Fermi surface reconstruction and enhanced spin fluctuations in strained La$_3$Ni$_2$O$_{7}$ on LaAlO$_3$(001) and SrTiO$_3$(001)

We explore the structural and electronic properties of the bilayer nickelate La3Ni2O7 on LaAlO3(001) and SrTiO3(001) by using density functional theory including a Coulomb repulsion term. For La$_3$Ni$2$O${7}$/LaAlO$3$(001), we find that compressive strain and electron doping across the interface result in the unconventional occupation of the antibonding Ni $3d{z^2}$ states. In sharp contrast, no charge transfer is observed for La$_3$Ni$2$O${7}$/SrTiO$3$(001). Surprisingly, tensile strain drives a metallization of the bonding Ni $3d{z^2}$ states, rendering a Fermi surface topology akin to superconducting bulk La$_3$Ni$2$O${7}$ under high pressure, yet with spin fluctuations enhanced considerably beyond pressure effects. Concomitantly, significant octahedral rotations are retained. We discuss the fundamental differences between hydrostatic pressure versus epitaxial strain and establish that strain provides a much stronger control over the Ni $e_g$ orbital polarization. The results suggest epitaxial La$_3$Ni$2$O${7}$, particularly under tensile strain, as interesting system to provide novel insights into the physics of bilayer nickelates and possibly induce superconductivity without external pressure.

💡 Research Summary

In this work the authors investigate whether the superconducting phase recently discovered in bulk La₃Ni₂O₇ (LNO) under high hydrostatic pressure (Tc ≈ 80 K) can be realized by epitaxial engineering rather than by applying external pressure. Using density‑functional theory with a Hubbard‑U correction (U = 4 eV) as implemented in Quantum ESPRESSO, they construct slab models that contain three bilayers of LNO sandwiched between either LaAlO₃(001) (LAO) or SrTiO₃(001) (STO) substrates. The in‑plane lattice constant a is fixed to the substrate value (a = 3.79 Å for LAO, a = 3.905 Å for STO), thereby imposing a compressive strain of ≈ –1 % on LAO and a tensile strain of ≈ +1 % on STO. The LAO interface is polar (LaO⁺/AlO₂⁻), which injects half an electron per unit cell into the nickelate, whereas the STO interface is non‑polar and does not transfer charge.

Structural relaxation shows that compressive strain on LAO expands the Ni–O bonds along the c‑axis (up to 4.71 Å) and contracts the A‑site spacing, while tensile strain on STO shortens the Ni–O bonds (down to 4.01 Å) and reduces the A‑site distance. These geometric changes strongly affect the Ni 3d e_g manifold. On LAO, the antibonding Ni 3d z² band is pulled down to the Fermi level and becomes partially occupied, a situation that does not occur in bulk LNO even at 30 GPa. The bonding z² band, in contrast, is fully filled and lies around –1 eV. On STO, the bonding Ni 3d z² band is flattened and crosses the Fermi level, creating a new hole pocket (labelled γ) that is essentially identical to the pressure‑induced γ pocket observed in superconducting bulk LNO.

To quantify the electronic consequences, the authors construct maximally‑localized Wannier functions for the full Ni 3d space and compute the dynamical spin susceptibility χ(q,ω) with the TRIQS‑TPRF impurity solver. The tensile‑strained STO heterostructure exhibits a spin‑fluctuation strength that exceeds the 30 GPa bulk case by roughly 30–40 %, indicating that epitaxial strain can enhance the magnetic glue more efficiently than hydrostatic pressure. The compressively strained LAO system, however, shows a reduction of low‑energy spin fluctuations because the injected electrons occupy the antibonding z² states, which are less prone to nesting.

Fermi‑surface analysis reveals that the LAO system displays a distorted α sheet (predominantly dₓ²₋ᵧ²) and a broadened β sheet (mixed dₓ²₋ᵧ²/d_z²) together with additional small pockets derived from the antibonding z² band. In the STO system the α and β sheets are similar to the bulk, but the emergence of the γ hole pocket reproduces the high‑pressure topology. Importantly, the octahedral rotation pattern (Glazer a⁻a⁻c⁰) remains robust under both compressive and tensile strain, unlike the pressure‑driven suppression of rotations in bulk LNO. This decoupling of orbital reconstruction from lattice‑rotation suppression provides a unique platform to test theories of superconductivity that rely on specific Fermi‑surface nesting conditions.

Figure 3 of the paper summarizes the orthogonal nature of strain versus pressure: hydrostatic pressure simultaneously reduces a and c, modestly decreasing the d_z² occupation (from 1.45 to 1.37 electrons) while keeping dₓ²₋ᵧ² nearly constant. In contrast, epitaxial strain drives a and c in opposite directions, allowing a much larger tuning of the e_g orbital polarization (Δ ≈ 8 % on STO, Δ ≈ 21 % on LAO). The authors argue that the tensile strain required to reproduce the high‑pressure Fermi surface corresponds to an effective pressure of ~10 GPa, but can be achieved simply by choosing an appropriate substrate.

Overall, the study identifies four key mechanisms: (i) charge transfer from polar LAO leading to occupation of antibonding d_z² states under compression; (ii) metallization of bonding d_z² states under tensile strain on STO, creating the γ pocket; (iii) preservation of octahedral rotations irrespective of strain sign; and (iv) a pronounced enhancement of spin fluctuations beyond what hydrostatic pressure can achieve. These findings suggest that epitaxially grown La₃Ni₂O₇ films, especially under tensile strain, could become superconducting at ambient pressure, offering a practical route to explore the pairing mechanism in bilayer nickelates without the technical challenges of high‑pressure experiments. Future work should focus on experimental verification via TEM (to measure A‑site and Ni‑site distances), ARPES (to map the reconstructed Fermi surface), and low‑temperature transport and magnetic susceptibility measurements to detect superconductivity and its interplay with the enhanced spin fluctuations.