Ligand topology effects on olefin oxidations by bio-inspired [Fe II(N2Py2)] catalysts

Article abstract of DOI:10.1002/chem.200600453

Linear tetradentate N₂Py₂ ligands can coordinate to an octahedral Fe^II center in three possible topologies (cis-α, cis-β, and trans). While for the N,N'-bis(2-pyridylmethyl)-l,2- diaminoethane (bpmen) complex, only the cis-α topology has been observed, for N,N'-bis(2-pyridylmethyl)-1,2-diamino-cyclohexane (bpmcn) both cis-α and cis-β isomers have been reported. To date, no facile interconversion between cis-α and cis-β topologies has been observed for iron(II) complexes even at high temperatures. However, this work provides evidence for facile interconversion in solution of cis-α, cis-β, and trans topologies for [Fe(bpmpn)X₂] (bpmpn = N, N'-bis(2-pyridylmethyl)-1,3- diaminopropane; X = triflate, CH₃CN) complexes. As reported previously, the catalytic behavior of cis-α and cis-β isomers of [Fe(bpmcn)(OTf)₂] with respect to olefin oxidation depends dramatically on the geometry adopted by the iron complex. To establish a general pattern of the catalysis/ topology dependence, this work presents an extended comparison of the catalytic behavior for oxidation of olefins of a family of [Fe(N₂py₂)] complexes that present different topologies. ¹⁸O labeling experiments provide evidence for a complex mechanistic land-scape in which several pathways should be considered. Complexes with a trans topology catalyze only non-water-assisted epoxidation. In contrast, complexes with a cis-α topology, such as [Fe(bpmen)X₂] and [Fe(α-bpmcn)-(OTf)₂], can catalyze both epoxidation and cis-dihydroxylation through a water-assisted mechanism. Surprisingly, [Fe(bpmpn)X₂] and [Fe(β-bpmcn)-(OTf)₂] catalyze epoxidation via a water-assisted pathway and cis-dihydroxylation via a non-water-assisted mechanism, a result that requires two independent and distinct oxidants.

Full text of DOI:10.1002/chem.200600453

FULL PAPER
DOI: 10.1002/chem.200600453
II
Ligand Topology Effectson Olefin Oxidationsby Bio-In sp ired [Fe
Catalysts
A
H
R
U
G
2
2
[
a]
[a, b]
[a]
RubØn Mas-BallestØ, Miquel Costas,
Tieme van den Berg, and
[
a]
Lawrence Que, Jr.*
1
8
Abstract: Linear tetradentate N Py li-
1,3-diaminopropane;
X=triflate,
O labeling experiments provide evi-
2
2
gands can coordinate to an octahedral
Fe center in three possible topologies
CH CN) complexes. As reported previ-
ously, the catalytic behavior of cis-a
dence for a complex mechanistic land-
scape in which several pathways should
be considered. Complexes with a trans
topology catalyze only non-water-as-
sisted epoxidation. In contrast, com-
plexes with a cis-a topology, such as
3
II
(
cis-a, cis-b, and trans). While for the
and cis-b isomers of [Fe
A
H
R
U
G
A
H
R
U
N,N’-bis(2-pyridylmethyl)-1,2-diamino-
ethane (bpmen) complex, only the cis-
a topology has been observed, for
N,N’-bis(2-pyridylmethyl)-1,2-diamino-
cyclohexane (bpmcn) both cis-a and
cis-b isomers have been reported. To
date, no facile interconversion between
cis-a and cis-b topologies has been ob-
served for iron(II) complexes even at
high temperatures. However, this work
provides evidence for facile intercon-
version in solution of cis-a, cis-b, and
with respect to olefin oxidation de-
pends dramatically on the geometry
adopted by the iron complex. To estab-
lish a general pattern of the catalysis/
topology dependence, this work pres-
ents an extended comparison of the
catalytic behavior for oxidation of ole-
[Fe
A
H
R
U
G
A
H
R
U
G
2
AHCTREUNG
(OTf) ], can catalyze both epoxidation
2
and cis-dihydroxylation through
water-assisted mechanism. Surprisingly,
[Fe(bpmpn)X ] and [Fe(b-bpmcn)-
(OTf) ] catalyze epoxidation via
a
fins of a family of [Fe
A
H
R
U
G
A
H
R
U
G
ACHTREUNG
2
2
2
plexes that present different topologies.
A
H
R
U
G
a
2
water-assisted pathway and cis-dihy-
droxylation via non-water-assisted
a
Keywords: bio-inspired catalysis ·
mechanism, a result that requires two
independent and distinct oxidants.
hydrogen peroxide
·
iron
·
trans topologies for [Fe
A
H
R
U
G
N ligands · olefin oxidation
(
bpmpn=N,N’-bis(2-pyridylmethyl)-
Introduction
to a porphyrin cofactor and is attached to the polypeptide
chain with an axial cysteinate ligand. Much effort has been
[5]
Many iron enzymes activate dioxygen and catalyze the ster-
invested in developing biomimetic iron porphyrin catalysts
[1,2]
[6,7]
eospecific oxidation of CꢀH or C=C bonds.
Prominent
which participate
for substrate oxidation.
More recently, nonheme iron en-
[3,4]
among these are the cytochromes P450,
zymes have become better characterized and been found to
[1,8]
in many important metabolic transformations and thus have
been the subject of intense study. The cytochrome P450
active site has a mononuclear iron center that is coordinated
carry out similar transformations.
A particularly interest-
ing class of nonheme iron enzymes are the Rieske dioxyge-
nases that are responsible for the cis-dihydroxylation of
arene double bonds, the first step in the biodegradation of
[9,10,11,12]
arenes in the environment.
The cis-dihydroxylating
active site of the Rieske dioxygenases consists of a mononu-
clear iron center that is coordinated to two histidines and a
[
a] Dr. R. Mas-BallestØ, Dr. M. Costas, T. van den Berg,
Prof. Dr. L. Que, Jr.
Department of Chemistry and Center for Metals in Biocatalysis
University of Minnesota
[9]
bidentate aspartate, in a variation of the so-called 2-His-1-
[13]
carboxylate facial triad.
This coordination environment
leaves two adjacent sites available for exogenous ligand
binding and indeed allows dioxygen to bind in a side-on
2
07 Pleasant St. SE, Minneapolis, MN 55455 (USA)
Fax : (+1)612-624-7029
E-mail: que@chem.umn.edu
[10]
mode, which presumably activates it for the cis-dihydroxy-
lation reaction. We have embarked on an effort to develop
biomimetic catalysts inspired by the Rieske dioxygenases
and discovered a family of nonheme iron complexes of tet-
[
b] Dr. M. Costas
Current address: Departament de Quimica, Universitat de Girona,
1
7071, Girona (Spain)
Chem. Eur. J. 2006, 12, 7489 – 7500
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7489

radentate N4 ligands that use H O efficiently as oxidant for
family as well as from new structures reported herein. For
2
2
[14,15]
highly stereoselective oxidations of olefins.
Both olefin
bpmen, there is crystallographic evidence for only the cis-a
[
18,19,20,21]
epoxidation and cis-dihydroxylation reactions are catalyzed,
with the diol-to-epoxide ratio apparently dependent on
ligand topology and the spin state of the metal center.
Herein we report further studies that demonstrate the sensi-
tivity of the olefin oxidation reaction to the nature and top-
ology of the tetradentate N4 ligand, revealing a complex re-
action surface that appears to involve more than one type of
metal-based oxidant.
topology.
This is exemplified by the structure of
(CH CN) ](ClO ) (6), in which the N-
II
low-spin [Fe
A
H
R
U
G
A
H
R
U
G
ACHTREUNG
3
2
4 2
methyl groups are oriented anti with respect to each other.
In solution, the complex exhibits a low-spin/high-spin equili-
brium, but the C symmetry of the complex is maintained,
2
[21,22]
as indicated by NMR spectroscopy.
In contrast, the
closely related bpmcn ligand can form two isomeric iron
complexes with the cis-a (7) and the cis-b (8) topology, as il-
II
lustrated by the crystal structures of [Fe
A
H
R
U
G
A
C
H
T
R
E
U
N
G
[23]
II
(
7)
and the closely related [Fe (b-5-Me -bpmcn)(OTf) ]
(8’). In these structures, the N-methyl groups are oriented
anti with respect to each other in the cis-a topology and syn
A
C
H
T
R
E
U
N
G
2
2
[24]
Results and Discussion
Structural and spectroscopic characterization of iron(II)
complexescontaining linear N 2Py2 ligands: The bpmen
ligand serves as the prototype of a family of linear tetraden-
tate N4 ligands that can in principle wrap around an iron
with respect to each other in the cis-b topology. No inter-
II
change between cis-a and cis-b isomers of [Fe
A
H
R
U
G
(bpmcn)-
A
H
R
U
G
[25]
at 508C in CD CN. Such an interchange would entail epi-
3
[16,17]
center in three different topologies (Scheme 1).
Table 1
merization at a tertiary amine that would require prior
compares data from previously published structures of this
cleavage of the FeꢀN bond.
The remaining topology not
represented in the bpmen
family until very recently
[26,27]
is the trans topology, with all
four ligating atoms of the tet-
radentate N4 ligand approxi-
mately arrayed in the equatori-
al plane. To fill this gap, we in-
8
itially designed the L py2
[26]
ligand, in which two propy-
lene straps connecting the two
2
-pyridylmethylamine
units
serve to enforce the desired
topology. In the recently re-
II
ported structure of [Fe -
8
A
H
R
U
G
A
H
R
U
G
A
H
R
U
G
2
iron center in fact adopts a
square-pyramidal arrangement
with only one bound triflate
occupying the apical position,
leaving a free coordination site
trans to the triflate anion.
To complete the compari-
sons within the bpmen ligand
family, we investigated the
iron coordination chemistry of
bpmpn, a ligand with only one
propylene strap connecting the
two
2-pyridylmethylamine
units. The bpmpn ligand ap-
pears to be more flexible than
the other ligands in this family
and gives rise to complexes
with all three possible topolo-
gies (Table 1). Furthermore,
unlike for the other ligands, in-
terconversion between com-
Scheme 1. Linear N4 ligands and topologies observed for their iron(II) complexes.
7490
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7489 – 7500

Ligand Topology Effects on Olefin Oxidations
FULL PAPER
II
Table 1. Selected bond lengths [Š] and angles [8] for [Fe (L)X
n
] complexes.
to 3 (cis-b) when the reaction
Complex number
L
1
2
3
4
5’
6
7
8’
bpmcn 5-Me
bpmcn
CN 2OTf 2OTf
9
is carried out in CH CN
(water ligand in 3 presumably
derives from residual water
3
8
bpmpn bpmpn bpmpn bpmpn bpmpn
bpmen
2
-
L Py
2
X
2Cl
2OTf CH
3
CN 1Cl
OTf
2CH
3
1OTf
present in CH CN solvent).
3
H
cis-b
anti
2
O
acetone
trans
syn
Furthermore, 2 also converts
to 3 when 2 is dissolved in
[
a]
[b]
topology
cis-a
anti
trans
syn
tbp
syn
cis-a
anti
cis-a
anti
cis-b
syn
sp
–
relative disposition of
methyl groups
FeꢀNamine
CH CN. Lastly, both 2 and 3
3
2.365
2.218
2.237
2.172
2.206
–
2.204
2.209
2.189
2.209
2.211
2.221
2.174
2.179
2.108
–
2.231
2.229
2.201
2.213
–
2.055
2.053
2.015
2.002
1.941
2.220
2.237
2.151
2.162
–
2.239
2.203
2.273
2.191
–
2.216
2.176
2.137
2.164
–
revert to 1 by the addition of
two equivalents of tetrabutyl-
2.365
FeꢀNpyridine
FeꢀNacetonitrile
FeꢀO(X)
2.187
ammonium
chloride.
The
2
–
.187
product iron complexes ob-
tained by these procedures are
readily identified by their char-
1
–
.950
–
2.126
2.111
–
2.093
2.131
2.159
2.083
2.190
2.011
2.128
A
H
E
G
1
acteristic H NMR patterns,
2
which are identical to those as-
sociated with the complexes
obtained by direct synthesis.
AHCTREUNG
FeꢀCl
2.448
.448
162.40 111.45 99.34
94.69 177.78 81.14
–
–
2.277
96.05
–
–
–
–
–
2
N
py-Fe-Npy
112.26
176.53
175.49 102.02
95.18 86.28
109.73
Although
the
observed
X-Fe-X’
Ref.
–
173.34
89.67
–
[19]
[c]
[c]
[c]
[c]
[21]
[24]
[23]
[26]
changes in ligand topology
may be construed as a shift of
one pyridine from one site on
an octahedron to an adjacent
site, the associated change in
[
a] tbp=trigonal bipyramidal. [b] sp=square pyramidal. [c] This work.
plexes of different topologies is surprisingly facile (vide
infra).
the relative orientations of the N-methyl groups is not so
simple and requires a facile mechanism for the epimeriza-
tion of the N-methyl groups.
Previously published data on [Fe
A
H
R
U
G
that the complex has a six-coordinate structure wherein the
The observed facile interconversion among the three com-
plexes suggests that more than one isomer is present in solu-
tions of 2 and 3. Indeed, evidence for the coexistence and
interconversion of the three possible topological isomers of
[18]
tetradentate ligand adopts a cis-a conformation. The two
chloride ligands are cis to each other, the two pyridine rings
are coordinated trans to each other, and the relative disposi-
tion of the N-methyl groups is anti. The reaction of bpmpn
with iron(II) salts of more weakly coordinating anions,
1
9
2 is provided by variable-temperature F NMR studies in
CD Cl (Figure 2). At ambient temperature, there are two
2
2
namely [Fe
A
C
H
T
R
E
U
N
G
(CH CN)
A
H
R
U
G
A
H
R
N
(ClO ) ·6H O
very broad features at d=ꢀ4 and ꢀ46 ppm (FWHM ~3950
and 5080 Hz, respectively). As the temperature is decreased,
these features sharpen and resolve into six signals at ꢀ708C.
Set 2a consists of a pair of peaks at d=23.7 and ꢀ62.2 ppm
of comparably large intensity, while set 2b represents anoth-
er pair of peaks at d=20.9 and 12.7 ppm of smaller but com-
parable intensity. The remaining two peaks, also of low in-
tensity, appear at d=ꢀ5.9 and ꢀ74.6 ppm. Sets 2a and 2b
correspond to species containing two nonequivalent fluorine
nuclei, consistent with complexes having trans or cis-b topol-
3
2
2
2
2
4
2
2
in CH CN led to the isolation of complexes [Fe
AHCTREUNG
3
A
C
H
T
R
E
U
N
G
(OTf) ] (2) or [Fe
A
C
H
T
R
E
U
N
G
(bpmpn)
2
4 2
spectively. In the crystal structure of 2 (Figure 1) the bpmpn
ligand adopts a trans topology with the N-methyl groups syn
relative to each other and the triflate ligands trans to each
other. The two axial FeꢀO
identical, suggesting that the sole propylene strap does not
impose a distortion on the six-coordinate iron center. In the
crystal structure of 3, the bpmpn ligand adopts instead a cis-
b topology where the two remaining coordination sites are
occupied by CH CN and H O. Interestingly, unlike those of
1
ogies. The H NMR spectrum of this solution at ꢀ708C
shows that the major species in solution has only eight para-
magnetically shifted resonances indicative of a symmetric
species in which two halves of the molecule are related by
means of a symmetry operation, thereby assigning set 2a
signals to the trans isomer. Set 2b signals must then corre-
spond to a minor isomer with cis-b topology. Since both
trans- and cis-b topologies have syn N-methyl groups, this
isomerization entails a simple shift of a pyridine to an adja-
cent site.
3
II
[
Fe (b-5-Me -bpmcn)
2
are anti with respect to each other, an arrangement that was
previously observed for [(bpmpn)Fe (Cl)(m-O)Fe Cl ].
Thus, the ligand bpmpn can adopt the three possible topolo-
gies around an iron(II) center in octahedral complexes of
the general formula [FeN X ], with the observed geometries
III
[28]
3
strongly dependent on the nature of the X ligands.
Despite having different ligand topologies, bpmpn com-
plexes 1–3 can be readily interconverted in solution at room
temperature. For example, as shown in Scheme 2, treatment
of 1 (cis-a) with two equivalents of silver triflate gives rise
to 2 (trans) when the reaction is carried out in CH Cl and
The low intensity signal at d=ꢀ74.6 ppm is quite broad
and has a chemical shift that corresponds to free triflate.
This assignment is confirmed by the addition of 0.3 or 2.5
equivalents of tetrabutylammonium triflate to CD Cl solu-
2
2
2
2
Chem. Eur. J. 2006, 12, 7489 – 7500
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7491

L. Que Jr. et al.
Figure 1. ORTEPrepresentations of the structures of complexes 2, 3, 4, and 5’, showing 50% probability thermal ellipsoids. Hydrogen and fluorine
atoms have been omitted for clarity.
tions of 2, resulting in a proportionate increase in intensity.
the triflate being a non-coordinated counterion. The iron
The appearance of free triflate suggests the dissociation of a
center is essentially trigonal bipyramidal having a t value of
triflate from the six-coordinate [Fe
plexes to form a five-coordinate species [Fe
To ascertain the NMR properties of such a five-coordi-
nate species, we employed the synthetic strategy outlined in
Scheme 3 to obtain a five-coordinate bpmpn complex. Com-
A
H
R
U
G
A
H
R
U
G
0.85 with the N2
angle (167.48). In this structure the N-methyl groups of the
bpmpn ligand have a syn relative disposition. Its H NMR
A
H
R
U
A
H
R
U
2
+
A
H
R
U
G
1
spectrum in CD Cl at room temperature indicates an asym-
2
2
II
metric coordination environment around the high-spin Fe
center showing at least 15 different resonances in the range
from d=110 to ꢀ60 ppm. F NMR measurements at room
temperature show a broad band centered at d=ꢀ62 ppm
with a bandwidth of about 760 Hz, suggesting an equilibrium
between the five-coordinate species 4 and the six-coordinate
plex [Fe
with one equivalent of tetrabutylammonium chloride or
by the reaction of [Fe(bpmpn)Cl ] with one equivalent of
silver triflate. The synthesis of 4 was established by ESI-MS
major peak at m/z 375.1 corresponding to the [Fe-
A
C
H
T
R
E
U
N
G
(bpmpn)Cl]OTf (4) was obtained by the reaction of
1
9
2
AHCTREUNG
2
(
+
1
19
A
C
H
T
R
E
U
N
G
(bpmpn)Cl] ion), H and F NMR spectroscopy, and X-ray
diffraction. Its solid-state structure (Figure 1) shows a five-
coordinate high-spin iron(II) center (d
(FeꢀN) ꢁ2.2 Š)
bound to a tetradentate bpmpn ligand and a chloride, with
species [Fe
Reaction with one equivalent of AgSbF6 converts 4 to
[Fe(bpmpn)(OTf)]SbF (5), as indicated by ESI-MS (major
peak at m/z 489.2 corresponding to the [Fe
A
H
R
U
G
A
H
R
U
G
A
H
R
U
G
A
H
R
U
G
ACHTREUNG
6
+
A
H
R
U
G
A
C
H
T
R
E
U
N
G
7492
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7489 – 7500

Ligand Topology Effects on Olefin Oxidations
FULL PAPER
1
19
ion) and by H and F NMR spectroscopy. A sample of 5
recrystallized from acetone/diethyl ether affords crystals of
[
Fe
ecule of acetone is coordinated trans to the triflate ligand
Figure 1). The bpmpn ligand adopts a trans topology, as in
, and the N-methyl groups are disposed syn relative to each
A
H
R
U
(bpmpn)
A
H
R
U
(OTf)
A
H
R
U
G
(
2
1
other. The H NMR spectrum in CD Cl of 5 at room tem-
2
2
perature exhibits seven broad peaks in the range from d=
1
50 to ꢀ40 ppm, indicative of a dynamic exchange under-
19
gone by this complex. F NMR measurements confirm the
fluxional behavior of 5 in CD Cl at room temperature,
2
2
showing a very broad signal (from d=20 to ꢀ5 ppm) corre-
sponding to coordinated triflate. However, at ꢀ708C the
1
9
F NMR spectrum of 5 displays seven different broad sig-
nals in the range from d=10 to 110 ppm. These results indi-
cate that 5 undergoes a dynamic interconversion among dif-
ferent isomers that are presumably the precursors for cis-a,
cis-b, and trans structures, including isomers with a syn or
anti relative disposition of the methyl groups of the bpmpn
ligand. Additional possible isomers of 5 include diiron spe-
cies similar to that recently reported for a [(N Fe
A
H
R
U
G
A
C
H
T
R
E
U
N
G
4
[29]
OTf) ] complex.
2
The observation of a free triflate signal in the low-temper-
ature NMR spectrum of 2 requires that the pentacoordinate
Scheme 2. Synthesis and interconversion of complexes 1, 2, and 3.
+
[
Fe
A
H
R
U
G
A
T
U
(OTf)] be present in this solution. However,
1
9
none of the signals observed in the F NMR spectrum of 5
at low temperature appears in that of 2. Considering that
the signal assigned to free triflate in the spectrum of 2 is
broad and of low intensity, the signal pattern corresponding
+
to [Fe
A
H
R
U
G
A
H
R
U
G
1
10 ppm may be of too low intensity to be detected. This
leaves the remaining unassigned sharp singlet at d=
1
9
ꢀ
5.9 ppm in the low-temperature F NMR spectrum of 2,
which can then be reasonably assigned to the two chemically
equivalent triflates of the cis-a isomer.
Taken together, the NMR data reported above indicate
that complex 2 in CH Cl undergoes dynamic interconver-
2
2
1
9
2 2
Figure 2. Variable-temperature F NMR studies of a CH Cl solution of
sion among isomers with cis-a, cis-b, and trans topologies by
2
.
+
formation of pentacoordinate [Fe
A
H
R
U
G
A
H
R
U
G
mediate species, as represented in Scheme 4. These equili-
bria entail epimerization of the
methyl groups of the bound
bpmpn ligand, which requires
the breaking of an FeꢀN-
A
H
R
U
G
the process. The occurrence of
this dynamic process contrasts
the situation for closely related
[
Fe
A
H
R
U
G
complexes
2
that can adopt both cis-a and
cis-b topologies but do not in-
terconvert in solution even at
[25]
5
08C in CD CN.
Perhaps
even more curious is the fact
that [Fe(bpmen)(X) ] com-
3
AHCTREUNG
2
plexes have so far only been
Scheme 3. Synthetic strategy to obtain complex 5.
found to adopt the cis-a topol-
Chem. Eur. J. 2006, 12, 7489 – 7500
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7493

L. Que Jr. et al.
3
in CH CN and give support to the interpretation proposed
3
19
for the behavior observed in CH Cl by means of F NMR
2
2
spectroscopy. That is, not only the simple exchange of tri-
flate anions occurs in solution but also interconversion
II
among the three different possible isomers of [Fe
A
H
R
U
G
A
H
R
U
G
may be responsible for the significant differences between
the catalytic behavior of 2 and those observed for related
[
N FeX ] complexes.
4 2
Catalytic behavior asa function of topology : The different
possible topologies associated with the coordination of a
II
linear N4 ligand to an octahedral Fe center result in signifi-
cant variations in the chemical environments of the remain-
ing coordination sites (Scheme 1) where activation of the
H O oxidant must take place. In the trans topology the two
2
2
labile sites are trans to each other, limiting the potential co-
operation between the two sites. However, in the cis-a and
cis-b topologies, the two labile sites are cis with respect to
each other, allowing the possibility for a cooperative interac-
tion between these sites during catalysis. In the cis-a topolo-
gy the two labile sites are chemically equivalent, being both
trans to an aliphatic amine-nitrogen atom of the linear
N Py ligand, while in the cis-b topology the two labile sites
Scheme 4. Spontaneous interconversion of cis-a, cis-b, and trans isomers
of 2 in CH Cl
2
2
2
2
.
are inequivalent, being trans to an aliphatic amine and a
pyridine, respectively. Overall, the three topologies give rise
to distinct chemical properties of the cis-a, cis-b, and trans
ogy. Thus, the fluxional behavior observed for the [Fe-
(bpmpn)(OTf) ] complex is a unique feature of the bpmpn
ligand.
The fluxional behavior of [Fe
isomers of a [Fe
ly illustrated by the different catalytic behaviors of the cis-a
and cis-b isomers of the [Fe
we extend this comparison of catalytic behavior to other
members of the linear N Py family (Scheme 1), including a
complex (9) that is constrained to have a trans topology
(Table 2). Here we present a systematic study of six com-
plexes that includes the effects of temperature and added
water on their catalytic behavior (Figures 3 and 4, Tables 2
and 4) as well as O-labeling experiments (Table 3). The
effect of added water has been investigated because such
conditions are necessary for O-labeling experiments that
have proven so useful in providing mechanistic insight.
Commercially available H2 O is typically a 2% solution in
2
water, so the use of 10 equivalents of H2 O entails concom-
itant addition of approximately 1000 equivalents of H O.
The mechanistic framework within which the results of
these experiments are evaluated is shown in Scheme 5 and
mainly derives from detailed studies of [Fe
This framework postulates the activation of H O2 by an
A
H
R
U
G
A
C
H
T
R
E
U
N
G
ACHTREUNG
2
[
25]
A
H
R
U
G
A
H
R
U
G
A
H
R
U
G
A
H
R
U
G
2
indicates a greater lability of the bpmpn ligand in compari-
son to bpmen or bpmcn. Unlike the latter two, bpmpn
forms a six-membered [FeN C ] ring upon binding to the
2
2
2
3
iron center, which may result in a lower coordination affini-
ty. This increase of the labilty of the bpmpn ligand also ac-
counts for the easy replacement of the iron by silver when
1
8
+
an excess of Ag is added. This displacement is only ob-
1
8
served for [Fe
A
C
H
T
R
E
U
N
G
(bpmpn)
A
H
R
U
G
2
[
14,15]
bpmen complex.
1
8
Since the catalysis experiments to be described below are
usually carried out in acetonitrile, in which the coordinated
triflate anions are displaced by solvent molecules, we stud-
ied the behavior of 3 in this solvent. Owing to the fact that
the coordination sites occupied by triflate ions in 2 are occu-
pied by solvent molecules in 3, useful information can not
1
8
2
2
[14]
A
H
R
U
G
1
9
1
be obtained from F NMR spectroscopy. Thus, H NMR
spectroscopy has been the spectroscopic technique used to
obtain a qualitative description of the behavior of 3 in solu-
2
iron(III) center with two cis-labile sites via either a non-
A
H
R
U
G
water-assisted (nwa) pathway or a water-assisted (wa) path-
way. There is strong indirect evidence for the wa pathway in
tion. In CD CN, 3 exhibits at room temperature a paramag-
3
1
netically shifted H NMR spectrum with 11 signals in a
[Fe(tpa)] catalysis, as both epoxide and cis-diol incorporate
A
T
U
G
1
8
range from d=150 to ꢀ120 ppm. This pattern, which cannot
be attributed solely to a pure trans, cis-a, or cis-b isomer, be-
comes even more complicated at ꢀ408C where the corre-
sponding spectrum exhibits at least 22 different signals in
the range from d=170 to ꢀ150 ppm. These observations in-
dicate that a fast spontaneous isomerization also occurs for
label from H2 O into the products. It is proposed that water
V
incorporation occurs via an HOꢀFe =O oxidant that is
formed through a water-assisted cleavage of the OꢀO bond
III
of an H OꢀFe ꢀOOH intermediate. The nwa pathway must
2
be invoked to account for the lack of label incorporation
from H2 O into the products for some members of the [Fe-
1
8
7494
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7489 – 7500

Ligand Topology Effects on Olefin Oxidations
FULL PAPER
Table 2. Catalytic efficiency of nonheme iron catalysts towards olefin oxidation at room temperature.
ACHTREUNG
8
Ligand
Complex number
Topology
cyclam
10
trans
L Py
9
2
bpmen
bpmcn
bpmcn
bpmpn
2
III
2
6
7
8
7 + 8
below, may involve an Fe ꢀh -
[
a]
[b]
[b]
trans
cis-a
cis-a
cis-b
OOH species. However, unlike
III
1
2+
cyclooctene
diol + epoxide (TN)
[
Fe
A
H
R
U
G
A
H
R
U
G
,
this
[
[
[
c]
c]
c]
3.5
0.03
3.9
0.1
8.4
0.1
6.5
0.1
7.7
1.8
7.4
0.6
6.0
0.7
species has not been detected.
A
C
H
T
R
E
U
N
G
[diol]/
-octene
diol + epoxide (TN)
[diol]/[epoxide]
A
C
H
T
R
E
U
N
G
[epoxide]
1
1.7
0.2
1.0
0.3
5.4
0.3
6.4
0.2
4.7
2.1
4.1
0.6
4.1
3.8
trans Complexes 9 and 10:
8
A
C
H
T
R
E
U
N
G
A
C
H
T
R
E
U
N
G
Complex 9, whose L Py ligand
2
cis-2-heptene
diol + epoxide (TN)
is designed to maintain a trans
topology in solution, exhibits a
strong preference for epoxida-
tion with high stereoselectivity,
as does 10, a complex of the
macrocyclic ligand cyclam pre-
viously studied by Nam, Ho,
4.6
0.04
93
2.9
0.1
83
6.4
0.1
92
7.0
0.1
>99
>99
6.8
0.8
67
4.2
0.8
60
A
C
H
T
R
E
U
N
G
[diol]/
RC epoxide
RC diol
A
C
H
T
R
E
U
N
G
[epoxide]
[
d]
%
%
[
d]
>99
>99
79
85
>99
trans-2-heptene
diol + epoxide (TN)
[
c]
2.2
0.1
>99
>99
4.2
0.04
96
5.7
0.2
84
5.7
5.6
0.6
86
2.2
0.5
88
A
C
H
T
R
E
U
N
G
[diol]/
RC epoxide
RC diol
A
C
H
T
R
E
U
N
G
[epoxide]
0.05
>99
>99
[
d]
%
%
[30]
[
d]
>99
96
81
>99
and Valentine (see Figure 3).
Addition of 1000 equivalents
of water reduces the activity of
[
a] Data from reference [14]. [b] Data from reference [25]. [c] TN=Turnover number. [d] %RC=100”(cis ꢀ
trans)/(cis + trans).
these complexes by 40–50%,
suggesting that water may
compete with H O in coordinating the iron center
2
2
(
Table 4). Interestingly, there is a small amount of cis-diol
(
0.1–0.3 TON) that is obtained under normal conditions.
The cis-diol product becomes undetectable when water is
added or in experiments carried out at low temperature
(
ꢀ158C) with or without added water. It is not clear how
the very minor cis-dihydroxylation product can be formed
by a complex with trans-labile sites, as the nature of the re-
action would appear to require two cis-labile sites. Access to
a ligand arrangement that affords cis-labile sites has been
shown to be possible for the cyclam ligand,
isomer is very likely thermodynamically much less favored.
[31,32,33]
but this
18
Experiments in the presence of isotopically labeled H2
O
for the oxidation of cyclooctene catalyzed by 9 and 10 show
1
8
Table 3. O-labeling results for cyclooctene oxidation by 2, and 6–10 in
CH CN.
3
Figure 3. Product yields for the oxidation of cyclooctene with H
lyzed by nonheme iron complexes.
2 2
O cata-
1
8
18
%
O-
%
O-diol
epoxide
1
8
18
16
18
18
No
O
O– O
O– O
1
1
1
1
8
8
8
8
9
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
2
2
2
2
2
2
2
2
2
60
<1
88
<1
73
30
90
11
66
15
57
16
31
21
70
31
48
55
–
–
–
–
–
–
–
–
–
–
1
4
7
1
93
0
94
0
90
1
96
1
95
0
1
6
7
8
7
2
0
–
–
[
[
[
a]
18
41
23
3
11
2
97
2
96
3
96
0
97
1
58
73
89
88
4
3
4
4
7
3
4
2
4
1
8
b]
b]
18
1
8
18
1
1
1
1
1
1
1
1
1
8
8
8
8
8
8
8
8
8
+ 8
(room temp)
under air
(room temp)
under Ar
(ꢀ408C)
under air
2
2
96
4
Figure 4. Effect of water addition in the oxidation of cyclooctene at room
temperature with H catalyzed by complexes 7, 8, 7+8, and 2.
2
O
2
[a] Data from reference [34]. [b] Data from reference [25].
Chem. Eur. J. 2006, 12, 7489 – 7500
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7495

L. Que Jr. et al.
ments for 7 show that the diol
product
incorporates
one
oxygen atom from H O2 and
2
one oxygen atom from water
(
Table 3), thereby strongly im-
plicating wa mechanism
Scheme 5). The epoxide prod-
a
(
18
uct also incorporates some
O
label from water, suggesting
that the wa pathway also con-
tributes to the epoxidation of
cyclooctene. However the dra-
matic differences in the D/E
ratios that depend on the pres-
ence of water suggest that a
different active species is re-
sponsible for the chemistry in
the absence of added water.
Under the latter conditions, we
propose that 6 and 7 follow a
nwa pathway, in which the
main active species is proposed
2 2
Scheme 5. Proposed mechanisms for oxidation of olefins with H O catalyzed by non-heme iron complexes.
III
to be LFe ꢀOOH that affords
no incorporation of the labeled water into the epoxide prod-
epoxide as a main product.
1
8
uct (Table 3). In addition, when H2 O is used, the epoxide
Experiments at low temperature (ꢀ158C) were carried
out to evaluate the relative energetics of the different path-
ways. In the absence of water, decreasing the temperature
diminishes the turnover yield of 7 by a factor of 5. Adding
2
1
8
product shows only partial incorporation of the O label, in-
dicating some involvement of radical processes, particularly
for 9, that generate product with air oxidation. While these
results clearly exclude a wa mechanism for the epoxidation,
the nature of the active oxidizing species cannot be ascer-
tained since no intermediate has been trapped in the reac-
1000 equivalents of H O decreases this yield by another
2
factor of 5. This comparison suggests that the wa mechanism
has a higher activation barrier than the nwa pathway for cat-
alysts with a cis-a topology.
III
tions of either 9 or 10 with H O . The Fe ꢀOOH intermedi-
2
2
ate proposed in the nwa pathway is a plausible candidate for
III
9
, as the related Fe ꢀOOR species has been observed and
cis-b Complex 8: In contrast to its cis-a isomer, 8 with a
bpmcn ligand in a cis-b topology exhibits a marked prefer-
ence for olefin cis-dihydroxylation, with a D/E ratio of 1.8:1
characterized in its low-temperature reaction with
[26]
tBuOOH.
[25]
in the case of cyclooctene oxidation (see Figure 3). Like
cis-a Complexes 6 and 7: Complexes with cis-a topology (6
and 7) are efficient catalysts for highly stereoselective olefin
epoxidation with diol/epoxide (D/E) ratios of 0.1:1 (see
its cis-a isomer, the epoxide product shows some amount of
1
8
label incorporation from H2 O, consistent with the involve-
ment of a wa pathway. Considering the fact that the two cis
labile sites are not equivalent in the structure of 8, one may
[14,25]
Figure 3).
Under the same conditions, 6 and 7 generally
V
afford about twice as much epoxide product as 9 and 10,
suggesting that the cis-a topology provides a superior ligand
arrangement for activating H O for epoxidation. However,
expect that label scrambling in the HOꢀFe =O species will
not be complete, so the wa pathway could be the dominant
mechanism for epoxidation. However, unlike for its cis-a
isomer, the cis-diol product derives both oxygen atoms only
from H O , strongly implicating a nwa pathway for cis-dihy-
2
2
as previously reported, the addition of 1000 equivalents of
H O in cyclooctene oxidation experiments catalyzed by 6 in-
2
2
2
creases the D/E ratio from 0.1:1 to 0.8:1, so cis-dihydroxyla-
droxylation. Within the constraints of the original mechanis-
tic framework (Scheme 5), these observations suggest the
generation of two different active species derived from 8:
one obtained through a wa mechanism responsible for epox-
idation of cyclooctene and the other one obtained through a
nwa pathway that accounts for cis-dihydroxylation.
In contrast to what is observed for 6 and 7, the addition
of water does not significantly affect the D/E ratio or the
turnover yield for the reactions catalyzed by 8 at room tem-
perature (see Figure 4). However, decreasing the reaction
temperature to ꢀ158C nearly doubles the D/E ratio, indicat-
tion becomes a more important reaction under these condi-
[34]
tions. Similarly, the addition of 1000 equivalents of H O
2
to a cyclooctene oxidation experiment catalyzed by 7 in-
creases the D/E ratio from 0.1:1 to 0.7:1 (see Figure 4).
These observations suggest that different active species may
be formed for complexes with cis-a topology depending on
the presence or absence of water in the reaction media.
18
O-labeling experiments provide clues into the nature of
the active species for olefin oxidation catalyzed by 6 or 7.
1
8
Consistent with our previous work on 6, O-labeling experi-
7496
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7489 – 7500

Ligand Topology Effects on Olefin Oxidations
FULL PAPER
ing that the cis-dihydroxylation process is the kinetically
preferred pathway. More interestingly, whereas the turnover
yield drops by a factor of 5 upon going to ꢀ158C in the ab-
sence of added water, as observed for 7, it is only halved in
the presence of added water. So the addition of water in
fact increases the activity of 8 nearly threefold at low tem-
perature. These results are puzzling. The labeling experi-
ments showing a nwa pathway for cis-dihydroxylation led us
to the expectation that the diol yield should not be affected
by the presence of water. So, it is not clear how addition of
water can have a positive effect on the diol yield at lower
temperature.
at least, two reactive species in the mixture that constitutes
2: one derived from a nwa mechanism responsible for cis-di-
hydroxylation and the other from a wa pathway, responsible
for epoxidation.
However 2 behaves differently from 7+8 when the tem-
perature is lowered. At ꢀ158C, the D/E ratio decreases sig-
nificantly, particularly in the presence of added water. So,
unlike for 7+8, nwa dihydroxylation is less favored than wa
epoxidation. This conclusion is supported by labeling experi-
ments at ꢀ408C where the diol product is formed by a nwa
mechanism and the epoxide product, on the other hand, is
formed solely by a wa mechanism, as it is 50% labeled by
1
8
18
2
either H2 O and H O labeling data.
2
Equimolar mixture of 7 and 8: The different pathways and
products derived from the reaction of 7 and 8 with H O led
Mechanistic overview: Based on the observations presented
above, the mechanistic landscape for the oxidation of olefins
by nonheme iron catalysts appears to be not as simple as
first described on the basis of detailed studies on the [Fe-
2
2
us to investigate the oxidation of olefins with equimolar
amounts of 7 and 8 (designated 7+8) to assess their relative
reactivities. Since 7 and 8 individually give rise to clearly dif-
ferent D/E ratios in olefin oxidation (see Figure 3), the dis-
tribution of oxidized products in these experiments should
allow us to estimate the extent of cis-a or cis-b catalysis in
olefin oxidation. For the oxidation of cyclooctene and 1-
octene, the 7+8 combination affords D/E ratios of about
[14,15]
A
H
R
U
G
Ligand topology plays a large role, as
demonstrated by complexes 6–10. The complexes with trans
(9 and 10) and cis-a (6 and 7) topologies strongly favor ep-
oxidation via a nwa pathway. However for the cis-a cata-
lysts, addition of water significantly changes the product
ratio and gives rise to comparable amounts of epoxide and
0.6:1 in the absence of added water, values intermediate be-
1
8
tween those observed for the individual cis-a or cis-b iso-
mers, suggesting that the two isomers are comparably reac-
tive. However, when the oxidation of cyclooctene is cata-
lyzed by an equimolar mixture of 7 and 8 in the presence of
added water, the product distribution resembles that for 8
alone (see Figure 4). Accordingly, labeling experiments
cis-diol. O-labeling experiments attribute this change to a
shift towards a wa pathway for both epoxidation and cis-di-
hydroxylation. These results are easily accommodated by
the mechanistic framework presented in Scheme 5, in which
III
the addition of water favors the proposed H OꢀFe ꢀOOH
2
V
aquated derivative from which is generated the HOꢀFe =O
(
which necessarily require an excess of water) for 7+8 give
active species that effects both epoxidation and cis-dihy-
results very close to those observed for 8 alone. A similar
conclusion can be reached in experiments at ꢀ158C. Scruti-
ny of the data suggests that 7 is deactivated by the presence
of water and 8 becomes the dominant catalyst under these
conditions.
droxylation. Indeed DFT calculations on [Fe
A
H
R
U
G
have found the activation barriers for OꢀO bond cleavage
ꢀ
1
to differ by only 1 kcalmol for the nwa and the wa path-
[34]
ways, well within the margin of error ascribed to DFT cal-
culations.
The experimental results obtained for 2 and 8 however
complicate the original mechanistic framework. Complex 8
is the only one in our study with an established cis-b topolo-
gy, while 2 consists of a mixture of isomers. Both 2 and 8
catalyze cis-dihydroxylation reactions that are not affected
by the addition of water and found to occur solely via a nwa
pathway. On the other hand, the incorporation of labeled
water into epoxide products requires the involvement of a
wa pathway that exclusively forms epoxide. Thus two dis-
tinct oxidizing agents must be formed. The presumed HOꢀ
Complex 2: The chemistry of the bpmpn complex 2 provides
an interesting twist to this study of topological effects, since
2
differs from the other complexes in consisting of a mixture
of cis-a, cis-b, and trans isomers that can interconvert in sol-
ution, despite solid-state evidence for discrete topological
isomers. Interestingly, the D/E ratios obtained using 2 gener-
ally resemble those observed for 8, except for cyclooctene
where autooxidation contributes significantly to the yield of
epoxide (see Table 2). These observations suggest that the
cis-b isomer is the more active form of 2 in acetonitrile.
Consistent with that observed for 8, labeling experiments
show that both diol oxygen atoms come from H O (see
V
Fe =O oxidant formed in the wa pathways of 2 and 8 must
differ from corresponding species associated with the wa
pathways of 6, 7, and [Fe
ACHTREUNG
2
2
V
Table 3), which implicates a nwa mechanism for cis-dihy-
V
2+
droxylation. However, labeling experiments show that cyclo-
cis-diol products. For the [Fe (tpa)(O)(OH)] intermediate,
A
H
R
U
G
1
2
8
octene oxide derives its oxygen 31% from H O and 70%
DFT calculations show that epoxidation and cis-dihydroxy-
1
8
from H2 O when the experiments are carried out under
lation have comparable activation barriers with respective
2
1
8
[35]
argon (see Table 3). (In air, the incorporation of O in the
olefin attacks occurring at the oxo and hydroxo atoms.
1
8
epoxide product is 21% in the presence of H O and 31%
However, given ligand backbones and topologies that are
different from those of tpa, the activation barriers for epoxi-
dation and cis-dihydroxylation do not necessarily have to be
2
18
using H2 O , so the other half of the epoxide product arises
2
from a radical autoxidation pathway.) So there appear to be,
Chem. Eur. J. 2006, 12, 7489 – 7500
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7497

L. Que Jr. et al.
III
isoenergetic. In fact, experimental results suggest that for 2
and 8 epoxidation is more favored than cis-dihydroxylation
by the putative [Fe (L)(O)(OH)] oxidant generated from a
iron(II) precursor complexes that give rise to [Fe (L)-
2
+
A
H
R
U
G
V
wa pathway. As a consequence a species different to
are catalytically ineffective for olefin oxidation. This leaves
V
[Fe (L)(O)(OH)] should be proposed as responsible of the
[Fe
ACHTREUNG
nwa cis-dihydroxylation. Within the proposed mechanistic
catalyst for olefin oxidation like the [Fe
ACHTREUNG
2
2
III
[14]
framework it seems reasonable suggest that it is the [Fe -
discussed in this paper. Namꢁs conclusion does not contra-
dict Scheme 5, as we have already reported compelling la-
2
A
C
H
T
R
E
U
N
G
(h -OOH)] species that accounts for generation of cis-diol.
III
2+
The mechanistic picture we have described thus far pres-
ents a reaction surface that is quite sensitive to the specifics
of the reaction. Comparative studies at 25 and ꢀ158C show
that the relative activation barriers for the two possible re-
actions with and without added water are dependent on the
nature of the N Py ligand (Table 4). The nwa pathway is fa-
beling results that exclude [Fe
A
H
R
U
G
A
H
R
U
G
dant. Instead we have postulated that it serves as the precur-
V
sor to the actual [Fe (O)(OH)] oxidant in the wa pathway.
For the [Fe(N Py )] catalysts discussed in this paper, corre-
A
H
R
U
G
2 2
III
2+
sponding [Fe (L)(OOH)] intermediates have to date not
A
H
R
U
G
been observed. Perhaps these do not exist or they are
simply too reactive to be trapped. In the absence of evi-
2
2
III
2
2+
dence to the contrary, the putative [Fe (L)(h -OOH)] ox-
A
H
R
U
Table 4. Effect of water and temperature on Fe
octene oxidation in CH CN.
A
H
R
U
G
2
Py
2
) catalysts in cyclo-
idants remain the best candidates to account for the cis-di-
hydroxylation of olefins by the nwa pathway. This mechanis-
tic notion is supported by the crystallographic observation
of a side-on bound dioxygen ligand in the iron active site of
3
Catalyst
Temperature
added water
258C
1000
[equiv]
ꢀ158C
0
0
1000
A
H
R
U
G
A
H
R
U
G
[10]
naphthalene dioxygenase as well as DFT calculations on
9
diol + epoxide (TN)
3.9
0.1
3.5
0.03
6.5
0.1
7.7
1.8
7.4
0.6
6.0
0.7
1.9
~0
2.1
~0
8.2
0.7
8.2
1.4
8.6
1.7
4.8
1.1
1.7
~0
2.3
~0
1.7
0.2
1.6
2.9
1.7
1.3
2.2
0.2
1.5
~0
1.5
~0
0.2
0.3
4.5
3.0
3.5
2.2
2.5
0.08
the arene cis-dihydroxylation reaction catalyzed by this
A
C
H
T
R
E
U
N
G
[diol]/
diol + epoxide (TN)
[diol]/[epoxide]
diol + epoxide (TN)
[diol]/[epoxide]
diol + epoxide (TN)
[diol]/[epoxide]
diol + epoxide (TN)
[diol]/[epoxide]
diol + epoxide (TN)
[diol]/[epoxide]
A
C
H
T
R
E
U
N
G
[epoxide]
[37]
enzyme.
1
7
8
7
2
0
A
C
H
T
R
E
U
N
G
ACHTREUNG
A
C
H
T
R
E
U
N
G
ACHTREUNG
Experimental Section
A
C
H
T
R
E
U
N
G
ACHTREUNG
+8
A
C
H
T
R
E
U
N
G
A
C
H
T
R
E
U
N
G
Materialsand methods : All reagents were purchased from Aldrich and
used as received unless noted otherwise. CH CN and CH Cl solvents
(90% O-enriched, 2% solu-
3
2
2
1
8
18
A
C
H
T
R
E
U
N
G
A
C
H
T
R
E
U
N
G
were distilled over CaH
before use. H
O) was obtained from Cambridge Isotope Lab., and H
95% O-enriched) was received from ICON isotop. The cis- and trans-
-heptenes were purified by passing through silica gel immediately
before the reactions. The syntheses of all metal complexes were carried
2 2
O
2
1
6
18
tion in H
(
2
2
O
1
8
2
vored at lower temperature for 7 and 8, affording predomi-
nantly epoxide product for 7 and cis-diol product for 8. In a
competition between 7 and 8, the major product is the cis-
diol formed by a nwa pathway, suggesting that the latter cat-
alyst is more effective at lower temperature. On the other
hand, the major product for 2 at lower temperature is the
epoxide formed by a wa pathway, so just the insertion of an
additional methylene group in the bridge that connects the
two pyridylmethylamine halves of the tetradentate ligand is
sufficient to affect significantly the relative energetics of the
two reactions.
II
out in a glove box under a dry N
2
atmosphere. The syntheses of [Fe -
II
A
H
R
U
G
A
H
R
U
G
2
1
] and [Fe
A
T
E
N
(L
8
py
2
)
A
T
E
N
2
[21,26]
19
13
reported.
H NMR, F NMR, and C NMR spectra were recorded
1
13
on a Varian Unity 300 or 500 spectrometer. H and C chemical shifts
1
9
(ppm) were referenced to the residual solvent peaks. F chemical shifts
were referenced to an CFCl external reference. Electrospray ionization
3
mass spectral experiments were carried out on a Bruker BioTOF II mass
spectrometer using the following conditions: spray chamber voltage=
4
000 V; gas carrier temperature=2008C. Samples for ESI-MS analysis
were prepared in a CH CN/H O 1:1 mixture. Product analyses were per-
3
2
formed on a Perkin-Elmer Sigma 3 gas chromatograph (AT-1701 column,
0 m) with a flame ionization detector. GC mass spectral analyses were
performed on a HP5898 GC (DB-5 column, 60 m) with a Finnigan MAT
3
Recently Nam and co-workers have reported experimen-
9
5
5 mass detector or a HP6890 GC (H P- 5 column, 30 m) with an Agilent
973 mass detector. A 4% NH /CH mix was used as the ionization gas
III
tal evidence that argues against [Fe (OOH)] species as oxi-
A
H
R
U
G
3
4
dants directly responsible for oxygen-atom transfer reac-
for chemical ionization analyses.
[Fe(bpmpn)(OTf) ] (2): A solution of bpmpn (0.300 g, 1.05 mmol) ligand
in CH Cl (5 mL) was added to a suspension of [Fe(CH CN) (OTf)
0.457 g, 1.05 mmol) in CH Cl (5 mL). The reaction mixture was stirred
for 3 h at room temperature and the resulting solution filtered. Solid
product 2 was obtained by concentrating the filtrate to 5 mL and adding
water until precipitation. The precipitate was filtered off and washed
[36]
tions.
They investigated the reactivities of the three
(OOH)] (L=tpa, N4Py, Bn-tpen) intermediates,
A
H
R
U
G
A
H
E
N
2
III
2+
2
2
A
T
E
N
3
2
A
C
H
T
R
E
U
N
G
2
]
[Fe (L)
all of which can be trapped at low temperature and charac-
A
C
H
T
R
E
U
N
G
(
2
2
.
ACHTREUNG
1
terized spectroscopically to be low-spin (S= = ) iron(III)
A
H
R
U
G
2
complexes, and found that the decay of these pre-formed in-
termediates is not accelerated by the addition of substrates.
At first glance, these recent results may appear to under-
mine the mechanistic framework presented in Scheme 5 for
2 2
with ether. Recrystallization was carried out in CH Cl /diethyl ether at
room temperature. Yield: 345 mg (51%). This procedure led to the isola-
tion of diffraction-quality crystals. Both amorphous and crystalline prod-
ucts are hygroscopic. Elemental analysis indicate that, during manipula-
tion, crystals of 2 incorporate one molecule of water for each molecule of
III
ACHTREUNG
2
the [Fe
[
A
C
H
T
R
E
U
N
G
(N Py )] catalysts that postulate [Fe
2
2
V
24 6 4 6 2 2
iron complex. Elemental analysis calcd (%) for C19H F FeN O S ·H O
7498
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7489 – 7500

Ligand Topology Effects on Olefin Oxidations
FULL PAPER
(
9
2): C 34.75, H 3.94, N 8.53, S 9.75; found: C 34.88, H 3.77, N 8.42, S
.73.
Fe(bpmpn)
.05 mmol) was added to a solution of bpmpn (0.300 g, 1.05 mmol) in
CN (15 mL). The reaction mixture was stirred for 3 h at room tem-
collection at 173(2) K. A preliminary set of cell constants was calculated
from reflections harvested from three sets of 20 frames. These initial sets
of frames were oriented such that orthogonal wedges of reciprocal space
were surveyed. The data collection was carried out using MoKa radiation
(graphite monochromator) with a frame time of 45 s and a detector dis-
tance of 4.9 cm. A randomly oriented region of reciprocal space was sur-
veyed to the extent of one sphere and to a resolution of 0.84 Š. Four
major sections of frames were collected with 0.308 steps in w at four dif-
ferent f settings and a detector position of ꢀ288 in 2q. The intensity data
[
1
CH
A
C
H
T
R
E
U
N
G
A
C
H
T
R
E
U
N
G
(CH
3
CN)
A
C
H
T
R
E
U
N
G
(H
2
O)]
A
H
R
U
G
4
)
2
A
H
R
N
4 2 2
(3): Solid Fe(ClO ) ·6H O (0.382 g,
3
perature. The reaction mixture was then filtered to remove insoluble in-
organic salts, and then concentrated to 5 mL. Addition of diethyl ether
led to the separation of 3 as an oil, which converted to an hydroscopyc
brown solid after being subject to vacuum over 5 h. This solid was
washed several times with diethyl ether before recrystallization from di-
[
38]
were corrected for absorption and decay (SADABS). Final cell con-
stants were calculated from the actual data collection after integration
ethyl ether/CH
(H O)](ClO were obtained by this method. Yield: 260 mg (41%). Ele-
mental analysis calcd (%) for FeN ·H ([(bpmpn)Fe-
(CH CN)(H O)](ClO ·H O (3)): C 37.01, H 5.03, N 11.36; found: C
7.06, H 4.91, N 11.42.
Fe(bpmpn)Cl]OTf (4): Silver triflate (125.5 mg, 0.486 mmol) was added
to a solution of [Fe(bpmpn)Cl ] (0.200 g, 0.486 mmol) in CH Cl (10 mL).
3
CN. X-ray quality crystals of [(bpmpn)Fe
A
H
R
U
G
3
(CH CN)-
[
39]
A
C
H
T
R
E
U
N
G
2
A
C
H
T
R
E
U
N
G
4
)
2
(SAINT). Pertinent crystallographic data and experimental conditions
are summarized in Table 5. Structures were solved and refined using
C
19
H
29Cl
2
5
O
9
2
O
A
C
H
T
R
E
U
N
G
3
A
C
H
T
R
E
U
N
G
2
A
C
H
T
R
E
U
N
G
4
)
2
2
[40]
Bruker SHELXTL. The space groups were determined based on sys-
3
tematic absences and intensity statistics. A direct-methods solution was
calculated which provided most non-hydrogen atoms from the E-map.
Full-matrix least squares/difference Fourier cycles were performed which
located the remaining non-hydrogen atoms. All non-hydrogen atoms
[
ACHTREUNG
A
C
H
T
R
E
U
N
G
2
2
2
The suspension was stirred for 6 h at room temperature until complete
precipitation of AgCl, which was re-
moved by filtration. Product 4 was
obtained by adding diethyl ether to
the previously concentrated (to
Table 5. Crystallographic data and structural refinement details for compounds 2, 3, 4 and 5’.
2
3
4
5’
5
mL) filtrate. X-ray quality crystals
were obtained by crystallization from
CH Cl /diethyl ether. Yield: 210 mg
82%). Elemental analysis calcd (%)
formula
C
19
H
24
F
6
FeN
4
O
6
S
2
C
21
H
32Cl
2
FeN
6
O
9
C
18
H
24ClF
3
FeN
4
O
3
S
C
21 30 9 4 4
H F FeN O SSb
M
r
638.39
monoclinic
P2 /n
13.687(2)
11.893(2)
16.813(3)
90
111.066(2)
90
2553.9(7)
4
639.28
524.77
triclinic
P1
8.9605(11)
9.3177(11)
13.8028(17)
86.142
84.883(2)
88.599(2)
1145.0(2)
2
783.15
monoclinic
P2 /c
9.7261(13)
21.476(3)
14.875(2)
90
108.254
90
2950.9(7)
4
2
2
crystal system
space group
a [Š]
b [Š]
c [Š]
a [8]
b [8]
g [8]
V [Š ]
triclinic
(
¯
1
P1
1
C
18 3 4 3
H24ClF FeN O S (4): C 41.18, H
9.5552(11)
12.0468(13)
13.8233(15)
72.447(2)
79.665
74.174(2)
1451.4(3)
2
4
.57, N 10.67, S 6.10; found: C 41.23,
H 4.19, N 10.60, S 6.07.
Alternative synthesis of 4: Treatment
of 2 (0.300 g, 0.47 mmol) in CH
10 mL) with [N(Bu) ]Cl (0.130 g,
.47 mmol). The suspension was stir-
2 2
Cl
(
4
0
3
red for 4 h, and after that time fil-
tered and concentrated to 5 mL.
Z
T [K]
l [Š]
173(2)
0.71073
1.660
173(2)
0.71073
1.463
173(2)
0.71073
1.522
173(2)
0.71073
1.763
Compound
4 was obtained as a
yellow solid after addition of diethyl
ether. Yield: 141 mg (57%). Addition
of an excess of AgOTf (>2 equiva-
lents) to a solution of 1 affords a
white solid that can be identified as
ꢀ3
1
[gcm
]
ꢀ1
m [cm
]
8.41
7.60
9.17
15.69
independent reflections 4519
6604
4052
6043
goodness-of-fit
1.078
0.0529
0.1317
1.036
0.0358
0.0932
1.047
0.0493
0.1266
1.081
0.0478
0.0934
[a]
R1 [I>2(I)]
[
3
Ag
91, corresponding to [Ag
Fe(bpmpn)(OTf)]SbF (5): An equi-
molar amount of AgSbF (0.100 g,
.29 mmol) was added to a solution
of product 4 (0.150 g, 0.29 mmol) in CH
A
C
H
T
R
E
U
N
G
(bpmpn)]
A
C
H
T
R
E
U
N
G
(OTf) by ESI-MS (m/z
[b]
wR2 [I>2(I)]
+
A
H
R
U
G
2
2
2
2
2
1/2
[
a] R1=jjF
o
jꢀjF
c
jj/jF
o
j. [b] wR2={[w
A
H
E
N
(F
o
ꢀF
c
A
H
R
N
o
) ]/[w(F ) ]} .
[
A
C
H
T
R
E
U
N
G
A
C
H
T
R
E
U
N
G
6
6
0
2
2
Cl (10 mL). After the mixture
were refined with anisotropic displacement parameters. All hydrogen
atoms were placed in ideal positions and refined as riding atoms with rel-
ative isotropic displacement parameters.
had been stirred for 3 h, the AgCl formed was removed by filtration, and
the resulting solution concentrated to 5 mL. Complex 5 was precipitated
by adding diethyl ether to the concentrated solution. The solid obtained
was washed several times with diethyl ether. Yield: 188 mg (89%). Ele-
CCDC-284391 (5’), CCDC-284392 (4), CCDC-284393 (3), CCDC-284394
(
2) contain the supplementary crystallographic data for this paper. These
mental analysis calcd (%) for C18
.73, S 4.42; found: C 29.97, H 3.42, N 7.76, S 4.36. X-ray quality crystals
were obtained from recrystallization in CH Cl /acetone. The found struc-
ture corresponds to the formula 5·CH C(O)CH (5’).
Fe(cyclam)(CH CN) (ClO (10): Solid Fe(ClO
.0 mmol) was added to a solution of cyclam (0.200 g, 1.0 mmol) in
CH CN (10 mL), and reaction mixture was stirred for 6 h at room tem-
24 9 4 3
H F FeN O SSb (5): C 29.80, H 3.31, N
data can be obtained free of charge from The Cambridge Crystallograph-
ic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
7
2
2
Reaction conditionsfor catalytic oxidations : Catalytic oxidations were
3
3
[
14]
carried out as reported elsewhere. 0.3 mL of a 70 mm H
(diluted from a 35 or 50% H solution) in CH CN was delivered by sy-
ringe pump over 30 min at 258C in air to a CH CN solution (2.7 mL)
2 2
O solution
[
A
C
H
T
R
E
U
N
G
A
C
H
T
R
E
U
N
G
3
2
]
A
C
H
T
R
E
U
N
G
4
)
2
A
H
R
N
4 2 2
) ·6H O (0.363 g,
1
2
O
2
3
3
3
perature. After this time 5 appeared as a purple solid, which was isolated
containing iron catalyst and olefin substrate. The final concentrations
by filtration. The solid obtained was washed several times with small por-
were 0.7 mm mononuclear iron catalyst, 7.0 mm H O , and 0.70m olefin.
2
2
tions of CH
analysis calcd (%) for C14
3
CN and with diethyl ether. Yield: 333 mg (62%). Elemental
FeN (5): C 31.28, H 5.58, N 15.64;
For competition experiment the final concentration of a-[Fe
A
C
H
T
R
E
U
N
G
(bpmcn)-
H
30Cl
2
6
O
8
A
H
R
U
G
A
H
R
U
G
ACHTREUNG
2
2
found: C 31.37, H 5.71, N 15.55.
for another 5 min after syringe pump addition was complete. The result-
ing solutions were treated with acetic anhydride (1 mL) together with 1-
methylimidazole (0.1 mL) to esterify the diol products. Organic products
were extracted with CHCl . To the organic extract was added naphtha-
3
lene as an internal standard and the solution was subjected to GC analy-
sis. The products were identified by comparison of their GC retention
Crystallographic studies
The structures were solved at the X-ray Crystallographic Laboratory of
the Chemistry Department of the University of Minnesota. A crystal of
complex 2, 3, 4, or 5’ was placed onto the tip of a 0.1-mm diameter glass
capillary and mounted on a CCD area detector diffractometer for a data
Chem. Eur. J. 2006, 12, 7489 – 7500
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7499

L. Que Jr. et al.
times. All experiments were run at least in duplicate, the reported data
being the average of these reactions.
[17] U. Knof, A. von Zelewsky, Angew. Chem. 1999, 111, 312; Angew.
Chem. Int. Ed. 1999, 38, 302.
[
18] S. Taktak, S. V. Kryatov, E. V. Rybak-Akimova, Inorg. Chem. 2004,
43, 7196.
19] N. Raffard, V. Balland, J. Simaan, S. LØtard, M. Nierlich, K. Miki, F.
Banse, E. AnxolabØh›re-Mallart, J.-J. Girerd, C. R. Chim. 2002, 5,
99.
Isotope labeling studies: Variations from the conditions described above
1
8
18
are the following: In experiments with H
2
O, 42 mL of H
2
O (70 mm,
[
1
000 equivalents) was added to the catalyst solution prior to the injection
1
8
18
of H
from a 2% H
2
O
2
. In experiments with H
2
O
2
, 7.0 mm H
2
O
2
(diluted by CH
3
CN
1
8
2
O
2
/H
2
O solution) was used instead of H O
2 2
. The organic
[
20] P. Mialane, L. Tchertanov, F. Banse, J. Sainton, J.-J. Girerd, Inorg.
Chem. 2000, 39, 2440.
products were subjected to GC/CI-MS analyses.
[
[
21] K. Chen, L. Que Jr., Chem. Commun. 1999, 1375.
22] K. P. Bryliakov, E. A. Duban, E. P. Talsi,Eur. J. Inorg. Chem. 2005,
1, 72.
Acknowledgements
[
23] M. Costas, J.-U. Rohde, A. Stubna, R. Y. N. Ho, L. Quaroni, E.
Münck, L. Que Jr., J. Am. Chem. Soc. 2001, 123, 12931.
This work was supported by a grant from the U.S. Department of Energy
FFG02–03ER15455 to L.Q.). R.M. thanks the Ministerio de Educación
(
[24] M. Costas, A. K. Tipton, K. Chen, D.-H. Jo, L. Que Jr., J. Am.
Chem. Soc. 2001, 123, 6722.
[25] M. Costas, L. Que Jr., Angew. Chem. 2002, 114, 2283; Angew.
Chem. Int. Ed. 2002, 41, 2179.
[26] M. R. Bukowski , H. L. Halfen, T. A. van den Berg , J. A. Halfen, L.
Que Jr., Angew. Chem. 2005, 117, 590; Angew. Chem. Int. Ed. 2005,
y Ciencia of Spain for postdoctoral support. The authors would like to
acknowledge the efforts of Benjamin E. Kucera and Dr. Victor G.
Young, Jr. of the X-Ray Crystallographic Laboratory, Department of
Chemistry, University of Minnesota in solving the crystal structures re-
ported.
4
4, 584.
27] G. J. P. Britovsek, J. England, A. J. P. White, Dalton Trans. 2006,
399.
28] N. Raffard, F. Banse, K. Miki, M. Nierlich, J.-J. Girerd, Eur. J. Inorg.
Chem. 2002, 1941.
[
[
[
1
[
[
1] M. Costas, M. P. Mehn, M. P. Jensen, L. Que Jr.,Chem. Rev. 2004,
04, 939.
2] D. T. Gibson, V. Subramanian, in Microbial Degradation of Organic
Compounds (Ed.: D. T. Gibson), Marcel Dekker, New York, 1984,
pp. 181.
3] P P. R. Ortiz de Montellano, Cytochrome P-450 Structure, Mecha-
nism and Biochemistry, 2nd ed., Plenum, New York, 1995.
4] W. Nam, “Cytochrome P-450” in Comprehensive Coordination
Chemistry II, Vol. 8, (Volume Eds.: L. Que, Jr., W. B. Tolman, Series
Eds.: J. McCleverty, T. J. Meyer), Elsevier, Amsterdam, 2004, p. 281.
5] I. G. Denisov, T. M. Makris, S. G. Sligar, I. Schlichting, Chem. Rev.
1
29] M. P. Jensen, S. J. Lange, M. P. Mehn, E. L. Que, L. Que Jr.,J. Am.
Chem. Soc. 2003, 125, 2113.
[
[
30] W. Nam, R. Ho, J. S. Valentine, J. Am. Chem. Soc. 1991, 113, 7052.
31] R. Guilard, O. Siri, A. Tabard, G. Broeker, P. Richard, D. J. Nurco,
K. J. Smith, J. Chem. Soc. Dalton Trans. 1997, 3459.
[
[
[
[
[
[
32] K. Meyer, E. Bill, B. Mienert, T. Weyhermuller, K. Weighardt, J.
Am. Chem. Soc. 1999, 121, 4859.
33] H. Chun, E. Bill, E. Bothe, T. Weyhermuller, K. Weighardt, Inorg.
Chem. 2002, 41, 5091.
34] D. QuiÇonero, K. Morokuma, D. G. Musaev, R. Mas-BallestØ, L.
Que Jr., J. Am. Chem. Soc. 2005, 127, 6548.
35] A. Bassan, M. R. A. Blomberg, P. E. M. Siegbahn, L. Que Jr.,
Angew. Chem. 2005, 117, 2999; Angew. Chem. Int. Ed. 2005, 44,
[
2
005, 105, 2253.
[
[
[
[
6] F. A. Walker, Chem. Rev. 2004, 104, 589.
7] H. Fujii, Coord. Chem. Rev. 2002, 226, 51.
8] E. Y. Tshuva, S. J. Lippard, Chem. Rev. 2004, 104, 987–1012.
9] B. Kauppi, K. Lee, E. Carredano, R. E. Parales, D. T. Gibson, H.
Eklund, S. Ramaswamy, Structure 1998, 6, 571.
10] A. Karlsson, J. V. Parales, R. E. Parales, D. T. Gibson, H. Eklund, S.
Ramaswamy, Science 2003, 299, 1039.
11] M. D. Wolfe, J. V. Parales, D. T. Gibson, J. D. Lipscomb, J. Biol.
Chem. 2001, 276, 1945.
12] D. T. Gibson, R. E. Parales, Curr. Opin. Biotechnol. 2000, 11, 236.
13] K. D. Koehntop, J. P. Emerson, L. Que Jr., J. Biol. Inorg. Chem.
2
939.
36] M. J. Park, J. Lee, Y. Suh, J. Kim W. Nam, J. Am. Chem. Soc. 2006,
28, 2630.
37] A. Bassan, M. R. A. Blomberg, P. E. M. Siegbahn, J. Biol. Inorg.
Chem. 2004, 9, 439.
[
[
[
[
1
[
[
38] R. Blessing, Acta Crystallogr. Sect. A 1995, 51, 33–38.
39] SAINT V6.2, Bruker Analytical X-Ray Systems, Madison, WI
[
[
(
2001).
40] SHELXTL V6.10, Bruker Analytical X-Ray Systems, Madison, WI
2000).
2
005, 10, 87.
[
[
[
[
14] K. Chen, M. Costas, J. Kim, A. K. Tipton, L. Que Jr., J. Am. Chem.
Soc. 2002, 124, 3026.
15] K. Chen, M. Costas, L. Que Jr., J. Chem. Soc. Dalton Trans. 2002,
72.
16] J. R. Aldrich-Wright, R. S. Vagg, P. A. Williams, Coord. Chem. Rev.
997, 166, 361.
(
6
Received: April 1, 2006
Published online: July 26, 2006
1
7500
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7489 – 7500