Copper(II) thenoyltrifluoroacetonate as acceptor matrix in design of heterospin complexes

Article abstract of DOI:10.1016/S0277-5387(01)00596-4

Coordination compounds of Cu(II) thenoyltrifluoroacetonate with stable 2-imidazoline nitroxides were synthesized; their structure and magnetic properties were investigated. The complex with a pyrazole nitroxide was isolated as two polymorphous modifications. For both polymorphs, the structure is comprised of Cu(tta)₂(NITPz^1.3Me) molecules shaped as nearly regular pyramids with four O atoms of the two tta ligands lying at the base. The solid polymorphs are constructed from formal binuclear fragments appearing when the pyramid bases of two neighboring molecules approach each other forming rather short contacts Cu - S (3.848 and 3.596 ?), due to which the polymorphs differ considerably in magnetic properties. In the modification with longer Cu - S distances, there are no exchange interactions between the paramagnetic centers of the neighboring Cu(tta)₂(NITPz^1.3Me) molecules whereas in the polymorph with shorter Cu - S distances the interactions are significant and analogous to those in real dimer molecules based on the Cu(AcO)₂ matrix with a nitroxide.

Full text of DOI:10.1016/S0277-5387(01)00596-4

Polyhedron 20 (2001) 1215–1222
www.elsevier.nl/locate/poly
Copper(II) thenoyltrifluoroacetonate as acceptor matrix in design
of heterospin complexes
Ilia Yeltsov b, Victor Ovcharenko a,*, Vladimir Ikorskii a, Galina Romanenko a,
Sergei Vasilevsky a
a International Tomography Center, Russian Academy of Sciences, Institutskaya Str. 3A, 630090 No6osibirsk, Russia
b No6osibirsk State Uni6ersity, Pirogo6a Str. 2, 630090 No6osibirsk, Russia
Received 17 September 2000; accepted 15 October 2000
Abstract
Coordination compounds of Cu(II) thenoyltrifluoroacetonate with stable 2-imidazoline nitroxides were synthesized; their
structure and magnetic properties were investigated. The complex with a pyrazole nitroxide was isolated as two polymorphous
modifications. For both polymorphs, the structure is comprised of Cu(tta)2(NITPz1,3Me) molecules shaped as nearly regular
pyramids with four O atoms of the two tta ligands lying at the base. The solid polymorphs are constructed from formal binuclear
fragments appearing when the pyramid bases of two neighboring molecules approach each other forming rather short contacts
,
Cu–S (3.848 and 3.596 A), due to which the polymorphs differ considerably in magnetic properties. In the modification with
longer Cu–S distances, there are no exchange interactions between the paramagnetic centers of the neighboring
Cu(tta)2(NITPz1,3Me) molecules whereas in the polymorph with shorter Cu–S distances the interactions are significant and
analogous to those in real dimer molecules based on the Cu(AcO)2 matrix with a nitroxide. © 2001 Elsevier Science Ltd. All rights
reserved.
Keywords: Copper(II); Nitroxide; Dimers; Polymorphous modifications
1. Introduction
chosen as an acceptor matrix and the widely used
NITpPy and NITmPy as well as the pyrazole derivative
NITPz1,3Me, structurally related to NITmPy, were cho-
sen as nitronyl nitroxides. Since Cu(tta)2 and Cu(hfac)2
exhibit similarity in their accepting abilities and in the
Metal hexafluoroacetylacetonates are widely used as
acceptor matrices for syntheses of heterospin systems
[1] (Scheme 1). Due to their stereochemical nonrigidity,
the matrices form heterospin complexes with polyfunc-
tional nitroxides varying in composition and structure
and showing the ability for magnetic ordering [2–6].
For the Cu(hfac)2 complex with nitronyl nitroxide con-
taining a pyridine substituent in the 2-position of the
imidazoline heterocycle (NITmPy), a new type of ther-
mally induced spin transition was found [7]. We at-
tempted to expand the scope of Cu-containing matrices
which are capable of forming heterospin complexes
with nitroxides (Schemes 1 and 2). Copper(II) thenoyl-
trifluoroacetonate, Cu(tta)2, related to Cu(hfac)2, was
Scheme 1.
* Corresponding author. Fax: +7-3832-331399.
E-mail address: ovchar@tomo.nsc.ru (V. Ovcharenko).
Scheme 2.
0277-5387/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0277-5387(01)00596-4

1216
I. Yeltso6 et al. / Polyhedron 20 (2001) 1215–1222
structure of coordination site, Cu(tta)2 was expected to
behave analogously to Cu(hfac)2 with nitroxides. It
appeared, however, that Cu(tta)2, which is stereochemi-
cally more rigid, reacts with pyridine-containing
nitronyl nitroxides typically forming much fewer types
of heterospin complex compared to Cu(hfac)2. In reac-
tion with a pyrazole-substituted nitronyl nitroxide,
Cu(tta)2 gave a heterospin compound with an unknown
structure whose magnetic behavior resembles that of
the binuclear complex [Cu(AcO)2]2(NITpPy)2. Here we
report on syntheses, structure, and magnetic properties
of all of the isolated Cu(tta)2 complexes with the above
nitronyl nitroxides.
Yield 200 mg (60%). Tm=90–92°C. IR spectrum
(cm−1): 3097; 2994; 2943; 1621; 1536; 1413; 1399; 1371;
1287; 1240; 1142; 779; 696. Anal. Found: C, 52.4; H,
4.7; N, 7.9; F, 11.0; S, 5.9. Calc. for CuC47H48N6O8
F6S2: C, 52.9; H, 4.5; N, 7.9; F, 10.7; S, 6.0%.
2.2.3. Cu(tta)2(NITPz1,3Me) (III and IV)
A mixture of Cu(tta)2 (100 mg, 0.2 mmol) and 50 mg
(0.2 mmol) NITPz1,3Me was dissolved in a (10/3) hex-
ane–ethylacetate mixture with stirring and heating to
50°C for 1 h. The solution was then cooled to r.t. and
allowed to stay for 2 days. Dichroic (blue–green) single
crystals (denoted by IV) shaped as long needles formed
in the solution. The mother solution was decanted and
allowed to stay, and crystals IV (25 mg) were collected
and dried on a paper filter. Rhomboid dark blue crys-
tals (III) (90 mg) precipitated from the mother solution
which was stored for 3 days more at r.t. Note that III
was the major product in all syntheses. In some cases,
IV was not detected at all, and the whole solid con-
sisted of crystals III. Solids III and IV have the same
composition since they are polymorphous modifications
of Cu(tta)2(NITPz1,3Me). Therefore, the results of analy-
sis are given for only one modification (III). Anal.
Found: C, 45.4; H, 3.8; N, 7.51; F, 15.8; S, 8.3. Calc.
for CuC28H27N4O6F6S2: C, 44.4; H, 3.6; N, 7.40; F,
15.1; S, 8.5%. For III: Tm=167–168°C; IR spectrum
(cm−1): 3086; 2991; 1593; 1543; 1411; 1360; 1316; 1233;
1190; 1147; 941; 786; 594; 540; 466. For IV: Tm=
(139–144°C, color change) 159–160°C, melting with
decomposition; IR spectrum (cm−1): 3439; 3062; 2996;
1613; 1586; 1542; 1413; 1354; 1316; 1232; 1198; 1178;
1149; 1067; 940; 787; 592; 540; 465.
2. Experimental
2.1. Nitronyl nitroxides
The nitronyl nitroxides used in this work were syn-
thesized by known procedures [8,9]. 2,3-Dihydroxy-
lamino-2,3-dimethylbutane sulfate needed for syntheses
of nitroxides was prepared according to [10,11].
2.2. Syntheses of complexes
2.2.1. Cu(tta)2(NITpPy)·0.5C6H6 (I)
Cu(tta)2 (216 mg, 0.43 mmol) was dissolved in ethy-
lacetate (5 ml). To the resulting solution was added a
solution of NITpPy (100 mg, 0.43 mmol) in benzene
(10 ml) and the mixture was allowed to stay in an open
container at room temperature (r.t.). The solution evap-
orated to half of its volume within three days, where-
upon a green precipitate settled out which was filtered
off and washed with a cool hexane–benzene mixture.
The product was recrystallized from benzene from
which it was separated as single crystals suitable for an
X-ray diffraction analysis. Yield 255 mg (77%). Tm=
166–167°C. IR spectrum (cm−1): 3436; 3106; 1593;
1532; 1410; 1316; 1144; 787; 730. Anal. Found: C, 46.6;
H, 3.3; N, 5.4; F, 15.3. Calc. for CuC31H35N3O6F6S2: C,
45.5; H, 3.3; N, 5.7; F, 15.4%. Using other starting
ratios of Cu(tta)2/(NITpPy) in the reaction mixture also
led to crystallization of I.
2.2.4. [Cu(AcO)2]2(NITpPy)2 (V)
Cu(AcO)2·H2O (85 mg, 0.43 mmol) was dissolved
with prolonged stirring at r.t. in methanol (15 ml).
Then NITpPy (100 mg, 0.44 mmol) was added to the
reaction mixture while stirring was continued. A finely
dispersed green precipitate started to settle from the
reaction mixture within a few moments. Stirring was
continued for further 3 h. The precipitate was filtered
off, washed with toluene to remove the excess of radical
(until the solvent ceased to show color), and dried in
air. Yield 120 mg (67%). Tm=218–219°C. IR spec-
trum (cm−1): 2991; 2942; 1623; 1427; 1312; 1215; 1167;
1066; 1016; 834; 681; 626; 542; 460. Anal. Found: C,
46.5; H, 5.3; N, 10.0. Calc. for CuC16H22N3O6: C, 46.2;
H, 5.3; N, 10.1%. To grow perfect single crystals of
Cu2(AcO)4(NITpPy)2, Cu(AcO)2·H2O was dissolved in
boiling methanol, the solution was then cooled to 30°C,
and an equal amount of NITpPy was added. Toluene
was added to the resulting dark blue solution until the
methanol/toluene ratio became 2/1, and the solution
was allowed to stay in an open flask at r.t. for 5 days.
2.2.2. Cu(tta)2(NITmPy)2·C7H8 (II)
Toluene (10 ml) was poured into a mixture of
Cu(tta)2 (216 mg, 0.43 mmol) and NITmPy (100 mg,
0.43 mmol). The mixture was stirred at r.t. for 2 h. The
resulting solution was filtered, heptane (5 ml) was
added to the filtrate, and the solution was allowed to
stay at r.t. After 4 days, concretions of greenish blue
crystals formed on the walls of the container, which
were separated and washed with heptane. Slow evapo-
ration of the mother solution gave more product as
single crystals suitable for an X-ray diffraction analysis.

I. Yeltso6 et al. / Polyhedron 20 (2001) 1215–1222
1217
Table 1
I
II
III
IV
V
Structural
motif
Chain ‘head-to-head’
Molecular
‘Dimer-like’
‘Dimer-like’
Dimer
Cu–ONIT
Cu–NNIT
Cu–Otta
2.458(2)
2.035(2)
1.924(1), 1.931(2),
2.020(2), 2.170(2)
1.261(2), 1.278(2)
2.033(2)
1.973(2),
2.262(2)
1.282(3),
1.277(3)
22.0
2.428(3)
2.366(12)
2.144(5)
1.956(5), 1.974(5), 1.966(4),
1.954(4)
1.921(3), 1.933(3),
1.925(3), 1.939(3)
1.281(6), 1.275(6)
1.924(8), 1.934(9),
1.943(8), 1.932(8)
1.268(15), 1.255(15)
N–O
1.293(6), 1.267(6)
ÚPy(Pz)–CN2 16.7
37.3
39.7
24.2
Fig. 1. Structure of the chain in I (benzene molecules and hydrogen atoms omitted for clarity).
The solution evaporated to half of its volume, and
perfectly shaped bright green plates of the complex
gradually grew in it. The plates were filtered off,
washed with toluene, and dried in air.
gen atoms. H atoms were partly located on difference
electron density maps (the others were calculated geo-
metrically) and refined isotropically. The crystal data
needed for further discussion are given in Table 1.
2.3. Magnetic measurements
3. Results and discussion
Magnetic susceptibilities of the complexes were mea-
sured on an MPMS-5S (Quantum design) SQUID mag-
netometer at temperatures of 2–300 K and in magnetic
fields of up to 10 kOe. The calculated temperature-de-
pendent paramagnetic susceptibility was corrected for
the diamagnetic term, which was 372×10−6 cm3
mol−1 for III, IV and 191×10−6 cm3 mol−1 for
[Cu(AcO)2]2(NITpPy)2, and for the temperature-inde-
pendent paramagnetic contribution of the Cu(II) ion,
which was 60×10−6 cm3 mol−1. The effective mag-
netic moment was calculated by the formula veff=(3k/
Our study of the reaction products of Cu(tta)2 with
NITpPy and NITmPy led to compounds of composi-
tion
Cu(tta)2(NITpPy)·0.5C6H6
and
Cu(tta)2-
(NITmPy)2·C7H8. No complexes with other stoi-
chiometries have been detected. Although NITpPy and
NITmPy are functionally related, in reaction with
Cu(tta)2 they form complexes which have essentially
different structures in the solid state. The two crystallo-
graphically independent Cu atoms in the structure of I
have a centrosymmetric distorted octahedral environ-
ment (Fig. 1). The Otta atoms lie in the equatorial plane
of a pulled tetragonally distorted Cu1 octahedron (Cu–
Ni2 · T)1/2
.
,
2.4. X-ray crystallography
Otta distances are 1.924(1) and 1.931(2) A); the nitroxyl
O atoms of NITpPy are in the axial positions forming
,
Crystals I–V are triclinic. The data were collected on
ENRAF NONIUS CAD4 and BRUKER AXS P4
automatic diffractometers using standard methods (Mo
Ka radiation, graphite monochromator, q/2q scan
mode, variable scan speed). The structures were solved
with SIR-97 and refined anisotropically for all nonhydro-
Cu–ONIT distances of 2.458(2) A and a Cu–O–N angle
of 146.1(1)°. For the Cu2 atom, the equatorial plane of
a nearly isometric octahedron is formed by two N
atoms of the pyridine heterocycle (Cu–NNIT 2.035(2)
,
,
A) and two Otta atoms (Cu–O 2.020(2) A), the other
two Otta atoms occupying the axial positions (Cu–O

1218
I. Yeltso6 et al. / Polyhedron 20 (2001) 1215–1222
Fig. 2. Structure of II.
,
On the −1(T) curve (Fig. 3), one can isolate two
linear segments corresponding to different parameters
of the Curie–Weiss equation (T)=C/(T−q):
2.170(2) A) (Table 1). The N–O distance in the coordi-
,
nated fragment is slightly shorter (1.261(2) A) than that
,
in the noncoordinated N–O group (1.278(2) A). The tta
fragments, which are trans to each other, are nearly
planar, the angle between the thenoyl heterocycle and
the respective metallocycle being up to 7.6°. The angle
between the planes of the Py ring and the CN2 frag-
ment of the imidazoline heterocycle is 16.7°. The para-
magnetic NITpPy performs the bridging function,
linking the Cu(tta)2 fragments into infinite zigzag chains
Scheme 3.
(
having a ‘head-to-head’ motif and running along [101].
The space between the chains is occupied by the ben-
zene molecules.
The structure of solid II is much simpler. It is built
from discrete Cu(tta)2(NITmPy)2 fragments alternating
with the toluene molecules (Fig. 2). The centrosymmet-
ric distorted octahedral environment of the copper
atom is formed in the equatorial plane by the N atoms
,
of the pyridine heterocycles (Cu–NNIT 2.033(2) A) and
,
by two Otta atoms (Cu–O 1.973(2) A). The axial posi-
tions are occupied by the other two Otta atoms (Cu–O
,
2.262(2) A).
Analyzing the magnetic properties of I, we correlated
the polymer chain to the following scheme of exchange
channels J1(Cu–O·–Nꢀ) and J2(Cu–···ꢁN–·O)
(Scheme 3).
Fig. 3. Dependence −1(T) for I.

I. Yeltso6 et al. / Polyhedron 20 (2001) 1215–1222
1219
and the coordinated nitroxyl oxygen (Table 1). The
exchange interactions by the J2 channel, i.e. through the
pyridine heterocycle are much more pronounced.
The possibility of strong exchange interactions of
odd electrons through the pyridine heterocycle is sup-
ported by the results of studies on molecular complex II
in which the shortest intermolecular contacts between
,
the O atoms of the O–Nꢀ fragments are up to 4.4 A.
Therefore, the crystal structure of II may be regarded
as comprised of quasiisolated exchange clusters O·–
Nꢀ···Cu2+···ꢁN–·O with indirect interactions via the
N atom of the pyridine ring. Theoretical treatment of
the experimental dependence veff(T) for II (Fig. 4) was
carried out on a three-center heterospin exchange clus-
ter model [13]. The energy levels of the cluster in an
external magnetic field were calculated using the
isotropic spin Hamiltonian:
Fig. 4. Theoretical (solid line) and experimental temperature depen-
dences of veff for II.
.
.
.
.
.
H= −2Jsˆsˆ%−i(gCuSz+gLSz)H−2J%zSzŽSz
where J and J%z are the intra and intercluster exchange
parameters, sˆ is the Cu2+ spin operator, sˆ=sˆ1+sˆ2 is
the operator of the total spin of two radicals, g and g%
are the g factors of Cu2+ and nitroxide, z is the number
of the nearest-neighbor molecules, i the Bohr magne-
.
ton, S=sˆ+sˆ% is the operator of the total spin of the
.
cluster, and ŽSz is its averaged projection onto the z
axis. The optimal J, J%z, and gCu (gLꢀ2) values were
(Ti)]2.
calc
eff
measd
eff
obtained by minimization of ꢀi[v (Ti)−v
The optimization gave the following spin Hamiltonian
parameters: g=2.0090.01, J=6.190.01 cm−1, J%z=
−0.3190.05 cm−1, indicative of appreciable ferro-
magnetic exchange interactions in the cluster. The
opposite sign of the exchange parameter (J for II and J2
for I) is in good agreement with spin polarization
effects.
Although NITPz1,3-Me is related to NITmPy in the
sequence of donor atoms in the paramagnetic ligand, in
reactions with Cu(tta)2 at different reagent ratios it
forms only a 1:1 complex Cu(tta)2(NITPz1,3Me). This
compound was isolated as two polymorphous modifica-
tions (III and IV). In both structures, the environment
of the Cu atom is a square pyramid with four O atoms
of two tta ligands at the base (Cu–Otta 1.921(3)–
Fig. 5. View of molecule III.
,
1.943(8) A) and the pyrazole N atom at the apex (Figs.
TB4 K, C=0.703, q=0.77 K
T\10 K, C=0.785, u= −3.3 K
5 and 6). Also note that the tta ligands are cis to each
other. The CuꢂN distances differ widely: 2.428(3) in III
,
and 2.366(12) A in IV. The greatest differences between
The exchange parameters were qualitatively estimated
by the formula q=2zJS(S+1)/3k [12] in a molecular
field approximation: J1=0.54 cm−1 and J2= −2.3
cm−1. The J1 parameter may be assigned to the head-
to-head fragment ꢁN–·O–Cu–O·–Nꢀ since ferro-
magnetic exchange is characteristic for Cu2+ complexes
with an axial coordination of the nitroxyl group. The
low value of ferromagnetic exchange may be the conse-
quence of the long distance between the copper atom
the molecular geometries of III and IV are observed in
the values of torsion angles (Table 2).
The Cu(tta)2NITPz1,3Me molecules are linked into
dimer-like fragments by weak Cu···S contacts, increas-
ing the coordination number of copper to 6. The differ-
ences in molecular geometry cause significant
differences in intermolecular contacts: Cu···S 3.848 and
,
,
3.596 A, N1-O1···O1%–N1% 4.372 and 3.454 A, N2–
O2···O1%–N1% 4.962 and 4.631 A, Cu···Cu 5.547 and
,

1220
I. Yeltso6 et al. / Polyhedron 20 (2001) 1215–1222
,
6.805 A for III and IV, respectively, and in the mag-
netic behavior of these modifications.
The absence of a temperature dependence for veff in
the range 2–300 K for III (Fig. 7) points to very weak
indirect exchange interactions between the odd elec-
trons of the copper ion and the nitroxyl group occur-
ring via the pyrazole heterocycle of a separate fragment
of Cu(tta)2(NITPz1,3Me) in agreement with considerably
longer Cu–NNIT distance compared to that in the
above complexes with pyridine-substituted nitroxides
(Table 1). The intermolecular exchange interactions in
solid III are not observed below 2 K. Magnetic suscep-
tibility of III is well approximated by the Curie equa-
Fig. 7. Dependences veff(T) for III–V. Solid lines: theoretical depen-
dences described in the text. Insert: Dependence (T) for V.
tion =Ni2ve2ff/3kT=C/T with a constant C=0.83.
In IV having much shorter Cu–S and Cu–Cu dis-
tances in the dimer-like fragment {(NITPz1,3Me)-
(tta)2Cu···Cu(tta)2(NITPz1,3Me)} compared to III, the
exchange interactions between the paramagnetic centers
are much more effective. The experimental curve veff(T)
for IV is formally divided into three sections. At low
temperatures, the effective magnetic moment slightly
increases with temperature. The region where the de-
pendence veff(T) is absent and veffꢁ1.8 BM is followed
by the section where veff increases with temperature up
to RT. An analysis of structural data and veff(T) makes
it possible to represent the experimental veff(T) curve as
the sum of two major exchange interaction terms, dif-
fering widely in magnitude. In the high-temperature
region, the curve is well approximated by the model of
the Cu(II)–Cu(II) dimer whose magnetic susceptibility
is defined by the Bleaney–Bowers equation:
−1
Ni2g2Cu
3kT
1
3
−2J
kT
ꢁ ꢂ ꢃn

=
1+ exp
dim
and at low temperatures depends on the weak exchange
interactions between the radicals J1 (J1ꢂJ). Magnetic
susceptibility in this region obeys the Curie-Weiss law
with a low Weiss constant =C/(T−q). The results of
the approximation of the experimental dependence by
the sum:
Fig. 6. View of molecule IV.
Table 2
Selected torsion angles
Atoms
Torsion angles (°)
=dim+C/(T−q)
III
IV
with the optimal parameters gCu=2.0, J= −45 cm−1
,
S1C17C16O4
S2C25C24O6
O4CuN3N4
O6CuN3N4
C12C7C7N2
C12C7C7N1
−4.1
7.2
29.6
−58.9
−141.8
39.0
9.1
−3.2
57.9
−28.6
34.8
−138.7
C=0.906, and q= −0.5 are presented in Fig. 7. The
value of veffꢁ1.8 BM in the ‘intermediate’ region
indicates that at TB20 K complete spin coupling takes
place in the exchange-coupled dimer Cu–Cu and that
the major contribution to veff is from the paramagnetic
centers of nitroxides.

I. Yeltso6 et al. / Polyhedron 20 (2001) 1215–1222
1221
The form of the veff(T) curve for V suggests that
there are stronger antiferromagnetic exchange interac-
tions in the Cu–Cu dimer compared to IV. The mag-
netic susceptibility (T) curve for V, however, is
distinctive at low temperatures, having a broad maxi-
mum (Fig. 7, insert) indicative of antiferromagnetic
ordering typical for a polymer chain. Using the crystal
data for V, one can represent the exchange interactions
between the paramagnetic centers in this complex as
shown in Scheme 4.
The above magnetic data for I, III, and IV are used
to make estimates JꢃJ2\J1, which are readily evi-
dent. In the framework of this exchange coupling
scheme, the dependence veff(T) for V is exclusively
dominated by the J2 interactions between the radicals
of different molecules in the chains at low temperatures
and by the exchange interactions between the Cu(II)
ions via the carboxylate bridges in the dimers at high
temperatures. Therefore, magnetic susceptibility of V is
represented as the sum of the magnetic susceptibilities
of the exchange-coupled Cu–Cu dimer and the ex-
change chain [14,15]:
In order to justify the use of the Cu–Cu dimer model
in treatment of the magnetic properties of IV, we
synthesized the authentic dimer V and investigated its
magnetic properties. This molecule (Fig. 8) is really a
dimer which is centrosymmetric relative to the middle
of the Cu–Cu bond.
The square bipyramid around the Cu atom consists
of four oxygen atoms of the acetate ligands in the
equatorial plane and the second Cu atom of the dimer
and the pyridine N atom in the axial positions. The
Cu–O, Cu–Cu, and Cu–N distances are 1.54(4)–
,
1.974(5), 2.617(2), and 2.144(5) A, respectively. The
bond lengths and angles in the bridging acetate ligands
agree with the values cited in the literature: C–O
,
1.244(7)–1.253(7), C–C 1.503(11)–1.532(1) A; ÚOCO
126.4(7) and 127.3(6)°. In the structure of nitronyl
nitroxide fragments, the N–O bond lengths differ con-
siderably (Table 1). The plane of the Py ring makes an
angle of 82.1° with the equatorial plane of the Cu
bipyramid and an angle of 24.2° with the N(1)C(3)N(2)
plane of the imidazoline heterocycle. The structure of V
in general is considered molecular. It has rather short
contacts between the O atoms of the N–O groups of
=chain+
dim
,
the neighboring dimer molecules (3.727 and 4.042 A).
where
This was taken into consideration in treatment of the
experimental dependence veff(T) for V (Fig. 7).
−1
Ni2g2Cu
3kT
1
3
−2J
kT
ꢁ ꢂ ꢃn

=
1+ exp
dim
Ni2g2R
0.25+Bx+Cx2
1+Dx+Ex2+Fx3

chain
=
·
kT
where x=ꢄJ2ꢄ/kT. This coupling scheme nicely approxi-
mates the experimental dependence (Fig. 7); the optimal
theoretical curve is characterized by the following
parameters: gR=2.07, J2= −2 cm−1; gCu=2.18, J=
−181 cm−1
.
4. Supplementary material
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC no. 149119 for complex I, CCDC
no. 149076 for complex II, CCDC no. 149077 for
complex III, CCDC no. 149078 for complex IV and
CCDC no. 149080 for complex V. Copies of this infor-
mation may be obtained free of charge from the Direc-
tor, CCDC, 12 Union Road, Cambridge, CB2 1EZ,
UK (fax: +44-1223-336033; e-mail: deposit@ccdc.cam.
ac.uk or http://www.ccdc.cam.ac.uk).
Fig. 8. Structure of dimer molecule V.
Acknowledgements
Research funded in part by US Civilian and Devel-
opment Foundation (grant REC-008), RFBR (grant
00-03-32987), and Integration Program.
Scheme 4.

1222
I. Yeltso6 et al. / Polyhedron 20 (2001) 1215–1222
[8] S. Vasilevsky, E. Tretyakov, O. Usov, Yu. Molin, S. Fokin, Yu.
References
Shwedenkov, V. Ikorskii, G. Romanenko, R. Sagdeev, V.
Ovcharenko, Mendeleev Commun. (1998) 216.
[9] E. Ulman, J. Osiecki, D.G.B. Boocock, R. Darcy, J. Am. Chem.
Soc. 94 (1972) 7049.
[10] V. Ovcharenko, S. Fokin, P. Rey, Mol. Cryst. Liq. Cryst. (A)
333 (1999) 109.
[11] V.I. Ovcharenko, S.V. Fokin, G.V. Romanenko, I.V. Korobkov,
P. Rey, Izv. Akad. Nauk. Ser. Khim. 108 (1999) 1539.
[12] J.S. Smart, Effective Fields Theories of Magnetism,
W.B.Saunders Company, Philadelphia, London, 1966.
[13] A.B. Burdukov, V.I. Ovcharenko, V.N. Ikorskii, N.V. Per-
vukhina, N.V. Podberezskaya, I.A. Grigor’ev, S.V. Larionov,
L.B. Volodarsky, Inorg. Chem. 30 (1991) 972.
[14] B. Bleaney, K.D. Bowers, Proc. R. Soc. A214 (1952)
451.
[15] Yu.V. Rakitin, V.T. Kalinnikov, Modern Magnetochemistry,
Nauka, St. Petersburg, 1994, p. 272.
[1] A. Caneschi, D. Gatteschi, P. Rey, Prog. Inorg. Chem. 39 (1991)
331.
[2] A. Caneschi, D. Gatteschi, J. Laugier, L. Pardi, P. Rey, C.
Zanchini, Inorg. Chem. 27 (1988) 2027.
[3] A. Caneschi, D. Gatteschi, J. Laugier, P. Rey, R. Sessolli, C.
Zanchini, J. Am. Chem. Soc. 110 (1988) 2795.
[4] A. Caneschi, D. Gatteschi, P. Rey, R. Sessolli, Inorg. Chem. 27
(1988) 1756.
[5] A. Caneschi, F. Ferraro, D. Gatteschi, P. Rey, R. Sessoli, Inorg.
Chem. 29 (1990) 4217.
[6] A. Caneschi, D. Gatteschi, J. Laugier, P. Rey, R. Sessolli, Inorg.
Chim. Acta 184 (1991) 67.
[7] F. Lanfranc de Panthou, E. Belorizki, R. Calemczuk, D.
Luneau, C.h. Marcenat, E. Ressouche, P.h. Turek, P. Rey, J.
Am. Chem. Soc. 117 (1995) 11247.
.