Modular synthesis of 1,2-Diamine derivatives by palladium-Catalyzed aerobic oxidative cyclization of allylic sulfamides

Article abstract of DOI:10.1002/anie.200906342

Chemical equation presented Allylic sulfamides undergo aerobic oxidative cyclization at room temperature, mediated by a Pd(O ₂CCF₃)₂/DMSO catalyst system in tetrahydrofuran. The cyclic sulfamide products are readily converted into 1,2-diamines, and substrates derived from chiral allylic amines cyclize with very high diastereoselectivity.

Full text of DOI:10.1002/anie.200906342

Angewandte
Chemie
DOI: 10.1002/anie.200906342
Oxidative Amination
Modular Synthesis of 1,2-Diamine Derivatives by Palladium-Catalyzed
Aerobic Oxidative Cyclization of Allylic Sulfamides**
Richard I. McDonald and Shannon S. Stahl*
Vicinal diamines are prevalent in biologically active mole-
cules and as ligands for transition metals. Consequently, these
structures have been the target of considerable synthetic
effort.[1] A number of novel palladium-catalyzed methods for
the synthesis of diamines from alkenes and dienes have been
reported recently.[2–6] Limitations of these reactions with
respect to substrate scope and identity of the stoichiometric
oxidant (e.g., benzoquinone, PhI(OAc)2, and di-tert-butyldia-
ziridinone), together with our interest in aerobic oxidation
reactions,[7] prompted us to consider whether analogous
diamines could be prepared using O2 as the oxidant. Herein,
we describe a versatile method for the stereocontrolled
synthesis of 1,2-diamine derivatives by the palladium-cata-
lyzed aerobic oxidative cyclization of allylic sulfamides (A,
Scheme 1), substrates which are readily prepared in multi-
gram quantities from their corresponding allylic and primary
amines.[8] These reactions were made possible by identifica-
tion of a very simple catalyst system consisting of a 2:1 ratio of
dimethyl sulfoxide and Pd(TFA)2 (TFA = trifluoroacetate),
which is capable of promoting the oxidative cyclization
reactions at room temperature with molecular oxygen as the
sole stoichiometric oxidant.
Initial reaction development efforts employed sulfamide 1
as the substrate (Table 1). We tested a number of catalyst
systems that have been reported previously for aerobic
oxidative heterocyclization reactions, the most prominent of
which include Pd(OAc)2 in dimethyl sulfoxide (Ac =
acetyl),[9] Pd(OAc)2/pyridine in toluene, and other closely
related variants.[10] These catalyst systems were only moder-
ately successful (Table 1, entries 1–3; for full screening data,
see the Supporting Information). For example, the Pd(OAc)2/
pyridine catalyst system we originally reported for tosyl-
substituted g-aminoalkenes[10a] afforded the desired cyclic
Table 1: Optimization of the palladium-catalyzed aerobic oxidative
cyclization of sulfamides.[a]
Entry
1
Catalyst
(5 mol%)
Additive
(mol%)
Base
(mol%)
Solvent
T
Yield 2 (3)
[8C] [%][b]
Pd(OAc)2
–
NaOBz DMSO 80 24
(200)
2
3
Pd(OAc)2
Pd(TFA)2
py (10)
py (20)
–
toluene 80 39 (25)
NaOAc toluene 80 68 (17)
(100)
4
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
Pd(TFA)2
py (20)
NaOBz
(20)
NaOBz
(20)
1,4-
dioxane
1,4-
dioxane
1,4-
80 35
80 48
80 92
25 99[c,d]
5
py (20)/
DMSO (10)
DMSO (10) NaOBz
(20)
DMSO (10) NaOBz
(20)
6
dioxane
THF
7
8
–
NaOBz
(20)
NaOBz
(20)
THF
THF
THF
25
9[c]
9
DMSO (5)
25 25[c]
25 93[c]
Scheme 1. Synthesis of 1,2-diamines by palladium-catalyzed aerobic
oxidative cyclization of allylic sulfamides.
10
11
12
13
DMSO (20) NaOBz
(20)
–
NaOBz DMSO 25 47
(20)
[*] R. I. McDonald, Prof. S. S. Stahl
ligand 4 (5) NaOBz
THF
25
9
Department of Chemistry, University of Wisconsin-Madison
1101 University Avenue, Madison, WI 53706 (USA)
Fax: (+1)608-262-6143
(20)
ligand 5 (5) NaOBz
(20)
THF
25 35
E-mail: stahl@chem.wisc.edu
Homepage: http://www.chem.wisc.edu/~stahl
[**] We thank Dr. Christopher C. Scarborough for helpful discussions,
Dr. I. A. Guzei and L. C. Spencer for X-ray crystallographic
assistance, and Dr. C. G. Fry for NMR spectroscopic assistance. We
are grateful for financial support from the NIH (R01GM67163) and
Abbott Laboratories (graduate fellowship for R.I.M.).
[a] Conditions: 1 (0.075 mmol), 3 ꢀ M.S. (20 mg), 1 atm O2, solvent
(0.75 mL), 24 h. [b] Determined by 1H NMR spectroscopy, internal
standard=1,3,5-trimethoxybenzene. [c] 10 h. [d] Yield of isolated prod-
uct (0.3 mmol scale). Bz=benzoyl, DMSO=dimethyl sulfoxide, THF=
tetrahydrofuran.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200906342.
Angew. Chem. Int. Ed. 2010, 49, 5529 –5532
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5529

Communications
sulfamide 2 in only 39% yield and generated the imine
studies, the large excess of dimethyl sulfoxide prevented us
from gaining fundamental insights into the palladium–
dimethyl sulfoxide interaction. The present catalyst system
is more amenable to characterization. Spectroscopic studies
draw attention to at least two properties of dimethyl sulfoxide
that are probably important in these reactions: 1) linkage
byproduct 3 in 25% yield (Table 1, entry 2). In the course of
testing other solvents and reaction conditions, we observed
that addition of catalytic quantities of dimethyl sulfoxide
improved the yield of 2 when Pd(TFA)2/pyridine was used as
the catalyst in 1,4-dioxane (Table 1, entries 4 and 5). Sub-
stantially better results were obtained by eliminating pyridine
altogether (Table 1, entry 6). The formation of palladium
black in these reactions prompted us to examine whether the
catalyst might be more stable, but retain good activity at lower
temperature. Additional optimization studies led to condi-
tions in which quantitative product formation could be
obtained in 10 hours at room temperature in tetrahydrofuran
(Table 1, entry 7). The stoichiometry of dimethyl sulfoxide
affected the reaction outcome (Table 1, entries 7–11). The
product yield was substantially lower at PdII/DMSO ratios of
less than 2:1 (Table 1, entries 8 and 9), and the yield also
diminished at higher [DMSO], falling to 47% when dimethyl
sulfoxide was used as the solvent (Table 1, entries 10 and 11).
The identity and quantity of the added base is also important,
and use of 20 mol% sodium benzoate was found to be
optimal.[11]
1
isomerism and 2) kinetic lability. H NMR spectra of the 2:1
DMSO/Pd(TFA)2 mixture in [D8]tetrahydrofuran revealed
several resonances for dimethyl sulfoxide, none of which
corresponded to free dimethyl sulfoxide. The chemical shifts
of these resonances, together with infrared spectroscopic
analysis of the PdII/DMSO complexes obtained under these
conditions, supported the presence of both S- and O-bound
dimethyl sulfoxide ligands.[11,14] These observations are con-
sistent with earlier studies of the coordination of dimethyl
sulfoxide to palladium(II).[15,16] As we have previously
speculated,[13] the ability of dimethyl sulfoxide to serve as a
“hard” (O) or “soft” (S) ligand could play an important role in
the interconversion between the relatively hard and soft PdII
and Pd0 redox states during the catalytic cycle. Variable-
1
temperature H NMR spectroscopic analysis (À60 to 408C)
revealed the coalescence of dimethyl sulfoxide ligand reso-
nances, the O-bound dimethyl sulfoxide resonances at À408C,
and the S-bound resonances at + 408C.[11] These observations
highlight the kinetically labile nature of dimethyl sulfoxide
coordination to palladium(II) under the reaction conditions.
This property contrasts with the behavior of pyridine as a
ligand[17] and probably facilitates substrate coordination to
palladium(II) and other ligand-exchange processes necessary
for efficient catalytic turnover at room temperature.
This oxidative cyclization reaction could proceed via two
different mechanisms (Scheme 2): aminopalladation of the
alkene followed by b-hydride elimination[10a] or allylic C H
À
activation to form a p-allyl–palladium(II) intermediate fol-
This simple catalyst system was also quite versatile in
reactions with other sulfamides (Table 2 and 3). Substrates
bearing both aliphatic or aryl N-substituents undergo efficient
cyclization, with the only exception being a substrate derived
from an electron-deficient aniline (7e). The reactions are
remarkably tolerant of functional groups, being compatible
with groups that are typically stable to oxidizing reaction
conditions, such as ester (7 f), aryl fluoride (7h), carbamate
(7i), primary chloride (7k), and ether groups (7d; see also,
Table 3, entry 7), whilst also tolerating groups that were
susceptible to oxidation in other palladium-catalyzed reac-
tions, including terminal alkene (7j) and furan substituents
(7l). A substrate with a silyl ether appended to the allyl amine
also underwent cyclization in high yield, affording a silyl enol
ether product that was stable under the reaction conditions
(Table 3, entry 3). In many of these reactions, analytically
pure products were obtained by simply filtering the reaction
mixture through a plug of activated basic alumina. Further-
more, all procedures were performed on the bench, and no
solvent purification was required prior to performing these
reactions.
Scheme 2. Possible mechanisms for the palladium-catalyzed oxidative
cylization reaction.
lowed by C N coupling.[12b,d,e] The chelating sulfoxides 4 and 5
À
were tested as replacements for dimethyl sulfoxide because
À
these ligands are known to facilitate allylic C H activa-
tion;[12b,d,e] however, only low yields of sulfamide product 2
were obtained in these reactions (Table 1, entries 12 and 13).
To further distinguish between these two mechanisms, the
homoallyl amine derivative 6 was synthe-
sized. Effective cyclization of this substrate
would provide evidence in favor of an allylic
À
C H activation pathway. However, subject-
ing this substrate to the optimized reaction
conditions resulted in complete recovery of starting material
Sulfamides derived from substituted allylic amines were
also excellent substrates (Table 3). The use of a trisubstituted
À
after 24 hours. This result suggests that allylic C H activation
does not occur under these reaction conditions.
À
alkene (Table 3, entry 1) led to quarternary C N bond
The beneficial effect of dimethyl sulfoxide and other
sulfoxides in palladium-catalyzed reactions has been noted
previously by a number of groups,[9,12] and we have already
reported kinetic studies of aerobic alcohol oxidation cata-
lyzed by Pd(OAc)2 in dimethyl sulfoxide.[13] In the latter
formation in quantitative yield. Cyclization onto an alkene
À
with remote C H bonds experienced little complication
associated with alkene isomerization (Table 3, entry 2). The
cyclization reactions exhibited good-to-excellent levels of
diastereoselectivity. A sulfamide derived from a-methylben-
5530
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5529 –5532

Angewandte
Chemie
Table 2: Aerobic oxidative cyclization of sulfamides with various N-
Table 3: Aerobic oxidative cyclization of sulfamides.[a]
substituents.[a,b]
Entry
Substrate
Product
Yield [%][b]
99
1
2
3
4
5
6
96[c]
7a 99%
7b 99%
7c 97%[c]
81[d,e]
7d 96%c
7g 92%
7e 30%[c,d]
7 f 70%[c,d]
90[d]
(d.r.=6:1)
7h 99%
7i 99%
93
(d.r.=11:1)
7j 99%
7k 91%c
7l 80%
95
[a] Conditions: substrate (0.3 mmol), Pd(TFA)2 (15 mmol), DMSO
(30 mmol), NaOBz (60 mmol), 3 ꢀ MS (75 mg), THF (3 mL), 258C,
24 h. [b] Yield of isolated product. [c] NaOAc used instead of NaOBz.
[d] 48 h.
(d.r.>30:1)
98
7
8
9
(d.r.=29:1)
zylamine as the primary amine (Table 3, entry 4) underwent
cyclization to afford a 6:1 ratio of diastereomers. Substrates
derived from chiral allylic amines exhibited substantially
higher levels of stereoselectivity: allylic substituents larger
than a methyl group provided products with diastereomeric
ratios > 29:1 (entries 5–9). The latter observation has impor-
tant implications for the synthetic utility of this cyclization
reaction because numerous procedures are available for the
enantioselective synthesis of chiral allylic amines.[18]
90[d]
(d.r.>30:1)
73
(d.r.>30:1)
The use of this method to prepare densely functionalized
chiral diamines is highlighted in the oxidative cyclization of
mannitol-derived substrate 8 (Scheme 3). This reaction was
carried out at room temperature on a 1 gram scale, and it
proceeded in quantitative yield, affording the product as a
single diastereomer. Facile removal of the sulfonyl group by
treatment of 9 with LiAlH4 in diethyl ether gave the acylic
diamine 10 in 89% yield.[11]
[a] Conditions: substrate (0.3 mmol), Pd(TFA)2 (15 mmol), DMSO
(30 mmol), NaOBz (60 mmol), 3 ꢀ M.S. (75 mg), THF (3 mL), 258C,
24 h. [b] Yield of isolated product; d.r. determined by 1H NMR spectros-
copy of the crude reaction mixture. [c] Isolated as a 12:1 mixture of
internal/terminal olefins. [d] 48 h. [e] Isolated as a 2.1:1 mixture of E/Z
olefins.
The reactions described here are complementary to other
metal-catalyzed routes to prepare vicinal diamines.[2–6]
Among the most synthetically useful methods currently
available are those reported recently by Shi and co-workers,
which employ di-tert-butylaziridinone as the oxidant and
diamination reagent [e.g., Eq. (1)].[3] The latter methods are
highly diastereoselective, they employ readily available
Scheme 3. Demonstration of 1,2-diamine synthesis by aerobic oxida-
tive cyclization of a mannitol-derived allylic sulfamide.
Angew. Chem. Int. Ed. 2010, 49, 5529 –5532
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5531

Communications
8, 2531 – 2534; b) J. A. Fritz, J. P. Wolfe, Tetrahedron 2008, 64,
6838 – 6852.
terminal alkenes as substrates, and enantioselective reactions
have been achieved.[3b] One of the few drawbacks of this
method is the requirement for tert-butyl groups as the N-
substituents of the diaziridinone reagent. Removal or replace-
ment of these groups requires additional steps in the synthesis
of a target molecule.[19] In contrast, the aerobic oxidation
method described here exhibits broad versatility with respect
to the N-substituents. This feature, together with the exten-
sive availability of enantiomerically pure allylic amines,[18]
could be highly advantageous in the synthesis of many target
molecules.
[6] For important methods employing metals other than palladium,
see: a) T. P. Zabawa, S. R. Chemler, Org. Lett. 2007, 9, 2035 –
2038; b) K. Muꢀiz, J. Streuff, C. H. Hꢁvelmann, A. Nfflꢀez,
Angew. Chem. 2007, 119, 7255 – 7258; Angew. Chem. Int. Ed.
2007, 46, 7125 – 7127; c) H. Du, B. Zhao, W. Yuan, Y. Shi, Org.
Lett. 2008, 10, 4231 – 4234; d) D. E. Olson, J. Du Bois, J. Am.
Chem. Soc. 2008, 130, 11248 – 11249; e) B. M. Trost, S. Malhotra,
D. E. Olson, A. Maruniak, J. Du Bois, J. Am. Chem. Soc. 2009,
131, 4190 – 4191.
[7] a) S. S. Stahl, Angew. Chem. 2004, 116, 3480 – 3501; Angew.
Chem. Int. Ed. 2004, 43, 3400 – 3420; b) S. S. Stahl, Science 2005,
309, 1824 – 1826.
Overall, these palladium-catalyzed aerobic oxidative
cyclization reactions provide a highly modular, efficient and
scalable approach for the preparation of 1,2-diamine deriv-
atives. The straightforward retrosynthetic disconnection evi-
dent in Scheme 1 and the excellent diastereoselectivity of the
reaction suggest that this reactivity should have broad utility
in the synthesis of complex molecules. Moreover, the simple
catalyst system identified in this work has clear advantages
over previously identified systems, and we anticipate it will
find application in numerous other oxidative transformations
that use molecular oxygen as the stoichiometric oxidant.
[8] A. Borghese, L. Antoine, J. P. Van Hoeck, A. Mockel, A.
Merschaert, Org. Process Res. Dev. 2006, 10, 770 – 775.
[9] The Pd(OAc)2/DMSO catalyst system was originally discovered
by the groups of Hiemstra and Larock. For early reports, see:
a) R. A. T. M. van Benthem, H. Hiemstra, J. J. Michels, W. N.
Speckamp, J. Chem. Soc. Chem. Commun. 1994, 357 – 359;
b) R. C. Larock, T. R. Hightower, L. A. Hasvold, K. P. Peterson,
J. Org. Chem. 1996, 61, 3584 – 3585; c) M. Rꢁnn, J.-E. Bäckvall,
P. G. Andersson, Tetrahedron Lett. 1995, 36, 7749 – 7752.
[10] a) S. R. Fix, J. L. Brice, S. S. Stahl, Angew. Chem. 2002, 114, 172 –
174; Angew. Chem. Int. Ed. 2002, 41, 164 – 166; b) R. M. Trend,
Y. K. Ramtohul, B. M. Stoltz, J. Am. Chem. Soc. 2005, 127,
17778 – 17788.
Received: November 10, 2009
Revised: March 18, 2010
[11] See the Supporting Information.
[12] a) H. Grennberg, A. Gogoll, J.-E. Bäckvall, J. Org. Chem. 1991,
56, 5808 – 5811; b) M. S. Chen, M. C. White, J. Am. Chem. Soc.
2004, 126, 1346 – 1347; c) D. Tanaka, S. P. Romeril, A. G. Myers,
J. Am. Chem. Soc. 2005, 127, 10323 – 10333; d) M. S. Chen, N.
Prabagaran, N. A. Labenz, M. C. White, J. Am. Chem. Soc. 2005,
127, 6970 – 6971; e) K. J. Fraunhoffer, M. C. White, J. Am. Chem.
Soc. 2007, 129, 7274 – 7276; f) K. L. Hull, M. S. Sanford, J. Am.
Chem. Soc. 2007, 129, 11904 – 11905; g) G. Brasche, J. García-
Fortanet, S. L. Buchwald, Org. Lett. 2008, 10, 2207 – 2210.
[13] a) B. A. Steinhoff, S. R. Fix, S. S. Stahl, J. Am. Chem. Soc. 2002,
124, 766 – 767; b) B. A. Steinhoff, S. S. Stahl, J. Am. Chem. Soc.
2006, 128, 4348 – 4355.
Published online: June 25, 2010
Keywords: cyclization · diamines · homogeneous catalysis ·
.
oxidation · palladium
[1] For recent reviews on 1,2-diamines, see: a) D. Lucet, T. Le Gall,
C. Mioskowski, Angew. Chem. 1998, 110, 2724 – 2772; Angew.
Chem. Int. Ed. 1998, 37, 2580 – 2627; b) K. Muꢀiz, New J. Chem.
2005, 29, 1371 – 1385; c) D. Savoia, Top. Organomet. Chem. 2005,
15, 1 – 58; d) S. R. S. Saibabu Kotti, C. Timmons, G. Li, Chem.
Biol. Drug Des. 2006, 67, 101 – 114; e) R. M. de Figueiredo,
Angew. Chem. 2009, 121, 1212 – 1215; Angew. Chem. Int. Ed.
2009, 48, 1190 – 1193.
[14] For previous NMR spectroscopic characterization of PdII- and
PtII-DMSO complexes, see: a) J. A. Davies, F. R. Hartley, S. G.
Murray, J. Chem. Soc. Dalton Trans. 1979, 1705 – 1708; b) G.
Annibale, L. Cattalini, V. Bertolasi, V. Ferretti, G. Gilli, M. L.
Tobe, J. Chem. Soc. Dalton Trans. 1989, 1265 – 1271.
[2] a) Y. Tamaru, M. Hojo, Z. Yoshida, J. Org. Chem. 1988, 53,
5731 – 5741; b) Y. Tamaru, M. Hojo, H. Higashimura, Z.
Yoshida, J. Am. Chem. Soc. 1988, 110, 3994 – 4002;
c) R. A. T. M. van Benthem, H. Hiemstra, G. R. Longarela,
W. N. Speckamp, Tetrahedron Lett. 1994, 35, 9281 – 9284; d) H.
Harayama, A. Abe, T. Sakado, M. Kimura, K. Fugami, S.
Tanaka, Y. Tamaru, J. Org. Chem. 1997, 62, 2113 – 2122;
e) G. L. J. Bar, G. C. Lloyd-Jones, K. I. Booker-Milburn, J. Am.
Chem. Soc. 2005, 127, 7308 – 7309; f) J. Streuff, C. H. Hꢁvel-
mann, M. Nieger, K. Muꢀiz, J. Am. Chem. Soc. 2005, 127, 14586 –
14587; g) K. Muꢀiz, C. H. Hꢁvelmann, J. Streuff, E. Campos-
Gꢂmez, Pure Appl. Chem. 2008, 80, 1089 – 1096; h) P. A.
Sibbald, F. E. Michael, Org. Lett. 2009, 11, 1147 – 1149.
[3] The intermolecular diamination of terminal olefins with di-tert-
butyldiaziridinones provides an attractive approach to analo-
gous 1,2-diamine products; see: a) H. Du, W. Yuan, B. Zhao, Y.
Shi, J. Am. Chem. Soc. 2007, 129, 7496 – 7497; b) H. Du, B. Zhao,
Y. Shi, J. Am. Chem. Soc. 2008, 130, 8590 – 8591; c) B. Zhao, H.
Du, S. Cui, Y. Shi, J. Am. Chem. Soc. 2010, 132, 3523 – 3532.
[4] B. Wang, H. Du, Y. Shi, Angew. Chem. 2008, 120, 8348 – 8351;
Angew. Chem. Int. Ed. 2008, 47, 8224 – 8227.
[15] a) B. B. Wayland, R. F. Schramm, Inorg. Chem. 1969, 8, 971 –
976; b) D. P. Bancroft, F. A. Cotton, M. Verbruggen, Acta
Crystallogr. Sect. C 1989, 45, 1289 – 1292.
[16] A crystal structure of [PdII(DMSO)2(TFA)2] exhibits one O- and
one S-bound dimethyl sulfoxide ligand (see Ref. [14b]). The
infrared spectroscopic data reported here confirmed that similar
coordination properties were retained in solution and were not
simply an artifact of crystallization.
[17] Previous studies of pyridine coordination to palladium-carbox-
ylates show that pyridine is comparatively non-labile: a) see
Ref. [13b]; b) M. J. Schultz, C. C. Park, M. S. Sigman, Chem.
Commun. 2002, 3034 – 3035.
[18] See the following, and the extensive list of references cited
therein: a) L. E. Overman, N. E. Carpenter, Org. React. 2005, 66,
1 – 107; b) K. Brak, J. A. Ellman, J. Am. Chem. Soc. 2009, 131,
3850 – 3851.
[19] The need for multiple N-substituent manipulations, following
the diamination reported by Shi and co-workers, is evident from
a recent synthesis of (+)-CP-99,994: R. Fu, B. Zhao, Y. Shi,
J. Org. Chem. 2009, 74, 7577 – 7580.
[5] Ureas, which were unreactive under our oxidative conditions,
have been used in Pd0-catalyzed intramolecular carboamination
reactions: a) J. A. Fritz, J. S. Nakhla, J. P. Wolfe, Org. Lett. 2006,
5532
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5529 –5532