Perovskite-like metal formates with weak ferromagnetism and as precursors to amorphous materials

Article abstract of DOI:10.1021/ic0498081

Three isomorphous compounds M(CHOO)₃[NH₂(CH ₃)₂] (M = Mn(1·Mn), Co(2·Co), Ni(3·Ni)) have been synthesized in solvothermal conditions. Single-crystal X-ray diffraction shows that they are all crystallized in the trigonal space group R 3c with small differences in the lattice parameters. Bridged by the three-atom single-bridge CHOO^-, M ions form a three-dimensional distorted perovskite-like structure with dimethylamine (DMA) cations located in the cages of the network. Based on the magnetic data, these three 3D compounds are weak ferromagnets with the critical temperature T _c = 8.5 K (1·Mn), 14.9 K (2·Co), and 35.6 K (3·Ni), and for 2·Co and 3·Ni, spin reorientation might take place at 13.1 and 14.3 K, respectively. At 1.8 K, hysteresis loops can be observed for all three compounds with the coercivity field ca. 90 Oe (1·Mn), 920 Oe (2·Co), and 320 Oe (3·Ni). The canting angles are estimated to be 0.08°, 0.5°, and 0.6° for 1·Mn, 2·Co, and 3·Ni, respectively. The magnetic coupling between Mn^II ions in 1·Mn was estimated based on the model developed by Rushbrook and Wood for a Heisenberg antiferromagnet on a simple cubic lattice and the best fit gives J = -0.23 cm^-1. At the same time, according to molecular field theory of antiferromagnetism, the J values for compounds 1·Mn, 2·Co, and 3·Ni were estimated to be -0.32 cm ^-1, -2.3 cm^-1, and -4.85 cm^-1, respectively. The spin cant in these compounds may originate from the noncentrosymmetric character of the three-atom single-bridge CHOO^-. Furthermore, amorphous materials 4·Mn238, 5·Mn450,6·Co320, and 7·Ni300 were prepared from precursors 1-3 under an argon atmosphere at different temperatures according to the thermogravimetric analyses. As an interesting result, 5·Mn450 was confirmed to be an amorphous form of Mn₃O₄ with a considerably large coercivity field H _C = 4.1 kOe at 30 K compared to that value (250 Oe) for bulk Mn ₃O₄.

Full text of DOI:10.1021/ic0498081

Inorg. Chem. 2004, 43, 4615−4625
Perovskite-like Metal Formates with Weak Ferromagnetism and as
Precursors to Amorphous Materials
Xin-Yi Wang, Lin Gan, Shi-Wei Zhang, and Song Gao*
State Key Laboratory of Rare Earth Materials Chemistry and Applications & PKU-HKU Joint
Laboratory on Rare Earth Materials and Bioinorganic Chemistry, College of Chemistry and
Molecular Engineering, Peking UniVersity, Beijing 100871, P. R. China
Received February 14, 2004
Three isomorphous compounds M(CHOO)3[NH (CH ) ] (M) Mn(1‚Mn), Co(2‚Co), Ni(3‚Ni)) have been synthesized
2
3 2
in solvothermal conditions. Single-crystal X-ray diffraction shows that they are all crystallized in the trigonal space
group R 3hc with small differences in the lattice parameters. Bridged by the three-atom single-bridge CHOO-, M
ions form a three-dimensional distorted perovskite-like structure with dimethylamine (DMA) cations located in the
cages of the network. Based on the magnetic data, these three 3D compounds are weak ferromagnets with the
critical temperature T ) 8.5 K(1‚Mn), 14.9 K (2‚Co), and 35.6 K (3‚Ni), and for 2‚Co and 3‚Ni, spin reorientation
c
might take place at 13.1 and 14.3 K, respectively. At 1.8 K, hysteresis loops can be observed for all three compounds
with the coercivity field ca. 90 Oe (1‚Mn), 920 Oe (2‚Co), and 320 Oe (3‚Ni). The canting angles are estimated
II
to be 0.08°, 0.5°, and 0.6° for 1‚Mn, 2‚Co, and 3‚Ni, respectively. The magnetic coupling between Mn ions in
1‚Mn was estimated based on the model developed by Rushbrook and Wood for a Heisenberg antiferromagnet on
a simple cubic lattice and the best fit gives J ) −0.23 cm-1. At the same time, according to molecular field theory
of antiferromagnetism, the J values for compounds 1‚Mn, 2‚Co, and 3‚Ni were estimated to be −0.32 cm-1, −2.3
-1
-1
cm , and −4.85 cm , respectively. The spin cant in these compounds may originate from the noncentrosymmetric
-
character of the three-atom single-bridge CHOO . Furthermore, amorphous materials 4‚Mn238, 5‚Mn450, 6‚Co320,
and 7‚Ni300 were prepared from precursors 1−3 under an argon atmosphere at different temperatures according
to the thermogravimetric analyses. As an interesting result, 5‚Mn450 was confirmed to be an amorphous form of
Mn3O with a considerably large coercivity field HC ) 4.1 kOe at 30 K compared to that value (250 Oe) for bulk
4
Mn3O .
4
Introduction
effectively mediate the magnetic coupling between the local
-
spin carriers. The bridging ligands including CN-, N3 ,
Molecule-based magnetic polymers have attracted intense
interest in recent years, due not only to the fundamental
research of magnetic interactions and magneto-structural
correlations but also to the development of new functional
molecule-based materials.1 The emphasis on synthesizing
these polymers is to search for the bridging ligands that can
-
C2O42-, N(CN)2 , and carboxyl groups have been widely
investigated, and numerous compounds bridged by these
ligands have been reported. In the context of this research
background, formate bridged compounds have received much
attention.2-10
As the smallest carboxylate, the formate ion has been
observed to display multiple bridging modes (see Scheme 1)
* Author to whom correspondence should be addressed. Tel.: 86-10-
62756320. Fax: 86-10-62751708. E-mail: gaosong@pku.edu.cn.
(1) (a) Niel, V.; Thompson, A. L.; Mun˜oz, M. C.; Galet, A.; Goeta, A.
E.; Real, J. A. Angew. Chem., Int. Ed. 2003, 42, 3760-3763. (b) Kahn,
O.; Martinez, C. J. Science 1998, 279, 44. (c) Gu¨tlich, P.; Garcia, Y.;
Woike, T. Coord. Chem. ReV. 2001, 219-221, 839. (d) Gu¨tlich, P.;
Hauser, A.; Spiering, H. Angew. Chem., Int. Ed. Engl. 1994, 33, 2024.
(e) Sato, O.; Iyoda, T.; Fujishima, A.; Hashimoto, K. Science 1996,
272, 704. (f) Sato, O.; Einaga, Y.; Fujishima, A.; Hashimoto, K. Inorg.
Chem. 1999, 38, 4405. (g) Sato, O.; Hayami, S.; Einaga, Y.; Gu, Z.-
Z. Bull. Chem. Soc. Jpn. 2003, 76, 443.
(2) (a) Wang, Z.-M.; Zhang, B.; Fujiwara, H.; Kobayashi, H.; Kurmoo,
M. Chem. Commun. 2004, 416-417. (b) Cornia, A.; Caneschi, A.;
Dapporto, P.; Faberetti, A. C.; Gatteschi, D.; Malevasi, W.; Sangre-
gorio, C.; Sessoli, R. Angew. Chem., Int. Ed. 1999, 38(12), 1780-
1781.
(3) (a) Viertelhaus, M.; Henke, H.; Anson, C. E.; Powell, A. K. Eur. J.
Inorg. Chem. 2003, 2283-2289. (b) Sapin˜a, F.; Burgos, M.; Escriva´,
E.; Folgado, J. V.; Marcos, D.; Beltra´n, A.; Beltra´n, D. Inorg. Chem.
1993, 32, 4337-4334.
10.1021/ic0498081 CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/26/2004
Inorganic Chemistry, Vol. 43, No. 15, 2004 4615

Wang et al.
Scheme 1. Bridging Mode for µ-CHOO
such as the common 2.11, the frequent 3.12, and the unusual
4.22 mode11 to link two or more transition-metal ions forming
a variety of zero-,7-9 one-,10 two-,4 and three-dimensional2,3,5,6
complexes. Depending on the geometry characters of formate
and the metal ions, formate can adopt different bridging
modes such as syn-syn, anti-anti, syn-anti, and mona-
tomic, they mediate ferro- or antiferromagnetic coupling
between metal ions in different situations.12 Although a
variety of compounds containing formate as the bridging
ligand has been structurally characterized, few reports
involved the magnetic studies on them.2-5 Until now, the
most extensively magnetic investigations have been mainly
concentrated on the simple isomorphous metal formate
complexes M(CHOO)2‚2H2O (M ) Mn, Fe, Co, Ni, Cu)5
and their anhydrous compounds,3 except for few other
formates: M(CHOO)2‚2(NH2)2CO (M ) Mn, Co),4b,c MnIII-
(CHOO)3‚1/2CO2‚1/4HCOOH‚2/3H2O2b (denoted below as
MnIII(CHOO)3), Mn3(CHOO)6‚G (G ) guests),2a and Co-
(4) (a) Rettig, S. J.; Thompson, R. C.; Trotter, J.; Xia, S.-H. Inorg. Chem.
1999, 38, 1360-1363. (b) Kubo, H.; Zenmyo, K.; Matsumura, M.;
Takeda, K.; Alhara, K.; Yamagata, K. J. Phys. Soc. Jpn. 1999, 68(1),
253-257. (c) Fujino, M.; Achiwa, N.; Koyano, N.; Shibuya, I.;
Ridwan; Yamagata, K. J. Magn. Magn. Mater. 1992, 104-107, 851-
852.
(5) (a) Kageyama, H.; Khomskii, D. I.; Levitin, R. Z.; Vasil’ev, A. N.
Phys. ReV. B 2003, 67, 224422. (b) Radhakrishnatt, P.; Gillonf, B.;
Chevrierts, G. J. Phys.: Condens. Matter 1993, 5, 6147. (c) Takeda,
K.; Kawasaki, K. J. Phys. Soc. Jpn. 1971, 31(4), 1026. (d) Burlet, P.;
Burlet, P.; Bertaut, E. F.; Roult, G.; Pillon, J. J. Solid State Commun.
1969, 7, 1403-1408. (e) Yamagata, K. J. Phys. Soc. Jpn. 1967, 22(2),
582-589. (f) Okada, K.; Kay, M. I.; Cromer, D. T.; Almodovar, I. J.
Chem. Phys. 1966, 44(4), 1648. (g) Hoy, G. R.; Barros, S. De S.;
Barros, F. De S.; Friedberg, S. A. J. Appl. Phys. 1965, 36(3), 936. (h)
Wagner, G. R.; Friedberg. S. Appl. Phys. Lett. 1964, 9, 11. (i) Martin,
R. L.; Waterman, H. J. Chem. Soc. 1959, 1359.
4a
(CHOO)2(HCONH2)2 showing long-range ordered states.
Here, we report three new 3D formate bridged compounds
M(CHOO)3[NH2(CH3)2] (M ) Mn(1‚Mn), Co(2‚Co),
Ni(3‚Ni)). They all crystallized as perovskite-like structure
and show spin-canted weak ferromagnetism at low temper-
ature. Their preparations, structures, and magnetic properties
are presented herein, and the magneto-structural correlations
are discussed.
On the other hand, one of the general applications of the
molecule-based materials is their use as precursors for solid-
state materials, by removing the organic ligands upon
chemical or thermal treatment. The reported examples
include the following: decomposition of single-molecule-
magnet Mn12 complexes to obtain oxide,13 decomposition
of nickel formate on sol-gel alumina to get nickel-supported
alumina,14 and conversion of a cyanide material into a metal
oxide by a mild solution route.15 Using the three compounds
1-3 as precursors, we obtained amorphous materials
4‚Mn238, 5‚Mn450, 6‚Co320, and 7‚Ni300 through simple
thermal treatment, and 5‚Mn450 had been chemically
analyzed and magnetically investigated in details.
(6) (a) Kaufman, A.; Afshar, C.; Rossi, M.; Zacharias, D. E.; Glusker. J.
P. Struct. Chem. 1993, 4, 191. (b) Bird, M. J.; Lomer, T. R. Acta
Crystallogr. 1971, B27, 859. (c) Kay, M. I.; Almodovar, I.; Kaplan,
S. F. Acta Crystallogr. 1968, B24, 1312-1316. (d) Okada, K.; Kay,
M. I.; Kromer, D. T.; Almodovar, I. J. Chem. Phys. 1966, 44, 1648.
(e) Strzyz˘ewska, M. B. Acta Crystallogr. 1965, 19, 357-362. (f)
Krogmann, V. K.; Mattes, R. Z. Kristallogr. 1963, 118, S. 291-302.
(g) Barclay, G. A.; Konnard, C. G. L. J. Chem. Soc, 1961, 3289. (h)
Kiriyama, R.; Ibamoto, H.; Matsuo, K. Acta Crystallogr. 1954, 7, 482.
(7) (a) Youngme, S.; Somjitsripunya, W.; Chinnakali, K.; Chantrapromma,
S.; Fun, H.-K. Polyhedron 1999, 18, 857. (b) Norman, R. E.; Leising,
R. A.; Yan, S.-P.; Que, L. Inorg. Chim. Acta 1998, 273, 393-396.
(c) Escriva`, E.; Carrio´, J. S.; Lezama, L.; Folgado, J. V.; Pezarro, J.
L.; Ballesteros, R.; Abarca, B. J. Chem. Soc., Dalton Trans. 1997,
2033-2038. (d) Brooker, S.; Mckee, V.; Metcalfe, T. Inorg. Chim.
Acta 1996, 246, 171. (e) Sapinˇa, F.; Burgos, M.; Escriva´, E.; Folgado,
J. V.; Beltra´n, D. Inorg. Chim. Acta 1994, 216, 185-190. (f) Sessler,
J. L.; Hugdahl, J. D.; Lynch, V.; Davis, B. Inorg. Chem. 1991, 30,
334. (g) Yamanaka, M.; Uekusa, H.; Ohba, S.; Saito, Y.; Iwata, S.
Acta Crystallogr. 1991, B47, 344-355. (h) Armstrong, W. H.; Spool,
A.; Papaefthymiou, G. C.; Frankel, R. B.; Lippard, S. J. J. Am. Chem.
Soc. 1984, 106, 3653. (i) Cotton, F. A.; Rice, G. W. Inorg. Chem.
1978, 17(3), 688-692.
Experimental Section
Materials and Methods. All starting materials were com-
mercially available, reagent grade, and used as purchased without
further purification. Elemental analyses of carbon, hydrogen,
oxygen, and nitrogen were carried out with an Elementar Vario
EL. The microinfrared spectroscopy studies were performed on a
Magna-IR 750 spectrophotometer in the 4000-500 cm-1 region.
(8) (a) Boyle, T. J.; Alam, T. M.; Tafoya, C. J.; Scott, B. L. Inorg. Chem.
1998, 37, 5588. (b) Scott, M. J.; Goddard, C. A.; Holm, R. H. Inorg.
Chem. 1996, 35, 2558-2567.
(9) Cadiou, C.; Coxall, R. A.; Graham, A.; Harrison, A.; Helliwell, M.;
Parsons, S.; Winpenny, R. E. P. Chem. Commun. 2002, 1106-1107.
(10) (a) Sanchis, M. J.; Go´mez-Romero, P.; Folgado, J. V.; Sapin˜az, R.;
Beltra´n, A.; Garc´ıa, J.; Beltra´n, D. Inorg. Chem. 1992, 31, 2915. (b)
Turner, P.; Gunter, M. J.; Hambley, T. W.; White, A. H.; Skelton, B.
W. Inorg. Chem. 1992, 31, 2297-2299. (c) Lis, T.; Trzebiatowska,
B. J. Acta Crystallogr. 1977, B33, 2112-2116.
(11) Harris notation: Harris notation describes the binding mode as
[X.Y1Y2Y3...Yn], where X is the overall number of metals bound by
the whole ligand, and each value of Y refers to the number of metal
atoms attached to the different donor atoms. The ordering of Y is
listed by the Cahn-Ingold-Prelog priority rules.
(12) (a) Colacio, E.; Ghaze, M.; Kiveka¨s, R.; Moreno, J. M. Inorg. Chem.
2000, 39, 2882-2890. (b) Yolanda, R. M.; Catalina, R. P.; Joaqu´ın,
S.; Francesc, L.; Miguel, J. Inorg. Chim. Acta 2001, 318, 159-165.
(c) Pe´rez, C. R.; Sanchiz, J.; Molina, M. H.; Lloret, F.; Julve, M. Inorg.
Chem. 2000, 39, 1363-1370, and references therein.
(13) (a) Larionova, J.; Cle´rac, R.; Boury, B.; Bideau, J. L.; Lecren, L.;
Willemin, S. J. Mater. Chem. 2003, 13, 795-799. (b) Liu, Y.; Liu,
Z.; Wang, G. Appl. Phys. A 2003, 76, 1117-1120.
(14) Kharat, A. N.; Pendleton, P.; Badalvan, A.; Abedini, M.; Amini, M.
M. J. Catal. 2002, 205, 7-15.
(15) Buchelew, A.; Gala´n-Mascaro´s, J. R.; Dunbar, K. R. AdV. Mater. 2002,
14(22), 1646-1648.
4616 Inorganic Chemistry, Vol. 43, No. 15, 2004

PeroWskite-like Metal Formates
Table 1. Elemental Analysis Results and Preparation Conditions of Compounds 4-8a
compound
starting compound
temp, °C
%M
%N
0 (0)
0 (0)
0.18 (0)
0.23 (0)
%C
%H
%O
4 Mn(CHOO)2
5 Mn3O4
6 (CoO‚Co)
7 (3Ni‚NiO‚C)
Mn(CHOO)3[NH2(CH3)2]
Mn(CHOO)3[NH2(CH3)2]
Co(CHOO)3[NH2(CH3)2]
Ni(CHOO)3[NH2(CH3)2]
238
450
320
300
38.7 (37.9)
72.9 (72.0)
88.3 (88.1)
86.4 (89.3)
16.35 (16.57)
1.27 (0)
0.35 (0)
1.85 (1.39)
0.56 (0)
0.61 (0)
0.28 (0)
15.7 (12.0)
5.1 (6.1)
5.85 (6.1)
a Calculated data are in parentheses.
Thermogravimetric analyses (TGA) were performed on a Du Pont
1090B thermal analyzer. Variable-temperature XRD pattern of the
samples were collected with a Rigaku D/Max2000 diffractometer
equipped with a Cu KR radiation source (λ ) 0.15418 nm) from
the temperature range 27 °C to 600 °C. Variable-temperature
magnetic susceptibility, zero-field ac magnetic susceptibility mea-
surements, and field dependence of magnetization were performed
on an Oxford Maglab 2000 System or Quantum Design MPMS
XL7 (SQUID) magnetometer. The experimental susceptibilities
were corrected for the diamagnetism of the constituent atoms
(Pascal’s tables). X-ray Photoelectron Spectroscopy (XPS) was
performed on an Axis Ultra XPS instrument.
Synthesis of Compounds 1-3. Compounds M(CHOO)3[NH2-
(CH3)2] (M ) Mn(1‚Mn), Co(2‚Co), Ni(3‚Ni)) were synthesized
by two different methods both under solvothermal conditions from
corresponding starting chlorides MCl2‚nH2O (M ) Mn, n ) 4; M
) Co, n ) 6; M ) Ni, n ) 6).
Method A. A mixture of MCl2‚nH2O (1 mmol), DMF (6 mL),
and H2O (6 mL) was heated in a Teflon-lined autoclave (20 mL)
at 140 °C for 3 days. After slow cooling to room temperature, some
black/red precipitation was removed from the solution through
filtering. Block crystals (colorless for Mn, pink for Co, and blue
for Ni) suitable for single-crystal X-ray crystallographic analysis
were obtained by evaporating the residual solution at room
temperature for about 1 week. The crystals were filtered from the
mother liquid and washed by ethanol (5 mL × 3). The output of
1‚Mn and 3‚Ni is 80 mg or so, about 30% of the ideal yield based
on the quantity of the chlorides. For 2‚Co, the main product is
another prismlike crystal (confirmed to be Co(CHOO)2‚2H2O) with
a small amount of the desired product 2‚Co. Fortunately, Co-
(CHOO)2‚2H2O can transfer to our desired 2‚Co by slowly
evaporating the mother liquid for a month.
Method B. A mixture of MCl2‚nH2O (1 mmol), NaCHOO‚2H2O
(3 mmol), DMA‚HCl (DMA‚HCl ) dimethylamine hydrochloride,
1 mmol), DMF (8 mL), and H2O (8 mL) was heated in a Teflon-
lined autoclave (20 mL) at 140 °C for 3 days. The succedent process
was similar to that presented for Method A. Crystals grew from
the mother liquid after 3-day evaporation. The output is 180 mg or
so, about 75% of the ideal yield based on the quantity of the
chlorides.
and 3NiO‚Ni‚C (7‚Ni300) were prepared by heating compounds
1-3, respectively, at different temperatures for 30 min under an
argon atmosphere. The yields of 4-7 from compounds 1-3 are
quantitative. Details of the preparation conditions and the elemental
analysis results of compounds 4-7 are listed in Table 1. The IR
spectrum of complex 4-6 are listed as below: 4‚Mn238: 3358m,
3285m, 1646sh, 1586s, 1396m, 1373m, 1358m, 1312w, 756w; 5‚
Mn450: 606s, 490s, 418m; 6‚Co320: 633m, 569s, 484m, 409w.
For 7‚Ni300, the presence of metals and amorphous carbon makes
the micro-IR spectrum a continuous one. The chemical composition
of 7‚Ni300 was deduced from elemental and thermal analyses
results. Detailed thermal analysis results are discussed in the Results
and Discussion section.
Single-Crystal X-ray Crystallography. The data collection of
1, 2, and 3 were made on a Rigak R-Axis RIPID IP with Mo-KR
radiation (λ ) 0.71073 Å) at 293 K. The structures were solved by
direct method and refined by a full matrix least squares technique
based on F2 using the SHELXL 97 program. The hydrogen atoms
+
belonging to NH2(CH3)2 were not added because of the disorder
of these cations in the structure.
Crystal Data for 1‚Mn. Mn(CHOO)3[NH2(CH3)2], C5H11NO6-
Mn, Mw ) 236.09, colorless block (0.4 × 0.4 × 0.2 mm), trigonal,
space group R 3hc, a ) b ) 8.3340(12) Å, c ) 22.895(5) Å, R )
â ) 90°, γ ) 120°, U ) 1377.0(4) Å3, Z ) 6, Dc ) 1.708 g‚cm-3
,
µ ) 1.433 cm-1, F(000) ) 726, GOF ) 0.969. A total of 467
unique reflections were collected in the range 3.34 < θ < 30.45°,
of which 383 were considered observed [I > 2σ(I)] and used in
the calculations. The refinement on F2 converged to R1 ) 0.0240
[I > 2σ(I)], wR2 ) 0.0600 (all data).
Crystal Data for 2‚Co. Co(CHOO)3[NH2(CH3)2], C5H11NO6-
Co, Mw ) 240.08, pink block (0.25 × 0.25 × 0.15 mm), trigonal,
space group R 3hc, a ) b ) 8.1989(12) Å, c ) 22.224(4) Å, R )
â ) 90°, γ ) 120°, U ) 1293.8(4) Å3, Z ) 6, Dc ) 1.849 g‚cm-3
,
µ ) 1.989 cm-1, F(000) ) 738, GOF ) 0.877. A total of 442
unique reflections were collected in the range 4.66 < θ < 30.44°,
of which 337 were considered observed [I > 2σ(I)] and used in
the calculations. The refinement on F2 converged to R1 ) 0.0247
[I > 2σ(I)], wR2 ) 0.0448 (all data).
Crystal Data for 3‚Ni. Ni(CHOO)3[NH2(CH3)2], C5H11NO6Ni,
Mw ) 239.86, blue block (0.35 × 0.35 × 0.25 mm), trigonal, space
group R 3hc, a ) b ) 8.1989(12) Å, 8.1989(12) Å, c ) 22.224(4)
Å, R ) â ) 90°, γ ) 120°, U ) 1293.8(4) Å3, Z ) 6, Dc ) 1.847
g‚cm-3, µ ) 2.249 cm-1, F(000) ) 744, GOF ) 1.042. A total of
434 unique reflections were collected in the range 3.41 < θ <
30.24°, of which 372 were considered observed [I > 2σ(I)] and
used in the calculations. The refinement on F2 converged to R1 )
0.0199 [I > 2σ(I)], wR2 ) 0.0482 (all data).
The elemental analysis results are as follows (%, calculated data
in parentheses). C5H11NO6Mn (1‚Mn): N 5.80 (5.93), C 25.41
(25.44), H 4.73 (4.70); C5H11NO6Co (2‚Co): N 5.84 (5.83), C 25.28
(25.01), H 4.43 (4.62); C5H11NO6Ni (3‚Ni): N 5.74 (5.84), C 24.92
(25.04), H 4.19 (4.62). The IR spectrum of complex 1-3 are listed
as below: 1‚Mn: 3062w, 3026w, 2851w, 2828w, 2496w, 1631sh,
1594s, 1474w, 1460w, 1443w, 1368m, 1345m, 1029w, 812m; 2‚
Co: 3058w, 3020w, 2943w, 2862w, 2844w, 2797w, 2498w,
1631sh, 1586s, 1473w, 1459w, 1443w, 1367m, 1348m, 1028w,
806m; 3‚Ni: 3021w, 2941w, 2866w, 2849w, 2797w, 2502w,
1633sh, 1585s, 1474w, 1460w, 1443w, 1368m, 1345m, 1029w,
812m. The detailed results of thermal analyze are discussed in the
Results and Discussion section.
Results and Discussion
Synthesis. The syntheses of 1‚Mn, 2‚Co, and 3‚Ni are
intriguing and out of the expectation of our original purpose,
which is to synthesize complexes with another ligand by
solvothermal method using H2O and DMF as solvent. To
our surprise, the title crystals grew from the mother liquid.
Preparation of Compounds 4-7. Compounds 4-7 formulated
as Mn(CHOO)2 (4‚Mn238), Mn3O4 (5‚Mn450), CoO‚Co (6‚Co320),
Inorganic Chemistry, Vol. 43, No. 15, 2004 4617

Wang et al.
As a matter of fact, it has long been known that DMF can
be hydrolyzed to formate and DMA and the rate increases
especially at basic conditions and at high temperature.16
Under our experiment conditions, the weak basic character
of the MII ions due to their hydrolyzation and the high
temperature and pressure favor the hydrolyzation of DMF.
We think that the MII ions behave as the catalyst for the
hydrolyzation of DMF because they promote the balance of
the hydrolyzation.
DMF played an important role in the synthesis of the
compounds. Besides behaving as the source of formate and
DMA, DMF also serves as the solvent for the crystals to be
separated out because the compounds are very soluble in
H2O, as is the reason that the crystals grow out after the
evaporation of H2O 1 week later rather than right after the
cooling of the reaction liquid. After we got the crystal
structures of these compounds, we synthesized them from
the starting materials NaCHOO‚2H2O and DMA‚HCl in the
same experiment conditions with a higher yield. Moreover,
the synthesis route with DMF as the starting material enables
us to synthesize the deuterated complexes easily from
commercial available d-7-DMF C3D7NO.
Figure 1. The perovskite-like structure of M(CHOO)3[NH2(CH3)2]
(M ) Mn(1‚Mn), Co(2‚Co), Ni(3‚Ni)) with DMA in the cage (1‚Mn as the
example). The hydrogen atoms in DMA were not added for clarity because
of the disorder of the DMA molecules.
with formate as the bridge ligand.2,4,7 The C-O bond length
and the angle of O-C-O have average values approximately
1.25 Å and 125°. The asymmetric parameter δ (δ ) d[C-
O] - d[C-O′]) is almost zero for all three compounds, which
means that the two C-O bonds are averaged compared with
HCOOH. Generally speaking, the carboxylate group in anti-
anti mode favors the antiferromagnetic couple between the
spin carriers,12 that has been convinced by our magnetic
investigation of these three compounds. The oxygen atoms
of formate are involved in strong hydrogen bonds with the
H-N from DMA cations residing in the cages, (N-O
Crystal Structure
Single-crystal X-ray investigations of 1-3 at 297 K
revealed that all these three compounds are isomorphism,
crystallized as a three-dimensional distorted perovskite-like
structure in the trigonal space group R 3hc with slight
differences in the lattice parameters. The details of their
structures can be found in CIFs in the Supporting Informa-
tion. The asymmetric unit of the structure contains one
independent metal cation on the special position (0,0,0), one
formate group CHOO- bonding to the metal cation, and one
distances are 2.933, 2.909, and 2.932
Å
for
+
disordered DMA cation NH2(CH3)2 . Each metal ion is
1‚Mn, 2‚Co, and 3‚Ni, respectively. The DMA cations in
the cages are 3-fold disordered with two carbon atoms lying
in the 3-fold axis (Figure 1). As can be seen from the
synthesis procedure, the DMA is very important to the
formation of the three-dimensional framework and can be
considered as the template for the structure.
connected to its six nearest neighbors through six formate
bridges (Figure 1). The Mn-O distance is 2.187 Å, longer
than those in Mn(CHOO)2 [2.138(2) Å]3a and MnIII(CHOO)3
[2.001(3) Å].2b The distances of Co-O and Ni-O are 2.109
and 2.095 Å, respectively; both are slightly longer than those
in the dihydrate salts and related compounds.4a,6e,f,9 From Mn
to Co and Ni, the decrescent distance of M-O might result
from the decrescent radii of the metal ions MII.
The coordination polyhedron around metal ion can be
described as a trigonally distorted octahedron (MO3O′3). The
M-O distances are all the same in one compound, but the
angles of O-M-O are divided into two groups: those bigger
than 90° (90.93(3)°) related by a 3-fold axis and their
supplementary angles (89.07(3)°). As a result of the distor-
tion, the octahedron loses its 4-fold axes and retains its 3-fold
axes. This could be the reason that those compounds are
crystallized in a trigonal rather than a cubic space group.
Comparatively, in the very similar structure MnIII(CHOO)3,2b
the Mn-O distances and the O-M-O angles are all the
same, as leads to the almost perfectly octahedral polyhedra
and the cubic space group.
The M-M distances are all the same, and the M-M-M
angles are divided into two groups. The nearest distances
between MII ions are 6.141, 6.011, and 6.011 Å for 1‚Mn,
2‚Co, and 3‚Ni, respectively. At the same time, each MII
ion has 12 next-near neighbors with the distance about 8-9
Å. The abundant neighbors and the short distances make the
magnetic interaction more efficient and may lead to the 3D
long-range order. The orientation of the octahedra of M and
the cages containing DMA can be seen from Figure 2. These
polyhedra screw with each other along the extending
directions in the crystal, which gives a distorted perovskite-
like structure. The smallest torsion angles of O-M-M-O
are 39.65°, 39.13°, and 38.58° for 1‚Mn, 2‚Co, and 3‚Ni,
respectively. As will be seen below, this distortion from the
perovskite structure has an important influence on the
magnetic properties of all these three compounds.
Each formate group connects two metal atoms with the
2.11-anti-anti mode similar to many reported compounds
TGA of 1-3 and Characterization of 4-7. The impor-
tance of the thermal behavior studies of transition-metal
formates arises from the possibility of their use for prepara-
(16) Juillard, J. Pure Appl. Chem. 1977, 49, 885-892.
4618 Inorganic Chemistry, Vol. 43, No. 15, 2004

PeroWskite-like Metal Formates
are two main peaks at 638.9 ev (fwhm ) 3.674 ev) and 527.4
ev (fwhm ) 1.630 ev), respectively, corresponding to the
Mn 2p3/2 and O 1s, in agreement with those recorded on the
commercial available Mn3O4 and reported value 641.4 eV
(Mn2p3/2) and 529.6 ev (O 1s) for Mn3O4.19 The Mn/O atomic
ratio (total manganese/oxygen in the lattice ) 1.26) is in
good agreement with the theoretical value (1.33). Actually,
the formation of Mn3O4 was not strange at all. M. Viertelhaus
et al. had mentioned in their report that Mn(CHOO)2‚2H2O
turned into Mn(CHOO)2 at 120 °C and then to Mn3O4 at
251 °C at nitrogen atmosphere.3a J. Larionova et al. also
reported that at 513 K Mn3O4 phase could be formed by
heating the famous SMM material Mn12.13a But the Mn3O4
they reported was crystallized rather than an amorphous
phase, which might lead to the differences in the magnetic
characters.
To make sure the presence of Ni and C in the final product
7‚Ni300, TGA was performed on it from room temperature
to 600 °C at air atmosphere (Figure S3). Due to the reaction
listed below
Figure 2. Three-dimensional structure of M(CHOO)3[NH2(CH3)2] (M )
Mn(1‚Mn), Co(2‚Co), Ni(3‚Ni)) with DMA in the cages. The neighbor
polyhedrons are twisted along the extending directions bridged by formate.
3NiO(s) + Ni(s) + C(s) + 3/2O2(g) f 4NiO(s) + CO2(g)
the weight should increase 13.4%. In fact, from temperature
319 °C to 427 °C, the weight increased 17.6%, which
together with the element analysis result confirms the
composition of 7‚Ni300.
tion of metal powders and metal oxide catalysts.14,17,18 As
examples, at inert atmosphere, Co(CHOO)2‚2H2O dehydrates
at the first stage and then decomposes to get CoO+Co;18b
and for Ni(CHOO)2‚2H2O, the final product is a finely
divided nickel metal which has widespread applications as
the nickel-supported catalysts.14,17 Detailed investigations
were focused on M(CHOO)2‚2H2O systems, but fewer data
can be found for other formate compounds. All three
compounds 1-3 undergo the similar two-step weight loss
in the same conditions (Figure S1). The weight losses are
as follows (%, the temperature ranges are in parentheses).
1‚Mn: 37.5 (208-238 °C), 28.0 (320-389 °C); 2‚Co: 54.0
(180-241 °C), 19.7 (297-311 °C); 3‚Ni: 41.8 (188-251
°C), 24.6 (251-281 °C). By the results of TGA, the final
products of 1-3 are confirmed to be Mn3O4, CoO‚Co, and
3NiO‚Ni‚C. Under TGA control, we heated 1-3 at corre-
sponding temperatures (Table 1) for 30 min under an argon
atmosphere to get materials 4-7. The IR spectrum of the
dehydrate Mn(CHOO)2(4‚Mn238) is quite similar with that
reported Mn(CHOO)2‚2H2O.3a We think it is because that
the dehydrate Mn(CHOO)2 is quite ready to absorb water
from the atmosphere. In fact, thermal analysis on 4‚Mn238
at the same condition as 1‚Mn (see Figure S2) shows a loss
of water (1.5%) from 76.5 °C to 98 °C. 5‚Mn450 was
confirmed to be amorphous Mn3O4 by chemical element
analysis, XPS, variable-temperature XRD (vide infra), and
magnetic measurement (vide infra) in addition to the TGA
result. The survey XPS spectrum recorded for 5‚Mn450
shows no significant presence of impurities, except for the
contaminant carbon. From binding energy 0-1200 eV there
Variable-Temperature XRD. From the crystal structures
and the TGA results, we notice that compounds 1-3 have
three-dimensional cages with DMA settled in and keep stable
up to 200 °C or so. Because of the increasing interests of
porous materials, we hope this kind of 3D framework with
cages can be maintained after removing those DMA mol-
ecules and can be refilled by other similar molecules. But
the results are far from our original thought. XRD patterns
for 1‚Mn and 2‚Co were recorded in the range 13° < 2θ <
45° at some selected temperatures from room temperature
to 600 °C. The results can be seen from Figures S4 and S5.
The peaks at 2θ ) 29.9° and 39.4° are the diffractions from
the substrate. At room temperature, the XRD spectrums for
both of them are in good agreement with the theoretical
patterns calculated from the single-crystal data. The intensi-
ties of diffraction peaks decrease as the temperature increases
and disappear at 250 °C for 1‚Mn and 225 °C for 2‚Co,
clearly indicating that the frameworks of these compounds
collapse at the corresponding temperatures and keep amor-
phous until the highest temperatures studied. These results
disappointed us at first, but they were reasonable after a
reflection on the nature of the compounds. As a matter of
fact, DMA is a cation lying in the cages of the 3D framework
and interacts with the formate through hydrogen bonds. The
simple removal of DMA cations would leave a framework
negatively charged, which is unfavorite. To keep the
framework neutrally charged, the loss of a DMA cation might
be accompanied by the loss of a formate anion. This
breakdown from the molecular level leads to the collapse of
(17) Edwards, A. B.; Garner, C. D. J. Phys. Chem. B 1997, 101, 20-26.
(18) (a) Masuda, Y.; Hatakeyama, M. Thermochim. Acta 1998, 308, 165-
170. (b) Ingier-Stocka, E.; Grabowska, A. J. Thermal Anal. 1998, 54,
115-123. (c) Galway, A. K.; Jamieson, D. M.; Brown, M. E. J. Phys.
Chem. 1974, 78, 2664.
(19) Ardizzone, S.; Bianchi, C. L.; Tirelli, D. Colloids Surf., A 1998, 134,
305-312.
Inorganic Chemistry, Vol. 43, No. 15, 2004 4619

Wang et al.
Figure 5. Field dependent isothermal magnetization M(T,H) for 1‚Mn at
1.8 K from 0 to 70 kOe. Insert: the hysteresis loop measured at 2 K.
Figure 3. Temperature dependence of øM of 1‚Mn at H ) 1000 Oe from
2 to 300 K. Points are experimental data; the line is the best-fit using the
model developed by Rushbrook and Wood. Inset: plots of øMT(T) and
øM-1(T) of 1‚Mn measured at 1 kOe from 2 to 300 K.
has a maximum at ca. 8.7 K, and no detected out-of-phase
signal and frequency dependence were observed.
To clarify the nature of the low-temperature phase, we
have measured the isothermal magnetization M(T,H) at 1.8
K with fields up to 70 kOe (Figure 5). The magnetization
increases almost linearly from 0 to 70 kOe and reaches 2.39
µB per MnII at 70 kOe, far from the saturation value MS )
5 µB for a spin only MnII ion, indicating the antiferromagnetic
coupling between Mn ions. In the low field range at 2 K, a
hysteresis loop can be observed with the coercivity field HC
about 90 Oe and the remnant magnetization MR ) 0.0069
µB. Till now, by considering the results of ZFC and FC
measurements, ac susceptibility measurement and isothermal
magnetization, we may safely come to the conclusion that
1‚Mn is a 3D weak ferromagnet originating from spin canting
below the critical temperature Tc ) 8.5 K. The canting angle
R is related to MR and MS through sin (R) ) MR/MS20 and is
estimated to be about 0.08°.
Figure 4. Temperature dependence of øM of 1‚Mn measured at 200 Oe
from 2 to 20 K (ZFC and FC).
the 3D framework and the form of amorphous phase of the
products. This kind of amorphous character makes the
magnetic properties of 5‚Mn450 more interesting and will
be discussed as follows. Although DMA cannot be removed
by thermal treatment, we think it may be exchanged by other
similar cations in the solvent. This idea is now on the
schedule of our further study.
On the basis of the structural analysis, the high-temperature
magnetic susceptibility of this compound may be analyzed
by the model21 developed by Rushbrook and Wood for a
Heisenberg antiferromagnet on a simple cubic lattice with
the Hamiltonian H ) -2J∑ i,j Si‚Sj, namely,
35Ng2â2
6
Magnetic Properties
ø )
(1 +
C xn)
(1)
∑
n
1‚Mn(CHOO)3[NH2(CH3)2]. Temperature dependence of
the magnetic susceptibility øM of 1‚Mn was measured using
a collection of some single crystals orientated randomly in
a magnetic field of 1 kOe, which shows a peak at about 8.0
K (Figure 3). The øMT value (inset of Figure 3) at 300 K is
4.15 cm3 mol-1 K, which decreases smoothly upon cooling
to ca. 50 K and then drops quickly to 0.29 cm3 mol-1 K at
1.9 K. The magnetic susceptibility in the range of 10-300
K obeys the Curie-Weiss law with a Curie constant C )
4.37 cm3 mol-1 K, which fits well to the expected value 4.375
cm3 mol-1 K for MnII ion, and a negative Weiss constant θ
of -16.3 K, which, together with the shape of the curves
øM(T) and øMT(T), indicates the presence of antiferromagnetic
coupling between Mn ions. ZFCM and FCM in a low field
of 200 Oe show abrupt increases at ca. 8.5 K and then
diverge at ca. 8.3 K with the decreasing of temperature
(Figure 4). This behavior may arise from the onset of long-
range ordering. Ac magnetic susceptibility with frequency
of 111, 199, 355, 633, and 1111 Hz was measured under
Hdc ) 0 Oe and Hac ) 2 Oe (Figure S6). The in-phase signal
12kT
n ) 1
with x ) J/kT, and C1 ) 35, C2 ) 221.67, C3 ) 608.22, C4
) 26049.66, C5 ) 210986, and C6 ) 8014980. The best
fitting of the magnetic data above 30 K (Figure 3) using eq
1 gives J ) -0.23 cm-1 by fixing g ) 2.00. At the same
time, according to the molecular field theory of antiferro-
magnetism,20,22
θ ) 2S(S + 1)zJ/3k
(2)
where θ, S, J, and k have their usual meanings, and z is the
magnetic coordination number of a lattice site. For this
5
sample 1‚Mn, θ ) -16.3 K, S ) /2 for MnII, and z ) 6.
Using eq 2, we get J ) -0.32 cm-1, as is comparable to
that deduced from eq 1 and consistent with that reported for
Mn-O-C(R)-O-Mn interactions (-0.2 to -0.3 cm-1).3a
(20) Kahn, O. Molecular Magnetism; VCH: New York, 1993.
(21) Rushbrook, G. S.; Wood, P. J. Mol. Phys. 1958, 1, 257-283.
(22) Carlin, R. L.; Van-Duyneveldt, A. J. Magnetic Properties of Transition
Metal Compounds; Springer-Verlag: New York, 1977.
4620 Inorganic Chemistry, Vol. 43, No. 15, 2004

PeroWskite-like Metal Formates
Table 2. Comparison of the Properties of Mn Formates
magnetic
coupling
connection mode
of CHOO-
structure
dimensionality
magnetic
dimensionality
Tc, K
I
Mn(CHOO)2‚2H2O
3.7
AF
2.11-anti-anti
2.11-syn-anti
2.11-anti-anti
2.11-anti-anti
3.12
2.11-anti-anti
2.11-syn-anti
3.12
3D
2D
Mn(CHOO)2‚2(NH2)2CO
1‚Mn
Mn3(CHOO)6‚G
Mn(CHOO)2
3.78
8.5
8a
AF
AF
AF/F
AF
2D
3D
3D
3D
2D
3D
3D
3D
II
8
III
MnIII(CHOO)3
27
F/AF
2.11-anti-anti
3D
3D
a Tc of this kind of materials relies on the guests in the molecular. The desolvated form Mn3(CHOO)6 has Tc ) 8.0 K. Others vary from 4.8 to 9.7 K.
Up to date, the reported 3D long-range ordered formates
of Mn ions are quite few. To the best of our knowledge, the
dihydrated formate Mn(CHOO)2‚2H2O,5b,c the anhydrous salt
Mn(CHOO)2,3a the manganese formate diruea Mn(CHOO)2‚
2b
2(NH2)2CO,3b,c Mn3(CHOO)6‚G,2a and MnIII(CHOO)3 are
the only several examples. Comparison of the magnetic
properties of our compound 1‚Mn with these five compounds
is of some interest. Some details are listed in Table 2.
Comparing 1‚Mn with MnIII(CHOO)3, the structure char-
acters, including the coordination environments of the central
Mn ions, the anti-anti connection mode of the bridging
formate, and the host-guest relationship of the entire
framework, are very similar, but the magnetic properties are
Figure 6. Temperature dependence of øMT and øM-1of 2‚Co at H ) 20
quite different. These differences could be mainly originated
kOe from 2 to 300 K. The line is the best-fit using Curie-Weiss law.
from the differences in the oxide state of Mn ions (+2 for
1‚Mn and +3 for MnIII(CHOO)3) and in the nuance of the
structures. Actually, a remarkable Jahn-Teller distortion and
the relative shorter distances between the MnIII ions (vide
supra) may lead to the higher critical temperature and the
coexistence of ferromagnetic and antiferromagnetic couple
between MnIII ions, which makes MnIII(CHOO)3 an anti-
ferromagnet with the AF coupled ferromagnetic-layer and
the spin-flip transtion. While for 1‚Mn, the magnetic coupling
between MnII is entirely antiferromagnetic. The distorted
arranged octahedra and the absence of symmetrical centers
between the neighboring MnII ions cause the nature of spin-
canted weak ferromagnetism (vide infra).
As to Mn(CHOO)2‚2H2O and Mn(CHOO)2‚2(NH2)2CO,
their crystal structures are quite different, but the critical
temperatures are almost the same. Careful examination of
the crystal structures reveals that there are some essential
connection and similarities between them, especially in the
viewpoint of magnetism. In Mn(CHOO)2‚2H2O, Mn2+ ions
may be divided into two groups, called A- and B-ions. The
network of A-ions may be regarded as a layer parallel to
the bc-plane bridged by anti-anti formate, while the
independent B-ions are connected to two adjacent A-ions.
Detailed studies reveal that half of the moments (B-ions)
are paramagnetic even below Ne´el temperature. In this
regard, Mn(CHOO)2‚2H2O has a simple quasi-two-dimen-
sional magnetic structure formed by A-ions such as that in
Mn(CHOO)2‚2(NH2)2CO, so it is not strange that they are
magnetic ordering at almost the same temperature. Similarly,
the magnetic properties of those 3D compounds Mn3-
(CHOO)6, Mn(CHOO)2, and 1‚Mn are very similar, although
the structure details are of great differences.
Insert: ZFC and FC curves at H ) 10 Oe from 2 to 25 K.
From the discussion above and in the view of magnetism,
these six compounds can be divided into three groups:
groups I, II, and III with the critical temperatures increase
from 3.7 to 8 K and to 27 K (Table 2). As for groups I and
II, the main difference is in the different dimensionality of
the magnetic structure. The increase of critical temperatures
indicates that increasing of the dimensionality will greatly
increase the critical temperature. While for groups II and
III, the main difference is in the local properties of the spin
carrier, and the comparison of them suggests that it is the
magnetic nature of the magnetic centers that dominates the
magnetic properties of the compounds. In fact, these conclu-
sions are not so new but give us some hints on getting
molecule-based magnets with high critical temperature.
2‚Co(CHOO)3[NH2(CH3)2]. Temperature dependence of
the magnetic susceptibility øM from 2 to 300 K of 2‚Co was
measured using a collection of some single crystals with a
weight of 14.0 mg orientated randomly at a magnetic field
of 1 kOe, which increases gradually as the temperature
decreases and shows a plateau below 15 K (Figure S7). The
figure of øMT(T) and øM-1(T) can be seen from Figure 6.
The øMT value at 300 K is 3.30 cm3 mol-1 K, which
decreases smoothly to 0.94 cm3 mol-1 K upon cooling to
ca. 16 K and abruptly increases to a maximum value of 2.19
cm3 mol-1 K at 13 K and then decreases until the lowest
temperature 2 K. The magnetic susceptibility in the range
of 100-300 K obeys the Curie-Weiss law with a Curie
constant C ) 3.80 cm3 mol-1 K, which is close to the
experimental value 3.0 cm3 mol-1 K for an octahedronal CoII
ion,22 and a negative Weiss constant θ of -49.4 K. The Curie
Inorganic Chemistry, Vol. 43, No. 15, 2004 4621

Wang et al.
Figure 7. Real (øM′) and imaginary (øM′′) ac magnetic susceptibilities in
zero applied dc field and an ac field of 5 Oe at different frequencies (277,
666, 1633, 4111 Hz) for 2‚Co.
Figure 9. Temperature dependence of øM of 3‚Ni measured at 200 Oe
from 2 to 65 K. Inset: plots of øMT(T) and øM-1(T) of 3‚Ni measured at 20
kOe from 2 to 300 K. The line is the best-fit according to Curie-Weiss
law.
but we consider that this phenomenon might come from the
reorientation of the ordered spin state due to the move of
the domain wall. In fact, this phenomenon of reorientation
is not strange and has once been found long ago in some
other magnetic system.23 The detailed measurements, includ-
ing the specific heat, the neutron diffraction on the deuterated
sample, and the careful investigation on single crystal, are
still needed to clearly understand the magnetic properties of
2‚Co. Works following this way are on the schedule.
Using the molecule field theoretical eq 2, we get the
exchange constant J ) -2.3 cm-1. The negative Weiss
constant θ and exchange constant J should include a
contribution of spin-orbit coupling.
Figure 8. Field dependent isothermal magnetization M(T,H) for 2‚Co at
1.8 K from 0 to 70 kOe. Insert: the hysteresis loop measured at 1.8 K.
constant for spin-only CoII is 1.875 cm3 mol-1 K; the
relatively bigger C value for 2‚Co should arise from the
significant spin-orbit coupling of CoII. Both ZFC and FC
curves at 10 Oe increase abruptly at 14.9 K with the
decreasing of temperature; then øM reaches a plateau before
13.1 K and keeps on increasing abruptly to another plateau
with the further decreasing temperature. Ac magnetic sus-
ceptibility with frequency of 277, 666, 1633, and 4111 Hz
was measured under Hdc ) 0 Oe and Hac ) 5 Oe (Figure 7).
The in-phase signal has two maximum at ca. 14.9 and 13.1
K. No visible out-of-phase signal and frequency dependence
were observed.
Besides the cobalt formate Co(CHOO)2‚2H2O,5c another
long-range ordered Co formate is Co(CHOO)2(HCONH2)2,4a
a compound in which cobalt ions are bridged by formate
ions with anti-anti mode in the same extended 2D layer as
that of Mn(CHOO)2‚2(NH2)2CO. As discussed above, the
magnetic structures of Co(CHOO)2‚2H2O and Co(CHOO)2-
(HCONH2)2 are two-dimensional, and the critical tempera-
tures are 5.1 and 9 K, respectively, lower than that (14.9 K)
of 2‚Co. In Co(CHOO)2‚2H2O, the magnetic-ordered A-ions
layers are separated by totally paramagnetic B-ions so that
the interlayer interaction might be neglectable. But in
Co(CHOO)2(HCONH2)2, hydrogen bonds existing between
the next two layers lead to weak interlayer interactions and
the higher critical temperature. The gradually increasing Tc
also indicates the importance of the dimensionality for the
critical temperature.
The isothermal magnetization M(T, H) at 2 K with field
up to 70 kOe and the hysteresis loop at 1.8 K have been
represented in Figure 8. The magnetization increases almost
linearly from 0 to 70 kOe and reaches 0.73 µB per CoII at 70
kOe, far from the saturation value MS ) 3 µB for a spin-
only CoII ion. This behavior strongly indicates the antifer-
romagnetic coupling between Co ions, together with the
negative Weiss constant. At 1.8 K, a hysteresis loop can be
observed with the coercivity field HC about 920 Oe and the
remnant magnetization MR ) 0.028 µB. From the results of
the whole set of experiments carried out above, we come to
a conclusion that the low-temperature state below 14.9 K of
2‚Co is weak ferromagnetic based on antiferromagnetic
interaction and spin canting. The canting angle R is related
to MR and MS through sin(R) ) MR/MS and is estimated to
3‚Ni(CHOO)3[NH2(CH3)2]. Temperature dependence of
the magnetic susceptibility øM from 2 to 300 K of 3‚Ni was
measured using a collection of some single crystals (27.0
mg) orientated randomly at a magnetic field of 20 kOe, which
increases gradually and shows a shoulder at about 31 K and
then continues increasing with the decreasing of temperature
(Figure S8). The curves of øMT(T) and øM-1(T) are presented
in the inset of Figure 9. The øMT value at 300 K is 1.03 cm3
mol-1 K, which fits the expected value 1.00 cm3 mol-1
K
for an uncoupled NiII and decreases gradually to 0.044 cm3
3
be about 0.5 ° with the efficient spin S ) /2 for CoII. The
(23) (a) Swu¨ste, C. H. W.; Botterman, A. C.; Millenaar, J.; De Jonge, W.
J. M. J. Chem. Phys. 1977, 66(11), 5021-5030. (b) Srinivasan, G.;
Seehra, M. S. Phys. ReV. B 1983, 28, 1.
origin of the two-step transfer (two-step increase of ZFC and
FC, two peaks in ac susceptibility) is not entirely understood,
4622 Inorganic Chemistry, Vol. 43, No. 15, 2004

PeroWskite-like Metal Formates
Figure 11. Hysteresis loop of 3‚Ni measured at 1.8, 15, and 30 K. Insert:
field dependent isothermal magnetization M(T,H) for 3‚Ni at 1.8 K from 0
to 70 kOe.
Figure 10. Real (øM′) and imaginary (øM′′) ac magnetic susceptibilities
in zero applied dc field and an ac field of 5 Oe at different frequencies
(199, 355, 633, 1111 Hz) for 3‚Ni.
are in the progress. Using the molecule field theoretical eq
2, we get the exchange constant J ) -4.85 cm-1.
mol-1 K upon cooling to ca. 2 K. The magnetic susceptibility
in the range of 60-250 K obeys the Curie-Weiss law with
a Curie constant C ) 1.27 cm3 mol-1 K, which is close to
the expected value 1.00 cm3 mol-1 K for NiII ion, and a
negative Weiss constant θ of -55.78 K, which, together with
the curve of øMT(T), indicates the presence of antiferromag-
netic coupling between Ni ions. FCM at a low field of 200
Oe shows an abrupt increase at ca. 35 K with the decreasing
temperature (Figure 9), suggesting that an onset of long-
range ordering may occur. Furthermore, øM shows another
little abrupt increase from 0.258 to 0.278 cm3 mol-1 at ca.
14 K. The origin of this phenomenon is not entirely
understood now, but as for 2‚Co we consider that it might
come from the spin reorientation of the ordered state. The
fine agreement between the experimental diffraction pattern
and that calculated from the single-crystal structure indicates
the purity of our sample 3‚Ni. Ac magnetic susceptibility
with frequency of 199, 355, 633, and 1111 Hz was measured
under Hdc ) 0 Oe and Hac ) 5 Oe (Figure 10). The in-phase
signal has two maximum at ca. 35.6 and 14.3 K. And there
is a nonzero out-of-phase signal below 35 K with slight
frequency dependence.
As mentioned above, the bridging carboxylic groups with
anti-anti mode mediate the antiferromagnetic coupling
between adjacent spin carriers due to the geometry of the
magnetic orbital. Given the neighboring spin carriers in a
unit cell are not related by an inversion center and coupled
antiferromagnetically, weak ferromagnetism may arise from
the canting of spins because of the Dzyaloshinsky-Moriya
(D-M) interaction.22 Although it was first observed in some
single-atom-bridged systems such as R-Fe2O3, â-MnS, NiF2,
and CsCoCl3‚2H2O, there is a growing interest in molecular
weak ferromagnets in recent years with a three-atom single-
bridge bonding mode. Typical three-atom single bridges are
-
azide µ-1,3-N3 , and formate µ-CHOO, and µ-1,3-NCNCN,
µ-OC(R)O, µ-imidazolate etc. can be viewed as quasi-three-
atom bridges.24 Despite our three compounds are crystallized
in the center symmetric space group (R 3hc), the inversion
centers are located in the metal ions, which makes the metals
bridged by formate unrelated by a symmetric center and the
D-M interaction can function. As can be seen from Figure
2, the octahedra are connected by formate and twist by about
39° (the torsion angle of O-M-M-O, vide supra) along
three extending directions. We think that nonsymmetry
antiferromagnetic coupling arising from the characteristic
structure for all three compounds and single-ion anisotropy
for 2‚Co and 3‚Ni, existing for most of the cobalt and nickel
complexes, lead to canted weak ferromagnetism. The inves-
tigation on the magnetic structures of this series of deuterated
complexes by neutron diffractions is now under way.
Interestingly, the formate series M(CHOO)2‚2H2O (M ) Mn,
The isothermal magnetization M(T, H) at 1.8 K with field
up to 70 kOe and the hysteresis loops at 1.8, 15, and 30 K
have been represented in Figure 11. The magnetization
increases almost linearly from 0 to 70 kOe and reaches 0.21
µB per NiII at 70 kOe, far from the saturation value MS ) 1
µB for a spin only NiII ion, indicating the strong antiferro-
magnetic coupling between Ni ions. In the low field range
at 1.8 K, a hysteresis loop can be observed with the coercivity
field HC about 320 Oe and the remnant magnetization MR )
0.0021 µB. From the results of the whole set of experiments
carried out above (dc, ac, M(T, H) and hysteresis loop), we
come to a conclusion that the low-temperature state below
35.6 K of 3‚Ni is weak ferromagnetic based on antiferro-
magnetic interaction and spin canting. The canting angle is
estimated to be about 0.6 °. Strangely, the HC and MR at 15
and 30 K are almost the same, namely 650 Oe and 0.0032
µB (Figure 11). These results indicate that the phase below
14 K is not exactly the same as that in the temperature range
14-35 K. Further investigations on the deuterated sample
3b
Fe, Co, Ni), Co(CHOO)2(HCONH2)2, and â-Cu(CHOO)2
are all canted weak ferromagnets below TN, and this gives
us an inspiration to synthesize the canted weak ferromagnets
using CHOO- as bridge.
5‚Mn450 (Mn3O4). It has long been known that Mn3O4
exhibits interesting magnetic properties.23b,25 Because of the
(24) For examples, see: (a) Han, S.; Manson, J. L.; Kim, J.; Miller, J. S.
Inorg. Chem. 2000, 39, 4182. (b) Manson, J. L.; Kmety, C. R.; Epstein,
A. J.; Miller, J. S. Inorg. Chem. 2000, 38, 2552. (c) Tian, Y.-Q.; Cai,
C.-X.; Ren, X.-M.; Duan, C.-Y.; Xu, Y.; Gao, S.; You, X.-Z. Chem.
Eur. J. 2003, 9, 5673 and references therein.
Inorganic Chemistry, Vol. 43, No. 15, 2004 4623

Wang et al.
from the disorder in the sample, although we do not entirely
understand it yet. It is widely known that high disorder and
presence of defects in a magnetic solid lead to higher
coercivities due to more domain walls in the materials.
Actually, the amorphous phase often leads to higher coer-
civities and low critical temperatures. As can be seen from
the variable temperature XRD analysis (Figure S4), since
the beginning of decomposition of 1‚Mn at about 200 °C,
the sample becomes amorphous until the generation of 5‚
Mn450. From the structural point of view, the decomposition
of 1‚Mn starts from the loss of CHOO‚NH2(CH3)2, which
leads to the entire collapse of the whole three-dimensional
crystalline structure. In recent investigations of magnetism,
one of the main objects is based on the generation of high
magnetic energy defined by E ) MR × HC. Amorphism,
due to its high disorder and presence of defects, may lead to
one way to find materials with high HC and high magnetic
energy.
6‚Co320 and 7‚Ni300. For these two samples, the
magnetic properties are difficult to investigate because of
the impurity. But the magnetic data measured for them give
us another evidence for the presence of the metal particles
in 6‚Co320 and 7‚Ni300. It is well-known that CoO and NiO
are all antiferromagnets with the critical temperatures 292
and 523 K, respectively, but Co and Ni are ferromagnets at
the room temperature. So we can know if there are metal
particles in our samples from the character of the curves of
M(H, T) vs H shown in Figures S11 and S12 for 6‚Co320
and 7‚Ni300, respectively. The magnetizations of these two
samples increase abruptly as the field increases and reach
their saturation values at low field (10 kOe for 6‚Co320 and
5 kOe for 7‚Ni300). These are the typical behaviors of
ferromagnets at the temperatures below their critical tem-
peratures and confirm the presence of Co and Ni in 6‚Co320
and 7‚Ni300 again, together with the evidences mentioned
above.
Figure 12. Field dependent isothermal magnetization M(T,H) for 5‚Mn450
at 1.8 K from 0 to 70 kOe. Insert: the hysteresis loop for 5‚Mn450 and
bulk Mn3O4 measured at 1.8 and 30 K.
triangular arrangement of the magnetic moments, bulk Mn3O4
behaves as a canted ferrimagnet with the critical temperature
approximately at 42 K. For 5‚Mn450 and bulk Mn3O4, the
øMT(T) and øM-1(T) at field of 5 kOe and the ZFC/FC curves
at 200 Oe of 5‚Mn450 can be seen from Figure S9. From
the good agreement of the two sets of data at high
temperatures and the evidences mentioned above, we could
come to the conclusion that our sample 5‚Mn450 is a pure
form of Mn3O4. Furthermore, ac magnetic susceptibility
under Hac ) 3 Oe and frequency of 355, 633, and 1111 Hz
was measured for 5‚Mn450 which confirms the magnetic
transition at 42 K (Figure S10).
Due to the difference of the øMT below 42 K between the
amorphous sample and the bulk Mn3O4, we measured the
field dependence of the magnetization for both samples in
the ordered state at 1.8 and 30 K, respectively. The results
are presented in Figure 12. The behavior of 5‚Mn450 is
somewhat different from that for the bulk Mn3O4 and those
reported for other Mn3O4 samples.15,23b,25 The significant
difference is in the hysteresis loop. The saturation value for
5‚Mn450 at 1.8 K up to 70 kOe is 1.80 µB, which is slightly
smaller than that for bulk Mn3O4 and the saturation value
1.88 µB reported by others.23b,25 The remnant magnetization
MR at zero field is 1.09 µB, lower than that 1.85 µB reported
by K. Dweight and N. Menyuk for a single-crystal sample25c
but much larger than that 0.29 µB reported by Aure´lie
Buckelew et al.15 The most intriguing property is the
coercivity field for the amorphous materials. At 1.8 K, the
coercive field HC for 5‚Mn450 is 11.9 kOe, much higher
than the value 8.8 kOe reported by Aure´lie Buckelew et al.,
which is already larger than the values reported by others
including the bulk Mn3O4 (2.8 kOe at 5 K) and the films
(3.5 kOe at 5 K). Actually we got the HC value 8.4 kOe for
bulk polycrystalline Mn3O4 at the same measurement condi-
tions. At 30 K, the difference between the HC values for 5‚
Mn450 and bulk Mn3O4 is even more obvious. As can be
seen from Figure 12, the HC for 5‚Mn450 is 4.1 kOe, about
16 times the value (ca. 250 Oe) for bulk Mn3O4!
Conclusion
In conclusion, we synthesized three novel three-dimen-
sional distorted perovskite-like metal formates 1-3 using
solvothermal method from two sets of different starting
materials. And based on the thermal behaviors of the three
compounds, another four amorphous materials (4-7) were
prepared by thermal treatment on the precursors 1-3,
respectively. Magnetic studies showed that compounds 1-3
behaved as canted weak ferromagnets with the critical
temperature Tc ) 8.5 K (1‚Mn), 14.9 K (2‚Co), and 35.6 K
(3‚Ni), and for 2‚Co and 3‚Ni, spin reorientation takes place
at 13.1 and 14.3 K, respectively. All of them show hysteresis
loops below their critical temperatures. Using the model
developed by Rushbrook and Wood for a Heisenberg
antiferromagnet on a simple cubic lattice and/or the molecular
field theory for antiferromagnetism, the magnetic coupling
parameters J were estimated to be -0.23/-0.32 cm-1, -2.3
cm-1, and -4.85 cm-1 for 1‚Mn, 2‚Co, and 3‚Ni, respec-
tively. The magneto-structure relations were discussed to
reveal the relationship between the canted magnetic structures
of the noncentrosymmetrical nature of the three-atom single-
As reported by Aure´lie Buckelew, we think the origin of
the large coercivity in the sample of 5‚Mn450 might stem
(25) (a) Chartier, A.; D’Arco, P. Phys. ReV. B 1999, 60, 14042. (b)
Srinivasan, G.; Sechra, M. S. Phys. ReV. B 1983, 28, 1. (c) Dwight,
K.; Menyuk, N. Phys. ReV. 1960, 119, 1470.
4624 Inorganic Chemistry, Vol. 43, No. 15, 2004

PeroWskite-like Metal Formates
bridge of µ-CHOO-, which might have some referenced
meaning to get the canted weak ferromagnets by a simple
route. Furthermore, we investigated the magnetic property
of the amorphous Mn3O4 (sample 5‚Mn450) and found the
amazing increase of the coercivity field of it. We attributed
it to the amorphous phase, and this might provide another
efficient way for chemists to synthesize materials with high
coercivity field and magnetic energy.
Research Fund for the Doctoral Program of Higher Education
(20010001020).
Supporting Information Available: TGA curves of 1-3, 4‚
Mn238, and 7‚Ni300, variable-temperature XRD results, curves of
ac susceptibility of 1‚Mn, 5‚Mn450 at zero applied dc field with
Hac ) 2 or 3 Oe at selected frequency, curve of øM vs T for 2‚Co
and 3‚Ni from 2 to 300 K, curves of øM vs T and ZFCM/FCM of
5‚Mn450 together with that of bulk Mn3O4 and curves of M(T,H)
vs H for 6‚Co320 and 7‚Ni300, and X-ray crystallographic files
(CIFs). This material is available free of charge via the Internet at
http://pubs.acs.org.
Acknowledgment. We acknowledge support from the
National Science Fund for Distinguished Young Scholars
(20125104), NSFC No. 20221101, 20490210, and the
IC0498081
Inorganic Chemistry, Vol. 43, No. 15, 2004 4625