Correlated electronic structures and unconventional superconductivity in bilayer nickelate heterostructures

Abstract

The recent discovery of ambient-pressure superconductivity in thin-film bilayer nickelates opens new possibilities for investigating electronic structures in this new class of high-transition-temperature ([Formula: see text]) superconductors. Here, we construct a realistic multi-orbital Hubbard model for the thin-film system based on structural parameters integrating scanning transmission electron microscopy measurements and ab initio calculations. The interaction parameters are calculated with the constrained random phase approximation (cRPA). Density functional theory (DFT) plus cluster dynamical mean-field theory (CDMFT) calculations, with cRPA-calculated on-site Coulomb repulsive [Formula: see text] and experimentally measured electron filling [Formula: see text], quantitatively reproduce Fermi surfaces from angle-resolved photoemission spectroscopy experiments. The distinct Fermi surface topology from simple DFT+[Formula: see text] results features the indispensable role of correlation effects. Based upon the correlated electronic structures, a modified random-phase-approximation (RPA) approach yields a pronounced [Formula: see text]-wave pairing instability, due to the strong spin fluctuations originating from the Fermi surface nesting between bands with predominantly [Formula: see text] characters. Our findings highlight the quantitative effectiveness of the DFT+cRPA+CDMFT approach that precisely determines correlated electronic structure parameters without fine-tuning. The revealed intermediate correlation effect may explain the same order-of-magnitude onset [Formula: see text] observed both in pressured bulk and strained thin-film bilayer nickelates.

Keywords: La3Ni2O7; cluster dynamical mean-field theory; correlated electronic structure; epitaxial thin film; unconventional pairing mechanism.

Figure 1.

Oxygen octahedral rotation analysis in the LaPrNiO film. (a) Schematic crystal structure of the bulk LaNiO with an oxygen octahedral rotation pattern. Displacements of oxygen atoms are indicated by . (b) Schematic crystal structure of the LaPrNiO film grown on an SrLaAlO substrate with an oxygen octahedral rotation pattern, showing no splitting or displacement in any direction. (c, d) Enlarged ABF images of the cross section of the LaPrNiO film, projected along [100] and [110], respectively. ABF images ([100], [010] and [110]) with larger fields of view are shown in Fig. S1. From top to bottom are the experimental results of the film, the simulation results of the film and the simulation results of the bulk. (e) The corresponding O atom column intensity profile on the vertical dashed line in the experimental and simulation results in panel (d). Only a column of O atoms on the right is annotated with dashed lines in panel (d). Crosses indicate the positions of the local minima corresponding to the positions of oxygen atoms. Dashed lines mark the same positions as the horizontal dashed lines in panel (d). Black arrows indicate the positions of the oxygen atoms deviating from the dashed lines.

Figure 2.

Crystal structure and DFT+ results ( eV,  eV). (a) The half-UC thin-film crystal structure of LaNiO constructed using structural parameters from the thin film of LaPrNiO. (b) The DFT+ bands (black solid line) and their Wannier interpolation (green dashed line) of the half-UC thin film. The thin (thick) dashed line marks the Fermi energy for the system with (without) Sr doping. The size of the red (blue) dots demonstrates the projected weight for Ni- (Ni-) orbitals. (c, d) Fermi surfaces of the Wannier TB model at and per Ni site, respectively. Names of all electron or hole pockets are labeled in (d). In panel (c), the extracted ARPES Fermi surface [40] is overlaid as the white dashed line. Only the right half is shown for clarity.

Figure 3.

The dependence of the CDMFT results at temperature  K, filling per Ni (in orbitals) and Hund coupling  eV. (a) The effective energy level for each bonding and anti-bonding orbital as a function of ( is the chemical potential). The dashed lines mark the bare energy levels from DFT. (b) The quasi-particle spectral weight (degenerate in spin). (c) The orbital occupancy per spin . (d–h) The momentum-resolved spectral function at indicated . We show in the large energy window for  eV in (d) and in the enlarged energy window for , 3.0, 3.6 and 3.77 eV in (e–h), respectively. The thin white dashed line in (d) indicates the DFT band of Ni- orbitals. The thick dashed line in (d–h) marks the Fermi energy. (i–l) The corresponding Fermi surface at indicated shown in (e–h). In panels (k–l), the extracted ARPES Fermi surfaces [37] are overlaid as the white dashed line. Only the right half is shown for clarity. The energy unit is in electronvolts (eV).

Figure 4.

The CDMFT+RPA calculation results on a square lattice at fixed temperature  eV (∼11.6 K), filling per Ni- and bare interaction strength  eV. (a) The distribution of the largest eigenvalue of the RPA-renormalized spin susceptibility matrix in the Brillouin zone for  eV. Here is the representative wave vector of eight equivalent distribution peaks related by the point group. (b) The largest eigenvalue of the effective pairing interaction vertex matrix as a function of the effective interaction strength for leading , and degenerate and wave pairings. (c) The distribution of the leading -wave pairing gap function near the Fermi surface for  eV. The black thin lines denote the Fermi surface. The wave vector in (a) and the Fermi surface nesting vector in (c) are identical. (d) The real space dependence of the -wave pairing gap function in the orbital basis for  eV. Here , the distance in units of the lattice, is constant between the two orbitals from two unit cells involved in the pairing. The inset shows the spatial distribution of . We denote by the pairing between the  orbital on the  layer and the  orbital on the  layer. The other input parameters are all from CDMFT.