國 立 交 通 大 學
環 境 工 程 研 究 所
博 士 論 文
聚氯化鋁水解物種之混凝行為：膠體去穩定機制及膠羽形成分析
COAGULATION BEHAVIOR OF HYDROLYZED PACl SPECIES:
COLLOID DESTABILIZATION MECHANISMS AND
FLOCS FORMATION ANALYSIS
研 究 生：林志麟
聚氯化鋁水解物種之混凝行為：膠體去穩定機制及膠羽形成分析
COAGULATION BEHAVIOR OF HYDROLYZED PACl SPECIES:
COLLOID DESTABILIZATION MECHANISMS AND
FLOCS FORMATION ANALYSIS
研 究 生：林志麟 Student：Jrlin Lin
指導教授：黃志彬 博士 Advisor：Dr. Chihpin Huang
國立交通大學
環境工程研究所
博士論文
A THESIS
SUBMITTED TO INSTITUTE OF ENVIRONMENTAL ENGINEERING
COLLEGE OF ENGINEERING
NATIONAL CHIAO TUNG UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN
ENVIRONMENTAL ENGINEERING
JULY 2008
HSINCHU, TAIWAN, REPUBLIC OF CHINA
聚氯化鋁水解物種之混凝行為：膠體去穩定機制及膠羽形成分析
研究生：林志麟 指導教授：黃志彬 博士
國立交通大學環境工程研究所
摘要
在水及廢水處理過程中，聚氯化鋁(PACl)是最常被使用於混凝程序以去穩定 顆粒之混凝劑。聚氯化鋁混凝劑之效用取決於水解物種與水中顆粒間之作用。水 解的鋁物種如聚合鋁及氫氧化鋁會嚴重地影響膠體顆粒之混凝機制，然後會影響 膠羽的形成。因為水解鋁物種會隨 pH 及總鋁濃度改變，故瞭解各種聚氯化鋁混 凝劑之鋁形態分佈及在各 pH 值與加藥量下之主要水解鋁物種對於 PACl 之應用 相當重要。 首先，以瓶杯試驗及Ferron法評估混凝過程中各種鋁水解物種對高嶺土顆粒 去穩定之影響，並利用即時的診斷技術探究膠羽的形成及表面結構。同時，進行 膠羽表面之鋁元素組成分析。此外，藉由輕敲式原子力顯微鏡及濕式掃描式電子 顯微鏡觀察Al13聚集體與氫氧化鋁之表面結構。 在中性條件下，無論加藥量多寡，PACl-C混凝膠羽之形成主要依賴氫氧化 鋁沉澱物。相對的，在鹼性條件下，具有高Al13含量之PACl-E主要以Al13聚集體 行電性補釘及電性中和之混凝機制。在鹼性及低加藥量條件下， 高純度聚氯化 鋁(PACl-Al13)混凝主要以電性補釘去穩定顆粒；在足夠加藥量下，由於具有電中 性之Al13聚集體形成促使顆粒間架橋變成主要的混凝機制。 在沉澱絆除或掃除機制及電性補釘機制下形成之膠羽，隨著加藥量增加，膠 羽結構變成較密實，此時膠羽遭遇破碎會增加膠羽之碎形維度。相反地，PACl-Al13 混凝之膠羽結構隨著加藥量增加而變鬆散。另一方面，沉澱絆除或掃除機制下形成之膠羽具有粗糙的外觀，而電性補釘及電性中和機制下形成之膠羽具有光滑的
表面。PACl-Al13混凝所形成之架橋作用會造成鬆散的結構且毛茸的外觀。此條件
下，存在一些由盤繞的Al13所構成之線條狀Al13聚集體與其他碎形結構之Al13 聚集
體。
本質上，富含氫氧化鋁的膠羽不具有良好的晶形結構，而含有Al13聚集體膠
羽則具有類似Al13的晶形結構。在富含氫氧化鋁的膠羽表面上，有許多無定形結
晶之氫氧化鋁沉澱物具有四面體或八面體結構，而膠體狀之氫氧化鋁具有凹陷的
Coagulation Behavior of Hydrolyzed PACl Species: Colloid
Destabilization Mechanisms and Flocs Formation Analysis
Student: Jrlin Lin Advisor: Dr. Chihpin Huang
Institute of Environmental Engineering
National Chiao Tung University
ABSTRACT
Polyaluminum chloride (PACl) is the most frequently used to destabilize particles for coagulation in water or wastewater treatment. Effective coagulation by PACl depends on the interaction between hydrolyzed Al species and particles in water. Hydrolyzed Al species, such as polymeric Al or Al(OH)3, affect significantly coagulation mechanisms of colloidal particles, which thereafter influence the formation of flocs. Since hydrolyzed Al species varies with pH as well as concentration of Al, it is very important to realize the Al speciation of various PACl coagulants, and their predominant hydrolyzed Al species at various pH values and dosages for coagulation in practice.
Effects of various hydrolyzed Al species on the destabilization of kaolin particles in coagulation were evaluated by jar test as well as Ferron method. Formation of and structure of flocs were also investigated via an in-situ diagnostic technology. In-situ morphology of the flocs formed after coagulation was viewed through a wet SEM assay, and the Al composition of these flocs were further surveyed by XPS. Moreover, in-situ configuration of the Al13 aggregates as well as Al(OH)3 precipitates
The formation of sweep flocs by PACl-C coagulation at neutral pH relied on Al(OH)3 precipitates regardless of the dosage applied. By contrast, the PACl-E containing a high percentage of Al13 caused either electrostatic patch or charge neutralization mechanisms with Al13 aggregates at alkaline pH. For high-purity PACl (PACl-Al13) coagulation, electrostatic patch was responsible for particle destabilization at alkaline pH and low dosage. Interparticle bridging becomes the major mechanism at sufficient dosage due to the formation of Al13 aggregates with nearly zero charge.
The structure of flocs formed by enmeshment or sweep flocculation and electrostatic patch becomes more compact with dosage, in which the breakage of flocs increases the fractal dimension of flocs. On the contrary, flocs coagulated by PACl-Al13 become looser with dosage. On the other hand, enmeshment or sweep flocculation caused sweep flocs with a rough and ragged contour, while electrostatic patch or charge neutralization induced flocs with a smooth and glossy surface. PACl-Al13 coagulation induced by interparticle bridging brought the flocs of a looser structure with a fluffy contour. At such condition, some larger linear Al13 aggregates composed of a chain of coiled Al13 and several coiled Al13 aggregates with different dimensions can be observed.
Intrinsically, the Al(OH)3-rich flocs do not possess well-formed crystalline structure, while the Al13-aggregate flocs possess a Al13-like crystalline structure. There are multitude of amorphous Al(OH)3 precipitates that involve either tetrahedral AlIV(O)4 or octahedral AlVI(O)6 center on the surface of Al(OH)3-rich flocs, while the colloidal Al(OH)3(s) has a sunken surface. It has been proved that the existence of Al13 aggregates on the surface of flocs coagulated by PACl-Al13 at alkaline pH.
誌 謝
「透過人生的孤寂，到達成功的彼岸」，這句話是中學時老師勉勵同學的話。 在五年的交大求學歲月中，孤寂陪伴我度過無數撰寫研究計畫、報告及研讀學術 論文的日子。在此過程中，感謝黃志彬教授於研究報告撰寫上的教導及鞭策，使 我能在學術研究的表達上有所進步。同時，感謝袁如馨老師對學生的細心關懷及 鼓勵，在學生求學過程中遭遇困難時，給予莫大的幫助。此外，要感謝秦靜如老 師在學術研究及論文撰寫上的協助，並能在忙碌的教學及研究外，撥空與學生討 論研究的方向及實驗的規劃。口試期間，承蒙葉宣顯教授、康世芳教授、李篤中 教授及張淑閔老師對本論文的撰寫提供寶貴的意見，使學生受益良多。 在博士班的修業過程中，感謝阿甘學長、佳欣、容忍、柏廷、肇毅、韋弘、 宏杰、欣彗、育俊、文彬、奕甫、嘉玲、靜逸、璧如、雅茹（OK）及怜秀（狗 狗）等學弟妹於研究工作上的協助與幫忙，使我能順利完成研究計畫及報告的工 作。在生活上，感謝昌郁、政倫及文善在研究之餘的陪伴及關心，使我能忘卻煩 惱，勇往直前。此外，感謝中國科學院生態環境研究中心曲久輝教授及王東升教 授熱心提供高純度聚氯化鋁混凝劑，特別感謝王東升教授在膠體混凝研究上的啟 發。同時，亦感謝中國科學院生態環境研究中心李濤博士對於本論文實驗規劃上 的幫忙。 在個人成長及求學過程中，最感謝爸爸及媽媽對我的支持及鼓勵，以及奶奶 的關心，讓我能一路從小到大無後顧之憂的唸書。感謝康世芳教授對學生的鼓 勵，使我能完成博士學位。感謝靖宜在博士班修業最後一年中於論文撰寫上的幫 忙及生活上的照顧，使我能順利完成學業。最後，僅將完成此論文的喜悅獻給我 的家人及身邊的朋友，願你們與我分享。CONTENTS
摘 要………I ABSTRACT………..III 誌 謝………V CONTENTS………..VI LIST OF TABLES………...………IX LIST OF FIGURES………...………X CHAPTER I INTRODUCTION……….……….………1 1.1 Background………11.2 Scope and objectives………..………4
1.3 Outlines………..5
CHAPTER II LITERATURE REVIEW……….……….………..8
2.1 Characterization of Aluminum Coagulants………8
2.1.1 Chemistry of Hydrolyzing Aluminum………9
2.1.2 Identification of Hydrolyzing Aluminum Species………12
2.1.3 Synthesis and Characteristics of Al13………....………...14
2.2 Alum Coagulation……….……….18
2.2.1 Colloidal Interaction Forces………...….……….19
2.2.2 Mechanisms of Coagulation………...….……….22
2.3 Coagulation Dynamics………...….……….28
2.3.1 Floc Formation……….………29
2.3.2 Floc Breakage and Restructuring………...…..………31
2.4 Floc Physical Characteristics………..…….………33
2.4.1 Floc Size……….…..………34
2.4.2 Fractal Dimension………...…….………35
2.4.3 Floc Strength……….…………...37
2.5 Surface Observation by Atomic Force Microscopy…….………38
CHAPTER III EXPERIMENTAL MATERIALS AND METHODS….………41
3.1 Materials………...41
3.1.1 Synthetic Water Sample………...………41
3.1.2 Coagulants………43
3.2.1 Ferron Assay………....………44
3.2.2 27Al-Nuclear Magnetic Resonance (NMR)………..………47
3.2.3 Preparation and Characterization of PACl-Al13………...………48
3.2.4 Coagulation Experiments………...……..54
3.2.5 Measurement of Floc Size and Structure………...……..57
3.2.6 Solid-state 27Al Magic-angle Spinning Nuclear Magnetic Resonance....59
3.2.7 Wet Scanning Electron Microscope………..………...…60
3.2.8 Atomic Force Microscope………....62
3.2.9 Field-Emission Electron Microscope………...…63
3.2.10 High-resolution X-ray Powder Diffractometer……..………..63
3.2.11 Fourier Transform Infrared Spectra……….63
3.2.12 X-ray Photoelectron Spectroscopy………..……….…64
CHAPTER IV EFFECT OF Al SPECIES TRANSFORMATION ON COLLOID DESTABILIZATION MECHANISMS.………...…………65
4.1 Effects of pH on Coagulation………...…65
4.1.1 Characterization of Coagulants……….66
4.1.2 Effects of pH on Turbidity Removal………….………...68
4.1.3 Effects of pH on Al Speciation in Coagulation…...70
4.1.4 Effects of Al Speciation on Particle Destabilization Mechanisms...72
4.2 Effects of Dosage on Coagulation Efficiency...74
4.2.1 Effects of Dosage on Particle Destabilization…...………..…….74
4.2.2 Reactive Al Species of Flocs……….…...79
4.3 Summary………...82
CHAPTER V FORMATION AND STRUCTURE OF FRACTAL FLOCS INDUCED BY VARIOUS DESTABILIZATION MECHANISMS………...83
5.1 Dynamic Growth of Al-Floc………84
5.2 Fractal Structure of Al-Flocs...………..…...88
5.2.1 Effects of Dynamic Growth of Floc on Fractal Structure……...…88
5.2.2 In-situ Observation on the Morphology of Flocs………...……94
5.3 In-situ Observation on the Morphology of Al13 Aggregates……...….……97
5.4 Predominant Destabilization Mechanisms Model…………..…….….…101
5.5 Summary………105
CHAPTER VI SURFACE Al COMPOSITION OF Al(OH)3-RICH AND Al13-AGGREGATE FLOCS………....…………..…..106
6.1.1 Surface structure of Flocs…………...………..107
6.1.2 Crystalline Structure of Flocs……….111
6.2 In-situ Observation of Al(OH)3 Precipitates………..…113
6.3 Surface Composition of Al-Flocs………...……115
6.4 Summary……….119
CHAPTER VII CONCLUSIONS AND RECOMMENDATIONS…...………120
7.1 Conclusions………120
7.2 Recommendations………..122
BIBLIOGRAPHY………...……….…123 VITA
LIST OF TABLES
Table 2.1 Summary of thermodynamic data for Al hydrolysis reactions (25℃).….10 Table 3.1 FTIR vibrational assignment of Al13 powder………53 Table 4.1 Characteristics of coagulants by Ferron assay and 27Al-NMR methods...67 Table 5.1 Fractal dimension (Ds) of various flocs during coagulation.……….93
LIST OF FIGURES
Figure 1.1 Schematic diagram of this thesis……..………..7
Figure 2.1 Solubility of amorphous aluminum hydroxide………11
Figure 2.2 Polymerization of Al2(OH)24+………..11
Figure 2.3 The ε-Keggin structure of Al13 polycation………...16
Figure 2.4 Schematic diagram of coagulation process: (a) coagulant hydrolysis (b) destabilization of colloidal particles, and (c) aggregation of destabilized colloidal particles………...………...…..18
Figure 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………...………..21
Figure 2.6 Electrostatic patch model of the interaction between negatively charged particles and adsorbed cationic polymers………...……….24
Figure 2.7 Fractal flocs formed by (a) particle-cluster aggregation (b) cluster-cluster aggregation………...……….………...……....…30
Figure 2.8 Two proposed mechanisms for the breakage of aggregates at different shear conditions: (a) Splitting (b) Surface Erosion………….………….32
Figure 2.9 Schematic diagram of atomic force microscope……….……….38
Figure 2.10 Illustration of the interaction force between tip-to-sample surface distance………...………...39
Figure 3.1 SEM micrographics of purified kaolin powders…………...42
Figure 3.2 Particle size distribution of the kaolin suspension ………...42
Figure 3.3 Illustration of Ferron colorimetric method……….………...46
Figure 3.4 The SEM images of Al13-(SO4)n precipitate (×700)….……...…...49
Figure 3.5 Solid-state 27Al-NMR spectrum of solid Al 13 powder.…………...52
Figure 3.6 HR-XRD pattern of solid Al13 powder……….…………...52
Figure 3.8 G curve for flat paddle in the gator jar……….…………...56 Figure 3.9 Construction of laboratory coagulation system incorporated with
SASLS……….…………...57 Figure 3.10 Setup of WSEM: (a) Top scheme of QX-102 capsule and (b)
Cross-section scheme of capsule……….…………...61 Figure 3.11 Setup of liquid tapping mode atomic force microscope…………...63 Figure 4.1 27Al-NMR spectra of (a) PACl-C (b) PACl-E (c) PACl-Al13…….……..67 Figure 4.2 The residual turbidity and the zeta potential for (a) PACl-C (7% Al13) (b)
PACl-E (60% Al13) and (c) PACl-Al13 (96% Al13) coagulation at various pH values and 1mg/L as Al...69 Figure 4.3 Al speciation of (a) PACl-C (b) PACl-E (c) PACl-Al13 at various pH
values under the concentration of 2×10-4 mol Al/L……….71 Figure 4.4 Dosage effects on the residual turbidity and the zeta potential for (a)
PACl-C (7% Al13) coagulation at pH 7.5 (b) PACl-E (60% Al13) and (c) PACl-Al13 (96% Al13) coagulation at pH 10………77 Figure 4.5 The zeta potential of Al species for PACl-Al13 coagulation at pH 10 and
various dosages without kaolin particles……….……78 Figure 4.6 Solid-state 27Al MAS-NMR spectra of freeze-dried kaolin flocs
coagulated by three coagulants at various dosages. PACl-C:(a) 1 mg/L (b) 8 mg/L; PACl-E:(c) 1 mg/L (d) 8 mg/L; PACl-Al13:(e) 1 mg/L (f) 8 mg/L...………..…..…..81 Figure 5.1 The particle aggregation dynamics induced by (a) PACl-C coagulation at
pH 7.5 and (b) PACl-E and (c) PACl-Al13 coagulation at pH 10.….…...87 Figure 5.2 The fractal dimensions (Ds) of flocs coagulated by (a) PACl-C at pH 7.5
and (b) PACl-E and (c) PACl-Al13 at pH 10 and various dosages……...92 Figure 5.3 The WSEM micrographics of flocs coagulated at various coagulants and
dosages under 1 min rapid mixing followed by 20 min slow mixing. PACl-C: (a) 1mg/L and (b) 8 mg/L; PACl-E: (c) 1 mg/L and (d) 8 mg/L; PACl-Al13: (e) 1 mg/L and (f) 8 mg/L………...96 Figure 5.4 TM-AFM topographic image of the Al13 adsorbed on the mica in liquid
Figure 5.5 Size distributions of Al13 in PACl-Al13 coagulation without kaolin particles at various dosages (rapid mixing: 200 rpm; reaction time: 1 min)………100 Figure 5.6 Schematic representation of the destabilization mechanism induced by
PACl-C coagulation at neutral pH. (a) Precipitation Charge
Neutralization (b) Enmeshment……….103 Figure 5.7 Schematic representation of the destabilization mechanism induced by
Al13 aggregates. (a) Electrostatic Patch (b) Interparticle Bridging…...104 Figure 6.1 SEM micrographics of Al(OH)3-rich and Al13-aggregate flocs formed by PACl-C and PACl-Al13 coagulation at 8 mg/L as Al……….….109
Figure 6.2 HR-TEM images of (a) Al(OH)3-rich and (b) Al13-aggregate flocs formed by PACl-C and PACl-Al13 coagulation at 8 mg/L as Al………110
Figure 6.3 HR-XRD patterns of (a) Kaolin powder (b) Al(OH)3-rich and (c)
Al13-aggregate flocs………...112 Figure 6.4 WSEM micrographics of Al(OH)3 precipitates formed after 1 min rapid
mixing at 200 rpm: (a) (×3,750); (b) (×50,000)…...114 Figure 6.5 Al 2p scans of XPS for (a) Kaolin powder (b) Al(OH)3-rich and (c)
CHAPTER I
INTRODUCTION
1.1 Background
In water or wastewater treatment, coagulation process is most commonly applied as an essential pretreatment process to destabilize colloidal material and suspended impurities, including organic and inorganic substances, for subsequent removal by separation processes such as sedimentation, flotation and filtration. Coagulation usually requires the addition of chemicals, namely coagulant, as an aid to the aggregation of particles for particle removal. Hydrolyzed metal salts such as aluminum and ferric salts are regularly used as coagulants, particularly alum and polyaluminum chloride (PACl). After the addition of aluminum coagulants, the hydrolysis reactions form a serious of soluble ionic species, including monomeric and polymeric aluminum species, and solid precipitates. Because an efficient coagulation is required to yield colossal and compact flocs through the destabilization of colloidal particles with various hydrolyzed aluminum species for subsequent rapid separation of flocs from water, the aggregation process such as growth and breakup of flocs affects the coagulation efficiency significantly. However, the production of flocs mainly depends on the interaction of destabilized colloidal particles in coagulation. Therefore, it is essential to understand the physicochemical reactions between the colloidal particles and hydrolyzed aluminum species of PACl in order to clarify the predominant mechanism of coagulation in practice.
Chemistry of alum and PACl coagulants have been extensively studied, and various hydrolyzed aluminum species in water are proposed and suggested. PACl is
wastewater treatment. Effective coagulation by PACl depends on the interaction between hydrolyzed aluminum species and colloidal particles in water. Many studies have proved that Al(III) species is driven into various hydrolyzed Al species such as [Al(OH)]+, [Al(OH)]2+, [Al2(OH)2]4+, [Al3(OH)4]5+, [Al13O4(OH)24(H2O)12]7+,
and aluminum hydroxide (Al(OH)3) in the hydrolysis process, which are governed by
the [OH-/Al] ratio (Akitt and Fathing, 1981; Bertsch et al., 1986b; Bottero et al., 1987). Poor performance of aluminum coagulants has been blamed on its low content in critical hydrolysis products, especially the polycationic aluminum (Al13).
Al13 is the highly charged polymeric aluminum species, AlO4Al12(OH)24(H2O)127+,
which has been commonly acknowledged as the critical species in colloidal particles coagulation by strong charge neutralization that facilitates the solid-liquid separation process (Gao et al., 2005; Wang and Hsu, 1994). In the past decade, a PACl with over 70% Al13 of total Al concentration has been produced by controlling the OH-/Al ratio during the preparation of PACl (Liu et al., 1999; Wang et al., 2004).
PACl containing over 90% Al13 has also been successfully prepared by sulfate
precipitation and nitrate metathesis (SO42-/Ba2+ separation method) from the
pre-hydrolyzed PACl solution produced by alkaline titration method (Shi et al., 2007). However, the stabilization of hydrolyzed Al species have a profound effect on coagulation mechanisms that is closely related to the formation of flocs as well as particles removal.
Different aqueous Al species possessing different specific hydroxo-Al structure deprotonate with pH. Al(III) species transforms into voluminous amorphous aluminum hydroxide (Al(OH)3(am)) at neutral pH (Van Benschoten and Edzwald,
1990), while Al13 is metastable with respect to Al(OH)3 and transforms to various
(Violante and Huang, 1985; Bradley et al., 1993). The condensation of Al13 could
form at a specific pH due to shear-induced mixing (Furrer et al., 2002). During condensation, Al13 can transform into various Al13 aggregates with increasing
[OH-]/[Al] ratio and total Al concentration (Bottero et al., 1987). On the contrary, Al13 aggregates can be decomposed by H+ as the pH value decreases (Furrer et al.,
1992). Furthermore, different oxygen sites of Al13 have different water exchange
rates which relate to decomposition and condensation of Al13 (Phillips et al., 2000),
resulting in various Al13 aggregates. However, there are many 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). Many researchers have clarified that coagulation mechanisms of various PACl coagulants for the removal of colloid particles or organic matter (Gao et al., 2005; Hu et al., 2006; Kazpard et al., 2006; Yan et al., 2007), involving charge neutralization, complexation, and enmeshment (sweep flocculation). Moreover, it has been inferred that Al13 aggregates could destabilize particles by either electrostatic patch or bridging
when PACl containing high Al13 contents is applied at high pH and sufficient dosage
(Chen et al., 2006; Ye et al., 2007; Wu et al., 2007). However, these assumptions have never been verified and could not accurately illustrate the coagulation behavior of Al13 aggregates because it is a formidable task to quantify the Al13 aggregates.
For the application of PACl in water or wastewater treatment, further investigations on the unknown hydrolyzed Al species, particularly Al13 aggregates, and their real
conformations and composition during coagulation are dearly necessary to identify the predominant interaction between various hydrolyzed Al species and colloidal particles.
1.2 Scope and Objectives
The overall goals of the research presented in this thesis are to provide in-depth understanding of particles interaction in coagulation with various PACl coagulants. Effects of various hydrolyzed Al species on the destabilization of kaolin particles in coagulation were evaluated by jar tests, and the growth of flocs was also studied via in-situ diagnostic technology. In-situ observation of flocs formed after coagulation was conducted through a novel wet scanning electron microscope (WSEM) assay. Configuration of the Al13 aggregates as well as Al(OH)3 precipitates were also
observed by using tapping mode atomic force microscope (TM-AFM) and WSEM in liquid system, respectively. Several surface analysis technologies, including scanning electron microscope (SEM), high-resolution transmission electron microscope (HR-TEM), high-resolution X-ray powder diffractometer (HR-XRD), X-ray photoelectron spectroscopy (XPS), were employed to further survey the Al composition on the surface of Al(OH)3-rich and Al13-aggregate flocs in order to
comprehend the characteristics of hydrolyzed Al species thoroughly. The specific objectives for this research are summarized as follows:
1. To investigate the effects of Al(III) speciation on the predominant destabilization
mechanism.
2. To study dynamic growth and structure of Al-flocs formed under various
destabilization mechanisms.
3. To determine the role of Al13 aggregates in colloid coagulation.
4. To identify the composition of reactive Al species on the surface of Al(OH)3-rich
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
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,
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
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
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) Al3+2H2O Al(OH)2+2H+ 9.39 (1) Monomer Al3+4H2O Al(OH)4-4H+ 23.0 (1) Polymer 2Al3+2H 2O Al2(OH)24+2H+ 6Al3+15H2O Al6(OH)153+15H+ 8 Al3+20H 2O Al8(OH)204+20H+ 13Al3+28H2O Al13(OH)247+32H+ 6.27 47.0 68.7 98.7 (2) (2) (2) (3) Precipitate Al(OH)3(am) Al
3+3OH
-Al(OH)3(s) Al3+3OH
-31.5
33.5 (3) (3) Source: (1) Baes and Mesmer (1976)
(2) Benefield et al. (1982) (3) Pouillot and Suty (1992)
pH 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 lo g [ s pe ci es c on ce rn tr at io n ] -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 Al(OH)4 -Al8(OH)203+ Al13(OH)247+ Al(OH) 2+ Al3+ Al2(OH)24+ Al6(OH)153+ Al(OH)3 (s)
Fig. 2.1 Solubility of amorphous aluminum hydroxide.
H2O H2O H2O H2O H2O H2O Al H2O H2O H2O H2O H2O H2O Al H2O H2O H2O H2O Al H2O OH H2O OH H2O H2O Al
Fig. 2.2 Polymerization of Al2(OH)24+.(Baes and Mesmer, 1976)
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
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, Al
13 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
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
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 operational parameters. The destabilization of inorganic colloidal particles is dictated by principal mechanisms that substantially affect coagulation dynamics and the properties of flocs such as size, structure and strength. Destabilization of inorganic colloidal particles will be discussed in this section as follows:
(1) Electrical Double-Layer (EDL) Compression
Particles in liquid system in which ions are present are usually charged due to adsorption, ionization, hydration, or isomorphic replacement on the surfaces. These surface charges create an electrostatic field tha causes a counter charge, and thus covers the electric surface charge. Some of the counter-ions might specifically adsorb near the surface and build an inner sub-layer, or so-called Stern layer; the outer part of the screening layer is called the diffuse layer (Lyklema, 1995). The electric potential has a maximum value at the particle surface and decreases with distance from the surface. The electric potential surrounding the particle is so-called zeta potential which is located in the diffuse layer. The electrical double layers is related to electrostatic interactions between particles, including surface potential and the electrical double layer thickness. Surface potentials can be influenced by the ionic strngth and adsorbed counterions in the system. The electrical double layer thickness is formed by the surface charge of particles and the associated counter-ions charge. According to the DLVO theory, higher ionic strngth can cause higher counter-ions concentration in the diffuse layer, which compresses the thickness of the diffused electrical double layer, lowering the surface potential and the energy barrier. Therefore, Al3+ is more efficient than Ca2+ or Na+ in the electrical double layer compression. The interaction between particles can be evaluated in terms of
potential energy. The overall interaction is unfavorable (repulsive) if the potential energy is positive. If the potential energy is negative, the net effect is attractive. Thus, the proper ionic strength in coagulation process is necessary to reduce the net potential energy for particle aggregation.
(2) Adsorption-Charge Neutralization
Inorganic particles such as alumino-silicates (clays) and silica particles are generally electrically charged due to the isomorphic substitution of ions in the bulk solid phase, or to the reactions of surface functional groups with dissolved ions in the aqueous phase (Sposito, 1993). The major inorganic colloids are generally negatively charged in water due to their low zero point of charge (Lerman, 1979). The positively charged Al(III) hydrolysis products such as Al137+ polycation can
specifically be adsorbed on the surface of colloids, which can modify the original charge on the surface of colloids. These interactions can neutralize the surface charge of the colloids and then facilitate destabilization of the colloids, which are referred to as adsorption-charge neutralization. In addition, it has been proposed that the colloids can be coated with Al(OH)3 precipitates and then the destabilization of
the colloids occurs by adsorption-charge neutralization since a local precipitation of positively charged Al(OH)3 on the colloids surface (Letterman et al., 1982). Dentel
(1988) also has proposed that different kinds of coagulants cause different coagulation mechanisms with precipitation charge neutralization (PCN) to explain that charge neutralization can be achieved by partial coverage of positively charged hydroxide on negatively charged particle surface. The adsorption-charge neutralization mechanism is characterized by stoichiometry between the coagulant and the concentration of colloids. Destabilization by charge neutralization generally occurs
at low concentrations of polymer where the repulsive electrostatic forces between the particles (Gregory, 1987). Overdosing of the coagulant can eventually reverse the surface charge on colloids, and a subsequent charge reversal of colloids surface and the restabilization of colloids possibly occur. Gregory (1996) has also demonstrated that some regional charge heterogeneity can occur that leads to the arrangement of the polymers in patches on the particle surface, which is referred to as electrostatic patch, as illustrated in Fig. 2.6. This mechanism is generally observed when particles
having a low charge are neutralized by the polymers with high charge densities. Electrostatic patch also could be caused by aluminum hydroxide precipitate for PACl coagulation in the presence of anion (Wang et al., 2002). Likewise, other study has suggested that the PACl containing high Al13 contents can bring electrostatic patch
coagulation at alkaline pH to destabilize kaolin particles when Al13 aggregates forms
(Ye et al., 2007).
Fig. 2.6 Electrostatic patch model of the interaction between negatively charged
(3) Enmeshment (Sweep flocculation)
Metal ions of high charge valence, e.g., Al3+, hydrolyze to form insoluble metal hydroxide precipitates after the addition of large amounts of aluminum or iron salts. The initial aluminum hydroxide precipitates are amorphous. Over time, the solid is concerted into a more stable crystalline phase. α-Al(OH)3 and -Al(OH)3 are two
ordinary forms. Enmeshment, or sweep flocculation, refers to physical mechanism where the colloids are entrapped in the aluminum hydroxide precipitates and the fine colloidal particles are swept together by sedimentation. The enmeshment is slower than other coagulation mechanisms because the Al(OH)3(s) require a few seconds to
form (Letterman et al., 1973). Coagulation of alum salts undergoes enmeshment mostly at around neutral pH where the particle can be then aggregated as sweep flocs by both surface precipitation and adsorption of amorphous hydroxide precipitate on the surface of particles (Chowdhury and Amy, 1991). Optimal sweep flocculation is identified in the pH range of 6 to 8 or higher depending on alum dosage. When pH value is less than the isoelectric point of Al(OH)3(am) (pH 8) and the alum
concentration is less than the dosage for sweep flocculation, the combination of sweep and adsorption occurs and the smaller positively charged aluminum hydroxide precipitates are adsorbed onto the larger negatively charged particles, which is called heterocoagulation (Amirtharajah and Mills, 1982). However, enmeshment or sweep flocculation also could occur for PACl coagulation of water when the PACl containing a considerable monomeric species that can hydrolyze into aluminum hydroxide precipitates. In enmeshment, the solid precipitates also increase the contact with colloidal particles and speed up the flocculation kinetics. This process is useful to remove particles from low turbidity water.
(4) Interparticle Bridging
Because of the stong adsorption of polymer toward colloidal particles, interparticle bridging occurs when the loops and tails of a polymer adsorbed to one colloidal particle attaches to one or more other colloidal particles. These processes are dictated by polymers which attach to several points of the colloidal sburface and are large enough to extend free segments (i.e., loops and tails) excess the zone of electrostatic double layers so that they are capable of binding to each other (Gregory, 1987). The surface coverage of adsorbed polymer plays a major role in controlling the probability of bridging. Flocculation can occur as polymers are at equilibrium with the surface of colloidal particles (Pelssers et al. 1989), while non-equilibrium flocculation occurs before the polymers are able to completely collapse on the colloidal surface (Gregory, 1988).
The dosage of polymers also affects the conformation and coverage surface of the polymer on colloidal surface, which can influence the probability of occurring bridging flocculation (Zhou and Franks, 2006). Moreover, the time that polymers remain before equilibrium is predominant over flocculation. The time required to adsorb the polymers on the colloidal surface will be influenced by the particle/polymer ratio, the size of the particle, the surface area of the particle, the adsorption energy of the polymer segments and the collision frequency among the particles (Gregory, 1988; Biggs et al., 2000), which substantially affects the flocculation efficiency. For charged polymers, the ionic strength plays an important part in bridging flocculation. With increasing the ionic strength, the polymer is deformed, decreasing the rigidity of polymer, resulting in worsened flocculation. However, the high ionic strength will also reduce the interparticle repulsion, allowing particles easily approach to each other. Therefore, optimal interparticle bridging
induced by polymers generally occurs at intermediate charge densities or salt concentrations (Matsumoto and Adachi, 1998).
On the other hand, study has advocated that the PACl containing large amounts of Al13 or giant polymeric aluminum species (Al30) may cause interparicle bridging
flocculation at high pH because both Al13 and Al30 can further hydrolyze and
restructure rapidly to form huger polymers when pH rises (Chen et al., 2006). However, it is difficult to directly verify the occurrence of interparticle bridging induced by Al13 or Al30 aggregates.
2.3 Coagulation Dynamics
During coagulation, the destabilized particles collide with each other to form clusters and continue to grow as flocs when they encounter other particles or clusters. For efficient coagulation, it is important to understand the process of particles movement during aggregation (i.e., dynamics of particles aggregation). Since particles aggregation is a kinetic and non-equilibrium growth process, the formation of clusters or flocs is through random collisions, and the structure of flocs is inherently related to the dynamics of aggregation, which in turn, is reflected by the time evolution of the cluster mass distribution. On the other hand, the breakage of flocs can be induced by a given shear force and the breakage behavior is affected by ambient physical and chemical conditions (Clark and Flora, 1991; Oles, 1992; Gregory and Dupont, 2001). As flocs grow larger, they increasingly become porous due to random collisions and collision efficiency comes lower (Brakalov, 1987). Once aggregate approach the length scales of turbulent eddies, hydrodynamic stress induced by mixing can fragment fragile flocs. However, the flocs strength is directly related to flocs structure and the inter-particle bonds between the components of flocs (Bache et al., 1997). Fractal dimension, which represents the scale-invariant branched structure formed as symmetric dilation (Mandelbrot, 1982), can be utilized to quantify the structure of flocs in coagulation process (Aubert and Cannell, 1986; Meakin, 1988; Lin et al., 1989). Intrinsically, the flocs with a fractal structure will form when the equilibrium between aggregation and breakage of flocs approach, which determines their strength during shearing. In this section, the studies about the dynamics of aggregation, the breakage of flocs and restructuring will be extensively reviewed.
2.3.1 Floc Formation
Witten and Sander (1981) have introduced the growth of flocs is dictated entirely by diffusion controlled aggregation model (i.e., diffusion-limited aggregation, DLA) in particle aggregation. In DLA, there is no repulsion between colloidal particles, so that particles attach permanently to other particles at the first contact and then grow to clusters or flocs via rapid random-walk trajectories (i.e., ‘rapid’ perikinetic aggregation) originating from outside of region occupied by the cluster. In this model, the sticking probability is equal to one and every collision between particles is effective. The fractal structures of flocs formed in this way are random ramified and insensitive to the sticking probability. Another model is the reaction-limited aggregation (RLA), for which a considerable number of collision (i.e., ‘slow’ orthokinetic aggregation) are needed between particles before sticking (Meakin and Witten, 1983). Obviously, in this model the sticking probability is small than one and only a small fraction of particles collisions lead to a permanent contact. This process produces more compact clusters or flocs whose density is irrelevant to distance. During coagulation, DLA and RLA will dominate over the fractal growth of particles, which results in different flocs structure (Elimelech et al., 1995). RLA produces a much narrower aggregate size distribution than DLA (Kostansek, 2004).
For difference in size and structure of flocs, particle-cluster and cluster-cluster aggregation provide a good explanation, illustrated in Fig. 2.7. In particle-cluster
aggregation, a particle is able to penetrate into a cluster before encountering another particle and sticking, which leads to a more compact structure, as seen in Fig. 2.7 (a).
As shown in Fig. 2.7 (b), tow cluster could collide and then stick at the first contact
before the clusters have interpenetrated to a significant extent, resulting in much more open structure. Torres et al. (1991) have found that flocs formed by cluster-cluster
aggregation are very similar to those formed by DLA. At RLA, particles need to collide many times before sticking occurs due to low collision efficiency, which increases opportunities for particle interpenetration. For this reason, flocs formed by RLA are more compact than those formed by DLA (Lin et al., 1989).
Moreover, flocs structure may change during coagulation, which gives more compact forms. Aubert and Cnnell (1986) have found that silica particle aggregation induced by DLA initially gives loose flocs with low fractal dimension, but denser flocs with higher fractal dimension are observed after a period of time. It is likely that shear induced by mixing can cause some deformation and rearrangement of flocs, leading to some compact flocs. On the other hand, it has been generally acknowledged that the turbulent energy dissipation with increasing floc size as a result of increasing porosity (Tambo and Watanabe, 1979; Kusters, 1991), and the porosity within the flocs could be depicted by use of fractal concepts in shearing (Sonntag and Russell, 1986).
(a) (b )
Fig. 2.7 Fractal flocs formed by (a) particle-cluster aggregation (b) cluster-cluster
aggregation.
2.3.2 Floc Breakage and Restructuring
Because of the inherent complexity, fragility of flocs and variation in floc size, shape and composition, the breakage behavior of flocs is not easily identified. It has been generally accepted that two model of flocs rupture induced by shear stress, including splitting and erosion (Francois, 1987; Mikkelsen and Keiding, 2002). These stresses have been manifested as splitting (i.e., large-scale fragmentation) and surface erosion, as demonstrated in Fig. 2.8.
Splitting is due to the instantaneous velocity differences across the body of the flocs, which produce several flocs fragments of a size similar to the parent flocs (Thomas, 1965). In addition, since fluid drag forces can strip primary particles or small clusters from the surface of flocs, called surface erosion, leading to an increase in the small particle size ranges. Glasgow and Luecke (1980) have observed experimentally that splitting is the dominant mechanism for flocs fragmentation. However, Williams et al. (1992) have suggested that more compact flocs structures were more likely to suffer surface erosion, while more open flocs would fracture by splitting. On the other hand, Potanin (1993) has modeled the shear-induced fragmentation of fractal flocs and advocated a combination of soft and rigid characteristics of actual flocs. From it, two or more fragments with mean daughter floc size are observed. These fragments are denser than parent flocs and there is no direct relationship between parent and daughter structure. During coagulation, irregular flocs with a fractal structure suffered any shearing is
probably to produce packed structure when particle-particle links shift to location with higher effective coordination numbers. This can not only occur in fragmentation, but also a change in fractal structure can often happen in the period of coagulation. Jiang and Logan (1996) have proposed that the fractal dimension of
flocs can increase with increasing floc size in coagulation. In other case, however, the fractal dimension of flocs can decrease over time in the initial stages of flocs formation for alum coagulation of latex mircrospheres (Chakraborti et al., 2003). The discrepancy can be attributed to restructuring of flocs that is the most prevalent compaction mechanism as a steady state is approached between aggregation and fragmentation during coagulation (Spicer and Pratsinis, 1996).
Tensile stress
S h ea r stress
(b) Surface Erosion (a) Sp l it t ing
Fig. 2.8 Two proposed mechanisms for the breakage of aggregates at different shear
conditions: (a) Splitting (b) Surface Erosion. Redrawn from Jarvis et al. (2005)
