Journal of Solid State Chemistry 149, 137}142 (2000) Article ID jssc.1999.8511, available online at http://www.idealibrary.com on Crystal Structure and Compressibility of Ba4Ru3O10 A. H. Carim,*,-,1 P. Dera,-,? L. W. Finger,- B. Mysen,- C. T. Prewitt,- and D. G. Schlom* *Department of Materials Science and Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802-5005; -Geophysical Laboratory, Carnegie Institution of Washington, Washington, D.C., 20015-1305; and ?Department of Crystal Chemistry, Adam Mickiewicz University, 60}780 Poznan, Poland Received June 17, 1999; accepted September 21, 1999 The crystal structure of Ba4Ru3O10 has been determined by single-crystal X-ray di4raction at room pressure. From re5nements to R 5 0.0203 at room temperature and ambient pressure, the material is orthorhombic with space group Cmca (space group No. 64) and has lattice parameters of a 5 5.7762(15) As , b 5 13.271(4) As , and c 5 13.083(3) As . The unit cell thus has a volume of V 5 1002.9(8) As 3 and contains four formula units (Z 5 4). Ba4Ru3O10 is therefore of higher symmetry than the previously reported monoclinic structure based on powder X-ray data. It is isostructural with the quaternary oxides Ba4 (Ti, Pt)3 O10 and Ba4Ir2AlO10 and the ternary 6uorides Cs4M3F10 (M 5 Mg, Co, Ni, Zn). Kinked chains of RuO6 octahedra run along the c direction, consisting of sets of three face-sharing units joined at the corners of the end units to additional similar sets. The two distinct Ba sites show 10-fold and 11-fold coordination. Compressibilities and bulk modulus have been determined from lattice parameter variations at pressures up to 5.4 GPa. No phase transition was observed up to this pressure. Compressibility is greatest along the c axis and the bulk modulus obtained from a weighted 5t to a Vinet equation of state is 113.3(47) GPa. ( 2000 Academic Press Key Words: ruthenates; crystal structure; compressibility; bulk modulus; high pressure. INTRODUCTION Although numerous barium ruthenates have been reported (BaRuO (1}4), BaRu O (4), BaRu O (5), 3 4 9 6 12 Ba RuO (3, 4, 6, 7), Ba RuO (3, 4), Ba Ru O (8), 2 4 3 5 4 3 10 Ba Ru O (8), Ba RuO (3), and Ba RuO (3)), crystal 5 3 12 4 6 9 11 structures have not been established for all of these compounds, and relatively few (speci"cally BaRuO , 3 BaRu O , and Ba Ru O ) have been synthesized and 6 12 5 3 12 analyzed in single-crystal form. The phases with Ba Ru O stoichiometries are particularly interesting n`1 n 3n`1 because of the possibility that they may form in layered 1To whom correspondence should be addressed. perovskite (Ruddlesden}Popper) structures (9}11), although this appears to require high pressures (7). In this report we describe single-crystal X-ray di!raction of Ba Ru O at pressures up to 5.4 GPa. The crystal struc4 3 10 ture is found to be similar to, but of higher symmetry than, the monoclinic structure derived earlier from powder diffraction (8). No phase transitions to the layered perovskite or other forms are observed within this pressure range. EXPERIMENTAL PROCEDURE Sample Synthesis The phase described here was obtained as crystals within multiphase samples, in the course of investigating ternary compounds in the Ba}Ru}O system. Initial powders of BaCO (Alfa Aesar, 99.997% metals basis) and RuO (Alfa 3 2 Aesar, 99.95% metals basis) were combined so as to provide a Ba : Ru ratio of 2 : 1. The mixture was ground (dry) in an agate mortar for 45}60 min and then pressed into a pellet at &180 MPa and 1203C for 10}15 min. Pellets were then suspended in a platinum wire cage and annealed for 95 h at 14003C and a f (O )"10~3 (for the single crystal analyzed) 2 or at 13003C and a f (O )"10~4 (giving similar results); 2 oxygen fugacity was controlled by mixing #owing CO and CO gases and monitored with an yttria-stabilized ZrO 2 2 sensor. The low oxygen fugacity was used to try to cause the reaction to proceed without loss of ruthenium via oxidation to volatile higher oxides such as RuO ; in the absence of 4 su$cient data for Ba}Ru}O systems, the f (O ) values used 2 were selected on the basis of extrapolations from earlier work in the Sr}Ru}O system (12). The compositions of the barium ruthenate grains in polished sample sections were analyzed by quantitative wavelength-dispersive spectroscopy in an electron microprobe. The crystals were homogeneous and found to have a nearly stoichiometric composition of Ba Ru 4.04(2) 2.98(1) O (averages and standard deviations from 7 grains, as10 suming 10 oxygen atoms per formula unit). The crystal used and those analyzed by a microprobe were chosen from 137 0022-4596/00 $35.00 Copyright ( 2000 by Academic Press All rights of reproduction in any form reserved. 138 CARIM ET AL. TABLE 1 Crystallographic Data, Collection Conditions, and Re5nement Parameters for Ba4Ru3O10 Crystallographic Data Orthorhombic (Cmca, No. 64) a"5.7762(15), b"13.271(4), c"13.083(3) Volume (As 3) 1002.9(8) Z 4 Calculated density (g cm~3) 6.71 Symmetry Cell parameters (As ) Crystal shape and size Di!ractometer Detector Source of radiation Data Collection (¹"203C) Parallelepiped, approx. 20]50]60 lm3 Bruker AXS P4 four-circle SMART 1000 CCD (1K) X-ray tube w/graphite monochromator (MoKa) 1650 (in 5 scans) 0.3 0.5 0.72 2251 684 No. of frames acquired Scan step size (3) Average peak half-width (3) Minimum d (As ) hkl No. of measured re#ections No. of re#ections used for re"nement No. of independent re#ections 684 Transmission coe$cient range 0.376}0.550 Structural Re"nement SAINT (Bruker's, Data processing) XPREP (Bruker's, absorption correction) SHELXS (structure solution) SHELXL (anisotropic re"nement) Extinction parameter 0.001689 No. of re"ned parameters 50 Merging R factor R 0.0764 */5 Reliability factor R 0.0203 (595 data with F '4p (F )) 1 0 0 0.0276 (all data) Reliability factor wR 0.0638 (all data) 2 Programs used sections away from the contact points of the Pt wire cage with the sample, and thus it is believed that little or no Pt is incorporated therein; Pt was not detected in energy-dispersive spectroscopy (EDS) or WDS scans of any grains. Ambient-Pressure Measurements and Structure Determination Data on the crystal selected for structure determination, the unit cell parameters measured at ambient pressure, and the experimental parameters are given in Table 1. The structural determination was carried out with an automated Bruker AXS P4 system equipped with a SMART 1000 CCD area detector. The instrument employs MoKa radiation and the generator was set at 50 kV and 40 mA. A hemisphere of data to 0.72 As was collected with 0.33 frames, with 88.1% coverage of this reciprocal space region and a mean I/p"13.32. Counting time for each frame was 10 s. The data were corrected for geometrical distortion, dark current, and #ood-"eld e!ects. Integrated intensities were extracted using Bruker software and the data were corrected for Lorentz and polarization e!ects. Data merging, both for redundant re#ections and the symmetrical equivalents, and an absorption correction using an ellipsoid model were performed with the program XPREP from the SHELXTL package supplied by Bruker AXS. Occupancies of the atomic sites have not been re"ned. A stoichiometric composition (Ba Ru O ) was assumed for the structural re"ne4 3 10 ment and provided a satisfactory "t. High-Pressure Measurements Elevated pressures were obtained by compressing the same single crystal described above within a modi"ed Merrill-Bassett diamond anvil cell (13). A 4 : 1 mixture of methanol : ethanol was used as a hydrostatic pressure medium. Pressure was calibrated using #uorescence from ruby chips enclosed within the cell. Lattice parameter measurements for the sample at a pressure within the diamond anvil cell were obtained on a Picker four-circle di!ractometer. The 8-re#ection positions (13) of 11}15 re#ections with 12.13(2h(25.43 were "tted to determine a, b, and c. The reported values are those obtained via constrained "ts in which the interaxial angles were presumed to be 903. Unconstrained "ts gave values within 1p (the estimated standard deviation) of 903 for each of the angles in most cases, with all but one within 2p. Full structural determinations were not repeated at high pressures. RESULTS AND DISCUSSION Crystal Structure Re"nement of the single-crystal data indicates an orthorhombic unit cell with the Cmca space group (No. 64) and lattice parameters of a"5.7762(15) As , b"13.271(4) As , and c"13.083(3) As . The atomic coordinates and isotropic thermal parameters are listed in Table 2. Anisotropic thermal parameters are provided in Table 3. TABLE 2 Atomic Coordinates and Isotropic Thermal Parameters for Ba4Ru3O10 Atom Site Ba(1) Ba(2) Ru(1) Ru(2) O(1) O(2) O(3) O(4) 8f 8f 4a 8f 8e 8f 16g 8f x 0 0 0 0 1 4 0 0.2708(7) 0 y z b (As 2) 0.23978(4) 0.53548(4) 0 0.87525(5) 0.3774(4) 0.0350(5) 0.3906(3) 0.7306(5) 0.11127(4) 0.13890(3) 0 0.14957(5) 1 4 0.1517(5) 0.0346(3) 0.1462(5) 0.90(2) 0.73(2) 0.54(2) 0.51(2) 1.08(10) 0.79(9) 0.78(6) 1.36(11) CRYSTAL STRUCTURE AND COMPRESSIBILITY OF Ba Ru O 4 3 10 TABLE 3 Anisotropic Thermal Parametersa for Ba4Ru3O10 Atom ; 11 ; 22 Ba(1) Ba(2) Ru(1) Ru(2) O(1) O(2) O(3) O(4) 0.0106(3) 0.0098(3) 0.0076(4) 0.0066(3) 0.013(3) 0.013(3) 0.009(2) 0.020(3) 0.0069(3) 0.0078(3) 0.0058(4) 0.0059(3) 0.016(3) 0.006(3) 0.009(2) 0.006(3) ; 33 ; 12 ; 13 ; 23 0.0167(3) 0 0 !0.0001(2) 0.0101(3) 0 0 !0.0006(2) 0.0072(4) 0 0 0.0004(3) 0.0070(3) 0 0 !0.0002(2) 0.012(3) 0 0.005(2) 0 0.011(2) 0 0 !0.000(2) 0.011(2) !0.002(1) !0.003(2) 0.003(1) 0.026(4) 0 0 !0.002(2) a Temperature factor"exp (!+ + 2n2h h a* a* ; ) , where h is a Miller i j i j ij i i j index and a* is the length of the ith reciprocal axis vector. i The structure reported here is of higher symmetry than the monoclinic (P2 /a) structure reported earlier (8) on the 1 basis of Rietveld re"nement of X-ray powder di!raction data. The reported monoclinic unit cell (a"5.776 As , b"13.076 As , c"7.234 As , b"113.533) is nearly identical to the primitive cell of the Cmca structure that we observed (a"5.7762(15) As , b"13.083(3) As , c"7.2373(19) As , b" 113.528(4)3).2 Calculated X-ray powder di!raction patterns based on the two structures are di$cult to distinguish from one another, as shown in Fig. 1. The intensities of the re#ections, the interaxial angles calculated from unconstrained re"nements at both ambient and high pressures, comparison to similar compounds, and the quality of the "nal re"nements all favor the orthorhombic assignment. The Inorganic Crystal Structure Database (ICSD) listing (14) based on the earlier monoclinic structural solution notes that whereas the temperature factors are self-consistent within the report, they are not all plausible or meaningful. Problems with re"ned thermal parameters are common when a structure is incorrectly re"ned in such a subgroup. The monoclinic form described in the earlier work on Ba Ru O (8) is isostructural with that determined for 4 3 10 Ba Ir O (15). Other reports on similar compounds 4 3 10 indicated an orthorhombic Cmca or Cmc2 unit cell 1 for Ba (Ru, Mn) O (16), Ba (Ti, Pt) O (17), 4 3 10 4 3 10 Ba (Ti, Ir) O (18), and Ba Ir AlO (19) and the ternary 4 3 10 4 2 10 139 #uorides Cs M F (M"Mg, Co, Ni, Zn) (20). Thus, al4 3 10 though the Cmca symmetry had not previously been observed for a ternary compound of A B O stoichiometry, 4 3 10 its presence in the case of Ba Ru O is consistent with 4 3 10 observations from the closely-related quaternary oxide and ternary #uoride compounds. The structural building blocks closely resemble those reported in the earlier work. As shown in Fig. 2, the structure contains short chains of three face-sharing RuO oc6 tahedra that are connected to other such groups at their terminal corners. Atoms of Ba are 10-fold and 11-fold coordinated by oxygen. There are, however, some minor di!erences. For example, the four O(3) atoms connecting the central RuO octahedron in each group to the adjacent 6 octahedra are equidistant from the central Ru(1) cation. The longest Ba}O bond distance of 3.661 As reported in the earlier work appears to be an error; for the monoclinic unit cell proposed there (8), that speci"c distance is actually 2.661 As and thus the longest Ba}O bond is 3.333 As . The Ba}O bond distances in our revised structure are rather di!erent and are listed in Table 4; they range from 2.592 to 3.391 As . Kafalas and Longo (7) were able to synthesize the end member (n"1) of the barium ruthenate Ruddlesden}Popper series (9}11), layered Ba RuO , under conditions of 2 4 P"6.5 GPa and ¹"12003C. They had previously suggested that the other (n"R) end member, BaRuO in the 3 perovskite form, would be expected to stabilize at approximately 12 GPa (7, 21). This would indicate that a phase transition for Ba Ru O might be expected between 6.5 4 3 10 and 12 GPa. In work by others, however, the synthesis of 2Values listed are for a monoclinic re"nement of our current data; they are within measurement error of values obtained from transformation of the derived orthorhombic parameters. The transformation matrix for conversion of the C-centered orthorhombic unit cell axes to the primitive monoclinic unit cell axes is C 1 0 0 0 !1 1 2 2 0 !1 0 D . FIG. 1. Calculated powder X-ray di!raction patterns for (a) the orthorhombic Cmca structure for Ba Ru O derived in the present work 4 3 10 and (b) the monoclinic (P2 /a) structure reported earlier (8) (CuKa source 1 radiation). 140 CARIM ET AL. FIG. 2. (a) Projected view of the crystal structure of Ba Ru O aligned nearly along [100] and (b) the crystal structure viewed along [010]. The 4 3 10 Ru coordination polyhedra are drawn in as semitransparent blocks with the oxygen positions at the vertices. Cations are represented as small spheres (Ru) within and larger spheres (Ba) outside the octahedra. Several unit cells are shown to clarify the interrelationships of the structural units. In (b), only the nearest sets of RuO octahedra are shown; in projection, the subsequent three-unit face-sharing chains along the viewing direction fall in the 6 positions between those seen here. layered Ba RuO having lattice parameters consistent with 2 4 the layered perovskite structure was reported without any apparent applied pressure during the synthesis (3, 6), suggesting that stabilization of Ruddlesden}Popper structures in barium ruthenates might not require as high a pressure as reported by Kafalas and Longo. In any event, no phase transition was observed in Ba Ru O in our work at 4 3 10 pressures up to 5.4 GPa. One might expect, however, that even if this transition were to be favored, the bondbreaking necessary to transform from the low-pressure to the high-pressure form would destroy a single crystal and thus maintenance (or recovery) of long-range order might not be feasible without a simultaneous increase of the temperature. Anisotropic Compressibility Figure 3 shows the variation of the lattice parameters as a function of pressure. The c axis is the direction of greatest compressibility, as is reasonable upon examination of the structure. In this direction, one can imagine that the zig}zag chains of RuO octahedra can be readily folded in an 6 accordion fashion at the corner-sharing linkages without substantial distortion of the individual coordination polyhedra. This type of behavior is similar to the polyhedral tilting that constitutes a type of displacive phase transition in other materials (22). In contrast, the lowest compressibility is along the b axis, which requires forcing the sheets of RuO octahedra and the intervening layers of Ba atoms 6 141 CRYSTAL STRUCTURE AND COMPRESSIBILITY OF Ba Ru O 4 3 10 TABLE 4 Selected Interatomic Distances and Bond Angles for Ba4Ru3O10 Distance (As ) TABLE 5 Compressibility Data (Measured Lattice Parameters as a Function of Pressure) Angle (3) Pressure (GPa) Ru(1)}O(2) (]2) Ru(1)}O(3) (]4) Ru(2)}O(1) (]2) Ru(2)}O(2) Ru(2)}O(3) (]2) Ru(2)}O(4) Ru(1)}Ru(2) (]2) Ba(1)}O(1) (]2) Ba(1)}O(2) Ba(1)}O(3) (]2) Ba(1)}O(3) (]2) Ba(1)}O(4) (]2) Ba(1)}O(4) Ba(1)}O(4) Ba(2)}O(1) (]2) Ba(2)}O(2) (]2) Ba(2)}O(3) (]2) Ba(2)}O(3) (]2) Ba(2)}O(4) Ba(2)}O(2) 2.038(6) 2.017(4) 1.953(1) 2.120(6) 2.014(4) 1.920(7) 2.563(1) 2.952(4) 2.769(6) 2.731(4) 2.896(4) 2.927(1) 3.175(6) 3.391(6) 2.932(4) 2.893(1) 2.829(4) 2.926(4) 2.592(7) 2.740(6) O(3)}Ru(1)}O(3) O(3)}Ru(1)}O(3) O(3)}Ru(1)}O(3) O(3)}Ru(1)}O(2) O(3)}Ru(1)}O(2) O(2)}Ru(1)}O(2) O(1)}Ru(2)}O(1) O(1)}Ru(2)}O(2) O(1)}Ru(2)}O(3) O(1)}Ru(2)}O(3) O(3)}Ru(2)}O(3) O(2)}Ru(2)}O(3) O(4)}Ru(2)}O(1) O(4)}Ru(2)}O(3) O(4)}Ru(2)}O(2) Ru(2)}Ru(1)}Ru(2) 82.1(2) 97.9(2) 180.00 86.9(2) 93.1(2) 180.00 95.4(1) 88.6(2) 90.8(1) 170.8(2) 82.2(2) 84.8(2) 91.7(2) 94.8(2) 180.00 180.00 0 0.18(3) 2.08(5) 2.72(2) 3.79(3) 4.77(2) 5.37(3) a (As ) b (As ) c (As ) < (As 3) 5.7762(15) 5.7738(9) 5.7413(12) 5.7303(18) 5.7134(15) 5.6985(13) 5.6902(26) 13.271(4) 13.263(1) 13.208(2) 13.187(2) 13.156(2) 13.132(1) 13.114(3) 13.083(3) 13.073(2) 12.987(5) 12.961(4) 12.930(5) 12.887(3) 12.848(7) 1002.9(8) 1001.2(2) 984.8(5) 979.4(4) 971.9(4) 964.4(3) 958.8(6) for a, b, and c in As , P in GPa, and where R is the correlation coe$cient for the "t. The corresponding linear compressibilities are b "0.0031, b "0.0024, and b "0.0034 in a b c units of GPa~1. A listing of the measured lattice parameters at each pressure is included as Table 5. Equation of State and Bulk Modulus c"13.082(3)!0.045(5)P#0.0008(9)P2 R"0.9973, A plot of the unit cell volume as a function of pressure is included as Fig. 4. Fitting of this data to a Vinet equation of state (23, 24) with a "xed zero-pressure unit cell volume of < "1002.91 As 3 (the measured value at ambient P) indi0 cates a bulk modulus of K "112.1(35) GPa with a deriva0 tive of K@ "3.8(17). Use of a Birch}Murnaghan expression 0 (13) with "xed < yields the same values (K " 0 0 112.1(35) GPa, K@ "3.8(16)), as expected for a material that 0 is not highly compressible. Weighted Vinet and Birch}Mur- FIG. 3. Variation of the unit cell parameters (a, b, and c) as a function of pressure (P). The polynomial "ts referenced in the text are shown for each data set. FIG. 4. Variation of the unit cell volume (< ) as a function of pressure (P). The estimated standard deviations in volume are as indicated by the bars, and the data have been "t to a Vinet equation of state as discussed in the text. closer together. The lattice parameters decrease nonlinearly with positive curvature, as expressed below: a"5.7769(2)!0.0179(3)P#0.00032(7)P2 R"0.9999, b"13.2693(8)!0.0313(9)P#0.00049(18)P2 R"0.9997, 142 CARIM ET AL. naghan "ts in which < is allowed to vary also provide 0 nearly identical results (Vinet: < "1002.911 (1) As 3, 0 K "113.3(47) GPa, and K@ "3.4(20); Birch}Murnaghan: 0 0 < "1002.911(3) As 3, K "113.6(47) GPa, and K@ " 0 0 0 3.4(18)). The standard deviation for K@ is large but not 0 surprising considering the limited pressure range examined. ACKNOWLEDGMENTS We gratefully acknowledge the assistance of C. Hadidiacos with the electron microprobe analyses and N. Boctor with microprobe sample preparation. Discussions with and experimental assistance from H. Yang are also appreciated. The CCD di!ractometer was purchased with a NSF Grant EAR9725354, with matching funds from the W. M. Keck Foundation. This work was performed while A. H. C. was on sabbatical leave at the Carnegie Institution of Washington. REFERENCES 1. J. J. Randall and R. Ward, J. Am. Chem. Soc. 81, 2629 (1959). 2. P. C. Donohue, L. Katz, and R. Ward, Inorg. Chem. 4, 306 (1965). 3. M. I. Gadzhiev and I. S. Shaplygin, Russ. J. Inorg. Chem. 29, 1230 (1984). 4. T. L. Popova, N. G. Kisel, V. I. Krivobok, and V. P. Karlov, Sov. Prog. Chem. 48, 8 (1982). 5. C. C. Torardi, Mater. Res. Bull. 20, 705 (1985). 6. I. I. Prosychev and I. S. Shaplygin, Russ. J. Inorg. Chem. 25, 489 (1980). 7. J. A. Kafalas and J. M. Longo, J. Solid State Chem. 4, 55 (1972). 8. C.Dussarrat, F. Grasset, R. Bontchev, and J. Darriet, J. Alloys Compd. 233, 15 (1996). 9. S. N. Ruddlesden and P. Popper, Acta Crystallogr. 10, 538 (1957). 10. S. N. Ruddlesden and P. Popper, Acta Crystallogr. 11, 54 (1958). 11. J. M. Longo and P. M. Raccah, J. Solid State Chem. 6, 526 (1973). 12. C. Mallika and O. M. Sreedharan, J. Alloys Compd. 191, 219 (1993). 13. R. M. Hazen and L. W. Finger, &&Comparative Crystal Chemistry.'' Wiley, New York, 1982. 14. Inorganic Crystal Structure Database. Release 98/2. CD-ROM. Fachinformationszentrum Karlsruhe and Gmelin-Institut, Karlsruhe, 1999. 15. J. Wilkens and Hk. MuK ller-Buschbaum, Z. Anorg. Allg. Chem. 592, 79 (1991). 16. M. Neubacher and Hk. MuK ller-Buschbaum, Monatsh. Chem. 121, 635 (1990). 17. R. Fischer and E. Tillmans, Z. Kristallogr. 157, 69 (1981). 18. Hk. MuK ller-Buschbaum and M. Neubacher, Z. Anorg. Allg. Chem. 586, 87 (1990). 19. M. Neubacher and Hk. MuK ller-Buschbaum, Z. Anorg. Allg. Chem. 594, 133 (1991). 20. R. E. Schmidt, J. Pebler, and D. Babel, Eur. J. Solid State Inorg. Chem. 29, 679 (1992). 21. J. M. Longo and J. A. Kafalas, Mater. Res. Bull. 3, 687 (1968). 22. R. M. Hazen and L. W. Finger, Phase ¹ransitions 1, 1 (1979). 23. P. Vinet, J. Ferrante, J. H. Rose, and J. R. Smith, J. Geophys. Res. 92, 9319 (1987). 24. P. Vinet, J. Ferrante, J. R. Smith, and J. H. Rose, J. Phys. C 19, L467 (1986).