J. MATER. CHEM., 1994, 4(5), 707-711 Crystal Structures of Two Sodium Yttrium Molybdates: NaY(MoO,), and Na,Y(MoO,), Nicola J. Stedman; Anthony K. Cheethamb and Peter D. Battle"* a lnorganic Chemistry Laboratory, South Parks Road, Oxford, UK OX7 3QR Materials Department, University of California, Santa Barbara, CA 93106, USA The crystal structure of NaY(MoO& has been determined from powder X-ray diffraction data [space group /4,/a, Z= 2, a =5.1 9890(5) A, c =11.3299(1) A]. The Na and Y atoms are disordered over the eight-coordinate sites of a scheelite structure. The crystal structure of Na,Y(MoO,), has been determined from single-crystal X-ray diffraction data [space group 14,/a, Z=4, a =11.374(3) A, c =11.440(5) A] and found to be scheelite related, with the tetrahedral sites occupied by an ordered arrangement of Mo and Na in a 4:1 ratio.Comparisons are drawn between the structures of these two closely related phosphor materials. Alkali-me tal rare-earth-me ta1 molybdates and tungs ta tes, which have been known since the 1880s,' were ascribed a wide range of formulae by early workers. The situation was clarified in 1943 by the elegant work of Sillen and Sundvall,, which showed that many of the preparations reported were, in fact, orientated overgrowths of two phases, AM(XO,), and A,M(XO,),, where A= Li, Na or K, M =La or Bi and X = Mo or W. A study of crystals of the sodium lanthanum molybdate NaLa(MoO,), and the sodium lanthanum tung- state Na,La( W04), suggested that the former adopts the scheelite structure with Na and La disordered over the eight- coordinate site (the calcium site in scheelite, CaW0,3) and that the latter adopts a related but more complex structure.However, although this work identified the structure type, it did not provide a detailed description of the crystal structure, including the accurate location of the anions. NaLa( has subsequently proved an ideal starting point for studies of the effect of small variations in host structure on luminescent properties4'' because each of the three types of cation can be replaced in turn (sodium by lithium, lanthanum by another lanthanide, yttrium or a combination of these, and molybdenum by tungsten), thus giving rise to a wide range of compositions whilst always retaining the same basic Relatively broad peaks were observed in the luminescence spectra of phosphors based on this type of structure, probably due to the inherent disorder of the lanthanide and alkali-metal cations, but little or no concentration quenching was ~een.~~,~' In contrast, phosphors based on AM(XO,), (A=K, Rb, Cs), which adopt structures in which the alkali-metal and rare-earth-metal ions are ordered, gave sharper spectra but were more prone to concen- tration quenching.,' Neodymium-or erbium-doped NaLa(MOO,), have lately proved promising as laser materials.21 The phases A,M(XO,), (A =Li, Na; M =La to Lu, Y; X = Mo or W) have also proved to be good phosphors, showing very little concentration quen~hing.'~,~~-~~ They are, again, isomorphous across the lanthanide series22v26 and since the early work by Sillen and Sundvall the crystal structures of the tungstates and molybdates Na,La(MO,),, Na,Tb( MO,), and Na,Lu(MO,), (M=Mo or W) have been fully deter- mined, confirming that the structure is scheelite related.27-29 Despite the high level of interest in the optical properties of these materials, no close structural comparison of the two series of compounds has been carried out.Curiously enough, this is not due to a lack of structural information regarding the more complex A,M(XO,), phases, but due to a lack of data regarding the simple scheelites AM(X04)2. We therefore undertook a study of the two sodium yttrium molybdates NaY(MoO,), and Na,Y(MoO,),, beginning our work at a time when no AM (XO,), compound had been fully character- ised. During the course of our work an account of the crystal structure of NaLa(MoO,), was published,21 but the quality of the refinement was not good enough to be considered definitive for the whole structure type.The results of our own work are described below. Experimental The initial synthesis of the two phases was entirely serendipit- ous; an attempt to crystallise a reduced yttrium molybdenum oxide by dissolution in a sodium molybdate flux at 1100 "C under an inert gas, followed by a slow cooling, led instead to crystals of molybdenum dioxide and Na,Y(MoO,), in a melt containing NaY(MoO,),. It was shown that the product contained two distinct Na-Mo-Y-0 phases by the analysis of individual crystallites using a JEOL 2000FX analytical electron microscope in transmission mode together with a Tracor Northern TN5500 energy-dispersive X-ray detector (Table 1).Thin crystallites (partly transparent to the electron beam) were analysed as the effects of fluorescence and absorp- tion can be considered negligible in the thin-crystal limit,30 thus enabling the composition of each crystallite to be inferred from the relative peak intensities in the X-ray spectra.Comparison of the results with those obtained on analysis of an yttrium molybdate of known composition showed that particles of type A contained yttrium and molybdenum in the ratio 1:4,and that particles of type B contained yttrium and molybdenum in the ratio 1:2. It was also apparent that particles of type A contained over twice as much sodium Table 1 Analytical electron microscopy results for individual crystal- lites of the two Na-Y-Mo oxides particles of Type A particles of Type B "a5Y(MoO4 )4 1 "aY (MOO, 121 Y-Ka/Mo-Ka Na-KalMo-Kr Y-KalMo-Ka Na-Ka/M o-Ka 0.33 0.50 0.63 0.21 0.36 0.47 0.60 0.23 0.33 0.52 0.67 0.24 0.31 0.47 0.64 0.25 0.35 0.53 0.65 0.23 0.33 0.57 0.30 0.46 Particles containing molybdenum only were also observed, corre- sponding to molybdenum dioxide.(relative to molybdenum) as particles of type B, suggesting the formulae given above. In order to confirm these formulae, and because good single crystals of NaY(MoO,), had not been obtained in the first experiment, a pure polycrystalline powder of each phase was prepared by standard solid-state techniques.In each case, the appropriate stoichiometric mix- ture of MOO, (Aldrich, 99.5%), Y203 (Aldrich, 99.99%, dried before use) and Na2C03 (BDH AnalaR, 99.9%, dried before use) was ground well in an agate mortar, pressed into a pellet and heated for 2 days at 500 "C in an alumina crucible, reground and repelleted, then heated for 5 days at 800 "C. The ratio of yttrium to molybdenum in each of these powders was checked by analytical electron microscopy, giving results which correlated well with those given in Table 1, and the sodium content of each was confirmed by atomic absorption spectroscopy (NaYMo,O,: calc.5.32%, expt. 5.87%. Na5Y(Mo04),: calc. 13.63%, expt. 13.97%). Na5Y(Moo4), was studied by single-crystal X-ray diffrac- tion. A small, approximately spherical (diameter z0.2 mm), colourless, transparent crystal suitable for data collection was selected and mounted on a four-circle kappa-geometry Enraf- Nonius CAD-4 automatic diffractometer controlled by .a PDPll/23 microprocessor. Mo-Ka radiation (2=0.710 69 A) was used. 25 reflections were located from a preliminary Polaroid photograph, centred and then indexed on a triclinic unit cell. However, examination of the Niggli matrix of this first unit cell suggested that the true symmetry was tetragonal. The unit cell was transformed accordingly and a number of additional reflections were then measured, centred and used in a least-squares refinement to obtain accurate unit-cell parameters.Shells of intensity data were collected in the angular range 0"<6< 37.5". A horizontal slit width of 4 mm was used with a &dependent o scan width of (1.10+0.35 tan 6)'. Three strong reflections were chosen as intensity standards and were monitored every hour, and three chosen as orientation con- trols were checked after the measurement of every 250 reflec- tions. At the end of data collection the intensities of two reflections with x close to 90" were measured as a function of @ (from 0 to 360'), and these data were used to correct for absorption (p= 73.51 cm-l; min/max =2.59/2.73). Powder diffraction data were collected on NaY(MoO,), by J.MATER. CHEM., 1994, VOL. 4 step scanning over the angular range 5"<28<100' using a Philips PW1710-based diffractometer operating in Bragg-Brentano geometry with a step size of 0.02 O 28 andoa count time of, 10 s per step. Cu-Ka radiation (& =1.540 51 A, A2 = 1.544 33 A) was used. Results Study of the systematic absences in the data collected on Na5Y(Mo04), led to the assignment of the space group as I4,/a (International Tables for Crystallography Volume A, no. 88, origin !t -131) with 274 and the unit-cell parameters a =11.374( 3) A, c = 11.440(5) A. The data were merged (R= 4.43%) and corrected for polarisation, the Lorentz factor and absorption. Friedel pairs were not averaged as a correction for anomalous scattering was made.Of the 2074 reflections remaining at this stage, 1403 with I >3a(I) were used for structure solution and refinement. The structure was solved using the automatic Patterson interpretation routine in the program SHELXS3,All the atoms were located. The model from SHELXS was then refined using the least-squares routines in the program CRYSTALS3, In the final stages of refinement, all atomic coordinates and anisotropic temperature factors were allowed to vary (a total of 59 parameters) and a three-term Chebyshev polynomial (coefficients 6.41, -3.13, 3.91) was introduced as a weighting scheme. The refined atomic positions and tempera- ture factors are listed in Table 2 and selected bond lengths and angles are given in Table 3.The final agreement factors were R=3.14% and R,=3.38%; the strongest pe$k in the last difference Fourier map had a height of 0.18 e A-,. The structure is represented in Fig. 1. The powder diffraction data collected on NaY (MOO,), were smoothed using the Philips Automated Powder Diffraction Software34 and sutsequently indexe$ on a tetra- gonal unit cell [a=5.19890(5) A, c= 11.3299( 1) A], consistent with space group I4Ja (International Tables Volume A, no. 88,l with 2 =2. The data were analysed using the program GSAS.35 In the starting model, yttrium and sodium were placed on the 4(b) site and constrained to half occupancy, and molybdenum was placed on the 4(a) site, as proposed originally by SillCn and Sundvall., The oxygen site was then Table 2 Atomic coordinates and temperature factors for Na,Y (Mo0J4 atom site X Y Z Ueauiv/A2 ~ ~ ~~ 0.0 0.25 0.625 0.023 1 0.7044(2) 0.5 0.1296(2) 0.25 0.5940( 2) 0.375 0.0138 0.006 1 0.8189( 3) 0.8512( 3) 0.61 1 l(2) 0.6798(2) 0.8240( 3) 0.34359( 2) 0.2 194 (2) 0.31 19( 2) 0.3343( 2) 0.4652( 3) 0.38685(2) 0.4727( 3) 0.5382( 2) 0.3 133 (2) 0.4791 (3) 0.0072 0.01 39 0.0105 0.01 10 0.0142 u22 u33 u23 u13 Na(1) 0.031 (1) 0.03 1 (1) 0.013(2) 0.0 0.0 0.0 W2) 0.0145( 7) 0.01 66( 7) 0.0116(6) O.O026( 5) -O.OOlO( 5) 0.001 1 (5) Y(1) O.O067( 1) 0.0067( 1) O.O052( 1) 0.0 0.0 0.0 Mo(1) 0.0075( 1) 0.0076( 1) 0.0067( 1) -O.O003( 1 ) 0.0000( 1) 0.0oO0( 1) O(1) 0.016( 1) 0.012( 1) 0.016(1) O.O04( 1) O.O03( 1) 0.002(1) O(2) 0.012( 1) 0.001( 1) 0.01 (2) -O.O006(8) -O.O033( 8) 0.0014(8) O(3) 0.0090( 9) 0.016( 1) 0.010(1) O.O024(8) -0.0007(8) -0.0019(8) (34) 0.023( 1) 0.013(1) 0.012( 1) -0.0059(9) O.Ool(1) -O.OOO( 1) T =[-27~~(h~a*~u,, +12c*2u33+k2b*2~2z +2hka*b*u12+2hla*c*uI3+2klb*c*u2,)].Uequiv=(U1U2U3)"3,where Ui are the principal components of the thermal displacement tensor. Structure factors for this compound have been deposited on magnetic tape at the Chemical Crystallography Laboratory, 9 South Parks Road, Oxford. J. MATER. CHEM., 1994, VOL. 4 Table3 Bond distances (‘/A)and bond angles (‘/degrees) in Na5Y(MOO,), Na( 1)-O( 1) 4 x 2.454( 3)O(1)-Na( 1)-O( 1) 4 x 120.3(l),2 x 89.5( 1) Na(2)-O( 1) 2.400(3)Na(2)-0( 1)’ 2.486( 3) Na(2)-O(2) 2.415(3)Na( 2)-0(3) 2.358( 3) Na( 2)-O(4) 2.459( 3) Na(2)-O(4)’ 2.310( 3) 4 x 2.362(2) 4 x 2.366( 2) Mo( 1)-O( 1) 1.759( 3) Mo( 1)-0(2) 1.775( 3) Mo( 1)-0( 3) 1.796 (2) Mo( 1)-O(4) 1.741(3) O(1)-Mo( 1)-0(2) 106.5(1) O(1)-Mo( 1)-0(3) 1133 1) O(1)-Mo( 1)-0(4) 107.0( 1) O(~)-Mo( 1)-0( 3) 112.5( 1) O(2)-Mo( 1)-0(4) 105.7( 1) 0(3)-Mo( 1)-0(4) 111.1( 1) C 0Nal”j F oy n Wh Fig.1 Crystal structure of Na5Y(MOO,),. MOO, tetrahedra are shaded. located using a difference Fourier synthesis. In the final cycles of refinement, a zero-point error, the unit-cell parameters a and c, the coordinates of the oxygen atom, an isotropic temperature factor for each atom (constrained to be equal for yttrium and sodium) and four terms for a pseudo-Voigt peak shape were refined simultaneously against the intensities of 82 Bragg reflections.The background was defined by linear interpolation between fixed points selected by eye. The final observed, calculated and difference diffraction profiles are shown in Fig. 2. The final atomic positions are given in 709 Table 4 Atomic parameters for NaY(MoO,j, atom site X Y z T ‘is,/A2 Nap 4(b) 0.5 0.75 0.125 0.0076(6)Mo 4(a) 0.0 0.25 0.125 0.0055(4)0 16(f) 0.1464(7) 0.4813(7) 0.2122(3) 0.012(2) Table 5 Bond distances (/A) and bond angles (/degrees) in NaY (MOO,), Na/Y-0 4 x 2.435(4), 4 x 2.511(4j Mo-0 4 x 1.733(3)0-Mo-0 4 x 108.9(l), 2 x I10.5(2) 0 Fig. 3 Crystal structure of NaY(MoO,),. MOO, tetrahedra are shown.Shaded circles represent a disordered distribution of Na and Y. Table 4. Selected bond angles and distances are given in Table 5 and the structure is drawn in Fig. 3. The final agree- ment factors were R,=5.49%, R, =9.37% and R, =12.52%, with x2=7.22. Discussion Na5Y(Mo04)4 was found to be isostructural with the other phases NaSLn(XO,), (Ln =La, Tb or Lu and X =Mo or W) for which the crystal structures have been deter~nined.~~-~~ The molybdenum site has an approximately regular tetra- hedral coordination geometry and the eight-coordinate yttrium site is approximately square prismatic. In contrast to the structure of N~Y(MOO,)~,two distinct sodium sites are found. Most of the sodium in the structure “a(?)] is accommodated in a highly irregular six-coordinate site, in which Na-0 bond lengths range from 2.310(3) to 2.486(3) A.The remainder [Na(l)] is found in a site of hiGher (S,) symmetry, coordinated by four oxygens at 2.454(3) A. Our refinements confirm that NaY(MoO,), adopts the scheelite structure with sodium and yttrium disordered over the 4(b) site. Molybtenum is coordinated by four oxygen atoms at 1.733(3) A in a tetrahedral site, and the yttrium/sgdium site is eight-coor$inate, with four oxygens at I 1 I I1-20 30 40 50 60 70 80 90 100 2.435(4)A and four at 2.511(4)A. The disorder of sodium and yttrium in this structure is also likely to affect the o~ygen site. Each oxygen atom forms part of the coordination sphere of three cation sites, one which is always occupied by molyb- denum and two which can be occupied by either sodium or 2Hdegrees Fig.2Observed (...), calculated (-) and difference X-ray diffraction profiles of NaY(Moo,),. Reflection positions are marked. J. MATER. CHEM., 1994, VOL. 4 yttrium; four different local environments are thus possible for each oxygen atom in the structure (the oxygen site is not equidistant from the two nearest sodium/yttrium sites, so there are two possible ways in which each oxygen may be coordinated by one molybdenum, one yttrium and one sodium). The site located in our refinements will thus be the centre of an averaged electron density distribution, the average being taken over the actual oxygen sites in the structure. Residual electron density near O(1) but displaced slightly towards each of the two nearest yttrium/sodium sites was observed in the final difference Fourier maps, but the disorder could not be modelled successfully. Despite this, the model obtained for NaY ( MOO,), appears to be significantly better than any published model for the series AM(X04)2.The most recent of these, that for NaLa(MoO,),, is clearly unsatisfac- tory; our calculations for the bond lengths in this phase using the parameters given in ref. 2b resulted in a remarkably short Mo-0 bond length (1.69 A) and a short 0-0 contact (1.91 A), which suggests that the oxygen position quoted is in error. In order to compare the structures of these two compounds it is useful to envisage the cation sublattice of Na5Y(Mo04), as being composed of layers perpendicular to the c axis, with four such layers, together with the anion sublattice, being contained in one unit cell.One layer is shown in Fig.4 together with the corresponding layer of the simple scheelite structure; the scheelite unit cell is also shown, lying with the axis a' at an angle of ca. 60 O to the a axis of Na,Y( MOO,),. In general, the structure of the scheelite-related phases A,M(XO,), can be derived from the structure of AM(XO,), by replacing one-fifth of each X with A, and altering the random 1: 1 distribution of A and M on the eight-coordinate sites such that every fifth site is now occupied by M, the remainder being occupied by A. This is then followed by small shifts in the positions of some of the cations, and larger changes in the anion sublattice, some oxygens being lost altogether.The cations X (Mo or W) remain tetrahedrally coordinated, but are displaced slightly out of the planes containing the cations M. Of the two new A sites, the six- (a1 0Na V V0 0 0 0 0 0 Fig.4 (a) Cation layer perpendicular to [OOl] in Na,Y(MoO,),. MOO, tetrahedra are shaded, the larger unfilled circles represent Na, the smaller circles represent Y. The unit cell is shown. (b)The same layer in NaY (MOO,),. coordinate sites, derived from an eight-coordinate M site in the scheelite structure, are also displaced slightly out of the planes containing the remaining M cations, while the A cations in the four-coordinate sites, derived from a tetrahedral X site in the scheelite structure, remain coplanar with the M cations.The disorder in the scheelite phase leads to ambiguity in the interpretation of the bond distances in Nay(M00,)~ as yttrium-oxygen and sodium-oxygen distances cannot be dis- tinguished. In the light of typical Na-0 and Y-0 bond lengths,36 use of the average oxygen position given in Table 3 is likely to lead to an Na/Y-0 bond length that on the one hand underestimates the true Na-0 distances and on the other overestimates the true Y -0 distances present. This would explain the apparent difference in Y-0 bond lengths in the two phases; the yttrium site appears smaller in Na,Y(MoO,), than in NaY(MoO,),. The two sodium sites in Na5Y(Mo0,), have lower coordination than the sodium/yttrium site in NaY(MoO,), and the Na-0 bond lengths are therefore shorter in the former.It is also possible that the Mo-0 bond lengths present in NaY(MoO,), are slightly underestimated as a result of the disorder as they are certainly smaller than those found in Na,Y (MOO,),. This may, however, be simply a result of the difference in the two structures; the sodium-rich phase adopts a much more open structure than that of the simple scheelite. Conclusions The structures of NaY MOO^)^ and Na,Y (MOO,), have been determined. Rietveld analysis of powder X-ray diffraction data showed that NaY(MoO,), adopts the scheelite structure with sodium and yttrium disordered over the eight-coordinate site.Residual electron density near the oxygen site suggested that this is also disordered, but it did not prove possible to model this. Na,Y( MOO,), was studied by single-crystal X-ray diffraction and was shown to adopt a scheelite-related structure. References 1 A. G. Hogbom, Ofvers. Akad. Stockholm, 1885,42, 1. 2 L. G. Sillen and H. Sundvall, Ark. Kemi, Mineral. Geol., 1943, 17A, 10. A. Zalkin and D. H. Templeton, J. Chem. Phys., 1964,40, 501. L. G. Van Uitert, J. Appl. Phys., 1960,31, 328. L. G. Van Uitert and R. R. Soden, J. Chem. Phys., 1962,36,517. L. G.Van Uitert and R. R. Soden, J. Chem. Phys., 1962,36,1289. L. G. Van Uitert, R. R. Soden and R. C. Linares, J. Chem. Phys., 1962,36,1793. 8 L. G.Van Uitert and R.R. Soden, J. Chem. Phys., 1962,36,1797. 9 L. G. Van Uitert, J. Chem. Phys., 1962,37,981. 10 L. G. Van Uitert and S. Iida, J. Chem. Phys., 1962,37,986. 11 L. G. Van Uitert, J. Electrochem. Soc., 1963,110,47. 12 M. M. Schieber and L. Holmes, J. Appl. Phys.. 1964,35,1004. 13 M. M. Schieber, Znorg. Chem., 1965,4, 762. 14 M. V. Mokhosoev, V. I. Krivobok, S. M. Aleikina, N. S. Zhingula and N. G. Kisel, Znorg. Muter., 1967,3, 1444. 15 M. V. Mokhosoev and E. I. Get'man, Russ. J. Inorg. Chem., 1969, 14, 1236. 16 R. Heindl, F. Damay, R. Der Agobian and J. Loriers, C. R. Acad. Sci. Paris, 1965,261, 3335. 17 S. Preziosi, R. R. Soden and L. G. Van Uitert, J.Appl. Phys., 1962, 33, 1893. 18 M. V. Savel'eva, I. V. Shakhno and V. E. Plyuschev, Znorg.Muter., 1970, 6, 1466. 19 B. F. Dzhurinskii, L. N. Zorina and G. V. Lysanova, Inorg. Muter., 1980, 16,86. 20 J. P. M. Van Vliet, G. Blasse and L. H. Brixner, J. Solid State Chem., 1988,76,160. 21 S. B. Stevens, C. A. Morrison, T. H. Allik, A. L. Rheingold and B. S. Haggerty, Phys. Rev. B Conden. Muter., 1991,43, 7386. J. MATER. CHEM., 1994, VOL. 4 71 1 22 23 24 25 26 27 28 29 30 V. K. Trunov, T. A. Berezina, A. A. Evdokimov, V. K. Ishunin and V. G. Krongauz, Russ. J. Inorg. Chem., 1978,23,1465. H. Y-P. Hong and K. Dwight, Muter. Res. Bull., 1974,9,775. J. Huang, J. Loners, P. Porcher, G. Teste de Sagey, P. Car0 and C. Levy-Clement,J. Less Common. Metals, 1983,94,251. J. Huang, J. Loners, P. Porcher, G. Teste de Sagey, P. Car0 and C. Levy-Clement, J. Chem. Phys., 1984,80,6204. M. V. Mokhosoev, F. P. Alekseev and E. I. Get’man, Russ J. Inorg. Chem., 1969,14,310. V. A. Efremov, T. A. Berezina, I. M. Averina and V. K. Trunov, Sou. Phys. Crystallogr., 1980,25, 146. V. A. Efremov, V. K. Trunov and T. A. Berezina, Sou. Phys. Crystallogr., 1982,27,77. R. F. Klevtsova, L. A. Glinskaya, L. P. Kozeeva and P. V. Klevtsov, Sou. Phys. Crystallogr., 1973, 17, 672. A. K. Cheetham and A. J. Skarnulis, Anal. Chem., 1981,53,1060. 31 32 33 34 35 36 International Tables for Crystallography, ed. T. Hahn. Kluwer, Dordrecht, 1989,vol. A. G. M. Sheldrick, Program for the Solution of Crystal Structures, University of Gottingen, Germany, 1986. D. J. Watkin, J. R. Carruthers and P. W. Betteridge, CRYSTALS User Guide, Chemical Crystallography Laboratory, University of Oxford, Oxford, 1985. Philips PW 1877 Automated Powder Diffraction Software, Version 3.50. A. C. Larson and R. B. von Dreele, Los Alamos Report LAUR R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect B, 1969, 25,925. 86-748,1987. Paper 3/06485I; Received 29th October, 1993