Solid State Ionics 59 ( 1993) 93-98 North-Holland
Barium nickel oxide with BaCu02 type structure Reiner Gottschall and Robert SchGllhorn Technische Universitdt Berlin, Institut ftir Analytische und Anorganische Chemie, StraJe des 17. Juni 135, 1000 Berlin 12. Germany Received 3 1 July 1992; accepted for publication 3 1 August 1992
Under appropriate conditions the reaction of NiO,,, with BaCOs leads to the formation of a cubic barium nickel oxide with the composition Ba7Ni60i2C03. X-ray powder studies revealed that this phase (“c-BaNiOz”) crystallizes in the unique structure type ofbarium copper oxide. Since BaNiOz has an orthorhombic structure if prepared by conventional synthesis we conclude that the cubic modification is stabilized by the CO:- ions. At 1050°C c-BaNiOz is unstable on extended exposure to nitrogen atmosphere (formation of orthorhombic BaNiOz); in oxygen atmosphere the formation of BaNi03+ is observed.
1. Introduction Polynary oxocuprates with electropositive main group cations and copper in mixed valence states Cu( II) /Cu( III) have been found to exhibit outstanding electronic properties [ 1 ] and high oxygen ionic mobility at rather low temperature [ 2-5 1. One interesting question related to this subject is whether copper is unique with respect to these properties or can be replaced by other transition metal ions with mixed valence states at high electrochemical potential. Several investigations on oxonickelates with K2NiF4 type have been reported recently [ 6,7]. In this respect we were interested in the properties of simple ternary alkali and alkaline earth oxonickelates; in the course of studies on the optimization of the preparation of orthorhombic barium nickel (II) oxide BaNiOl characterized by NiOl chain units we observed the formation of a cubic phase. We report here on the properties and characterization of this compound which was found to belong to an unusual structure type.
2. Experimental Barium carbonate and black nickel oxide (analytical grade) were mixed in a molar ratio 1: 1 and ground in a ball mill. Pressed pellets of the mixture
heated for various temperatures and times in nitrogen atmosphere. For the preparation of BaCuOZ we used BaC03 and CuO ( 1: 1 ratio) which were heated in oxygen up to 950°C. The structural characterization of the products was performed by X-ray powder data at 300 K (ENRAF Nonius diffractometer with primary monochromator and INEL position sensitive detector, transition mode, Cu Ku radiation ). The calculated peak positions and intensities were obtained by the computing program “lazy pulverix” [ 81. A thermal analysis unit DTA/TG (Netzsch STA 409) served for the determination of mass and thermal changes upon heating under different conditions. For the investigation of the magnetic properties an ac susceptometer (Lake Shore 7000) was used; analytical data were obtained by EDX measurements ( Hitachi S 520) and by redox titration.
3. Results and discussion The established way to obtain BaNiO, is based on the reaction of BaCOS or BaO with green NiO at 9501000°C under oil pump vacuum (100 Pa) [9,10]. The black product has an orthorhombic unit cell: the Ni/O sublattice is characterized by bent chains of edge sharing square planar NiO,-units (fig. 1). In deviation of this well-known method we tried to pre-
0167-2738/93/% 06.00 0 1993 Elsevier Science Publishers B.V. All rights reserved.
R. Gottschall. R. Schiillhorn /Barium nickel oxide with BaCuOz type structures
pared to vacuum) we increased the temperature to 1050°C and the sintering time to 48 h. As the product of the reaction, a black polycrystalline material is found, which dissolves easily in dilute mineral acids under release of a gaseous component. The X-ray powder diagram did not exhibit the characteristic pattern of orthorhombic BaNiOz as expected. A careful analysis of the diffraction lines yielded the surprising result that the line pattern observed is almost identical with that of the known cubic oxocuprate BaCu02 [ 121 (fig. 2). The diffractograms of BaCuOz and of the cubic barium oxonickelate (described in the following as “cBaNiO;‘) are shown in figs. 3 and 4, and table 1). Using the structure data reported by Miiller-Buschbaum et al. [ 121 the peak positions and intensities for BaNiOz with BaCuOz structure were calculated; it was assumed that the Ni atoms replace the copper ion positions completely and that the Ba 2a positions are vacant [ 13 ] . The result of this calculation (fig. 5) essentially demonstrates the equivalence of peak
Fig. I. Structure scheme of orthorhombic BaNiOz (a= 5.73 A, b=9.20Ac=4.73A) (ref. [ll]).
pare the compound in flowing nitrogen atmosphere using BaC03 and black NiO. Because of the higher stability of BaC03 under these conditions (as com-
Fig. 2. Structure scheme of BaCuO*; coordination polyhedra of the Cu/O sublattice (cubic a= 18.27 A) (ref. [ 121)
R. Gottschall, R. Schiillhorn /Barium nickel oxide with BaCu02 type structures Table 1 Peak positions oft-BaNiOZ *= nickel oxide. hkl
200 330 411 420 332 521 440 530 433 442 600 611 532 620 444 710 543 550 633 552 721 732 651 800 81 1 741 554 822 831 743 750 *(200 752 840 921 761 655 754 851 930 932 763 853 941 770 772 101 1 950 943 1022 666
Table 1 (continued). (cubic, a= 18.128
A, 1= 1.54056 A); hkl
4.053 3.865 3.311 3.205 3.109
4.054 3.865 3.310 3.205 3.109
5.8 7.8 8.2 13.4 92.2
2.866 2.616 2.563
2.866 2.616 2.564
13.0 16.7 7.2
d ( talc. )
2.266 2.23 1
2.266 2.23 1
2.088 2.052 2.027 1.955
2.052 2.027 1.955
4.4) 14.2 6.7 12.2
103 1 765 952 855 871 864 1040 961 103 3 1110 954 873 963 1121 105 1 880 11 32 972 776 875 1141 965 981 1150 1220
2 Theta 1.831
Fig. 3. X-ray diffraction
(cubic a = 18.128
positions and intensities between experimental and calculated values. A comparison of the intensities of representative lines of the observed and calculated diffraction diagrams of BaCuOz and c-BaNi02 (ta-
R. Gottschail, R. Sch6lihorn /Barium nickel oxide with BaCu02 type structures Table 2 Comparison of intensities Ni02 and BaCuO,.
Fig. 4. X-ray diffraction
pattern of BaCuOz (cubic a= 18.272 A).
ble 2) yields further support for a substitution of copper by nickel in the structure of BaCuOz since the intensity change for calculated and experimental values shows the same tendency. This observation is valid for all reflections observed. It is concluded therefore that c-BaNiO, exhibits a BaCu02 type structure. The X-ray diagram of c-BaNiO, given in fig. 3 shows that the material still contains traces of NiO; efforts to avoid this minor impurity by optimizing the preparation conditions have not been successful. The observation of a gaseous component upon dissolution of BaCuOz in aqueous acids as mentioned above was found to be due to (i) a slight oxygen excess in the lattice (0.25 wt% as determined by iodometric titration) leading to the formation of di-
2 Theta Fig. 5. Calculated
200 433 530 600 442 532 61 444 660 853
18.1 27.6 29.9
18.0 25.8 24.6
16.8 39.5 52.9
I 20.3 26.2 28.4
oxygen and (ii) to the release of CO* (detected by BaCOS precipitation upon reaction of the gas received with aqueous Ba(OH),). For the CO1 formation two explanations are possible: (i) presence of BaCO, as an impurity phase not detectable by Xray diffraction and (ii) CO:- presence as lattice constituent of c-BaNiO*. In order to answer this question DTA/TG measurements were performed with c-BaNi in nitrogen atmosphere. We observed that the compound transforms at 1050°C (these are the conditions of the preparation!) quantitatively to orthorhombic BaNiO, ( o-BaNiOz) as confirmed by X-ray diffraction data of the product. The mass loss due to the COZ release amounts to 3.38 wt%. If we assume that a potential BaC03 impurity is smaller than 5 wt% we can estimate the CO:- content in cBaNiO, to be in the order of 2.4-3.4 wt%. This value is in agreement with EDX data which yield a ratio Ba: Ni of 1.16. With the assumption that the Ba excess is directly related to the carbonate content the composition of the compound can be described by Ba7Ni60&03. In a subsequent literature search based on this composition we found a publication by Krischner et al. [ 141 who report on a compound 9BaNi0,*2 BaCOJ with the cubic lattice parameter a= 18.08 A; no structure model was proposed for this phase. If the composition of the phase described by us is transformed to the notation used by Krischner i.e. 6BaNi02 * BaC03 then we must conclude on grounds of a comparison of the lattice parameters and the
R. Gottschall, R. Schiilhorn /Barium nickel oxide with BaCuO, type structures
48 h, N2
48 h. N2 lOSO
I llSO°C (c_BaNiOz
30 h, lOSO
I BaNi OS_, Fig. 6. Scheme of formation and reactions of c-BaNi02.
similarity in composition that the two phases are closely similar. The small difference in composition may originate from a finite phase width of the compound. Replacing BaCO, by BaO in the synthesis step of c-BaNiOz under otherwise identical reaction conditions leads only to the formation of common oBaNi02. The DTA and X-ray investigations show that c-BaNiOz is melting at 1150°C in nitrogen atmosphere (under partial decomposition and CO1 release); the solidified melt consists of o-BaNiO* with traces of NiO. Preliminary measurements of the magnetic susceptibility of c-BaNiO, exhibits paramagnetic properties in the temperature range of 12- 150 K. The minor oxygen excess (6 BaNiO,+x* BaC03, x= 0.04, as calculated from analytical data) may be due to the fact that the starting material was black NiO, +x and due to traces of oxygen in the nitrogen atmosphere; in any case it must be concluded that the oxygen affinity of the compound is rather high. When c-BaNiOz is subjected to extended high temperature treatment under the condition of the preparation ( 1050” C, nitrogen atmosphere) one observes a quantitative transformation to o-BaNiOz in the course of 30 h; a limited reaction time is thus essential for the preparation procedure used. This instability is also apparent in the reaction with molec-
ular oxygen. At temperatures up to ca. 630°C cBaNiOz is inert towards 0,; at higher temperatures, however, the compound takes up oxygen and transforms into BaNiO,_, with BaNiO, type structure (hexagonal lattice parameters a= 5.584 ( 1) A, c=4.842( 1) A). Due to the contamination of the product with NiO the x value cannot be determined exactly; we estimate, however, that x is greater than 0.5. Thus the reaction with oxygen yields the same product as the reaction of o-BaNiO, with oxygen which has been described earlier [ 93. A summary of the processes investigated is given in fig. 6.
4. Conclusions According to the results of the present study the cubic nickel oxide c-BaNiOz can be considered as a carbonate ion stabilized variant of the unusual BaCuOz type structure. On a comparison of the chemical properties of BaCuOz and c-BaNiOz it becomes apparent that the behaviour of these phases is quite different: at 950 “C BaCuO, is stable in oxygen atmosphere but decomposes rapidly in nitrogen, while c-BaNiO, is unstable in oxygen atmosphere with respect to the formation to BaNi03_y but can be prepared in nitrogen atmosphere. A similar difference is observed in oxygen atmosphere at low
R. Gottschall, R. SchiiNhorn /Barium nickel oxide with BaCuO, type structures
temperatures: whereas c-BaNiOl exhibits no temperature dependent oxygen stoichiometry up to 630°C significant changes in oxygen content are found in the case of BaCuO, in the range from 300 to 1000°C [15,16]. The difference in behaviour cannot simply be explained by the different affinity of Ni and Cu, respectively, to oxygen. The experimental observations can, however, be understood on the assumption that those positions in the lattice of c-BaNiOl are occupied by CO:- which are correlated with the variable oxygen content of BaCuO*. Carbonate ions would thus compete with oxygen in the occupation of specific lattice sites depending upon ~(0~) and p(CO*). A proof of this model is difficult at present because of the uncertainties of positions and occupation factors of the excess oxygen in the complex BaCu02 structure [ 17,18 1. Obviously we have to consider also the possibility that BaCuOz may contain small amounts of CO:as a lattice constituent. It is of interest to notify that several reports have appeared recently on the existence of oxocuprates containing CO:- ions [ 19,201.
Acknowledgement This work was supported by the Bundesministerium fur Forschung und Technologie (BMFT).
References [ 1 ] J.G. Bednorz and K.A. Mtlller, eds., in: Superconductivity (Springer, Berlin, New York, Tokyo, 1990).
 G. Bums, F.H. Dacol, C. Feild and F. Holtzberg, Solid State Commun. 75 (1990) 893.  J.L. MacManus, D.J. Fray and J.E. Evetts, Physica C 190 (1992) 511. ] G. Cannelli, R. Cantelli, F. Cordero, M. Ferretti and F. Trequattrini, Solid State Commun. 77 ( 199 1) 429. ] H. Eickenbusch, W. Paulus, E. Gocke, J.M. March, H. Koch and R. Schollhorn, Angew. Chem. Int. Ed. Engl. 26 (1987) 1188;Angew.Chem.99(1987) 1201. ] Z. Kakol, J. Spalek and J.M. Honig, J. Solid State Chem. 79 (1989) 288. [ 7 ] J.C. Grenier, A. Wattiaux, J.P. Doumere, P. Dordor, L. Foumts, J.P. Chaminade and M. Pouchard, J. Solid State Chem. 96 (192) 31.  K. Yvon, W. Jeitschko and E. Parthe, J. Appl. Cryst. 10 (1977) 73. [ 9 ] M. Arjomand and D.J. Machin, J. Chem. Sot. Dalton Trans. 11 (1975) 1055. [lo] H. Krischner, K. Torkar and B.O. Kolbsen, J. Solid State Chem. 3 (1971) 349. [ 1 I] J.J. Lander, Acta Cryst. 4 (1951) 148. [ 121 R. Kipka and Hk. Milller-Buschbaum, Z. Naturforschung 32b (1977) 121. [ 131 W. Gutau and Hk. Mtiller-Buschbaum, J. Less Common. Met. 152(1989)Lll. [ 141 K. Torkar, H. Krischner and E. Will, M. Chem. 100 (969) 825. [ 15 ] S. Erikson, L.G. Johannsson, L. Borjesson and M. Kakhana PhysicaC 162-164 (1989) 59. [ 161 I.N. Dubrovina, R.G. Zakharov, G.K. Moiseev, E.G. Kostitsyn, O.V. Yavoiskaya, A.V. Antonov, V.F. Balakirev and N.A. Vatolin, Supercond.: Physics, Chem. Techn. (1989) 115. [ 171 E.F. Paulus, G. Miehe, H. Fuess, I. Yehia and U. Liichner, J. Solid State Chem. 90 ( 199 1) 17. [ 181 M.T. Weller and D.R. Lines, J. Solid State Chem. 82 (1989) 21. [ 191 D. Vemooy and A.M. Stacy, J. Solid State Chem. 95 ( 1991) 270. [ 201 A.R. Armstrong and P.P. Edwards, J. Solid State Chem. 98 (1992) 432.