Completeness of the nested Bethe ansatz for the RTT algebra of sp(4) type (extended Yangian of sp(4))

The algebraic nested Bethe ansatz of Burdík and Navrátil (Theor. Math. Phys. 198 (2019) 1) constructs off-shell Bethe vectors and Bethe equations for the transfer matrix of the RTT algebra of sp(4) type — the extended Yangian of sp(4) in the FRT presentation — assuming an arbitrary highest-weight module specified by its vacuum eigenvalues. Its stated goal is to find eigenvectors of the transfer matrix $H(x)$ for arbitrary such modules; the further directions the paper itself posed have since been carried out: the extension of the ansatz to sp(2n), o(2n), U_q(sp(2n)) and A_n was accomplished by the same authors and, independently, by Gerrard and Regelskis, and the construction of further highest-weight representations is largely covered by the classification of finite-dimensional irreducible representations of extended Yangians of types B, C, D together with explicit families of L-operators due to Karakhanyan and Kirschner. The natural open problem arising from the construction is completeness: determine whether the on-shell Bethe vectors, at solutions of the Bethe equations and with the multiplicities and degeneracies of the transfer matrix — including contributions of configurations in which rapidities coincide — correctly accounted, span the representation space. The question is sharpened by the authors' 2025 discovery that the off-shell Bethe vectors built through the auxiliary algebra are linearly dependent, the independent vectors being expressible through gl(2)-type Bethe vectors. Completeness of the Bethe ansatz is established for type-A (GL(N)-invariant) models, but no analogous theorem exists for chains of symplectic or orthogonal type.

Background

The quantum inverse scattering method diagonalizes one-dimensional quantum integrable models through the spectral problem of a commuting family of transfer matrices. In its algebraic (Bethe ansatz) form one encodes the dynamics in a monodromy matrix $T(x)=\sum_{i,k}E^k_i\otimes T^i_k(x)$, an operator-valued matrix depending on a spectral parameter $x$ whose entries generate an associative algebra fixed by the RTT relation $R_{12}(x,y)\,T_1(x)\,T_2(y)=T_2(y)\,T_1(x)\,R_{12}(x,y),$ where $R$ is a given solution of the Yang–Baxter equation acting on two auxiliary matrix spaces. The transfer matrix $H(x)=\mathrm{tr}\,T(x)$ then commutes with itself for all $x$ and generates the conserved charges. For the rational $R$-matrix invariant under $GL(N)$ this RTT algebra is the Yangian, and the eigenvectors of $H(x)$ are produced by the algebraic Bethe ansatz introduced by Kulish and Reshetikhin: one assumes a vacuum vector $\omega$ that is an eigenvector of the diagonal generators, $T^i_i(x)\omega=\lambda_i(x)\omega$, while one triangular half of the off-diagonal generators annihilates it; creation operators from the other half, applied to $\omega$, produce trial (off-shell) Bethe vectors, and demanding that such a vector be an eigenvector yields the Bethe equations on the auxiliary rapidities. For higher-rank symmetries the construction must be iterated (nested), introducing auxiliary Bethe ansätze of lower rank.

When $R$ is instead invariant under the symplectic group $Sp(4)$ (whose Lie algebra $sp(4)$ is isomorphic to $so(5)$), the same construction defines the RTT algebra of sp(4) type: the associative algebra generated by $T^i_k(x)$, $i,k\in\{+1,+2,-1,-2\}$, obeying the RTT relation with the rational sp(4) R-matrix. This algebra is the extended Yangian $X(sp(4))$ in the Faddeev–Reshetikhin–Takhtajan (FRT) presentation; quotienting by the ideals generated by its central series yields the ordinary Yangian of $sp(4)$. The associated nested Bethe ansatz is technically harder than in type A: historically the spectrum was obtained only for special representations — by Reshetikhin for the O(n)- and Sp(2k)-symmetric magnets, via composite transfer matrices in which the auxiliary problem becomes effectively one-dimensional, and by Martins and Ramos for the rational braid-monoid lattice models based on the Birman–Murakami–Wenzl algebra (the centralizer of the $Sp(2k)$ and $SO(n)$ actions). Burdík and Navrátil removed this limitation on the representation side: assuming an arbitrary highest-weight module with vacuum eigenvalues $\lambda_i(x)$, they constructed the off-shell Bethe vectors by nesting through a $gl(2)$-type auxiliary RTT subalgebra $\tilde{\mathcal A}_2$ and derived the Bethe equations, extending the nested algebraic Bethe ansatz to the sp(4) type in general. In a 2022 companion paper they further showed that these Bethe vectors can be rewritten in a form similar to the trace formula known for the RTT-algebra of $gl(3)$ type.

On the representation-theoretic side the landscape is by now well mapped: the finite-dimensional irreducible representations of the extended Yangians of types B, C, D were classified, with explicit constructions of all fundamental representations, by Arnaudon–Molev–Ragoucy, and extensive further families of L-operators (monodromy matrices with linear or quadratic dependence on the spectral parameter, the local building blocks of spin-chain monodromies) were built from oscillator, Clifford and spinor realizations by Karakhanyan and Kirschner. What is missing is spectral. For type-A models, completeness of the Bethe ansatz — the statement that the on-shell Bethe vectors span the representation space, so that the ansatz diagonalizes the transfer matrix completely — is an established result (Baxter for the six- and eight-vertex models; Tarasov for the periodic isotropic Heisenberg chain). For chains of symplectic or orthogonal type no analogous theorem exists, and the authors' own 2025 follow-up showed that the off-shell Bethe vectors of the sp(4) construction, built through $\tilde{\mathcal A}_2$, are linearly dependent — so the dimension of their span is not given by naive counting — with the linearly independent vectors expressible through gl(2)-type Bethe vectors.

Problem Statement

Let $\mathcal A$ be the RTT algebra of sp(4) type: the associative algebra with unit generated by the coefficients $T^i_k(x)$, $i,k\in\{+1,+2,-1,-2\}$, of the monodromy matrix $T(x)=\sum_{i,k}E^k_i\otimes T^i_k(x)$, where $E^k_i$ is the $4\times 4$ matrix unit with entry $1$ in the $i$-th row and $k$-th column, subject to the RTT relation $R_{12}(x,y)\,T_1(x)\,T_2(y)=T_2(y)\,T_1(x)\,R_{12}(x,y)$ with the rational sp(4)-invariant R-matrix of the source paper, in its conventions $R(x,y)=\frac{1}{f(x,y)}\Big(I\otimes I+g(x,y)\sum_{i,k=\pm1,\pm2}E^i_k\otimes E^k_i-h(x,y)\sum_{i,k=\pm1,\pm2}\epsilon_i\,\epsilon_k\,E^i_k\otimes E^{-i}_{-k}\Big),$ $f(x,y)=\frac{x-y+1}{x-y},\qquad g(x,y)=\frac{1}{x-y},\qquad h(x,y)=\frac{1}{x-y+3},\qquad\epsilon_i=\mathrm{sgn}(i)$ (equivalently, the extended Yangian $X(sp(4))$ in the Faddeev–Reshetikhin–Takhtajan presentation; the scalar prefactor $1/f$ drops out of the RTT relation and is recorded for conformity with the source). Let $W=\mathcal A\omega$ be a highest-weight module with vacuum vector $\omega$ satisfying $T^i_i(x)\omega=\lambda_i(x)\omega$ for $i=\pm1,\pm2$ and $T^i_k(x)\omega=0$ for $i<k$ (the triangular convention of the source construction; the opposite convention is handled in parallel), and let $H(x)=\mathrm{tr}\,T(x)$ be the commuting transfer matrix. The nested algebraic Bethe ansatz of Burdík and Navrátil produces off-shell Bethe vectors from $\omega$ — products of creation generators $T^i_k$ with $i>k$ at auxiliary rapidities, nested through the gl(2)-type auxiliary RTT subalgebra $\tilde{\mathcal A}_2$ — and Bethe equations on those rapidities whose solutions give on-shell eigenvectors of $H(x)$.

The open problem is the completeness of this construction: determine for which highest-weight modules $W$ — in particular for the finite-dimensional ones, whose highest weights are classified, and for the classical representations underlying Reshetikhin's Sp(2k) magnets and the Martins–Ramos braid-monoid models — the on-shell Bethe vectors, taken over all solutions of the Bethe equations and with the multiplicities and degeneracies of the transfer matrix (including contributions of configurations in which rapidities coincide) correctly accounted, span $W$, so that the nested ansatz diagonalizes $H(x)$ completely. Prove completeness where it holds, and exhibit the obstruction where it fails (an explicit module together with an eigenvalue or multiplicity of $H(x)$ not produced by any on-shell Bethe vector). Resolving the named classes — the finite-dimensional highest-weight modules and the classical Reshetikhin and Martins–Ramos representations — resolves the task; a full classification over all highest-weight modules is not required.

The source question is posed for sp(4); the analogous statements for the RTT algebras of sp(2n) and so(2n) type, whose nested ansätze the authors constructed in 2018 (refs 6 and 8), are a natural extension and may be settled by the same answer; these analogues are optional context and are not required to resolve the task.

References

1. Č. Burdík and O. Navrátil, "Nested Bethe ansatz for the RTT algebra of sp(4) type", Theor. Math. Phys. 198 (2019) 1–16. DOI 10.1134/S004057791901001X; arXiv:1708.05633. https://doi.org/10.1134/S004057791901001X
2. N. Yu. Reshetikhin, "Integrable models of quantum one-dimensional magnets with O(n) and Sp(2k) symmetry", Theor. Math. Phys. 63 (1985) 555–569. DOI 10.1007/BF01017501. https://doi.org/10.1007/BF01017501
3. M. J. Martins and P. B. Ramos, "The algebraic Bethe ansatz for rational braid-monoid lattice models", Nucl. Phys. B 500 (1997) 579–620. DOI 10.1016/S0550-3213(97)00342-8; arXiv:hep-th/9703023. https://doi.org/10.1016/S0550-3213(97)00342-8
4. P. P. Kulish and N. Yu. Reshetikhin, "Diagonalisation of GL(N) invariant transfer matrices and quantum N-wave system (Lee model)", J. Phys. A: Math. Gen. 16 (1983) L591–L596. DOI 10.1088/0305-4470/16/16/001. https://doi.org/10.1088/0305-4470/16/16/001
5. S. Belliard and E. Ragoucy, "The nested Bethe ansatz for 'all' closed spin chains", J. Phys. A: Math. Theor. 41 (2008) 295202. DOI 10.1088/1751-8113/41/29/295202. https://doi.org/10.1088/1751-8113/41/29/295202
6. Č. Burdík and O. Navrátil, "Nested Bethe ansatz for RTT-algebras of sp(2n) type", Phys. Part. Nucl. 49 (2018) 939–942. DOI 10.1134/S106377961805009X. https://doi.org/10.1134/S106377961805009X
7. Č. Burdík and O. Navrátil, "Nested Bethe ansatz for RTT-algebra of U_q(sp(2n)) type", Springer Proc. Math. Stat. 335 (2019) 319–328. DOI 10.1007/978-981-15-7775-8_22; arXiv:1911.05460. https://doi.org/10.1007/978-981-15-7775-8_22
8. Č. Burdík and O. Navrátil, "On nested Bethe ansatz for RTT algebra of o(2n) type", Phys. At. Nucl. 81 (2018) 810–814. DOI 10.1134/S106377881806008X. https://doi.org/10.1134/S106377881806008X
9. Č. Burdík and O. Navrátil, "Nested Bethe ansatz for RTT-algebra A_n", Springer Proc. Math. Stat. 396 (2022) 325–332. DOI 10.1007/978-981-19-4751-3_27. https://doi.org/10.1007/978-981-19-4751-3_27
10. Č. Burdík and O. Navrátil, "Dependence of Bethe vectors for RTT-type algebra sp(4)", in: Lie Theory and Its Applications in Physics (Springer, 2025), pp. 263–273. DOI 10.1007/978-981-97-6453-2_20. https://doi.org/10.1007/978-981-97-6453-2_20
11. D. Arnaudon, A. Molev and E. Ragoucy, "On the R-matrix realization of Yangians and their representations", Ann. Henri Poincaré 7 (2006) 1269–1325. DOI 10.1007/s00023-006-0281-9; arXiv:math/0511481. https://doi.org/10.1007/s00023-006-0281-9
12. D. Karakhanyan and R. Kirschner, "Orthogonal and symplectic Yangians — linear and quadratic evaluations", Nucl. Phys. B 933 (2018) 14–39. DOI 10.1016/j.nuclphysb.2018.05.022; arXiv:1712.06849. https://doi.org/10.1016/j.nuclphysb.2018.05.022
13. D. Karakhanyan and R. Kirschner, "Spinorial R operator and algebraic Bethe ansatz", Nucl. Phys. B 951 (2020) 114905. DOI 10.1016/j.nuclphysb.2019.114905; arXiv:1911.08385. https://doi.org/10.1016/j.nuclphysb.2019.114905
14. D. Karakhanyan and R. Kirschner, "Representations of orthogonal and symplectic Yangians", Nucl. Phys. B 967 (2021) 115402. DOI 10.1016/j.nuclphysb.2021.115402; arXiv:2010.06513. https://doi.org/10.1016/j.nuclphysb.2021.115402
15. A. Gerrard and V. Regelskis, "Nested algebraic Bethe ansatz for deformed orthogonal and symplectic spin chains", Nucl. Phys. B 956 (2020) 115021. DOI 10.1016/j.nuclphysb.2020.115021; arXiv:1912.11497. https://doi.org/10.1016/j.nuclphysb.2020.115021
16. A. Gerrard and V. Regelskis, "Nested algebraic Bethe ansatz for orthogonal and symplectic open spin chains", Nucl. Phys. B 952 (2020) 114909. DOI 10.1016/j.nuclphysb.2019.114909; arXiv:1909.12123. https://doi.org/10.1016/j.nuclphysb.2019.114909
17. R. Frassek, "Oscillator realisations associated to the D-type Yangian: Towards the operatorial Q-system of orthogonal spin chains", Nucl. Phys. B 956 (2020) 115063. DOI 10.1016/j.nuclphysb.2020.115063; arXiv:2001.06825. https://doi.org/10.1016/j.nuclphysb.2020.115063
18. R. J. Baxter, "Completeness of the Bethe ansatz for the six and eight vertex models", J. Stat. Phys. 108 (2002) 1–48. DOI 10.1023/A:1015437118218. https://doi.org/10.1023/A:1015437118218
19. V. Tarasov, "Completeness of the Bethe ansatz for the periodic isotropic Heisenberg model", in: Ludwig Faddeev Memorial Volume, World Scientific (2018), pp. 549–566. DOI 10.1142/9789813233867_0026. https://doi.org/10.1142/9789813233867_0026
20. A. Liashyk, S. Pakuliak and E. Ragoucy, "Scalar products and norm of Bethe vectors in o_{2n+1} invariant integrable models", SciPost Phys. 19 (2025) 023. DOI 10.21468/SciPostPhys.19.1.023; arXiv:2503.01578. https://doi.org/10.21468/SciPostPhys.19.1.023
21. Č. Burdík and O. Navrátil, "Trace formula for the RTT-algebra of sp(4) type", Theor. Math. Phys. 213 (2022) 1395–1405. DOI 10.1134/S0040577922100075. https://doi.org/10.1134/S0040577922100075