Lewis acid mediated [1,2]-rearrangement of ammonium ylides

Article abstract of DOI:10.1016/j.tetlet.2007.04.151

The first example of Lewis acid mediated [1,2]-rearrangement of various glycine derivatives has been developed. A brief study of steric and electronic properties of the migrating group is presented. The corresponding amides were obtained in good yields an

Full text of DOI:10.1016/j.tetlet.2007.04.151

Tetrahedron Letters 48 (2007) 4947–4949
Lewis acid mediated [1,2]-rearrangement of ammonium ylides
Pavel Tuzina and Peter Somfai*
Organic Chemistry, KTH Chemical Science and Engineering, SE-100 44 Stockholm, Sweden
Received 24 March 2007; revised 15 April 2007; accepted 26 April 2007
Available online 10 May 2007
Abstract—The first example of Lewis acid mediated [1,2]-rearrangement of various glycine derivatives has been developed. A brief
study of steric and electronic properties of the migrating group is presented. The corresponding amides were obtained in good yields
and in the case of substrate 4d, moderate diastereoselectivity was observed.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The [1,2] Stevens rearrangement of ammonium ylides is
a useful transformation, since it serves as a tool for con-
verting readily accessible C–N bonds into new C–C
bonds, and it has been used extensively in organic syn-
thesis.1 However, its asymmetric versions still remain a
challenge, probably reflecting the radical character of
the process.2 The rearrangement proceeds via a homo-
lytic cleavage of the carbon–nitrogen bond, generating
a radical pair, which is held tightly together by a solvent
cage. This is then followed by rapid recombination of
the radicals thus providing the product of the [1,2]-rear-
rangement.1 It has been shown, however, that for cyclic
ammonium ylides, having a stereogenic nitrogen nu-
cleus, high levels of stereoselectivity can be obtained.3,4
acid 2, which is sterically more demanding than BBr3,
increased amounts of the products derived from a
[1,2]-shift were sometimes obtained (Scheme 1). Indeed,
attempts at the [2,3]-rearrangement of allylic amine 1
(R = SiMe2Ph), having a bulky vinyl substituent, gave
only 3 in good yield and moderate enantioselectivity,
the product of a [1,2]-shift.6,9 These observations war-
ranted a more thorough study of the Lewis acid medi-
ated [1,2]-rearrangements of various glycine derivatives
and also an investigation of the possibilities of develop-
ing an asymmetric process. Herein, we report our preli-
minary findings.
In order to favor the [1,2]-shift over the [2,3] reaction
manifold, model substrate 4a, in which only the benzyl
group is capable of N!C migration, was employed
(Scheme 2).10 Based on the previous results,5 it was envi-
sioned that mixing 4a with BBr3 would result in the for-
mation of oxazaborolidine 5. Subsequent deprotonation
would then yield ylide 6, which could participate in a
[1,2]-rearrangement, ultimately yielding secondary
amine 7a (Scheme 2).
We have recently demonstrated that activation of N-
allyl-N-Bn amines derived from glycine with BBr3, fol-
lowed by deprotonation of the complex formed in situ,
results in the formation of the corresponding homoally-
lic amines, the products of [2,3]-sigmatropic rearrange-
ment.5 It was also shown that, when using a chiral
boron Lewis acid, this is easily transformed into an
asymmetric process.6 It is known that the [2,3]-rear-
rangement can be accompanied by formation of the
products derived from a [1,2]-rearrangement, and this
can be particularly pronounced when elevated reaction
temperatures are used or in cases where unfavorable ste-
ric interactions prevent the desired reaction pathway.7,8
In line with this, it was observed that the asymmetric
[2,3]-rearrangement of 1 (R = H), using chiral Lewis
Initial attempts were directed toward developing opti-
mal reaction conditions that would effect the rearrange-
ment of compound 4a. In the first attempt, BBr3 and the
non-ionic base Et3N were employed, since such
conditions were successfully used in our previous inves-
tigation. Thus, when subjecting amine 4a to BBr3
(1.2 equiv) and Et3N (5 equiv),5,11 secondary amine 7a
was obtained in low yield together with starting material
(Table 1, entry 1). Increasing the amount of BBr3
(2 equiv) resulted in a somewhat higher yield but a long
reaction time was still required, indicating that the
homolysis of the carbon–nitrogen is slow under these
Keywords: [1,2]-Rearrangement; Ammonium ylide; Lewis acid; Micro-
wave irradiation; Amide.
*
Corresponding author. Tel.: +46 8 790 6960; fax: +46 8 791 2333;
e-mail: somfai@kth.se
0040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.04.151

4948
P. Tuzina, P. Somfai / Tetrahedron Letters 48 (2007) 4947–4949
O
Bn
N
O
Ph
Ph
H
N
1)
2, CH2Cl2
N
N
2) Et3N, rt
3) HCl:MeOH (1:5)
TsN
NTs
B
SiPhMe2
R
Ph
Br
2
1
3
72%, 61% ee
Scheme 1. [1,2]-Rearrangement of allylic amine 1.
-
-
BBr4
BBr4
Ph
N
N
O
1) [1,2]
2) hydrolysis
BBr3
Et3N
Ph
Ph
N
N
N
N
H
Ph
O
N
N
B
B
O
O
Br Br
Br Br
4a
5
6
7a
Scheme 2. The Lewis acid mediated [1,2]-rearrangement of 4a.
Table 1. Optimization of the reaction conditions for [1,2]-rearrange-
ment of 4aa
of the migrating group was briefly studied. Substrates
4b–d were selected and their preparation is summarized
in Scheme 3.9,12
Entry Lewis acid Equiv Temp Reaction Yieldb 7a (%)
(ꢁC)
time (h)
1
2
3
4
5
BBr3
BBr3
BBr3
BF3ÆEt2O
BBr3
1.2
2
2
2
1.2
rt
rt
60c
60c
60c
48
24
1
1
1
39 (30)
48 (32)
52 (0)
R2
N
R2
Br
O
NH
N
N
n.r.d (96)
60 (12)
R1
O
K2CO3, CH3CN, o/n
72-86%
R1
a Reaction conditions: To 4a (1.0 equiv) in CH2Cl2 at 0 ꢁC was added
BBr3 (1.2 equiv) and Et3N (5 equiv) and the mixture was stirred at
the indicated temperature for the time given, see Ref. 13.
b Isolated yield. Yield of recovered starting material in parenthesis.
c Microwave-assisted heating.
4a R1= H, R2= H
b R1= -Bu, R2= H
t
c R1= F, R2= H
d R1= H, R2= Me
Scheme 3. Synthesis of rearrangement precursors 4a–d.
d n.r. = no reaction.
Rearrangement of glycine derivatives 4b–d provided the
corresponding products 7b–d in moderate yields
(Scheme 4, Table 2); steric and electronic effects do
not appear to have any dramatic impact on the reaction
outcome.13,14 It is also noted that for substrate 4d, which
has an a-methylbenzyl moiety, product 7d was obtained
as a 2:1 mixture of diastereomers.
reaction conditions (entry 2). The reaction time could be
decreased by heating the reaction mixture to 60 ꢁC for
1 h in a microwave reactor, providing 7a in 52% yield
(entry 3). Interestingly, even though only a moderate
yield was achieved, no starting material could be recov-
ered, suggesting that either 7a (in case of full conversion)
or starting material 4a decomposed during the reaction.
Attempts to promote the [1,2]-shift by using the weaker
Lewis acid BF3ÆEt2O resulted in no rearrangement and
only starting material was recovered (entry 4). This is
surprising since we have previously shown that BF3Æ
Et2O promotes the formation of oxazaborolidines corre-
sponding to 5, and the reasons why the subsequent C–N
bond homolysis is unfavored are not clear.11 Finally,
using BBr3 (1.2 equiv) together with Et3N (5 equiv)
and heating to 60 ꢁC in a microwave reactor for 1 h
proved to be superior and gave 7a in 60% yield together
with recovered starting material (12%, entry 5). It
should be noted that the yield obtained is in accordance
with the proposed reaction pathway (Scheme 2). When
using 1.2 equiv of BBr3 a maximum of 60% of 4a can
be converted into oxazaborolidine 5, the formation of
which is a prerequisite for the subsequent [1,2]-shift.
Somewhat surprisingly, using larger amounts of BBr3
(2 equiv), in an attempt to promote a complete conver-
sion of 4a into 5, resulted in only a low yield of product
7a.
R1
R2
R2
1) BBr3
N
N
2) NEt3, 60 ˚C, MW
3) HCl:MeOH
N
O
HN
R1
O
7a R1= H, R2= H
4a R1= H, R2= H
b R1= -Bu, R2= H
t
b R1= -Bu, R2= H
t
c R1= F, R2= H
c R1= F, R2= H
d R1= H, R2= Me
d R1= H, R2= Me
Scheme 4. Lewis acid mediated [1,2]-rearrangement of 4a–d.
Table 2. Lewis acid mediated [1,2]-rearrangement of amines 4a
Entry
Substrate
Product
Yieldb (%)
1
2
3
4
4a
4b
4c
4d
7a
7b
7c
7d
60
46
51
60c
a For reaction conditions, see Ref. 13.
b Isolated yield.
c Isolated as 2:1 mixture of diastereomers.
With the optimized reaction conditions in hand, the
influence of varying the steric and electronic properties

P. Tuzina, P. Somfai / Tetrahedron Letters 48 (2007) 4947–4949
4949
6. Blid, J.; Panknin, O.; Somfai, P. J. Am. Chem. Soc. 2005,
127, 9352–9353.
7. Jemison, R. W.; Laird, T.; Ollis, W. D.; Sutherland, I. O.
J. Chem. Soc., Perkin Trans. 1 1980, 1436–1449.
8. Jemison, R. W.; Laird, T.; Ollis, W. D.; Sutherland, I. O.
J. Chem. Soc., Perkin Trans. 1 1980, 1450–1457.
9. Blid, J.; Panknin, O.; Tuzina, P.; Somfai, P. J. Org. Chem.
2007, 72, 1294–1300.
In conclusion, we have presented the first example of a
Lewis acid mediated [1,2]-rearrangement of various gly-
cine derivatives. The effect when changing the steric and
electronic properties of the migrating group was briefly
studied and these parameters do not seem to have any
appreciable effect on the reaction outcome. Develop-
ment of an asymmetric version of the Lewis acid medi-
ated [1,2]-rearrangement is currently underway in our
laboratory.
10. Formation of the most stable potential carbon-centered
radical is favored.
11. Blid, J.; Somfai, P. Tetrahedron Lett. 2003, 44, 3159–3162.
12. N-Methyl-p-t-butylbenzylamine was prepared according
to the procedure by Pigge, F. C.; Coniglio, J. J.; Fang, S.
Organometallics 2002, 21, 4505–4512.
Acknowledgments
This work was supported financially by the Swedish
Research Council and the Knut and Alice Wallenberg
foundation.
13. Typical procedure: To 4b (113.8 mg, 0.434 mmol) in dry
CH2Cl2 (2 mL) at 0 ꢁC was added BBr3 (442 lL,
0.520 mmol, 1.178 M solution in CH2Cl2). The resultant
mixture was stirred for 1.5 h (0 ꢁC!rt) and then Et3N
(302 lL, 2.165 mmol) was added and the mixture was
heated to 60 ꢁC in a microwave reactor for 1 h. After
cooling to rt, MeOH–HCl (2 mL, 5:1) was added and the
resultant mixture was stirred overnight. The mixture was
then basified by addition of 2 M NaOH (2 mL) and aq
NaHCO3 (2 mL). The mixture was then extracted with
CH2Cl2 (3 · 10 mL), the organic phases dried (K2CO3),
and the resultant crude material purified by chromato-
graphy (SiO2, 40:1 CH2Cl2–MeOH with 0.5% of i-PrNH2)
to provide 7b (52.3 mg, 46%). 1H NMR (CDCl3,
500 MHz): dH 7.29 (d, 2H, J = 8.2 Hz), 7.12 (d, 2H,
J = 8.1 Hz), 3.65 (dd, 1H, J = 6.0, 8.5 Hz), 2.91 (dd, 1H,
J = 5.9, 13.3 Hz), 2.90 (s, 3H), 2.77 (dd, 1H, J = 8.6,
13.2 Hz), 2.54 (s, 3H), 2.31 (s, 3H), 1.93 (s, 1H), 1.31 (s,
9H); 13C NMR (CDCl3, 125 MHz): dC 174.2, 149.4, 134.7,
128.7, 125.0, 61.2, 39.8, 36.3, 35.4, 34.6, 31.3.
References and notes
1. For reviews see: (a) Vanecko, J. A.; Wan, H.; West, F. G.
Tetrahedron 2006, 62, 1043–1062; (b) Pine, S. H. Org.
React. (NY) 1970, 18, 403–464; (c) Lepley, A. R.;
Giumanini, A. G. In Mechanisms of Molecular Migrations;
Thygarajan, B. S., Ed.; Interscience: New York, 1971; Vol.
´
3, pp 297–440; (d) Marko, I. E. In Comprehensive Organic
Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon:
Oxford, 1991; Vol. 3, pp 913–973.
2. Ollis, W. D.; Rey, M.; Sutherland, I. O. J. Chem. Soc.,
Perkin Trans. 1 1983, 1009–1027.
3. Glaeske, K. W.; West, F. G. Org. Lett. 1999, 1, 31–33.
4. Tayama, E.; Nanbara, S.; Nakai, T. Chem. Lett. 2006, 35,
478–479.
14. All new compounds showed spectroscopic data (1H, 13C,
NMR, IR, HRMS) in accordance with their structure.
5. Blid, J.; Brandt, P.; Somfai, P. J. Org. Chem. 2004, 69,
3043–3049.