CHAPTER I INTRODUCTION
1.3 Outlines
Chapter I present the background and motivation of this research. The scope and objectives of this work are discussed. A literature review pertinent to this study, including Al hydrolysis chemistry, coagulation mechanisms, coagulation dynamics, characteristics of flocs and the application of AFM on surface observation, is presented in Chapter II.
Chapter III describes the experimental materials and methods, and the procedures of this study. A recipe for the analysis of Al speciation is presented. The preparation process of high-purity PACl was investigated. Liquid TM-AFM was developed to in-situ observe the aggregated Al13 molecules in an aqueous environment. The coagulation efficiency of various PACl coagulants as well as coagulation dynamics of fractal flocs was evaluated. Also, the surface Al composition of coagulated flocs was examined.
The major content of this thesis is shown as Fig. 1.1. Chapter IV highlights the effects of Al species transformation on the predominant destabilization mechanisms.
The effects of various parameters (i.e., pH, Al speciation, and dosage) on particle destabilization mechanisms were identified. Various destabilization mechanisms by PACl were postulated from the results of coagulation experiments and the analysis of Al species transformation.
Effects of coagulation dynamics on the formation and structure of flocs induced by various destabilization mechanisms were explored in Chapter V. Dynamic particles aggregation and breakages of flocs were evaluated by small angle static light scattering (SASLS) to monitor the growth profile of coagulated flocs.
Characteristics of flocs such as d50 size, mean size and fractal dimension during coagulation process were also determined, and the flocs were viewed by WSEM. In
addition, the morphology of Al13 aggregates was scanned by TM-AFM in liquid environment. Potential coagulation behavior of Al(OH)3 precipitates as well as Al13
aggregates was proposed.
Surface Al composition of Al(OH)3-rich and Al13-aggregate flocs were estimated by surface analysis technology in Chapter VI. The results of SEM and HR-TEM imaging were compared. The morphology of Al(OH)3 precipitates were in-situ observed by WSEM. The crystalline structure of Al(OH)3-rich and Al13-aggregate flocs were identified by HR-XRD. The tetrahedral and octahedral Al on the surface of Al(OH)3-rich and Al13-aggregate flocs were identified via XPS survey. Finally, conclusions and recommended future investigations are stated in Chapter VII.
Chapter IV
Effect of Al Species Transformation on Colloid Destabilization Mechanisms
Effects of pH on Coagulation
∗ Characterization of Coagulants
∗ Effects of pH on Turbidity Removal
∗ Effects of pH on Al Speciation in Coagulation
∗ Effects of Al Speciation on Particle Destabilization Mechanisms
Effects of Dosage on Coagulation
∗ Effects of Dosage on Particle Destabilization
∗ Reactive Al Species of Flocs
Chapter V
Formation and Structure of Flocs Induced by Various Destabilization Mechanisms
Dynamic Growth of Al-Flocs Fractal Structure of Al-Flocs
∗ Effects of Dynamic Growth of Flocs on Fractal Structure
∗ In-situ Observation on the Morphology of Flocs
In-situ Observation on the Morphology of Al13 aggregates
Predominant Destabilization Model
Chapter VI
Surface Al Composition of Al(OH)3-rich and Al13-aggregate Flocs
Structure of Al-Flocs
∗ Surface Structure of Flocs
∗ Crystalline Structure of Flocs
In-situ Observation of Al(OH)3 Precipitates Surface Composition of Al-Flocs
CHAPTER II
LITERATURE REVIEW
The aims of coagulation with aluminum coagulants in water or wastewater treatment are mainly focused on destabilization and aggregation of colloidal particles.
Aggregation of colloidal particles and the breakage of flocs are affected by the interaction forces between particles after the addition of coagulants. Colloidal interaction forces depend on the interactions of aluminum coagulant species present during coagulation with particles in water. In this section, the literatures related to the chemistry of aluminum coagulant in aqueous system, the behavior of particles aggregation and breakup of flocs during coagulation, the interparticle forces in liquid system and the application of atomic force microscope (AFM) on surface observation are reviewed.
2.1 Characterization of Aluminum Coagulants
Inorganic aluminum coagulants are commonly used to aggregate the colloidal particles in water or wastewater treatment. In general, they can be divided into aluminum salts and partially neutralized aluminum salts (i.e., polymeric aluminum coagulants). Aluminum coagulants speciation is governed by the hydrolysis characteristics of aqueous Al(III). Alum corresponds to a series of monomeric aluminum species in equilibrium with an amorphous or crystalline aluminum hydroxide precipitate (Al(OH)3). Polymeric aluminum coagulants such as polyaluminum chloride (PACl) are partially neutralized during prehydrolyzsis process, which promotes the formation of polymeric aluminum except a few monomeric aluminum species.
PACl are produced by controlling the titration of aluminum salts (e.g., AlCl3) with base and are characterized by the degree of neutralization (γ = [OH-/Al]) or basicity ((γ/3)×100%). The value of γ vary from zero to three, corresponding to basicities of 0 to 100%. The basicity affects the alkalinity consumption of coagulant as well as the aluminum species distribution.
2.1.1 Chemistry of Hydrolyzing Aluminum
The chemical behavior of Al(III) in dilute solution affects the hydrolysis of aluminum that restricts the quality of coagulant in the process of coagulant production.
In dilute solutions, the aluminum forms various monomeric and polymeric hydrolyzed aluminum species other than aluminum hydroxide precipitate (Al(OH)3). Studies have concluded that aluminum solubility can be described by the presence of monomeric and polymeric aluminum species in equilibrium with an amorphous Al(OH)3 solid phase (Baes and Mesmer, 1976; Benefield et al., 1982; Pouillot and Suty, 1992). The thermodynamic data for these species are summarized in Table 2.1.
From this, the solubility of monomeric and polymeric species and an amorphous precipitate can be plotted as a function of pH. The solubility diagram for amorphous aluminum hydroxide (Al(OH)3(am)) can be depicted as shown in Fig. 2.1. The hydrolyzed monomers include Al3+, [Al(OH)]2+, [Al(OH)]4-, and the familiar hydrolyzed polymer such as [Al2(OH)2]4+ as well as tridecamer ([Al13(OH)24]7+), in addition to Al(OH)3(am) or Al(OH)3(s) that occur near neutral pH. All of these polymeric aluminum species are formed by a series of bridging between hydroxyl of Al3+ caused by deprotonation of its bound water as pH rises, which is also called polymerization. For example, [Al2(OH)2]4+ formed through two deprotonated octahedral Al3+ bridges each other (Baes and Mesmer, 1976), as shown in Fig. 2.2.
Solubility of aluminum species is also influenced by the temperature of solution (Pernitsky and Edzwald, 2003), as well as the anions such as sulfate that can be incorporated into polymeric aluminum species to form visible precipitate (Wang et al., 2002). In fact, many aluminum species are metastable in hydrolysis process.
Therefroe, in-situ hydrolyzed aluminum species can further hydrolyze and vary with aging (Wang et al., 2004). It is very important to discover various unknown aluminum species to further understand the roles of various aluminum species on coagulation efficiency except these existent aluminum species proposed by theoretical calculation in the Al hydrolysis reactions.
Table 2.1 Summary of thermodynamic data for Al hydrolysis reactions (25)
Products Chemical Equation pK Ref.
Al3+H2O Al(OH)2+H+ 5.0 (1)
pH
Fig. 2.1 Solubility of amorphous aluminum hydroxide.
H2O
2.1.2 Identification of Hydrolyzing Aluminum Species
The identification of Al speciation is beneficial to further realize the reaction kinetics between particles and coagulants during coagulation. In the 1960s, various Al species have been proposed based on the theoretical physical chemistry, particularly polymeric aluminum species. However, the structures of these species were not further identified until the invention of nuclear magnetic resonance (NMR).
Up to now, many researchers continue to identify new polymeric aluminum species (e.g., Al30) or less prevalent polymers and to determine their formation mechanisms (Allouche et al., 2000; Chen et al., 2006; Rowsell et al., 2000). In most researches, Ferron colorimetric method and 27Al-NMR method were adopted to identify the Al speciation of coagulants.
Since 1971, Ferron colorimetric method developed by Smith (1971) has been widely adopted to quantify different aluminum species. Smith identified three forms of aluminum in partially neutralized solutions that exhibit differential reaction kinetics with ferron agent (8-hydroxy-7-iodo-5-quinoline-sulfonic acid). He suggested that the aluminum reacting instantaneously with ferron is mononuclear, while intermediate polymeric aluminum species displays pseudo-first-order reaction kinetics with ferron, and that the aluminum fraction possessing an imperceptible reaction rate with ferron is composed of giant polymeric species or incipient solid phases. Recently, the researches have proposed a modified Ferron colorimetric method to avoid some unstable features (Wang et al., 2004). Based on the kinetic difference of the reactions between the aluminum and Ferron agent, the hydrolyzed aluminum species can be divided into monomeric (Ala), polymeric (Alb), and colloidal (Alc), respectively. However, the structure of various aluminum species can not be clearly identified only by Ferron colorimetric method. Thus, there are still many debates on
the structure of these aluminum species determined by this method, especially polymeric aluminum species.
Direct 27Al-NMR investigations have cast doubt on such indirect method since
27Al-NMR spectroscopy has provided direct evidence for the existence of a number of polymeric species in a wide range of hydrolyzed Al solution (Bottero et al., 1980;
Bertsch et al., 1986a; Akitt and Elders, 1988). The 27Al-NMR can assist investigators in determining the coordination characteristics of the Al atoms contained in various polymers. By using 27Al-NMR method, many investigations have definitively proved that the Al13 polycations ([Al13O4(OH)24]7+) present generally in hydrolyzed Al solutions prepared from Al metal over various Al concentrations (from 10-5 to > 2 mol/L) by titration with bases (Bottero et al., 1987; Bertsch, 1987; Parker and Bertsch, 1992). In addition, the mononmeric Al species, oligomer, and other polymeric Al species also can be identified by 27Al-NMR method (Akitt and Farthing, 1981; Akitt, 1989).
However, there are many arguments on the correlation between the quantities of various species determined by Ferron and 27Al-NMR methods. Many researches have indicated that the quantity of Al13 determined by 27Al-NMR method can be roughly represented by that of Alb measured by Ferron colorimetric method (Liu et al., 1999; Shi et al., 2007). On the contrary, it is the case that the imperfect agreement still remains between the results of Ferron and 27Al-NMR methods, including the possibility that some implementation of Ferron method may overestimate monomeric Al fraction as well as polymeric species (Bertsch et al., 1986b). As a result, the comparison between the results of Ferron colorimetric method and 27Al-NMR method must be conducted for each research that needs the identification of Al species.
2.1.3 Synthesis and Characteristics of Al13
Tridecamer (i.e., Al13) is the major polymeric species of the PACl. The production of Al13 and its stability mainly depend on the synthesis parameters during hydrolysis process. In addition, the transformation of Al13 principally varies with pH or time. The synthesis, structure and transformation of Al13 are described as follows:
(1) Synthesis of Al13
The quantity of Al13 formed in hydrolysis process is controlled by many parameters, including basicity, total Al concentration (AlT), temperature, the addition rate of base and mixing intensity. Parker and Bertsch (1992) have reported that polymeric fraction decreases with decreasing AlT when AlT is in the range from 5 10-3 M to 110-4 M. By contrast, the formation of Al13 is not affected by Al concentration for AlT between 10-1 M and 10-3 M (Parthasarathy and Buffle, 1985).
However, the Al13 formed with the initial lower AlT was not stable due to higher pH of solution after preparation. For γ ([OH-/Al])= 2.4, Al13 is relatively insensitive to temperature below 85, while the Al13 transforms into Al13 aggregates as the temperature is over 85 (Kloprogge et al., 1992). In general, the maximum fraction of Al13 only can be produced when the value of γ is controlled between 2.2 and 2.4. Otherwise, the amount of Al13 formed during titration is increased as a function of base injection rate and reaches a maximum at higher base addition rate (Bertsch, 1987). The slower base addition rate leads to the formation of stable Al13, and the pH of solution decreases with aging (Parthasarathy and Buffle, 1985). On the other hand, mixing intensity could affect Al13 production, stability and precipitation during PACl preparation. Poorer mixing is required for the formation of Al13 during the preparation of PACl (Clark and Flora, 1991). Less intense mixing
chemical conditions for Al13 production to exist longer during base titration. On the contrary, other studies have indicated that the more mixing intensity results in higher content of Al13 after titration (Bertsch, 1987; Kloprogge et al., 1992). Because mixing efficiency depends on the mixing type and mixing energy, it is difficult to conclude that the more or less intense mixing is favorable for the Al13 production during base titration.
(2) Structure of Al13
The Al13 molecule, AlO4Al12(OH)24(H2O)127+, which is composed of a Al(O)4
tetrahedron surrounded by 12 octahedrally coordinated Al(O)6 sharing edges (Johansson, 1960) shown as Fig. 2.3 in which µ2-OH is a hydroxyl ligand bridging two Al atoms; µ4-Oµ is an oxo ligand bridging four Al atoms; η1-H2O is a bound and nonbridging water molecule. This kind of Al13 configuration has been commonly referred to as “Keggin” structure. The size of Al13 as determined by small angle X-ray scattering is approximately 1.2 nm (Bottero et al., 1982). However, there are many arguments about the formation mechanisms of Al13 in aqueous solution. One
"inhomogeneous model" has been proposed that Al13 is generated at localized high basic strength region during base addition that allows Al(III) to become four-fold coordination, e.g., Al(OH)4-, which is considered to be essential precursor for Al13
formation (Bertsch, 1987; Kloprogge et al., 1992).
η1− Ο Η 2
µ2− Ο Η ' µ2− Ο Η
µ4− Ο
Fig. 2.3 The ε-Keggin structure of Al13 polycation.(Johansson, 1960)
(3) Transformation of Al13
Al13 is metastable with respect to aluminium hydroxide (Al(OH)3), which transforms over time to crystalline Al(OH)3 such as gibbsite α-Al(OH)3 or bayerite
γ-Al(OH)3 in the hydrolysis processes (Violante and Huang, 1985; Bradley et al., 1993). Because different oxygen sites of Al13 have different water exchange rates which relate to decomposition of Al13 (Phillips et al., 2000), the transformation Al13
occur when pH changes. Al13 favors aggregation when pH is over 6 (Furrer et al., 1992), and the size of Al13 aggregates increase gradually from 570 nm at pH 4.35 to 1200 nm at round pH 6.5 (Dubbin and Sposito, 2005). Moreover, the condensation of Al13 could form at a specific pH due to shear-induced mixing (Furrer et al., 2002).
On the contrary, Al13 aggregates are easily decomposed by H+ as the pH value decreases (Furrer et al., 1999). The Al13 deprotonation varies with the concentration of Al13 in addition to the pH of solution. During condensation, Al13 can transform into various Al13 aggregates with increasing [OH-]/[Al] ratio and total Al concentration (Bottero et al., 1987). The aggregation of Al13 often occurs at higher total Al concentrations. In concentrated Al solution (>0.110-3 mol/L) synthesized
or aged at a given temperature, other larger polymeric Al species that are categorized as "Al13-like" polycations have been proposed and suggested (Akitt and Farthing, 1981; Fu et al., 1991). With aging, the Al13 structure is gradually lost and that they eventually take on the crystalline structure of gibbsite (Bottero et al., 1987; Bradley et al., 1993). Other study has suggested that the in-situ formed Al13 (i.e., metastable Al13) during hydrolysis do not condense with aging, but undergo direct structural rearrangement and then transforms into small polymers of the gibbsite fragment configuration, and may nucleate and commence the process of gibbsite crystallite growth (Hsu, 1988). On the other hand, organic matters and inorganic anion also have a profound effect on the structural transformations of the aggregated Al13 by complexing, ligand exchanging or hydrogen bonding (Molis et al., 1996; Wang et al., 2002). However, there are debates on the formation pathway of the Al13 aggregates, which depends on the reaction conditions such as Al concentration, pH, temperature, and aging processes (Sposito, 1996).
2.2 Alum Coagulation
Coagulation reaction with aluminum coagulants involves three continuous sequential steps which can form various flocs at different conditions such as pH, dosage, mixing intensity and temperature. Schematic illustration of coagulation process is shown in Fig. 2.4. (a) Addition and activation of the coagulant through a series of hydrolysis reactions in water and form active coagulant species. (b) Reactions of the active coagulant species with the stabilized colloidal particles to destabilize them through "rapid-mixing". The primary function of the rapid-mixing is to disperse the coagulants uniformly in a short period of time in order to efficiently destabilize colloidal particles. (c) Aggregations of the destabilized colloidal particles and collide to form microflocs. Through "slow-mixing", namely flocculation, these microflocs can form readily settleable or filterable flocs that can be removed by the subsequent sedimentation and filtration. In this section, the researches with respect to colloidal interaction forces, coagulation mechanisms and removal of colloidal particles will be reviewed.
Fig. 2.4 Schematic diagram of coagulation process: (a) coagulant hydrolysis (b) destabilization of colloidal particles, and (c) aggregation of destabilized colloidal particles.
2.2.1 Colloidal Interaction Forces
Interaction forces between colloidal particles affect significantly the stability of colloidal particles and their aggregation when colloidal particles approach each other.
The interparticle forces can be appropriately explained by the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory (Derjaguin and Landau, 1941;
Verwey and Overbeek, 1948). For DLVO theory, the magnitudes of the component and total interaction potentials at various distance of separation for two spheres with equal size are depicted in Fig. 2.5. In this theory, the van der Waals electrostatic forces were mainly considered, as indicated in Fig. 2.5 (a). When the potential energy turns into more and more attractive as colloids approach one another, colloids can be aggregated in the primary minimum of the potential energy, as illustrated in Fig. 2.5 (b). Fig. 2.5 (c) shows the second-minimum in the potential energy curve in which the potential energy becomes more and more attractive as colloids approach closer; however, the energy barrier arises after a certain separation distance. The aggregation of colloids also occurs by the long-range attraction under a long separation distance when the energy at second-minimum sink is favored, which can lead to the formation of fairly weak aggregates. By contrast, there is no aggregation between colloids and the colloids are stable when the potential energy keeps repulsive, as seen in Fig. 2.5 (d). In addition, non-DLVO forces, including hydrophobic attraction and steric repulsion, are also the considered component in the colloids interaction process. These interaction forces are important for the understanding of the properties of various colloid dispersions of practical interest.
When a surface has no polar or ionic groups or hydrogen-bonding sites, there is no affinity for water and the surface is said to be hydrophobic (Elimelech et al., 1995).
A hydrophobic surface is inert to water and can not bind water molecules by hydrogen
or ionic bonds (Israelachvili, 1991). The presence of a hydrophobic surface can restrict the natural structuring tendency of water, imposing a barrier which prevents the growth of clusters in a given direction. Water confined in a gap between two hydrophobic surfaces is unable to form clusters and result in an increased free energy of the water relative to bulk water. As a result, an attraction between hydrophobic surfaces shows up as a consequence of water molecules migrating from the gap to the bulk water. Such an attraction could be quite long range and play a major role in promoting flocculation. Steric repulsion often appears when two adsorbed layers of polymeric chain molecules overlap, where particle surface is saturated with polymer such that the polymer loops and trains form a relatively thick layer of adsorbed polymers. For a high density of adsorbed polymer molecules on the surface of particles, the steric repulsion is strong and stabilizes the colloid dispersions. Thus, the steric stabilization of a colloid dispersion is more easily achieved by attaching of long chain polymers to colloidal particles.
Fig. 2.5 Potential energy versus distance profiles of DLVO interactions. (a) van der Waals and electrostatic forces (b) primary-minimum of the potential energy (c) secondary-minimum of the potential energy and (d) potential energy of a stable suspended particles.
2.2.2 Mechanisms of Coagulation
The mechanisms of coagulation vary with the type of coagulants and the
The mechanisms of coagulation vary with the type of coagulants and the