Abstract Body

Following the copper and iron ages, the discovery of superconductivity in infinite layer nickelates _RE_NiO₂ (RE =Nd, La, etc.) [1] has kicked-off the “nickel age”[2] of unconventional superconductivity. Indeed, last year, superconductivity with a maximum T_c ~ 80K was discovered in a bilayer Ruddlesden-Popper nickelate La₃Ni₂O₇[3] under high pressure, and this was followed by the discovery of superconductivity with T_c = 20 ~ 30K in a trilayer La₄Ni₃O₁₀ also under high pressure [4], which was further confirmed by other experiments [5]. It is worth mentioning that for the bilayer nickelate La₃Ni₂O₇ two of the present authors discussed the possibility of superconductivity in this material even before its experimental discovery [6], and for the trilayer La₄Ni₃O₁₀ the structural transition to tetragonal symmetry under pressure as well as the occurrence of superconductivity with a T_c ~ ~comparable to those of the low T_c cuprates was predicted theoretically by the present authors [4].

As for the infinite layer nickelates, several authors, including us, have theoretically proposed a d-wave pairing scenario, since the electron configuration is considered to be close to d⁹ as in the cuprates [7-9]. Nonetheless, in a previous study, two of the present authors proposed a possibility of s±-wave superconductivity with an even higher Tc than for the d-wave pairing [10]. Here, s±-wave means that the sign of the superconducting gap function between 3d_x2-y2 and other four 3d orbitals are reversed. In Ref. [10], the existence of residual hydrogen was assumed so that the electron configuration is closer to d⁸.  When such an electron configuration is combined with a large energy level offset ΔE between 3d_x2-y2 orbitals and other 3d (especially 3d_3z2-r2 ) orbitals, as in the infinite-layer case, where there are no apical oxygens, a possibility of high T_c s±-wave superconductivity arises owing to a situation where the four 3d bands are incipient, namely, they lie close to, but does not intersect, the Fermi level,. The basic idea of the high T_c mechanism originates from the equivalence between two-orbital and bilayer Hubbard models [11, 12], where the interorbital energy level offset ΔE in the former is transformed to the interlayer hopping t_⊥ in the latter.

In the present study, we theoretically propose an alternative possibility of achieving s±-wave high T_c~  ~superconductivity in the infinite layer nickelates through realization of electron configuration close to d⁸. Namely, we consider doping large amount of holes in, e.g., LaNiO₂ by largely substituting the rare earth elements (La in this case) with alkaline-earth elements (such as Sr or Ba). The end materials SrNiO₂ and BaNiO₂ are known to exhibit orthorhombic crystal structures, but if the tetragonal structure can be achieved in thin films grown on substrates having tetragonal symmetry, there may be a chance of realizing the above mentioned scenario and hence high T_c s±-wave superconductivity.

In the actual calculation, we assume tetragonal symmetry of the lattice with the lattice constants fixed at those of SrTiO₃ LSAT, and LaAlO₃ substrates, and perform first principles band calculation, taking into account the effect of partial substitution through virtual crystal approximation.  Interestingly, our phonon calculations suggest that the P4/mmm symmetry as in LaNiO₂ is dynamically stable for La_1-xSr_xNiO₂ for the entire range of x and on all three substrates. Based on the obtained electronic band structure, we construct a five orbital model (including all Ni 3d orbitals). We also estimate the intraorbital and interorbital interaction values using constrained RPA. We apply the fluctuation exchange approximation to the models to take into account the effect of the spin fluctuations, and solve the linearized Eliashberg equation at a fixed temperature of T=0.01eV. The obtained eigenvalue of the linearized equation serves as a measure for the expected T_c which indeed suggests a possibility of high T_c superconductivity in the heavily hole-doped regime, especially for the substrates with small lattice constants.

References

[1] D. Li et al., Nature (London) 572, 624 (2019).

[2] M. R. Norman, Physics 13, 85 (2020).

[3] H. Sun et al., Nature 621, 493 (2023).

[4] H. Sakakibara et al., Physical Review B 109, 144511 (2024).

[5] Y. Zhu et al., arXiv:2311.07353.

[6] M. Nakata et al., Phys. Rev. B 95, 214509 (2017).

[7] H. Sakakibara et al., Phys. Rev. Lett. 125, 077003 (2020).

[8] X. Wu et al., Phys. Rev.B 101, 060504 (2020).

[9] M. Kitatani et al., npj Quantum Materials 5, 59 (2020).

[10] N. Kitamine et al.,Phys. Rev. Res. 2, 042032(R) (2020).

[11] H. Shinaoka et al., Phys. Rev. B 92, 195126 (2015).

[12] K. Yamazaki et al., Phys. Rev. Res. 2, 033356 (2020).

Acknowledgment

We are supported by JSPS KAKENHI Grant No. JP22K03512 (H. S.), JP22K04907 (K. K.), JP24K01333. The computing resource is supported by the supercomputer system (system-B) in the Institute for Solid State Physics, the University of Tokyo, and the supercomputer of Academic Center for Computing and Media Studies (ACCMS), Kyoto University.