Treatment of AMD using a combination of saw dust, bentonite clay and phosphate in the removal of turbid materials and toxic metals

South Africa is one of the countries rich in mineral resources such as gold, coal, uranium, steel and diamond. More than 60 percent of these minerals constitute acid mine drainage (AMD), an effluent generated from both operating and ceased mining sites. The acid state of the effluent is prone to dissolution of toxic metals and other soluble materials, which are detrimental to humans, organisms and aquatic life. Just as industrial wastewater is normally purified and re-used for essential or auxiliary processes, AMD can be treated through physico-chemical immobilization of toxic metals employing effective techniques. All types of contaminants present in wastewater can be treated according to their physico-chemical properties; that is, metabolization or decomposition of organic materials whereby metals transformed from labile toxic to form a more inert insoluble state which also exists in a less bio-available form in order to abate their detrimental effect (Nzihou & Sharrock 2009). A selection of cost-effective technology and a reagent(s) is imperative to encourage the mining authorities to employ such an approach of treating decanting AMD. However, a serious challenge associated with a majority of modern technologies includes high capital and operating costs (Ahmed et al. 2016; Fu et al. 2017), including difficulties encountered in the removal of toxic metals (Johnson & Hallberg 2003). A lot of studies conducted by Geldenhuys et al. (2001); Maree et al. (2004) and Petrik et al. (2003) revealed that most of the technologies characterized by high efficiency are sophisticated and costly. Research studies continued to explore some other mixed technologies, but the problem associated with costs still remains a challenge (Gitari et al. 2006, 2011; Nermen et al. 2009; Buzzi et al. 2011; Kopf et al. 2013). It was after various methods involving commercial reagents and new technologies showed unsuccessful outcomes that some other types of reagents such as waste materials and natural resources were investigated. The main problem regarding treatment of AMD is attributed to its complex chemical composition attributed to inorganic matter, dissolved and suspended solids, colour-and odour-producing materials and humic substances. Clay minerals and biomass were some of the techniques that were investigated using AMD and other synthetic complex wastewaters, where a majority of techniques exhibited high efficiency in the removal of contaminants (Tahir & Naseem 2007; Wei et al. 2008; Waanders & Brink 2010; Ntwampe et al. 2015a). The technique employed in this study includes treatment of the AMD using a combination of Na₃PO₄, bentonite clay and saw dust (flocculant). An advantage associated with this flocculant is that each reagent was added in reduced weight percentage ⁠, thus reducing the cost for commercial coagulants or reagents and plausible detrimental effect when the dosage of each reagent exists in a pure form (Ntwampe & Moothi 2018). An advantage of using clay in this study is based on its materials, which are mostly composed of heterogeneous mineral mixtures (Bloch & Hutcheon 1992), taking into account that South African bentonite clay contains both calcium and magnesium montmorillonite mineral, affording it optimal exchange capacity and adsorption (Syafalni Abdullah et al. 2013; Nel et al. 2014). Bentonite clay also has very reactive constituents, which provide it with net negative charge (Oladipo & Gazi 2014).

The selection of saw dust is based on the adsorption capacity due to the presence of lignocellulose, a combination of lignin, cellulose, hemicelluloses, which constitutes the primary structure of the plant wall protecting against attack from microbes and animals (Zhao et al. 2012). Its morphological structure provides an adsorption capacity when pulverized to establish a large surface area. On the other hand, the selection of Na₃PO₄ is based on its ability to react with multivalent cations and convert them into insoluble compounds, particularly orthophosphates. Various hydrogen and dihydrogen phosphates have been investigated to determine the reaction mechanism between soluble phosphate and metal ions, and the results showed easy formation of insoluble metal phosphate salts (Nriagu 1984). Na₃PO₄ was selected among other phosphates because of low costs and low reactivity of the sodium component compared to other components such as potassium, ammonium and tri-hydrogen. The larger amount of Pb normally reacts with PO₄²⁻ to form hydrargyrite (Pb₅(PO₄)₃Cl), a species linked to phosphate occurring during a reaction within a wide pH range. Notwithstanding its existence in a wide pH range, it can be soluble in an acidic medium, depending on the stability and the concentration of reactive metals/substances in the medium (Amrah-Bouali et al. 1994). The outer surface of hydrargyrite is negative in a basic medium and positive in an acidic medium; substitution is ubiquitous, particularly for apatite chemistry, whereby a variety of ions may be attached to the apatite structure.

The objective is to investigate the efficiency of a flocculant (saw dust, size-optimised bentonite clay and Na₃PO₄ at reduced mass % (⁠ %). The rationale being to optimise both the dosage and particle size of bentonite clay. The cost-effectiveness of a flocculant is the comparison between calculated costs to those obtained by Maree (1992). The reason to use cost analysis on the observations obtained by Maree (1992) with that obtained in this study was to optimize the system as the objectives are convergent; it also investigates cost-effective reagents which can replace costly conventional reagents. The novelty in this study is the use of natural and waste materials and low dosage of Na₃PO₄ without pre-processing of the reagents: a cost-saving approach.

In this study, the coagulation-flocculation treatment was applied to the AMD sample using 1.5 g clay (bentonite), 10, 20, 30, 40, 50 and 60 mL of 4.5 g/L of NaPO₄, and 3 g/L saw dust dosages respectively. Another set of experiments was conducted with a dosage of a flocculant consisting of the same reagents added to the AMD sample. The experiments were conducted in triplicate for reproducibility. The AMD samples were treated using a jar test at 250 rpm for 2 min and then at 100 rpm for 10 min. The samples were allowed to settle for 1 h, after which the pH, electrical conductivity (EC), turbid materials (turbidity), toxic metals content, dissolved oxygen (DO), and oxidation reduction potential (ORP) were measured.

The AMD was collected from the Western decant in Krugersdorp (SA) in a 25 L plastic drum, and modified by blending it with kitchen wastewater to introduce organic materials. The overall pH, conductivity, DO, ORP and turbidity of the composite AMD solution were 2.56, 4.43 mS/cm, 234 mV, 4.1 mg/L and 213 NTU. The concentration of turbid materials of the AMD (9.2 g/200 mL) was measured by filtering a 200 mL of the sample on a filter of a certain mass using a high vacuum filter and the residue dried in an oven at 110 °C for 4 h, and the toxic metals content in the AMD sample is listed in Table 1.

Bentonite clay was obtained from the Yellowstar Bentonite mine, a bentonite mining and supplying company situated in Parys in the Free State (SA). It was pulverised to 180 and 220 μm to an optimised and cost-effective size in order to curtail costs associated with energy and milling.

Toluidine blue dye purchased from Sigma Aldrich (South Africa) was used for coloration of the AMD; the rationale being to determine removal efficiency of colour using a flocculant employed in the study. 0.8 g of a dye was added to a litre of untreated AMD sample and stirred thoroughly.

A stock solution was prepared using the quantities of the reagents (saw dust, bentonite clay and Na₃PO₄ filled to 1 L of demineralized water filled to a litre of volumetric flasks each to prepare a flocculant. The selection of 4.5 g/L Na₃PO₄ was used to determine the efficiency of the salt at low concentration for cost savings. The calculation of the mass of metal salts was used to obtain various concentrations of Na₃PO₄ (163.94 g/moL, basicity 2.23), No. 157750, Grade 97%.

Table 2 shows the metal salts dosed into the AMD samples.

The utensils and experimental equipment were all rinsed three times using distilled water; the electrodes used were also thoroughly rinsed using distilled water. The samples were analysed at North-West University (SA). All equipment used for the experiments were calibrated as per laboratory standards, and the ICP-OES, XRD, and SEM analyses were conducted by qualified technicians. All the instruments were calibrated on a daily basis using appropriate standard solutions.

The equipment used for the jar tests was a BIBBY Stuart Scientific Flocculator (SW1 model), which has six adjustable paddles with rotating speeds between 0–350 rpm. A 200 mL sample of AMD (containing 9.3 g of turbid materials as measured by filtering 200 mL of the AMD sample, dehydrating it using an oven at 60 °C for 4 hours, and thereafter measuring the mass) was poured into each of the five 500 mL glass beakers for the test.

A Merck Turbiquant 3,000T Turbidimeter (Japan) was used to determine turbid materials (turbidity) in the supernatant, using NTU as a unit of measure. It was calibrated with 0.10, 10, 100, 1,000, and 10,000 NTU standard solutions. Turbidity was calculated by NTU conversion; that is, by multiplying the turbidity readings by 3.42.

The pH, EC, DO and ORP were measured using a SensoDirect Multimeter (South Africa) with an electrode filled with silver chloride solution and the outer glass casing with a small membrane covering the tip. The equipment was calibrated with standard solutions at pH of 4.0 and 7.0 before use. Parameters were measured using their respective probes, and the instrument was fitted with a ‘temperature correction’ device.

Experiment A: 200 mL of a sample was poured into each of the five 500 mL glass beakers and dosed with a synthetic flocculant as mentioned, and treated in a jar test. The sample was allowed to settle for 1 h, after which the pH, EC, turbid materials, DO, and ORP measurements were repeated.

Experiment B: A similar set of experiments, where a flocculant in 1 L of demineralized water was added to the AMD and treated in the same manner.

The X-ray diffraction (XRD) patterns of the turbid materials (salts content) adsorbed onto the sludge of the AMD samples dosed with a flocculant were recorded using a Rigaku Miniflex II Desktop X-ray diffractometer (Japan) with Cu Kα radiation. A step size of 0.02° at a speed of 4° (2θ)/min over 10–80° was applied. The minerals in VM treated between 950 and 1,300 °C were quantified by Siroquant software.

A KYKY-EM3200 digital scanning electron microscope (SEM; model EM3200) (China) was used to produce the SEM photomicrographs.

A Perkin Elmer Optima DV 7000 ICP-OES optical emission spectrometer (USA) was used to determine the concentration of metals and other pollutants in the supernatant of the AMD samples. The measurements were conducted by a qualified technician in the chemistry department following the standard operating procedure of the ICP-OES equipment.

Pseudo-second order model was applied in adsorption experiments to determine kinetic study of adsorption capacity of an adsorbate onto adsorbent; that is, turbidity onto the flocculant consisting of a combination of bentonite clay, saw dust and Na₃PO₄. The Freundlich model was applied for multilayer adsorption of turbidity on a surface of a heterogeneous adsorbent. Only the pseudo-second order model and Freundlich were selected based on the application and limited graphs required.

A plot of log q_e vs log C_e should be linear if the model fits the experimental data.

The determination of the ionic strength of the colloidal suspension can be experimentally or theoretically calculated using Equation (4). In this study, ionic strength was calculated due to a lack of the measuring equipment, and the results are shown in Table 3.

The effectiveness of a treatment process is shown by the rate of deprotonation during destabilization, which correlates with the hydrolysis, leading to the rate of adsorption (Ntwampe et al. 2013a). Figure 1 illustrates the relationship between the pH and the EC or DO as the function of the dosage employed during AMD treatment using a combination of clay, saw dust and (0.012, 0.025 and 0.05 M) of Na₃PO₄ in a flocculant.

The pH of the samples with a flocculant containing a dosage of 0.012, 0.025 and 0.05 M Na₃PO₄ dosage exhibited an inconsistent changing trend from 2.56 to the ranges of 2.62–2.4, 2.63–2.17 and 2.41–2.20 respectively (Figure 1). Such a changing pattern may be attributed to perturbation occurring between forward and reverse reactions. The pH and the EC of the samples dosed with a flocculant containing 0.012 M Na₃PO₄ show a correlation. On the contrary, the pH and the EC of the samples dosed with a flocculant containing 0.025 and 0.05 M Na₃PO₄ do not show correlation (Figure 1); that is, the EC shows an inconsistent changing trend. However, the high ionic strength of a solution (Table 3) is indicative of effective destabilization-hydrolysis. Notwithstanding the changing trend of the EC, the neutralizing effect of CaCO₃ is effective as the pH decreases consistently.

The DO of samples dosed with the flocculants containing 0.012, 0.025 and 0.05 M Na₃PO₄ changed from 4.1 mg/L (untreated AMD) to the ranges of 4.2–4.8, 4.1–5.7 and 2.1–4.6 mg/L respectively. The low DO range shown by the samples dosed with a flocculant containing 0.05 M Na₃PO₄ is probably attributed to high rate of oxidation of the cations contributing to high conductivity. This is exhibited by a low DO and conductivity of the samples dosed with a flocculant containing 0.012 M Na₃PO₄. The concentration of the Ca in the raw AMD (Table 1) was reduced from 188 to 59 ppm (Table A1, Supplementary material), and the removal is presumably through adsorption onto the flocs, co-precipitation or enmeshment in the precipitates (Moore et al. 1978). The pH, EC, DO, ORP, and turbid materials measurements of the 1.5 g of bentonite clay in a litre of demineralized water, are shown in Table 4. The metal concentrations of the untreated AMD are shown in Table 1.

The EC of the samples dosed with a flocculant containing 0.05, 0.025 and 0.012 M Na₃PO₄ (Figure 1) changed from 4.30 mS/cm (untreated AMD) to the ranges 5.09–7.66, 2.43–4.34 and 3.15–4.24 mS/cm respectively. On the other hand, AMD samples with 40–60 mL of 0.05 M Na₃PO₄ dosage exhibit an inconsistent changing pattern; only the samples dosed with a flocculant containing 0.025 M Na₃PO₄ exhibited a uniform decreasing pattern. The oxidation reduction potential (ORP) (Figure 2), is an indication of the rate at which redox occurs on the metals to form cationic and anionic metal; these are highly reactive compared to the parent molecule and indicative of the formation of precipitates which are removed by co-precipitation or sorption. Settling time of 1 hour allowed optimal nucleation, crystallization and sedimentation (Ntwampe et al. 2013b), resulting in effective physico-chemical reaction, which is predominantly induced by Brownian motion to form fully fledged flocs. Figure 2 shows the pH, conductivity and ORP changing trend with dosage.

The ORP changed from 236 mV in the raw AMD sample to a range 227–268 mV in treated samples (supernatant). The ORP of raw AMD sample showed a changing trend from 236 mV to the redox potential, which is in a range −6–32 mV in treated samples. The results show a correlation between ORP and the dosage, whereas the samples dosed with a flocculant containing 0.012 M Na₃PO₄ has ORP in a range −6 and −0.2 mV, which indicates that all the samples went through the oxidation process. On the other hand, ORP of the samples dosed with flocculants containing 0.025 and 0.05 M Na₃PO₄ exhibited an inconsistent changing trend in a range 5–21 and 13–32 mV respectively, indicating that the reduction reaction was a predominant phenomenon, the latter changing trend showing a high reducing strength. Figure 3 shows the pH, ORP and residual turbid materials of the AMD samples with varying concentrations of Na₃PO₄ in the flocculant. The effectiveness of turbid materials removal ranges from 1.32 to 2.74 mg/L (75–97.1%).

Amphotericity of water is another parameter that plays a role during destabilization-hydrolysis (Equations (4) and (5)), which then revealed an insignificant role played by pH adjustment in wastewater treatment using a flocculant containing Na₃PO₄. It is essential to note that pH adjustment is mainly to reduce the solubility of toxic metals in a solution with minimal effect during destabilization-hydrolysis as a majority of coagulants react within a wide range of pH (Flynn 1984). As a matter of fact, the process depends predominantly upon the physico-chemical properties of the solution (Moore et al. 1978; Flynn 1984) and the reactivity rate (electronegativity, acidity and distance between the neighbouring colloidal particles (Ntwampe 2013; Ntwampe et al. 2013a; Ntwampe et al. 2015b).

The flocculant showed optimal removal of toxic metals such as Co, Cu, Ni, Pb, As, Cr and Zn from 8.336, 14.45, 7.554, 17.71, 14.61, 8.434 and 12.564 ppm (Table 1) to 0.138, 0.144, 0.373, 0.189, 0.270, 0.795 and 0.315 ppm (Table A1) respectively. The removal efficiency of Pb corroborates with recorded observations (Nriagu 1984; Azihou & Sharrock 2010) by achieving the highest efficiency from 17.7 to 0.189, a removal efficiency of 98.9%. The observations show an immense mass transfer for various metal ions and considerable pore and surface diffusion through ion exchange at specific sites, and intercalation between the sheets formed by the reaction of hydroxyapatite, saw dust and bentonite clay. It is suggested that saw dust and ionized Na₃PO₄ firstly occupied the free sites on the surface of the bentonite clay to form new interfacial structures available for sorption and intercalation. Spontaneous dissolution-sorption mechanism cannot be considered as the only phenomenon attributable to optimal removal of turbidity; also hydration occurring around the central metals utilizes internal enthalpy of hydration to dehydrate aqueous metal prior to ionic exchange, which must be higher than the hydration enthalpy of adsorbable toxic metal ions. On the other hand, the dissolution-precipitation mechanism revealed that hydroxylapatite dissolves under a wide range of pH values and increases the mechanisms that govern the removal rate of toxic metal ions (Cheng et al. 1997), and such an inference is invoked by the results obtained in this study (Table A1). Some of the metals are largely removed compared to others, suggesting that the metals with higher enthalpy of hydration (Pb, Cu, Cr and As) exhibited optimal removal efficiencies. Removal efficiency is also influenced by surface and pore diffusion including intra-particle diffusion, which may at times be ineffective due to steadiness at room temperature. The efficiency of hydroxylapatite is also apparent when the metal ions precipitate due to low solubility on the surface of the hydroxylapatite during the secondary phase. Low removal efficiency of some toxic metals is more likely to be attributable to the interference of diffusion by clogging of the pathway during accumulation of amorphous products. Most of such metals are unstable at the primary stage, such as Cr²⁺ and Fe²⁺, among others. In addition to sorption mechanisms, the performance of ionic exchange is influenced predominantly by the cation-exchange-capacity (CEC) occurring when the metals attach to the surface of the flocculant and pushes away more soluble indigenous calcium from the flocculant, most of which is attached to hydroxylapatite. Alternatively, calcium may dissolve first followed by phosphate ions, allowing precipitation of toxic metal ions by homogeneous (in solution) or heterogeneous (on the remaining solid hydroxylapatite) nucleation (Manecki et al. 2000a, 2000b). The final removal of toxic metals is likely to be attributed to the crystallinity, specific surface area, density, pH, temperature, ligands, concentration of the metals and addition rate, as they also form part of the physico-chemical properties of the system. The removal efficiency of Ca²⁺ and Mg²⁺ ions (Table A1) shows that a large amount reacted with PO₄²⁻ to form a new metal-phosphate species, which is invoked by the available observations revealed in the study conducted by Cheng et al. (1997). Figure 4 shows the relationship between the pH and the removal efficiency of turbid materials using a flocculant containing 0.025 M Na₃PO₄. It can therefore be concluded that pollutant removal (Figure 3) is attributed to attraction between adsorbent and adsorbate (chemosorption), and attraction by neighbouring oppositely-charged particles (physicosorption) induced by van der Waal's forces of attraction (Ntwampe et al. 2015c), ion exchange (Sinha et al. 2013), and electrostatic repulsions on the basal planes or covalent bonds by amphoteric ligands on the edge of montmorillonite with the functional (carboxyl) group (van der Watt 2011), including its large specific surface area and exchange capacity (Syafalni Abdullah et al. 2013). Figure 4 shows the relationship between the pH and removal efficiency of turbid material using a flocculant containing 0.025 M Na₃PO₄. The selection of the molarity was to investigate the performance of the flocculant using optimized dosage (0.025 M being between 0.012 and 0.05 M).

Figure 4 yielded a high correlation regression (R²) of 99.9%, which is indicative of the accuracy of the experimental data used to plot the graph. Figure 5 is a Freundlich model showing a relationship between removal efficiency and the dosage of a flocculant containing 0.025 M Na₃PO₄. On the other hand, Figures A1 and A2 (Supplementary material) show the Freundlich model plotting a relationship between removal efficiency and dosage of a flocculant containing 0.05 M Na₃PO₄, and the pseudo-second order model showing the relationship between adsorption efficiency at equilibrium using a flocculant containing 0.025 M Na₃PO₄.

The experimental data (using a flocculant containing 0.025 M Na₃PO₄) used to plot a graph shows a best mathematical fit for Freundlich adsorption isotherms; with correlation regression of 0.998 (99.8%). On the other hand, the experimental data using a flocculant containing 0.05 M Na₃PO₄ (Figure A1) also shows a best fit with correlation regression of 0.993. The experimental data (using a flocculant containing 0.025 M Na₃PO₄) used to plot a graph (Figure A2) shows a best mathematical fit for pseudo-second order with coefficient regression (R²) of 0.999. Figure 6 illustrates the mineral compounds of the treated AMD with bentonite clay, saw dust and 0.025 M Na₃PO₄ dosage.

The XRD curves (Figure 6), where the AMD sample treated with a flocculant containing 0.025 and particle size 180 μm (bentonite clay) showed two peaks representing crystalline materials at 2θ positions of 20° and 37° respectively, with an intensity of 500 counts. The narrow diffraction peaks between 400 and 500 counts are indicative of the presence of trace metals disorderly stacked up by flocs. The XRD pattern of such minute peaks representing trace metals (Figure 6) confirms a high efficiency using that particular flocculant. SEM micrographs (Figure 7) show the results of the AMD sample dosed with a flocculant containing 0.025 and 0.05 M Na₃PO₄, and particle size of 220 μm during rapid mixing. On the other hand, Figure A3 shows the sludge of the sample dosed with a flocculant containing 0.025 and 0.05 M Na₃PO₄, and particle size of 220 μm during rapid mixing

The SEM micrograph of the sludge of a sample dosed with a flocculant containing 0.05 M Na₃PO₄ (Figure 7(a)) shows scattered dense flocs surrounded by a cluster of small flocs, indicating rapid floc formation due to over-dosage, which resulted in agglomeration of a dense sponge-like structure surrounded by large voids and smaller agglomerates. On the other hand, the SEM micrograph of a sample dosed with a flocculant containing 0.025 M Na₃PO₄ showed normal dispersed uniform-sized flocs joined together, an indication of a steady rate of sorption due to optimally pulverised bentonite clay with large surface area. The micrograph shows a normal distribution of saw dust and Na₃PO₄ particles throughout the surface of the clay to form a rigid adsorbent. SEM microgram sample dosed with a flocculant containing 0.025 and 0.05 M Na₃PO₄, and particle size 220 μm during rapid mixing (Figure A2), shows almost identical crystal morphology; that is, large flocs with perforated surface joined together by small flocs. It is suggested that the morphology of the micrographs is attributed to high sorption capacity of the pollutants by permanent negative surface charge and interlayer spaces resulting predominantly in higher ion exchange capacity and metal sorption, an observation already determined by Abbas et al. (2018); Chen et al. (2018); Delavernhe et al. (2015) and Khandelwal et al. (2019). Based on the constituents of bentonite clay, saw dust and chemical composition of Na₃PO₄, it is conceivable to suggest that optimal removal of the pollutants is dependent upon a large number of pores and abundant surface functional groups (Figure A2), utilized in metal adsorption. The tiny voids distributed throughout the micrograph and clustered morphological structure are indicative of high sorption capacity. Based on the high performance exhibited by the bentonite clay with 220 μm particle sizes, excessive milling using high energy consumption is costly and unnecessary, as larger particle size using low energy consumption during milling yields identical results.

The r-value obtained for the AMD samples (exp. A) in experiment (A) is 0.65 (65%) with the range of the correlation coefficient from −1 to 1. The r-value obtained for the AMD sample (exp. B) in experiment (B) is 0.55 (55%), showing a strong relationship. That is validated by the correlations of the samples dosed with a flocculant containing 0.025 and 0.05 M Na₃PO₄. The determination of natural organic matter (NOM) has not been a popular measurement in water treatment processes; however, it is significant as a majority of NOM consists of humic (hydrophobic) and fulvic (hydrophilic) acids such as high and low molecular weight compounds. These compounds are harmful to treated water, particularly when it has to be distributed for potable supplies and critical industrial activities (drinking, cleaning laboratory equipment, food production, etc.) as they react with disinfectant to form disinfection by-products (DBPs) (drinking, cleaning laboratory equipment, food production, etc.) (Chowdhury 2013). A sample of treated effluent dosed with a flocculant containing 0.025 Na₃PO₄ and particle size 220 μm (clay) was assessed to determine the amount of NOM (turbidity) at equilibrium employing the UV₂₅₄ technique. Adsorbable turbid materials remaining after coagulation could not be adsorbed due to equilibrium state with adsorbed turbid materials onto the flocs (Figure A4). The correlation regression of both the SUVA and UV254 are high, at 0.958 and 0.967 respectively (p > 0.05); high SUVA R2 shows optimal removal efficiency of NOM and TOC using a flocculant containing 0.025 Na₃PO₄ and particle size 220 μm.

The principal objective of this study was to determine the efficiency of a flocculant consisting of bentonite clay, saw dust and Na₃PO₄ in the removal of turbid materials present in the AMD sample. The reagents which were used to prepare a flocculant (bentonite clay, saw dust and Na₃PO₄) were mixed and dosed directly without processing as stated. The experimental results revealed that a combination of bentonite clay, saw dust and 0.012 and 0.025 M Na₃PO₄ without pre-processing yielded optimal removal of turbid materials, colour, NOM and toxic metals. The removal efficiency is indicative of a complete and effective destabilization-hydrolysis, a phenomenon that is indicated by the removal efficiency of the pollutants. The presence of Na₃PO₄ in a flocculant shows significant contribution in the removal of Ca and Mg from the AMD, which also acts as an ideal water softening agent. Bentonite clay and saw dust showed optimal removal of the turbid materials due to their porous and ion-exchange properties. On the other hand, the phosphate component of the flocculant shows optimal removal of the divant metals due to its ability to react with most of the metals and metalloids, such as Na, Mn and Fe. In addition, optimal removal efficiency associated with porosity and ionic exchange is an indication of adsorption of the pollutants through physico-chemical phenomenon. The flocculant, consisting of 0.025 M Na₃PO₄ and bentonite clay of 220 μm particle size, showed optimal removal efficiency of toxic metals and colour; that is, higher than 96%. The observation shows that optimization of the flocculant using Na₃PO₄ and particle size of the bentonite clay is significant for the removal of pollutants in the wastewater. Proliferation of electrical conductivity with dosage (increasing ionic strength of a solution) corresponds to optimal sorption of pollutants; this indicates that electrical conductivity plays a pivotal role during destabilization-hydrolysis. Optimal removal of pollutant still occurred using moderate particle size; this shows that over-pulverization of the reagents is unnecessary and costly. Optimal removal of NOM is indicative of their reactivity during physico-chemical reactions; that is an indication of their catalytic property during the treatment process. The SEM micrographs of the flocculants containing 0.025 and 0.05 M or 180 and 220 μm particle size all show dense sponge-like flocs surrounded are costly and unnecessary; bigger particle size (220 μm) yielded identical removal efficiency of pollutants as smaller particle size. The clusters exhibited by the sludge of treated samples (SEM micrographs) are indicative of high adsorption capacity of the adsorbent, and a high removal of the pollutants using moderate pulverised bentonite clay (220 μm). The bill of material (reagents) used in the treatment of 15 ML of the AMD amounts to R 33.3 (US$ 1.85) (Table A4), excluding labour costs; this is less compared to the costs obtained by Maree (1992). The practical interpretation of this comparison indicates that cost reduction on water treatment chemicals can be achieved by employing a mixture of commercial reagent(s) with waste materials with high sorption capacity.

All relevant data are included in the paper or its Supplementary Information.