Probing single-chain magnets in a family of linear chain compounds constructed by magnetically anisotropic metal-ions and cyclohexane-1,2- dicarboxylate analogues

Article abstract of DOI:10.1021/ic801498n

Five new metal-carboxylate chain-based laminated compounds, namely, _∞²[Fe^II(e,e-trans-1,2-chdc)] (3) (1,2-chdc = cyclohexane-1,2-dicarboxylate), _∞ ²[Ni^II(μ-OH₂)(e,a-cis-1,2-chdc)] (4), _∞²[Co^II(μ-OH₂)(1,2-chedc)] (5) (1,2-chedc = cyclohex-1-ene-1,2-dicarboxylate), _∞ ²[Co5^II(μ₃-OH)₂(OH ₂)₂(1,2-chedc)₄] (6), and _∞²[Co^II(4-Me-1,2-chdc)] (7) (4-Me-1,2-chdc = trans-4-methylcyclohexane-1,2-dicarboxylate) have been hydrothermally synthesized. In these series of magnetic chain-based compounds, 3 and 7 have the same dimeric paddle-wheel M(II)-carboxylate chain as the previously reported compound, _∞²[Co ^II(trans-1,2-chdc)] (2). However, compound 3 does not behave as a single-chain magnet (SCM) but simply an alternating ferro-antiferro magnetic chain. Compound 4 has the cis conformation of 1,2-chdc ligand, which leads to a uniform aqua-carboxylate-bridged Ni(II) chain. Such a Ni-O chain exhibits strong antiferromagnetic interactions, leading to a diamagnetic ground state. Compound 5 features a corner-sharing triangular chain, or Δ-chain, which is part of a Kagome lattice. However, 5 does not exhibit a spin-frustrated effect but simply spin competition. Compound 6 has a unique pentanuclear Co ^II cluster, which is further connected by the syn-anti carboxylate into a chain structure. Compound 6 exhibits antiferromagnetic interactions among the Co(II) ions, and no SCM behavior is observed. These results might indicate that the dimeric paddle-wheel Co(II)-carboxylate chain is essential in obtaining SCM behavior in this family of compounds. Although 2 and 7 have very similar SCM behavior, alternating current magnetic studies show that 7 has a higher energy barrier than that of 2. Such behavior is probably caused by the larger anisotropic energy barrier in 7.

Full text of DOI:10.1021/ic801498n

Inorg. Chem. 2008, 47, 11202-11211
Probing Single-Chain Magnets in a Family of Linear Chain Compounds
Constructed by Magnetically Anisotropic Metal-Ions and
Cyclohexane-1,2-Dicarboxylate Analogues
Yan-Zhen Zheng,† Wei Xue,† Ming-Liang Tong,*,† Xiao-Ming Chen,*,† and Shao-Liang Zheng‡
MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, School of Chemistry and Chemical
Engineering, Sun Yat-Sen UniVersity, Guangzhou 510275, China, and Department of Chemistry,
State UniVersity of New York at Buffalo, Buffalo, New York 14260-3000
Received August 7, 2008
Five new metal-carboxylate chain-based laminated compounds, namely, 2∞[FeII(e,e-trans-1,2-chdc)] (3) (1,2-chdc )
cyclohexane-1,2-dicarboxylate), 2∞[NiII(µ-OH2)(e,a-cis-1,2-chdc)] (4), ∞2 [CoII(µ-OH2)(1,2-chedc)] (5) (1,2-chedc )
cyclohex-1-ene-1,2-dicarboxylate), ∞2 [Co5II(µ3-OH)2(OH2)2(1,2-chedc)4] (6), and 2∞[CoII(4-Me-1,2-chdc)] (7) (4-Me-1,2-
chdc ) trans-4-methylcyclohexane-1,2-dicarboxylate) have been hydrothermally synthesized. In these series of
magnetic chain-based compounds, 3 and 7 have the same dimeric paddle-wheel M(II)-carboxylate chain as the
previously reported compound, [CoII(trans-1,2-chdc)] (2). However, compound 3 does not behave as a single-
2
∞
chain magnet (SCM) but simply an alternating ferro-antiferro magnetic chain. Compound 4 has the cis conformation
of 1,2-chdc ligand, which leads to a uniform aqua-carboxylate-bridged Ni(II) chain. Such a NisO chain exhibits
strong antiferromagnetic interactions, leading to a diamagnetic ground state. Compound 5 features a corner-sharing
triangular chain, or ∆-chain, which is part of a Kagome´ lattice. However, 5 does not exhibit a spin-frustrated effect
but simply spin competition. Compound 6 has a unique pentanuclear CoII cluster, which is further connected by the
syn-anti carboxylate into a chain structure. Compound 6 exhibits antiferromagnetic interactions among the Co(II)
ions, and no SCM behavior is observed. These results might indicate that the dimeric paddle-wheel Co(II)-carboxylate
chain is essential in obtaining SCM behavior in this family of compounds. Although 2 and 7 have very similar SCM
behavior, alternating current magnetic studies show that 7 has a higher energy barrier than that of 2. Such behavior
is probably caused by the larger anisotropic energy barrier in 7.
Introduction
applications, and therefore the exploration of SMMs and
SCMs with higher blocking temperatures (TB) becomes an
important topic for molecular magnets. As indicated by the
energy barriers of SMMs (∆A ) |D|S2 for integer spins,
hereafter D refers to the zero-field splitting parameter) and
SCMs (∆ )∆A + 2∆ꢀ ) |D|S2 + 8JS2), the larger sums of
the D and S vectors are essential for higher TB. However,
recent theoretical studies indicate that the D and S values
are extremely difficult to be simultaneously maximized, while
the ideal maximum situation is one in which all the Ising-
type spins are oriented parallel and ferromagnetically
coupled.4
One of the most active topics in today’s coordination
chemistry is the design of molecular magnets,1 which has
the final aim of creating qubits for quantum information
processing and storage. As highly expected candidates of
such molecular magnets, single-molecule magnets (SMMs)2
and single-chain magnets (SCMs)3 have received much
attention. In the past decade, many SMMs and a few of
SCMs have been made. However, the blocking temperatures
of up-to-date SMMs and SCMs are still too low for practical
* To whom correspondence should be addressed. E-mail: tongml@
mail.sysu.edu.cn (M.-L.T.), cxm@mail.sysu.edu.cn (X.-M.C.). Fax: (+) 86
20 8411-2245.
(2) For reviews, see (a) Christou, G.; Gatteschi, D.; Hendrickson, D. N.;
Sessoli, R. MRS Bull. 2000, 25, 66–71. (b) Gatteschi, D.; Sessoli, R.
Angew. Chem., Int. Ed. 2003, 43, 268–297. (c) Arom´ı, G.; Brechin,
E. K. Struct. Bonding (Berlin) 2006, 122, 1–67.
(3) For review, see. (a) Coulon, C.; Miyasaka, H.; Cle´rac, R. Struct.
Bonding (Berlin) 2006, 122, 163–206.
†
Sun Yat-Sen University.
State University of New York at Buffalo.
‡
(1) (a) Leuenberger, M. N.; Loss, D. Nature 2001, 410, 789. (b) Meier,
F.; Levy, J.; Loss, D. Phys. ReV. Lett. 2003, 90, 047901. (c) Meier,
F.; Levy, J.; Loss, D. Phys. ReV. B 2003, 90, 134417.
11202 Inorganic Chemistry, Vol. 47, No. 23, 2008
10.1021/ic801498n CCC: $40.75  2008 American Chemical Society
Published on Web 10/29/2008

Single-Chain Magnets
Scheme 1. Structures of 1,2-chdcH2 and Its Analogues
To obtain such an ideal compound, there are probably two
ways only. The first one is based on judicious ligand design.
Nevertheless, the current synthetic methods of metal clusters
are mostly based on self-assembly or the so-called “seren-
dipitous assembly”,5 because “unexpected” outcomes rather
than the “designed” structures can be generated. The second
one is based on the modifications of known clusters; such a
strategy has already been widely used in furnishing new
SMMs,6 SCMs,7 and other higher-dimensional molecular
networks with interesting magnetic properties.8
these layered compounds,12 since SCM behaviors are also
observed in compounds with these homo- or heterospin
systems. For this purpose, we extended our investigations
into other magnetically anisotropic transition metal ions and
trans-1,2-chdc analogues ligands (Scheme 1). The use of
these metals and ligands leads to a new family of layered
compounds that feature various metal-carboxylate chains:
∞2 [FeII(e,e-trans-1,2-chdc)] (3), ∞2 [NiII(µ-OH2)(e,a-cis-1,2-
We recently used a σ-bonded trans-1,2-cyclohexane-
dicarboxylate (trans-1,2-chdc) to link anisotropic CoII ions
into one-dimensional (1D) chains in a two-dimensional (2D)
polymer, [CoII(trans-1,2-chdc)] (2), showing SCM behav-
2
∞
ior.9 In contrast with some other CoII-based SCMs,10 the
chdc)](4), [CoII(µ-OH2)(1,2-chedc)](5), [Co5II(µ3-OH)2(OH2)2-
2
2
structure of 2 is somewhat unique because all the CosO
bonds are parallel and the CoII ions are ferromagnetically
coupled, which might fulfill the requirement of an “ideal
model” with a maximized energy barrier. It is therefore,
interesting to continue investigating such kind of SCMs.
Usually, only very similar ligands can lead to similar
structures, yet the experimental procedures may highly affect
the final outcomes from the same ligand and metal ions in
the same starting ratio.9,11 Moreover, it is also meaningful
to investigate the magnetic properties of other magnetically
anisotropic transition metal ions like MnIII, FeII, and NiII in
∞
∞
(1,2-chedc)4] (6), and 2∞[CoII(4-Me-1,2-chdc)] (7), where cis-
1,2-chdc ) cis-cyclohexane-1,2-dicarboxylate, 1,2-chedc )
cyclohex-1-ene-1,2-dicarboxylate, and 4-Me-1,2-chdc ) trans-
4-methylcyclohexane-1,2-dicarboxylate. The syntheses, struc-
tures, and magnetic properties of these compounds will
be discussed in relation with the SCM behavior observed
in 2.
Experimental Section
Materials and Physical Measurements. Commercially available
reagents were used as received without further purification. The
C, H, and N microanalyses have been carried out with an Elementar
Vario-EL CHNS elemental analyzer. The FT-IR spectra have been
recorded from KBr pellets in the range 4000-400 cm-1 on a Bio-
Rad FTS-7 spectrometer.
The magnetic susceptibility measurements on polycrystalline
samples of 3-7 have been carried out on a Quantum Design
MPMS-XL7 SQUID magnetometer between 1.8 and 300 K and (
7 T. Alternating current (ac) susceptibility measurements have been
performed at frequencies between 1 and 1500 Hz with an ac field
of 0.0005 T and with a zero or 0.1 T applied direct current (dc)
field. The zero-field cooled magnetization was obtained upon
warming after zero-field cooling from 300 to 1.8 K. The field-
cooled magnetization was obtained upon warming after constant
0.01 T applied field cooling from 300 to 1.8 K. Diamagnetic
corrections were calculated from Pascal constants and applied to
the observed magnetic susceptibility.
Typical Synthetic Procedure for 2∞[FeII(e,e-trans-1,2-chdc)]
(3). A mixture of FeSO4 ·7H2O (0.278 g, 1mmol), trans-1,2-chdcH2
or cis-1,2-chdcH2 (0.172 g, 1 mmol), triethylamine (0.200 g, 2.0
mmol), and deioned water (10 mL) was sealed in a 23 mL Teflon-
lined autoclave and heated at 160 °C for 2 days to give pale-green
plate-like crystals of 3 (yield 0.15 g, 65%). IR data (ν˜, cm-1):
2930(m), 2845(w), 1582(vs), 1522(s), 1450(s), 1420(s), 1356(w),
1308(m), 1220(w), 1029(w), 891(w), 836(w), 767(w), 729(w),
668(w), 588(w), 522(w), 438(w). Anal. Calcd (C8H10FeO4): C,
42.51; H, 4.46; Found: C, 42.80; H, 4.48%.
Typical Synthetic Procedure for 2∞[NiII(µ-OH2)(e,a-cis-1,2-chdc)]
(4). A mixture of NiCl2 ·6H2O (0.238 g, 1 mmol), cis-1,2-chdcH2
(0.172 g, 1 mmol), NaOH (0.200 g, 2.0 mmol) in deioned water
(10 mL) was sealed in a 23 mL Teflon-lined autoclave and heated
at 200 °C for 1 day to give pale-green plate-like crystals of 4 (yield
(4) (a) Waldmann, O. Inorg. Chem. 2007, 46, 10035. (b) Ruiz, E.; Cirera,
J.; Cano, J.; Alvarez, S.; Loosec, C.; Kortus, J. Chem. Commun. 2008,
52.
(5) Winpenny, R. E. P. Dalton Trans. 2002, 1.
(6) (a) Zheng, Y.-Z.; Xue, W.; Zhang, W.-X.; Tong, M.-L.; Chen, X.-M.
Inorg. Chem. 2007, 46, 6437. (b) Cornia, A.; Fabretti, A. C.; Garrisi,
P.; Mortalo`, C.; Bonacchi, D.; Gatteschi, D.; Sessoli, R.; Sorace, L.;
Wernsdorfer, W.; Barra, A.-L. Angew. Chem., Int. Ed. 2004, 43, 1136.
(7) (a) Lecren, L.; Wernsdorfer, W.; Li, Y.-G.; Vindigni, A.; Miyasaka,
H.; Cle´rac, R. J. Am. Chem. Soc. 2007, 129, 5045. (b) Lecren, L.;
Roubeau, O.; Coulon, C.; Li, Y.-G.; Goff, X. F. L.; Wernsdorfer, W.;
Miyasaka, H.; Cle´rac, R. J. Am. Chem. Soc. 2005, 127, 17353. (c)
Toma, L. M.; Lescoue¨zec, R.; Pasa´n, J.; Ruiz-Pa´rez, C.; Vaissermann,
J.; Cano, J.; Carrasco, R.; Wernsdorfer, W.; Lloret, F.; Julve, M. J. Am.
Chem. Soc. 2006, 128, 4842. (d) Bai, Y.-L.; Tao, J.; Wernsdorfer,
W.; Sato, O.; Huang, R.-B.; Zheng, L.-S. J. Am. Chem. Soc. 2006,
128, 16428. (e) Xu, H.-B.; Wang, B.-W.; Pan, F.; Wang, Z.-M.; Gao,
S. Angew. Chem., Int. Ed. 2007, 46, 7388. (f) Bernot, K.; Luzon, J.;
Sessoli, R.; Vindigni, A.; Thion, J.; Richeter, S.; Leclercq, D.;
Larionova, J.; van der Lee, A. J. Am. Chem. Soc. 2008, 130, 1619.
(8) (a) Lecren, L.; Roubeau, O.; Li, Y.-G.; Le Goff, X. F.; Miyasaka, H.;
Richard, F.; Wernsdorfer, W.; Coulon, C.; Cle´rac, R. Dalton Trans.
2008, 755. (b) Miyasaka, H.; Nakata, K.; Lollita, L.; Coulon, C.;
Nakazawa, Y.; Fujisaki, T.; Sugiura, K.-i.; Yamashita, M.; Cle´rac, R.
J. Am. Chem. Soc. 2006, 128, 3770. (c) Miyasaka, H.; Nakata, K.;
Sugiura, K.-i.; Yamashita, M.; Cle´rac, R. Angew. Chem., Int. Ed. 2004,
43, 707.
(9) Zheng, Y.-Z.; Tong, M.-L.; Zhang, W.-X.; Chen, X.-M. Angew. Chem.,
Int. Ed. 2006, 45, 6310.
(10) (a) Caneschi, A.; Gatteschi, D.; Lalioti, N.; Sangregorio, C.; Sessoli,
R.; Venturi, G.; Vindigni, A.; Rettori, A.; Pini, M. G.; Novak, M. A.
Angew. Chem., Int. Ed. 2001, 40, 1760. (b) Liu, T.-F.; Fu, D.; Gao,
S.; Zhang, Y.-Z.; Sun, H.-L.; Su, G.; Liu, Y.-J. J. Am. Chem. Soc.
2003, 125, 13976. (c) Lescoue¨zec, R.; Vaissermann, J.; Ruiz-Pe´rez,
C.; Lloret, F.; Carrasco, R.; Julve, M.; Verdaguer, M.; Draomzee, Y.;
Gatteschi, D.; Wernsdorfer, W. Angew. Chem., Int. Ed. 2003, 42, 1483.
(d) Toma, L. M.; Lescoue¨zec, R.; Pasa´n, J.; Ruiz-Pe´rez, C.; Vaisser-
mann, J.; Cano, J.; Carrasco, R.; Wernsdorfer, W.; Lloret, F.; Julve,
M. J. Am. Chem. Soc. 2006, 128, 4842. (e) Sun, Z.-M.; Prosvirin,
A. V.; Zhao, H.-H.; Mao, J.-G.; Dunbar, K. R. J. Appl. Phys. 2005,
97, 10B305/1. (f) Li, X.-J.; Wang, X.-Y.; Gao, S.; Cao, R. Inorg.
Chem. 2006, 45, 1508.
(11) Zheng, Y.-Z.; Tong, M.-L.; Zhang, W.-X.; Chen, X.-M. Chem.
Commun. 2006, 165.
(12) Zheng, Y.-Z.; Xue, W.; Tong, M.-L.; Chen, X.-M.; Grandjean, F.;
Long, G. J. Inorg. Chem. 2008, 47, 4077.
Inorganic Chemistry, Vol. 47, No. 23, 2008 11203

Zheng et al.
Table 1. Data Collection and Structural Refinement Parameters for 3-7
3
4
5
6
7
formula
f.w.
C8H10FeO4
226.01
C8H12NiO5
246.87
C8H10CoO5
245.09
C32H28Co5O20
1037.28
C9H12CoO4
243.12
T (K)
space group
a (Å)
b (Å)
c (Å)
298(2)
P1 (No. 2)
293(2)
288(2)
293(2)
P1 (No. 2)
293(2)
P1 (No. 2)
j
j
j
C2/c (No. 15)
25.881(7)
6.858(4)
10.376(6)
90
98.96(2)
90
1819.2(16)
8
C2/c (No. 15)
27.083(2)
6.7649(6)
9.4850(8)
90
95.844(2)
90
1728.7(3)
8
5.2211(6)
6.5621(8)
13.360(2)
99.450(2)
98.571(2)
102.775(2)
432.10(9)
2
7.2231(4)
9.1683(6)
14.5105(9)
74.563(1)
78.838(1)
82.749(1)
905.9(1)
2
5.1304(4)
6.5104(6)
14.984(1)
88.659(1)
81.184(1)
78.020(1)
483.76(7)
2
R (deg)
ꢁ (deg)
γ (deg)
V (Å3)
Z
Dc (g cm-3
)
1.737
1.721
1.803
2.125
1.883
1.976
1.901
2.328
1.669
1.757
µ (mm-1
)
data collected/unique
R1 (>2σ/all data)
wR2 (>2σ/all data)
GOF
3151/1664
0.0316/0.0343
0.0750/0.0766
1.071
6252/1734
0.0664/0.0885
0.1481/0.1591
1.027
4280/1661
0.0308/0.0375
0. 0743/0. 0769
1.037
7039/3511
0.0344/0.0374
0.0937/0.0959
1.026
3633/1849
0.0509/0.0587
0.1257/0.1298
1.074
residues (e Å-3
)
–0.272/0.508
–0.857/3.171
–0.308/0.389
–0.578/1.145
–0.583/1.437
0.092 g, 37%), IR data (ν˜, cm-1): 2958(m), 1603(vs), 1539(s),
1445(s), 1415(s), 1332(m), 1293(w), 1236(w), 1115(w), 1033(w),
933(w), 860(w), 775(w), 708(m), 615(w), 526(w), 445(w). Anal.
Calcd (C8H12NiO5): C, 38.92; H, 4.90; Found: C, 38.88; H, 4.98%.
Typical Synthetic Procedure for 2∞[CoII(µ-OH2)(1,2-chedc)] (5).
A mixture of CoCl2 ·6H2O (0.237 g, 1 mmol), 3,4,5,6-tetrahydro-
phthalic anhydride (0.168 g, 1 mmol) and triethylamine (0.200 g,
2 mmol) in deioned water (10 mL) was sealed in a 23 mL Teflon-
lined autoclave and heated at 170 °C for 3 days to give pink needle-
like crystals of 4 (yield 0.14 g, 57%), IR data (ν˜, cm-1): 3425(s),
2931(m), 1550(vs), 1375(s), 1328(s), 988(w), 880(w), 729(w),
450(w). Anal. Calcd (C8H10CoO5): C, 39.20; H, 4.11; Found: C,
39.22; H, 4.13%.
Typical Synthetic Procedure for 2∞[Co5II(µ3-OH)2(OH2)2(1,2-
chedc)4] (6). It was synthesized as for 5 by using KOH (0.112 g, 2
mmol) in place of triethylamine. Purple plate-like crystals of 6 were
isolated in 5% yield (0.011 g), IR data (ν˜, cm-1): 3424(m), 2931(s),
2857(w), 1547(vs), 1418(s), 1384(s), 1333(s), 1070(w), 968(w),
833(w), 805(w), 784(w), 734(w), 638(w), 446(w). Anal. Calcd
(C32H28Co5O20): C, 37.42; H, 2.75; Found: C, 37.38; H, 2.78%.
Typical Synthetic Procedure for 2∞[CoII(4-Me-1,2-chdc)] (7). A
mixture of CoCl2 ·6H2O (0.237 g, 1 mmol), hexahydro-4-methyl-
phthalic anhydride (cis- and trans- mixture) (0.186 g, 1 mmol) and
1,4-diazabicyclo[2,2,2]octane (0.056 g, 0.5 mmol) in deioned water
(10 mL) was sealed in a 23 mL Teflon-lined autoclave and heated
at 170 °C for 3 days to give dark-red plate-like crystals of 7 (yield
0.22 g, 89%), IR data (ν˜, cm-1): 2959(m), 2925(m), 1605(vs),
1543(s), 1449(s), 1418(s), 1337(m), 1295(w), 1239(w), 1118(w),
1035(w), 935(w), 861(w), 776(w), 710(m), 617(w), 528(w), 449(w).
Anal. Calcd (C9H12CoO4): C, 44.46; H, 4.97; Found: C, 44.42; H,
4.98%.
structural refinement parameters are given in Table 1, and selected
bond distances and angles are given in Table 2.
Results and Discussion
Syntheses. It is noteworthy that the conformations of
cyclohexane-1,2-dicarboxylate (1,2-chdc) observed in these
series of compounds are interesting. As shown in our
previous Density-Functional-Theory (DFT) studies,15 the
energy barrier of the trans conformation of 1,2-chdc is only
about 90 kJ mol-1 higher than that of the cis conformation,
and therefore, the conformations of 1,2-chdc can be signifi-
cantly affected by the synthetic conditions such as solvents,
temperature, pressure, and the cooperative effect of coordina-
tion with the metal ions. Thus, the final conformations of
1,2-chdc can not be predicted well. For example, we isolated
a MnII compound of cis-1,2-chdc by using ethylene glycol
as the solvent, while the trans conformation compound was
obtained if water was used instead.15 Moreover, no matter
which solvent was used and how long the reaction time was,
the trans products were always obtained in the presence of
CoII and FeII ions. In this work, we used shorter reaction
time to obtain the cis products in the presence of NiII ions,
but we were unable to obtain other well crystallized products
no matter how we prolonged the reaction time. To rationalize
the presence of the cis or trans form of 1,2-chdc requires
more experimental and theoretical study.
Nevertheless, this synthetic method can be applied to the
structurally related 1,2-chedcH2 and 4-Me-1,2-chdcH2 ligands
to obtain compounds with linear metal-oxygen chains. By
using 1,2-chedcH2, we isolated two new compounds, namely
X-ray Crystallography. Single-crystal X-ray diffraction intensi-
ties of 3-7 have been collected on a Bruker Apex CCD area-
detector diffractometer by using MoKR (λ ) 0.71073 Å) radiation.
Absorption corrections have been applied by using the multiscan
program SADABS.13 The structures were solved with direct
methods and refined with a full-matrix least-squares technique with
the SHELXTL program package.14 Anisotropic thermal parameters
have been assigned to all non-hydrogen atoms. The organic
hydrogen atoms were generated geometrically. Data collection and
aqua-bridged [CoII(µ-OH2)(1,2-chedc)] (5) and hydroxyl-
2
∞
II
bridged 2∞[Co5 (µ3-OH)2(OH2)2(1,2-chedc)4] (6) at lower and
higherpHcondition,respectively.Incontrast,themetal-oxygen
chains in these two compounds are different from those of
2. We believe that the structural change can be attributed to
the rigid CdC bond between two carboxylates in 1,2-chedc
rather than the flexible CsC bond in 1,2-chdc. If the flexible
4-Me-1,2-chdcH2 was employed, we were then able to obtain
(13) Sheldrick, G. M. SADABS 2.05; University Go¨ttingen: Go¨ttingen,
Germany, 2002.
(14) SHELXTL 6.10; Bruker Analytical Instrumentation: Madison, WI,
(15) Zheng, Y.-Z.; Xue, W.; Zheng, S.-L.; Tong, M.-L.; Chen, X.-M. AdV.
Mater. 2008, 20, 1534.
2000.
11204 Inorganic Chemistry, Vol. 47, No. 23, 2008

Single-Chain Magnets
Table 2. Selected Bond Lengths (Å) and Bond Angles (deg) for 2-7
2a
3b
4c
5d
Co1-O1
1.971(3)
2.210(3)
Co1-O2a#
1.986(3)
2.912(1)
Co1-O4b#
2.019(3)
3.269(1)
Co1-O3c#
2.044(3)
101.1(1)
Co1-O4d#
Co1 · · ·Co1a#
Co1 · · ·Co1f#
Co1b#-O4-Co1e#
Fe1-O1
2.229(2)
1.997(2)
Fe1-O2d#
2.079(2)
2.8419(7)
Fe1-O4b#
2.013(2)
3.3939(7)
Fe1-O1c#
2.072(2)
104.13(7)
Fe1-O3a#
Fe1 · · ·Fe1d#
Fe1 · · ·Fe1f#
Fe1c#-O1-Fe1
Ni1-O3a#
Ni1-O2b#
1.985(4)
1.991(4)
Ni1-O1
2.025(4)
2.033(4)
Ni1-O4c#
2.023(4)
3.457(2)
Ni1-O1w
2.210(3)
109.1(2)
Ni1-O1wb#
Ni1 · · ·Ni1d#
Ni1d#-O1w-Ni1
Co1-O1
2.009(2)
2.068(2)
96.81(7)
Co1-O4b#
Co2-O1w
2.079(2)
2.101(2)
Co1-O1w
Co1 · · ·Co2
2.214(2)
3.146(2)
Co2-O4c#
Co1-O1w-Co2
2.127(2)
93.59(7)
Co2-O3b#
Co1g#-O4-Co2g#
6e
Co1-O8a#
Co2-O2
Co2-O1w
Co3-O6c#
Co2-O5-Co1
2.016(2)
2.290(2)
2.084(2)
1.944(2)
91.08(7)
Co1-O5
2.173(2)
2.063(2)
2.099(2)
3.094(2)
117.2(1)
Co1-O9
2.084(2)
2.162(2)
1.944(2)
3.273(2)
96.50(8)
Co2-O7a#
Co3-O9
2.028(2)
1.947(2)
1.990(2)
3.423(2)
108.53(9)
Co2-O9
Co2-O5
Co2-O4d#
Co1 · · ·Co2
Co3-O9-Co2
Co3-O1
Co1 · · ·Co3
Co2-O9-Co1
Co3-O3b#
Co2 · · ·Co3
Co3-O9-Co1
7f
Co1)-O1
1.962(3)
2.206(3)
Co1-O2a#
1.981(3)
2.8974(9)
Co1-O4b#
2.013(2)
3.253(1)
Co1-O3c#
2.031(3)
100.8(1)
Co1)-O4d#
Co1 · · ·Co1a#
Co1 · · ·Co1f#
Co1b#-O4-Co1e#
a Symmetry codes for 2: (a#) -x + 2, -y + 1, -z + 1; (b#) -x + 1, -y + 2, -z + 1; (c#) -x + 2, -y + 2, -z + 1; (d#) x, y - 1, z; (e#) x, y + 1,
z; (f#) -x + 1,-y + 1, -z + 1. b Symmetry codes for3: (a#) x, y + 1, z; (b#) -x, -y - 1, -z; (c#) -x+1,-y,-z; (d#) -x,-y,-z. c Symmetry codes for
4: (a#) x, -y + 1, z - 1/2; (b#) -x + 1/2, y + 1/2, -z + 1/2; (c#) -x + 1/2, -y + 1/2, -z + 1; (d#) -x + 1/2, y - 1/2, -z + 1/2. d Symmetry codes
for 5: (b#) -x + 1, y - 1, -z + 1/2; (c#) x, y - 1, z; (g#) x, y + 1, z. e Symmetry codes for 6: (a#) -x, -y + 1, -z + 2; (b#) x + 1, y, z; (c#) -x + 1,
-y + 1, -z + 2; (d#) -x, -y + 2, -z + 2. f Symmetry codes for 7: (a#) -x + 2, -y + 1, -z + 1; (b#) -x + 1, -y + 2, -z + 1; (c#) -x + 2, -y +
2, -z + 1; (d#) x, y - 1, z; (e#) x, y + 1, z; (f#) -x + 1, -y + 1, -z + 1.
a
b
c
d
Figure 1. Coordination environment (a) (symmetry codes: x, -y + 1, z - 1/2; -x + 1/2, y + 1/2, -z + 1/2; -x + 1/2, -y + 1/2, -z + 1; -x +
e
1/2, y - 1/2, -z + 1/2; x, 1 - y, z + 1/2) and the layered structure (b) in 4.
2∞[CoII(4-Me-1,2-chdc)] (7), which is isomorphous to 2. These
results show that the flexible CsC bond and the trans
configuration between two carboxylates are critical to
maintain the framework structure of 2.
As shown in Figure 1a, each asymmetric unit contains one
NiII ion, one µ-aqua molecule, and one e,a-cis-1,2-chdc
ligand. The octahedral NiII ions are bridged by trans-related
axial aqua (NisOaquasNi 109.1(2)°) and two syn-syn-µ-η1:
η1 carboxylate ligands. As shown in Figure 1b, the NisO
chain goes along the b-axis with equal NiII · · · NiII distances
of 3.457 Å. However, the NiII ion does not sit on an inversion
center but a 21-screw axis running along the b-axis. The
function of e,a-cis-1,2-chdc is similar to those in 2 and 3,
which connects the NisO chains to form a 2D layer with a
shortest interchain NiII · · ·NiII distance of 5.539 Å. The layers
are also stacked only via van der Waals interactions with
the shortest NiII · · · NiIIdistance up to 12.92 Å.
Structural Characterization. The crystal structure of 3
is isomorphous with 2, as shown in Supporting Information,
Figure S1. The FesO chain is also constructed by the tetra-
carboxylate-bridged paddle-wheel FeII dimers. The intradimer
FeII · · · FeII distance is 2.842 Å, being slightly shorter than
that in 2. However, the interdimer FeII · · · FeII distance and
FesOsFe angle of bis(µ-η2)-carboxylate-bridged FeII ions
is 3.393(1) Å and 104.1(1)°, respectively. These values are
slightly larger than those in 2, as shown in Table 2. Such
FesO chains are further connected by the trans-1,2-chdc
into 2D layers, which are also stacked only by van der Waals
interactions with the shortest interlayer FeII · · · FeII distance
of 12.0 Å.
As shown in Figure 2, the asymmetric unit of 5 contains
two half-occupied CoII ions, one µ-aqua molecule, and one
1,2-chedc ligand. It is interesting that Co1 and Co2 have
elongated and compressed octahedral geometries, respec-
Inorganic Chemistry, Vol. 47, No. 23, 2008 11205

Zheng et al.
a
b
c
Figure 2. Coordination environments (a), CosO chain (b) and layer (c) in 5. Symmetry codes: -x + 1, y, -z + 1/2; -x + 1, y - 1, -z + 1/2; x, y
d
e
f
g
h
- 1, z; x, -y + 2, z + 1/2; -x + 1, -y + 1, -z + 1; -x + 1, -y + 2, -z + 1; x, y + 1, z; -x + 1, y + 1, -z + 1/2.
a
b
Figure 3. Coordination environments (a), CosO chain (b), and ab-plane layered structure (c) in 6. Symmetry codes: -x, -y + 1, -z + 2; x + 1, y, z;
c
d
-x + 1, -y + 1, -z + 2; -x, -y + 2, -z + 2.
tively. The elongated axial positions of Co1 are occupied
by two aqua molecules (CosO 2.214(2) Å) while the
equatorial plane is coordinated by four carboxylate oxygen
atoms (CosO 2.009(2)-2.079(2) Å). The compressed axial
positions of Co2 are coordinated by two carboxylate oxygen
atoms (CosO 2.068(2) Å) and the equatorial plane is
coordinated by a pair of trans-related carboxylate oxygen
atoms and a pair of trans-related aqua molecules (CosO
2.101(2)-2.127(2) Å). Being different with 4, the zigzag
CosO chain goes along the c direction rather than the b
direction. It should be noted that the equatorial planes of
the intrachain adjacent to Co2 ions have a dihedral angle of
47.2°. The CosO chain consists of edge-sharing octahedral
CoII ions with equal space (CoII · · · CoII 3.146(2) Å). The
bridging oxygen atoms are from one µ3-carboxylate end (O4)
and one µ-aqua ligand (O1w). The shortest CoII · · · CoII
distance between adjacent chains is 5.262 Å, and the layers
are stacked throught van der Waals interaction with the
shortest CoII · · · CoII distance up to 13.57 Å.
Å) and two hydroxide oxygen atoms (CosO 2.084(2) Å).
The µ3-OH group further connects with two other CoII ions,
octahedral Co2, and tetrahedral Co3. Co2 is coordinated by
six oxygen atoms from one aqua (CosO 2.084(2) Å), one
hydroxide (CosO 2.063(2) Å), and four carboxylates (CosO
2.028(2)s2.290(2) Å). Co3 is coordinated by four oxygen
atoms from one hydroxide (CosO 1.947(2) Å) and three
carboxylates (CosO 1.944(2)-1.990(2) Å). Thus, a penta-
nuclear [Co5(µ3-OH)2]8+ unit forms, which consists of three
edge-sharing octahedral CoII ions in the “body” with two
corner-sharing tetrahedral CoII ions at both sides. The
pentanuclear units are connected by the syn-anti carboxylates
into the chains along the b-axis (CoII · · ·CoII 5.500 Å), which
are connected by 1,2-chedc into a layer with the shortest
CoII · · · CoII distance of 5.034 Å. Such layers are further
stacked by van der Waals interactions with the shortest
CoII · · · CoII distance up to 10.25 Å.
Compound 7 is structurally very similar to that of 2 (Figure
4),9 which also features CoII-carboxylate chains based on
paddle-wheel dimers, being further linked by 4-Me-1,2-chdc
into 2D layers. These layers are also stacked only by van
der Waals interactions with the shortest CoII · · ·CoII distance
up to 13.71 Å, which is significantly larger than 12.08 Å in
2 because of the presence of a methyl group. More
importantly, all the CosO bonds, CoII · · ·CoII distances, and
CosOsCo angle in 7 are slightly shorter or smaller than
As shown in Figure 3, the asymmetric unit of 6 contains
two and a half CoII ions, one µ3-hydroxide, one terminal aqua,
and two 1,2-chedc ligands. The Co1 ion sits on the inversion
center in an elongated octahedral geometry, in which the
axial positions are occupied by two carboxylate oxygen
atoms (CosO 2.173(2) Å), and the equatorial positions are
occupied by two other carboxylate oxygen (CosO 2.016(2)
11206 Inorganic Chemistry, Vol. 47, No. 23, 2008

Single-Chain Magnets
Figure 4. Perspective view of the layer in 7.
Figure 6. Left: plot of ꢂT vs T at an applied field of 500 Oe from 2 to 310
K in 4. Solid line: the fitting curve with the Fisher model from 310 to 150
K. Dotted line: the extrapolation of the fitting curve from 150 to 2 K. Right:
the temperature-dependent inverse susceptibility of 4. Dashed line: the fitting
curve of the Curie-Weiss law.
ferromagnetic interaction within the chain, which could
probably be due to the bis(µ2-η2)-carboxylate bridge between
the FeII ions. However, we can not exclude the possiblity of
ferromagnetic interaction through the 4-fold syn-syn car-
boxylate bridge. The data below 30 K can not be fitted by
this Heisenberg model, which indicates that a more com-
plicated Hamiltonian including single-ion magnetic anisot-
ropy and spin-orbital coupling is necessary. As indicated
in the ac susceptibility data (Supporting Information, Figure
S2), such intrachain F-AF interaction does not lead to either
an SCM or a long-range magnetic ordering behavor in 3 since
no slow-relaxation or sharp peaks were observed. In addition,
the unsaturated M(H) plot even up to 7 T at 2 K confirms
the magnetic behavior of anisotropic iron(II) ions.
Figure 5. Plot of ꢂT vs T at an applied field of 500 Oe from 2 to 300 K
in 3. Solid line: the fitting curve with the Fisher model from 300 to 35 K.
Dotted line: the extrapolation of the fitting curve from 30 to 2 K.
the corresponding ones in 2 (Table 2). These small changes
of CosO bonds and CosOsCo angles have not led to a
marked change of the structure but have perturbations on
the magnetic properties of this structure (vide infra).
Non-SCM Behaviors in 3-6. As shown in Figure 5, the
ꢂT value of 3 at room temperature is 5.41 cm3 mol-1 K,
significantly large than the spin-only value of FeII (3.00 cm3
mol-1 K), indicating a large spin-orbital contribution of the
FeII ion in the square-pyramidal geometry. Upon cooling,
the ꢂT slightly decreases to a minimum of 5.37 cm3 mol-1
K at 240 K and then gradually increases, reaching a
maximum of 5.69 cm3 mol-1 K at 70 K. This shape of the
ꢂT plot is usually indicative of uncompensated spins similar
to those in ferrimagnets, weak ferromagnets, alternating
antiferromagnetic-ferromagnetic (AF-F) spin chains, and/or
compounds containing orbitally degenerate magnetic ions,
such as FeII ions herein. The large decrease of the ꢂT plot
below 30 K is expected for zero-field splitting of FeII ions
or significant interchain AF interaction. To study the
magnetic interactions in 3, a Heisenberg-Fisher model
(eq 1) was used to fit the data above 30 K,17
The dc susceptibility data of 4 (Figure 6) show a ꢂT value
of 1.08 cm3 mol-1 K at room temperature, which is very
close to that expected for a spin-only NiII ion (1.00 cm3 mol-1
K). Upon cooling, ꢂT continuously decreases. The inverse
molecular magnetic susceptibility, ꢂ-1, obeys the Curie-Weiss
law from 100-310 K and can be fitted with Curie-Weiss
constants, C and θ, of 1.23 cm3 mol-1 K and -47.5 K,
respectively. The C value corresponds to an effective
magnetic moment of 3.14 µB, and the large negative θ value
indicates significant AF interactions between the NiII ions.
A 1D Fisher chain model (for S ) 1),17 H ) sJ∑SiSi+1
,
ˆ
was tried to fit the data. However, this Heisenberg model
can not fit the data well below 150 K, as shown in Figure 6.
The dotted line represents the extrapolation of curve above
150 K with parameters of J/kB ) -23(1) K and g ) 2.22(1).
The failure of this model may indicate that the effect of zero-
field splitting and/or interchain interactions can not be
neglected at low temperature. Similar to that of 3, 4 does
not show either SCM or long-range magnetic ordering
behavior. The ꢂT plot decreases quickly below 30 K, which
may also indicate large zero-field splitting effect of the NiII
ions or interchain AF interaction, leading to a diamagnetic
ground state. As shown in the Supporting Information, Figure
S3, the ac susceptibility data of 4 also show only a
paramagnetic-like behavior down to 2 K. In addition, the 2
K M(H) plot increases almost linearly and does not show
saturation up to 7 T, which is probably consistent with the
observed intrachain AF interactions in 4.
N
N
ˆ
H ) -J
S S
- J
S S
2i 2i+2
(1)
∑
∑
1
2i 2i+1
2
i)1
i)1
where J1 and J2 are magnetic interactions in the chain (note
that this model can be reduced to the uniform chain if J1 )
J2). The best fitting gives J1/kB ) 9.7(5) K, J2/kB ) -3.7(1)
K, and g ) 2.623(5). The opposite signs of J indicate the
alternating F-AF exchange coupling interactions in the chain.
The larger magnitude of the positive J1 suggests a stronger
The dc susceptibility data (Figure 7) measured on a powder
sample of 5 show that the ꢂT value of 5 at 320 K is 3.20
cm3 mol-1 K per CoII, being significantly larger than the spin-
(16) (a) Kahn, O. Molecular Magnetism; VCH: New York, 1993. (b) Carlin,
R. L. Magnetochemistry; Springer: Berlin, 1986.
(17) Fisher, M. E. Am. J. Phys 1964, 32, 343.
Inorganic Chemistry, Vol. 47, No. 23, 2008 11207

Zheng et al.
significantly larger than the spin-only value of 9.38 cm3
mol-1 K, hence indicating a large orbital contribution. Upon
cooling, ꢂT continuously decreases to 2.0 K. The data in
the range of 20-320 K obeys the Curie-Weiss law with C
) 19.96 cm3 mol-1 K and θ ) -62.37 K. This large negative
θ value indicates not only spin-orbital coupling of the CoII
ion but also significant AF interactions between the CoII ions,
which are caused by the more effective magnetic com-
munication via the µ3-OH groups. The θ value is consistent
with those of reported CoII compounds with µ3-OH bridges.11
The ac susceptibility data (inset of Supporting Information,
Figure S5) also do not suggest any slow-relaxation behavior
in 6. In addition, the maximum value of the M(H) plot is
3.81 µB at 7 T (Supporting Information, Figure S5), being
significantly below the expected value for a CoII5 unit, which
is consistent with the AF interactions between the spins.
SCM Behavior in 7. The dc susceptibility data (Figure
9) measured on a powder sample of 7 show that the ꢂT value
at room temperature is 3.32 cm3 mol-1 K per CoII, which is
much larger than the spin-only value (1.88 cm3 mol-1 K),
as well as that (2.98 cm3 mol-1 K) of 2, hence indicating a
very large orbital contribution even larger than that of 2.
Upon cooling, it gradually decreases to a minimum of 3.28
cm3 mol-1 K at 176 K, and then increases. When fitting ꢂ-1
vs T data from 304 to 176 K with the Curie-Weiss law,
one gets C ) 3.39 cm3 mol-1 K and θ ) -6.73 K
(Supporting Information, Figure S6).
Figure 7. Plot of ꢂT vs T at an applied field 500 Oe from 2 to 320 K in
5. Inset: enlarged ꢂT vs T plot from 2-30 K.
Scheme 2. Spin-Coupling Scheme of the ∆-Chain in 5
Scheme 3. Spin-Chain Coupling Scheme of 6
only value of 1.88 cm3 mol-1 K and hence also indicates
large orbital contribution.16 Upon cooling, it gradually
decreases to about 30 K with the presence of a minimum of
2.74 cm3 mol-1 K, and then it increases to a maximum of
3.64 cm3 mol-1 K at 3.5 K. The data in the range of 40-320
Below 30 K, the ꢂT value increases quickly up to a
maximum value of 62.84 cm3 mol-1 K at 5.8 K. Because
the square-pyramidal CoII ions have the similar effective spin-
1/2 as that of the octahedral CoII ions because of the
combined action of spin-orbital coupling and noncubic
K obeys the Curie-Weiss law with C ) 3.29 cm3 mol-1
K
and θ ) –7.87 K. From the chain structure of 5 (Figure 2b),
one can get that the main magnetic exchange coupling ways
are a typical corner-sharing triangular chain, or called
∆-chain, which is part of a Kagome´ lattice,18 as shown in
Scheme 2. There are two main magnetic exchange coupling
ways. J1 is via the (µ-aqua)(µ-carboxylate)-bridge, and J2 is
through the syn-anti µ-carboxylate-bridge between the CoII
ions. However, the f value, which can be used to measure
the extent of spin-frustration, is defined as f ) |θ/Tc|. For 5,
f ) |7.87/3.5| ) 2.24 < 5 does not suggest any spin-
frustration effect.19 Magnetic competitions such as ferro- and/
or antiferromagnetic interactions are responsible for the non-
diamagnetic ground-state in 5.20 The M(H) plot gives the
maximum value of 2.40 µB at 7 T, which is consistent with
a typical value of one uncompensated octahedral CoII ion
(Supporting Information, Figure S4).16,21,22
Figure 8. Left: plot of ꢂT vs T at an applied field of 500 Oe from 2 to 320
K in 6. Right: the temperature-dependent inverse susceptibility of 6. Solid
line: the fitting curve with the Curie-Weiss law.
Compound 6 has a pentanuclear CoII-based spin-chain, as
shown in Scheme 3. The dc susceptibility data (Figure 8)
measured on a powder sample of 6 show that the ꢂT value
of 6 at 320 K is 16.90 cm3 mol-1 K per CoII , which is
5
(18) (a) Zheng, Y.-Z.; Xue, W.; Tong, M.-L.; Chen, X.-M.; Grandjean, F.;
Long, G. J. Angew. Chem., Int. Ed. 2007, 46, 6076. (b) Cheng, X.-N.;
Zhang, W.-X.; Zheng, Y.-Z.; Chen, X.-M. Chem. Commun. 2006,
3603. (c) Cave, D.; Coomer, F. C.; Molinos, E.; Klauss, H.-H.; Wood,
P. T. Angew. Chem., Int. Ed. 2006, 45, 803.
(19) (a) Ramirez, A. P. Annu. ReV. Mater. Sci. 1994, 24, 453. (b) Greedan,
J. E. J. Mater. Chem. 2001, 11, 37. (c) Harrison, A. J. Phys.: Condens.
Matter 2004, 16, S553.
(20) Chikazumi, S. Physics of Ferromagnetism; Clarendon Press, Oxford
Science Publications: Oxford, 1997.
Figure 9. Plot of ꢂT vs T at an applied field of 1 KOe from 35 to 304 K
in 7. Dashed line: fitted by eq 5. Inset: ꢂT vs T plot from 2-30 K. Solid
line: fitted by spin-1/2 alternating Ising chain model.
11208 Inorganic Chemistry, Vol. 47, No. 23, 2008

Single-Chain Magnets
crystal-field terms, giving six-Kramers doublets at low
temperatures,16,22 the data below 30 K can be treated as a
spin-1/2 Ising chain with alternating space sites, namely, with
two nearest neighbor interactions (J and j), and alternating
Lande´ factors (ga and gb). The full Hamiltonian is written as
follows23
N
N
H ) -J Sz Sz - j Sz Sz
-
ˆ
∑
∑
2i-1 2i
2i 2i+1
i)1
i)1
N
N
g
Sz + g
Sz µBH (2)
∑
∑
a
2i-1
b
2i
(
)
i)1
i)1
where Szi stands for the z component of the spin operator
located on site i, µB is the Bohr magneton, and H is the
external magnetic field. This Hamiltonian was solved by
using the transfer matrix method by Drillon et al.,23,24 giving
rise into the final analytical expressions, or eqs 3 and 4.
Nµ2B g2+ exp(J+ ⁄ 2kBT) + g-2 exp(J- ⁄ 2kBT)
Figure 10. Top: real part of ac susceptibility, inset: Cole-Cole diagram
at 4.2 K. Bottom: imaginary part of ac susceptibility; inset: the peak
temperatures of ꢂ′′ fitted by the Arrhenius law for 7.
ꢂ| )
(3)
2kBT
cosh(J- ⁄ 2kBT)
Ng2⊥µ2B
ꢂ⊥ )
[tanh(J ⁄ 4kBT) + (J ⁄ 4kBT) sech2(J ⁄ 4kBT)]
with the Ising model rather than a 3D long-range magnetic
ordering behavior.
2J
(4)
To better understand the ground-state of 7, the ꢂT data
were fitted by eq 5,25
where g( ) (ga ( gb)/2 and J( ) (J ( j)/2. In the case of
7, the alternating Lande´ factors (ga and gb) can be simplified
to ga ) gb ) g because all the CoII ions are crystallographi-
cally identical. For a powder sample of 7, ꢂm ) 1/3ꢂ| + 2/3ꢂ⊥
was applied to obtain an average value. During the fitting
process, we found that g⊥ tends to be zero, and the best fit
(from 30 to 9 K) gives g| ) 6.6(1), g⊥ ) 0, J ) 36(5) cm-1,
and j ) 25(1) cm-1 (see the inset of Figure 9). The large
value of g| indicates strong magnetic anisotropy of the CoII
ions, which is consistent with other reported CoII ions with
octahedral, square-pyramidal, or triangular-bipyramidal ge-
ometries.16,22,23 The unequal, large positive coupling con-
stants indicate strong alternating ferromagnetic interactions
in 7, which is consistent with those of 2. For comparison,
the same equations were applied to fit the ꢂT data of 2 from
30 to 9 K, giving the best fitting parameters of g| ) 5.7(4),
g⊥ ) 0, J ) 57(12) cm-1, and j ) 26.2(8) cm-1, as shown
in Supporting Information, Figure S7. These values show
that the ferromagnetic interactions in 2 are stronger than those
in 7, while the g| value of 2 is slightly smaller than that of
7. The larger g| value of 7 may suggest that the spin-orbital
coupling effect in 7 is enhanced in comparison to that of 2.
This conclusion is consistent with the observed decrease of
ꢂT vs T plot of 7 at higher temperatures. The well-fitted curve
confirms that the magnetic behavior of 7 is indeed consistent
ꢂT ) C1 exp(RJ ⁄ kBT) + C2 exp(ꢁJ ⁄ kBT)
(5)
where RJ < 0 indicates the spin-orbital coupling of the CoII
ions, being responsible for the initial high temperature decay
of ꢂT; ꢁJ > 0 denotes the ferromagnetic-like interaction,
leading to the increase of ꢂT at low temperatures. C1+C2
equals the Curie constant at high temperatures. Fitting the
data from 304 to 35 K (Figure 9) gives RJ/kB ) -265(11)
K, C1 ) 0.77(1) cm3 mol-1 K, ꢁJ/kB ) 15.1(2) K and C2 )
2.86(1) cm3 mol-1 K. To check the phase transition behavior
of 7, the curve of dlog(T)/dlog(ꢂmT)ferro vs T was plotted in
Supporting Information, Figure S8.25 Here (ꢂmT)ferro ) ꢂmT
s 0.77 × exp(-265/T) describes the ferromagnetic contribu-
tion, which is obtained by subtracting the spin-orbital
coupling component that dominates at the high temperature
regime from the total ꢂmT. A direct fit of the temperatures
from 30 to 8 K yields Tc ) 0 K and γ ) ∞, which is
consistent with the theoretical 1D-Ising value (see Supporting
Information, Figure S8).25,26 This behavior may rule out the
phase transition to long-range ordering of this system above
0 K and further confirms that 7 is indeed a real 1D system.
Thus, strongly frequency-dependent ac susceptibilities (ꢂ′
and ꢂ′′ are the real and imaginary components of ac
susceptibility, respectively) of 7 below 7 K were observed
(Figure 10), which is the typical Ising-type ferromagnetic
(21) Kapoor, R.; Kataria, A.; Venugopalan, P.; Kapoor, P.; Hundal, G.;
Corbella, M. Eur. J. Inorg. Chem. 2005, 3884.
(22) Bacˇo, R. Coord. Chem. ReV. 2004, 248, 757.
(23) Coronado, E.; Drillon, M.; Nugteren, P. R.; de Jongh, L. J.; Beltra´n,
D. J. Am. Chem. Soc. 1988, 110, 3907.
(24) Georges, R.; Borra´s-Almenar, J. J.; Coronado, E.; Cure´ly, J.; Drillon
M. One-dimensional Magnetism: An OVerView of the Models in
Magnetism: Molecules to Materials, Miller, J. S., Drillon, M., Eds.;
Wiley-VCH Verlag: Berlin, 2001; Vol I, pp 1-47.
(25) (a) Drillon, M.; Panissod, P.; Rabu, P.; Souletie, J.; Ksenofontov, V.;
Gu¨tlich, P. Phys. ReV. B 2002, 65, 104404. (b) Souletie, J. ; Rabu, P.;
Drillon, M. Scaling Theory Applied to Low Dimensional Magnetic
System in Magnetism: Molecules to Materials; Miller, J. S., DrillonM.,
Eds.; Wiley-VCH: Weinheim, Germany, 2005; Vol. V, pp 347-377.
(26) (a) Widom, B. J. Chem. Phys. 1965, 43, 3892. (b) Widom, B. J. Chem.
Phys. 1965, 43, 3898.
Inorganic Chemistry, Vol. 47, No. 23, 2008 11209

Zheng et al.
Figure 11. Logarithm of ꢂT vs 1/T of 7. The ac susceptibility was obtained
at a frequency of 1 Hz in an ac field of 5 Oe and a zero dc field. The dc
susceptibility was obtained in an applied field of 1 kOe. The red line is the
result of a fit between 5.6 and 40 K.
Figure 12. Hysteresis loop of 7 at 1.8 K. Inset: enlarged region in the
range of ( 0.5 kOe.
(ZFC) and field-cooled (FC) magnetization at 4.0 K (Sup-
porting Information, Figure S9). The coercive field of the
hysteresis loop is 110 Oe, which is similar with that of 2,
confirming a magnet-type behavior. Moreover, the maximum
magnetization value of 7 at 2.5 K reaches 2.41 µB at 7 T
(Supporting Information, Figure S10), which is also similar
with that of 2.
chain behavior predicted by Glauber.27 The peak tempera-
tures (TB) of ꢂ′′ can be fitted well by the Arrhenius plot
extracted from these data, showing the occurrence of a clear
crossover between two different activated regimes at T* )
4.7 K (1/T* ) 0.215 K-1), and giving the characteristic
relaxation time, τ0, of 3.7 × 10-13 and 1.2 × 10-10 s for the
high and low temperature regimes, as well as two corre-
sponding different energy barriers, ∆τ1/kB ) 109.2 and ∆τ2/
kB ) 82.1 K, respectively. This result is similar to the
observed finite-size effect28 halving the Glauber activation
barrier in 2. However, they are both about 29 K larger than
those of 2.
In addition, the peak temperatures (Tp) of ꢂ′ can be
measured by a parameter φ ) (∆Tp/Tp)/∆(log f) ) 0.13,29
which is in the range of a normal superparamagnet and
precludes the possibility of a spin-glass. Moreover, a shape
of the Cole-Cole diagram (inset of Figure 10) can be
obtained at 4.5 K, which can be fitted by a generalized Debye
model with an R value of 0.23, indicating a narrow-
distribution of relaxation time.30
2 and 7 actually have similar SCM behavior except that
the obtained energy barrier of 7 is larger than that of 2. If
considering the anisotropic energy barrier, ∆A, which equals
[∆τ - 2(∆τ1 - ∆τ2)]/kB, one gets the result ∆A7/kB ) 55 K
> ∆A2/kB ) 19.5 K. Therefore, the reason for the larger ∆τ
of 7 is probably due to the larger ∆A in 7, which is actually
related to the magnetic anisotropy of Co(II) ions in these
two compounds. As is known, the lowest orbital state of the
4
free CoII ion is F1. In the weak field of Oh symmetry, the
4F1 term splits into 4T1g,4T2g, and 4A2g terms. With reduction
4
of symmetry to D4h, further splitting occurs. First, the T1g
crystal field term splits into 4Eg and 4A2g, where the ground
4
state, A2g, can be further splits into Kramers doublets Γ6
and Γ7 when spin orbit coupling is switched on.16b The
difference between Γ6 and Γ7 states can be written as 2D.22
Therefore, the magnetic anisotropy of CoII ions is mainly
governed by the crystal field and the spin-orbital coupling
terms. Here, we estimated the electrostatic potentials of the
metal centers and ligands of 2, 5, and 7, which might explain
some of the differences of the crystal field between 2 and 7
(vide infra).
The temperature-dependent exponential behavior is evident
from the plot of ln(ꢂT) vs 1/T. Figure 11 shows ln(ꢂT)
increasing linearly between 40 and 5.6 K with an energy
gap of ∆ꢀ7/kB ) 20.4(3) K and Ceff ) 2.78 cm3 mol-1
K
(obtained by fitting the expression ꢂT ) Ceff exp(∆ꢀ/kBT)
between 5.6 to 40 K). This value of ∆ꢀ7/kB is smaller than
that of 2 (∆ꢀ2/kB ) 24.0 K). Moreover, from the saturation
value of ꢂT, we can estimate the number of spin unit (n) to
be about 23, and therefore, the real chain length (L) of the
7 to be about 11.6 nm, which is slightly longer than 10.8
nm of 2 (Ceff ) 2.47 cm3 mol-1 K, n ) 21).
Below the blocking temperature, the irreversibility effect
was observed in 7 with a clear hysteresis loop at 1.8 K
(Figure 12) as well as the discrepancy of zero-field-cooled
Electrostatic Potential Calculations. To further compare
the effect of crystal field, we have also performed electro-
static potential calculations based on Density-Function-
Theory (DFT) on 2, 5, and 7. To simplify DFT calculations
and give a clear-cut picture, a primitive model of one CoII
ion and one deprotonated dicarboxylate ligand was used at
the beginning. As shown in Figure 13, the resulting electronic
density maps of the three metal-ligand systems are similar,
having the positive density all on the CoII ion and negative
charge on the carboxylate oxygens. Meanwhile, the optimized
carboxylate conformations are different; those in 1,2-chdc
and 4-Me-1,2-chdc adopt a similar parallel arrangement,
while those in 1,2-chedc feature a perpendicular arrangement
owing to the CdC double bond between them. Hence 1,2-
chdc and 4-Me-1,2-chdc may lead to similar CoII compounds,
while 1,2-chedc probably does not.
(27) Glauber, R. J. J. Math. Phys. 1963, 4, 294.
(28) (a) Coulon, C.; Cle´rac, R.; Lecren, L.; Wernsdorfer, W.; Miyasaka,
H. Phys. ReV. B 2004, 69, 132408/1. (b) Bogani, L.; Caneschi, A.;
Fedi, M.; Gatteschi, D.; Massi, M.; Novak, M. A.; Pini, M. G.; Rettori,
A.; Sessoli, R.; Vindigni, A. Phys. ReV. Lett. 2004, 92, 207204/1. (c)
Bogani, L.; Sessoli, R.; Pini, M. G.; Rettori, A.; Novak, M. A.; Rosa,
P.; Massi, M.; Fedi, M. E.; Giuntini, L.; Caneschi, A.; Gatteschi, D.
Phys. ReV. B 2005, 72, 064406/1.
(29) (a) Ma, S. K. Phys. ReV. B 1980, 22, 4484. (b) Mydosh, J. A. Spin
Glasses: An Experimental Introduction; Taylor & Francis: London,
1993.
(30) Cole, K. S.; Cole, R. H. J. Chem. Phys. 1941, 9, 341.
11210 Inorganic Chemistry, Vol. 47, No. 23, 2008

Single-Chain Magnets
requires other techniques such as single-crystal magnetism,
high-frequency EPR,2b inelastic neutron diffraction studies,
and so forth.
Conclusions
Five new metal-carboxylate layered compounds containing
four kinds of metal-oxygen chains have been hydrother-
mally synthesized. In these series of magnetic chain-based
compounds, only the dimeric paddle-wheel cobalt(II)-car-
boxylate chain exhibit SCM behavior no matter whether CoII
ion coordinates to trans-1,2-chdc or 4-Me-1,2-chdc. The
presence of SCM behavior in these two CoII compounds is
mainly due to (i) the alternating ferromagnetic interactions
in the chain, (ii) the Ising-type magnetic anisotropy of CoII
ion, and (iii) the very weak interchain magnetic interactions
that are achieved by the CsC σ bond and the special steric
arrangement of the carboxylates. Although the FeII trans-
1,2chdc compound satisfies the last two requirements, the
intrachain magnetic interactions are no longer alternating
ferromagnetic interactions but alternating ferro- and antiferro-
magnetic. Therefore, 3 does not exhibit SCM behavior. Other
type of chains such as the uniform aqua-carboxylate-bridged
NiII chain in 4, CoII based ∆-chain in 5, and the pentanuclear
CoII chain in 6 also do not exhibit strong ferromagnetic
interactions, and no SCM behavior is observed in these
compounds. Although 2 and 7 have very similar SCM
behavior, ac magnetic studies show 7 has a higher energy
barrier than that of 2. Such behavior is probably caused by
the larger anisotropic energy barrier in 7.
Figure 13. Electrostatic potential mapped on the 0.05 au density isosurface
in the MsO from G03/B3LYP for Co(1,2-chdc) (a), Co(4-Me-1,2-chdc)
(b), and Co(1,2-chedc) (c). Color codes from blue to red: from +0.6244 to
-0.0202 in (a), from +0.6462 to -0.0295 in (b), from +0.6445 to -0.0107
in (c).
Table 3. DFT Results of the Simplest Co(II)-Dicarboxylate Model
1,2-chdc 4-Me-1,2-chdc 1,2-chedc
Mayer Bond Order of CosO
Mulliken atomic charge of Co
Mulliken atomic charge of O
Charge difference of Co and O
0.6135
0.4890
-0.5626
1.0516
0.6221
0.4992
-0.5783
1.0775
0.6192
0.5265
-0.5322
1.0587
More importantly, the 4-Me-1,2-chdc-CoII system has the
largest electrostatic potential difference between the CoII and
carboxylate oxygen, implying the strongest attraction be-
tween the CoII and carboxylate among these systems, which
is further supported by the Mayer bond order and Mulliken
atomic charge difference analyses (Table 3).31,32 This sug-
gests that a strongest ligand field effect can be found on the
energy level of the CoII ion of the 4-Me-1,2-chdc-CoII
system.16,21,22 Electrostatic potential calculations on a more
extended model of a four carboxylate-bridged CoII paddle-
wheel dimer of 1,2-chdc and 4-Me-1,2-chdc also indicates
stronger attraction between the CoII and carboxylates in the
4-Me-1,2-chdc-CoII system (see Supporting Information,
Figures S11 and S12 and Table S1), thus further implying
that similar results could occur in the fully extended
polymeric structures. As some reports show that the D value
of five-coordinated CoII has an increasing tendency as the
ligand field increased,16 at least we may expect to perturb
the magnetic properties of 2 by replacement of 1,2-chdc into
the subtly modified analogue 4-Me-1,2-chdc, although a
contrary result was observed for MnII compounds.33 There-
fore, the larger magnetic anisotropy of 7 might be partially
due to the stronger ligand field of 4-Me-1,2-chdc. However,
an accurate analysis of magnetic anisotropy of CoII ions
Acknowledgment. This work was supported by the “973
Project” (Grant No. 2007CB815302 and 2007CB815305),
NSFC (Grant No. 20525102) and NSF of Guangdong (Grant
No. 04205405). The authors thanks Prof. Marc Drillon
(IPCMS, 23 rue du Loess, B.P. 43, 67034 Strasbourg Cedex
2 France) for helpful discussions.
Supporting Information Available: Crystallographic data in
CIF file format, additional structural plots, additional magnetic data,
and the X-ray diffraction patterns. This material is available free
of charge via the Internet at http://pubs.acs.org.
(31) (a) Mayer, I. Chem. Phys. Lett. 1983, 97, 270. (b) Mayer, I. Int. J.
Quantum Chem. 1984, 26, 151. (c) Bridgeman, A. J.; Cavigliasso,
G.; Ireland, L. R.; Rothery, J. Dalton Trans. 2001, 2095.
IC801498N
(32) (a) Zheng, S.-L.; Messerschmidt, M.; Coppens, P. Angew. Chem., Int.
Ed. 2005, 44, 4614. (b) Zheng, S.-L.; Volkov, A.; Nygren, C. L.;
Coppens, P. Chem.sEur. J. 2007, 13, 8583.
(33) Duboc, C.; Phoeung, T.; Zein, S.; Pe´caut, J.; Collomb, M.-N.; Neese,
F. Inorg. Chem. 2007, 46, 4905.
Inorganic Chemistry, Vol. 47, No. 23, 2008 11211