Vanadium(IV and V) complexes of Schiff bases and reduced Schiff bases derived from the reaction of aromatic o-hydroxyaldehydes and diamines: Synthesis, characterisation and solution studies

Article abstract of DOI:10.1002/ejic.200400481

Several sal₂en-type reduced Schiff bases have been prepared from the reaction of 2 equiv. of salicylaldehyde (or derivatives) with ethylenediamine followed by reduction with NaBH₄ and subsequently characterised. The water-soluble ligands (SO₃-sal)₂en and R(SO₃-sal)₂en (SO₃-sal = salicylaldehyde-5- sulphonate) have been studied by pH potentiometry and ¹H NMR spectroscopy and the protonation constants of R(SO₃-sal) ₂en have been determined. The Schiff bases are more susceptible to hydrolysis than the corresponding reduced Schiff bases. The crystal structures of H₂(o-van)₂en and KH(o-van)₂en (o-van = o-vanillin) have been determined by X-ray diffraction. The vanadium(IV) complexes have been prepared and characterised by magnetic measurements and IR, UV/Vis and EPR spectroscopy. The colours, magnetic susceptibilities and IR spectra of the new vanadium(IV) complexes of the reduced Schiff bases now obtained suggest oligomeric/polymeric structures for the compounds while the corresponding V^IVO-Schiff-base complexes are monomeric. The complexation of [V^IVO]²⁺ and [V^VO ₂]⁺ with R(SO₃-sal)₂en in aqueous solution has been studied by pH potentiometry, UV/Vis as well as by EPR spectroscopy for the V^IVO system and ¹H and ⁵¹V NMR spectroscopy for the V^VO₂ system. Complex formation constants have been determined and binding modes have been proposed. N,N′-ethylenebis(pyridoxylaminato) (Rpyr₂en) has also been obtained and the structure of Na[V^VO₂(Rpyr ₂en)]·CH₃OH·3H₂O determined by X-ray diffraction. The Rpyr₂en^2- ligand coordinates through both the N-amine and O-phenolate moieties in a symmetrical α-cis type coordination mode, i.e. with both O-phenolato donors cis to the O-oxo atoms and trans to each other. Wiley-VCH Verlag GmbH & Co. KGaA, 2005.

Full text of DOI:10.1002/ejic.200400481

FULL PAPER
Vanadium(IV and ) Complexes of Schiff Bases and Reduced Schiff Bases
V
Derived from the Reaction of Aromatic o-Hydroxyaldehydes and Diamines:
Synthesis, Characterisation and Solution Studies
Isabel Correia,[a] João Costa Pessoa,*[a] M. Teresa Duarte,[a]
M. Fátima Minas da Piedade,[a] Tamás Jackush,[b] Tamás Kiss,*[b]
M. Margarida C. A. Castro,[c] Carlos F. G. C. Geraldes,[c] and Fernando Avecilla[d]
Keywords: Vanadium complexes / Schiff bases / Reduced Schiff bases / Crystal structures / Speciation
Several sal2en-type reduced Schiff bases have been pre-
pared from the reaction of 2 equiv. of salicylaldehyde (or de-
rivatives) with ethylenediamine followed by reduction with
NaBH4 and subsequently characterised. The water-soluble
ligands (SO3-sal)2en and R(SO3-sal)2en (SO3-sal = salicylal-
dehyde-5-sulphonate) have been studied by pH potentiome-
try and H NMR spectroscopy and the protonation constants
of R(SO3-sal)2en have been determined. The Schiff bases are
more susceptible to hydrolysis than the corresponding re- structure of Na[VVO2(Rpyr2en)]·CH3OH·3H2O determined
corresponding VIVO–Schiff-base complexes are monomeric.
The complexation of [VIVO]2+ and [VVO2]+ with R(SO3-sal)2en
in aqueous solution has been studied by pH potentiometry,
UV/Vis as well as by EPR spectroscopy for the VIVO system
and 1H and 51V NMR spectroscopy for the VVO2 system.
Complex formation constants have been determined and
binding modes have been proposed. N,NЈ-ethylenebis(pyri-
doxylaminato) (Rpyr2en) has also been obtained and the
1
duced Schiff bases. The crystal structures of H2(o-van)2en
and KH(o-van)2en (o-van = o-vanillin) have been determined
by X-ray diffraction. The vanadium(IV) complexes have been
prepared and characterised by magnetic measurements and
IR, UV/Vis and EPR spectroscopy. The colours, magnetic
susceptibilities and IR spectra of the new vanadium(IV) com-
plexes of the reduced Schiff bases now obtained suggest
oligomeric/polymeric structures for the compounds while the
by X-ray diffraction. The Rpyr2en2– ligand coordinates
through both the N-amine and O-phenolate moieties in a
symmetrical α-cis type coordination mode, i.e. with both O-
phenolato donors cis to the O-oxo atoms and trans to each
other.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
diamine) type ligands have been proposed for use as insulin-
enhancing agents and for treatment of obesity and hyper-
tension.[13–15] Saminathan and Durai[13] reported the ability
of VIVO(sal2en) to reverse the hyperglycaemic condition of
alloxan-induced diabetic rats, thereby exerting insulin-like
action. However, the rats tended to become hypoglycaemic
and withdrawal of treatment brought an immediate return
of hyperglycaemia.
The sal2en-type ligands present versatile steric, electronic
and lipophilic properties. They may be easily prepared by
the condensation of two compounds, namely an aromatic
o-hydroxyaldehyde and a diamine, the hydrophilic-lipophilic
balance being easily fine-tuned by choosing the appropriate
amine precursors and ring substituents of the aldehyde.
In solution, Schiff base (SB) ligands have the disadvan-
tage of their potential for hydrolysis. This instability can be
overcome by reduction of the SB to give an amine. This
presents interesting possibilities since the reduced SB will
be more flexible and not forced to remain planar when co-
ordinated to a metal centre. In this paper, we report the
preparation of the SB and reduced SB ligands of salicylal-
dehyde, o-vanillin (o-van) and salicylaldehyde-5-sulphonate
Introduction
The presence of vanadium in biological systems, its insu-
lin-enhancing action[1–4] and anticancer activity[5–7] has
driven a considerable amount of research. Particular inter-
est has been given to the study of the potential benefits
of vanadium compounds as oral insulin substitutes for the
treatment of diabetes.[8–12] Coordinated ligands should be
able to improve the absorption and intracellular mobility of
vanadium, reducing the dose necessary for producing equi-
valent effects. Several vanadium complexes of the dianionic
tetradentate Schiff base sal2en (N,NЈ-disalicylideneethylene-
[a]
Centro Química Estrutural, Instituto Superior Técnico,
Av. Rovísco Pais, 1049-001 Lisboa, Portugal
[b]
Biocoordination Chemistry Research Group of the Hungarian
Academy of Sciences, University of Szeged,
P. O. Box 440, 6701 Szeged, Hungary
[c]
Departamento de Bioquímica, Centro de RMN e Centro de
Neurociências e Biologia Celular, Universidade de Coimbra,
3001-401 Coimbra, Portugal
[d]
Departamento de Química Fundamental, Universidade da Co-
ruña,
Campus de A Zapateira, 15071 A Coruña, Spain
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
732
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200400481
Eur. J. Inorg. Chem. 2005, 732–744

Vanadium(iv and v) Complexes of Schiff Bases and Reduced Schiff Bases
FULL PAPER
(SO3-sal) with ethylenediamine and the preparation of their studies indicated the presence of various isomeric species in
[VIVO]2+ complexes which were characterised in the solid solution.
state and in solution.
Some of the vanadium(iv) complexes synthesised and
Most of vanadium’s biologically important reactions oc- characterised in the present and in the previous study[15]
cur in water-based environments such as blood plasma and have been tested in vitro for their toxicity and insulin-mi-
intracellular fluids. A knowledge of the distribution and metic behaviour, these are[16] Cs2[VIVO(SO3-sal)2en] (8), [VI-
chemical speciation of the vanadium compounds in aque- VO(Rsal2en)] (9), [VIVO(pyr2en)] and [VIVO(Rpyr2en)]. At
ous solution is, therefore, of the utmost importance. Most vanadium concentrations below 10 µm none of the com-
of the sal2en-type ligands and vanadium complexes pre- pounds were toxic and they all proved to be effective with
pared so far are not water-soluble. The compounds pre- respect to the insulin-mimetic activity which was tested as
pared using SO3-sal are, however, soluble in water and may, the ability of vanadium compounds to trigger glucose up-
therefore, be particularly useful for therapeutic use.
take by cells. In that study, compound 8 was considered one
The study of the moderately water-soluble Schiff base of of the most promising compounds with respect to its insu-
pyridoxal and ethylenediamine (pyr2en) and its hydrolyti- lin-mimetic behaviour and low toxicity in the physiological
cally more stable reduced derivative Rpyr2en was reported concentration range.[16]
in a previous paper.[15] Both proved to be efficient binders
of vanadium in the +iv and +v oxidation states. Several
VIVO– and VVO2–Rpyr2en complexes were prepared and
their properties studied. One of these, [VVO2(HRpyr2en)],
was isolated in a crystalline form and structurally character-
ised by X-ray diffraction. Both ligands pyr2en and Rpyr2en
Results and Discussion
Synthesis and Characterisation
The Schiff bases 1–3 (see Scheme 1) were prepared in
were found to form basically similar complexes coordinat- high yields (Ͼ75%) by the condensation of 2 equiv. of the
ing through two Ophenolate centres and two Namine/imine appropriate aldehyde derivative with 1 equiv. of ethylenedia-
centres. The solution speciation revealed that the reduced mine. Treatment of these condensation products with so-
SB formed more stable complexes with both [VIVO]2+ and dium borohydride resulted in the reduction of the two imine
[VVO2]+ than the SB itself. EPR (VIVO–Rpyr2en system), bonds, yielding the reduced Schiff bases 4–7 (Scheme 1). All
1H and 51V NMR (VVO2–Rpyr2en system) spectroscopic compounds gave satisfactory elemental analyses and were
Scheme 1. Ligands and [VIVO]2+ complexes prepared and characterised in this work
Eur. J. Inorg. Chem. 2005, 732–744
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
733

J. C. Pessoa, T. Kiss et al.
FULL PAPER
characterised by 1H NMR spectroscopy and mass spec-
trometry. Two types of crystals suitable for X-ray diffraction
studies were obtained for 2.
By treatment of 3–5 with oxovanadium(iv) salts, [VIVO]
complexes with an N2O2 binding mode were obtained in
good yields, namely [VIVO{(SO3-sal)2en}] (8),[17] [VIVO-
(Rsal2en)] (9) and [VIVO{R(o-van)2en}] (10). The com-
plexes were characterised by elemental analysis, IR and UV/
Vis spectroscopy. All our efforts to characterise the com-
pounds obtained from the reaction of 6 with [VIVO]2+ were
unsuccessful. Solids were obtained but no satisfactory for-
mulation could be deduced. Compound 11 was obtained
after slow concentration of a methanolic solution which
contained a VV salt and compound 7.
X-ray Diffraction Studies
H2(o-van)2en (2A) and KH(o-van)2en·2H2O (2B)
ORTEP diagrams of 2A and 2B are presented in Figure 1
and selected bond lengths and angles are given in Table 1.
A similar crystal structure for 2A has already been re-
ported[18] but we have included our data since it is of a
better resolution (taking into account the R and e.s.d. val-
ues; we also included the hydrogen atoms in the refine-
ment). The molecular structure of 2A agrees with the pub-
lished data[18] and both exhibit two geometry-determined
intramolecular H-bonds between the Nimine and the
Ophenolic centres in both sides of the molecule [O–H···N
2.576(3), 2.608(3) Å with angles of 147.4(2)° and 146.8(2)°
which dictate the molecular configuration and an N–C–C–N
torsion angle of –163.0(2)° (–160.91° in ref.[18])]. The bond
Figure 1. ORTEP diagrams of H2(o-van)2en (2A) and KH(o-van)
2en·2H2O (2B), showing the atom labelling scheme; thermal ellip-
soids of the non-hydrogen atoms have been drawn at the 40% prob-
ability level
The molecular structure of 2A is comparable with that
lengths and angles do not show any significant deviation of 2B, the latter being the same organic moiety but crystal-
from the expected values in similar compounds.[15,19,20] lised as an ionic pair with K+ and two water molecules of
Another structure for H2(o-van)2en has recently been re- crystallisation. In 2B, the presence of K+ changes the type
ported.[21] However, it is impossible to access this report to of intramolecular H-bonds formed, thus promoting a tot-
make adequate comparisons.[22]
ally different conformation of the organic moiety. One of
Table 1. Selected bond lengths [Å] and angles [°] for H2(o-van)2en (2A) and KH(o-van)2en·2H2O (2B); # symmetry operation: –x, –y, –z
H2(o-van)2en KH(o-van)2en·2H2O H2(o-van)2en KH(o-van)2en·2H2O
Bond lengths
1.461(3) N(2)–C(1)
Angles
N(1)–C(2)
1.509 (6) C(2)–N(1)–C(3)
119.0(2) C(1)–N(2)–C(10)
120.4(2) C(10)–C(11)–C(16)
121.4(2) N(2)–C(10)–C(11)
122.0(2) O(4)–C(16)–C(11)
135.2(4)
117.0(5)
135.0(4)
111.3(4)
123.4(3)
111.3(4)
110.8(4)
N(1)–C(3)
O(3)–C(8)
N(2)–C(10)
O(4)–C(16)
1.275(4)
1.347(3)
1.265(4)
1.355(3)
N(2)–C(10)
O(4)–C(16)
O(3)–C(8)
1.282(6)
1.305(6)
1.260(5)
1.156(6)
2.403(4)
2.497(3)
1.488(7)
1.413(7)
1.182(5)
1.422(7)
1.425(7)
1.156(7)
1.405(8)
1.768(7)
4.126(5)
2.554(5)
C(3)–C(4)–C(5)
N(1)–C(3)–C(4)
O(3)–C(5)–C(4)
N(1)–C(2)–C(1)
C(4)–C(9)–O(3)
N(1)–C(3)
O(3)–K(1)
O(4)#-K(1)
C(10)–C(11)
C(11)–C(12)
C(11)–C(16)
C(16)–C(15)
O(8)–C(18)
C(14)–C(13)
C(13)–C(12)
C(1)–C(2)
107(2)
122.3
N(2)–C(1)–C(2)
C(9)–C(8)–O(3)
C(11)–C(16)–O(4)
C(3)–C(4)
C(4)–C(5)
C(4)–C(8)
C(5)–C(8)
O(7)–C(17)
C(7)–C(6)
C(6)–C(5)
C(2)–C(1)
O(3),N(1)
O(4),N(2)
1.458(4)
1.402(4)
1.399(3)
1.411(3)
1.422(3)
1.388(5)
1.364(5)
1.527(4)
2.600(3)
2.567(3)
O(4),N(2)
O(3),N(1)
734
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 732–744

Vanadium(iv and v) Complexes of Schiff Bases and Reduced Schiff Bases
FULL PAPER
the intramolecular interactions is retained [O(4)-H···N(2) and only one of the pyridine N atoms is protonated. From
2.556(5) Å] but the other is lost due to the direct interaction methanolic solutions containing the ligand and a different
of the phenolic oxygen atom with the K+ cation. A com- metal precursor (Bu4NVO3), a new compound was ob-
pletely different conformation is thus obtained with an N– tained, namely Na[VVO2(Rpyr2en)]·CH3OH·3H2O (11), its
C–C–N torsion angle of –71.8(5)°. In the crystal structure molecular structure is shown in Figure 2 and some selected
of 2B, the molecules pack as pseudo dimers due to the in- bond lengths and angles are included in Table 2.Both 11
teractions with the K+ cations and an extensive H-bonding and 12 correspond to the symmetrical α-cis-type com-
system is promoted by the water molecules (see section ESI- plex.[15] The molecular structures are similar and the differ-
A of the Supporting information for details).Figure ESI-1 ence that must be emphasised is that in the present com-
(Supporting information) shows the cubic-like cage formed pound 11, both pyridinic N atoms are deprotonated. Thus,
by the hydroxyl-oxygen, the K+ ion and the O atoms of the both halves of the ligand become equivalent whereas they
water molecules. The coordination number of K+ is six and are different in 12. The complex molecule in 11 is anionic
it adopts a twisted trigonal-prismatic geometry (where the and there is an Na+ counterion in the crystal structure as
Ox···K+ interactions, with x = hydroxy or water, range from well as one methanol and three water molecules.
2.699 to 3.054 Å).
The V–Ooxo, V–Ophenolate and V–Namine bond lengths are
similar in 11 and 12 but (i) the difference between the two
V–Ooxo internuclear distances is 0.019 Å in 11 (not signifi-
cant within a 3σ criterion) while it is 0.056 Å in 12 (this is
probably due to different H-bonding interactions of the
Ooxo atoms in 11 and 12) and (ii) in the two halves of
[VVO2(Rpyr2en)]–, the two V–Ophenolate and V–Namine bond
lengths differ by 0.003 (not significant) and 0.02 Å, respec-
tively, while they differ by 0.07 and 0.06 Å, respectively, in
[VVO2(HRpyr2en)]. This is due to the deprotonation of
both pyridine N atoms in 11. The similarity of both halves
of the ligand can also be observed in the mean deviation
from the plane of the atoms of the chelating rings which
in 11 is the same in both V–O–C–C–C–N chelating rings
(0.292 Å) while they are different in 12 (0.282 and 0.300 Å).
The H-bonding network is quite complex. Several atoms
are involved in H-bonding with the neighbouring solvent
molecules and this is described in section ESI-A (Support-
ing information). The Na+ cation interacts strongly with
O(4) [Na–O(4) 2.424(6) Å] and the water molecules (see
Table 2).
Na[VVO2(Rpyr2en)]·CH3OH·3H2O (11)
Rpyr2en is stable to hydrolysis and H and 51V NMR
1
spectroscopic studies for the VVO2–Rpyr2en system indi-
cated the presence of various complexes in solution, namely
α-cis- and β-cis-type isomers[15] (Scheme 2). The pyridinium
N atoms of the pyridoxal rings do not take part in the coor-
dination but are involved in acid-base reactions which affect
the number, type and relative amount of the isomers of the
VVO2–Rpyr2en complexes present in solution.[15] DFT cal-
culations provided sensible molecular structures and the en-
ergy values for the various α-cis- and β-cis-type isomeric
complexes differ by less than 4 kcal mol–1, confirming that
both types may coexist in solution. The α-cis isomer is sym-
metric and corresponds to the structure presently deter-
mined by X-ray diffraction.
Infrared Spectra
Scheme 2. Schematic representation of the binding mode of the α-
and β-cis isomers in the present VO2 complexes
A set of selected IR bands for compounds 1–12 are sum-
marised in section ESI-B (Supporting information). The as-
+
The structure of [VVO2(HRpyr2en)] (12) obtained from signments were made based on refs.[15,23,24] and by compari-
aqueous solutions at pH Ϸ 7 was reported previously.[15] son of the spectra. While the spectrum of 2 in the powder
The ligand coordinates to the VV centre through the two and crystalline forms shows only one ν(C=N) strong band
Namine and two Ophenolate atoms. The molecule is neutral at 1633 cm–1, the IR spectrum of KH(o-van)2en shows two
Figure 2. ORTEP diagram of Na[VO2(Rpyr2en)]·CH3OH·3H2O (11) showing the atom labelling scheme; thermal ellipsoids have been
drawn at the 30% probability level
Eur. J. Inorg. Chem. 2005, 732–744
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
735

J. C. Pessoa, T. Kiss et al.
FULL PAPER
Table 2. Selected bond lengths [Å] and angles [°] for Na[VVO2(Rpyr2en)]·CH3OH·3H2O (11)
Bond lengths
Angles
V(1)–O(1V)
V(1)–O(2V)
V(1)–O(3)
V(1)–O(2)
V(1)–N(2)
V(1)–N(3)
O(3)–C(13)
O(2)–C(6)
N(2)–C(8)
N(3)–C(11)
N(2)–C(9)
N(3)–C(10)
1.643(5)
1.662(5)
1.957(4)
1.960(5)
2.284(6)
2.305(6)
1.347(7)
1.354(7)
1.472(8)
1.468(8)
1.464(9)
1.479(9)
O(1V)–V(1)–O(2V)
O(1V)–V(1)–O(3)
O(2V)–V(1)–O(3)
O(1V)–V(1)–O(2)
O(2V)–V(1)–O(2)
O(3)–V(1)–O(2)
O(1V)–V(1)–N(2)
O(2V)–V(1)–N(2)
O(3)–V(1)–N(2)
O(2)–V(1)–N(2)
O(1V)–V(1)–N(3)
O(2V)–V(1)–N(3)
O(3)–V(1)–N(3)
O(2)–V(1)–N(3)
N(2)–V(1)–N(3)
107.4(3)
96.5(2)
96.0(2)
98.4(2)
97.4(2)
155.9(2)
88.9(2)
163.8(2)
82.0(2)
79.6(2)
162.8(2)
89.6(2)
78.34(19)
81.87(19)
74.2(29)
Bond lengths and angles involving the water molecules
X1–H–X2
d(X1–H)
d(H–X2)
d(X1–X2)
angle(X1–H–X2)
O1–H10–O2S
0.90
2.33
2.802(8)
3.188(8)
2.843(8)
2.802(8)
3.288(9)
2.812(8)
2.868(7)
112.3
171(7)
166(9)
145(9)
149(6)
172(9)
154(9)
N2–H2N–O4S
O2S–H2SB–O2V
O2S–H2SA–O1
N3–H3N–O2S
O4S–H4SA–N4
O4S–H4SB–O3
1.000(2)
0.900(2)
0.900(2)
1.000(2)
0.900(2)
0.900(2)
2.196(16)
1.96(2)
2.02(6)
2.39(4)
1.918(14)
2.93(4)
strong bands at 1638 and 1617 cm–1. This may be explained the range 500–600 nm (band II: dxyǞdx2–y2) and a shoulder
by the different H-bonding networks.
at 600–710 nm (band I: dxyǞdxz, dyz).[29] Band III
The characteristic ν(V=O) of oxovanadium(iv) com- (dxyǞdz2) occurs below 500 nm and is under much stronger
plexes appears as a medium-strong band at 960 cm–1 in the LMCT bands (probably OphenolateǞd orbital).
IR spectrum of 8, in agreement with the literature.[17] For
the VIVO complexes of the reduced SB, the ν(V=O) is
shifted to lower wavenumbers by ca. 100 cm–1. These low
EPR Spectra
ν(V=O) values indicate V=O···V=O interactions and this is
The X-band EPR spectra of frozen solutions (77 K) of
in agreement with the magnetic susceptibility measurements
the vanadium complexes, dissolved in DMSO (except for 8
(see below). The binding of an Ooxo atom to an adjacent
vanadium atom trans to its vanadyl O atom lengthens and
weakens the bond, thereby lowering the V=O stretching fre-
which was in water), were measured and simulated with a
computer program by Rockenbauer.[30] The EPR parame-
ters obtained are listed in Table 3. All spectra show a small
quency.
rhombic distortion which can be observed in the perpendic-
ular lines corresponding to MI = 7/2, 5/2 and 3/2, indicating
a distorted geometry around the metal centre. Although the
additivity rule[31] was developed to allow the determination
of the identity of the equatorial donor groups in complexes
of square-pyramidal geometry (or octahedral with a weak
sixth ligand), it has been successfully applied to structurally
distorted molecules.[32,33] In the present [VIVO] complexes,
the EPR parameters fit well with a binding mode involving
either two equatorial Namine and two Ophenolate centres or
two equatorial Nimine and two Ophenolate centres. However,
since the contribution of equatorial DMSO ligands to Az is
Electronic Absorption Spectra
The electronic absorption spectra of bis(salicylaldimines)
and their transition metal complexes have been extensively
studied.[25–28] All free Schiff bases 1–3 show a broad band
centred at ca. 320 nm due to an nǞπ* transition associated
with the azomethine group and another band at around
260 nm which may arise from a transition involving mainly
the π molecular orbitals localised on the C=N group and/
or the aromatic ring.[26] In the spectra of the reduced SB
compounds, the nǞπ* transition is absent as expected and
the 260 nm transition is red-shifted and less intense. Section
ESI-C (Supporting information) summarises the UV/Vis
data.
ca. 42 × 10–4 cm–1, i.e. close to that of Namine [15] other bind-
,
ing modes cannot be ruled out.
Although three (or four) d-d bands can be expected to
appear in the spectra of the VIVO complexes in the visible
region, they are often overlapping or obscured by strong
Magnetic Properties
The magnetic susceptibilities of the vanadium complexes
LMCT bands. The spectra of 8–10 show broad bands in 8 and 9 were measured by the Faraday method in the tem-
736
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org Eur. J. Inorg. Chem. 2005, 732–744

Vanadium(iv and v) Complexes of Schiff Bases and Reduced Schiff Bases
FULL PAPER
Table 3. Spin-Hamiltonian parameters obtained by computer simulation[42]
8
9
10
gx, Ax (×104 cm–1)
gy, Ay (×104 cm–1)
gz, Az (×104 cm–1)
1.983, 45.6
1.980, 50.5
1.960, 157.7
1.982, 45.4
1.977, 50.2
1.957, 158.2
1.980, 48.5
1.977, 50.4
1.958, 158.0
perature range of 3 to 287 K. Data were corrected for dia- Speciation Studies
magnetic contributions using the Pascal’s constants.[34] For
both complexes the plot of 1/χP vs. T shows deviations from
(SO3-sal) en (3) and R(SO3-sal) en (6)
2
2
linearity which can be accounted for by assuming a Curie–
Compounds 3 and 6 are water-soluble and their acid-
1
Weiss law and a temperature-independent paramagnetism base behaviour was studied by pH potentiometry and H
(TIP) term. The magnetic susceptibility of compound 8 fits NMR spectroscopy. The protonation constants of 6 are
well the Curie–Weiss law. A µeff of 1.76 µB was obtained at listed in Table 4. For the SB 3, both the evaluation of the
1
room temperature which is typical for compounds with a titration curves and the H NMR studies indicate the oc-
spin ½, indicating that 8 is monomeric. This is in agreement currence of hydrolytic processes. For the reduced SB 6, no
with the ν(V=O) obtained. Evans[17] also reported a room- hydrolysis could be detected in the whole pH range. The SB
temperature magnetic moment of 1.76 µB for 8 in aqueous sal2en also undergoes hydrolysis in strongly acidic DMSO/
solution at pH = 8.4.
water [80:20 (w/w)] solutions.[38] In contrast, the SB N,NЈ-
Complex 9 shows temperature dependence for the µeffval- o-phenylenebis(salicylideneimine) is stable under the same
ues. A smaller µeff value than the theoretical for S = ½ was conditions at pH Ͼ 2.5.[39]
obtained (see section ESI-D of the Supporting infor-
In order to study the hydrolytic processes of compound
mation). This low magnetic moment indicates that there are 3, several samples were prepared by dissolving the SB (SO3-
V=O···V=O interactions which is in agreement with the low sal)2en and either 1H NMR or pH-potentiometric data were
ν(V=O) measured for this complex (see above). Low µeff collected and evaluated with the computer programs PSE-
and ν(V=O) values were also found for several sal2entype QUAD[40] and SUPERQUAD,[41] respectively. Our objec-
SB
vanadium
complexes[35–37]
including
[VIVO- tive was to obtain a reasonably quantitative approach for
the species which form as the pH is varied, but not very
(Rpyr2en)].[15]
Table 4. A) Protonation and formation constants of species MpLqHr formed in the VIVO and VVO2– H2R(SO3-sal)2en systems calculated
from the pH-potentiometric data with the PSEQUAD program;[40] B) protonation and formation constants of species (en)p(SO3-sal)q(H+)r
[the main products formed upon dissolving (SO3-sal)2en in water] obtained from the potentiometric and 1H NMR spectroscopic data
(see Figure 3 and Scheme ESI-1 in section ESI-E of Supporting information)
A) R(SO3-sal)2en
log β[a]
B) (SO3-sal)2en
log β (1H NMR)[d]
pKa
Composition[c] (p,q,r)
log β (potentiometry)[e]
HL
10.70(9)
19.50(9)
27.03(9)
33.11(18)
10.70
8.80
7.53
6.08
H2L
H3L
H4L
(101)
(102)
9.90(7)
16.67(10)
9.99[42]
17.18[42]
+ [a]
log β (VIVO2+ [a]
)
log β (VVO2
)
(011)
6.72(1)
6.78(2)
MLH2
MLH
ML
26.88(21)
–
18.97(3)
9.50(12)
45.2(3)
39.3(3)
32.45(12)
–
(111)
(112)
(121)
(122)
12.21(3)
20.30(2)
12.80(15)
22.40(3)
12.02(4)
20.45(2)
MLH–1
22.42(8)
Fitting
No. of points
5.19 × 10–3[b]
290
1.07 × 10–2[b]
304
Fitting
No. of points
7.83 × 10–4[f]
80 (4×20)
9.57 × 10–3[g]
373
[a] In order to account for the range of the log βpqr values obtained by pH-potentiometry in the several equilibrium models tested, the
standard deviation S.D. is multiplied by 3 and the values are included in brackets. [b] Average of the difference between the experimental
and calculated titration curves, expressed as the volume (mL) of titrant. [c] Other stoichiometries (110, 113, 120, 123 and 124) were not
possible to fit. [d] Ambient temperature (ca. 25 °C), ionic strength not fixed. The S.D. values presented are those obtained in the particular
model shown. [e] T = 25 °C, I = 0.20 m KCl. [f] Average of the difference between the experimentally determined and the calculated
concentrations of en, SO3-sal, HSB and SB. Some figures showing the fitting between the calculated and experimental chemical shifts
are given in section ESI-E (Supporting information). [g] Average difference between the experimental and calculated titration curves,
expressed in pH. Some figures showing the fitting between the calculated and experimental pH-potentiometric curves are given in section
ESI-E (Supporting information).
Eur. J. Inorg. Chem. 2005, 732–744
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
737

J. C. Pessoa, T. Kiss et al.
FULL PAPER
1
accurate pK values. From the H NMR spectra, all chemi-
cal shift values were assigned and then set as constants
which were used to determine all the other protonation and
formation constants from the pH concentration dependence
for each species. The pH dependence of the chemical shifts
assigned to the peaks of the free ethylenediamine and alde-
hyde protons were also used to determine their pK values
with PSEQUAD[40] (see ESI-E). The potentiometric data
were treated separately.
The species were expressed as (en)p(SO3-sal)q(H+)r and
the protonation and formation constants of the main pro-
ducts formed were determined. Table 4B contains the pro-
tonation and formation constants, section ESI-E (Support-
ing information) contains information on the calculations
and Figure 3 shows a concentration distribution diagram
showing the total concentrations of each main compound.
Figure 4. Concentration distribution curves of complexes formed
in solutions containing: a) [VIVO]2+ and R(SO3-sal)2en with CVO
1.0 mM and L/M = 2; b) [VVO2]+ and R(SO3-sal)2en with CV
=
=
3.0 mM and L/M = 2, calculated using the stability constants listed
in Table 4; the dashed lines indicate that the curves only represent
an approximate estimation of the relative concentrations (see text)
Figure 5 shows the high-field region of the X-band EPR
spectra of “frozen” solutions containing [VIVO]2+ and 6 at
several different pH values. The EPR parameters were ob-
tained by computer simulation[30] of the EPR spectra. At
pH Ͻ 4 the only species present is the (aqua)vanadyl ion.
At pH Ͼ 4 two new species form. [VOLH2] only forms in
Figure 3. Concentration distribution curves of hydrolytic species
formed in solutions containing compound 3 (see Scheme ESI-1 of
Supporting information); only the total concentrations of each
compound [aldehyde = (SO3-sal); HSB = (SO3-sal)en, SB = [SO3-
sal]2en] are represented
We can conclude that for SB 3, the hydrolysis is very
extensive, much more than for pyr2en and occurs in the
whole measurable pH range. With pyr2en, hydrolysis was
only detected to a small extent at pH Ͼ 7.[15]
Vanadium(IV) Complexes
Although potentiometric and spectroscopic studies were
carried out with SB 3, its extensive hydrolysis did not allow
a consistent evaluation of the system. For the correspond-
ing reduced SB 6, the interaction with [VIVO]2+ and [VVO2]+
was studied and the corresponding stability constants are
also listed in Table 4. With [VIVO]2+ the pH-metric data
were evaluated only up to pH Ϸ 9.5 since slow processes
occur at higher pH values (oxidation and/or hydrolysis).
The species distribution diagram is presented in Figure 4a.
The complex formation starts at pH Ϸ 4 with the formation
of the neutral complex [VOLH2]. The ligand forms quite
stable 1:1 complexes with [VIVO]2+ and the stoichiometry
[VOL]2– predominates in the pH range 4.5–9.0. In the basic
pH range, the process [VOL]2– i [VOLH–1]3– + H+ has a
pK of ca. 9.5.
Figure 5. High-field region of the EPR spectra (77K) of frozen
solutions containing [VIVO]2+ and R(SO3-sal)2en at several pH val-
ues with L/M = 1 and CVO Ϸ 0.004 m; four main components can
be detected a: [VO(H2O)5], b: VOL, c: VOLH2 and d: VOLH–1
(charges omitted)
738
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 732–744

Vanadium(iv and v) Complexes of Schiff Bases and Reduced Schiff Bases
FULL PAPER
low concentration (ca. 10%) and its EPR parameters are gz [VOLH–1]. The corresponding gz and Az parameters are ca.
Ϸ 1.945 and Az Ϸ 168 × 10–4 cm–1. The additivity rule[31] 1.948 and ca. 162 × 10–4 cm–1, respectively. Possible equato-
applied to bidentate coordination of the ligand, i.e. through rial binding modes are by two Namine, an Ophenolate and an
estim
equatorial binding of Ophenolate, Namine and two H2O groups OH– with Az
yields Az
binding mode of VOLH2.
= 158 × 10–4 cm–1 and by two Ophenolate
estim
estim
= 170 × 10–4 cm–1, so this is probably the centres, an Namine and an OH– with Az
=
156 × 10–4 cm–1. The binding mode for [VOLH–1] probably
For VOL, the EPR parameters[30]gx, gy and gz are 1.982, corresponds to the first case.
1.976 and 1.957, respectively, and Ax, Ay and Az are 52, 50
The visible absorption spectra for this system confirm
and 159 × 10–4 cm–1. This indicates an equatorial binding the speciation model obtained by pH potentiometry. At pH
mode involving two Ophenolate and two Namine centres, = 3.1 the spectrum corresponds to the spectrum of[VI-
estim
Az
= 157 × 10–4 cm–1. In all the EPR spectra above pH VO(H2O)5]2+. In the pH range 5–7 the spectra do not
Ϸ 4 a second minor component is present which probably change and the d-d bands at λmax = 535 nm (ε = 112
corresponds to a binding mode of two equatorial Ophenolate
–1 m–1) and 840 nm (ε Ϸ 50 m–1 m–1) correspond to VOL.
centres, one equatorial Namine and one equatorial Owater Below 400 nm there is a charge-transfer band probably due
m
estim
centre as well as one axial Namine centre, Az
=
to transitions between Ophenolate and empty d orbitals on
164 × 10–4 cm–1. Therefore, equilibrium between cis/trans the V atom.
isomers apparently occurs as observed in the VIVO–
Vanadium(V) Complexes
Rpyr2en system.[15] However, in the present system the posi-
tion of this equilibrium is shifted to the trans isomer (which
The interaction between R(SO3-sal)2en and VV was
presumably contains the H2O ligand trans to the Ooxo studied by pH potentiometry and H and 51V NMR spec-
atom) and the EPR spectra could be satisfactorily simulated troscopy. The species distribution diagram is shown in Fig-
assuming some rhombic distortion from a square-pyrami- ure 4B and Table 4 contains the complex formation con-
1
dal structure.
stants. The pH-potentiometric titrations were carried out
Above pH Ϸ 9 a new signal could be observed in the between pH = 11.5 and 2 but due to the slow formation of
EPR spectra corresponding to the hydroxo complex a small percentage of decavanadate species below pH Ϸ 5,
Figure 6. NMR spectra at different pH values of D2O of solutions containing VV and R(SO3-sal)2en with CV = 20 mM and L:M Ϸ 1.1;
a) aromatic region (8.5 Ͻ δ Ͻ 6.0 ppm) of the 1H NMR spectra and b) 51V NMR spectra; C = C1 + C2; inset: concentration distribution
curves for the conditions used in the NMR experiments
Eur. J. Inorg. Chem. 2005, 732–744
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
739

J. C. Pessoa, T. Kiss et al.
FULL PAPER
the model that showed the best fitting was obtained by con- the value for species C1 (Figure 6b). Although the 51V
sidering only data at pH Ͼ 5.7. The complexes formed be- NMR spectra for the present system are not as clear as
tween VV and 6 are quite stable and correspond to a 1:1 for the VVO2–Rpyr2en system, they also display an α-/β-cis
stoichiometry in several protonation states. For an L/M ra- isomeric equilibrium: the 51V NMR signal for species C1
tio of 2 (and CV = 3 mm), some hydrolysis occurs only shows a shoulder at higher frequency (assigned to C2 in
above pH Ϸ 10 with the formation of [HVO4]2– and [VO4] Figure 6b) but the signal overlap due to the proximity of
3–
(see Figure 4B).
the chemical shifts makes quantification of the equilibrium
The 51V NMR spectra (Figure 6B) are quite complex, difficult. However, in the H NMR spectra (Figure 6a), the
showing signals corresponding to a variety of VV species presence of the equilibrium is particularly clear for the aro-
including monovanadate (V1), its oligomers (V2 and V4) matic proton signals, e.g.at pH = 7.4–9.5, where VO2L (C1
and VV–ligand complexes. Although in the potentiometric and C2) is the only stoichiometry present, three separate 1H
studies, decavanadate is formed only below pH Ϸ 5, in these NMR resonances can be observed (besides the one corre-
51V NMR studies weak decavanadate signals were detected sponding to the excess of free ligand in solution, L). One
even at pH Ϸ 9 (data not shown). Nevertheless, the spectra- corresponds to C1, the α-cis isomer, and the other two to
clearly show complex formation between ligand 6 and C2, the β-cis isomer. Thus, as in the VVO2–Rpyr2en sys-
[VO2]+, e.g. species A, B, C1and C2in Figure 6B. These reso- tem,[15] while in the α-cis complex the protons of each half-
nances are broad due to quadrupolar relaxation and the molecule are equivalent, in the β-cis complex they give rise
1
asymmetric coordination environment of the VV.[43]
to two separate signals. Deconvolution of the 1H NMR
The assignment of these 51V NMR signals was done tak- peaks gave a rough estimate of the ratio between the α-cis
ing into account the respective chemical shift values and and β-cis isomers of ca. 80:20 at pH = 7.8 which increases
peak areas (the peak areas were determined with program at lower pH. The very low percentage of the C2 isomer pre-
NUTS[44] after deconvolution of the signals) and by com- cludes observation of its 51V NMR resonance at pH Ͻ 7.4
parison with the speciation diagram of Figure 4b. The reso- (upon formation of complexes with other stoichiometries)
nance C1 (δ Ϸ –565 ppm), observed above pH Ϸ 5, corre- but at higher pH it is detectable. The small percentage of
sponds to the α-cis-VO2L complex (see Scheme 2) which the β-cis species might be due to repulsion between the
–
predominates between pH Ϸ 7 and 9, and the C2 signal (δ negative SO3 groups which in the β-cis complex are closer
1
Ϸ –562 ppm) of lower intensity appears clearly as a shoul- in space. At pH = 6.9 and lower, the broad H resonances
der on signal C1 above pH Ϸ 7. This can be assigned to the of the protonated B and A forms can be observed. Thus, in
β-cis-VO2L isomer. The 51V NMR shifts of these reso- the present case, slow exchange conditions on the NMR
nances are in agreement with a binding mode involving time scale prevail, not only between the two isomers of
[43]
N2O2
and with previous data obtained with the VVO2– [VO2L] and L but also with [VO2LH] and [VO2LH2] be-
Rpyr2en system.[15] A low-intensity, broad resonance B (δ cause, in this case, ligand protonation causes a change of
Ϸ –550 ppm) appears at pH Ϸ 6.4–6.9 and can be assigned VV coordination which was not the case for VVO2–Rpyr2en
to a VO2LH species where the monoprotonated ligand co- in which ligand protonation only resulted in H and 51V
1
ordination is probably tridentate. Several isomers may coex- chemical shift changes.[15]
ist since the ligand can coordinate through one or two of
the Namine and Ophenolate atoms, this being a plausible expla-
nation for the broadness of the peak. Resonance A (δ Ϸ –
530 ppm), with a maximum intensity at pH Ϸ 5.4, is practi-
Conclusions
cally absent at pH = 6.4. It should correspond to a VO2LH2
The SB ligands sal2en, (o-van)2en, (SO3-sal)2en and their
complex, in which the coordination of the diprotonated lig- reduced derivatives H2Rsal2en, H2R(o-van)2en and
and is probably bidentate. Again several isomers may coex- H2R(SO3-sal)2en were synthesised in good yields and the
ist. Between pH = 6.9 and 8.2 a weak broad signal can be structures of H2(o-van)2en and KH(o-van)2en·2H2O were
observed at δ Ϸ –535 ppm but its assignment is not straight- determined by X-ray diffraction. As expected, the Schiff
forward. It is probably an isomer of complex B since it ap- bases are more susceptible to hydrolysis than the corre-
pears in the same pH range (see Figure 6b). The 51V shift sponding reduced SB and 1H NMR studies in aqueous
trend observed as the complex is protonated is in agreement solutions of (SO3-sal)2en indicated the presence of several
with a decreasing ligand coordination number.[45,46]
hydrolysis products, namely the “half SB”, the free ethylene-
1
51V and H NMR studies for the VVO2–Rpyr2en system diamine and the free SO3-salicylaldehyde.
have shown the existence of equilibria in solution between
For the present set of ligands, there is strong IR and
structural isomers of the α- and β-cis-type (see magnetic susceptibility evidence that the Schiff base VIVO
Scheme 2).[15] While the α-cis isomer is symmetric, in the β- complexes form monomeric structures in the solid state and
cis isomer the protons of the two half-molecules are not that the corresponding reduced Schiff-base complexes form
equivalent, with peaks appearing at different chemical shifts structures with V=O···V=O-type interactions. The potentio-
for each half-molecule. Complex 11, with the molecular metric and spectroscopic studies demonstrate the high af-
structure depicted in Figure 2, corresponds to an α-cis iso- finity of the reduced SB R(SO3-sal)2en for [VIVO]2+ and
mer with VVO2L stoichiometry. In aqueous solution its 51V [VVO2]+. The ligand forms stable 1:1 complexes with vana-
NMR chemical shift is at δ Ϸ –563 ppm which is close to dium in both oxidation states and for the predominant stoi-
740
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 732–744

Vanadium(iv and v) Complexes of Schiff Bases and Reduced Schiff Bases
FULL PAPER
chiometries in solution, ML (M = [VIVO]2+ and [VVO2]+),
there is considerable evidence that there are two types of
isomers.
The X-ray diffraction characterisation of both
[VVO2(HRpyr2en)][15] and Na+[VVO2(Rpyr2en)]– (11) con-
firmed the formation of VVO2–Rpyr2en complexes with dif-
ferent degrees of protonation at the N-pyridine sites, thus
confirming some data from the earlier report for this sys-
tem.[15]
R(o-van)2en (5): The procedure was similar to the one used for 4.
Yield: 0.57g, 75%. C18H24N2O4·0.2MeOH (459.49): calcd. C 64.52,
H 7.38, N 8.27; found C 64.6, H 7.5, N 8.1. MS: m/z = 333 [M]+.
1H NMR (500 MHz, CDCl3 25 °C): δ = 2.82 (s, 4 H,–CH2CH2–),
3.87 (s, 6 H, CH3–O), 3.99 (s, 4 H, CH2NH), 6.7 (m, 6 H,
CHaromatic) ppm.
R(SO3-sal)2en (6): The procedure was similar to the one for 4 but
the solvent used was water. Yield: 0.77g, 88.0%.
C16H20N2O8S2·1.5H2O (459.49): calcd. C 41.82, H 5.05, N 6.10, S
13.95; found C 41.7, H 5.1, N 6.0, S 14.3. 1H NMR (500 MHz,
D2O, pH Ϸ 5.7, 25 °C): δ = 3.35 (s, 4 H, –CH2CH2–), 4.23 (s, 4 H,
CH2NH), 6.97 (d, 2 H, CHaromatic), 7.67 (s, 4 H, CHaromatic) ppm.
Complexes 8, 9 and [VIVO(Rpyr2en)] were tested for tox-
icity and insulin-mimetic effects in vitro through studies in-
volving mice fibroblasts.[16] When compared with other
compounds tested, [VIVO(Rpyr2en)] showed relatively low
toxicity and complex 8 showed high efficiency in promoting
glucose uptake into cells as well as high aqueous solubility.
All these characteristics might be important factors for the
therapeutic action of the vanadium complexes.
Rpyr2en (7): The compound was prepared according to a literature
procedure.[15] Yield: 0.83g, 60%. C18H26N4O4 (362.43): calcd. C
59.65, H 7.23, N 15.35; found C 59.5, H 7.3, N 15.4. 1H NMR
(500 MHz, D2O, pH Ϸ 4.8, 25 °C): δ = 2.41 (s, 6 H, CH3), 3.35 (s,
4 H, –CH2CH2–), 4.35 (s, 4 H, CH2NH), 4.63 (s, 4 H, CH2O), 7.62
(s, 2 H, CHaromatic) ppm.
Synthesis of the Vanadium Complexes: All reactions with the VIV
O
complexes were carried out under nitrogen.
Experimental Section
Cs2[VIVO{(SO3-sal)2en}] (8): The synthesis followed a literature
procedure.[17]
A pink complex was obtained. Yield: 93%.
Preparation of Ligands and Complexes
C16H12N2O9S2VCs2·0.5H2O {formulation: Cs2[VIVO{(SO3-sal)2-
en}](H2O)0.5} (766.16): calcd. C 25.08, H 1.71, N 3.66, S 8.37;
found C 25.2, H 2.0, N 3.6, S 8.4.
Synthesis of the Schiff Bases and Reduced Schiff Bases: The Schiff-
base ligands were prepared according to literature procedures[47]
by reaction of the o-hydroxyaldehyde with ethylenediamine (2:1) in
methanol, except for 3 in which the solvent was water. All com-
pounds were characterised by elemental analysis, mass spectrome-
[VIVO(Rsal2en)] (9): Compound 4 (0.27g, 1.0 mmol) was dissolved
in methanol/chloroform (10:1, 20 mL) containing sodium acetate
(0.30g, 2.2 mmol). An aqueous solution of VOSO4·5H2O (0.25g,
1.0 mmol) was slowly added and a brownish-orange precipitate im-
mediately formed. This was collected by filtration and washed with
methanol:water (8:2), methanol and diethyl ether and then dried
under vacuum. Yield: 0.30g, 90%. C16H18N2O3V (337.27): calcd. C
56.98, H 5.38, N 8.31; found C 56.8, H 5.3, N 8.4.
1
try and H NMR spectroscopy.
sal2en (1): Yield: 0.77g, 95%. C16H16N2O2 (268.32): calcd. C 71.62,
H 6.01, N 10.44; found C 71.6, H 6.0, N 10.5. MS: m/z = 268
[M]+. 1H NMR (500 MHz, CDCl3 25 °C): δ = 3.95 (s, 4 H,
–CH2CH2–), 7.2 (m, 8 H, CHaromatic), 8.36 (s, 2 H, CH = N) ppm.
(o-van)2en (2): Yield: 0.89g, 76%. C18H20N2O4 (328.37): calcd. C
65.84, H 6.14, N 8.53; found C 65.8, H 6.1, N 8.5. MS: m/z = 228
[VIVO{R(o-van)2en}] (10): The procedure was similar to the one
used for 9. Yield: 0.36g, 94%. C18H22N2O5V·2.1H2O (435.17)
{formulation: [VIVO{R(o-van)2en}](H2O)2.1}: calcd. C 49.68, H
6.07, N 6.44; found C 49.5, H 6.0, N 6.4.
1
[M]+. H NMR (500 MHz, CDCl3 25 °C): δ = 3.89 (s, 6 H, CH3–
O), 3.97 (s, 4 H, –CH2CH2–), 6.8 (m, 6 H, CHaromatic), 8.33 (s, 2
H, CH = N) ppm.
Both the filtrate and the washings were kept at room temperature
for several weeks and yellow crystals of two different types were
collected and characterised by X-ray diffraction. Crystals of 2A
and 2B were obtained from the washings and filtrates from the
synthesis of the Schiff base, respectively. The source of the K+ ion
was the KOH solution used to increase the pH from 4 to ca. 7.
Na[VVO2(Rpyr2en)]·CH3OH·3H2O (11): A mixture of NH4VO3
(0.03g, 0.27mmol) and Bu4NBr (0.09g, 0.27mmol) in methanol
(10mL) was heated to reflux for 1 h. The solution turned pale yel-
low and was filtered to remove the solid materials. After cooling,
Rpyr2en (0.10g, 0.28mmol) was added and the solution became
red. As there was no precipitation after 1 h of stirring, the solution
was maintained at room temperature for 2 weeks. Yellow crystals
were collected and characterised by X-ray diffraction.
(SO3-sal)2en (3): The Schiff base was prepared according to a lit-
erature procedure.[48] Yield: 1.2 g, 87%.C16H14N2Na2O8S2 (472.40):
calcd. C 40.68, H 2.99, N 5.93, S 13.57; found C 40.7, H 3.0, N
5.9, S 13.4.
Physical and Spectroscopic Studies: 51V and 1H NMR spectra were
obtained with a Varian Unity-500 Spectrometer operating at
131.404 MHz and 499.824 MHz, respectively, using a 5-mm broad-
band probe at 25.0 0.5 °C. The 1H and 51V NMR chemical shifts
were referenced relative to TSS [sodium 3-(trimethylsilyl)propane
sulphonate] at δ = 0 ppm, and to an external [VOCl3] solution at δ
= 0 ppm, respectively. The 51V NMR acquisition parameters were
as follows: 33 kHz spectral width, 30 µs pulse width, 1 s acquisition
time and 10 Hz line broadening. A presaturation pulse sequence
Rsal2en (4): To a solution of the SB 1 (0.61 g, 2.27mmol) in dichlo-
romethane (10 mL) at ca. 0 °C was added a methanolic solution of
NaBH4 (0.076g, 2.0 mmol, 5 mL containing some drops of concd.
KOH solution. The pH was set to 6 and the solution stirred for
several hours until the yellow colour had disappeared. The solvent
was evaporated to dryness and water (10 mL) was added to the
residue. The pH was set to 4–5 by addition of 3 m HCl. The white
solid was filtered, washed with cold water, ethanol and diethyl ether
and dried under vacuum. Yield: 0.38g, 61%. C16H20N2O2 (272.35):
calcd. C 70.56, H 7.40, N 10.29; found C 70.8, H 7.0, N 10.4. MS:
1
was used for H NMR spectra to eliminate the residual water sig-
nal. IR spectra were recorded with a BioRad FTS 3000 MX FTIR
spectrometer. Vis spectra were recorded either with a Hitachi U-
m/z = 273 [M]+. 1H NMR (500 MHz, CDCl3 25 °C): δ = 2.85 (s, 4 2000 or with a Perkin–Elmer Lambda 9 UV/Vis/NIR spectropho-
H,–CH2CH2–), 4.00 (s, 4 H, CH2NH), 7.0 (m, 8 H, CHaromatic
ppm.
)
tometer. The EPR spectra were recorded at 77 K (on glasses pre-
pared by freezing solutions in liquid nitrogen) with a Bruker ESP
Eur. J. Inorg. Chem. 2005, 732–744
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
741

J. C. Pessoa, T. Kiss et al.
FULL PAPER
Table 5. Crystal and structure refinement data
H2(o-van)2en (2A)
KH(o-van)2en·2H2O (2B)
Na+[VO2(Rpyr2en)]–·CH3OH·3H2O
(11)
Empirical formula
Formula mass
Temperature [K]
Wavelength [Å]
Crystal system
Space group
a [Å]
C18H20N2O4
328.36
293(2)
1.5418
monoclinic
Pc
9.409(2)
6.257(2)
14.436(3)
93.19(2)
848.6(3)
4
1.285
0.753
C18H19N2O6K[a]
398.45
293(2)
1.5418
orthorhombic
Pbca
C19H34N4NaO10V
552.43
293(2)
0.71073
monoclinic
P21/c
19.141(2)
7.3419(8)
19.070(2)
114.109(2)
2446.1(4)
4
6.498(3)
b [Å]
c [Å]
27.155(3)
22.158(3)
90
3910(2)
8
1.354
2.701
β [°]
V[Å3]
Z
ρcalcd. [M gm–3]
1.500
0.484
µ [mm–1]
F(000)
θ [°]
Limiting indices
348
1664
1160
4.71 to 59.98
–10 Յ h Յ 10;
–1 Յ k Յ 7;
–1 Յ l Յ 16
1687/1399 [R(int) = 0.0179]
3.25 to 64.91
–1 Յ h Յ 7;
–31 Յ k Յ 0;
0 Յ l Յ 26
3495/3322 [R(int) = 0.0516]
2.15 to 28.26
–25 Յ h Յ 23;
–6 Յ k Յ 9;
–25 Յ l Յ 25
15098/3823 [R(int) = 0.0395]
Measured/unique reflections
Refinement method
Full-matrix least-squares on Full-matrix least-squares on Full-matrix least-squares on F2
F2
F2
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I Ͼ 2σ(I)]
R indices (all data)
Absolute structure parameter
Largest difference peak/hole
[e Å–3]
1399/2/242
1.054
R1 = 0.0304; wR2 = 0.0839
R1 = 0.0306; wR2 = 0.0844
0.1(2)
3322/2/332
1.118
R1 = 0.0735; wR2 = 0.1743
R1 = 0.1049; wR2 = 0.2085
–
5859/8/348
1.029
R1 = 0.0488; wR2 = 0.1223
R1 = 0.0904; wR2 = 0.1447
–
0.099 and –0.112
0.626 and –0.813
0.742 and –0.663
[a]
The formula of 2B includes only the hydrogen atoms actually used in the refinement, i.e. H19. Hydrogen atoms of the water molecules
were not found or inserted in calculated positions. Taking into account all H atoms, the correct formula for the compound is
C18H23KN2O6.
300E X-band spectrometer. The magnetic susceptibilities were mea-
sured in the range 5–296 K using a 7 T F Oxford Instruments sys-
tem coupled to a Sartorius S3D-V microbalance.
H(4SB) which were located from a difference electron density map
and then fixed to 0.9 Å from the corresponding heteroatom [except
H(4O) which was left to refine freely]. Refinement converged with
allowance for thermal anisotropy of all non-hydrogen atoms. Crys-
tal data and details of the data collection and refinement are also
summarised in Table 5. CCDC-235821, -235822 and -236365 con-
tain the supplementary crystallographic data for this paper for 2A,
2B, and 11, respectively. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
X-ray Crystal Structure Determinations: For compounds 2A and
2B data were collected with a TURBOCAD4 Enraf–Nonius dif-
fractometer using Cu-Kα graphite-monochromated radiation. Data
were corrected for Lorentz and polarisation effects. No absorption
correction was performed. The crystal structures were solved by
direct methods (program SIR97[49]
)
and refined with
SHELXL97[50] using the WinGX package Version 1.64.03b.[51] All
non-hydrogen atoms were refined anisotropically. In compound 2B,
hydrogen atoms located in the Fourier maps were refined iso-
tropically. In compound 2A, the hydrogen atoms were included in
calculated positions and allowed to refine while riding on the par-
ent C atom except for the hydrogen atoms on C(1), C(2), C(3) and
C(4) which were located and refined isotropically. The poor quality
of the crystals of compound 2B precluded better refinement results.
Further details of the crystal structure determinations are given in
Table 5. Graphical representations were prepared using ORTEP[49]
and Mercury 1.1.2.[52] The X-ray data for compound 11 were col-
Speciation Studies
Spectroscopic Measurements: For preparation of the solutions and
pH calibrations we used a special glass vessel with a double wall
with entries for the combined electrode (Radiometer “Red Rod”
pHC2015-8), thermometer, nitrogen and reagents (e.g. base). A
computerised system developed locally was used to control the ti-
tration conditions for pH calibrations. The emf measurements were
made with a Denver Model 15 pH meter.
pH-Potentiometric Titrations: The purity of the ligands was
lected with a Siemens Smart 1000 CCD operating at 25 °C. Data checked pH-potentiometrically and the exact concentrations of the
were corrected for Lorentz and polarisation effects and for absorp-
tion by semiempirical methods. Structures were refined by full-ma-
trix least squares on F2. Complex scattering factors were taken
from the program package SHELXTL. Of the 15098 reflections
measured, 3823 independent reflections exceeded the significance
level |F|/σ(|F|) Ͼ 4.0. Hydrogen atoms were included in calculated
positions and were refined in the riding mode except the hydrogen
solutions were determined by the Gran method.[53] A stock solu-
tion of [VIVO] was prepared and standardised as reported ear-
lier[54,55] and its H+ concentration was determined by pH potent-
iometry. The vanadate stock solution was prepared by dissolving
KVO3 in a KOH solution of known molarity. The vanadium con-
tent in the commercial KVO3 was determined as reported earlier[15]
and the H+ concentration in the stock VV solution was calculated.
atoms H(2N), H(3N), H(2SA), H(2SB), H(4O), H(4SA) and All solutions were manipulated under high-purity nitrogen or puri-
742 © 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjic.org
Eur. J. Inorg. Chem. 2005, 732–744

Vanadium(iv and v) Complexes of Schiff Bases and Reduced Schiff Bases
FULL PAPER
fied argon. The ionic strength was adjusted to 0.20 m KCl and the
temperature was 25.0 0.1 °C. The pH was measured with an
Orion 710A precision digital pH meter equipped with an Orion
Ross 8103BN type combined glass electrode, calibrated for H+ con-
centration as described earlier.[56] The ionic product of water is pKw
= 13.76. Stability constants were determined by pH-metric titration
of 10.0-mL samples. The ligand concentrations were in the range
0.002–0.004 m and the metal ion/ligand molar ratio varied from 1:1
to 1:4. Titrations of solutions containing [VIVO]2+ were normally
carried out with KOH solution of known concentration (ca. 0.2 m)
under purified argon from pH = 2.0 up to 11.5 unless very extensive
hydrolysis, precipitation or slow equilibration was detected.
Titrations of solutions containing VV were carried out by back-
titrations i.e., by titrating samples from basic pH with HCl solu-
tions of known concentration (ca. 0.2 m). The reproducibility of
titration points included in the evaluation was within 0.005 pH
units in the whole pH range. The protonation constants and the
ligand purity were determined with the program SUPERQUAD.[41]
The concentration stability constants βpqr = [MpLqHr]/[M]p[L]q[H]r
were calculated using the PSEQUAD computer program.[40] The
formation of the hydroxo complexes of [VIVO] was taken into ac-
count. The following species were assumed: [VIVO(OH)]+ (logβ10–1
= –5.94), [(VIVO)2(OH)2]2+ (log β20–2 = –6.95), with stability con-
stants calculated from the data of Henry et al.[57] and corrected for
the different ionic strength using the Davis equation; [(VIVO)2-
(OH)5]– (log β20–5 = –22.5) and [VIVO(OH)3]– (log β10–3 =–18.0).[29]
These data can be used to establish the most probable binding
mode of the ligand in the complexes formed, but care must be
taken since the contributions of the donor groups to the hyperfine
coupling may depend on their orientation[60] or the charge of the
ligand.[61] The influence of the axial donor groups (if any) was not
taken into account.
1H and 51V NMR Spectroscopy: All NMR samples were prepared
at room temperature immediately before NMR spectroscopic deter-
minations. Ligand solutions for the NMR pH titrations were pre-
pared in D2O (99.995% D) by weighing the appropriate amount of
the ligand in order to have the desired concentration. The pD val-
ues of these solutions were adjusted with DCl and CO2-free NaOD
solutions and measured with a Crison MicropH 2002 pH-meter
with an Ingold 405-M5 combined electrode calibrated at 20 1 °C
with standard buffers at pH = 4.0 and 7.0. The final pH values
were corrected for the deuterium isotope effect using pH = pD –
0.4.[62] Solutions containing the VV complexes were prepared by
weighing the appropriate amount of ligand and dissolving it in a
H2O solution of the vanadate salt of known concentration (with
10% D2O) to obtain 51V NMR spectra, or in a D2O solution of
the VV salt when the samples were analysed by 1H NMR spec-
troscopy, in order to have the desired L/M ratios. Several sets of
experiments as a function of pH with different L/M ratios and total
concentrations were carried out. The signal intensities of the NMR
resonances were obtained using the program NUTS.[44] After base-
line flattening, an iterative deconvolution subroutine (LF) was used
to fit the peaks of the spectrum. This subroutine adjusts several
parameters, namely the frequency, height, width at half height and
area of the selected peaks. The areas obtained may then be used to
calculate, for example, the concentrations of the different species.
For the VV systems the stability constants were defined as βpqr
=
[MpLqHr]/[M]p[L]q[H]r, where M refers to [VO2]+ (not to [HVO4]–
as is usual in the literature). The speciation of vanadate into mono-
meric, dimeric, tetrameric, pentameric and decameric species was
taken into account and the values used for the stability constants
were those of Petterson[59] (after correction for the different ionic
strength using the Davis equation, and recalculation using [VO2]+
as M): [HVO4]2– (log β1,0,–3 = –14.93), [H2VO4]–(log β1,0,–2 = –6.76),
[V2O7]4– (log β2,0,–6 = –29.71), [HV2O7]3–(log β2,0,–5 = –19.37),
[H2V2O7]2– (log β2,0,–4 = –10.87), [V4O13]6– (log β4,0,–10 = –37.02),
[HV4O13]5– (log β4,0,–9 = –27.67), [V4O12]4– (log β4,0,–8 = –17.80),
[V5O15]5– (log β5,0,–10 = –22.63), [V10O28]6– (log β10,0,–16 = –17.32),
Supporting Information (see also footnote on the first page of this
article): Hydrogen bonding in H2(o-van)2en (2A), KH(o-van)2en·
2H2O (2B) and Na[VVO2(Rpyr2en)]·CH3OH·3H2O (11) (ESI-A),
Some IR data for compounds 1–12 (ESI-B), UV/Vis data for the
ligands and VIVO complexes prepared (ESI-C), Magnetic suscep-
tibility data for complexes 8 and 9 (ESI-D), Study of the hydrolytic
processes for SB 3 in aqueous solution (ESI-E)
[HV10O28]5– (log β10,0,–15 = –10.70), [H2V10O28]4– (log β10,0,–14
–6.53) and [H3V10O28]3– (log β10,0,–13 = –4.67).
=
Acknowledgments
Spectroscopic Measurements
The authors are grateful for the financial support of the Fundo
Europeu para o Desenvolvimento Regional, Fundação para a
Ciência e Tecnologia, the POCTI Programme (project POCTI/
35368/QUI/2000), the Hungarian National Research Fund (OTKA
T31896/2000), the Hungarian Academy of Sciences and the Hung-
arian-Portuguese Intergovernmental S & T Co-operation Pro-
gramme. F. A. thanks the Universidade da Coruña for the financial
support. This work was carried out within the frame of a COST
D21 project. We also thank Rui T. Henriques for the measurement
of the magnetic susceptibilities.
UV/Vis Spectroscopy: For compounds 1–10 the spectra were re-
corded by dissolving the ligands or complexes in the appropriate
solvent (see text). For the VIVO–R(SO3-sal)2en system, the spectra
were recorded upon changing the pH, with approximately fixed
total vanadium and ligand concentrations. Two sets of measure-
ments were made with an L/M ratio of 1 and with CVO Ϸ 4 and
12 mM (CVO = total metal ion concentration). The temperature
was kept at 25.0 0.3 °C with circulating water. In the figures we
represent εm values vs. λ, [where εm = absorption/(b·CVO), b = op-
tical path (either 0.1- or 1-cm cells were used)]. The spectral range
covered was normally 350–900 nm.
EPR Spectroscopy: For the frozen (77 K) aqueous solution spectra
in the absence of ethylene glycol, a relatively broad background
was obtained. Most spectra were therefore run with solutions con-
taining 5% ethylene glycol. The VIVO EPR spectra were simulated
using a program from Rockenbauer.[30] The EPR spectra may help
[1] K. H. Thompson, J. H. McNeill, C. Orvig, Chem. Rev. 1999,
99, 2561–2571.
[2] K. H. Thompson, C. Orvig, J. Chem. Soc., Dalton Trans. 2000,
2885–2892.
[3] C. Orvig, K. H. Thompson, Coord. Chem. Rev. 2001, 219–221,
1033–1053.
[4] H. Sakurai, Y. Kojima, Y. Yoshikawa, K. Kawabe, H. Yasui,
Coord. Chem. Rev. 2002, 226, 187–198.
[5] A. E. Evangelou, Crit. Rev. Oncol. Hematol. 2002, 42, 249–265.
[6] P. Köpf-Maier, Eur. J. Clin. Pharmacol. 1994, 47, 1–16.
[7] C. Djorgevitz, in: Metal ions in biological systems (Eds.: H.
Sigel, A. Sigel), Marcel Dekker, New York, 1995, p. 596.
to elucidate which groups coordinate in solution.[31] For the VIV
O
systems, we used the additivity rule to estimate the hyperfine coup-
estim
ling constant Az
(and gzestim) based on the contributions Az,i
estim
(and g||,i) of each of the four equatorial donor groups [e.g. Az
4
estim
=
Az,i]. The estimated accuracy of Az
is 3 × 10–4 cm–1.[31]
͚
i = 1
Eur. J. Inorg. Chem. 2005, 732–744
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
743

J. C. Pessoa, T. Kiss et al.
FULL PAPER
[8]
[9]
[31] N. D. Chasteen, in Biological Magnetic Resonance (Eds.: J.
Lawrence, L. J. Berliner, J. Reuben), Plenum, New York, 1981,
vol. 3, pp. 53–119.
[32] C. R. Cornman, K. M. Geiser-Bush, S. P. Rowley, P. D. Boyle,
Inorg. Chem. 1997, 36, 6401–6408.
[33] B. J. Hamstra, L. P. Houseman, G. J. Colpas, J. W. Kampf, R.
Lobrutto, W. D. Frasch, V. L. Pecoraro, Inorg. Chem. 1997, 36,
4866–4874.
S. J. D. Karlish, Y. Shechter, Nature 1980, 284, 556–558.
G. R. Dubyak, A. Kleinzeller, J. Biol. Chem. 1980, 255, 5306–
5312.
C. E. Heyliger, A. G. Tahiliani, J. H. McNeill, Science 1985,
227, 1474–1477.
J. Meyerovitch, Z. Farfel, J. Sack, Y. Schechter, J. Biol. Chem.
1987, 262, 6658–6662.
K. H. Thompson, C. Orvig, J. Chem. Soc., Dalton Trans. 2000,
2885–2892.
N. Durai, G. Saminathan, J. Clin. Biochem. Nutr. 1997, 22, 31–
39.
J. M. Lacoste, J. Duhault, D. Ravel, EUR Pat. Appl., 1993, EP
521,787.
I. Correia, J. Costa Pessoa, M. T. Duarte, R. T. Henriques,
M. F. M. Piedade, L. F. Veiros, T. Jackush, A. Dornyei, T. Kiss,
M. M. C. A. Castro, C. F. G. C. Geraldes, F. Avecilla, Chem.
Eur. J. 2004, 10, 2301–2317.
D. Rehder, J. Costa Pessoa, C. F. G. C. Geraldes, M. M. C. A.
Castro, T. A. Kabanos, T. Kiss, B. Meier, G. Micera, L. Petters-
son, M. Rangel, A. Salifoglou, I. Turel, D. Wang, J. Biol. Inorg.
Chem. 2002, 7, 384–396.
[10]
[11]
[12]
[13]
[14]
[15]
[34] P. W. Selwood, Magnetochemistry, Interscience, New York,
1956.
[35] K. S. Patel, G. A. Kolawole, J. Coord. Chem. 1982, 11, 231–
237.
[36] C. J. Carrano, C. M. Nunn, R. Quan, J. A. Bonadies, V. L. Pe-
coraro, Inorg. Chem. 1990, 29, 944–951.
[37] Y. Kuge, S. Yamada, Inorg. Chim. Acta 1977, 21, 85–88.
[38] H. Kumagai, M. Arishima, K. Susumu, K. Ymada, S. Kawata,
S. Kaizaki, Inorg. Chem. 2002, 41, 1989–1992.
[39] A. Mederos, Polyhedron 1997, 16, 4191–4196.
[40] L. Zekany, I. Nagypál, in Computational Methods for the De-
termination of Stability Constants (Ed.: D.Leggett), Plenum,
New York, 1985, pp. 291–353.
[41] P. Gans, J. Chem. Soc., Dalton Trans. 1985, 1195–1200.
[42] A. Gergely, I. Sóvágó, Inorg. Chim. Acta 1976, 20, 27–32.
[43] D. Rehder, Inorg. Chem. 1988, 27, 584–587.
[44] NUTS, NMR spectroscopic data processing software, Acorn
NMR spectroscopy, 1999.
[16]
[17]
[18]
D. F. Evans, P. H. Missen, J. Chem. Soc., Dalton Trans. 1987,
1279–1281.
Y. Mo, G. Yang, G. Zhao, B. Li, Jinlin Daxue Ziran Kex. Xue.
1990, 93–1; this is the journal where the structure was pub-
lished but it was impossible to obtain so we refer to this struc-
ture with the reference code of the Cambridge Structural Data-
base: VOJSUI.
[45] D. C. Crans, P. K. Shin, J. Am. Chem. Soc. 1994, 116, 1305–
1315.
[46] A. S. Tracey, H. Li, M. J. Gresser, Inorg. Chem. 1990, 29, 2267–
2271.
[19] N. P. Pahor, M. Calligaris, D. Nardin, L. Randaccio, Acta
Crystallogr. Sect. B 1978, 34, 1360–1363.
[20] A. Palmer, F. Brisse, Acta Crystallogr., Sect. B 1980, 36, 1447–
1452.
[47] S. N. Poddar, K. Day, S. C. N. Sarkar, J. Indian Chem. Soc.
1963, 40, 489–490.
[48] D. F. Evans, P. H. Missen, J. Chem. Soc., Dalton Trans. 1987,
1279–1281.
[21] D. Cunningham, K. Gilligan, M. Hannon, C. Kelly, P.
McArdle, A. O’Malley, Organometallics 2004, 23, 984–994.
[22] Since no cif file is presently available for this crystal structure,
neither from CCDC nor from the supplementary data, it was
not possible to access these data to make adequate compari-
sons. We may deduce that the structure is different since only
one half of the molecule is found in the asymmetric unit. The
molecule itself seems more planar and so the conformation as
well as the packing are different.
[23] K. Nakamoto, Infrared and Raman Spectra of Inorganic Com-
pounds, Part B, 5th ed., John Wiley & Sons, New York, 1997,
pp. 73–271.
[24] G. Socrates, Infrared and Raman Characteristic Group Fre-
quencies. Tables and Charts, John Wiley & Sons, Chichester,
2001, pp. 108–113.
[49] L. J. Farrugia, J. Appl. Crystallogr. 1997, 30, 565.
[50] G. M. Sheldrick, SHELXL-97, Programs for Crystal Structure
Analysis, release 97-2, University of Göttingen, Germany, 1997.
[51] L. J. Farrugia, J. Appl. Crystallogr. 2002, 32, 837–838.
[52] I. J. Bruno, J. C. Cole, P. R. Edgington, M. Kessler, C. F. Mac-
rae, P. McCabe, J. Pearson, R. Taylor, Acta Crystallogr., Sect.
B 2002, 58, 389–397.
[53] G. Gran, Acta Chem. Scand. 1950, 4, 559–575.
[54] J. Costa Pessoa, L. F. Vilas Boas, R. D. Gillard, R. Lancashire,
Polyhedron 1988, 7, 1245–1262.
[55] I. Nagypál, I. Fábián, Inorg. Chim. Acta 1982, 61, 109–113.
[56] J. Costa Pessoa, T. Gajda, R. D. Gillard, T. Kiss, S. M. Luz,
J. J. G. Moura, I. Tomaz, J. P. Telo, I. Torok, J. Chem. Soc.,
Dalton Trans. 1998, 3587–3600.
[57] R. P. Henry, P. C. H. Mitchell, J. E. Prue, J. Chem. Soc., Dalton
Trans. 1973, 1156–1159.
[25] B. Bosnich, J. Am. Chem. Soc. 1968, 90, 627–632.
[26] M. Gullotti, A. Pasini, P. Fantucci, R. Ugo, R. D. Gillard,
Gazz. Chim. Ital. 1972, 102, 853–892.
[58] A. Komura, M. Hayashi, H. Imanaga, Bull. Chem. Soc. Jpn.
1977, 50, 2927–2931.
[27] H. E. Smith, J. R. Neergaard, E. P. Burrows, F. M. Chen, J.
Am. Chem. Soc. 1974, 96, 2908–2916.
[28] K. S. Patel, G. A. Kolawolen, A. Earnshaw, J. Inorg. Nucl.
Chem. 1981, 43, 3107–3112.
[29] L. F. Vilas Boas, J. Costa Pessoa, Vanadium, in Comprehensive
Coordination Chemistry (Eds.: G. Wilkinson, R. D. Gillard,
J. A. McCleverty), Pergamon Press, Oxford, 1987, pp. 453–583.
[30] A. Rockenbauer, L. Korecz, Appl. Magn. Reson. 1996, 10, 29–
43.
[59] L. Pettersson, B. Hedman, L. Andersson, N. Ingri, Chem. Scr.
1983, 22, 254–264.
[60] T. S. Smith, C. A. Root, J. W. Kampf, P. G. Rasmussen, V. L.
Pecoraro, J. Am. Chem. Soc. 2000, 122, 767–775.
[61] A. J. Tasiopoulos, A. N. Troganis, A. E. Evangelou, C. P. Rap-
topoulou, A. Terzis, Y. Deligiannakis, T. A. Kabanos, Chem.
Eur. J. 1999, 5, 910–921.
[62] P. K. Glasoe, F. A. Long, J. Phys. Chem. 1960, 64, 188–190.
Received: June 7, 2004
744
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 732–744