Iron-catalyzed kinetic resolution of N-sulfonyl oxaziridines

Article abstract of DOI:10.1039/c4sc01371a

We have developed a highly selective kinetic resolution of N-sulfonyl oxaziridines. This reaction utilizes an inexpensive and easily synthesized iron bis(oxazoline) catalyst to promote the efficient rearrangement of oxaziridines to the corresponding N-sulfonyl imides; no sacrificial reagents are required to effect this resolution. This process is readily translated to gram scale, which provides a practical method for the preparation of structurally diverse, enantiopure N-sulfonyl oxaziridines for use as reagents in organic synthesis.

Full text of DOI:10.1039/c4sc01371a

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Iron-catalyzed kinetic resolution of N-sulfonyl
oxaziridines†
Cite this: Chem. Sci., 2014, 5, 3524
*
Kevin S. Williamson, James W. Sawicki and Tehshik P. Yoon
We have developed a highly selective kinetic resolution of N-sulfonyl oxaziridines. This reaction utilizes an
inexpensive and easily synthesized iron bis(oxazoline) catalyst to promote the efficient rearrangement of
oxaziridines to the corresponding N-sulfonyl imides; no sacrificial reagents are required to effect this
resolution. This process is readily translated to gram scale, which provides a practical method for the
preparation of structurally diverse, enantiopure N-sulfonyl oxaziridines for use as reagents in organic
synthesis.
Received 10th May 2014
Accepted 13th June 2014
DOI: 10.1039/c4sc01371a
www.rsc.org/chemicalscience
these reaction conditions. Catalytic imine oxidations using a
chiral Lewis acid or hydrogen-bonding catalyst developed by
Introduction
Yamamoto10 and Ooi,11 respectively, provide high ee's only with
N-tosyl or N-mesityl imines. This presents a signicant limita-
tion in the context of our studies because of the large inuence
of the N-substituent on the reactivity of oxaziridines. Finally,
Feng and Liu recently reported a quite general kinetic resolu-
tion of racemic oxaziridines,12,13 but the need for a nontrivial
azlactone as a stoichiometric nucleophilic reagent renders this
method impractical for preparation of oxaziridines on gram
scale. Given the recent renewed interest in the chemistry of
these reactive heterocycles,14 we sought to develop a more
general approach to the preparation of enantiopure N-sulfonyl
oxaziridines that would tolerate a broad range of structural
variation and would be amenable to scale-up.
Oxaziridines, rst prepared by Emmons in 1957,1 are fasci-
nating three-membered heterocycles composed of nitrogen,
oxygen, and carbon. Much of the early work on the chemistry of
oxaziridines focused upon the unusual physical properties that
arise from the presence of two adjacent heteroatoms in a
strained three-membered ring.2 For example, the rate of pyra-
midal inversion in oxaziridines is slow enough to allow the
isolation of enantioenriched oxaziridines chiral only at
nitrogen.3 From a synthetic perspective, oxaziridines have long
been recognized as useful oxidants: they are generally air-stable
crystalline materials that mediate a wide variety of powerful
oxidative functional group transformations.4 Our laboratory has
a long-standing interest in developing catalytic methods that
exploit the unique reactivity of N-sulfonyl oxaziridines to
perform oxyamination, dipolar cycloaddition, and C–H amina-
tion reactions.5 We are particularly interested in investigating
stereocontrolled versions of these reactions,6 but our studies
have oen been hampered by the fact that only a few highly
enantioselective methods for the synthesis of chiral oxazir-
idines have been reported to date.
Our on-going investigations of structure-activity relation-
ships in stereoselective reactions of N-sulfonyl oxaziridines
require access to gram quantities of a wide variety of structurally
diversied oxaziridines in enantiopure form, but each of the
previously known methods have proven to be problematic for
our needs. Stoichiometric oxidations using chiral peracids7
generally afford low ee's. The organocatalytic phase-transfer
reactions recently reported by Jorgensen8 and Jin9 scaled up
poorly in our hands, which we attributed to the heterogeneity of
Results and discussion
Two well-established features of oxaziridine chemistry guided
our reaction design. First, the racemic synthesis of N-sulfonyl
oxaziridines by oxidation of the corresponding imines is robust,
tolerant of great structural diversity, and is scalable to hundreds
of grams.15 Thus, the racemates of a diverse range of oxazir-
idines are readily prepared in substantial quantities. Second,
oxaziridines are rapidly isomerized to the corresponding
amides in the presence of redox-active transition metal cata-
lysts.16 We speculated that an appropriate chiral Fe(II) complex17
might catalyze the kinetic resolution of racemic oxaziridines.
Such a process would satisfy the criteria for an ideal kinetic
resolution as delineated by Jacobsen:18 the resolution would
begin with an easily accessible racemate, it would involve an
inexpensive chiral catalyst, the products would be easily sepa-
rated by chromatography, and the rearrangement would rely on
the intrinsic reactivity of oxaziridines and thus require no
additional stoichiometric reagents.
Department of Chemistry, University of Wisconsin–Madison, 1101 University Avenue,
Madison, WI 53706, USA. E-mail: tyoon@chem.wisc.edu
† Electronic supplementary information (ESI) available: Experimental procedures
and characterization data for all compounds. See DOI: 10.1039/c4sc01371a
3524 | Chem. Sci., 2014, 5, 3524–3527
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
We began our investigation by examining the rearrangement difficult. Instead, apparent s values for the resolution were
of racemic oxaziridine 3 in the presence of series of chiral Fe(II) determined based upon the isolated yield of chromatographi-
bis(oxazoline) complexes (Table 1). This class of catalysts was cally puried oxaziridine. We felt that this would be a reason-
selected because our previous studies had shown that they able approximation because any errors arising from mechanical
provide an excellent stereocontrolling environment for oxazir- loss or uncatalyzed decomposition would result in a net
idine-mediated alkene oxyaminations.6b We observed that the decrease in the apparent s value; the calculated gure would
FeCl2 complex of isopropyl bis(oxazoline) ligand 6a catalysed thus provide a conservative lower limit for the genuine relative
the efficient rearrangement of 3 to imide 4 in a variety of rate.20
solvents (entries 1–4), and we were delighted to nd that the
Guided by this calculation, we performed a screen of other
remaining oxaziridine could be recovered in good to excellent bis(oxazoline) ligands in this reaction (entries 5–10) and
ee. The determination of the relative rate (s factor) of rear- observed optimal s factors using cyclopropane-bridged indanyl
rangement for the enantiomeric oxaziridines proved to be ligand 6g.21 We were interested in increasing the reaction
challenging; normally, s factor is calculated from the ee of the concentration and lowering the ligand loading of this process to
recovered reactant and the conversion according to the minimize the use of chiral material and solvent waste on larger
following relationship,19 where c is the conversion and ee is the scale; we found that these factors had little impact on selectivity
enantiomeric excess of the unreacted chiral substrate:
(entry 11). Finally, we observed that the initial addition of
catalyst to oxaziridine resulted in a modest but noticeable
evolution of heat. Controlling the temperature by beginning the
reaction at ꢀ78 ꢂC and then warming the solution to 0 ꢂC as the
reaction progressed improved the reproducibility and the
selectivity of the reaction (entry 12). Using this protocol, we
consistently observed good ee's of recovered oxaziridine aer
just 5 min of treatment with 5 mol% of the Fe(II) catalyst.
Table 2 summarizes experiments designed to probe the
effect of the oxaziridine N-substituent on the efficiency of the
kinetic resolution under these conditions. As we had hoped, the
reaction is relatively insensitive to the structure of the N-
sulfonyl moiety. A variety of substituted N-benzensulfonyl
groups were tolerated, and similar s values were obtained in
each case (entries 1–5), regardless of the electronic or steric
properties of the aryl group.22 N-t-Butylsulfonyl oxaziridines,
however, afforded somewhat more modest selectivity (entry 6),
and a resolution of an N-alkyl oxaziridine was essentially
unselective (entry 7).
ln½ð1 ꢀ cÞð1 ꢀ eeÞꢁ
s ¼
ln½ð1 ꢀ cÞð1 þ eeÞꢁ
However, several oxaziridines examined in our study undergo
slow, uncatalyzed decomposition in solution, which renders
accurate determination of the catalyst-mediated conversion
Table
1 Optimization of conditions for kinetic resolution of
oxaziridinesa
Table
2 Effect of N-substituents on the kinetic resolution of
oxaziridinesa
Entry
Solvent
Ligand
Time
Yieldb
eec
sd
1
2
3
4
5
6
7
8
Toluene
CH2Cl2
MeCN
THF
THF
THF
THF
THF
THF
THF
6a
6a
6a
6a
6b
6c
6d
6e
6f
2 h
2 h
2 h
2 h
1 h
1 h
1 h
1 h
30 min
15 min
5 min
5 min
43%
19%
29%
21%
95%
78%
79%
32%
38%
41%
36%
38%
70%
97%
60%
99%
4%
14%
17%
98%
80%
96%
98%
99%
7
5
3
7
8
Entry
R
Time
Yieldb
eec
sd
3
5
1
2
3
4
5
6
7
Ns (3)
Ts (7)
Bs (8)
2 Ns (9)
SO2Mes (10)
Bus (11)
i-Pr (12)
5 min
7 min
6 min
2 min
13 min
5 min
5 min
38%
41%
42%
40%
45%
55%
33%
99%
93%
91%
96%
89%
57%
48%
17
16
15
17
18
10
2
11
6
19
15
17
9
10
11e
12e,f
6g
6g
6g
THF
THF
a
Unless otherwise noted, reactions performed at 0 ꢂC on 0.5 mmol scale
b
a
(0.2 M) using 5 mol% FeCl2 and 10 mol% ligand. Isolated yield of
Reactions performed on 1.0 mmol scale (0.5 M). Ns
¼
4-
c
recovered oxaziridine. Enantiomeric excess determined by chiral SFC nitrobenzenesulfonyl; Ts ¼ 4-toluenesulfonyl; Bs ¼ benzenesulfonyl; 2
d
analysis. Calculated based upon yield of recovered oxaziridine Ns ¼ 2-nitrobenznesulfonyl; SO2Mes ¼ 2-mesitylenesulfonyl; Bus ¼ t-
e
b
c
according to ref. 17. Reaction conducted at 0.5 M concentration butylsulfonyl. Isolated yield of recovered oxaziridine. Enantiomeric
using 6 mol% ligand. f Reaction conducted on 1.0 mmol scale starting excess determined by chiral SFC analysis. Calculated based upon
d
at ꢀ78 ꢂC and warming to 0 ꢂC.
yield of recovered oxaziridine according to ref. 19.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 3524–3527 | 3525

View Article Online
Chemical Science
Edge Article
Table 3 outlines the scope of the reaction with respect to the
C-aryl substituent of the oxaziridine. The reaction is tolerant of
substitution at the para-, meta-, and ortho- positions of the aryl
ring (13–15); however, we observed that resolution of oxazir-
idines bearing ortho-substituted C-aryl groups required higher
catalyst loadings and reacted with slightly lower selectivity. The
reaction is tolerant of aryl rings bearing electron-withdrawing
groups (16–18). In general, we have found N-sulfonyl oxazir-
idines bearing electron-rich C-aryl groups to be unstable; those
bearing strongly electron-donating para substituents can
decompose violently at room temperature.23 Those bearing less
strongly electron-donating groups gave lower apparent s values,
which we rationalize as an artefact of the increased rate of their
uncatalyzed decomposition in solution compared to other more
stable oxaziridines (19 and 20). More extended condensed
aromatic groups also afforded somewhat lower selectivity (22),
as does a C-peruoroaryl oxaziridine (21). Finally, C-alkyl
Scheme 1 Gram-scale resolutions of N-sulfonyl oxaziridines.
oxaziridines (23) react at much slower rates and with more
modest selectivity than C-aryl oxaziridines, and this class of
oxaziridines represents a limitation of the method.
A principal motivation for this study was our need for a route
to structurally diverse enantiopure oxaziridines on signicant
scales. Gratifyingly, we found that this protocol for kinetic
resolution of oxaziridines is readily scalable (Scheme 1). The
gram-scale resolution of N-benzenesulfonyloxaziridine (ꢃ)-8
required somewhat higher reaction temperatures (0 ꢂC to
ambient temperature), but the enantioselectivity of the process
remained excellent even with lower catalyst loading, affording 8
on gram-scale in 98% ee. One of the most important advantages
of kinetic resolutions over other forms of asymmetric catalysis
is the ability to offset lower intrinsic stereoselectivity by
increasing ee at the expense of yield by running the resolution
to higher conversion. Thus, we examined the large-scale reso-
lution of racemic N-Bus oxaziridine (ꢃ)-11. Despite the lower s
factors observed for this oxaziridine compared to analogous N-
arenesulfonyl oxaziridines in our small-scale studies (Table 2,
entry 6), we were able to obtain 0.45 g of enantiomerically pure
oxaziridine aer a 20 min reaction time. Thus, by taking
Table 3 Scope of the kinetic resolutiona
advantage of
a feature inherent to kinetic resolutions,
substantial quantities of highly enantioenriched oxaziridines
can be obtained even when their small-scale resolutions
proceed with moderate s factors. These results suggest that this
methodology will indeed provide access to a wide range of
structurally diverse oxaziridines on synthetically relevant scales,
which will enable the investigation of the effect of stereo-
chemistry on a number of oxaziridine-mediated reactions.
Conclusions
To summarize, we have developed a scalable and robust iron-
catalyzed kinetic resolution of oxaziridines using an inexpen-
sive and readily available iron bis(oxazoline) catalyst. This
process is general for a broad range of N-sulfonyl oxaziridines,
the racemates are readily accessible via a robust and convenient
synthetic route, and the resolution is easily translated to the
gram scale without loss of selectivity. Importantly, this kinetic
resolution requires no sacricial reagents other than the oxa-
ziridine itself, which renders this a practical method for the
preparation of structurally diverse enantiopure oxaziridines as
reagents for organic synthesis. The use of these reagents in new
oxidative transformations is a continuing theme of research in
our laboratory.
a
Unless otherwise noted, all reactions were performed on 1.0 mmol
scale using 5 mol% FeCl2 and 6 mol% ligand 6g, ꢀ78 ꢂC to 0 ꢂC.
b
Reaction conducted using 10 mol% FeCl2 and 12 mol% ligand 6g.
c
Reaction conducted from 0 ꢂC to 23 ꢂC.
3526 | Chem. Sci., 2014, 5, 3524–3527
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
12 S. Dong, X. Liu, Y. Zhu, P. He, L. Lin and X. Feng, J. Am.
Chem. Soc., 2013, 135, 10026.
Acknowledgements
13 Kinetic resolution of an N-sulfonyl oxaziridine was also
observed during an enantioselective formal [3+2]
cycloaddition with ketenes. See: P.-L. Shao, X.-Y. Chen and
S. Ye, Angew. Chem., Int. Ed., 2010, 49, 8412.
14 For selected investigations of oxaziridine chemistry recently
reported by other laboratories, see: (a) A. M. Adams and J. Du
We acknowledge the NSF (CHE-1265613) for funding this
research.
Notes and references
1 W. D. Emmons, J. Am. Chem. Soc., 1957, 79, 5739.
2 (a) D. R. Boyd and R. Graham, J. Chem. Soc. (C), 1969, 2648;
(b) J. Bjørgo and D. R. Boyd, J. Chem. Soc., Perkin Trans. 2,
1973, 1575.
3 F. Montanari, I. Moretti and G. Torre, Chem. Commun., 1968,
1694.
¨
Bois, Chem. Sci., 2014, 5, 656; (b) H. F. Motiwala, B. Gulgeze
´
and J. Aube, J. Org. Chem., 2012, 77, 7005; (c) X. G. Peng,
Y. G. Zhu, T. A. Ramirez, B. G. Zhao and Y. Shi, Org. Lett.,
2011, 13, 5244; (d) A. Armstrong and R. S. Cooke, Chem.
Commun., 2002, 904.
4 For reviews, see: (a) F. A. Davis and A. C. Sheppard,
Tetrahedron, 1989, 45, 5703; (b) F. A. Davis and B.-C. Chen,
Chem. Rev., 1992, 92, 919; (c) V. A. Petrov and G. Resnati,
Chem. Rev., 1996, 96, 1809; (d) S. Andreae and E. Schmitz,
Synthesis, 1991, 327; (e) F. A. Davis and R. T. Reddy, Comp.
Heterocycl. Chem. II, ed., A. Padwa, Permagon Press,
Oxford, 1996, vol. 1, p. 365; (f) F. A. Davis, B.-C. Chen and
P. Zhou, Comp. Heterocycl. Chem III, ed. A. R. Katritzky, C.
A. Ramsden, E. F. V. Scriven and R. J. K. Taylor, Elsevier,
Oxford, 2008, vol. 1, p. 559; (g) K. S. Williamson,
D. J. Michaelis and T. P. Yoon, Chem. Rev., 2014, DOI:
10.1021/cr400611n.
5 (a) D. J. Michaelis, C. J. Shaffer and T. P. Yoon, J. Am. Chem.
Soc., 2007, 129, 1866; (b) K. M. Partridge, M. E. Anzovino and
T. P. Yoon, J. Am. Chem. Soc., 2008, 130, 2920; (c) C. P. Allen,
T. Benkovics, A. K. Turek and T. P. Yoon, J. Am. Chem. Soc.,
2009, 131, 12560; (d) K. M. Partridge, I. A. Guzei and
T. P. Yoon, Angew. Chem., Int. Ed., 2010, 49, 930; (e)
K. S. Williamson and T. P. Yoon, J. Am. Chem. Soc., 2010,
132, 4570.
15 L. C. Vishwakarma, O. D. Stringer and F. A. Davis, Org.
Synth., 1988, 66, 203.
16 (a) F. Minisci, R. Galli, V. Malatesta and T. Caronna,
Tetrahedron, 1970, 26, 4083; (b) K. Suda, M. Sashima,
M. Izutsu and F. Hino, J. Chem. Soc., Chem. Commun.,
1994, 949.
17 For a recent comprehensive review of chiral iron catalysts,
see: K. Gopalaiah, Chem. Rev., 2013, 113, 3248.
18 J. M. Keith, J. F. Larrow and E. N. Jacobsen, Adv. Synth. Catal.,
2001, 343, 5.
19 For leading references on kinetic resolutions, see: (a)
E. L. Eliel, S. H. Wilen and L. M. Mander, Stereochemistry
of Organic Compounds, Wiley-Interscience, New York, 1994,
p. 395; (b) H. B. Kagan and J. C. Fiaud, in Topics in
Stereochemistry, ed. N. L. Allinger and E. L. Eliel,
Interscience, New York, 1987, vol. 14, p. 249.
20 Jacobsen has also used isolated yield to calculate a lower
limit for s value. See: J. F. Larrow, S. E. Schaus and
E. N. Jacobsen, J. Am. Chem. Soc., 1996, 118, 7420.
21 A two-step, 26 kg scale synthesis of ligand 6g has been
reported. See: D. M. Barnes, J. Ji, M. G. Fickes,
M. A. Fitzgerald, S. A. King, H. E. Morton, F. A. Plagge,
M. Preskill, S. H. Wagaw, S. J. Wittenberger and J. Zhang,
J. Am. Chem. Soc., 2002, 124, 13097.
6 (a) D. J. Michaelis, K. S. Williamson and T. P. Yoon,
Tetrahedron, 2009, 65, 5118; (b) K. S. Williamson and
T. P. Yoon, J. Am. Chem. Soc., 2012, 134, 12370.
7 M. Bucciarelli, A. Forni, S. Marcaccioli, I. Moretti and
G. Torre, Tetrahedron, 1983, 39, 187.
22 The absolute congurations of oxaziridines 7 and 8 were
determined by comparison of their optical rotation to
´
8 L. Lykke, C. Rodrıguez-Escrich and K. A. Jørgensen, J. Am.
Chem. Soc., 2011, 133, 14932.
values reported in ref
8 and 7, respectively. The
9 T. Zhang, W. He, X. Zhao and Y. Jin, Tetrahedron, 2013, 69,
7416.
congurations of all other enantioenriched oxaziridines in
this study were assigned by analogy.
10 J. L. Olivares-Romero, Z. Li and H. Yamamoto, J. Am. Chem.
Soc., 2012, 134, 5440.
11 (a) D. Uraguchi, R. Tsutsumi and T. Ooi, J. Am. Chem. Soc.,
2013, 135, 8161; (b) D. Uraguchi, R. Tsutsumi and T. Ooi,
Tetrahedron, 2014, 70, 1691–1701.
23 F. A. Davis, J. Lamendola Jr, U. Nadir, E. W. Kluger,
T. C. Sedergran, T. W. Panunto, R. Billmers, R. Jenkins Jr,
I. J. Turchi, W. H. Watson, J. S. Chen and M. Kimura, J.
Am. Chem. Soc., 1980, 102, 2000.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 3524–3527 | 3527