Inverted diastereoselectivity in asymmetric aziridine synthesis via aza-Darzens reaction of (2S)-N-bromoacyl camphorsultam

Article abstract of DOI:10.1021/ol990874u

(formula presented) The aza-Darzens reaction of the chiral enolate derived from (2S)-bromoacetyl camphor sultam (1) with certain C-3-substituted N-diphenylphosphinyl imines gives mixtures of trans- and cis-aziridines. In some cases, only trans isomers are

Full text of DOI:10.1021/ol990874u

ORGANIC
LETTERS
1
999
Vol. 1, No. 9
339-1341
Inverted Diastereoselectivity in
Asymmetric Aziridine Synthesis via
Aza-Darzens Reaction of
1
(2S)-N-Bromoacyl Camphorsultam
Andrew B. McLaren and J. B. Sweeney*
Department of Chemistry, UniVersity of Reading, Reading RG6 6AD, U.K.
j.b.sweeney@reading.ac.uk
Received July 28, 1999
ABSTRACT
The aza-Darzens reaction of the chiral enolate derived from (2S)-bromoacetyl camphor sultam (1) with certain C-3-substituted N-diphenylphosphinyl
imines gives mixtures of trans- and cis-aziridines. In some cases, only trans isomers are observed. A steric repulsion between the enolate
halogen atom and this C-3-substituent is invoked to rationalize these observations.
There has been much attention of late on the synthesis and
1
uses of chiral aziridines. We recently reported the aza-
Darzens reaction (ADZ) of N-bromoacyl (2R)-camphorsul-
tam with N-diphenylphosphinylimines (“N-Dpp imines”) to
be a useful method for preparation of enantiopure cis-2-
Scheme 1. Variable Diastereocontrol in ADZ Reaction
2
carboxy aziridines. We here report our recent findings that
the presence of substituents in the C-3-position of certain
imines leads to altered diastereocontrol in the ADZ reaction
of bromoacyl (2S)-camphorsultam, in certain cases leading
exclusively to trans-aziridines.
N-Dpp benzaldimine reacted with the enolate derived from
(
2S)-sultam (1) under the conditions outlined in Scheme 1
3
to give exclusively the cis-aziridine 2 (R ) Ph) ( J ) 6.2,
typical of such aziridines ) in 71% yield, but the correspond-
3
ing imine derived from 2-chlorobenzaldehyde gave a mixture
(1) Davis, F. A.; McCoull, W. Tetrahedron Lett. 1999, 40, 249.
Sodergren, M. J.; Alonso, D. A.; Bedekar, A. V.; Andersson, P. G.
Tetrahedron Lett. 1997, 38, 6897. Osborn, H. M. I.; Sweeney, J.
Tetrahedron: Asymmetry 1997, 8, 1693. Sodergren, M. J.; Alonso, D. A.;
Andersson, P. G. Tetrahedron: Asymmetry 1997, 8, 3563.
of cis- and trans-aziridines 3 in 68% combined yield, with
3
the trans isomer dominating (cis:trans ) 1:4; Jcis ) 6.1 Hz,
3
J
trans ) 2.8 Hz) (Scheme 1). The absolute configuration of
(2) Cantrill, A. A.; Hall, L. D.; Jarvis, A. N.; Osborn, H. M. I.; Raphy,
J.; Sweeney, J. B. J. Chem. Soc., Chem. Commun. 1996, 2631.
the cis-aziridinyl sultam was deduced by comparison with
1
0.1021/ol990874u CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/02/1999

the H and 13C NMR obtained for authentic phenylaziri-
dine 2 (shown by X-ray analysis to be of 2R,2′R,3′R
configuration), while that of the trans-aziridine was con-
firmed as (2′S,3′R) by X-ray analysis of a single crystal
1
Table 1. Variation in Diastereoselectivity in ADZ of N-Dpp
Imines
R
yield/%
cis:trans
3J cis/trans/Hz
(Figure 1).
Ph (2)
71
84
68
67
73
60
65
87
65
74
72
77
67
100:0
50:50
20:80
0:100
0:100
100:0
100:0
50:50
0:100
100:0
100:0
100:0
100:0
6.2
2
2
2
2
3
4
2
2
4
2
4
2
-F-C6H4
6.2/2.8
6.1/2.8
2.6
2.8
6.2
6.2
6.0/2.8
3.1
6.2
5.8
-Cl-C6H4 (3)
-Br-C6H4
-I-C6H4
-Br-C6H4
-Br-C6H4
-Me-C6H4
-OMe-C6H4
-OMe-C6H4
-NO2-C6H4
-NO2-C6H4
-pyridyl
6.4
6.6
Interestingly, when the N-Dpp imines derived from ac-
rolein and 3,3-dimethylacrolein were used in the ADZ
reaction, a similar variable diastereoselectivity was ob-
served (Scheme 2). Thus, the unsubstituted vinyl imine
Figure 1. X-ray structure of trans-aziridine 3.
Scheme 2. Variable Diastereocontrol in ADZ Reaction of
Vinylimines
This inversion of the diastereopreference previously seen
was even more pronounced in the reaction of the N-Dpp
imine derived from 2-bromobenzaldehyde: in this reaction
3
only a trans-aziridine was observed (67% yield, J ) 2.6
Hz), an observation also made when N-Dpp-2-iodobenzald-
imine was employed in the reaction. When N-Dpp-2-
fluorobenzaldimine was reacted under the same conditions,
lower stereoselectivity was observed, as might be expected,
3
3
with the cis:trans ratio being unity ( Jcis ) 6.2 Hz, Jtrans
.8 Hz), although the yield of aziridine was good.
At first sight, these data seem to suggest that the bulk of
)
2
the 2-substituent is intimately related to the diastereoselec-
tivity of the reaction; to further probe the structural influ-
ences, N-Dpp-3-bromobenzaldimine was next subjected to
reacted in the “usual” fashion to give exclusively cis
aziridine, whereas the 3,3-dimethyl analogue gave trans-
configured aziridine.
3
the reaction conditions. In this case, only a cis-aziridine ( J
)
6.2 Hz) was isolated from the reaction. We then turned
our attention to nonhalogenated 2-substituted benzaldimines.
N-Dpp-2-methylbenzaldimine showed diastereoselectivity
similar to that of the analogous 2-fluoro imine, giving a 1:1
From these data, one may make a tentative suggestion as
to the sequence of events responsible for the variation in
(3) All previously unreported compounds exhibited satisfactory physical
3
3
mixture of cis- and trans-aziridines ( Jcis ) 6.0 Hz, Jtrans
)
data. The following procedure is representative: N-bromoacetyl-(2S)-
bornane-10,2-sultam (336 mg, 1.0 mmol) was dissolved in 20 mL of dry
THF under a nitrogen atmosphere and cooled to -78 °C. Lithium
2.8 Hz), while the corresponding 2- and 4-methoxy imines
reacted to give only one diastereoisomer in each case (a
(
hexamethyldisilyl)amide (1.1 mL, 1.0 M, 1.1 mmol) was added dropwise
3
trans-aziridine ( J ) 3.1 Hz) from the 2-isomer and a cis-
and the resulting yellow solution stirred for approximately 30 min. After
this time, N-diphenylphosphinyl 2′-chlorobenzaldimine (340 mg, 1.0 mmol)
was added as a solution in THF (15 mL) to the reaction mixture. The reaction
mixture was then left stirring at -78 °C for approximately 3-4 h and
followed by TLC. After this time the reaction was quenched via addition
of a saturated ammonium chloride solution (20 mL). The aqueous layer
was then extracted with diethyl ether (3 × 20 mL), the organic layers were
combined, washed with brine, dried (MgSO4), and filtered, and the solvent
was removed in vacuo to afford cis- and trans-(2S,2′S,3′S)-N-[(1-diphenyl-
phosphinyl-3-(2-chlorophenyl)-2-aziridinyl)carbonyl]bornane-10,2-sultam
aziridine from the 4-isomer). A 2-nitro substituent favored
exclusive formation of a cis-aziridine, perhaps due to lower
steric demand of this planar substituent. Once again, these
data do not point conclusively to a convincing rationalization
of the phenomena inducing this stereocontrol. These pre-
liminary data are collected in Table 1.
1340
Org. Lett., Vol. 1, No. 9, 1999

diastereoselectivity in this ADZ reaction (Scheme 3). Thus,
although there is a paucity of empirical data concerning
transition-state preferences in the reaction of imines with
Scheme 3. Mechanistic Possibilities in ADZ Reaction of
C-3-Substituted N-Dpp Imines
4
enolates, we suggest a closed transition state along Zim-
5
merman-Traxler guidelines. Since the aziridine C2 is
indubitably of (S)-absolute configuration, the reaction must
6
involve attack of the si-face of the Z-enolate upon the imine.
Where there is no imine C-3 substituent, we suggest a
pseudoequatorial positioning of the imine C-1 substituent
(
i.e., an E-configured imine) and attack upon the si-face.
When there is C-3-substitution, we postulate a repulsive
interaction between this group and the Br atom of the chiral
sultam enolate, leading to a preference for a transition state
in which the imine substituent adopts a pseudoaxial locus
7
(
a Z-imine ), leading to re-face attack. This arrangement leads
inevitably to a (2′S,3′R)- trans-aziridine, as shown in Scheme
. One may rationalize the lower selectivity shown for trans-
3
aziridine in the reaction of 2-methylbenzaldimine (whose C-3
substituent should engender at least as much steric demand
8
as a bromo substituent [which leads to an exclusively trans-
1
3
as a very pale yellow solid. From H NMR, the mixture was adjudged to
be approximately a 20:80 mixture of cis:trans isomers. After flash
chromatography (light petroleum ether:ethyl acetate (1:9)) cis-3 was obtained
20
as a colorless solid (81 mg, 0.13 mmol, 13%); Rf 0.63 (EtOAc); [R] D )
-
1
+
28.7 (c ) 1, CH2Cl2); νmax(CCl4)/cm 3057, 2964, 1703, 1440, 1341,
128, 1266, 1169, 745, 705, 644; δH (400 MHz, CDCl3) 0.78-0.83 (6H,
m), 1.12-1.20 and 1.72-1.97 (7H, m), 3.22 and 3.30 (2H, 2 × d, J )
3.9 Hz), 3.57-3.60 (1H, m), 4.11-4.50 (2H, 2 × dd, J ) 6.1 Hz, 15.9
1
1
Hz), 7.10-7.57 and 7.85-8.03 (14H, m); δC (100 MHz, CDCl3) 19.72,
2
4
1
1
0.55, 26.29, 32.49, 38.12, 40.69, 42.31 (2 × CH, J ) 5.6 Hz), 44.56,
7.70, 48.95, 52.46, 64.58, 126.01, 128.44, 128.57, 128.68, 129.07, 129.18,
29.58, 130.62, 130.82, 131.55, 131.68, 131.74, 131.77, 131.83, 131.94,
+
32.10, 132.27, 134.61, 163.74; m/z (CI) 595 ([MH] , 66), 531 (60), 419
100), 313 (43), 201 (91), 77 (32) (found: [MH] , 595.1594, C31H33ClN2O4-
PS requires [MH] , 595.1588). Similarly, trans-3 was obtained as a colorless
+
(
+
2
0
solid (325 mg, 0.55 mmol, 55%); Rf 0.56 (EtOAc); [R] D ) +59.4 (c )
-
1
1
1
1
, CH2Cl2); νmax(CCl4)/cm 3056, 2985, 1705, 1440, 1338, 1168, 1266,
126, 738, 706, 623 (Ar); δH (400 MHz, CDCl3) 0.75 (3H, s), 0.80 (3H, s),
.01-1.29 and 1.65-1.89 (7H, m), 3.23 (1H, d, J ) 13.6 Hz), 3.28 (1H,
d, J ) 13.6 Hz), 3.72 (1H, m), 4.16 (1H, dd, J ) 2.8 Hz, 13.3 Hz), 4.30
(
7
3
1H, dd, J ) 2.8 Hz, 13.3 Hz), 7.03-7.10 and 7.19-7.34 (10H, m), 7.63-
.67 and 7.79-7.83 (4H, m); δC (100 MHz, CDCl3) 19.65, 20.64, 26.29,
2.52, 37.75, 41.27 (aziridine CH, d, J ) 5.5 Hz), 43.96 (aziridine CH, d,
J ) 5.5 Hz), 44.37, 47.59, 48.76, 52.71, 65.11, 126.52, 127.95, 128.08,
configured aziridine]) by supposing that a Br-Br repulsion
(involving lone-pair:lone-pair interaction) is greater than a
Br-CH repulsion (involving a lone-pair:bonding-electron
repulsion). The precise details of the mechanism responsible
for these phenomena are at present under investigation in
our laboratory.
1
1
1
28.12, 128.24, 128.41, 128.94, 129.41, 131.19, 131.55, 131.65, 132.45,
+
33.00, 134.32, 135.49, 164.92 (CO, d, J ) 5.6 Hz); m/z (CI) 594 ([M] ,
9), 559 (49), 219 (68), 201 (100), 77 (27) (found: [M] , 594.1559, C31H32-
+
+
ClN2O4PS requires [M] , 594.1508).
(
4) Asymmetric aza-Darzens and Darzens-like reactions: Gennari, C.;
Pain, G. Tetrahedron Lett. 1996, 37, 3747. Fujisawa, T.; Hayakawa, R.;
Shimizu, M. Tetrahedron Lett. 1992, 33, 7903. Davis, F. A.; Zhou P.; Reddy,
G. V. J. Org. Chem. 1994, 59, 3243. Davis, F. A.; Zhou, P.; Liang, C. H.;
Reddy, R. E. Tetrahedron: Asymmetry 1995, 6, 1511. Florio, S.; Troisi L.;
Capriati, V. J. Org. Chem. 1995, 60, 2279.
Acknowledgment. We acknowledge the financial support
of the EPSRC (studenthship to A.B.McL.). J.B.S. thanks
ZENECA for an award from the Strategic Research Fund
and Mr. Glen Buchman for valuable advice.
(
5) Zimmerman, H. E. J. Am. Chem. Soc. 1957, 79, 1920.
(6) Calculations (UFF, Rapper, A. K.; Casewit, C. J.; Colwell, K. S.;
Goddard, W. A., III; Skiff, W. M. J. Am. Chem. Soc. 1992, 114, 10024)
show the Z-enolate to be more stable than the E-enolate by 4.1 kcal/mol.
3
(7) Although all imines used were of (E)-configuration ( JP-H ) 32 Hz)
in the pure state, (E)-(Z)-isomerism of imines under the influence of metal
ions is well-known; see, for example: Alshalaan, A. M.; Alshowiman, S.
S.; Alnajjar, I. M. Inorg. Chim. Acta 1986, 121, 127-129. For a discussion
of the factors underlying (E)/(Z)-preferences in imines, see: Bjørgo, J.;
Boyd, D. R.; Watson, C. G.; Jennings, W. B. J. Chem. Soc., Perkin Trans.
OL990874U
(8) Streitwieser, A.; Heathcock, C. H.; Kosower, E. M. Introduction to
Organic Chemistry; Macmillian Publishing Company: New York, 1992.
1
1974, 757.
Org. Lett., Vol. 1, No. 9, 1999
1341