Models for nonheme iron intermediates: Structural basis for tuning the spin states of Fe(TPA) complexes

Article abstract of DOI:10.1021/ja9638521

Our efforts to model the oxygen activation chemistry of nonheme iron enzymes have yielded transient intermediates with novel properties. These properties can be dramatically affected by the introduction of a 6-methyl substituent on the pendant pyridines of the tetradentate ligand TPA (TPA = tris(2-pyridylmethyl)amine). A series of Fe(TPA) complexes has thus been synthesized and characterized to provide the structural basis for these dramatic effects. The following complexes have been obtained: [Fe(L)(CH₃CN)₂](ClO₄)₂ (1, L = TPA; 2, L = 6-MeTPA; 3, L = 6-Me₂TPA; 4, L = 6-Me₃TPA) and [Fe(L)(acac)](ClO₄)₂ (5, L = TPA; 6, L = 5-Me₃TPA; 7, L = 6-MeTPA). As indicated by ₁H NMR and/or EPR, 1, 5, and 6 with no 6-methyl substituent are low spin, while complexes 2, 3, 4, and 7 with at least one 6- methyl substituent are all high spin, with higher redox potentials than their low-spin counterparts. The ligands with 6-methyl substituents thus favor a metal center with a larger ionic radius, i.e., Fe¹¹ over Fe(III) and high spin over low spin. Careful scrutiny of the crystal structures of 1, 4, 6, and 7 reveals that one hydrogen of the 6-methyl group is only 2.7 ? away from the metal center in the high-spin complexes. Its presence thus prevents the pyridine nitrogen from forming an Fe-N bond shorter than 2.1 ? as required for an iron center to adopt a low-spin configuration. This steric effect of the 6-methyl substituent serves as a simple but very useful ligand design tool to tune the electronic properties of the metastable alkylperoxoiron(III) species derived from the reactions of 1-4 with tert- butyl hydroperoxide. These intermediates serve as models for low-spin and high-spin peroxoiron(III) species in the reaction cycles of the antitumor drag bleomycin and lipoxygenase, respectively. Similar principles apply in the design of nonheme diiron(II) complexes that reversibly bind dioxygen and of high-valent bis(μ-oxo)diiron complexes that model the high-valent intermediates in the redox cycles of nonberne diiron enzymes such as methane monooxygenase and ribonucleotide reductase.

Full text of DOI:10.1021/ja9638521

J. Am. Chem. Soc. 1997, 119, 4197-4205
4197
Models for Nonheme Iron Intermediates: Structural Basis for
Tuning the Spin States of Fe(TPA) Complexes
Yan Zang,† Jinheung Kim,† Yanhong Dong,† Elizabeth C. Wilkinson,†
Evan H. Appelman,‡ and Lawrence Que, Jr.*,†
Contribution from the Department of Chemistry and Center for Metals in Biocatalysis,
UniVersity of Minnesota, Minneapolis, Minnesota 55455, and Chemistry DiVision,
Argonne National Laboratory, Argonne, Illinois 60439
ReceiVed NoVember 6, 1996X
Abstract: Our efforts to model the oxygen activation chemistry of nonheme iron enzymes have yielded transient
intermediates with novel properties. These properties can be dramatically affected by the introduction of a 6-methyl
substituent on the pendant pyridines of the tetradentate ligand TPA (TPA ) tris(2-pyridylmethyl)amine). A series
of Fe(TPA) complexes has thus been synthesized and characterized to provide the structural basis for these dramatic
effects. The following complexes have been obtained: [Fe(L)(CH3CN)2](ClO4)2 (1, L ) TPA; 2, L ) 6-MeTPA;
3, L ) 6-Me2TPA; 4, L ) 6-Me3TPA) and [Fe(L)(acac)](ClO4)2 (5, L ) TPA; 6, L ) 5-Me3TPA; 7, L ) 6-MeTPA).
As indicated by 1H NMR and/or EPR, 1, 5, and 6 with no 6-methyl substituent are low spin, while complexes 2, 3,
4, and 7 with at least one 6-methyl substituent are all high spin, with higher redox potentials than their low-spin
counterparts. The ligands with 6-methyl substituents thus favor a metal center with a larger ionic radius, i.e., FeII
over FeIII and high spin over low spin. Careful scrutiny of the crystal structures of 1, 4, 6, and 7 reveals that one
hydrogen of the 6-methyl group is only 2.7 Å away from the metal center in the high-spin complexes. Its presence
thus prevents the pyridine nitrogen from forming an Fe-N bond shorter than 2.1 Å as required for an iron center to
adopt a low-spin configuration. This steric effect of the 6-methyl substituent serves as a simple but very useful
ligand design tool to tune the electronic properties of the metastable alkylperoxoiron(III) species derived from the
reactions of 1-4 with tert-butyl hydroperoxide. These intermediates serve as models for low-spin and high-spin
peroxoiron(III) species in the reaction cycles of the antitumor drug bleomycin and lipoxygenase, respectively. Similar
principles apply in the design of nonheme diiron(II) complexes that reversibly bind dioxygen and of high-valent
bis(µ-oxo)diiron complexes that model the high-valent intermediates in the redox cycles of nonheme diiron enzymes
such as methane monooxygenase and ribonucleotide reductase.
Introduction
intradiol cleaving catechol dioxygenases) and oxidative decar-
boxylation of benzoylformate10 (mimicking the R-keto acid-
dependent enzymes). We have also trapped transient nonheme
iron-peroxo species at low temperature from the reactions of
peroxides with Fe(TPA) precursors.11,12 Both low-spin and
high-spin alkylperoxoiron(III) complexes can be obtained
depending on whether TPA or 6-Me3TPA is used in the
precursor complex. The TPA species has a low-spin iron(III)
center11 like that found in “activated bleomycin”,13 while the
6-Me3TPA species has a high-spin iron(III) center12 like that
associated with the purple form of soybean lipoxygenase.6
Using these ligands, we also obtained [FeIIIFeIV(µ-O)2L2]3+, the
first examples of complexes with a high-valent bis(µ-oxo)diiron
diamond core,14,15 a new structural motif that we propose for
the oxidizing intermediates in the reaction of O2 with nonheme
diiron enzymes.16 The properties of the TPA derivative indicate
that it is a valence-delocalized complex consisting of double
exchange-coupled low-spin FeIII and FeIV centers, while those
of the 6-MeTPA derivative show that it is a valence-localized
The tetradentate tripodal ligand TPA (tris(2-pyridylmethyl)-
amine)1 and its derivatives play an important and prominent
role in our efforts to synthesize functional models of mono-
nuclear and dinuclear nonheme iron oxygen activating en-
zymes2,3 such as the catechol dioxygenases,4 R-keto acid-
dependent enzymes,5 lipoxygenase,6 ribonucleotide reductase,7
and methane monooxygenase.8 To date, Fe(TPA) complexes
have been found to react quantitatively with O2 to afford
oxidative cleavage of 3,5-di-tert-butylcatechol9 (mimicking the
† University of Minnesota.
‡ Argonne National Laboratory.
X Abstract published in AdVance ACS Abstracts, April 15, 1997.
(1) Abbreviations: acac ) 2,4-pentanedione (acetylacetone); BF )
benzoylformate; bpy ) bipyridine; BPA ) bis(2-pyridylmethyl)amine; bpca
) N-(2-pyridinylcarbonyl)-2-pyridinecarboximidate monoanion; HPTP )
1,3-bis[N,N-bis(2-pyridylmethyl)amino]-2-hydroxypropane; NTB ) N,N,N-
tris(2-benzimidazolylmethyl)amine; phen ) 1,10-phenanthroline; salen )
N,N′-ethylenebis(salicylidenamine) dianion; 5-Me3TPA ) tris[(5-methyl-
2-pyridyl)methyl]amine; 6-MeTPA ) bis(2-pyridylmethyl)[(6-methyl-2-
pyridyl)methyl]amine; 6-Me2TPA ) bis[(6-methyl-2-pyridyl)methyl](2-
pyridylmethyl)amine; 6-Me3TPA ) tris(6-methyl-2-pyridylmethyl)amine;
TPA ) tris[(2-pyridyl)methyl]amine; tpy ) 2,2′:6′,2′′-terpyridine.
(2) Que, L., Jr.; Ho, R. Y. N. Chem. ReV. 1996, 96, 2607-2624.
(3) Wallar, B. J.; Lipscomb, J. D. Chem. ReV. 1996, 96, 2625-2657.
(4) Lipscomb, J. D.; Orville, A. M. Metal Ions Biol. Syst. 1992, 28, 243-
298.
(9) Jang, H. G.; Cox, D. D.; Que, L., Jr. J. Am. Chem. Soc. 1991, 13,
9200-9204.
(10) Chiou, Y.-M.; Que, L., Jr. J. Am. Chem. Soc. 1995, 117, 3999-
4013.
(11) Kim, J.; Larka, E.; Wilkinson, E. C.; Que, L., Jr. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 2048-2051.
(12) Zang, Y.; Elgren, T. E.; Dong, Y.; Que, L., Jr. J. Am. Chem. Soc.
1993, 115, 811-813.
(5) Abbott, M. T.; Udenfriend, S. In Molecular Mechanisms of Oxygen
ActiVation; Hayaishi, O., Ed.; Academic Press: New York, 1974; pp 167-
214.
(13) Sam, J. W.; Tang, X.-J.; Peisach, J. J. Am. Chem. Soc. 1994, 116,
5250-5256.
(6) Nelson, M. J.; Seitz, S. In ActiVe Oxygen in Biochemistry.; Valentine,
J. S., Foote, C. S., Greenberg, A., Liebman, J. F., Eds.; Blackie Academic
and Professional, Chapman and Hall: Glasgow, U.K., 1995; pp 276-312.
(7) Stubbe, J. AdV. Enzymol. 1989, 63, 349-419.
(14) Dong, Y.; Fujii, H.; Hendrich, M. P.; Leising, R. A.; Pan, G.;
Randall, C. R.; Wilkinson, E. C.; Zang, Y.; Que, L., Jr.; Fox, B. G.;
Kauffmann, K.; Mu¨nck, E. J. Am. Chem. Soc. 1995, 117, 2778-2792.
(15) Dong, Y.; Que, L., Jr.; Kauffmann, K.; Mu¨nck, E. J. Am. Chem.
Soc. 1995, 117, 11377-11378.
(8) Lipscomb, J. D. Annu. ReV. Microbiol. 1994, 48, 371-399.
S0002-7863(96)03852-8 CCC: $14.00 © 1997 American Chemical Society

4198 J. Am. Chem. Soc., Vol. 119, No. 18, 1997
Zang et al.
[Fe(6-Me2TPA)(CH3CN)2](ClO4)2‚4H2O (3). Anal. Calcd for C24-
H36Cl2FeN6O12: C, 39.62; H, 4.99; N, 11.55. Found: C, 39.82; H,
4.76; N, 11.29.
[Fe(6-Me3TPA)(CH3CN)2](ClO4)2 (4). Anal. Calcd for C25H30-
Cl2FeN6O8: C, 43.86; H, 4.52; N, 12.56. Found: C, 43.45; H, 4.81;
N, 12.71.
complex consisting of antiferromagnetically coupled high-spin
FeIII and FeIV centers. The properties of these complexes relate
them to transient species observed in the oxygen activation
cycles of methane monooxygenase17-19 and ribonucleotide
reductase.20-22
These TPA-based ligands exhibit a versatility that allows us
to alter the spin state of the iron center and dramatically affect
its spectroscopic properties by simply introducing a methyl
substituent on the pyridine rings of TPA. Here we report the
syntheses, structures, and properties of a series of iron(II) and
iron(III) complexes of TPA and its methyl-substituted deriva-
tives to provide the structural basis for the effects of introducing
6-methyl groups on the pyridine rings of TPA. These results
are interpreted in the context of the properties of metastable
peroxo and oxo species that have been generated as models for
intermediates of nonheme iron enzymes.
The three iron(III) acac complexes were synthesized using the
following procedure. Equimolar amounts of Fe(ClO4)3‚6H2O and ligand
were dissolved in methanol. To this solution was added 1 equiv each
of Hacac and Et3N in methanol, turning the solution deep purple in
color. After the solution stood overnight in the refrigerator, dark
crystalline needles were obtained.
[Fe(TPA)(acac)](ClO4)2 (5). Anal. Calcd for C23H25Cl2FeN4O10:
C, 42.88; H, 3.91; N, 8.70. Found: C, 42.69; H, 4.05; N, 8.63.
[Fe(5-Me3TPA)(acac)](ClO4)2‚H2O (6). Anal. Calcd for C26H33-
Cl2FeN4O11: C, 44.33; H, 4.72; N, 7.95. Found: C, 44.49; H, 4.81;
N, 7.97.
[Fe(6-MeTPA)(acac)](ClO4)2 (7). Anal. Calcd for C24H27Cl2Fe-
N4O10: C, 43.79; H, 4.13; N, 8.51. Found: C, 43.72; H, 4.48; N, 8.33.
tBuOOH-d9 was synthesized by adding an ethereal solution of
tBuMgCl-d9 via a syringe pump to a flask containing diethyl ether at
-78 °C while simultaneously bubbling dry O2 gas through the diethyl
ether, and purified according to the method of Walling and Buckler.25
The Grignard reagent was prepared according to standard methods26
Experimental Section
Syntheses. TPA‚3ClO4, 5-Me3TPA, and 6-Me3TPA were synthe-
sized according to literature methods.12,14,23,24 Commercially available
chemicals were purchased and used without further purification.
Elemental analyses were performed at MHW Laboratories (Phoenix,
AZ). Caution: Perchlorate salts are potentially explosiVe and should
be handled with care.
t
t
from BuCl-d9, which was prepared from BuOH-d9 and concentrated
HCl.
tBuO18OH was synthesized by the reaction of tBuOH with H18OF.27
The latter was prepared as a stabilized 0.6 M solution in acetonitrile
by passing molecular fluorine (20% in Ar) through 11 mL of a chilled
8% v/v solution of 18O-enriched water (96 atom % 18O, Monsanto
Research Corp.) in CH3CN.28-30 The resulting solution was shaken
with silica gel to remove excess HF and residual water. It was then
Bis[(6-methyl-2-pyridyl)methyl](2-pyridylmethyl)amine (6-Me2-
TPA). An aqueous NaOH solution (1 M, 5 mL) was added dropwise
to a solution of (2-pyridylmethyl)amine (0.01 mol) and 2-(chloro-
methyl)-6-methylpyridine hydrochloride (0.02 mol) in 15 mL of water
at 0 °C with stirring. The reaction mixture was then stirred for 24 h
at room temperature and then extracted with CHCl3. The extracts were
then washed with water to remove residual NaCl, and dried over Na2-
SO4. Removal of the solvent yielded a brown solid (80% yield). 1H
NMR (CDCl3): δ (ppm) 8.6 (1H, R-py), 7.9, 7.8, and 7.6 (6H, â-py),
7.3 (3H, γ-py), 4.5 (2CH2), 4.3 (CH2), 2.8 (2CH3).
Bis(2-pyridylmethyl)[(6-methyl-2-pyridyl)methyl]amine (6-MeT-
PA). An aqueous NaOH solution (1 M, 3 mL) was added dropwise to
a solution of bis(2-pyridylmethyl)amine (0.01 mol) and 2-(chloro-
methyl)-6-methylpyridine hydrochloride (0.01 mol) in 15 mL of water
at 0 °C with stirring. The reaction mixture was then stirred for 24 h
at room temperature and then extracted with CHCl3. The extracts were
then washed with water to remove residual NaCl, and dried over Na2-
SO4. Removal of the solvent yielded a dark yellow solid (70% yield).
1H NMR (CDCl3): δ (ppm) 8.8 (2H, R-py), 8.2, 7.8 (6H, â-py), 7.7
(3H, γ-py), 4.2 (CH2), 4.1 (2CH2), 2.7 (CH3).
t
mixed with 2 mL of BuOH and allowed to stand for several hours at
room temperature to ensure complete reaction. Residual HF was
removed by stirring overnight with calcium carbonate (Merck SupraPur)
and filtering. Concentration by evaporation at room temperature (with
t
some loss) yielded 3 mL of a ca. 0.7 M solution of labeled BuOOH.
Preparation of Intermediates. All intermediates were prepared
according to the general procedure outlined below. A 2-5 mM
acetonitrile solution of the iron(II)-acetonitrile complex was cooled
to -40 °C, and then 5 equiv of tBuOOH was added to this cold solution.
The deep blue or purple (depending on ligand) intermediate which forms
t
immediately upon addition of BuOOH was used for spectroscopic
measurements.
Physical Methods. 1H NMR spectra were recorded on a Varian
Unity 300 or 500 spectrometer at ambient temperature. Chemical shifts
(ppm) were referenced to the residual protic solvent peaks. EPR spectra
were obtained at liquid helium temperatures on a Varian E-109
spectrometer equipped with an Oxford ESR-10 cryostat. 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.
Electrochemical studies were carried out with a BAS 100 electro-
chemical analyzer (Bioanalytical Systems, Inc., West Lafayette, IN)
in acetonitrile using 0.1 M tetraethylammonium perchlorate as the
supporting electrolyte. Cyclic voltammograms (CV) were obtained by
using a three-component system consisting of a platinum disk working
electrode, a platinum wire auxiliary electrode, and a silver wire as the
reference electrode. Potentials were corrected to the NHE standard
by measuring the ferrocenium/ferrocene couple under the same condi-
tions (+400 mV vs NHE).
The syntheses of the FeII complexes were all carried out under Ar
by Schlenk line techniques following a common procedure. Equimolar
amounts of ligand and Fe(ClO4)2‚6H2O were dissolved in acetonitrile,
and the complex was precipitated by adding diethyl ether to this clear
solution. Recrystallization from CH3CN/ether afforded pure complexes
that gave satisfactory elemental analyses.
[Fe(TPA)(CH3CN)2](ClO4)2 (1). Anal. Calcd for C22H24Cl2Fe-
N6O8: C, 43.13; H, 3.86; N, 13.40. Found: C, 42.87; H, 3.86; N,
12.91.
[Fe(6-MeTPA)(CH3CN)2](ClO4)2 (2). Anal. Calcd for C23H26Cl2-
FeN6O8: C, 43.07; H, 4.06; N, 13.11. Found: C, 43.24; H, 4.31; N,
12.96.
(16) Que, L., Jr.; Dong, Y. Acc. Chem. Res. 1996, 29, 190-196.
(17) Lee, S.-K.; Fox, B. G.; Froland, W. A.; Lipscomb, J. D.; Mu¨nck,
E. J. Am. Chem. Soc. 1993, 115, 6450-6451.
(18) Lee, S.-K.; Nesheim, J. C.; Lipscomb, J. D. J. Biol. Chem. 1993,
268, 21569-21577.
(25) Walling, C.; Buckler, S. A. J. Am. Chem. Soc. 1955, 77, 6032-
6038.
(19) 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.
(26) Furniss, B. S.; Hannaford, A. J.; Smith, P. W. G.; Tatchell, A. R.
Vogel’s Textbook of Practical Organic Chemistry; Longman Scienticfic &
Technical: New York, 1989; pp 534-535.
(27) Appelman, E. H.; Thompson, R. C.; Engelkemeir, A. G. Inorg.
Chem. 1979, 18, 909-911.
(28) Rozen, S.; Brand, M. Angew. Chem., Int. Ed. Engl. 1986, 25, 554-
555.
(29) Rozen, S.; Brand, M.; Kol, M. J. Am. Chem. Soc. 1989, 111, 8325-
8326.
(20) Bollinger, J. M., Jr.; Edmondson, D. E.; Huynh, B. H.; Filley, J.;
Norton, J.; Stubbe, J. Science (Washington, D.C.) 1991, 253, 292-298.
(21) Ravi, N.; Bollinger, J. M., Jr.; Huynh, B. H.; Edmondson, D. E.;
Stubbe, J. J. Am. Chem. Soc. 1994, 116, 8007-8014.
(22) 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.
(23) Anderegg, G.; Wenk, F. HelV. Chim. Acta 1967, 50, 2330-2332.
(24) Gafford, B. G.; Holwerda, R. A. Inorg. Chem. 1989, 28, 60-71.
(30) Appelman, E. H.; Dunkelberg, O.; Kol, M. J. Fluorine Chem. 1992,
56, 199-213.

Models for Nonheme Iron Intermediates
J. Am. Chem. Soc., Vol. 119, No. 18, 1997 4199
Table 1. Crystallographic Data
1
4
6
7
formula
formula weight, amu
space group
a, Å
b, Å
c, Å
C70H64B2FeN6
1066.79
Pbca
20.030(8)
17.00(1)
33.22(3)
90
11312(10)
8
1.265
C73H80B2FeN6
1108.87
P212121
14.398(5)
17.172(7)
24.503(8)
90
6058(6)
4
1.227
C26H31Cl2FeN4O10
686.31
C24H27Cl2FeN4O10
658.25
P21/n (no. 14)
12.448(3)
17.704(4)
13.515(4)
96.77(3)
2958(2)
4
P21/c (no. 14)
8.7805(1)
9.0849 (3)
35.1580 (9)
96.500 (1)
2789.70 (12)
4
â, deg
V, Å3
Z
D(calc), g cm-3
1.541
1.567
temperature, K
172
0.710 69
3.12
0.094
0.091
172
0.710 73
3.93
0.089
0.083
172
0.710 69
7.5
0.073
0.072
172
λ, Å
µ, cm-1
Ra
0.710 73
0.795
0.0939b
Rw
0.1957c
2
2
4
a R ) (∑|(Fo - Fc)|)/(∑Fo); Rw ) {(∑w|Fo - Fc| )/(∑w(Fo)2)}1/2
.
b R1 ) ∑||Fo| - |Fc||/∑|Fo|. c wR2 ) (∑[w(Fo - Fc2)2]/∑[wFo ])1/2, where w
2
) q/σ2(Fo ) + (aP)2 + bP.
Table 2. Selected Bond Lengths and Bond Angles for
Resonance Raman spectra were recorded on a Spex 1403 double
monochromator interfaced with a Spex DM3000 data system using a
Spectra Physics 2030-15 argon ion laser to pump a 375B CW dye laser
(rhodamine 6G). Spectra were obtained at 77 K using a back-scattering
geometry with λex ) 599.5 nm. Samples were frozen onto a gold-
plated copper cold finger in thermal contact with a dewar containing
liquid nitrogen.31 Raman shifts of spectral features were referenced to
ν1 of frozen acetonitrile (922 cm-1) which was in turn referenced to
the emission spectrum of a low-pressure Hg lamp.
Electrospray ionization mass spectral data were collected using a
Sciex API III mass spectrometer (Sciex, Thornhill, Ontario, Canada).
Cooled CH3CN (dry ice on the syringe) was introduced by a syringe
infusion pump (model 22, Harvard Apparatus, South Natick, MA)
through a 76 µm internal diameter fused silica capillary line for about
20 min, and then the cooled sample solution was introduced. The
instrument was scanned in 0.2 Da steps and a 5-10 ms dwell time per
step; signal averaging was used to enhance the signal-to-noise ratio.
Crystallographic Studies. Crystals of the complexes were mounted
onto a glass fiber and cooled to -100 °C. For the FeII complexes
crystals were coated with a viscous hydrocarbon in an anaerobic box
before being mounted to prevent oxidation. Data collection and analysis
were conducted at the X-ray Crystallographic Laboratory of the
Chemistry Department of the University of Minnesota.
[Fe(TPA)(CH3CN)2]2+ (1) and [Fe(6-Me3TPA)(CH3CN)2]2+ (4)
Bond Lengths (Å)
1
4
Fe1-Namine
Fe1-Npy
(N1)
1.99(1)
1.97(1)
1.92(1)
1.95(1)
1.92(1)
1.93(1)
(N1A)
2.15(1)
2.25(1)
2.18(1)
2.24(1)
2.17(1)
2.17(1)
(N11)
(N21)
(N31)
(N41)
(N51)
(N11A)
(N31A)
(N21A)
(N41A)
(N51A)
Fe1-Nacetonitrile
Bond Angles (deg)
1
4
N1-Fe1-N11
N1-Fe1-N21
N1-Fe1-N31
N1-Fe1-N41
N1-Fe1-N51
N21-Fe1-N41
N41-Fe1-N51
84.7(5)
85.6(6)
82.8(5)
178.6(6)
93.8(5)
93.7(6)
86.8(6)
N1A-Fe1-N11A
N1A-Fe1-N31A
N1A-Fe1-N21A
N1A-Fe1-N41A
N1A-Fe1-N51A
N31A-Fe1-N41A
N41A-Fe1-N51A
76.9(5)
81.8(5)
76.2(5)
170.1(5)
87.7(5)
107.8(5)
82.5(5)
Table 3. Selected Bond Lengths and Bond Angles for
For crystals 1, 4, and 6, data were collected on an Enraf-Nonius
CAD-4 diffractometer with graphite monochromated Mo KR (λ )
0.710 69 Å) radiation. Cell constants and orientation matrices for data
collection were obtained from a least squares refinement using the
setting angles of 25 carefully centered reflections. The intensities of
three representative reflections were measured every 50 min of X-ray
exposure time throughout the data collection to ascertain crystal
integrity, and no decay in intensity was observed. All data were
corrected for empirical absorption and Lorentz and polarization effects.
Pertinent crystallographic data and experimental conditions are sum-
marized in Table 1. The structures were solved by direct methods.
All non-hydrogen atoms were refined either anisotropically or isotro-
pically, and hydrogen atoms were placed in calculated positions (dC-H
) 0.95 Å) and assigned thermal parameters that were 20% greater than
the Bequiv value of the atom to which they were bonded and not refined.
Refinement was carried out with full matrix least squares on F with
scattering factors from ref 32 and included anomalous dispersion terms.
Selected bond lengths and bond angles are listed in Tables 2 and 3.
For complex 7, data were collected on a Siemens SMART system.
An initial set of cell constants was calculated from reflections harvested
from three sets of 20-30 frames. These initial sets of frames are
oriented such that orthogonal wedges of reciprocal space are surveyed.
This produces orientation matrices determined from 50-300 reflections.
Final cell constants were calculated from a set that did not exceed a
number equal to 8192 of strong reflections from the actual data
collection. Pertinent crystallographic data and experimental conditions
[Fe(5-Me3TPA)(acac)]2+ (6) and [Fe(6-MeTPA)(acac)]2+ (7)
Bond Lengths (Å)
6
7
Fe-Namine
Fe-Npy
(N1)
(N2)
(N3)
(N4)
(O1)
(O2)
1.968(5)
1.980(5)
1.926(5)
1.958(5)
1.902(4)
1.872(4)
(N1)
(N4)
(N2)
(N3)
(O2)
(O1)
2.143(6)
2.161(7)
2.072(6)
2.144(6)
1.951(5)
1.909(5)
Fe-Oacac
Bond Angles (deg)
6
7
N1-Fe-N3
N1-Fe-N2
N1-Fe-N4
O1-Fe-N2
O1-Fe-N4
O2-Fe-N2
O2-Fe-N4
87.1(2)
N1-Fe1-N2
N1-Fe1-N4
N1-Fe-N3
O2-Fe1-N4
O2-Fe1-N3
O1-Fe1-N4
O1-Fe1-N3
82.4(2)
77.8(3)
76.5(3)
89.2(2)
95.0(2)
111.4(2)
94.1(2)
84.0(2)
81.7(2)
92.9(2)
90.8(2)
94.4(2)
99.8(2)
are summarized in Table 1. The structure was 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 parameters.
Selected bond lengths and bond angles are listed in Table 3.
(31) Czernuszewicz, R. C.; Johnson, M. K. Appl. Spectrosc. 1983, 37,
297-298.
(32) Cromer, D. T.; Waber, J. T. International Tables for X-ray
Crystallography; Kynoch Press: Birmingham, England, 1974; Vol. IV.
Results and Discussion
Our efforts to model the active sites of non-heme iron proteins
have focused mainly on the tripodal TPA ligand and its 6-methyl

4200 J. Am. Chem. Soc., Vol. 119, No. 18, 1997
Zang et al.
Figure 1. 1H NMR spectra of complexes 1, 4, 5, and 7 in CD3CN at room temperature. T1 values (ms) are shown on the top of the peak. Solvent
peaks are labeled “s”.
Table 4. 1H NMR, EPR, and Electrochemical Properties of Complexes
1H NMR (ppm)
tripodal ligand
acac
CH3
complexa
R-H (CH3)
â,â′-H (CH3)
γ-H
7.3
-7
-6
-11
-6
-5
-2.3
CH2
6.3
88
82, 37
92
86, 75
83
28.6
32
c
CH3
CH
EPR g values
E1/2 (mV)
1
2
10.9
95
(-24)
106
8.5, 8.4
48, 52
51, 50
56, 59
52, 57
860
940b
3
940b
(-25)
(-28)
28.6
32
4
5
6
7
49, 54
940b
70
20
24, 14.4
16, (1.6)
115, 108
128, 89
-18.5
-17.6
35
-18.5
-17.6
9
-65
-63.4
-21
2.57, 235, 1.71
2.57, 2.35, 1.71
4.50, 430, 4.15
-3.4
-6.5
1.2
c
175
(-5)
c
a 1 ) [Fe(TPA)(CH3CN)2](ClO4)2; 2 ) [Fe(6-MeTPA)(CH3CN)2](ClO4)2; 3 ) [Fe(6-Me2TPA)(CH3CN)2](ClO4)2; 4 ) [Fe(6-Me3TPA)(CH3CN)2](ClO4)2;
5 ) [Fe(TPA)(acac)](ClO4)2; 6 ) [Fe(5-Me3TPA)(acac)](ClO4)2; 7 ) [Fe(6-MeTPA)(acac)](ClO4)2. b Only an irreversible oxidation peak observed.
c Too broad to be observed.
derivatives. When 6-methyl substituents on the pyridine rings
are introduced, we observe significant changes in the properties
of the corresponding complexes.12,14,15 We have thus synthe-
sized a series of FeII and FeIII complexes with TPA, 6-MeTPA,
6-Me2TPA, and 6-Me3TPA in order to study systematically the
effects of increasing 6-methyl substitution on the properties of
the iron centers in these complexes and establish the basis for
these dramatic effects.
comparison to the spectra of other high-spin FeIITPA33 and
6-Me3TPA complexes.10 The 1H NMR spectra of 2 and 3 (Table
4) show that both are high-spin iron(II) complexes. Thus, the
introduction of even one 6-methyl substituent to TPA is
sufficient to change the iron(II) center from low spin to high
spin in the [FeIIL(CH3CN)2]2+ series.
Cyclic voltammetric measurements on 1-4 (Table 4) show
that the iron(II) oxidation state is greatly favored by these ligand
environments. Complex 1 shows a quasi-reversible cyclic
voltammogram with E1/2 ) 860 mV vs NHE (∆E ) 80 mV,
ipa/ipc ) 0.75), its high redox potential may be attributed to the
thermodynamic stability of the low-spin Fe(II) state by analogy
to [Fe(bpy)3]2+ and related complexes.34 Complexes 2-4 are
also oxidized at high potential (∼940 mV), but the oxidations
are irreversible in these cases. Since these three are all high-
spin complexes, factors other than spin state must disfavor the
iron(III) oxidation state in these complexes.
Iron(II) Complexes. The synthesis of the series of iron(II)
complexes [Fe(L)(CH3CN)2]2+ (L ) TPA, 6-MeTPA, 6-Me2-
TPA, and 6-Me3TPA) is straightforward; crystalline products
were obtained by simply mixing iron(II) perchlorate and the
appropriate ligand in acetonitrile and precipitating the desired
complex with ether. However, these iron(II) complexes have
very different properties. The 1H NMR spectrum of 1 in CD3-
CN at room temperature exhibits features of a diamagnetic
species, demonstrating that the FeII center in 1 is low spin
(Figure 1, Table 4). Unlike the low-spin complex 1, complex
Iron(III) Complexes. Unlike the iron(II) complexes, the
synthesis of a similar series of [FeIIIL(acac)]2+ complexes
1
4 has a H NMR spectrum typical of a high-spin FeII complex
(33) Zang, Y.; Que, L., Jr. Inorg. Chem. 1995, 34, 1030-1035.
(34) Fukuzumi, S.; Kochi, J. K. J. Am. Chem. Soc. 1982, 104, 7599-
7609.
with peaks paramagnetically shifted over a range of -30 to +80
ppm (Figure 1, Table 4). These peaks can be assigned by

Models for Nonheme Iron Intermediates
J. Am. Chem. Soc., Vol. 119, No. 18, 1997 4201
Figure 2. X-band EPR spectra recorded at 2 K. Typical EPR conditions: microwave frequency 9.23 GHz; microwave power 0.2 mW; modulation
amplitude 10 G; modulation frequency 100 KHz. (a) [Fe(TPA)(acac)](ClO4)2 (5). (b) [Fe(6-MeTPA)(acac)](ClO4)2 (7). (c) [Fe(TPA)(H2O)(OOt-
Bu)](ClO4)2 (1a). (d) [Fe(6-Me3TPA)(H2O)(OOtBu)](ClO4)2 (4a).
presented varying degrees of difficulty. The two ligands without
6-substituents, TPA and 5-Me3TPA, form stable iron(III)
complexes, but the stability of the complexes decreases sig-
nificantly as the number of 6-methyl substituents increases.
6-MeTPA forms an iron(III) complex that decomposes slowly
in solution over a period of weeks, while attempts to synthesize
[Fe(6-Me2TPA)(acac)]2+ and [Fe(6-Me3TPA)(acac)]2+ by mix-
ing the ligand, FeIII(ClO4)3, and Hacac have produced free ligand
and Fe(acac)3 instead. We tried to prepare [Fe(6-Me2TPA)-
(acac)]2+ and [Fe(6-Me3TPA)(acac)]2+ by oxidizing the corre-
sponding iron(II) complexes, but only observe the yellow
solutions of [FeL(acac)](ClO4) (L ) 6-Me2TPA, 6-Me3TPA)
convert to the red color characteristic of Fe(acac)3. These results
show that the introduction of two or three 6-methyl substituents
to the TPA ligand significantly destabilizes the [FeIIIL(acac)]2+
complexes.
three 5-methyl substituents. However, the opposite shift is
observed on going from 5 to 7, which has one 6-methyl
substituent, showing that the iron(II) oxidation state is favored
in 7. This shift is consistent with the observation that FeIIIacac
complexes of 6-Me2TPA and 6-Me3TPA could not be isolated.
Similar upshifts in the redox potentials of dimanganese com-
plexes have been noted by Hodgson and co-workers upon
introduction of a 6-methyl substituent.37 Crystal structures of
these iron complexes provide the rationale for this behavior.
Crystal Structures. ORTEP plots for the cations 1 and 4
are shown in Figure 3. The iron(II) center in each complex is
in a distorted octahedral environment ligated by the three
pyridine nitrogens and one amine nitrogen of the tripodal ligand
and two acetonitrile nitrogens. Although 1 and 4 have similar
coordination geometries, the Fe-N bond lengths and bond
angles differ greatly between these two complexes, reflecting
the difference in spin state (Table 2). For example, the Fe-
Namine distance of 1.99(1) Å in 1 is significantly shorter than
the Fe-Namine distance of 2.15(1) Å in 4. Likewise, the Fe-
Npy bond distances of 1 are 0.25-0.29 Å shorter than the
corresponding Fe-Npy bonds in 4. The average Fe-Npy
distance of 1.95 Å found in 1 is comparable to those typically
found in low-spin iron(II) complexes, such as [Fe(phen)3]2+
(1.97 Å),38 [Fe(bpy)3]2+ (1.96 Å),39 [Fe(tpy)2]2+ (1.96 Å),40 [Fe-
(BPA)2]2+ (1.97 Å),41 and [Fe(bpca)2] (1.94 Å).42 On the other
hand, the average Fe-Npy of 2.22 Å in 4 is comparable to those
found in [Fe2(6-Me3TPA)2F2]2+ (2.24 Å),43 [Fe(6-Me3TPA)-
(OBz)]+ (2.21 Å),12 and [Fe(6-Me3TPA)(BF)]+ (2.22 Å),10 but
slightly longer than those found in high-spin FeIITPA com-
plexes.10,33,43,44 The bond length difference between high-spin
TPA and 6-Me3TPA complexes is very likely due to the steric
effects of the 6-methyl substituents which prevent the pyridines
from approaching the metal center too closely. Similar elonga-
[Fe(TPA)(acac)]2+ (5) (Figure 2a) and [Fe(5-Me3TPA)-
(acac)]2+ (6) show EPR signals at g ) 2.57, 2.35, and 1.71,
typical of low-spin FeIII centers.35 On the other hand, 7 shows
EPR signals (Figure 2b) at g1 ) 4.50, g2 ) 4.30, and g3 )
4.15, which can be ascribed to the middle Kramers doublet of
an S ) 5/2 system with E/D ) 0.31, demonstrating that the
FeIII in 7 is high spin. 1H NMR shifts of the iron(III) complexes
are consistent with the EPR results (Table 4). [Fe(TPA)-
(acac)]2+ (5) (Figure 1) and [Fe(5-Me3TPA)(acac)]2+ (6) exhibit
1H NMR spectra with relatively sharp signals characteristic of
low-spin FeIII complexes. [Fe(6-MeTPA)(acac)]2+ (7), on the
1
other hand, exhibits a H NMR spectrum (Figure 1, Table 4)
typical of mononuclear high-spin FeIII complexes with larger
paramagnetic shifts and broader signals like those in [Fe(TPA)-
X2]+ (X ) Cl-, Br-) complexes.36 The peak assignments are
listed in Table 4.
The effects of introducing a methyl group are also reflected
in the redox properties of the FeIIIacac complexes. Complexes
5-7 each display a reversible redox wave with E1/2 values at
70 mV (∆E ) 75 mV, ipa/ipc ) 1.00), 20 mV (∆E ) 70 mV,
ipa/ipc ) 1.01), and 175 mV (∆E ) 90 mV, ipa/ipc ) 1.01) relative
to NHE, respectively. Methyl substitution on the pyridine ring
of the tripodal ligand should make the pyridine more basic and
lower the redox potential of the metal center as observed in the
50 mV decrease in potential on going from 5 to 6, which has
(37) Goodson, P. A.; Oki, A. R.; Glerup, J.; Hodgson, D. J. J. Am. Chem.
Soc. 1990, 112, 6248-6254.
(38) Zalkin, A.; Templeton, D. H.; Ueki, T. Inorg. Chem. 1973, 12,
1641-1646.
(39) Posse, G. M. E.; Juri, M. A.; Aymonino, P. J.; Piro, O. E.; Negri,
H. A.; Castellano, E. E. Inorg. Chem. 1984, 23, 948.
(40) Baker, A. T.; Goodwin, H. A. Aust. J. Chem. 1985, 38, 207-214.
(41) Butcher, R.; Addison, A. W. Inorg. Chim. Acta 1989, 158, 211-
215.
(42) Wocadlo, S.; Massa, W.; Folgado, J. Inorg. Chim. Acta 1993, 207,
199-206.
(43) Zang, Y.; Jang, H. G.; Chiou, Y.-M.; Hendrich, M. P.; Que, L., Jr.
Inorg. Chim. Acta 1993, 213, 41.
(44) Me´nage, S.; Zang, Y.; Hendrich, M. P.; Que, L., Jr. J. Am. Chem.
Soc. 1992, 114, 7786-7792.
(35) Burger, R. M.; Peisach, J.; Horwitz, S. B. J. Biol. Chem. 1981, 256,
11636-11644.
(36) Kojima, T.; Leising, R. A.; Yan, S.; Que, L., Jr. J. Am. Chem. Soc.
1993, 115, 11328-11335.

4202 J. Am. Chem. Soc., Vol. 119, No. 18, 1997
Zang et al.
Figure 4. ORTEP plots of [Fe(5-Me3TPA)(acac)]2+ (6) and [Fe(6-
MeTPA)(acac)]2+ (7). Ellipsoids are drawn at the 50% probability level.
Hydrogen atoms are omitted for clarity.
Figure 3. ORTEP plots of [Fe(TPA)(CH3CN)2]+ (1) and [Fe(6-Me3-
TPA)(CH3CN)2]+ (4). Ellipsoids are drawn at the 50% probability level.
Hydrogen atoms are omitted for clarity.
differ significantly between 6 and 7 (Table 3). In 6, where the
5-methyl substituents point away from the iron, there are no
steric effects. The average Fe-Npy distance of 1.96 Å and Fe-
Namine of 1.968(5) Å are about 0.2 Å shorter than those of high-
spin iron(III) TPA complexes.9,46,49 The Fe-O bonds of
1.872(5) and 1.902(5) Å in 6 are also 0.1 Å shorter than those
of Fe(acac)3 (av 1.99 Å).50 These results are consistent with
the low-spin state of the iron in 6. On the other hand, the
introduction of a 6-methyl group on only one pyridine in 7
results in the lengthening of the average Fe-Npy distance to
2.125 Å, comparable to corresponding values for [Fe(TPA)-
Cl2]+ (2.150 Å)36 and other high-spin FeIII(TPA) complexes.
The Fe-Oacac bonds in 7 have also lengthened to 1.909(5) and
1.951(5) Å. The average N1-Fe-Npy angle is 84.3° in 6
and 78.9° in 7.
The 6-methylpyridine pendant ligand has the longest of the
three Fe-Npy distances (2.161(7) Å), reflecting the steric effect
of the 6-methyl group. It is interesting to note that the tripodal
ligand does not adopt the most symmetric configuration in the
structure of 7. Such a configuration would entail having the
6-methylpyridine pendant ligand coordinate trans to an acac
oxygen and in the plane of the acac. This arrangement would
cause the 6-methyl substituent to run into an acac methyl group,
but is avoided by having the 6-methylpyridine pendant occupy
a position cis to the oxygen and perpendicular to the acac plane.
Other structural features also reflect the steric effects of the
6-methyl group in 7. The acac ligand is bent away from the
pendant 6-methylpyridine to avoid the steric interactions as
indicated by O1-Fe-N4 (111.3°) which is much bigger than
O1-Fe-N3 (94°). In 6 the corresponding angles are 99.8°
and 94.4°, respectively.
tion of the metal-to-nitrogen distances upon 6-methyl substitu-
tion has also been observed by Hodgson et al.37 and Suzuki et
al.45
The steric effects of the 6-methyl groups are also reflected
by the bond angles. The coordination geometry is more
distorted from an ideal octahedron in 4. The angles that differ
most significantly are N21-Fe-N41 in 1 (93.7(6)°) and the
corresponding N31A-Fe-N41A in 4 (107.8(5)°), because the
6-methyl substituent (C27, Figure 3) pushes the acetonitrile
away. This effect is also seen in N1-Fe-N41 which is
178.6(6)° in 1, very close to the ideal 180°, but only 170.1(5)°
in 4. The average N1-Fe-Npy angle is 84.4° in 1 and is
78.3° in 4, smaller than 90° because of the five-membered
chelate ring size imposed by the tripodal ligands. Other high-
spin Fe(TPA) and Fe(6-Me3TPA) complexes also exhibit small
Namine-Fe-Npy angles, averaging ca. 77°.43,46
ORTEP plots for the iron(III) complexes [Fe(5-Me3TPA)-
(acac)]2+ (6) and [Fe(6-MeTPA)(acac)]2+ (7) are shown in
Figure 4. The FeIII center is six-coordinate in each complex
with four nitrogens from the tripodal ligand and two oxygens
from the acac. The acac is chelated slightly asymmetrically to
the iron with the shorter Fe-O bond trans to the amine nitrogen
(∆rFe-O ) 0.03 Å for 7 compared to ∆rFe-O ) 0.05 Å for [Fe-
(salen)(acac)]47 and ∆rFe-O ) 0.09 Å for [Fe(NTB)(acac)]2+ 48
). Like the iron(II) complexes, the bond lengths and bond angles
(45) 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.
These structural features rationalize the relative stabilities of
the [FeIIIL(acac)] series of complexes. For 6-Me3TPA with
6-methyl groups on all three pyridines, one of the methyl groups
will run into a methyl group of the acac ligand; hence, [FeIII(6-
(46) Norman, R. E.; Yan, S.; Que, L., Jr.; Sanders-Loehr, J.; Backes,
G.; Ling, J.; Zhang, J. H.; O’Connor, C. J. J. Am. Chem. Soc. 1990, 112,
1554-1562.
(47) Lauffer, R. B.; Heistand, R. H.; Que, L., Jr. Inorg. Chem. 1983, 22,
50.
(48) Nazikkol, C.; Wegner, R.; Bremer, J.; Krebs, B. Z. Anorg. Allg.
Chem. 1996, 622, 329-336.
(49) Me´nage, S.; Que, L., Jr. New J. Chem. 1991, 15, 431-438.
(50) Iball, J.; Morgan, C. H. Acta Crystallogr. 1967, 33, 239-244.

Models for Nonheme Iron Intermediates
J. Am. Chem. Soc., Vol. 119, No. 18, 1997 4203
Figure 6. Absorption spectra of [Fe(TPA)(H2O)(OOtBu)](ClO4)2 (1a)
and [Fe(6-Me3TPA)(H2O)(OOtBu)](ClO4)2 (4a) in CH3CN at -40 °C.
Applications to the Transient Intermediates. Our system-
atic study of complexes 1-7 has demonstrated that the
introduction of a 6-methyl substituent on a pyridine of a TPA
ligand exerts a significant effect on the spin state of the iron
center, engendering dramatic changes in the properties of the
iron complex and its derivatives. Due to steric interactions with
the iron center, the 6-methyl-substituent prevents the pyridine
ligand from approaching the iron center too closely, making
the 6-methyl-substituted pyridine a much weaker ligand than
its intrinsic basicity may initially suggest. Thus, the 6-methyl-
substituted TPAs favor iron states with larger ionic radii, viz.,
high-spin over low-spin and FeII over FeIII. Consequently low-
spin Fe(TPA) complexes become high-spin upon the introduc-
tion of a 6-methyl substituent. The stabilization of the iron(II)
state for ligands with 6-methyl substituents is reflected in the
higher redox potentials of 2-4 and the instability of the iron-
(III)acac complexes of tripodal ligands with 6-methyl substit-
uents.
Figure 5. Steric interactions of the 6-methyl substituent with the iron
atom. The top structure is from the crystal structure of 7 with the acac
ligand omitted for clarity. The middle structure is generated with CSC
Chem3D with Fe-Npy bonds of 1.95 Å. A space-filling model to
illustrate the Fe-H interaction is shown on the bottom.
The effects of a 6-methyl substituent are also reflected in the
properties of transient alkylperoxoiron(TPA) species we have
characterized in our efforts to understand the oxygen activation
mechanisms of mononuclear nonheme iron active sites. Thus,
t
treatment of 1 with a slight excess of BuOOH at -40 °C in
Me3TPA)(acac)]2+ is unstable. The corresponding 6-Me2TPA
complex [FeIII(6-Me2TPA)(acac)]2+ is also unstable, even though
the two 6-methylpyridines can coordinate trans to each other
and avoid running into a methyl group of the acac; however,
with trans-6-methyl groups on the tripodal ligand, the acac has
no space to bend away from the 6-methylpyridine pendants to
relieve the steric interactions and yet remain coordinated to the
iron.
Careful scrutiny of space-filling models of 4 and 7 as well
as [Fe2O2(6-Me3TPA)2]2+ 51 and [Fe2O(OH)(6-Me3TPA)2]3+ 52
provides some insight into the steric effects of the 6-methyl
substituent. Figure 5 depicts the interaction between the
6-methyl group and the iron center. The top structure is from
the crystal structure of 7 with the acac ligand omitted for clarity.
It shows that a hydrogen on the 6-methyl group is only 2.7 Å
away from the iron center. Shortening the Fe-Npy bond by
0.2 Å as required by a low-spin iron center (middle structure in
Figure 5 as generated with CSC Chem3D) would increase the
unfavorable steric interactions between the 6-methyl substituent
and the iron center; these interactions prevent the pyridyl
nitrogen from getting any closer to the iron center than ca. 2.15
Å. Thus, the introduction of even one 6-methyl substituent
affords high-spin complexes. The addition of two other
6-methyl groups gives rise to additional steric interactions that
result in even longer Fe-N bonds (∼2.23 Å) as found in all
structurally characterized 6-Me3TPA complexes.43,51
CH3CN affords a transient blue intermediate, 1a, with spectro-
scopic properties essentially identical to those of the compound
derived from the reaction of [Fe2O(TPA)2(H2O)2]4+ with excess
tBuOOH.11 Species 1a exhibits an absorption spectrum (Figure
6) with a maximum near 600 nm (ꢀ ≈ 2200 M-1 cm-1) and
EPR signals (Figure 2c) at g ) 2.19, 2.14, and 1.98, which
indicate that it is a low-spin iron(III) complex. Double
integration reveals that the EPR signals account for all the iron,
indicating that the FeII center in 1 is completely oxidized to the
FeIII oxidation state. Electrospray ionization mass spectral
analysis shows that 1a is an alkylperoxoiron(III) complex. The
positive ion mass spectrum exhibits two intense peaks at m/z
) 452 and 534, corresponding to the ions {[FeIII(TPA)(O-
H)(OOtBu)]}+ and {[FeIII(TPA)(OOtBu)](ClO4)}+, respectively
(Figure 7). Similarly, its negative ion mass spectrum shows
two peaks at m/z ) 650 and 732 corresponding to the ions
{[FeIII(TPA)(OH)(OOtBu)](ClO4)2}- and {[FeIII(TPA)(OOt-
Bu)](ClO4)3}-, respectively. The formulations for these ions
are all corroborated by isotope distribution patterns that match
the calculated patterns. In this regard, the perchlorate counte-
rions are particularly useful as they produce isotope patterns
diagnostic of the number of anions and therefore the charge of
the complex cation.53 From the nature of the ions produced,
we deduce that 1a has the composition [FeIII(TPA)(OH2)(OOt-
Bu)]ClO4.
Related alkylperoxo species can also be obtained by the
reaction of tBuOOH with 2-4 in CH3CN at -40 °C. As with
(51) Zang, Y.; Dong, Y.; Que, L., Jr.; Kauffnann, K.; Mu¨nck, E. J. Am.
Chem. Soc. 1995, 117, 1169-1170.
(52) Zang, Y.; Pan, G.; Que, L., Jr.; Fox, B.; Mu¨nck, E. J. Am. Chem.
Soc 1994, 116, 3653-3654.
(53) Kim, J.; Dong, Y.; Larka, E.; Que, L., Jr. Inorg. Chem. 1996, 35,
2369-2372.

4204 J. Am. Chem. Soc., Vol. 119, No. 18, 1997
Zang et al.
Figure 8. Resonance Raman spectra of intermediates obtained with
599 nm excitation on frozen CH3CN solutions: (a) 1a (from tBuOOH),
(b) 1a (from tBuO18OH), (c) 1a (from tBuOOH-d9). Solvent bands are
labeled “s”.
Figure 7. Positive ion electrospray ionization mass spectra of [Fe-
(TPA)(H2O)(OOtBu)](ClO4)2 (1a) and [Fe(6-Me3TPA)(H2O)(OOtBu)]-
(ClO4)2 (4a).
complexes, the parent TPA ligand gives rise to a low-spin iron-
(III)-alkylperoxo species, while the 6-Me2TPA and 6-Me3TPA
ligands afford the corresponding high-spin iron(III) species. The
introduction of one 6-methyl substituent engenders an interme-
diate situation with both spin states accessible.
Table 5. Spectroscopic Properties of the Alkylperoxo
Intermediates
λmax
Raman shifts (cm-1
)
intermediatea (nm) EPR g values
The low-spin and the high-spin alkylperoxo intermediates
each show a distinctive pattern in their resonance Raman spectra.
Low-spin 1a exhibits three peaks at 796, 696, and 490 cm-1
(Figure 8), identical to those observed for the intermediate
derived from [Fe2O(TPA)2(H2O)2]4+ and tBuOOH11 and similar
to the low-spin iron(III) species obtained from [Fe(III)2O(bpy)4-
1a
600 2.19, 2.14, 1.98 490
696 796
1a(tBuO18OH)
1a(tBuOOH-d9)
2a
488 638, 657, 672 778
450
598 2.20, 2.12, 1.97 488
668, 700 803
682 790
4.3
552 4.3
562 4.3
469
470
468
463
422
464
648 842 878
3a
4a
647 843 881
637 842 877
612 829 867
t
(H2O)2](ClO4)4 and BuOOH.54 The 796 and 696 cm-1 peaks
4a(tBuO18OH)
4a(tBuOOH-d9)
8a
are sensitive to the introduction of both 18O into the terminal
622
860
2
peroxo oxygen and H into the alkyl methyl groups, while the
514 4.3
648 844 880
490 cm-1 peak is only sensitive to 2H substitution and shifts to
a [FeL(H2O)(OOtBu)](ClO4)2 in CH3CN (1a, L ) TPA; 2a, L )
6-MeTPA; 3a, L ) 6-Me2TPA; 4a, L ) 6-Me3TPA; 8a, [Fe(6-
Me3TPA)(OBz)(OOtBu)](ClO4)2 in CH2Cl2 (from ref 12).
t
450 cm-1 when BuOOH-d9 is used. The position of the 490
cm-1 peak (below 500 cm-1), the 40 cm-1 shift to lower energy
upon deuteration of the tert-butyl group, and the insensitivity
to 18O substitution all argue for the assignment of this band as
the C-C-C deformation of the tert-butyl group.55 The shifts
in the peaks at 796 and 696 cm-1 are less straightforward to
explain. The peroxide moiety is a strong Raman scatterer, and
so we expect to see an intense feature in the 800-900 cm-1
region resulting from the O-O stretch. The tert-butyl group
complicates the vibrational spectrum of tert-butyl peroxide
because the C-O linkage has a vibration of energy similar to
that of the O-O linkage; therefore, these coordinates couple
extensively. There can also be contributions from the carbon
skeleton as suggested by a normal coordinate analysis of tert-
butyl alcohol.56 This analysis demonstrates that the intense,
polarized vibration at 780 cm-1 in the Raman spectrum of liquid
(CH3)3COH involves approximately 67% C-O stretching and
33% symmetric CC4 stretching. The sensitivity of the peaks at
796 and 696 cm-1 to both 18O at the terminal oxygen and
deuterium in the methyl groups shows that these vibrations
involve motion of the alkyl group as well as the C-O and O-O
1a, intermediates 2a-4a each exhibit a strong absorption band
in the region of 550-600 nm with extinction coefficients of
ca. 2000 M-1 cm-1 (Figure 6, Table 5). The electrospray
ionization mass spectra of 4a (Figure 7) exhibit positive ions
at m/z ) 494 and 576 corresponding to {[FeIII(6-Me3TPA)(OH)-
(OOtBu)]}+ and {[FeIII(6-Me3TPA)(OOtBu)](ClO4)}+, respec-
tively, and negative ions at m/z ) 692 and 774 which correspond
to {[FeIII(6-Me3TPA)(OH)(OOtBu)](ClO4)2}- and {[FeIII(6-Me3-
TPA)(OOtBu)](ClO4)3}-, respectively. The mass difference of
42 units between corresponding ions of 1a and 4a reflects the
mass difference between TPA and 6-Me3TPA. The electrospray
ionization mass spectral data thus demonstrate that the inter-
mediates are alkylperoxoiron(III) complexes with the formula
[FeIII(L)(OH2)(OOtBu)]ClO4.
Unlike 1a, the EPR spectra of intermediates 2a-4a (Table
5) indicate the presence of high-spin FeIII species. Intermediates
3a and 4a exhibit isotropic EPR signals (Figure 2d) at g ) 4.3,
characteristic of a high-spin iron(III) complex in a rhombic
environment. Double integration of these signals reveals that
all the iron centers have been converted to this EPR-active form.
Interestingly, 2a exhibits EPR signals at g ) 4.3 and in the g
) 2 region, indicating that 2a is a mixture of high-spin and
low-spin FeIII components. Thus, as observed in the precursor
(54) Me´nage, S.; Wilkinson, E. C.; Que, L., Jr.; Fontecave, M. Angew.
Chem., Int. Ed. Engl. 1995, 34, 203-205.
(55) Dollish, F. R.; Fateley, W. G.; Bentley, F. F. Characteristic Raman
Frequencies of Organic Compounds; Wiley: New York, 1974; pp 1-6.
(56) Tanaka, C. Nippon Kagaku Zasshi 1962, 83, 398-405.

Models for Nonheme Iron Intermediates
J. Am. Chem. Soc., Vol. 119, No. 18, 1997 4205
6-methyl substituent on a pyridine of a TPA ligand significantly
affects the electronic properties of the metal center.
There are two other examples in the literature of the effects
of 6-methyl substituents on pyridine rings on the properties of
transient iron intermediates. The first example involves the
reversibility of dioxygen adduct formation from diiron(II)
complexes of dinucleating ligands with pendant pyridines. We
have shown that exposure of [FeII (HPTP)(O2CC6H5)]2+ to O2
2
affords an irreversible O2 adduct,59,60 while Suzuki et al. have
found that the introduction of 6-methyl substituents renders such
adducts reversible.45 For the parent complex, the diiron(III) state
is strongly favored, so O2 binding is irreversible. Introducing
the 6-methyl substituents destabilizes the diiron(III) state and
shifts the redox potential of the diiron center to more positive
values, thereby making O2 binding reversible.
The second striking example involves a pair of synthetic bis-
(µ-oxo)diiron(III,IV) complexes which serve as the only syn-
thetic precedents for high-valent intermediates in the reactions
of dioxygen with the reduced nonheme diiron enzymes such as
methane monooxygenase (MMO)17-19 and ribonucleotide re-
ductase (RNR).20-22 The TPA derivative has a valence-
delocalized, presumably double exchange-coupled low-spin
iron(III)-low-spin iron(IV) center with a novel S ) 3/2 ground
state.14 Introducing only one 6-methyl substituent on the ligand
makes the irons high spin, so the 6-MeTPA derivative15 has a
valence-localized antiferromagnetically coupled high-spin iron-
(III)-high-spin iron(IV) center. The TPA derivative has
Mo¨ssbauer parameters that resemble MMO intermediate Q
which is proposed to be a diiron(IV) species,17 while the
6-MeTPA derivative has EPR properties that resemble inter-
mediate X of ribonucleotide reductase.22 These similarities form
the basis for a mechanistic hypothesis in which the Fe2O2
diamond core serves as the common repository for the oxidizing
equivalents in the redox cycles of nonheme diiron enzymes.16
Figure 9. Resonance Raman spectra of intermediates obtained with
514.5 nm excitation on frozen CH3CN solutions: (a) 4a (from
tBuOOH), (b) 3a (from tBuOOH), (c) 2a (from tBuOOH; peaks marked
with an asterisk are from the low-spin isomer). Solvent bands are labeled
“s”.
linkages. We propose that the peak at 696 cm-1 is a vibration
t
of the BuO moiety but also involving some O-O stretching
component. The peak at 796 cm-1 actually moves to higher
energy (803 cm-1) when the tBuOO- ligand is deuterated, which
suggests that the symmetric CC4 stretching component has
become uncoupled from the O-O stretch. Thus, the 796 and
696 cm-1 peaks are assigned to be vibrations comprising O-O
stretching, C-O stretching, and symmetric CC4 stretching of
the tert-butyl group, with greater contribution of O-O stretching
to the 796 cm-1 peak.
The high-spin intermediate 4a shows a different pattern of
Raman shifts with peaks at 877, 842, 637, and 468 cm-1 (Figure
9); a similar pattern is observed for the intermediate 8a derived
from the reaction of [Fe(II)(6-Me3TPA)(OBz)]+ (8) and tBuOOH
(Table 5).12 Unlike the low-spin intermediates, all four peaks
of 4a are affected by the introduction of terminally 18O-labeled
tBuOOH or deuteration of the tert-butyl group (Table 5),
indicating that the entire molecule is involved in all four
vibrations. This greater complexity precludes a detailed as-
signment of the vibrations at present, and more detailed isotope
labeling experiments are needed to fully assign these peaks.57
Nevertheless, it is clear that 1a and 4a exhibit distinct Raman
spectra. Furthermore, high-spin 3a has the same Raman
spectrum as that of 4a, while 2a exhibits the features associated
with both high-spin and low-spin iron(III)-alkylperoxo com-
plexes. Thus, the spectral distinctions correlate with spin state
rather than a difference in peroxide coordination mode as earlier
suggested.12 Indeed the data accumulated thus far for this series
indicate that these complexes are all terminal η1-alkylperoxo
complexes.
In summary, this study provides the structural basis for using
6-methyl substituents on pyridine rings to tune the properties
of an iron center and demonstrates the utility of this simple
ligand design tool to obtain models for transient intermediates
in the catalytic cycles of nonheme iron enzymes.
Acknowledgment. This work was supported by the National
Institutes of Health (Grant GM 33162). The National Science
Foundation provided funds for the purchase of the Siemens
SMART system. We thank Dr. Albert Jache for assistance in
t
the preparation of the labeled BuOOH, which was carried out
at Argonne National Laboratories under the auspices of the
Office of Basic Energy Sciences, Division of Chemical Sciences,
U.S. Department of Energy. We also thank Professor Doyle
Britton and Dr. Victor Young, Jr., for solving the crystal
structures.
Supporting Information Available: Tables listing experi-
mental details of X-ray structure determinations, atomic coor-
dinates, isotropic and anisotropic thermal parameters, and all
bond lengths and angles for 1, 4, 6, and 7 (76 pages). See any
current masthead page for ordering and Internet access
instructions.
Intermediates 1a-4a represent the first characterized non-
heme alkylperoxoiron(III) complexes and serve as models for
metastable biologically relevant peroxoiron(III) complexes such
as “activated bleomycin” which has a low-spin iron(III)
center13,58 and the purple lipoxygenase complex which has a
high-spin iron(III) center.6 The above series exemplifies a
simple ligand design strategy in which the introduction of a
JA9638521
(57) In support of this complexity, only three higher frequency vibrations
of the intermediate derived from the reaction of 8 and cumene hydroperoxide
are sensitive to 18O-labeling. See ref 12.
(58) Burger, R. M.; Kent, T. A.; Horwitz, S. B.; Mu¨nck, E.; Peisach, J.
J. Biol. Chem. 1983, 258, 1559-1564.
(59) 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.
(60) Dong, Y.; Yan, S.; Young, V. G., Jr.; Que, L., Jr. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 618-620.