Trigonometric Gaudin Models: Integrability Insights

• Trigonometric Gaudin models are integrable many-body systems characterized by trigonometric r-matrices and Lax matrices that generate commuting Hamiltonians, bridging rational and XXZ-type formulations.
• Their analysis uses methods like quantum inverse scattering, boundary reflection equations, and algebraic Bethe ansatz to derive explicit Hamiltonians and Bethe equations with multiplicative spectral dependencies.
• These models offer practical insights into dissipative quantum dynamics, moduli space parameterizations, and dualities with systems such as Calogero–Sutherland and Ruijsenaars–Schneider.

Searching arXiv for recent and foundational papers on trigonometric Gaudin models. Trigonometric Gaudin models are integrable many-body systems whose interaction kernels are governed by trigonometric or hyperbolic dependence on the spectral or inhomogeneity parameters, rather than the rational dependence of the XXX-type Gaudin model. In the standard quantum-inverse-scattering perspective, they arise as quasi-classical limits of XXZ-type spin chains, while in classical form they are encoded by Lax matrices satisfying linear or boundary-modified $r$-matrix brackets with trigonometric $r$-matrices. Across the literature, the term includes bulk $sl(2)$ and $\mathfrak{gl}_N$ Gaudin magnets, open-boundary models obtained from reflection equations, higher-rank Richardson–Gaudin systems, quasi-trigonometric non-skew-symmetric deformations, and more geometric or dual formulations parameterized by moduli spaces or related to Calogero–Sutherland and Ruijsenaars–Schneider systems (António et al., 2013).

1. Algebraic definition and basic structures

The standard trigonometric Gaudin framework is characterized by a trigonometric classical $r$-matrix and a Lax matrix whose spectral invariants generate commuting Hamiltonians. For the classical $sl(2)$ trigonometric Gaudin magnet, the Lax matrix can be written as

$$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$

with

$$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$

where the site variables satisfy the direct-sum $sl(2)$ Poisson algebra and the Casimirs $(s_j^3)^2+s_j^+s_j^- = s_j^2$ fix the dynamics on a product of two-spheres (Ragnisco et al., 2010). The corresponding trigonometric $r$0-matrix is

$r$1

and the Lax matrix obeys the linear bracket

$r$2

The generating function of commuting integrals is

$r$3

whose residues yield the Gaudin Hamiltonians (Ragnisco et al., 2010).

For $r$4, the trigonometric model can be encoded in the current algebra $r$5 with current matrix $r$6 satisfying

$r$7

where

$r$8

Under tensor products of vector evaluation representations with parameters $r$9, one has

$sl(2)$0

and the residues of $sl(2)$1 reproduce the standard quadratic trigonometric Gaudin Hamiltonians (Molev et al., 2018).

A recurrent structural theme is that trigonometric models are multiplicative rather than additive in spectral dependence. This appears in the $sl(2)$2 current relation through $sl(2)$3, in multiplicative trigonometric $sl(2)$4-matrices, and in geometric parameter spaces such as $sl(2)$5 for trigonometric Gaudin subalgebras (Ilin et al., 2024). This suggests a precise sense in which trigonometric Gaudin theory interpolates between additive rational-current formulations and multiplicative XXZ-type constructions.

2. Classical and quantum integrability

A standard conceptual statement is that rational Gaudin models are quasi-classical limits of XXX-type chains, whereas trigonometric Gaudin models are quasi-classical limits of XXZ-type chains (António et al., 2013). In the periodic case, evaluating the transfer matrix at inhomogeneities and expanding in the quasi-classical parameter produces commuting Gaudin Hamiltonians. For the open XXZ chain, the same mechanism is implemented through the Sklyanin double-row transfer matrix

$sl(2)$6

whose quasi-classical expansion at $sl(2)$7 has the form

$sl(2)$8

yielding commuting open-boundary trigonometric Gaudin Hamiltonians (António et al., 2013).

In the boundary-free $sl(2)$9 case, the Hamiltonians extracted from

$\mathfrak{gl}_N$0

are

$\mathfrak{gl}_N$1

with $\mathfrak{gl}_N$2 and $\mathfrak{gl}_N$3 (Ragnisco et al., 2010). The model also admits a rationalized description via $\mathfrak{gl}_N$4, under which

$\mathfrak{gl}_N$5

so the trigonometric Gaudin magnet becomes a rational $\mathfrak{gl}_N$6-pole object with a $\mathfrak{gl}_N$7 reflection constraint (Ragnisco et al., 2010).

In the quantum $\mathfrak{gl}_N$8 setting, higher commuting Hamiltonians arise from the $\mathfrak{gl}_N$9 limit of XXZ Bethe subalgebra elements in $r$0. The generating series

$r$1

defines explicit higher Hamiltonians whose coefficients generate a commutative subalgebra of $r$2 (Molev et al., 2018). This is the trigonometric analogue of Talalaev-type higher Gaudin Hamiltonians and shows that the trigonometric theory is not restricted to quadratic Hamiltonians.

3. Boundary terms and reflection-algebra formulations

Open trigonometric Gaudin models are constructed through reflection equations. For the non-symmetric XXZ $r$3-matrix, the left boundary matrix $r$4 satisfies

$r$5

while the right boundary matrix $r$6 is related by

$r$7

With suitable normalization, the quasi-classical expansion of the open transfer matrix yields

$r$8

where the three terms are, respectively, a local boundary term, the usual bulk interaction, and a reflected interaction depending on $r$9 (António et al., 2013).

This reflected term is the hallmark of open trigonometric Gaudin systems. Compared with the bulk Hamiltonian, open boundaries add “image-spin” interactions dressed by the boundary matrix. The same paper also shows that a classical-reflection-equation construction based on

$sl(2)$0

does not reproduce the same Hamiltonians as the quasi-classical limit of the open XXZ transfer matrix, making the notion of “open trigonometric Gaudin model” construction-dependent (António et al., 2013).

For triangular boundary conditions, the trigonometric $sl(2)$1 model admits a full algebraic Bethe ansatz. The left and right reflection matrices are specialized to upper-triangular form, and in the Gaudin limit the condition

$sl(2)$2

ties the two boundaries together (Manojlović et al., 2017). The corresponding generating function is

$sl(2)$3

and the residues at $sl(2)$4 give the trigonometric Gaudin Hamiltonians with boundary terms (Manojlović et al., 2017). Appropriate Bethe vectors yield a simple off-shell action of $sl(2)$5, with Bethe equations

$sl(2)$6

which are the boundary trigonometric analogue of the rational Gaudin equations (Manojlović et al., 2017).

A broader generalization is the classical $sl(2)$7-reflection equation. Given a solution $sl(2)$8 of that equation, one defines

$sl(2)$9

which again satisfies the classical Yang–Baxter equation, though it is generally non-skew-symmetric (Caudrelier et al., 2018). This produces new boundary, folded, or cyclotomic Gaudin-type models and includes explicit trigonometric $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$0 and $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$1 solutions with Möbius transformations of the spectral parameter. The paper does not write explicit trigonometric Gaudin Hamiltonians in full detail, but it gives the exact framework for constructing them through residues of $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$2 (Caudrelier et al., 2018).

4. Separation of variables, higher structures, and deformations

A notable extension of the standard trigonometric framework is the quasi-trigonometric $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$3 Gaudin model governed by a non-skew-symmetric, non-dynamical classical $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$4-matrix

$$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$5

with $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$6 the distinguished one-parameter subfamily supporting two full separation-of-variables constructions (Skrypnyk, 2021). At $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$7 one recovers the usual skew-symmetric trigonometric $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$8-matrix, but for other values the model is genuinely non-skew-symmetric and satisfies the generalized classical Yang–Baxter equation (Skrypnyk, 2021).

For $$L(\lambda)= \begin{pmatrix} A(\lambda) & B(\lambda)\ C(\lambda) & -A(\lambda) \end{pmatrix},$$9 sites, the quasi-trigonometric Gaudin Lax matrix is

$$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$0

and the commuting Hamiltonians are the residues

$$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$1

Relative to the standard trigonometric Hamiltonians, the deformation adds the non-skew-symmetric $$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$2-matrix parameters $$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$3, on-site quadratic terms, and external magnetic-field terms $$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$4 (Skrypnyk, 2021).

The same model possesses an additional geometric integral

$$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$5

central in the separation-of-variables construction. For $$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$6, two one-parameter families of separating functions $$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$7 generate complete canonical separated variables when $$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$8, and for generic parameters the separation curves are not the ordinary spectral curve

$$A(\lambda)=\sum_{j=1}^{N}\cot(\lambda-\lambda_j)\,s_j^3,\qquad B(\lambda)=\sum_{j=1}^{N}\frac{s_j^-}{\sin(\lambda-\lambda_j)},\qquad C(\lambda)=\sum_{j=1}^{N}\frac{s_j^+}{\sin(\lambda-\lambda_j)},$$9

but rather $sl(2)$0-shifted spectral curves (Skrypnyk, 2021). This is one of the clearest demonstrations that trigonometric Gaudin theory naturally extends beyond standard Sklyanin-type separated coordinates.

The classical trigonometric Gaudin magnet also admits explicit Bäcklund transformations, interpreted as exact integrable time-discretizations. The map is defined by

$sl(2)$1

with a Darboux matrix preserving the reflection symmetry $sl(2)$2. In trigonometric variables the Darboux matrix takes the XXZ-type form

$sl(2)$3

and the interpolating Hamiltonian is

$sl(2)$4

The rational Gaudin Bäcklund transformation is recovered by a small-angle limit (Ragnisco et al., 2010).

A more indirect structural development is the bi-Hamiltonian approach to Gaudin algebras via compatible Lie brackets on quotient current algebras $sl(2)$5. This framework explicitly captures rational Gaudin algebras and certain limits, but it does not explicitly construct standard trigonometric Gaudin Hamiltonians; its relevance to trigonometric models is therefore structural and suggestive rather than direct (Yakimova, 2021).

5. Dualities, moduli spaces, and generalized parameter spaces

The trigonometric quantum-classical duality relates the twisted inhomogeneous XXZ chain to the classical trigonometric Ruijsenaars–Schneider model, and in its quasi-classical/nonrelativistic limit it produces a duality between the quantum trigonometric Gaudin model and the classical Calogero–Sutherland system (Beketov et al., 2015). On the quantum side, the XXZ transfer matrix

$sl(2)$6

is expanded with

$sl(2)$7

and the trigonometric Gaudin Hamiltonians appear as residues

$sl(2)$8

with explicit formula

$sl(2)$9

(Beketov et al., 2015). The dual classical spectrum forms strings centered at the twist parameters $(s_j^3)^2+s_j^+s_j^- = s_j^2$0, a trigonometric splitting phenomenon absent in the rational case.

A different geometric picture is provided by the family of trigonometric Gaudin subalgebras

$(s_j^3)^2+s_j^+s_j^- = s_j^2$1

defined for arbitrary complex simple $(s_j^3)^2+s_j^+s_j^- = s_j^2$2 as images of homogeneous Gaudin algebras under a quantum Hamiltonian reduction map $(s_j^3)^2+s_j^+s_j^- = s_j^2$3 (Ilin et al., 2024). Their quadratic Hamiltonians are

$(s_j^3)^2+s_j^+s_j^- = s_j^2$4

and the family is parameterized by $(s_j^3)^2+s_j^+s_j^- = s_j^2$5, compactifying over $(s_j^3)^2+s_j^+s_j^- = s_j^2$6 (Ilin et al., 2024). A central degeneration formula is

$(s_j^3)^2+s_j^+s_j^- = s_j^2$7

showing that rational inhomogeneous Gaudin subalgebras arise as limits of trigonometric ones (Ilin et al., 2024). The same work further identifies real loci with simple-spectrum regimes and interprets the resulting monodromy as an action of affine cactus-type groups.

For relativistic $(s_j^3)^2+s_j^+s_j^- = s_j^2$8 Gaudin-type systems, the elliptic theory admits a universal $(s_j^3)^2+s_j^+s_j^- = s_j^2$9-matrix formulation

$r$00

and the authors state explicitly that this description also provides trigonometric and rational analogues whenever the chosen $r$01-matrix satisfies the associative Yang–Baxter equation and the required additional properties (Trunina et al., 2022). The paper does not work out a separate trigonometric function-theoretic model in full detail, but it places trigonometric relativistic Gaudin analogues inside a uniform $r$02-matrix classification.

Some duality results with irregular singularities are likewise structurally relevant. They concern rational Gaudin algebras with Takiff data and do not directly construct trigonometric Hamiltonians, but they supply determinant/Berezinian duality mechanisms, oscillator realizations, and classical spectral identities that are closely comparable to the kinds of spectral correspondences appearing in trigonometric Gaudin theory (Cheong et al., 1 Jul 2025). This suggests a broader landscape in which trigonometric, cyclotomic, and irregular-rational models should be viewed as adjacent rather than isolated constructions.

6. Physical realizations and dynamical phenomena

Trigonometric Richardson–Gaudin models appear in a concrete open-system realization: a system of $r$03-level atoms coupled to a Markovian reservoir can be mapped, after vectorization, to a Liouvillian that is a linear combination of commuting trigonometric $r$04 Richardson–Gaudin charges (Lerma-Hernández et al., 2020). The trigonometric kernels are

$r$05

and in the two-copy realization the Liouvillian spectrum is determined by nested Bethe equations of trigonometric Gaudin type (Lerma-Hernández et al., 2020). For $r$06 this recovers the exact solution of the collective dissipative spin model, while for $r$07 the paper studies the steady state and dissipative gaps for finite systems and in the thermodynamic limit (Lerma-Hernández et al., 2020). This places trigonometric Gaudin theory directly inside non-Hermitian and dissipative quantum dynamics.

On the classical side, the $r$08 rational and trigonometric Gaudin models can be viewed as generalized coupled angular momenta systems on

$r$09

with momentum map

$r$10

The trigonometric subfamily is $r$11, with Hamiltonian

$r$12

(Henriksen, 2024). The rank-zero singularities $r$13 and $r$14 undergo Hamiltonian Hopf bifurcations at

$r$15

and the normal form computed up to sixth order determines when the bifurcation is non-degenerate, supercritical, or subcritical (Henriksen, 2024). In the subcritical case a flap appears in the image of the momentum map, and the figures in that work show additional global bifurcations involving pleats and cusp collisions (Henriksen, 2024). This suggests that even the low-rank trigonometric Gaudin model has a significantly richer singularity theory than its integrability alone would indicate.

Taken together, these developments show that “trigonometric Gaudin model” now denotes not a single formula but a coherent domain of integrable structures: XXZ quasi-classical limits, multiplicative current algebras, boundary reflection systems, higher Hamiltonian algebras, non-skew-symmetric quasi-trigonometric deformations, moduli-space families, quantum-classical dualities, and concrete realizations in dissipative and symplectic dynamics. A plausible implication is that future work will continue to treat trigonometric Gaudin systems less as isolated deformations of rational models and more as organizing centers connecting boundary integrability, geometric representation theory, and generalized $r$16-matrix frameworks (Ilin et al., 2024).