Aromatic Hydroxylation at a Non-Heme Iron Center: Observed Intermediates and Insights into the Nature of the Active Species

Article abstract of DOI:10.1002/chem.201002577

Mechanism of substrate oxidations with hydrogen peroxide in the presence of a highly reactive, biomimetic, iron aminopyridine complex, [Fe ^II(bpmen)(CH₃CN)₂][ClO₄] ₂ (1; bpmen=N,N'-dimethyl-N,N'-bis(2-pyridylmethyl)ethane-1,2- diamine), is elucidated. Complex 1 has been shown to be an excellent catalyst for epoxidation and functional-group-directed aromatic hydroxylation using H₂O₂, although its mechanism of action remains largely unknown.1, 2 Efficient intermolecular hydroxylation of unfunctionalized benzene and substituted benzenes with H₂O₂ in the presence of 1 is found in the present work. Detailed mechanistic studies of the formation of iron(III)-phenolate products are reported. We have identified, generated in high yield, and experimentally characterized the key Fe^III(OOH) intermediate (I_max=560 nm, rhombic EPR signal with g=2.21, 2.14, 1.96) formed by 1 and H₂O₂. Stopped-flow kinetic studies showed that Fe^III(OOH) does not directly hydroxylate the aromatic rings, but undergoes rate-limiting self-decomposition producing transient reactive oxidant. The formation of the reactive species is facilitated by acid-assisted cleavage of the O-O bond in the iron-hydroperoxide intermediate. Acid-assisted benzene hydroxylation with 1 and a mechanistic probe, 2-Methyl-1-phenyl-2-propyl hydroperoxide (MPPH), correlates with O-O bond heterolysis. Independently generated Fe^IV=O species, which may originate from O-O bond homolysis in Fe^III(OOH), proved to be inactive toward aromatic substrates. The reactive oxidant derived from 1 exchanges its oxygen atom with water and electrophilically attacks the aromatic ring (giving rise to an inverse H/D kinetic isotope effect of 0.8). These results have revealed a detailed experimental mechanistic picture of the oxidation reactions catalyzed by 1, based on direct characterization of the intermediates and products, and kinetic analysis of the individual reaction steps. Our detailed understanding of the mechanism of this reaction revealed both similarities and differences between synthetic and enzymatic aromatic hydroxylation reactions.

Full text of DOI:10.1002/chem.201002577

FULL PAPER
DOI: 10.1002/chem.201002577
Aromatic Hydroxylation at a Non-Heme Iron Center: Observed
Intermediates and Insights into the Nature of the Active Species
Olga V. Makhlynets and Elena V. Rybak-Akimova*[a]
Abstract: Mechanism of substrate oxi-
dations with hydrogen peroxide in the
presence of a highly reactive, biomim-
etic, iron aminopyridine complex, [FeII-
the key FeIII
G
heterolysis. Independently generated
(lmax =560 nm, rhombic EPR signal
with g=2.21, 2.14, 1.96) formed by 1
and H2O2. Stopped-flow kinetic studies
FeIV=O species, which may originate
III
ꢀ
from O O bond homolysis in Fe -
A
(OOH), proved to be inactive toward
A
N
G
showed that FeIII
aromatic substrates. The reactive oxi-
dant derived from exchanges its
rectly hydroxylate the aromatic rings,
but undergoes rate-limiting self-decom-
position producing transient reactive
oxidant. The formation of the reactive
species is facilitated by acid-assisted
1
N
oxygen atom with water and electro-
philically attacks the aromatic ring
(giving rise to an inverse H/D kinetic
isotope effect of 0.8). These results
have revealed a detailed experimental
mechanistic picture of the oxidation re-
actions catalyzed by 1, based on direct
characterization of the intermediates
and products, and kinetic analysis of
the individual reaction steps. Our de-
tailed understanding of the mechanism
of this reaction revealed both similari-
ties and differences between synthetic
and enzymatic aromatic hydroxylation
reactions.
Complex 1 has been shown to be an
excellent catalyst for epoxidation and
functional-group-directed aromatic hy-
droxylation using H2O2, although its
mechanism of action remains largely
unknown.[1,2] Efficient intermolecular
hydroxylation of unfunctionalized ben-
zene and substituted benzenes with
H2O2 in the presence of 1 is found in
the present work. Detailed mechanistic
cleavage of the O O bond in the iron–
hydroperoxide intermediate. Acid-as-
sisted benzene hydroxylation with 1
and a mechanistic probe, 2-Methyl-1-
phenyl-2-propyl
(MPPH), correlates with O O bond
studies of the formation of ironACTHNUTRGNE(UNG III)–
Keywords: hydrogen peroxide
hydroxylation · iron · kinetics ·
oxidation
phenolate products are reported. We
have identified, generated in high
yield, and experimentally characterized
Introduction
characterizations of several high-valent iron intermedi-
ates,[3–5] as well as discovery of synthetically useful, regiose-
lective epoxidations and hydroxylations catalyzed by bio-
mimetic iron complexes.[6] One of the most successful olefin
epoxidation catalysts is an iron(II) complex of a tetradentate
Development of regio- and stereoselective catalytic oxida-
tion of organic substrates is an important goal for organic
synthesis. Biomimetic oxidations are particularly attractive,
because they rely on cheap, nontoxic reactants (usually, O2
or H2O2 as oxidants, and Fe, Cu, or Mn complexes as cata-
lysts). Developing the chemistry of non-heme iron oxida-
tions proved to be intellectually rewarding and productive,
and impressive recent progress includes crystallization and
aminopyridine ligand: [FeII
ACHTUNGTREN(UNNG bpmen)ACHTNURTGEG(NNUN CH3CN)2]ACHTUNGRTEN(NUGN ClO4)2 (1;
bpmen=N,N’-dimethyl-N,N’-bis(2-pyridylmethyl)ethane-1,2-
diamine) that converts olefins into epoxides with high selec-
tivity and efficiency with hydrogen peroxide as the oxi-
dant.[1] It has also been shown to be a selective catalyst for
[7]
ꢀ
aliphatic C H oxidation, and regioselective aromatic hy-
droxylation.[2,8] Over the past decade the identity of the re-
active intermediate in the catalytic oxidation promoted by 1
became a subject of interest. Although there is no general
mechanism for a non-heme systems that activate oxygen,[9]
[a] O. V. Makhlynets, Prof. Dr. E. V. Rybak-Akimova
Department of Chemistry, Tufts University
62 Talbot Ave., Medford, MA 02155 (USA)
Fax : (+1)617-627-3443
ACTHNUTRGNEUNG
several intermediates, including FeIII(OOH), FeIV=O, and
E-mail: erybakak@tufts.edu
FeV=O, have been proposed as active species in catalytic
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002577.
cycles of enzymes and enzyme models.[6,10] A generally ac-
Chem. Eur. J. 2010, 16, 13995 – 14006
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
13995

cepted scheme of the reaction of 1 and H2O2 includes for-
which places two critical components of aromatic hydroxyl-
ation in separate molecules, decouples the reactivity of the
aromatic rings from the reactivity of carboxylic acids. Our
kinetic studies and isotope-labeling experiment show that
mation of the FeIII
ACHTUNGTRENNUNG(OOH) intermediate that undergoes het-
V
ꢀ
erolytic O O bond cleavage and yields active Fe =O
(Scheme 1B). The recent progress in understanding of the
FeIII
ACTHNUTRGNE(UNG OOH) is not the active ox-
idant in hydroxylation reaction,
but it produces the reactive
high-valent species in the rate-
limiting step. The reactive spe-
cies formation is facilitated by
protonation of the terminal
oxygen in FeIII
ACTHNUGRTENUNG(OOH), which
leads to heterolytic cleavage of
ꢀ
the O O bond and formation
of a putative FeV=O species.
Results and Discussion
Hydroxylation of benzene by
hydrogen peroxide in the pres-
ence of 1: The reactivity of 1 in
Scheme 1. A) ortho- and/or ipso-hydroxylation of benzoic acids with H2O2 promoted by 1; B) possible reactivi-
ty modes of FeIII(OOH) intermediate: FeIII
ACHTUNGTRENNUNG ACHTUGNTREN(NUGN OOH) can either directly transfer oxygen to substrate, or decom-
pose into a high-valent iron–oxo species that reacts with the substrate.
hydroxylation of aromatic hy-
drocarbons with hydrogen per-
oxide at room temperature was
mechanism of oxygen activation at 1 was based on indirect
explored in the present work. Benzene reacts with hydrogen
peroxide in the presence of 1 to form a blue species
(lmax=650 nm), which later decays (Figure 1). Quenching the
evidence and no reaction intermediates were reliably char-
acterized, although Talsi et al. detected FeIII
ACTHUNRTGNE(GNU OOH) by EPR
spectroscopy at ꢀ608C in very low yield.[11] In this work we
followed rapid reactions of 1 with H2O2 and found condi-
tions to generate FeIII
ACTHNUTRGNEU(GN OOH) in a high yield and character-
ized this intermediate by EPR spectroscopy. Furthermore,
we have tested the reactivity of this novel intermediate.
Aromatic hydroxylation presents an excellent opportunity
to gain detailed insights into the reactivity of complex 1. We
have recently reported that the hydroxylation of the aromat-
ic ring of substituted benzoic acids occurs exclusively in the
vicinity of the anchoring carboxylate functional group lead-
ing to ortho- or ipso-hydroxylated products (Scheme 1A).[2]
This carboxylic acid directed regioselectivity implies involve-
ment of a highly reactive, metal-based oxidant that attacks
the aromatic ring next to the directing group. One specific
question that needs to be addressed is the role of carboxylic
acid functionality in regioselective aromatic hydroxylations.
In addition to being an anchoring, directing group that coor-
dinates to the metal center of 1, carboxylic acid can also
serve as a source of protons. Other carboxylic acids, such as
Figure 1. UV/Vis spectral changes for the reaction of 1, benzene and
H2O2 in acetonitrile at 208C ([1]=0.5 mm, [H2O2]=5 mm, 200 equiv of
benzene vs. 1). Solid lines represent spectra of 1 (time 0 s) and a product
[(bpmen)FeIII-OPh]2+ (run time 90 s). Inset: kinetic trace at 650 nm
showing accumulation of [(bpmen)FeIII-OPh].
acetic acid, were used to improve performance of 1 as cata-
[1,7]
ꢀ
lyst of olefin epoxidation and aliphatic C H oxidation.
Que and co-workers proposed that coordinated carboxylic
reaction followed by a workup (see Experimental Section)
and GCMS identified the oxidation product as phenol. Co-
ordination of the deprotonated phenol to ironACTHNGUTERNN(UG III) in the
oxidized form of 1 generates colored products, similarly to
the formation of intensely colored salicylates upon hydroxyl-
ation of benzoic acids promoted by 1.[2,8] Screening and opti-
mization of reaction conditions revealed that the best yields
of hydroxylated products (phenolates) were obtained in ace-
ꢀ
acid promotes O O bond heterolysis of the Fe
III(OOH)
with formation of an FeV=O species.[12] In order to better
understand the mechanism of oxidations with hydrogen per-
oxide promoted by 1, we now report a detailed study of hy-
droxylation of non-coordinating aromatic hydrocarbons, and
the effects of externally added carboxylic acids on the rates
and mechanisms of aromatic hydroxylation. This approach,
13996
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13995 – 14006

Aromatic Hydroxylation
FULL PAPER
tonitrile at room temperature, with excess of H2O2 (10 equiv
vs. 1) and benzene (300 equiv vs. 1). Strongly coordinating
solvents, such as methanol and DMSO, suppress benzene
hydroxylation promoted by 1, presumably due to the substi-
tution of labile acetonitrile in the coordination sphere of
iron and blocking an access of the oxidant and the substrate
to the iron center. Excess of hydrogen peroxide dramatically
improves the conversion of benzene into phenol, but also
causes subsequent oxidation of phenol into dihydroxyben-
zenes (Figure S1 in the Supporting Information). We utilized
small excess of hydrogen peroxide (3 equiv) for kinetic and
deuterium retention studies, for which formation of dihy-
droxybenzene would make interpretation of results more
difficult, and larger quantities (10 equiv) to characterize
metal-based intermediates using EPR spectroscopy and
stopped-flow spectrophotometry. Time-resolved UV/Vis and
GCMS data show that hydroxylation reaction is complete in
5 min; GC and spectrophotometric yields of phenol match
closely (Figures S2 and S3 in the Supporting Information).[13]
Although the reaction is rapid and high-yielding, efficient
catalysis is precluded by strong coordination of the pheno-
in the presence of acetic acid: the yield of aromatic hydrox-
ylation product increased, while the yield of methylene hy-
droxylation product decreased (Figure S4 in the Supporting
Information). Presumably, the nature of the reactive inter-
mediates and/or their reactivity can be altered in presence
of carboxylic acids.
Both steric and electronic effects in aromatic hydroxyl-
ation by 1/H2O2 differ from non-selective reactivity of hy-
droxyl radicals (for example, nitrobenzene affords nitrophe-
C
nols in reactions with HO generated by photolysis of a-azo-
hydroperoxide).[14] Moreover, hydroxylation of chloroben-
zene with 1/H2O2 in the presence of a radical trap (galvin-
A
phenol. These observations suggest involvement of metal-
based intermediates in non-radical aromatic hydroxylation
with 1/H2O2. A metal-based intermediate was also implicat-
ed in a related aromatic hydroxylation of benzoic acids with
1/H2O2.[2]
Kinetic and mechanistic studies of benzene hydroxylation:
The formation of a phenolate upon mixing of 1 with ben-
zene/H2O2 appears as a two-exponential process (Figure S5
in the Supporting Information). The effective pseudo-first-
order rate constants k1obs and k2obs do not depend on concen-
tration of 1 indicating the reactionꢁs first-order in 1. The
first step of hydroxylation reaction was found to be first
order in H2O2, while the second step is independent of
[H2O2] (Figure S6 in the Supporting Information). Neither
k1obs nor k2obs depend on substrate concentration (Table S2
in the Supporting Information), but the spectroscopic yield
of phenol is proportional to the concentration of H2O2 (Fig-
ure S5 in the Supporting Information) and benzene. Further-
more, the rate of hydroxylation is not affected by the nature
of substrates (Table S3 in the Supporting Information), but
the spectroscopic yields of phenolate products under other-
wise identical conditions generally increase for electron-rich
substrates (Figure S7 in the Supporting Information). These
results suggest a mechanistic hypothesis: the first reaction
step affords an intermediate that decomposes in a subse-
quent rate-limiting step producing an active oxidant. To test
this assumption, we examined individual reaction steps and
identified reactive intermediates.
late products to the ironACHTNUTRGNE(NUG III) center.
The scope of the hydroxylation reaction: Although general,
aromatic hydroxylation catalyzed by 1 is sensitive to the
nature of substrate. Without an anchoring group it affords a
mixture of o-, m- and p-substituted phenols (Scheme 2 and
Table S1 in the Supporting Information). Sterically hindered
substrates (e.g., 1-tert-butyl-3,5-dimethylbenzene) and sub-
strates with strong electron-withdrawing groups (e.g., nitro-
benzene) give very low yield of phenol (Table S1 in the Sup-
porting Information).
Reaction of complex 1 with H2O2—formation of the FeIII-
Scheme 2. The scope of hydroxylation reaction.
AHCTUNGTREG(NNNU OOH) intermediate: Stopped-flow studies of the direct re-
action between 1 and H2O2 under optimal hydroxylation
conditions (room temperature, acetonitrile) identified a new
purple intermediate with lmax =560 nm (species 2, Figure 2)
that is short-lived at room temperature (half-life time 30 s).
The optical spectrum of 2 is similar to the spectra of known
The 1/H2O2 system favors aromatic hydroxylation over
hydroxylation of methyl group (phenols/benzyl alcohol
ꢁ11:1 in toluene hydroxylation). However, aliphatic chain is
preferentially hydroxylated when a methylene group is
available (Figure S4 in the Supporting Information). Addi-
tions of acetic acid can modulate both the activity and the
selectivity of oxidation reactions mediated by 1 and related
aminopyridine iron complexes. For example, olefin epoxida-
tion is strongly favored in the presence of acetic acid, while
dihydroxylation dominates in acid-free reactions.[1,12] Simi-
larly, we observed alteration of ethylbenzene hydroxylation
mononuclear FeIII
ACTHUNTGRENNG(U OOH) complexes, which typically have
an intense band at ꢁ500–550 nm (eꢁ1000mꢀ1 cmꢀ1).[15–22]
FeIII
ACHTNUTRGNEUNG(OOH) was proposed to be a plausible reaction inter-
mediate in reactions of 1 with H2O2;[8,12,23] however, owing
to the low yield of the intermediate (<3%) its absorption
spectrum could not be recorded.[11] Our stopped-flow experi-
ments confirmed that the addition of hydrogen peroxide to
Chem. Eur. J. 2010, 16, 13995 – 14006
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13997

E. V. Rybak-Akimova and O. V. Makhlynets
Figure 3. EPR spectrum (120 K) of FeIII
of FeIII
(OOH) vs. initial concentration of iron was determined using Cu-
AHCTUNTGREG(NUNN ClO4)2 as an external standard. Signals at 2.41 and 1.90 correspond to 3,
ACHTUNGTRNE(NUNG OOH) in acetonitrile. 25% yield
AHCTUNGTRENNUNG
Figure 2. Time-resolved UV/Vis spectra of the FeIII
ACTHNUTRGNE(NUG OOH) formation at
208C in acetonitrile ([1]=1 mm, [H2O2]=10 mm). Inset: kinetic trace at
which accompanies 2 at all temperatures.
560 nm. Solid lines are spectra of 1 and FeIII
ACTHNUTRGNE(UNG OOH) at its maximum for-
mation.
of total iron) is typical of low-spin ironACTHNUTRGNE(NUG III)–hydroperoxo
1 at lower temperatures (ꢀ80 to ꢀ308C) affords only yellow
species 3, and no purple species was detected (Figure S8 in
the Supporting Information). Unlike the majority of reactive
intermediates, that are generated in higher yields at low
temperature, the yield of 2 at the time of its maximum accu-
mulation increased as temperature increased from ꢀ208C to
+208C (Figure S9 in the Supporting Information), although
decomposition of 2 also accelerated with temperature. The
transient nature of this intermediate may have prevented its
observation and trapping in the past.
species (with characteristic range of g values of 1.93–2.26)[24]
and thus is assigned to 2. Simultaneous formation of two
low-spin iron
when FeII–BLM (BLM=bleomycin) reacted with O2,[25,26]
first signals of FeIII
(OOH)–BLM grew in (g=2.26, 2.17 and
ACHTUNGTRENUN(NG III) species has been previously observed
AHCTUNGTRENNUNG
1.94) and then the decay product FeIII–BLM accumulated
(g=2.45, 2.18 and 1.89). Complexes with aminopyridyl li-
gands similar to bpmen (L) also react with hydrogen perox-
ide to produce a mixture of low-spin iron species (Table S4
in the Supporting Information); one set of signals in EPR
The plot of the observed rate constants versus hydrogen
peroxide concentration under pseudo-first-order conditions
(ca. 10-fold excess H2O2 with respect to 1) is a straight line
with a nearly zero intercept (Figure 5,S10), indicating first-
order in H2O2 and the overall mixed second-order for the
spectrum corresponds to FeIII
ACTHNGUTERNN(UG OOH) and the other set was
assigned to [FeIII(L)X]n+ (n=2 or 3; X=Cl, OH, H2O,
OMe, HOMe),[18,19,27] which presumably serves as a presur-
sor to FeIII
ACHTUNGTRENNUNG
formation of FeIII
N
ACHTUNGTRENNUNG
vðFeIIIOOH formationÞ ¼ k½1ꢂ½H2O2ꢂ
AHCTUNGTRENNUNG
N
N
ACHTUNGTRENNUNG
Relatively high activation enthalpy (DH° =55.0 kJmolꢀ1,
Figure S11 in the Supporting Information) and large nega-
tive activation entropy (DS° =ꢀ29.3 Jmolꢀ1 Kꢀ1, Figure S12
in the Supporting Information) are consistent with a bimo-
lecular rate-limiting step, although the activation parameters
for a multistep conversion of 1 into 2 are likely to be com-
posite values, and straightforward interpretation may be
misguided.
N
ACHTUNGTRENNUNG
To directly probe whether [FeIII
ACTHNUTRGENN(UG bpmen)ACHTUNGTRENNUNG(OOH)-
Intermediates 2 and 3 were characterized by EPR spec-
troscopy and mass spectrometry. The main product of the
reaction between 1 and H2O2 at ꢀ308C was a low-spin
iron
N
in the Supporting Information). Most likely, an oxidation of
1 without subsequent coordination of hydroperoxide occurs
under these conditions. Freezing the reaction solution at the
maximum accumulation of 2 (room temperature, Figure 2)
gives a sample with two species: g values 2.21, 2.14, 1.96 and
2.41, 2.17 and 1.90 (Figure 3); the additional signal at 4.24
Iron–hydroperoxide is often considered to be a reaction in-
termediate in nonheme iron systems[15,16,28,29] that may either
directly transfer an oxygen atom to the substrate, or decom-
pose into a high-valent iron–oxo species that reacts with the
substrate.[12,30,31] However, the experimental data on the re-
activity of FeIII
ACTHNUGRTENUNG(OOH) are limited, and hardly support a uni-
belongs to a small amount of a high-spin rhombic iron
product. The set of signals with g=2.21, 2.14, 1.96 (ca. 30%
(III)
form view on their ability to directly oxidize substrates. For
example, Nam et al. tested the ability of several synthetic
13998
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13995 – 14006

Aromatic Hydroxylation
FULL PAPER
a two exponential process when 1 is mixed with benzene/
H2O2, in which k1obs corresponds to the formation of FeIII-
AHCTUNGTERN(GNUN OOH) (Figure 5) and k2obs is attributed to the hydroxyl-
ation of benzene. The rate constant of phenol formation
(k2obs in Figures 1 and 5) and the rate constant of FeIII-
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
series of double-mixing experiments.
The rate of phenolate accumulation did not depend on
the concentration of benzene added to a pre-generated FeIII-
AHCTUNGTREG(NNUN OOH) (Table S5 in the Supporting Information). Rate of
Figure 4. Time-resolved UV/Vis spectra of the reaction of FeIII
ACTHNUTRGNE(NUG OOH)
with benzene in acetonitrile at 208C ([1]=0.5 mm, [H2O2]=5 mm,
200 equiv of benzene vs. 1). Solid lines represent spectrum of 2 (lmax
hydroxylation also does not depend on the nature of sub-
strate: when different substrates (benzene, chlorobenzene,
toluene) were added to the pre-generated 2, the rate con-
stants of FeIII–phenolate formation were nearly identical
(Table S3 in the Supporting Information). Moreover, the
rate of aromatic hydroxylation did not depend on the con-
=
560 nm) at its maximum formation (age time 7 s) and a spectrum of a
product [(bpmen)FeIII-OPh] (lmax =650 nm, run time 80 s).
centration of excess H2O2: FeIII
ACTHNUTRGNEUNG(OOH) was generated by
mixing 1 and variable amounts of H2O2 (3, 10, and 20 equiv
vs. 1), followed by the addition of benzene at the maximum
accumulation of FeIII
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
the aromatic ring of the substrate, but this species decays
into another species (presumably, FeV=O or FeIV=O), which
rapidly hydroxylates the aromatic ring (Scheme 5).
Acetic acid facilitates FeIII
ACTHNUTRGNE(UNG OOH) decay in a concentra-
tion-dependent manner: larger amounts of acid lead to
faster hydroperoxide decomposition (Figure S16 in the Sup-
porting Information). Additionally, acetic acid accelerates
the rate of FeIII–phenolate formation (Figure 6), and the
rate constants of phenolate formation and those of self-
decay of 2 in the presence of one equivalent of acetic acid
are comparable (Figure S17 in the Supporting Information).
These data indicate that acetic acid accelerates the decay of
Figure 5. Rates of formation of phenol, FeIII
ACHTUNGTRENNUNG
(OOH) and decay of FeIII
ACHTUNGTERNNUNG -
constant of hydroxylation was determined from a single mixing experi-
ment, in which 1 was mixed with benzene and H2O2 ([1]=0.5 mm,
300 equiv benzene vs. 1). Kinetic traces at 650 nm were fitted in Kinetic
Studio program using A!B!C model. Formation and decay of FeIII
-
ACHTUNGTRENNUNG(OOH) was monitored at 560 nm after 1 (0.5 mm) was mixed with H2O2;
kinetic traces at 560 nm were fitted using A!B!C model, in which the
first process represents formation of FeIII
(OOH) and the second process
is a decay of FeIII
(OOH).
FeIII
ACTHNUTRGNE(NUG OOH) into a reactive intermediate responsible for aro-
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
iron–hydroperoxo complexes to oxidize sulfides or olefins,
and concluded that this intermediate is a sluggish oxidant.[32]
On the other hand, Solomon et al. reported that FeIII
ACTHNUTRGNE(NUG OOH)
in activated bleomycin (ABLM) directly attacks DNA.[26]
Kinetic data for benzene hydroxylation compared with ki-
netics of the reaction of 1 with H2O2 (Figure S15 in the Sup-
porting Information) suggest that 2 does not attack aromatic
ring, but decomposes into another, more active intermediate
(Scheme 5). The rate constant of the formation of FeIII-
ACHTUNGTRENNUNG(OOH) from 1 and H2O2 is comparable to the observed rate
constant of the first process in benzene hydroxylation stud-
ies (k1obs, Figure S6 in the Supporting Information). Both
rate constants are directly proportional to [H2O2]. Since ab-
Figure 6. Rate constant of phenolate formation is directly proportional to
[CH3COOH] ([1]=0.5 mm, [H2O2]=5 mm, 560 equiv of benzene vs. 1,
208C, acetonitrile).
sorption bands of FeIII(OOH) and FeIII–phenolates (lmax
ACHTUNGTERNUNNG =
560 and 650 nm, respectively) partially overlap, we observe
Chem. Eur. J. 2010, 16, 13995 – 14006
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13999

E. V. Rybak-Akimova and O. V. Makhlynets
matic hydroxylation. Addition of acetic acid is known to in-
ꢀ
crease catalytic activity of 1 in epoxidation and aliphatic C
H oxidation reactions.[1,7] We have also observed a some-
what increased yield of aromatic hydroxylation products in
reactions of benzenes with 1/H2O2 in the presence of acetic
acid (Figure S18 in the Supporting Information).
Products of decomposition of FeIII
hydroxylation reaction: Kinetic data suggest that decompo-
sition of FeIII
(OOH) yields an active oxidant and acetic acid
ACHTUNGTREN(NUNG OOH) and their role in
ACHTUNGTRENNUNG
accelerates this process. Therefore we trapped and charac-
terized the decay product of FeIII
tivity toward aromatic substrates.
ACHTUNGERTN(NUNG OOH) and probed its ac-
We followed the decay of 2 by UV/Vis spectrophotometry
and observed the formation of a green species 4 with lmax
Figure 8. Spectra of green species 4 obtained upon mixing of 1 with IBX
ester ([1]=1 mm; [IBX ester]=3.8 mm, acetonitrile, 108C) and 1 with
H2O2/CH3COOH ([1]=1 mm, [H2O2]=10 mm, [CH3COOH]=0.5 mm,
acetonitrile, 208C).
=
740 nm (Figure S19 in the Supporting Information). This
near-IR chromophore closely resembles low-spin L–FeIV=O
III
ꢀ
complexes generated by O O bond homolysis of L–Fe -
A
(OOH) intermediates supported by several polyamine- or
room temperature shows peaks at m/z=401, 425, 442 that
aminopyridine ligands (L) analogous to bpmen,[33] suggest-
ing that 4 can also be formulated as an FeIV=O intermediate.
This formulation is consistent with the EPR-silent nature of
the green intermediate 4: it was found that 2 decayed over
70 s (kꢁ0.01 sꢀ1, RT) at room temperature without forma-
tion of any new EPR-active species (Figure S20 in the Sup-
porting Information).
correspond to [Fe
(bpmen)(O)+OAc]+
ACHTUNGTNERNUNG , [FeCAHTUNGRTNEN(UGN bpmen)+
ClO4]+, [Fe(bpmen)(OH)+ClO4]+ respectively (Figure S21
AHCTUNGTRENNUNG
in the Supporting Information). Preliminary Mçssbauer
studies of the sample prepared by adding peracetic acid di-
rectly to 1 also confirms that the green species 4 has an
FeIV=O center.[34]
Additionally, we used an alternative, H2O2-free method to
directly generate the green species 4 from 1 and oxygen
atom donors. The isopropyl ester of 2-iodoxybenzoic acid
(IBX ester)[36] was recently developed as a stable, highly
soluble source of oxygen atoms in organic oxidations. This
reagent was also successfully used in metal-mediated oxida-
tions catalyzed by phthalocyanin complexes.[37] We applied
the IBX ester as an oxygen-atom transfer reagent in reac-
tions that generate high-valent iron(IV)–oxo species. Upon
mixing 1 and IBX ester (0.5–4 equiv vs. 1) no purple species
was observed, but instead a green species 4 formed, with an
optical spectrum very similar to the spectra of the species
generated from 1, H2O2, and HOAc (Figure 8). The differ-
ence in spectra of the species 4 derived from 1/IBX ester
and 1/H2O2/CH3COOH can be attributed to the binding of
HOAc or OAc to the FeIV=O center.[38]
The chemical nature of 4 suggested alternative methods
of generating this intermediate. Similarly to literature prece-
dents,[12] 4 can be obtained faster and in higher yield from
the reaction of 2 with acetic acid (Figure 7), or by adding a
mixture of H2O2 and acetic acid directly to 1 (Figure 8).
Acetic acid, added directly into EPR tube to a pre-generat-
ed 2, accelerated decomposition of 2 into an EPR-silent spe-
cies 4, while 3 (which initially accompanied 4—g values at
2.41, 2.17 and 1.90) remained intact (Figure 7, inset). The
ESI-MS spectrum of the mixture of 1/H2O2/CH3COOH at
To test the reactivity of the iron–oxo center, we generated
bpmen–FeIV=O by mixing 1, H2O2, and acetic acid at 208C.
After 4 was formed in its maximum yield (60 s), benzene
was added; however, only self-decay of 4 was observed. In a
similar experiment, 4 was prepared by mixing 1 and IBX
ester at 108C, addition of benzene (280 equiv vs. iron) did
not result in the formation of the FeIII–phenolate complex.
It can be concluded that the FeIV=O intermediate 4 is
unable to hydroxylate the aromatic ring, and cannot act as
kinetically competent intermediate in aromatic hydroxyl-
ation with H2O2 promoted by 1. We also used the stopped-
flow methodology to measure rates of FeIV=O decay in the
presence of toluene and without toluene and found that
they are comparable (Figure S22 in the Supporting Informa-
tion); therefore, bpmen–FeIV=O does not efficiently hydrox-
ylate benzylic CH3 groups.
Figure 7. Stopped-flow time-resolved UV/Vis spectra of the reaction of
FeIII(OOH) with one equivalent of acetic acid. FeIII
ACHTUNGTRENNUNG ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
addition of acetic acid (1 equiv vs. iron, dashed line) followed by slow
(over 80 s) accumulation of FeIV=O (solid line, lmax =740 nm).[35] Inset
shows EPR spectra of FeIII
ACHTUNGTRENNUNG
14000
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13995 – 14006

Aromatic Hydroxylation
FULL PAPER
ꢀ
Mechanistic insights into O O
bond cleavage in FeIII peroxides
supported by bpmen: Isotope-
labeling experiments (18O/16O)
are often useful in determining
the mechanistic pathways of
ꢀ
metal-assisted O O bond cleav-
age. Incorporation of oxygen
from water into the oxidation
product requires the formation
of FeIV=O or FeV=O species
and 16O/18O exchange with sol-
vent water prior to reaction
Scheme 3. Reaction of MPPH with iron center and products derived from MPPH.
with substrates.[39,40] In our
system, hydroxylation experi-
ments with labeled hydrogen peroxide (H218O2) in the pres-
ence of water (H216O) show the incorporation of 16O into
phenol product: 19% of phenol contained 16O and 81% of
phenol contained 18O. Hydroxylation of benzene with
H216O2 in the presence of H218O led to 78% H16O-C6H5 and
22% of H18O-C6H5. The results of isotope-labeling studies
are inconsistent with the mechanisms involving a direct
Pure recrystallized MPPH reacted with 1/benzene under
aerobic conditions to yield mainly benzaldehyde, which indi-
IV
ꢀ
cated the O O bond homolysis and the formation of Fe =
ꢀ
O (Scheme 3); the amount of MPPol, an indicator of O O
bond heterolysis, was small. Yield of phenol was low under
these conditions (Figure 9a). In contrast, oxidation of ben-
ACHTUNGTRENNUNG
attack of FeIII(OOH) on the aromatic ring, because FeIII-
ACHTUNGTRENNUNG
studies showed some loss of the label when H218O2 was used
as oxygen source for cis-dihydroxylation of naphthalene by
naphthalene 1,2-dioxygenase,[41] thus implicating oxygen ex-
change between the active intermediate and water, and sug-
gesting a reaction pathway that invokes FeV=O(OH) species.
In another example,[31] a cis-dihydroxylation reaction of al-
kenes with hydrogen peroxide in the presence of H218O cata-
lyzed by [FeII (CH3CN)2]2+ (tpa=tris(2-pyridylmethyl)-
ACHTUNGTRENNUNG(tpa)ACHTUNGTRENNUNG
amine) showed incorporation of 18O into product, which was
rationalized by water-assisted formation of FeV=O(OH). Co-
ordination of acetic acid to the iron–oxo moiety interfered
with this water-assisted 18O/16O exchange, decreasing the in-
Figure 9. Product yields (in mol per 1 mol of Fe) and effects of acetic
acid in aerobic reaction of benzene with 1/MPPH (Scheme 3): a) [1]=
2 mm, 280 equiv of benzene vs. 1, 1 equiv of MPPH; b) [1]=2 mm,
280 equiv of benzene vs. 1, 1 equiv of MPPH, 0.5 equiv of acetic acid;
control experiments:[46] c) MPPH in acetonitrile (same amount as in a
and b); d) MPPH and acetic acid (same amounts as in b). All samples
were stirred for 30 min, then acetylated, products extracted with di-
chloromethane and analyzed by GCMS. Error bars show standard devia-
tion.
corporation of the 18O from water in the products;[12,31]
a
similar effect was also observed in our experiments on ben-
zene hydroxylation with 1/H2O2/HOAc.
The 18O/16O scrambling in aromatic hydroxylations with
H2O2 promoted by 1 implies that the observed peroxo inter-
mediate, FeIII
ACHTUNGTRENNUNG(OOH), is not the sole oxidant in this reaction
ꢀ
and thus should undergo either a heterolytic O O bond
V
ꢀ
cleavage that affords Fe =O, or a homolytic O O bond
zene by MPPH in the presence of 1 and 0.5 equivalents of
acetic acid resulted in significant increase in the yields of
both MPPol and phenol (Figure 9b). These results are con-
cleavage to produce FeIV=O. In similar reactions, organic
peroxide MPPH (2-methyl-1-phenyl-2-propyl hydroperox-
ide) has been widely used as a mechanistic probe to distin-
ꢀ
sistent with O O bond heterolysis as a major pathway that
guish homolytic versus heterolytic cleavage of the alkylper-
generates a competent oxidant for benzene, implying that
[42–45]
oxide O O bond.
Homolysis of RO O bond would
FeV=O is the active species in the hydroxylation reaction.
ꢀ
ꢀ
IV
ꢀ
C
generate a Fe =O species and a transient RO radical that
decays into benzyl radical, eventually yielding benzyl alco-
hol under anaerobic conditions, or benzaldehyde in the pres-
ence of O2 (Scheme 3). On the other hand, heterolytic cleav-
age of the O O bond would generate Fe =O and 2-methyl-
Acetic acid catalyzes heterolytic O O bond cleavage and
the formation of FeV=O, therefore acetic acid dramatically
increased the yield of MPPol and the yield of aromatic hy-
droxylation product when MPPH is used as oxidant.
V
ꢀ
1-phenyl-2-propyl alcohol (MPPol). Analysis of organic
products provides the information on the mechanism of O
H/D kinetic isotope effects in substrate oxidations with 1/
H2O2: The mechanisms of interactions between the oxidant
and the organic substrate can be evaluated by comparing
ꢀ
O bond cleavage in oxidations with MPPH.
Chem. Eur. J. 2010, 16, 13995 – 14006
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14001

E. V. Rybak-Akimova and O. V. Makhlynets
the relative rates of hydroxylation of deuterated and unla-
beled substrates. While detailed calculations are necessary
for thorough interpretation of H/D kinetic isotope effects,
some qualitative conclusions can be derived from the direct
ꢀ
analysis of the experimental data. In particular, C H bond
breakage (e.g., in a hydrogen-atom abstraction or direct in-
sertion, Scheme 4 A,B) typically results in normal kinetic
isotope effects (KIE>1), but electrophilic attack on the aro-
matic ring accompanied by sp2/sp3 rehybridization often
leads to small inverse kinetic isotope effects (Scheme 4
C,D).[47]
Benzene hydroxylation in the presence of 1/H2O2 resulted
in inverse KIE kH/kD =0.8. The inverse kinetic isotope effect
in benzene hydroxylation indicates a change in hybridization
at a carbon bound to deuterium (sp2!sp3), which is consis-
tent with an electrophilic addition to the aromatic ring or
with an intermediate formation of epoxide.[48,49] The inverse
H/D kinetic isotope effect is inconsistent with an abstrac-
tion–recombination sequence, which therefore can be ex-
cluded for aromatic hydroxylation with 1/H2O2. Also tolu-
ene hydroxylation with H2O2 in the presence of 1 gave ki-
netic isotope effects of 0.7 for
Scheme 4. Mechanism of aromatic hydroxylation shown for 3-2H1-chloro-
benzene. A) Direct insertion; B) hydrogen abstraction; C) addition/rear-
rangement; D) formation of epoxide.
the aromatic ring hydroxylation
and 1.3 for the methyl group
hydroxylation. These data indi-
cate that a hydrogen-atom ab-
straction is probably involved
in benzylic hydroxylation of tol-
uene, but not in the hydroxyl-
ation of the aromatic ring. Very
similar values of H/D KIE in
toluene hydroxylation were ob-
served in the presence of acetic
acid, indicating an electrophilic
nature of the oxidant that at-
tacks the aromatic ring under
various reaction conditions. The
differences in isotope effects
observed for hydroxylation of
the aliphatic chain and the aro-
matic ring suggest that two dif-
ferent mechanisms are utilized
to oxidize these molecular frag-
ments, although both processes
(aliphatic and aromatic hydrox-
Scheme 5. Intermediates involved in aromatic hydroxylation catalyzed by 1.
ylation) may share the same
metal-based oxidant. Similarly, previously reported hydrox-
ylation of p-xylene catalyzed by toluene 4-monooxygenase,
a non-heme iron enzyme, gave normal isotope effect (2.22)
for methyl hydroxylation and inverse isotope effect (0.735)
for aromatic hydroxylation.[49] P450 was also shown to give
normal isotope effect (7–10) for benzylic hydroxylation of
xylene, in agreement with an oxygen rebound mechanism;
in contrast, aromatic hydroxylation of xylenes was character-
ized by an inverse isotope effect (0.83–0.94).[50]
tion. In aromatic hydroxylations, a hydrogen-atom shift (the
so-called NIH shift) is common for reactions that proceed
through an electrophilic attack of the oxidant on the aro-
matic ring with subsequent rearrangement of the cationic in-
termediate to a ketone intermediate (Scheme 4C).[51] We ob-
served a non-zero NIH shift for hydroxylation of 3,5-2H2-
chlorobenzene and 2,4,6-2H3-chlorobenzene with 1/H2O2;
these results corroborate electrophilic addition to the aro-
matic ring (Table S6 in the Supporting Information).
In addition to kinetic isotope effects, H/D migrations and
rearrangements provide meaningful mechanistic informa-
The results obtained for aromatic hydroxylations with hy-
drogen peroxide in the presence of 1 are in line with a
14002
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13995 – 14006

Aromatic Hydroxylation
FULL PAPER
number of related oxidations at the iron center that typically
show a small inverse kinetic isotope effect, and agree with
an electrophilic attack on the aromatic ring.[48,49,52–54]
tetrakis(2-pyridylmethyl)ethane-1,2-diamine))
analogues,
and hydroperoxo intermediates were proposed to be respon-
sible for catalytic activity, but no specific experiments that
support this hypothesis were described.[61] Pre-generated
FeIIIOOH species supported by several aminopyridine li-
gands (including tris-picolylamine, tpa) were shown to be
sluggish oxidants that did not react with olefins, sulfide, or
phosphines[32] at low temperatures. However, very little in-
formation is available about Fe–bpmen-derived intermedi-
ates, therefore we developed methods to generate bpmen-
Discussion
High activity of complex 1 in oxidation of organic substrates
with hydrogen peroxide[2,7,12] made this complex a promising
candidate for exploring intermolecular aromatic hydroxyl-
ation of substrates lacking directing groups. Similar to previ-
ously reported functional-group-assisted hydroxylation of
benzoic acids, we now found that 1 is equally efficient in
promoting hydroxylation of benzenes. Although biological
aromatic hydroxylations can be catalyzed by mononuclear
non-heme iron centers in pterin-dependent aromatic amino
acid hydroxylases,[55] and by non-heme diiron centers in bac-
terial multicomponent monooxygenases (BMMs), such as
methane and toluene monooxygenases,[56] synthetic biomim-
etic systems capable of hydroxylating unfunctionalized aro-
matic rings are still limited.[57] Aside from variations of clas-
sical Fenton reaction, which are usually non-selective and
often low-yielding, iron phthalocyanines[58] and iron com-
plexes with aminopyridine ligands[59] are among a handful of
recent examples of non-porphyrin complexes reactive in in-
termolecular aromatic hydroxylations. Complex 1 promotes
nearly quantitative conversion of aromatic hydrocarbons to
phenolates at room temperature within several minutes,
with H2O2 as the oxidant. Importantly, a range of substituted
benzenes, including electron-poor substrates (such as chloro-
benzene) readily undergoes hydroxylation with 1/H2O2.
However, strongly electron-withdrawing substituents (e.g.,
nitro groups) significantly decrease the yield of hydroxylated
product. Despite high reaction rates in aromatic hydroxyl-
ation with 1/H2O2, efficient catalysis is precluded by strong
supported FeIII
ACTHUNRTGNE(GNU OOH) in reasonable yield, and probed its
reactivity under ambient conditions where high-yielding hy-
droxylation takes place.
Formation of the [FeIII (bpmen)]2+ intermediate in-
ACTHNUTRGENNUG(OOH)ACHTUTGNRENNUGN
volved in hydroxylation was observed by stopped-flow ki-
netic measurements with spectrophotometric registration, by
EPR spectroscopy, and by mass spectrometry. The yield of
the FeIII
ACHTNUTRGNEUNG(OOH) intermediate significantly depends on tem-
perature; counterintuitively, it can be generated in relatively
high yield only at or near room temperature. Using stopped-
flow techniques we established that [FeIII (bpmen)]2+
ACTHNUTRGENNUG(OOH)ACHTUTGNRENNUGN
does not directly attack aromatic substrates: reaction rates
were independent on the concentration of benzene and on
the nature of substituent on the aromatic ring. Also, 18O-la-
beling experiments demonstrated that active intermediate
responsible for aromatic hydroxylation exchanges with
water and thus FeIII
These results are in line with generally low oxidizing reactiv-
ity of FeIII(OOH) intermediates.[62] A somewhat different
ACHTUNGTRNE(NUNG OOH) cannot be the sole oxidant.
AHCTUNGTRENNUNG
situation was found for iron bleomycin (BLM), a natural an-
tibiotic that causes oxidative DNA cleavage in the presence
of iron and dioxygen.[63] Although earlier studies indicated
nearly identical decay rates of activated BLM (ABLM) with
and without substrate present,[64] the more recent study by
Solomon and co-workers,[26] which took advantage of a dis-
tinct signature of ABLM in MCD spectra, provides evidence
coordination of the phenolate products to the ironACTHUNTGRNEUNG(III)
center. This coordination, however, has two beneficial ef-
fects: it prevents overoxidation of phenols, thus greatly im-
proving product selectivity, and it gives rise to intensely col-
in favor of direct interaction between this FeIII
mediate and the substrate. Differences in reactivity of syn-
thetic FeIII
(OOH) complexes (which are sluggish oxidants)
compared to a natural system, ABLM, can be attributed, at
least in part, to differences in spin states of iron(III), and to
ACHTUNGTRENUN(NG OOH) inter-
AHCTUNGTRENNUNG
ored ironACHTUNGTRENNUNG(III)–phenolate products, thus enabling spectro-
photometric monitoring of the reaction progress. The latter
feature was particularly useful for the present study, which
was focused on identifying metal-based intermediates that
are competent hydroxylating agents.
ACHTUNGTRENNUNG
enzyme-like proximity effects due to bleomycin intercala-
tion between the DNA base pairs.
Results presented herein, suggest that self-decomposition
Plausible intermediates that were proposed for non-heme
of FeIII
ACTHNGUTERN(UNG OOH) generates an active species capable of oxidiz-
iron systems include FeIII
N
ing an aromatic substrate. FeIII
ACTHNGUTERNN(UG OOH) can udergo heterolyt-
a fleeting FeV species; all of these species were previously
implicated in various iron-promoted oxidations, and most of
them were spectroscopically observed for some non-heme
iron complexes.[6] For intermolecular hydroxylations of non-
functionalized aromatic compounds, a number of active
non-heme iron reagents or catalysts is limited, and the
active species in these systems were not identified.[60,61] Al-
ic or homolytic O O bond cleavage to produce Fe =O or
FeIV=O respectively (Scheme 5). It is believed that mononu-
clear non-heme iron enzymes that hydroxylate aromatic sub-
strates, pterin-dependent oxygenases (PheH, TyrH, TrpH),
function by means of high-spin FeIV=O;[65] recently this in-
termediate was trapped in the reaction catalyzed by Tyro-
sine hydroxylase (TyrH) and characterized by rapid freeze–
quench Mçssbauer spectroscopy.[66] By now many synthetic
FeIV=O complexes have been generated and characterized
spectroscopically and, in several cases, crystallographical-
ly,[3–5] and in some cases, this species was able to carry out
V
ꢀ
though FeIII
ACHTUNGTRENNUNG(OOH) was proposed to participate in aromatic
hydroxylation,[61] the reactivity of this species was not stud-
ied in direct experiments. For example, anisol is hydroxylat-
ed with H2O2 in the presence of FeII–tpen (tpen=N,N,N’,N’-
Chem. Eur. J. 2010, 16, 13995 – 14006
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14003

E. V. Rybak-Akimova and O. V. Makhlynets
an oxidation of organic substrates (cyclooctene,[67,68] cyclo-
hexane,[43] thioanisole,[67] triphenylmethane,[60,69] alcohol,[60,70]
anthracene[53]). Intramolecular, regioselective hydroxylation
change with H218O and was formulated as an FeV=O species.
A similar oxidant has been proposed for the iron-catalyzed
self-hydroxylation of aryl peroxyacids,[74] and for the intra-
molecular hydroxylation of the aromatic ring covalently ap-
pended to the analogue of 1.[75] Although only one non-
heme iron(V)–oxo complex has been directly observed at
low temperature as a transient species and characterized by
a variety of spectroscopic methods,[76] and recent (albeit pre-
liminary)[77] EPR results suggested transient, low-yield gen-
of the appended aromatic ring in reaction of [FeII
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
(CH3CN)2]2+
ACHTUNGTRENNUNG
a homolytic pathway that generates an FeIV=O species as a
competent oxidant.[42] However, we recently showed that
the tpa–FeIV=O intermediate does not hydroxylate benzoic
acids.[2] In addition, an FeIV=O intermediate supported by a
substituted TPEN did not oxidize anisole (although efficient
anisole hydroxylation with H2O2 was promoted by this iron
complex).[61] In the present work, we were able to generate,
in high yield, an FeIV=O intermediate supported by bpmen,
and confirmed that bpmen–FeIV=O does not hydroxylate ar-
ACTHNUTRGNEUNG
eration of FeV from [Fe(bpmen)]2+ and H2O2,[78] indirect evi-
dence supporting participation of non-heme FeV–oxo inter-
mediates in oxidations of alcohols,[79] olefins,[79–81] and
arenes[60,75,82] continues to accumulate.
In summary, [FeACHTUGNTRNE(NNUG bpmen)ACHTUNRGTEG(NNUN CH3CN)2]CAHTNGUTREN[UNGN ClO4]2 (complex 1) is
efficient in promoting hydroxylation of aromatic hydrocar-
bons with hydrogen peroxide, and the reaction is greatly ac-
celerated by additions of carboxylic acids. The system re-
ported herein is a rare example of an efficient aromatic hy-
droxylation with non-heme iron complex and hydrogen per-
oxide that allows for an unambiguous interpretation of the
role of the FeIII–hydroperoxo intermediates. A new bpmen–
omatic rings. Similarly to ironACTHNUTRGNEUGN(III)–hydroperoxo intermedi-
ates, the differences in reactivity of ferryl(IV) species in en-
zymes and in synthetic complexes can be attributed to dif-
ferent spin states of iron(IV)[63,71] (high spin in mononuclear
non-heme iron enzymes,[72] and low spin in synthetic inter-
mediates supported by aminopyridine ligands[73]), and by
proximity effects in enzymes (and also in intramolecular
ligand hydroxylations by iron(IV) species in synthetic sys-
tems).[42,63] An alternative mechanistic pathway that starts
ꢀ
FeIII
ACTHNUTRGNE(NUG OOH) intermediate is definitely involved in the reac-
tion pathway, but it does not attack the aromatic rings di-
rectly, generating a different, metal-based active oxidant in-
stead. Another new intermediate, bpmen–FeIV=O, was gen-
erated and shown to be unreactive with aromatic substrates.
C
with O O bond homolysis would utilize HO radicals in the
hydrogen atom abstraction from the aromatic ring, immedi-
ately followed by an oxygen atom incorporation from the
adjacent FeIV=O species. This scenario is inconsistent with
inverse H/D kinetic isotope effects in aromatic hydroxyl-
ation (an electrophlic attack rather than a hydrogen atom
abstraction must occur). In addition, use of an hydroxyl rad-
ical trap (galvinoxyl radical) did not effect the yields of phe-
ꢀ
An acid-assisted O O bond heterolysis of coordinated per-
oxides is likely involved in the reaction pathway (Scheme 5).
Inverse H/D kinetic isotope effects and the NIH shift in hy-
droxylations of selectively deuterated substituted benzenes
imply an electrophilic attack of the aromatic ring by the oxi-
dant and exclude a hydrogen atom abstraction as the key
step in the hydroxylation pathways.
C
nols. Observed NIH shift also argues against HO radicals
playing a major role in the aromatic ring hydroxylation.
We have shown that acetic acid facilitates decay of FeIII-
ACHTUNGTRENNUNG
(OOH) and leads to faster FeIII–phenolate formation, thus
Experimental Section
acetic acid promotes formation of reactive species responsi-
ble for aromatic hydroxylation. Moreover, acetic acid assist-
All chemicals and solvents were purchased from Aldrich, Acros Organics
or Fisher Scientific and were used without additional purification unless
otherwise noted. H218O2 (90% isotopic purity, 2% solution in water) was
obtained from ICON Isotopes, H218O (98%) was purchased from Cam-
bridge Isotope Laboratories. Use of anhydrous HPLC grade acetonitrile
ꢀ
ed O O bond heterolysis of a mechanistic probe MPPH in
the presence of 1 and benzene was accompanied by a signifi-
cant growth in yields of hydroxylated products (phenolates),
these results suggest that hydroxylation is performed by
ꢀ
is imperative in order to obtain FeIII
ACTHNGUTERN(UNG OOH) intermediate. GCMS experi-
V
Fe =O species. Carboxylic acid assisted O O bond heteroly-
sis pathway, leading to an as-of-yet unobserved FeV=O oxi-
dant, has been proposed[12] by Mas-Ballestꢂ and Que in the
efficient epoxidation of olefins by H2O2 in the presence of
ments were carried out using a Shimadzu GC-17 A gas chromatograph
(Rtx-xLB column) with a GCMS-QP 5050 mass detector. Time-resolved
spectra of rapid hydroxylation reactions were acquired with TgK Scientif-
ic (formerly Hi-Tech Schientific, Salisbury, Wiltshire, UK) SF-61DX2
cryogenic Stopped-flow system equipped with J&M Diode array (Spec-
tralytics).
acetic acid and [FeII (CH3CN)2]2+, and it also appears to
ACHTUNGTRENNUNG(tpa)ACHTUTGNRENNUGN
be a likely mechanism of regioselective benzoic acid hydrox-
ylations.[2] Recently, another example of aromatic hydroxyl-
ation involving a proposed FeV=O intermediate was report-
ed. Upon addition of CH3CO3H to a solution containing
Materials: Isopropyl ester of 2-iodoxybenzoic acid (IBX ester),[36] N,N’-
dimethyl-N,N’-bis(2-pyridylmethyl)ethane-1,2-diamine (bpmen)[83] and
(ClO4)2 (1)[84] were prepared using published pro-
[FeACHTUNGRTNE(NUNG bpmen)ACHTNURTGEG(NNUN CH3CN)2]ACHTUNGTRENNUGN
cedures. 2-Methyl-1-phenyl-2-propyl hydroperoxide (MPPH), generously
provided by Prof. John Caradonna, was recrystallized from pentane
(208C) prior to use until iodometric titration proved> 98% activity.[44]
For iodometric titration solid KI (210 mg) was added to recrystallized
MPPH (30 mg) dissolved in aqueous acetic acid (20 mL, 25% v/v). The
reaction was allowed to proceed for 10 h in a sealed flask followed by
quick titration of the resulting brown solution with Na2S2O3 (0.1m aque-
ous solution).
[FeII
(bqen)]2+
(bqen=N,N’-dimethyl-N,N’-bis(8-quinolyl)-
ethane-1,2-diamine) and benzyl alcohol, aromatic ring was
hydroxylated at the ortho-position.[60] Interestingly, the cor-
responding FeIV=O species did not hydroxylate aromatic
ring, but produced benzaldehyde. The intermediate respon-
sible for aromatic hydroxylation was found to rapidly ex-
14004
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13995 – 14006

Aromatic Hydroxylation
FULL PAPER
Determination of phenol yield using GCMS: Mixture of 1 with benzene
in CH3CN was prepared in a glove box. Hydrogen peroxide was prepared
aerobically and delivered at room temperature; the resulting solution
was stirred for 30 min for reaction to complete. Reaction mixtures were
acetylated (0.1 mL of 1-methylimidazole and 1 mL of acetic anhydride)
for 30 min followed by addition of 1m HCl (2 mL) and extraction with di-
chloromethane (1 mL). The organic layer was washed with saturated
aqueous NaHCO3 (2 mL), water (2 mL) and finally dried over MgSO4.
Internal standard (nitrobenzene or naphthalene) was added prior to the
extraction. Yields of hydroxylated products were established by GCMS
relative to the internal standard and converted to absolute yields using a
calibration curve with phenyl acetate and the corresponding standard, de-
termined prior to each run. All experiments were run at least in tripli-
cate, reported yields are average of these trials.
MPPH cleavage: O2 was bubbled through anaerobically prepared solu-
tion of 1 (2 mm) and benzene (280 equiv vs. iron) in acetonitrile (1 mL)
for 20 s followed by injection of MPPH (60 mL, 2 mmol, 1 equiv vs. iron)
at room temperature. After addition of MPPH, solutions were stirred for
ꢁ30 min and prepared for GCMS analysis as described above. When
looking at the effects of H+ on the reactivity, acetic acid (10 mL, 1 mmol,
0.5 equiv vs. iron) was added to 1/benzene mixture before purging it with
O2.
Blank solutions containing only pure acetonitrile (1 mL) were treated
same way as the 1/benzene mixtures. Control samples with pure bibenzyl,
benzyl alcohol acetate, benzaldehyde, 2-methyl-1-phenyl-2-propyl alcohol
acetate (MPPol), phenyl acetate and nitrobenzene were run to find de-
tector response for each product against nitrobenzene.
Determination of H/D KIE: An solution of 1 (5 mm) in acetonitrile and
a 1:1 mixture of toluene/[D8]toluene (150 equiv each; the ratio of protio/
deuterio isotopomers was analyzed by GCMS and adjusted to 1:1 if nec-
essary) was prepared in a glove box. H2O2 (3 equiv vs. 1) was added
dropwise at room temperature and the mixture was left to react for
30 min. Products were acetylated and subjected to GCMS analysis; the
amounts of the 2H and 1H products were quantified by integrating the
peaks of their corresponding ions. The o-, m-, and p-cresols were well
separated in the chromatogram; however, the GC peak from meta cresol
partially overlapped with benzyl alcohol. Therefore we used an ion
unique to benzyl alcohol (91 for protio and 98 for deutero) to determine
KIE for the hydroxylation of CH3 group; relatively low intensity of this
peak limits the accuracy of KIE for benzylic hydroxylation. Additional
details are provided in Supporting Information (Figure S24).
Acknowledgements
We thank Prof. Dr. L. Que, Jr. for fruitful discussion and Dr. Ivan V. Ko-
rendovych and Dr. Rubꢂn Mas-Ballestꢂ for critical reading of the manu-
script. This work was supported by the US Department of Energy (Grant
DE-FG02–06ER15799). The NMR facility, the ESI-MS spectrometer,
and rapid kinetic instrumentation at Tufts were supported by the NSF
(grants CHE-9723772, MRI CHE 0821508, MRI CHE 0320783, and
CHE 0639138).
[1] M. C. White, A. G. Doyle, E. N. Jacobsen, J. Am. Chem. Soc. 2001,
123, 7194.
Stopped-flow experiments: Kinetic measurements were performed in the
diode array mode using a stopped-flow instrument. Reactivity of 1 with
hydrogen peroxide in the presence of benzene was studied in acetonitrile
at 208C. In a single mixing experiment, an acetonitrile solution of 1 and
benzene was prepared in a glove box and mixed with H2O2 in the
stopped-flow. In a double mixing experiment 1 was mixed with H2O2, the
[2] O. V. Makhlynets, P. Das, S. Taktak, M. Flook, R. Mas-Balleste,
E. V. Rybak-Akimova, L. Que Jr. , Chem. Eur. J. 2009, 15, 13171.
[3] J.-U. Rohde, J.-H. In, M. H. Lim, W. W. Brennessel, M. R. Bukow-
ski, A. Stubna, E. Mꢃnck, W. Nam, L. Que Jr. , Science 2003, 299,
1037.
[4] J. England, Y. Guo, E. R. Farquhar, V. G. Young, Jr., E. Mꢃnck, L.
Que Jr. , J. Am. Chem. Soc. 2010, 132, 8635.
[5] E. J. Klinker, J. Kaizer, W. W. Brennessel, N. L. Woodrum, C. J.
Cramer, L. Que, Jr., Angew. Chem. 2005, 117, 3756; Angew. Chem.
Int. Ed. 2005, 44, 3690.
reaction mixture was incubated to reach the highest yield of FeIII
ACTHNUTRGNE(NUG OOH)
(age time), followed by addition of benzene. The effect of acetic acid on
hydroxylation rate was studied in a double mixing experiment, in which
FeIII
ACHTUNGTRENNUNG(OOH) was pre-generated by mixing 1 and H2O2 at 208C followed
by addition of benzene with variable amounts of acetic acid (1–4 equiv
vs. 1). Concentrations of all reagents are reported in the figure captions
for the onset of the reaction (after mixing). Kinetic parameters were de-
termined in SPECFIT (global fitting), Kinetic Studio or IgorPro using
single-exponential (A!B) or double exponential (A!B!C) models.
[6] L. Que, Jr., W. B. Tolman, Nature 2008, 455, 333.
[7] M. S. Chen, C. White, Science 2007, 318, 783.
[8] S. Taktak, M. Flook, B. M. Foxman, L. Que, Jr., E. V. Rybak-Aki-
mova, Chem. Commun. 2005, 5301.
[9] L. Que, Jr., J. Biol. Inorg. Chem. 2004, 9, 684.
[10] I. V. Korendovych, S. V. Kryatov, E. V. Rybak-Akimova, Acc. Chem.
Res. 2007, 40, 510.
Single-exponential fit was calculated to A=Ainf +DAexp
(ꢀkt), in which
Ainf is absorbance of the reaction mixture after reaction is complete,
DA=A0ꢀAinf ; A0 is initial absorbance. Similarly, double-exponential fit
[11] E. A. Duban, K. P. Bryliakov, E. P. Talsi, Eur. J. Inorg. Chem. 2007,
852.
was calculated using A=Ainf +DA1 exp
N
U
[12] R. Mas-Ballestꢂ, L. Que Jr. , J. Am. Chem. Soc. 2007, 129, 15964.
[13] Although blue species decays over time this process is not due to
overoxidation reaction, GCMS analysis showed that the blue solu-
tion (maximum accumulation of phenolate) and the yellow solution
(1 h after the onset of the reaction) contain the same amount of
phenol (Figure S2 in the Supporting Information).
[14] H. Marusawa, K. Ichikawa, N. Narita, H. Murakami, K. Ito, T.
Tezuka, Bioorg. Med. Chem. 2002, 10, 2283.
[15] C. Kim, K. Chen, J. Kim, L. Que, Jr., J. Am. Chem. Soc. 1997, 119,
5964.
[16] M. Lubben, A. Meetsma, E. C. Wilkinson, B. Ferinda, L. Que, Jr.,
Angew. Chem. 1995, 107, 1610; Angew. Chem. Int. Ed. Engl. 1995,
34, 1512.
[17] P. Mialane, A. Nivorojkine, G. Pratviel, L. Azema, M. Slany, F.
Godde, A. Simaan, F. Banse, T. Kargar-Grisel, G. Bouchoux, J. Sain-
ton, O. Horner, J. Guilhem, L. Tchertanova, B. Meunier, J.-J.
Girerd, Inorg. Chem. 1999, 38, 1085.
[18] B. Bernal, I. M. Jensen, K. B. Jensen, C. J. McKenzie, H. Toftlund,
J.-P. Tuchagues, J. Chem. Soc. Dalton Trans. 1995, 3667.
[19] K. B. Jensen, C. J. McKenzie, N. P. Nielsen, J. Z. Pedersen, H. M.
Svendsen, Chem. Commun. 1999, 1313.
ments were run in at least triplicate and averaged rate constants are re-
ported. The activation enthalpy and entropy for FeIII
(OOH) formation
were calculated from linear Arrhenius and Eyring plots.
AHCTUNGTRENNUNG
Isotope-labeling studies: A mixture of 1 (2 mm) and benzene (280 equiv
vs. 1) was prepared in a glove box. In experiments with labeled hydrogen
peroxide, H218O2 (11 mL, 0.53m, 3 equiv vs. 1) was added to the 1/benzene
mixture (1 mL) at room temperature. In an experiment with labeled
water, H218O (11 mL, 98%) was added to solution of the 1/benzene
(1 mL) prior to the addition of H2O2 (60 mL of 0.1m H2O2 diluted by
CH3CN from 70% stock solution). After 30 min the reaction mixture was
subjected to GCMS analysis as described above.
EPR studies: Purple species 2 was generated directly in the EPR tube by
mixing 1 (3 mm, 0.15 mL) with H2O2 (30 mm, 0.15 mL) in acetonitrile at
room temperature. The tube was frozen immediately after all hydrogen
peroxide was injected (5 s). Complex 2 decays fast when treated with
acetic acid: Complex 2 was pre-generated in the EPR tube as described
above and then acetic acid (1.5 mm, 0.11 mL, <0.5 equiv vs. iron) was in-
jected into the tube and reaction quenched in liquid nitrogen (reaction
time 5 s). All EPR spectra were acquired at 120 K. EPR spectra simula-
tion were done using SimFonia (Bruker).
Chem. Eur. J. 2010, 16, 13995 – 14006
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14005

E. V. Rybak-Akimova and O. V. Makhlynets
[20] T. J. Mizoguchi, S. J. Lippard, J. Am. Chem. Soc. 1998, 120, 11022.
[21] V. Balland, F. Banse, E. Anxolabehere-Mallart, M. Ghiladi, T. A.
Mattioli, C. Philouze, G. Blondin, J.-J. Girerd, Inorg. Chem. 2003,
42, 2470.
[22] A. J. Simaan, S. Dopner, F. Banse, S. Bourcier, G. Bouchoux, A.
Boussac, P. Hildebrandt, J.-J. Girerd, Eur. J. Inorg. Chem. 2000,
1627.
[51] D. R. Boyd, J. W. Daly, D. M. Jerina, Biochemistry 1972, 11, 1961.
[52] M.-J. Kang, W. J. Song, A.-R. Han, Y. S. Choi, H. G. Jang, W. Nam,
J. Org. Chem. 2007, 72, 6301.
[53] S. P. de Visser, K. Oh, A.-R. Han, W. Nam, Inorg. Chem. 2007, 46,
4632.
[54] G. R. Moran, A. Derecskei-Kovacs, P. J. Hillas, P. F. Fitzpatrick, J.
Am. Chem. Soc. 2000, 122, 4535.
[23] D. QuiÇonero, K. Morokuma, D. G. Musaev, R. Mas-Ballestꢂ, L.
Que, Jr., J. Am. Chem. Soc. 2005, 127, 6548.
[24] J.-J. Girerd, F. Banse, A. J. Simaan, Struct. Bonding (Berlin) 2000,
97, 145.
[25] R. M. Burger, S. B. Horwitz, J. Peisach, J. B. Wittenberg, J. Biol.
Chem. 1979, 254, 12299.
[26] M. S. Chow, L. V. Liu, E. I. Solomon, Proc. Natl. Acad. Sci. USA
2008, 105, 13241.
[27] I. Lippai, R. S. Magliozzo, J. Peisach, J. Am. Chem. Soc. 1999, 121,
780.
[28] C. Nguyen, R. J. Guajardo, P. K. Mascharak, Inorg. Chem. 1996, 35,
6273.
[29] W. Nam, R. Ho, J. S. Valentine, J. Am. Chem. Soc. 1991, 113, 7052.
[30] X. Shan, L. Que Jr. , J. Inorg. Biochem. 2006, 100, 421.
[31] K. Chen, M. Costas, L. Que Jr. , J. Chem. Soc. Dalton Trans. 2002,
672.
[32] M. J. Park, J. Lee, Y. Suh, J. Kim, W. Nam, J. Am. Chem. Soc. 2006,
128, 2630.
[33] Y. Zhou, X. Shan, R. Mas-Balleste, M. R. Bukowski, A. Stubna, M.
Chakrabarti, L. Slominski, J. A. Halfen, E. Mꢃnck, L. Que, Jr.,
Angew. Chem. 2008, 120, 1922; Angew. Chem. Int. Ed. 2008, 47,
1896.
[55] P. C. A. Bruijnincx, G. van Koten, R. J. M. Klein Gebbink, Chem.
Soc. Rev. 2008, 37, 2716.
[56] L. J. Murray, S. J. Lippard, Acc. Chem. Res. 2007, 40, 466.
[57] D. A. Alonso, C. Nꢄjera, I. M. Pastor, M. Yus, Chem. Eur. J. 2010,
16, 5829.
[58] E. V. Kudrik, A. B. Sorokin, Chem. Eur. J. 2008, 14, 7123.
[59] A. Thibon, J.-F. Bartoli, R. Guillot, J. Sainton, M. Martinho, D.
Mansuy, F. Banse, J. Mol. Catal. A 2008, 287, 115.
[60] J. Yoon, S. A. Wilson, Y. K. Jang, M. S. Seo, K. Nehru, B. Hedman,
K. O. Hodgson, E. Bill, E. I. Solomon, W. Nam, Angew. Chem. 2009,
121, 1283; Angew. Chem. Int. Ed. 2009, 48, 1257.
[61] A. Thibon, J.-F. Bartoli, S. Bourcier, F. Banse, Dalton Trans. 2009,
9587.
[62] E. V. Rybak-Akimova in Physical Inorganic Chemistry (Ed.: A.
Bakac), Wiley, Hoboken, 2010, pp. 109.
[63] E. I. Solomon, S. D. Wong, L. V. Liu, A. Decker, M. S. Chow, Curr.
Opin. Chem. Biol. 2009, 13, 99.
[64] R. M. Burger, J. Peisach, S. B. Horwitz, J. Biol. Chem. 1981, 256,
11636.
[65] E. G. Kovaleva, J. Lipscomb, Nat. Chem. Biol. 2008, 4, 186.
[66] B. E. Eser, E. W. Barr, P. A. Frantom, L. Saleh, J. M. Bollinger, C.
Krebs, P. F. Fitzpatrick, J. Am. Chem. Soc. 2007, 129, 11334.
[67] M. H. Lim, J.-U. Rohde, A. Stubna, M. R. Bukowski, M. Costas,
R. Y. N. Ho, E. Mꢃnck, W. Nam, L. Que Jr. , Proc. Natl. Acad. Sci.
USA 2003, 100, 3665.
[34] R. Mas-Ballestꢂ, E. Mꢃnck, L. Que, Jr., unpublished results.
ꢀ
III(OOH) (purple)
[35] Acetic acid facilitates O O bond cleavage of Fe CAHTUNGRTENNUGN
and formation of very unstable active species that quickly converts
into iron
oxide (in excess) to produce FeIV=O. To test this hypothesis we gen-
erated FeIII
(OOH) by mixing 1 and 1.5 equiv of H2O2 (stoichiomet-
ric amount) and then added acetic acid to the preformed 2, as ex-
pected we saw a decay of FeIII(OOH) and very low yield of FeIV=O
G
[68] M. Martinho, F. Banse, J.-F. Bartoli, T. Mattioli, P. Battioni, O.
Horner, S. Bourcier, J.-J. Girerd, Inorg. Chem. 2005, 44, 9592.
[69] J. Kaizer, E. J. Klinker, N. Y. Oh, J.-U. Rohde, W. J. Song, A.
Stubna, J. Kim, E. Mꢃnck, W. Nam, L. Que, Jr., J. Am. Chem. Soc.
2004, 126, 472.
(Figure S23 in the Supporting Information).
[70] N. Y. Oh, Y. Suh, M. J. Park, M. S. Seo, J. Kim, W. Nam, Angew.
Chem. 2005, 117, 4307; Angew. Chem. Int. Ed. 2005, 44, 4235.
[71] G. Xue, R. De Hont, E. Mꢃnck, L. Que Jr. , Nat. Chem. 2010, 2,
400.
[72] J. M. Bollinger, Jr., C. Krebs, J. Inorg. Biochem. 2006, 100, 586.
[73] W. Nam, Acc. Chem. Res. 2007, 40, 522.
[74] N. Y. Oh, M. S. Seo, M. H. Lim, M. B. Consugar, M. J. Park, J.-U.
Rohde, J. Han, K. M. Kim, J. Kim, L. Que, Jr., W. Nam, Chem.
Commun. 2005, 5644.
[75] A. Nielsen, F. B. Larsen, A. D. Bond, C. J. McKenzie, Angew. Chem.
2006, 118, 1632; Angew. Chem. Int. Ed. 2006, 45, 1602.
[76] F. T. de Oliveira, A. Chanda, D. Banerjee, X. Shan, S. Mondal, L.
Que, Jr., E. Bominaar, E. Mꢃnck, T. J. Collins, Science 2007, 315,
835.
[36] V. V. Zhdankin, D. N. Litvinov, A. Y. Koposov, T. Luu, M. J. Fergu-
son, R. McDonald, R. R. Tykwinski, Chem. Commun. 2004, 106.
[37] I. M. Geraskin, O. Pavlova, H. M. Neu, M. S. Yusubov, V. N. Nemy-
kin, V. V. Zhdankin, Adv. Synth. Catal. 2009, 351, 733.
[38] J.-U. Rohde, A. Stubna, E. Bominaar, E. Mꢃnck, W. Nam, L.
Que Jr. , Inorg. Chem. 2006, 45, 6435.
[39] M. S. Seo, J.-H. In, S. O. Kim, N. Y. Oh, J. Hong, J. Kim, L. Que Jr. ,
W. Nam, Angew. Chem. 2004, 116, 2471; Angew. Chem. Int. Ed.
2004, 43, 2417.
[40] B. Meunier, J. Bernadou, Struct. Bonding (Berlin) 2000, 97, 1.
[41] M. D. Wolfe, J. D. Lipscomb, J. Biol. Chem. 2003, 278, 829.
[42] I. M. Jensen, S. J. Lange, M. P. Mehn, E. L. Que, L. Que, Jr., J. Am.
Chem. Soc. 2003, 125, 2113.
[43] T. L. Foster, J. Caradonna, J. Am. Chem. Soc. 2003, 125, 3678.
[44] G. T. Rowe, E. V. Rybak-Akimova, J. P. Caradonna, Chem. Eur. J.
2008, 14, 8303.
[45] P. A. MacFaul, K. U. Ignold, D. D. M. Wayner, L. Que Jr. , J. Am.
Chem. Soc. 1997, 119, 10594.
[46] To check stability of MPPH during the work up and GCMS experi-
ments we set up two control experiments containing 1) MPPH and
2) MPPH with equimolar amount of acetic acid. In both cases the
main product was bibenzyl (product of benzyl radical coupling) with
small amount of MPPol. For hydroxylation experiments we used
one equivalent of MPPH to minmize any contribution of MPPH
self-decomposition.
[47] R. P. Hanzlik, G. O. Shearer, Biochem. Pharmacol. 1978, 27, 1441.
[48] K. R. Korzekwa, D. C. Swinney, W. F. Trager, Biochemistry 1989, 28,
9019.
[77] We have not been able to trap the intermediate described in the
paper.
[78] O. Y. Lyakin, K. P. Bryliakov, G. J. P. Britovsek, E. P. Talsi, J. Am.
Chem. Soc. 2009, 131, 10798.
[79] S. H. Lee, J. H. Han, H. Kwak, E. Y. Lee, H. J. Kim, J. H. Lee, C.
Bae, S. N. Lee, Y. Kim, C. Kim, Chem. Eur. J. 2007, 13, 9393.
[80] K. Chen, M. Costas, J. Kim, A. K. Tipton, L. Que Jr. , J. Am. Chem.
Soc. 2002, 124, 3026.
[81] A. Company, Y. Feng, M. Gꢃell, X. Ribas, J. M. Luis, L. Que, Jr., M.
Costas, Chem. Eur. J. 2009, 15, 3359.
[82] Y. Feng, C. Ke, G. Xue, L. Que Jr. , Chem. Commun. 2009, 50.
[83] S. Poussereau, G. Blondin, M. Cesario, J. Guilhem, G. Chottard, F.
Gonnet, J.-J. Girerd, Inorg. Chem. 1998, 37, 3127.
[84] K. Chen, L. Que Jr. , Chem. Commun. 1999, 1375.
[49] K. H. Mitchell, C. E. Rogge, T. Gierahn, B. G. Fox, Proc. Natl.
Acad. Sci. USA 2003, 100, 3784.
Received: September 6, 2010
[50] R. P. Hanzlik, K.-H. J. Ling, J. Am. Chem. Soc. 1993, 115, 9363.
Published online: November 29, 2010
14006
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13995 – 14006