Science - USA (2021-07-09)

(Antfer) #1
SCIENCE sciencemag.org

GRAPHIC: C. BICKEL/

SCIENCE

By Eva Benckiser

S

uperconductivity is one of the most
fascinating, macroscopically observ-
able quantum phenomena. A notable
aspect of its description is the exis-
tence of an attractive interaction, a
“glue,” between electrons. This can
bind them into Cooper pairs, which then
condense into a common superconducting
ground state. In contrast to conventional su-
perconductors, which are described by the
theory of Bardeen, Cooper, and Schrieffer, a
primary phonon-mediated pairing mecha-
nism is unlikely in the high-temperature
(Tc) superconducting cuprates, and there-
fore they are referred to as unconventional.
One characteristic of all high-Tc cuprates are
the strong, quasi–two-dimensional antifer-
romagnetic correlations, with spin waves or
magnons as possible alternative pairing glue.
On page 213 of this issue, Lu et al. ( 1 ) report
on antiferromagnetic magnon-like spin ex-
citations in superconducting infinite-layer
nickelate thin films.
The recent discovery of this new class of
potentially unconventional superconductors,
Nd1–xSrxNiO 2 ( 2 ), has sparked great interest.
Although the Tc of 15 K is lower than the
Tc≤ 130 K in the cuprates, their structural
and electronic similarity immediately raised
the possibility of magnetic correlations and
a related pairing mechanism. To this point,
the authors measured the spin wave disper-
sion of Nd1–xSrxNiO 2 with resonant inelastic
x-ray scattering. A nearest-neighbor super-
exchange coupling of J = 64 meV was deter-
mined, which quantifies the coupling of two
neighboring magnetic nickel ions through
the nonmagnetic oxygen ion in between.
This large superexchange interaction is unex-
pected because it is not much smaller than
the J found in cuprates of over 100 meV.
The outstanding question is whether the
new nickelate superconductors are just like
the high-Tc cuprates. At first glance, the
isostructural transition-metal-oxide planes
contain monovalent Ni1+ ions that are iso-
electronic to Cu2+, with a single spin ½ in the

planar 3dx (^2) – y 2 orbital on a square lattice (see
the figure). In both materials, strong elec-
tron-electron correlations cause the partially
occupied transition-metal d-band to split
up into so-called lower and upper Hubbard
subbands. One major difference in the elec-
tronic structure is the relative energy of the
oxygen-2p–derived bands with respect to the
Hubbard subbands. Whereas in cuprates, the
2 p bands lie in between, in nickelates, they
are located below both subbands. In a local
picture, this translates into a strongly re-
duced hybridization of nickel 3d- and oxygen
2 p-orbitals with possibly important conse-
quences for the magnetic superexchange in-
teraction. Reduced hybridization reduces the
effective hopping t between magnetic nickel
ions by way of the oxygen ligand. Because the
coupling constant J is proportional to t^2 /U,
where U is the on-site Coulomb repulsion,
values 10 times smaller than those in cu-
prates have been estimated ( 3 , 4 ). Quantum
chemical calculations ( 5 ), however, provide
an alternative argument that both t and U are
reduced in the nickelates. These calculations
predict a superexchange coupling J = 77 meV
for the bulk structure of NdNiO 2 , which is rel-
atively close to the one reported by Lu et al.
A notable aspect of the authors’ infinite-
layer nickelate superconductors is that they
are synthesized as epitaxial thin films grown
on a single-crystalline SrTiO 3 substrate.
Neither superconductivity nor antiferro-
magnetic order have been observed in bulk
samples ( 6 ). This calls into consideration
the extent to which heteroepitaxial modi-
fications from the substrate play a decisive
role in phase stabilization. Effects to be con-
sidered are, for example, the biaxial lattice
mismatch with the underlying substrate and
possible structural and electronic reconstruc-
tions at the polar substrate-film interface ( 7 ,
8 ). Because the hopping integral depends
inversely on some power of the distance be-
tween the ions, a shrinkage of the in-plane
lattice as provided by the epitaxy with the
underlying SrTiO 3 substrate could increase
the superexchange coupling. Should hetero-
epitaxial influences prove to be decisive, ex-
tremely interesting perspectives arise from
this for a targeted modification of the elec-
tronic structure as well as the magnetic in-
teractions and thus the possibility to stabilize
superconductivity at higher temperatures.
Addressing the coupling to rare-earth itin-
erant electrons will also need to be sorted
out. This might be accomplished by examin-
ing the spin waves in the analogous lantha-
num compound. The self-doped Ruddlesden-
Popper natural superlattices with varying
numbers of infinite-layer nickelate slabs are
another possible material system for gain-
ing insight into the electronic structure. A
recent x-ray spectroscopy study found siz-
able antiferromagnetic superexchange but
also indicates differences in nickel-oxygen
hybridization ( 9 ). Further information on
the symmetry of the superconducting wave
function is desired to assess the possibility of
a magnon-mediated Copper-pairing mecha-
nism. High-Tc cuprates show a characteristic
d-wave symmetry of the superconducting gap
function, which can be examined, for exam-
ple, with scanning tunneling spectroscopy. j
REFERENCES AND NOTES



  1. H. Lu et al., Science 373 , 213 (2021).

  2. D. Li et al., Nature 572 , 624 (2019).

  3. M. Jiang, M. Berciu, G. A. Sawatzky, Phys. Rev. Lett. 124 ,
    207004 (2020).

  4. Z. Liu et al., NPJ Quantum Mater. 5 , 31 (2020).

  5. V. M. Katukuri, N. A. Bogdanov, O. Weser, J. van den Brink,
    A. Alavi, Phys. Rev. B 102 , 241112 (2020).

  6. B.-X. Wang et al., Phys. Rev. Mater. 4 , 084409 (2020).

  7. F. Bernardini, A. Cano, J. Phys. Mat. 3 , 03LT01 (2020).

  8. B. Geisler, R. Pentcheva, Phys. Rev. B 102 , 020502 (2020).

  9. J. Q. Lin et al., Phys. Rev. Lett. 126 , 087001 (2021).
    10.1126/science.abi6855


SUPERCONDUCTIVITY

A touch more unconventional


Strong antiferromagnetic interactions are revealed


in superconducting nickelates


Max Planck Institute for Solid State Research,
Heisenbergstraße 1, 70569 Stuttgart, Germany.
Email: [email protected]

A structure on a substrate
The lattice mismatch with the SrTiO 3
substrate may afect the antiferromagnetic
superexchange coupling (J).

Spin

Nd/Sr

NiO 2

SrTiO 3

Ni-3dx2–y2

O–2p

J

Orbital

Ni

Magnetic interactions in the NiO 2 planes
Antiferromagnetic coupling of neighboring Ni
spins arises from exchange interaction involving
specifc oxygen orbitals.

Superconducting nickelates


9 JULY 2021 • VOL 373 ISSUE 6551 157

0709Perspectives.indd 157 7/1/21 6:34 PM

Free download pdf