Does Simple Two-Dimensional Hubbard Model Account for High-Tc Superconductivity in Copper Oxides?

TL;DR

This study uses advanced Monte Carlo simulations to evaluate the doped Hubbard model's ability to explain high-Tc superconductivity in cuprates, finding that the model's superconducting correlations are too weak to account for experimental observations.

Contribution

It provides an unbiased estimate of superconducting correlations in the Hubbard model, challenging previous theoretical extensions that overestimated these correlations.

Findings

Superconducting correlations decay as inverse cubic of distance.

Upper bound of correlations is too weak for high-Tc explanation.

Recent theories overestimate the strength of correlations.

Abstract

We reexamine whether the essence of high-Tc superconductivity is contained in doped Hubbard models on the square lattice by using recently developed pre-projected Gaussian-basis Monte Carlo method. The superconducting correlations of the dx2-y2 wave symmetry in the ground state at distance r decays essentially as inverse cubic of r. The upper bound of the correlation at long distances estimated by this unbiased method is 0.001, indicating that recent extensions of dynamical mean-field theories and variational methods yielded at least an order of magnitude overestimates of it. The correlations are too weak for the realistic account of the cuprate high-Tc superconductivity.