Single-molecule-magnet behavior and spin changes affected by crystal packing effects.

Abstract

Five Mn 3Zn 2 heterometallic complexes have been synthesized and structurally and magnetically characterized. Spin ground states up to S = 6 have been observed for these complexes and are shown to depend on the cocrystallizing cation and on the terminal ligand. Large axial zero-field interactions ( D = -1.16 K) are the result of near-parallel alignment of the Mn (III) Jahn-Teller axes. High-frequency electron paramagnetic resonance, single-crystal magnetization hysteresis, and alternating current susceptibility measurements are presented to characterize [NEt 4] 3[Mn 3Zn 2(salox) 3O(N 3) 6X 2] [X (-) = Cl (-) ( 1), Br (-) ( 2)] and [AsPh 4] 3[Mn 3Zn 2(salox) 3O(N 3) 6Cl 2] ( 3) and reveal that 1 and 2 are single-molecule magnets ( U eff = 44 K), while 3 is not.

Full text

Page 1

Single-Molecule-Magnet Behavior and Spin Changes Affected by Crystal
Packing Effects
Patrick L. Feng,†Changhyun Koo,‡John J. Henderson,§Motohiro Nakano,|Stephen Hill,‡Enrique del
Barco,§and David N. Hendrickson*,†
Department of Chemistry and Biochemistry, UniVersity of California, San Diego, La Jolla,
California 92093-0358, Department of Physics, UniVersity of Florida, GainesVille, Florida 32611,
Department of Physics, UniVersity of Central Florida, Orlando, Florida 32816-2385, and DiVision
of Applied Chemistry, Osaka UniVersity, Suita, Osaka 565-0871, Japan
Received July 1, 2008
FiveMn3Zn2heterometalliccomplexeshavebeensynthesizedand
structurallyandmagneticallycharacterized.Spingroundstatesup
toS) 6havebeenobservedforthesecomplexesandareshown
todependonthecocrystallizingcationandontheterminal ligand.
Largeaxial zero-fieldinteractions(D) -1.16K)aretheresult of
near-parallel alignment of the MnIIIJahn-Teller axes. High-
frequency electron paramagnetic resonance, single-crystal mag-
netization hysteresis, and alternating current susceptibility
measurements are presented to characterize [NEt4]3[Mn3Zn2-
(salox)3O(N3)6X2] [X-) Cl-(1), Br-(2)] and [AsPh4]3[Mn3Zn2-
(salox)3O(N3)6Cl2] (3) andreveal that 1and2aresingle-molecule
magnets (Ueff) 44 K), while 3 is not.
The observance of slow magnetization relaxation and quan-
tum effects in single-molecule magnets (SMMs) has led to
intense research interest in the area.1-7The basic requirements
for a SMM include an appreciable spin ground state and a
negative uniaxial anisotropy.7One synthetic strategy has been
to maximize the spin by facilitating ferromagnetic interactions
among anisotropic metal ions such as Mn3+. Impressively large
spin ground states up to83/2have been achieved but have not
resulted in increased barriers due to corresponding decreases
in |D|.8A related approach involves maximizing the magnitude
of |D| while maintaining an appreciable spin. This method has
been particularly successful for the synthesis of SMMs that are
supported by phenolic oximes.9-12In fact, the largest barrier
to magnetization reversal to date has been observed in the
complex [Mn6O2(Et-sao)6(O2CPh(Me)2)2(EtOH)6] as a result of
a large negative anisotropy (D ) -0.618 K) and an S ) 12
spin ground state.9
There is also significant interest in understanding the
specific factors that govern ferromagnetic exchange in oxo-
centered Mn3triangles, which serve as molecular building
blocks for complexes such as Mn6.10,13Variations in the
Mn-N-O-Mn angle in oximate-supported complexes have
been shown to affect the nature and magnitude of magnetic
exchange interactions, where larger torsion angles correspond
to stronger ferromagnetic interactions. These studies, how-
ever, have been based on comparisons between nonidentical
Mn3 triangles differing in coordination environment and
chelating ligands.10,13Indeed, the S ) 12 Mn6SMMs were
also obtained via chemical modification of the oxime
ligand.12Our complexes 1 and 3 have the formula
[cation]3[Mn3Zn2(salox)3O(N3)6X2], with X-) Cl-, and
switch from S ) 6 to a strongly mixed low-spin ground state
as a result of a cocrystallizing cation change from [NEt4]+
to [AsPh4]+. Likewise, complexes 2, 4, and 5 have [cation]+
) [NEt4]+and exhibit similar spin-state changes as a result
* To whom correspondence should be addressed. E-mail: dhendrickson@
ucsd.edu. Fax: 858-534-5383.
†University of California, San Diego.
‡University of Florida, Gainesville.
§University of Central Florida.
|Osaka University.
(1) Sessoli, R.; Tsai, H.-L.; Schake, A. R.; Wang, S.; Vincent, J. B.;
Folting, K.; Gatteschi, D.; Christou, G.; Hendrickson, D. N. J. Am.
Chem. Soc. 1993, 115, 1804.
(2) Sessoli, R.; Gatteschi, D.; Caneschi, A.; Novak, M. A. Nature 1993,
365, 141.
(3) Thomas, L.; Lionti, F.; Ballou, R.; Gatteschi, D.; Sessoli, R.; Barbara,
B. Nature 1996, 383, 145.
(4) Wernsdorfer, W.; Aliaga-Alcalde, N.; Hendrickson, D. N.; Christou,
G. Nature 2002, 416, 406.
(5) Hill, S.; Edwards, E. S.; Aliaga-Alcalde, N.; Christou, G. Science 2003,
302, 1015.
(6) Wernsdorfer, W.; Bhaduri, S.; Boskovic, C.; Christou, G.; Hendrickson,
D. N. Phys. ReV. B 2002, 65, 180403.
(7) Gatteschi, D.; Caneschi, A.; Pardi, L.; Sessoli, R. Science 1994, 265,
1054.
(8) Ako, A. M.; Hewitt, I. J.; Mereacre, V.; Clerac, R.; Wernsdorfer, W.;
Anson, C. E.; Powell, A. K. Angew. Chem., Int. Ed. 2006, 45, 4926.
(9) Milios, C. J.; Vinslava, A.; Wernsdorfer, W.; Moggach, S.; Parsons,
S.; Perlepes, S. P.; Christou, G.; Brechin, E. K. J. Am. Chem. Soc.
2007, 129, 2754.
(10) Cano, J.; Cauchy, T.; Ruiz, E.; Milios, C. J.; Stoumpos, C. C.;
Stamatatos, T. C.; Perlepes, S. P.; Christou, G.; Brechin, E. K. J. Chem.
Soc., Dalton Trans. 2008, 234.
(11) Yang, C.-I.; Wernsdorfer, W.; Lee, G.-L.; Tsai, H.-L. J. Am. Chem.
Soc. 2007, 129, 456.
(12) Milios, C. J.; Inglis, R.; Vinslava, A.; Bagai, R.; Wernsdorfer, W.;
Parsons, S.; Perlepes, S. P.; Christou, G.; Brechin, E. K. J. Am. Chem.
Soc. 2007, 129, 12505.
(13) Viciano-Chumillas, M.; Tanase, S.; Mutikainen, I.; Turpeinen, U.; de
Jongh, L. J.; Reedijk, J. Inorg. Chem. 2008, 47, 5919.
Inorg. Chem. 2008, 47, 8610-8612
8610 Inorganic Chemistry, Vol. 47, No. 19, 2008 10.1021/ic801208z CCC: $40.75
 2008 American Chemical Society
Published on Web 09/05/2008

Page 2

of a change in the terminal X-ligand from Br-to I-to
N3-, respectively. These observations highlight the impor-
tance of size and shape upon the crystal packing of these
molecules and their resulting magnetic properties.
Complexes 1-3 crystallize in the trigonal space groups
R3c, R3c, and Rj 3c, respectively, as racemic mixtures of C3-
symmetric chiral molecules. Complexes 4 and 5 do not
exhibit C3symmetry and crystallize in the monoclinic space
groups P21/c and P21/n, respectively. The metallic cores
for all molecules are comprised of a µ3-oxo-centered triangle
of Mn3+ions with two Zn2+ions located above and below
the Mn3 plane. Magnetic exchange interactions between
Mn3+ions are propagated by the central µ3-oxo ion and
through the coordinating oxime. Six µ1,1-azido ligands bridge
the Mn3+ions to the two tetrahedral Zn2+ions, and a terminal
halide/azide binds to the axial coordination site of the Zn2+
ions. Complexes 1-3 crystallize in one molecular orientation
with no solvate molecules, whereas 4 and 5 exhibit multiple
molecular orientations and each cocrystallize with a MeOH
solvate molecule.
Complexes 1 and 2 have large Mn-N-O-Mn torsion
angles of 32.28° and 32.08°, as is also evident by the
puckered phenyl ring orientation (Figure 1, bottom). This
results in S ) 6 spin ground states (vide supra) and arises
from the crystal packing in the noncentrosymmetric R3c
space group. The metal clusters are arranged in columns
along the c axis, with the molecular C3 axis (easy-axis)
coinciding with the c axis. Adjacent molecules along the c
axis exist as alternating stereoisomers in a “head-to-tail”
arrangement, while adjacent molecules in the a-b plane are
identical. [NEt4]+cations fit into the chiral void spaces of
the metal cluster and also exhibit alternating handedness
when viewed along c.
A similar packing arrangement is not possible in 3 because
the [AsPh4]+cation is larger and does not have the same
packing flexibility. Instead, complex 3 crystallizes in the
centrosymmetric space group Rj 3c, where adjacent molecules
along the c axis and a-b plane exist as different stereoiso-
mers in the “head-to-tail” orientation. Inspection of the
structure reveals a reduced Mn-N-O-Mn torsion angle of
11.93° and virtually coplanar phenyl rings (Figure 1, bottom),
as enforced by the presence of bulky [AsPh4]+cations. The
result of this crystalline packing is a low-spin ground state
arising from strongly mixed low-lying spin states (Figure
S4 in the Supporting Information). It is also notable that the
individual-ion Jahn-Teller (JT) cant angles in 3 (5.06°) are
close to those of 1 (8.43°) and 2 (8.09°), indicating that the
oximate ligand geometry more likely determines the nature
of magnetic exchange interactions, not changes in the
direction of the JT elongations.
The magnetostructural effects of terminal-ion substitution
have also been investigated. In spite of the terminal-ion
coordinating to diamagnetic Zn2+ions, there are significant
structural and magnetic effects associated with keeping the
cation the same but changing the terminal ligand. Complexes
1 and 2 have Cl-and Br-as terminal ligands, respectively,
and possess similar trigonal crystal structures and S ) 6 ground
states. Complex 4 contains the larger I-ion and exhibits a
monoclinic crystal packing and smaller torsion angles, resulting
in a strongly mixed low-spin ground state. These observations
may be attributed to the [NEt4]+cation geometry with respect
to the cavity formed by adjacent bridging N3-ions and the
terminal halide. Complexes 1 and 2 have smaller halide groups
and a larger void space for an ethyl fragment to fit, when
compared to 4. Further crystal packing differences are evident
from the lower site symmetry and nonaxially symmetric
orientation of terminal ligands in 4.
A related situation is observed for complex 5, where bent
terminal N3-ligands prevent C3symmetry and require a lower-
symmetryspacegroup.Theresultingmonocliniccrystalpacking
exhibits small Mn-N-O-Mn angles and leads to an S ) 2
ground state (Figure S6 in the Supporting Information).
In addition to the above structural analysis, magnetic
susceptibility data were collected on complexes 1-5 from
300 to 1.8 K and from 0.01 to 5 T. These data were fit via
block diagonalization of the microscopic spin Hamiltonian
schematized by eq 1 and explicitly defined in Figure S1 in
the Supporting Information.14
Hˆ)Hˆexchange+Hˆzfs+HˆZeeman
(1)
These data were fit by means of an uncoupled basis set for
three MnIIIions (125 × 125 Hamiltonian matrix), yielding
best-fit parameters for complex 1 of g ) 1.93, JMn-Mn )
+2.44 K, and DMnIII) -4.76 K; the θ ) 8.43° single-ion
JT cant angles were also taken into account. Zero-field-
splitting (zfs) parameters D and B40for a molecular S ) 6
ground state were then extracted from the eigenvalue
spectrum of the microscopic Hamiltonian for complex 1,
resulting in values of D ) -1.192(9) K and B40) -1.110(3)
× 10-4K (Figure S2 in the Supporting Information). In
comparison, the low-spin complex 3 could not be ap-
proximated by the giant spin Hamiltonian because of low-
(14) Wilson, A.; Lawrence, J.; Yang, E.-C.; Nakano, M.; Hendrickson,
D. N.; Hill, S. Phys. ReV. B 2006, 74, 140403.
Figure 1. (top) Molecular structure of complex 1. (bottom) Space-filling
model for the [Mn3(salox)3O]+magnetic core of complexes 1 and 3, when
viewed in the Mn3IIIplane.
COMMUNICATION
Inorganic Chemistry, Vol. 47, No. 19, 2008 8611

Page 3

lying excited states. The best-fit single-ion parameters for
complex 3 were determined as g ) 1.94, JMn-Mn) -4.07
K, DMnIII) -6.20 K, and θ ) 5.06°. Additional magnetic
data and fits are provided in the Supporting Information.
Alternating current (ac) susceptibility measurements were
taken between 10 and 1000 Hz and from 1.8 to 5 K for
complexes 1-5. Clear peaks in the out-of-phase susceptibility
were observed for 1 and 2 and indicate slow relaxation of a
SMM (Figure S7 in the Supporting Information). Fitting to
the Arrhenius equation resulted in a magnetization reversal
barrier of Ueff) 44 and 46 K for 1 and 2, respectively. No
out-of-phase ac susceptibility signal was observed for
complexes 3-5.
High-frequency electron paramagnetic resonance (HFEPR)
measurements were taken on a single crystal of 1 to directly
probe the transitions between spin states and more accurately
determine the spin Hamiltonian parameters. Very sharp
absorption peaks were observed and indicate a monodisperse
crystalline environment and a very small distribution of local
microenvironments. Furthermore, a single molecular orienta-
tion made it possible to apply the field precisely along the
molecular easy-axis and hard-plane. Variable-temperature
spectra for the hard-plane orientation are given in Figure 3
and confirm a large negative value of the axial zfs parameter
D. Exceptional simulations of the easy-axis and hard-plane
multifrequency data gave the following unique parameter set:
gz) 1.97(2), gx) gy) 1.96(2), D ) -1.16(1) K, and B40
) -7.6(5) × 10-5K. No measurable transverse anisotropy
was detected in these initial studies.
HFEPR experiments on complex 2 yielded spectra and
simulation parameters similar to those of 1, while a broad
transition peak was observed for complex 3. The observance
of a broad peak is consistent with the strongly mixed spin
states in 3.
Magnetization hysteresis measurements were conducted
on a single crystal of 1 using micro-Hall-bar probe mag-
netometry, with the field applied along the molecular easy
axis. The resulting hysteresis loop is shown in Figure 4 and
is characteristic of a SMM, i.e., sharp vertical steps corre-
sponding to quantum tunneling of the magnetization (QTM).
The sharp QTM steps reflect the single molecular orientation,
an absence of solvate molecules, and high crystal quality.
The QTM step positions are in excellent agreement with the
zfs parameters obtained from magnetization fitting and
HFEPR. Similar hysteresis data were found for 2.
A series of closely related Mn3Zn2complexes have been
synthesized and shown to undergo spin-state changes as a
result of changes in the cocrystallizing cation and terminal
ligand. These reactions produce well-formed crystalline
products in high yields (75-90%) and provide an unprec-
edented platform from which to study the origins of
ferromagnetic exchange in Mn3 triangles with chemically
identical magnetic cores. Investigations are underway to
elucidate the synthetic requirements for tuning the magnetic
properties of new structures based on these triangular Mn3
subunits.
Acknowledgment. This work was supported by the
National Science Foundation.
Supporting Information Available: Experimental details, crys-
tallographic information in CIF format, fits of dc susceptibility data,
and ac susceptibility data. This material is available free of charge
via the Internet at http://pubs.acs.org. The atomic coordinates for
these structures (CCDC 692288, 692289, 699692, 699694, 699695)
have been deposited with the Cambridge Crystallographic Data
Centre. The coordinates can be obtained, upon request, from the
Director, Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, U.K.
IC801208Z
Figure 2. Fit of ?MT vs T data for complexes 1 (squares) and 3 (circles),
taken at 0.01, 0.1, and 1 T from 300 to 1.8 K. Fitting parameters are
described in the text.
Figure 3. Hard-plane variable-temperature HFEPR for an oriented single
crystal of 1, taken at 104.1 GHz from 2 to 20 K.
Figure 4. Hysteresis loop measurement for a single crystal of 1. The
magnetization is normalized by the saturation value MS.
COMMUNICATION
8612 InorganicChemistry, Vol. 47, No. 19, 2008