Boundary bound states and integrable Wilson loops in ABJM

We derive an integrable reflection matrix for the scattering of excitations from a boundary with a degree of freedom when the reflection process preserves an $SU(1|2)$ symmetry. As this residual symmetry is not sufficient to fully determine the reflection matrix, we use the boundary remnant of the Yangian symmetry invariance and obtain a family of integrable solutions. A concrete realization of this setup is found when studying insertions in the 1/2 BPS Wilson loop in ABJM theory. The boundary degree of freedom appears as a boundary bound state due to poles in the dressing phase of the reflection matrix. We also compare our results with those obtained from the boundary bound state bootstrap procedure. The ABJM Wilson loop example enables us to perform perturbative verifications of our results.

💡 Research Summary

In this work the authors investigate the integrable structure underlying operator insertions on the 1/2‑BPS Wilson line of the three‑dimensional N = 6 super‑Chern‑Simons‑matter (ABJM) theory. The Wilson line can be viewed as a one‑dimensional defect whose local excitations are described by magnons propagating on an open spin chain. The bulk scattering of these magnons is governed by the well‑known centrally‑extended su(2|2) S‑matrix, while the presence of the line introduces a boundary that reflects magnons. For the fundamental magnon the residual symmetry after the insertion of a supersymmetric reference operator is SU(1|2). This symmetry alone fixes the reflection matrix up to an overall scalar factor, and the resulting matrix satisfies the boundary Yang‑Baxter equation (BYBE), confirming integrability in the simplest setting.

The novelty of the paper lies in two extensions. First, the authors consider higher‑rank bound‑state magnons (Q‑magnon bound states) and, more importantly, a boundary that itself carries a dynamical degree of freedom – a “boundary bound state” that appears as a pole in the dressing phase of the fundamental reflection matrix. The pole at x⁻ = −i (with x⁺ fixed by the magnon dispersion) signals that a magnon can become trapped at the endpoint of the spin chain, forming a new boundary excitation whose energy is E_B = ½(a_B d_B + b_B c_B). This phenomenon mirrors the familiar boundary bound states in the N = 4 SYM Wilson‑line problem, but here it occurs in the ABJM context and for both symmetric and antisymmetric representations.

Second, the authors show that the SU(1|2) symmetry is insufficient to uniquely determine the full reflection matrix when the boundary hosts such a degree of freedom. To close the system they invoke the Yangian extension Y(su(1|2)) that survives at the boundary. By demanding that the level‑one Yangian charges Q^{(1)} and S^{(1)} are conserved in the scattering process, they obtain additional algebraic constraints that fix the previously undetermined scalar functions. The resulting reflection matrix is a one‑parameter family (the parameter being a phase associated with the boundary state) and automatically satisfies the BYBE. The construction is performed explicitly for the fundamental magnon, for Q‑magnon bound states, and for the four‑dimensional representation describing the boundary bound state itself.

The paper also compares this Yangian‑based solution with the traditional boundary bound‑state bootstrap approach. In the bootstrap one builds the reflection matrix by imposing crossing symmetry, unitarity, and the existence of bound‑state poles, without reference to Yangian symmetry. The authors demonstrate that both methods lead to identical pole/zero structures and identical expressions for the dressing factors, confirming that the Yangian constraints are equivalent to the bootstrap consistency conditions in this setting.

To validate their all‑loop proposal, the authors perform explicit weak‑coupling calculations using the perturbative open‑spin‑chain Hamiltonian derived from the ABJM dilatation operator. They compute the one‑loop correction to the energy of the boundary bound state and to the reflection amplitudes, and they verify that these match the small‑g expansion of the exact reflection matrix obtained from Yangian invariance. This perturbative check provides a non‑trivial test of the integrable framework and of the identification of the pole as a genuine boundary bound state.

In the concluding section the authors summarise their findings: (i) the residual SU(1|2) symmetry alone does not fix the reflection matrix when the boundary carries a non‑trivial representation; (ii) the Yangian remnant supplies the missing constraints, preserving integrability; (iii) the pole in the dressing phase corresponds to a physical boundary bound state whose spectrum can be read off from the exact matrix; (iv) the Yangian‑based construction is fully consistent with the bootstrap method and with weak‑coupling perturbation theory. They suggest several future directions, including the study of more general representations, multiple boundary degrees of freedom, and the extension to finite‑coupling regimes via thermodynamic Bethe ansatz or quantum spectral curve techniques.