After receiving a number of comments and reviews from our colleagues who have suggested that our results could be greatly improved by other methods of extrapolation, we have decided to withdraw our paper entitled "Absence of superconductivity in the 2D Hubbard model" until we can perform these calculations.
The numerical studies of d x 2 - y 2 -wave pairing in the two-dimensional (2D) and the 2-leg Hubbard models are reviewed. For this purpose, the results obtained from the determinantal Quantum Monte Carlo and the Density-Matrix Renormalization-Group calculations are presented. These are calculations which were motivated by the discovery of the high- T c cuprates. In this review, the emphasis is placed on the microscopic many-body processes which are responsible for the d x 2 - y 2 -wave pairing correlations observed in the 2D and the 2-leg Hubbard models. In order to gain insight into these processes, the results on the effective pairing interaction as well as the magnetic, density and the single-particle excitations will be reviewed. In addition, comparisons will be made with the other numerical approaches to the Hubbard model and the numerical results on the t - J model. The results reviewed here indicate that an effective pairing interaction which is repulsive at ( ~ , ~ ) momentum transfer, and enhanced single-particle spectral weight near the ( ~ ,0) and (0, ~ ) points of the Brillouin zone, create optimum conditions for d x 2 - y 2 -wave pairing. These are two effects which act to enhance the d x 2 - y 2 -wave pairing correlations in the Hubbard model. Finding additional ways is an active research problem.
We have studied the pair susceptibilities for all possible pair wave functions that fit on a two-dimensional (2D) eight-site Hubbard cluster by exact diagonalization of the Hamiltonian. Band fillings corresponding to four and six electrons were studied (two or four holes in the half-filled band) for a wide range of Hubbard interaction strengths and temperatures. Our results show that all pairing susceptibilities are suppressed by the Hubbard repulsion. We have also carried out perturbation-theory calculations which show that the leading-order U2 contributions to the d-wave pair susceptibility suppresses d-wave pairing over a significant temperature range. These results are consistent with recent Monte Carlo results and provide further evidence suggesting that the 2D Hubbard model does not exhibit superconductivity.
We show that the phase boundary between the paramagnetic metal and the nonmagnetic Mott insulator for the Hubbard model on a triangular lattice obtained by Yoshioka et al. in arXiv:0811.1575 does not correctly represent that of the thermodynamic limit but is an artifact of the 6 by 6 lattice they rely on. After the system size extrapolation, the phase boundary is located at U/t=5.2 as proposed by Morita et al., J. Phys. Soc. Jpn. 71 (2008) 2109 and in contrast to Yoshioka et al. Here, U is the onsite Coulomb repulsion and t is the nearest-neighbor transfer.