A tetranuclear CrIIINiII3 cyano-bridged complex based on M(tacn) derivative building blocks

Article abstract of DOI:10.1021/ic0514391

The reaction of Cr(Bztacn)(CN)₃ (Bztacn is 1,4,7-trisbenzyl-1,4, 7-triazacyclononane) with Ni(iPrtacn)Cl₂ (iPrtacn is 1,4,7-trisisopropyl-1,4,7-triazacyclononane) affords a CrNi₃ tetranuclear complex. Variable temperature and magnetization versus field measurements show a S = 9/2 ground state and an appreciable magnetic anisotropy with a negative D_9/2 value equal to -0.54 cm^-1. Magnetization studies on one single crystal using a micro-SQUID show a fast tunneling process at zero field at very low temperature.

Full text of DOI:10.1021/ic0514391

Inorg. Chem. 2005, 44, 8194−8196
A Tetranuclear CrIIINiII Cyano-Bridged Complex Based on M(tacn)
3
Derivative Building Blocks
Jean-Noe1
l Rebilly,† Laure Catala,† Eric Rivie`re,† Re´gis Guillot,† Wolfgang Wernsdorfer,‡ and
Talal Mallah*,†
Laboratoire de Chimie Inorganique, ICMMO, CNRS UMR 8613, UniVersite´ Paris-Sud, 91405
Orsay, France, and Laboratoire Louis Ne´el, CNRS, 25 AVenue des Martyrs, BP166, 38042
Grenoble, Cedex 9, France
Received August 24, 2005
The reaction of Cr(Bztacn)(CN)3 (Bztacn is 1,4,7-trisbenzyl-1,4,7-
triazacyclononane) with Ni(iPrtacn)Cl2 (iPrtacn is 1,4,7-trisisopropyl-
1,4,7-triazacyclononane) affords a CrNi3 tetranuclear complex.
Variable temperature and magnetization versus field measurements
9
show a S
)
/
ground state and an appreciable magnetic
2
1
anisotropy with a negative D9 value equal to
−
0.54 cm-
.
/
2
BztacnCr(CN)3 and Ni(iPrtacn)Cl2 in methanol (Bztacn is
1,4,7-trisbenzyl-1,4,7-triazacyclononane and iPrtacn is 1,4,7-
trisisopropyl-1,4,7-triazacyclononane).4,5 Because the Bztac-
nCr(CN)3 complex is almost insoluble in common solvents,4
it was added as a solid to the Ni-containing solution. A color
change (yellow to green) followed by a complete dissolution
of the chromium complex occurred. Green needlelike single
crystals suitable for structure determination grew within a
Magnetization studies on one single crystal using a micro-SQUID
show a fast tunneling process at zero field at very low temperature.
Since the report of the first polynuclear high-spin com-
plexes using cyanides as the bridging ligand to transmit the
electronic information between the paramagnetic ions,1 much
effort has been invested to prepare new complexes mainly
using a stepwise approach.2 New bimetallic complexes
including some showing single-molecule magnet behavior
have been discovered during the past few years.3 Recently,
using a two-step synthetic approach, Dunbar and co-workers
obtained a trimetallic complex thanks to the presence of the
chemically inert Co(III) ions.2s,t
(2) (a) Heinrich, J. L.; Berseth, P. A.; Long, J. R. Chem. Commun. 1998,
1231. (b) Marvilliers, A.; Pei, Y.; Cano Boquera, J.; Vostrikova, K.
E.; Paulsen, C.; Rivie`re, E.; Audie`re, J.-P.; Mallah, T. Chem. Commun.
1999, 1951. (c) Vostrikova, K. E.; Luneau, D.; Wernsdorfer, W.; Rey,
P.; Verdaguer, M. J. Am. Chem. Soc. 2000, 122, 718. (d) Berseth, P.
A.; Sokol, J. J.; Shores, M. P.; Heinrich, J. L.; Long, J. R. J. Am.
Chem. Soc. 2000, 122, 9655. (e) Sokol, J. J.; Shores, M. P.; Long, J.
R. Angew. Chem., Int. Ed. 2001, 40, 236. (f) Heinrich, J. L.; Sokol, J.
J.; Hee, A. G.; Long, J. R. J. Solid State Chem. 2001, 159, 293. (g)
Smith, J. A.; Galan-Mascaros, J. R.; Cle´rac, R.; Sun, J. S.; Xiang, O.
Y.; Dunbar, K. R. Polyhedron 2001, 20, 1727. (h) Rogez, G.; Parsons,
S.; Paulsen, C.; Villar, V.; Mallah, T. Inorg. Chem. 2001, 40, 3836.
(i) Sokol, J. J.; Shores, M. P.; Long, J. R. Inorg. Chem. 2002, 41,
3052. (j) Lescoue¨zec, R.; Vaissermann, J.; Lloret, F.; Julve, M.;
Verdaguer, M. Inorg. Chem. 2002, 41, 5943. (k) Podgajny, R.;
Desplanches, C.; Sieklucka, B.; Sessoli, R.; Villar, V.; Paulsen, C.;
Wernsdorfer, W.; Dromzee, Y.; Verdaguer, M. Inorg. Chem. 2002,
41, 1323. (l) Rogez, G.; Marvilliers, A.; Sarr, P.; Parsons, S.; Teat, S.
J.; Ricard, L.; Mallah, T. Chem. Commun. 2002, 1460. (m) Si, S. F.;
Tang, J. K.; Liu Liao, D. Z.; Jiang, Z. H.; Yan, S. P.; Cheng, P. Inorg.
Chem. Commun. 2003, 6, 1109. (n) Marvaud, V.; Decroix, C.; Scuiller,
A.; Guyard-Duhayon, C.; Vaissermann, J.; Gonnet, F.; Verdaguer, M.
Chem. Eur. J. 2003, 9, 1678. (o) Wang, S.; Zuo, J. L.; Zhou, H. C.;
Choi, H. J.; Ke, Y.; Long, J. R.; You, X. Z. Angew. Chem., Int. Ed.
2004, 43, 5940. (p) Lee, I. S.; Long, J. R. Dalton Trans. 2004, 3434.
(q) Karadas, F.; Schelter, E. J.; Prosvirin, A. V.; Basca, J.; Dunbar,
K. R. Chem. Commun. 2005, 1414. (r) Beltran, L. M. C.; Long, J. R.
Acc. Chem. Res. 2005, 38, 325. (s) Berlinguette, C. P.; Dunbar, K. R.
Chem. Commun. 2005, 19, 2451. (t) Berlinguette, C. P.; Dragulescu-
Andrasi, A.; Sieber, A.; Gudel, H. U.; Achim, C.; Dunbar, K. R. J.
Am. Chem. Soc. 2005, 127, 6766. (u) Withers, J. R.; Ruschmann, C.;
Bojang, P.; Parkin, S.; Holmes, S. M. Inorg. Chem. 2005, 44, 352.
(v) Ni, Z. H.; Kou, H. Z.; Zhao, Y. H.; Zheng, L.; Wang, R. J.; Cui,
A. L.; Sato, O. Inorg. Chem. 2005, 44, 2050.
In this Communication, we report the preparation, the
crystal structure, and the magnetic behavior of a tetranuclear
CrIIINiII complex using triazacyclononane derivatives as
3
blocking ligands. Triazacyclononane-based complexes had
first been successfully used by Long and co-workers, and a
variety of complexes with different architectures were
stabilized.2r
The tetranuclear complex of formula [BztacnCr(CNNi-
(iPrtacn)Cl)3]Cl3‚10H2O (1; see scheme for Bztacn and
iPrtacn) was obtained from the stoichiometric reaction of
* To whom correspondence should be addressed. E-mail: mallah@
icmo.u-psud.fr.
† Laboratoire de Chimie Inorganique, ICMMO, CNRS UMR 8613.
‡ Laboratoire Louis Ne´el, CNRS.
(1) (a) Mallah, T.; Auberger, C.; Verdaguer, M.; Veillet, P. J. Chem. Soc.,
Chem. Commun. 1995, 61. (b) Scuiller, A.; Mallah, T.; Verdaguer,
M.; Nivorozkhin, A.; Tholence, J.-L.; Veillet, P. New J. Chem. 1996,
20, 1. (c) Langenberg, K. V.; Batten, S. R.; Berry, K. J.; Hockless, D.
C. R.; Moubaraki, B.; Murray, K. S. Inorg. Chem. 1997, 36, 5006.
8194 Inorganic Chemistry, Vol. 44, No. 23, 2005
10.1021/ic0514391 CCC: $30.25
© 2005 American Chemical Society
Published on Web 10/11/2005

COMMUNICATION
Figure 2. øMT ) f(T) for 1: (O) experimental data, (s) best fit including
D, and (- - -) best fit including zJ (see text). Inset: M ) f(H/T), (O) T ) 6
K, (0) T ) 4 K, (∆) T ) 2 K, and (s) best fit.
Figure 1. (left) View of the molecular structure of [BztacnCr(CNNi(iPrt-
acn)Cl)3]3+. (right) View of the structure of a Ni unit belonging to the
tetranuclear complex. Hydrogen atoms were omitted for clarity.
bridging three Ni ions each capped by the iPrtacn ligand.
The coordination sphere of the Ni atoms is pentacoordinated
because only one chloride completes the coordination sphere
(Figure 1, right). The geometries around the three crystal-
lographically independent Ni atoms of the complex are
similar and close to a distorted trigonal bipyramid.
few days by a slow diffusion of tert-butyl methyl ether into
the mother solution.6
Compound 1 crystallizes in the monoclinic P1h space
group.7 The molecular structure of the tetranuclear complex
(Figure 1, left) shows the central chromium atom linked to
the Bztacn ligand and three cyanide groups in facial position
The pseudotrigonal axis is along two nitrogen atoms: one
from the bridging cyanide (N50) and the other (N11)
belonging to the iPrtacn ligand; the N50-Ni1-N11 angle
is almost linear and is equal to 176.8°. For the other two Ni
units, this angle is mainly the same (176.6 and 176.7°). The
base of the bipyramid is formed by the remaining nitrogen
atoms of the organic ligand (N13 and N12) and the chloride
atom. Because of the bite angle of the tacn-based ligand,
the N12-Ni1-N13 angle is equal to 87.5° and the two
N12-Ni1-Cl1 and N13-Ni1-Cl1 angles are equal to 144.0
and 128.5°, respectively. It is important to note that the
pentacoordinated environment around the Ni atoms is prob-
(3) (a) Sokol, J. J.; Hee, A. G.; Long, J. R. J. Am. Chem. Soc. 2002, 124,
7656. (b) Berlinguette, C. P.; Vaughan, D.; Canada-Vilalta, C.; Galan-
Mascaros, J. R.; Dunbar, K. R. Angew. Chem., Int. Ed. 2003, 42, 1523.
(c) Palii, A. V.; Ostrovsky, S. M.; Klokischner, S. I.; Tsukerblat, B.
S.; Berlinguette, C. P.; Dunbar, K. R.; Galan-Mascaros, J. R. J. Am.
Chem. Soc. 2004, 126, 16860. (d) Wang, S.; Zuo, J. L.; Zhou, H. C.;
Choi, H. J.; Ke, Y.; Long, J. R.; You, X. Z. Angew. Chem., Int. Ed.
2004, 43, 5940. (e) Choi, J.; Sokol, J. J.; Long, J. R. Inorg. Chem.
2004, 43, 1606. (f) Schelter, E. J.; Prosvirin, A. V.; Reiff, W. M.;
Dunbar, K. R. J. Am. Chem. Soc. 2004, 126, 15004. (g) Choi, H. J.;
Sokol, J. J.; Long, J. R. J. Phys. Chem. Solids 2004, 65, 839. (h)
Miyasaka, H.; Takahashi, H.; Madanbashi, T.; Suguira, K.-I.; Cle´rac,
R.; Nojiri, H. Inorg. Chem. 2005, 44, 5969.
(4) The tacn-based ligands were prepared as already reported.5 The
BztacnCr(CN)3 was prepared from its bromide precursor5 BztacnCrBr3
as described: 1.10 g (1.61 × 10-3 M) of BztacnCrBr3 was dissolved
in 10 mL of degassed DMSO and heated to 60 °C for 30 min under
argon. A total of 5.81 g (1.2 × 10-1 M) of NaCN was added, and the
reaction mixture was heated to 80 °C for 12 h. After cooling, the
compound was precipitated as a pink-yellow solid upon the addition
of 100 mL of water, filtered, washed with a small amount of cold
water, and dried under vacuum. BztacnCr(CN)3 is insoluble in water,
methanol, ethanol, acetonitrile, acetone, THF, chloroform, and dichlo-
romethane and very slightly soluble in DMSO and DMF. IR (KBr
pellet): ν(CN) 2125 cm-1. Yield: 90%. øMT(295): 1.90 cm3 mol-1
K. The iPrtacnNiCl2 complex was prepared as follows: 0.46 g (2 ×
10-3 M) of NiCl2‚6H2O was dissolved in 10 mL of methanol, 2 mL
of trimethylorthoformate was added, and the solution was mixed with
30 mL of DME and 100 mL of THF. It was then heated under reflux
for 10 min. Then 0.5 g (2 × 10-3 M) of iPrtacn was dissolved in 20
mL of THF, and this solution was added dropwise to the first solution.
A green-yellowish precipitate appeared upon cooling of the solution.
The precipitate was filtered, then dissolved in a minimum amount of
chloroform, and precipitated by the addition of an excess of THF.
The obtained solid was filtered, thouroughly washed with THF, and
dried under vacuum. Anal. Calcd for C15H33Cl2Ni (%): C, 46.79; H,
8.64; N, 10.91; Cl, 18.42; Ni, 15.24. Found: C, 46.45; H, 8.77; N,
10.99; Cl, 18.39; Ni, 14.96. Yield: 67%. øMT(295): 1.15 cm3 mol-1
K.
i
ably due to the presence of the three Pr groups on the
capping ligand, which induces steric hindrance and prevents
the coordination of the sixth ligand.
The N-C-Ni angles (where N-C is a cyanide ligand)
are very close to linearity and range from 169 to 174°. Within
the crystal, the molecules are quite far apart, with the shortest
Ni-Ni distance belonging to two different molecules larger
than 7.6 Å. However, the presence of water molecules creates
a 2D H-bond network linking the clusters together within
the crystal (Figure S1 in the Supporting Information).
The magnetic studies within the temperature range 300-2
K and for an applied field of 2000 Oe above 100 K and 100
Oe below (to avoid saturation effects) show that the øMT )
f(T) curve increases upon cooling (Figure 2).
This is due to the presence of a ferromagnetic exchange
3
coupling interaction between the central Cr(III) [S ) /2;
(dxy)1(dxz)1(dyz)1 electronic configuration] and the three Ni(II)
1
1
2
2
2
[S ) 1; (dz ) (dx -y ) configuration] ions as expected from
the orthogonality of the magnetic orbitals.8 The øMT value
at room temperature (5.8 cm3 mol-1 K) is slightly higher
(5) Haselhorst, G.; Stoetzel, S.; Strassburger, A.; Walz, W.; Wieghardt,
K.; Nuber, B. J. Chem. Soc., Dalton Trans. 1993, 83.
(6) Anal. Calcd for 1 (%): C, 48.53; H, 8.36; N, 11.09; Cl, 11.31; Ni,
9.36; Cr, 2.76. Found: C, 48.41; H, 7.91; N, 11.09; Cl, 9.88; Ni, 9.24;
Cr, 3.00. IR (KBr pellet): ν(CN) 2131 cm-1. Yield: 70%.
than the expected one for uncoupled three Ni(II) (S ) 1)
3
and one Cr(III) (S ) /2) metal ions (4.875 cm3 mol-1
K
(7) Crystal data for 1: C75H168Cl6CrN15Ni3O18, fw ) 2009.0, triclinic P1h,
T ) 293 ( 2 K, a ) 17.2543(3) Å, b ) 17.8334(4) Å, c ) 19.3734(4)
Å, R ) 68.4360(10)°, â ) 66.7440(10)°, γ ) 82.9780(10)°, V )
5091.73(18) Å3, Z ) 2, Dcalc ) 1.287 g cm-3, F000 ) 2082.00, µ(Mo
KR) ) 8.67 cm-1, final R1 ) 0.0574 [I > 4.00σ(I)], R ) 0.0911 (all
assuming the same g value equal to 2). The maximum value
of øMT (11.3 cm3 mol-1 K) at T ) 12 K is slightly lower
data), wR2 ) 0.1545 (all data), GOF ) 1.016, Fmax ) 0.832 e Å-3
,
(8) Gadet, V.; Mallah, T.; Castro, I.; Verdaguer, M.; Veillet, P. J. Am.
Chem. Soc. 1992, 114, 9213.
Fmin ) -0.674 e Å-3
.
Inorganic Chemistry, Vol. 44, No. 23, 2005 8195

COMMUNICATION
than that corresponding to S ) 9/2 (12.375 cm3 mol-1 K with
an average g value of 2) arising from the ferromagnetic
interaction. Below 12 K, øMT decreases. The lower value of
øMT and the decrease at low temperature may be ascribed
to zero-field splitting of the spin ground state and/or
antiferromagnetic intermolecular interactions. The magnetic
data can be modeled using the following Heisenberg spin
Hamiltonian: H ) -JCr-Ni[SCr‚SNi1 + SCr‚SNi2 + SCr‚SNi3].
The data were fitted for two cases: (i) considering only the
9
Figure 3. Micro-SQUID magnetization vs field studies at T ) 0.04 K
and between -1 and +1 T (left) and 0 and 1 T (right): (s) sweep rate
0.56 T s-1; (- - -) sweep rate 0.008 T s-1. Only the data between -0.3 and
+0.3 T are shown for clarity (see the Supporting Information).
zero-field splitting parameter D of the ground state S ) /2
and (ii) considering only intermolecular interaction (zJ)
within the near-field approximation. The two approaches give
satisfactory results (JCr-Ni ) 20 cm-1, g ) 2.07, |D| ) 2.5
cm-1 and JCr-Ni ) 20 cm-1, g ) 2.03, zJ ) -0.07 cm-1).
The quality of the fit is not improved by considering more
than one J parameter because the J value is almost
independent from the N-C-Ni angle for angles larger than
170°.9 The D value found does not reflect the real magnitude
of the zero-field splitting of the ground state; it is overes-
timated because it reflects the decrease of øMT at low
temperature, which may be due to antiferromagnetic inter-
molecular interactions as well. The best way to get a better
idea of the D value is to fit the magnetization data at low
temperature and a field higher that 0.5 T, where the
intermolecular interactions are overcome. The magnetization
data for 1 were measured at T ) 2, 4, and 6 K in the field
range 0.6-5.5 T (Figure 2, inset). The three curves are not
superimposable, which is the signature of the presence of
appreciable magnetic anisotropy within the compound. The
magnetization value (8.5 µB) at H ) 5.5 T for the lowest
temperature curve (T ) 2 K) is slightly lower than that for
9 µB (expected value for a S ) 9/2 ground state with g ) 2).
the system shows a strong transition at H ) 0 due to fast
relaxation. To check whether the H ) 0 transition is due to
the tunneling of the magnetization, the sample was cooled
from high temperature to 40 mK in zero field, getting a
magnetization state with half of the spins up and the others
down. We, then, ramped the field up to 1 T at different field
sweep rates, observing a larger hysteresis effect than when
the field was set at a negative field and cycled. This is a
signature of the tunneling process at zero field because when
starting at a negative field, all of the molecules are in the
MS ) S state and when the field approaches zero, most of
them tunnel, leading to a strong transition at zero field
(Figure 3, left). While when the field is set at zero (Figure
3, right), only half of the molecules are in the MS ) S state,
upon an increase in the field, half of them are blocked in
the up state and they are not able to tunnel (H > 0), which
leads to a larger hysteresis loop.10
For a half-integer spin state, the degeneracy of the Kramers
doublets cannot be lifted by crystal field effects. So, no
tunneling at zero field should be observed; however, very
weak dipolar or exchange fields due to intermolecular
interactions may do so. This is probably the case in the
present compound because a 2D H-bond network linking
the molecules together is indeed present.
9
This is coherent with a S ) /2 ground state experiencing
some magnetic anisotropy. The energy difference between
9
7
the ground S ) /2 state and the first excited S ) /2 spin
state is around 30 cm-1 (3JCr-Ni/2). It is thus reasonable to
assume that, even at T ) 6 K, only the ground state is
populated. The M ) f(H/T) data were then fitted by
calculating the magnetization versus field at the three
temperatures for different orientations of the magnetic field.
The best fit leads to D ) -0.54 cm-1, E/D ) 0.29, and g )
2.00 with an agreement factor R ) 5 × 10-4. Magnetization
studies give the magnitude of the anisotropy parameters; only
detailed EPR studies may give accurate values. However,
because a negative D value was obtained for complex 1, we
investigated its behavior at low temperature, where a
blocking of the magnetization may be observed.
In summary, metal complexes containing tacn derivatives
allow one to build simple polynuclear complexes with well-
defined ground states. Detailed EPR studies on 1 and on the
mononuclear Ni(iPrtacn)Cl2 complex are underway in order
to determine accurately the anisotropy parameters and the
relationship between local anisotropy terms (DNi and ENi)
9
and those related to the S ) /2 ground state.
Acknowledgment. The authors thank the CNRS (Centre
National de la Recherche Scientifique) and the European
community for financial support (Contract MRTN-CT-2003-
504880/RTN Network “QuEMolNa”).
The magnetization studies were performed using an array
of micro-SQUIDs on one single crystal oriented with its
anisotropy axis parallel to the applied magnetic field. The
magnetization versus field values were measured at T ) 40
mK for different sweep rates of the magnetic field (only two
sweep rates are shown below). The field was cycled between
-1 and +1 T (Figure 3, left). A hysteresis is observed as a
result of the slow relaxation of the magnetization. However,
Supporting Information Available: X-ray data for 1 in CIF
format and Figures S1-S4. This material is available free of charge
via the Internet at http://pubs.acs.org.
IC0514391
(9) Toma, L.; Toma, L. M.; Lescoue¨zec, R.; Armentano, D.; De Munno,
G.; Andruh, M.; Cano, J.; Lloret, F.; Julve, M. J. Chem. Soc., Dalton
Trans. 2005, 1357.
(10) Such fast tunneling at zero field prevents the observation of dynamical
magnetic behavior, so no ac signal was observed in the absence of a
dc applied magnetic field.
8196 Inorganic Chemistry, Vol. 44, No. 23, 2005