Refinement of the structure of trilanthanum trichlorohexaoxotungstate, La3WO6Cl3, from neutron powder diffraction data

M r = 802.9, P63/m, a = 9.4092 (2), c = 5 .4276(2)A, V = 416.15 (2) A 3, z = 2 , highresolution neutron powder diffraction data ( 2 = 1.5416 (3) A, T = 295 K), collected from 10 to 122 ° in 20 with a 0.05 ° step size; final weighted profile R = 9.16% was compared to an expected R E = 7.41% (conventional R based on integrated intensities= 6-00%). The refinement confirms the unusual trigonal prismatic coordination found for the WO6 group in the X-ray study [Brixner, Chen & Foris (1982). J. Solid State Chem. 44, 99-107]. The W O and La-CI bond lengths determined in this study are slightly shorter, by 0.006 (2)-0.014 (4)A, than those found in the X-ray determination. Introduction. We have recently published (Brixner, Chen & Foris, 1982) the structure of the new compound La3WO6Cl 3, where W 6÷ finds itself in a highly unusual trigonal prismatic coordination, surrounded by six equidistant 02atoms. An extended Hfickel molecular-orbital calculation (Pensak & McKinney, 1979), comparing the D3h environment for W 6+ in La3WO6Cl 3 with the more common O h symmetry in Ba2CaWO 6 indicated a 146 kJ difference in favor of the octahedral arrangement. A calculation of the Madelung energy contribution (Hoppe, 1982)to the lattice energy also showed that W O bonds do not contribute their normal share of energy in this D3h environment. These facts raised some doubt about the structure of La3WO6CI 3 and prompted us to undertake a neutron diffraction study and unambiguously establish the correct symmetry of the WO 6 group. Experimental. Preparation followed procedures previously described (Brixner, Che n& Foris, 1982); 10 g sample in vanadium can, ~10 mm diameter, highresolution five-detector diffractometer (Prince & Santoro, 1980) at the National Bureau of Standards Reactor; monochromator: 220 reflection of Cu; * Contribution No. 3193. 0108-2701/83/101326-03501.50 horizontal divergences: 10' in-pile collimator, 20 ° monochromatic beam collimator, 10' diffracted beam collimator; monochromator mosaic spread: -,.15'; angular ranges scanned by detectors: 10-42, 30-62, 50-82, 70-102, 10-122 ° . Data analyzed with the Rietveld technique (Rietveld, 1969a,b) as modified by Prince (1980) for data collected from the multi-detector neutron powder diffractometer; Fourier and bond length/bond angle calculations performed using a package of programs described by Finger & Prince (1975); b (O)=5-803 , b (La )=8 .27 , b ( W ) = 4 . 7 7 , b(Cl) = 9.5792 fm (Koester, 1977); background fitted by assuming a straight line, and the two parameters describing it were refined for each of the five counters; a common polynomial curve to describe halfwidths and a common zero point was assumed for the counters; corrections for counter efficiency carried out by comparison to an A120 3 standard prior to Rietveld analysis; peak shape considered to be strictly Gaussian with no corrections made for either low-angle peak asymmetry or preferred orientation effects; no corrections for absorption; peaks due to unreacted LaOC1 starting material in the sample excluded from refinement; unit-cell constants and symmetry checked using the program of Visser (1969) and were found to be in agreement with the single-crystal analysis; in particular, no evidence of a superlattice was detected. Using the atomic parameters of Brixner, Chen & Foris (1982), refinement of all positional parameters, individual isotropic thermal parameters, cell constants, three half-width parameters, and diffractometer zero point led to the final discrepancy factors listed in Table 1. Fig. 1 shows the final fit between observed and calculated neutron powder diffraction patterns for La3WO6Cl3.t No significant details were observed in the final difference Fourier map which was calculated Numerical data corresponding to all 5 counters have been deposited with the British Library Lending Division as Supplementary Publication No. SUP 38697 (6 pp.). Copies may be obtained through The Executive Secretary, International Union of Crystallography. 5 Abbey Square, Chester CH1 2HU, England. © 1983 International Union of Crystallography J. B. PARISE, L. H. B R I X N E R A N D E. P R I N C E 1327 T a b l e 1. Results of the Rietveld refinement of the s t ruc ture of La3WO6CI 3 x y z B ( /~ ) La 0.0896 (3) 0-6843 (3) ~ 0.50 (5) W ~ ] 1⁄4 1.42(21) CI 0-2399 (2) 0.0474 (3) ~ 0.42 (4) O 0.3627 (3) 0.8370 (3) 0.0231 (5) 0.56 (5) Function minimized: Z ~ = Z i w l [ y i ( o b s ) ( I / c ) y i ( c a l c ) ] ~ R u = 100 [ 2. I I ( obs ) I ( c a l c ) I/~_J(obs)] = 6 .00 R e = 100 [~_ l y ( o b s ) y ( c a l c ) l / ) s y ( o b s ) ] = 7.00 R w = 100 { Z w [ y ( o b s ) y ( c a l c ) 1 2 / Z w y 2 ( o b s ) } ~n = 9 . 1 6 R e = 100 [ ( N P ) / ) w y 2 ( o b s ) ] ~/2 = 7.41 In the above formulas, N is the number of statistically independent observations, P the number of parameters refined, I the integrated intensities, y(obs) and y(calc) the observed and calculated intensities, and w= l /y(obs) the weights associated with the data points y(obs). using structure factors derived from the integrated intensities determined according to the approximation of Rietveld (1969a,b). T a b l e 2 . Comparison of interatomic distances ( A ) f o r the s t ruc ture ofLaaWO6C13 Derived from (a) neutron and (b) X-ray data (Brixner, Chen & Foris, 1982).