LATTICE PARAMETERS
Oct. 20, 1954
sorption factor for a cylindrical sample with pR = 3.0. The structure found for KAm02F2 is of the CaU0202type. l1 The interatomic distances are Am Am K K
- 2 0 = (1.93 A.) - 6 F = 2.47 A. - 6 0 = 2.73 A. - 2 F = (2.74 A.)
the values in parentheses being assumed. The AmF and K-0 distances are not much affected by moderate variations in the parameter values u and V , and are therefore quite accurate. The great difference between the primary Am-0 bond length and the secondary Am-F bond length is worthy of notice. Table I1 gives results for primary and secondary bond lengths of X O aradicals in various crystals. I n KPu02C03,the six secondary bonds all lie in a plane, and the bond length is for this reason longer than otherwise might be expected. This substance (11) W.H.Zachariasen, Acta Cryst., 1, 281 (1948).
[CONTRIB UTION
FROM THE
OF
RAREEARTH COMPOUNDS
5237
TABLE I1
Compound UOtFz" KAmO%Fs MgUOsOz18 CaU0,OZ KPuOzCOa'
Radical [U01]t* [AmOn]+l [UOz]+z [UO,]tr [Pu01]+1
Bond length, A. Primary Secondary U 2 0 = (1.91) U 6F 2.50 Am 2 0 (1.93) Am 6 F = 2.47 4 0 = 2.18 U 2 0 = 1.93 f 0.03 U u 2 0 = 1.91 f 0.10 U 6 0 2.29 Pu 2 0 = (1.94) PU 6 0 = 2.55
-
-
-
-
apart, the experimental results indicate that the secondary X-F bonds are appreciably longer than the secondary X-0 bonds, and this in turn suggests that the primary bond lengths in U02F, and KAmOzFz may be somewhat smaller than assumed. The KAm02F2structure is built up of hexagonal layers [Am02F2]-held together by the potassium ions. These layers are isostructural with the UOZF 2 layers in uranyl fluoride and with the [ U O ~ O Z ] - ~ layers in CaUO2OZ. The period is 4.22 A.for the [Am02F2]- l a y p and 4.20 A. for the UO2Fz layer as against 3.87 A. for the [ U 0 2 0 ~ ] layer. -~ (12) W.H. Zachariasen, ibid., 1, 277 (1948). (13) W, H.Zachariasen, unpublished result.
Los ALAMOS, NEW MEXICO CHICAGO, ILLINOIS
DEPARTMENT OF CHEMISTRY AND RADIATION LABORATORY, UNIVERSITY OF CALIFORNIA, BERKELEY 1
Lattice Parameters of Some Rare Earth Compounds and a Set of Crystal Radii BY D. H. TEMPLETON AND CAROL H. DAUBEN RECEIVED JUNE 24, 1954 Unit cell dimensions are given for the compounds CeC13, PrC13, SmCla, EuCla and GdC13 (hexagonal UCls type), Smd&, EU& GdzOa, Dy208, Hog03, Er203, Tmz03, Yb203 and Lu203 (cubic MnzOs type), TbF4 (monoclinic UF, type), and TbOF (rhombohedral LaOF type). A set of empirical crystal radii for the trivalent rare earth ions is proposed.
In the course of our studies of the crystal chemistry of rare earth compounds we have determined the lattice dimensions of several compounds of known structure type. In certain cases no lattice dimensions have been reported previously, and in others the new values are believed to be more reliable because of better purity of the rare earth compounds. These results are derived from powder diffraction patterns obtained with Cu KCY(A 1.5418 A.) or Cr K a (A 2.2909 A.)radiation in cameras of 9 cm. diameter. Hexagonal Trich1orides.-The trichlorides of the elements lanthanum through gadolinium were shown to be isostructural by Bommer and Hohmann.' Zachariasen2 showed that they are hexagonal and worked out the atomic positions for the isostructural uranium trichloride. No lattice dimensions seem to have been reported for SmC13, EuC& or GdC13. Our results (Cr K a radiation) are listed in Table I, where the prior values of Zachariasen2 are given in parentheses for comparison. For CeC& and PrC13, which were studied in both researches, the agreement is good. Our sample of SmC13 was prepared by Dr. H. R. Lohr, and the others by Dr. C. W. Koch by the reaction of the. respective oxides with hydrogen chloride gas a t 800 to 900°K. (1) H.Bommer and E. Hohmann, Z anorg. allgem. Chez? , 248, 373 (1941). (2) W.H. Zachariasen, J. Chcm. Phyr., 16, 254 (1948).
TABLE I LATTICEPARAMETERS OF HEXAGONAL TRICHLORIDES a, R. c, A. V, A.r Lacla (7.483 f 0.003)" (4.375 & 0.003)" 212.2 CeC13 7.450 f ,004 4.315 f ,002 207.4 (4.313 & .004)" (7.451 Z!Z ,004)" 4.275 f .004 203.9 PrC13 7.422 f ,005 (7.42 i .01)" (4.26 f .01)" rl'dCla (7.396 f ,004)' (4.239 f . 0 0 3 ) O 200.8 SmC13 7.378 f ,007 4.171 i ,004 196.6 EuC13 7.369 f ,004 4.133 f ,002 194.4 GdC13 7.363 f ,004 4.105 f ,002 192.7 a W. H. Zachariasen (reference 2). Changed from kX. units.
Cubic Sesquioxides.-Most of the sesquioxides of the rare earth elements as commonly prepared have the cubic MnzO3 type structure3. The atomic positions are given for the mineral bixbyite, (Fe,Mn)203, by Pauling and S h a ~ p e l l . ~ We have calculated cell dimensions (Cu KCYradiation) for samples whose purity is greater than 99.9% according to spectrographic analysis for metallic impurities. These samples, purified by ion-exchange methods, were originally obtained from the Institute for Atomic Research, Iowa State College. The results are compared in Table I1 with some of the prior values found in the literature. In several (3) Strukturbcrichl, Vol. 11, p. 38. (4) L. Pauling and M . D. Shappell, 2. Krist.. 78, 128 (1030).
D. H. TEMPLETON AND CAROL H. DAUBEN
5238
cases the differences are significant. The new values, when plotted as a function of atomic number, fall somewhat better on a curve which is smooth except for a cusp a t gadolinium. For the atomic parameters given by Pauling and S h a ~ p e l l ,which ~ we have found also to be suitable for AmZ03,5each metal atom has six oxygen neighbors a t an average distance of 0.21441a, where a is the cell dimension. Terbium Tetrafluoride.-Dr. D. C. Feays prepared TbF4 by the reaction of TbFa with fluorine gas a t 320". The powder patterns, while not excellent, were identified as those of TbF4 by the similarity with those of CeF4 and UF4.'08 The pattern was indexed (Table 111) according to a monoclinic unit cell with the dimensions a =
TABLE 111
POWDERDIFFRACTION DATAFOR TbF4 (Cr K a , X = 2.2909A.)
=
Com-
Obsd.
110
0.0260 ,0331
3ii 112 220 312 002 130 222 i3i 310 422 313 42i 040 112 330 333
12.11 -f 0.06 A. 126.1 f 0.5"
TABLE I1 LATTICE DIMENSIONS OF CL-BIC RAREEARTH OXIDES pound
hkl
1 ii 021 111 302
b = 10.15 f 0.05 c = 7.92 f 0.04
B
Vol. 76
Cell dimensions A . This research P;evious w o r k e d
.1362 .2009 ' .2021 .2037 % .2038 .2041 ,
I, obsd." 1v \\'
.0826
S
.0936
m
.I052
VS
.I160
111
.1287
w
.1347
m
.2015
W
,2382 vw 10.932f 0,009 10.922' 10.915b ,3034 V \v EU2O.i 10.866i .005 10.862" 10.864b ,3257 \v 602 C7di0 10.813f ,005 10.820" 10.81~~ 042 ,3320' r)yyo:i 10.667f .006 10.650* ,3385 w 150 ,3322 Ho20.1 10.607 f .005 10.6OC 222 ,3330 Er&, 10.547 i ,003 10.526' 532 ,3386 TmpO.$ 10.488 f. ,006 10.476' .3395 023 .3404 m 10.42gb Yb?O:i 10.439 f .007 ,3896' i5i 10.396b LuzOa 10.391 f .005 241 ,3403, a A . Iandelli, Gam. chim. ital., 7 7 , 312 (1947). H. O v = very, s = strong, m = medium, w = weak. Bommer, 2.anorg. nllgem. Chem., 241, 273 (1939). W. H .Zachariasen, Norske Vid. Akad. Oslo, I, Mat. Nat. K1. 4, 1 (1928). The values cited from a,.b and c have been clearly that the TbOF was rhombohedral. The diffraction data (Table IV) correspond to a rhomchanged from kX. units. Stll?O.i
I
bohedral cell with
The compounds ZrF4, HfF4,9 ThF4, SpF4, PuF4' and AmF410 also have this structure. Zachariasen7 determined approximate metal atom positions for UF4. Burbank,s with single crystal data for UF4, refined the structure given by Zachariasen and obtained also the fluorine positions. The space group is c i h - c 2 / c with twelve molecules in the unit cell. For TbF4 the cell volume is 786 and the density calculated from the X-ray data is 5.95 g. cm.-3. Terbium Oxyfluoride.-Feay6 also prepared TbOF by pyrohydrolysis of TbF4 a t 400" in a muffle furnace for about ten hours. The pattern (Cu K a ) was pseudo-cubic with considerable broadening of some of the lines. With chromium Kcu radiation the doubling of several of these lines was resolved. Comparison of the intensities and line structure with the data of Zachariasen" for the tetragonal and rhombohedral forms of YOF and LaOF showed (5) D. H. Templeton and C. H. Dauben, THIS JOURNAL, T6, 4560 (1953). (fi) B. B. Cunningham, D. C . Feay and hf. A. Rollier, ibid., 76, 3361 (1954). (7) W. H. Zachariasen, A d a CryJt., 2 . 388 (1949). (8) R. D. Burbank, Atomic Energy Commission Declassified Document, ABCD-3216, August, 1951. (9) G. E. R . Schulze, 2. Krist., 89, 477 (1934). (10) L . B. Asprey. THISJ O V R N A L , 76, 2019 (1854). ( 1 1 ) W: H. Zachariasen, Ariii Cry.?/., 4, 231 (1951).
a = 6.751 I0.005 A. cy = 33.09 f 0.03"
Zachariasen" found the space group D,6d - R h with two molecules in the unit cell and reported atomic positions for LaOF which were satisfactory for YOF also. The parameters cannot be very different for TbOF as the intensities agree well with those observed by Zachariasen for LaOF. With these parameters, each terbium atom has four oxygen neighbors a t 2.45 A. and four fluorine neighbors a t 2.30 The structure is a superlattice based on a slightly distorted CaFz type structure. For the undistorted structure a i ~ ~ 3 3 . 5 6 " . For ~ * TbOF the unit cell volume is 244.8 A.3 and the X-ray density 7.89 g. ~ m . - ~ . Empirical Crystal Radii.-One of the chief uses of crystal radii is the correlation of various thermodynamic properties. In the rare earth series this is not satisfactory if the radii are expressed to two decimal places as is customary, since the differences between adjacent elements are of the same order of magnitude as the possible rounding-off errors.la For this reason, we list in Table V a Set of empirical crystal radii for the trivalent rare earth ions, given
A.
(12) The value 33.22O given in reference 11 is incorrect. 4 recent example of such difficulties io given by Wheelwright, Speddlrlg .'rid Schwarzenbach, Tars JOURNAL, 76, 4186 (1853). (17)
OCt. 20, 1954
TABLE IV POWDER DIFFRACTION DATAFOR TbOF (Cr K a , X = 2.2909 A.) hkl"
hkl, rhomb.
006 102 104 009 107 108 110 201 1, 0,I O 116 202 204 1, 0, 11 119 207 208 1, 0,13 0, 0, 15 1. 0. 14
222 110 21 1
hex.
sin' 8
Calcd.
I,
obsd.
ms m
.2919
vw
.3482 .3550
m m+
433 321 200 220 443
'2g05} .2941 .3480 .3551 .4771 .4771 .4842 .4878 .5309 ' .5524
.4768
W
,485p
m
,5310 .5519
vw
432 331 422 544 555 554
'6456} .6493 .7031 ,7246 ,8071 .8215
.6478
W
.7037 .7232 .8095 .8215
mw vw trace
ioi 1 ii
0.1291 .1327 .1758
Obsd.
0.1284 .1325 .1759
333 322 332
5239
X-RAYAND DIFFERENTIAL THERMAL ANALYSES OF PR-4SEODYMIUM OXIDES
i
W
W
in this structure (coordination six). The cell dimensions of the tetragonal oxychlorides,14the second most extensive isostructural set available, were used in an empirical way to help fix the values for the elements near lanthanum. The less extensive data on the rnonoclinicl6 and hexagonal chlorides and orthorhombic fluorideslBwere used to test the curvature of the plot of radius against atomic number. About as many deviations occurred in one direction as in the other. TABLE V CRYSTAL. RADIIO F TRIVALEXT RARE EARTHI O N S
A. 1.061 1 ,034 1.013 0.995 .979 Sm+++ .964 Eu+++ ,950 a+++ .938 Ion
La+++ Ce+++ Pr+++ Nd+++ Pm+++
Radius,
Ion
Tb+++ Dy+++ Ho+++ Er+++ Tm+++ Yb+++
A. 0.923 ,908 ,894
Radius,
LU+++
.881
.869 .858 .848
We wish t o thank Mrs. Helena W. Ruben who took the diffraction photographs and performed some of the calculations, and Professor B. B. Cun2;0; 10 442 '8322} .8316 W 211 20i .8323 ningham and his students who supplied us with the 2ii .8430' 212 .8430 m+ rare earth trichlorides and the terbium com,8713 W 543 .8716 1, 1, 12 pounds. The completion of this study was aided by 214 310 .8861 .8857 m a John Simon Guggenheim Memorial Fellowship a For an hexagonal cell with a = 3.844, c = 19.13,Z = 6. (to D.H.T.) and by the kind hospitality of Professor * Diffuse. G. Hagg and the Institute of Chemistry, Univert o three decimal places. It should be remembered sity of Uppsala, Uppsala, Sweden. This research that the second decimal is in doubt on an absolute was supported by the Atomic Energy Commission. scale, but the third decimal is of significance in the (14) D. H. Templeton and C. H. Dauben, THISJ O U R H A L , 75, 6069 diferences of adjacent radii. For most correla- (1953). (15) D. H. Templeton and G. F. Carter, J . Phys. Chcm., 68, Nov. tive purposes, the absolute scale is of no conse- (1954). quence. (16) A. Zalkin and D. H. Templeton, TEIS JOURNAL, I S , 2453 These radii are based primarily on the cubic ox- (1953). ides, with the radius of oxygen taken as 1.380 A. BERKELEY,CALIFORNIA W
(CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, STATE UNIVERSITY OF IOWA]
Praseodymium Oxides. 11. X-Ray and Differential Thermal Analyses' BY E. DANIELGUTH,*J. R. HOLDEN, N. c. BAENZIGER AND LEROYEYRING RECEIVEDMAY24, 1954 This is a study of oxides of praseodymium in the composition range PraOa to PrsOlb. X-Ray powder diagrams of quenched samples were taken in order t o investigate the structures of oxides of several compositions. High temperature X-ray powder, diagrams were taken for a check on the quenching procedure. Differential thermal analyses were made to further illustrate the stepwise nature of oxidation and reduction in this system. The results indicate that under equilibrium conditions oxides with hexagonal, body-centered cubic, rhombohedral or face-centered cubic lattices are stable a t various compositions. As Pr~,Ollis reduced t o PrlOs it undergoes changes in lattice from face-centered cubic t o rhombohedral to body centered cubic t o hexagonal. Each of these lattices is found t o be stable over a range of compositions. The rhombohedral lattice has not previously been reported.
Introduction Dissociation pressure measurements on praseodymium oxidesS indicate that in addition t o the well-known oxides, PrzOs and Pr8011,other oxides (1) Part of the data reported here was taken from a dissertation submitted by E. Daniel Guth t o the Graduate College of the State University of Iowa. (2) Research Division, Phillips Petroleum Co., Bartlcnrille, Oklahoma. (3) R. Ferguson, E. Daniel Guth and L. Eyring, THIS JOURNAL, 76, 3890 (1854).
intermediate to them exhibit stability under equilibrium conditions at particular pressures up t o one atmosphere of oxygen, and temperatures up to 1050'. Many previous workers have reported PrzOs and PrsOll, but the nature of the region between these compositions has not been clearly understood. Through the techniques of X-ray diffraction and differential thermal analysis a study of this system has been carried out to further elucidate it.
