Quasi-particle Statistics and Braiding from Ground State Entanglement
Yi Zhang, Tarun Grover, Ari Turner, Masaki Oshikawa, Ashvin Vishwanath
TL;DR
The paper develops a framework to extract quasi-particle statistics from ground-state entanglement on a torus by exploiting the nontrivial topology of entanglement cuts. It introduces Minimum Entropy States (MESs) as the ground-state basis that maximizes the topological entanglement entropy for a given cut, enabling the construction of modular matrices $\mathcal{S}$ and $\mathcal{U}$ that encode self- and mutual-statistics and quantum dimensions. Through analytical and variational Monte Carlo analyses of a SU(2) spin-symmetric Chiral Spin Liquid and analytic treatment of a $Z_2$ toric code spin liquid, the authors show that MESs correspond to definite flux sectors and that the $\mathcal{S}$-matrix can be extracted from rotations acting on the MES basis, up to a global phase fixed by the identity particle. They also provide explicit algorithms to obtain $\mathcal{S}$ and related data from a degenerate ground-state manifold without requiring edge states or finite-size extrapolations. The approach offers a practical, ground-state–based route to diagnosing topological order and quasi-particle statistics in lattice models and beyond, with potential generalizations to higher dimensions.
Abstract
Topologically ordered phases are gapped states, defined by the properties of excitations when taken around one another. Here we demonstrate a method to extract the statistics and braiding of excitations, given just the set of ground-state wave functions on a torus. This is achieved by studying the Topological Entanglement Entropy (TEE) on partitioning the torus into two cylinders. In this setting, general considerations dictate that the TEE generally differs from that in trivial partitions and depends on the chosen ground state. Central to our scheme is the identification of ground states with minimum entanglement entropy, which reflect the quasi-particle excitations of the topological phase. The transformation of these states allows for a determination of the modular S and U matrices which encode quasi-particle properties. We demonstrate our method by extracting the modular S matrix of an SU(2) spin symmetric chiral spin liquid phase using a Monte Carlo scheme to calculate TEE, and prove that the quasi-particles obey semionic statistics. This method offers a route to a nearly complete determination of the topological order in certain cases.
