Coulomb Branch and Integrability
Frank Coronado, Shota Komatsu, Konstantin Zarembo
TL;DR
This work addresses one-point functions on the Coulomb branch of planar $\mathcal{N}=4$ SYM by leveraging integrability and holography. The authors formulate an integrable boundary bootstrap for a probe D3-brane in $AdS_5\times S^5$, derive a finite-coupling asymptotic formula for vacuum condensates, and express it as a determinant-like overlap of spin-chain states with a boundary state, including a kinematical factor $\mathcal{C}_{\mathbf{K}}$ and a boundary dressing phase $\sigma_B(u)$. They validate the prediction by explicit tree-level and one-loop field-theory computations for several operator families (Konishi, BMN-like two-magnon states, and dimension-4 scalars), finding perfect agreement and thus providing nontrivial evidence for integrability on the Coulomb branch. The results connect the Coulomb-branch condensates to overlaps governed by the Gaudin norm and two-particle form factors, and they outline future directions to extend the framework to descendants, other observables, and a string-theory proof of the boundary integrability. Overall, the paper advances a concrete, testable integrability program for non-conformal vacua in a maximally symmetric gauge theory with holographic duality, potentially informing wider non-perturbative analyses beyond conformal settings.
Abstract
We study one-point functions of non-BPS single-trace operators on the Coulomb branch of planar $\mathcal{N}=4$ supersymmetric Yang-Mills theory. Holography relates them to overlaps between on-shell closed string states and a boundary state describing a probe D3-brane in $AdS_5\times S^{5}$. Assuming that the D-brane preserves integrability, we formulate and solve integrable bootstrap equations satisfied by the boundary state at finite 't Hooft coupling. This leads to a closed-form determinant expression for one-point functions at finite coupling, valid for sufficiently long operators. We test the result against direct field theory computations at tree level and one loop, finding perfect agreement.
