3. Ion exchangers and their use in waste water treatment

Since the discovery of ion exchange, it has been used for water purification. However, the method is not extensively adapted to metals removal in industrial wastewater treatment. Industry has shunned ion exchange processes even if ion exchange seems to offer an ideal solution for a particular problem, such as the removal of metal ions from electroplating effluents. Bolto offers a few reasonable explanations for this anomaly i.e: Ion exchange receives little attention in technical education courses and it is not normally considered a promising research area. Ion exchange equipment more resembles a manufacturing plant than a waste water treatment facility. Finally, ion exchange has been regarded more as a method to produce pure water than as a method to purify waste (Bolto).

3.1 Ion exchange materials and their applications in waste water treatment

Ion exchange materials comprise two main groups: organic and inorganic exchangers. Both groups include synthetic and natural materials. Ion exchangers form a very heterogeneous group of materials, their only common feature is that they contain a fixed electric charge which can bind counter ions with an opposite charge.

Practically all organic exchangers used in waste management have a synthetic polymer backbone, although natural polymers like cellulose, alginic acid and chitin offer an endless source of raw materials (Allan). Especially chitin and chitosan have been studied for their good metal sorption properties and found to have potential for waste management (Inoue). One probable cause for the lack of their applications is that natural polymers are commonly biodegradable. Their chemical resistance is rather good but their microbiological sensitivity restricts their use in hydrometallurgy.

3.1.1 Ion exchange resins

Organic ion exchange resins having cross linked hydrocarbon matrix and derivatized with inorganic group are the most common ion exchange materials used in industrial applications. The majority of the commercial resins are based on the styrene-divinylbenzene structure because of its good resistance against chemical and physical stress. The structure is stable at relatively high temperatures and over the whole pH range. Styrene polymer chains (Figure 2) are crosslinked with divinylbentzene (DVB) and the elasticity of resin can be adjusted by varying the amount of DVB it contains.

Figure 3

Figure 2. Styrene-divinylbentzene co-polymer X= ionogenic/functional group

The ion exchange properties of organic resins are mainly based on ionogenic groups which can be attached to practically all the styrene rings in the styrene-divinylbentzene co-polymer. Thus, resins with very low crosslinking (DVB portion 1-2%) can have the maximum theoretical concentration of the ionogenic groups of approximately 9 mmol/g. Usually, commercial resins contain ionogenic groups at a concentration in the range of 2.5-5 mmol/g. The nature of ionogenic groups can vary from strong acidic cation exchangers (-SO3- ) to strong basic anion exchangers (-N+(CH3)3 ) and to chelate forming exchangers. There are also active groups which have no electric charge but donor atoms attract cations by donating free electron pairs to form coordination bonds. Ionogenic, chelating, and complex forming active groups may be called functional groups.

The ion exchange systems used in the metal plating industry are mainly based on conventional technology i.e., strong acid cation or medium base anion resin in fixed bed column systems. This technology has been applied in many special applications like the two-step metal-cyanide recovery treatment of acidic copper or zinc solutions (Cushnie, Fravel) and the renovation of chromating baths (removing of cationic impurities) (Pajunen) but major waste streams are usually treated by other methods. Strong base anion resin can be used for removing anionic metal complexes from acidic waters like ZnCl42- in spent pickle solution or Cr(VI) in rinse water after chromating (Tan). Weakly acidic exchangers (-COO- ) have shown good separation performance for Zn in plating waste (Uy) and for nickel (Halle), however weakly acidic resins have no widespread applications in the plating industry. Carboxylic resins can be manufactured directly from copolymer components, so a functionalization step is unnecessary and it is easy to generate a high theoretical specific capacity for the product. The carboxylic group dissociates in a higher pH range than the sulphonic group and exchanger is easily regenerated with acid. On the other hand, the complex forming characteristics of carboxylic resins (Kononowa) are too weak for competition with complex forming agents in waste solutions.

Chelating ion exchange resins have ionogenic groups which can form coordination bonds with metals, their donor atoms are usually sulphur, nitrogen or both in the same group (Figure 3). The bonds formed in this kind of metal sorption usually have both covalent and ionic characteristics. The sorption properties of chelating resins are well known and documented (Sahni). Probably the best property of the chelating resins is their selectivity towards transition metals and the weakly acidic nature of chelating groups makes the regeneration step with mineral acid quite easy. There are also resins which do not have negatively charged ionogenic groups but which form complexes with metals. An important example of these are the picolylamine resins (Figure 3) which form rather stable complexes at low pH ranges (Grinstead). Sengupta recommends these kind of complex forming resins for removing trace amounts of metal cations from the background of very high concentrations of competing alkaline- and alkaline earth metal ions at acidic pH (Sengupta 1991).

Figure 3

Figure 3. Divalent metal binding onto chelating exchangers A) iminodiacetic acid B) picolylamine

The applications of chelating materials for hydrometallurgic waste treatment are however few. There is an example of an elegant method for processing stable metal complexes. The method is mostly based on the ability of special resins to adsorb metals in a low pH range (Masahide). Waste solutions containing Cu as a EDTA complex are made acidic (pH» 2.5) with sulphuric acid to break the complex and the metal is separated utilising a chelating resin. Cosmen has shown that it is possible to use chelating aminophosphonate resin for continuous ion exchange in a fluidized bed and to remove metals used in the electroplating industries (Cosmen). Also resins with iminodiacetic acid or with picolylamine groups have been shown to have good properties in practice for Cu recovery (Brown).

Chelating resins seem to have several favourable properties, but there are some important reasons why these resins are not used. First, compared to conventional resins, chelating resins are expensive. Second, chelating resins are kinetically much slower in their action and consequently large volumes of the resin are required compared to conventional resins.

3.1.2 Inorganic ion exchangers

Natural inorganic exchangers can be classified into three main categories: zeolites, oxides and clay minerals. Synthetic inorganic exchangers can be classified into the following categories: zeolites, hydrous oxides, acidic salts of polyvalent metals, salts of heteropoly acids, hexacyanoferrates and other ionic compounds (Baacke, Lieser, Weiss). Inorganic exchangers have no applications for waste management in the hydrometallurgic industry, but they have interesting properties such as resistance to decomposing in the presence of ionizing radiation or at high temperatures, which have made them interesting for the treatment of nuclear waste. There are also examples of their high selectivity towards certain ions (Komarneni, Lehto1990), which indicates that they may be suitable for certain hydrometallurgic applications.

Zeolites and clay minerals are crystalline aluminosilicates. The structure of zeolites is based on tetrahedral SiO4 and AlO4 units, which are connected by shared oxygen atoms (Figure 4). This kind of three-dimensional structure has small pores where the exchangeable ions are located and where the ion exchange reaction takes place. Silicon is tetravalent and aluminium is trivalent, which results in negatively charged framework structures and thus each mole of aluminium produces one equivalent of cation exchange capacity for the zeolite framework. In contrast to zeolites, clay mineral exchangers have two-dimensional layered structures.

Figure 4

Figure 4. The structure of the Faujasite (zeolite X and Y)

Zeolites have a rigid pore structures but the layered structures of clay minerals may have some elasticity depending on the ionic form in which the mineral exists (Scott). The ion exchange properties of zeolites and clay minerals are mainly based on the charge density and pore size of the materials.

Some studies of natural zeolites and some hydrous oxides have been made for their profitable use in waste treatment. Clinoptilolite and chabazite have been investigated for the separation of transition metals from mixed metal contaminated effluents (Ouki), also phlogopite mica and mordenite have been studied for Cs and Sr sorption (Liang, Komarneni). Clinoptilolite has been rather extensively used in radioactive waste decontamination (Hutson).

Synthetic inorganic exchangers form heterogeneous groups but there are a few selective hydrous oxides, hexacyanoferrates and zeolites which play an essentially part in nuclear waste management. Investigations of modified titanates (Anthony, R.G.1993, Bortun, Yamazaki), synthetic hydroxyapatite (Lazic) and hexacyanoferrates (Lehto 1990, Lehto 1991) have been targeted to find a method for the selective separation of Sr, Cs and Ag from nuclear wastes. Nuclear waste management is very expensive compared to waste management in the hydrometallurgic industry, so, very expensive synthetic exchangers can be used if the over-all process is still cost-effective. On the other hand, there are some studies to find inexpensive selective inorganic materials, which can compete with chelating exchangers: Modified titanates (Anthony, R.G.1993, Davis) and composite of iron oxyhydroxides and gypsum (Sengupta, 1995). Granular hexacyanoferrate has been utilised in several industrial scale applications for Cs separation (Harjula,1998)

3.1.3 Activated carbons

Activated carbons are chemically stable materials and they are known to take up metal complexes from solutions and thus they could be utilised for waste purification in certain chemical environments for the removal of metal complexes, which are commonly used in plating baths (Bansal et. al.). Especially, anionic complexes cannot be removed with ordinary cation exchange resins and metal complexation in the solution phase reduces the performance of chelating resins in metal removal processes. Besides surface area and pore structure, the sorption behaviour of activated carbon is characterised also by the raw material, preparation and activation methods.

There are only a few special applications for activated carbon in the waste management of hydrometallurgy e.g., gold extraction from cyanide solution (Dean) is advantageous compared to ion exchange. Activated carbon has been shown to have very effective sorption properties for Cd, Cr, Zn and Cu in sewerage waters ( Argo) which can be considered to be as difficult a matrix as wastewater from the plating industry.

3.1.4 Industrial ion exchange processes

Although ion exchangers are used mainly in packed bed columns, a variety of different techniques can be utilised in the ion exchange with industrial processes. Continuous operation and integration with the production process are important for effective production in plating shops. There are numerous variations of semi continuous moving-bed systems in which the resin is periodically transferred from a loading column to a regeneration column, Cosmen has successfully used this kind of technique for the separation of metals used in the plating industry (Cosmen). A step forward is the truly continuous moving-bed system, in which the resin is continuously moving from loading column to regeneration column. These continuous systems need rather large amounts of resin compared to reciprocating flow systems, in which the resin is in a packed bed column and the column height is very short but the bed has a large diameter (Brown), the reaction zone in the column covers a larger portion of the resin than in conventional packed bed column or in moving bed systems. The ion exchanger is always loaded to a low level and that results in kinetic improvements. Both the reduction in resin volume and kinetic enhancement are advantageous for the use of chelating resins. Klein has reported the successful application of a short-bed ion exchange system. In the zinc plating process, which produces effluents at 3.5-4.5m3/h, the RecofloÒ system can recover Zn and sulphuric acid cost-effectively (Klein).

3.2 Ion exchange equilibrium

An ion exchanger is an insoluble material (R) that has ionogenic groups to which counter ions (A) can be bound (IUPAC). In the ion exchange reaction (1), liquid is contacted with the ion exchanger, and ions C and A are exchanged:

Thermodynamic formulation and prediction of binary A/C ion exchange equilibrium is rather simple compared to the handling of a multicomponent equilibrium which becomes progressively more complicated as the number of exchanging components is increased (Towsend). Every exchanger has a maximum amount of charged sites per gram of exchanger; this value is called the theoretical specific capacity (Q) [meq/g]. The capacity is identical to the charge density due to ionogenic groups.

3.2.1 Selectivity

Selectivity is a characteristic of an ion exchanger, which makes the exchanger prefer one counterion to another; thus selectivity drives the reaction either to the left hand or right hand side. The selectivity coefficient for A/C ion exchange in reaction 1 can be written as follows.

Equation (2) is obtained when the mass-action law is applied to the ion exchange without activity corrections. The selectivity coefficient depends on experimental conditions (Helfferich) , and the value of the coefficient can vary in the range of a few orders of magnitude.

Helfferich lists several properties of an exchanger on which selectivity depends; the most important factors for selectivity towards metal cations are: electric interactions between ion and exchanger, pore structure and elasticity of the exchanger and other interactions between exchanger and ion. Most often selectivity is a result of a combination of two or more factors.

Electric interactions between metal cation and exchanger involve two main interactions, they are electroselectivity and electrostatic attractions.

Electroselectivity is due to the Donnan potential, which is the electric field between the exchanger and the electrolyte solution. The concentration of counter ions (cations) is larger in the exchanger and the concentration of co ions (anions) is larger in the solution. The migration of the ions from one phase to another causes the electric potential difference between the two phases. The Donnan potential increases with increasing exchange capacity, and with dilution of the solution the potential pulls counter ions to the exchanger and excludes the co ions from the solution. The force that interacts with the ions is proportional to the valence of the ions and this causes a preference for ions with higher valences.

Electrostatic attraction is interaction between counterion and ionogenic groups in situations where the character of the interaction is only weakly chemical. The force is proportional to the ion exchange capacity, which produces a higher charge density in the exchanger. Electrostatic attraction prefers ions with higher valence, smaller ions and ions with stronger polarisation.

The structural elasticity of exchanger causes selectivity partly on the basis of the hydrated ion diameter and partly on the basis of the stability of the hydration shell. Every metal cation dissolved in water has a hydrate layer, which has a characteristic thickness and stability. The model presented by Marcus describes these factors (Marcus 1987). Whenever fully hydrated large counter ions are in an elastic exchanger they produce a pressure that tends to make the solid matrix swell, the phenomenon is known as the swelling pressure. If a small counter ion replaces the large one, the swelling pressure becomes lower and this causes a preference for ions with small hydrated diameters. The effect of the swelling pressure gets stronger in parallel with increasing crosslinking of resins. Some inorganic ion exchanger materials have a similar character. The swelling of clay minerals is affected by the hydration properties of the counter ion and parallel selectivity depends on the degree of swelling (Laudelout).

Zeolites have a practically rigid structure with a regular pore size; characteristically in the range of 4-7Å, and in the case of a large cation, small pores can totally exclude the ion, because the swelling pressure can be considered to be an infinite factor. Thus, exchangers with rigid lattices work as molecular sieves. On the other hand, zeolites with high exchange capacities (high charge density) can strip the hydration shell from the metal cation. Most metals have cationic diameters less than 4 Å. In the case of the zeolites the character of selectivity changes as a function of ion exchange capacity. With a low capacity zeolite, a sieve-like action is dominant and the increasing exchange capacity emphasises the role of the hydration shell. Zeolites with pore diameters less than the fully hydrated metal prefer cations with a hydrate layer which has the lowest possible stability. Finally, the significance of the hydration shell is at its minimum with zeolites with very high capacities; High charge density can strip the hydration shell regardless of hydration energy and the selectivity depends on the association of cation to zeolite (Baacke).

Other interactions between cation and ionogenic groups are mainly the formation of chemical bonds, which may have a character in the range from a pure ionic bond to a more covalent bond, called a coordination bond in the case of transition metals. The character of the bond is due to the ability of the exchanger and of the cation to donate and to accept electrons. Inorganic exchangers form ion bonds with all metals, there are only few exceptions, like organic complexing agents, which are immobilised in inorganic substructure (Motojima, Sugawara). Immobilised ligands form sparingly soluble compounds or stable complexes with sorbed metal. Organic ion exchangers form a wide variety of bonds from ion bonds to coordination bonds. The ionic character of the bond is stressed when an exchanger has only an oxygen donor in an ionogenic group. A more covalent character appears when the exchanger has chelate forming groups or donor atoms without an electric charge such as sulphur or nitrogen. The main rule is that ion exchangers prefer ions which form stronger bonds with the solid phase.

3.2.2 The distribution coefficient and practical capacity

The distribution of metals between solution and solid depends of the selectivity of the ion exchanger, in certain experimental conditions. For trace ion exchange a useful measure for metal distribution is the distribution coefficient (KD), which is defined by:

Under the special condition that ion C is present at trace level([C] [A] ,

[RC] [RA], [RA]≈Q) selectivity coefficient KC/A is practically constant and

The KD values are thus independent of the trace ion concentration [C] and inversely proportional to [A]Zc/Za. Although in general the experimental conditions, such as metal concentration and pH, have a strong effect on the distribution coefficient, the KD value can be used as a comparative measure of the efficiencies of various exchangers and sorbents.

The capacity that can be loaded with the metal involved, in certain experimental conditions, is called the practical specific capacity QA [meq/g], which is usually experimentally determined. From this practical capacity it is possible to calculate the maximum processing capacity QP [ml/g] of the column operation, in terms of solution volume that that can be purified from ion C with a given mass of exchanger.

where [C]i = initial feed concentration of solution [eq/l].

The KD value can be used for estimation of the solution volume passed through the column at 50% breakthrough, V50. A volume VS of solution is eluted through a column, which has a mass me of exchanger R. The feed solution has a concentration [Ci] and eluate has a concentration [C] for a metal C.

The amount of substance, metal ion C, in exchanger (R) is NR after volume VS of solution has been eluted through the column. The amount of substance in the eluate is NL. The total amount of substance originally in solution volume VS is NTOT , i.e.

NR=me [RC]



Because practically all the metal is exchanged NL NR ; NTOT = NR and the capacity is totally used QA » [RC]. One obtains from equation (3)

At equilibrium the exchanger in the column does not absorb the metal any more so [C] = [Ci] and

in case of symmetrical breakthrough curve.

3.2.3 Other chemical and physical phenomena involved in ion exchange

In the waste effluents, beside exchangeable harmful metals, there are species which interfere and compete for the ion exchange reaction. Metal separation from complex waste solution by selective ion exchange is chemically rather complicated. In addition to basic ion exchange equilibria one must consider several chemical equilibria.

Hydrolysis of the exchanger (8) takes place when hydronium ion produced in the autoprotolysis of water exchanges to counter-ions loaded in the resin. This is a common phenomenon with weakly acidic resins and with inorganic exchangers. Dilution of the electrolyte drives the reaction to the right and the equilibrium pH value in batch exchange may rise even to 10-11 in dilute solution with a chelating resin (Lehto 1994). Considerable conversion to the H+-form may take place which will suppress the uptake of metal ions. A high pH value in solution does not directly affect the cation exchange reaction but changes in metal speciation in a solution may interfere with ion exchange equilibria.

Kaava 8-12

The effect of competing cations M in equation (9) transports the equilibrium to the right and depresses the KD value of metal C. This negative effect can be avoided by using an exchanger that is highly selective towards metal C.

The complexing agents L in solution suppress the ion exchange and move the equilibrium in equation (10) to the right. The ligand may have originated from the production process or it may be diluted atmospheric carbon dioxide. Hydrolysis of cations (11) has a similar effect. Complex formation has a negative effect on the distribution coefficient. If analytical methods are used to determine the total concentration of metal species (e.g. atomic absorption spectrophotometry) it is observed that the distribution KD/obs (12) of metal C is lower than the distribution of ionic C (Lehto 1995).

One possible interfering factor is also the peptization of exchanger which produces colloidal exchanger particles in the solution. Colloids act like ion exchangers, they bind ions and keep small portions of metal in solution (Harjula 1993). Peptization mostly concerns inorganic ion exchangers.