Simultaneous sorption of As, B, Cr, Mo and Se from coal fly ash leachates by Al³⁺-pillared bentonite clay: implication for the construction of activated geo-synthetic clay liner

South Africa relies mainly on coal combustion for power generation (Vadapalli et al. 2010; Somerset et al. 2011; Vadapalli et al. 2012; Gitari et al. 2013c; Muriithi et al. 2014). Low grade bituminous coal mainly used for power generation and it has been documented to generates huge volumes of by-products such as bottom ash, fly ash (FA), boiler slag, flue gas desulfurization sludge and non-captured particles (Adriano et al. 1980; Steenari et al. 1999; Sherly & Kumar 2011; Seshadri et al. 2013; Wang et al. 2014). Amongst the generated waste, of most concern is FA which is mainly collected from flue gases by means of mechanical devices such as electrostatic precipitators, bag house, scrubbers and cyclones (Ahmaruzzaman 2010).

After collection, FA is hydraulically transported to holding ponds, lagoons, landfills and slag heaps where it can be reacquired for treatment purposes or discarded or conveyed to ash heaps in dry disposal systems. The transportation water is collected as penstock overflow and recirculated for more ash slurry uptake in wet disposal systems (Ahmaruzzaman 2010; Mahlaba et al. 2011; Akar et al. 2012; Akinyemi et al. 2012). Approximately 27 MT of FA is produced annually in South African (RSA) power utilities (Gitari et al. 2008). Of this, only 5% is utilized beneficially for construction (Yao et al. 2015), agricultural purposes (Zevenbergen et al. 2000; Basu et al. 2009; Kishor et al. 2010; Yunusa et al. 2014), treatment of wastewater (Gupta & Torres 1998; Wang & Viraraghavan 1998; Ahmaruzzaman 2011; Wang et al. 2013; Visa & Chelaru 2014) and beneficiated into new products such as zeolite (Ríos et al. 2008; Vadapalli et al. 2010; Zhang et al. 2011; Xie et al. 2014; Yao et al. 2015).

The remainder is discarded in surface impoundments such as ash retention dams and ponds where it commence with environmental degradation through leaching to underground and surface water resources thus impairing the aquatic and terrestrial ecosystems (Zevenbergen et al. 1999; Zevenbergen et al. 2000; Abbott et al. 2001; Mahlaba et al. 2011; Akinyemi et al. 2012; Muriithi et al. 2013). Coal FA is a heterogeneous mixture of inorganic species which were incorporated in coal during the coalification stages. Post ash disposal and on contact with aqueous medium, the toxic inorganic pollutants such as Cl^-, SO₄^2-, oxyanion species of As, B, Cr, Mo, Se and Al, Ba, Ca, Cd, Cu, Fe, K, Mg, Na, Sr and Zn are leached to terrestrial and aquatic ecosystems (Reardon et al. 1995; Steenari et al. 1999; Iyer 2002; Praharaj et al. 2002a; Vítková et al. 2009; Zheng et al. 2011; Tsiridis et al. 2012; Vítková et al. 2013).

Literature studies have reported that availability of As, B, Cr, Mo and Se in the biosphere are of ecological concern due to possible toxicity that they pose to living organisms. Epidemiological studies have reported that long-term exposure to oxyanions may interfere with metabolic activities of living organisms, failure of multisystem organ by allosteric inhibition of sulfhydryl enzymes that are essential for regulating metabolic activities hence leading to death as an end point of toxicity. Conjunctivitis, hyperkeratosis, hyper pigmentation, cardiovascular diseases, disorder of central nervous system and peripheral vascular system, skin cancer and gangrene of the limbs are some of the toxicity that are triggered by oxyanions (Kashiwa et al. 2000; Hassett et al. 2005; Tajer Mohammad Ghazvini & Ghorbanzadeh Mashkani 2009; He & Yao 2010; Puranen et al. 2010). Moreover, coal FA leachates can render surface and groundwater unsuitable for numerous defined uses (Iyer 2002; Praharaj et al. 2002b; Lapa et al. 2007; Prasad & Mondal 2009; Izquierdo & Querol 2012; Tsiridis et al. 2012; Neupane & Donahoe 2013; Reinik et al. 2014).

As such, there is a need to come up with a sustainable and pragmatic technology that will be used to prevent the contamination of underground water resources by coal FA leachates. The most prominent technology relies on the use of clay liners and exchange resins beds on the ash retention pond linings (Lange et al. 2005; Mapanda et al. 2007; Lange et al. 2010a, 2010b, Lu et al. 2011; Yang et al. 2012; Fan 2014b, 2014a; Musso et al. 2014). Abundancy, high surface area, cation exchange capacity (CEC) and magnificent swelling capacity of clays made clay and clay minerals to anchor the depollution science (Bhattacharyya & Gupta 2008; Bhattacharya et al. 2013). Clays are widely utilized for remediation of wastewater as geo-synthetic clay liners, however, they are dependent on pH (Dukić et al. 2015). Due to low density of cations on the clay matrices, clays can be modified with polycations to increase their inorganic contaminants retention abilities (Gitari et al. 2013a; Masindi et al. 2014b). Trivalent species of aluminium has strong affinity to oxyanionic species (Gitari et al. 2013b; Masindi 2013; Masindi et al. 2014b). This will reduce the use of virgin material by modifying the clay with species that has high affinity to oxyanionic species. The present study reports the optimum conditions for intercalating the Al³⁺ species to bentonite clay interlayers and evaluate the potential of using Al³⁺-modified bentonite clay for removal of oxyanionic species of As, B, Cr, Mo and Se from coal FA leachates.

Bentonite clay was obtained from ECCA holdings Pty, Ltd and coal FA was collected from Hendrina power utility in Mpumalanga, South Africa. All reagents used in this study were of analytical grade.

For ion exchange reaction, Al₂(SO₄)₃.18H₂O was used. The optimum conditions for loading Al³⁺ polycations onto bentonite clay lattice were observed to be 60 min of retention time, 1 g of adsorbent dosage and 100 mg/l of adsorbate concentration (Masindi 2013). Bentonite clay, 10 g, was transferred into a 1,000 mL High Density Polyethylene (HDPE) container containing 500 ml of 100 mg/l Al³⁺. The mixture was shaken for 60 min on a table shaker. This was repeated 25 times to make up a total of 250 g of adsorbent (Masindi 2013).

Elemental composition was ascertained using X-ray fluorescence (XRF). Surface are by Brunauer–Emmett–Teller (BET: A Tristar II 3020, Micrometrics BET from Norcross, GA, USA). pH_pzc was determined using solid addition method (Masindi et al. 2014a). CEC was determined using ammonium acetate method (Masindi et al. 2014a).

pH, total dissolved solids (TDS – mg/l) and electrical conductivity (μS/cm) were monitored using CRISON MM 40 portable pH/EC/TDS/Temperature multimeter probe. Aqueous samples were analysed using inductively coupled plasma mass spectrometer (ICP-MS) (7500ce, Agilent, Alpharetta, GA, USA). The accuracy of the analysis was monitored by analysis of National Institute of Standards and Technology water standards. Three replicate measurements were made on each sample and results are reported on average.

One thousand mg/l standard solutions of As, B, Cr, Mo and Se from Lab Consumables Supply, Johannesburg, South Africa were used to prepare the combined working solutions. From 1,000 mg/l stock solution, 10 mg/l solution was prepared by 100 time dilution of 1,000 mg/l oxyanion standard solution and transferred into 1,000 ml volumetric flask. The volumetric flask was topped to the mark by adding ultra-pure water.

Adsorption of oxyanions onto Alum-bent was evaluated in batch experimental procedure. Stock solutions of metal ions were prepared in ultra-pure water using the obtained standard solutions. Several operational and optimization parameters were evaluated and they include effects of time, dosage, concentration and pH. All the experiments were carried out using 250 ml HDPE containers and mixed at 250 rpm using Labcon model 3100E table shaker.

Effects of time were investigated by varying time from 1–180 min. Other conditions include: 1:100 S/L ratios, 10 mg/l adsorbate concentration, pH 7.5 and 26°C room temperature. After shaking, the mixtures were filtered through a 0.45 μm pore nitrate cellulose filter membrane and refrigerated at 4°C prior to analysis by an ELAN 6000 ICP-MS. The pH before and after agitation was measured using the CRISON multimeter probe (model MM 40).

Effects of Alum-Bent dosage on adsorption of oxyanions were evaluated by varying the dosages from 0.5–5 g. Other conditions include: 30 min of equilibration, 10 mg/l adsorbate concentration, pH 7.5 and 26°C room temperature. The pH and metal content were measured as described in the preceding section.

Effects of ions concentration on removal of oxyanions by Alum-Bent were investigated by varying concentration ranges from 0.1–15 mg/l. Other conditions include: 30 min of equilibration, 1:100 S/L ratios, pH 7.5 and 26°C room temperature. The pH and metal content were measured as described in the preceding section.

Effects of pH were evaluated by adjusting the pH of aqueous media to required pH using HCl or NaOH at the beginning of batches and not controlled afterwards. Other conditions were 30 min of equilibration, 1:100 S/L ratios, 15 mg/l and 26°C temperatures. Optimum adsorption conditions obtained were used for treatment of coal FA leachates. The pH and metal content were measured as described in the preceding section.

Coal FA leachates were generated at 10 g: 100 ml S/L ratio using ultra-pure water. Three 10 g of coal FA was measured and transferred to individual 250 ml HDPE plastic with 100 ml of ultra-pure water. The mixtures were equilibrated for 60 min on a reciprocating table shaker. After agitation, the leachates were filtered through a 0.45 μm pore nitrate cellulose filter membrane and acidified with two drops of concentrated HNO₃ acid to prevent aging and immediate precipitation of As, B, Cr, Mo and Se ions. The samples were refrigerated at 4°C until analysis by ICP-MS.

100 ml of coal FA leachates was transferred into three 250 ml HDPE container and 4 g of Alum-Bent added. The mixtures were equilibrated for 60 min using a reciprocating orbital shaker. After shaking, the mixture was filtered through 0.45 μm cellulose membrane. The filtrates were acidified with three drops of nitric acid and stored in the refrigerator until analysis for As, B, Cr, Mo and Se using ICP-MS.

This equation is applied to obtain K₂, the second order rate from the plots t vs t/qe.

The mechanisms and intensity of adsorption was described by use of two common adsorption models: Langmuir and Freundlich adsorption isotherms these models describe adsorption processes on a homogenous or heterogeneous surface respectively.

The linear plot of log versus log indicates if the data is described by Freundlich isotherm. The value of implies that the energy of adsorption on a homogeneous surface is independent of surface coverage and n is an adsorption constant which reveals the rate at which adsorption is taking place. These two constants are determined from the slope.

The chemical composition of raw and Al³⁺ modified bentonite clay is shown in Table 1.

Results in Table 1 show that bentonite clay is composed largely of SiO₂ and Al₂O₃ indicating that it's an aluminosilicate material. High levels of Fe₂O₃ could indicate presence of iron oxides which might have been adsorbed during deposition processes. There is insignificant proportion of Mn and P impurities which might have been incorporated in clay fractions during deposition. The increased Al content in the bentonite clay indicate successful introduction of Al³⁺ in the bentonite clay interlayers. The contents of MgO, Na₂O, CaO and K₂O were observed to decrease in the modified bentonite indicating that they are exchanged for Al³⁺.

The physicochemical properties of raw and Al³⁺ modified bentonite clay are presented in Table 2.

Determination of exchangeable base cations by the ammonium acetate method at different pHs (pH 5.4 and 7.4) revealed that Mg²⁺, Ca²⁺, Na⁺ and K⁺ are the exchangeable base cations. This collaborates XRF results which showed a decrease of Ca²⁺, Mg²⁺, K⁺, and Na⁺ in Al³⁺- modified bentonite clay (Table 1). The CEC results show that South African bentonite clay is characterized by high CEC of 262 meq/100 g at pH 5.4 and 265.5 meq/100 g at pH 7.4. The CEC decreased to 186.9 meq/100 g at pH 5.4 and 183.3 meq/100 g at pH 7.4 on introducing Al³⁺. The results show that CEC is independent of pH for the tested pH ranges. Surface area plays a pivotal role in adsorption of chemical species from the solution, the higher the surface area, the higher the adsorption and vice versa (Masindi et al. 2014a). BET results indicated that the loading of Al³⁺ onto bentonite clay interlayers increased the specific surface area from 16 m²/g to 44.3 m²/g. pHpzc is the pH at which a surface has a net charge of zero. There was a slight difference in pH_PZC of Al³⁺ modified bentonite clay (8.2) as compared to raw bentonite (8.0). A slight increase in pH_PZC will also contribute to the enhanced adsorption efficiencies of anions from aqueous media (Gitari et al. 2013b, 2015; Masindi et al. 2015a, 2015c).

Batch experiments were carried out to evaluate the effects of contact time, adsorbent dosage, adsorbate concentration and pH on adsorption of As, B, Cr, Mo and Se on Al³⁺-modified bentonite clay.

Figure 1 shows sorption of As, B, Cr, Mo and Se onto Al³⁺ bentonite clay increases with an increase in contact time to reach a steady state at 15 min. Within 1^st min of contact adsorption of B was very rapid and gradually approached a steady state after 5 minutes. At 30 min the system seems to reach a steady state for all chemical species. Maximum uptake of B was attained in 5 minutes reaching approximately 99.9% removal (Figure 1). Adsorption of other chemical species was below 80% at equilibrium but followed the same adsorption trend as B. This indicates similar adsorption mechanism for all the chemical species. The affinity of the Al³⁺ modified bentonite clay for the investigated chemical species showed the following order: B>Cr>Mo>As>Se. The adsorption capacity trends were similar to the trends of % removal for all oxyanions investigated. The result shows that modified bentonite clay is more effective and fast for all chemical species. This is attributed to higher surface area, micro-pore volume, and surface charge of modified bentonite clay as compared to raw bentonite clay (Table 2). Other reasons may be an increase in density of sorption sites due to introduction of Al³⁺ which has high charge density. The higher adsorption of most of the species by Al³⁺-bentonite clay could also be due to increase in interlayer spacing compared to raw bentonite. The maximum adsorption was observed to occur at 30 min hence this was used as the optimum contact time for subsequent experiments.

The % adsorption of As, B, Cr, Mo and Se was observed to increase with adsorbent dosage for both Al³⁺-modified bentonite clay and raw bentonite clay (Figure 2). Increase of adsorbent dose provides more sites for adsorption hence increase in % adsorption. At 3 g adsorbent dosage >91% adsorption was observed for all chemicals species for the Al³⁺ modified bentonite clay (Figure 2). The adsorption affinity for Al³⁺ modified clay for various species varied as follows: B>Cr>As>Se > Mo. pH was observed to increase in final process water for raw bentonite. This could be due to dissolution of alkaline metal oxides (Table 1). High adsorption capacity of the modified clay is attributed to high surface area, micro-pore volume and surface charge of modified bentonite clay (Table 2). Masindi et al. (2014a) observed that Al³⁺/Fe³⁺ modified bentonite clay had higher adsorption capacity for arsenic species as compared to unmodified bentonite clay. At 4 g adsorbent dosage both systems recorded >40% adsorption for all chemical species, consequently, 4 g was taken as the optimum adsorbent dosage and was applied for subsequent experiments.

Figure 3 shows that, as the concentration of As, B, Cr, Mo and Se increased the adsorption capacity of the modified bentonite clay decreased. The modified bentonite clay was observed to adsorb more than 98% of the chemical species over the concentration range of 0.1–10 mg/l (Figure 3) its adsorption capacity decreased at concentrations greater than 10 mg/l but was still >80%. At low concentration, more adsorption sites are available and adsorption is high but as the concentration increases the adsorption sites become the limiting factor hence resulting in lower adsorption. At concentration range 0.1–1 mg/l, raw bentonite clay recorded % adsorption >98% for all chemical species. As the concentration increased the % adsorption decreased dropping to below 60% at 6 mg/L for As, Se, Mo, and Cr (Figure 3). Physicochemical characterization of the modified bentonite clay showed increased surface area and pore volume (Table 2) which could account for the increased adsorption capacity. The final pH of the media in the Al³⁺-bentonite reactions was observed to be dependent on adsorbate concentration, with 0.1–1 mg/l showing a final pH of 4–5 while 3–15 mg/l solutions having a final pH of 2–3. The results show that 10 mg/l would be the optimum adsorbate concentration at which Al³⁺modified bentonite clay would have optimum adsorption. However 10 mg/l was selected as the optimum adsorbate concentration for the subsequent experiment.

Figure 4 shows the effect of varying the pH on adsorption (% removal) of As, B, Cr, Mo and Se for Al³⁺ bentonite clay. The % adsorption of all the chemical species increases with increasing pH but is maximum at pH 4–10 except for Se and B (Figure 4). The % adsorption decreases at pH 12 for B, Se and Mo while an increase is observed for As and Cr (Figure 4). At pH range of 4–10 the % adsorption for all chemical species was >70 for the Al³⁺ modified bentonite clay. The Al³⁺ modified bentonite clay recorded >90% adsorption for B and Se over the pH range of 2–12. Both raw and Al³⁺ modified bentonite clay exhibit close and high % adsorption capacity and trends for B and Se. This indicates the adsorption mechanism for B and Se is similar for both adsorbents and could involve other surface complexation of the anionic species that does not involve the Al³⁺ in the interlayers (Jain & Jayaram 2009; Jiang et al. 2009; Mohapatra et al. 2009). At the pH 2, As, Cr, Mo, B and Se will most likely be existing as positively charged ionic species and are likely to be exchanged with the exchangeable alkali metals in the interlayers hence the high adsorption. The decrease in adsorption at pH > 10 is mainly due to competition of adsorption sites with OH^- as most of these chemical species will be existing as oxyanions. This is attributed to adsorption of hydroxyl group as the pH increases. Optimum adsorption of As, B, Cr Mo and Se occurred over the acidic, circum-neutral to alkaline pH for both adsorbents. pH 10 was selected as the optimum pH for adsorption of the oxyanionic species and was applied in decontamination experiments of the FA leachates.

The adsorption isotherms of As, B, Cr, Mo and Se were developed by varying the initial As, B, Cr, Mo and Se concentration over the range 0.1–15 mg/l. The adsorption conditions were 30 minutes of contact time, 250 rpm agitation speed, 26°C and 4 g/100 ml S/L ratio. Two widely used adsorption isotherms: the Langmuir and Freundlich were used to describe the adsorption data as discussed in previous section.

The Langmuir and Freundlich adsorption isotherm constants for As, B, Cr, Mo and Se are presented in Table 3.

The Langmuir model gave a strong fit for As, B, Cr and Se (R² > 0.8) and Mo (R² = 0.2) was for Al-bentonite. The Q_m (mg/g) values for most of the species were higher for Al-bentonite. Computed R_L values for both systems were between 0 and 1 showing the adsorption process is favorable. The calculated R² values for the Freundlich adsorption isotherms show a strong correlation for most of the species for both systems (Table 4). This would indicate heterogeneity of the adsorbent surface. The higher values of K_f for the Al-modified bentonite indicate that the Al³⁺-bentonite system offered a much higher adsorption capacity. This could indicate that the adsorption occurred as a multilayer adsorption on the heterogeneous clay surface. The smaller values of 1/n for Al³⁺-bentonite indicate higher affinity of these species to Al-bentonite.

The correlation coefficients for all the oxyanion species ranged from 0.999–1 indicating a strong fit to the pseudo-second order model (Figure 5). The rate constants k₂ (g/mg/min) were as follows: As = 0.035, Mo = 0.061, Cr = 3.03, Se = 0.016 and B = 0.076. The fact that the adsorption kinetics of the Al³⁺-modified bentonite can be described by the pseudo second order indicates the adsorption is chemisorption.

Results of treatment of coal FA leachates using Al³⁺ modified bentonite clay at optimized conditions are presented in Table 4.

Table 4 shows the adsorption of As, B, Cr, Mo and Se from coal FA leachates onto Al³⁺ modified bentonite clay at optimized conditions. Al³⁺ modified bentonite clay exhibited superior adsorption of oxyanionic species from aqueous solution at optimized conditions. This was proved by the reduction in the concentration of oxyanions in the product water quality. The high pH value for CFA leachates may be attributed to high content of Ca in coal FA as reported in literature (Ahmaruzzaman 2010; Gitari et al. 2013c; Masindi 2013). Based on these results, it can be concluded that Al³⁺ modified bentonite clay is an effective adsorbent for As, B, Cr, Mo and Se from coal FA leachates. Similar results were obtained by Masindi et al. (2016) using Fe³⁺ modified bentonite clay. This implies that Al³⁺ modified bentonite clay can be used as a reactive barrier in ash retention ponds for removal of oxyanionic species.

Adsorption of most of the oxyanions was observed at pH ≥ pHpzc which indicates that both electrostatic and chemical interaction occurred with the clay surface and interlayers. Chemical interaction could have occurred through monodentate or multidentate ligands with the surface hydroxyl groups. The interlayer Al(H₂O)₆³⁺ could also chemically interact with the oxyanions through the OH- groups on hydrolysis [Equation (8)]. Increase in the adsorption dosage was observed to increase the pH for raw bentonite clay mixtures. This was attributed to dissolution of alkaline oxides from bentonite which is a natural geological material. Increase in adsorption with adsorbent dosage (Figure 2) pointed to similar adsorption mechanisms of the oxyanionic species in both materials. Introduction of Al³⁺ into bentonite clay interlayers was observed to increase adsorption capacity for the oxyanion species. High charge density Al³⁺ has high affinity for negatively charged species. At low pH and very low concentrations of the oxyanion species (0.1–1 mg/l) both adsorbents exhibited high adsorption capacity, it is postulated that ion exchange process for bentonite clay contributed to this high capacity. At low pH the oxyanion species would be positively charged and hence can exchange with positively charged alkali metals in the interlayers. Different selectivity of the oxyanion species was observed for the two adsorbents as the pH was varied from 2–12 (Figure 4). Al³⁺-bentonite showed low adsorption for Mo while raw bentonite showed low adsorption for As, Mo and Cr. This indicates that apart from interaction with Al³⁺ species in the interlayers and ion exchange, another factor could have contributed to the adsorption patterns observed for both adsorbents such as the size of the oxyanions. Bhattacharyya & Gupta (2008) and Dukić et al. (2015) reported that introduction of mono, di and tri-alkyl surfactants into Na-montmorillonite increased the interlayer spacing. Subsequently it's postulated that the introduction of hydrated Al³⁺ ions increased the interlayer spacing of the bentonite clay which allowed more of the oxyanion species to access the interlayer space of Al³⁺-bentonite.

Bentonite clay was successfully modified by introduction of Al³⁺ in the interlayers through ion exchange. Introduction of Al³⁺ led to decrease in CEC from 262–265.5 meq/100 g to 186.9–183.3 meq/100 g at pH 5.4–7.4. CEC was observed to be independent of pH and that Ca²⁺, Na⁺, Mg²⁺ and K⁺ were the main exchangeable cations. Modification with Al³⁺ increased the total surface area and micro-pore area. A slight increase in pH_PZC of Al³⁺ modified bentonite clay (8.2) compared to raw bentonite (8.0) was observed, this would enhance adsorption of anions from aqueous media over a wider pH range. XRF showed SiO₂, Al₂O₃, Fe₂O₃ as the major chemical constituents confirming that bentonite clay is an aluminosilicate material. The adsorption affinity of the modified bentonite clay for the oxyanionic species exhibited the following order: B = Se > Mo = Cr > As. Modified bentonite clay showed greater total adsorption capacity as compared to raw bentonite. Modification of the bentonite clay with Al³⁺ increased the adsorptive sites for the oxyanionic species. The adsorption data fitted well to Freundlich adsorption isotherm than Langmuir adsorption isotherm suggestion heterogeneous surface coverage by the adsorbate species. The kinetic data fitted well to Lagergren pseudo-second order kinetic model suggesting chemisorption as the main mechanism of adsorption. The Al³⁺ modified bentonite clay successfully removed As, B, Cr, Mo and Se from generated coal FA leachates. The results show that Al³⁺ modified bentonite clay is a potentially effective adsorbent for removal of oxyanionic species from coal FA leachates and could be applied as a reactive barrier in ash retention ponds.

The authors wish to convey their sincere gratitude to University of Venda, Council for Scientific and Industrial Research (CSIR), SASOL-INZALO, National Research Foundation (NRF) for funding this project.