Synthesis, structure, and highly efficient copper-catalyzed aziridination with a tetraaza-bispidine ligand

Article abstract of DOI:10.1002/chem.200802682

The distorted trigonal-bipyramidal Cu" complex [Cu(L ¹)(NCCH₃)J²⁺ of the novel tetradentate bispidine-derived ligand L¹ with four tertiary amine donors (L ¹ = 1,5-diphenyl-3-methyl-7-(l,4,6-trimethyl-l,4-diazacycloheptane-6- yl)diazabicyclo[3.3.1]nonane-9-one) is a very efficient catalyst for the aziridination of olefins in the presence of a nitrene source. In agreement with the experimental data (in situ spectroscopy, product distribution, and its dependence on the geometry of the substrate and of the nitrene source), a theoretical analysis based on DFT calculations indicates that the active catalyst has the Cu center in its + 11 oxidation state, that electron transfer is not involved, and that the conversion of the olefin to an aziridine is a stepwise process involving a radical intermediate. The striking change of efficiency and reaction mechanism between classical copper-bispidine complexes and the novel L¹-based catalyst is primarily attributed to the structural variation, enforced by the ligand architecture.

Full text of DOI:10.1002/chem.200802682

DOI: 10.1002/chem.200802682
Synthesis, Structure, and Highly Efficient Copper-Catalyzed Aziridination
with a Tetraaza-Bispidine Ligand
Peter Comba,*[a] Christina Haaf,[a] Achim Lienke,[b] Amsaveni Muruganantham,[a] and
Hubert Wadepohl[a]
Abstract: The distorted trigonal-bipyr-
amidal complex
CuII [Cu(L1)-
(in situ spectroscopy, product distribu-
tion, and its dependence on the geome-
try of the substrate and of the nitrene
source), a theoretical analysis based on
DFT calculations indicates that the
active catalyst has the Cu center in its
+II oxidation state, that electron trans-
fer is not involved, and that the conver-
sion of the olefin to an aziridine is a
stepwise process involving a radical in-
termediate. The striking change of effi-
ciency and reaction mechanism be-
tween classical copper–bispidine com-
plexes and the novel L1-based catalyst
is primarily attributed to the structural
variation, enforced by the ligand archi-
tecture.
ACHTUNGTRENNUNG
(NCCH3)]2+ of the novel tetradentate
bispidine-derived ligand L1 with four
tertiary amine donors (L1 =1,5-diphen-
yl-3-methyl-7-(1,4,6-trimethyl-1,4-diaza-
cycloheptane-6-yl)diazabicycloACTHNUTRGNE[UNG 3.3.1]-
nonane-9-one) is a very efficient cata-
lyst for the aziridination of olefins in
the presence of a nitrene source. In
agreement with the experimental data
Keywords: aziridination · copper ·
homogeneous catalysis · N ligands ·
reaction mechanisms
Introduction
The chemistry of three-membered ring heterocycles, espe-
cially that of the aziridines, has inspired organic chemistry
for decades,[1–3] and the development of suitable methods for
their synthesis has a long history.[4,5] Aziridines may be pre-
pared by stoichiometric or catalytic reactions (see
Scheme 1). In stoichiometric reactions, the efficiency strong-
ly depends on the nitrogen source and the alkene precursor;
the advantage is a good accessibility of chiral azirdines.[6,7]
The transition-metal-catalyzed route is often used due to its
efficiency, and many examples have been studied in
detail.[1,3,5,8–13] Due to its high reactivity, PhINTs ([N-(p-tol-
uenesulfonyl)imino]phenyliodinane) is generally used as the
Scheme 1. Copper-catalyzed preparation of aziridines.
nitrene source, although others such as chloramine T have
economic and ecological advantages.[4,5,8,14,15]
Various reaction mechanisms have been proposed for the
transition-metal-catalyzed processes, and these include step-
wise pathways with radical intermediates as well as concert-
ed reactions, and nitrene complex intermediates with the
metal ions in various oxidation states (e.g., for copper: CuI,
CuII, CuIII) have been suggested; mechanisms that do not in-
clude any electron-transfer steps at the metal center and
Lewis acid based mechanisms with non-redox-active metal
ions such as Zn2+ and Mg2+ have also been pro-
posed.[5,8,16–20] We have characterized a series of CuII/I cou-
ples with tetra- and pentadentate bispidine-based ligands
(see L2 for a tetradentate derivative),[21–28] and these were
shown to be reasonably active aziridination catalysts.[19,29]
We now describe the synthesis and characterization of the
novel bispidine-type ligand L1 and its CuII complex, as well
as the highly efficient copper-catalyzed aziridination and a
[a] Prof. P. Comba, C. Haaf, Dr. A. Muruganantham, Prof. H. Wadepohl
Universitꢀt Heidelberg
Anorganisch-Chemisches Institut
INF 270, 69120 Heidelberg (Germany)
Fax : (+49)6226-546617
E-mail: peter.comba@aci.uni-heidelberg.de
[b] Dr. A. Lienke
Unilever R&D Vlaardingen
Olivier van Noortlaan 120
3133AT Vlaardingen (The Netherlands)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200802682.
10880
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10880 – 10887

FULL PAPER
mechanistic analysis, based on spectroscopy, product analy-
ses, and DFT calculations.
Scheme 2. Synthesis of [Cu(L1)
N
.
The transition-metal coordination chemistry of the tradi-
tional bispidines is rich in its structural variety and broad in
its applications.[30] The majority of bispidine ligands reported
so far has two or more aromatic nitrogen donor groups and
enforces coordination geometries derived from cis octahe-
dral with metal–donor distances and angular geometries
around the metal center strictly enforced by the rigid ada-
mantanoid bispidine backbone. Although complexes with
specifically distorted cis-octahedral geometries, imposed by
the first-generation bispidine ligands, for example, L2, have
produced exciting results,[30,31] specifically in the areas of
ligand-enforced molecular geometries and the emerging
properties,[32–35] and metal-ion-mediated oxygen activa-
tion,[36–40] we have designed a similarly rigid ligand system
that may impose different coordination geometries to the
metal center and offer a large variety of donor sets. We
report here the first example of a second generation of bis-
pidine ligands (L1 versus L2), in which the extra donors are
exclusively due to substitution at the amine components of
the Mannich reaction. This enabled us to prepare a ligand
(L1), which enforces a (distorted) trigonal-bipyramidal struc-
ture with tertiary amines as the only donor groups.
Figure 1. Molecular structure of [Cu(L1)
are omitted for clarity).
(NCCH3)]2+ (hydrogen atoms
ACHTUNGTRENNUNG
data are given in Table 1. The coordination geometry of the
CuII center is distorted trigonal bipyramidal, with N3 of the
bispidine ligand and the coordinated MeCN at axial posi-
tions. As expected from other bispidine complexes,[30,32] the
bond lengths between CuII and the two tertiary amine nitro-
gen atoms of the bispidine backbone are approximately
2.0 ꢂ. The bond lengths between the CuII center and the
other equatorial amine nitrogen atoms (N1 and N2) are sig-
nificantly longer. This may partially be due to the rather
small angle of just under 778 between N1 and N2, enforced
by the 5-membered chelate ring. As a result, the other two
Results and Discussion
Synthesis and characterization of the copper(II) catalyst:
The new bispidine ligand L1 is based on the cyclic triamine
6-amino-1,4,6-trimethyl-1,4-diazacycloheptane, obtained in a
simple two-step reaction from bulk chemicals.[41,42] In a two-
step Mannich condensation, the cyclic triamine is trans-
formed into the piperidone P1 in acceptable yields (see
Scheme 2). This general scheme gives access to a wide range
of derivatives with differing donor sets and geometries. The
tetraaza ligand L1 is the first member of this novel family of
ligands and is obtained in very good yield from P1. A water-
Table 1. Selected
(BF4)2·CH3CN.
structural
parameters
of
[Cu(L1)
ACHTUNGTRENNUNG(NCCH3)]-
AHCTUNGTRENNUNG
Bond lengths [ꢂ]
Valence angles [8]
free sample of [CuII(L1)
ACHTNUGTRENNGU(NCCH3)]AHCTUNGTREN(NUGN BF4)2 was obtained from
ꢀ
Cu1 N7
2.032(3)
1.999(3)
2.169(4)
2.195(4)
1.969(4)
N3-Cu1-N4
176.75(15)
86.15(13)
83.85(14)
87.50(13)
76.56(14)
144.61(14)
137.14(14)
the corresponding CuI complex, which reacted in situ with
AgBF4 to give the product.
ꢀ
Cu1 N3
N3-Cu1-N1
N3-Cu1-N2
N3-Cu1-N7
N1-Cu1-N2
N7-Cu1-N1
N7-Cu1-N2
ꢀ
Cu1 N1
Single crystals of [Cu(L1)
ACTHUNRGTNNEUG(NCCH3)]AHCTUNGTRENN(GUN BF4)2·CH3CN were
ꢀ
Cu1 N2
ꢀ
Cu1 N4
obtained from a solution of the complex in MeCN, and the
structure was solved by X-ray diffraction. A plot of the mo-
lecular cation appears in Figure 1 and selected structural
Chem. Eur. J. 2009, 15, 10880 – 10887
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10881

P. Comba et al.
angles in the trigonal plane are much larger (137 and 1458);
the angles between the axial and equatorial ligands are all
around 908. The bond from CuII to the axial MeCN coligand
is 1.97 ꢂ, thus in the expected range.
activity. In contrast, smaller yields were observed in these
experiments (entries 8–10). Therefore, it appears that in the
aziridination reaction studied here, the solubility of the cat-
ionic catalyst and the poorly soluble PhINTs substrate are
responsible for the observed reactivity trends. Catalytic
aziriACTHUNRTGNEUNGdination experiments with trifluoromethanesulfonate in-
stead of tetrafluoroborate as counterion did not show en-
hanced catalytic efficiency (entry 11).
For a trigonal-bipyramidal geometry a (dz2)1 ground state
is expected, and this should lead to the following spin Ham-
iltonian parameters: g factors g? >gjj and hyperfine struc-
ture constants jA? j~jAjj j (the A values are expected to be
around 100 G, and the sign of Ajj is usually positive, where-
as that of A? depends on the electronics and may vary).[43]
Several observations suggested that the {CuII(L1)}2+-cata-
lyzed process follows a novel pathway with a CuII–nitrene
intermediate, which initiates a stepwise conversion of sty-
rene to the corresponding aziridine, and therefore involves a
radical intermediate and no change in the CuII oxidation
state. The latter assumption is supported by UV/Vis spectra
of {CuII(L1)}2+/PhINTs mixtures, in which no {CuI(L1)}+ was
detected under conditions that allowed spectrophotometric
trapping of CuI intermediates in other systems:[44] the spec-
trum in Figure 2 taken after the addition of PhINTs shows
The X-band EPR spectrum of [Cu(L1)
(NCCH3)]2+ has been
ACHTUNGTRENNUNG
measured and reveals a dx2ꢀy2 ground state (gjj =2.21, g?
=
2.08, Ajj =170 G, A? =156 G; see the Supporting Informa-
tion). This unconventional behavior is due to the relatively
large distortion from trigonal-bipyramidal geometry. Specifi-
cally, the angular distortion within the “trigonal” plane
(CuII, N1, N2, N7) leads to nondegenerate dx2ꢀy2 and dxy or-
bitals and, in particular, to a destabilization of the dx2ꢀy2 or-
bital. This also follows from DFT calculations, which unam-
biguously show the dx2ꢀy2 orbital to be highest in energy of
the d orbital manifold (see below and the Supporting Infor-
mation).
ACTHNUTRGNEUNG
Experimental mechanistic studies: [Cu(L1)(NCCH3)]2+ is
one of the most efficient catalysts for the aziridination of
styrene with PhINTs.[17–19,29] Various solvents and catalyst/
substrate/nitrene ratios were studied (see Table 2), and
linear pseudo-first-order kinetic behavior for styrene con-
centrations up to 20 equivalents was observed (entries 5–7; a
variation of the amount of PhINTs did not affect the aziridi-
nation reactivity). Nonpolar solvents are expected to favor
radical reactions. However, changing the solvent from
CH3CN to noncoordinating solvents such as CH3NO2 or
CH2Cl2/toluene, which were expected to accelerate the addi-
tion of PhINTs to the precatalyst, did not lead to higher re-
Figure 2. UV/Vis experiments: a={CuII(L1)}2+ (10ꢀ4
c={CuII(L1)}2++PhINTs (10ꢀ4 m in CH3CN).
m in CH3CN);
two bands in the UV region with increasing intensity but no
significant change of the d–d transition in the visible region
at l=630 nm. This indicates that in the {CuII(L1)}2+/PhINTs
system, there is no reduction of CuII to CuI. Attempts to
trap and characterize a {CuI(L1)}+ complex failed because
{CuI(L1)}+ readily disproportionates to {CuII(L1)}2+ and
{Cu0}. This is also evident from the irreversible electrochem-
ical behavior of {CuII(L1)}2+ in MeCN (Eir,CuI,CuII =ꢀ377 mV
Table 2. Aziridination of styrene with PhINTs, catalyzed by [Cu(L1)-
ACHTUNGTRENNUNG
(NCCH3)]2+ in MeCN.[a]
Entry Equiv Solvent [ml]
styrene
Catalyst t [min] Yield TON[b]
[mol%]
[%]
1
2
3
4
5
6
7
8
22
22
22
1
20
10
1
22
1
1
CH3CN (2)
CH3CN (2)
CH3CN (2)
CH3CN (0.5)
CH3CN (2)
5
5
40
200
27
11
28
47
25
30
30
88
85
85
70
92
91
87
64
17
3
18
168
339
70
92
91
87
13
17
3
0.5
0.25
1
1
1
1
5
2
2
and Eir,CuII
=+1720 mV, respectively, versus Ag/AgNO3).
,CuIII
The absence of any formation of {Cu0} (disproportionation
of CuI, see above) is further support for our interpretation
that {CuII(L1)}2+ is the active catalyst and does not change
the oxidation state in the catalytic cycle.
CH3CN (2)
CH3CN (2)
CH3NO2 (2)
CH3NO2 (0.5)
CH2Cl2 (0.5)/toluene (2)
CH3CN (0.5)
CH3CN (2)
9
10
11[c]
12
ACTHNUTRGNEUNG
The [Cu(L1)(NCCH3)]2+-catalyzed conversion of cis-b-
1
22
1
0
30
420
40
0
40
0
methylstyrene with PhINTs produces both cis- and trans-
aziridine products (19% trans, see Table 3), which suggests
that the aziridination is a stepwise process with a radical in-
termediate of significant life time.[5] Due to this radical,
there is stereochemical scrambling at the former double
bond as shown by the value of 62% for the diastereomeric
ACHTUNGTRENNUNG
[a] [Cu(L1)(NCCH3)]2+ is the most efficient bispidine-derived aziridina-
tion catalyst reported so far and only slightly less efficient than recently
reported CuII complexes.[18] The maximum efficiency observed with Cu
complexes of bispidine ligands derived from L2 are TON 270 in 24 h
(5 mol% cat., 46% yield, 24 h).[19,29] [b] TON=n
G
ACHTUNGTRENUN(NG catalyst).
[c] Experiments carried out with [Cu(L1)
ACTHNUGRTEN(NUNG NCCH3)]ACHTUGNTRENNNUG
10882
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10880 – 10887

Copper-Catalyzed Aziridination
FULL PAPER
Table 3. Stereoselective aziridination of cis-b-methylstyrene with
adduct reveals that PhI is immediately dissociated to form a
CuII nitrene species in a fast and facile reaction. This is in
contrast to observations in similar computational studies
with other bispidine-based aziridination catalysts[19,29] but is
consistent with the data presented in Table 3.
ACTHNUTRGNEUNG
PhINTs, catalyzed by [Cu(L1)(NCCH3)]2+ in MeCN.
Nitrene source
cis/trans ratio
de[a] [%]
t [min]
Yield [%]
68
81:19
62
85
Due to the rigidity and steric demand of the bispidine
ligand system and the enforced trigonal-bipyramidal geome-
try, but in contrast to similar reactions with other bispidine-
derived catalyst systems,[19,29] the nitrene only acts as a
monodentate ligand. The catalytically active CuII nitrene
species has two unpaired electrons on the nitrene N and one
on the CuII center, and a quartet as well as a doublet spin
state is possible. The doublet is the ground state (direct ex-
change between Cu (dx2ꢀy2) and Nnitrene orbitals), with the
quartet state at 35.7 kJmolꢀ1. The doublet ground state has
a distorted trigonal-bipyramidal structure with the Nnitrene
trans to N3 on the pseudo-trigonal axis (see Figure 3), and
84:16
68
120
100
[a] de=diastereomeric excess.
excess (de). To confirm the proposed radical-based mecha-
nism, a sterically more demanding PhINTs derivative was
also used as a nitrene source. PhI is assumed to be fully dis-
sociated from the nitrene source in the radical intermediate
(see also “Computational studies” below) but covalently
bound in a Lewis acid process.[5] Bulky substituents on the
PhI group are therefore expected to change the cis/trans
ratio in the aziridination product of cis-b-methylstyrene in
the case of a Lewis acid based process but not in a radical
mechanism. The fact that there is no significant change of
the product ratio as a function of steric bulk (62 vs. 68% de,
see Table 3) indicates that the {CuII(L1)}-catalyzed aziridina-
tion with PhINTs derivatives follows a radical-based mecha-
nism.
Computational studies: DFT calculations were used to con-
firm the mechanistic conclusions derived from the experi-
ments and to understand the fundamental reasons for the
change of mechanism between the complexes discussed here
and other Cu-based aziridination catalyst systems.[19,20,29,45,46]
Based on published experimental and computational obser-
vations, Scheme 3 was adopted for the DFT calculations. In
ACTHNUTRGNEUNG
the initial step, [Cu(L1)(NCCH3)]2+ reacts with PhINTs to
Figure 3. Potential-energy diagram of the aziridination reaction with opti-
mized structures of the relevant species (rad-int= radical intermediate):
doublet (S=1= , lower energy), quartet (S=3= , higher energy). The opti-
form a {CuII(L1)···PhINTs} precursor. Optimization of this
2
2
mized structures shown are those of the S=1= pathway, those on the S=
2
3= reaction are similar (see Table 3) and not shown for clarity.
2
the distribution of the spin density as expected for a doublet
state (Cu: ꢀ0.255, Nnitrene: 1.329; the spin densities on the
CuII center and all nitrogen atoms in the coordination
sphere are given in Table 5; Table 4 reports selected struc-
tural parameters of the computed structures). The excess
spin density on Nnitrene facilitates the oxidation of the olefin.
The spin on the CuII center is located in the dx2ꢀy2 orbital, as
ACTHNUTRGNEUNG
observed in the EPR spectrum of the [Cu(L1)(NCCH3)]2+
precursor complex (see above and the Supporting Informa-
tion), with the lobes of the orbital directed towards N2, N3,
N7, and Nnitrene, in which spin delocalization is dominating.
ꢀ
The now axial N1 Cu bond is significantly elongated
(2.409 ꢂ) and therefore gains spin density mainly through
spin polarization (see Tables 4 and 5). The drastic reduction
of spin density on Nnitrene in the doublet as compared to the
quartet state is due to strong spin delocalization, that is, in
Scheme 3. Proposed mechanistic scheme for the aziridination reaction
due to {CuII(L1)}2+
.
Chem. Eur. J. 2009, 15, 10880 – 10887
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10883

P. Comba et al.
Table 4. Selected B3LYP-computed structural parameters of the reactants, transition states (ts), intermediates (int), and products in the aziridination of
ACTHNUTRGNEUNG
alkenes, catalyzed by [Cu(L1)(NCCH3)]2+ (see Scheme 3).
CuII–complex
2CuII–nitrene
4CuII–nitrene
17.4
2ts1
4ts1
2int1
4int1
8.4
CuII–aziridine
E [kJmolꢀ1
]
0.0
ꢀ18.3
36.9
44.1
ꢀ6.9
ꢀ196.0
bond lengths [ꢂ]
ꢀ
Cu N7
2.004
2.033
2.182
2.144
–
–
–
–
2.096
2.006
2.409
2.166
1.948
3.316
–
2.112
2.069
2.470
2.120
2.032
3.768
–
2.148
2.091
2.424
2.229
2.202
2.976
1.994
2.777
2.154
2.090
2.434
2.221
2.205
3.013
2.039
2.805
2.156
2.096
2.514
2.168
2.242
3.076
1.496
2.507
2.162
2.101
2.462
2.211
2.224
3.007
1.481
2.504
2.201
2.103
2.179
2.627
2.306
2.706
1.557
1.545
ꢀ
Cu N3
ꢀ
Cu N1
ꢀ
Cu N2
ꢀ
Cu Nnitrene
ꢀ
Cu O
ꢀ
Nnitrene C1
ꢀ
Nnitrene C2
–
–
valence angles [8]
N7-Cu-Nnitrene
N3-Cu-Nnitrene
N1-Cu-Nnitrene
N2-Cu-Nnitrene
Cu-Nnitrene-S
Nnitrene-C1-C2
N7-Cu-N1
–
–
–
–
–
–
92.2
173.4
99.3
88.3
114.6
–
94.9
170.6
90.5
95.7
136.7
–
95.9
177.9
97.4
95.8
178.3
96.9
96.8
174.6
96.1
97.5
176.8
97.1
95.6
168.3
99.8
93.2
100.3
60.8
94.5
95.7
95.1
94.2
104.7
107.7
143.1
139.2
106.1
107.0
142.8
139.3
107.2
113.3
137.1
145.4
105.9
113.5
139.1
142.7
137.9
143.4
141.6
143.0
137.0
147.1
152.2
131.1
N7-Cu-N2
The stepwise pathway is consistent with the observed cis/
Table 5. Calculated spin densities of various species along the potential-
energy surface for the aziridination of alkenes, catalyzed by [Cu(L1)-
ꢀ
trans product mixture. Calculations to estimate the C C ro-
ACHTUNGTRENNUNG
(NCCH3)]2+ (see Scheme 3).
tation barrier of the doublet radical intermediate were per-
formed by a relaxed coordinate scan and gave an energy
barrier of approximately 14 kJmolꢀ1, which supports a possi-
ble cis/trans conversion during the ring-closure step (the
energy barrier for the product-formation step is approxi-
mately 48 kJmolꢀ1, see below). To examine the effect of dif-
ferent substrates, we have also calculated the rotational bar-
riers for other radical intermediates, such as those derived
from other common substrates, that is, but-2-ene and pent-
2-ene in addition to cis-b-methylstyrene. The results suggest
that the substitution pattern has only a small influence on
the rotation-barrier height at the doublet radical intermedi-
ate stage (differences of the maxima are less than
5 kJmolꢀ1, see the Supporting Information, i.e., the rotation-
al barrier is always significantly lower than the barrier for
product formation). We have not been able to locate a true
transition state for the product-formation step (decay of the
radical intermediate to the aziridination product but a re-
laxed coordinate scan reveals that the barrier height on the
doublet surface is approximately 40 kJmolꢀ1. The final prod-
CuII– 2CuII– 4CuII–
complex nitrene nitrene
2ts1
4ts1 2int1 4int1
CuII–
aziridine
Cu
N7
N3
N1
N2
Nnitrene
O1
O2
C1
0.549 ꢀ0.255 0.421 ꢀ0.513 0.504 ꢀ0.501 0.485
0.099 ꢀ0.026 0.142 ꢀ0.105 0.119 ꢀ0.106 0.123
0.277 ꢀ0.101 0.189 ꢀ0.177 0.189 ꢀ0.192 0.196
ꢀ0.002 0.080 ꢀ0.148 ꢀ0.008 0.048 ꢀ0.001 0.038
0.015 ꢀ0.060 0.116 ꢀ0.047 0.050 ꢀ0.062 0.062
0.531
0.109
0.175
0.098
0.003
0.034
0.003
–
–
–
–
–
–
1.329 1.726 1.195 1.273 1.008 1.028
0.030 0.156 0.059 0.088 0.033 0.039
0.029 0.057 0.025 0.034 ꢀ0.003 0.002 ꢀ0.000
–
–
– ꢀ0.003 0.029 ꢀ0.085 ꢀ0.078
– 0.637 0.645 0.893 0.945
0.001
0.001
C2
S
ꢀ0.049 ꢀ0.053 ꢀ0.045 ꢀ0.043 ꢀ0.010 ꢀ0.005 ꢀ0.001
the doublet state the negative spin density on the copper
center is transferred to Nnitrene
.
Once the CuII–nitrene species is formed and reacts with
an alkene, the conversion to the aziridine product may
occur either by a concerted or a stepwise process. All at-
tempts to optimize a transition state for a concerted reaction
lead to the intermediate of the stepwise pathway, and a re-
laxed coordinate scan had an energy barrier much higher
than that of the stepwise reaction (138 vs. 55 kJmolꢀ1; the
computed bond lengths of the species on the potential-
energy surface are given in Table 4). The formation of the
doublet nitrene precursor is exothermic by ꢀ18.3 kJmolꢀ1
but the quartet nitrene precursor is endothermic by
+17.4 kJmolꢀ1. Starting with the nitrene precursor, the first
energy barrier (transition state 1, ts1) is 55.2 kJmolꢀ1 on the
doublet and 26.7 kJmolꢀ1 on the quartet spin surface. This
leads to the radical intermediate, with the doublet radical
being endothermic by 11.4 kJmolꢀ1, relative to the nitrene
precursor; formation of the quartet radical is exothermic by
ꢀ9.0 kJmolꢀ1.
uct with
a doublet ground state is exothermic by
ꢀ196.0 kJmolꢀ1 from the reactant and by ꢀ177.7 kJmolꢀ1
from the nitrene precursor. The cleavage of the Cu–aziridine
complex to regenerate the CuII catalyst is an exothermic
process by ꢀ27.8 kJmolꢀ1.
The structural parameters presented in Table 4 and the
spin densities in Table 5 show the development of the
course of the reaction along the potential-energy surface. In
ts1, when the alkene approaches the nitrene complex, the
Cu–Nnitrene distance is significantly elongated from 1.948 to
ꢀ
2.202 ꢂ, and therefore is weakening the Cu Nnitrene bond.
Also, the spin density on the Nnitrene atom decreases signifi-
cantly in ts1 (1.329 to 1.195 ꢂ), as expected when a new
bond is formed. The transfer of spin density from Nnitrene to
10884
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10880 – 10887

Copper-Catalyzed Aziridination
FULL PAPER
AS-200 or Bruker DRX-200 instruments, respectively, with deuterated
solvents as reference. IR measurements were performed with a Perkin–
Elmer 16 PC FTIR instrument as KBr pellets. Mass spectra were record-
ed on a JEOL JMS-700 or a Finnigan TSQ 700/Bruker ApexQe hybrid
9.4 FT-ICR instrument; nibeol=4-nitrobenzyl alcohol. Electronic spectra
were obtained from a Tidas II J&M spectrophotometer. EPR spectra
were measured on a Bruker ELEXSYS-E-500 instrument at 125 K, spin
Hamiltonian parameters were obtained by simulation of the spectra with
XSophe.[47,48] For electrochemical measurements, a BAS-100B worksta-
tion with a three-electrode setup, consisting of a glassy carbon working
electrode, a Pt-wire auxiliary electrode, and Ag/AgNO3 reference elec-
the newly forming nitrogen–carbon bonds occurs as expect-
ed through spin polarization. Moreover, the spin density on
the carbon atoms of the alkene in ts1 is significant and sug-
gests that the alkene already has radical character in the
transition state. In the successive steps the radical intermedi-
ꢀ
ate has a further elongated Cu Nnitrene bond and a decreased
ꢀ
spin density. These trends continue until the Cu Nnitrene
bond stretches to 2.31 ꢂ with spin densities close to zero in
the CuII–aziridination product. At the very long distance of
trode (0.01m Ag+, 0.1m (Bu4N)
ACTHNUGTRNEUGN(PF6) in CH3CN, dry, degassed, under
ꢀ
2.31 ꢂ the Cu Nnitrene bond is susceptible to cleavage and re-
Ar), was used; the potential of the Fc+/Fc couple had a value of 91 mV
in CH3CN . Elemental analyses were obtained from the analytical labora-
tories of the chemical institutes at the University of Heidelberg, using a
Vario EL instrument of Elementar.
sults in the release of the aziridination product and regener-
ation of the catalyst.
Caution: Although no difficulties were observed with the perchlorate
salts of the complexes described here, these are potentially explosive and
should be handled with extreme care. Heating, especially when dry, must
be avoided.
Conclusion
With the novel tetradentate bispidine ligand system L1, we
have been able to enforce a distorted trigonal-bipyramidal
CuII coordination geometry, and this produces a very effi-
cient aziridination catalyst that is shown to follow a mecha-
nistic scheme that has not been observed before. The
PhINTs moiety couples under cleavage of PhI with the CuII
precatalyst. This leads to a relatively low-energy CuII–ni-
trene complex that produces a radical intermediate with
long enough lifetime to allow for structural rearrangement.
Experimentally, the fact that CuII is not reduced along the
catalytic cycle, is supported by electronic spectroscopy and
electrochemistry, and a radical intermediate with significant
lifetime emerges from the stereochemical scrambling with
cis-b-methylstyrene as substrate. The computed energetics
(see Figure 3) suggest that the entire aziridination reaction
occurs on the doublet spin surface. However, as the gap is
small at the first transition state, there might be some in-
volvement of the quartet state in the early stages of the re-
action. As expected from the experimental observations, the
reaction is thermodynamically favorable and the energy bar-
riers are small compared with other aziridination reac-
tions.[19,20,29,45,46] The important differences between this and
other Cu-catalyzed aziridination reactions are 1) the ligand-
enforced structure of the catalytically active Cu–nitrene
complex with an unusual electronic ground state; 2) energet-
ically unfavorable electron transfer (i.e., the reduction of
the CuII–nitrene to a CuI–nitrene precursor does not occur;
the computed energy difference of 13.9 kJmolꢀ1 indicates
that, in contrast to other systems, this is an endothermic pro-
cess, as observed experimentally); 3) lower energy barriers
and higher thermodynamic stability of the products, driving
the reaction faster along the potential-energy surface;
4) facile regeneration of the active catalyst.
1-(1,4,6-Trimethyl-1,4-diazacycloheptane-6-yl)-3,5-diphenylpiperidine-4-
one (P1): 1,4,6-trimethyl-6-amino-1,4-diazacycloheptane[41,49] (6.0 g,
38.1 mmol) was dissolved in 1,2-dimethoxyethane (DME; 20 mL). At
08C glacial acetic acid (7.5 mL) and formaldehyde solution (6.4 g,
79.1 mmol) were added, followed by 1,3-diphenylpropane-3-one (8.0 g,
38.1 mmol) at RT. The reaction mixture was stirred at 908C for 6 h. After
cooling to RT all solvents were removed in vacuo and the resulting red
oil was suspended in diethyl ether (500 mL). 70% HClO4 was added
slowly with rigorous stirring to precipitate the piperidone perchlorate.
The precipitate was washed with 1:1 EtOH/H2O and crystallized from
CH3CN/diethyl ether (1:1) to yield the monoperchlorate salt of the piper-
idone. This was suspended in diethyl ether and stirred with 2m KOH.
After isolation of the organic phase, the aqueous solution was extracted
again with diethyl ether, until no solid remained. The organic solvent was
dried with Na2SO4, and the solvent was removed under reduced pressure,
to yield the piperidone P1 (6.2 g, 12.5 mmol, 33%). 1H NMR (200 MHz,
[D3]CH3CN, 258C, TMS, perchlorate): d=1.0 (s, 3H; CH3), 2.6 (m, 6H;
CH3), 2.8–3.4 (m, 12H; CH2), 4.2 (dd, 2J(H,H)=12 Hz, 3J
ACTHNUTRGENNUG ACHUTGNTREN(NUNG H,H)=6 Hz,
2H; CH), 7.2–7.4 ppm (m, 10H; CHPh); 13C NMR (50 MHz, [D3]CH3CN,
258C, TMS, perchlorate): d=15.1 (1C; CH3), 45.9 (2C; CH3), 52.9 (2C;
C
Ph-C-CH2), 54.7 (2C; N-CH2), 56.0 (2C; CH), 58.1 (1C; C.), 62.8 (2C;
C-CH2), 126.9 (2C; CHPh/p), 128.0 (4C; CHPh/o), 129.1 (4C; CHPh/m), 136.5
(2C; CPh), 205.7 ppm (1C; CO); IR (KBr): n˜ =3419 (b), 3087 (m), 3059
(m), 3027 (m), 2967 (s), 2963 (s), 2843 (s), 2801 (s), 1717 (s), 1599 (m),
1496 (m), 1452 (s), 1376 (m), 760 (w), 748 cmꢀ1 (w); MS (FAB+)
(nibeol): m/z (%): 392.3 (100) [P1+H]+; elemental analysis calcd (%) for
C25H34N3ClO5: C 61.03, H 6.97, N 8.54, Cl 7.21; found: C 60.95, H 6.94, N
8.56, Cl 6.99.
1,5-Diphenyl-3-methyl-7-(1,4,6-trimethyl-1,4-diazacycloheptane-6-
yl)diazabicycloACTHNUTRGNEUNG
[3.3.1]nonane-9-one (L1): Methylamine (41%, 230 mg,
3.0 mmol), formaldehyde (536 mg, 6.6 mmol), and glacial acetic acid
(0.7 mL) were mixed at 08C in MeOH (6 mL). The ice bath was removed
and P1 (1.173 g, 3.0 mmol) was added. The reaction mixture was stirred
for 8 h at 658C. After complete removal of the solvents, the resulting
orange solid was suspended in diethyl ether. 2m KOH was added and the
two phases were stirred until no insoluble solids remained. The organic
phase was isolated and the aqueous phase was extracted twice with dieth-
yl ether. The combined organic phases were dried with Na2SO4. The sol-
vent was removed under reduced pressure, yielding L1 as a white foam,
which was used without further purification (1.161 g, 2.6 mmol, 87%).
Analytical data were obtained by crystallization from MeOH. 1H NMR
(200 MHz, [D1]CHCl3, 258C, TMS): d=1.1 (s, 3H; CH3), 2.2 (m, 6H; N-
CH3), 2.2 (d, 2J=13.4 Hz, 2H; C-N-CH2ax), 2.4 (m, 3H; N-CH3), 2.5 (m,
Experimental Section
4H; CH2-CH2), 2.8 (d, 2J
(H,H)=10.6 Hz, 2H; CH2ax-N-CH3/N-CH2ax-C), (d, 2J
2H; N-CH2ax-C/CH2ax-N-CH3), 3.4 (H,H)=10.6 Hz, 2H; CH2eq-N-
(d, 2J
CH3/N-CH2eq-C), 3.8 (d, 2J
(H,H)=11.0 Hz, 2H; N-CH2eq-C/CH2eq-N-
CH3), 7.2 ppm (m, 10H, CHPh); 13C NMR (50 MHz, [D1]CHCl3, 258C,
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
Materials and measurements: Chemicals and absolute solvents (ABCR,
Aldrich, Fluka) were used without further purification if not otherwise
noted. Styrene was distilled and stored under argon at ꢀ338C. NMR
spectra were recorded at 200.13 (1H) and 50.33 MHz (13C) on Bruker
A
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Chem. Eur. J. 2009, 15, 10880 – 10887
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10885

P. Comba et al.
TMS): d=18.9 (1C; C-CH3), 46.0 (1C; N-CH3), 49.2 (2C; CH3), 55.8
(1C; C-CH3), 59.2 (2C; C-CPh), 60.0 (2C; N-CH2-C), 62.6 (2C; N-CH2),
66.7 (2C; N-CH2-C/CH2-N-CH3), 67.3 (2C; N-CH2-C/CH2-N-CH3), 126.8
(2C; CHPh/p), 127.3 (4C; CHPh/o), 128.2 (4C; CPh/m), 143.5 (2C; CPh),
212.1 ppm (1C; CO); IR (KBr): n˜ =3022 (m), 2937 (s), 2802 (s), 1733 (s),
1599 (m), 1575 (m), 1495 (m), 1456 (s), 1320 (m), 761 (w), 693 cmꢀ1 (w);
MS (FAB+) (nibeol): m/z (%): 447.4 (100) [L1+H]+; elemental analysis
calcd (%) for C28H38N4O: C 75.30, H 8.58, N 12.54; found: C 75.00, H
8.53, N 12.52.
X-ray crystal structure determination: Crystal data and structure refine-
ment for [Cu(L1)
ACTHUNRTGENN(GU NCCH3)]AHCUTTGNNERN(GUN BF4)2·CH3CN: C32H44B2CuF8N6O; Mr =
¯
765.89; T=296(2) K; triclinic; space group P1; a=10.0530(9), b=
11.9894(11), c=15.4074(14) ꢂ; a=100.629(2), b=99.844(2), g=
93.785(2)8; V=1789.2(3) ꢂ3; Z=2; 1calcd =1.422 Mgmꢀ3; m=0.688 mmꢀ1
;
FACTHNUTRGNEUNG
(000)=794; crystal size 0.25ꢃ0.20ꢃ0.05 mm3; q range for data collec-
tion 1.37–25.098; 29116 reflections collected; 6340 independent reflec-
tions (Rint =0.0541); 4175 observed reflections with I>2s(I); index
ranges (independent set): h=ꢀ11 to 11, k=ꢀ14 to 14, l=0 to 18; com-
pleteness to q: 25.098, 99.7%; max/min transmission factors: 0.7452/
0.6765; data/restraints/parameters 6340/19/485; GOF on F2 =1.092;
R(F)=0.0524, wR(F2)=0.1272 (observed data, Fo >4s(Fo)); R(F)=
0.1016, wR(F2)=0.1606 (all data); difference density: root-mean-square/
max/min 0.082/0.541/ꢀ0.448 eꢂꢀ3. Intensity data were collected with a
Bruker AXS Smart 1000 CCD diffractometer (MoKa radiation, graphite
monochromator, l=0.71073 ꢂ). Data were corrected for air and detector
absorption, and Lorentz and polarization effects;[52] absorption by the
crystal was treated with a semiempirical multiscan method.[53,54] The
structure was solved by the heavy-atom method combined with structure
expansion by direct methods applied to difference structure factors[55,56]
and refined by full-matrix least squares methods based on F2 against all
unique reflections.[57, 58] All non-hydrogen atoms were given anisotropic
displacement parameters. Hydrogen atoms were input at calculated posi-
tions and refined with a riding model. The tetrafluoroborate anions were
A
ACHTUNGTRENNUNG(NCCH3)]ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
a
0.36 mmol) in CH3CN (3 mL). For the oxidation of CuI to CuII, a 5 mm
solution of AgBF4 in CH3CN was added to the reaction mixture
(0.5 mL). After being stirred at RT overnight, the precipitated silver was
removed by filtration. To the resulting blue solution diethyl ether was al-
lowed to diffuse, yielding blue crystals (101 mg, 0.14 mmol, 39%). IR
(KBr): n˜ =3450 (b), 3064 (w), 3029 (w), 2977 (m), 2935 (m), 2248 (s),
1742 (s), 1635 (m), 1479 (m), 1450 (m), 1048 (s), 766 (m), 699 cmꢀ1 (m);
UV/Vis (CH3CN): l (e)=903 nm (341 Lmolꢀ1 cmꢀ1); lmax (e)=627 nm
(684 Lmolꢀ1 cmꢀ1); MS (ESI+, CH3CN) (water due to measurement in
ESI machine): m/z (%): 568 (100) [CuII
(H3O)(L1)(NCCH3)]+, 510 (6)
ACHUTGTNRENNUG CAHTUNGTRENNUGN
[CuII(H)(L1)]+, 275 (62) [CuII(L1)(NCCH3)]2+, 255 (14) [CuII(L1)]2+; ele-
AHCTUNGTRENNUNG
mental analysis calcd (%) for C30H41N5B2CuF8O: C 50.18, H 5.79, N
10.97; found: C 50.09, H 5.82, N 10.93; EPR (CH3CN/toluene 1:1, T=
105 K): gjj =2.21 (Ajj =70 G), g? =2.08 (A? =15 G); CV: E1/2 (CH3CN,
T=258C, 100 mVsꢀ1): ꢀ377 (CuI/CuII), +1720 mV (CuII/CuIII).
ꢀ
found to be disordered. A split atom model was used and/or B F and
F···F distances were restrained to sensible values during refinement.
CCDC-725812 contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
A
ACHTUNGTRENNUNG(NCCH3)]ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Computational details: DFT calculations were performed by using Jaguar
6.5.[59] Two different basis sets were used, the LACVP basis set BS1[60–63]
and the LACV3P**++ basis set BS2.[60–63] Full structural optimizations
were carried out with BS1, and BS2 was then used for single-point calcu-
lations. For the location of transition states, either the usual transition-
state search with a starting geometry obtained with BS1 and B3LYP[64,65]
or the synchronous transit-guided quasi-Newton (STQN) method or a
combination of both were used. Frequency calculations were performed
on all optimized structures to verify their nature and to obtain zero-point
corrections to the energies. Unless otherwise stated, all DFT energies and
spin densities quoted in this manuscript are derived from B3LYP/
LACV3P**++ (BS2) calculations, with solvation (acetonitrile), zero-
point, and entropy corrections included.
solved in CH3CN (2 mL) and then added to a solution of L1 (150 mg,
0.33 mmol) in CH3CN (3 mL). After being stirred at RT overnight, dieth-
yl ether was added to the resulting blue solution and allowed to diffuse
to yield blue crystals (91.3 mg, 0.11 mmol, 33%). MS (ESI+, CH3CN):
ACTHNUTRGNEUNG
m/z (%): 658 (100) [CuII(L1)(O3SCF3)]+, 255 (9) [CuII(L1)]2+; elemental
analysis calcd (%) for C32H41N5CuF6O7S2: C 45.25, H 4.87, N 8.25; found:
C 45.21, H 4.93, N 8.30.
1-tert-Butyl-2(diacetoxyiodo)-3,4,5-trimethylbenzene:[50]
1-tert-Butyl-2-
iodo-3,4,5-trime-thylbenzene (1.33 g, 4.4 mmol) was dissolved in glacial
acetic acid (40 mL) at 408C. After 20 min, sodium perborate tetrahydrate
(6.82 g, 44.3 mmol) was added slowly under stirring. After 7 h of being
stirred at 408C, half of the solvent was removed in vacuo, and water
(50 mL) was added. The precipitate was collected and washed twice with
cyclohexane to yield (6-tert-butyl-2,3,4-trimethylphenyl)methylene diace-
tate (774 mg, 2.6 mmol, 60%) as a white solid. 1H NMR (200 MHz,
[D1]CHCl3, 258C, TMS): d=1.5 (s, 9H; C-(CH3)3), 1.9 (s, 6H; OCO-
CH3), 2.3 (s, 3H; CPh/C3;C4 -CH3), 2.3 (s, 3H; CPh/C3;C4-CH3), 2.8 (s, 3H;
CPh/C5-CH3), 7.4 ppm (s, 1H; CHPh).
N-Tosylimino(1-tert-butyl-3,4,5-trimethylbenzene)iodinane:[14] p-Toluene-
sulfonamide (416 mg, 2.43 mmol) and KOH (325 mg, 5.79 mmol) were
dissolved in MeOH (8 mL) and cooled to 08C. Under stirring, 1-tert-
butyl-2(diacetoxyiodo)-3,4,5-trimethylbenzene (682 mg, 2.32 mmol) was
added at 108C. After 1 h of being stirred at 08C and 3 h at RT, water
(8 mL) was added slowly at 08C. After one night in the fridge, a light
yellow precipitate was isolated by filtration and washed with a small
amount MeOH and cyclohexane, to yield N-tosylimino(1-tert-butyl-3,4,5-
trimethylbenzene)iodinane (764 mg, 1.62 mmol, 70%). Elemental analy-
sis calcd (%) for C20H26NIO2S: C 50.96, H 5.56, N 2.97; found: C 50.94,
H 5.52, N 2.96.
Acknowledgements
Generous financial support by the German Science Foundation (DFG) is
gratefully acknowledged.
[1] D. Tanner, Angew. Chem. 1994, 106, 625–646; Angew. Chem. Int.
Ed. Engl. 1994, 33, 599–619.
[2] H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. 2001, 113,
2056–2075; Angew. Chem. Int. Ed. 2001, 40, 2004–2021.
[3] R. S. Coleman, J.-S. Kong, J. Am. Chem. Soc. 1998, 120, 3538.
[4] J. B. Sweeney, Chem. Soc. Rev. 2002, 31, 247.
[5] D. A. Evans, M. M. Faul, M. T. Bilodeau, J. Org. Chem. 1991, 56,
6744.
Aziridination experiments: For the standard conditions used, see also
Table 2;[17,19] the catalyst and styrene or cis-b-methylstyrene were dis-
solved in the dry solvent (see Table 2 for the solvents and concentrations
used) before PhINTs was added. When the reaction mixture became
clear (the corresponding reaction times are given in Table 2), it was
passed through a short neutral aluminia column with ethyl acetate as
added solvent (25 mL). The solvents were evaporated under reduced
pressure. Anthrone (38.8 mg, 2.0 mmol) was added as internal standard
[6] Z. Lu, W. Zhang, W. D. Wulff, J. Am. Chem. Soc. 1994, 116, 2742.
[7] P. Mꢄller, C. Fruit, Chem. Rev. 2003, 103, 2905.
[8] D. A. Evans, M. M. Faul, M. T. Bilodeau, J. Am. Chem. Soc. 1994,
116, 2742.
[9] H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. 2001, 113,
2056; Angew. Chem. Int. Ed. 2001, 40, 2004.
[10] A. Yudin in Aziridines and Epoxides in Organic Synthesis Wiley-
VCH, Weinheim, 2006, p. 492.
[11] P. A. S. Lowden in Aziridines and Epoxides in Organic Synthesis
Wiley-VCH, Weinheim, 2006, p. 399.
1
to the resulting solid and the yield was determined by H NMR spectros-
copy[51] in CDCl3.
10886
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10880 – 10887

Copper-Catalyzed Aziridination
FULL PAPER
[12] S.-M. Au, J. S. Huang, W.-Y. Yu, W.-H. Fung, C.-M. Che, J. Am.
Chem. Soc. 1999, 121, 9120.
[13] J.-L. Liang, J.-S. Huang, X.-Q. Yu, N. Zhu, C.-M. Che, Chem. Eur. J.
2002, 8, 1563.
[14] Y. Yamada, T. Yamamoto, M. Okawara, Chem. Lett. 1975, 361.
[15] T. P. Albone, P. S. Aujnla, P. C. Taylor, S. Challenger, A. M. Derrick,
J. Org. Chem. 1998, 63, 9569.
[16] Z. Li, R. W. Quan, E. N. Jacobsen, J. Am. Chem. Soc. 1995, 117,
5889.
[17] J. A. Halfen, J. K. Hallman, J. A. Schultz, J. P. Emerson, Organome-
tallics 1999, 18, 5435.
[18] F. Mohr, S. A. Binfield, J. C. Fettinger, A. N. Vedernikow, J. Org.
Chem. 2005, 70, 4833.
[19] P. Comba, C. Lang, C. Lopez de Laorden, A. Muruganantham, G.
Rajaraman, H. Wadepohl, M. Zajaczkowski, Chem. Eur. J. 2008, 14,
5313.
[20] P. Brandt, M. J. Sçdergren, P. G. Andersson, P.-O. Norrby, J. Am.
Chem. Soc. 2000, 122, 8013.
[21] P. Comba, B. Nuber, A. Ramlow, J. Chem. Soc. Dalton Trans. 1997,
347.
[22] H. Bçrzel, P. Comba, C. Katsichtis, W. Kiefer, A. Lienke, V. Nagel,
H. Pritzkow, Chem. Eur. J. 1999, 5, 1716.
[23] H. Bçrzel, P. Comba, K. S. Hagen, C. Katsichtis, H. Pritzkow, Chem.
Eur. J. 2000, 6, 914.
[24] H. Bçrzel, P. Comba, H. Pritzkow, J. Chem. Soc. Chem. Commun.
2001, 97.
[38] J. Bautz, M. Bukowski, M. Kerscher, A. Stubna, P. Comba, A.
Lienke, E. Mꢄnck, L. Que, Jr., Angew. Chem. 2006, 118, 5810;
Angew. Chem. Int. Ed. 2006, 45, 5681; Angew. Chem. 118, 5810.
[39] J. Bautz, P. Comba, C. Lopez de Laorden, M. Menzel, G. Rajara-
man, Angew. Chem. 2007, 119, 8213; Angew. Chem. Int. Ed. 2007,
46, 8067.
[40] P. Comba, M. Maurer, P. Vadivelu, J. Phys. Chem. A 2008, 112,
13028.
[41] A. C. M. Appel, R. Hage, S. W. Russell, D. Tetard, Unilever, WO A
0185717, 2000; [Chem. Abstr. 2001, 135, 359431].
[42] R. A. Peralta, A. Neves, A. J. Borloluzzi, A. Casellato, A. dos Anjos,
A. Greatti, F. R. Xavier, B. Szpoganicz, Inorg. Chem. 2005, 44, 7690.
[43] R. Barbucci, M. J. M. Campell, Inorg. Chim. Acta 1975, 15, L15.
[44] K. M. Gillespie, C. J. Sanders, P. OꢅShaughnessy, I. Westmoreland,
C. P. Thickitt, P. Scott, J. Org. Chem. 2002, 67, 3450.
[45] M. M. Diaz-Requejo, P. J. Perez, M. Brookhart, J. L. Templeton, Or-
ganometallics 1997, 16, 4399.
[46] K. M. Gillespie, E. J. Crust, R. J. Deeth, P. Scott, J. Chem. Soc.
Chem. Commun. 2001, 785.
[47] D. Wang, G. R. Hanson, J. Magn. Reson. Ser. A 1995, 117, 209.
[48] D. Wang, G. R. Hanson, Appl. Magn. Reson. 1996, 11, 401.
[49] “Bispidonliganden und ihre Metallkomplexe”: P. Comba, C. Haaf,
M. Kerscher, A. Lienke, PCT/EP 2008/004506, 2008.
[50] A. McKillop, D. Kemp, Tetrahedron 1989, 45, 3299.
[51] S. Gandhi, A. Bisai, B. A. B. Prasad, V. K. Singh, J. Org. Chem.
2007, 72, 2133.
[25] H. Bçrzel, P. Comba, K. S. Hagen, M. Kerscher, H. Pritzkow, M.
Schatz, S. Schindler, O. Walter, Inorg. Chem. 2002, 41, 5440.
[26] P. Comba, C. Lopez de Laorden, H. Pritzkow, Helv. Chim. Acta
2005, 88, 647.
[27] K. Born, P. Comba, A. Daubinet, A. Fuchs, H. Wadepohl, J. Biol.
Inorg. Chem. 2006, 12, 36.
[28] M. Atanasov, C. Busche, P. Comba, F. El Hallak, B. Martin, G. Ra-
jaraman, J. van Slageren, H. Wadepohl, Inorg. Chem. 2008, 47, 8112.
[29] P. Comba, M. Merz, H. Pritzkow, Eur. J. Inorg. Chem. 2003, 1711.
[30] P. Comba, M. Kerscher, W. Schiek, Prog. Inorg. Chem. 2007, 55,
613.
[52] SAINT, area detector control and integration software, Bruker
AXS, Madison, 2007.
[53] R. H. Blessing, Acta Crystallogr. Sect. A 1995, 51, 33.
[54] SADABS-2004–2008, G. M. Sheldrick, Bruker AXS, Madison, 2004–
2008.
[55] P. T. Beurskens in Crystallographic Computing 3 (Eds.: G. M. Shel-
drick, C. Krꢄger, R. Goddard), Clarendon Press, Oxford, 1985,
p. 216.
[56] P. T. Beurskens, G. Beurskens, R. de Gelder, J. M. M. Smits, S.
Garcia-Granda, R. O. Gould DIRDIF-2008, Radboud University
Nijmegen, The Netherlands, 2008.
[31] P. Comba, A. Lienke, Inorg. Chem. 2001, 40, 5206.
[32] P. Comba, M. Kerscher, M. Merz, V. Mꢄller, H. Pritzkow, R. Reme-
nyi, W. Schiek, Y. Xiong, Chem. Eur. J. 2002, 8, 5750.
[33] P. Comba, A. Hauser, M. Kerscher, H. Pritzkow, Angew. Chem.
2003, 115, 4675; Angew. Chem. Int. Ed. 2003, 42, 4536.
[34] P. Comba, M. Kerscher, G. A. Lawrance, B. Martin, H. Wadepohl, S.
Wunderlich, Angew. Chem. 2008, 120, 4818; Angew. Chem. Int. Ed.
2008, 47, 4740.
[35] A. Bentz, P. Comba, R. J. Deeth, M. Kerscher, H. Pritzkow, B. Sei-
bold, H. Wadepohl, Inorg. Chem. 2008, 47, 9518.
[36] M. R. Bukowski, P. Comba, C. Limberg, M. Merz, L. Que, Jr., T.
Wistuba, Angew. Chem. 2004, 116, 1303; Angew. Chem. Int. Ed.
2004, 43, 1283.
[57] SHELXL-97, G. M. Sheldrick, University of Gçttingen, Gçttingen,
1997.
[58] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64.
[59] JAGUAR 5.5, Schrçdinger LLC, New York, 2005.
[60] T. H. Dunning, Jr., P. J. Hay in Modern Theoretical Chemistry, Vol. 3
(Ed.: H. F. Schꢀfer III), Plenum Press, New York, 1976, p. 1.
[61] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 270.
[62] R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54, 724.
[63] V. A. Rassolov, M. A. Ratner, J. A. Pople, P. C. Redfern, L. A. Cur-
tiss, J. Comput. Chem. 2001, 22, 976.
[64] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[65] A. D. Becke, J. Chem. Phys. 1993, 98, 5648.
[37] M. R. Bukowski, P. Comba, A. Lienke, C. Limberg, C. Lopez de
Laorden, R. Mas-Balleste, M. Merz, L. Que, Jr., Angew. Chem.
2006, 118, 3524; Angew. Chem. Int. Ed. 2006, 45, 3446.
Received: December 19, 2008
Revised: July 3, 2009
Published online: September 11, 2009
Chem. Eur. J. 2009, 15, 10880 – 10887
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10887