Rigorous foundation, multi-species generalization, multi-time extension, and convergence domain of the LeClair-Mussardo expansion for the Lieb-Liniger gas

The Lieb-Liniger gas — one-dimensional bosons with repulsive contact interactions — is the paradigmatic exactly solved interacting quantum many-body system, yet exact closed representations of its correlation functions at finite density remain incomplete. Goldstein and Andrei (arXiv:1507.00077, 2015) derived a LeClair-Mussardo-type series expressing multi-point correlation functions in arbitrary Bethe-ansatz eigenstates (ground, thermal, and generalized-Gibbs states) as a sum of multiple integrals whose integrands are products of the state's filling function $f(k)$ times kernels built from algebraic-Bethe-ansatz determinant identities. Four issues remain open in that work, and they jointly constitute the problem of establishing this expansion as a rigorous, general, convergent representation: (i) the derivation is not mathematically rigorous — the supporting sinh-Gordon nonrelativistic-limit argument is motivational, and the LeClair-Mussardo formula has never been rigorously proved for general correlators even in relativistic integrable quantum field theory; (ii) the conjectured multi-species generalization (their Eq. (18)) for arbitrary integrable models with several particle types is unproven; (iii) the extension to genuinely time-dependent (multi-time) correlation functions, announced as ongoing work, was never delivered; (iv) the convergence domain beyond the short-distance, low-density (large-coupling) regime is unknown. Later literature has resolved parts of the surrounding landscape — two-point LeClair-Mussardo series at space-like separations in relativistic integrable field theory, nested-Bethe-ansatz series for two-species ground states, a systematic $1/c$ dynamical expansion, and an exact resummation at zero distance — but a rigorous derivation, the general multi-species formula, a multi-time series for the Lieb-Liniger gas, and the finite-distance convergence domain all remain open.

Background

The Lieb-Liniger model describes bosons moving on a line and interacting via a repulsive delta-function potential of strength $c$; it is the simplest exactly solved interacting quantum many-body system and a quantitative description of cold-atom experiments in highly anisotropic traps. Its exact eigenstates are given by the Bethe ansatz: each $N$-particle state is labelled by a set of quasi momenta ('rapidities') $k_1,\ldots,k_N$ fixed by the Bethe equations. In the thermodynamic limit a macroscopic state is characterized not by individual rapidities but by the filling function $f(k)=\rho_p(k)/\rho_t(k)$, the ratio of the particle rapidity density to the total (particles plus holes) rapidity density; ground states, thermal states, and generalized Gibbs ensembles (GGE — the ensemble built from all local conserved charges of an integrable model, which describes the long-time steady state after a quantum quench, i.e. a sudden change of a system parameter) are all encoded this way.

A form factor is a matrix element of a local operator between exact eigenstates. Integrable scattering is elastic and factorized, so the natural spectral representation of a correlation function is a series over multi-particle intermediate states; the delicate objects are the diagonal form factors, where incoming and outgoing rapidities coincide, which are singular limits requiring a prescription (connected or symmetric evaluation, subtracting kinematical pole contributions). In relativistic integrable quantum field theory, LeClair and Mussardo proposed in 1999 that the expectation value of a local operator in a finite-density (thermal) state is given by a universal series: a sum over $n$ of multiple integrals, each weighted by filling factors $(1+e^{\varepsilon(\theta)})^{-1}$ of the Bethe-ansatz rapidity distribution, multiplying the $2n$-particle connected diagonal form factor. This LeClair-Mussardo (LM) formula was later extended to arbitrary excited Bethe states, and its Lieb-Liniger counterpart was developed through the identification of the Lieb-Liniger gas as the nonrelativistic limit of the sinh-Gordon model (the simplest relativistic integrable quantum field theory, describing a single scalar particle with an exponential self-interaction): Kormos, Mussardo and Trombettoni used this mapping to compute one-point functions in thermal and highly excited states, building on the exact two-point series for the delta-function Bose gas found by Creamer, Thacker and Wilkinson in the 1980s.

Goldstein and Andrei (arXiv:1507.00077) derived, entirely within the algebraic Bethe ansatz of the Lieb-Liniger model, a series of the LM type for arbitrary multi-point correlation functions with respect to arbitrary eigenstates: schematically, $\langle\{k\}|O|\{k\}\rangle=\sum_n \frac{1}{n!}\int\prod_j f(k_j)\frac{dk_j}{2\pi}\,F^{O}_{2n}(k_1,\ldots,k_n)$, with kernels $F^{O}_{2n}$ expressible through determinants (Gaudin-matrix-type objects — matrices of derivatives of the Bethe equations with respect to the rapidities, which encode the normalization and scalar products of Bethe states) and depending only on the momenta, not on the state. They state that the series converges rapidly only at short distances and low particle density (equivalently large coupling $c$). They conjecture the analogous multi-species formula for arbitrary integrable models with several particle types (their Eq. (18)), in which one additionally sums over the particle species $p_j$ and the kernels carry species indices, $F^{O,p_1\ldots p_n}_{2n}$. The authors explicitly flag as open the lack of a fully rigorous derivation (noting that even the original relativistic LM formula has never been rigorously proved for general correlation functions), the unproven multi-species conjecture, and the multi-time extension, which they announce as work in progress; they claim rapid convergence only at short distances and low particle density, leaving the convergence behavior at larger distances or higher densities undetermined.

The subsequent literature treats these questions inside a broader active program. Pozsgay and Szécsényi derived an LM series for two-point functions in relativistic integrable field theory, valid only at space-like separations, and state that they have no decisive answer for time-like separations. Hutsalyuk, Pozsgay and Pristyák proved expansion theorems for nested Bethe-ansatz models with algebra symmetries $\mathfrak{gl}(2|1)$ and $\mathfrak{gl}(3)$ (two-species systems such as the Gaudin-Yang model of spin-1/2 fermions), yielding LM series for one- and two-point functions, but only for ground states with real Bethe roots, with string solutions (bound-state patterns of complex rapidities), finite temperature and dynamical correlators deferred to future work. For the Lieb-Liniger gas itself, Granet and Essler developed a systematic $1/c$ expansion of form-factor sums for dynamical correlations in arbitrary eigenstates, uniform in space and time — an asymptotic expansion at large coupling rather than the convergent LM filling-function series — while exact closed expressions for local (zero-distance) correlations in arbitrary states follow from the conjectural Negro-Smirnov formula for the sinh-Gordon model transported through the nonrelativistic mapping. None of these delivers a rigorous foundation, the general multi-species statement, a multi-time series, or a finite-distance convergence theory for the LM expansion.

Problem Statement

Establish the LeClair-Mussardo expansion for correlation functions of the Lieb-Liniger gas — and, as conjectured, of general integrable models — as a rigorous, general, and convergent spectral representation of finite-density correlation functions. The common objective is that a single framework, built from the bare Bethe-ansatz form factors and the filling function $f(k)$ of the macroscopic state, yields provably valid series for correlation functions in all states, species structures, and space-time regimes; the following four requested results decompose that objective and are scientifically inseparable from it, since each one closes a distinct failure mode (uncontrolled derivation, restricted model class, restricted time dependence, unknown convergence) of the same series:

(a) Rigor: provide a mathematically rigorous derivation of the LeClair-Mussardo series for the Lieb-Liniger gas (Goldstein-Andrei Eq. (2)) for multi-point correlation functions of local operators in arbitrary macroscopic Bethe states (ground, thermal, and generalized-Gibbs states) — with full control of the thermodynamic limit, a well-defined and proven prescription for the diagonal limits of the form-factor kernels, and explicit remainder or convergence statements — without relying on the motivational nonrelativistic-limit argument from the sinh-Gordon model.

(b) Multi-species: derive the conjectured generalization (Goldstein-Andrei Eq. (18)) in its stated generality: for arbitrary integrable models with multiple particle types, correlation functions are given by the analogous series with an additional sum over particle species, weights $f^{p_j}(k_j)$, and kernels $F^{O,p_1\ldots p_n}_{2n}$.

(c) Multi-time: extend the expansion to genuinely time-dependent correlation functions of the Lieb-Liniger gas (operator insertions at unequal times), with a well-defined summation prescription for the kinematical singularities and a statement of validity in the space-time separation.

(d) Convergence domain: determine in which region of interparticle distance, density, coupling $c$, and filling function the multi-integral series converges beyond the short-distance, low-density regime in which only rapid convergence is claimed.

A complete answer establishes the shared framework for all four components. Rigorous resolutions of individual components (for example, a rigorous single-species derivation, or a proven convergence domain at fixed distance) constitute partial progress toward the overall objective rather than a complete solution, provided the reduction in scope is explicitly stated rather than silently substituted for the full objective.

References

1. G. Goldstein and N. Andrei, 'Correlation functions of the Lieb-Liniger gas and the LeClair-Mussardo formula', arXiv:1507.00077 (2015, unpublished preprint), doi:10.48550/arXiv.1507.00077, https://arxiv.org/abs/1507.00077
2. A. LeClair and G. Mussardo, 'Finite temperature correlation functions in integrable QFT', Nucl. Phys. B 552 (1999) 624-642, doi:10.1016/S0550-3213(99)00280-1, https://doi.org/10.1016/S0550-3213(99)00280-1
3. M. Kormos, G. Mussardo and A. Trombettoni, 'Expectation Values in the Lieb-Liniger Bose Gas', Phys. Rev. Lett. 103 (2009) 210404, doi:10.1103/PhysRevLett.103.210404, https://doi.org/10.1103/PhysRevLett.103.210404
4. M. Kormos, G. Mussardo and A. Trombettoni, 'One-dimensional Lieb-Liniger Bose gas as nonrelativistic limit of the sinh-Gordon model', Phys. Rev. A 81 (2010) 043606, doi:10.1103/PhysRevA.81.043606, https://doi.org/10.1103/PhysRevA.81.043606
5. D. B. Creamer, H. B. Thacker and D. Wilkinson, 'Some exact results for the two-point function of an integrable quantum field theory', Phys. Rev. D 23 (1981) 3081-3084, doi:10.1103/PhysRevD.23.3081, https://doi.org/10.1103/PhysRevD.23.3081
6. B. Pozsgay and G. Takács, 'Form factors in finite volume II: disconnected terms and finite temperature correlators', Nucl. Phys. B 788 (2008) 209-251, doi:10.1016/j.nuclphysb.2007.07.008, https://doi.org/10.1016/j.nuclphysb.2007.07.008
7. B. Pozsgay, 'Mean values of local operators in highly excited Bethe states', J. Stat. Mech. (2011) P01011, doi:10.1088/1742-5468/2011/01/P01011, https://doi.org/10.1088/1742-5468/2011/01/P01011
8. B. Bertini, L. Piroli and P. Calabrese, 'Quantum quenches in the sinh-Gordon model: steady state and one point correlation functions', J. Stat. Mech. (2016) 063102, doi:10.1088/1742-5468/2016/06/063102, https://doi.org/10.1088/1742-5468/2016/06/063102
9. S. Negro and F. Smirnov, 'On one-point functions for sinh-Gordon model at finite temperature', Nucl. Phys. B 875 (2013) 166-185, doi:10.1016/j.nuclphysb.2013.06.023, https://doi.org/10.1016/j.nuclphysb.2013.06.023
10. A. Bastianello, L. Piroli and P. Calabrese, 'Exact local correlations and full counting statistics for arbitrary states of the one-dimensional interacting Bose gas', Phys. Rev. Lett. 120 (2018) 190601, doi:10.1103/PhysRevLett.120.190601, https://doi.org/10.1103/PhysRevLett.120.190601
11. A. Bastianello and L. Piroli, 'From the sinh-Gordon field theory to the one-dimensional Bose gas: exact local correlations and full counting statistics', J. Stat. Mech. (2018) 113104, doi:10.1088/1742-5468/aaeb48, https://doi.org/10.1088/1742-5468/aaeb48
12. B. Pozsgay and I. M. Szécsényi, 'LeClair-Mussardo series for two-point functions in Integrable QFT', J. High Energy Phys. 2018 (5) 170, doi:10.1007/JHEP05(2018)170, https://doi.org/10.1007/JHEP05(2018)170
13. A. Hutsalyuk, B. Pozsgay and L. Pristyák, 'The LeClair-Mussardo series and nested Bethe Ansatz', Nucl. Phys. B 964 (2021) 115306, doi:10.1016/j.nuclphysb.2021.115306, https://doi.org/10.1016/j.nuclphysb.2021.115306
14. A. Cortés Cubero and M. Panfil, 'Thermodynamic bootstrap program for integrable QFT's: form factors and correlation functions at finite energy density', J. High Energy Phys. 2019 (1) 104, doi:10.1007/JHEP01(2019)104, https://doi.org/10.1007/JHEP01(2019)104
15. M. Panfil and R. M. Konik, 'Extending the thermodynamic form factor bootstrap program: multiple particle-hole excitations, crossing symmetry, and reparameterization invariance', J. High Energy Phys. 2023 (10) 072, doi:10.1007/JHEP10(2023)072, https://doi.org/10.1007/JHEP10(2023)072
16. E. Granet and F. H. L. Essler, 'A systematic $1/c$-expansion of form factor sums for dynamical correlations in the Lieb-Liniger model', SciPost Phys. 9 (2020) 082, doi:10.21468/SciPostPhys.9.6.082, https://doi.org/10.21468/SciPostPhys.9.6.082
17. B. Doyon, 'Exact large-scale correlations in integrable systems out of equilibrium', SciPost Phys. 5 (2018) 054, doi:10.21468/SciPostPhys.5.5.054, https://doi.org/10.21468/SciPostPhys.5.5.054
18. N. A. Slavnov, 'Algebraic Bethe ansatz approach to the correlation functions of the one-dimensional bosons with attraction', J. High Energy Phys. 2024 (6) 061, doi:10.1007/JHEP06(2024)061, https://doi.org/10.1007/JHEP06(2024)061
19. M. Borsi, B. Pozsgay and L. Pristyák, 'Current operators in integrable models: a review', J. Stat. Mech. (2021) 094001, doi:10.1088/1742-5468/ac0f6b, https://doi.org/10.1088/1742-5468/ac0f6b
20. L. Piroli and P. Calabrese, 'Local correlations in the attractive one-dimensional Bose gas: From Bethe ansatz to the Gross-Pitaevskii equation', Phys. Rev. A 94 (2016) 053620, doi:10.1103/PhysRevA.94.053620, https://doi.org/10.1103/PhysRevA.94.053620 (preprint arXiv:1609.03854, https://arxiv.org/abs/1609.03854)