Intercalation of Zr(IV) into the acid-activated (HK ) and untreated (UnK) kaolin.

UnK or HK (10 g) was added to 10 mL of distilled H₂O, impregnated for a week in a stoppered 250 mL glass conical flask to guarantee sufficient hydration of kaolinite layers, then ZrOCl₂·8H₂O (19.69 g ≡ 0.061) was added and mixed thoroughly. Using a high concentration of ZrOCl₂ (6.1 mol L^–1) was important to avoid hydrolysis (Solovkin & Tsvetkova, Reference Solovkin and Tsvetkova1962). The slurry was sonicated for 60 min and impregnated again for 6 days with frequent shaking. For analysis purposes, 1 g was separated after this stage, dried at 70°C, and labeled as HKZS. The mixture was dried gradually at 100°C for 24 h (labeled as UnKZS100 and HKZS100, respectively), then heated again at 200°C for 60 min, washed with 10 mL of 0.1 mol L^–1 HCl and then with 50 mL of distilled water, dried at 105°C, and labeled as UnKZS200 and HKZS200, respectively. Finally, the UnKZS200 or HKZS200 was heated at 300 and 500°C for 3 h and labeled as UnKZS300 and UnKZS500 or HKZS300 and HKZS500, respectively. For the sake of comparison, the same procedure for the synthesis of HKZS200 was repeated without HK to evaluate the effect of wet aging, sonication, and thermal treatment on ZrOCl₂·8H₂O. The samples produced were labeled as ZrOCl₂-S, ZrOCl₂-S100, and ZrOCl₂-S200 for the samples heated at 70, 100, and 200°C, respectively. In addition, the synthesis of HKZS-100 was repeated, replacing ZrOCl₂·8H₂O and water with 10 mL of concentrated HCl (36–37%) to resemble the acidity produced during the hydrolysis of ZrOCl₂·8H₂O. The samples produced were labeled as HK-S-HCl, HK-S-HCl100, and HK-S-HCl200 and were heated at 70, 100, and 200°C, respectively. HK was also heated at 300 and 500°C, and similarly labeled as HK300 and HK500, respectively. Detailed names, labels, and synthesis parameters are given in the Supplementary Material (Table S1).

Adsorption of Cd ²⁺ , Cu ²⁺ , or Pb ²⁺ .

The adsorption experiments were carried out at room temperature (~25°C). To investigate the effect of the initial pH on the adsorption behavior of Cd²⁺, Cu²⁺, or Pb²⁺, a 10 mL volume (v) with an initial concentration of 10 mg L^–1 (C _i) of the metal ions (as nitrate) was added to a mass (m) of 10 mg of HK or HKZS300. The resulting solutions containing adsorbent were agitated for 60 min. The initial pH values, 1.87, 3.0, 5.0 (acetate-buffered), and 8.0 (ammonia-buffered), were controlled using 0.1 M NaOH and HNO₃ solutions. The suspensions were filtered and the final concentration of metal ions in solution (C _f) was determined by FAAS. The removal efficiencies (E, %) of the metal ions onto HK or HKZS300 were calculated using the following equation:

(1) $E=\frac{\left(C_i-C_f\right)}{C_i}100\left(\%\right)$

The distribution coefficient of the metal ions (K _{d, M} ) and the selectivity coefficient (S _M1/M2) were calculated using the following equations:

(2) $K_{d,M}=\frac{\left(C_i-C_f\right)·v}{C_f·m}\left(mL{g}^{-1}\right)$

(3) $S_{M1/M2}=\frac{K_{d,M1}}{K_{d,M2}}$

XRD/TG studies.

X-ray diffraction (Fig. 1) traces indicated the effect of zirconium oxychloride on kaolinite in the raw inactivated kaolin (UnK) and the HCl-activated kaolin (HK) after homogenizing mixtures by a sonication-impregnation technique and heating the products at 100, 200, 300, or 500°C. The XRD pattern of the parent materials (UnK and HK) were typical of kaolin and similar to those obtained previously with d ₀₀₁ values for kaolinite of 7.19 Å (Drweesh et al., Reference Drweesh, Fathy, Wahba, Hanna, Akarish, Elzahany, El-Sherif and Abou-El-Sherbini2016). The XRD pattern of ZrOCl₂·8H₂O was also typical of a pure substance (ICSD 00-032-1498).

Fig. 1. XRD patterns of UnK, HK, ZrOCl₂·8H₂O, and their ZrO₂-modified products. K, Q, Z, and T denote kaolinite, quartz, ZrOCl₂·8H₂O, and tetragonal zirconia, respectively

Within the present XRD parameters, no decisive evidence of the intercalation of ZrO₂ into kaolinite layers of the ZrOCl₂-treated kaolins could be observed. A remarkable decrease in the intensity of the kaolinite reflections was observed, however, especially that of the basal spacing (d ₀₀₁) of kaolinite phases compared with the d ₁₀₁ reflection of quartz at 3.34 Å in the parent kaolins (UnK and HK) and concentrated HCl-treated HK blanks (HK-S-HCl100 and HK-S-HCl200). Although new, broad reflections were observed, with greater d ₀₀₁ values of 10.7–13.6 Å, for the ZrOCl₂-treated samples dried at 100–200°C, which are close to the reported basal spacing of meta-kandite intercalated with ZrO₂– 14.5 Å (Vaughan, Reference Vaughan1994), they may also be due to the residues of incompletely decomposed ZrOCl₂·8H₂O when compared with the diffraction patterns of ZrOCl₂-S100 and ZrOCl₂-S200. Relative broadening as well as the decrease in intensity of basal spacing peaks, compared to the parent clay mineral, were reported to be good evidence for the existence of successful intercalation, and/or an irregular-packing of intercalated and non-intercalated layers (Valverde et al., Reference Valverde, de Lucas, Sánchez, Dorado and Romero2003; Majd et al., Reference Majd, Davoudi, Ramezanzadeh, Ghasemi, Ramezanzadeh and Mahdavian2020). In addition, the intercalation of kaolinite, in particular, may reduce XRD quality due to the generation of lattice perturbation, defects, and disorders in the structure (Grzybek et al., Reference Grzybek, Klinik, Olszewska, Papp and Smarzowski2001).

Zr(IV)-intercalation in other clays was reported (Ohtsuka et al., Reference Ohtsuka, Hayashi and Suda1993; Chaabene et al., Reference Chaabene, Bergaoui and Ghorbel2004; Valášková et al., Reference Valášková, Tokarský, Hundáková, Zdrálková and Smetana2013). Two models of Zr-tetrameric cations ([Zr₄(OH)₁₄(H₂O)₁₀]²⁺ and [Zr₄(OH)₈(H₂O)₁₆]⁸⁺, and one model of Zr-octameric cation ([Zr₈(OH)₂₀(H₂O)₂₄]¹²⁺) were built; only the tetrameric models were studied for intercalation into vermiculite because the octameric cation was excluded due to its thickness (1.0 nm) being greater than that of the vermiculite layer (0.86 nm) (Valášková et al., Reference Valášková, Tokarský, Hundáková, Zdrálková and Smetana2013). The insertion of Zr-tetrameric cations into vermiculite was suggested to yield a d ₀₀₂ of 14.2 Å. Successful Zr(IV)-intercalation in montmorillonite was also reported to have a van der Waals’ spacing of 11.4 or 7.0 Å (Ohtsuka et al., Reference Ohtsuka, Hayashi and Suda1993; Chaabene et al., Reference Chaabene, Bergaoui and Ghorbel2004), assigned to hosting of octamer [Zr₈(OH)₂₀(H₂O)₂₄Cl₁₂]⁴⁺ or tetramer [Zr₄(OH)₈(H₂O)₁₆Cl₆]²⁺, respectively.

According to the length and thickness of the [Zr₄(OH )₁₂(H₂O)₁₂]⁴⁺ polycation estimated to be 9.1 and 5.5 Å, respectively (Miehé-Brendlé et al., Reference Miehé-Brendlé, Khouchaf, Baron, Le Dred and Tuilier1997), the intercalation form of Zr(IV) in the Zr-modified kaolin samples is supposed to proceed in two steps: (1) ZrOCl₂·8H₂O undergoes hydrolysis which increases with temperature (Matsui & Ohgai, Reference Matsui and Ohgai2002), to form tetramers [Zr₄(OH)₈(H₂O)₁₆]⁸⁺ or more complex polyoxide polymers. (2) The resultant hydrated tetramers are bonded to the kaolinite layers with face-to-face or edge-to-edge positioning parallel to the xy plane as the unit kaolinite layer thickness is ~5.4 Å (Zhang et al., Reference Zhang, Chen and Xing2017).

Comparing the diffraction patterns of UnKZS100-300 (ZrOCl₂-treated inactivated kaolin) with HCl-activated analogs HKZS100–300, the ratio of intensities of the 00l lines of kaolinite to the quartz reflection at 3.34 Å, accounting for the d ₁₀₁ of quartz, were observed to decrease in the HK-activated kaolin compared with untreated kaolin, which may indicate that the ZrO₂ binding process is enhanced by the acid-activation of the kaolinite surface. This may be attributed to the release of more surface OH groups upon acid activation (Drweesh et al., Reference Drweesh, Fathy, Wahba, Hanna, Akarish, Elzahany, El-Sherif and Abou-El-Sherbini2016).

The ~13 Å d value was not observed by XRD for the samples heated at 300 or 500°C (Fig. 1), while sustaining a remarkable reduction in the parent d ₀₀₁ intensity relative to the d value at 3.34 Å accounting for the d ₁₀₁ reflection of quartz. The increase of background intensity at >8°2θ for both UnKZS300 and HKZS300 may be due to the polymerization of intercalated zirconia building units with increasing temperature yielding an amorphous kaolinite with an enlarged perturbed structure. Polymerization of the intercalated Zr(IV) species was suggested to form zirconia sandwiched face-to-face or edge-to-edge between less ordered stacking of the kaolinite layers. The polymerization of tetramer and octamer forms of Zr(IV) was reported to be enhanced with calcination yielding crystalline zirconia (Farfan-Torres et al., Reference Farfan-Torres, Sham and Grange1992; Chaabene et al., Reference Chaabene, Bergaoui and Ghorbel2004).

A loss of the uniform layered packing in the xy plane of Zr-modified kaolinite layers may accompany thermal treatment that causes aggregation of zirconia particles and weakens bonding between kaolinite layers with ZrO₂. This explanation is supported by XRD (Fig. 1) via the disappearance of kaolinite reflections (due to the dehydration of the kaolinite and the formation of metakaolinite (Gil et al., Reference Gil, Vicente, Lambert and Gandιa2001)) and the development of a ZrO₂ metastable tetragonal phase [PDF80-0965] at 500°C. Obtaining the tetragonal ZrO₂ metastable phase at such a low temperature is interesting due to its catalytic importance (Vera et al., Reference Vera, Pieck, Shimizu and Parera2002), knowing that it crystallizes at a temperature of 1205°C (Chevalier et al., Reference Chevalier, Gremillard, Virkar and Clarke2009), or at lower temperature by using special calcination conditions such as the presence of silica or alumina as supporting materials (Nagarajan & Rao, Reference Nagarajan and Rao1990).

TGA. In order to understand the mechanism further, a TGA measurement (Fig 2) was performed for the ZrOCl₂·8H₂O and HKZS sample that was dried at 70°C without further thermal treatment. The TGA of ZrOCl₂·8H₂O showed the presence of five mass-loss stages, with T _max at 113, 162, 286, 343, and 460°C, which corresponded to mass-loss stages of 25.3, 18.3, 10.2, 3.1, and 3.4%, respectively. These results are close to the thermal decomposition of ZrOCl₂·8H₂O, performed in air, reported to occur in two major mass-loss stages (Takagi, Reference Takagi1954; Gorodylova et al., Reference Gorodylova, Šulcová, Bosacka and Filipek2014; Gorodylova & Šulcová, Reference Gorodylova and Šulcová2018), which help to explain the ZrOCl₂·8H₂O/HK reaction. A release of crystalline water was reported within the range 70–230°C, leading to the formation of XRD-amorphous ZrOCl₂ as illustrated in Eq. 4. In the temperature range of 230–500°C, ZrOCl₂ reacts with atmospheric water to form amorphous ZrO₂ and HCl is released as shown by Eq. 5. A small exothermic effect was observed with a maximum rate at 500°C that can be associated with the crystallization of monoclinic ZrO₂(Beden & Guillaum, Reference Beden and Guillaum1969) as represented by Eq. 6.

(4) ${\text{ZrOCl}}_2·8H_{2}O\stackrel{70-230°C}{\to }{\text{ZrOCl}}_2+8H_{2}O$

(5) ${ZrOCl}_2+H_{2}O\stackrel{230-500{}^{○}C}{\to }amorphous{ZrO}_2+2HCl$

(6) $amorphous{ZrO}_2\stackrel{500{}^{○}C}{\to }monoclinic{ZrO}_2$

Fig. 2. TGA and DTG of ZrOCl₂·8H₂O and impregnated ZrOCl₂·8H₂O + HK (HKZS) at a heating rate of 10°C/min under N₂

The TGA of the HKZS (Fig. 2) showed the presence of four principal mass-loss stages at T _max 48.9, 130.8, 261.0, and 436.6–472.6°C, which correspond to 2.90, 11.92, 9.15, and 9.37% of mass-loss stages, respectively.

On the other hand, TGA of HK was reported to show two main mass-loss stages at 25–75°C and 490°C, which were attributed to elimination of the adsorbed H₂O and dehydroxylation of kaolinite aluminols, respectively (Drweesh et al., Reference Drweesh, Fathy, Wahba, Hanna, Akarish, Elzahany, El-Sherif and Abou-El-Sherbini2016), as explained in Eq. 7.

(7) ${\mathrm{Al}}_2{\mathrm{Si}}_2{O}_5{\left(\mathrm{OH}\right)}_4\to {\mathrm{Al}}_2{\mathrm{Si}}_2{O}_7+{\text{2H}}_{2}O$

Accordingly, the first and second stages in the present TGA study of HKZS (Fig. 2) were attributed to the loss of water (either adsorbed on kaolin or crystallized in ZrOCl₂·8H₂O) and HCl. The formation of white clouds in the ammonia solution test indicated the evolution of HCl gas, beginning from 70°C. The third stage may have been related to the loss of most chloride content, as revealed from the EDX measurement of the HKZS300 (Fig. S2) and the elemental mapping results (Fig. S3), which differed from the reported thermal decomposition of ZrOCl₂·8H₂O to amorphous ZrO₂(Eq. 6), that may be physically or chemically bonded to perturbed kaolinite layers as suggested by the XRD study.

The last TGA stage, which showed a remarkable reduction in the offset-temperature of the stage of HK (490°C) (Drweesh et al., Reference Drweesh, Fathy, Wahba, Hanna, Akarish, Elzahany, El-Sherif and Abou-El-Sherbini2016), was composed of multiple overlapped thermal decomposition reactions. During this stage, dehydroxylation of aluminol groups clearly took place. Accordingly, the existence of the dehydroxylation stage below 500°C and its complex structure are evidence of the bonding of the kaolinite layers with the intercalated ZrO₂(Kristó et al., Reference Kristó, Frost, Felinger and Mink1997; Cheng et al., Reference Cheng, Liu, Cui, Zhang, Zhang and Frost2012).

The last stage was accompanied by the crystallization of the tetragonal ZrO₂ as concluded from the XRD pattern of the HKZS500 (Fig. 1), contrary to the expected monoclinic form reported in the literature for the thermal treatment of ZrOCl₂.8H₂O in air (Gorodylova et al., Reference Gorodylova, Šulcová, Bosacka and Filipek2014). This may be attributed to the influence of kaolinite as a support for the crystallization process of ZrO₂.

XRF. Analysis by XRF of UnK and HK and their zirconia-modified kaolins UnKZS300 and HKZS300 (Table 1) gave results which were very similar to those from EDX measurements (Fig. S3), indicating the homogeneity of the sample despite the presence of separate phases of silica-, titania-, and zirconia-rich kaolinite. The XRF indicated that Si/Al atomic ratios increased to 1.383 and 1.631 in UnKZS300 and HKZS300 compared with their initial values (1.226 and 1.258) in the parent samples UnK and HK, respectively. On the other hand, Cl/Zr atomic ratios decreased from the theoretical 2:1 in ZrOCl₂·8H₂O to only 0.329 and 0.272 for UnKZS300 and HKZS300, respectively. These results suggest a partial release of Al as AlCl₃ from the kaolinite lattice as a result of the thermal treatment of zirconyl chloride with kaolin at temperatures up to 300°C according to Eq. 8.

(8) $5/2x\mathrm{ZrO}{\mathrm{Cl}}_2+{\mathrm{Al}}_2{\mathrm{Si}}_2{O}_5{\left(\mathrm{OH}\right)}_4\stackrel{300°C}{\to }\text{amorphous}\mathrm{Zr}{O}_2/{\mathrm{Al}}_{2-x}{\mathrm{Si}}_2{O}_5{\left(\mathrm{OH}\right)}_{4–2x}+x\mathrm{Al}{\mathrm{Cl}}_3+2x\mathrm{HCl}$

Table 1. XRF analysis of UnK and HK and their zirconia-modified kaolins

As determined by loss on ignition (LOI), the kaolinite percentages were 89.28 and 83.90% in UnK and HK, respectively

The atomic ratios of Zr/Al in UnKZS300 and HKZS300 (0.637 and 0.691) were also smaller than the starting values of UnK and HK (0.912 and 0.920). These results mean that the loading efficiency of zirconia with respect to kaolinite in HKZS300 is greater than that of UnKZS300 where zirconia loading efficiencies were 69.8 and 75.1%, respectively.

On the other hand, metal contents in the acid-activated samples were slightly less than those for the untreated samples, indicating the leaching effect of HCl that releases the OH groups engaged in bonding with the leached metal ions; this was also reported previously (Drweesh et al., Reference Drweesh, Fathy, Wahba, Hanna, Akarish, Elzahany, El-Sherif and Abou-El-Sherbini2016).

FTIR. The FTIR spectra (Fig. 3) obtained for the ZrOCl₂·8H₂O, UnK, HK, HK-S-HCl, HK300, and HK500 samples and their kaolin-ZrO₂-modified products showed, for UnK, HK, and HK-S-HCl, three strong absorption bands centered at 3697, 3654, and 3622 cm^–1, corresponding to the inter- and intra-hydroxyl groups (Farmer & Russell, Reference Farmer and Russell1964; Frost et al., Reference Frost, Kristof, Paroz, Tran and Kloprogge1998; Abou-El-Sherbini et al., Reference Abou-El-Sherbini, Elzahany, Wahba, Drweesh and Youssef2017). The bands at ~1626 and ~3430 cm^–1 correspond to the bending and stretching modes of the OH groups of water molecules. The bands centered at 1111, 1034, 1005, 938, 913, 792,753, 694, 537, 469, and 428 cm^–1 are assignable to the apical Si–O stretching, two in-plane Si–O–Si stretching, inner Al–OH bending, inner surface Al–OH bending, three OH translational, Si–O–Al in-plane bending deformation, Si–O bending, and Si–O vibrations, respectively, which are typically the IR absorption bands of kaolinite (Frost & Vassallo, Reference Frost and Vassallo1996). The FTIR spectra of HK300 showed a 1–2 cm^–1 blue shift in the wavenumber positions of stretching modes of the inner and inner surface hydroxyl-group vibrations while in HK500 all Al–O–H-including vibration bands disappeared due to the dehydroxylation of aluminols and the formation of the amorphous metakaolinite phase (Frost & Vassallo, Reference Frost and Vassallo1996). Also, new broad bands appeared at 1080 and 799 cm^–1, ascribed to the Si–O and AlO₄ vibrations, respectively, in the newly formed metakaolinite (Ilić et al., Reference Ilić, Mitrović and Miličić2010). The FTIR spectra of both ZrOCl₂.8H₂O and ZrOCl₂-S showed strong absorption bands at 3000–3500 cm^–1(broad) and 1620–1624 cm^–1 due to ν_OH and δ_OH vibrations of water molecules. Much weaker absorption bands at 545–548, 438–439, and 407–415 cm^–1 are due to Zr–O (Štefanić et al., Reference Štefanić, Musić, Popović and Furić1996).

Fig. 3. FTIR spectra of HK, HK300, HK500, and their ZrO₂-modified products

By comparing the FTIR spectra of HK and HK300 with those of HKZS, HKZS300, and UnKZS300, the effect of the Zr(IV) coating on the kaolinite structure was confirmed by the observed 2 cm^–1 red shift in the stretching modes of the inner and inner-surface hydroxyl-group vibrations (especially in HKZS300), the remarkable weakness of the surface stretching AlO–H vibrations, and a red-shift of 15 cm^–1 of the stretching mode of the OH groups of water molecules at 3430 cm^–1. Also, the bands due to Zr–O at 407–415 and 438–439 cm^–1 in ZrOCl₂·8H₂O and ZrOCl₂-S were blue-shifted to 420 and 449 cm^–1 in the FTIR spectra of HKZS but could not be distinguished in the other zirconia-modified kaolin samples. The interaction of OH, Si–O, and aluminol groups with the Zr(IV) species confirmed the intercalation of zirconia and the presence of strong bonding with kaolinite surface as concluded elsewhere with another aluminosilicate support; SBA-15(Gong et al., Reference Gong, Sun, Sun, Li and Liu2011).

In addition, the FTIR spectra of UnKZS500 and HKZS500 support the intercalation process of zirconia through the remarkable decrease of bands due to AlO₄ vibrations in the metakaolinite, in addition to the appearance of multiple weak bands in the region of the inner and inner-surface OH-group vibrations for HKZS500, which confirmed the strong interaction between aluminols and zirconia. The blue shift of hydroxyl-stretching vibrations was reported to indicate weaker hydrogen bonding, corresponding to the structural OH groups with the higher energies (i.e. wavenumbers) and the weaker participation in the hydrogen bonding (Libowitzky & Beran, Reference Libowitzky and Beran2004). Accordingly, the observed red shift accompanying zirconia upload may be correlated with O–H stretching frequencies and O–H…..O stronger hydrogen bond strength due to the coating with zirconia (Libowitzky, Reference Libowitzky, Schuster and Mikenda1999). A summary (Table 2) was given for the FTIR absorption spectra (cm^–1) of kaolin and its zirconia-modified samples.

Table 2. Locations and assignment of FTIR absorption bands (cm^–1) of kaolin and its zirconia-modified samples

vw: very weak, w: weak, def: deformation

SEM. Examination of HK by SEM (Fig. 4a) indicated that kaolinite is found as discrete, coarse aggregates, poorly crystallized with irregular geometry that might reflect its detrital nature. In addition, silt-sized quartz grains were present. The SEM image of HKZS300 (Fig. 4b) revealed improvement in the lamellar crystalline nature of the clay, as well as a uniform distribution of Zr over all of the face-to-face- and edge-to-edge-ordered clay layers as confirmed by the EDX measurement (Fig. S3), despite the presence of minor constituents of ZrO₂-rich/kaolinite. Zeta size mapping (Fig. S4) confirmed these large differences in the morphologies of HK and HKZS300. Although HK was polydispersed, mainly due to the presence of sedimentary particles, it showed a high population of particles having a diameter of 1 μm, while the HKZS300 sample was not suitable for DLS measurements due to its polydispersed nature. The HKZS300 contained large particles and aggregates with an average size larger than the upper size display limit of the Zeta Sizer instrument (10 μm). The presence of zirconia in the HKZS300 led to binding of the layers with kaolinite particles.

Fig. 4 SEM images of aHCl-treated kaolinite (HK) and b its Zr-intercalated sample HKZS300

TEM. Transmission electron microscopy images for HKZS300 (Fig. 5a–h) support the presence of zirconia-intercalated kaolinite phases with basal spacings of ~11 to 25 Å, as illustrated in Fig. 5h, while the 7.16 Å d ₀₀₁ value of the parent kaolinite is present also. Both kaolinite phases, the pristine and the enlarged, could be recognized easily even in the same particle (regions A and B in Fig. 5c). The modified kaolinite crystallites were observed to be darker on their edges than in their centers (Fig. 5d, f). A TEM-SAED pattern of kaolinite layers (Fig. 5d, e) also showed well-defined spots with d spacings of ~21.49 Å. These observations may be explained by assuming gradient contents of intercalated zirconia into kaolinite interlayer space that decrease toward the kaolinite crystallite center. The observed darker edges compared to centers of particles (Fig. 5d, f) support the suggested cement-like action of zirconia to the kaolinite particles, which in turn led to enhancement of the stacking order of the kaolinite particles. In addition, the kaolinite layers appeared thicker by variable levels than their ideal 5.4 Å thickness of the silica-gibbsite layer (Fig. 5g, h). The outer surface of the kaolinite crystallites in the XY plane was observed to be coated with a layer of nano-sized dark particles of 1.8 nm diameter, capturing nanopores within them (Fig. 5g); these may be zirconia nano-particles formed by dimerization or higher polymerization of zirconia tetramers. The enlarged kaolinite phase was observed to be wavy and strongly perturbed, which explains the poor-quality XRD patterns of zirconia-modified kaolin samples. Partial intercalation of ZrOCl₂·8H₂O into kaolinite occurred at temperatures of up to 500°C (Fig. 6).

Fig. 5. TEM images a–h of Zr-intercalated sample HKZS300. TEM-SAED pattern of kaolinite layers in part d is shown in part e. The arrows indicate examples of enlarged interlayer spacings while the yellow lines show some perturbed and enlarged layers. Calculations written in yellow were made by ImageJ software

Fig. 6. Schematic representation of ZrOCl₂·8H₂O intercalation into kaolinite layers and its temperature dependence

Texture study.

The N₂adsorption/desorption isotherms and pore-size distributions (Fig. 7a, b , respectively) were performed for HK and its zirconia-modified product, HKZS300. Three well distinguished regions of the adsorption isotherm were evident: (1)monolayer-multilayer adsorption; (2) capillary condensation; and (3) multilayer adsorption on the outer surfaces of the particles. HK samples have the type II isotherm with a distinct capillary condensation step, which is characteristic of mesoporous materials according to the classification of the International Union of Pure and Applied Chemistry (IUPAC). Nitrogen adsorption-desorption isotherms of HK illustrated a clear H4 type hysteresis loop in the relative pressure range between 0.4 and 1.0, implying that this material had very regular mesoporous channels. Based on the surface texture data (Table 3), HK was observed to have a mesopore area of 24.17 m² g^–1 and mesopore volumes of 0.125 cm³ g^–1, which are in good agreement with the results reported by Drweesh et al. (Reference Drweesh, Fathy, Wahba, Hanna, Akarish, Elzahany, El-Sherif and Abou-El-Sherbini2016).

Fig. 7. a N₂ isothermal adsorption and desorption profiles and b pore-size distribution plots for HK and HKZS300 samples

Table 3. Surface texture parameters of HK and its zirconia-modified product HKZS300

The adsorption isotherm of HKZS300 (Fig 7a) was estimated to be Type-II, based on the IUPAC classification of sorption isotherms, indicating the existence of mesoporosity. In addition, an obvious type H4 hysteresis loop can also be seen in the relative vapor pressure range of 0.4–1.0 bar, indicating the presence of a hierarchical micro/mesoporous system (Valášková et al., Reference Valášková, Tokarský, Hundáková, Zdrálková and Smetana2013). This was further ascertained by the BJH pore-size distribution plot (Fig. 5b) which confirmed the presence of micropores with an average size of ~1.83 nm and mesopores with an average size of ~3.98 nm. The HKZS300 sample showed a total surface area of 59.74 m² g^–1 with micro- and mesoporous surface areas of 23.68 and 36.06 m² g^–1, respectively. The total pore volume of the sample was 0.072 cm³ g^–1 in which the micro- and mesoporous volumes were 0.009 cm³ g^–1 and 0.063 cm³ g^–1, respectively. In other words, the bulk of the total surface area and pore volume of HKZS300 was derived from mesopores. Such a mesoporous structure would be favorable for adsorption purposes in terms of improved accessibility of the adsorbent to the active binding centers within the porous structure of HKZS300, as well as high diffusion rates. The increment in S _BET after zirconia intercalation represents an increase of ~250% over the parent surface area, which is in agreement with the morphological changes observed in TEM images. According to previous studies, the acid-activated samples HK and HKZS300 were selected for comparing their properties in the adsorption study of the Cd²⁺, Cu²⁺, and Pb²⁺, due to their confirmed reduced trace-metal contents, and the remarkable influence of zirconia on the kaolinite structure compared with the inactivated samples.