Crystal structures and magnetic and gas-occlusion properties of microporous materials containing infinite chains of mononuclear metal (Cu(II), Zn(II), and Ni(II)) dicarboxylates unit

Article abstract of DOI:10.1246/bcsj.76.1387

A new series of microporous materials containing infinite chains of a mononuclear metal (Cu(II), Zn(II), Ni(II))-pyridine unit bridged by dicarboxylates (copper(II) terephthalate-pyridine 1a, copper(II) terephthalate-4-phenylpyridine 1b, zinc(II) fumarate-pyridine 2a, zinc(II) terephthalate-pyridine 2b, nickel(II) fumarate-pyridine 3a, and nickel(II) terephthalate-pyridine 3b) have been prepared. These complexes, which have a regular one-dimensional structure, occlude large amounts of gases. The maximum amount of N₂ gas is 0.7-10.7 mole per mole of metal(II) salt, indicating the presence of a large number of micropores. A porous structure, which is formed by stacking and self-assembly of the linear metal(II) dicarboxylates, was determined by X-ray crystallography. We have elucidated the paramagnetic properties of 1a, 1b, 3a, and 3b by magnetic measurement (SQUID), whereas 2a and 2b were found to be diamagnetic. The magnetic susceptibilities of 1a and 1b obey the Curie-Weiss law over the range of 70-300 K (θ = -1. 4 K and θ = +5.5 K, respectively); the obtained Weiss constants (θ) indicate the existence of small antiferromagnetic interactions (1a) and ferromagnetic interactions (1b). The different magnetic behaviors between 1a and 1b demonstrate that hydrogen bonding between the carboxylate groups and coordinated water molecules plays an important role in determining the bridge geometries and super-exchange interaction between the Cu(II) ions through the Cu-O-C-O-H-O-Cu pathways. In contrast to those of the copper(II) complexes, the magnetic susceptibilities of 3a and 3b obey the Curie law.

Full text of DOI:10.1246/bcsj.76.1387

ꢀ 2003 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 76, 1387–1395 (2003) 1387
Crystal Structures and Magnetic and Gas-Occlusion Properties of
Microporous Materials Containing Infinite Chains of Mononuclear
Metal (Cu(b), Zn(b), and Ni(b)) Dicarboxylates Unit
ꢀ
Tetsushi Ohmura, Wasuke Mori, Mari Hasegawa, Tohru Takei, Toshiya Ikeda,
and Eiko Hasegawa
Department of Chemistry, Faculty of Science, Kanagawa University, Hiratsuka, Kanagawa 259-1293
(Received December 20, 2002)
A new series of microporous materials containing infinite chains of a mononuclear metal (Cu(b), Zn(b), Ni(b))-
pyridine unit bridged by dicarboxylates (copper(b) terephthalate-pyridine 1a, copper(b) terephthalate-4-phenylpyridine
1b, zinc(b) fumarate-pyridine 2a, zinc(b) terephthalate-pyridine 2b, nickel(b) fumarate-pyridine 3a, and nickel(b) tere-
phthalate-pyridine 3b) have been prepared. These complexes, which have a regular one-dimensional structure, occlude
large amounts of gases. The maximum amount of N2 gas is 0.7–10.7 mole per mole of metal(b) salt, indicating the
presence of a large number of micropores. A porous structure, which is formed by stacking and self-assembly of the
linear metal(b) dicarboxylates, was determined by X-ray crystallography. We have elucidated the paramagnetic proper-
ties of 1a, 1b, 3a, and 3b by magnetic measurement (SQUID), whereas 2a and 2b were found to be diamagnetic. The
magnetic susceptibilities of 1a and 1b obey the Curie-Weiss law over the range of 70–300 K (ꢀ ¼ ꢁ1:4 K and ꢀ ¼ þ5:5
K, respectively); the obtained Weiss constants (ꢀ) indicate the existence of small antiferromagnetic interactions (1a) and
ferromagnetic interactions (1b). The different magnetic behaviors between 1a and 1b demonstrate that hydrogen
bonding between the carboxylate groups and coordinated water molecules plays an important role in determining the
bridge geometries and super-exchange interaction between the Cu(b) ions through the Cu–O–C–O–H–O–Cu pathways.
In contrast to those of the copper(b) complexes, the magnetic susceptibilities of 3a and 3b obey the Curie law.
The mineral zeolite was reported by Cronstedt in 1756 and
Damour in 1840.1 In the 20th century, numerous porous com-
pounds (either natural or artificial) became recognized. Zeo-
lite, porous silica, clay, activated carbon, carbon fabric, etc.,
are well-known classical porous materials.2 They are classi-
fied by their components (organic and inorganic), the degree of
crystallization, pore size, and function.
into two types, based on their crystal structures, as represented
schematically by A and B. A preliminary report on the micro-
porous structures, gas-occlusion properties, and magnetic
properties of 1a with the type-A structure has appeared.10 In
this paper, we report on the synthesis of novel adsorbent co-
ordination polymers of mononuclear metal(b) dicarboxylates
and their magnetic and gas occlusion properties.
In recent years, metal-organic frameworks have received
considerable attention as new and attractive materials in gas
storage, ion exchange, and as potential catalysts. Previously,
we reported that copper(b) dicarboxylates,3 molybdenum(b)
dicarboxylates,4 and ruthenium(b,c) dicarboxylates5 reversi-
bly occluded large amounts of gases, such as N2, Ar, O2,
CH4, and Xe. The uniform linear micropores of these adsor-
bents were assumed as having been constructed by the stacking
or bonding of two-dimensional lattices of dinuclear transition-
metal dicarboxylates. Many excellent studies of adsorbent
porous complexes have also been reported by Yaghi,6
Kitagawa,7 Williams,8 and Fujita.9
Recently, it was found that a microporous dicarboxylate
polymer, a linear-chain complex of mononuclear copper(b)
terephthalate-pyridine 1a, was followed by a similar system,
copper(b) terephthalate-4-phenylpyridine 1b, zinc(b) fuma-
rate-pyridine 2a, zinc(b) terephthalate-pyridine 2b, nickel(b)
fumarate-pyridine 3a, and nickel(b) terephthalate-pyridine 3b
(Chart 1). When guest solvents were released, these com-
plexes, except for 3a, occluded large amounts of gases, such
as N2 and Ar. The complexes described here can be classified
Experimental
Synthesis of 1a. Copper(b) terephthalate-pyridine was syn-
thesized by using a carboxylate-exchange procedure, which we
reported previously.10 Deep-blue, plate-like crystals precipitated
out of the solution. [Cu(b)(ꢁ-O2CC6H4CO2)(C5H5N)2(H2O)]
C5H5N H2O 1a. Anal. Found: C, 53.32; H, 3.82; N, 6.54%.
Calcd for C18H16CuN3O6: C, 55.14; H, 4.63; N, 8.39%. TG:
16.4% initial weight loss between 23 and 64 ꢃC (16.4% Calcd
for C5H5N).
Synthesis of 1b. Copper(b) terephthalate-4-phenylpyridine
was synthesized by using a carboxylate-exchange procedure.
To a methanol solution (153 mL) of 4-phenylpyridine (155 mg,
0.222 mmol) was added an aqueous solution (10.0 mL) of
copper(b) formate tetrahydrate (50.0 mg, 0.222 mmol), an aque-
ous solution (10.0 mL) of terephthalic acid disodium salt (46.6
mg, 0.222 mmol), and formic acid (18.6 mL). Slow evaporation
of the pale-blue eluate gave deep blue, plate-like crystals.
ꢂ
ꢂ
[Cu(b)(ꢁ-O2CC6H4CO2)(C11H9N)2(H2O)] 2MeOH 1b. Anal.
Found: C, 62.43; H, 4.98; N, 4.54%. Calcd for C32H32CuN2O7:
C, 61.98; H, 5.20; N, 4.52%. TG: 9.2% initial weight loss between
ꢂ
Published on the web July 15, 2003; DOI 10.1246/bcsj.76.1387

1388 Bull. Chem. Soc. Jpn., 76, No. 7 (2003)
Gas-Occlusion Properties of Metal Carboxylate
Chart 1.
ꢃ
23 and 116 C (10.3% Calcd for 2MeOH).
at room temperature, after which sky-blue plate-like crystals pre-
cipitated out of the solution. [Ni(b)(ꢁ-O2CC2H2CO2)(C5-
H5N)3] C5H5N 2H2O 3a. Anal. Found: C, 54.53; H, 5.12; N,
10.99%. Calcd for C24H26N4NiO6: C, 54.89; H, 4.99; N, 10.67%.
TG: 22.0% initial weight loss between 23 and 100 ꢃC (21.9%
Calcd for C5H5N).
Synthesis of 3b. Another nickel(b) complex 3b was synthe-
sized in the same manner as 3a, except that terephthalic acid was
used instead of fumaric acid. Sky-blue, plate-like crystals pre-
cipitated out of the solution. [Ni(b)(ꢁ-O2CC6H4CO2)(C5H5-
N)3]6 (C5H5N)2 6DMF 9H2O 3b. Anal. Found: C, 56.62; H,
Synthesis of 2a. A form of zinc(b) fumarate-pyridine was
synthesized as follows. To an aqueous solution (15.0 mL) of
zinc(b) nitrate hexahydrate (200 mg, 0.673 mmol) was added a
pyridine solution (50.0 mL) of fumaric acid (116 mg, 1.00 mmol).
The mixture was then allowed to stand for several days at room
temperature, after which colorless, plate-like crystals precipitated
out of the solution. [Zn(b)(ꢁ-O2CC2H2CO2)(C5H5N)3] C5H5-
N 0.5H2O 2a. Anal. Found: C, 57.48; H, 4.62; N, 10.73%. Calcd
for C24H23N4O4:5Zn: C, 57.10; H, 4.59; N, 11.10%. TG: 15.2%
initial weight loss between 23 and 116 ꢃC (16.0% Calcd for
C5H5N).
Synthesis of 2b. A form of zinc(b) terephthalate-pyridine was
synthesized as follows. To an aqueous solution (15.0 mL) of
zinc(b) acetate dihydrate (147 mg, 0.673 mmol) was added a pyr-
idine solution (40.0 mL) of terephthalic acid (166 mg, 1.00
mmol). The mixture was then allowed to stand for several days
at room temperature, after which colorless plate-like crystals pre-
cipitated out of the solution. [Zn(b)(ꢁ-O2CC6H4CO2)(C5H5N)2-
(H2O)2] (C5H5N)2 2H2O 2b. Anal. Found: C, 54.40; H, 5.57; N,
8.95%. Calcd for C28H32N4O8Zn: C, 54.42; H, 5.22; N, 9.07%.
TG: 31.3% initial weight loss between 23 and 83 ꢃC (31.4% Calcd
for (C5H5N)2 2H2O).
Synthesis of 3a. A form of nickel(b) fumarate-pyridine was
synthesized as follows. To a pyridine solution (30.0 mL) of fu-
maric acid (23.3 mg, 0.201 mmol) was added a DMF solution
(20.0 mL) of nickel(b) acetate tetrahydrate (50.0 mg, 0.201
mmol). The mixture was then allowed to stand for several days
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
5.17; N, 10.58%. Calcd for C166H184N26Ni6O39: C, 56.65; H,
5.27; N, 10.35%. TG: 22.4% initial weight loss between 23 and
ꢃ
105 C (21.7% Calcd for (C5H5N)2 6DMF 9H2O).
ꢂ
ꢂ
Physical and Chemical Measurements. Carbon, hydrogen,
and nitrogen analyses were conducted using a Perkin-Elmer PE
2400 seriesbCHNS/O analyzer. Infrared (IR) spectra were mea-
sured with a Jasco. Co., FT/IR-300 spectrophotometer in the re-
gion of 4000–400 cmꢁ1 using KBr discs. Thermogravimetric/
differential thermal analysis was conducted using TG8101D and
TAS300 with an aluminum pan in an ordinary atmosphere. Sam-
ples were heated between room temperature and 450 ꢃC at 1–4 ꢃC/
min. The magnetic susceptibilities at room temperature were
measured by the Gouy method (22000 Oe). The temperature
dependence of the magnetic susceptibilities was measured by a
SQUID magnetometer (Quantum Design, MPMS-5S) in a tem-
perature range of 2–300 K (10000 Oe). The susceptibilities were
corrected for the diamagnetism of the constituent atoms by using
ꢂ
ꢂ
ꢂ

T. Ohmura et al.
Bull. Chem. Soc. Jpn., 76, No. 7 (2003) 1389
Pascal’s constants.11 The effective magnetic moments (ꢁeff) were
calculated from the equation, ꢁeff ¼ 2:83ðꢂATÞ1=2, where ꢂA is
the atomic magnetic susceptibility.
was measured. After the buoyancy had been corrected to the
obtained amount of the weight change, the absorbed amount of
the sample was calculated.
X-Ray Crystal Structure Determinations. The crystal struc-
tures of 1a, 1b, and 2a were determined by X-ray diffraction using
a Rigaku R-AXIS RAPID diffractometer with graphite-monochro-
mated Mo Kꢃ radiation. Temperature control was achieved with
a temperature control unit. X-ray diffraction data of 2b were
collected on a Rigaku R-AXIS RAPID diffractometer by using
graphite-monochromated Cu Kꢃ radiation. The structures were
solved by a direct method using the program SIR-92 and were
refined by full-matrix least-squares iterations. Anisotropic ther-
mal parameters were assigned to the Cu, Zn, C, O, and N atoms.
The H atoms, except for those of water molecules, were placed at
the calculated positions. No disorder was observed in any of the
structures.
Adsorption isotherms were measured using an ASAP 2010 vol-
umetric adsorption apparatus provided by Micromeritics Co., Ltd.
Before adsorption measurements, the samples were degassed un-
der a vacuum at 273 K. Adsorption isotherms were measured in
the relative pressure (p=p0, p0 = saturation pressure) range from
10ꢁ6 to 0.65.
Results and Discussion
Structure Description. Crystallographic data and other
pertinent information are summarized in Table 1. For each
structure, we present an ORTEP drawing of the unit containing
the copper, zinc, and nickel atoms and all attached atoms, in
Figs. 1–6. In each figure, the atomic numbering scheme for the
species is defined.
The microporous structures, each of which has a large num-
ber of micropores, are shown in Fig. 7. The results suggest
that these complexes, except for 3a, form capillary structures.
In all of the complexes, the van der Waals interaction between
the pyridine rings of the ligands, the so-called p-p stack, may
The crystal structures of 3a and 3b were determined by X-ray
diffraction using a Bruker APEX diffractometer with graphite-
ꢀ
monochromated Mo Kꢃ radiation (ꢄ ¼ 0:7107 A). The structures
were solved by a direct method using the program SHELXS-97,
and were refined by full-matrix least-squares iterations. Aniso-
tropic thermal parameters were assigned to the Ni, C, O, and N
atoms. The H atoms, except for those of water molecules, were
placed at the calculated positions. There is disorder at the C, N, O
atoms of the solvent molecules (DMF) in the structure of 3b,
which are divided into two positions with the same weight, re-
spectively.
X-ray powder diffraction data were collected on a Mac Science
M18XHF22-SRA target Cu diffractometer by using Cu Kꢃ
radiation. The simulated powder patterns were obtained by using
a Cerius2 DIFFRACTION module.
Gas-Adsorption Measurement. The temperature dependence
of the amounts of occluded N2 gas for all complexes were deter-
mined by a Cahn 1000 electric balance at 20 torr, where 1 torr =
(101325/760) Pa, in a temperature range of 77–220 K. After each
sample was set in the apparatus, the guest solvents in the pore and
on the surface were removed by a treatment in a vacuum. After
confirming that a constant weight was reached, nitrogen was
dosed into the adsorption chamber. After the equilibrium of ad-
sorption was established, the change in the weight of the sample
Fig. 1. ORTEP view of 1a showing the numbering scheme.
For clarity, the hydrogen atoms and the solvent molecules
(pyridine) have been omitted.
Table 1. Crystallographic Data Collection and Structure Refinement Parameters for 1a, 1b, 2a, 2b, 3a, and 3b
1a
1b
2a
2b
3a
3b
Chemical formula CuC26H21N4O5 CuC30H24N2O5 ZnC25H13N3O4 ZnC28H24N4O4 Ni2C48H44N8O8 Ni6C174H156N24O38
Formula mass/Da 533.02
Space group
556.08
C2=c
484.78
P21=n
545.9
P21=n
978.33
P1
3543.49
P21212
16.1359(11)
57.178(4)
19.4167(13)
90.000
90.000
90.000
17914(2)
1.314
Cc
ꢀ
a/A
17.8636(9)
5.9305(3)
22.001(2)
90.000
114.239(2)
90.000
2125.3(2)
1.666
296.0
0.71069
0.055
0.122b)
4.0
10.79
24.908(3)
5.8332(5)
22.075(2)
90.000
105.628(3)
90.000
3088.8(5)
1.196
123.0
0.71069
0.074
0.124b)
4.0
7.44
9.9050(8)
17.877(1)
13.427(1)
90.000
91.309(2)
90.000
2376.9(3)
1.693
203.0
0.71069
0.049
0.057b)
4.0
13.35
12.368(5)
15.382(5)
23.340(8)
90.000
91.41(2)
90.000
4439.0(2)
1.225
103.0
1.5419
0.039
0.051b)
6.0
14.508
9.1818(8)
9.7958(9)
12.8848(12)
90.244(2)
90.017(2)
90.381(2)
1158.86(18)
1.402
90.0
0.71073
0.0336
0.0835c)
1.0
8.75
ꢀ
b/A
ꢀ
c/A
ꢃ/deg
ꢅ/deg
ꢆ/deg
ꢀ 3
V/A
ꢇ
T/K
calcd/g cmꢁ3
90.0
ꢀ
ꢄ/A
0.71073
0.0906
0.2588c)
4.0
Ra)
Rw
b)
Z
ꢁ/cmꢁ1
6.97
2
2
2 1=2
a) R ¼ ꢁjjF0j ꢁ jFCjj=ꢁjF0j. b) Rw ¼ ½ꢁwðjF0j ꢁ jFCjÞ =ꢁwF02ꢄ1=2. c) Rw ¼ ½ꢁðwðF02 ꢁ FC2 Þ Þ=ꢁðF02Þ ꢄ
.

1390 Bull. Chem. Soc. Jpn., 76, No. 7 (2003)
Gas-Occlusion Properties of Metal Carboxylate
Fig. 5. ORTEP view of 3a showing the numbering scheme.
For clarity, the hydrogen atoms and the solvent molecules
(pyridine) have been omitted.
Fig. 2. ORTEP view of 1b showing the numbering scheme.
For clarity, the hydrogen atoms and the solvent molecules
(methanol) have been omitted.
act as a driving force for the self-assembly of the one-dimen-
sional metal(b) dicarboxylates to form two-dimensional lattice
structures.
The hydrogen bonding of 1a and 1b with the type-A struc-
ture are clearly characterized by X-ray crystallography; hydro-
gen bonding occurs between the two-dimensional layers,
through the O atoms of the carboxylate groups and the coor-
dinated water molecules, increasing the structural dimension-
ality of 1a and 1b (Fig. 8).
The crystal structures of 2a, 2b, 3a, and 3b with the type-B
structure contain no coordinated water molecules, and thus no
water hydrogen bonding. These structures are constructed by
the stacking of two-dimensional lattices (intermolecular inter-
action), and form very similar capillary structures to that of the
type-A (Fig. 9).
In the process of crystallization, the micropores are occu-
pied by molecules of the solvent, as shown in Fig. 10. The
guest solvent, such as pyridine and/or water molecules, were
removed from the micropores by a treatment in a vacuum. No
structural decomposition occurred in the removal of crystal
solvents in a vacuum, as illustrated by sharp peaks observed
in the X-ray powder diffraction patterns, which is in a good
agreement with that of the simulated structures, as shown in
Fig. 11. This result suggests no deformation of the crystal
frameworks.
Fig. 3. ORTEP view of 2a showing the numbering scheme.
For clarity, the hydrogen atoms and the solvent molecules
(pyridine) have been omitted.
Stability of Complexes. All of the synthesized complexes
were hygroscopic and relatively sensitive to moisture. Ana-
lytical data for all of the complexes are in good agreement with
a composition having few water molecules.
A TG-DTA analysis showed that the guest solvent was
eliminated from the network as the temperature rose
(Fig. 12). When the guest solvents were released, the deep-
blue 1a and 1b changed to the pale-blue 1a0 and 1b0, capable
of occluding gases.12 These networks with the type-A struc-
ture were stable to the initial weight loss of the guest solvent,
whereas in the loss of coordinated molecules (water forming
hydrogen bonding and pyridine), these complexes exhibited no
gas-occlusion properties. On the other hand, when the guest
solvents were released, the colorless 2a and 2b changed to
white 2a0 and 2b0, capable of occluding gases. Similarly,
the sky-blue 3a and 3b changed to pale-sky-blue 3a0 and
Fig. 4. ORTEP view of 2b showing the numbering scheme.
For clarity, the hydrogen atoms and the solvent molecules
(pyridine and water) have been omitted.

T. Ohmura et al.
Bull. Chem. Soc. Jpn., 76, No. 7 (2003) 1391
Fig. 7. View of crystal structure of 1b. For clarity, the
hydrogen atoms and the solvent molecules (methanol)
have been omitted.
Fig. 8. Hydrogen bonding structure of 1b. The water hy-
drogen bonding between the H atoms (H1 and H2) of the
coordinated water molecules and the O atoms (O3 and
O30) of the carboxylate groups also form a three-dimen-
sional network structure. For clarity, the hydrogen atoms
and the solvent molecules (methanol) have been omitted.
Fig. 6. ORTEP view of 3b showing the numbering scheme.
For clarity, the hydrogen atoms and the solvent molecules
(pyridine, DMF, and water) have been omitted.
3b0. Complex 3b0 exhibited gas-occlusion properties. These
networks with the type-B structure are stable to the initial
weight loss of the guest solvent, whereas in the loss of coor-
dinated molecules (pyridine and/or water) these complexes
exhibited no gas-occlusion properties.
IR Spectra. The IR spectra of all the complexes exhibited
very strong-intensity bands in the 1200 and 1700 cmꢁ1
regions. Selected IR spectral data are summarized in
Table 2. The bands for 3b at 1591 and 1381 cmꢁ1 are assigned
to the ꢈasym(CO2) and ꢈsym(CO2) stretching modes, respective-
ly, of the bound monodentate end of the terephthalate ligand,
as shown in Fig. 6. The 1568 cmꢁ1 band is assigned to the
Fig. 9. View of crystal structure of 2b. For clarity, the
hydrogen atoms and the solvent molecules (pyridine and
H2O) have been omitted.
ꢈ
asym(CO2) stretching mode of the bidentate portion of the

1392 Bull. Chem. Soc. Jpn., 76, No. 7 (2003)
Gas-Occlusion Properties of Metal Carboxylate
Fig. 10. View of crystal structure of 1a. The micropores
are occupied by molecules of the solvent (pyridine). For
clarity, the hydrogen atoms have been omitted.
Fig. 12. TG-DTA curves of 1a. Sample of 11.328 mg,
heating in an aluminum paꢃn at 1.0 ꢃC/min. An initial
weight loss of 16.4% at 64 C for evacuation of the pyr-
idine guests from the micropores, with one pyridine re-
moved per formula unit (16.4% calculated), followed by
second weight loss of 4.8% at 89 ꢃC, corresponding to the
removal of one coordinated water forming hydrogen bond-
ing per formula unit (3.7% calculated).
Table 2. Selected IR Spectral Data (cmꢁ1) of 1a, 1b, 2a, 2b,
3a, and 3b
Complex
ꢈ
asym(CO2)
ꢈ
sym(CO2)
ꢉðꢈasym ꢁ ꢈsym
Þ
1a
1b
2a
2b
3a
3b
1599
1593
1554
1574
1551
1361
1362
1385
1394
1389
1381
238
231
169
180
162
210, 187
1591, 1568
Table 3. Selected Bond Angles (deg) of 1a, 1b, 2a, 2b, 3a,
and 3b
1a
3a
ꢀ
O(2)–C(11)–O(4)H bonds 122(1) O(1)–C(49)–O(2)
O(3)–C(13)–O(5)H bonds 128(1) O(5)–C(43)–O(6)
O(7)–C(47)–O(8)
120.3(8)
122.9(8)
121.4(8)
1b
Fig. 11. Simulated (top) and observed (botton) X-ray pow-
der diffractions patterns of 3b0 at room temperature in a N2
gas flow.
O(2)–C(1)–O(3)H bonds 125.3(7) 3b
O(1)–C(1)–O(2)
128.0(8)
121.0(7)
121.0(8)
124.7(8)
123.1(8)
ꢀ
ꢀ
2a
O(3)–C(17)–O(4)
124.0(5) O(5)–C(53)–O(6)
124.5(5) O(7)–C(15)–O(8)
O(9)–C(7)–O(10)
O(2)–C(16)–O(6)
O(1)–C(18)–O(3)
terephthalate ligand, with the ꢈsym(CO2) mode overlapping
that of the monodentate end of the terephthalate ligand. The
splitting between the bands of the bound monodentate end is
ꢉ ¼ 210 cmꢁ1, larger than that for the bidentate end (ꢉ ¼ 187
cmꢁ1). This is consistent with the larger carboxylate O–C–O
bond angle of the terephthalate ligand for the monodentate
end, as compared to the carboxylate end bidentate to Cu(b),
Zn(b), and Ni(b), as shown in Tables 2 and 3. The IR spectrum
of 1a and 1b with the type-A structure displays ꢉ ¼ 238 and
231 cmꢁ1, respectively, indicative of a monodentate binding
mode where H-bonding occurs with the unbound CO group.13
Gas-Occlusion Properties. The occlusion of N2 gas by
1a0 occurs at temperatures below 180 K, with the amount of
occluded N2 gas nearly reaching saturation at 130 K.10 The
maximum amount of occluded N2 gas for 1a0 was 1.1 mole per
ꢀ
2b
O(11)–C(13)–O(12) 119.0(8)
ꢀ
O(8)–C(24)–O(9)
O(4)–C(13)–O(5)
O(2)–C(6)–O(3)
126.3(6) O(13)–C(73)–O(14) 121.3(9)
126.0(5) O(15)–C(3)–O(16)
128.7(8)
126.1(6) O(17)–C(20)–O(18) 129.2(8)
ꢀ
O(21)–C(56)–O(22) 120.3(8)
O(19)–C(66)–O(20) 122.1(8)
ꢀ
O(23)–C(2)–O(24)
122.9(8)
ꢀBond bidentate end of the terephthalate ligand.
mole of copper(b) salt. Remarkably, 1b0, 2a0, 2b0, and 3b0,
with the exception of 3a0, displayed the ability to occlude large
amounts of gases. X-ray crystallographic analyses revealed
that 3a had “zero-dimensional micropores”, as shown in

T. Ohmura et al.
Bull. Chem. Soc. Jpn., 76, No. 7 (2003) 1393
Fig. 13. Crystal structure of 3a. The probability distribu-
tion of loaded gas molecules is pictured as a cloud of dots.
For clarity, the hydrogen atoms and the solvent molecules
(pyridine) have been omitted.
Fig. 15. Temperature dependence of the effective magnetic
moments and magnetic susceptibilities for 1a( ) and
1b( ).
ceptibilities of 1a, 1b, 3a, and 3b were measured. In the type-
A complexes, in which one copper(b) ion is linked by one
dicarboxylate and one coordinated water molecule, two-path-
ways are provided for a superexchange interaction between
copper(b) ions. Fig. 15 shows the temperature dependence
of the magnetic susceptibilities and effective magnetic
moments. The magnetic susceptibilities of 1a and 1b with
the type-A structure obey the Curie-Weiss law over the range
of 70–300 K (ꢀ ¼ ꢁ1:4 K, and ꢀ ¼ þ5:5 K, respectively); the
obtained Weiss constants (ꢀ) indicate the existence of small
antiferromagnetic interactions for 1a, and ferromagnetic inter-
actions for 1b. The relatively high value of ꢀ for 1b suggests
significant ferromagnetic coupling arising from the hydrogen
bonding. Also, similar values (ꢀ ¼ þ9:0 K) have been report-
ed for ferromagnetic interactions through the hydrogen bond-
ed pathways in [Cu(b)(ꢁ-O2CC6H5)2(C5H5N)2(H2O)].15
The different magnetic behaviors between 1a and 1b demon-
strate that the hydrogen bonding between the bridging carbox-
ylate groups and the coordinated water molecules plays an
important role in determining the bridge geometries and su-
per-exchange interaction between the Cu(b) ions through the
Cu–O–C–O–H–O–Cu pathways. The dihedral angles between
the [Cu, O2, O4] and [Cu, O3, O5] planes and between the
[Cu, O2, O3] and [Cu, O20, O30] planes are 3.4ꢃ and 38.4ꢃ in
1a and 1b, respectively. Conversely, it might be suggested
that no magnetic interactions exist between the Cu(b) ions
through the bridges of terephthalate ligands. The dihedral
angles between the [O2(O3), N1, N2] and [C12(C14),
C15(C16), C17(C18)] planes and between the [O2, N3, N30]
and [C2(C20), C30(C3), C4(C40)] planes are 69.9–75.5ꢃ and
70.6ꢃ in 1a and 1b, respectively. These results suggest that
1a and 1b are not affected by magnetic interactions arising
from a large distortion of these dihedral angles (69.9–75.5ꢃ
and 70.6ꢃ). Consequently, the net ferromagnetic interaction
observed for 1b is due to a weaker antiferromagnetic contribu-
tion derived from a large distortion of the Cu–O–C–O–H–O–
Cu super-exchange pathway.
Fig. 14. Adsorption isotherm of 2a0 obtained with an argon
gas in the relative pressure range from 10ꢁ6 to 0.65 at
temperature of liquid argon. Adsorption isotherm of 1a0
(type I) is shown in the inset.
Fig. 13, and thus no gas-occlusion properties. Also, the
amounts of occluded gas increase with a decrease in the tem-
perature, and finally reached a maximum point at the tempera-
ture of liquid nitrogen. The maximum amounts of occluded
N2 gas for 1b0, 2a0, 2b0, and 3b0 were 4.8, 10.6, 2.0, and 0.7
mole per mole of metal(b) atoms, respectively. The crystal
structures and gas-occlusion properties suggest that 1b0, 2b0,
and 3b0 form very similar capillary structures to that of 1a0.
For 2a0, an argon adsorption experiment was carried out in
the relative pressure range from 10ꢁ6 to 0.7 (Fig. 14). This
complex shows an isotherm of type V according the BDDT
classification.14 The volume adsorbed increased linearly with
an increase in the relative pressure, and exhibited a very high
adsorbed volume, thus, indicating a large gas-occlusion ca-
pability.
Magnetic Properties. To investigate the structure in fur-
ther detail, the temperature dependences of the magnetic sus-
Fig. 16 shows the temperature dependence of the magnetic
susceptibilities and effective magnetic moments for 3a and 3b.

1394 Bull. Chem. Soc. Jpn., 76, No. 7 (2003)
Gas-Occlusion Properties of Metal Carboxylate
(SQUID), whereas 2a and 2b were found to be diamagnetic.
It is expected that isomorphous complexes will be synthe-
sized using chromium, tungsten, ruthenium, rhenium, rhodium,
etc. instead of copper, zinc, and nickel. Similar systems may
also be designed and constructed using other sets of ligands
and transition metals. Furthermore, from the perspective of
molecular architecture, this strategy advances functional ma-
terial science based upon complex chemistry and the nature of
transition metals.
We thank Nihon Bruker AXS K.K. for its support. We also
acknowledge Dr. Kenji Yoza for his assistance in the X-ray
crystal examination of 3a and 3b.
This work was supported by a grant-in-aid from the New
Energy and Industrial Technology Development Organization
(NEDO).
Fig. 16. Temperature dependence of the effective magnetic
moments and magnetic susceptibilities for 3a( ) and
3b( ).
References
1
a) A. F. Constedt, Akad. Hankle. Stockholm, 18, 120
(1756). b) A. Damour, Ann. Mines, 17, 191 (1840).
a) R. K. Iler, “The Chemistry of Silica,” Wiley, New York
(1979); R. Szostack, “Molecular Sieves,” 2nd ed., Blackie Aca-
demic and Professional, London (1998). b) L. L. Henchi and D. R.
Ulrich, “Ultrastructure Processing of Ceramics, Glasses, and
Composites,” Wiley, New York (1984). c) R. E. Grim, “Applied
Clay Mineralogy,” McGraw-Hill, New York (1962). d) R. F. Lobo
et al. (Eds.), “Microporous and Macroporous Materials,” Materi-
als Research Society, Pittsburgh (1996).
In contrast to the copper(b) complexes (1a and 1b), the mag-
netic susceptibilities of 3a and 3b with the type-B structure
obey the Curie law. Therefore, the magnetic behaviors of 3a
and 3b with no hydrogen bonding are barely affected by the
magnetic interaction between the Ni(b) ions.
2
Conclusion
In the present work, we succeeded to synthesize novel ad-
sorbent coordination polymers of mononuclear metal(b) dicar-
boxylates: copper(b) terephthalate-pyridine 1a, copper(b) tere-
phthalate-4-phenylpyridine 1b, zinc(b) fumarate-pyridine 2a,
zinc(b) terephthalate-pyridine 2b, nickel(b) fumarate-pyridine
3a, and nickel(b) terephthalate-pyridine 3b. When guest sol-
vents were released, these complexes (1a0, 1b0, 2a0, 2b0, and
3b0) exhibited a gas-occlusion capability similar to that of
copper(b) dicarboxylates, as reported previously.3 The max-
imum amounts of occluded N2 gas for 1a0, 1b0, 2a0, 2b0, and
3b0 were 1.1, 4.8, 10.6, 2.0, and 0.7 mole per mole of the
metal(b) atoms, respectively. Uniform capillaries constructed
by the stacking and self-assembly of the linear metal(b) dicar-
boxylates were determined by X-ray crystallography. Large
amounts of gas may be occluded in the capillaries (in the case
2a0 may be also occluded in other site), as shown in Figs. 7, 9–
11, 14. In comparison with industrially important adsorbents,
such as the well-known zeolite, silica, and activated carbon,16
the system discovered in the present work is the simplest and
most advantageous adsorbent for studying pure and applied
chemistry. This statement is true in the following respects:
(1) The amounts of occluded gas are very large. (2) No dis-
turbances, such as counterions, are required. (3) The homo-
geneity of the micropores, which are most probably linear
capillaries, and the adsorbed molecules would be distributed
as a column. (4) Simplicity to obtain single crystals for mi-
croporous polymer of mononuclear metal(b) dicarboxylates.
(5) It is possible to synthesize novel adsorbent coordination
polymers of mononuclear Ni(b) dicarboxylates, but not dinu-
clear Ni(b) dicarboxylates.
3
W. Mori, F. Inoue, K. Yoshida, H. Nakayama, S.
Takamizawa, and M. Kishita, Chem. Lett., 1997, 1219.
S. Takamizawa, W. Mori, M. Furihata, S. Takeda, and K.
Yamaguchi, Inorg. Chem. Acta, 283, 268 (1998).
a) S. Takamizawa, K. Yamaguchi, and W. Mori, Inorg.
4
5
Chem. Commun., 1, 177 (1998). b) S. Takamizawa, T. Ohmura,
K. Yamaguchi, and W. Mori, Mol. Cryst. Liq. Cryst., 342, 199
(2000).
6
M. Eddaoudi, J. Kim, N. Rosi, D. Vodak, J. Wachter, M.
O’Keeffe, and O. M. Yaghi, Science, 295, 18 (2002).
a) S. Noro, R. Kitaura, M. Kondo, S. Kitagawa, T. Ishii, H.
7
Matsuzaka, and M. Yamashita, J. Am. Chem. Soc., 124, 2568
(2002). b) R. Kitaura, S. Kitagawa, Y. Kubota, T. C. Kobayashi,
K. Kindo, Y. Mita, A. Matsuo, M. Kobayashi, H.-C. Chang, T. C.
Ozawa, M. Suzuki, M. Sakata, and M. Takata, Science, 298, 2358
(2002).
8
S.-Y. Chui, S. M.-F. Lo, S. S.-Y. Chui, L.-Y. Shek, Z. Lin,
X. X. Zhang, G.-H. Wen, and I. D. Williams, J. Am. Chem. Soc.,
122, 6293 (2000).
9
10 T. Ohmura, W. Mori, M. Hasegawa, T. Takei, and A.
Yoshizawa, Chem. Lett., 32, 34 (2003).
`
11 W. R. Angus, J. Favede, J. Hoarau, and A. Pacault,
“LANDOLT-BORNSTEIN,” Eigenschaften Der Materie in Ihren
K. Biradha and M. Fujita, Chem. Commun., 2002, 1866.
¨
Aggregatzustanden, 10. Teil, Magnetische Eigenschaften b, Ber-
¨
lin, Heidelberg, New York, Springer-Verlag (1967).
12 S. Noro, R. Kitaura, M. Kondo, S. Kitagawa, T. Ishii, H.
Matsuzaka, and M. Yamashita, J. Am. Chem. Soc., 124, 2568
(2002).
13 G. B. Deacon and R. J. Phillips, Coord. Chem. Rev., 33,
227 (1980).
14 S. Brunauer, L. S. Deming, W. E. Deming, and E. Teller, J.
We have demonstrated the paramagnetic properties of 1a,
1b, 3a, and 3b, as elucidated by magnetic measurements

T. Ohmura et al.
Bull. Chem. Soc. Jpn., 76, No. 7 (2003) 1395
Am. Chem. Soc., 62, 1723 (1940).
15 R. E. Del Sesto, A. M. Arif, and J. S. Miller, Inorg. Chem.,
39, 4894 (2000).
16 R. E. Lobo, J. S. Beck, S. L. Snib, D. R. Corbin, M. E.
Davis, L. E. Iton, and S. I. Zones (eds.), “Microporous and
Macroporous Materials,” Materials Research Society Symp. Proc.,
Vol. 431, Materials Research Society, Pittsburgh, PA (1996), and
refs. therein; F. Rodriguez-Reinoso, J. Rouquerol, K. S. W. Sing,
and K. K. Unger (eds.), “Characterization of Porous Solids b,”
Studies in Surface Science and Catalysis, Vol. 62, Elsevier,
Amsterdam (1991), and refs. therein.