The coordination chemistry of FeCl3 and FeCl2 to Bis[2-(2,3-dihydroxyphenyl)-6-pyridylmethyl](2-pyridylmethyl)amine: Access to a diiron(III) compound with an unusual pentagonal-bipyramidal/square-pyramidal environment

Article abstract of DOI:10.1002/chem.200600276

Coordination of FeCl₃ to the title ligand yields a mononuclear iron-(III) complex 1, which was characterized by spectroscopic techniques and X-ray diffraction. The ligand is (Κ³-N) tridentate and the metal, which lies in a pseudo-oct

Full text of DOI:10.1002/chem.200600276

DOI: 10.1002/chem.200600276
The Coordination Chemistry of FeCl3 and FeCl2 to Bis[2-(2,3-
dihydroxyphenyl)-6-pyridylmethyl](2-pyridylmethyl)amine: Access to a
Diiron(iii) Compound with an Unusual Pentagonal-Bipyramidal/Square-
N
Pyramidal Environment
Ahmed Machkour,[a, b] Nasser K. Thallaj,[a] Laila Benhamou,[a, b]
Mohammed Lachkar,[b] and Dominique Mandon*[a, b]
Abstract: Coordination of FeCl3 to the
title ligand yields a mononuclear iron-
10 mV (vs. SCE). Reduction of com-
pound 1 with Zn/Hg yields 2’, which
displays identical properties to 2.
Taken together, these findings indicate
that in spite of the different oxidation
state of the metal in 2, no major geo-
metrical/structural change is observed
at the metal center with respect to 1.
The reaction of 2 with dioxygen in the
absence of organic substrates proceeds
extremely rapidly and yields compound
3, which is diiron(iii) derivative
whose X-ray crystal structure is also re-
ported. The possibility of a radical-
based mechanism is discussed. Com-
pound 3 displays an unusual geometry:
a
ACHTREUNG
A
ized by spectroscopic techniques and
X-ray diffraction. The ligand is (k3-N)
tridentate and the metal, which lies in
a pseudo-octahedral environment, is
bound to a phenolate group from the
catechol substituent. The dichloroiro-
n(ii) complex 2 was easily obtained by
metalation of the ligand with FeCl2 and
characterized by various spectroscopic
techniques. In their cyclic voltammo-
grams both 1 and 2 display the same
one iron(iii) center is seven-coordinate,
G
whereas the other lies in a square-pyra-
midal environment. The two iron
atoms are bridged by the catecholato
substituents. To the best of our knowl-
edge, 3 is the first example of a seven-
Keywords: dioxygen activation
·
coordinate iron
A
iron complexes · N ligands · oxida-
tion · radical reactions
tris(2-pyridylmethyl)amine ligands.
reversible FeII/FeIII wave at E1/2
=
Introduction
lished, was formerly introduced from studies involving alkyl
or simple aryl substituents.[2–4] Later on, more sophisticated
ligands with non-innocent groups such as crown ethers or
amido-, phenoxy-, or quinone-appended ligands, to mention
only a few examples, were reported with more or less ex-
pected properties.[5–10] As a particularly salient illustration of
this approach, ligands with 2,5-dimethoxyphenyl, 2,5-hydro-
quinone, or 2,5-quinone substituents have recently been pre-
pared and their copper complexes studied in detail to gain
insight into the mechanism by which protons can be dis-
placed at the same time that the metal center undergoes
redox modification.[11]
Tris(2-pyridylmethyl)amine-based ligands (TPAs) form one
of the most widely investigated family of compounds within
the class of tripodal tetraamine ligands.[1] The concept of
tuning the reactivity within a series of metal complexes by
substitution of the tripod skeleton, which is now well estab-
[a]Dr. A. Machkour, N. K. Thallaj, L. Benhamou, Dr. D. Mandon
Laboratoire de Chimie Biomimetique des Metaux de Transition,
Institut de Chimie, UMR CNRS no. 7177 - LC3
UniversitØ Louis Pasteur, Institut Le Bel
4 rue Blaise Pascal, 67070 Strasbourg cedex (France)
Fax : (+33)390-245-001
TPA ligands are easily accessible, and their iron com-
plexes have been extensively used as tools to understand the
chemistry at the active site of some iron proteins, thus
making this series of compounds one of the most widely in-
vestigated classes of complexes for inorganic chemists in-
E-mail: mandon@chimie.u-strasbg.fr
[b]Dr. A. Machkour, L. Benhamou, Prof. Dr. M. Lachkar,
Dr. D. Mandon
Laboratoire dꢀIngenierie des Materiaux Organometalliques et
Moleculaires (L.I.M.O.M.), Universite Sidi Mohamed Ben Abdellah,
Faculte des Sciences, B.P. 1796 (Atlas), 30000, Fez (Morocco)
G
volved in the field of biomimetics.[1,12–14] Iron
(iii) derivatives
A
are particularly stable, easy to handle, and, among other
properties, their ability to coordinate catechol derivatives
Supporting Information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
6660
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 6660 – 6668

FULL PAPER
has resulted in many studies aimed at understanding the
mechanisms at the active site of intradiol dioxygenases,
which are an important class of iron proteins.[15–17]
Catechol fragments form stable complexes with iron and
are also involved in iron uptake and transport by sidero-
phores.[18] Catechols, and phenolates in general, are consid-
ered as “hard ligands” and therefore would stabilize iron in
the iron(iii) state, whereas nitrogen-containing ligands are
G
“soft” and would stabilize the iron(ii) state. Following these
principles, an elegant example of redox-driven translocation
of iron in a bicompartmental ligand has already been descri-
bed.[19]
We have been involved in the chemistry of stable iron(ii)
derivatives with various substituted aminomethylpyridyl-
containing ligands, and have reported that some of them ex-
hibit an interesting reactivity towards dioxygen.[20–22] As an
extension of our work with substituted TPA derivatives, we
decided to prepare a ligand with two very different coordi-
nation sites, namely bis[2-(2,3-dihydroxyphenyl)-6-pyridyl-
methyl](2-pyridylmethyl)amine (BCATTPA), whose struc-
ture is shown in Scheme 1, and embarked upon an investiga-
tion of the coordination chemistry of this ligand with iron-
ACHTREUNG
A
ligand occurs through the nitrogen-containing site, as con-
firmed by X-ray diffraction analysis of the complex. The di-
chloroiron(ii) complex displays a similar geometry, and can
be synthesized either by metalation of the ligand with
iron(ii) chloride or chemical reduction of the iron(iii) deriva-
T
tive. This complex is extremely oxygen sensitive and in the
absence of substrate yields a dinuclear compound with a
very unusual structure: two metals bridged by phenoxo
groups lie in seven- and five-coordinate environments, re-
spectively. To the best of our knowledge, this is the first ex-
ample of a seven-coordinate iron(iii) derivative in the field
N
Scheme 1. Mononuclear iron(ii) and iron(iii) complexes of BCATTPA de-
N
of tris(2-pyridylmethyl)amine ligands. We report here the
coordination chemistry of this new ligand and the prepara-
tion and structural characterization of an original diiron
complex.
scribed in this work.
Metalation of BCATTPA with FeCl3: Metalation of the cat-
echol ligand in the absence of a base was expected to occur
due to the vicinity of the nitrogen of the pyridyl group with
the proton of the hydroxy function. This reaction actually
worked very well, and treatment of anhydrous FeCl3 with
one equivalent of the ligand in THF or CH3CN yielded a
dark violet solution. Once the reaction was complete,
workup was carried out without the need for any particular
precautions and the compound was obtained as a thermally
and oxygen stable solid.
Results and Discussion
Preparation of the ligands: This is straightforward and fol-
lows a procedure involving a Suzuki-type cross-coupling re-
action between a boronic acid and aryl halide.[4] Although
condensation from simple pyridyl-based precursors of the li-
gands may be carried out,[5,23] we started directly with the
preformed Br2TPA tripod, for convenience, and treated it
with 2,3-dimethoxyphenylboronic acid. The tetramethoxy
derivative BDMPTPA obtained was then treated with BBr3
to yield the catechol-containing ligand BCATTPA. Note
that the BBr3/ligand ratio should not be too high (typically
6–7/1) in order to minimize side reactions between the
Lewis acid and the nitrogen-containing base. We obtained
the ligand as a crystalline material with typical yields of 55
to 60%.
The UV/Vis spectrum of this complex is shown in
Figure 1 and reveals a broad band at l4 =617 nm (e4 =0.82”
103 mmolcm2), which we assigned to a LMCT phenolate!
FeIII transition. This transition is responsible for the dark
color of the complex. In general, complexes with iron(iii)
N
chloride coordinated to pyridyl-containing tripods exhibit a
moderately intense LMCT Cl!FeIII absorption in the
region 360–390 nm.[24,25] In the present case, a shoulder is
observed at l3 =354 nm. Although it lacks a well-defined
Chem. Eur. J. 2006, 12, 6660 – 6668
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6661

D. Mandon et al.
Figure 1. UV/Vis spectrum of [(BCATTPA)FeIIICl2] (1) in CH3CN at
room temperature. Inset: cyclic voltammogram of the same compound in
CH3CN (0.1m (TBA)PF6, 200 mVsꢀ1, scan to negative values).
Figure 2. Structure of [(BCATTPA)FeIIICl2] (1; ORTEP diagram showing
a partial numbering scheme). Selected bond lengths [Š]and angles [ 8]:
pattern, this signal might be this expected LMCT band, thus
suggesting that the tripod is indeed involved in coordination
to the iron, as outlined in Scheme 1. Finally, two other
ligand-based absorptions are observed at l2 =297 and
263 nm (e2 =13.01 and 13.67”103 mmolcm2, respectively).
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Fe O1 1.870(3), Fe N1 2.164(3), Fe N2 2.182(3), Fe N3 2.259(3), Fe
ꢀ
ꢀ
Cl2 2.3093(13), Fe Cl1 2.3581(12), Ho3 N4: 1.835(4); O1-Fe-N1
87.17(12), O1-Fe-N2 92.50(13), N1-Fe-N2 91.25(13), O1-Fe-N3
157.87(13), N1-Fe-N3 74.23(12), N2-Fe-N3 76.36(13), O1-Fe-Cl2 99.41(9),
N1-Fe-Cl2 172.47(9), N2-Fe-Cl2 84.85(10), N3-Fe-Cl2 98.56(9), O1-Fe-
Cl1 99.61(10), N1-Fe-Cl1 88.50(9), N2-Fe-Cl1 167.86(10), N3-Fe-Cl1
91.92(9). The ellipsoids enclose 50% of the electronic density.
1
We have previously reported H NMR spectroscopic data
for trichloroiron
gands.[24] The resonances in the paramagnetic region are
generally extremely broad for high-spin iron(iii) derivatives.
(iii) complexes with substituted TPA li-
G
To understand any further reactivity studies, the question
of whether the structure is retained in solution must be con-
sidered. Within the dichloroiron(ii) series of complexes with
this kind of ligand, this point can be addressed by combining
NMR, UV/Vis and molecular conductivity data.[20,21,26,27] In
In the present case, only two extremely weak and poorly de-
fined signals were detected at around d=95 and 60 ppm.
Broadened signals, which might be due to the uncoordinated
part of the ligand, were also observed in the diamagnetic
region. The spectrum is given in the Supporting Informa-
tion.
1
the present case, however, the H NMR spectrum is uninfor-
mative, and one has to rely on UV/Vis and electrochemical
techniques. The molecular conductivity (L=39 Scm2 molꢀ1)
was measured in CH3CN, and its value indicates that the
complex remains undissociated in solution. This observation
strongly suggests that the geometry of 1 is retained in solu-
tion.
Single crystals of [(BCATTPA)FeIIICl2]·Et2O (1·Et2O)
could be obtained by layering an acetonitrile solution of 1
with dry diethyl ether. An ORTEP diagram of 1 is displayed
in Figure 2, in which it can clearly be seen that the metal is
coordinated to the nitrogen atoms of the ligand. The nitro-
gen-containing tripod coordinates in a tridentate fashion
(k3-N), with one substituted pyridyl unit remaining remote
from the coordination center. The coordination environment
is completed by two chloride ligands from the starting metal
salt and one phenolato group from a singly deprotonated
substituent of one coordinated pyridine. The metal lies in a
Cyclic voltammetry experiments were carried out within
the range expected for reduction of iron
A
shown in Figure 1, a reversible wave is observed at E1/2
=
10 mV (vs. SCE). This suggests that no major structural
change is observed upon reduction of compound 1.
ꢀ
distorted octahedral environment. All the N Fe distances
are longer than 2.1 Š, in line with a high-spin state for the
Metalation of BCATTPA with FeCl2: This reaction was car-
ried out following previously reported methods, under strict-
ly anaerobic conditions, in THF or CH3CN. Iron(ii) chloride
reacts immediately with the ligand to produce an intense
orange color. The compound formed (2) is one of the most
oxygen-sensitive dichloroiron(ii) complexes that we have
ever come across within the series of complexes of substitut-
ed TPA ligands—it becomes dark violet within a fraction of
a second upon exposure to air, even in the solid state. Un-
fortunately, we have not yet been able to isolate crystals
suitable for an X-ray diffraction analysis, therefore the struc-
ture of 2 shown in Scheme 1 is based on spectroscopic data
only.
ꢀ
metal. The O1 Fe distance (1.870 Š) is short and is consis-
tent with an anionic bond between a deprotonated phenol
and the iron(iii) center. The chloride to iron distances lie in
A
ꢀ
ꢀ
the expected range (Cl1 Fe 3.358 and Cl2 Fe 2.309 Š). It is
noteworthy that the HO3 proton interacts with the N4 nitro-
gen of the vicinal pyridine, the corresponding distance being
only 1.835 Š. As a consequence, the aromatic rings in the
uncoordinated pyridine are almost co-planar.
The nOH vibration is observed at 3412 cmꢀ1 in the IR spec-
trum, a value perfectly in line with the presence of a bound
hydroxy group.
6662
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 6660 – 6668

A Diiron(iii) Compound with an Unusual Coordination Environment
A
FULL PAPER
The UV/Vis spectrum (Figure 3) contains, together with
the ligand-centered transitions observed at l1 =260 and
290 nm (e2 =11.72”103 mmolcm2 and 13.05”103 mmolcm2,
respectively), two absorptions at l=355 and 446 nm (e=
2.98”103 mmolcm2 and 0.72”103 mmolcm2, respectively).
along with at least one phenolate, ii) a very likely octahedral
geometry around the metal, iii) an ionic character of the
compound, and iv) an identical FeII/FeIII redox wave to that
obtained with compound 1. When viewed together, these
observations do not leave many possibilities for a structure
different to that drawn in Scheme 1, that is, two coordinated
chloride anions and one phenolate from a catechol-substitut-
ed pyridine for a ligand coordinated in a “hypodentate”
mode. Elemental analysis was expected to confirm this
structure by direct fit of the percentages with those calculat-
ed. However, the extreme sensitivity of 2 to oxygen imped-
ed any satisfactory result. However, we were at least able to
see that the Cl/Fe molar ratio (2.0) indicates that two chlor-
ide ligands per iron atom are indeed present in the complex.
Reduction of [(BCATTPA)FeCl3] with Zn/Hg: indirect
characterization of 2: Because of the identical nature of the
electrochemical data for 1 and 2, we decided to reduce 1
with zinc amalgam, as shown in Scheme 1. The reaction was
carried out under rigorously anaerobic conditions in CH3CN
or THF, and the medium, which was dark violet at the be-
ginning, turned bright orange within a few tens of minutes.
Filtration and precipitation afforded an orange solid whose
1H NMR and UV/Vis spectroscopic data were identical to
those obtained for an authentic sample of 2. Obviously,
going from the neutral molecule 3 to the charged compound
2’ requires the presence of a countercation. In this case it is
likely to be zinc. The metal center is, however, unaffected
by this cation. We therefore strongly believe that the struc-
ture of 2 drawn in Scheme 1 is the correct one.
Figure 3. UV/Vis spectrum of [(BCATTPA)FeIICl2] (2) in CH3CN at
room temperature. Inset: cyclic voltammogram of the same compound in
CH3CN (0.1m (TBA)PF6, 200 mVsꢀ1, scan to positive values).
Both the wavelength and intensity features of the absorption
at 446 nm strongly suggest that it corresponds to a MLCT
FeII!N(py) transition for which the ligand coordinates in a
so-called “hypodentate” mode, that is, in a tridentate (k3-N)
fashion.[20,28] The absorption at 355 nm is absent from the
spectrum of any other simple dichloroiron(ii) complex re-
ported in the literature, such as [(TPA)FeCl2]for instance,
and is therefore assigned to a LMCT phenolate!FeII transi-
tion, thereby indicating that, as is the case for 1, the catechol
coordinates to the metal.
Oxygenation of [(BCATTPA)FeCl2]: As already mentioned
above, 2 is extremely sensitive to dioxygen, even in the solid
state, where the orange solid turns dark within a second. In
solution this oxygenation reaction can be monitored by UV/
Vis spectroscopy upon lowering the temperature to 263 K,
as shown in Figure 4. The reaction is complete within one
hour at this temperature, and no intermediate species could
be detected.
Unlike the iron(iii) derivative, paramagnetically shifted
A
signals appear in the 1H NMR spectrum of 2. Relatively
sharp signals with full widths at half-height (FWHHs) of be-
tween 120 and 150 Hz are observed at d=47, 46, and
18 ppm, reminiscent of coordinated b-b’ pyridyl protons in
an octahedral geometry.[20] Other broad resonances with
FWHHs of more than 300 Hz are observed at d=127, 50,
35, 20, and ꢀ7 ppm. The presence of moderately broadened
signals in the diamagnetic region supports the “hypoden-
tate” coordination of the ligand; these resonances are as-
signed to protons belonging to the uncoordinated arm of the
1
ligand. The H NMR spectrum is given in the Supporting In-
formation.
The molecular conductivity (L=107 Scm2 molꢀ1) was
measured in CH3CN and its value indicates an ionic behav-
ior in solution, in line with the presence of a singly charged
species.[29] The cyclic voltammetric behavior was also investi-
gated and, interestingly, the same wave as already seen with
the iron(iii) compound 1 was observed (E1/2 =10 mV vs.
G
SCE; Figure 3).
Figure 4. UV/Vis spectral changes observed upon oxygenation of
[(BCATTPA)FeIICl2] (2) at 263 K in acetonitrile solution (one trace
every five minutes for 75 minutes).
To summarize, we have observed: i) coordination of the
ligand in a tridentate (k3-N) fashion by nitrogen atoms
Chem. Eur. J. 2006, 12, 6660 – 6668
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6663

D. Mandon et al.
The most striking feature is the progressive appearance of
a new broad absorption centered at l3 =590 nm (e3 =1.10”
103 mmolcm2). This absorption is slightly blue shifted and
more intense than that observed for compound 1, but can
also be assigned to a LMCT phenolate!FeIII transition. The
ligand-centered bands are also slightly red shifted to l1 =
266 nm and l2 =300 nm (e1 =12.67”103 mmolcm2 and e2 =
14.64”103 mmolcm2 respectively).
,
On a preparative scale, a black solid was obtained upon
addition of diethyl ether to the oxygenated medium. The
1H NMR spectrum of this compound displays no obvious
signals in the paramagnetic region. Unlike 1, nothing sug-
gesting that any uncoordinated fragment is present could be
detected—only two broad and unresolved resonances were
observed at d=3 and 1 ppm. The 1H NMR spectrum is
given in the Supporting Information section.
Figure 5. Structure of [(BCATTPA)FeIII2Cl2] (3; ORTEP diagram show-
An X-ray diffraction analysis study showed that the reac-
ꢀ
ing a partial numbering scheme). Selected bond lengths [Š]: Fe2 N1
tion product (3) is
a dinuclear species, as shown in
ꢀ
ꢀ
ꢀ
ꢀ
2.503(4), Fe2 N2 2.246(4), Fe2 N4 2.477(4), Fe2 O3 2.024(3), Fe2 O2
Scheme 2.
ꢀ
ꢀ
ꢀ
ꢀ
2.066(3), Fe2 Cl2 2.2781(13), Fe2 N3 2.103(4), Fe1 O2 2.057(3), Fe1
O3 2.029(3), Fe1 O4 1.906(3), Fe1 O1 1.887(3), Fe1 Cl1 2.2407(15),
Slow evaporation of the solvent from an acetonitrile solu-
tion of 3 afforded single crystals. An ORTEP diagram of the
ꢀ
ꢀ
ꢀ
ꢀ
Fe1 Fe2 3.281(4). The ellipsoids enclose 50% of the electronic density.
The present work reports
some preliminary coordination
chemistry of FeCl3 and FeCl2
with a ligand that contains two
different coordination sites.
We expected that the presence
of two catechol moieties and
nitrogen-containing
bases
Scheme 2. Oxygenation of [(BCATTPA)FeCl2] (2).
within the same ligand would
allow coordination of the
metal to the “hard” catechol
complex is shown in Figure 5 and the main geometrical fea-
tures are detailed in Figure 6.
site and leave the tripod free (or at least protonated, with
the nitrogen atoms acting as bases). However, a “normal”
coordination of the TPA tripod was observed, although with
one chloride ion being replaced by a phenolato group from
the substituent. The structure of 1 in the solid state indicates
that the ligand coordinates in a tridentate (k3-N) fashion,
with a dangling substituted pyridine. This coordination
mode is now well established in the chemistry of substituted
TPA ligands and obviously reflects a gain in stability with
respect to a more sterically hindered environment.[12,24,30]
The molecular structure of compound 3 reveals the pres-
ence of two iron atoms for each ligand. Fe2 lies in an unusu-
al seven-coordinate environment, with the tripod coordinat-
ing in a (k4-N) mode; one chloride ligand coordinates trans
to a pyridine, and two bridging phenolates from the catechol
ꢀ
substituents complete the coordination sphere. All Fe2 N
distances are above 2.2 Š, with a maximum for the two
ꢀ
ꢀ
equatorial pyridines (Fe2 N1 2.503, Fe2 N4 2.477 Š). The
1
ꢀ
ꢀ
Fe–catecholato distances (Fe2 O2 2.066, Fe2 O3 2.024 Š)
are consistent with the values expected for a bridging phe-
nolate. The angles between two adjacent atoms in the equa-
torial plane do not differ much from the ideal value of 728,
with a minimum for N2-Fe2-N4 (68.068) and a maximum for
O2-Fe2-N1 (74.118). The axial angle is defined by N3-Fe2-
Cl2 (168.828).
The H NMR spectra of the FeCl3 complexes generally dis-
play broad and unresolved signals in the paramagnetic
region.[24,25] Complex 1, with an almost silent trace, is no ex-
ception to this rule. Uncoordinated substituents of the
tripod can, however, be observed as broadened resonances
in the diamagnetic region, thus indicating that the geometry
is retained in solution. The very low value for the molecular
conductivity also supports this. Electrochemical studies car-
ried out on FeCl3 and FeCl2 complexes of simple TPA deriv-
atives generally display an FeIII/FeII wave at markedly posi-
tive values.[25,31] The presence of an anionic ligand can signif-
icantly lower the corresponding potential, with consequen-
ces for the reactivity towards dioxygen, for instance. In the
On the other hand, Fe1 lies in a severely distorted square-
pyramidal environment with two long catechol–Fe bonds
ꢀ
ꢀ
ꢀ
(Fe1 O2 2.057, Fe1 O3 2.029 Š), two short ones (Fe1 O1
ꢀ
ꢀ
1.887, Fe1 O4 1.906 Š), and an Fe1 Cl1 distance of
2.2407 Š. The distance between the two metal centers is
3.281 Š.
6664
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 6660 – 6668

A Diiron(iii) Compound with an Unusual Coordination Environment
A
FULL PAPER
cyclic voltammogram in the potential range corresponding
to the FeII/FeIII couple led us to suspect that their coordina-
tion environments might be similar, at least at the metal
center. Confirmation of this hypothesis came from reduction
of 1 with Zn/Hg, as complex 2’ was obtained.
With respect to activation of dioxygen, those mononuclear
iron(ii) TPA derivatives that have been found to react rapid-
ly with dioxygen are either complexes in which the metal is
coordinated to a noninnocent ligand or derivatives with a
coordinatively
unsaturated
environment
at
the
metal.[2,20,21,30] An iron(ii)–TPA complex in which a thiolate
is coordinated to the metal has been found to react rapidly
with dioxygen.[32] The redox potential for this complex is not
available, but as it is a cationic species the value is likely to
be positive. In fact, the metal ion is coordinatively unsatu-
rated, and the spectroscopic detection of an intermediate
led the authors to postulate a mechanism similar to that
published a long time ago in the chemistry of porphyrins.[33]
Subsequently, we reported the formation of m-oxo dimers by
reaction of dioxygen with dichloroiron(ii) complexes. The
fact that all complexes exhibit a positive FeII/FeIII couple,
which makes outer sphere reduction of dioxygen unlikely, is
noteworthy.[21,31] A mechanism similar to that occurring
within the chemistry of porphyrins was also postulated. In
our hands, complex 2 is certainly the one that is the most
difficult to handle in the TPA series—it becomes dark im-
mediately, even in the solid state. This is probably due to
the anionic character of the metal center in 2, which actually
decreases the redox potential of the FeII/FeIII couple. No in-
termediate is detected, and we measure a redox couple
close to zero. The potential of the first reduction of molecu-
lar dioxygen is negative (E8=ꢀ0.73 V vs. SCE in DMSO),
although this value is extremely dependant on the condi-
tions, especially organic solvents with different dielectric
constants and/or the presence or not of protons, and can be
shifted to less negative values.[34,35] In our case, we have no
spectroscopic evidence that dioxygen coordinates and
cannot argue this point, although we believe that the low
potential of the metal center, together with the presence of
accessible protons from the catechol, might induce reduction
of dioxygen to superoxide. Superoxide is known to react
with metal salts, sometimes leading to the formation of
metal peroxides.[35,36] In the chemistry of metalloporphyrins,
the presence of amido residues close to the coordination
center can even stabilize such complexes.[37] In the present
case, catechols definitely provide acidic protons that would
react with the superoxide ion. As a consequence, the free
hydroxy group of the ligand is likely to be deprotonated,
Figure 6. Side (A) and bottom (B) views of the coordination environment
in [(BCATTPA)FeIII2Cl2] (3). Selected bond angles [8]: N3-Fe2-Cl2
168.82(11), N1-Fe2-N2 72.00(14), N2-Fe2-N4 68.06(14), N4-Fe2-O3
73.01(14), O3-Fe2-O2 71.93(12), O2-Fe2-N1 74.11(14), O2-Fe1-O3
72.03(12), O3-Fe1-O4 83.21(13), O4-Fe1-O1 103.09(15), O1-Fe1-O2
81.33(14), Cl1-Fe1-O2 129.89(10), Cl1-Fe1-O3 99.59(9), Cl1-Fe1-O4
110.00(12), Cl1-Fe1-O1 103.30(12).
present case the redox wave is reversible, and the value of
10 mV (vs. SCE) that we observe, although still positive, is
not too far from the value of ꢀ38 mV found for [Fe-
(BPG)Cl2], another complex in which a carboxylate sub-
stituent binds to the metal with a di(pyridylmethyl)amine
environment.[25]
Unlike FeCl3, coordination of FeCl2 to the nitrogen-con-
taining part of the ligand was expected. In addition to the
common spectroscopic features observed for dichloroiron(ii)
complexes with simple TPA ligands, complex 2 displays ad-
ditional patterns (LMCT absorption in the UV/Vis spec-
1
C
trum; H NMR spectrum) that indicate that a phenolate is
yielding the highly reactive HO2 radical and leaving no
coordinated, leaving one substituted pyridyl arm uncoordi-
nated, with the metal center lying in an pseudo-octahedral
environment. The molecular conductivity value indicates
that 2 is a charged molecule, in line with our expectation of
coordination of two chloride anions and a phenolate. This
requires a pyridyl unit to be protonated, as drawn in
Scheme 1. At this point we needed a structural confirma-
tion. The fact that 1 and 2 exhibit an identical reversible
chance for any putative intermediate to be detected.
Formation of 3 from 2 requires either decoordination or
decomposition of one ligand. In any case, the reaction is
very fast and is perfectly reproducible, the amount of oxi-
dized material recovered being constant over a series of
more than ten experiments. Analysis of the work-up sol-
vents showed that no BCATTPA was present, only decom-
position products. Whereas free ligand could be recovered
Chem. Eur. J. 2006, 12, 6660 – 6668
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6665

D. Mandon et al.
evaporated to dryness and the remaining oil was dissolved in dichlorome-
thane. This organic phase was washed with aqueous sodium carbonate
and distilled water, and dried over magnesium sulfate. After filtration
and concentration by rotary evaporation, the medium was deposited at
the top of a column filled with aluminum oxide and mounted with di-
chloromethane. The column was washed with diethyl ether, and the com-
pound eluted with a 75/25 diethyl ether/acetone mixture. A pale-yellow
oil (837 mg, 84%) was obtained. Elemental analysis calcd (%) for
C34H34N4O4: C 72.60, H 6.05; found: C 72.85, H 6.21; 1H NMR (CDCl3):
d=8.5 (d, 3J=4.9 Hz, 1Harom), 7.7–7.5 (m, 6Harom), 7.4–6.8 (m, 9Harom),
4.0 (s, 4H), 3.9 (s, 2H), 3.8 (s, 6H), 3.6 ppm (s, 6H).
by oxygenation of the iron(ii) complex in the fluorinated
TPA series,[21] the present result suggests that the formation
of the final species proceeds via a different mechanism, and
C
supports the formation of the reactive HO2 radical. In gen-
eral, m-oxo dinuclear complexes are obtained upon reaction
of dichloroiron(ii) species with molecular dioxygen or organ-
ic peroxides.[21,38] The present situation, however, is differ-
ent: because of the presence of protons and radicals, the
oxygenation conditions are more drastic and lead to some
ligand decomposition. As a consequence, free iron
generated and is trapped by the catechol substituents.
A
Bis[2-(2,3-dihydroxyphenyl)-6-pyridylmethyl)(2-pyridylmethyl)amine
(BCATTPA): BBr3 (10 mL of a 0.8m CH2Cl2 solution) was added, under
argon, to BDMPTPA (680 mg, 1.21 mmol) dissolved in CH2Cl2 (100 mL)
in a Schlenk tube. The medium was refluxed for 4 h under argon. The re-
action mixture was then poured into methanol (200 mL), evaporated to
dryness on a rotary evaporator, and five such cycles of addition of metha-
nol and evaporation of solvent were performed in order to eliminate any
volatile compounds. Finally, the brownish oil was suspended in dichloro-
methane. This suspension was extracted with dilute hydrochloric acid at
pH 1. The acidic phase was washed several times with dichloromethane,
and the organic layer discarded. The pH of the aqueous phase was then
raised to 9, at which a pale yellow compound could be extracted with di-
chloromethane. Addition of hexane, followed by evaporation, yielded a
light brownish solid (350 mg, 57.7%). The 1H NMR spectrum indicated
that this compound was clean. Recrystallization from dichloromethane/
diethyl ether gave the analytically pure ligand as a yellow, microcrystal-
line powder. Elemental analysis calcd (%) for C30H26N4O4: C 71.14, H
5.14; found: C 71.45, H 5.51; 1H NMR (CDCl3): d=12.8 (brs, 2OH), 5.9
(s, 2OH), 8.5 (d, 3J=4.7 Hz, 1Harom), 7.9–7.6 (m, 6Harom), 7.5–6.7 (m,
9Harom), 4.0 ppm (s, 6H).
Only a few examples can be found in literature where the
metal in iron(iii) complexes lies in a seven-coordinate envi-
ACHTREUNG
ronment. In most of these cases, the ligands are potentially
penta- or even heptadentate.[39–41] The occurrence of such a
geometry in our case was unexpected, and means that, to
the best of our knowledge, compound 3 is the only example
of a seven-coordinate iron
TPAs.
A
In conclusion, the new BCATTPA ligand stabilizes coor-
dination of iron(iii) and iron(ii) chloride at the nitrogen-con-
ACHTREUNG
taining site. One hydroxy group from the catechol substitu-
ents is involved in the coordination, which leaves one substi-
tuted pyridyl arm uncoordinated. Coordination of an anionic
phenol residue enhances the kinetics of reaction with O2,
and the presence of free phenol protons is certainly not in-
nocent as it induces a different reactivity to that observed
with neutral derivatives. A radical-based mechanism is very
likely, but this point still needs to be confirmed. Finally, and
to the best of our knowledge, the dinuclear compound
whose structure is reported here is the first example of a
Metalation of BCATTPA with FeCl3: Preparation of 1: A solution of
BCATTPA (100 mg, 0.2 mmol) in THF (50 mL) was added to a THF sol-
ution (100 mL) of anhydrous FeCl3 (32.5 mg, 0.2 mmol). THF or acetoni-
trile can be equally used as reaction solvent. The medium was stirred for
16 h. It was then concentrated under vacuum, and precipitation was in-
duced by addition of diethyl ether. Subsequent solubilization and precipi-
tation with diethyl ether yielded a dark violet solid (120 mg, yield 90%).
Elemental analysis calcd (%) for C30H25Cl2FeN4O4: C 56.96, H 3.95;
seven-coordinate iron
A
tris(2-pyridylmethyl)amine ligands.
found:
C 56.58, H 3.52; UV/Vis (room temperature): l (e)=263.2
(13.67), 297.5 (13.01), 354.0 (shoulder, close to 4), 617.0 nm (0.82
[”103 mmolꢀ1 cm2]). MS FAB+: m/z 560 (BCATTPAFe+). The 1H NMR
spectrum is supplied in the Supporting Information section. No well-de-
fined signals can be observed in the paramagnetic region. Molecular con-
Experimental Section
General: 1H NMR spectra were recorded in CD3CN for the complexes
and CDCl3 for the ligands at ambient temperature on a Bruker AC 300
spectrometer at 300.1300 MHz using the residual signal of CD2HCN
(CHCl3) as a reference for calibration.
ductivity (0.39 mm, 295 K): L=38 Scm2 molꢀ1
. Cyclic voltammetry
(CH3CN, 0.1m (TBA)PF6, 200 mVsꢀ1, room temp.): E1/2 =10 mV (vs.
SCE), along with other waves at E1/2 =460, Ea =1050, and Ea =1300. IR
(Nujol): nOH =3412 cmꢀ1
.
The UV/Vis spectra were recorded with a Varian Cary 05 E UV/Vis NIR
spectrophotometer equipped with an Oxford instrument DN1704 cryo-
stat, using optically transparent Schlenk cells.
Metalation of BCATTPA by FeCl2: preparation of 2: This reaction was
carried out according to an already published procedure.[20] THF or ace-
tonitrile can be equally used as reaction solvent. The bright orange com-
pound obtained with a yield of 85 to 90% is extremely air sensitive, even
in the solid state. We believe that this is the reason why a satisfactory ele-
mental analysis could not be obtained. Elemental analysis calcd (%) for
C30H26Cl2FeN4O4: C 56.87, H 4.11, Cl 11.21, Fe 8.85; found: C 51.97, H
3.75, Cl 8.56, Fe 6.52; C/Cl, Cl/Fe, C/Fe, H/Cl ratios [molar]: calcd: 5.07
[15.0], 1.26 [2.0], 6.42 [30.0], 0.36 [13.0]; found: 5.98 [17.7], 1.31 [2.0],
7.85 [36], 0.44 [15.6]; UV/Vis (room temperature): l (e)=260.0 (11.72),
Conductivity measurements were carried out under argon at 208C with a
CDM 210 Radiometer Copenhagen Conductivity Meter, using a Tacussel
CDC745-9 electrode.
Cyclic voltammetry measurements were performed with a PAR 173 A
potentiostat in a 0.1m acetonitrile solution of (TBA)PF6 (supporting elec-
trolyte), using platinum electrodes and the saturated calomel electrode as
reference.
1
Elemental analyses were carried out by the Service Central dꢀAnalyses
du CNRS in Vernaison, France.
290.0 (13.05), 355.0 (2.98), 446.0 nm (0.72 [”103 mmolꢀ1 cm2]); H NMR
(CD3CN, room temp.): d (FWHH)=127 (br), 50.2 (305 Hz), 47.7 (150),
46.7 (129), 35.6 (452), 20.4 (312), 17.7 (128), ꢀ6.9 (342), 9.0–3.0 ppm
(sharp signals, uncooordinated catecholatopyridyl protons). The spectrum
is supplied in the Supporting Information. Molecular conductivity
(0.09 mm, 295 K): L=107 Scm2 molꢀ1. Cyclic voltammetry (CH3CN, 0.1m
(TBA)PF6, 200 mVsꢀ1, room temperature): E1/2 =10 mV (vs. SCE). We
also observed the presence of other waves at potentials similar to those
reported for 1 (E1/2 =460, Ea =929, Ea =1300).
Bis[2-(2,3-Dimethoxyphenyl)-6-pyridylmethyl](2-pyridylmethyl)amine
(BDMPTPA): Br2TPA[5] (800 mg, 1.78 mmol) and [Pd
(PPh3)4](140 mg,
G
0.12 mmol) were suspended in degassed toluene (100 mL) under argon.
Sodium carbonate (7.2 mL of a 2m aqueous solution) and commercially
available 2,3-dimethoxyphenylboronic acid (1 g, 5.5 mmol) dissolved in
ethanol (5 mL) were added under argon. The reaction mixture was re-
fluxed under argon for 24 h. The orange-colored solution was then
6666
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 6660 – 6668

A Diiron(iii) Compound with an Unusual Coordination Environment
A
FULL PAPER
Reduction of 1 with Zn/Hg: A solution of 1 in dry and degassed CH3CN
(100 mg in 100 mL) was transferred into a Schlenk tube containing a few
drops of 10% Zn/Hg. The medium was stirred and the color turned
bright orange within 10 min. It was then filtered, concentrated, and a
bright orange solid was obtained upon addition of diethyl ether. The
solid was washed several times with dry and degassed diethyl ether and
dried under vacuum. All spectroscopic features (UV/Vis, cyclic voltam-
metry, 1H NMR) are identical to those obtained from a clean sample of
2.
Both organisations are gratefully acknowledged for their support. The
authors thank Professor Richard Welter and Laboratoire DECOMET for
the X-ray diffraction studies. We wish to give special thanks to Dr. Remy
Louis, for his constant encouragement and support. The Institut de
Chimie, CNRS and ULP are gratefully acknowledged for financial sup-
port.
[1]A. G. Blackman, Polyhedron 2005, 24, 1–39.
[2]Y. Zang, J. Kim, Y. Dong, E. C. Wilkinson, E. H. Appelman, L.
Que, Jr., J. Am. Chem. Soc. 1997, 119, 4197–4205.
[3]K. Chen, L. Que, Jr., J. Am. Chem. Soc. 2001, 123, 6327–6337.
[4]M. P. Jensen, S. J. Lange, M. P. Mehn, E. L. Que, L. Que, Jr., J. Am.
Chem. Soc. 2003, 125, 2113–2128.
[5]C. L. Chuang, O. Dos Santos, X. Xu, J. W. Canary, Inorg. Chem.
1997, 36, 1967–1972.
[6]M. Harata, K. Jitsukawa, H. Masuda, H. Einaga, Chem. Lett. 1995,
61–62.
[7]S. Ogo, S. Wada, Y. Watanabe, M. Iwase, A. Wada, M. Harata, K.
Jitsukawa, H. Masuda, H. Einaga, Angew. Chem. 1998, 110, 2198–
2200; Angew. Chem. Int. Ed. 1998, 37, 2102–2104.
[8]K. Rudzka, A. M. Arif, L. M. Berreau, Inorg. Chem. 2005, 44, 7234–
7242.
Oxygenation of 2: preparation of 3: Complex 2 (50 mg) was dissolved in
dry and degassed CH3CN (50 mL). Upon exposure to dry dioxygen, the
medium turned dark violet within a fraction of a second. The solution
was concentrated and a black solid was obtained upon addition of diethyl
ether. The solid was washed with three portions (50 mL each) of diethyl
ether, dried under vacuum, and recrystallized from CH3CN/diethyl ether
to give 20 mg of product The three portions of diethyl ether used for
washing were collected, the solvent evaporated. The corresponding
1H NMR spectrum showed no evidence for the presence of any free
ligand amongst the numerous and unassignable signals in the range d=
3–8 ppm. Elemental analysis calcd (%) for C30H22Cl2Fe2N4O4: C 52.55, H
3.21; found: C 52.90, H 3.62; UV/Vis (room temperature): l (e)=266.0
(12.67), 300.0 (14.64), 350.0 (shoulder), 590.0 nm (1.10[”103 mmolꢀ1 cm2]).
The 1H NMR spectrum is supplied in the Supporting Information. No
well defined signals can be observed in the paramagnetic region.
[9]Z. He, P. J. Chaimungkalanont, D. C. Craig, S. B. Colbran, J. Chem.
Soc. Dalton Trans. 2000, 1419–1429.
X-ray analysis
Single crystals of [(BCATTPA)FeIIICl2]·Et2O and [(BCATTPA)-
FeIII2Cl2]·CH3CN were mounted on a Nonius Kappa-CCD area detector
diffractometer (MoKa radiation; l=0.71073 Š). Quantitative data were
obtained at 173 K for both complexes. The complete conditions of data
collection (Denzo software) and structure refinements are given as Sup-
porting Information. The cell parameters were determined from reflec-
tions taken from one set of 10 frames (1.08 steps in phi angle), each at
20 s exposure. The structures were solved by direct methods (SIR97) and
refined against F2 using the SHELXL97 software suite (Kappa CCD Op-
eration Manual, Nonius B.V., Delft, The Netherlands, 1997; G. M. Shel-
drick, SHELXL97, Program for the refinement of crystal structures, Uni-
versity of Gçttingen, Germany, 1997). The absorption was not corrected.
All non-hydrogen atoms were refined anisotropically. Hydrogen atoms
were generated according to stereochemistry and refined using a riding
model in SHELXL97.
[10]Z. He, D. C. Craig. , S. B. Colbran, J. Chem. Soc. Dalton Trans.
2002, 4224–4235.
[11]Z. He, S. B. Colbran, D. C. Craig, Chem. Eur. J. 2003, 9, 116–129.
[12]A. Diebold, K. H. Hagen, Inorg. Chem. 1998, 37, 215–223.
[13]M. Costas, K. Chen, L. Que, Jr., Coord. Chem. Rev. 2000, 517–544.
[14]M. Costas, M. Mehn, M. P. Jensen, L. Que, Jr., Chem. Rev. 2004,
104, 939–986.
[15]E. I. Solomon, T. C. Brunold, M. I. Davis, J. N. Kemsley, S.-K. Lee,
N. Lehnert, F. Neese, A. J. Skulan, Y. S. Yang, J. Zhou, Chem. Rev.
2000, 100, 235–350.
[16]H. G. Jang, D. D. Cox, L. Que, Jr., J. Am. Chem. Soc. 1991, 113,
9200–9204.
[17]J. H. Lim, T. H. Park, H. J. Lee, K. B. Lee, H. G. Jang, Bull. Korean
Chem. Soc. 1999, 20, 1428–1432.
[18]F. Pattus, M. A. Abdallah, J. Chin. Chem. Soc. 2000, 47, 1–20.
[19]C. Belle, J.-L. Pierre, E. Saint-Aman, New J. Chem. 1998, 22, 1399–
1402.
[20]D. Mandon, A. Machkour, S. Goetz, R. Welter, Inorg. Chem. 2002,
41, 5364–5372.
[21]A. Machkour, D. Mandon, M. Lachkar, R. Welter, Inorg. Chem.
2004, 43, 1545–1550.
[22]A. Machkour, D. Mandon, M. Lachkar, R. Welter, Eur. J. Inorg.
Chem. 2005, 158–161.
[23]M. M. Makowska-Grzyska, E. Szajna, C. Shipley, A. M. Arif, M. H.
Mitchell, J. A. Halfen, L. M. Berreau, Inorg. Chem. 2003, 42, 7472–
7488.
[24]D. Mandon, A. Nopper, T. Litrol, S. Goetz, Inorg. Chem. 2001, 40,
4803–4806.
[25]M. C. Rodrigez, F. Lambert, I. Morgenstern-Badarau, M. Cesario, J.
Guilhem, B. B. Keita, L. Nadjo, Inorg. Chem. 1997, 36, 3525–3531.
[26]A. Machkour, D. Mandon, M. Lachkar, R. Welter, Inorg. Chim.
Acta 2005, 358, 839–843.
Crystal data for [(BCATTPA)FeIIICl2]·Et2O (1·Et2O): Violet crystals,
0.10”0.10”0.10 mm3; C34H35Cl2FeN4O5, M=706.41 gmolꢀ1. Monoclinic,
space group P21/c, a=14.360(5), b=13.589(5), c=16.874(5) Š, b=
94.654(5)8, V=3281.9(19) Š3, 1calcd =1.430 gcmꢀ3
,
Z=4, 1.428<q<
29.178. Of 8835 total reflections, 4546 were considered to be observed
[I>2s(I)], with 405 parameters. Final results: R=0.0623 and Rw =0.1553;
GOF=0.879; maximum residual electronic density: 1.068; minimum re-
sidual electronic density: ꢀ0.913.
Crystal data for [(BCATTPA)FeIII2Cl2]·CH3CN (3·CH3CN): Dark red
crystals, 0.10”0.07”0.03 mm3; C32H25Cl2Fe2N4O5, M=726.17 gmolꢀ1. Or-
thorhombic, space group P212121, a=12.6320(10), b=14.631(2), c=
16.550(2) Š, V=3058.8(6) Š3, 1calcd =1.577 gcmꢀ3
, Z=4, 1.868<q<
30.048. Of 8900 total reflections, 5877 were considered to be observed
[I>2s(I)], with 406 parameters. Final results: R=0.0567 and Rw =0.1218;
GOF=0.936; Flack x: 0.00(2); maximum residual electronic density:
0.544; minimum residual electronic density: ꢀ0.717.
CCDC-298849 and CCDC-298850 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of charge
from the Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif.
[27]N. K. Thallaj, A. Machkour, D. Mandon, R. Welter, New J. Chem.
2005, 29, 1555–1558.
[28]A. G. Blackman, C. R. Chim. 2004, 7, 107–119.
[29]T. J. Hubin, J. M. McCormick, S. R. Collinson, M. Buchalova, C. M.
Perkins, N. W. Alcock, P. K. Kahol, A. Raghunatan, D. H. Busch, J.
Am. Chem. Soc. 2000, 122, 2512–2522.
[30]D.-H. Jo, J.-M. Chiou, L. Que, Jr., Inorg. Chem. 2001, 40, 3181–
3190.
Acknowledgments
[31]The values of the redox potentials of the many dichloroiron( ii) com-
plexes studied by us are all systematically positive and will be pub-
lished later.
This work was supported as a collaborative action between CNRS (Proj-
ect no. 17703-France) and CNRST (Project no. chimie 08/05-Morocco).
[32]Y. Zang, L. Que, Jr., Inorg. Chem. 1995, 34, 1030–1035.
Chem. Eur. J. 2006, 12, 6660 – 6668
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6667

D. Mandon et al.
[33]D. H. Chin, J. Del Gaudio, G. La Mar, A. Balch, J. Am. Chem. Soc.
1977, 99, 5486–5488.
[38]S. J. Lange, H. Miyake, L. Que, Jr., J. Am. Chem. Soc. 1999, 121,
6330–6331.
[39]D. S. Marlin, M. M. Olmstead, P. Mascharak, Inorg. Chim. Acta
2000, 297, 106–114.
[40]I. Morgenstern-Badarau, F. Lambert, J.-P. Renault, M. Cesario, J.-D.
Marechal, F. Maseras, Inorg. Chim. Acta 2000, 297, 338–350.
[41]I. Ivanovic-Burmazovic, M. S. Hamza, R. van Eldik, Inorg. Chem.
2002, 41, 5150–5161.
[34]A. D. Goolsby, D. T. Sawyer, Anal. Chem. 1968, 40, 83–86.
[35]J. S. Valentine, A. E. Quinn, Inorg. Chem. 1976, 15, 1997–1999.
[36]E. McCandlish, A. R. Miksztal, M. Nappa, A. Q. Sprenger, J. S. Val-
entine, J. D. Stong, T. G. Spiro, J. Am. Chem. Soc. 1980, 102, 4268–
4271.
[37]M. Schappacher, R. Weiss, R. Montiel-Montoya, A. X. Trautwein,
A. Tabard, J. Am. Chem. Soc. 1985, 107, 3736–3738.
Received: February 28, 2006
Published online: June 21, 2006
6668
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 6660 – 6668