One-carbon ring expansion of azetidines via ammonium ylide [1,2]-shifts: A simple route to substituted pyrrolidines

Article abstract of DOI:10.1021/jo9001323

Simple N-substituted azetidines were heated with diazocarbonyl compounds in the presence of catalytic Cu(acac)₂ to furnish substituted pyrrolidines via Stevens [1,2]-shift. In all but two examples, complete selectivity was seen for ring expansion rather than migration of the other exocyclic group on the azetidinium nitrogen. The two exceptions, observed with ylides substituted with two carbonyl groups and lacking a stabilizing group at the 2-position of the azetidine, underwent exocyclic benzyl migration in preference to ring expansion.

Full text of DOI:10.1021/jo9001323

One-Carbon Ring Expansion of Azetidines via Ammonium Ylide
[1,2]-Shifts: A Simple Route to Substituted Pyrrolidines
Tina M. Bott,† John A. Vanecko,§,‡ and F. G. West*,†
Department of Chemistry, UniVersity of Alberta, E3-43 Gunning-Lemieux Chemistry Centre, Edmonton,
AB T6G 2G2, Canada, and Department of Chemistry, UniVersity of Utah, 315 S. 1400 East,
Rm 2020, Salt Lake City, Utah 84112
frederick.west@ualberta.ca
ReceiVed January 22, 2009
Simple N-substituted azetidines were heated with diazocarbonyl compounds in the presence of catalytic
Cu(acac)2 to furnish substituted pyrrolidines via Stevens [1,2]-shift. In all but two examples, complete
selectivity was seen for ring expansion rather than migration of the other exocyclic group on the azetidinium
nitrogen. The two exceptions, observed with ylides substituted with two carbonyl groups and lacking a
stabilizing group at the 2-position of the azetidine, underwent exocyclic benzyl migration in preference
to ring expansion.
Introduction
especially those containing one or more piperidine rings.6 More
recently, this methodology has been applied to the synthesis of
the pyrrolizidine alkaloids turneforcidine and platynecine.7 As
part of that work, a preliminary model study was carried out in
which intermolecular addition of a catalytically generated
metallocarbene to a simple azetidine furnished a protected
3-carboxyproline 4a as an inseparable mixture of diastereomers
in good yield via an intermediate azetidinium ylide 3a (eq 1).
Regioselective migration of the ester-substituted carbon of 3a
(with concomitant ring expansion)8 in preference to benzyl
migration was noteworthy, as it suggested that azetidine ring
strain could overcome the inherent ammonium ylide migratory
preferences seen in simple acyclic examples.9 Moreover, ef-
ficient intermolecular trapping of the metallocarbene by a simple
azetidine, if general, could offer a novel and convenient method
for the construction of substituted pyrrolidines from two readily
available building blocks. Here we describe the results from a
The pyrrolidine skeleton is found in a variety of settings. For
example, many naturally occurring alkaloids possess one or
more pyrrolidine rings.1 A large number of unnatural, biologi-
cally active substances also contain the pyrrolidine substructure.2
Finally, pyrrolidines are prominently featured in a number of
organocatalytic processes.3 Given their importance, it is not
surprising that many elegant methods have been developed for
the synthesis of pyrrolidines.4
One-carbon ring expansion of ammonium ylides via the
Stevens rearrangement5 has proven to be a convenient and
versatile entry to various nitrogen-containing heterocycles,
† University of Alberta.
§ Current address: Amgen Inc., One Amgen Center Drive, Thousand Oaks,
CA, 91320-1799, USA.
‡ University of Utah.
(1) (a) Michael, J. P. Nat. Prod. Rep. 2008, 25, 139–165. (b) Vanecko, J. A.;
West, F. G. Heterocycles 2007, 74, 125–143. (c) Daly, J. W.; Spande, T. F.;
Garraffo, H. M. J. Nat. Prod. 2005, 68, 1556–1575. (d) Pyne, S. G.; Davis,
A. S.; Gates, N. J.; Hartley, J. P.; Lindsay, K. B.; Machan, T.; Tang, M. Synlett
2004, 2670–2680. (e) Felpin, F.-X.; Lebreton, J. Eur. J. Org. Chem. 2003, 369,
3–3712.
(6) (a) Glaeske, K. W.; Naidu, B. N.; West, F. G. Tetrahedron: Asymmetry
2003, 14, 917–920. (b) Naidu, B. N.; West, F. G. Tetrahedron 1997, 53, 16565–
16574. (c) West, F. G.; Naidu, B. N. J. Am. Chem. Soc. 1993, 115, 1177–1178.
(7) Vanecko, J. A.; West, F. G. Org. Lett. 2005, 7, 2949–2952.
(2) (a) Bellina, F.; Rossi, R. Tetrahedron 2006, 62, 7213–7256. (b) Cheng,
X.-C.; Wang, Q.; Fang, H.; Xu, W.-F. Curr. Med. Chem. 2008, 15, 374–385.
(3) (a) Gruttadauria, M.; Giacalone, F.; Noto, R. Chem. Soc. ReV. 2008, 37,
1666–1688. (b) Ting, A.; Schaus, S. E. Eur. J. Org. Chem. 2007, 5797–5815.
(c) Guillena, G.; Najera, C.; Ramon, D. J. Tetrahedron: Asymmetry 2007, 18,
2249–2293.
(8) For other examples of azetidinium ylide ring expansions, see:(a) Couty,
F.; Durrat, F.; Evano, G.; Marrot, J. Eur. J. Org. Chem. 2006, 4214–4223. (b)
Wills, M. T.; Wills, I. E.; Von Dollen, L.; Butler, B. L.; Porter, J.; Anderson,
A. G. J. Org. Chem. 1980, 45, 2489–2498. (c) Hatu, Y.; Watanabe, M.
Tetrahedron Lett. 1972, 46, 4659–4660.
(9) West, F. G.; Glaeske, K. W.; Naidu, B. N. Synthesis 1993, 977–980.
(10) For other applications of azetidinium ylides, see: (a) Alex, A.; Larmanjat,
B.; Marrot, J.; Couty, F.; David, O. Chem. Commun. 2007, 2500–2502. (b) Couty,
F.; David, O.; Larmanjat, B.; Marrot, J. J. Org. Chem. 2007, 72, 1058–1061.
(4) (a) Minatti, A.; Muniz, K. Chem. Soc. ReV. 2007, 36, 1142–1152. (b)
Wolfe, J. P. Eur. J. Org. Chem. 2007, 57, 1–582.
(5) Vanecko, J. A.; Wan, H.; West, F. G. Tetrahedron 2006, 62, 1043–1062.
2832 J. Org. Chem. 2009, 74, 2832–2836
10.1021/jo9001323 CCC: $40.75 2009 American Chemical Society
Published on Web 03/05/2009

One-Carbon Ring Expansion of Azetidines
TABLE 1. Survey of Catalyst and Conditions for the Formation
of 4aa
more extensive study of the azetidinium ylide ring expansion
process,10 in which a range of azetidine and diazocarbonyl
reactants undergo successful one-step conversion to pyrrolidine
products.
entry
catalyst (mol %)
yield of 4a (%)b
1
2
3
4
5
Cu powder (50)
Rh2(OAc)4 (5)
Cu(acac)2 (10)
Cu(tfacac)2 (10)
Cu(hfacac)2 (10)
39
45
56
50c
61
a Standard procedure: a solution of 1a and catalyst in PhCH3 (0.025
M) was heated to reflux, and a solution of 2a (0.67 equiv) in PhCH3
(0.1 M) was added via syringe pump over a 12 h period. Following
completion of addition, the reaction mixture was cooled to rt then
washed with an equivalent volume of 0.5 M aqueous K2CO3 solution
and brine, dried over MgSO4, filtered, concentrated under reduced
pressure, and purified by flash chromatography. b All yields given are
for isolated product after chromatographic purification. In all cases 4a
was obtained as a 1:1 mixture of diastereomers. c Yields up to 82%
were obtained on occasion, but these results were not consistently
reproducible.
Results and Discussion
Azetidines 1a-e were selected as substrates.11 Variation of
both the exocyclic nitrogen substituent and substitution adjacent
to N (at C-2) within the ring was expected to address the scope
of the methodology. Likewise, a range of diazocarbonyl partners
2a-d was examined, possessing either one or two electron-
withdrawing groups.12 Initial optimization studies employed
methyl N-benzylazetidine-2-carboxylate 1a and ethyl diazoac-
etate 2a to provide carboxyproline diester13 4a (Table 1).
Rhodium(II) acetate dimer and various copper(II) acetylaceto-
nate complexes (10 mol %) proved to be effective catalysts for
the carbene-transfer process, with slow addition of 2a to a
solution of 1a and catalyst in toluene at reflux. Copper powder
(0.5 equiv) also catalyzed the reaction, but with decreased
efficiency. Although Cu(tfacac)2 furnished the highest yields
on occasion, we found this catalyst to be capricious. On the
other hand, Cu(acac)2 offered consistent and reproducible results,
and was therefore selected for use with the other substrate pairs.
The generality of the ring expansion protocol was examined
next, employing all possible combinations of 1a-e and 2a-d
(Table 2). Treatment of azetidine 1a with diazomalonate 2b
under the standard conditions from Table 1 furnished pyrrolidine
triester 4b, but at an inconveniently slow rate. We therefore
investigated microwave heating as an alternative, and found that
this method afforded 4b in high yield and with minimal
formation of side products. Microwave heating also led to
improved yields of 4a (entry 2) and these alternative conditions
were employed in a number of other cases. Both methods were
investigated for each reactant pair, and the conditions that
furnished products in the highest yield are given in the table.
While microwave heating reduced the formation of ylide
decomposition products and metallocarbene solvent adducts, it
did frequently lead to greater amounts of olefinic products
derived from “dimerization” of 2, since slow addition is
precluded under those conditions. With diazoacetophenone 2c,
the only isolable product was pyrroline 5a, formed in modest
yields (entry 10). Apparent dehydrogenation products of this
sort were also obtained from the reaction of 2c with azetidines
1b,c (entries 11 and 12).14 Variation of the azetidine exocyclic
nitrogen substituent, replacing N-benzyl with N-pentyl or N-allyl,
had little effect on the overall efficiency of the process. For
example, 1a-c furnished 4a,d,g in 62-71% yield with ethyl
diazoacetate 2a (entries 2, 4, and 7) and 4b,e,h in 60-81%
yield with diazomalonate 2b (entries 1, 5, and 8). However, in
the case of N-allyl substrate 1c, it is especially notable that
complete selectivity for ring expansion over allyl migration was
seen (see below for further discussion).
(11) The known azetidine substrates 1a,b,d,e were prepared via the following
literature procedures: (a) 1a: Rodebaugh, R. M.; Cromwell, N. H. J. Heterocycl.
Chem. 1968, 5, 309–311. (b) 1b : Wasserman, H. H.; Lipshutz, B. H.; Tremper,
A. W.; Wu, S. W. J. Org. Chem. 1981, 46, 2991–2999. (c) 1d : Leonard, N. J.;
Durand, D. A. J. Org. Chem. 1968, 33, 1322–1333. (d) 1e: Lai, G. Synth.
Commun. 2001, 31, 565–568. (e) Methyl N-allylazetidine-2-carboxylate 1c was
prepared using method B from ref 11b. For a similar procedure, see: (f) Starmans,
W. A. J.; Doppen, R. G.; Thijs, L.; Zwanenburg, B. Tetrahedron: Asymmetry
1998, 9, 429–435.
(12) (a) Ethyl diazoacetate 2a was purchased from Sigma-Aldrich. (b) Diethyl
diazomalonate 2b was prepared via the standard diazotransfer procedure described
in: Doyle, M. P.; McKervey, M. A.; Ye, T. Modern Catalytic Methods for
Organic Synthesis with Diazo Compounds; Wiley-Interscience: New York, 1998;
p 10. (c) Diazoacetophenone 2c : Jung, M. E.; Min, S.; Houk, K. N.; Ess, D. J.
Org. Chem. 2004, 69, 9085–9089. (d) 2-Diazo-3-oxo-3-phenylpropionic acid
ethyl ester 2d: Lall, M. S.; Ramtohul, Y. K.; James, M. N. G.; Vederas, J. C. J.
Org. Chem. 2002, 67, 1536–1547.
(13) Carboxyproline derivatives are of interest as rigid aspartate analogues: (a)
Flamant-Robin, C.; Wang, Q.; Chiaroni, A.; Sasaki, N. A. Tetrahedron 2002,
58, 10475–10484. (b) Carpes, M. J. S.; Miranda, P. C. M. L.; Correia, C. R. D.
Tetrahedron Lett. 1997, 38, 1869–1872. (c) Cotton, R. J. A. N. C.; North, M.
Tetrahedron 1995, 51, 8525–8544. (d) Baldwin, J. E.; Adlington, R. M.; Gollins,
D. W.; Godfrey, C. R. A. Tetrahedron 1995, 51, 5169–5180. (e) Sasaki, N. A.;
Pauly, R.; Fontaine, C.; Chiaroni, A.; Riche, C.; Potier, P. Tetrahedron Lett.
1994, 35, 241–244.
All of the examples discussed so far possessed an ester
substituent on the migrating center. Azetidines 1d,e offered an
important probe of the extent to which ring strain can drive
migratory selectivity. A significant early observation by Hatu
and Watanabe using 1e and 2a suggested that the ring expansion
(14) Khlebnikov and co-workers observed a similar oxidative process upon
heating a polycyclic 3-carboxyproline diester with ethyl diazoacetate in the
presence of copper powder. The mechanism of this transformation is unclear
and under current study. Khlebnikov, A. F.; Novikov, M. S.; Kostikov, R. R.;
Kopf, J. Russ. J. Org. Chem. 2005, 41, 1341–1348.
J. Org. Chem. Vol. 74, No. 7, 2009 2833

Bott et al.
TABLE 2. Effect of Structural Variation of 1 and 2a
azetidine
R1
Bn
Bn
Bn
C5H11
C5H11
C5H11
allyl
allyl
allyl
Bn
C5H11
allyl
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
R2
R3
diazocarbonyl
R4
R5
method
product
yield (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1a
1a
1a
1b
1b
1b
1c
1c
1c
1a
1b
1c
1d
1d
1d
1d
1e
1e
1e
1e
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
Me
H
H
H
H
H
H
H
H
H
H
H
H
Me
Me
Me
Me
H
2b
2a
2d
2a
2b
2d
2a
2b
2d
2c
2c
2c
2a
2b
2c
2d
2a
2b
2c
2d
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
OEt
Ph
CO2Et
H
COPh
H
CO2Et
COPh
H
CO2Et
COPh
H
B
B
A
B
B
A
B
B
A
B
B
B
B
B
B
B
Ad
A
A
A
4b
4a
81
66
c
-
4d
4e
4f
4g
4h
4i
5a
5b
5c
4j
62
60
20
71
69
32
34
65
48
62
Ph
Ph
H
H
H
OEt
OEt
Ph
OEt
OEt
OEt
Ph
c
Me
Me
Me
CO2Et
H
COPh
H
CO2Et
H
COPh
-
4l
24
c
-
H
H
H
H
4n
6a
4o
6b
67e
59
H
H
H
39f
21
OEt
a Method A: See Table 1, footnote a. Method B: A solution of 2 (1.5 equiv) in ClCH2CH2Cl (0.5 M) was added to a solution of 1 and Cu(acac)2 (10
mol %) in ClCH2CH2Cl (0.05M) in a 5 mL Biotage microwave vial. The vial was sealed and the resulting mixture was subjected to microwave
irradiation (150-180 °C) in a Biotage Initiator microwave reactor for 1 h. After cooling, the reaction mixture was washed with an equivalent volume of
0.5 M aqueous K2CO3 and brine, dried over MgSO4, filtered, concentrated, and purified by flash chromatography. b All yields given are for isolated
product after chromatographic purification. Under the conditions of method B, 9-20% of alkene product from dimerization of diazo compounds 2 was
also isolated. c Diazo compound 2d was consumed very slowly, in most cases furnishing a complex mixture containing diazo dimers and unreacted 1
(entries 3, 14, and 16). d For this reaction, Cu powder (50 mol %) was used in place of Cu(acac)2 and an additional filtration through celite to remove
residual Cu powder was added to the workup procedure. e Following an unspecified procedure, Hatu and Watanabe reported a yield of 96% in this
transformation, using Cu(acac)2 (see ref 8c). f A crude yield of 39% could be obtained in the case of pyrrolidine 4o, but purification was complicated by
its rapid conversion to lactam 7.
SCHEME 1. Oxidative Decomposition of
Benzoyl-Substituted Products
process could still be viable in the absence of a stabilizing
substituent on the migrating carbon.8c In the event, both 1d and
1e underwent reaction with 2a and 2c to give pyrrolidines
4j,l,n,o in modest to good yields (entries 13, 15, 17, and 19).
However, none of the desired pyrrolidine product was obtained
with 2b,d. In the case of 1d, only uncharacterizable polar
materials were formed, while with 1e the benzyl [1,2]-shift
products 6a,b were formed (entries 18 and 20).
Adducts derived from diazoacetophenone 2c showed unusual
lability. As noted above, with ester-substituted azetidines 1a-c,
only pyrrolines 5 were isolated, in moderate yields. In the case
of the simple N-benzylazetidine 1e, the expected pyrrolidine
4o was isolated, but it underwent rapid and apparently quantita-
tive conversion to N-benzylpyrrolidinone 7 during handling and
attempted chromatographic purification. This process, an ap-
parent debenzoylative oxidation, may involve reaction of the
ylide tautomer of 4o (4o′) with ambient oxygen (Scheme 1).15
It is also possible that the dehydrogenation products 5a-c may
result from an alternative elimination pathway following reaction
(15) (a) Facile aerobic debenzoylative oxidation of R-aminoketones under
basic conditions has been observed by Garc´ıa-Valverde and co-workers: Garc´ıa-
Valverde, M.; Pedrosa, R.; Vicente, M. Synlett 2002, 2092-2093. (b) For a
of intermediate pyrrolidines 4p-r with oxygen. Intervention
of these unexpected processes in benzoyl-substituted examples
related oxidative deacylation, see: Yijima, C.; Hino, F.; Suda, K. Synthesis 1981,
610–611.
2834 J. Org. Chem. Vol. 74, No. 7, 2009

One-Carbon Ring Expansion of Azetidines
SCHEME 2. Radical Pair Mechanism for the Stevens
[1,2]-Shift
than one competent migrating group on the ammonium nitrogen,
selective rearrangement is possible if there is a significant
energetic difference between the two possible radical pairs. In
the specific case of azetidine 1a, exclusive formation of
pyrrolidines 4a,b indicates that release of ring strain is a
significant factor in bond homolysis selectivity, since benzyl
groups show greater migratory aptitude than CH2CO2R in
acyclic substrates. The n-pentyl group of 1b is expected to have
limited migratory aptitude, so successful ring expansion in those
cases is not surprising. The presence of ꢀ-hydrogens on the side
chain does permit a possible R′,ꢀ-fragmentation process by
the intermediate ylide; however, no evidence was seen for the
formation of the simple dealkylation products 8c,d. As with
1a, rapid homolytic ring opening appears to be the predominant
fate of the azetidinium ylide.
Observation of exclusive ring expansion by N-allyl substrate
1c merits further comment. While the allyl group is capable of
undergoing Stevens [1,2]-shift via the usual radical pair mech-
anism, the alternative concerted [2,3]-shift process is also
possible. In those cases where both rearrangement pathways
are possible, the [2,3]-shift appears to occur more readily for
ammonium,20 sulfonium,21 and oxonium22 ylides, presumably
due to a lower activation barrier for the concerted process. The
absence of any [2,3]-shift product from 1c again indicates the
important effect of ring strain in azetidinium ylide rearrangements.
Azetidines 1d,e lacking a conjugating group on the migrating
carbon were chosen to test the limits of ring strain as a
predisposing factor in migratory aptitude. gem-Dimethyl sub-
stitution in the case of 1d was expected to provide moderate
stabilization of the radical intermediate, consistent with other
examples of [1,2]-shift by tertiary radicals lacking any conjugat-
ing groups.23 In fact, a surprisingly good yield of N-benzyl-
3,3,-dimethylproline ethyl ester 4j was obtained, with no
evidence for competing benzyl shift. Formation of the corre-
sponding benzoyl-substituted 4l in only modest yield is disap-
pointing, but oxidative decomposition analogous to that seen
for 4o may be occurring.24 The failure to observe comparable
results with diazo partners 2b,d is puzzling, and requires further
investigation. The behavior of the simple N-benzylazetidine 1e
is especially intriguing. As mentioned above, Hatu and Wa-
tanabe had previously reported successful ring expansion to
provide protected proline 4n, a result that we confirmed.
Likewise, treatment with diazoacetophenone also leads to
pyrrolidine product (4o). However, doubly stabilized ylides
derived from diazomalonate or the corresponding ketoester
provide only the benzyl [1,2]-shift products 6a,b. Failure to react
via the ring expansion pathway in these cases indicates the limits
of ring strain as a deciding factor in migratory selectivity. It is
possible that doubly stabilized ammonium ylides, which are
generally slower to rearrange than their monostabilized coun-
terparts, are more sensitive to the stability of the radical on the
may be due to the greater acidity of the adjacent R proton in
comparison to carboethoxy-substituted cases (e.g., 4a,d,g,j,n),
and no such reactivity is expected from malonate- or ketoester-
derived products, due to the absence of an acidic proton adjacent
to the ring nitrogen.
The Stevens [1,2]-shift of ammonium ylides is believed to
involve a stepwise homolytic mechanism (Scheme 2).16 In light
of the likely intermediacy of a biradical, it is not surprising that
pyrrolidines 4a,d,f,g,i were formed as diastereomeric mixtures.
In the case of spirocyclic ammonium ylides resulting from
intramolecular metallocarbene addition high diastereoselectivity
was seen, with high levels of retention of configuration during
migration of an ester-substituted center.6 On the other hand,
with monocyclic ylides such as 3, although face-selective
metallocarbene addition cis to the neighboring ester is expected
to predominate,17 both rotamers of the exocyclic ylide N-C
bond are likely to be present. Even if the subsequent [1,2]-shift
occurred with retention, both diastereomers would be expected.
Moreover, if the biradical intermediate persists long enough to
randomize, little diastereoselectivity would be expected in the
eventual radical recombination step.18
The radical center, residing at the former ylide carbon, is
stabilized by both the electron-withdrawing substituent and the
basic nitrogen.19 Typically, the migrating group is substituted
with a moiety able to stabilize the other radical center through
conjugation (e.g., aryl, alkenyl, or carbonyl). If there is more
(16) (a) Heard, G. L.; Yates, B. F. Aust. J. Chem. 1995, 48, 1413–1423. (b)
Ollis, W. D.; Rey, M.; Sutherland, I. O. J. Chem. Soc., Perkin Trans. 1 1983,
1009–1027.
(21) Vedejs, E.; Arnost, M. J.; Hagen, J. P. J. Org. Chem. 1979, 44, 3230–
3238.
(22) (a) West, F. G.; Eberlein, T. H.; Tester, R. W. J. Chem. Soc., Perkin
Trans. 1 1993, 2857–2859. (b) Marmsa¨ter, F. P.; Vanecko, J. A.; West, F. G.
Tetrahedron 2002, 58, 2027–2040. (c) For the role of transition metal catalysts
in periselectivity, see: Marmsa¨ter, F. P.; Vanecko, J. A.; West, F. G. Org. Lett.
2004, 6, 1657–1660.
(23) (a) Tester, R. W.; West, F. G. Tetrahedron Lett. 1998, 39, 4631–4634.
(b) West, F. G.; Naidu, B. N. J. Org. Chem. 1994, 59, 6051–6056.
(24) None of the 3,3-dimethyl-2-pyrrolidinone product analogous to 7 was
isolated from reaction of 1d with 2c.
(25) For a related example of exclusive benzyl migration rather than ring
expansion of an azetidinium ylide stabilized by ester and trifluoromethyl groups,
see: Osipov, S. N.; Sewald, N.; Kolomiets, A. F.; Fokin, A. V.; Burger, K.
Tetrahedron Lett. 1996, 37, 615-618.
(17) (a) Vedejs, E.; Arco, M. J.; Powell, D. W.; Renga, J. M.; Singer, S. P.
J. Org. Chem. 1978, 43, 4831–4837. (b) Kawanishi, N.; Fujiwara, K.; Shirai,
N.; Sato, Y.; Hatano, K.; Kurono, Y. J. Chem. Soc., Perkin Trans. 1 1997, 3013–
3016. (c) Glaeske, K. W.; West, F. G. Org. Lett. 1999, 1, 31–33.
(18) Epimerization of one or both stereocenters may also occur during the
reaction. However, since the diastereomeric mixtures were inseparable, it was
not possible to probe for this by resubjecting pure diastereomers to the reaction
conditions.
(19) (a) Welle, F. M.; Beckhaus, H.-D.; Ru¨chardt, C. J. Org. Chem. 1997,
62, 552–558. (b) Schulze, R.; Beckhaus, H.-D.; Ru¨chardt, C. Chem. Ber. 1993,
126, 1031–1038.
(20) (a) Roberts, E.; Sanc¸on, J. P.; Sweeney, J. B. Org. Lett. 2005, 7, 2075–
2078. (b) Jemison, R. W.; Laird, T.; Ollis, W. D.; Sutherland, I. O. J. Chem.
Soc., Perkin Trans. 1 1980, 1450–1461.
J. Org. Chem. Vol. 74, No. 7, 2009 2835

Bott et al.
migrating center.25 Further competition experiments with other
nonconjugating substituents on the migrating carbon may help
clarify this issue.
Supporting Information for a description of temperature monitoring
in microwave reactions.) After cooling, the reaction mixture was
washed with an equivalent volume of 0.5 M aqueous K2CO3 and
brine, dried over MgSO4, filtered, concentrated, and purified by
flash chromatography.
Conclusion
1-Benzylpyrrolidine-2,3-dicarboxylic acid 3-methyl 2-ethyl
ester 4a (1:1 mixture of diastereomers as determined by
integration of the OMe singlets in the H NMR spectrum): Rf
A general method for the preparation of substituted pyrro-
lidines has been described. Heating two readily available
reactants, an N-substituted azetidine and a diazocarbonyl
compound, in the presence of catalytic Cu(acac)2 results in the
formation of an azetidinium ylide via addition of a transient
metallocarbene to the basic nitrogen of the azetidine. Microwave
heating usually gives cleaner reactions and higher yields over
much shorter reaction times. In most cases the ylide undergoes
regioselective [1,2]-shift of an azetidine ring carbon, leading
to ring expanded products, even when competent migrating
groups such as benzyl or allyl are present as exocyclic nitrogen
substituents. A range of substitution patterns is tolerated on the
migrating center, including (in the case of monostabilized
metallocarbenes) CH2. For examples leading to two adjacent
stereocenters no diastereoselectivity is observed. Future efforts
will focus on control of relative and absolute configuration, as
well as the delineation of factors controlling migratory aptitude
in ylides lacking a strongly stabilizing group at the azetidine
C-2 position.
1
0.17 (2:3 EtOAc/hexanes); IR (thin film) 2979, 2954, 2839, 1740,
1603, 1495, 1453 cm-1; 1H NMR (300 MHz, CDCl3) δ 7.34-7.22
(m, 10H), 4.21-4.09 (m, 4H), 3.99 (d, J ) 12.9 Hz, 1H), 3.83-3.70
(m, 3H), 3.71 (s, 3H), 3.65 (s, 3H), 3.58 (d, J ) 2.9 Hz, 1H), 3.54
(d, J ) 3.6 Hz, 1H), 3.33-3.21 (m, 2H), 3.06-2.98 (m, 2H), 2.71
(ddd, J ) 8.8, 7.4, 7.4 Hz, 1H), 2.56-2.47 (m, 1H), 2.35 (dddd, J
) 12.7, 9.2, 9.2, 7.4 Hz, 1H), 2.22-2.02 (m, 3H), 1.26 (t, J ) 7.2
Hz, 3H), 1.25 (t, J ) 7.2 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ
174.5, 172,8, 172.7, 171.5, 138.4, 138.4, 129.2, 129.1, 128.4, 128.4,
127.4, 127.3, 68.4, 66.2, 61.2, 60.7, 58.8, 57.2, 52.8, 52.4, 52.1,
51.5, 47.3, 46.6, 27.7, 26.7, 14.5, 14.4.
1-Benzylpyrrolidine-2,2,3-tricarboxylic acid 3-methyl 2,2-
diethyl ester 4b: IR (thin film) 2982, 2842, 1731, 1495, 1454, 1367,
1214, 1098, 1028 cm -1; 1H NMR (500 MHz, CDCl3) δ 7.37-7.35
(m, 2H), 7.29-7.26 (m, 2H), 7.22-7.19 (m, 1H), 4.34-4.20 (m,
4H), 3.97 (d, J ) 13.5 Hz, 1H), 3.85 (d, J ) 13.5 Hz, 1H), 3.71
(dd, J ) 9.0, 8.4 Hz, 1H), 3.68 (s, 3H), 2.92 (app td, J ) 8.8, 4.7
Hz, 1H), 2.81 (app td, 8.5, 6.8 Hz, 1H), 2.24-2.13 (m, 2H), 1.30
(app t, J ) 7.1 Hz, 3H), 1.29 (app t, J ) 7.2 Hz, 3H); 13C NMR
(125 MHz, CDCl3) δ 172.3, 169.1, 168.7, 139.6, 128.3, 128.2,
126.9, 76.9, 61.6, 61.3, 54.8, 52.0, 51.1, 50.4, 26.1, 14.2, 14.1;
HRMS calcd for C19H25NO6Na (M + Na) 386.1574, found
386.1577.
Experimental Section
General Procedures for Ring Expansions. Method A: a
solution of azetidine 1 and catalyst in PhCH3 (0.025 M) was heated
to reflux, and a solution of diazo compound 2 (0.67 equiv) in PhCH3
(0.1 M) was added via syringe pump over a 12 h period. Following
completion of addition, the reaction mixture was cooled to rt then
washed with an equivalent volume of 0.5 M aqueous K2CO3
solution and brine, dried over MgSO4, filtered, concentrated under
reduced pressure, and purified by flash chromatography. Method
B: A solution of 2 (1.5 equiv) in ClCH2CH2Cl (0.5 M) was added
to a solution of 1 and Cu(acac)2 (10 mol %) in ClCH2CH2Cl (0.05
M) in a 5 mL Biotage microwave vial. The vial was sealed and the
resulting mixture was subjected to microwave irradiation (150-180
°C) in a Biotage Initiator microwave reactor for 1 h. (See the
Acknowledgment. This work was supported by NSERC
(Canada) and the donors of the Petroleum Research Fund,
administered by the American Chemical Society. T.B. thanks
NSERC for a PGSM award and the University of Alberta for a
Queen Elizabeth II scholarship.
Supporting Information Available: NMR spectra for
4b,d-j,l,o, 5a-c, and 6a,b. This material is available free of
charge via the Internet at http://pubs.acs.org.
JO9001323
2836 J. Org. Chem. Vol. 74, No. 7, 2009