Dioxygen binding to complexes with FeII2(μOH) 2 cores: Steric control of activation barriers and O 2-adduct formation

Article abstract of DOI:10.1021/ic0485312

A series of complexes with [Fe^II₂(μ-OH) ₂] cores has been synthesized with N3 and N4 ligands and structurally characterized to serve as models for nonheme diiron(II) sites in enzymes that bind and activate O₂. These complexes react with O₂ in solution via bimolecular rate-limiting steps that differ in rate by 10 ³-fold, depending on ligand denticity and steric hindrance near the diiron center. Low-temperature trapping of a (μ-oxo)(μ-1,2-peroxo) diiron(III) intermediate after O₂ binding requires sufficient steric hindrance around the diiron center and the loss of a proton (presumably that of a hydroxo bridge or a yet unobserved hydroperoxo intermediate). The relative stability of these and other (μ-1,2-peroxo)diiron(III) intermediates suggests that these species may not be on the direct pathway for dioxygen activation.

Full text of DOI:10.1021/ic0485312

Inorg. Chem. 2005, 44, 85−99
Dioxygen Binding to Complexes with FeII2(µ-OH)2 Cores: Steric Control
of Activation Barriers and O2-Adduct Formation
Sergey V. Kryatov, Sonia Taktak, Ivan V. Korendovych, and Elena V. Rybak-Akimova*
Department of Chemistry, Tufts UniVersity, 62 Talbot AVenue, Medford, Massachusetts 02155
Jo´zsef Kaizer, Ste´phane Torelli, Xiaopeng Shan, Sanjay Mandal, Vicki L. MacMurdo,
Antoni Mairata i Payeras, and Lawrence Que, Jr.*
Department of Chemistry and Center for Metals in Biocatalysis, UniVersity of Minnesota,
207 Pleasant Street SE, Minneapolis, Minnesota 55455
Received October 19, 2004
A series of complexes with [FeII
2(µ-OH)2] cores has been synthesized with N3 and N4 ligands and structurally
characterized to serve as models for nonheme diiron(II) sites in enzymes that bind and activate O2. These complexes
react with O2 in solution via bimolecular rate-limiting steps that differ in rate by 103-fold, depending on ligand
denticity and steric hindrance near the diiron center. Low-temperature trapping of a (µ-oxo)(µ-1,2-peroxo)diiron(III)
intermediate after O2 binding requires sufficient steric hindrance around the diiron center and the loss of a proton
(presumably that of a hydroxo bridge or a yet unobserved hydroperoxo intermediate). The relative stability of these
and other (µ-1,2-peroxo)diiron(III) intermediates suggests that these species may not be on the direct pathway for
dioxygen activation.
Introduction
genases.8 Extensive studies have identified diiron(III)-
peroxo and higher-valent iron-oxo intermediates in the
catalytic cycles of some of these enzymes. A terminal end-
on hydroperoxide is established for oxyhemerythrin,4 while
bridged peroxo species are observed for MMO,1,7 ∆9D,9 and
a mutant RNR.10 An iron(III)iron(IV) intermediate called X
is directly responsible for tyrosyl radical formation in RNR,5
while a diiron(IV) intermediate called Q is the key oxidant
that hydroxylates methane in MMO.1,7
Model complexes have played critical roles in providing
synthetic precedents for the proposed structures of these
catalytic intermediates.3,11-15 For the peroxo intermediates
that are the focus of this paper, several O2 adducts of
Diiron(II) centers participate in the binding and activation
of dioxygen at a number of metalloprotein active sites.1-3
The reactions that occur include reversible oxygenation of
hemerythrin4 and the oxidations of various substrates, such
as the formation of a catalytically essential tyrosyl radical
from an endogenous tyrosine in class I ribonucleotide reduc-
tases (RNR),5 the dehydrogenation of fatty acid alkyl chains
by stearoyl-ACP ∆9-desaturase (∆9D),6 and the hydroxyl-
ation of methane and alkanes by soluble methane monooxy-
genase (MMO)7 and membrane bound alkane monooxy-
* To whom correspondence should be addressed. E-mail:
elena.rybak-akimova@tufts.edu (E.V.R.-A.); que@chem.umn.edu (L.Q.).
(1) Wallar, B. J.; Lipscomb, J. D. Chem. ReV. 1996, 96, 2625-2657.
(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) Lee, D.; Lippard, S. J. Nonheme Di-iron Enzymes. In ComprehensiVe
Coordination Chemistry II; McCleverty, J. A., Meyer, T. J., Eds.;
Elsevier: Oxford, U.K., 2004; Vol. 8, pp 309-342.
(4) Stenkamp, R. E. Chem. ReV. 1994, 94, 715-726.
(5) Stubbe, J.; Riggs-Gelasco, P. Trends Biochem. Sci. 1998, 23, 438-
443.
(8) Shanklin, J.; Whittle, E. FEBS Lett. 2003, 545, 188-192.
(9) Broadwater, J. A.; Ai, J.; Loehr, T. M.; Sanders-Loehr, J.; Fox, B. G.
Biochemistry 1998, 37, 14664-14671.
(10) Moe¨nne-Loccoz, P.; Baldwin, J.; Ley, B. A.; Loehr, T. M.; Bollinger,
J. M., Jr. Biochemistry 1998, 37, 14659-14663.
(11) Que, L., Jr.; Dong, Y. Acc. Chem. Res. 1996, 29, 190-196.
(12) Que, L., Jr.; Tolman, W. B. Angew. Chem., Int. Ed. 2002, 41, 1114-
1137.
(13) Suzuki, M.; Furutachi, H.; Okawa, H. Coord. Chem. ReV. 2000, 200-
202, 105-129.
(6) Broadwater, J. A.; Haas, J. A.; Fox, B. G. Fett/Lipid 1998, 100, 103-
(14) Du Bois, J.; Mizoguchi, T. J.; Lippard, S. J. Coord. Chem. ReV. 2000,
200-202, 443-485.
113.
(7) 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.
(15) Tolman, W. B.; Que, L., Jr. J. Chem. Soc., Dalton Trans. 2002, 653-
660.
10.1021/ic0485312 CCC: $30.25
Published on Web 12/14/2004
© 2005 American Chemical Society
Inorganic Chemistry, Vol. 44, No. 1, 2005 85

Kryatov et al.
Scheme 1. Catalytic Cycle of the Diiron Center in the R2 RNR
Protein (Top) and Its Functional Model, Complex 1 (Bottom)
Scheme 2. Structural Formulas of the Ligands Used in This Work
TPA)2]2+ (1) was synthesized and structurally characterized.26
This complex reacted with O2 at low temperature to generate
a metastable peroxo intermediate, which could undergo an
acid-promoted conversion to an Fe(III)Fe(IV) species capable
of oxidizing phenols.26 This system thus mimics all essential
steps in the formation of the diiron(III)-tyrosyl radical
cofactor of RNR (Scheme 1).12,26 However, for 1, the rates
of the individual reactions shown in Scheme 1 are substan-
tially slower than those of the corresponding enzymatic
transformations.26,27 Detailed studies of the oxygenation of
this complex suggested that the steric bulk imposed by the
6-methyl substituents on the pyridines shielded the diiron
site from access by O2.27 We anticipated that removing
methyl substituents and/or replacing CH3 groups with planar
aromatic substituents would significantly influence the
accessibility of the iron centers and the rates of key reactions
in O2 activation cycle. We have thus introduced systematic
changes in the ligand structure (Scheme 2) to prepare a series
diiron(II) complexes have been characterized.3,13-29 Efforts
in the mid-1990s led to the crystallization of three O2 adducts
whose structures revealed the formation of (µ-1,2-peroxo)-
diiron(III) units supported by additional alkoxo and/or
carboxylate bridges.16,18,19 For these examples, the peroxo
intermediate was either quite stable or underwent subsequent
decay to iron(III) products without evidence of any higher-
valent iron species that may result from O-O bond cleavage.
Subsequently, the diiron(II) complex [Fe2(µ-OH)2(6-Me3-
(16) 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.
(17) Sugimoto, H.; Nagayama, T.; Maruyama, S.; Fujinami, S.; Yasuda,
Y.; Suzuki, M.; Uehara, A. Bull. Chem. Soc. Jpn. 1998, 71, 2267-
2279.
(18) Kim, K.; Lippard, S. J. J. Am. Chem. Soc. 1996, 118, 4914-4915.
(19) Dong, Y.; Yan, S.; Young, V. G., Jr.; Que, L., Jr. Angew. Chem., Int.
Ed. Engl. 1996, 35, 618-620.
(20) Kitajima, N.; Tamura, N.; Amagai, H.; Fukui, H.; Moro-oka, Y.;
Mizutani, Y.; Kitagawa, T.; Mathur, R.; Heerwegh, K.; Reed, C. A.;
Randall, C. R.; Que, L., Jr.; Tatsumi, K. J. Am. Chem. Soc. 1994,
116, 9071-9085.
(21) Hayashi, Y.; Kayatani, T.; Sugimoto, H.; Suzuki, M.; Inomata, K.;
Uehara, A.; Muzitani, Y.; Kitagawa, T.; Maeda, Y. J. Am. Chem. Soc.
1995, 117, 11220-11229.
(22) (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.
of complexes with [FeII (µ-OH)2] cores. These complexes
2
react with O2 at rates that differ by 103-fold. Detailed
structural characterization of this series of complexes and
electrochemical studies of their redox properties allow us to
evaluate the relative importance of steric and electronic
factors in the oxygenation kinetics. This first detailed study
of a series of [FeII (µ-OH)2] complexes re-emphasizes the
2
importance of steric control and coordinative unsaturation
in facilitating O2 attack at a diiron center and points out a
requirement for the loss of a proton after initial adduct
formation in order for the (µ-1,2-peroxo)diiron(III) adduct
to be trapped.
(23) Herold, S.; Lippard, S. J. J. Am. Chem. Soc. 1997, 119, 145-156.
(24) LeCloux, D. D.; Barrios, A. M.; Lippard, S. J. Bioorg. Med. Chem.
1999, 7, 763-772.
(25) Mizoguchi, T. J.; Kuzelka, J.; Spingler, B.; DuBois, J. L.; Davydov,
R. M.; Hedman, B.; Hodgson, K. O.; Lippard, S. J. Inorg. Chem. 2001,
40, 4662-4673.
Experimental Section
Syntheses. Ligands TPA, 6-Me3-TPA, and BQPA (Scheme 2)
were synthesized according to literature methods.30-32 Commercially
available chemicals were purchased and used without further
(26) MacMurdo, V. L.; Zheng, H.; Que, L., Jr. Inorg. Chem. 2000, 39,
2254-2255.
(27) Kryatov, S. V.; Rybak-Akimova, E. V.; MacMurdo, V. L.; Que, L.,
Jr. Inorg. Chem. 2001, 40, 2220-2228.
(28) Costas, M.; Cady, C. W.; Kryatov, S. V.; Ray, M.; Ryan, M. J.; Rybak-
Akimova, E. V.; Que, L., Jr. Inorg. Chem. 2003, 42, 7519-7530.
(29) Kryatov, S. V.; Chavez, F. A.; Reynolds, A. M.; Rybak-Akimova, E.
V.; Que, L., Jr.; Tolman, W. B. Inorg. Chem. 2004, 43, 2141-2150.
(30) Tyeklar, Z.; Jacobson, R. R.; Wei, N.; Murthy, N. N.; Zubieta, J.;
Karlin, K. D. J. Am. Chem. Soc. 1993, 115, 2677-2689.
(31) Da Mota, M. M.; Rodgers, J.; Nelson, S. M. J. Chem. Soc. A 1969,
2036-2044.
86 Inorganic Chemistry, Vol. 44, No. 1, 2005

FeII2(µ-OH)2 Cores and O2-Adduct Formation
purification. Syntheses and manipulation of iron(II) complexes were
conducted in a glovebox under inert atmosphere. Elemental analyses
were performed at MHW Laboratories (Phoenix, AZ). Caution:
Perchlorate salts are potentially explosiVe and should be handled
with care!
afforded brown crystals suitable for X-ray diffraction. Solid 3(OTf)2
is stable under inert atmosphere at room temperature for a long
time. Anal. Calcd for 3(OTf)2‚2CH3CN or C58H52F6Fe2N10O8S2
(Found): C, 53.30 (53.55); H, 4.01 (4.12); N, 10.72 (10.58). UV-
vis (CH3CN): λmax, nm (ꢀ, M-1 cm-1), 445 (2900), 550 (2100).
[Fe2(µ-OH)2(BnBQA)2(CH3CN)2](ClO4)2, 4(ClO4)2. To a solu-
tion of BnBQA (0.39 g, 1.0 mM) in CH3CN (3 mL) under argon
was added a solution of Fe(ClO4)2‚6H2O (0.36 g, 1.0 mM) in
CH3CN (2 mL) dropwise. The yellow-orange solution immediately
turned dark red. NEt3 (1 equiv, 1 mM, 130 µL) was then added,
and the color changed to dark red-purple. The solution was stirred
for 2 h, layered with ether (5 mL), and left for crystallization to
yield a purple solid. The solid was recrystallized from CH3CN/
ether affording purple crystals suitable for X-ray diffraction. Solid
4(ClO4)2 slowly decomposes at room temperature even under inert
atmosphere and should be kept in a freezer. Anal. Calcd for 4(ClO4)2
or C58H54Cl2Fe2N8O10 (Found): C, 57.78 (57.61); H, 4.51 (4.55);
Cl, 5.88 (5.86); N, 9.29 (9.45). UV-vis (CH3CN): λmax, nm (ꢀ,
M-1 cm-1), 518 (2400).
N-Benzyl-N,N-di(quinolin-2-ylmethyl)amine, BnBQA. 2-
Chloromethylquinoline hydrochloride (3.64 g, 17 mmol), benzyl-
amine (0.86 g, 8.02 mmol), and Na2CO3 (24 g, 280 mmol) dissolved
in water (20 mL) were combined in 200 mL of CH3CN. The
resulting suspension was heated to reflux for 30 h. The mixture
was then filtered, and the filtrate was reduced in volume. After
addition of 30 mL of water to the concentrated filtrate, the product
was extracted by CH2Cl2 (3 × 50 mL) and dried over Na2SO4.
The filtrate was evaporated to yield a yellow oil, to which Et2O
was added. Partial evaporation of the solvent resulted in precipita-
tion of the title compound as yellow crystals (2.3 g, 74%). 1H NMR
(CDCl3): δ (ppm) 8.15 (2H, d), 8.07 (2H, d), 7.79 (2H, m), 7.77
(2H, d), 7.70 (2H, m), 7.52 (2H, m), 7.47 (2H, d), 7.34 (2H, m),
7.24 (1H, m), 4.03 (4H, s), 3.78 (2H, s).
[Fe(BnBQA)(CH3CN)(H2O)](ClO4)2, 5(ClO4)2. BnBQA (0.39
g, 1 mmol) dissolved in CH2Cl2 (2 mL) was added to a vigorously
stirred suspension of Fe(ClO4)2‚6H2O (0.36 g, 1.0 mmol) in CH2Cl2
(3 mL). The mixture was stirred overnight and dried in vacuo to
afford a yellow powder. Recrystallization by layering ether onto a
CH3CN solution affords the product as large pale yellow plates
(0.67 g, 93%). Anal. Calcd for C29H28Cl2FeN4O9 (Found): C, 49.52
(49.82); H, 4.01 (3.94); N, 7.97 (8.35); Cl, 10.08 (10.39).
[Fe3(µ-O)3(TPA)3](OTf)3, 6(OTf)3. Dark gray-green crystals of
this trinuclear complex were obtained by exposure of 1(OTf)2 in
CH3CN to O2 and slow diffusion of diethyl ether into this solution.
Yield: 95%. ES-MS: m/z 1384 (M - CF3SO3)+. Anal. Calcd for
6(OTf)3 or C57H54F9Fe3N12O12S3 (Found): C, 44.63 (43.83); H, 3.55
(3.46); N, 10.96 (11.17). UV-vis (CH3CN): λmax, nm (ꢀ, M-1
cm-1): 490 (320), 530 (350), 595 (250).
[Fe2(µ-OH)2(6-Me3-TPA)2](CF3SO3)2, 1(OTf)2, was synthesized
by a procedure analogous to that published elsewhere for 1(ClO4)2.26
The anaerobic reaction of equimolar amounts of Fe(OTf)2‚2CH3CN,
6-Me3-TPA, and NaOH in methanol, followed by recrystallization
of the crude complex from CH3CN/diethyl ether at 4 °C, afforded
1(OTf)2 as an orange solid in 56% yield. Solid 1(OTf)2 is stable
under inert atmosphere at room temperature. UV-vis (CH3CN):
λ
max, nm (ꢀ, M-1 cm-1), 420 (2400). Anal. Calcd for 1(OTf)2‚
2CH3CN or C48H56F6Fe2N10O8S2 (Found): C, 48.41 (48.57); H,
4.74 (4.87); N, 11.76 (11.59).
[Fe2(µ-OH)2(TPA)2](ClO4)2, 2(ClO4)2. Equimolar amounts of
Fe(ClO4)2‚6H2O (0.36 g, 1.0 mmol) and TPA (0.29 g, 1 mmol)
were mixed together in degassed methanol (10 mL), forming a
yellow solution. Addition of 1 equiv of solid NaOH (0.040 g, 1
mmol) formed a red solution. Complex 2(ClO4)2 could then be
crystallized as a dark-red solid from the solution by layering with
Et2O. Anal. Calcd for 2(ClO4)2 or C36H38Cl2Fe2N8O10 (Found): C,
46.73 (46.49); H, 4.14 (4.04); N, 12.11 (11.86); Cl, 7.66 (7.94).
UV-vis (CH3CN): λmax, nm (ꢀ, M-1 cm-1), 465 (2800).
[Fe2(µ-OH)2(TPA)2](CF3SO3)2, 2(OTf)2. This compound was
prepared as a dark-red solid following the method for 2(ClO4)2 and
substituting Fe(ClO4)2‚6H2O for an equivalent amount of Fe(OTf)2‚
2CH3CN and 6 equiv of H2O. Yield: 57%. Crystals suitable for
the X-ray analysis formed directly from the reaction mixture. Anal.
Calcd for 2(OTf)2, C38H38F6Fe2N8O8S2 (Found): C, 44.55 (44.31);
H, 3.74 (4.09); N, 10.94 (10.82). UV-vis (CH3CN): λmax, nm (ꢀ,
M-1 cm-1), 465 (2800). Due to its better solubility, the triflate salt
2(OTf)2 was used for the kinetic investigation. Solids 2(ClO4)2 and
2(OTf)2 slowly decompose at room temperature even under inert
atmosphere and should be kept in a freezer.
[Fe2(µ-OH)2(BQPA)2](CF3SO3)2, 3(OTf)2. To a solution of
BQPA (0.15 g, 0.385 mmol) in methanol (10 mL) under argon
was added dropwise Fe(OTf)2‚2CH3CN (0.17 g, 0.385 mmol) in
methanol (5 mL). The yellow orange solution immediately turned
red. NaOH (1 equiv, 0.016 g, 0.39 mmol) was then added as solid
(or as a solution in methanol), and the color changed to dark red.
After stirring overnight, the solvent was removed under vacuum.
The resulting dark red-brown solid was dissolved in CH2Cl2, and
the solution was filtered to remove insoluble impurities. The solvent
was removed again, and the red brown powder that was obtained
was dried under vacuum. Recrystallization from CH3CN/ether
Generation of Dioxygen Adducts. A solution of a diiron(II)
complex (1, 3, or 4) was put into a gastight cuvette, cooled to low
temperature, and slowly purged with dioxygen. The progress of
the reactions can be monitored by the development of a deep-green
color. The optimal conditions found for the generation of the O2
adducts were the following: CH2Cl2 solution and T ) -80 °C for
1‚O2; CH2Cl2 solution with 2 equiv of Et3N (vs 3) and T ) -80
°C for 3‚O2; MeCN solution and T ) -40 °C for 4‚O2. UV-vis,
λ
max, nm (ꢀ, M-1 cm-1), 1‚O2 490 (1100), 640 (1100); 3‚O2 480
(1000), 620 (1000); 4‚O2 502 (1300), 650 (1300). The relative yields
of the O2 adducts under other conditions were estimated on the
basis of these extinction coefficients.
Titration of Diiron(II) Complexes with O2. A solution of a
diiron(II) complex (1-4, typically 0.3 mM) was put into a gastight
optical cuvette with a septum and titrated at room temperature with
air-saturated solvent (1.2 mM O2 in CH2Cl2 or 1.7 mM O2 in
MeCN)33-35 using a graduated syringe. The progress of titration
was monitored by visible spectrophotometry.
Physical Methods. Room temperature UV-vis spectra were
acquired on a Hitachi 2000 spectrophotometer. Low temperature
visible spectra were recorded on a Hewlett-Packard 8452 diode array
spectrophotometer using an immersion Dewar equipped with quartz
windows and filled with methanol chilled with liquid N2. 1H NMR
spectra were recorded on a Varian Unity 300 or 500 spectrometer
(33) Battino, R. Oxygen and Ozone; Battino, R., Ed.; Pergamon Press: New
York, 1981; Vol. 8.
(34) Achord, J. M.; Hussey, C. L. Anal. Chem. 1980, 52, 601-602.
(35) Sawyer, D. T.; Chiericato, G.; Angelis, C. T.; Nanni, E. J.; Tsuchiya,
T. Anal. Chem. 1982, 54, 1720-1724.
(32) Wei, N.; Murphy, N. N.; Narasimba, N.; Chen, Q.; Zubieta, J.; Karlin,
K. D. Inorg. Chem. 1994, 33, 1953-1965.
Inorganic Chemistry, Vol. 44, No. 1, 2005 87

Kryatov et al.
Table 1. Crystallographic Data
1(OTf)2‚CH3CN
(6-Me3-TPA)
2(ClO4)2‚2CH3OH
2(OTf)2
(TPA)
3(OTf)2‚2CH3CN
4(ClO4)2
(BnBQA)
(TPA)
(BQPA)
formula
fw, amu
cryst habit, color
cryst syst
space group
a, Å
C46H53F6Fe2N9O8S2
1149.79
block, orange
triclinic
P1h
13.128(1)
13.816(2)
15.971(2)
72.905(2)
68.213(2)
76.613(2)
2546.4(5)
2
C38H46Cl2Fe2N8O12
989.43
plate, orange
monoclinic
P21/n
9.9451(3)
13.7240(3)
16.5473(3)
90
104.012(1)
90
2191.3(1)
2
1.500
173(2)
0.853
C38H38F6Fe2N8O8S2
1024.58
prism, orange
monoclinic
P21/n
11.6706(11)
12.1665(11)
15.8301(14)
90
108.096(2)
90
2136.5(3)
2
C58H52F6Fe2N10O8S2
1306.92
cube, red-black
monoclinic
P21/c
11.6282(6)
18.890(1)
13.3536(8)
90
98.956(1)
90
2897.4(3)
2
C58H54Cl2Fe2N8O10
1205.69
block, red
orthorhombic
Pbca
13.518(8)
18.87(1)
21.64(1)
90
90
90
5519(6)
4
1.451
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
D(calc), Mg m-3
T, K
1.500
173(2)
0.734
0.0489
1.593
173(2)
0.864
0.0516
1.498
173(2)
0.656
0.0413
173(2)
1.451
0.0587
0.1066
µ, mm-1
a
R1
0.0356
0.0737
b
wR2
0.1264
0.1361
0.1029
a R ) (∑|(Fï - Fc)|)/(∑|Fï|). b Rw ) (∑[w(Fï - Fc )2]/(∑[w(Fï4])1/2, where w ) q/σ2(Fo ) + (aP)2 + bP.
2
2
2
at ambient temperature. Chemical shifts (in ppm) were referenced
to the residual protic solvent peaks.
was introduced into the solvent. Ferrocene was used as an internal
standard.38 In most experiments, the range of potentials from ca.
-1.6 to +0.4 V versus Fc+/Fc was scanned with a rate of 100 mV
Crystallographic Studies. Crystals of the complexes 1(OTf)2‚
CH3CN, 2(ClO4)2‚2CH3OH, 2(OTf)2, 3(OTf)2‚2CH3CN, 4(ClO4)2,
and 6(OTf)3‚2CH3CN were mounted onto glass fibers, and data
were collected at -100 °C on a Siemens SMART system equipped
with graphite-monochromated Mo KR (λ ) 0.71073 Å) radiation
and a CCD detector (or Bruker P4 SMART CCD system for
2(OTf)2). An initial set of cell constants was calculated from
reflections harvested from three sets of 20-30 frames. These initial
sets of frames were oriented such that orthogonal wedges of
reciprocal space were surveyed. Final cell constants for all six
compounds were calculated from sets of 3560, 6719, 9587, 2999,
3525, and 3570 strong reflections from the actual data collections
for 1(OTf)2‚CH3CN, 2(ClO4)2‚2CH3OH, 2(OTf)2, 3(OTf)2‚2CH3CN,
4(ClO4)2, and 6(OTf)3‚2CH3CN, respectively. All calculations were
performed using an SGI INDY R4400-SC or Pentium computer
with the SHELXTL V5.0 program suite.36,37 Pertinent crystal-
lographic data and experimental conditions are summarized in Table
1. All six structures were solved by direct methods. All non-
hydrogen atoms were refined anisotropically, and hydrogen atoms
were placed in ideal positions and refined as riding atoms with
individual (or group if appropriate) isotropic displacement param-
eters. For 2(ClO4)2‚2CH3OH both the hydroxide proton and the
solvate methanol O-H proton were located from difference maps
and were refined isotropically. Because the perchlorate anion was
disordered over two positions, bond length constraints (SADI) were
applied to the Cl-O bonds. Hydrogen atoms bonded to the bridging
oxygen atoms were also located on the difference Fourier map for
2(OTf)2 and refined isotropically.
min-1
.
ESI Mass Spectrometry. Products of oxygenation of complexes
1-4 at room temperature were identified by electrospray mass
spectrometry using a Finnigan LTQ or a Finnigan-MAT (San Jose,
CA) LCQ ion trap mass spectrometer. Acetonitrile solutions of 1-4
(0.1 mM) were exposed to air and subjected to analysis immediately
upon a change of color to pale-yellow, characteristic of oxo-bridged
oligomeric iron(III) complexes. Assignment of the major peaks in
the mass spectra was confirmed by the comparison of observed
values of m/z and isotopic satellite patterns with those calculated
by the IsoPro 3.0 software.
Kinetic Measurements. The kinetics of the oxygenation of
complexes 1-4 was studied using a Hi-Tech Scientific (Salisbury,
Wiltshire, U.K.) SF-43 multimixing anaerobic cryogenic stopped-
flow instrument combined with either a monochromator (low
intensity light irradiation of the sample) or a diode array rapid
scanning unit (strong UV-vis irradiation of the sample). No
difference in spectral and kinetic data was observed for the
monochromator and diode array modes in this study, indicating
the absence of photosensitivity. All manipulations with complexes
1-4 and their solutions were done using an argon atmosphere
glovebox, airtight syringes, and the anaerobic stopped-flow instru-
ment to avoid contamination with air. Saturated solutions of O2 in
CH2Cl2 and CH3CN were prepared by bubbling the dry O2 gas for
20 min in a septum-closed cylinder with the solvent at a constant
temperature (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 acetonitrile at
25 °C.33-35 Solutions with lower O2 concentrations were prepared
from graduated mixtures of oxygen and nitrogen gases assuming
the linear dependency of solubility on the partial pressure of gas.
The solutions of a diiron(II) complex and O2 were cooled to a preset
temperature in the stopped-flow instrument before mixing. The
concentrations of the reactants were corrected for the 1:1 mixing
ratio and the temperature contraction of solvent (Tables S1 and
S2). The densities of CH3CN and CH2Cl2 at different temperatures
were determined experimentally as described elsewhere.27 The
density of CH3CN at temperatures from -40 to +25 °C could be
Electrochemistry. Cyclic voltammetry measurements were
carried out in a Vacuum Atmosphere glovebox with an argon
atmosphere using an EG&G PAR 273 potentiostat and a three-
electrode cell (pyrolytic carbon working electrode, platinum wire
auxiliary electrode, and silver pseudoreference electrode). Dry
acetonitrile or dichloromethane with 0.1 M n-Bu4NPF6 supporting
electrolyte was used as solvent for complexes 1-4 (ca. 2 mM). In
additional experiments, a small amount of triethylamine (ca. 4 mM)
(36) SHELXTL 5.10 (PC/NT-Version). Program library for Structure
Solution and Molecular Graphics; Bruker Analytical X-ray Systems:
Madison, WI, 1998.
(37) Sheldrick, G. M. SHELXL-97. Program for the Refinement of Crystal
Structure; University of Gottingen: Gottingen, Germany, 1997.
(38) Gagne´, R. R.; Koval, C. A.; Lisensky, G. C. Inorg. Chem. 1980, 19,
2854-2855.
88 Inorganic Chemistry, Vol. 44, No. 1, 2005

FeII2(µ-OH)2 Cores and O2-Adduct Formation
satisfactorily described by a linear equation: dT (g/cm3) ) 0.805(5)
- 0.00108(2) × T (T in °C). The density of CH2Cl2 could be
satisfactorily described by the following linear equation: dT (g/
cm3) ) 1.370(5) - 0.00180(2) × T (T in °C).27
For the kinetic experiments, dioxygen was always taken in large
excess so that its concentration does not change significantly during
the reaction with a diiron(II) complex. The kinetic traces obtained
under the pseudo-first-order conditions over 3-5 half-lives were
fit to eq 1.
At ) A∞ - (A∞ - Ao)‚exp(-kobst)
(1)
The fit of experimental data to eq 1 and the independence of
kobs on [Fe2] served as proof of the first-order behavior of the
reactions in the diiron(II) complex. A linear dependence of kobs on
[O2] served as proof of the first-order behavior of the reactions in
dioxygen. Equation 2 was used to calculate the second-order rate
constant k.
Figure 1. UV-vis spectra of the diiron(II) complexes 1 (O), 2 (]), 3
(0), and 4 (4) in acetonitrile solution.
k ) kobs/[O2]
(2)
(3)
charge-transfer transitions. The absorbance maximum of 1
at 420 nm is observed to red-shift to 465 nm in 2, consistent
with the expected decrease in the Lewis acidity of the metal
center upon replacement of 6-Me3-TPA with TPA.39 The
absorbance maxima of 3 and 4 are even more red-shifted
due to the presence of the quinoline ligands that have lower
energy π-acceptor orbitals for the MLCT transitions. Inter-
estingly, 3, a complex with mixed pyridine and quinoline
ligation, exhibits a spectrum with two absorbance features
at 440 and 560 nm; these features may be assigned,
respectively, to the iron(II)-to-pyridine and the iron(II)-to-
quinoline charge-transfer transitions.
-d[Fe2]/dt ) kobs[Fe2] ) k[O2][Fe2]
The values of k (M-1 s-1) were determined at different
temperatures and fit to the Eyring eq 4 to obtain the activation
parameters ∆Hq and ∆Sq.
ln(k/T) ) 23.76 + ∆Sq/R - ∆Hq/RT
(4)
In some cases, the formation of an oxygen adduct from the
diiron(II) precursor and O2 was closely followed by the decomposi-
tion of the adduct. Then the kinetic traces were fit to a two-
exponential eq 5.
At ) A∞ + ∆A1‚exp(-kobst) + ∆A2‚exp(-kdecompt)
(5)
Structural Characterization of the Complexes with
FeII (OH)2 Diamond Cores. Single crystals of 1(OTf)2‚
The values of kobs (corresponding to the formation of the oxygen
adduct) were treated as described above. Kinetic data are sum-
marized in Tables S3-S18 and Figures S1-S26.
2
CH3CN, 2(OTf)2, 2(ClO4)2‚2CH3OH, 3(OTf)2‚2CH3CN, and
4(ClO4)2 suitable for X-ray diffraction studies were grown
from CH3CN/Et2O or CH3OH/Et2O. Figure 2 shows the
structures of the complex cations in 1-4, and Table 2 lists
selected bond lengths and angles. The structure of complex
1 as the perchlorate salt 1(ClO4)2 has been published26 and
is very similar to that found for the triflate salt 1(OTf)2‚
CH3CN reported here. Likewise, the geometric parameters
of complex 2 are very close in the crystals of 2(OTf)2 and
2(ClO4)2‚2CH3OH. Examination of the structures reveals that
Results
Diiron(II) Precursors. Synthesis and Solution Charac-
terization. The reaction of iron(II) triflate or perchlorate with
an equimolar amount of a tetradentate aminopyridine ligand
L (L ) 6-Me3-TPA, TPA, or BQPA) and an equimolar
amount of NaOH in methanol afforded complexes [FeII (µ-
2
OH)2L2]2+ (1-3). Using a similar procedure, with the
tridentate ligand BnBQA in acetonitrile and Et3N as base,
gave the complex [Fe2(µ-OH)2(BnBQA)2(CH3CN)2]2+ (4).
These four diiron(II) complexes were characterized by X-ray
crystallography as triflate or perchlorate salts (vide infra).
Complex 4(ClO4)2 had limited stability in the solid state even
under anaerobic conditions. It was then found that the
corresponding mononuclear complex [Fe(BnBQA)(CH3CN)-
(H2O)](ClO4)2, 5(ClO4)2, obtained from a combination of
Fe(ClO4)2‚6H2O and BnBQA in CH2Cl2/CH3CN by omitting
the addition of base, was a convenient and stable precursor
to 4, which could then easily be prepared in situ in nearly
quantitative yields (as determined by UV-vis spectropho-
tometry) by treatment of a CH3CN solution of 5 with 1 equiv
of triethylamine.
1-4 all contain a centrosymmetric FeII (µ-OH)2 diamond
2
core (Scheme 3). Four nitrogen atoms of the tetradentate
ligand in 1-3 or the tridentate ligand plus CH3CN in 4
complete the distorted octahedral coordination environment
about each iron center. The FeII (µ-OH)2 diamond cores have
2
C2h symmetry with asymmetric Fe-O(H)‚‚‚Fe units; a
similar asymmetry is observed in related diiron(II) complexes
reported by Lippard and co-workers having a single hydroxo
or methoxo bridge, [Fe2(BPMAN)(µ-OH or OMe)(µ-O2CR)]-
(OTf)2.40,41 The asymmetry appears to be determined by the
nature of the trans ligand; the shorter FeII-µ-OH bond
(39) Zang, Y.; Kim, J.; Dong, Y.; Wilkinson, E. C.; Appelman, E. H.; Que,
L., Jr. J. Am. Chem. Soc. 1997, 119, 4197-4205.
(40) He, C.; Lippard, S. J. Inorg. Chem. 2001, 40, 1414-1420.
(41) Kuzelka, J.; Mukhopadhyay, S.; Spingler, B.; Lippard, S. J. Inorg.
Chem. 2003, 42, 6447-6457.
Figure 1 shows the visible spectra of complexes 1-4,
which are dominated by iron(II)-to-pyridine (or quinoline)
Inorganic Chemistry, Vol. 44, No. 1, 2005 89

Kryatov et al.
Figure 2. ORTEP representations of the X-ray crystal structures of the bis(µ-hydroxo)diiron(II) complex cations 1-4 showing 50% probability thermal
ellipsoids. Hydrogen atoms are omitted for clarity.
Table 2. Selected Bond Lengths (Å) and Bond Angles (deg) for Bis(µ-hydroxo)diiron(II) Complexes 1-4a
b
1(ClO4)2
1(OTf)2‚CH3CNc
2(ClO4)2‚2CH3OH
2(OTf)2
3(OTf)2‚2CH3CN
4(ClO4)2
BnBQA
ligand
6-Me3-TPA
6-Me3-TPA
TPA
TPA
BQPA
Fe-O
1.973(2) (O1)
2.168(2) (O1*)
2.222(2) (N1)
2.285(2) (N3)
1.995(2)/2.032(2) (O1)
2.182(2)/2.208(2) (O1*)
2.230(3)/2.213(3) (N1)
2.333(3)/2.307(3) (N3)
2.0059(15) (O1*)
2.1595(16) (O1)
2.253(2) (N1)
2.191(2) (N4)
1.9726(17) (O1)
2.1344(18) (O1*)
2.264(2) (N1)
1.9837(14) (O1)
2.1007(15) (O1*)
2.2421(18) (N1)
2.1698(18) (N3)
1.985(2) (O1)
2.090(2) (O1*)
2.254(3) (N1)
Fe-Nam (in-plane)
Fe-Npy (in-plane)
Fe-NMeCN (in-plane)
Fe-Npy (out-of-plane)
2.189(2) (N3)
2.290(3) (N4)
2.252(3) (N2)
2.254(3) (N3)
3.135
100.54(11)
79.46(11)
2.326(3) (N2)
2.290(3) (N4)
3.187
100.55(9)
79.45(9)
2.306(3)/2.336(3) (N2)
2.282(3)/2.287(3) (N4)
3.175/3.221
98.82(10)/98.75(10)
81.18(10)/81.25(10)
2.226(2) (N2)
2.235(2) (N3)
3.087
95.57(6)
84.43(6)
2.205(2) (N2)
2.192(2) (N4)
3.010
94.20(7)
85.80(7)
2.3057(19) (N2)
2.2880(19) (N4)
3.086
98.11(6)
81.89(6)
Fe‚‚‚Fe*
Fe-O-Fe*
O-Fe-O *
a Diiron core is depicted in Scheme 2; ORTEP views of complex cations are shown in Figure 2. b Data from ref 26. c There are two crystallographically
independent dimers in the formula unit, both with inversion centers at their molecular centers.
Scheme 3. Coordination Environment of the Iron(II) Centers in
Complexes 1-4
(1.97-2.01 Å in 1-4) is associated with the hydroxide trans
to the tertiary amine, while the longer FeII-(µ-OH) bond
(2.09-2.21 Å in 1-4) is associated with the hydroxide trans
to the pyridine or CH3CN ligand. Indeed, in the FeII (µ-OH)2
2
complexes with capping ligands having identical donor atoms
the FeII-(µ-OH) bonds are almost equivalent as in [Fe2(µ-
OH)2(κ2-PhTptBu)2] (1.974 and 1.978 Å),42 [Fe2(µ-OH)2(κ3-
43
PhTpiPr )2] (2.016 and 2.04 Å), and [Fe2(µ-OH)2(κ4-
2
(42) Kisko, J. L.; Hascall, T.; Parkin, G. J. Am. Chem. Soc. 1998, 120,
10561-10562.
BPMCN)2] (2.05 and 2.08 Å).44 The shorter FeII-(µ-OH)
90 Inorganic Chemistry, Vol. 44, No. 1, 2005

FeII2(µ-OH)2 Cores and O2-Adduct Formation
Table 3. FeIIIFeII/FeIIFeII Redox Potentials (mV vs Fc+/Fc0) for
Bis(µ-hydroxo)diiron Complexes
diiron(II,II) species
E1/2 (∆E) reversibility
ref
1, [Fe2(OH)2(6-Me3-TPA)2]2+
(in CH2Cl2)
-150 (95)
-100a
QR
IRR
QR
IRR
QR
IRR
IRR
IRR
QR
IRR
QR
IRR
QR
IRR
R
this work
1, [Fe2(OH)2(6-Me3-TPA)2]2+
(in CH2Cl2/Et3N)
this work
this work
this work
this work
this work
this work
this work
this work
this work
this work
this work
this work
this work
44
1, [Fe2(OH)2(6-Me3-TPA)2]2+
(in MeCN)
-160 (100)
-120a
1, [Fe2(OH)2(6-Me3-TPA)2]2+
(in MeCN/Et3N)
2, [Fe2(OH)2(TPA)2]2+
(in CH2Cl2)
-460 (65)
-430a
2, [Fe2(OH)2(TPA)2]2+
(in CH2Cl2/Et3N)
2, [Fe2(OH)2(TPA)2]2+
(in MeCN)
-450a
Figure 3. Cyclic voltammogram of 2 mM [FeII2(µ-OH)2(BQPA)2](OTf)2
(3) in MeCN with 0.1 M [n-Bu4N][PF6] as a supporting electrolyte (scan
rate 100 mV s-1).
2, [Fe2(OH)2(TPA)2]2+
(in MeCN/Et3N)
-450a
3, [Fe2(OH)2(BQPA)2]2+
(in CH2Cl2)
-370 (70)
-330a
bonds in complexes 1-4 (1.97-2.01 Å) are comparable to
those in other known (µ-hydroxo)diiron(II) complexes
(1.95-2.08 Å),40,42-45 but much longer than that for an
iron(II) complex with a terminal hydroxide, [Fe(OH)(κ3-
PhTptBuiPr)] (1.83 Å).46 The longer FeII-(µ-OH) bonds in 1-4
(2.09-2.21 Å) are in fact among the longest of such bonds.
As noted previously, the coordination environments of the
iron(II) centers are completed by the tetradentate tripodal
ligand in 1-3 and by the meridional tridentate ligand and
CH3CN in 4. In 3 and 4, the quinoline ligands occupy the
axial positions perpendicular to the FeN2O2 plane. The Fe-
Npy bond lengths observed (2.17-2.34 Å) are consistent with
the complexes having high-spin iron(II) ions.39 The variations
in length reflect the presence or absence of an R substituent,
which introduces a steric impediment that leads to the
lengthening of the Fe-Npy bond by an average of 0.09 Å.39
3, [Fe2(OH)2(BQPA)2]2+
(in CH2Cl2/Et3N)
3, [Fe2(OH)2(BQPA)2]2+
(in MeCN)
-340 (65)
-310a
3, [Fe2(OH)2(BQPA)2]2+
(in MeCN/Et3N)
4, [Fe2(OH)2(BnBQA)2(MeCN)2]2+ -310 (95)
(in MeCN)
4, [Fe2(OH)2(BnBQA)2(MeCN)2]2+ -280a
(in MeCN/Et3N)
[Fe2(OH)2(BPMCN)2]2+
(in CH2Cl2)
-360
O2/O2- (in MeCN)b
O2/O2- (in CH2Cl2)b
-1250
-1265
R
R
35, 56, 57
58, 59
-
a Anodic wave only. b The O2/O2 potential is given for comparison.
Cyclic voltammetry experiments show that the diiron(II)
complexes 1-4 are electrochemically active and can be
oxidized at potentials of -460 to -150 mV versus Fc+/Fc
(Figure 3, Table 3). The E1/2 values for 1-4 are similar to
the FeIIFeIII/FeIIFeII potential reported for the closely related
The dimensions of the FeII (µ-OH)2 diamond core vary little
2
among the four complexes, as indicated by the small range
of Fe-Fe distances (3.010-3.187 Å) and the Fe-O-Fe
angles (94.20-100.55°). All the hydroxo bridges in 1-4 are
hydrogen bonded to oxygen atoms of corresponding anions
with O‚‚‚O distances in the range of 2.77-3.15 Å.
Electrochemistry. Cyclic voltammetry (CV) experiments
on complexes 1-4 were performed in dichloromethane and
acetonitrile solutions (with 0.1 M n-Bu4NPF6) under inert
atmosphere (Ar), except for 4, which decomposes in CH2Cl2.
In additional CV experiments, a small amount of triethyl-
amine (2 equiv with respect to 1-4) was introduced into
the freshly prepared solutions. The range of potentials from
-1.6 to +0.4 V (vs Fc+/Fc) was scanned to look for the
FeIIFeIII/FeIIFeII redox couple, because for all previously
reported diiron complexes with N/O-donor atoms such a
couple had been found within the range (Tables 3 and
S19).40,41,44,45,47-59
diiron complex [FeII (µ-OH)2(BPMCN)2]2+ (-360 mV vs
2
Fc+/Fc), whose one-electron oxidation product [FeIIFeIII(µ-
OH)2(BPMCN)2]3+ has been isolated and structurally char-
acterized.44 The electrochemical oxidations are quasirevers-
ible in CH2Cl2 and MeCN and essentially solvent-independent,
except for 2, where the simple change of solvent from
CH2Cl2 to MeCN caused the loss of reversibility. Apparently,
the oxidized FeIIFeIII form of 2 is unstable in MeCN. A
byproduct, the mononuclear complex [Fe(TPA)(MeCN)2]2+,
showed its prominent redox feature (a reversible wave at
+750 mV vs Fc+/Fc) after repeated electrochemical cycling
of 2 in MeCN. A similar instability in MeCN has been
observed for the mixed valence complex [FeIIFeIII(OH)2-
(BPMCN)2]3+, which could be isolated and crystallized only
from CH2Cl2.44 In these cases, the formation of highly stable
low-spin mononuclear iron(II) complexes may serve as a
thermodynamic sink that destroys the mixed-valent FeIIFeIII
species and leads to irreversibility of the diiron redox
couples.45,60,61 It should be noted that the TPA complex is
(43) Kitajima, N.; Tamura, N.; Tanaka, M.; Moro-oka, Y. Inorg. Chem.
1992, 31, 3342-3343.
(44) Stubna, A.; Jo, D.-H.; Costas, M.; Brenessel, W. W.; Andres, H.;
Bominaar, E. L.; Mu¨nck, E.; Que, L., Jr. Inorg. Chem. 2004, 43, 3067-
3079.
(45) 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.
(46) Hikichi, S.; Ogihara, T.; Fujisawa, K.; Kitajima, N.; Akita, M.; Moro-
oka, Y. Inorg. Chem. 1997, 36, 4539-4547.
(47) Stassinopoulos, A.; Schulte, G.; Papaefthymiou, G. C.; Caradonna, J.
P. J. Am. Chem. Soc. 1991, 113, 8686-8697.
(48) Mukerjee, S.; Stassinopoulos, A.; Caradonna, J. P. J. Am. Chem. Soc.
1997, 119, 8097-8098.
(49) Suzuki, M.; Uehara, A.; Oshio, H.; Endo, K.; Yanaga, M.; Kida, S.;
Saito, K. Bull. Chem. Soc. Jpn. 1987, 60, 3547-3555.
Inorganic Chemistry, Vol. 44, No. 1, 2005 91

Kryatov et al.
unique in its properties in the series of related mononuclear
iron(II) species, [Fe(L)(MeCN)n]2+ (L ) TPA, BQPA,
6-Me3-TPA, BnBQA). Only the TPA complex is low-
spin39,62,63 and may therefore possess exceptional stability.
quasireversibility is the deprotonation of the OH-ligand upon
oxidation to the FeIIFeIII state.45 In our experiments, addition
of a small amount of Et3N to the solvent (either CH2Cl2 or
CH3CN) did not shift the anodic peak of complexes 1-4,
but it caused the cathodic counter peaks to disappear.
Apparently, triethylamine does not remove protons from the
FeIIFeII complexes but may deprotonate the FeIIFeIII species
leading to the disintegration of the oxidized complexes. As
a result, the FeIIFeIII/FeIIFeII couples lose their reversibility.
Extending the anodic range of potentials in CH2Cl2 solutions
of 1-3 revealed an additional irreversible wave at +520 mV
versus Fc+/Fc, attributed to the oxidation of Et3N.
The FeIIFeIII/FeIIFeII redox potentials increase in the series
2 < 3 ∼ 4 < 1, which may be explained by the weakening
donor properties of the N-ligands with the introduction of
substituents at the pyridine 6-position. Indeed, BQPA is
derived from TPA by adding two electron-withdrawing benzo
groups. Ligand BnBQA contains the same two additional
benzo groups and lacks one pyridine donor, which is
substituted for a weaker donor, MeCN, in complex 4. Finally,
ligand 6-Me3-TPA contains three 6-methyl substituents that
interfere sterically with the coordination of the pyridine
N-atoms.39 There is a positive correlation between the
average Fe-N bond lengths in complexes 1-4 and their
oxidation potentials (Figure S27), indicating that longer (and
weaker) Fe-N bonds destabilize the higher oxidation state
of the diiron complexes. The connection between long
metal-ligand bonds and high redox potentials has also been
noted for (hexaamine)cobalt(II/III) complexes64 and may be
true in general for complexes with predominantly σ-donor
ligands.65,66
As mentioned above, incorporating two hydroxo bridges
into diiron complexes decreases the FeIIFeIII/FeIIFeII redox
potentials and makes diiron(II) complexes 1-4 more sus-
ceptible to oxidation. Comparing these redox potentials
(-460 to -150 mV vs Fc+/Fc) to the much lower potential
for a one-electron reduction of O2 into superoxide (ca. -1250
mV vs Fc+/Fc in MeCN or CH2Cl2, Table 335,56-59) shows,
however, that the outer-sphere one-electron oxidation of 1-4
with O2 in these solvents is a thermodynamically uphill
process. Moreover, the redox potentials for other known
diiron(II,II) complexes (Table S19) show that none of them
can be an effective outer-sphere reducing agent for molecular
oxygen in aprotic organic solvents. Coordination of the O2
molecule to an iron center with subsequent inner-sphere
electron transfer events (eventually yielding a coordinated
peroxide) is expected to be the likely mechanism for the
oxygenation of complexes 1-4 and related species.
Reactions with O2. Overview. Table 4 summarizes the
results from the reactions of 1-4 with O2 in different
solvents; a detailed discussion for each complex follows in
the subsections below. Complexes 1, 3, and 4 (but not 2)
formed O2 adducts in nearly quantitative yield at low
temperature under the appropriate conditions. The adducts
exhibit two comparably intense absorbance maxima near 450
The half-wave potentials for the one-electron oxidation
of complexes 1-4 fall in the middle of the range reported
in the literature for the FeIIFeIII/FeIIFeII redox pairs (Tables
3 and S19), which vary significantly depending on the charge,
donor properties, and topology of the ligands. The lowest
potentials (about -1000 mV vs Fc+/Fc) were found for com-
plexes [Fe2(H2Hbab)2(DMF)2(N-MeIm)]0 and [Fe2(H2bamb)2-
(N-MeIm)2]0 with two negatively charged and strongly
donating phenolate groups on each tetradentate ligand.47,48
At the other limit, a relatively high potential of about +300
mV versus Fc+/Fc was observed for complex [Fe2(BPMAN)-
(O2CR)2]2+ with a neutral polydentate aminopyridine ligand
BPMAN and weakly donating bridging carboxylates.40,41
Hydroxide bridges modulate the redox properties of
dinuclear iron complexes. For example, the substitution of
a carboxylate ligand in [Fe2(BPMAN)(O2CPhCy)2]2+ with
a better-donating hydroxide caused the shift of the FeIIFeIII/
FeIIFeII potential from +296 mV versus Fc+/Fc in [Fe2-
(BPMAN)(O2CPhCy)2]2+ to -22 mV versus Fc+/Fc in
[Fe2(BPMAN)(O2CPhCy)(OH)]2+.40 As expected, even more
negative potentials (ranging from -150 to -460 mV) were
found in our work for complexes 1-4 that contain tetra-
dentate aminopyridine ligands similar to BPMAN but have
two hydroxo-bridges between two iron(II) centers. The effect
is especially dramatic if one compares an FeIIIL/FeIIL
potential for a monomeric complex [Fe(TPA)(MeCN)2]2+
(+750 mV) to an FeIIIFeII/FeIIFeII potential for the corre-
sponding bis-hydroxo bridged complex 2 (-450 mV).
Although reversible oxidation and stable mixed-
valent(II,III) species were observed for several diiron(II,II)
complexes of dinucleating ligands with bridging car-
boxylates,40,49-53 introduction of the bridging OH-ligands into
the diiron(II,II) complexes usually led to quasireversible
electrochemical oxidation.45,49,50 A possible reason for the
(50) Suzuki, M.; Oshio, H.; Uehara, A.; Endo, K.; Yanaga, M.; Kida, S.;
Saito, K. Bull. Chem. Soc. Jpn. 1988, 61, 3907-3913.
(51) Suzuki, M.; Fujinami, S.; Hibino, T.; Hori, H.; Maeda, Y.; Uehara,
A.; Suzuki, M. Inorg. Chim. Acta 1998, 283, 124-135.
(52) Mashuta, M. S.; Webb, R. J.; Oberhausen, K. J.; Richardson, J. F.;
Buchanan, R. M.; Hendrickson, D. N. J. Am. Chem. Soc. 1989, 111,
2745-2746.
(53) Borovik, A. S.; Papaefthymiou, V.; Taylor, L. F.; Anderson, O. P.;
Que, L., Jr. J. Am. Chem. Soc. 1989, 111, 6183-6195.
(54) Buchanan, R. M.; Mashuta, M. S.; Richardson, J. F.; Oberhausen, K.
J.; Hendrickson, D. N.; Webb, R. J.; Nanny, M. A. Inorg. Chem. 1990,
29 (9), 1299-1301.
(55) Lee, D.; Krebs, C.; Huynh, B. H.; Hendrich, M. P.; Lippard, S. J. J.
Am. Chem. Soc. 2000, 122, 5000-5001.
(56) Ohsaka, T.; Tsushima, M.; Tokuda, K. Bioelectrochem. Bioenerg.
1993, 31, 289-300.
(57) Vasudevan, D.; Wendt, H. J. Electroanal. Chem. 1995, 192, 69-74.
(58) Peover, M. E.; White, B. S. J. Chem. Soc., Chem. Commun. 1965,
183-184.
(59) Ruiz, J.; Astruc, D. C. R. Acad. Sci., Ser. IIc: Chim. 1998, 1, 21-27.
(60) Armstrong, W. H.; Spool, A.; Papaefthymiou, G. C.; Frankel, R. B.;
Lippard, S. J. J. Am. Chem. Soc. 1984, 106, 3653-3667.
(61) Costas, M.; Que, L., Jr. Angew. Chem., Int. Ed. 2002, 41, 2179-
2181.
(64) Comba, P.; Sickmu¨ller, A. F. Inorg. Chem. 1997, 36, 4500-4507.
(65) Gutmann, V. Struct. Bonding 1973, 15, 141-166.
(66) Lever, A. B. P. Inorg. Chem. 1990, 29, 1271-1285.
(62) Diebold, A.; Hagen, K. S. Inorg. Chem. 1998, 37, 215-223.
(63) Chen, K.; Que, L., Jr. J. Am. Chem. Soc. 2001, 123, 6327-6337.
92 Inorganic Chemistry, Vol. 44, No. 1, 2005

FeII2(µ-OH)2 Cores and O2-Adduct Formation
Table 4. Properties of the Diiron(II) Peroxo Intermediates Derived from the Oxygenation of Complexes 1-4 in Different Solvents
UV-vis λmax
,
resonance Raman
oxidation
productsb
complex
solvent (temp)
yield, %a
nm (ꢀ, M-1 cm-1
)
ν(Fe-O), ν(O-O)/ cm-1
1 (L ) 6-Me3-TPA)
CH2Cl2 (<-50 °C)
CH3CN (-40 °C)
∼100%
490 (1100), 640 (1100)
n/a
465, 844c
462, 848
470, 843
[FeIII2(O)2(L)2]2+ and
[FeIII2(O)(OH)(L)2]3+
[FeIII3(O)3(L)2]3+
25%
2 (L ) TPA)
CH2Cl2 (<-50 °C)
CH3CN (-40 °C)
not observed
3 (L ) BQPA)
CH2Cl2 (<-50 °C)
CH2Cl2/NEt3d (<-50 °C)
CH3CN (-40 °C)
not observed
∼100%
35%
70%
∼100%
480 (1000), 620 (1000)
[FeIII2(O)2(L)2]2+ and
[FeIII2(O)(OH)(L)2]3+
CH3CN/NEt3d (-40 °C)
CH3CN (-40 °C)
4 (L ) BnBQA)
502 (1300), 650 (1300)
[FeIII2O2(BnBQA)2]2+ and
[BnBQA‚H]+
a The yields were estimated on the basis of extinction coefficients obtained from visible spectral data on samples evaluated by Mo¨ssbauer spectroscopy.
b Products obtained upon warming of the peroxo intermediates to room temperature or directly upon oxygenation of diiron(II) complexes at room temperature.
c From ref 26. d 2 mol of triethylamine per mol of 3 is required; NEt3 may be substituted for other noncoordinating amines.
The peroxo intermediates, however, were thermally un-
stable and yielded oxo-bridged oligoiron(III) complexes upon
warming. The same products were obtained if oxygenation
was carried out at room temperature. The final products from
the reaction of 1 or 3 and dioxygen can be identified as
[FeIII2(µ-O)2(L)2]2+ complexes, which have been previously
characterized in detail.69 The dications were readily observed
by electrospray mass spectral analyses of the reaction
solutions (Table 4, Figures S28-29). In the case of 3, an
additional feature was sometimes observed at m/z 463,
corresponding to [FeII(BQPA)(OH)]+. This feature is also
present in the ES-MS spectrum of [FeIII2(O)(OH)(BQPA)2]3+
and is attributed to a reduction product formed under the
ionization conditions during mass spectrometry. UV-vis
titration experiments (Figures S5 and S12) indicate a 1:2
stoichiometry for the oxygenation reactions leading to the
Figure 4. Visible spectra of the peroxo diiron(III) complexes 1‚O2 (O),
3‚O2 (0), and 4‚O2 (4).
oxo-bridged diiron(III) complexes (eq 7).
2[FeII (µ-OH)2Ln]2+ + O2 f 2[FeIII2(µ-O)2Ln]2+ + 2H2O
2
and 650 nm (Figure 4) that have been previously associated
with (µ-oxo)(µ-1,2-peroxo)diiron(III) complexes.26,67,68 The
(7)
presence of a µ-1,2-peroxo ligand in 3‚O2 and 4‚O2 was also
confirmed by the observation of characteristic Fe-O and
O-O vibrations by resonance Raman spectroscopy (Table
4), as previously detailed for 1‚O2.26 Of particular interest
is the observation that 3 gave rise to a peroxo intermediate
in CH2Cl2 only in the presence of added base (e.g.,
triethylamine). Formation of the peroxo complexes proceeds
according to the generalized eq 6.
Because water is formed as a byproduct in reaction 7,
protonation of the rather basic FeIII2(µ-O)2 core can occur to
afford its conjugate acid.69 Indeed, scrutiny of the visible
spectra of the product solutions from the oxidations of 1 and
3 indicates the formation of some fraction of the conjugate
acid, as evidenced by significant absorbance at 550 nm that
is characteristic of complexes with the FeIII2(µ-O)(µ-OH)
core.69
The reaction of 4 with O2 also affords [FeIII O2(BnBQA)2]2+,
2
[FeII (µ-OH)2Ln]2+ + O2 f [Fe2(µ-O)(µ-O2)Ln]2+ + H2O
which can be observed as a prominent feature in the
electrospray mass spectrum of the product solution (Figure
S30). Also observed was a feature corresponding to
[FeII(BnBQA)OH]+. A third prominent feature can be
assigned to [BnBQA‚H]+, suggesting the facile breakdown
of the FeIII2(µ-O)2 complex. The [FeIII2O2(BnBQA)2]2+
complex has not been isolated and may be less stable because
of the tridentate nature of the ligand, which cannot protect
the FeIII2(µ-O)2 core as well as the sterically more hindered
tetradentate 6-Me3-TPA and BQPA ligands.
2
(6)
The accumulation of peroxo intermediates suggests that
these species are chemically fairly stable at low temperatures
and do not react with their diiron(II) precursors at an
appreciable rate. Indeed, in a separate experiment we
observed that preformed [Fe2(µ-O)(µ-O2)(6-Me3TPA)2]2+ did
not react with added diiron(II) precursor 1; similar observa-
tions were made for 4.
In the reaction of 2 with O2, no evidence for the
(67) 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.
(68) Kodera, M.; Taniike, Y.; Itoh, M.; Tanahashi, Y.; Shimakoshi, H.;
Kano, K.; Hirota, S.; Iijima, S.; Ohba, M.; Okawa, H. Inorg. Chem.
2001, 40, 4821-4822.
corresponding dimeric [FeIII2(µ-O)2(TPA)2]2+ complex could
(69) Zheng, H.; Zang, Y.; Dong, Y.; Young, V. G., Jr.; Que, L., Jr. J. Am.
Chem. Soc. 1999, 121, 2226-2235.
Inorganic Chemistry, Vol. 44, No. 1, 2005 93

Kryatov et al.
Table 5. Kinetic Parameters for the Oxygenation of Diiron(II) Complexes 1-4 at Low Temperatures
kobs, s-1
(-40 °C, 100% O2)
k, M-1 s-1
(-40 °C)
∆Hq,
∆Sq,
complex (solvent/additive)
kJ mol-1
J mol-1 K-1
1 (CH2Cl2)
1 (MeCN)
2 (CH2Cl2)
2 (MeCN)
3 (CH2Cl2)
3 (CH2Cl2/2 equiv NEt3)
3 (MeCN)
3 (MeCN/2 equiv NEt3)
4 (MeCN)
2.2(2) × 10-3
8.6(5) × 10-3
3.8(2) × 10-2
3.8(3) × 10-2
1.01(5) × 10-2
8(1) × 10-3
4.7(2) × 10-3
4.5(2) × 10-3
11.7(5)
0.70(7)
1.94(11)
12.1(6)
8.6(6)
3.2(2)
2.6(3)
1.07(5)
1.03(5)
2670(100)
16(2)
16(2)
30(4)
32(4)
36(4)
36(4)
34(4)
33(4)
16(2)
-177(10)
-167(10)
-94(10)
-87(10)
-80(10)
-81(10)
-95(10)
-101(10)
-108(10)
be obtained although we believe that it is likely to be the
initial product (reaction 7). Instead, the more stable trimer
[FeIII3(µ-O)3(TPA)3]3+ (6) was isolated in 95% yield and
identified in solution by ES-MS (Figure S31) and its visible
spectrum with maxima at 490 nm (ꢀ ) 320 M-1 cm-1), 530
nm (ꢀ ) 350 M-1 cm-1), and 595 nm (ꢀ ) 250 M-1 cm-1),
features very similar to those of previously reported [FeIII3(µ2-
O)3(5-Et3-TPA)3]3+.69 Complex 6 was crystallographically
characterized as its triflate salt (see Supporting Information,
Figure S32). We postulate that the initially formed dimer
product isomerizes rapidly to the more stable trimer that is
observed.
Oxygenation of 2. Complex [Fe2(µ-OH)2(TPA)2]2+ (2)
reacts with dioxygen in solution (CH3CN or CH2Cl2) as seen
from the color change of the solution from bright orange to
pale yellow upon O2 bubbling or mixing with O2-containing
solvent. Spectrophotometric titrations revealed that the
stoichiometry of the reaction between 2 and O2 is 2:1 in either
CH3CN or CH2Cl2 (Figure S5). The isolated product was
the more stable [Fe3(µ-O)3(TPA)3]3+ (6) presumably derived
from an initially formed [Fe2(µ-O)2(TPA)2]2+.
The oxidation kinetics of 2 with O2 was studied in
acetonitrile (T ) -40 to -5 °C) and dichloromethane (T )
-50 to -20 °C). No intermediates were detected during the
reaction at low temperatures (Figures S5 and S6). The
measurements of the pseudo-first-order rate constant at
different concentrations of the reagents, [2]0 ) 0.1-0.3 mM
and [O2]0 ) 1.0-4.2 mM, showed that the reaction is first
order in 2 and O2 (Tables S5-S7, Figures S7-S11). The
values of the second-order rate constant are quite similar in
the two studied solvents, 8.6(6) M-1 s-1 in MeCN and
12.1(6) M-1 s-1 in CH2Cl2 at -40 °C. The activation
parameters of the reaction are also nearly solvent-independent
(Table 5).
Oxygenation of 3. Diiron(II) complex [Fe2(µ-OH)2-
(BQPA)2]2+ (3) reacts with O2 with the formation of different
products depending on the nature of solvent (Table 4). In
neat CH2Cl2, the reaction proceeds similarly to the oxygen-
ation of 2 yielding [Fe2(µ-O)2(BQPA)2]2+ with no detectable
intermediates even at -80 °C (Figures S12 and S13). The
reaction is accompanied by a color change of the solution
from dark red to brownish-yellow. However, in the presence
of triethylamine (2 equiv) in CH2Cl2, the oxygenation of
complex 3 at low temperatures (at or below -50 °C) gives
nearly quantitatively a dark green dioxygen adduct 3‚O2
(Figure S14). At higher temperatures (-45 °C to -25 °C),
the yield of 3‚O2 is smaller, and its decomposition becomes
evident (Figures S15 and 16). The oxygenation of 3 also
proceeds in neat MeCN at -40 °C, resulting in a low yield
(30-35%) of 3‚O2 (Figure S21). A higher yield (ca. 70%)
of the dioxygen adduct was obtained in MeCN in the
presence of Et3N (2 equiv).
Oxygenation of 1. Earlier we reported a detailed study
on the oxygenation of the perchlorate salt of complex 1,
[FeII (µ-OH)2(6-Me3-TPA)2](ClO4)2.26,27 At low temperature
2
in dichloromethane the reaction yielded quantitatively a (µ-
peroxo)diiron(III) complex 1‚O2, [FeIII2(µ-O)(µ-O2)(6-Me3-
TPA)2]2+, which was characterized by a number of spectro-
scopic techniques including an EXAFS structural deter-
mination.26,67 The triflate salt of complex 1, [FeII (µ-OH)2-
2
(6-Me3-TPA)2](CF3SO3)2, demonstrates the same chemistry.
The reaction of bright-yellow 1(OTf)2 with O2 in neat CH2Cl2
at low temperatures (-70 to -40 °C) gave in high yield the
dark green dioxygen adduct 1‚O2 (Figures 4 and S2). The
reaction of 1(OTf)2 and O2 is first order in each of the
reagents with a second-order rate constant k ) 0.7(1) M-1s-1
in CH2Cl2 at -40 °C. The dependence of k on temperature
(T ) -70 to -40 °C) gave the activation parameters ∆Hq
) 16 ( 2 kJ mol-1 and ∆Sq ) -177 ( 10 J mol-1 K-1
(Table S3, Figure S3). These kinetic parameters are almost
identical to those obtained previously for the perchlorate
salt 1(ClO4)2.27 Thus, no change in the course or kinetics of
the oxygenation reaction in nonpolar solvents was found
-
upon the substitution of the counterion from ClO4 to
-
CF3SO3 .
In MeCN the oxygenation of 1 also gave the peroxo
complex 1‚O2, but its formation even at -45 °C (near the
freezing point of MeCN) was closely followed by its
decomposition, and the yield of 1‚O2 was small (25% or
less). The formation of 1‚O2 in MeCN was first order in
both 1 and O2 with a second-order rate constant k ) 1.9(1)
M-1s-1 at -40 °C and the activation parameters ∆Hq ) 16
( 2 kJ mol-1 and ∆Sq ) -167 ( 10 J mol-1 K-1 (Table
S4, Figure S4). The rate of oxygenation of complex 1 was
somewhat higher in MeCN than in CH2Cl2, but the activation
parameters are indistinguishable (within experimental error)
for both solvents (Table 5).
Complex 3‚O2 can also be generated from 3 and O2 in
CH2Cl2 at low temperature in the presence of other amines
(Figures S14 and S18). Diisopropylethylamine (i-Pr2EtN),
1,4-diazabicyclo[2.2.2]octane (DABCO), 1,8-diazabicyclo-
[4.3.0]non-5-ene (DBU), and N,N,N′,N′-tetramethylguanidine
(TMG) were effective in the amount of 2 equiv (vs 3). TMG,
which is the strongest base in the series, was also effective
94 Inorganic Chemistry, Vol. 44, No. 1, 2005

FeII2(µ-OH)2 Cores and O2-Adduct Formation
in a substoichiometric amount (0.5 equiv). It should be noted
that the addition of the amines does not change the UV-vis
spectrum of the diiron(II) precursor 3 (except for the case
of DBU, where a slow decomposition process leading to the
formation of a brown precipitate occurred at room temper-
ature).
The kinetics of the reaction between 3 and O2 in neat
CH2Cl2 was studied in the temperature range from -65 to
+15 °C. Dioxygen (0.60-3.2 mM) was always taken in
excess relative to complex 3 (0.10-0.60 mM). Spectral
changes observed by stopped-flow spectrophotometry were
nearly the same as those obtained by regular spectropho-
tometry in the “batch” experiments, confirming that in neat
CH2Cl2 the reaction product is the oxo-bridged diiron(III)
complex [Fe2(µ-O)2(BQPA)2]2+ with no observable inter-
mediates (Figures S12 and S13). The reaction between 3 and
O2 is first order in each of the reagents and second order
overall (Figure S19, Table S12). The temperature dependence
of k gave the activation parameters of the reaction as ∆Hq
) 36 ( 4 kJ mol-1 and ∆Sq ) -80 ( 10 J mol-1 K-1
(Figure S20, Table S10).
The above-mentioned experiments were carried out in
rigorously dry, nonstabilized dichloromethane. No change
in the reaction rate was observed when nonstabilized
dichloromethane was substituted for the more usual com-
mercial brand of dichloromethane, which contains ca. 50 mM
of pentenes. Addition of small amounts of water (2-4 mM),
which is in excess to the 0.5 mM concentration of complex
3, did not change significantly the rate of oxygenation of 3
(it increased by about 20-30%), similarly to what was
observed earlier for complex 1.27 It was impossible to
increase the concentration of water in CH2Cl2 further without
causing ice precipitation from the solution at low tempera-
tures, which interfered strongly with the stopped-flow
experiments.
The kinetics of the reaction between 3 and O2 in CH2Cl2
solution in the presence of triethylamine (Et3N, 2 equiv vs
3) was studied by stopped-flow spectrophotometry in the
temperature range from -65 to -25 °C (Figures S13 and
S14). Spectral changes observed by stopped-flow spectro-
photometry in the course of the reaction between 3 and O2
in the presence of Et3N were very similar to those obtained
by regular spectrophotometry in the batch experiments and
characteristic of the peroxo complex 3‚O2 (Figures S14 and
S15). The average rate of the reaction in the presence of
triethylamine decreased by ca. 20% relative to that in neat
dichloromethane (Tables S10 and S11). The addition of Et3N
changed neither the rate law of the reaction, first order in 3
and O2 (Figure S19), nor the activation parameters of the
process (Figure S20).
An exploratory kinetic study was done with the other
amines: i-Pr2NEt, DABCO, DBU, and TMG. Kinetic traces
of the reaction between 3 and excess O2 in the presence of
all studied amines (2 equiv) at -35 °C could be fit well to
single-exponential eq 1 with similar values of the pseudo-
first-order rate constant (Table S15). The rate of oxygenation
did not change significantly in the presence of different
amines, so no further studies were carried out.
Figure 5. Time-resolved visible spectra acquired by stopped-flow
technique during the reaction of complex 4 (0.5 mM) with O2 (4.4 mM) in
MeCN at T ) -40 °C. Inset: kinetic data points superimposed with a
pseudo-first-order fit (k ) 12.0 s-1).
The reaction between complex 3 and excess O2 in
acetonitrile was studied by stopped-flow spectrophotometry
in the temperature range from -40 °C to -15 °C. The
reaction has two stages, which correspond to the formation
and decay of the oxygen adduct 3‚O2, respectively. The
maximum yield of the 3‚O2 intermediate in the course of
the reaction is about 35% at -35 °C (Figure S21). The
kinetic parameters for the reaction between 3 and O2 in
MeCN are very similar to those determined in CH2Cl2 (Table
5).
Addition of amine to the starting solution of complex 3
in acetonitrile significantly improved the yield of peroxo
complex 3‚O2. In the presence of triethylamine or DBU (2
equiv) at -35 °C, the yield of 3‚O2 was about 70% (Figure
S21). The reaction rate was only about 10% slower compared
to that in neat MeCN (Table S15), and the activation
parameters were not influenced significantly by the addition
of amine (Table 5, Figure S22).
Oxygenation of 4. The red-purple complex [Fe2(µ-OH)2-
(BnBQA)2(MeCN)2](ClO4)2 (4) reacts with O2 in MeCN
solution very rapidly at -40 °C forming a dark green
dioxygen adduct 4‚O2 (λmax ) 502 nm, ꢀ ) 1300 M-1 cm-1;
λmax ) 650 nm, ꢀ ) 1300 M-1 cm-1) (Figures 5, S23).
Because the solid complex 4 had a limited shelf life, an
alternative method of preparing solutions of 4 was developed
by mixing monomeric complex 5(ClO4)2 with 2 equiv of
triethylamine in acetonitrile. Kinetics of oxygenation of
solutions of 4 prepared from a solid dimeric complex and
solutions of 4 generated in situ from 5 and NEt3 were found
to be identical. In most kinetic measurements, solutions of
4 generated in situ were used. Kinetic studies at low
temperature (-45 to -20 °C) showed that the reaction was
first order in each of the reagents (4 and O2) with a second-
order rate constant of 2700 M-1 s-1 at -40 °C and the
Inorganic Chemistry, Vol. 44, No. 1, 2005 95

Kryatov et al.
Scheme 4. Proposed Reaction Pathway of the Oxygenation of
Complexes 1-4
The Rate-Limiting Dioxygen Binding Step. The rate of
oxygenation of the diiron(II) complexes increases in the order
1 ∼ 3 < 2 , 4 (Table 5). The initial oxidation of the
diiron(II) center may involve either an inner-sphere or outer-
sphere process. However, an outer-sphere one-electron
transfer would appear to be quite unlikely, because the
Figure 6. Arrhenius plots of the second-order rate constants (M-1 s-1
for the oxygenation of diiron(II) complexes 1-4 in MeCN at low
temperatures (-45 to -5 °C).
)
-
potential for the O2/O2 redox pair in organic solvents (ca.
-1250 mV vs Fc+/Fc in MeCN and CH2Cl2, Table 3)35,56-59
is about 1 V more negative than those determined for the
FeIIFeIII/FeIIFeII redox pairs for 1-4 (-460 to -150 mV vs
Fc+/Fc, Table 3). Furthermore, there is no correlation
between the half-wave redox potentials for 1-4 with the rates
of their oxygenation (Table 5). In particular, 3 and 4 exhibit
nearly identical redox potentials, but the oxygenation rate
of 4 is about 103 times faster than that of complex 3. Thus,
an outer-sphere mechanism is highly unlikely.70
activation parameters ∆Hq ) 16 ( 2 kJ mol-1 and ∆Sq )
-108 ( 10 J mol-1 K-1 (Tables S16-S18, Figures S23-
S25). The main difference between 4 and the other diiron(II)
complexes (1-3) is that the oxygenation of 4 proceeds about
3 orders of magnitude faster in MeCN (Figures 5 and 6).
Complex 4 dissolves in CH2Cl2 with decomposition, and
therefore, the oxygenation reaction could not be studied in
that solvent. It should be noted that, for other complexes
(1-3), the change of solvent from MeCN to CH2Cl2 did not
alter significantly the kinetic parameters of the oxygenation
reaction (Table 4).
A more probable mechanism involves the rate-limiting
coordination of the O2 molecule to the diiron(II) center. The
trend in the rate constants of the oxygenation reactions (1 ∼
3 < 2 , 4) follows the order of decreasing ligand bulk and
increasing access to the diiron core, fully consistent with an
inner-sphere mechanism. Such a mechanism is proposed in
Scheme 4.
Complex 1 has the slowest rate of oxygenation of the four
complexes studied (Table 5). The results reported here for
the triflate salt of 1 agree with previous observations on the
perchlorate salt,27 showing that the counterions play no role
in the oxygenation reaction. The slow kinetics can be
attributed to the 6-methyl substituents of the 6-Me3-TPA
ligand that severely restrict access of O2 to the diiron(II)
core,27 as illustrated by the space filling model shown in
Figure 7A. Because of the congestion about the diiron(II)
core and the coordinative saturation of the iron(II) ions, it
seems highly unlikely that O2 could coordinate to both
iron(II) centers in one step. The O2 molecule would more
likely bind to one metal ion first to form a transient
Fe(II)Fe(III)-superoxo species as the first intermediate.
Similar intermediates have been proposed for the oxygenation
of other diiron(II) complexes22,27-29 (including MMO1,7), but
a superoxo intermediate has not yet been characterized. For
Discussion
Diiron(II) complexes 1-4 all react with dioxygen. Under
carefully selected conditions of solvent and temperature, the
oxygenation of complexes 1, 3, and 4 yields diiron(III)
peroxo species in high yield (Table 4). Under other condi-
tions, the reaction products are oxo-bridged di- or triiron(III)
complexes without any observable intermediates. The kinet-
ics of the oxygenation of 1-4 in solution were studied by
stopped-flow techniques at low temperatures. In all cases,
the reaction is first order in diiron(II) complex and first order
in O2 (second order overall) and characterized by low
enthalpies of activation (∆Hq from 16 to 36 kJ mol-1) and
strongly negative entropies of activation (∆Sq from -177 to
-80 J mol-1 K-1). The values of rate constants and activation
parameters of the reactions are compared in Table 5 and
Figure 6. For any particular diiron(II) complex (1-4), the
kinetic parameters do not depend substantially on the solvent
and the observable reaction product (peroxo vs oxo-bridged
oligoiron(III) complexes). The kinetic data we have obtained
suggest that the first and rate-limiting step of the reaction is
the association of the diiron(II) complex and O2 leading to
the oxidation of the diiron(II) center.
(70) Marcus, R. A. Angew. Chem., Int. Ed. Engl. 1993, 32, 1111-1121.
96 Inorganic Chemistry, Vol. 44, No. 1, 2005

FeII2(µ-OH)2 Cores and O2-Adduct Formation
be a result of stronger bonds for the candidate ligands for
the dissociation step. Specifically, the Fe2(µ-OH)2 core in 2
is less asymmetric than in 1, with the weaker Fe-OH bond
having a distance of ∼2.15 Å for 2, compared to ∼2.18 Å
for 1 (Table 2). The in-plane Fe-Npy bond of 2 (2.19 Å) is
also shorter than that in 1 (∼2.30 Å). Thus, the strengthening
of either bond can account for the observed increased
activation enthalpy of the oxygenation reaction.
Interestingly, the reaction of 2 with O2 does not result in
the observation of a peroxo intermediate. Instead the reaction
affords [FeIII3(µ-O)3(TPA)3]3+, the more stable oligomer of
the presumed initial [FeIII2(µ-O)2(TPA)2]2+ product. It is
likely that the complete absence of 6-methyl substituents on
the pendant pyridines results in a much more exposed diiron
core that cannot stabilize a peroxo intermediate.
The ligand BQPA provides a degree of steric shielding
around the diiron core that is intermediate between TPA and
6-Me3-TPA. Structural studies of complexes 1-3 demon-
strated that a 6-substituent on a pyridine ring introduces
sufficient steric hindrance in the vicinity of the iron center
to cause a lengthening of the Fe-N bond. The BQPA ligand
has only two pendant arms with 6-substituted pyridine rings,
and these substituents consist of aromatic sp2 carbons that
impose less steric bulk around the metal center than the sp3-
carbon substituents in 6-Me3-TPA. The decrease in steric
shielding about the diiron(II) core of 3 relative to that of 1
is evident from the space filling model of the X-ray structure
shown in Figure 7B. However, 3 reacts with O2 at rates
comparable to 1. As observed for 2, there is a favorable
change in activation entropy that is counterbalanced by an
unfavorable increase in activation enthalpy. Similarly, the
candidate ligands for dissociation in 3 have stronger bonds
than corresponding ones in 1; the longer Fe-µ-OH distance
is 2.10 Å, and the in-plane Fe-Npy distance is 2.17 Å.
Like 2, 3 reacts with O2 to afford an oxo-bridged diiron(III)
complex even at low temperature. However, the peroxo
intermediate 3‚O2 can be trapped if the oxygenation is carried
out in the presence of an added base. Interestingly, the kinetic
parameters obtained under the two reaction conditions are
essentially identical. This result shows that the oxygen
binding step that is probed by the kinetic measurements is
the same irrespective of the presence of base and that adding
base affects the chemistry after the formation of the initial
O2 adduct. (Further discussion on the chemistry associated
with the presence of added base can be found in the next
section.)
Figure 7. Space-filling representations of the X-ray structures of the
complexes 1 (a), 3 (b), and 4 (c): C, gray; H, white; N, blue; O, red; Fe,
green.
1, its superoxo intermediate (1A) must then undergo a series
of faster transformations to yield the first detected reaction
product, the (µ-oxo)(µ-peroxo)diiron(III) intermediate (1C),
and then a bis(µ-oxo)diiron(III) complex upon thermal decay
(Scheme 4).
The first step of the mechanism requires ligand dissociation
to make at least one metal center coordinatively unsaturated.
Examination of the crystal structure of 1 (Figure 2) identifies
two plausible candidates for the dissociable ligand. Complex
1 has an asymmetric Fe2(µ-OH)2 core with one short (∼1.99
Å) and one long (∼2.18 Å) Fe-OH bond based on the
structures of 1(ClO4)2 and 1(OTf)2 (Table 2). Breaking the
weaker Fe-OH bond would cost little energy and could be
compensated for by concomitant O2 binding in the transition
state. It is also conceivable for one of the pendant pyridine
arms to dissociate, as observed for some mononuclear iron(II)
complexes of 6-substituted TPA ligands.71 For 1, the in-plane
Fe-Npy bond is quite weak (∼2.30 Å) and could thus
dissociate to allow O2 binding. The activation parameters
found for O2 binding to 1 (Table 5) support this ligand
dissociation scenario: the activation enthalpy (16 kJ mol-1)
is relatively small, while the activation entropy is very
unfavorable (ca. -170 J mol-1 K-1) due to the steric
hindrance. These results suggest that oxygenation of di-
iron(II) complexes may be facilitated by mitigation of the
steric bulk from the 6-Me3-TPA ligand to afford a more
favorable activation entropy.
The last complex in our series, 4, uses the tridentate ligand
BnBQA, which is related to the tetradentate BQPA, except
for the replacement of the 2-pyridylmethyl pendant arm with
a sterically analogous, but noncoordinating, benzyl group.
The resulting complex differs from the other bis(µ-hydroxo)-
diiron(II) complexes in having a tridentate meridional ligand
(BnBQA) and a molecule of coordinated MeCN instead of
a tripodal tetradentate ligand (Figures 2 and 7c). Notably, 4
reacts with O2 3 orders of magnitude faster than 1 in the
studied range of temperatures (Figure 6, Table 5), with an
activation enthalpy comparable to that of 1 and an activation
entropy comparable to those of 2 and 3. While the alleviation
At the other extreme of steric bulk is 2, a complex of the
unsubstituted TPA ligand. It reacts with O2 at a rate
severalfold faster than 1 in the studied temperature range.
An examination of its activation parameters shows that, while
there is a favorable change in activation entropy (as
expected), this gain is counterbalanced by an unfavorable
increase in activation enthalpy (from 16 kJ mol-1 for 1 to
ca. 31 kJ mol-1 for 2, Table 5). Analysis of the structural
data suggests that this increase in activation enthalpy may
(71) Mandon, D.; Machkour, A.; Goetz, S.; Welter, R. Inorg. Chem. 2002,
41, 5364-5372.
Inorganic Chemistry, Vol. 44, No. 1, 2005 97

Kryatov et al.
of congestion around the diiron core readily rationalizes the
more favorable activation entropy relative to that of 1, the
comparable activation enthalpies of 1 and 4 suggest that
metal-ligand bonds of comparable strength are broken in
the rate-determining step. Comparison of the two candidate
bonds in 4 suggests that it must be the in-plane Fe-NCMe
bond that is broken, as its distance of 2.29 Å is comparable
to the ∼2.30 Å distance for the in-plane Fe-Npy bond in 1.
In contrast, the weaker Fe-µ-OH bond (2.09 Å) of 4 is
significantly shorter than the corresponding bond in 1 (2.18
Å). Thus, both enthalpic and entropic factors accelerate the
oxygenation kinetics of 4. Taken together, the kinetic results
for the four complexes strongly suggest that the attack of
O2 on the diiron(II) core occurs by displacement of an in-
plane nitrogen ligand, rather than the weaker hydroxo bridge
in the Fe2(µ-OH)2 core.
Product Determining Steps Subsequent to Dioxygen
Binding. The nature of the polydentate ligand also affects
the fate of the O2 adduct after the rate-limiting step, affording
either the peroxo intermediate or the bis(µ-oxo)diiron(III)
complex. The peroxo intermediate is favored by the presence
of at least two pendant pyridines with 6-substituents, low
temperature, and addition of base in the case of 3. The steric
hindrance introduced by the 6-substituents prolongs the
lifetime of the peroxo intermediate at low temperature.
However, the effect of added base in the case of 3 suggests
a requirement for the loss of a proton to form the peroxo
intermediate. It is clear that the added amine affects only
the reaction steps after the rate-limiting step, because there
are no changes in the UV-vis spectrum of the starting
diiron(II) complex or the kinetics of its reaction with O2.
The role of the added amine is most probably that of a
general base, since DBU and especially TMG acted more
efficiently in generating the peroxo intermediate than Et3N,
i-Pr2NEt, or DABCO, consistent with the stronger basicities
of the former group of amines. There thus appears to be a
proton-sensitive branching point after O2 binding that
determines the nature of the product.
The mechanism proposed in Scheme 4 can account for
these experimental observations. As discussed in detail in
the previous section, the rate-limiting step is a bimolecular
reaction between the diiron(II) complex and O2 yielding a
superoxo Fe(II)Fe(III) species designated A, which does not
accumulate in the reaction mixture and exists as a steady-
state intermediate with a low concentration. The next step
is a hydrogen-atom transfer or a proton-coupled electron
transfer to form intermediate B, which is analogous to that
proposed for the formation of oxyhemerythrin after the initial
O2 binding step.72 Loss of H2O then leads to peroxo
intermediate C. Alternatively, a second equivalent of di-
iron(II) complex may be oxidized by A or B to form 2 equiv
of the bis(µ-oxo)diiron(III) product. The product is deter-
mined by the relative rates of the two steps, with the latter
clearly favored at ambient temperature for all four complexes.
There are two possibilities for the base-sensitive species,
either the hydroxo bridge in A or the coordinated hydro-
peroxide in B. Loss of either proton should facilitate
formation of a 1,2-peroxo bridge. The pKa of this proton is
determined by the Lewis acidity of the metal center, as
modulated by the polydentate ligand. As indicated by its low
energy iron(II)-to-pyridine charge transfer band (Figure 1)
and rather negative FeIIFeIII/FeIIFeII redox potential (Table
3), 2 has the least Lewis acidic metal center of the four
complexes. The consequent higher pKa of the hydroxo proton
in A or the hydroperoxo proton in B would disfavor
formation of the peroxo intermediate, while the more
negative redox potential of the diiron(II) precursor would
promote formation of the bis(µ-oxo)diiron(III) product. On
the other hand, 1 has the most Lewis acidic metal center,
which lowers the pKa of the hydroxo proton in A or the
hydroperoxo proton in B sufficiently to promote formation
of the peroxo intermediate even in the absence of added base.
At the same time, its higher redox potential disfavors the
bimolecular self-oxidation pathway at low temperature.
Complex 3 represents an intermediate case that allows
added base to determine the course of its reaction with O2.
In neat CH2Cl2, self-oxidation occurs, but in neat MeCN,
the peroxo complex 3‚O2 forms in a relatively low yield due
to the weakly basic nature of MeCN. Addition of a strong
base in either solvent favors formation of the peroxo
intermediate.
The clean formation of the peroxo intermediate in the case
of 4 suggests that its diiron center is more Lewis acidic than
that in 3. The greater Lewis acidity of the metal center in 4
may be rationalized by the replacement of the pyridine ligand
in 3 with the less basic MeCN ligand in 4 and is corroborated
by two observations: (a) that the iron(II)-to-quinoline charge
transfer band in 3 is red-shifted relative to that of 4 (Figure
1), and (b) that the peroxo-to-iron(III) charge transfer band
of 3‚O2 is blue-shifted relative to that of 4‚O2 (Table 4,
Figure 4). Thus, the effect of the 6-substituents on the
pyridine ligands is not only steric but also electronic. In
addition, peroxo intermediate formation may also be pro-
moted by the facile displacement of MeCN on the second
iron, a feature not available for the other complexes.
Summary and Perspective
A series of bis(µ-hydroxo)diiron(II) complexes 1-4 with
polydentate aminopyridine ligands has been synthesized and
structurally characterized. The ligands used in the series
(Scheme 2) are tetradentate TPA (2), tetradentate BQPA (3),
and tridentate BnBQA (4), and tetradentate 6-Me3TPA (1),
listed in order of increasing steric congestion around the
diiron(II) core as indicated by the number of pyridine ligands
with 6-substituents. Electrochemical studies show that this
relative order also corresponds to the order of FeIIIFeII/FeIIFeII
redox potentials ranging from -460 to -150 mV versus Fc+/
Fc (in MeCN or CH2Cl2), confirming that 6-substituted
pyridines favor the iron(II) state.39 All four complexes react
readily with O2 in similar bimolecular associative processes.
Interestingly, the three complexes with tetradentate ligands
(1-3) have comparably slow oxygenation rates despite
having ligands with different steric requirements. Oxygen-
ation rates of these complexes are constrained by a very
(72) Brunold, T. C.; Solomon, E. I. J. Am. Chem. Soc. 1999, 121, 8288-
8295.
98 Inorganic Chemistry, Vol. 44, No. 1, 2005

FeII2(µ-OH)2 Cores and O2-Adduct Formation
unfavorable activation entropy in the case of 1 due to steric
congestion around the diiron core and by relatively large
activation enthalpies for less hindered complexes 2 and 3
arising from the required dissociation of a more strongly
bound ligand in the rate-limiting step. In contrast, the
complex with the tridentate ligand (4) reacts with O2 much
faster. The use of tridentate BnBQA introduces MeCN as
the sixth ligand on each iron in place of the third arm of the
tetradentate ligands, and its kinetic lability lowers the
activation enthalpy for oxygenation. At the same time, the
reduced steric crowding around the diiron core of 4 also
makes the activation entropy less unfavorable. So enthalpic
and entropic factors combine to give rise to the 1000-fold
rate acceleration observed for 4.
The remarkable effect of added base in promoting peroxo
intermediate formation in the case of 3 has revealed the
involvement of a proton-sensitive step that controls the fate
of the initial O2 adduct A (Scheme 4). In the absence of
base, O2 adduct 3A or its isomer 3B can oxidize residual 3.
Interestingly, 3‚O2 does not react with added 3; neither do
1‚O2 and 4‚O2 with their respective diiron(II) precursors.
These results suggest that the (µ-1,2-peroxo)diiron(III) state
is relatively unreactive and thus may not be directly involved
in oxidation reactions. This relative stability may account
for the fact that almost all O2 adducts of synthetic diiron
centers have been characterized to be (µ-1,2-peroxo)-
diiron(III) complexes.73 Indeed, the oxidative reactivities of
synthetic (µ-1,2-peroxo)diiron(III) intermediates24,28,29 and
related systems74 investigated so far are generally poor; they
only work on easily oxidizable functional groups and often
require substrate coordination to the metal center.
Dioxygen adducts of diiron enzymes are also characterized
to have (µ-1,2-peroxo)diiron(III) centers such as those found
for D84E RNR R275 and chemically reduced ∆9D desatu-
rase;9 these intermediates are also quite stable and similarly
unreactive. However, it has been demonstrated that the
conversion of diiron(III)-peroxo intermediate MMO-P to
methane oxidizing diiron(IV) intermediate MMO-Q requires
a proton.76 So protonation of the (µ-1,2-peroxo)diiron(III)
intermediate C (Scheme 4) may be required to activate the
O-O bond for cleavage in order to attack aliphatic C-H
bonds such as those oxidized by methane monooxygenase
and fatty acid desaturases. This protonation strategy is
utilized by heme enzymes to promote cleavage of the O-O
bond.77-79 Efforts to trap an HOO-FeIII-O-FeIII intermedi-
ate (species B of Scheme 4) and characterize its reactivity
are ongoing.
Acknowledgment. This work was supported by grants
from the National Science Foundation (CHE-0111202 to
E.V.R.-A.), the National Institutes of Health (GM-38767 to
L.Q.), and the Research Corporation (RI0223 to E.V.R.-A.).
The CCD based X-ray diffractometer at Tufts University was
purchased through Air Force DURIP Grant F49620-01-1-
0242, and the ESI mass spectrometer was funded by NSF
Grant MRI 0320783.
Supporting Information Available: Crystallographic data in
CIF format and additional figures and tables. This material is
available free of charge via the Internet at http://pubs.acs.org.
IC0485312
(75) Voegtli, W. C.; Khidekel, N.; Baldwin, J.; Ley, B. A.; Bollinger, J.
M., Jr.; Rosenzweig, A. C. J. Am. Chem. Soc. 2000, 122, 3255-3261.
(76) Lee, S.-K.; Lipscomb, J. D. Biochemistry 1999, 38, 4423-4432.
(77) Sono, M.; Roach, M. P.; Coulter, E. D.; Dawson, J. H. Chem. ReV.
1996, 96, 2841-2888.
(78) Harris, D. L.; Loew, G. H. J. Am. Chem. Soc. 1998, 120, 8941-8948.
(79) Loew, G. H.; Harris, D. L. Chem. ReV. 2000, 100, 407-420.
(73) Girerd, J.-J.; Banse, F.; Simaan, A. J. Struct. Bonding 2000, 97.
(74) Carson, E. C.; Lippard, S. J. J. Am. Chem. Soc. 2004, 126, 3412-
3413.
Inorganic Chemistry, Vol. 44, No. 1, 2005 99