Chemical reactivity between CaCO 3 and Ca ( OH ) 2 in acid mine drainage ( AMD ) with mixing and shaking techniques during the destabilization-hydrolysis of the AMD

The acid mine drainage (AMD) was poured into five 500 ml glass beakers. The samples were dosed with synthetic af-PFCl of Ca(OH)2 and af-PFCl of CaCO3 polymers respectively. The samples were treated in a jar test and a shaker at 250 rpm for 2 min, and thereafter were allowed to settle for an hour after which the pH, conductivity and turbidity (TSS) were measured. A similar second set of experiments was conducted by placing the samples in a shaker at 250 rpm for 2 min, after which three measurements were conducted after 1, 2 and 6 h. Similar third and fourth set of experiments was conducted dosing the AMD with 0.043 M of Ca 2+ in Ca(OH)2, and 0.043 M Ca 2+ in CaCO3 respectively. A fifth set of experiment was conducted by dosing the AMD sample with 0.021 and 0.043 M Ca 2+ in Ca(OH)2 respectively and treated in a jar test, shaker and without mixing. The synthetic acid free PFCl of Ca(OH)2 or CaCO3 exhibited a high TSS removal efficiency. Both polymers also show a similarly identical TSS removal efficiency, which depict Fe 3+ ions as the principal role player during destabilization-hydrolysis. Effective sedimentation of the turbid materials in the AMD sample with af-PFCl polymers of both Ca(OH)2 or CaCO3 occurs after 2 h of settling. The TSS removal values in the AMD sample which were treated in a shaker at 200 rpm are slightly lower than those treated in a jar test at 200 rpm. The residual TSS values in the AMD samples stirred at 350 rpm during rapid mixing are slightly higher compared to those stirred at 200 rpm rapid mixing.


INTRODUCTION
Wastewater treatment is a process which is very essential, more especially during this era where climate change affects the entire ecosystem.The treatment process includes the removal of particulate matter from the colloidal suspension to meet the requirements where it can be re-used.In order to understand the behaviour of particles in water and to develop and design efficient treatment facilities, the characteristics of particles has to be known on the basis of individual solids and of whole particle populations.In water treatment, particles are of extremely heterogeneous nature with respect to size, density, shape, chemical composition, shear strength, surface charge, amongst others.Acid mine drainage (AMD), is the type of wastewater which requires a very intensive research and understanding.This is because it is caused by a naturally occurring compound, namely pyrite (FeS 2 ), which could not be prevented or avoided.Mining is not the cause of pyrite formation, but it only exposes it to oxygen and water during and after excavation to form acidic solution.The reaction that takes place during the oxidation of pyrite is shown in Equation (1).
FeS 2 + 7/2 O 2 + H 2 O → Fe 2+ + 2 H 2 SO 4 (1) The H 2 SO 4 which has been formed during the reaction is detrimental to the ecosystem, that is, biota and fauna, whereas the Fe 2+ can combine with OH -ions to form either unstable ferrous hydroxide or stable ferric hydroxide (Fe(OH) 2 or Fe(OH) 3 ) species respectively (Jones et al., 1988;Maree, 1994;Feng and Nansheng, 2000;Geldenhuys et al., 2001;Naicker et al., 2003;Maree, 2004;Semerjian and Ayoub, 2003;Watten et al., 2005;Maree, 2006;Akcil and Koldas, 2006;Kurniawan et al., 2006;Herrera et al., 2007;Sibrell et al., 2009).A survey conducted by Naicker et al. (2003) revealed that the ground water in the mining district of the Witwatersrand (RSA) was heavily contaminated with heavy metals and acidified due to the oxidation of pyrite containing tailings.It is also confirmed that mine water from the coal mining industry is also periodically discharged directly into local streams (Geldenhuys et al., 2001).It is immaterial as to whether these mine waters originate from the flooded mines, sand and slimes dumps, or direct discharge effluent; the condition remains an environmental hazard.The acidic character (pH as low as 2) has been reported by Feng and van Deventer (2004), that is that a high sulphate content and dissolved heavy metal content in the AMD prohibit its discharge into public streams as it poses a real threat to aquatic life.The costs for AMD treatment are high, and the trend to submit to mine closures rather than the treatment in South Africa has become a matter of great concern (Labuschagne et al., 2005).Lime has been a conventional reagent used in AMD treatment to increase the pH and concomitantly precipitate heavy metals.The use of limestone as a cheaper alternative was explored by Maree (1994), a combination of lime and limestone by Maree (1994) and Geldenhuys et al. (2001).Other agents investigated are blast furnace slag (Feng and van Deventer, 2004), whereas for fly ash (Petrik et al., 2003;Xenidis et al., 2000).Feng and van Deventer (2004) reported that the use of slag as a potential sorbent for metal ions in AMD increased the pH to neutral, and most of the heavy metals were removed successfully.Petrik et al. (2003) reported that neutralization of various sources of AMD with fly ash or fly ash leachate was possible, without the additional use of liming agents.Some novel applications to control AMD at the source with the use of thiocyanate compounds and phosphate derivatives are under indefinite discussion (Mudder et al., 2005).Several new processes have been developed, based on the use of precipitated calcium carbonate or lime pre-treatment for neutralization of AMD and partial desalination (Maree, 1992;Maree, 1994;Abdessemed et al., 2000;Lee, 2001;Scherrenberg et al., 2008;Maree et al., 2004;Chang and Yu, 2004;Carballa et al., 2005;Hankins, 2006;Pratt et al., 2007;Edwards and Withers, 2007;Kempkes et al., 2007;Pinto, 2008;Navratil et al., 2008;Moussas and Zouboulis, 2009;Suarez et al., 2009;van der Graaf et al., 2010;Jiang et al., 2012).A common by-product of lime neutralisation is gypsum, which precipitates when the AMD is often rich in sulphate and the calcium in lime will bring the solubility product well above saturation.The reaction of calcium hydroxide with sulphuric acid from the oxidation of FeS 2 is as follows: Limestone (CaCO 3 ) reacts with strong acids to form water, carbon dioxide and calcium salts as shown by Equation (3).
Although not shown, the product water molecules are attached to the CaSO 4 to form gypsum (CaSO 4 .2H 2 O).
In view of the fact that the reagents which are used, [Ca(OH 2 ) and CaCO 3 ] in this study have been used before, the difference in this regard is that the former coagulant was prepared by decarbonization of the latter and hydrated as shown in the reaction by Equation (4).
CaCO 3 + energy → CaO + CO 2 (4) When the water (H 2 O) is added to CaO, the formation of an aqua-Ca(OH) 2 occurs.In principle, the Ca 2+ ions in both reagents share physico-chemical properties, but different ligands (OH -or CO 3 2-).The other different factor is pH, that is, adjustment in AMD was conducted in previous studies, whereas it is not included in this study.Maree et al. (2004) stated that during the treatment of acid leachate from coal discard using CaCO 3 and biological sulphate removal, a portion of sulphate tends to crystallize out in the coal processing plant and form scaling on the equipment and pipeline, a problem which can be investigated in this study.Maree et al. (2004) also found in their study on the optimization of effluent treatment at a coal mine by process modelling that gypsum has a tendency of forming scaling on the equipment and pipeline.The focus of this present study is to investigate the chemical reactivity between the metal ions on both reagents in the destabilization of the double layer (DL) of the aqua-colloids; Figure 1 shows that the ionic strength of the metal ions influences the rate of the destabilization of the colloidal suspension through double layer compression.The study also investigates the effect of the ligands on the formation of hydrolysis species using conductivity as an indicator.
Mixing and the concentration of the dosage expressed in mass per volume (g/L) have been employed in various research projects.The former is commonly known for its disruption in the formation of smaller and larger flocs, thus weakening turbidity removal efficiency.This happens when the shear forces induced during rapid mixing rupture the flocs formed during coagulationflocculation, thus causing re-stabilization (Oldshue, 1983;Heath et al., 2002;Binnie et al., 2003).In the present study the AMD sample treated with mixing and shaking, and without mixing to determine the effect of mixing when compared with shaking and no mixing is investigated.This also determines the reactivity of metal ions (Ca 2+ ) of the reagents (electronegativity) during the destabilizationhydrolysis process.Destabilization is a nonlinear function of the concentration of the coagulant and the minimum concentration of a coagulant to cause effective destabilization which is termed the critical coagulation concentration (CCC) (Mark et al., 2013).Apart from the concentration of a coagulant, ionic strength plays a role in the destabilization.A study by Baran and Teslenko (1990) revealed that destabilization of nonionic polymers and anionic polyelectrolytes cause their aggregation by the bridging mechanism, whereas destabilization in the highly charged cationic polyelectrolytes occurs predominantly via the neutralization of the surface charge and potential of the particles.The present study the physico-chemical dynamics which are involved during destabilization-hydrolysis using synthetic polymers prepared by af-PFC of Ca(OH) 2 or CaCO 3 are also investigated.The hydrolysis reaction with metal ions (inorganic coagulants) differs from that with a polymer.This is primarily because hydrolysis with the latter involves a polymeric compound, whereas in the former it involves only a single metal ion.If the polymers are unionized, one part gains a hydrogen atom (H + ) and the other gains a hydroxyl group (OH -) from a split water molecule.If the polymer is ionized, one part receives an oxygen atom, O 2-and the other 2 hydrogen atoms, 2 H + (Cain et al., 2010).Hydrolysis reaction with a polymer entails the consumption of a water molecule as a result of breaking the covalent bond holding two components of a polymer together.The breaking of the bonds results in the release of energy.Gibbs free energy (Equation 5) also plays a pivotal role during the destabilizationhydrolysis process.The value is an indication of the spontaneity or induced mode of the reaction; negative values indicate that the process is spontaneous, and when positive it is induced by energy input.

ΔG = ΔH-TΔS (5)
The inference is that the combination of the chemical energies stored in the reagent/coagulant and the electronegativity of the metal ion have a spontaneous reaction rate.Dosage in mass per volume (g/L) is not a stoichiometric expression between a destabilizinghydrolysis agent (M n+ ) in a reagent/coagulant and the amount/concentration of colloids adsorbed.The concentration of the coagulants in the present study is expressed as molarity (number of moles per volume, m/L).This approach illustrates the number of moles of the metal ions in a reagent/coagulant which reacts with a stabilized colloidal suspension to form hydrolysis species which are adsorption substrates.
Effective wastewater treatment depends on the physico-chemical properties of both the coagulant and the colloidal suspension (O'Melia and Shen, 2001;Tambo and Watanabe, 1984;Adams et al., 2002).Notwithstanding all those properties, the type and size of the turbid materials and the electronegativity of the metal ions of a coagulant are the main key role players (Hubbell et al., 2003;Meghzili, 2008;Water Specialist Technology, 2003;Aboulhassan et al., 2006;Scholtz, 2010).Colloids can either be hydrophilic (water-loving) or hydrophobic (water-hating), where the latter are easy to treat (Moore and Moore, 1976).Contaminated water contains particles of different sizes which can be classified as dissolved (< 0.08 μm), colloidal (0.08 -1 µm), supracolloidal (> 1 -100 µm) and settleable (> 100 µm) (van Niewenhuijzen et al., 2004).The type of treatment selected depends on the size of particles present in the wastewater; where very fine particles of a colloidal nature (colloids of size < 1 µm) with high stability are significant pollutants.The stability is caused by the electrostatic surface charges of the same sign possessed by these particles (usually negative).This means that repulsive forces are created between them which prevent their aggregation and subsequent settling.It has therefore proved to be impossible to separate them by settling or flotation, and physico-chemical treatment (addition of coagulants) is the only option.It changes the physical state of the colloids allowing them to remain in an indefinitely stable form and therefore form into particles or flocs with settling properties (Tambo and Watanabe, 1984).The treatment occurs in two stages such as coagulation and flocculation.The aggregation of submicron particles during rapid mixing is relatively fast if their surface chemistry is ideally suited and their concentration is high enough (> 108/ml).Their transport is brought about by Brownian motion, also known as perikinetic flocculation which is influenced by the thermal condition of the colloidal system and induced by the coagulants (Dobias, 1993).The agglomerates still stay small and cannot be removed by sedimentation or filtration until further agglomeration during flocculation termed orthokinetic flocculation occurs (Dobias, 1993).Flocculation is classified either as micro-flocculation or macro-flocculation.The former is significant for particles in the size range from 0,001 to 1 m and the latter for particles of size greater than 1 or 2 m (Metcalf and Eddy, 2003).

MATERIALS AND METHODS
In the present study, coagulation-flocculation treatment has been employed to AMD solution dosed with 0.043 M of af-PFCl of Ca(OH)2 or CaCO3 respectively in a jar test at 350 rpm for 2 min with mixing or shaking.The samples were allowed to settle for 1, 2 and 6 h respectively after which the pH and turbidity were measured.A second set of experiments was conducted with the same flocculants at 200 rpm with mixing or shaking.The measurements were conducted after 1, 2 and 6 h.A third set of experiment was conducted with 0.021 and 0.043 M of Ca 2+ in Ca(OH)2 dosage respectively at 250 rpm with mixing and shaking for 2 min.The samples settled for 1 h, after which similar measurements were conducted.

Acid mine water sample
The AMD sample was collected from the Western Decant in Krugerdorp in a 25 L plastic drum.The sample was air-tied and stored at room temperature.The pH and turbidity of the AMD solution were 2.56 and 100 NTU respectively, but the conductivity was not used in the present study.The solid content of the sample was 6.8 g in a 200 mL.The sample contained the following ions: As, Ba, Be, Cd, Co, Cr, Cu, K, Li, Mn, Mo, Na, Ni, Pb, Sb, Se, Si, Tl and Zn.

Coagulants
Pulverized CaCO3 was used in the experiment, where Ca(OH)2 was prepared by heating, that is, decomposing the CaCO3 at 900°C to form CaO and CO2.The CaO was then hydrated to form Ca(OH)2.The concentration of the metal hydroxide of 0.043 M was used to prepare 0.043 M of Ca 2+ in Ca(OH)2, 0.043 M Ca 2+ in CaCO3, or polymers of 0.043 M Ca in PFCl of Ca(OH)2 or CaCO3 dosed in the present study was chosen as per a study which was conducted by Fasemore (2004) on paint wastewater treatment.
The calculation of the mass of metal salt to obtain 0.021 and 0.043 M of M 3+ Ca is as follows: In Table 1 the metal hydroxide dosed into AMD sample is shown.

Jar test procedure
The equipment used for a jar tests was a BIBBY Stuart Scientific Flocculator model (made in Japan), which has six adjustable paddles with rotating speeds between 0 and 250 rpm.The AMD solution containing 6.8 g colloid in 200 mL of the solution was poured in each of the five 500 mL glass beakers for the test.Synthetic polymers were prepared by mixing 0.043 M of FeCl3 and 0.043 M of CaCO3 or Ca(OH)2 to produce af-PFCl polymers of CaCO3 or Ca(OH)2.The coagulants were added in five 500 mL glass beakers containing 200 mL of AMD samples and were mixed in a jar test at 200, 250 and 350 rpm for 2 min, respectively.The samples were allowed to settle for 1, 2 and 6 h respectively after which the pH and turbidity were measured.

Shaking procedure
The equipment used for shaking was a Labotec model 261 (made in South Africa).A similar approach to that in sub-section 3.1 was employed except that the samples were placed in a shaker.

Performance evaluation
The pH was used as a determinant to assess the rate of hydrolysis and hydrolytic potential of the coagulants at different mixing duration, whereas turbidity was used to determine the removal of colloidal particles from the samples.

pH measurement
A MetterToledo Seven Multimeter (made in Germany) pH meter with an electrode filled with silver chloride solution and the outer glass casing with a small membrane covering at the tip was used.
The equipment was calibrated with standard solutions with the pH of 4.0 and 7.0 before use.

Turbidity measurement
A Merck Turbiquant 3000T Turbidimeter (made in Japan) was used to determine turbidity or the suspended particles in the supernatant using NTU as a unit of measurement.It was calibrated with 0.10, 10, 100, 1000 and 10000 NTU standard solutions.

Inductively coupled plasma (ICP)
A Perkin Elmer Optima DV 7000 ICP-OES Optical Emission Spectroscopy (made in USA) was used to determine the metals in the supernatant using ppm as a unit of measurement.It was calibrated with the standard solution between 2 and 50 ppm of the salts mentioned previously.

RESULTS AND DISCUSSION
The addition of the reagent/coagulant to a colloidal suspension gives rise to a number of mixing-induced factors (physico-chemical) some of which include vibrational and collision frequencies; which have not been deliberated on in the previous AMD treatment studies.
The preliminary reaction such as destabilizationhydrolysis is an attribute to the concentration of the H + ions to suppress the pH of the solution.The pH values in AMD samples with af-PFCl polymers of both Ca(OH) 2 and CaCO 3 after 1, 2 and 6 h settling do not show significant change (range of 2.21-2.38).This indicates that protons discharged during hydrolysis and the OH ions released by the Ca(OH) 2 were in equilibrium due to neutralization reaction.The latter were initially dispersed throughout the solution during dissolution reaction and reacted with the protons from hydrolysis reaction.The pH of all the samples with 20 mL dosage is slightly higher than the succeeding samples, which all exhibit a decreasing trend.There are three inferences which have been identified that attribute to chemical reaction of the coagulant to cause destabilization, which are as follows: 1. Destabilization reaction is induced by the release of the energies stored within the colloidal suspension (AMD) after it has been added to the system.2. The size and energy of each colloid contributes to the effectiveness of the destabilization process.3. The distance between the central colloid and charges surrounding it influences the rate of destabilization.4. The distance separating the central colloid and the charges has a correlation with the ionic strength (electronegativity) of the metal ions of a coagulant.These afore-mentioned inferences are influenced by vibrational and collision frequencies to effectively yield an optimal turbidity/TSS removal.Ca 2+ compounds (Ca(OH) 2 or CaCO 3 ) are significantly indispensable in AMD due to their ability to form gypsum.The main problem is that if a plant does not recycle sludge and significant sulphate is present, some precipitation will occur on the walls of the reactor.The gypsum will then accumulate and end up causing a thick layer of gypsum (Aubé and Arseneault, 2003).Such a problem can be resolved by employing High Density Sludge (HDS) processes, a unit which adds to extra costs.
The removal of the pollutants (turbidity) was calculated according to the following formula: Removal efficiency = ×100 (9) Ti and Tf = the initial and final turbidity of the pollutants.
In the present study a conversion of turbidity to TSS was used, because the former is the measurement of the light deflected by the concentration of the particles in the solution, whereas the latter is the measure of the mass concentration of particles in the solution.Equation ( 14) shows that the growing rate of the mass of the floc within a specific time-frame is directly proportional to the surface area and the rate at which the density changes with its radius.This equation has a direct relation with the rate of destabilization-hydrolysis and it expresses nucleation and agglomeration processes.The relationship between Equation ( 10) and the pH changing rate is that there is a varying growth rate between af-PFCl polymers of Ca(OH) 2 and CaCO 3 within a specific time-frame as a function of their surface areas and densities.This can be noticed by the changing pattern of the residual turbidity (Figure 3), which is an indication of the rate of adsorption.Notwithstanding the poor solubility of Ca(OH) 2 or CaCO 3 in water at room temperature, the rate of turbidity removal revealed when electrolytes can destabilize the suspension particles merely through two different mechanisms; charge neutralization and polymer bridging, or both of them simultaneously (Finch, 1996).In polymer bridging, the molecular weight plays a pivotal role, whereas in case of charge neutralization, the charge density plays a pivotal role.Furthermore the presence of charged segments along a polymer chain could have a significant effect on bridging flocculation.
The pH in the AMD sample with af-PFCl polymers of both Ca(OH) 2 and CaCO 3 dosage with mixing after 1 hour settling (Figure 2) decreased from 2.56 to a range of 2.21 to 2.39; the pH after 2 h is in a range of 2.26 to 2.28, and after 6 h is in a range of 2.30 to 2.31.The slight pH changing trend is influenced by both acidic FeCl 3 and basic Ca(OH) 2 or CaCO 3 .Increasing the quantity of the flocculants did not contribute to a significant pH changing trend because of the neutralization effect by the protons and hydroxides released during hydrolysis.It was envisaged that the pH changing trend in the samples with af-PFCl polymer of Ca(OH) 2 to be higher than the af-PFCl of CaCO 3 (Figure 2) due to OH -ions from the Ca(OH) 2; instead it is lower than in the latter, in the ranges of 2.21-2.37 and 2.23-2.31,respectively.It is suggested that the surplus OH -ions reacted with the Fe 2+ (FeS 2 ) to form unstable Fe(OH) 2 precipitates, which are further oxidized (Fe 2+ ) to form Fe(OH) 3 precipitates.The pH exhibits a decreasing trend with increasing dosage of both polymers, indicating that the H + ions released during hydrolysis of Fe 3+ control the pH in the system.The residual TSS in the AMD sample dosed with an af-PFCl polymer of Ca(OH) 2 with 350 rpm rapid mixing after 1 h settling (Figure 3) is relatively high, that is, it is reduced from 235 g to a range of 93.1-98.9g; whereas the TSS in the samples after 2 and 6 h shows a significant decreasing trend.The former yielded residual TSS in a range of 36.0 to 47.7 g, whereas the latter a very small range of 3.69 to 4.54 g.Equation 2shows that Ca(OH) 2 has the potential of removing sulphates from the AMD sample through precipitation to form gypsum.The pH of the AMD sample (2.56) treated in the present study shows a high concentration of FeS 2 , a main pollutant in AMD.Based on the preparatory methods of synthetic polymers, it is suggested that their reaction dynamics which are prevalent during destabilization-hydrolysisadsorption in this study may differ from those in commercial polymers.This inference is due to the fact that the preparation employed with synthetic polymers is just a simple mixing technique, whereas the commercial polymers are prepared in specialized reactors with certain conditions (temperature, pressure, etc).The changing trend of the TSS in the AMD sample after 1 h settling is inconsistent, which indicates that the system was subjected to a high shear force that caused a randomly breakage of flocs.As mentioned about the reaction dynamics of synthetic polymers, it is suggested that the heat mechanical energy emitted during rapid mixing (350 rpm) partially breaks the hydrogen bonds holding the polymer together because the viscosity is not as high as that of the commercial polymers.The dipolarity of neighboring water molecules also has an impact on destabilization-hydrolysis as it keeps the colloids apart, and the net charge of the polymers/polyelectrolyte causes destabilization of the suspension.The mechanism by which a commercial polymer (flocculant) reacts differs from coagulants, because it reacts in three different ways during destabilization-flocculation.That entails coating the colloids, adsorption-and depletion-flocculation.Particles dispersion (mono-, hetero-or poly-dispersion) also plays a role during coagulation-flocculation.All these factors play a pivotal role in the TSS or turbidity removal, hence a reagent/coagulant with metal ions of high electronegativity or valence is required.Apart from the effect of the physico-chemical properties of the colloidwater in the process, mechanical agitation is another essential factor.The high rapid mixing speed causes a rupture on the flocs to a reduced size, which is prone to induce re-stabilization and de-flocculation due to charge reversal at the colloidal surface (Jiang and Graham, 1997;Jiang and Lloyd, 2003;Amuda et al., 2006;Ghaly et al., 2006).Ineffective mixing is also a challenge as it would not disperse the coagulant/flocculant throughout the colloidal suspension.
The samples which settled for 2 h show an improvement in TSS removal by almost 50% as compared to those which settled for 1 hour (93.1-98 9 to 36.0-47.7 g).Such a considerable change is a signal of the re-organization of the flocs into large flocs due to the internal forces within the system.Since there is no mechanical energy applied to the system, it indicates that the van der Waals forces of attraction are greater than the net repulsive forces, which include electrostatic forces of repulsion, ion-ion repulsive and nucleus-nucleus repulsive forces including hydrodynamic forces.The impact of the internal forces within the system is complemented by the residual TTS values attained by the samples which settled for 6 h (3.69-4.54g).The uniform decreasing trend (Figure 3) indicates that all the forces and energies associated with the repulsion were at minimal strength.This also indicates that the zeta potential, which indicates the degree of electrostatic repulsion between adjacent, similarly charged particles in dispersion, was far below van der Waals forces of attraction, a condition which induces agglomeration of the flocs formed during coagulation.The fact is that a high zeta potential will confer stability to a colloidal suspension which consists of molecules and particles that are small enough (solution or dispersion resists aggregation).When the potential is small, attractive forces may exceed this repulsion, and the dispersion may break and flocculate as stated when explaining the residual TSS value of the AMD samples after 6 h settling (Figure 3).Colloids with a high zeta potential are electrically stabilized, while colloids with low zeta potentials tend to coagulate or flocculate (Derjaguin and Landau, 1941).It is postulated that the entropy in the system during rapid mixing (350 rpm) was high and positive, resulting in negative Gibbs energy (spontaneous reaction), a condition which resulted in a high TSS removal after 6 hours settling (Figure 3).This indicates that the spontaneity of the system enabled further nucleation, crystallization and sedimentation to achieve optimal TSS removal.Apart from a further series of reactions which occur during extended settling, the ability of the coagulant/polymer to reduce the equilibrium between counter-forces, Stern and double layers, and induce activation energy from the compression pressure generated during double layer compression is significant.
The inference about the generation of compression pressure during destabilization is based on the DOVL theory as developed by Derjaguin and Landau (1941), stating the existence of disjoining pressure due to explosion of the bubbles during the weakening of the double layer and van der Waals forces in a solution.Notwithstanding that compression pressure does not build up because the system is open to atmospheric pressure, it inevitably causes a rise in temperature which causes an increase in the rate of destabilization reaction.The change in pressure (ΔP) is directly related to the decrease in the double layer surface area (ΔA) (Nishioka et al., 1981a).The combination of Stern and diffuse layers from each pair of neighbouring aqua-colloids in a stable condition gives rise to equilibrium film of thickness, He (Nishioka et al., 1981a).He plays a pivotal role because it occurs on both colloid-colloid and colloidwater, and both bondages store chemical energy.He decreases sharply with increasing concentration of both af-PFCl polymers of Ca(OH) 2 and CaCO 3 as shown by the decreasing trend of residual TSS (Figure 4).The high ionic strengths created by the high ionic density of the polymers with increasing concentration weaken both He the double layer compression.
Unlike commercial polymers, af-PFCl polymers are synthesized in a simplified reaction of mixing acid-free FeCl 3 with Ca(OH) 2 or CaCO 3 , which are more likely to have a different morphological structure.The dissolution of the metal salts is shown by Equations ( 11) and ( 12): According to Equations ( 11) and ( 12), the polymers consist only of both reagents per polymer, which after dissociation form ionic components.Each polymer consists of a combination of 5+ electron valence (Fe 3+ and Ca 2+ ), which gives it a high electronegative potential.This property provides the polymers with a strong destabilization efficiency as stated (Peavy et al., 1985;van Nieuwenhuijzen et al., 2004;Spellman, 2009) that high electron valence possesses a high ability to neutralize ionic charge of the colloid by reducing surface potential of the colloid.The high electron valence (5+) of the polymer indicates that the disruption and regrowth of aggregates resulting from 350 rpm rapid mixing is determined by the strength of the interparticle contacts, the ability of flocculent macromolecules desorbed from the particle surface, and the change conformation in an adsorption layer.It can be concluded that destabilization by af-PFCl polymers occurs mainly through charge neutralization, thus enhancing the rate of coagulationflocculation as stated by Juttner et al. (2000).This statement also concurs with Decher and Schlenoff (2003) when stating that the rate of adsorption is governed by the ionic charge and mass transport.An inference can be derived (Equations 11 and 12) that the rate of the destabilization-hydrolysis-sedimentation with af-PFCl polymer dosage is higher compared to that of the coagulants of the same concentration, mainly due to extra 2+ ionic charges (Equations 11 and 12).The ionic configuration of aqua-colloid (Figure 1), solvated metal ions (Equations 11 and 12), and the electronegativity of the metal ions of the salt or hydroxide determine the effectiveness of the destabilization-hydrolysis-adsorption.The metal ions, similarly to colloids, are surrounded by primary and secondary water molecules occupying "1 st and 2 nd solvation shells".The monomers are then formed during hydrolysis which results in the formation of polymers during the continuous replacement of water molecules by the ligands, L (Equations 13).This replacement starts on the 1 st solvation shell to form a complex which is described as an inner-sphere complex, [ML] (p-q)+ (Atkins and de Paula, 2006) as shown by Equation ( 13).It is formed by a direct contact ion pair which increases in size to form larger flocs of stable polymeric af-PFCl complexes.These polymers (Equation 13) are adsorption substrates and adsorb the sulphates and heavy metals from the AMD sample.
The rate of adsorption of the complex (Equation 13) depends on the interrelation between the colloidochemical parameters of the particles, molecular mass and charge density of a polymer and susceptibility of the flocs to disruption and subsequent regrowth (Baran and Teslenko, 1996), supporting the results obtained in Figures 3 and 6.It is suggested that the behaviour of the flocs in the course of disruption/regrowth depends on the mechanism of particle aggregation because it is perceived that flocculation is induced by neutralization of surface charges as shown by Equations ( 11) and ( 12) (Baran and Teslenko, 1996).Destabilization by polymers is not limited to neutralization as particle bridging (Gregory and Duan, 2001;Goldberg, 2002;Duan and Gregory, 2003).It is suggested in this study that particle bridging continues during agglomeration.Aggregation/crystallization is another process which has never been explicitly elucidated physico-chemically.It is therefore suggested that intensive and extended rapid mixing are not ideal methods in wastewater treatment using polymers as that cause rupturing of the chain of particles bonded by polymers (Aguilar et al., 2002;Heath et al., 2002;Binnie et al., 2003;Aguilar et al., 2005).Dehydration is one of the reaction depicted as a principal key player during the aggregation/crystallization process and has not been investigated during the AMD treatment.However, Maree (2003) revealed that the rate of gypsum crystallization is directly proportional to the surface area of the gypsum.The author even corroborated the statement by illustration of the following equation: A postulate is added to the expression that the rate of dehydration of the water surrounding the metal ions determines the rate of agglomeration/crystallization (flocculation).The prediction is based on the inference stated in literature (Atkins and de Paula, 2006) about the 1 st and 2 nd solvation shells.It has been mentioned that the 1 st solvation shell is active during hydrolysis, of which another inference from this study states that the 2 nd solvation is active during aggregation/crystallization.This process plays a vital role in the removal of turbidity or TSS.
The residual TSS (Figure 4) in the AMD samples with af-PFCl polymer of Ca(OH) 2 dosage with 200 rpm rapid mixing and settling for 1, 2 and 6 h respectively, shows lower values after 1 h compared to the corresponding AMD samples with 350 rpm rapid mixing (Figure 3).This concurs with the statement by Swartz and Ralo (2004), Amuda et al. (2006) and Ghaly et al. (2006) that rapid mixing causes rupturing of the flocs which results in the re-stabilization or de-flocculation reactions.This indicates that re-stabilization is not only caused by over-dosing as stated by Swartz and Ralo (2004) and Aboulhassan et al. (2006), but extended and intensive rapid mixing as well as overdosing.Such a changing trend is shown by the AMD samples which settled for 2 and 6 h respectively.The effect that various particle sizes on the destabilization-neutralization rate was also deliberated in this study.The lower residual TSS values in the samples which settled for 2 and 6 h indicate that the surface area of the primary flocs plays a pivotal role during agglomeration/crystallization.This is explained by the effect of the increasing size of the flocs due to the velocity gradient and differential velocity, where the rate of adsorption is optimized.It is suggested that the higher valence cations also play a major role by forming bridges between anionic groups of different macromolecules and thereby enhance their associations.This allows the macromolecules to approach one another more closely, so that optimal intermolecular attractive forces predominate and coagulation or precipitation can occur.It is suggested that the optimal removal of TSS is due to the effect of hydrolysis and polymerization by forming gelatinous charged hydroxo-cationic complexes, which are able to remove pollutants through adsorption and charge neutralization as stated by Barash et al. (2009) and Ge et al. (2004).The metal hydroxide flocs are produced with a large surface area to enhance adsorption.These valuable characteristics are very useful for the removal of non-settleable pollutants by rapid adsorption of soluble organic compounds and trapping of colloidal particles.The extended settling allows easy removal of these flocs from the effluent by sedimentation or flotation (Roa-Morales et al., 2007;Dehghani et al., 2011;Farhadi et al., 2012).
Solubility in wastewater treatment is another factor which plays a pivotal role as it deters the solubilisation of the metal ions component in the precipitates (adsorbent) formed during hydrolysis.However, it is more unlikely to play any significant role in commercial organic flocculent, because they are in a gelatinous form.Although the synthetic inorganic polymers are not entirely gelatinous, it is alleged that solubility plays an inconsiderable role in their effectiveness (agglomeration/crystallization).The effectiveness of the polymers in coagulation depends on the change in ionic concentration and increases exponentially as the charge of the ions in the system increases (Binnie, 2003), which supports the results obtained in this study.Moore and Moore (1976) recommended the colloidal charge in most wastewater coagulation requires positively charged inorganic polymers.They react with wastewater through neutralization or particles bridging, whereas the coagulant undergoes hydrolysis, which are in the gel form.On the contrary, the inorganic polymers are in either aqueous or semi-gelatinous form and react with wastewater in a similar manner as the inorganic coagulants as shown in the study conducted by Ntwampe et al. (2013) using paint wastewater.The solubility does not play a prominent role during acidic wastewater or AMD treatment using inorganic flocculants/polymers (Ntwampe et al., 2015).Re-stabilization and deflocculation are predominant reactions which occur during AMD treatment using inorganic polymers as a result of the shift in equilibrium.Effective dispersion of the coagulant/flocculent throughout the colloidal suspension without causing rupturing of the flocs can optimize the removal of turbidity.
The 200 rpm rapid mixing experiment also showed a uniform decreasing trend with increasing concentration of Fe 3+ and Ca 2+ in the af-PFCl polymer, except in a few samples after 6 h.It is postulated that such deviations might be caused by the influence of the atmospheric gaseous conditions (O 2 and CO 2 ).The overall uniformity of the changing trend shows that the reactions within the AMD samples with 200 rpm rapid mixing occurred at a steady state without interruptions (chemical or mechanical).
The residual TSS in the AMD samples with 200 rpm shaking (Figure 5) shows a similarly equal changing values and pattern compared to the corresponding samples with 200 rpm rapid mixing (Figure 4).The residual TSS changing trend in the AMD samples with 200 rpm mixing after 1, 2 and 6 hours (Fig. 4) is in a range of 93.1-98.9,35.9-35.7 and 3.7-4.5 g respectively, whereas in the AMD samples with 200 rpm shaking (Figure 5) is in a range of 45.2-57.8, 20.9-33.4 and 3.5-4.8g respectively.As indicated that one of the objectives in the present study is to determine the possibility of replacing rapid mixing with shaking, this changing trend (Figures 4 and 5) corroborates our investigation.
The residual TSS in the AMD samples with af-PFCl of CaCO 3 dosage rapid mixing at 350 rpm (Figure 6) shows a slightly lower changing trend, but similar changing pattern, that is, settling for 1 h is high, 2 h low, and 6 h the lowest.It is expected from the chemistry perspective that the pH of the samples with af-PFCl polymer of CaCO 3 would be reduced due to the CO 2 as shown by Equation ( 15): It is suggested that the increased buffer capacity of the CO 2 contributes to stabilizing the pH of the water, including the assumption that a fraction of CO 2 evaporated due to an open system.
The values after 1 hour are in a range of 82.9-89.5 g, after 2 h is in a range of 37.5 to 44.9 g, and 6 h is in a range of 3.59 to 4.93 g.Similarly, the reaction by CaCO 3 reagent to the sulphates which is formed during the oxidation of FeS 2 , also form gypsum as Ca(OH) 2 as shown by Equation (3).A similarly equal residual TSS in AMD samples with af-PFCl polymers of Ca(OH) 2 and CaCO 3 , give rise to a conclusion that the metal salt (FeCl 3 ) and the ligands (Ca(OH) 2 and CaCO 3 ) behaved independently.The slight difference shown between the two polymers were influenced by the concentration of the ligands in the AMD during hydrolysis.The common denominator between the two polymers is the reaction of  Fe 3+ and Ca 2+ ions in a polymer during destabilizationhydrolysis. The hydrolysis of both polymers is identical, that is, the speciation occurs with the release of protons to cause pH suppression.Invariably, hydrolysis of Ca 2+ occurs in a similar pattern, except for the CO 2 released in the reaction of a polymer of CaCO 3 .Presumably, the pH in the AMD should decrease due to the release of CO 2 , this then indicates that it evaporated to the atmosphere in a gas state.
The residual TSS in the AMD sample with af-PFCl of CaCO 3 dosage with 20 rpm rapid mixing after 1 h settling (Figure 7) is lower compared to the corresponding samples with 350 rpm rapid mixing after 1 h settling, in a range of 45.2 to 57.8 g.Invariably, this changing trend occurred in the corresponding samples after 2 and 6 h, in a range of 20.9-33.4 and 3.6-4.8g respectively.
The residual TSS in the AMD sample with af-PFCl of CaCO 3 dosage (the sum of the mass of Fe 3+ and Ca 2+ in the af-PFCl) with shaking (Figure 8) after 1 h settling is slightly lower than the corresponding samples with 350 rpm rapid mixing and 200 rpm rapid mixing (Figures 6  and 7), in a range of 45.1 to 51.9 g.On the other hand,  the residual TSS in the corresponding samples with 200 rpm shaking after 2 and 6 h is also lower than both 350 rpm rapid mixing and 200 rpm shaking in a range of 20.6-33.4 and 3.6-4.8g respectively.A firm conclusion is drawn based on the experimental results comparing the residual TSS between the corresponding samples with rapid mixing and shaking, that the latter method can be an ideal replacement of the former.
The experimental results of the residual TSS in AMD sample with 0.021 and 0.043 M of Ca(OH) 2 dosage (Figure 9) with mixing and shaking show that the 0.021 M Ca(OH) 2 dosage yielded lower pH values and higher residual TSS in both mixing and shaking.All the samples (mixing and shaking) with 20 mL of Ca(OH) 2 dosage showed lower pH values and higher residual TSS values (3.05-3.23 and 49.6-54.3g respectively) compared to the samples with 30-60 mL dosages.The pH and residual TSS are in a range of 3.60-6.56and 41.1-49.3g respectively, excluding the samples with 20 mL dosage.On the contrary, the corresponding samples with 0.043 M Ca(OH) 2 dosage with both mixing and shaking yielded relatively higher pH and lower residual TSS.The pH and residual TSS values are in a range of 5.10-9.99 and 4.7-7.99g respectively, excluding the samples with 20 mL dosage.However, the samples with shaking still exhibit similarly equal residual TSS compared to the samples with mixing.This corroborates the investigation to determine the possibility of replacing the mixing with shaking.
Both SEM images of the af-PFCl of Ca(OH) 2 and CaCO 3 ) with shaking (Figure 12) exhibit a crystal morphological matrix with identical intermolecular bridging pattern, that is, the flocs are interconnected in clusters which are separated by solid structure (more on the right hand image).Although they both yielded almost identical TSS removal in corresponding samples, the crystal morphological matrix indicates that they are prone to show a dissimilar performance in wastewater with high TSS.This inference is based on the fact that the image of af-PFCl of Ca(OH) 2 (left) shows an even distribution of small and larger flocs, whereas the image of af-PFCl of CaCO 3 (right) shows small and large flocs at the bottom half of the image, whereas the top part is occupied by a solid cake-like structure.The SEM image (Figure 13) of af-PFCl of Ca(OH) 2 (left) and CaCO 3 (right) with mixing shows a distinctive crystal morphological matrix between the two images.The image on the left (Figure 13) shows widely distributed flocs with intermolecular bridges clustered unevenly, whereas the image on the right shows clusters of intermolecular bridges with compacted flocs in a block shape.The image depicts that there was a greater force of attraction which held the flocs together with some cavities in-between.
Both the samples dosed with af-PFCl polymer of Ca(OH) 2 or CaCO 3 with shaking (Figure 14) show copper (Cu) at 2000 counts and some metals at very low peaks.Both polymers show identical peaks, which indicates that their turbidity removal efficiencies are identical.There are some minor unidentified peaks which appear around 1020 and 1030 counts.
The sample dosed with af-PFCl polymer of Ca(OH) 2 or CaCO 3 rapid mixing (Figure 15) shows Cu counts of 1500, whereas it is at 2000 in the samples dosed with Mg(OH) 2 rapid mixing.The latter performed similarly to the polymers in Figure 10.Similarly, there are some minor unidentified peaks which appear around 1020 and 1030 counts.
Pearson correlation coefficient (r) is used to calculate the relation between pH and residual turbidity using Equation (15).The results obtained with AMD sample with af-PFCl of Ca(OH) 2 and CaCO 3 mixing respectively (Figures 10 and  11), yielded the r-values of 0.926 and 0.626 (92.6 and 62.6.3%)respectively.The range of the correlation coefficient is from -1 to 1.Our correlation coefficients in samples with both mixing and shaking fall within a range of strong and very strong relationship.This is validated by R 2 of the pH vs E % of those polymers with mixing as it is  shown in Figures 11 and 12 of 0.969 and 0.941 (96.9 and 94.1%) respectively.

Conclusion
The experimental results revealed that the synthetic acid free PFCl of Ca(OH) 2 or CaCO 3 exhibit a high TSS removal efficiency which is comparable to costly commercial organic or inorganic polymers.Both polymers

Figure 1 .
Figure 1.A structure of an aqua-colloid showing the layers.

Figure 2 .
Figure 2. pH in AMD with af-PFCl of Ca(OH)2 or CaCO3 at varying settling times with mixing.1pH = pH after 1 h, 2pH = pH after 2 h and 3pH = pH after 6 h.
d[CaSO 4 •2H 2 O]/dt = rate of crystallisation, k = reaction rate constant, [CaSO 4 •2H 2 O](S) = surface area of the seed crystals, C = initial concentration of calcium sulphate in solution and C 0 = saturated concentration of calcium sulphate in solution.

Figure 9 .
Figure 9. pH and residual TSS of AMD sample with 0.021 and 0.043 M Ca(OH)2 with mixing and shaking.
0.70 or higher is very strong relationship, 0.40-0.69 is strong relationship and 0.30-0.39 is moderate relationship.The results of the experiments are used in the calculations as follows: Σx = sum of pH, Σy =sum of turbidity removal efficiency, Σx 2 = sum of the square of pH, Σy 2 = sum of the square of turbidity removal efficiency and Σxy = sum of product of pH and turbidity removal efficiency, af-PFCl-Ca(OH) 2 = acid-free polymer of FeCl 3 -Ca(OH) 2 and af-PFCl = acidfree polymer of FeCl 3 -CaCO 3 .

Figure 10 .
Figure 10.pH vs E % of AMD sample with af-PFCl of Ca(OH)2 mixing.

Figure 11 .
Figure 11.pH vs E % of AMD sample with af-PFCl of CaCO3 mixing.

Figure 12 .
Figure 12.Comparison of SEM images between the sludge of the AMD with FeCl3-Ca(OH)2 and FeCl3-CaCO3 dosage with shaking (25000x).

Figure 13 .
Figure 13.Comparison of SEM images between the sludge of the AMD with CaCO3 and Ca(OH)2 with mixing (25000x).

Figure 14 .Figure 15 .
Figure 14.Comparison of XRD graphs between the sludge of the AMD with FeCl3-Ca(OH)2 and FeCl3-CaCO3 dosage with shaking.

Table 1 .
Metal hydroxide dosed into AMD sample.