Stereoselective Lewis acid mediated (3+2) cycloadditions of N-H- and N-sulfonylaziridines with heterocumulenes

Article abstract of DOI:10.1002/chem.201303699

Alkyl and aryl isothiocyanates and carbodiimides are effective substrates in (3+2) cycloadditions with N-sulfonyl-2-substituted aziridines and 2-phenylaziridine for the synthesis of iminothiazolidines and iminoimidazolidines. Additionally, the stereoselec

Full text of DOI:10.1002/chem.201303699

DOI: 10.1002/chem.201303699
Full Paper
&
Organic Synthesis
Stereoselective Lewis Acid Mediated (3+2) Cycloadditions of
N-H- and N-Sulfonylaziridines with Heterocumulenes
Robert A. Craig, II, Nicholas R. O’Connor, Alexander F. G. Goldberg, and Brian M. Stoltz*[a]
Abstract: Alkyl and aryl isothiocyanates and carbodiimides
are effective substrates in (3+2) cycloadditions with N-sulfo-
nyl-2-substituted aziridines and 2-phenylaziridine for the syn-
thesis of iminothiazolidines and iminoimidazolidines. Addi-
tionally, the stereoselective (3+2) cycloaddition of N-H- and
N-sulfonylaziridines with isothiocyanates can be accom-
plished, allowing for the synthesis of highly enantioenriched
iminothiazolidines. Evidence for an intimate ion-pair mecha-
nism is presented herein in the context of these chemo-,
regio-, and diastereoselective transformations. The demon-
strated ability to remove the sulfonyl group from the hetero-
cyclic products displays the utility of these compounds for
further derivatization and application.
Introduction
aziridines in (3+2) cycloadditions has been explored minimal-
ly.[6a–e,k] In fact, only the work of Nadir and co-workers has ex-
amined the (3+2) cycloadditions of N-sulfonyl-2-arylaziridines
with heterocumulenes.[6k,9] Their reaction system has a narrow
scope and is only able to accommodate aryl isocyanates and
aryl isothiocyanates, resulting in similarly limited product deri-
vatization options. This transformation depends on the use of
an alkali metal iodide as a noninnocent reaction partner; Nadir
and co-workers explicitly demonstrate the formation of the
ring-opened iodide intermediate prior to product formation.
Additionally, only a single example is known for the synthe-
sis of enantioenriched iminothiazolidines by a stereoselective
(3+2) cycloaddition[6i] despite readily available enantioenriched
aziridine starting materials.[1,2,6h] This method, however, has an
extremely narrow substrate scope, requiring the use of N-alkyl-
or N-arylaziridines and aryl heterocumulenes. There are no ex-
amples of this transformation with N-sulfonyl-protected aziri-
dines or more synthetically versatile heterocumulenes.
Aziridines are versatile intermediates and reaction partners for
the preparation of a structurally diverse assortment of nitro-
gen-containing architectures.[1] These heterocycles are charac-
terized by a unique reactivity profile, in part due to the large
strain energy (27 kcalmolÀ1) contained within their three-mem-
bered ring structure,[2] rendering them susceptible to nucleo-
philic ring opening,[3] carbonylation,[4] and ring expansion.[5]
Previous work has shown the utility of 2-arylaziridines in transi-
tion metal-mediated and -catalyzed (3+2) cycloadditions with
heterocumulenes for the formation of imidazolines,[6a–e] oxazoli-
dines,[6d–e] iminoazolidinones,[6f–i] iminothiazolidines,[6i–k] and imi-
noimidazolidines.[6f,i] Iminothiazolidines and iminoimidazoli-
dines have enjoyed broad application as effective organic cata-
lysts in asymmetric transformations including Strecker reac-
tions,[7a] O- and N-acylations,[7b–c] and Michael additions,[7d] and
as highly active pharmacophores for the treatment of a wide
range of medical conditions including obesity, diabetes, cancer,
and arthritis.[8]
From this foundation, we sought to develop the first stereo-
selective Lewis acid mediated (3+2) cycloaddition reaction of
N-sulfonyl-2-substituted aziridines and alkyl heterocumulenes.
The Lewis acid mediated conditions would likely enable the
use of a broad variety of heterocumulenes and consequently
furnish readily derivatizable, highly enantioenriched heterocy-
clic building blocks.
The critical limitation of the majority of the existing (3+2)
cycloaddition manifolds is the requirement that aziridine start-
ing materials bear either alkyl or aryl N-substitution. The harsh
conditions necessary for the removal of such robust groups se-
verely limits the potential for derivatization and, thus, the utili-
ty of the products. Despite this, the use of N-sulfonyl-protected
Results and Discussion
[a] R. A. Craig, II, N. R. O’Connor, Dr. A. F. G. Goldberg, Prof. B. M. Stoltz
Warren and Katharine Schlinger Laboratory for Chemistry
and Chemical Engineering
Optimization of reaction conditions
Initial reaction development focused on the cycloaddition of
N-tosyl-2-phenylaziridine (1a) with allyl isothiocyanate
(Table 1). In contrast to our previous work on (3+2) cycloaddi-
tions of donor–acceptor cyclopropanes with heterocumu-
lenes,[10] tin(II) triflate was found to be an ineffective Lewis acid
mediator for the desired transformation (Table 1, entry 1). Alter-
Division of Chemistry and Chemical Engineering
California Institute of Technology
1200 E California Blvd, MC 101-20
Pasadena, CA 91125 (USA)
E-mail: stoltz@caltech.edu
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201303699.
Chem. Eur. J. 2014, 20, 4806 – 4813
4806
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Optimization of reaction conditions.[a]
Table 2. Substrate scope of isothiocyanate (3+2) cycloadditions.[a]
Entry
Lewis acid
t [h]
Yield [%][b]
1
2
3
4
5
6
7
Sn(OTf)2
Zn(OTf)2
ZnCl2
1.0
60.0
6.0
3.0
1.3
72.0[d]
72.0[d]
0
79
95
95
99
7
Znl2
ZnBr2
LiBr[c]
ZnBr2
[e]
4
[a] Conditions: aziridine 1a (0.40 mmol), isothiocyanate (0.80 mmol),
Lewis acid (0.50 mmol), CH2Cl2 (0.80 mL). [b] Isolated yield. [c] 1.00 mmol
of LiBr. [d] Starting material was not fully consumed. [e] 0.12 mmol of
ZnBr2 with 0.40 mmol of tetra(n-butyl)ammonium bromide.
natively, zinc(II) salts proved competent for the formation of
iminothiazolidine 2a. While zinc(II) triflate furnished the prod-
uct in good yield, the reaction times were greatly reduced
when zinc(II) halides were employed (entries 2–5). Lithium bro-
mide mediated reaction conditions resulted in low conversion
(entry 6). Attempts to develop a catalytic system with zinc(II)
bromide proved unsuccessful, even in the presence of an addi-
tional bromide ion (entry 7). Ultimately, the use of 1.25 equiva-
lents of zinc(II) bromide and 2.00 equivalents of allyl isothiocya-
nate in dichloromethane at ambient temperature proved opti-
mal (entry 5).[11]
[a] Conditions: aziridine 1 (0.40 mmol), isothiocyanate (0.80 mmol), ZnBr2
(0.50 mmol), CH2Cl2 (0.80 mL). [b] Isolated yield. [c] The product of nucleo-
philic ring opening of the starting material by a bromide ion was isolated
in 35% yield. [d] 0.90 mmol of ZnBr2.
Exploration of 2-aziridine and heterocumulene substitution
With optimal conditions identified, we examined the substrate
scope of the reaction. We found that a variety of N-tosyl-2-aryl-
substituted aziridines participated effectively in the zinc(II) bro-
mide mediated (3+2) cycloaddition with allyl isothiocyanate to
yield the corresponding iminothiazolidine products with com-
plete chemo- and regioselectivity (Table 2).[12,13] Altering the C-
aryl substitution from phenyl to mesityl allowed for the forma-
tion of the corresponding heterocycle (2b) in a shorter reac-
tion time despite the increased steric bulk.[14] Similarly, (p-tol-
yl)thiazolidine 2c was successfully furnished with a slightly de-
creased yield. Acetoxy substitution was compatible with the re-
action conditions as well, generating 2d in excellent yield.
Compared to the p-chlorophenyl- or o-chlorophenylthiazoli-
dines (2e and 2g, respectively), the more electronically deacti-
vated m-chlorophenylthiazolidine 2h was produced more
slowly, albeit without any significant reduction in
also be synthesized in approximately the same reaction time
as phenyl product 2a. Additionally, primary and secondary
alkyl isothiocyanates were found to be highly compatible
under the reaction conditions as were electron-neutral, -rich,
and -deficient aryl isothiocyanates, all furnishing the desired
iminothiazolidines in excellent yields (2k–o, respectively).
In contrast to the C-aryl-substituted aziridines, C-alkyl-substi-
tuted aziridine 4 reacted with allyl isothiocyanate under the re-
action conditions to furnish two isomeric (3+2) adducts
(Scheme 1). Formation of 5-alkyl-substituted iminothiazolidine
5 was accomplished in only 18% yield, whereas 4-alkyl-substi-
tuted product 6 was furnished in 56% yield. While C-alkyl-sub-
yield.[15] The highly electron-deficient p-nitrophenylth-
iazolidine 2 f was formed in modest yield, with a sig-
nificant portion of starting material lost to nucleo-
philic ring opening of the aziridine by the bromide
counterion.[16] Substrates bearing coordinating C-aryl
substituents served to slow the rate of reaction (e.g.,
2d, 2 f) or require additional ZnBr2 to drive the reac-
tion to completion (2i). Styrenyl thiazolidine 2j could Scheme 1. (3+2) cycloaddition with 2-alkylaziridine 4.
Chem. Eur. J. 2014, 20, 4806 – 4813
www.chemeurj.org 4807
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
stituted aziridines are suitable reaction partners in the (3+2)
cycloaddition and the heterocyclic products are formed with
complete chemoselectivity, they are not formed with the regio-
fidelity exhibited by aziridines substituted at carbon with aryl
groups or other conjugated systems.
Table 4. Substrate scope of carbodiimide (3+2) cycloaddition.[a]
Effect of aziridine N-substitution
As an extension of the substrate scope, we investigated the
effect of N-substitution on the aziridine (Table 3). Aziridines
protected with N-sulfonyl groups provided the desired imino-
thiazolidines (2a, 8a–c) in excellent yields, generally showing
Table 3. N-Substitution scope of isothiocyanate (3+2) cyclo-
addition.[a]
[a] Conditions: aziridine 1 (0.40 mmol), carbodiimide (0.41 mmol), ZnBr2
(0.50 mmol), CH2Cl2 (0.80 mL). [b] Isolated yield. [c] Bis(trimethylsilyl)carbo-
diimide was used.
imidazolidine 9b. In accordance with our observations in the
isothiocyanate (3+2) cycloadditions, a variety of 2-arylaziridines
were suitable reaction partners with diphenylcarbodiimide
(9c–e).
Entry
R
Product
t [h]
Yield [%][b]
1
2
3
4
5
6
7
8
tosyl
p-NO2-C6H4-SO2
mesyl
p-OMe-C6H4-SO2
H
n-C10H21
pivaloyl
benzoyl
2a
8a
8b
8c
8d
8e
8 f
1.3
1.2
1.8
2.0
0.5
96.0[c]
96.0[c]
96.0[c]
99
94
91
90
75
0
Cycloaddition of disubstituted N-sulfonylaziridines
0
0
8g
During our investigation of the substrate scope of the (3+2)
cycloaddition of N-sulfonylaziridines with isothiocyanates, we
discovered that trans-2,3-disubstituted aziridine 10 was an ac-
ceptable reaction partner, furnishing cis-thiazolidine 11 in high
yield in the shortest reaction time observed for any N-tosyl-
substituted substrate (Scheme 2). Iminothiazolidine 11 was
formed as a single diastereomer, and the relative stereochemis-
try was confirmed by single-crystal X-ray diffraction. The cis
configuration of product 11 led to the hypothesis that the
mechanism of the reaction involves inversion at the benzylic
position of the aziridine starting material.[19]
[a] Conditions: aziridine 7 (0.40 mmol), isothiocyanate (0.80 mmol), ZnBr2
(0.50 mmol), CH2Cl2 (0.80 mL). [b] Isolated yield. [c] Starting material was
not consumed.
reaction times that were slightly shorter for electron-deficient
sulfonyl groups (Table 3, entry 2) and slightly longer for more
electron-rich sulfonyl groups (entries 3–4) in comparison to the
N-tosyl-substituted substrate (entry 1). Interestingly, unprotect-
ed 2-phenylaziridine showed an improved reaction time yet
a partially decreased yield of the heterocyclic product 8d,
whereas N-(n-decyl)-substituted aziridine was unreactive under
the reaction conditions (entries 5–6). N-Acyl aziridines also
failed to furnish any of the desired heterocycles 8 f and 8g (en-
tries 7–8).
To confirm that the chemo-, regio-, and diastereoselective
formation of cis-thiazolidine 11 was not substrate dependent,
we exposed trans-2,3-disubstituted acroylaziridine 12 to identi-
cal reaction conditions and were pleased to find that cis-ac-
roylthiazolidine 13 was similarly formed as the sole product
and as a single diastereomer (Scheme 3).
Extension of heterocumulene scope
Subsequently, we turned our attention to broadening
the scope of competent heterocumulenes and found
that while isocyanates gave only trace yields of the
(3+2) cycloaddition adducts,[17] carbodiimides were
compatible with the conditions, furnishing iminoimi-
dazolidines (Table 4).[18] Our unique zinc-mediated
conditions enabled the (3+2) cycloadditions of N-sul-
fonylaziridines with dialkyl- and disilyl-carbodiimides,
unlike any previously published systems,[6,9] furnish-
ing diisopropyliminoimidazolidine 9a and secondary Scheme 2. Diastereoselective (3+2) cycloaddition with aziridine 10.
Chem. Eur. J. 2014, 20, 4806 – 4813
www.chemeurj.org 4808
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Contrastingly, the (3+2) cycloaddition of allyl isothiocyanate
with cis-2,3-disubstituted aziridine 15 resulted in the nondias-
tereoselective formation of both trans-thiazolidine 16 and the
cis isomer 17 (Scheme 4a). Interestingly, exposure of cis-aziri-
dine 15 to the reaction conditions in the absence of heterocu-
mulene resulted in the rapid formation of pyrrolines 18 and 19
(Scheme 4b).[20] Finally, we found that geminally disubstituted
N-tosyl-2-methyl-2-phenylaziridine failed to provide any (3+2)
cycloaddition product.[21]
Table 5. Optimization of stereoselective reaction conditions.[a]
Entry
Lewis acid
t [h]
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
8
ZnBr2
Znl2
ZnCl2
Zn(OTf)2
LiBr
1.3
3.0
6.0
60.0
96[d]
72[d]
1.3
99
95
95
79
10
67
98
99
42
32
63
31
À76
69
Stereoselective (3+2) cycloaddition
The formation of cis-thiazolidines 11 and 13 with excellent
chemo-, regio-, and diastereoselectivity intimated the potential
to develop a stereoselective reaction manifold. Initial develop-
ment focused on the (3+2) cycloaddition of (R)-N-tosyl-2-phe-
nylaziridine ((R)-1a) with allyl isothiocyanate (Table 5).[22] The
optimized zinc(II) bromide mediated reaction conditions fur-
nished (S)-2a in excellent yield and with 42% enantiomeric
excess (ee) (Table 5, entry 1). Other zinc(II) halide salts furnished
the desired product in similar yield with increased ee when
zinc(II) chloride was employed (entries 2–3). Zinc(II) triflate pro-
vided no improvement in the ee of (S)-2a (entry 4). Lithium
bromide mediated and catalytic zinc(II) bromide conditions
both exhibited incomplete conversion of (R)-1a (entries 5–6).
Interestingly, while the lithium bromide conditions furnished
the opposite enantiomer (R)-2a, the catalytic zinc(II) bromide
conditions produced (S)-2a with an improved 69% ee. Inspired
by this result, we were pleased to find that increasing the
[e]
ZnBr2
[f]
ZnBr2
65
94
[g]
ZnCl2
2.0
[a] Conditions: aziridine (R)-1a (0.40 mmol, 99% ee), isothiocyanate
(0.80 mmol), Lewis acid (0.50 mmol), CH2Cl2 (0.80 mL). [b] Isolated yield.
[c] Determined by analytical chiral SFC. [d] Starting material was not fully
consumed. [e] 0.12 mmol of ZnBr2. [f] 6.00 equiv allyl isothiocyanate.
[g] 10.0 equiv allyl isothiocyanate.
equivalents of isothiocyanate in the presence of stoichiometric
zinc(II) bromide could provide a similar boost in ee while main-
taining full conversion of the starting material (entry 7). Ulti-
mately, the use of 1.25 equivalents of zinc(II) chloride and
10.0 equivalents of allyl isothiocyanate in dichloromethane at
ambient temperature proved optimal, furnishing (S)-2a in 99%
yield and with 94% ee (entry 8).
Exploration of isothiocyanate
substitution
With optimal conditions identi-
fied, we examined the scope of
heterocumulene substitution in
the reaction.[23] We found that,
along with allyl isothiocyanate,
primary and secondary alkyl iso-
thiocyanates were all highly
compatible under the reaction
conditions, furnishing desired
enantioenriched iminothiazoli-
dines (S)-2a, (S)-2k, and (S)-2l in
uniformly excellent yields and ee
(Table 6).[13] The use of a tertiary
isothiocyanate, however, extend-
ed the reaction time and provid-
ed thiazolidine (S)-2p in de-
creased yield and ee. Additional-
ly, aryl isothiocyanates were
competent cycloaddition reac-
tion partners under the zinc(II)
chloride mediated conditions,
providing thiazolidine products
(S)-2m, (S)-2n, and (S)-2o in ex-
Scheme 3. Diastereoselective (3+2) cycloaddition with aziridine 12.
Scheme 4. (3+2) cycloaddition with aziridine 15.
Chem. Eur. J. 2014, 20, 4806 – 4813 www.chemeurj.org
4809
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 6. Substrate scope of stereoselective isothiocyanate (3+2) cycload-
dition.[a]
Table 7. N-substitution scope of stereoselective isothiocyanate (3+2)
cycloaddition.[a]
Entry
R
Product
t [h]
Yield [%][b]
ee [%][c]
1
2
3
4
5
tosyl
p-NO2-C6H4-SO2
mesyl
p-OMe-C6H4-SO2
H
(S)-2a
(S)-8a
(S)-8b
(S)-8c
(S)-8d
2.0
1.8
4.0
5.0
0.5
99
>99
95
94
31
94
95
90
91
34
[a] Conditions: aziridine (R)-7 (0.40 mmol, 99% ee), isothiocyanate
(4.00 mmol), ZnCl2 (0.50 mmol), CH2Cl2 (0.80 mL). [b] Isolated yield. [c] De-
termined by analytical chiral SFC.
Proposed mechanism of stereoselective (3+2) cycloaddition
We hypothesize that the mechanism of the (3+2) cycloaddi-
tions presented herein proceeds through a stereoselective inti-
mate-ion-pair mechanism similar to that invoked in our previ-
ous work[10] and by Johnson[24] and Kerr[25] in related work on
the cycloadditions of donor–acceptor cyclopropanes
(Scheme 5). Our observations including lack of reactivity in the
[a] Conditions: aziridine (R)-1 (0.40 mmol, 99% ee), isothiocyanate
(4.00 mmol), ZnCl2 (0.50 mmol), CH2Cl2 (0.80 mL). [b] Isolated yield. [c] De-
termined by analytical chiral SFC.
cellent yields. While phenyliminothiazolidine (S)-2m was isolat-
ed with good ee, the use of a more electron-rich isothiocya-
nate resulted in the formation of a thiazolidine product
((S)-2n) with an excellent 90% ee, whereas the use of a more
electron-deficient isothiocyanate provided a product with a sig-
nificantly lower ee ((S)-2o).
Scheme 5. Proposed general mechanism for stereoselective (3+2) cycloaddi-
tion.
absence of Lewis acid,[11,26] inversion at the benzylic position,
greater reactivity of aziridines with electron-rich aryl substitu-
ents, and shorter reaction times of N-substituted aziridines
with more electron-withdrawing groups are all consistent with
this mechanistic hypothesis. The formation of (R)-2a under lith-
ium bromide mediated conditions strongly suggests we have
developed a Lewis acid mediated process in contrast to the re-
lated alkali metal halide mediated system reported by Nadir
and co-workers,[6k,9] who observe overall stereoretention as
a result of a double inversion pathway, which proceeds
through an iodinated intermediate, and the palladium(II)-cata-
lyzed reaction conditions of Alper and co-workers[6i] who also
observe the stereoretentive product as the major enantiomer.
This hypothesis also accounts for the observed nondiastereose-
lective formation of thiazolidines 16 and 17 and pyrrolines 18
and 19, considering the fully separated ion-pair intermediate
that likely results from destabilization of the polarized CÀN
bond by the steric interaction between the cis substituents on
aziridine 15 (see Scheme 4).
Effect of aziridine N-substitution on the transfer of chiral in-
formation
We subsequently investigated the effect of N-substitution on
the aziridine in the stereoselective (3+2) cycloaddition. Aziri-
dines with N-sulfonyl substitutions were extremely well tolerat-
ed under the reaction conditions, furnishing the desired thiazo-
lidines ((S)-2a, (S)-8a–c) in excellent yield with uniformly high
ee’s, exhibiting reaction times that increased slightly moving
from more electron-deficient (Table 7, entry 2) to more elec-
tron-rich (entry 4) sulfonyl groups. While unprotected 2-phenyl-
aziridine showed an improved reaction time, unfortunately
both the yield and ee of thiazolidine (S)-8d were reduced
(entry 5).
Chem. Eur. J. 2014, 20, 4806 – 4813
www.chemeurj.org
4810
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Cycloaddition of a diester aziridine
imine, allyl group, and secondary nitrogen could be envisioned
to provide rapid access to highly enantioenriched oxothiazoli-
dines, thiazoles, and a variety of polyclic scaffolds.[34] Critically,
access to the enantioenriched secondary thiazolidine and imi-
dazolidine products enabled by our (3+2) cycloaddition allows
for their use as asymmetric organic catalysts, as the free secon-
dary nitrogen functions as a necessary hydrogen bond donor
for the majority of these applications.[7a–d]
Noting the apparent mechanistic similarities with our previous
work,[10] we synthesized diester aziridine 22 to assess the po-
tential for selective activation of the CÀC or CÀN bond under
either our tin(II)- or zinc(II)-mediated conditions, respectively.[27]
Unfortunately, Sn(OTf)2 failed to provide any cycloaddition
product.[28] Alternatively, use of ZnBr2 provided thiolactam 23
as the sole (3+2) adduct (Scheme 6a). This is the only thiolac-
Conclusion
We have disclosed the first ste-
reoselective Lewis acid mediated
(3+2) cycloaddition of N-H- and
N-sulfonylaziridines with alkyl
heterocumulenes. These zinc(II)-
mediated conditions offer broad
tolerance of alkyl, silyl, and aryl
heterocumulenes, as well as azir-
idine substitution, enabling the
formation of iminoimidazolidines
and enantioenriched iminothia-
zolidines in overall excellent
yields from enantioenriched azir-
idines, which are easily accessi-
ble from their amino acid pre-
cursors. Combined with the ex-
Scheme 6. (3+2) cycloaddition with diester aziridine 22.
hibited ability to simply and or-
thogonally remove the sulfonyl
tam (3+2) cycloaddition product observed during our stud-
ies.[13] The ability of the malonate group to stabilize the nega-
tive charge and the nitrogen to further stabilize the benzylic
positive charge allows for the formation of zwitterion 24, likely
resulting in the observed divergent reactivity (Scheme 6b).[29]
and allyl protecting groups, this reaction system enables the
installation of a broad number of functional group handles for
further derivatization of these biologically and catalytically im-
portant heterocyclic scaffolds.
Deprotection of N-sulfonyl heterocycles
Having explored a range of substrates, heterocumulenes, and
N-protecting groups, we next sought to assess the potential to
derivatize our heterocyclic products. As expected, the N-sulfo-
nyl protecting group removal was trivial. Secondary iminothia-
zolidine (S)-8d could be accessed rapidly in an excellent 91%
yield without any loss of enantiomeric excess through detosy-
lation of thiazolidine (S)-2a (Scheme 7a)[30] or through the de-
sulfonylation of p-nosyl-protected thiazolidine (S)-8a with
slightly increased yield (Scheme 7b), furnishing thiazolidine
(S)-8d in 87% yield and with 94% ee over two steps from (R)-
N-(p-nitrobenzenesulfonyl)-2-phenylaziridine.[31]
Alternatively, cleavage of the allyl imine CÀN bond of het-
erocycle (S)-2a in the presence of palladium(0) enabled access
to secondary iminothiazolidine (S)-25 with some loss of enan-
tiomeric excess (Scheme 7c). While the cleavage of N-allyl
bonds has been shown in the literature for functional groups
including amines[32] and amides,[33] this is the first known exam-
ple of the cleavage of an imino N-allyl bond.
Iminothiazolidines (S)-8d and (S)-25 are extremely versatile
heterocycles. Derivatization and synthetic manipulations of the
Scheme 7. Desulfonylation and deallylation of iminothiazolidine products.
Chem. Eur. J. 2014, 20, 4806 – 4813
www.chemeurj.org
4811
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[6] a) M. K. Ghorai, K. Ghosh, K. Das, Tetrahedron Lett. 2006, 47, 5399–5403;
b) M. K. Ghorai, K. Das, A. Kumar, K. Ghosh, Tetrahedron Lett. 2005, 46,
4103–4106; c) B. A. B. Prasad, G. Pandey, V. K. Singh, Tetrahedron Lett.
2004, 45, 1137–1141; d) S. Gandhi, A. Bisai, B. A. B. Prasad, V. K. Singh, J.
Org. Chem. 2007, 72, 2133–2142; e) V. K. Yadav, V. Sriramurthy, J. Am.
Chem. Soc. 2005, 127, 16366–16367; f) M. Sengoden, T. Punniyamurthy,
Angew. Chem. 2013, 125, 600–603; Angew. Chem. Int. Ed. 2013, 52,
572–575; g) T. Munegumi, I. Azumaya, T. Kato, H. Masu, S. Saito, Org.
Lett. 2006, 8, 379–382; h) M. S. Kim, Y.-W. Kim, H. S. Hahm, J. W. Jang,
W. K. Lee, H.-J. Ha, Chem. Commun. 2005, 3062–3064; i) J.-O. Baeg, C.
Bensimon, H. Alper, J. Am. Chem. Soc. 1995, 117, 4700–4701; j) J.-Y. Wu,
Z.-B. Luo, L.-X. Dai, X.-L. Hou, J. Org. Chem. 2008, 73, 9137–9139;
k) U. K. Nadir, S. Joshi, Indian J. Chem. Sect. B: Org. Chem. Incl. Med.
Chem. 2003, 42B, 1760–1764.
[7] a) E. J. Corey, M. J. Grogan, Org Lett. 1999, 1, 157–160; b) V. B. Birman,
X. Li, Org. Lett. 2006, 8, 1351–1354; c) V. B. Birman, H. Jiang, X. Li, L.
Guo, E. W. Uffman, J. Am. Chem. Soc. 2006, 128, 6536–6537; d) W. Ye, D.
Leow, S. L. M. Goh, C.-T. Tan, C.-H. Chian, C.-H. Tan, Tetrahedron Lett.
2006, 47, 1007–1010.
[8] a) V. S. Jain, D. K. Vora, C. S. Ramaa, Bioorg. Med. Chem. 2013, 21, 1599–
1620; b) Y. Pandey, P. K. Sharma, N. Kumar, A. Singh, Int. J. PharmTech
Res. 2011, 3, 980–985; c) M. D’hooghe, N. De Kimpe, Tetrahedron 2006,
62, 513–535; d) E. H. Schweizer, F. Mꢁrki, C. Lehmann, H. Dietrich, J.
Med. Chem. 1983, 26, 964–970.
Experimental Section
Representative general procedure for the (3+2) cycloaddi-
tion of N-H- and N-sulfonylaziridines with heterocumulenes
To an oven-dried 1-dram vial equipped with a magnetic stir bar
was added zinc(II) bromide (113 mg, 0.50 mmol, 1.25 equiv), freshly
powdered with a mortar and pestle, in an inert atmosphere glove-
box. The vial was sealed with a screw cap fitted with a Teflon
septum, removed from the glovebox, and placed under an inert at-
mosphere. To a separate, oven-dried 1-dram vial was added
N-tosyl-2-phenylaziridine (1a, 109 mg, 0.40 mmol, 1.00 equiv). The
vial was then sealed with a screw-cap fitted with a Teflon septum
and anhydrous CH2Cl2 (0.60 mL) and allyl isothiocyanate (79 mL,
0.80 mmol, 2.00 equiv) were added. The mixture was transferred to
the first vial with a rinse of anhydrous CH2Cl2 (0.20 mL). The hetero-
geneous reaction mixture was then allowed to stir at ambient tem-
perature. Upon consumption of the aziridine (determined by TLC
analysis), the reaction solution was diluted with CH2Cl2 (3 mL) and
CH3OH (1 mL), adsorbed onto Celite, and purified by silica gel
column chromatography (20% acetone in hexanes eluent) to fur-
nish thiazolidine 2a (147 mg, 99% yield) as an amorphous white
solid.
[9] a) U. K. Nadir, N. Basu, Tetrahedron 1993, 49, 7787–7792; b) U. K. Nadir,
N. Basu, Tetrahedron Lett. 1992, 33, 7949–7952.
[10] A. F. G. Goldberg, N. R. O’Connor, R. A. Craig,II, B. M. Stoltz, Org. Lett.
2012, 14, 5314–5317.
[11] No reaction was observed in the absence of Lewis acid. For full details,
see the Supporting Information.
Acknowledgements
The authors wish to thank NIH-NIGMS (R01M080269—01),
Amgen, and Caltech for financial support. R.A.C. gratefully ac-
knowledges the support of this work provided by a fellowship
from the National Cancer Institute of the National Institutes of
Health under Award Number F31A174359. A.F.G.G. thanks the
Natural Sciences and Engineering Research Council (NSERC) of
Canada for a PGS D scholarship. L. Henling (Caltech) and Dr. M.
Takase (Caltech) are gratefully acknowledged for X-ray crystal-
lographic structural determination. The Bruker KAPPA APEXII
X-ray diffractometer was purchased via an NSF CRIF:MU award
to the California Institute of Technology, CHE-0639094.
[12] The Z-imino stereochemistry of the iminothiazolidine products was as-
signed by analogy to the structure of thiazolidine 11, which was unam-
biguously established by single-crystal X-ray diffraction. For full details,
see the Supporting Information.
[13] Thiourea product 3 was not observed in any case during this study.
Similarly, the analogous thiolactam product was not observed during
our previous studies of (3+2) cycloadditions of donor–acceptor cyclo-
propanes with heterocumulenes, see ref. [10].
[14] This observation is in stark contrast to our previous studies of (3+2) cy-
cloadditions of donor–acceptor cyclopropanes with heterocumulenes,
in which increased steric bulk on the ring caused a large increase in re-
action time, see ref. [10].
[15] a) C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195; b) L. P.
Hammett, J. Am. Chem. Soc. 1937, 59, 96–103.
[16] This was the only substrate found to be susceptible to nucleophilic ring
opening by the zinc(II) counterion during our studies. For full details,
see the Supporting Information.
Keywords: cycloadditions · heterocumulenes · heterocycles ·
Lewis acids · stereoselectivity
[17] A variety of isocyanates were tested under zinc(II) bromide mediated
conditions, including trimethylsilyl, phenyl, and allyl isocyanates.
[18] The assignment of the E-imino-stereochemistry of the iminoimidazoli-
dine products was assigned by analogy to a structure related to imida-
zolidine 9a, which was unambiguously established by single-crystal
X-ray diffraction of the corresponding imidazolidinium ion. For full de-
tails, see the Supporting Information.
[19] Inversion at the benzylic position was observed in our previous studies
of stereoselective (3+2) cycloadditions of donor–acceptor cyclopro-
panes with heterocumulenes, see ref. [10].
[1] a) A. Padwa, in Comprehensive Heterocyclic Chemistry III, Vol. 1 (Eds: A. R.
Katritzky, C. A. Ramsden, E. F. V. Scriven, R. J. K. Taylor), Elsevier, Oxford,
2008, pp. 1–105; b) W. McCoull, F. A. Davis, Synthesis 2000, 1347–1365;
c) D. Tanner, Angew. Chem. 1994, 106, 625–646; Angew. Chem. Int. Ed.
Engl. 1994, 33, 599–619.
[2] a) W. H. Pearson, B. W. Lian, S. C. Bergmeier, in Comprehensive Heterocy-
clic Chemistry II, Vol. 1 A (Eds: A. R. Katritzky, C. W. Rees, E. F. V. Scriven),
Elsevier, Oxford, 1996, p. 1; b) T. L. Gilchrist, Heterocyclic Chemistry, 2nd
ed., Longman, Essex, 1992, p. 38.
[20] Alternatively, trans-aziridine 12 did not undergo this transformation
under the same conditions, although both styryl- and acroylaziridines
are known to isomerize to pyrrolines, see: a) M. Brichacek, M. N. Villalo-
bos, A. Plichta, J. T. Njardarson, Org. Lett. 2011, 13, 1110–1113; b) A.-H.
Li, L.-X. Dai, X.-L. Hou, J. Chem. Soc. Perkin Trans. 1 1996, 2725–2729;
c) R. S. Atkinson, C. W. Rees, J. Chem. Soc. Chem. Commun. (London)
1967, 1232.
[21] N-Tosyl-2-methyl-2-phenylaziridine (26) proved susceptible to isomeriza-
tion under the reaction conditions, furnishing no (3+2) adduct, but
rather enamine 27 as a mixture of E and Z isomers. For full details, see
the Supporting Information.
[3] a) R. Chawla, A. K. Singh, L. D. S. Yadav, RSC Adv. 2013, 3, 11385–11403;
b) M. K. Ghorai, D. P. Tiwari, N. Jain, J. Org. Chem. 2013, 78, 7121–7130;
c) X. E. Hu, Tetrahedron 2004, 60, 2701–2743.
[4] a) P. Davoli, I. Moretti, F. Prati, H. Alper, J. Org. Chem. 1999, 64, 518–521;
b) S. V. Ley, B. Middleton, Chem. Commun. 1998, 1995–1996; c) K.
Khumtaveeporn, H. Alper, Acc. Chem. Res. 1995, 28, 414–422.
[5] a) A. L. Cardoso, T. M. V. D. Pinho e Melo, Eur. J. Org. Chem. 2012, 6479–
6501; b) D. M. Hodgson, P. G. Humphreys, J. G. Ward, Org. Lett. 2006, 8,
995–998; c) W. Zhu, G. Cai, D. Ma, Org. Lett. 2005, 7, 5545–5548; d) J.
Huang, P. O’Brien, Chem. Commun. 2005, 5696–5698; e) E. A. Acar, F.
Glarner, U. Burger, Helv. Chim. Acta 1998, 81, 1095–1104; f) G. Cardillo,
L. Gentilucci, A. Tolomelli, C. Tomasini, Tetrahedron Lett. 1997, 38, 6953–
6956.
[22] Absolute stereochemistry of the iminothiazolidine products were as-
signed by analogy to the structure of thiazolidine (S)-8a, which was un-
Chem. Eur. J. 2014, 20, 4806 – 4813
www.chemeurj.org
4812
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ambiguously established by single-crystal X-ray diffraction. For full de-
tails, see the Supporting Information.
[29] The same mode of reactivity has been previously observed in (3+2) cy-
cloadditions with the N-alkylated and N-arylated derivatives of aziridine
22 under thermolysis conditions with isothiocyanates, see: H. Ben-
haoua, F. Texier, Tetrahedron 1978, 34, 1153–1161.
[30] D. A. Alonso, P. G. Andersson, J. Org. Chem. 1998, 63, 9455–9461.
[31] a) R. S. Narayan, M. S. VanNieuwenhze, Org. Lett. 2005, 7, 2655–2658;
b) T. Kan, T. Fukuyama, Chem. Commun. 2004, 353–359; c) T. Fukuyama,
C.-K. Jow, M. Cheung, Tetrahedron Lett. 1995, 36, 6373–6374.
[32] a) S. Chandrasekhar, C. R. Reddy, R. J. Rao, Tetrahedron 2001, 57, 3435–
3438; b) M. Honda, H. Morita, I. Nagakura, J. Org. Chem. 1997, 62,
8932–8936; c) F. Garro-Helion, A. Merzouk, F. Guibe, J. Org. Chem. 1993,
58, 6109–6113.
[23] As with the zinc(II) bromide mediated conditions, isocyanates were not
suitable cycloaddition partners under our stereospecific reaction condi-
tions. Diphenylcarbodiimide was compatible with the zinc(II) chloride
mediated conditions, furnishing the iminoimidazolidine product in ex-
cellent yield, but with comparatively reduced ee. For full details, see the
Supporting Information.
[24] a) P. D. Pohlhaus, S. D. Sanders, A. T. Parsons, W. Li, J. S. Johnson, J. Am.
Chem. Soc. 2008, 130, 8642–8650; b) P. D. Pohlhaus, J. S. Johnson, J.
Am. Chem. Soc. 2005, 127, 16014–16015.
[25] a) A. Karadeolian, M. A. Kerr, J. Org. Chem. 2007, 72, 10251–10253; b) K.
Sapeta, M. A. Kerr, J. Org. Chem. 2007, 72, 8597–8599; c) D. Wanapun,
K. A. Van Gorp, N. J. Mosey, M. A. Kerr, T. K. Woo, Can. J. Chem. 2005, 83,
1752–1767.
[33] a) T. Matsumura, M. Akiba, S. Arai, M. Nakagawa, A. Nishida, Tetrahedron
Lett. 2007, 48, 1265–1268; b) T. Koch, M. Hesse, Synthesis 1992, 931–
932.
[34] a) M. Jasinski, G. Mloston, H. Heimgartner, J. Heterocycl. Chem. 2010, 47,
1287–1293; b) K. Keiji, K. Takanobu, K. Masaki, S. Hiroki, Thiazoline De-
rivative and Use of the Same. Eur. Pat. Appl. 1669352, 2006; c) P. Pecor-
ari, M. Rinaldi, L. Costantino, A. Provvisionato, C. Cermelli, M. Portolani,
Farmaco 1991, 46, 899–911; d) R. Crossley, 2,3-Dihydro- Thiazolo- and
Thiazino- Benzimidazoles as Anti-Hyper Secretion Agents. U. S. Patent
4,873,237, October 10, 1989; e) V. Skaric, D. Skaric, A. Cizmek, J. Chem.
Soc. Perkin Trans. 1 1984, 2221–2225; f) G. R. Brown, J. Chem. Soc.
Perkin Trans. 1 1973, 2022–2024.
[26] In the absence of heterocumulene, aziridine (R)-1a was completely
racemized in 10 min. Concurrently, no reduction in ee of (S)-2a was ob-
served upon re-exposure to the reaction conditions. For full details, see
the Supporting Information.
[27] The selective CÀC bond cleavage of N-tosyl-3-arylaziridine-2,2-dicarbox-
ylates has been explored previously in the literature, see: a) L. Li, J.
Zhang, Org. Lett. 2011, 13, 5940–5943; b) Z. Jiang, J. Wang, P. Lu, Y.
Wang, Tetrahedron 2011, 67, 9609–9617; c) P. D. Pohlhaus, R. K.
Bowman, J. S. Johnson, J. Am. Chem. Soc. 2004, 126, 2294–2295.
[28] Conditions: aziridine 22 (0.40 mmol), allyl isothiocyanate (0.80 mmol),
Sn(OTf)2 (0.44 mmol), CH2Cl2 (1.3 mL). See reference [10] and general
procedure D in the associated Supporting Information.
Received: September 21, 2013
Published online on March 6, 2014
Chem. Eur. J. 2014, 20, 4806 – 4813
www.chemeurj.org
4813
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim