Bergaya F. Handbook of Clay Science

Подождите немного. Документ загружается.

but is laterally uncorrelated despite the highly ordered stacking of the layers (3R).

This random anionic layer stacking was attributed to the high degree of degeneracy

(26-fold) of carbonate ion sites. Bellotto et al. (1996a) examined the layer stacking

arrangement of MgAl, MgGa, and NiAl by XRD analysis with the DIFFAX pro-

gram. The occurrence of faults is quantiﬁed by a ‘fault pro bability’, describing the

probability for a speciﬁc layer sequence to be arranged in rhombohedral or hex-

agonal stacking. The stacking arrangement is random for the solids investigated,

except for MgAl2CO

2

, which shows a preference for the rhombohe dral polytype.

This feature is assigned to a predominantly Coulombic layer–interlayer bondin g.

The stacking sequence is not affected by hydrother mal treatments at 470 K, which

only increased crystallinity. The higher entropy of random sequences can account for

their therm al stability.

D. Guest– Host Interactions

Interactions between guests (anions) and host (the brucite-like layer) are controlled

by electrostatic interactions and hydrogen bonding between the hydroxyl surface

groups and the anions. The interaction depends on the ordering of the metal cations

in the layer, and the chemical nature of the anion. For ordered LDH structures

such as [Li–Al], the anions are aligned perpendicular to the Li

–Li

axis (Fogg and

Hare, 1999), showing the key role of electrostatic interactions in this structure.

Generally, oxoanions strongly interact with the hydroxyl surface groups and a

dense packing of the anions is often achieved. Thi s can be explained in terms of

symmetry compatibility between the oxoanions polyedra and the octahedral layers

of the host structure. The various hexagonal or rhomboedral stacking of the LDH

layers deﬁne octahedral, prismatic, and tetrahedral interlayer crystallographic sites,

which can ideally accommodate monomeric oxoanions. Optimal hydrogen bonds are

formed when the Td or Oh units are arranged face to face or corner to corner. As an

example, the structure of the [Zn–Cr–SO

] interlayer domains is described as an

ordered arrangement of alternate inverse interlayer SO

tetrahedra. The tetrahedra

retain their C

-axis perpendicular to the layer, with one oxygen pointing to a metal

cation of the octahedral layer and the other three oxygens facing three OH grou ps of

the opposite layer. The short hydrogen bond lengths of 0.293 and 0.271 nm indicate

strong interactions betw een the layers and sulphate anions, resulting in a 0.892 nm

basal spacing (Khaldi et al., 1997). The structural series of XO

-containing LDH

([Zn–Cr–SO

], [Zn–Al–SO

], [Cu–Cr–SO

], [Zn–Al–CrO

], [Cu–Cr–CrO

[Zn–Cr–Cr

], and [Cu–Cr–Cr

]) shows basal spacings close to the d value of

[Zn–Cr–SO

]. In all these phases, the oxoanions are orientated similar to the SO

2

ions in the reference material. Similar d values obtain even with the more ﬂexible

because the anions can lie ﬂat and parallel to the layers (Bigey et al., 1997).

When interlayer anions strongly interact with the layers, modiﬁcation of molec -

ular dynamic is expected. Intercalation of CO

2

; SO

2

; and ClO



into [Mg–Al]

LDH lowers the symmetry of the anions due to hydrogen bonding with intercalated

13.1.5. Structure 1041

water molecules and OH groups as shown by Fourier transform infrared (FTIR) and

Raman spectroscopy (Kloprogge et al., 2002). On the other hand, the NO



anions

seem to be randomly distributed in the interlayer space because their vibration bands

remain unaffected during intercalation.

Cell parameter reﬁnement for intercalat ed silicate anions [Zn–Cr–SiO

] gives the

typical hexagonal 3R lattice as for [Zn–Cr–Cl] (Schu

tz and Biloen, 1987; Depe

et al., 1996). Additional reﬂections of the [Zn–Al–SiO

] phase are attributed to a

superstructural arrangement of the silicate species in the LDH interlayer space. The

XRD pattern can be indexed in the 1H stacking hexagonal mode with a unit

cell parameter a ¼ a

ﬃﬃﬃ

. The

Si NMR chemical shifts are similar to those in

phyllosilicates, suggesting that the SiO

4

groups in the LDH phases are arranged

similarly, and that the silicon environment is either of type Si(OSi)

3y

(OM)

(OH)

or Si(OSi)

3y

(OM)

(OM

) (with typically M ¼ Al and M

in octahedral site)

(Fig. 13.1. 4 ).

Very short basal spacings are obtained with voluminous anions such as V

6

because of a ﬂat orientation of the decaoctahedra building block of V

6

. The

limited size of the oxometallate anions and their low charge density do not allow the

formation of interlayer microporosity. Indeed, the basal spacing of LDH containing

oxopolyanions ne ver surpasses 1.4 nm, whereas intercalation of Al

hydroxocations

into clay minerals yields a distinctly greater interlayer expansion (Pinnavaia, 1987).

Organic anions always interact with their anionic groups (2CO



; 2SO



;

2PO

2

; 2OSO



; 2OPO

2

) being strongly hydrogen bonded to the surface hy-

droxyl groups, while their hydrophobic hydrocarbon chains are pushed far away

from the hydrophilic layer surface, and adopt the lowest energy conformation (Meyn

et al., 1990). a, o-Dianionic molecules can bridge two adjacent layers, giving a basal

spacing proportional to the length of the hydrocarbon chain. With monovalent long-

chain anions, either a bilayer or an interdigitated arrangement is formed. As a result,

the basal spacing is usually greater than 1.5 nm (Sanz and Serratosa, 1984; Meyn et

al., 1990; Bonnet et al., 1996; Pre

vot et al., 1998; Costantino et al., 1999).

4.6 Å

2.0 Å

2.7 Å

d calc = 12.0 Å

d exp = 11.85 Å

2.0 Å

4.6 Å

d calc = 6.6 Å

d exp = 7.65 Å

Fig. 13.1.4. Polycondensation of silicate layers in double layered hydroxides.

Chapter 13.1: Layered Double Hydroxides1042

In particular, the interlayer packing depends on the relation between the charge

and size of the organic anions and the equivalent area of the layer. When the cross-

sectional area of the anions is similar to the equivalent area, optimal packing is

obtained. This is the case for dodecylsulphate-containing [Zn

Al] LDH. The anions

adopt the same packing as in their sodium salt, with electrostatic and hydrogen

bonding between anionic head groups and the surface OH groups and van der Waals

bonding between lateral hydrocarbon chains. Guest–guest interactions are then de-

termined by the layer cha rge density, which is a determining factor in the adsorption

of neutral molecules (alkyl amines and alcohols) in alkylsulphate-containing LDH

(Boehm et al., 1977 ; Kopka et al., 1988; Meyn et al., 1990, 1993), catalysis reactions,

and in situ polymerization (Leroux and Besse, 2001).

E. Analytical Methods

Since synthetic LDH are commonly not highly crystalline, various and co mplemen-

tary techniques are needed to get structural information.

XRD

Diffraction peaks for syn thetic LDH are usually reﬁned in R-3m space group in

rhomboedral symmetry, and the parameters calculated according to the equation

1=d

¼½4ðh

þ hk þ k

Þ=3a

þl

. The basal spacing is expected to decrease with

increasing layer charge density (and vice versa) due to enhanced electrostatic

attraction between hydroxide layer s and interlayer anions. For an ideal in-plane

octahedral assembly, the size of the octahedra sharing edges is related to the metal-

oxygen distance (d) by the relation a ¼

ﬃﬃﬃ

dðM2OÞ and the latter to the ionic radius

(r) by the relation d(M–O) ¼ (1x) r(M

)+xr(M

). The variation of the lattice

parameter a with the M(III) content is given by da=dx ¼

ﬃﬃﬃ

½rðM

2þ

ÞrðM

3þ

Þ.

This relation is often used to assess the cation substitution in LDH.

The coherence length along the stacking direction may be estimated from the (0 0 l)

FWHM using the Scherrer formula D

hkl

¼ Kl/b

1/2

cos y, where l is the X-ray wave-

length, y the diffraction angle; b

1/2

the full-width at half-maximum intensity; and K is

a constant usually taken to be equal to 0.9 (CuK

radiation). The values are highly

dependent on the LDH composition. For synthetic hydrotalcite (Mg

Al) a value of

90 nm is calculated, corresponding roughly to 100 layers (Oriakhi et al., 1997), 37 nm

for Zn

Al (Pre

vot et al., 1998), and 18 nm for Ni

Al (Oriakhi et al., 1997).

EDX is useful for studying the kinetics of formation or exchange reaction. When

time resolved, it allows staging phenomena in some LDH to be observed (Khan and

O’Hare, 2 002).

Neutron Diffraction

Structures of LiAl

(OH)

X(X¼ Cl, Br, NO

) dehydrated phases may be determined

by neutron powder diffraction technique (Besserguenev et al., 1997). These phases

are isomorphous and crystallize in the P6

/mcm space group. This technique is also

13.1.5. Structure 1043

applied to improve understanding of the thermal evolution of the Pd/Mg/Al system

(segregation and oxide formation) (Basile et al., 2000).

Inelastic neutron scattering was used to study the interlayer water (Kagunya et al.,

1998). Quasi-elastic neutron scattering showed previously that the water molecules

are not ﬁxed in one location, but exhibi t translational diffusion as well as reorien-

tation motions. The basal spacing of the LDH inﬂuences water mobility (Kagunya,

1996).

XAS

Because of the poorly crystalline nature of LDH, XAS was widely used to investigate

LDH formation (Roussel et al., 2001; Yamaguchi et al., 2002). XAS indicates local

ordering for pyroaurite (Vucelic et al., 1997) and other synthetic LDH such as

1y

(OH)

Cl nH

O(Intissar et al., 2002).

EXAFS shows structural changes arising from substitution s as in

(Co

1y

)

Al(OH)

Cl nH

O, where for y ¼ 1 the local structure is similar to that

of bottalackite (Leroux et al., 2002). When the M(II)/M(III) ratio is greatly changed

as in Co

Al(OH)

2(1+x)

Cl nH

O, the decrease of layer charge density is compensated

by oxidation Co(II) to Co(III) (Leroux et al., 2001a)(Fig. 13.1.5).

EXAFS is also used to determine the interlayer structure and possible grafting

reactions. For example, the W LIII-edge and Nb K-edge absorption shows that

hexametallate anions, Nb

6x

(x+2)

, orient with their C

-axis perpendicular to

the host layers (Evans et al., 1996). For CrO

2

and Cr

2

, Cr–K edge absorption

Fig. 13.1.5. Moduli of the Fourier transform for Co

Al(OH)

2(n+1)

Cl nH

O: (a) n ¼ 3 and

(b) n ¼ 2. The distances are not corrected from atomic phases. The ideal cation intralamellar

arrangement is displayed. Co K-edge EXAFS spectra were recorded at the beam line D44

(LURE, Orsay).

Chapter 13.1: Layered Double Hydroxides1044

shows that the decrease in basal spacing is related to grafting of these oxoanions to

the surface OH groups (Malherbe et al., 1999).

Sulphur K-edge absorption was used to determine the degree of interaction of

styrene sulphonate with the ZnAl LDH surface. To illustrate the high selectivity of

the XAS technique, an ill-deﬁned ZnS-like phase was detected after heating to 600 1C

under nitrogen (Moujahid et al., 2003).

Raman and FTIR

The application of both Raman and FTIR spectroscopy to the characterization of

LDH has been reviewed by Kloprogge and Frost (2001) .

FTIR and Raman spectroscopy were used to investigate the local structures of

LDH such as M

(OH)

nH

O(M¼ Mg, Ni, Co) (Kloprogge and Frost,

1999) and Mg

6x

(OH)

(CO

) 4H

O(Johnson et al., 2002). Raman micro-

spectroscopy of hydrocalumites reveal the dynami c disorder resulting from the free

rotation of the nitrate groups around the [0 0 1] axis, and from the half-bonded and

half-free H

O molecules (Renaudin and Franc-ois, 1999). The orientation of Co

and Ni

phthalocyanine in the interlayer space was also studied by the Raman

spectroscopy. The results indicate inter actions of the p electrons with the surface,

causing the organic molecules to adopt a ﬂat orientation with respect to the surface

(De Fria et al., 1998).

Raman spectroscopy was also used to investiga te the kinetics and mechanisms of

catalytic acitivity, including the decomposition of H

by a molybdate-exchanged

LDH (Sels et al., 2001), and the epoxidation of geraniol by a vanadium-pillared

LDH (Villa et al., 2001).

FTIR spectro scopy is largely used to follow the alteration of the vibration bands

after intercalation and/or thermal treatment (Rives, 2002; Guo et al., 2002; Villegas

et al., 2002). This technique is also useful for assessing catalytic activity (e.g., cat-

alytic oxidation of sulphides (Choudary et al., 2002) and wet air oxidation of phenol

(Alejandre et al., 2001)), as well as surface acid–base properties. In the latter case, the

shift of a particular vibration band of a probe molecule, such as CDCl

or CO

,is

used to scale the interactions with basic surface sites, while the absorption bands for

pyridine and NH

provide an estimate of Brønsted and Lewis surface acidity (Lopez-

Salinas et al., 1997a, Chisem et al., 1998: Kooli et al., 1997c; Prinetto et al., 2000a).

Solid-State NMR

Different nuclei located either in the inorganic layers or the interlayer space may be

investigated by solid-state NMR spectroscopy.

Al NMR is commonly used to

determine the degree of conversion of Al

from octahedral to tetrahedral envi-

ronment during the thermal decomposition or reconstruction of LDH phases

(Rocha et al., 1999).

Al and

Ga MAS NMR were used to investigate the thermal

decomposition of Mg/Al and Mg/Ga LDH (Aramendia et al., 1999).

Characterization of the interlayer species and their interactions with the host

structure are the subject of numerous NMR studies. The ordering of interlayer water

13.1.5. Structure 1045

and carbonate was studied using

H and

C NMR techniques (van der Pol et al.,

1994).

Incorporation of C

into LDH was established by measuring the spin-lattice

relaxation time ( T

)by

C NMR. The C

guest molecule does not freely rotate as in

the pure solid form (Tseng et al., 1996).

Other notable NMR studies are concerned with intercalated silicates by

(Depe

ge et al., 1996), phenylphosphonic acid by

P(Carlino et al., 1996), styrene

sulphonate and poly(styrene) sulphonate by

C CP MAS (Mo ujahid et al., 2002a),

and borate by

B(del Arco et al., 2000). Static

V NMR spectra indicate hyperﬁne

magnetic interaction exerted by unpaired electrons of Ni

on the inserted vanadium

nuclei in NiCo LDH (Me

trier et al., 1997).

NMR experiments on uncommon nuclei were also reported. For example, Hou

and Kirkpatrick (2000) investigated the structures and dynamics of seleno-oxoanions

in LDH phases by solid-state

Se NMR. Although selenate is a tetrahedral anion,

the observed

Se chemical shift anisotropy is uniaxial when the anion is incorpo-

rated between the LDH layers. Static

Cl NMR spectroscopy was performed on

hydrotalcite and hydrocalumite samples (Hou et al., 2002). The collected data in-

dicate that the chlorine environment varies signiﬁcantly with the hydration state.

Al,

Cl, and

6,7

Li nuclei in LiAl

–Cl LDH phases were studied by Hou and

Kirkpatrick (2001).

Further developments in this area would be expecte d with the application of two-

dimensional heteronuclear-correlated MAS NMR spectroscopy, e.g.,

H-

Al and

Al-

P correlations (Fernandez et al., 2001; Alba et al., 2001).

Fe Mo

ssbauer Spectroscopy

This technique provided information on the mechanisms of oxidation of

Ni(II)–Fe(II) hydroxides in chloride-containing aqueous media leading to a Ni–Fe

pyroaurite-type material (Refait and Ge

nin, 1997).

Fe Mo

ssbauer spectroscopy

was also us ed to study the in situ behaviour of a hydrated iron-substituted nickel

hydroxide in a Ni/Cd electrochemical cell (Guerlou–Demourgues et al., 1996).

Computer Modeling

By comparison with clay minerals, only a few computer-modelling studies on LDH

materials were reported so far (Newman et al., 2001).

Molecular dynamics computer simulations of Mg/Al hydrotalcite containing Cl



were performed to improve understanding of the structure, hydration, and swelling

behaviour of LDH (W ang et al., 2001; Hou et al., 2002). Similar approache s were

used by Kalinichev and Kirkpatrick (2002) on Friedel salt (Ca

Al(OH)

Cl 2H

hydrocalumite) and ettringite to model some components of crystalline cement

phases. Bonapasta et al. (2001, 2002) modelled the c ross-linking of polyacrylate and

poly(vinyl alcohol) with Ca or Al atoms via the carboxylate or OH groups. This

reaction may be a key factor determining the ﬁlling of large voids in macro defect-

free cements.

Chapter 13.1: Layered Double Hydroxides1046

The surface adsorption and intercalation of ﬂuorescein anions into ZnAl hy-

drotalcite were computed based on the structure of the host and the dimensions of

the guest compound (Costantino et al., 2000).

Other models were used to estimate the relative layer rigidity of LDH materials in

comparison to other 2D structures. The dependence on composition of the basal

spacing, determined by XRD, was simulated using an extended version of the dis-

crete ﬁnite-layer rigidity model including both intralayer and interlayer rigidity

effects (Solin et al., 1995).

13.1.6. MORPHOLOGY

A. Size and Shape

The shape of LDH particles is obviously de pendent on the method of preparation. In

the standard coprecipitation method both nucleation and growth steps oc cur si-

multaneously during synthesis. Particles are obtained with a high degree of crys-

tallinity leading to well-resolved XRD patterns but with a wide range of sizes. The

control of particle size is more effective by hydrothermal treatment or homogeneous

precipitation using urea hydrolysis. Well-deﬁned plate-like hydrotalcite particles

with regular hexagonal shape and sizes between 2 and 20 mm were prepared (Ogawa

and Kaiho, 2002; Oh et al., 2002). The urea method always gives rise to larger

particles.

Yun and Pinnavaia (1995) compared the effect of coprecipitation conditions on

particle morphology and porosity. Fine-grained particles with rough surfaces, rel-

atively high surface areas, and mesopores of size in the range 5–30 nm in size are

obtained by copr ecipitation at varia ble pH. On the other hand, at constant pH,

larger, well-formed hexagon al particles with a narrower mesopore size distribution

(about 2 nm radius) are obtained. Edge-face particle aggregation and co-facial layer

stacking create interparticle mesopores. Structural defects in [Mg–Ga] were ob-

served by TEM ( Lopez-Salinas et al., 1998). Layers located nea r the edges of the

particles are more loose ly held together than those in the bulk of the particles.

Adachi-Pagano et al. (2003) studie d the effect of preparat ion procedure on particle

size and aggregation of Cu

Cr(OH)

Cl 2H

OandMg

Al(OH)

(CO

)

0.5

2H

(Fig. 13.1.6). Coprecipitation with long ageing time often favours a sand-rose

morphology.

Morphology is also affected by chemical composition. Partial substitution of

or Cu

for Mg

in synthetic hydrotalcite leads to a distinct change in

textural properties (Carja et al., 2001). The Cu

-substituted LDH displays a rel-

atively uniform porous structure with enlarged mesopores and decreased speciﬁc

surface area compared with the unsubstituted Mg

compound. The Fe

-substi-

tuted LDH develops microporosity, non-uniform particles, decreas ed pore sizes, and

increased speciﬁc surface area.

13.1.6. Morphology 1047

Hydrotalcite-like materials with a unique sheet-like morphology and a diameter/

thickness ratios >100 were prepared by reacting MgO, AlOOH, and a mon-

ocarboxylic acid in aqueous media at 80–90 1C for 6 h at pH>7 (Kelkar an d Schutz,

1997a; Kelkar, et al., 1997b). In contrast to conventional hydrotalcites with hex-

agonal particle morphology, these materials have high mechanical strength making

them useful industrially. Acicular particles of [Mg–Al] hydrotalcite were prepared by

coprecipitation, and they show a high endothermic DSC peak and a wide temper-

ature range of decomposition (Ren et al., 2002).

B. Speciﬁc Surface Area and Porosity

The speciﬁc surface area of a single LDH layer with known composition and struc-

ture may be derived from the formula

S ¼ a

ﬃﬃﬃ

18

N=M

where N is the Avogadro’s number ð¼ 6:022  10

Þ, a the cell parameter (nm), and

M the molecular weight of the unit formula. For Zn

Cr(OH)

Cl 2H

O and

Al(OH)

(CO

)

0.5

2H

O, the calculated S values are 817 and 1285 m

/g, respec-

tively. Such high values are not measured because the internal surfaces are often

2 µm

Fig. 13.1.6. SEM of Cu

Cr(OH)

Cl 2H

O prepared by (a) CuO/CrCl

-induced hydrolysis

and (b) coprecipitation (V. Pre

vot, unpublished results) and Mg

Al(OH)

(CO3)

0.5

2H

prepared by (c) coprecipitation and (d) urea method (Adashi-Pagano, et al., 2003).

Chapter 13.1: Layered Double Hydroxides1048

inaccessible. Typical values of the speciﬁc surface area measured by the BET tech-

nique range from 20 to 85 m

/g (del Arco et al., 1994). For Mg–Al LDH containing

simple anions such as carbonate, chloride, and nitrate, the surface area is mostly less

than 100 m

/g ( Pesic et al., 1992; Di Cosimo et al., 1998; Hussein et al., 2001; Inacio

et al., 2001). Hydrothermal crystallization decreases the surface area (Labajos et al.,

1992). Synthesis by coprecipitation in mixed solutions of water and organic solvents

increases the surface area (Malherbe et al., 1997; Inacio et al., 2001).

Nitrogen adsorption–desorption isotherms (type II with a narrow hysteresis) in-

dicate that LDH-containing simple anions are mesoporous with no microporosity

even when large POM pillars are intercalated (Labajos et al., 1992; Pesic et al., 1992;

Inacio et al., 2001). However, [Mg–Al] LDH intercalated by [Fe(CN)

] complexes

have an exceptionally high speciﬁc surface area of 499 m

/g (Nijs et al., 1999) and a

narrow pore size distribution with the majority of pores o0.71 nm. A correlation is

found between the Mg

/Al

ratio and the resulting microporosity after pillaring.

The optimal ratio is about 3.3, resulting in a pillared [Fe(CN)6]–MgA l–LDH with a

speciﬁc Langmuir surface area of 499 m

/g and a micr opore volume between 0.158

and 0.177 mL/g.

13.1.7. FUNDAMENTAL PROPERTIES

A. Chemical Stability

Chemical stability is of importance to many practical applications in which LDH

feature. For example, this parameter needs to be assessed when LDH are used as a

sink of radioactive metal ions from nuclear waste repositories (Allda et al., 2002). In

geochemical assessments, the stability of LDH is evaluated in terms of its solubility

in water (Ford et al., 1999; Scheckel et al., 2000). Solubility data for metal hydrox-

ides and LDH were derived from pH titration (Boclair and Braterman, 1998, 1999;

Boclair et al., 1999). The stability of LDH increases in the order Mg

oMn

E Ni

oZn

for divalent cations, and Al

oFe

for trivalent cations

(Boclair and Braterman, 1999). This trend accords with the pK

values of the cor-

responding metal hydroxides (K

is the solubility product). Because the pK

Mg(OH)

is smaller than that of Zn(OH)

, Mg-based LDH are more soluble. Thus,

the aqueous solution of a Mg-based LDH is more basic than the corresponding

aqueous solution of a Zn-based LDH. Saturated aqueous solutions of Mg–Al–Cl

LDH and Zn–Al–Cl LDH, deaerated with Ar, have pH values of 8.91 and 6.97,

respectively (Shaw et al., 1990).

In addition to direct measurements of solubi lity, Allda et al. (2002) calculated the

aqueous solubility of an LDH from therm ochemical data. The heat of formation and

the free energy of formation of the carbonate form of an LDH are equal to

the values for a physical mixture of the binary compounds, i.e., the metal hydroxides

and carbonates. Their simple physical mixture model indicates that the aqueous

13.1.7. Fundamental Properties 1049

solubilities of LDH are affected by the interlayer anions. Relative to carbonate,

silicate an d borate decrease solubility, while nitrate and sulphate increase it.

B. Thermal Stability

The thermal stability of LDH was intensely investigated because the thermal de-

composition products are interesting as catalysts. Despite the diversity in compo-

sition most LDH exhibit a similar thermal decomposition behaviour. When heated,

LDH release interlayer water up to ca. 250 1C, followed by dehydroxylation of the

hydroxide layers, and decomposition of the interlayer anions at higher temperatures.

The differential thermal analysis (DTA) curves of Mg–Al–CO

LDH show two

major endothermic peaks (Fig. 13.1.7). The ﬁrst peak (150–2 50 1C) corresponds to

the loss of interlayer water, while the seco nd (250–450 1C) arises from the decom-

position of the hydroxide layers and carbonate anions. In some cases, the second

peak splits into two peaks, the ﬁrst of which corresponds to the loss of hydroxyl

groups bound to Al, while the second of the split peaks represents the loss of hy-

droxyl groups bound to Mg together with decomposition of carbonate anions

(Miyata, 1980). These weight-loss steps are observed in the corres ponding TG curve.

The XRD patterns indicate that the layer structure doe s not collapse with the loss of

interlayer wat er (Fig. 13.1.8). Valente et al. (2000) measured the decomposition

temperatures of carbonate forms of LDH containing various metal cations under the

same experimental conditions. They found that the thermal stability increases in the

order Co–AloZn–AlE Cu–AloMg–FeENi–AloMg–AlEMg–Cr. The highest de-

composition temperature is ca. 400 1C for Mg–Al LDH and Mg–Cr LDH, and the

lowest decomposition tempe rature is 220 1C for Co–Al LDH.

Fig. 13.1.7. DTA curves of Mg–Al–CO

LDHs having various Mg/Al atomic ratios. The Mg/

Al ratio is indicated on each curve.

Chapter 13.1: Layered Double Hydroxides1050