Integrability of the inhomogeneous central-spin XX model for arbitrary central-spin quantum number

The inhomogeneous central-spin XX Hamiltonian $H=\omega_0 S_0^z + S_0^+ G^- + S_0^- G^+$, with $G^\pm=\sum_{j=1}^L g_j S_j^\pm$, describes a central spin of quantum number $s_0$ coupled by flip-flop (XX) interactions of arbitrary strengths $g_j$ to a bath of $L$ surrounding spins, as realized for electron spins in quantum dots and nitrogen-vacancy centers in diamond. Integrability — an extensive set of mutually commuting conserved charges enabling exact Bethe-ansatz eigenstates — has been proven only for $s_0=1/2$ and $s_0=1$. Whether the model is integrable for arbitrary $s_0$ (first open case $s_0=3/2$), and how the known conserved charges embed into the Richardson-Gaudin/Yang-Baxter framework of algebraic integrability, remains open. The conjecture is supported by Poissonian level statistics at $s_0=3/2$ and by an integrable large-detuning effective Hamiltonian, but follow-up work has shown the classical limit to be nonintegrable away from zero energy, leaving the question genuinely undecided.

Background

Central-spin models describe a single well-controlled quantum degree of freedom coupled to a mesoscopic environment of $L$ surrounding spins. They model the hyperfine interaction between an electron spin in a quantum dot and the nuclear-spin bath, nitrogen-vacancy (NV) defect centers in diamond coupled to surrounding nuclei, and — with a cavity mode playing the central role — ensembles of two-level emitters in cavity quantum electrodynamics. A central-spin Hamiltonian is called integrable if it possesses an extensive set of mutually commuting conserved quantities (conserved charges), which typically allows all eigenstates to be constructed exactly by a Bethe ansatz, i.e. by parametrizing eigenstates through algebraic variables (rapidities) satisfying coupled algebraic equations. With isotropic XXX couplings $\vec S_0\cdot\vec S_j$ between central spin and bath, the model belongs to the Richardson-Gaudin class of exactly solvable long-range spin models (which also contains the reduced BCS pairing Hamiltonian) and is integrable for any central-spin value $s_0$. The experimentally important XX case is different: the flip-flop terms $S_0^+S_j^-+S_0^-S_j^+$ arise from dipolar couplings in NMR, NV centers and certain quantum dots, and for generic (inhomogeneous) couplings $g_j$ the model lies outside the standard Richardson-Gaudin family. Since the total magnetization is conserved, a rotating-frame transformation leaves the Hamiltonian in the form $H=\omega_0 S_0^z + S_0^- G^+ + S_0^+ G^-$ with $G^\pm=\sum_{j=1}^L g_j S_j^\pm$ and detuning $\omega_0$; the bath spins $s_j$ and couplings $g_j$ are arbitrary. Villazon, Chandran and Claeys (2020) proved integrability for a spin-1/2 central spin, with two classes of eigenstates: bright states, in which the central spin is entangled with the bath, and dark states, in which it is fully polarized along $z$ and unentangled. Tang, Long, Polkovnikov, Chandran and Claeys (2023) proved integrability for a spin-1 central spin: for each bath spin they constructed a conserved charge $\tilde Q_j$ built from Richardson-Gaudin-type charges $Q_j$ of the operator $G^+G^-+G^-G^+$, the projector $P_0 = 1-(S_0^z)^2$ onto the $m_0=0$ level of the central spin, and terms of up to four spin operators, and they obtained all eigenstates from Bethe-type constructions (on resonance the bright states split further into double and triple classes, with observable consequences for quench dynamics). Beyond $s_0=1$, exact results exist only in special limits: for homogeneous couplings the model is solvable through collective spin operators for any $s_0$, and for inhomogeneous couplings a subset of eigenstates (separable states and single-spin-flip bright states) is known via an operator-product ansatz. Tang et al. nevertheless conjectured that the inhomogeneous model is integrable for every $s_0$, supported by three pieces of evidence: Poissonian level-spacing-ratio statistics in the $s_0=3/2$ model (integrable systems show uncorrelated, Poisson-distributed level spacings, chaotic ones Wigner-Dyson statistics); the Schrieffer-Wolff effective Hamiltonian at large detuning $\omega_0$ (a Schrieffer-Wolff transformation is a perturbative unitary rotation eliminating far off-resonant flip-flop terms to obtain an effective Hamiltonian) being integrable through second order in the interaction strength for any $s_0$; and numerics in the classical limit $s_0,s_j\to\infty$ showing residual memory of the initial state. The same paper leaves open how the charges $\tilde Q_j$ are to be incorporated into the general framework of Richardson-Gaudin integrability — generalized Gaudin algebras (algebraic structures whose commuting elements generate solvable spin models), Yang-Baxter-based constructions such as the algebraic Bethe ansatz, or generalized classical Yang-Baxter equations with a modified Bethe ansatz — noting that the charges are closely related to, but not contained in, the XXZ Richardson-Gaudin family. The question is directly tied to the inhomogeneous Tavis-Cummings model of cavity QED (a single cavity mode coupled to two-level emitters with couplings $g_j$), because in the fixed-excitation sector $N$ that model is equivalent to the central-spin XX problem with effective central spin $s_0=N/2$.

Problem Statement

Determine whether the inhomogeneous central-spin XX Hamiltonian $H = \omega_0 S_0^z + S_0^+ G^- + S_0^- G^+, \qquad G^\pm = \sum_{j=1}^L g_j S_j^\pm,$ with arbitrary bath size $L$, arbitrary inhomogeneous couplings $g_j$ and arbitrary bath spins $s_j$, is integrable for arbitrary central-spin quantum number $s_0$ — the first open case being $s_0=3/2$ — in the same operational sense as the proven cases $s_0=1/2$ and $s_0=1$: either construct an extensive set of mutually commuting conserved charges (one per bath spin, generalizing the spin-1 charges $\tilde Q_j$), commuting with $H$ and functionally independent, together with the exact eigenstates they enable, or establish that no such set exists for some $s_0$. The source poses companion open questions that a resolution is expected to inform but that can be settled separately and are facets of the same program rather than additional required deliverables: how the conserved charges of the solvable cases are to be incorporated into the general framework of Richardson-Gaudin integrability — a generalized Gaudin algebra, an $r$-matrix/Yang-Baxter construction with its (possibly modified) Bethe ansatz, or another algebraic structure such as a supersymmetric charge algebra — given that the spin-1 charges $\tilde Q_j$ are closely related to, but not contained in, the XXZ Richardson-Gaudin family; whether the Schrieffer-Wolff effective Hamiltonian remains integrable beyond second order in the interaction strength (the two quartic terms not known to be integrable cancel at the computed order, but the integrability of the full correction series is unsettled); and how the conserved charges quantitatively constrain late-time nonequilibrium observables after quenches. Through the excitation-sector equivalence, integrability for arbitrary $s_0$ with spin-1/2 bath spins implies the integrability of the inhomogeneous Tavis-Cummings model (two-level emitters) at arbitrary excitation number, currently proven only in the two-excitation sector; the source also conjectures, as a consequence of integrability of $H$ at arbitrary $s_0$, integrability of a generalized Tavis-Cummings limit that retains arbitrary bath spins $s_j$.

References

1. L.-H. Tang, D. M. Long, A. Polkovnikov, A. Chandran, P. W. Claeys, "Integrability and quench dynamics in the spin-1 central spin XX model", SciPost Phys. 15, 030 (2023), doi:10.21468/SciPostPhys.15.1.030, arXiv:2212.04477, https://arxiv.org/abs/2212.04477
2. T. Villazon, A. Chandran, P. W. Claeys, "Integrability and dark states in an anisotropic central spin model", Phys. Rev. Research 2, 032052 (2020), doi:10.1103/PhysRevResearch.2.032052, arXiv:2001.10008, https://arxiv.org/abs/2001.10008
3. E. De Nadai, N. Maestracci, A. Faribault, "Integrability and dark states of the XX spin-1 central spin model in a transverse field", Phys. Rev. B 110, 205427 (2024), doi:10.1103/PhysRevB.110.205427, arXiv:2407.03029, https://arxiv.org/abs/2407.03029
4. A. Wierzchucka, F. Piazza, P. W. Claeys, "Integrability, multifractality, and two-photon dynamics in disordered Tavis-Cummings models", Phys. Rev. A 109, 033716 (2024), doi:10.1103/PhysRevA.109.033716, arXiv:2312.03833, https://arxiv.org/abs/2312.03833
5. J. van Tonder, J. Links, "Supersymmetry and integrability for a class of XY central spin models", J. Phys. A: Math. Theor. 56, 47LT01 (2023), doi:10.1088/1751-8121/ad043a, arXiv:2303.17837, https://arxiv.org/abs/2303.17837
6. J. van Tonder, J. Links, "Dark states in an integrable XY central spin model", J. Stat. Mech. (2025) 063104, doi:10.1088/1742-5468/adde3d, arXiv:2502.04651, https://arxiv.org/abs/2502.04651
7. H. Kim, R. Schäfer, D. M. Long, A. Polkovnikov, A. Chandran, "Confined and deconfined chaos in classical spin systems", arXiv:2507.07168 (2025), https://arxiv.org/abs/2507.07168
8. P. S. Isaac, J. Links, I. Lukyanenko, J. L. Werry, "Explicit conserved operators for a class of integrable bosonic networks from the classical Yang-Baxter equation", Journal of Geometry and Physics 225, 105837 (2026), doi:10.1016/j.geomphys.2026.105837, arXiv:2507.00483, https://arxiv.org/abs/2507.00483
9. A. Foerster, J. Links, "Superintegrable bosonic star networks", arXiv:2509.05590 (2025), https://arxiv.org/abs/2509.05590
10. N. Wu, X.-W. Guan, J. Links, "Separable and entangled states in the high-spin XX central spin model", Phys. Rev. B 101, 155145 (2020), doi:10.1103/PhysRevB.101.155145, arXiv:2002.04774, https://arxiv.org/abs/2002.04774
11. R. I. Nepomechie, X.-W. Guan, "The spin-s homogeneous central spin model: exact spectrum and dynamics", J. Stat. Mech. (2018) 103104, doi:10.1088/1742-5468/aae2d9, arXiv:1810.03937, https://arxiv.org/abs/1810.03937
12. C. Dimo, A. Faribault, "Strong-coupling emergence of dark states in XX central spin models", Phys. Rev. B 105, L121404 (2022), doi:10.1103/PhysRevB.105.L121404, arXiv:2112.09557, https://arxiv.org/abs/2112.09557
13. J. Dukelsky, S. Pittel, G. Sierra, "Colloquium: Exactly solvable Richardson-Gaudin models for many-body quantum systems", Rev. Mod. Phys. 76, 643 (2004), doi:10.1103/RevModPhys.76.643, https://doi.org/10.1103/RevModPhys.76.643
14. M. Tavis, F. W. Cummings, "Exact Solution for an N-Molecule—Radiation-Field Hamiltonian", Phys. Rev. 170, 379 (1968), doi:10.1103/PhysRev.170.379, https://doi.org/10.1103/PhysRev.170.379