arXiv:1402.2057
Order parameter for the Mott transition in two dimensions: Small-world network
Chung-Pin Chou
Abstract
We show how to define the appropriate network measure that characterizes the nonmagnetic Mott transition starting from the Hubbard model on a square lattice. To simulate the Mott transition, we employ the variational Monte Carlo method based on the Gutzwiller wave function with nearest-neighbor doublon-holon binding correlations. In order to search the Mott transition point, we use the doublon-holon correlation function as the weighted network representation for analyzing the half-filled electronic system. We then show that a network quantity, small-worldness, can be applied to defining the order parameter describing a hidden symmetry broken in the network representation, which allows us to capture the Mott transition point. As a result, we suggest that the picture of the Mott transition can be obtained using a mass of data entirely contained in topologies of the proper weighted networks.