Materials

Natural kaolinite was collected from the Jabal Al-Harad kaolin deposit (Batn El-Ghoul area). It is located in southern Jordan, 280 km south of Amman, ~70 km southeast of Ma’an city, and 4 km to the east of the Ma’an–Mudawwara road at latitude 29°37′52″N and longitude 35°20′57″E (Fig. 1; Gougazeh & Buhl Reference Gougazeh and Buhl2010). According to Rietveld refinement of X-ray diffraction (XRD) data, the material used in this study for synthesis reactions is ~80 wt.% kaolinite and ~16 wt.% quartz and has a chemical composition of 53.86 wt.% SiO₂, 0.74 wt.% TiO₂, 32.45 wt.% Al₂O₃, 0.65 wt.% Fe₂O₃, 0.08 wt.% MgO, 0.13 wt.% CaO, 0.06 wt.% Na₂O, 0.54 wt.% K₂O and 11.21% LOI (Gougazeh & Buhl Reference Gougazeh and Buhl2010).

Fig. 1 Location map of the Jabal Al-Harad, Jordan, kaolin deposit

Sodium hydroxide (NaOH, 99%) as pellets was purchased from Merck Chemical Company (Darmstadt, Germany), citric acid and sodium silicate solution (Na₂Si₂O₅) were provided by Fluka Chemie (Buchs, Switzerland), and analytical-grade Basic Blue 41 (BB-41) was purchased from Sigma–Aldrich (Munich, Germany) and used without further purification. BB-41 has the molecular formula C₂₀H₂₆N₄O₆S₂ (molecular weight = 482.57 g/mol) with a color index number 11105. The chemical structure of BB-41 is shown in Fig. 2 (Tajul Islam et al. Reference Tajul Islam, Aimone, Ferri and Rovero2015). Distilled water was used in the standard purification methods.

Fig. 2 Chemical structure of the dye Basic Blue 41

Preparation of Zeolites

The zeolite A was prepared following the method described by Gougazeh & Buhl (Reference Gougazeh and Buhl2014). 1.0 g of raw kaolinite (RK) was treated in 8 M NaOH at 120°C for 20 h. The product was washed three times using distilled water and dried overnight at 80°C. Then, 1 g of the product, hydroxysodalite, was mixed with diluted citric acid (20 mL 1 M) and heated to ~70°C for 2–4 h. To this solution, solid NaOH powder (3.2 g) was added with vigorous stirring at room temperature for homogenization. The resulting zeolite precursor gel was placed in an autoclave (50 mL) and kept at ~100 °C for 3 h. After hydrothermal treatment, the product was filtered and washed three times with distilled water to remove excess alkali, then oven-dried at 80 °C for 4 h.

The process to prepare zeolite X consisted, firstly, of the alkaline hydrothermal transformation of kaolinite into hydroxysodalite, as described above for zeolite A, followed by the same treatment with diluted citric acid (20 mL 1 M) and heating to ~70°C for 2–4 h. To this solution, 3.2 g (8.0 M) of solid NaOH powder and 0.5 g of solid Na₂SiO₃ powder were added with vigorous stirring at room temperature for homogenization; the pH of the aluminosilicate solution was adjusted to ~12 to obtain a gel; it was then placed in an autoclave (50 mL) and kept at ~80°C for 16 h. After hydrothermal treatment, the product was filtered and washed three times with distilled water to remove excess alkali. Next, the samples were oven-dried at 80°C for 4 h.

Batch Mode Adsorption Studies

The adsorption experiments were performed in a batch process (Kooli et al. Reference Kooli, Yan, Al-Faze and Al-Sehimi2015b). Preliminary experiments demonstrated that equilibrium was established in ~8 h, but the experiments were allowed to run for 18 h. To obtain adsorption isotherms, a fixed amount (0.100 g) of kaolinite or zeolite was mixed with 10 mL of dye solution of different concentrations from 25 to 1000 mg/L. The supernatant was collected by centrifugation for 10 min at 2380×g using a Labofuge 200 centrifuge. The equilibrium BB-41 concentrations were determined by using a Cary 100 Conc spectrophotometer from Varian (Sydney, Australia) at 610 nm.

The removal capacity of dye was calculated as follows:

(1) $q_e=\left(C_i-C_e\right)\times \frac{V}{W}$

and the removal percentage of dye was calculated as

(2) $R\%=\frac{\left(C_i-C_e\right)}{C_i}\times 100$

where q _e (mg/g) is the amount of dye adsorbed, C _i (mg/L) is the initial dye concentration, C _e is the dye concentration at equilibrium, W (g) is the mass of adsorbent added to the BB-41 solution, and V (L) is the volume of the dye solution.

Regeneration Process

The spent materials were regenerated as reported previously (Kooli et al. Reference Kooli, Yan, Al-Faze and Al-Sehimi2015b). The ZX, ZA, or HS samples were added separately to 10 mL of BB-41 solution (200 ppm) overnight. Each solid was collected by centrifugation and was then dispersed into 10 mL of Co(NO₃)₂ ·6H₂O aqueous solution. The Co²⁺cations served as a homogeneous catalyst. The oxidant, 10 mg of oxone (2KHSO₅ ·KHSO₄ ·K₂SO₄) was added to the mixture to degrade the adsorbed BB-41, via sulfate radical oxidation. The mixture was stirred for 30 min, separated by centrifugation, washed four or five times with deionized water, and then recycled for the next run. The same oxone and Co(NO₃)₂ aqueous solution was used throughout the recycle runs.

Characterization

The identification of mineralogical phases comprising the raw kaolinite and the zeolite products was achieved using XRD patterns from a Bruker AXS D4 ENDEVOR diffractometer (Hannover, Germany) with Ni-filtered CuKα radiation at 40 kV and 40 mA. The powder samples were mounted on the sample holder with very little pressure so as to minimize preferred orientation of particles in the sample. The measurements were carried out with a step width of 0.03°2θ and a measuring time of 1 s per step. The powder data were evaluated with the Stoe WinXPOW software package. Fourier-transform infrared spectroscopy (FTIR) measurements were performed using a Bruker IFS66v FTIR spectrometer in the 4000–370 cm⁻¹ region with samples as KBr pellets. A JEOL JSM-6390A scanning electron microscope (SEM) equipped with a field electron gun (20 kV) was used to study the morphology of the samples. The textural characteristics, such as specific surface area (S _BET), total pore volume (TPV), and average pore diameter of the different samples, were determined from nitrogen adsorption isotherms using a Micromeritics (Norcross, Georgia, USA) ASAP 2040. All samples were degassed at 150°C overnight prior to measurement. The concentrations of the BB-41 dye after treatment with the kaolinite and zeolites were measured by UV-Visible spectroscopy at 610 nm.

X-ray Diffraction

The powder XRD data indicated that the kaolinite exhibited characteristic reflections at 12.35°2θ (001) and 24.64°2θ (002) with additional ones related to quartz (Fig. 3a), and the H-S material showed several common reflections at 14.10 (110), 19.98 (200), 24.54 (211), 31.85 (310), and 34.99°2θ (222) (Fig. 3b), as reported in previous studies (Lee et al. Reference Lee, Han, Park, Park and Choy2003; Gougazeh & Buhl Reference Gougazeh and Buhl2014). The products synthesized products (ZA and ZX) exhibited high crystallinity and purity, as no impurities were detected. ZX exhibited several common reflections detected at 2θ values of 6.10^o (111), 9.99^o (220), 11.74° (311), 15.45° (331), 18.44° (511), 20.10° (440), 23.32° (537), 26.70° (642), 30.47° (660), 31.15° (751), 32.03° (840), and 33.62° (864) (Fig. 3c). ZA exhibited characteristic reflections located at 7.14° (100), 10.10° (110), 12.38° (111), 16.38° (210), 21.58° (300), 23.92° (311), 27.00° (321), 29.82° (410), and 34.08° (332) °2θ (Fig. 3d; Gougazeh & Buhl Reference Gougazeh and Buhl2014).

Fig. 3 Powder XRD patterns of the samples, including (a) Jordanian kaolinite (JK), (b) synthesized hydroxysodalite (HS), (c) synthesized zeolite X (ZX), and (d) synthesized zeolite A (ZA)

FTIR Spectroscopy

The structural changes that occurred during synthesis were monitored by FTIR spectroscopy (Fig. 4). The IR spectra show the prominent peaks of raw kaolinite (Fig. 3a): 3620–3700 cm⁻¹ for -OH stretching, 1000–1120 cm⁻¹ for Si–O stretching, 910–940 cm⁻¹ for -OH deformation modes, and 400–550 cm⁻¹ for Si–O–Si bending vibrations (Farmer Reference Farmer, Olphen and Fripiat1979; Gougazeh & Buhl Reference Gougazeh and Buhl2010). The doublet between 3694 and 3619 cm⁻¹ is indicative of a well ordered kaolinite structure (Farmer Reference Farmer, Olphen and Fripiat1979; Russell Reference Russell and Wilson1987; Gougazeh & Buhl Reference Gougazeh and Buhl2010; Pentrak et al. Reference Pentrak, Madejová and Komadel2009, Reference Pentrak, Madejová, Andrejkovičová, Uhlík and Komadel2012). For the synthesized products (Fig. 4b, c, d), the bands with maxima at 3435 and 1655 cm⁻¹ were characteristic of vibrations of functional groups of the OH type and were ascribed to water molecules in the zeolitic materials. Similar behavior was observed in previous studies on the synthesis of crystalline zeolites HS, X, and A (Flanigen et al. Reference Flanigen, Khatami and Szymanski1971; Baren et al. 1999a; Alkan et al. Reference Alkan, Hopa, Yilmaz and Guler2005; Garrido-Pedrosa et al. Reference Garrido Pedrosa, Souza, Melo and Araújo2006), while obtaining the highly crystalline zeolites HS, X, and A.

Fig. 4 FTIR spectra of the samples, including (a) Jordanian kaolinite, (b) synthesized zeolite X (ZX), (c) synthesized hydroxysodalite (HS), and (d) synthesized zeolite A (ZA)

The strong signal at ~1000 cm⁻¹ characterized the asymmetrical T–O–T valence vibrations (T–Si, Al) of the aluminosilicate zeolite A framework, and the band at ~550 cm⁻¹ represented the double ring vibration of the structure (Flanigen et al. Reference Flanigen, Khatami and Szymanski1971). The bands at ~664 cm⁻¹ (symmetric stretching mode) and 460 cm⁻¹ (T–O bending mode) corresponded to the internal vibrations of the TO₄ tetrahedra (Fig. 3b, c, d). The broader signals at ~550 cm⁻¹ and 1000 cm⁻¹ of the reaction products characterized silica gel in the mixture with zeolite A, and were another indication of the formation of an amorphous material from the quartz content of the kaolinite (see the XRD paragraph). The additional vibrations in the 720–660 cm⁻¹ range were in good agreement with the symmetric stretching modes of hydroxysodalite (729, 701, and 660 cm⁻¹) reported by Flanigen et al. (Reference Flanigen, Khatami and Szymanski1971) (Fig. 3c, d). In the same manner, the band near 420 cm⁻¹ and the intensive mode at 464 cm⁻¹ also indicated the presence of hydroxysodalite in this sample (Fig. 3c); these findings were in good agreement with the reported values of the T–O bending modes of the hydroxysodalite framework structure (Flanigen et al. Reference Flanigen, Khatami and Szymanski1971). The detailed FTIR assignments for hydroxysodalite were summarized by Barnes et al. (Reference Barnes, Addai-Mensah and Gerson1999) and later by Zhao et al. (Reference Zhao, Deng, Harsh, Flury and Boyle2004). The broad bands at ~3435 cm⁻¹ and 1655 cm⁻¹ were attributed to zeolite water (Fig. 3c, d). The synthesized zeolite X showed adsorption bands at 972, 560, and 460 cm⁻¹ related to external linkage of the zeolite structure (Somerset et al. Reference Somerset, Petrik, White, Klink, Key and Iwuoha2005). Furthermore, a strong medium band at ~1650 cm⁻¹was attributed to the H₂O deformation mode due to incomplete dehydration of the zeolite samples (Fig. 4b). Moreover, the observed single strong band at 3435 cm⁻¹ was ascribed to the presence of hydroxyl groups in the faujasite (FAU) supercages (Garrido Pedrosa et al. Reference Garrido Pedrosa, Souza, Melo and Araújo2006). Ultimately, the intensities of these adsorption bands were proportionate to the purity of the samples, and were in agreement with the interpretation of the XRD results.

The band at 458 cm⁻¹ was close to the band at 456 cm⁻¹ (bending vibrations of the TO₄) of FAU zeolite. The bands at 560 and 972 cm⁻¹ corresponded to the 6-membered double-ring vibration: to the symmetric stretching and to the asymmetric stretching, respectively. The band at 557 cm⁻¹ could represent the beginning of the crystallization of a zeolite with double rings (Gougazeh & Buhl Reference Gougazeh and Buhl2014).

Scanning Electron Microscopy

SEM photomicrographs were utilized to determine the microscale structure and morphology of the raw kaolinite and zeolite samples. The progress of the zeolitization reactions was observed through changes in the morphology of the starting material and synthesized products. The SEM photomicrograph of raw kaolinite consisted predominantly of the plate-like morphology and pseudo-hexagonal crystals characteristic of kaolinite (Gougazeh & Buhl Reference Gougazeh and Buhl2014; Fig. 5a). The occurrences of hydroxysodalite (HS) crystals growing at the surface of zeolite A and spherical morphologies corresponding to HS associated with cubic crystals of zeolite A can be clearly identified (Gougazeh & Buhl Reference Gougazeh and Buhl2014; Fig. 5b). The morphology of the synthesized samples can be observed in the images of zeolite A, with characteristic cubic morphology (Fig. 5c). In contrast, zeolite X exhibited well defined crystals (Fig. 5d).

Fig. 5 SEM images of: (a) of pseudohexagonal crystals of Jordanian kaolinite; (b) spherical (spheroidal) agglomerates of hydroxysodalite grown on the surface of well developed cubes of zeolite A; (c) very well defined cubes of zeolite A; and (d) well defined crystals of zeolite X

Adsorption Isotherms

The capacity of raw kaolinite to remove BB-41 was very small and did not exceed 6 mg/g. However, the zeolitic materials exhibited larger removal percentages and capacities (40.9 mg/g for HS, 27.3 mg/g for ZA, and 16.9 mg/g for ZX). For all the samples, the removal percentage of BB-41 decreased as the initial concentration of BB-41 increased. However, the removal capacity of the different zeolites increased as the initial concentrations did (Fig. 6). Similar results were reported using different adsorbents (Adeyemo et al. Reference Adeyemo, Adeoye and Bello2015).

Fig. 6 Basic Blue 41 adsorption isotherms for Jordanian kaolinite, hydroxysodalite, zeolite A, and zeolite X

The experimental adsorption data were subjected to Langmuir and Freundlich adsorption isotherm models (Freundlich Reference Freundlich1906; Langmuir Reference Langmuir1916) to determine the adsorption capacity of the zeolitic materials, and to develop an equation that could be used for design purposes (Ismadji & Bhatia Reference Ismadji and Bhatia2000). The Langmuir isotherm represents the equilibrium distribution of dye molecules between the solid and liquid phases. The linearized form allows the calculation of the maximum adsorption capacity (q _max) and the Langmuir constant (K_L), and is given by eq. 3:

(3) $\frac{C_e}{q_e}=\frac{1}{q_{\max }.K_L}+\frac{C_e}{q_{\max }}$

where q _e is the amount adsorbed at equilibrium (mg/g), C _e is equilibrium concentration of the adsorbate (mg/L), q _max is the maximum amount of dye removed (mg/g), and K_L is related to the energy of adsorption (Langmuir constant, L/mg). These parameters can be calculated from the intercept and slope of the linear plot for the experimental data of C _e/q _e vs. C _e. The isotherms examined fitted well with this model, with R² values close to 1 for all of the samples, except for the raw kaolinite. The largest calculated values for q _max (Table 1) (the maximum amount or monolayer capacity of BB-41 removed) was for hydroxysodalite (close to 39 mg/g), the smallest value was for the ZX sample (17.69 mg/g), and for ZA the value was intermediate at 26.9 mg/g. The K_L values indicate the affinity of BB-41 molecules for the zeolite surfaces, and it increased as the q _max values increased.

Table 1 Langmuir parameters for the removal of BB-41 by the samples studied

RK, Raw kaolinite; HS, Hydroxysodalite; ZA, Zeolite A; ZX, Zeolite X

The Freundlich isotherm is an empirical equation which assumes that the adsorption process occurs on heterogeneous surfaces, with a non-uniform distribution of heat of adsorption over the surface (Freundlich Reference Freundlich1906). The amount of BB-41 adsorbed (q _e) increased curvilinearly with increasing equilibrium concentration (C _e) but the data did not fit the linear form very well, expressed as

(4) $ln{q}_e=ln{K}_F+\frac{1}{n}ln{C}_e$

where K_F (mg/g) and n are the Freundlich constants related to the adsorption capacity, and adsorption intensity of adsorbents, respectively. Values of n in the range 0 < 1/n < 1 represent favorable adsorption conditions. If a plot of lnq _e vs. lnC _e is linear, values of the intercept K_F and the slope 1/n can be obtained. Because the experimental data did not fit the Freundlich equation well for ZA and ZX samples, the R² values obtained were far from 1 (Table 2).

Table 2 Freundlich parameters for the removal of BB-41 by the samples studied

RK, Raw kaolinite; HS, Hydroxysodalite; ZA, Zeolite A; ZX, Zeolite X

Adsorbents such as zeolites consist primarily of silicon and aluminum oxides, the hydroxylated surface of which developed negative charges in aqueous solution (Bertolini et al. Reference Bertolini, Izidoro, Magdalena and Fungaro2013). The removal of BB-41 occurred through interaction forces between the negatively charged sites of the adsorbent and the positively charged dye molecules. The removal might have been achieved by cation exchange, when the dye molecule size was appropriate for accessing the pores. The BB-41 molecule is considered a large molecule, however, and could not access the pores of the prepared zeolites (Selim et al. Reference Selim, El-Makkawi and Ibrahim2018). Indeed, the primary crystalline spherical cavities of a-cages in zeolite A and supercages in zeolite X are 11.4 and 12.6 Å, respectively (Job Reference Job and Fahrner2005). Meanwhile, zeolite materials generally display high affinity for cationic dyes with very small molecular size, such as methylene blue (Awala et al. Reference Awala, Leite, Saint-Marcel, Clet, Retoux, Naydenova and Mintova2016).

A comparison of the maximum removal capacity of BB-41 by various zeolites with similar aluminosilicate structures (Table 3) revealed that the hydroxysodalite prepared has a maximum removal capacity of 39 mg/g, larger than the reported value of 27 mg/g (Selim et al. Reference Selim, El-Makkawi and Ibrahim2018). This difference could be related to the accessibility of removal sites within the structures of the various zeolite materials or to the specific surface areas of the zeolites used.

Table 3 Comparison of BB-41 removal by different adsorbents

RK, Raw kaolinite; HS, Hydroxysodalite; ZA, Zeolite A; ZX, Zeolite X

The well packed structure of kaolinite makes it difficult for the particles to be broken down, and the layers are not easily separated, so the adsorption properties of kaolinite are likely determined by the edges of the layers (Miranda-Trevinol & Coles Reference Miranda-Trevinol and Coles2003).The CEC is confined primarily to the external surface, in contrast to smectites where most of the CEC belongs to interior sites (Miranda-Trevinol & Coles Reference Miranda-Trevinol and Coles2003). Kaolinite, consequently, removed less BB-41 than smectite. In addition, the values obtained were smaller than those reported for some raw clay minerals and acid-activated derivatives (Roulia & Vassiliadis Reference Roulia and Vassiliadis2005; Kooli et al. Reference Kooli, Yan, Al-Faze and Al-Sehimi2015a). This difference could be related to the specific surface area or to the accessibility of the adsorbent sites for the BB-41 dye molecules. Indeed, the textural properties of the starting raw kaolinite (RK) and the derived zeolites revealed that the zeolite X and zeolite A exhibited larger surface areas (160 m²/g and 120 m²/g, respectively) than hydroxysodalite (26 m²/g, Table 4). However, small amounts of BB-41 were removed. In this case, the small removal capacities (of BB-41) using both ZX and ZA were probably due to its large size (Selim et al. Reference Selim, El-Makkawi and Ibrahim2018).

Table 4 Textural properties of raw kaolinite and synthesized zeolites

PV: pore volume; APD: average pore diameter; RK: Raw kaolinite; HS: Hydroxysodalite; ZA: Zeolite A; ZX: Zeolite X

Regeneration Data

The feasibility of applying the adsorbent systems to large-scale operations is determined by either disposal costs of spent adsorbents or regeneration costs. The commonly reported regeneration methods are thermal treatment (Tamon & Okazaki Reference Tamon and Okazaki1997: Wang et al. Reference Wang, Li, Xie, Liu and Xu2006; Vimonses et al. Reference Vimonses, Jim, Chow and Saint2009), chemical and solvent regeneration (Martin & Ng Reference Martin and Ng1978), electrochemical regeneration (Narbaitz & Karimi-Jashni Reference Narbaitz and Karimi-Jashni2008), ultrasonic regeneration (Lim & Okada Reference Lim and Okada2005), and wet air oxidation (Shende & Mahajani Reference Shende and Mahajani2002). An economical method was used to regenerate the spent zeolites, consisting of treating the spent zeolites with a solution in which Co²⁺ cations served as an homogeneous catalyst and the oxidant oxone was used to degrade the BB-41 removed (Anipsitakis et al. Reference Anipsitakis, Dionysiou and Gonzalez2006; Kooli et al. Reference Kooli, Yan, Al-Faze and Al-Sehimi2015a).

The regeneration of the spent zeolites (Fig. 7) revealed that the removal efficiency was preserved, but decreased slightly as the number of cycles increased from 1 to 4. A reduction of 10% was observed compared to the fresh zeolite, for hydroxysodalite. The regeneration efficiencies decreased by 30% after five cycles, and reached a lower value of 60% after seven cycles. Meanwhile, the efficiency of spent ZX and ZA was reduced after three cycles by an average of 25%. The reduction of efficiency might be related to deactivation of removal sites, and indicates that the BB-41 could not be released easily from the spent materials after many cycles, especially for ZX and ZA zeolites (Kooli et al. Reference Kooli, Yan, Al-Faze and Al-Sehimi2015b).

Fig. 7 Regeneration efficiency for (a) zeolite A, (b) zeolite X, and (c) hydroxysodalite after successive regeneration cycles