Nickel-Catalyzed Enantioselective Reductive Cross-Coupling of Styrenyl Aziridines

Article abstract of DOI:10.1021/jacs.7b03448

A Ni-catalyzed reductive cross-coupling of styrenyl aziridines with aryl iodides is reported. This reaction proceeds by a stereoconvergent mechanism and is thus amenable to asymmetric catalysis using a chiral bioxazoline ligand for Ni. The process allows

Full text of DOI:10.1021/jacs.7b03448

Communication
pubs.acs.org/JACS
Nickel-Catalyzed Enantioselective Reductive Cross-Coupling of
Styrenyl Aziridines
Brian P. Woods,† Manuel Orlandi,‡ Chung-Yang Huang,† Matthew S. Sigman,*,‡
and Abigail G. Doyle*,†
†Department of Chemistry, Princeton University, Princeton, New Jersey 08544, United States
‡Department of Chemistry, University of Utah, 315 S 1400 E, Salt Lake City, Utah 84112, United States
S
* Supporting Information
ABSTRACT: A Ni-catalyzed reductive cross-coupling of
styrenyl aziridines with aryl iodides is reported. This
reaction proceeds by a stereoconvergent mechanism and is
thus amenable to asymmetric catalysis using a chiral
bioxazoline ligand for Ni. The process allows facile access
to highly enantioenriched 2-arylphenethylamines from
racemic aziridines. Multivariate analysis revealed that
ligand polarizability, among other features, influences the
observed enantioselectivity, shedding light on the success
of this emerging ligand class for enantioselective Ni
catalysis.
ziridines are versatile synthetic intermediates for the
1
Apreparation of bioactive targets.
Similar to epoxides,
aziridines experience significant ring strain that enhances their
susceptibility to ring opening with a wide array of nucleophiles.2
Because these reactions are typically stereospecific, the
enantioselective synthesis of amines from aziridines is reliant
on the availability of enantioenriched aziridines. However,
methods for asymmetric catalytic aziridination typically suffer
from modest scope and product enantioselectivity, significantly
lagging behind methods for asymmetric catalytic epoxidation.3
Alternative preparations of enantioenriched aziridines require
multistep routes originating from either the chiral pool (amino
acids) or chiral building blocks (aminoalcohols, epoxides).4
Therefore, catalytic strategies that permit the conversion of
racemic aziridines to enantioenriched amine products in high
yield would represent an attractive alternative.5
In 2012, the Doyle lab introduced the catalytic cross-
coupling of aziridines with organozinc partners (Figure 1a),6
building on seminal stoichiometric precedent from Hillhouse7
and Wolfe.8 Use of an electron-deficient olefin (EDO) as a
ligand for Ni proved critical to achieving high selectivity for C−
C bond formation over competitive β-H elimination.
Subsequently, the groups of Michael,9 Minakata,10 and
Jamison11 described unique and complementary systems for
cross-coupling with aziridines.12 However, these systems, as
well as the stoichiometric precedent from Hillhouse and Wolfe,
share one feature in common: the reactions are stereospecific.13
For example, Minakata demonstrated that enantioenriched
phenethylamine products can be obtained by stereospecific
cross-coupling with enantioenriched styrenyl aziridines. By
contrast, stereochemical studies of our Ni/EDO system
Figure 1. Cross-coupling of aziridines.
revealed that C−C bond formation is stereoablative, presum-
ably via an open shell intermediate.14
In light of this stereochemical outcome, we were intrigued
that a stereoconvergent synthesis of enantioenriched 2-
arylphenethylamines might be possible from racemic styrenyl
aziridines. These products have seen important biological
applications as dopamine receptor agonists15 and have been
generally prepared by asymmetric catalytic conjugate additions
of organometallic reagents to nitroolefins.16 Although enabling,
this approach requires two steps and organometallic reagents.
Our approach would be complementary as the products could
be accessed through a single operation using readily accessible
coupling partners as well as a simple chiral catalyst.
While we were able to obtain proof-of-concept using our Ni/
EDO system, efforts directed at the design of a chiral EDO
ligand were met with limited success.17 As an alternative, we
envisioned that a Ni-catalyzed asymmetric reductive cross-
coupling reaction could deliver stereoconvergent C−C bond
formation with racemic aziridines. Reductive cross-coupling has
come to define an important class of reactions wherein two
organic electrophiles are combined using both a metal catalyst
and stoichiometric reductant. Although a reductive cross-
Received: April 5, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b03448
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
coupling of aziridines has yet to be developed, a number of
features of a reductive cross-coupling approach suggested its
application to the goal at hand. First, efficient coupling has been
demonstrated with secondary alkyl electrophiles.18 Amine- and
imine-based ligands, of which numerous chiral variants are
available, are generally most effective for these transformations.
Finally, it has been possible to render some reductive coupling
reactions of alkyl electrophiles stereoconvergent, as they
presumably occur via open shell intermediates.19 In seminal
studies, the Reisman lab demonstrated that chiral bisoxazoline
and phosphinooxazoline ligands can be applied in asymmetric
reductive coupling of secondary benzylic halides with acid
chlorides, and secondary allylic halides with aryl halides.20
Using a chiral titanocene catalyst in combination with an achiral
nickel catalyst, the Weix lab reported an enantioselective cross-
coupling of meso- epoxides with aryl bromides.21 Herein we
describe the development of a racemic and enantioselective
reductive cross-coupling of aryl iodides and styrenyl aziridines.
Investigation into ligand structure enantioselectivity trends
shed light on nonintuitive factors of chiral BiOx substitution
that influence asymmetric catalysis.
Supporting Information (SI)). Tetrakis(dimethylamino)-ethyl-
ene (TDAE) also works as a reductant,23 albeit in poorer
reaction efficiency (entry 5). In the absence of ligand, only trace
product was detected (entry 10), and control reactions without
Ni or reductant did not yield product (entries 11−12). In all
cases, C−C bond formation takes place at the benzylic carbon
of 1a.
Under the optimal reaction conditions, the aziridine ring-
opening reaction proved amenable to a wide array of aryl and
heteroaryl iodides (Table 2). Electron-rich (2), electron-neutral
a
Table 2. Reaction Scope
We first set out to probe the feasibility of a reductive cross
electrophile coupling with aziridines in racemic fashion. Initial
reaction optimization was conducted with 2-phenyl-N-
tosylaziridine 1a and 4-iodoanisole (2 equiv) using 10 mol %
NiCl2·glyme and Mn(0) as the stoichiometric reductant (Table
1). A broad range of ligands were evaluated to achieve
Table 1. Reaction Optimization
a
Yield of isolated product (2 equiv of iodide, 14−24 h).
a
a
entry deviation from standard conditions yield ( )-2 (%) yield 3 (%)
1
none
80
35
55
86
17
35
40
<5
42
<5
0
57
7
(4, 5), and electron-poor (6−8) aryl iodides undergo reaction
in high yield. Cross-coupling occurs with aryl iodides bearing
chloro- and pinacol boronate groups, generating products 9 and
12 with functional group handles for subsequent metal-
catalyzed cross-coupling. Furthermore, nitrogen-containing
heteroaromatics (14, 15) as well as oxygen-containing
heterocycles (16, 17) are well tolerated, making this an
attractive method for preparing potential bioactive agents.
While meta-substitution is well tolerated (10−12), ortho-
substitution does curtail reactivity (see SI).
Broad functional group compatibility is also displayed for the
styrenyl aziridine coupling partner.24 Electron-rich and -poor
para- and meta-substituted aziridines function well (18−25).
Cross-coupling takes place with an aziridine bearing ortho-
substitution, highlighting a complementarity to the aryl iodide
scope (25). Overall, the transformation has advantages over our
previously described Negishi coupling, since it avoids the need
for pregeneration of relatively harsh organozinc reagents.
Additionally, the arylations reported here are especially
productive in comparison to the previous system where 3
equiv of the PhZnBr nucleophile were needed to achieve only a
moderate yield.6
2
5 mol % Ni, 6 mol % L
1 equiv of ArI
Zn instead of Mn
TDAE instead of Mn
bpy (12 mol %)
QuinOx (12 mol %)
PyBox (12 mol %)
BiOx (12 mol %)
no ligand
3
20
52
<5
62
59
14
76
13
0
4
5
6
7
8
9
10
11
12
no Ni
no reductant
0
0
a
Determined by GC using dodecane as an internal standard.
selectivity for the desired product 2 over biaryl 3 arising from
homocoupling of the aryl iodide. In previous work from the
Doyle laboratory on Ni-catalyzed coupling of acetals and aryl
iodides, we identified the tridentate 2,6-bis(N-pyrazolyl)-
pyridine (bpp) ligand as effective in increasing selectivity for
product over biaryl formation.22 A similar benefit was observed
here; bipyridine, QuinOx, PyBOX, and BiOx ligands gave poor
to moderate yields of 2 (entries 6−9) whereas the bpp ligand
delivered 2 in 80% yield (entry 1). Reducing the catalyst
loading or lowering the equivalents of aryl iodide led to the
diminished yield of 2 (entries 2−3). However, with other aryl
iodides such as p-Ac-Ph-I, use of 1.1 equiv of the reaction
partner gave the corresponding product in 81% yield (see
We next proceeded to examine the feasibility of pursuing a
stereoconvergent coupling. Subjecting enantiopure aziridine
(+)-1a to the coupling conditions produced 7 as a racemate,
while recovered (+)-1a remained enantiopure (Scheme S3).
Among a collection of chiral amine- and phosphine-based
B
DOI: 10.1021/jacs.7b03448
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
ligands examined for the enantioselective coupling with rac-1a,
the BiOx ligand framework proved most promising. As
expected from the optimization studies of the racemic reaction,
yields initially suffered without the bpp ligand. However, we
found that inclusion of NaI and catalytic TMSCl as additives
improved the cross selectivity (see SI).25 Notably, ee’s ranging
from 10−78% were observed with seemingly minor perturba-
tions to the BiOx substituent. Use of the unique 4-heptyl BiOx
ligand L7 afforded 7 in 78% ee under these conditions, and in
90% ee under optimized conditions (Figure 2).
dispersion forces or CH−π interactions.33 Additionally, ligands
containing longer alkyl chains present higher NBOO (see SI for
more details), which may affect catalysis by modifying the
electron density at the metal center. We anticipate that the
correlations discovered here, particularly regarding NCIs, will
be informative for the development of additional trans-
formations as the field progresses.
The reaction with L7 as ligand is highly enantioselective,
producing various phenethylamine products in up to 94% ee at
−10 °C in THF (Table 3). Both electron-rich and -deficient
a b
,
Table 3. Scope of Stereoconvergent Reaction
Figure 2. Impact of BiOx ligand substituent on asymmetric cross-
coupling.
Since the enantioselectivity presented a nonobvious depend-
ence on the ligands’ structural features and the BiOx ligand
class has seen extensive recent application in the development
of both asymmetric reductive cross-couplings20a,d as well as Ni/
photoredox catalysis,26 we sought to apply the multivariate
analysis tools27 developed by the Sigman lab to understand the
origins of enantioselectivity. Molecular vibrations,28 Sterimol
descriptors,29 and molecular charges30 were computed for a
total of 17 BiOx ligands31 as parameters to describe electronic,
steric, and electrostatic effects. Additionally, polarizability was
calculated to probe if noncovalent interactions (NCIs) were
potentially involved.32 Linear regression modeling was applied
to relate the enantioselectivity (expressed as ΔΔG‡) and the
collected parameters. Only three parameters, condensed in two
terms, were required to obtain the correlation depicted in
Figure 2 (R2 = 0.85): B1 (minimum width of the ligand
substituent R) accounting for steric effects, NBOO (NBO
charge of the oxazoline oxygen atoms) mainly representing the
electronic character of the catalyst, and Pol (polarizability of the
ligand), which suggests the presence of NCIs. The single-
parameter correlations reported in Figure S2 support the
model, highlighting the importance of each one of these
parameters with qualitative trends (see SI). Through
interpretation of the model’s equation, a long, branched, alkyl
chain is suggested as the preferred substituent on the ligand.
Such a substituent would display enhanced steric effects while
providing a nonpolar surface for possible NCIs, such as
a
b
Yield of isolated product (3 equiv of iodide). Determined by chiral
c
d
HPLC. 10 mol % Ni, 12 mol % L7. No TMSCl added.
aryl iodides are tolerated, with meta-substituted aryl iodides
delivering the highest levels of enantioinduction. Furthermore,
substituted styrenyl aziridines undergo C−C bond formation in
moderate yield and ee (19, 20, 29−35). As evidence of the
stereoconvergent nature of the reaction, racemic aziridine 1a
was subjected to two reactions of identical conditions, except
for the enantiomer of ligand involved. As expected, the
coupling with ligands (S)-L7 and (R)-L7 generated enantioen-
riched product 7 of equal and opposite sign, both in high yield
(Scheme S5). In light of the stereospecific aziridine cross-
coupling reactions already reported,9−12 we expect that this
enantioselective variant will offer a valuable complement where
C
DOI: 10.1021/jacs.7b03448
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Jr.; Bayburt, E. K.; Asin, K. E.; Britton, D. R.; Lin, C. W.; Shiosaki, K. J.
Med. Chem. 1997, 40, 1585. (d) Paterson, I. A.; Juorio, A. V.; Boulton,
A. A. J. Neurochem. 1990, 55, 1827.
either enantiomer of product can be accessed from racemic
starting material.
(16) (a) Yamashita, M.; Yamada, K.-I.; Tomioka, K. J. Am. Chem. Soc.
2004, 126, 1954. (b) Wang, Z.-Q.; Feng, C.-G.; Zhang, S.-S.; Xu, M.-
H.; Lin, G.-Q. Angew. Chem., Int. Ed. 2010, 49, 5780. (c) Jumde, V. R.;
Iuliano, A. Adv. Synth. Catal. 2013, 355, 3475.
(17) Huang, C.-Y. D.; Doyle, A. G. J. Am. Chem. Soc. 2015, 137, 5638.
(18) Everson, D. A.; Jones, B. A.; Weix, D. J. J. Am. Chem. Soc. 2012,
134, 6146.
(19) Biswas, S.; Weix, D. J. J. Am. Chem. Soc. 2013, 135, 16192.
(20) (a) Cherney, A. H.; Kadunce, N. T.; Reisman, S. E. J. Am. Chem.
Soc. 2013, 135, 7442. (b) Cherney, A. H.; Reisman, S. E. J. Am. Chem.
Soc. 2014, 136, 14365. (c) Kadunce, N. T.; Reisman, S. E. J. Am. Chem.
Soc. 2015, 137, 10480. (d) Poremba, K. E.; Kadunce, N. T.; Suzuki, N.;
Cherney, A. H.; Reisman, S. E. J. Am. Chem. Soc. 2017, DOI: 10.1021/
jacs.7b01705.
(21) Zhao, Y.; Weix, D. J. J. Am. Chem. Soc. 2015, 137, 3237.
(22) Arendt, K. M.; Doyle, A. G. Angew. Chem., Int. Ed. 2015, 54,
9876.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b03448.
■
S
Experimental procedures, additional reaction optimiza-
tion, and spectroscopic data for all new compounds
(PDF)
AUTHOR INFORMATION
■
Corresponding Authors
*agdoyle@princeton.edu
*sigman@chem.utah.edu
ORCID
Matthew S. Sigman: 0000-0002-5746-8830
Notes
The authors declare no competing financial interest.
(23) Broggi, J.; Terme, T.; Vanelle, P. Angew. Chem., Int. Ed. 2014,
53, 384.
(24) Although the reductive coupling is currently limited to tosyl-
protected styrenyl aziridines, the tosyl group can be deprotected in
high yield and with good conservation of ee. 1,2-Disubstituted styrenyl
aziridines and (hetero)aryl bromides have thus far proven unreactive
under the optimal conditions.
(25) Tables S1 and S2 indicate that these additives enhance yield.
Control experiments are not consistent with the intermediacy of a
benzyl iodide, although additional studies are ongoing: (a) Zhao, Y.;
Weix, D. J. J. Am. Chem. Soc. 2014, 136, 48. (b) Fagnou, K.; Lautens,
M. Angew. Chem., Int. Ed. 2002, 41, 26.
(26) (a) Tellis, J. C.; Primer, D. N.; Molander, G. A. Science 2014,
345, 433. (b) Ahneman, D. T.; Doyle, A. G. Chem. Sci. 2016, 7, 7002.
(27) Sigman, M. S.; Harper, K. C.; Bess, E. N.; Milo, A. Acc. Chem.
Res. 2016, 49, 1292.
(28) Milo, A.; Bess, E. N.; Sigman, M. S. Nature 2014, 507, 210.
(29) Harper, K. C.; Bess, E. N.; Sigman, M. S. Nat. Chem. 2012, 4,
366.
ACKNOWLEDGMENTS
■
We thank the Reisman laboratory for introducing us to L7 and
for sharing material and procedures for its preparation. We also
thank Amanuella Mengiste for assistance with substrate
preparation. Financial support was provided by an F32 Ruth
L. Kirschtein NRSA Fellowship under Award No.
F32GM115024 (B.P.W.) and Amgen. M.S.S. thanks the NSF
for partial support of this work (CHE-1361296). M.O. thanks
the Ermenegildo Zegna Group for a postdoctoral fellowship.
REFERENCES
■
(1) Aziridines and Epoxides in Organic Synthesis; Yudin, A. K., Ed.;
Wiley-VCH: Weinheim, 2006.
(2) (a) Hu, X. E. Tetrahedron 2004, 60, 2701. (b) Stankovic, S.;
́
D’hooghe, M.; Catak, S.; Eum, H.; Waroquier, M.; Van Speybroeck,
V.; De Kimpe, N.; Ha, H. Chem. Soc. Rev. 2012, 41, 643.
(3) (a) Zhu, Y.; Wang, Q.; Cornwall, R. G.; Shi, Y. Chem. Rev. 2014,
114, 8199. (b) Xia, Q. H.; Ge, H. Q.; Ye, C. P.; Liu, Z. M.; Su, K. X.
Chem. Rev. 2005, 105, 1603.
(30) Mougel, V.; Santiago, C. B.; Zhizhko, P. A.; Bess, E. N.; Varga,
J.; Frater, G.; Sigman, M. S.; Coper
6699.
́
et, C. J. Am. Chem. Soc. 2015, 137,
(31) We also examined some 4,5-disubstituted and heteroatom
substituted bioxazoline ligands which were not included in the model
(see SI) but have been effective in other Ni couplings: Do, H.-Q.;
Chandrashekar, E. R. R.; Fu, G. C. J. Am. Chem. Soc. 2013, 135, 16288.
(32) (a) Chen, Z.-M.; Hilton, M. J.; Sigman, M. S. J. Am. Chem. Soc.
2016, 138, 11461. (b) Yamamoto, E.; Hilton, M. J.; Orlandi, M.; Saini,
V.; Toste, F. D.; Sigman, M. S. J. Am. Chem. Soc. 2016, 138, 15877.
(33) Biedermann, F.; Schneider, H.-J. Chem. Rev. 2016, 116, 5216.
(4) (a) Pellissier, H. Tetrahedron 2010, 66, 1509. (b) Pellissier, H.
Adv. Synth. Catal. 2014, 356, 1899.
(5) (a) Bhat, V.; Welin, E. R.; Guo, X.; Stoltz, B. M. Chem. Rev. 2017,
117, 4528. (b) Lu, Z.; Wilsily, A.; Fu, G. C. J. Am. Chem. Soc. 2011,
133, 8154.
(6) Huang, C.-Y. D.; Doyle, A. G. J. Am. Chem. Soc. 2012, 134, 9541.
(7) Lin, B. L.; Clough, C. R.; Hillhouse, G. L. J. Am. Chem. Soc. 2002,
124, 2890.
(8) Ney, J. E.; Wolfe, J. P. J. Am. Chem. Soc. 2006, 128, 15415.
(9) Duda, M. L.; Michael, F. E. J. Am. Chem. Soc. 2013, 135, 18347.
(10) (a) Takeda, Y.; Ikeda, Y.; Kuroda, A.; Tanaka, S.; Minakata, S. J.
Am. Chem. Soc. 2014, 136, 8544. (b) Takeda, Y.; Kuroda, A.; Sameera,
W. M. C.; Morokuma, K.; Minakata, S. Chem. Sci. 2016, 7, 6141.
(11) Jensen, K. L.; Standley, E. A.; Jamison, T. F. J. Am. Chem. Soc.
2014, 136, 11145.
(12) Huang, C.-Y. D.; Doyle, A. G. Chem. Rev. 2014, 114, 8153.
(13) Additional examples of stereospecific addition of Ni to C−N
bonds: (a) Maity, P.; Shacklady-McAtee, D. M.; Yap, G. P. A.; Sirianni,
E. R.; Watson, M. P. J. Am. Chem. Soc. 2013, 135, 280. (b) Basch, C.
H.; Cobb, K. M.; Watson, M. P. Org. Lett. 2016, 18, 136.
(14) Nielsen, D. K.; Huang, C.-Y.; Doyle, A. G. J. Am. Chem. Soc.
2013, 135, 13605.
(15) (a) Zhang, A.; Neumeyer, J. L.; Baldessarini, R. J. Chem. Rev.
2007, 107, 274. (b) Zhang, J.; Xiong, B.; Zhen, X.; Zhang, A. Med. Res.
Rev. 2009, 29, 272. (c) Michaelides, M. R.; Hong, Y.; DiDomenico, S.,
D
DOI: 10.1021/jacs.7b03448
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX