Role of Carboxylate Bridges in Modulating Nonheme Diiron(II)/O2 Reactivity

Article abstract of DOI:10.1021/ic034359a

A series of diiron(II) complexes of the dinucleating ligand HPTP (N, N, N′,N′-tetrakis(2-pyridylmethyl)-2-hydroxy-1,3-diaminopropane) with one or two supporting carboxylate bridges has been synthesized and characterized. The crystal structure of one member of each subset has been obtained to reveal for subset A a (μ-alkoxo)(μ-carboxylato)-diiron(II) center with one five- and one six-coordinate metal ion and for subset B a coordinatively saturated (μ-alkoxo)bis(μ-carboxylato)diiron(II) center. These complexes react with O₂ in second-order processes to form adducts characterized as (μ-1,2-peroxo)diiron(III) complexes. Stopped-flow kinetic studies show that the oxygenation step is sensitive to the availability of an O₂ binding site on the diiron(II) center, as subset B reacts more slowly by an order of magnitude. The lifetimes of the O₂ adducts are also distinct and can be modulated by the addition of oxygen donor ligands. The O₂ adduct of a monocarboxylate complex decays by a fast second-order process that must be monitored by stopped-flow methods, but becomes stabilized in CH₂Cl₂/DMSO (9:1 v/v) and decomposes by a much slower first-order process. The O₂ adduct of a dicarboxylate complex is even more stable in pure CH₂Cl₂ and decays by a first-order process. These differences in adduct stability are reflected in the observation that only the O₂ adducts of monocarboxylate complexes can oxidize substrates, and only those substrates that can bind to the diiron center. Thus, the much greater stability of the O₂ adducts of dicarboxylate complexes can be rationalized by the formation of a (μ-alkoxo)(μ-1,2-peroxo)diiron(III) complex wherein the carboxylate bridges in the diiron(II) complex become terminal ligands in the O₂ adduct, occupy the remaining coordination sites on the diiron center, and prevent binding of potential substrates. Implications for the oxidation mechanisms of nonheme diiron enzymes are discussed.

Full text of DOI:10.1021/ic034359a

Inorg. Chem. 2003, 42, 7519−7530
Role of Carboxylate Bridges in Modulating Nonheme Diiron(II)/O
2
Reactivity
Miquel Costas,† Clyde W. Cady,† Sergey V. Kryatov,‡ Manabendra Ray,† Meghan J. Ryan,†
,‡
,†
Elena V. Rybak-Akimova,* and Lawrence Que, Jr.*
Department of Chemistry and Center for Metals in Biocatalysis, UniVersity of Minnesota,
Minneapolis, Minnesota 55455, and Department of Chemistry, Tufts UniVersity,
Medford, Massachusetts 02155
Received April 2, 2003
A series of diiron(II) complexes of the dinucleating ligand HPTP (N,N,N′,N′-tetrakis(2-pyridylmethyl)-2-hydroxy-1,3-
diaminopropane) with one or two supporting carboxylate bridges has been synthesized and characterized. The
crystal structure of one member of each subset has been obtained to reveal for subset Aa(µ-alkoxo)(µ-carboxylato)-
diiron(II) center with one five- and one six-coordinate metal ion and for subset B a coordinatively saturated (µ-
alkoxo)bis(µ-carboxylato)diiron(II) center. These complexes react with O in second-order processes to form adducts
2
characterized as (µ-1,2-peroxo)diiron(III) complexes. Stopped-flow kinetic studies show that the oxygenation step
is sensitive to the availability of an O binding site on the diiron(II) center, as subset B reacts more slowly by an
2
order of magnitude. The lifetimes of the O adducts are also distinct and can be modulated by the addition of
2
oxygen donor ligands. The O adduct of a monocarboxylate complex decays by a fast second-order process that
2
must be monitored by stopped-flow methods, but becomes stabilized in CH Cl2/DMSO (9:1 v/v) and decomposes
2
by a much slower first-order process. The O adduct of a dicarboxylate complex is even more stable in pure
2
CH Cl2 and decays by a first-order process. These differences in adduct stability are reflected in the observation
2
that only the O adducts of monocarboxylate complexes can oxidize substrates, and only those substrates that can
2
bind to the diiron center. Thus, the much greater stability of the O adducts of dicarboxylate complexes can be
2
rationalized by the formation of a (µ-alkoxo)(µ-1,2-peroxo)diiron(III) complex wherein the carboxylate bridges in
the diiron(II) complex become terminal ligands in the O adduct, occupy the remaining coordination sites on the
2
diiron center, and prevent binding of potential substrates. Implications for the oxidation mechanisms of nonheme
diiron enzymes are discussed.
Introduction
center is proposed to bind oxygen to form a diiron(III)-per-
oxo adduct, which then undergoes O-O bond cleavage in a
subsequent step. Spectroscopically observable but distinct
high valent intermediates have been observed for RNR4-6
and MMO.7-10 The observation of such species in the enzyme
Investigations of dioxygen binding and activation at dinu-
clear iron complexes have been motivated by the presence
of carboxylate bridged dinuclear iron sites in dioxygen
activating enzymes such as ribonucleotide reductase (RNR),
methane monooxygenase (MMO), and fatty acid desatu-
rases.1-3 In the first step of oxygen activation, the diiron(II)
(4) Bollinger, J. M., Jr.; Edmondson, D. E.; Huynh, B. H.; Filley, J.;
Norton, J.; Stubbe, J. Science 1991, 253, 292-298.
(5) Sturgeon, B. E.; Burdi, D.; Chen, S.; Huynh, B.-H.; Edmondson, D.
E.; Stubbe, J.; Hoffman, B. M. J. Am. Chem. Soc. 1996, 118, 7551-
7557.
(6) Riggs-Gelasco, P. J.; Shu, L.; Chen, S.; Burdi, D.; Huynh, B. H.; Que,
L., Jr.; Stubbe, J. J. Am. Chem. Soc. 1998, 120, 849-860.
(7) Lee, S.-K.; Nesheim, J. C.; Lipscomb, J. D. J. Biol. Chem. 1993, 268,
21569-21577.
(8) Lee, S.-K.; Fox, B. G.; Froland, W. A.; Lipscomb, J. D.; Mu¨nck, E.
J. Am. Chem. Soc. 1993, 115, 6450-6451.
(9) 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.
* To whom correspondence should be addressed. E-mail: erybakak@
tufts.edu (E.V.R.-A.); que@chem.umn.edu (L.Q.).
† University of Minnesota.
‡ Tufts University.
(1) Wallar, B. J.; Lipscomb, J. D. Chem. ReV. 1996, 96, 2625-2658.
(2) 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.
(3) Merkx, M.; Kopp, D. A.; Sazinsky, M. H.; Blazyk, J. L.; Mu¨ller, J.;
Lippard, S. J. Angew. Chem., Int. Ed. 2001, 40, 2782-2807.
10.1021/ic034359a CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/18/2003
Inorganic Chemistry, Vol. 42, No. 23, 2003 7519

Costas et al.
Scheme 1
reactions has stimulated efforts to generate and characterize
analogous intermediates in model systems to serve as syn-
thetic inorganic precedents for the chemistry of these non-
heme diiron enzymes.11,12
The study of synthetic diiron(III)-peroxo intermediates13
has greatly enhanced our understanding of the corresponding
species in proteins. Three diiron(III)-peroxo complexes have
been crystallographically characterized,14-16 and the structure
of a fourth one has been deduced by EXAFS analysis.17
While detailed thermodynamic18 and kinetic19-23 studies on
dioxygen binding to some diiron(II) complexes have been
reported, the factors that affect subsequent decomposition
of these peroxo complexes and their oxidative reactivities
with substrates are much less well understood.24,25 In contrast,
detailed studies have been reported of related copper and
nickel complexes that generate peroxo intermediates that
effect intramolecular ligand oxidation26,27 and, in some cases,
intermolecular oxidation of added substrates.28-32
(O2CC6H5)]2+ reacted with PPh3 and 2,4-di-tert-butylphenol,
unlike its structurally characterized benzimidazole analogue
(N-Et-HPTB, Scheme 1),15 which was unreactive toward
these substrates. Thus a change from pendant benzimidazoles
to pyridines elicits a dramatic change in reactivity. In this
paper, we have extended our initial studies to a series of
Previously, we reported the generation and characterization
of a (µ-1,2-peroxo)diiron(III) complex from the diiron(II)
precursor [Fe2(HPTP)(O2CC6H5)]X233 (see Scheme 1).24 Our
initial studies showed that the O2 adduct of [Fe2(HPTP)-
[FeII (HPTP)] complexes bridged by one or two carboxylates.
2
The crystal structures of two of these precursor complexes
are reported, as well as kinetic studies for the formation and
decay of their O2 adducts. The reactivities of the O2 adducts
depend on the number of carboxylate bridges present. These
differences provide insight into some of the factors that
modulate the stability and oxidative power of the (µ-1,2-
peroxo)diiron(III) unit in chemistry and biology.
(10) Shu, L.; Nesheim, J. C.; Kauffmann, K.; Mu¨nck, E.; Lipscomb, J. D.;
Que, L., Jr. Science 1997, 275, 515-518.
(11) Que, L., Jr. J. Chem. Soc., Dalton Trans. 1997, 3933-3940.
(12) Du Bois, J.; Mizoguchi, T. J.; Lippard, S. J. Coord. Chem. ReV. 2000,
200-202, 443-485.
(13) Girerd, J.-J.; Banse, F.; Simaan, A. J. Struct. Bonding 2000, 97, 143-
177.
(14) Ookubo, T.; Sugimoto, H.; Nagayama, T.; Masuda, H.; Sato, T.;
Tanaka, K.; Maeda, Y.; Okawa, H.; Hayashi, Y.; Uehara, A.; Suzuki,
M. J. Am. Chem. Soc. 1996, 118, 701-702.
Experimental Section
Synthesis. The dinucleating ligand H-HPTP was synthesized
following the published procedure.34 The diiron(II) complexes 1-5
(15) Dong, Y.; Yan, S.; Young, V. G., Jr.; Que, L., Jr. Angew. Chem., Int.
Ed. Engl. 1996, 35, 618-620.
were prepared following the published procedure for [FeII (HPTP)-
(16) Kim, K.; Lippard, S. J. J. Am. Chem. Soc. 1996, 118, 4914-4915.
(17) Dong, Y.; Zang, Y.; Shu, L.; Wilkinson, E. C.; Que, L., Jr.; Kauffmann,
K.; Mu¨nck, E. J. Am. Chem. Soc. 1997, 119, 12683-12684.
(18) Sugimoto, H.; Nagayama, T.; Maruyama, S.; Fujinami, S.; Yasuda,
Y.; Suzuki, M.; Uehara, A. Bull. Chem. Soc. Jpn. 1998, 71, 2267-
2279.
(19) (a) Feig, A. L.; Becker, M.; Schindler, S.; van Eldik, R.; Lippard, S.
J. Inorg. Chem. 1996, 35, 2590-2601. (b) Feig, A. L.; Becker, M.;
Schindler, S.; van Eldik, R.; Lippard, S. J. Inorg. Chem. 2003, 42,
3704
2
24
(O2CC6H5)](BPh4)2 by combining 2 equiv of Fe(O3SCF3)2‚
2CH3CN, 1 equiv of H-HPTP‚4HClO4, 1 equiv of ArCOOH, and
6 equiv of Et3N in methanol, stirring for 30 min, and then precip-
itating the complexes with the addition of 2 equiv of NaBPh4.
Recrystallization from CH3CN/ether afforded analytically pure
complexes. Tetramethylammonium salts of benzoic acids were
prepared by neutralizing the acid using (Me4N)OH in MeOH/H2O
and subsequently removing the solvent under reduced pressure. The
(20) Feig, A. L.; Masschelein, Z.; Bakac, A.; Lippard, S. J. J. Am. Chem.
Soc. 1997, 119, 334-342.
1
purity of the salts was checked using H NMR.
(21) LeCloux, D. D.; Barrios, A. M.; Mizoguchi, T. J.; Lippard, S. J. J.
Am. Chem. Soc. 1998, 120, 9001-9014.
[FeII2(HPTP)(O2CC6H5)](BPh4)2. Anal. for (1)(BPh4)2 or
C82H74B2Fe2N6O3, Calcd (Found): C, 74.34 (74.35); H, 5.63 (5.73);
N, 6.34 (6.29). λmax (ꢀ, M-1 cm-1) in MeCN: 326 (2200), 400 sh.
[FeII2(HPTP)(O2CC6H4-p-Cl)](BPh4)2. Anal. for (2)(BPh4)2
orC82H73B2ClFe2N6O3, Calcd (Found): C, 72.46 (72.45); H, 5.41
(5.48); N, 6.18 (6.31). λmax (ꢀ, M-1 cm-1) in MeCN: 324 (2400),
375 sh.
(22) Kryatov, S. V.; Rybak-Akimova, E. V.; MacMurdo, V. L.; Que, L.,
Jr. Inorg. Chem. 2001, 40, 2220-2228.
(23) Chavez, F. A.; Ho, R. Y. N.; Pink, M.; Young, V. G., Jr.; Kryatov, S.
V.; Rybak-Akimova, E. V.; Andres, H. P.; Mu¨nck, E.; Que, L., Jr.;
Tolman, W. B. Angew. Chem., Int. Ed. 2002, 149-152.
(24) Dong, Y.; Me´nage, S.; Brennan, B. A.; Elgren, T. E.; Jang, H. G.;
Pearce, L. L.; Que, L., Jr. J. Am. Chem. Soc. 1993, 115, 1851-1859.
(25) LeCloux, D. D.; Barrios, A. M.; Lippard, S. J. Bioorg. Med. Chem.
1999, 7, 763-772.
(26) Tolman, W. B. Acc. Chem. Res. 1997, 30, 227-237.
(27) Que, L., Jr.; Tolman, W. B. Angew. Chem., Int. Ed. 2002, 41, 1114-
1137.
(33) Abbreviations used: BIPhMe ) 2,2′-bis(1-methylimidazolyl)phenyl-
methoxymethane; BPMP
) 2,6-bis[[bis(2-pyridylmethyl)amino]-
methyl]-4-methylphenolate; ∆9D ) stearoyl acyl carrier protein ∆9-
desaturase; HPTP ) anion of N,N,N′,N′-tetrakis(2-pyridylmethyl)-2-
hydroxy-1,3-diaminopropane; Me3TACN ) 1,4,7-trimethyl-1,4,7-tri-
azacyclononane; 6-Me4-HPTP ) anion of N,N,N′,N′-tetrakis(6-methyl-
2-pyridylmethyl)-2-hydroxy-1,3-diaminopropane; MMO ) methane
monooxygenase, N-Et-HPTB ) anion of N,N,N′,N′-tetrakis[(N-ethyl-
2-benzimidazolyl)methyl]-2-hydroxy-1,3-diaminopropane; RNR, ri-
bonucleotide reductase; TpiPr2 ) hydrotris(3,5-diisopropylpyrazolyl)-
borate anion; TPPDO ) anion of N,N,N′,N′-tetrakis(6-pivalamido-2-
pyridylmethyl)-2-hydroxy-1,3-diaminopropane.
(28) Mahadevan, V.; DuBois, J. L.; Hedman, B.; Hodgson, K. O.; Stack,
T. D. P. J. Am. Chem. Soc. 1999, 121, 5583-5584.
(29) Mahadevan, V.; Henson, M. J.; Solomon, E. I.; Stack, T. D. P. J. Am.
Chem. Soc. 2000, 122, 10249-10250.
(30) Taki, M.; Itoh, S.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 6203-
6204.
(31) Taki, M.; Itoh, S.; Fukuzumi, S. J. Am. Chem. Soc. 2002, 124, 998-
1002.
(32) Zhang, C. X.; Liang, H.-C.; Kim, E.-i.; Shearer, J.; Helton, M. E.;
Kim, E.; Kaderli, S.; Incarvito, C. D.; Zuberbuhler, A. D.; Rheingold,
A. L.; Karlin, K. D. J. Am. Chem. Soc. 2003, 125, 634-635.
(34) McKee, V.; Zvagulis, M.; Dagdigian, J. V.; Patch, M. G.; Reed, C.
A. J. Am. Chem. Soc. 1984, 106, 4765-4772.
7520 Inorganic Chemistry, Vol. 42, No. 23, 2003

Carboxylates in Nonheme Diiron(II)/O2 ReactiWity
[FeII2(HPTP)(O2CC6F5)](BPh4)2. Anal. for (3)(BPh4)2 or
C82H69B2F5Fe2N6O3, Calcd (Found): C, 69.61 (69.53); H, 4.91
(4.91); N, 5.94 (5.97). λmax (ꢀ, M-1 cm-1) in MeCN: 323 (1700),
375 sh.
[FeII2(HPTP)(O2CC6H4-p-OMe)](BPh4)2. Anal. for (4)(BPh4)2‚
3H2O or C83H83B2Fe2N6O7, Calcd (Found): C, 70.71 (70.45); H,
5.96 (5.79); N, 5.93 (6.02). λmax (ꢀ, M-1 cm-1) in MeCN: 338
(2300), 450 sh.
[FeII2(HPTP)(O2CC6H4-p-Me)](BPh4)2. Anal. for (5)(BPh4)2‚
2H2O or C83H80B2Fe2N6O5, Calcd (Found): C, 72.51 (72.35); H,
5.86 (5.62); N, 6.11 (6.24). λmax (ꢀ, M-1 cm-1) in MeCN: 324
(2500), 375 sh.
[FeII2(HPTP)(O2CC6H5)2](BPh4), (6)(BPh4). This complex was
synthesized by the addition of a solution of Me4N(O2CC6H5) (0.024
g, 0.123 mM) in MeCN to a stirred solution of 1 (0.161 g, 0.121
mM) in MeCN. Immediately, the light yellow solution of 1 changed
to red, and 6 was precipitated as a red crystalline powder within
10 min. The precipitate was filtered and dried under vacuo (0.118
g, 96%). The complex was recrystallized from CH2Cl2 and diethyl
ether as red needles. Anal. for (6)(BPh4)2‚MeCN‚0.5H2O or C67H63-
BFe2N7O5.5, Calcd (Found): C, 68.38 (68.31); H, 5.40 (5.42); N,
8.33 (8.56). λmax (ꢀ, M-1 cm-1) in CH2Cl2 at 298 K: 367 (2600),
480 sh.
[FeII2(HPTP)(O2CC6H4-p-Cl)2](BPh4). Anal. for (7)(BPh4)‚
MeCN‚H2O or C67H62BCl2Fe2N7O6, Calcd (Found): C, 64.14
(64.22); H, 4.98 (4.79); N, 7.81 (7.54); Cl, 5.64 (5.94). λmax (ꢀ,
M-1 cm-1) in CH2Cl2 at 233 K: 384 (2000).
[FeII2(HPTP)(O2CC6H4-p-C6H5)2](BPh4). Anal. for (8)(BPh4)‚
MeCN‚H2O or C79H72BFe2N7O6, Calcd (Found): C, 70.92 (70.55);
H, 5.42 (5.34); N, 7.33 (7.15). λmax (ꢀ, M-1 cm-1) in CH2Cl2 at
233 K: 364 (2300).
holographic super-notch filter with a Princeton Instruments liquid
N2-cooled (LN-1100PB) CCD detector with 4 cm-1 spectral
resolution.
Generation of Peroxo Adducts and Conventional Kinetic
Experiments. The peroxo adduct intermediates were generated in
an immersion dewar maintained at the appropriate low temperature
by bubbling pure oxygen at atmospheric pressure through a
CH2Cl2/DMSO-d6 (9:1 v/v) (1-5) or CH2Cl2 (6-10) solution
(approximately 2 mL) of the diiron(II) precursor for 30 s to produce
a blue-violet color within seconds. The decomposition of the peroxo
intermediates was monitored for at least 5 half-lives, and data were
analyzed using the program Specfit (BioLogic Science Instruments,
Grenoble, France). As decay started within a few seconds in the
presence of substrate, no attempts were made to remove excess
O2. The iron(II) solutions before and after addition of substrate were
1
checked using H and 31P NMR and UV-vis spectroscopy, and
no evidence was found for substrate binding to the diiron(II)
precursor.
Stopped-Flow Studies. All manipulations of 1 and 6 and their
solutions were done inside a Vacuum Atmospheres glovebox with
argon or using Hamilton airtight syringes. Graduated mixtures of
oxygen and nitrogen were prepared using Bel-Art gas flow meters.
Saturated solutions of O2 and graduated O2-N2 mixtures were
prepared by bubbling the gas through a liquid for 20 min in a
cylinder closed with a septum and thermostated at 20 or 25 °C.
The solubility of O2 was accepted to be 5.8 mM in dichloromethane
at 20 °C and 8.1 mM in acetonitile at 25 °C.35,36 Oxygen solubility
in dichloromethane-acetonitrile mixtures was estimated by linear
interpolation. Kinetic measurements were performed using a Hi-
Tech Scientific (Salisbury, Wiltshire, U.K.) SF-43 multimixing
anaerobic cryogenic stopped-flow instrument combined with (a) a
Hi-Tech Scientific M300 monochromator or (b) a Hi-Tech Scientific
Kinetascan diode array rapid scanning unit. No difference in spectral
and kinetic data was observed for the monochromator and diode
array modes in this study. The mixing cell (1 cm) was maintained
to (0.1 K, and mixing time was 2-3 ms. In all kinetic experiments,
a series of 5-7 shots gave standard deviations within 5%, with
overall reproducibility within 10%. Data analysis was performed
with IS-2 Rapid Kinetics Software (Hi-Tech Scientific) for single-
wavelength kinetic traces and with the program Specfit (BioLogic
Science Instruments, Grenoble, France) for global fitting of the
spectral changes acquired in a diode array mode. Solvent purifica-
tion proved to be essential for obtaining reproducible data (par-
ticularly with complex 1). Acetonitrile was distilled over CaH2
followed by another distillation over P2O5. Even though the
oxygenation reactions were not affected by small quantities of water,
this treatment was necessary to remove other impurities. Using
anhydrous acetonitrile (Aldrich) or passing the solvent through a
column with activated alumina proved insufficient. Dichloromethane
(Fisher, >99.9%, GS/MS grade, not stabilized) was additionally
treated with basic alumina.
[FeII2(HPTP)(O2CC6H4-p-Me)2](BPh4). Anal. for (9)(BPh4) or
C67H63BFe2N6O5, Calcd. (Found): C, 69.69 (69.43); H, 5.50 (5.51);
N, 7.28 (7.40). λmax (ꢀ, M-1 cm-1) in CH2Cl2 at 233 K: 360 (2200),
460 sh.
[FeII2(HPTP)(O2CC6H4-p-OMe)2](BPh4). Anal. for (10)(BPh4)‚
2MeCN‚0.5H2O or C71H70BFe2N8O7.5, Calcd. (Found): C, 66.73
(66.81); H, 5.52 (5.37); N, 8.77 (8.51). λmax (ꢀ, M-1 cm-1) in
CH2Cl2 at 233 K: 336 (2500), 464 (1600).
Physical Methods. UV-vis spectra were recorded on a HP
8452A diode array spectrometer. Low-temperature visible spectra
were obtained using an immersion dewar equipped with quartz
windows. GC analyses were performed with a Perkin-Elmer
Autosystem equipped with an Alltech 1701 30 m × 0.25 mm ×
0.25 µm capillary column. GC-MS analysis were carried out in a
HP-5898 GC equipped with a DB-5 capillary column and a
Finnigan MAT 95 mass detector, using NH3 as ionization gas. NMR
spectra were recorded on a Varian VXR 300 spectrophotometer.
1
The H NMR spectra were recorded using a 90° pulse with 16K
data points. An inversion-recovery pulse sequence (180°-τ-90°-
AQ) was used to obtain nonselective proton relaxation times (T1)
with carrier frequencies set at several different positions to ensure
the validity of the measurements.
Oxidation of Substrates. A 3 mM solution of the diiron(II)
complex in CH2Cl2/DMSO-d6 (9:1 v/v) (1) or CH2Cl2 (6) was trans-
ferred, in a drybox, to a 5 mL Schlenk flask which was subsequently
sealed with a septum. The Schlenk flask was placed in an acetone
bath cooled to -80 °C, and dry O2 was bubbled through the solution
for 10 s. After the appropriate amount of substrate was added via
syringe, the complete reaction mixture was purged by three vac-
uum-argon cycles and maintained at -40 °C with stirring. After
2 h at -40 °C, the acetone bath was removed, and the reaction
Infrared spectra were obtained from KBr pellets on a Nicolet
Avatar 320 FT-IR instrument. The KBr pellet of 6‚O2 was made
within a 3-min period at room temperature after its isolation as a
solid sample at -40 °C to minimize its decomposition. Resonance
Raman spectra were collected using a backscattering geometry on
sample solutions frozen on a coldfinger cooled to 77 K with 568-
nm laser excitation (30 mW at the sample) from a Spectra Physics
2030-15 argon ion laser and a 375B CW dye (Rhodamine 6G) laser.
Spectra were calibrated with indene and recorded with an Acton
AM-506 spectrometer (2400-groove grating) using a Kaiser Optical
(35) Battino, R. In Solubility Data Series Vol. 7. Oxygen and Ozone;
Battino, R., Ed.; Pergamon: New York, 1981.
(36) Achord, J. M.; Hussey, C. L. Anal. Chem. 1980, 52, 601-602.
Inorganic Chemistry, Vol. 42, No. 23, 2003 7521

Costas et al.
Table 1. Crystallographic Data for (3‚CH3CN)(BPh4)2‚CH3CN and
(6)(BF4)‚CH3CN
Table 2. Selected Bond Lengths (Å) and Bond Angles (deg) for
(3‚CH3CN)(BPh4)2‚CH3CN and (6)(BF4)‚CH3CN
(3‚CH3CN)(BPh4)2‚CH3CN
(6)(BF4)‚CH3CN
(3‚CH3CN)(BPh4)2‚CH3CN (6)(BF4)‚CH3CN
formula
fw
cryst color
T
cryst syst
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V
C86H75B2F5Fe2N8O3
1496.86
yellow
173(2) K
triclinic
C67H62BFe2N7O5
1167.75
reddish brown
173(2) K
triclinic
Fe1-O1
1.9390(17)
2.0636(18)
2.2410(19)
2.101(2)
Fe1-O1
Fe1-O2
Fe1-O4
Fe1-N2
Fe1-N1
Fe1-N5
Fe2-O1
Fe2-O3
Fe2-O5
Fe2-N3
Fe2-N4
Fe2-N6
Fe1‚‚‚Fe2
Fe1-O1-Fe2
1.9906(14)
2.131(2)
2.052(2)
2.278(2)
2.166(2)
2.185(2)
1.9972(14)
2.197(2)
2.0585(14)
2.256(2)
2.159(2)
2.247(2)
3.30
Fe1-O2
Fe1-N4
Fe1-N5
Fe1-N6
Fe2-O1
Fe2-O3
Fe2-N1
Fe2-N2
Fe2-N3
Fe2-N7
2.096(2)
P1h
P1h
2.0061(17)
2.0917(18)
2.254(2)
2.150(2)
2.175(2)
13.674(5)
14.212(6)
22.138(9)
102.342(7)
96.252(7)
90.270(7)
4176(3) Å3
2
9.9497(2)
16.9298(2)
17.5250(2)
80.050(1)
89.116(1)
86.714(1)
2902.80(8) Å3
2
2.218(2)
Fe1‚‚‚Fe2
Fe1-O1-Fe2
3.47
123.18(8)
Z
1.190 cm-3
4.09 cm-1
0.0465
0.1065
1.022
1.336 g cm-3
5.58 cm-1
0.0370
0.0777
1.038
111.74(6)
D (calcd)
µ
R1a
For (6)(BF4)•CH3CN, the space group P1h was determined on
the basis of systematic absences and intensity statistics. A Patterson
solution was calculated, which provided the positions of the heavy
atoms. Several full-matrix least-squares/difference Fourier cycles
were performed to locate the remaining non-hydrogen atoms. All
non-hydrogen atoms were refined with anisotropic displacement.
All hydrogen atoms were found from the difference map and refined
for all parameters. The structure was found with one solvent
molecule, acetonitrile, per asymmetric unit. Selected bond lengths
and angles are listed in Table 2.
wR2a
GOF on F2
a R1 ) ∑||Fo| - |Fc||/∑|Fo| and wR2 ) (∑[w(Fo2 - Fc )2]/∑[wFo ])1/2
,
2
4
2
where w ) q/σ2(Fo ) + (aP)2 + bP.
mixture was allowed to warm to room temperature. Naphthalene
was added as an internal standard, and the products were analyzed
by GC. Oxidation products were identified by GC-MS analysis and
by comparing their retention time with authentic samples.
Crystallographic Experiments. Data collection and structure
solution were conducted at the X-ray Crystallographic Laboratory
of the University of Minnesota Chemistry Department. All calcula-
tions were performed using SGI INDY R4400-SC or Pentium
computers using the SHELXTL V5.0 and V5.10 suite of programs.
Crystals of (3‚CH3CN)(BPh4)2‚CH3CN and (6)(BF4)‚CH3CN were
placed onto the tips of glass capillaries and mounted on a Bruker
SMART system for data collection at 173(2) K. Preliminary sets
of cell constants were calculated from reflections harvested from
three sets of 20 and 60 frames, respectively. These initial sets of
frames were oriented such that orthogonal wedges of reciprocal
space were surveyed. Initial orientation matrices were determined
from 53 and 570 reflections, respectively. Final cell constants were
calculated from a set of 1024 and 14878 reflections, respectively.
Additional crystallographic data and refinement information are
summarized in Table 1.
For (3‚CH3CN)(BPh4)2‚CH3CN, the space group P1h was deter-
mined on the basis of systematic absences and intensity statistics.
A direct-methods solution was calculated which afforded positions
of most non-hydrogen atoms from the E-map. Full-matrix least-
squares/difference Fourier cycles were performed which located
the remaining non-hydrogen atoms. All non-hydrogen atoms were
refined with anisotropic displacement parameters. All hydrogen
atoms were placed in ideal positions and refined as riding atoms
with individual (or group if appropriate) isotropic displacement
parameters.
Results and Discussion
The monocarboxylate-bridged diiron(II) complexes
[FeII (HPTP)(O2CAr)](BPh4)2 (1-5, Scheme 1) were syn-
2
thesized following the method of Dong et al.,24 while their
bis(carboxylate) analogues [FeII (HPTP)(O2CAr)2](BPh4)
2
(6-10, Scheme 1) were obtained by reaction of the cor-
responding monocarboxylate complex with 1 equiv of
(Me4N)(O2CAr) in acetonitrile. Yields were somewhat de-
pendent on the nature of the carboxylate, ranging from 34%
for Ar ) p-OMeC6H4 to practically quantitative for Ar )
C6H5. These complexes serve as precursors to metastable
dioxygen adducts that are the focus of this investigation.
Crystal Structures of (3‚CH3CN)(BPh4)2 and 6(BF4).
ORTEP plots based on the solid state structures of the cations
of (3‚CH3CN)(BPh4)2 and 6(BF4) are shown in Figure 1. For
both complexes, the diiron(II) center is bridged by the alkoxo
oxygen atom, and each iron is facially capped by the
tridentate bis(2-pyridylmethyl)amine moieties on either end
of the 2-hydroxypropane backbone of the HPTP ligand. For
3, there is one bidentate carboxylate bridge, which coordi-
nates trans to the amine nitrogen of each iron (Figure 1a).
Interestingly, the two iron atoms have different coordination
geometries. Fe1 is trigonal bipyramidal (τ ) 0.86), with the
amine nitrogen and the carboxylate oxygen defining the
trigonal axis. Fe2 is six-coordinate but has a similar ligand
arrangement as Fe1 except that the trigonal plane has
expanded to accommodate an acetonitrile ligand trans to one
of the pyridines. Complex cation 6 has two carboxylate
bridges, and each iron atom is six-coordinate (Figure 1b).
Although there is no crystallographically imposed symmetry,
the molecule has approximate Cs symmetry with a mirror
plane passing perpendicular to the Fe-Fe vector. One
benzoate coordinates trans to the amine nitrogen atoms of
Because an unknown amount of disordered solvent was present
in the structure, the data set was modified using the program
PLATON/SQUEEZE.37 An investigation of the void space in the
structure showed that 116 electrons are located in one major void
of approximately 635.1 Å3. The refinement using the corrected data
set improved the overall structure and the R1 value. Some of the
given values in the CIF file are known to be incorrect due to the
unknown amount of solvent (e.g., F(000), formula, formula weight).
Selected bond lengths and angles are listed in Table 2.
(37) Spek, A. L. PLATON. A multipurpose crystallographic tool; Utrecht
University: Utrecht, The Netherlands, 2002.
7522 Inorganic Chemistry, Vol. 42, No. 23, 2003

Carboxylates in Nonheme Diiron(II)/O2 ReactiWity
transfer transitions. Likewise, the UV-vis spectra of the
dicarboxylate complexes 6-10 in CH2Cl2 exhibit an absorp-
tion at 335-385 nm (ꢀ ) 2000-2800 M-1 cm-1); the red
shift of the MLCT transition is attributed to the presence of
the second carboxylate, which diminishes the Lewis acidity
of the metal center (Figure 2).
1
Figure 3 shows the H NMR spectra of 1 and 6, which
are representative of mono- and dicarboxylate complexes,
1
respectively. The H NMR spectra of the complexes show
well-resolved, relatively sharp signals that span over 220 ppm
in chemical shift. The high resolution and the relatively
narrow widths of the signals observed in the spectra are
typical of high spin Fe(II) complexes and are similar to those
of previously described structurally related diiron(II) com-
plexes.24,43 The chemical shifts for all 10 complexes are listed
in Table S1; most of these can be assigned by their T1 values
and a comparison of the spectra of related complexes. The
geminal proton to the alkoxide bridge is readily assigned on
the basis of its single-proton integration, short T1, and large
shift due to its proximity to the metal centers. Most, if not
all, of the methylene protons appear as well-resolved signals
that integrate as two protons each. Since the CH2 groups
are in five-membered metal-chelate rings, the protons
comprising each geminal pair can be designated as axial or
equatorial and are affected differently by the paramagnetic
centers.43 Thus, the appearance of six distinct resonances
demonstrates that the two halves of the dinucleating ligand
are magnetically equivalent on the NMR time scale. This
notion is also supported by the number of pyridyl resonances
observed. A comparison of the spectra of 1, 2, and 3 allows
the assignment of the aromatic proton signals arising from
the carboxylate. Finally, the signals of the different protons
in the pyridine ring can be identified on the basis of the
dramatic difference in their respective relaxation times.
Generation and Characterization of the Dioxygen
Adducts. The monocarboxylate bridged diiron(II) complexes
1-5 react with O2 in CH2Cl2/DMSO (9:1 v/v) to generate
deep purple species, which are stable only for minutes at
low temperature. These O2 adducts 1‚O2-5‚O2 are best
described as diiron(III)-peroxo complexes on the basis of
their spectroscopic properties and comparisons with related
complexes. The O2 adducts exhibit one intense and broad
visible absorption band with λmax ) 573-580 nm (ꢀ )
2100-2800 M-1 cm-1) (Table S2). As for previously
reported diiron(III) peroxo complexes,24 this chromophore
Figure 1. ORTEP plots of (A) cation of 3, [FeII2(HPTP)(O2CC6F5)-
(MeCN)]2+, and (B) the cation of 6, [FeII2(HPTP)(O2CC6H5)2]+, showing
50% probability ellipsoids. Hydrogen and fluorine atoms have been omitted
for clarity.
HPTP, as in 3, while the second benzoate binds trans to two
of the pyridine rings.
Metal-ligand bond lengths for 3 and 6 listed in Table 2
are characteristic of high spin FeII complexes. For 3, the
metal-ligand distances for 5-coordinate Fe1 are not surpris-
ingly shorter than those for 6-coordinate Fe2. For example,
the average Fe-Npy distance is 2.10 Å for Fe1 and 2.16 Å
for Fe2. For 6, both iron ions are 6-coordinate, and the
average Fe-Npy distance is 2.19 Å, slightly longer than that
for the 6-coordinate Fe2 of 3. More interesting are the two
carboxylate bridges, which are distinct from each other. The
carboxylate bridge trans to the amine nitrogens has an
average Fe-O bond (Fe1-O2 and Fe2-O3) of 2.06 Å,
comparable to that for 3 (2.08 Å), but the other carboxylate
has much longer Fe-O bonds (Fe1-O4 and Fe2-O5),
averaging 2.16 Å, suggesting that this carboxylate is more
weakly bound. Conversion of 3 to 6 by the introduction of
a second carboxylate group results in the closing of the
Fe-O-Fe angle from 123° to 112° and a shortening of the
Fe-Fe distance from 3.47 to 3.30 Å. As shown in Table 3,
these changes are mirrored in the structural differences
between other complexes with [Fe2(µ-OR)] cores having one
or two carboxylate bridges.
(38) (a) Hayashi, Y.; Kayatani, T.; Sugimoto, H.; Suzuki, M.; Inomata,
K.; Uehara, A.; Mizutani, Y.; Kitagawa, T.; Maeda, Y. J. Am. Chem.
Soc. 1995, 117, 11220-11229. (b) Arii, H.; Nagatomo, S.; Kitagawa,
T.; Miwa, T.; Jitsukawa, K.; Einaga, H.; Masuda, H. J. Inorg. Biochem.
2000, 82, 153-162.
(39) Kitajima, N.; Tamura, N.; Tanaka, M.; Moro-oka, Y. Inorg. Chem.
1992, 31, 3342-3343.
(40) Hartman, J. R.; Rardin, R. L.; Chaudhuri, P.; Pohl, K.; Wieghardt,
K.; Nuber, B.; Weiss, J.; Papaefthymiou, G. C.; Frankel, R. B.;
Lippard, S. J. J. Am. Chem. Soc. 1987, 109, 7387-7396.
(41) Borovik, A. S.; Hendrich, M. P.; Holman, T. R.; Que, L., Jr.;
Papaefthymiou, V.; Mu¨nck, E. J. Am. Chem. Soc. 1990, 112, 6031-
6038.
Physical Properties of Diiron(II) Complexes. Complexes
1-10 are air-sensitive solids. The UV-vis spectra of
monocarboxylate complexes 1-5 in acetonitrile show no
significant absorption in the visible region and intense
features in the UV region at λmax ) 323-338 nm (ꢀ ) 1700-
2500 M-1 cm-1) arising from iron(II)-to-pyridine charge
(42) Rardin, R. L.; Tolman, W. B.; Lippard, S. J. New J. Chem. 1991, 15,
417-430.
(43) Ming, L.-J.; Jang, H. G.; Que, L., Jr. Inorg. Chem. 1992, 31, 359-
364.
Inorganic Chemistry, Vol. 42, No. 23, 2003 7523

Costas et al.
Table 3. Structural Parameters of Complexes with [FeII2(L)(µ-OR)(O2CR′)x] Cores
diiron(II) complexa
R
x
Fe-µ-OR (av), Å
Fe-O-Fe, deg
Fe‚‚‚Fe, Å
ref
3
6
alkyl
alkyl
alkyl
alkyl
alkyl
H
H
aryl
formyl
1
2
1
1
1
1
2
2
2
1.97
1.9939(10)
1.97
2.02
2.07
123.18(8)
111.74(6)
124.0(3)
131.2(1)
3.47
3.30
b
b
[FeII2(N-Et-HPTB)(OBz)]2+
[FeII2(6-Me4-HPTP)(OBz)]2+
[FeII2(TPPDO)(O2CC6H4-4-Cl)]2+
[FeII2(OH)(TpiPr2)2(OBz)]
3.473(7)
3.684(1)
3.732(1)
3.681(3)
3.32(1)
3.348(2)
3.573(8)
24
38a
38b
39
40
41
42
1.98
137.0(3)
113.2(2)
108.93(6)
113.0(1)
[FeII2(Me3TACN)2(OH)(OAc)2]+
[FeII2(BPMP)(OPr)2]+
1.987(8)
2.057(1)
2.142(1)
[FeII2(O2CH)4(BIPhMe)2]
a For abbreviations used, see footnote 33. b This work.
Scheme 2
Figure 2. UV-vis spectra, at -60 °C, of 1 (- - -, left) and 1‚O2 (- - -,
right) in CH2Cl2/DMSO (9:1 v/v) and of 6 (s, left) and 6‚O2 (s, right) in
CH2Cl2.
1‚O2 adduct has been observed in the absence of DMSO
under stopped-flow conditions19 and presumably has the
structure depicted in Scheme 2A. So the presence of DMSO
must stabilize the peroxo complex, allowing its concentration
to build up for characterization without the need for fast
detection techniques. The addition of OPPh3 has a similar
effect on the stability of the O2 adduct. Indeed, the addition
of OPPh3 to the O2 adduct of the corresponding N-Et-HPTB
complex stabilized it sufficiently to allow diffraction quality
crystals to be obtained and resulted in its crystallographic
characterization.15 That structure revealed a (µ-alkoxo)-
(µ-1,2-peroxo)diiron(III) complex in which the carboxylate
bridge had been displaced by two molecules of OPPh3
(Scheme 2B, X ) Y ) Ph3PO). A similar scenario may be
considered for the HPTP complexes in this study (Scheme
2B, X ) Y ) DMSO), but the observation that the λmax’s of
the dioxygen adducts in the presence of DMSO remain
sensitive to the basicity of the carboxylate group (Table S2)
suggests that the latter is still coordinated to the iron after
the oxygen binding step. Thus, we propose that only one
DMSO molecule coordinates to the adduct, converting the
carboxylate bridge into a terminal ligand on one iron (Scheme
2B, X ) OOCR, Y ) DMSO).
Figure 3. 1H NMR spectra of 1 in CD3CN and 6 in CD2Cl2 at room
temperature.
In contrast, O2 adducts of 6-10 in CH2Cl2 (λmax
)
572-575 nm; ꢀ ) 3000-3500 M-1 cm-1) (Table S2) remain
indefinitely stable at -60 °C. The second carboxylate group,
like DMSO, stabilizes the adduct, suggesting that both
molecules may act in a similar mode. Raman spectra of 6‚
O2, obtained with 568.2 nm laser excitation, show resonance
enhanced features at 456, 482, 888, and 902 cm-1, which
are quite similar to those observed for 1‚O2. This similarity
in ν(Fe-O) and ν(O-O) values for 1‚O2 and 6‚O2 suggests
that the peroxide is coordinated in a similar fashion in both
complexes and that the Fe-O and O-O bond strengths and
the geometry of the Fe-O-O-Fe unit are not significantly
is assigned to a peroxo-to-Fe(III) charge transfer transition.
Consistent with this assignment, resonance Raman spectra
of 1‚O2 with 568 nm laser excitation shows a Fermi doublet
at 453 and 481 cm-1 assigned to ν(Fe-O), and another Fermi
doublet at 877 and 893 cm-1 assigned to ν(O-O).
The formation of 1‚O2-5‚O2 cannot be detected by
conventional spectrophotometry in the absence of DMSO38,44
under our low temperature conditions.24 However, a transient
(44) Hayashi, Y.; Suzuki, M.; Uehara, A.; Mizutani, Y.; Kitagawa, T. Chem.
Lett. 1992, 91-94.
7524 Inorganic Chemistry, Vol. 42, No. 23, 2003

Carboxylates in Nonheme Diiron(II)/O2 ReactiWity
isotopomer. Compared to its precursor, the O2 adduct exhibits
νasym features upshifted by 10-15 cm-1 and νsym features
downshifted by about 50 cm-1, leading to a dramatic increase
in the ∆ν values (284 and 269 cm-1). These results
unequivocally demonstrate that the carboxylates of 6 have
become monodentate upon formation of the O2 adduct
-
(Scheme 2B, X ) Y ) OOCR).
Oxygenation Kinetics. Choice of Solvents and Addi-
tives. A detailed kinetic analysis has been performed on the
oxygenation of 1 and 6. Oxygenation kinetics data were
previously reported in propionitrile for 1 and related mono-
carboxylate-bridged complexes.19 Although no substantial
difference between oxygenation rates in propionitrile and
acetonitrile was expected, control experiments on oxygen
binding to 1 in acetonitrile were performed. Complex 6 was
insoluble in polar solvents but soluble in chlorinated
hydrocarbons, making noncoordinating CH2Cl2 the solvent
of choice. For direct comparisons of 1 and 6, oxygenation
in CH3CN/CH2Cl2 mixtures was also investigated. Peroxo
intermediates derived from 1 were greatly stabilized by the
addition of oxygen donors (DMSO or Ph3PO), which coor-
dinate to the metal centers in both the diiron(II) complexes
and corresponding diiron(III) peroxo adducts.15,24 These
additives however could retard the oxygenation rates by
blocking vacant iron(II) coordination sites in 1. Direct mea-
surements were needed in order to determine the influence
of oxygen donors on the oxygenation rates.
Kinetic measurements were performed under pseudo-first-
order conditions (large excess of dioxygen with respect to
diiron complex) over the temperature range from -40 to 0
°C in CH3CN, and from -80 to 0 °C for CH2Cl2. As ex-
pected, the oxygenation of 1 in CH3CN or CH3CN/CH2Cl2
mixtures proceeded similarly to oxygenation of 1 in pro-
pionitrile,19 with rapid accumulation of the peroxo species
(over ca. 200 ms at -40 °C, CH3CN) immediately followed
by its decomposition, but reproducible kinetic parameters
could be obtained only in doubly distilled (over CaH2 and
over P2O5) CH3CN. The oxygenation of 1 in the presence
of oxygen donors (DMSO and Ph3PO) and of 6 proceeded
as essentially one-exponential reactions on the stopped-flow
time scale, as the decomposition of peroxo adducts was much
slower than oxygen binding to diiron(II) complexes 1 and 6
(Figure 5). Spectral changes upon oxygenation observed by
stopped-flow spectrophotometry were identical to those
obtained by regular spectrophotometry in the “batch” experi-
ments and corresponded to clean peroxide formation in the
first stage of the reactions.
Figure 4. IR spectra of (A) 6 and (B) 6‚O2. Black spectra correspond to
samples with natural abundance benzoate, and red spectra correspond to
samples with benzoate 13C-labeled at the carboxylate group. Asterisks
designate peaks arising from decayed product.
Table 4. Comparison of Carboxylate Stretches (cm-1) for 6 and 6‚O2
asym
νCOO
sym
sample
νCOO νCOO νCOO ∆ν(asym - sym)
6 (O2n.a.CAr)
1622, 1613 1565 1401 1378
1590 1541 1370 1351
1636, 1623 1574 1351 1333
216, 187
220, 190
279, 241
284, 269
154
6 (O213CAr)
6‚O2 (O2n.a.CAr)
6‚O2 (O213CAr)
1594
1561 1310 1292
Me4N+C6H5n.a.COO- 1553
Me4N+C6H513COO- 1516
1399
1368
148
affected by the introduction of the second carboxylate
group.45
FT-IR spectroscopy was used to show that the carboxylate
bridges of 6 became monodentate upon formation of the
peroxo intermediate. Carboxylates exhibit characteristic sym-
metric and asymmetric stretching modes, the separation (∆ν)
of which is indicative of the binding mode of the carboxy-
late.46 The ∆ν value for a bidentate carboxylate bridge should
be comparable to that of the ionic carboxylate but smaller
than that for a monodentate carboxylate. The IR spectrum
of a KBr pellet of 6 shows features at 1622, 1613, 1565,
1401, and 1378 cm-1, which shift to 1590, 1541, 1370, and
1351 cm-1 with 13C labeling of the carboxylate carbon on
the benzoate group (Figure 4). The presence of an intense
13C-insensitive feature at 1602 cm-1 and the observation of
three 13C-sensitive peaks in the νasym region complicate the
interpretation of the natural abundance data. However, the
13C data are more straightforward to analyze, affording ∆ν
values of 220 and 190 cm-1 for 6. For comparison,
(Me4N)(O213CC6H5) has features at 1516 and 1368 cm-1,
giving a ∆ν of 148 cm-1 (Table 4). Thus, the ∆ν values
observed for 6 are consistent with the presence of bridging
carboxylates. The IR spectrum of 6‚O2, on the other hand,
shows features at 1636, 1623, 1574, 1351, and 1333 cm-1,
which shift to 1594, 1561, 1310, and 1292 cm-1 in the 13C
Rate Law and Activation Parameters. The observed
pseudo-first-order rate constants for peroxo complex forma-
tion were independent of the concentration of the diiron(II)
complexes but increased linearly with dioxygen concentration
(Figure 6). Consequently, the oxygenation of complexes 1
and 6 is a second-order process (first-order in diiron(II)
complex, and first-order in O2):
(45) Brunold, T. C.; Tamura, N.; Kitajima, M.; Moro-oka, Y.; Solomon,
E. I. J. Am. Chem. Soc. 1998, 120, 5674-5690.
(46) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coor-
dination Compounds, 4th ed.; John Wiley: New York, 1986.
V ) kobs[Fe2] ) k[O2][Fe2]
The rate law was identical for all reactions studied (Table
Inorganic Chemistry, Vol. 42, No. 23, 2003 7525

Costas et al.
equiv of Ph3PO, and by a factor of 3 in the presence of 10%
DMSO. These effects, however, are relatively small as
compared to the dramatic increase observed in the stability
of the peroxo adducts. Thus, coordination of labile oxygen
donors is not detrimental for the oxygenation step and highly
effective for suppressing peroxide decomposition pathways.
The oxygenation of the diiron(II) complexes with HPTP is
much faster than the oxygenation of a six-coordinate,
sterically hindered [Fe2(µ-OH)2(6-Me3-TPA)2]2+ complex
investigated previously.22 These results suggest that the HPTP
scaffold does not block access of O2 to the diiron(II) center
and allows for rapid and efficient oxygenation.
The oxygenation rates of diiron(II) HPTP complexes
depend on the polarity of the solvent (Table 5). For 1, diiron-
(III)-peroxo formation is fastest in CH3CN and 7-fold slower
in proprionitrile.19 Similarly for 6, the rate is an order of
magnitude faster in 1:1 CH2Cl2/CH3CN than in CH2Cl2 alone.
The possible coordination of CH3CN to the metal center(s)
as the rationale for the rate acceleration observed for 6 can
be discarded, as the presence of 2.5 vol % CH3CN in
CH2Cl2 ([CH3CN]/[6] ∼ 1000) does not change the oxy-
genation rate as compared to pure CH2Cl2, but larger volume
fractions of CH3CN in CH2Cl2 (up to 50 vol %) do lead to
substantially faster (up to 10-fold) oxygenation rates (Table
5). The increase in oxygen binding rates with an increase in
solvent polarity supports a polar transition state for the
oxygenation process, consistent with the notion that O2
coordination to iron(II) is accompanied by electron transfer.
Oxygen binding to diiron(II) HPTP complexes studied in
this work is a low-barrier process, with an activation enthalpy
of only 8-20 kJ/mol (Table 5). The differences in reaction
rates originate primarily from the differences in the activation
entropies, which become more unfavorable for coordinatively
saturated species, so steric effects are responsible for the
differences in oxygenation rates. This pattern of small
activation enthalpies and large negative activation entropies
is commonly found for the association of O2 and a reduced
metal center when formation of a metal-superoxo adduct
is believed to be the rate-limiting step. Examples include
myoglobin (∆Hq ) 21 kJ mol-1) and monomeric hemoglobin
(∆Hq ) 24 kJ mol-1),50,51 as well as dicopper(I) complexes
of dinucleating ligands related to HPTP.52 An extreme exam-
ple is the slow oxygenation of [Fe2(OH)2(6-Me3TPA)2]2+,
where the bulky ligand prevented easy access of the O2
molecule to the diiron(II) center, giving rise to a large neg-
ative activation entropy.22 It can be concluded that a delicate
balance is required in designing dinuclear iron complexes
for oxygen binding and activation. On one hand, steric
protection of the metal center is necessary to prevent the
Figure 5. Spectral changes upon oxygenation of 1 in CH3CN at -40 °C.
Individual spectra were collected at 10.0 ms intervals. [1] ) 2.23 ×
10-4 M, [O2] ) 4.05 × 10-3 M. The inset shows a kinetic trace at
573 nm that corresponds to a pseudo-first-order formation of 1‚O2 (kobs
)
30.5 s-1).
Figure 6. Plot of the observed pseudo-first-order rate constant, k(obs), vs
dioxygen concentration for the oxygenation of 6 in CH2Cl2. Temperature
-20.9 °C, initial concentration of 6 is 4.58 × 10-5 M.
5) and did not change with temperature; the oxygenation of
6 in methylene chloride remained a second-order process
even at -80 °C. Activation parameters of the oxygenation
reactions were determined from linear Arrhenius and Eyring
plots (Figure 7, Table 5). Small activation enthalpies (∆Hq
) 8-19 kJ/mol) and large negative activation entropies (∆Sq
) -140 to -90 J/K mol) are typical of associative
reactions.47 Thus, ligand dissociation to generate a binding
site is facile and cannot be the rate-limiting step in the
oxygenation of any of the complexes studied in this work.
For example, in the case of 6, a carboxylate shift42 from a
bridging bidentate mode to a terminal monodentate mode is
quite plausible and suggested by the longer Fe1-O2 and
Fe2-O3 bridging carboxylate distances observed in its
crystal structure (Table 2).
While oxygen binding to 1 and 6 under different conditions
shares the same rate-limiting step, the oxygenation rates are
somewhat sensitive to the nature of additional ligands and
solvents (Figure 7 and Table 5). Oxygen binding to 6 is
almost an order of magnitude slower than to its coordina-
tively unsaturated counterpart 1 in the same solvent mix-
ture (1:1 CH3CN/CH2Cl2). Similarly, oxygenation of 1 in
CH3CN is retarded by a factor of 2 in the presence of 10
(48) Valentine, A. M.; Stahl, S. S.; Lippard, S. J. J. Am. Chem. Soc. 1999,
121, 3876-3887.
(49) Broadwater, J. A.; Achim, C.; Mu¨nck, E.; Fox, B. G. Biochemistry
1999, 12197-12204.
(50) Warburton, P. R.; Busch, D. H. In PerspectiVes on Bioinorganic
Chemistry; JAI Press: London, 1993; Vol. 2, pp 1-79.
(51) Wilkins, R. G. In Oxygen Complexes and Oxygen ActiVation by
Transition Metals; Martell, A. E., Sawyer, D. T., Eds.; Plenum Press:
New York, 1987; pp 49-60.
(52) Karlin, K. D.; Kaderli, S.; Zuberbu¨hler, A. D. Acc. Chem. Res. 1997,
30, 139-147.
(47) Tobe, M. L.; Burgess, J. Inorganic Reaction Mechanisms; Longman:
London, 1999.
7526 Inorganic Chemistry, Vol. 42, No. 23, 2003

Carboxylates in Nonheme Diiron(II)/O2 ReactiWity
Table 5. Kinetic Parameters for the Oxygenation of 1, 6, and Other Diiron(II) Complexes
k
obs, s-1
k, M-1 s-1
(-40 °C)
∆Hq,
kJ/mol
∆Sq,
J/K mol
complex, solvent/additive
1, CH3CH2CN
(-40 °C, 100% O2)
ref
4.22(13)
29.4(8)
11.4(4)
18.3(8)
18.8(8)
0.195(5)
0.203
0.250
0.470
2.04(6)
4.40(2)
2.08(15) × 10-3
9.6(3) × 102
7.3(5) × 103
2.8(2) × 103
4.5(3) × 103
5.4(4) × 103
67(3)
16.5 ( 0.4
15.8 ( 4
8.0 ( 3
15.9 ( 3
19.0 ( 4
16.7 ( 2
N/A
-114 ( 2
-101 ( 10
-143 ( 10
-104 ( 10
-90 ( 10
-132 ( 8
N/A
19
a
a
a
a
a
a
a
a
1, CH3CN
1, CH3CN/DMSO (9:1 v/v)
1, CH3CN/10 equiv Ph3PO
1, CH3CN/CH2Cl2 (1:1 v/v)
6, CH2Cl2
6, CH3CN/CH2Cl2 (1:40 v/v)b
6, CH3CN/CH2Cl2 (1:10 v/v)b
6, CH3CN/CH2Cl2 (1:2 v/v)b
6, CH3CN/CH2Cl2 (1:1 v/v)
[Fe2(N-Et-HPTB)(OBz)]2+, CH3CH2CN
[Fe2(6-Me3TPA)2(µ-OH)2]2+, CH2Cl2
MMOH (H2O)
70.0
86.2
N/A
N/A
N/A
N/A
162
5.9(1) × 102
16.3 ( 4
15.4 ( 0.6
17 ( 2
92 ( 17
22
-120 ( 10
-121 ( 3
-175 ( 20
88 ( 42
-134
a
103(12)
19
22
48
49
0.66(6)
∆9D (H2O)
a This work. b Oxygen solutions were prepared in pure CH2Cl2.
Table 6. Kinetic Parameters for the Decay of Diiron(III)-Peroxo
Complexes
∆Hq,
∆Sq,
complex, solvent
1, CH3CH2CN
order
KJ mol-1
J K-1mol-1 ref
2nd
2nd
2nd
1st
1st
1st
51.9 ( 2.7
44 ( 5
22 ( 6b
19
a
a
a
a
1, CH3CN
-21 ( 16
-97 ( 16
-79 ( 12
-50 ( 8
59 ( 15
1, CH3CN, 20 equiv Ph3PO
1, CH2Cl2/DMSO (9:1 v/v)
3, CH2Cl2/DMSO (9:1 v/v)
6, CH2Cl2
40 ( 4
49 ( 4
60 ( 4
103 ( 4
80.6 ( 3.9
a
19
[Fe2(N-Et-HPTB)(OBz)]2+
CH3CH2CN
,
2nd
74 ( 14
[Fe2(N-Et-HPTB)(OBz)]2+
CH2Cl2
,
1st
66.6 ( 3.3
-40 ( 20
15
MMOH (H2O)
∆9D (H2O)
1st
1st
114
80
163
-41
48
49
a This work. b ∆Sq was reported by Feig et al.19a as -47 ( 11 J K-1
mol-1, but a reexamination of the data provided in the supporting material
shows that the activation entropy should be changed to the value listed in
this table.19b
Figure 7. Arrhenius plots for the oxygenation of diiron(II) complexes:
(a) 1 in CH3CN; (b) 1 in CH3CN/CH2Cl2 (1:1 v/v); (c) 1 in CH3CN in the
presence of Ph3PO (2.5 × 10-4 M solution of 1, 5 × 10-3 M solution of
Ph3PO); (d) 1 in CH3CN/DMSO (9:1 v/v); (e) 6 in CH3CN/CH2Cl2 (1:1
v/v); (f) 6 in CH2Cl2.
rapid second-order process.19 In complete agreement with
these previous findings, our stopped-flow experiments per-
formed in the -40 to 0 °C temperature range revealed a
very similar second-order decomposition of 1‚O2 in aceto-
nitrile (Figure S1). However, the solvent had to be doubly
distilled to obtain the second-order behavior; otherwise, the
kinetic traces were more complex and poorly reproducible,
possibly arising from a combination of first- and second-
order processes. The unusual second-order nature of the
decomposition reaction was confirmed by excellent fits of
the experimental kinetic traces to a second-order equa-
tion, and by the independence of the second-order rate
constant on the initial concentration of the peroxo species.
As expected, the initial rates of the peroxide decomposi-
tion depended dramatically on the complex concentration.
These reaction rates and the activation parameters derived
from them were also similar to those reported by Feig et al.
(Table 6).19
Figure 8. Relationship between the rates of decay (k, s-1) of the peroxo
adducts of 1-5 (2) at -40 °C and the peroxo adducts of 6-10 at 0 °C (9)
versus the Hammett σ value of the para substituent in the bridging benzoate
(p-R-C6H4CO2-).
As previously observed for [Fe2(N-Et-HPTB)(O2)-
(O2CR)]2+, the addition of Ph3PO to 1‚O2 greatly stabilized
the intermediate and allowed its decay to be monitored by
conventional UV-vis spectroscopy. Kinetic analysis revealed
that the decay of 1‚O2 remains a second-order process, whose
rate is inversely proportional to Ph3PO concentration (Figure
S2). This observation strongly suggests that added Ph3PO is
unwanted autoxidation processes; on the other hand, exces-
sive steric hindrance would substantially limit the oxygen-
ation rates.
Peroxide Decomposition Kinetics. The decomposition of
the peroxo species 1‚O2 in propionitrile was previously
investigated by stopped-flow methods and reported to be a
Inorganic Chemistry, Vol. 42, No. 23, 2003 7527

Costas et al.
involved in a binding preequilibrium that limits the concen-
tration of the species that undergoes second-order decay.
Stopped-flow studies of the reaction of 1 with O2 in
CH2Cl2 indicated that the decay of 1‚O2 in this solvent was
similar to that in CH3CN and thus not investigated in detail.
However, when a 9:1 (v/v) mixture of CH2Cl2/DMSO was
used as solvent, the 1‚O2 adduct was significantly stabilized,
and the decomposition of 1‚O2 became a first-order process
with a rate constant independent of the concentration of the
complex. Considering the importance of using purified
solvent in the CH3CN study, the first-order behavior observed
in the presence of DMSO may be due to the introduction of
an impurity present in DMSO or due to a change in mech-
anism. In any case, it is clear that the nature of the solvent
can play an important role in determining the mechanism of
peroxide decomposition.
The activation parameters for the decay of 1‚O2 under
different conditions as well as that for 3‚O2 in CH2Cl2/DMSO
are collected in Table 6, and Eyring plots are collected in
Figure S3. While the activation enthalpies for peroxo decay
in acetonitrile, propionitrile, and CH2Cl2/DMSO fall within
a relatively narrow range of 40-60 kJ mol-1 K-1, the
obtained activation entropies show much greater variability,
ranging from +22 to -97 J mol-1 K-1. It is difficult to
rationalize the observed differences in the light of the distinct
molecularities of the decay processes, but the decomposition
of these peroxo intermediates is very likely a multistep
reaction, thereby complicating the molecular interpretation
of the activation parameters.
The effects of benzoate ring substituents on the rates of
decomposition of a series of O2 adducts of mono- and
dicarboxylate-bridged diiron(II) complexes were also inves-
tigated. As shown in Figure 8, the decay rates for the peroxo
intermediates from monocarboxylate precursors correlate
with Hammett σ values to give a F value of -1.1. Hence,
increasing the donating ability of the carboxylate diminishes
the stability of the peroxide, as previously observed for
related O2 adducts of [Fe2(N-Et-HPTB)(O2CR)] complexes.15
In sharp contrast, the decay rates for the dicarboxylate O2
adducts were not significantly affected by the different
substituents on the carboxylates. This strongly suggests that
the stability conferred to the peroxide intermediate by the
second carboxylate is due not to electronic factors but instead
to the coordinative saturation of the metal center.
Substrate Oxidation Studies. The oxidative abilities of
the O2 adducts versus different substrates were investigated.
Due to its short lifetime in CH3CN solvent, the reaction of
1‚O2 with Ph3P was monitored with double-mixing stopped-
flow methodology. The short-lived 1‚O2 was generated in
the first mixing loop, and the substrate was added in the
second mixing step with a fixed time delay that ranged from
500 ms at -40 °C to 40 ms at 0 °C. In all experiments, the
rate of decomposition of 1‚O2 was accelerated 10- to 100-
fold upon addition of Ph3P (Figure 9). The reaction kinetics,
however, were complicated and could not be adequately fit
to either a pseudo-first-order or a second-order process. The
initially rapid reaction slowed with time. Most likely, this
behavior could be attributed to product inhibition. The
oxidation of Ph3P yields Ph3PO that effectively competes
with PPh3 for the labile coordination sites of 1‚O2 (X and Y
in Scheme 2).
Relative to the O2 adducts of the monocarboxylate-bridged
complexes, the 6‚O2 intermediate is even more stable. It
decomposes via a first-order process in pure CH2Cl2, with a
rate constant that is independent of the concentration of
complex. At 0 °C, the rate of decomposition is 1.43 × 10-4
s-1, compared to a calculated value of 1.45 × 10-1 s-1 for
1‚O2 in CH2Cl2/DMSO (9:1 v/v) obtained by extrapolation
of the corresponding Eyring plot. In other words, the
introduction of the second carboxylate group enhances the
stability of the peroxide by 3 orders of magnitude over that
of 1‚O2, already stabilized by the presence of DMSO. The
higher stability of 6‚O2 is reflected by a 50 kJ mol-1 higher
activation enthalpy. Opposite to 1‚O2, the activation entropy
for the decay of 6‚O2 is positive and indicative of a
dissociative mechanism. Given the instability of the mono-
carboxylate peroxo adduct in the absence of DMSO, it is
reasonable to propose that the complete dissociation of the
second carboxylate may be the rate determining step in the
decomposition of 6‚O2. The observed high barrier can easily
be rationalized by the charge separation required for complete
dissociation. However, the rate of decay of 6‚O2 was found
The kinetic behavior of 1‚O2 with PPh3 was simpler in
CH2Cl2/DMSO (9:1 v/v). Figure 10 shows that the rate of
first-order decay of 1‚O2 increases linearly with PPh3
concentration at low concentrations but plateaus at higher
concentration. This saturation behavior indicates a preequi-
librium binding of PPh3 to the diiron center prior to its
oxidation by 1‚O2, i.e.
[(HPTP)Fe2(OBz)(O2)]2+ + PPh3 yk1z
k-1
[(HPTP)Fe2(OBz)(O2)(PPh3)]2+ (1)
k2
[(HPTP)Fe2(OBz)(O2)(PPh3)]2+ 9
8
[(HPTP)Fe2O(OBz)]2+ + OPPh3 (2)
By incorporating a pre-equilibrium step, the rate law can be
expressed as eq 3:
-
-d[1‚O2]/dt ) k2K1[1‚O2][PPh3](1 + K1[PPh3])-1 (3)
to be independent of added ArCO2 . Furthermore, the decay
of 1‚O2 in CH3CN could be dramatically suppressed in a
double mixing stopped-flow experiment by adding either 1
or 20 equiv of benzoate ion after formation of 1‚O2 but prior
to its decay. This result shows that binding of the added
benzoate to 1‚O2 is rapid and a rapid pre-equilibrium between
1‚O2 and 6‚O2 is quite unlikely to be a factor in the
mechanism of decomposition.
From the double reciprocal plot shown in Figure 10 inset,
the K1 of this pre-equilibrium is estimated to be 120 M-1 at
-40 °C. Thus, PPh3 must bind to the diiron-peroxo
intermediate prior to being oxidized.
Studies of the reactivity of 1‚O2 with other substrates in
CH2Cl2/DMSO (9:1 v/v) at -40 °C are listed in Table 7.
7528 Inorganic Chemistry, Vol. 42, No. 23, 2003

Carboxylates in Nonheme Diiron(II)/O2 ReactiWity
weaker than those of benzyl alcohol (78 vs 81 kcal/mol).53
Such a lack of hydrogen atom abstraction ability argues
against the participation of high valent species in the
decomposition of the peroxo intermediate, which would be
expected to be good hydrogen atom abstracting agents and
thus effective in oxidizing substrates such as dihydroan-
thracene. What distinguishes dihydroanthracene from the
other substrates listed in Table 7 is its inability to coordinate
to the diiron center. We thus propose that the oxidation of
the substrates listed in Table 7 involves initial coordina-
tion to the peroxo intermediate and subsequent reaction
with the metal-bound peroxide. In support of this inner
sphere oxidation mechanism, we note that the coordinatively
saturated 6‚O2 is unable to oxidize PPh3, benzyl alcohol,
methyl phenyl sulfide, nor 9,10-dihydroanthracene, presum-
ably because these potential substrates lack sufficient affinity
for iron(III) to displace one of the bound carboxylates in
6‚O2.
Figure 9. Decomposition of 1‚O2 generated at -40 °C in a stopped-flow
mixing cell: (a) in pure CH3CN; (b) in the presence of PPh3 added in the
second mixing loop. Initial concentrations of reagents in the mixing cell:
[1] ) 0.250 mM; [O2] ) 2.03 mM; [PPh3] ) 1.00 mM; solvent, CH3CN.
Conclusions
In this paper, we have synthesized and characterized a
series of diiron(II) complexes of the dinucleating ligand
HPTP supported by one or two carboxylate bridges, in order
to examine the role these carboxylates play in the reaction
of the diiron(II) complex with O2. The monocarboxylate-
bridged complexes have at least one available coordination
site, while the dicarboxylate-bridged complexes are coordi-
natively saturated. Nevertheless, all complexes bind O2 in a
second-order reaction to form (µ-1,2-peroxo)diiron(III) in-
termediates. However, the addition of the second bridge
results in a decrease of the rate of oxygenation by an order
of magnitude, presumably due to a required carboxylate shift
from bridging to terminal to open up a site for O2 binding.
The decomposition of peroxo intermediates derived from
monocarboxylate precursors is somewhat complex. It exhibits
second-order behavior with respect to diiron complex in
organonitrile solvents but becomes first-order in CH2Cl2/
DMSO (9:1 v/v). Electron withdrawing substituents on the
benzoate slow the rate of decomposition, in agreement with
an observation made previously for a related series of
[Fe2(O2)(N-Et-HPTB)(O2CAr)]2+ intermediates.15 Substrate
oxidation follows an inner sphere mechanism, requiring prior
coordination of the substrate to an open coordination site
on the diiron center and subsequent reaction with the metal-
bound peroxide.
Figure 10. Dependence of the rate of decay of 1‚O2 in CH2Cl2/DMSO
(9:1 v/v) at -40 °C with added PPh3.
a
Table 7. Oxidation Reactions of 1‚O2 and 6‚O2
% yield
% yield
substrate
product
OPPh3
methyl phenyl
sulfoxide
from [1‚O2] from [6‚O2]
PPh3
100
59
n.r.b
n.r.b
methyl phenyl
sulfide
thiophenol
3,5-di-tert-butyl-
catechol
diphenyl disulfide
3,5-di-tert-butyl-
quinone
100
100
benzyl alcohol
cyclohexanol
9,10-dihydroanthracene anthracene
benzaldehyde
cyclohexanone
100
23
n.r.b
n.r.b
n.r.b
a The 60 mM substrate reacted with 3 mM 1‚O2 in CH2Cl2/DMSO (9:1
v/v) or 6‚O2 in CH2Cl2 at -40 °C. b n.r. ) no reaction.
Complex 1‚O2 effects efficient oxidation of methyl phenyl
sulfide to the sulfoxide, without any overoxidation to the
sulfone. Similarly to what was observed with the addition
of PPh3, the rate of decay of 1‚O2 is accelerated by the
increase of the concentration of substrate, suggesting that
the peroxo complex itself is responsible for the oxygen
transfer to the substrate. The peroxo complex also quanti-
tatively oxidizes thiophenol to the corresponding disulfide
and 3,5-di-tert-butylcatechol to the quinone. With respect
to alcohols, benzyl alcohol is oxidized quantitatively to
benzaldehyde, but cyclohexanol with the stronger R-C-H
bond is oxidized to cyclohexanone only in 23% yield. In
contrast, 1‚O2 does not oxidize 9,10-dihydroanthracene at
all, despite that the fact that its C-H bonds are somewhat
The unexpected finding of this work is the remarkable
stability of the peroxo species generated from the dicar-
boxylate-bridged precursors 6-10. IR experiments indicate
that the carboxylate bridges in the precursor have become
terminal ligands in the O2 adduct. However, unlike their
monocarboxylate analogues, the stability of the peroxo
intermediates is not significantly affected by substituents on
the benzoates. The stability of the (µ-1,2-peroxo)diiron(III)
unit in these complexes is thus determined by the acces-
(53) Bordwell, F. G.; Cheng, J.-P.; Ji, G.-Z.; Satish, A. V.; Zhang, X. J.
Am. Chem. Soc. 1991, 113, 9790-9795.
(54) Broadwater, J. A.; Ai, J.; Loehr, T. M.; Sanders-Loehr, J.; Fox, B. G.
Biochemistry 1998, 37, 14664-14671.
Inorganic Chemistry, Vol. 42, No. 23, 2003 7529

Costas et al.
Scheme 3
sibility of potential reducing agents to the diiron center. By
effectively blocking access, the terminal carboxylates of
6‚O2 and, by extension, the Ph3PO ligands of the crystallo-
graphically characterized [Fe2(O2)(N-Et-HPTB)(Ph3PO)2]3+
intermediate15 limit the reactivity of the diiron(III)-peroxo
species and enhance the thermal stability of these intermedi-
ates.
the synthetic analogues, since such substrates cannot coor-
dinate to the diiron center. Thus, their oxidation may occur
by proximity to the diiron centers or may require some
activation step, perhaps protonation of the bound peroxide,
that promotes cleavage of the O-O bond prior to substrate
oxidation. Indeed, such a protonation step is strongly
implicated by the pH dependence observed for the conversion
of peroxo intermediate P to high valent intermediate Q in
single turnover studies of methane monooxygenase.55
These observations can be rationalized by the mechanistic
picture shown in Scheme 3. The monocarboxylate-bridged
diiron(II) precursor reacts with O2 to form adduct A, which
decays rapidly via a second-order process. In the presence
of an oxygen donor ligand like Ph3PO, DMSO, or a second
carboxylate, the O2 adduct becomes stabilized as structure
B. Adduct decay can be initiated by dissociation of Y. In
polar CH3CN, the dissociation of Ph3PO is facile, and the
peroxo decomposition retains its second-order character.
However, in less polar CH2Cl2, ligand dissociation is less
facile and can become partially rate determining. Such a
situation is suggested by the more complex kinetics of peroxo
decay in the presence of 20 equiv of Ph3PO. When Y is a
second carboxylate, dissociation of Y becomes kinetically
and thermodynamically unfavorable, and a slow first-order
decomposition of a coordinatively saturated peroxo complex
B dominates. Furthermore, substrates that can occupy site
Y are oxidized, and the 1‚O2‚substrate adduct decays in a
first-order process with the peroxide attacking the substrate
in the metal coordination sphere.
Acknowledgment. This work was supported by NIH
Grant GM-38767 to L.Q. and NSF Grant CHE0111202 and
Research Corporation Grant RI0223 to E.V.R.-A. M.C.
thanks Fundacio La Caixa for a postdoctoral fellowship that
supported in part his stay at the University of Minnesota.
M.J.R. was supported by the Lando Summer Research
Fellowship Program, while C.W.C. was supported in part
by a grant from the University of Minnesota’s Undergraduate
Research Opportunities Program.
Supporting Information Available: X-ray crystallographic file
for [FeII2(HPTP)(O2CC6F5)(CH3CN)](BPh4)2‚CH3CN (3) and
[FeII (HPTP)(O2CC6H5)2](BF4)‚CH3CN (6) in CIF format and a pdf
2
file containing Tables S1 and S2, respectively, listing NMR shifts
of diiron(II) complexes 1-10 and the absorption maxima of
corresponding O2 adducts, Figure S1 showing kinetic data for the
decomposition of 1‚O2 in CH3CN, Figure S2 showing the effect of
added OPPh3 on the decay of the O2 adduct of 1, and Figure S3
showing Eyring plots for the decomposition of the diiron(III)-
peroxo intermediates. This material is available free of charge via
the Internet at http://pubs.acs.org.
Extrapolation of these notions to the reactivity of diiron-
(III)-peroxo species in biology suggests the following
insight. The hydrocarbon substrates of methane monooxy-
genase and fatty acid desaturases are unlikely to be oxidized
by their respective (µ-1,2-peroxo)diiron(III) intermedi-
ates48,49,54 by a direct coordination mechanism as found for
IC034359A
(55) Lee, S.-K.; Lipscomb, J. D. Biochemistry 1999, 38, 4423-4432.
7530 Inorganic Chemistry, Vol. 42, No. 23, 2003