Novel iron(III) complexes with imidazole containing tripodal ligands as model systems for catechol dioxygenases

Article abstract of DOI:10.1016/s0020-1693(99)00129-2

The iron (III) complexes [Fe(bpia)Cl₂][FeCl₄] (1), [Fe(bipa)Cl₂](ClO₄) (2), and [Fe₂O(bpia)₂Cl₂]Cl₂(4MeOH (3) (bpia: bis[(2-pyridyl)methyl][(1-methylimidazol-2-yl)methyl]amine, bipa: bis[(1-methylimidazol-2-yl)methyl][(2-pyridyl)methyl]amine) were synthesized. 1 and 2 are structural and functional models for catechol 1,2-dioxygenase. All compounds are characterized by spectroscopic methods and X-ray structure analysis. 3 was also investigated by EXAFS. The coordination environment around all Fe(III) cores is distorted octahedral by the tripodal ligand, chloride, and in the case of 3, additionally, oxygen. Chloride and oxygen are bonded labile in the functionally essential cis positions to the iron core. These are the binding sites of the catechol. The cation in 3 is formed as a dimerization product of 1 so that the oxo ligand acts as a linear bridge with a FeFe separation of 3.5756(8) ?. These properties of 3 have been used to study the amplitude enhancement in the EXAFS signal due to multiple scattering. In situ prepared iron(III) complexes with the ligands bpia and bipa show significant intradiol catechol dioxygenase activity with respect to the cleavage of 3,5-di-tert-butylcatechol and pyrocatechol.

Full text of DOI:10.1016/s0020-1693(99)00129-2

www.elsevier.nl/locate/ica
Inorganica Chimica Acta 291 (1999) 289–299
Novel iron(III) complexes with imidazole containing tripodal
ligands as model systems for catechol dioxygenasesꢀ
Matthias Pascaly a, Mark Duda a, Annette Rompel a, Bernd H. Sift a,
Wolfram Meyer-Klaucke b, Bernt Krebs a,*
a Anorganisch-Chemisches Institut der Westfa¨lischen Wilhelms-Uni6ersita¨t, Wilhelm-Klemm-Str. 8, D-48149 Mu¨nster, Germany
b EMBL Outstation Hamburg c/o DESY, Notkestr. 85, D-22603 Hamburg, Germany
Received 27 November 1998; accepted 1 March 1999
Abstract
The iron(III) complexes [Fe(bpia)Cl2][FeCl4] (1), [Fe(bipa)Cl2](ClO4) (2), and [Fe2O(bpia)2Cl2]Cl2(4MeOH (3) (bpia: bis[(2-
pyridyl)methyl][(1-methylimidazol-2-yl)methyl]amine, bipa: bis[(1-methylimidazol-2-yl)methyl][(2-pyridyl)methyl]amine) were syn-
thesized. 1 and 2 are structural and functional models for catechol 1,2-dioxygenase. All compounds are characterized by
spectroscopic methods and X-ray structure analysis. 3 was also investigated by EXAFS. The coordination environment around
all Fe(III) cores is distorted octahedral by the tripodal ligand, chloride, and in the case of 3, additionally, oxygen. Chloride and
oxygen are bonded labile in the functionally essential cis positions to the iron core. These are the binding sites of the catechol.
The cation in 3 is formed as a dimerization product of 1 so that the oxo ligand acts as a linear bridge with a Fe···Fe separation
,
of 3.5756(8) A. These properties of 3 have been used to study the amplitude enhancement in the EXAFS signal due to multiple
scattering. In situ prepared iron(III) complexes with the ligands bpia and bipa show significant intradiol catechol dioxygenase
activity with respect to the cleavage of 3,5-di-tert-butylcatechol and pyrocatechol. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Iron complexes; Tripodal ligand complexes; Crystal structures; Catechol dioxygenase model complexes; X-ray absorption spectroscopy
1. Introduction
extradiol enzymes according to their catalysis of aro-
matic ring cleavage between or outside the two ortho-
hydroxo groups. Intradiol dioxygenases contain
iron(III), whereas extradiol dioxygenases contain
iron(II) in their active sites.
Non-heme iron proteins with mono- and dinuclear
active sites have attracted considerable attention.
Mononuclear enzymes are involved in processes includ-
ing the biosynthesis of isopenicillin N (isopenicillin N
synthases) [1], oxidation of unsaturated fatty acids
(lipoxygenases) [2] and oxidative cleavage of aromatic
diols (catechol dioxygenases) [3]. The important class of
catechol dioxygenases takes part in nature’s mechanism
for metabolizing aromatic substrates to aliphatic prod-
ucts by insertion of both atoms of O2 into the substrate.
These metalloproteins are subdivided into intradiol and
Complexes with various ligands have been used for
modeling the properties of catechol 1,2-dioxygenases
[4–6]. Especially tetradentate tripodal ligands have
proven to be able to regulate the dioxygenase reactivity
of biomimetic iron model complexes [7–12]. Tripodal
ligands form distorted complexes with two unoccupied
cis coordination sites for binding catecholate. The hete-
rocyclic organic compounds bpia (bis[(2-pyridyl)
methyl][(1-methylimidazol-2-yl)methyl]amine) and bipa
(bis[(1-methylimidazol-2-yl)methyl][(2-pyridyl)methyl]
amine) belong to the class of tripodal ligands that
provide N4 donor sets. These are the first tripodal
ligands with methylimidazole moieties used for model-
ing catechol 1,2-dioxygenases. The methylimidazole
moieties are closely related to the amino acid histidine
which coordinates the metal cores in enzymes. There-
ꢀ Dedicated to Professor Heinz Dieter Lutz on the occasion of his
65th birthday.
* Corresponding author. Tel.: +49-251-83 331 31; fax: +49-251-
83 383 66.
E-mail address: krebs@nwz.uni-muenster.de (B. Krebs)
0020-1693/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0020-1693(99)00129-2

290
M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
fore, the ligands model the coordinative aspects in
metalloenzymes better than ligands synthesized before.
Usually 3,5-di-tert-butylcatechol (dbc) has been used
as the substrate in functional investigations. The ob-
served cleavage product is quite stable and has been
identified as 3,5-di-tert-butyl-1-oxacyclohepta-3,5-dien-
2,7-dione (anhydride of the cis,cis-muconic acid), the
immediate product of oxygen insertion into dbc (Fig. 1)
[8,9,13]. In previous studies Duda et al. have examined
the oxygenation reactivity of catechols depending on
their substituents [14].
In this paper we report the synthesis, structural, and
spectroscopic characterization of a series of novel
iron(III) compounds as functional and structural model
complexes for catechol 1,2-dioxygenases: [Fe(bpia)-
Cl2][FeCl4] (1) and [Fe(bipa)Cl2](ClO4) (2) providing
two cis coordination sites for substrate binding which
are occupied with labile chloro ligands. For comparison
we also report the structure of the m-oxo bridged dimer
[Fe2O(bpia)2Cl2]Cl2·4MeOH (3) in which one chloro
ligand is replaced by the labile m-oxo bridge. This
dinuclear species is postulated to be formed at the end
of the reaction of the mononuclear bpia iron complex
with catechol and O2 [14]. In analogy to previous
investigations it can be assumed that the complexes
transform into the reactive species by substitution of
the chloro ligands (for 1 and 2) and by the cleavage of
the oxo bridge (for 3) so that these compounds function
as precursor complexes for reactive species in solution
[15].
2. Experimental
2.1. Materials
All chemicals were of reagent grade and used as
received.
2.2. Physical measurements
IR spectra were obtained as KBr pellets on a Bruker
IFS 48 spectrometer in the range of 4000–400 cm−1
.
1H and 13C NMR spectra were measured on a Bruker
WH300 spectrometer. The electronic absorption spectra
were recorded on
a Kontron spectrophotometer
UVIKON 933 using quartz cuvettes (1 cm) and
methanol as solvent (c=2.5×10−4 mol l−1). Time
dependent UV–Vis spectroscopy was performed on a
Hewlett Packard Diode Array spectrometer. Elemental
analyses were carried out at the Organisch-Chemisches
Institut der Universita¨t Mu¨nster on a Heraeus CHN-O-
Rapid analyzer. Cyclic voltammograms were measured
on a BAS CV-50W electrochemical analyzer under Ar
atmosphere using a graphite working electrode with 0.1
M tetrabutylammonium tetrafluoroborate as support-
ing electrolyte and 10−4 mol l−1 of the analyst at 25°C.
Voltages are recorded relative to Ag/AgCl. The Mo¨ss-
bauer spectrum was measured in transmission at 293 K
at the Physikalisch-Chemisches Institut der Universita¨t
Mu¨nster using a 28 mCi source of 57Co in Co metal.
Velocity calibration was carried out by the resonance
lines of a-iron, and the isomer shift was given relative
to a-iron at room temperature. Magnetic susceptibility
of a powdered sample was recorded on a Faraday-type
magnetometer using a Cahn RG electrobalance in the
temperature range 4.5–320 K. The applied magnetic
field was about 1.5 T. Details of the apparatus have
been described elsewhere [18]. The experimental suscep-
tibility data were collected for the underlying diamag-
netism in the usual manner using Pascal’s constants
[19]. Correction for diamagnetism was estimated as
−546.6×10−6 cm3 mol−1. From the magnetic suscep-
tibility the effective magnetic moment veff was derived
following veff[vB]=ꢀ8T.
The oxo complex 3 presents a significant challenge
for multiple scattering (MS) EXAFS analysis (extended
X-ray absorption fine structure). MS effects are ex-
pected to become especially significant for a mono-
atomic bridge which approaches linearity [16,17]. Since
there are not many complexes with
a
linear
monoatomic oxo bridge we performed EXAFS analysis
on compound 3 to study the significant amplitude
enhancement in the EXAFS signal.
For the investigation of catechol-cleaving activity of
in situ prepared mononuclear complexes UV–Vis spec-
troscopic studies were carried out with dbc and pyro-
catechol (cat) as substrates to study the influence of the
ligand and the catechol on the reaction rate.
2.3. X-ray absorption spectroscopy
2.3.1. EXAFS measurements
The sample of 3 was enclosed in an appropriate cell
with kapton foil windows and measured at beamline E4
(DORIS III storage ring at HASYLAB) with an elec-
tron beam energy of 4.5 GeV and an average current of
80 mA. Fe K-edge XAS experiments resembled those
previously described for Fe [20] and Cu [21] model
complexes measured in absorption mode. For energy
calibration, the spectrum of a-iron was measured simul-
taneously, the first inflection point of which was as-
Fig. 1. Oxidative cleavage of catechols.

M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
291
signed to the value 7112.0 eV. Reproducibility of the Fe
K-edge energies in the above conditions has been found
previously to be 90.1 eV [20]. Spectra were collected
from 6800 to 8120 eV. Four scans were recorded and
the average analyzed.
2.3.2. Data analysis
To analyze Fe EXAFS data, we followed a procedure
similar to the one used for other Fe [20] and Cu [21]
model compounds. The k-space data were truncated
Fig. 2. The tripodal ligands bpia and bipa used in this work.
near the zero crossing (k=3.5–16 A−1) before the
,
Fourier transform was applied. Raw data were ana-
lyzed. The back-transformed, Fourier isolated data in
k-space were then subjected to curve fitting using the
full curved wave approach from the program EX-
CURV92 [22].
The wide-shell filtered data were fitted using con-
strained refinement of the imidazole groups including
their multiple scattering contributions as described in
Ref. [23]. The imidazole rings are treated as a geometri-
cally rigid unit.
hyde was heated at 160°C for 2 h. After cooling to
room temperature the mixture was solved in methanol.
The product precipitated as a light brown powder at
−20°C. This crude product was recrystallized from
CHCl3 to yield yellow crystals (16.2 g, 0.144 mol, 39%);
1
m.p. 114°C. H NMR (CDCl3) l(ppm): 3.6 (s, 3H), 4.4
(s, 2H), 4.7–5.3 (m, 1H), 5.8 (d, 1H), 6.5 (d, 1H).
2.4.2. Bis((2-pyridyl)methyl)amine
A total of 15.0 g (0.139 mol) of ((2-pyridyl)methyl)
amine and 15.0 g (0.140 mol) of pyridine-2-carboxalde-
hyde were stirred together in 60 ml methanol. After 2 h
1.9 g (0.054 mol) NaBH4 were added at 0°C. The
resulting red solution was stirred overnight and addi-
tionally refluxed for 1 h. The solution was allowed to
cool to room temperature diluted with 10 ml water and
acidified (pH 1) with hydrochloric acid to yield the
amine as a hydrochloride. To separate impurities and
side products this mixture was extracted with CHCl3
(5×20 ml). In order to neutralize the hydrochloride the
pH was adjusted to 8 with an aqueous solution of
NaOH. The free amine was extracted with CHCl3
(4×20 ml). The combined organic layers were dried
(MgSO4) and concentrated under reduced pressure.
High vacuum distillation yielded 17.6 g (0.089 mol,
A value of 0.7 was used for the amplitude reduction
factor and a value of −1.0 eV for the constant imagi-
nary potential which describes the lifetime of the pho-
toelectrons and the value E0 is refined for all shells
together.
The quality of the fit is represented by the R-factor
which is given as
kin
N
R= %
ꢁexp(ki)−th(ki)ꢁ×100%
i
i
N
i=1
% knj ꢁexp(kj)ꢁ
j
j=1
where exp(k) and th(k) are the experimental and
theoretical EXAFS [22].
The maximum number of parameters allowed to vary
simultaneously in the curve-fitting analysis was esti-
mated by Nfree=2DRFFDkfit/y where DRFF is the width
of the filter window and Dkfit is the length of the data
set in k space.
1
64%) of the product; b.p. 145–150°C (1 mmHg). H
NMR (CDCl3) l(ppm): 2.6 (s, 1H), 4.0 (s, 4H), 7.13 (t,
2H), 7.37 (d, 2H), 7.63 (t, 2H), 8.58 (d, 2H). 13C NMR
(CDCl3) l(ppm): 54.79 (CH2), 121,92 (Car), 122.27
(Car), 136.41 (Car), 149.31 (Car), 159.73 (Car).
CAUTION: Perchlorate salts of metal complexes
with organic ligands are potentially explosive.
2.4. Synthesis of the ligands bis[(2-pyridyl)methyl]-
[(1-methylimidazol-2-yl)methyl]amine (bpia) and bis
[(1-methylimidazol-2-yl)methyl][(2-pyridyl)methyl]amine
(bipa)
2.5. Synthesis of [Fe(bpia)Cl2][FeCl4] (1)
The ligands (Fig. 2) were prepared according to a
procedure described in the literature [24]. Modifications
were made in the syntheses of the preliminary products
(1-methylimidazol-2-yl)methanol and bis((2-pyridyl)me-
thyl)amine to improve the conversion to the product.
To a solution of 50 mg (0.18 mmol) bpia in 5 ml
methanol a solution of 29 mg (0.18 mmol) FeCl3 in 1
ml methanol was added and refluxed together for 5
min. After vapor diffusion of ether into the hot solution
red crystals suitable for X-ray diffraction analysis were
deposited after a few days. Yield: 45 mg (0.074 mmol,
41%); m.p. 153°C. Anal. Calc. for C17H19N5Cl6Fe2: C,
33.1; H, 3.1; N, 11.3. Found: C, 33.9; H, 3.9; N, 10.9%.
2.4.1. (1-Methylimidazol-2-yl)methanol
The followed procedure was first described by Reese
and Pei-Zhou [25]. A mixture of 30.0 g (0.365 mol)
1-methylimidazole and 30.0 g (1 mol) paraformalde-

292
M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
2.8. Crystallography
2.6. Synthesis of [Fe(bipa)Cl2](ClO4) (2)
The unit-cell data and diffraction intensities of com-
pounds 1 and 3 were collected on a STOE imaging-
plate diffraction system at 213 K using Mo Ka
A
total amount of 9 mg (0.017 mmol) of
Fe(ClO4)3·9H2O and 9 mg (0.035 mmol) of FeCl3 were
solved together in 5 ml of a methanol+ethanol mix-
ture (4:3). After addition of 15 mg (0.052 mmol) bipa in
15 ml of the same solvent the solution was heated until
reflux. Upon standing for 24 h red crystals suitable for
X-ray diffraction were obtained. Yield: 15 mg (0.028
mmol, 53%); decomposition over 210°C. Anal. Calc. for
C16H20N6Cl3FeO4: C, 36.8; H, 3.9; N, 16.1. Found: C,
36.7; H, 3.7; N, 15.8%.
,
radiation (graphite monochromated, u=0.71073 A).
The sample-to-plate distance was fixed at 70 mm for 1
(60 mm for 3) with a scan range from 05q5300° for
1 (05q5180° for 3) with an exposure time of 2 min
and an oscillating angle of Dq=3° for 1 and Dq=2°
for 3, respectively. Intensity data for complex 2 were
collected on an Enraf–Nonius CAD4 (Mo Ka, 0.71073
,
A, graphite monochromator), by using the ꢀ-scan tech-
nique with a variable scan rate of 2.93–29.30° min−1
.
2.7. Synthesis of [Fe2O(bpia)2Cl2]Cl2·4MeOH (3)
Further crystal data and experimental parameters are
listed in Table 1.
To a solution of 74 mg (0.3 mmol) bpia and 8.8 ml
(0.0064 mg, 0.063×10−6 mol) triethylamine in 5 ml
methanol a solution of 82 mg (0.3 mmol) FeCl3 · 9H2O
in 1 ml methanol was added. After vapor diffusion of
diethyl ether into this solution 43 mg (0.044 mmol,
29%) black crystals precipitated suitable for X-ray anal-
ysis; m.p. 144°C. Anal. Calc. for C38H50N10Cl4Fe2O5: C,
46.5; H, 5.1; N, 14.3. Found: C, 44.5; H, 5.5; N, 14.7%.
The structures of 1 and 3 were solved by direct
methods, structure 2 was solved by a Patterson synthe-
sis using the program system SHELXS 86. All other
non-hydrogen atoms were taken from a series of full-
matrix least-squares refinement cycles based on F2 with
the SHELXL 93 program followed by difference
Fourier syntheses [26].
Table 1
Crystallographic data and experimental details
[Fe(bpia)Cl2][FeCl4] (1)
[Fe(bipa)Cl2](ClO4) (2)
[Fe2O(bpia)2Cl2]Cl2·4MeOH (3)
Formula
C17H19N5Fe2Cl6
617.77
C16H20N5Cl3FeO4
522.58
C38H50N10Cl4Fe2O5
980.39
M (g mol−1
)
Crystal system
monoclinic
monoclinic
triclinic
(
Space group
P21 (No. 4)
9.538(2)
11.699(2)
11.225(2)
90
101.57(3)
90
1227.1(4)
P21/n (No. 14)
9.110(2)
15.450(3)
15.869(3)
90
106.18(3)
90
2145.1(4)
P1 (No. 2)
9.244(2)
,
a (A)
,
b (A)
11.002(2)
12.507(3)
68.74(3)
77.62(3)
83.40(3)
1156.9(4)
,
c (A)
h (°)
i (°)
k (°)
3
,
V (A )
Formula units/uc
2
4
1
zc (g cm−3
T (K)
)
1.672
213(2)
1.850
620
1.618
298(2)
1.113
1068
1.407
213(2)
0.771
376
Absorption coefficient (mm−1
F(000)
)
Data collection range (°)
Reflections measured
Independent reflections
8.752q552.2
14478
4724 (Rint=0.0571)
3.852q551.9
4476
4204 (Rint=0.0241)
10.852q552.0
8085
4191 (Rint=0.0403)
Reflections observed (I\2|(I))
4539
3577
3808
Variables
Goodness-of-fit (Fo
Final R value (I\2|(I))
271
1.084
271
1.039
271
1.009
2
)
R1=0.0313, wR2=0.0764a
R1=0.0342, wR2=0.0813a
0.385
R1=0.0423, wR2=0.1148b
R1=0.0529, wR2=0.1294b
0.805
R1=0.0365, wR2=0.0904c
R1=0.0429, wR2=0.1120c
0.754
Final R value (all data)
(D/z)max (e−
(D/z)min (e−
A
)
−3
,
−3
,
A
)
−0.413
−0.430
−0.381
2
a Calc. w=1/[|2(Fo2)+(0.0267P)2+1.2528P], where P=(Fo +2Fc2)/3.
2
b Calc. w=1/[|2(Fo2)+(0.0589P)2+3.0419P], where P=(Fo +2Fc2)/3.
2
c Calc. w=1/[|2(Fo2)+(0.0410P)2+0.6786P], where P=(Fo +2Fc2)/3.

M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
293
Table 2
All hydrogen atoms were placed on calculated posi-
,
Selected bond lengths (A) and angles (°) for 1
tions, except for the positions of the protons bonded to
the oxygen atoms O(2) and O(3) of methanol in 3
which were taken from the Fourier map, and allowed
to ride on their corresponding carbon atoms. The
isotropic thermal parameters for the methyl protons
were refined 1.5 times and for all other hydrogen atoms
1.2 times the Ueq value of the bonding atom. The
hydroxo protons of the solvate molecules in 3 were
located in the difference Fourier map and refined 1.5
times the Ueq value of the corresponding atom. The
methyl protons were placed on calculated positions
(UH=1.5Ueq of the bonded atom) in staggered confor-
mation with respect to the hydroxo groups. All other
non-hydrogen atoms of the complexes were refined
anisotropically.
,
Bond lengths (A)
Fe(1)–Cl(1)
Fe(1)–N(1)
Fe(1)–N(3)
2.2939(10)
2.125(3)
2.116(3)
Fe(1)–Cl(2)
Fe(1)–N(2)
Fe(1)–N(4)
2.2442(10)
2.271(3)
2.104(3)
Bond angles (°)
Cl(1)–Fe(1)–Cl(2)
Cl(1)–Fe(1)–N(2)
Cl(1)–Fe(1)–N(4) 170.20(8)
Cl(2)–Fe(1)–N(2) 167.77(8)
Cl(2)–Fe(1)–N(4)
N(1)–Fe(1)–N(3) 152.20(11)
N(2)–Fe(1)–N(3)
N(3)–Fe(1)–N(4)
99.59(4)
92.61(8)
Cl(1)–Fe(1)–N(1)
Cl(1)–Fe(1)–N(3)
Cl(2)–Fe(1)–N(1) 102.55(8)
Cl(2)–Fe(1)–N(3) 104.73(8)
N(1)–Fe(1)–N(2)
N(1)–Fe(1)–N(4)
N(2)–Fe(1)–N(4)
90.62(9)
90.09(8)
90.21(8)
76.24(11)
86.97(11)
77.59(10)
75.96(11)
87.65(11)
Table 3
,
Selected bond lengths (A) and angles (°) for 2
2.9. Catechol 1,2-dioxygenase acti6ity
,
Bond lengths (A)
Fe(1)–Cl(1)
Fe(1)–N(1)
Fe(1)–N(3)
2.2717(10)
2.136(3)
2.077(3)
Fe(1)–Cl(2)
Fe(1)–N(2)
Fe(1)–N(5)
2.2490(10)
2.304(3)
2.108(2)
The catechol cleaving activity of the iron(III) com-
plexes containing 3,5-di-tert-butylcatechol (dbc) and
pyrocatechol (cat) and the ligands bpia and bipa were
performed on in situ prepared complexes. To 2 cm3 of
Bond angles (°)
Cl(1)–Fe(1)–Cl(2) 100.64(4)
Cl(1)–Fe(1)–N(2) 92.61(7)
Cl(1)–Fe(1)–N(5) 168.05(7)
Cl(2)–Fe(1)–N(2) 166.61(7)
Cl(2)–Fe(1)–N(5)
N(1)–Fe(1)–N(3) 151.70(11)
N(2)–Fe(1)–N(3)
N(3)–Fe(1)–N(5)
Cl(1)–Fe(1)–N(1)
Cl(1)–Fe(1)–N(3)
Cl(2)–Fe(1)–N(1) 105.58(8)
Cl(2)–Fe(1)–N(3) 102.05(8)
N(1)–Fe(1)–N(2)
N(1)–Fe(1)–N(5)
N(2)–Fe(1)–N(5)
89.09(8)
91.79(8)
a
5×10−4 mol dm−3 solution consisting of
iron(III)perchlorate nonahydrate and the ligand were
added 0.02 cm3 of a 5×10−2 mol dm−3 (1 equiv.)
solution of the catechol and 0.04 cm3 of a 5×10−2 mol
dm−3 (2 equiv.) solution of piperidine. Air saturated
methanol was used as solvent. The disintegration of the
iron(III)–catecholate complexes was monitored by
UV–Vis spectroscopic measurements at ambient tem-
perature (25°C). Kinetic analyses of the intradiol cleav-
ing reactions were carried out by time dependent
measurements of the lower energy catecholate–iron(III)
charge-transfer band.
90.29(7)
76.30(10)
83.29(10)
76.68(9)
75.40(10)
90.60(10)
Table 4
,
Selected bond lengths (A) and angles (°) for 3
,
Bond lengths (A)
Fe(1)···Fe(1a)*
Fe(1)–O(1)
Fe(1)–N(1)
Fe(1)–N(3)
3.5756(8)
1.7878(6)
2.156(2)
2.154(2)
Fe(1)–Cl(1)
Fe(1)–N(2)
Fe(1)–N(4)
2.2991(9)
2.266(2)
2.157(2)
3. Results and discussion
Bond angles (°)
Fe(1)–O(1)–Fe(1a) 180.0
O(1)–Fe(1)–Cl(1)
O(1)–Fe(1)–N(2)
O(1)–Fe(1)–N(4)
Cl(1)–Fe(1)–N(2)
Cl(1)–Fe(1)–N(4)
N(1)–Fe(1)–N(3)
N(2)–Fe(1)–N(3)
N(3)–Fe(1)–N(4)
101.40(3)
93.71(5)
169.32(5)
164.85(5)
88.37(6)
152.59(7)
77.07(7)
84.28(7)
O(1)–Fe(1)–N(1)
O(1)–Fe(1)–N(3)
92.88(6)
89.15(5)
The ligands bpia and bipa (Fig. 2) have been pre-
pared following a modified literature method [24]. On
addition of a solution of iron(III) salts to a solution of
the ligand the compounds [Fe(bpia)Cl2][FeCl4] (1), [Fe-
(bipa)Cl2](ClO4) (2), and [Fe2O(bpia)2Cl2]Cl2·4MeOH
(3) have been obtained. The iron cores have a distorted
octahedral coordination sphere in common. The ligands
consist of pyridine and imidazole moieties so that N4
donor sets are provided. The cis coordination sites,
necessary for catalytic activity, are either occupied by
two labile chloro ligands or by a chloro and a bridging
oxo ligand. Crystallographic and experimental details
are given in Table 1. Relevant distances and angles of
the complexes are shown in Tables 2 to 4.
Cl(1)–Fe(1)–N(1) 102.45(6)
Cl(1)–Fe(1)–N(3) 103.92(6)
N(1)–Fe(1)–N(2)
N(1)–Fe(1)–N(4)
N(2)–Fe(1)–N(4)
75.55(7)
89.13(7)
76.64(7)
* Symmetry transformation: 1−x, 1−y, 1−z.
viously [11,12]. Table 5 gives an overview of related
complexes.
3.1. Crystal structures of [Fe(bpia)Cl2][FeCl4] (1) and
[Fe(bipa)Cl2](ClO4) (2)
These mononuclear and binuclear motifs with N4
donor sets of different ligands have been reported pre-
Compound 1 crystallizes with two complex cations[-
Fe(bpia)Cl2]+ and two complex counter ions [FeCl4]−

294
M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
Fig. 3. ORTEP plot of [Fe(bpia)Cl2]+; hydrogen atoms omitted for
clarity.
per unit cell. The structure of 2 is closely related to that
of 1. The ligand bpia is replaced by the ligand bipa and
a perchlorate ion acts as a counter ion. The unit cell of
2 contains four formula units. Figs. 3 and 4 show the
cations of 1 and 2. Both complexes exhibit a N4Cl2
coordination sphere. The N4 donor sets are provided by
the ligands bpia (1) and bipa (2). Due to the limiting
bite of both ligands the coordination polyhedron is
distorted octahedral. With respect to the nature of the
N donor moieties the largest Fe–N bond length in
these mononuclear complexes occurs between Fe(1) and
N(2). In general, compared to the tertiary amine nitro-
gen the nitrogen atoms of the aromatic heterocycles
bond more strongly to the iron center because the
tertiary amine has a lower ability to act as a p-donor.
In contrast to this the N donor atoms of the aromatic
heterocycles exhibit shorter bond lengths with an aver-
,
,
age value of 2.115 A for 1 and 2.107 A for 2, respec-
Fig. 4. ORTEP plot of [Fe(bipa)Cl2]+; hydrogen atoms omitted for
clarity.

M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
295
Fig. 5. ORTEP plot of [Fe2O(bpia)2Cl2]2+; hydrogen atoms omitted for clarity.
inversion. Consequently the m-oxo dimer has a linear
tively. In compound 2 the trans effect becomes appar-
ent in the extended bond length of one imidazole part
(N(5)), whereas the same bond length in compound 1
shows an even shorter value than the others. In this
case steric interactions between the large FeCl4− coun-
ter ions and the complex cation play a role. The chloro
ligands have in both complexes approximately the same
mean bond length. As in compound 2 the similar tpa
(tris(2-pyridylmethyl)amine) complex exhibits a trans
effect of the chloride ligands that results in a lengthened
Fe–Npyridine bond [11]. In this case also a small counter
ion (ClO4−) has been used. Due to the higher Lewis
acidic character of the tpa-iron core the chloro bond
lengths are shortened.
Fe–O–Fe structure (180°). The Fe(1)–O(1) bond
,
length is 1.788(1) A resulting in an Fe···Fe separation of
,
3.5756(8) A. The tetradentate ligand bpia consists of
one methylimidazole and two pyridine moieties con-
nected to a tertiary amine. The sixth coordination site is
occupied by a terminally bonded chloride atom. The
two pyridine nitrogen atoms (N1,N3) and the tertiary
amino group are cis oriented while the imidazole (N4)
part binds trans to the bridging oxo ligand.
The same reasons as described for 1 and 2 cause the
expanded bond length of the tertiary amine donor. This
is reflected in the bond lengths of the coordinated
The octahedron is distorted towards the tertiary
amino group. This is reflected in the low average value
of the angles between the tertiary amine donor and the
pyridine/imidazole moieties (76.6° for 1; 76.1° for 2).
Consequently the angles between the two chloro ligands
are expanded (99.6° for 1; 100.6° for 2).
3.2. Crystal structure of [Fe2O(bpia)2Cl2]Cl2·4MeOH
(3)
The unit cell of compound 3 consists of one complex
cation [Fe2O(bpia)2Cl2]2+, two chloride anions and
four methanol solvate molecules. The cation is shown
in Fig. 5. Both iron cores have a N4OCl donor set
forming a distorted octahedron. They are bridged by
the oxygen atom O(1) which is located on the center of
Fig. 6. Cell plot of compound 3.

296
M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
,
wave that can be assigned to one-electron processes
(Fe(III)/Fe(II)). The E1/2 values are 0.614 V for 1
and 0.330 V for 2. These potentials are quasi-re-
versible by definition [27]. The observed values reflect
the order expected from the UV–Vis spectroscopic
experiments. The spectrum of 3 shows a single irre-
versible reduction wave at −0.9 V and a coupled
oxidation–reduction wave at 0.293 and −0.086 V.
The resulting E1/2 value is 0.095 V. The redox process
is irreversible by definition. The irreversible wave at
−0.9 V can be either assigned to a one-electron re-
duction process resulting in a mixed valence oxidation
state or a cleavage of the m-oxo bridge due to a
one-electron reduction [28]. The first case corresponds
to the redox couple Fe2(III,III)/Fe2(III,II), the second
to Fe(III)/Fe(II).
heterocycle atoms (2.156(2) A for Fe(1)–N(1), 2.154(2)
,
,
A for Fe(1)–N(3) and 2.157(2) A for Fe(1)–N(4))
exhibiting almost the same bond lengths. One exception
,
is the bond length of the tertiary amine (2.266(2) A for
Fe(1)–N(2)). With respect to steric interactions between
the imidazole moiety of the ligand and the counter ion
no trans effect of the oxo bridge appears. The distortion
of the octahedron occurs towards the tertiary amino
group. The cis angles are decreased due to the limiting
bite of the ligand (average value of 76.4°).
The distortion of the coordination polyhedron is
caused partly by interactions between opposite ligands.
In the similar ntb (tris(2-benzimidazolylmethyl)amine)
complex steric effects result in a linear oxo bridge [12].
The cocrystallization with its mononuclear pendant
leads in the case of the tpa compound to a complex of
lower crystallographic symmetry. Thus, the angle
Fe(1)–O(1)–Fe(1a) has a value of 179.98(0)° [11].
There are hydrogen bonds between two methanol
solvate molecules and one chloride anion as shown in
the cell plot (Fig. 6).
Species related to the complexes show similar elec-
trochemical properties [12]. The dinuclear tpa complex
described by Que and coworkers is an exception as
compared to the analogous ntb and bpia complexes
due to the cocrystallization with its mononuclear pen-
dant [11].
3.3. IR and electronic spectra
3.5. Mo¨ssbauer spectroscopy and magnetic
susceptibility
For further characterization the IR and electronic
absorption spectra have been determined.
A pair of Lorentzians with identical amplitude and
linewidth was used to fit the 57Fe Mo¨ssbauer spectrum
of 1. The existence of octahedral high-spin iron(III)
ions is confirmed by one distinct symmetrical quadru-
pole doublet exhibiting a l1 value of 0.50 mm s−1 and
a DEQ1 value of 1.87 mm s−1. The value l1 is in the
normal range observed for high-spin iron(III) com-
plexes (0.3BlB0.6 mm s−1) [29]. The quadrupole
splitting reflecting the field gradient at the nucleus
suggests an equivalent N4OCl coordination sphere at
both metal cores. The value of the quadrupole split-
ting indicates considerable distortions from octahedral
symmetry.
The UV–Vis spectroscopic properties of the com-
plexes are listed in Table 5. For 1 and 2 absorption
features ascribed to the charge transfer transitions from
Cl− to Fe(III) are observed at 326 nm (m=3240 cm2
mol−1), and 372 nm (m=3025 cm2 mol−1), respectively.
The UV–Vis spectrum of complex 3 exhibits a band at
367 nm (m=3840 cm2 mol−1) and a shoulder at 312 nm
(m=4880 cm2 mol−1) which can be assigned to oxo
Fe(III) and Cl− “Fe(III) charge transfer transitions.
The bands below 250 nm are caused by p–p* transitions
of the ligand. The slight differences in wavelengths of
the similar complexes (Table 5) can be correlated with
the electron donor abilities of the tetradentate ligands.
Thus, conclusions about the Lewis acidity of the iron
cores can be drawn. A higher electrophilic character of
the metal ion shifts the charge transfer transitions to
higher wavelengths.
The susceptibility shows a minimum at 45 K. veff
has a value of 0.45 vB at 4.5 K which rises with
increasing temperature to 2.91 vB at 320 K indicating
antiferromagnetic coupling between the iron centers.
The exchange constant J was obtained by least-squares
fit of the theoretical expression with H= −2JS1S2 to
the experimental (T) data assuming a TIP of 400×
10−6 emu/mol per Fe(III) ion and a fixed g-factor of
The IR spectra of the complexes show the typical
aromatic and aliphatic bands. A strong band in the
spectrum of 3 can be assigned to w(O–H, CH3OH) at
3321 cm−1. The coordination of the imidazole and
pyridine moieties causes a shift of the aromatic CꢀN
bands to higher wavenumbers.
2.00, giving a value of −J=116 cm−1
.
For similar compounds with a linear bridging, oxo
ligand values of 109.7 cm−1 (tpa) and 100 cm−1 (ntb)
for −J have been observed (Table 5) [11,12], in accor-
dance with the overall range of −J=80–120 cm−1
for both linear and slightly bent m-oxo bridged diiron
complexes [30,31].
3.4. Electrochemistry
The cyclovoltammograms of the two mononuclear
complexes each exhibit a reduction and oxidation

M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
297
Fig. 7. k3-weighted fine structures of 3.
Fig. 8. Fourier transformations of 3.
the crystallographic analysis (Table 6), including a
3.6. Extended X-ray absorption fine structure (EXAFS)
spectroscopy
,
short Fe–O bond at 1.78 A. The angle Fe(1)–N(4)–
C(15) determined by crystallographic analysis is 137°.
The multiple scattering refinement performed in this
case gives an angle of 137° which agrees perfectly with
the crystallographically determined angle (Fig. 9).
Compound 3 is the one of the first dinuclear iron(III)
compounds containing a 180° oxo bridge that is investi-
gated by EXAFS spectroscopy [17]. Figs. 7 and 8 show
the comparison of the k3-weighted fine structures and
their corresponding Fourier transforms. The first peak
3.7. Catechol 1,2-dioxygenase acti6ity
,
in the Fourier transform at about 2 A is broad, suggest-
ing different metal ligand distances. The intense peak at
In the first step of the catechol dioxygenase reaction
catecholate binds to the iron core. This is displayed in
a color change of the solution from yellow to dark
purple. In the UV–Vis spectra two bands appear in the
range from 500 to 900 nm that can be assigned to
charge transfer transitions from catecholate to iron.
The transformation of the substrate by reaction with
oxygen to a derivative of cis,cis-muconic acid (Fig. 10)
was observed by the decrease of the lower energy
charge transfer band with time. A band at 400 nm
corresponding to the oxidation product 3,5-di-tert-
butyl-ortho-quinone or ortho-quinone, respectively,
does not appear in the spectra. Fig. 10 shows the
progress of the reaction of the bpia complex with dbc
and O2. The inset displays the logarithm of the absorp-
,
3.5 A suggests that a Fe–Fe interaction and multiple
scattering contributions arising from the 180° oxo
bridge are involved. Peaks I, II, and III were isolated
together to minimize any possible distortions from
Fourier filtering and to take into account multiple
scattering contributions. It is impossible to fit the three
peaks with reasonable values applying single scattering
calculations. In contrast to that a good fit was obtained
by applying multiple scattering calculations assuming
the presence of one Fe–O, one Fe–Cl, one Fe–Nim,
three Fe–N and one Fe–Fe interaction. Bond lengths
and structure of the complex indicated by EXAFS
analysis agree in a perfect manner with those shown by
Table 6
Curve-fitting results for Fe–K EXAFS of 3a
2
2
,
,
,
Window width (A)
Nfree
N and scatterer
R (A)
|
(A )
Eo (eV)
b (°)
R-factor (%)
1.3–4.7
30
1, O
1, Cl
3, N
1, N(imi)
1, C(imi)
1, C(imi)
1, C(imi)
1, N(imi)
1, Fe
1.789(3)
2.321(4)
2.111(3)
2.361(7)
3.016(7)
3.381(7)
4.232(7)
4.349(7)
3.648(3)
0.004(1)
0.001(1)
0.005(1)
0.001(3)
0.001(1)
0.001(1)
0.017(5)
0.017(5)
0.003(0)
17.3(3)
27.6
137.3(1)
a N is the number of scatterers per metal, R is the metal-scatterer distance, |2 is a mean square deviation in R. The fit was performed over the
−1
,
k=2–16 A
range.

298
M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
Table 7
Kinetic and spectroscopic properties of the in situ prepared com-
plexes
Reaction rate constant Lower energy c.t.
(M−1 s−1
)
band (nm)
[Fe(bpia)(dbc)]+
[Fe(bipa)(dbc)]+
[Fe(bpia)(cat)]+
[Fe(bipa)(cat)]+
4.3
865
842
729
718
0.98
0.089
0.015
with pyrocatechol. The reaction rate constants for the cat
complexes are significantly smaller (about 50 times)
which results from the lower electron donating character
of cat. Therefore, the energies of the LMCT bands
observed for the complexes with cat are higher (729 nm
for bpia; 718 nm for bipa) than those observed for dbc.
This is due to the electron donating character of the
tert-butyl groups in dbc.
In former investigations by Que et al. the degradation
of the isolated compound [Fe(tpa)(dbc)]BPh4 has been
determined [9]. This complex exhibits an even higher
reactivity (k=10 M−1 s−1) than the bpia compound as
expected from the spectroscopic data of the chloro
complexes (Table 5).
In summary, large reaction rate constants result from
a high electrophilic, i.e. high Lewis acidic, character of
the iron(III) complex caused by the ligand. The reaction
rate constants are additionally influenced by a high
nucleophilic character of the substrate caused by the
substituents of the catechol.
Fig. 9. Definition of the imidazole group.
tion versus time. The reactions show pseudo first order
kinetics due to the excess of dioxygen dissolved in meth-
anol (2.12×10−3 mol l−1 at 25°C; 25 times the amount
of complex). Kinetic analyses have been carried out by
calculation of the initial rates of the reaction rate cons-
tant. For the reaction of dbc the bpia complex exhibits
a more than four times greater value compared to the
bipa complex (Table 7). Using pyrocatechol (cat) as
substrate the bpia complex shows an even higher activity
(ca. six times) in contrast to the bipa complex [14].
From the position of the catecholate“iron(III) charge
transfer band conclusions on the influence of the ligand
properties can be drawn (Table 7). The UV–Vis absorp-
tion maxima shift to higher wavelengths, i.e. lower
energy, with higher Lewis acidity of the iron core. The
spectrum of the bipa-dbc complex exhibits a c.t. band at
842 nm, for the bpia–dbc complex the corresponding
band is found at 865 nm. Therefore, the coordination of
the ligand bpia leads to an iron(III) complex with a
higher Lewis acidity. The higher electron accepting
ability favors the formation of the second step in the
reaction mechanism – a semiquinone species [8,9,13,32].
Similar conclusions can be drawn from the experiments
4. Conclusions
A series of complexes as functional and structural
models for catechol 1,2-dioxygenases have been pre-
pared. In all three compounds the iron centers are
distorted octahedrally coordinated by the tetradentate
tripodal ligands bpia and bipa that provide N4 donor
sets and either two chloro ligands (1, 2) or a chloro and
a bridging oxo ligand resulting in a dinuclear species
(3). The structures of these complexes could be ob-
tained by X-ray crystallography.
We investigated the effects of these ligand sets with
imidazole moieties on the reactivity of catechol 1,2-
dioxygenase. For the in situ prepared complexes the
catechol cleaving activity of the bpia complexes is sig-
nificantly higher than for the bipa complexes. 3,5-Di-
tert-butylcatechol exhibits a higher reactivity than
pyrocatechol. The differences in reactivity are reflected
in the UV–Vis spectra of the catecholato complexes.
The position of the dbc/cat“Fe(III) c.t. band is influ-
enced by the ligands and by the substituents of the
catechols. Therefore, the bpia–dbc complex exhibits the
highest Lewis acidity of the iron core, the highest
Fig. 10. Progress of the reaction of [Fe(bpia)(dbc)]+ with O2 (mea-
surement interval: 15 s). Inset: Plot of ln(Abs) vs.
[Fe(bpia)(dbc)]+ reaction at 25°C.
t for the

M. Pascaly et al. / Inorganica Chimica Acta 291 (1999) 289–299
299
[3] (a) D.H. Ohlendorf, A.M. Orville, J.D. Lipscomb, J. Mol. Biol.
244 (1994) 586. (b) T. Senda, K. Sygiyama, H. Narita, T.
Yamamoto, K. Kimbara, M. Fukuda, M. Sato, K. Yano, Y.
Mitsui, J. Mol. Biol. 255 (1996) 735. (c) S.J. Lange, L. Que Jr.,
Curr. Opin. Chem. Biol. 2 (1998) 159.
electron donating character of the catechol, and conse-
quently the lowest energy of the LMCT band among
the examined complexes. The conclusions support
the substrate activation mechanism by Que et al.
[8,9,13,32].
[4] R.B. Lauffer, R.H. Heistand II, L. Que Jr., Inorg. Chem. 22
(1983) 50.
In spite of some minor structural differences these
three compounds fit perfectly into the context of similar
complexes (Table 5) [11,12]. Within the mononuclear
species a trend in Lewis acidity of the iron core becomes
apparent. This is reflected in the position of c.t. bands
and in the catechol 1,2-dioxygenase reactivity. Similar
relationships for the UV–Vis spectra of the three dinu-
clear species were found; the antiferromagnetic coupling
correlates with the separation of the iron cores.
The EXAFS work contributes to the general problem
of linear monoatomic linear bridges that may occur in
synthetic and biological assemblies. From our multiple
scattering calculations accurate metric determinations of
distances and angles were obtained.
[5] W.O. Koch, H.-J. Kru¨ger, Angew. Chem. 107 (1995) 2928;
Angew. Chem. Int. Ed. Engl. 34 (1995) 2671.
[6] M. Ito, L. Que Jr, Angew. Chem. 109 (1997) 1401; Angew.
Chem. Int. Ed. Engl. 36 (1997) 1342.
[7] Y. Nishida, H. Shimo, S. Kida, J. Chem. Soc., Chem. Commun.
(1984) 1611.
[8] D.D. Cox, L. Que Jr, J. Am. Chem. Soc. 110 (1988) 8085.
[9] H.G. Jang, D.D. Cox, L. Que Jr, J. Am. Chem. Soc. 113 (1991)
9200.
[10] R. Viswanathan, M. Palaniandavar, T. Balasubramanian, T.P.
Muthiah, Inorg. Chem. 37 (1998) 2943.
[11] T. Kojima, R.A. Leising, S. Yan, L. Que Jr, J. Am. Chem. Soc.
115 (1993) 11328.
[12] R.M. Buchanan, R.J. O’Brien, J.F. Richardson, J.-M. Latour,
Inorg. Chim. Acta 214 (1993) 33.
[13] L. Que Jr, R.C. Kolanczyk, L.S. White, J. Am. Chem. Soc. 109
(1987) 5373.
The catechol 1,2-dioxygenase reactivity of iron(III)
complexes of further ligand systems with new substrates
is under investigation.
[14] M. Duda, M. Pascaly, B. Krebs, J. Chem. Soc., Chem. Com-
mun. (1997) 835.
[15] R. Viswanathan, M. Palaniandavar, J. Chem. Soc., Dalton
Trans. (1995) 1259.
[16] B.-K. Teo, J. Am. Chem. Soc. 103 (1981) 3990.
[17] M.S. Co, W.A. Hendrickson, K.O. Hodgson, S. Doniach, J. Am.
Chem. Soc. 105 (1983) 1144.
5. Supplementary material
[18] (a) L. Merz, W. Haase, J. Chem. Soc., Dalton Trans. (1980) 875.
(b) S. Gehring, P. Fleischhauer, H. Paulus, W. Haase, Inorg.
Chem. 32 (1993) 54.
[19] O. Kahn, Molecular Magnetism, VCH, NY, 1993.
[20] S. Priggemeyer, P. Eggers-Borkenstein, F. Ahlers, G. Henkel, M.
Ko¨rner, H. Witzel, H.-F. Nolting, C. Hermes, B. Krebs, Inorg.
Chem. 34 (1995) 1445.
Further details of the crystal structure investigation,
atomic coordinates, thermal parameters, complete bond
distances and angles can be obtained from the Cam-
bridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK, on quoting the full journal
citation.
[21] F. Zippel, F. Ahlers, R. Werner, W. Haase, H.-F. Nolting, B.
Krebs, Inorg. Chem. 35 (1996) 3409.
[22] N. Binsted, J.W. Campbell, S.J. Gurman, P.C. Stephenson,
EXCURV92 program, SERC Daresbury Laboratory, UK, 1992.
[23] N. Binsted, R.W. Strange, S.S. Hasnain, Biochemistry 31 (1992)
12117.
[24] N. Wei, N.N. Murthy, Z. Tyekla´r, K.D. Karlin, Inorg. Chem. 33
(1994) 1177.
[25] C.B. Reese, Z. Pei-Zhou, J. Chem. Soc., Perkin Trans. (1993)
2291.
[26] (a) G.M. Sheldrick, SHELXTL PLUS, Siemens Analytical X-
Ray Instruments, Madison, WI, 1990. (b) G.M. Sheldrick,
SHELXL 93, Program for Crystal Structure Determination,
University of Go¨ttingen, 1993.
Acknowledgements
Financial support from the Henkel KGaA, the
Deutsche Forschungsgemeinschaft, the Volkswagen
Foundation, the Bundesministerium fu¨r Bildung und
Forschung and the Fonds der Chemischen Industrie is
gratefully acknowledged. We thank Professor W. Haase
and K. Falk (Technische Universita¨t Darmstadt) for
magnetic susceptibility measurements. M.P. thanks the
Ministerium fu¨r Schule und Weiterbildung, Wis-
senschaft und Forschung des Landes Nordrhein-West-
falen for a Graduiertenstipendium.
[27] A.J. Bard, L.R. Faulkner, Electrochemical Methods, Wiley, New
York, 1980.
[28] J.R. Hartman, R.L. Rardin, P. Chaudhuri, K. Pohl, K.
Wieghardt, B. Nuber, J. Weiss, G.C. Papaefthymiou, R.B.
Frankel, S.J. Lippard, J. Am. Chem. Soc. 109 (1987) 7387.
[29] P. Gu¨tlich, R. Link, A.X. Trautwein, Mo¨ssbauer Spectroscopy
and Transition Metal Chemistry, Springer, NY, 1978.
[30] D.M. Kurtz Jr., Chem. Rev. 90 (1990) 585.
[31] R. Hotzelmann, K. Wieghardt, U. Flo¨rke, H.-J. Haupt, D.C.
Weatherburn, J. Bonvoisin, G. Blondin, J.-J. Girerd, J. Am.
Chem. Soc. 114 (1992) 1681.
References
[1] P.L. Roach, I.J. Clifton, C.M.H. Hensgens, N. Shibata, C.J.
Schofield, J. Hajdu, J.E. Baldwin, Nature 387 (1997) 827.
[2] W. Minor, J. Steczko, B. Stec, Z. Otwinowski, J.T. Bolin, R.
Walter, B. Axelrod, Biochemistry 35 (1996) 10687.
[32] R.Y.N. Ho, L. Que Jr., Chem. Rev. 96 (1996) 2607.
.