New imino-wittig rearrangement of benzyl and allyl hydroximates

Full text of DOI:10.1021/jo961768y

9078
J . Org. Chem. 1996, 61, 9078-9079
Ta ble 1. 1,2-Wittig Rea r r a n gem en t of (Z)-Hyd r oxim a tes
3
New Im in o-Wittig Rea r r a n gem en t of
Ben zyl a n d Allyl Hyd r oxim a tes
Okiko Miyata, Tomoko Koizumi,
Ichiya Ninomiya, and Takeaki Naito*
Kobe Pharmaceutical University,
Motoyamakita, Higashinada, Kobe 658, J apan
yield
(%)
entry substrate
R1
R2
T (°C)
-23
Received September 17, 1996
1
2
3
4
5
3a
3b
3b
3c
3d
3e
3f
Ph
Ph
Ph
Ph
p-tolyl
styryl
Et
Ph
89
45
We now report the first example of an imino-Wittig
rearrangement of benzyl and allyl hydroximates by
treatment of benzyl and allyl hydroximates 1 with a base,
where the N-alkoxy imino group migrates from an oxygen
to a carbon atom (eq 1). The 1,2-rearrangement of
CHdCH2 -23
CHdCH2 -40
COOMe -50 f 0
60 (82)a
b
76
79
64
Ph
Ph
Ph
Ph
-23
-23
-23
-23
6
7
8
9
10
3g
3h
3i
PhCH2CH2
PhCH2CH2
CH2dCH(CH2)3 Ph
75 (89)a
64
CHdCH2 -40
-23
70
a
b
Based on recovery of the starting material. Complex mixture
observed on TLC.
R-lithio ethers to the corresponding alcohols is the Wittig
rearrangement (eq 2),1 where R1 is generally an alkyl
group such as a benzyl or tert-alkyl.2 However, migration
Ta ble 2. 1,2-Wittig Rea r r a n gem en t of
(E)-Hyd r oxim a tes 5
of an imino group (R1: RCdNR) to a negatively charged
carbon has not been reported so far. We have now found
that 1,2-Wittig rearrangement of benzyl and allyl (Z)-
hydroximates 3 proceeds smoothly to give 2-hydroxy
oxime ethers 4 (eq 3) and provides a new entry to
carbon-carbon bond formation.
ratio
T (°C) yield (%) (E)-6:(Z)-4
entry substrate
R1
Ph
Ph
styryl Ph
R2
1
2
3
5a
5b
5e
Ph
CHdCH2 -40 19 (34)a
-23 58 (71)a
-23 41 (51)a
3:1
1.7:1
1:1
a
Based on recovery of the starting material.
We first investigated the rearrangement of hydroxi-
mates 33 having the Z-geometry at the CdN bond. The
results of those studies are summarized in Table 1.
Treatment of benzyl (Z)-benzohydroximate (3a ) with 2
equiv of LDA in THF at -23 °C afforded the 2-hydroxy
oxime ether having the Z-geometry 4a 4 as the sole
product in 89% yield (Table 1, entry 1). This result shows
that the rearrangement occurs with retention of config-
uration at the N-methoxy imino group. In the rear-
rangement of allyl (Z)-hydroximate 3b, lowering the
temperature (-40 °C) improved the yield from 45% to
60% (Table 1, entries 2 and 3). When R2 is a methoxy-
carbonyl group, many spots are observed on TLC due to
the formation of a complex reaction mixture (Table 1,
entry 4). The rearrangements of p-toluo- and cinnamo-
hydroximates 3d and 3e proceeded smoothly to give the
rearranged products 4d 4 and 4e4, respectively (Table 1,
entries 5 and 6).
In order to compare the acidity of the two methylene
groups at the R- and R′-positions (eq 1), we next inves-
tigated the reaction of the hydroximates 3f-i having
active methylene groups at the R-position of the N-
methoxyimino group with LDA. Although the presence
of two active methylene groups in these substrates was
expected to complicate the reaction, the rearrangements
of the hydroximates 3f, 3g, 3h , and 3i proceeded cleanly
to give the rearranged products 4f-i4 as the sole isolated
product under similar conditions (Table 1, entries 7-10).
This rearrangement was then applied to the (E)-
hydroximate 5a 3 under the same reaction conditions (eq
4, Table 2). However, the yield decreased markedly,
giving a 3:1 mixture of (E)-6a 4 and (Z)-4a with recovery
of starting material (Table 2, entry 1). Since the equili-
bration between (E)-6a 4 and (Z)-4a was not observed
under the reaction conditions, (E)-6a and (Z)-4a would
be the kinetic products. Similarly, the rearrangements
of (E)-hydroximates 5b,e3 gave a mixture of (E)-6b,e4 and
(Z)-4b,e4 (Table 2, entries 2 and 3).
(1) Reviews on the 1,2-Wittig rearrangement: (a) Scho¨llkopf, U.
Angew. Chem., Int. Ed. Engl. 1970, 9, 763. (b) Marshall, J . A. In
Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon: Oxford, 1991; Vol. 3, pp 975-1014.
(2) A few papers have been published on the migration of alkenyl,
carbonyl, and thiocarbonyl groups. (a) Rautenstrauch, V.; Bu¨chi, G.;
Wu¨est, H. J . Am. Chem. Soc. 1974, 96, 2576-2580. (b) Hayashi, T.;
Baba, H. J . Am. Chem. Soc. 1975, 97, 1608-1609. (c) Crooks, P. A.;
Galt, R. H. B.; Matusiak, Z. S. Chem. Ind. 1976, 693-694. (d) Lee, S.
D.; Chan, T. H.; Kwon, K. S. Tetrahedron Lett. 1984, 25, 3399-3402.
(3) The substrates 3a -i and 5a ,b,e were prepared according to the
reported procedure. J ohnson, J . E.; Springfield, J . R.; Hwang, J . S.;
Hayes, L. J .; Cunningham, W. C.; McClaugherty, D. L. J . Org. Chem.
1971, 36, 284-294.
(4) 4a and 6a were identical with the authentic samples5a that were
prepared by treatment of benzoin with methoxyamine hydrochloride.
The stereochemistries of other rearranged products 4b,d -i and 6b,e
were deduced by comparison of the chemical shifts of the R′-methine
protons in 1H-NMR spectra according to ref 5b, which states that
signals due to R′-methine protons of cis-O-methyloxime appear at
appreciably lower fields than those of trans-isomers.
(5) (a) Creary, X.; Wang, Y.-X.; J iang, Z. J . Am. Chem. Soc. 1995,
117, 3044-3053. (b) Karabatsos, G. J .; Hsi, N. Tetrahedron 1967, 23,
1079-1095.
In order to clarify the reaction pathway, we investi-
gated the cross reaction. The fact that a mixture of 3d
and 3b that was treated with LDA gave only a mixture
of two products 4d and 4b suggests that the newly found
rearrangement of hydroximates proceeds via an intramo-
lecular process. From the above results, we propose two
S0022-3263(96)01768-9 CCC: $12.00 © 1996 American Chemical Society

Communications
J . Org. Chem., Vol. 61, No. 26, 1996 9079
Sch em e 1
Sch em e 2
conformational isomers B-1 and B-3, both of which gave
the rearranged products (Z)-4 and (E)-6, respectively.
The alternative is a radical mechanism as shown in
Scheme 2. It is known that 1,2-Wittig rearrangement
of the ethers proceeds not in a concerted fashion but via
the radical dissociation-recombination mechanism.7-9
The cyclic intermediate A, formed from (Z)-3, dissociates
to the radical pair of the imidoyl radical D10 and the
oxygen radical E, whereupon the oxygen radical E
isomerizes to carbon radical F . Recombination of the
resulting radical pair of imidoyl radical D and carbon
radical F occurs more rapidly than inversion of the
imidoyl radical center D to the geometrical isomer G,
judging from the high degree of retention at the oxime
ether group. On the rearrangement of (E)-hydroximate
5, it is suggested that the isomerization of the imidoyl
radical center G, formed from intermediate C, isomerizes
partially to the isomer D due to steric hindrance between
the R1 group and the methoxy group in G during the
course of the reaction. Elucidation of the precise mech-
anism involving stereoselectivity of the imino group
requires further detailed investigation.
possible reaction pathways for this rearrangement. The
first pathway proceeds by an ionic addition-elimination
process as proposed for the related 1,2-rearrangements
(Scheme 1).2a,d,6 Grovenstein6 suggested that in the 1,2-
Wittig rearrangement of benzyl propenyl ether initial
cyclization of the lithio ether, generated from the benzyl
propenyl ether, proceeds by anti-addition to give the
epoxide, and subsequent ring opening of the epoxide
occurs by syn-elimination to give the product with the
same configuration at the trisubstituted olefin. Accord-
ing to this mechanism, we propose a possible reaction
pathway of the imino-Wittig rearrangement as shown in
Scheme 1. Treatment of (Z)-3 with LDA gives the lithio
ether A, which would be stabilized by chelation with a
methoxy group. Intramolecular addition of the resulting
carbanion A to the imino double bond then proceeds in
an anti-fashion to give the epoxide B-1 as shown in the
Newman projection. Finally, the epoxide B-1 undergoes
ring-opening reaction in anti-periplanar manner involv-
ing the nitrogen lone pair and C-O bond to afford the
rearranged product (Z)-4 with retention of configuration
at the methoxy imino group.
On the other hand, rearrangement of (E)-hydroximate
5 proceeded slowly and nonstereoselectively to give a
mixture of (E)-6 and (Z)-4. As in the case of (Z)-3,
treatment of (E)-5 with LDA gives the lithio ether C,
which would not be stabilized by chelation with the
methoxy group and then undergoes anti-addition to the
imino group to give the epoxide B-2. The Newman
projection shows that in the conformation B-2 the C-O
bond and lone pair are not situated anti-periplanar
suitable for the E2 type of the final ring-opening reaction.
Therefore, there would be two possible reaction path-
ways. One is E1cB-type elimination from the conforma-
tion B-2, which gives a mixture of (E)-6 and (Z)-4
products. The other is E2-type elimination from the
In conclusion, we have shown for the first time that
benzyl and allyl (Z)-hydroximates smoothly rearrange to
give 2-hydroxy oxime ethers, which would serve as
precursors of diols and amino alcohols widely found in
biologically active natural products. The application of
the imino-Wittig rearrangement to the synthesis of
biologically active natural products is now under study.
Ack n ow led gm en t. We wish to thank the Ministry
of Education, Science, Sports and Culture of J apan and
the Science Research Promotion Fund of the J apan
Private School Promotion Foundation for research grants.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedure and compound characterization data (5 pages).
J O961768Y
(7) Tomooka, K.; Igarashi, T.; Nakai, T. Tetrahedron 1994, 50, 5927-
5932 and references cited therein.
(6) Grovenstein, E., J r.; Black, K. W.; Goel, S. C.; Hughes, R. L.;
Northrop, J . H.; Streeter, D. L.; VanDerveer, D. J . Org. Chem. 1989,
54, 1671-1679. Grovenstein, et al. have discussed the reaction mech-
anism of the 1,2-rearrangement of the benzyl and methallyl propenyl
ethers leading to the formation of the carbinols with retention of
stereochemistry in the propenyl group, which was previously reported
by Bu¨chi’s group.2a
(8) Verner, E. J .; Cohen, T. J . Am. Chem. Soc. 1992, 114, 375-377.
(9) Lansbury, P. T.; Pattison, V. A.; Sidler, J . D.; Bieber J . B. J .
Am. Chem. Soc. 1966, 88, 78-84.
(10) We could not isolate cyclopentanone O-methyloxime as
a
product that is expected to be formed via radical cyclization of an
intermediary imidoyl radical D in the reaction of 3i.