New pyrochlore-like compound Cu2OCl2 with S = 1/2

Article abstract of DOI:10.1143/JPSJ.75.123705

We report here on the magnetic properties of Cu₂OCl₂, the structure of which can be regarded either as a pyrochlore-like corner-shared tetrahedral network with S = 1/2 or cross-linking chains of edge-sharing CuO₂Cl₂ squares. The magnetic susceptibility exhibits a broad maximum at approximately 140 K, indicating a low-dimensional feature, and drops at 70 K accompanied with a small specific-heat jump; however, it does not follow the typical S = 1/2 linear chain model. The magnetic susceptibility can be explained by a mean field corrected chain model with a closed value of the intra- and interchain exchange constants; thus, magnetic frustration would be dominant over a low-dimensional feature in Cu ₂OCl₂. From the band calculation results, we found that the Cu 3d on-site Coulomb interaction is essential to describe the electronic structure of Cu₂OCl₂. We conclude that Cu ₂OCl₂ should be regarded as a 3D material that has the potential to combine the pyrochlore-like geometrical frustration and the strong correlation among the Cu 3d electrons. 2006 The Physical Society of Japan.

Full text of DOI:10.1143/JPSJ.75.123705

Journal of the Physical Society of Japan
Vol. 75, No. 12, December, 2006, 123705
#2006 The Physical Society of Japan
LETTERS
New Pyrochlore-like Compound Cu OCl with S ¼ 1=2
2
2
H. OKABE, K. SUZUKI, K. KAWASHIMA,
T. MURANAKA and J. AKIMITSU
Department of Physics and Mathematics, Aoyama Gakuin University,
-10-1 Fuchinobe, Sagamihara, Kanagawa 229-8558
5
(
Received July 6, 2006; accepted October 24, 2006; published December 11, 2006)
We report here on the magnetic properties of Cu2OCl2, the structure of which can be regarded either as
a pyrochlore-like corner-shared tetrahedral network with S ¼ 1=2 or cross-linking chains of edge-sharing
CuO2Cl2 squares. The magnetic susceptibility exhibits a broad maximum at approximately 140 K,
indicating a low-dimensional feature, and drops at 70 K accompanied with a small specific-heat jump;
however, it does not follow the typical S ¼ 1=2 linear chain model. The magnetic susceptibility can be
explained by a mean field corrected chain model with a closed value of the intra- and interchain exchange
constants; thus, magnetic frustration would be dominant over a low-dimensional feature in Cu2OCl2.
From the band calculation results, we found that the Cu 3d on-site Coulomb interaction is essential to
describe the electronic structure of Cu2OCl2. We conclude that Cu2OCl2 should be regarded as a 3D
material that has the potential to combine the pyrochlore-like geometrical frustration and the strong
correlation among the Cu 3d electrons.
KEYWORDS: pyrochlore, cuprate, oxychloride, frustration, low-dimensional magnetism, susceptibility, specific
heat, electronic structure, LSDA þ U
DOI: 10.1143/JPSJ.75.123705
1
5)
Much attention has been focused on a 3D network of Cu OCl . Filatov and Vergasova suggested that the
corner-sharing tetrahedra, which are found in pyrochlore and structure of melanothallite shows a large negative thermal
2
2
spinel structures, because the geometrical frustration in these expansion along the b axis (b ¼ ꢀ26:7 ꢁ 10ꢀ ) and a large
6
ꢀ6
structures amplifies the electronic correlation and quantum positive expansion along the a axis (a ¼ 50:6 ꢁ 10
þ
ꢀ
9
effects. In pyrochlore compounds R2M2O7 (R: divalent or 25:2 ꢁ 10 ) on the basis of a high-temperature X-ray
trivalent cation, M: transition metal), the magnetic ions powder diffraction study and gave a reasonable explanation
occupy the vertices of network-linked tetrahedra and no spin for the contraction mechanism.16) Despite these detailed
configuration exists that simultaneously satisfies all of the studies, the physical properties of Cu2OCl2 have not been
1)
interactions between the spins. Pyrochlore compounds reported so far. In this study, we investigate the magnetic
2
,3)
show numerous intriguing phenomena such as spin ice,
the anomalous Hall effect, metal–insulator transition,
colossal magnetoresistance,
susceptibility, specific heat, and electronic structure of
Cu OCl and reveal an anomalous magnetic ground state.
The polycrystalline sample was prepared by a conven-
4
)
5)
2
2
6
)
7)
and superconductivity.
Since the ground state of the S ¼ 1=2 quantum Heisenberg tional solid-state reaction of high-purity CuCl2 (99.995%,
antiferromagnet on the pyrochlore lattice is theoretically Sigma-Aldrich) and CuO (99.995%, Sigma-Aldrich) pow-
8
,9)
expected to be a quantum spin liquid (QSL) state, there ders. The stoichiometric powder mixtures were pressed
has been much interest in the discovery of a QSL state in into pellets and sintered at 623 K for 18 h in CO2 flow. The
pyrochlore compounds. This is because they might exhibit a sample was examined by powder X-ray diffraction using
macroscopic quantum phenomenon similar to superconduc- an X-ray spectrometer with a graphite monochromator
tivity. The pyrochlore lattice with S ¼ 1=2 is an excellent (MultiFlex, Rigaku). The magnetic susceptibility was mea-
candidate to realize a QSL state and test the validity of the sured using a SQUID magnetometer (MPMSR2, Quantum
recent theories on the quantum fluctuations in frustrated Design), and the specific heat was measured by a standard
systems. The fact that few real materials can be considered thermal relaxation method (PPMS, Quantum Design).
to posses S ¼ 1=2 spins in pyrochlore compounds is a
Electronic structure calculations were carried out by
however, these materials do not the full-potential LAPW method as implemented in the
complex problem;1
0–12)
17)
exhibit the QSL state. It is essential to realize a QSL state in WIEN2k program package within the local spin density
a
1
8)
which the S ¼ 1=2 spins are preferable because the quantum approximation (LSDA) and its LSDA þ U extension;
fluctuation of spins is more significant for small spins as Hubbard-type on-site Coulomb interaction is added to the
compared with large spins; thus, intense efforts have been LSDA. The valence electrons were used in a scalar-rela-
made to discover and propose a novel candidate for this tivistic approximation and the Perdew–Burke–Ernzerhof
state.
generalized-gradient approximation (GGA) potential was
19)
Here we report on a new material, copper oxychloride used for the exchange correlation potential. The lattice
Cu2OCl2), which has a pyrochlore-like corner-shared constants and internal coordinates were taken from the
(
16)
tetrahedral network with S ¼ 1=2. The material is the detailed structural data reported by Filatov and Vergasova;
mineral named melanothallite, consisting of Cl penetrating the LAPW sphere radii were 1.8 a.u. for Cu, 2.0 a.u. for Cl,
2
into the Cu O structure (Fig. 1). Melanothallite was first and 1.8 a.u. for O. The basis set cut-off was chosen to be
2
13)
discovered by Scacchi from the fumaroles of the Vesuvius RKmax ¼ 9:0, where R is the radius of the Cu sphere. Self-
volcano, and its detailed structure was refined by Arpe consistency was imposed on k-point meshes of approx-
and M u¨ ller-Buschbaum using the synthetic crystals of imately 171 points (10 ꢁ 10 ꢁ 10) in the irreducible wedge
14)
123705-1

J. Phys. Soc. Jpn., Vol. 75, No. 12
LETTERS
H. OKABE et al.
μ0H = 0.01 T
(
a)
1
.8
.6
.4
1
1
5
0
5
0
0
0
0
1
.8
.6
.4
c
a
b
0
0
0
μ0H = 5.5 T
0
50 100 150 200 250 300
T (K)
10
20
30
40
50
60
70
80
80 (b)
2
θ
(degrees)
μ0H = 0 T
6
4
2
0
0
0
0
Fig. 1. Schematic crystal structure and powder X-ray diffraction pattern
ꢀ
2þ
of Cu2OCl2. The large and small balls represent Cl and Cu ions,
4
3
3
0
5
0
respectively. The O2 ion is located at the center of the tetrahedron.
ꢀ
μ0H = 0 T
μ
0H = 8 T
of the Brillouin zone, which gives a convergence of 0.1
mRyd/f.u. for the total energies and charge densities. The
spin-orbit interaction was incorporated using a second
variational procedure.
60
70
T (K)
0
50
100
150
T (K)
200
250
300
Figure 1 shows the powder X-ray diffraction (XRD)
pattern of Cu OCl at room temperature. There are no
impurities, and all the observed peaks are indexed as an
Fig. 2. (a) Temperature dependence of magnetic susceptibility ꢃðTÞ of
Cu2OCl2 measured in fields, ꢀ0H ¼ 0:01 and 5.5 T (inset). The solid and
open circles represent the data for zero-field cooling and field cooling,
respectively. (b) Temperature dependence of specific heat CðTÞ of
Cu OCl . The inset shows a magnified view of C at 55–75 K. Solid and
2
2
orthorhombic lattice with space group Fddd (No. 140).
˚
˚
The lattice constants—a ¼ 9:542 A, b ¼ 9:712 A, and c ¼
2
2
˚
:396 A—are consistent with the previously reported
open symbols represent the data for ꢀ0H ¼ 0:01 and 8.0 T, respectively.
7
values.
16)
The synthesized Cu2OCl2 is a reddish brown
highly hygroscopic insulator that quickly decomposes to
.
greenish yellow Cu3Cl4(OH)2 2H2O and CuO in ambient
3
.06 Å
.
air. Cu Cl (OH) 2H O exhibits ferromagnetic behavior at
7.5 K, so we were careful to avoid the contamination of the
3
4
2
2
3.04 Å
1
3.41 Å
3.41Å
20)
3.04 Å
measured sample.
3
.06 Å
The magnetic susceptibility of Cu2OCl2 as a function of
temperature [Fig. 2(a)] shows a broad maximum at approx-
imately 140 K and suddenly drops at 70 K under both the
field cooling and zero-field cooling processes. The behavior
changes little upon varying the external fields, ꢀ0H ¼ 0:01
and 5.5 T, as shown in Fig. 2(a). A broad maximum of the
magnetic susceptibility is often found in low-dimensional
Cu
(
1)
J
inter
(
2)
J
O
inter
Cl
(2)
Jintra
(
1)
J
intra
O
21)
0
22)
materials, for example, CuGeO ,
ꢁ -NaV O ,
and
b
3
2
5
TiOCl,2 which have linear chain structures. Cu2OCl2 can
also be described as the cross-linking chains of the edge-
sharing CuO2Cl2 squares (Fig. 3) instead of an oxocentered
3)
a
16)
copper tetrahedral lattice. Therefore, the broad maximum
at approximately 140 K may be derived from the low
dimensionality of the CuO Cl square chain. Figure 2(b)
c
Fig. 3. Cross-linking chain structure of Cu2OCl2. The large and small
2
2
ꢀ
2ꢀ
2þ
white balls represent Cl and O ions, respectively. Cu ions are
represented by black balls and located at the center of the CuO2Cl2
squares.
shows the temperature dependence of the specific heat under
a zero field and an applied field of 8 T. We observed a ꢂ-type
anomaly at 70 K corresponding to the magnetization-drop
temperature, as shown in Fig. 2(a). It is seen that there is no
difference between the zero field and the applied field data temperature,16) as shown in Fig. 3. The intrachain spin
ð1Þ
from the inset of Fig. 2(b). Following these findings, the exchange has two superexchange paths, J (Cu–O–Cu) and
ð2Þ
intra
(Cu–Cl–Cu), and the interchain spin exchange has two
anomaly at 70 K is interpreted as an antiferromagnetic
transition. The field dependence of the critical temperature kinds of Cu–O–Cu superexchange paths, J (103.8 ) and
J
intra
ð1Þ
ꢂ
inter
(122.6 ). From these data, it is difficult to conclude that
Cu OCl can be regarded as a quasi-1D material because
2 2
ð2Þ
ꢂ
in CuGeO is known to be an indicator of the dimerization
3
J
inter
21)
(
spin-Peierls transition).
In the case of Cu2OCl2, the three neighboring Cu–Cu there is little difference between the intrachain and inter-
˚
distances have similar values—3.04, 3.06, and 3.41 A—and chain exchanges, even if the Cu–O–Cu angles are consid-
ꢂ
the Cu–O–Cu angles are 102.9, 103.8, and 122.6 at room ered. Thus we tried to fit the magnetic susceptibility (ꢃ) to
123705-2

J. Phys. Soc. Jpn., Vol. 75, No. 12
LETTERS
H. OKABE et al.
1
5
2
0
0
0
μ0H = 0.01 T
(a)
Total
Cu
Cl
(b)
Total
Cu
Cl
1
O
10
5
O
1
0
0
.8
.6
χ1D Jintra/k = 93.1(4) K
B
0
χMF Jintra/k = 112.6(1) K
B
Jinter/k = 108(9) K
B
-5
0
50
100
150
T (K)
200
250
300
-10
-20
-
-
10
Fig. 4. (Color online) Magnetic susceptibility ꢃðTÞ of Cu2OCl2 in field of
0
.01 T fitting by the equation for S ¼ 1=2 1D Heisenberg model (see text).
15
The solid and broken lines represent the fitting results of equations (1)
and (2), respectively.
-8 -6 -4 -2
0
2
4
6
8
-8 -6 -4 -2
0
2
4
6
8
E (eV)
E (eV)
Fig. 5. (Color online) Total and partial spin-polarized densities of states
of Cu2OCl2, calculated by LSDA (a) and LSDA þ U (b) methods. The
dotted lines denote the Fermi level.
the following equations for the classical S ¼ 1=2 linear chain
model (1) and the mean field corrected chain model (2),
4,25)
2
which involves the interaction between chains:
closer to the limit of magnetic frustration, such that the
magnetic frustration would be more important than the low-
dimensional magnetism in Cu OCl . A similar situation is
2
2
B
2
Ng ꢀ
A þ BX þ CX
ꢃ1D ꢀ ꢃ ¼
ꢃ
;
ð1Þ
ð2Þ
0
2
3
kBT
1 þ DX þ EX þ FX
2
2
ꢃ1D
realized in the temperature dependence of the magnetic
0
ꢃMF ꢀ ꢃ ¼
;
0
27)
28)
2
zJinterðꢃ1D ꢀ ꢃ0Þ
susceptibility of Cu4O3 and Ag2Cu2O3; these structures
consist of alternating chains of edge-shared square planar
CuO4. However, these materials have smaller Jinter=Jintra
1
ꢀ
2
2
B
Ng ꢀ
where X ¼ jJintraj=kBT, ꢃ0 and ꢃ0 are temperature-inde- values (0.36 for Cu O and 0.21 for Ag Cu O ) than that of
0
4
3
2
2
3
pendent terms (sum of the diamagnetic and van Vleck Cu OCl , so that these materials are affected more by the
2
2
paramagnetic parts), g is the powder-averaged g-factor, Jintra low-dimensional magnetic feature than the magnetic frus-
2
6)
and Jinter are the effective intrachain and interchain exchange tration. Another example of the pyrochlore-like corner-
2
þ
constants, and z is the number of nearest neighbor chains sharing Cu tetrahedral material is the mineral clinoataca-
(
2
9)
z ¼ 2 in this case). The coefficients A ¼ 0:25, B ¼ 0:14995, mite Cu Cl(OH) . Cu Cl(OH) transits into a disordered
2
3
2
3
C ¼ 0:30094, D ¼ 1:9862, E ¼ 0:68854, and F ¼ 6:0626 spin-glass-like structure at a low temperature (6.4 K) with a
were obtained from the theoretical calculation for the S ¼ slight drop in entropy (0:31R ln 2), suggesting the geometri-
24)
1
=2 Heisenberg system. The results are shown by the lines cally frustrated spin system to be pyrochlores. Cu Cl(OH)3
2
in Fig. 4. The solid line indicates the best fitting result has a 3D structure consisting of an alternating edge-shared
[
Jintra=kB ¼ 93:1ð4Þ K, g ¼ 2:00 (fixed), ꢃ0 ¼ 4:27ð2Þ ꢁ CuO and CuO Cl octahedron; it is quite different from
6
4
2
ꢀ4
10
without interchain interaction, and the broken line indicates Consequently, the magnetism of Cu OCl , should be
emu/Cu-mol] by the S ¼ 1=2 linear chain model Cu OCl , which consists of edge-shared square chains.
2
2
2
2
the best fitting result [J =k ¼ 112:6ð1Þ K, J =k ¼ regarded as a geometrically frustrated system induced by
intra
B
inter
B
0
ꢀ4
1
08ð9Þ K, g ¼ 2:00ð6Þ, ꢃ ¼ 3:2ð2Þ ꢁ 10 emu/Cu-mol] by the pyrochlore-like corner-shared tetrahedral structure and
0
the mean field corrected chain model with interchain not a low-dimensional system.
interaction. We only used the data above 85 K to fit, because
Figure 5 illustrates the spin-polarized total and partial
these equations are not valid for a long range ordered density of states (DOS) of Cu OCl using the LSDA (a) and
2
2
magnetic state. The ꢃ obeys the mean field corrected chain LSDA þ U (b) procedures. Using the conventional spin-
model instead of the classical linear chain model, as shown polarized calculations of LSDA, we found that Cu OCl was
2
2
in Fig. 4. This means that Cu2OCl2 tends to be a 3D in a metallic ground state, although the experimental results
antiferromagnetic system with a strong intrachain interac- indicated that it was a good insulator. To understand the
tion, instead of a S ¼ 1=2 1D antiferromagnetic Heisenberg magnetic insulating ground states, we took into account the
linear chain system. Indeed, J =k and J =k show a on-site Coulomb interaction in the LSDA calculation. The
intra
B
inter
B
closed value within the standard deviation [112.6(1) and parameter U for the Coulomb energy and J for the Hund
d
H
1
08(9) K]. The increase in the interchain interaction for the exchange were chosen to be 6.5 and 0.98 eV, respectively,
intrachain interaction prevents their most stable spin states; which were obtained from a reasonable value for copper
3
0)
thus, the extent of the magnetic frustration would be strongly oxides. In the LSDA þ U result [Fig. 5(b)], the occupied
2
6)
enhanced if the Jinter=Jintra value is increased toward 1. In and unoccupied bands are separated by ꢄ2 eV, and the
the opposite case, i.e., the Jinter=Jintra value is reduced toward Fermi level lies closer to the valence band. The occupied
0
, the magnetic structure would be dominated by the parts of the Cu 3d band are strongly mixed with the O 2p
intrachain interaction, thereby inducing a low-dimensional and Cl 2p bands. On the other hand, the lowest-lying
magnetic feature. The best fitted Jinter=Jintra value (ꢄ0:96) is unoccupied bands are mainly of the Cu 3d character. This
123705-3

J. Phys. Soc. Jpn., Vol. 75, No. 12
LETTERS
H. OKABE et al.
clearly indicates that Cu OCl2 is a charge-transfer-type Quantum Materials’’ (No. 16076212) and the ‘‘High-Tech
2
insulator, typical of late 3d transition-metal oxides, i.e., the Research Center’’ project for private universities from the
parent materials of high-Tc superconductors. However, we Ministry of Education, Culture, Sports, Science and Tech-
should not eliminate a possibility that the degree of orbital nology of Japan.
freedom plays an important role in Cu2OCl2 because there
is a case of KCuF3, which indicates a low-dimensional
magnetism due to orbital-ordering by the cooperative Jahn-
1
) A. P. Ramirez: Annu. Rev. Mater. Sci. 24 (1994) 453.
Teller distortion.3 In the case of Cu OCl , the Cu ions are
located at the center of the distorted square plane formed by
1)
2
2
2) S. T. Bramwell and M. J. P. Gingras: Science 294 (2001) 1495.
3) M. J. Harris, S. T. Bramwell, D. F. McMorrow, T. Zeiske and K. W.
Godfrey: Phys. Rev. Lett. 79 (1997) 2554.
14)
two Cl and two O ions; therefore, the crystal field does not
split the 3d orbitals into the usual t2g triplet and an excited eg
4
)
S. Yoshii, S. Iikubo, T. Kageyama, K. Oda, Y. Kondo, K. Murata and
M. Sato: J. Phys. Soc. Jpn. 69 (2000) 3777.
2
2
doublet. Instead, the x –y orbital slightly degenerates with
the xy, yz, and zx orbitals such that the hole carrier cannot be
adequately confined within the edge-sharing CuO2Cl2 square
chain even if carrier doping is possible. However, if one can
dope holes into the CuO Cl square chain, the bond angles
5
)
T. Takeda, R. Kanno, Y. Kawamoto, M. Takano, F. Izumi, A. W.
Sleight and A. W. Hewat: J. Mater. Chem. 9 (1999) 215.
6) Y. Shimikawa, Y. Kubo and T. Manako: Nature 379 (1996) 53.
7
8
9
)
)
)
M. Hanawa, Y. Muraoka, T. Tayama, T. Sakakibara, J. Yamaura and
Z. Hiroi: Phys. Rev. Lett. 87 (2001) 187001.
A. B. Harris, A. J. Berlinsky and C. Bruder: J. Appl. Phys. 69 (1991)
2
2
ꢂ
will be much smaller (102.9 ) than those of the CuO square
2
5200.
B. Canals and C. Lacroix: Phys. Rev. Lett. 80 (1998) 2933.
ꢂ
2
2
plane (ꢄ180 ) such that the two neighboring Cu x –y
orbitals rarely couple via the same ligand orbital (2px or 2py
of O or Cl). They cannot be expected to provide the low-
dimensional transport properties such as the CuO2 plane
in cuprate superconductors. Nevertheless, Cu2OCl2 should
be treated as a 3D material that has the potential to combine
the pyrochlore-like geometrical frustration and the strong
correlation among the Cu 3d electrons.
10) S. Inagaki and R. Kubo: Int. J. Magn. 4 (1973) 139.
11) D. Yanagishima and Y. Maeno: J. Phys. Soc. Jpn. 70 (2001) 2880.
12) H. Fukazawa and Y. Maeno: Phys. Rev. B 67 (2003) 054410.
13) A. Scacchi: Rend. R. Accad. delle Scienze Fis. eMat. di Napoli 9
(
1870) No. 5, 86.
14) R. Arpe and H. M u¨ ller-Buschbaum: Z. Naturforsch. B 32 (1977) 380.
15) S. Filatov and L. Vergasova: Mineral. Zh. 5 (1983) No. 3, 84.
1
6) S. Krivovichev and S. Filatov: Can. Mineral. 40 (2002) 1185.
7) P. Blaha, K. Schwarz, G. K. H. Madsen, K. Kvasnicka and J. Luitz:
WIEN2k, An Augmented Plane Wave + Local Orbitals Program for
Calculating Crystal Properties (Karlheinz Schwarz, Tech. Universit a¨ t
Wien, Austria, 2001).
1
In this study, we have investigated the magnetic properties
of Cu2OCl2 which has a pyrochlore-like corner-shared
tetrahedral network with S ¼ 1=2. The magnetic suscepti-
bility exhibits a low-dimensional feature; however, it does
not follow the typical S ¼ 1=2 linear chain model. We
suggest that the magnitude of the interchain interaction is
close to that of the intrachain interaction; thus, Cu OCl
18) V. I. Anisimov, I. V. Solovyev, M. A. Korotin, M. T. Czyzyk and
G. A. Sawatzky: Phys. Rev. B 48 (1993) 16929.
1
9) J. P. Perdew, K. Burke and M. Ernzerhof: Phys. Rev. Lett. 77 (1996)
865.
3
2
2
2
0) U. Asaf, D. Hechel and I. Felner: Solid State Commun. 98 (1996) 571.
1) M. Hase, I. Terasaki and K. Uchinokura: Phys. Rev. Lett. 70 (1993)
3651.
should be regarded as a pyrochlore-like spin frustrated
material rather than a low-dimensional magnetic material.
From the results of the band calculation results, we found
that the Cu 3d on-site Coulomb interaction is essential for
describing the electronic structure of Cu2OCl2, i.e., Cu2OCl2
is a potential material with the pyrochlore-like geometrical
frustrated state and a strongly correlated electron state.
If Cu OCl maintains the potential for instability of the
2
2
2) M. Isobe and Y. Ueda: J. Phys. Soc. 65 (1996) 1178.
3) A. Seidel, C. A. Marianetti, F. C. Chou, G. Ceder and P. A. Lee:
Phys. Rev. B 67 (2003) 020405.
2
2
2
4) W. E. Hatfield: J. Appl. Phys. 52 (1981) 1985.
5) J. N. McElearney, S. Merchant and R. L. Carlin: Inorg. Chem. 12
(1973) 906.
2
2
6) M. H. Whangbo and H. J. Koo: Inorg. Chem. 41 (2002) 3570.
7) L. Pinsard-Gaudart, J. Rodr ´ı guez-Carvajal, A. Gukasov and P. Monod:
Phys. Rev. B 69 (2004) 104408.
2
2
magnetic interaction even at low temperatures, it might be
premature to conclude that a long-range magnetic order
appears at 70 K. It is highly desirable that NMR, ꢀSR, and
more detailed diffraction studies be performed to understand
the characteristics of the magnetic structure.
2
8) E. M. Tejada-Rosales, J. Rodr ´ı guez-Carvajal, N. Casa n˜ -Pastor, P.
Alemany, E. Ruiz, M. S. El-Fallah, S. Alvarez and P. Gomez-Romero:
´
Inorg. Chem. 41 (2002) 6604.
2
3
3
9) X. G. Zheng, T. Kawae, Y. Kashitani, C. S. Li, N. Tateiwa, K. Takeda,
H. Yamada, C. N. Xu and Y. Ren: Phys. Rev. B 71 (2005) 052409.
0) V. I. Anisimov, J. Zaanen and O. K. Andersen: Phys. Rev. B 44
Acknowledgments
(1991) 943.
1) S. Kadota, I. Yamada, S. Yoneyama and K. Hirakawa: J. Phys. Soc.
Jpn. 23 (1967) 751.
This work was financially supported by a Grant-in-Aid for
Scientific Research in priority area ‘‘Invention of Anomalous
123705-4