Enantioselective allylation of β,γ-unsaturated aldehydes generated via Lewis acid induced rearrangement of 2-vinyloxiranes

Article abstract of DOI:10.1021/ol016946a

(matrix presented) 2-Vinyloxiranes have been found to be excellent surrogates to β,γ-unsaturated aldehydes. These valuable electrophiles, generated in situ by treatment of a 2-vinyloxirane with a catalytic amount of Sc(OTf)₃, are effectively tr

Full text of DOI:10.1021/ol016946a

ORGANIC
LETTERS
2002
Vol. 4, No. 1
83-86
Enantioselective Allylation of
â,γ-Unsaturated Aldehydes Generated
via Lewis Acid Induced Rearrangement
of 2-Vinyloxiranes
Mark Lautens,* Matthew L. Maddess, Effiette L. O. Sauer, and
Ste´phane G. Ouellet
DaVenport Research Laboratories, Department of Chemistry, UniVersity of Toronto,
80 St. George Street, Toronto, Ontario M5S 3H6, Canada
mlautens@alchemy.chem.utoronto.ca
Received October 22, 2001
ABSTRACT
2-Vinyloxiranes have been found to be excellent surrogates to â,γ-unsaturated aldehydes. These valuable electrophiles, generated in situ by
treatment of a 2-vinyloxirane with a catalytic amount of Sc(OTf)3, are effectively trapped by the chiral allylating agents based on r-pinene,
affording bishomoallylic alcohols in high yield and excellent selectivity.
â,γ-Unsaturated aldehydes have rarely been used as building
blocks in synthetic organic chemistry, presumably due to
their very low stability with respect to olefin isomerization.1,2
As such, the potential utility of these substrates as electro-
philes remains largely unexplored. One practical solution to
this instability is to form and react the â,γ-unsaturated
aldehyde in situ; a few strategies have been reported along
these lines.3,4 We recently disclosed5 a high-yielding protocol
for the allylation or crotylation of various â,γ-unsaturated
aldehydes, generated by the treatment of 2-vinyloxiranes with
a Lewis acid (LA). To extend the synthetic utility of this
methodology, we sought to develop an asymmetric version
of this transformation (Scheme 1). In this report we describe
Scheme 1. Enantio-Strategy for Preparation of Chiral
Bishomoallylic Alcohols
(1) For examples, see: (a) Saha, G.; Basu, M. K.; Kim, S.; Jung, Y.;
Adiyaman, Y.; Powell, W. S.; FitzGerald, G. A.; Rokach, J. Tetrahedron
Lett. 1999, 40, 7179. (b) Crimmins, M. T.; Choy, A. L. J. Am. Chem. Soc.
1999, 121, 5653. (c) Paquette, L. A.; Braun, A. Tetrahedron Lett. 1997,
38, 5119. (d) Roush, W. R. J. Org. Chem. 1991, 56, 4151. (e) Ahmar, M.;
Bloch, R.; Mandville, G.; Romain, I. Tetrahedron Lett. 1992, 33, 2501.
(2) (a) Bachmann, W. E.; Horwitz, J. P.; Warzynshi, R. J.J. Am. Chem.
Soc. 1953, 75, 3268. (b) Serramedan, D.; Marc, F.; Pereyre, M.; Filliatre,
C.; Chabardes, P.; Delmond, B. Tetrahedron Lett. 1992, 33, 4457. (c) Marc,
F.; Soulet, B.; Serramedan, D.; Delmond, B. Tetrahedron 1994, 50, 3381.
(3) (a) Hertweck, C.; Boland, W. J. Org. Chem. 2000, 65, 2458. (b)
Bideau, F. L.; Gilloir, F.; Nilsson, Y.; Aubert, C.; Malacria, M. Tetrahedron
1996, 52, 7487.; (c) Wipf, P.; Xu, W. J. Org. Chem. 1993, 53, 825.
(4) For an alternative to the use of â,γ-unsaturated aldehydes, see:
Hughes, G.; Lautens, M.; Wen, C. Org. Lett. 2000, 2, 107.
the successful realization of this objective, allowing ready
access to enantiomerically enriched bishomoallylic alcohols.
We initiated our studies by examining the use of chiral
allylboron compounds6 as nucleophiles. This approach
(6) Roush, W. R. Methods of Organic Chemistry (Houben-Weyl) 4th
Edition; Helmchen, G., Hoffmann, R. W., Mulzer, J., Schaumann, E., Eds.;
Georg Thieme Verlag Stuttgart; New York; 1995; Vol. E21b, pp 1410-
1486.
(5) Lautens, M.; Ouellet, S. G.; Raeppel, S. Angew. Chem., Int. Ed. 2000,
39, 4079.
10.1021/ol016946a CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/15/2001

seemed especially promising since as a result of the empty
p-orbital at the boron atom, these species are also Lewis
acidic. We reasoned this feature could remove the need for
an external LA since the nucleophile could play the dual
role of ring opening the oxirane and effecting allylation. The
reagents developed by Roush7 (1), Corey8 (2), and Brown9
(3 produced from (R)-(+)-R-pinene, or ent-3 from (S)-(-)-
R-pinene), were prepared according to literature procedures
and reacted with our test substrate, 2-styryl-oxirane (4). A
summary of the results is presented in Table 1.
the reaction mixture, and notably (MeO)MgBr has been
previously reported to retard the reaction rate and lower the
ee.9 Thus, Brown’s reagent (3) was prepared salt free
according to a published report9 from basic starting materials.
Although this procedure requires careful manipulation, once
prepared 3 may be stored for more than 1 month as a solution
in THF.10 The initial results were very satisfying, and the
observed selectivity was excellent (Table 1, entries 4 and
5).
A variety of Lewis acids were examined (Table 2), and
Table 1. Effect of the Nucleophilea
Table 2. Effect of the Lewis Acida
entry nucleophile (equiv)
LA (mol %)
yield (%) ee (%)e
entry
LA
yield (%)b
ee (%)b
note
1
2
3
4
5
1 (1.5)
1 (1.5)
2 (1.5)
3 (2.3)
3 (2.3)
40
27
47
37
93
96
BF3‚OEt2 (30)
BF3‚OEt2 (30)
BF3‚OEt2 (20)
BF3‚OEt2 (20)
70
1
2
3
4
5
6
Sm(OTf)3
SnCl4
Ag(OTf)
Y(OTf)3
Sc(OTf)3
BF3‚OEt2
28
18
33
65
92
59
96
83
94
96
96
96
incomplete
33b
48b,c
80b,d
incomplete
incomplete
a All reactions performed by slow addition of the oxirane over 1 h.
b HPLC yield. c NH4Cl workup. d NaOH/H2O2 workup. e Determined by
HPLC.
incomplete
a All reactions performed by slow addition of the oxirane over 1 h to a
solution of 3 (2.3 equiv) and LA (10%) in THF at -78 °C (0.34 mmol
scale). b Yield and ee determined by HPLC. Absolute configuration predicted
by analogy.12
all gave the desired product (5) with comparable levels of
induction with the exception of SnCl4 (Table 2, entry 2)
where a slightly reduced ee was observed. By far, the cleanest
and highest yielding reaction was obtained with Sc(OTf)3
(Table 2, entry 5). The majority of the other Lewis acids
suffered from incomplete conversion (Table 2, entries 1, 3,
4, and 6) or caused decomposition of the vinyloxirane.
Although the use of BF3‚OEt2 is economically attractive for
large-scale work, decomposition (Table 1, entry 5) or
incomplete conversion (Table 2, entry 6) resulted in lower
isolated yields. As a control experiment, (MeO)MgBr11 (2.3
equiv) was added to “salt free” 3 along with Sc(OTf)3 (7.5
mol %). In analogy with the failure of the one-pot reaction
from commercially available (MeO)B(Ipc)2, only starting
material (4) was recovered.
Roush’s reagent (1) was studied extensively, and the results
reported are the best following optimization studies. Of the
three boron-based allylating agents, 1 was the only one to
effect ring opening without the addition of an external LA
(Table 1, entry 1). However, both the yield and selectivity
of the bishomoallylic alcohol 5 were improved by the
presence of BF3‚OEt2 (Table 1, entry 2). Corey’s allylating
reagent (2) was briefly examined, but provided inferior results
for our system (Table 1, entry 3) and was significantly more
difficult to prepare then the tartrate-based allylboron com-
pound 1. Consequently, our focus shifted to Brown’s reagent
(3), and initial studies were performed using the com-
mercially available precursors (MeO)B(Ipc)2 or DIPCl (Al-
drich) to prepare the active trialkyl reagent (either 3 or ent-
3) in situ. When combined with a LA and treated with the
vinyloxirane (4), the reaction originating from (MeO)B(Ipc)2
failed to give any product whatsoever and primarily starting
material was recovered. Alternatively, reactions that em-
ployed Brown’s reagent (3) prepared from DIPCl were very
promising, but surprisingly inconsistent. In both cases the
presence of the magnesium salts significantly complicated
A brief examination of reaction stoichiometry showed that
the ee remained constant but the best yield was obtained
with 2.3 equiv of the allylation agent (3) and 7.5-10 mol
% of Sc(OTf)3. Examination of the effect of varying the
temperature (Table 3) indicated that the ee remained above
90% (Table 3, entries 1-3) as long as the reaction was kept
< -50 °C. Even at room temperature the level of selectivity
was good (ee ) 86%) (Table 3, entry 5).
(10) Solution stored in a Wheaton bottle at -20 °C to ensure consistency.
It is highly recommended that the methanol be freshly distilled from
magnesium and iodine.
(11) This magnesium salt was obtained from drying the filtrate from the
synthesis of 3 under high vacuum for 12 h.
(7) Roush, W. R.; Walts, A. E.; Hoong, L. K. J. Am. Chem. Soc. 1985,
107, 8186.
(8) Corey, E. J.; Yu, C.; Kim, S. S. J. Am. Chem. Soc. 1989, 111, 5495.
(9) Racherla, U. S.; Brown, H. C. J. Org. Chem. 1991, 56, 401.
84
Org. Lett., Vol. 4, No. 1, 2002

yield and purity after a simple aqueous wash. This approach
was advantageous, as purification of these often-sensitive
epoxides could be avoided. When purification was required,14
the isolated chemical yield was typically sacrificed in order
to obtain the desired 2-vinyloxiranes free from the original
impurities and decomposition products that arise during their
purification.
Table 3. Effect of Temperaturea
entry
temp (°C)
yield (%)a
ee (%)a
In general, both the yield and enantiomeric excess for most
of the substrates were excellent (Table 4). Either antipode12
of the desired bishomoallylic alcohols is readily available
(Table 4, entries 1 versus 2) depending on whether Brown’s
reagent was synthesized from (R)-(+)- or (S)-(-)-R-pinene
(i.e., use of 3 versus ent-3). 2-Vinyloxiranes of moderate
reactivity (Table 4, entries 1-3, 8-10, and 11) gave the best
yields with no appreciable difference between aryl or alkyl
substituents. It appears that a balance between stabilization
of the proposed carbocation intermediate and the propensity
for the requisite 1,2-hydride shift to occur is responsible.
Electron rich substrates (Table 4, entry 5) were very reactive
toward ring opening but gave lower yields in the carbonyl
addition with the remainder of the mass balance being
decomposition products. This is potentially explained as a
consequence of an increased lifetime for the carbocation
intermediate. On the other hand, electron poor substrates
(Table 4, entries 4, 6, and 7) were much more sluggish to
react. However, with extended reaction times the o-nitro
product (ent-13) was available in high yield and excellent
enantioselectivity (Table 4, entry 7 versus 6). A vinyloxirane
with a bromo substituent R to the epoxide (8) failed to react
(Table 4, entry 4) except when BF3‚OEt2 at significantly
higher loadings of the LA (1.1 equiv versus 7.5%) was
employed. The ee of 9 was high (90%), but the yield was
poor (17%) due to significant decomposition under these
1
2
3
4
5
-96
-78
-50
0
94
96
100
100
100
>98
97
95
89
86
20
a All reactions performed by slow addition of the oxirane over 1 h
to a solution of 3 (2.3 equiv) and LA (7.5%) in THF (0.34 mmol scale).
b Yield and ee determined by HPLC. Absolute configuration predicted by
analogy.12
We next prepared a series of 2-vinyloxiranes through the
two-step sequence described by Matteson13 (Scheme 2). In
Scheme 2. Approach to Synthesis of 2-Vinyloxiranesa
a (a) ClCH2I (1.5 equiv), n-BuLi (1.5 equiv), THF, -78 °C, ∼1
h; (b) NaH (95%, 1.1 equiv), NaI (10%), THF, 0 °C, ∼1 h.
most cases, careful purification of the chlorohydrin inter-
mediates allowed the desired epoxides to be isolated in high
Table 4. Substrate Scopea
substrate
product
entry no.
R1
R2 R3 R4
Nu c
temp (°C)
-78
en t-3 -78
time (h)
no.
yield (%)b
ee (%)c
96
96
96
1
2
4
4
Ph
Ph
Ph
Ph
H
H
H
H
H
H
H
Ph
H
H
H
H
Me
Br
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
3
6
6
6
6
6
6
5
88 (92)d
96d
83
SM
68
35
88
94
92
en t-5
7
3
6
3
3
3
3
-78
-78
-78
-78
4
8
9
5
6
7
8
10
12
12
14
o-(MeO)C6H4
o-(NO2)C6H4
o-(NO2)C6H4
Ph
11
13
96
96
96
95
en t-3 -78, then -70 4, then 36 en t-13
3
3
3
3
-78
-78
-78
-78
6
6
6
6
15
17
19
21
9
1615 Ph
>95,e de ∼0
10
11
18
20
CH3(CH2)4
H
80
92
97f
96f
R1 to R3 ) (CH2)4, R2 and R4 ) H
a All reactions performed by the slow addition of the oxirane over 1 h on a 0.34-0.68 mmol scale. b Yields are isolated yields. c ee determined by chiral
HPLC against racemic material.5 d HPLC yield. e ee of both diastereomers greater than 95, inseparable. f ee determined by chiral GC. Absolute configuration
predicted by analogy.12
Org. Lett., Vol. 4, No. 1, 2002
85

and excellent selectivity (Table 4, entry 9). No control over
the R-methyl stereocenter was observed, and a nearly 1:1
mixture of diastereomers (inseparable) was isolated (17). Our
current studies are focused on addressing this issue.
In conclusion, we have shown that 2-vinyloxiranes are
excellent surrogates for â,γ-unsaturated aldehydes in the
preparation of bishomoallylic alcohols. These valuable
electrophiles, generated in situ by treatment of a 2-vinylox-
irane with a catalytic amount of Sc(OTf)3, are effectively
trapped by chiral allylating agents based on R-pinene,
affording the products in high yield and excellent selectivity.
The reaction is highly versatile and allows access to a useful
class of products that would be difficult to synthesize by
other means. Mechanistic details, further examples, and
applications will be reported in due course.
Scheme 3. Palladium-Catalyzed Cross Coupling-Allylation
Sequencea
a Pd(PPh3)4 (5%), CuI (10%), THF/Et3N (1:1), rt, 3 h. b ClCH2I
(1.5 equiv), n-BuLi (1.5 equiv), THF, -78 °C, 1 h. c NaH (95%,
1.1 equiv), NaI (10%), THF, 0 °C, 2 h. d (1) ent-3 (2.3 equiv),
Sc(OTf)3 (7.5%), THF, -78 °C, 6 h; (2) 3 N NaOH, 30% H2O2,
rt, 14 h. e (1) ent-3 (2.3 equiv), Sc(OTf)3 (7.5%), THF, -78 °C, 4
h, -70 °C 36 h; (2) 3 N NaOH, 30% H2O2, rt, 14 h.
Acknowledgment. We thank Merck Frosst Canada, the
Ontario Research and Development Challenge Fund (OR-
DCF), NSERC (Canada), and the University of Toronto for
financial support of this work. M.M. thanks NSERC (Canada)
for financial support in the form of a postgraduate fellowship.
conditions. It is possible to perform a palladium cross
coupling reaction as a first step and subsequently carry the
intermediate on to a bishomoallylic alcohol (25) (Scheme
3).
Finally, a 2-vinyloxirane substituted at the 2-position (16)
performed very well, giving the desired products in high yield
Supporting Information Available: Representative ex-
perimental procedures, and full characterization of all novel
compounds within 4-25 as well as related intermediates.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(12) The absolute configuration may be predicted from analogy with
previous results. 3 is known to give si face attack while ent-3 provides the
enantiomer. See: (a) Brown, H. C.; Jadhav, P. K. J. Am. Chem. Soc. 1983,
105, 2092; (b) Vulpetti, A.; Gardner, M.; Gennari, C.; Bernardi, A.;
Goodman, J. M.; Paterson, I. J. Org. Chem. 1993, 58, 1711.
OL016946A
(14) When necessary the 2-vinyloxiranes may be purified by distillation
or in some cases flash chromatography (5% Et3N, EtOAc/Hex).
(13) Sadhu, K. M.; Matteson, D. S. Tetrahedron Lett. 1986, 27, 795.
86
Org. Lett., Vol. 4, No. 1, 2002