Origin of the pseudogap end point in high-Tc cuprate superconductors

There are two seemingly unrelated puzzles about the cuprate superconductors. The first puzzle concerns the strong non-BCS behavior around $x_c$, the end point of the superconducting dome on the overdoped side, where the cuprate is believed to be well described by the Fermi-liquid theory. This is the most evident in the observed ${\rho }_s\left(0\right)-T_c$ scaling and the large amount of uncondensed optical spectral weight at low energy. The second puzzle concerns the remarkable robustness of the $d$-wave pairing against the inevitable disorder effect in such a doped system, which is also totally unexpected from the conventional BCS picture. Here we show that these two puzzles are deeply connected to the origin of a third puzzle about the cuprate superconductors, namely, the mysterious quantum critical behavior observed around $x^{*}$, the so called pseudogap end point. Through a systematic variational Monte Carlo (VMC) study of the disordered 2D $t-J$ model from the resonating valence bond (RVB) perspective, we find that the $d$-wave pairing in this model is remarkably more robust against the disorder effect than that in a conventional $d$-wave BCS superconductor. We find that such remarkable robustness can be attributed to the spin-charge separation mechanism in the RVB picture, through which the $d$-wave RVB pairing of the charge-neutral spinons becomes essentially immune to the disorder potential except for the secondary effect related to the modulation of the local doping level by the disorder. We propose that there exists a Mott transition at $x^{*}$, where the RVB pairing in the underdoped regime is transmuted into the increasingly more BCS-like pairing for $x>x^{*}$, whose increasing fragility against the disorder effect leads to the non-BCS behavior and the ultimate suppression of superconductivity around $x_c$.