Oxidizing intermediates from the sterically hindered iron salen complexes related to the oxygen activation by nonheme iron enzymes

Article abstract of DOI:10.1021/ic051377e

Oxidizing intermediates are generated from nonheme iron(III) complexes to investigate the electronic structure and the reactivity, in comparison with the oxoiron(IV) porphyrin π-cation radical (compound I) as a heme enzyme model. Sterically hindered iron salen complexes, bearing a fifth ligand Cl (1), OH ₂ (2), OEt (3), and OH (4), are oxidized both electrochemically and chemically. Stepwise one-electron oxidation of 1 and 2 generates iron(III)-mono- and diphenoxyl radicals, as revealed by detailed spectroscopic investigations, including UV-vis, EPR, Moessbauer, resonance Raman, and ESIMS spectroscopies. In contrast to the oxoiron(IV) formation from the hydroxoiron(lll) porphyrin upon one-electron oxidation, the hydroxo complex 4 does not generate oxoiron(IV) species. Reaction of 2 with mCPBA also results in the formation of the iron(III)-phenoxyl radical. One-electron oxidation of 3 leads to oxidative degradation of the fifth EtO ligand to liberate acetaldehyde even at 203 K. The iron(III)-phenoxyl radical shows high reactivity for alcoxide on iron(III) but exhibits virtually no reactivity for alcohols including even benzyl alcohol without a base to remove an alcohol proton. This study explains unique properties of mononuclear nonheme enzymes with Tyr residues and also the poor epoxidation activity of Fe salen compared to Mn and Cr salen compounds.

Full text of DOI:10.1021/ic051377e

Inorg. Chem. 2005, 44, 8156−8166
Oxidizing Intermediates from the Sterically Hindered Iron Salen
Complexes Related to the Oxygen Activation by Nonheme Iron Enzymes
Takuya Kurahashi,† Yoshio Kobayashi,‡ Shigenori Nagatomo,† Takehiko Tosha,† Teizo Kitagawa,† and
Hiroshi Fujii*,†
Institute for Molecular Science & Okazaki Institute for IntegratiVe Bioscience, National Institutes
of Natural Sciences, Myodaiji, Okazaki, Aichi 444-8787, Japan, and The Institute of Physical and
Chemical Research (RIKEN), Hirosawa, Wako, Saitama 351-0198, Japan
Received August 12, 2005
Oxidizing intermediates are generated from nonheme iron(III) complexes to investigate the electronic structure and
the reactivity, in comparison with the oxoiron(IV) porphyrin -cation radical (compound I) as a heme enzyme
model. Sterically hindered iron salen complexes, bearing a fifth ligand Cl (1), OH2 (2), OEt (3), and OH (4), are
oxidized both electrochemically and chemically. Stepwise one-electron oxidation of 1 and 2 generates iron(III)
mono- and diphenoxyl radicals, as revealed by detailed spectroscopic investigations, including UV vis, EPR,
π
−
−
Mo¨ssbauer, resonance Raman, and ESIMS spectroscopies. In contrast to the oxoiron(IV) formation from the
hydroxoiron(III) porphyrin upon one-electron oxidation, the hydroxo complex 4 does not generate oxoiron(IV) species.
Reaction of 2 with mCPBA also results in the formation of the iron(III)
−phenoxyl radical. One-electron oxidation of
3 leads to oxidative degradation of the fifth EtO ligand to liberate acetaldehyde even at 203 K. The iron(III)
−
phenoxyl radical shows high reactivity for alcoxide on iron(III) but exhibits virtually no reactivity for alcohols including
even benzyl alcohol without a base to remove an alcohol proton. This study explains unique properties of mononuclear
nonheme enzymes with Tyr residues and also the poor epoxidation activity of Fe salen compared to Mn and Cr
salen compounds.
Introduction
soluble diiron enzyme methane monooxygenase3 retains two
oxidizing equivalents divided on two iron centers, generating
the FeIV2O2 diamond core.4
A number of important oxidative transformations are
carried out by iron enzymes utilizing dioxygen as an oxygen
source.1 A wide variety of coordination environments around
the iron center in those enzymes generate distinct oxidizing
intermediates, which are supposed to result in their intrinsic
reactivities. The most important feature of oxidizing inter-
mediates in determining reactivities is how oxidizing equiva-
lents from dioxygen are distributed around an iron center.
In the case of heme enzymes, for example, the electronic
structure of the oxidizing intermediate is well documented
to be the oxoiron(IV) porphyrin π-cation radical called com-
pound I, with two oxidizing equivalents divided into the
central iron and the surrounding porphyrin ligand.2 On the
other hand, the Fe2O2 diiron active site observed for the
In contrast, mononuclear nonheme iron enzymes have long
been a challenge on how oxidizing equivalents are held
without an electron pool such as a porphyrin π-conjugation
and a nearby iron atom. From this viewpoint, the iron-
containing glycopeptide bleomycin has been extensively
studied for a long time, and the only oxidizing intermediate
detected to date is the FeIII-OOH form.5 But recent works
on the taurin:R-ketoglutarate-dependent dioxygenase (TauD),
which has a mononuclear nonheme iron center facially
coordinated by two histidines (His) and one asparaginic acid
(3) Rosenzweig, A. C.; Frederick, C. A.; Lippard, S. J.; Nordlund, P.
Nature 1993, 366, 537-543.
(4) (a) Lee, S.-K.; Nesheim, J. C.; Lipscomb, J. D. J. Biol. Chem. 1993,
268, 21569-21577. (b) Lee, S.-K.; Fox, B. G.; Froland, W. A.;
Lipscomb, J. D.; Mu¨nck, E. J. Am. Chem. Soc. 1993, 115, 6450-
6451. (c) Liu, K. E.; Valentine, A. M.; Wang, D.; Huynh, B. H.;
Edmondson, D. E.; Salifoglou, A.; Lippard, S. J. J. Am. Chem. Soc.
1995, 117, 10174-10185. (d) Shu, L.; Nesheim, J. C.; Kauffmann,
K.; Mu¨nck, E.; Lipscomb, J. D.; Que, L., Jr. Science 1997, 275, 515-
518.
* To whom correspondence should be addressed. E-mail: hiro@ims.ac.jp.
† National Institutes of Natural Sciences.
‡ RIKEN.
(1) See, for example: Lippard, S. J.; Berg, J. M. Principles of Bioinorganic
Chemistry; University Science Books: New York, 1994.
(2) Sono, M.; Roach, M. P.; Coulter, E. D.; Dawson, J. H. Chem. ReV.
1996, 96, 2841-2887.
8156 Inorganic Chemistry, Vol. 44, No. 22, 2005
10.1021/ic051377e CCC: $30.25
© 2005 American Chemical Society
Published on Web 10/01/2005

Sterically Hindered Iron Salen Complexes
(Asp), provided compelling evidence for the oxoiron(IV)
formation during oxygen activation.6 In addition, several
synthetic oxoiron(IV) species were successfully prepared
from biomimetic nonheme model complexes,7 which share
a redox-innocent nitrogen coordination motif together with
TauD. These epoch-making examples suggest that an non-
heme oxoiron(IV) species could possibly be stabilized by
the His coordination.
Another group of nonheme iron enzymes bear tyrosine
(Tyr) residues, which also play an important role in catalysis
in a more diverse manner due to a redox-noninnocent nature.8
For example, in the case of the ribonucleotide reductase R2
protein,9 the uncoordinated Tyr in close proximity to the
diiron(III) core is converted to the tyrosyl radical upon dioxy-
gen activation.10 A coordinated Tyr residue is also found
for mononuclear nonheme iron(III) enzymes such as proto-
catechuate 3,4-dioxygenase (3,4-PCD) bearing the His2Tyr2
coordination environment.11 Strikingly different from most
of the iron enzymes activating dioxygen on the iron center,
3,4-PCD is proposed to activate an iron-bound substrate for
reaction with dioxygen.12 A rather unique reaction sequence
of 3,4-PCD is possibly induced by the electronic effect
caused by the Tyr coordination to the iron(III) center.
Model studies concerning both coordinated and uncoor-
dinated Tyr in nonheme iron enzymes have been quite limited
so far.13-15 To examine the Tyr coordination to the iron(III)
center, Wieghardt et al. prepared a series of hexadentate iron
complexes containing a 1,4,7-triazacyclononane backbone
and three N-bound phenolate moieties.14 Stepwise one-
electron oxidation of their model complexes was reported
to result in three successive ligand-centered, phenolate-to-
phenoxyl radical conversions. Their phenoxyl radical com-
plexes are excellent models, which are stable at room
temperature. However, their model complexes do not neces-
sarily reproduce the biological system in all aspects, because
their model system does not have a vacant coordination site,
which is critical for accommodation of a biologically
important ligand such as H2O and OH. Indeed, 3,4-PCD has
a vacant coordination site, which is occupied with a solvent-
derived water molecule.11 The vacant coordination site on
the iron(III) center would be of considerable importance for
substrate-binding, considering the examples of the copper-
(II)-phenoxyl radical as a model for galactose oxidase
(GO).16 At present, reactivity of the iron(III)-phenoxyl
radical intermediates is not clear.14
In this context, we constructed a biologically more relevant
system based on a iron salen complex. A salen ligand well
reproduces the coordination environment by two His and two
Tyr. Most importantly, an iron salen complex has a vacant
fifth coordination site. Following the strategy utilized in the
heme enzyme model,17 we introduced bulky mesityl groups
to the salen framework to stabilize the monomeric iron center
and successfully prepared a H2O-coordinated iron salen
complex, which not only duplicates the structural features
around the iron(III) active site but also mimics the spectral
characteristics of 3,4-PCD.18 Herein, we carried out elec-
trochemical and chemical oxidation of the iron salen
complexes bearing Cl (1), OH2 (2), OEt (3), and OH (4) as
a fifth ligand (Figure 1). We also attempted mCPBA
(5) (a) Burger, R. M.; Kent, T. A.; Horwitz, S. B.; Mu¨nck, E.; Peisach, J.
J. Biol. Chem. 1983, 258, 1559-1564. (b) Burger, R. M.; Blanchard,
J. S.; Horwitz, S. B.; Peisach, J. J. Biol. Chem. 1985, 260, 15406-
15409. (c) Sam, J. W.; Tang, X.-J.; Peisach, J. J. Am. Chem. Soc.
1994, 116, 5250-5256. (d) Absalon, M. J.; Wu, W.; Kozarich, J. W.;
Stubbe, J. Biochemistry 1995, 34, 2076-2086. (e) Neese, F.; Zaleski,
J. M.; Zaleski, K. L.; Solomon, E. I. J. Am. Chem. Soc. 2000, 122,
11703-11724.
(6) (a) Price, J. C.; Barr, E. W.; Tirupati, B.; Bollinger, J. M., Jr.; Krebs,
C. Biochemistry 2003, 42, 7497-7508. (b) Proshlyakov, D. A.;
Henshaw, T. F.; Monterosso, G. R.; Ryle, M. J.; Hausinger, R. P. J.
Am. Chem. Soc. 2004, 126, 1022-1023. (c) Riggs-Gelasco, P. J.; Price,
J. C.; Guyer, R. B.; Brehm, J. H.; Barr, E. W.; Bollinger, J. M., Jr.;
Krebs, C. J. Am. Chem. Soc. 2004, 126, 8108-8109.
(7) (a) Grapperhaus, C. A.; Mienert, B.; Bill, E.; Weyhermu¨ller, T.;
Wieghardt, K. Inorg. Chem. 2000, 39, 5306-5317. (b) Rohde, J.-U.;
In, J.-H.; Lim, M. H.; Brennessel, W. W.; Bukowski, M. R.; Stubna,
A.; Mu¨nck, E.; Nam, W.; Que, L., Jr. Science 2003, 299, 1037-1039.
(c) Lim, M. H.; Rohde, J.-U.; Stubna, A.; Bukowski, M. R.; Costas,
M.; Ho, R. Y. N.; Mu¨nck, E.; Nam, W.; Que, L., Jr. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 3665-3670. (d) Balland, V.; Charlot, M.-F.;
Banse, F.; Girerd, J.-J.; Mattioli, T. A.; Bill, E.; Bartoli, J.-F.; Battioni,
P.; Mansuy, D. Eur. J. Inorg. Chem. 2004, 301-308. (e) Kaizer, J.;
Klinker, E. J.; Oh, N. Y.; Rohde, J.-U.; Song, W. J.; Stubna, A.; Kim,
J.; Mu¨nck, E.; Nam, W.; Que, L., Jr. J. Am. Chem. Soc. 2004, 126,
472-473. (f) Decker, A.; Rohde, J.-U.; Que, L., Jr.; Solomon, E. I. J.
Am. Chem. Soc. 2004, 126, 5378-5379. (g) Rohde, J.-U.; Torelli, S.;
Shan, X.; Lim, M. H.; Klinker, E. J.; Kaizer, J.; Chen, K.; Nam, W.;
Que, L., Jr. J. Am. Chem. Soc. 2004, 126, 16750-16761. (h) Ghosh,
A.; Tiago de Oliveira, F. Yano, T.; Nishioka, T.; Beach, E. S.;
Kinoshita, I.; Mu¨nck, E.; Ryabov, A. D.; Horwitz, C. P.; Collins T. J.
J. Am. Chem. Soc. 2005, 127, 2505-2513. (g) Kim, S. O.; Sastri, C.
V.; Seo, M. S.; Kim, J.; Nam, W. J. Am. Chem. Soc. 2005, 127, 4178-
4179.
(13) Chaudhuri, P.; Wieghardt, K. Prog. Inorg. Chem. 2001, 50, 151-
216.
(14) (a) Hockertz, J.; Steenken, S.; Wieghardt, K.; Hildebrandt, P. J. Am.
Chem. Soc. 1993, 115, 11222-11230. (b) Adam, B.; Bill, E.; Bothe,
E.; Goerdt, B.; Haselhorst, G.; Hildenbrand, K.; Sokolowski, A.;
Steenken, S.; Weyhermu¨ller, T.; Wieghardt, K. Chem.sEur. J. 1997,
3, 308-319. (c) Snodin, M. D.; Ould-Moussa, L.; Wallmann, U.;
Lecomte, S.; Bachler, V.; Bill, E.; Hummel, H.; Weyhermu¨ller, T.;
Hildebrandt, P.; Wieghardt, K. Chem.sEur. J. 1999, 5, 2554-2565.
(15) (a) Goldberg, D. P.; Watton, S. P.; Masschelein, A.; Wimmer, L.;
Lippard, S. J. J. Am. Chem. Soc. 1993, 115, 5346-5347. (b) Goldberg,
D. P.; Koulougliotis, D.; Brudvig, G. W.; Lippard, S. J. J. Am. Chem.
Soc. 1995, 117, 3134-3144.
(16) (a) Wang, Y.; Stack, T. D. P. J. Am. Chem. Soc. 1996, 118, 13097-
13098. (b) Halfen, J. A.; Jazdzewski, B. A.; Mahapatra, S.; Berreau,
L. M.; Wilkinson, E. C.; Que, L., Jr.; Tolman, W. B. J. Am. Chem.
Soc. 1997, 119, 8217-8227. (c) Wang, Y.; DuBois, J. L.; Hedman,
B.; Hodgson, K. O.; Stack, T. D. P. Science 1998, 279, 537-540. (d)
Itoh, S.; Taki, M.; Takayama, S.; Nagatomo, S.; Kitagawa, T.;
Sakurada, N.; Arakawa, R.; Fukuzumi, S. Angew. Chem., Int. Ed. 1999,
38, 2774-2776. (e) Chaudhuri, P.; Hess, M.; Mu¨ller, J.; Hildenbrand,
K.; Bill, E.; Weyhermu¨ller, T.; Wieghardt, K. J. Am. Chem. Soc. 1999,
121, 9599-9610. (f) Thomas, F.; Gellon, G.; Gautier-Luneau, I.; Saint-
Aman, E.; Pierre, J.-L. Angew. Chem., Int. Ed. 2002, 41, 3047-3050.
(17) (a) Groves, J. T.; Haushalter, R. C.; Nakamura M.; Nemo, T. E.; Evans,
B. J. J. Am. Chem. Soc. 1981, 103, 2884-2886. (b) Balch, A. L.;
Latos-Grazynski, L.; Renner, M. W. J. Am. Chem. Soc. 1985, 107,
2983-2985. (c) Sugimoto, H.; Tung, H.-C.; Sawyer, D. T. J. Am.
Chem. Soc. 1988, 110, 2465-2470. (d) Fujii, H. J. Am. Chem. Soc.
1993, 115, 4641-4648.
(8) Solomon, E. I.; Brunold, T. C.; Davis, M. I.; Kemsley, J. N.; Lee,
S.-K.; Lehnert, N.; Neese, F.; Skulan, A. J.; Yang, Y.-S.; Zhou, J.
Chem. ReV. 2000, 100, 235-349.
(9) Nordlund, P.; Sjo¨berg, B.-M.; Eklund, H. Nature 1990, 345, 593-
598.
(10) Stubbe, J.; Donk, W. A. v. d. Chem. ReV. 1998, 98, 705-762.
(11) (a) Ohlendorf, D. H.; Lipscomb, J. D.; Weber, P. C. Nature 1988,
336, 403-405. (b) Ohlendorf, D. H.; Orville, A. M.; Lipscomb, J. D.
J. Mol. Biol. 1994, 244, 586-608.
(12) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L., Jr. Chem. ReV. 2004,
104, 939-986.
(18) Fujii, H.; Funahashi, Y. Angew. Chem., Int. Ed. 2002, 41, 3638-
3641.
Inorganic Chemistry, Vol. 44, No. 22, 2005 8157

Kurahashi et al.
Table 1. Oxidation Potentials of the Iron Salen Complexesa
compd
E1
E2
1/2
1/2
1
2
3
4
0.85
0.80
0.82
0.79
0.96
1.00
1.00
0.95
a See the caption in Figure 2 for the measurement conditions.
Figure 1. Series of iron salen complexes.
Figure 3. UV-vis spectral changes upon electrochemical oxidation of 1
in CH2Cl2 at 203 K. Conditions: 0.5 × 10-3 M 1; 0.10 M tetrabutylam-
monium perchlorate supporting electrolyte; controlled-potential oxidation
at 1.00 V for [1]+ and 1.30 V for [1]2+ vs Fc′/Fc. 1, [1]+, and [1]2+ are
depicted in a solid line.
Figure 2. Cyclic voltammogram of 1 in CH2Cl2 at 233 K. Conditions: 1
× 10-3 M 1; 0.10 M tetrabutylammonium perchlorate supporting electrolyte;
a saturated calomel electrode as a reference electrode; a glassy carbon
working electrode; a platinum-wire counter electrode; scan rate 50 mV s-1
.
The first (E11/2) and second oxidation potential (E21/2) is 0.85
and 0.96 V versus the ferrocenium/ferrocene couple, respec-
tively (Table 1).
The potentials are referenced versus the ferrocenium/ferrocene (Fc′/Fc)
couple.
oxidation of 2 to generate oxidizing intermediates, which is
often utilized to generate high-valent metal-oxo species in
porphyrin chemistry.17 Surprisingly, there are only two
reports concerning oxidizing intermediates from iron salen
complexes.19 Rajagopal and Ramaraj proposed a compound-
I-like oxoiron(IV) salen radical from iron salen complexes
and iodosylbenzene.19a But later Bryliakov provided spec-
troscopic observations against such active species and
proposed the iodosylbenzene-iron(III) salen adduct.19b Herein,
we did obtain intermediates of a higher oxidation state from
iron salen complexes. And we characterized the electronic
structures and investigated the reactivity of the oxidizing
intermediates, to get insight into the unique reaction sequence
for the nonheme iron(III) center coordinated by Tyr.
Consistent with the two reversible oxidation waves, a
controlled-potential electrochemical oxidation of 1 at 203
K shows two-stage UV-vis spectral changes with clear
isosbestic points, generating the one- and two-electron-
oxidized intermediates [1]+ and [1]2+ as shown in Figure 3.
[1]+ is generated through controlled-potential oxidation at
1.0 V for 2 h. An additional 2 h and an increased voltage of
1.3 V are required to generate [1]2+. The starting iron(III)
complex 1 displays a characteristic phenolate-to-iron(III)
charge-transfer band around 500 nm. Upon the first one-
electron oxidation to [1]+, the intensity of this band is
decreased, while an intense absorption at 346 nm and a broad
absorption at 780 and 870 nm grow. The second one-electron
oxidation from [1]+ to [1]2+, on the other hand, results in a
doubled intensity and a bathochromic shift (346 to 353 nm,
780 and 870 nm to 913 nm) in these absorption maxima.
These spectral changes are reversible, and the starting iron-
(III) complex is regenerated when the reversed potential
(-0.9 V) is applied at 203 K. The intense absorption around
350 nm for [1]+ and [1]2+ is indicative of the phenoxyl
radical formation, because phenoxyl radicals reported to date
usually exhibit an absorption around 400 nm.14,16 Chemical
oxidation of 1 with tris(2,4-dibromophenyl)aminium hexachlo-
roantimonate (5)20 also generates [1]+ and [1]2+.21 It was
reported that phenoxyl radicals from Cu22 and Ni salen23
Results
Detailed Electronic Structures of Oxidizing Intermedi-
ates from 1. We investigate in detail electrochemically
generated oxidizing intermediates from 1, which is a five-
coordinate monomeric iron salen complex bearing only one
Cl ligand as clarified by X-ray crystallographic analysis.18
The electrochemical behavior of 1 is investigated with cyclic
voltammetry. Cyclic voltammetry of 1 does not give any
clear oxidation wave at room temperature but gives two
reversible oxidation waves at 233 K (Figure 2), indicating
that oxidizing intermediates are stable at low temperature.
(20) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877-910.
(21) [1]+ is generated in high yield utilizing just 1.0 equiv of the oxidant
5, but the yield of [1]2+ is low utilizing even an excess of 5. See
Figure 5.
(19) (a) Sivasubramanian, V. K.; Ganesan, M.; Rajagopal, S.; Ramaraj, R.
J. Org. Chem. 2002, 67, 1506-1514. (b) Bryliakov, K. P.; Talsi, E.
P. Angew. Chem., Int. Ed. 2004, 43, 5228-5230.
8158 Inorganic Chemistry, Vol. 44, No. 22, 2005

Sterically Hindered Iron Salen Complexes
Figure 4. Resonance Raman spectra of (a) 1, (b) 1 + 1.0 equiv of the
oxidant 5, and (c) 1 + 2.0 equiv of the oxidant 5 in CD2Cl2 at 193 K. The
excitation wavelength is 351.4 nm. The band from CD2Cl2 is denoted with
“s”. The band designated with an asterisk is observable for the reaction
product from the oxidant 5 alone in the absence of 1. A low signal-to-
noise ratio for the spectrum of 1 is due to a smaller ꢀ value of 1 at
351.4 nm.
display intense absorptions (1750 and 1103 nm, respectively)
in the near-infrared (NIR) region. But no additional spectral
feature from 1100 to 2500 nm is observed for both [1]+ and
[1]2+ generated by use of 5.
To obtain more definitive evidence for the phenoxyl radical
formation, resonance Raman spectra are measured at 193 K
for [1]+ and [1]2+ generated by use of 5 in CD2Cl2, using an
excitation wavelength of 351.4 nm. As shown in Figure 4,
one of the C-O stretching band of two phenolates in 1 is
observed at 1306 cm-1.18 When the 1306 cm-1 band is
deleted by the radical formation, the nearby band at 1338
cm-1 is shifted to a lower frequency, indicating that the 1338
cm-1 mode is interacting with the C-O stretching. Upon
stepwise one-electron oxidation, the 1306 cm-1 band exhibits
decrease in intensity and instead a new band appears at 1485
cm-1. The 1485 cm-1 band slightly increases in intensity in
[1]2+ compared to [1]+. This band is assigned to Y7a’ of
the phenoxyl radical. Both 1338 and 1485 cm-1 bands appear
as a single band, suggesting that the two phenoxyl groups
have a similar structure and environment from the aspect of
resonance Raman spectroscopy.
Then, Mo¨ssbauer spectra of 1, [1]+, and [1]2+ dissolved
in CH2Cl2 are measured at 11 K, to investigate the elec-
tronic state of the central iron. As shown in Figure 5 a,
the iron(III) complex 1 gives a rather broad signal with an
isomer shift δ of 0.21 ( 0.07 mm s-1 and a quadrupole
splitting ∆EQ of 0.86 ( 0.14 mm s-1, indicative of high-
spin iron(III).24 Addition of 1.0 equiv of the oxidant 5
produces the spectrum shown in Figure 5b, which is fitted
with two subspectra i and ii. Upon addition of 4.0 equiv of
Figure 5. Zero-field Mo¨ssbauer spectra of (a) 1, (b) 1 + 1.0 equiv of the
oxidant 5, and (c) 1 + 4.0 equiv of the oxidant 5 in CH2Cl2 at 11 K. The
spectra in (b) and (c) were fitted with two subspectra i and ii. Mo¨ssbauer
parameters: δ ) 0.21 ( 0.07 mm s-1 and ∆EQ ) 0.86 ( 0.14 mm s-1 for
1; δ ) 0.43 ( 0.01 mm s-1 and ∆EQ ) 1.36 ( 0.02 mm s-1 for i; δ )
0.31 ( 0.03 mm s-1 and ∆EQ ) 0.42 ( 0.06 mm s-1 for ii.
the oxidant 5, the spectrum shown in Figure 5c is obtained.
The component ii, which is a minor component in Figure
5b, is considerably increased in Figure 5c. We thus as-
signed the component i with δ ) 0.43 ( 0.01 mm s-1 and
∆EQ ) 1.36 ( 0.02 mm s-1 to [1]+ and the component ii
with δ ) 0.31 ( 0.03 mm s-1 and ∆EQ ) 0.42 ( 0.06 mm
s-1 to [1]2+.
The isomer shifts δ of 0.43 and 0.31 mm s-1 clearly
indicate that both [1]+ and [1]2+ have a high-spin iron(III)
center. Iron(IV) complexes reported to date usually dis-
play very small isomer shifts (0.01-0.17 mm s-1).7,25 There-
fore, both the first and second one-electron oxidations are
not metal-centered, indicating formation of the iron(III)-
(22) Pratt, R. C.; Stack, T. D. P. J. Am. Chem. Soc. 2003, 125, 8716-
8717.
(23) Shimazaki, Y.; Tani, F.; Fukui, K.; Naruta, Y.; Yamauchi, O. J. Am.
Chem. Soc. 2003, 125, 10512-10513.
(24) (a) Kennedy, B. J.; McGrath, A. C.; Murray, K. S.; Skelton, B. W.;
White, A. H. Inorg. Chem. 1987, 26, 483-495. (b) Shyu, H.-L.; Wei,
H.-H.; Lee, G.-H.; Wang, Y. J. Chem. Soc., Dalton Trans. 2000, 915-
918.
Inorganic Chemistry, Vol. 44, No. 22, 2005 8159

Kurahashi et al.
Figure 6. X-band EPR spectra of (a) 1, (b) electrochemically generated
[1]+, and (c) [1]2+ at 4 K in frozen CH2Cl2 containing 0.1 M tetrabutyl-
ammonium perchlorate. Conditions: microwave frequency, 9.58 GHz;
microwave power, 101 µW; modulation amplitude, 7 G.
monophenoxyl and iron(III)-diphenoxyl radical complexes
for [1]+ and [1]2+, respectively.
To verify the above assignment, X-band EPR spectra are
recorded at 4 K for electrochemically generated [1]+ and
[1]2+ as shown in Figure 6. The neutral complex 1 displays
two sets of ferric high-spin signals at g ) 8.1, 3.5, and ca.
1.8 (E/D ≈ 0.10) and g ) 6.7, 5.2, and ca. 2 (E/D ≈ 0.03)
in frozen CH2Cl2 at 4 K.18 One-electron oxidation of 1 results
in disappearance of both signals. The silent EPR signal for
[1]+ is consistent with the monophenoxyl radical formation,
as a result of a strong intramolecular antiferromagnetic
coupling between the ferric ion (S ) 5/2) and the phenoxyl
radical (S ) 1/2). In contrast to two species observed for 1,
[1]2+ shows one set of EPR signals at g ) 4.10, 3.80, and
2.00, typical of St ) 3/2 (E/D ≈ 0.02). The observed St )
3/2 spin system is well accounted for by the diphenoxyl
radical formation, considering an antiferromagnetic coupling
between the ferric ion (S ) 5/2) and the two phenoxyl
radicals (S ) 1/2) on the salen ligand.
Figure 7. ESI mass spectra of (a) 1 + 1.2 equiv of the oxidant 5 and (b)
1 + 4.0 equiv of the oxidant 5 in CH2Cl2. The signal denoted with an
asterisk comes from the oxidant 5. (c, d) Observed (top) and calculated
(bottom) isotope distribution patterns. Calculations are done for (c) [1,
C70H74ClFeN2O2]2+ and (d) [1, C70H74ClFeN2O2]+.
explore a possibility of a high-valent iron-oxo formation
from the OH complex 4, following the successful examples
in the heme chemistry.26
Cyclic voltammograms are measured for 2 and 3 under
exactly the same experimental conditions for 1. Closely
similar to 1, both 2 and 3 exhibit two reversible oxidation
waves. The E1 and E2 values are almost identical with
1/2
1/2
those of 1 (0.80 and 1.00 V for 2, 0.82 and 1.00 V for 3) as
shown in Table 1.
Electrochemical oxidation of 2 in CH2Cl2 at 203 K
successfully generates [2]+ and [2]2+ as shown in Figure 8.
In the case of 2, chemical oxidation by 5 cannot be employed,
because the H2O ligand is easily replaced with Cl ions in
5.20 Formation of phenoxyl radicals similar to 1 is evident
from the characteristic absorption around 350 and 850-900
nm, compared to the spectra in Figure 3. EPR further
supports formation of the monophenoxyl radical for [2]+ and
the diphenoxyl radical for [2]2+ (Figure 9). One-electron
oxidation of 2 results in disappearance of the ferric signal
from 2, just in the same way as 1. The following one-electron
oxidation produces a St ) 3/2 EPR signal with g ) 4.30,
3.58, and 1.98 (E/D ≈ 0.07). It is interesting to note that
To establish the chemical structures for [1]+ and [1]2+,
ESI mass spectra are measured at low temperature in a
positive mode. The solution of [1]+ generated by 1.2 equiv
of 5 gives a single signal at m/z 1065.47 as a singly charged
ion (Figure 7a). The solution of [1]2+ generated by 4.0 equiv
of 5 produces predominantly a signal at m/z 532.71 as a
doubly charged ion (Figure 7b). These results are consistent
with formation of [Fe(salen)Cl]+ and [Fe(salen)Cl]2+, in
which one and two electrons are removed from the starting
1, with no concomitant chemical reaction (Figure 7c,d).
Effect of the Fifth Ligand. The fifth ligand could
considerably influence the electronic state and the geometry
of the resting iron(III) center. Then, the iron salen complexes
bearing H2O (2) and EtO (3) are investigated. We also
(26) (a) Lee, W. A.; Calderwood, T. S.; Bruice, T. C. Proc. Natl. Acad.
Sci. U.S.A. 1985, 82, 4301-4305. (b) Calderwood, T. S.; Lee, W. A.;
Bruice, T. C. J. Am. Chem. Soc. 1985, 107, 8272-8273. (c) Groves,
J. T.; Gilbert, J. A. Inorg. Chem. 1986, 25, 123-125. (d) Calderwood,
T. S.; Bruice, T. C. Inorg. Chem. 1986, 25, 3722-3724. (e) Swistak,
C.; Mu, X. H.; Kadish, K. M. Inorg. Chem. 1987, 26, 4360-4366. (f)
Groves, J. T.; Gross, Z.; Stern, M. K. Inorg. Chem. 1994, 33, 5065-
5072.
(25) (a) Collins, T. J.; Kostka, K. L.; Mu¨nck, E.; Uffelman, E. S. J. Am.
Chem. Soc. 1990, 112, 5637-5639. (b) Collins, T. J.; Fox, B. G.; Hu,
Z. G.; Kostka, K. L.; Mu¨nck, E.; Rickard C. E. F.; Wright, L. J. J.
Am. Chem. Soc. 1992, 114, 8724-8725. (c) Kostka, K. L.; Fox, B.
G.; Hendrich, M. P.; Collins, T. J.; Rickard C. E. F.; Wright, L. J.
Mu¨nck, E. J. Am. Chem. Soc. 1993, 115, 6746-6757.
8160 Inorganic Chemistry, Vol. 44, No. 22, 2005

Sterically Hindered Iron Salen Complexes
Figure 10. UV-vis spectral changes upon electrochemical oxidation of
3 in CH2Cl2-EtOH (95:5) at 203 K. Conditions: 0.5 × 10-3 M 3; 0.10 M
tetrabutylammonium perchlorate supporting electrolyte; controlled-potential
oxidation at 0.83 V vs Fc′/Fc.
Figure 8. UV-vis spectral changes upon electrochemical oxidation of 2
in CH2Cl2 at 203 K. Conditions: 0.5 × 10-3 M 2; 0.10 M tetrabutylam-
monium perchlorate supporting electrolyte; controlled-potential oxidation
at 1.00 V for [2]+ and 1.30 V for [2]2+ vs Fc′/Fc. 2, [2]+, and [2]2+ are
depicted in a solid line.
oxidized. An EPR measurement is consistent with this
interpretation, and the ferric high-spin signal remains intact
in intensity after the irreversible one-electron oxidation. As
expected, quantitative analysis of the resulting solution after
controlled-potential oxidation of 3 shows generation of
acetaldehyde in ca. 25% net yield. Notably, the present iron
salen complex exhibits a considerably high reactivity for the
coordinated alcohol.
To examine the reactivity of [2]+, 2 equiv of benzyl
alcohol is introduced at 203 K to the solution of the
electrochemically generated [2]+. Surprisingly, the UV-vis
spectrum of [2]+ remains intact, and benzyl alcohol is not
oxidized. However, upon addition of 2,6-di-tert-butylpyri-
dine, the characteristic UV-vis spectrum of [2]+ is dimin-
ished very quickly. Indeed, analysis of the resulting solution
with GC-MS shows formation of benzaldehyde in 56 ( 2%
yield. Considering the high reactivity for EtO group upon
one-electron oxidation of 3, a base to remove a proton from
alcohol OH is required to initiate oxidation sequences from
alcohol to aldehyde. Reactivity of [2]2+ is investigated in
the same manner, but the yield of benzaldehyde is not
changed at all (57 ( 5%), indicating that the one-electron
oxidized state from the iron(III) complex is enough for
oxidation of alcohols. Reaction with cyclohexene is also
examined, but cyclohexene oxide, 2-cyclohexene-1-ol, and
2-cyclohexene-1-one are not detected at all (<0.5%).
The OH complex 4 is produced by addition of tetrabutyl-
ammonium hydroxide to the CH2Cl2 solution of 2 through
acid-base equilibrium but under an excess of tetrabutylam-
monium hydroxide (2.5 equiv). Cyclic voltammetry of 4
gives two reversible oxidation waves at 0.79 and 0.95 V
(Table 1), with an additional broad irreversible oxidation
wave around 0.45 V, which is assigned to oxidation of
tetrabutylammonium hydroxide. Upon electrochemical oxi-
dation of 4, the UV-vis spectral change follows acid-base
equilibrium back to 2, probably due to oxidation of an excess
of tetrabutylammonium hydroxide. The further oxidation of
the resulting solution then generates an EPR-silent species,
which is an iron(III)-phenoxyl radical, judging from the
characteristic UV-vis spectrum at λ (ꢀ/104 M-1 cm-1) )
345 (1.87) and 850 (0.94) nm.
Figure 9. X-band EPR spectra of (a) 2, (b) electrochemically generated
[2]+, and (c) [2]2+ at 4 K in frozen CH2Cl2 containing 0.1 M tetrabutyl-
ammonium perchlorate. Conditions: microwave frequency, 9.58 GHz;
microwave power, 101 µW; modulation amplitude, 7 G.
[2]2+ (E/D ≈ 0.07) is more rhombic than the [1]2+ (E/D ≈
0.02), while the neutral complex 2 (E/D ≈ 0.21) has also a
more rhombic iron center than 1 (E/D ≈ 0.03, 0.10).
In contrast to 1 and 2, a UV-vis spectral change upon
electrochemical oxidation of 3 shows no isosbestic points,
indicative of more than two oxidation pathways. Upon
addition of 5% of EtOH to the CH2Cl2 solution of 3, one of
these pathways seems to prevail as suggested by the clear
isosbestic points (Figure 10). However, the spectral change
is distinctively different from those of 1 and 2. The
absorption at 350 and 530 nm is increased, while the
absorption at 410 nm is decreased. The resulting spectrum,
which is rather close to that of the aqua complex 2, is not
affected at all upon the reversed-potential electrochemical
oxidation (-0.9 V) at 203 K, suggesting that the coordinated
EtO group is oxidized upon one-electron oxidation. Then,
the EtO group is replaced with the CD3O group, which is
more resistant to oxidation. But the UV-vis spectral change
is exactly the same, suggesting that even the CD3O group is
Inorganic Chemistry, Vol. 44, No. 22, 2005 8161

Kurahashi et al.
Figure 13. Zero-field Mo¨ssbauer spectra of 2 + 3 eqiuv of mCPBA in
toluene-CH2Cl2 (3:7) at 5 K. The spectrum was fitted with three subspectra
a-c. Mo¨ssbauer parameters: δ ) 0.66 ( 0.01 mm s-1 and ∆EQ ) 1.33 (
0.04 mm s-1 for a; δ ) 0.52 ( 0.06 mm s-1 and ∆EQ ) 0.86 ( 0.01 mm
s-1 for b. For c, the Mo¨ssbauer parameters could not be determined.
Figure 11. UV-vis spectra of 2 and 2 + 1 equiv of mCPBA in CH2Cl2
at 193 K. Conditions: 1.5 × 10-4 M 1; 1.5 × 10-4 M mCPBA.
utilization of a large excess of 2 (20 equiv), 1 equiv of
mCPBA generates 2 equiv of the monophenoxyl radical,
suggesting that all the oxidizing equivalents from mCPBA
are retained in the blue-green solution.
The Mo¨ssbauer spectrum, on the other hand, indicates that
the blue-green solution contains equimolar amounts of the
components a and b as shown in Figure 13. The component
a is fitted with δ ) 0.66 ( 0.01 mm s-1 and ∆EQ ) 1.33 (
0.04 mm s-1, and the component b is fitted with δ ) 0.52
( 0.06 mm s-1 and ∆EQ ) 0.86 ( 0.01 mm s-1. In addition,
a trace amount of the component c is observed, but the
Mo¨ssbauer parameters could not be determined. As indi-
cated by the δ values, both the component a and b contain
an iron(III) center. The ∆EQ value for the component a is
almost identical with that for [1]+, supporting formation of
the monophenoxyl radical. But the ∆EQ value for the
component b is rather small.
To get insight into the fifth ligand of oxidizing inter-
mediates in the blue-green solution, the ESI mass spec-
trum is measured at low temperature in a positive mode. The
CH2Cl2 solution of 2 gives a single signal at m/z 1030.51 as
a singly charged ion, which corresponds to [FeIII(salen)]+
with loss of H2O and ClO4- from 2. The blue-green solution
generated from 2 and 3 equiv of mCPBA gives a predominant
signal at m/z 1046.43 as a singly charged ion. The mass value
of m/z 1046.43 is increased by 16 Da from [FeIII(salen)]+.
By use of the 18O-labeled mCPBA, this signal is shifted by
2.0 Da, indicating that the predominant species detected with
a mass spectrometer is [Fe(salen)O]+ and that the oxygen
atom comes from mCPBA. Reduction of the blue-green solu-
tion with ferrocene results in complete disappearance of the
signal at m/z 1046.43. Instead, the signal at m/z 1030.64 is
regenerated, together with an unidentified signal at m/z
1231.67, indicating that the signal at m/z 1046.43 arises from
a transient intermediate and not from a ligand-oxidation
product.
Figure 12. X-band EPR spectra of (a) 2 and (b) 2 + 3 equiv of mCPBA
at 1.6 K in frozen toluene-CH2Cl2 (3:7). Conditions: microwave frequency,
9.59 GHz; microwave power, 101 µW; modulation amplitude, 7 G.
Oxidizing Intermediates from the Iron Salen Complex
and mCPBA. With the fully characterized iron(III)-phe-
noxyl radical from the iron salen complex on hand, we then
investigate oxidizing intermediates generated by mCPBA,
which has been frequently utilized to generate high-valent
metal-oxo species in the porphyrin chemistry.17 Addition
of 1 equiv of mCPBA to the CH2Cl2 solution of 2 generates
a blue-green solution with absorption maxima at 344 and
860 nm shown in Figure 11. Titration of 2 with mCPBA
shows a saturation behavior, and the absorption at 344 and
860 nm does not increase any more upon addition of more
than ca. 0.8 equiv of mCPBA. Compared to the UV-vis
spectra for [1]+ and [1]2+ in Figure 3, the absorption at 344
and 860 nm is indicative of the monophenoxyl radical
formation on the salen ligand. Indeed, using an excitation
wavelength at 413.1 nm, a broad resonance Raman band
appears at 1485 cm-1, exactly the same wavenumber as [1]+.
An EPR measurement also supports the monophenoxyl
radical formation around the iron(III) center. The ferric high-
spin signal for 2 disappears in the blue-green solution (Figure
12). An additional feature is observed at g ≈ 2, but the origin
of this signal is not identified. Stoichiometry of mCPBA to
the monophenoxyl radical formation is roughly evaluated
Discussion
Phenoxyl Radicals from the Iron Salen Complex.
Oxidizing intermediates generated through stepwise one-
using the ꢀ855
value for [2]+ (13 100 M-1 cm-1). With
nm
8162 Inorganic Chemistry, Vol. 44, No. 22, 2005

Sterically Hindered Iron Salen Complexes
electron oxidation of the iron(III) salen complex is subjected
to a thorough spectroscopic investigation. The N2O2 coor-
dination environment of the salen ligand is indeed found for
the active site in nonheme iron(III) enzymes, such as 3,4-
PCD.11 In addition, the metal salen complex is one of the
most important catalysts in synthetic chemistry.27 Therefore,
the electronic structure of oxidizing intermediates from the
iron salen complex assumes a fundamental importance, not
only from a bioinorganic aspect but also from a design of
effective oxidation catalysts.
signal, suggesting that one of the species is converted to the
other upon stepwise one-electron oxidation.
The phenoxyl radical complexes from 1 display a π f
π* transition of the phenoxyl radical at λmax ≈ 346 nm, which
is largely shifted from the reported values of 400 ( 20 nm
for phenoxyl radicals.13,14,16 The observed hypsochromic shift
is probably due to introduction of the electron-withdrawing
imino groups to the ortho position of the phenolate ring,
causing a consierable energy lowering of the HOMO level.
Indeed, 1 exhibits a considerably high oxidation potential
(E1 ) 0.85 V), compared to other examples such as the
Stepwise one-electron oxidation of 1 bearing Cl as a fifth
ligand generates the one- and two-electron-oxidized inter-
mediates [1]+ and [1]2+, having an intense absorption around
350 nm and a characteristic broad absorption around 500-
1100 nm. This study clearly indicates that [1]+ and [1]2+
are best represented as the mono- and diphenoxyl radicals
coordinated to the iron(III) center, respectively. It is not trivial
to determine oxidation states of iron centers only from
Mo¨ssbauer parameters, when noninnocent ligands are coor-
dinated to the iron center, as discussed recently by Walker
and Trautwein.28 But in the present case isomer shifts δ for
both [1]+ and [1]2+ are rather increased from the starting
iron(III) complex 1, indicating that the iron center in both
[1]+ and [1]2+ bears iron(III) character. In addition, the
phenoxyl radical formation on the salen ligand is suggested
by the intense absorption around 350 nm. Indeed, resonance
Raman measurements upon excitation at 351.4 nm give a
resonance Raman band around 1485 cm-1, which could be
assigned to the phenoxyl radical C-O stretching.
The complex 1 shows two sets of EPR signals from high-
spin iron(III) centers in frozen CH2Cl2 at 4 K, major
component with g ) 8.1, 3.5, and ca. 1.8 (E/D ≈ 0.10) and
minor component with g ) 6.7, 5.2, and ca. 2 (E/D ≈ 0.03).
This suggests that two iron(III) species of different rhom-
bicity are present in solution. The EPR parameter of the
major component is close to that of the previous report.19b
On the other hand, we assigned the minor component to a
six-coordinated form resulting from binding of ethanol to 1
in our previous paper,18 but this species is still present when
EPR spectra are measured in amylene-stabilized CH2Cl2 for
the well-dried 1. We thus reinterpret that 1 in a five-
coordinated form may have a metastable structure due to
steric repulsion from mesityl groups. In frozen solution, the
stable structure, which has been shown by the X-ray crystal
structure, would provide the EPR signals for the major
component and the metastable structure would be the EPR
signals for the minor component. On the other hand, 1H NMR
in CD2Cl2 at 193 K shows only one set of signals for the
phenolate rings, indicating that these two species are inter-
converted with each other very fast compared to the NMR
time scale. Upon stepwise one-electron oxidation, both
species are oxidized in the same manner as indicated by EPR.
In contrast to 1, [1]2+ exhibits only one set of a S ) 3/2
1/2
Wieghardt’s macrocylic iron(III) complex (E11/2 ) 0.11 V).14c
The other characteristic feature in the UV-vis spectra of
the present phenoxyl radical is an intense broad absorption
around 500-1100 nm (ꢀ ) 10 000, 23 000 M-1 cm-1 for
[1]+ and [1]2+, respectively). The best assignment for this
absorption feature is the charge transfer between the iron-
(III) center and the phenoxyl radical, because [1]2+, in which
all the two phenolates are converted to the phenoxyls,
displays the broad absorption of a doubled intensity in this
region, compared to [1]+ bearing one phenolate and one
phenoxyl.
Quite interestingly, the active form of galactose oxidase
(GO) does exhibit a similar feature (ꢀ ) ∼5000 M-1 cm-1)
around 600-1200 nm, but Whittaker and Spiro proposed
that this feature arises from a interligand charge transfer from
the tyrosinate to the thiolate-modified tyrosyl radical.29
Synthetic phenoxyl radicals from the Cu and Ni salen
complexes were reported to show a NIR absorption feature
at 1750 nm (ꢀ ) 2400 M-1 cm-1)22 and 1103 nm (ꢀ ) 3000
M-1 cm-1),23 respectively. In the case of the Cu salen, the
absorption at 1750 nm is assigned to the phenolate-phenoxyl
charge transfer. But [1]+ exhibits no additional absorption
in the NIR region from 1100 to 2500 nm, indicating that the
phenolate-phenoxyl charge-transfer absorption in [1]+ is not
present or hidden behind the intense absorption around 500-
1100 nm. Considering that 1 has two distinct phenolate rings
due to the distortion,18 the phenolate-phenoxyl charge-
transfer absorption in [1]+ is possibly shifted to higher
energy, compared to the nondistorted Cu salen bearing two
phenolate rings of a similar environment. The absorption
feature around 780 nm in [1]+ might be a candidate, because
this feature disappears in [1]2+.
It is also interesting to note that a quadrupole splitting
∆EQ is much larger for [1]+ (1.36 mm s-1) than 1 (0.86 mm
s-1). The ∆EQ value of 0.86 mm s-1 for 1 clearly indicates
high-spin iron(III), because low-spin iron(III) salen com-
plexes exhibit larger ∆EQ value (>2 mm s-1).24 The larger
∆EQ is indicative of lower symmetry around the iron center.
In the previous report by Koikawa et al. on one-electron
oxidation of a similar iron(III)-phenolate system, a large
∆EQ is ascribed to the iron(IV) formation.30 However, as
already pointed out by Wieghardt,13,14 a large ∆EQ for
iron(III)-phenolate systems is rather consistent with the
(27) McGarrigle, E. M.; Gilheany, D. G. Chem. ReV. 2005, 105, 1563-
1602.
(29) McGlashen, M. L.; Eads, D. D.; Spiro, T. G.; Whittaker, J. W. J.
Phys. Chem. 1995, 99, 4918-4922.
(30) Koikawa, M.; Okawa, H.; Maeda, Y.; Kida, S. Inorg. Chim. Acta 1992,
194, 75-79.
(28) Zakharieva, O.; Schu¨nemann, V.; Gerdan, M.; Licoccia, S.; Cai, S.;
Walker, F. A.; Trautwein, A. X. J. Am. Chem. Soc. 2002, 124, 6636-
6648.
Inorganic Chemistry, Vol. 44, No. 22, 2005 8163

Kurahashi et al.
Scheme 1
phenoxyl radical formation, when it is assumed that the
radical is localized to one of the phenolate rings of the salen
ligand on the Mo¨ssbauer time scale, leading to an electroni-
cally unsymmetrical environment. Our study provides further
support for the interpretation on the ∆EQ value, because
oxidation of both phenolate rings on the salen ligand,
generating more symmetrical environment than [1]+, indeed
results in the smaller ∆EQ value for [1]2+ (0.42 mm s-1).
Vacant Coordination Site on the Iron Salen Complex.
A fifth ligand on the vacant coordination site could be an
important factor to determine the electronic state and the
geometry of the resting iron(III) center. The effect of a fifth
coordination ligand is worth discussing here, because previ-
ous iron(III)-phenoxyl radical complexes have no vacant
coordination site on the iron(III) center. From this viewpoint,
particularly interesting for the present model system is a
drastic structural change upon exchange of the fifth ligand
from Cl (1) to OH2 (2).18 As revealed by X-ray crystal-
lographic analysis, the ferric aqua complex 2 displays a
distorted trigonal-bipyramidal iron(III) center, while the
structure of 1 is rather close to the common square-
pyramidal. Such a structural difference is seemingly retained
in solution, because EPR measurements indicate that 2 (E/D
≈ 0.21) has a much more rhombic iron center than 1 (E/D
≈ 0.03, 0.10). However, the structural difference around the
iron center does not exert any observable influence on the
electrochemical behavior of 1 and 2, and both 1 and 2
generate the iron(III)-phenoxyl radical complex upon
electrochemical oxidation. Quite interestingly, the structural
distortion caused by the coordination of H2O is retained in
the oxidized form of 2, because [2]2+ (E/D ≈ 0.07) is also
more rhombic than [1]2+ (E/D ≈ 0.02)
benzyl alcohol upon addition of 2,6-di-tert-butylpyridine.
High reactivity of the present iron(III)-phenoxyl radical for
alcohols may imply that the copper ion in the active site of
GO could be replaced with other metals, just when the metal
center binds a substrate in close vicinity to the phenoxyl
radical and also functions as a Lewis acid. Indeed, phenoxyl
radical complexes with other metals such as zinc and alkaline
earth metal ions have already been reported to exhibit
oxidation reactivity.16d,e,32
In contrast to the oxoiron(IV) formation from the hydroxo-
iron(III) porphyrin upon one-electron oxidation, the iron salen
complex 4 bearing OH does not generate a high-valent iron
species. Upon the initial stage of electrochemical oxidation,
the fifth OH ligand in 4, which is retained under an excess
of tetrabutylammonium hydroxide, is supposed to be pro-
tonated to a considerable extent due to oxidation of an excess
of tetrabutylammonium hydroxide. This is supported by the
UV-vis spectral change, which suggests that an acid-base
equilibrium is considerably shifted from 4 to 2. Further
oxidation generates an iron(III)-phenoxyl radical, but the
fifth ligand is not pure OH.
Oxidizing Intermediates from the Iron Salen Complex
and mCPBA. In the case of the heme chemistry, a model
intermediate relevant to the oxygen activation by heme
enzymes was first synthesized from the iron(III) porphyrin
complex and mCPBA and was identified to be the oxoiron-
(IV) porphyrin π-cation radical by Groves.17a On the other
hand, reaction of 2 with mCPBA at 193 K generates the blue-
green solution whose UV-vis spectrum is closely similar
to the iron(III)-monophenoxyl radicals such as [1]+ and [2]+.
The EPR spectrum further supports the monophenoxyl
radical formation, because the ferric high-spin signal for 2
disappears in the blue-green solution. But the Mo¨ssbauer
spectrum reveals that the blue-green solution contains two
iron(III) intermediates of an equimolar amount.
We postulate the reaction sequence shown in Scheme 1,
based on the experimental observation that all the oxidizing
equivalents are retained in the blue-green solution. Reaction
of 2 with mCPBA gives a two-electron-oxidized intermediate
6 at first. But 6 is readily reduced by another molecule of 2
to give two molecules of monophenoxyl radicals 7 and 8.
This accounts well for the saturation behavior by addition
of less than 1 equiv of mCPBA. An excess of mCPBA
The present model system is the first example of the
iron(III)-phenoxyl radical complex bearing the vacant coor-
dination site and, thus, provides an opportunity to examine
reactivity relevant to iron enzymes having the phenolate
group coordinated to the iron(III) center. Particularly remark-
able is one-electron oxidation of the EtO or CD3O complexes
causes conversion of alcoxide to aldehyde even at low
temperature. The iron(III) porphyrin cation radical does not
have such reactivity at all, but coordination of the electron-
donating CH3O groups results in formation of a rather stable
iron(IV) intermediate.31 One of interesting findings from a
mechanistic perspective is a clear demonstration of necessity
of a base to remove an alcohol proton. As already described,
both [2]+ and [2]2+ do not oxidize even benzyl alcohol at
all, unless 2,6-di-tert-butylpyridine is added to the solution.
Once benzyl alcohol is converted to alcoxide on the iron-
(III) center, conversion to benzaldehyde proceeds smoothly
without a base, as suggested by oxidation of EtO in 3.
Another point to be noted is that a negligible difference is
observed for reactivity of [2]+ and [2]2+, indicating that the
monophenoxyl radical complex is the key intermediate for
oxidation of alcohols and that one more oxidizing equivalent
does not contribute to effectiveness. This might be due to
very fast reduction of [2]2+ to [2]+ before reaction with
(31) Groves, J. T.; Quinn, R.; McMurry T. J.; Nakamura, M.; Lang, G.;
(32) Itoh, S.; Kumei, H.; Nagatomo, S.; Kitagawa, T.; Fukuzumi, S. J.
Am. Chem. Soc. 2001, 123, 2165-2175.
Boso, B. J. Am. Chem. Soc. 1985, 107, 354-360.
8164 Inorganic Chemistry, Vol. 44, No. 22, 2005

Sterically Hindered Iron Salen Complexes
converts 8 to 9 by exchange of the fifth ligand from H2O to
m-chloroperoxybenzoate.
A considerably high reactivity for coordinated substrate on
the iron(III) center might be a hint to understand a rather
unique reaction sequence upon oxygen activation.
Among two major components observed in the Mo¨ssbauer
spectrum, the component a with δ ) 0.66 mm s-1 and ∆EQ
) 1.33 mm s-1 would correspond to 7, because the ∆EQ
value, which is almost identical with that for [1]+, indicates
the unsymmetrical monophenoxyl radical. The larger isomer
shift δ of 0.66 mm s-1 compared to [1]+ could be explained
by coordination of the more electron-donating ligand OH as
a fifth ligand. We tentatively assign the component b with
δ ) 0.52 mm s-1 and ∆EQ ) 0.86 mm s-1 to 9. The smaller
∆EQ value might be due to an electron-accepting nature of
the m-chloroperoxybenzoate ligand to increase symmetry
around the iron(III) center. ESI mass spectrometry supports
the above scenario, and the signal at m/z 1046.43 could be
considered to originate from 9 through a homolytic cleavage
of the weak O-O bond during an ionization process. This
signal apparently matches the oxoiron(IV) salen phenoxyl
radical complex or the oxoiron(V) salen complex, but the
presence of such species in solution is denied by the other
spectroscopic data.
Implications. We employ two independent routes to the
nonheme high-valent iron-oxo intermediate from the iron
salen complex, electrochemical oxidation of the OH complex
4 and mCPBA oxidation of 2, both of which are the
established pathways to the high-valent metal-oxo species
in the porphyrin chemistry. Unexpectedly, both routes lead
to generation of very unstable iron(III)-phenoxyl radicals,
although the salen ligand is harder to be oxidized than the
porphyrin ligand bearing the large π-conjugation. It is also
in sharp contrast to the heme chemistry that electrochemical
oxidation of 3 bearing an electron-donating EtO ligand results
in oxidative degradation of the EtO group and not in the
iron(IV) formation. These results indicate that the phenolate
(Tyr) coordination to the iron(III) center brings about a
considerable preference for the iron(III) state over the iron-
(IV), irrespective of a high oxidation potential of the
surrounding coordination environment.
A bunch of mononuclear nonheme iron(II) enzymes, which
share the common His2Asp/Glu motif around the iron(II)
center referred to as the 2-His-1-carboxylate facial triad, are
proposed to employ similar mechanistic strategies with heme
enzymes in oxidation catalysis.33 Indeed, a compound-II-
like oxoiron(IV) intermediate was detected in the oxidation
cycle by TauD, a mononuclear nonheme iron enzyme of this
category.6 Furthermore, several iron(II) model complexes
bearing the similar nitrogen coordination environment were
demonstrated to generate an oxoiron(IV) species,7 suggesting
that the His-coordinated iron(II) center utilizes an oxoiron-
(IV) species to effect oxidation reactions. On the other hand,
a mechanistic scenario seems to be utterly different for
mononuclear iron(III) enzymes with the His2Tyr2 motif
around the iron(III) center, such as 3,4-PCD. As clarified in
this manuscript, the phenolate-coordinated iron(III) center
sticks to the iron(III) state and, therefore, follows a reaction
pathway mechanistically distinct from the heme paradigm.
Finally, we describe another important implication for this
study. The salen ligand is one of the most successful
frameworks for development of effective transition metal
catalysts, but disappointedly low reactivity has been repeat-
edly encountered for the iron metal. The poor catalytic
activity for epoxidation compared to Mn and Cr salens can
be now understood very well by a considerable preference
for the iron(III) state in the salen framework. By analogy to
the heme chemistry, the iron salen complex has been
elucidated to give a compound I analogue as a reactive
intermediate, but this might be not true for most iron salen
complexes in light of the present comprehensive investiga-
tion. Recently, Rajagopal and Ramaraj reported a mechanistic
study on oxygenation of organic sulfides by iron salen
complexes.19a They elucidated the oxoiron(IV) salen radical
formation from several iron salen complexes and PhIO on
the basis of limited spectroscopic data and discussed the
oxygenation mechanism. But we have to point out that their
mechanistic elucidation remains to be investigated.
Conclusion
The model complex for mononuclear iron(III) enzymes,
bearing two His and two Tyr coordinated to the iron(III)
center, is investigated in detail to reveal the relationship
between the electronic structure and the reactivity of oxidiz-
ing intermediates. The model complex employed in this study
is the sterically hindered iron salen complex, which well
reproduces a coordination environment by His2Tyr2 and most
importantly a vacant coordination site on the iron(III) center.
In contrast to the oxoiron(IV) formation from heme models
and His-coordinated nonheme iron(II) centers, the iron salen
complex generates the iron(III)-phenoxyl radical irrespective
of the fifth ligand. The nonheme iron center coordinated with
the phenolate group is resistant to adopt the iron(IV) state
even in a highly oxidizing environment. The vacant coor-
dination site also enables investigations on reactivity, which
reveals a highly oxidizing nature of the iron(III)-phenoxyl
for coordinated alcohols. This study explains unique proper-
ties of mononuclear nonheme enzymes with Tyr residues
and also the poor epoxidation activity of Fe salen compared
to Mn and Cr salens.
Experimental Section
Instrumentation. Cyclic voltammograms were measured with
a ALS612A electrochemical analyzer (BAS). UV-vis spectra were
recorded on an Agilent 8453 (Agilent Technologies) equipped with
a USP-203 low-temperature chamber (UNISOKU). UV-vis spectra
in the NIR region were measured with Hitachi U-3500. EPR spectra
were recorded in a quartz cell (d ) 5 mm) at 4 K on an E500
continuous-wave X-band spectrometer (Bruker) with an ESR910
helium-flow cryostat (Oxford Instruments). Resonance Raman
spectra were measured with a quartz spinning cell kept at 193 K,
using a MC-100DG 100-cm single polychromator (Ritsu Ohyo
Kogaku) equipped with a LN/CCD-1100-PB liquid- nitrogen-cooled
CCD detector (ROPER Scientific). A BeamLok 2080 argon ion
(33) Hegg, E. L.; Que, L., Jr. Eur. J. Biochem. 1997, 250, 625-629.
Inorganic Chemistry, Vol. 44, No. 22, 2005 8165

Kurahashi et al.
laser (Spectra Physics) was utilized as an excitation source. The
laser power at the sample was about 10 mW. Raman shifts were
calibrated with indene. ESI mass spectra were obtained with a LCT
time-of-flight mass spectrometer equipped with an electrospray
ionization interface (Micromass).
2 h. Then, [1]2+ and [2]2+ were generated after an additional 2 h
by controlled potential oxidation at the voltage that was higher by
300 mV than the second oxidation potential.
Preparative Electrochemical Oxidation. Exactly the same
method described above was employed except for utilization of a
0.1 cm thin-layer quartz cell. After formation of one- and two-
electron-oxidized intermediates was confirmed with UV-vis
spectroscopy, the electrodes were removed from the quartz cell.
The intermediate solution was carefully transferred via a cooled
gastight syringe to an EPR sample tube or a reaction tube to measure
EPR or investigate reactivity.
Mo1ssbauer Measurements. Mo¨ssbauer spectra were measured
with a conventional spectrometer in the constant-acceleration mode.
Isomer shifts were reported relative to metallic iron (R-Fe) at room
temperature. Mo¨ssbauer spctra were measured for 57Fe-enriched
complexes. The sample solution was prepared at 203 K in a test
tube and then transferred via a precooled gastight syringe to a Teflon
Mo¨ssbauer cell. Immediately, a Mo¨ssbauer cell was immersed in a
liquid-nitrogen bath and then attached to the sample holder. To
obtain clear spectra for samples in CH2Cl2, data were accumulated
for 3-7 days. Simulation was conducted without any restriction.
Low-Temperature ESIMS. To obtain mass spectra of transient
species, both desolvation and nebulizer gas were cooled with liquid
nitrogen. The sample solution was prepared in a test tube at 203
K, and a portion of the solution was diluted with CH2Cl2 in a
gastight syringe cooled with dry ice. The sample solution was then
transferred to the inlet of the mass spectrometer through precooled
capillary tubing.
Chemical Oxidation of 1 by 5. For Mo¨ssbauer, resonance
Raman, and ESIMS measurements of [1]+ and [1]2+, chemical
oxidation by 5 was employed. Preparation of the resonance Raman
sample is described as a representative procedure: To the solution
of 1 (1.92 mg, 1.80 µmol) in CH2Cl2 (0.6 mL) in an acetone-CO2
bath was added 1.0 or 2.0 equiv of 5 (2.11 mg, 1.80 µmol, or 4.21
mg, 3.60 µmol) in CH2Cl2 (0.4 mL). [1]+ and [1]2+ were generated
immediately
mCPBA Oxidation of 2. Preparation of the UV-vis sample is
described as a representative procedure: To the CH2Cl2 solution
of 2 (1.67×10-4 M, 1.8 mL) in a UV-vis cell at 193 K was added
a CH2Cl2 solution of mCPBA (1.5 × 10-3 M, 0.2 mL). The mixture
was stirred at 193 K for 5 min, and then a UV-vis spectrum was
measured.
Materials. CH2Cl2 was purchased from Kanto as anhydrous
solvent and was stored in the presence of 4A molecular sieves. To
remove a trace of HCl, CH2Cl2 was passed through activated
alumina under an Ar atmosphere just before use. Tetrabutylam-
monium perchlorate was purchased from Kanto and was dried over
P2O5 in vacuo. CD3OD was purchased from ACROS and was dried
over 3A molecular sieves. CD2Cl2 was purchased from ACROS
and was stored in the presence of 4A molecular sieves. CD2Cl2
was also passed through activated alumina under an Ar atmosphere
just before use. 57Fe metal (95.40%) was purchased from ISOFLEX.
The oxidant 5 was prepared according to the reported method34
and was assayed with titration by ferrocene. mCPBA was purchased
from Nacalai and was purified by washing with a phosphate
buffer.35 18O-labeled mCPBA was prepared according to the
reported method.36 Purity of mCPBA was checked with iodometry.
Other reagents were purchased from Kanto or Aldrich and were
used as received.
Synthesis. Preparations of 1 and 2 have been already reported.18
3. To the complex 1 (50 mg, 45 µmol) in anhydrous THF (0.5
mL) was added NaOEt (3.1 mg, 45 µmol) in anhydrous EtOH (5
mL). The solution was stirred at room temperature for 10 min,
yielding a reddish precipitate. The resulting precipitate was filtered
out and washed with a small amount of EtOH. The precipitate was
dissolved in hexane and was passed through a membrane filter (pore
size 0.5 µm) to remove a trace of NaCl. Recrystalization from hot
EtOH gave a red crystalline product of 3 (35.7 mg, 33 µmol). Anal.
Calcd for [C72H79FeN2O3](H2O): C, 79.03; H, 7.46; N, 2.56.
Found: C, 78.68; H, 7.31; N, 2.59.
4. The solution of 4 was prepared by dissolving 2 (4.37 mg,
3.69 µmol) in 0.1 M tetrabutylammonium perchlorate/CH2Cl2 (3.7
mL) containing 2.5 equiv of tetrabutylammonium hydroxide (7.38
mg, 9.23 µmol) and utilized without purification.
The iron salen complex bearing the CD3O group was prepared
in exactly the same manner using NaOCD3/CD3OD instead of
NaOEt/EtOH. Anal. Calcd for [C71H74D3FeN2O3](H2O)2: C, 77.43;
H, 7.69; N, 2.54. Found: C, 77.71; H, 7.33; N, 2.64.
57Fe-enriched 1 was synthesized from 57Fe(OAc)2, which was
prepared from 57Fe metal (95.40%) and acetic acid. 57Fe-enriched
2 was prepared from 57Fe-enriched 1 according to the preparation
of 2.18
UV-Vis Measurements upon Controlled-Potential Electro-
chemical Oxidation. Controlled-potential electrochemical oxi-
dation was conducted in a thin-layer quartz cell (0.05 cm) using a
gold-mesh working electrode, a platinum-wire counter elec-
trode, and an Ag/AgCl reference electrode, which are connected
to a HA-151 potentiostat-galvanostat (Hokuto Denko). The quartz
cell was cooled at 203 K, and then the constant potential was applied
at the voltage which was higher by 150 mV than the first oxidation
potential. UV-vis spectral changes were monitored during elec-
trochemical oxidation to confirm generation of [1]+ and [2]+ after
Quantitative Analysis. The yield of bezaldehyde was determined
with a QP5000 GC-MS (Shimadzu), using benzophenone as an
internal standard. Quantitative analysis of acetaldehyde was carried
out by the reported method.37 The net yield of acetaldehyde was
calculated, considering the blank test under exactly the same
conditions. Each quantitative analysis was repeated three times, and
the average values were reported.
Acknowledgment. We thank Mr. S. Makita in our
Institute for elemental analyses. We also thank Equipment
Development in our Institute for preparation of the Mo¨ss-
bauer cell. This work was supported by a Grant-in Aid for
Scientific Research from the Ministry of Education, Culture,
Sports, Science, and Technology of Japan.
(34) Schmidt, W.; Steckhan, E. Chem. Ber. 1980, 113, 577-585.
(35) Schwartz, N. N.; Blumbergs, J. H. J. Org. Chem. 1964, 29. 1976-
1979.
IC051377E
(37) Miyake, T.; Shibamoto, T. J. Agric. Food Chem. 1993, 41, 1968-
(36) Wagner, W. R.; Rastetter, W. H. J. Org. Chem. 1983, 48, 402-404.
1970.
8166 Inorganic Chemistry, Vol. 44, No. 22, 2005