Anti-Carbometalation of Homopropargyl Alcohols and Their Higher Homologues via Non-Chelation-Controlled Syn-Carbometalation and Chelation-Controlled Isomerization

Full text of DOI:10.1021/jo9622688

784
J . Org. Chem. 1997, 62, 784-785
the reaction mixture for 72 h, however, a complete
reversal of the stereochemistry from >98% E to >98% Z
took place to produce, after iodinolysis, a 60% yield of
4a (E ) I). Similarly, 1- and 2-methyl-substituted
homopropargyl alcohols 5 and 6 were converted to 7 and
8, both of which were >98% Z, in 61 and 50% yields,
respectively.7 These products appear to represent a class
of compounds not readily accessible by any of the previ-
ously known reactions.8
An ti-Ca r bom eta la tion of Hom op r op a r gyl
Alcoh ols a n d Th eir High er Hom ologu es via
Non -Ch ela tion -Con tr olled
Syn -Ca r bom eta la tion a n d
Ch ela tion -Con tr olled Isom er iza tion †
Shengming Ma and Ei-ichi Negishi*
Department of Chemistry, Purdue University,
West Lafayette, Indiana 47907
Received December 4, 1996
Highly stereoselective anti-carbometalation reactions
of alkynes are rare,1 the great majority of the syntheti-
cally useful carbometalation reactions of alkynes, such
as Zr-catalyzed carboalumination2 and carbocupration,3
being syn-addition processes. One notable example of
stereoselective anti-carbometalation is the Cu-catalyzed
carbomagnesiation reaction of propargylic alcohols.4 Un-
fortunately, the scope of this reaction does not generally
extend to homopropargyl alcohols and higher homo-
logues.
We report herein a novel strategy for achieving net
anti-carbometalation of homopropargyl alcohols and even
some higher homologues. This strategy critically hinges
on our finding that the syn- and stereorandom-carbo-
alumination products obtained from ω-hydroxyalkynes
represented by HORCtCZ, where Z ) H, Si, or Ge, can
be thermally isomerized to give nearly exclusively or
predominantly alkenylalanes that correspond to anti-
carboalumination of alkynes (Scheme 1). The reactions
shown in Scheme 1 are fundamentally different from the
carbotitanation reaction of homopropargyl alcohols and
longer alkynols,5 which displays the opposite regioselec-
tivity of carbometalation, producing 1, and must therefore
be chelation-controlled in the addition step itself. The
reactions presented here are also critically different from
the Cu-catalyzed carbomagnesiation4 of propargylic al-
cohols producing 2 in that these two classes of reactions
display essentially nonoverlapping and hence comple-
mentary scopes.
It is important to note that, in the absence of the
homopropargylic hydroxy group, the E-to-Z isomerization
does not occur. Thus, the corresponding reaction of
1-decyne merely gave (E)-1-iodo-2-methyl-1-decene (>98
E) in about 80% yield, and neither the stereochemistry
nor the yield detectably changed even after 72 h at the
refluxing temperature of 1,2-dichloroethane. In fact, no
detectable isomerization was observed even with 4-pen-
tyn-1-ol. The observed isomerization of 3 must therefore
be chelation-controlled. We tentatively propose a Lewis
acid-induced chelation-controlled mechanism producing
9 as the product as shown in Scheme 2. Although we
have thus far been unsuccessful in obtaining definitive
structural data on 9, the corresponding reaction of the
terminally Me3Ge-substituted derivative 10 was faster
(36-48 h, 25 °C, CD2Cl2) and cleaner, producing a >80%
NMR yield of 11: 1H NMR (500 MHz, CD2Cl2) δ -0.58
(bs, 3 H), 0.22 (s, 9 H), 1.84 (s, 3 H), 2.0-2.5 (m, 2 H),
3.9-4.25 (m, 2 H); 13C NMR (50 MHz, CD2Cl2) δ -6.50,
1.15, 28.29, 41.02, 64.63, 110.31, 110.62. Hydrolysis with
aqueous Na2CO3 and iodinolysis gave 12a (E ) H, >98%
E) and 12b (E ) I, >95% Z) in 77 and 73% isolated yields,
respectively. Furthermore, treatment of 11 with
ClCOOMe gave 13, albeit 12% yield.
Substitution of the terminal alkynyl hydrogen atom
with a metal-containing group, such as Si and Ge, not
only accelerates stereoisomerization9 but also expands
the scope of reaction. As summarized in Table 1, the
reaction of a series of ω-(trimethylsilyl)alkynols with
Me3Al (3 equiv) and Cp2ZrCl2 (1 equiv) at the refluxing
temperature of CH2Cl2 gave stereoselectively the anti-
methylalumination products. The anti/syn ratios were
>98/<2 for 3-butyn-1-ol derivatives, g97/e3 for 4-pentyn-
1-ols, and 88/12 for 5-hexyn-1-ols. However, those for the
6-heptyn-1-ol and 10-undecyn-1-ol derivatives were
roughly in the 40/60-60/40 range and were very similar
to that observed with 1-(trimethylsilyl)-1-octyne.10 These
latter results reinforce our view that the high anti/syn
ratios observed with the C4 and C5 ω-alkynols must be
chelation-controlled, involving the formation of six- and
seven-membered aluminacycles, respectively. The anti/
syn ratio of 88/12 observed with 6-(trimethylsilyl)-5-
hexyn-1-ol is very intriguing. It suggests that the
reaction may be largely chelation-controlled. If so, it
Specifically, treatment of 3-butyn-1-ol with Me3Al (3
equiv) and 25 mol % of Cp2ZrCl2 in (CH2Cl)2 at 23 °C
produced the expected syn-methylalumination product 3
(>98% E) as previously reported by us.6 Upon refluxing
† We wish to dedicate this paper to Professor D. Seebach of ETH,
Zu¨rich, on the occasion of his 60th birthday.
(1) For a review, see: Normant, J . F.; Alexakis, A. Synthesis 1981,
841.
(2) (a) Van Horn, D. E.; Negishi, E. J . Am. Chem. Soc. 1978, 100,
2252. (b) Negishi, E.; Van Horn, D. E.; Yoshida, T. J . Am. Chem. Soc.
1985, 107, 6639. (c) Negishi, E. Pure Appl. Chem. 1981, 53, 2333.
(3) (a) Normant, J . F.; Bourgain, M. Tetrahedron Lett. 1971, 2583.
(b) Normant, J . F. J . Organomet. Chem. Lib. 1976, 1, 219.
(4) (a) J ousseaume, B.; Duboudin, J . G. J . Organomet. Chem. 1975,
91, C1. (b) Duboudin, J . G.; J ousseaume, B. J . Organomet. Chem. 1979,
168, 1.
(7) Conversion of 6 to 8 was performed by Ms. F. Liu in our
laboratories.
(5) (a) Coleman, R. A.; O’Doherty, C. M.; Tweedy, H. E.; Harris, T.
V.; Thompson, D. W. J . Organomet. Chem. 1976, 107, C15. (b) Smedley,
L. C.; Tweedy, H. E.; Coleman, R. A.; Thompson, D. W. J . Org. Chem.
1977, 42, 4147. (c) Brown, D. C.; Nichols, S. A.; Gilpin, A. B.; Thompson,
D. W. J . Org. Chem. 1979, 44, 3457.
(6) Rand, C. L.; Van Horn, D. E.; Moore, M. W.; Negishi, E. J . Org.
Chem. 1981, 46, 4093.
(8) See, however: Kocienski, P.; Wadman, S.; Cooper, K. J . Am.
Chem. Soc. 1989, 111, 2363.
(9) Facile stereoisomerization of Si-substituted alkenylalanes and
alkenyllithiums has been previously reported and discussed. (a) Miller,
J . A.; Negishi, E. Israel J . Chem. 1984, 24, 76. (b) Negishi, E.;
Takahashi, T. J . Am. Chem. Soc. 1986, 108, 3402.
(10) Snider, B. B.; Karras, M. J . Organomet. Chem. 1979, 179, C37.
S0022-3263(96)02268-2 CCC: $14.00 © 1997 American Chemical Society

Communications
J . Org. Chem., Vol. 62, No. 4, 1997 785
Sch em e 1
Sch em e 2
Sch em e 3
Ta ble 1. An ti-Ca r boa lu m in a tion of Hom op r op a r gyl
Alcoh ols a n d Th eir High er Hom ologu esa
tion-initiated and that the latter benefit must be prac-
tically limited only to the cases of 3-butyn-1-ols. On the
other hand, isomerization of the Si-substituted deriva-
tives must kinetically rely on their intrinsic configura-
tional instability,10 and chelation is significant only in a
thermodynamic sense. Regardless of their mechanism,
these favorable isomerization reactions observed even
with 4-pentyn-1-ol and 5-hexyn-1-ol derivatives were
quite unexpected but of considerable synthetic potential.
For example, treatment of 14e (Z/E ) 88/12) with NaOMe
in MeOH at 65 °C for 12 h gave an 85% yield of the
desilylated product, i.e., 6-iodo-5-methyl-5-hexen-1-ol
(15), which was also 88% Z.
The conversion of 4a into (3Z)-R-farnesene11 (16) as
summarized in Scheme 3 provides an example of natural
products synthesis employing the newly developed strat-
egy.
a
Unless otherwise stated, the reaction was carried out in
CH2Cl2 by using 1 equiv of Cp2ZrCl2. b Isolated yield with the NMR
yield in parentheses. c The thermally equilibrated ratio of the anti-
Ack n ow led gm en t. We thank the National Science
Foundation (CHE-9402288) for support of this research.
We also thank Ms. F. Liu for providing some experi-
mental data.
d
carboalumination to syn-carboalumination products. The reac-
tion was carried out in (CH2Cl)2 using 25% of Cp2ZrCl2. e 5 )
4-pentyn-2-ol. f 6 ) 2-methyl-3-butyn-1-ol. 19 ) 5-(trimethylsi-
g
h
lyl)-4-pentyn-2-ol. Not determined.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for the preparation of 4a , 13, 14a , 14b, 14e, and 16 as
well as spectral data of 7, 8, 12a , 12b, 13, 14d , 14f, 14g, 14h ,
and 15 (7 pages).
must involve the preferential formation of an eight-
membered aluminacycle. The differences in the ease and
scope between stereoisomerization of terminally unsub-
stituted ω-alkynols and that of ω-silyl-substituted ana-
logues are striking. One plausible explanation for the
observed differences might be that isomerization of
unsubstituted ω-alkynols must be not only thermody-
namically driven by chelation but also kinetically chela-
J O9622688
(11) For a recent synthesis of (3Z,6E)-R-farnesene, which was 80%
stereoselective, see: Ramaiah, P.; Pegram, J . J .; Millar, J . G. J . Org.
Chem. 1995, 60, 6211.