Interplay of two $E_g$ orbitals in Superconducting La$_3$Ni$_2$O$_7$ Under Pressure

TL;DR

This paper investigates the roles of two $E_g$ orbitals in the high-$T_c$ superconductivity of La$_3$Ni$_2$O$_7$, revealing that the $3d_{x^2-y^2}$ orbital primarily drives superconductivity, while the $3d_{z^2}$ orbital's role is limited.

Contribution

The study introduces a two-orbital bilayer $t$-$J$ model to analyze the interplay of $E_g$ orbitals, providing new insights into their distinct contributions to superconductivity in La$_3$Ni$_2$O$_7$.

Findings

$3d_{x^2-y^2}$ orbital facilitates high-$T_c$ superconductivity.

$3d_{z^2}$ orbital is less effective in inducing phase coherence.

Doping affects $T_c$ and the orbital contributions, with a crossover from BCS to BEC regimes.

Abstract

The discovery of high-$T_c$ superconductivity (SC) in La$_3$Ni$_2$O$_7$ (LNO) has aroused a great deal of interests. Previously, it was proposed that the Ni-$3d_{z^2}$ orbital is crucial to realize the high-$T_c$ SC in LNO: The preformed Cooper pairs therein acquire coherence via hybridization with the $3d_{x^2-y^2}$ orbital to form the SC. However, we held a different viewpoint that the interlayer pairing $s$-wave SC is induced by the $3d_{x^2-y^2}$ orbital, driven by the strong interlayer superexchange interaction. To include effects from both $E_g$-orbitals , we establish a two-orbital bilayer $t$-$J$ model. Our calculations reveal that due to the no-double-occupancy constraint, the $3d_{x^2-y^2}$ band and the $3d_{z^2}$ bonding band are flattened by a factor of about 2 and 10, respectively, which is consistent with recent angle-resolved-photo-emission-spectroscopy measurements. Consequently, a high temperature SC can be hardly induced in the $3d_{z^2}$-orbital due to the difficulty to develop phase coherence. However, it can be easily achieved by the $3d_{x^2-y^2}$ orbital under realistic interaction strength. With electron doping, the $3d_{z^2}$-band gradually dives below the Fermi level, but $T_c$ continues to enhance, suggesting that it is not necessary for the high-$T_c$ SC in LNO. With hole doping, $T_c$ initially drops and then rises, accompanied by the crossover from the BCS to BEC-type superconducting transitions.