Textural and Structural Modification of Homo-ionic Montmorillonite

761

*E-mail: [email protected].

Textural and Structural Modification of Homo-ionic Montmorillonite

Mervat S. Hassan* Central Metallurgical R & D Institute, Helwan, P.O. Box 87, Cairo, Egypt.

(Received 21 December 2008; revised form accepted 12 January 2010)

ABSTRACT: The textural and structural modifications of calcium bentonite(CaB) induced by ion exchange, heat treatment and acid leaching were studiedby infrared spectroscopy (IR), scanning electron microscopy (SEM) andnitrogen adsorption/desorption methods. The cation-exchange capacities (CEC)and the spontaneous and mechanical dispersion of the heat-treated samples werealso measured. CaB taken from the northern coast of Egypt consisting ofcalcium montmorillonite (85%) and kaolinite, in addition to traces of quartz, wasused in this study.

CaB and its sodium and lithium forms, NaB and LiB, were calcined at 600 °Cfor times ranging from 2 h to 20 h, while selected calcined samples were leachedwith H2SO4 for 1 h. Upon heating, the decrease in the layer charge caused by thefixation of Ca2+, Na+ and Li+ ions in the structure, as revealed by the CEC values,led to an upward shift of the Si–O stretching band in the IR spectrum to thefrequency at which the Si–O groups of pyrophyllite absorb. The CEC valuesshowed that the layer charge of the homo-ionic montmorillonite decreased withincreasing thermal treatment time and that the amounts of fixed cations were inthe order: Li+ > Na+ > Ca2+.

The dispersability of the thermally-treated CaB was higher than that of NaBand LiB, both before and after mechanical dispersion. The high dispersability ofCaB arises from the inability of the large Ca2+ cation to penetrate further into thelattice. The textural parameters of heated samples showed that the zig-zagalignment of the silicate chains varied as a function of the heat-treatment time,irrespective of the nature of the exchangeable cations. A significant increase inthe amount of pores with diameters in the range 13–38 Å was detected in the caseof LiB. Heating LiB further for up to 20 h led to a slight increase in the specificsurface area and pore volume, but with a considerable reduction in microporosity.On the other hand, thermal treatment after leaching led to a considerable increasein the specific surface areas, pore volumes and microporosities for those samplessubjected to heating for a short time, but decreased progressively with increasingheating time, with no difference being observed to the exchangeable cations. Thisincrease in surface area was mainly due to the presence of micropores whoseformation was attributed to the leaching of octahedral cations. On the other hand,following calcination for a long time, the magnitude of the textural parametersdiminished due to the gradual collapse of the microporous structure.

INTRODUCTION

The names bentonite, smectite and montmorillonite are used interchangeably. However, by nomeans do they refer to the same material. Montmorillonite is by far the most abundant mineral of

the smectite clay minerals. It is the predominant mineral which includes bentonite as well asFuller’s Earth worldwide and is a major component of the soils formed in arid and semi-arid areas.Bentonite is used widely in many fields of application, e.g. drilling mud, foundry clay, refiningand bleaching purposes, and as a backfill material for the construction of high-level nuclear wasterepositories.

When used in iron ore compaction, the ceramics and boundary industries, as powder catalystsand in the preparation of pillared clays, bentonite may be subjected to high temperatures. Inaddition, before the construction of bridges and buildings on smectite soils, the ground below thefoundations may be heat-treated up to 600 oC to harden the clays and reduce their swellingcharacteristics (Wang et al. 1990). Some physicochemical properties of bentonite such asswelling, plasticity, cohesion, compressibility, strength, cation-exchange capacity, particle size,adsorptive properties, pore structure, surface area, surface acidity and catalytic activity, as well asits mineralogy, are greatly affected by thermal treatment, acid activation and the nature of theexchanged cation, as well as by the pH of the clay suspension (Gates et al. 2000; Emmerich et al.2001; Venaruzzo et al. 2002; Neaman et al. 2003; Wu et al. 2005; Madejová et al. 2006; Volzone2007; Volzone and Ortiga 2009; Steudel et al. 2009).

The purpose of the present work was to shed some light on (1) the influence of thereorganisation of the structural phase in calcined homo-ionic bentonite on its swelling andexpandability properties, and (2) the influence of acid activation on the creation of a nanoporousstructure in dehydroxylated homo-ionic bentonite. It is assumed that such a material couldserve as a precursor for obtaining porous silicate ceramics as well as adsorbents and catalyticmaterials.

MATERIALS AND METHODS

A commercial Egyptian bentonite was selected for this study. This particular bentonite is minedon the northern coast of Egypt and is Miocene in age. About 85% of the crude clay consists ofmontmorillonite with a high iron content, with some substitution of Al3+ sites by Si4+ sites. Theprincipal exchangeable cation in this material is Ca2+ with a cation-exchange capacity of 93mequiv/100 g material. The main impurity in this bentonite is kaolin, with quartz in addition as anon-clay mineral.

Removal of iron oxide from the bentonite

The bentonite was fractionated by sedimentation and centrifugation to obtain a fraction with aparticle size < 2 µm. Free iron oxides were removed using the procedure of Mehra and Jackson(1960).

Preparation of homo-ionic clay

Homo-ionic samples containing Li+ and Na+ ions were prepared by saturating 5 g of the < 2 µmfraction for 12 h, once with 240 m� of a 0.2 M chloride solution of the respective salt, followedby three times with a 0.1 M chloride solution of the respective salt. The saturated fractions weredialysed using distilled water until the washings were free from chloride ions. The suspendedfractions were dried overnight at 60 °C.

762 Mervat S. Hassan/Adsorption Science & Technology Vol. 27 No. 8 2009

Thermal and acid treatment of homo-ionic clay

Purified calcium montmorillonite and the Na+ and Li+ ion-saturated samples were calcined at600 °C over a wide time range between 2 h and 20 h. Some of the calcined samples were treatedwith 25 wt% H2SO4 for 1 h at 98 °C according to the procedure of Duarte et al. (1995). Theactivated samples were washed several times until all traces of aluminium and sulphate ions hadbeen removed. The samples were dried overnight at 60 °C.

Mechanical and spontaneous dispersion of calcined fractions

The amount of spontaneously dispersible clay (calcium montmorillonite and its Na+ and Li+

forms) was determined by dispersing 1 mg of an appropriate sample in 50 m� distilled water andmeasuring the amount of clay remaining in suspension after 30 min using a turbidimeter. Inaddition, the amount of mechanically dispersed clay was determined by adding 1 mg of anappropriate sample to 50 m� distilled water and shaking the suspension overnight. The amount ofclay suspended in the solution after such treatment was again measured turbidimetrically using theabove procedure. A calibration curve was employed to convert the turbidity measurements into theamounts of dispersed clay.

Cation-exchange capacities

The cation-exchange capacities of the unmodified and modified montmorillonites were measuredusing the Methylene Blue technique according to Inglethorpe et al. (1993).

Surface areas and porosities

Nitrogen adsorption/desorption isotherms were measured employing a Nova-1200 instrument(Quantachrome, Boynton Beach, FL, U.S.A.) after degassing the various samples overnight at100 °C. The BET (Brunauer–Emmett–Teller) surface areas (SBET) were calculated from nitrogenadsorption isotherms measured at –196 oC over the relative pressure range 0.01–0.1P/P0. Theexternal surface areas (Sext) — to which the main contribution was from mesopores — and themicropore volumes (Vt) were calculated by the t-plot method (de Boer 1958). The differencebetween the specific surface area (SBET) and the external surface area is referred to as themicropore surface area (St). The total volume of pores with diameters < 500 Å (Vp) was obtainedfrom the adsorption data at a relative pressure of 0.98P/P0 as calculated by the BJH method(Barrett et al. 1951).

Characterisation

A Philips PW 1730 powder X-ray diffractometer employing Fe-filtered Co Kα radiation andoperating at 30 kV and 20 mA was used to examine the calcium montmorillonite and the homo-ionic montmorillonites. Infrared vibrational spectra were recorded at the University of Florida ona Nicolet Magna 750 Fourier-transform spectrometer. For each sample, 28 scans wereaccumulated over the 4000–400 cm–1 spectral range employing the transmittance mode and aresolution of 4 cm–1. The pressed KBr discs employed for this purpose were prepared using 0.4 mgof sample and 200 mg of KBr. To examine the morphology of the bentonite samples, the fracture

Textural and Structural Modification of Homo-ionic Montmorillonite 763

surfaces of selected samples coated with Au–Pd were observed at the University of Florida witha JSM-6400 scanning electron microscope.

RESULTS AND DISCUSSION

Sample composition

The principal component of the Egyptian bentonite clay was calcium montmorillonite, but it alsocontained kaolinite and quartz. The most prominent feature of the FT-IR spectrum of thisbentonite (Figure 1) was a wide band at 1027 cm–1, corresponding to the Si–O stretch of thephyllosilicate clay structure (Farmer 1974). This band was not fully resolved because of theconsiderable thickness of the clay film. Other structural vibrations were also observed. Thus,the bands at 3627 cm–1 and 3692 cm–1 correspond to the stretching vibrations of structural OHgroups, the band at 3692 cm–1 being characteristic of kaolinite while the 3627 cm–1 band iscommonly found in many different phyllosilicate minerals (Madejová et al. 2002). The band at690 cm–1 corresponds to the structural vibration of the OH group, while the band at 919 cm–1

corresponds to the deformation mode of the Al2OH group. The band at ca. 880 cm–1 may beassigned to AlFeOH vibrations, thereby demonstrating the presence of iron in the octahedralsheets of montmorillonite (Farmer 1974). Such AlMgOH vibrations are not observed in thespectrum of a homo-ionic montmorillonite (Andrejkovicová et al. 2006), thus proving that theoctahedral sheets had a relatively low Mg content. The remaining bands may be assigned tovibrations arising from water molecules. Thus, the bands at 3426 cm–1 correspond to the stretchingvibrations (ν3 and ν1 modes), whereas the 1642 cm–1 band corresponds to H–O–H bending in thewater molecule (ν2 mode).

Structural modification of tetrahedral sheets due to the presence of small cations either inhexagonal holes and/or in previously vacant octahedral sites induces changes in the Si–O vibrationmodes. As the negative charge associated with the layers decreases, the structure of the mineralbecomes more akin to pyrophyllite (Farmer 1974). On heating NaB and LiB to 600 oC, the broadband at ca. 1027 cm–1 assigned to complex Si–O stretching vibrations in the tetrahedral sheetmoved to 1046 cm–1 for all samples (see Figures 2 and 3), with no difference in the exchangeable


4000 3000

3426

3627

3692

1642

919

534

690

789

2000 1000 400

Wavenumber (cm−1)

Tra

nsm

ittan

ce (

a.u.

)

Figure 1. FT-IR spectrum of calcium bentonite (CaB).


4000 3000

343936

33

1643

362720 h

2 h

3426

1642

913

791

688

534

472

-731

798

482

560

1046

3692

2000 1000 400

Wavenumber (cm−1)

Tra

nsm

ittan

ce (

a.u.

)

Figure 2. FT-IR spectrum of sodium bentonite (NaB) and of NaB subjected to thermal treatment at 600 oC for 2 h and20 h, respectively.

4000 3000 2000 1000 400

1136

1644

798

1044 69

247

356

273

363

5

3633

3444

2 h

20 h

Wavenumber (cm−1)

Tra

nsm

ittan

ce (

a.u.

)

Figure 3. FT-IR spectrum of lithium bentonite (LiB) subjected to thermal treatment at 600 oC for 2 h and 20 h, respectively.

cations, thereby approaching the frequency at which pyrophyllite absorbs (Farmer 1974). Thedecrease in the layer charge is assumed to cause the collapse of some of the expandable interlayersin the montmorillonite structure, with the subsequent development of pyrophyllite-like (non-expandable) layers. This leads to the formation of mixed-layer pyrophyllite-like montmorillonitecrystals. It has been suggested (Emmerich et al. 2001) that the substantial increase in the non-expandable (non-swelling) layers in the order Li > Na > Ca leads to a decrease in the CEC value.

A more detailed analysis of the various spectra led to the following conclusions:

OH-stretching region: The absorption band at 3627 cm–1, assigned to the stretching vibrations ofoctahedral OH groups, was found to be independent of the interlayer cation species for unheatedhomo-ionic montmorillonite. Also, heating at 600 °C caused similar shifts to higher wavenumbers(ca. 3633 cm–1) for all the samples.

Vibration of molecular water: The vibrations corresponding to adsorbed water at 3426 cm–1 and1642 cm–1 diminished upon heating the samples up to 600 °C, and diminished even further whenthe heating time for the samples was increased to 20 h. However, the broader band at ca.3439 cm–1 probably arises from perturbed OH vibrations caused by the close approach of Ca2+ andNa+ ions to the lattice OH groups. It is possible that the oxygen of the OH group forms acoordinate bond with the cation. The thermal stability of the band in the Ca2+ form results fromthe inability of this large ion to penetrate further into the lattice. In contrast, the disappearance ofthe band in LiB arises when this small ion migrates into the octahedral layer.

OH-bending: No significant differences occur in the regions of the spectra corresponding to thisvibration for any of the forms of bentonite studied after heating up to 600 °C. The band at789 cm–1, corresponding to OH associated with Mg–Mg pairs (Calvet and Prost 1971), apparentlymoves to 798–800 cm–1 after heating the samples. The band at 880 cm–1, assigned to AlFeOHvibrations, disappeared for all the samples. A weak Si–O absorption near 731 cm–1 for all thesamples confirmed the local tri-octahedral character of their structures.

Morphological studies

The calcium montmorillonite sample [Figure 4(a)] exhibited the typical morphology for thismineral, consisting of complex aggregates of grains of a few microns in size. The edges of thesegrains were sharp and rugged. The granular aggregates were, as a rule, not transparent to anelectron beam. The tactoid morphology (face-to-face aggregation) of montmorillonite may berelated to its moderate CEC value and the relative large particle size (Maes et al. 1997). Thismorphology also affects the surface area and porosity of this clay. Heating the sample to 600 oCled to a “loosening” of this morphology, with the complex aggregates exhibiting a wavy leafappearance after heating for 6 h, except for the (001) plane crystal face which was notwell-developed [Figure 4(b)]. The characteristic morphological features of the homo-ionicmontmorillonite structures after heat treatment for 20 h consisted of thin and short leaves, onion-shaped, honeycomb, lath-shaped and dispersed thin flakes [Figures 4(c) and (d)].

Dispersability of the heated-treated fractions

Figures 5 and 6 show the dispersability of the thermally-treated CaB, NaB and LiB samples beforeand after mechanical dispersion. The unheated samples formed stable colloidal dispersions.



(c) (d)

(a) (b)

Figure 4. Scanning electron micrographs of bentonite samples: (a) CaB; (b) LiB (after heat treatment at 600 oC for 6 h);(c) CaB (after heat treatment at 600 oC for 20 h); (d) LiB (after heat treatment at 600 oC for 20 h).

2500

2000

1500

1000

500

00 10 20

Time (min)

Tur

bidi

ty (

NT

Y)

2500

2000

1500

1000

500

0

Tur

bidi

ty (

NT

Y)

30 40 0 10 20

Time (min)

30 40

Time (min)

Tur

bidi

ty (

NT

Y)

1800

1600

1400

1200

1000

800

600

400

200

00 10 20 30 40

(a) (b)

(c)

Figure 5. Spontaneous dispersion of thermally-treated homo-ionic montmorillonite as measured by nephelometry (NTY):(a) CaB, (b) NaB and (c) LiB, respectively. In each part, the data points relate to the following: (�) initial sample, andafter heating at 600 oC for (�) 2 h, (�) 6 h, (�) 12 h and ( ) 20 h, respectively.

(a) (b)

(c) (d)

However, the degree of dispersion of heated-treated CaB, NaB and LiB depended largely on theinterlayer cations as well as on the heat-treatment time. A comparison of the decrease in the degreeof dispersion upon heating CaB, NaB and LiB showed that the three different cations wereincorporated differently into the bentonite structure. The extent to which CaB dispersedspontaneously was higher than that of LiB and NaB. After heating at 600 oC for 2 h, the CEC valueof CaB was reduced by 66%. However, heating for 20 h at the same temperature reduced the valueof the CEC by 78% [see Figure 7(a)]. On the other hand, NaB and LiB almost completely lost theircation-exchange capabilities on increasing the length of thermal treatment up to 20 h. The thermalstability of the CaB results from the inability of this large cation to penetrate further into the lattice(Russell and Farmer 1964). The lower polarising power of Na+ would account for its failure toinfluence the lattice OH groups.

The spontaneous dispersion of all the samples after thermal treatment at 600 oC for 6 h was higherthan that after 2 h (Figure 5), this increase arising from the deformability of the solid clay structure(Boek et al. 1995; Steudel et al. 2009). Hetzel et al. (1994) have confirmed that the arrangement ofthe multi-layer particles and the sizes of the inter-particle pores play a major role in the developmentof swelling strain. However, the attachment of hydroxy groups at the sheet edges as well as theremoval of organic matter (OM) and sesquioxides could also be another reason. The presence of OMand sesquioxides could restrict both the collapse of the expandable clay interlayers on dehydrationand their swelling on hydration; they may also effectively block some interlayer spaces. In addition,


3500

3000

2500

2000

1500

1000

500

00 10 20 30 40

Tur

bidi

ty (

NT

Y)

Tur

bidi

ty (

NT

Y)

Time (min)

0 10 20 30 40

Time (min)

2500

2000

1500

1000

500

0

350030002500200015001000

5000

0 5 10 15 20 25 30 35

Tur

bidi

ty (

NT

Y)

Time (min)

(a) (b)

(c)

Figure 6. Mechanical dispersion of thermally-treated homo-ionic montmorillonite as measured by nephelometry (NTY):(a) CaB, (b) NaB and (c) LiB, respectively. In each part, the data points relate to the following: (�) initial sample, andafter heating at 600 oC for (�) 2 h, (�) 6 h, (�) 12 h and ( ) 20 h, respectively.


the adsorption of water by OM itself may result in a decrease in the effect of the clay structure andan apparent increase in the exchangeable cation effect (Katerina et al. 2004).

During thermal treatment at 600 °C, dehydrated cations with a smaller radius (Li+) begin toleave the interlayer spaces and move into vacancies in the octahedral sheets and/or hexagonalholes in the tetrahedral sheets. However, larger-sized cations (Ca2+ and Na+) are unable to enterthe structure of the octahedral sheets due to space position resistance effects (Emmerich et al.2001) but may be inserted at the bottom of a tetrahedron and form a covalent bond. Thus, the layercharge is apparently reduced as a result of the stronger bonding of these cations in the hexagonalcavities of the tetrahedral sheets. Increasing the thermal treatment time up to 20 h virtuallycompletes the interlayer transfer of positive ions with a smaller radius and the insertion of ionswith a larger radius, thereby causing a loss in the cation-exchange capability. The interlayerdistance in the thermal products is reduced with increasing thermal treatment time, but changes inthe CEC value occur later than those in the interlayer distance. The reason for this may be that theinterlayer distance is sensitive to hydration situations, while the CEC is affected by the completelydehydroxylated status of the interlayer ions.

Following mechanical dispersion, full expandability of CaB and NaB was recorded after thermaltreatment up to 12 h; however, the expandability was reduced slightly by increasing the length ofthermal treatment up to 20 h. CaB regained ca. 81% of its initial CEC value on thermal treatmentup to 12 h, with only ca. 51% being regained after thermal treatment up to 20 h [Figure 7(b)]. Inthe case of NaB, 83% of the initial CEC value was regained after thermal treatment up to 6 h andca. 40% after thermal treatment up to 20 h. However, in the case of LiB, the improvement inexpandability was lower than that for CaB and NaB, especially on increasing the thermal treatmenttime [Figure 7(b)]. Green-Kelly (1955), in one of the earliest papers ever published on Li+ ion

100

CaBNaBLiB

90

80

70

60

50

40

CE

C (

meq

uiv/

100

g)

30

20

10

00 2

Heat-treatment time (h)

6 12 20

CE

C (

meq

uiv/

100

g)

100(a) (b)

90

80

70

60

50

40

30

20

10

00 2 6 12 20

Heat-treatment time (h)

CaBNaBLiB

Figure 7. Cation-exchange capacity (CEC) of initial homo-ionic montmorillonites and of samples subjected to variousheat-treatment times as measured after dispersion for 30 min: (a) spontaneous dispersion; (b) mechanical dispersion.

fixation in montmorillonite, reported that cation migration can take place from the octahedral sitesto the interlayer positions when LiB is heated at 250 °C in 0.1 M NaCl solution. This result helpsto explain the expandability obtained following NH3 vapour treatment and the increasedexchangeable in the Li+ ion content after heat treatment (Farmer and Russell 1967).

Surface area and porosity

Effects of exchangeable cations on surface area and porosity

The shapes of the adsorption isotherms for CaB, NaB and LiB depicted in Figure 8 are types IIand IV in the BDDT classification, with a large uptake being observed close to the saturationpressure where capillary condensation in the large voids between the aluminosilicate sheetscommenced. However, the adsorption isotherm of LiB is of type I at low pressure, indicating amicropore-filling process. All the desorption isotherms exhibit obvious capillary condensation at


0

20

40

60

80

100

120

140

LiBLiB 12 hLiB 20 h L

(c)

Ads

orbe

d vo

lum

e @

ST

P (

cm3 /

g)

0 0.2 0.4 0.6 0.8Relative pressure, P/P0

(a)

0

50

100

150

200

250

0 0.2 0.4 0.6 0.8

CaBCaB 12 hCaB 20 hCaB 2 h LCaB 20 h L

Ads

orbe

d vo

lum

e @

ST

P (

cm3 /

g)

Relative pressure, P/P0

0 0.2 0.4 0.6 0.8Relative pressure, P/P0

0

20

40

60

80

100

120

140

160NaBNaB 20 hNaB 20 h L

(b)

Ads

orbe

d vo

lum

e @

ST

P (

cm3 /

g)

Figure 8. Effects of exchangeable cations, thermal treatment and acid leaching on the nitrogen adsorption/desorptionisotherms of (a) CaB, (b) NaB and (c) LiB, respectively.

an intermediate pressure. As demonstrated in Figure 8, the shapes of the hysteresis loops for allthe samples were similar. The increases or decreases in these hysteresis loops may be mainlyattributed to changes in the crystallinity of the samples, as well as to the homogeneity orheterogeneity of the pores inside the samples (Pérez-Vidal et al. 2006). It will be noted from thefigure that the adsorption and desorption branches of the isotherms exhibited steep slopes over therelative pressure range 0.98–1.00P/P0, suggesting the presence of macropores that were not filledby nitrogen. In addition, the decrease exhibited in the low-pressure region confirms previous data.For LiB, the values of the micropore surface area and micropore volume were both zero, possiblydue to the fact that the pores in this material exhibited sizes on the borderline between microporesand mesopores.

In general, all the samples contained broad and asymmetrical pore-size distributions. Figure 9also depicts the effect of exchangeable cations on the micropore/mesopore distributions of CaB,NaB and LiB. Thus, both CaB and NaB had pores which showed a broad pore-size distribution inthe range 100–370 Å, although narrow mesopores with sizes between 35 Å and 55 Åpredominated. For LiB, a significant decrease in the amount of mesopores of size > 100 Å wasobserved. On the other hand, the number of pores with sizes on the border between microporesand mesopores increased.


0

50

100

150

200

250

300

350

400

(a) (b)

(c)(d)

0 100 200 300 400 500 600 700 800 900

Ads

orbe

d vo

lum

e (c

m3 /

g)

CaBCaB 12 hCaB 20 h

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700 800 900

Por

e vo

lum

e (c

m3 /

g)

NaBNaB 2 hNaB 20 hNaB 12 h L

0

200

400

600

800

1000

1200

1400

0 100 200 300 400 500 600 700 800 900

Ads

orbe

d vo

lum

e (c

m3 /

g)

LiBLiB 12 hLiB 20 hLiB 20 h L

0

50

100

150

200

250

300

350

400

< 2 µm 2 h L< 2 µm 20 h L

0 200 400 600 800 1000

Ads

orbe

d vo

lum

e (c

m3 /

g)

Figure 9. Effects of exchangeable cations, thermal treatment and acid leaching on the micropore/mesopore distribution in(a,b) CaB, (c) NaB and (d) LiB, respectively.

Effect of thermal treatment on the surface area and porosity

Both the open and micropore surface areas, as well as the micropore volumes, exhibited “zig-zag”variations as a function of treatment time at constant temperature (see data listed in Table 1). Thezig-zag change in the values of SBET and the pore volume (Vp) arose from the dehydration anddehydroxylation of homo-ionic bentonite. The isotherm shapes show little change with treatmenttime, although there is an evident reduction in the adsorption capacity. In all cases, irrespective ofthe exchangeable cations, the adsorption/desorption branches of the isotherms remain unchanged(Figure 8).

The pore-size distribution curves of the heated samples depicted in Figure 9 indicate asignificant increase in the amount of pores on the micropore/mesopore border for LiB withheat-treatment time. However, under the same circumstances, no significant difference occurredin this distribution for NaB and CaB. According to Zhu et al. (1997), the effect of heating on the


TABLE 1. Textural Parametersa of Homo-ionic Bentonite Before and After Modification

Sample SBET Sex St % St V0.98 V0.1 Vt Vl Ap (Å)heating time (m2/g) (m2/g) (m2/g) (cm3/g) (cm3/g) (cm3/g) (cm3/g)

CaB0 h 89.89 70.72 19.17 21.32 108.37 22.66 0.008 0.167 74.482 h 80.86 66.61 14.25 17.62 107.03 20.22 0.006 0.165 81.786 h 86.78 71.86 14.91 17.18 112.24 21.76 0.006 0.173 79.91

12 h 80.86 66.61 14.25 17.62 107.03 20.22 0.006 0.165 81.7820 h 98.46 78.38 20.08 20.39 123.71 24.68 0.0085 0.191 77.622 h L 245.90 111.78 134.111 54.53 194.72 67.55 0.06 0.300 48.926

20 h L 161.01 89.82 71.19 44.21 144.95 43.85 0.03 0.223 55.62

NaB0 h 81.12 73.41 7.71 9.50 118.21 21.38 0.003 0.182 90.0282 h 59.12 56.28 2.84 4.80 96.16 14.47 0.001 0.148 100.526 h 66.76 60.60 6.02 9.01 123.61 16.54 0.002 0.190 114.62

12 h 71.67 61.34 10.33 14.41 91.21 17.80 0.004 0.140 78.6220 h 86.79 73.49 13.30 15.32 109.91 21.71 0.005 0.169 78.212 h L 137.04 88.89 48.14 35.12 139.90 36.12 0.02 0.21 63.07

LiB0 h 85.62 90.90 0.00 0.00 121.35 20.32 0.00 0.187 87.562 h 77.04 67.52 9.52 12.35 105.13 19.22 0.004 0.162 84.316 h 76.32 66.41 9.91 12.98 109.86 19.00 0.004 0.169 88.92

12 h 87.53 80.53 6.99 7.98 121.33 21.62 0.002 0.187 85.6420 h 90.07 78.78 11.29 12.53 118.61 22.35 0.004 0.183 81.3512 h L 160.35 85.46 74.88 46.69 129.95 43.43 0.034 0.200 50.07

aSBET = BET surface area; V0.98 = adsorbed volume of N2 at P/P0 = 0.98; V0.1 = adsorbed volume of N2 at P/P0 = 0.1; Sex =external surface area by de Boer t-method; St = open surface area by de Boer t-method; % St = percentage surface areacontributed by micropores; Vt = micropore volume by de Boer t-method; Vp = average pore volume; Ap = average porediameter. Note that the designation “L” relates to leached samples.

textural parameters of homo-ionic bentonite is due both to dehydration and dehydroxylation of theclay sheets and to an initial increase in the homogeneity of the mineral crystallites. Thisinterpretation is in agreement with that reported by Bojemueller et al. (2001) who concluded thatthe desorption of hydration and interlayer water during calcination for a short time led to closureof the interlayer spaces and a denser packing of the particles. However, heating for 20 h causedan increase in the surface area and pore volume for all the samples due to dehydroxylation of theedges of some layers.

Effect of acid leaching following thermal treatment

The effect of acid leaching following thermal treatment was studied using selected thermally-treatedsamples and determining their nitrogen adsorption/desorption isotherms (Figure 8). The isothermsfor CaB and NaB were not significantly influenced by acid leaching following thermal treatment.However, the shape of that for LiB changed considerably after acid leaching. In this case, theadsorption/desorption isotherms were nearly horizontal over the relative pressure range0.98–1.00P/P0, suggesting that virtually all the pores were filled with nitrogen and that LiBcontained no macropores [see Figure 8(c)]. The enlargement of the hysteresis loop upon acidleaching indicates a slight change in the average pore width. This enlargement in the pore width isexpected to be due to destruction of the original structure as a result of acid attack.

The effect of acid leaching following thermal treatment on the micropore/mesopore distributionof selected samples is shown in Figure 9. For LiB, a significant decrease in the amount ofmesopores larger than 100 Å was observed. Simultaneously, however, the amount of pores in thesize range 35–50 Å increased [see Figure 9(c)]. In the case of CaB, acid leaching followingthermal treatment for 2 h led to a significant increase in amount of pores in the size range 13–38 Å,with a corresponding significant increase in surface area. Most of this increase arose from thecontribution of micropores, i.e. the open surface area of CaB increased significantly. However,acid leaching following thermal treatment for a long time (20 h) caused a decrease in the opensurface area, the micropore volume and the total pore volume, irrespective of the exchangeablecation involved. From the data listed in Table 1, it is clear that acid leaching of samples calcinedfor 2–6 h caused abrupt increases in the surface area (up to 245.90 m2/g), the micropore volume(up to 0.06 cm3/g), the average pore volume (up to 0.30 cm3/g) and the reduced average pore size(up to 4.8 nm). However, acid leaching of samples calcined for 12–20 h led to a reduction in thesurface area (down to 137.04–160.35 m2/g), the micropore volume (down to 0.02–0.034 cm3/g),the average pore volume (down to 0.20–0.22 cm3/g) and the average pore size (down to 5.0–6.3 nm),respectively.

CONCLUSIONS

Exchangeable cations, heat treatment and acid leaching of homo-ionic montmorillonites led toconsiderable variations in their structural and textural properties. Thus, decreases in the layercharge and local structural modification of the tetrahedral and octahedral sheets due to theintroduction of Ca2+, Na+ and Li+ cations also influenced the fundamental IR vibrations of theSi–O and OH groups.

The degree of dispersion of heated homo-ionic montmorillonite was found to depend largely onthe interlayer cations as well as on the heating time. The high dispersability of Cab resulted fromthe inability of the large Ca2+ cation to penetrate the lattice to any great effect.


Ion exchange led to the simultaneous widening of the pores and an increase in both the porevolume and connectivity due to the insertion of Li+ and Na+ cations between the clay sheets duringthe exchange process. However, the introduction of the Li+ cation created more pores in the borderregion between micropores and mesopores.

The effect of heating on the textural parameters of homo-ionic montmorillonite arose fromdehydration and dehydroxylation of the clay sheets and to an initial increase in the homogeneityof the mineral crystallites. Similarly, the considerable changes in the textural parameters observedduring acid leaching following heat treatment were caused by the removal of cations from theoctahedral holes in the layer sheets.

ACKNOWLEDGMENTS

We are grateful to the Major Analytical Instrumentation Center (MAIC) and the ParticleEngineering Research Center (PERC) and the Department of Materials Science at the Universityof Florida for the use of their equipment during this investigation. Partial financial support fromthe PERC and the industrial partners of PERC, NSF Grant No. INT0115417 and NSF GrantNo. 0211253 are gratefully acknowledged.

REFERENCES

Andrejkovicová, S., Madejová, J., Czímerová, A., Galko, I., Dohrmann, R. and Komadel, P. (2006) Geol.Zborn. Geol. Carpath. 57, 371.

Barrett, E.P., Joyner, L.G. and Halenda, P.H. (1951) J. Am. Chem. Soc. 73, 373.Boek, E.S., Convey, P.V. and Skipper, N.T. (1995) Langmuir 11, 4629.Bojemueller, E., Nennemann, A. and Lagaly, G. (2001) Appl. Clay Sci. 18, 277.Calvet, R. and Prost, R. (1971) Clays Clay Miner. 19, 175.de Boer, J.H. (1958) in The Structure and Properties of Porous Materials, Everett, D.H., Stone, F.S., Eds,

Academic Press, New York/London.Duarte, M.A.I., Maccedo, J.C.D. and Alves, D.B. (1995) “Generation of Pores in Metakaolinite by Acid

Dealumination”, 10th Int. Clays Conf., Adelaide, Australia, July 18–23.Emmerich, K., Michael, P. and Günter, K. (2001) Appl. Clay Sci. 19, 143.Farmer, V.C. (1974) “Layer Silicates”, in Infrared Spectra of Minerals, Farmer, V.C., Ed, Mineralogical

Society of London, pp. 331–363. Farmer, V.C. and Russell, J.D. (1967) “Infrared Absorption Spectrometry in Clay Studies”, in Proc. 15th Nat.

Conf. on Clays and Clay Minerals, Pittsburgh, PA, U.S.A., pp. 121–147.Gates, W.P., Komadel, P., Madejová, J., Bujdák, J., Stucki, J.W.E. and Kirkpatrick, R.J. (2000) Appl. Clay

Sci. 16, 257.Green-Kelly, R. (1955) Mineral. Mag. 30, 604.Hetzel, F., Tessier, D., Jaunet, A.M. and Doner, H. (1994) Clay Miner. 42, 242.Inglethorpe, S.D.J., Morgan, D.J., Highley, D.E. and Bloodworth, A.J. (1993) Industrial Minerals Laboratory

Manual “Bentonite”, British Geological Survey Technical Report WG/93/20, Mineralogy and PetrologySeries, Natural Environment Research Council, London.

Katerina, M.D., Darrell, L.N., Cliff, T.J. and Jerry, M.B. (2004) Soil Sci. Soc. Am. J. 68, 1218.Madejová, J., Pálková, H. and Komadel, P. (2002b) Clay Miner. 37, 377.Madejová, J., Pálková, H. and Komadel, P. (2006) Vib. Spectrosc. 40, 80.Maes, N., Heylen, I., Cool, P. and Vansant, E.F. (1997) Appl. Clay Sci. 12, 43.Mehra, O.P. and Jackson, M.I. (1960) Clays Clay Miner. 7, 317.


Neaman, A., Manuel, P. and Frédéric, V. (2003) Appl. Clay Sci. 22, 153.Pérez-Vidal, H., Custodio-García, E., López-Alejandro, E., Morales-Hidalgo, J. and Frías-Márquez, D.M.

(2006) Sol. Energy Mater. Sol. Cells 90, 841.Russell, J.D. and Farmer, V.C. (1964) Clay Miner. Bull. 5, 443.Steudel, A., Batenburg, L.F., Fischer, H.R., Weidler, P.G. and Emmerich, K. (2009) Appl. Clay Sci. 44, 105.Venaruzzo, J., Volzone, C., Rueda, M.L. and Ortiga, J. (2002) Microporous Mesoporous Mater. 56, 73.Volzone, C. (2007) Appl. Clay Sci. 36, 191.Volzone, C. and Ortiga, J. (2009) Appl. Clay Sci. 44, 251.Wang, M.C., Benway, J.M. and Arayssi, A.M. (1990) ASTM STP 1095, ASTM, Philadelphia, PA, U.S.A.,

pp. 1139–1158.Wu, P., Wub, H. and Li, R. (2005) Spectrochim. Acta A 61, 3020.Zhu, H.Y., Vansant, E.F., Xia, J.A. and Lu, G.Q. (1997) J. Porous Mater. 4, 17.


Copyright of Adsorption Science & Technology is the property of Multi-Science Publishing Co Ltd and its

content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.

Documents

Textural and Structural Modification of Homo-ionic Montmorillonite