A practical, fast, and high-yielding aziridination procedure using simple Cu(II) complexes containing N-donor pyridine-based ligands

Article abstract of DOI:10.1021/jo050485f

Four-coordinate dichlorocopper(II) complexes derived from di(2-pyridyl)methanes or pyridine itself exhibit high catalytic activity in aziridination of regular olefins with PhINTs in weakly coordinating chloroform in the presence of 1-2 equiv of NaBAr

Full text of DOI:10.1021/jo050485f

A Practical, Fast, and High-Yielding Aziridination Procedure
Using Simple Cu(II) Complexes Containing N-Donor
Pyridine-Based Ligands
Fabian Mohr, Seth A. Binfield, James C. Fettinger, and Andrei N. Vedernikov*
Department of Chemistry and Biochemistry, University of Maryland, College Park, Maryland 20742
avederni@umd.edu
Received March 10, 2005
Four-coordinate dichlorocopper(II) complexes derived from di(2-pyridyl)methanes or pyridine itself
exhibit high catalytic activity in aziridination of regular olefins with PhINTs in weakly coordinating
chloroform in the presence of 1-2 equiv of NaBArF
(BAr
F -
) tetra[3,5-di(trifluoromethyl)phenyl]-
4
4
borate). High yields of aziridines exceeding 90% can be obtained with a 1:1 olefin/PhINTs ratio
and 1-5 mol % catalyst loading for such reactive olefins as styrene, tri- and tetramethylethylene.
For cis-cyclooctene, indene, methyl acrylate, methyl methacrylate, vinyl methyl ketone, tert-
butylethylene, and neopentylethylene, as well as for 1-hexene and cyclopentene, yields of
corresponding aziridines vary from 44% to 83%. The catalytic activity and efficiency of the reported
copper complexes decrease moderately in the absence of NaBArF
.
4
Introduction
(10-3) are still the most important limitations of cur-
rently employed protocols for olefin aziridination. These
limitations are especially important when the olefinic
substrate and/or catalyst are expensive. In addition to
Aziridines are important synthetic intermediates in
modern organic synthesis, which can be obtained, in
particular, by reacting olefins with suitable sources of
5
11-15
1
theoretical modeling and experimental study
of the
nitrenes such as iminoiodinanes. The importance of the
reaction mechanism, a better understanding of the
structure-reactivity relationship for copper catalysts
could contribute significantly to the solution of these
problems. Until recently, most copper-based catalysts for
latter reaction steadily increases as more and more
efficient catalytic routes are discovered.2
-10
High catalyst
loadings (up to 5 mol %) and high olefin/PhINTs ratios
olefin aziridination were prepared in situ and have rarely
(
1) Dauban, P.; Dodd, R. H. Synlett 2003, 1571.
been structurally characterized,1
3-16
thus making a sys-
tematic search of any structure-reactivity relationships
difficult. Recently we have shown that copper complexes
(
2) Evans, D. A.; Faul, M. M.; Bilodeau, M. T. J. Org. Chem. 1991,
5
1
1
6, 6744.
(
3) Evans, D. A.; Bilodeau, M. T.; Faul, M. M. J. Am. Chem. Soc.
994, 116, 2742.
(
4) Perez, P. J.; Brookhart, M.; Templeton, J. L. Organometallics
n
LCuX (X ) Cl, OTf; n ) 1, 2) containing the macrocyclic
993, 12, 261.
5) Brandt, P.; Soedergren, M. J.; Andersson, P. G.; Norrby, P.-O.
J. Am. Chem. Soc. 2000, 122, 8013.
6) Mairena, M. A.; Diaz-Requejo, M. M.; Belderrain, T. R.; Nicasio,
M. C.; Trofimenko, S.; Perez, P. J. Organometallics 2004, 23, 293.
7) Albone, D. P.; Aujla, P. S.; Taylor, P. C.; Challenger, S.; Derrick,
A. M. J. Org. Chem. 1998, 63, 9569.
(
(11) Li, Z.; Quan, R. W.; Jacobsen, E. N. J. Am. Chem. Soc. 1995,
(
117, 5889.
(12) Gillespie, K. M.; Crust, E. J.; Deeth, R. J.; Scott, P. Chem.
Commun. 2001, 785.
(
(13) Comba, P.; Merz, M.; Pritzkow, H. Eur. J. Inorg. Chem. 2003,
1711.
(
8) Li, Z.; Conser, K. R.; Jacobsen, E. N. J. Am. Chem. Soc. 1993,
15, 5326.
9) Kantam, M. L.; Neeraja, V.; Kavita, B.; Haritha, Y. Synlett 2004,
25.
1
(14) Gillespie, K. M.; Sanders, C. J.; O’Shaughnessy, P.; Westmo-
reland, I.; Thickitt, C. P.; Scott, P. J. Org. Chem. 2002, 67, 3450.
(15) Halfen, J. A.; Fox, D. C.; Mehn, M. P.; Que, L., Jr. Inorg. Chem.
2001, 40, 5060.
(
5
(10) Langham, C.; Piaggio, P.; Bethell, D.; Lee, D. F.; McMorn, P.;
Page, P. C. B.; Willock, D. J.; Sly, C.; Hancock, F. E.; King, F.;
Hutchings, G. J. Chem. Commun. 1998, 1601.
(16) Thorhauge, J.; Roberson, M.; Hazell, R. G.; Jorgensen, K. A.
Chem. Eur. J. 2002, 8, 1888.
1
0.1021/jo050485f CCC: $30.25 © 2005 American Chemical Society
Published on Web 05/06/2005
J. Org. Chem. 2005, 70, 4833-4839
4833

Mohr et al.
tripyridine ligand L ) [2.1.1]-(2,6)-pyridinophane are
very active catalysts for aziridination of a variety of
olefins, allowing fast and high yielding reactions when
3
1
-5 equiv of olefin per 1 equiv of PhINTs was used (eq
1
7
):
Since dpm ligand contains potentially vulnerable ben-
zylic CH bonds, its dimethylated analogue, 2,2-di(2-
pyridyl)propane (dpp), was expected to be more robust
toward highly oxidizing PhINTs. Similarly, as a result
of the electron-withdrawing nature of the chlorine sub-
stituent, the benzylic CH bonds in (6-chloro-2-pyridyl)2-
pyridylmethane (Cl-dpm) were expected to be of lower
reactivity than those in dpm, whereas complexes derived
from (6-methyl-2-pyridyl)2-pyridylmethane (Me-dpm)
might be less robust. Pyridine (py) and dipyridyl (dipy)
complexes with no benzylic CH bonds or bridging carbons
at all might be of superior efficiency as compared with
dpm-based copper complexes. Finally, for the reason of
better solubility in reaction mixtures, a copper derivative
High activity of the catalysts was achieved in poorly
coordinating dichloromethane solutions when the anionic
ligands (X) were removed with NaBArF (BArF
4
4
) tet-
rakis(3,5-di(trifluoromethyl)phenyl)borate) creating the
coordinatively unsaturated cationic copper species
n+
m
[LCuX ]
(n + m ) 2 or 1). Two factors, the coordination
unsaturation that promotes faster coordination of PhINTs
to copper and the macrocycle ring constraints that make
the catalysts less vulnerable to oxidative degradation,
may be responsible for the enhanced catalytic activity.
In support of the latter idea copper complexes based on
perbrominated hydridotris(pyrazolyl)borate were recently
of 4-tert-butylpyridine (tBupy) was also included in our
19
study. A series of (L′)CuCl
2
complexes, (dpm)CuCl
2
,
20
,21
(
Me-dpm)CuCl
2
, (Cl-dpm)CuCl
2
, (dpp)CuCl
2 2 2
, (py) CuCl
6
found to be very active in olefin aziridination. In this
22
(
dipy)CuCl
2
,
and (tBupy)
2
CuCl
2
, have been prepared,
work we show that the presence of a facially chelating
ligand attached to copper is not a requirement for
creating highly active catalysts and that coordinative
unsaturation at copper alone can be responsible for the
and their catalytic activity in olefin aziridination was
studied in this work. A related cationic four-coordinate
+
-
2 2
copper(I) analogue, [(Cl-dpm) Cu] [CuCl ] , was also
synthesized and tested to emphasize the effect of the
metal coordination unsaturation on the catalytic activity.
To determine the degree of coordination unsaturation of
copper in these complexes, we analyzed results of their
X-ray crystal structure determination.
Preparation and Solid-State Structure Charac-
terization of Copper Catalysts. A series of copper(II)
very high catalyst activity. Furthermore, we show here
F
that simple BAr
4
-free coordinatively unsaturated copper-
(
II) complexes can serve as efficient aziridination cata-
lysts when a 1:1 olefin/PhINTs ratio is used. In combi-
nation with the simplicity of preparation and low cost of
the catalysts reported here, these results imply develop-
ment of very practical and versatile catalytic systems.
complexes of the type (L′)CuCl
Me-dpm, Cl-dpm, dipy, (py) , and (tBupy)
in 75-94% yield by reacting CuCl with a 10% excess of
2
, where L′ ) dpm, dpp,
2
2
, was prepared
Results and Discussion
2
On the basis of the results from our previous work,17,18
strong pentacoordinate copper(II) or tetracoordinate cop-
per(I) complexes such as (η -L)CuCl and (η -L)CuCl are
2
ligand in THF or benzene. The resulting green or blue
complexes were characterized by elemental analysis and,
in the case of (Cl-dpm)CuCl and [(Cl-dpm) Cu] [CuCl ] ,
2 2 2
by X-ray crystallography. X-ray crystal structures of all
of the other complexes studied here with the exception
+
-
3
3
poor catalysts for olefin aziridination. However, replace-
ment of chloride by much more labile triflate ligands or
F
of (dpp)CuCl
2
and (tBupy)
2
CuCl
According to the data available, (dpm)-
exists in a solid state as a bis(µ-chloro)-bridged
dimer containing five-coordinate copper atoms in a
2
have been determined
removal of Cl with NaBAr
4
increases the catalytic
previously.1
CuCl
9-22
activity significantly. At this point, it was reasonable to
assume that four-coordinate copper(II) or thee-coordinate
copper(I) complexes might be efficient catalysts for olefin
aziridination in weakly coordinating solvents. Coordinat-
ing solvents such as acetonitrile are not compatible with
low-coordinate copper species. Potentially four-coordinate
2
19
distorted trigonal-bipyramidal environment. We assume
that the analogous (dpp)CuCl
2
complex has a similar
dinuclear structure.
copper(II) complexes of the type (L′)CuCl
2
, where L′
In contrast, (Me-dpm)CuCl complex is monomeric with
2
denotes a bidentate N-donor ligand such as di(2-pyridyl)-
a copper atom that adapts highly distorted tetrahedral
geometry presumably due to steric crowding caused by
methane (dpm) or two pyridine ligands, can be considered
3
20
as unsaturated analogues of (η -L)CuCl
2
in which one of
the o-methyl substituents. Indeed, a similar metal
the three pyridine rings as well as some or all of its
alkylydene bridges have been removed.
coordination geometry was found by us for the chloro-
(
20) Garcia, A. M.; Manzur, J.; Spodine, E.; Baggio, R. F.; Garland,
(
17) Vedernikov, A. N.; Caulton, K. G. Org. Lett. 2003, 5, 2591.
18) Vedernikov, A. N.; Huffman, J. C.; Caulton, K. G. Inorg. Chem.
M. T. Acta Crystallogr., C 1994, C50, 1882.
(
(21) Morosin, B. Acta Crystallogr., Sect. B 1975, B31, 632.
(22) Garland, M. T.; Grandjean, D.; Spodine, E.; Atria, A. M.;
Manzur, J. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1988,
C44, 1209.
2
002, 41, 6244.
(
19) Spodine, E.; Manzur, J.; Garland, M. T.; Fackler, J. P., Jr.;
Staples, R. J.; Trzcinska-Bancroft, B. Inorg. Chim. Acta 1993, 203, 73.
4834 J. Org. Chem., Vol. 70, No. 12, 2005

Aziridination Procedure Using Simple Cu(II) Complexes
Finally, the Cl-dpm copper(I) complex synthesized
contains both unsaturated and saturated copper(I) atoms
and therefore might exhibit some catalytic activity.
x
Catalytic Olefin Aziridination in the PhINTs/Cu /
NaBArF
4 3
/CHCl Systems. (a) Copper(II) Complexes.
Catalytic studies with cis-cyclooctene as a test substrate
showed that all (L′)CuCl complexes at 5 mol % loading
activated with 2 equiv of NaBAr
2
F
4
in chloroform solutions
exhibited high catalytic activity at the 3:1 olefin/PhINTs
ratio with almost quantitative yields of corresponding
aziridines for the all the copper(II) catalysts. These
results allowed us to decrease the olefin/PhINTs ratio to
1
:1, which still gave high yields of corresponding reaction
products (Table 1, entry 1). Yields ranging from 66%
dpm) to 86% (L′ ) dpp and (tBupy) ) are very similar to
the results obtained with the macrocyclic [LCuX ]-
catalyst (87%). Remarkably, one of the best
(
2
m
(
BArF
)
4 n
17
results in entry 1 was obtained with the simplest catalyst
used in this work, the dichlorodi(pyridine)copper(II)
complex. It is also worth noting that the efficiency of the
catalysts derived from dpm and tested here with cis-
cyclooctene and other substrates increases in the same
order as their expected ability to withstand oxidative
degradation increases, L′: Me-dpm ≈ dpm < Cl-dpm ≈
analogue (Cl-dpm)CuCl 23
2
(see Figure 1 in Supporting
Information).
The dichloro(dipyridyl)copper(II) complex (dipy)CuCl
2
containing the more compact dipy ligand compared to
dpm is polymeric in the solid state with copper atoms
linked by bridging chlorides.22 The copper(II) pyridine
dpp < (tBupy)
2
≈ (py)
2 2
. Interestingly, (dipy)CuCl exhib-
complex, (py)
coordinate planar copper atom.
2 2
CuCl , is monomeric and contains a four-
21
ited the same or slightly lower activity than dpm and
Me-dpm complexes. Possible reason for this behavior may
The product obtained by reacting copper(I) chloride and
Cl-dpm taken in 1:1 ratio in dichloromethane turned out
to be ionic with one copper atom in its cationic part being
four-coordinate, in a highly distorted tetrahedral envi-
be lower solubility of the (dipy)CuCl
form caused by its polymeric nature.
To learn more about the scope and limitations of this
catalytic system we studied other olefinic substrates
containing one, two, three and four substituents, both
2
complex in chloro-
+
24
ronment, (Cl-dpm)
2
Cu
(see Figure 2 in Supporting
Information), and another copper atom in the anionic
-
activating (Ph, o-C
6 4 2
H CH , Me, n-Bu, t-Bu, neo-Am,
2
part being two-coordinate, CuCl .
(
CH ) and deactivating (COMe, COOMe), attached to
2 3
)
Because the copper atom in the cationic part of this
copper(I) complex forms much longer and therefore
weaker bonds with the nitrogen atoms of the chloro-
substituted pyridine fragments, N1 and N12, it is reason-
able to expect relatively easy displacement of the chlo-
ropyridine fragments by stronger electron donors if such
are present in solution.
the CdC bond. Aziridination of the most reactive of them,
styrene, was performed in all of the dpm-based systems
with indistinguishably high, almost quantitative yields
in less than 1-2 min (Table 1, entry 2). Remarkably, as
2 2
little as 1% of (py) CuCl -based catalyst was enough to
obtain the same yield of aziridine though the reaction
took more time.
Thus, according to the solid-state structure studies, we
High yields of aziridines exceeding 90% were also
observed for tetra- (entry 3) and trimethylethylene (entry
4). Worse though still satisfactory performance was
exhibited by indene (83%, entry 5), substrates with
electron-withdrawing COOMe (methyl methacrylate, 69%,
entry 6; methyl acrylate, 44%, entry 7) or COMe group
2
could expect that all (L′)CuCl complexes investigated
here might exhibit high catalytic activity in olefin aziri-
dination in weakly coordinating solvents when NaBArF
4
is used to remove the chloride ligands. More importantly,
we could also expect that the four monomeric four-
coordinate copper(II) complexes, (Me-dpm)CuCl
dpm)CuCl , (py) CuCl , and (tBupy) CuCl may be active
aziridination catalysts in the absence of NaBAr
nuclear copper complexes (dpm)CuCl and (dpp)CuCl
2
, (Cl-
(
1-butene-2-one, 65%, entry 8), bulky t-Bu group (tert-
2
2
2
2
2
F
butyl ethylene, 55%, entry 5), or alkyl groups with
4
. Di-
secondary allylic CH bonds,7 cyclopentene (52%, entry
,17
2
2
1
0) and 1-hexene (49%, entry 11). The importance of the
containing longer than regular and therefore weaker Cu-
Cl bonds might also be active enough in NaBArF
presence of the electron-rich CdC bond in olefinic sub-
strates is clearly evidenced, in particular, by the fact that
methyl methacrylate performed much better than methyl
acrylate. On the other hand, the steric effect of substit-
uents directly attached to the CdC bond on its reactivity
is evident when one compares the better performance of
4,4-dimethyl-1-pentene (67%, entry 12) with less bulky
neo-pentyl group attached to the CdC bond and the worse
performance of tert-butyl ethylene where the steric effect
of the alkyl group is more pronounced.
4
-free
systems.
(
23) Greenish crystals of (Cl-dpm)CuCl
2
(C11
H
9
Cl
3 2
CuN ) are mono-
clinic, space group C2/c, with a ) 21.008(16) Å, b ) 8.111(7) Å, c )
3
1
5.895(13) Å, â ) 99.683(13)°, V ) 2670(4) Å and Z ) 8 at -100 °C.
2
The full-matrix least-squares refinement on F provided residuals R
1
)
2
0.0467, wR
σ(I)] and 191 variables.
2
) 0.1145 and GOF ) 1.047 for 4266 reflections [I >
(
6
24) Yellowish crystals of [(Cl-dpm)
2
2 2 2 2 20 4
Cu][CuCl ]‚CH Cl (C23H N -
Cl
Cu ) are centrosymmetric monoclinic, space group P21/n (No. 14)
with a ) 11.6293(15) Å, b ) 13.7395(18) Å, c ) 16.846(2) Å, â )
3
1
05.510(2)°, V ) 2593.6(6) Å and Z ) 4 at -100 °C. The full-matrix
2
(
b) Copper(I) Complex. With cis-cyclooctene as a
least-squares refinement on F provided residuals R
0.0833 and GOF ) 1.050 for 4564 reflections [I > 2σ(I)] and 396
variables.
1 2
) 0.0341, wR
substrate we tested also the catalytic activity of Cl-dpm-
)
+
-
derived copper(I) catalyst, [(Cl-dpm)
2 2
Cu] [CuCl ] , in the
J. Org. Chem, Vol. 70, No. 12, 2005 4835

Mohr et al.
TABLE 1. Results of Olefin Aziridination with 1 equiv of PhINTs in the Presence of 5 mol % of (L′)CuCl2 and 2 equiv of
NaBArF4 in CHCl3 at 296 K
a
yield of aziridine, % (time, min)
entry
substrate
dpm
dpp
Cl-dpm
Me-dpm
(py)2
dipy
(tBupy)2
1
2
3
4
5
6
7
8
9
cis-cyclooctene
styrene
tetramethylethylene
trimethylethylene
indene
methyl methacrylate
methyl acrylate
1-butene-2-one
tert-butylethylene
cyclopentene
66 (1)
95 (1)
76 (5)
82 (1)
95 (1)
81 (5)
83 (1)
98 (1)
91 (2)
67 (5)
93 (2)
75 (1)
83 (1)
97 (10)
94 (2)
90 (5)
83 (5)
69 (10)
44 (5)
65 (8)
55 (3)
52 (5)
49 (10)
67 (5)
73 (10)
98 (5)
50 (5)
86 (5)
94 (2)
85 (1)
94 (1)
85 (2)
56 (10)
32 (10)
72 (8)
50 (3)
51 (5)
47 (10)
65 (2)
b
35 (30)
49 (10)
45 (30)
60 (3)
45 (20)
63 (5)
38 (25)
31 (40)
1
1
1
0
1
2
1-hexene
4,4-Dimethyl-1-pentene
a
NMR yields on PhINTs. The isolated yields were 1-5% lower. b 1 mol % of catalyst.
TABLE 2. Results of Olefin Aziridination (n ) 1 or 5 equiv) with PhINTs in the Presence of 5 mol % of (L′)2CuCl2 and
NaBArF4 (x ) 0 or 1 equiv) in CHCl3 at 296 K
a
yield of aziridine, % (time, min)
n ) 1; x ) 1
n ) 1; x ) 0
n ) 5; x ) 0
entry
substrate
styrene
tetramethylethylene
trimethylethylene
cis-cyclooctene
indene
methyl methacrylate
methyl acrylate
1-butene-2-one
4,4-dimethyl-1-pentene
1-hexene
tert-butylethylene
cyclopentene
(py)2
dpm
Cl-dpm
Me-dpm
(py)2
(py)2
1
2
3
4
5
6
7
8
9
93 (5)
84 (5)
88 (5)
73 (3)
75 (7)
68 (6)
69 (60)
65 (9)
64 (5)
50 (5)
30 (10)
37 (5)
97 (20)
51 (30)
81 (5)
67 (2)
65 (20)
55 (30)
92 (5)
68 (2)
71 (5)
62 (2)
75 (12)
72 (5)
33 (15)
59 (20)
54 (5)
34 (9)
31 (15)
39 (6)
98 (5)
70 (2)
98 (5)
98 (2)
84 (30)
65 (2)
72 (2)
83 (5)
74 (4)
44 (9)
39 (15)
42 (6)
50 (15)
57 (1)
36 (60)
1
1
1
0
1
2
21 (30)
19 (15)
a
NMR yields on PhINTs. The isolated yields were 1-5% lower.
presence of 2 equiv of NaBArF
PhINTs ratio. While the reaction set up under the same
conditions was equally fast in the case of (Cl-dpm)CuCl
and at the 3:1 olefin/
increased if a similar level of catalytic activity were possi-
ble at a reduced content or in the absence of NaBAr
4
.
4
F
Catalytic Olefin Aziridination in NaBArF
2
-Poor
4
+
-
2 2
and [(Cl-dpm) Cu] [CuCl ] catalysts (1-2 min), the yield
and NaBArF
4
-Free Systems. (a) Systems with 1
F
of aziridine obtained in the latter case was almost 20%
lower. Assuming that the catalytic activity of the copper-
equiv of NaBAr . Results of the aziridination of styrene
4
and tetra- and trimethylethylene given in Table 2
(
I) catalyst may be due to the presence of low-coordinate
(column 3) show that in the presence of only 1 equiv of
-
NaBArF high yields of aziridines exceeding 84% can still
copper anion CuCl
NaBAr
2
we tried pure CuCl + 1 equiv of
4
F
4
system and observed only very slow reaction
be obtained for the three most reactive substrates with
(120 min) with the aziridine yield of 50%. Therefore, the
the (py) CuCl catalyst (entries 1-3), whereas the results
2
2
observed activity of the Cl-dpm copper(I) complex can be
more likely ascribed to the cationic part of this copper(I)
compound, which has relatively weak Cu-chloropyridine
bonds and therefore can readily open up a coordination
vacancy at the copper atom.
for cis-cyclooctene (73%, entry 4), indene (75%, entry 5),
methyl methacrylate (68%, entry 6), methyl acrylate
(69%, entry 7), 1-butene-2-one (65%, entry 8), 4,4-
dimethyl-1-pentene (64%, entry 9), and 1-hexene (50%,
entry 10) are less satisfactory. The performance of tert-
butylethylene (30%, entry 11) and cyclopentene (37%,
The catalytic results obtained here allowed us to con-
entry 12) is poor. The yields and/or reaction rates for all
clude that simplified analogues of a previously reported
F
macrocyclic copper complex systems17 show almost the
NaBAr -poor systems are slightly worse in comparison
4
F
with systems activated with 2 equiv of NaBAr
4
.
same level of catalytic activity. Thus, coordinative unsat-
uration of the copper atom is the most important attrib-
ute of a complex that is highly active in olefin aziridina-
tion with PhINTs in weakly coordinating solvents. This
conclusion opens up the possibility for further developing
highly active metal catalysts for this reaction. Though
the dichlorodi(pyridine) copper(II) based systems reported
here are good examples of such practical catalysts, the
value of the catalytic systems described could be further
These results confirm again that the degree of coordi-
nation unsaturation of copper-based systems is an im-
portant factor determining their activity as aziridination
catalysts. In accordance with this conclusion, the perfor-
mance of the NaBArF
-free systems including dpm- or
pyridine-based catalysts was found to be lower.
(b) NaBArF
-Free Systems. Reflected in longer reac-
tion times (up to 60 min) and/or lower yields of aziridines
4
4
4836 J. Org. Chem., Vol. 70, No. 12, 2005

Aziridination Procedure Using Simple Cu(II) Complexes
II
25,26
SCHEME 1
of Cu -nitrene adducts)
or concerted (as in the case
transfer to olefin. A consecu-
tive nitrene transfer from the radical-like nitrene adduct
I
11,25
of Cu -nitrene adducts)
II
26
B has been proposed for Cu -catalyzed reactions, and
this also may be the case of the copper-pyridinophane
systems, which showed a partial loss of the initial CdC
bond configuration.3 We obtained a similar result in
this work with isomerically pure cis-2-butene and PhINTs
,17
taken in the 5:1 ratio in the presence of 5% of CuCl
and 2 equiv of NaBArF
. The experiment showed fast and
practically quantitative conversion of the olefin into the
.7:1 mixture of the cis- and trans-isomeric 2,3-dimethyl-
2 2
(py)
4
4
N-tosyl-aziridines and thus the partial loss of the olefin
(
up to 28%), this trend can be seen by comparing the
effectiveness of the dpm, Cl-dpm, Me-dpm, and di-
pyridine) dichlorocopper(II) complexes given in Tables
and 2 (columns 4-7). Despite the decreased efficiency,
still high yields of aziridine of 92-97% can be obtained
for styrene (entry 1). The yields with the (py) CuCl
configuration.
To figure out if the radical-like behavior can be
(
1
2
expected for nitrene adducts such as trans-[CuCl(py) -
+
(
NSO
2
Ar)] , which can form in PhINTs-(py)
2
CuCl
2
-
F
NaBAr systems as a result of chloride ligand abstrac-
tion with Na and nitrene coordination to copper(II), or
4
+
2
2
complex as a catalyst (column 5) are still satisfactory for
tetramethylethylene (68%, entry 2), trimethylethylene
F
2 2 4
trans-CuCl (py)(NSO Ar), which can form in NaBAr -
free systems and be viewed as a result of pyridine ligand
substitution by nitrene, we performed DFT calculations
(71%, entry 3), cis-cyclooctene (62%, entry 4), indene
(
75%, entry 5), methyl methacrylate (72%, entry 6),
-butene-2-one (59%, entry 8), and 4,4-dimethyl-1-pen-
+
for two model intermediates, trans-[CuCl(py)
see Figure 3a in Supporting Information) and trans-
CuCl (py)(NSO Ph) (see Figure 3b in Supporting Infor-
2 2
(NSO Ph)]
1
(
tene (54%, entry 9) substrates. The yields for three
remaining substrates are poor and range from 31% to
2
2
mation). The results of our DFT calculations show that
both adducts have slightly distorted square planar
geometry typical for d metal complexes. The resonance
3
9%.
To learn if we can improve yields of aziridines in this
very simple and therefore attractive NaBArF
-free cata-
8
III •
2
structure C, LCu ( NSO Ar), has an important contribu-
4
lytic system we changed the olefin/PhINTs ratio from 1:1
used everywhere in this work (all data in Table 1 and
columns 3-7 in Table 2) to 5:1 (Table 2, last column).
Modest (3-10%) to significant (36%, the case of cis-
cyclooctene) yield increase was observed. These results
tion to the copper-nitrene adduct structure in both cases.
Indeed, the Mulliken spin density distribution in the
cationic complex is 1% on copper and 99% on the NSO
Ph ligand including 74% on the nitrene nitrogen and 2%
on copper and 98% on the NSO Ph ligand including 75%
2
-
2
imply that simple (py)
ficient in olefin aziridination if greater than 1:1 olefin/
PhINTs ratios are affordable (olefin substrate is cheap).
The differences in catalyst performance are clearly
visible from the data given in Table 2 and follow the
trends discussed for data in Table 1. In particular,
2
CuCl
2
complex may be also ef-
-NaBArF
(
2 2
py) CuCl complex performs better than other complexes
for all substrates, while the Me-dpm complex is signifi-
cantly less efficient. Another observation based on com-
parison of dpm and Me-dpm complexes is that the ability
of a ligand to withstand oxidative degradation (L′ ) dpm)
might be more important in determining catalyst ef-
ficiency than monomeric structure (L′ ) Me-dpm) of (L′)-
2
CuCl complexes. Thus, four-coordinate dichlorocopper(II)
complexes alone can catalyze aziridination of activated
olefins efficiently, and dichlorodi(pyridine)copper(II) com-
plex in CHCl solvent represents an active and inexpen-
3
sive catalytic system for olefin aziridination with PhINTs.
Mechanistic Considerations. The use of low-coor-
dinate copper species as aziridination catalysts for reac-
tive olefinic substrates dramatically accelerates the
otherwise slow PhINTs coordination to copper leading
presumably to a soluble copper-iminoiodinane adduct A
(Scheme 1, step 1) in a poorly coordinating solvent.
Coordination to the metal may be required for an
intramolecular electron transfer leading to liberation of
PhI and formation of a plausible highly reactive copper-
nitrene adduct B (step 2). In general, the copper-nitrene
adducts may be involved in a consecutive (as in the case
(
25) Brandt, P.; S o¨ dergren, M. J.; Anderson, P. G.; Norrby, P.-O. J.
Am. Chem. Soc. 2000, 122, 8013 and references therein.
26) D ´ı az-Requejo, M. M.; Rerez, P. J.; Brookhart, M.; Templeton,
J. L. Organometalics 1997, 16, 4399.
(
J. Org. Chem, Vol. 70, No. 12, 2005 4837

Mohr et al.
Conclusions
(dpm)CuCl
Cl
43.56, H, 3.21; N, 8.87.
(dpp)CuCl : blue solid (yield: 82%). Anal. Calcd for C13
Cu (332.7): C, 46.93; H, 4.24, N, 8.42. Found: C, 46.73; H,
.27; N, 8.02.
Cl-dpm)CuCl
for C11 Cl Cu (339.1): C, 38.96; H, 2.68; N, 8.26. Found:
C, 38.56; H, 2.42; N, 7.90.
Me-dpm)CuCl : lime green solid (yield 89%). Anal. Calcd
for C12 Cl Cu (318.7): C, 45.23; H, 3.80; N, 8.79. Found:
C, 44.95; H, 3.76; N, 8.70.
tBupy) CuCl : sky blue solid (yield 85%). Anal. Calcd for
Cl Cu (404.9): C, 53.40; H, 6.47; N, 6.92. Found: C,
3.50; H, 6.32; N, 6.87.
6-Chloro-2-pyridyl)(2-pyridyl)methane Copper(I) Com-
plex, [(Cl-dpm) Cu][CuCl ] CH Cl . A suspension of CuCl
2
: green solid (yield 85%). Anal. Calcd for
C
11
H
10
N
2
2
Cu (304.7): C, 43.37; H, 3.31; N, 9.19. Found: C,
We have shown that unsaturated four-coordinate
dichlorocopper(II) complexes are very active catalysts for
aziridination of activated and regular olefins, which
perform well at 1-5% loading when 1:1 olefin/PhINTs
is used in chloroform solutions. An enhancement of their
2
14 2
H N -
Cl
4
2
(
2
: bright green solid (yield 75%). Anal. Calcd
catalytic activity is observed in the presence of NaBArF
H
9
N
2
3
4
.
Thus, simple, cheap, and readily available four-coordinate
copper(II) compounds such as (py) CuCl can be recom-
mended as efficient catalysts for aziridination of a variety
of olefins in weakly coordinating solvents.
(
2
2
2
H
12
N
2
2
(
H
2
2
C
18
26
N
2
2
5
Experimental Section
(
Computational Details. Theoretical calculations including
2
2
2
2
Mulliken spin population analysis in this work have been
(0.100 g, 1.00 mmol) in dichloromethane (ca. 5 mL) was treated
with a dichloromethane solution of (6-chloro-2-pyridyl)(2-
pyridyl)methane (1.1 equiv), and the mixture was left for
2
7
performed using density functional theory (DFT) method,
2
8
specifically functional PBE, implemented in an original
2
9
program package “Priroda”. In PBE calculations relativistic
Stevens-Basch-Krauss (SBK) effective core potentials (ECP)
overnight. The resulting yellowish solid was isolated by
1
filtration, washed well with Et
NMR (CD Cl
2
O, and dried (yield 90%). H
30-32
optimized for DFT calculations have been used. The basis
2
2
, 400 MHz) 4.41 (br s, 2H), 7.33 (br m, 1H), 7.40
set was 311-split for main group elements with one additional
polarization p-function for hydrogen, additional two polariza-
tion d-functions for elements of higher periods. Full geometry
optimization has been performed without constraints on
symmetry. For all species under investigation frequency
analysis has been carried out. All minima have been checked
for the absence of imaginary frequencies.
(br d, J ) 7.1 Hz, 1H), 7.55 (d, J ) 7.6 Hz, 1H), 7.61 (br d, J
) 7.6 Hz, 1H), 7.81 (t, J ) 7.8 Hz, 1H), 7.86 (dt, J ) 1.5, 7.8
20 4 6 2
Hz, 1H), 8.43 (br s, 1H). Anal. Calcd for C23H N Cl Cu
(692.2): C, 39.91; H, 2.91; N, 8.09. Found: C, 40.52; H, 2.32;
N, 8.27.
Aziridination Experiments. In a drybox the copper
catalyst (10 µmol, 5 mol %) and NaBArF if any (17.6 mg, 20
4
(
6-Chloro-2-pyridyl)(2-pyridyl)methane (Cl-dpm). An
µmol or 8.9 mg, 10 µmol) were placed into a small sample vial
equipped with a magnetic stirrer bar. The olefin (200 µmol)
ice-cooled solution of 2-picoline (9.9 mL, 0.1 mol) in THF (40
n
mL) was treated with BuLi (10 mL, 10 M solution in hexanes).
3
dissolved in CDCl (0.5 mL) was added with stirring, and
After 30 min of stirring, a solution of 2,6-dichloropyridine (7.4
g, 0.05 mol) in THF (20 mL) was added, and the mixture was
heated to reflux for 1 h. After cooling to room temperature
the reaction mixture was hydrolyzed with water, and the
organic phase was separated. The aqueous phase was ex-
PhINTs (74.6 mg, 200 µmol) was introduced immediately. The
reaction time was defined as the time required for all PhINTs
to dissolve. All resulting reaction mixtures were ultimately
bright green, leaving no doubt that the catalyst resting state
II
is Cu .
tracted with Et
were dried (Na
ing dark oil was distilled in a vacuum to afford 6.5 g (64%) of
2
O (3 × 20 mL) and the combined organic layers
The NMR yields on PhINTs were calculated from NMR
integrals of the aziridine and iodobenzene resonances. In a
number of cases we used also dichloromethane as an internal
standard. The latter (5.0 µL) was added after the reaction was
complete. In all the aziridination reactions iodobenzene liber-
ated quantitatively based on PhINTs and therefore it could
be considered and used as an “internal standard” by itself. To
ensure that there is no errors associated with integration of
the signals of aromatic and aliphatic protons due to significant
difference in their relaxation times, when taking NMR spectra
we used the NMR relaxation delay of 8.0 s. Greater values of
the delay had no effect on the integral ratios.
2
SO ). After removal of the solvents the remain-
4
1
Cl-dpm with a bp of 107-110° at 0.23 mmHg. H NMR (CDCl
3
,
4
)
1
4
1
2
00 MHz) 4.31 (s, 2H), 7.13 (dd, J ) 1, 4 Hz, 1H), 7.17 (dd, J
4, 7 Hz, 2H), 7.28 (dd, J ) 1, 8 Hz, 1H), 7.55 (t, J ) 8 Hz,
H), 7.61 (dt, J ) 2, 8 Hz, 1H), 8.55 (m, 1H); 13C NMR (CDCl
3
,
00 MHz) 46. 4, 121.6, 121.8, 121.9, 123.5, 136.6, 139.0, 149.4,
50.6, 158.5, 160.2; HRMS (FAB+) calcd for C11
10 2
H N Cl m/z
05.0533, found m/z 205.0538.
Dichlorocopper(II) Complexes, (L′)CuCl
of CuCl (0.100 g, 0.744 mmol) in THF or benzene (in the case
of tBupy ligand) (ca. 5 mL) was treated with a THF or benzene
solution of the ligand (1.1 equiv), and the mixture was left for
ca. 18 h. The resulting green or blue solids were isolated by
filtration, washed well with THF or benzene and Et
dried. Thus obtained were
2
. A suspension
2
To confirm the identity of the aziridine by 1H NMR
17,33
13
spectroscopy,
and in a few cases by C NMR spectroscopy,
and estimate its isolated yield, the reaction mixture was
filtered through a short column filled with alumina and eluted
with small amount of dichloromethane. This method allowed
reliable separation of copper catalyst, any “inorganic” compo-
nents of the reaction mixture and efficient purification of the
aziridine. The NMR-pure aziridine was isolated from the
filtrate after removal of solvent, unreacted olefin and iodo-
benzene under high vacuum. The yield of the isolated aziridine
was no more than 5% lower compared with the NMR yield.
2
O, and
(
py)
10
2
N
CuCl
Cl Cu (292.7): C, 41.04; H, 3.44; N, 9.57. Found: C,
1.14; H, 3.42; N, 9.27.
dipy)CuCl : turquoise solid (yield 93%). Anal. Calcd for
Cl Cu (290.6): C, 41.33; H, 2.77; N, 9.64. Found: C,
2
: sky blue solid (yield 94%). Anal. Calcd for
C
4
10
H
2
2
(
2
C
4
10
H
8
N
2
2
1.24; H, 2.54; N, 9.29.
3
4
1
2
-Acetyl-N-(p-toluenesulfonyl)-aziridine.
H NMR
3
(
CDCl
3
, 22 °C) δ: 2.00 (s, 3H), 2.39 (s, 3H), 2.42 (d, JH-H
)
)
(
27) Parr, R. G.; Yang, W. Density-Functional Theory of Atoms and
3
3
4
.2 Hz, 1H), 2.72 (d, JH-H ) 7.4 Hz, 1H), 3.21 (dd, JH-H
3 3
Molecules; Oxford University Press: Oxford, 1989.
4
(
.2 Hz, JH-H ) 7.4 Hz, 1H), 7.29 (d, JH-H ) 8.4 Hz, 2H), 7.76
(
28) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996,
3
13
d, JH-H ) 8.4 Hz, 2H). C NMR (CDCl
3
, 22 °C) δ: 21.8, 26.0,
7
7, 3865.
(
(
29) Laikov, D. N. Chem. Phys. Lett. 1997, 281, 151.
30) Stevens, W. J.; Basch, H.; Krauss, M. J. Chem. Phys. 1984, 81,
32.0, 42.1, 128.3, 130.1, 134.0, 145.5.
6
026.
(31) Stevens, W. J.; Basch, H.; Krauss, M.; Jasien, P. Can. J. Chem.
(33) Chanda, B. M.; Vyas, R.; Bedekar, A. V. J. Org. Chem. 2001,
66, 30.
(34) Pak, C. S.; Kim, T. H.; Ha, S. J. J. Org. Chem. 1998, 63, 10006.
1
992, 70, 612.
(32) Cundari, T. R.; Stevens, W. J. J. Chem. Phys. 1993, 98, 5555.
4838 J. Org. Chem., Vol. 70, No. 12, 2005

Aziridination Procedure Using Simple Cu(II) Complexes
1
2
-neo-Pentyl-N-(p-toluenesulfonyl)-aziridine. H NMR
spectrum was recorded. Yield based on the NMR integrals of
the aziridine and iodobenzene liberated was 94% on PhINTs
(an average of two experiments). To confirm the identity of
the aziridine, the reaction mixture was filtered through a 2
cm column filled with alumina and eluted with 6 mL of
dichloromethane. NMR-pure aziridine was isolated from the
filtrate after removal of solvent, the unreacted olefin, and
iodobenzene under high vacuum. Isolated yield 47.0 mg (93%).
(
3
CDCl , 22 °C) δ: 0.90 (s, 9H), 1.25 (m, 1H), 1.45 (m, 1H), 1.94
3
3
(
d, JH-H ) 4.6 Hz, 1H), 2.41 (s, 3H), 2.57 (d, JH-H ) 7.0 Hz,
3
3
1
H), 2.78 (m, 1H), 7.30 (d, JH-H ) 8.4 Hz, 2H), 7.79 (d, JH-H
1
3
)
3
8.4 Hz, 2H). C NMR (CDCl , 22 °C) δ: 21.7, 29.5, 30.7,
3
4.1, 37.7, 45.5, 128.1, 129.8, 135.4, 144.6.
Mixture of cis- and trans-2,3-Dimethyl-N-tosyl-aziri-
1
7
2 2
dines. In the experiment with 5% CuCl (py) , 2 equiv of
F
4
NaBAr , and 5:1 olefin/PhINTs the ratio of the cis- and trans-
isomeric products was 4.7:1.The reaction took ca. 5 min. Both
Acknowledgment. We thank the University of
Maryland for financial support of this work.
NMR and isolated yields of the two aziridines on PhINTs are
1
quantitative. H NMR (CDCl
3
, 22 °C) select peaks δ: 1.15 (vd,
3
3
J
H-H ) 5.0 Hz, 6H, Me, cis-aziridine), 1.40 (d, JH-H ) 5.0
Hz, 6H, Me, trans-aziridine), 2.70 (m, 2H, CH, trans-aziridine),
Supporting Information Available: Crystallographic
2
.84 (m, 2H, CH, cis-aziridine).
experiments, description and full crystallographic details of
1
Example of Aziridination of Tetramethylethylene
(Cl-dpm)CuCl
2
and [(Cl-dpm)
2
Cu][CuCl
2
] in CIF format; H
1
3
with 1 equiv of PhINTs Catalyzed with 5% of (py)
2
CuCl
CuCl
4
2.9 mg, 10 µmol, 5mol %) and NaBArF (17.6 mg, 20 µmol)
2
-
2
and C NMR spectra of (Cl-dpm), 2-neo-pentyl-N-(p-toluene-
sulfonyl)aziridine and [(Cl-dpm) Cu][CuCl ]; Cartesian coor-
dinates and results of the Mulliken spin density calculations
2
(
NaBΑrF
4
.Thedichlorodi(pyridine)copper(II)complex,(py)
2
,
2
2
+
were placed into a vial with a magnetic stirrer bar. Tetram-
2 2 2 2
for trans-[(py) CuCl(NSO Ph)] and trans-(py)CuCl (NSO Ph);
ethylethylene (16.8 mg, 200 µmol) was weighed in a separate
and electronic absorption spectrum of the sample reaction
mixture. This material is available free of charge via the
Internet at http://pubs.acs.org.
3
vial, diluted with 0.5 mL of CDCl , and added with stirring to
the mixture above. Immediately after that PhINTs (74.6 mg,
2
00 µmol) was introduced. In 2 min all PhINTs dissolved to
JO050485F
give clear green solution. UV-vis spectrum recorded im-
mediately after that showed the presence of two broad bands
II
35
at 736 and 846 nm, in the region typical for Cu compounds.
(
35) Lever, A. B. P. Inorganic Electronic Spectroscopy; Elsevier:
1
The liquid was transferred into a NMR tube, and the H NMR
Amsterdam, New York, 1984.
J. Org. Chem, Vol. 70, No. 12, 2005 4839