Non-heme iron complexes for stereoselective oxidation: Tuning of the selectivity in dihydroxylation using different solvents

Article abstract of DOI:10.1002/ejic.200300667

A new class of functional models for non-heme iron-based dioxygenases, including [(N3Py-Me)Fe(CH₃CN)₂](ClO₄) ₂ and [(N3Py-Bn)Fe(CH₃CN)₂](ClO ₄)₂ {N3Py-Me = [di(2-pyridyl)methyl]methyl(2-pyridyl) methylamine; N3Py-Bn = [di(2-pyridyl)methyl]benzyl(2-pyridyl)methylamine}, is presented here. NMR, UV and X-ray analyses revealed that six-coordinate low-spin Fe^II complexes with the pyridine N-atoms and the tertiary amine functionality of the ligand bound to Fe are formed. The two remaining coordination sites located cis to each other are occupied by labile CH ₃CN groups that are easily exchanged by other ligands. We demonstrate that the reactivity and stereoselectivity of the complexes investigated depend on the choice of the solvent. The complexes have been examined as catalysts for the oxidation of both alkanes and olefins in CH₃CN. In this solvent alkanes are oxidized to alcohols and ketones and olefins to the corresponding cis-epoxides and cis-diols. In acetone as solvent a different reactivity pattern was found, with, as the most striking example, the trans-dihydroxylation of cis-olefins. ¹⁸O-labeling studies in CH₃CN establish incorporation of ¹⁸O from H₂¹⁸O₂ and H₂¹⁸O in both the epoxide and the diol implicating an HO-Fe^V=¹⁸O active intermediate originating from an H ₂¹⁸O-Fe^IIIOOH species. These results are in full agreement with mechanistic schemes derived for other dioxygenase model systems. Based on labeling studies in acetone an additional oxidation mechanism is proposed for this solvent, in which the solvent acetone is involved. This is the first example of a catalyst that can give cis- or trans-dihydroxylation products, just by changing the solvent. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejic.200300667

FULL PAPER
Non-Heme Iron Complexes for Stereoselective Oxidation: Tuning of the
Selectivity in Dihydroxylation Using Different Solvents
Marten Klopstra,[a] Gerard Roelfes,[a] Ronald Hage,[a] Richard M. Kellogg,[a] and
Ben L. Feringa*[a]
Keywords: Dihydroxylation / Enzyme models / Epoxidation / Iron / N ligands / Oxidation
A new class of functional models for non-heme iron-based
dioxygenases, including [(N3Py-Me)Fe(CH3CN)2](ClO4)2
and [(N3Py-Bn)Fe(CH3CN)2](ClO4)2 {N3Py-Me = [di(2-pyrid-
yl)methyl]methyl(2-pyridyl)methylamine; N3Py-Bn = [di(2-
pyridyl)methyl]benzyl(2-pyridyl)methylamine}, is presented
epoxides and cis-diols. In acetone as solvent a different react-
ivity pattern was found, with, as the most striking example,
the trans-dihydroxylation of cis-olefins. 18O-labeling studies
in CH3CN establish incorporation of 18O from H218O2 and
H218O in both the epoxide and the diol implicating an HO-
here. NMR, UV and X-ray analyses revealed that six-coord- FeV=18O active intermediate originating from an H218O-
inate low-spin FeII complexes with the pyridine N-atoms and
the tertiary amine functionality of the ligand bound to Fe are
FeIIIOOH species. These results are in full agreement with
mechanistic schemes derived for other dioxygenase model
formed. The two remaining coordination sites located cis to systems. Based on labeling studies in acetone an additional
each other are occupied by labile CH3CN groups that are
easily exchanged by other ligands. We demonstrate that the
reactivity and stereoselectivity of the complexes investigated
depend on the choice of the solvent. The complexes have
been examined as catalysts for the oxidation of both alkanes
and olefins in CH3CN. In this solvent alkanes are oxidized to
alcohols and ketones and olefins to the corresponding cis-
oxidation mechanism is proposed for this solvent, in which
the solvent acetone is involved. This is the first example of a
catalyst that can give cis- or trans-dihydroxylation products,
just by changing the solvent.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
The study of synthetic complexes that model the active
site of these enzymes has been an important tool in the
study of their functions. Some of these model complexes
have proven to be very promising catalysts for the selective
oxidation of organic substrates using H2O2 as the terminal
The catalytic oxidation of hydrocarbons using green oxi-
dants like hydrogen peroxide or dioxygen is a topic of con-
siderable interest.[1] Even though significant progress has
been made, in particular with regard to epoxidation of al-
kenes using H2O2,[2Ϫ5] many challenges still remain. Among
others, the hydroxylation of alkanes and the dihydroxyl-
ation of alkenes are highly interesting reactions from a syn-
thetic point of view.
In nature an impressive variety of enzymes capable of
performing these reactions with high selectivity under mild
conditions exists. An important subset of this class of en-
zymes are the non-heme iron oxygenases, like, for example,
methane monooxygenase,[6] which selectively oxidizes meth-
ane to methanol, the Rieske dioxygenases,[7,8] which are
capable of stereospecifically dihydroxylating arenes, and
iron-bleomycin, a metallo-glycopeptide that is capable of
oxidative DNA cleavage,[9] but which can also oxidize a
wide variety of organic substrates.[10]
oxidant.
Stereoselective
hydroxylation[11Ϫ14]
epoxi-
dation[4,14Ϫ16] and cis-dihydroxylation,[17Ϫ20] a reaction pre-
viously only available using osmium-,[21] ruthenium-[22] and
manganese-based[23] catalysts, have been reported. Recently,
even the first examples of enantioselective dihydroxylation
have appeared in the literature.[24]
Recent studies have shed some light on the mechanism
of these oxidation reactions. In the case of [Fe(T-
PA)(CH3CN)2](ClO4)2, a representative member of this
class of catalysts, a strong and compelling case has been
built in favor of the involvement of a formally HO-FeVO
species that is formed by heterolysis of the OϪO bond of
a low spin FeIIIOOH intermediate.[13,20,25] This HO-FeVO
species is postulated to be the active species responsible for
the formation of both the epoxide and the cis-diol. The lat-
ter product is thought to be formed via a cyclic intermediate
similar to reactions with KMnO4.[15]
[a]
Department of Organic and Molecular Inorganic Chemistry,
Stratingh Institute, University of Groningen,
Nijenborgh 4, 9747 AG Groningen, Netherlands
Fax: (internat.) ϩ 31-50-363-4296
We have previously reported on the [Fe(N4Py)(CH3CN)]-
(ClO4)2 complex, which contains a pentadentate N ligand
and reacts with H2O2 to form a well-defined and well-
E-mail: B.L.Feringa@chem.rug.nl
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
846
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200300667
Eur. J. Inorg. Chem. 2004, 846Ϫ856

Non-Heme Iron Complexes for Stereoselective Oxidation
FULL PAPER
characterized FeIIIOOH intermediate.[26Ϫ28] This intermedi-
ate was demonstrated to have an unusually weak OϪO
bond,[29,30] which is thought to be the reason for its high
activity as an oxidation catalyst. Reactivity studies using a
broad spectrum of mechanistic probes revealed that radical-
type oxidizing species are involved and, consequently, a
mechanism involving homolysis of the FeIIIOOH species to
give FeIVO and ·OH was proposed.[31] This notion was later
supported by calculations, which indeed showed homolysis
to be the favored reaction pathway.[32] A strong effect of the
solvent was observed. In acetone higher kinetic isotope ef-
fect (KIE) values of cyclohexane oxidation and higher C3/
C2 ratios in adamantane oxidation were found compared
to acetonitrile. This was attributed to the ability of acetone
to act as an ·OH trap.
Following the TPA precedent we decided to redesign the
N4Py ligand into a tetradentate ligand in order to obtain
iron complexes capable of stereoselective oxidation. In this
paper we present a new class of iron complexes with tetra-
dentate ligands based on N4Py, in which one of the picolyl
groups is replaced with a non-coordinating moiety like a
methyl or a benzyl group (Figure 1). The properties of the
corresponding iron complexes and their application as cata-
lysts in oxidation reactions with H2O2 will be discussed,
with particular emphasis on the effect of the solvent used,
i.e. acetonitrile or acetone.
Scheme 1. Synthesis of ligands 1 and 2 and the corresponding
iron() complexes 4 and 5
Figure 1. Ligands used for stereoselective epoxidation and cis-dihy-
droxylation of alkenes
Results and Discussion
Figure 2. ORTEP plot of the cation of 4
Synthesis and Properties of Iron Complexes
˚
˚
Ligands 1 and 2 were prepared from a common precur- tances vary from 1.909(8) A for Fe1ϪN5 to 1.965(9) A for
sor, N3Py (3).[33] N3Py-Me was obtained in high yield by Fe1ϪN3. These distances are similar to those found for
reductive amination with formaldehyde and NaBH(OAc)3 other low-spin iron() complexes like [(N4Py)-
(Scheme 1).[34] N3Py-Bn was prepared by alkylation of Fe(CH3CN)](ClO4)2,[19]
and
[(TPA)Fe(CH3CN)2]-
N3Py with benzyl chloride in the presence of K2CO3.
Complexation of the ligands and
(ClO4)2.[35]
1
1
2
with
The H NMR spectra in CD3CN exhibit signals between
Fe(ClO4)2·6H2O in methanol/acetonitrile resulted in the δ ϭ 0 and 15 ppm, typical of a low-spin iron() complex.
formation of red crystals, characterized as [(N3Py-R)- However, in contrast to the spectra of other low-spin
Fe(CH3CN)2](ClO4)2, based on 1H NMR and ES/MS iron() complexes like [(TPA)Fe(CH3CN)2](ClO4)2 and
analysis and confirmed by X-ray analysis for this complex [(N4Py)Fe(CH3CN)2](ClO4)2,[28,35] which show very sharp
with an N-methyl substituent (Figure 2).
signals, significant line broadening was observed for some
The six-coordinate FeII center has a distorted octahedral of the signals of 4 and 5. An example is the spectrum of 4
geometry, with the iron bound to the three pyridine N shown in Figure 3A. The line broadening is probably due
atoms, the tertiary amine functionality of the ligand and to an equilibrium between the low-spin FeII bis-acetonitrile
two acetonitrile molecules, originating from the solvent complex [(N3Py-R)Fe(CH3CN)2]2ϩ and the paramagnetic
used for crystallization. The open coordination sites, where high-spin FeII mono-acetonitrile complex [(N3Py-R)-
acetonitrile molecules are bound, are located cis with re- Fe(CH3CN)]2ϩ, i.e. a reversible dissociation of one of the
spect to each other, as was expected. The FeϪNligand dis- labile CH3CN ligands. This was confirmed by the 1H NMR
Eur. J. Inorg. Chem. 2004, 846Ϫ856
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
847

M. Klopstra, G. Roelfes, R. Hage, R. M. Kellogg, B. L. Feringa
FULL PAPER
spectrum in [D6]acetone, in which dissociation of aceto-
nitrile is irreversible and results in signals in the δ ϭ
0Ϫ180 ppm range, characteristic for a high-spin FeII com-
plex (Figure 3B).[28] A similar solvent effect was found in
the UV/Vis spectra of 4 in CH3CN and acetone. In CH3CN
4 exhibits a UV/Vis spectrum with absorption peaks at 352
(ε ϭ 5.3 ϫ 103

sorption peak at 331 nm (ε ϭ 4.2 ϫ 103 Ϫ1cmϪ1) with
shoulders at 459 (ε ϭ 3.2 ϫ 102 Ϫ1cmϪ1) and 506 nm (ε ϭ
2.2 ϫ 102 Ϫ1cmϪ1), suggesting the formation of a different
species in acetone.

Ϫ1cmϪ1) and 428 nm (ε ϭ 4.3 ϫ 103
Ϫ1cmϪ1), whereas in acetone the complex exhibits an ab-
Figure 4. Cyclic voltammogram of 4 in acetone: (a) scan rate 50
mV/s; (b) Scan rate 100 mV/s; (c) scan rate 500 mV/s; (d) 600 mV
used as vertex potential
Combined with the NMR spectroscopic data the results
can be rationalized in the following manner (Scheme 2). In
acetone loss of a labile acetonitrile ligand can occur afford-
ing 7, which is assumed to have a redox potential of 940
mV. Oxidation of this complex will lead to the formation of
the unstable mono-acetonitrile iron() complex 8. Loss of
the final acetonitrile ligand will result in the formation of
an [(N3Py-Me)Fe(X)]3ϩ species 9, in which X can be ace-
tone or water. This species will exhibit a different electro-
chemical behavior, presumably giving rise to the reduction
wave at 350 mV.
1
Figure 3. H NMR spectrum of 4 (a) in CD3CN and (b) in [D6]-
acetone
Cyclic voltammetry of 4 and 5 in CH3CN showed one
single quasi-reversible redox wave for the FeII/FeIII couple
at E1/2 ϭ 1180 mV (∆E ϭ 100 mV; ia/ip ϭ 0.95) and 1200
mV vs. SCE (∆E ϭ 160 mV; ia/ip ϭ 0.98), respectively.
These values are significantly higher than those found for
[(N4Py)Fe(CH3CN)](ClO4)2 (E1/2 ϭ 1010 mV vs. SCE).[27]
Apparently, in acetonitrile the low-spin iron() state is even
more favored in the case of bis-acetonitrile complexes like
4
than mono-acetonitrile complexes like [(N4Py)-
Fe(CH3CN)](ClO4)2.
In acetone as solvent a different redox behaviour was ob-
served for 4 (Figure 4). At a scan rate of 50 mV/s an oxi-
dation wave was found at 1150 mV vs. SCE accompanied
by a small reduction wave at 940 mV (ia/ip ϭ 0.35) and a
strong reduction wave at 350 mV vs. SCE. Increase of the
scan rate leads to an increase of the reduction wave at 940
Scheme 2. Redox behavior of 4 in CH3CN and acetone
Catalytic Oxidation with N3Py Iron Complexes
Complexes 4 and 5 were tested in catalytic oxidations of
mV (ia/ip ϭ 0.44 at 100 mV/s; ia/ip ϭ 0.58 at 500 mV/s). both alkanes and alkenes. The results are summarized in
When 600 mV vs. SCE was used as the vertex potential, i.e. Table 1 and 2. In view of the different properties of the
without going through the oxidation at 1150 mV, no re- complexes in acetone and acetonitrile (vide supra), com-
duction wave was observed at 350 mV (Figure 4d). These bined with the fact that it is known from catalytic oxidation
results indicate that oxidation of the complex in acetone with [(N4Py)Fe(CH3CN)](ClO4)2 that the solvent can have
leads to an unstable species that decomposes to give a new a pronounced effect on the observed reactivity and selec-
species with a reduction potential at 350 mV.
tivity,[31] both solvents were employed in catalytic oxidation.
848
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 846Ϫ856

Non-Heme Iron Complexes for Stereoselective Oxidation
FULL PAPER
Table 1. Oxidation of alkanes by 4 and 5; reactions performed under argon
Entry
Substrate
Products
Acetonitrile[a]
Acetone[a]
4
5
4
5
1
cyclohexane
cyclohexanol
cyclohexanone
A/K
trans-1,2-dimethylcyclohexanol
cis-1,2-dimethylcylcohexanol
stereoselectivity (%)
trans-1,2-dimethylcyclohexanol
cis-1,2-dimethylcylcohexanol
stereoselectivity (%)
1-adamantanol
7.9
8.1
1.4
5.8
2.7
0.6
73
0.3
0.6
8.8
2.0
4.4
4.8
1.8
82
1.6
1.5
50
5.7
1.5
3.8
4.1
1.9
68
1.1
1.1
50
1.7
4.6
2.3
0.4
85
0.4
1.1
73
2[b]
3[b]
4
cis-1,2-dimethylcyclohexane
trans-1,2-dimethylcylohexane
adamantane
48
13.1
2.5
13.3
1.1
2-adamantanol
2-adamantanone
tertiary/secondary
1.3
10.3
0.5
24.9
[a]
[b]
Turnover number ϭ mol product/mol catalyst. Secondary alcohols and ketones not quantified.
Table 2. Oxidation of alkenes by 4 and 5 with 50 equiv. H2O2; reactions carried out under air, unless noted otherwise
Acetonitrile[a][b] Acetone[a][c]
Entry Complex Substrate
cis-epoxide trans-epoxide meso-diol rac-diol cis-epoxide trans-epoxide meso-diol rac-diol
1
2
3
4
4
5
4
4
cyclooctene
cyclooctene
cyclohexene[d][e]
c-stilbene
stereoselectivity (%) 93
t-stilbene
stereoselectivity (%)
c-4-octene
stereoselectivity (%) 95
t-4-octene
stereoselectivity (%)
norbornene
14
Ϫ
Ϫ
Ϫ
0.7
22
14
13
8
93
0.5
1
0
0.7
0.6
17
19
0.8
7
70
0
Ϫ
Ϫ
Ϫ
3
0.2
3
0
4
10
2.4
7
54
3
16
11[f]
9[g]
6
5
6
7
8
4
4
4
4
0[h]
8
Ͼ99
7
93
1
6
Ͼ99
11
79
0
19
1
17
94
1
11
79
2
3
1
Ͼ99
1
2
4
67
Ϫ
1
50
2
6
75
Ϫ
1
50
exo 0
endo 1
36
exo 2
endo 1
2
43
[a] Turnover number ϭ mol product/mol catalyst. [b] Reaction time 60 min. [c] Reaction time 30 min. [d] Reactions carried out under argon.
[e]
[f]
In acetone many other products were found, which were not identified. 28 turnovers towards cyclohexenol and 7 turnovers towards
cyclohexenone were observed. [g] 11 turnovers towards benzaldehyde were observed. [h] 13 turnovers towards benzaldehyde were observed.
Reactions were carried out under air, except in the case [(bpmen)Fe(CH3CN)2](ClO4)2,[12,13] the most active non-
of substrates that are prone to undergo autoxidation in the heme iron cyclohexane oxidation catalyst known to date.
presence of O2, and an excess of substrate (1000 equiv.) was The major product was cyclohexanol, with an alcohol/ke-
used to maximize turnover numbers. Hydrogen peroxide (50 tone (A/K) ratio of 3.8Ϫ5.8, significantly higher than those
equiv.) was introduced into the reaction mixture at 25 °C reported for radical chain autoxidation reactions (A/K ഠ
by syringe pump to reduce the loss of hydrogen peroxide 1).[37Ϫ39]
due to catalase activity. Typical addition times are 60 and
30 min for acetonitrile and acetone, respectively. After the
In acetone a similar reactivity and A/K ratio was ob-
addition of H2O2 was complete the reaction mixture was tained as in acetonitrile (Table 1, entry 1). However, one
stirred for another 5 min before a sample was taken and clear difference is seen with the KIE values for cyclohexanol
analyzed by GC.
formation, determined in a competition experiment be-
tween cyclohexane and [D12]cyclohexane. In CH3CN a kH/
kD value of 3.5 was found for complex 4, similar to the
Oxidation of Alkanes
Complexes 4 and 5 show a fair activity in the oxidation value observed for [(TPA)Fe(CH3CN)2](ClO4)2,[36] which
of cyclohexane in CH3CN, with yields up to 19% based on was proposed to react via a formally FeVO species. Complex
hydrogen peroxide for complex 4 (Table 1, entry 1). These 5 gives a lower kH/kD value of 2.4. In acetone the kH/kD
yields are slightly lower than those observed for [(N4Py)- ratio increased significantly to 4.4 for 4 and 3.9 for 5. The
Fe(CH3CN)](ClO4)2,[28]
and
[(TPA)Fe(CH3CN)2]- latter values are in the range of those typically observed in
(ClO4)2,[36] and significantly lower than those found for oxidations catalyzed by non-heme iron complexes using
Eur. J. Inorg. Chem. 2004, 846Ϫ856
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
849

M. Klopstra, G. Roelfes, R. Hage, R. M. Kellogg, B. L. Feringa
FULL PAPER
tert-butyl hydroperoxide as oxidant,[40] which were demon- whereas 5 displayed a small preference for epoxide forma-
strated to proceed via alkoxy radicals.[36Ϫ38]
tion.
Cyclohexene as substrate also gave the epoxide and the
The oxidation of cis- and trans-1,2-dimethylcyclohexane
to the corresponding tertiary alcohols was studied to assess
the stereoselectivity of the oxidation reaction. Although
several isomeric secondary alcohols were also formed, these
were not quantified. In CH3CN the oxidation of cis-1,2-
dimethylcyclohexane afforded rather low yields of tertiary
alcohols with 85% and 82% retention of configuration at
the tertiary carbon (i.e. formation of trans-1,2-dimethylcy-
clohexanol) for complexes 4 and 5, respectively (Table 1, en-
try 2). Using trans-1,2-dimethylcyclohexane low yields of
the tertiary alcohols, with 73% and 50% retention of con-
figuration (i.e. cis-dimethylcyclohexanol) were also found
for 4 and 5 (Table 1, entry 3). For comparison, more than
99% retention of stereochemistry was observed with
meso-diol as main products (Table 2, entry 3). However,
with this substrate allylic oxidation products — cyclohex-
enol (28 turnovers) and cyclohexenone (seven turnovers) —
were also formed.
With cis-stilbene and cis-4-octene the cis-oxide and meso-
diol were also formed as major products, with a stereoselec-
tivity of 93Ϫ95% (Table 2, entries 4 and 6). With trans-stil-
bene the trans-oxide, was the only detectable epoxidation
product (Table 2, entry 5). The cis-dihydroxylation product,
i.e. rac-1,2-diphenylethanediol, was obtained with a stereo-
selectivity of 94%. Besides the oxides and diols substantial
amounts of benzaldehyde were also formed with the stil-
benes — 11 and 13 turnovers for cis- and trans-stilbene,
respectively. With trans-4-octene and norbornene a prefer-
ence for epoxidation was found. For both substrates the
corresponding diols were only formed in small amounts.
In acetone as solvent a very different reactivity pattern
was observed. Generally, more side-products were found
compared to oxidations in CH3CN as solvent. Using cyclo-
octene as substrate, the epoxide and the 1,2-diols were ob-
tained as major products (Table 2, entries 1 and 2). How-
ever, in contrast to the results in CH3CN now the trans-
dihydroxylation product, i.e. rac-1,2-cyclooctanediol, was
obtained as major product, with a selectivity of 95% in the
case of complex 4. Complex 5 showed a little bit lower sel-
ectivity towards rac-diol (76%), but a significantly in-
creased yield.
Cyclooctene oxide is known to be very robust so it is
unlikely that the rac-diol is formed by epoxide ring-open-
ing.[48] Furthermore, when cyclooctene oxide was subjected
to the standard reaction conditions no diol formation was
observed, which demonstrates that the rac-diol is formed
independently and is not a decomposition product from the
epoxide. Thus, by changing the solvent from acetonitrile to
acetone the stereochemical outcome of the dihydroxylation
reaction is changed from cis to trans.
With cis- and trans-4-octene and cis- and trans-stilbene
the epoxides were formed in similar amounts as for the re-
action in CH3CN, with the cis-epoxidation product as
major product (Table 2, entries 4Ϫ7). Hence, the stereo-
chemistry of epoxidation is not dependent on the solvent.
With both cis- and trans-stilbene only small amounts of
diol were found. However, substantial amounts of the cor-
responding acetonides were obtained, which are the prod-
ucts of the reaction of the diols with the solvent to form an
acetal. Similar amounts of acetal were found by indepen-
dent reaction of the corresponding diol under the same re-
action conditions, indicating strongly that the acetonide is
derived from the diol. When the acetonides are taken into
account cis-stilbene gives the trans-dihydroxylation
products — the meso-diol — as the main product and trans-
[(TPA)Fe(CH3CN)2](ClO4)2,[41]
[(bpmen)Fe(CH3CN)2]-
[12]
(ClO4)2
and cytochrome P-450 models,[42] whereas a
complete lack of stereoselectivity was observed for [(N4Py)-
Fe(CH3CN)](ClO4)2 [31] and in free-radical autoxidation.[43]
In acetone higher yields of tertiary alcohols were ob-
tained, but generally with a lower degree of stereoselectivity.
In the case of cis-1,2-dimethylcyclohexane 73% and 68%
retention of configuration (Table 1, entry 2), and in the case
of trans-1,2-dimethylcyclohexane 48% and 50% retention of
configuration were obtained (Table 1, entry 3) with 4 and
5, respectively.
Adamantane as substrate was oxidized both at the sec-
ondary and tertiary carbons. Using 4 a tertiary/secondary
ratio (normalized on a per hydrogen basis) of 10.3 was ob-
tained in acetonitrile (Table 1, entry 4). In acetone this ratio
increased to 24.9 (Table 1, entry 4). For TPA and related
tetradentate nitrogen-donor ligands values between 15 and
33 have been reported in CH3CN as solvent.[13] For com-
parison, tertiary/secondary ratios of about 2 have been
found for oxidation of alkanes by ·OH,[11] values of 3.1Ϫ3.3
for [(N4Py)Fe(CH3CN)](ClO4)2,[31] 9.5Ϫ10 for oxidations
by tBuOOH utilizing catalysts like [Fe2O(bipy)4(H2O)2]-
(ClO4)4 and [FeCl2(TPA)](ClO4),[44,45] and 11Ϫ48 for oxi-
dations with PhIO catalyzed by P-450 mimics.[46]
Oxidation of Alkenes
The oxidation of symmetrical alkenes in CH3CN with
H2O2, catalyzed by 4 and 5, gave both epoxidation and cis-
dihydroxylation products (Table 2). Based on the results
with Fe-TPA complexes this would be expected for iron
complexes with two open coordination sites cis to each
other.[15] All reactions could be carried out under air, except
with cyclohexene as substrate, since this led to a strongly
increased formation of the allylic alcohol and ketone. This
indicates involvement of autoxidation pathways under
air.[47]
Cyclooctene as a substrate afforded a very clean reaction stilbene affords mainly the corresponding cis-dihydroxyl-
giving only the epoxide and the cis-dihydroxylation prod- ation products, i.e. the rac-diol.
uct, i.e. the meso-diol (Table 2, entries 1 and 2). A small
With norbornene as substrate the main product was
preference for cis-diol formation was observed with 4 found to be rac-norbornanediol (Table 2, entry 8). When
850
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 846Ϫ856

Non-Heme Iron Complexes for Stereoselective Oxidation
FULL PAPER
the epoxide was tested under these reaction conditions no
diol formation was observed, again suggesting that the diol
is not formed as a decomposition product of the epoxide.
With cyclohexene as substrate many products were formed,
including low yields of cyclohexene oxide and rac-1,2-cyclo-
hexanediol (Table 2, entry 3). No evidence for cis-dihydrox-
ylation products was found. However, many of the side-
products that were formed have not been identified up to
now, except the allylic oxidation products cyclohexenol (11
turnovers) and cyclohexenone (one turnover), which makes
it difficult to compare the results for this substrate.
The most remarkable feature of the reactions in acetone
is the fact that, in all cases, the rac-diol is the preferred
product. For cis-alkenes like norbornene, cyclooctene,
cyclohexene, and cis-stilbene, this amounts to a trans-dihy-
droxylation reaction. This is the first example of a catalyst
that can give both cis- and trans-dihydroxylation just by
changing the solvent. Catalytic cis-dihydroxylation is
mostly achieved by OsO4 complexes in the presence of a
stoichiometric oxidant like K3Fe(CN)6 or N-methylmorph-
oline N-oxide.[21] trans-Diols are usually prepared from al-
kenes in two steps: first epoxidation followed by epoxide
ring-opening.[49]
Figure 5. Incorporation of 18O from H218O: • diol; ᭜ epoxide
At low concentration there is a more or less linear re-
lationship between the concentration of water and the de-
gree of incorporation of 18O into the diol. At higher con-
centrations the extent of oxygen incorporation relative to
the enhanced concentration decreases. This result is indica-
tive of a pre-equilibrium between water and the active spec-
ies of the catalyst. Similar results were found for the model
systems [(bpmen)Fe(CH3CN)2](ClO4)2, [(TPA)Fe(CH3-
CN)2](ClO4)2, and several Fe-TPA derivatives for the oxi-
dation of alkanes and alkenes.[50] Based on the cumulative
Labeling Studies
data,
which
are
in
accordance
with
To obtain more information on the mechanism of the [(TPA)Fe(CH3CN)2](ClO4)2, the same mechanism for the
epoxidation and dihydroxylation reaction 18O labeling oxidation in CH3CN is proposed as was postulated for the
experiments were performed. The oxidation of cyclooctene oxidation with [(TPA)Fe(CH3CN)2](ClO4)2 (Scheme 3). In
in acetonitrile was monitored by using 10 equivalents of a this mechanism the active (HO)FeVϭO species 13 is formed
2% solution of H218O2 in H216O under a N2 atmosphere by heterolysis of the peroxide bond in intermediate 12.
Table 3). Analysis was performed by GC/CI-MS. Only 50%
These labeling experiments were repeated in acetone,
of the epoxide contained 18O. For the meso-diol predomi- using cis-stilbene as substrate since it gives higher yields of
nantly the singly labeled product was found. In a comp- trans-dihydroxylation products than cyclooctene. Here, in-
lementary experiment using a 2% solution of H216O2 in corporation of 18O-oxygen into both epoxides and diols was
H218O 28% of the epoxide contained an 18O atom. For the investigated. The results are summarized in Table 4.
cis-diol the main product was again singly labeled cis-cyclo-
Even though a different substrate was used for the labe-
octanediol. When a solution of 1% H218O2 in a 1:1 mixture ling studies and the level of 18O incorporation is generally
of H216O and H218O was used, both singly and doubly lab- lower than in CH3CN, the same pattern as in CH3CN is
eled cis-cyclooctanediols were formed. These observations discernible for the epoxidation and cis-dihydroxylation.
lend support to a proposed catalytically active species in
About one third of the label is built into the epoxide
which both oxygen atoms from hydrogen peroxide and when H218O2 is used and about 40% incorporation is found
water are present. Further evidence for such a mechanism when H218O is used. Also, in the case of the cis-diol the
was obtained by examining the incorporation of labeled main fraction is singly labeled by H218O2. With H218O
oxygen as a function of concentration of water in the reac- about one third of the product contains a single labeled
tion mixture. The results are shown in Figure 5.
oxygen, whereas with a combination of H218O2 and H218O
Table 3. Incorporation of 18O into the oxidation products of cyclooctene in acetonitrile[a]
Product
H218O2 in H2O
H2O2 in H218O
H218O2 in H218O/H2O (1:1)
epoxide, labeled
50 (7)
3 (1)
94 (1)
3 (1)
28 (1)
23 (1)
76 (1)
1 (1)
53 (4)
3 (1)
37 (1)
60 (1)
cis-diol, not labeled
cis-diol, single labeled
cis-diol, double labeled
[a]
Margins of error in parentheses.
Eur. J. Inorg. Chem. 2004, 846Ϫ856
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
851

M. Klopstra, G. Roelfes, R. Hage, R. M. Kellogg, B. L. Feringa
FULL PAPER
Scheme 3. Proposed mechanism for the catalytic oxidation by N3Py-MeFe(ClO4)2(CH3CN)2 in acetonitrile
Table 4. Incorporation of 18O into the oxidation products of cis-stilbene in acetone[a]
Product
H218O2 in H2O
H2O2 in H218O
H218O2 in H218O/H2O (1:1)
cis-epoxide, labeled
trans-epoxide, labeled
cis-diol, not labeled
cis-diol, single labeled
cis-diol, double labeled
trans-diol, not labeled
trans-diol, single labeled
trans-diol, double labeled
32 (1)
21 (1)
10 (2)
67 (8)
23 (10)
19 (8)
63 (6)
18 (2)
40 (1)
17 (5)
70 (5)
29 (5)
1 (1)
70 (3)
17 (6)
13 (3)
49 (6)
23 (5)
15 (1)
42 (4)
43 (4)
29 (1)
49 (1)
22 (1)
[a]
Margins of error in parentheses.
about 40% of the product is singly labeled and 43% doubly
labeled. Taken together the data suggest that the epoxide
and the cis-diol are formed by the same mechanism as with
Mechanistic Interpretation
N3Py-R Fe complexes (4 and 5) are clearly very promis-
CH3CN as solvent, involving a catalytically active species ing catalysts for stereoselective catalytic oxidation with
in which oxygen atoms from both hydrogen peroxide and H2O2. The reactivity profile in acetonitrile, not unexpec-
water are present.
tedly, is very characteristic of an iron complex with two
The labeling data for the rac-diol are somewhat harder cis open coordination sites: stereoselective hydroxylation of
to interpret. One complication is the fact that the rac-diol alkanes and stereoselective epoxidation and dihydroxylation
was obtained in two forms, the ‘‘free’’ diol and the aceton- of alkenes. The reactivity pattern of alkane oxidation, in-
ide. Since the acetonide is a secondary product, derived cluding the observed KIE values in cyclohexane oxidation,
from the diol, only the free diol was considered in the the 3:2 ratio in adamantane oxidation and the high degree
experiments. In the experiment with H218O2 63% of the of stereoselectivity in the oxidation of cis- and trans-1,2-
product contains a single label whereas roughly equal dimethylcyclohexane observed with 4 and 5 is strongly
amounts of non-labeled and doubly labeled diol were reminiscent of those reported for other iron complexes of
[36]
found. This appears to suggest that at least one oxygen of tetradentate N ligands like [(TPA)Fe(CH3CN)2](ClO4)2
the diol comes from H2O2. The use of H218O resulted and [(bpmen)Fe(CH3CN)2](ClO4)2.[12] Furthermore, the
mainly in a product without a label whereas the experiment oxidation of alkenes gives rise to the corresponding epox-
with H218O2 and H218O gave only a marginally higher ides and cis-dihydroxylation products with high stereoselec-
amount of doubly labeled product. Thus, the second oxygen tivity. These results are also similar to what was reported
in the trans-diol does not have a clear origin and can stem for [(TPA)Fe(CH3CN)2](ClO4)2 in CH3CN.[15,36] These fin-
from multiple sources like H2O2, H2O or, most probably, dings suggest that a common mechanistic pathway for these
the solvent acetone.
complexes exists. A mechanism involving heterolysis of the
Since the experiments were performed under an inert at- OϪO bond of a low-spin FeIIIOOH intermediate 12 to give
mosphere, incorporation of oxygen from O2 can be ex- a formally FeVO species 13 is proposed, similar to that for
cluded.
[(TPA)Fe(CH3CN)2](ClO4)2-catalyzed hydroxylation[12][50a]
852
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 846Ϫ856

Non-Heme Iron Complexes for Stereoselective Oxidation
FULL PAPER
and cis-dihydroxylation (Scheme 3).[15,17] Addition to a stereoselectivity in acetone, or the rac-diol by reaction with
double bond presumably produces a radical species 14 that another oxygen radical species.
is too short lived to isomerize and thus rapidly reacts with
the coordinated OH to form the cis-diol. The same mecha-
Conclusions
nism is expected to apply to the N3Py-R complexes dis-
cussed here. This is clearly supported by the labeling experi-
ments.
In this paper we have presented a new class of tetraden-
tate N-ligands for application in iron-catalyzed oxidations
using H2O2: N3Py-R, with R being Me or Bn. Compared to
the parent ligand of this class of catalysts, TPA, the N3Py-R
ligands presented here have the added advantage that the R
group can be readily modified, for example by the introduc-
tion of chiral groups for use in catalytic asymmetric oxi-
dations.
The electrochemistry, NMR features and UV/Vis spectra
of the N3Py-Fe complexes are solvent dependent. In ace-
tone the labile acetonitrile ligands are rapidly displaced,
presumably by the solvent, leading to the formation of a
high-spin complex with different spectroscopic and electro-
chemical behaviour.
The oxidation results obtained in acetone as solvent are
quite different from those in CH3CN, and it is hard to
rationalize them in a single mechanistic proposal. Based on
the results it is more likely that several competing path-
ways exist.
Based on the labeling studies it appears that an active
species similar to that proposed in the mechanism in
CH3CN, resulting from heterolysis of a FeIIIOOH inter-
mediate, is responsible for the formation of epoxide and cis-
diol. The fact that the second O atom in the product is
clearly not only derived from water molecules is presumably
caused by scrambling with the solvent.
However, from the data it is clear that this cannot be the
sole pathway. In particular, the fact that rac-diols are an
important product of the oxidation of alkenes suggests that
another mechanistic pathway exists as well. Furthermore,
the higher KIE in the oxidation of cyclohexane and the
stronger preference for oxidation at tertiary carbons in ada-
mantane suggest the presence of a more selective oxidizing
species but could also indicate the presence of alkoxy-type
radicals (vide supra). The lower stereoselectivity in the hy-
droxylation of cis- and trans-1,2-dimethylcyclohexane, al-
kene epoxidation and alkene dihydroxylation, combined
with the fact that generally more side products are formed
in acetone than in CH3CN as solvent, are certainly more
indicative of the involvement of radical species.
One possibility for radical species to be formed is through
the formation of an iron-acetonylperoxy intermediate (15),
by reaction of the iron center with the hydrogen peroxide
adduct of acetone, followed by homolysis of the OϪO bond
to generate an FeIVO complex (1) and an acetonyloxy rad-
ical species 17 (Scheme 4). Iron-alkylperoxo complexes have
been demonstrated to be particularly predisposed to un-
dergo OϪO bond homolysis.[51]
In both solvents good oxidation activity with H2O2 as
sacrificial oxidant was found, but the reactivity, and in par-
ticular the selectivity, are strongly solvent dependent.
Not surprisingly the reactivity pattern in acetonitrile is
quite similar to that of [(TPA)Fe(CH3CN)2](ClO4)2. The
N3Py-Fe complexes are capable of stereoselective hydroxyl-
ation, epoxidation and cis-dihydroxylation. Mechanistic
studies support a mechanism similar to that described for
TPA-Fe, involving formally FeVO species.
A major finding reported in this paper is the very differ-
ent reactivity pattern observed in acetone, a solvent never
reported for [(TPA)Fe(CH3CN)2](ClO4)2. The most striking
example is the oxidation of alkenes to give the correspond-
ing epoxides and the rac-diol as the main products. Mech-
anistic studies revealed that most likely more than one com-
peting mechanistic pathway is available in acetone. Besides
the heterolysis of a tentative FeIIIOOH intermediate to give
formally FeVO species — the dominant mechanism in
CH3CN — a pathway involving FeIVO species and acetony-
loxo radicals, formed by homolysis of an intermediate iron-
acetonylperoxy species, is proposed. The FeIVO species may
be responsible for the formation of trans-diols, following
the TMP-Fe precedent.
The work described here constitute the first examples of
a new generation of iron-based oxidation catalysts that can
give either cis-dihydroxylation or trans-dihydroxylation of
alkenes, simply by switching from CH3CN to acetone as
solvent.
Scheme 4. Proposed formation of radical species by homolysis of
the OϪO bond of a iron-acetonyl peroxy species
Experimental Section
Such radical species could also explain the formation of
rac-diols: analogous to the proposal for TMP-Fe complexes
(TMP ϭ tetramesitylporphyrin),[52] the FeIVO species could
add to the double bond, leaving a carbon radical intermedi-
ate that is sufficiently long-lived to undergo isomerization
to the thermodynamically more stable trans form before
General Information: Commercially available chemicals were used
without further purification unless noted otherwise. H218O2 (90%
18O-enriched, 2% solution in H216O) and H218O (95% 18O-en-
riched) was obtained from ICON Services Inc. N3Py[33] and cis-4-
octene[53] were prepared according to literature procedures.
1H NMR and 13C NMR spectra were recorded on a Varian Gemini
formation of either the epoxide, which explains the reduced 300 spectrometer at ambient temperature (1H at 300 MHz and 13
C
Eur. J. Inorg. Chem. 2004, 846Ϫ856
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
853

M. Klopstra, G. Roelfes, R. Hage, R. M. Kellogg, B. L. Feringa
at 75 MHz). Chemical shifts (in ppm) are referenced to the residual added in small portions. After stirring for 7 h at room temperature
FULL PAPER
protic solvent peaks. X-band EPR spectra were obtained at liquid
nitrogen temperature on a Bruker ECS106 instrument or at liquid
helium temperatures on a Bruker E500 equipped with an Oxford
Instruments ESR-10 cryostat. UV/Vis spectra were recorded on a
Hewlett Packard 8453 diode array spectrophotometer, using a cell
equipped with a home-built temperature-control device. Electro-
chemical studies were carried out with a PAR 273A potentiostat
in acetonitrile with 0.1  tetrabutylammonium perchlorate as the
saturated NaHCO3(aq) (35 mL) was added and the 1,2-dichloroe-
thane layer was separated. The aqueous layer was extracted with
CH2Cl2 (3 ϫ 20 mL). The combined organic layers were washed
with 1  NaOH (20 mL) and brine (20 mL), dried (Na2SO4) and
the solvent removed in vacuo to give 1 (1.235 g, 4.27 mmol, 93%)
1
as a slightly yellow oil. H NMR (CDCl3): δ ϭ 2.19 (s, 3 H), 3.72
(s, 2 H), 4.96 (s, 1 H), 7.14 (m, 3 H), 7.71 (m, 6 H), 8.56 (m, 3 H)
ppm. 13C NMR (CDCl3): δ ϭ 40.32 (q), 61.04 (t), 77.87 (d), 121.76
supporting electrolyte. Cyclic voltammograms (CV) were obtained (d), 122.10 (d), 122.92 (d), 123.21 (d), 136.29 (d), 136.43 (d), 148.86
by using a three-component system consisting of a glassy carbon
working electrode, a platinum wire auxiliary electrode and a satu-
rated calomel reference electrode. Mass spectra were recorded on a
NERMAG R 3010 triple quadrupole mass spectrometer (Nermag,
Argenteuil, France) equipped with a home-built atmospheric-press-
ure ionization source. GC analyses were performed on a Hewlett
Packard 6890 Gas Chromatograph using a HP-1 dimethyl polysi-
loxane column or a HP-5 5%-phenyl methyl siloxane column, or a
Hewlett Packard 5890 II Gas Chromatograph using a CP-wax 52
CB column or a CP-wax 57 CB column.
(d), 149.22 (d), 159.37 (s), 160.59 (s) ppm. MS (CI): m/z ϭ 291 [M
ϩ 1].
N-Benzyl-N-[di(2-pyridinyl)methyl]-N-(2-pyridinylmethyl)amine
(N3Py-Bn, 2): A mixture of benzyl chloride (0.46 mL, 3.8 mmol),
3 (740 mg, 2.7 mmol) and K2CO3 (370 mg, 2.7 mmol) in CH3CN
(20 mL) was stirred under an argon atmosphere. The mixture was
heated under reflux overnight and benzyl chloride (0.23 mL,
1.9 mmol) was again added. After heating under reflux for 24 h the
solvent was evaporated and water (20 mL) was added to the resi-
due. The product was extracted with ethyl acetate (3 ϫ 30 mL) and
the combined ethyl acetate layers were washed with brine (15 mL)
and dried (Na2SO4). Evaporation of the solvent followed by col-
umn chromatography (Alox. neutral, akt. I, ethyl acetate/hexane/
triethylamine, 10:6:1) afforded 2 (670 mg, 1.83 mmol, 68%) as a
yellow oil. 1H NMR (CDCl3): δ ϭ 3.79 (s, 2 H), 3.90 (s, 2 H), 5.30
(s, 1 H), 7.14 (m, 4 H), 7.29 (m, 2 H), 7.40 (m, 2 H), 7.66 (m, 6
H), 8.48 (m, 1 H), 8.56 (m, 2 H) ppm. 13C NMR (CDCl3): δ ϭ
54.93 (t), 56.47 (t), 71.24 (d), 121.62 (d), 121.90 (d), 122.60 (d),
123.75 (d), 126.77 (d), 128.14 (d), 128.70 (d), 136.11 (d), 136.23 (d),
138.94 (s), 148.81 (d), 149.15 (d), 160.04 (s) ppm. MS (CI): m/z ϭ
367 [M ϩ 1].
Caution: Perchlorate salts are potentially explosive and should be
handled with care.
Crystallographic Studies. Crystal Structure Determination of 4:
(C22H24FeN6)2ϩ·(ClO4Ϫ)2, Mr
ϭ
627.22, triclinic, P1,
a
ϭ
¯
˚
10.504(8), b ϭ 11.685(9), c ϭ 12.449(9) A, α ϭ 77.20(1)°, β ϭ
3
˚
65.71(1)°, γ ϭ 66.34(1)°, V ϭ 1272.7(17) A , Z ϭ 2, Dx ϭ 1.637
gcmϪ3, λ(Mo-Kα) ϭ 0.71073 A, µ ϭ 8.63 cmϪ1, F(000) ϭ 644, T ϭ
˚
90 K, GooF ϭ 0.970, wR(F2) ϭ 0.2377 for 4307 reflections and 355
parameters and R(F) ϭ 0.0856 for 2378 reflections obeying Fo Ն
4.0 σ(Fo) criterion of observability.
The asymmetric unit consists of three moieties, a cationic Fe com-
plex and two ClO4Ϫ anions. Crystals were obtained by recrystalliza-
tion from a mixture of acetonitrile and pentane. The crystals were
very fragile: contact with a glass fiber caused damage as crystal
cracks. The recorded frames showed (broad) anisotropic mosaicity
and duplicated reflections. A red parallelepiped-shaped crystal with
the dimensions of 0.220 ϫ 0.150 ϫ 0.120 mm mounted on a glass
fiber was aligned on a Bruker[54] SMART APEX CD dif-
fractometer (Platform with full three-circle goniometer). The dif-
fractometer was equipped with a 4 K CCD detector set 6.0 cm from
the crystal. The crystal was cooled to 90 K using the Bruker KRY-
OFLEX low-temperature device. Intensity measurements were per-
formed using graphite-monochromated Mo-Kα radiation from a se-
aled ceramic diffraction tube (SIEMENS). Generator settings were
50 kV/40 mA.
[(N3Py-Me)Fe(CH3CN)2](ClO4)2
(4):
A
solution
of
Fe(ClO4)2·6H2O (250 mg, 0.69 mmol) in CH3CN (3 mL) was added
to a solution of 1 (198 mg, 0.68 mmol) in MeOH (3 mL). The dark-
red solution was placed in an ethyl acetate bath and after 3 days 4
(344 mg, 0.55 mmol, 81%) was isolated as dark-red crystals. 1H
NMR (CD3CN): δ ϭ 3.68, 5.12, 6.83, 7.40 (m, 1 H), 7.63 (m, 1 H),
8.02 (m, 1 H), 8.52, 8.65, 8.77, 8.95, 11.21 ppm. C22H24Cl2FeN6O8:
calcd. C 42.13, H 3.86, N 13.40; found C 41.98, H 3.78, N 13.27.
UV/Vis (in CH3CN): λmax (ε) ϭ 352 nm (5288 Ϫ1cmϪ1), 428
(3261).
[(N3Py-Bn)Fe(CH3CN)2](ClO4)2 (5): Following the procedure as
for 4, starting from 2 (95 mg, 0.26 mmol) and Fe(ClO4)2·6H2O
(105 mg, 0.29 mmol), 5 (152 mg, 0.22 mmol, 85%) was obtained as
dark red crystals. 1H NMR (CD3CN): δ ϭ 2.82, 4.44, 5.84, 6.98
(m, 3 H), 7.35 (m, 1 H), 7.49 (m, 3 H), 8.02 (m, 1 H), 9.40, 9.50,
9.64, 9.72, 9.79, 9.96, 10.78, 13.17 ppm. ES/MS: m/z ϭ 521 [M Ϫ
The structure was solved by Patterson methods (DIRDIF) and re-
fined with SHELXL against F2 of all reflections. Non-hydrogen
atoms were refined freely with anisotropic displacement param-
eters. Hydrogen atoms were refined in rigid mode. Checking for
higher symmetry and structure calculations were performed with
the PLATON package.
CCDC-213726 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieving.html [or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ, UK; Fax: (internat.) ϩ44-1223/336-033; E-mail:
deposit@ccdc.cam.ac.uk].
(ClO4)Ϫ Ϫ 2(CH3CN)]ϩ, 231.5 [M Ϫ 2(ClO4)Ϫ Ϫ (CH3CN)]2ϩ
.
Catalytic Oxidations: All experiments were carried out under air,
unless noted otherwise, in a water bath thermostatted at 25 °C. In
a typical procedure cyclohexane (0.19 mL, 1000 equiv.) was added
to a solution of 9 (8.75 ϫ 10Ϫ4 , 2 mL) and a known amount
of bromobenzene (internal standard) in acetone. The reaction was
started by syringe pump addition of 1  H2O2 (88 µL, 50 equiv.)
over 30 min. After the addition was complete the reaction mixture
was stirred for an additional 5 min. An aliquot (1 mL) was taken
from the reaction and filtered through a small silica column. The
silica was thoroughly washed with diethyl ether or diethyl ether/
10% methanol. The sample was concentrated to a volume of 2 mL
by passing an Ar stream over the solution and then analyzed by
GC.
N-[Di(2-pyridinyl)methyl]-N-methyl-N-(2-pyridinylmethyl)amine
(N3Py-Me, 1): Formaldehyde (37% solution in water, 0.45 mL,
6.0 mmol) was added to a solution of 3 (1.262 g, 4.59 mmol) in
1,2-dichloroethane (35 mL). NaBH(OAc)3 (3.92 g, 18.5 mmol) was
854
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 846Ϫ856

Non-Heme Iron Complexes for Stereoselective Oxidation
FULL PAPER
Y. Mekmouche, S. Menage, C. Toia-Duboc, M. Fontecave, J.-
[14]
´
Kinetic Isotope Effect Determination: Essentially the same pro-
cedure was used as described above, but now a mixture of cyclohex-
ane and [D12]cyclohexane was used (ratio cyclohexane/[D12]cy-
clohexane varied from 1:1 to 1:4). The KIE was determined by
comparing the turnover numbers for cyclohexanol and [D11]cy-
clohexanol (determined by GC using the CP-wax 52 CB column)
and corrected for relative concentration of cyclohexane and
[D12]cyclohexane.
´
B. Galey, C. Lebrun, J. Pecaut, Angew. Chem. Int. Ed. 2001,
40, 949Ϫ952.
[15]
[16]
[17]
[18]
[19]
[20]
K. Chen, L. Que Jr., Angew. Chem. Int. Ed. 1999, 38,
2227Ϫ2229.
W. Nam, R. Ho, J. S. Valentine, J. Am. Chem. Soc. 1991, 113,
7052Ϫ7054.
K. Chen, M. Costas, J. Kim, A. K. Tipton, L. Que Jr., J. Am.
Chem. Soc. 2002, 124, 3026Ϫ3035.
J. Y. Ryu, J. Kim, M. Costas, K. Chen, W. Nam, L. Que Jr.,
Chem. Commun. 2002, 1288Ϫ1289.
M. Costas, L. Que Jr., Angew. Chem. Int. Ed. 2002, 41,
Labeling Experiments: In a typical procedure cyclooctene (0.23 mL,
1000 equiv.) was added to a solution of 4 (8.75·10Ϫ4 , 2 mL) in
acetonitrile under a N2 atmosphere. The reaction was started by
syringe pump addition of H218O2 (10 equiv., 30 µL of 2% H218O2
in H216O, diluted with 220 µL acetonitrile) over 60 min. After the
addition was complete the reaction mixture was stirred for an ad-
ditional 5 min. The reaction mixture was filtered through a small
silica column. The silica was thoroughly washed with diethyl ether
or diethyl ether/10% methanol. The sample was concentrated to a
volume of 2 mL by passing an Ar stream over the solution and
then analyzed by GC/CI-MS.
2179Ϫ2181.
K. Chen, M. Costas, L. Que Jr., J. Chem. Soc., Dalton Trans.
2002, 672Ϫ679.
[21] [21a] H. C. Kolb, M. S. VanNieuwenhze, K. B. Sharpless, Chem.
[21b]
´
I. E. Marko, J. S. Svendsen,
Rev. 1994, 94, 2483Ϫ2548.
in Comprehensive Asymmetric Catalysis, Vol. II (Eds.: E. N.
Jacobsen, A. Pfaltz, H. Yamamoto), Springer-Verlag, Berlin,
Heidelberg, 1999, pp. 713Ϫ790.
[22]
[23]
T. K. M. Shing, E. K. W. Tam, I. H. F. Chung, Q. Jiang, Chem.
Eur. J. 1996, 2, 50Ϫ57.
In the H218O experiment, H216O2 (50% in H216O, 2 µL) and H218O
(90%, 48 µL) were dissolved in acetonitrile (50 µL) and used for
the oxidation as mentioned above.
J. Brinksma, L. Schmieder, G. van Vliet, R. Boaron, R. Hage,
D. E. de Vos, P. L. Alsters, B. L. Feringa, Tetrahedron Lett.
2002, 43, 2619Ϫ2622.
M. Costas, A. K. Tipton, K. Chen, D.-H. Jo, L. Que Jr., J.
Am. Chem. Soc. 2001, 123, 6722Ϫ6723.
A. Bassan, M. R. A. Blomberg, P. E. M. Siegbahn, L. Que Jr.,
J. Am. Chem. Soc. 2002, 124, 11056Ϫ11063.
M. Lubben, A. Meetsma, E. C. Wilkinson, B. Feringa, L. Que
Jr., Angew. Chem. Int. Ed. Engl. 1995, 34, 1512Ϫ1514.
G. Roelfes, M. Lubben, K. Chen, R. Y. N. Ho, A. Meetsma,
S. Genseberger, R. M. Hermant, R. Hage, S. K. Mandal, V. G.
Young Jr., Y. Zang, H. Kooijman, A. L. Spek, L. Que Jr., B.
L. Feringa, Inorg. Chem. 1999, 38, 1929Ϫ1936.
G. Roelfes, V. Vrajmasu, K. Chen, R. Y. N. Ho, J.-U. Rohde,
C. Zondervan, R. M. la Crois, E. P. Schudde, M. Lutz, A. L.
Spek, R. Hage, B. L. Feringa, E. Munck, L. Que Jr., Inorg.
Chem. 2003, 42, 2639Ϫ2653.
[24]
[25]
[26]
[27]
In the H218O2/H218O experiment, H218O2 (2% in H216O, 25 µL) and
H218O (90%, 25 µL) were dissolved in acetonitrile (50 µL) and used
for the oxidation as mentioned above.
Acknowledgments
Financial support for this research from NWO-CW (project supra-
molecular catalysis) (M. K., B. L. F.) is gratefully acknowledged.
We wish to thank Prof. L. Que Jr. (University of Minnesota) for
helpful discussions and A. Meetsma for performing the X-ray
analysis.
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
R. Y. N. Ho, G. Roelfes, B. L. Feringa, L. Que Jr., J. Am.
Chem. Soc. 1999, 121, 262Ϫ265.
N. Lehnert, R. Y. N. Ho, L. Que Jr., K. Fujisawa, E. I. Solo-
mon, J. Inorg. Biochem. 2001, 86, 313.
G. Roelfes, M. Lubben, R. Hage, L. Que Jr., B. L. Feringa,
Chem. Eur. J. 2000, 6, 2152Ϫ2159.
N. Lehnert, F. Neese, R. Y. N. Ho, L. Que Jr., E. I. Solomon,
J. Am. Chem. Soc. 2002, 124, 10810Ϫ10822.
G. Roelfes, M. E. Branum, L. Wang, L. Que Jr., B. L. Feringa,
J. Am. Chem. Soc. 2000, 122, 11517Ϫ11518.
A. F. Abdel-Magid, K. G. Carson, B. D. Harris, C. A. Maryan-
off, R. D. Shah, J. Org. Chem. 1996, 61, 3849Ϫ3862.
Y. Zang, J. Kim, Y. Dong, E. C. Wilkinson, E. H. Appelman,
L. Que Jr., J. Am. Chem. Soc. 1997, 119, 4197Ϫ4205.
C. Kim, K. Chen, J. Kim, L. Que Jr., J. Am. Chem. Soc. 1997,
119, 5964Ϫ5965.
P. A. MacFaul, K. U. Ingold, D. D. M. Wayner, L. Que Jr., J.
Am. Chem. Soc. 1997, 119, 10594Ϫ10598.
P. A. MacFaul, I. W. C. E. Arends, K. U. Ingold, D. D. M.
Wayner, J. Chem. Soc., Perkin Trans. 2 1997, 135Ϫ145.
S. Evans, J. R. Lindsay Smith, J. Chem. Soc., Perkin Trans. 2
2000, 7, 1541Ϫ1552.
[1] [1a]
R. A. Sheldon, J. K. Kochi, in Metal Catalyzed Oxidations
[1b]
of Organic Compounds, Academic Press, New York, 1981.
D. A. Goetsch, L. D. Schmidt, Science 1996, 271, 1560Ϫ1562.
[1c]
G. B. Shul’pin, J. Mol. Catal. A 2002, 189, 39Ϫ66.
[2]
B. S. Lane, K. Burgess, Chem. Rev. 2003, 103, 2457Ϫ2473.
[3] [3a]
J. Brinksma, R. Hage, J. Kerschner, B. L. Feringa, Chem.
Commun. 2000, 537Ϫ538. [3b] R. Hage, J. E. Iburg, J. Kershner,
J. H. Koek, E. L. M. Lempers, R. J. Martens, U. S. Racherla,
S. W. Russell, T. Swarthoff, M. R. P. van Vliet, J. B. Warnaar,
L. van der Wolf, B. Krijnen, Nature 1994, 369, 637Ϫ639.
[4]
M. C. White, A. G. Doyle, E. N. Jacobsen, J. Am. Chem. Soc.
2001, 123, 7194Ϫ7195.
K. Kamata, K. Yonehara, Y. Sumida, K. Yamaguchi, S. Hi-
[5]
kichi, N. Mizuno, Science 2003, 300, 964Ϫ966.
M. Merkx, D. A. Kopp, M. H. Sazinsky, J. L. Blazyk, J. Muller,
[6]
S. J. Lippard, Angew. Chem. Int. Ed. 2001, 40, 2782Ϫ2807.
M. D. Wolfe, J. V. Parales, D. T. Gibson, J. D. Lipscomb, J.
[7]
Biol. Chem. 2001, 276, 1945Ϫ1953.
A. Karlsson, J. V. Parales, R. E. Parales, D. T. Gibson, H. Ek-
[8]
lund, S. Ramaswamy, Science 2003, 299, 1039Ϫ1042.
[9]
R. M. Burger, Chem. Rev. 1998, 98, 1153Ϫ1169.
[10] [10a] D. L. Boger, T. M. Ramsey, H. Cai, S. T. Hoehn, J. Stubbe,
[10b]
´
R. A. Leising, J. Kim, M. A. Perez, L. Que Jr., J. Am. Chem.
J. Am. Chem. Soc. 1998, 120, 9139Ϫ9148.
brook, S. A. Carr, M. A. Mentzer, E. C. Long, S. M. Hecht,
Inorg. Chem. 1987, 26, 3835Ϫ3836.
D. C. Heim-
Soc. 1993, 115, 9524Ϫ9530.
C. Kim, K. Chen, J. Kim, L. Que Jr., J. Am. Chem. Soc. 1997,
119, 5964Ϫ5965.
J. R. Lindsay Smith, P. R. Sleath, J. Chem. Soc., Perkin Trans.
2 1983, 1165Ϫ1169.
S. Miyajima, O. Simamura, Bull. Chem. Soc. Jpn. 1975, 48,
526Ϫ530.
[10c]
N. Murugesan, S. M.
Hecht, J. Am. Chem. Soc. 1985, 107, 493Ϫ500.
M. Costas, K. Chen, L. Que Jr., Coord. Chem. Rev. 2000,
200, 517Ϫ544.
K. Chen, L. Que Jr., Chem. Commun. 1999, 1375Ϫ1376.
K. Chen, L. Que Jr., J. Am. Chem. Soc. 2001, 123, 6327Ϫ6337.
[11]
[12]
[13]
Eur. J. Inorg. Chem. 2004, 846Ϫ856
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
855

M. Klopstra, G. Roelfes, R. Hage, R. M. Kellogg, B. L. Feringa
FULL PAPER
[44]
´
S. Menage, J. M. Vincent, C. Lambeaux, G. Chottard, A.
606Ϫ607. [50c] W. Nam, J. S. Valentine, J. Am. Chem. Soc. 1993,
115, 1772Ϫ1778.
Grand, M. Fontecave, Inorg. Chem. 1993, 32, 4766Ϫ4773.
T. Kojima, R. A. Leising, S. Yan, L. Que Jr., J. Am. Chem.
[51] [51a] S. J. Lange, H. Miyake, L. Que Jr., J. Am. Chem. Soc. 1999,
121, 6330Ϫ6331. [51b] M. P. Jensen, S. J. Lange, M. P. Mehn, E.
L. Que, L. Que Jr., J. Am. Chem. Soc. 2003, 125, 2113Ϫ2128.
[45]
Soc. 1993, 115, 11328Ϫ11335.
[46]
J. T. Groves, T. E. Nemo, J. Am. Chem. Soc. 1983, 105,
6243Ϫ6248.
[52] [52a] J. T. Groves, Z. Gross, M. K. Stern, Inorg. Chem. 1994, 33,
[47]
[52b]
H. Weiner, A. Trovarelli, R. G. Finke, J. Mol. Catal. A: Chemi-
5065Ϫ5072.
Z. Gross, S. Nimri, J. Am. Chem. Soc. 1995,
[52c]
cal 2003, 191, 217Ϫ252.
117, 8021Ϫ8022.
Z. Gross, J. Biol. Inorg. Chem. 1996, 1,
[48] [48a]
[52d]
J. Rudolph, K. L. Reddy, J. P. Chiang, K. B. Sharpless, J.
368Ϫ371.
Z. Gross, J. Org. Chem. 1997, 62, 5514Ϫ5521.
[48b]
[53]
[54]
Am. Chem. Soc. 1997, 119, 6189Ϫ6190.
A. C. Cope, J. M.
D. J. Hastings, A. C. Weedon, Can. J. Chem. 1991, 69,
Grisar, P. E. Peterson, J. Am. Chem. Soc. 1959, 81, 1640Ϫ1642.
1171Ϫ1182.
[49] [49a]
M. Tokunaga, J. F. Larrow, F. Kakiuchi, E. N. Jacobsen,
Bruker (2000). SMART, SAINT, SADABS, XPREP and
SHELXTL/NT. Smart Apex Software Reference Manuals.
Bruker AXS Inc., Madison, Wisconsin, USA.
Received September 25, 2003
[49b]
Science 1997, 277, 936Ϫ938.
B. Rickborn, in Comprehen-
sive Organic Synthesis (Ed.: B. M. Trost), Pergamon Press, Ox-
ford, 1991, pp. 733Ϫ775.
[50] [50a]
J. Bernadou, B. Meunier, Chem. Commun. 1998,
Early View Article
Published Online January 19, 2004
[50b]
2167Ϫ2173.
K. A. Lee, W. Nam, Inorg. Chem. 1998, 37,
856
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 846Ϫ856