JBCS

INTRODUCTION

Organic Pillared clays, also known as organoclays or nanoclays, are obtained by cation exchange of interlayer cations, component of 2:1 type clay minerals, with quaternary ammonium cations. This process leads to the formation of materials with hydrophobic characteristics that can be applied in a variety of processes, such as catalytic processes, rheological control agents in paintings and lubricants; polymer and plastics matrix reinforcing; adsorbents for effluent treatments, oil spilling, releasing active matrix, etc.^1-6

The application of these materials depends strongly on the obtained structure, which, at the same time, conditions its properties. Therefore, understanding the relation between both the method and synthesis parameters used, and the developed structure, is of the utmost importance.

Among variables that affect the synthesis process, the properties of the employed clay must be considered: chemical composition, cation exchange capacity (CEC), charge density, specific surface area, etc.; and also those related to the synthesis process itself: kind and concentration of surfactant, time of contact, temperature, etc.

Some authors have demonstrated that the structure and properties developed by these materials are correlated directly with the organization of organic cations in the interlayer space,³ showing the importance of determining the type of arrangement developed in organoclays.

The purpose of this work was to synthesize and characterize organoclays from an Argentinian montmorillonite (Bent) with hexadecyltrimethylammonium bromide (HDTMA-Br) as intercalation agent. The influence of the amount of intercalated hexadecyltrimethylammonium cation (HDTMA⁺) characterizing the obtained products was evaluated. Their structures were correlated with models of distribution and ordering of HDTMA⁺ molecules intercalated in the intra and extra layer space.

EXPERIMENTAL

Materials

The used clay was an Argentinian bentonite sample (Bent) from Rio Negro Province, with mineralogy: 84% Na-montmorillonite, 4% quartz and 12% feldspars, determined by Rietveld method;⁷ the CEC was 0.837 mol kg^-1, determined by the Cu(EDA)₂²⁺ method.⁸ HDTMA-Br was supplied by Merck, (99%), used without further purification. All solutions and dispersions were prepared using deionized water obtained with SQC 3 Reverse Osmosis, from Water Factory System^® equipment.

Samples preparation

Organoclays were obtained by putting in contact a 5% w/v suspension of starting clay with a determined volume of HDTMA-Br 5 × 10^-3 mol L^-1 solution for 2 h with orbital shaking at room temperature."Surfactant loading" (SL) is defined as the x fraction of the CEC that is replaced by HDTMA⁺:

SL = (mol HDTMA⁺ / mol exchange cations) CEC = x CEC

To obtain different SL, different volumes of HDTMA-Br 5 × 10^-3 mol L^-1 solution were used. Solid products were isolated by centrifugation and repeatedly washed with distilled water until they were free of bromide ions (determined by AgNO₃ test). Furthermore, the solids were dried at 60 ºC over night, milled in agate mortar, kept under conditions of controlled humidity, and denoted as Bent followed by the SL (Bent-0.4; Bent-0.8, etc.).

Desorption of the surfactant was analyzed through leaching tests. A suspension of each organoclay was prepared in distilled water and then shaken for 0.5 h on an orbital shaker at room temperature, after which it was separated by centrifugation. The leaching process was repeated "n" times. Solids obtained in this way are indicated with the name of the initial organoclay followed by "Ln" (Bent-0.4-L3, Bent-0.8-L7, etc).

Materials characterization

Differential thermal analysis (DTA) was performed with a Rigaku TAS 1100 from room temperature to 1200 ºC, using about 20 mg samples, with a heat temperature rate of 20 ºC min^-1 in static air atmosphere.

X-ray diffraction (XRD) studies were obtained on oriented samples by spreading the sample suspension on glass slides and further drying (48 h) at room temperature with a relative humidity of 0.47. Analyses were performed using a Philips PW 1710 diffractometer using CuKα radiation (0.154 nm). Measuring conditions were: a power supply of 40 kV and 30 mA; 1º (2θ) divergence and detector slits, 0.02º (2θ) step size, counting time of 10 s step^-1 and patterns collected from 3º to 70º (2θ). In cases where overlapping peaks were detected, mathematical deconvolutions were performed with OriginPro 8 software. The positions and peak areas were calculated using the Gaussian functions, applying a fitting algorithm for nonlinear least squares.

KBr pressed discs of dried montmorillonite and organoclays with a sample to KBr relationship of 1:100 approximately, ground in agate mortar and pressed at a pressure of 3 ton, were analyzed by FTIR in a Spectrum GX Perkin Elmer Infrared spectrometer, between 4000 and 400 cm^-1.

N₂ adsorption isotherms were obtained on a Micromeritics sorptometer ASAP 2020 V3.03 E, at -196 ºC, previously outgassing solids at 100 ºC for 10 h (or until a pressure lower than 8 × 10^-6 mmHg was reached). BET surface was calculated with at least 5 points with a minimum linear correlation coefficient of 0.999. Micropore Volume (V_µP) was calculated with the t-plot method using Harkins-Jura-Boer equation; Total Pore Volume (V_TP) by Gurvitch method, interpolating at p/p₀ = 0.98; and the Mesopore Volume (V_mP) by the difference between V_TP and V_µP. Pore size distribution was conducted using Halsey equation over desorption branch.

Using Mopac 2009 software, a computational simulation of the surfactant molecule in all-trans conformation was performed. A semi-empirical method with a PM6 base was used for its optimization. In this way, molecular dimensions and surface area were calculated. Thickness values for the multilayer structures were also calculated through techniques of computational calculation.

RESULTS AND DISCUSSION

TGA

Thermal analysis of Bent samples showed the two typical regions associated with water-loss in smectites (Figure. 1).⁹ The water adsorbed by external surfaces and hydration water of the interlayer cations were removed between 80 and 150 ºC, while structural water was released at temperatures higher than 500 ºC. Thermal treatment of organoclays samples produced further mass loss at temperatures of 180 to 780 ºC,^10,11 corresponding to organic matter oxidation and charcoal formation.¹² Temperature ranges were directly related with the SL of each sample.

Figure 1. TGA (A) and DTA (B) of indicated samples. Temperature range limits are indicated with vertical lines. Note different scales in y axis of figure 1-B

To determine the mass loss scheme in each sample, Δm_i,α was defined as the mass loss produced in the α temperature range for the organoclays with an i surfactant load.

Range 1: Room Temperature - 180 ºC. Within this range of temperatures, physisorbed and interlayer water and/or of water of hydration of exchangeable cations (non structural water) were eliminated,¹³ both in the starting clay and in the organoclays. This mass loss was associated with the presence of endothermic peaks in the DTA curve.

Range 2: 180 ºC - 800 ºC. In this range, a mass loss of 3% was observed in Bent clay that corresponds to dehydroxylation of clay layers.^10,14,15

For organoclays, three mass losses in this range of temperatures were observed, associated with several (from 3 to 5) exothermic events (Figure 1). Initial and final temperatures of each one of these mass losses varied with the surfactant loading, but fell among the following subdivisions of range 2 (Table 1).

Range 2-a: Mass loss was associated with a very important exothermic event (Figure 1-B); its position depends on the SL. The amount of mass eliminated in this stage increased with the SL up to 2.0 CEC (Δm_0.4,_2-a = 2.5%, Δm_0.8,_2-a = 4.0% and Δm_2.0,_2-a = 22.5%) and then kept constant (Δm_3.0,_2-a = 22.0%).

Range 2-b: The second stage of mass loss was associated with various exothermic events, whose positions depended on the SL (Figure 1). The amount of mass eliminated in this stage increased with surfactant loading up to 0.8 CEC (Δm_0.4,_2-b = 4.5%, Δm_0.8,_2-b = 8.2%,) and kept constant up to 2.0 CEC, whereas for higher SL mass, loss decreased slightly (Δm_2.0,_2-b = 8.5% and Δm_3.0,_2-b = 7.5%). This slight decrease could be attributed to the fact that as SL increased, a higher amount of heat was released by combustion of organic groups in the 2-a range, which provided enough energy for the combustion of molecules that should be eliminated in range 2-b.

Range 2-c: A mass loss associated with a weak exothermic peak was observed (Figure 1). In this stage, mass loss due to dehydroxylation of layers and the third stage of mass loss associated with surfactant combustion were overlapped.

To calculate the mass of organic compounds eliminated in this stage, the fact that mass corresponding to dehydroxylation of layers represents a 3% of Bent clay mass was taken into account.

The amount of organic compounds eliminated in this stage follows the same trend that the one observed in Range 2-b. The proportion of mass loss at this range of temperatures corresponding only to combustion of organic compounds (after subtracting the contribution of the dehydroxylation of layers) was 3.6%, 6.6%, 6.9% and 6.8% for Bent-0.4, Bent-0.8, Bent-2.0 and Bent-3.0 respectively.

Range 3: 800 ºC - 1200 ºC. For Bent sample, an endothermic maximum at 867 ºC and two exothermic maximums at 906 and 1145 ºC were found, all of them without any associated mass loss (Figure 1). These peaks were attributed to structural rearrangements and/or formation of new phases such as spinel, cristobalite or mullite.¹⁰

To perform a deeper analysis of the processes of mass loss observed at the different temperature ranges, the relation R_i(α/β)=(Δm_i,α)/( Δm_i,β) was defined, where i and Δm_i,α have the meaning indicated in section 3-1, and β is a range of temperatures different from α.

R_i(2-a/2-b) and R_i(2-a/2-c), increased with the surfactant loading, whereas R_i(2-b/2-c) kept practically constant (Figure 2). The slight decrease of R_i(2-b/2-c) with SL was due to the strong release of heat that took place in stage 2-a, which was discussed before.

Figure 2. Ri(α/β) (equation 4) in function of surfactant loading. (О): Ri(2-a/2-b), (Δ): Ri(2-a/2-c), (□): Ri(2-b/2-c)

Thermograms of leached organoclays were also obtained. In general, they showed a similar behavior to starting organoclays. For surfactant loadings up to 0.8 CEC, values of Δm_2-a, Δm_2-b and Δm_2-c were not modified with leaching treatments. For superior SL, the Δm_2-a decreased with the lecheates, whereas Δm_2-b and Δm_2-c kept constant (Figure 3).

Figure 3. Washing effect from a Bent-2.0 sample. Squares indicate mass loss in range 2-a, circles in range 2-b and triangles in range 2-c

Taking into account Figures 2 and 3, Δm_2-b and Δm_2-c showed the same behavior. To facilitate the discussion of the results, stages 2-b and 2-c will be analyzed together, and be named as stage 2-(b+c).

The surfactant mass that was eliminated in stage 2-a decreased with lecheates and led to the conclusion that the Δm_2-a corresponds to molecules of surfactant weakly attached to the clay. On the other hand, in stage 2-(b+c), the loss of surfactant molecules more strongly attached to the surface was produced, as Δm_2-(b+c) did not decrease with lecheates.

XRD studies

X-ray diffraction patterns (Figure 4) of different organoclays showed values of the X-Ray diffraction peak corresponding to d(₀₀₁) distance that increased from 12.1 Å for the starting clay up to 19.2 Å for the organoclay with higher SL (3.0 CEC). This increase (Table 2) indicated that at least a fraction of the cationic surfactant has replaced the hydrated interlayer cations.⁷

Figure 4. X-ray diffraction pattern of indicated samples: Bent-0.4 (a), Bent-0.8 (b), Bent-2.0 (c) and Bent-3.0 (d)

In Bent-0.8 the X-Ray diffraction peak d(₀₀₁) was actually composed by the overlapping of more than one component. In Bent-2.0 and in Bent-3.0 the presence of other X-Ray diffraction peaks of lower intensity was also observed. This behavior is an indicator of interstratification. The area of those peaks is related to the amount of layers with a determined value of d(₀₀₁).

In table 2, the positions of d(₀₀₁) X-Ray diffraction peaks obtained by deconvolution of XRD pattern are shown. From these distances, the basal spacing was calculated, taking into account that the thickness of the layer was 9.5 Å.¹⁶ In the cases in which there are more than one d(₀₀₁) line, the percentages of the area of each one of them in respect to the total area were calculated.

To determine the type of structure developed by the surfactant intercalated within clay layers, the molecule of HDTMA⁺ was modelled in an all-trans conformation.

Considering both the computational models and the interlaminar distances experimentally obtained, different structures that the surfactant could present in the interlamelar space were proposed (Table 2).

N₂ adsorption

For expandable clays, N₂ adsorption studies give a sub-evaluation of the specific surface area value because they only provide information about the external surface.^7,17 They also provide information about the meso and macropores that could be generated by the ordering of tactoids in the structures of the "house of cards" type¹⁸ or, in the case of mixture of clays with other materials of bigger grain size, by the spatial ordering of particles.¹⁹

The N₂ adsorption isotherm of Bent sample presented a hysteresis loop type B, indicating mesopores presence (Figure 5).²⁰ This hysteresis loop decreased with the SL until it disappeared for Bent-2.0 and Bent-3.0 samples. Pore size distribution (Figure 6-A) displayed a maximum of pore size of 20.6 Å for Bent sample, increasing up to 25.1 Å for Bent-0.4 and Bent-0.8 samples. A maximum in the curve of pore size distribution for Bent-2.0 and Bent-3.0 samples was not observed. The value of BET surface area decreased from Bent (56 m² g^-1) to almost 10 m² g^-1 for clays with higher SL (Table 3).

Figure 5. N₂ adsorption-desorption isotherms. Symbols indicate: (□) Bent; (О) Bent-0.4; (Δ) Bent-0.8; (◊) Bent-2.0 and () Bent-3.0 samples

Figure 6. Pore size distribution (A) and Fractal dimension (B). Symbols indicate: (□) Bent; (О) Bent-0.4; (Δ) Bent-0.8; (◊) Bent-2.0 and () Bent-3.0 samples

Total pore volume (V_TP) increased with the SL from Bent to Bent-0.8, and then decreased for organoclays with higher SL (Table 3). Studies of scanning electron microscopy (not shown) have shown that the starting clay presents a rather compact structure with a great fraction of the layers presenting a face to face ordering. On the other hand, organoclays present a more opened structure called "corn flakes". This change in the structure generated the biggest porosity detected in organoclays. When the SL exceeded a certain value, the excess of surfactant occupied mesopores or blocked their entries, and, in this way, porosity decreased.

XRD results indicated that the interlayer space was never higher than 10 Å, so that observed mesopores (within 20.6 and 25.1 Å) corresponded to the pores generated by the "house of cards" ordering type of tactoids. This indicate that, apart from the amount of HDTMA⁺ adsorbed in the interlayer space, there was another fraction of surfactant adsorbed in the external surface, occupying and/or blocking the mesopores.

To examine if an effect of surface coverage was present, fractal dimensions (D) were calculated using equations developed by Frenkel, Halsey and Hill (FHH Theory).²¹ D can be considered as an operational measure of surface roughness. Generally, D is placed between 2 (soft and regular surface) and 3 (surface extremely irregular).

In this work, as the starting solid (Bent) presents hysteresis, the equation used for calculating the fractal dimension was D = 3 - 1/m, where (1/m) was the absolute value of the slope obtained in the Ln(Ln(P₀/P)) vs Ln(V_ads/V_M) graph (Figure 6-B).

Starting clay presented a high value for the fractal dimension (D_Bent = 2.71), similar to the value obtained by Wang for Saz-1 clay (D_Wang = 2.74),²¹ indicating a very irregular surface. The D value of organoclays decreased as surfactant loading increased. This decrease was produced in two stages: the first stage for Bent-0.4 and Bent-0.8 (D_Bent-0.4 = 2.51, D_Bent-0.8 = 2.50), that depends on the elimination of micropores as well as on the "softness" of the surface by the surfactant and the second stage for Bent-2.0 and Bent-3.0 (D_Bent-2.0 = 2.46, D_Bent-3.0 = 2.45) due to the increase in the amount of molecules of surfactant in the organoclay and the blocking of mesopores.

As the fractal dimension decreased and Total Pore Volume increased together with the SL for Bent-4.0 and Bent-0.8 solids, the un-intercalated surfactant covered the external surface as well as the internal surfaces of the walls of pores formed by tactoids grouping in "house of cards" type structures.

Consequently, from nitrogen adsorption isotherms it could be concluded that cationic surfactant, besides intercalating within clay layers, was adsorbed over the external surface and over the internal walls of mesopores, coating and decreasing surface roughness. It also fills or blocks mesopores, decreasing hysteresis and making it disappear from organoclays with higher SL.

FTIR

IR spectra of samples (Figure 7) showed that the absorption bands corresponding to the starting clay were also seen in organoclays, approximately in the same positions.

Figure 7. IR spectra of indicated samples. Bent (a), Bent-0.4 (b), Bent-0.8 (c), Bent-2.0 (d), Bent-3.0 (e) and HDTMA-Br (f)

Symmetric bending bands, rocking bands, and symmetric and asymmetric stretching of methylene (ν_s(CH₂) and ν_as(CH₂) respectively) were observed in all organoclays. ν(N-H) peak starts to be seen from a SL of 2.0 CEC, as corresponds to free HDMTA⁺ molecules.²² This result agrees with the results obtained in N₂ adsorption studies, indicating that for Bent-2.0 and Bent-3.0 organoclays, part of the surfactant is adsorbed out of the interlayer space, filling or blocking mesopores.

The more intense bands in the spectrum of surfactant correspond to symmetric (ν_s (CH₂)) and asymmetric (ν_as (CH₂)) stretching of methylene groups of the carbon chain of surfactant. As SL increases, the intensity of the bands increases and the frequency decreases, according to literature.^21,23-27

Modelling the developed structures

Molecular modelling calculation showed that the wider surface that a HDTMA⁺ molecule can cover is of approximately 100 Ų. Thermogravimetry indicates the amount of molecules that were absorbed for each SL, and from DRX results the structure of the interlayer space of each organoclay was obtained. Using these data, the surface that can cover those surfactant molecules (S_HDTMA+) could be calculated in the following way:

where S_HDTMA+ is the specified surface (in m² g^-1) that n moles of adsorbed surfactant (n_HDTMA+) can cover, taking into account the surfactant structure in the interlayer space (monolayer, bilayer, etc); N_A is the Avogadro number; s_HDTMA+ is the surfactant molecule area (Ų); 10^-20 is the conversion factor between m² and A²; M_clay is the clay mass (g), and f is a factor that takes into account the amount of surfactant layers present in the interlayer space of the clay. The f factor is obtained knowing that, for example, if a bilayer is formed, the surface covered by the surfactant will be equivalent to the clay surface, as the surface of the inferior layer and the superior layer will both be considered, and therefore f_bilayer = 1. In the same way, if a monolayer is formed the factor must be f_monolayer = 2, and for a trilayer f_trilayer = 2/3. For the intermediate structures, factors are calculated as the average of the structures that make them up. So, f_ps-bilayer = (2+1)/2 = 1.5, and f_ps-trilayer = (1+2/3)/2 = 5/6.

In the case of Bent-0.4 organoclays, the structure that was formed is a monolayer and the factor that was applied is f_monolayer = 2. For the other cases, the structures were interstratified, so that the factor that was applied is a weighted average, taking into account the percentages of the different structures present in each organoclay. For Bent-2.0 and Bent-3.0 cases the percentage of collapsed layers were also taken into account (Table 2).

Equation (1) can be used with the total mass of absorbed organic compounds or with the mass of any of the fractions determined by thermogravimetry. In each case, the surface that the surfactant molecules of the employed range can cover will be obtained. The results obtained employing the total amount of the adsorbed organic compounds are shown as S_{HDTMA+, Total}, and the amount absorbed in range 2-(b+c) as S_{HDTMA+, 2-(b+c)} (Table 4). The percentage that this surface represents in respect to the total surface of starting clay (621 m² g^-1)¹⁰ is also indicated.

For SL higher than 0.8, S_{HDTMA+, Total} exceeded 100% of covered surface, which corroborated the fact that a fraction of surfactant is adsorbed out of the interlayer surface in the mesopores generated by tactoids ordering, causing the decrease of the pore volume and the elimination of the hysteresis loop that was observed in the N₂ adsorption isotherm. On the other hand, S_{HDTMA+, 2-(b+c)} never exceeded 100%, achieving 99% of surface covering for a SL of 0.8. The result confirmed that the surfactant corresponding to stage 2-(b+c) was adsorbed between layers. S_{HDTMA+, 2-(b+c)} decreased for the organoclays with higher SL (2.0 and 3.0) which could be due to the excessive release of heat that was generated by the combustion of surfactant in range 2-a, as was explained before.

CONCLUSIONS

Obtained organoclays resulted to be more hydrophobic than the starting clay.

Surfactant, both in molecular and cationic form, was adsorbed in the Bent clay in at least two kinds of different sites, indicated by the different thermal stabilities.

From characterization results and molecular modelling calculation, a mechanism of adsorption is proposed. The surfactant was adsorbed initially, in his cationic form, in sites placed in the interlayer space of the clay through cation exchange. Adsorption in these kind of sites continued until the interlayer space was saturated. Depending on the quantity introduced in the intercalation process, different organizations of surfactant in the interlayer, both in molecular and cationic form, were obtained, varying from monolayer in Bent-0.4 to Bilayer/Pseudo-trilayer in Bent-3.0.

Further adsorption of surfactant, principally in molecular form, occurred in the mesopores generated by tactoids ordered in the "house of cards" type structure. This process left surfactant molecules relatively free, out of the interlayer space. This fact could be favourable or unfavourable, and must be analyzed for each potential application in particular (catalytic processes, rheological control agents in paintings and lubricants; polymer and plastics matrix reinforcing; adsorbents for effluent treatments, oil spilling, releasing active matrix, etc.).

ACKNOWLEDGEMENTS

The funding for this work was granted by "Consejo de Investigación de la UNSa" (Project Nº 1632) and SECyT FONCyT-ANCyP (Project 1360). Authors acknowledge Lic. L. Davies, Ing. S. Locatelli, Dra. D. Acosta and Ing. J. Villarroel Rocha for their technical assistance and clarifying discussions. Pablo Naranjo thanks CONICET and Chubut Province for their fellowships.

REFERENCES

1. Azejjel, H.; del Hoyo, C.; Draoui, K.; Rodríguez-Cruz, M. S.; Sánchez-Martín, M. J.; Desalination 2009, 249, 1151. DOI: http://dx.doi.org/10.1016/j.desal.2009.02.066

2. Carrado, K. A.; Appl. Clay Sci. 2000, 17, 1. DOI: http://dx.doi.org/10.1016/S0169-1317(00)00005-3

3. de Paiva, L. B.; Morales, A. R.; Valenzuela Díaz, F. R.; Appl. Clay Sci. 2008, 42, 8. DOI: http://dx.doi.org/10.1016/j.clay.2008.02.006

4. Delbem, M. A.; Valera, T. S.; Valenzuela-Díaz, F. R.; Demarquette, N. R.; Quim. Nova 2010, 33, 309. DOI: http://dx.doi.org/10.1590/S0100-40422010000200015

5. Cavalcanti, J. V. F. L.; de Abreu, C. A. M.; Sobrinho, M. A. M.; Baraúna, O. S.; Portella, L. A. P.; Quim. Nova 2009, 32, 2051. DOI: http://dx.doi.org/10.1590/S0100-40422009000800013

6. Teixeira-Neto, E.; Teixeira-Neto, A. A.; Quim. Nova 2009, 32, 809. DOI: http://dx.doi.org/10.1590/S0100-40422009000300023

7. Reid-Soukup, U.; SSSA Book Series: 7 Soil Mineralogy with Environmental Applications, Soil Science Society of America, Inc.: Madison, 2002.

8. Bergaya, F.; Vayer, M.; Appl Clay Sci. 1997, 12, 275. DOI: http://dx.doi.org/10.1016/S0169-1317(97)00012-4

9. Hedley, C. B.; Yuan, G.; and Theng, B. K. G.; Appl. Clay Sci. 2007, 35, 180. DOI: http://dx.doi.org/10.1016/j.clay.2006.09.005

10. Xi, Y.; Frost, R. L.; He, H.; J. Colloid Interface Sci. 2007, 305, 150. DOI: http://dx.doi.org/10.1016/j.jcis.2006.09.033 PMID: 17045287

11. Yariv, S.; Lapides, I.; J. Therm. Anal. Calorim. 2005, 80, 11. DOI: http://dx.doi.org/10.1007/s10973-005-0608-7

12. Mackenzie, R. C.; Differential Thermal Analysis, Academic Press: London, 1970.

13. Moronta, A.; Solano, R.; Ferrer, V.; Sánchez, J.; Choren, E.; Ciencia 2003, 11, 130.

14. Xie, W.; Gao, Z.; Pan, W. P.; Hunter, D.; Singh, A.; Vaia, R.; 979.

15. Magnoli, A. P.; Tallone, L.; Rosa, C. A. R.; Dalcero, A. M.; Chiacchiera, S. M.; Torres Sanchez, R. M.; Appl. Clay Sci. 2008, 40, 63. DOI: http://dx.doi.org/10.1016/j.clay.2007.07.007

16. He, H.; Frost, R. L.; Bostrom, T.; Yuan, P.; Duong, L.; Yang, D.; Xi, Y.; Kloprogge, J. T.; Appl. Clay Sci. 2006, 31, 262. DOI: http://dx.doi.org/10.1016/j.clay.2005.10.011

17. Michot, L. J.; Villieras, F. In Handbook of Clay Science; Bergaya, F.; Theng, B. K.; Lagaly, G., eds.; Elsevier: Amsterdam, 2006, chap. 12.9.

18. Tessier, D.; Doctoral Thesis, Institut National de la Recherche Agronomique, France, 1984.

19. Přikryl, R.; Weishauptová, Z.; Appl. Clay Sci. 2010, 47, 163. DOI: http://dx.doi.org/10.1016/j.clay.2009.10.005

20. Gregg, S. J.; Sing, K. S. W.; Adsorption, Surface Area and Porosity, Academic Press: London, 1982.

21. Wang, C. C.; Juang, L. C.; Hsu, T. C.; Lee, C. K.; Lee, J. F.; Huang, F. C.; J. Colloid Interface Sci. 2004, 273, 80. DOI: http://dx.doi.org/10.1016/j.jcis.2003.12.028 PMID: 15051435

22. Wang, C. C.; Juang, L. C.; Lee, C. K.; Hsu, T. C.; Lee, J. F.; Chao, H. P.; J. Colloid Interface Sci. 2004, 280, 27. DOI: http://dx.doi.org/10.1016/j.jcis.2004.07.009 PMID: 15476770

23. Mandalia, T.; Bergaya, F.; J. Phys. Chem. Solids 2006, 67, 836. DOI: http://dx.doi.org/10.1016/j.jpcs.2005.12.007

24. Patel, H. A.; Somani, R. S.; Bajaj, H. C.; Jasra, R. V.; Appl. Clay Sci. 2007, 35, 194. DOI: http://dx.doi.org/10.1016/j.clay.2006.09.012

25. Praus, P.; Turicová, M.; Studentová, S.; Ritz, M.; J. Colloid Interface Sci. 2006, 304, 29. DOI: http://dx.doi.org/10.1016/j.jcis.2006.08.038 PMID: 17007866

26. Zhu, J.; He, H.; Zhu, L.; Wen, X.; Deng, F.; J. Colloid Interface Sci. 2005, 286, 239. DOI: http://dx.doi.org/10.1016/j.jcis.2004.12.048 PMID: 15848422

27. Zhu, R.; Zhu, L.; and Xu, L.; Colloids Surf., A 2007, 294, 221. DOI: http://dx.doi.org/10.1016/j.colsurfa.2006.08.016