Synthesis and structure of a new oxynitride Ba3W2O6N2 † P. Subramanya Herle,a M. S. Hegde*a and G. N. Subbannab aSolid State and Structural Chemistry Unit and bMaterials Research Centre, Indian Institute of Science, Bangalore, 560012, India A new oxynitride Ba3W2O6N2 has been synthesised from the ammonolysis of Ba3W2O9. This compound crystallises in a hexagonal structure with a=5.993(2) and c=21.40(4) A ° .Transmission electron microscopy (TEM) studies were carried out to elucidate the structure of this new compound. IR and Raman data are consistent with the C3v site symmetry of the (WO3N)3- unit. This compound is isostructural with Ba3V2O8 reported in the literature. In recent years there has been a spurt of interest in exploring Raman spectra of the samples were recorded using a Spectra Physics SPEX 1403 double spectrometer (Ar-ion laser, l= new oxynitrides because of interesting structural and catalytic properties.1,2 It has been known that oxygen atoms can 514.5 nm) series 2000.Electron diVraction and microscopy (TEM) were carried out using a JEOL 200CX transmission substitute nitrogen atoms in monometallic nitrides due to similarity in their ionic radius as well as in polarizability.electron microscope to elucidate the microstructural features. Of late a number of new ternary nitrides and oxynitrides have been discovered, e.g., LiMN2 (M=Mo,W),3,4 Results and Discussion Mn2(MnTa3)N6-d (0d1)5 and Ln2Ta2O5N2 (Ln=lanthanide). 6 Bimetallic oxynitrides containing strongly electroposi- BaWO4 was heated in ammonia at diVerent temperatures. The product obtained when heated in ammonia at 900 °C for 12 h, tive elements such as LaTaO2N, Na3WO3N7 have substantial ionic character.Their limiting compositions have been was black and did not contain any starting material. The powder X-ray diVraction pattern of this sample (Fig. 1) was described by the normal rule of valency.8 There are reports on the ternary oxynitrides of rare-earths such as Ln2WO6-xNx9 diVerent from that of BaWO4.When heated in O2 atmosphere in the TPR system, this black material yielded only N2 as the and alkali-metal ions with tungsten. However, there are no reports on alkaline-earth metal and tungsten oxynitride phases. gaseous product. Since the sample was not highly crystalline and also because there was an indication of a small amount We wondered if partial substitution of N3- for O2- in the Ba–W–O system would yield any interesting material.Here of a W2N impurity phase, we were not sure whether the N2 came from the W2N phase alone or from any other unknown we report our work on the synthesis and structure of the new oxynitride Ba3W2O6N2. nitride product. To understand this reaction further, Ba2WO5 was heated in ammonia at 900 °C for 12 h.Interestingly, the powder pattern strongly resembles that of the black product Experimental obtained from BaWO4 heated in ammonia. A small amount of W2N impurity was also observed. From this it is clear that The previously reported oxides BaWO4, Ba2WO5, Ba3WO6 and Ba3W2O910 were synthesised by taking stoichiometric the starting oxide is decomposing to give rise to a new phase.When Ba3WO6 was heated in ammonia, a white compound amounts of BaO2 (99.5%, Fluka) and WO3 (99%, Koch Light) and heating in a muZe furnace and checking for product was obtained. The powder diVraction pattern of this compound contained the same unknown phase and BaO as impurity but formation by powder X-ray diVraction.About 1.5 g of these oxides was loaded in a quartz tube in an alumina boat. The there was no W2N impurity. The TPR of oxidation of this white product gave N2 as the gaseous product emanating samples were heated in flowing ammonia gas (flow rate ca. 120 ml min-1) at diVerent temperatures. The products were above 375 °C. The amount of N2 liberated was not as high as would be expected for a pure nitride phase indicating that analysed employing a JEOL-8P powder X-ray diVractometer (Cu-Ka radiation).The samples were heated in O2 atmosphere the new phase is an oxynitride. From this it is clear that the unknown phase has a Ba/W composition between that of in a temperature programmed reaction (TPR) system attached to a VG QXK300 quadrupole mass spectrometer11 to check Ba2WO5 and Ba3WO6.When Ba3W2O9 was heated in ammonia, a colourless single phase compound was obtained. for nitride phase formation. In a typical experiment about 200 mg of the sample was loaded and the reactor was evacuated This also liberated N2 on heating in O2. Experimental details to 10-6 Torr. Oxygen gas was admitted at ca. 20–25 mmol s-1 and the reactor was heated from 30 to 700 °C at a rate of 15 °C min-1 and the gaseous products were analysed.Nitrogen estimation was carried out using a home-built thermogravimetric analyser. EDX analyses of these samples were conducted using a Cambridge scanning electron microscope (SEM) S- 360, equipped with a LINK systems AN10000 X-ray analyser. FTIR spectra of the samples were recorded in polyethylene pellets in the range 100–700 cm-1 employing a Bruker IFS- 113V FTIR spectrometer and a Nicolet Impact 400D FTIR instrument in the range 700–1400 cm-1 using KBr pellets.Fig. 1 Powder X-ray diVraction pattern of BaWO4 heated in ammonia. † Contribution no. 1263 from Solid State and Structural Chemistry Unit. The asterisks indicate W2N impurity. J. Mater. Chem., 1997, 7(10), 2121–2125 2121Table 1 Summary of ammonolysis of diVerent ternary oxides phase in oxygen atmosphere is shown in Fig. 4. The oxidised product contained a mixture of Ba2WO5 and BaWO4 as the ammonolysis major phases and small amounts of Ba3W2O9. There was a starting compound conditionsa products 2.2% mass gain in the TG experiment, which could be attributed to the loss of N2 along with the uptake of oxygen. The BaWO4 900 °C Ba3W2O6N2+W2N Ba2WO5 900 °C Ba3W2O6N2+W2N molecular formula of the oxynitride from the TG studies is Ba3WO6 900 °C Ba3W2O6N2+BaO Ba3W2O9 800 °C Ba3W2O6N2 Ba3V2O8 900 °C no reaction aDuration 12 h.of all these studies are summarised in Table 1. When Ba3W2O9 was heated at diVerent temperatures, we found that the colour of the sample changed slowly from white to grey above 850 °C for 12 h.This colour change may be due to a partial reduction of WVI. From this point, we focused our attention on the white product obtained from Ba3W2O9. The powder X-ray diVraction pattern of this new phase is shown in Fig. 2 and the TPR of the new product in an oxygen atmosphere is shown in Fig. 3. We can see that the N2 was liberated at 350 °C with simultaneous uptake of oxygen.The TG of this oxynitride Fig. 2 Powder X-ray diVraction pattern of Ba3W2O6N2 Fig. 3 TPR of oxidation of Ba3W2O6N2 in O2 Fig. 5 (a) Electron diVraction pattern along the [0001] zone axis for Ba3W2O6N2; (b) electron diVraction pattern along the [01190] zone Fig. 4 TG of oxidation of Ba3W2O6N2 axis; (c) bright-field image of Ba3W2O6N2 crystallites 2122 J.Mater. Chem., 1997, 7(10), 2121–2125Ba3W2O6N2.00(±0.01). It is interesting that after the 2.2% mass Ba3W2O9 belongs to the B cation vacancy-ordered perovgain, the sample started losing mass above 500 °C. Although skite system with 2/3 of the octahedral sites in every layer the mass gain was expected, the mass loss could not be filled with tungsten ions.10 It can be recalled that, for Ba3W2O9, accounted for.To clarify this, Ba3W2O9 was heated in flowing the (W2O9)6- unit is in D3 symmetry and the tungsten trigonal O2 and it was found that this oxide does indeed lose mass prisms share faces. However, in the case of Ba3V2O8, vanadium above 500 °C, and gains its original mass on cooling. ions are in tetrahedral coordination and there is no bridging Scanning electron microscopy (SEM) of these samples was oxygen between the two tetrahedra.In Ba3W2O6N2, the conducted to confirm their metal composition. Spot mode (WO3N)3- unit should be in tetrahedral coordination and analysis showed that the Ba/W ratio is 352. The above therefore the local site group of tungsten tetrahedra is assumed observations lead to the following chemical equations for the to be C3v.To elucidate the local site geometry of oxide and ammonolysis of Ba3W2O9 and subsequent heating of the nitride ligands around the tungsten atom, FTIR and Raman product in oxygen atmosphere: spectroscopic investigations were carried out for these samples. Fig. 6 shows the FTIR spectra of the Ba3W2O9 and Ba3W2O9+4NH3 �Ba3W2O6N2+3H2O+N2+3H2 (1) Ba3W2O6N2 phases.The IR spectrum of Ba3W2O9 matches 2Ba3W2O6N2+3O2�2Ba2WO5+2BaWO4+2N2 (2) very well with the reported spectrum.10 The bands in the region 1000–650 cm-1 have been assigned to the terminal All the peaks in the powder X-ray diVraction pattern (Fig. 2) stretching modes and the region 650–450 cm-1 contains bridg- could be indexed to a hexagonal cell with a=5.993(2) and c= ing stretching modes.The far-IR bands can be attributed to 21.40(4) A ° . The intensity pattern of this sample was generated the bending modes. using the Lazy-Pulverix program with Ba3V2O812 as the struc- Fig. 7(a), (b) and (c) show Raman spectra of Ba3W2O9, tural model. The observed and calculated intensity patterns of Ba3V2O8 and Ba3W2O6N2 respectively. The FTIR and Raman Ba3W2O6N2 are given in Table 2.The calculated intensities spectra of the oxynitride closely resemble those of the Ba3V2O8 match the observed intensities quite well. For comparison phase. There are no bands in the region 650–450 cm-1, which powder diVraction data of Ba3V2O8 is also given in Table 2. means that there are no bridging modes in the oxynitride. This Electron diVraction and microscopy was carried out in order observation clearly confirms that the bridging oxygen atoms to confirm the crystal structure of this phase.Fig. 5(a) and (b) were lost during the reaction and that the nitridation process show electron diVraction along [0001] and [01190] zone axes, is not topotactic in nature because the local coordination confirming the hexagonal symmetry of the phase. The d-values geometry around tungsten is completely changed.On the basis obtained from electron diVraction agree with the observed dof intensities and on comparison with the spectra of (MO3N)n- parameters from the powder X-ray diVraction. Fig. 5(c) shows molecules (M=Re and Os),13,14 we assign these bands to the bright field image of the powder particles. The particles fundamentals of the (WO3N)3- unit.According to the irreduc- are in the range of submicrometre regime with irregular morphology. ible representations of the C3v point group, C=3A1+3E. All Table 2 X-Ray powder diVraction data for Ba3W2O6N2 [cell parameters a=5.993(2), c=21.40(4) A ° ] and Ba3V2O8 [cell parameters a=5.7845(1), c=21.317(1) A ° ]a Ba3W2O6N2 Ba3V2O8 h k l dobs/A ° dcalc/A ° I/I0(obs.) I/I0 (calc.) dobs/A ° I/I0 (obs.) 0 0 3 7.131 7.136 1 <1 7.09 4 1 0 1 5.039 5.057 5 6 4.878 11 0 1 2 4.670 4.680 2 2 4.537 3 0 1 4 3.738 3.731 <1 <1 3.651 12 0 0 6 3.560 3.568 3 3 — — 0 1 5 3.302 3.306 100 100 3.247 100 1 1 0 2.998 3.005 89 90 2.893 75 0 2 1 2.585 2.583 2 1 2.487 4 2 0 2 — 2.528 — <1 2.439 7 0 1 8 — 2.380 — 1 2.353 3 0 0 9 2.370 2.378 2 2 2.369 9 0 2 4 — 2.340 — <1 2.267 11 1 1 6 — 2.298 — <1 2.243 6 2 0 5 2.222 2.224 37 41 2.16 40 0 2 7 — 1.982 — <1 1.934 4 1 0 10 1.980 1.980 21 24 1.962 25 2 1 1 1.951 1.959 6 9 1.886 3 1 2 2 1.923 1.934 2 5 — — 1 1 9 — 1.865 — <1 1.833 10 2 1 4 — 1.846 — <1 1.785 3 1 2 5 1.787 1.787 19 17 1.731 25 3 0 0 1.733 1.735 14 18 1.670 13 0 2 10 1.653 1.653 11 12 1.624 12 1 2 8 — 1.585 — <1 1.543 1 3 0 6 — 1.560 — <1 1.511 2 2 0 11 1.559 1.558 1 3 — — 2 2 0 1.503 1.502 11 15 1.446 10 2 1 10 1.448 1.448 12 9 1.416 13 1 3 1 — 1.440 — <1 1.386 1 0 0 15 — 1.427 — 2 1.421 4 3 0 9 — 1.401 — <1 1.3647 3 1 3 4 — 1.393 — <1 1.3495 1 3 1 5 1.366 1.368 7 7 1.3208 9 aThe values for Ba3V2O8 are from JCPDS file 29–211.J. Mater. Chem., 1997, 7(10), 2121–2125 2123these six modes are IR- and Raman-active.There are six IR and Raman bands in the region 100–1400 cm-1 clearly con- firming the C3v site symmetry of the (WO3N)3- unit. We can assign the strong absorption band at 857 cm-1 [in Fig. 7(c)] to the symmetric stretching mode (n2). The less intense bands at 976, 707, 433, 370 and 296 cm-1 can be attributed to stretching (n1), antisymmetric stretching (n4) and three bending modes n6, n3 and n5.In the IR spectrum of Ba3W2O6N2, the broad unsymmetrical band centred around 350 cm-1 is composed of bands at 290, 340 and 391 cm-1 and three bands at 760, 815 and 983 cm-1 are assigned to n5, n3, n6, n4, n2 and n1 modes, respectively. The diVerence between FTIR and Raman spectra of the corresponding modes may be due to lack of coincidence arising from a coupling between the tetrahedral ions in the unit cell.15 Our assignments also agree with the general rule that symmetrical stretch vibration gives the most intense Raman line.16 Table 3 presents a comparison of IR spectrum of K2ReO3N and the Raman spectrum of [OsO3N]1- ion in solution with the FTIR and Raman spectrum of Ba3W2O6N2.It is clear from the table that the transition-metal and the nitrogen ligand bond is stronger than the corresponding metal–oxygen bond. The general usefulness of vibration spectroscopy for confirmation and identification Fig. 6 FTIR spectra of (a) Ba3W2O9 and (b) Ba3W2O6N2 of structural features in polytypes of mixed metal oxides such as Ba3(B,B¾)2O9-y (B,B¾=Mo, W, V and Ti) has been systematically studied.17 On the basis of above studies, we propose a structural model for this new oxynitride as shown in Fig. 8. From the literature we observe that in most of the oxynitrides which contain alkali/alkaline-earth metal the tetrahedrally coordinated transition metal in the highest possible oxidation state (i.e. NbV, TaV, MoVI, WVI, ReVII and OsVIII). The stability Fig. 8 Projection of Ba3W2O6N2 structure on the (010) plane Fig. 7 Raman spectra of (a) Ba3W2O9, (b) Ba3V2O8 and (c) Ba3W2O6N2 emphasising the (WO3N)3- tetrahedral polyanions. The unit cell is indicated by the dashed lines. Table 3 Fundamental frequencies of vibration (cm-1) of (MO3N)n- (M=W, Re and Os) groupsa assignment and description n1(A1) n2(A1) n3(A1) n4(E) n5(E) n6(E) compound n(MMN) ns(MO3) d(MO3) nas(MO3) d(OMO) d(OMN) K2(ReO3N) (IR solid) 1022 878 315 830 273 380 Na(OsO3N) (R soln.) 1021 897 309 871 309 372 Ba3W2O6N2 (IR solid) 983 815 340 760 290 391 Ba3W2O6N2 (R solid) 976 857 370 707 296 433 R=Raman; soln.=solution. 2124 J. Mater. Chem., 1997, 7(10), 2121–2125of Mn+ (M=transition metal) ion in the tetrahedral environ- W6+ atom changes from six to four and correspondingly, the local site group symmetry changes from D3 to C3v.This ment could be due to eVective p bonding formation by the nitrogen atom in contrast with this type of bonding by the oxynitride is isostructural with the Ba3V2O8 phase. The corresponding molybdenum analogue Ba3Mo2O6N2 was obtained oxygen atoms. This is in agreement with predictions based on the simple consideration that, owing to the formal negative but could not be synthesised in a pure form.charge of N3- and lower electronegativity of nitrogen than The authors thank Dr H. N. Vasan for collection of FTIR and oxygen, N3- will have greater tendency to donate electrons to Raman data of the samples. One of the authors (P. S. H.) the central metal atom through back bonding.13 thanks the Council of Scientific and Industrial Research It is interesting that when Ba3V2O8 is heated in ammonia, (CSIR), New Delhi for a fellowship.Financial support from there is no change in the powder diVraction pattern and also the Department of Science and Technology, Govt. of India is that there is no change in colour. This indicates a high gratefully acknowledged. thermodynamic stability of this oxide.The TPR of the product also did not show N2 released from the sample. It is worth noting that, in the case of ammonolysis of Ba3W2O9 and other References barium tungsten oxide systems, Ba3W2O6N2 was the only 1 Rainer Niewa and Herbert Jacobs, Chem. Rev., 1996, 96, . stable phase obtained. 2 S. Sellem, C. Potvin, J. M. Manoli, R. Contant and G. Dje�ga- Attempts were made to obtain isostructural oxynitride Mariadassou, J.Chem. Soc., Chem. Commun., 1995, 359. phases containing molybdenum, Sr3W2O9, Sr3Mo2O9 and 3 S. H. Elder, L. H. Doerrer, F. J. DiSalvo, J. B. Parise, D. Guyomard Ba3Mo2O9 oxide phases not having been reported in the and J. M. Tarascon, Chem. Mater., 1992, 4, 928. literature, in order to explore new possible oxynitrides belong- 4 P. Subramanya Herle, N.Y. Vasanthacharya, M. S. Hegde, ing to Ba3W2O6N2 family. We prepared a series of oxide J. Gopalakrishnan and G. N. Subbanna, J. Solid State Chem., 1994, 112, 208. mixtures containing BaO2, Sr(NO3)2 (Fluka, 99.5%), MoO3 5 J. Grins, P.-O. Ka�ll and G. Svensson, J. Solid State Chem., 1995, (Aldrich 99.9%) and WO3 with (Ba,Sr):(Mo,W) in the ratio 117, 48. 352, and heated them to 600 °C for 24 h with intermittent 6 F. Pors, R. Marchand and Y. Laurent, J. Solid State Chem., 1993, grindings. Ammonolysis of these oxides was subsequently 107, 39. carried out at diVerent temperatures and the products were 7 S. H. Elder, F. J. DiSalvo, J. B. Parise, J. A. Hriljac and analysed. Only for Ba–Mo oxide did X-ray patterns show the J. W. Richardson, Jr., J.Solid State Chem., 1994, 108, 73. 8 D. S. Bem and H.-C. Zur Loye, J. Solid State Chem., 1993, 104, 467. presence of an oxynitride phase related to Ba3W2O6N2 and 9 R. Marchand, P. Antoine and Y. Laurent, J. Solid State Chem., BaMoO3. In the other cases a perovskite-like phase was the 1993, 104, 34. major phase. It is known in the literature that the oxynitride 10 K.R. Poeppelmeier, A. J. Jacobson and J. M. Longo, Mater. Res. SrMoO2.6N0.418 can be synthesised by heating SrMoO4 in Bull., 1980, 15, 339. ammonia. For the BaMoO3 related phase which we obtained 11 M. S. Hegde, S. Ramesh and G. S. Ramesh, Proc. Indian Acad. Sci. as one of the product phases, it is quite possible that partial (Chem. Sci.), 1992, 104, 591. 12 P. Su� sse and M. J. Buerger, Z. Kristallogr., 1970, 131, 161. substitution of oxygen with nitrogen can take place. 13 B. Krebs and A. Mu� ller, J. Inorg. Nucl. Chem., 1968, 30, 463. 14 L. A.Woodward, J. A. Creighton and K. A. Taylor, T rans. Faraday Soc., 1960, 56, 1267. Conclusions 15 R. H. Busey and O. L. Keller, Jr., J. Chem. Phys., 1964, 41, 215. We have synthesised a new oxynitride Ba3W2O6N2 from the 16 K. Nakamoto, Infrared and Raman Spectra of Inorganic and Co-ordination Compounds, John Wiley & Sons Inc., New York, 3rd ammonolysis of Ba3W2O9 at 800 °C for 12 h. This confirms edn., 1978. the generally observed trend that the ternary oxides containing 17 B. Mo� ssner and S. Kemmler-Sack, J. L ess-Common Met., 1985, the most electropositive ions (alkaline, alkaline earth, rare 114, 333. earth) will not form ternary nitrides by ammonolysis. They 18 Guo Liu, Xinhua Zhao and H. A. Eick, J. Alloys Compd., 1992, will perhaps form oxynitrides, or decompose into the electro- 187, 145. positive metal oxide and binary transition-metal nitride. From Ba3W2O9 to Ba3W2O6N2 the coordination number around the Paper 7/02969A; Received 30th April, 1997 J. Mater. Chem., 1997, 7(10), 2121&ndash