determined by the shorter l
crown ethers in the salt 1 is l
longer than l
2
. The distance between the adjacent
1
= 8.13 Å, which is quite a bit
2
.
In conclusion, we have constructed a novel molecular spin
ladder using a spool-type supramolecular cation. According to
theory,10 hole doping to two-leg spin ladders will lead to
superconductivity due to that effective attraction between extra
holes may arise from the magnetic interactions. In fact,
superconducting transitions have been reported in hole-doped
spin ladders of calcium cuprates.11 In the case of salt 1, the hole
doping will be accomplished without significant lattice distor-
tion by replacing a fraction of the ammonium cations with the
corresponding amines. Studies on this are in progress.
The authors thank Dr K. Ichimura and Professor K. Nomura
for the use of the SQUID magnetometer. This work was partly
supported by a Grant-in-Aid for Science Research from the
Ministry of Education, Culture, Sports, Science and Technol-
ogy of Japan.
Fig. 2 Temperature dependence of the magnetic susceptibility of salts 1 (1)
and 2 (8). The solid lines are fitted curves of eqn. (1) ( < 80 K) and eqn. (3)
(
> 150 K) (see text), and the dashed line is a fit by eqn. (4).
Notes and references
Fig. 2 shows the temperature dependence of the magnetic
†
Crystal data: for 1: C24
H
32NO
6
S10Ni, M = 809.82, T = 296 K,
susceptibility (c) of the salts 1 and 2 after subtracting the Curie
¯
2
1
triclinic, space group P1, a = 14.217(1), b = 16.666(2), c = 8.1271(6) Å,
component, C = 0.0060 and 0.0063 emu K mol , for salts 1
and 2, respectively. The magnetic susceptibility was measured
on a Quantum Design MPMS-XL SQUID susceptometer in a
field of 1 T. The susceptibility of salt 1 had a maximum at 120
3
c
a = 97.120(2), b = 92.599(3), g = 112.846(2)°, V = 1751.6(3) Å , D =
23
21
1
=
7
.535 g cm , Z = 2, F(000) = 838, m(Mo-Ka) = 11.87 cm , final R, R
0.051, 0.118. I
594 unique. GOF = 1.69.
w
o
= 5677 ‘observed’ [I > 3.0s(I)] reflections out of N =
2
1
K, and reached 0 emu mol
temperature-limit equation of a two-leg spin ladder,9
Low = aT21/2exp(2D/T)
around 15 K. The low-
For 2: C46
H
64
O
12
N
4
S
20Ni
2
, M = 1623.63, T = 296 K, triclinic, space
group P 1¯ , a = 13.1321(4), b = 16.3032(5), c = 17.6903(5) Å, a =
3
73.5786(9), b = 76.2712(8), g = 82.426(1)°, V = 3520.6(2) Å , D
c
=
c
(1)
23
21
1
.531 g cm , Z = 2, F(000) = 1680, m(Mo-Ka) = 11.82 cm , final R,
reproduced the observed susceptibility below 80 K with the spin
gap D = 190 K. Since the gap is related to J and JA as,
Rw = 0.055, 0.094. Io = 8095 ‘observed’ [I > 3.0s(I)] reflections out of N
= 16043 unique. GOF = 1.11. Crystallographic data collected on a Rigaku
RAXIS-RAPID Imaging Plate. All H atoms placed in calculated posi-
tions.
1
2
DªJ' -J + J / J'
(2)
2
CCDC reference numbers 176980 and 176981. See http://www.rsc.org/
suppdata/cc/b1/b110368g/ for crystallographic data in CIF or other
electronic format.
we can estimate the magnetic exchange energies J = 20 K and
J’ = 200 K by assuming JA/J = 10. The magnetic susceptibility
at higher temperatures ( > 150 K) is well reproduced by the
high-temperature equation,9
1
P. W. Anderson, Science, 1987, 235, 1196.
2 E. Dagotto and T. M. Rice, Science, 1996, 271, 618; D. J. Scalapino,
Nature, 1995, 377, 12; Z. Hiroi and M. Takano, Nature, 1995, 337,
È
Î
-1 1 Ê
1
ˆ
-2
3
-3˘
4
1.
cHigh = C T
-
J + J¢ T
+
JJ¢T
(3)
Í
˙
Ë
¯
3 C. Rovira, J. Veciana, E. Ribera, J. Tarrés, E. Candell, R. Rousseau, M.
Mas, E. Molins, M. Almeida, R. T. Henriques, J. Morgado, J. P.
Schoeffel and J. P. Pouget, Angew. Chem., Int. Ed. Engl., 1997, 109,
2418; H. Imai, T. Inabe, T. Otsuka, T. Okuno and K. Awaga, Phys. Rev.
B, 1996, 54, R6338; T. Komatsu, N. Kojima and G. Saito, Solid State
Commun., 1997, 103, 519; M. Fourmigué, B. Domerq, I. V. Jourdain, P.
Molinié, F. Guyon and J. Amaudrut, Chem. Eur. J., 1988, 4, 1714.
2
2
16
˚
using these parameters and C = 0.394 emu K mol2 determined
1
from EPR.
In the case of salt 2, c increased with the decrease in
temperature down to 210 K, and decreased exponentially at
lower temperatures. A fitting by the singlet–triplet thermal
excitation model,
4
5
C. Rovira, Chem. Eur. J., 2000, 6, 1723.
N. Takamatsu, T. Akutagawa, T. Hasegawa, T. Nakamura, T. Inabe, W.
Fujita and K. Awaga, Inorg. Chem., 2000, 39, 870; T. Akutagawa, S.
Nishihara, N. Takamatsu, T. Hasegawa, T. Nakamura and T. Inabe, J.
Phys. Chem. B, 2000, 104, 5871.
C
4exp(-2J / T)
cSinglet-Triplet
=
(4)
T 1+ 3exp(-2J / T)
6
7
E. Weber, L. J. Toner, I. Goldberg, F. Vögtle, A. D. Laidler, F. J.
Stoddart, A. R. Bartsch and L. C. Liotta, Crown Ethers and Analogs, S.
Patai and Z. Rappoport, ed., John Wiley & Sons, New York, 1989.
G. C. Pimentel and A. L. McClellan, The Hydrogen Bond,Freemann,
San Francisco, 1960.
using C = 0.394 emu K mol2 gave a good result, and the
magnetic exchange interaction 2J was around 340 K. The ratio
of JA/J = 43 may be too large to realize the spin-ladder state in
salt 2. It should be noted that eqn. (4) does not give a good fit in
the case of salt 1.
1
8 T. Mori, A. Kobayashi, Y. Sasaki, H. Kobayashi, G. Saito and H.
Inokuchi, Bull. Chem. Soc. Jpn., 1984, 57, 627.
9 M. Troyer, H. Tsunetsugu and D. Würtz, Phys. Rev. B, 1994, 50,
The stronger face-to-face p–p interaction within the [Ni(d-
2
mit)
2
]
dimer reflects the larger value of JA/J in salt 2. The
1
3 515.
0 T. M. Rice, S. Gopalan and M. Sigrist, Europhys. Lett., 1993, 23,
45.
1 M. Uehara, T. Nagata, J. Akimitsu, H. Takahashi, N. Mori and K.
Kinoshita, J. Phys. Soc. Jpn., 1996, 65, 2764; J. H. Schön, M. Dorget,
F. C. Beuran, X. Z. Xu, E. Arushanov, M. Lagues and C. Deville
Cavellin, Science, 2001, 293, 2430.
2
interaction depends on the distance between the [Ni(dmit) ]
2
1
1
units forming the dimer, which should be related to the distance
between crown ethers. The distances between the 18-crown-6
4
2+
molecular planes sandwiching the [p-Ph(NH
are l = 7.49 and l = 8.80 Å (Fig 1). Since each dimer is on
3 2 3
both of [p-Ph(NH ) ] and CH CN, the interaction should be
3
)
2
]
3
and CH CN
2
3
2+
CHEM. COMMUN., 2002, 408–409
409