Acid p Ka Dependence in O-O Bond Heterolysis of a Nonheme FeIII-OOH Intermediate to Form a Potent FeVa? O Oxidant with Heme Compound I-Like Reactivity

Article abstract of DOI:10.1021/jacs.9b08442

Protons play essential roles in natural systems in controlling O-O bond cleavage of peroxoiron(III) species to give rise to the high-valent iron oxidants that carry out the desired transformations. Herein, we report kinetic and mechanistic evidence that acids can control the mode of O-O bond cleavage for a nonheme S = 1/2 Fe^III-OOH species [(BnTPEN)Fe^III(OOH)]²⁺ (2, BnTPEN = N-benzyl-N,N′,N′-tris(2-pyridylmethyl)-1,2-diaminoethane). Addition of acids having pK_a values of >8.5 in CH₃CN results in O-O bond homolysis, leading to the formation of hydroxyl radicals that give rise to alcohol/ketone (A/K) ratios of around 1 in the oxidation of cyclohexane. However, the introduction of acids with pK_a values of <8.5 elicits a different outcome, namely the achievement of A/K ratios of as high as 9, the observation of rapid and catalytic hydroxylation of cyclohexane, and a million-fold acceleration in the decay rate of the Fe^III-OOH intermediate at -40 °C. These results implicate the generation of a highly reactive Fe^V? O species via proton-assisted O-O bond heterolysis of the Fe^III-OOH intermediate, which is unprecedented for nonheme iron complexes supported by neutral pentadentate ligands and serves as a nonheme analogue for heme enzyme compounds I.

Full text of DOI:10.1021/jacs.9b08442

Subscriber access provided by Uppsala universitetsbibliotek
Article
Acid pKa Dependence in O–O Bond Heterolysis of a
Nonheme FeIII–OOH Intermediate to Form a Potent
FeV=O Oxidant with Heme Compound I-Like Reactivity
Shuangning Xu, Apparao Draksharapu, Waqas Rasheed, and Lawrence Que
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b08442 • Publication Date (Web): 12 Sep 2019
Downloaded from pubs.acs.org on September 13, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
III
Acid pK Dependence in O–O Bond Heterolysis of a Nonheme Fe –OOH Intermediate
a
V
to Form a Potent Fe =O Oxidant with Heme Compound I-Like Reactivity
Shuangning Xu, Apparao Draksharapu, Waqas Rasheed, and Lawrence Que, Jr.*
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Department of Chemistry and Center for Metals in Biocatalysis, University of Minnesota, 207
Pleasant St. SE, Minneapolis, MN 55455, USA
ABSTRACT
Protons play essential roles in natural systems in controlling O–O bond cleavage of
peroxoiron(III) species to give rise to the high-valent iron oxidants that carry out the desired
transformations. Herein we report kinetic and mechanistic evidence that acids can control
III
the mode of O–O bond cleavage for a nonheme S = ½ spin Fe –OOH species
III
2+
[
(BnTPEN)Fe (OOH)] (2, BnTPEN = N-benzyl-N,N’,N’-tris(2-pyridylmethyl)-1,2-diamino-
ethane). Addition of acids having pK values of > 8.5 in CH CN results in O–O bond homolysis,
a
3
leading to the formation of hydroxyl radicals that give rise to alcohol/ketone (A/K) ratios of
around 1 in the oxidation of cyclohexane. However, the introduction of acids with pK values
a
of < 8.5 elicits a different outcome, namely the achievement of A/K ratios of as high as 9, the
observation of rapid and catalytic hydroxylation of cyclohexane, and a million-fold
III
acceleration in the decay rate of the Fe –OOH intermediate at -40 °C. These results implicate
V
the generation of a highly reactive Fe =O species via proton-assisted O–O bond heterolysis
III
of the Fe –OOH intermediate, which is unprecedented for nonheme iron complexes
supported by neutral pentadentate ligands and serves as a nonheme analog for heme
enzyme Compounds I.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Introduction
Iron-containing enzymes are versatile catalysts that utilize oxygen to carry out a variety
of oxidative transformations, the most challenging of which is the functionalization of very
strong C–H bonds.1-5 Activation of these inert C–H bonds through hydrogen atom transfer
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
HAT) is often accomplished by high-valent iron-oxo species, which can be generated by
cleaving the O–O bonds of corresponding peroxoiron intermediates derived from O or H O .
2
2 2
One of the best characterized high-valent species is generated by heme-containing enzymes
6
IV
+
and known as “Compound I” (Cpd I), an Fe (O)(por• ) species proposed to be the active
oxidant in the catalytic cycles of peroxidases and cytochromes P450 (Scheme 1).7, 8 Its
III
precursor, called “Compound 0” (Cpd 0), is an S = ½ Fe –OOH species that gives rise to Cpd
I through heterolytic scission of its O–O bond.9-12 This key step requires the participation of
one proton to allow the distal oxygen to leave as water. For H O -utilizing heme peroxidases,
2
2
introduction of the proton to the distal oxygen is proposed to be accomplished by amino acid
residue pairs, which serve as general acid-base catalysts and are located at the distal pockets
of the heme active sites of the enzymes.13-16 For dioxygen-utilizing cytochrome P450-type
enzymes, threonine has been identified as the key residue in the proton delivery networks
required to generate Cpd 0 and Cpd I.17-20
Heme systems
H
OH
O
O
H
O
O
N
N
H+
N
N
-H2O
N
N
FeIII
FeIII
FeIV
N
N
N
N
N
N
S
S
S
Cys
Cys
Cys
Cpd 0
Cpd I
Water-assisted mechanism in nonheme systems
N
N
N
N
FeV
FeV OH
H2O
FeIII
O
-H O
O
N
2
FeIII OOH
N
CH3CN
N
OH
O
N
O
O H
NCCH3
H
H
This work
O
H
H
H
O
O
N
N
FeIII
O
FeIII
N
N
O
N
FeV
H+
N
-H O
N
2
N
N
N
N
N
N
N
N
Scheme 1. Proposed mechanism for proton-assisted activation of the O–O bonds of
hydroperoxoiron(III) species.
2
ACS Paragon Plus Environment

Page 3 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
How a proton is involved in the activation of ferric hydroperoxo species remains a
fundamental mechanistic question that needs further investigation. Although there are
extensive studies on the O–O bond cleavage of heme ferric acylperoxo and alkylperoxo
2
1
III
complexes, there are only a few reports that elucidate the effect of protons on heme Fe –
OOH species, and most only provide indirect evidence by either tracking the hydroperoxo-
derived high-valent iron species or the outcome of oxidation reactions.22-24 Van Eldik and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
III
coworkers have studied an Fe –OOH species supported by an octa-anionic porphyrin ligand
in aqueous medium and found that the outcome of O–O bond cleavage switches at pH 8.5,
forming either the two-electron oxidized Cpd I at lower pH or the one-electron oxidized
IV
24
Fe (O) species at higher pH. Similar behavior is observed for the manganese counterpart
of Cpd 0 in a study by Groves and coworkers, where it is suggested that the protonation and
V
release of the distal oxygen as water occur concomitantly with the formation of the Mn (O)
species.25
In nonheme iron systems, the action of the anti-cancer drug bleomycin (BLM) has been
III
found to involve an S = ½ Fe –OOH species known as activated bleomycin (ABLM), which is
proposed to initiate double-stranded DNA cleavage by attacking the C4’–H bond of the target
nucleotide in DNA.26-28 However, to what extent the mechanism of action for this Fe –OOH
species parallels that of its heme counterpart Cpd 0 remains unclear. While calculations
show that protonation of the peroxo moiety can lower the barrier for O–O bond cleavage and
III
2
9
promote the formation of high-valent iron species, it has been suggested that direct HAT
III
by this nonheme Fe –OOH intermediate may be the most favorable pathway for initiating
double-stranded DNA cleavage.30-32
III
Many of the synthetic nonheme Fe –OOH complexes reported thus far are those
supported by neutral pentadentate ligands such as N4Py (N4Py = N,N-bis(2-pyridylmethyl)-
N-bis(2-pyridyl)methylamine) and R-TPEN (R-TPEN = N-R-N,N’,N’-tris(2-pyridylmethyl)-
1
,2-diaminoethane), which are generated by adding a large excess of H O to their iron(II)
2 2
precursors at low temperatures.33-41 These Fe –OOH species themselves have been
described as “sluggish oxidants” of hydrocarbons,42 and the lack of oxidative selectivity
exhibited by these species has often been attributed to the action of hydroxyl radicals
III
III
41-44
generated via homolysis of the O–O bond of the Fe –OOH intermediate.
Currently,
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
III
experimental evidence for proton-assisted O–O bond heterolysis of nonheme Fe –OOH
species has been found only for those supported by tetradentate ligands such as TPA (TPA =
tris(pyridyl-2-methyl)amine) and Me2PyTACN (Me2PyTACN = 1-(pyridyl)-2’-methyl]-4,7-
dimethyl-1,4,7-triazacyclononane). For these complexes, the availability of an additional
coordination site allows water to bind cis to the hydroperoxo ligand and promote the
heterolytic cleavage of the O–O bond via a so-called “water-assisted” mechanism to generate
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
V
45-54
a potent Fe (O)(OH) oxidant (Scheme 1).
More recent studies on two other tetradentate
III
2+
ligand complexes, namely [(PyNMe )Fe (OOH)] (PyNMe = 3,6,9-trimethyl-3,6,9-triaza-
3
3
(2,6)-pyridina-cyclodecaphane) and [(β-BPMCN)Fe (OOH)]2+ (β-BPMCN = N,N’-
III
1
bis(pyridyl-2-methyl)-N,N’-dimethyl-trans-1,2-diaminocyclohexane) indicate that O–O bond
V
55, 56
cleavage can be facilitated by strong acids to form Fe =O species.
While the specific role
the sixth site plays in the O–O bond cleavage step and the precise formulation of the actual
oxidant remain unclear, these observations strongly suggest that protons may also play a
crucial role in O–O bond cleavage in nonheme iron chemistry.
In our continuing effort to elucidate the role protons may play in O–O bond cleavage of
III
Fe –OOH species, we have turned to the previously well-characterized
III
2+
[
(BnTPEN)Fe (OOH)] complex (2, BnTPEN = N-benzyl-N,N’,N’-tris(2-pyridylmethyl)-1,2-
diaminoethane)35, 36, 38 and investigated the effect of adding various acids on the fate of the
III
Fe –OOH species. We demonstrate that the mode of O–O bond cleavage depends on the pKa
value of the added acid and present indirect evidence for the formation of an unprecedented
V
3+
[
(BnTPEN)Fe (O)] species via acid-assisted O–O bond heterolysis (Scheme 1), which is
capable of facile hydroxylation of saturated hydrocarbons as well as benzene. This study on
III
the effect of adding acid to the metastable Fe –OOH species may shed light on the
mechanisms of proton-dependent formation of high-valent iron-oxo species in nature.
4
ACS Paragon Plus Environment

Page 5 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Results and Discussion
Formation of 2 with HOAc and H O at 25 °C
2
2
II
Intermediate 2 can be generated by the reaction of [(BnTPEN)Fe (OTf)](OTf) (1) with
5
7
excess H O in methanol or acetone as previously reported; these are conditions typical for
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
III
41
the formation of Fe –OOH species supported by pentadentate ligands. Its formation is
associated with the appearance of an absorption maximum at 525 nm arising from a peroxo-
3
3
to-iron(III) charge transfer transition and an EPR signal with g = 2.22, 2.16 and 1.96 that is
III
57
typical of an S = ½ Fe center. In the absence of other additives, a large excess of H O (100
2
2
equiv) is needed to obtain a maximum absorbance of ~0.3 from a 1-mM solution of the
ferrous precursor in CH CN at 25 °C (Fig S1). Under these conditions, the EPR spectrum of
3
III
this solution shows a mixture of the desired Fe –OOH intermediate 2 and the corresponding
III
33
Fe –OH complex (g = 2.40, 2.18, and 1.90), together accounting for only about 20% of the
iron in solution based on double integration of the low-spin signals (Fig S1). However, we
have found that the addition of 10 equiv HOAc (pK = 23.5 in CH CN) to the solution prior to
a
3
the introduction of H O allows 2 to be generated using only 10 equiv 90% H O at an amount
2
2
2 2
representing about 50% of the iron in the reaction mixture with no evidence for any other
low-spin ferric moieties in the EPR spectrum (Fig S1). This species exhibits a somewhat
sharper UV-vis spectrum with about twice as large an absorbance of ~0.6 at 525 nm (Fig 1).
The absence of other S = ½ ferric species under these conditions has enabled us to reliably
-1
-1
quantify the S = ½ EPR signals of 2 and establish an extinction coefficient of 1100 M cm at
525 nm when used in conjunction with the UV-vis absorption spectrum of the same sample
(
Fig 1, Fig S1). To the best of our knowledge, this result represents the first time that a molar
extinction coefficient has been determined for 2, which agrees well with the value of 1000
-1
-1
III
2+
M cm at 520 nm for the related [(MeTPEN)Fe (OOH)] species (MeTPEN = N-methyl-
N,N’,N’-tris(2-pyridylmethyl)-1,2-diaminoethane) as determined by Banse and coworkers.58
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. (a) UV-vis spectral changes and (b) time traces associated with the formation
and self-decay of 2 in the presence of HOAc (red) and the reaction of 2 with 1 equiv HOTf to
form 3 (blue). Reaction conditions: 10 equiv 90% H O is added to 1 mM 1 (black) in CH CN
2
2
3
in the presence of 10 equiv HOAc; 1 equiv HOTf introduced to this reaction mixture ~30 s
after H O addition in CH CN at 25 °C.
2
2
3
This notable change in speciation prompted us to further investigate what role acetic acid
1
may play in this chemistry. In our efforts, we determined that H-NMR spectroscopy should
1
be quite a useful tool for addressing this question. The H-NMR spectrum of 1 in CD CN only
3
shows features in the diamagnetic region (Fig. 2b), which are broader than those of the free
ligand (Fig 2a). This broadening is likely the result of a rapid equilibrium between the
II
2+
59
predominant diamagnetic low-spin (S = 0) [Fe (BnTPEN)(NCCD )] complex in solution
3
II
+
60
and a minor amount of the paramagnetic high-spin (S = 2) [Fe (BnTPEN)(OTf))] species.
6
ACS Paragon Plus Environment

Page 7 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
In contrast, the NMR spectrum of 1 in CD Cl (Fig 2e) shows much broader peaks spread over
2
2
a much larger chemical shift range, as would be expected for the high-spin (S = 2)
II
+
[Fe (BnTPEN)(OTf))] complex.
No significant change in chemical shifts or linewidths is observed when 10 equiv HOOCCD3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
is added to a solution of 1 in CD CN, suggesting that acetic acid has no significant effect on
3
the S = 0/S = 2 equilibrium (Fig 2c). However, there is a new peak at ~9 ppm (Fig 2c), which
corresponds to the OH proton of acetic acid (Fig S2). Under these conditions, there is also no
evidence for the dissociation of a pyridyl arm. However, evidence for pyridyl arm
dissociation is observed upon addition of 1 equiv DClO , where a doublet at 8.5 ppm
4
corresponding to the α-pyridyl proton of the dissociated arm appears (Fig 2d and S2). Thus
the BnTPEN ligand remains pentadentate in 1 in the presence of 10 equiv acetic acid in
CD CN, and the iron(II) complex has only one available coordination site.
3
*
(a)
(b)
(c)
(d)
(e)
#
#
*
10
9
8
7
6
5
4
3
2
1
5
0
60 55 50 45 40 35 30 25 20 15 10
0
δ(ppm)
1
Figure 2. Top Panel: H-NMR spectra in the range of 10.5 to -0.5 ppm in CD CN (unless
3
specified otherwise) of (a): the BnTPEN ligand, (b): 2.5 mM 1, (c): (b) + 25 mM CD COOH,
3
1
(
d): (b) + 10 mM CD COOH and 6 mM DClO , and (e): 2.5 mM 1 in CD Cl . Bottom panel: H-
3
4
2
2
NMR spectrum of 2.5 mM 1 in CD Cl in the range of 62 to -4 ppm. # indicates the OH signal
2
2
from CD COOH. * indicates the α-H signal of a dissociated pyridyl arm upon acidification; see
3
Fig S2 for more spectra illustrating this feature.
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
As for the low-spin ferric species in solution, addition of 1-400 equiv acetic acid has no
-1
effect on the decay rate of 2 (0.011(1) s in CH CN at 25 °C; see Table S1), suggesting that
3
III
acetic acid does not interact with the Fe –OOH species. However, given that the addition of
III
acetic acid eliminates the signals corresponding to the S = ½ Fe –OH species and increases
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
III
the amount of S = ½ Fe –OOH species (2) formed (Fig S1), we propose that the role of acetic
III
acid is to provide the weakly acidic medium needed to protonate the S = ½ Fe –OH species,
III
converting it into the S = 5/2 Fe –OH species (Scheme 2), by analogy to the closely related
2
III
3+
61
[
(N4Py)Fe (OH )] complex described by Browne and coworkers. In this process, the
2
III
kinetically more inert low-spin Fe –OH species is replaced by the kinetically more labile
III
high-spin Fe -OH species that then reacts rapidly with H O to achieve the higher steady-
2
2 2
III
III
state concentration of 2 observed (Fig 1). The corresponding Fe –O–Fe species may also
be present in the reaction solutions and be involved in these acid-base transformations.
Scheme 2. Proposed role of acetic acid in the formation of 2.
H O
slow
2
2
.5 equiv (N5)FeIII OH
0
H O
S = 1/2
2
2
(N5)FeIII OOH
H+
(
N5)FeII NCCH3
(
S = 1/2
H O
2
2
N5)FeIII OH2
fast
S = 5/2
III
Formation and Characterization of the Intramolecular Phenolate Species Fe (OAr) (3)
Interestingly, when 2 generated with H O /HOAc reaches its maximum concentration at
2
2
2
5 °C, the addition of 1 equiv HOTf (pK = 2.6 in CH CN) with respect to 1 results in the decay
a 3
of 2 within a second (Fig 1a, dashed blue trace), concomitant with formation of 3, which
reaches maximum formation in ~300 s (Fig 1). The latter has an intense broad absorption
feature spanning from 500-1000 nm with a λmax at 680 nm and can be identified as the
intramolecular phenolate product 3. This is a result of ortho-hydroxylation of the pendant
benzyl ring on the BnTPEN ligand, which converts the ligand into a hexadentate ligand with
5 neutral N donors and 1 anionic O donor. Coordination of the nascent hexadentate ligand to
the iron(III) center gives rise to the phenolate-to-iron(III) charge-transfer band associated
8
ACS Paragon Plus Environment

Page 9 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
with 3 (λmax = 680 nm), which is significantly red-shifted relative to that observed for the (µ-
oxo)diferric complex supported by the pentadentate ligand 2-(((di(pyridin-2-yl)methyl)-
(pyridin-2-ylmethyl)amino)methyl)phenolate) with λmax ~540 nm.62, 63 This red shift reflects
III
the higher Lewis acidity of the Fe center of 3 due to the absence of the highly basic µ-oxo
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
bridge found in the Feringa complex. The rate of formation of 3 and its maximum yield are
found to be approximately linearly dependent on the concentration of 1 (Fig S3, left panel),
indicating that the hydroxylation reaction is unimolecular in nature. Doubling the amount of
HOAc or H O added does not significantly affect the formation rate of 3, but doubling the
2
2
amount of HOTf added clearly has a deleterious effect on the formation of 3 (Fig S3, right
panel).
Given the constraints of the ligand framework, intramolecular hydroxylation at the ortho-
position of the pendant benzyl ring is the most likely site for attack by an iron-oxo species
generated at the active site. This conclusion is supported by electrospray ionization (ESI)
mass spectrometry (Fig 3, Fig S4), which shows a molecular ion at m/z 643.0 and an isotope
III
+
distribution pattern consistent with its formulation as [(BnTPEN)Fe + OTf + O – H] . We
did not find ions that would support the presence of the corresponding oxo-bridged dimer
in solution.
Furthermore, its resonance Raman spectrum is typical of iron(III) centers with an ortho-
substituted phenolate ligand (Fig 3, Fig S5),63-66 which exhibit a feature around 600 cm that
-1
-1
is associated with an Fe–OAr mode and vibrations in the 1100 to 1600 cm region that
6
6
correspond to phenolate ring deformation modes. This spectrum exhibits more phenolate
III
ring deformations than found for Fe –OPh complexes due to the loss of two-fold symmetry
III
–
upon ortho-substitution of the phenolate ligand, as observed for [Fe (EDDHA)] (H EDDHA
4
=
ethylenediamine-di(2-hydroxyphenylacetic acid)).67 The molar absorptivity of
III
-
-1
-1 68
[
Fe (EDDHA)] , which has two phenolates bound to the iron(III) center, is 4800 M cm ,
III
-
about twice the number found for [Fe (o,p-EDDHA)] (H -o,p-EDDHA = ethylenediamine-
4
-1
-1
N(o-hydroxyphenylacetic)-N’(p-hydroxyphenylacetic) acid) (2130 M cm ), where there is
only one iron(III)-phenolate moiety.69 Based on the molar absorptivity coefficients
associated with a number of iron(III)-ortho-phenolate compounds, we estimate the molar
-1
-1
absorptivity coefficient of the 680-nm chromophore of 3 to be approximately 2000 M cm .
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Although intramolecular aromatic hydroxylation has been demonstrated for related
nonheme iron complexes,70-75 this report is, to the best of our knowledge, the first time that
intramolecular aromatic hydroxylation has been demonstrated for 1.
The effects of other acid additives with pK values ranging from 0.2 to 23.5 in CH CN have
a
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
been further investigated. These results reveal that the acids studied can be subdivided into
two subsets based on the yield of the intramolecular phenolate product formed (Fig S6, Table
S2). The introduction of HOTs•H O (pK = 8.45 in CH CN) or stronger acids generates 3 in
2
a
3
high yields (A680 ~1 derived from a solution of 1 mM 1) with no observed accumulation of 2,
whereas the addition of HNO (pK = 8.8 in CH CN) or weaker acids leads to much lower
3
a
3
yields of 3 (A680 = ~0.1). In the latter cases, the initial accumulation of 2 is also observed in
the presence of Cl CHCOOH (pK = 13.2) or weaker acids. The fact that 2 does not accumulate
2
a
in the presence of strong acids is consistent with the high reactivity observed between 2 and
HOTf (Fig 1), suggesting that the strong acid activates the initially formed but unobserved
III
Fe –OOH species to facilitate the formation of a potent oxidant capable of electrophilic
aromatic hydroxylation (EAS) of the pendant phenyl group.
1
0
ACS Paragon Plus Environment

Page 11 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
III
Fe (OPh)
635
1282
1483
1165
1175
1300
613
3
6
58
6
1158
65
1244
1266
1
229
1134
641
1
114 1193
608
1486
4
00 500 600 1100 1200 1300 1500 1600
-1
Raman Shift (cm )
Figure 3. Top: The expanded molecular ion region of the ESI-MS spectrum of the
intramolecular phenolate product 3. See Fig S4 for the full mass spectrum of this species.
Bottom: comparison of resonance Raman spectra of the intramolecular phenolate product 3
III
2+
and the intermolecular phenolate product [(BnTPEN)Fe (OPh)] formed in the presence of
benzene. Resonance Raman spectra were obtained in CH CN at 77 K with 561-nm laser
3
excitation. Reaction conditions: for 3: 1 mM 1 for ESI-MS or 2 mM 1 for resonance Raman,
III
1
0 equiv HOAc, 1 equiv HOTf in CH CN at room temperature. Fe (OPh) is generated in the
3
same fashion at 1 mM 1 and 100 equiv benzene in solution. See Fig S5 for full spectra.
Effect of Acid pK on Cyclohexane and Benzene Hydroxylation at Room Temperature
a
The highly oxidizing nature of this new oxidant generated in the presence of strong acids
is demonstrated by the facile intermolecular oxidation of cyclohexane, a hydrocarbon
-1 76
substrate with very strong C–H bonds (BDE = 99.5 kcal.mol ) (Fig 4, Table 1). The use of
1
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HOTf (pK = 2.6 in CH CN) results in the catalytic formation of cyclohexanol (3.6 turnovers
a
3
from 10 equiv 90% H O ) with an alcohol-to-ketone (A/K) ratio of 9. The high A/K ratio
2
2
V
argues for a mechanism involving the generation of a highly reactive Fe =O oxidant that
performs HAT on the substrate C–H bond directly followed by rapid rebound of the nascent
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
IV
77
alkyl radical with the incipient Fe –OH moiety. Acids with comparable pK values, such as
a
HBF •Et O (0.2) and HClO (1.57), give outcomes with similarly high A/K ratios (Fig. 4 and
4
2
4
Table 1). The observed high A/K ratio is not significantly affected by the presence of O as
2
shown in the case of HOTf addition (Fig S7), indicating that the alkyl radicals if formed must
rebound too quickly to be intercepted by O . Under these conditions, the reaction mixtures
2
turn light blue during the course of the reactions, suggesting that intramolecular aromatic
hydroxylation occurs to some extent. On the other hand, the addition of weaker acids having
pK ’s >8.5 affords smaller amounts of cyclohexanol and cyclohexanone but in comparable
a
yields, so the A/K ratios approach 1; such an outcome is typically associated with the
formation of long-lived carbon-centered radicals.78 Therefore, as the pK decreases
a
from >8.5 to <2, a transition from low to high A/K selectivity is observed that generally
correlates with the pK of the added acid (Fig 4).
a
1
2
ACS Paragon Plus Environment

Page 13 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table 1. Cyclohexane oxidation by 1 with H O in the presence of acids at room temperature.
2
2
10 eq HOAc
OH
O
10 eq 90%H O2
2
1
eq acid
II
III
Fe (BnTPEN) +
+
+ Fe (OAr)
CH CN, RT, 2h
1
3
3
Acid
pKa*
Alcohol Ketone A/K
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(1 equiv)
HBF •OEt
(TON)
3.2(2)
(TON)
0.4(1)
Ratio
8
0.2
4
2
HClO (70%)
1.57
2.6
3.2(1)
3.6(1)
3.3(3)
3.5(1)
3.3(1)
2.9(1)
4.2(2)
3.8(1)
2.3(2)
3.3(1)
2.9(5)
2.1(1)
1.6(1)
2.2(1)
1.5(2)
2.4(3)
1.6(1)
1.4(2)
1.5(2)
1.5(1)
0.7(2)
0.5(1)
0.4(1)
0.5(1)
0.4(1)
0.4(1)
0.6(1)
0.9(1)
0.8(1)
0.8(1)
0.8(1)
0.7(2)
0.7(1)
1.0(3)
0.6(1)
1.5(1)
3.3(4)
1.4(1)
1.4(1)
1.3(1)
1.2(1)
1.1(1)
6
4
HOTf
9
+
+
2
-Cl-4-NO -An
3.22
3.68
4.8
7
2
5-Cl-2-NO -An
9
2
+
2
-NO -An
8
2
+
2
,5-Cl -An
6.21
6.79
7.2
5
2
2-Cl-pyH+
5
H SO
2
5
4
2-Cl-An+
7.86
8.03
8.39
8.45
8.8
3
+
4
-CF -An
4
3
+
2
,4-F -An
4
2
HOTs•H O
3
2
HNO3
4-Br-An+
2
9.43
10.75
12.53
13.2
14.13
18.8
23.5
—
4
Cl CCOOH
1
3
PyH+
0.7
1
Cl CHCOOH
2
+
2
,6-CH -PyH
1
3
ClCH COOH
1
2
HOAc only
1
No acid additive
0.7
An = anilinium. BF is the counteranion of all pyridinium and anilinium salts. *pK values
4
a
obtained from ref.79, 80 TON = turnover number with respect to the catalyst. A/K ratio
calculated by dividing cyclohexanol TON by cyclohexanone TON. The intramolecular
phenolate product 3 is not quantified for catalysis reactions. Maximum A680 values
representing the yields of 3 obtained without cyclohexane as the competing substrate are
listed in Table S2. Reaction conditions: 1 mM 1, 10 equiv HOAc, 1 equiv various acids, 10
equiv 90% H O and 1000 equiv cyclohexane in CH CN at room temperature. The solutions
2
2
3
were prepared in a N glovebox except for acid and H O stock solutions, which together
2
2 2
1
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
account for 2% of the total volume of the reaction mixture. The results are the average of at
least two trials.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Cyclohexane oxidation product distributions catalyzed by 1 in the presence of
various acids. See reaction conditions under Table 1.
The O-based acids in this study can also be replaced by N-based acids such as pyridinium
and anilinium salts, which exert similar effects as the O-based acids with comparable pKa
values (Table 1), indicating that the observed reactivity depends mainly on the acidity of
protons added. This pK -dependent behavior in cyclohexane oxidation reactivity also
a
coincides with the observation of significant intramolecular aromatic hydroxylation
reactivity for acids with pK values below 8.5 (Fig S6), indicating that this putative metal-
a
based oxidant is capable of cleaving strong C–H bonds as well as electrophilic aromatic
hydroxylation.
All the experiments discussed above have been carried out with dry reagents and solvent,
with the only water introduced coming from the 10 equiv 90% H O added, which contains
2
2
about 2 equiv H O. With this small amount of water present, the A/K ratio from cyclohexane
2
oxidation in the presence of HOAc/HOTf is not significantly affected by the presence of O2
(
Fig S7). However, increasing the amount of water present in the reaction mixture lowers
the A/K selectivity of 1 in cyclohexane oxidation. For example, the A/K ratio of 8 obtained
1
4
ACS Paragon Plus Environment

Page 15 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
from a reaction mixture containing HOAc/HOTf and 10 equiv 90% H O reduces to 5 when
2
2
1
00 equiv water is added (Fig S8), which approximates the amount of water present in 10
equiv 30% H O (88 equiv H O). Increasing the amount of water present further decreases
2
2
2
the A/K ratio such that it becomes unity with 1000 equiv H O, showing that the nature of the
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
oxidant is altered dramatically in the presence of larger amounts of water. The deactivating
effect of water observed in the case of 1 may arise from its leveling effect, which attenuates
the electrophilicity of the proton and favors O–O bond homolysis of 2 over its heterolysis.
V
The generation of an Fe =O oxidant in the presence of strong acids is further corroborated
1
8
18
by labeling studies with H O or H O (Table 2). Cyclohexane hydroxylation performed
2
2
2
1
8
18
16
under an inert atmosphere with 10% H O (90% O-enriched in H O, a 10-equiv aliquot
2
2
2
1
8
16
of which would contain 10 equiv H O and ~200 equiv H O), 1 equiv HOTf, and 10 equiv
2
2
2
18
HOAc affords cyclohexanol products with 93(1)% and 6(1)% O-incorporation respectively
1
8
18
from H O and H O, together accounting for 100% of the oxygen atom found in the
2
2
2
1
8
alcohol product. Under air, the reaction affords cyclohexanol with 69% O-incorporation
1
8
18
18
from H O and 4% O-incorporation from H O, the remaining 27% product presumably
2
2
2
incorporating its O-atom from O . When the 1-equiv-HOTf additive is excluded, only 10%
2
1
8
18
O-incorporation from H O into cyclohexanol is observed, with negligible incorporation
2
2
1
8
18
16
of O from H O in the complementary experiment with 10 equiv 90% H O (containing
2
2
2
1
6
18
18
18
2
equiv H O) and 200 equiv H O. The high degree of O-labeling from H O in the
2 2 2 2
alcohol product in the presence of HOTf is consistent with a C–H bond hydroxylation
7
7
mechanism involving HAT by a metal-oxo species followed by rapid rebound. However, the
1
8
larger amount of water present in the 10% H O solution diminishes the electrophilicity of
2
2
the HOTf proton and opens the door for a competing pathway that generates a longer lived
alkyl radical that can be intercepted by O , hence decreasing the A/K ratio.
2
1
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
18
Table 2. Effect of added water on the A/K product ratios and the % O-incorporation into
the cyclohexanol product.
A/K
Equiv H218O2
H O
2
H218O
Ref
added
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
+ HOTf + HOAc under Ar
+ HOTf + HOAc
5
9
5
4
3
1
200
0 a
93(1)
6(1)
–
–
–
1 + HOTf + HOAc
100
200
400
200
1.4(2)
4.2(1)
6.9(2)
0.9(1)
This work
This work
69(4)
-
1 + HOAc
10(1)
II
Fe (CH CN) (OTf)
2
1
200
200
–
–
0.5(3)
0.6(1)
3
2
2
II
Fe (CH CN) (OTf) + HOTf + HOAc
3
2
2
b
Fe(N4)(CH CN) (ClO )
4 2
3
2
4
4
5
5
N4 = TPA
5
9
1000
1000
400
70
69
–
27
38
30
18
42
N4 = 5-Me TPA
3
45
-
8
1
N4 = BPMEN
N4 = Me2PyTACN
8
1000
1000
84
47
47
12
a
16
16
The only water present is the 2 equiv H O in the 10 equiv of 90% H O added. This entry
2
2
2
has been included to show the effect of added water in the reaction. Reaction conditions are
listed under Table 1.
b
45, 81
47
18
For these reactions, either 10 equiv of 30%
or 35% H O was used for H O labeling
2 2 2
studies. About 100 equiv of H O was present in each of these reaction mixtures. BPMEN =
1
6
2
N,N′-bis(pyridylmethyl)ethylenediamine. Reaction conditions: 1 mM Fe complex or salt,
with or without 1 equiv HOTf and/or 10 equiv HOAc, 10 equiv 10% H O (90% 18O
enriched in H O, contains about 200 equiv H O when adding 10 equiv) or 10 equiv 90%
H O along with 200 equiv H O, and 1000 equiv cyclohexane in CH CN at room
1
8
2
2
1
6
16
2
2
16
18
2
2
2
3
temperature in air.
This deleterious effect of added water in the case of 1 is quite different from the observed
positive effect of added water in Fe(TPA)-catalyzed oxidations.45, 46, 50 For the latter, addition
1
6
ACS Paragon Plus Environment

Page 17 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
of 400 equiv water accelerates the rate of heterolytic O–O bond cleavage by as much as 7-
5
0
fold and results in a 30% incorporation of oxygen label from water into the cyclohexanol
4
5
product. In significant contrast, the presence of just 50 equiv water cuts the rate of HOTf
-1
(
1 equiv)-assisted heterolytic O–O bond cleavage of 2 at -40 °C from kobs = 0.9(1) s without
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
added H O to 0.015(1) s with 50 equiv H O, for a 60-fold difference (Fig S9, detailed kinetics
2
2
discussions of the decay of 2 in the following section). Also, as shown in Table 2, only a small
1
8
fraction of the alcohol product from 1-catalyzed cyclohexane hydroxylation incorporates O
1
8
label from the added H O. These results suggest the participation of a minor pathway
2
related to the water-assisted mechanism established for Fe(TPA)-catalyzed hydrocarbon
V
oxidations to generate an Fe (O)(OH) oxidant. Such a pathway would likely require
dissociation of a pyridyl arm, as previously documented for complexes of TPEN and related
ligands,82-84 in order to open up a coordination site for the water to bind. However, given the
1
8
18
small extent of O incorporation from H O, such a pathway would likely not be the major
2
pathway for cyclohexane hydroxylation by 1.
We have also investigated the use of benzene as an intermolecular substrate for EAS to
further probe the nature of this oxidant (Fig 5). About 5 TON of phenol product is obtained
in the presence of HOAc/HOTf, consistent with the intramolecular EAS reactivity exhibited
by the complex. Competition experiments between benzene and cyclohexane at a 1:1 molar
ratio shows a 5:1 preference for hydroxylation of benzene over cyclohexane. Similar
V
preferences for EAS over HAT have been reported for related systems where Fe =O species
have been proposed to be the active oxidant.56, 85 Reactions using a 1:1 mixture of benzene
and perdeuterated benzene (Table 3) reveal a product kinetic isotope effect (PKIE) of 0.9,
demonstrating preferential oxidation of the deuterated isotopomer. Such an inverse PKIE is
V
consistent with electrophilic attack of the aromatic ring by an Fe =O species, where the PKIE
2
would originate from a change in hybridization of the hydroxylated carbon atom from sp to
3
58, 86-93
sp during the formation of a σ-complex.
Comparable amounts of phenol products, about 5 TON, are also found in the oxidation of
benzene by 1 with no acid added or with only HOAc introduced (Fig 5). Even when 1 is
II
replaced with an iron salt Fe (CH CN) (OTf) to induce Fenton-type reactions, about 4 TON
3
2
2
of phenol can still be obtained (Table S3). These observations show that benzene
1
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
18
hydroxylation to phenol can be readily carried out by hydroxyl radicals. O-labeling studies
suggest that the majority of oxygen atoms found in the phenol product originates primarily
from H O regardless of whether the reaction is performed in the presence of strong acids
2
2
(
Table S4). However, in sharp contrast to the reactions in the presence of strong acids,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
competition experiments between benzene and cyclohexane with no acid or HOAc result in
the preferential oxidation of cyclohexane, with A/K product ratios similar to that obtained
in the absence of benzene (Fig 5, Table 1). A similar preference towards cyclohexane
II
oxidation is observed for Fe (CH CN) (OTf) (Table S3). Furthermore, PKIE values of 1.0-1.2
3
2
2
II
are observed for benzene oxidation with Fe (CH CN) (OTf) or with 1 in the absence of acid
3
2
2
or in the presence of HOAc, resembling that obtained for benzene hydroxylation by
Fe(MeTPEN) and H O reported by Banse and coworkers, where the hydroxyl radicals are
2
2
proposed to be the active oxidant.40, 58 Taken together, the drastic differences in HAT/EAS
V
reactivity and PKIE clearly distinguish the Fe =O oxidant generated in the presence of strong
acids and radicals generated in the presence of no acid or weak acids.
phenol
6
-
-
ol
one
5
4
3
2
1
0
no HOAc HOAc/
acid HOTf
no acid
HOAc HOAc/HOTf HOAc/
HOTf
1000 equiv
benzene
500 equiv cyclohexane
and 500 equiv benzene
1000 equiv
cyclohexane
Figure 5. Benzene hydroxylation by 1/H O under different conditions and in competition
2
2
with cyclohexane. Reaction conditions: 1 mM 1 and 10 equiv 90% H O with or without 10
2
2
equiv HOAc and 1 equiv HOTf in CH CN at room temperature in air. A small amount of phenol
3
product can be observed in the presence of 500 equiv cyclohexane and 1000 equiv benzene
in the absence of any acid or upon addition of HOAc only, see Table S5.
1
8
ACS Paragon Plus Environment

Page 19 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table 3. PKIEs for hydrocarbon oxidations by 1 and related complexes
Benzene Cyclohexane
Reactions at 25 °C
Ref
(EAS)
(HAT)
a
a
1.0
1.8
1
+ H O
2 2
1
+ HOAc + H O
1.2
1.9
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
a
1
+ HOAc + H O + HOTs•H O
0.9
0.9
0.8
0.9
2.1
2.3
3.2
-
2
2
2
1
+ HOAc + H O + HOTf
2 2
II
45, 90
91
Fe (BPMEN)(CH CN) (ClO ) + H O
3
2
4 2
2
2
II
b
[
Fe (Py NHC )(CH CN) ](PF )
2
2
3
2
6 2
III
III
[(PDP*)Fe (μ-OH) Fe (PDP*)](OTf)
2
4
9
3
0
1
.9
.2
-
-
+
CH CO H/H O + HOAc
3 3 2 2
II
[
(MeTPEN)Fe (Cl)]Cl
58
d
+ H O + ArSH
2
2
a
Fe(CH CN) (OTf) + HOAc + H O + HOTf
1.2
1.2–1.3
2.2
-
3
2
2
2 2
e
94
(
FePctBu ) N + H O
4 2 2 2
anthracene ethylbenzene
.86 50
IV
2+
89, 95
[(BnTPEN)Fe (O)]
0
a
This work. Reactions conditions for this work: 1 mM 1, 10 equiv HOAc, 10 equiv 90% H O ,
2
2
1
equiv HOTf, 500 equiv unlabeled and 500 equiv perdeuterated benzene or cyclohexane in
CH CN at room temperature in air.
3
b
Py NHC = bis(o-imidazol-2-ylidenepyridine)methane.
2
2
c
N,N′-bis(3,5-dimethyl-4-methoxypyridyl-2-methyl)-(S,S)-2,2′-bipyrrolidine.
d
e
ArSH serves as a radical scavenger and is added to improve catalytic conversion.58
PctBu = tetrakis(tert-butyl)phthalocyanine dianion.
4
1
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Kinetic Studies of the Reaction of 2 with Acids at -40 °C
To systematically assess the effect of acids with a range of pK values on the decay of the
a
III
Fe –OOH species, 2 was first generated at room temperature with 10 equiv HOAc and 10
equiv 90% H O for 1 minute to achieve a yield of ~50%, and then cooled to -40 °C for kinetic
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
III
studies. Under these conditions, the Fe –OOH species formed eventually achieves a
maximum absorbance at 525 nm corresponding to ~70% conversion (Fig 6). The benefits of
using the H O /HOAc combination to generate the intermediate are three-fold: 1) it
2
2
III
circumvents the need for a large excess of H O typically required to generate this Fe –OOH
2
2
III
species, as only 10 equiv H O and 10 equiv HOAc are needed to generate the Fe –OOH
2
2
species in more than double the yield obtained when 100 equiv H O is used alone (Fig 1);
2
2
III
2) the EPR spectrum of the H O /HOAc reaction mixture shows the S = ½ Fe –OOH species
2
2
to be the only EPR-active iron species in solution, whereas significant amounts of signals
III
originating from ferric byproducts such as the Fe –OH species are also observed when using
H O alone (Fig S1); 3) HOAc is the weakest acid in our study (pK = 23.5 in CH CN) and likely
2
2
a
3
III
does not directly interact with the Fe –OOH species, as the decay rate of 2 is independent of
HOAc concentration (Table S1), thereby allowing it to be used as a baseline for reactivity
III
studies with other acids. Unlike experiments at room temperature where the Fe –OOH
species forms rapidly, 2 does not regenerate appreciably on the time scale of its reactions
with strong acids at -40 °C, which only take seconds. Therefore, its reactions with the strong
acids can be treated as stoichiometric transformations between the acid and 2 at its steady-
state concentration.
2
0
ACS Paragon Plus Environment

Page 21 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. (a) Representative spectra obtained from the reaction of 2 with 1 equiv HOTf at
-40 °C taken over the course of 5 sec. (b) First-order decay rates of 2 plotted as a function of
acid and its concentration. Reaction conditions: acid added to a solution of 2 at -40 °C, which
was first generated with 1 mM 1, 10 equiv HOAc and 10 equiv 90% H O in CH CN at room
2
2
3
temperature and then cooled to -40 °C. Errors for the second order rate constants obtained
for the slopes of the linear plots were estimated to be less than 10%.
The solution of 2 formed as described above is quite stable at -40 °C. It persists for 1 hr
with hardly any change in absorbance and decays completely over the course of 16 h. A kdecay
-5
-1
value of 2.4 x 10 s at -40 °C can be extrapolated from an Eyring plot based on rate data
obtained at higher temperatures (Fig S10). The addition of strong acids to 2 results in the
immediate decay of 2, dramatically decreasing the lifetime of 2 by 4 orders of magnitude,
from hours to seconds. Simultaneously 3 is formed with an apparent isosbestic point at 625
nm (Fig 6). Cyclohexane and benzene both can intercept the active oxidant, as their
2
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
introduction results in a decrease in yield of 3 in the case of cyclohexane, or the formation of
III
the Fe –OPh chromophore in the case of benzene (Fig S11 and S12). Excess H O in solution
2
2
may also serve as a substrate, as the introduction of additional H O leads to a decrease in
2
2
the yield of 3 (Fig S13).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Intermediate 2 decays exponentially, and the first order rate constants derived therefrom
increase as a function of acid concentration to afford second rate constants from the slopes
of the linear plots (Fig 6). The decay rates of 2 also depend on the pK of the added acid. Thus,
a
III
the addition of HClO (70%) to a solution of 2 gives rise to an Fe –OOH decay rate of 1300
4
-1
-1
-1 -1
M s , the fastest reported thus far (Table 4, Fig 7), followed by HOTf (870 M s ) and
-1 -1
HOTs•H O (340 M s ). In the latter case, the addition of HOTs•H O to a solution of 2 at -40
2
2
°C results in the formation of a different chromophore at 650 nm, which likely derives from
hydroxylation of HOTs. These three k values differ by only a factor of 4 despite having pK ’s
2
a
in CH CN that range from 1.6 to 8.5. Surprisingly, HNO with a pK of 8.8 in CH CN reacts with
3
3
a
3
-1 -1
2
with a second order rate constant of 2.6 M s , which is two orders of magnitude slower
than that for HOTs•H O, even though the two acids differ by only ~0.3 of a pK unit in CH CN.
2
a
3
At the same time, the A/K ratio observed for cyclohexane hydroxylation for this entire series
of acids decreases from 9 to 1 upon replacing a strong acid with an acid having pK in CH CN
a
3
of more than 8.5 (Table 1), and the amount of 3 that is formed also drops significantly (Fig
S6, Table S2). These results suggest a change in mechanism in which the addition of an acid
as strong or stronger than HOTs•H O favors a proton-assisted O–O bond heterolysis of the
2
III
Fe –OOH intermediate 2, which is in competition with the O–O bond homolysis pathway
inherent to 2. A similar pH-dependent switch at pH ~ 9 in the outcome of O–O bond cleavage
has been reported by van Eldik and coworkers in the reaction of a water-soluble porphyrin
2
4
with H O .
2
2
2
2
ACS Paragon Plus Environment

Page 23 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table 4. Decay of 2 in the presence of various acids at -40 °C.a
-1 -1
Acid
pKa
1.57
—
k (M s )
2
HClO (70% in H O)
1300
1500
870
340
2.6
4
2
DClO (68% in D O)
4
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
HOTf
2.6
HOTs•H O
8.45
8.8
2
HNO3
a
Second order rate constants were extracted from Fig 6b. Rates obtained for 3 formation at
25 nm and 2 decay at 580 nm were almost identical (Fig S14) and thus were treated as
8
equivalent information. Rates for reactions of 2 with HNO were obtained by monitoring the
3
decay of 2 at 580 nm. Rates for reactions of 2 with HOTs•H O were obtained by monitoring
2
6
50 nm chromophore formation at 825 nm.
log kobs
0.001
0.01
0.1
1
-
1
1
equiv HClO (70%) 1.3
s
4
-
1
1
equiv HOTf 0.9s
-
1
1 equiv HOTsH O 0.3
s
2
-
1
0.0026
s
1 equiv HNO3
-
5
-1
2
.4 x 10 s self-decay*
00 equiv HOAc 0.17
BnTPEN
-1
2
s
-
1
a
3
00 mM H O 0.0066
s
2
TPA
-
1
b
1
.1 equiv HOTf 0.02
s
PyNMe3
c
3
+
-1
1 equiv Sc or HClO4 0.3s -BPMCN
III
Figure 7. First-order decay rates for (L)Fe –OOH species at -40 °C.
*The self-decay rate of 2 at -40 °C was estimated from an extrapolation from the Eyring
plot for the decay of 2 in the presence of HOAc measured at higher temperatures (Fig S10).
a
96,50 b
55 c
56
refs
; ref ; ref .
2
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
We have also investigated whether a higher concentration of a weak acid such as 100 equiv
V
HNO could facilitate the decay of 2 and give rise to the Fe =O oxidant. However, the decay
3
rate of 2 under these conditions remains considerably slower than with addition of 1 equiv
HOTs•H O (Fig 6). The reactivity of 2 towards cyclohexane also does not change significantly
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
upon addition of 100 equiv HNO , suggesting no change in the reaction mechanism despite
3
the higher concentration of this weak acid (Fig S15). On the other hand, adding an excess of
strong acid is found to be detrimental to the reaction, with the total yield of cyclohexane-
derived products dropping to only 2 TON and the A/K ratio approaching unity (Fig S15).
Some ligand dissociation from the iron complex could occur under such conditions, allowing
for a Fenton-type reaction to dominate instead.
The use of DClO in place of HClO results in a 10% faster rate of decay, or an inverse H/D
4
4
isotope effect of 0.9 (Table 4). This effect differs from the solvent KIE (KSIE) values obtained
for reactions where the proton delivered to the peroxo species likely originates from a weak
acid, such as a water molecule in the formation of Cpd I from Cpd 0 in horseradish peroxidase
(1.7),97 the conversion of the diferric-peroxo intermediate of soluble methane
monooxygenase (sMMO) to the diiron(IV) intermediate Q (1.4 and 1.8),98, 99 or the water-
III
50
assisted O–O heterolysis of the (TPA)Fe –OOH species (2.5). In these latter examples, the
KSIE observed likely originates from the cleavage of the water O–H/(D) bond prior to or
concomitant with O–O bond cleavage of the hydroperoxo moiety. In our reactions with
perchloric acid, the slightly faster rate obtained with DClO may be the result of an
4
III
III
equilibrium isotope effect, favoring the formation of an Fe –OOH---D adduct over an Fe –
OOH---H adduct, as deuterium can form a stronger bond with oxygen than protium.100-102
Taken together, the data strongly favors a mechanism where a proton is directly involved in
III
O–O bond heterolysis of the Fe –OOH species.
Eyring analysis has been carried out to gain further insight into the rate determining
O–O bond cleavage step (Table 5, Fig S10). In the presence of HOAc, the decay of 2 has Eyring
⧧
-1
⧧
-1
-1
parameters of ΔH = 52(2) kJ.mol and ΔS = -108(10) J K .mol . These values are very close
3
5
to those found for the decay of 2 in CH OH and almost identical to that obtained for
3
III
(
N4Py)Fe –OOH in CH OH, both of which have been proposed to undergo rate-determining
3
homolytic O–O bond scission.103 The lack of a dramatic change in Eyring parameters upon
2
4
ACS Paragon Plus Environment

Page 25 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
addition of HOAc further supports our hypothesis that HOAc does not play a significant role
in the O–O bond cleavage step. In stark contrast, the addition of 1 equiv HOTf to the solution
⧧
-1
-1
of 2 drastically lowers its ΔH by 22 kJ mol to 30(2) kJ mol , reflecting the significantly
more rapid decay of 2 in the presence of HOTf. This activation enthalpy is the smallest ΔH⧧
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
III
-1
value we have determined for the decay of an Fe –OOH species and at least 14 kJ mol
III
smaller than values associated with Fe –OOH complexes supported by tetradentate ligands
such as TPA, PyNMe and TMC that undergo heterolytic cleavage of the O–O bond, even upon
3
treatment with strong acid (Table 5). Thus, the proton-induced heterolytic O–O bond
III
cleavage of the (BnTPEN)Fe –OOH intermediate 2 is distinct from the other complexes and
V
more facile, unleashing quite a powerful Fe =O oxidant capable of rapidly cleaving strong C–
H bonds even at -40 °C.
III
Table 5. Eyring Activation Parameters for the Decay of (L)Fe -OOR Complexes
ΔH⧧,
kJ/mol
ΔS⧧,
J/(K·mol)
Proposed
O–O Bond
Ref
Cleavage Mode
a
a
2
2
+ HOAc + HOTf in CH CN
30(2)
52(2)
53(2)
81
-118(20)
-108(10)
-72(8)
Heterolysis
Homolysis
Homolysis
Homolysis
Homolysis
Homolysis
Homolysis
Homolysis
Heterolysis
Homolysis
Heterolysis
3
+ HOAc in CH CN
3
3
5
2 in CH OH
3
III
58
(
(
(
(
MeTPEN)Fe –OOH in CH CN
-1
3
III
103
104
105
106
50
N4Py)Fe –OOH in acetone/CF CH OH
53(1)
53(1)
52(1)
54(3)
45(2)
44(3)
46(4)
-121(2)
-68(4)
3
2
III
bbpc)Fe –OOH in CH CN
3
III
t
TPA)Fe -OO Bu in CH CN
–74(3)
–35(13)
-95(10)
-138(12)
-79(13)
3
III
2
(TPA)Fe (κ -O C(CH ) OH) in acetone
2
3 2
III
(
TPA)Fe –OOH in CH CN
3
III
55
(PyNMe )Fe –OOH in CH CN
3
3
III
55
(
PyNMe )Fe –OOH in CH CN
3
3
with 1.1 equiv HOTf
III
103
107
(
14-TMC)Fe –OOH
56(2)
44(2)
-75(2)
Homolysis
Heterolysis
in acetone/CF CH OH
3
2
III
(
14-TMC)Fe –OOH + HClO in CH CN
-90(10)
4
3
a
This work. bbpc = N,N′-dibenzyl-N,N′-bis(2-pyridylmethyl)-1,2-cyclohexanediamine. 14-
TMC = 1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane. See Fig S10 for Eyring plots
and reaction conditions.
2
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Mechanistic Insights
Scheme 3. Proposed mechanism for the reactivity of 2.
H+
FeIII OAr
H
pKa < 8.5
O
H
intramolecular
aromatic hydroxylation
O
(CH3CN)
O
N
N
N
N
FeV
N
N
N
FeIII
H2O
N
OH
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
H
N
O
O
proton-assisted
O–O bond heterolysis
N
N
N
N
intermolecular
aromatic hydroxylation
FeIII
N
2
H
OH
O
O
OH
O
O
+
N
N
N
N
N
N
FeIII
FeIV
N
N
alkane hydroxylation
self-decay
N
N
substrates
spontaneous
O–O bond homolysis
hydroxyl radical-
mediated reactions
V
We propose that an Fe =O species is responsible for the oxidation reactions observed
when 2 is treated with strong acids (Scheme 3). Strong acids promote heterolytic scission of
III
V
the O–O bond of the Fe –OOH species, producing an Fe =O species that hydroxylates
cyclohexane to cyclohexanol and benzene to phenol within seconds at -40 °C, which is
III
unprecedented for 2 and related Fe(N5) complexes. The decay rate of the Fe –OOH species
is linearly correlated with the concentration of the strong acid used and is faster for acids
III
with lower pK values (Fig 6b and Table 4). An Eyring plot obtained for Fe –OOH decay in
a
-1
⧧
the presence of HOTf shows a dramatic decrease of 22 kJ mol in ΔH relative to the 52-kJ
-1
35, 103
mol value typical for O–O-bond homolysis (Table 5),
suggesting that O–O bond
heterolysis is greatly facilitated by strong acids. Taken together, the data strongly support
III
the direct involvement of a proton in the O–O bond cleavage of the Fe –OOH species by
facilitating the departure of the distal OH group by pulling it away as water, thereby
V
promoting the formation of an Fe =O species without compromising its electrophilicity.
Prior to the current work, the best-supported examples of proton-assisted O–O bond
III
heterolysis of Fe –OOH species in nonheme model systems occur via what is known as the
water-assisted pathway, which is associated with bio-inspired nonheme iron catalysts for
hydrocarbon oxidations that are supported by tetradentate N4 ligands such as TPA, BPMEN
and Me2PyTACN.50, 51, 85 In this class of catalysts, an iron-bound water molecule is
III
conveniently positioned cis to the Fe –OOH moiety, allowing for hydrogen bonding
interactions between a proton of the cis-bound water to the distal oxygen of the
hydroperoxide, thereby forming a five-membered ring transition state that facilitates O–O
2
6
ACS Paragon Plus Environment

Page 27 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
V
45, 46, 50, 51
bond heterolysis and the formation of the putative Fe (O)(OH) oxidant (Scheme 1).
Subsequent oxo/hydroxo tautomerism allows the oxygen atom from H O to be incorporated
2
V
into the Fe =O unit. This mechanistic scenario was conceived 20 years ago for the Fe(BPMEN)
1
8
catalyst to account for the observed partial O-labeling of hydroxylated products when the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
18 81
reactions were carried out in the presence of H O and has since been confirmed for other
2
45
Me2
47
Fe(N4) catalysts like Fe(TPA) and Fe( PyTACN) among others. Subsequently, evidence
V
49, 52
for the Fe (O)(OH) oxidant was obtained from mass spectrometry experiments,
as well
as more recent helium tagging infrared photodissociation (IRPD) spectroscopic experiments
of Roithová, Costas and coworkers.53
III
In contrast, the pentadentate BnTPEN ligand framework forms a terminal S = ½ Fe –OOH
species with a coordinately saturated coordination sphere, much like the Cpd 0 species found
in heme systems10, 11 and requires the added strong acid to react directly with the Fe –OOH
moiety in an intermolecular fashion. The effects of water on 1-catalyzed reactions in the
presence of strong acids are drastically different from those found for Fe(TPA). At the same
water concentration, the incorporation of water oxygen into cyclohexanol is observed to a
III
2
+ 45, 46, 50
much lower extent for 1 compared to [Fe(TPA)(NCCH ) ] .
Most significant of all,
3 2
-1
strong acids greatly accelerate the decay of 2 even at 1 mM (kobs ~ 1 s at -40 °C), resulting
in decay rates at least two orders of magnitude faster than the fastest rate for water-assisted
III
-1
50
Fe –OOH decay (k ~ 0.007 s at 300 mM H O at -40 °C). However, the presence of 50
obs
2
equiv of water can slow down the decay of 2 by about 60-fold (Fig S9). These observations
clearly establish the proton-assisted O–O bond cleavage pathway of 1 as a separate pathway
from the water-assisted pathway.
In this study, we have taken advantage of the pendant benzyl group of the BnTPEN
framework as a mechanistic probe for the reactivity of 2, as the hydroxylation of this
aromatic ring gives rise to the intramolecular phenolate product 3. Indeed, intramolecular
aromatic hydroxylation has been observed for a number of nonheme iron complexes, for
IV
V
70-75
which either Fe =O or Fe =O species are proposed to be the arene hydroxylating agent.
IV
2+
95, 108
In the case of 2, we can exclude [(BnTPEN)Fe (O)] as a possibility.
Although HAT
reactivity has been studied in detail for this complex,95, 108 to the best of our knowledge there
is no report that it is capable of intramolecular aromatic hydroxylation. Nam and Fukuzumi
2
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 44
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
have reported that metal-oxo species in the presence of strong protic acids can exhibit
8
3
enhanced reactivity via a proton-coupled electron transfer (PCET) pathway, and this effect
IV
2+
109,
is seen in the oxidation of α-methylstyrene by [(N4Py)Fe (O)] in the presence of HOTf.
1
10
Under our reaction conditions, however, addition of 1 equiv HClO does not significantly
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
IV
2+
accelerate the decay of [(BnTPEN)Fe (O)] nor lead to the formation of 3 (Fig S16); even
high concentrations of perchloric acid do not accelerate its decay nor is pyridyl arm
IV
2+
dissociation found (Fig S17). These observations show that [(BnTPEN)Fe (O)] is not
activated by the presence of strong acids and can be ruled out as the active species, thus
V
leading us to invoke an Fe =O oxidant that is formed by the heterolytic cleavage of the O–O
bond in 2.
Thus far there is only one example, reported by Hitomi and coworkers, that presents
V
substantial evidence for an Fe =O species formed in a catalytic nonheme iron system
supported by a pentadentate ligand. They have demonstrated hydroxylation of cyclohexane
III
2+
using H O with [(DPAQ)Fe (OH )] (DPAQ = 2-[bis(pyridine-2-ylmethyl)]amino-N-
2
2
2
quinolin-8-yl-acetamidate anion) as the catalyst without requiring any additives.111 Ions
III
+
V
+
consistent with [(DPAQ)Fe (OOH)] (m/z 471.1) and {[(DPAQ)Fe (O)]ClO } (m/z 553.1)
4
have been observed by cryospray ionization mass spectroscopy (CSI-MS). Furthermore, A/K
ratios as high as 11 are found for cyclohexane oxidation at room temperature, consistent
with metal-based oxidation. In this ligand framework, the introduction of the monoanionic
amidate donor lowers the redox potential of the iron center, while its location trans to the
hydroperoxo ligand provides a “push” towards heterolytic cleavage of the O–O bond and
V
stabilizes the incipient Fe =O oxidant, analogous to the role ascribed to the thiolate ligand in
cytochrome P450.112-114 It is also proposed that the proton released upon H O binding to
2
2
III
2+
the [(DPAQ)Fe (OH )] catalyst provides the “pull” to cleave the O–O bond heterolytically.
2
Unfortunately, these catalytic experiments were carried out at room temperature with the
H O being delivered by syringe pump over a 30-min period, so there is no kinetic data
2
2
III
2+
available with which to compare O–O bond cleavage rates of [(BnTPEN)Fe (OOH)] (2) and
III
+
[
(DPAQ)Fe (OOH)] . However, given that substrate hydroxylation by 2 is shown to occur at
-40 °C within seconds after the introduction of the strong acid, it would seem likely that the
2
8
ACS Paragon Plus Environment

Page 29 of 44
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
V
O–O bond cleavage step to generate the Fe =O oxidant is significantly more facile in the case
of 2.
In an effort to demonstrate the feasibility of generating a metastable oxoiron(V) complex,
Collins has incorporated four highly electron-donating amido donors into a macrocyclic
ligand called TAML, which stands for tetraamido macrocyclic ligand, leading to the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
V
115, 116
characterization of the first bona fide Fe =O complex.
A family of relatively stable
V
(TAML)Fe (O) complexes has been obtained by introducing different substituents on the
macrocycle. Perhaps not surprisingly, these complexes exhibit cyclohexane oxidation rates
-3
-1 -1
117, 118
with k < 10 M s at -40 °C,
which are two orders of magnitude slower than those
2
found for the two most reactive oxoiron(IV) complexes described to date.119, 120 In contrast,
V
cyclohexane oxidation by the putative Fe =O oxidant derived from the addition of 1 equiv
HClO to 2 in the presence of cyclohexane occurs within seconds at this temperature. In the
4
latter case, substrate oxidation occurs at a rate independent of substrate concentration, as it
takes place subsequent to the rate-determining decay of 2 upon its reaction with strong acids.
Thus, substrate oxidation must occur at a rate faster than the reaction of 2 with HClO . The
4
-1
kobs value for 1 mM 2 and 1 mM HClO is 1.3 s at -40 °C, which is at least three orders of
4
V
–
magnitude faster than the oxidation rate of 1 M cyclohexane by [(TAML)Fe (O)] at -40 °C
IV
and 3 – 5-fold faster than the two most reactive nonheme Fe =O complexes reported thus
far.119, 120
When the reactivity of 2 in the presence of strong acid is compared with reported rates for
cyclohexane oxidation (assuming a substrate concentration of 1 M as used in our catalysis
studies), the cyclohexane oxidation rate for 2 is calculated to be at least an order of
magnitude faster than that found for a synthetic Cpd I model, the one-electron oxidized
IV
derivative of [(TDCPP)Fe (O)] (H TDCPP = meso-tetrakis(2,6-dichlorophenyl)porphin) (k
2
obs
-1
-1
120
=
1.1 x 10 s for 1 M cyclohexane at -40 °C). However it is much less reactive than
AaeAPO-I, the Cpd I intermediate of a heme-thiolate peroxygenase, which exhibits a kobs of
4
-1
121
10 s for the oxidation of 1 M cyclohexane carboxylic acid at 4 °C. The rate constant for
the decay of 2 in the presence of 1 mM HOTf increases by a factor of 15 when its rate at
-
40 °C is extrapolated to 4 °C based on an Eyring analysis of its temperature dependence
(
Figure S13), making it about three orders of magnitude less reactive than AaeAPO-I. These
2
9
ACS Paragon Plus Environment