1016
COMAIUSICATIOSS TO
major product. Our work is continuing in order to define the scope of this reaction, and to assess the implications for the mechanisms of other radiation-induced reactions in organic media. Acknowledgment.-We thank Professors 11:. H. Fletcher and A. D. Melaven for the use of spectroscopic apparatus. DEPARTMENT O F CHEMISTRY DOSALDH. MARTIS USIVERSITY OF TESNESSEE FFRASCOS ~VILLIAMS KNOXVILLE, TENSESSEE RECEIVED JASUARY 11, 1963
DE CAPHENYLCY CLOPENTA SILANE
Sir: The reaction of dichlorodiphenylsilane with or lithiumyb produces several perphenylated cyclosilanes. One of them, designated Compound “B” by Kipping,la was first proposed to be octaphenylcyclotetrasilane ( 1 ) l a s b . and, more recently, to be dodecaphenylcyclohesasilane (11). Both of these proposals were based on molecular weight de terminations; the SiPhz SiPhz PhzSi PhzSi’ ‘SiPhz PhzSi’ ‘SiPhz
I
PhzSi-
I
I
Sipha
I
I
I
PhzSi,
.,SiPhz PhzSiSiPhz SlPhz I11 I I1 former on values determined cryoscopically in benzene and camphol-Iasband the latter on values determined cryoscopically in pery1ene.j It now has been shown that both proposals were in error and that Compound “B” is decaphenylcyclopentasilane (111). Derivatives of this compound, which had previously been designated as h e x a s i l a n e ~ , ~are in fact l,5-disubstituteddecaphenylpentasilanes. Chemical proof for the structure of decaphenylcyclopentasilane has been obtained by two different means starting with decaphenylcyclopentasilane. In a recent p ~ b l i c a t i o n i, t~ has been shown t h a t the a,w-dihydroxy derivative of Compound “B” (decaphenylcyclopentasilane), when chromatographed on basic alumina, provides 1,112,2,3,3-hexaphenyltrisilane. This reaction indicates that the dihydroxy compound was a pentasilane and t h a t Compound “B” is the cyclopentasilane. Compound “B”
ILiI __3
THE
EDITOR
Vol. 85
plete. Formation of derivatives, using trimethyl or tri-n-butyl phosphate6 or chlorotrimethylsilane, led to derivatives of the corresponding 1,5-disubstituted compounds, in yields approaching 83yC, The constitutions of these compounds were established by elemental analysis and by proton magnetic resonance spectra determinations to obtain aliphatic t o aromatic hydrogen atom ratios. 1,5-Dimethyldecaphenylpentasilanewas also obtained in high yield from the reaction of methylmagnesium iodide with 1,5dibromodecaphenylpentasilane. Molecular weight determinations have been obtained for decaphenylcyclopentasilane (molecular weight 91 1) by three different methods which gave consistent results. Ebullioscopic determinations in toluene‘ using octaphenylcyclotetrasilane as a standard gave a value of 921 + i(four determinations). =In X-ray diffraction study and density determination of decaphenylcyclopentasilane, crystallized from benzene-ethanol, provided a value of 983, in good agreement with a value of 989 calculated for decaphenylcyclopentasilane plus one molecule of benzene, which was subsequently shown to be present. The molecular weight was also determined in benzene using a “vapor phase a t several concentrations. Extrapolation to zero concentration gave a value of 912, in excellent agreement with the theoretical value of 91 1 for decaphenylcyclopentasilane. Experimental details for the reactions and data presented here and for other polysilanes derived from decaphenylcyclopentasilane will be forthcoming. Acknowledgments.-The authors wish to express their gratitude to Drs. C . .A. Glover, H . II-.Patton and K-. D. Kennedy, of the Tennessee Eastnian Co., for the ebullioscopic and X-ray determinations. This research was supported in part by the United States Air Force under Contract -IF33(616)-G463 monitored by Materials Laboratory, Directorate of Laboratories, U’right *Iir Development CeRter, IVright-Patterson * I F R , Ohio. (0) H.Gilman and B. J. G a j , ibid., a6, 447 f l Y i i l j . (7) C A . Glover a n d R. R. Stanley, A d . Chern., 33, 447 (1901) ( 8 ) .\Iechrolab, Inc., Mountain Vier-, Calif.
CHEMISTRY DEPARTMEST H E Y R YGILMAN GER.4LD L. SCHIVEBKE IOWASTATEUKIVERSITY AMES, I O W A RECEIVED JASUARY 28, 1963
Li(SiPh2)sLi
THE REDUCTIONS OF cis- AND tuens--Co(NHa),(N,j2’ AND C O ~ N H ~ ) ~ BY N ~Fe++ *Br( SiP1if)jBr
[CH&lgI]
---f
CHa(SiPh2)aCHs
[alumina I -
H(SiPh2)rH
[HDI
HO( SiPhr)$OH
The lithium cleavage products of Compound “B” were examined to throw more light on the structure of the cyclic polysilane. I t was found t h a t when the cyclosilane was allowed to react with lithium in tetrahydrofuran for 2 hr., the cleavage was essentially com( 1 ) ( a ) 1‘. S. Kipping a n d J , E . Sands, J . C h e m . Soc., 119, 830 (1921); ( b ) F. S. K i p p i n g , ibid., 123, 2.590 (1923); (c) 126, 2291 (1924); (d) 2719 (1‘327). (2) (a) H. Gilman, I ) . J Peterson, .4. W. Jarvie a n d H. J . S. Winkler, J , A m . C h c m Soc., 82, 2076 (1900); (h) A. W. P. Jarvie, H. J . S. Winkler, I ) . J. Peterson a n d €I. Gilman, ibid.,83, 1921 (1961). ( 3 ) (a) 13. Gilman. L). J. Peterson, A . XV. Jarvie a n d H. J . S. Winkler, Z’eiinhedion L e f l e v s , 23, 5 (1960); ( b j €1. J. S. Winkler, A . \V. P. Jarvie, U . J. Peterson a n d H. Gilman, J . A n i . C h e m . Soc., 85, 4089 (1961). (4) (a) A . W. P. Jarvie and H. Gilman, C h e m Ind. (London), 1271 (1960); (h) A . \V. P.Jarvie, H. J . S. Winkler a n d H.Gilman, J . Ovg. C h e m . , 27, 614 ( i M 2 ) ; (c) H.J. S . IVinkler a n d H. Gilman, ibid., 27, 254 (1902). Lj) G . R . Chainani, S. Cooper and H. Gilman, ibid., in press.
Sir : We have measured spectrophotometrically the rates ~)~+ of reduction of cis- and t r a n ~ - C o ( K ; H ~ ) ~ ( Xand co(h”3)&3++ by Fe++ in aqueous perchlorate solutions (no other anions present) a t 25’. The results, summarized in Table I , show t h a t the reduction of the cis complex is acid-independent for (H’) varying from 0.072 to 0.219 M, whereas that of the trans complex is, under the same conditions, strongly acid-dependent. The rate laws (time in minutes) are given by the equations 11.1(Fe++) ( C ~ ~ - C O ( S H ~ ) ~ and ( X ~ )[4.4 ~+) 82 ( H + ) ] ( F e + + ) (truns-C0(PI;H~)~(~\1’3)2+). The reduction of C O ( N H ~ ) ~ Nis~ +acid-independent + and obeys the rate law 0.52(Fe++)(Co(NH3)5N3++). Because of the substitution-lability of F e ( I I I ) , i t cannot be ascertained whether these reactions proceed via a bridged or outer-sphere activated complex.’ However, assuming that bridged activated complexes are operative, reasonable explanations can be advanced for the present observations.
+
(1) H T a u h e , Adz’an. I ? m i p . Chem. Radiochem., I , 1 (1959)
COMMUNICATIONS TO THE EDITOR
April 5 , 19G3
TABLE I THEKEAClIONS OF Cis- AND trans-Coi S H d , ( SJ)?' and C O ( N H ~ ) & ~ ' VITH + Fe*? AT 25", B (Clod-) = 0.26 (Co(SHa)i-
kz,' .Lf-i min.-I
(s3)Z-l
Complex
X 108, .\I ( F e + * ) ,.lf
11.8 1.8 1.8 7.41 1.8 1.8 3.87 3.87
cis CIS
ClS
0.0186 .I1186 0186 ,0185 0093 ,0370 ,0186 0185 0185 . 0 185 009:3 .03;0 ,148
( H + ) , JI
0.072 131 189 2 19 183
A",', min:'
0.203 200
11.0 10.8 11.2 11.3 11.6 10.8 19Ud . 281d . 36gd
20;
210 108 CiS 144 40 1 trans 072 0 190,O 193 trans 0 280 131 3.87 380 trans 189 .114d trans 3.87 2 19 (1 116,O411 20id trans 3.8i 0 206 219 2.87 ~ ) trans-Co(NH3)4(N3)2+ ~+ >> C O ( N H ~ ) ~ N ~cannot ++, be explained invoking solely the effect of ligands trans to the bridging group2: according to the trans effect, one would expect cisCo("3)4(N3)2+ to react a t approximately the same rate as C O ( N H ~ ) ~ N ~ Alternately, ++.~ it could be assumed that replacement of one NH3 in Co(NH3)bN3++ by K3-, whether in the cis or trans positions, results in an increased rate of reaction with Fe++, perhaps by stabilization of the Co orbital which accepts the e l e c t r ~ n . ~However, ,~ in view of the efficiency of the acid-catalyzed path for the reaction of trans-Co(NH3)4(N3)2+with Fe++, it becomes difficult to understand why acid-catalysis is not observed for the cis isomer. We suggest, therefore, that the reaction of cis-Co(NH3)4(N3)2+with Fe++ proceeds zlia a double-bridged activated complex entirely analogous to the one recently demonstrated for the reaction between Cr(N&+ and Cr + + . j This interpretation receives further support when the ratio of the rate constants for the reactions of C ~ ~ - C O ( N H ~ ) ~and ( N Co(NH3)5N3++ ~)~+ with F e + + is compared with the corresponding ratio for the reactions of ~ i s - C r ( X ~ )and ~ + CrN3++ with Cr++. The observed ratio for the Co(II1) complexes is approximately 20, and, when corrected for the difference in ionic strengths,6this ratio will be close to the value of 50 observed5for the chromium system. The increased reactivity of trans-Co(YH3)4(N3)2+ as conipared to Co(hTH3)jhT3++ is consistent with a trans-effect in electron-transfer reactions, * and the observed acid-catalysis further supports this interpretation : the N3- trans to the bridging N3- has a site available for a proton attachment, and movement of the ( 2 ) I,. Orgel, Report of t h e Tenth Solvay Conference, Brussels, 195fi. p 289. (3) I t is n o t likely t h a t t h e increased reactivity of t h e cis complex as compared t o Co(SHa)sSz" i s a consequence of t h e decrease in charge. T h u s , a t a faster rate t h a n cis- or trans-CrFz (Y. T . Chia C r F - - reacts with Cr and E. L. King, Discu:sio?rs Faraday Soc., 29, 109 (1960)). Furthermore, an equilibrium mixture of cis- and Irans-Co(SHa)aOH? N 3 + + reacts with F e + +in 1.2 M HCIOI 70 times faster t h a n C O ( N H J ) ~ S( ~A .+ Hairn, ~ unpublished results). See also ref. 4 , ( 4 ) K . D. K o p p l e a n d R . R . Miller, P r o c . Chern. Soc., 306 (1962). (,j)R . Snellgroveand E. I,. King, J . A m . Chem. S o c . , 84, 41310(1962). (fi) 'The reactions of cis-Co(SHa)r(Na)z and Co(l\'Ha)aN\'a++were studied a t 2 (Clod-) values of 0.26 and 0.89. respectively + +
1017
trans N3- away from the Co center as electron transfer occurs' is facilitated by protonation.' Acid-catalyzed paths for electron-transfer reactions between metal ions, although not unprecedented, are not a common occurrence. Such paths have been observed previously in systems where addition of a proton improves the conjugation of two metal centers connected by an unsaturated bridging ligand,' and for the reactions of carboxylatotetraamminecobalt(111) complexes with CrA+.? (7) Facilitation of t h e removal of K3- b y protonation has been observed in o t h e r systems [A. Haim a n d W K . Wilmarth, litovg, Chein., 1 , 583 (19F2)) a n d is f u r t n e r supported by unpublished experiments showing t h a t t h e ) ; ( S ~ ) ~0.2 - a n d 1 . 2 41 H - are aquations of cis- a n d I ~ . Q ~ z ~ - C O ( N H ~between first order in complex and first order in H -, with undetectable contributions of H - independent paths.
DEPARTMEST OF CHEMISTRY THEPESNSYLVATIA STATE USIVERSITY UNIVERSITY PARK,PEXSSYLVASIA RECEIVED FEBRUARY 9, 1963
ALBERT HAIM
ON THE MECHANISM OF PHOTOCHEMICAL FORMATION OF CYCLOBUTANOLS
Sir: Aliphatic ketones containing r-hydrogen upon irradiation with ultraviolet light in solution undergoes two concurrent reactions, the type I1 cleavage1 and the formation of cyclobutanols.* Both reactions appear to be intramolecular with little or no detectable side react i o n ~ . In ~ the cases of 2-octanone and 20-ketosteroids, two isomeric cyclobutanols'are ~ b t a i n e d . ~ Two mechanisms have been proposed for the photochemical formation of cyclobutanols, one a step-wise mechanism2 (eq. 2) and the other a concerted mechanism5 (eq. 3 ) . On the basis of available experimental data, these two mechanisms cannot be differentiated.
The irradiation of 6-hepten-2-one [I ] was undertaken in order to provide direct evidence as to the mechanism of the cyclobutanol formation. Should the reaction proceed by a concerted mechanism, the products would contain only cyclobutanols. However, if the intermediate is a free radical, an allyl radical [11] in this nlJ
.A
.-
+
(1) R. G. W . Xorrish, T r a n s . F a r a d a y SOC.,33, 1321 11937). (2) N. C . Yang a n d D . H . Yang, J . A m . C h e m . .Soc., 80, 2913 (1958). (3) W. Uavis, J r . , and W. A . S o y e s , J r . , ibid., 69, 2153 (1947); P. Ausloos and R. E. Rebbert, ibid., 83, 1897 (1961). (4) N. C . Yang and I) H . Yang. Telrahedioii Letters, 4, 10 (1960); 0 . Jeger, e l al., H e h . Chim. A d a , 43, 3 5 %(1960). ( 5 ) 0. Jeger, el al., ibicl., 42, 2122 (1939).
