Turning on single-molecule magnet behavior in a linear {Mn3} compound

Article abstract of DOI:10.1021/ic301820w

The synthesis, structure, and magnetic properties are reported for a new manganese compound with a mixed-valent {Mn₃} core arranged in a linear fashion. The previously reported complex 1, [Mn^IV ₃(dpo)₆]·2MeCN, where H₂dpo is (E)-1-hydroxy-1,1-diphenylpropan-2-one oxime, served as a starting point for the isolation of a {Mn₃} compound with an analogous core arrangement through the reaction of Mn(OAc)₂·4H₂O, H ₃oxol ((E)-2,5-dihydroxy-2,5-dimethylhexan-3-one oxime), and NaOH in MeOH and MeCN. By using these reaction conditions, compound 2, Na[Mn ^IV₂Mn^III(Hoxol)₆] _n·MeOH·H₂O, was successfully isolated revealing a central Mn^III ion thereby introducing structural and magnetic anisotropy to the system. The structure of 2 reveals linear trinuclear Mn^IV-Mn^III-Mn^IV units connected through Na ⁺ ions forming a linear one-dimensional coordination polymer. The Jahn-Teller axes of each trinuclear unit are aligned parallel within the same chain and form a 75 angle between the two symmetry related chains. Magnetic susceptibility measurements of 1 and 2 in the temperature range 1.9-300 K reveal that only the reduced compound, 2, is a single-molecule magnet (SMM) largely due to the anisotropy introduced by the Jahn-Teller distortions on the Mn ^III ions, which effectively induce this magnet behavior. Weak antiferromagnetic interactions along the chains through the Na⁺ cations lead to a modulation of the intrinsic properties of the Mn ^IV-Mn^III-Mn^IV SMMs.

Full text of DOI:10.1021/ic301820w

Article
pubs.acs.org/IC
Turning on Single-Molecule Magnet Behavior in a Linear {Mn3}
Compound
†
†
†
†
†
Fatemah Habib, Gabriel Brunet, Francis Loiseau, Thushan Pathmalingam, Tara J. Burchell,
M. Beauchemin,† Wolfgang Wernsdorfer, Rodolphe Cler
,‡
§
ac,* and Muralee Murugesu*
,∥,⊥
,†,‡
Andre
†
́
́
Department of Chemistry, University of Ottawa, 10 Marie-Curie, Ottawa, Canada K1N6N5
Centre for Catalysis Research and Innovation, 30 Marie-Curie, Ottawa, Canada
‡
§
∥
Institut Nee
́ ́
l, CNRS and Universite Joseph Fournier, BP 166, 25 Avenue des Martyrs, 38042 Grenoble Cedex 9, France
CNRS, CRPP, UPR 8641, F-33600 Pessac, France
⊥
Univ. Bordeaux, CRPP, UPR 8641, F-33600 Pessac, France
S Supporting Information
*
ABSTRACT: The synthesis, structure, and magnetic properties are reported for a new
manganese compound with a mixed-valent {Mn } core arranged in a linear fashion. The
3
IV
3
previously reported complex 1, [Mn (dpo) ]·2MeCN, where H dpo is (E)-1-hydroxy-
,1-diphenylpropan-2-one oxime, served as a starting point for the isolation of a {Mn3}
6
2
1
compound with an analogous core arrangement through the reaction of Mn-
OAc) ·4H O, H oxol ((E)-2,5-dihydroxy-2,5-dimethylhexan-3-one oxime), and
(
2
2
3
NaOH in MeOH and MeCN. By using these reaction conditions, compound 2,
Na[Mn Mn (Hoxol) ] ·MeOH·H O, was successfully isolated revealing a central
Mn ion thereby introducing structural and magnetic anisotropy to the system. The
IV
III
2
6 n
2
III
IV
III
IV
+
structure of 2 reveals linear trinuclear Mn −Mn −Mn units connected through Na
ions forming a linear one-dimensional coordination polymer. The Jahn−Teller axes of
each trinuclear unit are aligned parallel within the same chain and form a 75° angle
between the two symmetry related chains. Magnetic susceptibility measurements of 1
and 2 in the temperature range 1.9−300 K reveal that only the reduced compound, 2, is a single-molecule magnet (SMM) largely
III
due to the anisotropy introduced by the Jahn−Teller distortions on the Mn ions, which effectively induce this magnet behavior.
+
Weak antiferromagnetic interactions along the chains through the Na cations lead to a modulation of the intrinsic properties of
IV
III
IV
the Mn −Mn −Mn SMMs.
6
INTRODUCTION
anisotropy, thereby obtaining SMMs with higher barriers.
■
Another, more methodical, approach is to modify high-spin
complexes to induce anisotropy by controlling the oxidation
states of the metal centers. The former strategy has been shown
In recent years, metal oxime chemistry has become an intense
research focus in the field of molecular magnetism as they
provide systems, which can be methodically fine-tuned to form
single-molecule magnets (SMMs), with applications as high
7
to yield SMMs with larger energy barriers, whereas the latter
1
remains quite challenging and is being actively pursued, since a
certain degree of control over structural distortions is highly
desirable.
density storage materials and quantum computing. Lately, it
has been demonstrated that ferromagnetic interactions can be
promoted between Mn ions via oximate ligands.
reported oximate-based {Mn } and {Mn } complexes, a slight
ligand modification has generated a larger Mn−N−O−Mn
torsion angle that in turn induced the parallel alignment of the
spins. Such a careful systematic approach has led to complexes
with large spin ground states (S). When large spins are
combined with uniaxial Ising-like magnetic anisotropy, slow
relaxation of the magnetization can be observed. Molecules
with such magnet-like behavior are termed SMMs. To date,
the use of tunable oxime ligands has led to the largest energy
barrier for a 3d metal ion based SMM. However, more
research is necessary to induce and control magnetic anisotropy
for obtaining complexes with even larger energy barriers. One
approach is to incorporate highly anisotropic lanthanide ions in
transition-metal complexes in order to increase the Ising-type
III
1b,c
In the
In manganese chemistry, magnetic anisotropy can be
3
6
III
obtained through the isolation of complexes with Mn ions
where the magnetic anisotropy arises from the observed Jahn−
8
2
Teller (J−T) distortions. The spin vector naturally aligns along
3
one axis, known as the easy axis, and that corresponds to the
III
III
Mn J−T axis. In such molecules, controlled isolation of Mn
ions in order to increase the Ising-type anisotropy is difficult as
often mixed Mn /Mn /Mn oxidation states are encountered
in these complexes. With this in mind, our aim is to induce
II
III
IV
4
9
5
magnetic anisotropy in a known high-spin manganese complex
by simply stabilizing Mn ions with a +3 oxidation state, while
Received: August 20, 2012
Published: January 23, 2013
©
2013 American Chemical Society
1296
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303

Inorganic Chemistry
Article
increasing the spin ground state of the complex simultaneously.
a sealed tube. The mixture was heated in a wax bath for 24 h at 100 °C.
After cooling to room temperature, the mixture was concentrated
IV
We have recently reported a linear Mn
complex,
3
IV
under reduced pressure and recrystallized from hot hexanes. The
[
Mn 3(dpo) ]·2MeCN, 1, where H dpo is (E)-1-hydroxy-
6 2
,1-diphenylpropan-2-one oxime (Scheme 1), bridged solely by
1
H oxol ligand was isolated as white powder (2.89 g, 55% yield); H
3
1
NMR (DMSO-d , 300 MHz) δ 10.55 (s, 1H), 5.67 (s, 1H), 5.48 (s,
6
13
a
1H), 2.63 (s, 2H), 1.29 (s, 6H), 1.17 (s, 6H); C NMR (DMSO-d ,
5 MHz) δ 161.9, 71.9, 70.3, 36.9, 30.7, 29.9; IR (film, cm ) 3423,
970, 1674, 1601, 1464, 1355, 1179, 1153, 1114, 1042, 963, 895, 864,
72, 612; HRMS (EI): exact mass calculated for C H NO [M −
OH] : 158.1181; found: 158.1200.
Preparation of [Mn 3(H dpo) ]·2MeCN (1). The reaction of
6
Scheme 1
−
1
7
2
6
+
8
16
2
+
IV
2
6
Mn(ClO ) ·6H O (0.25 mmol, 0.091 g) with H dpo (0.5 mmol, 0.121
4
2
2
2
a
g) and Et N (0.25 mmol, 0.035 mL) in 10:10 mL of MeOH/MeCN
H dpo: (E)-1-hydroxy-1,1-diphenylpropan-2-one oxime. H oxol:
3
2
3
(
E)-2,5-dihydroxy-2,5-dimethylhexan-3-one oxime.
gave a dark-brown opaque solution. The filtrate was left undisturbed
for a period of 4 days, after which, dark-brown rectangular crystals
suitable for X-ray crystallography were isolated in 45% yield with
9
10
−1
oxime ligands with an S = / spin ground state. Our initial
respect to Mn. Selected IR (KBr pellet, cm ): 3451 (br), 3074 (m),
2
studies along with electrochemical experiments indicated that
1600 (m), 1501 (m), 1450 (m), 1077 (m), 1026 (s), 1000 (s), 949
(m), 916 (w), 775 (w), 750 (w), 705 (s), 690 (s).
IV
the central Mn ion in the trinuclear complex can be reduced
IV
III
III
Preparation of Na[Mn 2Mn (Hoxol) ] ·(MeOH, H O) (2).
to Mn thus inducing magnetic anisotropy. Consequently, our
6 n
2
Compound 2 was synthesized by the reaction of Mn(OAc) ·4H O
synthetic strategy consisted of employing similar oxime
chelating ligands in order to retain the same superexchange
pathway between the metal centers, thus giving a way to a
comparable structural core. Initial tests revealed that the
addition of reducing agents resulted in rapid decomposition of
2
2
(
0.25 mmol, 0.061 g) with H oxol (0.25 mmol, 0.045 g) and NaOH
3
(1.0 mmol, 0.040 g) in 5 mL of MeOH and 10 mL of MeCN giving a
dark-brown opaque solution similar to complex 1. Crystallization of
the compound was performed over a period of 1 week to provide dark-
brown needle-like crystals suitable for X-ray crystallography with a
IV
the metal complex. Other means of reducing the central Mn
−1
yield of 64% with respect to Mn. Selected IR (KBr pellet, cm ): 3440
br), 2975 (m), 2924 (m), 1605 (w), 1465 (w), 1350 (m), 1155 (s),
118 (s), 1056 (s), 974 (s), 890 (w), 865 (w). Anal. Calcd for
C H Mn N NaO : C, 46.98%; H, 7.41%; N, 6.85%. Found: C,
ion, such as cyclic voltammetry, were investigated; however, no
trinuclear complex could be isolated. Herein, we report the
synthesis, structure, and magnetic properties of an analogous
(
1
IV
III
48 90
3
6
18
linear {Mn } compound, Na[Mn Mn (Hoxol) ] ·Me-
3
2
6 n
4
6.84%; H, 7.35%; N, 6.81%.
OH·H O, 2, where H oxol is the (E)-2,5-dihydroxy-2,5-
dimethylhexan-3-one oxime ligand (Scheme 1), with a central
Mn ion. Our strategy proved successful in inducing magnetic
2
3
Physical Measurements. X-ray crystallographic data were
collected on single dark-brown crystals mounted on a glass fiber for
complexes 1 and 2 (Figure 1). Unit cell measurements and intensity
data collections were performed on a Bruker-AXS SMART 1 k CCD
and Bruker APEX II diffractometer for 1 and 2, respectively, using
graphite monochromatized Mo Kα radiation (λ = 0.71073 Å) for 1
and Cu Kα radiation (λ = 1.54178 Å) for 2. The data reduction
included a correction for Lorentz and polarization effects, with an
applied multiscan absorption correction (SADABS). The reflection
III
anisotropy in an oxime bridged linear high-spin manganese
unit.
EXPERIMENTAL SECTION
General Considerations. All chemical reagents and solvents used
in these syntheses were obtained commercially and without further
■
purification. “Caution! Due to the high energy content of NH OH,
2
data was consistent with triclinic P-1 and monoclinic P2 /c systems for
1
appropriate care should be taken when conducting these experiments. The
1
and 2, respectively. The crystal structure was solved and refined
hydroxylamine concentration should not be increased beyond 5−10 wt %
19
using the SHELXTL program suite. Direct methods yielded all non-
hydrogen atoms that were refined with anisotropic thermal parameters.
All hydrogen atom positions were calculated geometrically and were
riding on their respective atoms. For compound 2, the option squeeze
was used to correct the data for the presence of disordered solvent
(
i.e., the typical reaction conditions), and appropriate safety controls should
be performed before scaling up this chemistry above the gram scale,
especially at very high temperatures.”
Synthesis of (E)-1-Hydroxy-1,1-diphenylpropan-2-one
Oxime (H dpo). An oven-dried microwave tube (10 mL), equipped
2
molecules (H O, MeOH). Crystallographic data for 1 and 2 are
with a magnetic stir bar and a rubber septum, was purged with an
argon balloon for 5 min. 1,1-Diphenylprop-2-yn-1-ol (1.25 g, 6.00
mmol), aqueous hydroxylamine (920 μL of a 50 wt % solution, 15.0
mmol), and isopropanol (freshly distilled, 6 mL) were added to the
reaction vessel while keeping a constant flow of argon. The rubber
septum was quickly replaced by a microwave aluminum cap. The
mixture was heated in a CEM microwave for 4 h at 140 °C. After
cooling the mixture to room temperature, the crude was concentrated
under reduced pressure and purified through column chromatography
2
presented in Table S1 of the Supporting Information.
IR analyses were obtained by using a Nicolet Nexus 550 FT-IR
−1
spectrometer in the 4000−650 cm range. The spectra were obtained
by preparing KBr pellets.
NMR spectroscopic analyses were conducted on a Bruker Avance
300 MHz spectrometer with a 5 mm autotuning broadband probe with
Z gradient.
The magnetic susceptibility measurements were obtained using a
Quantum Design superconducting quantum interference device
(
7.5% EtOAc in toluene). The H dpo ligand was isolated as a white
2
1
powder (1.05 g, 78% yield); H NMR (DMSO-d , 300 MHz) δ 10.90
6
(SQUID) magnetometer MPMS-XL7 that works between 1.8 and
(
s, 1H), 7.31−7.30 (m, 8H), 7.28−7.18 (m, 2H), 6.41 (s, 1H), 1.81 (s,
3
00 K for direct current (dc) applied fields ranging from −7 to 7 T.
3
1
1
H); 13C NMR (DMSO-d , 75 MHz) δ 159.6, 145.1, 127.4, 127.3,
6
Measurements were performed on polycrystalline samples of 4.58 mg
for 1 and 31.3 mg for 2. Alternating current (ac) susceptibility
measurements were performed under an oscillating ac field of 3 Oe
and ac frequencies that ranged from 10 to 1500 Hz. Ferromagnetic
impurities that were absent in both samples were investigated by
collecting magnetization data at 100 K. All magnetic data were
corrected for the sample holder as well as diamagnetic contributions.
−1
26.5, 80.9, 12.0; IR (film, cm ): 3332, 1659, 1602, 1488, 1443, 1374,
051, 1013, 968, 880, 758, 728, 694; HRMS (EI): exact mass
+
+
calculated for C H NO [M − OH] : 224.1070; found: 224.1053.
15
14
Synthesis of (E)-2,5-Dihydroxy-2,5-dimethylhexan-3-one
Oxime (H oxol). 2,5-Dimethylhex-3-yne-2,5-diol (4.27 g, 30.0
3
mmol), aqueous hydroxylamine (2.76 mL of a 50 wt % solution,
45.1 mmol), and isopropanol (freshly distilled, 30 mL) were added to
1
297
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303

Inorganic Chemistry
Article
attempts included replacing the phenyl groups of H dpo with
2
diethyl groups, which did not yield the desired complex. Other
strategies included the addition of a moderate strength reducing
agent, such as sodium borohydride, which resulted in complex
decomposition as well as reduction through cyclic voltammetry.
However, the reduced complex could not be isolated
preventing further characterization. Consequently, the reaction
conditions were varied in terms of ligand, metal ion source,
base, and solvent.
Compound 2 was obtained by changing the base from Et N
3
to NaOH as well as the metal ion source from Mn(ClO ) to
4
2
Mn(OAc) . Crystallization of 2 occurred from the mother
2
liquor in a period of 1 week resulting in small needle-like
crystals suitable for single crystal X-ray diffraction. It is
noteworthy that an excess of base was used; however, only
two hydroxy groups were deprotonated leaving the remaining
hydroxy group to form hydrogen bonds within the coordina-
III
tion sphere of the central Mn ion. Furthermore, the remaining
base in the reaction allowed for the slow oxidation of the
manganese atoms. Its dual role as a base, as well as a source of
+
the bridging Na ions linking the trinuclear units (Figure 1,
bottom; and Figure 2), was confirmed when other potential
Figure 1. (top) Molecular structure of complex 1. Symmetry
equivalent labels are denoted by an additional “a” in the label
(
Mn1a). (bottom) Structure of the trinuclear unit in compound 2
revealing the J−T distortion (black bond) on the central Mn atom.
IV
III
Color code: green (Mn ), purple (Mn ), red (O), blue (N), gray
Figure 2. (top) Core structure of the linear trinuclear complex 1
showing the triple oximate superexchange pathways between the
manganese centers. (bottom) Core structure of 2 showing the
trinuclear unit bridged by Na atoms with the J−T distortion on the
central Mn(2) atom indicated in black bonds. Color code: green
(
C), black (H). Most H atoms are omitted for clarity.
RESULTS AND DISCUSSION
Syntheses. Employing polydentate ligands such as H dpo
■
IV
III
(Mn ), purple (Mn ), red (O), blue (N), yellow (Na).
2
and H oxol is essential for the synthesis of multinuclear
3
complexes due to their ability to coordinate multiple metal ions
simultaneously. The coordination of the ligands to metal ions is
promoted by the deprotonation of the hydroxyl and oximate
sodium sources, such as sodium chloride, were added but no
crystallized product was observed. Moreover, the sodium ions
are a key contributing factor in the formation of 2 when
forming a stable one-dimensional coordination polymer.
Structural Analysis. The structural analysis of both
groups by a base, such as NaOH or Et N, forming stable five- or
3
six-membered coordination rings. The previously published
1
0
complex, 1, was synthesized through the reaction of
compounds reveals a trinuclear {Mn } core with a triple
3
Mn(ClO ) with H dpo and Et N in MeOH and MeCN
yielding dark-brown rectangular crystals (Figure 1, left). The
oximate superexchange pathway between the manganese
centers as shown in Figures 1 and 2. Complex 1 crystallizes
in the triclinic P-1 space group while 2 crystallizes in
4
2
2
3
resulting complex consisted of a trinuclear {Mn } unit where all
3
Mn ions were in the +4 oxidation state (Figure 1, top). In order
to introduce magnetic anisotropy and hence potentially induce
monoclinic P2 /c space group. Compound 1 exhibits a
1
centrosymmetric geometry in which the three Mn centers are
linked solely via oximate groups with a 180° Mn(1)−Mn(2)−
Mn(1a) angle while 2, with the reduction of its central
manganese atom (Mn2), is no longer centrosymmetric nor
perfectly linear (Figure 3). The angle between the three metal
centers, Mn(1)−Mn(2)−Mn(3), was determined to be 177°.
In both complexes 1 and 2, the Mn oxidation states are
confirmed by bond-valence sum (BVS) calculations (Table 1)
and charge considerations. Furthermore, in compound 2, the
IV
SMM behavior, we set out to reduce the central Mn ion to
III
Mn that was shown to be accessible through density
10
functional theory calculations.
Initially, the synthesis of a reduced analogue of 1 consisted of
employing a less bulky ligand. This approach would increase
the flexibility of the complex and possibly create an opening for
potentially distorting the coordination sphere of the central Mn
ion leading to elongation/compression of one axis. These
1
298
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303

Inorganic Chemistry
Article
Figure 3. Na ions in 2 linking the trinuclear {Mn } units forming a
3
IV
III
one-dimensional structure. Color code: green (Mn ), purple (Mn ),
red (O), blue (N), yellow (Na), gray (C).
Table 1. Bond Valence Sum Calculations with Suggested
Oxidation States for 1 and 2
terminal Mn ion
central Mn2
ox
ox
MnII
MnIII
MnIV
MnII
MnIII
MnIV
Figure 4. Molecular structure of the trinuclear unit in 2 illustrating the
compd
state
state
intramolecular hydrogen bonding as well as the J−T axis on Mn2
1
2
4.26
4.36
4.00
4.10
4.02
4.13
4
4
4.03
3.12
3.69
3.87
4
3
III
(
central Mn ion) in black. The shorter H bonds are indicated in red
2.85
2.99
and involve the participation of O3 and O18, which form the J−T axis.
IV
III
Color code: green (Mn ), purple (Mn ), red (O), blue (N), gray
C).
oxidation of the central manganese atom from Mn to MnIII is
also confirmed by the elongation of the Mn(2)−O(3) and
Mn(2)−O(18) bonds (Table 2) indicating a J−T distortion
II
(
arranged in a linear or quasi-linear fashion, respectively, where
Table 2. Selected Bond Distances (Å) and Torsion Angles
2−
the metal ions are bridged exclusively by six dpo ligands for 1
(deg) for 1 and 2
2−
and six Hoxol ligands for 2 (Figure 2). Furthermore, the
central metal ion (Mn2) is coordinated to six oxygen atoms
while the peripheral Mn ions (Mn(1) and Mn(1a) for 1;
Mn(1) and Mn(3) for 2) are coordinated to three nitrogen
atoms and three oxygen atoms. Selected bond distances and
angles are presented in Table 2. In 1, the average Mn(1)−O,
Mn(2)−O, and Mn(1)−N distances are 1.83, 1.91, and 1.99 Å,
respectively, while in 2, the average Mn(1)/Mn(3)−O, Mn(1)/
Complex 1
Mn(1)···Mn(2)
Mn(1)···Mn(1a)
Mn(1)−O(1)
Mn(1)−O(3)
Mn(1)−O(5)
Mn(1)−N(1)
Mn(1)−N(2)
3.54
7.08
Mn(1)−N(3)
Mn(2)−O(2)
1.826(7) Mn(2)−O(4)
1.997(9)
1.911(6)
1.910(7)
1.915(6)
53.5(2)
51.7(1)
48.8(2)
1.834(8) Mn(2)−O(6)
1.826(8) Mn(1)−N(1)−O(2)−Mn(2)
1.996(8) Mn(1)−N(2)−O(4)−Mn(2)
1.989(8) Mn(1)−N(3)−O(6)−Mn(2)
Complex 2
Mn(3)−N, Mn(2)−O
, and Mn(2)−O
distances are
axial
equatorial
1.85, 1.98, 1.97, and 2.15 Å, respectively.
Mn(1)···Mn(2)
Mn(2)···Mn(3)
Mn(1)−O(1)
Mn(1)−O(4)
Mn(1)−O(7)
Mn(1)−N(1)
Mn(1)−N(2)
Mn(1)−N(3)
Mn(2)−O(3)
Mn(2)−O(6)
Mn(2)−O(9)
3.548
3.535
Mn(2)−O(18)
Mn(3)−N(4)
2.140(2)
1.981(3)
1.988(3)
1.981(3)
1.852(2)
1.852(2)
1.852(2)
44.7(3)
Other important distances involve those separating the Mn
centers in the compounds. In complex 1, the Mn(1)···Mn(2)
and Mn(1)···Mn(1a) distances of 3.54 and 7.08 Å, respectively,
are comparable to those in 2 with Mn(1)···Mn(2),
Mn(2)···Mn(3), and Mn(1)···Mn(3) distances being 3.54,
3.53, and 7.07 Å, respectively (Figure 3). The closest interunit
distances in 2 are between Mn(1)···Mn(1) through the Na
bridges with distances of 5.85 and 5.92 Å for Na(1) and Na(2),
respectively, occurring along the chain as opposed to those in 1
that occur between well-isolated complexes with separations of
12.11, 11.58, and 11.37 Å along the a axis (Figure 5), b axis
1.842(2) Mn(3)−N(5)
1.847(2) Mn(3)−N(6)
1.852(2) Mn(3)−O(10)
1.979(3) Mn(3)−O(13)
1.977(3) Mn(3)−O(16)
1.973(3) Mn(1)−N(1)−O(3)−Mn(2)
2.156(2) Mn(1)−N(2)−O(6)−Mn(2)
1.964(2) Mn(1)−N(3)−O(9)−Mn(2)
48.8(3)
50.6(3)
1.982(2) Mn(2)−O(12)−N(4)−
47.1(3)
Mn(3)
(
Supporting Information, Figure S1), and c axis (Supporting
Mn(2)−O(12)
Mn(2)−O(15)
1.929(2) Mn(2)−O(15)−N(5)−
48.7(3)
47.5(3)
Information, Figure S2), respectively. This structural feature
renders the magnetic interactions between trinuclear units in 2
more relevant and can potentially influence the magnetic
properties of the compound.
Mn(3)
1.991(2) Mn(2)−O(18)−N(6)−
Mn(3)
The presence of bulky groups on the ligands induces large
torsion angles, which in turn influence the nature and strength
of magnetic interactions between the metal ions. The average
torsion angle between the terminal and the central manganese
ions (Mn(1)−N−O−Mn(2)) was calculated to be 51.3° for 1
and 47.8° for 2 for Mn(1)−N−O−Mn(2) and Mn(2)−N−O−
Mn(3), which is much higher than for other trinuclear oximate-
characteristic of MnIII ions. It is noteworthy that the redox
properties of the central Mn metal ion toward the +3 state are
certainly favored by the ligand choice. Additionally, this +3
oxidation state may have also been facilitated by the
intramolecular hydrogen bonds, which help by stabilizing the
axially elongated Mn2−O3 and Mn2−O18 bonds as illustrated
in Figure 4. For 1 and 2, the octahedral Mn centers are
11
bridged complexes (0−20°).
1
299
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303

Inorganic Chemistry
Article
Figure 6. Packing arrangement of 2 along the c axis showing chains
aligned in an antiparallel fashion highlighting the J−T axes in black.
Figure 5. Packing arrangement of 1 along the a axis showing well-
isolated complexes.
Moreover, the magnetic anisotropy of the {Mn } system in 2
3
is derived from a prominent J−T elongation in the axial
III
positions of the central Mn ions. It is noteworthy that these
axial oxygen atoms are participating in intramolecular hydrogen
bonds with the nearby protonated hydroxyl groups as
illustrated in Figure 4. The oxygen atoms that form the J−T
III
axes on Mn ions are shown to participate in shorter H bonds
than the equatorial oxygen atoms (distances are indicated on
the structure in Figure 4). The J−T distortion induces
structural and consequently magnetic anisotropy in the system
satisfying the second requirement of SMMs and potentially
promoting SMM behavior in the trinuclear complex. It is worth
mentioning that, to our knowledge, there are only two other
Figure 7. Plot of χT vs T for compounds 1 and 2 at 1000 Oe (with χ
defined as molar magnetic susceptibility equal to M/H per mole of
complex). The solid lines are the best fits obtained with the
Heisenberg trinuclear model (see text).
previously reported linear trinuclear {Mn } complexes that
3
1
2
behave as SMMs. The packing diagram for 1 along the a axis
shows well-isolated trinuclear complexes that are organized in a
parallel fashion (Figure 5). For 2, it can be seen from the
packing arrangement (Figure 6) that the trinuclear units are
3
−1
product at room temperature is 6.7 cm ·K·mol for 1 and 7.2
3
−1
cm ·K·mol for 2, which are comparable to the theoretical
+
3
−1
linked via Na ions into chains along the c axis (this can also be
values (for g = 2) of 5.625 cm ·K·mol for 1 and 6.75
3
−1
seen in the packing diagrams along the a and b axes in the
Supporting Information, Figures S3 and S4, respectively). The
chains are arranged in an antiparallel fashion with the closest
interchain distance being 3.44 Å between the two carbon atoms
of the dangling ligand arms. The J−T elongations within each
chain are aligned parallel to each other; however, they form a
cm ·K·mol for 2. As the temperature decreases, the χT values
increase steadily up to 100 K and then rapidly reaching maxima
3
−1
3
−1
of 13.0 cm ·K·mol at 7.5 K for 1 and 12.5 cm ·K·mol at 19
K for 2 indicating dominant intramolecular ferromagnetic
interactions. The maximum value of χT for 1 is in accordance
with a spin ground state of S = / (theoretical value of 12.375
cm ·K·mol when g = 2). For 2, the maximum of the χT value
is slightly lower but also in agreement (vide infra) with the
9
T
2
3
−1
7
5° angle (noted α thereafter) to neighboring J−T axes in
III
IV
other chains (Figure 6). While the Mn ···Mn distance in 2 is
IV
IV
3
−1
comparable to Mn ···Mn distances in 1, the shortest metal−
theoretical value (15.0 cm ·K·mol when g = 2) expected for
IV
metal distance between intrachain Mn atoms occurs through
an S = 5 spin ground state.
T
the sodium ions (5.85 Å for Na(1) and 5.92 Å for Na(2),
The final decrease at low temperature of the χT product
suggests the presence of intermolecular antiferromagnetic
interactions and/or the presence of magnetic anisotropy. It is
evident that the main magnetic interactions between the metal
centers occur through superexchange pathways formed by the
Figure 3) and is significantly shorter than the distance
IV
III
separating two Mn atoms through a Mn ion (7.08 Å,
Figure 3).
Magnetic Properties. The dc magnetic susceptibility
measurements for both 1 and 2 were performed between 1.9
and 300 K under an applied dc field of 1000 Oe. The χT versus
T plots of 1 and 2 are presented in Figure 7, where the χT
oxime groups from the dpo2 (1) and Hoxol (2) ligands
(Figure 2). Fitting of the magnetic susceptibility data for 1 was
carried out in order to determine the sign and magnitude of the
−
2−
1
300
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303

Inorganic Chemistry
Article
intramolecular interactions between MnIV ions. The following
exchange interaction now between MnIV and MnIII ions while
⃗
⃗
⃗
⃗
Heisenberg Hamiltonian was employed: H = −2J {S S + S S }
the intercomplex magnetic interaction, J′, is dominated by the
1
2
1 3
IV
IV
where J is the intramolecular magnetic exchange interaction
⃗
Mn −Mn interaction via sodium cations (Scheme 2), S is
i
III
IV
IV
⃗
between a terminal Mn ion and a central Mn ion, S is the
the spin operator for each Mn with S = 2 (Mn ) and S = S =
1 2 3
i
3
IV
3
IV
spin operator for each Mn with S = / (Mn ) as seen in
/ (Mn ) as seen in Scheme 2, bottom. The best fit to the
2
2
Scheme 2, top. The low field magnetic susceptibility was
experimental data yielded the following parameters: J/k =
B
+
7.1(2) K, g = 2.07(3), and J′/k = −0.09(2) K. In comparison
B
Scheme 2. Schematic Diagram Showing the Intra- and
Intermolecular Magnetic Interactions between Mn ions in 1
to that of 1, it is evident that the antiferromagnetic
intermolecular interaction is stronger in 2 which explains the
somewhat low χT value at 19 K relative to the theoretical S =
(top) and 2 (bottom)
T
5
value. It should be noted that the J′ value should be taken
with caution as it also contains phenomenologically the
contribution from the magnetic anisotropy. By using the
calculated J value, the energies of the excited states were
determined resulting in a spin ground state of S = 5 with the
T
first (S = 4) and second (S = 3) excited states being 27.3 and
5
4.9 K higher in energy, respectively.
The M versus H and M versus H/T plots for 2 (Figure 8)
show nonsaturation of the magnetization at low temperature
and high field (1.9 K and 7 T) as well as nonsuperposition of
the curves on a single master curve indicating the presence of
1
3
calculated by applying the van Vleck equation to Kambe’s
magnetic anisotropy. Attempts to fit this data with an S
T
= 5
14
2
2
2
vector coupling scheme, and the experimental data were fitted
yielding the following parameters: J/k = +11.5(1) K, g =
macro-spin approximation (H = DST,z + E(ST,x − ST,y )) using
16
the program Magnet were fruitless as intermolecular exchange
interactions are present between trinuclear units. The signature
of these intercomplex interactions along the chain is indeed
very clear in the M versus H data at 1.9 K that display a typical
“S” shape below 1.0 T (Figure 8).
B
2
.06(3), and J′/k = −0.06(1) K; where J′ is the intercomplex
B
magnetic interaction introduced in the frame of the mean field
1
0,11
theory.
This data indicates ferromagnetic interactions
between metal centers and thus a spin ground state of S =
T
9
7
5
/2
with the first (S = / ) and second (S = / ) excited states
In order to investigate the effect of added anisotropy to the
2
2
10
III
being 34.2 and 69.7 K higher in energy, respectively. For such
linear, triply bridged oxime systems with torsion angles such as
in 1 and 2, ferromagnetic interactions between metal centers
system through the presence of Mn ions in 2, ac magnetic
susceptibility measurements were performed in the temperature
range 1.8−5 K and frequency range 10−1500 Hz. The out-of-
phase susceptibility χ″ versus T plot is shown in Figure 9, top
(χ′ vs T plot is shown in the inset), and displays a tail of a
relaxation peak evident at low temperature. This feature is
15
and thus high-spin ground states are commonly observed.
The field dependence of the magnetization plot at varying
temperatures for 1 is shown in Figure S5 of the Supporting
Information. The clear saturation of the curve at 1.8 K indicates
the absence of magnetic anisotropy, with a saturation value of
.6 μ confirming a spin ground state of S = / . The M versus
indicative of slow relaxation of the magnetization and thus of
IV
potential SMM behavior that is absent in the parent {Mn
}
3
9
9
compound 1. This difference of properties can be directly
B
T
2
9
III
H/T data was also fitted to a Brillouin function with S = /
attributed to the anisotropy introduced by the Mn ions in 2.
T
2
and g = 2.11(3) further confirming the lack of anisotropy. This
was expected as all three Mn ions in 1 are in the +4 oxidation
state that is known to be isotropic preventing the complex from
behaving as an SMM.
For 2, the same Heisenberg Hamiltonian was employed with
the following modifications: J is the intramolecular magnetic
In order to confirm this behavior, single-crystal dc magnet-
ization measurements were carried out on a micro-SQUID
17
apparatus at temperatures below 5 K (Figure 9, bottom, and
Supporting Information, Figure S6). At temperatures below 1 K
for an applied dc field oriented in the easy magnetic direction of
a single crystal, M versus H hysteresis loops are observed that
Figure 8. Magnetization vs field measurements at 1.9, 3, 5, and 8 K for a polycrystalline sample of 2 plotted as M vs H (left) and M vs H/T (right).
The nonsaturation of the magnetization curves at high fields indicates the presence of magnetic anisotropy in the system. The “S” shape M vs H plot
highlights the presence of weak antiferromagnetic interactions between trinuclear complexes.
1
301
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303

Inorganic Chemistry
Article
crystal (found along the b axis, i.e., at an α/2 angle from the J−
T axes of the two different chains, Figure 6), the two symmetry
related chains experience the same effective dc field: Heff = H
cos(α/2). Therefore J′/k is accurately estimated at −0.039(5)
B
K, in good agreement with the rough mean-field value
estimated from the χT versus T fit (vide supra).
The absence of full peaks with maxima in the χ″ versus T
plot above 1.8 K and the complicated exchange-bias relaxation
see in Figure 9 prevent the estimation of the SMM energy
barrier. It is noteworthy that performing the ac measurements
under a weak dc field above 1.8 K did not shortcut the QTM
significantly to allow the observation of the maximum of the
frequency-dependent ac peaks.
CONCLUSION
■
Through careful synthetic strategy, two analogous linear {Mn }
3
compounds were isolated with high-spin ground states and very
different magnetic properties. In both compounds, the metal
centers are coupled ferromagnetically through triple-oxime
9
superexchange pathways yielding S = / and S = 5 high-spin
T
2
T
compounds. By employing the oxol ligand, we were able to
promote the formation and isolation of a previously predicted
IV
III
IV
III
yet elusive Mn −Mn −Mn unit. The incorporation of Mn
ions into the system introduces magnetic anisotropy that
subsequently “turns on” the SMM behavior. The latter strategy
provides a new avenue/methodology to fine-tuning a simple
system in order to control the switching on of their magnet-like
behavior.
Figure 9. (top) Temperature dependence of the out-of-phase ac
magnetic susceptibility plotted as χ″ vs T for 2 in zero applied dc field
with an ac oscillating field of 3 Oe at the indicated frequencies; inset
shows the corresponding in-phase susceptibility, χ′, vs T for 2.
ASSOCIATED CONTENT
■
*
S
Supporting Information
(
bottom) Plot of M/M and dM/dH vs H obtained from oriented
s
Crystallographic data (including CIF file), packing diagrams,
and additional magnetic measurement plots of 1 and 2. This
material is available free of charge via the Internet at http://
pubs.acs.org.
single-crystal (in the easy magnetic direction) measurements at 30 mK
and a field sweep rate of 0.14 T/s for compound 2.
confirm the slow relaxation of the magnetization detected in 2
by ac susceptibility (Figure 9, top). There are three major steps
AUTHOR INFORMATION
■
*
(
in the M/M versus H plot at H = 0 and ±0.73 T (±H in
s
1
Corresponding Author
Phone: +33 5 56 84 56 50 (R.C.), 613-562-5800 ext 2733
M.M.). Fax: +33 5 56 84 56 00 (R.C.), 613-562-5170 (M.M.).
E-mail: clerac@crpp-bordeaux.cnrs.fr (R.C.), m.murugesu@
uottawa.ca (M.M.).
Figure 9, bottom) that are better seen on the dM/dH versus H
plot. This type of behavior has been previously described in
1
8
detail and is typical of a chain of weakly antiferromagnetically
coupled SMMs. Therefore, in order to understand the M/Ms
versus H behavior of 2 shown in Figure 9, bottom, the whole
chain as trinuclear SMM units interacting through Na bridges
must be considered. When a strong negative dc field is applied,
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
the {Mn } magnetic moments are all aligned along the
3
direction of the applied field. As the applied field reaches ±
H corresponding to the intrachain exchange, the spins of the
1
Notes
trinuclear unit experience an effective zero field and can
therefore flip by quantum tunneling of the magnetization
The authors declare no competing financial interest.
(
QTM). Between H = −H and +H , the intrachain magnetic
1
1
ACKNOWLEDGMENTS
interactions are dominating, but some of the trinuclear unit
moments in the chain are trapped in a state where they are not
regularly antiparallel, creating magnetic defects (in an effective
zero exchange field) along the chain. The minor steps observed
at H = ± 0.36 T, which are labeled as ± H /2, correspond to
the expected quantum relaxation seen for open-ended chains of
■
We thank the University of Ottawa, the Canada Foundation for
Innovation (CFI), FFCR, NSERC (Discovery and RTI grants),
the Conseil Reg
Aquitaine (COMET Project), the Universite
́
ional d′Aquitaine, GIS Advanced Materials in
́
de Bordeaux, the
1
CNRS, and the ANR (NT09_469563, AC-MAGnets project)
for financial support. We also thank Dr. Ilia Korobkov
18
a finite size where the last spin flips. From the H value, the
1
intrachain interactions, J′, can be estimated equalizing the
(University of Ottawa) and Francine Bel
(Universite de Montreal) for their help with the single-crystal
X-ray crystallography.
́ ́
anger-Gariepy
2
T
18
Zeeman and exchange energies: gμ H S = 4|J′|S . As the
́
́
B
eff
T
measurements have been performed in the easy direction of the
1
302
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303

Inorganic Chemistry
Article
(
19) (a) SHELXTL, v. 6.12; Bruker Analytical X-ray Systems Inc.:
REFERENCES
■
Madison, WI, 2001; pp 53719−1173. (b) Sheldrick, G. M. Acta
Crystallogr. 2008, A64, 112.
(
1) (a) Zaleski, C. M.; Depperman, E. C.; Kampf, J. W.; Kirk, M. L.;
Pecoraro, V. L. Angew. Chem., Int. Ed. 2004, 43, 3912. (b) Milios, C. J.;
Vinslava, A.; Whittaker, A. G.; Parsons, S.; Wernsdorfer, W.; Christou,
G.; Perlepes, S. P.; Brechin, E. K. Inorg. Chem. 2006, 45, 5272.
(
c) 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.
(
2) Milios, C. J.; Inglis, R.; Bagai, R.; Wernsdorfer, W.; Collins, A.;
Moggach, S.; Parsons, S.; Perlepes, S. P.; Christou, G.; Brechin, E. K.
Chem. Commun. 2007, 3476.
(
3) (a) Ferbinteanu, M.; Miyasaka, H.; Wernsdorfer, W.; Nakata, K.;
Sugiura, K.-i.; Yamashita, M.; Coulon, C.; Clerac, R. J. Am. Chem. Soc.
005, 127, 3090. (b) Scott, R. T. W.; Parsons, S.; Murugesu, M.;
Wernsdorfer, W.; Christou, G.; Brechin, E. K. Angew. Chem., Int. Ed.
005, 44, 6540. (c) Wang, W.-G.; Zhou, A.-J.; Zhang, W.-X.; Tong,
́
2
2
M.-L.; Chen, X.-M.; Nakano, M.; Beedle, C. C.; Hendrickson, D. N. J.
Am. Chem. Soc. 2007, 129, 1014.
(
4) (a) Christou, G.; Gatteschi, D.; Hendrickson, D. N.; Sessoli, R.
MRS Bull. 2000, 25, 66. (b) Aromi, G.; Brechin, E. K. Struct. Bonding
Berlin, Ger.) 2006, 122, 1.
5) 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.
(
6) (a) Sessoli, R.; Powell, A. K. Chem. Rev. 2009, 253, 2328 (and
references therein). (b) Hołynska, M.; Premuzic, D.; Jeon, I.-R.;
Wernsdorfer, W.; Cler
ac, R.; Dehnen, S. Chem.Eur. J. 2011, 17,
605.
7) Stamatatos, T. C.; Teat, S. J.; Wernsdorfer, W.; Christou, G.
Angew. Chem., Int. Ed. 2009, 48, 521.
8) Glaser, T. Chem. Commun. 2011, 47, 116 (and references
therein).
́
9
(
(
(
9) (a) Zaleski, C. M.; Weng, T.-C.; Dendrinou-Samara, C.; Alexiou,
M.; Kanakaraki, P.; Hsieh, W.-Y.; Kampf, J.; Penner-Hahn, J. E.;
Pecoraro, V. L.; Kessissoglou, D. P. Inorg. Chem. 2008, 47, 6127.
(
1
b) Mukhopadhyay, S.; Armstrong, W. H. J. Am. Chem. Soc. 2003, 125,
3010.
10) Pathmalingam, T.; Gorelsky, S. I.; Burchell, T. J.; Bed
Beauchemin, A. M.; Clerac, R.; Murugesu, M. Chem. Commun. 2008,
782.
11) (a) Price, D. J.; Batten, S. R.; Berry, K. J.; Moubaraki, B.;
(
́
ard, A.-C.;
́
2
(
Murray, K. S. Polyhedron 2003, 22, 165. (b) Johnson, B. F. G.; Sieker,
A.; Blake, A. J.; Winpenny, R. E. P. J. Chem. Soc., Chem. Commun.
1
993, 1345. (c) Birkelbach, F.; Florke, U.; Haupt, H. J.; Butzlaff, C.;
Trautwein, A. X.; Wieghardt, K.; Chaudhuri, P. Inorg. Chem. 1998, 37,
000.
12) (a) Scott, R. T. W.; Parsons, S.; Murugesu, M.; Wernsdorfer,
W.; Christou, G.; Brechin, E. K. Chem. Commun. 2005, 2083.
b) Zhou, C.-L.; Wang, Z.-M.; Wang, B.-W.; Gao, S. Dalton Trans.
012, 41, 13620.
13) van Vleck, J. H. The Theory of Electric and Magnetic Susceptibility;
Oxford University Press: Oxford, U.K., 1932.
2
(
(
2
(
(
14) Kambe, K. J. Phys. Soc. Jpn. 1950, 5, 48.
(
15) (a) Birkelbach, F.; Florke, U.; Haupt, H.-J.; Butzlaff, C.;
̈
Trautwein, A. X.; Wieghardt, K.; Chaudhuri, P. Inorg. Chem. 1998, 37,
000. (b) Alexiou, M.; Dendrinou-Samara, C.; Karagianni, A.; Biswas,
2
S.; Zaleski, C. M.; Kampf, J.; Yoder, D.; Penner-Hahn, J. E.; Pecoraro,
V. L.; Kessissoglou, D. P. Inorg. Chem. 2003, 42, 2185. (c) Price, D. J.;
Batten, S. R.; Moubaraki, B.; Murray, K. S. Polyhedron 2003, 22, 165.
(
d) Audhya, A.; Maity, M.; Abtab, S. M. T.; Mathonier
Clerac, R. Polyhedron 2012, 33, 353. (e) Mukherjee, C.; Weyhermu
T.; Wieghardt, K.; Chaudhuri, P. Dalton Trans. 2006, 2169.
f) Pavlishchuk, V.; Birkelbach, F.; Weyhermuller, T.; Wieghardt, K.;
Chaudhuri, P. Inorg. Chem. 2002, 41, 4405.
16) Davidson, E. R. Magnet; Indiana University: Bloomington, IN,
999.
̀
e, C.; Kalisz, M.;
́
̈
ller,
(
̈
(
1
(
17) Wernsdorfer, W. Supercond. Sci. Technol. 2009, 22, 064013.
(
18) Lecren, L.; Wernsdorfer, W.; Li, Y.-G.; Vindigni, A.; Miyasaka,
́
H.; Clerac, R. J. Am. Chem. Soc. 2007, 129, 5045.
1
303
dx.doi.org/10.1021/ic301820w | Inorg. Chem. 2013, 52, 1296−1303