Apr 7, 2025
1:45pm - 2:15pm
Summit, Level 4, Room 439
Hideki Yamamoto1,Yoshiharu Krockenberger1,Ai Ikeda1,Yoshitaka Taniyasu1,Michio Naito2
NTT Basic Research Labs, NTT Corp.1,Tokyo University of Agriculture and Technology2
While a wide variety of high-
Tc superconducting cuprates have been discovered, the search for new materials remains at the forefront of research, and we have been striving to create novel superconductors by means of molecular beam epitaxy (MBE) [1]. Here, we introduce two families of cuprates newly synthesized by MBE, in which the playground of superconductivity is exclusively within CuO
2 planes with square-planar-coordinated Cu. Our results may challenge the widely accepted notion that the high-
Tc superconductivity is induced by carrier doping into Mott-insulating parent compounds.
We start with the Nd
2CuO
4 (T') structure. T'-
Ln2CuO
4 (
Ln = La, Pr, Nd) are known as the parent compounds of electron-doped cuprates and heterovalent substitution (Ce
4+ for
Ln3+) is required to induce superconductivity in bulk specimens. However, our T'-
Ln2CuO
4 thin films show superconductivity without the heterovalent substitution [1]. The stark contradiction between our results and commonly achieved data likely originates from the complicated oxygen chemistry. Unlike that hole doping into CuO
2 planes with octahedral- and/or pyramidal-coordinated Cu induces superconductivity, electron doping into CuO
2 planes with square-planar-coordinated Cu alone is insufficient for the induction of superconductivity. Instead, treatment of the as-grown specimens of T’-
Ln2-xCe
xCuO
4 under reducing environments is necessary, irrespective of the dopant (Ce) concentration
x. Such a reduction treatment is vital for elimination of defects, for example, removal of excess oxygen at apical sites while simultaneously regular oxygen sites remain occupied. As the annealing process is a diffusion process, thin-film samples are advantageous for achieving an ideal and uniform oxygen configuration. Nonetheless, the two regular oxygen sites in T’-cuprates—O(1) in the CuO
2 plane and O(2) in the
Ln2O
2 layer—may favor oxygen off-stoichiometries, and therefore unintentional doping.
This can be tested more rigorously by infinite-layer (IL) cuprates as the oxygen sites are confined within the CuO
2 planes. Single-crystalline specimens of such cuprates can only be prepared by MBE. We focus on IL-CaCuO
2, the common ingredient of cuprates with
Tc > 100 K. While the bare IL-CaCuO
2 thin films are insulating, our in-depth crystallographic analysis revealed the cause. Particularly, in-plane scanning transmission electron microscopy (STEM) shows that cationic stripes are formed in the CuO
2 planes and their density is sufficient to quench metallic conduction [2]. This stripe formation is considered to buffer charge imbalances, which are inevitably introduced during the growth triggered by point defect formation. To eliminate those stripes, we inserted Ca
2Fe
2O
5 layers after every 13 or 17 layers of CaCuO
2 by MBE. These superstructures drive the system metallic and superconducting [3]. As superconductivity is induced when CaCuO
2/Ca
2Fe
2O
5 superlattices are cooled from growth temperatures under strong oxidizing conditions, we also explored the possibility of hole-doping by excess oxygen introduced into the Ca
2Fe
2O
5 layers. We find this to be unlikely for the following reasons: (1) excess oxygen is not observed in STEM measurements; (2) Ca
2Fe
2O
5+δ only accepts small amount of excess oxygen (
δ ≤ 0.08) as far as the brownmillerite structure is preserved [4]; (3) a large Meissner signal excludes the possibility of interface superconductivity and the carrier concentration would be too small if the carriers were injected into the 13-17 layers of CuO
2. After all, the square-planar CuO
2 planes have a metallic and superconducting ground state
per se and nearly disorder-free CuO
2 planes are the only prerequisite for high-
Tc superconductivity.
[1] H. Yamamoto
et al., in “Epitaxial growth of complex metal oxides” (Elsevier, 2022).
[2] Y. Krockenberger
et al., ACS Omega 6 (2021) 21884.
[3] A. Ikeda
et al., ACS Appl. Electron. Mater. 4 (2022) 2672.
[4] A. Shaula
et al., DOI: 10.1109/OMEE.2012.6464764.