Magnetic properties of 1:4 complexes of CoCl2 and pyridines carrying carbenes (S0 = 4/2, 6/2, and 8/2) in diluted frozen solution; Influence of carbene multiplicity on heterospin single-molecule magnets

Article abstract of DOI:10.1039/c2dt31288c

The microcrystalline sample of a parent complex, [CoCl₂(py) ₄], showed a single-molecule magnet (SMM) behavior with an effective activation barrier, U_eff/k_B, of 16 K for reversal of the magnetism in the presence

Full text of DOI:10.1039/c2dt31288c

View Online / Journal Homepage
Dynamic Article Links
Dalton
Transactions
Cite this: DOI: 10.1039/c2dt31288c
www.rsc.org/dalton
PAPER
Magnetic properties of 1 : 4 complexes of CoCl2 and pyridines carrying
carbenes (S0 = 4/2, 6/2, and 8/2) in diluted frozen solution; influence of
carbene multiplicity on heterospin single-molecule magnets†
Satoru Karasawa, Kimihiro Nakano, Jun-ichi Tanokashira, Noriko Yamamoto, Takahito Yoshizaki and
Noboru Koga*
Received 15th June 2012, Accepted 18th July 2012
DOI: 10.1039/c2dt31288c
The microcrystalline sample of a parent complex, [CoCl2(py)4], showed a single-molecule magnet
(SMM) behavior with an effective activation barrier, Ueff/kB, of 16 K for reversal of the magnetism in the
presence of a dc field of 3 kOe. Pyridine ligands having 2–4 diazo moieties, DYpy; Y = 2, 3l, 3b, and 4,
were prepared and confirmed to be quintet, septet, septet, and nonet in the ground state, respectively, after
irradiation. The 1 : 4 complexes, CoCl2(DYpy)4; Y = 2, 3l, 3b, and 4 in frozen solutions after irradiation
showed the magnetic behaviors of SMMs with total spin multiplicity, Stotal = 17/2, 25/2, 25/2, and 33/2,
respectively. Hysteresis loops depending on the temperature were observed and the values of coercive
force, Hc, at 1.9 K were 12, 8.4, 11, and 8.1 kOe for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, respectively. In
dynamic magnetic susceptibility experiments, ac magnetic susceptibility data obeyed the Arrhenius law to
give Ueff/kB values of 94, 92, 93, and 87 K for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, respectively, while
the relaxation times for CoCl2(CYpy)4; Y = 2 and 3l, obtained by dc magnetization decay in the range of
3.5–1.9 K slightly deviated downward from Arrhenius plots on cooling. The dynamic magnetic behaviors
for CoCl2(CYpy)4 including [CoCl2(py)4] and CoCl2(C1py)4 suggested that the generated carbenes
interacted with the cobalt ion to increase the relaxation time, τq, due to the spin quantum tunneling
magnetization, which became larger with increasing Stotal of the complex.
magnetism presents an interesting challenge in the field of the
Introduction
molecule-based magnet. For the construction of SMM, we have
been using a heterospin system5,6 consisting of the 3d spins of
One of the striking characteristics of the single-molecule
magnets (SMMs)1–4 exhibiting a slow magnetic relaxation is the
the metal ions and the 2p organic spins. In the heterospin
size as a magnet. Because in SMMs, an individual molecule
functions as a magnet, the size of the magnet depends on the
molecular size. Furthermore, an important physical property of
SMMs is the quantum tunneling of magnetization (QTM)2
observed at extremely low temperature. From these unique mag-
netic properties, SMMs are expected to allow the development
of new magnetic materials. To date, many complexes and clus-
ters containing various kinds of 3d3 and 4f4 metal ions have
been reported as SMM. At the present stage, the construction of
a SMM with a high activation barrier for the reversal of
system, the combination of the anisotropic high-spin cobalt(II)
ion and aminoxyl- and carbene-pyridine ligands (4NOpy and
C1py, respectively) were selected and those complexes success-
fully provided heterospin SMM with relatively large effective
activation barriers, Ueff/kB. In the simple 1 : 4 cobalt(II) com-
plexes, [Co(X)2(Y)4]; X = counter anion and Y = 4NOpy and
C1py, the magnetic properties of heterospin SMMs were found
to depend on the axial ligand, X, and the equatorial one, Y. The
5a
complexes, CoX2(4NOpy)4 and CoX2(C1py)4,5b behaved as
SMMs in frozen solution; the Ueff/kB values were 50, 31, and
20 K for X = NCO−, NCS−, and Br−, respectively, in
CoX2(4NOpy)4 and 130, 89, and 91 K for X = NCO−, NCS−,
and Cl−, respectively, in CoX2(C1py)4.
In order to understand the heterospin SMM properties, this
time, the influence of the spin multiplicity of carbene spin, S0, at
the equatorial ligands on the SMM properties was investigated.
In the cobalt-carbene complex, it is possible to vary the total
S value, Stotal, of the SMM complex systematically by altering
the number of carbenes. Carbene, being a divalent carbon, is a
reactive intermediate in fundamental organic chemistry, and
especially diphenylcarbene, having a triplet ground state, has
Graduate School of Pharmaceutical Sciences, Kyushu University, 3-1-1
Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan.
E-mail: koga@fc.phar.kyushu-u.ac.jp; Fax: +81-92-642-6590;
Tel: +81-92-642-6590
†Electronic supplementary information (ESI) available: UV-Vis spectra
changes of D2bpy in solution on photolysis at 10 K (S1), M vs.
irradiation time plot for D2py and CoCl2(D2py)4 (S2), data of slow
relaxation of magnetization for [CoCl2(py)4] (S3), and plots of
hysteresis loops (S4), χ′mol T vs. T (S5), dc decay (S6), and τ vs. T−1(S7)
for 1 : 4 mixtures of CoCl2 and CYpy; Y = 2, 3l, 3b, and 4. See DOI:
10.1039/c2dt31288c
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Online
been used as a building block for high-spin organic molecules.
Based on polycarbene studies,7 the diazo-pyridine derivatives,
D2py, D3py, and D4py, connecting two, three, and four diazo
moieties with ferromagnetic m-phenylene linkers, were designed
as precursors of high-spin carbene. The carbene centers gener-
ated by photolysis of D2py, D3py, and D4py interact ferromag-
netically to form high-spin carbenes, C2py, C3py, and C4py,
with S0 = 4/2, 6/2, and 8/2, respectively. Since the carbene at the
4-position of the pyridine ring ferromagnetically interacts with
the high-spin cobalt(II) ion with the effective spin, Seff = 1/2,8
the 1 : 4 complexes CoX2(CYpy)4, where Y = 1, 2, 3l, 3b, and 4
3-formyldiphenylmethane to give a carbinol 6. After reduction
to methylene with trimethylsililchloride–sodium hydride,10
bromide 7 was lithiated with n-butyllithium followed by the
reaction with 4-formylpyridine to afford 8. Hydroxy derivatives,
4, 5, and 8, were oxidized with sodium perchromate(VI) to the
corresponding ketones, OYpy; Y = 2, 3l, and 4. The dilithiation
of 1,3,5-tribromobenzene with tert-butyllithium followed by the
reaction with 4-tert-butylbenzaldehyde gave a mixture of dihy-
droxy derivative 9 containing a monoketo–monohydroxy deriva-
tive. A mixture was reduced with sodium borohydride to
dihydroxy derivative 9, and then by the reaction with trimethyl-
sililchloride–sodium hydride to bromide 10. Similarly, bromide 10
was lithiated and was reacted with 4-formylpyridine to afford 11.
Hydroxy derivative 11 was oxidized stepwise; first, with manga-
nese(IV) dioxide to monoketone 12, and after the purification,
secondly oxidized with sodium perchromate(VI) to triketone
O3bpy. The direct oxidation with sodium perchromate(VI) from
11 to triketone gave a poor yield. Bisdiazo-, tridiazo-, and tetra-
diazo-pyridine ligands, DYpy; Y = 2, 3l, 3b, and 4, were pre-
pared from the corresponding ketone derivatives by a standard
procedure;6f,7 hydrazone formation, followed by oxidation with
freshly prepared active manganese dioxide. Diazo derivatives
DYpy were purified by column chromatography (Al2O3) and
then by recrystallization from Et2O, Et2O–CH2Cl2, and Et2O, to
afford red single crystals for Y = 1, 2, and 3b, and reddish
powders for Y = 3l and 4, which were stable in the dark.
In this study, for the investigation of the equatorial ligand
effect, the chloride ion was used at the axial ligand. The 1 : 4
cobalt(II) complexes, CoCl2(DYpy)4; Y = 1, 2, 3l, 3b, and 4,
were prepared by mixing the solution of CoCl2 in EtOH with the
solution of DYpy in CH2Cl2 at 1 to 4 ratios. Attempts to obtain
the single crystals of [CoCl2(DYpy)4] were carried out under
various conditions. However, all attempts gave the 1 : 4 com-
plexes as red powders. The crystal of the parent 1 : 4 complex,
[CoCl2(py)4], which was an elongate octahedron in the molecu-
lar structure, was obtained by the procedure reported
previously.5g
are expected to produce the high-spin ground state with Stotal
=
9/2, 17/2, 25/2, 25/2, and 33/2, respectively.
In this study, the chloride ion was selected as the axial ligand
and a frozen solution condition was used for the effective pho-
tolysis of diazo moieties in the complex. The ac and dc magnetic
susceptibility measurements for the 1 : 4 complexes of CoCl2-
(DYpy)4; Y = 2, 3l, 3b, and 4, in frozen solution after irradiation
were carried out. In addition, the magnetic property of the parent
1 : 4 complex, [CoCl2(py)4], having no diazo unit was also
investigated in the microcrystalline state.
Results and discussion
Preparations of di-, tri-, tetradiazo-pyridine derivatives and
their cobalt(II) complexes
Based on the studies of the high-spin polycarbenes,7 mono-, di-,
tri-, and tetracarbene-monopyridine derivatives were designed, in
which a pyridine and carbene centers served as binding ligands
for the metal ion and organic spin source, respectively. Each
carbene center was connected with m-phenylene ferromagnetic
coupler and the carbene center was located at the 4-position of
the pyridine ring. The diazo-pyridine derivatives, DYpy, which
were precursors of corresponding high-spin carbenes, were pre-
pared by the modified procedure reported previously.6f,7 Two iso-
meric tridiazo-pyridine derivatives, D3bpy and D3lpy, which
had branched and linear structures, respectively, were designed
and prepared. In the branched one, tert-butyl groups were intro-
duced at the p-position of the benzene ring to increase the solu-
bility in organic solvent after the complexation with cobalt ion.
Preparations of DYpy; Y = 2, 3l, 3b, and 4, are summarized in
Scheme 1.
Measurements of Vis-NIR spectra
(A) DYpy; Y = 1, 2, 3l, 3b and 4. UV-Vis spectra for diazo-
pyridine, D1py in MTHF solution (1 mM) at room temperature
showed two absorptions at 283 and 500 (ε = 1.2 × 102) nm. The
latter one is characteristic of the n–π* transition due to the diazo
group. When the irradiation (λ = 514 nm) started at 10 K, this
absorption decreased and two new absorptions appeared at 508
and 471 nm due to π–π* transition. Similarly, diazo-pyridines,
D2py, D3bpy, D3lpy, and D4py had absorptions at 508 (ε = 2.1
× 102), 512 (3.1 × 102), 510 (3.6 × 102), and 510 (4.2 × 102) nm
due to the n–π* transition and their absorption coefficiencies,
(ε, were proportional to the number of diazo moieties). After
irradiation under similar conditions, new broad absorptions were
observed at 411 and 476 nm for C2py, 406 and 471 nm for
C3lpy, 411 and 476 nm for C3bpy, and 420 and 472 nm for
C4py. The spectra after irradiation closely resembled that for
diphenylcarbene (465 nm) reported previously.5g,6f All new
absorptions observed after irradiation disappeared at tempera-
tures higher than 60 K. The absorption change for D2py on
photolysis is shown in Fig. S1.†
The lithiation of starting bromo derivatives, 1 and 2, which
were prepared by the procedure reported previously,5g,7,9 with
n-butyllithium was followed by the reaction with 4-formylpyri-
dine to afford carbinol 4 and 5. Grignard reagent prepared from
3,3′-bromodiphenylmethane 3 and magnesium reacted with
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Online
Scheme 1 Preparation routes for DYpy; Y = 2, 3l, 3b, and 4.
(B) CoCl2(DYpy)4; Y = 1, 2, 3b, 3l, and 4. In order to inves-
tigate the coordination structure of the complexes formed in
frozen solutions, Vis-NIR spectra of 1 : 4 mixtures (5 mM) of
CoCl2 and DYpy in MTHF were measured in the temperature
range 280–175 K. The spectra of the 1 : 4 mixtures at room
temperature showed two sets of absorptions at 600–700 and
∼1000 nm and at ∼500 nm. The latter absorption maximum was
not clear due to overlapping with that (510 nm) due to n–π*
transition of diazo moieties of DXpy. The former and the latter
absorptions are characteristic of the d–d transitions of the cobalt(II)
ion in tetrahedral and octahedral structure, respectively.11 On
cooling to 200 K, absorptions for the former gradually decreased
and the latter at ∼500 nm increased (Fig. 1a). At 200 K, the
former absorptions disappeared completely. The observed spec-
tral change indicated that the tetrahedral and octahedral species
were equilibrated with the former and the latter dominant at
room temperature and temperature below 200 K, respectively.5b
The observed thermal spectral changes on cooling suggested that
the 1 : 4 mixture in frozen solution can be safely considered to
be octahedral as observed in X-ray crystallography.5b,12 The sol-
ution of the sample was cooled down to 10 K and was irradiated
by argon laser (λ = 514 nm). After irradiation for 2 h, a new
absorption at 400–600 nm appeared at the expense of the one at
∼500 nm for diazo moieties of DYpy. This new absorption
observed after irradiation was characteristic for the carbene and
disappeared after annealing at 70 K.
The changes of Vis spectra for the sample of a 1 : 4 mixture of
CoCl2 and D2py in MTHF solution on cooling from 280 to
175 K and on photolysis at 10 K are shown in Fig. 1(a) and
1(b), respectively. The other combinations of CoCl2 and DYpy
showed similar spectral changes on cooling and on photolysis at
10 K.
Magnetic properties
Photolysis of the frozen solution samples. Photolysis of the
frozen solution samples were performed in SQUID apparatus by
the light from an argon laser (50 mW, λ = 514 nm) through the
optical fiber and were followed by magnetization (Mmol
)
measurement at the constant field of 5 kOe and at 5 K. When the
irradiation started, the Mmol values gradually increased and
levelled off after ca. 10–30 h. The observed increase in the Mmol
value indicates that the diazo moieties were photolyzed to form
the carbene. The plots of Mmol vs. irradiation time for the
samples of D2py and the 1 : 4 mixture of CoCl2 and D2py are
shown in Fig. S2.† The degrees of the photolysis in frozen solu-
tion were determined by the consumption of diazo moiety in IR
spectra after SQUID measurements. DYpy and CoCl2(DYpy)4
showed quantitative and over ca. 95% photolysis, respectively.
(A) DYpy; Y = 2, 3l, 3b and 4. 2–10 mM solutions of DYpy
in MTHF were used for the samples. In order to determine the
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Online
Fig. 2 Mmol vs. H/T plots at 2.0 (open) and 5.0 K (filled) for D2py
(red circle), D3lpy (blue square), D3bpy (green triangle), and D4py
(orange reversed triangle) in MTHF frozen solution after irradiation.
Solid lines show the theoretical curves with S = 4/2, 6/2, and 8/2.
Fig. 1 Vis spectra changes of a 1 : 4 mixture of CoCl2 and D2py in
MTHF solution (a) on cooling in the temperature range of 280–175 K
before irradiation and (b) on photolysis at 10 K. Inset in (a) shows the
spectra in the near IR region. Arrows indicate the increase and decrease
of the absorption.
ground state multiplicity after irradiation of DYpy, the field
dependencies of magnetization before and after irradiation at
2 and 5 K were measured. The Mmol vs. H/T plot for DYpy; Y =
2, 3l, 3b, and 4, in frozen solution, in which Mmol = Ma − Mb
(Ma and Mb are the molar magnetization after and before
irradiation, respectively), are shown in Fig. 2 together with theor-
etical curves with S = 4/2, 6/2, and 8/2, respectively.
All experimental data nearly traced on the theoretical curves7,13
for the corresponding S values, indicating that carbene centers
generated by photolysis interacted ferromagnetically to form the
high-spin species in the quintet, septet, septet, and nonet ground
state for CYpy; Y = 2, 3l, 3b, and 4, respectively.14,15 After
annealing above 80 K, the magnetization data completely
returned to the levels before irradiation.
(B) CoCl2(py)4 and CoCl2(DYpy)4; Y = 2, 3l, 3b, and 4. The
magnetic properties of [CoCl2(py)4] having no organic spin in
the crystalline state and those before and after irradiation of 1 : 4
complexes, CoCl2(DYpy)4; Y = 2, 3l, 3b, and 4, in frozen solu-
tions were investigated by SQUID magneto/susceptometry. The
magnetic property for CoCl2(D1py)4 under frozen conditions
was already reported.5b
Fig. 3 Plots of (a) χmol T vs. T at dc field of 5 kOe and (b) χ′′mol vs. T
with a 5 Oe ac field oscillating at 1000 (red), 750 (blue), 500 (black),
400 (green), 250 (purple) and 100 (brown) Hz in the presence of a
3 kOe dc field for a microcrystalline sample of [CoCl2(py)4].The solid
lines for (a) and (b) indicate the fitting curves and the visual guides,
respectively.
(B-1) CoCl2(py)4. The values of dc molar magnetic suscepti-
bility, χmol, for a microcrystalline sample of [CoCl2(py)4] were
collected at a constant field of 5 kOe. The obtained χmol values
were plotted as a function of temperature. The ac magnetic sus-
ceptibilities were measured in the absence and presence of a 3
kOe dc field with a 5.0 Oe ac field at the frequencies of 1000,
750, 500, 400, 250 and 100 Hz in the temperature range of
1.9–6.5 K. The plots of χmolT vs. T and χ′′mol vs. T in the pres-
ence of dc field are shown in Fig. 3.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Online
In the χmolT vs. T plot, the χmolT values were constant (2.7 cm3
K mol−1) in the temperature range of 300–150 K, gradually
decreased below 150 K, reached nearly a constant of 2.0 cm3 K
mol−1 in the range 20–5 K, and then decreased below 5 K. The
observed thermal profile of the χmolT was analyzed by the fol-
lowing Hamiltonian (eqn (1)) in the MAGSAKI program16,17 to
afford Δ = −197 K, κ = 0.79, and λ = −887 K.
H ¼ ΔðL2z ꢀ 2=3Þ ꢀ ð3=2ÞκλL ꢁ S þ β½ꢀð3=2ÞκL þ geSꢂ ꢁ H
ð1Þ
in which Δ is the axial splitting parameter, κ is the orbital
reduction factor, and λ is the spin-orbit coupling parameter. The
obtained values were comparable to those for a high spin cobalt(II)
having an octahedral structure.5,8,16,17 The fitting result is shown
in Fig. 3a.
Fig. 4 χmolT vs. T plots below 30 K before (cross) and after irradiation
of 1 : 4 mixtures of CoCl2 and DYpy; Y = 15b (red circle), 2 (blue
square), 3l (green reversed triangle), 3b (black triangle), and 4 (purple
filled circles) in frozen MTHF solutions.
In ac magnetic susceptibility measurements, the χ′mol and
χ′′mol signals (in-phase and out-of-phase components of ac mag-
netic susceptibilities, respectively) and their frequency depen-
dence were observed, indicating that slow magnetic relaxation
existed in [CoCl2(py)4]. In the χ′′mol vs. T plots, frequency-
dependent χ′′mol signals with no maximum were observed above
1.9 K. To suppress the pathway due to the spin quantum tunnel-
ing magnetization (QTM), the dc external field was applied and
the ac magnetic susceptibility was measured under similar con-
ditions. The maxima of χ′′mol signals with individual frequency
were observed in the presence of dc field of 3 kOe (Fig. 3b). The
observed peak-top temperature of χ′′mol signals depended on the
frequency and it shifted to the higher temperature with increasing
the frequency. Because each frequency at the peak-top tempera-
ture for χ′′mol is consistent with τ−1, the effective activation
energy, Ueff/kB, for reversal of the magnetism and the pre-expo-
nential factor, τ0, can be obtained in terms of the equation, τ =
1/(2πν) = τ0exp(Ueff/kBT). From the Arrhenius plot, the Ueff/kB
value of 16 K in the presence of 3 kOe was obtained. These
results indicated that a 1 : 4 cobalt complex essentially had a
SMM property showing slow magnetic relaxation, which was
strongly affected by the QTM effect, and also suggested the role
of organic spin, carbene, in the following heterospin SMM com-
plexes. The χ′molT vs. T plot and Arrhenius plot for [CoCl2(py)4]
are shown in Fig. S3.†
Fig. 5 Hysteresis loops at 1.9 K after irradiation of 1 : 4 mixtures of
CoCl2 and DYpy; Y = 1 (red), 2 (blue), 3l (green), 3b (black), and 4
(purple) in frozen MTHF solution with a sweeping rate of 0.36 kOe s−1
.
(Cl)2(D1py)4 reported previously.5b After irradiation, the χmol
T
(B) CoCl2(DYpy)4; Y = 2, 3l, 3b, and 4. Solutions (5, 2.5, 2.5
and 1.5 mM) of the 1 : 4 mixtures of CoCl2 and DYpy; Y = 2,
3l, 3b, and 4, respectively, in MTHF were used as samples for
SQUID measurements.
values increased greatly and those at 10 K were 29, 46, 44, and
83 cm3 K mol−1 for CoCl2(DYpy)4; Y = 2, 3l, 3b, and 4,
respectively. The values at 10 K were much larger than the theor-
etical values (6, 14, 26, 42 cm3 K mol−1 for Y = 2, 3l, 3b, and
4, respectively) calculated by a spin-only equation with four iso-
lated high-spin carbenes and one high-spin cobalt(II) ion5b,8 with
effective spin S′eff = 1/2 and geff = 4.0, suggesting that the car-
benes and the cobalt ions in the complex interacted ferromagne-
tically to form a high-spin ground state. On cooling, the χmol
values were nearly constant in the range 10–20 K and gradually
decreased at lower temperatures. The decrease in the χmol
values below 5 K indicated an effect of the zero-field splitting
caused by spin-orbit coupling in the cobalt ion.8 The observed
thermal profiles were similar to that for CoCl2(C1py)4.
Static magnetic properties
(i)
χmolT vs. T plots. The values of χmol, before and after
irradiation of CoCl2(DYpy)4; Y = 2, 3l, 3b, and 4, in frozen solu-
tion were collected at a constant field of 5 kOe below 20–30 K.
The obtained χmol values were plotted as a function of tempera-
ture. The plots of χmolT vs. T for CoCl2(DYpy)4; Y = 2, 3l, 3b,
and 4, before and after irradiation are summarized in Fig. 4
together with the plot for CoCl2(D1py)4.
Before irradiation, the χmolT values for four samples were
comparable and were nearly constant (2.0 cm3 K mol−1) in the
temperature range of 1.9–20 K. The constant value of 2.0 cm3 K
mol−1 was consistent with those for [CoCl2(py)4] and Co
T
T
(ii) Hysteresis loops. The field dependencies of molar mag-
netization, Mmol, for all the samples of CoCl2(CYpy)4; Y = 2, 3l,
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Online
3b, and 4, were measured at a field scanning rate of 0.36 kOe
s−1 in the range of −50 to 50 kOe and in the temperature range
of 1.9–3.5 K. When the dc field increased from 0 to 50 kOe, the
Mmol values gradually increased until 50 kOe. The Mmol values
at 50 kOe were 6.3 × 104, 9.5 × 104, 9.7 × 104, and 1.2 × 105
cm3 Oe mol−1 for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, respecti-
vely, and they did not reach the saturation magnetization, Ms,
even at 50 kOe, suggesting that the complexes had a large mag-
netic anisotropy. All of the cobalt complexes, CoCl2(CYpy)4,
exhibited hysteresis with respect to the applied field. The loops
of hysteresis depended on the temperature; they appeared below
ca. 3.5 K and the area inside the loop increased upon decreasing
the temperature. The observed temperature-dependent hysteresis
loop is one of the magnetic behaviors typical to a SMM. The
temperature dependencies of hysteresis loops for CoCl2(CYpy)4;
Y = 2, 3l, 3b, and 4, are shown in Fig. S4† including a Hc, and
Mr vs. T plot (inset). The hysteresis loops at 1.9 K for CoCl2-
(CYpy)4; Y = 2, 3l, 3b, and 4, are summarized in Fig. 5 together
with CoCl2(C1py)4. The values of coercive force, Hc, and
remnant magnetization, Mr, at 1.9 K were 12 and 3.0 × 104, 8.4
and 3.6 × 104, 11 and 3.9 × 104, and 8.1 kOe and 5.1 × 104 cm3
Oe mol−1 for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, respectively,
and were larger than those (7.0 kOe and 1.1 × 104 cm3 Oe
mol−1, respectively) for CoCl2(C1py)4.5b
(DYpy)4; Y = 2, 3l, 3b, and 4 were investigated by ac and dc
magnetic susceptibility measurements.
(i) Ac magnetic susceptibility measurements. The ac mag-
netic susceptibilities of the MTHF solution samples for the 1 : 4
complexes, CoCl2(DYpy)4; Y = 2, 3l, 3b, and 4, after irradiation
were measured in a zero dc field with a 5.0 Oe ac field in the
temperature range of 1.9–10 K. After irradiation, the χ′mol and
χ′′mol signals with frequency dependence were clearly observed.
The intensities of both signals increased in the order of CoCl2-
(CYpy)4; Y = 2, 3l, 3b, and 4 and the relative intensity of χ′′mol
signals with χ′mol signal were in the region of 0.2–0.3, which is
consistent with that reported for SMM previously.9 Observations
of χ′mol and χ′′mol signals indicated that cobalt complexes have
slow magnetic relaxations for reversal of the magnetism.
The plots of χ′molT vs. T and χ′′mol vs. T after irradiation of the
1 : 4 mixtures of CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, are shown
in Fig. S5† and Fig. 6, respectively. In the plots of χ′molT vs. T,
the values of χ′molT for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, are
37.5, 55.1, 59.3, and 85.5 cm3 K mol−1 at 10 K, respectively,
and remained essentially constant on cooling until the onset of
the contribution of χ′′mol signals. The constant χ′molT values
were larger than those obtained in χmolT vs. T plots, suggesting
that in static measurements and saturation of the magnetic sus-
ceptibility might take place in the constant dc field of 5 kOe.
From the constant χ′molT value of 37.5 cm3 K mol−1
,
CoCl2(C2py)4 was estimated to be Stotal = 17/2 and g = 2.0. In
CoCl2(CYpy)4; Y = 3l, 3b, and 4, however, the obtained con-
stant χ′ molT values deviated from the theoretical ones (84.4 and
144 cm3 K mol−1 for CoCl2(CYpy)4; Y = 3 and 4, respectively,
Dynamic magnetic properties
The magnetization relaxation time in a high and a low tempera-
ture region after irradiation of the 1 : 4 complexes, CoCl2-
Fig. 6 χ′′mol vs. T plots after irradiation of a 1 : 4 mixture (5.0–1.5 mM) of CoCl2 and DYpy; Y = (a) 2, (b) 3l, (c) 3b, and (d) 4, in frozen MTHF sol-
ution with a 5 Oe ac field oscillating at 1000 (red), 500 (blue), 100 (black), 10 (green), 5 (deep red) and 1(brown) Hz. The solid lines are visual
guides.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Online
assuming g = 2.0). The observed discrepancy might be caused
by the intramolecular antiferromagnetic interaction between the
polycarbene chains and/or the contribution of the thermal
excited state, which is often observed in the complexes with
large spin multiplicity.2c,18 Below 8 K, χ′molT values decreased
with frequency-dependence, while the χ′′mol signals appeared at
the same temperature. In χ′′mol vs. T plots for CoCl2(CYpy)4;
Y = 2, 3l, 3b, and 4, the χ′′mol signal intensities increased with
increasing Stotal and the peak-top temperatures for each fre-
quency were nearly constant. From χ′′mol vs. T plots, the values
of Ueff/kB and τ0 were estimated to be 94 and 6.2 × 10−11, 92
and 6.6 × 10−11, 93 and 4.2 × 10−11, and 87 K and 1.0 × 10−10
s, for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, respectively. Arrhe-
nius plots for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, are shown in
Fig. 8 and Fig. S7.† In addition, the blocking temperature, TB,
defined by the peak-top temperature at 0.01 Hz in the Arrhenius
plot was estimated to be 3.3, 3.2, 3.3, and 3.2 K for CoCl2-
(CYpy)4; Y = 2, 3l, 3b, and 4, respectively. These temperatures
were consistent with those starting to appear in the hysteresis
loops (Fig. S4†).
All decay data for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4 were
analyzed by a stretched exponential decay (eqn (2)).5b,f,19
B
lnðMÞ ¼ lnðM0Þ ꢀ ðt=τÞ
ð2Þ
where M0 is the initial magnetization, τ is the average relaxation
time, and B is the width of the distribution; B = 1 is a single-
exponential decay. The experimental decay data in the region of
M0/M0(1.9) > 0.1 and M/M0 < 0.9, where M0(1.9) is M0 at 1.9 K,
at the given temperature were selected and used for the fitting.
Under this condition, the measurable value of τ were ca. 5 × 102
< τ <ca. 5 × 105 s. The obtained τ values at the given tempera-
tures are listed in Table S1† and the fitting curves are shown in
Fig. 7 and Fig. S6† as solid lines.
(iii) Magnetic relaxations in CoCl2(CYpy)4; Y = 2, 3l, 3b,
and 4. In SMM, the characteristic slow magnetic relaxation for
the reverse of the magnetism takes place in two pathways due to
the thermal activation barrier and the spin quantum tunneling
magnetization (QTM). The relaxations due to the former are
temperature-dependent and obtained by the ac magnetic suscep-
tibility technique, while those due to the latter are temperature-
independent and determined by the temperature dependence of
the dc magnetization decay at extremely low temperature. In
CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, the dc magnetization
decays were combined with those given by means of the ac mag-
netic susceptibility technique and plotted as a function of inverse
T. The τ vs. T−1 plots are shown in Fig. 8 for CoCl2(C2py)4
together with data for CoCl2(C1py)4 and in Fig. S7† for the
others.
At higher temperatures, the decay times obeyed the Arrhenius
law affording the Ueff/kB values. From the ac magnetic suscepti-
bility experiments, the values of Ueff/kB were estimated to be 94,
92, 93, and 87 K for CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4,
respectively. In the lower temperature region of 1.9–3.0 K, the
relaxation time for CoCl2(CYpy)4; Y = 2, and 3l collected by
the dc magnetization decay slightly deviated downward from the
extrapolation of the linear Arrhenius plot obtained by the ac
(ii) Dc magnetization decay measurements. To characterize
the slow magnetization relaxation at even lower temperatures, dc
magnetization decay experiments were carried out at 7–8 temp-
eratures in the range of 1.9–3.3 K for CoCl2(CYpy)4; Y = 2, 3l,
3b, and 4. After the cycle of applying the external field of 50
kOe at 10 K, cooling down to the desired temperature, and then
reducing to 0 kOe, the magnetizations were measured as a func-
tion of time. The magnetization decay was followed for 2 h. For
each measurement, the temperature was raised to 10 K once and
then cooled to the desired temperature. The magnetization data,
M, were normalized against, M0, at t = 0 at each temperature and
the M/M0 values were plotted as a function of time. The M/M0
vs. time plots for CoCl2(CYpy)4 within 3 × 103 s (50 min) at the
given temperatures are shown in Fig. 7 for Y = 2, and in
Fig. S6† for Y = 3l, 3b, and 4. The decay for CoCl2(CYpy)4; Y
= 2, 3l, 3b, and 4, became slower on cooling until 1.9 K without
reaching the temperature independent decay due to the QTM.
The observed decays at the given temperatures for CoCl2-
(CYpy)4; Y = 2, 3l, 3b, and 4, were slower than those for
CoCl2(C1py)4.
Fig. 8 τ vs. T−1 plots of the data collected by ac magnetic suscepti-
bility technique (open) and dc magnetization decay (filled) after
irradiation of a 1 : 4 mixture of CoCl2 and DYpy; Y = 15b (triangle) and
2 (circle) in MTHF frozen solution. The solid line is the least-squares fit
of the ac data for CoCl2(C2py)4, according to the Arrhenius equation.
Fig. 7 Dc magnetization decays after irradiation of 1 : 4 mixtures
(5.0 mM) of CoCl2 and D2py at the given temperatures in frozen sol-
ution. Solid lines show the fittings by the stretched exponential equation.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Online
magnetic susceptibility technique and those for CoCl2(CYpy)4;
Y = 3b, and 4, showed no deviation. The observed deviations
suggested that in the low-temperature region (<3.0 K), the relax-
ation for CoCl2(CYpy)4; Y = 2, and 3l were weakly affected by
the QTM process and those for CoCl2(CYpy)4; Y = 3b, and 4,
were not. Although the values of τq for CoCl2(CYpy)4 could not
be directly determined, those must be longer than 105 s. On the
other hand, It has been reported that the relaxation time for
CoCl2(C1py)4 collected by the ac magnetic susceptibility tech-
nique and the dc magnetization decay were clearly distinguish-
able and the Ueff/kB and τq values were determined to be 91 K
from the ac magnetic susceptibility experiment and 3.5 × 103 s
from the temperature independent dc magnetization decay,
respectively.5b These results might suggest that the τq value
depended on the spin multiplicity of the complex and increased
with increasing the Stotal value. All of the obtained physical mag-
netic values for the 1 : 4 complexes, CoCl2(CYpy)4 including
CoCl2(C1py)4 in frozen solution are summarized in Table 1.
These results of the dc magnetization decay might suggest that
the carbenes strongly affected the transverse interaction, mainly,
the rhombic ZFS parameter E, relating to QTM and deduced the
τq rate. In CoCl2(CYpy)4, the transverse anisotropy of the cobalt
ion might become small by the magnetic interaction between the
high-spin carbene and the cobalt ion, which increased the τq
values. The τq for CoCl2(C1py)4 having the smallest Stotal value
of 9/2 in CoCl2(CYpy)4 was already slow enough not to affect
the Ueff/kB values. Therefore, the increase of τq in CoCl2-
(CYpy)4; Y = 2, 3l, 3b, and 4, which had τq values over 102
times larger than that for CoCl2(C1py)4, did not reflect to the
Ueff/kB value any more and gave almost the same Ueff/kB values.
The increase of τ value from CoCl2(C1py)4 to CoCl2(C2py)4
might lead to the increase in Hc of the hysteresis loop.
Ueff/kB values were independent of the Stotal values and were 94,
92, 93, and 87 K, respectively. On the other hand, the relaxation
time of QTM affecting the U/kB depended on the Stotal values
and was suggested to increase with increasing the Stotal value.
The τq values of QTM time were estimated to be over 105 s,
which did not reflect the Ueff/kB value any more.
This study suggests that organic spins, carbenes, attaching
at equatorial positions in the octahedral cobalt complexes,
CoCl2-(CYpy)4, ferromagnetically interacted with the cobalt
ion through the pyridine ring to increase the relaxation time
of QTM. In the 1 : 4 complexes CoCl2(CYpy)4, CoCl2(C1py)4
having the smallest Stotal value of 9/2 already had
a
large τq value of 103 s. To reduce the Stotal value of the
complex and further clarify the role of organic spin in
heterospin SMM, investigations of the magnetic properties of the
1 : 2 complexes of bidentate cobalt complex with CYpy are in
progress.
Experimental methods
General methods
Infrared spectra were recorded on a JASCO 420 FT-IR spec-
trometer. UV-Vis and visible-near infrared spectra were recorded
on a JASCO V570 spectrometer. 1H NMR spectra were
measured on a JEOL 270 Fourier transform spectrometer using
CDCl3, CD2Cl2, and CD3OD as solvent and referenced to tetra-
methylsilane. FAB mass spectra (FAB MS) were recorded on a
JEOL JMS-SX102 spectrometer. ESI mass spectra (ESI MS)
were recorded on a Bruker Daltonics microTOF spectrometer.
Melting points were obtained with a MEL-TEMP heating block
and are uncorrected. Elemental analyses were performed in the
Analytical Center of the Faculty of Science in Kyushu
University.
Conclusion
In SMM, the thermodynamic activation barrier U/kB for the
reversal of the magnetism is affected by the spin quantum tun-
neling magnetization (QTM), leading to the effective activation
barrier, Ueff/kB; U/kB > Ueff/kB. In [CoCl2(py)4] having no
organic spin, interestingly, frequency-dependent χ′′mol signals
were observed and the thermal profiles of χ′′mol in the presence
of a dc field for suppression of the QTM pathway afforded
Ueff/kB of 16 K, indicating that [CoCl2(py)4] was an SMM and
was strongly affected by the contribution of QTM. The 1 : 4
complexes, CoCl2(CYpy)4; Y = 2, 3l, 3b, and 4, with Stotal = 17/2,
25/2, 25/2, and 33/2, respectively, were also SMMs, whose
Visible-near infrared spectra measurements at cryogenic
temperature
Visible-near infrared spectra were recorded on a JASCO V570
spectrometer attached with an NACC cryo-system LTS-22X for
the low temperature measurements. The solution samples were
placed in separable quartz cells (path length; 1 mm or 1 cm)
under a helium atmosphere. Photolysis of diazo samples were
performed by an argon ion laser (514 nm, 200 mW, Omnichrome
543–200 M).
Table 1 Values of Stotal, Ueff/kB, τ0, τq, Hc, and Mr, for CoCl2(CYpy)4; Y = 1, 2, 3l, 3b, and 4, in MTHF frozen solution
Complexes
Stotal
Ueff/kB (K)
τ0 (s)
τq (s)
Hc (kOe)
Mr (cm3 Oe mol−1
)
1.2 × 10−10
6.2 × 10−11
6.6 × 10−11
4.2 × 10−11
1.0 × 10−10
3.5 × 103
>5 × 105
>5 × 105
>5 × 105
>5 × 105
7.0
12
8.4
11
1.1 × 104
3.0 × 104
3.6 × 104
3.9 × 104
5.1 × 104
a
CoCl2(C1py)4
CoCl2(C2py)4
9/2
17/2
25/2
25/2
33/2
91
94
92
93
87
CoCl2(C3lpy)4
CoCl2(C3bpy)4
CoCl2(C4py)4
8.1
a Ref. 5b.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Online
Magnetic measurements
O2py To a solution of 3.6 g (7.6 mmol) of 4 in acetic acid
(25 ml) was added sodium perchromate dihydrate (9.1 g,
30.4 mmol). The reaction mixture was stirred and refluxed for
5 h. The suspension solution was poured into ice water, and
KOH (5.0 g), CHCl3 (20 ml), and MeOH (10 ml) were added
into the suspension. After filtration, the solution was extracted
with CHCl3 (50 ml × 5) and dried with MgSO4. After filtration,
the organic solvent was evaporated with a rotary evaporator and
the obtained yellow crude solid was chromatographed on silica
gel with CHCl3 (only) ∼ CHCl3 : MeOH = 100 : 1 as eluents to
give O2py as a white solid in a 69% yield (1.6 g, 5.2 mmol).
Ac and dc magnetic susceptibility data were obtained on
Quantum Design MPMS2 (0 10 kOe) and MPMS-5S (0 50
kOe) SQUID magneto/susceptometries, respectively.
(A) Preparation of solution samples. Solutions (1.5–5 mM)
of the samples for the SQUID measurements were obtained as
follows: a solution (1 ml) of CoCl2·6H2O (3–10 mmol) in
MTHF and a solution (1 ml) of DYpy (12–40 mmol) in MTHF
were mixed and used as samples.
Mp 124–127 °C; IR (KBr disc) 1661, 1597 cm−1 1H NMR
;
(B) General procedure for SQUID measurement in frozen
solution. The sample solution was lightly degassed by bubbling
nitrogen gas and 50 μl of sample solution was placed into the
capsule with a microsyringe (Hamilton Co). The capsule was
attached to a handmade SQUID sample rod for the photolysis by
a straw. The ac and dc magnetic susceptibility measurements
were performed in the temperature range of 1.9–10 and 1.9–20
or 30 K, respectively, before and after irradiation. The irradiation
(CDCl3, 270 MHz) δ 8.84 (d, J = 6.2 Hz, 2H), 8.22 (s, 1H),
8.10–8.04 (m, 2H), 7.82 (d, J = 6.2, 2H); Mass spectrum (ESI)
m/z Calcd for (C19H13NO2Na)+ 310.0838. Found: 310.0828;
Anal. Calcd for C19H13NO2: C, 79.43; H, 4.56; N, 4.88. Found:
C, 79.44; H, 4.51; N, 4.67.
5 This was prepared in a manner similar to the procedure for 4
using 2 (1.2 g, 3.6 mmol) in place of 1. The crude mixture was
chromatographed on silica gel with CHCl3 : MeOH = 49 : 1 as
eluents to give carbinol 5 (0.7 g, 1.9 mmol) as yellow oil in a
system was reported previously.5g Before irradiation, the χmol
T
values for all samples in the range of 1.9–20 K were nearly con-
stant at ca. 1.9–2.1 cm3 K mol−1 for CoCl2(DYpy)4 respectively.
The magnetic susceptibility of MTHF solution itself was
measured under similar conditions and was used as the control
data.
The samples were taken out of the apparatus after SQUID
measurements and were measured by IR spectra on KBr pellets.
The quantitative photolysis of the samples were confirmed by
the difference of the absorption due to the diazo moiety
(∼2050 cm−1) before and after irradiation.
53% yield. IR (NaCl plate) 3359, 1600 cm−1 1H NMR
;
(CD3OD, 270 MHz) δ 8.42 (d, J = 6.0 Hz, 2H), 7.40 (d, J = 6.0
Hz, 2H), 7.23–7.12 (m, 10H), 7.01 (br, 3H), 5.72 (s, 1H), 3.90
(s, 2H), 3.89 (s, 2H), 1.10 (br, 1H); Mass spectrum (ESI) m/z
Calcd for (C26H23NONa)+ 388.1672. Found 388.1670.
O3lpy This was prepared in a manner similar to the procedure
for O2py by using 5 (1.0 g, 2.7 mmol) in place of 4. The crude
mixture was chromatographed on silica gel with CHCl3 : MeOH
= 49 : 1 as eluents to give triketone O3lpy (0.8 g, 2.2 mmol) as a
white solid in a 78% yield. Mp 188–192 °C; IR (KBr disc)
1
1661, 1598 cm−1; H NMR (CDCl3, 270 MHz) δ 8.84 (d, J =
Materials
6.1 Hz, 2H), 8.22 (s, 2H), 8.13–8.04 (m, 4H), 7.83 (d, J = 6.7
Hz, 2H), 7.72–7.59 (m, 3H), 7.62 (d, J = 6.1 Hz, 2H), 7.51 (t,
J = 7.4 Hz, 2H); Mass spectrum (ESI) m/z Calcd for
(C26H17NO3Na)+ 414.1101. Found 414.1095; Anal. Calcd for
C26H17NO3: C, 79.78; H, 4.38; N, 3.58. Found: C, 79.26; H,
4.35; N, 3.56.
6 To a solution of 3,3′-dibromodiphenylmethane 3 (6.0 g,
18.5 mmol) in anhydrous THF (60 ml) were added a 1.57 M
solution of n-butyllithium (13 ml, 20.4 mmol) in n-hexane at
−78 °C and stirred for 2 h. The solution of 3-benzylbenzalde-
hyde (4.0 g, 20.4 mmol) in anhydrous THF was added dropwise.
The reaction mixture was stirred for 1 h at −78 °C and then left
to stand overnight at room temperature. The reaction was
quenched with saturated aqueous ammonium chloride. The solu-
tion was extracted with ether for three times. The organic layers
were combined and dried on sodium sulfate. After filtration,
organic solvent was evaporated with a rotary evaporator. The
obtained residue was chromatographed on silica gel with
n-hexane : AcOEt = 10 : 1 as eluents to give carbinol 6 (3.7 g,
8.5 mmol) as a pale yellow oil in a 46% yield. IR (NaCl plate)
Unless otherwise stated, the preparative reactions were carried
out under a high purity dry nitrogen atmosphere. Diethylether,
tetrahydrofuran (THF), and 2-methyltetrahydrofuran (MTHF)
were distilled from sodium benzophenone ketyl. Dichloro-
methane and dimethylsulfoxide were distilled under high-
purity N2 after drying with calcium hydride. [CoCl2(py)4] was
prepared in a manner similar to the procedure reported
previously.12
4 To the solution of 3-bromodiphenylmethane 1 (3.0 g,
12.1 mmol) in anhydrous THF (30 ml) was added 1.6 M solu-
tion of n-butyllithium (14.7 ml, 24.3 mmol) in n-hexane at
−78 °C and was stirring for 2 h. Subsequently, 4-formylpyridine
(2.5 g, 24.3 mmol) in anhydrous THF (5 ml) was added drop-
wise and the reaction mixture was stirred for 4 h at −78 °C. The
reaction was quenched with saturated aqueous ammonium chlor-
ide. The solution was extracted with CHCl3 for three times. The
organic layers were combined and dried on sodium sulfate. After
filtration, organic solvent was evaporated with a rotary evapor-
ator. The obtained residue was chromatographed on silica gel
with n-hexane : CHCl3 = 1 : 1, and then CHCl3 : MeOH = 100 : 1
as eluents to give carbinol 4 (2.1 g, 7.6 mmol) as pale yellow oil
3391 cm−1 1H NMR (CD3OD, 270 MHz) δ 7.20–6.96 (m,
;
17H), 5.59 (s, 1H), 3.81 (s, 4H), 3.90 (br, 1H); Mass spectrum
(ESI) m/z 465.08, 467.08 (C27H23BrONa)+; Anal. Calcd for
C27H23BrO: C, 63.18; H, 4.49. Found: C, 4.35; H, 4.35.
7 To the solution of 6 (3.7 g, 8.4 mmol) in n-hexane (50 ml)
and CH3CN (2.7 ml) were added sodium iodide (7.6 g,
50.6 mmol) and chlorotrimethylsilane (6.4 ml, 50.6 mmol). The
in a 63% yield; IR (NaCl plate) 3421, 1599 cm−1 1H NMR
;
(CDCl3, 270 MHz) δ 8.50 (d, J = 6.0 Hz, 2H), 7.31(s, 1H), 7.27
(d, J = 6.0 Hz, 2H), 7.20–7.14 (m, 8H), 5.75 (s, 1H), 3.97 (s,
2H), 2.80 (br, 1H); Mass spectrum (ESI) m/z Calcd for
(C19H17NONa)+ 298.1202. Found: 298.1197.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Online
solution was stirred for 6 h at room temperature. The reaction
mixture was poured into water and extracted by ether (500 ml)
three times. The combined ether solutions were washed with
sodium thiosulfate and then dried over sodium sulfate. After fil-
tration, ether was removed by reduced pressure. The obtained
crude mixture was chromatographed on silica gel with n-hexane
as an eluent to afford 7 (2.9 g, 6.7 mmol) as a colorless oil in
MeOH = 49 : 1 to give carbinol 9 (11.5 g, 24.2 mmol) as a white
solid in a 76% yield. Mp 167–171 °C; IR (KBr disc)
1
3310 cm−1; H NMR (CDCl3, 270 MHz) δ 7.43 (d, J = 5.4 Hz,
2H), 7.36 (d, J = 8.7 Hz, 4H), 7.34 (s, 1H), 7.25 (d, J = 8.0 Hz,
4H), 5.75 (s, 2H) 1.30 (s 18H); Mass spectrum (ESI) m/z
Calcd for (C28H33O2BrNa)+ 503.1556. Found: 503.1554; Anal.
Calcd for C28H33O2Br: C, 69.85; H, 6.91. Found: C, 70.21;
H, 6.98.
1
79.6% yield. IR (NaCl plate) 3025, 2905, 2838 cm−1; H NMR
(CDCl3, 270 MHz) δ 7.31–7.10 (m, 11H), 7.01 (d, J = 7.5 Hz,
6H), 3.94 (s, 2H), 3.91 (s, 2H), 3.89 (s, 2H); Mass spectrum
(ESI) m/z Calcd for (C27H23BrNa)+: 449.0881, 451.0860.
Found: 449.0890, 451.0866; Anal. Calcd for C27H23Br: C,
75.88; H, 5.42. Found: C, 76.24; H, 5.41.
10 This was prepared in a manner similar to the procedure for
7 using 9 in place of 6. Reprecipitation from CHCl3–Et2O gave
10 as a white powder in 85% yield. Mp 68–70 °C; IR (KBr disc)
1
3024, 2978 cm−1; H NMR (CDCl3, 270 MHz) δ 7.31 (d, J =
8.0 Hz, 4H), 7.16 (s, 2H), 7.08 (d, J = 8.0 Hz, 4H), 7.00 (s, 1H),
3.87 (s, 4H), 1.30 (s, 18H); Mass spectrum (ESI) m/z 471.16
(C28H33BrNa)+; Anal. Calcd for C28H33Br: C, 74.82; H, 7.40.
Found: C, 74.84; H, 7.43.
11 This was prepared in a manner similar to the procedure for
5 using 10 in place of 2. The crude mixture including 11 and 12
was obtained as a yellow oil and used for the next oxidation pro-
cedure. Mass spectrum (ESI) m/z Calcd for (C34H39NO)+:
477.3032. Found: 478.3104.
8 To anhydrous THF solution (30 ml) stirring at −78 °C was
added 1.57 M n-butyllithium (3.5 ml, 5.3 mmol) in n-hexane
and then the solution of 7 (1.5 g, 3.5 mmol) in anhydrous THF
(15 ml). After stirring for 1.5 h, the solution of 4-formylpyridine
(5 ml, 5.3 mmol) in anhydrous THF (15 ml) was added drop-
wise. The dry ice–acetone bath was removed and the reaction
mixture was stirred overnight at room temperature. Saturated
ammonium chloride solution was added and the reaction mixture
was extracted by ether (200 ml) three times. The combined ether
solutions were washed with sodium thiosulfate and then dried
over sodium sulfate. After filtration, ether was removed by
reduced pressure. The obtained crude mixture was chromato-
graphed on silica gel with CHCl3 : MeOH (100 : 1–10 : 1) as
eluents to afford 8 (0.63 g, 1.4 mmol) as a yellow oil in 39.4%
12 Activated MnO2 (10.4 g, 119.2 mmol) was added to the
crude mixture of 11 and 12 (7.6 g) in CHCl3 (65 ml) and the
mixture was refluxed for 6 h. After filtration, the solvent was
removed on a rotary evaporator. The crude mixture was chro-
matographed on silica gel with CHCl3 (only) ∼ CHCl3 : MeOH
= 100 : 1 as eluents to afford 12 (4.0 g, 8.4 mmol) as a
1
yield. IR (NaCl plate) 3330 cm−1; H NMR (CDCl3, 270 MHz)
white solid in 76% yield in two steps. IR (KBr disc) 1663 cm−1
;
δ 8.47 (d, J = 6.0 Hz, 2H), 7.28–7.10 (m, 13H), 7.01 (d, J = 8.1
Hz, 6H), 5.72 (s, 1H), 3.93 (s, 2H), 3.91 (s, 2H), 3.89 (s, 2H),
2.90 (br, 1H); Mass spectrum (ESI) Calcd for (C33H30NONa)+.
456.2322. Found: 456.2224; Anal. Calcd for C33H29NO:
C, 87.00; H, 6.42; N, 3.07. Found: C, 87.87; H, 6.22; N, 2.67.
O4py This was prepared in a manner similar to the procedure
for O2py using 8 in place of 4. Reprecipitation from CHCl3–
Et2O gave tetraketone O4py (0.22 g, 2.0 mmol) as a white
powder in 32.6% yield. Mp 233–235 °C; IR (KBr disc) 1661,
1H NMR (CDCl3, 270 MHz) δ 8.74 (d, J = 6.0 Hz, 2H), 7.49
(d, J = 6.0 Hz, 2H), 7.46 (s, 3H), 7.31 (d, J = 8.4 Hz, 4H), 7.09
(d, J = 8.4 Hz, 4H), 3.87 (s, 4H), 1.30 (s, 18H); Mass
spectrum (ESI) Calcd for (C33H30NONa)+: 456.2322. Found:
456.2224.
O3bpy Sodium perchromate (9.1 g, 30.4 mmol) was added to
a solution of 12 (3.6 g, 7.6 mmol) in acetic acid (45 ml). The
suspension was stirred and refluxed for 10 h. The reaction
mixture was poured into ice water. After filtration, the residue of
pale yellow solid was washed with water (50 ml). The filtrate
was centrifuged at 3500 rpm for 10 min and the sticky precipi-
tate on the bottom of the test tube was collected. The sticky pre-
cipitate was dissolved in MeOH (30 ml), KOH was added until
the color of the solution changed to pale green, and the solution
was stirred at room temperature overnight. Water (100 ml) was
added and the solution was extracted with Et2O (50 ml) three
times, ether layers were combined and dried with MgSO4. After
filtration, ether was evaporated by a rotary evaporator to afford
O3bpy (2.6 g, 5.2 mmol) as a white solid in a 69% yield.
1
1598 cm−1; H NMR (CDCl3, 270 MHz) δ 8.84 (d, J = 6.1 Hz,
2H), 8.24 (s, 1H), 8.23(d, J = 7.4 Hz, 2H), 8.12–8.03 (m, 6H),
7.83 (d, J = 7.4 Hz, 2H), 7.71 (d, J = 2.0 Hz, 1H), 7.67 (d, J =
6.1 Hz, 2H), 7.63–7.59 (m, 3H), 7.50(t, J = 7.4 Hz, 2); Mass
spectrum (ESI) m/z 518.14 (C33H21NO4Na)+; Anal. Calcd for
C33H21NO4: C, 79.99; H, 4.27; N, 2.83. Found: C, 79.28;
H, 4.54; N, 2.76.
9 To a solution of 1,3,5-tribromobenzene (3.0 g, 9.5 mmol) in
anhydrous THF (120 ml) was added a 1.6 M solution of n-butyl-
lithium (17.8 ml, 28.5 mmol) in n-hexane at −78 °C. After stir-
ring for 4 h, 1-formyl-4-tert-butylbenzene (3.0 g, 28.5 mmol) in
anhydrous THF (30 ml) was added dropwise. The reaction
mixture was stirred for 1 h at −78 °C. Saturated ammonium
chloride solution was added and the reaction mixture was
extracted by ether (200 ml) three times. The combined ether sol-
utions were dried over sodium sulfate. After filtration, ether was
removed by reduced pressure. Crude carbinol (15.2 g, ca. 98%
yield) containing monoketone derivative was obtained as a
brown solid. Sodium borohydride (4.8 g) was added to a solution
of crude carbinol in methanol (40 ml). The suspension was
stirred and refluxed for 4 h. The solvent was evaporated and the
crude mixture was chromatographed on silica gel with CHCl3 :
1
Mp 230–234 °C; IR (KBr disc) 1663 cm−1; H NMR (CDCl3,
270 MHz) δ 8.86 (d, J = 4.0 Hz, 2H), 8.41 (s, 3H), 7.79 (d,
J = 8.0 Hz, 4H), 7.64 (d, J = 4.7 Hz, 2H), 7.54 (d, J = 8.0 Hz,
4H) 1.37 (s, 18H); Mass spectrum (ESI) m/z 526.25
(C34H33NO3Na)+;
Anal.
Calcd
for
C34H33NO3·
0.1CHCl3·0.1CH3CN: C, 79.27; H, 6.48; N, 2.96. Found: C,
79.26; H, 6.76; N, 2.96.
D2py This was prepared by using O2py in a manner similar
to the procedure reported previously.6f Crude D2py was chro-
matographed on Al2O3 with CH2Cl2 as eluent and then recrystal-
lized from Et2O–CH2Cl2 to afford D2py as red thin plate
crystals. Mp (dec.) 108–110 °C; IR (KBr disc), νCvN2
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012

View Online
2038 cm−1; UV-Vis in CH2Cl2, λmax(ε): 294 (3.5 × 104) and
J. Am. Chem. Soc., 2006, 128, 6975–6989; (c) C. J. Milios, A. Vinslava,
P. A. Wood, S. Parsons, W. Wernsdorfer, G. Christou, S. P. Perlepes and
E. K. Brechin, J. Am. Chem. Soc., 2007, 129, 8–9; (d) C. J. Milios,
A. Vinslava, W. Wernsdorfer, S. Moggach, S. Parsons, S. P. Perlepes,
G. Christou and E. K. Brechin, J. Am. Chem. Soc., 2007, 129, 2754–
2755.
3 (a) S. Osa, T. Kido, N. Matsumoto, N. Re, A. Pochaba and J. Mrozinski,
J. Am. Chem. Soc., 2004, 126, 420–421; (b) J. Rinck, G. Novitchi,
W. Van den Heuvel, L. Ungur, Y. H. Lan, W. Wernsdorfer, C. E. Anson,
L. F. Chibotaru and A. K. Powell, Angew. Chem., Int. Ed., 2010, 49,
7583–7587; (c) V. Baskar, K. Gopal, M. Helliwell, F. Tuna,
W. Wernsdorfer and R. E. P. Winpenny, Dalton Trans., 2010, 39, 4747–
4750.
4 (a) N. Ishikawa, M. Sugita, T. Ishikawa, S. Koshihara and Y. Kaizu,
J. Am. Chem. Soc., 2003, 125, 8694–8695; (b) N. Ishikawa, M. Sugita
and W. Wernsdorfer, J. Am. Chem. Soc., 2005, 127, 3650–3651;
(c) F. Pointillart, K. Bernot, R. Sessoli and D. Gatteschi, Chem.–Eur. J.,
2007, 13, 1602–1609; (d) P.-H. Lin, T. J. Burchell, L. Ungur,
L. F. Chibotaru and W. Wernsdorfer, Angew. Chem., Int. Ed., 2009, 48,
9489–9492; (e) S.-D. Jiang, B.-W. Wang, G. Su, Z.-M. Wang and S. Gao,
Angew. Chem., Int. Ed., 2010, 49, 7448–7451.
5 (a) S. Kanegawa, S. Karasawa, M. Nakano and K. Koga, Bull. Chem.
Soc. Jpn., 2006, 79, 1372–1382; (b) S. Karasawa and N. Koga, Inorg.
Chem., 2011, 50, 5186–5195; (c) S. Kanegawa, S. Karasawa,
M. Maeyama, M. Nakano and K. Koga, J. Am. Chem. Soc., 2008, 130,
3079–3094; (d) S. Karasawa, G. Zhou, H. Morikawa and N. Koga,
J. Am. Chem. Soc., 2003, 125, 13676–13677; (e) D. Yoshihara,
S. Karasawa and N. Koga, J. Am. Chem. Soc., 2008, 130, 10460–10461;
(f) D. Yoshihara, S. Karasawa and N. Koga, Polyhedron, 2011, 30,
3211–3217; (g) N. Koga and S. Karasawa, Bull. Chem. Soc. Jpn., 2005,
78, 1384–1400; (h) S. Karasawa, D. Yoshihara, N. Watanabe, M. Nakano
and N. Koga, Dalton Trans., 2008, 1418–1421; (i) S. Kanegawa,
S. Karasawa, M. Nakano and N. Koga, Chem. Commun., 2004, 3590–
3591.
1
508 (2.1 × 102) nm; H-NMR (270 MHz, CDCl3), δ 8.43 (dd,
J = 4.6 and 1.7 Hz, 2H), 7.51–7.10 (m, 9 H), 7.06 (dd,
J = 4.6 and 1.7 Hz, 2H); Mass spectrum (ESI) m/z Calcd for
(C19H13N5Na)+ 334.1063. Found 334.1048; Anal. Calcd for
C19H13N5: C, 73.30; H, 4.21; N, 22.49. Found: C, 73.16; H,
4.19; N, 22.44.
D3lpy This was prepared by using O3lpy in a manner similar
to the procedure reported previously.6f Crude D3lpy was chro-
matographed on Al2O3 with CH2Cl2 as eluent and then re-preci-
pitated from CH2Cl2–Et2O to afford D3lpy as red thin plate
crystals. Mp (dec.) 74–75 °C; IR (KBr disc), νCvN2 2039 cm−1
;
UV-Vis in CH2Cl2, λmax(ε): 302 (4.6 × 104) and 512 (3.1 × 102)
1
nm; H-NMR (270 MHz, CD2Cl2), δ 8.48 (d, J = 4.9 Hz, 2),
7.50–7.10 (m, 13 H), 7.07 (dd, J = 4.6 and 1.7 Hz, 2H); FAB
mass (in NBA matrix), (M + 1)+ 428; Anal. Calcd for
C26H17N7: C, 73.05; H, 4.01; N, 22.94. Found: C, 73.09; H,
4.01; N, 22.78.
D3bpy This was prepared by using O3bpy in a manner
similar to the procedure reported previously.6f Crude D3bpy was
chromatographed on Al2O3 with CH2Cl2 as eluent and then
recrystallized from ether to afford D3bpy as red crystals. Mp
(dec.) 95–105 °C; IR (KBr disc) νCvN2 2039 cm−1; UV-Vis in
1
CH2Cl2 λmax(ε) = 294 (6.2 × 104) and 510 (3.6 × 102) nm; H
NMR (CD2Cl2, 270 MHz) δ 8.44 (dd, J = 6.4, 1.7 Hz, 2H), 7.44
(d, J = 8.7 Hz, 4H), 7.25 (d, J = 8.7 Hz, 4H), 7.07 (dd, J = 6.4,
1.7 Hz, 2H), 7.02 (d, J = 1.7 Hz, 2H), 6.95 (s, 1H), 1.32 (s,
18H); Mass spectrum (ESI) m/z Calcd for (C34H34N7)+
540.2876. Found 540.2884; Anal. Calcd for C34H33N7:
C, 75.67; H, 6.16; N, 18.17. Found: C, 75.60; H, 6.21; N, 17.95.
D4py This was prepared by using O4py in a manner similar
to the procedure reported previously.6f Crude D4py was chro-
matographed on Al2O3 with CH2Cl2 as eluent and then repreci-
pitated from Et2O : n-hexane (10 : 1) to afford D4py as a red
powder in 56.3% yield. Mp (dec.) 96–100 °C; IR (KBr disc)
2039, 1587 cm−1; UV-Vis in CH2Cl2 λmax(ε) = 300 (9.2 × 104),
6 (a) K. Murashima, T. Watanabe, S. Kanegawa, D. Yoshihara, Y. Inagaki,
S. Karasawa and N. Koga, Inorg. Chem., 2012, 51, 4982–4993;
(b) S. Karasawa and N. Koga, Inorg. Chem., 2011, 50, 2055–2057;
(c) S. Karasawa, Y. Sano, T. Akita, N. Koga, T. Itoh and H. Iwamura,
J. Am. Chem. Soc., 1998, 120, 10080–10087; (d) S. Karasawa,
H. Kumada, N. Koga and H. Iwamura, J. Am. Chem. Soc., 2001, 123,
9685–9686; (e) Z. Zhu, S. Karasawa and N. Koga, Inorg. Chem., 2005,
44, 6004–6011; (f) H. Morikawa, F. Imamura, Y. Tsurukami, T. Itoh,
H. Kumada, S. Karasawa, N. Koga and H. Iwamura, J. Mater. Chem.,
2001, 11, 493–502.
7 K. Matsuda, N. Nakamura, K. Takahashi, K. Inoue, N. Koga and
H. Iwamura, J. Am. Chem. Soc., 1995, 117, 5550–5560.
1
510 (ε = 4.2 × 102) nm; H-NMR (270 MHz, CD2Cl2), δ 8.48
8 (a) O. Kahn, Molecular Magnetism, Wiley-VCH Publishers, Weinheim,
1993; (b) A. C. Rizzi, C. D. Brondino, R. Calvo, R. Baggio,
M. T. Garland and R. E. Rapp, Inorg. Chem., 2003, 42, 4409–4416;
(c) E. Pardo, R. Ruiz-García, F. Lloret, J. Faus, M. Julve, Y. Journaux,
M. A. Novak, F. S. Delgado and C. Ruiz-Pérez, Chem.–Eur. J., 2007, 13,
2054–2066; (d) M. Sarkar, G. Aromí, J. Cano, V. Bertolasi and D. Ray,
Chem.–Eur. J., 2010, 16, 13825–13833; (e) A. Bencini, C. Benelli,
D. Gatteschi and C. Zanchini, Inorg. Chem., 1980, 19, 3027–3030;
(f) J. Faus, M. Julve, F. Lloret, J. A. Real and J. Sletten, Inorg. Chem.,
1994, 33, 5535–5540; (g) A. Caneschi, D. Gatteschi, N. Lalioti,
R. Sessoli, L. Sorace, V. Tangoulis and A. Vindigni, Chem.–Eur. J., 2002,
8, 286–292.
(d, J = 4.9 Hz, 2H), 7.50–7.10 (m, 17 H), 7.07 (dd, J = 4.6
and 1.7 Hz, 2H); Mass spectrum (ESI) m/z Calcd for
(C33H21N9Na)+ 566.1812. Found 566.1788; Anal. Calcd for
C33H21N9: C, 72.92; H, 3.89; N, 23.19. Found: C, 72.85;
H, 3.90; N, 23.02.
Acknowledgements
This work was supported by a Grant-in-Aid for Scientific
Research (B)(2) (No. 17350070) from the Ministry of Education,
Science, Sports and Culture, Japan, and by the “Nanotechnology
Support Project” of the Ministry of Education, Culture, Sports,
Science and Technology (MEXT), Japan.
9 H. Tobinaga, M. Suehiro, T. Itoh, G. Zhou, S. Karasawa and N. Koga,
Polyhedron, 2007, 26, 1905–1911.
10 G. A. Olah, S. C. Narang, B. G. Balaram Gupta and R. Malhotra, J. Org.
Chem., 1979, 44, 1247–1251.
11 (a) F. A. Cotton, D. M. L. Goodgame and D. Goodgame, J. Am. Chem.
Soc., 1961, 83, 4690–4699; (b) A. B. P. Lever, Inorganic Electronic Spec-
troscopy, Elsevier, Amsterdam, 2nd edn, 1984; (c) J. M. Holland,
C. A. Kilner, M. Thornton-Pett and M. A. Halcrow, Polyhedron, 2001,
20, 2829–2840; (d) I. Kuzniarska-Biernacka, A. Bartecki and K. Kurzak,
Polyhedron, 2003, 22, 997–1007.
References
12 A. Bencini, C. Benelli, D. Gatteschi and C. Zanchini, Inorg. Chem.,
1980, 19, 1301–1304.
1 D. Gatteschi, R. Sessoli and J. Villain, Molecular Nanomagnets, Oxford
University Press, 2006.
13 A. Rajca, J. Wongsriratanakui and S. Rajca, Science, 294, 1503–
1505.
14 H. Iwamura and N. Koga, Acc. Chem. Res., 1993, 26, 346–
351.
2 (a) L. Lecren, W. Wernsdorfer, Y. G. Li, O. Roubeau, H. Miyasaka and
R. Clérac, J. Am. Chem. Soc., 2005, 127, 11311–13117;
(b) N. E. Chakov, S.-C. Lee, A. G. Harter, P. L. Kuhns, A. P. Reyes,
S. O. Hill, N. S. Dalal, W. Wernsdorfer, K. A. Abboud and G. Christou,
This journal is © The Royal Society of Chemistry 2012
Dalton Trans.

View Online
15 (a) K. Hirai, E. Kamiya, T. Itoh and H. Tomioka, Org. Lett., 2006, 8,
1847–1850; (b) T. Itoh, K. Hirai and H. Tomioka, J. Am. Chem. Soc.,
2004, 126, 1130–1140.
16 H. Sakiyama, J. Chem. Software, 2001, 7, 171–178.
17 (a) H. Sakiyama, Inorg. Chim. Acta, 2006, 359, 2097–2100;
(b) J. Suárez-Varela, H. Sakiyama, J. Cano and E. Colacio, Dalton Trans.,
2007, 249–256.
W. Wernsdorfer, K. Folting, M. Pink and G. Christou, J. Am. Chem. Soc.,
2004, 126, 2156–2165.
19 (a) L. Thomas, A. Caneschi and B. Barbara, Phys. Rev. Lett., 1999, 83,
2398–2398; (b) S. M. J. Aubin, Z. M. Sun, L. Pardi, J. Krzystek,
K. Folting, L. C. Brunel, A. L. Rheingold, G. Christou and
D. N. Hendrickson, Inorg. Chem., 1999, 38, 5329–5340;
(c) D. N. Hendrickson, G. Christou, H. Ishimoto, Y. Yoo, E. K. Brechin,
A. Yamaguchi, E. M. Rumberger, S. M. J. Aubin, Z. M. Sun and
G. Aromi, Polyhedron, 2001, 20, 1479–1488.
18 (a) M. Murugesu, M. Habrych, W. Wernsdorfer, K. A. Abboud and
G. Chrisou, J. Am. Chem. Soc., 2004, 126, 4766–4767; (b) M. Soler,
Dalton Trans.
This journal is © The Royal Society of Chemistry 2012