NOTES
Dec. 5 , 1954
6029
wed1 by the hollow prism method. The method hexafluoride was found to be 1.3580 f 0.0004, has proved to be useful not only for examining whereas the value 1.367 has been previously recorrosive interhalogen compounds, but for examin- ported in the literature.2p8 The earlier value, ing highly volatile substances as well. I n the which was obtained by measuring the critical angle present instance, an electrically heated hollow of reflection a t the surface of liquid uranium hexaprism has been used to measure the refractive fluoride, is believed to be in error. indices of the systems uranium hexafluorideExperimental bromine trifluoride and uranium hexafluorideThe apparatus used in making these measurements and bromine pentafluoride. The data obtained at the method of purification of the bromine fluorides have 70" with the sodium-D lines are given in Tables I been previously described.' In the present work, the origiand 11. The values of n70D for the pure compon- nal hollow prism was modified by the addition of electric coils to the top and bottom jackets, so that a tements are averages of three or more determinations, heating perature of 70.0 f 0.5" could be maintained. This temwhereas the values for the binary solutions are perature was chosen in order to keep the uranium hexafluothe results of single determinations. ride in the liquid state over the entire range of composition. The temperature was measured to f 0 . 1 " by means of a calibrated copper-constantan thermocouple and a Leeds REFRACTIVE INDICES OF URANIUMHEXAFLUORIDE-BRO-and Northrup model K potentiometer. The uranium hexafluoride was obtained from Carbide and MINE TRIFLUORIDE SOLUTIONS Carbon Chemicals Company. I t s freezing point was found Mple Mole t o be 64.0°, in good agreement with the value of Oliver and fraction of fraction of Grisard.' uranium ~ 7 0 ~ uranium n70~
TABLE I
hexafluoride
Expt.
Calcd.
hexafluoride
Expt.
Calcd.
0.0000 ,0487 .0498 ,1566 .3484
1.4302 1.4248 1.4247 1.4131 1.3968
1.4302 1.4248 1.4247 1.4138 1.3968
0.5303 ,6420 .8181 1.0000
1.3836 1.3758 1.3656 1.3580
1.3830 1.3758 1.3662 1.3580
(2) J. J. Katz and E. Rabinowitch, "The Chemistry of Uranium," National Nuclear Energy Series, Div. VIII, Vol. 5, McGraw-Hill Book Co., New York, N. Y., 1951, p. 432. (3) D. R. Llewellyn, J . Chcm. SOC., 28 (1953). (4) G. D. Oliver and J. Grisard, THISJOURNAL, 76, 2827 (1958).
CHEMICAL ENGINEERING DIVISION ARGONNENATIONAL LABORATORY P. 0. Box 299 TABLEI1 LEMONT, ILLINOIS REFRACTIVE INDICES OF URANIUM HEXAFLUORIDE-BROMINE PENTAFLUORIDE SOLUTIONS Mole fraction of uranium hexafluoride
0.0000
,1684 ,3153
n7o~ Expt. Calcd.
1.3275 1.3338 1.3388 1.3428 1.3449
1.3276 1.3337 1.3389 1.3427 1.3452
Mole fraction of uranium hexafluoride
Expt.
0.5965 ,7607 ,9069 1.0000
1.3477 1.3522 1.3560 1.3580
n7oD
Calcd.
1.3477 1.3522 1.3560 1.3580
Yttrium Chromium Oxide, a New Compound of the Perowskite Type BY JOHN T. LOOBY AND LEWISKATZ JULY 19, 1954 RECEIVED
The search for suitable systems in which to study the distribution of trace ions between a molten salt and a solid led us to consider the possibility of An empirical equation has been fitted to the the replacement of lanthanum in LaCrO3 with ytdata in each table, representing n 7 0 as ~ a function trium using sodium chloride as the liquid phase. of N , the mole fraction of uranium hexafluoride. This appeared to be a convenient system since suitEquation 1, below, applies to the uranium hexa- able radioactive isotopes of both yttrium and lanfluoride-bromine trifluoride solutions, and equa- thanum are available. In the course of this work tion 2 applies to the uranium hexafluoride-bromine a finely divided green solid was prepared from pentafluoride solutions. The equations have been the oxides Cr203and Y203. Casual inspection of made to yield the experimental values of n70~ the X-ray diffraction pattern suggested that the a t N =' 0 and N = 1, since the data are most structure was similar to that of perowskite. This reliable a t these points. The calculated values of result is perhaps not surprising in view of the fact that the tolerance factor t in the Goldschmidt relan 7 0are ~ listed in column 3 of Tables I and 11. tionship RA RO = t d Z ( R B Ro) has the value n7'D = 1.4302 - 0.1123N + 0.0484N' - 0.0083N' (1) of 0.86. According to NBray-Szab6,' this would %'OD = 1.3275 .f 0.0380N - 0.0065Na - O.OO1ONa (2) lead one to expect a monoclinic distortion of the The statistical probable error of n7'D computed perowskite structure or according to Megaw2 an from equation 1 is f0.0004, and that computed orthorhombic distortion. from equation 2 is fO.0001. The smaller value is Experimental believed to be fortuitous, and a single measurement Chromic oxide was prepared by ignition of reagent grade is believed to have a precision of approximately fO.0004. Since n 7 0 of ~ bromine trifluoride differs chromic nitrate until there was no further loss of weight. The yttrium oxide was obtained from Fisher Scientific Comfrom that of uranium hexafluoride by 0.0722, it pany and was labeled 99% pure. Mixtures of the two should be possible to analyze the binary solutions oxides, with the yttrium oxide to chromic oxide molar ratio to within f0.6% by refractive index measurements. always greater than one, were heated in a sodium chloride Analysis of uranium hexafluoride-bromine penta- flux a t 900" in a hydrogen atmosphere. The flux was removed from the water-insoluble product by leaching with fluoride solutions within f1.3y0should be possible, water; excess yttrium oxide was removed by leaching with since t h e difference in n 7 0 is ~ 0.0305. very dilute hydrochloric acid. In all cases the X-ray diffracI n the present investigation, n701) for uranium tion patterns appeared the same. The density of the com.4354 .5134
+
(1) L. Stein, R. C. Vogel and W. H. Ludewig. THISJOWRNAL, 76, 4287 (1954).
+
(1) I. Niray-Szabb, Mucgyctcmi K&dcmcnyck, 1, 30 (1947). (2) H. D. Megaw, Proc. Phys. Soc., 68, 133 (1946).
NOTES
6030
pound was determined pycnometrically by water displacement and an average value of 5.78 st 0.05 g./cc. was obtained. Chromium was determined by dissolving the sample in boiling, concentrated perchloric acid and titrating the resulting chromium( VI) in potassium iodide solution with sodium thiosulfate. Yttrium was determined as the oxalate. For the sample for which the X-ray data are given in Table I, analysis showed: chromium, 28.75 =k 0.05%; yttrium, 46.1 f 0.1%. Theoretical percentages for YCrOs are: chromium, 27.54%; yttrium, 47.06%. If one accepts the formula YCr08 for the reaction product, the above figures would indicate an excess of about 2.5% Cr*Oa. All but two lines of the X-ray powder pattern were indexed on the basis of a monoclinic cell (Table ) with dimensions a = c = 7.61 f 0.01 A.; b = 7.54 0.01 A.; fl = 92'56' &6'. One of the two very weak extra lines matched a medium reflection of the chromic oxide pattern. The remaining chromic oxide lines were effectively covered by the yttrium chromium oxide pattern. Thus a small excess of chromium oxide was indicated by the X-ray evidence and also could be implied from the chemical analysis.
VOl.
70
The monoclinic cell has each of its axes doubled as compared t o the fundamental perowskite unit and thus corresponds to eight molecules. Using the density of 5.78 g./cc., the calculated X-ray molecular weight is 190 f 2 molecular weight units as compared to the formula weight for YCrO; of 188.9. In itself, this agreement is not conclusive proof for the simple formula, because the molecular weight is not a sensitive test of formula. As examples, Y1.08Cro.~20? has rr formula weight of 192; Yu.s,Crl.oaOahas a formula weight of 188. When we consider the X-ray and chemical evidence together, however, the simple formula seems well established. I t cannot be stated with certainty that the a and c axe3 are exactly equal although they are not likely to differ by more than 0.015 A. If they are exactly equal, then choo+ ing the diagonals of the a,c face as new a and c axes gives :L cell for which all the angles are 90'. This cell would imply that the crystal is orthorhombic, rather than monoclinic.
Discussion The reason for the distortion from cubic symmetry,as indicated by the tolerance factor, is the inadequate size of the yttrium ion. The ratio of ytTABLE I trium to oxygen univalent radii would indicate a COMPARISON OF OBSERVED AND CALCULATED INTERPLANAR coordination number of eight3 for yttrium rather SPACINGS than the twelve of the undistorted perowskite f hkl dobs. dcalod. I dobid. dmlod. structure. The atomic positions have not as yet vvw 5 . 5 9 6,52 101 U' 1.310 1 . 3 1 0 441 been determined. However, simply shortening one v\.w 5 . 2 1 6.24 101 1 . 3 1 1 401 axis of the cubic cell could reduce the number of 4.31 W 111 4.29 ,793 1 . 2 9 2 414 nearest neighbors about yttrium from twelve to tn 3.775 3.770 020 1 . 2 9 2 531 200 3.801 vw , 2 7 3 1.272 3.51 eight. The attendant adjustment of atomic posivw 3 . 6 2 4 (3.62) cr203 ,531 1.273 tions to give the most satisfactory coordination 5 3.387 3 . 3 9 3 210 W ,249 1.249 442 polyhedra could then result in a puckering of the 120 3.377 1 . 2 4 8 582 structure as evidenced by the doubled celL4 vw 3.054 3 . 0 5 9 121 1.250 610 211 3.047 W ,206 1.205 812 Acknowledgment.-The authors wish to thank 2.972 2 . 9 6 1 vw 220 0 ,193 1 . 1 9 3 260 W+ Professors Roland Ward and William C. Orr for S 2.756 2 . 7 5 9 102 ,183 1 . 1 8 3 540 W helpful discussions. This work was supported in vvs 2.674 2 . 6 7 7 220 261 1.182 s 2 . 6 2 1 2.619 202 part by the Atomic Energy Commission under 1.184 602 m 2 . 5 8 7 2.591 21" vw 1.171 1 . 1 7 0 612 Contract No. AT-(30-1)-1154. W
W
vw vw m
mm m W +
W
S
m+ mm
2,472 2.365 2.312 2,263
2.223 2.152 2.095 2,049 2.005 1.902 1.883 1.858 1.842
vw
1.815
vv w S
1.737 1.722 1 690
W
1.670
W +
1.606
vvw
1,607
vw m+ m+
1,593
VI'W
S
mi vw vw Wf
nk
+
VYW
2.474 2.368 2.323 2.259 2.266 2,227 2.152 2.097 2,050 2.005 2.007 1.901 1.885 1.859 1.843 1.839 1.813 1.812 1.736 1.723 1.689 1.692 1.672 1.666 1.627 ?
1.594 1.577 1.577 1 . 5 5 5 1.557 1.525 1,524 1.527 1.428 1.429 1.404 1.402 1.380 1 . 3 8 0 1.382 1 , 3 5 7 1.357 1.339 1.339 1.325 1.325 1.326 1.324
212 301 311 311 131 222 222 230 321 321 231 400 040 232 -410 303 232 411 402 33 1 240 412 321 402 412 9
332 422 542 41"
!I3 432 250 a04 423 414 440 333 44 1 252
YW
1.159
1.159 1.159 1.137
524 ,504 533
W
1.137
W
1.139
552
630 542 $61 4.52 832
CHEMICAL LABORATORIES UNIVERSITY OF CONNECTICUT STORRS,CONNECTICUT
W
1 . 1 3 2 1.132 1.117 1.117 1.118 1.118 1.098 1.098
613
W
1.077
1.099
Some Observations on the 8-Quinolinol and 5,7Dihalo-8-quinolinol Chelates of Scandium
vw
W W
mm-
1.076 444 1 . 0 7 7 070 1 . 0 7 0 1.069 E14 1 . 0 1 8 1 . 0 1 8 462 1.018 z54 1.017 703 1.002 1.003 642 0.9936 0.9938 624 0.9940 722
(3) Linus Pauling, "Nature of the Chemical Bond," Cornell U n i versity Press Ithaca, N. Y., 1940, p. 382. (4) H D. Megaw, Acta Cryst., 6, 739 (1952).
BY THERALD MOELLERASD M. VESKATA RAMANIAH RECEIVED JULY 6, 1954
Reaction between scandium ion and 8-quinolinol has been shown' to yield the 1 to 4 chelate, Sc(C9H6N0)&9H6NOH, a compound which is useful for the gravimetric determination of scandium2but one which, unlike the comparable thorium and uranium(V1) c o m p o ~ n d s ,cannot ~ be converted to the normal chelate by heating.' The absorption spectrum of this compound in toluene amounts to an intense peak a t 317 mp (same wave length as is characteristic of 8-quinolinol) and a much less intense peak a t 375 mp.4 Absorption spectra data are interpreted in terms of bonding of the extra mole of 8-quinolinol in the solid by lattice forces4 The similarity of the reported absorption spect m m 4 to the spectra of the partially hydrolytically (1) L. Pokras and P. M. Bernays, THIS JOURNAL, 73, 7 (19.51). (2) L. Pokras and P. M. Bernays, Anal. Chcm., 53, 757 (1451). (3) F. 1. Frere. THIS JOURNAL, 55, 4362 (1933). (4) I,. Pokras, M. Kilpatrick and P. M. Bernays, { b i d , 75, 12J4 (1953).
