The Ba
Ru
O
single crystal was grown in a mirror furnace (
HKZ,Scidre). Therefore, the starting materials were prepared by thorough grinding of BaCO
and RuO
with a Ba to Ru ratio of 2:1 and sintering this mixture at 1000
C for one day in air. The crystal was grown with a speed of 6 mm/h under 16 bar of an Argon/Oxygen atmosphere (with a ratio of 1:6). Thus, black, shiny single crystals of Ba
Ru
O
could be finally obtained, see
Figure 1. As known for other ruthenates, also during the growth of Ba
Ru
O
a larger evaporation of Ru oxide could be observed during the crystal growth which explains the finally obtained Ba:Ru ratio in the grown single crystal. Also other Ba:Ru ratios of 2:1.1 and 2:1.05 have been tried for a compensation of the Ru loss, but only the ratio of 2:1 results in the growth of large impurity-free single crystals of Ba
Ru
O
. Its composition has been confirmed with EDX (energy dispersive X-ray spectroscopy) measurements in a scanning electron microscope yielding a molar Ba:Ru ratio of 2.07(8):1.00(5) which is within the error bars in agreement with the composition of Ba
Ru
O
.
For powder X-ray diffraction (XRD) measurements parts of the grown single crystals have been ground into fine powders. The XRD measurements have been performed using Cu
radiation on a
Bruker D8 Discover A25 powder X-ray diffractometer. The
FullProf program package [
18] was used for Rietveld refinements, see
Figure 2.
Using Mo
radiation single crystal X-ray diffraction measurements have been performed on a
Bruker D8 VENTURE single crystal X-ray diffractometer equipped with a bent graphite monochromator and a
Photon III detector. A crystal of roughly 10
m size has been measured, see
Figure 3. Due to the small size of the sample we were able to find an almost untwined single crystal with regard to the possible orthorhombic twin domains. Nevertheless, the inversion twin could not be avoided. (For distinctly larger sample sizes we could always observe the appearance of all six possible twin domains that might appear for this structure.) The
Jana2006 program suite [
19] was used for the crystal structure refinement. The refinement was based on
F(obs)
with an instability factor of 0.01. For this tiny sample no extinction correction was necessary. The volume fractions of the main domain (
id) and the corresponding twin domain obtained by inversion (
-id) amount to 49(3)% and 40(1)% whereas the remaining four twin domains related to rotations (plus inversions) are very small (i.e., 1.6(1.6)%, 2.4(1.6)%, 2.5(1.6)% and 4.1%) with a volume fraction almost comparable to their error bars. The reduced amount (size) of twin domains in this sample with pseudo-hexagonal crystal structure underlines the reliability of the refinement results. Only for the heavier atoms Ba and Ru the anisotropic displacement parameters
have been refined. For the oxygen atoms
has been refined. Goodness of fit, R- and weighted R-values and the obtained structural parameters and bond lengths are listed in
Table 1,
Table 2,
Table 3 and
Table 4.
Structure
The crystal structure of Ba
Ru
O
alternatingly consists of layers of single Ru ions and of layers of dimerized Ru-ions located in face sharing octahedra which are denoted in
Figure 4 as layers ’A#’ (Ru single layers) and ’B#’ (Ru double layers) respectively. Within these Ru-layers the Ru ions arrange in a pseudohexagonal manner, see
Figure 5. Two third of the Ru ions within the ’A#’ layers have a trigonal bipyramidal oxygen coordination—see
Figure 5a—with an enhanced oxidation state compared to the other Ru ions, see
Table 5. These Ru ions form a distorted Kagome lattice. The other Ru ions are octahedrally coordinated by the oxygen ions, see
Figure 5a and are situated in the free space of the Kagome lattice. This difference in the Ru sites becomes also apparent in their Ba coordination which can be seen in
Figure 5b. The RuO
octahedra are located in quite compact RuBa
octahedra (
red) wheras the RuO
bipyramids are situated in much larger RuBa
icosahedra (
yellow). The dimerized Ru ions within the ’B#’ layers all form Ru
O
polyhedra consisting of two face sharing RuO
octahedra. The corresponding Ru-Ba polyhedra (
orange) exhibit more uniform sizes compared to the ones within the ‘A#’ (single-)layers, see
Figure 5d. It is the different stacking (‘A1-B1-A2-B2-A3-B1
-A4-B2
-A1’) of these layers that is responsible for the large value of the lattice constant in
b-direction, compare
Figure 4. Note, that also the heavy Ba-ions follow this stacking.
For many Ba-containing compounds the Ba ions are twelve-fold coordinated by anions as is observed for Ba7, Ba10 and Ba13 in Ba
Ru
O
. An example in literature is BaNiO
[
20]. But also a plethora of other oxygen coordinations have been reported for Ba in literature. The Ba1 ions in our new crystal structure exhibit an 8-fold oxygen coordination that can be also found e.g., in Ba
OSiO
[
21]. A different 8-fold oxygen coordination in our new crystal structure can be observed for the Ba4, Ba5 and Ba6 ions. This latter oxygen coordination was reported for Ba
(BO
)
[
22]. Also an octahedral anionic coordination is known for Ba (although an octahedral oxygen coordination is much more typical for somewhat smaller but still very large Sr
ions) and has been reported for Ba
CuO
Cl [
23]. The oxygen coordination of the Ba ions is also summarized in
Table 5. The Ba-O polyhedra are characterized by three values given in brackets. The central values denotes the amount of oxygen ions that are situated very roughly in the plane of the Ba-ion (perpendicular to the pseudohexagonal axis). The first and the last values denote the number of oxygen ions far above and below this plane. [e.g., an octahedral coordination would be abbreviated by (3-0-3).] A cif-file is provided as a
Supplementary Materials File for a detailed visualization.
The nominal Ru oxidation state should be slightly higher than 5+ according to the composition of Ba
Ru
O
. Note, that the absolute values of the bond valence sums (listed in
Table 5) are not necessarily equal (or even close) to the real oxidation states, especially also because of the huge complexity of the crystal structure of Ba
Ru
O
. Note that there is a scattering in the BVS values of the Ba
ions (oxygen ions) of
(
) around 2.085+ (1.888−) which provides an indication for the reliability of the BVS in this very complex structure. A similar scattering range of
of the BVS values around 4.305+ can be observed for the Ru ions. If one would take the values serious, the bond valence sum (BVS) formalism would indicate that the oxidation state of the Ru ions with trigonal bipyramidal oxygen coordination is higher than that of the other all octahedrally oxygen coordinated Ru ions. Such a higher oxidation state of the Ru ions with trigonal bipyramidal oxygen coordination would be in agreement with the total oxygen composition. Note, that a trigonal bipyramidal oxygen coordination is known to exist for Ru ions e.g., in K
Ru(OH)
O
[
25]. Nevertheless, the scattering of all the BVS values is somewhat enhanced (see above) and future X-ray absorption spectroscopy measurements are required to study the real Ru valencies in Ba
Ru
O
.
Besides floating zone grown single crystals also conventional solid state reaction under ambient pressure conditions (air) was successful. We conclude that Ba
Ru
O
is the ambient pressure phase within the phase diagram. Note, that high pressures of several GPa are needed to stabilize Ba ruthenates with layered perovskite (K
NiF
) structure for Ba to Ru ratios of 2:1 [
17]. For smaller Ba:Ru ratios (like 2:1.1), we observed the formation of a Ba
Ru
O
impurity phase that is growing under the same growth conditions during our floating zone growth.