Synergistic modification of bentonite by acid activation and hydroxyl iron pillaring for enhanced dye adsorption capacity

Nowadays, dyes are widely used in many fields, such as textile, paper, plastic, leather, guest-host liquid crystal displays, solar cells, food and mineral processing industries. However, the excessive discharge of dyes will hinder the photo penetration of water, inhibit the photosynthesis of aquatic plants, the growth of biota, and thus destroy the ecological balance (Han et al. 2009). Furthermore, many non-biodegradable dyes in water will be harmful to human health and even cause damage to central nervous and reproductive systems due to their high toxicity and bioaccumulation. For example, the typical dye rhodamine B (RhB) is well known as a food additive and sometimes fluorescent water tracer, but its large harmful effects on the human body and ecosystems have forced people to use it with care (Marnani & Shahbazi 2019).

Some methods have already been developed for removing dyes, including precipitation, flocculation, coagulation, incineration, ion exchange, biological degradation and adsorption (Singh et al. 2018). Among them, adsorption is very competitive for separating pollutants from water bodies with the advantages of easy operation, relatively low cost and recyclability. Bentonite has been considered to be a promising adsorbent because of its lamellar structure, high cation exchange capacity, and low price with resource abundance (Pereira et al. 2017). Many researchers have tried to modify bentonite to enhance the adsorption capacity for cationic dyes (Selvam et al. 2008; Farhan et al. 2014). For example, Javed et al. (2018) prepared activated bentonite using different organic and inorganic acids and conducted batch experiments to investigate its adsorption removal efficiency for Mordant Red 73. The results show that acid activation leached out the Ca²⁺ and Mg²⁺ ions and increased the porosity of bentonite, thus elevating the adsorption capacity to 149 mg/g. Chinoune et al. (2016) prepared Mg(OH)₂ coated bentonite and found it is efficient for adsorption treatment of reactive blue 2 and reactive blue 19. The iron-pillared bentonite was also prepared by ion exchange using natural bentonite, with which the adsorption capacity of 98.6 mg/g for RhB was obtained.

In this paper, a new facile modification strategy to enhance the adsorption performance of bentonite utilizing synergistic acid activation and hydroxyl iron pillaring was conducted. Herein, different adsorbents were obtained by various combinations of the modification methods, whose adsorption capacities were investigated and compared for a typical organic dye, RhB. To explore the adsorption mechanism of these adsorbents for RhB, the kinetic models and adsorption isotherms were also simulated. Overall, the enhanced adsorption capacity of the modified adsorbents for RhB was validated, suggesting the present strategy is promising for structural modification of dye adsorbents.

The pristine sodium BT powder was obtained from the Hebei Lingshou Dehang Mine products Co., Ltd, China. All the reagent grade chemicals, such as RhB, H₂SO₄, Fe(NO₃)·9H₂O and Na₂CO₃, were purchased from Sinopharm Chemical Reagent Co. Ltd, China, and used without any further purification.

Herein, 8 g BT powder was added into 50 mL H₂SO₄ solution with a concentration of 1 mol/L, and stirred at a speed of 200 rpm for 24 hours. After that, the mixture was centrifuged at a speed of 8,000 rpm for 8 min and washed with distilled water until the pH reached 6–7. Finally, the mixture was dried in an oven at 60 °C for 12 hours and passed through a 200-mesh sieve to obtain the final powder, labelled as S-BT.

The mixture was stirred for 2 hours at room temperature and aged in a water bath at a temperature of 60 °C for 24 hours. By adjusting the amount of Na₂CO₃, iron pillaring liquid with different ratios of OH⁻ to Fe³⁺(OH/Fe) was obtained.

The BT or S-BT powder, 6 g in weight, was stirred at 200 rpm in distilled water at 60 °C for 24 hours, and poured into the hydroxyl iron pillaring liquid, in which the molar ratio of Fe to bentonite was fixed at 100 mmol/g. This mixed solution was also stirred at 200 rpm and 60 °C for 2 hours, and aged for 24 hours at room temperature. The supernatant was centrifuged at 8,000 rpm for 8 min and washed with distilled water three times. Finally, the modified bentonite was dried at 60 °C for 12 hours, ground and passed through a 200-mesh sieve.

For comparison, different adsorbents were also obtained by combining and adjusting the modification order of sulphuric acid activation and hydroxyl iron pillaring. For convenience, the abbreviations for the adsorbents used in this study are listed as follows. Schematic illustration for the fabrication process of S-Fe-BT is shown in Figure 1.

BT: Pristine bentonite

S-BT: Sulphuric acid-activated BT

Fe-BT: Hydroxyl iron pillared BT

Fe-S-BT: (a) Sulphuric acid activation + (b) hydroxyl iron pillaring for BT

S-Fe-BT: (b) Hydroxyl iron pillaring + (a) sulphuric acid activation for BT

The X-ray diffraction patterns (XRD) analysis was conducted using a Rigaku 2000 automated diffractometer with CuKα radiation from 5° to 90° at a rate of 5° min⁻¹. Fourier transform infrared (FTIR) spectra were taken on a Nexus 410 spectrometer (Nicolet, USA) and scanned in the wavelength range of 4,000–400 cm⁻¹. The microstructural morphologies were observed by field emission scanning electron microscopy (FESEM, JSM-7600F, JEOL Ltd, Japan). Specific surface area was determined by using the Brunauer-Emmett-Teller (BET) apparatus (Quantachrome Noca Win, USA). The thermogravimetric (TG) analysis was conducted by a STA449C apparatus (Netzsch, Germany) from room temperature to 800 °C at a rate of 10 °C min⁻¹ in a nitrogen atmosphere.

The adsorption experiments were carried out by a batch method with various RhB concentrations, adsorbent dosages, temperatures and pH. The experiment was performed in 250 mL glass flasks, in which 200 mL RhB solution of different concentrations and 0.2 g adsorbent were added. The solution in the glass flask was agitated by a stirring rod at room temperature at 200 rpm for 5 hours to reach adsorption equilibrium. To explore the effect of initial pH value, 1 mol/L NaOH or 1 mol/L HCl was added into the solution in order to reach the initial expected pH.

At pre-determined time intervals, the used adsorbents were removed from the solution by centrifugation. Then the RhB concentration in the solution was determined by measuring the absorbance of the solution at 554 nm (λ_max for RhB), using a TU1901 UV-Vis spectrophotometer (Purkinje General Instrument Co. Ltd, China). Batch experiments were conducted in triplicate to ensure repeatability and the average values are presented here.

To determine the interactive behavior between the adsorbent and the adsorbate, the adsorption kinetics and isotherm curves were also fitted for investigation, whose detailed experimental procedure and related models can be found in the Supplementary information file.

The XRD spectra (Figure 2(a)) show that the pristine BT is mainly composed of montmorillonite, quartz and feldspar (Li et al. 2015). After modification, the (001) peaks of montmorillonite were found to apparently shift lower, and the d(001) spacing value of S-BT, Fe-BT, S-Fe-BT and Fe-S-BT was calculated to be 1.484 nm, 1.506 nm, 1.543 nm and 1.543 nm, respectively, all higher than that of BT (1.406 nm), indicating the hydroxyl iron was successfully intercalated into the BT layer. More importantly, FTIR spectra were employed to further confirm the structural differences of the BT and S-Fe-BT (Figure 2(b)). Comparing the two spectra of the modified bentonite and pristine bentonite, they showed almost the same peak shape, which indicated that the basic skeleton of bentonite had no obvious change during the modification process. The absorption bands at 3,460 cm⁻¹ (H-O-H stretching) and 1,638 cm⁻¹ (H-O-H bending) can be ascribed to the water molecules (Hou et al. 2011; Wu et al. 2014). The three absorption bands at 1,034, 793 and 518 cm⁻¹ belong to Si-O stretching vibrations (Hajjaji & El Arfaoui 2009). The characteristic band at 1,383 cm⁻¹ reveals the existence of due to the adoption of Fe(NO₃)₃ (Hou et al. 2011). Two small new peaks were observed after the modification of bentonite at 539 cm⁻¹ and 470 cm⁻¹, which correspond to the vibration of Si-O-Fe and Fe-O bonds (Nie et al. 2009). The change of the spectrum after modification proved that Fe-OH molecules were inserted into the bentonite layers, which is important for RhB adsorption.

Figure 3 shows the N₂ adsorption-desorption isotherm plots for (a) BT and (c) S-Fe-BT, and the corresponding pore size distribution plots ((b) for BT and (d) for S-Fe-BT). The type IV isotherms are observed with apparent hysteresis loops in the P/Po range of 0.4–1.0, indicating the presence of mesopores (Liu et al. 2014). The BET specific surface area and total pore volume of S-Fe-BT are both higher than those of BT. In addition, it should also be noticed that the synthetic modification treatment also increased the average pore size. Such a microstructural altering is therefore beneficial for enhanced adsorption performance of BT, which will be validated in the following adsorption testing results (Ranđelović et al. 2014).

The TG-DSC curves of BT and S-Fe-BT are shown in Figure S1(a) and S1(b) (Supplementary information). The first endothermic peak of BT is 112.6 °C, which is attributed to the dehydration of bentonite. The first and second endothermic peaks of S-Fe-BT are 94.3 °C and 130.3 °C, which is attributed to the loss of surface water and interlaminar water of bentonite (Chinoune et al. 2016), and the first stage weight loss of BT and S-Fe-BT is determined to be 2.54% and 6.59% respectively. The second endothermic peak at 681.3 °C, with a weight loss of 4.36%, is ascribed to the dehydroxyl reaction of aluminum silicate in BT. However, this completely disappeared for S-Fe-BT due to the enhanced thermal stability after modification.

Figure 4 shows the SEM images of (a) BT, (b) S-BT, (c) Fe-BT, (d) S-Fe-BT and (e) Fe-S-BT, respectively. The pristine BT exhibits a lamellar microstructure with a rough and uneven surface (Figure 4(a)). After acid activation, the pores of BT grow larger and the surface turns smooth (Figure 4(b)). Due to hydroxyl iron intercalation, the spacing of the BT layer (montmorillonite) increases obviously, and the layered structure shows curly morphology (Figure 4(c)). After the synthetic acid activation and Fe intercalation, some of the lamellae of bentonite peel off, and the pores increase obviously (Figure 4(d) and 4(e)). The OH-Fe particles are well dispersed on the surface of the bentonite with no agglomeration. The mapping of S-Fe-BT is supplied in Figure S2, indicating that the Fe ions are uniformly intercalated into the interlayers of BT. Furthermore, the element contents of BT and S-Fe-BT is also shown in the table embedded in Figure S2(f), where the Fe element was identified due to the successful intercalation of the Fe ion into the bentonite interlayer.

To investigate the effect of OH/Fe on the adsorption capacity of Fe-BT, RhB solutions (200 mL) with different concentrations of 50 and 100 mg/L were prepared, in which 0.2 g Fe-BT powders were used for adsorption tests. Figure 5(a) and 5(c) indicate that the initial adsorption process happened very fast, and slowed down after 10 hours. Figure 5(b) shows the final removal efficiency of RhB reached higher than 99% for all the Fe-BT powders with a ratio of OH/Fe from 0.05 to 1.0 when the RhB concentration was 50 mg/L. However, when the RhB concentration increased to 100 mg/L (Figure 5(d)), the highest removal efficiency of only 95.7% was obtained at a molar ratio of 0.2 for OH/Fe. A molar ratio less than 0.2 leads to insufficient active adsorption sites, while more than 0.2 results in sites stacking and decreasing removal efficiency.

To compare with adsorption capacity of modified BT with pristine BT, 200 mL RhB solution was prepared with different concentrations from 50 to 300 mg/L, after which 0.2 g adsorbent was added to test the adsorption effect, as shown in Figure 6(a). The adsorption capacity of the pristine BT for RhB was very limited and the removal efficiency was less than 30%. Compared with BT, the adsorption capacity of four modified BT adsorbents was apparently enhanced. When the RhB concentration was 50 mg/L, the removal efficiencies of the four kinds of modified bentonite were all close to 100%. With increasing RhB concentration, the adsorption capacity of these adsorbents decreased, whereas the S-Fe-BT exhibited the highest value. Its adsorption capacity increased from 99.6 mg/g to 246.1 mg/g, after which it declined to 237.8 mg/g with the initial concentration of RhB increased from 100 to 600 mg/L, while the RhB removal efficiency declined from 99.6% to 39.6% due to the saturated active adsorption sites (Figure 6(b)).

To go further, the effects of some typical experimental parameters on RhB removal by S-Fe-BT were investigated. With increasing S-Fe-BT dosage from 0.025 to 0.2 g, an initial rapid increase in color removal from 14.98% to 99.20% was detected (Figure 7(a)), which may also be ascribed to the availability of more adsorption sites. As the adsorbent dosage increased further (>0.2 g), the color removal remained almost the same, indicating the splitting effect of flux (concentration gradient) occurred between the adsorbent and RhB (Zhang et al. 2012). Figure 7(b) shows the effect of adsorption time on the removal efficiency of RhB at 200 mg/L. At the first hour, the color removal increased sharply, after which it gradually slowed down, and approached equilibrium at about 3 hours due to the slow intraparticle diffusion into the interlayer of S-Fe-BT. After 3 hours, a dynamic equilibrium was reached between the adsorption and desorption of the dye.

Figure 7(c) shows that the removal of RhB dye increased with decreasing pH value. As shown in Figure S3 of the Supplementary information file, RhB is a kind of xanthene dye with three forms of protonated species with different charges (Hou et al. 2011) of zwitterion (RhB^±) and two positively charged species: RhBH⁺ and ⁠. Below pH 3.5, RhB molecules can easily go into adsorbent pores because of its monomeric form. Above pH 3.5, however, RhB in water becomes its zwitterionic form, which is easily aggregated to larger molecules. Therefore, the transport of the molecule inside the pores was hindered, thus reducing the adsorption capacity of the adsorbent. The basic form and lactone form of RhB are predominant in the solution at higher pH values, unfavorable for their adsorption onto adsorbents. In view of the actual situation of the printing and dyeing wastewater and the prevention of acid pollution of the water body, the subsequent experiment still used the neutral dye solution.

With increasing temperature from 20 to 60 °C, the color removal ratio also increased from 89.8% to 98.7% (Figure 7(d)). At lower temperature, the decrease of adsorption capacity may be due to the weakening of the physical interaction between dyes and adsorbents. Since there is only a slight change in the color removal with temperature, it is more practicable to adsorb the dye at room temperature, indicating that the as-obtained S-Fe-BT adsorbent is promising for practical applications.

So far, there are many adsorbents for adsorption of RhB as listed in Table 1. It can be seen that the adsorption capacity of mineral materials is less than 50 mg/g (Khan et al. 2012; Bhattacharyya et al. 2014; Cai et al. 2014). The fabrication of waste-based adsorbents is beneficial for secondary resource utilization, but their adsorption capacity still needs to be improved (Kadirvelu et al. 2005; Sureshkumar & Namasivayam 2008). New adsorption materials, such as nanomaterials (Qin et al. 2018; Thi Phuong Minh et al. 2019) and polymer materials (Huang et al. 2008; Wang et al. 2015; Saleh & Islam 2018), have received extensive attention. Bentonite with different modifications has also been investigated (Selvam et al. 2008; Hou et al. 2011). Obviously, in comparison with those reported adsorbents in literature of RhB, the S-Fe-BT adsorbent as a low cost and rapid adsorbent shows comparable adsorption capacity, which is desirable for the treatment and purification of wastewater samples containing similar organic dye.

To determine the interactive behavior between the adsorbent and the adsorbate, we analyzed adsorption isotherm models, including the Langmuir, Freundlich and Temkin models (Supplementary information S1).

As can be seen from Figure S4 and Table S1, the regression coefficient (R²) of RhB adsorption isotherms on to S-Fe-BT obtained from the Langmuir isotherm under different temperatures is higher (R² = 0.9834, 0.7729 and 0.8720, respectively) than those from the Freundlich and Temkin isotherms, while the root mean square (RMS) is lower. This indicates that the Langmuir isotherm is a better fit and that adsorption phenomenon is in monolayer. The separation factor (R_L) values are 0.0030–0.0750 for S-Fe-BT, suggesting that the adsorption of RhB on S-Fe-BT is a favorable adsorption process (Muinde et al. 2017; Venkatesan & Narayanan 2018).

In order to understand the mechanism of the adsorption kinetic process, the experimental data were fitted by the following three kinetic models (Supplementary information S2), and the results are shown in Figure S5. Tables S2 and S3 summarize the pseudo-first-order (Figure S5(a) and S5(b)), pseudo-second-order (Figure S5(c) and S5(d)) and intraparticle diffusion kinetics (Figure S5(e) and S5(f)) constants obtained from a linear regression analysis. For the pseudo-second-order kinetics model, high correlation coefficients (R² > 0.999) were obtained with low RMS (<0.0002) for all RhB concentrations and temperatures. It is well known that the pseudo second model is based on the rate control step of chemical reaction or chemisorption through electron sharing or electron gain and loss, so we consider that the adsorption process was primarily controlled by chemisorption. Moreover, the experimental data matched well with the theoretical q_e values, indicating that this adsorption process conformed to pseudo-second-order kinetics. When the initial RhB concentration increased, the q_e values increased while the rate constants (k₂) decreased. This could be attributed to the relationship between the competition of adsorption sites and RhB concentration. At lower concentrations, less competition occurred for the sorption surface sites, while at higher concentrations, the competition for the active sites between RhB molecules was high, and consequently lower k₂ values were obtained. The rise of temperature was accompanied by the rise of k₂, indicating that the reaction rate was accelerated at higher temperatures.

The Weber intraparticle diffusion model was also employed to identify the steps that occurred during the adsorption process. The linear plots of various initial concentrations did not pass through the origin point (Figure S5(e) and S5(f)), suggesting that the adsorption rate was not controlled by intraparticle diffusion. The intercept value, C, was used to represent the thickness of the boundary layer and played a significant role in the removal of the dye. Generally, the larger the C value is, the more significant the boundary layer effect is. The C values increased from 84.6 to 134.2 mg/g with increasing initial dye concentration from 100 to 250 mg/L (Table S3), suggesting a boundary layer diffusion effect. From above, it can be concluded that pseudo-second-order and intraparticle diffusion took place simultaneously, which effectively enhanced the adsorption capacity of S-Fe-BT.

As a cationic dye, the RhB molecules are positively charged in neutral and alkaline solution, which can produce electrostatic attraction with the negatively charged bentonite layer. However, the protons on the hydroxyl groups of RhB molecules can be ionized under acidic conditions, which leads the dye molecules to be negatively charged and cause electrostatic attraction with the Fe³⁺ of S-Fe-BT. Furthermore, the adsorption isotherms fitted the Langmuir model, indicating that electrostatic attraction is not the dominant mechanism of adsorption. Besides, hydrogen bonding interaction may occur between RhB and S-Fe-BT (Deng et al. 2019). In more detail, the hydrogen bonding interaction may happen between the hydroxyl group of S-Fe-BT and the oxygen of the xanthenes ring as well as the aromatic ring of RhB molecule (Mohammadi et al. 2010). The interaction could also occur between the hydroxyl groups from the adsorbed H₂O in the interlayer of S-Fe-BT and all of the N and O atoms in the RhB molecule (Yu et al. 2013). In summary, as shown in Figure 8, the adsorption mechanisms of RhB on the surface of the adsorbent are chemical adsorption through hydrogen bond action with monolayer and uniform active sites as well as physical adsorption through electrostatic attraction (De Castro et al. 2018; de Queiroga et al. 2019).

In this paper, a facile modification strategy for BT was proposed by synergistic acid activation and hydroxyl iron pillaring, after which the interlayer spacing was enlarged with increased BET specific surface area and total pore volume. The maximum adsorption capacity of S-Fe-BT was found to be 246.06 mg/g without pH adjustment when the RhB concentration was 400 mg/L. The equilibrium adsorption isotherm followed the Langmuir model better than the Freundlich or Temkin model, suggesting that the adsorption phenomenon is monolayer and the adsorption sites of adsorbents are uniform. The experimental data fits well with the pseudo-second-order adsorption model (R₂² = 0.9999), indicating that the adsorption was controlled by a chemical process. Adsorption mechanisms of RhB on S-Fe-BT were electrostatic attraction and formation of hydrogen bonds. Therefore, the modified BT obtained in this study is highly proposed as a promising adsorbent for the removal of RhB from water environments.

The authors would like to acknowledge the financial supports from National Key R&D Program of China (2018YFC0408003,2018YFC1508704).

The Supplementary Material for this paper is available online at https://dx.doi.org/10.2166/wst.2020.239.