Lewis acid mediated asymmetric [2,3]-sigmatropic rearrangement of allylic amines. Scope and mechanistic investigation

Article abstract of DOI:10.1021/jo062178v

(Chemical Equation Presented) The first asymmetric [2,3]-sigmatropic rearrangement of achiral allylic amines has been realized by quaternization of the amines with an enantiomerically pure diazaborolidine and subsequent treatment with Et₃N. The

Full text of DOI:10.1021/jo062178v

Lewis Acid Mediated Asymmetric [2,3]-Sigmatropic Rearrangement
of Allylic Amines. Scope and Mechanistic Investigation
Jan Blid, Olaf Panknin, Pavel Tuzina, and Peter Somfai*
Organic Chemistry, KTH Chemical Science and Engineering, SE-100 44 Stockholm, Sweden
somfai@kth.se
ReceiVed October 19, 2006
The first asymmetric [2,3]-sigmatropic rearrangement of achiral allylic amines has been realized by
quaternization of the amines with an enantiomerically pure diazaborolidine and subsequent treatment
with Et3N. The resultant homoallylic amines were obtained in good yields and excellent ee’s. The observed
diastereo- and enantioselectivities were rationalized by invoking a kinetically controlled process, and
support for this model was obtained from an NMR spectroscopic investigation of the chiral Lewis acid-
substrate complex. The structure of the Lewis acid-product complex was established by X-ray
crystallographic analysis and supported the proposed mechanism.
SCHEME 1. [2,3]-Sigmatropic Rearrangement of Ylidesa
Introduction
Carbon-carbon bond-forming reactions are of central im-
portance to the organic chemist, and of particular interest are
those transformations that proceed with high regio- and stereo-
selectivity. Although many such reactions have been success-
fully developed, pursuit of asymmetric versions is still a
challenging goal. In this respect, the [2,3]-sigmatropic rear-
rangement of allylic onium salts (Scheme 1),1-3 such as
ammonium,4,5 sulfonium,6 oxonium,3,7 selenonium,8 and iodo-
a Key: R, R′ ) alkyl, Ar. X ) NR2 , SR2, OR2, SeR2, I. G ) Ar, CO2Me,
2
CN, CCTMS.
nium salts,5,9 have been investigated, establishing the rearrange-
ment as a powerful C-C bond-forming reaction. The rearrange-
ment is believed to proceed via a five-membered cyclic transition
state,10,11 and the stereochemical outcome is dependent on steric
and electronic interactions between the migrating allyl moiety
and the anion-stabilizing group (G, Scheme 1) in the exo and
endo transition states.1
(1) Marko´, I. E. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 3, pp 913-974.
(2) Li, A.-H.; Dai, L.-X.; Aggarwal, V. K. Chem. ReV. 1997, 97, 2341-
2372.
(3) Hodgson, D. M.; Pierard, F. Y. T. M.; Stupple, P. A. Chem. Soc.
ReV. 2001, 30, 50-61.
(4) (a) Clark, J. S.; Hodgson, P. B.; Goldsmith, M. D.; Blake, A. J.;
Cooke, P. A.; Street, L. J. J. Chem. Soc., Perkin Trans. 2001, 1, 3325-
3337. (b) Clark, J. S.; Hodgson, P. B.; Goldsmith, M. D.; Street, L. J. J.
Chem. Soc., Perkin Trans. 2001, 1, 3312-3324. (c) Clark, J. S.; Middleton,
M. D. Org. Lett. 2002, 4, 765-768. (d) Smith, R. S.; Bentley, P. D.
Tetrahedron Lett. 2002, 43, 899-902. (e) Heath, P.; Roberts, E.; Sweeney,
J. B.; Wessel, H. P.; Workman, J. A. J. Org. Chem. 2003, 68, 4083-4086.
(5) Simonneaux, G.; Galardon, E.; Paul-Roth, C.; Gulea, M.; Masson,
S. J. Organomet. Chem. 2001, 617-618, 360-363.
(6) (a) Baldwin, J. E.; Hackler, R. E.; Kelly, D. P. J. Chem. Soc., Chem.
Commun. 1968, 537-538. (b) Blackburn, G. M.; Ollis, W. D.; Plackett, J.
D.; Smith, C.; Sutherland, I. O. J. Chem. Soc., Chem. Commun. 1968, 186-
188. (c) Chappie, T. A.; Weekley, R. M.; McMills, M. C. Tetrahedron
Lett. 1996, 37, 6523-6526. (d) Aggarwal, V. K.; Ferrara, M.; Hainz, R.;
Spey, S. E. Tetrahedron Lett. 1999, 40, 8923-8927. (e) Gulea, M.;
Marchand, P.; Masson, G.; Saquet, M.; Collignon, N. Synthesis 1999, 1635-
1639. (f) Zhou, C.-Y.; Yu, W.-Y.; Chan, P. W. H.; Che, C.-M. J. Org.
Chem. 2004, 69, 7072-7082.
(7) (a) Doyle, M. P.; Griffin, J. H.; Chinn, M. S.; van Leusen, D. J.
Org. Chem. 1984, 49, 1917-1925. (b) Pirrung, M. C.; Werner, J. A. J.
Am. Chem. Soc. 1986, 108, 6060-6062. (c) Roskamp, E. J.; Johnson, C.
R. J. Am. Chem. Soc. 1986, 108, 6062-6063.
(8) (a) Gassman, P. G.; Miura, T.; Mossman, A. J. Org. Chem. 1982,
47, 954-959. (b) Nishibayashi, Y.; Ohe, K.; Uemura, S. Chem. Commun.
1995, 1245-1246. (c) Kurose, N.; Takahashi, T.; Koizumi, T. J. Org. Chem.
1997, 62, 4562-4563.
(9) Doyle, M. P.; Tamblyn, W. H.; Bagheri, V. J. Org. Chem. 1981, 46,
5094-5102.
(10) (a) Mageswaran, S.; Ollis, W. D.; Sutherland, I. O.; Thebtaranonth,
Y. Chem. Commun. 1971, 1494-1495. (b) Rautenstrauch, V. HelV. Chim.
Acta 1972, 55, 2233-2240.
(11) Mageswaran, S.; Ollis, W. D.; Sutherland, I. O. J. Chem. Soc., Perkin
Trans. 1 1981, 1953-1962.
10.1021/jo062178v CCC: $37.00 © 2007 American Chemical Society
Published on Web 01/17/2007
1294
J. Org. Chem. 2007, 72, 1294-1300

[2,3]-Sigmatropic Rearrangement of Allylic Amines
TABLE 1. Optimization of the Asymmetric [2,3]-Sigmatropic
Rearrangement of 1aa
SCHEME 2. Preparation of the Chiral Lewis Acids 4a,b
by complexation with a Lewis acid.15 Having realized this, our
attention was directed toward the possibility of employing chiral
Lewis acids, and recently, we communicated the first asym-
metric [2,3]-sigmatropic rearrangement of achiral allylic amines
mediated by a chiral diazaborolidine.14 The transformations
proceeded in good yields and provided various secondary
homoallylic amines with moderate to good diastereoselectivities
and excellent ee’s; a rationalization of the observed selectivities
was also presented.15,16 It was envisaged that either a kinetically
or thermodynamically controlled process was responsible for
the stereochemical outcome. Herein, we report the full details
of this investigation.
ligand/
entry (equiv) (equiv)
BBr3/
base/
(equiv)
conversion
ee
solvent of 2a (%)b
(%)c
1d
2f
3f
4f
5f
6f
7
3a/2.0
3a/1.2
3a/1.2
3a/1.2
3a/1.2
3a/1.2
3a/1.2
3a/1.2
3a/2.0
2.4
1.4
1.4
1.4
1.4
1.4
1.2
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
5/1.0
5/1.0
Et3N/1.0
Et3N/3.0
K2CO3/5.0
NaHCO3/12 CH2Cl2
5/2.0
5/2.0
5/2.0
5/2.0
5/1.0
5/2.0
5/2.0
PhMe
29(∼30e) 68 (R)
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
22(55)
17(59)
43(32)
0(83)
85 (R)
69 (R)
89 (R)
-
0(83)
39(59)
79(0)
-
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
PhMe
Et2O
THF
CH2Cl2
CH2Cl2
CH2Cl2
97 (R)
48 (R)
96 (R)
97 (R)
97 (R)
96 (R)
96 (R)
84 (R)
97 (R)
25 (S)
28 (S)
8
9
79(0)
10g 3a/2.0
71(11)
18(78)
70(30)
27(33)
15(83)
87(3)
11
12
3a/2.0
3a/2.0
13h 3a/2.0
14
15
16
17
3a/2.0
3a/2.0
3b/2.0
3b/2.0
5/2.0
Results and Discussion
Et3N/5.0
Et3N/5.0
5/2.0
84(0)
87(nd)
The optimization of the asymmetric rearrangement of 1
mediated by 4a,b is summarized in Table 1.
a
Reaction conditions: to a premade solution of 4a at room temperature
The initial attempts to effect an asymmetric rearrangement
were performed by mixing the chiral ligand 3a and BBr3 to
generate Lewis acids 4a (Scheme 2).17 Subsequent removal of
the solvent and the formed HBr under a vacuum was followed
by addition of 1a, and after 1-2 h, the resultant 1a/4a complex
was treated with base. When a solution of amine 1a in PhMe
was added to 4a, a precipitate was formed, which prevented
efficient stirring of the reaction mixture. Although promising
results were obtained in this solvent (Table 1, entry 1), changing
to CH2Cl2 avoided this problem and resulted in a homogeneous
solution throughout the reaction. Consequently, employment of
1.2 equiv of 3a and 1.4 equiv of BBr3 together with phosphazene
base 5 in CH2Cl2 furnished 2a with high ee, although the yield
was still low (entry 2). Using instead a stoichiometric amount
of Et3N as base gave a comparable yield but a slight decrease
in ee (entry 3), and it was shown that an excess of Et3N was
advantageous, producing 2a in 43% yield and with high ee (entry
4). The weaker bases K2CO3 and NaHCO3 did not promote the
rearrangement (entries 5 and 6). Surprisingly, the ee’s obtained
in the rearrangement were inconsistent, regardless of the amount
or type of base employed,18 and this irregularity was attributed
to the slight excess of BBr3 used in the preparation of Lewis
acid 4a.19 Because BBr3 is a stronger Lewis acid than 4a, amine
1a will preferentially complex to BBr3 and react through an
achiral pathway, thus lowering the obtained ee. Consequently,
by using an equimolar amount of BBr3, we obtained an excellent
was added 1a (1 equiv), and the resultant solution was stirred for 1 h. The
base was added, and the reaction mixture was stirred for 18-22 h, unless
b
otherwise stated. Conversion determined by HPLC. Unreacted 1a in
c
parentheses. Determined by HPLC analysis of the crude product on a
Chiracel OJ column (hexane/iPrOH). The absolute configuration was
d
established via chemical correlation; see ref 14. Reaction performed at
-20 °C. e Estimated from 1H NMR spectroscopy. f Reaction was run for 3
g
h
days. Reaction was run for 3 h. Reaction was run for 40 h.
A salient feature of the [2,3]-sigmatropic rearrangement is
that two new vicinal stereocenters can be created (C2′ and C3,
see Scheme 1). Furthermore, the olefin moiety and the anion-
stabilizing group provide possibilities for further derivatization,
thus turning the rearrangement into a useful tool in the
construction of complex organic compounds. An inherent
drawback in the [2,3]-sigmatropic rearrangement of ylides is
that the heteroatom moiety in the resultant product is fully
substituted, making further transformations of this moiety
problematic. Asymmetric variants of this rearrangement have
typically relied on intramolecular chirality transfer, either via
chirality transfer within the cyclic transition state or by auxiliary
control.1,2,12 An elegant example of the latter type is Sweeney’s
report of a highly diastereoselective rearrangement of various
glycine-derived allylic ammonium salts using Oppolzer’s sul-
tam.13 Although it would be desirable to use chiral catalysts in
these rearrangements, such strategies have been deemed prob-
lematic as the nitrogen atom must be quaternary, which seems
to exclude coordination of the substrate to a chiral Lewis acid.13
To address this issue, a [2,3]-sigmatropic rearrangement of
ammonium ylides was designed in which the ylid was generated
(14) Blid, J.; Panknin, O.; Somfai, P. J. Am. Chem. Soc. 2005, 127,
9352-9353.
(15) Blid, J.; Brandt, P.; Somfai, P. J. Org. Chem. 2004, 69, 3043-
3049.
(16) For prior Lewis acid mediated [2,3]-sigmatropic rearrangements of
allylic amines, see: (a) Murata, Y.; Nakai, K. Chem. Lett. 1990, 2069-
2072. (b) Kessar, S. V.; Singh, P.; Kaul, V. K.; Kumar, G. Tetrahedron
Lett. 1995, 36, 8481-8484. (c) Coldham, I.; Middleton, M. L.; Taylor, P.
L. J. Chem. Soc., Perkin Trans. 1998, 1, 2817-2821.
(17) Corey, E. J.; Imwinkelried, R.; Pikul, S.; Xiang, Y. B. J. Am. Chem.
Soc. 1989, 111, 5493-5495.
(18) The rearrangements were repeated but gave differing ee’s.
(19) Excess BBr3 was trapped as BBr4 anions during preparation of the
chiral Lewis acid.
(12) For recent publications, see: (a) Clark, J. S.; Hodgson, P. B.
Tetrahedron Lett. 1995, 36, 2519-2522. (b) Gulea-Purcarescu, M.; About-
Jadet, E.; Collignon, N.; Saquet, M.; Masson, S. Tetrahedron 1996, 52,
2075-2086. (c) Glaeske, K. W.; West, F. G. Org. Lett. 1999, 1, 31-33.
(d) Arbore´, A. P. A.; Cane-Honeysett, D. J.; Coldham, I.; Middleton, M.
L. Synlett 2000, 236-238. (e) Aggarwal, V. K.; Fang, G.; Charmant, J. P.
H.; Meek, G. Org. Lett. 2003, 5, 1757-1760.
(13) Workman, J. A.; Garrido, N. P.; Sancon, J.; Roberts, E.; Wessel,
H. P.; Sweeney, J. B. J. Am. Chem. Soc. 2005, 127, 1066-1067.
J. Org. Chem, Vol. 72, No. 4, 2007 1295

Blid et al.
TABLE 2. Asymmetric [2,3]-Sigmatropic Rearrangement of
Amines 1a
FIGURE 1. Structure of complex 6.
ee, although the conversion was still moderate (entry 7).
Likewise, employment of a large excess of BBr3 significantly
lowered the ee, although the conversion of 1a was complete
(entry 8). Although the ee of the rearrangement was gratifying,
the low conversion had to be addressed. This issue could be
solved by adding 2 equiv of the Lewis acid 4a to 1a in CH2Cl2
at room temperature, adding 2 equiv of base 5, and stirring the
resultant mixture overnight. These reaction conditions satisfac-
torily afforded 2a in 79% yield and with excellent ee (entry 9).
In the initial experiments, an unidentified byproduct could
yield
(%)b
ee anti/ee syn
(%)d
entry
amine
anti/sync
1
2
3
4
5
6
7
8
9
10
11
12
13
1b
1b
1b
1c
1d
1ej
1ej
1f
1g
1h
1i
1j
1k
2b 82
2b 46f
2b <10h
2c 85
2d 92
2e 57k
2e 65f,l
2f 70
2g 71
2h 52
7i 72m
2j 80
79:21
80:20
67:33
20:80
67:33
37:63
30:70
88:12i
29:71i
95:5i
-
96 (2R,3R):75 (2R,3S)e
96 (2R,3R):nde,g
nd
82 (2R,3R):98 (2R,3S)e
97:77i
92:93i
94:88i
93:ndg,i
98:99i
99:ndg,i
61
1
be observed via H NMR spectroscopy, which proved to be
compound 6, the complex between the product from the
rearrangement and the Lewis acid (Figure 1). When subjecting
the crude reaction mixture to saturated NaHCO3 followed by
chromatography, we could isolate complex 6 in 93% yield.
Recrystallization of this material followed by X-ray crystal-
lographic analysis established the structure of 6 and, conse-
quently, the absolute configuration of 2a (see Supporting
Information). The complex was shelf-stable for at least several
months.
Consequently, it was decided to run the reaction overnight
because it was not complete after 3 h (entry 10). Furthermore,
it was evident that an excess of base was required; applying 1
equiv of base gave a poor conversion of 1a (entry 11). Using
PhMe as solvent with the optimized reaction conditions gave
comparable results (entry 12), but Et2O and THF gave lower
yields and/or ee’s (entries 13 and 14). As expected, an excess
of Et3N could be employed as base without compromising yield
or ee (entry 15). By using the chiral Lewis acid 4b, derived
from the ligand 3b, we obtained good yields of ent-2a. However,
the ee’s were poor, and the product had the opposite absolute
configuration (entries 16 and 17).
-
-
99i
2k 64
96i
a
Reaction conditions: to a premade solution of 4a (2 equiv) at room
temperature was added 1 (1 equiv), and the resultant solution was stirred
for 1 h. Et3N (5 equiv) was added, and the reaction mixture was stirred for
b
18-22 h. Isolated yield of the diastereomeric mixture. Conversions in
italics. c Determined by 1H NMR spectroscopic analysis of the crude product.
d Determined by HPLC analysis of crude products; see ref 14. e The absolute
f
configuration was established via chemical correlation; see ref 14. The
g
h
reaction was performed at -20 °C. nd ) not determined. The reaction
i
was performed in PhMe. Stereochemistry assigned by analogy with the
j
k
rearrangements of 1b. E/Z 1:10. A byproduct was formed; see Table 3.
14% recovered 1e. See Table 3.
l
m
TABLE 3. [1,2]-Sigmatropic Rearrangement of (Z)-Olefins 1c,e,g,ia
With optimized rearrangement conditions at hand, the focus
was turned to examine the scope and limitations of the
asymmetric rearrangement as well as to establish a stereochem-
ical rationale accounting for the observed outcome. A series of
allylic amines were selected to reflect different electronic
properties and the steric bulk of the allyl moiety (Table 2). Four
different (E)/(Z)-olefin pairs (1b-i) were prepared as the anti/
syn selectivity obtained in the rearrangements of such pairs
might provide valuable information about steric and electronic
factors effecting the transition state. As an additional advantage,
silyl-substituted olefins 1h and 1i would furnish interesting
products containing an allylsilane moiety. The (E)-olefin 1b gave
a smooth rearrangement under the optimized conditions, af-
fording 82% of 2b in a 79:21 anti/syn ratio and an excellent ee
of the major diastereomer (Table 2, entry 1). An attempt to
improve the selectivity by performing the reaction at -20 °C
resulted in a low conversion (entry 2). Using PhMe as solvent
gave poorer diastereoselectivity and low conversion of 1b (entry
3). The isomeric (Z)-olefin 1c gave the opposite major diaste-
reomer, syn-2c, with a similar ee and an almost identical but
reversed anti/syn ratio (20:80, entry 4). Pursuing the asymmetric
rearrangement with the (E)/(Z)-isomeric pair of cinnamyl amines
1d,e and the benzyloxy methylene olefins 1f,g provided the
conversions (%)b
T
(°C)
entry
amine
7
2
1
2
3
4
5
1i (R ) SiMe2Ph)
1e (R ) Ph)
rt
∼100c
-
77
82
95
97
rt
-20
rt
20
7
1e (R ) Ph)
1c (R ) Me)
5
3
1g (R ) CH2OBn)
rt
a
b
Reaction conditions: see Table 2.
Determined by 1H NMR
spectroscopic analysis of the crude product. Yield 72%, 61% ee.
c
corresponding homoallylic amines in good yields and excellent
ee (entries 5-9). The diastereoselectivities showed the same
trend as that for 1b,c: the (E)-olefins gave the corresponding
anti homoallylic amines as major diastereomers, and the (Z)-
olefins afforded the corresponding syn isomers. Moreover,
compound 1e had to be rearranged at -20 °C to suppress the
formation of a byproduct (see Table 3), which resulted in slightly
better diastereoselectivity than at room temperature (entries 6
and 7).
Although the asymmetric rearrangement of the (E)-olefin 1h
provided essentially one diastereomer in 99% ee and 52% yield
1296 J. Org. Chem., Vol. 72, No. 4, 2007

[2,3]-Sigmatropic Rearrangement of Allylic Amines
SCHEME 3. Asymmetric Rearrangement of 1la
SCHEME 5. Attempted Asymmetric Rearrangement of 10
and 12a
a For reaction conditions, see Table 1, entry 15.
SCHEME 4 Attempt to Rearrange the
1,2,5,6-Tetrahydropyridine 8a
a For reaction conditions, see Table 1, entry 15.
material and ligand 3a were recovered. This can be rationalized
by examining structure 9, the deprotonated complex between 8
and 4a. [2,3]-Sigmatropic rearrangement of 9 would impose a
considerable amount of strain that effectively prevents the
desired bond formation.
a For reaction conditions, see Table 1, entry 15.
(entry 10), the isomeric (Z)-olefin 1i failed to give the
corresponding homoallylic amine, and instead, compound 7i,
the product of a [1,2]-sigmatropic rearrangement, was formed
in 72% yield and 61% ee (entry 11 and Table 3, entry 1).
Attempts to suppress the [1,2]-pathway by performing the
reaction at -20 °C did not affect the outcome.20 Finally,
rearrangement of amines 1j,k afforded the homoallylic amines
2j,k in 99% and 96% ee, respectively (Table 2, entries 12 and
13).
To further extend the scope of the asymmetric rearrangement,
(Z)-olefin 1l containing a dimethylamide moiety was prepared.
Subjecting this material to the optimized rearrangement condi-
tions furnished a mixture of 2l (62%), starting material (3%),
and benzylic [1,2]-shift product 7l (17%) (Scheme 3). As
expected, the syn diastereomer was the major isomer under these
conditions (anti/syn 28:72 for the crude product). After purifica-
tion, an anti/syn ratio of 26:74 with at least 97% ee for the syn
diastereomer was obtained.
In the rearrangement of (Z)-olefins 1c,e,g minor amounts of
the corresponding [1,2]-products could be detected, although
the major products in all cases were derived from the [2,3]-
rearrangement (Table 3, entries 2-5).21 With 1e, the [1,2]-
rearrangement became a concern when the reaction was
performed at room temperature (entry 2), but by lowering the
reaction temperature to -20 °C, the [2,3]-sigmatropic rear-
rangement prevailed (entry 3). The (E)-olefins 1b,d,f,h gave
insignificant amounts of the corresponding [1,2]-rearranged
products.
By introduction of a substituent at the R-carbon of the
rearrangement substrate, homoallylic amines with a quaternary
stereogenic center should be obtained. Initial attempts to realize
this focused on rac- and (S)-N-allyl-N-benzyl alanine derivatives
10. However, subjecting 10 to the standard rearrangement
conditions gave only 11, the product of a benzylic [1,2]-shift,
and recovered 10 (Scheme 5). Instead, treatment of diallyl
derivatives rac- and (S)-12 under the optimized rearrangement
conditions provided the R-allyl methyl alanines 13 in 59% and
78% yields and excellent enantioselectivities, respectively. It
should be noted that rearrangement of the allyl group via either
a [1,2]- or a [2,3]-rearrangement would, in this case, provide
identical products.
Reaction of (E)-1 can give complexes A and A′, which, when
subjected to base, would afford B and B′, respectively (Scheme
6). An equilibration between B and B′ can be envisioned,
possibly assisted by Br- or a base, and similar processes have
been suggested previously.23 As a result, the stereochemical
outcome will be determined by the energy difference between
transition states C and C′. In C′, an N-Ts moiety efficiently
blocks the si-face of the enolate and thus interferes with the
rearrangement, and no such interactions are present in C. Thus,
the rearrangement will be channeled through transition state C
affording (2R,3R)-2, which is in concord with the experimental
findings.
In this model, the absolute configuration at C2′ is established
via re- or si-face selectivity in the C/C′ transition states, whereas
the C2′/C3 relative stereochemistry is generated by the exo/
endo orientation of the allyl moiety (Scheme 7). The rearrange-
ment of (E)-olefins afforded the anti products as the major
diastereomers, and consequently, the reaction proceeds via the
exo-C transition state. In exo-C, the allyl moiety experiences
less steric interactions compared to endo-C. The increased steric
bulk of the R substituent enhances the diastereoselectivity, as
both the R substituent and the allylic C2 carbon are oriented
away from the bulky interior of the oxazaborocycle.
The rearrangement conditions were also attempted on the
N-substituted 1,2,5,6-tetrahydropyridine 8, which would give
an efficient entry to substituted pyrrolidines (Scheme 4).11,22
Unfortunately, no product could be detected and only the starting
(20) (a) Jemison, R. W.; Laird, T.; Ollis, W. D.; Sutherland, I. O. J.
Chem. Soc., Perkin Trans. 1980, 1, 1450-1457. (b) Jemison, R. W.; Laird,
T.; Ollis, W. D.; Sutherland, I. O. J. Chem. Soc., Perkin Trans. 1980, 1,
1436-1449.
(21) The structures of [1,2]-rearranged 1c,e,g were confirmed by H,H-
COSY.
(22) (a) Burns, B.; Coates, B.; Neeson, S. J.; Stevenson, P. J. Tetrahedron
Lett. 1990, 31, 4351-4354. (b) Sweeney, J. B.; Tavassoli, A.; Carter, N.
B.; Hayes, J. F. Tetrahedron 2002, 58, 10113-10126. (c) Roberts, E.;
Sancon, J. P.; Sweeney, J. B.; Workman, J. A. Org. Lett. 2003, 5, 4775-
4777. (d) Roberts, E.; Sancon, J.; Sweeney, J. B. Org. Lett. 2005, 7, 2075-
2078.
The rearrangements of the (Z)-olefins afforded lower and
opposite diastereoselectivities than the corresponding (E)-olefins.
(23) Vedejs, E.; Fields, S. C.; Lin, S.; Schrimpf, M. R. J. Org. Chem.
1995, 60, 3028-3034. Equilibration before deprotonation and rearrangement
can also be envisioned.
J. Org. Chem, Vol. 72, No. 4, 2007 1297

Blid et al.
SCHEME 6. Kinetically Controlled Stereoselectivitya
SCHEME 8. Thermodynamically Controlled Reaction
Pathwaya
a Only rearrangement from the exo transition state considered, and the
charges are omitted for clarity. For structures B and B′, see Scheme 6.
when the R moiety becomes larger. With 1i (R ) SiMe2Ph),
severe steric interactions in both the exo and the endo transition
result in a preferential reaction through a benzylic [1,2]-
rearrangement (see Table 3).
a The charges are omitted in C and C′ for clarity.
SCHEME 7. Proposed Transition States in the Kinetically
Controlled Rearrangement of 1a
The stereochemical outcome can also be rationalized by
diastereoselective complexation of Lewis acid 4a to 1, generat-
ing complex A. However, this kinetically controlled scenario
appears less likely, as the benzyl and allyl groups are sterically
and electronically relatively similar.
Alternatively, the stereoselectivities observed in the rear-
rangements could be under thermodynamic control. In this case,
the focus would be on the rearranged complexes D and D′
(Scheme 8). The basic conditions applied in the rearrangement
can lead to deprotonation of D and D′, thus establishing
equilibriums between D and E and between D′ and E′,
respectively. Because of unfavorable steric interactions with the
N-Ts moiety, complexes D′ and E will be destabilized compared
to D and E′. Thus, upon acidic hydrolysis, complex D will give
the observed major product (2R,3R)-anti-2, whereas the ther-
modynamically less stable complex E will provide the minor
product (2S,3R)-syn-2. Likewise, complex E′ gives (2R,3S)-
syn-2 and D′ yields minor product (2S,3S)-anti-2. Although this
rationale concerns rearrangement of B and B′ via an exo
transition state, the analogous event via an endo transition state
will merely give the opposite product outcome (D will instead
provide (2R,3S)-syn-2, whereas E′ will give (2R,3R)-anti-2, i.e.,
the experimentally observed products). In this scenario, the
absolute configuration at C2′ is established under thermody-
namic control during a base-induced equilibrium. The C2′/C3
relative stereochemistry is established by the relative rate of
formation of complexes D/D′ and the exo/endo preference in
the transition states.
a The charges are omitted for clarity.
This observation can be explained by the fact that both the exo
and endo transition states experience unfavorable steric interac-
tions (Scheme 7). In the exo transition state, unfavorable steric
interactions between the R substituent and the bulky interior of
the oxazaborocycle arise, whereas the endo transition state is
destabilized by interaction of the allylic C2 carbon with the
interior of the oxazaborocycle. There is a slight preference for
the exo transition state, however, though this seems to decrease
To investigate this, DFT calculations were performed on
complexes D (R ) H, same as E′) and D′ (R ) H, same as
E).24 These calculations established that complex D (R ) H) is
about 10 kcal/mol more stable than D′ (R ) H), which is in
(24) For computational details, see ref 14.
1298 J. Org. Chem., Vol. 72, No. 4, 2007

[2,3]-Sigmatropic Rearrangement of Allylic Amines
arising from the rearranged complexes D and E′. This signal is
in agreement with the 11B NMR spectrum of complex 6.
Moreover, by inspection of the 1H integrals, it could be
established that the relative amount between the complexes and
amine 1b remained constant during the acquisitions of NMR
2-6.
Our rationalization of the data from the NMR spectroscopic
investigation is as follows. Complexes B and B′ existed in a
slow equilibrium in a 2:1 relationship (NMR time scale, NMR
2-4).23 Nevertheless, in the kinetic scenario, only complex B
rearranged, either via the exo- or endo-C transition state, to
provide an 80:20 diastereomeric ratio of rearranged complexes
D and E′. After the rearrangement was complete (NMR 5),
acidic hydrolysis and workup provided (2R,3R)-2b/(2R,3S)-2b
in an 80:20 diastereoselectivity and >97 and 82% ee, respec-
tively, together with recovered starting material (NMR 6).
Consequently, the NMR investigation suggests that a kinetic
pathway such as the one described may well be responsible for
the stereochemical outcome.
FIGURE 2. 1H NMR spectroscopic investigation of the rearrangement
(δ 5.4-6.4 ppm). (1) Complex(es) formed by mixing 1b (1 equiv) and
chiral Lewis acid 4a (2 equiv). (2) After addition of Et3N (>2 equiv).
(3) After 90 min. (4) After 2 h. (5) After 5-6 h. (6) After acidic
hydrolysis and workup.
Conclusions
The first asymmetric [2,3]-sigmatropic rearrangement of
achiral allylic amines has been realized by employing the chiral
Lewis acid 4a. The scope of the rearrangement was demon-
strated on several amines, affording the corresponding homoal-
lylic secondary amines 2 in excellent ee’s and good diastere-
oselectivities. A study of the diastereoselectivity was provided
by four different (E)/(Z)-olefinic pairs, suggesting an exo
preference in the transition state of the rearrangement. The
obtained stereoselectivities were rationalized by invoking either
a kinetic or a thermodynamic pathway. The kinetic pathway
was favored in light of the results from a D2O-quenching
experiment of the isolated rearranged complex 6 and a NMR
spectroscopic investigation of the process.
agreement with the observed reaction outcome. However, a
deuterium exchange experiment conducted on complex 6 (Et3N,
D2O, CH2Cl2) resulted in no deuterium incorporation, and
consequently, the kinetic process seems more likely.
With the objective to clarify the mechanistic details of the
rearrangement, an NMR spectroscopic investigation of the
complex formed between 4a and 1b along with the formation
of the anticipated deprotonated and rearranged complexes B/B′
and exo-D/endo-D, respectively, was initiated (Figure 2). Thus,
the complex between 4a and 1b was formed as previously
1
described and monitored by H and 11B NMR spectroscopy.
By the upfield change in the chemical shift of the signals from
the diastereotopic N-allyl and N-benzyl protons of 1b, it
appeared that 1b was converted into complexes A and/or A′.
From the 1H NMR spectroscopic data, it was difficult to
establish whether a mixture of the complexes formed or whether
an equilibrium between them existed (Figure 2, NMR 1). 11B
NMR revealed a broad singlet at δ 10.4 ppm and two broader
signals at 23.5 and 27.3 ppm, respectively. The upfield boron
signal most likely arose from the complexes A/A′, and the latter
two probably originated from residual chiral Lewis acid. To
Experimental Section
Representative Experimental Description for the Asymmetric
[2,3]-Sigmatropic Rearrangement of Amines 1. (2R,3R)-2-
(Benzylamino)-3-((benzyloxy)methyl)-1-(pyrrolidin-1-yl)pent-4-
en-1-one (anti-2d). To a solution of the ligand 3a (115 mg, 0.22
mmol) in CH2Cl2 (2 mL) was added BBr3 (0.22 mL, 1 M in CH2-
Cl2), and the resultant mixture was stirred for 1.5 h at room
temperature. The solvent was removed in vacuo, and fresh CH2Cl2
(2 × 2 mL) was added and removed in vacuo to provide an off-
white solid.25 To the solid was transferred 1f (42 mg, 0.11 mmol)
in CH2Cl2 (2 mL), and the resultant solution was stirred for 1 h
before Et3N (80 µL, 0.56 mmol) was added. After 20 h, the reaction
mixture was treated with conc. HCl/MeOH (1:5, 2 mL) overnight
at room temperature.26 To the cooled solution was added 2 M NaOH
(1 mL) and saturated NaHCO3 (5 mL), and the water phases were
extracted with CH2Cl2 (2 × 5 mL). The combined organic phases
were dried (K2CO3) and concentrated, and the residue was triturated
with Et2O (3 × 1 mL) to provide the ligand 3a (100 mg, 87%) as
a solid. The Etheral triturate was concentrated and purified by
1
that solution was added >2 equiv of Et3N, and the H NMR
spectrum was instantly recorded (NMR 2). Two sets of signals
at approximately δ 6.0 ppm, the vinylic protons in B and B′,
had emerged along with a broad signal at δ 5.6 ppm attributed
to starting material 1b. How 1b formed is unclear, but the
relative integral stayed constant during the experiment (NMR
3-6). The signals corresponding to B and B′ had a 2:1 ratio
throughout the NMR experiment and were slowly diminishing
relative to the other signals and disappeared within 5-6 h. In
the 11B NMR spectrum, the signal at δ 10.4 ppm had endured,
but now only one broad signal appeared at δ 22.8 ppm. While
the signals in the 1H NMR spectrum corresponding to complexes
B/B′ diminished, two new signals at δ 6.2 and δ 6.3 ppm
appeared and were interpreted as allylic C2 protons of the
rearranged complexes D and E′, respectively (NMR 3). The
ratio between the two signals was 4:1 and remained constant
during the experiment (NMR 3-5). During this time, the signal
at δ 10.4 ppm in the 11B NMR spectrum also diminished and
was replaced by a signal at approximately δ 9 ppm, probably
i
chromatography (acetone/pentane 0:1f1:2 + 0.3% PrNH2) to
provide 2f (29.2 mg, 70%) as a diastereomeric mixture (anti/syn
1
88:12) and the remaining ligand 3a (14.5 mg, 13%). H NMR
(CDCl3, 400 MHz) δH 7.31-7.18 (m, 10H), 5.83 (ddd, J ) 17.4,
10.5, 8.1 Hz, 1H), 5.15 (d, J ) 10.5 Hz, 1H), 5.07 (d, J ) 17.4
Hz, 1H), 4.41 (s, 2H), 3.83 (d, J ) 13.3 Hz, 1H), 3.67 (dd, J )
(25) This procedure was necessary to remove all HBr from the reaction
vessel.
(26) The reaction time could be shortened to 1-2 h by heating at 60 °C.
J. Org. Chem, Vol. 72, No. 4, 2007 1299

Blid et al.
8.9, 7.8 Hz, 1H), 3.61 (d, J ) 4.8 Hz, 1H), 3.52 (d, J ) 13.3 Hz,
1H), 3.52 (m, 2H), 3.40 (m, 2H), 3.16 (m, 1H), 2.55 (m, 1H), 2.22
(bs, 1H), 1.81 (m, 4H). 13C NMR (CDCl3, 100 MHz) δC 172.4,
140.4, 138.4, 135.2, 128.32, 128.25, 128.2, 127.48, 127.46, 126.8,
117.5, 73.2, 71.0, 58.6, 52.3, 46.6, 45.9, 45.6, 26.1, 24.1. IR (neat)
3323, 3030, 2873, 1637, 1452, 1101 cm-1. HRMS (FAB+) calcd
lenberg foundation. We are grateful to Dr. T. Privalov for ab
initio calculations and to Dr. Andreas Fischer for X-ray structure
determination.
Supporting Information Available: Experimental procedures
and spectral data characterizations of all new compounds. X-ray
data of compound 6 in CIF format. This material is available free
of charge via the Internet at http://pubs.acs.org.
20
for C24H31N2O2 [M + H]+: 379.2386. Found: 379.2387. [R]D
+15.4 (c ) 0.49, CH2Cl2).
Acknowledgment. This work was supported financially by
the Swedish Research Council and the Knut and Alice Wal-
JO062178V
1300 J. Org. Chem., Vol. 72, No. 4, 2007