Enantioselective palladium-catalyzed diamination of alkenes using N-fluorobenzenesulfonimide

Article abstract of DOI:10.1021/ja4043406

An enantioselective Pd-catalyzed vicinal diamination of unactivated alkenes using N-fluorobenzenesulfonimide as both an oxidant and a source of nitrogen is reported. The use of Ph-pybox and Ph-quinox ligands afforded differentially protected vicinal diamines in good yields with high enantioselectivities. Mechanistic experiments revealed that the high enantioselectivity arises from selective formation of only one of four possible diastereomeric aminopalladation products of the chiral Pd complex. The aminopalladation complex was characterized by X-ray crystallography.

Full text of DOI:10.1021/ja4043406

Communication
pubs.acs.org/JACS
Enantioselective Palladium-Catalyzed Diamination of Alkenes Using
N‑Fluorobenzenesulfonimide
Erica L. Ingalls, Paul A. Sibbald, Werner Kaminsky, and Forrest E. Michael*
Department of Chemistry, University of Washington, Box 351700, Seattle, Washington 98195-1700, United States
S
* Supporting Information
The development of an enantioselective diamination reaction
began with a screen of the chiral pyridinebis(oxazoline)
(pybox) ligands depicted in Figure 1 under standard
ABSTRACT: An enantioselective Pd-catalyzed vicinal
diamination of unactivated alkenes using N-fluorobenze-
nesulfonimide as both an oxidant and a source of nitrogen
is reported. The use of Ph-pybox and Ph-quinox ligands
afforded differentially protected vicinal diamines in good
yields with high enantioselectivities. Mechanistic experi-
ments revealed that the high enantioselectivity arises from
selective formation of only one of four possible
diastereomeric aminopalladation products of the chiral
Pd complex. The aminopalladation complex was charac-
terized by X-ray crystallography.
Figure 1. Chiral ligands.
hiral 1,2-diamines are important moieties found in
Cbiologically active compounds, organocatalysts, asymmet-
ric ligands, and auxiliaries.1,2 Direct difunctionalization of
unactivated alkenes constitutes a valuable and powerful method
for creating such useful chiral 1,2-diamine motifs. Despite the
many uses of asymmetric 1,2-diamine scaffolds, efficient
methods for their direct synthesis from alkenes are limited in
comparison with conceptually similar dihydroxylation and
aminohydroxylation transformations.3 Though several transi-
tion-metal-catalyzed diamination reactions have recently been
developed,4 enantioselective variants are still rare. Notable
examples include the intermolecular enantioselective diamina-
Table 1. Chiral Ligand Screen and Optimization
a
b
entry
alkene
ligand
cond.
% yield
% ee
1
2
3
4
5
6
7
8
9
1a
1a
1a
1a
1a
1a
1a
3a
3a
(R,R)-L1
(S,S)-L2
(S,S)-L3
(S,S)-L4
(S,S)-L5
(R)-L6
A
A
A
A
A
A
75
56
62
59
92
−32
−10
−20
−62
36
tions reported by Muniz5,6 and Shi.7 Although these methods
̃
c
provide high yields and enantioselectivities, there are still
limitations on these transformations, particularly that they
require stoichiometric amounts of osmium5 or chiral iodine
reagent6 or are limited to diene substrates.7
60
42
80
66
50
d
(R)-L7
B
80
(R)-L7
B
A
93
We recently disclosed a novel palladium-catalyzed diamina-
tion of alkenes using N-fluorobenzenesulfonimide (NFBS) as
both an oxidant and a source of nitrogen.8 This method is
useful for the generation of a variety of differentially protected
cyclic 1,2-diamines (Scheme 1). Herein we report a highly
enantioselective version of this reaction using chiral oxazolines
as ligands along with the results of mechanistic experiments
that shed light on the origin of the enantioselectivity in this
system.
(R,R)-L1
80
a
Conditions A: 15 mol % ligand, 10 mol % Pd(TFA)2, 20 mol %
TEMPO, EtOAc, reflux. Conditions B: 12 mol % ligand, 10 mol %
b
Pd(TFA)2, 20 mol % TEMPO, 1,4-dioxane, rt. Determined by chiral
c
d
HPLC. 1H NMR yield versus an internal standard. NFBS was added
over a 3 h period.
diamination conditions using substrate 1a (Table 1). It was
immediately clear that among the pybox ligands L1−L5, only
the phenyl-substituted pybox ligand L1 gave sufficiently high
enantioselectivity (entry 1). Although ligand L1 gave a good
yield of product 2a, it (like all of the pybox ligands) resulted in
a substantial decrease in catalyst reactivity. Though the
Scheme 1. Pd-Catalyzed Diamination Using NFBS
Received: May 1, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja4043406 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
conversion in the absence of ligand was rapid at room
temperature, in the presence of a tridentate ligand, complete
consumption of the starting material was seen only upon
refluxing in EtOAc overnight. This drop in reactivity was also
apparent in the attempted reaction of substrate 3a using ligand
L1, which gave the desired product in a yield of only 50%
(entry 9). We reasoned that perhaps the tridentate ligands
donate too much electron density to the metal center and
therefore decrease the rate of aminopalladation. If this
hypothesis was true, the use of a bidentate ligand should
alleviate this problem. Gratifyingly, the use of an analogous
bidentate ligand, the phenyl-substituted quinolineoxazoline
(quinox) ligand L7, resulted in a much more active catalyst
with only slighly diminished enantioselectivity. After a short
optimization (see the Supporting Information), we found that
the use of L7 allowed the reaction to be run at room
temperature while giving higher yields and very high ee (entry
8).
Scheme 2. Substrate Scope
a
Conditions: 12 mol % ligand, 10 mol % Pd(TFA)2, 20 mol %
Amide and carbamate protecting groups were tested under
the optimized reaction conditions (Table 2). For amides
b
TEMPO, 2 equiv of NFBS, 1,4-dioxane, rt. Determined by chiral
HPLC.
Table 2. Scope of the Amine Protecting Group
Scheme 3. Formation of Alkylpalladium Intermediate 9
a
entry
alkene
R
% yield
% ee
1
2
3
4
5
6
7
3b
3c
3d
3e
3f
Ph
75
71
70
71
60
34
50
91
96
94
91
93
82
91
p-MeOC6H4
p-BrC6H4
CH3
OBn
3g
3h
Ot-Bu
OCH2CCl3
a
Determined by chiral HPLC.
(entries 1−4), both electron-withdrawing and -donating groups
gave high enantioselectivities and yields similar to that for
substrate 3a. Carbamates (entries 5−7) also afforded high
enantioselectivities, albeit with somewhat lower yields com-
pared with the amides. Substrates with different substitution
patterns were also subjected to the optimized conditions
(Scheme 2). Products with geminal disubstitution on the
backbone (2f, 6a) could generally be made in good yields with
high enantioselectivity. Monosubstituted substrate 7 also gave
excellent enantioselectivity, but the yield and diastereoselectiv-
ity were modest.
In our previous mechanistic work on the diamination
reaction,9 an intermediate alkylpalladium complex was isolated
by trapping with a bipyridine ligand. To learn more about the
origin of the enantioselectivity in this reaction, a Pd(Ph-quinox)
complex was generated by mixing Pd(TFA)2 (TFA =
trifluoroacetate) and ligand L7 and treating the mixture with
substrate 3e (Scheme 3). Full conversion to alkylpalladium
complex 9 was observed.10 Remarkably, only one of the four
possible stereoisomeric products could be detected by 1H
NMR spectroscopy. Complex 9 was isolated in 56% yield,
crystallized, and analyzed by X-ray crystallography (Figure 2).
The structure of 9 shares several important features with our
previously reported bipy−Pd−alkyl complex,9 including the
strong chelation of the amide carbonyl to the Pd center and the
presence of the H(OCOCF3)2 complex counterion. The very
Figure 2. Crystal structure of (R)-Ph-quinox−Pd−alkyl complex 9.
The complex counterion has been omitted for clarity. Bond distances
(Å): Pd−C, 1.998; Pd−O, 2.039; Pd−N1, 2.233; Pd−N2, 2.021.
large difference in the Pd−N bond lengths (2.02 vs 2.23 Å) in
this complex is noteworthy. Two factors could be responsible
for this difference. First, the large difference in trans influence
by O and C should result in some lengthening of the
quinoline−Pd bond. Second, the quinoline−Pd bond should
also be longer because of steric interference between the peri
hydrogen of the quinoline and the ligand cis to the quinoline.
Two existing X-ray crystal structures confirm that both factors
are operative and that the latter factor is more important. In the
bipy−Pd−alkyl complex, which should be affected only by the
trans influence, the Pd−N bond trans to C is only 0.07 Å longer
than the Pd−N bond trans to O (2.03 vs 2.10 Å). In the (t-Bu-
quinox)PdCl2 complex reported by Yang,11 which should
display only the effects of quinoline sterics, the quinoline−Pd
bond is 0.15 Å longer than the oxazoline−Pd bond (2.16 vs
B
dx.doi.org/10.1021/ja4043406 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Soc., Perkin Trans. 1 2002, 2733−2746. (c) O’Brien, P. Angew. Chem.,
Int. Ed. 1999, 38, 326−329.
2.01 Å). The 0.21 Å difference in bond lengths in complex 9 is
very nearly the sum of those two factors.12 The failure of pyrox
ligand L6 to give high enantioselectivity (Table 1, entry 6),
indicates that the steric effect of the quinoline plays a crucial
role in determining the stereoselectivity.
To establish further the intermediacy of complex 9 in the
catalytic diamination, 9 was treated with NFBS and
triethylammonium benzenesulfonimide (Scheme 4).13 Under
(4) Examples of transition-metal-catalyzed diaminations: (a) Li, G.
G.; Wei, H. X.; Kim, S. H.; Carducci, M. D. Angew. Chem., Int. Ed.
2001, 40, 4277−4280. (b) Bar, G. L.; Lloyd-Jones, G. C.; Booker-
Milburn, K. I. J. Am. Chem. Soc. 2005, 127, 7308−7309. (c) Streuff, J.;
Hovelmann, C. H.; Muniz, K. J. Am. Chem. Soc. 2005, 127, 14586−
̈
̃
14587. (d) Muniz, K.; Hovelmann, C. H.; Streuff, J. J. Am. Chem. Soc.
̃
̈
2008, 130, 763−773. (e) Muniz, K. J. Am. Chem. Soc. 2007, 129,
14542−14543. (f) Hovelmann, C. H.; Streuff, J.; Brelot, L.; Muniz, K.
̃
̈
̃
Chem. Commun. 2008, 2334−2336. (g) Du, H.; Zhao, B.; Shi, Y. J. Am.
Chem. Soc. 2007, 129, 762−763. (h) Sequeira, F. C.; Turnpenny, B.
W.; Chemler, S. R. Angew. Chem., Int. Ed. 2010, 49, 6365−6368.
(i) Zhang, H.; Pu, W.; Siong, T.; Li, Y.; Zhou, X.; Sun, K.; Liu, Q.;
Zhang, Q. Angew. Chem., Int. Ed. 2013, 52, 2529−2533.
Scheme 4. Amination of the Alkylpalladium Complex
(5) Asymmetric stoichiometric Os-promoted diaminations:
(a) Muniz, K.; Neiger, M. Synlett 2003, 211−214. (b) Muniz, K.;
̃
̃
Nieger, M.; Mansikkiamaki, H. Angew. Chem., Int. Ed. 2003, 42, 5958−
5961. (c) Muniz, K.; Atsushi, I.; Nieger, M. Chemistry 2003, 9, 5581−
̃
5596. (d) Muniz, K. Tetrahedron Lett. 2003, 44, 3547−3549.
̃
(e) Muniz, K. New J. Chem. 2005, 29, 1371−1385. (f) Muniz, K.;
̃
̃
Nieger, M. Chem. Commun. 2005, 2729−2731.
these conditions, the diamination product was isolated in 55%
yield with 98% ee, which is very nearly the same as the catalytic
reaction affords. Furthermore, the major enantiomer produced
in this reaction matched that observed in the catalytic reaction,
and its absolute configuration was determined to be S,
matching what was observed in complex 9 (see the Supporting
Information). This is consistent with aminopalladation serving
as the enantiodetermining step of the catalytic cycle.
In conclusion, an enantioselective method for the diamina-
tion of alkenes to create a differentally protected diamination
product has been developed. The palladium-catalyzed reaction
provided products in moderate yields with up to 99% ee using
the (R)-Ph-quinox ligand. Isolation of a single stereoisomer of
the intermediate alkylpalladium complex established that
aminopalladation is the enantiodetermining step of this
transformation.
(6) Enantioselective metal-free diamination: Roben, C.; Souto, J. A.;
̈
Gonzalez, Y.; Lishchynskyi, A.; Muniz, K. A. Angew. Chem., Int. Ed.
2011, 50, 9478−9482.
́
̃
(7) Enantioselective diaminations of dienes: (a) Du, H.; Yuan, W.;
Zhao, B.; Shi, Y. J. Am. Chem. Soc. 2007, 129, 11688−11689. (b) Xu,
L.; Shi, Y. J. Org. Chem. 2008, 73, 749−751. (c) Du, H.; Zhao, B.;
Yuan, W.; Shi, Y. Org. Lett. 2008, 10, 4231−4234. (d) Zhao, B.; Du,
H.; Shi, Y. J. Org. Chem. 2009, 74, 8392−8395. (e) Cornwall, R. G.;
Zhao, B.; Shi, Y. Org. Lett. 2013, 15, 796−799.
(8) Sibbald, P. A.; Michael, F. E. Org. Lett. 2009, 11, 1147−1149.
(9) Sibbald, P. A.; Rosewall, C. F.; Swartz, R. D.; Michael, F. E. J. Am.
Chem. Soc. 2009, 131, 15945−15951.
(10) Previously, the stereochemistry of this step was determined to
be anti. The formation of 9 also proceeds through anti-amino-
palladation (see the Supporting Information).
(11) He, W.; Yip, K. T.; Zhu, N. Y.; Yang, D. Org. Lett. 2009, 11,
5626−5628.
(12) See Table S2 in the Supporting Information for details on bond
lengths.
(13) As reported in ref 8, excess benzenesulfonimide was required to
prevent competitive incorporation of the trifluoroacetate counterion.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental conditions, spectroscopic data, crystallographic
information (CIF), and data for determining the enantiomeric
excesses of the diamination products. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
michael@chem.washington.edu
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank the University of Washington and the National
Science Foundation for financial support of this project.
■
REFERENCES
■
(1) Vicinal diamine reviews: (a) Leucet, D.; Gall, T. L.; Mioskowski,
C. Angew. Chem., Int. Ed. 1998, 37, 2580−2627. (b) Saibabu Kotti, S.
R. S.; Timmons, C.; Li, G. Chem. Biol. Drug Des. 2006, 67, 101−114.
(2) Transition-metal-catalyzed diamination reviews: (a) Marcia de
Fueiredo, R. Angew. Chem., Int. Ed. 2009, 48, 1190−1193. (b) Carcona,
F.; Goti, A. Nat. Chem. 2009, 1, 269−275. (c) Martínez, C.; Muniz, K.
̃
J. Org. Chem. 2013, 78, 2168−2174.
(3) (a) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem.
Rev. 1994, 94, 2483−2547. (b) Bodkin, J. A.; MacLeod, M. D. J. Chem.
C
dx.doi.org/10.1021/ja4043406 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX