Organometallics 1983, 2, 906-908
906
Reaction of (CH,),SnNa and Ph,MLi (M = C, Si, Ge, Sn) with 6-Bromo- 1-heptene: Are Intermediate 1-Methyl-5-Hexenyl Radicals Involved?' Kang-Wook Lee and Joseph San Filippo, Jr.' Department of Chemistry, Rutgers University, New Brunswick, New Jersey 08903 Received December 14, 1982
Collective observations lead to the conclusion that the formation of (2-methylcyclopentyl)methyl-derived products cannot be considered as prima facie evidence for the intermediacy of 1-methyl-5-hexenylradicals during the alkylation of (CH&SnNa, Ph,SiLi, and PhBGeLiby 6-bromo-1-heptene. The cyclization of the 5-hexenyl radical, 1, is a welldocumented process that occurs with a rate constant of ca 1.07 X lo5 s-l (at 25 As a consequence of this oC).293
-.
k
1.07 x
:
lo5 s - '
25°C
1
2
behavior, the 5-hexenyl system has become a standard probe, diagnostic of the intermediacy of (kinetically) free alkyl radicals., The related 1-methyl-5-hexenyl system, 3, has been less studied; however, the rate of cyclization of 3 is known to be essentially the same as that observed for 1, i.e., k = 1 X lo5 It follows from a consideration
the reaction of (CH,),SnM with 6-bromo-1-heptene. The former authors reported (for M = Na) a 71% (absolute) yield of the cyclic product 5 and a 4% (absolute) yield of the unrearranged compound 6 while the latter authors reported this same reaction (for M = Li) affords a 48% (absolute) yield of 5 and a 12% (absolute) yield of 6. On this basis both groups concluded that alkylation proceeds through the general intermediacy of 1-methyl-5-hexenyl radicals. (CH313SnNa
t d/
B
r
2 0% °C
d
S
"
(
C
-
Results and Discussion In this vein, the recent observations and proposals of Ashby and DePrieste and the related findings of Kitching and co-workers' are noteworthy. Both groups examined (1) Supported by the National Science Foundation, Grant 80-17045, and DOE, Contract DE-AS05-80-ER-1062. (2) Schmid, P.; Griller, D.; Ingold, K. U.Int. J. Chem. Kinetc. 1979, 11, 333. Lal, D.; Griller, S.;Ingold, K. U.Ibid. 1974, 96,6355. (3) (a) Garst, J. F.; Smith, C. D. J. Am. Chem. SOC. 1976,98,1520 and references therein. (b) Wilt, J. W. In "Free Radicals";Kochi, J. K., ed.; Wiley-Interscience: New York, 1973; Vol. 1, Chapter 8. (c) Noyes, R. M. h o g . React. Kinet. 1961,1, 129. (4) Beckwith, A. L. J.; Blair, I.; Phillipou, G. J. Am. Chem. SOC. 1974, 96,1613 and references therein. (5) Greene, F. D.; Berwick, M. A.; Stowell, J. C. J . Am. Chem. SOC. 1970,92,867. Kopecky, K. R.; Gillan, T. Can. J . Chem. 1969,47,2371. (6) Ashby, E. C.; DePriest, R. J . Am. Chem. SOC.1982, 104, 6144.
3
+
S n(CH3)3
i;5
of the magnitude of these rate constants that the rearrangement of 1 and 3 is Z104 times too slow to compete with geminate coupling processes which are generally recognized to proceed with effective rate constants kdd, 2 1O'O s-l.% Indeed, the virtue of both the 5-hexenyl and the 1-methyl-Bhexenylsystems as diagnostic probes in the elucidation of reaction mechanism rests squarely on the understanding that their characteristic rearrangement occurs during the time that the parent radical is kinetically free. By comparison, the loss of configuration associated with alkyl radical center occurs primarily within the parent solvent cage and is complete for kinetically free radicals.s It follows, therefore, that under equivalent circumstances, the extent to which the rearrangement of 3 4 takes place, can (at most) equal but never exceed the extent of racemization associated with the equivalent reaction involving a radical generated from a chiral precursor.
W
6
(3)
Ashby and DePriest also examined the corresponding reaction of (CH,),SnNa with optically active 2-bromooctane and observed, as previously noted by other investigators18that this alkylation proceeds with overall net (-50% in their hands) inversion of configuration a t carbon. However, in contrast to earlier studies8 which concluded that the stereoselectivity associated with this process results from a partitioning between S N 2 and a oneelectron-transfer process (the latter process resulting in the production of free 2-octyl radicals that in turn yield racemic product), these authors proposed that inversion of configuration results from an electron-transfer process followed by a carbon-tin bond-forming step involving a stereospecificreaction between (CH3),Sn. and the resulting 2-octyl radical, all within the same solvent cage. These conclusions pose a dilemma. Specifically,they require that a significant fraction (233%) of the production of 5 must occur in the same solvent cage in which the electrontransfer step takes place.g However, since 1-methyl-5hexenyl radicals do not have time to cyclize under such conditions, cyclization under these circumstances cannot be invoked as prima facie evidence for the intermediacy of 1-methyl-5-hexenyl radicals. Further evidence that 5 does not constitute prima facie evidence for the intermediacy of 1-methyl-5-hexenylradicals in eq 3 is recognized in the caveats posed by two (7) Kitching, W.; Olszowy, H. A.; Harvey, K. J . Org. Chem. 1982,47, 1893. (8) San Filippo, J., Jr.; Silbermann, 3. J.Am. Chem. SOC.1981, 103, 5588; 1982,104, 2831. (9) The normalized yields of 5 and 6 (Table I) are, respectively, 88% and 12%. Since -50% of the total yield of coupling product (Le., 5 + 6) must occur (according to ref 6) within the same solvent cage as the rate-limiting electron-transfer step, it follows (assuming all or less than all of 6 is also derived in this initial solvent cage) that a minimum of (50-12)/88 or 33% of 5 must also be formed in this initial solvent cage.
0276-7333/83/2302-0906$01.50/0 0 1983 American Chemical Society
Reaction of (CH3)&Na and Ph,MLi with 1-Heptene Table I.
Organometallics, Vol. 2, No. 7, 1983 907
Reactions of 6-Halohept-1-ene with (CH,),SnNa and Ph,MLi (M = C, Si, Ge, S n ) in THF
yield,a % SnlCH3)3
entry
- 1 2 3 4 5 6 7 a
M
X
Na
c1
Na Na Na Na Na Li
Br Br Br Br Br Br
additive
DCPH
Absolute yields; relative yields are in parentheses.
order o f additn
d
S
n
(
C
H31
55 (62) 4 (12) 11 (13)
inverse normal inverse inverse normal inverse normal
33 71 72 14
1(7)
(61 (32) 12 (21)
-
(10) Newcomb, M.; Courtney, A. R. J. Org. Chem. 1980, 45, 1707. (11) Bock,P. L.; Whitasides, G. M. J.Am. Chem. SOC.1974,96,2826. (12) San Filippo, J., Jr.; Silbermann,J.; Fagan, P. J. J.Am. Chem. SOC. 1978,100,4034(13) Kuivila, H. G.; Alnajjar, M. S. J.Am. Chem. SOC.1982,104,6146. (14) (a) Mishra. S.P.: Neilson. G. W.: Svmons. M. C. R. J.Chem. SOC.. Faraday, 1974, 70, 1280. Symons, M. C. R . J. &em. SOC. Chem. Corn: mun. 1977,408. Hasegawa, A.; Williams, F. Chem. Phys. Lett. 1977,45, 275. Riederer, H.; Hottermann, J.; Symons, M. C. R. J. Chem. SOC., Chem. Commun. 1978, 313. (b) However, see: Garst, J. F.; Roberta, R. D.; Pacifci, J. A. J.Am. Chem. SOC.1977,99,3528 and references therein.
ref
1.5 3.7 3.2 3.7
6 6 6 6
(38) (88)
(87) (93) (94)
C
(68)
C
48 ( 7 9 )
2.8
7
This work.
Dicyclohexylphosphine.
additional observations: (i) the finding that the closely related rearrangement of 1 2 does not occur during the corresponding reaction of 6-bromohex-5-ene7J0 (more sensitive probes"J2 suggest that ca. 20% of the carbon-tin coupling product afforded by the reaction of (CH,),SnM with primary bromides arises via a pathway consistent with a free radical process) and (ii) the finding (entries 2 and 3, Table 11) that the extent of rearrangement observed during the reaction of both Ph3SiM and Ph3GeM with 6-bromo-1-heptene exceeds the extent of racemization associated with the corresponding reaction between these same reagents and optically active 2-bromooctane. The reaction of (CH,),SnNa with 6-halo-1-heptene exhibits other qualities that are inconsistent with the intermediacy of 1-methyl-5-hexenyl radicals. Thus, prior studies4 have established that the cis-trans ratio of the cyclic product produced from 3 is invariant: cis/trans = 2.3. Moreover, this value is, as it should be, independent of reaction conditions (i.e., solvent, concentration, additives, and leaving group). By contrast, the cis/trans ratios seen in Table I range between 1.5 and 3.7 and, in addition, are substantially influenced by the presence of the additive dicyclohexylphosphineand the nature of the leaving group. Collectively, these findings vitiate the conclusions of Ashby and DePriest6 along with the corollary conclusions of Kitching, Kuivila, and co-workers.'J3 They do not, however, offer a clear explanation to the origin(s) of the cyclic products produced in the reaction of 6-bromo-lheptene with (CH3),SnM, Ph,SiLi, and Ph3GeLi. A number of speculative possibilities suggest themselves, including the likelihood that rearrangement originates through the agency of (i) olefin activation initiated by metal coordination, (ii) the intermediacy of the corresponding radical anion, or (iii) a breakdown in the principles (extrapolated from the 6-halo-1-hexene systems) which provide the basis for the interpretations associated with reactions of the 6-halo-1-hexenesystem. With respect to ii, the existence of such intermediate species has been suggested on the basis of physical however,
cis/trans
Table 11. A Comparison of the Reactions of 2-Bromohept-6-ene and (S)-a-Octyl Bromide with Ph,MLi (M = C, Si, Ge, S n ) d
B
r
+
Ph3MLi
S-Z-CsH17Br
+
15-25 T H F 'C
Ph3MLi
d
M
P
h
3
+
THF
2-CeH17MPh3
yield,a %
1
M C
2
Si
3
Ge
entry
4 5 6
Ge Sn Sn
order of additn normal inverse normal inverse normal inverse normal normal normal
> 99 > 99 11
37 66 84 68 (18) > 99 >99(83)
ref -
