Stereochemistry and mechanism of vinyl-migrating [1,2]-Wittig rearrangement of α-lithioalkyl vinyl ethers

Article abstract of DOI:10.1246/cl.2000.418

Enantiomerically defined α-stannylalkyl or α-methylbenzyl vinyl ethers, when treated with butyllithium, are shown to undergo the 1,2-vinyl migration to afford the allylic alcohols in almost racemic form in low or high yield, respectively, thereby proposin

Full text of DOI:10.1246/cl.2000.418

418
Chemistry Letters 2000
Stereochemistry and Mechanism of Vinyl-migrating [1,2]-Wittig Rearrangement
of α-Lithioalkyl Vinyl Ethers
Katsuhiko Tomooka, Tadashi Inoue, and Takeshi Nakai*
Department of Chemical Technology, Tokyo Institute of Technology, Meguro-ku, Tokyo 152-8552
(Received January 5, 2000; CL-000005)
migration (Scheme 1). Thus, (R)-1 (95% ee)6 was first treated
Enantiomerically defined α-stannylalkyl or α-methylben-
with RhCl(PPh3)3 to afford the vinyl ether 28 in Z-riched form
7
zyl vinyl ethers, when treated with butyllithium, are shown to
undergo the 1,2-vinyl migration to afford the allylic alcohols in
almost racemic form in low or high yield, respectively, thereby
proposing the radical cleavage-recombination pathway.
(Z/E=2/1) without appreciable loss of enantio-purity.9
Interestingly, treatment of 1 with the hydride catalyst generated
in situ from Ir(cod)(PMePh2)210 was found to result in the exclu-
sive formation of (E)-2,8 again, without loss of enantio-purity.9
When (E)-2 was transmetalated with n-BuLi (1.2 equiv) in THF
at –78 ˚C to generate the lithio species (E)-3 with complete
retention of cofiguration11 and then stirred at that temperature
for 0.5 h followed by hydrolysis with D2O, no vinyl-migrated
product (4) was detected, but the deuterated ether (E)-5a was
obtained in 72% yield. However, when the reaction temperature
was raised up to –25 ˚C over a period of 14 h followed by
hydrolysis with D2O, the rearrangement product (E)-412 was
obtained in 19% yield, together with 43% yield of the protonat-
ed ether (E)-5b; surprisingly, none of the deuterated one was
formed.13 Significantly enough, alcohol (E)-412 thus obtained
was found to be almost racemic (9% ee),14 while the E-geometry
was completely retained as previously observed.2,15 A similar
reaction of (Z)-2 (>95% Z) also afforded 16% yield of the
racemic alcohol (Z)-4, again, with complete retention of Z-
geometry, together with 18% recovery of (Z)-5b13 and 16%
yield of byproduct 6.16 Thus, these results reveal that the 1,2-
vinyl migration onto an α-oxyalkyllithium occurs with much
smaller facility, compared with the reported benzylic and allylic
counterparts2 and, more importantly, the present vinyl migration
proceeds in almost non-stereospecific fashion at the Li-bearing
terminus. Obviously, these findings are suggestive of the radical
nature of the 1,2-vinyl migration.
Over decades the 1,2-alkyl migration on α-lithiated ethers,
now called [1,2]-Wittig rearrangement, has attracted consider-
able interest from mechanistic and synthetic points of view.1
This type of 1,2-alkyl migration is now well-recognized to pro-
ceed via the radical cleavage-recombination pathway with
slight inversion of configuration at the Li-bearing terminus (eq
1). Meanwhile, another interesting variant has been reported
which involves a 1,2-vinyl migration on α-vinyloxy benzylic or
allylic organolithiums to yield the allylic alcohols (eq 2),2 while
its scope remains largely unexplored. This type of [1,2]-Wittig
variant is of particular interest, since structurally related α-allyl-
oxy and α-homoallyloxy organolithiums are known to undergo
the concerted [2,3]-sigmatropy rearrangement3 and the car-
bolithiative cyclization,4 respectively. However, the mecha-
nism of the 1,2-vinyl migration remains totally unsolved. The
key question is whether the 1,2-vinyl migration proceeds via a
radical mechanism similar to that proposed for the usual 1,2-
alkyl migration or via an intramolecular carbolithiation mecha-
nism similar to that recently proposed for the 1,2-carbamoyl
migration (eq 3).5 To answer this question, we investigated the
steric course of the vinyl-migrating [1,2]-Wittig variant using
enantiomerically-defined substrates, with the reasonable predic-
tion that the carbolithiation pathway should result in almost
complete retention of configuration at the Li-bearing stereocen-
ter. Disclosed herein are the stereochemical outcomes of the
1,2-vinyl migrations of the enantio-defined α-stannylalkyl and
α-methylbenzyl vinyl ethers, which provide solid evidence in
support of the radical cleavage-recombination mechanism.
In order to gain more evidence supporting the radical
mechanism, we next carried out the rearrangement of enantio-
defined benzylic vinyl ethers. Thus, the (R)-α-deuterobenzyl
vinyl ether 8, prepared as an E-riched mixture (E/Z=12/1)8 in
89% yield from the enatio-enriched allyl ether 7 (96% ee) by
using the aforementioned iridium catalyst, was treated with n-
BuLi (1.2 equiv.) in THF-TMEDA (20% vol) at –78 ˚C to
At the outset, our attention was focused on the rearrange-
ment of the enantiomerically defined α-stannylalkyl vinyl ether
(E)- or (Z)-2 prepared from the allyl ether (R)-1 via olefin
Copyright © 2000 The Chemical Society of Japan

Chemistry Letters 2000
419
afford, after hydrolysis, the vinyl-migration product 9 in 59%
yield (eq 4). Alcohol 9 thus obtained was essentially racemic
as revealed by chiral HPLC assay. However, this outcome,
while consistent with the radical mechanism, does not necessar-
ily serve as solid evidence for the radical mechanism, since α-
oxybenzyllithiums in general are known to be configurationally
labile.17 Thus, we next examined the rearrangement of (R)-α-
methylbenzyl vinyl ether (11) of which the lithium species
should be configurationally more stable.17 Thus, treatment of
(R)-11 (100% ee), prepared from (R)-10 by the literature
method,18 with n-BuLi in THF at –78 ˚C was found to afford
the vinyl-migration product (R)-1219 in 87% yield in slightly
inverted form (22% ee)20 (eq 5).
Promotion of Science. T.I. thanks JSPS for a postdoctoral fel-
lowship.
References and Notes
1
Reviews: a) U. Schöllkopf, Angew. Chem., Int. Ed. Engl., 9, 763
(1970). b) J. A. Marshall, in “Comprehensive Organic Synthesis”, ed
by B. M. Trost and I. Fleming, Pergamon Press, London (1991), Vol.
3, p. 975. c) K. Tomooka and T. Nakai, J. Synth. Org. Chem. Jpn.,
54, 1000 (1996). d) K. Tomooka, H. Yamamoto, and T. Nakai,
Liebigs Ann./Recueil, 1997, 1275.
2
3
V. Rautenstrauch, G. Büchi, and H. Wuest, J. Am. Chem. Soc., 96,
2576 (1974).
Reviews: a) R. W. Hoffmann, Angew. Chem., Int. Ed. Engl., 18, 563
(1979); Angew. Chem., 91, 625 (1979). b) T. Nakai and K. Mikami,
Chem. Rev., 86, 885 (1986). c) R. Brücker “Comprehensive Organic
Synthesis” ed by B. M. Trost and I. Fleming, Pergamon Press,
London (1991), Vol. 6, p. 873. d) T. Nakai and K. Mikami, Org.
React., 46, 105 (1994).
4
a) C. A. Broka, W. J. Lee, and T. Shen, J. Org. Chem., 53, 1336
(1988). b) C. A. Broka and T. Shen, J. Am. Chem. Soc., 111, 2981
(1989). c) K. Tomooka, N. Komine, and T. Nakai, Tetrahedron Lett.,
38, 8939 (1997).
5
6
K. Tomooka, H. Shimizu, T. Inoue, H. Shibata, and T. Nakai, Chem.
Lett., 1999, 759.
K. Tomooka, T. Igarashi, M. Watanabe, and T. Nakai, Tetrahedron
Lett., 33, 5795 (1992). Note that the [2,3]-Wittig rearrangement of
(R)-1 itself has been reported to proceed with complete retention of
configuration at the Li-bearing terminus.
7
8
E. J. Corey and J. W. Suggs, J. Org. Chem., 38, 3224 (1973).
Selected spectral data for (E)-2: colorless oil; [α]D –47.5 (c = 0.56,
27
This result provides solid evidence for the radical mecha-
nism and against the carbolithiation mechanism as well.
Overall, it is safe to conclude that the present 1,2-vinyl migra-
tion proceeds via the radical pathway, wherein the configura-
tional integrity at the Li-bearing terminus is mostly lost, while
the geometry of the migrating vinyl group is completely
retained (Scheme 2).
CHCl3); (300 MHz, CDCl3) δ 7.32-7.15 (m, 5H), 6.14 (dd, J = 12.3,
1.5 Hz, 1H), 4.74 (dq, J = 12.3, 6.9 Hz, 1H), 4.10 (dd, J = 9.9, 4.5 Hz,
1H) 2.78 (m, 1H), 2.63 (m, 1H), 2.20 (m, 1H), 1.98 (m, 1H), 1.56 (dd,
J = 6.9, 1.5 Hz, 3H), 1.48 (m, 6H), 1.30 (m, 6H), 0.92-0.86 (m, 15H);
13C NMR (75 MHz, CDCl3) δ 147.2,142.1, 128.6, 128.3, 125.7, 98.9,
27
75.4, 37.2, 33.9, 29.1, 27.5, 13.7, 12.6, 9.4. (Z)-2: colorless oil; [α]D
+24.3 (c = 0.72, CHCl3); (300 MHz, CDCl3) δ 7.32-7.15 (m, 5H),
5.90 (dq, J = 6.3, 1.5 Hz, 1H), 4.34 (qq, J = 6.6 Hz, 1H), 4.10 (dd, J =
8.7, 4.5 Hz, 1H) 2.80-261 (m, 2H), 2.20 (m, 1H), 2.01 (m, 1H), 1.60
(dd, J = 7.2, 2.1 Hz, 3H), 1.50 (m, 6H), 1.31 (m 6H), 0.95-0.87 (m,
15H); 13C NMR (75 MHz, CDCl3) δ 146.8, 142.1, 128.5, 128.3,
125.7, 100.1, 77.9, 37.4, 33.8, 29.3, 29.2, 27.5, 13.7, 9.3.
Confirmed by hydrolysis of 2 which gave the α-stannyl alcohol in
>95% ee.
9
10 J. J. Oltvoort, C. A. A. van Boeckel, J. H. de Koning, and J. H. van
Boom, Synthesis, 1981, 305.
11 The Sn/Li transmetallation is known to proceed with complete reten-
tion of configuration: J. S. Sawyer, A. Kucerovy, T. L. Macdonald,
and G. J. McGarvey, J. Am. Chem. Soc., 110, 842 (1988), and refer-
ences cited therein.
12 H. J. Reich and S. Wollowitz, J. Am. Chem. Soc., 104, 7051 (1982).
13 The exact mechanism of the protonated byproduct 5b is unclear at
present.
14 The % ee of (E)-4 was determined by the 19F NMR assay of the
MTPA ester. 19F NMR (400 MHz, CDCl3) δ 95.3 (45%), 95. 2 (55%).
15 The complete retention of the olefin geometry is of special interest
because vinylic radicals are known to have low inversion barriers in
general. : D. P. Curran, N. A. Porter, and B. Giese, in
“Stereochemistry of Radical Reactions”, VHC, New York (1995).
16 This byproduct is likely to arise from the dimerization of the car-
benoid species generated from the α-alkoxy organolithium involved.
Also see ref.5.
In summary, we have proved that the vinyl-migrating
[1,2]-Wittig rearrangement on α-lithiated alkyl and benzyl
vinyl ethers, like the well-known alkyl-migrating variant, pro-
ceeds via the radical cleavage-recombination pathway based on
the elucidation of the steric course of the asymmetric versions
using enantio-defined substrates. Furthermore, we have point-
ed out the importance of a radical stabilizing element in the car-
banion part as a key factor to facilitate the 1,2-vinyl migration.
Further works on synthetic applications of the present [1,2]-
Wittig variant as well as the development of the enantioselec-
tive version are in progress.
17 D. Hoppe and T. Hense, Angew. Chem., Int. Ed. Engl., 36, 2282
(1997).
18 H. Lussi, Helv. Chem. Acta, 49, 1681 (1966).
19 D. R. Dimmel and S. Huang, J. Org, Chem., 38, 2756 (1973).
20 The % ee was determined by the HPLC analysis using a Daicel CHI-
RALPAK AD [hexane:i-PrOH = 200:1 v/v; tR = 26.8 min (R) and
29.3 min (S)]. The R configuration was determined based on the opti-
cal rotation after conversion to the known (R)-2-phenyl-2-butanol via
hydrogenation: B. Weber and D. Seebach, Tetrahedron, 50, 6117
(1994).
We thank Professor Hiroharu Suzuki of our department for
a gift of the iridium complex and his helpful suggestions on the
olefin migration process. This work was supported by the
Research for the Future Program, Japan Society of the