A different branch of the high T_c family?

In PNAS, Li et al. (1) suggest that ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$ is a member of a different branch of high-$T_c$ cuprate superconducting materials. This branch is characterized as heavily overdoped with an exceptionally short Cu apical O spacing and O vacancies that are located in the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ planes. These characteristics, illustrated in Fig. 1, differ in significant ways from those of the traditional cuprate superconducting materials (2).

Fig. 1.

(A) A schematic cuprate phase diagram as a function of hole doping $x$. On the left at low doping are the familiar antiferromagnetic (AF), pseudogap (PG), and $d$-wave superconducting (SC) regions of ${\text{La}}_{2-x}{\mathrm{S}\mathrm{r}}_x{\mathrm{C}\mathrm{u}\mathrm{O}}_4$. On the right is the heavily overdoped $x\sim 0.4$ region of superconductivity Li et al. (1) discuss for ${\mathrm{B}\mathrm{a}}_2{\mathrm{C}\mathrm{u}\mathrm{O}}_{4-y}$. (B) The center panel shows a fictitious ${\mathrm{B}\mathrm{a}}_2{\mathrm{C}\mathrm{u}\mathrm{O}}_4$ compound with the same ${\mathrm{K}}_2{\mathrm{N}\mathrm{i}\mathrm{F}}_4$ structure as ${\mathrm{L}\mathrm{a}}_2{\mathrm{C}\mathrm{u}\mathrm{O}}_4$. The structures shown on the left and right have the compressed $c$-axis structure of ${\mathrm{B}\mathrm{a}}_2$${\text{Cu}\mathrm{O}}_{4-y}$. In this case the hole doping is controlled by the O vacancies. For the structure on the right, the O vacancies are on the apical sites and the ${\mathrm{C}\mathrm{u}\mathrm{O}}_4$ corner-shared ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ sheets remain intact. However, according to Li et al. (1) in ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$ the O vacancies are in the plane. In this case, as shown by the structure on the left, the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ sheets are destroyed and randomly oriented Cu–O chain segments are likely formed.

For example, consider ${\text{La}}_{2-x}{\mathrm{S}\mathrm{r}}_x{\mathrm{C}\mathrm{u}\mathrm{O}}_4$ (LSCO), which has the same ${\mathrm{K}}_2{\mathrm{N}\mathrm{i}\mathrm{F}}_4$ structure as ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$. Using a valence count with ${\text{La}}^{+2}$, ${\text{Sr}}^{+3}$, and ${\mathrm{O}}^{-2}$, the number of holes in the Cu $3d$ shell of LSCO is $1+x$. For $x=0$, LCO is antiferromagnetic. As Sr is added the antiferromagnetic Néel temperature decreases as shown in Fig. 1A and superconductivity onsets with $T_c$ peaking at a small doping $x\sim 0.15$. At larger doping, as the system moves further away from the antiferromagnetic Mott–Hubbard phase, $T_c$ decreases and vanishes for $x>0.25$. For ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$, the valence of Ba is $+2$ so that the hole doping $x=2\left(1-y\right)$. Thus, the $y=0.8$ sample Li et al. (1) discuss is heavily overdoped with $x=0.4$. Nevertheless, the $T_c$ of ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$ is approximately 30 K higher than that of optimally doped LSCO.

Like the previously reported highly overdoped cuprate materials (3–8) ${\mathrm{S}\mathrm{r}}_2$Cu${\mathrm{O}}_{4-y}$, ${(\text{Sr},\text{Ba})}_2$Cu${\mathrm{O}}_{4-y}$, and ${\text{Cu}}_{0.75}$${\text{Mo}}_{0.25}{\mathrm{S}\mathrm{r}}_2{\mathrm{Y}\mathrm{C}\mathrm{u}}_2{\mathrm{O}}_{7.54}$, ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$ is synthesized under high pressure and high temperature in the presence of a strong oxidizing agent. While the crystal has the ${\mathrm{K}}_2{\mathrm{N}\mathrm{i}\mathrm{F}}_4$ structure illustrated in the center of Fig. 1B, the ${\mathrm{C}\mathrm{u}\mathrm{O}}_6$ octahedron structure is highly compressed with a much shorter Cu–O apical distance (1.86 Å) than the traditional cuprates (2.42 Å for ${\mathrm{L}\mathrm{a}}_2{\mathrm{C}\mathrm{u}\mathrm{O}}_4$). This is schematically illustrated at the left and right in Fig. 1B. The shortening of the Cu–O apical distance raises the energy of the $3d_{3z^2-r^2}$ orbital so that in ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$ the Cu states near the Fermi level have both $3d_{3z^2-r^2}$ and $3d_{x^2-y^2}$ orbital character. As the authors note, a shortened apical Cu–O distance and the admixing of $3d_{3z^2-r^2}$ orbital weight in the states near the Fermi energy are found to reduce $T_c$ in the traditional cuprate superconductors (9).

Finally, the authors note at the end of the legend for figure 3 in ref. 1 that while the exact positions of the O vacancies are not known at present, they are in the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ planes. This agrees with the conclusions of Geballe and Marezio (5) regarding the O vacancies in ${\mathrm{S}\mathrm{r}}_2$Cu${\mathrm{O}}_{4-y}$. A number of earlier studies Li et al. (1) cite assume that the oxygen vacancies were at the apical O sites. In this case, while there would be ${\mathrm{C}\mathrm{u}\mathrm{O}}_5$ pyramidal structures and square fourfold coordinated ${\mathrm{C}\mathrm{u}\mathrm{O}}_4$ along with the ${\mathrm{C}\mathrm{u}\mathrm{O}}_6$ octahedral units, the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ planes would survive intact as illustrated in the right-hand structure of Fig. 1B. However, if the oxygen vacancies are in the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ planes one will have Cu–O chain structures rather than the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ planes as indicated in the left-hand structure of Fig. 1B. The 2D ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ planes, consisting of corner-shared ${\mathrm{C}\mathrm{u}\mathrm{O}}_4$ units, are considered the key structural elements of the traditional cuprate superconductors. They are the defining characteristic of these superconductors and imperfections in these layers are known to reduce $T_c$.

As noted, ${\mathrm{B}\mathrm{a}}_2$Cu${\mathrm{O}}_{4-y}$ is made under high pressure and temperature resulting in polycrystalline samples and the precise location of the oxygen vacancies in the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ plane remains open. Nevertheless, the remarkably high $T_c$ of this highly overdoped cuprate with its short Cu apical O separation and its O vacancies in the ${\mathrm{C}\mathrm{u}\mathrm{O}}_2$ plane suggest that it is a member of a different branch of high-$T_c$ cuprate materials, which challenges the basic tenants of many high-T_c theories.