Intercalation of a nitronyl nitroxide radical into layered inorganic hosts. Preparation and physico-chemical characterization

Article abstract of DOI:10.1016/S0020-1693(02)01026-5

The synthesis of new intercalation compounds obtained introducing the organic radical 2-(4-carboxyphenyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazol-1-oxyl-3- oxide (NITpBAH) in layered inorganic hosts via topotactic anion exchange reactions or exchange of anionic ligands is presented. Three different inorganic hosts were employed, two LDH-phases (Mg-Al-CO₃ and Zn-Al-Cl) and the butanoxy derivative of λ-zirconium phosphates. In all cases the intercalation procedure was successfully accomplished. The percentage of exchanged anions obtained was 27, 45, 40% for Mg-Al-LDH, Zn-Al-LDH and λ-ZrPO₄[O(CH₂)₃CH₃]DMSO, respectively. The magnetic properties of the intercalated compounds were measured with a SQUID magnetometer and compared to those of a polycrystalline powder of the pure radical. The intercalation process does not change the nature of magnetic intermolecular interactions which remain antiferromagnetic but reduces their magnitude.

Full text of DOI:10.1016/S0020-1693(02)01026-5

Inorganica Chimica Acta 338 (2002) 127Á132
/
www.elsevier.com/locate/ica
Intercalation of a nitronyl nitroxide radical into layered inorganic
hosts.
Preparation and physico-chemical characterization
A. Caneschi a, D. Gatteschi a,*, C. Sangregorio a, M.G.F. Vaz a, U. Costantino b,
M. Nocchetti b, R. Vivani b
a Department of Chemistry, University of Florence, via della Lastruccia 3, 50019 Sesto F.no. FI, Italy
b Department of Chemistry, University of Perugia, via Elce di Sotto 8, 06123 Perugia, Italy
Received 31 December 2001; accepted 8 April 2002
Abstract
The synthesis of new intercalation compounds obtained introducing the organic radical 2-(4-carboxyphenyl)-4,4,5,5-tetramethyl-
4,5-dihydro-1H-imidazol-1-oxyl-3-oxide (NITpBAH) in layered inorganic hosts via topotactic anion exchange reactions or
exchange of anionic ligands is presented. Three different inorganic hosts were employed, two LDH-phases (Mgꢀ
AlꢀCl) and the butanoxy derivative of l-zirconium phosphates. In all cases the intercalation procedure was successfully
accomplished. The percentage of exchanged anions obtained was 27, 45, 40% for MgꢀAlꢀLDH, ZnꢀAlꢀLDH and l-
/
Alꢀ
/
CO3 and Znꢀ
/
/
/
/
/
/
ZrPO4[O(CH2)3CH3]DMSO, respectively. The magnetic properties of the intercalated compounds were measured with a SQUID
magnetometer and compared to those of a polycrystalline powder of the pure radical. The intercalation process does not change the
nature of magnetic intermolecular interactions which remain antiferromagnetic but reduces their magnitude.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Nitronyl nitroxide radicals; Magnetism; Intercalation compounds
1. Introduction
devices [2] and molecular recognition [3] can be found in
the recent literature.
The ability of many layered solid structures to
incorporate guest species in a constrained environment
represents a useful tool for materials chemistry. By
intercalation procedures it is possible to prepare materi-
als in which a species having peculiar properties is
present within the layers of a host in an ordered array
that is usually different from that found in a pure phase
From this point of view, we found it of interest to
study the change of the properties of magnetic organic
molecules when intercalated into lamellar hosts. The
guest species selected for this study was a substituted a-
nitronyl nitroxide radical. This class of compounds has
been extensively investigated since the discovery that
some members of the family undergo a transition to a
ferromagnetic ordered state at low temperature [4]. The
nature and the magnitude of the intermolecular inter-
actions which lead to the magnetic ordering depend on
the crystal packing and on the relative orientation of the
molecules. For example, the p-nitrophenyl nitronyl
nitroxide radical crystallizes in four different phases of
which only one, the b phase, orders ferromagnetically
[4]. The intercalation of these compounds into layered
inorganic hosts is expected to change the intermolecular
arrangement with respect to those of the pure solid
phases therefore affecting the magnetic properties of the
or a solution. Due to hostÁ
/
and guestÁguest interac-
/
tions, the properties of the material can be modified for
specific purposes. Some studies on the effects of inter-
calation into layered hosts (e.g. double hydroxides and
zirconium phosphates and phosphonates) of species of
interest in the field of non-linear optics [1], sensing
* Corresponding author. Tel.: ꢀ
4573372
E-mail address: dante.gatteschi@unifi.it (D. Gatteschi).
/
39-055-4573285; fax: ꢀ39-055-
/
0020-1693/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 0 - 1 6 9 3 ( 0 2 ) 0 1 0 2 6 - 5

128
A. Caneschi et al. / Inorganica Chimica Acta 338 (2002) 127Á132
/
compounds. Among the different organic radicals which
have been synthesized in the last decade, we chose the 2-
(4-carboxyphenyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-
imidazol-1-oxyl-3-oxide (NITpBAH in the following)
[5]. Bearing a carboxylic group, the above compound
was expected to be one of the best candidates for
intercalation in layered hosts via topotactic anion
exchange reactions or exchange of anionic ligands.
Among inorganic layered solids, very few possess
anionic exchange capabilities. LDH are the most known
and commonly employed inorganic anion exchangers
[6]. Their general formula is M(II)1ꢁxM(III)x(OH)2-
ethyl alcohol, once whit CO2-free distilled water and
finally dried at room temperature (r.t.) (75% R.H.).
Chemical analysis gave the following composition:
[Mg0.67Al0.33(OH)2](NITpBA)0.09(CO3)0.12
TGA: % weight loss at 1000 8C: 54.5 (calc. 54.7,
assuming the formation of 0.505 MgOꢀ0.165 MgAl2O4
at 1000 8C).
ZnꢀAlꢀLDH NITpBA derivative (II) was prepared
as follows: 1 g of [Zn0.67Al0.33(OH)2]Cl0.33 0.6H2O was
×
/
0.3H2O.
/
/
/
×
/
equilibrated, for 1 day at r.t., with 180 ml of 0.04 M of
hydro-alcoholic (50% v/v) solution of NITpBAH. The
pH of this solution was previously set to 8, by NaOH
addition. The solid was then separated by centrifugation
and washed once with ethyl alcohol, once whit CO2-free
distilled water and finally dried at r.t. (75% R.H.).
Chemical analysis gave the following composition:
[Anꢁ
]
×
/
mS, where M(II) may be Mg, Zn, Co, Ni;
x/n
M(III) Al, Cr, Fe, V; Anꢁ is the charge compensating
ꢁ
ꢁ
anion, CO3 , SO4 , Clꢁ, organic anions; m the number
of co-intercalated solvent S, generally water, per for-
mula weight of compound. The layerÁ
actions are basically ionic.
/
counterion inter-
[Zn0.67Al0.33(OH)2](NITpBA)0.15(Cl)0.18
% weight loss at 1000 8C: 51 (calc. 50.3, assuming the
formation of 0.505 ZnOꢀ0.165 ZnAl2O4 at 1000 8C).
The hosts, [Mg0.67Al0.33(OH)2](CO3)0.165 0.4H2O and
[Zn0.67Al0.33(OH)2](CO3)0.165 0.4H2O were prepared
×
/
0.5H2O. TGA:
Recently, in the Perugia laboratory a new family of
anionic ligand exchangers has been developed, that is l-
zirconium phosphates. Although the first synthesized
and structurally characterized l-type compound was
ZrPO4FDMSO [7], the first compound of this class
showing anion exchange capabilities is ZrPO4ClDMSO
[8], in which chloride can be easily replaced, by a
topotactic anion exchange reaction, with other organic
or inorganic monoanionic ligands. In l-ZrP derivatives
the exchangeable anionic ligands are covalently bonded
to zirconium atoms, and hence they occupy a specific
site in the layer structure.
/
×
/
×
/
with a procedure accomplished by the thermal hydro-
lysis of urea [6]. Before NITpBAH intercalation, the
carbonate form of Znꢀ
/
Alꢀ/LDH was converted into the
chloride form via ion exchange at pH 5. The solid was
washed with CO2-free distilled water and dried at r.t.
over P4O10.
l-ZrP NITpBA derivative (III) was obtained as
follows: 0.3 g l-ZrPO4[O(CH2)3CH3]DMSO were dis-
persed in 54 ml of 0.05 M NITpBAH solution and
maintained under stirring at 90 8C for 2 days. The solid
was then separated by centrifugation and washed twice
with 50 ml dioxane each time. The compound obtained
presented the same violet color of NITpBAH solution,
even after washing. Chemical analysis gave the following
composition: ZrPO4(NITpBA)0.4(OH)0.6DMSO. Anal.
Found (Calc.): C, 23.78 (24.04); N, 2.64 (2.11); H, 3.64%
(3.01). TGA: % weight loss at 1000 8C: 49.00 (calc.
48.85, assuming the formation of (ZrP2O7)0.5(ZrO2)0.5 at
1000 8C).
This paper reports the preparation and physico-
chemical characterization of NITpBAH intercalation
compounds with both classes of inorganic layered hosts
described above; specifically, two LDH-phases (Mgꢀ
/
Alꢀ
/
CO3 and Znꢀ
/
Alꢀ
/
Cl LDH), and the butanoxy
derivative of l-zirconium phosphate were employed.
2. Experimental
2.1. Synthesis
The host, l-ZrPO4[O(CH2)3CH3]DMSO, was pre-
pared as follows: 1 g of l-ZrPO4ClDMSO (prepared
as previously reported [8]) was suspended in 85 ml of a
0.1 M n-Bu3N in BuOH as solvent for 5 days at 80 8C.
The solid was washed with BuOH and dried at 70 8C.
The organic radical 2-(4-carboxyphenyl)-4,4,5,5-tetra-
methyl-4,5-dihydro-1H-imidazol-1-oxyl-3-oxide was
synthesized according to the procedure previously
described [9].
Mgꢀ
as follows: the hydrotalcite [Mg0.67Al0.33(OH)2]-
(CO3)0.165 0.4H2O was calcined in air at 500 8C for 18
/
Alꢀ
/
LDH (NITpBA) derivative (I) was prepared
2.2. Analytical and instrumental procedures
×
/
h. One gram of the calcined oxides thus obtained was
suspended in 80 ml of CO2-free distilled water at
refluxing temperature for 45 min. A solution obtained
by dissolving 2.08 g of NITpBAH in 100 ml of ethyl
alcohol was added to the above cold suspension of re-
hydrated oxides. The suspension was stirred for 1 day.
After the equilibration the solid was washed once with
C, N and H elemental analysis was obtained by a C.
Erba 1106 Analyzer.
Mg(II), Zn(II) and Al(III) content was determined as
follows. Weighed amounts of the samples (approxi-
mately 100 mg) were dissolved in a few drops of concd.
HCl and diluted with water to 50 ml. These solutions
were then titrated with standard EDTA.

A. Caneschi et al. / Inorganica Chimica Acta 338 (2002) 127Á
/
132
129
TGA measurements were performed with a Stanton
Redcroft STA780 thermoanalyzer. Ion Chromatogra-
phy was carried out with a Dionex 2000 i/sp instrument,
AS4A column and 1.7ꢂ
/
10ꢁ3 M NaHCO3/1.8ꢂ10ꢁ3
/
M Na2CO3 solution as eluent.
X-ray powder diffraction (XRD) measurements were
used for phase identification and for monitoring the
evolution of topotactic reactions of insertion of NITp-
BAH into the hosts. It was accomplished by evaluating
the change of the interlayer distance, determined from
the first strong reflection of the XRD patterns, recorded
with a Philips X’PERT diffractometer using the Cu Ka
radiation.
J-modulated 13C liquid NMR spectra were taken with
a Bruker DPX 200: about 50 mg of solid sample were
dissolved in 1 ml of 6 M HF using a proper deuterated
solvent.
Fig. 2. Temperature dependence of xT measured at 1 T on a
polycrystalline sample of NITpBAH. The full line represents the fit
to Eq. (1).
Static magnetic susceptibilities were measured with a
Metronique Ingenie´rie MS02 SQUID magnetometer
with an applied field of 1 T. Magnetic data were
corrected for diamagnetic contributions which were
directly measured on the non-intercalated inorganic
hosts.
discussed by some authors which reported different
behaviors and used different models to reproduce
experimental data (dimer model [5], chain model [9]
and 2D square lattice model [10]). We found the best
model to fit our data was the Bonner and Fisher model
for antiferromagnetic chains of Sꢃ1/2 Heisenberg spins
[11]:
/
EPR spectra were recorded on a Varian E-9 spectro-
meter working at X-band (9.25 GHz) and equipped with
an He flux cryostat ESR9 built by Oxford Instruments,
which allows temperature variation from 4 K to r.t.
Ng2m2B
xT ꢃ
kB
0:25 ꢀ 0:074975x ꢀ 0:075235x2
1:0 ꢀ 0:9931x ꢀ 0:172135x2 ꢀ 0:757825x3
ꢀ
(1)
3. Results and discussion
where xꢃ
/
jJj/kBT. The best fit parameters we found
/
were gꢃ
/
2.03(1) and Jꢃ
/
3.38(8) cmꢁ1 with Rꢃ
/
6.9ꢂ
Fig. 1 shows the structure formula and the approx-
imate dimensions of NITpBAH. In Fig. 2 the tempera-
ture dependence of the xT product of the pure radical is
shown. A continuous decrease of xT upon decreasing
temperature is observed indicating the presence of
intermolecular antiferromagnetic interactions. The mag-
netic properties of the compound were previously
10ꢁ3. The data are in good agreement with those
reported in Ref. [9] (the g value is larger than 2 due to
an experimental artifact) and can be explained on the
basis of a magnetic interaction operating through the
chain-like structure formed by hydrogen bonds between
the oxygen atom of the nitronyl nitroxide groups and
the carboxylic groups, where a small amount of spin
density is expected to be delocalized.
The procedures used to intercalate NITpBAH and the
products obtained for the different hosts, together with
their magnetic properties, will be discussed separately in
the following.
3.1. LDH-NITpBA derivatives
A powerful synthetic route for inserting inorganic and
organic anions into LDH is the so-called memory effect,
that is the ability of Mgꢀ
/
Alꢀ
/
CO3ꢀ
/
LDH calcined at
500 8C to reconstruct its layered structure in the
presence of water and anions [12].
NITpBAꢁ was first intercalated into Mgꢀ
using this effect. The mixture of magnesium aluminium
oxides formed after heat treatment was regenerated in
/
Alꢀ
/
LDH
Fig. 1. Structure formula of NITpBAH. Van der Waals dimensions
are also shown.

130
A. Caneschi et al. / Inorganica Chimica Acta 338 (2002) 127Á132
/
CO2-free distilled water, obtaining the following com-
pound: [Mg0.67Al0.33(OH)2](OH)0.33 nH2O, in which the
balancing anions are hydroxyl groups. In the presence of
NITpBAH an acidÁbase reaction takes place and lead
to I where 27% of OH groups where exchanged by
NITpBA anions. The residual hydroxyl groups were
spontaneously converted into carbonate anions by CO2
from air. Owing to the high reactivity of LDH in OH
form and to its strong selectivity towards carbonate, all
our attempts to increase the amount of intercalated
NITpBA were unsuccessful.
molecular interactions arising from the different packing
of the radical in the interlayer are of the same nature and
not as strong as those observed in the free radicals. The
EPR spectrum of the two compounds were recorded at
30 K. The two spectra consist of a single line centered at
×
/
/
gꢃ/2.00 for both samples. The linewidths are 10.5 and
12.5 G for Mgꢀ
/
AlꢀLDH and ZnꢀAlꢀLDH, respec-
/
/
/
tively. Within the resolution limit of our instrument no
significant differences were observed with the spectrum
of the pure radical where a single line with linewidth
12.5 G is present.
XRD patterns showed an interlayer distance of 20.7
A, and the typical reflection of the carbonate form (7.6
˚
3.2. l-ZrP NITpBA derivative
˚
A).
To obtain an increased amount of NITpBAꢁ in the
interlayer region we used a direct anionic exchange
Preliminary experiments to intercalate NITpBAH
directly into l-ZrPO4ClDMSO were unsuccessful, prob-
ably due to the high steric hindrance of the guest species.
To facilitate the topotactic reaction, it was therefore
used, as host, the l-butanoxy derivative, of formula l-
ZrPO4[O(CH2)3CH3]DMSO. This derivative has a lar-
ger interlayer distance than the chloride one (15.7 vs.
reaction path. In this case the host used was Znꢀ
/
Alꢀ
/
LDH in chloride form, and the following reaction was
carried out:
[Zn0:67Al0:33(OH)2]Cl0:33×0:6H2Oꢀ0:15 NITpBAꢁ
0 Zn0:67Al0:33(OH)2](NITpBA)0:15(Cl)0:18×0:5H2O
ꢀ0:15 Clꢁ ꢀ0:1 H2O
˚
10.2 A) and it is more reactive towards the anion
exchange reactions, because of the high basic character
of the alkoxy anions [13]. Using this phase, the inter-
calation of NITpBAꢁ was easier than with the l-
ZrPO4ClDMSO phase, following the scheme:
As indicated from the composition of II about 45% of
the chloride anions were exchanged with NITpBAꢁ.
˚
The interlayer distance was 19.7 A, very close to that
observed for I.
ZrPO4[O(CH2)3CH3]DMSOꢀ0:4 NITpBAH
ꢀ0:6 H2O 0 ZrPO4(NITpBA)0:4(OH)0:6DMSO
ꢀCH3(CH2)3OH
3.1.1. Magnetic properties
The temperature dependence of the xT product per
mol of radical of the Mgꢀ
/
Alꢀ
/
LDH and Znꢀ
/
Alꢀ
/
LDH
Note that only 40% of anions were exchanged by
NITpBA anions, and all our attempts to increase this
amount were unsuccessful. Owing to the presence of
intercalated compounds are shown in Fig. 4(a and b),
respectively. For both compounds xT slowly decreases
on lowering temperature indicating that weak antiferro-
magnetic intermolecular interactions are still present in
four methyl terminal groups, the cross section of
2
NITpBAH was evaluated to be roughly 45 A , a value
˚
both compounds. The data can be fit to a CurieÁ
law with Cꢃ 2.3 K for
0.376 emu K molꢁ1 and uꢃ
MgꢀAlꢀLDH and Cꢃ
0.356 emu K molꢁ1 and uꢃ
3.0 K for ZnꢀAlꢀLDH. For both compounds at
/
Weiss
close to the free cross section per site available in the
2
˚
/
/
ꢁ
/
interlayer region of a l-derivative (43.5 A ). Further-
/
/
/
/
more, the cross section of NITpBA groups is not
uniform, the terminal part being larger than the body;
the extra free room in the interlayer space hence created,
is probably occupied by NITpBA groups coming from
adjacent layers, in an interdigitated arrangement as
depicted in the hypothetical model of Fig. 3. The
interlayer distance derived from this model is consistent
with the experimental interlayer distance for sample III
ꢁ
/
/
/
low temperature a small deviation from this law is
observed, which indicates that low dimensional interac-
tions are operating, as it should be expected. An analysis
of magnetic data considering a dimer, an antiferromag-
netic chain and a 2D square lattice models was
attempted but none of the models could satisfactory
reproduce the experimental data. The failure of the
above models can be the consequence of the incomplete-
ness of the radical intercalation into the layered
structure or of a not fully ordered packing of the radical
which can make the interpretation of the magnetic
behavior more complex.
˚
(18.7 A), when the organic group is tilted by an angle of
748 with respect to the layers plane.
The butanoxy groups not exchanged by NITpBA
anions were not found in the final compound, as
detected by 13C NMR measurements. Preliminary
experiments carried out in pure l-butanoxy derivative
have shown that butanoxy groups can be easily replaced
by hydroxyl groups, because of the strong basic
character of butanoxy anions. Therefore, it is presum-
able that the butanoxy groups not exchanged by
The intercalation of the organic radical in the
inorganic host destroys the main path for exchange
along the hydrogen bond between the carboxylic group
and the nitronyl nitroxide oxygen. However, the inter-

A. Caneschi et al. / Inorganica Chimica Acta 338 (2002) 127Á
/
132
131
NITpBA were replaced by OH groups coming from
water vapour. Chemical and thermogravimetric analysis
are in agreement with this assumption.
3.2.1. Magnetic properties
The xT versus T curve of the compound is shown in
Fig. 5. If compared to the other two compounds a
smaller decrease of xT on lowering temperature is
observed, suggesting even weaker antiferromagnetic
intermolecular interactions. The data could be satisfac-
tory fit either to a CurieÁ
K molꢁ1and uꢃ
1.1 K, by a 2D square lattice model
with Jꢃ
0.63 cmꢁ1 or by the AF chain model with Jꢃ
1.2 cmꢁ1, the best R value (Rꢃ
0.0025) being obtained
/
Weiss law with Cꢃ/0.38 emu
/
ꢁ
/
/
/
/
for the last model. However, due to the small value of
involved magnetic interactions no conclusions on their
dimensionality can be drawn from the data fit. Anyway
the data confirm that the different radical arrangement
in this host results in weaker intermolecular interactions.
The EPR spectrum of this compound recorded at 30
K is similar to that of the other intercalated compounds
and consists of a single line at gꢃ2.00 with linewidth
/
13.5 G.
Fig. 3. Structural model, viewed along two directions perpendicular to
each other, showing the hypothetical arrangement of NITpBAH in the
interlayer region of l-ZrP.
4. Conclusions
The intercalation of the organic radical NITpBAH
into inorganic hosts was successfully accomplished. In
all the investigated compounds intermolecular interac-
tions are still antiferromagnetic but are weaker than
those observed in the free radical. Despite the results
obtained with this radical are on the opposite directions
than that we hoped for, the proposed method has shown
to provide an effective route to modify magnetic
interactions in organic radicals.
Fig. 4. Temperature dependence of the xT product of NITpBAꢀ
AlꢀLDH (a) and NITpBAꢀZnꢀAlꢀLDH (b).
/
Mgꢀ
/
Fig. 5. Temperature dependence of the xT product of l-ZrP NITpBA
derivative.
/
/
/
/

132
A. Caneschi et al. / Inorganica Chimica Acta 338 (2002) 127Á132
/
(b) G. Alberti, M. Casciola, U. Costantino, R. Vivani, Adv.
Mater. 8 (1996) 291.
Acknowledgements
[4] (a) K. Agawa, Y. Maruyama, Chem. Phys. Lett. 158 (1989) 556;
(b) M. Tamura, Y. Nakazawa, D. Shiomi, K. Nozawa, Y.
Hosokoshi, M. Ishikawa, M. Takahashi, M. Kinoshita, Chem.
Phys. Lett. 186 (1991) 401;
This work was supported by MURST under the
national project ‘Molecules for nanostructured func-
tional materials’ 1999 and by EC network 3MD (con-
tract no. ERBFMRX-CT98-0181) and CNR PF MSTA
II. M.G.F.V. gratefully acknowledges for a 1-year grant
the Brazilian Conselho Nacional de Desenvolvimento
Cientifico e Tecnologico (CNPq).
(c) A. Caneschi, F. Ferraro, D. Gatteschi, A. le Lirzin, M.A.
Novak, E. Renstschler, R. Sessoli, Adv. Mater. 7 (1995) 476.
[5] K. Inoue, H. Iwamura, Chem. Phys. Lett. 207 (1993) 551.
[6] (a) F. Trifiro`, A. Vaccari, in: G. Alberti, T. Bein (Eds.), Solid-state
Supramolecular Chemistry: Two and Three-Dimensional Inor-
ganic Networks Comprehensive Supramolecular Chemistry, vol. 7
(Chapter 8), Pergamon, Oxford, 1996;
(b) U. Costantino, F. Marmottini, M. Nocchetti, R. Vivani, Eur.
J. Inorg. Chem. (1998) 1439.
References
[7] D.M. Poojary, B. Zhang, A. Clearfield, J. Chem. Soc., Dalton
Trans. (1994) 2453.
[1] (a) S. Cooper, K. Dutta, J. Phys. Chem. 94 (1990) 114;
(b) R. Hoppe, G. Alberti, U. Costantino, C. Dionigi, G. Schulz-
Ekloff, R. Vivani, Langmuir 13 (1997) 7252;
[8] (a) G. Alberti, M. Bartocci, M. Santarelli, R. Vivani, Inorg.
Chem. 36 (1997) 3574;
(c) G.G. Aloisi, U. Costantino, F. Elisei, M. Nocchetti, C. Sulli,
Mol. Cryst. Liq. Cryst. 311 (1998) 245;
(b) G. Alberti, S. Masci, R. Vivani, Inorg. Chem. 41 (2002) 1913.
[9] N.C. Schiødt, F. Fabrizi de Biani, A. Caneschi, D. Gatteschi,
Inorg. Chim. Acta 248 (1996) 139.
(d) U. Costantino, N. Coletti, M. Nocchetti, G.G. Aloisi, F.
Elisei, L. Latterini, Langmuir 15 (2000) 4454.
[10] J. Cirujeda, C. Rovira, J.-L. Stanger, P. Turek, J. Veciana, in: O.
Kahn (Ed.), Magnetism: A Supramolecular Function NATO ASI
Ser. C, vol. 484, Kluwer, Dordrecht, 1996, p. 219.
[2] (a) G. Alberti, in: G. Alberti, T. Bein (Eds.), Two and Three-
Dimensional Inorganic Networks Comprehensive Supramolecu-
lar Chemistry, vol. 7, Pergamon, Oxford, 1996, pp. 151Á187;
/
(b) A. de Roy, J.P. Besse, Solid State Ionics 46 (1991) 95.
[3] (a) T. Ikeda, H. Amoh, T. Yasunaga, J. Am. Chem. Soc. 106
(1984) 5772;
[11] J.C. Bonner, M.E. Fisher, Phys. Rev. A 135 (1964) 640.
[12] F. Rey, V. Fornes, J. Chem. Soc., Faraday Trans. 88 (1992) 2233.
[13] G. Alberti, S. Masci, R. Vivani, to be published.