Acid–base properties of pillared interlayered clays with single and mixed Zr–Al oxide pillars prepared from Tunisian-interstratified illite–smectite

3.2.1 Mass titration curves

The mass titration analysis was performed at three NaCl concentrations 10⁻³, 10⁻² and 10⁻¹ mol L⁻¹. Figs. 1 and 2 show the evolution of pH values according to the addition of solid samples Na-G, Al-G, Zr₁₀Al₉₀-G, Zr₅₀Al₅₀-G, Zr₉₀Al₁₀-G and Zr-G respectively. The pH gradually changes with the addition of solid minerals and asymptotically approaches a limiting value. The direction of pH variation depends on the pH of the starting NaCl solutions. Therefore, the pH where solid addition does not produce any change in the pH of the initial NaCl solution can be estimated by interpolation. The PZCs estimated in this way are marked with arrows in the Figures. These values are reported in Table 3 for all ionic strengths. The results show that PZCs are not influenced by NaCl concentration. In the case of purified clay Na-G the pH_PZC is about 7.3 for the three ionic strengths. This value is in agreement with the PZC of smectitic clay ranging between 6 and 8.1 according to the experimental conditions selected by the authors (Helmy et al., 1994; Wanner et al., 1994; Avena and De Pauli, 1998; Zhuang and Gui-Rui, 2002; Ganor et al., 2003; Kriaa et al., 2007).

Close

Figure 1

Mass titration curves for Na-G bentonite before and after pillaring with single Al and Zr precursor for each ionic strength.

Export to PPT

Close

Figure 2

Mass titration curves for Na-G bentonite before and after pillaring with mixed Al-Zr precursor for each ionic strength.

Export to PPT

For pillared clays the PZC are shifted to acidic pH, this can be explained by the propriety of the polycations used in the pillaring process (Avena et al., 1990). The PZC values of Zr pillared clays are more acidic compared to Al pillared samples. The PZC value of Al-PILCs found by us is in good agreement with the PZC value of Al-PILCs reported in the literature (Avena et al., 1990; Mrad et al., 1997). The introduction of zirconium in the aluminum intercalated solution decreases the PZC values of all mixed solids in comparison to the PZC of Al-G toward the PZC of single Zr-G which is equal to 4.7. These values are near the PZC of ZrO₂ found in the literature which range from 4 to 5 (Haussonne et al., 2005). The decrease of PZC correlates well with the number of Zrp calculated as atom per cell (Table 2). Except for the Zr₉₀Al₁₀-G which contains the highest amount of Zrp, its PZC is lower than Zr-G simple. This is very much due to the existence of a significant amount of introduced aluminum (Alp = 1.30 atom/cell) that increases the PZC.

3.2.2 Potentiometric titration curves

The acid–base potentiometric titration curves at different salt concentrations were used to measure the proton adsorption or proton charge. All experiments for charge determination were carried out under ambient laboratory conditions. The experimental method employed was similar to that used by Huertas et al. (1998) and Sposito (1998). The NaCl electrolyte concentration was adjusted to 10⁻³, 10⁻² and 10⁻¹M. The proton surface charge density σ_H (mol m⁻²), determined from potentiometric titration was calculated as the difference between total amounts of H⁺ or (OH⁻) added to the dispersion and required bringing a blank solution of the same NaCl concentration to the same pH (Schroth and Sposito, 1997).

Figs. 3 and 4 show the acid–base potentiometric titration curves of the starting and modified bentonite dispersed in different NaCl concentrations. The pure and pillared clay surfaces undergo protonation followed by deprotonation. For Na-G, below pH < PZC and in acidic range, the cation exchange at layer sites and protonation of edge sites (≡AlOH groups) occur simultaneously. For pH > PZC and in alkaline pH range, deprotonation of surface hydroxyl groups exposed at the edge sites (≡SiOH and ≡A1OH at high pH) of smectite platelets causes an overall negative charge. The behavior of these curves is almost the same in the considered pH range and resembles those in the literature (Avena and De Pauli, 1998; Kriaa et al., 2007) with the difference that they had weak slope σ_H vs. pH, in the pH near to crossing the pH axis. This can be explained, in agreement with Avena and De Pauli (1998) by an edge-to-face interaction between positively charged edges and negatively charged faces or H⁺ adsorption exchanging some Na⁺ ions in faces that would slow the protonation of hydroxylated sites at the edges. In addition, the increase in the slopes of the titration curves at pHs < 5 is also of interest. This increase in the adsorption of H⁺ ions for the bentonite sample can be explained by the Na⁺–H⁺ exchange according to the following equation, reported by Delgado et al. (1986): $$\text{Na-smectite}+{\text{H}}^{+}\to \text{H}-\text{smectite}+{\text{Na}}^{+}$$

Close

Figure 3

Potentiometric titration curves at I = 10⁻¹, 10⁻² and 10⁻³ (M) for Na-G and Zr and Al-pillared samples.

Export to PPT

Close

Figure 4

Potentiometric titration curves at I = 10⁻¹, 10⁻² and 10⁻³ (M) for Zr–Al-pillared samples.

Export to PPT

Comparing the behavior of acid–base potentiometric titration curves of pillared samples with those of Na-G bentonite we can conclude that the acid–base properties of starting and modified samples are similar. All studied surfaces undergo protonation and deprotonation reaction in the pH range 3–10. However the differences in the shape of the curves may be explained by the differences in the composition of all samples. In fact, the creation of new sites in the case of pillared clay must be considered (i) protonation of pillar sites ≡AlOH or/and ≡ZrOH before the PZC (ii) deprotonation of these sites after the PZC. In addition, we should note that the weakly acidic surface functional groups (≡XH) accounting for ion exchange reaction can be neglected in the case of the different pillared samples because pillared samples have a low cation exchange capacity. Beside, the point of interception of σ_H curves of single and mixed pillared bentonite indicated that the surface became more negative. The pH_PZC decreases from 7.3 to 5 after pillaring.

To describe the behavior of starting and modified bentonite, our approach is similar to that used by Huertas et al. (1998) for kaolinite. We assume that one protonation-deprotonation occurs at solid surface. The protonation and deprotonation reactions of a surface site ≡MOH are represented by the following equations:

The acid dissociation constant $K_{{\text{a}}_1}$ and $K_{{\text{a}}_2}$ of ≡MOH and $\text{≡}{\text{MOH}}_2^{+}$ respectively can be written as follows:

X_i is donated by the following equation: $X_{\text{i}}=\frac{{\theta }_{\text{i}}}{{\theta }_{\text{M}}}$

In which θ_M and θ_i are the total surface density site and the surface density of any site respectively.

The mass balance for each kind of surface sites is given by the sum of the positive, neutral, and negative species:

If we consider, as an approximation, that the positive and negative complexes are the only charged species, respectively, below and above the pH_PZC; Eq. (6) may be simplified as follows:

And consequently, after rearranging expressions (4) and (5), the acid dissociation constant $K_{{\text{a}}_1}$ and $K_{{\text{a}}_2}$ can be obtained:

The surface site density ${\theta }_{\text{i}}({\theta }_{\text{I}}={\theta }_{\text{M}}{X}_{{\text{MOH}}_2^{+}}\text{and}{\theta }_{\text{II}}={\theta }_{\text{M}}{X}_{{\text{MO}}^{-}})$ was estimated by fitting the sections of the potentiometric titration curves using least-square nonlinear regression analysis.

From $K_{{\text{a}}_1}$ and $K_{{\text{a}}_2}$ value we calculated the PZC value for each sample using the following equation:

In the previous calculation, we assumed the existence of two types of sites: a homogenous acidic site (pH < PZC) and a homogenous basic site (pH > PZC). Here, the two sites-two pK_as model (Model I) is considered. This model simplifies the heterogeneous Na-G bentonite and pillared sample surface as a system with two uniform sites.

To better explain the surface properties of our samples and based on the difference in the shape of σ_H versus pH curves, we also tested the multisites model (Model II). In the case of Na-G and Al-G, results of this study lead to two sites-two pK_as in the acidic domain and two sites-two pK_as in the basic one. However; by using this model, we got three sites-three pK_as in the protonation and in the deprotonation sections for pillared sample containing zirconium showing that modification with zirconium creates new active sites. pK_a and θ_i values estimated with Model II are summarized in Table 5. We see clearly that the pK_a values obtained from Model II as from Model I (Table 4) show the stronger acidic surface sites of samples rich in zirconium. However; Model II considers the heterogeneity of the sample surface and gives more detailed description of the acid-base behavior of modified and starting bentonite. We can see that Model II allows the distinction of the stronger basic sites (pK_a > 9). In fact, using Model I, the pK_a2 of modified sample ranges from 6 to 8. But when using Model II, we can obtain pK_a ≈ 10.