Coordination Chemistry Reviews

181 (1999) 61 – 120

Layered double hydroxides (LDH) intercalated

with metal coordination compounds and
oxometalates
Vicente Rives a,*, Marı́a Angeles Ulibarri b
a
Departamento de Quı́mica Inorgánica, Uni6ersidad de Salamanca, Salamanca, Spain
b
Departamento de Quı́mica Inorgánica e Ingenierı́a Quı́mica, Facultad de Ciencias,
Uni6ersidad de Córdoba, Córdoba, Spain
Received 24 April 1998; received in revised form 11 September 1998

Contents

Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2. Systems hosting single-metal anionic complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1. Layered double hydroxides intercalated with halocomplexes. . . . . . . . . . . . . . . . 64
2.2. Layered double hydroxides intercalated with cyanocomplexes . . . . . . . . . . . . . . . 66
2.3. Layered double hydroxides intercalated with oxocomplexes . . . . . . . . . . . . . . . . 73
2.4. Layered double hydroxides intercalated with macrocyclic
ligand-containing complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3. Systems hosting oxometalates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.1. Layered double hydroxides intercalated with low-nuclearity oxometalates . . . . . . . . 87
3.2. Layered double hydroxides intercalated with medium-nuclearity oxometalates:
vanadates and molybdates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.2.1. Vanadates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.2.2. Thermal decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.2.3. Molybdates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.3. Layered double hydroxides intercalated with high-nuclearity oxometalates:
iso and hetero-polyoxometalates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4. Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

* Corresponding author. Tel.: +34-923-294400/1545; fax: +34-923-294574; e-mail:

vrives@gugu.usal.es.

0010-8545/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S 0 0 1 0 - 8 5 4 5 ( 9 8 ) 0 0 2 1 6 - 1
62 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Abstract

This paper reviews the synthesis, properties and applications of layered double hydroxides
(LDHs), also known as anionic clays or hydrotalcite-like materials, containing intercalated
anions constituted by metal complexes or oxometalates. After an introduction describing the
main features of these compounds, emphasis is put on the synthesis methods, characteriza-
tion and applications. © 1999 Elsevier Science S.A. All rights reserved.

Keywords: Anionic clays; Hydrotalcite-like materials; Layered double hydroxides; LDH; Oxometalates;
Anionic compounds of metal ions

1. Introduction

Layered double hydroxides (LDHs), also known as anionic clays, are a family of
compounds which are deserving much attention in recent years ([1–3] and refer-
ences therein). The structure of most of them corresponds to that of hydrotalcite,
a natural magnesium – aluminum hydroxycarbonate, discovered in Sweden around
1842, which occurs in nature in foliated and contorned plates and/or fibrous
masses. Its formula is Mg6Al2(OH)16CO3 · 4H2O, although due to the relationship
between its structure and that of brucite, Mg(OH)2, it is usually formulated as
[Mg0.75Al0.25(OH)2](CO3)0.125 · 0.5H2O. Brucite shows the well-known CdI2-type
structure, i.e. an hexagonal close-packing of hydroxyl ions, with all octahedral sites
every two interlayers occupied by Mg2 + ions. Partial Mg2 + /Al3 + substitution gives
rise to positively charged layers, thus leading to location of anions in the unoccu-
pied interlayers. In natural hydrotalcite these interlayer anions are carbonate, and
water molecules also exist in the interlayer space. Stacking of the layers can be
accomplished in two ways, leading to two polytypes with a rombohedral (3R
symmetry) or an hexagonal cell (2H symmetry); hydrotalcite corresponds to sym-
metry 3R, Fig. 1, while the 2H analogous is known as manasseite [2]. On the other
hand, the electric charge of the layers and the interlayer ions is just the opposite of
that found in silicate clays (cationic clays), and for all these reasons, these materials
are usually known as layered double hydroxides (LDH), anionic clays or hydrotal-
cite-like materials.
As with cationic clays, the interlayer anions are easily exchanged, and carbonate
has been exchanged for many different anions [4–9], including even hydroxyl
groups (meixnerite). The nature of the layer cations can be also changed, and,
although most of the studies reported in the literature refer to systems with
M2 + /M3 + cations in the layers, other are known with M2 + /M4 + [10,11] or the
rather well studied Li + /Al3 + system [12]. The value of the M3 + /M2 + ratio is
limited if pure materials are desired, and such a ratio, in addition to being
important, also determines the concentration of interlayer anions.
Also as cationic clays, hydrotalcites can be pillared with polynuclear anions
[1,13–17], although the thermal stability of anionic clays is markedly lower than
that of cationic clays. Its thermal decomposition leads to mixed oxides.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 63

An important feature of these hydrotalcites is that they can be obtained, in

addition to direct synthesis from soluble salt precursors or by anionic exchange, by
recovering of the layered structure, once a hydrotalcite has been calcined at
moderate temperatures; this treatment leads to mostly amorphous materials, which,
in contact with solutions containing anions, recover the layered structure, hosting
the anions in the interlayer space.
Interest in hydrotalcites and derived materials arises from the widely use they can
be given: catalysts or catalyst supports, processing of selective chemical nanoreac-
tors, separation and membrane technology, filtration, scavenging and controlled
release of anions, electroactive and photoactive materials, etc. [1].
Most of the applications correspond to the field of heterogeneous catalysis
[3,18,19], where the choice of different metal cations (hydrotalcites have been also
prepared containing two or more different M2 + or M3 + cations in the brucite-like
layers) make these materials rather suitable for a fine modulation of chemical
composition and hence catalytic properties. On the other hand, as the cations in
hydrotalcites are well dispersed, heterogeneous catalysts obtained upon thermal
decomposition usually show a high dispersion of the metal sites. The use of
microporous and mesoporous molecular sieve materials in catalysis, including
LDHs and pillared LDHs, has been recently reviewed [20].

Fig. 1. Idealized structure of a layered double hydroxide, with interlayer carbonate anions. Several
parameters are defined.
64 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Intercalation of metal-containing anions in the interlayer space of hydrotalcites

provides several advantages for an improved use: in the case of catalysts, it permits
incorporation of, at least, a third metal component in the mixed oxide obtained
upon thermal decomposition. In other materials, the interlayer space provides
unique reaction conditions, thus permitting some reactions to take place under
softer experimental conditions than in its absence.
In the present review we have collected the chemistry of hydrotalcite-like
materials with metal-containing interlayer anions, well as halocomplexes, oxocom-
plexes, cyanocomplexes or oxometalates. In order to systematize the study, we have
chosen a classification based on the nuclearity of the interlayer anion, rather than
on the nature of the interlayer metal atom. Although we refer mostly on the
preparation and characterization of these materials (and so experimental details are
often given), changes induced in their properties (as compared with those displayed
by the same anions in the bulk form), and their applications, are also considered.
Literature until the end of 1997 has been reviewed.

2. Systems hosting single-metal anionic complexes

2.1. Layered double hydroxides intercalated with halocomplexes

LDHs intercalated with halo-complexes have been investigated as catalysts for

chloride-exchange reactions, and also as modified electrodes. The reactivity of some
halo-complexes in the interlayer space has been also investigated.
López-Salinas and Ono [21] have prepared Mg,Al–LDH with [NiCl4]2 − or
[CoCl4]2 − in the interlayer by anion exchange (from a nitrate-containing precursor)
in non-aqueous media. However, exchange is not complete (IR spectra of the LDH
halo-complexes show a band at 1376 cm − 1, ascribed to mode n1 of NO3− anions),
and the complexes partially decompose during the exchange process or the washing/
drying steps. In addition, the IR spectra evidence the formation of a small amount
of nickel nitrato-complexes, where the nitrato anions coordinate to nickel as
bidentate ligands, as concluded from the positions of the n5 and n1 nitrate bands
[22]. The nature of the non-aqueous solvent (ethanol or nitromethane) used during
synthesis seems to play no major role on the nature of the interlayer species (for the
Ni-compound), nor on the exchange degree (38% when synthesis was carried out in
ethanol, and 26% in nitromethane).
The measured interlayer thickness of the LDH–MCl4 (M=Ni, Co) compounds
were almost coincident with that measured for the Mg,Al–NO3 precursor, with the
first (003) harmonic being recorded at 10.8 Å, corresponding to nitrate anions
located in vertical position between the brucite-like layers [23]. Taking also into
account that the interlayer width (3.3 Å) is smaller than the ionic diameter for
[NiCl4]2 − in [Me4N]2[NiCl4], 4.6 Å [24], these findings throw some doubts about if
the complex is located in the interlayer or simply adsorbed on the external surface.
Outgassing of LDH – CoCl4 at 200°C (a treatment that changed the color of the
sample from green – blue to turquoise blue), and exposure to ammonia vapor, led to
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 65

a change to a violet color, but the UV–vis bands expected for tetrahedral
[Co(NH3)4]2 + or octahedral [Co(NH3)6]2 + species did not appear; instead, all the
bands shifted to the high energy side of the spectrum. Gentle evacuation at room
temperature (r.t.) restored the original spectrum, indicating that the interaction
between ammonia and the Co(II) ions is rather weak. A similar experiment with
pyridine (a stronger field ligand than ammonia in the spectrochemical series) gave
rise to no color change, even after 1 week exposure at r.t. These results suggest
pyridine does not reach the complex, which should be located in the interlayer
space, and not simply adsorbed on the external surface of the particles.
UV–vis spectroscopy shows further evidence of reactivity in the interlayer space.
In the solid state, the UV – vis/DR (Diffuse Reflectance) spectrum of the LDH–
[NiCl4] compound prepared in ethanol shows, in addition to a doublet at 665 and
706 nm characteristic of [NiEt4]2[NiCl4] salt and of [NiCl4]2 − in solution, two bands
at 550 and 750 nm that have been ascribed by these authors [21] to the presence of
three-coordinated planar [NiCl3] − species [25], while the presence of hexacoordi-
nated Ni species, similar to those existing in NiCl2 [26], can be concluded from the
presence of a band at 450 nm.
In the interlayer space, chloride ligands in the coordination shell of Ni(II) ions
can be reversibly exchanged by water molecules, as concluded from UV–vis/DR
measurements. Outgassing the LDH – [NiCl4] compound under reduced pressure up
to 150–250°C leads to development of bands characteristic of [NiCl4]2 − and
[NiCl3] − , and the color of the sample changes from pale green to pale violet. These
changes can be related to stripping of ligand molecules (residual water or solvent
molecules) from hexacoordinated species NiCl4L2 (L=ligand). Exposure of the
outgassed solid to water vapor at r.t. results in a pale green solid, which spectrum
shows a broad band at 700 nm and a weak band at 390 nm, which may be ascribed
to [Ni(H2O)6]2 + ions. Reevacuation at 200°C for 2 h results in a spectrum similar
to that recorded before water adsorption, the pale-violet color being restored.
The green – blue colored Co-complex shows a somewhat different behavior. Its
UV–vis/DR spectrum shows the characteristic split band at 610 and 658 nm,
typical of tetrahedral Co(II) species, as in the precursor salt [27]; in addition, a
broad feature at 500 nm suggests the presence of Co(II) species with a higher
coordination number, probably [Co(H2O)6]2 + . Outgassing leads to a change in the
color of the sample to turquoise blue, removal of the band at 500 nm, and shift of
the split band to 570 and 608 nm, while a sharp band at 756 nm, with shoulders at
740 and 715 nm, develops. The anomalous position of the sharp band at 756 nm
suggests, from comparison with the spectra of halo-cobalt(II) complexes (halide=
Cl − , Br − , I − ) and other compounds, a decrease in the field strength around the
Co(II) cation in evacuated LDH – [CoCl4]. Exposure to water vapor removes the
756 nm band, suggesting formation of both tetrahedral and octahedral Co species,
coordinated by chloride and aquo ligands, contrary to the results with the Ni
analogous, where only octahedral species existed in the presence of water vapor.
Tetrachloronickelate(II) has been also introduced in the interlayer space of a
Li,Al–LDH by anionic exchange from the nitrate form [28]. Again the interlayer
space calculated (2.9 Å) from the spacing (7.7 Å) was too small for tetrahedral
66 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

species. However, from EXAFS measurements, the first neighbor Ni–Cl distances
were 2.10 Å, compatible with 4-fold coordination. In addition, no EPR signal was
recorded for the LDH – [NiCl4] compound, despite a strong signal was reported
both for octahedral (NiCl2) and tetrahedral ([Et4N]2[NiCl4] in nitromethane) spe-
cies, thus suggesting that a change in the geometry has occurred in the LDH
interlayer, forming square planar [NiCl4]2 − species which size, if the C4 axis is
perpendicular to the brucite-like layers, is compatible with the measured interlayer
space.
The LDH – [NiCl4] compound has been tested for chloride/bromide exchange in
butyl bromide in toluene suspension [29]. The exchange rate has been found to be
slightly lower than that for LDH in the chloride form. This rate sharply increases
with the reaction temperature; thus, only 18% exchange was achieved at 50°C after
150 min, while at 100°C an exchange degree of 84% was reached; the reaction is
inhibited in n-butanol and ethanol [30]. The exchanged bromide anion enters the
coordination shell of interlayer Ni(II) ions, as concluded from UV–vis/DR results.
This same compound, LDH[NiCl4], also catalyzes the halide exchange reaction
between benzyl chloride and butyl bromide [29,30], but only in DMF, and not in
toluene.
Anion-exchanged hydrotalcite-like clay-modified electrodes containing IrCl26 − in
the interlayer have been reported by Itaya et al. [31]. These were prepared by anion
exchange of the LDH in the chloride form (carbonate was not exchanged) by
shaking the electrodes (films of LDH –Cl on SnO2) with a 20 mM solution of the
iridium complex for ca. 1 h, leading to a solid with a basal spacing of 10.8 Å,
corresponding to an interlayer space of 6.03 Å. Steady voltammgrams recorded
after an initial potential of 0 V (vs. SCE) for 10 s indicate that all [IrCl6]2 − ions
incorporated into the film were reduced to [IrCl6]3 − at 0 V.

2.2. Layered double hydroxides intercalated with cyanocomplexes

Layered double hydroxides intercalated with cyano-complexes of iron, cobalt,

molybdenum and some other metals have been synthesized and their properties for
hydrocarbon adsorption and electrochemical behavior have been studied. In addi-
tion to X-ray diffraction, so useful to characterize these layered materials, IR and
Mössbauer (in the case of iron and cobalt-containing cyanides) spectroscopies have
been also widely used. The methods of synthesis used have been coprecipitation and
anionic exchange in most of the cases, although reconstruction of the layered
structure from a carbonate – LDH precursor calcined at 500°C has been also used
[32].
Miyata and Hirose [33] reported the synthesis of a Mg,Al–[Fe(CN)6]4 − LDH by
coprecipitation; its X-ray diffraction diagram is consistent with an hexagonal cell
with a= 3.06 Å and c = 33.62 Å, i.e. the value for a is very close to that of the
hydrotalcite with the same Mg/Al ratio, but the value for c has increased due to the
larger ionic radii of hexacyanoferrate(II) with respect to that of carbonate. How-
ever, a small amount of carbonate also existed in the interlayer space, probably
through adsorption from atmospheric carbon dioxide during preparation. Even so,
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 67

Fig. 2. X-ray powder diffractions and IR spectra of [Fe(CN)6]4 − – LDH. R-values correspond to molar
percentage of aluminum. Reprinted from S. Kikkawa, M. Koizumi, Ferrocyanide anion bearing Mg,Al
hydroxide, Mater. Res. Bull. 17 (1982) 191–198, © 1982, with permission from Elsevier Science.

the amount of carbonate is markedly lower than that found as an impurity in

hydrotalcites with monovalent anions, such as chloride or nitrate.
On heating, this material losses interlayer water continuously between 100 and
300°C, the thickness of the layer, d(003), decreasing from 11.18 to 8.04 Å at 250°C.
At 300°C the material is almost amorphous, and at 400°C the layered structure
collapses, with simultaneous removal of structural water (from condensation of
layer hydroxyl groups) and decomposition of the interlayer anions. As observed
with most of this sort of materials, nitrogen adsorption at − 196°C for specific
surface area assessment indicates its increase from 80 m2 g − 1 (original sample) to
348 m2 g − 1 for the sample calcined at 200°C, the maximum value (419 m2 g − 1)
being measured for the sample calcined at 150°C. Once interlayer water molecules
have been removed, interstitial sites in the interlayer are able to adsorb molecules
such as O2 or CO2.
Similar compounds, but with varying Mg/Al ratios (between 2:1 and 5:1) have
been prepared by Kikkawa and Koizumi [34], who have observed that intercalation
of hexacyanoferrate(II) is favored as the Al content increases, especially for a
Mg:Al ratio equal or lower than 3; for Mg:Al =5:1 the major interlayer anion was
carbonate and for intermediate values (Mg/Al = 4:1) both anions coexist, as
concluded from the values for d(003) spacings, and also from the relative intensities
of the IR bands recorded close to 2000 cm − 1, characteristic of n(CN) in hexacyano-
ferrates [35] and the n3 band of carbonate close to 1350 cm − 1, Fig. 2.
68 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

One of the questions raised with these hexacyanoferrate-containing LDHs has

been the orientation of the anion in the interlayer. In the case of carbonate it is
widely accepted, from X-ray diffraction, ionic size and charge density, that its C3
axis is perpendicular to the brucite-like layers, while in the case of nitrate it may be
oriented with the C3 axis parallel to the layers [23], due to the larger number of
(monovalent) nitrate anions required to balance the positive charge of the layers,
than of divalent carbonate anions; even for a Mg/Al ratio fairly low (i.e. with
relatively high positive layers), a basal spacing of 8.8 Å (the same observed for a
nitrate LDH) for a carbonate LDH, has been attributed to ‘upright’ oriented
carbonate anions [36]. For hexacyanoferrate(II) LDHs, Kikkawa and Koizumi [34]
suggest an orientation where the C3 axis of the anion octahedron is perpendicular
to the brucite-like layers, the face-to-face distance in the octahedron being 6.5 Å,
thus leading to a total thickness of 11.3 Å (from the thickness of the brucite layer,
4.8 Å), a value fairly close to that determined by X-ray diffraction (ca. 11.2 Å), Fig.
3. This conclusion has been also reached by Braterman et al. [37] through a detailed
study of the IR spectra of these compounds in the n(CN) region using oriented and
randomly oriented [Mg2Al(OH)6]4[Fe(CN)6] (nominal composition) prepared by
anionic exchange from a chloride-containing LDH. For a regular octahedron (Oh
symmetry) with cyanide ligands in the corners, the C–N vibrational modes corre-
spond to A1g +Eg +T1u. The first two modes are IR-forbidden, although they are
recorded as very weak bands, and the band corresponding to mode T1u splits into
two bands, assigned to modes Eu and A2u. These data support a decrease of
symmetry from Oh to D3d, with opposite triangular faces of the octahedron parallel
to the brucite-like layers. In addition, the environment of the interlayer hexacyano-
ferrate anions is closer to that observed in aqueous solution than that observed in
metallic salts, where the terminal nitrogen atom is bonded to a cation, so support-
ing a model where the anion becomes hydrogen-bonded to OH groups from the
layers or from interlayer water molecules, but not directly bonded to Mg2 + or
Al3 + . However, Wang et al. [38] attribute an interlayer spacing of 6.6 Å in an

Fig. 3. Calculated end-to-end distances in a hexacyanoferrate ion octahedron. Adapted from Ref. [34].
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 69

hexacyanoferrate(III) Mg,Al – LDH prepared by direct synthesis to the octahedron

oriented with its C4 axis perpendicular to the brucite-like layers, although crystallo-
graphic data do not support his conclusion.
A decrease in the symmetry of the [Fe(CN)6]4 − octahedron in the interlayer
space of LiAl2 – [Fe(CN)6]4 − (a material with a structure similar to that of hydrotal-
cite) prepared by ionic exchange of a chloride precursor, has been also concluded
by Dutta and Puri [39] from its Raman and IR spectra, which shows five bands
between 2094 and 2027 cm − 1 (n(CN) region) instead of the three bands at 2098
(Raman, A1g), 2062 (Raman, Eg), and 2044 (IR, T1u) cm − 1 recorded in solution. As
only four bands would be expected for a C3 site symmetry, a lower symmetry or
intermolecular coupling effects may be responsible for the spectrum observed.
The electronic state of iron in cyanoferrate-containing LDHs has been studied by
IR and Mössbauer spectroscopies by Idemura et al. [40]. These authors have
prepared different samples by anionic exchange of a nitrate–LDH (nominal Mg:Al
ratio 3:1) with decarbonated water under nitrogen atmosphere with K3[Fe(CN)6],
K4[Fe(CN)6] · 3H2O, and Na2[Fe(CN)5(NO)] · 2H2O. Exchange with [Fe(CN)6]3 −
reached only 84%, but its IR spectrum shows absorption bands at 2120 and 2040
cm − 1, due to n(CN) of [Fe(CN)6]3 − and [Fe(CN)6]4 − , respectively, suggesting a
partial reduction from Fe(III) to Fe(II) during intercalation. When anionic ex-
change was carried out with [Fe(CN)6]4 − , a single band was recorded at 2040 cm − 1
if the sample had been dried under vacuum, but an additional band at 2120 cm − 1
(characteristic of the hexacyanoferrate(III) complex) is recorded in addition when
the sample was dried in air, i.e. partial Fe(II) “ Fe(III) oxidation should take place.
This behavior has been also observed in some other cases, although an unambigu-
ous explanation is still lacking in the literature.
These redox processes have been also concluded from Mössbauer spectroscopy
studies. Two different states for iron have been found in the [Fe(CN)6]3 − ex-
changed LDH: one corresponding to [Fe(CN)6]3 − , although distorted along the C3
axis, probably because of the stress from the host layers, and another correspond-
ing to [Fe(CN)6]4 − , formed upon reduction of hexacyanoferrate(III), although the
mechanism for this reduction remains unclear. Correspondingly, the Mössbauer
spectrum of the sample originally exchanged with [Fe(CN)6]4 − is identical to the
second species above cited, confirming oxidation in this case.
The IR spectrum of the [Fe(CN)s(NO)]2 − containing LDH shows an absorption
at 1940 cm − 1 (n(NO) stretching) and four bands in the n(CN) stretching region
(2143–2040 cm − 1). From the relative intensities of the bands, if compared to those
of the spectrum of the corresponding sodium salt, Idemura et al. conclude [40] that
most of the NO ligands are removed along the exchange process. Detection of
bands at 2040 and 2050 cm − 1 (due to hexacyanoferrate(II)), together with the
hexacyanoferrate(III) one at 2100 cm − 1, suggests a series of reactions in the layered
host, as follows:

[FeIII(CN)5(NO)]2 − “[FeII(CN)5(NO)]3 − “ [FeII(CN)5(NO)]3 − + NO (1)

[FeII(CN)5]3 − +H2O “[FeII(CN)5(H2O)]3 − “ [FeIII(CN)5(H2O)]2 − (2)

70 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Carrado et al. [7] have found differences in the electronic properties of a

Mg,Al–[Fe(CN)6]4 − LDH as synthetized (green crystals), or after being ground to
a grey powder. In the former case, the Mössbauer spectrum shows isomer shifts of
1.8 and − 0.14 mm s − 1 (close to that of K3[Fe(CN)6] at − 0.124 mm s − 1), while
the powder shows a single peak with an isomer shift of − 0.13 mm s − 1, i.e.
oxidation takes place already during synthesis, and is completed after grinding.
These data have been confirmed by these same authors [7] by EPR, the spectra
showing a signal near g = 4.1 due to Fe(III) in all cases. These results would suggest
that the Fe(III)“ Fe(II) reduction is achieved through pressure during grinding,
and in fact, Larsen and Drickamer [41] have reported reduction of K3[Fe(CN)6]
under high pressure. On the other hand, in most cases reduction is concluded from
IR spectra, where usually samples are prepared under high pressure as well in KBr
pellets. However, IR spectra of Mg,Al–[Fe(CN)6]3 − LDHs recorded following the
DRIFTS technique (where the samples are not submitted to high pressure) also
show the presence of both oxidation states of iron [42].
Hansen and Koch [43] have synthetized a Mg,Al–[Fe(CN)6]4 − LDH from the
carbonate form, via an intermediate nitrate form; carbonate is not completely
expelled, and 30% of the layer positive charge is still balanced by interlayer
carbonate. Partial oxidation (up to 20%) to [Fe(CN)6]3 − is also observed by these
authors (as concluded from a IR spectroscopy study of the n(CN) region and from
Mössbauer spectroscopy studies) when the wet solid obtained during synthesis is
dried in an oven at 70 – 100°C or is washed with oxygen-containing, non-aqueous
solvents; however, if solvents without oxygen in their composition are used and the
samples are dried at r.t., oxidation is prevented. These authors conclude that
oxidation takes place not because of oxygen existing to the solvent molecule, but
because of oxygen dissolved in ethanol or acetone, and when the solid is dried
above r.t., the fast removal of water from the interlayer favors migration of oxygen
to the interlayer space, where Fe(II) becomes oxidized. Miyata and Hirose [33] had
previously reported that the interlayer space of a Mg,Al–[Fe(CN)6]4 − LDH is
accesible to gases such as N2 or O2 after being dried at 100°C.
In any case, oxidation from Fe(II) to Fe(III) requires a change in the interlayer
anions to keep balanced the positive charge of the brucite-like layers. These authors
[43] also report that the IR band recorded at 2080 cm − 1 is due to free cyanide
ligands electrostatically held in the interlayer, and formed according to the reaction:
[Fe(CN)6]4 − +H2O “[Fe(CN)5(H2O)]3 − + CN − (3)
The IR band is not recorded in samples dried at 100°C, when water molecules are
not available for such an exchange reaction, and the cyano–aquo complex is more
prone to oxidation by dioxygen than the hexacyanoferrate(II) ion [44]. Attempts for
a direct exchange to introduce cyanide anions in the interlayer, however, failed.
Hexacyanoferrate(II) and (III)-containing Mg,Al–LDHs have been also prepared
by Holgado et al. [42] by ionic exchange of carbonate containing samples at pH 4.5
(with aqueous HCl), where intermediate exchange of chloride ions presumably
takes place. The X-ray diffraction diagrams for these samples show two series of
harmonics due to diffraction by (00l) planes, indicating the presence of mixed
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 71

phases, with interlayer hexacyanoferrate and carbonate. Again, partial oxidation/re-

duction has been observed by these authors by IR spectroscopy, and their degree
has been quantified from first derivative XANES spectra, concluding that 30% of
Fe(II) becomes oxidized, while 20% of Fe(III) is reduced. When the samples are
calcined in air at 600°C, X-ray diffraction and temperature-programmed reduction
results suggest formation of MgO and MgFe2O4, whichever the starting hexacyano-
ferrate ((II) or (III))-containing LDH.
Crespo et al. [45] have reported a careful study of different methods (including
direct synthesis, reconstruction from a calcined precursor, anionic exchange of a
nitrate LDH, and via intermediate terephthalate) to prepare Zn,Al–LDH with
hexacyanoferrate(II) and hexacyanoferrate (III). Although in all methods used, the
redox process above observed also develops, these authors report that to obtain
carbonate-free phases the best method was direct synthesis for hexacyanoferrate(II),
while anionic exchange of a nitrate precursor for hexacyanoferrate(III).
Hexacyanocobaltate(III) anions have been introduced in the interlayer space of
LDHs with different layer cations (Mg, Zn, Al, Cr) by anionic exchange of
precursors with different anions (nitrate, sulphate, chloride, chromate) [46]. Ex-
change levels of 79 – 90% were reached in all cases, except for Zn,Cr–NO3, where it
was only 21%. The expansion of the interlayer space from 3.0–4.0 Å (depending on
the precursor anion) to 6.0 – 6.5 Å suggests the [Co(CN)6]3 − anions are oriented
with their C3 axis perpendicular to the brucite-like layers, as previously reported for
hexacyanoferrate complexes. Expansion of the interlayer space allows also access of
nitrogen molecules during specific surface area measurements, which shows a value
of 13 m2 g − 1 for the Mg,Al – NO3 precursor, but 330 m2 g − 1 for the exchanged
LDH. A similar increase in the specific surface area had been also reported by
Cavalcanti et al. [47] for a [Fe(CN)6]3 − containing LDH. In addition, the shape of
the isotherm changes from type II in the IUPAC classification [48], due to
adsorption on non-porous solids, to a Langmuir type isotherm [47,49], due to
adsorption on microporous solids. Expansion makes also the interlayer space
accessible to linear hydrocarbons and, in a lesser extent, to branched hydrocarbons.
The accessibility of the interlayer space of hexacyanoferrate (both Fe(II) and
Fe(III))-intercalated LDHs has been studied by Cavalcanti et al. [47]. For a given
Mg/Al ratio (3.0), the population of [Fe(CN)6]4 − should be lower than that of
[Fe(CN)6]3 − (because of the larger formal negative charge in the former), and thus,
taking into account the close size of both anions, the void interlayer space accesible
for adsorption would be larger. However, the measured specific surface areas were
246 m2 g − 1 for the [Fe(CN)6]4 − – LDH, and 355 m2 g − 1 for the [Fe(CN)6]3 − –
LDH. This unexpected result is explained on the basis of the redox process
occurring in these materials, as concluded from IR spectroscopy studies, and widely
discussed above, as well as by incorporation of carbonate anions from the atmo-
sphere to the interlayer space, for charge balance after such redox process. Results
on adsorption of hydrocarbons (C5 – C7) is consistent with the specific surface area
values measured, and changes in the Mg/Al ratio are also consistent with a decrease
in void interlayer space as the Al content is decreased. A variation in the charge
density of the layers in Mg,Al – [Fe(CN)6] LDHs via the M2 + /M3 + ratio optimizes
72 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

the porosity properties [50], with a value of 499 m2 g − 1 for a M2 + /M3 + ratio of
3.33.
The change in the capacity for carbon dioxide adsorption on a Mg,Al-
[Fe(CN)6]4 − LDH with the Mg/Al ratio (between 1 and 7) has been studied by
Mao et al. [51]. Maximum adsorption was observed for Mg:Al= 1.7, i.e. it depends
not only on the width of the interlayer space (ranging, although not steadily, from
10.64 to 10.96 Å), but also on the layer charge. The isosteric heat of adsorption was
calculated to be 43.3 kJ mol − 1, a value similar to that reported by Miyata and
Hirose [33], who found the adsorption capacity for CO2 of a Mg,Al–[Fe(CN)6]4 −
LDH to be 60% of that of zeolite 5A [52].
The interlayer space of LDHs provides a reaction medium for chemical reactions.
However, its utility is limited because of the interlayer space and the size of the
reagents. So, reactions have been studied on LDHs previously expanded with large
anions, such as hexacyanoferrates or terephthalate. Challier and Slade [53] have
reported the oxidative (due to Cu2 + ) polymerization of aniline between the layers
of a Cu,Al – [Fe(CN)6]4 − LDH.
Metalocyano-containing LDHs have been also used to prepare modified elec-
trodes, as the redox behavior of the interlayer anions makes them electrochemically
active. Itaya et al. [31] have reported on LDHs containing [Mo(CN)8]4 − or
[Fe(CN)6]3 − prepared by anionic exchange of a commercial Mg,Al–Cl precursor.
A film (thickness ca. 100 nm) of the chloride hydrotalcite was prepared on SnO2,
and exchange was achieved by shaking the electrodes in dilute (20 mM) solutions of
the metalocyano complexes. Interlayer spacings of 11.2 Å were measured, in
agreement with previous results described above. The peak currents for the
[Mo(CN)8]4 − /3 − couple incorporated in the interlayer are clearly observed at ca.
0.5 V (vs. SCE), which is almost in complete agreement with the half-wave potential
of this couple in the same solution. In addition, a steady voltammogram was
obtained after 10 – 20 cycles, without further decrease after 60 min cycling, indicat-
ing the anion is strongly held in the interlayer space of the LDH. Similar results
were obtained for [Fe(CN)6]3 − .
Hexacyanoferrate(III) was used by Shaw et al. [54] as an electroactive probe to
determine the availability of Zn2 + and Al3 + at the surface of a Zn,Al–Cl LDH
applied at the surface of a glassy carbon electrode. When [Fe(CN)6]3 − is reduced at
the electrode surface to [Fe(CN)6]4 − , this becomes anchored forming a multilayered
Prussian blue-like film, the nitrogen atom binding to Zn2 + and Al3 + ions from the
layers (as concluded from XPS measurements for the N(1s) energy level), so
demonstrating such an availability. Phenol was not electro-oxidized on this elec-
trode, while such an oxidation was observed on the Mg,Al analogue, where the
Prussian blue-like layer was not formed. From these results, and taking into
account that oxidation of phenol is pH-dependent, these authors were able to
assign effective pH values of \11.2 and 8.3, respectively, for the surfaces of
Mg,Al–Cl and Zn,Al – Cl.
Hexacyanoferrate(II) modified carbon paste electrodes have been also studied by
Labuda and Hudáková [55], who observe oxidation of ascorbic acid at a potential
0.4 V less positive than the oxidation potential at a bare carbon paste electrode, at
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 73

pH\ 4. According to these authors, this electrode is more sensitive and more
selective than carbon paste electrodes modified with organic salts and that hexa-
cyanoferrate(II) bound to poly(4-vinylpyridine), and permits good storage and
operational stability.
Qiu and Villemure [56] have found an enhanced reduction current when
[Fe(CN)6]4 − or [Mo(CN)8]4 − are exchanged in Ni,Al or Ni,Fe–LDH-modified
electrodes. These authors found that Ni2 + in these LDHs can be oxidized up to
4 + if sufficiently positive potentials are used, but they can be reduced only to
2.7+ ; reversible reduction to 2+ is attained only up to a maximum oxidation
value of 3.6+, the layered structure being stable, at least, as far as X-ray
diffraction diagram concerns. The enhancement in the cathodic current has been
attributed to electron transfer between the intercalated anions and the oxidized Ni
sites in the brucite-like layer, as no enhancement was observed when the oxidized
forms of the cyanocomplexes were intercalated. Similar results were obtained in
systems containing Mn or Co in the brucite-like layers, the voltammetric waves
being smaller, however, if the interlayer anions were carbonate or chloride, instead
of [Fe(CN)6]4 − , [Mo(CN)8]4 − or [Ru(CN)6]4 − . When the potentials of Ni, Co or
Mn LDH modified electrodes were scanned in mixtures of the iron and ruthenium
or molybdenum cyanocomplexes, electron transfer from [Fe(CN)6]4 − to electro-
chemically oxidized [Ru(CN)6]3 − or [Mo(CN)8]3 − , mediated by the LDH electroac-
tive metal centres, was observed, but such a transfer was not observed for the redox
inactive Zn,Al – Cl LDH.

2.3. Layered double hydroxides intercalated with oxocomplexes

Corma et al. have reported the preparation of Zn,Al–LDHs containing Mo

oxo-complexes in the interlayer, and their role as heterogeneous catalysts for the
oxidation of thiols [57 – 60]. [MoVIO2(O2CC(S)Ph2)2]2 − ([O2CC(S)Ph2]2 − being the
sterically hindered 2,2-diphenyl-2-mercaptoethanoate ligand) is able to oxidize
aliphatic thiols under homogeneous conditions [61,62], it being reduced to a
monomeric complex [MoVO(O2CC(S)Ph2)2] − . However, this is clearly a disadvan-
tage, as this second complex does not undergo oxidation with Me2SO, so cancelling
the possibility to design catalytic cycles.
Reduction of the MoVIO2 complex takes place [63] through Reactions (4),
reduction by the thiol substrate, and (5), comproportination of the MoVIO2 species
and its MoIVO reduced form, to yield a MoVO species:
[MoVIO2{O2CC(S)Ph2}2]2 − +2RSH
“[MoIVO{O2CC(S)Ph2}2]2 − +RSSr + H2O (4)
[MoVIO2{O2CC(S)Ph2}2]2 − +[MoIVO{O2CC(S)Ph2}2]2 − + 2H +
“ 2[MoVO{O2CC(S)Ph2}2] − +H2O (5)
So, immobilization of the Mo complex in the interlayer space of an LDH
prevents from Reaction (5) to take place, due to steric hinderance, and thus
74 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Fig. 4. Structure of intercalated [MoO2{O2CC(S)Ph2}2]2 − based on basal spacing measurements (left),

and from EXAFS results (right). Adapted from Ref. [60].

permiting species [MoIVO{O2CC(S)Ph2}2]2 − to be oxidized to close a catalytic cycle

[59].
The LDH – MoVI intercalation compound was prepared by anionic exchange of
the LDH in its nitrate form. The measured spacing for the (00l) reflection was 17.6
Å, corresponding to an interlayer space close to 13 Å, consistent with the anionic
complex oriented with its C2 axis parallel to the brucite-like layers; actually, the size
of the anion in the direction perpendicular to such a C2 axis was calculated to be
12.6 Å from X-ray crystallography [61]. Exchange, however, is not complete, as
concluded from a FT-IR band at 1367 cm − 1, due to nitrate species (mode n3). The
LDH–MoVI complex could not be obtained through the so-called reconstruction
method. It is worthwhile to be mentioned that the intercalation compound is
indefinitely stable in air.
The complex was active for thiophenol oxidation to disulphide by dioxygen, and
the conversion was much lower when using air as the oxidizing agent. According to
these authors [58], a catalytic cycle can be proposed, where the coupling of electron
and proton transfer allows a direct formation of products without the formation of
less stable intermediates, which would then have to be protonated by hydrolysis.
Moreover, another advantage is the use of water as solvent.
Although X-ray diffraction and IR and UV–vis/DR spectroscopies data are
consistent with the structure shown in Fig. 4(left), recent results [60] obtained by
XAS indicate a change in the local coordination around the MoVI ions, as shown
in Fig. 4(right) (both situations would lead, however, to similar values for the
height of the interlayer space), from the isolated compound, to the situation when
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 75

intercalated in the Zn,Al – LDH. The increase in the number of terminal oxygen
atoms bonded to the MoVI ion, as concluded from XANES studies [60], follows
decoordination of the carboxylate ligands, this being the first step which induces
sufficient positive charge on the MoVI site to allow an oxygen atom transfer from
a contiguous water molecule without any change in the oxidation state of Mo.
Mg,Al– LDH has been also intercalated with peroxomolybdate(VI) anions con-
taining tartrate as ligand [64], [Mo2O2(O2)4(C4H2O6)]4 − . Intercalation was achieved
after swelling the LDH with terephthalate, a method widely used to intercalate
large anions [65], leading to samples with spacings of 12.4 and 14.3 Å, respectively,
for the terephthalate and peroxomolybdate-containing compounds. When sus-
pended in water, this compound releases the peroxo group at 80°C to yield
oxomolybdate, the process being reversible by addition of hydrogen peroxide;
oxygen is released when heated in the solid state at ca. 120°C.
Pinnavaia et al. [66] have studied the effect of intercalation in different hosts on
the properties of the excited-state of dioxorhenium(V) ions. From previous studies
with polypyridyl complexes, it was concluded that the excited-state properties are
preserved upon intercalation [67 – 70]. However, in the case of metal–oxo com-
pounds and, particularly, d2 trans-dioxo species [71], it was expected that its
location in the interlayer space of LDHs and layered silicate clays would probably
modify the orientation of the trans-ReO2+ core, thus affecting luminiscence proper-
ties. A Mg,Al – ReO2(CN)4 LDH was prepared by coprecipitation of the hydrotal-
cite in the presence of the complex anion, while intercalates in layered silicate clays
were prepared with [ReO2py4] + and [ReO2en2] + (py= pyridine; en=ethylendi-
amine) in hectorite and fluorhectorite. In all three cases, exchange corresponded to
ca. 15% of the exchange capacity of the clay.
Location of these complexes in the interlayer space of the clays gives rise only to
minor structural distortions, as revealed by IR and Raman (in the resonance
Raman effect mode) spectroscopies. Basal spacings (from XRD studies) are en-
larged, consistenly with incorporation of the complexes. The luminescent hectorite
intercalate is largely unperturbed, emission from the fluorhectorite intercalate is
significantly attenuated, and no luminiscence was observed from the ReO2 –LDH
intercalate. From basal spacing measurements and calculations of charge density of
the layers, these differences have been related by these authors [66] to the different
possibilities of orientation of the ReO2 complexes in the interlayer space, Fig. 5. So,
in the case of the hectorite intercalate, the complex should be located with its
O–Re–O axis perpendicular to the layers, and the oxygen atoms can then be
‘keyed’ in the hexagonal cavities of the layered silicate clay; for the fluorhectorite
intercalate, the higher charge density (27 vs. 80 Å2 per charge unit for hectorite)
precludes such an orientation, but data are consistent with the complex oriented
with its C4 axis parallel to the layers. Finally, for the LDH intercalate, the
d-spacing is consistent with the effective C3 axis of the pesudooctahedral
ReO2(CN)34 − complex perpendicular to the layers, in agreement with previous
studies which have demonstrated that the preferred orientation of intercalated
anions in LDHs either maximizes the hydrogen bonding interactions of the protons
of the brucite-like layers with the guest species, and/or minimizes the charge
76 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

separation distance between the positive layers and the gallery anions [33,72,73].
Contrary to the structure of hectorite, in LDHs there are no cavities in which the
ReO2+ core can key, and oxygen atoms of this core are probably hydrogen-bonded
directly to the hydroxide layer. Because proton-donating solvent effects efficiently
quench ReO2+ excited states by hydrogen bonding interactions [74], the non-radi-
active decay rates of electronically-excited ReO2(CN)34 − ions in LDH are expected
to be exceedingly fast.

2.4. Layered double hydroxides intercalated with macrocyclic ligand-containing

complexes

One of systems most widely studied in recent years concerning LDHs, corre-
sponds to those where the interlayer anion is a coordination compound with
macrocyclic ligands; also, where the interlayer anion is the anionic ligand itself.
These nanocomposite materials prepared with intercalated metalloporphyrines and

Fig. 5. Proposed orientations of the trans-dioxorhenium(V) core in (a) hectorite, (b) fluorhectorite, and
(c) a layered double hydroxide. Adapted from Ref. [66].
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 77

metallophtalocyanines have shown interesting properties as heterogeneous catalysts

for alkane hydroxylation, alkane epoxidation, alcohol and hydrocarbon oxidation
[75–80]. They find also applications as electrochemical microsensors in the form of
clay modified electrodes or as gas or optical sensors [81,82]. On the other hand,
stacked metal-over-metal structures of phtalocyanines are being also studied as
conducting materials [83], and the interlayer space of the LDHs provides a unique
location for such a stacking arrangement. A review has been recently published on
these [84] and related systems with other layered hosts [85], and their role as
biomimetic catalysts.
One of the first papers [86] corresponds to the intercalation of 5,10,15,20-tetra(4-
sulfonatephenyl)porphyrin (TSPP) in a Mg,Al–LDH. The interest of this anion
arises from the high temperature photochemical hole burning (PHB) of its sodium
salt above liquid nitrogen in a polyvinyl alcohol matrix [87], thus leading to
investigations about the interaction of the anion with different host substances from
the point of view of developing new PHB materials. The Mg,Al–TSPP LDH
intercalation compound was obtained by anion exchange using a Mg,Al–LDH
precursor in the chloride form, after soaking at 60°C for 1 week. The measured
basal spacing increased from 8.0 Å for the chloride LDH to 22.4 Å for the TSPP
form, i.e. a gallery height of 17.6 Å, fairly close to the side of the almost perfect
square molecule of TSPP (18 Å). From X-ray diffraction and chemical analysis it
was concluded, however, that exchange was not complete and that ca. 15% of the
anion exchange capacity of the anionic clay corresponded to intercalated carbonate.
Basal spacing, molecular size, and layer charge density suggested thus that the
intercalated porphyrin anions were arranged with the molecular plane perpendicu-
lar to the brucite-like layers, Fig. 6.
Intercalation of meso-tetrakis(p-carboxyphenyl)porphyrin (hereafter pTCPP) [88]
in Zn,Al – LDH, and of the atropoisomer (aaaa) of meso-tetrakis(o-car-
boxyphenyl)porphyrin (oTCPP) and the ammonium salt of the meso-tetrakis(p-sul-
fonatophenyl)porphyrin (pTSPP) [89] has been reported by Besse et al.
Intercalation was conducted by coprecipitation at constant pH or anion exchange
of a Zn,Al – LDH in the chloride form, leading in both cases to pure LDH phases,
while by the calcination (300°C for 18 h)–reconstruction method significant
amounts of ZnO were found. The diffraction patterns are characteristic of a
structural disorganization, probably arising from a turbostatic effect induced by
both the highly separated sheets, and the weak bond interactions between the
interlayer species and the host lattice. Also, crystallinity is lower than for the
chloride LDH, although it was improved by hydrothermal treatment at autogenous
pressure.
Pure LDHs were obtained only for Zn/Al ratios of 2 and 3; larger values lead to
coprecipitation of ZnO, while lower values lead to formation of amorphous
aluminum hydroxide. Basal spacings ranged randomly (i.e. with no relation to the
Zn/Al ratio) from 22.24 to 23.09 Å, probably because of different hydration
degrees. However, the a dimension of the unit cell (equivalent to the average
distance between two closest metal cations in the brucite-like layers [3]) steadily
increased from 3.02 to 3.10 Å for the samples prepared by coprecipitation, as the
78 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Fig. 6. Proposed orientation of 5,10,15,20-tetra-(4-sulfonatophenyl)porphyrin (TSPP) in the interlayer

space of a Mg,Al layered double hydroxide. Adapted from Ref. [86].

Zn/Al ratio was increased. From the basal spacings and the size of the intercalated
anions [90], these authors conclude [88,89] that the anion (pTCPP and pTSPP)
should be located in a perpendicular orientation relatively to the layers, with the
four anionic groups in tight interactions with the hydroxylated sheets, in a fashion
similar to that above described for the Mg,Al–TSPP LDH [86] material. Such an
arrangement would also be consistent with the layer charge density, that corre-
sponds to 33 Å2 per unit charge for the sample with a Zn/Al ratio of 3; an
arrangement of the interlayer anion in horizontal position would correspond to 49
Å2 per unit charge, while in the vertical position it corresponds to 15 Å2 per unit
charge. On the contrary, although the basal spacing measured for the LDH–
oTCPP compound (18.5 Å) could be fitted also with a vertical disposition (this
molecule is ca. 3.2 Å smaller than the other two), such an arrangement would
decrease the hydrogen bonding interaction between the carboxylate groups and the
hydroxyl anions in the brucite-like layers. Consequently, a parallel disposition of
two molecules has been claimed in this case [89].
The photochemical properties of this sort of macromolecules can be modified
when located in the interlayer space of an LDH. Tagaya et al. [91] have studied the
intercalation of colored organic anions (pTSPP and pTCPP) in Mg,Al and Zn,Al
LDHs by the reconstruction method. As in the cases above described [86,88,89], the
basal space increase is consistent with a vertical orientation of the anions (spacings
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 79

up to 30.7 Å, depending on the nature of the porphyrin), although the guest/host
ratio varied from 48 to 100%. The absorption maxima of these colored organic
anions are solvent-dependent, although no clear correlation exists between the
wavelengths of the absorption maxima and the dielectric constant of the solvents
[91]. When intercalated, a shift towards the red was observed for both anions,
which has been explained on the basis of a close packing of anion molecules in the
interlayer space, an effect similar to that previously reported for methylene blue
adsorbed on a clay [92].
Incorporation of metal cations coordinated by macroligands in the interlayer
space of LDHs has also deserved very much attention. In the case of the
Zn,Al–pTCPP system, complexation of copper in the interlayer nanospace was
attained simply by suspending the organic-LDH solid in a solution of copper
nitrate [88], and metallation was confirmed by UV–vis and EXAFS spectra. In the
UV–vis region, metallation leads to a slight red-shifting of the Soret band recorded
at 400–450 nm for the free anion, while only two Q bands are recorded between
500 and 650 nm, and these results were observed for the Zn,Al–pTCPP–Cu(II)
system. From EXAFS spectra, radial distributions around the Cu(II) were similar
for Cu(NH3)4SO4 · H2O (with four equatorial nitrogen atoms and two oxygen atoms
from water molecules in trans geometry) and for Cu(II)–pTCPP, with four nitrogen
atoms in a planar environment around the copper cation at ca. 2.10 Å and Cu–Cu
interactions along the direction perpendicular to the chelating ligand. For the
Zn,Al–pTCPP – Cu(II) system, the radial distribution reveals a nearly regular
octahedral environment, probably due to completion of the copper coordination
sphere by two water molecules, thus further confirming the perpendicular arrange-
ment of the porphyrin pillars in the interlayer space, as well as the absence of
Cu–Cu interactions.
Intercalation of tetracarboxyphthalocyanine cobalt (II), TPC–C, in Zn,Al and
Mg,Al–LDHs was carried out by the reconstruction method by Tagaya et al. [91].
The guest/host ratio was 100% for the Zn,Al system, with a basal spacing of 24.8
Å, i.e. the plane of the guest anion perpendicular to the plane of the host layers.
However, for the Mg,Al system the basal spacing was 7.9 Å. Although these
authors claim this value being consistent with a small amount (not quantified) of
intercalated TPC – C, this basal spacing coincides with the value reported in the
literature for carbonate containing Mg,Al–LDHs [3], and so it is possible that no
intercalation of the macrocyclic was attained.
Complexes with macrocyclic ligands can be also incorporated in the interlayer
space of LDHs via anionic exchange. Dutta and Puri [39] have reported complete
ion exchange of nickel(II) phtalocyaninetetrasulfonate ion in the Al2Li–LDH in the
chloride form. Quite surprisingly, the basal spacing was 10.61 Å, a value quite close
to that reported for phosphate and sulphate-containing LDHs, indicating that the
phtalocyanines are parallel (‘flat’) to the aluminate layer and not arranged in a
stacked fashion, which would require a spacing close to 22 Å.
Cobalt(II) phtalocyanines, specifically Co(II) phtalocyanine-3,4%,4%%,4%%%-tetrasul-
fonate (hereafter CoPcTs) have been also introduced in the galleries of Mg,Al
LDHs by reconstruction of the hydrotalcite precursor calcined at 500°C and
80 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

exposed to excess aqueous Pc salt at 60°C [76] or hydrothermally at 100°C [93], the
solid displaying a basal spacing of 23.3–23.7 Å, Fig. 7. This is in agreement with
a tilted orientation with respect to the hydroxide layers, with the axis joining the
non-coordinating nitrogen atoms of the CoPcTs molecule oriented almost perpen-
dicular to the hydroxide layers. These Co(II)–PcTs-containing LDHs have been
tested as catalysts for autoxidation of 1-decanethiol [76] and 2,6-di-tert-butylphenol
[93,94] by Pinnavaia et al., who have concluded that the catalyst becomes extremely
stabilized in the gallery space of the LDH, if compared to its stability under
homogeneous catalysis conditions; deactivation of the LDH-supported catalysts
during 1-decanethiol oxidation was not observed even after five catalytic cycles for
a total of more than 770 turnovers, while the homogeneous catalyst was deactivated
after 25 turnovers; such a stabilization was even greater during 2,6-di-tert-butylphe-
nol oxidation (3200 vs. 25 turnovers).
Carrado et al. [95] have synthetized Mg,Al LDHs intercalated with CuPcTs by
hydrolysis of mixed aqueous salt solutions in the presence of NaOH, a method
previously proposed by Park et al. [96], who first reported the direct synthesis of
organic dyes into LDHs. Anionic exchange was achieved only when starting from
a freshly prepared slurry (‘wet’ anionic exchange) of the LDHs (in the nitrate form),
exchanged under carbon dioxide-free conditions, and an aqueous CuPcTs solution;
however, no satisfactory products were obtained when the exchange was performed
with dry LDH (also in the nitrate form) redispersed in water. Basal spacing for

Fig. 7. X-ray diffraction pattern (Cu–Ka ) of an oriented film sample of Mg,Al – [Co(PcTs)] LDH.
Reprinted from M. Chibwe, T.J. Pinnavaia, Stabilization of a cobalt(II) phtalocyanine oxidation catalyst
by intercalation in a layered double hydroxide host, J. Chem. Soc. Chem. Commun. (1993) 278 – 280, ©
1993, with permission from The Royal Society of Chemistry.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 81

both samples (prepared by direct synthesis and by anionic exchange) was 22.5–23.0
Å, but crystallinity was better for the sample prepared by wet anionic exchange.
This spacing is markedly larger than that obtained (14–16 Å) for hectorite
interlayered with cationic copper(II)-containing dyes, such as alcian blue. The
difference arises from the different layer charge density for these cationic and
anionic clays. The anionic clay prepared by these authors had a layer charge density
close to 34 Å2 charge − 1, a value only reached by micas among the cationic clays.
From elemental chemical analysis and molecular modelling, and taking into ac-
count the layer density charge, Carrado et al. have concluded [95] that the
phtalocyanine molecules are oriented in a tilted arrangement, in agreement with
similar previous results by Pinnavaia et al. [76,93]. On the contrary, in the case of
the hectorite clays, their lower layer charge density leads to a flat orientation of the
phtalocyanine anions.
In addition to the increased stability (probably because the immobilization
process inhibits the deactivating dimerization and self-oxidation reactions occurring
in the homogeneous catalyst), it should be noted the large reactivity of the
heterogeneous Co(PcTs) – LDH catalyst, despite the extremely low specific surface
area, 28 m2 g − 1, as determined from nitrogen adsorption, i.e. the nitrogen
molecules are merely adsorbed on the external surface of the crystallites, without
accessing the gallery space. So, 2,6-di-tert-butylphenol, with a much larger Van der
Waals radii, does not access either, and thus only the Co(II) ions held at crystallite
edge sites and external basal surface sites are able of participating in the oxidation
reaction. This conclusion is supported by the finding that a similar system, but with
a lower charge layer density (with Mg/Al ratio of 4, instead of 2), and thus with a
greater separation between the intercalated cobalt centers, shows a 5-fold increase
in activity.
This same cobalt phtalocyaninetetrasulphonate has been incorporated into a
Zn,Al LDH by coprecipitation at constant pH from a Zn and Al nitrates solution
and the Co complex. The solid isolated showed an interlamellar spacing of 23.0 Å,
in close agreement with data previously reported [76]. This compound has been
used for in situ studies of cyclohexene oxidation by combined EXAFS/XRD
techniques studies [97].
Similar compounds containing increasing amounts (2–90 mmol g − 1 of LDH
support) of Co(II) phtalocyanine tetracarboxylate or Co(II) phtalocyanine tetrasul-
fonate have been tested by Iliev et al. for 2-mercaptoethanol oxidation [98].
Incorporation of the complexes into the interlayer space was achieved in this case
by soaking the LDH (previously calcined at 450°C for 24 h in argon) with a
solution of the sodium salts of the phtalocyanines, at 60°C for 7 days in argon.
However, while the basal spacing was close to 22.7 Å in both series of samples (in
agreement with the values reported by the authors previously cited), in the case of
the carboxylate phtalocyanine, crystallization of its sodium salt, most likely on the
external surface of the LDH crystallites, was also observed. The ESR spectra of
these samples showed, for low Co and Cu(II) phtalocyanine concentrations (5 mmol
g − 1 of LDH support), a hyperfine splitting from 57Co (I =7/2) and from 63,65Cu
(I= 3/2), together with superhyperfine splitting from 14N (I=1) in the case of the
82 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Cu(II)-containing sample. These results are consistent with magnetically distributed

complexes in these low-concentrated samples. In addition, when the complex
concentration was raised to 60 mmol g − 1 of LDH support, spectra characteristic of
aggregated and microcrystalline complexes were also recorded, specially a narrow
singlet at g =2.004, specific of the sodium salt of Co(II) phtalocyanine tetracar-
boxylate crystallites having deffects in the crystal lattice [99–101]. This signal did
not disappear even after repeated heating–evacuation–purging cycles, thus mean-
ing that the singlet is not due to superoxo complexes formed by the monomeric
Co(II) phtalocyanine molecules.
The catalytic effectiveness of the Co(II) containing LDHs for 2-mercaptoethanol
oxidation (to the disulfide) decreases as the concentration of the complexes is
increased, a fact probably related to aggregation and crystallization of the phtalo-
cyanine complexes, similarly to previous results for thiol oxidation using cobalt
complexes anchored on ion exchange resins or other supports [99,102,103]. The
turnover frequency was higher when water:DMF (1:3) was used instead of water as
solvent for the Co(II) phtalocyanine complex during preparation of the compounds.
Moreover, it is well known [104] that m-peroxo complexes between neighboring ion
exchanged molecules are low active for this catalytic reaction, and a similar effect
would possibly account for differences found in the series of compounds studied by
Iliev et al. [98]; this would nicely explain the small differences found by these
authors when using water or water:DMF as solvent for preparing the samples, as
the water:DMF mixture would permit a better molecular dispersion of the complex
in the hydrotalcite host.
Further evidences of the effect of layer charge density on the orientation of the
hosted phtalocyanine and porphyrine complexes has been obtained from UV–vis
and ESR studies of these complexes intercalated in hectorite, fluorhectorite and a
hydrotalcite-like LDH, three layered materials with markedly different charge layer
densities [105]; it has been also found that the hydration state of the galleries
modifies the orientation of the hosted anions. So, Co(II) tetrakis(N-methyl-4-pyri-
diniumyl)porphyrin (CoTMPyP) was intercalated in low layer charge density (80 Å2
charge − 1) hectorite and in large layer charge density (27 Å2 charge − 1) fluorhector-
ite by reacting the clay with an aqueous solution of the complex, under nitrogen,
and different loadings were achieved. Co(II) tetrasulphophthalocyanine (CoPcTs)
anion was intercalated in Mg,Al LDH (layer charge density 25 Å2 charge − 1 [106])
following a method similar to that described by Pinnavaia et al. [76,93]. The basal
spacing for the hectorite intercalate was 14.05 Å, changing to 14.00 Å after
dehydration and outgassing in vacuum; thus, the complex plane should be very
likely lying parallel to the clay layers. The basal spacing was slightly lower
(12.6–13.8 Å) for the low loaded complex-clay systems, values similar to those
reported for montmorillonite-hosted CoTMPyP [80]; such a basal spacing decrease
at low loadings of large cations is commonly observed for smectite clays and
attributed to irregular stratification and/or domain formation in the interlayers
[107]. For the fluorhectorite compound, the basal spacing was 19.6 Å, decreasing to
17.6 Å upon dehydration. These values are consistent with either a face-to-face
bilayer or a monolayer with the porphyrin being tilted at an angle (35° has been
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 83

Fig. 8. Arrangement of Co(II)-tetrakis(N-methyl-4-pyridinium)porphyrin (CoTMPyP) in (a) hectorite

and (b) fluorhectorite, and (c) of cobalt(II) tetraulfophtlalocyanine (CoPcTs) in a layered double
hydroxide. Adapted from Ref. [110].

calculated [108]) to the clay layers. Finally, the basal spacing was 23.5 Å, and did
not change upon dehydration, for the LDH-hosted CoPcTs, a value similar to that
reported by these and other authors [76,93,95,98], Fig. 8.
While the UV – vis spectra of the hectorite and fluorhectorite systems do not
change sensibly upon exposure to O2 (suggesting that intercalated Co macrocycles
do not form adducts with O2, as observed in solution [109]), small changes in the
positions of the Soret and Q bands for the CoPcTs–LDH compound have been
related to changes in the electron density of the conjugated p orbitals of the
macrocycle ring caused by confinement in the interlamellar space of the LDH, by
interactions between neighboring CoPcTs and by changes in the inductive effect of
the sulfonate groups.
ESR results are consistent with the XRD results, for the CoTMPyP–fluorhector-
ite compound in the hydrated state, if the molecular plane is tilted 27° with respect
to the clay layers, and with one or two water molecules coordinating to the Co(II)
along the z-axis. Upon dehydration, water molecules are removed, and the hosted
84 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

molecules form a bi-layer, with the molecular plane oriented parallel to the clay
layers, a similar disposition as that concluded for CoTMPyP–hectorite, although in
this case only a monolayer of hosted molecules is formed, without coordinating
water molecules. The orientation of CoPcTs in the LDH gallery could not be
concluded from anisotropic ESR measurements, as no oriented films could be
obtained. From XRD measurements, similar conclusions to those by Carrado et al.
[95] were achieved, i.e. a perpendicular arrangement of the complex in the LDH
gallery; similar spacing values could be explained by a non-favorable trilayer
arrangement of the anions parallel to the LDH plane, but in this case the charge
layer density would not be matched, and the spacing would be expected to decrease
upon dehydration, but this was not observed. The ESR spectrum of CoPcTs–LDH
was similar to that of air-dried CoTMPyP–hectorite, so further confirming the
slightly tilted ‘upright’ orientation; however, it changes dramatically upon vacuum
dehydration, with a weak ESR signal at g: 2, probably because aggregation of the
hosted molecules upon water removal; this process is reversible upon rehydration
and water diffusion into the interlayers.
From these results, these authors conclude [105] on the suitability of the
fluorhectorite and LDH compounds, but not the hectorite-hosted one, for catalytic
reactions taking place on the Co(II) sites, accesible to reactant molecules for
electron-transfer reactions, behaving as biomimetic catalysts, as observed for 2,6-di-
tert-butylphenol oxidation [94].
These three host – guest systems have been also tested for reductive dechlorination
of carbon tetrachloride [110], and the results compared with those obtained for the
same Co compounds in homogeneous conditions, and by silica-supported
CoTMPyP. While under homogeneous conditions, both chloroform and
dichloromethane were formed, only the former was observed under heterogeneous
conditions, accounting for less than 30% of degradated CCl4, which is consistent
with previous studies of degradation by Co macrocycles [111–115]. Degradation of
CCl4 follows a first-order kinetics. Under homogeneous conditions, the initial rate
rapidly decreases, indicating deactivation of the catalyst, probably because aggrega-
tion in aqueous solution [116], but as aggregation is hindered in the clays gallery,
activity is maintained. The lack of formation of dimethylmethane has been at-
tributed [110] to a change in the reduction potential of the supported macrocycle
because of the charged layers. Initial rate constants for heterogeneous CCl4
degradation (as calculated along the first 30 min of reaction) decreases when using
the supports silica-gel\LDH \ fluorhectorite\hectorite. This decrease is in agree-
ment with accessibility of the reactant molecules to the Co(II) sites, as concluded
from XRD and ESR studies [94,105] for the layered materials. Overall, the
dechlorination reaction in these heterogeneous systems is very similar to enzyme-
catalyzed reactions, and the initial degradation rate (R0) can be fitted by the
Michaelis-Menten kinetic model:
R0 = nmax[CCl4]0/([CCl4]0 +Km) (6)
where nmax is the maximum reaction rate for a specified initial Co macrocycle
concentration and Km is the Michaelis constant. nmax values decrease for the
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 85

catalysts supported on silica-gel \LDH\ fluorhectorite\ hectorite, i.e. as the ini-

tial reaction rate does, and in agreement with studies on the orientation and
hydration of the macrocycles in the interlayers [105].
Zikmund et al. have recently proposed an alternative route to the synthesis of
LDH-hosted coordinatively unsaturated Co(II) complexes, consisting in the in situ
synthesis of the complex inside the LDH gallery [117]. The route is rather different
to those described above. First of all, these authors prepare the Mg,Al–carbonate
LDH by reaction of solid magnesium carbonate hydrate with an aqueous solution
of sodium aluminate and sodium hydroxide. After calcination at 500°C in air for 5
h of the washed solid, it was suspended in carbon dioxide-free distilled water,
leading to the reconstructed OH-containing LDH (meixnerite). Exchange of the
interlayer hydroxyl groups by amminocarboxylates (L-glutamate and isonicotinate)
was achieved in a glycol:water (2:1) mixture, to facilitate swelling of the solid.
Suspension of this solid in an ethylene glycol solution of CoCl24 − led to scavenging
of Co ions in the galleries in the form of ion pair species, and hopefully intercalated
CoCl3 –glutamate (or CoCl3 – isonicotinate) species. Mixing with Schiff bases salen
or salophen under argon led to intercalated coordinatively unsaturated square
planar Co(II) chelates. The d – d diffuse reflectance spectra of these compounds
were similar to those of the analogous complexes in solution and in Y-zeolite cages.
Their basal spacings were 13.1 and 11.6 Å, respectively, for the salen and salophen
derivatives, thus suggesting the anions have their molecular plane parallel to the
LDH layers. When air was contacted with the Co(salen) or Co(salophen) deriva-
tives, the ESR spectrum showed two signals: a narrow one (signal I) with dH = 120
mT showing hyperfine structure, and a second signal (II) with g = 2.02 and
dH = 950 mT. On heating at 300°C, signal II is removed, but it is restored after
decreasing the temperature. This oxygenation–deoxygenation process was undefin-
itely reversible. As no 59Co hyperfine splitting was observed in signal I, it was
attributed to superoxide species interacting with Al3 + ions, while signal II strongly
resembles that observed for other immobilized Co(salen)–dioxygen complexes.
Macrocyclic complexes of other first-series transitions cations have been also
exchanged into the interlayer space of LDHs. Mansuy et al. [75] have reported the
preparation and catalytic properties of Mn(III) porphyrins (TSPP, dianion of
meso-tetrakis-(4-sulfonatophenyl)porphyrin, and TDCSPP, dianion of meso-te-
trakis-(2,6-dichloro-3-sulfonato-phenyl)porphyrin) supported on LDHs and other
supports, for alkane and alkene oxidation. The complexes were hosted via the
reconstruction method of previously calcined (450°C) Mg,Al–LDH in its carbonate
form; the complex was not released in the presence of organic solvents (e.g.
dichloromethane, methyl cyanide, methanol), and its UV–vis spectra showed the
expected Soret band for Mn(III) porphyrins. Upon comparison with these same
Mn(III) porphyrins supported on alumina, the LDH-supported one showed a
rather low catalytic activity for heptane hydroxylation, but it gave the highest
regioselectivity for hydroxylation at the terminal positions. Manganese porphyrins
intercalated-LDHs have been also studied as electrode modifiers [118].
A rather complex system containing LDHs has been described by Robins and
Dutta for photochemical processes [119]. These authors have prepared a LDH
86 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

photochemical assembly consisting of a Li,Al–LDH hosting titanium dioxide and

myristate anions, together with [tetrakis-(4-carboxyphenyl)porphyrinato] zinc(II),
(ZnTPPC). First, the Li,Al – LDH anionic clay interlayered with myristate
(CH3(CH2)12COO − ) was prepared; the interlayer spacing was 21 Å, consistent with
a monolayer of anions in all-trans configuration. Upon treatment with an hexane
solution of titanium butoxide the interlayer spacing increased to ca. 25 Å, and the
XRD peaks became slightly broader; controlled hydrolysis led to interlayer TiO2
particles, the diffraction peaks sharpening up somewhat and with similar spacings
as to the original Li,Al – myristate. Finally, incorporation of Zn(TPPC) was
achieved by partial anionic exchange of the myristate anions; the diffraction pattern
was similar as for the Li,Al – myristate material, but, considering that the porphyrin
dimension is ca. 18 Å (smaller than the spacing measured for the myristate LDH),
the porphyrin molecule could be held with its plane perpendicular to the metal-hy-
droxide layer, as previously reported [76,95,105] for phtalocyanines in Mg,Al
LDHs. The emission spectra of the solid (excitation wavelength 406.7 nm) resem-
bles that of Zn(TPPC) in water and ethanol, where no aggregation of the porphyrin
molecules exists [120,121], suggesting the complex is dispersed, and not forming
aggregates in the interlayer space of the LDH. The Soret band broadens and
slightly shifts from 425 to 420 nm, but this changes cannot be due to aggregation
(as concluded from the emission spectra results), and so should be due to interac-
tion in the interlayers, probably by hydrogen bonding to the framework hydroxyl
groups, and costrained orientation of the benzoic acid rings, in a similar fashion as
that described by Pinnavaia et al. [105] for clay-hosted cationic porphyrins. Further
shift to 410 nm is observed upon incorporation of TiO2, probably because hydrogen
bonding, as also concluded from Raman spectroscopy measurements.
This intercalated porphyrin promotes viologen (heptyl viologen and propyl
viologen sulfonate, PVS) reduction upon excitation, in the presence of EDTA as
sacrificial electron donor, Fig. 9. This scheme implies that the photochemical
reaction takes place near the edge of the LDH particle, quite reasonably, as TiO2
is formed by hydrolysis of titanium tetrabutoxide by ambient moisture, and water
is not expected to penetrate deep into the LDH, that had become rather hydropho-
bic upon exchange with myristate anions.

3. Systems hosting oxometalates

Polyoxometalates with several nuclearities have been incorporated into the

interlayer space of different LDHs. These studies have been carried out on LDHs
with a wide variety of cations in the layers, but, in relation to the nature of the
intercalated anion, it is limited by the chemistry of metal ions forming oxometa-
lates. Derivatives of chromium and vanadadium are those most widely studied,
although several reports have been also published on molybdenum and manganese
compounds. On the other hand, the number of papers published on systems
containing iso- and hetero-polyanions with a Keggin-like or related structure, is
rapidly increasing in recent years. General reports on the preparation of polyox-
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 87

ometalates – intercalated LDHs and their applications have been published [84,122–
124]. In this section, classification has been made according to the nuclearity and
nature of the interlayer anion.

3.1. Layered double hydroxides intercalated with low-nuclearity oxometalates

The reactivity of simple oxometalates, such as chromate and dichromate, inside

interlamellar domains of Zn,Al or Zn,Cr, or Cu,Cr–LDHs after ageing under
moderate thermal treatment, has been studied. Pillaring and grafting processes have
been put into evidence from structural and spectroscopic data for these compounds,
and their application in several catalytic processes has been also tested.
Layered double hydroxides hosting intercalated CrO24 − have been prepared by
Miyata and Okada [72] by coprecipitation (direct synthesis). The bright-yellow solid
prepared, Mg,Al – CrO24 − , contains, however, ca. 2% intercalated carbonate, incor-
porated during the synthesis or drying step. As already mentioned, total exclusion
of carbonate is rather difficult to achieve, especially when metals forming highly
basic hydroxides (e.g. Mg2 + ) are used. Consequently, the ratio between the total
negative charge provided by interlayer anions (carbonate and chromate) and total
positive charge in the layers (due to aluminum) was 1.13, instead of the expected
value of 1.0, suggesting that the excess anions are strongly adsorbed on the
positively charge surfaces. The XRD spacing for planes (003) was 8.94 Å, corre-
sponding to an interlayer space of 4.06 Å. This value is lower than that reported by
Miyata et al. [125] for an LDH intercalated with perchlorate (interlayer space 4.38
Å), despite the size of CrO24 − and ClO4 − are almost the same; the difference has

Fig. 9. Scheme of a layered double hydroxide photochemical assembly. Adapted from Ref. [119].
88 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

been attributed by these authors [72] to a stronger interaction of the divalent

chromate anions than of the monovalent chlorate anions with the layers. The IR
spectrum band n1 of chromate, IR forbidden for a pure Td symmetry, is recorded
at 865 cm − 1, while band n3 splits into three absorptions at 845, 885, and 950 cm − 1,
confirming a symmetry close to C26 for the CrO24 − anion in the interlayer space of
the LDH. On calcination, removal of interlayer water up to 350°C gives rise to a
small decrease in d(003), but at 400°C the layered structure collapses, forming
MgO, and above 800°C the spinel MgAl2 − x Crx O4 is detected.
The reconstruction method, from a Mg,Al–LDH calcined at 450°C, has been
applied by Chibwe and Jones [32] to intercalate Cr2O27 − by suspending the powder
in a decarbonated solution of K2Cr2O7. No interference of carbonate is reported,
and the interlayer space was 10 Å, with the C2 axis of the guest species oriented
perpendicular to the brucite layers. However, a value of 8.4 Å has been reported by
Misra and Perrotta [126] for a Mg,Al –LDH prepared following the same method,
but using Na2Cr2O7 instead of the potassium salt, as the precursor; this value is
fairly close to that reported by Miyata and Okada [72] for intercalation of
chromate, and so it is possible that dedimerization of the dichromate has taken
place during synthesis.
A similar, although slightly lower, value of 8.25 Å has been reported by Besse et
al. [127] for a Cu,Cr – LDH intercalated with CrO24 − . Layered double hydroxides
containing Cu or Zn in the brucite like layers are not usually prepared by the
conventional methods described in the literature [2,3] because their hydroxides do
not possess the brucite-type structure; instead, they are prepared by adding an
aqueous solution of a salt of the trivalent cation (Cr(III) chloride in this case) to a
suspension of the oxide of the divalent cation (CuO), and further Cl − /CrO24 −
anionic exchange (No pH control is mentioned by Besse et al. [127]). The thermal
properties of this compound are worthwhile to be analyzed in detail. Upon
calcination at 300°C, the weight loss recorded is attributed to dehydration (removal
of interlayer water molecules) and also dehydroxylation (from the brucite-like
layers), the layered structure collapsing, and forming CuCr2O4 and CuO at 520°C.
This behavior is rather similar to that observed for other LDHs. However, when
the calcination is performed at 80°C, the interlayer spacing decreases from 8.25 to
7.32 Å, while only a small decrease (due to dehydration) is observed for the chloride
analogous in the same temperature range. For the chromate compound, the
decrease in spacing is not consistent with dehydration only, taking also into account
that the size of the chromate tetrahedron is larger than that of the chloride anion,
and so these authors suggest some bonding (grafting) between chromate and the
layer. This was confirmed after observing that despite CrO24 − /Cl − exchange is
observed for the uncalcined Cu,Cr – CrO24 − , such an exchange is not observed after
calcination at 80°C.
Grafting has been also observed for the dichromate analogous [128] prepared by
ionic exchange at pH 4.5 (the chromate compound is obtained at pH 8.5). EXAFS
and XANES studies [129] show that the local environment of the copper cations is
not affected by the exchange, and consists of a strongly tetragonal elongated
distorted octahedron. Storing conditions of the exchanged materials are important
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 89

Fig. 10. Effect of heating on interlayer spacings of Cu,Cr – X LDHs (X = Cl − , CrO24 − , Cr2O27 − ).
Reprinted from C. Depège, C. Forano, A. de Roy, J.P. Besse, [Cu – Cr] layered double hydroxides
pillared by CrO24 − and Cr2O24 − , Mol. Cryst. Liq. Cryst. 244 (1994) 161 – 166, © 1994, with permission
from Gordon & Breach Science Publishers SA.

in determining the interlayer space, due to the ability to adsorb water in the
interlayer space, as confirmed by thermogravimetric and elemental chemical analy-
ses. The spacings determined from X-ray diffraction were 8.95 Å and 8.42 Å,
respectively, for the dichromate and chromate compounds, these values being
markedly lower than those reported by Chibwe and Jones [32] for a Mg,Al–
Cr2O27 − LDH. These values, however, are close to that reported by Miyata and
Okada [72] for the chromate form, and so a dedimerization could happen. Both
phases undergo a slow evolution to contracted forms with basal spacings of 7.68 Å
(chromate) and 7.87 Å (dichromate) upon water removal, Fig. 10, and these
extremely low values cannot account for free oxoanions, but evidence an effective
pillaring on the hydroxylated sheets. The difference with the behavior observed with
other chromate-containing LDHs has been attributed [128] to the specific chemical
properties of copper, which can form a wide range of lamellar copper hydroxides
Cu2(OH)3A (A =Cl − , ClO4− , NO3− , MnO4− , etc.) [130] with short interlayer
distances. However, such pillaring is reversible, as evidenced by the ability to
exchange by chloride in the presence of an excess of KCl. A decrease in the basal
spacing to 7.10 and 7.30 Å is observed for the chromate and dichromate forms,
respectively, upon heating at 80°C [128], the decrease is not reversible, and the
anions cannot be further exchanged, indicating that this rather soft treatment has
been able to anchor the anions to the layers. Further heating at 200–300°C destroys
the lamellar structure, as confirmed by X-ray diffraction and XAS measurements
[129].
Similar spontaneous contraction of the layers has been also described for Zn,Al
and Zn,Cr – LDHs intercalated with CrO24 − and/or Cr2O27 − , but only after ageing
90 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

or thermal treatment [131]. For Zn,Al–CrO4 and Zn,Cr–Cr2O7 a decrease in the

basal spacings from 8.61 to 7.86 Å and from 9.20 to 7.34 Å, respectively, is
observed upon ageing the solid it its mother liquor, values too low to be compatible
with free interlayer anions, despite they were easily exchanged for chloride. How-
ever, after treatment at 150°C for 24 h, exchange was not achieved, and the basal
spacings decreased to values in the range 6.80–7.20 Å. From these and former
studies [127,128], Besse et al. conclude that in the contracted phases formed prior
to thermal treatment, also described as pre-grafted phases, the interlayer anions
become close to the metal cations in the layers, by approaching one oxygen atom
of the interlayer tetrahedron to one of the triangular voids of the hydroxide layers,
so decreasing the interlayer space. On heating, as water molecules are removed,
some of the hydroxyl groups are removed as well, effective anchoring of the anions
taking place, Fig. 11.
Aldol condensation between formaldehyde and acetone to yield methylvinyl
ketone has been studied by Suzuki and Ono [132] on several hydrotalcites calcined
at 500°C. Although the most effective precursor was that containing Mg,Al–CO3,
a selectivity of 100% was reached upon CO23 − /CrO24 − exchange.
Oxidation of alcohols in non-polar media at r.t. has been achieved by perman-
ganate anions intercalated in Mg,Al – LDH [133]. The reaction proceeds selectively
to the corresponding aldehyde or carboxylic acid, depending on the starting

Fig. 11. Two ways of grafting of ditetrahedral anions (Cr2O27 − , V2O27 − ) to layered double hydroxides.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 91

alcohol. The selectivity towards benzaldehyde during oxidation of benzyl alcohol

has been explained based on the low ability of the LDH to adsorb the aldehyde.

3.2. Layered double hydroxides intercalated with medium-nuclearity oxometalates:

6anadates and molybdates

3.2.1. Vanadates
Among all metal-containing anions incorporated into the interlayer space of
hydrotalcites and other LDHs, oxovanadates represent the widest studied group.
Although mostly as decavanadate, some studies have been also devoted to interca-
lation of lower oligovanadates.
Oxometalate-pillared LDHs are in many cases prepared by a two-step anion
exchange method, through intermediate intercalation of a large organic anion, to
swell the brucite-like layers [65,134].
The polymerization degree of oxovanadates is pH-dependent, nuclearity increas-
ing as pH decreases. For a 0.1 M aqueous vanadate solution, these equilibria are as
follows:
V10O26(OH)2− 4, V10O27(OH)5 − , V10O628− , decavanadate,
pH =1 −3 l
lVO(OH)3, VO2(OH), V3O39 − , V4O412− , metavanadate,
pH = 4 −6 l
VO3(OH)2 − , HV2O37 − ,V2O47 − , pyrovanade, pH= 8−11l
lVO34 − , vanadate, pH \12.
In one of the pioneering works on LDH-intercalated vanadates, Twu and Dutta
[135] prepared Li,Al – LDH with different oxovanadate oligomers by ion exchange
of the Li,Al – Cl precursor with NH4VO3 aqueous solutions at different pH; it was
expected anionic exchange and intercalation of the oxovanadate more stable at each
pH. However, at pH larger than 13, selective Cl − /OH − exchange, but not
chloride/vanadate, was observed. When the pH was lowered to 8–11, complete
chloride/vanadate exchange was achieved, the basal spacing increasing from 7.8 to
10.8 Å. Although a compound with the same gallery height (6.0 Å, once the
thickness of the brucite-like layer, 4.8 Å, is considered) is obtained after exchange
at pH 5– 6, the Raman spectra of both samples were completely different. Finally,
at pH 3– 4, only partial exchange was achieved, as a vanadate/chloride competition
exists, because of the use of HCl to attain this low pH values. As different interlayer
anions could give rise to the same gallery height (ca. 6 Å), thus making impossible
to ascertain the actual nature of the interlayer anion, this was studied by Raman
spectroscopy, based on results by Griffith et al. [136,137] for oxovanadate species.
The most prominent Raman band of vanadate species, due to V–O stretching,
occurs in the 800 – 1000 cm − 1 range, shifting to higher wavenumbers as polymeriza-
tion and/or branching increases. From this, it was concluded that at pH 10 the
92 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Fig. 12. Two possible orientations of V2O27 − in layered double hydroxides. Adapted from Ref. [135].

interlayer species is V2O47 − (also the major component of the solution at this pH),
with the V – V axis parallel to the layers, Fig. 12(b), and also hosting interlayer
water molecules; at pH 5 – 6 the interlayer species is V4O412− , with a similar gallery
height, while at lower pH the predominant species still is V4O412− , with only a minor
amount of V10O628− , despite the decavanadate can be completely exchanged in other
hydrotalcites (see below).
On the other hand, the average charge per V atom changes from one oxovana-
date to another. Taking into account that the negative charge of the interlayer
anion should balance the positive charge of the Al-containing brucite-like layers,
Bhattacharyya et al. [138] have prepared, through a careful control of pH during
reaction and of the NaVO3/NaOH ratio in the starting solution (from 1:7 to 1:3),
Mg,Al and Mg – Zn,Al – LDHs containing different oxovanadate oligomers by a
single-step method. In all cases, an aqueous solution of the Mg and Al (or Mg, Zn,
and Al) cations was dropwise added to the NaVO3/NaOH solution at the desired
pH (10.8 for V2O47 − and 8.3 for V4O412− ), the gelatinous mixture being heated for
several hours at 80 – 90°C. The basal spacings determined by X-ray diffraction were
10.5 Å for the pyrovanadate, and 9.7 Å for the cyclotetravanadate.
The V2O47 − and HV2O37 − anions are constituted by two [VO4] tetrahedra sharing
a vertex, and in the interlayer space, two different arrangements are possible [139],
one with the V – V edge perpendicular to the brucite-like layers, with a calculated
basal spacing of 12.6 Å (from the size of the V2O47 − anion derived from single
crystal data for Mg2V2O7 [140]), and another where it is parallel to the layers, with
a calculated value of 9.8 Å, Fig. 12, a situation similar to that described by Chibwe
and Jones [32] for intercalated Cr2O27 − ; in addition, the presence of water molecules
can enlarge such a spacing. On the other hand, the smallest dimension of V4O412− is
expected to be the same as V2O47 − [138], and these authors conclude the anions
should be in all cases in a ‘parallel’ disposition.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 93

Similar studies have been carried out by Besse et al. [141] in a Cu,Cr–LDH,
analyzing also the role of swelling agents to favor the exchange and incorporation
of polyoxometalate species. At pH 10 –11, V2O47 − is selectively exchanged, whereas
as the pH is lowered, the major interlayer species are V2O412− (pH 6–7) and V10O628−
(pH 4–5). These authors conclude that the size of the precursor anion used to swell
the layers (terephtalate or dodecylsulphate), and hence the values of the interlayer
distances, must favor the intercalation of vanadates of similar hinderance.
Delmas et al. [142,143] have recently proposed an alternative route, by the
so-called ‘chimie douce’ (soft chemistry), to insert metavanadate in an LDH, Fig.
13. These authors claim the method overcomes the problem, usually found, that
chemical composition of LDH shows fluctuations due to the different pH at which
precipitation of M(OH)2 and M(OH)3 hydroxides occurs, thus leading to a chemical
composition determined by the intrinsic stability of the solid, rather than by the
composition of the starting solution. By high temperature synthesis methods these
authors obtained a layered NaNi1 − y Coy O2 solid, with a layer–layer distance of
5.18 Å; an oxidizing hydrolysis with NaClO and KOH leads to an expansion to
7.08 Å and insertion of potassium ions, followed by reduction in H2O2/NH4VO3
solution. The asymmetry of the X-ray diffraction lines usually indexed as due to
(101) and (111) planes, in some cases attributed to a turbostratic-like character
(parallel and equidistant layers disoriented with regard to one another along the
c-axis), has been attributed by these authors to local distortions within each layer,
probably related to a misfit of the oxygen in the layers (O–O distance 3.04 Å) and
in the metavanadate chains (2.91 Å). The interlayer spacing of the vanadium-con-

Fig. 13. Scheme showing the successive reaction steps involved in the preparation of a LDH by ‘chimie
douce’. Reprinted from K.S. Han, L. Guerlou-Demourgues, C. Delmas, A new metavanadate inserted
layered double hydroxide prepared by ‘chimie douce’, Solid State Ionics 84 (1996) 227 – 238, © 1996, with
permission from Elsevier Science.
94 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Fig. 14. Orientation of decavanadate anion, V10O628− , in the interlayer space of layered double
hydroxides, with its ‘main’ C2 axis parallel to the brucite-like layers.

taining material was 9.15 Å, and from IR data, the presence of polymeric,
metavanadate species, (VO3)nn − (consisting of [VO4] groups with a C26 symmetry)
was concluded, in agreement also with chemical analysis, which suggest one
negative charge per vanadium ion, while the presence of cyclic metavanadate
entities, such as V3O39 − or V4O412− , was excluded.
Pinnavaia et al. have shown [134] that it is possible to introduce polyoxovanadate
ions as pillars in LDHs containing Zn,Al or Zn,Cr or Ni,Al in the layers, in the
chloride form. At pH 5.5 – 10 a byproduct containing V4O412− was formed, but at pH
4.5 total exchange with V10O628− was observed, with solids possessing a basal
spacing of 11.9 Å, corresponding to a gallery height of 7.1 Å and a decavanadate
orientation in which the C2 axis is parallel to the host layers, Fig. 14. Retention of
the structure of the interlayer anions was confirmed by 51V MAS-NMR spec-
troscopy, while EXAFS studies at the Zn K-edge have shown [144] that no
structural distortion occurs for the brucite-like host lattices upon intercalation, in
agreement with electron microscopy studies which show that the exchange reactions
are topotactic. Exchange from the nitrate form has been reported by Woltermann
[145]. Photocatalytic oxidation of isopropyl alcohol to acetone was achieved on the
Zn,Al–V10O28 LDH in the presence of oxygen, this catalyst being more active than
the homogeneous catalyst, despite scattering of light by the host particles [134].
LDHs containing Zn and Al in the layers have been also prepared with interlayer
vanadates, by ionic exchange of chloride or carbonate precursors [146]. It has been
found that the Zn/Al ratio in the final Zn,Al–vanadate LDH decreases when the
pH is lowered, probably due to selective dissolution of Zn2 + . Vanadium K-edge
XAS data show that in samples prepared at low pH the major interlayer species is
V10O628− , while as the pH is increased, interlayer vanadates consist of tetrahedral
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 95

[VO4] units, and it has been even possible to estimate the fraction of V5 + ions
existing as decavanadate, tetravanadate or tetrahedra chains.
Drezdzon [65] has proposed an alternative method to prepare LDHs exchanged
with large oxometalates, by intermediate preparation of organic-exchanged materi-
als. It was expected that, since hydrotalcite-like materials have higher charge
density than cationic clays, they would be more difficult to swell and exchange. A
Mg,Al–terephthalate LDH was obtained from Mg and Al nitrates and terephtha-
late in basic (NaOH) medium, with a basal spacing of 14.4 Å (7.8 Å for the
carbonate form). After addition of this solid to an aqueous NH4VO3 solution at pH
4.5, the basal spacing of the layered material obtained was 11.9 Å (coincident with
the value reported by Pinnavaia et al. [134]). Acidification plays a double role: (i)
polymerization of the metalate, and (ii) protonation of the organic anions, so
making easy its removal from the interlayer space; however, it has been also
claimed [147] that the process is inhibited by the poor solubility of the organic acid
in water, and the resulting difficulty in its removal from the clay matrix.
A third method already mentioned to prepare other LDHs, different from
exchange and pre-swelling with organic anions, was used by Jones et al. [148], from
the known ability of calcined LDHs to recover the layered structure. An Mg–Al
LDH calcined at 450°C for 18 h, suspended in an aqueous solution of NaVO3,
recovers the layered structure, hosting decavanadate species, as concluded from a
basal spacing of 11.8 Å, after acidification with HCl at pH 4.5.
All these methods require the use of carbon dioxide-free conditions, in order to
avoid incorporation of carbonate anions in the interlayer region. For acidic
systems, such as the Zn,Cr – LDH system, Kooli and Jones have reported a direct
method for the synthesis of decavanadate-containing Zn,Cr–LDH [149], at a pH
where carbonate is not present in the solution. A solution containing Zn and Cr
chlorides was added to an aqueous solution of NaVO3 (pH 4.5), and the slurry
obtained aged overnight at 55°C. The basal spacing for (003) planes was 11.89 Å,
suggesting again an orientation of the decavanadate anion with the main C2 axis
parallel to the host layers. Expanded structures with interlayer V10O628− anions were
obtained for solutions with Zn/Cr ratios between 1 and 5, although the a parameter
(related to the average cation – cation distance in the brucite-like layers, and hence,
on the ionic radii of these cations and their nature and concentration) remains close
to 3.11 Å, whichever the starting Zn/Cr ratio, a result similar to that previously
reported by de Roy et al. [1]. However, decavanadate was incorporated even at pH
6.5 (if the decavanadate solution had been prepared at pH 4.5), but not at higher
pH, probably due to the preferred formation of other oxovanadate oligomers.
These authors also succeed to prepare by this direct method decavanadate interca-
lated Zn,Al LDHs, avoiding the use of ZnO.
Exchange of decavanadate by carbonate in the interlayer space of Mg–Al LDH
was easily achieved by ultrasonic treatment of a suspension of the carbonate form
in a vanadate solution at the pH required to polymerize VO3− , but without any
further pH control [150]. It is likely that decavanadate exchange is facilitated by the
high dispersion of the agglomerated particles following ultrasonic treatment, and
also by enhanced diffusion of decavanadate on temporary delaminated particles.
96 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Conductivity measurements have shown [151] that chloride/decavanadate exchange

follows a first-order mechanism, without defoliation of the layered material.
In many of the synthesis leading to oxometalate–LDH compounds, in addition
to the layered material containing decavanadate and characterized by a basal
spacing close to 12 Å, an additional non-layered material, characterized by a broad
diffraction peak close to 10 Å (in some cases, obscuring the first harmonic of the
layered material), is formed [149,152 – 154], and is thought to be due to a compound
formed as a result of the reaction between the basic hydrotalcite and the acidic
polyoxometalate.
Anionic exchange of decavanadate for carbonate in LDHs containing Ni–Al or
Mg–Al, with variable layer charge (MII/MIII ratio ranging from 2 to 6) was
attempted by Kooli et al. [155] by direct exposure of the LDH to a vanadate
solution at pH 4.5. For the carbonate precursors, increasing amounts of Al led to
a decrease in the a-parameter (lower ionic radii for Al3 + than for Mg2 + or Ni2 + )
and a decrease in the c-parameter (lower electrostatic interaction between the
increased positive layer charge and the carbonate anion [156]). Upon exchange, the
crystallinity of the material decreases for M(Mg or Ni)/Al ratios larger than 4, and
in the case of the Mg,Al LDH, the non-layered by-product, characterized by the
diffraction peak close to 10 Å, also develops. It is expected that the V content
depends on the M/Al ratio, and this is true for the Ni–Al series; however, it
remains constant for the Mg – Al system, i.e. the exchange process results in a
change in the Mg/Al ratio. From the value of the a-parameter, a value of 2 was
calculated for the Mg/Al ratio, suggesting a dissolution of octahedral cations until
a layer composition close to this value is obtained.
The synthesis of decavanadate-containing Mg–Al LDHs was studied by Ulibarri
et al. [157,158] following different routes, Fig. 15: exchange of the initial carbonate
or therephtalate anions, as well as reconstruction (directly to the vanadate form, or
via intermediate terephthalate) of the layered structure from the carbonate form
previously calcined at 550°C, also checking the role of preswelling with glycerol. In
all cases a layered material containing V10O628− was obtained (and, when used,
therephtalate or glycerol were completely removed), together with the non-layered
solid characterized by a XRD peak at 10 Å, but direct reconstruction of the
calcined carbonate led also to formation of a fibrous material, and partial dissolu-
tion of Mg. The main difference in the properties of the solids obtained by the
different routes corresponds to the pore size distribution, a narrow distribution
being present in samples prepared via the terephthalate intermediate. The use of
large organic anions as pre-swelling purposes to intercalate polyoxometalates has
been claimed [50] as the most promising method for creating stable pores, avoiding
the formation of sidephases. Similar methods were successfully used also by Chisem
and Jones [159] to prepare decavanadate-containing Li–Al LDHs.
The Ni – Al LDH containing carbonate has been also pillared with polyvanadate
at pH 4.5 – 5.5 by anionic exchange, without any swelling agent [160,161], and also
by reconstruction of the carbonate form after calcination. However, exchange was
only achieved when the Ni,Al LDH was maintained in solution (not dried) and the
vanadate solution at pH 4.5 – 5.5 added, and not when a dried Ni–Al LDH was
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 97

used. The same effect of the drying step was observed by Carrado et al. [95] for
CuPcTs exchange in a Mg,Al – LDH. At pH 8.5 no exchange was observed, but at
intermediate pH a biphasic material, with characteristic XRD peaks at 7.54 Å
(carbonate) and 11.7 Å (decavanadate), was obtained. With regards to the recon-
struction method, this was valid only if the sample was submitted to hydrothermal
treatment at autogenous pressure, and when the precursor had been calcined at
300°C, but for higher calcination temperatures, NiO was also formed. According to
Clause et al. [162], as the calcination temperature is increased, Al3 + ions migrate
from the mixed oxide phase formed upon carbonate removal, to the crystallite
surface, where from they are dissolved when the solid is suspended at pH 4.5; so,
the nickel in excess does not enter the reconstructed LDH structure, but remains as
an independent NiO phase. The intensities of the NiO XRD peaks decrease (and
those of the LDH material increase) as the temperature during hydrothermal
treatment is increased in the 80 – 150°C range. While at pH 4.5–5.5 reconstruction
leads to the decavanadate – LDH phase (together with the byproduct responsible for
the XRD peak at 10 Å), when the pH during reconstruction was increased the IR
and XRD data indicate formation of (VO3) chain-like polymeric metavanadate,
composed of [VO4] tetrahedra with C26 symmetry, and with the longitudinal chain
axis parallel to the host layers.
Similar studies were also carried out with the Mg,Cr and Ni,Cr systems [161,163],
but decavanadate-containing LDHs were obtained only when following the ex-
change method using pre-wet carbonate–LDH, and at pH lower than 6.5 for

Fig. 15. Diagram showing various routes followed for the synthesis of terephthalate and vanadate
materials. Reprinted from M.A. Ulibarri, F.M. Labajos, V. Rives, R. Trujillano, W. Kagunya, W.
Jones, Comparative study of the synthesis and properties of vanadate-exchanged layered double
hydroxides, Inorg. Chem. 33 (1994) 2592–2599, © 1994, with permission from The American Chemical
Society.
98 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Mg,Cr, but 5.5 for Ni,Cr. In other words, not only pH, but also the intrinsic nature
of the layer cations seems to play an important role on the ability to exchange
decavanadate for carbonate. With regards to vanadate-containing LDHs prepared
following the reconstruction method, differences are also observed depending on
the nature of the cations in the brucite-like layers. So, the Ni,Al and Mg,Cr systems
are easily reconstructed if the carbonate precursor had been calcined below 300°C,
while after calcination at 400°C only an amorphous material was obtained after
contacting the calcined product with the vanadate solution. However, the Ni,Cr
calcined precursor does not reconstruct at all, and the spinel (NiCr2O4) phase is
detected by XRD in the precursor calcined at 500°C. Probably, the additional
stability of the calcined product because of crystal field effects when containing
transition metal cations somewhat hinders recovering of the layered structure with
interlayer vanadates.

3.2.2. Thermal decomposition

Evolution of intercalated decavanadate to other species by thermal treatment has
been widely studied, as interest in these polyoxometalate –hydrotalcite materials
also stems from their potential use as catalytic materials, as prepared or after
thermal decomposition [123,164]; already in 1984, they were reported to be useful
for exhaust gas and hydrocarbon conversion processes [145]. Kagunya and Jones
[165] have reported aldol condensation of acetaldehyde on a Mg,Al–vanadate LDH
calcined at 450°C, although the activity is lower than in solids prepared, at the same
temperature, from a carbonate precursor, probably because of a decrease in surface
basicity, required for self-condensation of acetaldehyde.
Formation of cyclic and chain-like oxovanadates in the interlayers of a Mg,Al–
LDH on heating at moderate temperatures has been reported by Twu and Dutta
[139]. For the synthesis, these authors followed the two-step method [65], first
preparing a Mg,Al – terephthalate LDH, then exchanging with NaVO3 at pH 4.5.
The nature of the intercalated anion was concluded not only from swelling, as
measured by XRD, but also from Raman and XANES spectra. As the calcination
temperature is increased, the gallery height, calculated from the position of the
basal XRD lines, decreases; changes in the Raman spectra are consistent with the
following series of reactions:
1. at 160 – 280°C: V10O628− +3H2O “3V3O39 − + HVO24 − + 5H + although bands
due to VO34 − (formed through deprotonation of HVO24 − ) were also detected.
2. between 220 and 450°C: nV3O39 − “ 3n(VO3)nn −
3. between 450 and 650°C: MgO+(VO3)nn − “ a-Mg2V2O7 + Mg3(VO4)2
The XANES spectra [139] in the vanadium K-edge shows a pre-edge absorption
(which intensity increases with the calcination temperature), assigned to a dipole
forbidden 1s“3d transition, strengthed by the mixing of the 3d orbitals with 4p
metal and 2p oxide orbitals; this mixing is promoted by the lowering of symmetry
around the vanadium atom from strictly octahedral (as that existing in decavana-
date) to tetrahedral, existing in other vanadates.
The nature of the magnesium vanadate species formed upon calcination at high
(700°C) temperature depends markedly on the nature of the precursor layered
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 99

materials. So, calcination of a carbonate LDH containing Mg2 + , Al3 + and V3 + in

the layers [166,167] leads mostly to formation of Mg3(VO4)2 species, containing
isolated [VO4] tetrahedra, while starting from a Mg,Al–LDH containing interlayer
decavanadate [158,167] leads to formation of MgV2O6, containing pairs of edge-
sharing VO6 octahedra [168]. These results may be important for catalytic purposes,
as it has been claimed [169,170] that the catalytic properties of the V–Mg–O
system in the oxidative dehydrogenation of propane depend on the nature of the
species existing, the highest selectivity being related to the presence of isolated [VO4]
tetrahedra, the lack of V – O – V species (where oxygen can be released easily
through reduction of V5 + to V4 + ) decreasing the activity for combustion [171].
When the nature of the cations existing in the brucite-like layers is changed, the
trends found are similar, although the calcination temperatures required for decom-
position and crystallization of new phases may vary from one case to another: so,
alumina-supported decavanadate maintains its structure up to 450°C [172,173];
when intercalated in Mg,Al LDH its decomposition starts at 100°C [139,174], but
at 200°C in Ni,Al LDH prepared by ionic exchange, and at 150°C in the case of
vanadate-containing Ni,Al – LDH prepared by reconstruction [160], and in Mg,Cr
or Ni,Cr LDHs. Such a behavior can be related to hydrolysis of decavanadate
species by water molecules remaining in the interlayer space, leading to polymers
with lower charge density; the presence of cations with different polarizing power in
the brucite-like layers would also modify the interaction between water molecules
and the decavanadate anions. Surface acidity of the solids obtained upon calcina-
tion at 300 or 500°C of NiAl, NiCr, and MgCr LDHs containing decavanadate, has
been assessed from IR monitoring of pyridine adsorption [175]. It has been found
that samples NiAlV contain both Lewis and Brönsted acid sites, while Lewis sites
are lacking in the presence of Cr. Adsorption of 2-propanol on these oxides takes
place dissociatively, leading to oxidation to acetone at 300°C via a Mars and van
Krevelen mechanism, more extensively in the less acidic, Cr-containing catalysts,
although in this case further oxidation to acetate species takes place.
Twu and Dutta [135] have followed the evolution of V2O47 − , exchanged at pH 10
in a Li,Al LDH, upon heating. In this case, heating at 80°C is enough to yield
dimerization to V4O412− , but the process is reversible. Upon heating at 200°C
hydroxide-mediated polymerization/depolymerization reactions occur, leading to
oligomeric chains of [VO4] tetrahedra linked by oxygen atoms, which presence has
been also ascertained by diffuse reflectance spectroscopy. Finally, above 450°C the
Li aluminate framework is destroyed, forming Li3VO4 and LiVO3. Oxidation of
o-xylene on the material calcined at 450°C led to formation of o-tolualdehyde,
indicating some sort of selectivity in oxidation of one of the methyl groups.
Thermal treatment may lead, in other cases, to different processes. So, on heating
a NiIICoIII – LDH containing interlayer (VO3)nn − chains at 190°C (a temperature
low enough to avoid collapsing of the layered structure), the basal spacing decreases
from 9.15 to 7.21 Å, a value of the same order as that found with intercalated
chromate and dichromate [127,128,131], and slightly smaller than that found for
materials containing planar, interlayer carbonate; in other words, only a layer of
oxygens, from the vanadate species, should be located between the hydroxide
100 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

layers. This behavior has been ascribed to grafting of the intercalated anions to the
layer upon heating, Fig. 11 [176]. The grafting process has been followed also by
51
V MAS-NMR, even from the very first stages of exchange in a NiIICoIII –LDH
[177]; as soon as they are inserted in the interlayer to compensate the positive
charge in excess in the layers, the isolated diperoxovanadate ions (formed in the
NH4VO3/H2O2 medium) undergo a competition between polycondensation and
grafting: if the solid is maintained with the reducing solution for a long time, partial
grafting occurs, leading to dehydroxylation of the layers; however, if the solid is
removed early from the solution, polycondensation is favored, together with a low
extended grafting. Thermal treatment favors further grafting and defragmentation
of the polyoxovanadate, finally leading to collapsing of the layered structure.
Grafting (as concluded from an abnormal short layer–layer distance), even without
any thermal treatment, has been also observed for pyrovanadate species, V2O47 − ,
onto a Cu,Cr LDH [141]; in this case, however, two adjacent hydroxyl groups of
one OH layer are substituted by two oxygen atoms of the ditetrahedra, as proposed
also for a Cu,Cr – Cr2O7 LDH phase [128].
Isopolyvanadate has been also exchanged at pH 4.5 in Mg–Al LDHs in the
nitrate form. A detailed study by XRD and IR spectroscopy of the species formed
upon heating in air at increasing temperatures has been carried out by López-Sali-
nas and Ono [174]. The results were similar to those reported by Twu and Dutta
[139], despite the difference in the Mg/Al ratio, 2 or 3. The changes can be also
easily followed by IR spectroscopy: The decavanadate gives rise to a strong, sharp
absorption band at 957 cm − 1, together with weaker bands at 557, 598, 660, 740,
and 820 cm − 1, while the presence of metavanadates gives rise to bands at 840 and
920 cm − 1 (terminal VO stretching), and 550 and 680 cm − 1 (V–O stretching in
bridging V – O – V bonds) [178]. If decomposition is carried out under vacuum,
partial reduction of V5 + to V4 + species takes place, the highest concentration (6
V4 + /100 Vtotal) of reduced species being reached at 150°C [174]; such a reduction is
not reversed by oxygen treatment at 200°C during 1 h, probably because of the
hindered access of even small molecules into the interlayer space due to very close
adjacent isopolyanions.
The presence of decavanadate anions in the interlayer space of LDHs has also
important effects on the decomposition process of the material. First of all,
calcination of a Mg – Al LDH in the carbonate form usually leads to a weight loss
close to 35 – 50% of the initial weight, due to removal of water physically adsorbed
on the external surface of the crystallites (usually at low temperature), removal of
interlayer water molecules, and, finally, dehydration/dehydroxylation, due to con-
densation of hydroxyl groups from the brucite-like layers, and decarbonation, from
the interlayer carbonate anions [179,180]. If decavanadate, instead of carbonate, is
present in the interlayer, the second weight loss corresponds only to water removal
through condensation of layer hydroxyl groups, and the total weight loss is usually
lower than 20 – 25% of the initial sample weight. The nature of the solids formed
upon calcination also depends on the nature of the interlayer anion. This effect has
been particularly studied for Co,Cr [181] and Zn,Cr [182] LDHs by del Arco et al.
For Co,Cr – V10O628− and Zn,Cr – V10O628− LDHs prepared from a carbonate precur-
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 101

Fig. 16. Cr–K edges XANES spectra for Zn,Cr – CO3 (a) and Zn,Cr – V10O28 (b) LDHs calcined at 400
and 800°C. Inset: Cr K-XANES spectrum for crystalline K2CrO4. Reprinted from M. del Arco, V.
Rives, R. Trujillano, P. Malet, Thermal behaviour of Zn – Cr layered double hydroxides with hydrotal-
cite-like structures containing carbonate or decavanadate, J. Mater. Chem. 6 (1996) 1419 – 1428, © 1996,
with permission from The Royal Society of Chemistry.

sor by ion exchange at pH 4.5, the XANES at the vanadium K-edge features
(e.g. pre-edge peak intensity and position, main edge position, and post-edge
structure) are almost identical as for a crystalline decavanadate (n-
C6H13NH3)6(V10O28) · 2H2O, confirming the structure of the intercalated anion. It
has been also observed, from XAS studies [147], that the zinc shell is in accordance
with Lowenstein’s rule, which states that trivalent cations in aluminosilicates should
not be in adjacent metal sites [183]. On calcination, while the DTA profiles for these
carbonate LDHs showed the expected endothermic peaks due to water removal,
when recorded in air, an additional exothermic peak was recorded for the Zn,Cr–
LDH at 435°C (335°C for the Co,Cr–LDH), but was absent when recorded in
nitrogen, indicating an oxidation process, involving the Cr ions; according to Fuda
et al. [184], the presence of carbonate in the interlayer of hydrotalcites favors the
Cr3 + “ Cr6 + oxidation. The peak was absent in the DTA profiles of the vanadate-
containing LDHs. When the samples were calcined in air at increasing tempera-
tures, oxidation of Co2 + to Co3 + takes place at 400°C in the Co,Cr–CO3 LDH,
but not in the decavanadate analogue. In addition, a weak pre-edge peak character-
istic of chromate ions appears at the chromium K-edge XANES spectrum, Fig. 16.
Depolymerization of decavanadate species follows trends similar to those reported
by Twu and Dutta [139] for Mg,Al –decavanadate LDHs. Calcination at 650°C
leads to crystallization of CoIICoIIICrIIIO4 from the carbonate precursor, but of
102 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

CoIICr2O4 and CoII 2 V2O7 from the decavanadate precursor. In other words, the
presence of decavanadate and the absence of carbonate hinders oxidation of Co2 +
ions, thus modifying the nature of crystalline phases formed at high temperature.
For the Zn,Cr analogues, the same behavior is observed with respect to Cr
oxidation, and the crystalline phases formed were ZnO and ZnCr2O4 from the
carbonate precursor (through formation of chromate species at intermediate calci-
nation temperatures), and ZnV2O6 and Zn2V2O7 at 400 and 750°C, respectively,
from the vanadate precursor, together with ZnCr2O4. Formation of chromate and
Co3 + species in the case of the carbonate precursors, but not in those containing
interlayer decavanadate, has been also concluded from temperature-programmed
reduction studies [181,182], a technique that has been proved to be adequate to
determine redox processes in layered double hydroxides containing reducible
cations in the layers or in the interlayer anions [185].

3.2.3. Molybdates
While the chemistry of vanadates in the interlayer space of LDHs has been
throughout studied, that of molybdates has been restricted to the heptamolybdate,
Mo7O624− , and the papers published are rather scarce. The preswelling technique
with terephthalate was used by Drezdzon [65] to intercalate Mo7O624− in the
interlayer space of a Mg,Al carbonate precursor. The method was similar to that
above described for decavanadate, i.e. direct preparation of the Mg,Al–terephtha-
late precursor (from Mg2 + and Al3 + nitrates and terephthalic acid in NaOH
medium), and mixing of the slurry with a Na2MoO4 · 2H2O aqueous solution, and
further lowering of the pH to 4.4 – 4.7 with HNO3. The basal spacing was 12.2 Å,
corresponding to a heptamolybdate orientation in which the C2 axis is perpendicu-
lar to the brucite-like layers. Exchange reactions using other organic precursors did
not succeed (2,5-dihydroxy-1,4-benzendisulphonate; lauryl sulphate) or proceed
with difficulty (1,5-naphtalenedisulphonate), because of competition with complex
formation between the organic species and the metalate [186].
The so-called reconstruction method, from a carbonate precursor calcined at
450°C and an acidified (pH 4.5) solution of ammonium heptamolybdate, led also to
intercalation of heptamolybdate species [148,187] with the same spacing as that
obtained by Drezdzon. However, Misra and Perrotta have reported [126] the
preparation of a Mg,Al – Mo7O24 LDH with a basal spacing of 9.6 Å from a
carbonate precursor calcined at 500°C, but without acidification. This spacing is
markedly lower than that reported by Drezdzon for samples prepared by the
terphthalate intermediate [65], and by Chibwe and Jones [148] for samples prepared
by reconstruction. The difference has been attributed by these authors [126] to the
different Mg/Al ratio in their sample (1.88) and in Drezdzon’s sample (2.0).
However, the difference is very small and, moreover, the value by Misra and
Perrotta is almost coincident with that reported by Chibwe and Jones (1.82), which
led an interlayer space of 12.0 Å. So, the difference should be more probably due
to: (i) a different orientation of the interlayer anion, (ii) the presence of a different
oxomolybdate anion (due to partial depolymerization), or (iii) some sort of
grafting, as reported in the case of vanadate [141] and chromate [127,128,131], as
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 103

remaining of organic molecules in the interlayer space should be ruled out since
these were not used in the method followed by Chibwe and Jones [148]; unfortu-
nately, these authors [126] provide no other experimental data, in addition to XRD,
to support their conclusions. The same value, 12.0 Å, has been recently reported by
Hibino and Tsunashima [188] for samples prepared in an ethanol–water solution
by anionic exchange, to avoid partial dissolution of the brucite-like layers, due to
the acidic medium provided by the molybdate solution. A value of 12.2 Å has been
also reported [189] by Twu and Dutta in materials (Mg/Al = 2.0) prepared follow-
ing the Drezdzon’s method; although the layer structure is destroyed at 300°C, the
Mo7O624− moiety is stable up to 400°C, then forming MgAl2O4, MgMo2O7 and
MgMoO4, as concluded from XRD and Raman spectroscopy studies. However,
intercalation of MoO24 − in a Li,Al LDH failed, due to hydrolysis of the molybdate
anion, even at r.t.
Levin et al. [190,191] have reported synthesis of a layered ammonium zinc
molybdate using as a precursor a Zn,Al–LDH calcined at 500°C and Mo7O624−
(although this depolymerizes along the reaction) at r.t. These authors find a similar
reactivity with LDHs containing, in addition to Zn and Al, Cu2 + , Co2 + or Ni2 + ,
but not with LDHs containing Ni2 + or Mg2 + only as the divalent cation. From
27
Al MAS-NMR studies, it was concluded that a high fraction of tetrahedral Al3 +
ions are required in the calcined precursor to yield the layered molybdate.

3.3. Layered double hydroxides intercalated with high-nuclearity oxometalates: iso

and hetero-polyoxometalates

In this chapter we summarize the studies reported on the synthesis, characteriza-

tion, and properties of LDHs where the anions existing in the interlayer space is a
polyoxometalate (POM) possessing the Keggin-type, or related, larger structures,
such as Dawson and Finke [192] or Preyssler types [193]. One of the aims of these
studies was to expand the layers further than the values obtained with decavana-
date; it has been also observed improved catalytic properties of these systems. In a
Keggin-type structure, twelve metaloxygen octahedra form a shell surrounding a
tetrahedrally coordinated heteroatom (P, Si, B, etc.); the shape is close to spherical,
with a diameter of ca. 10 Å
Kwon and Pinnavaia [194] reported in 1989 the preparation of LDHs containing
intercalated [XM12O40]n − anions. These pillared intercalates are intrinsically
difficult to synthetize in highly crystalline form, in part, because LDH hosts are
basic, whereas the anions are highly acidic. Actually, these authors succeed in
preparing intercalates with a Zn2Al– NO3 precursor by ion exchange, but not with
the more basic Mg3Al– NO3 one. The gallery height, as determined by XRD, was
close to 10 Å, and the intercalation by ion exchange seems to depend on the net
charge and on the polyhedral form of the anion. So, intercalation with a hot
suspension of Zn,Al – NO3 LDH (Zn:Al=2:1) with a-[H2W12O40]6 − and a-
[SiV3W9O40]7 − led to complete exchange, while exchange was lower than 20% for
Keggin-type structures, such as [PCuW11O39(H2O)]5 − , and no exchange at all was
observed with a-[PW12O40]3 − or a-[SiW12O40]4 − . These findings are in agreement
104 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

with charge balance reasons: as the diameter of the Keggin unit is close to 9.8 Å,
an area of 83 Å2 is required to accomodate a Keggin unit; as the charge density in
the LDH used was 16.6 Å2 (although this value has been corrected to 25 Å2 by
Clearfield et al. [122]), those Keggin units with a formal charge lower than − 5 will
be unable to enter in the interlayer space to compensate the positive charge of the
layers.
In any case, the products obtained, Zn,Al–a-[H2W12O40]6 − and Zn,Al–a-
[SiV3W9O40]7 − , are crystallographically well-ordered phases, with a basal spacing
of 14.5 Å, corresponding to a gallery height of 9.8 Å, in agreement with crystallo-
graphic data for Keggin units, and up to six diffraction harmonics were recorded in
the XRD diagram. The extremely large swelling of the layers is accompanied by a
substantial increase in the specific surface area, from 26 m2 g − 1 for the nitrate
precursor, to 63 and 155 m2 g − 1, respectively, for the solids intercalated with
a-[H2W12O40]6 − and a-[SiV3W9O40]7 − , and then the term ‘pillared’ seems to be
more adequately used than in other cases where, despite intercalation of large
anions, such an increase is not observed. Further evidence for the retention of the
Keggin structure inside the layers was attained by IR and 29Si and 51V MAS-NMR
spectroscopies. With respect to the orientation of the anion in the interlayer, these
authors [194] conclude that the C2 axis of the oxygen framework should be
perpendicular to the brucite-like layers, Fig. 17, as in this orientation the number of
hydrogen bonds to layer OH groups is maximized. This orientation has been also
proposed by Liu et al. [195] for [PW11O39Cr(H2O)]4 − , [PW11TiO40]5 − and
[PW11VO40]4 − intercalated in Zn,Al – LDHs prepared by ion exchange.
In order to intercalate Keggin-type anions in rather basic LDHs, Dimotakis and
Pinnavaia have proposed [196,197] an alternative method, consisting in preparation
of the Mg3Al – OH LDH (meixnerite) by reconstruction of a calcined carbonate
precursor; meixnerite is then exchanged with p-toluensulfonate or adipate in the
presence of glycerol as a swelling agent, yielding a well-crystallized phase with
extremely well ordered organic anions, which gallery height is very close (14.4 Å for
the adipate) to that of the Keggin derivative, from which anion exchange led to
microporous, single crystalline phases with d= 14.8 Å, after exchanging with
[H2W12O40]6 − or [SiW11O39]8 − .

Fig. 17. A Keggin unit, [XM12O40]n − , formed by 12-corner-sharing octahedra, in two different
orientations.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 105

The reconstruction method originally used by Chibwe and Jones to prepare

LDHs by reconstruction of calcined precursors [148] has been tested by Narita et al.
[152] to prepare intercalates with Keggin anions, starting from a Zn,Al–CO3
precursor. Addition of the amorphous solid obtained after calcination at 500°C to
a boiling solution of [SiW11O39]8 − at pH 6 led to a LDH with a basal spacing of
14.6 Å, almost coincident with the value reported for the solid obtained by ionic
exchange [194]. However, in addition, a broad reflection is also recorded at ca. 10.2
Å, which has been ascribed by these authors [152] to a quasicrystalline phase
consisting of Zn2 + and Al3 + salts, and similar results were obtained for
[SiV3W9O40]7 − . Nevertheless, the method [148] seems to be general both for acidic
and basic LDHs, although with acidic LDHs better ordered systems are obtained.
These authors have also prepared Zn,Al–LDHs intercalated with the Keggin ion
a-[SiW11O39]8 − by direct synthesis [153], i.e. by adding a solution of Zn2 + and
Al3 + to a vigorously stirred solution (100°C, pH 6) of the Keggin species. Although
the XRD diagram shows several sharp harmonics corresponding to (00l) diffrac-
tions (the first one at 14.6 Å, a value similar to that obtained for solids prepared by
ion exchange or reconstruction [152,194,197]), the broad feature at ca. 10.8 Å is
also observed for the solid prepared by reconstruction [152] by these same authors.
The use of swelling organic molecules (glycerol or triethylenglycol) to prepare the
Mg,Al–H2W12O640− LDH has been also followed by Pinnavaia et al. [198]. The
advantage of the method is the direct use of the carbonate form of the LDH
precursor. Hansen and Taylor [199] had reported that the reaction of Mg,Al–CO3
LDHs with glycerol at 160°C resulted in swelling and well dispersed LDH suspen-
sions. Well crystallized LDHs have been obtained by Pinnavaia et al. following this
method [198], with up to seven basal harmonics in their XRD diagram, and with a
basal spacing of 14.9 Å, characteristic of LDHs intercalated with Keggin ions, Fig.
18. Although a byproduct, characterized by a broad diffraction at ca. 11 Å, is also
formed, its content is lower when using triethyleneglycol than when using glycerol
or in samples prepared by reconstruction [153]. Also, the particles show an irregular
glass-like morphology, very different from the hexagonal morphology of the
starting carbonate LDH, thus indicating the reaction is not topotactic (as observed
in samples prepared by ionic exchange of a LDH–OH precursor), but probably
proceeds via dissolution and recrystallization of the LDH–CO3 at high tempera-
ture. Differences are observed also in the specific surface area and porosity of the
samples: those prepared following the polyol route display a value of ca. 160 m2
g − 1 (83 m2 g − 1 when prepared by ion exchange of meixnerite), 65% of which
corresponding to micropores. However, Weir et al. [200] have reported a total
blocking of the micropore system of Mg,Ga and Mg,Al–LDH containing
[PW11O39]7 − and [H2W12O40]6 − , prepared via terephthalate intermediates, probably
because of the large charge density of the LDH–terephthalate precursors (Mg2 +
:M3 + = 2:1).
Clearfield et al. have carried out a detailed study on the experimental conditions
to intercalate Keggin anions in LDHs [122,154]. These authors have prepared
different LDHs precursors, Mgm Al– X (m= 2–5; X= Cl − , NO3− ) pillared with
[PVn W12 − n O40](3 + n) − (n = 0 – 4), and have shown that exchange can be attained in
106 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Fig. 18. XRD patterns (Cu–Ka ) of oriented film samples of Mg,Al – LDH metatungstate reaction
products obtained from (A) meixnerite, (B) LDH glycerolate, and (C) LDH triethyleneglycolate
precursors. All diffraction patterns were recorded at r.t. For the LDH – POM intercalated derived from
the triethyleneglycolate precursors, the XRDs were taken at r.t. after 1 h preheating under nitrogen at
(D) 100, (E) 200, and (F) 250°C. Reprinted from S.K. Yun, V.R.L. Constantino, T.J. Pinnavaia, New
polyol route to keggin ion-pillared layered double hydroxides, Microporous Mater. 4 (1995) 21 – 29, ©
1995, with permission from Elsevier Science.

aqueous solution without any organic swelling agent. Direct anion exchange in
aqueous solution was also used by Guo et al. [201] to prepare Zn2Al–LDH
intercalated with [PVn W12 − n O40](3 + n) − (n=1–4). Exchange is easier if the LDH
precursor is thoroughly wet [122,154], either by preparing wet solids or by soaking
the dried product for 3 – 4 h (a similar role of the wetness state of the precursor has
been also shown by Carrado et al. [95] and Kooli and Jones [149] to incorporate
phtalocyanines or decavanadate in the interlayer region, respectively). Also the
exchange is easier with soft powders than with glassy particles. Moreover, if the
solid is slurried enough, pillaring seems to be almost independent on the charge of
the Keggin unit: for a wet Ni2Al– NO3 precursor soaked for at least 3 h, total
exchange was obtained with all Keggin ions with net charge ranging from − 3 to
− 7 (a 90% exchange was already attained after 5–10 min). This apparent indepen-
dence on the net charge of the Keggin unit is rather surprising, as the area required
to accomodate a Keggin unit is 83 Å2, and the area per layer charge unit is 25 Å2
(16.6 Å2 according to Kwon and Pinnavaia [194]); then the charge per Keggin unit
should be, at least, −4 (or − 5, according to Kwon and Pinnavaia). The experi-
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 107

mental finding that even Keggin units with a net charge of − 3 are also exchanged,
can be only explained assuming an alteration of the layers during the exchange
process, leading to an increase in the layer charge density.
On the other hand, the interlayer spacing required to host the Keggin unit is close
to 14.2 Å, but the interlayer spacing measured for the pillared Ni,Al–LDHs was
only ca. 12 Å. For the Mg,Al – LDHs the spacing was 14.7 Å (the corresponding
harmonics being also recorded), and an additional broad reflection, similar to that
reported by Narita et al. [152], was also recorded at 11–13 Å. These authors
suggest a partial dissolution of the divalent anion in the acidic medium (pH 4–5)
during exchange, together with partial removal of hydroxyl groups from the layers,
thus creating vacancies where the Keggin ions could fit, leading to interlayer
spacings lower than expected, and thus explaining intercalation of the [PW12O40]3 −
Keggin unit. Computer graphics models support such relationship between ‘pene-
tration’ of the Keggin units into the brucite-like layers, and the spacing.
To overcome the difficulty in intercalation of Keggin units with low negative
charge, Serwicka et al. [202] have subjected the [PMo12O40]3 − anion to electrochem-
ical reduction prior to exposure to the LDH; it is known [192] that these anions
undergo facile reduction to give so-called heteropoly-blues, and that reduction
renders the anions less acidic, so helping also to overcome decomposition by the
reaction with the rather basic Mg,Al –NO3 LDH. In this case, reduction led to a
transfer of four electrons to the Keggin unit, and the authors find that the
unreduced heteropolyanion actually reacts with the LDH, XRD show diffraction
maxima at 11.1 Å, too small to correspond to the intercalated Keggin unit,
probably corresponding to decomposed species; the atomic Mo:P ratio was 4.5.
However, for the reduced heteropolyanion, the spacing was 14.8 Å, with corre-
sponding harmonics, although an intense reflection was also recorded at 10.8 Å,
ascribed to formation of a non-layered byproduct.
Direct exchange reaction in aqueous solution has been used by Hua et
al. [203,204] to intercalate peroxoheteropolyanions with the Keggin structure,
such as [SiW11(TiO2)O39]6 − , [SiW9(TiO2)3O37]10 − , [PW11(TiO2)O39]5 − and
[PW9(TiO2)3O37]9 − , in Zn2Al – LDHs in the nitrate form, obtaining solids with
basal spacings of 14.7 Å.
Weber et al. [205] have studied by TEM the partial exchange of [SiV3W9O40]7 −
in a Mg2Al – LDH. These authors find that the average crystallite size of the
exchanged product was larger than that of the original LDH, suggesting exchange
proceeds via dissolution and topotactic reprecipitation of the exchanged LDH, as
also proposed by Pinnavaia et al. when following formation of triethylenglycol
intermediates [198]. On the other hand, the results obtained on the local chemical
composition (Mg/Al and W/Al ratios) indicate that in partially intercalated solids,
the resulting structures consist of stacks of completely substituted layers superposed
on unchanged layers.
The synthesis of many other LDHs containing intercalated Keggin-type anions
has been described in the literature. Hu et al. [206,207] have reported the intercala-
tion in Zn2Al – LDH of different POMs, such as [PVW11O40]4 − ,
[XW11O39Z(H2O)]n − (X = Si, B; Z =Co, Ni, Cu, Al), [Ln(XW11O39)2]n − (Ln= La,
108 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

CeIII; X= Si, P, B) by ion exchange. As expected, basal spacings close to 14.5 Å

were observed and, from 31P and 27Al MAS-NMR, orientation of the Keggin unit
with its C2 axis perpendicular to the layers was concluded, validating the prediction
made by Pinnavaia et al. based on the formation of hydrogen bonds between the
layers and the oxide ions of the Keggin unit [194]. [SiW11O39Co(H2O)]6 − was
intercalated in the interlayer space of Zn2Al–NO3 LDH by the action of ultrasound
[208], a method previously used to intercalate decavanadate [151].
López-Salinas et al. have reported [209] the intercalation of [SiW11O39X(H2O)]6 −
(X= Co2 + , Mn2 + ) in a Mg2Al– NO3 LDH by ion exchange, obtaining a layered
material with a basal spacing of 15.2 Å, although a byproduct, characterized by the
broad reflection close to 11 Å, is also formed. The structure is stable up to 350°C,
and the water ligand can be reversibly removed below 200°C. The coordinatively
unsaturated cation (Co or Mn) can reversibly adsorb small molecules, such as
water, methanol, ethanol or ammonia, as concluded from UV–vis/DR studies,
while no spectral changes were observed upon adsorption of pyridine (vapor or
liquid phase) or THF, probably because of steric hinderance to access the adsorp-
tion site.
Clearfield et al. [38] have described the synthesis of a Cu,Al–LDH intercalated
with the [PV3W9O40]6 − anion by a controlled increase of the pH of a solution
containing the POM and the metal cations (Cu2 + and Al3 + ), via hydrolysis of urea
on mild heating (ca. 50°C). The method permits a completely homogeneous pH
environment throughout the solution, avoiding local changes in concentration, pH,
etc.
Intercalation of heteropolyanions larger than Keggin-type ones has been reported
by Yun and Pinnavaia [210]. These authors report intercalation of anions with
Dawson, a-[P2W18O62]6 − , and Finke, [Co4(H2O)2(PW9O34)2]10 − , structures in a
Mg3Al–LDH, and compare the results with intercalation of the Keggin ion
a-[H2W12O40]6 − . In addition to obtaining larger gallery heights, differences are also
found due to different orientations that these anions can achieve in the interlayer
space, as, contrary to the spherical shape shown by the Keggin ions, the Dawson
and Finke ions have a structure close to cylindrical (D3h and C2h symmetry,
respectively), so the size of micropores could be hopefully modulated. The Dawson
and Finke derivatives were prepared using a LDH–adipate as the precursor at
100°C, following the method by Dimotakis and Pinnavaia [196], and also using a
meixnerite precursor and ambient pH conditions at r.t. [210]. With respect to the
Dawson derivatives, the first method yielded a layered material with a basal spacing
of 17.6 Å, suggesting the anion is oriented with its C2 axis perpendicular to the
layers; only small decreases in the basal spacings are observed when the solid is
heated at 25 – 200°C, probably due to dehydration. However, via the meixnerite
method, the basal spacing was 19.3 Å, in agreement with the C3 axis of the anion
perpendicular to the layers; this value decreases irreversibly to 17.4 Å upon heating
at 100°C, suggesting a re-orientation of the Dawson unit from end-on position to
an horizontal position, Fig. 19; the former orientation seems so to be unstable, and,
actually, while only six oxygen atoms from the Dawson unit are oriented to form
hydrogen bonds with the layers in the end-on position, in the horizontal position 14
oxygen ions are able.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 109

Contrary to the results with the Dawson’s anion, intercalation of the Finke anion
led to layered materials with a basal spacing of 17.79 0.3 Å, whichever (adipate or
meixnerite) the precursor used, suggesting that both intercalated products have the
C2 axis of the heteropolyanion perpendicular to the LDH layers. The microporous
structure is maintaned up to 200°C when heated in N2. Upon heating, the gallery
height decreases by 2 – 3 Å, by removal of water molecules, resulting in stronger
electrostatic and hydrogen bonding interactions between the oxygen ions of the
heteropolyanion and the LDH hydroxyl groups, leading to reorientation of the
former. Retention of the Dawson and Finke structures in the intercalated state was
verified by FTIR spectra, which showed the expected bands due to P–O linkages
close to 1100 and 1030 cm − l, and W –O linkages close to 960, 930, and 750 cm − l,
the precise positions depending on the particular POM [211,212].
As in most of the cases previously described [153,154,205], the XRD diagrams of
these materials showed, in addition to the lines due to the layered material, a broad
peak close to 11 Å. Although different explanations have been previously proposed
for the nature and origin of the material responsible for this reflection, these
authors conclude that salt formation from cations depleted from the layers and
non-gallery POM remains the favored explanation, and the XRD features of this
impurity were the same as of the solid obtained by grinding a carbonate–LDH or
even Mg(OH)2 and a POM in the solid state. This report [210] represents the most
widely study on this byproduct, detected upon reaction of POMs with LDHs.
Larger POMs have been introduced in the gallery space of Mg,Al and Zn,Al
LDHs by Evans et al. [147] by ion exchange and direct synthesis; solids have been
prepared with gallery heights ranging from 7.1 to 16 Å including species such as
[Nbx W6 − x O19](x + 2) − (x = 2 – 4), [V2W4O19]4 − , [Ti2W10PO40]7 − , and
l4 −
[NaP5W30O110] . The largest yields to highly crystalline solids were obtained
following the exchange method, with precursors containing chloride or nitrate.

Fig. 19. Reaction scheme for intercalation of Dawson polyoxometalate in a LDH.

110 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

These POMs, in addition, are stable in a wide pH range, and so can be incorpo-
rated into the interlayer space of strongly basic Mg,Al–LDH. Probably, one of the
most interesting products prepared by these authors [147] is that containing the
Preyssler anion, [NaP5W30O110]14 − , one of the largest known anions, and a rather
abnormal chemical assembly with a C5 axis. Its structure consists of a cyclic
arrangement of five [PW6O22] units, each formally derived from the Keggin-type ion
[PW12O40]3 − by removal of two sets of corner-shared [WO6] groups. The basal
spacing for the Zn,Al – Preyssler derivative obtained from the nitrate precursor was
21 Å, in agreement with the ion orienting its shortest dimension parallel to the host
layers. When ion exchange was performed from benzenecarboxylate-containing
precursors (YC6H4COO − , Y =COO − , OH, CH3), low (ca. 3.5–4.5) pH values
were required to protonate the carboxylate group, to induce removal of the free
organic acid. Direct synthesis led to incorporation of the Preyssler ion in a
Zn,Al–LDH, with a spacing of 21 Å. Retaining of the POM structure was checked
by EXAFS [147]. As expected, a by-product responsible for a reflection at 7–11°
(2u, Cu– Ka ) and which nature has been discussed above is formed in most of the
cases.
Studies on the thermal decomposition of Zn2Al–LDHs intercalated with differ-
ent Keggin ions have shown [213] that after dehydration (200°C), dehydroxylation
is completed at 350°C, leading to amorphous solids; the layer structure is destroyed
between 200 and 250°C, depending on the precise nature of the interlayer Keggin
ion. The layered material can be rehydrated by immersion of the calcined solid in
water, but the pillared, layered structure is not reconstructed by simply rehydration.
Crystallization of anhydrous mixed oxides (e.g. ZnWO4) is observed after dehy-
droxylation, leading to an exothermic peak in the DSC trace. Similar results have
been reported by Guo et al. [214].
The structure of salts of Keggin ions with bulk cations (e.g. Cs3[PWl2O40]) is
stable even up to 700°C. The results by Kwon and Pinnavaia [213] suggest that the
mixed oxide formed upon decomposition of the brucite-like layers reacts easily with
the hosted Keggin ion. This low thermal stability in some sort of way limits the
application of these materials to catalytic processes taking place at rather low or
even r.t., such as photocatalytic oxidation of isopropanol to acetone [213].
Although LDHs are mostly basic solids, incorporation of POMs in the interlayer
space not only increases the gallergy height and the thermal stability, but also may
provide electron acceptor sites and acid sites. Thereof, these LDH–POM com-
pounds are interesting because of their acid–base properties. In these LDH–POM
systems, basic sites are located on the layers, while acid sites are on the interlayer
anions [215]. As a consequence, the relative acid–base strength could be changed by
different anionic exchange ratios, or even thermal treatments to yield materials with
tailored acid – base properties. Putyera et al. [216] have prepared a series of
Mg,Al–LDHs intercalated with molybdate and tungstate, in order to assess the
relationship existing between the composition and the acid–base properties of the
pillared materials and of their derivatives obtained upon calcination. As expected,
bearing in mind the close relationship existing between the nuclearity of the POM
and the pH, (WO24 − is stable at pH \ 8, W12O12 42
−
at pH 7.8, and W12O639− at pH
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 111

5.7, while for the oxomolybdate species, MoO24 − is stable at pH\7, but Mo7O624−
at lower pH values), a change in pH during synthesis gives rise to different acid
sites/basic sites ratio in the intercalated derivatives. Upon calcination, formation of
high nuclearity oxometalates (Mo7O624− or W12O10 −
41 ) again produces changes in the
acid–base properties.
Polyoxometalates are well known as oxidation catalysts, and so it is expected that
POM-catalyzed oxidation processes on POM hosted in the interlayer region of an
LDH can be constrained in a shape-selective environment. Tatsumi et al. [217,218]
have reported the epoxidation of alkenes (e.g. 2-hexene, cyclohexene and b-methyl-
styrene) with H2O2, catalyzed by LDHs intercalated with POM derivatives of Mo
and W. When the catalytic activity results are compared with those for the
unhosted POM, an steric effect is observed, and so epoxidation of large alkenes is
less favored on the LDH – POM catalyst, than when smaller alkenes are used. On
the other hand, the steric hinderance for the alkenes to access the interlayer region
seems to be less important for LDH – W12O41 than for LDH–Mo7O24, in agreement
with a larger basal spacing in the former than in the latter (12.2 vs. 9.9 Å,
respectively). These results suggest a possible shape selectivity control on changing
the size of the hosted species. In addition, hydrolysis of epoxides to yield di-ol
species is slowed down if compared with that observed for the unhosted catalysts,
probably because of the basic properties of the brucite-like layers. However,
Gardner and Pinnavaia have recently pointed out [219] that the co-product
generally formed (characterized by a broad diffraction maximum close to 10–11 Å)
when preparing LDH – POM intercalates can have important catalytic conse-
quences, and be even catalytically more important than the LDH phase.
The catalytic oxidation of benzaldehyde to benzoic acid using H2O2 in a biphasic
liquid–solid system, has been studied by Hu et al. [220] on Zn,Al–LDHs interca-
lated with [SiW11O39]8 − and [SiW11O39Z(H2O)]6 − (Z=Co2 + , Ni2 + , Cu2 + ). The
largest catalytic activity has been observed for the cobalt-derivative, and this
finding has been tentatively related to the easy change in the oxidation state of Co,
suggesting that Co2 + becomes oxidized to Co3 + by H2O2, and Co3 + oxidizes
benzaldehyde to benzoic acid.
Guo et al. [221] have studied the O2-oxidation of cyclohexene on LDHs interca-
lated with [XW11O39Z(H2O)]n − (X = P, Si; Z= Mn2 + , Fe3 + , Co2 + , Ni2 + , Cu2 + ).
The LDH – POM system is more active than the LDH in its nitrate form, and also
more active than the alkaline salts of the POM. An effect of the transition metal
cation existing in the POM moiety (Z) has been observed. It was concluded that the
cations in the layers or in the POM group behave as active sites for oxygen transfer
in cyclohexene oxidation, and so the catalytic properties could be hopefully tuned
by precise changes in the nature and concentration of transition metal cations in
both types of sites (layers and POM units).
Zheng et al. [222] have reported alkylation of iso-butane with butene on
Zn,Al–LDHs intercalated with [SiW12O42]8 − , and of Ni,Al–LDH intercalated with
[PW12O42]7 − , with good results with respect to activity and selectivity. For this
reaction, alkylation may proceed on basic (LDH) or acid (POM) sites, i.e. the
POM-pillared LDH can behave as a bifunctional acid–base catalyst. On calcination
112 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

at 300°C, Ni2Al– PW12O42 shows a much higher butene conversion than the
uncalcined material, as well as a higher content of C12 and C16 products, probably
due to a lower steric hinderance upon removal of interlayer water molecules.
Clearfield et al. [38] have used LDH–POM materials for catalytic conversion of
isopropanol to acetone or propene. This is a test reaction widely used for catalyst
characterization, as it proceeds to propene on acid sites, while to acetone on basic
sites. According to these authors, LDHs such as Mg,Al–CO3 behave as basic
catalysts, leading to acetone [215]. However, insertion of heteropolyacid anions
alters the selectivity drastically towards propene, showing that the acid character of
the POM predominates. Minor changes in acid/base catalytic properties have been
also correlated to the precise nature of the cations (Co2 + , Fe3 + , Ni2 + , Mg2 + ,
Al3 + ,...) in the brucite-like layers. Studies by Kagunya and Jones [165] on the aldol
condensation of acetaldehyde on Mg,Al[SiW12O40] have correlated the catalytic
activity with the surface area available, while selectivity seems to be related to the
number and strength of the basic sites, responsible for the activity in this reaction.
Ethanolysis of propene oxide to yield glycol ether (a reaction that can proceed
catalytically both on acid and basic sites) has been studied by Jones et al. [223] as
a way to assess the nature of active sites in Mg,Al–LDHs intercalated with POMs
such as [PW12O40]3 − and [SiW12O40]4 − . While for the LDHs lacking interlayer
POMs these authors report the hydroxyl groups as being the active sites, in the
LDH–POM systems, up to three sites are present: (i) oxide anions directly linked
to metal atoms (strong basicity), (ii) oxide ions bonded to atoms adjacent to metal
centers (medium basicity), and (iii) surface hydroxyl groups (weak basicity).
The acid – base functionality of the Mg,Al–LDH–[H2W12O640− ] association,
which synthesis has been described above, has been examined by Pinnavaia et al.
[198] using 2-methyl-3-butyn-2-ol (MBOH) as a reactive probe. When the catalyst
was obtained via a triethylenglycolate intermediate, a high reactivity for the
base-catalyzed disproportionation was observed, whereas when obtained via the
glycolate or meixnerite it was rather inactive. The difference has been attributed to
the different porosity in the samples obtained when using alternative intermediates.
Keita et al. [224,225] have prepared oxometalate-clay-modified electrodes con-
taining metatungstate. Glassy carbon electrodes have been modified with LDHs
containing Zn and Al in the brucite-like layers, and the interlayer anions have been
substituted for [H2W12O40]6 − under mild acid conditions. Clay films were prepared
by dropping a colloidal solution of the LDH onto the glassy carbon surface;
incorporation of the POM was accomplished readily by cycling the electrode in the
solution containing the oxometalate in the potential domain of the first redox
system of [H2W12O40]6 − to monitor the progress of the incorporation. Migration of
the highly-charged POM in the interlayer region of the clay results in the final
LDH–POM system. The best results were obtained with a Zn,Al–LDH precursor
containing the relatively large (if compared to chloride or nitrate) terephthalate
dianion at pH 5, acidic enough to ensure protonation of the organic anion (thus
favoring its removal from the interlayer region), but basic enough to avoid
dissolution of the layers.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 113

4. Miscellaneous

Adsorption of several oxometalates on mixed oxides prepared by calcination of

hydrotalcites (originally containing carbonate in the interlayer space) has been
studied as a way to remove these anions from waste water. Sato et al. [226] have
reported the adsorption of CrO24 − , HVO24 − , and MnO4− , among other oxoanions,
on a Mg,Al – LDH previously calcined at 500°C. The degree of adsorption of
divalent anions was greater than that of monovalent anions, although longer
equilibrium times were required to reach the equilibrium. Similar results have been
more recently reported by Rhee et al. [227] for chromate sorption on a Mg,Al–
LDH calcined at 560°C. Chatêlet et al. [228] have found that the amount of
chromate adsorbed on a Mg,Al – LDH calcined at 450°C is larger than its anionic
exchange capacity (AEC), with a change in the shape of the adsorption isotherm
exactly at a value corresponding to such AEC. These authors conclude that
adsorption above the AEC takes place on external sites with partially variable
charge and, from specific surface area measurements, conclude that the surface
density of adsorbed chromate is ca. 1 nm2, a value fairly close to the average value
of site density on oxides [229].
The ability of Mg,Al – LDH to adsorb TcO24 − and ReO24 − has been studied to
test their applicability for removal of Tc7 + from radioactive wastes [230]. The
sorption has been found to lead to grafted intercalates, substituting one or two
hydroxyl groups from the layers, depending on the TcO24 − /LDH ratio.
Similar studies have been also reported by Yamagishi et al. [231] on the
adsorption of chromate and permanganate on a thermally decomposed Mg,Al–
LDH. The intercalates, with a basal spacing of 8.7 Å, are then calcined at
800–900°C to avoid leaching of the heavy cations to aqueous solutions.
The strong basicity of hydrotalcites provides unexpected properties to modulate
the catalytic properties of supported catalysts. Pinnavaia et al. [232] have supported
Ru3(CO)12 on a Mg,Al – LDH by suspending the hydrotalcite in a degassed CH2Cl2
solution of the carbonyl complex. This undergoes reductive decarbonylation to the
−
HRu3(CO)11 , as confirmed by IR spectroscopy; equivalent reductive decarbonyla-
tion in homogeneous solution requires the presence of very strong bases [233],
provided in the LDH by hydrolysis of the carbonate anion. After exposure to air,
oxidation leads to grafted [Ru(CO)x (OM)2]n, analogous to grafted cationic com-
plexes formed on aluminum-pillared montmorillonite [234]. Reduction at 275°C
yields Ru crystallizes active for Fischer-Tropsch reaction, forming a high fraction of
oxygenates (mainly methanol and lesser amounts of C2 –C4 alcohols).These authors
[232] propose decoration of the metal crystallizes by the basic support to explain
oxygenates formation.
As with complexes of macrocyclic ligands, photophysical properties of other
anionic photocatalysts may be also modified when intercalated in LDHs. Pinnavaia
et al. [67] have reported immobilization of the luminescent anion Ru(BPS)43 −
(BPS= 4,7-diphenyl-1,10-phenantrolinedisulphonate) in a Mg,Al–LDH. Incom-
plete chloride exchange led to a layered material with a basal spacing of 22 Å,
indicating the complex is intercalated with its C3 axes normal to the layers.
114 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

Although marginal perturbations in the visible absorption profile are similar to

those observed for Ru(bipy)23 + intercalated in smectite clays, the IR and Raman
spectra of intercalated Ru(BPS)43 − and in solution are almost coincident, thus
discarding such low-symmetry-ligand distortions. The luminescence decay is com-
plex, a behavior that in the case of smectites has been ascribed to quenching of the
excited state by impurity ions (e.g. Fe3 + , Cr3 + ) isomorphically substituting Al3 +
or Si4 + in the clay structure [235], but this is precluded in the LDH. On the
contrary, the multiphasic emission decay in hydrotalcites arises from self-quench-
ing, as these authors have concluded from studies with systems simultaneously
containing Ru(BPS)43 − and nonemissive Zn(BPS)43 − .
CdS and CdS – ZnS have been incorporated into the interlayer of a Mg,Al–LDH
[236] and their photocatalytic properties for hydrogen evolution from Na2S,
Na2SO3 and/or 2-aminoethanol [237] under visible light (l\ 400 nm) irradiation
has been studied. The particle size of CdS should be very small (ca. 1 nm), as the
band gap of the Mg,Al – LDH/CdS was 2.64 eV, slightly larger than that of
unsupported CdS (2.40 eV), while for alloyed CdS–ZnS it was even smaller. The
photoactivity in the named reaction was larger for intercalated Cdl − x Znx S than for
the corresponding, unsupported sol, although the use of semiconducting supports
(e.g. H2Ti4O9 or H4Nb6Ol7) further improved the photoactivity.

5. Conclusions

From the results in this review, it is obvious the interest that layered double
hydroxides have deserved in recent years. Its structure, similar to that of layered
silicates, but with a change in the sign of the electric charges of the layers and the
interlayer ions, makes them true companions in systematizing the study of these
solids. On the other hand, as the layer cations and the interlayer anions can be
almost chosen from any one in the Periodic Table, the opportunities for synthesis
chemistry are enormous. This obviously constitutes a challenge for chemists. In
addition, the promising role that these materials, as obtained or after adequate
thermal treatments, can play as catalysts, sensors, electrodes, etc., makes them
worthwhile to be studied in a systematic way to modulate and to improve their
properties. Probably, in the next few years we will witness a lot of new work on
these compounds, expanding the nature of intercalated metal-containing anions.

Acknowledgements

The authors would like to thank the collaboration of their co-workers in the
Universities of Córdoba and Salamanca (Spain), as well as of Dr W. Jones
(University of Cambridge, UK), Dr P. Malet (Universidad de Sevilla, Spain), and
Dr F. Kooli (currently at NIRIM, Tsukuba, Japan). Finantial support by Junta de
Castilla y León (Consejerı́a de Educación y Cultura, grant SA45/96), Junta de
Andalucı́a (grant FQM-214) and Ministerio de Educacion y Cultura (grant PB96-
1307-C03) is acknowledged.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 115

References

[1] A. de Roy, C. Forano, K. El Malki, J.-P. Besse, in: M.L. Occelli, H.E. Robson (Eds.), Synthesis
of Microporous Materials, vol. 2, Expanded Clays and Other Microporous Systems, Van Nos-
trand Reinhold, New York, 1992, pp. 108 – 169.
[2] F. Trifirò, A. Vaccari, in: J.L. Atwood, J.E.D. Davies, D.D. MacNicol, F. Vögtle, J.-M. Lehn, G.
Alberti, T. Bein (Eds.), Comprehensive Supramolecular Chemistry, vol. 7, Solid-State Supramolec-
ular Chemistry: Two- and Three-Dimensional Inorganic Networks, Pergamon, Oxford, 1996, pp.
251 – 291.
[3] F. Cavani, F. Trifirò, A. Vaccari, Catal. Today 11 (1991) 173.
[4] M. Zikmund, K. Hrnciarová, Chem. Listy 91 (1997) 169.
[5] W.T. Reichle, Solid State Ionics 22 (1986) 135.
[6] M. Chibwe, J.B. Valim, W. Jones, in: C.A.C. Sequeira, M.J. Hudson (Eds.), Multifunctional
Mesoporous Solids, Kluwer, Amsterdam, 1993, pp. 191 – 206.
[7] K.A. Carrado, A. Kostapapas, S.L. Suib, Solid State Ionics 26 (1988) 77.
[8] M. Meyn, K. Beneke, G. Lagaly, Inorg. Chem. 29 (1990) 5201.
[9] W. Jones, M. Chibwe, in: I.V. Mitchell (Ed.), Pillared Layered Structures: Current Trends and
Applications, Elsevier, London, 1990, pp. 67 – 77.
[10] R.M. Taylor, Clay Miner. 19 (1984) 591.
[11] S. Velu, V. Ramaswamy, A. Ramani, B.M. Chanda, S. Sivasanker, J. Chem. Soc. Chem.
Commun. (1997) 2107.
[12] C.J. Serna, J.L. Rendón, J.E. Iglesias, Clays Clay Miner. 10 (1982) 180.
[13] M.A. Drezdzon, ACS Symp. Ser. 437 (1990) 140.
[14] A. Clearfield, M. Kuchenmeister, J. Wang, K. Wade, in: P.A. Jacobs, N.I. Jaeger, L. Kubelková,
B. Wichterlová (Eds.), Zeolite Chemistry and Catalysis, Elsevier, Amsterdam, Stud. Surface Sci.
Catal., vol. 69, 1991, pp. 485–497.
[15] A. Clearfield, in: C.A.C. Sequeira, M.J. Hudson (Eds.), Multifunctional Mesoporous Solids,
Kluwer, Amsterdam, 1993, pp. 159–178.
[16] S. Yamanaka, Mater. Sci. Forum 152–153 (1994) 69.
[17] K. Ohtsuka, Chem. Mater. 9 (1997) 2039.
[18] R. Szostak, C. Ingram, in: H.K. Beyer, H.G. Karge, I. Kiricsi, J.B. Nagy (Eds.), Catalysis by
Microporous Materials, Elsevier, Amsterdam, Stud. Surface Sci. Catal., vol. 94, 1995, pp. 13 – 38.
[19] T.J. Pinnavaia, M. Chibwe, V.R.L. Constantino, S.K. Yun, Appl. Clay Sci. 10 (1995) 117.
[20] A. Corma, Chem. Rev. 97 (1997) 2373.
[21] E. López-Salinas, Y. Ono, Microporous Mater. 1 (1993) 33.
[22] A.B.P. Lever, E. Montovani, B.S. Ramaswany, Can. J. Chem. 49 (1971) 1957.
[23] L.E. Alzamora, J.R.H. Ross, E.C. Kruissink, L.L. van Reijden, J. Chem. Soc. Faraday Trans. I
77 (1981) 665.
[24] J.R. Weisner, R.C. Srivastava, C.H.L. Kennard, M. di Vaira, E.C. Lingafelter, Acta Crystallogr.
23 (1967) 565.
[25] J. Brynestad, G.P. Smith, J. Am. Chem. Soc. 92 (1970) 3198.
[26] A.B.P. Lever, Inorganic Electronic Spectroscopy, 2nd edn, Elsevier, Amsterdam, 1984, p. 507.
[27] F.A. Cotton, D.M.L. Goodgame, M. Goodgame, J. Am. Chem. Soc. 83 (1961) 4690.
[28] K. Okada, F. Matsushita, S. Hayashi, Clay Min. 32 (1997) 299.
[29] E. López-Salinas, N. Tomita, T. Matsui, E. Suzuki, Y. Ono, J. Mol. Catal. 81 (1993) 397.
[30] E. López-Salinas, Y. Ono, E. Suzuki, Mater. Res. Symp. Proc. 368 (1995) 363.
[31] K. Itaya, H.-C. Chang, I. Uchida, Inorg. Chem. 26 (1987) 624.
[32] K. Chibwe, W. Jones, J. Chem. Soc. Chem. Commun. (1989) 926.
[33] S. Miyata, T. Hirose, Clays Clay Min. 26 (1978) 441.
[34] S. Kikkawa, M. Koizumi, Mater. Res. Bull. 17 (1982) 191.
[35] L.H. Jones, Inorg. Chem. 2 (1963) 777.
[36] F.M. Labajos, V. Rives, M.A. Ulibarri, Spectrosc. Lett. 24 (1991) 499.
[37] P.S. Braterman, C. Tan, J. Zhao, Mater. Res. Bull. 29 (1994) 1217.
116 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

[38] J.D. Wang, G. Serrette, Y. Tian, A. Clearfield, Appl. Clay Sci. 10 (1995) 103.
[39] P.K. Dutta, M. Puri, J. Phys. Chem. 93 (1989) 376.
[40] S. Idemura, E. Suzuki, Y. Ono, Clays Clay Min. 37 (1989) 553.
[41] H.G. Drickamer, S.C. Fung, G.K. Lewis Jr., Adv. High Pressure Res. 3 (1969) 1.
[42] M.J. Holgado, V. Rives, M.S. San Román, P. Malet, Solid State Ionics 92 (1996) 273.
[43] H.C.B. Hansen, C.B. Koch, Clays Clay Min. 42 (1994) 170.
[44] J.A. Olabe, H.O. Zerga, Inorg. Chem. 22 (1983) 4156.
[45] I. Crespo, C. Barriga, V. Rives, M.A. Ulibarri, Solid State Ionics 101 – 103 (1997) 729.
[46] E. Suzuki, S. Idemura, Y. Ono, Clays Clay Min. 37 (1989) 173.
[47] F.A.P. Cavalcanti, A. Schutz, P. Biloen, in: B. Delmon, P. Grange, P.A. Jacobs, G. Poncelet
(Eds.), Preparation of Catalysts IV, Elsevier, Amsterdam, 1987, pp. 165 – 174.
[48] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, J. Rouquerol, T. Sieminiewska, Pure Appl.
Chem. 57 (1985) 603.
[49] J.M. Fernández, M.A. Ulibarri, F.M. Labajos, V. Rives, J. Mater. Chem. 8 (1998) 2507.
[50] H. Nijs, P. Cool, E.F. Vansant, Interface Sci. 5 (1997) 83.
[51] G. Mao, M. Tsuji, Y. Tamaura, Clays Clay Min. 41 (1993) 731.
[52] D.W. Breck, W.G. Eversole, R.M. Milton, T.B. Reed, T.L. Thomas, J. Am. Chem. Soc. 78 (1956)
5963.
[53] T. Challier, R.C.T. Slade, J. Mater. Chem. 4 (1994) 367.
[54] B.R. Shaw, Y. Deng, F.E. Strillacci, K.A. Carrado, M.G. Fessehaie, J. Electrochem. Soc. 137
(1990) 3136.
[55] J. Labuda, M. Hudáková, Electroanalysis 9 (1997) 239.
[56] (a) J. Qiu, G. Villemure, J. Electroanal. Chem. 395 (1995) 159. (b) J. Qiu, G. Villemure, J.
Electroanal. Chem. 428 (1997) 165.
[57] A. Cervilla, A. Corma, V. Fornés, E. Llopis, P. Palanca, F. Rey, A. Ribera, J. Am. Chem. Soc.
116 (1994) 1595.
[58] A. Cervilla, E. Llopis, A. Ribera, A. Corma, V. Fornés, F. Rey, J. Chem. Soc. Dalton Trans.
(1994) 2953.
[59] A. Corma, V. Fornés, F. Rey, A. Cervilla, E. Llopis, A. Ribera, J. Catal. 152 (1995) 237.
[60] A. Corma, F. Rey, J.M. Thomas, G. Sankar, G.N. Greaves, A. Cervilla, E. Llopis, A. Ribera, J.
Chem. Soc. Chem. Commun. (1996) 1613.
[61] P. Palanca, T. Picher, V. Sanz, P. Gómez-Romero, E. Llopis, A. Doménech, A. Cervilla, J. Chem.
Soc. Chem. Commun. (1990) 531.
[62] V. Sanz, T. Picher, P. Palanca, E. Llopis, J.A. Ramı́rez, D. Beltrán, A. Cervilla, Inorg. Chem. 30
(1990) 3113.
[63] E. Llopis, A. Doménech, J.A. Ramı́rez, A. Cervilla, P. Palanca, T. Picher, V. Sanz, Inorg. Chim.
Acta 189 (1991) 29.
[64] H. Kominami, S. Kurimoto, M. Kubota, R. Shiozaki, Y. Kera, J. Ceram. Soc. Jpn. 105 (1997) 707
(in Japanese).
[65] M.A. Drezdzon, Inorg. Chem. 27 (1988) 4628.
[66] M.D. Newsham, E.P. Giannelis, T.J. Pinnavaia, D.G. Nocera, J. Am. Chem. Soc. 110 (1988) 3885.
[67] E.P. Giannelis, D.G. Nocera, T.J. Pinnavaia, Inorg. Chem. 26 (1987) 203.
[68] P.K. Ghosh, A.J. Bard, J. Phys. Chem. 88 (1984) 5519.
[69] R.A. Della Guardia, J.K. Thomas, J. Phys. Chem. 87 (1983) 990.
[70] H. Nijs, J.J. Fripiat, H. Van Damme, J. Phys. Chem. 87 (1983) 1279.
[71] J.R. Winkler, H.B. Gray, J. Am. Chem. Soc. 105 (1983) 1373.
[72] S. Miyata, A. Okada, Clays Clay Min. 25 (1977) 14.
[73] (a) S. Miyata, Clays Clay Min. 23 (1975) 369. (b) S. Miyata, Clays Clay Min. 31 (1983) 305.
[74] J.R. Winkler, H.B. Gray, Inorg. Chem. 24 (1985) 346.
[75] L. Barloy, J.P. Lallier, P. Battioni, D. Mansuy, Y. Piffard, M. Tournous, J.B. Valim, W. Jones,
New J. Chem. 16 (1992) 71.
[76] M.E. Pérez–Bernal, R. Ruano-Casero, T.J. Pinnavaia, Catal. Lett. 11 (1991) 55.
[77] S. Fukuzumi, S. Mochizuki, T. Tanaka, Isr. J. Chem. 23 (1987 – 1988) 29.
[78] L. Gaillon, F. Bedioui, J. Devinck, P. Battioni, J. Electroanal. Chem. 347 (1993) 435.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 117

[79] H. Van Damme, M. Crespin, F. Obrecht, M.I. Cruz, J.J. Fripiat, J. Colloid Interface Sci. 66 (1978)
43.
[80] H. Kameyama, H. Suzuki, A. Amano, Chem. Lett. (1988) 1117.
[81] C. Mousty, S. Therias, C. Forano, J.P. Besse, J. Electroanal. Chem. 374 (1994) 63.
[82] S. Therias, C. Mousty, Appl. Clay Sci. 10 (1995) 147.
[83] J. Martinsen, J.L. Stanton, R.L. Greene, J. Tanaka, B.M. Hoffman, J.A. Ibers, J. Am. Chem. Soc.
107 (1985) 6915.
[84] T.J. Pinnavaia, Adv. Chem. Ser. 245 (1995) 283.
[85] F. Bedioui, Coord. Chem. Rev. 144 (1995) 39.
[86] I.Y. Park, K. Kuroda, C. Kato, Chem. Lett. (1989) 2057.
[87] K. Sakoda, K. Kominami, M. Iwamoto, Jpn. J. Appl. Phys. 27 (1988) L1304.
[88] S. Bonnet, L. Bigey, C. Forano, A. de Roy, J.P. Besse, P. Maillard, M. Momenteau, in: M.L.
Occelli, H. Kessler (Eds.), Synthesis of Porous Materials: Zeolites, Clays and Nanostructures,
Marcel Dekker, New York, 1997, pp. 627 – 640.
[89] S. Bonnet, C. Forano, A. de Roy, J.P. Besse, P. Maillard, M. Momenteau, Chem. Mater. 8 (1996)
1962.
[90] A. Stone, E.B. Fleisher, J. Am. Chem. Soc. 90 (1968) 2735.
[91] H. Tagaya, A. Ogata, T. Kuwahara, S. Ogata, M. Karasu, J. Kadokawa, K. Chiba, Microporous
Mater. 7 (1996) 151.
[92] R.A. Schoonheydt, L. Heughebaert, Clay Miner. 27 (1992) 91.
[93] M. Chibwe, T.J. Pinnavaia, J. Chem. Soc. Chem. Commun. (1993) 278.
[94] M. Chibwe, L. Ukrainczyk, S.A. Boyd, T.J. Pinnavaia, J. Mol. Catal. A: Chem. 113 (1996) 249.
[95] K.A. Carrado, J.E. Forman, R.E. Botto, R.E. Winans, Chem. Mater. 5 (1993) 472.
[96] I.Y. Park, K. Kuroda, C. Kato, J. Chem. Soc. Dalton Trans. (1990) 3071.
[97] I.J. Shannon, T. Maschmeyer, G. Sankar, J.M. Thomas, R.D. Oldroyd, M. Sheehy, D. Madill,
A.M. Waller, R.T. Townsend, Catal. Lett. 44 (1997) 23.
[98] V.I. Iliev, A.I. Ileva, L.D. Dimitrov, Appl. Catal. A: Gen. 126 (1995) 333.
[99] V. Iliev, J. Mol. Catal. 85 (1993) L269.
[100] J.M. Assour, W.K. Kahn, J. Am. Chem. Soc. 87 (1965) 207.
[101] J.F. Boas, P.E. Fielding, A.G. McKay, Austr. J. Chem. 27 (1974) 7.
[102] V. Iliev, A. Andreev, D. Wohrle, G. Schulz-Ekloff, J. Mol. Catal. 66 (1991) L5.
[103] A. Skorobogaty, T.D. Smith, J. Mol. Catal. 16 (1982) 131.
[104] J. Zwart, H.C. van der Weide, N. Broeker, C. Rummens, G.C.A. Schuit, A.L. German, J. Mol.
Catal. 3 (1977–1978) 151.
[105] L. Ukrainczyk, M. Chibwe, T.J. Pinnavaia, S.A. Boyd, J. Phys. Chem. 98 (1994) 2668.
[106] R. Allmann, Chimia 24 (1970) 99.
[107] M.B. McBride, Clays Clay Miner. 27 (1979) 97.
[108] E.P. Giannelis, Chem. Mater. 2 (1990) 627.
[109] J. Subramanian, in: K.M. Smith (Ed.), Porphyrins and Metalloporphyrins, Elsevier, Amsterdam,
1975, p. 568.
[110] L. Ukrainczyk, M. Chibwe, T.J. Pinnavaia, S.A. Boyd, Environ. Sci. Technol. 29 (1995) 439.
[111] U.E. Krone, R.K. Thauer, H.P.C. Hogenkamp, Biochemistry 28 (1989) 4908.
[112] U.E. Krone, K. Laufer, R.K. Thauer, H.P.C. Hogekamp, Biochemistry 28 (1989) 10061.
[113] C.J. Gantzer, L.P. Wackett, Environ. Sci. Technol. 25 (1991) 715.
[114] N. Assaf-Amid, K.F. Hayes, T.M. Vogel, Environ. Sci. Technol. 28 (1994) 246.
[115] G.M. Klecka, S.J. Gonsior, Chemosphere 3 (1984) 391.
[116] Y.C. Yang, J.R. Ward, R.P. Seiders, Inorg. Chem. 24 (1985) 1765.
[117] M. Zikmund, K. Putyera, K. Hrnciarova, Chem. Papers 50 (1996) 262.
[118] L. Gaillon, F. Bedioui, J. Devynck, P. Battioni, J. Electroanal. Chem. 347 (1993) 435.
[119] D.S. Robins, P.K. Dutta, Langmuir 12 (1996) 402.
[120] R.F. Pasternack, L. Francesconi, D. Raff, E. Spiro, Inorg. Chem. 12 (1973) 2606.
[121] K. Kalyanasundaram, M. Naumann-Spallart, J. Phys. Chem. 86 (1982) 5163.
[122] J. Wang, Y. Tian, R.-C. Wang, J.L. Colón, A. Clearfield, in: R.L. Bedard, T. Bein, M.E. Davis,
J. Garces, V.A. Maroni, G.D. Stucky (Eds.), Synthesis/Characterization and Novel Applications
of Molecular Sieve Materials, Materials Research Society, Pittsburgh, 1991, pp. 63 – 80.
118 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

[123] T. Tatsumi, K. Yamamoto, Trans. Mater. Res. Soc. Jpn. 15 (1994) 141.
[124] C.W. Hu, Q.L. He, E.B. Wang, Progr. Nat. Sci. 6 (1996) 524.
[125] S. Miyata, T. Kumura, H. Hattori, K. Tanabe, Nippon Kagaku Zasshi 92 (1971) 514 (in
Japanese).
[126] C. Misra, A.J. Perrotta, Clays Clay Min. 40 (1992) 145.
[127] K. El Malki, A. de Roy, J.P. Besse, Eur. J. Solid State Inorg. Chem. 26 (1989) 339.
[128] C. Depège, C. Forano, A. de Roy, J.P. Besse, Mol. Cryst. Liq. Cryst. 244 (1994) 161.
[129] L. Bigey, C. Depège, A. de Roy, J.P. Besse, J. Phys. IV France, C2, 7 (1997) 949.
[130] S. Yamanaka, T. Sako, K. Seti, M. Hattori, Solid State Ionics 53 – 56 (1992) 527.
[131] C. Forano, A. de Roy, C. Depège, M. Khaldi, F.Z. El Metoui, J.P. Besse, in: M.L. Occelli, H.
Kessler (Eds.), Synthesis of Porous Materials: Zeolites, Clays and Nanostructures, Marcel Dekker,
New York, 1997, pp. 607–625.
[132] E. Suzuki, Y. Ono, Bull. Chem. Soc. Jpn. 61 (1988) 1008.
[133] H. Shimada, K. Saito, Nippon Kagaku Kaishi (1997) 335 (in Japanese).
[134] T. Kwon, G.A. Tsigdinos, T.J. Pinnavaia, J. Am. Chem. Soc. 110 (1988) 3653.
[135] J. Twu, P.K. Dutta, J. Phys. Chem. 93 (1989) 7863.
[136] W.P. Griffith, T.D. Wolkins, J. Chem. Soc. A (1966) 1087.
[137] W.P. Griffith, P.J.B. Lesniak, J. Chem. Soc. A (1969) 1066.
[138] A. Bhattacharyya, D.B. Hall, T.J. Barnes, Appl. Clay Sci. 10 (1995) 57.
[139] J. Twu, P.K. Dutta, J. Catal. 124 (1990) 503.
[140] R. Gopal, C. Calvo, Acta Crystallogr. Sect. B 30 (1974) 2491.
[141] C. Depège, L. Bigey, C. Forano, A. de Roy, J.P. Besse, J. Solid State Chem. 126 (1996) 314.
[142] K.S. Han, L. Guerlou-Demourgues, C. Delmas, Solid State Ionics 84 (1996) 227.
[143] K.S. Han, L. Guerlou-Demourgues, C. Delmas, Solid State Ionics 98 (1996) 85.
[144] M. Doeuff, T. Kwon, T.J. Pinnavaia, Synthetic Metals 34 (1989) 609.
[145] G.M. Woltermann, US patent 4,454,244, Ashland Oil Co., June 12, 1984.
[146] C. Barriga, W. Jones, P. Malet, V. Rives, M.A. Ulibarri, Inorg. Chem. 37 (1998) 1812.
[147] J. Evans, M. Pillinger, J. Zhang, J. Chem. Soc. Dalton Trans. (1996) 2963.
[148] K. Chibwe, W. Jones, Chem. Mater. 1 (1989) 489.
[149] F. Kooli, W. Jones, Inorg. Chem. 34 (1995) 6237.
[150] F. Kooli, W. Jones, V. Rives, M.A. Ulibarri, J. Mater. Sci. Lett. 16 (1997) 27.
[151] F. Kooli, M.J. Holgado, V. Rives, S. San Roman, M.A. Ulibarri, Mater. Res. Bull. 32 (1997) 977.
[152] E. Narita, P. Kaviratna, T.J. Pinnavaia, Chem. Lett. (1991) 805.
[153] E. Narita, P.D. Kaviratna, T.J. Pinnavaia, J. Chem. Soc. Chem. Commun. (1993) 60.
[154] J. Wang, Y. Tian, R.-C. Wang, A. Clearfield, Chem. Mater. 4 (1992) 1276.
[155] F. Kooli, V. Rives, M.A. Ulibarri, W. Jones, Mater. Res. Soc. Symp. Proc. 371 (1995) 143.
[156] G. Mascolo, O. Marino, Miner. Mag. 43 (1980) 619.
[157] M.A. Ulibarri, F.M. Labajos, V. Rives, W. Kagunya, W. Jones, R. Trujillano, Mol. Cryst. Liq.
Cryst. 244 (1994) 167.
[158] M.A. Ulibarri, F.M. Labajos, V. Rives, W. Kagunya, W. Jones, R. Trujillano, Inorg. Chem. 33
(1994) 2592.
[159] I.C. Chisem, W. Jones, J. Mater. Chem. 4 (1994) 1737.
[160] F. Kooli, V. Rives, M.A. Ulibarri, Inorg. Chem. 34 (1995) 5114.
[161] F. Kooli, V. Rives, M.A. Ulibarri, Mater. Sci. Forum. 152 – 153 (1994) 375.
[162] O. Clause, B. Rebours, E. Merlen, F. Trifiro, A. Vaccari, J. Catal. 133 (1992) 231.
[163] F. Kooli, V. Rives, M.A. Ulibarri, Inorg. Chem. 34 (1995) 5122.
[164] J. Guo, Q.Z. Jiao, G. Xiong, H.J. Lu, D.Z. Jiang, E.Z. Min, Chin. Chem. Lett. 7 (1996) 531.
[165] W. Kagunya, W. Jones, Appl. Clay Sci. 10 (1995) 95.
[166] V. Rives, F.M. Labajos, M.A. Ulibarri, P. Malet, Inorg. Chem. 32 (1993) 5000.
[167] P. Malet, J.A. Odriozola, F.M. Labajos, V. Rives, M.A. Ulibarri, Nucl. Instr. Methods Phys. Res.
B 97 (1995) 16.
[168] F. Kooli, I. Crespo, C. Barriga, M.A. Ulibarri, V. Rives, J. Mater. Chem. 6 (1996) 1199.
[169] A. Corma, J.M. López-Nieto, N. Paredes, M. Pérez, Appl. Catal. 97 (1993) 159.
[170] A. Corma, J.M. López-Nieto, N. Paredes, Appl. Catal. 104 (1993) 161.
 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120 119

[171] M.A. Chaar, D. Patel, M.C. Kung, H.H. Kung, J. Catal. 105 (1987) 483.
[172] F. Roozeboom, J. Medema, P.J. Gellings, Z. Phys. Chem. (Wiesbaden) 111 (1978) 215.
[173] R.L. Coustumer, B. Taouk, M. le Meur, E. Payen, M. Guelton, J. Grimblot, J. Phys. Chem. 92
(1988) 1230.
[174] E. López Salinas, Y. Ono, Bull. Chem. Soc. Jpn. 65 (1992) 2465.
[175] F. Kooli, C. Martin, V. Rives, Langmuir 13 (1997) 2303.
[176] C. Delmas, Y. Borthomieu, J. Solid State Chem. 104 (1993) 345.
[177] M. Ménétrier, K.S. Han, L. Guerlou-Demourgues, C. Delmas, Inorg. Chem. 36 (1997) 2441.
[178] S. Onodera, Y. Ikegami, Inorg. Chem. 19 (1980) 615.
[179] M. del Arco, C. Martin, I. Martin, V. Rives, R. Trujillano, Spectrochim. Acta Part A 49 (1993)
1575.
[180] L. Pesic, S. Salipurovic, V. Markovic, D. Vucelic, W. Kagunya, W. Jones, J. Mater. Chem. 2
(1992) 1069.
[181] M. del Arco, M.V.G. Galiano, V. Rives, R. Trujillano, P. Malet, Inorg. Chem. 35 (1996) 6362.
[182] M. del Arco, V. Rives, R. Trujillano, P. Malet, J. Mater. Chem. 6 (1996) 1419.
[183] L. Lowenstein, Am. Miner. 39 (1954) 92.
[184] K. Fuda, K. Suda, T. Matsunaga, Chem. Lett. (1993) 1479.
[185] V. Rives, M.A. Ulibarri, A. Montero, Appl. Clay Sci. 10 (1995) 83.
[186] B.I. Intorre, A.E. Martell, J. Am. Chem. Soc. 82 (1960) 358.
[187] F. Rey, V. Fornés, J.M. Rojo, J. Chem. Soc. Faraday Trans. 88 (1992) 2233.
[188] T. Hibino, A. Tsunashima, Chem. Mater. 9 (1997) 2082.
[189] J. Twu, P.K. Dutta, Chem. Mater. 4 (1992) 398.
[190] D. Levin, S.L. Soled, J.Y. Ying, Chem. Mater. 8 (1996) 836.
[191] D. Levin, S.L. Soled, J.Y. Ying, ACS Symp. Ser. 622 (1996) 237.
[192] M.P. Pope, Heteropoly and Isopoly Oxometalates, Springer-Verlag, New York, 1983.
[193] C. Preyssler, Bull. Soc. Chim. Fr. 1 (1970) 30.
[194] T. Kwon, T.J. Pinnavaia, Chem. Mater. 4 (1989) 381.
[195] Y.Y. Liu, C.W. Hu, Z.P. Wang, J.Y. Zhang, E.B. Wang, Sci. China Ser. B Chem. 39 (1996) 86 (in
Chinese).
[196] E.D. Dimotakis, T.J. Pinnavaia Jr., in: E.W. Corcoran Jr., M.J. Ledoux (Eds.), Synthesis and
Properties of New Catalysts: Utilization of Novel Materials Components and Synthetic Tech-
niques, Materials Research Society, Pittsburgh, 1990, pp. 77 – 80.
[197] E.D. Dimotakis, T.J. Pinnavaia, Inorg. Chem. 29 (1990) 2393.
[198] S.K. Yun, V.R.L. Constantino, T.J. Pinnavaia, Microporous Mater. 4 (1995) 21.
[199] H.C.B. Hansen, R.M. Taylor, Clay Miner. 26 (1991) 311.
[200] M.R. Weir, J. Moore, R.A. Kydd, Chem. Mater. 9 (1997) 1686.
[201] J. Guo, T. Sun, J.P. Shen, D.Z. Jiang, E.Z. Min, Chem. J. Chin. Univ. 16 (1995) 512 (in Chinese).
[202] E. Serwicka, P. Nowak, K. Bahranowski, W. Jones, F. Kooli, J. Mater. Chem. 7 (1997) 1937.
[203] R.N. Hua, Q.J. Shan, J.A. Gong, L.Y. Qu, Chem. J. Chin. Univ. 17 (1996) 1500 (in Chinese).
[204] R.N. Hua, Q.J. Shan, B.Z. Zhao, Y.H. Wang, B.L. Li, L.Y. Qu, Acta Chim. Sin. 55 (1997) 773
(in Chinese).
[205] R.S. Weber, P. Gallezot, F. Lefebvre, S.L. Suib, Microporous Mater. 1 (1993) 223.
[206] C.-W. Hu, Q.-L. He, Y.-H. Zhang, Y.-Y. Liu, Y.-F. Zhang, T.-D. Tang, J.-Y. Zhang, E.-B. Wang,
J. Chem. Soc. Chem. Commun. (1996) 121.
[207] C.W. Hu, Y.Y. Liu, Z.P. Wang, E.B. Wang, Acta Chim. Sin. 55 (1997) 49 (in Chinese).
[208] C.W. Hu, X. Zhang, Q.L. He, E.B. Wang, S.W. Wang, Q.L. Guo, Transit. Metal Chem. 22 (1997)
197.
[209] E. López Salinas, P. Salas Castillo, Y. Ono, Mater. Res. Soc. Symp. Proc. 371 (1995) 163.
[210] S.K. Yun, T.J. Pinnavaia, Inorg. Chem. 35 (1996) 6853.
[211] R.G. Finke, M.V. Droege, P.J. Domaille, Inorg. Chem. 26 (1987) 3886.
[212] C. Rocchiccioli-Deltcheff, R. Thouvenot, R. Franck, Spectrochim. Acta Part A 32 (1976) 587.
[213] T. Kwon, T.J. Pinnavaia, J. Mol. Catal. 74 (1992) 23.
[214] J. Guo, Q.Z. Jiao, J.P. Shen, H.J. Lu, D. Liu, D.Z. Jiang, E.Z. Min, Acta Chim. Sin. 54 (1996)
357 (in Chinese).
120 V. Ri6es, M. Angeles Ulibarri / Coordination Chemistry Re6iews 181 (1999) 61–120

[215] M.A. Drezdzon, in: R. Terry, K. Baker, L.L. Murrel (Eds.), Novel Materials in Heterogeneous
Catalysis, American Chemical Society, Washington, DC, 1990, pp. 141 – 148.
[216] K. Putyera, J. Jagiello, T.S. Bandosz, J.A. Schwarz, J. Chem. Soc. Faraday Trans. 92 (1996) 1243.
[217] T. Tatsumi, K. Yamamoto, H. Tajima, H. Tominaga, Chem. Lett. (1992) 815.
[218] T. Tatsumi, H. Tajima, K. Yamamoto, H. Tominaga, in: L. Guczi, F. Solymosi, P. Teteny (Eds.),
New Frontiers in Catalysis, Elsevier, Amsterdam, 1993, pp. 1703 – 1706.
[219] E. Gardner, T.J. Pinnavaia, Appl. Catal. A: Gen. 167 (1998) 65.
[220] C. Hu, Q. He, Y. Zhang, E. Wang, T. Okuhara, M. Misono, Catal. Today 30 (1996) 141.
[221] J. Guo, Q.Z. Jiao, J.P. Shen, D.Z. Jiang, G.H. Yang, E.Z. Min, Catal. Lett. 40 (1996) 43.
[222] X. Zheng, W. Yue, H. Heming, J. Dazben, in: H. Hattori, M. Misono, Y. Ono (Eds.), Acid-Base
Catalysis II, Elsevier, Amsterdam, Stud. Surface Sci. Catal. 90 (1994) 279.
[223] W. Kagunya, Z. Hassan, W. Jones, Inorg. Chem. 35 (1996) 5970.
[224] B. Keita, A. Belhouari, L. Nadjo, J. Electroanal. Chem. 314 (1991) 345.
[225] B. Keita, A. Belhouari, L. Nadjo, J. Electroanal. Chem. 355 (1993) 235.
[226] T. Sato, T. Wakabayashi, M. Shimada, I&EC Prod. Res. Develop. 25 (1986) 89.
[227] S.W. Rhee, M.J. Kang, H. Kim, C.H. Moon, Environ. Tech. 18 (1997) 231.
[228] L. Châtelet, J.Y. Bottero, J. Yvon, A. Bouchelaghem, Colloids Surf. A. Physicochem. Eng. Asp.
111 (1996) 167.
[229] W. Stumm, J.J. Morgan, Aquatic Chemistry — An Introduction Emphasizing Chemical Equilibria
in Natural Waters, 2nd ed., Wiley, New York, 1981, Ch. 5.
[230] M.J. Kang, S.W. Rhee, H. Moon, V. Neck, T. Fanghanel, Radiochim. Acta 75 (1996) 169.
[231] T. Yamagishi, Y. Oyanagi, E. Narita, Nippon Kagaku Kaishi, (1993) 329 (in Japanese).
[232] T.J. Pinnavaia, M. Rameswaran, E.D. Dimotakis, E.P. Giannelis, E.G. Rightor, Faraday Discuss.
Chem. Soc. 87 (1989) 227.
[233] C.R. Eady, P.F. Jackson, B.F.G. Johnson, J. Lewis, M.C. Malatesta, M. McPartlin, J.H. Nelson,
J. Chem. Soc. Dalton Trans. (1980) 383.
[234] E.P. Giannelis, E.G. Rightor, T.J. Pinnavaia, J. Am. Chem. Soc. 110 (1988) 3880.
[235] F. Bergaya, H. Van Damme, J. Chem. Soc. Faraday Trans. II 79 (1983) 505.
[236] T. Sato, H. Okuyama, T. Endo, M. Shimada, React. Solids 8 (1990) 63.
[237] T. Sato, K. Masaki, T. Yoshiaki, A. Okuwaki, J. Chem. Tech. Biotechnol. 58 (1993) 315.