Kinetic analysis of the formation and decay of a non-heme ferric hydroperoxide species susceptible to O-O bond homolysis

Article abstract of DOI:10.1021/ic5003786

The formation of a ferric hydroperoxide species from [Fe(bbpc)(MeCN) ₂]²⁺ (bbpc = N,N-dibenzyl-N,N-bis(2-pyridylmethyl)-1,2- cyclohexanediamine) and its subsequent decomposition were analyzed kinetically. The rate of decay is not strongly influenced by the presence of either water or substrate, suggesting that the ferric hydroperoxide degrades through O-O bond homolysis and is not the relevant metal-based oxidant in the observed catalysis of C-H activation. The rate law corresponding to the complexs formation from O₂ is consistent with the intermediacy of a mononuclear ferric superoxo species.

Full text of DOI:10.1021/ic5003786

Article
pubs.acs.org/IC
Kinetic Analysis of the Formation and Decay of a Non-Heme Ferric
Hydroperoxide Species Susceptible to O−O Bond Homolysis
Qiao Zhang and Christian R. Goldsmith*
Department of Chemistry and Biochemistry, Auburn University, Auburn, Alabama 36849, United States
S
* Supporting Information
ABSTRACT: The formation of a ferric hydroperoxide species
from [Fe(bbpc)(MeCN)2]2+ (bbpc = N,N′-dibenzyl-N,N′-
bis(2-pyridylmethyl)-1,2-cyclohexanediamine) and its subse-
quent decomposition were analyzed kinetically. The rate of
decay is not strongly influenced by the presence of either water
or substrate, suggesting that the ferric hydroperoxide degrades
through O−O bond homolysis and is not the relevant metal-
based oxidant in the observed catalysis of C−H activation.
The rate law corresponding to the complex’s formation from
O2 is consistent with the intermediacy of a mononuclear ferric
superoxo species.
We recently generated an FeIII−OOH species with the ligand
INTRODUCTION
■
N,N′-dibenzyl-N,N′-bis(2-pyridylmethyl)-1,2-cyclohexanedi-
C−H bonds, particularly those on aliphatic carbons, are noto-
riously difficult to modify chemically. The key difficulty in
amine (bbpc; Scheme 1).8,24 The ferric complex can be
activating these functional groups is that the harsh oxidants and
reaction conditions generally needed for such chemical trans-
Scheme 1
formations tend to overoxidize hydrocarbon substrates. This
has prompted much research into developing synthetic options
that work under milder conditions.1−3 Among these are pro-
cesses that rely on nonheme iron catalysts, which have been
designed to mimic metalloenzymes capable of catalyzing alkane
oxidation under ambient conditions.4−7 These catalysts are
particularly attractive since they rely upon an inexpensive and
naturally abundant metal for activity. In these systems, the
general consensus is that the terminal oxidant reacts with an
prepared from [Fe(bbpc)(MeCN)2](SbF6)2 (1; Scheme 1)
iron(II) precursor to convert it to a higher-valent species that
through reactions with either H2O2 or O2. The O2 reactivity
is ultimately responsible for C−H activation. Ferric hydro-
requires a substrate with a weak C−H bond, and the obser-
peroxide and ferryl species are commonly proposed as inter-
vation of a primary kinetic isotope effect with 9,10-dihydro-
mediates in this chemistry and have been amply observed in
anthracene (kH/kD = 6.8) for the formation of FeIII−OOH
both enzymatic and small-molecule systems.8−24
suggests that it may be preceded by a ferric superoxide oxidant
One difficulty has been identifying the metal-based oxidant
(Scheme 2).8 The assignment of the FeIII−OOH intermediate
responsible for C−H activation. Whether ferric hydroperoxide
species can directly activate C−H bonds is still debated.17−19,25,26
Complicating matters is that the O−O bonds of these species
can break either homolytically or heterolytically; this may give
rise to fundamentally different sorts of reactivity.12,18 Que has
found strong evidence against the direct oxidation of alkene
and alkane substrates by low-spin Fe(III)-OOH species sus-
ceptible to heterolytic O−O cleavage.6,12,19 Nam, conversely,
has found evidence for the direct involvement of a high-spin
Fe(III)-OOH complex that decomposes through homolytic
O−O cleavage, although the analysis is complicated by the
rapid degradation of the complex to a ferryl oxidant in the
absence of substrate.17,18
was supported by UV/vis, EPR, and resonance Raman data.8,24
Although mass spectrometry suggests that the composition of
Scheme 2
Received: February 17, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic5003786 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
the complex is [Fe(bbpc)(MeCN)(OOH)]2+,8 we cannot rule
out that a portion of the Fe(III) exists as [Fe(bbpc)(OOH)]2+,
with the metal center being either penta- or hexacoordinate.
Pentacoordination, which would correspond to a terminally
bound hydroperoxide ligand, has been observed in the crystal
structure of the related [Fe(bbpc)Cl]+ (Figure S1, Supporting
Information). Strong steric repulsions between the ligands
were evident in the metrical parameters of [Fe(bbpc)Cl2], and
similar interactions could potentially spur dissociation of the
MeCN from [Fe(bbpc)(MeCN)(OOH)]2+.24 Electron para-
magnetic resonance (EPR) spectroscopy suggests that the ferric
complex undergoes a spin crossover and is likely high spin at
room temperature;8 regrettably, the complex is not sufficiently
stable at room temperature (t1/2 ≈ 2 min) to allow the spec-
troscopic measurements needed to confirm this assignment.
For the sake of simplicity, we refer to the ferric hydroperoxide
species as [Fe(bbpc)(OOH)]2+ (2) throughout this article,
with the caveat that this may well be a mixture of closely
related ferric hydroperoxide complexes in equilibrium with each
other.
Here, we have used stopped-flow kinetics to investigate the
formation and decay of 2. We have determined the influence
of several additives, notably acid, water, and hydrocarbon
substrates, on the rates of formation and decomposition of
the FeIII−OOH species. The results are consistent with
the complex decaying through O−O homolysis, rather than
heterolysis. Additionally, the presence of substrates that un-
dergo oxidation catalyzed by the ferrous precursor does not
impact the rate of decomposition of the ferric hydroperoxide
species, suggesting that it is not the relevant oxidant for C−H
activation.
Reactivity: Oxidation of 1 by O2. The reactions involving the
oxidation of 1 by O2 proceeded in a manner analogous to those in-
volving H2O2. For each experiment, 0.20 mL aliquots from two
syringes, one containing an aerobic solution of 1 (A) and one
containing an aerobic solution of cyclohexene (B), were simulta-
neously injected into the stopped-flow instrument. The solutions were
mixed for 1.0 s prior to the start of data acquisition. Additives, if any,
were introduced via syringe B. The aerobic solutions were prepared by
bubbling pure O2 through anhydrous MeCN for 20 min at room
temperature, resulting in stock solutions containing 8.1 mM O2.27 The
concentration of O2 was controlled and varied via dilution with solu-
tions made with degassed anhydrous MeCN. The initial concentration
of cyclohexene (C6H10) was 100 mM unless stated otherwise. The
stopped-flow spectrometer was set to 535 nm to monitor the changes
in absorbance, since intermediate 2 does not form cleanly when
generated from O2 and C6H10.8 The previously observed side reactivity
prompted us to limit the analysis to an initial rates analysis of the
formation of 2. All reactions involving O2 as a reagent were run at
298 K.
Data Analysis. All kinetic data were modeled using the GraphPad
Prism 6 program. All reactions were repeated at least three times in
order to confirm their reproducibility and to assess the precision of
the measurements. All first-order or pseudo-first-order processes
were allowed to proceed for at least 5 half-lives. All calculated acti-
vation parameters were obtained from measurements taken at four
temperatures. Data points were taken at each temperature, and the
entire experiment was repeated two additional times with fresh stock
solutions in order to confirm the reproducibility of the obtained values
of ΔH⧧ and ΔS⧧. Whenever a rate or a rate constant was correlated to
the concentration of a reagent, at least four different concentrations of
that reagent were investigated. Initial rates were estimated using the
11 data points taken from 1.0 to 3.0 s. In each case, the change
in absorbance scaled linearly with time, validating the initial rate
approximation.
RESULTS
EXPERIMENTAL SECTION
■
■
Generation of [Fe(bbpc)(OOH)]2+ from H2O2. An FeIII−
OOH species (2) was previously generated from the reaction
between H2O2 and [Fe(bbpc)(MeCN)2](SbF6)2 (1) in
acetonitrile (MeCN).8,24 When the reaction is followed by
UV/vis, the changes in absorbance can be fit satisfactorily to an
A → B → C model, with species B corresponding to 2. The
concentration of 2 is directly correlated to the intensity at
690 nm.24 Under normal conditions, the intensity peaks between
20 and 30 s after the reaction begins. The absorbance at 690 nm
increases linearly until 4−6 s after the reagents are combined, the
exact time being dependent upon the reaction conditions. The
linearity suggested that an initial rate analysis of these data would
be feasible. Consequently, we varied the starting concentrations
of the reagents and measured the changes in absorbance at
690 nm from 1 to 3 s to get an approximation of the initial
rates. Prior to 1 s, the data were noisy, likely due to lingering
turbulence from the mixing with our particular instrument and
set-up; this necessitated the truncation of the data.
Materials. Except where noted otherwise, all chemicals were
purchased from Sigma-Aldrich and used as received. Dry dioxygen
(O2) was purchased from Airgas. Anhydrous acetonitrile (MeCN) was
purchased from Acros Organics. Hydrogen peroxide (H2O2, 50 wt %)
was bought from Fisher; its concentration was confirmed to be
50.5 wt % via a titration with KMnO4 and H2SO4 in water solution.
The ligand N,N′-dibenzyl-N,N′-bis(2-pyridylmethyl)-1,2-cyclohexane-
diamine (bbpc) and its ferrous complex [Fe(bbpc)(MeCN)2](SbF6)2
(1) were synthesized and identified as described previously.24
Instrumentation. A Hi-Tech SF-51 stopped-flow spectrophotom-
eter was used for the described stopped-flow kinetic studies. The
reactions were monitored at either 690 or 535 nm with data points
taken every 0.2 s. These wavelengths were chosen since they displayed
the greatest changes in absorbance during the reactions corresponding
to the formation and decay of [Fe(bbpc)(OOH)]2+ (2). A Hi-Tech
C-400 circulator was used to control and maintain the temperature.
The program Olis 4300 was used for data acquisition. GraphPad Prism
6 was used for data analysis. A Varian Cary 50 spectrophotometer was
used to collect routine optical data; software from the WinUV Analysis
Suite was used to process and analyze these data.
The first two series of experiments were run in MeCN at
298 K. The initial concentration of 1 was varied between 0.10
and 0.80 mM with a set 5.0 mM initial concentration of H2O2.
Subsequently, the concentration of H2O2 was varied from 0.50
to 5.0 mM with a set initial concentration of 0.50 mM 1. In
both series, the changes in absorbance from 1 to 3 s scale
linearly with higher concentrations of the investigated reagent,
indicating that the formation of 2 is first order with respect to
both 1 and H2O2. Under such conditions, the formation of 2
follows the rate law described by eq 1.
Reactivity: Oxidation of 1 by H2O2. For each stopped-flow
kinetics experiment, 0.20 mL aliquots from two syringes, one filled
with 1 in MeCN (A) and one filled with H2O2 in MeCN (B), were
simultaneously injected into the instrument. Additives, if present, were
introduced via syringe B. The solutions were mixed for 1.0 s before the
data acquisition began. The spectrophotometer was set to 690 nm,
which corresponds to the peak absorbance of a strong ligand to metal
charge transfer feature associated with 2.24 For most experiments, the
initial concentration of 1 after mixing was 0.50 mM. The concentration
of 1 was controlled and varied by diluting a 1.0 mM stock solution
with pure MeCN. Unless stated otherwise, the initial concentration of
H2O2 after mixing was 5.0 mM. Except for the variable-temperature
experiments, the reactions were run at 298 K.
The formation of 2 from 0.50 mM 1 and 5.0 mM H2O2 was
studied in MeCN from 294 to 324 K. The temperature-dependent
B
dx.doi.org/10.1021/ic5003786 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
noticeable decrease in the kobs value is observed when over
5 equiv of HClO4 (2.5 mM) are added. The addition of NaClO4
had a similar, but smaller, effect. The kobs value for the for-
mation of 2 from 1 and 5.0 mM H2O2 was 0.095 s−1 in the
presence of 50 mM NaClO4 and 0.083 s−1 in the presence of
HClO4.
d[2]/dt = k2[1][H2O2]
(1)
second-order rate constants for the formation of 2 were
calculated from values of kobs obtained from the fits of the
A → B → C model to the data. Subsequently, these were used
to prepare an Eyring plot (Figure 1). ΔH⧧ and ΔS⧧ were
Decomposition of [Fe(bbpc)(OOH)]2+. The decomposi-
tion of 2 formed from H2O2 and 1 can be followed by the loss
of the 690 nm UV/vis band associated with 2 and fit to a simple
B → C step. When the initial concentration of H2O2 was varied,
it was found that excess terminal oxidant hastens the dis-
appearance of 2, as confirmed from the higher kobs value for the
B → C step from the fits to the data. Similar observations have
been made in other mononuclear non-heme iron systems and
have been used to explain the lessened oxidative efficiency for
hydrocarbon oxidation catalysis with higher loadings of
terminal oxidant.28 At higher concentrations of H2O2, the
relationship between [H2O2] and the observed rate constant for
the decay appears to be linear, suggesting that H2O2 is reacting
directly with 2. The large y intercept would suggest other
pathways for decomposition that are independent of [H2O2].
The addition of C6H12, which is oxidized to cyclohexanol and
cyclohexanone under these conditions,24 fails to significantly
alter the rate of decomposition of 2 (Figure 3). A 10% increase
Figure 1. Eyring plot for the formation of 2 from a reaction between
0.50 mM 1 and 5.0 mM H2O2 in MeCN at temperatures ranging from
294 to 324 K. The reactions were followed using the absorbance at
690 nm. All of the second-order rate constants from three independent
sets of experiments are plotted. The calculated ΔH⧧ value was 50.5
( 0.5) kJ mol−1. The calculated ΔS⧧ value was −50 ( 2) J K−1 mol−1.
calculated to be 50.5( 0.5) kJ mol−1 and −50( 2) J mol−1 K−1,
respectively. The negative entropy of activation is consistent with
an associative process, corroborating the bimolecular rate law
suggested by the initial rate analysis (eq 1).
Various concentrations of H2O and two oxidizable substrates,
cyclohexane (C6H12) and cyclohexene (C6H10), were added,
but none of these additives influenced the rate of formation of
2 from 1 and H2O2 to a significant degree (Figure 2). Water
Figure 3. Influence of additives on the kobs value for the
decomposition of 2 formed from 0.50 mM 1 and 5.0 mM H2O2 in
MeCN at 298 K. The data from three independent sets of experiments
are plotted for each additive. The kobs value for the decomposition of 2
by itself was found to be 0.0044 s−1.
in the rate of decay is observed as the concentration of C6H12 is
increased from 5.0 to 50 mM, but this is unlikely to result from
a direct reaction between 2 and C6H12. The addition of C6H10,
which has more readily oxidized allylic C−H bonds, would be
anticipated to increase kobs if substrate oxidation were involved
in the rate-determining step (RDS). This substrate, however,
likewise fails to increase the rate of the disappearance of 2
relative to the substrate-free baseline, even at concentrations as
high as 100 mM (Figure 3).
Figure 2. Influence of additives on the kobs value for the formation of 2
from 0.50 mM 1 and 5.0 mM H2O2 in MeCN at 298 K. The data from
three independent series of experiments are plotted for each additive.
It had been previously observed that H2O hastens the
disappearance of some, but not all, ferric hydroperoxide
species.10,12 The differences were attributed to water’s ability
to act as an acid and hasten heterolytic cleavage of the O−O
bond. The FeIII−OOH species that were not water-sensitive
were proposed to decay instead through homolytic cleavage of
the O−O bond. The presence of additional H2O beyond that
added with the H2O2 fails to accelerate the rate of decom-
position of 2, even when 100 equiv of H2O relative to iron is
present (50 mM, Figure 3).
might have been expected to inhibit the formation of 2 by
acting as a competing ligand, but the data instead suggest that
H2O cannot compete effectively with H2O2 for binding sites on
the iron. Conversely, the kobs value for the formation of 2 from
0.50 mM 1 and 5.0 mM H2O2 increases slightly (∼10%) as the
concentration of water increases from 5.0 to 50 mM (Figure 2).
Aliquots of HClO4 were also added to determine if the
presence of an acid influenced the rate of formation. A
C
dx.doi.org/10.1021/ic5003786 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
The O−O bond likely cleaves in the RDS of the decom-
position reaction. The temperature dependence of the kobs
value from the decay portion of the reaction between 0.50 mM
1 and 5.0 mM H2O2 was assessed from 294 to 324 K. An
Eyring plot prepared from these data yields the following
kinetic parameters: ΔH⧧ = 53.2( 1.1) kJ mol−1 and ΔS⧧ =
−68( 4) J mol−1 K−1 (Figure 4).
DISCUSSION
■
Prior studies of ferric hydroperoxide species have found that
their formation from ferrous precursors and H2O2 occurs in
two steps: (1) the first /2 equiv of H2O2 oxidizes 1 equiv of
1
Fe(II) to Fe(III), and (2) the remaining H2O2 substitutes for
one of the ligands on the initially generated ferric product to yield
an FeIII−OOH complex.23 The initial rate analysis of the forma-
tion of [Fe(bbpc)(OOH)]2+ (2) from [Fe(bbpc)(MeCN)2]2+
(1) and H2O2 suggests that, at higher concentrations of the
reagents, the rate-determining step (RDS) involves a bimolec-
ular collision between 1 and H2O2. This conclusion is sup-
ported by the highly negative ΔS⧧ value obtained from kinetic
data acquired over several temperatures (Figure 1). Although
the reaction stoichiometry suggests that 1 equiv of H2O2 will
lead to the one-electron oxidation of two Fe(II) centers, the
rate law determined by the initial rate analysis suggests that the
RDS involves only 1 equiv of Fe(II). This would be inconsis-
tent with formation of a diiron peroxide-bridged species, which
may be conceived as the simplest means to oxidize two iron
centers with one molecule of H2O2. The data at low concen-
trations of 1, however, deviate from the linear relationship
established at higher concentrations. One possible explanation
for this deviation is that a subsequent reaction between the 1:1
iron−H2O2 adduct and another 1 equiv of iron may become
the RDS under such conditions.
The influences of various additives on the rates of formation
and decay were assessed. These included two substrates that are
readily oxidized by mixtures of 1 and H2O2: cyclohexane and
cyclohexene.8,24 The results suggest that H2O does not have a
major role in the oxidation of the iron (Figure 2). Indeed, the
only tested additives that influenced the formation of 2 were
HClO4 and NaClO4, which both slowed the rate of formation
of 2 from 1 and H2O2. Excess H2O2 appears to promote the
more rapid degradation of 2 (Figure 3). Although H2O2 could
potentially be a substrate, with O−H bond dissociation energies
of approximately 88 kcal mol−1,29 the inability of cyclohexane
(C6H12) or cyclohexene (C6H10) to hasten the decomposition
of 2 may suggest that the excess H2O2 reacts with 2 by other
means, perhaps through the formation of a more reactive species.
None of the other additives, including the two hydrocarbon
substrates, affect the rate of decomposition of 2 (Figure 3). The
data therefore suggest that 2 is not directly responsible for the
catalyzed C−H activation and that the hydrocarbon oxida-
tion instead occurs during a step subsequent to the RDS of the
FeIII−OOH compound’s decay. This contrasts with recent
results from Nam’s group, who found that the lifetime of a
high-spin FeIII−OOH complex generated with the TMC ligand
decreases as the concentration of an added hydrocarbon sub-
strate increases.17 The addition of H2O or acid likewise fails to
accelerate the disappearance of 2; this is consistent with a decay
process involving homolysis, rather than heterolysis, of the O−
O bond.10,12 Nam’s FeIII−OOH complex likewise decomposes
through O−O homolysis.17
Figure 4. Eyring plot for the decomposition of 2 formed from a
reaction between 0.50 mM 1 and 5.0 mM H2O2 in MeCN at
temperatures ranging from 294 to 324 K. The reactions were followed
using the absorbance at 690 nm. The data from three independent
experiments are plotted. The calculated ΔH⧧ value was 53.2( 1.1) kJ
mol−1. The calculated ΔS⧧ value was −68( 4) J K−1 mol−1.
Generation of [Fe(bbpc)(OOH)]2+ from O2. In studies
using 1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane
(TMC) as a ligand, Nam and co-workers had observed the
formation of ferryl species, prompting them to posit that
the initially generated oxidant was a ferric superoxo species.9
Subsequent work from ourselves with the bbpc ligand had
found that the rate of formation of 2 was first order with
respect to the concentration of C6H10 or an analogous allylic
or benzylic substrate; the rate of formation also slowed when
a deuterated substrate was used.8 On the basis of these
observations plus the detection of organic radicals in these
reaction mixtures, we proposed that the formation of 2
proceeds through a ferric superoxo species, which abstracts a
hydrogen atom from the hydrocarbon to yield 2 and an organic
radical.
The formation of 2 from mixtures of 1, O2, and cyclohexene
was further investigated, with an interest in determining the
reaction’s dependence on O2 and 1. A kinetic analysis was
performed, with the changes in absorbance at 535 nm from 1 to
3 s serving as approximations of the initial rates. The 690 nm
feature does not develop strongly enough in most of these
reactions to serve as a reliable spectroscopic tag for the
concentration of 2, necessitating that we instead monitor the
reactions at 535 nm.8 As the concentration of O2 was increased
with set concentrations of cyclohexene and 1, the changes in
the absorbance from 1 to 3 s are most consistent with a first-
order dependence on [O2]. The changes in absorbance from
1 to 3 s likewise scale linearly with the initial concentration of 1.
The rate law can be described by eq 2.
Additional, albeit weaker, support for homolytic cleavage is
provided by the kinetic parameters (Figure 4). The ΔH⧧ value
for the decay of 2 is more similar to those of non-heme FeIII−
OOH species that undergo homolytic O−O cleavage than to
those that react through O−O heterolysis.12,17,18 Although the
ΔS⧧ value for the decomposition of 2 falls comfortably within
the range previously observed for nonheme FeIII−OOH
species, these values do not distinguish homolytic from hetero-
lytic O−O cleavage.12 The negative ΔS⧧ is curious, since 2
d[2]/dt = k3[1][O2][C6H10]
(2)
D
dx.doi.org/10.1021/ic5003786 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
presumably degrades through a unimolecular reaction. Groves
and Watanabe had invoked the coordination of an additional
ligand to the iron in the transition state to explain the negative
entropies of activation calculated for the O−O bond cleavage
associated with heme peroxo complexes.30,31 We currently have
no data, however, that would corroborate this in our own
system.
Prior results from our laboratory found that O atom ex-
change appears to occur between 2 and H2O, suggesting that
the cleavage of the O−O bond is reversible.8 The reversibility
of the O−O cleavage may indicate that the hydroxyl radical
generated from homolytic cleavage of the OOH ligand remains
associated with the iron-containing product, which is iso-
electronic with a ferryl species (Scheme 3). Whether the two
CONCLUSIONS
■
We have determined that the FeIII−OOH species with the bbpc
ligand decomposes through homolysis of the O−O bond. The
ferric hydroperoxide species itself does not appear to be the
relevant oxidant for C−H activation. The previously observed
regioselectivity of the alkane oxidation and the reversibility of
the O−O cleavage are inconsistent with freely diffusing radicals,
suggesting that any generated hydroxyl radicals remain associ-
ated with the higher-valent iron byproduct. We have also
determined the rate law associated with the formation of 2 from
O2; this rate law is consistent with the previously proposed
intermediacy of a mononuclear ferric superoxo species.
ASSOCIATED CONTENT
■
S
* Supporting Information
Scheme 3
Text, figures, tables, and a CIF file giving crystal structure data
for [Fe(bbpc)Cl](SbF6), initial rate analyses of the formation of
2 from 1, O2, and H2O2, and the peak absorbance at 690 nm as
a function of added HClO4. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail for C.R.G.: crgoldsmith@auburn.edu.
Notes
The authors declare no competing financial interest.
combine into an Fe(V) species prior to re-formation of the
O−O bond remains unclear, and we cannot preclude it as a
possibility at this time. The close association of the hydroxyl
radical with the ferryl species would be a possible explanation
for the observed regioselectivity of the alkane oxidation
catalyzed by 1, which shows a stronger than usual preference
for oxidizing C−H bonds on secondary carbons over those on
tertiary carbons.24 The regioselectivity is inconsistent with the
agency of freely diffusing hydroxyl radicals.6
ACKNOWLEDGMENTS
■
The authors thank Auburn University and the American
Chemical Society Petroleum Research Fund (Grant #49532-
DNI3) for financial support and Prof. David M. Stanbury and
his research group for their assistance with the stopped-flow
experiments and a presubmission review of this manuscript.
The crystal structure of [Fe(bbpc)Cl](SbF6) in the Supporting
Information was solved by Prof. John D. Gorden from a crystal
grown by Mr. Christopher F. Cain.
Reversible O−O cleavage may also explain the discrepancy
between our results and those of Nam’s research group. A rapid
interconversion between ferric species and higher-valent
oxidants would render the FeIII−OOH susceptible to reactions
with oxidizable C−H bonds, at rates approaching those of the
FeIVO oxidant. An unidentified equilibrium between these
two may explain the previously noted similarities in the rates of
hydrocarbon oxidation reactions involving FeIII−OOH and
FeIVO complexes with TMC.17,18
REFERENCES
■
(1) Gunay, A.; Theopold, K. H. Chem. Rev. 2010, 110, 1060−1081.
(2) Bruckl, T.; Baxter, R. D.; Ishihara, Y.; Baran, P. S. Acc. Chem. Res.
̈
2012, 45, 826−839.
(3) Shilov, A. E.; Shul’pin, G. B. Chem. Rev. 1997, 97, 2879−2932.
(4) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L., Jr. Chem. Rev.
2004, 104, 939−986.
The FeIII−OOH species can also be generated from reactions
between 1, O2, and substrates with weak C−H bonds.8 Prior
work from our laboratory found that the rate of the ferric
intermediate’s formation scales with the concentration of the
hydrocarbon substrate. The observation of a primary kinetic
isotope effect with 9,10-dihydroanthracene (kH/kD = 6.8) sug-
gested that the hydrocarbon serves as a hydrogen atom donor,
possibly indicating a ferric superoxo species as an initially
generated oxidant. The formation of 2 from O2 was probed
further, with an interest in determining the nuclearity of the
immediate precursor to 2. The rate law suggests that the oxi-
dation of 1 by O2, with concomitant oxidation of C6H10, pro-
ceeds through a mononuclear, rather than binuclear, iron
species; otherwise, the rate law would be second order in iron.
The rate law is consistent with our previously hypothesized
mechanism, in which a ferric superoxide species abstracts a
hydrogen atom from an allylic substrate in the RDS to yield 2
(Scheme 2).8
(5) Que, L., Jr.; Ho, R. Y. N. Chem. Rev. 1996, 96, 2607−2624.
(6) Chen, K.; Que, L., Jr. J. Am. Chem. Soc. 2001, 123, 6327−6337.
(7) Nam, W. Acc. Chem. Res. 2007, 40, 522−531.
(8) He, Y.; Goldsmith, C. R. Chem. Commun. 2012, 48, 10532−
10534.
(9) Lee, Y.-M.; Hong, S.; Morimoto, Y.; Shin, W.; Fukuzumi, S.;
Nam, W. J. Am. Chem. Soc. 2010, 132, 10668−10670.
(10) Li, F.; Meier, K. K.; Cranswick, M. A.; Chakrabarti, M.; Van
Heuvelen, K. M.; Munck, E.; Que, L., Jr. J. Am. Chem. Soc. 2011, 133,
̈
7256−7259.
(11) Nam, E.; Alokolaro, P. E.; Swartz, R. D.; Gleaves, M. C.; Pikul,
J.; Kovacs, J. A. Inorg. Chem. 2011, 50, 1592−1602.
(12) Oloo, W. N.; Fielding, A. J.; Que, L., Jr. J. Am. Chem. Soc. 2013,
135, 6438−6441.
(13) Rohde, J.-U.; In, J.-H.; Lim, M. H.; Brennessel, W. W.;
Bukowski, M. R.; Stubna, A.; Munck, E.; Nam, W.; Que, L., Jr. Science
̈
2003, 299, 1037−1039.
(14) Wada, A.; Ogo, S.; Nagatomo, S.; Kitagawa, T.; Watanabe, Y.;
Jitsukawa, K.; Masuda, H. Inorg. Chem. 2002, 41, 616−618.
E
dx.doi.org/10.1021/ic5003786 | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(15) Ho, R. Y. N.; Roelfes, G.; Hermant, R.; Hage, R.; Feringa, B. L.;
Que, L., Jr. Chem. Commun. 1999, 2161−2162.
(16) Martinho, M.; Blain, G.; Banse, F. Dalton Trans. 2010, 39,
1630−1634.
(17) Cho, J.; Jeon, S.; Wilson, S. A.; Liu, L. V.; Kang, E. A.; Braymer,
J. J.; Lim, M. H.; Hedman, B.; Hodgson, K. O.; Valentine, J. S.;
Solomon, E. I.; Nam, W. Nature 2011, 478, 502−505.
(18) Liu, L. V.; Hong, S.; Cho, J.; Nam, W.; Solomon, E. I. J. Am.
Chem. Soc. 2013, 135, 3286−3299.
(19) Park, M. J.; Lee, J.; Suh, Y.; Kim, J.; Nam, W. J. Am. Chem. Soc.
2006, 128, 2630−2634.
(20) Price, J. C.; Barr, E. W.; Glass, T. E.; Krebs, C.; Bollinger, J. M.,
Jr. J. Am. Chem. Soc. 2003, 125, 13008−13009.
(21) Krebs, C.; Fujimori, D. G.; Walsh, C. T.; Bollinger, J. M., Jr. Acc.
Chem. Res. 2007, 40, 484−492.
́
(22) Galonic, D. P.; Barr, E. W.; Walsh, C. T.; Bollinger, J. M., Jr.;
Krebs, C. Nat. Chem. Biol. 2007, 3, 113−116.
(23) Roelfes, G.; Lubben, M.; Chen, K.; Ho, R. Y. N.; Meetsma, A.;
Genseberger, S.; Hermant, R. M.; Hage, R.; Mandal, S. K.; Young, V.
G., Jr.; Zang, Y.; Kooijman, H.; Spek, A. L.; Que, L., Jr.; Feringa, B. L.
Inorg. Chem. 1999, 38, 1929−1936.
(24) He, Y.; Gorden, J. D.; Goldsmith, C. R. Inorg. Chem. 2011, 50,
12651−12660; Inorg. Chem. 2012, 51, 7431 (correction).
(25) Chen, K.; Costas, M.; Kim, J.; Tipton, A. K.; Que, L., Jr. J. Am.
Chem. Soc. 2002, 124, 3026−3035.
(26) Decker, A.; Chow, M. S.; Kemsley, J. N.; Lehnert, N.; Solomon,
E. I. J. Am. Chem. Soc. 2006, 128, 4719−4733.
(27) Sawyer, D. T.; Chiericato, G., Jr.; Angelis, C. T.; Nanni, E. J., Jr.;
Tsuchiya, T. Anal. Chem. 1982, 54, 1720−1724.
(28) Britovsek, G. J. P.; England, J.; White, A. J. P. Inorg. Chem. 2005,
44, 8125−8134.
(29) Litorja, M.; Ruscic, B. J. Electron Spectrosc. 1998, 97, 131−146.
(30) Groves, J. T.; Watanabe, Y. J. Am. Chem. Soc. 1986, 108, 7834−
7836.
(31) Groves, J. T.; Watanabe, Y. J. Am. Chem. Soc. 1988, 110, 8443−
8452.
F
dx.doi.org/10.1021/ic5003786 | Inorg. Chem. XXXX, XXX, XXX−XXX