magnetic fields — the open-chain analogue of the Mukhin–Tarasov–Varchenko theorem. The common objective is to put the Q-system method for this class of open chains on the same rigorous footing as the periodic chain: a proven completeness criterion, its behavior on the degenerate boundary-parameter manifold, and an operator-level realization of the same functional relations. Specifically, for the open XXX (rational) chain of $N$ sites with diagonal boundary parameters $\alpha,\beta$ — and, for the completeness criterion of component (1), correspondingly also for the open XXZ (trigonometric) chain with Laurent-polynomial conditions in $t=e^{u}$, whose Q-systems and Wronskian relation the source treats in parallel:

(1) Completeness/admissibility equivalence (core). Prove, or refute, Nepomechie's conjecture: a solution $\{u_1,\ldots,u_M\}$ of the Bethe equations

$\frac{g(u_j-\tfrac{i}{2})}{f(u_j+\tfrac{i}{2})}\left(\frac{u_j+\tfrac{i}{2}}{u_j-\tfrac{i}{2}}\right)^{2N}=\prod_{k\ne j}^{M}\frac{(u_j-u_k+i)(u_j+u_k+i)}{(u_j-u_k-i)(u_j+u_k-i)},\qquad f(u)=(u-i\alpha)(u+i\beta),\; g(u)=f(-u),$

is admissible (i.e. the roots of a Q-function $Q(u)=\prod_k(u-u_k)(u+u_k)$ belonging to a genuine transfer-matrix eigenvalue and eigenvector) if and only if the dual Q-function $P(u)$ defined through the boundary-deformed Wronskian relation

$g\,P^{+}Q^{-}-f\,P^{-}Q^{+}=u\,Q_{0,0},\qquad Q_{0,0}(u)=u^{2N},\qquad F^{\pm}(u)=F(u\pm i/2),$

is a polynomial in $u$; equivalently, that the two boundary-deformed Q-systems of Nepomechie (the one deforming the $Q_{1,n}$ recursion and the one deforming the $Q_{0,n}$ recursion, each with its $D^{n}$-recursion) generate polynomial Q-functions exactly on the admissible solutions, so that each Q-system produces the complete transfer-matrix spectrum and nothing else. The statement must hold for all chain lengths $N$, all $M=0,\ldots,N$, and generic boundary parameters, and must settle the degenerate cases.

(2) Special manifold. On the manifold $\alpha-\beta=+1$ one has $f^{+}=g^{-}$, and the Wronskian vanishes on class-I configurations — those whose Bethe roots contain a special root $u=\pm i(\alpha-\tfrac{1}{2})$; class-II configurations contain no special root, and their remaining roots obey the SU(2)-invariant chain's Bethe equations. On the manifold $\alpha-\beta=-1$ one has $f^{-}=g^{+}$, and the dual Bethe equations degenerate in the analogous way, with dual special root $\tilde{u}=\pm i(\alpha+\tfrac{1}{2})$ (class-I dual configurations contain one). Prove (or refute) the factorization conjectures $Q(u)=f^{+}(u)P(u)$ (eq. (A.4), for class-I configurations at $\alpha-\beta=+1$) and $P(u)=f^{-}(u)Q(u)$ (eq. (A.9), for class-I configurations at $\alpha-\beta=-1$).

(3) Operator-level realization. For the open chain with diagonal boundary fields, derive the boundary-deformed Wronskian relation and the two Q-system recursions (including the deformed $D^{n}$ recursion) as operator identities for Q-operators whose eigenvalues are $Q(u)$ and $P(u)$ (such operators have already been constructed in the rational case by Frassek and Szécsényi) — relations so far established only at the level of eigenvalues.

Components (2) and (3) are integral parts of the same objective rather than independent add-ons: (2) is forced by (1), since the completeness criterion degenerates exactly on this manifold and a theorem excluding it would be incomplete; (3) is the operator-level counterpart of the same Wronskian and QQ relations that underlie the completeness criterion. The further extensions raised in the source paper — non-diagonal K-matrices and rank greater than one — together with trigonometric (XXZ) analogues of components (2) and (3), which the source poses only within the rational setup, are open generalization directions beyond this problem's completion boundary.

References

1. R. I. Nepomechie, "Q-systems with boundary parameters", J. Phys. A: Math. Theor. 53 (2020) 294001. DOI: 10.1088/1751-8121/ab9386; arXiv:1912.12702. https://doi.org/10.1088/1751-8121/ab9386
2. E. Mukhin, V. Tarasov, A. Varchenko, "Bethe algebra of homogeneous XXX Heisenberg model has simple spectrum", Commun. Math. Phys. 288 (2009) 1–42. DOI: 10.1007/s00220-009-0733-4; arXiv:0706.0688. https://doi.org/10.1007/s00220-009-0733-4
3. V. Tarasov, "Completeness of the Bethe ansatz for the periodic isotropic Heisenberg model", Rev. Math. Phys. 30 (2018) 1840018. DOI: 10.1142/S0129055X18400184. https://doi.org/10.1142/S0129055X18400184
4. E. K. Sklyanin, "Boundary conditions for integrable quantum systems", J. Phys. A: Math. Gen. 21 (1988) 2375–2389. DOI: 10.1088/0305-4470/21/10/015. https://doi.org/10.1088/0305-4470/21/10/015
5. G. P. Pronko, Yu. G. Stroganov, "Bethe equations 'on the wrong side of the equator'", J. Phys. A: Math. Gen. 32 (1999) 2333–2340. DOI: 10.1088/0305-4470/32/12/007; arXiv:hep-th/9808153. https://doi.org/10.1088/0305-4470/32/12/007
6. Z. Bajnok, E. Granet, J. L. Jacobsen, R. I. Nepomechie, "On generalized Q-systems", J. High Energy Phys. 2020 (2020) 177. DOI: 10.1007/JHEP03(2020)177; arXiv:1910.07805. https://doi.org/10.1007/JHEP03(2020)177
7. R. Frassek, I. M. Szécsényi, "Q-operators for the open Heisenberg spin chain", Nucl. Phys. B 901 (2015) 229–248. DOI: 10.1016/j.nuclphysb.2015.10.010; arXiv:1509.04867. https://doi.org/10.1016/j.nuclphysb.2015.10.010
8. R. Frassek, I. M. Szécsényi, "Algebraic Bethe ansatz for Q-operators of the open XXX Heisenberg chain with arbitrary spin", J. Phys. A: Math. Theor. 55 (2022) 505201. DOI: 10.1088/1751-8121/aca5d3; arXiv:2208.02197. https://doi.org/10.1088/1751-8121/aca5d3
9. Z. Tsuboi, "Universal Baxter TQ-relations for open boundary quantum integrable systems", Nucl. Phys. B 963 (2021) 115286. DOI: 10.1016/j.nuclphysb.2020.115286; arXiv:2010.09675. https://doi.org/10.1016/j.nuclphysb.2020.115286
10. Y. Jiang, Y. Miao, "Spectrum-preserving deformations of integrable spin chains with open boundaries", Phys. Rev. B 110 (2024) L121406. DOI: 10.1103/PhysRevB.110.L121406; arXiv:2406.17171. https://doi.org/10.1103/PhysRevB.110.L121406
11. Y.-J. He, J. Hou, Y.-C. Liu, Z.-X. Tan, "Spin-s Q-systems: Twist and open boundaries", SciPost Phys. Core 8 (2025) 057. DOI: 10.21468/SciPostPhysCore.8.3.057; arXiv:2502.15636. https://doi.org/10.21468/SciPostPhysCore.8.3.057
12. Y. Jiang, Y.-C. Liu, Y. Miao, Z.-X. Tan, "Rational Q-systems for integrable spin chains without U(1) symmetry", SciPost Phys. 20 (2026) 175. DOI: 10.21468/SciPostPhys.20.6.175; arXiv:2512.01551. https://doi.org/10.21468/SciPostPhys.20.6.175
13. R. I. Nepomechie, "Wronskian-type formula for inhomogeneous TQ-equations", Theor. Math. Phys. 204 (2020) 1195–1200. DOI: 10.1134/S004057792009007X; arXiv:2005.08326. https://doi.org/10.1134/S004057792009007X
14. R. I. Nepomechie, "The A_m^(1) Q-system", Mod. Phys. Lett. A 35 (2020) 2050260. DOI: 10.1142/S0217732320502600; arXiv:2003.06823. https://doi.org/10.1142/S0217732320502600
15. D. Chernyak, S. Leurent, D. Volin, "Completeness of Wronskian Bethe Equations for Rational gl(m|n) Spin Chains", Commun. Math. Phys. 391 (2022) 969–1045. DOI: 10.1007/s00220-021-04275-9; arXiv:2004.02865. https://doi.org/10.1007/s00220-021-04275-9