Ferromagnetic dinuclear mixed-valence Mn(II)/Mn(III) complexes: Building blocks for the higher nuclearity complexes. Structure, magnetic properties, and density functional theory calculations

Article abstract of DOI:10.1021/ic302731z

A series of six mixed-valence Mn(II)/Mn(III) dinuclear complexes were synthesized and characterized by X-ray diffraction. The reactivity of the complexes was surveyed, and structures of three additional trinuclear mixed-valence Mn(III)/Mn(II)/Mn(III) spec

Full text of DOI:10.1021/ic302731z

Article
pubs.acs.org/IC
Ferromagnetic Dinuclear Mixed-Valence Mn(II)/Mn(III) Complexes:
Building Blocks for the Higher Nuclearity Complexes. Structure,
Magnetic Properties, and Density Functional Theory Calculations
Mikko M. Hanninen,† Juha Valivaara,† Antonio J. Mota,‡ Enrique Colacio,*,‡ Francesc Lloret,§
̈
̈
and Reijo Sillanpaa*,†
̈
̈
†Department of Chemistry, University of Jyvaskyla, P.O. Box 35, 40014 Jyvaskyla, Finland
̈
̈
̈
̈
‡Departamento de Química Inorgan
§Instituto de Ciencia Molecular (ICMOL), Universitat de Valen
́
ica, Facultad de Ciencias, Universidad de Granada, Avda. Fuentenueva s/n, 18002 Granada, Spain
̀
cia, 46980 Paterna, Valencia, Spain
̀
S
* Supporting Information
ABSTRACT: A series of six mixed-valence Mn(II)/Mn(III) dinuclear complexes were synthesized and characterized by X-ray
diffraction. The reactivity of the complexes was surveyed, and structures of three additional trinuclear mixed-valence Mn(III)/
Mn(II)/Mn(III) species were resolved. The magnetic properties of the complexes were studied in detail both experimentally and
theoretically. All dinuclear complexes show ferromagnetic intramolecular interactions, which were justified on the basis of the
electronic structures of the Mn(II) and Mn(III) ions. The large Mn(II)−O−Mn(III) bond angle and small distortion of the
Mn(II) cation from the ideal square pyramidal geometry were shown to enhance the ferromagnetic interactions since these
geometrical conditions seem to favor the orthogonal arrangement of the magnetic orbitals.
Mn(III) centers connected by oxo-bridging ligands, exhibiting slow
relaxation of the magnetization and magnetic hysteresis below the
so-called blocking temperature (TB) without undergoing three-
dimensional (3D) magnetic ordering.11 From that moment onward
numerous metal complexes exhibiting slow relaxation of the mag-
netization were prepared, mainly being Mn clusters containing at
least some MnIII centers. These nanomagnets, called single-molecule
magnets (SMMs),12 straddle the quantum/classical interface show-
ing quantum effects such as quantum tunnelling of the magnet-
ization and quantum phase interference. They are also potential can-
didates for magnetic information storage and quantum computing.13
The origin of the SMM behavior is the existence of an energy
barrier (Δ) that prevents reversal of the molecular magnetization
and causes slow relaxation of the magnetization at low temperature.
This energy barrier depends on the large-spin multiplicity of the
ground state (ST) and the easy-axis (or Ising-type) magnetic
anisotropy of the entire molecule (D < 0).
INTRODUCTION
■
Polynuclear manganese complexes have been profusely studied
for many years due to their relevance in fields such as metallo-
enzymes and molecular magnetism. As far the former field is con-
cerned, it should be pointed out that many biological processes are
catalyzed by di- and tetramanganese enzymes. These include catal-
ases (hydrogen peroxide disproportionation),1−3 arginase (L-arginine
hydrolysis),4 sulfur-oxidizing enzyme (energy metabolism in
thiobacilli),5 xylose isomerase (isomerization of, e.g., glucose to
fructose),6−8 hydrolases (phosphodiester, -triester and phos-
phoamide hydrolysis), class Ib ribonucleotide reductase,9 and
water oxidase (photosynthetic oxygen production).10 The work
in this area has been focused on the preparation of polynuclear
manganese model compounds, including mixed-valence com-
plexes, to mimic the structure, physical properties, and function
of the active center of the metalloenzyme with the ultimate goal
of elucidating its mechanism of action.
With regard to the field of molecular magnetism, it has
undergone a renaissance with the discovery two decades ago of the
paradigmatic complex Mn12Ac, containing Mn(IV) and anisotropic
Received: December 12, 2012
Published: January 30, 2013
© 2013 American Chemical Society
2228
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
Proceeding from serendipitous synthesis14 to a more rational
design of these complexes (with potential SMM behavior) we
have to be able to control the nuclearity and the bridging atoms
and/or molecules to produce compounds with desired properties.
This is not a straightforward task, especially when the syntheses
begin with a large amount of smaller building blocks. One solution
to this problem is a stepwise synthesis of the SMMs which can be
done by synthesizing complexes with only few ferromagnetically
coupled metal ions and potential reactive sites which can be utilized
in follow-up reactions to increase the nuclearity (and thus total
spin) of the compounds.15 By carefully choosing the reactive
components, these reactions could be furthermore used to acquire
the appropriate symmetry for the product.16
graphite monochromated Mo-Kα radiation or with Agilent SuperNova
dual wavelength diffractometer equipped with Atlas CCD area
detector using Cu−Kα radiation. Variable-temperature (2−300 K)
magnetic susceptibility measurements on polycrystalline samples were
carried out with a Quantum Design Squid MPMSXL-5 device
operating at 1 T from room temperature to 50 K and at 0.05 T
from this latter temperature to 2 K. The experimental susceptibilities
were corrected for the diamagnetism of the constituent atoms by using
Pascal’s tables. Reflectance diffuse spectra were recorded with a Varian
Cary 5E UV−vis-NIR spectrophotometer and the XPS spectra with a
Kratos AXIS Ultra spectrophotometer.
Ligand Syntheses. Ligands H2L1 and H2L2 were synthesized
using our recently published thermal oven method23 following the
general procedure. Phenol (eugenol or 2-methoxy-4-methylphenol),
paraformaldehyde, and 4-(2-aminoethyl)morpholine were weighed
into a 25 mL beaker which was placed inside a 100 mL beaker and
covered with a watch glass. This reaction container was placed in
thermal oven at 120 °C for 1.5 h. The formed yellow oil was dissolved
in ethanol (H2L1) or ethanol−water mixture (H2L2) and was left to
crystallize for a few days. The white (slightly yellowish for H2L2)
precipitate was filtered and air-dried to get synthesis-pure products
To apply this strategy in practice, we have prepared novel
compartmental aminobis(phenolate) ligands which can have up
to seven donor atoms (five oxygens and two nitrogens) and
several different coordination modes for one or more metal
ions (Figure 1). In these ligands the donor atoms are
1
with good yields. For H NMR and elemental analysis of the ligands,
see Supporting Information.
Complex Syntheses. Complex 1 was synthesized by dissolving 1.0
mmol (198 mg) of MnCl2·6H2O into 15 mL of methanol and 0.5 mmol
(242 mg) of H2L1 was added to the solution. Then, 1.0 mmol (140 μL)
of triethylamine was added to neutralize the formed hydrochloric acid, and
the mixture was heated at 65 °C for 45 min. After the solution had cooled
to room temperature, the formed dark green precipitate was separated by
filtration. X-ray diffraction quality crystals were grown in six days by
placing solutions with the above-mentioned molar amounts into a fridge
without preheating the solution. Yield: 241 mg (66%). Elemental analysis:
C30H43Cl2Mn2N2O7 (724.45): calcd. C 49.74, H 5.98, N 3.87; found C
49.50, H 6.03, N 3.91.
Complex 2 was synthesized by dissolving 0.6 mmol (119 mg) of
MnCl2·6H2O into 15 mL of methanol, and 0.5 mmol (130 mg) of
H2L2 was added to the solution. Then, 0.6 mmol (84 μL) of
triethylamine was added to neutralize the formed hydrochloric acid,
and the mixture was heated at 65 °C for 30 min. After the solution had
cooled to room temperature, the solution was concentrated to ca. 10
mL by rotary evaporator. X-ray diffraction quality crystals were formed
in two days and were separated by filtration. Yield: 138 mg (69%).
Elemental analysis: C26H41Cl2Mn2N2O7 (672.38): calcd. C 46.44, H
5.85, N 4.17; found C 46.40, H 5.91, N 4.05.
Complex 3 was synthesized by dissolving 0.5 mmol (99 mg) of
MnCl2·6H2O into 13 mL of ethanol, and 0.25 mmol (121 mg) of
H2L1 was added to the solution. Then, 0.5 mmol (70 μL) of
triethylamine was added to neutralize the formed hydrochloric acid
and the mixture was refluxed for 5 h. After the solution had cooled to
room temperature, a dark green precipitate formed and was separated
by filtration. X-ray diffraction quality crystals were grown from the
solution with the above-mentioned molar amounts at room temper-
ature without heating the solution at any point. Yield: 138 mg (69%).
Elemental analysis: C32H47Cl2Mn2N2O7(752.51): calcd. C 51.07, H
6.30, N 3.72; found C 50.54, H 6.20, N 3.61.
Complex 4 was synthesized by dissolving 0.5 mmol (99 mg) of
MnCl2·6H2O into 13 mL of ethanol, and 0.25 mmol (108 mg) of
H2L2 was added to the solution. Then, 0.6 mmol (84 μL) of
triethylamine was added to neutralize the formed hydrochloric acid.
The reaction vessel was placed in ultrasound bath for half an hour and
then left to crystallize at room temperature. Dark green X-ray
diffraction quality crystals were formed to solution in two days and
were separated by filtration. Yield: 124 mg (76%). Elemental analysis:
C26H37Cl2Mn2N2O6(654.36): calcd. C 47.72, H 5.70, N 4.28; found C
47.65, H 5.84, N 4.08.
Figure 1. Two examples of the possible coordination modes of
aminobis(phenolate)-based ligands.
positioned around the whole molecule. Thus, they can act
also as hydrogen bond acceptors and direct the crystallization
process when they are not coordinated to metal ions. This
family of aminobis(phenolate) ligands was made to react with
Mn2+ in basic conditions under air to obtain mixed-valence
MnIIMnIII complexes. It should be noted that only a few
examples of magnetically and structurally characterized mixed-
valence MnIIMnIII complexes have been reported so far, which
can exhibit either weak ferro- or antiferromagnetic interactions
between the MnII and MnIII metal ions.
In this study, we report the syntheses, reactivity, and magnetic
behavior of six new phenoxo-alkoxo heterobridged mixed-valence
MnIIMnIII dinuclear complexes of general formula [MnIIMnIII(μ-
L)μ-OR)Cl2D] (R = Me or Et and D = MeOH or EtOH). In
addition, we have surveyed the reactivity of the dinuclear
complexes with either different anions (acetate and nitrate) or
ethylene glycol giving rise to three novel MnIIMnIIIMnII trinuclear
complexes. The aim of this work is 2-fold: (i) to disclose how the
structural factors of the bridging network affect the magnetic
exchange interaction in this type of mixed-valence MnIIMnIII
complexes, which would allow the rational preparation of
MnIIMnIII complexes with ferromagnetic exchange interaction,
and (ii) to figure out if the MnIIMnIII complexes containing highly
anisotropic MnIII ions and ferromagnetic coupling between metal
ions leading to a high spin ground state that can exhibit SMM
behavior.
EXPERIMENTAL SECTION
■
General Procedures. All starting materials were reagent grade,
purchased from Sigma Aldrich and used as received. Solvents were
HPLC grade and used without any additional drying. All syntheses
were performed under ambient laboratory atmosphere. Elemental
analysis was performed by using a Vario El III elemental analyzer.
Single-crystal X-ray measurements were performed by using either
Enraf Nonius Kappa CCD area detector diffractometer with the use of
Complex 5 was synthesized by dissolving 0.4 mmol (115 mg) of
MnBr2·4H2O into 10 mL of methanol and 0.2 mmol (97 mg) of H2L1
was added to the solution. Then, 0.4 mmol (56 μL) of triethylamine was
added to neutralize the formed hydrobromic acid and the mixture was
heated at 65 °C for 4 h. After the solution had cooled to room
2229
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
2230
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
Scheme 1. Synthetic Route to Dinuclear Mn-Complexes
temperature it was concentrated to ca. 5 mL by rotary evaporator and
after which the formed dark green precipitate was separated by filtration.
X-ray diffraction quality crystals were formed from slow evaporation of
solution with the above-mentioned molar amounts at room temperature.
Yield: 81 mg (50%). Elemental analysis: C30H43Br2Mn2N2O7 (813.36):
calcd. C 44.30, H 5.33, N 3.44; found C 43.68, H 5.28, N 3.33.
Complex 6 was synthesized by dissolving 0.4 mmol (115 mg) of
MnBr2·4H2O into 10 mL of methanol and 0.2 mmol (86 mg) of H2L2
was added to the solution. Then, 0.4 mmol (56 μL) of triethylamine
was added to neutralize the formed hydrobromic acid and the mixture
was heated at 65 °C for 4 h. After the solution had cooled to room
temperature it was concentrated to ca. 5 mL by rotary evaporator and
after which the formed dark green precipitate was separated by
filtration. X-ray diffraction quality crystals were formed from slow
evaporation of solution with the above-mentioned molar amounts at
room temperature. Yield: 81 mg (53%). Elemental analysis:
C26H39Br2Mn2N2O7 (761.28): calcd. C 41.02, H 5.16, N 3.68; found
C 41.14, H 5.13, N 3.57.
atoms (1.2 or 1.5 times of the host atom). The OH hydrogens were located
from the electron density map and refined isotropically.
RESULTS AND DISCUSSION
■
Synthesis. The ligands H2L1 and H2L2 were prepared in
good to excellent yields by using a recently reported thermal oven
method.23 In general, the mixed-valence complexes 1−6 were
synthesized according to Scheme 1, in which 2 equiv of
MnCl2·6H2O (1−4) or MnBr2·4H2O (5−6) react with 1 equiv
of the corresponding ligand in the presence of 2 equiv of
triethylamine in a selected alcohol. As indicated previously by other
authors,24 in basic solutions under air, the coordination of the
tetradentate tripodal ligands similar to H2L1 and H2L2 to Mn(II)
species favors formation of Mn(III) species. The further reaction in
a 1:1 molar ratio between the mononuclear Mn(III) intermediate
and Mn(II)X2 species present in the reaction medium gave rise to
the formation of the MnIIMnIII mixed-valence complexes.
Complexes 7 and 8: The single crystals of these complexes could be
prepared by dissolving 0.02 mmol of 1 into methanol and adding 0.04
mmol of NH4NO3 (7) or NaOAc (8) to the solution and stirring the
solutions for a few minutes. Crystals formed after slow evaporation of
the solvent at room temperature. Despite several attempts we were
unable to prepare these complexes in a larger scale.
In the syntheses of complexes 1 and 2, best yields were
obtained by heating the reacting solutions in methanol for 30−
45 min, after which the solutions were allowed to cool to room
temperature and the precipitate (for 1) was filtered or the
complexes (for 2) were allowed to crystallize for a certain time
to increase the yield and then were separated by filtration. A
similar procedure as for 1 was also used for synthesis of 3. For
complex 4, however, a better quality product was achieved with
ultrasound assisted synthesis, when the starting materials (in
ethanol) were placed in a ultrasound bath at room temperature
for half an hour, took off from the bath and left to crystallize for
two days after which the formed crystals were filtered.
Complexes 5 and 6 were prepared by similar procedures as
for 1 and 2 by using MnBr2·4H2O as the metal salt. The
reactivity of 1−6 is discussed in detail later in the text.
The diffuse reflectance spectra of complexes 1−6 (Figure S8)
show intense bands with a shoulder in the ultraviolet region at
∼270 and ∼320 nm, which are more likely due to intraligand
and/or Mn(III)-ligand charge transfer bands. In the visible
region these compounds exhibit a shoulder and a wide band at
∼420 and ∼700 nm, respectively. Since the high-spin Mn2+ ions
have a d5 configuration without spin-allowed bands in the
visible region, these two bands come from the Mn3+ ion with
tetragonally elongated octahedral geometry due to the presence
of a strong Jahn−Teller effect. The band with a maximum at
∼700 nm is assigned to the 5A1g ← 5B1g transition, whereas the
shoulder at ∼420 nm can be tentatively assigned to the non-
resolved 5B2g ← 5B1g and 5Eg ← 5B1g transitions, which have their
origin in the splitting of the 5T2g excited state of the Oh geometry.
The positions of these bands are similar to those found for other
Complex 9: 0.09 mmol (68 mg) of complex 1 was dissolved into 6
mL of acetonitrile, and 0.3 mmol (17 μL) of ethylene glycol and 0.6
mmol (84 μL) of triethylamine were added. The mixture was left to
crystallize at room temperature for two days and after which formed X-
ray quality crystals were separated by filtration. Yield: 36 mg (60%).
Elemental analysis: C62H82Cl2Mn3N6O12 (1339.07) (with 2 MeCN):
calcd. C 55.61, H 6.17, N 6.28; found C 54.97, H 6.00, N 6.41.
X-ray Crystallography. The crystal data for compounds 1−10 are
summarized in Table 1 along with other experimental details. The
crystallographic data for 1−6 and 9 were collected at 123 K with an
Enraf Nonius Kappa CCD area-detector diffractometer with the use of
graphite-monochromated Mo-Kα radiation (λ = 0.71073 Å). Data
collection was performed by using φ and ω scans, and the data were
processed by using DENZO-SMN v0.93.0.17 SADABS18 absorption
correction was applied for the compounds. For complexes 7, 8, and 10
the data collection was performed with Agilent SuperNova dual
wavelength diffractometer equipped with an Atlas CCD area detector
using Cu-Kα radiation using CrysAlisPro program package.19 The
empirical (7 and 8) absorption correction with SCALE3 ABSPACK
scaling algorithm or analytical numeric one using multifaceted crystal
(10) was performed as implemented in the CrysAlisPro program.
The structures were solved by direct methods by using the SHELXS-
9720 program or the SIR-9721 program, and the full-matrix least-squares
refinements on F2 were performed using the SHELXL-9720 program.
Figures were drawn with Diamond 3.22 For all compounds the heavy atoms
were refined anisotropically. The CH hydrogen atoms were included at the
calculated distances with fixed displacement parameters from their host
2231
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
tetragonally distorted Mn3+ complexes.25 The diffuse reflec-
tance spectrum of 9 (Figure S8) exhibits one intense band at
250 nm and a band without clear maximum at 640 nm. The
blue shift of the A1g ← B1g transition with respect to its
position in complexes 1−6 may be due to the higher Jahn−
Teller distortion affecting complex 9. It should be noted that no
strong absorption assignable to an intervalence charge transfer
was observed in the visible or near-IR region.
A direct evidence of the existence of Mn(II) and Mn(III)
species is provided by XPS measurement. The deconvoluted
XPS spectra of 1−6 are all almost identical, and, therefore, only
that of compound 3 is given as an example in Figure 2. The
coordinated phenoxo oxygen, two tripodal nitrogens, one μ2-
bridging phenoxo group from the ligand, and one μ2-bridging
alkoxo oxygen. The sixth coordination site of Mn(III) ion is
occupied either by a coordinated alcohol molecule (1−3, 5 and 6)
or by the morpholine oxygen (4) coming from a neighboring
complex (see below). In the six-coordinated Mn(III) ions O1, N8,
O2, and O5 donor form equatorial bonds to the metal ion (1.840−
2.114 Å), and O6 and N18 form axial bonds (2.318−2.423 Å). The
calculation of the degree of distortion of the Mn3+ coordination
polyhedron with respect to ideal six-vertex polyhedra, by using the
continuous shape measure theory and SHAPE software,26 led to
shape measures relative to the octahedron (OC-6) and trigonal
prism (TPR-6) with values in the 1.6−2.1 and 14.2−15.4 ranges,
respectively (see Table S3) Therefore, the Mn3+ coordination
spheres are found in the OC-6 ↔ TPR-6 deformation pathway and
are close to the octahedral geometry (∼90%) somewhat distorted
to trigonal prismatic (see Table 3).
5
5
The pentacoordinated Mn(II) ions has in turn a distorted square
based pyramidal coordination (Table 3), and it is surrounded by
two terminal chlorides (1−4) or bromides (5−6), one μ2-bridging
phenoxo (O2), one μ2-bridging alkoxo (O5), and one methoxy
oxygen (O4) from the ligand. The other methoxy group (on the
side of the hexacoordinated manganese) does not coordinate to the
metal ion in any of the studied complexes.
Important geometrical parameters of complexes 1−6 are
summarized in Tables 2 and S1. The bond distances around
manganese atoms in all complexes are comparable to earlier
studies concerning similar complexes.15,24 The bonds from
Mn(III) ion to phenoxo and alkoxo oxygens are about 0.2 Å
shorter than those from Mn(II) ion to the same atoms, which
furthermore confirms the different oxidation states between the
two ions. The Mn1−O1 distance is slightly shorter than the
Mn1−O2 one because the O2 is the bridging atom between the
two manganese ions. In addition, the Mn1−N18 bond is
considerably longer (usually more than 0.3 Å) compared to the
other manganese nitrogen bonds (Mn1−N8) which is due to
the Jahn−Teller effect of the Mn(III) ion.
Figure 2. XPS spectrum for 3.
spectrum shows peaks of Mn2P3/2 and Mn2P1/2 at 638.4 and
650.4 eV attributable to Mn(II), whereas the peaks at 639.9 and
651.6 eV indicate the presence of Mn(III). The contour area of the
peaks clearly indicates the same content of Mn(II) and Mn(III). On
the other hand, the XPS spectrum of compound 9 (Figure S9)
shows the Mn(II) and Mn(III) peaks at bonding energies very close
to those found for compounds 1−6, but in this case the contour
area of the peaks indicates a 2:1 Mn(III)/Mn(II) ratio.
X-ray Diffraction Analysis. The molecular structures of all
studied complexes were determined by single crystal X-ray
diffraction, and that of complex 1 is depicted in Figure 3 (for
The change of bridging alkoxo group from methoxo (1) to
ethoxo (3) certainly has influence on the Mn−O5 bond lengths
but also to other bonds around manganese ions which in turn
causes variation on the Mn−Mn distance and the Mn−O−Mn
angle. Generally the bond lengths between different complexes
are similar whereas the bond angles have more deviation. Since
complexes 5 and 6 are similar to 1 and 2 with terminal
chlorides substituted with bromides, the structural parameters
of the corresponding complexes are almost identical (excluding
the Mn−Br bonds). However, there are small differences in
Mn−O−Mn bond angles of 1 and 5 because of the different
packing of the complexes (see below).
Solid State Packing. As the complexes provide several
hydrogen bond acceptor atoms (chlorides coordinated to the
Mn(II) ion, methoxy oxygens and aromatic rings in ligands) the
packing of the complexes is mainly directed by these bonds and/or
weak interactions. Interestingly, one of the most common donor
atoms for these bonds is methylene or aromatic methoxy carbon
atoms, whose charge distribution is polarized by neighboring
nitrogen or oxygen atoms and aromatic rings, thus allowing these
carbons to act as donor atoms in weak C−H···X hydrogen bonds.27
In complex 1 the OH proton of the coordinated methanol
molecule has a strong intramolecular hydrogen bond to chloride
(OH···Cl = 2.50 Å), whereas the same hydrogen in 2 has even
stronger (2.39 Å) intermolecular hydrogen bond to the chloride ion
of the neighboring molecule. Furthermore, complex 2 possesses an
Figure 3. Molecular structure of complex 1. Thermal ellipsoids have
been drawn at the 30% probability level and C−H hydrogen atoms are
omitted for clarity.
the molecular structures of 2−6, see Figures S1−S3). The basic
structure of all complexes 1−6 contains Mn(II) and Mn(III)
ions, one aminobis(phenolato) ligand, one bridging alkoxo
group, and two terminal halides. The Jahn−Teller distorted
Mn(III) ion (D4h) in all complexes is hexacoordinated with
elongated octahedral coordination sphere which comprises one
2232
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
Table 2. Selected Bond Lengths and Angles of 1−6
a
a
1
2
3
4a
4b
5
6
Mn1−O1
Mn1−O2
Mn1−O5
Mn1−N8
Mn1−O6
Mn1−N18
Mn1−O21′
Mn2−O2
Mn2−O4
Mn2−O5
Mn2−Cl1
Mn2−Cl2
Mn1−Mn2
Mn1−O2−Mn2 (θ)
Mn1−O5−Mn2 (θ)
Mn1−O2−O5−Mn2
O2−Mn1−Mn2−O5
1.846(2)
1.915(2)
1.902(2)
2.092(3)
2.319(2)
2.413(3)
1.847(2)
1.934(2)
1.900(2)
2.095(2)
2.320(2)
2.423(2)
1.857(2)
1.903(2)
1.912(2)
2.090(2)
2.323(2)
2.402(2)
1.861(5)
1.924(5)
1.889(6)
2.110(7)
1.858(5)
1.905(5)
1.927(6)
2.082(6)
1.840(5)
1.912(5)
1.902(5)
2.114(6)
2.318(5)
2.408(6)
1.853(4)
1.933(4)
1.904(4)
2.095(5)
2.329(4)
2.411(4)
2.357(6)
2.414(5)
2.160(5)
2.422(6)
2.152(6)
2.340(2)
2.332(2)
3.2188(16)
103.9(2)
105.4(2)
176.5(3)
175.5(4)
2.413(6)
2.484(5)
2.146(6)
2.379(6)
2.155(6)
2.349(2)
2.345(2)
3.243(2)
106.2(2)
105.1(2)
177.9(3)
177.2(4)
2.109(2)
2.433(2)
2.127(2)
2.309(1)
2.358(1)
3.1726(6)
104.0(1)
103.8(1)
171.8(1)
169.5(1)
2.112(2)
2.464(2)
2.118(2)
2.338(1)
2.360(1)
3.1777(5)
103.4(1)
104.4(1)
177.1(1)
176.2(1)
2.127(2)
2.416(2)
2.147(2)
2.385(1)
2.326(1)
3.2061(4)
105.3(1)
104.2(1)
171.8(1)
169.4(1)
2.123(5)
2.409(5)
2.106(5)
2.497(2)
2.480(2)
2.100(4)
2.476(4)
2.115(4)
2.484(1)
2.501(1)
b
c
b
c
3.1848(15)
104.1(2)
105.1(2)
175.1(2)
3.1733(12)
103.7(2)
104.2(2)
176.5(2)
173.7(3)
175.6(3)
c
a
b
The geometrical parameters of two different dinuclear complexes in the asymmetric unit of 4 are presented as 4a and 4b. Mn1−Br1. Mn1−Br2.
Figure 4. Solid state packing of complex 2 showing the intermolecular hydrogen bonding (red) and π−π interaction (blue).
Figure 5. The packing diagram of 4 showing the 1D polymeric structure.
example of parallel offset aromatic−aromatic (π−π) interaction with
two phenyl rings lying partly on top of each other with a separation
of 3.38 Å. The preceding interaction and/or strong intermolecular
OH···Cl hydrogen bonds mainly determines the spatial arrangement
of 1 and 2 in the lattice (Figure 4). Both complexes have also several
rather weak intermolecular interactions between chlorides and
methylene, aromatic and/or methoxy methyl hydrogens, which
further facilitate the crystallization by stabilizing the entire structure.
The packing of complex 3 is fairly similar to that of 1 with
strong intramolecular hydrogen bonds from coordinated ethanol to
chloride and slightly weaker intermolecular interactions to chloride
and morpholine oxygen from methylene and aromatic hydrogens.
The solid state ordering of 4 differs from the preceding
complexes notably since the coordinated alcohol molecule observed
in previous complexes has been replaced by a morpholine oxygen
from the adjacent complex filling the coordination sphere of
Mn(III) ion (Figure 5). This induces significant changes in the
2233
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
Figure 6. The solid state ordering of complex 5 showing the intermolecular OH···Br- (red) and CH···π-hydrogen bonds (blue).
Scheme 2. Schematic Pathway for the Formation of Complexes 7−9
orientation of the whole morpholine ring compared to complexes
1−3 in which the morpholine ring has settled nearly parallel
compared to the Mn−O−Mn−O plane, whereas in 4, the ring is
virtually orthogonal to this plane. The coordination of the mor-
pholine oxygen produces a 1D coordination polymer with fairly few
interactions to two other directions. Besides the aforementioned
interaction, one notable intermolecular contact is between a
hydrogen from the methoxy group and the methoxy oxygen from
an adjacent complex (with C···O separation of 3.38 Å).
Rather surprisingly, the packing of complex 5 is reminiscent
of complex 2 more than 1 (which has the same ligand, H2L1),
with strong intermolecular H···Br hydrogen bonds instead of
intramolecular ones. The packing of complexes 6 and 2 is
almost identical (although the aromatic−aromatic interaction is
not as clear as in 2). The stacking of aromatic rings in both
complexes 5 and 6 is probably best described as CH/π
hydrogen bond than actual π−π interaction (Figure 6).
Reactivity of the Complexes. We were interested in the
reactivity of the synthesized complexes in order to be able to
prepare more complicated molecular structures with larger spin
ground state and easy-axis (Ising type) magnetic anisotropy, the
most sought-after properties of SMMs. From the solid state
structures of 1−6, chemical intuition would suggest that the
coordinated halides would be the most reactive groups in these
complexes; hence the lability of the complexes toward anion
exchange reactions using sodium and ammonium salts of some
common anions was investigated.
solution was not observed even at −20 °C. However, small
amounts of single crystals suitable for X-ray diffraction were
formed after very slow evaporation of methanol solution. The
solid state structure revealed that the nitrate group has indeed
replaced the chloride ions. However, since the coordination
capability of the nitrate as a monodentate ligand is weaker than
the chloride, instead of the formation of the assumed Mn2L1μ2-
OMe(NO3)2 complex, the reaction unexpectedly proceeded
further to produce the mixed-valence trinuclear Mn(III)/
Mn(II)/Mn(III) complex 7 (Scheme 2). The distorted
octahedral coordination sphere of both Mn(III) ions in 7 is
similar to 1 and 2 by means of amino(phenolate) ligand
coordinated in a tetradentate manner and Mn(III)/Mn(II) μ2-
bridging methoxo group. The μ2-bridging nitrate anions have
replaced the chlorides, aromatic methoxy substituents, and/or
coordinated alcohol/morpholine oxygens around the manga-
nese ions, thus producing a distorted octahedral coordination
sphere for all metal ions in the complex (for geometrical
parameters, see Table S2).
The above-described procedure with complex 1 and sodium
acetate produced complex 8 (Figure 7) with acetate anions as
bridging ligands. The coordination sphere around metal ions is
similar to 7, although the geometrical parameters around them
are somewhat different (Table S2). In spite of our numerous
efforts with several solvents and starting materials, we have not
been able to produce pure enough bulk material from these
reactions, and hence the magnetic properties of complexes were
studied only by computational methods (see below). However,
the obtained crystal structures demonstrate that coordinated
halides are reactive and can be replaced by different anions.
The reactivity was initially tested with ammonium nitrate by
stirring complex 1 and 2 equiv of NH4NO3 in methanol for 4 h.
Unfortunately, precipitation or crystal formation from the
2234
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
Figure 7. Molecular structure of 8. Thermal ellipsoids have been drawn at 30% probability level and C−H hydrogen atoms are omitted for clarity.
Another interesting reaction for the prepared dinuclear
complexes would be the substitution reaction of the halides by
an alcoholate with the aim to get a new complex with an
additional alkoxo ligand coordinated to the metal ion. Thus, we
explored the possible reaction between complexes 1−4 and 10-
fold excess of ethylene glycol in the presence of triethylamine
expecting that the ethylene glycolate dianion coordinates to the
Mn(II) ion in a bidentate manner by replacing the chloride
ions. The reaction in acetonitrile rapidly produced a green
precipitate, but we were not able to recrystallize it to achieve
any X-ray diffraction quality crystals. However, the same reaction
with complex 3 proceeded more slowly, thus producing green,
good quality crystals of 9 from which we were able to determine
the structure of the product (Figure 8).
trinuclear Mn(III)/Mn(II)/Mn(III) complex (9), with a Mn(1)−
Mn(2)- and Mn(2)−Mn(3)-μ2-bridging ethylene glycolate,
Mn(1)−Mn(3)-μ2-bridging chloride anion and one additional
terminal chloride coordinated to the central Mn(2)-ion (Scheme 2,
Figure 8). Both Mn(III) cations have Jahn−Teller distorted octa-
hedral coordination spheres with tetradentally coordinated ligand
L32− replenished with bridging chloride and alkoxo oxygen, while
the surroundings of the Mn(II) ion vaguely resemble those in 1−6
with two pairs of bridging alkoxo and phenoxo oxygens and
terminal chloride anion. The bridging chloride anion between the
Mn(III) cations causes the whole structure to be folded (the
Mn(1)−Cl(2)−Mn(3) angle is 111.8°) compared to the rather
linear arrangement in 7 and 8. Selected geometrical parameters of
9 can be found in Table S2.
In the absence of triethylamine the reaction produces
complex 10 where the original alkoxo bridge has been replaced
by just singly deprotonated ethylene glycol (Figure S5); hence
triethylamine is presumably needed for the double deprotona-
tion of ethylene glycol enabling the trinuclear product. Later,
we were able to grow crystals from the reaction of 1 and
ethylene glycol, thus confirming that also the methoxo bridged
complexes undergo the preceding reaction and the bulk
material was prepared using 1 as the starting material.
On considering the remarkable reactivity of these dinuclear
Mn(II)/Mn(III) complexes, we should be able to push these
complexes to react with another alcohol (or acids) to produce
differently bridged systems. It seems that we can transform 1 into 3
by dissolving it in ethanol and exchange the methoxo bridge to
monodentate μ2-bridging ethylene glycol (10), which gives us
additional support on the reactivity of the alkoxo bridge. The
previously mentioned reactions pointed out how rich the chemistry
of complexes 1−6 can be, and we are currently working to
selectively change the bridging group and exploring new reactions
to produce complexes with higher nuclearities.
Magnetic Properties. Variable-temperature magnetic
susceptibility studies were carried out on powdered samples
of the crystalline complexes 1−6 and 9 in the temperature
range 2−300 K and under an applied field of 1000 Oe. The
magnetic properties of compounds 1−6 and 9 in the form of
χMT vs T plots (χM being the magnetic susceptibility per
MnIIMnIII unit for compounds 1−6, and per MnIIIMnIIMnIII
unit for compound 9) are shown in Figure 9.
Figure 8. Molecular structure of 9. Thermal ellipsoids have been
drawn at 30% probability level and hydrogen atoms and non-
coordinating solvents are removed for clarity.
Nevertheless, crystals did not consist of the expected ethylene
glycolate adduct, but quite surprisingly we found again a
2235
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
sharply to reach a value of 4.75 cm3 mol−1 K at 2 K. The shape of
this curve is typical of a dominant antiferromagnetic coupling in
the trinuclear complex 9.
In keeping with the dinuclear structure of the MnIIMnIII
mixed-valence complexes the magnetic susceptibilities of
complexes 1−6 were analyzed by means of the following the
spin exchange Hamiltonian:
H = −JS1S2 − DMn(S2z − 2) + zJ′ < Sz > Sz
with (S1 = 2 and S2 = 5/2). The second term describes the local
anisotropy of the Mn3+ ion and the third term describes the
intermolecular interactions in the molecular field approximation.28
Because D and zJ′ are strongly correlated, as they produce the
same effect on the experimental curves (a decrease in the χMT pro-
duct at low temperature), their independent contributions cannot
be reliably evaluated from the magnetic susceptibility data. In view
of this, we decided to fit the experimental magnetic susceptibility
with the above Hamiltonian but fixing either D or zJ′ to zero. How-
ever, in the first case, the corresponding fits lead to unacceptable
high values of D. Therefore, we will keep the magnetic parameters
extracted with DMn fixed to zero, which are gathered in Table 3, for
further discussions. It should be noted that the zJ′ parameters
obtained from the fitting procedure are overestimated as they
include the effects of both the weak intermolecular interactions
and the local anisotropy of the Mn3+ ion. In the case of the
chain complex 4, the binuclear subunits are far away from each
other, so that the magnetic interaction between the two
dinuclear units can also be analyzed with the molecular field
approximation. To avoid overparametrization, the two different
dinuclear units present in 4 were considered to have the same
magnetic coupling.
Figure 9. Temperature dependence of the χMT product for complexes
1−6 and 9.
For compounds 1−6, the value of χMT at room temperature
is slightly higher than the theoretical value (7.375 cm3 mol−1 K)
expected for a dinuclear system with magnetically isolated Mn2+
(S = 5/2, g = 2) and Mn3+ ions (S = 2, g = 2). When the
temperature is lowered, the χMT steadily increases reaching a
maximum (a shoulder in the case of 4) of ∼12 cm3 mol−1 K at
∼10 K. Below this temperature, the χMT decreases rapidly to
2 K (in the case of 4 the χMT sharply increases to reach a value
of 16.6 cm3 K mol−1 at 2 K). This behavior is due to a
significant intradinuclear ferromagnetic coupling between Mn2+
and Mn3+ ions leading to a S = 9/2 ground state. The decrease in
χMT at low temperatures for compounds 1−3 and 6 suggests the
existence of zero field splitting (ZFS) effects due to the presence
of an anisotropic Mn3+ ion (the ZFS for the isotropic Mn2+ ion is
assumed to be comparatively negligible) and/or intermolecular
antiferromagnetic interactions. In the case of 4, the sharp increase
in χMT at low temperature is due to intermolecular ferromagnetic
interactions. At room temperature, the χMT value for 9 (11.23
cm3 K mol−1) is very close to the expected value for a magne-
tically uncoupled system containing two Mn3+ ions and one Mn2+
ion with g = 2.11 (11.42 cm3 mol−1 K). On lowering tem-
perature, the χMT first decreases slowly until ∼100 K and then
In order to support the experimental J values found for
compounds 1−6 and to determine some magnetic interactions
that could not be extracted from the fitting of the magnetic
data, density functional theory (DFT) calculations using the
broken-symmetry approach (see Supporting Information) were
Table 3. Magnetostructural Data for the Mixed-Valence MnIIMnIII Complexes 1−9
complex
1
2
3
4
5
6
7
8
9
J calcd/cm−1
a
c
e
f
g
J1−2
J2−3
J1−3
J1−4
+10.4
+13.4
+10.0
+10.7/+12.4
+12.7
+12.6
+5.9
+2.4
+7.9
+7.9
+3.1
b
−0.02
+0.52
−8.6
+0.02/−0.01
−0.37
−0.44
d
<0.01
J exp/cm−1
J1−2 (J1′-2′)
J2−2′
+3.44(6)
+4.2(1)
+2.15(6)
+7.9(7)
+3.7(1)
+2.9(1)
0.04(7)
−5.3(1)
2.11(1)
g
2.045(2)
−0.10(1)
103.88
3.179
2.012(4)
−0.06(1)
103.90
3.178
2.032(3)
−0.07(1)
104.73
3.206
1.993(1)
−0.04(1)
105.15
3.231
2.049(5)
−0.08(1)
104.61
3.185
2.042(3)
−0.04(1)
103.93
3.173
zJ′
i
θ/(°)
Mn2+-Mn3+/Å
100.68
3.189
102.65
3.181
101.63
3.191
17.3
i
h
distortion
9.3
7.2
14.4
2.4
2.3
6.8
from C4v (δ, %)
11.7
i
i
distortion
9.3
10.3
9.6
11.5
10.3
10.5
9.4
11.6
13.1
j
from Oh (%)
a
The pair of values refers to both dinuclear subunits forming the tetranuclear complex (see Figure S10). The first value is for the dinuclear moiety
b
with MnII−MnIII distance at 3.219 Å, and the second, to the other subunit, with MnII−MnIII distance of 3.243 Å. This value is for the coupling
c
between the two dinuclear subunits connected throughout the MnIII metallic centers. This value refers to the two possible 1,3 MnII/MnIII magnetic
d
e
coupling. This value is for the 1,4 MnII/MnII magnetic coupling. This value is for the pair MnIII/MnII at 3.182 Å (below appears the value for the
f
g
MnIII/MnII pair at 3.196 Å). This complex is centrosymmetric. This value is for the pair MnIII/MnII at 3.216 Å (below appears the value for the
h
MnIII/MnII pair at 3.166 Å). Distortion from SPY-5 toward TPBY-5 calculated using the Addison’s method: δ = (β−α)/60, α and β are the largest
trans bond angles; see ref 29. Average values. Distortion from OC-6 toward TPR-6 calculated using the SHAPE program.26
i
j
2236
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
carried out on the X-ray structures as found in the solid state.
As it can be observed in Table 3, the calculated values are in
good agreement in sign and magnitude with those obtained
from the experimental results.
have square pyramidal (C4v) and ideal elongated octahedral
(D4h) geometries, respectively. The five d-type orbitals span the
irreducible representations of the corresponding symmetry
group and split as shown in Figure 12. For the Mn2+ ion all the
The field dependences of the molar magnetization at 2 K for
compounds 1−6 (the M vs H plot for compound 3 is given as an
example in Figure 10) are below the Brillouin function for S = 9/2,
Figure 10. M vs H plot for 3. The solid and dashed lines represent the
Brillouin functions for a S = 9/2 ground state and for the sum of a
S = 2 and a S = 5/2 spin states with g = 2.03, respectively.
which is more likely due to the combined effects of interdimer
antiferromagnetic interactions and/or zero-field splitting of the
S = 9/2 ground state. However, the experimental values are well
above the sum of the magnetization contributions expected for the
isolated ions, thus supporting the intramolecular ferromagnetic
interaction in these compounds.
Variable-temperature magnetization data were collected
between 2 and 10 K at applied fields ranging from 0.5 to
7 T. The M vs H/T plots for 1−6 (the plot for compound 3 is
given as an example in Figure 11) are not superimposed on a
Figure 12. Magnetic exchange pathways in MnIIMnIII mixed-valence
dinuclear complexes. Red and blue lines represent the ferromagnetic
and antiferromagnetic interactions, respectively.
d orbitals are half-filled orbitals, whereas in the Mn3+ ion the
dxy(b1) is an empty orbital and the rest are half-filled d orbitals.
In the almost planar MnII(O)2MnIII bridging fragment, the
interaction between the magnetic orbitals of the Mn2+ and
Mn3+ ions gives rise to a number of ferromagnetic (F) and
antiferromagnetic (AF) competing contributions, the most
significant occurring in the xy plane. Thus, the interaction
between the empty dxy orbital of the Mn3+ ion and the half-
filled, close in energy magnetic orbitals of the Mn2+ ion, the so-
called crossed interactions (Jb1b2, Jb1a1 and Jb1b1 being the
leading terms),28 as well as the interaction between orthogonal
orbitals (the leading terms being Jb2b1, Ja1b1, Jegb1, and Jegb2) is
ferromagnetic. For θ bridging angles larger than 100°, the
interaction between magnetic orbitals of the same symmetry is
AF, such as Jb2b2, Ja1a1, Jegeg. This latter contribution concerns
orbitals where the xy is a nodal plane, which are weakly
delocalized toward the bridging oxygen atoms through π-type
metal-bridging ligand overlap and therefore is expected to be
very weak AF. The Ja1a1 involves orbitals that are weakly
delocalized in the xy plane, and so a weak AF interaction is also
expected. Finally, the Jb2b2 contribution deals with the dx2−y2
orbitals pointing toward each other along the x-axis and are
weakly delocalized toward the bridges in a π-manner.
Therefore, for θ > 100° a medium AF contribution is expected
(this includes the direct overlap which is favored when θ is
small). Although usually AF contributions are stronger than the
F ones, in complexes 1−6 the relatively higher number of F
contributions and the weakness of the AF contributions (all
these complexes have θ > 100°) could lead to a situation where
Figure 11. M vs H/T plots for 3. Solid lines are guides for the eye.
single master curve, clearly indicating the presence of a
significant magnetic anisotropy in these complexes. However,
all attempts to obtain the D and E anisotropic parameters of the
S = 9/2 ground state by fitting the reduced magnetization data
were unsuccessful.
The sign and magnitude of the ferromagnetic exchange
interactions in MnIIMnIII mixed valence pairs can be under-
stood using the known Goodenough−Kanamori rules,30 which
state that the magnetic exchange interactions between pairs of
overlapping magnetic orbitals are antiferromagnetic, those
between pairs of orthogonal magnetic orbitals are ferromag-
netic, and those due to the overlapping of the magnetic orbital
of an ion with an empty orbital of another ion are
ferromagnetic through the spin polarization mechanism. For
the sake of simplicity, let us assume that Mn2+ and Mn3+ ions
2237
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
a
Table 4. Magnetostructural Data for Dinuclear and Trinuclear Mixed-Valence MnIIMnIII Complexes
geometry
compound
Jexp/cm−1
θ/°
β/°
MnII/MnIII
CCDC code
ref
Dinuclear
Double triazolate bridges
b
[MnIIMnIII(L)2(CH3OH)3(OCN)]·(H2O)2
[MnIIMnIII(L)2(H2O)2(phba)]·(DMF)2·(CH3OH)0.5
[MnIIMnIII(L)2(DMF)1.5(H2O)1.5Cl]·(H2O)2.5
+1.3
+2.3
+1.7
+1.6
132.7
16.1
36.3
27.6
OC-6/OC-6
OC-6/OC-6
OC-6/OC-6
OC-6/OC-6
OHIXEJ
OHIYAG
OHIXUZ
OHIXOT
31
30
30
30
131.6
132.2
{[MnIIMnIII(L)2(CH3OH)(H2O)2]2(μ-fua)}{[MnIIMnIII (L)2(CH3OH)
(H2O)2](μ-fua)}·4CH3OH·8H2O
132.1/132.1
26.02/31.04
{[MnIIMnIII(L)2(DMSO)2(H2O)]2(μ-suc)}·10H2O
{[MnIIMnIII(L)2(CH3OH)3]2(μ-phth)}·CH3OH·5H2O
+1.2
+2.2
+1.5
133.5
27.8
OC-6/OC-6
OC-6/OC-6
OC-6/OC-6
OHIYEK
OHIXIN
OHIYIO
30
30
30
131.9
1.4
b
{[MnIIMnIII (L)2(H2O)2][MnIIMnIII (L)2(CH3OH)(H2O)(DMF)]}
128.9/132.3
55.84/27.30
·5H2O·DMF·CH3OH
Phenoxo-alkoxo bridges
[MnIII(L1)(CH3OH)(OCH3)MnIICl2]·CH3OH
Double diphenoxo bridges
+6.0
106.2/103.3
9.8
SPY-5/OC-6
CAFHOG
24a
[MnIIMnIII(biphen)2(biphenH)(bpy)2]
[MnIIMnIII(L2)(H2O)3](ClO4)2·1/2EtOH·1/4H2O
Double dialkoxo bridges
+1.8
+2.9
97.0/102.4
99.1
10.8
9.8
SPY-5/OC-6
OC-6/OC-6
GAVWOP
LOBMOE
32
33
(TEA)[MnIIMnIII(2-OH(5-Clsal)pn)2]
Triple alkoxo/hydroxo/phenoxo dicarboxylate bridges
[MnIIMnIII(μ-OH)-(μ-piv)2(Me3tacn)2](ClO4)2
[MnIIMnIII(bpmp)(μ-OAc)2](ClO4)2·H2O
[MnIIMnIII (bcmp)(μ-OAc)2](ClO4)2·Cl2CH2
Triple diphenoxochloride bridges
[MnIIMnIII (L3)(μ-Cl)ClBr]
+0.5
103.9
5.1
OC-6/OC-6
ROCRAC
34
−8.5
−6.0
−7.7
119.3
114.4
112.3
OC-6/OC-6
OC-6/OC-6
OC-6/OC-6
TIPFED
35
36
37
FOSCAR
GACCOC
−1.7
93.6/102.5
34.2
OC-6/OC-6
FOSCEV
35
Trinuclear
Bisdialkoxo and bisalkoxophenoxo bridges
[MnIIIMn2II(Hcht)2(bpy)4](ClO4)3·Et2O·2MeCN
+3.2
+3.6
102.0
0.0
6.9
9.1
OC-6/OC-6
OC-6/OC-6
WANHUP
QAXZUK
38
39
[MnIIIMn2 (L4)(μ-OMe)2](ClO4)3
96.8/108.0
96.6/107.7
II
Triple Phenoxodicarboxylate
[MnIII2MnII(L5)2(ba)4]
−6.7
−2.5
113.2/114.8
102.1/97.8
OC-6/OC-6
OC-6/OC-6
CAFNAY
PUYMEB
24a
40
Triple Phenoxo-hydroxo-carboxylate
[MnIII2MnII(L5)2(dcba)2(OMe)2]
Triple Phenoxo-alkoxo-carboxylate
[MnIII2MnII(L6)2(CH3CH2COO)2(OMe)2]·H2O
[MnIII2MnII(L7)2(mcba)2(OMe)2]
22.5
−0.13
+3.8
99.9/100.9
102.9/99.2
26.5/24.9
24.5
OC-6/OC-6
OC-6/OC-6
FORVEO
NUJPEN
41
42
a
H2L = 3-(2-phenol)-5-(pyridin-2-yl)-1,2,4-triazole. Phba = p-Hydroxybenzoic. Fua = fumaric acid. Suc = succinic acid. Phth = terephthalic acid.
H2biphen = 2,2′-biphenol. Piv = pivalic acid. Me3tacn = 1,4,7-trimethyltriazacyclonane. Bpmp = it was prepared from the condensation of bis[(2-
pyridylmethyl)methyl]amine with 2,6-bis(chloromethyl)-p-cresol. H2L1 = N-(3,5-di-t-buthyl-2-hydroxybenzyl)-N-(2-hydroxy-3-methoxybenzyl)-
N′,N′-dimethylethylenediamine. H2L2 = it was prepared form the condensation of 2,6-diformyl-4-methylphenol with diethylenetriamine and
ClCH2COOEt. 2-OH(5-Clsal)pn = 1,3-bis(salicylideneamino)-2-propanol. H2L3 = it was prepared form the condensation of 2,6-diformyl-4-t-
butylphenol with 1,3-diaminepropane. Bcmp = prepared by the reaction of N,N′-bis(p-tolylsulfonyl)-1,4,7-tacn with 2,6-bis(chloromethyl)-p-cresol
followed by deprotection. Hcht = cis,cis-1,3,5-cyclohexanetriol. H2L4 = it was prepared from the metal-templated Schiff-base condensation of 2,6-
diformyl-4-methylphenol with 1,11-diamino-3,9-dimethyl-3,9-diazo-6-oxaundecane. H2L5 = N,N-bis(2-hydroxybenzyl)-N′,N′-dimethylethylenedi-
amine and dcba = 2,6-dichlorobenzoic acid. H2L6 = N,N′-dimethyl-N,N′-bis(2-hydroxy-3,5-dimethylbenzyl)-ethylenediamine. H2L7 = N-(3,5-di-t-
b
buthyl-2-hydroxybenzyl)-N-(2-hydroxybenzyl)-N′,N′-dimethylethylenediamine. Hmcba = m-chlorobenzoic. Mean value of the MnII−N−N and
MnIII−N−N angles.
the F contributions were dominant. In fact, all the dinuclear and
trinuclear MnIIMnIII complexes containing a planar bridging
network (diphenoxo, phenoxo/alkoxo, and dialkoxo bridges)
exhibit ferromagnetic coupling (Table 4).
In those cases where the magnetic coupling is AF, there are
additional bridges (usually syn-syn carboxylate groups) that slightly
fold the structure affecting the magnitude of the AF and F
competing interactions in such a way that the AF contributions
become dominant. This AF behavior can be alternatively explained
by considering the counter-complementarity exerted by the
coordination of an additional syn-syn carboxylate group to the
square planar MnII(O)2MnIII bridging fragment, which induces a
decrease in the ferromagnetic J value. The relatively low calculated
J values for the trinuclear MnIII−MnII−MnIII complexes 7 and 8,
which contain triple alkoxo/phenoxo/acetate bridging groups,
compared to those complexes 1−6, support this hypothesis.
It should be pointed out that the magnetic coupling might also
contain a contribution due to the spin-dependent delocalization
(also called double exchange), which is always positive.43 Whereas
the superexchange takes place by electron transfer via an excited
state, which is mixed into the ground state by second-order
perturbation, the spin-dependent delocalization occurs via ground
state configurations. In complexes 1−6, the spin-dependent
delocalization term, if it exists, is very small as the valences of
the Mn2+ and Mn3+ are essentially trapped (class I compounds
according to Robin and Day classification).
2238
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
After analyzing the structural parameters for a series of
dinuclear and trinuclear MnIIMnIII mixed valence complexes,
one can draw the following conclusion: Although the J values
are not very sensible to the θ value, it seems that they increase
when θ increases and the distortion of the Mn2+ polyhedron
from ideal square pyramidal geometry decreases. Thus
compound 3 with a comparatively large θ value and the larger
distortion (δ = 0.14; δ is zero for SPY-5 and 1 for TBPY-5)
exhibits the lower J value. However, compound 4 with the
bigger θ values and less distortion in one of their different
dinuclear units exhibits the stronger magnetic coupling. The
calculated J values for 1−6 seem to follow the same trend. This
conclusion must be taken with caution as the variation in
structural parameters and magnetic coupling values takes place
in a short-range for all these compounds.
that the J values are less positive (the F interaction decreases)
when δ increases and becomes negative (AF interaction) for δ
values above ∼0.20. These theoretical results are in full
agreement with the conclusions drawn from the experimental
results for compounds 1−6.
The magnetic data for compound 9 were analyzed with the
following isotropic Hamiltonian for an isosceles trinuclear
triangular system
H = −J (SMn1SMn2 + SMn2SMn3) − J23(SMn1SMn1)
12
The fit of the experimental susceptibility data with the above
Hamiltonian using the full matrix diagonalization MAGPACK
program44 afforded the following set of parameters: J12
=
+0.04(7) cm−1, J23 = −5.3(1), g = 2.11(1). It should be noted
that in this case, with a dominant antiferromagnetic interaction,
the influence of local anisotropy on the magnetic properties is
very difficult to be evidenced from the magnetic susceptibility
data. This is the reason why we have not considered the local
anisotropy in the Hamiltonian used for analyzing the magnetic
data of 9. The interaction between the MnII and MnIII ions
through the phenoxo/alkoxo bridging groups is very weak
compared to that found in complexes 1−6, which can be due to
the shift of the central MnII atom with respect to the plane
containing the four oxygen bridging atoms. This distortion
would reduce the more significant ferromagnetic contribution
to the magnetic coupling (dxy ↔ dx2−y2).
To support the above conclusion we have performed DFT
calculations (see Supporting Information) on the model
compound given in Figure 13, in which the Cphenyl−O bond
The field dependence of the magnetization matches well with
the Brillouin function for an S = 5/2 ground state with g = 2.11,
which would appear as a consequence of the stronger
antiferromagnetic interaction through the chloride bridge as
compared with that mediated by the alkoxo/phenoxo pathway
(Figure 14). The fact that the magnetization is not fully
Figure 13. Calculated J values vs the Mn2+−O−Mn3+ angle (θ) and
calculated J values vs δ (inset) for the model compound indicated in
the inset (top left).
involving the methoxy group coordinated to the Mn2+ ion and
linked to the phenyl ring has been broken and the methoxy
group has been replaced by a coordinated water molecule.
Moreover, the Mn2+−O and the Mn3+−O bonds in the
MnII(O)2MnIII bridging fragment have been set to the values
Mn(III)−O = 1.917 Å and Mn(II)−O = 2.115 Å. In these
calculations, the MnII(O)2MnIII bridging fragment was kept
with the same angle as in complex 1 and the θ angle varied in
the 90−105° range. In all cases, the phenyl ring was turned
away from the MnII(O)2MnIII bridging plane by 70° to avoid
steric hindrance with neighboring parts of the molecule. The
DFT results (see Figure 13) clearly show that a relationship
exists between J and θ, so that for angles larger than ∼97° (the
crossover point below which the magnetic interaction changes
from ferromagnetic to antiferromagnetic) the magnetic
exchange interaction is ferromagnetic in nature and its
magnitude increases when θ increases reaching a value of
+20.9 cm−1 at 105°. It should be pointed out that the imposed
geometrical changes in the model lead to calculated J values
that are of the same order of magnitude but higher than the
experimental and calculated values for compounds 1−6. To
know how the distortion of the SP geometry toward the
TBPY-5 geometry of the Mn2+ coordination polyhedron affects
the J values we have carried out calculations on the model
compound for θ = 100° by gradually changing the δ para-
meter29 for the Mn2+ coordination polyhedron from δ = 0
(ideal SPY-5) to δ = 0.63 (geometry closer to TBPY-5). The
three model compounds with different δ values are given in
Figure S11. The DFT results (Figure 13, inset) clearly show
Figure 14. M vs H plot for complex 9. The solid line corresponds to
the Brillouin function for S = 5/2.
saturated at high field may be due to the presence of a
significant magnetic anisotropy and/or more likely the presence
of low-lying excited states that are partially (thermally and field-
induced) populated.
The spin density distributions for the decaplet ground state
in 1−6 and the sextuplet ground state in 9 (the spin density of
2, 4, 8, and 9 are given as examples in Figures S12−S15)
confirm the predominance of the delocalization mechanism
through the σ type exchange pathways involving the magnetic
orbitals of the Mn2+ and Mn3+ atoms and the p orbitals of the
phenoxo/alkoxo bridging groups (and the chloride bridging group
in the case of 9). It should be noted that there indeed exists spin
density on the two semicoordinated methoxo groups in compound
9, with long Mn−O distances of 2.891 Å and 2.872 Å. A similar
but lower spin density is delocalized on the methoxy groups that
are located at ∼3 Å in the trinuclear compounds 7 and 8.
2239
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
Dynamic ac magnetic measurements were performed to know if
complexes 1−6 exhibit single-molecule magnet behavior. In spite
of the significant anisotropy and S = 9/2 ground state for these
compounds, they do not show any maximum in out-of-phase
susceptibility signals (χM″) above 2 K, even in the presence of an
small dc applied field. Hence, these complexes do not exhibit slow
relaxation of the magnetization and, therefore, SMM behavior.
In this study we have prepared several novel mixed-valence
di- and trinuclear manganese(II,III) complexes which hold
interesting magnetic properties. In all dinuclear complexes the
exchange interaction between Mn3+ and Mn2+ ions is proven to
be ferromagnetic. The prepared complexes can be further
modified to produce new complexes with higher nuclearity
and/or different bridging groups and diverse magnetic behavior
depending on the structural parameters of the bridging
network. The analysis of the magnetostructural data for the
complexes reported in this paper and other closely related
diphenoxo, phenoxo/alkoxo, and dialkoxo doubly bridged
dinuclear and trinuclear MnIIMnIII mixed valence complexes
seems to indicate that, even though the J values are not very
sensible to the Mn2+−O−Mn3+ angle (θ), for θ > ∼100°
ferromagnetic interactions increase when θ increases and the
distortion of the Mn2+ polyhedron from ideal square pyramidal
toward trigonal bipyramidal geometry decreases. These
magnetostructural correlations have been supported by DFT
calculations on model compounds. Further work is in progress
to use the obtained MnIIMnIII mixed valence complexes as
tunable building blocks in the difficult task of synthesizing
single molecule magnets and other magnetic materials.
REFERENCES
■
(1) Dismukes, G. C. Chem. Rev. 1996, 96, 2909.
(2) Kono, Y.; Fridovich, I. J. Biol. Chem. 1983, 258, 13646.
(3) Barynin, V. V.; Whittaker, M. M.; Antonyuk, S. V.; Lamzin, S. V.;
Harrison, P. M.; Artymiuk, A. J.; Whittaker. J. Mol. Struct. 2001, 9, 725.
(4) Reczkowski, R. S.; Ash, D. E. J. Am. Chem. Soc. 1992, 114, 10992.
(5) Cammack, R.; Chapman, A.; Lu, W.-P.; Karagouni, A.; Kelly, D.
P. FEBS Lett. 1989, 253 (1−2), 239.
(6) Carrel, H. L.; GLusker, J. P.; Burger, V.; Manfre, F.; Tritsch, D.;
Biellmann, J.-F. Proc. Natl. Acad. Sci 1989, 86, 4440.
(7) Whitlow, M.; Howard, A. J; Finzel, B. C.; Poulos, T. L.;
Winborne, E.; Gilliland, G. L. Proteins 1991, 9, 153.
(8) Collyer, C. A.; Henrick, K.; Blow, D. M. J. Mol. Biol. 1990, 212,
211.
(9) (a) Cox, N.; Ogata, H.; Stolle, P.; Reijerse, E.; Auling, G.; Lubitz,
W. J. Am. Chem. Soc. 2010, 132, 11197. (b) Cotruvo, J. A.; Stubbe, J.
Biochemistry 2010, 49, 1297.
(10) Dismukes, G. C.; Siderer, Y. Proc. Natl. Acad. Sci. U.S.A. 1981,
78 (1), 274.
(11) Sessoli, R.; Gatteschi, D.; Caneschi, A.; Novak, M. A. Nature
1993, 365, 141.
(12) For some recent reviews see: (a) Gatteschi, D.; Sessoli, R.
Angew. Chem., Int. Ed. 2003, 42, 268. (b) Christou, G. Polyhedron
2005, 24, 2065. (c) Gatteschi, D.; Sessoli, R.; Villain, J. Molecular
Nanomagnets; Oxford University Press: Oxford, UK, 2006. (d) Aromí,
G.; Brechin, E. K. Struct. Bonding (Berlin) 2006, 122, 1. (e) Rebilly, J.-
N.; Mallah, T. Struct. Bonding (Berlin) 2006, 122, 103. (f) Cornia, A.;
Costantino, A. F.; Zobbi, L.; Caneschi, A.; Gatteschi, D.; Mannini, M.;
Sessoli, R. Struct. Bonding (Berlin) 2006, 122, 133. (g) Milios, C. J.;
Piligkos, S.; Brechin, E. K. Dalton Trans. 2008, 1809. (h) Bagai, R.;
Christou, G. Chem. Soc. Rev. 2009, 38, 1011. (i) Sessoli, R.; Powell, A.
K. Coord. Chem. Rev. 2009, 253, 2328. (j) Andruh, M.; Costes, J. P.;
Diaz, C.; Gao, S. Inorg. Chem. 2009, 48, 3342 (Forum Article). (k)
Molecular Magnets themed issue: Brechin, E. K. Ed.; Dalton Trans.
2010. (l) Bagai, R.; Christou, G. Chem. Soc. Rev. 2009, 38, 1011.
(13) (a) Wernsdorfer, W.; Sessoli, R. Science 1999, 284, 133.
(b) Leuenberger, M. N.; Loss, D. Nature 2001, 410, 789. (c) Meier, F.;
Loss, D. Physica B 2003, 329, 1140.
ASSOCIATED CONTENT
■
S
* Supporting Information
X-ray crystallographic data for all complexes in CIF format
(CDCC 911571−911580), summary of the crystallographic
data and selected geometrical parameters for 7−9. Spectro-
1
(14) Winpenny, R. E. P. J. Chem. Soc., Dalton Trans. 2002, 1.
(15) Mandal, S.; Rosair, G.; Ribas, J.; Bandyopadhyay, D. Inorg. Chim.
Acta 2009, 362, 2200.
scopic data for complexes 1−6 and 9, H NMR spectra and
elemental analysis results for ligands are included. This material
is available free of charge via the Internet at http://pubs.acs.org.
(16) Glaser, T. Chem. Commun. 2011, 47, 116.
(17) Otwinowski, Z.; Minor, W. In Methods in Enzymology: Part A;
Carter, C. W., Sweet, R. M., Eds.; Academic Press: New York, 1997;
Vol. 276, pp 307−326.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: ecolacio@ugr.es (E.C.), resillan@jyu.fi (R.S.).
(18) Sheldrick, G. M. SADABS; University of Gottingen: Germany,
̈
2002.
Notes
(19) CrysAlisPro; Agilent Technologies Ltd: Yarnton, England, 2011.
(20) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112.
(21) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.;
Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.;
Spagna, R. J. Appl. Crystallogr. 1999, 32, 115.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to Inorganic Materials Chemistry Graduate
Program for financing the project and to Elina Hautakangas for
performing the elemental analysis. E.C. and A.J.M. thanks to
(22) DIAMOND, J. Appl. Crystallogr.1999, 32, 1028.
(23) Riisio, A.; Wichmann, O.; Sillanpaa, R. Lett. Org. Chem. 2010, 7,
̈
̈
̈
298.
(24) (a) Hirotsu, M.; Kojima, M.; Mori, W.; Yoshikawa. Bull. Chem.
Soc. Jpn. 1998, 71, 2873. (b) Hirotsu, M.; Kojima, M.; Yoshikawa, Y.
Bull. Chem. Soc. Jpn. 1997, 70, 649. (c) Schmitt, H.; Lomoth, R.;
the Ministerio de Educacion Cultura y Deporte (Spain)
́
(Projects CTQ-2008-02269/BQU and CTQ2011-24478), the
́
Junta de Andalucia (FQM-195) and the Universidad de
Magnuson, A.; Park, J.; Fryxelius, J.; Kritikos, M.; Martensson, J.;
̊
Granada for financial support. We would like to thank the
Centro de Supercomputacion de la Universidad de Granada for
́
Hammarstrom, L.; Sun, L.; Åkermark, B. Chem.Eur. J. 2002, 8, 3757.
̈
(d) Anderlund, M. F.; Zheng, J.; Ghiladi, M.; Kritikos, M.; Riviere, E.;
Sun, L.; Girerd, J.-J.; Akermark, B. Inorg. Chem. Commun. 2006, 9,
1195.
computational resources. We would like also to thank Victor K.
Abdelkader, University of Granada, for deconvoluting the XPS
spectra. F.L. thanks to the MICINN (Spain) (Project
CTQ2010-15364), the University of Valencia (Project UV-
INVAE11- 38904) and the Generalitat Valenciana (Spain)
(Projects PROMETEO/2009/108, GV/2012/051 and ISIC/
2012/002) for financial support.
(25) Gom
́
ez-Alcan
́
tara, M. M.; Aranda, M. A. G.; Olivera-Pastor, P.;
̃
Beran, P.; García-Munoz, J. L.; Cabeza, A. Dalton Trans. 2006, 577.
(26) Llunell, M.; Casanova, D.; Cirera, J.; Bofill, J. M.; Alemany, P.;
Alvarez, S.; Pinsky, M.; Avnir, D. SHAPE v1.1b; University of
Barcelona: Barcelona, 2005.
2240
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241

Inorganic Chemistry
Article
(27) (a) Taylor, R.; Kennard, O. J. Am. Chem. Soc. 1982, 104, 5063.
(b) Desiraju, G. R. Acc. Chem. Res. 1991, 24, 290. (c) Steiner, T. New J.
Chem. 1998, 1099.
(28) Kahn, O. Molecular Magnetism; VCH: Weinheim, Germany,
1993.
(29) Addison, A. W.; Rao, T. N.; Reedijk, J.; Rijn, J.; Verschoor, G. C.
J. Chem. Soc., Dalton Trans. 1984, 1349.
(30) (a) Goodenough, J. B. Phys. Rev. 1955, 100, 564. (b) J.
Kanamori, J. Phys. Chem. Solids 1959, 10, 87.
(31) Liu, H.; Tian, J.; Kou, Y.; Zhang, J.; Feng, L.; Li, D.; Gu, W.; Liu,
X.; Liao, D.; Cheng, P.; Ribas, J.; Yan, S. Dalton Trans. 2009, 10511.
(32) Bashkin, J. S.; Schake, A. R.; Vincent, J. B.; Chang, H.-R.; Li, Q.;
Huffman, J. C.; Christou, G.; Hendrickson, D. N. J. Chem. Soc. Chem.
Commum. 1988, 700.
(33) Gou, S.; Zeng, Q.; Yu, Z.; Qian, M.; Zhu, J.; Duan, C.; You, X.
Inorg. Chim. Acta 2000, 175.
(34) Gelasco, A.; Kirk, M. L.; Kampf, J. W.; Pecoraro, V. L. Inorg.
Chem. 1997, 36, 1829.
(35) (a) Bossek, U.; Hummel, H.; Weyhermuuller, T.; Wieghardt, K.;
̈
Russell, S.; van der Wolf, L.; Kolb, U. Angew. Chem. 1996, 108, 1653.
(b) Cox, N.; Ames, W.; Epel, B.; Kulik, L. V.; Rapatskiy, L.; Neese, F.;
Messinger, J.; Wieghardt, K.; Lubitz, W. Inorg. Chem. 2011, 50, 8238.
(36) Diril, H.; Chang, H.-R.; Zhang, X.; Larsen, S. K.; Potenza, J. A.;
Pierpont, C. G.; Schugar, H. J.; Isied, S. S.; Hendrickson, D. N. J. Am.
Chem. Soc. 1987, 109, 6207.
(37) Chang, H.-R.; Diril, H.; Nilges, M. J.; Zhang, X.; Potenza, J. A.;
Schugar, H. J.; Hendrickson, D. N.; Isied, S. S. J. Am. Chem. Soc. 1988,
110, 625.
(38) Prescimone, A.; Wolowska, J.; Rajaraman, G.; Parsons, S.;
Wernsdorfer, W.; Murugesu, M.; Christou, G.; Piligkos, S.; McInnes,
E. J. L.; Brechin, E. K. Dalton Trans. 2007, 5282.
(39) Yoshino, A.; Miyagi, T.; Asato, E.; Mikuriya, M.; Sakata, Y.;
Sugiura, K.; Iwasaki, K.; Hino, S. Chem. Commun. 2000, 1475.
(40) Hirotsu, M.; Aoyagi, M.; Kojima, M.; Mori, W.; Yoshikawa, Y.
Bull. Chem. Soc. Jpn. 2002, 75, 259.
(41) Mandal, D.; Chatterjee, P. B.; Bhattacharya, S.; Choi, K.-Y.;
́
Clerac, R.; Chaudhury, M. Inorg. Chem. 2009, 48, 1826.
(42) Hirotsu, M.; Kojima, M.; Yoshikawa, Y. Bull. Chem. Soc. Jpn.
1997, 70, 649.
(43) (a) Zener, C. Phys. Rev. 1951, 82, 403. (b) Girerd, J.-J. J. Chem.
Phys. 1983, 79, 1766.
(44) (a) Borras
Tsukerblat, B. S. Inorg. Chem. 1999, 38, 6081. (b) Borras
J. J.; Clemente, J.; Coronado, E.; Tsukerblat, B. S. J. Comput. Chem.
2001, 22, 985.
́
-Almenar, J. J.; Clemente, J.; Coronado, E.;
́
-Almenar,
2241
dx.doi.org/10.1021/ic302731z | Inorg. Chem. 2013, 52, 2228−2241