Preparation, Identification and Chemical Properties of the Thorium

Preparation, Identification and Chemical Properties of the Thorium Silicides13. By E. L. Jacobson, Robert D. Freeman, A. G. Tharp and. Alan W. Searcy1...
0 downloads 0 Views 405KB Size
4850

E. L. JACOBSEN,

[CONTRIBUTION FRON

THE

R.D. FREEMAN, A. G. THARP A N D A. %’.

VOl. 7s

SE.\KCY

DEPARTMENT OF CHEMISTRY, PURDUE UNIVERSITY, AND THE CERAMIC LABORATORIES, DIVISION OF MINERALTECHNOLOGY, UNIVERSITY OF CALIFORNIA]

Preparation, Identification and Chemical Properties of the Thorium Silicides’” BY E. L.

JACOBSON,

ROBERT D. FREEMAN, A. C,. THARP AND .ALAN W. S E A R C Y ’ ~ RECEIVED APRIL 2, 1956

An X-ray diffraction investigation of the thorium-silicon,system establishes the existence of “p-ThSi,,” ThSi and Th3Si4 p-ThSi2” is hexagonal, space group D1gh - PB/mmm, with ao = in addition t o a-ThSizwhich has been previously reported. 3.985 =!= 0.002 A. and GO = 4.220 f 0.002 A. The phase contains a higher ratio of thorium to silicon atoms than does a-ThSiZ a?d apparently has a real composition of about ThSil.s*o 2 . ThSi is orthorhombic, space group Dl6>h- Pbnm, ith ao = 5.89A., bo = 7.88 A, and C; = 4.16 8. ThsSi2is primitive tetragonal, space group D54h - P4/mbm, n i t h a0 = 7 835 f 0.003 and co = 4.154 f 0.005 A. These phases are isomorphous with corresponding uranlum silicides. Reactibities of the thorium silicides with twelve common chemical reagents are reported.

A study of thorium-silicon compounds was undertaken because, although thorium and uranium usually form similar compounds with metalloids, only one silicide of thorium, a-ThSi2,2 was known compared to six silicides of The known uranium silicides are USi3,4 ~ 5a-USi2, p-USi2,4J USi,4b6 U$3i24,6 and U3Si.4v6 Another phase, ?-US&, has been reported,’ but the cell constants and reported structure are so nearly identical with those for USi3 that the existence of ?-Usit is questionable. 4p6

+

Experimental Mixtures of powdered thorium and silicon in the desired atomic ratios were heated in graphite, tungsten or aluminum oxide containers in Z ~ U C U O to temperatures of 10001700” for varying periods of time. Weights of samples varied from 0 . 5 to 4.0 g. Crucible reaction was not noted unless the Si:Th atomic ratio was greater than 3 : l or less than 0.33: 1. The thqrium powder which was available at the commencement of our investigations contained considerable thorium dioxide. Samples with initial compositions in the range ThSia.o to ThSil.22 (neglecting oxide content) when heated to 1400-1700” for about one hour showed only faint diffraction patterns of thorium dioxide. Lower temperatures or higher thorium contents always gave preparations containing considerable oxide contaminant. It is apparent that in high silicon content samples heated to high temperatures, oxygen was removed by reactions of the type ( n 2)Si T h o , = ThSi, -I- 2SiO(g). T o eliminate oxygen from high silicon content samples prepared a t lower temperatures a few preparations were made using the aluminum solvent technique of Brauer and Mitius.2 These preparations, after dissolution of the aluminum by hydrochloric acid, yielded thorium dioxide-free platelets of a-ThSi2 which if heated to 1460’ for one hour or more were converted t o a phase is:; structural with p-USia and hence designated “p-ThSi2. \Vhen thorium of lower oxide content became available, samples of all desired compositions were prepared successfully by direct synthesis from the elements. Analysis of each preparation was made by standard X-

+

ray powder diffraction techniques. Diffraction patterns were obtained on 114.6 mm. cameras using nickel-filtered copper E;a radiation. The samples were mounted on silica fibers. The “p-ThSi2” Phase.-When samples of a-ThSin were heated above 1400’ and cooled t o room temperature a new phase was found. Diffraction data, which are given in Table I for the first 21 planes of this phase, cor:espond to a hexagonal unit cell with a0 = 3.985 =I=0.002 A. and GO = 4.220 f 0.002 8. Comparison of the spacings and intensities with those for p-USi2 showed the phases to be isostructural and the new phase was accordingly named pThSip. Intensities for each reflection were calculated by using the relationship 1 cos228 = (F2) sin20 cos0 P with the symbols having their usual meaning. The space group is Dl6h-P6/mmm with one thorium in (O,O,0 ) and two silicon in 2 / 3 , z ) with z = The calculated X-ray density is 8.23 g. TABLE I CRYSTALLOGRAPHIC DATAFOR “p-ThSL”

+

(1) (a) Work supported hy t h e Officeof Kava1 Research. Contains material from t h e M.S. thesis of E. L. Jacobson, Purdue Cniversity, June, 1952. (h) Author t o whom inquiries should he addressed. ( 2 ) G. Brauer and A. Mitius, Z . a n o r g . allgem. Chem., 249, 325 (1942). (3) I n a paper t h a t came t o our attention a f t e r structure assignments were completed, G. F . H a r d y and J . K. H u h , Phys.Rev., 93, 1004 (1964), i t is stated Zachariasen has found a p-ThSiz phase and a ThlSi? phase and has made t h e same space group assignments as we find for two of t h e phases reported here. (4) P. Gordon, B. Cullity, h l . Cohen, G. Bitsianes, A. R . Kaufmann and R . B. Bostian, quoted h y J. J. Katz and E. Rabinowitch in “The Chemistry of Uranium,” McGraw-Hill Book Co., Inc., New York, N . Y . , 1951, pp. 206-228. ( 3 ) B. R. T . Frost and J . T . IIackrey, J . 11zs1. M e t a l s , 82, 177 (l!lt53). ( 0 ) W. H . Zachariasen, Acta Crysf , 2, 94 (1949). (7) G. Brauer a n d H. Hoag, Z. anorg. aligem. Chew., 269, 197 (1949).

I‘

sin2B Ilk1

Obsd.

001 100 101 002 110

0.0334 ,0499 ,0835 ,1336 ,1492

Calcd.

Obsd.

Calcd.

0.0334 in 177 ,0499 S 752 ,0833 VS 2480 ,1333 w+ 346 ,1497 S744 ,1834 lo2) ,1836 257’/ 77 111 ,1833 200 ,2000 ,1996 IV + 112 ,2330 S48 1 20 1 ,2329 ,2832 S506 112 ,2834 ,3331 w 20 202 3323 ,3302 S210 lo’) ,3502 90 ,3493 211 .3826 ,3827 S430 .4491 ,4494 ,4500 rn112 212 ,4828 n,+ ,4826 26 ,4825 301 203’ ,4996 ,4999 m148 ,5826 188) ’02) 104 ,6836 ,5838 49 vs, very strong; s, strong; m , medium; w, weak.

2461 1

The X-ray structure determination and the formation of 0-ThSL b y heating wThSi2 suggested p-ThSi2 to be a high temperature modification of ThSi,. More thorough investigation, however, showed tliaf a-ThSi? resulted from heating samples with Si:Th atom ratlos 2 : 1 or higher a t any temperature in the range 1000 to 1670°, while samples of lower silicon content heated in the s m i e range yielded p-ThSiz. A series of experiments mas accordingly devised to p r o w whether or not the elemental compositions of the 01 and p phases are identical.

PREPARATION AND PROPERTIES OF THORIUM SILICIDES

Oct. 5, 1956

4851

Samples of initial composition ThSi3.0 and ThSi2.o were TABLE I1 heated simultaneously a t 1520". After heating, the diffracCRYSTALLOGRAPHIC DATAFOR ThSi tion patterns were those of pure a-ThSiz.o and of pure 8sin% Ia ThSi2.0, respectively. These samples were crushed to fine hkl Obsd. Calcd. Obsd. Ca1cd.b powders, and enough silicon was added to the p-ThSiz to 0.0267 110 0.0267 W 38 raise the silicon-thorium ratio to 3 : 1. Both samples were then reheated to 1670' for one hour and cooled to room tem.0382 020 .0383 m84 perature. Diffraction analysis showed that the a-ThSiz had 101 .0517 ,0516 m 167 been converted to B-ThSiz, and, that the p-ThSi2 had re120 .0554 m 112 .0554 acted with the excess silicon to yield a-ThSil. A sample of S 285 .0611 111 .0611 P-ThSis heated at 1000° for three days was not converted to a-ThSi2. .0684 200 Absent .. 0 These results are consistent with the conclusion that p.0727 02 1 0.0727 S 174 ThSi2 contains a lower concentration of silicon than does a210 S 214 .0780 .0779 ThSi2 and that a-ThSiz is converted to P-ThSiz a t high tem121 .OS98 .OS98 m 111 peratures by silicon vaporization, perhaps partly as SiO. Comparison of diffraction patterns with compositions, after .lo32 ,1032 130 rn 102 allowance for Tho2 impurity, fixes the composition of a.lo67 Absent 220 .. 0 ThSiz a t about the ideal composition for ThSil. The actual ,1124 0.1123 211 m 82 composition of the P-ThSi? phase appears to be ThSil.5 A 0.2. ,1376 Both P-USiz4 and B-PuSi28 have been assigned compositions .1380 of MSi1.5 from composition studies, although Zachariasene 002 .1378 63 discounted the uranium-silicon composition studies because 22 1 Absent .1411 . . 0' of his structure results. 040 Absent ,1530 .. 0 The ideal 8-phase structure consists of close packed sheets Absent .1545 .. 4 230 of metal atoms placed directly over each other with an intermediate sheet of silicon atoms in each of the two holes Absent 310 .1636 .. 5 per metal atom. A hexagonal close packed lattice of uni112 .1645 w8 0.1652 form spheres differs from this structure in that in packing .1699 140 ,1701 W 44 of uniform spheres only half the holes between the top and 022 .1760 W 21 .1765 bottom layers are filled. All positions in the intermediate layer can be filled without distortion only if the atoms of the .1874 m+ ,1882 intermediate layer are much smaller than those of the top 231 .1889 107 and bottom layers. It is probable that silicon vacancies w, weak; m, medium; s, strong. These are Zachariaoccur in the ideal P-MSis lattices because of crowding. The silicon-silicon atoms in B-ThSiz are 0.10 .&. closer to- sen's values for the intensities of USi6 and are within the error of visual observation since the scattering factor of gether than in a-ThSi2. The ThSi Phase.-Samples of composition near ThSil.o thorium differs little from that of uranium. when heated to 1500-1700' gave, in addition to P-ThSi2, TABLE I11 strong diffraction patterns of another phase. The phase was also obtained by prolonged heating of a sample of BCRYSTALLOGRAPHIC DATAFOR Th3Siz ThSi2 in vucuo a t 1450". sin% I5 The X-ray data for the first 23 reflections of this phase hkl Obsd. Calcd. Obsd. Calcd. are shown in Table 11. The sin% values can be fitted to an 100 Absent 0,0097 . . 0.0 orthorhombic unit cell with ao = 5.89 A., bo = 7.88 A. and 110 0.0194 .0194 w 27.2 co = 4.15 A. Comparison of the lattice spacings and in001 ,0343 ,0344 \\' 16.3 tensities for this phase and USi indicated that they are isostructural. The only other Group I V b element that has 200 .. 3.0 Absent .0387 been shown to form a monosilicide of this structure is zir101 Absent .044 1 . . 0.0 conium.g ThSi, along with USi and ZrSi, contains four 210 0.0485 .0484 m' 117.8 molecules per unit cell. Four thorium atoms are located 111 .0537 .0538 in 55.9 1 / 2 , l / 4 ) and four silicon atoms in f (x,y, 1/4), ( l / ~- x , y l / ~ ,l / d ) . The are located in & ( u , v , l/4), ( l / ~- u , v 20 1 ,0730 ,0732 S+ 259.2 parameters assigned to x and y are 0.13 and 0.18, respec220 .0774 ,0774 w+ 74.9 tively, and those assigned to u and v are 0.61 and 0.03, 211 ,0829 ,0828 S 190.3 respectively. The space group is Di62h - Pbnm. The cal300 Absent ,0871 .. 0 0 culated density is 9.03 g. cm.-3. The Th& Phase.-Preparations with Si: T h atomic 310 0.0968 ,0968 m+ 117.1 ratios of about 0.82 yield, in addition to ThSi, a diffraction 22 1 .. 0.8 Absent ,1119 pattern very similar to that for U3Si2. Calculated sin20 30 1 Absent .1216 .. 0.0 values were fitted readily to a unit cell with tetragonal sym320 0.1259 ,1258 w -8.5 metry. The space group of U3St is D54h - P4/mbm with t w o metal atoms in (0, 0, O ) , (I/?; 1 / 2 , O ) , four metal atoms 311 Absent ,1312 .. 3.4 in f ( u , u l/?, I / z ) , ( l / ~- u , u , / 2 ) , and four silicon atoms 002 0.1371 ,1378 m51.5 in f ( v , v 1/2,,0)? (I/? 7 v , v , 0). Intensities calculated 102 Absent .I474 .. 0.0 for the thorium silicide using Zachariasen's values6 of u and v ( u = 0.18, v = 0.39) for Upsisgave satisfactory agreement w , weak; m, medium; s, strong; vs, very stroiig. with observed intensities. The silicide, therefore, is Th3Si2, isostructural with U3Si2. The cell is primitive tetragonal initial silicon content than ThSi3.o gave diffraction with uo = 7.835 0.003 b. and GO = 4.154 f 0.005 h. and has a calculated density of 9.80 g. Table I11 gives patterns of cu-ThSiz and silicon, and those of lower silicon content than ThSio.5 gave patterns of Th3data for the first 18 crystallographic spacings of this phase.

+

+

+ +

Discussion Since phases identified as USi3 and UaSi have been found in the uranium-silicon system, corresponding phases were intensively sought in the thorium-silicon system. Preparations of higher (8) 0. J. C. Runnalls and R . R . Boucher, A c f a Cryst., 8 , 592 (1955). (9) H. Schachner, H. Nowotny and H. Kudielka, Monulsh. Chem., 86, 1140 (1954).

Si2 and thorium as well as considerable amounts of Tho2. Thorium silicides of formulas ThSia and Th3Si either do not exist or are of limited stability. Apparently plutonium also does not form an MSi3 phase.8 Chemical Properties.-Finely powdered samples of the thorium silicides were tested with twelve common chemical reagents. Samples were allowed to stand for a t least 12 hours in each reagent.

4852

R. H. SANBORN AND E. F. ORLEMANN

VOl. 78

Solutions and samples were then heated to the ThaSi2 were attacked by 5% hydrochloric acid. boiling point of the solutions for several minutes. All the thorium silicides were unstable on proIn no instance was a reaction observed on heating longed standing in air. They disintegrated slowly if none had occurred with cold solutions. to form a fine black p o d e r which was aniorphous -411 thorium silicides reacted with concentrated to X-rays. Samples of ThSi sometimes reacted hydriodic acid, concentrated hydrochloric acid with air rather rapidly. *Melting points were not concentrated hydrofluoric acid and aqua regia. determined, but a-ThSin and “fl-ThSiZ” sample9 Sone of the silicides reacted with 20y0 sodium shoced no sign of melting when heated to 1Gi0” hydroxide, 30% hydrogen peroxide, 0.1 AT potas- while samples of ThsSiz were melted a t 1700”. sium permanganate or 5To nitric acid. Only Samples of ThSi were never obtained free of Th3Siz reacted appreciably with 36 N HzS04 al- “@-ThSiz”or Th$iz and may be stable over only a though both Th3Siz and ThSi reacted with 6 -Ir limited temperature range. HZSO4. Only ThSi was even slightly attacked by LAFAYETTE, INDIANA concentrated nitric acid. A11 compounds except BERKELEY,CALIFORUIA

[ CONTRIBUTION

FROM THE

DEPARTMEST

O F CHEMISTRY A S D CHEDlICAL EXGINEERISC A S D THE vNIVERSITY O F CALIFORNIA, BERKELEY]

RADIATIOV LAEORATORL,

The Formation of Unipositive Nickel by Electrolysis in Concentrated Salt Solutions1’ BY RUSSELL H. SANBORN AXD E. F. ORLEMANN RECEIVED MAY4, 1956 The reduction of Si(I1) a t the dropping mercury electrode leads to the forination of S i ( 1 ) where the concentration of various salts, such as XaC101, LiCIOI, Ca(C101)2, S a C l and KCl, is made sufficiently high. In general, the formation of X ( I ) becomes clearly evident a t salt concentrations of about one molar and becomes the principal process where the salt coiicentration is made two t o three molar. Evidence for the formation of Ki(I) lies in a comparison of the observed diffusioii currents with directly measured diffusion coefficients, the existence of a n intermediate state of nickel that reacts with bromate, and the fact that the amount of metallic nickel produced per Faraday of electricity in the concentrated solutions is much less than that obtained in dilute salt solutions. Various experiments t o characterize the plus one state of nickel are described and some possible explanations of the phenomenon are presented.

tion of two Ni(I1) species, that have different activation energies of reduction, which are not maim tained in equilibrium. Curves 3 and 4 show the effect of increasing the NaCl concentration to !2 and 3 MI respectively, The change in slope becomes more marked as the salt concentration is increased and the height of the first section of the c.v curve decreases. The limiting current of the first section will be referred to as il. It is measured approximately by the intersection of the extrapolated slopes of the initial and second “linear” -1 Sargent Model XI1 Polarograph was used to record sections of the c v curves. TTalues of il so obthe current-voltage curves. The electrolysis cell was of the tained are quite uncertain where a curve such as 2 €1-type, with a sintered glass disk and an agar plug separat- is analyzed and become reasonably precise for ing the two arms. Oxygen was swept out of the solutions with purified nitrogen. -411 measurements were made a t curves such as 3 and 4. In the light of data presented subsequently ill 25 i.0.03”. .ill reagents mere reagent grade chemicals except LiCIOa this paper the curves in Fig. 1 are to be inter;itid Ca( Clod)>,which were crystallized out of solution prepreted as follows. The reduction of Ni(I1) is a pared by the action of perchloric acid on the corresponding two-step process. In 0.1 -11 NaCl the Xi(1) cnrbonates. initially produced disproportionates rapidly or is Results Current-Voltage Curves of Nickel in Concen- immediately further reduced to metal a t the electrode and the total diffusion current corresponds to trated Solutions of Various Salts.-Current-voltage complete reduction to the metal. I n I 51 NaC1, curves of ?Ji(II) in NaCl solutions from 0.1 to 3.0 il.(r containing lop3 Jf HCl are shown in Fig. 1. curve 2, the activation energy for electrode reducCurve 1 in Fig. 1 shows a single continuous C.V. tion of Ni(1) ha5 increased so that the stepwise curve corresponding to the reduction of Ni(I1) to nature of the reduction becomes observable the metal. Curve 2 obtained in 1 M NaCl sho,ws Tl‘here the NaC1 is made 3 M this kinetic “stabilia distinct change in slope which suggests either a zation” of Ki(Ii has become almost cumplete and stepwise reduction of Ni(I1) or possibly the reduc- the limiting current il is due almost entirely to rcduction to Ni(1) with only a very small contribu(1) This paper comprises a portion of t h e P h . D . Thesis of Russell H. tion by further reduction to the metal. It should Sanborn, L‘niversity of California a t Berkeley, December, 1955. T h e research was supported in part by A E C Contract W-7405-eng-48 be noted that the Xi(I), which is about -IT, a t t h e Radiation Laboratory, Berkeley, California. need only have a half-life greater than a few im1(2) Presented before the Physical and Inorganic Division of t h e liseconds to account for curves 3 and 4. 128th h-ational ACS Meeting, Minneapolis, M i n n . , September, 1966.

In the course of a study of the mechanism of reduction of the first transition group metals a t the dropping mercury electrode, it has been found that nickel shows a marked difference in behavior from the other members of this group in concentrated salt solutions. This paper is concerned with a study of the reduction of Ni(I1) in concentrated salt solutions and the evidence that Xi(1) is obtained as the primary product under certain conditions. Experimental