Synthesis and characterization of [Cu(Me2oxpn)Ni(NO2)(tmen)](ClO4): a single ferrimagnetic dinuclear CuII-NiII complex acting as weak molecule-based magnet.

Abstract

The new heterodinuclear complex [Cu(Me2oxpn)Ni(NO2)(tmen)](ClO4), that exhibits strong antiferromagnetic intramolecular coupling between CuII and NiII ions (ferrimagnetic behavior), shows ferromagnetic ordering at low temperature, due likely to a small canting phenomenon; it is one of the very few compounds made from isolated molecules that lead to cooperative magnetic behavior.

Full text

Page 1

Synthesis and characterization of [Cu(Me2oxpn)Ni(NO2)(tmen)](ClO4): a
single ferrimagnetic dinuclear CuII–NiIIcomplex acting as weak
molecule-based magnet
Javier Tercero,aCarmen Diaz,aJoan Ribas,*aJosé Mahíaband Miguel Maestrob
aDepartament de Química Inorgànica, Universitat de Barcelona, Martí Franquès 1-11 08028. Barcelona,
Spain. E-mail: joan.ribas@qi.ub.es
bServicios Xerais de Apoio á Investigación, Facultade de Ciencias, Universidade da Coruña, Campus da
Zapateira s/n, A Coruña, Spain
Received (in Cambridge, UK) 23rd November 2001, Accepted 11th January 2002
First published as an Advance Article on the web 25th January 2002
The new heterodinuclear complex [Cu(Me2oxpn)Ni(NO2)-
(tmen)](ClO4), that exhibits strong antiferromagnetic intra-
molecular coupling between CuIIand NiIIions (ferrimag-
netic behavior), shows ferromagnetic ordering at low
temperature, due likely to a small canting phenomenon; it is
one of the very few compounds made from isolated
molecules that lead to cooperative magnetic behavior.
More and more researchers in coordination chemistry are
interested by bulk magnetic properties.1One of the main
challenges of this field is to design molecular-based compounds
exhibiting spontaneous magnetization. Spectacular results have
been reported in past years.1,2Additionally, molecules with
high-spin multiplicity in the ground state have also been
described.3To design molecular materials displaying zero-field
magnetization below a critical temperature Tc one must
assemble the magnetic molecular unit in such a way that the
interaction between the local spin carriers at the scale of the
lattice leads to a non-zero resulting spin. These results can be
achieved, for example, by 3D-ferromagnetism, 3D-ferrimagnet-
ism and, also, by the weak ferromagnetism derived from canted
antiferromagnetic spins. If this is the case, a weak ferromagnetic
transition is expected at Tc. Obviously, the onset of spontaneous
magnetization results from a three-dimensional transition.4It is
becoming evident that magnetic ordering is rather common in
materials with extended structures but not in species formally
derived from isolated molecules. Only very few examples have
been reported: {K2[CoO3PCH2N(CH2COO)2]2}6·xH2O,5a
[(Ni(bpm)2)3(Fe(CN)6)2]·7H2O5band some pure organic radi-
cals.6Evidently, this kind of molecular entity must be linked by
hydrogen bonds or van der Waals forces to give a three-
dimensional network with magnetic ordering. Our strategy
arises from using strong metal-ligand bonds and incorporates
more flexible bonds between molecules. We have successfully
used this synthetic strategy starting from ferrimagnetic mole-
cules.7
We report here the synthesis of a new compound [Cu-
(Me2oxpn)Ni(NO2)(tmen)](ClO4),† 1, (Me2oxpn = N,NA-bi-
s(3,amino-2,2A-dimethylpropyl)oxamido; tmen = N,N,NANA-
tetramethylethylenediamine) which has been characterized by
single-crystal X-ray structure analysis.‡ The drawing of the
cationic part, with atom labeling scheme is given in Fig. 1(a).
The Ni atom is in a distorted octahedral environment and the
coordination polyhedron of CuIIcan be considered as a square-
planar geometry. Its crystal structure is constituted by cationic
[Cu(Me2oxpn)Ni(NO2)(tmen)]+and ClO42anions giving su-
pramolecular one-dimensional assemblies [Fig. 1(b)]. The
chains show an alternation of [Cu–Ni] cation units linked by
hydrogen bonds between the hydrogen atoms of the N(1) and
the hydrogen atoms of the N(3) of the neighboring unit through
oxygen atoms of the perchlorate ion. The [Cu–Ni] units are
situated in a zigzag fashion along the chain.
Susceptibility measurements for complex 1 are shown in Fig.
2(a) as cMT vs. T per Cu–Ni unit.§ From room temperature
down to 40 K there is a clear fall in the cMT, which corresponds
to the classical behavior for an isolated [Cu–Ni] entity coupled
antiferromagnetically. 1 shows a clear decrease in cMT from 40
to 2 K. This decrease can be explained by magnetic interactions
through hydrogen bonds, as described above. The fit was
performed according to the Hamiltonian
H = 2JSn/2
by irreducible tensor operator formalism (ITO) using the
CLUMAG program.8The results were: J1(intra) = 2115.0
cm21, J2(inter) = 22.16 cm21, g = 2.27 and R = 1.14 3
1024.
A closer examination of the magnetism of 1 revealed quite
exceptional magnetic behavior for discrete magnetic molecules.
The study of cMT vs. T at different fields, [Fig. 2(b)], shows a
maximum in cMT at 11 K at lower fields. In a 100 G field, for
instance, cMTrises to a maximum of 0.66 cm3K mol21and then
falls rapidly as the temperature approaches 2 K. This transition
at 11 K could indicate the onset of long-range magnetic
ordering, although this phenomenon is not as abrupt as in other
i = 1[S2iS2i 2 1+ aS2iS2i + 1],
Fig. 1 (a) Drawing of the cationic part, with the atom labeling scheme of 1.
Ellipsoids at the 50% probability level. (b) Projection down the zx plane of
the chains, showing the alternation of [Cu–Ni] cations linked by hydrogen
bonds (dotted lines) through oxygen atoms of the perchlorate anion.
Fig. 2 (a) Experimental and calculated variations of the product cMT vs. T
for 1 (10.000 gauss; at low fields there are cooperative phenomena, see (b);
(b) temperature dependence of the product cMT vs. T for 1, at the field
values shown.
This journal is © The Royal Society of Chemistry 2002
364
CHEM. COMMUN., 2002, 364–365
DOI: 10.1039/b110747j

Page 2

cases reported.9Further, to confirm the occurrence of magnetic
phase transition, the DC magnetization was measured with the
sample cooled in zero-field (ZFCM) and in a field of 100 gauss
(FCM). The FCM gave a rapid increase below 15 K and the
ZFCM diverged from FCM from about 11 K [Fig. 3(a)]. The
difference ZFCM 2 FCM corresponds to the remnant magneti-
zation (RM) induced on cooling within the field below ca. 11 K.
This remnant magnetization is very weak (5 cm3G mol21at 2
K). At 2 K the magnetization curve exhibits a very small
hysteresis cycle, with a coercive field of ca. 30 Oe (the possible
error of the measurement is less than 10 Oe). Even very small,
there is a remnant magnetization. In some few cases, these
molecular-based magnets do not exhibit hysteresis, demonstrat-
ing weak anisotropy.10Finally, to complete the characterization
of the magnetic ordering, the ac susceptibility measurements at
100 and 1000 Hz show an weak out-of-phase signal which has
a maximum close to 11 K (the temperature of the onset of the
magnetic ordering) [Fig. 3(b)].
The origin of this behavior, very rare for discrete molecules,
should be emphasized that is due to the interactions between the
molecules giving a three-dimensional net by hydrogen or van
der Waals forces. Taking into account the strong intradinuclear
antiferromagnetic coupling between CuIIand NiII(J = 2115
cm21) the ground state at 11 K, where the ferromagnetic
ordering begins to be operative, is S = 1⁄2. The phase change
from paramagnet to ferromagnet could be due to a canted
antiferromagnet indicating the significant influence of either the
single-ion anisotropy or the antisymmetric part of the spin
Hamiltonian, dij[Si3 Sj].11In 1, D and E are, in principle, zero
because the ground state is S = 1⁄2. The antisymmetric part is
proportional to [(g 2 2)/g]J where the intermolecular coupling,
J, is 22.16 cm21and the average g value is 2.27 (which agrees
with the value obtained by EPR). Thus, although small, this
term is not zero. This is the reason given in the literature for
MnIIor even CuIIcanted complexes.4c–e
It is important to point out that we have also prepared and
fully characterized the isostructural complex [Cu(Me2oxpn)Ni-
(NO2)(trimen)](ClO4) 2 (trimen = N,N,NA-trimethylethylene-
diamine), which does not show any signal of weak ferromagne-
tism (possible canted antiferromagnetism) neither in cMT
values at different fields nor remnant magnetization or cMB out-
of-phase signal in ac susceptibility measurements. The only
apparent difference between 1 and 2 is the presence of four
methyl groups in the blocking amine of 1 (tmen) and three
methyl groups in 2 (trimen). Thus, very fine subtleties could
explain why complex 2 does not exhibit weak ferromagnetism,
while complex 1, does. Further studies are necessary to interpret
the weak ferromagnetism in 1 and clarify the unexpected
difference between the two isostructural complexes, 1 and 2.
This work was supported by the Spanish Dirección General
de Investigación Científica y Técnica (project BQU2000-
0791).
Notes and references
† [Cu(Me2oxpn)Ni(NO2)(tmen)](ClO4) 1: an ethanolic solution (20 mL) of
N,N,NA,NA-tetramethylethylenediamine (tmen) (0.32 g, 2.73 mmol) was
added to a stirred solution of Ni(NO3)2·6H2O (1 g, 2.73 mmol) in ethanol
(20 mL). A suspension of [Cu(Me2oxpn)]12(0.87 g, 2.73 mmol) in ethanol
(40 mL) was then added. The blue solution was filtered, and NaNO2(0.189
g, 2.73 mmol) was added. The resulting blue solution was left to evaporate
slowly at room temperature. Violet monocrystals suitable for X-ray
determination were collected after 1 week (yield ca. 65%) and were
identified by IR spectroscopy and elemental analysis. Calc. (found): C,
33.70 (33.9); H, 6.30 (6.3); N, 15.31 (15.1)%.
‡ Crystal data for 1: C18H40ClCuN7NiO8, crystal dimensions 0.45 3 0.40
3 0.35 mm, orthorhombic, space group Pbca, a = 11.616(1), b =
17.996(1), c = 26.613(1) Å, V = 5563.0(1) Å3, Z = 8, Dc= 1.529 g cm23,
T = 298 K, 36060 reflections, 6878 independent reflections (Rint =
0.0296). Lorentz polarization and absortion corrections were applied, m =
1.590 mm21, 370 parameters refined, R1 = 0.0385, wR2 = 0.0936 (for all
data). Mo-Karadiation (l = 0.71073 Å). The structure was solved by direct
methods using SHELXS-87 computer program for crystal structure
determination and refined by full-matrix least-squares methods on F2, with
the SHELXL-97 computer program. Hydrogen atoms were included in
calculated positions and refined in riding mode. Ionic perchlorate shows
high thermal parameters and less than ideal geometry.CCDC reference
number 166398. See http://www.rsc.org/suppdata/cc/b1/b110747j/ for
crystallographic data in CIF or other electronic format.
§ Magnetic susceptibility measurements were carried out on polycrystalline
samples with a SQUID apparatus working in the range 2–300 K. To be sure
of the magnetic measurements (magnetic ordering) the measurements have
been performed in two different SQUIDS in two different Centers
(Barcelona and Toulouse) with different samples. Results were con-
cordant.
1 Magnetism: Molecules to Materials. Models and Experiments, vol. 1;
Molecule-based Magnets, vol. 2; and Nanosized Magnetic Materials,
vol. 3; ed. J. S. Miller and M. Drillon, Wiley-VCH, 2001.
2 Metal–Organic and Organic Molecular Magnets, ed. P. Day and A. E.
Underhill, The Royal Society of Chemistry, Cambridge, 1999.
3 S. M. J. Aubin, Z. Sun, H. J Eppley, E. M. Rumberger, I. A. Guzei, K.
Folting, P. K. Gantzel, A. L. Rheinhold, G. Christou and D. N.
Hendrickson, Inorg. Chem., 2001, 40, 2127 and references therein.
4 (a) M. L. Kirk, M. S. Lah, C. Raptopoulou, D. P. Kessissoglou, W. E.
Hatfield and V. L. Pecoraro, Inorg. Chem., 1991, 30, 3900; (b) J. Ribas,
M. Monfort, X. Solans and M. Drillon, Inorg.Chem., 1994, 33, 742; (c)
E. Bakalbassis, P. Bergerat, O. Kahn, S. Jeannin, Y. Jeannin, Y.
Dromzee and M. Guillot, Inorg. Chem., 1992, 31, 625; (d) C. R. Kmety,
Q. Huang, J. W. Lynn, R. W. Erwin, J. L. Manson, S. McCall, J. E.
Crow, K. L. Stevenson, J. S. Miller and A. J. Epstein, Phys. Rev. B,
2000, 62, 5576; (e) P. Jensen, D. J. Price, S. R. Batten, B. Moubaraki and
K. S. Murray, Chem. Eur. J., 2000, 6, 3186.
5 (a) S. O. H. Gutschke, D. J. Price, A. K . Powell and P. T. Wood, Angew.
Chem., Int. Ed., 1999, 38, 1088; (b) K. V. Langenberg, S. R. Batten, K.
J. Berry, D. C. R. Hockless, B. Moubaraki and K. S. Murray,
Inorg.Chem., 1997, 36, 5006.
6 M. Mito, H. Nakano, T. Kawae, M. Hitaka, S. Takagi, H. Deguchi, K.
Suzuki, K. Mukai and K. Takeda, J. Phys. Soc. Jpn., 1997, 66, 2147.
7 J. Tercero, C. Diaz, M. S. El Fallah, J. Ribas, X. Solans, M. A. Maestro
and J. Mahía, Inorg. Chem., 2001, 40, 3077 and references therein.
8 D. Gatteschi and L. Pardi, Gazz. Chim. Ital., 1993, 123, 231.
9 S. R. Batten, P. Jensen, C. J. Kepert, M. Kurmoo, B. Moubaraki, K. S.
Murray and D. J. Price, J. Chem. Soc., Dalton Trans., 1999, 2987 and
references therein.
10 Z. J. Zhang, H. Seino, Y. Mizobe, M. Hidai, M. Verdaguer, S. Ohkoshi
and K. Hashimoto, Inorg. Chem., 2000, 39, 5095.
11 (a) R. C. Carlin, Magnetochemistry, Springer-Verlag, Berlin, 1986, p.
148; (b) T. Moriya, in Magnetism, ed. G. T. Rado and H. Suhl,
Academic Press, London, 1963, vol. 1, p. 85; (c) T. Moriya, Phys. Rev.,
1960, 117, 635.
12 J. Ribas, A. Garcia and M. Monfort, Polyhedron, 1991, 10, 103.
Fig. 3 (a) Temperature dependence of the magnetization, M, for 1 cooled in
zero-field (ZFCM) and in field of 100 G (FCM). Remnant magnetization
(RM) is also shown; (b) Real or in phase (cA) and imaginary or out-of-phase
(cB) components of the zero-field ac susceptibility (cac) as a function of
temperature, measured in an oscillating field of amplitude, Hac = 3 G and
frequencies, f = 100 and 1000 Hz.
365
CHEM. COMMUN., 2002, 364–365