Superconducting dome in doped quasi-two-dimensional organic Mott insulators: A paradigm for strongly correlated superconductivity

Layered organic superconductors of the BEDT family are model systems for understanding the interplay of the Mott transition with superconductivity, magnetic order, and frustration, ingredients that are essential to understand superconductivity also in the cuprate high-temperature superconductors. Recent experimental studies on a hole-doped version of the organic compounds reveals an enhancement of superconductivity and a rapid crossover between two different conducting phases above the superconducting dome. One of these phases is a Fermi liquid, the other not. Using plaquette cellular dynamical mean field theory with state-of-the-art continuous-time quantum Monte Carlo calculations, we study this problem with the two-dimensional Hubbard model on the anisotropic triangular lattice. Phase diagrams as a function of temperature T and interaction strength U/t are obtained for anisotropy parameters t' = 0.4t, t' = 0.8t and for various fillings. As in the case of the cuprates, we find, at finite doping, a first-order transition between two normal-state phases. One of theses phases has a pseudogap while the other does not. At temperatures above the critical point of the first-order transition, there is a Widom line where crossovers occur. The maximum (optimal) superconducting critical temperature T-c(m) at finite doping is enhanced by about 25% compared with its maximum at half filling and the range of U/t where superconductivity appears is greatly extended. These results are in broad agreement with experiment. Also, increasing frustration (larger t'/t) significantly reduces magnetic ordering, as expected. This suggests that for compounds with intermediate to high frustration, very light doping should reveal the influence of the first-order transition and associated crossovers. These crossovers could possibly be even visible in the superconducting phase through subtle signatures. We also predict that destroying the superconducting phase by a magnetic field should reveal the first-order transition between metal and pseudogap. Finally, we predict that electron doping should also lead to an increased range of U/t for superconductivity but with a reduced maximum T-c. This work also clearly shows that the superconducting dome in organic superconductors is tied to the Mott transition and its continuation as a transition separating pseudogap phase from correlated metal in doped compounds, as in the cuprates. Contrary to heavy fermions for example, the maximum T-c is definitely not attached to an antiferromagnetic quantum critical point. That can also be verified experimentally.