Comment on "Absence of a Slater transition in the two-dimensional Hubbard model".

Abstract

In a recent paper, Phys. Rev. Lett. 87, 167010/1-4 (2001), Moukouri and Jarrell presented evidence that in the two-dimensional (d=2) Hubbard model at half-filling there is a metal-insulator transition (MIT) at finite temperature even in weak coupling. While we agree with the numerical results of that paper, we arrive at different conclusions: The apparent gap at finite-temperature can be understood, at weak-coupling, as a crossover phenomenon involving large (but not infinite) antiferromagnetic (AFM) correlation length. Phase-space effects on the self-energy in d=2 are crucial, as are the ratio between AFM correlation length and single-particle thermal de Broglie wavelength. In weak coupling, d=2, there is in general no finite-temperature MIT transition in the thermodynamic sense.