Lewis acid catalyzed [1,3]-sigmatropic rearrangement of vinyl aziridines

Article abstract of DOI:10.1021/ol802123e

(Chemical Equation Presented) This paper details the copper-catalyzed ring expansion of vinyl aziridines to 3-pyrrolines. Broad substrate scope (24 examples) using tosyl-and phthalimide-protected vinyl aziridine substrates is observed. Cu(hfacac)₂

Full text of DOI:10.1021/ol802123e

ORGANIC
LETTERS
2008
Vol. 10, No. 21
5023-5026
Lewis Acid Catalyzed [1,3]-Sigmatropic
Rearrangement of Vinyl Aziridines
Matthew Brichacek, DongEun Lee, and Jon T. Njardarson*
Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell
UniVersity, Ithaca, New York 14853-1301
jn96@cornell.edu
Received September 10, 2008
ABSTRACT
This paper details the copper-catalyzed ring expansion of vinyl aziridines to 3-pyrrolines. Broad substrate scope (24 examples) using tosyl-
and phthalimide-protected vinyl aziridine substrates is observed. Cu(hfacac)2 was determined to be superior to all other catalysts tested.
3-Pyrrolines (2,5-dihydropyrroles), while important in their
own right, are synthetically versatile heterocycles which
allow ready access to pyrrolidines and pyrroles.1 Collectively,
these ring systems occur in countless natural products and
pharmaceuticals.2 Not surprisingly, a number of different
synthetic approaches toward pyrrolines have been developed,
but the most popular are Birch reduction of pyrroles,3 [3 +
2] cyclization,4 and ring-closing metathesis.5 Despite these
and other creative approaches, a requisite exists for a simple
and selective method which allows access to structurally
complex 3-pyrrolines. Herein, we demonstrate that vinyl
aziridines can be efficiently converted to 3-pyrrolines using
commercially available copper(II) catalysts.
observed in the late 1960s.7 Harsh thermal conditions were
employed, and little substrate scope was observed. Since
these initial experiments, several attempts have been made
to investigate the reaction further, but limitations such as
competing side reactions and the necessity of additional
functionality were found.8 Therefore, applications of this
rearrangement to the synthesis of complex molecules have
been lacking.9,10 Clearly, there is a need to develop condi-
tions that expand the scope of the rearrangement and allow
functional group compatibility.
Vinyl aziridines are most easily synthesized by one of three
methods: the addition of a nitrene to a diene, the addition of
an allylic ylide to an imine, or the cyclization of unsaturated
amino alcohols.11 Each method has limitations often directly
related to the nitrogen protecting group. This paper focuses
Analogous to the vinyl cyclopropane rearrangement,6 the
thermolysis of vinyl aziridines to 3-pyrrolines was first
(1) (a) Walker, D.; Hiebert, J. D. Chem. ReV. 1967, 67, 153. (b) Jones,
R. A.; Marriott, T. P.; Rosenthal, W. P.; Arques, J. S. J. Org. Chem. 1980,
45, 4515. (c) Shim, Y. K.; Youn, J. I.; Chun, J. S.; Park, T. H.; Kim, M. H.;
Kim, W. J. Synthesis 1990, 753.
(7) (a) Atkinson, R. S.; Rees, C. W. Chem. Commun. 1967, 1230. (b)
Scheiner, P. J. Org. Chem. 1967, 32, 2628. (c) Logothetis, A. L. J. Am.
Chem. Soc. 1965, 87, 749.
(8) (a) Somfai, P.; Ahman, J. Targets Heterocycl. Syst. 1999, 3, 341.
(b) Mente, P. G.; Heine, H. W.; Scharoubim, G. R. J. Org. Chem. 1968,
33, 4547. (c) Hoesch, L.; Dreiding, A. S. Chimia 1972, 26, 629. (d) Fugami,
K.; Morizawa, Y.; Oshima, K.; Nozaki, H. Tetrahedron Lett. 1985, 26, 857.
(2) (a) Pinder, A. R. Nat. Prod. Rep. 1989, 6, 67. (b) Le Quesne, P. W.;
Dong, Y; Blythe, T. A. Alkaloids: Chem. Bio. Persp. 1999, 13, 237. (c)
Kidwai, M.; Venkataramanan, R.; Mohan, R.; Sapra, P. Curr. Med. Chem.
2002, 9, 1209. (d) Higashio, Y.; Shoji, T. Appl. Catal. A 2004, 260, 251.
(3) Donohoe, T. J.; Guyo, P. M. J. Org. Chem. 1996, 61, 7664.
(4) (a) Huisgen, R.; Scheer, W.; Huber, H. J. Am. Chem. Soc. 1967, 89,
1753. (b) Coldham, I.; Hufton, R. Chem. ReV. 2005, 105, 2765. (c) Pandey,
G.; Banerjee, P.; Gadre, S. Chem. ReV. 2006, 106, 4484.
(9) Hirner, S.; Somfai, P. Synlett 2005, 3099
.
(10) (a) Hudlicky, T.; Frazier, J. O.; Seoane, G.; Tiedje, M.; Seoane,
A.; Kwart, L. D.; Beal, C. J. Am. Chem. Soc. 1986, 108, 3755. (b) Pearson,
W. H.; Bergmeier, S. C.; Degan, S.; Lin, K. C.; Poon, Y. F.; Schkeryantz,
J. M.; Williams, J. P. J. Org. Chem. 1990, 55, 5719
.
(5) Fu, G. C.; Grubbs, R. H. J. Am. Chem. Soc. 1992, 114, 7324.
(6) Baldwin, J. E. Chem. ReV. 2003, 103, 1197.
(11) Ohno, H., In Aziridines and Epoxides in Organic Synthesis; Yudin,
A. K., Eds.; Wiley-VCH: Weinheim, Germany, 2006; p 37.
10.1021/ol802123e CCC: $40.75
Published on Web 10/09/2008
2008 American Chemical Society

Table 1. Copper-Catalyzed Ring Expansion of Vinyl Aziridinesa
Table 2. 3-Pyrroline Synthesisa
entry
catalyst
time (h)
yieldb (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
none
CuI
16
16
16
16
16
16
16
16
16
16
16
16
16
11
7.5
1d
0c
0
0
0
0
0
10
0
35
20
0
Cu[(S,S)-Ph-box]Cl2
Cu[(S,S)-Ph-box]OTf2
Cu(acac)2
Cu(tfacac)2
Cu(dbm)2
Cu(ptfm)2
Cu(accy)2
Cu(tfaccy)2
Cu(fod)2
Cu(3-[NO2]acac)2
Cu(3-[CO2Et]acac)2
Cu(hfacac)2· H2O
Cu(hfacac)2 (dry)
Cu(hfacac)2 (dry)
Cu(hfacac)2 (dry)
Cu(hfacac)2 (dry)
0
20
70
99
99
99
99
22e
36f
a Conditions: 5 mol % of catalyst, toluene, 150 °C, 0.1 M. b NMR yield.
c Decomposes at 170 °C. d 5 M. e 1 mol %. f 130 °C. {hf ) hexafluoro,
acac ) acetylacetonate, (S,S)-Ph-box ) (S,S)-2,2′-Isopropylidene-bis(4-
phenyl-2-oxazoline), tf ) trifluoro, dbm ) dibenzoylmethanate, ptfm )
1,3-bis[4-(trifluoromethyl)phenyl]-1,3-propanedionate, accy ) 2-acetylcy-
clohexanone, fod ) heptafluorobutanoyl)pivaloylmethanate, 3-[NO2]acac
) 3-nitroacetylacetonate, 3-[CO2Et]acac ) ethyl diacetoacetate}
on the p-toluenesulfonyl (Ts) and phthalimido (Phth) protect-
ing groups. Tosyl aziridines are easily accessed by employing
an aziridination protocol developed by Sharpless12 or by
catalytic decomposition of N-phenyliodinanes.13 Although
phthalimide-substituted aziridines are less commonly used,
they offer an attractive option based on their ease of synthesis
utilizing a stabilized singlet nitrene and broader substrate
scope.14
Vinyl aziridine 1 was chosen as a model substrate due to
the growing interest in dehydroprolines.15 Based on previous
work by our group on the catalytic ring expansion of vinyl
oxiranes and vinyl thiiranes,16 copper catalysts were tested
(Table 1). The more electrophilic copper(II) hexafluoro-
acetylacetonate (entries 14-18) proved far superior to other
commercially available and synthetic catalysts. Interestingly,
the reaction proceeds faster and with a higher yield when
the Cu(hfacac)2 hydrate is dried prior to use. Optimized
reaction conditions were found to be 150 °C, 0.1 M
[substrate], and 5 mol % catalyst loading.
a Conditions: 5 mol % of dry Cu(hfacac)2, 0.1 M. b Racemic. c Isolated
yields. d 100 °C benzene. e 150 °C toluene. f 120 °C toluene. g Cu(hfacac)2.H2O
does not work. h 0.2 M. i 20 mol % of Cu(hfacac)2. j >20:1 dr. k 4:1 dr.
l Thermal component to reaction (see the Supporting Information).
(12) Jeong, J. U.; Tao, B.; Sagasser, I.; Henniges, H.; Sharpless, K. B.
J. Am. Chem. Soc. 1998, 120, 6844.
(13) Evans, D. A.; Faul, M. M.; Bilodeau, M. T. J. Org. Chem. 1991,
56, 6744.
(14) (a) Carpino, L. A.; Kirkley, R. K. J. Am. Chem. Soc. 1970, 92,
1784. (b) Siu, T.; Yudin, A. K. J. Am. Chem. Soc. 2002, 124, 530.
(15) (a) Werner, S.; Kasi, D.; Brummond, K. M. J. Comb. Chem. 2007,
9, 677. (b) Calaza, M. I.; Cativiela, C. Eur. J. Org. Chem. 2008, 3427. (c)
Mauger, A. B. J. Nat. Prod. 1996, 59, 1205.
Table 2 displays the results of applying these optimized
conditions to a variety of phthalimide- and tosyl-protected
vinyl aziridines. These products can be readily deprotected
to give the N-H pyrrolines.17 Simple substrates (entries
(16) (a) Batory, L. A.; McInnis, C. E.; Njardarson, J. T. J. Am. Chem.
Soc. 2006, 128, 16054. (b) Rogers, E.; Araki, H.; Batory, L. A.; McInnis,
C. E.; Njardarson, J. T. J. Am. Chem. Soc. 2007, 129, 2768.
5024
Org. Lett., Vol. 10, No. 21, 2008

1-11) demonstrate that all substitution patterns are well
tolerated resulting in excellent yields. Entries 4, 5, and 8
display an important synthetic aspect of this rearrangement,
as each 3-pyrroline can originate from two regioisomeric
vinyl aziridine precursors or even from their mixture. Some
of the simple substrates can rearrange thermally at more
extreme temperatures and often at dramatically diminished
yield. However, the thermal rearrangement of more complex
substrates is either nonexistent or very poor.
The copper-catalyzed rearrangement has been shown to
be quite functional group tolerant. These groups include enol
ethers, esters, protected alcohols, sulfonamides, and imides.
This compatibility is important because some of these
functional groups likely coordinate to the electrophilic copper
center and yet do not prevent the reaction from occurring.
Next, fused and bridged bicyclic compounds (entries 15-23)
can be rearranged to create complex ring systems. With
respect to protecting groups, we observe that in general the
more electron-rich phthalimide aziridines rearrange faster
than the corresponding tosyl-protected aziridines. Product
yields vary between protecting groups, but only phthalimide
aziridines are competent for very strained systems (entries
17-20).18 Entries 23 and 24 are particularly noteworthy,
since they establish the potential for diastereoselective
rearrangements.19 Lastly, the diastereomer of the substrate
in entry 18 and the Z-isomer of the substrate in entry 10 do
not rearrange under a variety of conditions tested. These
substrates display the importance of the conformation and
sterics of the substrate when it is bound to the catalyst.
Investigation into this reaction has revealed that the
rearrangement is copper specific and with a few exceptions
distinct to copper(II) acetylacetone ligands. Determination
of the mechanism has proved elusive, but the synthesis of
two new catalysts has probed what the active copper species
might be. The substrate from entry 3 was chosen as a model
system for investigation, and the results are shown in Figure
1. The reaction was conducted in sealed NMR tubes under
the standard reaction conditions (0.1 M of vinyl aziridine, 5
mol % of catalyst), except CDCl3 was used as the solvent.
The reactions with Cu(hfacac)2 (3) have a characteristic
sigmoidal shape suggesting an autocatalytic process. When
the catalyst loading is changed a dramatic effect on the rate
is observed. When a mixed acac catalyst is used, Cu(dbm-
)(hfacac) (4),20 a slightly slower reaction is observed with a
similar profile. Careful analysis of the slope of the line
suggests the maximum rate of the reaction is higher than
that of Cu(hfacac)2. This suggests that a more active catalyst
species is produced during the course of the reaction. The
Figure 1. Kinetics plot of 2-methyl-N-tosyl-2-vinylaziridine.
reaction profile of Cu(hfacac)(TMEDA)Br (5)21 is completely
different with little to no induction period or autocatalytic
behavior. This suggests that the catalyst is already in its most
active form, unlike Cu(hfacac)2. This catalyst is inferior likely
due to a less electrophilic copper(II) center. The mixed
complexes, 4 and 5, are peculiar since no other catalysts have
even approached the effectiveness of Cu(hfacac)2 as dem-
onstrated in Table 1. Finally, similar kinetic experiments were
performed in benzene, and those experiments confirmed the
same behavior; however, improved solubility in CDCl3
created more reproducible results.
Using all results obtained during this investigation, we
have produced a catalytic cycle shown in Scheme 1. We
propose that substrate (7) binds in the axial site of Cu(hfa-
cac)2 to form complex 8. The substrate induces dispropor-
tionation of the Cu(hfacac)2 to form cationic copper(II)
complex 9 and anionic copper(II) complex X-, which serves
as a counterion for the active cationic complex. This in situ
generated cationic copper(II) catalyst through a Lewis acid
activation weakens the C-N bond of the aziridine. During
the ring expansion, it is possible the copper center may
chelate to the nearby olefin producing the transition state
10. We believe that an ordered transition state, without
complete heterolysis or homolysis of the C-N bond, is
necessary to account for the diastereoselectivity observed in
the rearrangement of substrates 23 and 24. The rearrangement
produces complex 11, which then associatively undergoes
an interchange with more substrate to reform complex 9.
Complexes 9 and 11 constitute the autocatalytic cycle that
we propose is in good agreement with the observed kinetics
presented in Figure 1. Complexes 9 and 11 cannot be
observed during the course of the reaction because of the
paramagnetic behavior of copper(II).
(17) (a) Egli, M.; Hoesch, L.; Dreiding, A. S. HelV. Chim. Acta 1985,
68, 220. (b) Koohang, A.; Stanchina, C. L.; Coates, R. M. Tetrahedron
1999, 55, 9669. (c) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in
Organic Synthesis; Wiley-Interscience: New York, 1999; p 604.
(18) The substrates in entries 11, 13, and 24 under the thermal conditions
form azomethine ylides, and products from these intermediates are observed.
In the reactions of entries 19 and 20, the pyrroles can be isolated from the
retro-Diels-Alder reaction of the products.
This proposed mechanism accounts for a number of other
experimental observations not yet mentioned. First, product
12 was found to inhibit the reaction dramatically. Second,
(19) Unfortunately, when compound 1 was synthesized asymmetrically,
chirality was not transferred to the product. Compound 1 was transesterified
with a chiral alcohol in order to assess enantiomeric purity of the starting
material and product. The product was found not to epimerize under the
given reaction conditions.
(21) Fukuda, Y.; Yasuhira, M.; Sone, K. Bull. Chem. Soc. Jpn. 1985,
58, 3518.
(20) Dash, G. C.; Mohapatra, B. K. J. Indian Chem. Soc. 1984, 61, 830.
Org. Lett., Vol. 10, No. 21, 2008
5025

catalyst. Next, copper(II) species are observable (UV-vis
spectroscopy, NMR) throughout the reaction, but no cop-
per(I) intermediates could be detected, nor were copper(I)
catalysts competent in the rearrangement. Finally, using
Singleton’s method,23 we have obtained preliminary kinetic
isotope results for two different vinyl aziridine substrates.
These studies show significant heavy atom isotope effects
at all three carbon centers involved in the rearrangement,
suggesting the reaction is concerted in nature.
Scheme 1. Proposed Catalytic Cycle
In summary, we have demonstrated that 3-pyrrolines can
be readily accessed from vinyl aziridines using commercially
available copper(II) catalysts. The scope of this new catalytic
rearrangement is very broad, providing access to a greater
range of diverse products than other currently available
3-pyrroline-forming methods. Mechanistic analysis revealed
unique autocatalytic behavior and can be summarized as a
Lewis acid catalyzed [1,3]-sigmatropic rearrangement. New
catalysts were synthesized on the basis of these mechanistic
insights. These catalysts approach the effectiveness of
Cu(hfacac)2 but allow more options for electronic tuning and
induction of asymmetry. Further studies on the mechanism
and scope of this ring expansion are currently underway.
Acknowledgment. We thank Cornell University and the
NIH (CBI training grant to M.B.) for financial support. Emil
Lobkovsky (Cornell University) is gratefully acknowledged
for X-ray crystal structure analysis.
X- was found to not be catalytically competent when
synthesized independently.22 Also, reactions with cationic
copper(II) complex 5 do not possess an induction period
because it mimics the structure of 9 which is the active
Supporting Information Available: Experimental details
and spectral data for all new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL802123E
(22) Truter, M. R.; Fenton, D. E.; Vickery, B. L. J. Chem. Soc. D 1971,
93.
(23) Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117,
9357.
5026
Org. Lett., Vol. 10, No. 21, 2008