7516
C2BjHi and CH3CBjHs,other major by-products in the diborane-CzB3Hj reaction, are possibly formed by direct assimilation of BzHG by C2B3H5,i.e.
+
CzB3H5 BzHs +CH3CBsH8 CzB& BzH6 +CZBLH? 2Hz
+
+
(4) (5)
Alternatively, C2BjH7 might arise by loss of BH3 from C2B6H10. CzBaHio +CzBsH,
+ BH3
(6)
Acknowledgments. We acknowledge the support of this research by the Office of Naval Research, the President's Fund of the California Institute of Technology, and NASA Contract NAS 7-100. We are also grateful to Professor A. B. Burg of the University of Southern California for obtaining the 32-MHz IlB nmr spectra. Dr. Burg, with T. F. Reilly, has independently discovered a higher yield, lower temperature route for the production of CzB6H~owhich will be reported by them at a later date. (1 1) University of California, Irvine. (12) Send requests for reprints to Chemical Systems Inc., Irvine, Calif. 92705.
i
Alan J. GOtcher,11j12 J. F. Ditter,l2 Robert E. Williams*lZ Department of Chemistry, University of California Irvine, California 92664 Chemical Systems Incorporated Imine, California 92705 Receioed March 13, 1973
I
Chemistry of Thiocarbonyl Complexes of Chromium, Molybdenum, and Tungsten
Sir: Several metal thiocarbonyl complexes have previously been but little is known about the chemistry of the thiocarbonyl ligand. We now report the preparation of the first group VI metal thiocarbonyl complexes and some novel reactions of these compounds. Anions of Cr, Mo, and W were prepared by stirring the hexacarbonyls, M(CO)6, with excess sodium amalgam in refluxing tetrahydrofuran for 12 hr. Figure 3. The 60-MHz lH nmr spectrum of 4,7-CZB&. Studies of Hayter4 and Kaskaj indicate that M2(CO)L~2and possibly some M(CO)j2- are produced i n these though it is a nido-borane according to its empirical reductions. These solutions were cooled to room formula,6and/or equivalent electron counting s ~ h e m e , ~ ' ~temperature and added rapidly to a tetrahydrofuran has an open arachno structure. solution of thiophosgene (C1,CS). After stirring the A tentative reaction mechanism which may account mixture several minutes, it was evaporated to dryness for the production of C ~ B G Hand ~ Oits by-products folunder reduced pressure. Sublimation of the residue lows. In the thermal decomposition of diborane one at 50-60" under high vacuum yielded a mixture of of the transient products is generally conceded to be M(CO)6 and M(CO)j(CS). The M(CO), complexes [B3H7],present in pseudoequilibrium with BzH6, lo and were removed from the mixtures by repeated crystalsince diborane is thermally Iess stable than CPBBHB,a lization at -20" from pentane or hexane solutions, in plausible mechanism for formation of C2B6H10 is the which the thiocarbonyl complexes were appreciably reaction more soluble. Chromatography of the thiocarbonylcontaining solutions on Florisil with pentane followed CZBSHS [BzHd +CzB6Hio Hz (2) by sublimation yielded the pure air- and moistureCzB6Hsis produced as a by-product, and at slightly stable yellow M(CO),(CS) complexes.6 higher reaction temperatures the ratio of CzB6Ha to (1) I. S. Butler and A. E. Fenster, J. Chem. SOC.D, 933 (1970), and C2B6H10increases proportionately with the temperature therein. increase, suggesting that C ~ B G evolves H~ from C2BsH10 references (2) A. E. Fenster and I. S. Butler, Can. J. Chem., 50, 598 (1972). by simple loss of hydrogen, i.e. (3) M . J. Mays and F. P. Stefanini, J . Chem. SOC.A , 2747 (1971).
i
+
+
CZBGHIO +CnB6H8
+ Hz
(3)
(8) I&C=&)R
+-+
' I
I
"H1
0
* '&-t-CO,R
\ ,Si=C-CO,R
c
When ethyl trimethylsilyldiazoacetate is photolyzed with a high-pressure mercury lamp in alcohols, four products, 1-IV, are obtained in approximately 60-90 % ,SiMe,
K !C OC'
+
ROH
Et
hv
H
I
+
Me,%-C-C0,Et
1
OR I
H
I
Me Si-C-CO-R
I
OR
H
Me
I + Me,Si-CC-CO,R I
I
+ MeLSi-C-CO,Et I
I1
I
RO
OEt
H
N
I11
yield. All four products were structurally characterized by ir, nmr, and elemental analyses. Product 1may be rationalized in terms of a trimethylsilyl(carbeth0xy)carbene reaction, and product 111 has been explained by a Wolff rearrangement, probably not involving carbene ir~termediate.~,~ Product I1 may be derived via both carbene and ion pair4 j paths. The formation of IV, with a methyl migration, is quite interesting and is the main concern of the present communication. This reaction is quite general since rearranged products were formed in each system studied. The efficiency of migration varies with respect to solvent alcohol in the order t-BuOH > i-PrOH > EtOH > MeOH, in contrast with 0-H insertion products (1 and 11) (Table I). Table I. Photolysis of Ethyl Trimethylsilyldiazoacetate in Alcohols Alcohol MeOH EtOH= i-PrOHb t-BuOH
---Relative I 64 87 35 40
product yields, %---
I1
111
IV
14
12
10
22 17
22 8
13 21 35
a Products I, 11, and III are all identical in ethanol. yields of I and I1 are obtained by nmr.
Relative
This implies that when the carbene does not show sufficient reactivity toward an oxygen atom (formation of products 1 and I1 via oxygen ylide), probably owing to steric hindrance of a neighboring bulky group,6 then
J . Amer. Chem. SOC.,94, 7926 (1972).
(3) U. Schollkopfand N. Rieber, Afzgew. Chem., 79, 905 (1967). (4) 0 . P. Strausz, T. DoMinh, and H. E. Gunning, J . Amer. Chem. SOC.,90,1660 (1968). (5) T. DoMinh, 0. P. Strausz, and H. E. Gunning, ibid., 91, 1261
(2) P. Boudjouk and L. H. Sommer, J . Chem. SOC.,Chem. Commrrtl., 54 (1973).
(1969). (6) W. Ando, I. Imai, and T. Miglta, J . Org. Chem., 37,3596 (1972).
(1) P. Boudjouk, J. R. Roberts, C. M. Golino, and L. H. Soinmer,
Journal o j t h e American Cheniicuf Society
/ 95:22 / October 31, 1973
