- Email: [email protected]
Enhanced removal of heavy metal ions bound to humic acid by polyelectrolyte flocculation
Separation and Purification Technology 51 (2006) 48–56
Enhanced removal of heavy metal ions bound to humic acid by polyelectrolyte flocculation Nicholas P. Hankins a,∗ , Na Lu b , Nidal Hilal b a
b
Department of Engineering Science, The University of Oxford, Parks Road, Oxford OX1 3PJ, UK School of Chemical, Environmental and Mining Engineering, University of Nottingham, University Park, Nottingham NG7 2RD, UK Received 16 October 2005; received in revised form 14 December 2005; accepted 16 December 2005
Abstract The enhanced removal of heavy metal ions from solution, such as Pb2+ and Zn2+ , was studied by binding the ions to humic acid (HA) and then coagulating–flocculating with the cationic polyelectrolyte polydiallyldimethylammonium chloride (PolyDADMAC). The effect of the dosage of PolyDADMAC, the pH level and the concentrations of HA and metal ions were studied. Ultrafiltration was used to separate bound metal ions from free ions in solution. The removal of bound metal ions was found to increase with the extent of coagulation–flocculation of the HA by PolyDADMAC. The pH affects the removal efficiency of bound (complexed) metal, in so much as it affects the binding strength of metal ions to the HA. The removal efficiency of metal also increases with the initial concentration of HA. The effective coagulation–flocculation region of the HA by PolyDADMAC is affected by the initial concentration of the metal ions; an increase in the concentration leads to a decrease in the amount of PolyDADMAC required. Humic substances have the advantage of being naturally occurring, and the results indicate that such a complexation–flocculation process is of potential interest for the removal of heavy metals during water treatment. © 2006 Elsevier B.V. All rights reserved. Keywords: Humic acid; Heavy metal; Coagulation; Flocculation; Flocculant; Cationic polyelectrolyte; Complexation–flocculation; Waste water treatment
1. Introduction As major components of natural organic matter, humic substances (HS) are involved in many chemical and physicochemical interactions in soil and aqueous systems [1]. HS include humic and fulvic acids, and have a wide range of molecular weights, with both hydrophilic and hydrophobic moieties. The interactions of HS with heavy metals have already been established and studied [2]. The binding of heavy metal ions by HS is controlled mainly by the pH, the concentrations of the HS and heavy metal ions and by other parameters [3–6]. Ultrafiltration and other methods have been used to study the binding of metal ions to HS [3,7]. In the current work, the soil-derived humic acid (HA) material used had a sufficiently large molecular weight to be trapped by ultrafiltration; on the other hand, if aquatic HA material is used, a normal UF membrane would trap only about half the material and a nanofiltration membrane might well be required.
∗
Corresponding author. Tel.: +44 1865 273 027; fax: +44 1865 283 273. E-mail address: [email protected] (N.P. Hankins).
1383-5866/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2005.12.022
Because of their appreciable solubility in water, HS may be regarded as anionically charged macromolecular colloids (as opposed to colloidal dispersions, which are generally insoluble). As such, HS can be removed by coagulation–flocculation processes; this can be achieved by the addition of an inorganic salt and/or or a cationic polyelectrolyte (PE) [8]. In the context of the present work, coagulation essentially refers to charge neutralization and hence solute destabilisation, while flocculation refers to the process of floc growth whereby more and more macromolecular material becomes incorporated into the growing flocs. For instance, the removal of HS from upland waters during potable water treatment is frequently carried out by the addition of hydrolysed ferric or aluminium salts; in this case, the removal occurs through a combination of charge neutralization and possible enmeshment (also known as ‘sweep flocculation’). In the case of polyelectrolyte addition, the mechanism of removal has not yet been fully elucidated or agreed upon, but it seems likely to be dominated by charge neutralization [9]; at optimum dosage, the resulting HS–PE complex is uncharged and highly hydrophobic, and so can easily grow (flocculate) into large flocs. Note, however, that the initial complexes are macromolecular in nature, and cannot easily be characterised for zeta
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
potential, charge density or hydrodynamic size, in the same way as for colloidal particle dispersions. The role, if any, of polymer bridging by PE remains uncertain. For the coagulation–flocculation process, a PE with a higher charge density leads to a higher removal efficiency, while molecular weight may have little influence [9]. The actual removal efficiency of HS by PE depends on the dosage of PE, the pH and the other flocculation conditions [8,10–12]. Given that HS can bind with other pollutant species in aqueous solution and then be removed by coagulation–flocculation, HS can be considered as a separating agent to remove bound pollutants in conjunction with PE coagulation–flocculation. However, such a complexation–flocculation process, involving HS as a substrate and PE as a coagulant–flocculant, has received little prior attention for the removal of heavy metals during water treatment. HA has been used to remove bound organic contaminants (OC) including pyrene, fluoranthene and anthracene, by coagulation–flocculation with Al(III) and Fe(III) salts at pH 6 [13,14]. The removal of OC occurs simultaneously with coagulation–flocculation of the OC–HA system. The removal efficiency of OC depends mostly on the stability of the OC–HA complexes. Cationic PEs have also been used as flocculants to remove cations, anions and organic pollutants bound to negatively charged organic chelating agents or surfactants. Mixtures of PolyDADMAC and 4,5-dihyhroxy-1,3-benzenedisulphonic acid (Titron) were found to be effective in removing cations, such as Cu2+ and Pb2+ , while expelling non-complexed alkalineearth ions such as Ca2+ [15]. Mixtures of PolyDADMAC and diethylenetriaminepentaacetic acid (DTPA) were discovered that can remove divalent cations, such as Cu2+ and Pb2+ , and anions such as CrO42− at pH values higher than 4, and can also expel interfering cations, such as Ca2+ [16]. In the complexation–flocculation process, the effect of additional coagulant–flocculant on the stability of the HS complexes needs to be taken into account. For the coagulation–flocculation of HA with inorganic substances to remove bound organic compounds, since Al(III) or Fe(III) and the organic compounds react with different sites on the humic material, a competition between them is not expected. The bound organic compounds in the complex are, in effect, precipitated, flocculated and entrapped in Al or Fe flocs in the same way and to the same extent as the HA itself. On the other hand, inorganic coagulant–flocculants such as aluminium and other trivalent electrolytes may have significant effects on the stability of complexes of humic acid and bound heavy metals. The competitive effect of Al(III) and other trivalent electrolytes on HS–La(III) complexes has been assessed [17]. The possibility of significant competitive effects was confined to Al(III). The amount of complexed (bound) La at pH 4 decreased significantly after an equilibration time of 48 h if the metals were introduced in two steps, such as adding Al to La + HA solution, or adding La to Al + HA solution, or adding a mixture of Al and La to the HA solution. Thus, inorganic salts are not suitable for removing heavy metal ions bound to the HS by coagulation–flocculation.
49
The effect of additional metal ions on HS coagulation– flocculation also needs to be considered. On the one hand, a competition between Al(III) or Fe(III) coagulant–flocculant and organic pollutant is not expected during complexation of the latter to the HS, so the optimum Al(III) or Fe(III) dosages for the complexation–flocculation process are expected to be the same as the optimum dosages for simple HS coagulation–flocculation. On the other hand, in tests of the coagulation–flocculation of HA by Al(III) in the presence of bound La(III), although the HA removal efficiency can be kept at the same level, the optimum dosage of Al(III) actually decreased with participation of La(III) [17]. Using the polyelectrolyte flocculant PolyDADMAC, Tuncay et al. also found that the removal efficiencies of DTPA and Titron become higher at lower ratios of DTPA to PolyDADMAC or Titron to PolyDADMAC [15,16]. The simultaneous removal of HS and bound heavy metal ions by complexation–flocculation has not been studied previously. Because of the chemical features of polyelectrolyte (PE) as a coagulant–flocculant, it has a good propensity for HS removal. Moreover, there is no evidence of it having a strong competitive effect on metal–HS binding. Therefore, its application to the complexation–flocculation process for the removal of bound heavy metal ions in water treatment looks promising. Such a process has received no previous attention. In a report released by the U.S. EPA, heavy metals such as lead, cadmium, mercury, chromium, copper, zinc and nickel have been placed at the top of a priority pollutant ranking [18]. The interactions between HS and trivalent metals have been studied [19], and zinc and lead were chosen as typical bivalent metals. In this work, the effect of a number of factors is studied, and Zn and Pb are used as model heavy metals. Poly(diallyldimethylammonium) chloride, PolyDADMAC, is used as a high charge cationic PE. It has been shown to have a stable, highly positive charge over a wide pH range [20]. 2. Materials HA solution was prepared as follows: 4 g of HA (Aldrich, soil derived and used as the sodium salt) were dissolved in 20 mL of 1 M NaOH, then diluted into 1.8 L of deionised water and stirred for 2 h. The pH was initially adjusted to 7; at this value of pH, the HA was soluble. No precipitation of HA was observed during any test in which pH was subsequently adjusted to a lower (or higher) value. The total volume was adjusted to 2 L, and the solution was stirred for an additional 3 h. The solution was finally filtered through Whatman GF-A filter paper, followed by filtration through a 0.45 m membrane (Millipore). HA concentrations were determined throughout by measuring the UV absorption at a wavelength of 254 nm in a Jenway 6405 UV/vis spectrophotometer (note that preliminary tests of total organic carbon in solution showed an agreement with UV absorption measurements of HA, indicating an absence of nonabsorbing compounds). Aqueous solutions of Zn and Pb were purchased from Sigma–Aldrich. One percent PolyDADMAC solutions were prepared daily by diluting 20% commercial PolyDADMAC product (Sigma–Aldrich; medium molecular weight around 50 000, with a high charge density around 5–6 mequiv./g)
50
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
with deionised water. Analytical grade reagents were employed throughout the experiments. Milli-Q plus (Millipore) water was used in preparing all solutions. A stirred ultrafiltration Cell (Millipore, Model 8010, 10 mL) was used, in conjunction with a YM1 membrane with a cut-off of 1 kDa (Millipore). A Waterproof pH Tester (3 Double Junction, OAKTON Instruments) was employed for pH determination. A floc-tester (Bibby Stuart flocculator SW6) was used for flocculation jar testing. An Orbitar incubator-shaker was used to bring solutions to equilibrium during binding measurements. Adjustments in pH were carried out with HCl or NaOH. All the experiments were performed at an ambient temperature of 22 ◦ C. 3. Methods 3.1. HA–metal binding determination The binding between HA and heavy metals was determined in batch studies. A series of solutions with a fixed or variable HA concentration and a fixed or variable Pb2+ or Zn2+ concentration were prepared as feed solutions. Hundred milliliters of feed solution were placed in a 100-mL conical flask and closed with lab foil. Finally, the pH of the solutions was adjusted to a fixed or variable value in the range between 3 and 9. Prior to ultrafiltration to remove the bound ions, solutions were kept shaking at 25 ◦ C for 12 h to reach an equilibrium state for the HA–metal system. Then 10 mL of sample were ultra-filtered, with a stirring rate of 300 rpm and an operating pressure of 50 psi. This ensured that only free (non-bound) ions were present in the permeate: measurements of metal removal thus measured the true effects of binding (complexation). The concentration of metal ions in the permeate solution was measured by a flame atomic absorption spectrophotometer (AAS). 3.2. HA–PolyDADMAC coagulation–flocculation tests A series of standard jar-test experiments were carried out to determine the required dosage of PolyDADMAC to efficiently precipitate HA. A series of 250 mL HA samples with a fixed or variable concentration were placed in 500 mL beakers, and the pH adjusted to a fixed or variable value in the range between 3 and 9. One percent PolyDADMAC solution was added during fast-stirring at 250 rpm for 2 min, in variable mass ratios of PolyDADMAC to HA, followed by slow stirring at 30 rpm for 10 min. The pH level was maintained at the desired level during fast-stirring. After overnight settling, 40 mL of solution were centrifugally separated at 4000 rpm for 20 min. The residual HA content in the supernatant was determined by UV absorption at 254 nm.
desired final concentrations of each, were prepared by mixing the appropriate amounts of HA and metal ion stock solution and diluting with deionised water. The pH was adjusted to a fixed or variable value in the range between 3 and 9. One percent PolyDADMAC solution was added during fast-stirring at 250 rpm for 2 min, in variable mass ratios of PolyDADMAC to HA, followed by slow stirring at 30 rpm for 10 min. The solutions were shaken at 25 ◦ C for 12 h to reach an equilibrium state of complexation. After overnight settling and centrifugation of 40 mL of solution at 4000 rpm for 20 min, 10 mL of supernatant were ultra-filtered through a 1 kDa membrane, as in 3.1, to ensure that only free (unbound) ions were present in the permeate. The concentration of metal ions in the permeate solution was measured by AAS, and HA concentration in the supernatant was determined by UV absorption at 254 nm. The conditions for the complete set of tests are summarized in Table 1. 4. Results and discussion 4.1. HA–metal binding determination Since the binding between HA and metal ions is the basis for the complexation–flocculation process, the effects of pH and the ratio of HA to metal ions on the binding were studied. 4.1.1. Effect of pH on HA–metal binding According to Fig. 1, the removals of Zn and Pb are lower than 50% at pH 3, since the major functional groups of HA responsible for the metal binding (weak acidic groups, which dissociate into negatively charged forms) have low dissociation at low pH levels. This inhibits the complexation. The quantities of bound metal ions are reduced, as a result of proton competition. With an increase in pH level, the removal of metal ion is improved. The dissociation of HA functional groups increases with the pH level, leading to an increase in negative charge density and hence in the binding of metal ion to HA.
3.3. Complexation–flocculation tests The mechanism of HA–metal complex removal by the complexation–flocculation process was tested using a series of flocculation jar tests, followed by centrifugation and ultrafiltration. A series of 250 mL solutions of HA and metal ions, of the
Fig. 1. The effect of pH on metal removal by binding during the complexation of HA and metal ions. Initial concentration of HA is 20 mg/L; initial concentration of metal ions is 5 mg/L.
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
51
Table 1 Experimental conditions for tests Test
HA (mg/L)
Pb2+ /Zn2+ (mg/L)
pH
Mass ratio PolyDADMAC:HA
Effect of pH on the metal binding Effect of HA–metal ratio on the metal binding Effect of PolyDADMAC dosage and pH on HA removal Effect of HA concentration on HA removal Effect of PolyDADMAC dosage on heavy metal removal Effect of pH Effect of HA concentration Effect of metal ion concentration on the ion removal
20 20 20 10, 20, 40 20 20 10, 20, 40 20
5 0.6–20 – – 1.2 1.2 1.2 1.2, 2.5, 5
3–9 7 3–9 7 7 3–9 7.0 7.0
– – 0.1–0.5 0.1–0.5 0.1–0.5 0.1–0.5 0.1–0.5 0.1–0.5
Thus, the pH level significantly affects the binding of metal ions to HA. 4.1.2. Effect of the HA–metal ratio on the metal removal after ultrafiltration Fig. 2 presents Zn and Pb-HA binding tests, which were made at pH 7 under the same conditions. The removal for Pb is higher than that for Zn. Between the two bivalent ions, Pb has a larger ionic radius than Zn. But Zn2+ has higher ionic potential, is easier to hydrolyse, and exists as monovalent (ZnOH)+ , which consequently binds less strongly than bivalent lead. For a given HA concentration, Fig. 2 shows there is little change in metal ion removal as the HA–metal ratio varies above 1 or 2; presumably, an excess of binding sites are already available. From these results, we can calculate the maximum binding ratio between heavy metal and HA on a mass basis: for lead this is 0.98, and for zinc this is 0.65. Binding constants of various metal–HA complexes have been reported [21]. A descending series of log K values may be presented as follows: Fe3+ (7.5) > Al3+ (5.1) > Cu2+ (4.6) > Ni2+ (4.4) > Co2+ (4.2) > Pb2+ = Ca2+ (3.5) > Zn2+ (3.1) > Mn2+ (2.8) > Mg2+ (2.6). Pb2+ has a stronger binding ability than Zn2+ , so Pb2+ removal is more effective than Zn2+ .
Fig. 2. The effect of HA–metal ratio on metal removal by binding during complexation of HA and metal at pH 7. HA initial concentration is 20 mg/L.
4.2. HA–PolyDADMAC coagulation–flocculation tests 4.2.1. Effect of PolyDADMAC dosage and pH on HA removal The coagulation–flocculation of HA by PolyDADMAC is expected to be dominated by charge neutralization [9]; the ‘bridging’ plays an uncertain role. The process mainly depends on the ionic dissociation of HA, and on the oppositely charged character of PolyDADMAC. As shown in Fig. 3, with a fixed HA concentration of 20 mg/L and pH 7, PolyDADMAC is effective in removing HA when the mass:mass ratio lies in the range between 0.3 and 0.4. The flocculation curve has a classical shape, comprising underdose (under neutralized), optimum dose (neutralized) and overdose (re-charged) regions. Similar results have been reported in previous work [8,22]. In the underdose region at lower concentrations of PolyDADMAC, some charge neutralization occurs, but this is not sufficient to allow aggregation and hence the creation of flocs large enough to settle. In the optimum dose region, the surface charge of the HA is significantly neutralized by the PolyDADMAC; the resulting HA–PolyDADMAC complex is thus rendered highly hydrophobic in an aqueous medium, floc growth can then occur readily. As
Fig. 3. Removal of HA by PolyDADMAC flocculation at different pH levels. HA initial concentration is 20 mg/L.
52
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
a cationic, medium molecular weight polyelectrolyte, polymer bridging may possibly also assist in forming flocs large enough to settle. In the overdose region, with a ratio higher than 0.4, the removal efficiency again decreases. The HA molecules are surrounded by excess PolyDADMAC, leading to high charge and both electrostatic and steric repulsion, which in turn leads to stable dispersion of the HA. Fig. 3 also presents the removal efficiency of HA from the supernatant after coagulation–flocculation, as a function of the mass dosage ratio of PolyDADMAC:HA, at different pH levels. For all the pH levels tested, the highest removal efficiencies are above 85%. The optimum PolyDADMAC dosage ratios for different pH levels are shown in Table 2. Table 2 clearly shows that the optimum mass dosage ratio increases with the pH level. Since the charge density of PolyDADMAC is relatively constant at varying pH [20], the increasing dissociation of HA functional groups at increasing pH, and hence the negative charge density of the HA, is the main factor in controlling coagulation–flocculation. At low pH, the low dissociation of HA yields a low charge density, and it can be neutralized by a smaller amount of PolyDADMAC. Also, it might also be speculated that PolyDADMAC has a good polymer bridging ability, so that the flocculation is still effective at low pH values. In a similar fashion, less PolyDADMAC is required to redisperse the HA at low pH. As pH is increased, the HA becomes more dissociated, and a larger amount of the positively charged PolyDADMAC is required to neutralize the more negatively charged HA. The actual removal efficiency at optimum dosage shows no significant difference as pH level varies. 4.2.2. Effect of HA concentration on HA removal Fig. 4 demonstrates the removal efficiency of HA after coagulation–flocculation from the supernatant, as a function of the dosage mass ratio of PolyDADMAC to HA, with different initial HA concentrations at pH 7. The highest removal efficiencies, with varying initial concentrations of HA, are all above 85%. The effective range becomes broader at the higher ratio end with an increase in HA concentration. In solutions with higher initial concentrations of HA, it might be speculated that bridging flocculation can occur more easily under shear at the same dosage ratio of PolyDADMAC to HA. In such solutions, the total amount of PolyDADMAC and HA are higher, such that bridging causes flocculation to occur over a broader region of dosage ratios.
Fig. 4. Removal of HA by PolyDADMAC at pH 7 with different initial concentrations of HA: (a) 10 mg/L, (b) 20 mg/L, (c) 40 mg/L.
4.3. Complexation–flocculation tests 4.3.1. Effect of PolyDADMAC dosage on heavy metal removal Figs. 5 and 6 show the removal efficiency by binding for Pb2+ and Zn2+ , following both complexation and flocculation (and finally centrifugation and ultrafiltration of the supernatant), as a function of the dosage mass ratio of PolyDADMAC to HA. The removal efficiency for metal ions remained at a lower level (95% and 62%, respectively) in the low PolyDADMAC dosage region (0–0.1), it increased (to 99% and 72%, respectively) in the effective range of coagulation–flocculation (0.25–0.3), and decreased back to a lower level (90% and 62%, respectively) in the overdose range (0.35–0.5). Because metal ions and PolyDADMAC are both positively charged, an interaction between them is not expected. The increase in removal efficiency must therefore occur because of an improvement in the complexation
Table 2 Optimum mass dosage ratio of PolyDADMAC to HA for HA removal at different pH values pH
Mass ratio of PolyDADMAC to HA
Removal efficiency (%)
3 5 7 9
<0.1 0.2–0.3 0.3–0.4 0.35–0.45
90 92 87 89
Fig. 5. Removal of Pb2+ and HA at pH 7. Pb2+ initial concentration is 1.2 mg/L. HA initial concentration is 20 mg/L.
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
53
Fig. 6. Removal of Zn2+ and HA at pH 7. Zn2+ initial concentration is 1.2 mg/L. HA initial concentration is 20 mg/L.
of metal ions to the functional groups of the HA during flocculation. The interaction of PolyDADMAC with the functional groups of the HA molecule partially or completely neutralizes its surface charge; under this altered electrostatic environment, it might be speculated that it would lead to a transformation in the overall molecular structure of the HA (analogous to tertiary structure in a protein molecule) and a change in the spatial distribution of the HA surface functional groups responsible for binding heavy metal ions. In essence, the structure ‘opens up’, and the functional groups become more sterically accessible to the heavy metal ions. It could then be hypothesized that structural changes lead to an enhancement in the binding of metal ions to HA. In the context of a water treatment process, such a result is highly promising. However, this hypothesis awaits experimental verification lying outside the scope of the present work. For the sake of experimental rigour, it should be pointed out that the separation of HA and heavy metal ions differs during the analysis. The HA is separated out by settling and centrifugation, while the heavy metal ions receive a further ultrafiltration step. Preliminary tests revealed that the molecular weight cut-off of the ultrafiltration membrane (10 kDa) was sufficient to trap all the HA; TOC measurements revealed no HA in the permeate. Indeed, the soil-derived HA of the Aldrich type has about double the MW of aquatic types, and is further increased after complexation with metal ions. Therefore, the additional UF step is only necessary to determine the free (uncomplexed) metal ion concentration. Clearly, in a conventional treatment process involving coagulation–flocculation followed by sedimentation and coarse filtration, only the HA removal data are strictly applicable, while the removal efficiency results for heavy metal ion would incorporate complexed but unflocculated material. However, the results give an indication of the power and efficacy of the complexation–flocculation process. 4.3.2. Effect of pH on the removal As seen earlier, the pH level can significantly affect the binding of heavy metal ions to HA through the dependence
Fig. 7. Removal of HA at different pH levels. HA initial concentration is 20 mg/L, Pb2+ initial concentration is 2.5 mg/L.
of HA functional group dissociation on pH level [2]. Further, the removal of HA by PolyDADMAC remains a pH-dependent process in the presence of metal ions, Fig. 7. As seen previously, it is the pH-dependent dissociation of HA which makes the effective flocculation region dependent on pH level. In Fig. 8, it is also apparent that the removal of Pb2+ increases with both pH and with flocculant:HA ratio within the optimum coagulation–flocculation region. The removal of Pb2+ increases significantly above pH 5, and for a given pH the maximum in removal corresponds roughly to the mass ratio of PolyDADMAC:HA for maximum flocculation in Fig. 7, i.e. 0.1–0.2 at pH 3, 0.2 at pH 5, 0.25 at pH 7, and 0.3–0.4 at pH 9. Therefore, the explanation previously given for the maximum in HA flocculation with PolyDADMAC:HA mass ratio can also be applied to the maximum in heavy metal removal.
Fig. 8. Removal of Pb2+ at different pH levels. HA initial concentration is 20 mg/L, Pb2+ initial concentration is 2.5 mg/L.
54
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
Fig. 9. Removal of Zn2+ with different HA initial concentrations: (a) 10 mg/L, (b) 20 mg/L, (c) 40 mg/L; Zn2+ initial concentration is 1.2 mg/L, pH 7.
4.3.3. Effect of HA concentration on the removal of heavy metals The removal efficiency for Zn2+ in the supernatant after complexation and flocculation remained at relatively low levels in the low and high range of PolyDADMAC dosage ratios, and increased in the effective flocculation region. This was true for the three different initial concentrations of HA examined (Fig. 9); the removal efficiency of Zn2+ ions also increased slightly with initial concentration of HA. A similar (though less pronounced) result was found for Pb2+ removal. The amount of bound metal ions increases with increasing HA concentration, essentially as more binding sites become available in solution for the heavy metal. 4.3.4. Effect of metal ion concentration on the removal of heavy metals and humic acid The removal efficiency curves for Pb2+ in Fig. 10 have a similar shape to those for Zn2+ in Fig. 9, showing two lower regions and a higher plateau of enhanced removal in the floc-
Fig. 10. Removal of Pb2+ with different Pb2+ initial concentrations: (a) 1.2 mg/L, (b) 2.5 mg/L, (c) 5.0 mg/L; HA initial concentration is 20 mg/L, pH 7.
Fig. 11. Removal of HA with different Pb2+ initial concentrations: (a) 1.2 mg/L, (b) 2.5 mg/L, (c) 5.0 mg/L; HA initial concentration is 20 mg/L, pH 7.
culation region. Again, the initial concentration of heavy metal ion seems to have little (in this case even less) effect on removal efficiency. For HA removal, the highest removal efficiencies are above 85%, Fig. 11. However, the effective flocculation range is affected by the initial concentration of heavy metal ion. The negative charge of the HA is partially neutralized by the metal ions after complexation, and less PolyDADMAC is needed to neutralize the remaining charge and bring about flocculation. The heavy metal ions and the flocculant do not seem to compete, but instead bind to different sites on the HA. Hence, less flocculant is needed when heavy metal binding is increased. Such auto-flocculating action is advantageous in a water treatment process. Equally, less PolyDADMAC is required for redispersion. The effect of increasing initial concentrations of metal ion is thus similar to decreasing pH. A proposed mechanism for the complexation–flocculation process for the removal of heavy metal ions bound to HS is shown in Fig. 12. In this figure, the HS are represented only schematically as spheres, whereas in reality they are complex, soluble macromolecules. Ideally, only the unbound heavy metal ions would exist in the separated supernatant after complexation–flocculation treatment; bound ions would remain associated with the flocculated phase. Coarse filtration to remove the flocculated material, rather than the ultrafiltration necessary to remove unflocculated HS, would be sufficient. It should be pointed out that in previous, unpublished work carried out in our laboratory, ferric, zinc and aluminium salts were used as inorganic coagulant–flocculants to remove 2,4D bound to HA. The binding mass-ratio between 2,4-D and HA was found to increase with pH level. Significant differences in the binding mass-ratio were also found in the presence of the different coagulant–flocculants; for the sequence of ferric, zinc and aluminium salts, the values were 0.05, 0.1 and 0.15, respectively. This suggests that there is an interaction between the ions and the bound materials. A significant effect of time on removal was not found, implying rapid action.
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
55
sion in the present paper, this result may well arise because the hydrophobicity of HA increases when the surface charge is neutralized, and because the internal hydrophobic core in the HA micellar-like structure becomes exposed on the exterior and enhances the adsorption of 2,4-D [1]. 5. Conclusions The simultaneous removal of HA and bound heavy metal ions by complexation–flocculation has received little or no previous attention. Using PolyDADMAC as a cationic flocculant, it has been shown that the complexation–flocculation process is effective for the removal of metal ions, such as Zn2+ and Pb2+ , bound to humic acid. Such a complexation–flocculation process is therefore of potential interest for the removal of heavy metals during water treatment. In particular, the results have shown that:
Fig. 12. Mechanism of complexation–flocculation process for HS and cation removal (Note: the various species are not necessarily univalent; the HS and the coagulant–flocculant in particular have multiple charges).
The significant finding of this previous work was the enhancement by previous (rather than subsequent) coagulation–flocculation of HA on the removal of material bound to HA: coagulation–flocculation was used before the complexation (binding) process between 2,4-D and HA. Fig. 13 shows that the flocculation–complexation approach is more effective in this case than the conventional complexation– flocculation procedure. A difference is not seen at lower HA concentrations. However, when the HA concentration increases, the mass of bound 2,4-D is higher for the flocculation–complexation procedure, rising to more than double that using the conventional complexation–flocculation procedure. In view of the discus-
Fig. 13. Differences in the mass of bound 2,4-D between the complexation– flocculation procedure and the flocculation–complexation procedure. Aluminium sulphate was used as the flocculant at pH 6.
• The metal ion binding increases with pH. • The removal efficiency of HA (and bound metal) by flocculant passes through a maximum at intermediate values of the PolyDADMAC:HA ratio. This ratio increases as pH increases. • Higher concentrations of HA broaden the effective flocculation region, and lead to an increase in metal ion removal efficiency after flocculation. • The binding of metal ions to the HA (and hence subsequent removal) is actually enhanced during flocculation, possibly through structural rearrangements in the HA molecule. Polymeric flocculant and heavy metal ions do not seem to compete for the same sites on the HA molecule. • Increasing the heavy metal concentration leads to a decrease in the required amount of PolyDADMAC for flocculation; in this fashion, the metal ions have an auto-flocculating action, which is again clearly advantageous in a water treatment process. References [1] M. Essington, Soil and Water Chemistry, CRC Press, 2004. [2] E. Tipping, Cation Binding by Humic Substances, Cambridge University Press, 2002. [3] A. Lpatova, S. Verbych, M. Bryk, R. Nigmatullin, N. Hilal, Ultrafiltration of water containing natural organic matter: heavy metal removal in the hybrid complexation-ultrafiltration process, Sep. Purif. Technol. 40 (2004) 155–162. [4] M. Haitzer, G. Aiken, J. Ryan, Binding of mercury(II) to aquatic humic substances: influence of pH and source of humic substances, Environ. Sci. Technol. 37 (2003) 2436–2441. [5] D. Kinniburgh, C. Milne, M. Benedetti, J. Pinheiro, J. Filius, L. Koopalm, W. Riemsdijk, Metal ions binding by humic acid: application of the NICA-Donnan model, Environ. Sci. Technol. 30 (1996) 1687–1698. [6] E. Tipping, C. Rey-Castro, S. Bryan, J. Hamilton-Taylor, Al(III) and Fe(III) binding by humic substances in fresh waters and implication for trace metal speciation, Geochim. Cosmochim. Acta 66 (2002) 3211–3224. [7] T. Nifant’eva, P. Burba, O. Fedorova, V. Shkinev, B. Spivakov, Ultrafiltration and determination of Zn– and Cu–humic substances complexes stability constants, Talanta 53 (2001) 1127–1131. [8] J. Yu, D. Sun, J. Tay, Characteristics of coagulation–flocculation of humic acid with effective performance of polymeric flocculant and inorganic coagulant, Water Sci. Technol. 47 (2002) 89–95.
56
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
[9] S.K. Kam, J. Gregory, The interaction of humic substances with cationic polyelectrolytes, Water Res. 35 (2001) 3557–3566. [10] B. Bolto, D. Dixon, S. Gray, H. Chee, P. Harbour, The use of soluble organic polymers in waste treatment, Water Sci. Technol. 34 (1996) 117–124. [11] B. Bolto, G. Abbt-Braun, D. Dixon, R. Eldridge, F. Frimmel, S. Hesse, S. King, M. Toifl, Experimental evaluation of cationic polyelectrolytes for removing natural organic matter from water, Water Sci. Technol. 40 (1999) 71–79. [12] B. Bolto, D. Dixon, R. Eldridge, S. King, Cationic polymer and clay or metal oxide combinations for natural organic matter removal, Water Res. 35 (2001) 2669–2676. [13] Y. Laor, M. Rebhun, Complexation–flocculation: a new method to determine binding coefficients of organic contaminants to dissolved humic substances, Environ. Sci. Technol. 31 (1997) 3558– 3564. [14] M. Rebhun, S. Meir, Y. Laor, Using dissolved humic acid to remove hydrophobic contaminants from water by complexation–flocculation process, Environ. Sci. Technol. 32 (1998) 981–986. [15] M. Tuncay, S. Christian, E. Tucker, R. Taylor, J. Scamehorn, Ligand-modified polyelectrolyte-enhanced ultrafiltration with electrostatic attachment of ligands. 1. Removal of Cu(II) and Pb(II) with expulsion of Ca(II), Langmuir 10 (1994) 4688–4692.
[16] M. Tuncay, S. Christian, E. Tucker, R. Taylor, J. Scamehorn, Ligand-modified polyelectrolyte-enhanced ultrafiltration with electrostatic attachment of ligands. 2. Use of diethylenetriaminepentaacetic acid/cationic polyelectrolyte mixtures to remove both cations and anions from aqueous streams, Langmuir 10 (1994) 4693–4697. [17] H. Lippold, A. Mansel, H. Kupsch, Influence of trivalent electrolytes on the humic colloid-borne transport of contaminant metals: competition and flocculation effects, J. Contam. Hydrol. 76 (2004) 337–352. [18] P. Burba, B. Aster, T. Nifant’eva, V. Shkinev, B. Spivakov, Membrane filtration studies of aquatic humic substances and their metal species: a concise overview. Part 1. Analytical fraction by means of sequentialstage ultrafiltration, Talanta 45 (1998) 977–988. [19] H. Lippold, A. Mansel, H. Kupsch, Influence of trivalent electrolytes on the humic colloid-borne transport of contaminant metals: competition and flocculation effects, J. Contam. Hydrol. 76 (2004) 337–352. [20] S. Kam, J. Gregory, Charge determination of synthetic cationic polyelectrolyte by colloid titration, Colloid Surf. A: Physicochem. Eng. Aspects 159 (1999) 165–179. [21] M. Fukushima, K. Nakayasu, S. Tanaka, H. Nakamura, Chromium(III) binding abilities of humic acids, Anal. Chim. Acta 317 (1995) 195–206. [22] M. Lurie, M. Rebhun, Effect of properties of polyelectrolytes on their interaction with particulates and soluble organics, Water Sci. Technol. 36 (1997) 93–101.
Enhanced removal of heavy metal ions bound to humic acid by polyelectrolyte flocculation Nicholas P. Hankins a,∗ , Na Lu b , Nidal Hilal b a
b
Department of Engineering Science, The University of Oxford, Parks Road, Oxford OX1 3PJ, UK School of Chemical, Environmental and Mining Engineering, University of Nottingham, University Park, Nottingham NG7 2RD, UK Received 16 October 2005; received in revised form 14 December 2005; accepted 16 December 2005
Abstract The enhanced removal of heavy metal ions from solution, such as Pb2+ and Zn2+ , was studied by binding the ions to humic acid (HA) and then coagulating–flocculating with the cationic polyelectrolyte polydiallyldimethylammonium chloride (PolyDADMAC). The effect of the dosage of PolyDADMAC, the pH level and the concentrations of HA and metal ions were studied. Ultrafiltration was used to separate bound metal ions from free ions in solution. The removal of bound metal ions was found to increase with the extent of coagulation–flocculation of the HA by PolyDADMAC. The pH affects the removal efficiency of bound (complexed) metal, in so much as it affects the binding strength of metal ions to the HA. The removal efficiency of metal also increases with the initial concentration of HA. The effective coagulation–flocculation region of the HA by PolyDADMAC is affected by the initial concentration of the metal ions; an increase in the concentration leads to a decrease in the amount of PolyDADMAC required. Humic substances have the advantage of being naturally occurring, and the results indicate that such a complexation–flocculation process is of potential interest for the removal of heavy metals during water treatment. © 2006 Elsevier B.V. All rights reserved. Keywords: Humic acid; Heavy metal; Coagulation; Flocculation; Flocculant; Cationic polyelectrolyte; Complexation–flocculation; Waste water treatment
1. Introduction As major components of natural organic matter, humic substances (HS) are involved in many chemical and physicochemical interactions in soil and aqueous systems [1]. HS include humic and fulvic acids, and have a wide range of molecular weights, with both hydrophilic and hydrophobic moieties. The interactions of HS with heavy metals have already been established and studied [2]. The binding of heavy metal ions by HS is controlled mainly by the pH, the concentrations of the HS and heavy metal ions and by other parameters [3–6]. Ultrafiltration and other methods have been used to study the binding of metal ions to HS [3,7]. In the current work, the soil-derived humic acid (HA) material used had a sufficiently large molecular weight to be trapped by ultrafiltration; on the other hand, if aquatic HA material is used, a normal UF membrane would trap only about half the material and a nanofiltration membrane might well be required.
∗
Corresponding author. Tel.: +44 1865 273 027; fax: +44 1865 283 273. E-mail address: [email protected] (N.P. Hankins).
1383-5866/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2005.12.022
Because of their appreciable solubility in water, HS may be regarded as anionically charged macromolecular colloids (as opposed to colloidal dispersions, which are generally insoluble). As such, HS can be removed by coagulation–flocculation processes; this can be achieved by the addition of an inorganic salt and/or or a cationic polyelectrolyte (PE) [8]. In the context of the present work, coagulation essentially refers to charge neutralization and hence solute destabilisation, while flocculation refers to the process of floc growth whereby more and more macromolecular material becomes incorporated into the growing flocs. For instance, the removal of HS from upland waters during potable water treatment is frequently carried out by the addition of hydrolysed ferric or aluminium salts; in this case, the removal occurs through a combination of charge neutralization and possible enmeshment (also known as ‘sweep flocculation’). In the case of polyelectrolyte addition, the mechanism of removal has not yet been fully elucidated or agreed upon, but it seems likely to be dominated by charge neutralization [9]; at optimum dosage, the resulting HS–PE complex is uncharged and highly hydrophobic, and so can easily grow (flocculate) into large flocs. Note, however, that the initial complexes are macromolecular in nature, and cannot easily be characterised for zeta
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
potential, charge density or hydrodynamic size, in the same way as for colloidal particle dispersions. The role, if any, of polymer bridging by PE remains uncertain. For the coagulation–flocculation process, a PE with a higher charge density leads to a higher removal efficiency, while molecular weight may have little influence [9]. The actual removal efficiency of HS by PE depends on the dosage of PE, the pH and the other flocculation conditions [8,10–12]. Given that HS can bind with other pollutant species in aqueous solution and then be removed by coagulation–flocculation, HS can be considered as a separating agent to remove bound pollutants in conjunction with PE coagulation–flocculation. However, such a complexation–flocculation process, involving HS as a substrate and PE as a coagulant–flocculant, has received little prior attention for the removal of heavy metals during water treatment. HA has been used to remove bound organic contaminants (OC) including pyrene, fluoranthene and anthracene, by coagulation–flocculation with Al(III) and Fe(III) salts at pH 6 [13,14]. The removal of OC occurs simultaneously with coagulation–flocculation of the OC–HA system. The removal efficiency of OC depends mostly on the stability of the OC–HA complexes. Cationic PEs have also been used as flocculants to remove cations, anions and organic pollutants bound to negatively charged organic chelating agents or surfactants. Mixtures of PolyDADMAC and 4,5-dihyhroxy-1,3-benzenedisulphonic acid (Titron) were found to be effective in removing cations, such as Cu2+ and Pb2+ , while expelling non-complexed alkalineearth ions such as Ca2+ [15]. Mixtures of PolyDADMAC and diethylenetriaminepentaacetic acid (DTPA) were discovered that can remove divalent cations, such as Cu2+ and Pb2+ , and anions such as CrO42− at pH values higher than 4, and can also expel interfering cations, such as Ca2+ [16]. In the complexation–flocculation process, the effect of additional coagulant–flocculant on the stability of the HS complexes needs to be taken into account. For the coagulation–flocculation of HA with inorganic substances to remove bound organic compounds, since Al(III) or Fe(III) and the organic compounds react with different sites on the humic material, a competition between them is not expected. The bound organic compounds in the complex are, in effect, precipitated, flocculated and entrapped in Al or Fe flocs in the same way and to the same extent as the HA itself. On the other hand, inorganic coagulant–flocculants such as aluminium and other trivalent electrolytes may have significant effects on the stability of complexes of humic acid and bound heavy metals. The competitive effect of Al(III) and other trivalent electrolytes on HS–La(III) complexes has been assessed [17]. The possibility of significant competitive effects was confined to Al(III). The amount of complexed (bound) La at pH 4 decreased significantly after an equilibration time of 48 h if the metals were introduced in two steps, such as adding Al to La + HA solution, or adding La to Al + HA solution, or adding a mixture of Al and La to the HA solution. Thus, inorganic salts are not suitable for removing heavy metal ions bound to the HS by coagulation–flocculation.
49
The effect of additional metal ions on HS coagulation– flocculation also needs to be considered. On the one hand, a competition between Al(III) or Fe(III) coagulant–flocculant and organic pollutant is not expected during complexation of the latter to the HS, so the optimum Al(III) or Fe(III) dosages for the complexation–flocculation process are expected to be the same as the optimum dosages for simple HS coagulation–flocculation. On the other hand, in tests of the coagulation–flocculation of HA by Al(III) in the presence of bound La(III), although the HA removal efficiency can be kept at the same level, the optimum dosage of Al(III) actually decreased with participation of La(III) [17]. Using the polyelectrolyte flocculant PolyDADMAC, Tuncay et al. also found that the removal efficiencies of DTPA and Titron become higher at lower ratios of DTPA to PolyDADMAC or Titron to PolyDADMAC [15,16]. The simultaneous removal of HS and bound heavy metal ions by complexation–flocculation has not been studied previously. Because of the chemical features of polyelectrolyte (PE) as a coagulant–flocculant, it has a good propensity for HS removal. Moreover, there is no evidence of it having a strong competitive effect on metal–HS binding. Therefore, its application to the complexation–flocculation process for the removal of bound heavy metal ions in water treatment looks promising. Such a process has received no previous attention. In a report released by the U.S. EPA, heavy metals such as lead, cadmium, mercury, chromium, copper, zinc and nickel have been placed at the top of a priority pollutant ranking [18]. The interactions between HS and trivalent metals have been studied [19], and zinc and lead were chosen as typical bivalent metals. In this work, the effect of a number of factors is studied, and Zn and Pb are used as model heavy metals. Poly(diallyldimethylammonium) chloride, PolyDADMAC, is used as a high charge cationic PE. It has been shown to have a stable, highly positive charge over a wide pH range [20]. 2. Materials HA solution was prepared as follows: 4 g of HA (Aldrich, soil derived and used as the sodium salt) were dissolved in 20 mL of 1 M NaOH, then diluted into 1.8 L of deionised water and stirred for 2 h. The pH was initially adjusted to 7; at this value of pH, the HA was soluble. No precipitation of HA was observed during any test in which pH was subsequently adjusted to a lower (or higher) value. The total volume was adjusted to 2 L, and the solution was stirred for an additional 3 h. The solution was finally filtered through Whatman GF-A filter paper, followed by filtration through a 0.45 m membrane (Millipore). HA concentrations were determined throughout by measuring the UV absorption at a wavelength of 254 nm in a Jenway 6405 UV/vis spectrophotometer (note that preliminary tests of total organic carbon in solution showed an agreement with UV absorption measurements of HA, indicating an absence of nonabsorbing compounds). Aqueous solutions of Zn and Pb were purchased from Sigma–Aldrich. One percent PolyDADMAC solutions were prepared daily by diluting 20% commercial PolyDADMAC product (Sigma–Aldrich; medium molecular weight around 50 000, with a high charge density around 5–6 mequiv./g)
50
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
with deionised water. Analytical grade reagents were employed throughout the experiments. Milli-Q plus (Millipore) water was used in preparing all solutions. A stirred ultrafiltration Cell (Millipore, Model 8010, 10 mL) was used, in conjunction with a YM1 membrane with a cut-off of 1 kDa (Millipore). A Waterproof pH Tester (3 Double Junction, OAKTON Instruments) was employed for pH determination. A floc-tester (Bibby Stuart flocculator SW6) was used for flocculation jar testing. An Orbitar incubator-shaker was used to bring solutions to equilibrium during binding measurements. Adjustments in pH were carried out with HCl or NaOH. All the experiments were performed at an ambient temperature of 22 ◦ C. 3. Methods 3.1. HA–metal binding determination The binding between HA and heavy metals was determined in batch studies. A series of solutions with a fixed or variable HA concentration and a fixed or variable Pb2+ or Zn2+ concentration were prepared as feed solutions. Hundred milliliters of feed solution were placed in a 100-mL conical flask and closed with lab foil. Finally, the pH of the solutions was adjusted to a fixed or variable value in the range between 3 and 9. Prior to ultrafiltration to remove the bound ions, solutions were kept shaking at 25 ◦ C for 12 h to reach an equilibrium state for the HA–metal system. Then 10 mL of sample were ultra-filtered, with a stirring rate of 300 rpm and an operating pressure of 50 psi. This ensured that only free (non-bound) ions were present in the permeate: measurements of metal removal thus measured the true effects of binding (complexation). The concentration of metal ions in the permeate solution was measured by a flame atomic absorption spectrophotometer (AAS). 3.2. HA–PolyDADMAC coagulation–flocculation tests A series of standard jar-test experiments were carried out to determine the required dosage of PolyDADMAC to efficiently precipitate HA. A series of 250 mL HA samples with a fixed or variable concentration were placed in 500 mL beakers, and the pH adjusted to a fixed or variable value in the range between 3 and 9. One percent PolyDADMAC solution was added during fast-stirring at 250 rpm for 2 min, in variable mass ratios of PolyDADMAC to HA, followed by slow stirring at 30 rpm for 10 min. The pH level was maintained at the desired level during fast-stirring. After overnight settling, 40 mL of solution were centrifugally separated at 4000 rpm for 20 min. The residual HA content in the supernatant was determined by UV absorption at 254 nm.
desired final concentrations of each, were prepared by mixing the appropriate amounts of HA and metal ion stock solution and diluting with deionised water. The pH was adjusted to a fixed or variable value in the range between 3 and 9. One percent PolyDADMAC solution was added during fast-stirring at 250 rpm for 2 min, in variable mass ratios of PolyDADMAC to HA, followed by slow stirring at 30 rpm for 10 min. The solutions were shaken at 25 ◦ C for 12 h to reach an equilibrium state of complexation. After overnight settling and centrifugation of 40 mL of solution at 4000 rpm for 20 min, 10 mL of supernatant were ultra-filtered through a 1 kDa membrane, as in 3.1, to ensure that only free (unbound) ions were present in the permeate. The concentration of metal ions in the permeate solution was measured by AAS, and HA concentration in the supernatant was determined by UV absorption at 254 nm. The conditions for the complete set of tests are summarized in Table 1. 4. Results and discussion 4.1. HA–metal binding determination Since the binding between HA and metal ions is the basis for the complexation–flocculation process, the effects of pH and the ratio of HA to metal ions on the binding were studied. 4.1.1. Effect of pH on HA–metal binding According to Fig. 1, the removals of Zn and Pb are lower than 50% at pH 3, since the major functional groups of HA responsible for the metal binding (weak acidic groups, which dissociate into negatively charged forms) have low dissociation at low pH levels. This inhibits the complexation. The quantities of bound metal ions are reduced, as a result of proton competition. With an increase in pH level, the removal of metal ion is improved. The dissociation of HA functional groups increases with the pH level, leading to an increase in negative charge density and hence in the binding of metal ion to HA.
3.3. Complexation–flocculation tests The mechanism of HA–metal complex removal by the complexation–flocculation process was tested using a series of flocculation jar tests, followed by centrifugation and ultrafiltration. A series of 250 mL solutions of HA and metal ions, of the
Fig. 1. The effect of pH on metal removal by binding during the complexation of HA and metal ions. Initial concentration of HA is 20 mg/L; initial concentration of metal ions is 5 mg/L.
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
51
Table 1 Experimental conditions for tests Test
HA (mg/L)
Pb2+ /Zn2+ (mg/L)
pH
Mass ratio PolyDADMAC:HA
Effect of pH on the metal binding Effect of HA–metal ratio on the metal binding Effect of PolyDADMAC dosage and pH on HA removal Effect of HA concentration on HA removal Effect of PolyDADMAC dosage on heavy metal removal Effect of pH Effect of HA concentration Effect of metal ion concentration on the ion removal
20 20 20 10, 20, 40 20 20 10, 20, 40 20
5 0.6–20 – – 1.2 1.2 1.2 1.2, 2.5, 5
3–9 7 3–9 7 7 3–9 7.0 7.0
– – 0.1–0.5 0.1–0.5 0.1–0.5 0.1–0.5 0.1–0.5 0.1–0.5
Thus, the pH level significantly affects the binding of metal ions to HA. 4.1.2. Effect of the HA–metal ratio on the metal removal after ultrafiltration Fig. 2 presents Zn and Pb-HA binding tests, which were made at pH 7 under the same conditions. The removal for Pb is higher than that for Zn. Between the two bivalent ions, Pb has a larger ionic radius than Zn. But Zn2+ has higher ionic potential, is easier to hydrolyse, and exists as monovalent (ZnOH)+ , which consequently binds less strongly than bivalent lead. For a given HA concentration, Fig. 2 shows there is little change in metal ion removal as the HA–metal ratio varies above 1 or 2; presumably, an excess of binding sites are already available. From these results, we can calculate the maximum binding ratio between heavy metal and HA on a mass basis: for lead this is 0.98, and for zinc this is 0.65. Binding constants of various metal–HA complexes have been reported [21]. A descending series of log K values may be presented as follows: Fe3+ (7.5) > Al3+ (5.1) > Cu2+ (4.6) > Ni2+ (4.4) > Co2+ (4.2) > Pb2+ = Ca2+ (3.5) > Zn2+ (3.1) > Mn2+ (2.8) > Mg2+ (2.6). Pb2+ has a stronger binding ability than Zn2+ , so Pb2+ removal is more effective than Zn2+ .
Fig. 2. The effect of HA–metal ratio on metal removal by binding during complexation of HA and metal at pH 7. HA initial concentration is 20 mg/L.
4.2. HA–PolyDADMAC coagulation–flocculation tests 4.2.1. Effect of PolyDADMAC dosage and pH on HA removal The coagulation–flocculation of HA by PolyDADMAC is expected to be dominated by charge neutralization [9]; the ‘bridging’ plays an uncertain role. The process mainly depends on the ionic dissociation of HA, and on the oppositely charged character of PolyDADMAC. As shown in Fig. 3, with a fixed HA concentration of 20 mg/L and pH 7, PolyDADMAC is effective in removing HA when the mass:mass ratio lies in the range between 0.3 and 0.4. The flocculation curve has a classical shape, comprising underdose (under neutralized), optimum dose (neutralized) and overdose (re-charged) regions. Similar results have been reported in previous work [8,22]. In the underdose region at lower concentrations of PolyDADMAC, some charge neutralization occurs, but this is not sufficient to allow aggregation and hence the creation of flocs large enough to settle. In the optimum dose region, the surface charge of the HA is significantly neutralized by the PolyDADMAC; the resulting HA–PolyDADMAC complex is thus rendered highly hydrophobic in an aqueous medium, floc growth can then occur readily. As
Fig. 3. Removal of HA by PolyDADMAC flocculation at different pH levels. HA initial concentration is 20 mg/L.
52
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
a cationic, medium molecular weight polyelectrolyte, polymer bridging may possibly also assist in forming flocs large enough to settle. In the overdose region, with a ratio higher than 0.4, the removal efficiency again decreases. The HA molecules are surrounded by excess PolyDADMAC, leading to high charge and both electrostatic and steric repulsion, which in turn leads to stable dispersion of the HA. Fig. 3 also presents the removal efficiency of HA from the supernatant after coagulation–flocculation, as a function of the mass dosage ratio of PolyDADMAC:HA, at different pH levels. For all the pH levels tested, the highest removal efficiencies are above 85%. The optimum PolyDADMAC dosage ratios for different pH levels are shown in Table 2. Table 2 clearly shows that the optimum mass dosage ratio increases with the pH level. Since the charge density of PolyDADMAC is relatively constant at varying pH [20], the increasing dissociation of HA functional groups at increasing pH, and hence the negative charge density of the HA, is the main factor in controlling coagulation–flocculation. At low pH, the low dissociation of HA yields a low charge density, and it can be neutralized by a smaller amount of PolyDADMAC. Also, it might also be speculated that PolyDADMAC has a good polymer bridging ability, so that the flocculation is still effective at low pH values. In a similar fashion, less PolyDADMAC is required to redisperse the HA at low pH. As pH is increased, the HA becomes more dissociated, and a larger amount of the positively charged PolyDADMAC is required to neutralize the more negatively charged HA. The actual removal efficiency at optimum dosage shows no significant difference as pH level varies. 4.2.2. Effect of HA concentration on HA removal Fig. 4 demonstrates the removal efficiency of HA after coagulation–flocculation from the supernatant, as a function of the dosage mass ratio of PolyDADMAC to HA, with different initial HA concentrations at pH 7. The highest removal efficiencies, with varying initial concentrations of HA, are all above 85%. The effective range becomes broader at the higher ratio end with an increase in HA concentration. In solutions with higher initial concentrations of HA, it might be speculated that bridging flocculation can occur more easily under shear at the same dosage ratio of PolyDADMAC to HA. In such solutions, the total amount of PolyDADMAC and HA are higher, such that bridging causes flocculation to occur over a broader region of dosage ratios.
Fig. 4. Removal of HA by PolyDADMAC at pH 7 with different initial concentrations of HA: (a) 10 mg/L, (b) 20 mg/L, (c) 40 mg/L.
4.3. Complexation–flocculation tests 4.3.1. Effect of PolyDADMAC dosage on heavy metal removal Figs. 5 and 6 show the removal efficiency by binding for Pb2+ and Zn2+ , following both complexation and flocculation (and finally centrifugation and ultrafiltration of the supernatant), as a function of the dosage mass ratio of PolyDADMAC to HA. The removal efficiency for metal ions remained at a lower level (95% and 62%, respectively) in the low PolyDADMAC dosage region (0–0.1), it increased (to 99% and 72%, respectively) in the effective range of coagulation–flocculation (0.25–0.3), and decreased back to a lower level (90% and 62%, respectively) in the overdose range (0.35–0.5). Because metal ions and PolyDADMAC are both positively charged, an interaction between them is not expected. The increase in removal efficiency must therefore occur because of an improvement in the complexation
Table 2 Optimum mass dosage ratio of PolyDADMAC to HA for HA removal at different pH values pH
Mass ratio of PolyDADMAC to HA
Removal efficiency (%)
3 5 7 9
<0.1 0.2–0.3 0.3–0.4 0.35–0.45
90 92 87 89
Fig. 5. Removal of Pb2+ and HA at pH 7. Pb2+ initial concentration is 1.2 mg/L. HA initial concentration is 20 mg/L.
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
53
Fig. 6. Removal of Zn2+ and HA at pH 7. Zn2+ initial concentration is 1.2 mg/L. HA initial concentration is 20 mg/L.
of metal ions to the functional groups of the HA during flocculation. The interaction of PolyDADMAC with the functional groups of the HA molecule partially or completely neutralizes its surface charge; under this altered electrostatic environment, it might be speculated that it would lead to a transformation in the overall molecular structure of the HA (analogous to tertiary structure in a protein molecule) and a change in the spatial distribution of the HA surface functional groups responsible for binding heavy metal ions. In essence, the structure ‘opens up’, and the functional groups become more sterically accessible to the heavy metal ions. It could then be hypothesized that structural changes lead to an enhancement in the binding of metal ions to HA. In the context of a water treatment process, such a result is highly promising. However, this hypothesis awaits experimental verification lying outside the scope of the present work. For the sake of experimental rigour, it should be pointed out that the separation of HA and heavy metal ions differs during the analysis. The HA is separated out by settling and centrifugation, while the heavy metal ions receive a further ultrafiltration step. Preliminary tests revealed that the molecular weight cut-off of the ultrafiltration membrane (10 kDa) was sufficient to trap all the HA; TOC measurements revealed no HA in the permeate. Indeed, the soil-derived HA of the Aldrich type has about double the MW of aquatic types, and is further increased after complexation with metal ions. Therefore, the additional UF step is only necessary to determine the free (uncomplexed) metal ion concentration. Clearly, in a conventional treatment process involving coagulation–flocculation followed by sedimentation and coarse filtration, only the HA removal data are strictly applicable, while the removal efficiency results for heavy metal ion would incorporate complexed but unflocculated material. However, the results give an indication of the power and efficacy of the complexation–flocculation process. 4.3.2. Effect of pH on the removal As seen earlier, the pH level can significantly affect the binding of heavy metal ions to HA through the dependence
Fig. 7. Removal of HA at different pH levels. HA initial concentration is 20 mg/L, Pb2+ initial concentration is 2.5 mg/L.
of HA functional group dissociation on pH level [2]. Further, the removal of HA by PolyDADMAC remains a pH-dependent process in the presence of metal ions, Fig. 7. As seen previously, it is the pH-dependent dissociation of HA which makes the effective flocculation region dependent on pH level. In Fig. 8, it is also apparent that the removal of Pb2+ increases with both pH and with flocculant:HA ratio within the optimum coagulation–flocculation region. The removal of Pb2+ increases significantly above pH 5, and for a given pH the maximum in removal corresponds roughly to the mass ratio of PolyDADMAC:HA for maximum flocculation in Fig. 7, i.e. 0.1–0.2 at pH 3, 0.2 at pH 5, 0.25 at pH 7, and 0.3–0.4 at pH 9. Therefore, the explanation previously given for the maximum in HA flocculation with PolyDADMAC:HA mass ratio can also be applied to the maximum in heavy metal removal.
Fig. 8. Removal of Pb2+ at different pH levels. HA initial concentration is 20 mg/L, Pb2+ initial concentration is 2.5 mg/L.
54
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
Fig. 9. Removal of Zn2+ with different HA initial concentrations: (a) 10 mg/L, (b) 20 mg/L, (c) 40 mg/L; Zn2+ initial concentration is 1.2 mg/L, pH 7.
4.3.3. Effect of HA concentration on the removal of heavy metals The removal efficiency for Zn2+ in the supernatant after complexation and flocculation remained at relatively low levels in the low and high range of PolyDADMAC dosage ratios, and increased in the effective flocculation region. This was true for the three different initial concentrations of HA examined (Fig. 9); the removal efficiency of Zn2+ ions also increased slightly with initial concentration of HA. A similar (though less pronounced) result was found for Pb2+ removal. The amount of bound metal ions increases with increasing HA concentration, essentially as more binding sites become available in solution for the heavy metal. 4.3.4. Effect of metal ion concentration on the removal of heavy metals and humic acid The removal efficiency curves for Pb2+ in Fig. 10 have a similar shape to those for Zn2+ in Fig. 9, showing two lower regions and a higher plateau of enhanced removal in the floc-
Fig. 10. Removal of Pb2+ with different Pb2+ initial concentrations: (a) 1.2 mg/L, (b) 2.5 mg/L, (c) 5.0 mg/L; HA initial concentration is 20 mg/L, pH 7.
Fig. 11. Removal of HA with different Pb2+ initial concentrations: (a) 1.2 mg/L, (b) 2.5 mg/L, (c) 5.0 mg/L; HA initial concentration is 20 mg/L, pH 7.
culation region. Again, the initial concentration of heavy metal ion seems to have little (in this case even less) effect on removal efficiency. For HA removal, the highest removal efficiencies are above 85%, Fig. 11. However, the effective flocculation range is affected by the initial concentration of heavy metal ion. The negative charge of the HA is partially neutralized by the metal ions after complexation, and less PolyDADMAC is needed to neutralize the remaining charge and bring about flocculation. The heavy metal ions and the flocculant do not seem to compete, but instead bind to different sites on the HA. Hence, less flocculant is needed when heavy metal binding is increased. Such auto-flocculating action is advantageous in a water treatment process. Equally, less PolyDADMAC is required for redispersion. The effect of increasing initial concentrations of metal ion is thus similar to decreasing pH. A proposed mechanism for the complexation–flocculation process for the removal of heavy metal ions bound to HS is shown in Fig. 12. In this figure, the HS are represented only schematically as spheres, whereas in reality they are complex, soluble macromolecules. Ideally, only the unbound heavy metal ions would exist in the separated supernatant after complexation–flocculation treatment; bound ions would remain associated with the flocculated phase. Coarse filtration to remove the flocculated material, rather than the ultrafiltration necessary to remove unflocculated HS, would be sufficient. It should be pointed out that in previous, unpublished work carried out in our laboratory, ferric, zinc and aluminium salts were used as inorganic coagulant–flocculants to remove 2,4D bound to HA. The binding mass-ratio between 2,4-D and HA was found to increase with pH level. Significant differences in the binding mass-ratio were also found in the presence of the different coagulant–flocculants; for the sequence of ferric, zinc and aluminium salts, the values were 0.05, 0.1 and 0.15, respectively. This suggests that there is an interaction between the ions and the bound materials. A significant effect of time on removal was not found, implying rapid action.
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
55
sion in the present paper, this result may well arise because the hydrophobicity of HA increases when the surface charge is neutralized, and because the internal hydrophobic core in the HA micellar-like structure becomes exposed on the exterior and enhances the adsorption of 2,4-D [1]. 5. Conclusions The simultaneous removal of HA and bound heavy metal ions by complexation–flocculation has received little or no previous attention. Using PolyDADMAC as a cationic flocculant, it has been shown that the complexation–flocculation process is effective for the removal of metal ions, such as Zn2+ and Pb2+ , bound to humic acid. Such a complexation–flocculation process is therefore of potential interest for the removal of heavy metals during water treatment. In particular, the results have shown that:
Fig. 12. Mechanism of complexation–flocculation process for HS and cation removal (Note: the various species are not necessarily univalent; the HS and the coagulant–flocculant in particular have multiple charges).
The significant finding of this previous work was the enhancement by previous (rather than subsequent) coagulation–flocculation of HA on the removal of material bound to HA: coagulation–flocculation was used before the complexation (binding) process between 2,4-D and HA. Fig. 13 shows that the flocculation–complexation approach is more effective in this case than the conventional complexation– flocculation procedure. A difference is not seen at lower HA concentrations. However, when the HA concentration increases, the mass of bound 2,4-D is higher for the flocculation–complexation procedure, rising to more than double that using the conventional complexation–flocculation procedure. In view of the discus-
Fig. 13. Differences in the mass of bound 2,4-D between the complexation– flocculation procedure and the flocculation–complexation procedure. Aluminium sulphate was used as the flocculant at pH 6.
• The metal ion binding increases with pH. • The removal efficiency of HA (and bound metal) by flocculant passes through a maximum at intermediate values of the PolyDADMAC:HA ratio. This ratio increases as pH increases. • Higher concentrations of HA broaden the effective flocculation region, and lead to an increase in metal ion removal efficiency after flocculation. • The binding of metal ions to the HA (and hence subsequent removal) is actually enhanced during flocculation, possibly through structural rearrangements in the HA molecule. Polymeric flocculant and heavy metal ions do not seem to compete for the same sites on the HA molecule. • Increasing the heavy metal concentration leads to a decrease in the required amount of PolyDADMAC for flocculation; in this fashion, the metal ions have an auto-flocculating action, which is again clearly advantageous in a water treatment process. References [1] M. Essington, Soil and Water Chemistry, CRC Press, 2004. [2] E. Tipping, Cation Binding by Humic Substances, Cambridge University Press, 2002. [3] A. Lpatova, S. Verbych, M. Bryk, R. Nigmatullin, N. Hilal, Ultrafiltration of water containing natural organic matter: heavy metal removal in the hybrid complexation-ultrafiltration process, Sep. Purif. Technol. 40 (2004) 155–162. [4] M. Haitzer, G. Aiken, J. Ryan, Binding of mercury(II) to aquatic humic substances: influence of pH and source of humic substances, Environ. Sci. Technol. 37 (2003) 2436–2441. [5] D. Kinniburgh, C. Milne, M. Benedetti, J. Pinheiro, J. Filius, L. Koopalm, W. Riemsdijk, Metal ions binding by humic acid: application of the NICA-Donnan model, Environ. Sci. Technol. 30 (1996) 1687–1698. [6] E. Tipping, C. Rey-Castro, S. Bryan, J. Hamilton-Taylor, Al(III) and Fe(III) binding by humic substances in fresh waters and implication for trace metal speciation, Geochim. Cosmochim. Acta 66 (2002) 3211–3224. [7] T. Nifant’eva, P. Burba, O. Fedorova, V. Shkinev, B. Spivakov, Ultrafiltration and determination of Zn– and Cu–humic substances complexes stability constants, Talanta 53 (2001) 1127–1131. [8] J. Yu, D. Sun, J. Tay, Characteristics of coagulation–flocculation of humic acid with effective performance of polymeric flocculant and inorganic coagulant, Water Sci. Technol. 47 (2002) 89–95.
56
N.P. Hankins et al. / Separation and Purification Technology 51 (2006) 48–56
[9] S.K. Kam, J. Gregory, The interaction of humic substances with cationic polyelectrolytes, Water Res. 35 (2001) 3557–3566. [10] B. Bolto, D. Dixon, S. Gray, H. Chee, P. Harbour, The use of soluble organic polymers in waste treatment, Water Sci. Technol. 34 (1996) 117–124. [11] B. Bolto, G. Abbt-Braun, D. Dixon, R. Eldridge, F. Frimmel, S. Hesse, S. King, M. Toifl, Experimental evaluation of cationic polyelectrolytes for removing natural organic matter from water, Water Sci. Technol. 40 (1999) 71–79. [12] B. Bolto, D. Dixon, R. Eldridge, S. King, Cationic polymer and clay or metal oxide combinations for natural organic matter removal, Water Res. 35 (2001) 2669–2676. [13] Y. Laor, M. Rebhun, Complexation–flocculation: a new method to determine binding coefficients of organic contaminants to dissolved humic substances, Environ. Sci. Technol. 31 (1997) 3558– 3564. [14] M. Rebhun, S. Meir, Y. Laor, Using dissolved humic acid to remove hydrophobic contaminants from water by complexation–flocculation process, Environ. Sci. Technol. 32 (1998) 981–986. [15] M. Tuncay, S. Christian, E. Tucker, R. Taylor, J. Scamehorn, Ligand-modified polyelectrolyte-enhanced ultrafiltration with electrostatic attachment of ligands. 1. Removal of Cu(II) and Pb(II) with expulsion of Ca(II), Langmuir 10 (1994) 4688–4692.
[16] M. Tuncay, S. Christian, E. Tucker, R. Taylor, J. Scamehorn, Ligand-modified polyelectrolyte-enhanced ultrafiltration with electrostatic attachment of ligands. 2. Use of diethylenetriaminepentaacetic acid/cationic polyelectrolyte mixtures to remove both cations and anions from aqueous streams, Langmuir 10 (1994) 4693–4697. [17] H. Lippold, A. Mansel, H. Kupsch, Influence of trivalent electrolytes on the humic colloid-borne transport of contaminant metals: competition and flocculation effects, J. Contam. Hydrol. 76 (2004) 337–352. [18] P. Burba, B. Aster, T. Nifant’eva, V. Shkinev, B. Spivakov, Membrane filtration studies of aquatic humic substances and their metal species: a concise overview. Part 1. Analytical fraction by means of sequentialstage ultrafiltration, Talanta 45 (1998) 977–988. [19] H. Lippold, A. Mansel, H. Kupsch, Influence of trivalent electrolytes on the humic colloid-borne transport of contaminant metals: competition and flocculation effects, J. Contam. Hydrol. 76 (2004) 337–352. [20] S. Kam, J. Gregory, Charge determination of synthetic cationic polyelectrolyte by colloid titration, Colloid Surf. A: Physicochem. Eng. Aspects 159 (1999) 165–179. [21] M. Fukushima, K. Nakayasu, S. Tanaka, H. Nakamura, Chromium(III) binding abilities of humic acids, Anal. Chim. Acta 317 (1995) 195–206. [22] M. Lurie, M. Rebhun, Effect of properties of polyelectrolytes on their interaction with particulates and soluble organics, Water Sci. Technol. 36 (1997) 93–101.