1804 [CONrRIBUTION FROM
V. I. MATKOVICH RESEARCH LABORATORIES, ALLIS-CHALMERS MANUFACTURING COMPANY,
Vol. 83 MILWAUKEE,
WISCOSSIS]
Interstitial Compounds of Boron BY V. I. MATKOVICH* RECEIVED NOVEMBER 12, 1960
A model of the structure of boron compounds is presented which explains such compounds as BI3O2,B4C, B&, BI3PZ, BL~ASZ and B4Si as a group of interstitial compounds of boron. Formation of a new boron sulfide, E&, belonging t o the same group is reported.
Introduction Recent developments indicate that such conipounds as B4Si1-3 and B13P24as well as both boron carbides B4C and B13C26'6can be derived from a recently discovered boron modification7,*by placing the corresponding elements in its interstitial openings. A study of the boron-sulfur system has been carried out as a part of investigation of compounds related to boron or boron carbides. Mechanism of Formation of Interstitial Boron Compounds.-According to Zhdanov and Sevastyanovg and Clark and Hoardlo the B4C structure consists of linar chains of three carbon atoms and groups of twelve boron atoms placed a t the vertices of nearly regular icosahedra. These are distributed in a rhombohedron in a NaCl type arrangement. The smallest rhombohedron representing the unit cell contains three B4C molecules corresponding to twelve boron and three carbon atoms (see Fig. 1). The three carbon atoms lie on the trigonal axis of the rhombohedron which is the c axis of the hexagonal representation. With respect to the symmetry and coordination, the two outside positions of the carbon chain 2c are identical and different from the central one 1b. l b is the only unique position in the boron carbide unit cell and was found to accommodate boron as well as carb0n.l' Thus the 4 - C - C - chain can transform into -C-B-Carrangement. l 2 Complete substitution of carbon by boron in l b position results in the ratio of B13C2.5J The homogeneity range in the boron-carbon system is thus explained as a complete series of solid solutions between B4Cand BI3C2. Closely related to the above described structui e is one of the recently reported modifications of b o r ~ n . ~While , ~ the framework of boron icosahedra appears to be identical with that found in boron carbide, the three carbon positions are vacant and represent interstitial openings. The rhombo* T h e Carborundum C o m p a n y , Research a n d Development Division, Niagara Falls, N.Y. (1) V. I. Matkovich, Acta C ~ y s t .13, , 679 (1960). (2) E. Colton, J . A m . Chem. Soc., 82, 1002 (1960). ( 3 ) C . F. Cline a n d D. E. Sands, Natuve, 185, 456 (1960). (4) V. 1. Matkovicb, A c t a Cryst., 14, 93 (1961). (5) G. S. Zhdanov, N. K. Zhuravlev a n d L. S. Zevin, D o k l a d ~A k a d . X a u k S.S.S.R., 92, 767 (1953). ( 6 ) G. V. Samsonov, N. N . Zhuravlev and I. G. Amnuel, Piz. Metal. i Mplallncd. A k a d . .Voz?k S.S.S.K. Ural Filial, 3, 309 (19.56). (7) I,. V. NcCarty, J. S. Kasper, F. H. H o r n , B. F. Decker and A. E . Newkirk, J . A m . Cizem. Soc., 80, 2595 (1958). (8) B. F. Decker a n d J. S. Kasper, Acta Cryst., 12, 503 (1959). (9) G . S. Zhdanov a n d N. G . Sevastyanov, Compt. rend. A c a d . Sci. U.R.S.S., 32, 432 (1941). (10) H. E;. Clark a n d J. L. H o a r d , J . A m . Chem. Soc., 65, 2115 (1949). (11) H. A. Silver and P. J . Bray, J . C k e m . Phys.. 11, 247 (1959). (12) R . D . Allen, J . A m . Chem. SOL.,75, 3582 (1953).
hedral unit cell thus contains 12 rather than 13 atoms of boron. The large interstitial openings in boron suggest a possibility that the elements other than carbon may be accommodated in these positions forming a series of isostructural boron compounds. The interstitial compounds (interstitial compounds are often referred to as interstitial solid solutions) are described in a simple way as involving the arrangement of relatively large metal atoms in closest packing or in some other simple structure and the insertion of the small non-metal atoms into interstices of the lattice of metal atoms.13 The small atoms usually are found to be boron, carbon. nitrogen or oxvgen. For examde. the series of intervtitial tita&m compounds lncludes TiB, T i c , TiN and TiO. The interstitial compounds of boron can be understood using the same description by considering the boron icosahedra as single units. Thus. the arrangement of boron icosahedra forms a somewhat distorted face centered cube, equivalent to the close packed atomic structure of large metal atoms. Since the icosahedra are large compared with the metal atoms, we find that as many as three non-metal atoms may be accommodated in the interstitial openings and the openings can accommodate larger elements of the second and third row of the periodic system as well as those of the first row. Once the mechanism of formation of interstitial boron compounds is recognized, the isostructural compounds are identified easily by means of Xray power photographs. In fact the identification of a series of these compounds consists in corrections of the known data rather than in thc discovery of new compounds. The lower boron phosphide prepared by high temperature decomposition of BP, reported to have a chemical formula of BcPI4,is properly identified on the basis of cheinical analysis and X-ray data as B13P2.4 Similarly, the high temperature decomposition product of BAS reported to have proportions of Bg7As14 and an orthorombic unit cell15 has been correctedI6 aiid found to have a corresponding rhombohedral unit cell containing one molecule of B13-ils2. The substitution system appears to include also the elements of the sixth group of the periodic system. Thus a lower boron oxide analyzed to be B6.60 and reported'' as an orthorhombic B70 also has been corrected1fiand found (13) G . H B g g , Z. p h y s i k . Ciaem., B6, 221 ( 1 9 2 9 ) ; B2, 33 (19313. (14) F. V. Williams a n d R . A . Ruehrwein, J . A m . Clzem. Soc.. 8 2 , 1330 (1960). (15) J. A. Perri, S. LaPlaca a n d B. Post, Acta Cf'ysf., 1 1 , 310 (1958). (16) Personal Communication from B. P o s t , Polytechnic Institute of Brooklyn. ( l i ) R . rl. I'asternak. . 4 d u Cryst., 12, til2 (l%!tj.
INTERSTITIAL COMPOUNDS OF BORON
April 20, 1961
rhombohedral with the appropriate formula of BI3O2. Thus the interstitial positions in boron structure are found to accommodate silicon, phosphorus, arsenic and oxygen as well as carbon, forming compounds isostructural either with BqC or B13C2, depending on whether the l b interstitial position is occupied by boron or the substitution element. The unit cell sizes of the above mentioned compounds are compared in Table I.
0
1805
BORON -6h,
BORON
0
-6h,
CARBON
-2c
TABLE I COMPARISON OF THE UNIT CELLSIZES OF THE INTERSTITIAL BORONCOMPOUNDS Hexagonal cell
A. 4.908 a,
Boron BiaOt BizC3(B*C) B13CZ
5.37 5.60 5.67 B13P2 5.984 BUASP 6.142 BlnSia(BSi) 6.330
c,
A.
12.567 12.31 12.12 12.19 11.850 11.892 12.736
c/a
2.56 2.30 2.17 2.15 1.98 1.935 2.015
Rhombohedral cell a, A, u
5.057 5.14 5.175 5.218 5.248 5.319 5.602
58" 4' 62'56' 65' 18' 65'49' 69'31' 70'32' 68' 49'
Fig. 1.-Rhombohedral unit cell of boron carbide (after Silver and Bray, ref. 11).
a t above 1700' results in further evolution of sulfur leaving boron as the residue. Measurement of temperatures was found to be exceedingly difficult as evolution of sulfur vapors The compounds in Table I are arranged in the interfered with the reading of the optical pyincreasing order of the atomic sizes as we go down rometer. The temperatures therefore were estithe table.l* It is interesting that the expansion mated by running a blank under the same condiof the unit cell takes place in the direction of the tions. Black samples obtained by above described prohexagonal a-axis only. At the same time the cedure contained less than 1% and up to 2% unit cell shrinks in the direction of the c-axis until aocertain minimum value is reached (about carbon, with unreacted boron in some samples. Chemical analysis of samples indicated a boron 11.85 A.). From there on the expansion takes place in all directions. The rhombohedral unit to sulfur ratio of 1 2 : l . Somewhat higher ratios, cell expands accordingly by adjustment of the up to 13: 1, also were obtained, but they are explained by the presence of unreacted boron. The angle rather than by change of the axis size. From the positions in the periodic table and on compound hence is assigned the formula of B12S. The X-ray powder photograph of B12Sis very the basis of the atomic sizes i t appears that a t similar to that of B13P2 and is readily indexed least two more interstitial boron compounds should exist: namely, a lower boron nitride and a lower (Table 11) assuming a rhombohedral ynit cell with hexagonal dimensions of a = 5.80 A. and c boron sulfide. Preparation of Boron Sulfide ( B U S ).-Boron sul- = 11.90 A. The corrpponding rhombohedral fide was prepared by the reaction of amorphous boron dimensions are a = 5.19 A. and CY = 67' 56'. Denand sulfur powders a t about 1600-1700". The sity was determined by floating the powder in heavy boron powder obtained from the United States liquids and by the pycnometer method. The Borax and Chemical Corporation was 95-97% measured value was found to be about 2.40 g./cm.3 pure. The sulfur powder was of U.S.P. quality which is somewhat higher than the theoretical .~ from X-ray data, and obtained from Merck and Company, Inc. value of 2.33 g . / ~ m calculated assuming one BlzS molecule per rhombohedral The mixture of elements containing equal amounts of boron and sulfur by weight was pressed into a unit cell. Presence of about 2% carbon corpellet, placed in a graphite crucible and fired in an responds approximately to 9% BIC or about 13% induction furnace. The rate of heating is very B&2. As there is no evidence for the presence of important as the slow heating results in sublima- boron carbide, i t is assumed that a small amount tion of sulfur, either in the elemental state or as of carbon is placed in the vacant interstitial openB&, leaving boron as the residue. However, ings in the BlzSstructure. when a temperature of 1600-1700' is obtained Discussion within one or two minutes, a reaction between The new lower boron sulfide, B12S, while of boron and sulfur takes place forming a fine brown or black powder. Upon further heating a t 1600- different proportions belongs, on the basis of its 1650°, the brown powder releases some sulfur and structure, to the formerly described group of inturns black. I n order to avoid formation of boron terstitial compounds of boron. The single atom carbide from reaction with the graphite crucible, of sulfur should logically be placed in the l b posithe heating was discontinued as soon as a tempers- tion, since that is the only uniquely defined positure of 1600-1700' was reached. The samples tion in the structure. The two 2c positions therethen were fired again in zirconia crucibles a t about fore appear vacant and could possibly accommo1650' until evolution of sulfur vapors ceased and date other elements such as carbon. The fact that the sample appeared uniformly black. Heating the X-ray powder photographs do not show the presence of boron carbide even when up to 2% (18) Boron silicide is placed below boron arsenide, because t h e s u m carbon is present indicates that this is the case. of sizes of three silicon atoms is larger t h a n t h a t of tmo atoms of arsenic a n d one of boron. However, a defect structure also is possible.
1806
D. J. PHILLIPS AND S. Y. TYREE, JR.
With the discovery of BlzS there appear to be three types of interstitial compounds of boron, namely, BI2X,B12X3and B13X2. The 2c positions in the structure are found to accommodate carbon, silicon, phosphorus and arsenic while the l b position can accommodate boron, carbon, silicon and sulfur. A homogeneity range similar to that found in the boron-carbon system may also exist in some other boron systems. However, with the present methods of preparation they have not been observed. Though the proposed mechanism explains satisfactorily the formation of a series of boron compounds involving elements of the third, fourth, fifth and sixth groups, a direct determination of the actual atomic arrangement would be desirable. l9 The X-ray scattering power of silicon, phosphorus, etc., is sufficiently different from that of boron to enable definite placement of these elements in the structure. Acknowledgments.-The author wishes to express his appreciation for technical assistance provided by J. L. Peret, of this Laboratory, who prepared the samples of B12S,and to R. G. Greenler for helpful discussions and the preparation of the manuscript.
TABLE I1 X-RAYPOWDER DIFFRACTION DATAFOR B12S Obtained by use of 140 mm. camera and filtered CrKn radiation
A.
I/I;
20
1 5 1 6 3 5 0
3.97 3.84 2.90 2.566 2.454 2.159 1.875 1.641 1.545 1.486 1.449
1'
1.383
20
1.359
20
h
k
0 0
0 1 1 0 2 1 1 1 0 2 2
3 2 0
3 2 1 0 0 0
1 1 0 0 2 1 3 1 2
d,
4
TO 10 100 100 10 10 10
5 30 20 20
1 2 3 2 3 4
50 1.251
Vol. 83
1
(19) Such work is in progress at the Polytechnic Institute of Brooklyn by B. Post, on BnPz crystals supplied by this Laboratory.
5
[CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, UNIVERSITY OF NORTH CAROLINA, CHAPEL HILL, NORTHCAROLINA 1
The Donor Properties of Triphenylarsine Oxide BY D. J. PI-IILLIPS AND S.Y. TYREE, JR. RECEIVED SEPTEMBER 9, 1960 The coordination chemistry of triphetij-larsine oside with :i variety of acceptors has been iiivestigated. The new substances CrC1?.4R, CrCl2C1O4.4R,MnCi2.2R, Ain{ClC~,),.4R,FeCl3.2R, Fe(Cl( i4),.4P, CoC12.2R,CoBr:.2P., Co(C104)?.4F, NiC12.2R, N I ( C ~ O ~ ) Z ..Cu(NOs)t4R, ~R, 211(C10~)>4?, HgCI,.2RI IlgCl?.K, SnC14.2K,SnBra.2R, SbCI8.2R, and ShCl3.R (where R = triphenylarsine oxide) have been isolated and characterized by molecular conductallce measurenients in nitrobenzene, cryoscopic molecular weight measurements in freezing nitrobenzene, magnetic susceptibility measurements and infrared spectra. Most of the halides appear to be non-electrolytes, whereas the nitrate and perchlorates appear to be
salt-like.
Introduction In the course of some previous work we became aware of the class of compounds formed between triphenylphosphine oxide and metal halide accept0rs.l Now we wish to report extensions of the earlier work to a wider variety of metal compounds and using triphenylarsine oxide. During the course of the work two other laboratories2 reported analogous series of phosphine oxide compounds. Also Lindquist3 has prepared and studied some additional similar phosphine oxide complexes in the interim. Workers in Prof. Hieber's laboratory a t Munich also report addition compounds of iron and cobalt carbonyls with triphenylarsine oxide. 4t6
(1) R. H.Pickard and J. Kenyan, J. Chcm. SOL.,262 (1906). (2) (a) F. A. Cotton, e l ai., Proc. Chem. SOL.(London), 158 (1958); (b) F. A. Cotton, et ol., J. Chcm. SOL.,1873 (1960); (c) 1878 (1960); (d) 1959 (1960); (e) 2266 (1960); (f) K. Issleib and B. Mitscherling, 2. onorp. ullgrm. Chcm., 304, 73 (1960). (3) I. Lindquist, private communication. (4) W. Hieber and A. Lipp, Chem. Bcr., 92, 2075 (1959). (6) G. Franz, Doctoral Dissertation. Technischen Hochschule Munchen, 1959.
We believe the first examde of an arsine oxide complex to be (C3H7)&0:2HgCl2, reported by Partheil and co-workers.6 The only other examples we are aware of are Cu [(C6H~)2CH3As0 ]&, reported by Nyholm,? and the chelates bis-(apicolyldirnethylarsine oxide) -copper(I) perchlorate and bis- (a-picolyldimethylarsine oxide) -copper (II) perchlorate, reported by Goodwin and Lions Experimental Reagents.-Reagent grade chemicals were used without further purification except in the cases noted. Triphenylarsine oxide was prepared from Eastman "White Label" triphenylar~ine,~ m.p. 194.5-196°; literature values, 189°'oJ1 and 191-193".'* A n a l . Calcd. for ( C B H S ) B X ~As, O : 23.25. Found1$: AS, 23.0. (6) A. Partheil, et ul., Archru. Phorm., 157, 136 (1899). (7) R. S. Nyholm, J. Chcm. SOL.,1767 (1961). (8) H. A. Goodwin and F. Lions. 1.A m . Chcm. So., 81, 311 (1959). (91 R. L. Shriner and C. N. Wolf, 07g. Syn., SO, 97 (1950). (10) A. Michaelis, A n n . (Liebig), 201, 244 (1880). (11) F. Zuckerkandl and M. Sinai, Be?'., 64, 2485 (1921). (12) F. G.Manu. J . Chem. Sac., 970 (1932).
