Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Surface-chemical interactions between apatite and hematite in aqueous suspensions Ghaffari Touran, Naeimeh 2020

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2020_may_ghaffari_touran_naeimeh.pdf [ 7.64MB ]
Metadata
JSON: 24-1.0388358.json
JSON-LD: 24-1.0388358-ld.json
RDF/XML (Pretty): 24-1.0388358-rdf.xml
RDF/JSON: 24-1.0388358-rdf.json
Turtle: 24-1.0388358-turtle.txt
N-Triples: 24-1.0388358-rdf-ntriples.txt
Original Record: 24-1.0388358-source.json
Full Text
24-1.0388358-fulltext.txt
Citation
24-1.0388358.ris

Full Text

 SURFACE-CHEMICAL INTERACTIONS BETWEEN APATITE AND HEMATITE IN AQUEOUS SUSPENSIONS   by   Naeimeh Ghaffari Touran B.Sc., University of Tehran, Iran, 2007 M.Sc., University of Tehran, Iran, 2009    A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT  OF THE REQUIREMENTS FOR THE DEGREE OF   DOCTOR OF PHILOSOPHY   in    THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (MINING ENGINEERING)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)    January 2020   © Naeimeh Ghaffari Touran, 2020     ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Surface-chemical interactions between apatite and hematite in aqueous suspensions  submitted by Naeimeh Ghaffari Touran in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Mining Engineering  Examining Committee: Dr. Marek Pawlik Supervisor  Dr. Maria Holuszko Supervisory Committee Member  Dr. Wenying Liu Supervisory Committee Member Dr. Bern Klein University Examiner Dr. Edouard Asselin University Examiner    iii Abstract In relation to separation between apatite and hematite by froth flotation, surface-chemical interactions between apatite and hematite were studied as a function of pH and constituent ion concentrations. The surface charge characteristics of the minerals and their mixtures were determined using zeta potential measurements. Aggregation and dispersion phenomena were followed using laser light scattering techniques. Measurements of the amount of fine hematite attached to a large apatite crystal were conducted to assess the extent of slime coatings in the system. The results were supplemented by apatite solubility studies, and by measurements of calcium and phosphate adsorption on hematite under selected conditions. Calcium and phosphate released by apatite into supernatant strongly affected the zeta potential and aggregation-dispersion of hematite. The various phosphate species were attracted towards the positively charged hematite surface below the iso-electric point of the mineral (pH 6.8) while calcium cation showed high affinity towards the hematite surfaces above the iep value. Extensive coating of apatite by fine hematite occurred in the pH range from 7 to 9, and no coatings formed at pH 10-11 in the absence of calcium and phosphate ions. As the constituent ion concentrations increased in background solution, the amount of fine hematite on the crystal surface became independent of pH. These results were explained by a dispersing effect of phosphate at lower pH and by a coagulating effect of calcium at high pH. Co-adsorption of phosphate and calcium ions on the hematite surface was also observed, and it was proposed that calcium cations at pH 10-11 adsorbed in the inner Helmholtz plane while phosphate co-adsorption proceeded into the outer Helmholtz plane. Calcium adsorption caused hematite aggregation, while co-adsorption of phosphate led to partial dispersion.  In apatite-hematite mixtures, addition of apatite resulted in hematite dispersion, while removal of apatite caused hematite aggregation. It was recognized that the tested minerals systems were under non-equilibrium conditions in terms of apatite dissolution. The aggregation-dispersion phenomena were interpreted in terms of variations in calcium and phosphate ion concentrations in solution in the presence and absence of apatite, and the resulting changes in the adsorption of those ions on hematite.   iv Lay Summary The presence of iron oxide minerals, basically hematite, as impurities in some phosphate ores, represents one of the difficulties facing the phosphate beneficiation by flotation. The current work focuses on the assessing stability of minerals, as one of the main factors affecting the flotation separation performance, in mixed systems. The data indicated that constituent ions released by apatite, calcium and phosphate, co-adsorb on hematite, and depending on their concentrations, coagulation or dispersion of hematite occurs. In alkaline solutions, phosphate had a dispersing effect on hematite while calcium produced a coagulating effect. The hematite dispersion behaviour is greatly affected by apatite dissolution kinetics. As concentration of ions increases with time due to dissolution of apatite, hematite stability changes due to the adsorption of calcium and phosphate. Implications of these fundamental findings to controlling slime coatings in the hematite-apatite flotation system is discussed.   v Preface All the experiments were performed by the author, N. Ghaffari Touran, with the exception of mineralogical (XRD) and chemical (ICP) analyses that were performed by the Earth, Ocean and Atmospheric Sciences Department of the University of British Columbia as a commercial service. The selection of the techniques and procedures and the analysis of the research data were made by the author under the mentorship of the academic supervisor.                  vi Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface ............................................................................................................................................ v Table of Contents ......................................................................................................................... vi List of Tables ................................................................................................................................. x List of Figures ............................................................................................................................... xi Nomenclature ............................................................................................................................ xvii List of Abbreviations ............................................................................................................... xviii Acknowledgements .................................................................................................................... xix Dedication .................................................................................................................................... xx Chapter 1: Introduction ........................................................................................................... 1 1.1 Dispersion and aggregation phenomena ..................................................................... 1 1.2 Scope of the dissertation ............................................................................................. 2 1.3 Research objectives ..................................................................................................... 3 Chapter 2: Literature Review .................................................................................................. 4 2.1 Flotation separation of hematite and apatite ............................................................... 5 2.2 Aggregation/dispersion in apatite-hematite system .................................................. 11 2.3 Solubility of apatite ................................................................................................... 13 2.4 Surface charge at the mineral-solution interface ...................................................... 15   vii 2.4.1 Surface charge ................................................................................................... 15 2.4.2 Origin of charges on salt-type minerals ............................................................ 16 2.4.3 Origin of charges on oxide minerals ................................................................. 17 2.4.4 Development of electrical double layer ............................................................ 18 2.4.5 The electrokinetic (zeta) potential .................................................................... 22 2.4.6 Point of zero charge and iso-electric point ....................................................... 23 2.4.7 Electrokinetic phenomena ................................................................................. 24 2.4.8 Zeta potential distribution measurements on mineral mixtures ........................ 25 2.5 Stability of mineral suspensions ............................................................................... 27 2.5.1 Forces involved in aggregation/dispersion ....................................................... 27 2.5.2 Structural forces ................................................................................................ 30 2.5.3 Steric effect ....................................................................................................... 31 2.5.4 Slime coating .................................................................................................... 32 2.5.5 Influence of the ionic species on aggregation/dispersion ................................. 32 Chapter 3: Experimental Program ....................................................................................... 41 3.1 Reagents and materials ............................................................................................. 41 3.1.1 Model minerals ................................................................................................. 41 3.1.2 Chemical reagents ............................................................................................. 43 3.2 Experimental methodology, procedures and equipment ........................................... 44 3.2.1 Dissolution experiments .................................................................................... 44 3.2.2 Stability measurements ..................................................................................... 45 3.2.3 Zeta potential measurements ............................................................................. 54   viii 3.2.4 Direct measurements of formation of slime coatings on a single apatite crystal….. .......................................................................................................................... 59 3.2.5 Atomic absorption spectroscopy (AAS) ........................................................... 62 Chapter 4: Results and Discussion ........................................................................................ 64 4.1 Apatite solubility ....................................................................................................... 64 4.2 Hematite suspensions ................................................................................................ 67 4.2.1 Correlation between zeta potential of hematite and its dispersion behaviour in background electrolyte ...................................................................................................... 67 4.2.2 The effect of apatite supernatant on zeta potential and dispersion behaviour of hematite… ......................................................................................................................... 69 4.2.3 The effect of calcium and phosphate on surface properties of hematite in alkaline solution…. ......................................................................................................................... 79 4.2.4 The relationship between zeta potential and stability of hematite .................... 90 4.3 Apatite suspensions ................................................................................................... 93 4.3.1 Correlation between zeta potential of apatite and its dispersion behaviour ...... 93 4.3.2 The effect of hematite and apatite supernatant on zeta potential and dispersion behaviour of apatite ........................................................................................................... 95 4.4 Interactions between fine hematite and fine apatite particles ................................... 98 4.4.1 Stability measurements ..................................................................................... 98 4.4.2 Zeta potential distribution measurements ....................................................... 103 4.5 Slime coating in apatite/hematite system ................................................................ 107 4.5.1 Direct measurement of fine hematite coatings on single apatite crystals ....... 107   ix 4.5.2 Interactions between coarse apatite and fine hematite - Turbidity measurements… .............................................................................................................. 116 4.5.3 Interactions between coarse apatite and fine hematite - Transmission measurements .................................................................................................................. 119 4.6 The effect of apatite on stability of hematite in alkaline solutions ......................... 122 4.6.1 The effect of apatite to hematite ratio on stability and electrokinetic properties of hematite… ....................................................................................................................... 122 4.6.2 The composition of solutions in the presence and absence of apatite and hematite… ....................................................................................................................... 124 4.6.3 The reversibility of dispersing effect of apatite on hematite .......................... 128 4.6.4 The effect of apatite-hematite mixing time on the stability of hematite ......... 131 Chapter 5: Summary and Conclusions ............................................................................... 134 Chapter 6: Recommendations for Future Work ................................................................ 137 Bibliography .............................................................................................................................. 138 Appendices ................................................................................................................................. 153 Appendix A Determination of the composition of samples used for zeta potential measurements based on AAS results .............................................................................. 153 Appendix B Calibration curves for calcium and iron concentrations using AAS .......... 155  	   x List of Tables Table 3.1 Results of quantitative phase analysis (wt.%). ............................................................. 42 Table 3.2 Turbidity of coarse apatite (– 300 +150 µm) suspension after different settling time. 51 Table 3.3 Effect of calcium and phosphate ions on determination of iron concentration with AAS in a 5.0 × 10–3 g/L standard iron solution. ..................................................................................... 57 Table 4.1 Turbidity of solutions with different concentrations of calcium and phosphate at pH 10........................................................................................................................................................ 84 Table 4.2 Zeta potential of calcium phosphate precipitates at pH 10 prepared by mixing different concentrations of calcium and phosphate. .................................................................................... 84 Table 4.3 The amount of hematite and apatite (%vol) in the samples used for zeta potential distribution measurement experiments at pH 7 and pH 10. Volume of hematite was obtained from iron concentrations using stoichiometric coefficients and the density of hematite (5300 g/L). Similarly, volume of apatite was obtained from calcium concentrations using stoichiometric coefficients and the density of apatite (3200 g/L). ..................................................................... 104 Table 4.4 The composition of apatite supernatant conditioned in the presence and absence of fine hematite and coarse apatite at pH 10. ......................................................................................... 125  Table A. 1 Measurement of the volume of apatite (µl/L) in the tested mixtures used for zeta potential distribution measurements ........................................................................................... 153 Table A. 2 Measurement of the volume of hematite (µl/L) in the tested mixtures used for zeta potential distribution measurement ............................................................................................. 154 Table A. 3 Calculation of the amount of hematite and apatite (%vol) in the tested mixtures used for zeta potential distribution measurements .............................................................................. 154    xi List of Figures Figure 2.1 Schematic diagram of electrical double layer (above) and potential gradient across the EDL (below) (Zhu et al., 2017). ................................................................................................... 20 Figure 2.2 Schematic zeta potential distributions for a binary particulate component system that can be interpreted for particle interactions. ................................................................................... 26 Figure 2.3 Total potential energy of interaction between two particles according to DLVO theory (Bellmann, 2004). ......................................................................................................................... 29 Figure 3.1 One of the clean hexagonal apatite crystals tested in this thesis. ................................ 41 Figure 3.2 Probability and cumulative functions of the particle size distribution of apatite and hematite. ........................................................................................................................................ 43 Figure 3.3 Schematic of the Hach turbidimeter optical system (Pavanelli and Bigi, 2005). ........ 46 Figure 3.4 Schematic of the Turbiscan optical system (Buron et al., 2004). ................................ 47 Figure 3.5 Example of transmission and backscattering profiles obtained by Turbiscan and the change in clarified layer thickness with time. A schematic of the measurement cell is also shown........................................................................................................................................................ 48 Figure 3.6 Illustration of the calculation of clarification rate for hematite suspension in background electrolyte at different pH values. The slope of each line represents the clarification rate. ......... 49 Figure 3.7 Schematic of the method to assess aggregation of fine hematite with coarse apatite. 52 Figure 3.8 Graphical illustration of the definition of the width of the zeta potential distribution used in this study. Fine hematite in apatite supernatant at pH 10. ................................................ 58 Figure 3.9 Experimental procedure for direct measurement of hematite coating on single apatite crystal. ........................................................................................................................................... 59   xii Figure 3.10 Observed images of apatite crystal. A: clean crystal before measurement, B: crystal conditioned in hematite suspension (5µm) at pH 8, and C: crystal after applying ultrasound for 10 minutes. ......................................................................................................................................... 60 Figure 3.11 Experimental setup for direct imaging of slime coating phenomenon. ..................... 61 Figure 4.1 The conductivity of apatite suspensions prepared by mixing different amount of apatite in deionized water at natural pH (6.8) as a function of time. ........................................................ 65 Figure 4.2 Dissolution curve (the composition of apatite supernatant as a function of conditioning time) for 1 g/L apatite suspension in background electrolyte (0.01 mol/L NaCl) at natural pH (6.8)........................................................................................................................................................ 66 Figure 4.3 The concentration of calcium and phosphate after 60 minutes of mixing different amount of apatite in 0.01 mol/L NaCl solution at natural pH (6.8). ............................................. 67 Figure 4.4 Electrokinetic properties and clarification rate of hematite as a function of pH in background electrolyte. ................................................................................................................. 68 Figure 4.5 Zeta potential of hematite as a function of pH in background electrolyte and apatite supernatant. ................................................................................................................................... 69 Figure 4.6 Zeta potential of hematite in the background electrolyte as a function of pH, in the presence and absence of calcium and phosphate ions. The dotted line shows the hematite zeta potential in apatite supernatant. .................................................................................................... 70 Figure 4.7 Species distribution diagram for a total calcium concentration of 10–3 mol/L. ........... 71 Figure 4.8 Phosphate species distribution diagram as a function of pH (Liu et al., 2012). .......... 73 Figure 4.9 Surface properties of hematite (zeta potential and clarification rate) in apatite supernatant as a function of pH. ................................................................................................... 74   xiii Figure 4.10 Width of zeta potential distributions (90% of population) as a function of pH. The zeta potential distributions of hematite in apatite supernatant and background electrolyte (0.01 mol/L NaCl) are also shown in the graph. ............................................................................................... 77 Figure 4.11 Schematic of calcium ions bridging negatively charged colloids. ............................ 79 Figure 4.12 The effect of calcium on the zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte. The dotted curve shows the stability of hematite in apatite supernatant. ................................................................................................ 80 Figure 4.13 The effect of phosphate on zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte. .................................................... 82 Figure 4.14 The effect of calcium on zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte and in the presence of 2.0 × 10–5 mol/L phosphate. The dotted curve shows the stability of hematite in apatite supernatant. ................... 83 Figure 4.15 The stability (clarification rate) of hematite (0.07%vol solids) as a function of calcium concentration Triangle: in the absence of phosphate and Circle: in the presence of 2.0 × 10–5 mol/L phosphate. ..................................................................................................................................... 87 Figure 4.16 The effect of phosphate on zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte and in the presence of 2.0 × 10–5 mol/L calcium. The dotted curve shows the stability of hematite in apatite supernatant. ............ 88 Figure 4.17 Schematic of the hematite/water interface with adsorbed calcium ions in the inner Helmholtz plane and phosphate ions in the outer Helmholtz plane. ............................................. 90 Figure 4.18 The stability (clarification rate) of hematite (0.07%vol solids) as a function of zeta potential in Square: background electrolyte, Star: apatite supernatant, and in the presence of different concentrations of Circle: calcium, Cross: phosphate, Triangle: calcium at fixed amount   xiv of phosphate, Diamond: phosphate at fixed amount of calcium. A single trend line was drawn through the sets of data. ................................................................................................................ 91 Figure 4.19 The stability (clarification rate) of hematite (0.07%vol solids) as a function of zeta potential in background electrolyte, apatite supernatant, and in the presence of different concentrations of ions at high pH. ................................................................................................ 93 Figure 4.20 Surface properties of apatite as a function of pH in background electrolyte: Electrokinetic properties and clarification rate. ............................................................................ 94 Figure 4.21 Surface properties of apatite as a function of pH in background electrolyte and hematite supernatant: Electrokinetic properties and clarification rate. ZP and CR stands for zeta potential and clarification rate, respectively. ............................................................................................... 95 Figure 4.22 Surface properties of apatite as a function of pH in background electrolyte and apatite supernatant: Electrokinetic properties and clarification rate. ZP and CR stand for zeta potential and clarification rate, respectively. ...................................................................................................... 96 Figure 4.23 Zeta potential of apatite as a function of pH in CaCl2 and Na3PO4 solutions. .......... 98 Figure 4.24 Stability of apatite, hematite, and the 1:1 mixture (by volume) in apatite supernatant as a function of pH. ....................................................................................................................... 99 Figure 4.25 Stability (turbidity) of apatite, hematite, and the 1:1 mixture (total 0.2%vol) in apatite supernatant as a function of pH. ................................................................................................. 101 Figure 4.26 Zeta potential distributions of hematite, apatite, and hematite-apatite mixture in apatite supernatant prepared at solids content of 0.2%vol at pH 7. ........................................................ 103 Figure 4.27 Zeta potential distributions of hematite, apatite, and hematite-apatite mixture in apatite supernatant prepared at solids content of 0.2%vol at pH 10. ...................................................... 106   xv Figure 4.28 Hematite slime coating on apatite crystal (milligrams of hematite per gram of apatite crystal) in background electrolyte and apatite supernatant solutions. Supernatants A, B, and C were prepared respectively by mixing 1 g, 3 g, and 5 g of fine apatite in 1 L of 0.01 mol/L NaCl solution...................................................................................................................................................... 108 Figure 4.29 Images of different apatite crystals (1.2 × 0.6 cm) before and after conditioning with hematite in background electrolyte at different pH values. ........................................................ 109 Figure 4.30 Zeta potential of hematite in background electrolyte and in different supernatant solutions as a function of pH. Supernatant A, B, and C were prepared respectively by mixing 1 g, 3 g, and 5 g fine apatite in 1 L background electrolyte. .............................................................. 111 Figure 4.31 Images of apatite crystals after conditioning with hematite under different conditions. Left image:  at pH 9 and in background electrolyte. Right image: at pH 11 and in supernatant A...................................................................................................................................................... 112 Figure 4.32 Hematite slime coating on apatite crystal (milligrams of hematite per grams of apatite crystal) in background electrolyte, and in the presence of calcium and phosphate ions. ........... 114 Figure 4.33 Turbidity of hematite suspensions (0.07%vol solids) at pH 7, 10, and 11 for single and mixed systems in 0.01 mol/L NaCl as the background electrolyte (Test Group 1) and apatite supernatant (Test Group 2). H and A stands for hematite and apatite, respectively ................... 116 Figure 4.34 Stability of hematite suspension, transmission profiles, (0.07%vol solids) at pH 7, 10, and 11 for single and mixed systems in 0.01 mol/L NaCl as the background electrolyte. H and A stands for hematite and apatite, respectively .............................................................................. 120 Figure 4.35 Stability of hematite suspension, transmission profiles, (0.07%vol solids) at pH 7, 10, and 11 for single and mixed systems in apatite supernatant (prepared in 0.01 mol/L NaCl). The   xvi results for stability of hematite in background electrolyte (0.01 mol/L NaCl) are also shown in the graphs. H and A stands for hematite and apatite, respectively. .................................................. 121 Figure 4.36 Stability of hematite, transmission profiles, (0.07%vol solids) conditioned in the presence of different amount of coarse apatite (–300 +150 µm) at pH 10. All tests were conducted with apatite supernatant. ............................................................................................................. 123 Figure 4.37 Stability (transmission profiles) and zeta potential of hematite (0.07%vol solids) conditioned in the presence and absence of 0.9 g coarse apatite (–300 +150 µm) at pH 10. H and A stands for hematite and apatite, respectively. ......................................................................... 129 Figure 4.38 The stability (clarification rate) of hematite (0.07%vol solids) as a function of zeta potential at pH 10 in the presence of Triangle: calcium at fixed amount of phosphate, Diamond: phosphate at fixed amount of calcium, Star: coarse apatite, and as a result of adding and removing Circle: 1.5 g coarse apatite. ......................................................................................................... 131 Figure 4.39 Stability (transmission profiles) of hematite (0.07%vol solids) as a function of time in the presence and absence of 0.9 g coarse apatite (–300 +150 µm) at pH 10. H and A stands for hematite and apatite, respectively. .............................................................................................. 132 Figure B. 1 Absorbance versus concentration for calcium and iron at the wavelength of respectively 422.67 nm and 248.33 nm…………….. ................................................................ 155    xvii Nomenclature  𝐶–:  :  Concentrations of negative ions 𝐶$  :  Concentrations of positive ions 𝐶%  :  Bulk concentration of ions in solution 𝑉'  :  Electrophoretic mobility of particles 𝑉( 	 :  Hydrophobic interaction Ψ+   :  Stern potential A  :  Absorbance measured with atomic absorption spectrometer 𝐴  :  Hamaker constant a(𝝀)  :  Wavelength-dependent absorptivity coefficient b  :  Path length e :  Electronic charge 𝑓(k𝑎)  :  Henry equation I  :  Ionic strength k   :  Boltzmann constant K  :  Equilibrium constant KSP  :  Solubility product NA  :  Avogadro constant T  :  Absolute temperature VA  :	 Attractive interaction energy	Vhydration 	 :  Hydration interaction VR  :	 Electrical repulsive interaction energy	VT 	 :	 Total potential energy of interaction	z  :  Ion valency 𝜀  :  Permittivity 𝜁	  :  Zeta potential 𝜂  :  Viscosity 𝜅  :  Debye-Hückel parameter 𝝀  :  Wavelength ρ  :  Net volume charge density 𝛹	  :  Electrostatic potential   xviii List of Abbreviations AAS : Atomic Absorption Spectroscopy/Spectrometer BET : Brunauer- Emmett-Teller model (Quantachrome Autosorb analyzer) CCC : Critical Coagulation Concentration CMC : Carboxymethyl Cellulose CR : Point of Charge Reversal EDL : Electrical Double Layer ESA : Electrokinetic Sonic Amplitude HAP : Hydroxyapatite ICP : Inductively Coupled Plasma (Spectroscopy/ Spectrometer) iep : Iso-electric Point IHP : Inner Helmholtz Plane NTU : Nephelometric Turbidity Unit (Turbidimeter) OHP : Outer Helmholtz Plane PAM : Polyacrylamide–Acrylate Flocculant PZC : Point of Zero Charge UVP : Ultrasonic Vibration Potential XPS : X-ray Photoelectron Spectroscopic    xix Acknowledgements First and foremost, I would like to express my sincere gratitude to Dr. Marek Pawlik for his immense knowledge and excellent guidance which helped me in all the time of research and greatly contributed to the successful completion of this program. I am also deeply grateful to my supervisory committee members, Dr. Maria Holuszko and Dr. Wenying Liu, for their insightful comments and feedback that were helpful in completing this work. My sincere thanks should also go to Ms. Sally Finora for always being so kind and helpful. I would also like to acknowledge Mr. Aaron Hope and Mr. Libin Tong for their support in the coal and mineral processing laboratory. A warm word for my friends in the surface chemistry laboratory, Dr. Jophat Engwayu, Ryan MacIver, Avery Payne, Jose Moreno, and Haixing (Frank) Yan that created a cordial working environment. Finally, I would like to acknowledge the people who mean world to me. I thank with love my husband, Mohsen, who has always been there for me. Thank you for your continuous love, help and support. My deep and sincere gratitude to my mom, dad, and my siblings. This journey would not have been possible without your support.   xx Dedication To my beloved mom, dad, sisters, brothers, and to my love, Mohsen.   1 Chapter 1: Introduction Apatite and hematite are examples of two major groups of minerals, i.e., sparingly soluble minerals and metal oxides, respectively. They may occur together in phosphate and iron oxide ores. The presence of hematite or other non-magnetic iron oxide minerals in phosphate ores can cause separation issues with obtaining a low iron apatite concentrate. Therefore, the separation of apatite and hematite is necessary. The separation of apatite and iron oxides is also required in the phosphate-containing hematite ores. The phosphorus in iron ore deposits may occur primarily as apatite, or in the form of weathered apatite (secondary apatite), or in the crystal structures of hydrated iron oxides, (Barbour, 1973; Ler and Stanforth, 2003). Regardless of whether the ore is a high-iron phosphate type or high phosphate iron type, separation of apatite from hematite is carried out in a very similar way. The separation is carried out by anionic flotation with fatty acids collectors to float apatite while iron oxides are depressed using starch, starch derivatives, and other types of polysaccharides (e.g., guar gum).    However, in practice, the flotation performance is not always satisfactory for separation of apatite and iron oxides. Most research on the subject has been focused on examining the effect of different reagents and operating variables on the flotation separation of apatite and iron oxides in order to enhance the selectivity of flotation. Optimization of the reagents and experimental conditions is not the only way to get better performance in ore flotation. To be the most selective, investigating the mechanisms of misplacement of minerals and finding the best condition under which selectivity could be achieved is required. There are different factors which can cause misplacement of minerals in the flotation process, including inadequate liberation of minerals, entrainment, interference from dissolved species, and aggregation phenomena between different ore components.  1.1 Dispersion and aggregation phenomena  The dispersion/aggregation state of particles in suspension mainly depends on the particle-particle interactions. When aggregation takes place an unstable colloidal system results, whereas dispersion occurs when fine mineral particles in suspension are prevented from aggregating and stay freely dispersed. The colloidal stability of mineral particles against aggregation plays an   2 important role in the flotation separation of mixed mineral suspensions. There are generally two forms of aggregation: if particles of the same mineral type aggregate, aggregation is referred to as homo-aggregation, while it is referred to as hetero-aggregation if two different minerals aggregate. Even in a well liberated mineral system, aggregation among particles of different minerals inhibits the selectivity of flotation separation. Also, entrapment and therefore recovery of unwanted mineral in aggregates of another floating mineral may occur. The most troublesome manifestation of hetero-aggregation in froth flotation is the slime coating phenomenon, the coating of coarse valuable particles of one mineral by fines of another gangue mineral, which can detrimentally affect separation of multi-component ores (Mitchell et al., 2005; Parsonage, 1985; Peng and Zhou, 2011). In mineral suspensions containing fine particles, depending on the degree of dispersion of the minerals, higher or lower coverage of coarser particles with fines may take place depending on process water chemistry, particle size distributions of the involved minerals, and their surface charging characteristics. The consequence of slime coating is the loss of separation efficiency. Highly dispersed suspensions, on the other hand, may also decrease separation selectivity, as the probability of recovery of fine particles due to entrainment increases. Controlled dispersion was shown to be an important factor influencing the efficiency of flotation separation (Fuerstenau, 1980) and other separation methods such as shear flocculation  (Hu and Yu, 1988).  1.2 Scope of the dissertation  Understanding the interaction between valuable and gangue minerals is of both fundamental and practical importance in the field of flotation. Searching through literature indicates that systematic studies about interparticle interactions in the apatite-hematite system are rare. Therefore, this research focused on investigating the interactions in a mixed mineral system of apatite and hematite under different physicochemical conditions. Apatite as a sparingly-soluble salt type mineral releases constituent ions into solution; primarily calcium and phosphate when contacted with water, whereas hematite dissolution under the usual flotation conditions (pH, temperature) proceeds so slowly that it can be considered as an insoluble mineral (Hunter, 1981). The study commenced with a detailed determination of the interfacial properties of single minerals in the presence and absence of calcium and phosphate. Subsequently, the aggregation/dispersion behaviour of the minerals in a mixture was examined. Since in phosphate flotation operations   3 apatite particles are usually coarse, fine particles of other minerals may cover the surface of apatite particles and hinder the selectivity of phosphate flotation. Therefore, slime coating (attachment of fine hematite to coarse apatite particles) was also investigated. 1.3 Research objectives The general objective of this research is to provide new insights to improving separation efficiency of phosphates and iron oxides by investigating aggregation/dispersion behaviour of the minerals. The following detailed objectives were set: 1) To understand the effect of dissolved ions released by apatite (calcium and phosphate ions), on the electrokinetic properties of apatite, hematite, and their mixtures. 2) To examine aggregation/dispersion behaviour of apatite and hematite in single and mixed minerals system and correlate their behavior with their electrokinetic properties.  3) To propose the governing mechanism of aggregation/dispersion of minerals (apatite and hematite). 4) To study the slime coating phenomenon and identify key variables which control particle-particle interactions leading to the formation of slime coatings.       4 Chapter 2: Literature Review Phosphate ores are major resources for the production of agricultural fertilizers and phosphorus-based chemicals. Phosphate deposits can be divided into three groups (Kawatra and Carlson, 2013). The most abundant phosphate deposits are the sedimentary phosphates, containing significant amounts of carbonate minerals (north of Africa and Florida). Considerable phosphate resources are also associated with igneous rocks in several countries (e.g. Brazil, Republic of South Africa, Sweden, and Russia). The third type of phosphate deposits are biogenetic deposits (Guimarães et al., 2005). Apatite, as the principle primary phosphate mineral, is a calcium phosphate mineral that includes hydroxyapatite (Ca5(PO4)3OH), fluorapatite (Ca5(PO4)3F), and chlorapatite (Ca5(PO4)3Cl). The most common gangue minerals associated with phosphate ores are silica, calcareous minerals (mainly calcite and dolomite), aluminosilicates (clays), and metal oxides (Zafar et al., 1995). The beneficiation techniques for phosphate ores depend mainly on the type of associated gangue minerals. Flotation is the predominant method for recovering phosphate minerals, producing more than half of the world’s commercial phosphate (Abouzeid, 2008; Sis and Chander, 2003) and is commonly carried out using anionic or cationic surfactants as collectors. Fatty acids or their salts are the most widely used collectors in the flotation of phosphate ores because of their availability and lower cost (Sis and Chander, 2003). However, other collectors have been tested and are available for the flotation of phosphate ores. Other concentration methods including calcinations, acid leaching, and magnetic separation, with certain drawbacks and limitations compared to flotation, are also used (Gu, 2002; Sis and Chander, 2003; Zafar et al., 1995). Comprehensive reviews on treatment of phosphate ores were given by Zhang, Albarelli, & Stewart  (1999) and Abouzeid (2008). The type of phosphate deposit affects the flotation performance. Sedimentary deposits can be upgraded successfully by flotation when the gangue is essentially siliceous. The Crago process is the dominant flotation process for the beneficiation of siliceous phosphates in Florida (Guan, 2009; Zhang et al., 1997). In this technique, phosphate is floated using fatty acids and fuel oil, leaving the majority of silica and silicate minerals as tailings. The phosphate concentrate is acid-scrubbed with sulfuric acid in order to remove the collector coating around phosphate particles. Then, the cationic flotation with amine is carried out to remove silica entrapped within the   5 phosphate concentrate. The ore types containing a high amount of carbonates, however, are extremely difficult to upgrade via flotation due to the similar physicochemical surfaces properties of the carbonates and phosphates (Abouzeid et al., 2009; Elgillani and Abouzeid, 1993). An extensive work at laboratory and pilot-plant scales have been carried out on carbonate–phosphate separation by flotation. The reverse anionic flotation, in acidic media, is the most commonly used technique for beneficiation of phosphate ores containing carbonate impurities (El-Midany et al., 2011; Elmahdy et al., 2009; Santana et al., 2008). The literature on the processing of sedimentary phosphate ores by flotation, and reagents used in the process, was reviewed by Sis and Chander (2003). Igneous phosphate deposits usually contain well-formed apatite crystals, which is beneficial to froth flotation because of their low porosity and therefore low specific surface areas (El-Shall et al., 2004). Igneous reserves of different types are rich in apatite but can vary significantly in gangue minerals. Depending on the types of gangue minerals, the beneficiation of igneous phosphate ores can range from a relatively simple method to complex processing methods. Iron oxides are among the commonly encountered gangue minerals in igneous phosphate deposits. Tall oil flotation of phosphate minerals after depressing iron oxides has been employed. A review of the reagents used in flotation of igneous phosphates can be found in Guimarães et al. (2005), and the key features of the process will be discussed in the next section. 2.1  Flotation separation of hematite and apatite Considerable research was conducted on the removal of iron minerals from phosphate concentrates (Assis et al., 1996; Dasgupta and Taluja, 1988; Gong et al., 1993, 1992; Moudgil and Somasundaran, 1986; Nanthakumar et al., 2009; Oliveira et al., 2011; Rubio et al., 2015; Wang and Heiskanen, 1990). The beneficiation of high-iron phosphate ores is carried out by depression of hematite followed by direct flotation of apatite with fatty acids or their salts (El-Shall et al., 2004).  The depressing action of a variety of depressants towards iron oxides was studied by different researchers. The fatty acid flotation of phosphate-iron oxide-silicate sample from Sokli, Finland was investigated through batch flotation tests in order to obtain low iron-containing apatite concentrate (Wang and Heiskanen, 1990). Different organic and inorganic depressants were   6 examined for their depressing action towards iron oxides and silicate minerals. High phosphate recovery and acceptable grades were achieved with corn starch. The grade of silica and iron was low, but their recoveries were still high. The depression action of carboxymethyl cellulose (CMC) towards silica and hematite was better than that of corn starch. However, the recovery of phosphate in the presence of CMC was lower than that obtained with starch. Tannic acid, and dextrin showed promising selective depressing action towards silicate and iron oxide minerals. Sodium silicate was the most selective reagent acting simultaneously as depressant and dispersant in fatty acid flotation of phosphates. The mechanism of selective action of sodium silicate was suggested to be due to the adsorption of polysilicate polymer onto the hydroxylated sites on the surface of silicates and iron oxides, leading to an increase in surface hydrophilicity and slime dispersion. However, the results showed that the iron content in the concentrate was still high and other factors rather than true flotation of iron oxides were, therefore, suggested to cause the high iron content in concentrate. Wang and Heiskanen (1990) did not investigate the possible origins of iron in concentrate. In another study (Gong et al., 1992), the mechanism of the selective depressing action of sodium silicate towards iron oxides was studied. It was found that the modulus of sodium silicate, pH and time of standing of the stock sodium silicate solutions, and the presence of polyvalent metal ions (CaCl2 and AlCl3), all played an important role in the depressing effect of sodium silicate. A comparison between the performance of sodium silicate and starch as depressants for hematite in the fatty acid flotation of the ore showed that sodium silicate was more promising as a selective depressant in the separation of apatite from hematite by oleate flotation. Apatite concentrate with high grade and recovery (81% P2O5 with 2.7% Fe2O3) was obtained from high-iron Mt. Weld phosphate ore (22% P2O5 and 38% Fe2O3) by selective depression of iron oxide with sodium silicate. Gong et al. (1992) also investigated the origin of iron oxides in concentrate. The presence of iron in the concentrate, in the form of hematite and goethite, was largely attributed to the insufficient depression of the liberated iron particles and the interlocking of some apatite particles with iron oxides. The mechanism of the depressing action of sodium silicate in flotation of apatite from hematite was also examined by Gong et al. (1993) through determining the nature of the silicate species in solution and on the surfaces of apatite and hematite. Also, the effect of the silicate species on the flotation separation of apatite and hematite as the gangue mineral was   7 measured. From X-ray photoelectron spectroscopic (XPS) analyses, the adsorption of polymerized silicates on hematite from silicate solutions with concentrations of 10–3 to 10–2 mol/L (close to concentrations used in flotation) was observed. The selective depression of hematite with sodium silicate observed in the earlier studies by Gong et al. (1992) was suggested to be due to formation of moderately polymerized silicates and/or colloidal silica particles of very small sizes in solution and subsequently their adsorption on the surface of minerals. It was suggested that the adsorption of these species on hematite would prevent oleate adsorption whereas the adsorbed silicate on apatite would be displaced by the strongly adsorbed oleate. Guar gum and corn starch were evaluated for their effectiveness as iron depressant for a difficult-to-float igneous phosphate ore (Nanthakumar et al., 2009). It was observed that the performance of guar gum was superior to that of corn starch, producing better selectivity and higher phosphate grade. Nanthakumar et al. (2009) conducted batch flotation in the presence of soda ash. The results showed that the fatty acid flotation of phosphate minerals was affected by the high concentration of multivalent cations, mainly calcium ions, present in process water. The phosphate flotation was markedly improved by addition of soda ash which precipitated the cations from the solution.  The problem of iron in phosphate flotation was also explored by determining the effect of different collectors and operating conditions on flotation of apatite from iron oxides. Assis et al. (1996) studied the bench scale flotation of two different phosphate ores, Arafertil and Fosfertii ores, with hydroxamates as collectors with the aim of increasing P2O5 content and reducing iron grade in the concentrates. The results indicated a preferential flotation of apatite with respect to iron bearing minerals, though the iron/hydroxamate complex is more stable than calcium hydroxamate. It was suggested that two factors, complexing ability and solubility of minerals, affect the hydroxamates adsorption and, therefore, the selectivity of flotation of minerals with hydroxamates. Oliveira et al. (2007) studied the selectivity of apatite flotation from a phosphate ore sample from the Barreiro carbonatite complex (Araxa ́, Minas Gerais state, Brazil) with the aim of finding the experimental condition, reagent (collector and depressant) dosages and air flow rate, for obtaining the highest recovery and P2O5 content in concentrate. Collector and depressant dosages of respectively 126 g/ton and 186 g/ton and air flow rate of 126 L/h resulted in the highest recovery, while the best condition for a high P2O5 content was found to be a collector dosage of 74g/ton, depressant dosage of 314 g/ton and air flow rate of 126 L/h. The possible recovery of   8 apatite from phosphate flotation tailings (from a Brazilian fertilizer manufacturer) containing 9.52% of P2O5 and high amount of Fe2O3 and SiO2 (26.20% of Fe2O3 and 22.69% of SiO2) was examined by Oliveira et al. (2011). The effect of reagent dosages and operating variables including air flow rate, recycle flow rate, and conditioning time on the flotation performance was assessed using a special flotation apparatus (an acrylic cylindrical flotation column). The effect of a mixture of two types of collectors, rice oil soap and a synthetic anionic collector (KE), with varying percentages of rice oil soap (ROS), on the P2O5 grade and recovery was measured. Increasing percentage of KE in the mixture was found to increase the P2O5 grade in the concentrate. The higher P2O5 grade by using a mixture of collectors, was attributed to the higher selectivity of KE towards calcium containing minerals. However, very high concentration of KE to ROS ratio decreased the P2O5 recovery. It was explained in accordance with the frothing properties of fatty acid collectors. The decrease in the concentration of ROS was suggested to result in the decrease in the foam layer and, therefore, recovery of minerals. A concentrate with a grade of 29.4 wt.% P2O5 and a recovery of 46.2 wt.% were obtained under selected operating conditions (500 g/t corn starch as depressant, 80 g/t collector consisting of 80 wt.% of rice oil soap and 20 wt.% of KE). The influence of different particle size distributions of the feed, pH, type and dosage of collectors, and depressant dosages was assessed for the beneficiation of apatite from a low-grade magnetite–apatite ore (Sung Valley, Shillong, Meghalaya State, India) (Dasgupta and Taluja, 1988). Wet magnetic separation followed by flotation was used in this study. The results demonstrated that using sodium oleate collector and sodium silicate depressant, a concentrate, assaying 37.18% P2O5 with 80% P2O5 recovery, suitable for the phosphate industry was obtained at pH of 7 to 7.5 (natural pH). The fatty acid flotation of a phosphate sample, containing high amount of Fe2O3 and SiO2, from the concentrate plant at Araxá (Vale Fertilizers-Southeast Brazil) was investigated using the modified three-product column (3PC) flotation cell, at laboratory and pilot plant scales (Rubio et al., 2015). The 3PC was shown to have a higher potential in rougher-flash or cleaner flotation circuits, than other flotation cells. High grades of P2O5 (39%) and lower contents of Fe2O3 and SiO2 in concentrate were obtained in the 3PC when compared with conventional column cell. However, the recovery of P2O5 in the 3PC was lower. The separation of apatite and iron oxides is also important in the high phosphorus iron oxide ores. In phosphorus-rich (high phosphorus) iron ores, phosphorus may exist in the form of   9 apatite, disseminated in iron oxide mineral structure and is locked with gangue minerals such as quartz and carbonate minerals (Barbour, 1973; Ler and Stanforth, 2003; Szekely and Poveromo, 1975). Basically, two different methods can be used for decreasing the phosphorus content of the iron ores based on the extent to which the phosphorus bearing minerals are liberated from iron minerals. The separation is performed using magnetic separation and flotation if phosphorus bearing minerals are completely liberated. The second approach, suitable for iron ores with larger particle size, is a wet chemical method in which the ore is leached with a suitable solution and the phosphorus is removed by the dissolution of the apatite (Muhammed and Zhang, 1989).Flotation method has been the subject of most of the studies on phosphorus removal from iron ores (Gaudin, 1957; Hanna and Anazia, 1983; Leite Nunes et al., 2012; Potapova et al., 2011; Ranjbar, 2002; Rao et al., 2010; Siirak and Hancock, 1988; Subramanian et al., 2002; Tohry and Dehghani, 2016). In the case of high-phosphate iron ores, the flotation of apatite from iron using fatty acid-based surfactants is performed while the depression of iron minerals is achieved using starch, sodium silicate, or guar gum. Tohry & Dehghani (2016) studied the reduction of the silica and phosphorus contents in iron concentrate of Chador Malu iron ore (Iran) by batch flotation tests. The main valuable minerals in the flotation feed sample were reported to be hematite and magnetite, and the main gangue minerals were apatite and quartz. Different types of depressants with varying dosages were tested and it was found that sodium silicate provided better results in terms of depressing iron oxides than other reagents. The effect of multivalent cations (Zn2+, Ca2+, Al3+, Mg2+), different dosages and modulus of sodium silicate, pulp pH, and conditioning time of sodium silicate solution on flotation separation of apatite and silica from iron oxides were examined. An effective separation of apatite from iron oxides was achieved by using 800 g/t of sodium silicate with a modulus of 2.5 at pH 10 and after conditioning/aging the sodium silicate for 10 min. However, the silica content in iron concentrate was still high. In the presence of multivalent cations, the separation selectivity of quartz and apatite from iron oxides was improved. Iron concentrate with acceptable grades (67% Fe, 1.5% SiO2, and 0.034% P) and recovery of 93.6% of Fe was obtained by using silicate–CaCl2.2H2O mixture with Me/Si ratio of 0.4 as a depressant.  Using in-situ ATR- FTIR spectroscopy, zeta-potential, and contact angle measurements, Potapova et al. (2011) investigated the consecutive adsorption of sodium silicate and an anionic   10 surfactant (maleic acid ester) on magnetite in the presence of calcium ions at different pH values. It was suggested that calcium ions could have a detrimental effect on dispersing action of sodium silicate in flotation of apatite from magnetite, as calcium ions were found to adsorb on the magnetite surface and decrease the negative surface charge of magnetite treated by sodium silicate. In the presence of calcium, the amount of sodium silicate adsorbed on magnetite did not have any notable effect on the adsorption of maleic acid ester over the pH range from 7.5 to 9.5, suggesting that they adsorbed on different surface sites: silicate on magnetite surface hydroxyl groups and collector on the calcium ions adsorbed on magnetite. It was found that the treatment of the magnetite sample with calcium and sodium silicate lowered the contact angle between water and magnetite, but subsequent adsorption of maleic acid ester increased the hydrophobicity of magnetite (contact angle of 40◦–50◦) to a degree detrimental to the strength of the pellets, but not to the reverse flotation of iron ores. The main difficulty encountered in the reverse flotation of iron oxide fines from Chogart, Gol-e-Gohar, Chador Malu  iron ore (Iran) was reported to be the reduction in concentrate phosphorus content, apatite, to an acceptable level of less than 0.045 wt% (Ranjbar, 2002). In order to find the experimental conditions for achieving a low-phosphorus concentrate, various process parameters such as reagent (collector and depressant) type and dosage, pH, temperature were optimized. The results indicated that a concentrate with a phosphorus content of 0.040 wt% and an iron oxide recovery of more than 82% in the concentrate could be achieved. In order to study the froth flotation potential for producing concentrates with a low phosphorus content, Leite Nunes et al. (2012) performed microflotation experiments of wavellite [Al3(PO4)2(OH)3 5H2O], the most common secondary phosphate present in iron ores, using a variety of anionic, cationic and amphoteric collectors over the pH range from 8 to 12, followed by bench flotation experiments on a high-phosphorus iron ore at pH 10.5 (Minas Gerais – Brazil). The results from microflotation tests showed that among collectors tested, dodecylamine produced the highest floatability of wavellite, which were around 100% above pH 8. The high efficiency of dodecylamine in floating wavellite was suggested to be due to its strong adsorption, both chemical and electrostatic, on the wavellite surface. According to the bench flotation results, anionic and amphoteric collectors were less effective than amine collectors (Flotigam EDA, Flotigam 2835 2L, Gemul NCD, Gemul NSD). Flotation with amine as a collector gave the lowest phosphorus content in the final iron   11 concentrate. The collector Flotigam 2835 2L was shown to be more selective towards phosphates and aluminum silicates than the other amine. An iron concentrate containing 0.312% phosphorus with a mass recovery of 90.24% of iron oxides was obtained with the Flotigam 2835 2L.  2.2 Aggregation/dispersion in apatite-hematite system  Adequate liberation and dispersion of the minerals from each other is one of the key factors affecting the separation of minerals by physical and physicochemical processes. Even in a well liberated mineral system, aggregation among different minerals can lead to low separation selectivity. The literature on the dispersion behaviour of minerals in apatite-hematite system is very limited. The main difficulty encountered in the flotation of phosphate ores from Western Australian Mt. Weld was reported to be the removal of iron from concentrate. The presence of iron in the concentrate can generally be attributed to several factors, including: insufficient depression of liberated iron oxide particles, the presence of composite apatite-iron mineral particles, entrainment of liberated iron oxide particles into the concentrate, and surface coating of apatite particles by iron oxides (Belton, 1985; Dyall, 1985; Gong et al., 1992; Mcpheat, 1984; Yernberg, 1985). However, in the flotation study on Western Australian Mt. Weld phosphate ore by Gong et al. (1992), the presence of iron in the concentrate, mainly in the form of hematite and goethite, was shown to be due to the inadequate depression of the liberated iron particles and to the interlocking of some apatite particles with iron oxides. In the study by Wang and Heiskanen (1990), the poor selectivity of flotation separation of a phosphate-iron oxide-silicate sample in the absence of a depressant was attributed to the iron oxide contamination on the apatite surface. Also, the presence of Ca2+ ions dissolved from apatite, caused the non-selective bulk precipitation of calcium fatty acid soap onto all the minerals, and therefore activating the flotation of silicates and iron oxides. It is important to assess the effect of the main ions in process water on the selectivity of the flotation process. The literature data on the concentrations of calcium ions in process water of phosphate ores vary considerably. Calcium levels in plant water from Florida phosphate operations were reported to be on the order of 36-70 ppm (Gruber et al., 1995), and could reach 1000 ppm of calcium if more soluble salts of calcium (e.g., gypsum) were present in phosphate ores (Nanthakumar et al., 2009).   12 The stability characteristics of apatite, hematite, phlogopite [KMg3AlSi3O10(F,OH)2], and the selective coagulation and/or selective dispersion between the minerals in binary mixtures, apatite-hematite and apatite-phlogopite, using sodium silicate and phosphate dispersants was determined by Wang & Heiskanen (1992a, 1992b) through particle size measurement, sedimentation tests, and electrophoretic measurements. In deionized water, minerals behaved differently in terms of dispersion; apatite coagulated strongly, hematite coagulated moderately, and phlogopite fully dispersed. The effects of the apatite dissolved species on the dispersion of hematite and phlogopite was also examined. A background solution prepared by conditioning apatite in water (supernatant) had a strong effect on the stability of hematite, reducing hematite dispersion over the pH range from 7 to 12 (Wang & Heiskanen, 1992a). However, the mechanism of destabilization and the role of ions released by apatite, calcium and phosphate, in destabilizing hematite was not clearly presented in those studies. No effect of apatite supernatant was observed on the phlogopite dispersion. The effects of different inorganic dispersants, sodium silicate, sodium polyphosphates, sodium metasilicate, sodium orthophosphate, and sodium hexametaphosphate, on the dispersion of the coagulated apatite and hematite samples were examined. Sodium silicate and sodium polyphosphates showed to be the most effective dispersants towards apatite and hematite, with hematite needing much smaller amounts of the dispersants for dispersion compared to apatite. The single minerals showed very different stabilities and dispersibilities in the presence of the dispersants (Wang & Heiskanen, 1992a).  The experiments on the binary mineral mixtures in the presence and absence of selective dispersants, sodium silicate and sodium pyrophosphate, demonstrated that coagulation/dispersion selectivities in mixture were much poorer than those predicted from individual mineral tests (Wang & Heiskanen, 1992b). It was observed that the presence of a more dispersed mineral, hematite or phlogopite, significantly enhanced apatite dispersion and, therefore, decreased the amount of dispersant needed for apatite dispersion. Also, the presence of apatite in a mixture, notably decreased the stability of hematite and phlogopite, even though these minerals were predicted to be stable based on tests on single minerals. Zeta potential distribution measurements on the individual and mixed minerals at around pH 9.5 showed the presence of a peak that could be attributed to hetero-aggregation (Wang & Heiskanen, 1992b). However, the composition of the tested mixture was not verified. Therefore, the assumption that the peak originated from aggregates   13 may not be accurate. The loss of the dispersion selectivities was also studied on the basis of colloid stability theories and was explained by the hetero-aggregation between minerals. It was suggested that the hetero-aggregation between ultrafine particles of different minerals and the adhesion of fine apatite particles onto coarser hematite or phlogopite particles would decrease the stability of more dispersed hematite or phlogopite. In contrast, the adhesion of hematite or phlogopite ultrafines to coarser apatite particles would change the effective surface potential of apatite and stabilize the apatite particles. The low dispersion selectivity of mineral mixtures was also suggested to arise from the interference of dissolved ions and the entrapment of more stable particles into the loose aggregates of the less stable mineral (Wang & Heiskanen, 1992a, 1992b).  In the study on the selective flocculation behaviour of apatite-hematite mixtures, (Wang and Heiskanen, 1992c), it was observed that in alkaline solutions, above pH 8, sodium oleate acted as a flocculant of apatite but a dispersant of hematite in single minerals suspensions, whereas the results for a 1:1 apatite-hematite mixture at pH 10.1-10.4 showed very low selectivity only with high concentrations of the flocculant. The non-selective flocculation at low concentrations of sodium oleate was suggested to be a result of the hetero-aggregation between apatite and hematite. The flocculation selectivity was enhanced by using sodium silicate and sodium pyrophosphate as effective dispersant towards minerals. The use of a dispersant was shown to be necessary for separation by selective flocculation due to hetero-aggregation between apatite and hematite. Using this method, dispersion followed by flocculation with sodium oleate, higher selectivity was only obtained from mixtures with a high apatite-to-hematite ratio (Wang and Heiskanen, 1992c). Although in these studies the aggregation phenomena were suggested to lead to the lower selectivity of the separation of apatite-hematite mixtures, direct measurements of aggregation between fine minerals and/or fine particles of one mineral and coarse particles of the other (slime coating) were not conducted. 2.3 Solubility of apatite Solubility properties of minerals are of great importance in flotation, because of their role in determining both the chemical composition of the aqueous phase and the surface characteristics of the mineral-solution interface. Salt-type minerals are characterized by solubilities that are lower than those of simple salt minerals, but higher than those of most oxides and silicate minerals. The   14 solubility of apatite, as a sparingly soluble mineral, and the time needed to reach equilibrium was researched in many studies (Chaïrat et al., 2007b; Clark, 1955; Fulmer et al., 2002; McDowell et al., 1977; Snoeyink and Jenkins, 1980; Valsami-Jones et al., 1998b). However, there are many disagreements between the findings of these investigations. Fluorapatite and hydroxyapatite dissolve according to the following reactions: Ca=	(PO@)AOH ⇋ 5CaE$ + 3PO@AH + OHH (2.1) Ca=	(PO@)AOH ⇋ 5CaE$ + 3PO@AH + FH (2.2) The dissolution process of salt-type minerals is reversible and the amount of mineral in solution in equilibrium with the solid is given by the solubility product (KSP). KSP is the product of the concentrations of the ions in moles per liter (mol/L). For example, the solubility product of hydroxyapatite can theoretically be calculated from the assayed ion concentration of Ca2+, PO43– and OH– (Equation 2.3): KKL	MNOPQRNSTSUVUW = [CaE$]=[PO@AH\A[OHH]	 (2.3) Fluorapatite is less soluble than hydroxyapatite, however literature values of the solubility products vary considerably for both minerals. Values for fluorapatite range between 10–70 (Valsami-Jones et al., 1998a) to 10–58.13 (Jaynes et al., 1999) while the reported solubility product values of hydroxyapatite were shown to vary even more: from 10− 88.5 (McDowell et al., 1977) to 10−55.9 (Snoeyink and Jenkins, 1980). The source of the minerals used in the various experiments can be expected to play a significant role in this discrepancy. The solubility of apatite was found to be dependent on the stoichiometry and porosity of the mineral surface. (Lin et al., 1981; Pugh and Stenius, 1985; Rao et al., 1990; Zhong et al., 1993). The slower diffusion of ions through the pores was suggested to result in lower dissolution rates of apatite (Zhong et al., 1993). The variation can also be attributed to the differences in experimental conditions and to the failure to reach the equilibrium condition even after long condition times (Fulmer et al., 2002; Valsami-Jones et al., 1998a). The apatite dissolution rate is strongly pH dependent. The concentrations of ions released by apatite in solutions markedly increase with increasing acidity (Chaïrat et al., 2007b, 2007a; Levinskas and Newman, 1955). The apatite dissolution was reported to be non-stoichiometric in the early stages of reaction. As dissolution proceeds, however, the dissolution process becomes congruent (Chaïrat et al., 2007a, 2007b; Guidry and Mackenzie, 2003).    15 2.4 Surface charge at the mineral-solution interface Particle surfaces acquire an electrical charge when exposed to water. Accordingly, a cloud of oppositely charged counter ions is developed around the surface in order to maintain the electrical neutrality. These oppositely charged regions on the surface and in the solution phase are referred to as the electrical double layer. The electrical properties of fine particles in aqueous solution play a significant role in adsorption of inorganic and organic species at the solid/solution interface. Also, surface charging characteristics directly affect the dispersion behaviour of minerals (Ersoy and Çelik, 2002; Fuerstenau and Pradip, 2005). 2.4.1 Surface charge In general, the interface can be charged through one or more of the following charge mechanisms (Cosgrove, 2005; Delgado, 2002; Kitchener, 1969). Preferential dissolution of surface ions The surface charge can result from the preferential dissolution of lattice ions from the surface of particles. The tendency of the lattice ions to transfer to solution can be determined by comparing their hydration energies with their lattice energies. The ions with the greater negative hydration energy have stronger tendency to hydrate and leave the surface. Accordingly, the proportion of cations and anions transferred into solution is not stoichiometric. Multivalent cations have a greater hydration energy than anions and therefore, in most cases, leave the surface and go to solution to a greater extent and make the surfaces negatively charged (Lu et al., 2005). Ionization of surface groups Minerals which have ionizable groups will be charged as a result of dissociation of those surface groups. These groups, such as hydroxyl (OH), depending on pH, dissociate to form positively or negatively charged sites on the surface of a mineral. The typical example is the ionization of the surface hydroxyl groups on oxides and oxidized minerals. This mechanism can also be induced by the polar nature of the mineral surface which makes it capable of adsorbing either H+ or OH– (basic-hydroxyl) from solution to form charged sites. For example, amine (-NH2) and alcohol-hydroxyl (-OH) groups adsorbed on the surface of minerals, tend to adsorb H+ or OH– groups due to their polar nature.   16 Lattice substitution In crystal lattices of certain minerals, one atom may be substituted by another atom of similar size but different valency. The replacement of aluminum for silicon within the faces of clays, for example, is responsible for permanent negatively charged surface sites, while the charge on the edges of clay particles strongly depends on the pH of solution. Preferential adsorption of ions Adsorption of multivalent ions, as well as surfactants and polymers, on the mineral surfaces leads to the generation of surface charges. The charge of the adsorbing ionic species would determine the sign of the surface charge. For example, anionic polymers will make mineral surfaces negatively charged at sufficiently high adsorption densities, while cationic ones will render the surfaces positively charged.  Specifically adsorbing ions have chemical or specific affinity for the mineral surface in addition to purely Coulombic interactions and affect the charge distribution at the solid/liquid interface but are not constituent ions of the solid. The ions interacting with the solid surface through electrostatic forces only are referred to as indifferent ions. These ions do not directly affect the surface charge of the mineral (Dobiáš, 2005). Indifferent ions can affect the system by compressing the electrical double layer, and hence by changing the magnitude of the zeta potential. Those ions that are constituents of the solid matrix and can adsorb on the mineral surface and affect the surface charge are known as potential determining ions. Since protons and hydroxyl ions possess high affinity for many surfaces, these ions are often treated as potential determining ions for many solids, including oxide minerals (Lyklema, 1987). 2.4.2 Origin of charges on salt-type minerals  Preferential dissolution of surface ions and surface hydrolysis are two major mechanisms responsible for charge development of salt-type minerals. Generally, it is believed that for salt-type minerals, surface charging occurs primarily due to preferential diffusion of lattice ions to the solution phase under the influence of water molecules (Chander and Fuerstenau, 1984; Somasundaran et al., 1985). Due to hydration of their surface, all salt-type minerals such as calcite, dolomite, apatite are to some extent soluble in water and their dissolution leads to the formation of surface charges on their surface. When an excess of anions is present in solution, cations will   17 preferentially leave the surface and go into solution to satisfy the solubility product, which results in a net negative charge on the mineral surface. Similarly, when solution contains an excess of cations, anions are released to a greater extent and leave the surface, whereby a positively charged surface is formed. Formation of surface hydroxyl groups was also found to play a role in the surface charging phenomena on salt-type minerals (Pokrovsky et al., 2000). It is well established that phosphate ions function as potential determining ions for apatites (Amankonah and Somasundaran, 1985; Bell et al., 1973; Mishra et al., 1980; Mishra, 1978; Smani et al., 1975; Somasundaran, 1968; Somasundaran and Agar, 1972), while calcium ions act as either potential determining ions (Bell et al., 1973; Eigeles, 1958; Saleeb and De Bruyn, 1972) or specifically adsorbed ions (Aplan and Fuerstenau, 1962; Hanna and Somasundaran, 1976; Mishra, 1978; Somasundaran, 1968) for apatites. A comprehensive review of electrophoretic studies of phosphates and associated calcareous gangue minerals was given by (Smani et al., 1975). This review shows that H+, OH−, Ca2+, HPO42– act as potential determining ions for apatite.  2.4.3 Origin of charges on oxide minerals  The surface charge of simple oxide minerals such as hematite and silica is caused mainly by the dissociation of ionizable surface hydroxyl groups (Parks and De Bruyn, 1962; Yopps and Fuerstenau, 1964). Two models, by Parks & De Bruyn (1962) and Yopps & Fuerstenau (1964), describe oxide hydrolysis processes. According to Yopps & Fuerstenau (1964), surface hydroxyl groups are produced at the oxide surface thorough its interaction with water molecules. These hydroxyl groups undergo acid-base reactions as a function of the pH of the suspension, according to the following reactions, Equation 2.4 and Equation 2.5, and produce the surface charge. 𝑀𝑂𝐻(`abc) ⇌ 𝑀𝑂(`abc)– +	𝐻(ef)$  (2.4) 	𝑀𝑂𝐻E	(`abc)$ ⇌ 𝑀𝑂𝐻(`abc) +	𝐻(ef)$  (2.5) Both reactions are reversible and are described by their respective dissociation constants:  𝐾– = [𝑀𝑂(`abc)– \ × [𝐻(ef)$ \[𝑀𝑂𝐻(`abc)\  (2.6)   18 𝐾$ = [𝑀𝑂𝐻(`abc)\ × [𝐻(ef)$ \[	𝑀𝑂𝐻E	(`abc)$ \  (2.7) These equations indicate that the values of the dissociation constants and the concentrations of the positively and negatively charged sites are required for calculating the surface charge density on an oxide surface. Different models have been suggested for the calculation of these dissociation constants and the dissociation constants values for some oxides are available (Atkinson et al., 1967; Drzymala et al., 1979; Dzombak and Morel, 1990; Westall and Hohl, 1980). For example, the dissociation constants of pK–=10.2±0.3 and pK+=3.2±0.3 were obtained for hematite by Watanabe and Seto (1990). The charging mechanism of oxides implies that the surface charge density of all oxide minerals strongly depends on pH and the hydroxyl (OH–) and hydrogen (H+) ions function as the potential determining ions for the oxide minerals (Fuerstenau, 1970; Parks, 1965; Wood, 1946). Another mechanism suggested by Parks & De Bruyn (1962) involves partial dissolution of oxide and formation of hydroxyl complexes in solution and subsequently, adsorption of these hydroxyl complexes on the mineral surface: 𝑀E𝑂A	(`ijk+) + 	3𝐻E𝑂 ⇌ 2𝑀(𝑂𝐻)A	(ef) (2.8) 𝑀(𝑂𝐻)A	(ef) ⇌ 𝑀(𝑂𝐻)m					(ef)AHm + (3 −𝑚)𝑂𝐻(ef)–  (2.9) 𝑀(𝑂𝐻)m					(ef)AHm ⇌ 𝑀(𝑂𝐻)m					(`abc)AHm  (2.10) where M is the metal cation such as iron (Fe) in hematite, and m is the valency of the cation. 2.4.4 Development of electrical double layer When solid particle surfaces are brought into contact with aqueous solution, the solid surface is charged, by preferential dissolution of surface ions, ionization of surface groups, lattice substitution, or preferential adsorption of ions, and acquires a potential with respect to the solution. Consequently, an ion concentration profile develops around the particle/solution interface and the surface charge is compensated by equal charge distribution in the solution phase. Such an ion concentration profile that develops around a charged particle is referred to as the electrical double   19 layer (EDL). Various models including Helmholtz (1879) and Perrin (1904), Gouy (1910) and Chapman (1913), and Stern (1924) models (Grahame, 1947; Parsons, 1990) have been developed to describe the structure and characteristics of the electrical double layer. The Stern model as the most popular model is a combination of earlier models, Gouy-Chapman model (for the diffuse layer) and Helmholtz and Perrin model (for the inner layer). Stern introduced a modification to the Gouy-Chapman model by replacing the point charges with ions of finite size, and introduced the concept of the compact layer (also known as the Stern layer), with thickness of d, essentially equal to the size of a hydrated ionic radius. The thickness d is defined as the closest distance to which an ion can approach the surface. Stern also introduced the concept of specific adsorption of ions at the solid interface and hydration of ions.  Figure 2.1 shows the structure of the electric double layer and the potential decrease across it. The electric double layer is formed from two regions: the inner compact region, Stern layer, containing the adsorbed ions, and the outer region, diffuse or Gouy–Chapman layer, including ions which are distributed according to the influence of electrical forces and random thermal motion. The Stern layer is subdivided into the inner Helmholtz plane (IHP) and the outer Helmholtz plane (OHP). The IHP runs through the centers of the specifically adsorbed ions, whereas the OHP runs through the centers of the more weakly adsorbed hydrated ions and determines the beginning of diffuse layer (Shaw, 1980).    20  Figure 2.1 Schematic diagram of electrical double layer (above) and potential gradient across the EDL (below) (Zhu et al., 2017). The concentration profiles of ions, the numbers of positive and negative ions per unit volume, follow the Boltzmann distribution function. Therefore, for a positively charged particle, the concentrations of positive and negative ions, 𝐶$ and 𝐶– (number/volume), at a point away from the particle surface, where the electrostatic potential is 𝛹 (V), are given by the following equations (Shaw, 1980):   21 𝐶$ = 𝐶%𝑒𝑥𝑝 s−𝑧𝑒𝛹𝑘𝑇 w (2.11) 𝐶– = 𝐶%𝑒𝑥𝑝 s+𝑧𝑒𝛹𝑘𝑇 w (2.12) where z is the ion valency, e is the electronic charge (C), k is the Boltzmann constant and T is the absolute temperature (K).	𝑧𝑒𝛹 and 𝑘𝑇 are the free electrochemical energy and thermal energy, respectively. 𝐶% is the concentration of anions or cations at a point away from the solid surface where the electrostatic potential is zero. The electrostatic potential is assumed to be zero at a very large distance from the surface. Therefore, 𝐶% is considered as the bulk concentration of ions in solution.  The net volume charge density, 𝜌, at any point of solution is given by Equation 2.13 (Shaw, 1980):  𝜌 = 𝑧𝑒(𝐶$ − 𝐶–) = 𝑧𝑒𝐶% y𝑒𝑥𝑝 s−𝑧𝑒𝛹𝑘𝑇 w − 𝑒𝑥𝑝 s+𝑧𝑒𝛹𝑘𝑇 wz = −2𝑧𝑒𝐶%𝑠𝑖𝑛ℎ y𝑧𝑒𝛹𝑘𝑇 z (2.13) The relationship between the net volume charge density, ρ (C.m-3), the electrostatic potential, Ψ (V), for a flat surface is given by the Poisson equation:  𝑑E𝛹𝑑𝑥E = −𝜌𝜀  (2.14) where 𝜀 and x are the permittivity of water (Fm-1) and the distance from the surface (m), respectively. Combination of Equations (2.13) and (2.14) gives: 𝑑E𝛹𝑑𝑥E = −𝑧𝑒𝐶%𝜀 	𝑠𝑖𝑛ℎ y𝑧𝑒𝛹𝑘𝑇 z (2.15) By taking the boundary conditions into consideration (𝛹= 𝛹% at x=0 and 𝛹 = 0 at x = ∞), the following exact solution can be obtained:  𝑡𝑎𝑛ℎ y𝑧𝑒𝛹4𝑘𝑇z = 𝑡𝑎𝑛ℎ y𝑧𝑒𝛹%4𝑘𝑇 z 𝑒𝑥𝑝(−𝜅𝑥) (2.16) where 𝜅 is the Debye-Hückel parameter with units of reciprocal length (m–1) and its value is given by Equation 2.17:   22 𝜅 = „2𝑒E𝑁†𝐶𝑧E𝜀𝑘𝑇  (2.17) Where NA is the Avogadro constant,	𝐶𝑧E is the ionic strength of a 1:1 symmetric electrolyte which is an electrolyte with equal number of moles of constituent ions of similar valency. The parameter 1/	𝜅 is referred to as the thickness of the electrical double layer.  Using the Debye- Hückel approximation for low surface potentials, 𝑧𝑒𝛹%/2𝑘𝑇 ≪ 1, the Equation 2.16 reduces to Equation 2.18: 𝛹 = 𝛹%𝑒𝑥𝑝(−𝜅𝑥) (2.18) This expression is known as the Debye-Hückel approximation which indicates the exponential decay of the potential across the diffuse part of the electrical double layer for a flat surface. The approximate equation for the decay of potential in a spherical double layer is given by the following equation: 𝛹 = 𝛹% (𝑎 + 𝑑)𝑟 𝑒𝑥𝑝(−𝜅[𝑟 − (𝑎 + 𝑑)]) (2.19) where a is particle radius (m), d is the thickness of the Stern layer (m), and r is the distance from the center of the particle (m).  2.4.5 The electrokinetic (zeta) potential The electrokinetic or zeta potential, 𝜁, is the potential at the plane of shear where slip with respect to bulk solution is suggested to occur and is assumed to be equal to the Stern potential (Hunter, 1966). The exact position of the shear plane is uncertain. The shear plane is defined as a boundary between the hydrodynamically mobile and immobile fluid when liquid flows tangentially along a charged solid surface. The ions in the Stern layer will remain adsorbed with the surface. The liquid motion may be hindered in the region where ions strongly interact with the surface. It is, therefore, possible that the shear plane is located at a small distance further away from the surface than the beginning of the diffuse part of the EDL (Stern plane) which implies that the zeta potential is equal to or lower (in magnitude) than the diffuse layer potential. At a high electrolyte concentration, compression of the diffuse layer causes a steeper potential decrease  from 𝛹+ to the shear plane and therefore, smaller zeta potential values than the diffuse layer   23 potential, while at low electrolyte concentration, the decay of the potential is small and 𝜁 ≅ 𝛹+ (Delgado et al., 2007; Shaw, 1980). Also, the adsorption of  large macromolecules (e.g., surfactants and polymers) can cause the shear layer to be moved farther away from the Stern plane and result in zeta potential values lower than diffuse layer potential (Shaw, 1980). 2.4.6 Point of zero charge and iso-electric point  The point of zero charge (pzc) refers to the pH where the net surface charge density is zero while the iso-electric point (iep) refers to the pH where the zeta potential is zero. Below the iep, the surface is positively charged due to the fact that the concentration of positive species on the surface is higher than that of negative species, while above the iso-electric point, the surface is negatively charged due to the dominance of negative sites at the interface. For a given mineral, the iso-electric point is equal to the point of zero charge, if solution contains only indifferent electrolytes (in the absence of specifically adsorbing ions). Specific adsorption of ions on a mineral can lead to a shift of the iep of the mineral and even charge reversal at the diffuse layer (Dobiáš, 2005), whereas indifferent ions can only change the magnitude of the zeta potential and do not affect the pzc/iep of the mineral. As mentioned above, at low ionic strength, the decay of the potential as a function of distance is small which results in high zeta potential values, while at high ionic strength, the potential drop is sharper  and therefore low zeta potentials are obtained (Hunter and Wright, 1971). The zeta potential of apatite and hematite as a function of pH was investigated by different researchers and various pzc or iep values were reported. The pzc and iep values of hematite reported in the literature vary in a wide pH interval from 3 to 9.5 (Cornell & Schwertmann, 1996; Kosmulski, 2011, 2014; Parks, 1965). The values reported for the pzc of apatite also vary widely, from 4.35 to 7.6 for hydroxyapatite, and from 4 to >12 for fluorapatite  (Bell et al., 1973). The variation in the reported pzc values of the minerals in different studies can be due to the presence of specifically adsorbed ions and impurities, different measurement techniques, and the failure to achieve equilibrium in terms of mineral solubility and ion adsorption. Also, the source of the minerals used in the various experiments and the temperature of system can be expected to play a significant role in this discrepancy (Cerovic et al., 2009; Hunter, 1966; Kosmulski, 2004, 2002).   24 2.4.7 Electrokinetic phenomena  Electrokinetic phenomena arise when a charged surface is set in a relative motion with respect to the adjacent liquid phase, leading to disturbance of the equilibrium of charges within the electrical double layer, followed by charge redistribution at the interface in order to restore electrical equilibrium. Basically, there are four types of electrokinetic effects: electrophoresis, streaming potential, electro-osmosis and sedimentation potential (Leja, 1982). Electrophoresis considers the movement of charged particles relative to the surrounding liquid under an applied field. Streaming potential involves the measurement of the electrical potential generated by liquid flow in a single capillary or in a system of capillaries. Electro-osmosis involves the movement of liquid through porous plug or capillary bed under an applied potential. Sedimentation potential determines the magnitude of the potential developed across a column of a colloidal suspension settling under the influence of gravity. Overall, electrokinetic effects arise when the liquid moves and the solid is stationary (streaming potential and electro-osmosis), or when the liquid is stationary and the solid moves (electrophoresis and sedimentation potential). Of these four, electrophoresis and the streaming potential are the techniques of greatest practical interest (Kosmulski, 2011; Parks, 1965). Electrophoresis is most often used for non-settling fine particles, but the streaming potential method, in which a bed of coarse particles is stationary and the liquid flows through the bed, can be used for zeta potential measurements on coarse particles. Recently, some new techniques of electrokinetic potential measurements have been developed among which the electroacoustic technique has received the most attention and has proved to be an effective method for the measurement of electrokinetic potentials in concentrated aqueous suspensions. There are two main types of electroacoustic phenomena: the Ultrasonic Vibration Potential (UVP) where sound waves pass through a colloidal suspension forcing charged particles to vibrate and produce a macroscopic potential difference, and the Electrokinetic Sonic Amplitude (ESA) where an alternating electric field is applied to a colloidal suspension and a macroscopic acoustic wave is generated by oscillations of charged particles as they try to follow changes in the electrical field. Extensive reviews of electroacoustic technology can be found in Hunter (1998) and Greenwood (2003).   25 2.4.8 Zeta potential distribution measurements on mineral mixtures  In mixed mineral systems, the interfacial behavior of various minerals can be different from what would be expected on the basis of their individual characteristics. Interactions of dissolved cations or anions from one mineral with the other in the pulp can be considered to cause changes in the surface properties of minerals. Changes in the electrokinetic behavior of minerals in the presence of other minerals was reported by Le Bell and Lindstrom (1982) who observed considerable changes in the point of zero charge of fluorite when it was conditioned in carbonate solutions (Le Bell & Lindstrom, 1982). In a study by Amankonah and Somasundaran (1985) it was observed that calcite and apatite, which individually had completely different electrokinetic properties, exhibited almost identical behavior in a mixture (Amankonah and Somasundaran, 1985). The precipitation of one mineral over another can occur depending on the solution conditions. The surface conversion of apatite into fluorite (Lin et al., 1981), apatite into calcite (Stumm and Morgan, 1981), and smithsonite into cerussite (Yuehua et al., 1995) were reported. The different electrokinetic behavior of the minerals in a mixture from their behaviour in single mineral systems can also arise from slime coatings or aggregation between particles of different minerals.  Zeta potential values were shown to display multimodal distribution for multi-component mixtures (Connah et al., 2002). The study by  Liu et al. (2002) on a bitumen-clay system and Xu et al. (2003) on coal-clay mixtures revealed that the zeta potential distribution measurement can be an effective method for investigating slime coating phenomena in a complex colloidal system. These researchers provided a methodology for interpretation of the zeta potential distributions in a mixture of two types of mineral particles in terms of interactions between the two components. Figure 2.2 shows schematic zeta potential distributions for a two-component system. A zeta potential distribution of a single component, measured separately, typically forms a well-defined peak around an average zeta potential value. The extent of interaction could be predicted from the relative shift of the peaks and widths of the zeta potential distributions in mineral mixtures (Liu et al., 2002;  Xu et al., 2003).    26  Figure 2.2 Schematic zeta potential distributions for a binary particulate component system that can be interpreted for particle interactions. The black and white circles represent mineral matter (M) and coal (C) particles, respectively. (a) Zeta potential distribution of the two components measured separately; (b) binary mixture without attraction; (c) weak attraction (coal partially covered by mineral matter with some remnant free mineral matter particles); (d) strong attraction (coal fully covered with and possibly with some remnant free mineral matter particles); (e) strong attraction (coal partially covered with some but not sufficient mineral matter particles available for full surface coverage) (Xu et al., 2003).   27 In a dispersed suspension of two minerals, with no interaction, the original peaks are expected to remain unaffected and two separate zeta potential peaks corresponding to individual minerals would appear. In this case, a slight shift in position of peaks of minerals towards the other mineral may be observed due to the hydrodynamic interaction of moving particles with different electrokinetic potential. In the case of weak interaction between two minerals, a notable shift in the position of the peaks can be expected. Strong interaction, however, could result in a monomodal peak or in more closely-spaced bimodal peaks. The suitability of zeta potential distribution to study the interactions in mixed mineral systems has been demonstrated by other researchers (Deng et al., 2013a; Ding et al., 2006; Engwayu, 2015; Forbes et al., 2014; Kusuma et al., 2014; Liu et al., 2005, 2004, 2003; Wu et al., 2015; Zhao et al., 2006). However, the interpretation of zeta potential distributions in mixtures of minerals without knowing the actual composition of the tested suspension to identify the minerals remaining in suspension could be  highly misleading (Engwayu, 2015). 2.5 Stability of mineral suspensions When particles are suspended within a medium, random particle movements due to Brownian motion and mixing effects and subsequent collisions between particles are inevitable. As a result of these collisions, particles may rebound off one another or aggregate (Overbeek, 1977). When aggregation takes place an unstable colloidal system results, whereas a colloidal dispersion is considered to be stable when particles remain freely dispersed. Colloidal stability plays an important role in different processes such as selective flocculation of minerals and the flotation separation of mixed mineral suspensions.  2.5.1 Forces involved in aggregation/dispersion When two particles approach each other, a balance of attractive and repulsive forces controls the aggregation and dispersion processes between the particles. The DLVO theory (Derjaguin and Landau, 1941; Verwey and Overbeek, 1948) calculates the total interaction between particles (VT) by adding the attractive van der Waals interaction (VA) and repulsive electrostatic forces (VR). Van der Waals forces are attractive forces which arise from the interaction of atomic and molecular dipoles. These dipoles arise from the spontaneous polarization of one molecule by the fluctuations of the charges in the electric clouds of the second (Hunter,   28 1987). The attractive van der Waals interaction energy (J) at a short separation distance of 𝑑 (nm) between equal-sized spheres is calculated from Equation 2.20 (Hamaker, 1937). 𝑉† = − 𝐴𝑅12𝑑 (2.20) where 𝑅 and 𝐴 are the radius of particles (nm) and the Hamaker constant (typically of the order of 10-20 J), respectively. The negative sign, by convention, means attraction. The attractive interaction for two dissimilar spherical particles of different radii,	𝑅and 	𝑅E, is given by Equation 2.21 (Addai-Mensah and Prestidge, 2005; Takeo, 1999). 𝑉† = − 𝑅𝑅E𝐴6(𝑅 + 𝑅E)𝑑 (2.21) The electrostatic forces result from interactions of the double layer, which forms around all the particles in an aqueous suspension. The electrical repulsive interaction energy, (VR), between two identical spherical particles separated by a distance of d, is given by Equation 2.22 ( Hunter, 1993):  𝑉 = 2𝜋𝜀𝑅Ψ+E ln(1 + exp(−k𝑑)) (2.22) or by Equation 2.23 (Derjaguin, 1934): 𝑉 = 𝜀𝑅Ψ+E2 ln(1 + exp(−k𝑑)) (2.23) where Ψ+  is the Stern potential (V), k is the Debye-Hückel parameter (m-1), and 𝜀 is the permittivity of the medium (Fm-1).  The energy of repulsion between two spherical particles of two different sizes, 𝑅and 	𝑅E, and potentials, Ψ+–𝑎𝑛𝑑	Ψ+—,	is calculated from (Hogg et al., 1966): 𝑉 = 𝜋𝜀𝑅𝑅E(Ψ+–E + Ψ+—E )𝑅+𝑅E ™ 2Ψ+–	Ψ+—(Ψ+–E + Ψ+—E )	ln š1 + exp(−k𝑑)1 − exp(−k𝑑)› + ln(1 − exp[−2k𝑑])œ (2.24) Equation 2.24 applies for potentials of less than 50-60 mV and under the condition that the double layer thickness is small compared to particle size. The total potential energy of interaction is given by the following equation:   29 𝑉 = 𝑉† + 𝑉 = 	−𝐴	.		𝑅12𝑑 + 2𝜋𝜀𝑅Ψ+E ln(1 + exp(−k𝑑))										 (2.25) 									A typical variation of van der Waals, electrical, and total interaction energy as a function of the separation distance is depicted in Figure 2.3. As the separation distance between two particles approaching each other reduces to zero, electrical interaction energy increases exponentially, but van der Waals energy approaches negative infinity. The total potential energy curve shows an energy maximum which acts as energy barrier that must be exceeded for aggregation into primary minimum to occur. In the case of a small energy barrier, comparable to the energy of thermal motion of particles, particles may be brought to a very close distance and aggregate into the primary minimum. If the energy barrier is large, particles cannot aggregate into the primary minimum, but the total potential energy may fall into the secondary minimum. The aggregation at the primary minimum is strong and irreversible while at the secondary minimum, the aggregation is weak and reversible (Bijsterbosch, 1987; Lu et al., 2005).  Figure 2.3 Total potential energy of interaction between two particles according to DLVO theory (Bellmann, 2004). According to the DLVO theory of stability, fine particles can be expected to aggregate at their iso-electric point (coagulation), as the electrostatic forces are zero at the iep and only Primary MaximumVTSecondary MinimumPotential EnergyVRPrimary MinimumVAr  30 attractive van der Waals interaction forces act between particles. However, it was postulated that there is a critical absolute zeta potential value (25 - 30 mV) below which the colloids coagulate (Lu et al., 2005; Pashley, 1992; Shibata and Fuerstenau, 2003). There are two types of aggregation associated with electrical forces. When the surface charge of particles of the same type approach zero (by compression of the electrical double layer or by charge neutralization), particles undergo rapid aggregation (coagulation). This sticking together of particles of the same material is termed as homo-aggregation. Different minerals with the surface charge of opposite sign can also aggregate through electrostatic attraction, and this process is called hetero-aggregation. 2.5.2 Structural forces Besides electrostatic and van der Waals forces, additional particle interactions were also  identified (Christenson, 1988; Claesson, 1987; Greene et al., 1994; Shaw, 1980). These forces arise from rearrangement of water molecules at the solid-liquid interface, and are generally referred to as structural forces. The structural forces cause either attraction when the solid is poorly wetted by liquid or repulsion when the solid is strongly hydrated. Strongly hydrophilic surfaces characterized by contact angles lower than 20o show short-range repulsive hydration forces. Repulsive hydration forces were first extensively studied between clay surfaces (Van Olphen, 1977). Such repulsive forces were also measured in detail between mica and silica surfaces (Horn et al., 1989; Israelachvili and Adams, 1978; Leng, 2012; Pashley, 1982, 1981). These hydration forces arise when a hydration sheath is formed around the colloids due to adsorption of hydrated ions and water molecules at the solid–water interface. It was found that the interaction between molecularly smooth mica surfaces in very low electrolyte concentrations is consistent with the DLVO theory of stability, but at high electrolyte concentrations, hydrated cations bind to the negatively charged surfaces and cause an additional short-range hydration force (Pashley, 1982, 1981). The hydration forces can lead to stable dispersions, even at the pzc/iep when the colloids would be expected to aggregate. The dispersion effect of hydration forces at pzc value was observed by Yotsumoto & Yoon (1993a, 1993b) in their stability studies of rutile (TiO2) and silica (SiO2) dispersions. The magnitude of the hydration forces, Vhydration, follows a double exponential function (Yotsumoto and Yoon, 1993a):    31 𝑉Ÿ +be¡ki¢ = −𝑎2 £𝐶𝐷𝑒𝑥𝑝(−𝐷/𝐷) + 𝐶E𝐷E𝑒𝑥𝑝(−𝐷/𝐷E)¥ (2.26) where a is the particle radius (m), C1 and C2 are constants (Nm-1), and D is the distance from the surface (m). D1 and D2 are decay lengths (m) which characterize the forces over larger separation distances.  For hydrophobic surfaces, their affinity for water is very low and the surface does not strongly attract water molecules. Highly hydrophobic surfaces characterized by large water contact angles (>80-90 deg), exhibit an additional attractive force which is much stronger than van der Waals forces (Liang et al., 2007; Xu and Yoon, 1990, 1989; Yoon et al., 1997). Xu & Yoon  (1989, 1990) showed that attractive hydrophobic forces between coal and methylated silica dominated over the repulsive electrostatic forces and promoted coagulation even when the zeta potential was as high as –40 mV for coal particles, and –51 mV for methylated silica. Attraction between hydrophobic surfaces can be of surprisingly long range up to about 80 nm (Claesson and Christenson, 1988). The magnitude of hydrophobic interaction, 𝑉(, can be quantified by the following equation (Claesson et al., 1986): 𝑉( = − 𝐾𝐷E (2.27) where D is the distance from the surface (m) and K is the hydrophobic force constant of the order of 10-10 to 10-20 J and can be measured through contact angle measurements or direct atomic force measurement (Yoon et al., 1997).  The strength of the attractive hydrophobic forces and repulsive hydrophilic forces is respectively related to the degree of hydrophobicity and hydrophilicity of the surfaces. 2.5.3 Steric effect The adsorbed polymers or surfactants on the surface of particles can also affect the interaction between particles. Polymers are often electrostatically charged, so they influence the electrostatic interaction. In addition, the steric effect, arises when adsorbed polymer chains on approaching particles interact with each other. When two polymer covered surfaces approach each other, at some distance the polymer chains start to overlap. So, steric repulsion between interacting surfaces is caused by an increase in free energy associated with compression of polymer chains   32 against the surfaces (Napper, 1977). Other interactions mechanisms including polymer bridging, depletion, and hydrodynamic forces play an important role in dispersion behaviour of colloids (Addai-Mensah and Prestidge, 2005). 2.5.4 Slime coating Slimes are ubiquitous in flotation systems, originating from weathering of ores, the brittle nature of valuable and associated gangue minerals, comminution processes, mineral surface oxidation, and chemical precipitation (Yu et al., 2017). The presence of very fine particles (below 5 microns) known as slimes can negatively affect the separation of coarser particles from mixed mineral suspensions. Because of their relatively high specific surface area and surface energy, slimes consume reagents more rapidly than coarse particles, leading to non-selective activation or collection of the fines and insufficient activator or collector coverage of the larger sizes. In addition, the small mass of slimes leads to their recovery through water recovery or entrainment. Another major problem with fine particles is that slimes attach themselves to coarser mineral particles, a phenomenon known as slime coating.  Most slime coating phenomena are a direct result of aggregation between particles of different types, often through electrostatic attraction, since under a given set of conditions (pH, ion concentration) different minerals exhibit different surface charging characteristics. The consequence of slime coatings is a great loss of separation efficiency of the desired coarser particles, as their surfaces are coated with unwanted fines and become inaccessible to adsorption of flotation reagents. A slime coating therefore decreases flotation selectivity and lowers the recovery of metallic minerals by preventing them from adhering to air bubbles. The detrimental effect of slime coatings on flotation separation of minerals was demonstrated by many researchers (Bandini et al., 2001; Feng et al., 2018, 2012; Fornasiero and Ralston, 2006; Liu et al., 2019; Ni et al., 2018; Yang et al., 2018; Zhou and Feng, 2015). An extensive review on slime coatings in froth flotation, including particle adhesion mechanisms, slime coating detection and quantification techniques, influencing parameters, control methods and mitigation measures was given by Yu et al. (2017). 2.5.5 Influence of the ionic species on aggregation/dispersion  In a flotation system, a variety of species exist in process water which may introduce   33 undesirable effects. These species most come from dissolution of minerals. The amount of ions dissolved into solution depends on the mineral type, i.e. properties of surface ions, and solution conditions such as ionic strength, pH, the presence of other species in solution, and temperature. The ionic species can undergo the reactions such as hydrolysis, complexation, adsorption, and surface precipitation or precipitation in the bulk solution. The interfacial characteristics of minerals and therefore the efficiency of interfacial processes such as flotation is controlled by the complex interactions among these reactions. The effects of ions present in process water on flotation performance of minerals were studied extensively (Agey et al., 1963; Celik and Somasundaran, 1986; Deng et al., 2017; Guimarães and Peres, 1999; Ikumapayi et al., 2012; Kirjavainen et al., 2002; Scotti and Smith, 1993; Tian et al., 2018; Weedon et al., 2007; Xing et al., 2016).  The preceding discussion of the DLVO theory highlights the important role of electrostatic forces in the stability of colloids. The ionic species can influence electrostatic interactions and therefore modify forces acting between charged particles in aqueous suspensions. Such effects are essential in different areas such as flotation, flocculation, and water purification (Bolto and Gregory, 2007; DiFeo et al., 2001; Henderson and Wheatley, 1987; Leong et al., 2012; Valmacco et al., 2016). Ionic species affect electrostatic interactions in different ways. Indifferent ions affect the electrostatic interaction through compression of the electrical double layer. These ions cannot change the iep of the minerals but they affect the magnitude of the zeta potential since the electrokinetic potential of minerals depends on the concentration of the background electrolyte. High zeta potential values are often obtained at low concentration of indifferent electrolyte and as concentration increases, zeta potentials decrease due to compression of the EDL.  For particles dispersed in an indifferent electrolyte solution, an increase of electrolyte concentration screens the surface charges and leads to aggregation. The electrolyte concentration required to cause coagulation is known as the critical coagulation concentration (CCC), and is strongly dependent on the valence (z) of the counterion of the electrolyte (the Schulze-Hardy rule) (Verwey and Overbeek, 1948). Polyvalent ions such as Ca2+ and Mg2+, SO42–, PO43– are able to specifically adsorb on solid surfaces and strongly affect the electrostatic interactions and interaction forces through direct change of the surface potential (Besteman et al., 2004; Dishon et al., 2011; Pashley, 1984; Sinha et al., 2013; Trefalt et al., 2017; Valmacco et al., 2016). Polyvalent ions can shift the iep of minerals and lead to charge reversal of surfaces. Polyvalent metal cations form a range of   34 hydrolysis products as a function of pH. The polyvalent ion (Mn+) may hydrolyze according to the following reactions, depending on pH and concentration (Hunter, 1981): 𝑀¢$ + 𝑥𝑂𝐻H ⟷ 	𝑀(𝑂𝐻)§(¢H§)$ (2.28) 𝑀(𝑂𝐻)§(¢H§)$ + 𝑂𝐻H ⟷ 𝑀(𝑂𝐻)§$(¢H§H)$ (2.29) 𝑀(𝑂𝐻)§(¢H§)$ + 𝐻$ ⟷ 	𝑀(𝑂𝐻)§H(¢H§$)$ + 𝐻E𝑂 (2.30) The hydrolysis products were shown to have high adsorption affinity for charged surfaces (the hydrolyzed species are highly surface-active) (Wolstenholme and Schulman, 1950). For each metal ion, there is a critical pH range over which the hydrolysis and adsorption of metal ion species increases significantly (Atesok et al., 1988; Groppo and Parekh, 1996; James and Healy, 1972a, 1972b, 1972c; Manukonda and Iwasaki, 1987; Mpofu et al., 2003). The adsorption of metal ions and their hydrolysis products onto metal oxide/hydroxide particles was widely investigated (James et al., 1975; James and Healy, 1972a, 1972b, 1972c; Wiese and Healy, 1975a). Gaudin & Fuerstenau (1955) found that specific adsorption of Ba2+ on quartz led to a charge reversal. Similar effect was observed by Walsch & Dultz (2010) for the adsorption of Ca2+ and SO42– ions on hematite. It was reported that Ca2+ could reverse the surface charge of δMnO2 from negative to positive and therefore enhance phosphate adsorption onto δMnO2 (Yao and Millero, 1996). Potapova et al. (2014) observed that addition of calcium ions reversed the hematite zeta potential from negative to positive over pH 8. Similar results for the magnetite-calcium system and kaolinite-calcium have been reported by other researchers (Atesok et al., 1988; Dixon, 1985; Mpofu et al., 2003; Potapova et al., 2011; Su, 1998). Charge reversal and shift in pzc of goethite and ferrihydrite in the presence of 10-6 to 10-3 mol/L phosphate were also reported in several studies (Antelo et al., 2005; Arai and Sparks, 2001; Hansmann and Anderson, 1985). In these studies, the dispersions were equilibrated for long periods of time, 15-18 hours to 48 hours, before the zeta potential measurements. Specifically adsorbing ions can affect particle stability towards aggregation in two ways: 1) specific chemical effect; changing surface potential and charge through chemical bonding to the surface and 2) ionic strength effect; compressing the EDL. Many studies demonstrated a significant effect of specific chemical interactions between multivalent cations and anions and the   35 surfaces of minerals on their stability behaviour. Xu et al. (2015) investigated the effect of different ions on aggregation of positively charged hematite and goethite under the conditions of pH < pHpzc (pzchematite = 7.5, pzcgoethite = 7.9) or when the surfaces of the minerals were positively charged. NO3– and Cl– ions decreased the positive zeta potential values of hematite and goethite and promoted their aggregation by compressing their electrical double layer (EDL). High concentrations of NO3– and Cl– decreased the zeta potential to near zero but could not reverse the zeta potential values. By contrast, for SO42– and F–, the zeta potential decreased quickly even at very low electrolyte concentration. F– and SO42– anions promoted the aggregation of hematite and goethite mainly through specific adsorption and therefore neutralization of the positive surface charges of the particles. The compression of electrical double layer, as another mechanism responsible for decrease in zeta potential, was shown to play a less important role at low concentration of F– and SO42–. At high concentrations, F– and SO42– ions reversed the surface charge of hematite and goethite from positive to negative. In contrast, specific adsorption of cations, Ca2+ and La3+, increased positive surface charges of hematite and goethite particles and therefore improved their stability. In the study by Liang and Morgan (1990) it was observed that at pH 6.5, where hematite was positively charged, phosphate induced hematite aggregation by charge neutralization. As the solution equilibrium concentration reached 2.0 × 10–5 mol/L, the net surface charge was reduced to zero, leading to minimum stability of particles. A further increase in phosphate concentration in solution increased the stability of particles, as the surface potential was reversed to negative values and repulsive electrostatic forces arose between the negatively charged surfaces. The stability of hematite was also measured in a SO42– solution at pH 6.5, and sulfate ions showed specific interactions with the surface of hematite at that pH. Xu et al. (2017) demonstrated that specific adsorption of phosphate (PO43–) could promote or inhibit aggregation of hematite and goethite nanoparticles through changing the surface charge of particles. At low phosphate concentrations (0.1 × 10–5 mol/L and 5.0 × 10–5 mol/L for respectively goethite and hematite), the specific adsorption of phosphates and the neutralization of surface charges was shown to be responsible for hematite and goethite aggregation while at high phosphate concentrations (above 10–2 mol/L), destabilization of suspensions was attributed to the charge screening by accompanying cations (Na+ and K+) and compression of the EDL. At medium phosphates concentration (10–2 mol/L), phosphate improved the stability of hematite and goethite   36 suspensions, as specific adsorption of phosphates made the goethite and hematite negatively charged and increased electrostatic repulsive force between the nanoparticles.  It  was shown that the adsorption of multivalent anions such as sulfates and phosphates increased the negative charge density on the surface of clay minerals, affecting flocculation-dispersion phenomena (Lagaly and Ziesmer, 2003; Lima et al., 2000; Penner and Lagaly, 2001). A dispersing effect of phosphate and citrate ions was found to decrease the interparticle attractive force between clay slurries in the pH region of 4-12 (Leong et al., 2012). Teo et al. (2009) observed that a sufficiently high phosphate concentration completely deflocculated clay slurries. In the study on  slime coatings in the silica-titania system (Huynh et al., 2000), phosphate was shown to adsorb selectively on titania and remove silica particles from titania surfaces by changing the electrostatic forces between the silica and titania from attractive to repulsive. Polyvalent metal ions are known to promote aggregation through electrical double-layer compression and surface charge reduction or neutralization, especially when present as products of hydrolysis. A significant effect of calcium and cupric ions on  interactions between zinc sulphide and silicate minerals  was reported by Xu et al. (2000) who found that the addition of calcium ions, reduced the long-range repulsive force between sphalerite and silica, although no adhesion took place between the surfaces at contact. This reduction in repulsive force was attributed to the adsorption of calcium ions and therefore some degree of charge neutralization at the minerals surface. Compared with calcium, cupric ions had a more significant effect on the interaction forces, changing the interaction forces from repulsive to attractive. The same effect was observed in the study by Ren et al.  (2012) who showed that significant aggregation occurred between sphalerite and silica particles when minerals were conditioned in the presence of calcium ions. A strong effect of calcium on the dispersion behaviour of sphalerite and silica  was reported by DiFeo et al. (2001). Both minerals showed a tendency to coagulate in the presence of calcium ions. Adsorption of calcium hydroxy species (CaOH+) on minerals was suggested to be responsible for the decrease in the magnitude of the zeta potential values and therefore, increase in their aggregation. The stability of hematite in the presence of Na+, Ca2+ or Mg2+ at  pH 10.5 (above the pzc of the hematite sample) was studied by Liang & Morgan (1990). Sodium and calcium ions caused particle coagulation by electrical double layer compression and followed the Schulze-Hardy rule which states that the critical coagulation concentration (CCC) of electrolytes is extremely related to the valence of the   37 counterions. However, Mg2+ ions exhibited specific interactions with the hematite surface and therefore, the critical coagulation concentration value was lower than the value predicted by the Schulze-Hardy rule. In the froth flotation of apatite from magnetite, the dispersing effect of sodium silicate on iron oxides was weakened in the presence of calcium ions, as adsorption of calcium ions reduced the negative charge of the magnetite surface treated with sodium silicate (Potapova et al., 2011). Mpofu et al. (2003) investigated the effect of Mn2+ and Ca2+ metal ions on flocculation of kaolinite dispersions. Addition of Mn2+ and Ca2+ ions to the kaolinite dispersions had a dramatic effect on electrokinetic behavior of kaolinite and on the rheological behavior of kaolinite dispersions at pH values higher than 7 and 10, respectively. The effect was attributed to the formation of positively charged manganese and calcium hydrolysis species and subsequently their specific adsorption on the kaolinite surface. The adsorption of anionic polyacrylamide–acrylate flocculant (PAM) and therefore flocculation and dewatering behavior of kaolinite was also observed to improve significantly in the presence of metal ions at pH 7.5 for Mn2+ and pH 10.5 for Ca2+, where hydrolyzed species are formed. A similar effect of hydrolyzable metal ions on flocculation and dewatering behavior of kaolinite and other mineral dispersions was reported in other studies (Atesok et al., 1988; Groppo and Parekh, 1996; Manukonda and Iwasaki, 1987; Sworska et al., 2000). A number of stability experiments on mixtures of oxide minerals such as SnO2–TiO2, and Al2O3–TiO2 demonstrated that preferential solubility of one mineral and adsorption of hydrolysed metal species on the surface of the other mineral affected surface properties of minerals and the stability of the entire system. In a mixed system of SnO2– Al2O3, Healy et al. (1973) observed that soluble species from minerals influenced the electrical surface properties of the other mineral and hetero-aggregation in the mixed system occurred around their new iep values obtained in supernatant of the other mineral. The same solubility-re-adsorption phenomena were observed in a titania-alumina mixture by (Wiese and Healy, 1975b). It was found that hetero-aggregation in mixed TiO2–Al2O3 system depended significantly on the conditions of mixing and equilibration of the dispersions. Under equilibrium conditions, dissolved A1(III) ions from Al2O3 formed a complete coating of aluminum hydroxide on the TiO2 surface, and the iep of the TiO2 was shifted towards the iep of Al2O3. Consequently, the coagulation behavior of the system was identical to that of a Al2O3 dispersion; rapid coagulation in the mixed dispersion was observed in the pH region around the iep of Al2O3 (pH 9).    38 In some cases, soluble hydrolyzed species may convert to colloidal particles and precipitate as hydroxides. Therefore, in addition to charge neutralization and double layer compression due to adsorption of polyvalent ions, another aggregation mechanism, sweep coagulation, may be induced due to aggregation (polymerization) of the hydroxide precipitates. In this case, abundant precipitation of metal hydroxides occurs which entrap suspended particles (Amirtharajah and Mills, 1982; Johnson and Johnson, 1983; Packham, 1965). Bridging through hydroxide species  was proposed as another aggregation mechanism in the presence of hydroxide species (Healy and Jellett, 1967; Krishnan and Iwasaki, 1986). Aggregation of ZnO particles under alkaline condition, observed by Healy and Jellet (1967), was considered to be due to release of Zn ions to solution and formation of Zn(OH)2 above pH 7, which polymerizes and flocculates the ZnO particles. The same aggregation mechanism was proposed for sphalerite; aggregation in sphalerite-water dispersions was attributed to the flocculating action of the polymeric Zn(II) hydrolyzed species (Mirnezami et al., 2003). El-Ammouri et al., (2002) investigated the role of Mg in  sulphide slurries (plant slurries collected from Cu and Zn circutes), and Mg ions were shown to cause significant aggregation of the sulfide particles (sphalerite/chalcopyrite) above pH 9. Aggregation was considered to be due to the formation of Mg(OH)2 precipitate and consequent electrostatic bridging mechanism (attractive electrostatic interactions between positively charged Mg(OH)2 patches on one particle and the precipitate-free negatively charged regions on the second particle). A similar observation has been reported for quartz in the presence of Mg ions (Krishnan and Iwasaki, 1986). Also, a "bridging effect" of multivalent counterions between charged surfaces can promote aggregation (Israelachvili, 1992). Particle bridging occurs when part of the counterion (adsorbed on one particle) is adsorbed on a second particle. Particle bridging due to adsorbed metal ions was suggested to be responsible for aggregation of particles in different systems (Chen et al., 2006; Kloster et al., 2013; Leong, 2005; Liu et al., 2002; Liu et al., 2004; Mylon et al., 2004; Pashley, 1984; Vergouw et al., 1998). Vergouw et al. (1998) observed that in the presence of calcium, the dispersion behaviour of sphalerite was not consistent with zeta potential values. The greatest agglomeration occurred over the pH range 8-12, where the zeta potential was –20 mV, but not around the iep (pH 6). The aggregation behaviour was suggested to be due to either hydrophobic forces (associated with the change in surface speciation on the zinc sulfide) or bridging by calcium ions between the surface of particles. In a study on interactions between bitumen and clay minerals,   39 Liu et al. (2002) showed that in the presence of 10–3 mol/L calcium, aggregation between bitumen and montmorillonite increased significantly. The strong hetero-aggregation was attributed mainly to the bridging action of calcium between bitumen and montmorillonite surfaces and to a lesser extent to reduction in repulsive electrostatic forces. The bridging effect of calcium between montmorillonite clay and bitumen surfaces was also suggested by Liu et al. (2004). The results from force measurements between mica surfaces in LaC13 and Cr(NO3)3 solutions demonstrated that adsorbed hydrated Cr3+ ions were displaced from the mica surfaces by protons before final (primary minimum) contact between surfaces, while La3+ ions appeared to remain adsorbed and formed a strong bridge between two negatively charged mica sheets which led to a strong final contact adhesion (Pashley, 1984). A bridging action of metal ions was also observed by Leong (2005) who examined the effects of hydrolysable Cu2+, Al3+ and Th4+ ions on the zeta potential and yield stress behaviour of silica dispersions. The silica dispersion displayed three points of charge reversal (CR1, CR2 and CR3) in the presence of adsorbed hydrolysis products of Cu2+ and Al3+, but Th4+ caused the dispersion to display only one point of charge reversal (CR3). CR1, CR2, and CR3 refer to the pH of charge reversal (reversal of the sign of the electrokinetic potential) in order of increasing pH (James and Healy, 1972b). CR1, the positive to negative charge reversal, is a result of changes in the surface charge of the solid as pH increases. CR1 is, therefore, the iso-electric point of the solid. CR2, the negative to positive charge reversal, occurs at a pH below the bulk hydroxide precipitation pH. Hydrolysable species are responsible for the second charge reversal. CR3, the positive to negative charge reversal, is a result of the coating of the metal hydroxide on the solid and is equivalent to the iso-electric point of the hydroxide at complete surface coverage (James and Healy, 1972b). In the study by Leong (2005), CR2 appeared near the pH for the formation of the first hydrolysis (positively charged) product while CR3 occurred near the pzc of the formed metal hydroxides. The dispersion showed a high yield stress at CR2 at pH 5.2 for Cu2+. The appearance of this yield stress was attributed to particle bridging by adsorbed hydrolysis products, charged patch attraction and van der Waals forces. However, a very low yield stress at CR2 of 2.8 for Al3+ ions was found, because silica had a very small zeta potential (almost zero) at pH 2.8 and accordingly, particle bridging and charge patch attraction did not occur due to very low adsorption of hydrolysis products at this pH (2.8). The effect of calcium and phosphate ions on stability of hematite was investigated in a   40 number of studies. However, very few works examined the stability of hematite in the solutions containing both ions, or in mixtures of apatite and hematite. Also, hetero-aggregation of these minerals and coating of apatite crystals by fine hematite (slime coating) were not extensively studied. Most studies focused mainly on assessing the effect of different reagents and operating variables on the flotation separation of apatite and iron oxides in order to increase the selectivity of flotation separation. Therefore, fundamental data on the underlying phenomena are missing and the overall purpose of this thesis is to enhance our understanding of apatite-hematite interactions.   41 Chapter 3: Experimental Program In this study, a combination of techniques was used to investigate surface-chemical interactions between apatite and hematite in aqueous suspensions. First, the surface characteristics of single minerals in the presence and absence of dissolved mineral species was assessed. Secondly, the aggregation/dispersion properties of minerals and the mechanisms responsible for their behaviour in the mixed systems were examined.  3.1 Reagents and materials 3.1.1 Model minerals  Highly pure samples of the natural minerals, hematite and apatite, obtained from Brazil and Mexico respectively, were used for the study. Fine minerals were prepared by pulverizing the samples in stages in a ring and puck pulverizer to obtain a final d50 of around 5μm. Also, coarse samples in the size range from 300 to 150 µm were prepared for investigating slime coatings and the effect of one mineral on the stability of the other. Prior to experiments, the coarse minerals were washed and ultrasonicated in distilled water to remove any fines attached to their surfaces. Very clean apatite crystals of the same size (1.2 × 0.6 cm) and same shape (hexagonal) were selected for direct measurements of slime coating (Figure 3.1).  Figure 3.1 One of the clean hexagonal apatite crystals tested in this thesis.   42 The quantitative x-ray diffraction analysis showed that the natural hematite and apatite samples used in this study were respectively of 97.3% and 98.4% purity. According to the XRD results, shown in Table 3.1, apatite was in the form of hydroxyapatite (HAP). The analysis indicated the presence of traces of corundum and quartz in both samples. Corundum is suggested to be likely contamination from the micronizer mill grinding elements used for reducing the sample to the optimum grain-size range for quantitative x-ray analysis. Table 3.1 Results of quantitative phase analysis (wt.%). Mineral Ideal Formula Apatite Hematite Hematite a-Fe2O3 - 97.3 Hydroxyapatite Ca5(PO4)3(OH) 98.4 - Corundum Al2O3 0.4 1.9 Quartz SiO2 1.2 0.8 Total  100.0 100.0 The particle size analysis of the samples was performed by Malvern Mastersizer 2000. The Malvern Mastersizer 2000 is a laser diffraction particle size analyzer which has a wide measuring range of 0.02 - 2000 μm. Figure 3.2 shows the particle size distribution of fine minerals. As shown, apatite displayed a broader particle size distribution compared with hematite. For apatite, eighty percent was below 10.1 μm and ten percent was below 0.54 μm while for the hematite sample, eighty percent was below 4.6 μm and ten percent was below 0.9 μm. The median diameter of apatite and hematite was found to be 3.2 μm and 2.25 μm, respectively. The maximum particle size of about 56.3 μm and 50.2 μm was observed for apatite and hematite, respectively.  A Quantachrome Autosorb 1MP analyzer was used for determination of specific surface area of samples through gas adsorption analysis based on the Brunauer- Emmett-Teller (BET) model of adsorption. The specific surface area measured by N2 adsorption was found to be 2.84 m2/g for hematite and 2.90 m2/g for apatite.   43  Figure 3.2 Probability and cumulative functions of the particle size distribution of apatite and hematite. 3.1.2 Chemical reagents  Iron (Fe) and calcium standard solutions (1 g/L) were purchased from Fisher Scientific and used for preparing the standards for atomic absorption spectroscopy analysis. Trisodium phosphate (Na3PO4) and calcium chloride (CaCl2) were obtained from Fisher Scientific and utilized for controlling the concentration of ions in the solutions. Stock solutions of trisodium phosphate (0.1 mol/L) and calcium chloride (0.1 mol/L) were prepared by mixing desired amount of salt with 1 L of sodium chloride solution (0.01 mol/L). Dilute sodium chloride electrolyte solution (0.01 mol/L NaCl) was used for all measurements as background electrolyte in order to maintain a constant ionic strength. The pH of aqueous solutions was controlled using analytical grade hydrochloric acid and sodium hydroxide. Concentrated hydrochloric acid (HCl) was used for dissolving hematite and apatite, when the amount of minerals (apatite and/or hematite) in the suspension needed to be measured. Also, a 0204060801000123456780.01 0.1 1 10 100Cumulative volume (%)Volume (%)Particle size (μm )ApatiteHematite  44 stock solution of nitric acid (2%) was prepared by dilution of concentrated nitric acid and used for preparation of iron and calcium calibration standards. All these chemical reagents were ACS certified chemicals obtained from Fisher Scientific.  3.2 Experimental methodology, procedures and equipment  The experimental program included apatite dissolution experiments, zeta potential measurements, stability experiments, direct measurements of slime coatings on single crystals, atomic absorption spectroscopy (AAS) analyses, and inductively coupled plasma (ICP) analyses. The details are given in the following sections. 3.2.1 Dissolution experiments As the literature review section showed, apatite samples from different sources will produce different levels of ions in solution. As a result, to assess the effect of ions released by a specific apatite sample on surface properties of other minerals, the solubility information of the apatite sample should be collected. Therefore, in spite of many studies on apatite solubility, the solubility of apatite used in this study was examined by conductivity tests and direct measurement of ion concentrations in apatite supernatant. It was previously shown that the experimental conditions such as composition of solution, solid concentration, temperature, and mixing rate significantly affected the dissolution of apatite and the concentration of ions in solution (Dorozhkin, 2002). Therefore, all the experiments were conducted at natural pH (6.8), under the same temperature and mixing rate conditions.  For conductivity tests, a mass equivalent to different volume concentration of apatite (0.02%, 0.1%, 0.25%, and 0.5% vol) was mixed with 100 ml of deionized water in a 150 ml Pyrex Erlenmeyer flask. Then, the sample was conditioned in a temperature-controlled orbital auto shaker (IKA KS 4000 ic Control) at a temperature of 25oC and 250 rpm for 2 hours. Conductivity was measured at different conditioning times. In a study on the dissolution characteristics of apatite and calcite, it was observed that atmospheric carbon dioxide affected the solubility of calcite significantly but had no noticeable effect on the solubility of apatite (Amankonah et al., 1985). However, to minimize any possible effect of carbon dioxide, the flasks were covered by parafilm.    45 To determine the concentration range of ions in apatite supernatant, the composition of apatite supernatants was measured as a function of time. A desired amount of apatite (0.1, 0.2, and 0.5 g) was placed in 150 mL Pyrex Erlenmeyer flask. 100 mL of 0.01 mol/L NaCl solution was added into flasks. The flasks were covered by parafilm and conditioned at natural pH in a temperature-controlled orbital auto shaker (IKA KS 4000 ic Control) at a temperature of 25oC and 250 rpm. 8ml samples were taken from each flask at different conditioning times (0.5, 1, 3, 7, 24, and 48 hours). The samples were centrifuged at a centrifugal force of 10,000 g for 15 minutes with a Heraeus Biofuge Primo centrifuge and then were filtered using 0.2 μm pore diameter membrane filters. The ICP technique was used to determine the concentration of calcium and phosphorus in supernatant solutions. In this study, apatite supernatant solution used in the experiments was prepared by mixing 1 g fine apatite in 1 L background electrolyte at natural pH (pH 6.8) for one hour, unless otherwise stated. 3.2.2 Stability measurements Aggregation/dispersion states of suspensions are controlled by the interactions between particles. The stability of suspension towards aggregation and settling is therefore an indicator of inter-particle interactions. The techniques used to evaluate the aggregation/dispersion states of suspension include sedimentation methods, particles counting, pulp rheology methods, light scattering methods, electrical properties and filtration methods (Lu et al., 2005; Vincent, 1974). Two techniques based on light scattering principles were used to measure the aggregation/dispersion states of particles in suspension.  Turbidity measurements were performed using a Hach 2100 AN nephelometric turbidimeter. Turbidity is an optical property that results when light interacts with suspended particles in solution. Suspended solids scatter the light passing through a sample, and the intensity of scattered light is proportional to the solids content in suspension. This light scatter results in turbidity. A dispersed system is, therefore, characterized by high turbidity. As the extent of aggregation increases, the aggregates become larger and settle faster and therefore turbidity of suspension decreases. The turbidimeter measures the extent of light scattering by the particles in the suspension at an angle of 90o to the direction of the incident beam (nephelometric method) and provides a turbidity value in nephelometric turbidity unit (NTU). The optical system, shown in   46 Figure 3.3, is comprised of a tungsten-filament lamp, lenses and apertures to focus the light, a 90° detector to monitor scattered light, a forward-scatter light detector, a transmitted-light detector, and a back-scatter light detector. The instrument can measure turbidity using only the 90o scattered-light detector or the complete set of detectors. When using all detectors, the instrument’s microprocessor uses a mathematical calculation to ratio signals from each detector. The ratio detection system significantly reduces color interference and corrects for lamp fluctuations. Therefore, in the case of coloured minerals, the ratio detection system is recommended. Since in this study, one of the minerals (hematite) had a red colour, the ratio system was used to compensate for the colour. The turbidimeter has a wide measurement range from 0 to 10,000 NTU. The instrument was calibrated using <0.1, 20, 200, 1000, 4000 and 7500 NTU formazin standards.  Figure 3.3 Schematic of the Hach turbidimeter optical system (Pavanelli and Bigi, 2005). The measurement of stability of suspensions was also carried out with an optical dispersion analyser, Turbiscan Lab Expert (Formulaction, France), which provides backscattering/ transmission values versus time. The instrument uses the multiple light scattering theories to scan the turbidity profile of the entire height of the sample in a 5-cm cell. The main part of the Turbiscan Lab Expert is a detection head which is composed of a near-infrared light source (λ=880 nm) and two detectors. The transmission detector receives the light, which goes through the sample, while the backscattering detector receives the light scattered backward by the sample (Figure 3.4).  suspended solid concentration but different compositionmay not scatter the same amount of light. Also watercolour (Malcolm, 1985) and temperature (Sadar &Engelhardt, year not specified) may also bias turbiditymeasurements, but their effect is les severe.Another failing of turbidimetric analysis is the lowrepeatability of diluted sample measurement results,turbidity being an apparent optical property of water.Optical instruments measuring transmitted light arenamed transmissiometers and are different from theoptical backscatter sensors (OBS), which measure lightscattered by particles at right angles. All analyticalmeasurements require a primary standard upon whichthe calibration is based: one nephelometric turbidityunit (NTU) was established as the turbidity resultingfrom a suspension of one part per million of silica.Formazine is used as a primary turbidity standard(method 2130B, APHA, 1999) to calibrate the labora-tory nephelometer and the measurement unit is thenephelometric turbidity unit. The instrument used forthis study is a laboratory turbidimeter Hach 2100ANequipped with four photosensors for transmitted andscattered light and ranging from 0 to 10 000 NTU(Fig. 1).When the sample has an over-range turbidity, it ispossible to have a NTU measurement by diluting thesample. In order to analyse water samples, the instru-ment requires a sub-sample of 30ml extracted from theoriginal sample after a proper agitation. This proceduremight produce a non-representative sub-sample, parti-cularly when a dilution is needed.When a sample needs dilution, turbidity is calculatedusing the following relationship (US EPA, 1999):TR ¼ TD ðw þ sÞs: (1)where TR is the turbidity of the undiluted sample inNTU; TD is the turbidity of the diluted sample in NTU;w is the content of clear water in the dilution in ml; and sis the content of sample water in ml.To convert NTU data records to suspended sedimentconcentration in samples, it is necessary to establish amathematical correlation between nephelometric unitsand SSC on a sufficiently wide sample population.2.3. Settleable solidsSettleable solids are defined as the solids that settle inan undisturbed sample of liquid after a specific timeperiod. Water samples are thoroughly mixed and pouredin Imhoff cones: settleable matter is measured volume-trically after 1 and 24 h; the readings are standardised ona 1 l sample. This technique is very common in waste-water analysis and its units are usually expressed in ml/l.Measurement precision of the graduated cone is 0.1mlfor volumes less than 2ml; it increases to 0.5ml forvolumes of less than 10ml and it is 1ml for volumesranging between 10 and 100ml. In the remainder of thepaper, settled solids are given as free settled solids (FSS)to indicate that the material settled is mostly inorganicmatter and subject to free settling (method 2540F:APHA, 1999).3. Results3.1. Turbidity-suspended sediment concentrationrelationshipThree to four turbidity measurements were made oneach laboratory sample. Readings might result within oroutside the instrumental NTU range: in the former case,readings had high repeatability and small deviations; inthe latter case it has been verified as lower reliability ofthe results according to the increase in dilution. Table 3presents the results of the turbidimetric analysis forsamples 1–6: samples 1–4 did not need any dilution,whereas samples 5–6 required dilution because theturbidity was slightly outside the instrumental range.Samples 7–12 required high dilution rates and thismanipulation affected the quality of the results.To further test the existence of a difference inconsistency between the diluted and undiluted samples,ARTICLE IN PRESSLED source90˚ detectorBack scatter detectorForward scatter detectorTransmitted light detectorSample cellFig. 1. Sche a cross-se tion of a labora ory turbidimeter; LED, li ht-emitting diodeD. PAVANELLI; A. BIGI78  47  Figure 3.4 Schematic of the Turbiscan optical system (Buron et al., 2004). The vertical motion of the detection head along the cylindrical cell enables transmission and backscattering data to be obtained at different time intervals from which localized phenomena such as particle size variation (aggregation, flocculation) and particle migration (sedimentation) can be identified and quantified. The instrument is able to detect real-time changes in transmission and light scattering intensities and acquire data in less than 20 seconds for each full scan thus quickly monitoring changes in stability of the tested dispersion/emulsion. This instrument can, therefore, detect changes in the stability of suspensions long before they become visible. The measurements are performed with a very high vertical resolution since data points are obtained every 40 μm along the height of the cell (55mm). A great advantage of the Turbiscan is that it can measure the long-term stability of opaque and concentrated colloidal dispersion and does not require any dilution of the sample, so it is appropriate to investigate the actual phenomena and the dispersion state of the system.  Figure 3.5 shows an example of Turbiscan spectra of the suspensions. X-axis and Y-axis represent, respectively, the height of the cell and the transmitted or backscattered light percentage. The sedimentation rate and clarification rate as an indication of pulp stability can be obtained from backscattering spectrum and transmission spectrum, respectively.  HOW IT WORKS ?TURBISCANma 2000Coalescence and sedimentation of a concentrated cosmetic emulsion (O/W, Φ = 40%)The TurbiScan MA 2000 detects the destabilisation20 times earlier than the naked eye. Moreover, it allows to fully understand the destabilisa-tion causes : here the coalescence phenomenonoccurs first, resulting in big droplets which sediment.Compared sedimentation of two latex suspensions (Φ = 10 %)The  TurbiScan MA 2000 gives an easy to accesspicture of products behaviour comparison. The drawn kinetics give the thickness evolution ofthe clarification phase (in the sample top) as afunction of time. Due to the flocculation of the par-ticles in the B product, their settling rate is biggerthan for the A product.Vs (A) = 8.3 10-8 ms-1  : d (A) = 2.0 µmVs (B) = 42 10-8 ms-1 : d (B) = 4.7 µmd (B) is the equivalent diameter of the sphere whichsettles at the same speed than the floc.The TurbiScan MA 2000 allows to calculate the particle mean diameter :• by measuring λ* in the sample heart (BS = 95 %, λ* = 96 µm) : d ≈ 1.9 µm• with the settling rate measurement (shift velocity of dispersion/continuous phase interface ,Vs ≈ 8.3 10-8 ms-1) : d ≈ 2.1 µm (General Law of Sedimentation, Snabre, Mills, 1994)Detects concentrated dispersion nascent destabilisation’s phenomena and unravels their mechanismsto improve formulations, shorten and document ageing tests.Without dilution, it operates on emulsions, suspensions and foams:Up to 60% v/v concentratedFrom 0.1 µm to 1 mm particle sizeThis vertical scan macroscopicanalyser consists of a readinghead moving along a flat-bottomedcylindrical cell, while scanning theentire sample height. The readinghead itself consists of a pulsednear infrared light source and twosynchronous detectors:-The transmission detector picksup the light transmitted throughthe product,-The backscattering detectorreceives the light backscatteredby the product (135°).The reading head acquires trans-mission and backscattering dataevery 40 µm on a maximum heightof 80 mm. The profile obtainedcharacterise the product homoge-neity, particles concentration andmean diameter. It is representedon the software screen by a curveshowing the percentage of backs-cattered or transmitted light as afunction of the sample height (inmm). The acquisition along the productis then repeated with a program-mable frequency to obtain a super-imposition of product fingerprintscharacterising the stability orinstability of the product, whetherthey are identical or not.Multiple light scattering measurement for concentrated dispersion analysisComparisonQuantificationParticle size variationParticule size variations (flocculation orcoalescence) induce λ* or λ changes,and therefore BS & T variations on thewhole height of the sample.Particles migrationParticules migration phenomena (creaming or sedimentation) induceparticle volume fraction changes at the extremities of the sample. By following the migration front,Turbiscan MA 2000 allows the calculation of the migration rate. Stability If no particule size or volume fractionchange occurs, BS & T remain constant(all the profiles superimpose).FLOCCULATIONCREAMINGSEDIMENTATIONCOALESCENCEThe measurement performed allows the quantification of the physical processes involved : backscattered (BS) and transmitted (T) light fluxes measured depend respectively on the mean path length of photons in the dispersion λ and λ*. These physical absolute parameters, depending on particle diameter d and volume fraction Φ, give information on the real state of the dispersion (no dilution required). Dispersions instability is often the result of two different physical processes :Particle size increase (droplets or aggregates) due to coalescence or flocculation phenomena,Particles migration within the samples leading to creaming or sedimentation.The TurbiScan MA 2000 performs a kinetic analysis allowing the detection of these phenomena at an early stage.dh = detection area heightg(d) = Assymetry factorQS(d) = Scattering efficiency factorri = measurement cell internal radiusSTABILITYSedimentationSedimentationClarificationClarificationCoalescenceCoalescenceBackScatteringTransmissionSample Height(mm)Intensity (%)Application examplesMultiple Light Scattering TheoryDestabilisation Phenomena CharacterisationDestabilisation understandingFunctions The acquisition programallows the analysis of pro-ducts which destabilisevery quickly (1 scan every20 seconds) and qualitycontrol of stable products(1 scan per day).Integration modes are avai-lable to draw the destabili-sation kinetics : BS and Tmean value variations as afunction of time to analysedestabilisation intensities,peak thickness (par ticlemigration distance) as afunction of time to analysesediment or cream layerthickness evolution.An easy operation to direct-ly overlay many kineticsallows the comparison ofdifferent products destabili-sations.ConvivialityAll treatments can be saved(zoom, kinetic curves,...).Kinetics of reference pro-ducts can be saved as tem-plates, and easily compa-red with others analysis(ex : visualisation and selec-tion of formula more or lessstable than the reference).User friendly interfaceLatex suspension analysis(Φ = 10 %, d(manufacturer) = 2.03 µm)λ*dd VsVs(B)Vs(A)MeasurementcellResult of one scan of the sample  48  Figure 3.5 Example of transmission and backscattering profiles obtained by Turbiscan and the change in clarified layer thickness with time. A schematic of the measurement cell is also shown. In this research, a single average transmission value (percentage of light transmitted) was measured from the transmission profiles at the top 10 mm layer of the sample as a function of time and used for evaluating the stability of suspensions. High transmission values indicate strong aggregation, and low transmission is characteristic of a dispersed system.  To enable a comparison between the results for individual and mixed mineral systems, the clarification rate was used. To obtain the clarification rate, the evolution of the clarified layer thickness (transmission spectrum) at a transmission of 5% was followed as a function of time (Figure 3.5), or how fast a clear liquid layer developed, and then a clarification rate was computed from the slope of the curve obtained between 120 and 720 seconds (Figure 3.6). As aggregation increases, aggregates settle to the bottom of vial faster and clarification rate increases. Therefore, a higher clarification rate indicates stronger aggregation and a low clarification rate is characteristic of a non-settling dispersed system.  Change of clarified layer thickness5%Clarified layerSuspensionSedimentMigration of the clarification frontHeight mm)H= 0 mmH= 45 mmClarified layer thickness00h:00m:00s00h:02m:00s00h:04m:00s00h:06m:00s00h:08m:00s00h:10m:00s00h:12m:00s00h:14m:00s00h:16m:00s00h:18m:00s00h:20m:00s00h:22m:00s00h:24m:00s00h:26m:00s00h:28m:00s00h:30m:00s  49  Figure 3.6 Illustration of the calculation of clarification rate for hematite suspension in background electrolyte at different pH values. The slope of each line represents the clarification rate. The turbidimeter and the Turbiscan should provide equivalent data. However, because turbidity is an average for the entire height of the sample, it may not be sensitive to local changes in aggregation. Therefore, transmission measurement for the top layer, which is the sensitive layer if aggregation occurs, was used.  To determine the effect of pH and dissolved mineral ions on the aggregation behaviour of minerals, separate stability measurements of single and mixed minerals in background electrolyte and apatite supernatant were performed at different pH values. To establish a direct correlation between the results from single mineral systems and those from a mixture, the total volumetric solids content used in single and mixed systems was the same. For single minerals, a mass equivalent to 0.2% volume fraction of the fine mineral (530 mg of hematite and 320 mg of apatite) was conditioned in 50 ml of background electrolyte or apatite supernatant for 60 minutes in the temperature-controlled auto shaker at a temperature of 25oC and 250 rpm at a given pH value. A y = 0.0029x - 0.8436y = 0.0122x - 1.8867y = 0.017x - 1.8365y = 002468100 200 400 600 800Clarified layer thickness (mm)Time (sec) pH 5pH 6pH 7pH 9  50 conditioning time of 60 minutes was enough for the solution to reach steady-state pH values, though the concentration of calcium and phosphate ions in solution did not stabilize after 60 min when apatite was present in the system. A mixture was prepared by mixing equal volumes of hematite and apatite (265 mg of hematite and 160 mg of apatite) to reach a total solid content of 0.2%vol. 30 ml of the suspension was pipetted into a borosilicate glass vial for turbidity and subsequently zeta potential distribution measurements. To mask minor imperfections and scratches that may contribute to light scattering, the outside of the vial was coated with a thin layer of silicone oil and the oil was uniformly spread with a clean lint-free cloth (the silicone oil has the same refractive index as the vial). The turbidity was measured after the sample was settled for 5 minutes. The settling time was selected based on the turbidity values of different suspensions. At very short settling times, the turbidity of some suspensions was very high, above the detection limit of the instrument. At long settling times, some suspensions settled fast and therefore differences between them diminished. For the measurements using Turbiscan, 30 ml of the suspension was pipetted into the glass vial. The vial was placed in the instrument and the sample was scanned for 30 minutes at time interval of 1 minute. From each measurement, 30 scans were collected from which data as a function of time were obtained. For assessing the extent of the attachment of fine hematite particles to coarse apatite crystals (slime coatings), transmission profiles and turbidity values were used. The measurements were performed for fine hematite suspension (0.07%vol) and the mixture of fine hematite and coarse apatite (–300 +150 µm). All the turbidity measurements were conducted after the hematite suspension was settled for 5 minutes in the glass vial. 5 minutes was selected based on the turbidity values of the suspensions at different conditions, so that all turbidity values were in the measuring range of the instrument after 5 minutes. The sedimentation time (5 minutes) was long enough for coarse apatite crystals (in the mixed system of fine hematite and coarse apatite) to settle to the bottom and not to contribute to the turbidity or transmission values in the upper layers of the samples. It was experimentally verified by measuring the turbidity of suspension containing only coarse apatite (0.9 g in 30 ml of background electrolyte) after different settling times (Table 3.2). Suspension was mixed in a vial for 5 minutes using auto-shaker, then the vial was placed in the instrument and the turbidity values were measured as a function of time. The results show very low turbidity values even after 1 minute of settling, indicating that all apatite particles settled to   51 the bottom and did not contribute to the turbidity. For the measurement of transmission, the vial was placed in Turbiscan and the sample was monitored for 30 minutes at time interval of 1 minute, after 2 minutes of settling (long enough for coarse apatite particles to settle). Table 3.2 Turbidity of coarse apatite (– 300 +150 µm) suspension after different settling time. Settling time (minutes) Turbidity (NTU) 1 1.71 2 1.72 3 1.71 5 1.71  Figure 3.7 shows how the extent of aggregation of fine hematite with coarse apatite can be evaluated from turbidity values. The method is based on slime coating measurements performed by Uribe et al. between clays and chalcopyrite (Uribe et al., 2016) and is based on the assumption that if a slime coating forms, or when aggregation of fine hematite on coarse apatite occurs, the fine hematite particles would be taken down by coarse apatite and therefore fewer hematite particles would be left in suspension, leading to lower turbidity. When using Turbiscan, aggregation of hematite fines with coarse apatite particles is expected to result in higher transmission values for the mixed system than the value for fine hematite alone. It is noteworthy that in the mixed system, calcium and phosphate ions released by apatite can affect aggregation/dispersion of fine hematite and therefore the turbidity and transmission values. In order to make a correct interpretation based on the results obtained for single and mixed systems, the role of these ions needs to be taken into account. Two sets of experiments were carried out for evaluating aggregation of fine hematite with coarse apatite and the role of ions was considered in both sets of tests.     52  Figure 3.7 Schematic of the method to assess aggregation of fine hematite with coarse apatite. In the first set of stability tests, 112 mg of fine hematite was mixed with 0.9 g of coarse apatite (–300 +150 µm) in 30 ml background electrolyte in a vial using auto-shaker (25oC and 250 rpm) for 60 minutes at a given pH value. After mixing the minerals, the vial was placed in the instrument and suspension was measured (first measurement). Then, coarse apatite was removed by screening the suspension through a 150-mesh sieve (106 µm), and hematite suspension was conditioned again for 60 minutes using the auto-shaker (25oC and 250 rpm) before measurement (second measurement). Using this procedure, the effect of ions on stability of hematite in single and mixed system were expected to be the same. In the first measurement (fine hematite plus coarse apatite), any hematite attached to apatite would be taken to the bottom and therefore, only the hematite particles left in suspension are measured. In the second measurement (hematite only), all the hematite particles already attached to apatite should be separated and go back to suspension before removing coarse apatite and conditioning hematite alone. So that all the particles contribute to the turbidity or transmission results. Therefore, ultrasound treatment was applied to suspension before removing apatite by screening in order to separate hematite particles attached to apatite crystals.  t=0 minHematite particlesHematite particles + Apatite crystalst=5 minTurbidity T1Turbidity T2Hematite ApatiteIf slime coating occursthen T2<T1Fine hematite particles taken to the bottom by coarse apatite crystalsdue to slime coating  53 In the second set of tests, all experiments were conducted in apatite supernatant. For the mixture, 112 mg of fine hematite (–38 µm) and 0.9 g coarse apatite (– 300 +150 µm) were mixed together in 30 ml apatite supernatant in a vial. The suspension was conditioned for 60 minutes at a given pH value in the shaker (25oC and 250 rpm). Then, the vial was placed in the instrument and suspension was measured. Next, coarse apatite was removed by screening the suspension through a 150-mesh sieve (106 µm), and hematite suspension was conditioned again for 60 minutes using the auto-shaker (25oC and 250 rpm) before measurement. It should be noted that the main difference between the two sets of experiments was in the type of background solution. In the first set, dilute NaCl was used, in the second test, an apatite supernatant was used to prepare the suspensions. Although in both types of tests coarse apatite was present, it was felt that the results would potentially provide information about the effect of dissolution kinetics on the measured stability profiles.  The effect of different amounts of coarse apatite on the stability behaviour of fine hematite in background NaCl electrolyte and supernatant (prepared at natural pH) was also assessed using Turbiscan. Different size fractions of the minerals were used to facilitate their subsequent separation. For these tests, the samples (30 ml) contained an initial volume fraction of 0.07%vol of fine hematite (112 mg hematite) with or without different amounts of coarse apatite particles of the other mineral (–300µm + 150µm). The procedure for preparing the suspensions was the same as that used for evaluating aggregation of fine hematite with coarse apatite. The suspension (with or without coarse particles) was conditioned for 60 minutes at a given pH value in the shaker (25oC and 250 rpm) and then measured with Turbiscan for 30 minutes after 2 minutes of settling. In some tests, the zeta potential of the particles in suspension was measured using the same sample after stability measurements. Also, water analysis was performed for the supernatant in the presence and absence of fine hematite and coarse apatite at pH 10. The supernatant was prepared at natural pH (pH 6.8), then hematite or/and apatite was conditioned at pH 10 for 60 minutes. These conditions represent a processing scenario in which phosphate ores are ground at natural pH and then pH is raised to 10-11 for the flotation process. Then, the suspension was centrifuged at a centrifugal force of 10,000g for 15 minutes with a Heraeus Biofuge Primo centrifuge, and filtered using 0.2 μm pore diameter membrane filters. The solutions were assayed for their calcium and phosphorus content using the ICP technique.   54  In addition, the stability and zeta potential of hematite at pH 10 in the presence of different concentrations of calcium, phosphate, and mixtures of calcium/phosphate ions was measured with Turbiscan. The suspension was prepared by mixing a volume fraction of 0.07%vol of fine hematite in solutions containing the ions for 60 minutes (25oC and 250 rpm) and then measured. 3.2.3 Zeta potential measurements Electrokinetic properties of single minerals, hematite and apatite, and their mixtures were assessed with zeta potential measurements. Different techniques are available for the zeta potential measurement. The ZetaView® PMX100 (Particle Metrix, Germany), an electrophoretic instrument capable of measuring distributions of zeta potential values for a large population of particles, was used in this study. In a micro-electrophoresis configuration, the particles in a suspension move within an applied electrical field. Depending on the surface charge of the particles they either move to the anode or cathode. To capture the movement of particles in the ZetaView, a laser scattering microscope with a video camera is used. Laser light is directed into the focal point of the microscope lens, which enables the tracking of particles as they scatter light in this region. From the video data a velocity distribution of the particles is derived at the two stationary layers where the electroosmotic flow is cancelled by an opposing flow of the suspension, and therefore the particle motion is affected only by the applied electrical field. The stationary layer was positioned through the alignment of equipment using a colloidal suspension of polystyrene. The measured electrophoretic mobility is converted to the zeta potential using the Smoluchowski equation, Equation 3.1, which is valid for low zeta potential (𝜁≤ 50 mV) and therefore low concentration polarization effects (Delgado et al., 2007), low electrical field, and thin electrical double layer compared to particle size (Hunter, 1966, 1981; Lyklema and Overbeek, 1961): 𝑽𝑬 = 𝜺𝜻	𝜼  (3.1) Where, 𝜁, 𝑉',	𝜀, and 𝜂 represent the zeta potential (V), electrophoretic mobility of particles (m2 V -1s-1), permittivity of water (Fm-1) and viscosity of the electrolyte solution (Pa.s), respectively. The correction function 𝑓(k𝑎) in Henry equation (Henry, 1931), Equation 3.2, was calculated to validate the Smoluchowski approach.   55 𝑽𝑬 = š𝟐𝜺𝜻𝒇(k𝒂)𝟑𝜼 › (3.2) In 𝑓(k𝑎), 	𝑎 is particle radius (nm) and k is the Debye-Hückel parameter (nm-1). The parameter 1/k is referred to as the thickness of the electrical double layer. Henry’s function 𝑓(k𝑎) varies smoothly from 1 to 1.5 as k𝑎 changes from 0 to infinity. In the limit of a thin double layer compared to particle size (i.e. k𝑎≥300), Henry’s function approaches a value of 1.5 and the Smoluchowski mobility expression (Equation 3.1) is obtained. The Debye-Hückel parameter is determined from the following equation: k = 𝟑. 𝟐𝟖𝟖𝟏. √𝑰 (3.3) Where I is the ionic strength (mol/L). In this study, the background electrolyte is 0.01 mol/L NaCl solution (1:1 symmetric electrolyte with equal valency and equal number of moles of cations and anions). Therefore, the ionic strength is 0.01 mol/L and k equals 0.329 nm-1. The value of ka for hematite and apatite, with a median diameter of hematite at 2.25 μm and that of apatite at 4.7 μm, is 368.3 and 772, respectively which are larger than 300. Also, in the ZetaView a constant low electrical field of 30 V is applied to the suspension and the maximum measured zeta potential values for the minerals were around 45 mV. Therefore, the Smoluchowski equation was suitable for the measurements in this study. The samples from the stability tests were also used in zeta potential measurements. To explore the possibility of ion transfer between minerals due to their direct contact, the same samples in stability measurement of fine/coarse mineral mixtures were used for zeta potential measurements. Also, to investigate aggregation between fine hematite and fine apatite, the samples used in the stability tests on individual and the mixed mineral systems were measured with the ZetaView and zeta potential distributions were analyzed.  In this study, to make a valid interpretation of zeta potential distributions obtained for mineral mixtures, a procedure was developed for determining the composition of the tested suspensions using Atomic Absorption Spectroscopy. Appendix A shows how the amount of hematite and apatite (%vol) in tested mixtures used for zeta potential measurements were determined based on AAS results. In order to enable a direct correlation, the samples for AAS were collected from the same samples used for zeta potential tests; 15 ml sample suspension was   56 carefully pipetted from the sample container. It should be recalled that all experiments were conducted in apatite supernatant and solutions contained calcium and phosphate ions. Also, in the mixture, apatite undergoes dissolution and releases more ions into solution. In order to take this amount of calcium into consideration and correct the data, the concentration of calcium in the background solution was measured. Phosphate ions present in solution may cause strong chemical interference when determining calcium with AAS. To correct for this chemical interference, lanthanum can be used, as it forms a thermally stable compound with phosphate and, therefore, calcium absorption will not be affected. Thus, 5 ml of sample suspension was filtered using 0.2 μm pore diameter nylon syringe filter to remove all apatite and hematite particles, then 9 mg lanthanum chloride was added to reach a concentration of around 1 g/L lanthanum and the solution was used for determination of calcium concentration in background solution. The rest 10 ml of sample suspension was used for measuring the amount of apatite and hematite in the suspension.  The dissolution behaviour of goethite and hematite has been investigated intensively (Cornell and Giovanoli, 1993; Cornell and Schindler, 1987; Schwertmann, 1984; Sidhu et al., 1981; Torrent et al., 1987). Among various mineral acids used, HCl was found to be the most effective acid for the dissolution of hematite and goethite. The efficiency of HCl extraction is attributed to the relatively greater complexing ability of Cl– with Fe3+ compared to other ferric salts (Abdus-Salam and Adekola, 2006; Parida and Das, 1996; Sidhu et al., 1981). The formation of Fe-Cl surface complexes increases the dissolution rate by reducing the repulsion between oxide surface and protons in solution and the attraction between surface Fe3+/Fe2+ and O2– ions (Sidhu et al., 1981). Therefore, concentrated hydrochloric acid was used for dissolution of hematite. The suitable condition for dissolving hematite was found by processing a specific amount of hematite at different conditions (acid concentration, temperature, conditioning time, and mixing rate) and measuring the percent of dissolved hematite based on the Atomic Absorption Spectroscopy results. Mixing hematite (250 rpm) in 5 mol/L hydrochloric acid for 24 hours at 60oC was found to result in complete dissolution of hematite. Apatite is a sparingly soluble mineral and dissolves readily in strong acids. Therefore, all apatite particles in the sample dissolve along with hematite after 24 hrs at 60oC.     57 Concentrated hydrochloric acid was added to the remaining 10 ml of sample suspension to reach a final concentration of around 5 mol/L hydrochloric acid. The suspension was then mixed for 24 hours in the temperature-controlled auto shaker at a temperature of 60oC and 250 rpm, so that all hematite and apatite particles were dissolved. For measuring hematite concentration, 5 ml of the solution was diluted with a 2% nitric acid solution, and the iron concentration was determined with AAS. As mentioned above, the background solution contains calcium and phosphate ions. In addition, apatite dissolves along with hematite and consequently, more calcium and phosphate are released into solution. Therefore, the possibility of the interference of calcium and phosphate ions with iron determination was explored. Solutions (each 10 ml) containing 5.0 × 10–3 g/L of ferric ions (14.5 10–3 g/L FeCl3) and different concentrations of calcium and phosphate were prepared. Then, the solutions were assayed for their iron content. As can be seen in Table 3.3, the presence of calcium and phosphate ions, with the concentration levels tested, had no notable effect on the iron concentration measured with AAS.  Table 3.3 Effect of calcium and phosphate ions on determination of iron concentration with AAS in a 5.0 × 10–3 g/L standard iron solution. Calcium added to the  solution (× 10–3 g/L) Iron measured with AAS (× 10–3 g/L) Phosphate added to the  solution (× 10–3 g/L) Iron measured with AAS (× 10–3 g/L) 0.0 5.0  0.0 5.0  2.0 4.9  2.0 4.9  5.0 5.1  5.0 5.0  8.0 4.9  8.0 5.0  10.0 5.0  10.0 4.9   For measuring apatite concentration, nitric acid solution containing lanthanum chloride was added to the rest of solution to avoid the chemical interference from phosphate ions. The solution was assayed for the calcium content using AAS. To calculate the amount of apatite, the calcium concentration in the background solution was deducted from calcium concentration measured after dissolution of apatite particles.  The suspensions for zeta potential measurements of single minerals were prepared by conditioning very fine mineral in 0.01 mol/L sodium chloride background electrolyte at a given pH value for 60 minutes. In order to determine the effect of dissolved mineral species on the   58 electrokinetic properties of minerals, the zeta potential measurement of apatite and hematite was also conducted in supernatant solutions. The supernatant was used for conditioning of mineral prior to zeta potential measurements. To identify the effect of individual ions on zeta potential of minerals, fine minerals were conditioned in solutions containing dissolved species, primarily calcium (CaCl2) and phosphate (Na3PO4), and zeta potential values were measured. After conditioning the mineral for 60 minutes, the samples were allowed to settle for different settling times, depending on their stability, to reach the concentration levels suitable for electrophoretic measurements. Then, about 5 ml of the suspension was inserted into the quartz cell of the instrument. To produce reliable statistics, twenty measurement cycles were conducted for each sample. The results reported in this study are the averages of the results from three experiments on separate samples. The width of the zeta potential distribution, whenever needed, was obtained from the cumulative zeta potential distribution curve. The difference between the zeta potential value at a cumulative frequency of 95% and the value at a cumulative frequency of 5% was defined as the width of the distribution (Figure 3.8).   Figure 3.8 Graphical illustration of the definition of the width of the zeta potential distribution used in this study. Fine hematite in apatite supernatant at pH 10. 0204060801001200102030405060708090100-70 -57 -44 -32 -19 -6 7 20FrequencyCumulative frequency (%)Zeta potential (mV)Width of distribution  59 3.2.4 Direct measurements of formation of slime coatings on a single apatite crystal To explore the interaction of hematite slimes with the apatite surface, a method of exposing an apatite crystal to hematite fines under mixing conditions was developed. A schematic of the experimental procedure for direct measurement of hematite coatings on an apatite crystal is shown in Figure 3.9. The procedure involved putting a stationary coarse apatite crystal, attached to a capillary, in a suspension of hematite slimes under varying conditions. Stirring was continued for 60 minutes, then the stirrer was switched off and the apatite crystal was taken out of the solution. In order to remove the hematite particles deposited on apatite due to settling or transferred mechanically along with apatite, the crystal was mixed in a water solution of the same composition as the tested suspension (pH, ion concentrations) for 2 minutes. Finally, ultrasonic treatment was applied to the apatite crystal immersed in deionized water for 10 min to detach hematite fines from the apatite surface.   Figure 3.9 Experimental procedure for direct measurement of hematite coating on single apatite crystal. Fixed CapillaryMagnetic stirrerHematite particles attached on apatite surfaceTemperature-Controlled Auto Shaker Atomic Absorption SpectrometerUltrasonic Cleaner   60 Concentrated hydrochloric acid (12 mol/L) was added to the solution containing hematite fines to reach a final concentration of 5 mol/L hydrochloric acid. The solution was then mixed for 24 hours in the temperature-controlled auto shaker at a temperature of 60oC and 250 rpm to ensure the complete dissolution of hematite. The solution was diluted with 2% nitric acid to keep the iron content in the concentration range of 0-5 mg/L, over which the relationship between the absorbance and concentration of iron was linear. Then, the quantitative determination of iron concentration was performed using AAS in order to quantify the attachment of hematite slimes onto apatite surface. To ensure that 10-minute ultrasound was long enough to remove all hematite particles from apatite surface, one apatite crystal was imaged before experiment (clean crystal), after experiment (crystal covered by hematite fines) and after ultrasound treatment. The images are presented in Figure 3.10. The images show that the crystal surface was completely clean after applying ultrasounds for 10 minutes. Furthermore, to verify that the ultrasound treatment is effective enough in separating the hematite fines from apatite surface, three apatite crystals, coated by hematite particles, were cleaned by ultrasonic treatment (for 10 minutes) and then were immersed in 5 mol/L hydrochloric acid for 5 minutes. As mentioned, apatite dissolves readily in strong acid, therefore the hematite particles left on the apatite crystals, if any, were expected to transfer to the acid solution. The solutions were conditioned for 24 hours in the temperature-controlled auto shaker at a temperature of 60oC and 250 rpm, and assayed using AAS. The results showed that no iron was detectable in the solutions, which confirmed that 10-minute ultrasound was enough for complete separation of hematite particles from the apatite crystal.  Figure 3.10 Observed images of apatite crystal. A: clean crystal before measurement, B: crystal conditioned in hematite suspension (5µm) at pH 8, and C: crystal after applying ultrasound for 10 minutes. A B C  61 For these experiments, different crystals, very similar in shape and size, were used at different conditions. However, for each condition, the same crystal was utilized for repeating the test (after being cleaned with ultrasonic treatment). The average weight of the apatite crystals used in these experiments was 470 mg. The results are reported as the ratio of the mass of attached hematite (mg) to the mass of apatite crystal (g). In order to eliminate the mechanical attachment and examine the effect of only surface chemistry on the coating phenomenon, and to increase the reproducibility of the experiments, only the crystals with flat surfaces were selected for the measurements. Using this technique, the effect of pH on slime coatings in background electrolyte was assessed. Also, apatite crystals were imaged before and after mixing with hematite particles at different pH in background electrolyte. After conditioning, the stirrer was switched off to allow hematite particles to settle and then the apatite crystal immersed in the clear solution was imaged with a digital camera (Figure 3.11). In order to examine slime coatings under the conditions of real systems, in the presence of dissolved ions, the experiments were also conducted with apatite supernatant solutions. Three different supernatant solutions were prepared by mixing different amounts of fine apatite (1 g, 3 g, 5 g) in 1 L background electrolyte for one hour, followed by separation of the solids from the liquid.   Figure 3.11 Experimental setup for direct imaging of slime coating phenomenon. Current techniques employed to study slime coatings such as scanning electron microscopy, electrokinetic measurements, atomic force microscopy, induction time measurement, Fixed CapillaryMagnetic stirrerLight sourceCamera  62 etc. are mainly focused on the interparticle interactions in a static and ideal condition which may not give an accurate account of true slime coatings under real dynamic conditions. An advantage of the present set-up is that the interaction between particles is subject to the ‘natural’ action of the fluid flow in a dynamic system, based on a realistic combination of hydrodynamic and chemical factors. Also, since this approach is non-intrusive and basically in-situ, the modification of the coating structure and the artifacts caused by sub-sampling, drying, and preparation are eliminated. The effect of different operational and chemical parameters could be readily examined by direct measurement of their impact on the attachment of hematite slimes to the surface via analytical analysis. This experimental approach is not limited to apatite/hematite, and other mineral systems could be investigated using this simple methodology. 3.2.5 Atomic absorption spectroscopy (AAS) Atomic absorption spectroscopy (AAS) is a spectroanalytical procedure for the quantitative determination of chemical elements using the absorption of optical radiation (light) by free atoms in the gaseous state. The main principle of atomic absorption spectroscopy is that atoms of different elements absorb and re-emit light in different ways. A beam source emits a set of known wavelengths or a continuous spectrum. As the different wavelengths of light pass through the sample, they encounter different elements that either absorb or pass along the light, depending on the characteristic wavelength of the sample atoms. An electronic detector of light measures the intensity of different wavelengths of light after they pass through the sample. Regions of the spectrum with decreased intensity indicate the absorption of specific wavelengths. These specific wavelengths correspond to specific atoms/ions. Quantitative measurements in atomic absorption are based on Beer-Lambert law, which states that concentration is proportional to absorbance. The Beer-Lambert law (or Beer's law), equation is the linear relationship between absorbance and concentration of an absorbing species. The general Beer-Lambert law is usually written as (Robinson et al., 2004): 𝑨 = 𝒂(𝝀)𝒃𝒄 (3-4) where A is the measured absorbance, a(𝝀) is a wavelength-dependent absorptivity coefficient, b is the path length (cm), and c is the analyte concentration (mol/L). The technique needs standards with known analyte content to establish the relation between   63 the measured absorbance and the analyte concentration. It is well known, however, that for some elements, particularly at high concentrations, the relationship between concentration and absorbance deviates from Beer's Law and is not linear. In this study, iron and calcium concentrations were measured with AAS at a wavelength of respectively 248.33 nm and 422.67 nm (Robinson et al., 2004). The relationship between concentration and absorbance for these two elements, shown in Appendix B, showed to be linear over the concentration range of elements in diluted standard solutions.        64 Chapter 4: Results and Discussion  This section contains the results from zeta potential and stability experiments for single mineral suspensions (apatite and hematite suspensions) and mixed mineral suspensions (mixture of apatite and hematite). Also, the results from apatite solubility experiments, atomic absorption spectroscopy and ICP are presented and discussed in this section. All experiments were carried out in triplicate and the error bars were displayed in the figures. Error bars indicate the standard deviation of triplicate samples. Experimental data points are typically connected with straight lines in order to highlight the trends. 4.1 Apatite solubility The conductivity of apatite suspensions, prepared by mixing different amount of fine apatite in deionized water at natural pH (6.8), as a function of stirring time is given in Figure 4.1. As can be seen, the conductivity sharply increased in the first 10 minutes and then increased slowly with time. The continual increase of conductivity during mixing time indicates that the dissolution equilibrium was still not reached after 2 hours. Based on the results, as apatite concentration was raised, the conductivity increased, meaning that more ions were released into solution. This effect can be attributed to an increase of the reactive surfaces of the solid. These results are in agreement with those reported by Greenwald (1942) and Levinskas and Newman (1955) who studied the solubility of model calcium phosphate and synthetic hydroxyapatite and determined the effect of varying solid-to-solution ratios on the concentration of calcium and phosphate in the solution. It was shown that there was an obvious tendency for more calcium and phosphate ions to dissolve at higher solids contents.    65  Figure 4.1 The conductivity of apatite suspensions prepared by mixing different amount of apatite in deionized water at natural pH (6.8) as a function of time.  The aqueous concentrations and element ratios during the dissolution of 1 g/L of apatite at 25°C as a function of time are shown in Figure 4.2. According to the results, the release of calcium into the solution was higher than the release of phosphate, compared to a stoichiometric ratio. The concentration of ions increased continually with time and the system did not reach an equilibrium after 48 hours. Even though the absolute concentrations increased with time, calcium to phosphate molar ratios (Ca:P) decreased, meaning that when dissolution proceeded, the release rate of calcium decreased, while the release rate of phosphate increased. During early stages of the hydroxyapatite dissolution reaction, mineral component ions were released in non-stoichiometric ratios, with ratios of dissolved Ca:P being higher than the mineral stoichiometric ratio of Ca5(PO4)3OH, i.e., 1.67. This suggests that calcium was preferentially released compared to phosphate from the mineral structure. As the dissolution progressed, the aqueous Ca:P molar ratios decreased and became closer to the theoretical stoichiometric value. It was shown that dissolution of hydroxyapatite and fluorapatite in aqueous medium was always non-stoichiometric at the 051015202530350 20 40 60 80 100 120Conductivity (µS/cm)Time (min)0.02 %vol0.10 %vol0.25 %vol0.50 %vol  66 beginning, but when the mineral approached equilibrium conditions at longer times, at any given pH, the solution Ca:P ratio approached a limiting value of 1.67 (Chaïrat et al., 2007a, 2007b; Guidry and Mackenzie, 2003; Mika et al., 1976). Once this value was reached, the mineral only maintained this ratio and continued to dissolve stoichiometrically (Mika et al., 1976).   Figure 4.2 Dissolution curve (the composition of apatite supernatant as a function of conditioning time) for 1 g/L apatite suspension in background electrolyte (0.01 mol/L NaCl) at natural pH (6.8).  Figure 4.3 shows the concentration of ions after 60 minutes of mixing different amount of apatite in 0.01 mol/L NaCl solution at natural pH (6.8). It can be seen that as more apatite was added to solution, the calcium and phosphate concentrations in solution became greater which is consistent with conductivity results. The most important conclusion from these results is that the apatite-solution system used in this thesis is not at equilibrium over the timescale of the experiments. The apatite supernatant prepared by mixing 1 g of fine apatite in 1 L background electrolyte (1 g/L) was used in the experiments in the following sections, unless otherwise stated. All experiments were carried out within 60 minutes and in the presence of low levels of ions. It 01234024681012140 10 20 30 40 50Calcium/Phosphorus ratioConcentration of ions (×10⁻⁵ mol/L)Time (hrs)CalciumPhosphorusCa/P  67 should be noted that there would be variation of concentrations as the conditioning time or the amount of apatite changed.  Figure 4.3 The concentration of calcium and phosphate after 60 minutes of mixing different amount of apatite in 0.01 mol/L NaCl solution at natural pH (6.8).  4.2 Hematite suspensions The results from electrokinetic properties and stability experiments for hematite in single mineral suspensions in background electrolyte as well as apatite supernatant are presented in this section.  4.2.1 Correlation between zeta potential of hematite and its dispersion behaviour in background electrolyte Results obtained for the zeta potential and stability of hematite sample in background electrolyte solutions are given in Figure 4.4. It can be seen that the iso-electric point (iep) of the hematite sample was found to be around 6.8 which is in the range of the literature values. The iep 012345670 1 2 3 4 5 6Concentration of ions (×10⁻⁵ mol/L)Apatite concentration (g/L)CalciumPhosphorusSupernatant ASupernatant BSupernatant C  68 values of hematite reported in the literature vary in a wide pH interval from below 3 to 9.5 (Cerovic et al., 2009; Kosmulski, 2014).  Figure 4.4 Electrokinetic properties and clarification rate of hematite as a function of pH in background electrolyte.  According to the stability results, the greatest aggregation, indicated by the highest clarification rate, occurred at around pH 7, which corresponds to the iep of the hematite sample. As pH moves towards acidic and basic regions, the magnitude of positive and negative zeta potential values of hematite, and therefore the repulsive electrostatic forces, increase. Therefore, the clarification rate decreases as pH moves towards acidic and basic range, meaning that hematite dispersion increases, and the clarification rate reaches zero at pH 4 and 9, where the hematite zeta potential is very high. The stability behaviour of hematite suspension in background electrolyte is in agreement with the classical DLVO theory of colloid stability, suggesting that electrostatic forces control the dispersion behaviour of hematite particles.  020406080-50-40-30-20-10010203040504 5 6 7 8 9 10 11 12Clarification Rate (mm/hr)Zeta Potential (mV) pHZeta potentialClarification rate  69 4.2.2 The effect of apatite supernatant on zeta potential and dispersion behaviour of hematite The effect of apatite supernatant on the electrokinetic properties of hematite can be seen in Figure 4.5. Examination of the results shows that the zeta potential of hematite is affected markedly by supernatant. In the supernatant, hematite became less negatively charged in alkaline media while its positive zeta potential values decreased to negative values in acidic and neutral solutions.   Figure 4.5 Zeta potential of hematite as a function of pH in background electrolyte and apatite supernatant.  To identify the ion responsible for the changes in hematite zeta potential in the supernatant at different pH values, the effect of individual ions, calcium and phosphate, on hematite electrokinetic properties was measured. Figure 4.6 shows the hematite zeta potential in solutions containing two concentration levels of calcium and phosphate. Concentrations of 3.7 × 10–5 mol/L calcium and 1.6 × 10–5 mol/L phosphate were selected based on the composition of supernatant (Figure 4.3). The effect of higher concentrations of calcium (10–4 mol/L) and phosphate (10–4 mol/L), closer to equilibrium condition, were also examined. -50-40-30-20-1001020304 5 6 7 8 9 10 11 12Zeta potential (mV) pH0.01 mol/L NaClApatite supernatant  70    Figure 4.6 Zeta potential of hematite in the background electrolyte as a function of pH, in the presence and absence of calcium and phosphate ions. The dotted line shows the hematite zeta potential in apatite supernatant.  The data presented for the zeta potential of hematite in CaCl2 solutions show that the magnitude of the zeta potential of hematite decreased notably in the presence of calcium ions, at both concentrations tested, but only above the iep. Calcium is a cation and so its adsorption on the positive surface (below the iep) is not electrostatically favored. As the mineral surface develops an increasingly negative charge above the iep, attractive electrostatic interactions favor more calcium adsorption.  The increase in adsorption of calcium with pH can also be attributed to the formation of the calcium-hydroxy species, Ca(OH)+. Equations 4.1 to 4.2 show ionic equilibria in calcium solutions. The equilibrium constants, K1 and K2 for the reactions are also given. CaOH+ + H+ ⇌ Ca2+ + H2O K1 = 5.0 × 1012 (Davies and Hoyle, 1951) (4.1) Ca(OH)2 (solid) ⇌ Ca2+ + 2OH– K2 = 5.5 × 10–6 (Butler, 1964)  (4.2) Using the actual dissociation constants for calcium, the following diagram, Figure 4.7, can be constructed for the total concentration of 10–3 mol/L. -50-40-30-20-1001020304 5 6 7 8 9 10 11 12Zeta potential (mV) pH0.0 3.7 10.0Calcium concentration(×10⁻⁵ mol/L)-50-40-30-20-1001020304 5 6 7 8 9 10 11 12Zeta potential (mV) pH0.0 1.6 10.0Phosphate concentration (×10⁻⁵ mol/L)  71  Figure 4.7 Species distribution diagram for a total calcium concentration of 10–3 mol/L. It can be seen that at pH lower than around 10, Ca2+ ions are predominant. As pH increases, Ca(OH)+ ions are increasingly produced while the concentration of Ca2+ decreases. The concentration of Ca(OH)+ reaches a maximum at pH 13.2. As pH increases further, Ca(OH)2 precipitates forms in the solution and Ca(OH)+ concentration decreases. Therefore, at high pH, calcium is present not only as calcium cation but also as the calcium hydroxy complex which exhibits a high affinity towards mineral surfaces. The diagram shows that the concentration of Ca(OH)+ ions is low at pH 10, but these hydroxy complexes are known to be very active in adsorption on other oxide surfaces. It is generally believed that hydrolyzed cationic species are more strongly adsorbed on negative surfaces than the free hydrated metal cations (Matijevic, 1973). Therefore, as pH increases, more adsorption on the surface occurs and the hematite surface becomes less negative.  The data in Figure 4.6 suggest that the adsorption of calcium on the hematite surface is not purely electrostatic and can be described as specific, as the cation has a tendency to reverse the sign of the zeta potential at higher pH and higher concentrations. Specific adsorption of calcium   72 on hematite and other iron oxides (magnetite) at high pH was previously reported (Potapova et al., 2014, 2011; Talebi Atouei et al., 2016). The zeta potential results for the effect of phosphate on hematite surface charge indicates that phosphate adsorbs strongly onto hematite over a wide pH range. In the presence of phosphate, the zeta potential values became more negative and the iep of hematite shifted from 6.8 to 4, if the data were extrapolated to lower pH values. Charge reversal and shift in iep of iron oxides in the presence of phosphate were mentioned in previous studies (Tejedor-tejedor and Anderson 1990; Hansmann and Anderson 1985; Arai and Sparks 2001). In the pH range 7 to 10, even though hematite is negatively charged, and the adsorption of phosphate is not electrostatically favored, phosphate ions still adsorb on hematite and make the zeta potential values more negative. This observation indicates that interaction between phosphate and the surface of hematite is not only electrostatic. A number of researchers/earlier studies (Elzinga and Sparks, 2007; Hayes et al., 1988; Huang, 2004; Li et al., 2006) demonstrated that phosphate was strongly adsorbed on hematite by formation of inner-sphere surface complexes. There are no significant changes in the zeta potential values above pH 10, indicating very low, if any, adsorption of phosphate over this pH range.  According to the results, the adsorption of phosphate on hematite decreases as pH increases. This pH effect is well-known for anion adsorption onto iron oxides and was reported by other researchers (Antelo et al., 2010, 2005; Hiemstra and Van Riemsdijk, 1996). The decrease of phosphate adsorption by increasing pH could be due to the fact that as pH is raised, the hematite surface charge becomes more negative and a stronger electrostatic repulsion between hematite and phosphate ions hinders phosphate adsorption. The lower phosphate adsorption at higher pH values could also be explained in accordance with the ligand exchange mechanism (Lin et al., 2017). As pH is raised, the concentration of phosphate species in solution changes. Equations 4.3 to 4.5 show the equilibrium equations and the stability constant for speciation of phosphate as a function of pH (Powell et al., 2005). Also, the distribution diagram for phosphate species is given in Figure 4.8. As the equations and graph suggest, H2PO4– is dominant at lower pH values, between pH 2 (pK1) and pH 7 (pK2), while the dominant species above pH 7 are HPO42– and PO43–. H2PO4– is more easily adsorbed on the surface than the two other species (Chubar et al., 2005; Li et al., 2016; Lin et al., 2017). As pH increases, the concentration of H2PO4– and thus the adsorption on hematite   73 decreases. It should also be noted that at pH 10, at which most of subsequent stability tests were carried out, the HPO42– anion is the most important ion in solution.  H3PO4 ⇌ H2PO4− + H+ pK1 = 2.1 (4.3) H2PO4− ⇌ HPO42− + H+ pK2 = 7.2 (4.4) HPO42− ⇌ PO43− + H+ pK3 = 12.3 (4.5)  Figure 4.8 Phosphate species distribution diagram as a function of pH (Liu et al., 2012). Comparison of the zeta potential results in supernatant (Figure 4.5) and those in CaCl2 and Na3PO4 solutions (Figure 4.6) demonstrates that the data for the supernatant at high pH are very similar to the results for calcium, while at lower pH, the supernatant results are very close to the phosphate data. However, it seems that in the presence of phosphate, a small amount of calcium adsorbs on the surface below the iep of hematite and makes the zeta potential slightly less negative. Likewise, in the presence of calcium, phosphate adsorption on the hematite surface occurs to a limited extent at high pH, making the zeta potential slightly more negative. It seems that at low   74 pH, phosphate strongly adsorbs on hematite and then, a limited amount of calcium ions start interacting with the adsorbed phosphate ions. At high pH, where phosphate has very low affinity for negatively charge hematite, calcium strongly adsorbs on hematite, then some phosphate interacts with adsorbed calcium on the surface. The interaction between adsorbed calcium and phosphate ions on the mineral surface can be explained by electrostatic interactions; the interaction of the negative charge of phosphate ions with the positive charge of adsorbed calcium ions and vice versa (Rietra et al., 2001; Stachowicz et al., 2008; Talebi Atouei et al., 2016). The stability results for hematite in supernatant are shown in Figure 4.9. The results for hematite zeta potential are also given in the figure in order to determine a correlation with the surface properties of hematite. The dispersion behaviour of hematite in supernatant was found to be completely different from that in background electrolyte. These effects were due to the dissolved species, calcium and phosphate, released from apatite into solution.  Figure 4.9 Surface properties of hematite (zeta potential and clarification rate) in apatite supernatant as a function of pH.  020406080-50-30-101030504 5 6 7 8 9 10 11 12Clarification rate (mm/hr)Zeta Potential (mV) pHZeta potentialClarification rate  75 Hematite aggregated strongly at pH 5-6, where the zeta potential of hematite in supernatant was low and the clarification rate was relatively high. The aggregation behaviour of hematite at low pH is related to phosphate adsorption, as phosphate adsorbs strongly on hematite and reduces the high positive zeta potential values. As pH was raised to pH 8, the zeta potential became more negative and therefore, clarification rate decreased, meaning that hematite dispersion improved. Above pH 8, the magnitude of the zeta potential decreased slightly while the clarification rate increased sharply, reaching the highest value at pH 11, indicating strong aggregation of the mineral under alkaline conditions. The dispersion behaviour over the pH range from 4 to 8 is consistent with zeta potential values and correlates with the DLVO theory of colloid stability. This suggests that electrostatic forces control the dispersion behaviour of hematite at lower pH values. However, above pH 8, the behaviour of hematite does not agree well with the DLVO model and points to mechanisms other than electrostatic attraction/repulsion since aggregation progresses even though the zeta potential remains unaffected. It is suggested that above pH 8, calcium starts to play an important role and affects the stability of suspension.  A similar effect of calcium on mineral aggregation was observed in other studies (DiFeo et al., 2001; Vergouw et al., 1998b). In the study by DiFeo et al. (2001), it was found that in the presence of calcium, homo-aggregation of sphalerite and silica particles was significantly enhanced at high pH values. The species responsible for a decrease in negative zeta potential values and strong aggregation of silica and sphalerite at high pH was suggested to be CaOH+ which is a more powerful coagulant than the calcium cation. Vergouw et al. (1998) observed that in the presence of calcium, the dispersion behaviour of sphalerite was not consistent with zeta potential values. The greatest agglomeration occurred over the pH range from 8 to 12, where the zeta potential was –20 mV, but not around iep (pH 6). The aggregation behaviour was suggested to be due to either hydrophobic forces (associated with the change in surface speciation on sphalerite) or bridging by calcium ions between the surfaces of particles (Vergouw et al., 1998b). Liu et al. (2002) studied the interactions between bitumen and clay minerals (kaolinite and montmorillonite) by zeta potential distribution measurements and found that calcium ions promoted strong aggregation between bitumen and montmorillonite at pH 8, even though a relatively dispersed system was expected based on the zeta potential values of individual minerals. The role of calcium was suggested to be decreasing the repulsive electrostatic forces and to a greater extent acting as   76 a bridge between bitumen and montmorillonite surfaces. The weak correlation between dispersion and zeta potentials was explained by calcium adsorption mechanism on montmorillonite. Calcium ions were suggested to adsorb on montmorillonite by a cation-exchange mechanism having a weaker effect on the net surface charge density. The bridging effect of calcium between montmorillonite clay and bitumen surfaces was also mentioned by Liu et al. (2004). The electrostatic patch coagulation mechanism was shown to play an important role in stability of suspensions (Cheng et al., 2010; Gregory, 1973; Wang et al., 2008). Gregory (1973) proposed that the coagulation of particles, by cationic polymers, before neutralization of the surface charges could be attributed to uneven charge distribution (patches) on the particles. In this mechanism, part of the charged surface is neutralized by adsorbing oppositely charged ions and an uncharged patch is formed on the surface. Consequently, coagulation could be induced by local contact between neutralized ‘‘patches’’ on the surface of particles. Also, charged ions can adsorb on some of the sites on the surface and inverse the surface charge, forming charged patches. In this case, the surface of particles contains positive and negative sites. When particles approach each other, there can be a significant electrical attraction between the oppositely-charged surface regions. A bridging force can be formed between charged surfaces by multivalent counterions. Calcium ions are capable of bridging two negatively charged surfaces together and induce aggregation (Israelachvili, 1992).  Analysis of zeta potential distributions is a simple method for evaluating the adsorption of ions on particles (Deng et al., 2013a, 2013b; Engwayu, 2015). Figure 4.10 shows the width of zeta potential distributions for hematite in background electrolyte and in apatite supernatant. The results show very broad zeta potential distributions in background electrolyte. This wide distribution of zeta potential can be attributed to the charge nonuniformity on the surface of individual particles (Feick and Velegol, 2002, 2000; Velegol et al., 2000), causing a variously charged population of particles.   77  Figure 4.10 Width of zeta potential distributions (90% of population) as a function of pH. The zeta potential distributions of hematite in apatite supernatant and background electrolyte (0.01 mol/L NaCl) are also shown in the graph.  In background electrolyte, the width of distribution is maximum at pH 7 which is very close to iep; as pH moves towards very low and high values, the distribution becomes narrower. The change in the width of zeta potential distribution with pH can be explained in accordance with the surface charge heterogeneity. The wide distribution around iep shows a varied population of particles with positive and negative zeta potentials. As pH moves towards low (or high) pH values, particles become more positively (or negatively) charged, however there is a limit to what the particles can achieve in terms of the surface charge density due to the limited number of sites on their surface, and the zeta potential reaches a plateau. Therefore, moving towards low pH, the most positively charged particles achieve the limit quickly and no longer become more positive, while the negative or less positive particles become more positively charged until they reach the plateau region, resulting in a narrower zeta potential distribution over this area. Likewise, as pH is raised to high values, particles become more negative but due to the limit near the plateau area, highly 0102030404 5 6 7 8 9 10 11 12Width of zeta potential distribution (mV)pH0.01 mol/L NaClSupernatant0102030405060708090100-69 -58 -48 -37 -27 -16 -6 5 15 26 36Frequency (%)Zeta potential (mV)pH 7- Apatite supernatant020406080100-69 -58 -48 -37 -27 -16 -6 5 15 26 36Frequency (%)Zeta potential (mV)pH 7- 0.01 mol/L NaCl  78 negative surfaces no longer change in terms of surface charge whereas the positive or less negative particles become more negative until they also finally approach a limiting charge density. There was no notable effect of supernatant on the width of the zeta potential distributions except at pH 7, where the width of distribution decreased by 5 mV in supernatant. As suggested by the zeta potential data, Figure 4.6, phosphate adsorption decreases with increasing pH, as the repulsive electrostatic forces between negatively charged hematite and phosphate anions at higher pH inhibit dense phosphate adsorption. It is suggested that the decreases in width of distribution at pH 7 is due to preferential adsorption of phosphate ions onto the more positively charged sites on the particles, which make the particles more uniformly charged. At low and high pH values (acidic and basic solutions), the fact that the effect of supernatant on the average zeta potential values of hematite is notable (Figure 4.5), while the width of distributions is not affected by supernatant to any considerable extent (Figure 4.10) suggests that the adsorption of phosphate (at low pH) and calcium (at high pH) proceeds on all particles rather than selectively on some particles. Also, since the concentration of calcium in supernatant is low, the negatively charged hematite particles are not likely to be fully covered by calcium at high pH. Accordingly, the particles are only partially covered by calcium and contain both positive calcium sites and negative unaffected regions. Under such conditions, an attractive electrostatic interaction between the positive calcium site on one particle and negatively charged site on another particle may develop and lead to aggregation. This coagulating effect of calcium is basically consistent with the bridging effect of calcium on interparticle aggregation proposed in other works (Liu et al., 2004, 2002). By analogy to graph presented by Tang et al. (2016) on calcium and magnesium bridging of muscovite, bridging of hematite by calcium ions can be shown in the following way.   79  Figure 4.11 Schematic of calcium ions bridging negatively charged colloids. According to the zeta potential results in Figure 4.6, as pH changes from 8 to 11, the adsorption of calcium species on hematite, and therefore the number of positive sites on the negatively charged particles, increases. The increase in the concentration of calcium sites, even by small amounts, on particles enhances the probability of attractive interactions between oppositely charged regions on different particles leading to greater aggregation. 4.2.3 The effect of calcium and phosphate on surface properties of hematite in alkaline solution Since most phosphate flotation separations are conducted under alkaline conditions, the behaviour of hematite at high pH was the major focus of further testing. The surface characteristics of hematite at pH 10 in the presence of calcium, phosphate, and both ions together were evaluated. The results for stability and zeta potential of hematite at pH 10 in the presence of different concentrations of calcium are given in Figure 4.12. The stability of hematite in supernatant (dotted line) is also given in the figure for comparing the effects produced by calcium and supernatant.  ++ ++––+–––––––––– –––––––––––––––––––––––– ––––––+Negatively charged hematite––––––––Calcium ion+++++++++++++––––––––––––––+++++++  80  Figure 4.12 The effect of calcium on the zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte. The dotted curve shows the stability of hematite in apatite supernatant. The results show that in the absence of calcium, the transmission values are zero, indicating that hematite is fully dispersed because, as shown in section 4.2.1, the zeta potential of hematite is highly negative in alkaline solutions. As calcium concentration increases, the negative zeta potential value of hematite decreases which indicates that the adsorption of calcium on hematite takes place. The addition of calcium promoted hematite aggregation, demonstrating a coagulating effect of calcium ions. According to the results, a calcium concentration of 2.0 × 10–5 mol/L produces the same result as that obtained in apatite supernatant. The concentration of calcium in supernatant is around 3.7 × 10–5 mol/L. This significantly higher level of calcium indicates that for reaching the same transmission values, the concentration of calcium in the presence of phosphate (in supernatant) needs to be approximately twice the concentration in the absence of phosphate (in dilute NaCl), which suggests that there is an interaction between calcium and phosphate ions from supernatant during adsorption on hematite. 0102030405060700 500 1000 1500 2000Transmission (%)Time (sec)0.0 1.01.2 1.72.0 3.0 !Ave = -32.9!Ave = -36.0!Ave = -31.7!Ave = -34.2!Ave = -42.5!Ave = -29.0Calcium concentration (× 10⁻⁵ mol/L)  81 By addition of 1.0 × 10–5 mol/L calcium to background electrolyte, the zeta potential changed by 6 mV and the aggregation of hematite was moderately enhanced. Raising the concentration from 1.0 × 10–5 mol/L to 2.0 × 10–5 mol/L had almost the same effect on the zeta potential (changing by 4.3 mV), while the effect on the stability was much stronger than the effect of the first addition of calcium. Moving from 2.0 × 10–5 mol/L to 3.0 × 10–5 mol/L, produced a weaker effect on both the zeta potential and stability compared to the previous additions. These results can be explained in terms of the calcium bridging mechanism, and dependence of hematite stability on the calcium concentration on the surface of particles. At lower calcium concentrations (below 1.0 × 10–5 mol/L), hematite still has a high zeta potential value and the bridging effect is probably weaker, and electrical attraction forces between the oppositely-charged surface sites of particles are not strong enough to bring highly charged particles together and induce significant aggregation. As the concentration of calcium ions in solution, and therefore on the hematite surface increases, the magnitude of the hematite zeta potential decreases. In addition, the higher calcium concentration leads to more bridging interactions and a stronger attraction between the hematite particles. The notable increase in aggregation with calcium addition could be a result of changes in the zeta potential and/or the bridging action of calcium ions. Therefore, the bridging effect of calcium must be the dominant factor in hematite aggregation in apatite supernatant, or even at very low levels of calcium in background solution. This sharp increase in aggregation of hematite is similar to the considerable change in hematite aggregation in apatite supernatant at high pH, where zeta potential changed slightly but stability decreased appreciably as pH was raised from pH 8 to pH 11. As discussed, it seems that when pH increases from 8 to 11, a stronger bridging effect between particles occurs in the presence of calcium, even though the zeta potential value is still high. This result again indicates the strong coagulating effect of calcium ions on hematite, and that aggregation of hematite is not simply a result of charge neutralization. Figure 4.13 shows the results for stability and the zeta potential of hematite at pH 10 in the presence of different concentrations of phosphate. Phosphate showed no effect on the stability behaviour of hematite and suspension remained fully dispersed. Also, in the presence of increasing concentrations of phosphate, the average zeta potential value is –42.6 mV, very close to the zeta potential of hematite in background electrolyte only. These results are consistent with the results for the zeta potential of hematite in the presence of phosphate in Figure 4.6 and demonstrate that   82 phosphate ions have no effect on the zeta potential and aggregation of hematite, suggesting that adsorption of phosphate on the hematite particles is insignificant at pH 10.  Figure 4.13 The effect of phosphate on zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte. The surface properties of hematite were also examined in the presence of calcium and phosphate together. The stability and zeta potential of hematite at pH 10 were measured at a fixed amount of one ion, and different concentrations of the other. Figure 4.14 shows the results in the presence of 2.0 × 10–5 mol/L phosphate and different concentrations of calcium. The stability of hematite in supernatant (dotted line) is also given in the figure. It can be seen that hematite is completely dispersed in the phosphate solution. As calcium was added into solution, aggregation of hematite increased. The results show that in the presence of phosphate, the curve reached the values obtained in supernatant with the addition of 3.8 × 10–5 mol/L calcium, while in the absence of phosphate (Figure 4.12) the same curve was obtained with addition of 2.0 × 10–5 mol/L calcium. Again, the results demonstrate that when solution contains phosphate, more calcium is required for enhancing hematite aggregation. 0102030405060700 500 1000 1500 2000Transmission (%)Time (sec)0.01.01.25.0!Ave = -42.6Phosphate concentration (× 10⁻⁵ mol/L)  83  Figure 4.14 The effect of calcium on zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte and in the presence of 2.0 × 10–5 mol/L phosphate. The dotted curve shows the stability of hematite in apatite supernatant. Based on the apatite solubility results, Figure 4.2, these concentrations of calcium and phosphate in solution should not lead to precipitation of calcium phosphate and both ions are very likely in solution. However, solutions with different concentrations of calcium and phosphate were also prepared by adding CaCl2 and Na3PO4 salts to background electrolyte. The solutions were mixed for 15 minutes at pH 10 and then their turbidity values were measured with the Hach turbidimeter. The turbidity results, given in Table 4.1, show very small values for all solutions indicating the absence of a colloidal precipitate. The turbidity of solution did not change as the concentration of calcium and phosphate increased to the level used in the experiments.    0102030405060700 500 1000 1500 2000Transmission (%)Time (sec)0.0 1.52.0 2.53.8 6.0!Ave = -34.9!Ave = -33.9!Ave = -42.4!Ave = -36.2!Ave = -31.7!Ave = -28.9Calcium concentration (× 10⁻⁵ mol/L)  84 Table 4.1 Turbidity of solutions with different concentrations of calcium and phosphate at pH 10. Calcium  (mol/L) Phosphate (mol/L) Calcium/phosphate  (Molar ratio) Turbidity  (NTU) 0.0 × 10–5 0.0 × 10–5 0.00 0.20 2.0 × 10–5 2.0 × 10–5 1.00 0.07 3.0 × 10–5 2.0 × 10–5 1.50 0.09 6.0 × 10–5 2.0 × 10–5 3.00 0.18 2.0 × 10–5 13.0 × 10–5 0.15 0.10 In addition, to test the possibility of formation of calcium phosphate precipitate on the hematite surface, precipitates were produced by introducing high concentrations of calcium and phosphate into background electrolyte and mixing the solution at pH 10 for 60 minutes. Then, the zeta potential of precipitates was measured. The results are given in Table 4.2. According to the data, a small amount of calcium mixed with a high concentration of phosphate produces a negatively charged calcium phosphate precipitate, but when the concentration of calcium is higher than the concentration of phosphate, a precipitate with a positive zeta potential is formed. Table 4.2 Zeta potential of calcium phosphate precipitates at pH 10 prepared by mixing different concentrations of calcium and phosphate. Calcium  (mol/L) Phosphate (mol/L) Zeta Potential  (mV) 3×10-3 2×10-3 1.1 2×10-3 2×10-3 -5.3 2×10-3 1×10-5 9.9 5×10-5 8.9 1×10-4 8.1 3×10-4 5.2 5×10-4 4.6 1×10-3 3.4 1×10-5 2×10-3 -28.2 5×10-5 -28.3 1×10-4 -28.0 3×10-4 -19.6 5×10-4 -17.3 1×10-3 -12.0   85 Another interesting observation from this experiment is that the zeta potential for a precipitate produced by stoichiometric amounts of ions in the solution (3 × 10–3 mol/L calcium and 2 × 10–3 mol/L phosphate) is around zero which was expected based on the charging mechanism of salt type minerals, where the ions are released into solution and leave an oppositely charged site on the surface of the mineral. Comparing the zeta potential values for calcium phosphate precipitates (Table 4.2) and the values in Figure 4.14, it appears that no precipitates form on the hematite surface by addition of calcium into solution, because the zeta potential values of hematite in Figure 4.14 are not consistent with the values obtained for any of the precipitates. When the concentration of calcium added to solution is higher than phosphate, ([Ca2+] > 2.0 × 10–5 mol/L), the zeta potential of hematite is negative, while precipitates formed at those proportions of calcium to phosphate are positively charged. When the calcium concentration is less than or equal to the phosphate concentration ([Ca2+]=1.5 × 10–5 mol/L and [Ca2+]=2.0 × 10–5 mol/L), the zeta potential is around –36 mV and –34 mV, while the most negative zeta potential value of a calcium phosphate precipitate was around –28 mV, which was obtained when the concentration of phosphate was ten times higher than the calcium concentration. These results show that a calcium phosphate precipitate is not forming under the experimental conditions. Thus, the difference in calcium concentration in the presence and absence of phosphate for reaching the same aggregation level (Figure 4.12 and Figure 4.14), is not related to precipitation but due to competing action between calcium and phosphate towards the hematite surface. One ion cancels out the effect of the other. Based on the results in the preceding section, at high pH, phosphate does not seem to adsorb on hematite in the absence of calcium. When calcium is introduced into solution, calcium ions adsorb strongly onto hematite surface and subsequently, phosphate ions start interacting with the adsorbed calcium ions and neutralize their coagulating effect. Therefore, in the presence of phosphate, more calcium than in the case in Figure 4.12 is needed to get to the same level of aggregation. At high calcium concentrations, the effect of phosphate ions is eliminated by calcium ions and hematite aggregates. It is interesting to note that the ratio of calcium to phosphate, when the transmission values reached the values obtained in supernatant, was 3.8:2, which is very close to the composition of supernatant, 3.7:1.6 (Figure 4.3). This indicates that the behaviour of supernatant can be replicated by controlled tests using known ion concentrations. These experiments using model compositions   86 explain that in supernatant there is a competing effect between calcium and phosphate. Phosphate neutralizes the coagulating effect of calcium as it electrostatically interacts with the adsorbed calcium ions on the surface. However, at pH 10, at the calcium and phosphate concentrations in supernatant, calcium overcomes the dispersing effect of phosphate and induces aggregation. It can be seen that as the concentration of calcium increased from 1.5 × 10–5 mol/L to 3.8 × 10–5 mol/L, the change in zeta potential was found to be small by about 4.5 mV but the increase in aggregation was substantial. Again, this significant change in hematite dispersion can be attributed to the decrease in the magnitude of zeta potential and/or the powerful coagulating effect of calcium ions. The increase in the aggregation level at low concentrations of calcium (below 1.5 × 10–5 mol/L) is not significant most likely due to the high zeta potential and strong repulsive electrostatic forces, and due to a weak bridging effect of calcium.  To compare the effect of different calcium concentrations in the presence and absence of phosphate on hematite stability, the clarification rate as a function of ion concentration was plotted based on the results in Figure 4.12 and Figure 4.14, and is shown in Figure 4.15. The figure clearly shows that different amounts of calcium are required in the presence and absence of phosphate for reaching the same aggregation level; and to compensate for the presence of phosphate, the calcium concentration should increase. It can be seen that above a certain critical calcium concentration, i.e., 1.0 × 10–5 mol/L in the absence of phosphate and 1.5 × 10–5 mol/L in the presence of phosphate, a small change in calcium concentration produces large changes in aggregation. Also, the graph shows that at very high calcium levels, the effect on stability becomes weaker.    87   Figure 4.15 The stability (clarification rate) of hematite (0.07%vol solids) as a function of calcium concentration Triangle: in the absence of phosphate and Circle: in the presence of 2.0 × 10–5 mol/L phosphate.  Figure 4.16 shows the results for stability and zeta potential of hematite in the presence of 2.0 × 10–5 mol/L calcium and different concentration of phosphate. As the concentration of phosphate increases, the negative zeta potential value increases and dispersion of hematite improves. These results demonstrate a dispersing effect of phosphate on hematite at high pH. The turbidity results in Table 4.1 show that calcium phosphate precipitates do not form at the calcium and phosphate concentrations tested. Furthermore, a comparison between the zeta potential values in Figure 4.16 and the values for calcium phosphate precipitates (Table 4.2) strongly suggest that calcium phosphate precipitates do not form on the hematite surface under the experimental conditions. 015304560750 1 2 3 4 5 6Clarification rate (mm/hr)Concentration of calcium (×10⁻⁵ mol/L)In the absence ofphosphateIn the presence ofphosphate  88  Figure 4.16 The effect of phosphate on zeta potential and stability of hematite (0.07%vol solids) at pH 10. All tests were conducted in background electrolyte and in the presence of 2.0 × 10–5 mol/L calcium. The dotted curve shows the stability of hematite in apatite supernatant. By addition of 12.0 × 10–5 mol/L phosphate, the negative zeta potential value increased to –42 mV, and hematite became fully dispersed. None of the zeta potential values of calcium phosphate precipitates are of this order. Thus, it is not the zeta potential of calcium phosphate (coating) that defines the zeta potential of hematite; it is a result of phosphate adsorption rather than phosphate precipitation. Phosphate ions adsorb onto the surface and interact with adsorbed calcium ions and therefore neutralize the coagulating effect of calcium ions. As phosphate concentration increases, more calcium sites bind to phosphate and consequently, the surface becomes more negative and also, the number of the available calcium sites on particles for interacting with negative sites on other particles decreases and hematite becomes more dispersed. Again, these results indicate that calcium and phosphate produce an opposing effect on the dispersion behaviour of hematite. As shown in the Figure 4.16, the most significant effect on dispersion was observed at lower phosphate concentrations, from 0 to 2.0 × 10–5 mol/L. As the 0102030405060700 500 1000 1500 2000Transmission (%)Time (sec)0.0 1.02.0 3.05.0 12.0!Ave = -33.3!Ave = -34.4!Ave = -35.2!Ave = -32.1!Ave = -42.2!Ave = -37.1Phosphate concentration (× 10⁻⁵ mol/L)  89 phosphate concentration further increases, a less pronounced effect on the dispersion of hematite is produced even though the effect on the zeta potential is notable. Also, it seems that as the concentration of phosphate on the surface increases, the rate of adsorption decreases and therefore much more phosphate is required for increasing dispersion. All the results in Figure 4.12, 4.14, and 4.16 indicate that there is a critical value of calcium concentration in the solution, and therefore on the hematite surface, above which a small change in concentration results in a significant change in hematite stability. Below this critical value, a change in calcium concentration on the surface has a smaller effect on stability. It seems that the bridging effect of calcium starts to play an important role when the calcium concentration on the hematite surface reaches a critical value. Comparing the results in Figure 4.14 and Figure 4.16 shows that the curves for a 1:1 ratio of calcium to phosphate concentration (2.0 × 10–5 mol/L calcium and 2.0 × 10–5 mol/L phosphate) are comparable and the net effect on stability and zeta potential is the same even though the ions were added in a different order. In other words, no effect of altering the order of addition of calcium and phosphate was found, which suggests that in both scenarios whether calcium or phosphate is added first, the end effect is the same and consistent with calcium adsorption and compensation by phosphate co-adsorption. The results indicate that calcium, as the specifically adsorbing ion, adsorbs directly onto the surface into the inner Helmholtz plane. Then, phosphate adsorbs on top of calcium most likely into outer Helmholtz plane, as it cannot adsorb directly on the negatively charged hematite. At high pH, adsorption of phosphate can only take place when calcium is already on the surface. A schematic of adsorption of ions on the hematite surface is given in Figure 4.17. When calcium is fixed and the phosphate concentration increases, phosphate is able to completely reverse the effect of calcium in terms of dispersion/aggregation. The same applies to the case when phosphate is fixed, and the calcium concentration is raised; calcium fully reverses the dispersion of the system and hematite aggregates. This observation highlights the importance of interactions between calcium and phosphate ions on the surface in defining the aggregation behavior of hematite particles. It seems that, in terms of aggregation/dispersion at high pH, aggregation is caused by calcium ions and prevention of aggregation by phosphate ions. Phosphate ions alone do not disperse fine hematite at high pH and an insignificant effect of phosphate ions on the zeta potential of hematite is observed.   90  Figure 4.17 Schematic of the hematite/water interface with adsorbed calcium ions in the inner Helmholtz plane and phosphate ions in the outer Helmholtz plane. 4.2.4 The relationship between zeta potential and stability of hematite All data in section 4.2.3 showed that over a certain concentration range of ions, despite the notable effect of calcium and phosphate on the hematite dispersion, only a minor effect on the zeta potential of hematite was observed. It is important to clarify how the ions affect the stability of hematite and whether aggregation/dispersion is driven by only zeta potential value or other mechanisms also play a role. To define the relationship between zeta potential and stability and to evaluate whether this small change in zeta potential causes the notable change in stability, all data in section 4.2.3 and the results for hematite (0.07%vol solids) in background electrolyte and apatite supernatant were replotted on one graph (Figure 4-18).  –––––––––––––––+ –++––IHP OHP Shear plane+++Hematite +–Calcium ionPhosphate ionIHP: Inner Helmholtz PlaneOHP: Outer Helmholtz Plane  91  Figure 4.18 The stability (clarification rate) of hematite (0.07%vol solids) as a function of zeta potential in Square: background electrolyte, Star: apatite supernatant, and in the presence of different concentrations of Circle: calcium, Cross: phosphate, Triangle: calcium at fixed amount of phosphate, Diamond: phosphate at fixed amount of calcium. A single trend line was drawn through the sets of data.  As shown in Figure 4.18, all data points fall on two curves. The data for hematite in background electrolyte follow curve 1, showing the effect of pH or basically the effect of potential determining ions. The curve shows a maximum around the iep of hematite where the highest aggregation level is expected to occur due to the effect of only pH (potential determining ions) and in the absence of specifically adsorbing ions. This is equivalent to the results in Figure 4.4 which shows a maximum clarification rate near iep. When pH is fixed and calcium and phosphate ions of different concentrations are introduced into the system, data points initially fall on curve 1 but then go in different direction on curve 2. These results demonstrate the effect of specific adsorbing ions, calcium and phosphate. In the initial part of graph, when the effect of calcium overlaps with the effect of pH, it seems that the interaction forces controlling the dispersion behaviour of hematite are electrostatic forces and van der Waals interactions and calcium causes aggregation 020406080100-60 -50 -40 -30 -20 -10 0 10 20 30 40 50Clarification rate (mm/hr)Zeta potential (mV)0.01 mol/L NaCl-pH from 4 to 11Apatite supernatant-pH from 5 to 11Adding Ca-at fixed amount of P-pH 10Adding P-at fixed amount of Ca-pH 10Adding Ca-pH 10Adding P-pH 10Curve 1Curve 2Adding phosphateAdding calciumCorresponding to: 1.7 × 10–5 mol/L calciumand 0.0 mol/L phosphate Corresponding to: 2.5 × 10–5 mol/L calcium and 2.0 × 10–5 mol/L phosphate   92 by decreasing the magnitude of zeta potential. However, at higher concentration of calcium, aggregation is enhanced compared to the effect of pH, indicating that it is no longer just electrostatic and van der Waals interaction forces acting between particles. There is an additional factor that takes the clarification rate to higher values compared to the values obtained based on only DLVO forces. It is clear that points for lower calcium dosages are following the trend with pH but at high calcium concentrations, an additional mechanism plays a role resulting in stronger aggregation. It is most likely related to the bridging action of calcium ions; a bridging effect between particles seems to be introduced at higher calcium concentrations. At high calcium concentrations, above 1.7 × 10–5 mol/L in the absence of phosphate and 2.5 × 10–5 mol/L calcium in the presence of 2.0 × 10–5 mol/L phosphate, higher specific adsorption of calcium on hematite takes place and accordingly, the bridging effect of calcium is enhanced, leading to stronger aggregation compared to the aggregation achieved due to the effect of the zeta potential only. In real systems with additional sources of calcium ions, these aggregation effects will be even stronger. The trend given by curve 2 can probably continue towards less and less negative values, and even turn to positive zeta potential values, by further increasing the calcium concentration. The results for hematite in supernatant in Figure 4.18 show that when the hematite surface is positively charged or near iep (pH 5 and 7), the data follow the trend with pH (curve 1) rather than with calcium because at lower pH, calcium has no notable effect on hematite, as discussed in section 4.2.2. Whereas, the points for alkaline solutions fall on curve 2, indicating the important role of calcium ions in controlling hematite dispersion. The effect of calcium on aggregation increases as pH is raised from 8 to 11; the effect is much more significant at pH 11 than it is at pH 8 because at high pH, calcium hydroxy species, which are much more active than calcium cations, form in the solution. Accordingly, the data switch from a curve determined by the electrostatic and van der Waals interactions to the curve reflecting the bridging effect of calcium when pH increases from low values to high values. According to the results, the effect of pH and calcium are similar when the concentration of calcium is low but they are different at higher calcium concentration at a given pH, or at higher pH at a lower calcium dosage; either very high calcium concentration or high pH (at lower calcium dosages) results in an enhanced bridging effect of calcium. Figure 4.18 confirms the powerful coagulating effect of calcium ions and explains the strong aggregation of hematite in supernatant at high pH values, where hematite still has negative zeta potential values.   93 As the focus of this work is on alkaline conditions, the data from curve 2 of the graph is shown separately in Figure 4.19. The figure clearly shows that at pH 10, calcium and phosphate ions produce opposing effects on the zeta potential and aggregation of hematite. Addition of calcium always increases the clarification rate (moving up the curve) by decreasing the magnitude of zeta potential and introducing the bridging effect while phosphate causes higher stability (moving down the curve) by making the zeta potential more negative and essentially neutralizing the effect of calcium. As can be seen, the effect of phosphate becomes visible only when calcium is present in the system.   Figure 4.19 The stability (clarification rate) of hematite (0.07%vol solids) as a function of zeta potential in background electrolyte, apatite supernatant, and in the presence of different concentrations of ions at high pH. 4.3 Apatite suspensions 4.3.1 Correlation between zeta potential of apatite and its dispersion behaviour  Figure 4.20 shows the zeta potential and stability of apatite suspensions in background electrolyte as a function of pH. The iep of the apatite sample in background electrolyte was found 020406080100-45 -40 -35 -30 -25Clarification rate (mm/hr)Zeta potential (mV)0.01 mol/L NaClApatite supernatantAdding Ca-at fixed amount of PAdding P-at fixed amount of CaAdding CaAdding PAdding phosphateAdding calcium  94 to be about 6.7, by extrapolation of the data to lower pH values, which agrees with the most commonly accepted literature values (Amankonah and Somasundaran, 1985; MacKenzie and Mishra, 1970; Mishra et al., 1980; Mishra, 1978). Generally, however, apatites are known to exhibit a wide range of iep values (Bell et al., 1973), so the iep value reported here is sample-specific. It can be seen that apatite is negative over the entire pH range tested. The zeta potential steadily decreases as pH increases, reaching a moderate zeta potential (around –22) at pH 11 (alkaline media). The highest clarification rate, strongest aggregation, occurred in the pH range from 7 to 9, where the apatite zeta potential was low. As the pH increased from 9 to 11, the magnitude of the zeta potential increased, so particles stayed more dispersed by electrostatic repulsion and the clarification rate therefore decreased. The stability behaviour of apatite is, therefore, consistent with zeta potential values.  Figure 4.20 Surface properties of apatite as a function of pH in background electrolyte: Electrokinetic properties and clarification rate.  020406080-30-20-10010206 7 8 9 10 11 12Clarification Rate (mm/hr)Zeta potential (mV)pHZeta potentialClarification rate  95 4.3.2 The effect of hematite and apatite supernatant on zeta potential and dispersion behaviour of apatite Figure 4.21 shows the results obtained for the zeta potential and stability of apatite in hematite supernatant as a function of pH. The hematite supernatant was prepared by conditioning 1 g/L hematite in background electrolyte (0.01 mol/L NaCl) for 60 minutes. The surface properties of apatite in background electrolyte are also given in the figure for comparison. The surface properties of apatite are not expected to be affected by hematite supernatant because hematite is an oxide mineral and its solubility is very low (Hunter, 1981); hematite supernatant should not contain ions released by the oxide. As can be seen in Figure 4.21, no effect of the type of background solution - dilute NaCl or hematite supernatant - was found in the tested pH range.   Figure 4.21 Surface properties of apatite as a function of pH in background electrolyte and hematite supernatant: Electrokinetic properties and clarification rate. ZP and CR stands for zeta potential and clarification rate, respectively.  The results obtained for the zeta potential and stability of apatite in apatite supernatant along with the surface properties of apatite in background electrolyte as a function of pH are given 020406080-30-20-10010206 7 8 9 10 11 12Clarification Rate (mm/hr)Zeta potential (mV)pHZP-0.01 mol/L NaClZP-hematite supernatantCR-0.01 mol/L NaClCR-hematite supernatant  96 in Figure 4.22. The apatite surface is not expected to be affected to any substantial extent by its own supernatant, since the concentration of dissolved ions from apatite should be very similar regardless of whether apatite is treated in background electrolyte or in its own supernatant. The results show that there is a slight change in the zeta potential and clarification rate values, but the trends are similar. These small differences in apatite surface properties are most likely related to the difference in the conditioning time (60 minutes in the tests with background electrolyte versus 120 minutes in the tests with apatite supernatant) and in the extent of dissolution of apatite. As was shown in section 4.1, the timescales of the experiments were insufficient to reach a solubility equilibrium. The impact of conditioning time (mineral-water contact time) on the zeta potential of apatite and other salt type minerals was previously noted (Mishra, 1978; Somasundaran, 1970, 1968).  Figure 4.22 Surface properties of apatite as a function of pH in background electrolyte and apatite supernatant: Electrokinetic properties and clarification rate. ZP and CR stand for zeta potential and clarification rate, respectively.   -15525456585-30-20-10010206 7 8 9 10 11 12Clarification rate (mm/hr)Zeta potential (mV)pHZP-0.01 mol/L NaClZP-apatite supernatantCR-0.01 mol/L NaClCR-apatite supernatant  97 In fact, the background electrolyte in contact with apatite particles is no longer composed of just dilute sodium chloride, because apatite starts dissolving once it is immersed in the aqueous solution. As shown in Figure 4.2, apatite dissolution progresses continuously, and the concentration of ions increases until an equilibrium state is attained. The concentration of ions in supernatant tests are higher than their concentrations in background electrolyte tests. According to the results of apatite solubility, calcium transfers to solution much faster than phosphate, meaning that in supernatant tests, the molar concentration of calcium in solution is higher than the phosphate concentration. As a result, calcium produces a stronger effect on the surface properties of apatite than phosphate does, and decreases the zeta potential to less negative values and consequently decreases the stability of apatite. Also, since stability and zeta potential measurements were conducted over the pH range from 7 to 9, where apatite possesses negative zeta potential values, calcium was expected to exhibit higher affinity for the apatite surface than phosphate. It was shown that zeta potential of apatite was not dramatically changed in either hematite or apatite supernatants (containing 3.7 × 10–5 mol/L calcium and 1.6 × 10–5 mol/L phosphate). However, it was felt that the impact of individual ions on the zeta potential and aggregation of apatite should be assessed. The results obtained in the presence of 10–4 mol/L CaCl2 and Na3PO4 (about 3 and 6 times higher than the concentration of respectively calcium and phosphate present in apatite supernatant) are presented in Figures 4.23. It was observed that additions of calcium and phosphate ions caused a significant change in zeta potential values almost over the entire pH range tested. The surface of apatite became more positively charged in CaCl2 solutions at all pH values, while the zeta potential values became more negative in the presence of phosphate. These results are in agreement with the data reported previously for hydroxyapatite (Amankonah and Somasundaran, 1985; Somasundaran and Wang, 1984; Wang, 1975). At pH 7 (close to the iep), calcium has a weak effect on the apatite zeta potential while the effect of phosphate on the zeta potential of apatite is stronger, which is similar to the effect produced by these ions on the hematite zeta potential at pH 7 (Figure 4.6).   98  Figure 4.23 Zeta potential of apatite as a function of pH in CaCl2 and Na3PO4 solutions.  At high pH, the effect of phosphate on the apatite zeta potential is different from its effect on the hematite zeta potential; at pH 10 and pH 11, hematite is not affected by phosphate (Figure 4.6), while the impact of phosphate on apatite (Figure 4.23) is significant, which points to the role of phosphate as a potential determining ion for apatite. In contrast to phosphate, the effect of calcium at high pH, pH 10 and pH 11, is very similar for apatite and hematite. It can be seen that the zeta potential curve for phosphate is almost parallel to the curve in background electrolyte (NaCl solution), but the curve for calcium is not, except for high pH values. Thus, calcium behaves differently than phosphate, though both ions are potential determining ions for apatite. 4.4 Interactions between fine hematite and fine apatite particles 4.4.1 Stability measurements The stability, clarification rate and turbidity, of hematite, apatite and a 1:1 mixture (by volume) of hematite and apatite suspensions in apatite supernatant at a 0.2%vol total concentration -35-25-15-55156 7 8 9 10 11 12Zeta potential (mV)pH10⁻² mol/L NaCl10⁻⁴ mol/L Calcium10⁻⁴ mol/L phosphate  99 of solids as a function of pH is shown in Figures 4.24 and 4.25. Figure 4.24 provides the clarification rate values obtained with Turbiscan. The results show very good reproducibility of these experiments. The dispersion behaviour of apatite and hematite was discussed in the preceding sections. The results for the mixture indicate that the clarification rate generally decreases with increasing pH. The clarification rate was high at pH 7 and pH 8, indicating low stability of the mixture, and slightly decreased over the pH range 8 to 10. A dramatic change in the stability of the mixture occurred above pH 10 as the clarification rate decreased sharply between pH 10 and pH 11, meaning that the mixture was much more stable at pH 11.   Figure 4.24 Stability of apatite, hematite, and the 1:1 mixture (by volume) in apatite supernatant as a function of pH.  It should be noted that the low clarification rate indicates high dispersion. Comparing the results for the mixture and those for the individual minerals, it can be seen that at high pH, the clarification rate is lower than the values expected from the results for individual minerals, assuming absence of any interactions between the minerals. The results for single minerals show 0204060801006 7 8 9 10 11 12Clarification rate (mm/hr)pHApatiteHematite1:1 Mixture  100 that at pH 11, hematite aggregates and settles quickly while apatite remains dispersed. The mixture is fully dispersed at pH 11, even though only a half of the mixture is a component (apatite) that should be dispersed while the other half (hematite) should aggregate. The clarification rate for the mixture should be in between the values for apatite and hematite, if there was no interaction between apatite and hematite particles. At pH 10, the single minerals show similar dispersion behaviour. However, the clarification rate for the mixture is lower than the values for the single minerals; mixture is more dispersed than what was expected from the values for the individual minerals. At pH 9, apatite aggregated and settled, but hematite was more dispersed. The mixture, however, behaved like hematite in terms of dispersion behaviour. At lower pH values, pH 7 and pH 8, the results for the mixture are in between the values for hematite and apatite. Since clarification rate is not an additive parameter, the results for the mixture at low pH cannot be used as evidence of aggregation between apatite and hematite. The data given in Figure 4.25 show the turbidity values collected by turbidimeter. As can be seen the reproducibility of turbidity measurements was not as good as the reproducibility of measurements with the Turbiscan. The results for apatite systems showed low turbidity values over the pH range 7 to 10, indicating aggregation of apatite. As pH increased from 10 to 11, turbidity increased, and apatite became more dispersed. In the hematite-supernatant system, the most dispersed state was observed at pH 8. As pH moved towards higher or lower values, turbidity decreased, meaning that hematite aggregation increased. These results are in agreement with the clarification rates of single mineral systems obtained with the Turbiscan.   101  Figure 4.25 Stability (turbidity) of apatite, hematite, and the 1:1 mixture (total 0.2%vol) in apatite supernatant as a function of pH.  Turbidity is an optical property that results when light is scattered and absorbed by suspended solids in solution rather than transmitted through the sample. The light attenuation coefficient (c) of a suspension which is a measure of the energy removed from a light beam by both scattering and absorption was shown to depend not only on particle concentration but also on other physical properties of the suspended material including size distribution, index of refraction and particle shape. Beam attenuation will be a linear function of the concentration of the suspended matter if the variations in size, shape, and index of refraction are negligible. Similarly, the relationship between turbidity and suspended particles was proven to depend on the physical properties of the suspended material (Baker and Lavelle, 1984; Gipple, 1989; Packman et al., 1999). Using a calibrated instrument, a linear relationship between turbidity and solids concentration occurs if the suspended particles do not differ in their physical characteristics as their concentration changes (Gipple, 1989). In this study, the experiments were performed on pure minerals under constant experimental conditions. Thus, a linear relationship should exist between 020004000600080006 7 8 9 10 11 12Turbidity (NTU)pHApatiteHematite1:1 Mixture  102 turbidity of hematite or apatite suspensions and their concentration. However, since the refractive indices of hematite and apatite are different, 2.87 for hematite and 1.64 for apatite, the turbidity for the mixture may not exactly equal the sum of the values for the single minerals. The mixed systems are 1:1 by volume and only a half of the amount of minerals used in single mineral systems were used in mixtures. Thus, the contribution of each mineral in mixture to the total turbidity value is expected to be approximately a half of their contribution in the single minerals systems.  The results for the mixed mineral systems indicate that at pH 10, and pH 11, turbidity is much higher than the values obtained for individual minerals. At pH 10, the turbidity value for apatite and hematite suspensions is 1650 and 2700 NTU, respectively. The turbidity value for the mixture should not be above 2700 NTU if there was no interaction between apatite and hematite particles, while the mixture has turbidity values of 4200-4500 NTU. The same behaviour was observed at pH 11. However, the difference between the expected and observed values was much more pronounced at pH 11, where the turbidity of the mixture (7000 NTU) was remarkably higher than those of single minerals (3200 NTU for apatite and 2500 NTU for hematite). It should be recalled that turbidity indicates the amount of solid remaining in the suspension. The high turbidity values of mixtures at high pH suggest that there is interaction between these two solids in the mixed system; the mixture is much more dispersed than the turbidity values for the single minerals would suggest. At pH 8 and pH 9, the results for the mixture fall between the results for hematite and apatite single systems. Therefore, the behaviour of the mixture is consistent with the behaviour of the two components. This observation may apparently suggest that interactions between apatite and hematite (hetero-aggregation) are very weak as there are no obvious signs of either enhanced aggregation (very low turbidity) or enhanced dispersion (very high turbidity) in this pH range.  The turbidity results correlate closely with the results from Turbiscan (clarification rate). According to the stability results from Turbiscan and turbidimeter (Figure 4.24 and Figure 4.25), it seems that at pH 10 and pH 11, the dispersion behaviour of minerals in the mixture is not the same as their behaviour in single mineral suspensions, and one or both of the minerals become dispersed when mixed with the other mineral.     103 4.4.2 Zeta potential distribution measurements The zeta potential distributions for 1:1 (by volume) apatite-hematite mixtures along with the zeta potential distributions for single minerals at pH 7 are shown in Figure 4.26.   Figure 4.26 Zeta potential distributions of hematite, apatite, and hematite-apatite mixture in apatite supernatant prepared at solids content of 0.2%vol at pH 7. 020406080100120-60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 2 !Mixture = –25.20020406080100120-60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 1 !Mixture = –26.29020406080100120-60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 4 !Mixture = –26.17020406080100120-60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 5 !Mixture = –28.15020406080100120-60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 3 !Mixture = –28.02  104 The experiments on mixtures were performed by mixing equal volumes of apatite and hematite to match a dispersion concentration of 0.2%vol. The graphs show one type of test conducted five times under the same conditions. The figure shows that at pH 7, the zeta potential distribution of the mixture is monomodal (only one peak is present in the data), and the peak for the mixture overlaps with the peak for hematite alone. This type of distribution can be a result of strong aggregation between hematite and apatite, as discussed in the literature review chapter (Figure 2.2). It can also be a result of dominant presence of hematite in the tested sample. The interpretation of the zeta potential distribution results without knowing the composition of the tested population of particles based only on aggregation phenomenon may not be correct, as other effects such as faster sedimentation of one component are ignored. The results for the direct assay of the tested samples are given in Table 4.3. As can be seen, samples used in the zeta potential measurements were mainly composed of hematite, with between 84% and 95% of hematite by volume.  Table 4.3 The amount of hematite and apatite (%vol) in the samples used for zeta potential distribution measurement experiments at pH 7 and pH 10. Volume of hematite was obtained from iron concentrations using stoichiometric coefficients and the density of hematite (5300 g/L). Similarly, volume of apatite was obtained from calcium concentrations using stoichiometric coefficients and the density of apatite (3200 g/L). PH Test No. Calcium conc. in sample (mg/L) Apatite volume in sample (µl/L) Iron conc. in sample (mg/L) Hematite volume in sample (µl/L) Total volume of minerals in sample (µl/L) Apatite (%vol) Hematite (%vol) pH 7 Test 1 1.53 1.20 32.97 8.89 10.09 11.86 88.14 Test 2 2.50 1.96 42.35 11.42 13.39 14.65 85.35 Test 3 0.89 0.70 47.23 12.74 13.44 5.17 94.83 Test 4 2.54 1.99 38.79 10.47 12.45 15.98 84.02 Test 5 1.61 1.26 51.69 13.94 15.20 8.27 91.73 pH 10 Test 1 2.88 2.26 43.70 11.79 14.05 16.06 83.94 Test 2 2.33 1.82 49.30 13.30 15.12 12.06 87.94 Test 3 1.43 1.12 55.31 14.92 16.04 6.96 93.04 Test 4 3.54 2.77 46.90 12.65 15.43 17.99 82.01 Test 5 1.64 1.28 56.24 15.17 16.45 7.80 92.20    105 Figure 4.27 presents the zeta potential distributions for apatite/hematite mixtures and single minerals at pH 10. The graphs show that the reproducibility of distribution at pH 10 is poorer but still indicate the same trend as at pH 7; The zeta potential distribution for the mixture is still positioned around the zeta potential distribution for hematite. Again, it may apparently point to strong hetero-aggregation between hematite and apatite and suggest that the peak originated from the unsettled apatite-hematite aggregates. But the assay, Table 4.3, showed that the tested sample contained mostly hematite, more than 82%vol. Based on the results for single minerals, both apatite and hematite were expected to aggregate, but as the assays indicate hematite does not aggregate in a mixture with apatite, which seems to explain the unusually high dispersion of a hematite-apatite mixture shown in Figure 4.25: the higher turbidity values at pH 10 and 11 for the mixture originate from dispersed hematite. Therefore, the appearance of a single peak for a mixture overlapping with the peak for hematite originated from the residual hematite but not from unsettled apatite-hematite aggregates.   106  Figure 4.27 Zeta potential distributions of hematite, apatite, and hematite-apatite mixture in apatite supernatant prepared at solids content of 0.2%vol at pH 10. It is noteworthy that a very low solids content is required for zeta potential measurements using the microelectrophoretic technique, therefore after mixing, the suspension was given a long settling time. Based on the assay, the hematite left in the suspension represented only 4% to 5% of the total amount of hematite that was used for the experiment and around 95% of hematite settled. Therefore, the nearly complete disappearing of apatite from the suspension can be a result of either 020406080100120-75 -60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 1!Mixture = –32.55020406080100120-75 -60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 5!Mixture= –35.37020406080100120-75 -60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 2!Mixture = –33.38020406080100120-75 -60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 3!Mixture = –35.25020406080100120-75 -60 -45 -30 -15 0 15Frequency (%)Zeta potential (mV)ApatiteHematiteMixtureTest 4!Mixture = –32.63  107 a long settling time or its selective aggregation. According to the particle size distribution results for apatite and hematite (Figure 3.2), the samples are of comparable particle size, however apatite contains more fines in the very fine fraction, below 1.5 µm, compared to hematite. Furthermore, the density of apatite (3200 g/L) is lower than the density of hematite (5300 g/L). Hence, apatite would remain in suspension if there was no interaction between these two minerals and sedimentation was the only mechanism for the particles to report to the settled fraction.  The results suggest that apatite settles out from the mixture, and hematite remains in the supernatant as a result of selective dispersion. Hence, the results from zeta potential distribution measurements can be used to reinforce the earlier observations from the Turbiscan and turbidimeter; there seems to be an interaction between apatite and hematite at pH 10. Hematite and apatite seem to interact also at pH 7, as all graphs at both pH values present the same trends. The position of the peak for the mixture and its distribution is almost identical in all the tests. Also, the composition of the tested sample was quite reproducible - considering the very small amount of material available for the assay - as 82% to 95% of hematite was consistently detected in the samples. For the apatite/hematite system, aggregation between minerals (hetero-aggregation) cannot be predicted from the zeta potential distribution results for the mixture because almost all apatite and 95% of hematite settled out from suspension before the samples were taken for zeta potential measurements.  4.5 Slime coating in apatite/hematite system 4.5.1 Direct measurement of fine hematite coatings on single apatite crystals The results of analysis of slime coatings on a coarse apatite crystal by hematite fines in background electrolyte and apatite supernatant solutions as a function of pH are presented in Figure 4.28. The data in background electrolyte show that the most extensive coating occurs around pH 7, which is very close to the iso-electric points of both apatite and hematite. Over the pH range from 7 to 8.5, attachment decreases slightly, even though the negative zeta potential of hematite increases considerably (Figure 4.4). As shown in section 4.3.1, the zeta potential of apatite is still very low over this pH range (between –4 mV and –11 mV), and therefore, repulsive electrostatic forces between apatite and hematite apparently are not strong enough to disperse the minerals. Above pH 8.5, the hematite surface charge becomes highly negative and stronger   108 repulsive forces start acting between the minerals even though the zeta potential of apatite becomes only slightly more negative. As a result, the amount of hematite attached to the apatite surface in dilute NaCl decreases as pH is raised above 8.5, and becomes insignificant at pH 10 and pH 11, where hematite has a very high negative zeta potential (–43mV). The aggregation behaviour of minerals (slime coating) in background electrolyte, therefore, follows closely the DLVO theory since dispersion seems to be controlled by electrostatic forces.   Figure 4.28 Hematite slime coating on apatite crystal (milligrams of hematite per gram of apatite crystal) in background electrolyte and apatite supernatant solutions. Supernatants A, B, and C were prepared respectively by mixing 1 g, 3 g, and 5 g of fine apatite in 1 L of 0.01 mol/L NaCl solution.  Figure 4.29 presents some images of an apatite crystal, conditioned with fine hematite in background solution at different pH values. The images show the same trend; very high coverage at pH 7 and 8, much lower coverage at pH 9, and an almost clean crystal at pH 10.  0.00.20.40.60.81.06 7 8 9 10 11 12Attached hematite (mg/g)pH0.01 mol/L NaClSupernatant ASupernatant BSupernatant C  109        Figure 4.29 Images of different apatite crystals (1.2 × 0.6 cm) before and after conditioning with hematite in background electrolyte at different pH values. It is noteworthy that the experiment involved just one apatite crystal in a hematite suspension. It is unlikely that a single crystal is capable of releasing significant concentrations of the constituent ions into a large volume of solution over a short period of time, and thus the concentrations of calcium and phosphate ions in solution were most likely very low during the Clean crystal pH 7pH 8 pH 9pH 10  110 tests performed in dilute sodium chloride as the background electrolyte. Therefore, the aggregation behaviour in supernatant was expected to be different from the behaviour in the background electrolyte. As seen in Figure 4.28, the extent of slime coatings was affected significantly by supernatant solutions, and the change in coating was clearly due to an effect of dissolved ions, i.e., calcium and phosphate. The three supernatants, A, B and C, were prepared by mixing different masses of fine apatite with 1 L of the background NaCl solution for 60 minutes. The masses of fine apatite were 1 g, 3 g, and 5 g, for supernatant A, B, and C, respectively. In the pH range from 7 to 8.5, the mass of fine hematite attached to the apatite crystal decreased when supernatant was used in the experiment, compared to dilute NaCl, but above pH 8.5 the results in supernatant showed higher masses of hematite present on the crystal surfaces than in dilute NaCl. It can be seen that as the concentration of apatite for preparing supernatant increases from supernatant A to C, more ions are released into solution, as shown in Figure 4.3, and the effect of supernatant composition on slime coatings is more pronounced. In supernatant C, the effect of pH disappears, and the mass of hematite attached to apatite is practically constant regardless of pH. It is postulated that at lower pH values, from pH 7 to pH 8.5, the amount of attached hematite decreases in apatite supernatant due to a dispersing effect of increasing amounts of phosphate ions released by apatite, while at higher pH values, above pH 8.5, coating is enhanced as a result of a coagulating effect of higher and higher amounts of calcium. At some points, the dispersing effect of phosphate and coagulating effect of calcium cancel out, and a flat line is observed. This happens in supernatant C whose composition is probably closer to steady-state in terms of apatite dissolution and adsorption of ions on hematite. Steady-state dissolution at a given temperature can be reached either by using a small amount of apatite and conditioning for a longer period of time, or using increasing amounts of apatite over a shorter conditioning time (assuming the same particle size of the apatite particles). Basically, supernatants B and C are most likely closer to steady-state dissolution conditions than supernatant A since larger amounts of the mineral were used to prepare them. Based on the previous results in section 4.2.2 and 4.3.2, the zeta potential of apatite is not affected by its own supernatant to any appreciable extent, but the surface properties of hematite are expected to be affected markedly by these supernatant solutions. The subsequent discussion is based on the assumption that higher masses of apatite used to prepare the supernatants would   111 release higher concentrations of calcium and phosphate ions into the solution (see also Figure 4.3, page 86). Figure 4.30 demonstrates a significant effect of the supernatant solutions on the zeta potential of hematite as a function of pH. It can be seen that at low pH, as concentration of ions in the solution increases from supernatant A to C, hematite becomes more negatively charged due to adsorption of phosphate ions, but at high pH, its negative zeta potential decreases because of adsorption of mainly calcium.   Figure 4.30 Zeta potential of hematite in background electrolyte and in different supernatant solutions as a function of pH. Supernatant A, B, and C were prepared respectively by mixing 1 g, 3 g, and 5 g fine apatite in 1 L background electrolyte.  The data are consistent with the previous results in preceding sections and confirm that the decrease in the mass of fine hematite attached to the apatite crystal below pH 8.5, as the system transitions from dilute NaCl to supernatant C, is related to the effect of increasing concentration of phosphate ions since the adsorption of phosphate on hematite increases the repulsive electrostatic forces between the minerals. At the same time, the increase in hematite attachment to apatite at high pH from supernatant A to C is a result of the coagulating effect of increasing -50-40-30-20-1001020304 5 6 7 8 9 10 11 12Zeta potential (mV) pH0.01 mol/L NaClSupernatant ASupernatant BSupernatant C  112 amounts of calcium cations. The effect of calcium at high pH is two-fold. Firstly, calcium adsorbs on hematite and decreases the repulsive electrostatic forces between apatite and hematite by making hematite less negative. Secondly, the bridging action of calcium can increase aggregation between hematite fines and the apatite crystal. Also, as discussed before, calcium induces aggregation (by a bridging action) between hematite particles, which can lead to the attachment of entire hematite aggregates on the apatite crystal.  To get a better insight into the possibility of attachment of entire hematite aggregates onto the apatite crystal, the appearance of the hematite coating was compared under two sets of conditions (Figure 4.31). The left image in Figure 4.31 shows the hematite coating at pH 9 in background electrolyte, while the right figure shows the hematite coating at pH 11 in supernatant A. In these two cases, the amount of hematite attached to apatite was more or less the same (~ 0.2 mg/g, Figure 4.28), but as can be seen from the pictures, the texture of the coating is quite different. At pH 9 and in background electrolyte, the apatite surface appears to be uniformly coated by fine hematite particles, while at pH 11 in supernatant A, the apatite surface appears to be coated by hematite aggregates.         Figure 4.31 Images of apatite crystals after conditioning with hematite under different conditions. Left image:  at pH 9 and in background electrolyte. Right image: at pH 11 and in supernatant A. Hematite particles should be dispersed in background electrolyte at pH 9 (Figure 4.4), so it is not surprising to find fine particles on the apatite crystal. At pH 11 in supernatant A, hematite should be partly aggregated (Figure 4.9), which correlates with the presence of aggregates on the crystal surface. These qualitative observations strongly suggest that the aggregation/dispersions   113 state of fine hematite particles defines whether the slime coating is formed by individual particles or by larger hematite aggregates.  To confirm the marked dispersing effect of phosphate at lower pH, pH 7, and the strong coagulating effect of calcium at high pH, controlled measurements as a function of calcium and phosphate concentration were performed at pH 7, 10, and 11. The results, given in Figure 4.32, clearly demonstrate that at pH 7 slime coating was affected by phosphate, and decreased notably as the phosphate concentration increased, but higher calcium concentrations did not cause a significant effect at pH 7. At pH 10 and pH 11, increasing phosphate concentrations had no effect on coating while calcium ions considerably increased the amount of hematite slimes on the apatite crystal. The amount of calcium required to result in the same level of coating at pH 10 is lower than the amount of calcium at pH 11, which most likely results from the fact that both minerals are more negatively charged at higher pH and therefore, higher amounts of calcium are needed to reduce the magnitude of the zeta potential.  The results in Figure 4.32 show that in the absence of phosphate, calcium has no noticeable effect on the coating at pH 7. Likewise, phosphate did not affect the coating in the absence of calcium at pH 10 and pH 11. A comparison of the results in supernatant in Figure 4.28 with those in CaCl2 and Na3PO4 solutions in Figure 4.32 shows that at pH 7 the coating was reduced from 0.81 mg to 0.64 mg hematite by adding 1.6 ´ 10–5 mol/L phosphate but it decreased  from 0.81 mg to 0.76 mg hematite in supernatant A in Figure 4.28, which contains the same amount of phosphate (1.6 ´ 10–5 mol/L). Similarly, the attachment of hematite to the phosphate crystal in background NaCl solution in the presence of 2.3 ´ 10–5 and 2.7 ´ 10–5 mol/L phosphate is lower than the attachment in, respectively, supernatant B (containing 2.3 ´ 10–5 phosphate) and supernatant C (containing 2.7 ´ 10–5 mol/L phosphate).   114   Figure 4.32 Hematite slime coating on apatite crystal (milligrams of hematite per grams of apatite crystal) in background electrolyte, and in the presence of calcium and phosphate ions.  0.00.20.40.60.81.00 1 2 3 4 5 6 7 8Attached hematite (mg/g)Ion concentration (×10⁻⁵ mol/L)CalciumPhosphatepH 7Concentration in supernatant AConcentration in supernatant BConcentration in supernatant C0.00.20.40.60.81.00 1 2 3 4 5 6 7 8Attached hematite (mg/g)Ion concentration (×10⁻⁵ mol/L)pH 10-CalciumpH 10-PhosphatepH 11-CalciumpH 11-PhosphatepH 10 and pH 11Concentration insupernatant AConcentration insupernatant BConcentration insupernatant C  115 On the other hand, at high pH, the amount of the hematite coating on the apatite crystal in background NaCl solution in the presence of 3.7 ´ 10–5, 5.6 ´ 10–5, and 6.4 ´ 10–5 mol/L calcium is higher than the mass of coating in, respectively, supernatant A (containing 3.7 ´ 10–5 calcium), supernatant B (containing 5.6 ´ 10–5 mol/L calcium), and supernatant C (containing 6.4 ´ 10–5 mol/L calcium). The dispersing effect of phosphate in background NaCl is weakened in supernatants due to the presence of calcium cations. Analogously, the coagulating effect of calcium is counteracted by the presence of phosphate ions in supernatants. These observations indicate that phosphate adsorption on hematite at pH 7 stimulates calcium adsorption since calcium alone does not affect the zeta potential and aggregation of hematite at that pH. By analogy, at pH 10-11, adsorption of calcium on hematite induces phosphate adsorption as no indications of phosphate interaction with hematite were found from the zeta potential and dispersion/aggregation studies. It can be seen that at pH 10, the most significant change in the amount of hematite attached to apatite is observed at calcium concentrations up to 3.7 × 10–5 mol/L and increasing the concentration from 3.7 × 10–5 mol/L to 8.0 × 10–5 mol/L had a less pronounced effect. Whereas, at pH 11, the attachment increases notably as calcium concentration increases to 6.4 ´ 10–5 mol/L. These results correlate well with the results in Figure 4.28, where the increase in concentration of ions in the solution from supernatant A to C affected slime coating at pH 11 significantly but had only a small effect at pH 10. According to literature, considerable effort has been made into depressing hematite in order to remove iron minerals from phosphate concentrates. A variety of depressants were examined for their depressing actions towards hematite, with the objective of preventing collector adsorption onto hematite rather than inhibiting formation of slime coatings, so the problem of high iron content in concentrates still persists. The role of slime coatings as a possible mechanism responsible for reporting hematite to the concentrate in the flotation separation of apatite and iron oxides is not well-researched and direct evidence is lacking. However, the data in this section indicate that it is possible to minimize slime coatings by pH adjustment when the concentrations of ions in process water are low. This condition may, however, be difficult to achieve under plant conditions because the phosphate content in phosphate ores is relatively high, and long conditioning times are employed for feed preparation ahead of flotation.   116 4.5.2 Interactions between coarse apatite and fine hematite - Turbidity measurements  The results for the turbidity of hematite suspension in the presence and absence of coarse apatite at pH 7, 10, and 11 are given in this section. Figure 4.33 shows the results for the first set of tests (open symbols), where minerals were conditioned in background electrolyte, as well as the second set of tests (closed symbols), where all experiments were conducted in apatite supernatant (prepared in 0.01 mol/L NaCl solution). In both groups of tests, fine hematite (112 mg) was mixed with coarse apatite (0.9 g) in 30 ml background electrolyte (first group of tests) or apatite supernatant (second group of tests) for 60 minutes using an auto-shaker and then the mixture was tested (fine hematite-coarse apatite mixed system). Next, coarse apatite was removed by screening, and the suspension of fine hematite was re-tested for turbidity (hematite alone).  Figure 4.33 Turbidity of hematite suspensions (0.07%vol solids) at pH 7, 10, and 11 for single and mixed systems in 0.01 mol/L NaCl as the background electrolyte (Test Group 1) and apatite supernatant (Test Group 2). H and A stands for hematite and apatite, respectively. Mixed system: mixing fine hematite and coarse apatite (–300 + 150 µm) at desired pH for 60 minutes and measuring the turbidity of the entire suspension of fine hematite with coarse apatite. Hematite alone: conditioning hematite alone for 60 minutes after apatite was removed and then measuring the solution. 02,0004,0006,0008,0006 7 8 9 10 11 12Turbidity (NTU)pHH+A-Test Group 1H-Test Group 1H+A-Test Group 2H-Test Group 2Increase in turbidityof mixture at pH 7 Decrease in turbidity of mixture at pH 10 and 11   117  According to the results, the turbidity of hematite suspensions after removing apatite is relatively low at all pH values, meaning that hematite is at least partly aggregated. As pH is raised from 7 to 11, turbidity slightly decreases which shows that aggregation increases over this pH range. This set of data correlates well with the turbidity values for hematite in supernatant in Figure 4.25 (data points at pH 7, 10, and 11) and follows the same trend. This is due to the fact that in the single mineral tests in Figure 4.25, the testing conditions were somewhat similar to those from Figure 4.33 since mixing hematite in apatite supernatant is equivalent to mixing hematite and apatite together in the same background solution.  The turbidity values for hematite suspensions in the presence and absence of apatite are similar at pH 7, but there is a large difference between those two types of samples at pH 10 and pH 11. It was expected that aggregation of fine hematite onto coarse apatite, in the form of slime coatings, would remove a fraction of fine hematite from suspension, decreasing the solids content in suspensions. As a result, the turbidity of the mixture should be lower than the turbidity of a suspension of fine hematite alone whenever hematite aggregates onto apatite. According to the data in Figure 4.28, in apatite supernatant, hematite attaches to apatite crystals at all pH values. Coarse apatite particles (–300 +150 µm) did not contribute to turbidity because they quickly settled out of suspension (based on turbidity results in Table 3.2) and they were undetectable by the turbidimeter. The results, however, indicate an opposite trend at high pH. At pH 10 and pH 11, when hematite is in combination with apatite, turbidity is much higher compared to hematite alone. The higher turbidity of the mixture could be due to mechanical phenomena induced by the large apatite particles (abrasion, attrition, etc.), and once coarse apatite is removed fine hematite starts re-aggregating. However, the same trend would be observed at all pH values if it was a purely mechanical effect. As seen in Figure 4.33, the results for the single and mixed systems were identical at pH 7. According to the results, hematite and apatite need to be in suspension at the same time for the enhanced dispersion of fine hematite to be observed under the experimental conditions. It is important to note the order of the test: first a suspension of fine hematite with coarse apatite is tested, then apatite is removed, and the suspension is re-tested. In other words, dispersion of fine hematite apparently does not take place when coarse apatite is removed. It should be remembered that when hematite and apatite are mixed together in background electrolyte, apatite releases calcium and phosphate into solution. The effect of apatite on dispersion of hematite   118 is suggested to be related to the equilibrium of ions between these two minerals and the equilibrium seems to be disturbed once apatite is removed. As shown in section 4.2.3, at high pH, calcium produces a coagulating effect on hematite by adsorbing on the surface while phosphate disperses hematite by canceling the effect of calcium ions. Such results strongly suggest that the higher turbidity and stronger dispersion of hematite in the presence of apatite at high pH is a result of lower adsorption of calcium on the hematite surface and/or interaction of phosphate ions with adsorbed calcium ions and neutralizing their coagulating effect (co-adsorption on the hematite surface).  At pH 10 and 11, hematite is completely dispersed in background electrolyte, with turbidity values beyond the measuring range of the instrument. When coarse apatite is added to fine hematite, the turbidity decreases, which suggests that two phenomena took place at the same time. Fine hematite particles formed slime coatings around coarse apatite particles, and hematite particles themselves underwent an aggregation process, as also seen in the tests with large apatite crystals. Both the formation of slime coatings and hematite aggregation can be attributed to increasing calcium levels released by apatite. Once coarse apatite is removed, the turbidity values should no longer be affected by slime coatings, but it is rather clear that aggregation of hematite continues in the absence of apatite since turbidity decreases even further. Similar trends can qualitatively be seen for the experiments in supernatant. However, the hematite particles are much more aggregated in supernatant than in background NaCl, therefore the turbidity of the mixture is quite low. At the same time, once coarse apatite is removed, aggregation of hematite continues and turbidity decreases to the same levels as in background NaCl. The final aggregation state of hematite in both testing scenarios appears to be the same, and the effect of pH under those conditions is much less significant.  The arrows in figure 4.33 indicate relative changes in turbidity of the mixture at the start of the two-test sequence, when tested in background NaCl and in supernatant. At pH 10 and 11, hematite particles are much more dispersed in dilute NaCl than in supernatant, but at pH 7, the situation is reversed: hematite appears to be slightly more dispersed in supernatant than in background NaCl. These results agree with the trends in the amount of attached fine hematite to single apatite crystals (Figure 4.28), and are again consistent with a dispersing effect of phosphate ions at pH 7, and with a coagulating effect of calcium at high pH. Also, whether in single crystal   119 tests or in mixtures of coarse apatite and fine hematite, the effect of pH is very strong in background NaCl, and the role of pH becomes less significant in supernatant. Given enough time, hematite tends to aggregate to the same state regardless of the initial experimental conditions, such as pH or the type of background electrolyte (dilute NaCl versus apatite supernatant). This observation suggests that all the changes in turbidity and aggregation occur as a result of changes in the dissolution degree of apatite with time and the resulting changes in phosphate and calcium levels. Once apatite is removed, hematite continues to aggregate in both experimental systems, most likely as a result of on-going ion adsorption from solution.  4.5.3 Interactions between coarse apatite and fine hematite - Transmission measurements Measuring transmission is another way of monitoring aggregation as a function of time. Hematite-apatite interactions were also investigated using transmission profiles. Transmission is opposite to turbidity; higher transmission indicates a lower amount of solids left in suspension. The sample preparation procedure in these experiments was the same as in turbidity measurements. Two groups of tests were carried out as before, one in background electrolyte and one in supernatant (prepared in 0.01 mol/L NaCl solution). In both group of tests, fine hematite was mixed with coarse apatite in background electrolyte or apatite supernatant for 60 minutes and then the mixture was tested (fine hematite-coarse apatite mixed system). Next, coarse apatite was removed, and the suspension of fine hematite alone was re-tested for turbidity (single hematite system). The transmission results for hematite alone and a mixture with apatite in background electrolyte at pH 7, 10, and 11 are presented in Figure 4.34.  The results for the hematite-apatite mixture show that at pH 7 transmission values are high and little solids is left in suspension, while at high pH (pH 10 and pH 11), the low transmission values suggest that the suspension is more stable. For single mineral suspensions (hematite only), the transmission values are high at all pH values (pH 7, 10, and 11) and show that hematite is aggregated. The graphs generally follow the same trend as in turbidity tests and correlate with the results presented in Figure 4.33. These results also show that after removing coarse apatite, aggregation of hematite continues. At pH 7, hematite is near the iep so van der Waals attraction dominates and therefore, the difference between the hematite-only and hematite-apatite mixed systems is very small. However, at pH 10 and 11 calcium acts as a strong coagulant. These types   120 of aggregation phenomena taking place far from steady-state lead to interesting effects of the coarse apatite content on stability of fine hematite.   Figure 4.34 Stability of hematite suspension, transmission profiles, (0.07%vol solids) at pH 7, 10, and 11 for single and mixed systems in 0.01 mol/L NaCl as the background electrolyte. H and A stands for hematite and apatite, respectively. Mixed system: mixing fine hematite and coarse apatite (–300 +150 µm) at desired pH for 60 minutes and measuring the stability of the entire suspension of fine hematite with coarse apatite. Single hematite system: conditioning hematite alone for 60 minutes after apatite was removed and then measuring the solution. The transmission profiles for hematite only and mixture (fine hematite-coarse apatite) in supernatant at pH 7, pH 10 and pH 11 are given in Figure 4.35. The results in the background electrolyte are also shown in the graphs as the baseline. 0102030405060700 400 800 1200 1600 2000Transmission (%)Time (sec)Fine H+coarse A-pH 7Fine H+coarse A-pH 10Fine H+coarse A-pH 11Fine H-pH 7Fine H-pH 10Fine H-pH 11  121   Figure 4.35 Stability of hematite suspension, transmission profiles, (0.07%vol solids) at pH 7, 10, and 11 for single and mixed systems in apatite supernatant (prepared in 0.01 mol/L NaCl). The results for stability of hematite in background electrolyte (0.01 mol/L NaCl) are also shown in the graphs. H and A stands for hematite and apatite, respectively.  As can be seen, the results correlate very well with the turbidity results, showing the higher dispersion of hematite in the mixture of fine hematite and coarse apatite. At pH 7, the result in background electrolyte showed high transmission values and hematite aggregated, as pH 7 is very close to the iep of hematite. In supernatant, when hematite was conditioned alone at pH 7, transmission values decreased slightly indicating stronger dispersion due to the adsorption of mainly phosphate and to a lesser extent calcium. These results are consistent with the results in 0102030405060700 500 1000 1500 2000Transmission (%)Time (sec)Fine H-0.01 mol/L NaClFine H+coarse A-supernatantFine H-supernatantpH 70102030405060700 500 1000 1500 2000Transmission (%)Time (sec)Fine H-0.01 mol/L NaClFine H+coarse A-supernatantFine H-supernatantpH 100102030405060700 500 1000 1500 2000Transmission (%)Time (sec)Fine H-0.01 mol/L NaClFine H+coarse A-supernatantFine H-supernatantpH 11  122 section 4.2.2. The stability of hematite suspension at pH 7 was greater when apatite was present in suspension, probably because apatite provided more phosphate for adsorption on hematite. At pH 10 and pH 11, hematite is totally dispersed in background electrolyte producing zero transmission, most likely by electrostatic repulsion resulting from the highly negative zeta potential value (–43 mV) in alkaline solutions. When hematite is mixed with apatite in supernatant, transmission of suspensions increases markedly from zero in NaCl to 40% due to coagulation effect of calcium. Once coarse apatite is removed from mixture, transmission increases from 40% to around 60%, indicating again that aggregation of hematite continues after removing apatite due to adsorption of ions from solution. 4.6 The effect of apatite on stability of hematite in alkaline solutions Since the flotation operation of phosphate ores are routinely performed under alkaline conditions, the experiments in this section were focused on the behaviour of hematite in the presence and absence of apatite at high pH, (pH 10).  4.6.1 The effect of apatite to hematite ratio on stability and electrokinetic properties of hematite The results for the stability of hematite suspension conditioned in the absence and presence of different amounts of coarse apatite at pH 10 are given in Figure 4.36. In order to determine changes in electrokinetic properties of hematite caused by ion transfer from apatite, and to explore the correlation between the dispersion behaviour of hematite and its electrokinetic properties, the hematite zeta potential was also measured. The results for zeta potential measurements were also added to Figure 4.36. It can be seen that as the apatite mass increases, the dispersion of hematite improves notably which confirms that there is an interaction between these two minerals. For 2 g and 4 g apatite, the curves are identical, and no further dispersion was observed. It should also be noted that transmission levels of zero, as observed in background electrolyte at pH 10, were not reached showing that only partial re-dispersion was possible by adding higher and higher amounts of apatite.  As discussed, the higher dispersion of hematite in the presence of apatite is suggested to be due to lower adsorption of calcium on the hematite surface and/or co-adsorption of phosphate   123 ions on the hematite surface. The fact that the zeta potential and stability of hematite do not change by increasing the amount of apatite from 2 g to 4 g suggest that there is a limit to what apatite can release and, therefore, what hematite can simultaneously adsorb.  Figure 4.36 Stability of hematite, transmission profiles, (0.07%vol solids) conditioned in the presence of different amount of coarse apatite (–300 +150 µm) at pH 10. All tests were conducted with apatite supernatant. By addition of more apatite, hematite was expected to become more negatively charged due to ion transfer (adsorption of phosphate). The increase in the magnitude of the negative zeta potential value of hematite by addition of apatite was within a narrow margin of only 2-3 mV. It is very important to note that for zeta potential measurements, the sample was taken from the top layer of suspension which contained only fine hematite particles and the zeta potential measurements were performed on the hematite particles in the absence of apatite. The dispersing effect of apatite on hematite was observed only when hematite and apatite were in contact and once apatite was removed, the dispersing effect ended. Accordingly, it is also possible that the 0102030405060700 500 1000 1500 2000Transmission (%)Time (sec)0.0 g Apatite0.7 g Apatite2.0 g Apatite4.0 g ApatitepH 10 !Ave = -31.7!Ave = -32.8!Ave = -34.9!Ave = -34.7  124 effect of apatite on the zeta potential of hematite is not so pronounced once apatite is removed from system. When the samples for zeta potential measurements are taken from the suspension, coarse apatite is separated from hematite and therefore the calcium ions present in the sample solution may adsorb onto the hematite surface, as discussed earlier. To be able to measure the effect of apatite on electrokinetic characteristics of hematite, zeta potential measurements of hematite need to be carried out in the presence of apatite. Since only samples with fine particles can be used in the ZetaView meter, the effect of apatite on hematite surface charge can be determined only by measuring the zeta potential of fine apatite-fine hematite mixtures. However, in these measurements, the apatite particles would also be measured along with the hematite particles, and could obscure the zeta potential results of hematite. However, the zeta potential results for mixtures at pH 10, given in Figure 4.27, can be used for examining the effect of apatite on the zeta potential of hematite, as the direct assay of the tested samples showed that those mixtures contained 82% to 95% of hematite and 5% to 18% apatite. This means that apatite and hematite were still in contact during measurement but the contribution of apatite to the zeta potential value should not be significant. The zeta potential distributions on mixtures and the direct assay of the tested samples showed that hematite was the dispersed mineral, remaining in suspension, and therefore the data directly relate to the results in Figure 4.36. The average zeta potential values for hematite in the mixtures (Figure 4.27), ranging from –32.6 mV to –35.4 mV, are comparable with the values in Figure 4.36. 4.6.2 The composition of solutions in the presence and absence of apatite and hematite  It was shown that 2 g and 4 g apatite caused the highest level of hematite dispersion. It seems that there is a limit to re-dispersion of hematite by apatite, as increasing the mass of apatite from 2 g to 4 g had no notable effect on the hematite zeta potential and dispersion. To measure the concentration of ions in equilibrium with the minerals, water analysis was performed for the supernatant in the presence and absence of fine hematite and coarse apatite (2 g and 4g) at pH 10. The supernatant was prepared at natural pH (pH 6.8), then mineral (minerals) was mixed in supernatant and the suspension was conditioned at pH 10 for 60 minutes. The difference between the pH of solution in preparing supernatant (pH 6.8) and the solution pH in conditioning minerals in supernatant (pH 10) may cause a change in concentration of ions. Therefore, apatite supernatant prepared at natural pH was conditioned at pH 10 in the absence of minerals and was assayed for   125 ions. Also, to determine the effect of pH on the concentration of ions released by apatite, apatite supernatants prepared at pH 10 by fine and coarse apatite were assayed. The results are given in Table 4.4. Table 4.4 The composition of apatite supernatant conditioned in the presence and absence of fine hematite and coarse apatite at pH 10. Solution Number Sample Solution pH while conditioning minerals Mixing time (minutes) Total calcium (mol/L) Total phosphate (mol/L) Hematite zeta potential (mV) 1 Fine apatite supernatant prepared at natural pH (pH 6.8) - 60  3.7 × 10 –5 1.6 × 10 –5 - 2 Fine apatite supernatant prepared at natural pH, then conditioned at pH 10 - 60  3.7 × 10 –5 1.6 × 10 –5 - 3 Fine hematite in supernatant prepared at natural pH 10 60  1.9 × 10 –5 1.4 × 10 –5 –31.7 4 2 g coarse apatite in supernatant prepared at natural pH 10 60  3.0 × 10 –5 2.1 × 10 –5 - 5 4 g coarse apatite in supernatant prepared at natural pH 10 60  3.2 × 10 –5 2.2 × 10 –5 - 6 Hematite+2g coarse apatite in supernatant prepared at natural pH 10 60  2.0 × 10 –5 2.0 × 10 –5 –34.9 7 Hematite+4g coarse apatite in supernatant prepared at natural pH 10 60  2.1 × 10 –5 2.2 × 10 –5 –34.7 8 Fine apatite supernatant prepared at pH 10 - 60  0.4 × 10 –5 0.6 × 10 –5 - 9 Coarse apatite (2 g) supernatant prepared at pH 10 - 60  0.3 × 10 –5 0.5 × 10 –5 - 10 Coarse apatite (4 g) supernatant prepared at pH 10 - 60  0.9 × 10 –5 0.9 × 10 –5 -     126 The results for solutions 1 and 2 indicate that the concentration of ions does not change as pH of solution increases from 6.8 to 10, though the speciation of ions changes. It can be seen that when hematite was added to supernatant (solution 2), calcium concentration decreased notably, changing from 3.7 × 10–5 to 1.9 × 10–5 mol/L, while the decrease in phosphate concentration was minor, changing from 1.6 × 10–5 to 1.4 × 10–5 mol/L. The change in concentrations of ions due to addition of hematite to supernatant is a result of adsorption of those ions on hematite. The adsorption of calcium on hematite is much higher than the adsorption of phosphate. The adsorption results support the zeta potential and stability data from sections 4.2.2 and 4.2.3 and confirm that at high pH, calcium has a higher affinity towards hematite than phosphate and preferentially adsorbs over phosphate. These data also support the idea proposed earlier that at high pH, the high specific adsorption of calcium proceeds into the inner Helmholtz plane and low adsorption of phosphate occurs in the outer Helmholtz plane.  Comparing the results for solution 2 (supernatant conditioned at pH 10) with those for solution 4 and 5, by adding 2 g or 4 g of coarse apatite to supernatant, calcium concentration decreased while the concentration of phosphate increased. These changes in ion concentration show that calcium also adsorbs on apatite (not only on hematite), but phosphate is released by the mineral rather than being adsorbed. These data indicate that both coarse apatite and fine hematite should adsorb calcium from supernatant, but only hematite should adsorb phosphate while apatite will tend to release phosphate into solution. This hypothesis was verified by adding 2 g (or 4 g) of coarse apatite and fine hematite into supernatant (solutions 6 and 7). As a result, the concentration of calcium decreased considerably from 3.7 × 10–5 mol/L to around 2.0 × 10–5 mol/L (or 2.1 × 10–5 mol/L). The phosphate concentration, however, simultaneously increased under these conditions from 1.6 × 10–5 mol/L to 2.0 × 10–5 mol/L (or 2.2 × 10–5 mol/L).  Because calcium adsorbs on both apatite and hematite, the amount of calcium adsorbed on hematite in a mixture is not as high as the amount of calcium adsorbed on hematite from supernatant in the absence of apatite. In the former case, the difference in calcium concentration is only about 1 × 10-5 mol/L (solutions 4 and 6, and 5 and 7), while in the latter case the difference is about 1.8 × 10-5 mol/L (solutions 2 and 3). Simultaneously, the concentration of phosphate does not measurably change. The adsorption of phosphate on hematite cannot be clearly detected by monitoring concentrations because phosphate adsorption on hematite at pH 10 is facilitated by   127 calcium adsorption. With lower adsorption of calcium, changes in phosphate concentration would also be much smaller. For example, in solutions 2 and 3, a change in calcium concentration by 1.8 × 10–5 mol/L was accompanied by a change in phosphate concentration by only 0.2 × 10–5 mol/L, or nine times less. In solutions 5 and 7, calcium concentration changes by 1.1 × 10–5 mol/L, which would suggest that phosphate concentration should change by about 0.1 × 10–5 mol/L. Such a small change is probably equal to the experimental error, as seen from the range of nearly equal total phosphate concentrations found in solutions 4 through 7. However, it is also possible that in the mixture of coarse apatite and fine hematite, co-adsorption of phosphate on hematite increases, as apatite provides a higher phosphate environment compared to the solution 2 (supernatant in the absence of apatite). Since apatite continuously resupply phosphate to counteract the adsorption on hematite, the change in concentration cannot be detected by measuring the concentrations. Therefore, the zeta potential of hematite becomes more negative in the presence of apatite (–31.7 mV in the absence of apatite versus – 34.9 mV and –34.7 mV in the presence of 2 g and 4 g apatite) due to lower calcium adsorption and/or higher co-adsorption of phosphate. The concentrations of ions in the presence of 2 g and 4 g of apatite (solution 4 and 5), are nearly identical. Both solutions contain about 3.0 × 10–5 mol/L calcium and 2.0 × 10–5 mol/L phosphate, meaning that both amounts (2 g and 4 g) tend to produce effectively the same type of supernatant. These data explain why the same zeta potential and stability results were produced in the presence of 2 g and 4 g coarse apatite. As can be seen, the concentration of ions in the supernatant prepared at natural pH (solution 1) is much higher than the concentrations in the supernatant prepared at pH 10 (solution 8). Also, the results demonstrate that at high pH, phosphate is preferentially released compared to calcium whereas at natural pH (6.8), apatite introduces more calcium than phosphate. Solution 9 and solution 10 were prepared by respectively 2 g and 4 g coarse apatite at pH 10. With increasing the amount of apatite from 2 g to 4 g, the concentration of both ions (calcium and phosphate) increased but the amount of calcium increased more and a solution with the same concentration of ions (1:1) was obtained. As can be seen from compositions of solutions 6 and 7, the molar ratio of calcium to phosphate is close to 1, and this ratio is nearly equal to the ratio of calcium to phosphate in solution 10, which is essentially a supernatant prepared at pH 10. The concentrations in solution 10 are much lower than in solutions 6 and 7 since longer total conditioning times were used in   128 those tests (60 minutes of mixing fine apatite plus 60 minutes of mixing coarse apatite with hematite) compared to a shorter time to prepare solution/supernatant 10 (60 minutes of mixing coarse apatite). It seems that the presence of apatite allows this ratio to be maintained. However, after removing apatite, this ratio would be no longer expected to remain constant. 4.6.3 The reversibility of dispersing effect of apatite on hematite In this section, apatite was added and removed to/from hematite suspension a few times to examine how the addition and removal of coarse apatite affect stability and electrokinetic properties of hematite and determine the reversibility of the dispersion/aggregation phenomena. These cycled experiments can also be viewed as kinetic experiments in which ion concentrations in solution change as a result of repeated apatite addition. First, hematite stability in background electrolyte at pH 10 was measured. For the second experiment, coarse apatite was added to the suspension and the mixture was conditioned for 60 minutes and then measured. In the third experiment, apatite was removed from the suspension, and suspension was conditioned again for 60 minutes before the measurement. Then, two more measurements were performed on the same sample as the fourth and fifth experiments after adding and then removing apatite to close the cycle. The transmission results for these five stages of experiments at pH 10 are given in Figure 4.37. The results for zeta potential measurements were also added to the figure.  The data indicate that hematite in background electrolyte was fully dispersed, as expected. In the second test, when hematite was mixed with apatite, the hematite stability and the magnitude of its zeta potential decreased slightly. Once apatite was removed, hematite aggregation was significantly enhanced, and the transmission values increased (“test 3” in the legend) indicating that the concentration of solids decreased as a result of aggregation and settling. The magnitude of the hematite zeta potential was also affected under these conditions. By re-introducing apatite in the next step (test 4), now the dispersion of suspension and the magnitude of zeta potential increased, but transmission values were not as low as in the second stage. In the last stage, when apatite was removed, hematite still again aggregated, and transmission values increased.    129  Figure 4.37 Stability (transmission profiles) and zeta potential of hematite (0.07%vol solids) conditioned in the presence and absence of 0.9 g coarse apatite (–300 +150 µm) at pH 10. H and A stands for hematite and apatite, respectively. When hematite and apatite are mixed together in the background electrolyte, apatite starts releasing ions into solution to satisfy its solubility product. As discussed in preceding section, at pH 10, apatite prevents the high adsorption of calcium ions on hematite because apatite also tends to adsorb calcium. At the same time, apatite releases more phosphate which probably results in slightly higher adsorption of phosphate on hematite. As a result, the small amount of calcium adsorbed on hematite is sufficient to only slightly aggregate hematite, an effect that is to some degree counteracted by the small amount of phosphate ions from apatite.  The large change in the state of aggregation and in the magnitude of the zeta potential of the fine hematite particles, as the experiment moves from test 2 to test 3, suggest that the adsorption of calcium cations on hematite is high. Basically, when apatite is removed, only hematite interacts with calcium ions in solution and continues adsorbing calcium. The change in the zeta potential is equivalent to the change produced by 2 × 10–5 mol/L of Ca2+ (Figure 4.12), which is also very close to the overall effect of supernatant (or calcium plus phosphate). Re-introduction of apatite (test 4) 0102030405060700 400 800 1200 1600 2000 2400Transmission (%)Time (sec)Test 1:Test 2:Test 3:Test 4:Test 5:!Ave = -32.5!Ave = -33.5!Ave = -38.1!Ave = -42.2!Ave = -32.7H in background electrolyteIntroducing coarse AH in supernatant with A present Screening out AH in supernatantRe-introducing AH in supernatant with A present Screening out AH in supernatant  130 results in weak re-dispersion of hematite and the zeta potential simultaneously becomes slightly more negative. This net-dispersing effect of solid apatite is most likely a result of continued apatite dissolution and a release of phosphate ions into solution. Although additional calcium cations are also released, the presence of apatite decreases the level of calcium adsorption on hematite. Also, the high adsorption density of calcium in the previous stage appears to decrease the calcium adsorption rate, while the specific nature of the interaction between calcium cations and the hematite surface make calcium desorption highly unlikely. It should be noted that higher calcium levels can make the zeta potential of hematite even less negative, as seen in the data in Figure 4.12, but as Figure 4.37 shows, additional calcium ions at this stage of the test do not decrease the magnitude of the zeta potential. When apatite is removed one more time (test 5), aggregation of hematite can be observed although a limit is apparently reached in terms of changes in the zeta potential and in the transmission values. This aggregation step is again a result of continued calcium adsorption although the amount adsorbed is most likely very low since the greatest changes in adsorption on hematite, as judged by changes in the zeta potential, occurred in the early stages of the experiment. Although hematite appears to undergo a cycle of aggregation-dispersion phenomena in the absence and presence of apatite, the final result is an aggregated state, showing that long-term (or repeated) exposure of fine hematite to solid apatite leads to hematite aggregation. Such partly reversible aggregation-dispersion processes take place in early stages of the test, which suggests that it is the departure from steady-state dissolution of apatite (in terms of changes in calcium and phosphate ion concentrations) that allows those processes to take place. Once the hematite-apatite system approaches steady-state, the net result is irreversible hematite aggregation, and only calcium and phosphate ions from sources other than apatite could potentially change this balance. The clarification rate as a function of zeta potential was plotted based on the results in Figure 4.36, Figure 4.37, and section 4.2.3 and is shown in Figure 4.38. All the results for the reversibility tests, and those for adding different masses of apatite fall on the curve which shows the effect of specifically adsorbing ions on the zeta potential and clarification rate. It can be seen that even a small change of 2-3 mV in zeta potential values is sufficient to measurably disperse or aggregate hematite. Basically, the graph indicates that the range of values obtained for zeta potential and clarification rate by reversibility tests (Figure 4.37) are in the same range of values   131 produced by adding more apatite (Figure 4.36). Accordingly, the reversibility tests must produce the same ions concentration levels as the levels observed by adding more apatite (Table 4.4). According to the graph, once apatite is introduced into the system, the effect on the zeta potential and stability of hematite is highly consistent with increasing phosphate concentration, while by removing apatite an effect consistent with increasing calcium concentration is produced. It can be suggested that apatite can disperse hematite by decreasing calcium adsorption and increasing co-adsorption of phosphate on hematite. When apatite is removed, now calcium ions continue adsorbing on hematite and increase hematite aggregation.   Figure 4.38 The stability (clarification rate) of hematite (0.07%vol solids) as a function of zeta potential at pH 10 in the presence of Triangle: calcium at fixed amount of phosphate, Diamond: phosphate at fixed amount of calcium, Star: coarse apatite, and as a result of adding and removing Circle: 1.5 g coarse apatite. 4.6.4 The effect of apatite-hematite mixing time on the stability of hematite Keeping a constant amount of apatite with hematite in the solution for a long time would be expected to produce the same effect as adding more apatite into system in stages. To investigate 020406080100-45 -40 -35 -30 -25Clarification rate (mm/hr)Zeta potential (mV)Adding Ca-at fixed amount of PAdding P-at fixed amount of CaAdding apatiteReversibility testsAdding apatiteRemoving apatite  132 the effect of mixing time, fine hematite (0.07%vol solids) and a fixed amount of coarse apatite were conditioned at pH 10 and the stability was measured as a function of the conditioning time (60, 120, and 180 minutes). In the final stage, apatite was removed and hematite suspension was conditioned for a further 60 minutes before the measurement. The results for these experiments at pH 10 are given in Figure 4.39. Hematite in background electrolyte is fully dispersed. When hematite is mixed with apatite for 60 minutes, the hematite stability decreases marginally. As the mixing time increases from 60 minutes to 120 and 180 minutes, the hematite aggregation increases gradually, indicating that the concentration of ions in the solution and therefore their adsorption on hematite increases with time. The net effect is continuing towards aggregation because the coagulating effect of calcium overcomes the dispersing effect of phosphate due to its specific adsorption. Once apatite is removed in the final step, a significant shift towards aggregation is observed.   Figure 4.39 Stability (transmission profiles) of hematite (0.07%vol solids) as a function of time in the presence and absence of 0.9 g coarse apatite (–300 +150 µm) at pH 10. H and A stands for hematite and apatite, respectively. 0102030405060700 500 1000 1500 2000Transmission (%)Time (sec)Fine H in 0.01 mol/L NaCl-1 h mixingFine H+coarse A-total 1 h mixingFine H+coarse A-total 2 h mixingFine H+coarse A-total 3 h mixingFine H after removing coarse A  133 The trend suggests that aggregation of hematite continues with prolonged mixing, most likely as a result of continuous calcium adsorption. However, even after a long period of mixing, aggregation of hematite in the presence of apatite is not as advanced as the data from Figure 4.37 would suggest (test 4). When apatite is present, the extent of calcium adsorption on hematite is slowed down, as apatite competes with hematite for calcium. Also, in the presence of hematite, more phosphate is available to adsorb on hematite. When coarse apatite is removed, hematite becomes strongly aggregated in the same fashion as in Figure 4.37. This sudden change in the aggregation state is equivalent to continued calcium adsorption on hematite. It is noteworthy that the transmission values can also be affected by aggregation between fine hematite and apatite crystals. However, according to Figure 4.28, the amount of hematite attached to the apatite surface is the same in all three supernatants at pH 10. In other words, the extent of aggregation between these two minerals is a constant factor regardless of the composition of the supernatant. Thus, the amount of apatite aggregated with hematite should be theoretically the same in all stages.  These two sets of tests, where apatite is added or removed a few times and where hematite and apatite are mixed for different periods of time, are comparable to kinetic experiments. In both cases, the test is started with no ions in solution. By whether adding and removing apatite or mixing the hematite-apatite mixture for a longer period of time, apatite is basically allowed more time to release more ions into solution. Consequently, the concentration of ions increases and the solution gets closer to the steady-state condition in terms of apatite dissolution and, therefore, the final results are identical. These results point again towards the possibility of controlling slime coatings phenomena by pH adjustment when hematite-apatite mixture is far from dissolution equilibrium. Slime coatings will not take place at high pH to any considerable extent if the conditioning time is short. However, for a solution that is at steady-state condition with respect to apatite dissolution, slime coatings will form regardless of pH, and adjusting pH will have little effect.    134 Chapter 5: Summary and Conclusions Constituent ions released by apatite into the supernatant solution have a strong effect on the surface charge and aggregation characteristics of hematite. Phosphate ions show high affinity towards the hematite surface at pH values lower than 9. Generally, the various phosphate species are attracted towards the positively charged hematite surface below the iso-electric point of the mineral (pH 6.8) although specific interactions with the negatively charged hematite surface were also seen under weakly alkaline conditions (pH 6.8 ~ 9). No evidence was found of phosphate interactions with hematite under highly alkaline conditions, pH 10-11, or at pH values much higher than the iep of the mineral. In contrast, the calcium cation showed high affinity towards the hematite surfaces under alkaline conditions, or above the iep value, while no interactions were detected below the iep.  In the case of apatite, although only a minor effect of the constituent ions in supernatant was observed on the zeta potential of the mineral, measurable calcium adsorption at high pH was detected. Apatite was found to strongly aggregate at pH values near the iso-electric point, when the zeta potential was almost zero or slightly negative (-12 mV or so), while strong dispersion was found at higher pH values, when the surface became more negatively charged (-20 mV or lower). Attachment of fine hematite to large apatite crystals, analogous to formation of slime coatings, was found to be a strong function of pH in dilute sodium chloride as a background electrolyte, or when the constituent ion concentrations were low. Extensive coating took place in the pH range from 7 to 9, and no coatings formed at pH 10-11. As the constituent ion concentrations increased in different types of supernatant, the effect of pH disappeared and the amount of fine hematite on the crystal surface was practically constant. These results indicated that at lower pH values (pH 7-9), where fine hematite coatings were extensive in background electrolyte, phosphate ions from the supernatant acted as a dispersing agent gradually decreasing the amount of fine hematite attached to the apatite crystal. An opposite trend was found under alkaline conditions, where no hematite coatings formed in background electrolyte. The amount of fine hematite gradually increased with increasing constituent ion concentrations in the supernatant solution, and this overall change was attributed to a coagulating effect of calcium.    135 Based on zeta potential and aggregations studies on model hematite, co-adsorption of phosphate and calcium ions on the hematite surface was also evident, especially under conditions when the adsorption of a given ion (calcium or phosphate) would normally not take place (e.g., calcium adsorption at pH 7 or phosphate adsorption at pH 10). For example, no effect of phosphate ions on the zeta potential of hematite was found at pH 10, but an effect of phosphate on the zeta potential of hematite was measurable in the presence of calcium. Based on the magnitude of the co-adsorption phenomena and their effect on the zeta potential, it was proposed that calcium cations at pH 10-11 adsorbed within the inner Helmholtz plane of the electrical double layer around the hematite particles, while phosphate co-adsorption proceeded onto the outer Helmholtz plane. Although calcium adsorption at high pH invariably resulted in coagulation and aggregation of fine hematite, the co-adsorption of phosphate led to partial dispersion of the fine hematite particles. The dispersing effect of phosphate ions was proportional to their concentration, so at sufficiently high phosphate concentrations, the full dispersion of hematite could be restored. The combined aggregation and zeta potential results also showed that aggregation or dispersion of hematite is generally a function of the magnitude of the zeta potential. However, at higher calcium concentrations and at high pH, the aggregation of hematite was enhanced and did not quite follow the DLVO model, suggesting that a factor other than electrostatic repulsion and van der Waals attraction played a role in hematite aggregation. This departure from the purely electrostatic arguments behind aggregation was attributed to the “bridging effect” of calcium ions, in which calcium was capable of bridging hematite particles together despite negative zeta potential values on individual particles. Measurements of solubility of apatite as a function of time showed that the mineral continuously dissolved in background electrolyte over a period of 48 hours, which meant that all the experiments involving mixtures of apatite and hematite were performed under non-equilibrium conditions in terms of constituent ion concentrations in the solution phase. As a result of this departure from steady-state, several previously undescribed phenomena were reported. It was generally found that the addition of apatite to a hematite suspension resulted in hematite dispersion. At the same time, removal of apatite from a hematite suspension led to strong coagulation of hematite. The apparent dispersion of hematite by apatite was attributed to a tendency of apatite to release phosphate ions into solution at higher pH and to adsorption of calcium on the added apatite   136 particles. Under these conditions, the amount of calcium adsorbed on hematite was lowered, which in combination with the higher levels of phosphate ions in solution (available for co-adsorption) led to hematite dispersion. Removal of apatite from a suspension directed calcium adsorption onto hematite only, thus enhancing the aggregation of the iron oxide. Although varying amounts of phosphate ions were always present in these solutions, the data clearly showed that the coagulating effect of calcium was generally stronger than the dispersing effect of phosphate in the ion concentration range naturally found in apatite supernatants. These aggregation-dispersion transitions became less pronounced when either larger amounts of apatite were used in the tests or longer mixing times of apatite-hematite suspensions were employed. In other words, such phenomena took place under conditions that were far from steady-state in terms of apatite dissolution and ion adsorption processes. Once these systems approached steady-state, hematite generally aggregated in apatite supernatants, compared to full dispersion in the background electrolyte and in the absence of calcium and phosphate ions, which again confirmed the stronger coagulating power of calcium over the dispersing abilities of phosphate ions. All the fundamental data also indicated that slime coating phenomena in the apatite-hematite flotation system can be controlled by simple pH adjustments as long as the conditioning times are relatively short and the mineral system is far from steady-state conditions. At typical pH values of 10-11 in flotation, hematite should not form extensive slime coatings at low levels of constituent ions in the flotation pulp, but prolonged mixing and gradual dissolution of the salt-type mineral will release calcium and phosphate ions into the pulp, and under these conditions the formation of slime coatings should be enhanced. In this case, pH control will be insufficient, and the use of a dedicated dispersant will most likely be necessary.           137 Chapter 6: Recommendations for Future Work The direct measurement of slime coating on a single crystal can be conducted by in-situ imaging using a powerful high-resolution camera and subsequent image analysis. This method would eliminate any possible artifacts caused by removal of the coarse particle from solution and its handling. Also, this high-resolution imaging can be used to evaluate the textures of hematite on the surface. The current study was performed using a fixed amount of fine hematite. The mass of fine mineral would be expected to have an effect on the extent of slime coating. Also, the effect of different operational parameters (such as mixing rate) and chemical factors (ionic species and dispersants) can be examined.  In order to discover a correlation between the stability results and surface forces, direct measurement of surface forces using colloid probe atomic force microscopy (AFM) can be used to measure the force of interaction between a pair of colloids (apatite and hematite) in aqueous solutions in the presence and absence of dissolved ions and dispersants at different pH values. Flotation experiments can be performed on the different ratios of apatite-hematite mixture for different periods of time to explore how interactions between hematite and apatite and apatite dissolution kinetic affect flotation performance. Different dispersants can be evaluated with respect to flotation performance in terms of grade and recovery but more importantly, the effect of those dispersants on the actual phenomena of slime coatings and entrainment in the concentrate should be examined. Adsorption and co-adsorption of ions on minerals should be measured. This is a very difficult measurement because the system is not at equilibrium and the experimental conditions such as mixing time, solid content in suspension, temperature, and pH will affect the data.   138 Bibliography Abdus-Salam, N., Adekola, F.A., 2006. Comparative dissolution of natural goethite samples in HCl and HNO₃. J. Appl. Sci. Environ. Manag. 10, 11–17. Abouzeid, A.Z.M., 2008. Physical and thermal treatment of phosphate ores - An overview. Int. J. Miner. Process. 85, 59–84. Abouzeid, A.Z.M., Negm, A.T., Elgillani, D.A., 2009. Upgrading of calcareous phosphate ores by flotation: Effect of ore characteristics. Int. J. Miner. Process. 90, 81–89. Addai-Mensah, J., Prestidge, C.A., 2005. Structure formation in dispersed systems, in: Stechemesser, H., Dobiáš, B. (Eds.), Coagulation and Flocculation. Taylor & Francis, pp. 135–216. Agey, W.W., Salisbury, H.B., Pacek, P.L., 1963. Effect of ions common to natural water on flotation of limestone. Amankonah, J.O., Somasundaran, P., 1985. Effects of dissolved mineral species on the electrokinetic behavior of calcite and apatite. Colloids and Surfaces 15, 335–353. Amirtharajah, A., Mills, K.M., 1982. Rapid-mix design for mechanisms of alum coagulation. J. Am. Water Work. Assoc. 74, 210–216. Antelo, J., Avena, M., Fiol, S., López, R., Arce, F., 2005. Effects of pH and ionic strength on the adsorption of phosphate and arsenate at the goethite – water interface. J. Colloid Interface Sci. 285, 476–486. Antelo, J., Fiol, S., Pérez, C., Mariño, S., Arce, F., Gondar, D., López, R., 2010. Analysis of phosphate adsorption onto ferrihydrite using the CD-MUSIC model. J. Colloid Interface Sci. 347, 112–119. Aplan, F.A., Fuerstenau, D.W., 1962. Principles of nonmetallic mineral flotation, in: Fuerstenau, D.W. (Ed.), Froth Flotation. AIME, 50th Ann, pp. 170–178. Arai, Y., Sparks, D.L., 2001. ATR – FTIR spectroscopic investigation on phosphate adsorption mechanisms at the ferrihydrite – water interface. J. Colloid Interface Sci. 241, 317–326. Assis, S.M., Montenegro, L.C.M., Peres, A.E.C., 1996. Utilisation of hydroxamates in minerals froth flotation. Miner. Eng. 9, 103–114. Atesok, G., Somasundaran, P., Morgan, L.J., 1988. Adsorption properties of Ca2+ on Na-kaolinite and its effect on flocculation using polyacrylamides. Colloids and Surfaces 32, 127–138. Atkinson, R.J., Posner, A.M., Quirk, J.P., 1967. Adsorption of potential-determining ions at the ferric oxide-aqueous electrolyte interface. J. Phys. Chem. 71, 550–558. Baker, E.T., Lavelle, J.W., 1984. The effect of particle size on the light attenuation coefficient of natural suspensions. J. Geophys. Res. 89, 8197–8203. Bandini, P., Prestidge, C.A., Ralston, J., 2001. Colloidal iron oxide slime coatings and galena particle flotation. Miner. Eng. 14, 487–497. Barbour, A.P., 1973. Distribution of phosphorus in the iron ores deposits of Itabira, Minas Gerais, Brazil. Econ. Geol. 68, 52–64. Bell, L.C., Posner, A.M., Quirk, J.P., 1973. The point of zero charge of hydroxyapatite and fluorapatite in aqueous solutions. J. Colloid Interface Sci. 42, 250–261. Bellmann, C., 2004. Stability of dispersions. Chem. Eng. Technol. 27, 937–942. Belton, G.R., 1985. Report on the beneficiation study of Mt. Weld phosphate. Wallsend, NSW. Besteman, K., Zevenbergen, M.A.G., Heering, H.A., Lemay, S.G., 2004. Direct observation of charge inversion by multivalent ions as a universal electrostatic phenomenon. Phys. Rev. Lett. 93.   139 Bijsterbosch, B., 1987. Stability of solid/liquid dispersions, in: Tadros, T.F. (Ed.), Solid/Liquid Dispersions. Academic Press, London, pp. 91–109. Bolto, B., Gregory, J., 2007. Organic polyelectrolytes in water treatment. Water Res. 41, 2301–2324. Buron, H., Mengual, O., Meunier, G., Cayré, I., Snabre, P., 2004. Optical characterization of concentrated dispersions: Applications to laboratory analyses and on-line process monitoring and control. Polym. Int. 53, 1205–1209. Butler, J.N., 1964. Ionic Equilibrium - A Mathematical Approach. Reading, Mass. : Addison-Wesley Pub. Co. Celik, M.S., Somasundaran, P., 1986. The effect of multivalent ions on the flotation of coal. Sep. Sci. Technol. 21, 393–402. Cerovic, L., Lefèvre, G., Jaubertie, A., Fédoroff, M., Milonjic, S., 2009. Deposition of hematite particles on polypropylene walls in dynamic conditions. J. Colloid Interface Sci. 330, 284–291. Chaïrat, C., Oelkers, E.H., Schott, J., Lartigue, J.E., 2007a. Fluorapatite surface composition in aqueous solution deduced from potentiometric, electrokinetic, and solubility measurements, and spectroscopic observations. Geochim. Cosmochim. Acta 71, 5888–5900. Chaïrat, C., Schott, J., Oelkers, E.H., Lartigue, J.E., Harouiya, N., 2007b. Kinetics and mechanism of natural fluorapatite dissolution at 25 °C and pH from 3 to 12. Geochim. Cosmochim. Acta 71, 5901–5912. Chander, S., Fuerstenau, D., 1984. Solubility and interfacial properties of hydroxyapatite: a review, in: Misra, D.N. (Ed.), Adsorption on and Surface Chemistry of Hydroxyapatite. Springer, pp. 29–49. Chapman, D.L., 1913. LI. A contribution to the theory of electrocapillarity. London, Edinburgh, Dublin Philos. Mag. J. Sci. 25, 475–481. Chen, K.L., Mylon, S.E., Elimelech, M., 2006. Aggregation kinetics of alginate-coated hematite nanoparticles in monovalent and divalent electrolytes. Environ. Sci. Technol. 40, 1516–1523. Cheng, Y.L., Wong, R.J., Lin, J.C. Te, Huang, C., Lee, D.J., Mujumdar, A.S., 2010. Water coagulation using electrostatic patch coagulation (EPC) mechanism. Dry. Technol. 850–857, 850–857. Christenson, H.K., 1988. Non-DLVO forces between surfaces- solvation, hydration and capillary effects. Dispers. Sci. Technol. 9, 171. Chubar, N.I., Kanibolotskyy, V.A., Strelko, V. V, Gallios, G.G., Samanidou, V.F., Shaposhnikova, T.O., Milgrandt, V.G., Zhuravlev, I.Z., 2005. Adsorption of phosphate ions on novel inorganic ion exchangers. Colloids Surfaces A Physicochem. Eng. Asp. 255, 55–63. Claesson, P.M., 1987. Experimental evidence for repulsive and attractive forces not accounted for by conventional DLVO theory. Prog. Colloid Polym. Sci. 74, 48. Claesson, P.M., Blom, C.E., Herder, P.C., Ninham, B.W., 1986. Interactions between water-stable hydrophobic Langmuir-Blodgett monolayers on mica. J. Colloid Interface Sci. 114, 234–242. Claesson, P.M., Christenson, H.K., 1988. Very long range attractive forces between uncharged hydrocarbon and fluorocarbon surfaces in water. J. Phys. Chem. 92, 1650–1655. Clark, J.S., 1955. Solubility criteria for the existence of hydroxyapatite. Can. J. Chem. 33, 1696. Connah, M.T., Kaszuba, M., Morfesis, A., 2002. High resolution zeta potential measurements: Analysis of multi-component mixtures. J. Dispers. Sci. Technol. 23, 663–669. Cornell, R.M., Giovanoli, R., 1993. Acid dissolution of hematites of different morphologies. Clay   140 Miner. 28, 223–232. Cornell, R.M., Schindler, P.W., 1987. Photochemical dissolution of goethite in acid/oxalate solution. Clays Clay Miner. 35, 347–352. Cornell, R.M., Schwertmann, U., 1996. The iron oxides : structure, properties, reactions, occurrence, and uses. VCH, Weinheim. Cosgrove, T., 2005. Colloid Science: Principles, Methods and Applications. Blackwell Publishing. Dasgupta, H.K., Taluja, S.C., 1988. Beneficiation of a magnetite-apatite ore from Sung Valley, Shillong, Meghalaya. Bull. Mater. Sci. 10, 453–460. Davies, C.W., Hoyle, B.E., 1951. The dissociation constant of calcium hydroxide. J. Chem. Soc. 233–234. Delgado, A.V., 2002. Interfacial Electrokinetics and Electrophoresis. Marcel Dekker, Inc., New York. Delgado, A.V., González-Caballero, F., Hunter, R.J., Koopal, L.K., Lyklema, J., 2007. Measurement and interpretation of electrokinetic phenomena. J. Colloid Interface Sci. 309, 194–224. Deng, M., Liu, Q., Xu, Z., 2013a. Impact of gypsum supersaturated water on the uptake of copper and xanthate on sphalerite. Miner. Eng. 49, 165–171. Deng, M., Liu, Q., Xu, Z., 2013b. Impact of gypsum supersaturated solution on surface properties of silica and sphalerite minerals. Miner. Eng. 46–47, 6–15. Deng, R., Hu, Y., Ku, J., Zuo, W., Yang, Z., 2017. Adsorption of Fe(III) on smithsonite surfaces and implications for flotation. Colloids Surfaces A Physicochem. Eng. Asp. 533, 308–315. Derjaguin, B.., Landau, L., 1941. Theory of the stability of strongly charged lyophobic sols and of the adhesion of strongly charged particles in solution of electrolytes. Acta Physico-Chimica Sin. USSR 14, 633–662. Derjaguin, B.V., 1934. Untersuchungen über die reibung und adhäsion, IV. Colloid Polym. Sci. 69, 155–164. DiFeo, A., Finch, J.A., Xu, Z., 2001. Sphalerite-silica interactions: effect of pH and calcium ions. Int. J. Miner. Process. 61, 57–71. Ding, X., Repka, C., Xu, Z., Masliyah, J., 2006. Effect of illite clay and divalent cations on bitumen recovery. Can. J. Chem. Eng. 84, 643–650. Dishon, M., Zohar, O., Sivan, U., 2011. Effect of cation size and charge on the interaction between silica surfaces in 1:1, 2:1, and 3:1 aqueous electrolytes. Langmuir 27, 12977–12984. Dixon, D.R., 1985. Interaction of alkaline-earth-metal ions with magnetite. Colloids and Surfaces 13, 273–286. Dobiáš, B., 2005. Surfactant adsorption and dispersion stability in mineral flotation, in: Stechemesser, H., Dobiáš, B. (Eds.), Coagulation and Flocculation. Taylor & Francis, pp. 663–765. Dorozhkin, S.V., 2002. A review on the dissolution models of calcium apatites. Prog. Cryst. Growth Charact. Mater. 44, 45–61. Drzymala, J., Lekki, J., Laskowski, J., 1979. Surface dissociation constants for solid oxide/aqueous solution systems. Colloid Polym. Sci. 257, 768–772. Dyall, H., 1985. An investigation of the beneficiation of carbonatite phosphate ore from Mr. Weld. Bentley, WA. Dzombak, D.A., Morel, F.M.M., 1990. Surface complexation modeling: hydrous ferric oxide. Wiley, New York.   141 Eigeles, M.A., 1958. Selective flotation of non-sulfide minerals, progressin mineral dressing, in: Progressin Mineral Dressing, Trans. 4th Int. Mineral Dressing Congress. Almqvist and Wiksell, pp. 591–609. El-Ammouri, E., Mirnezami, M., Lascelles, D., Finch, J.A., 2002. Aggregation index and a methodology to study the role of magnesium in aggregation of sulphide slurries. CIM Bull. 95, 67–72. El-Midany, A.A., El-Shall, H., Svoronos, S., 2011. Modeling the PVA-coated dolomite floatability in acidic media. Powder Technol. 209, 25–28. El-Shall, H., Zhang, P., Abdel Khalek, N., El-Mofty, S., 2004. Beneficiation technology of phosphates: challenges and solutions. Miner. Metall. Process. 21, 17–26. Elgillani, D.A., Abouzeid, A.Z.M., 1993. Flotation of carbonates from phosphate ores in acidic media. Int. J. Miner. Process. 38, 235–256. Elmahdy, A.M., El-Midany, A.A., Abdel-Khalek, N.A., El- Mofty, S.E., 2009. Effect of oleate/bacteria interactions on dolomite separation from phosphate ore. Tenside Surfactants Deterg. 46, 340–345. Elzinga, E.J., Sparks, D.L., 2007. Phosphate adsorption onto hematite : An in situ ATR-FTIR investigation of the effects of pH and loading level on the mode of phosphate surface complexation. J. Colloid Interface Sci. 308, 53–70. Engwayu, J., 2015. Aggregation and dispersion phenomena in the quartz-hematite system in the presence of polymers. The University of British Columbia. Ersoy, B., Çelik, M.S., 2002. Electrokinetic properties of clinoptilolite with mono- and multivalent electrolytes. Microporous Mesoporous Mater. 55, 305–312. Feick, J.D., Velegol, D., 2002. Measurements of charge nonuniformity on polystyrene latex particles. Langmuir 18, 3454–3458. Feick, J.D., Velegol, D., 2000. Electrophoresis of spheroidal particles having a random distribution of zeta potential. Langmuir 16, 10315–10321. Feng, B., Feng, Q., Lu, Y., 2012. A novel method to limit the detrimental effect of serpentine on the flotation of pentlandite. Int. J. Miner. Process. 114–117, 11–13. Feng, B., Peng, J., Zhang, W., Luo, G., Wang, H., 2018. Removal behavior of slime from pentlandite surfaces and its effect on flotation. Miner. Eng. 125, 150–154. Forbes, E., Davey, K.J., Smith, L., 2014. Decoupling rehology and slime coatings effect on the natural flotability of chalcopyrite in a clay-rich flotation pulp. Miner. Eng. 56, 136–144. Fornasiero, D., Ralston, J., 2006. Effect of surface oxide/hydroxide products on the collectorless flotation of copper-activated sphalerite. Int. J. Miner. Process. 78, 231–237. Fuerstenau, D., 1970. Interfacial processes in mineral/water systems. Pure Appl. Chem. 24, 135–164. Fuerstenau, D.W., 1980. Fine particle flotation, in: Somasundaran, P. (Ed.), Fine Particles Processing. AIME, New York, pp. 669–705. Fuerstenau, D.W., Pradip, 2005. Zeta potentials in the flotation of oxide and silicate minerals. Adv. Colloid Interface Sci. 114–115, 9–26. Fulmer, M.T., Ison, I.C., Hankermayer, C.R., Constantz, B.R., Ross, J., 2002. Measurements of the solubilities and dissolution rates of several hydroxyapatites. Biomaterials 23, 751–755. Gaudin, A., Fuerstenau, D., 1955. Quartz flotation with anionic collectors. Trans. AIME 202, 66–72. Gaudin, A.M., 1957. Flotation, 2nd editio. ed. McGraw-Hill, New York.   142 Gipple, C., 1989. The use of turbidimeters in suspended sediment research. Hydrobiologia 176/177, 465–480. Gong, W.Q., Klauber, C., Warren, L.J., 1993. Mechanism of action of sodium silicate in the flotation of apatite from hematite. Int. J. Miner. Process. 39, 251–273. Gong, W.Q., Parentich, A., Little, L.H., Warren, L.J., 1992. Selective flotation of apatite from iron oxides. Int. J. Miner. Process. 34, 83–102. Gordon, H.R., Smith, R.C., Ronald, J., Zaneveld, V., 1980. Introduction To Ocean Optics. SPIE Ocean Optics VI. Gouy, G., 1910. Constitution of the electric charge at the surface of an electrolyte. J. Phys. 9, 457–467. Grahame, D.C., 1947. The electrical double layer and the theory of electrocapillarity. Chem. Rev. 41, 441–501. Greene, M.R., Hammer, D.A., Olbricht, W.L., 1994. The effect of hydrodynamic flow field on colloidal stability. J. Colloid Interface Sci. 167, 232–246. Greenwald, I., 1942. The solubility of calcium phosphate I. The effect of pH and amount of solid phase. J. Biol. Chem. 143, 703–710. Greenwood, R., 2003. Review of the measurement of zeta potentials in concentrated aqueous suspensions using electroacoustics. Adv. Colloid Interface Sci. 106, 55–81. Gregory, J., 1973. Rates of flocculation of latex particles by cationic polymers. J. Colloid Interface Sci. 42, 448–456. Groppo, J.G., Parekh, B.K., 1996. Surface chemical control of ultra fine coal to improve dewatering. Coal Prep. 17, 103–116. Gruber, G.A., Moudgil, B.M., Somasundaran, P., 1995. Understanding the basics of anionic conditioning in phosphate flotation. Gu, Z., 2002. Fine particle flotation for florida dolomitic phosphate pebbles. West Virginia University. Guan, C., 2009. Theoretical Backgound of the Crago Phosphate Flotation Process. Miner. Metall. Process. 26, 55–64. Guidry, M.W., Mackenzie, F.T., 2003. Experimental study of igneous and sedimentary apatite dissolution: control of pH, distance from equilibrium, and temperature on dissolution rates. Geochim. Cosmochim. Acta 67, 2949–2963. Guimarães, R.C., Araujo, A.C., Peres, A.E.C., 2005. Reagents in igneous phosphate ores flotation. Miner. Eng. 18, 199–204. Guimarães, R.C., Peres, A.E.C., 1999. Interfering ions in the flotation of a phosphate ore in a batch column. Miner. Eng. 12, 757–768. Hamaker, H.C., 1937. The London-van der Waals attraction between spherical particles. Physica IV, 1058–1072. Hanna, H.S., Anazia, J., 1983. Beneficiation of low grade Birmingham iron ores to recover enriched concentrates, in: Proceedings of the SME-AIME Fall Meeting. Littleton, CO: SME., Salt Lake City, Utah. Hanna, H.S., Somasundaran, P., 1976. Flotation of salt-type minerals, in: Fuerstenau, M.C. (Ed.), Flotation, A.M. Gaudin Memorial Volume 1. AIME, New York, pp. 197–272. Hansmann, D.D., Anderson, M.A., 1985. Using electrophoresis in modeling sulfate, selenite, and phosphate adsorption onto goethite. Environ. Sci. Technol. 19, 544–551. Hayes, K.F., Papelis, C., Leckie, J.O., 1988. Modeling ionic strength effects on anion adsorption   143 at hydrous oxide / solution interfaces. J. Colloid Interface Sci. 125, 717–726. Healy, T., Wiese, G., Yates, D., Kavanagil, B., 1973. Heterocoagulation in mixed oxide colloidal dispersions. J. Colloid Interface Sci. 42, 647–649. Healy, T.W., Jellett, V.R., 1967. Adsorption-coagulation reactions of Zn (II) hydrolyzed species at the zinc oxide-water interface. J. Colloid Interface Sci. 24, 41–46. Helmholtz, H., 1879. The Double Layer. Wied.Ann 7. Henderson, J.M., Wheatley, A.D., 1987. Factors effecting a loss of flocculation activity of polyacrylamide solutions: Shear degradation, cation complexation, and solution aging. J. Appl. Polym. Sci. 33, 669–684. Henry, D., 1931. The cataphoresis of suspended particles. Part I. the equation of cataphoresis. Hiemstra, T., Van Riemsdijk, W.H., 1996. A surface structural approach to ion adsorption : The charge distribution (CD) model. J. Colloid Interface Sci. 179, 488–508. Hogg, R., Healy, T.W., Fuerstenau, D.W., 1966. Mutual coagulation of colloidal dispersions. Trans. Faraday Soc. 62, 1638–1651. Horn, R.G., Smith, D.T., Haller, W., 1989. Surface forces and viscosity of water measured between silica sheets. Chem. Phys. Lett. 162, 404–408. Hu, Y., Yu, M., 1988. Controlled dispersion-shear flocculation separation of ultrafine apatile from rhodochrosite, in: Plumpton, A.J. (Ed.), Production and Processing of Fine Particles. Pergamon, New York, pp. 353–362. Huang, X., 2004. Intersection of isotherms for phosphate adsorption on hematite. J. Colloid Interface Sci. 271, 296–307. Hunter, R., 1966. The interpretation of electrokinetic potentials. J. Colloid Interface Sci. 22, 231–239. Hunter, R.J., 1998. Recent developments in the electroacoustic characterisation of colloidal suspensions and emulsions. Colloids Surfaces A Physicochem. Eng. Asp. 141, 37–66. Hunter, R.J., 1993. Introduction to Modern Colloid Science. Oxford, England ; New York: Oxford University Press. Hunter, R.J., 1987. Foundations of Colloid Science. Oxford Science Publications. Hunter, R.J., 1981. Zeta Potential in Colloid Science: Principles and Applications. Academic Press, London. Hunter, R.J., Wright, H.J.L., 1971. The dependence of electrokinetic potential on concentration of electrolyte. J. Colloid Interface Sci. 37, 564–580. Huynh, L., Feiler, A., Michelmore, A., Ralston, J., Jenkins, P., 2000. Control of slime coatings by the use of anionic phosphates: A fundamental study. Miner. Eng. 13, 1059–1069. Ikumapayi, F., Makitalo, M., Johansson, B., Rao, K.H., 2012. Recycling of process water in sulphide flotation: Effect of calcium and sulphate ions on flotation of galena. Miner. Eng. 39, 77–88. Israelachvili, J.N., 1992. Adhesion forces between surfaces in liquids and condensable vapours. Surf. Sci. Rep. 14, 109–159. Israelachvili, J.N., Adams, G.E., 1978. Measurement of forces between two mica surfaces in aqueous-electrolyte solutions in range 0-100 nm. J. Chem. Soc. Trans. I 74, 975–1001. James, R.O., Healy, T.W., 1972a. Adsorption of hydrolyzable metal ions at the oxide-water interface. I. Co(II) adsorption on SiO 2 and TiO 2 as model systems. J. Colloid Interface Sci. 40, 42–52. James, R.O., Healy, T.W., 1972b. Adsorption of hydrolyzable metal ions at the oxide-water   144 interface. II. Charge reversal of SiO2 and TiO2 colloids by adsorbed Co(II), La(III), and Th(IV) as model systems. J. Colloid Interface Sci. 40, 53–64. James, R.O., Healy, T.W., 1972c. Adsorption of hydrolyzable metal ions at the oxide-water interface. III. A Thermodynamic Model of Adsorption. J. Colloid Interface Sci. 40, 65–81. James, R.O., Stiglich, P.J., Healy, T.W., 1975. Analysis of models of adsorption of metal ions at oxide/water interfaces. Faraday Discuss. Chem. Soc. 59, 142–156. Jaynes, W.F., Moore, P.A., Miller, D.M., 1999. Solubility and ion activity products of calcium phosphate minerals. J. Environ. Qual. 28, 530–536. Johnson, P.N., Johnson, P.N. and A.A., 1983. Ferric chloride and alum as single and dual coagulants and dual coagulants. J. Am. Water Work. Assoc. 75, 232–239. Kawatra, S.K., Carlson, J.T., 2013. Beneficiation of Phosphate Ore. Society for Mining, Metallurgy, and Exploration. Kirjavainen, V., Schreithofer, N., Heiskanen, K., 2002. Effect of calcium and thiosulfate ions on flotation selectivity of nickel-copper ores. Miner. Eng. 15, 1–5. Kitchener, J.A., 1969. Colloidal minerals: Chemical aspects of their dispersion, flocculation and filtration. Fltration Sep. 6, 553–560. Kloster, N., Brigante, M., Zanini, G., Avena, M., 2013. Aggregation kinetics of humic acids in the presence of calcium ions. Colloids Surfaces A Physicochem. Eng. Asp. 427, 76–82. Kosmulski, M., 2014. The pH dependent surface charging and points of zero charge. VI. Update. J. Colloid Interface Sci. 426, 209–212. Kosmulski, M., 2011. The pH-dependent surface charging and points of zero charge V. Update. J. Colloid Interface Sci. 353, 1–15. Kosmulski, M., 2004. pH-dependent surface charging and points of zero charge II. Update. J. Colloid Interface Sci. 275, 214–224. Kosmulski, M., 2002. The pH-dependent surface charging and the points of zero charge. J. Colloid Interface Sci. 253, 77–87. Krishnan, S. V., Iwasaki, I., 1986. Heterocoagulation vs. surface precipitation in a quartz-Mg(OH)₂ system. Environ. Sci. Technol. 20, 1224–1229. Kusuma, A.M., Liu, Q., Zeng, H., 2014. Understanding interaction mechanisms between pentlandite and gangue minerals by zeta potential and surface force measurements. Miner. Eng. 69, 15–23. Lagaly, G., Ziesmer, S., 2003. Colloid chemistry of clay minerals: The coagulation of montmorillonite dispersions. Adv. Colloid Interface Sci. 100–102, 105–128. Le Bell, J.C., Lindstrom, L., 1982. Electrophoretic characterization of some calcium minerals. Finnish Chem. Lett. 6, 134–138. Leite Nunes, A.P., Lopes Pinto, C.L., Sales Valadão, G.E., De Magalhães Viana, P.R., 2012. Floatability studies of wavellite and preliminary results on phosphorus removal from a Brazilian iron ore by froth flotation. Miner. Eng. 39, 206–212. Leja, J., 1982. Surface chemistry of froth flotation. Plenum, New York. Leng, Y., 2012. Hydration force between mica surfaces in aqueous KCl electrolyte solution. Langmuir 28, 5339–5349. Leong, Y.K., 2005. Yield stress and zeta potential of nanoparticulate silica dispersions under the influence of adsorbed hydrolysis products of metal ions - Cu(II), Al(III) and Th(IV). J. Colloid Interface Sci. 292, 557–566. Leong, Y.K., Teo, J., Teh, Ej., Smith, J., Widjaja, J., Lee, J.X., Fourie, A., Fahey, M., Chen, R.,   145 2012. Controlling attractive interparticle forces via small anionic and cationic additives in kaolin clay slurries. Chem. Eng. Res. Des. 90, 658–666. Ler, A., Stanforth, R., 2003. Evidence for surface precipitation of phosphate on goethite. Environ. Sci. Technol. 37, 2694–2700. Levinskas, G.J., Newman, W.F., 1955. The solubility of bone mineral I. Solubility studies of synthetic hydroxyapatite. J. Phys. Chem. 59, 164. Li, R., Wang, J.J., Zhou, B., Awasthi, M.K., Ali, A., Zhang, Z., Gaston, L.A., Lahori, A.H., Mahar, A., 2016. Enhancing phosphate adsorption by Mg/Al layered double hydroxide functionalized biochar with different Mg/Al ratios. Sci. Total Environ. 559, 121–129. Li, W., Zhang, S., Jiang, W., Shan, X., 2006. Effect of phosphate on the adsorption of Cu and Cd on natural hematite. Chemosphere 63, 1235–1241. Liang, L., Morgan, J.J., 1990. Chemical aspects of iron oxide coagulation in water: Laboratory studies and implications for natural systems. Aquat. Sci. 52, 32–55. Liang, Y., Hilal, N., Langston, P., Starov, V., 2007. Interaction forces between colloidal particles in liquid: Theory and experiment. Adv. Colloid Interface Sci. 134–135, 151–166. Lima, J.M., Anderson, S.J., Curi, N., 2000. Phosphate-induced clay dispersion as related to aggregate size and composition in Hapludoxs. Soil Sci. Soc. Am. J. 64, 892–897. Lin, J., Raghavan, S., Fuerstenau, D.W., 1981. The adsorption of fluoride ions by hydroxyapatite from aqueous solution. Colloids and Surfaces 3, 357–370. Lin, J., Zhan, Y., Wang, H., Chu, M., Wang, C., He, Y., Wang, X., 2017. Effect of calcium ion on phosphate adsorption onto hydrous zirconium oxide. Chem. Eng. J. 309, 118–129. Liu, C., Song, S., Li, H., Li, Y., Ai, G., 2019. Elimination of the adverse effect of calcite slimes on the sulfidization flotation of malachite in the presence of water glass. Colloids Surfaces A Physicochem. Eng. Asp. 563, 324–329. Liu, J., Xu, Z., Masliyah, J., 2005. Interaction forces in bitumen extraction from oil sands. J. Colloid Interface Sci. 287, 507–520. Liu, J., Xu, Z., Masliyah, J., 2004. Role of fine clays in bitumen extraction from oil sands. AICHE J. 50, 1917–1927. Liu, J., Xu, Z., Masliyah, J., 2003. Studies on bitumen-silica interaction in aqueous solutions by atomic force microscopy. Langmuir 19, 3911–3920. Liu, J., Zhou, Z., Xu, Z., Masliyah, J., 2002. Bitumen-clay interactions in aqueous media studied by zeta potential distribution measurement. J. Colloid Interface Sci. 252, 409–418. Liu, Y., Sheng, X., Dong, Y., Ma, Y., 2012. Removal of high-concentration phosphate by calcite: Effect of sulfate and pH. Desalination 289, 66–71. Lu, S., Pugh, R.J., Forssberg, E., 2005. Interfacial separation of particles. Elsevier Publications. Lyklema, J.., 1987. Structure of the solid-liquid interface and the electrical double layer, in: Thadros, T.F. (Ed.), Solid/Liquid Dispersions. Academic Press, London, pp. 63–90. Lyklema, J., Overbeek, J.T.G., 1961. On the interpretation of electrokinetic potentials. J. Colloid Sci. 16, 501–512. MacKenzie, J.M.W., Mishra, S.K., 1970. Zeta potential studies on apatite, in: 4th National Convention, Australian Chemical Institute. Canberra, Australia, p. 43. Manukonda, V.R., Iwasaki, I., 1987. Control of calcium ion via chemical precipitation-ultrasonic treatment in selective flocculation. Miner. Metall. Process. 4, 217–222. Matijevic, E., 1973. Colloid stability and complex chemistry. J. Colloid Interface Sci. 43, 217–245.   146 McDowell, H., Gregory, T.M., Brown, W.E., 1977. Solubility of Ca₅(PO₄)₃OH in the system Ca(OH)₂-H₃PO₄-H₂O at 5, 15, 25, and 37 °C. J. Res. Int. Bur. Stand. Sect. A—Physics Chem. 81, 273–281. Mcpheat, I.W., 1984. Beneficiation of Mt. Weld phosphate. Frewville, South Australian. Mika, H., Bell, L.C., Kruger, B.J., 1976. The role of surface reactions in the dissolution of stoichiometric hydroxyapatite. Arch. Oral Biol. 21, 1976. Mirnezami, M., Restrepo, L., Finch, J.A., 2003. Aggregation of sphalerite: Role of zinc ions. J. Colloid Interface Sci. 259, 36–42. Mishra, R.K., Chander, S., Fuerstenau, D.W., 1980. Effect of ionic surfactants on the electrophoretic mobility of hydroxyapatite. Colloids and Surfaces 1, 105. Mishra, S.K., 1978. The electrokinetics of apatite and calcite in inorganic electrolyte environment. Int. J. Miner. Process. 5, 69–83. Mitchell, T.K., Nguyen, A. V., Evans, G.M., 2005. Heterocoagulation of chalcopyrite and pyrite minerals in flotation separation. Adv. Colloid Interface Sci. 114, 227–237. Moudgil, B.M., Somasundaran, P., 1986. Advances in phosphate flotation, in: SME/AIME. Mpofu, P., Addai-Mensah, J., Ralston, J., 2003. Influence of hydrolyzable metal ions on the interfacial chemistry, particle interactions, and dewatering behavior of kaolinite dispersions. J. Colloid Interface Sci. 261, 349–359. Muhammed, M., Zhang, Y., 1989. A hydrometallurgical process for the dephosphorization of iron ore. Hydrometallurgy 21, 277–292. Mylon, S.E., Chen, K.L., Elimelech, M., 2004. Influence of natural organic matter and ionic composition on the kinetics and structure of hematite colloid aggregation: Implications to iron depletion in estuaries. Langmuir 20, 9000–9006. Nanthakumar, B., Grimm, D., Pawlik, M., 2009. Anionic flotation of high-iron phosphate ores-Control of process water chemistry and depression of iron minerals by starch and guar gum. Int. J. Miner. Process. 92, 49–57. Napper, D.H., 1977. Steric stabilization. J. Colloid Interface Sci. 58, 390–407. Ni, C., Bu, X., Xia, W., Peng, Y., Xie, G., 2018. Effect of slimes on the flotation recovery and kinetics of coal particles. Fuel 220, 159–166. Oliveira, M.S., Queiroz, G.M., Guimarães, R.C., Ataíde, C.H., Barrozo, M.A.S., 2007. Selectivity in phosphate column flotation. Miner. Eng. 20, 197–199. Oliveira, M.S., Santana, R.C., Ataíde, C.H., Barrozo, M.A.S., 2011. Recovery of apatite from flotation tailings. Sep. Purif. Technol. 79, 79–84. Overbeek, J.T.G., 1977. Recent developments in the understanding of colloid stability. Colloid Interface Sci. 58, 408–422. Packham, R.F., 1965. Some studies of coagulation of dispersed clays with hydrolyzing salts. J. Colloid Sci. 20, 81–92. Packman, J.J., Comings, K.J., Booth, D.B., 1999. Using turbidity to determine total suspended solids in urbanizing streams in the Puget Lowlands. Manag. Chang. Water Resour. Environ. Can. Water Resour. Assoc. 158–165. Parida, K.M., Das, N.N., 1996. Reductive dissolution of hematite in hydrochloric acid medium by some inorganic and organic reductants: A comparative study. Indian J. Eng. Mater. Sci. 3, 243–247. Parks, G., De Bruyn, P., 1962. The zero point of charge of oxides. J. Phys. Chem. 66, 967–973. Parks, G.A., 1965. The isoelectric points of solid oxides, solid hydroxides, and aqueous hydroxo   147 complex systems. Chem. Rev. 65, 177–198. Parsonage, P.G., 1985. Effects of slime and colloidal particles on the flotation of galena, in: Flotation of Sulphide Minerals. Elsevier, Amsterdam, pp. 111–140. Parsons, R., 1990. Electrical double layer: Recent experimental and theoretical developments. Chem. Rev. 90, 813–826. Pashley, R.M., 1992. Interparticulate forces, in: Laskowski, J., Ralston, J. (Eds.), Colloid Chemistry in Mineral Processing. Elsevier, Amsterdam, pp. 97–114. Pashley, R.M., 1984. Forces between mica surfaces in La3+ and Cr3+ electrolyte solutions. J. Colloid Interface Sci. 102, 23–35. Pashley, R.M., 1982. Hydration forces between mica surfaces in electrolyte solutions. Adv. Colloid Interface Sci. 16, 57–62. Pashley, R.M., 1981. Hydration forces between mica surfaces in aqueous electrolyte solutions. J. Colloid Interface Sci. 80, 153–162. Pavanelli, D., Bigi, A., 2005. Indirect methods to estimate suspended sediment concentration: Reliability and relationship of turbidity and settleable solids. Biosyst. Eng. 90, 75–83. Peng, Y., Zhou, S., 2011. The effect of surface oxidation of copper sulfide minerals on clay slime coating in flotation. Miner. Eng. 24, 1687–1693. Penner, D., Lagaly, G., 2001. Influence of anions on the rheological properties of clay mineral dispersions. Appl. Clay Sci. 19, 131–142. Perrin, J., 1904. Mécanisme de l’électrisation de contact et solutions colloidales. J. Chim. Phys. 2, 601–665. Pokrovsky, O.S., Mielczarski, J.A., Barres, O., Schott, J., 2000. Surface speciation models of calcite and dolomite/aqueous solution interfaces and their spectroscopic evaluation. Langmuir 16, 2677–2688. Potapova, E., Jolstera, R., Holmgren, A., Grahn, M., 2014. The effect of inorganic ions on dodecylbenzenesulfonate adsorption onto hematite : an ATR-FTIR study. J. Surfactants Deterg. 17, 1027–1034. Potapova, E., Yang, X., Grahn, M., Holmgren, A., Forsmo, S.P.E., Fredriksson, A., Hedlund, J., 2011. The effect of calcium ions, sodium silicate and surfactant on charge and wettability of magnetite. Colloids Surfaces A Physicochem. Eng. Asp. 386, 79–86. Powell, K.J., Brown, P.L., Byrne, R.H., Gajda, T., Hefter, G., Sjöberg, S., Wanner, H., 2005. Chemical speciation of environmentally significant heavy metals with inorganic ligands Part 1: The Hg2+– Cl−, OH−, CO₃2−, SO₄2−, and PO₄3− aqueous systems. Pure Appl. Chem. 77, 739–800. Pugh, R., Stenius, P., 1985. Solution chemistry studies and flotation behaviour of apatite, calcite and fluorite minerals with sodium oleate collector. Int. J. Miner. Process. 15, 193–218. Ranjbar, M., 2002. Dephosphorization of iranian iron oxide fines by flotation. Erzmetall J. Explor. Min. Metall. 55, 612–616. Rao, K.H., Antti, B.M., Forssberg, E., 1990. Mechanism of oleate interaction on salt-type minerals part II. Adsorption and electrokinetic studies of apatite in the presence of sodium oleate and sodium metasilicate. Int. J. Miner. Process. 28, 59–79. Rao, K.H., Dwari, R., Lu, S., Vilinska, A., Somasundaran, P., 2010. Flotation of phosphate gangue from magnetite fines - non-ionic surfactant as atrac collector modifier, in: XXV International Mineral Processing Congress, IMPC. Australasian Institute of Mining and Metallurgy, Brisbane, QLD, Australia, pp. 1933–1943.   148 Ren, H., Ning, J., Song, G.L., 2012. Study on the interaction in silica-sphalerite-mixture (silica:sphalerite =4:1)-calcium carbonate systems. Adv. Mater. Res. 524–527, 1105–1108. Rietra, R.P., Hiemstra, T., Van Riemsdijk, W.H., 2001. Interaction between calcium and phosphate adsorption on goethite. Environ. Sci. Technol. 35, 3369–3374. Robinson, J.W., Skelly Frame, E.M., Frame II, G.M., 2004. Undergraduate instrumental analysis. CRC Press; 6th edition, New York. Rubio, J., Matiolo, E., De Paiva, M., 2015. Recovery of phosphate ores in the modified three-product column (3PC) flotation cell. Miner. Eng. 72, 121–128. Saleeb, F.Z., De Bruyn, P.L., 1972. Surface properties of alkaline earth apatites. J. Electroanal. Chem. Interfacial Electrochem. 37, 99–118. Santana, R.C., Farnese, A.C.C., Fortes, M.C.B., Ataíde, C.H., Barrozo, M.A.S., 2008. Influence of particle size and reagent dosage on the performance of apatite flotation. Sep. Purif. Technol. 64, 8–15. Schwertmann, U., 1984. The influence of aluminium on iron oxides: IX. Dissolution of Al-goethites in 6 M HCl. Clay Miner. 19, 9–19. Scotti, J.L., Smith, R.W., 1993. Calcium ion effects in amine flotation of quartz and magnetite. Miner. Eng. 6, 1245–1255. Shang, C., Stewart, J.W.B., Huang, P.M., 1992. pH effect on kinetics of adsorption of organic and inorganic phosphates by short-range ordered aluminum and iron precipitates. Geoderma 53, 1–14. Shaw, D.J., 1980. Introduction to Colloid and Surface Chemistry, 3rd ed. Butterworth and Company (publishers) Limited., London. Shibata, J., Fuerstenau, D.W., 2003. Flocculation and flotation characteristics of fine hematite with sodium oleate. Int. J. Miner. Process. 72, 25–32. Sidhu, P.S., Gilkes, R.J., Cornell, R.M., Posner, A.M., Quirk, J.P., 1981. Dissolution of iron oxides and oxyhydroxides in hydrochloric and perchloric acids. Clays Clay Miner. 29, 269–276. Siirak, J., Hancock, B.A., 1988. Progress in developing a flotation phosphorus reduction process at the tilden iron ore mine, in: XVI International Mineral Processing Congress. Elsevier, Stockholm, Sweden. Sinha, P., Szilagyi, I., Montes Ruiz-Cabello, F.J., Maroni, P., Borkovec, M., 2013. Attractive forces between charged colloidal particles induced by multivalent ions revealed by confronting aggregation and direct force measurements. J. Phys. Chem. Lett. 4, 648–652. Sis, H., Chander, S., 2003. Reagents used in the flotation of phosphate ores: A critical review. Miner. Eng. 16, 577–585. Smani, M.. ., Blazy, P., Cases, J.M., 1975. Beneficiation of sedimentary Moroccan phosphate ores, Parts 1-4. AIME Trans. 258, 168. Snoeyink, V.L., Jenkins, D., 1980. Water Chemistry. Wiley and Sons, New York. Somasundaran, P., 1970. Pretreatment of mineral surfaces and its effect on their properties, in: Goldfinger, G. (Ed.), Clean Surfaces: Their Preparation and Characterization for Interface Studies. Marcel Dekker, New York, p. 285. Somasundaran, P., 1968. Zeta potential of apatite in aqueous solutions and its change during equilibration. Colloids and Surfaces 27, 659. Somasundaran, P., Agar, G.E., 1972. Further streaming potential studies of hydroxyapatite in inorganic solutions. Trans. AIME 252, 348–352. Somasundaran, P., Amankonah, J.O., Ananthapadmabhan, K.P., 1985. Mineral—solution   149 equilibria in sparingly soluble mineral systems. Colloids and Surfaces 15, 309–333. Somasundaran, P., Wang, Y.W., 1984. Surface chemical and adsorption properties of apatite, in: Adsorption and Surface Chemistry of Hydroxyapatite Plenum. New York, pp. 129–149. Stachowicz, M., Hiemstra, T., Van Riemsdijk, W.H., 2008. Multi-competitive interaction of As(III) and As(V) oxyanions with Ca2+, Mg2+, PO₄3−, and CO₃2− ions on goethite. J. Colloid Interface Sci. 320, 400–414. Stern, O., 1924. Theory of the electrical double layer. Electrochemistry 30, 508–516. Stumm, W., Morgan, J.J., 1981. Aquatic Chemistry, 2nd ed. Wiley-Interscience, New York. Su, F.W., 1998. Dephosphorization of magnetite fines. Luleå University of Technology, Luleå. Subramanian, S., Rao, K.H., Forssberg, K.S.E., 2002. Dispersion characteristics of apatite and magnetite fines in the presence of inorganic and organic reagents and its influence on the dephosphorization of magnetite ore, in: Zhang, P., El- Shall, H., Somasundaran, P., Stana, R. (Eds.), Beneficiation of Phosphates III –Fundamentals and Technology. SME, Littleton, pp. 21–31. Sworska, A., Laskowski, J., Cymerman, G., 2000. Flocculation of the Syncrude fine tailings Part I. Effect of pH, polymer dosage and Mg2+ and Ca2+ cations. Int. J. Miner. Process. 60, 143–152. Szekely, J., Poveromo, J.J., 1975. A mathematical and physical representation of the raceway region in the iron blast furnace. Metall. Mater. Trans. B 6, 119–130. Takeo, M., 1999. Disperse systems. Wiley-VCH. Talebi Atouei, M., Rahnemaie, R., Goli Kalanpa, E., Davoodi, M.H., 2016. Competitive adsorption of magnesium and calcium with phosphate at the goethite water interface : Kinetics, equilibrium and CD-MUSIC modeling. Chem. Geol. 437, 19–29. Tang, J., Zhang, Y., Bao, S., 2016. The effect of Ca2+ and Mg2+ on the dispersion and flocculation behaviors of muscovite particles. Minerals 6. Tejedor-tejedor, M.I., Anderson, M.A., 1990. Protonation of phosphate on the surface of goethite as studied by CIR-FTIR and electrophoretic mobility. Langmuir 6, 602–611. Teo, J., Liew, W.K., Leong, Y.K., 2009. Clay, phosphate adsorption, dispersion, and rheology. Water, Air, Soil Pollut. Focus 9, 403–407. Tian, M., Liu, R., Gao, Z., Chen, P., Han, H., Wang, L., Zhang, C., Sun, W., Hu, Y., 2018. Activation mechanism of Fe (III) ions in cassiterite flotation with benzohydroxamic acid collector. Miner. Eng. 119, 31–37. Tohry, A., Dehghani, A., 2016. Effect of sodium silicate on the reverse anionic flotation of a siliceous-phosphorus iron ore. Sep. Purif. Technol. 164, 28–33. Torrent, J., Schwertmann, U., Barron, V., 1987. The reductive dissolution of synthetic goethite and haematite in dithionite. Clay Miner. 22, 329–337. Trefalt, G., Palberg, T., Borkovec, M., 2017. Forces between colloidal particles in aqueous solutions containing monovalent and multivalent ions. Curr. Opin. Colloid Interface Sci. 27, 9–17. Uribe, L., Gutierrez, L., Jerez, O., 2016. The depressing effect of clay minerals on the floatability of chalcopyrite. Miner. Process. Extr. Metall. Rev. 37, 227–235. Valmacco, V., Elzbieciak-Wodka, M., Herman, D., Trefalt, G., Maroni, P., Borkovec, M., 2016. Forces between silica particles in the presence of multivalent cations. J. Colloid Interface Sci. 472, 108–115. Valsami-Jones, E., Ragnarsdottir, K.W.W., Putnis, A., Bosbach, D., Kemp, A.J., Cressey, G.,   150 1998a. Solubility and ion activity products of calcium phosphate minerals. Chem. Geol. 151, 215–233. Valsami-Jones, E., Ragnarsdottir, K. V., Putnis, A., Bosbach, D., Kemp, A.J., Cressey, G., 1998b. The dissolution of apatite in the presence of aqueous metal cations at pH 2-7. Chem. Geol. 151, 215–233. Van Olphen, H., 1977. An Introduction to Clay Colloid Chemistry, 2nd ed. Wiley, New York. Velegol, D., Feick, J.D., Collins, L.R., 2000. Electrophoresis of spherical particles with a random distribution of zeta potential or surface charge. J. Colloid Interface Sci. 230, 114–121. Vergouw, J.M., Difeo, A., Xu, Z., Finch, J.A., 1998a. An agglomeration study of sulphide minerals using zeta potential and settling rate. Part I: pyrite and galena. Miner. Eng. 11, 159–169. Vergouw, J.M., Difeo, A., Xu, Z., Finch, J.A., 1998b. An agglomeration study of sulphide minerals using zeta potential and settling rate. Part II: sphalerite/pyrite and sphalerite/galena. Miner. Eng. 11, 605–614. Verwey, E.J.W., Overbeek, J.T.G., 1948. Theory of the stability of lyophobic colloids. Elsevier, Amsterdam. Vincent, B., 1974. The effect of adsorbed polymers on dispersion stability. Adv. Colloid Interface Sci. Interface Sci. 4, 193–277. Walsch, J., Dultz, S., 2010. Effects of pH, Ca- and SO₄-concentration on surface charge and colloidal stability of goethite and hematite - Consequences for the adsorption of anionic organic substances. Clay Miner. 45, 1–13. Wang, D., Gregory, J., Tang, H., Wang, D., Gregory, J., Tang, H., 2008. Mechanistic difference of coagulation of kaolin between PACl and cationic polyelectrolytes : A comparative study on zone 2 coagulation. Dry. Technol. 26, 1060–1067. Wang, Q., Heiskanen, K., 1992a. Dispersion selectivity and heterocoagulation in apatite-hematite- phlogopite fine particle suspensions. I. Dispersion of single minerals. Int. J. Miner. Process. 35, 121–131. Wang, Q., Heiskanen, K., 1992b. Dispersion selectivity and heterocoagulation in apatite-hematite-phlogopite fine particle suspensions. II. Dispersion selectivities of the mineral mixtures. Int. J. Miner. Process. 35, 133–145. Wang, Q., Heiskanen, K., 1992c. Selective hydrophobic flocculation in apatite-hematite system by sodium oleate. Miner. Eng. 5, 493–501. Wang, Q., Heiskanen, K., 1990. Batch flotation tests by fatty acid on a phosphate-iron oxide- silicate regolith ore sample from sokli, finland 3, 473–481. Wang, Y.H., 1975. Zeta potential studies on hydroxyapatite. Columbia University, New York. Watanabe, H., Seto, J., 1990. The intrinsic equilibrium constants of the surface hydroxyl groups of maghemite and hematite. Bull. Chem. Soc. Jpn. Weedon, D., Grano, S., Akroyd, T., Goncalves, K., Moura, R., 2007. Effects of high magnesium ion concentration on KCl flotation: Part II - Plant research. Miner. Eng. 20, 716–721. Westall, J., Hohl, H., 1980. A comparison of electrostatic models for the oxide/solution interface. Adv. Colloid Interface Sci. 12, 265–294. Wiese, G.R., Healy, T.W., 1975a. Adsorption of AI (III) at the TiO₂-H₂O Interface. J. Colloid Interface Sci. 51, 434–442. Wiese, G.R., Healy, T.W., 1975b. Solubility effects in Al₂O₃ and TiO₂ colloidal dispersions. J. Colloid Interface Sci. 52, 452–457. Wolstenholme, G.W., Schulman, J.H., 1950. Metal-monolayer interactions in aqueous systems   151 Part 1. The interaction of monolayers of long- chain polar compounds with metal ions in the underlying solution. Trans. Faraday Soc. 46, 475. Wood, L.A., 1946. The measurement of the potential at the interface between vitreous silica and pure water. J. Am. Chem. Soc. 68, 437–441. Wu, C., Wang, L., Harbottle, D., Masliyah, J., Xu, Z., 2015. Studying bubble–particle interactions by zeta potential distribution analysis. J. Colloid Interface Sci. 449, 399–408. Xing, Y., Gui, X., Cao, Y., 2016. Effect of calcium ion on coal flotation in the presence of kaolinite clay. Energy and Fuels 30, 1517–1523. Xu, C. yang, Deng, K. ying, Li, J. yu, Xu, R. kou, 2015. Impact of environmental conditions on aggregation kinetics of hematite and goethite nanoparticles. J. Nanoparticle Res. 17, 1–13. Xu, C.Y., Xu, R.K., Li, J.Y., Deng, K.Y., 2017. Phosphate-induced aggregation kinetics of hematite and goethite nanoparticles. J. Soils Sediments 17, 352–363. Xu, Z., Chi, R., Difeo, T., Finch, J.A., 2000. Surface forces between sphalerite and silica particles in aqueous solutions. J. Adhes. Sci. Technol. 14, 1813–1827. Xu, Z., Liu, J., Choung, J.W., Zhou, Z., 2003. Electrokinetic study of clay interactions with coal in flotation. Int. J. Miner. Process. 68, 183–196. Xu, Z., Yoon, R.H., 1990. A study of hydrophobic coagulation. J. Colloid Interface Sci. 134, 427–434. Xu, Z., Yoon, R.H., 1989. The role of hydrophobia interactions in coagulation. J. Colloid Interface Sci. 132, 532–541. Yang, S., Xie, B., Lu, Y., Li, C., 2018. Role of magnesium-bearing silicates in the flotation of pyrite in the presence of serpentine slimes. Powder Technol. 332, 1–7. Yao, W., Millero, F.J., 1996. Adsorption of phosphate on manganese dioxide in seawater. Environ. Sci. Technol. 30, 536–541. Yernberg, W.R., 1985. Mr. Weld phosphate beneficiation study. Yoon, R.H., Flinn, D.H., Rabinovich, Y.I., 1997. Hydrophobic interactions between dissimilar surfacesYoon,. J. Colloid Interface Sci. 185, 363–370. Yopps, J.A., Fuerstenau, D.W., 1964. The zero point of charge of alpha-alumina. J. Colloid Sci. 19, 61–71. Yotsumoto, H., Yoon, R., 1993a. Application of extended DLVO theory: I. Stability of rutile suspensions. J. Colloid Interface Sci. 157, 426–433. Yotsumoto, H., Yoon, R., 1993b. Application of extended DLVO theory: II. Stability of silica suspensions. J. Colloid Interface Sci. 157, 434–441. Yu, Y., Ma, L., Cao, M., Liu, Q., 2017. Slime coatings in froth flotation: A review. Miner. Eng. 114, 26–36. Yuehua, H., Lin, L., Guangzhou, Q., 1995. Solution chemistry of electrokinetic behavior of carbonate minerals. Trans. Nonferrous Met. Soc. China J. 5, 26–30. Zafar, Z.I., Anwar, M.M., Pritchard, D.W., 1995. Optimization of thermal beneficiation of a low grade dolomitic phosphate rock. Int. J. Miner. Process. 43, 123–131. Zhang, P., Albarelli, G.E., Stewart, K., 1999. Phosphate beneficiation bibliography. Florida Institute of Phosphate Research. Zhang, P., Yu, Y., Bogan, M., 1997. Challenging the “Crago” double float process II. Amine-fatty acid flotation of siliceous phosphates. Miner. Eng. 10, 983–994. Zhao, H., Long, J., Masliyah, J.H., Xu, Z., 2006. Effect of divalent cations and surfactants on silica-bitumen interactions. Ind. Eng. Chem. Res. 45, 7482–7490.   152 Zhong, K., Vasudevan, T. V., Somasundaran, P., 1993. Floatability of apatites of different type and origin: role of surface area and porosity. Int. J. Miner. Process. 38, 177–188. Zhou, X., Feng, B., 2015. The effect of polyether on the separation of pentlandite and serpentine. J. Mater. Res. Technol. 4, 429–433. Zhu, Y., Free, M.L., Woollam, R., Durnie, W., 2017. A review of surfactants as corrosion inhibitors and associated modeling. Prog. Mater. Sci. 90, 159–223.                        153 Appendices Appendix A  Determination of the composition of samples used for zeta potential measurements based on AAS results Table A. 1 Measurement of the volume of apatite (µl/L) in the tested mixtures used for zeta potential distribution  Test Solution  Calcium concentration in sample (mg/L)Apatite concentration in sample (mg/L)Apatite volume in sample (µl/L)Test 1 - pH 7 Background 2.64Sample 4.16Corrected value 1.53 3.83 1.20Test 2 - pH 7 Background 2.50Sample 5.01Corrected value 2.50 6.27 1.96Test 3 - pH 7 Background 2.81Sample 3.70Corrected value 0.89 2.22 0.70Test 4 - pH 7 Background 2.64Sample 5.18Corrected value 2.54 6.37 1.99Test 5 - pH 7 Background 3.02Sample 4.63Corrected value 1.61 4.02 1.26Test 1 - pH10 Background 0.38Sample 3.26Corrected value 2.88 7.22 2.26Test 2 - pH 10 Background 0.32Sample 2.65Corrected value 2.33 5.83 1.82Test 3 - pH 10 Background 0.48Sample 1.90Corrected value 1.43 3.57 1.12Test 4 - pH 10 Background 3.88Sample 0.33Corrected value 3.54 8.88 2.77Test 5 - pH 10 Background  0.57Sample 2.21Corrected value 1.64 4.10 1.28  154 Table A. 2 Measurement of the volume of hematite (µl/L) in the tested mixtures used for zeta potential distribution measurements.  Table A. 3 Calculation of the amount of hematite and apatite (%vol) in the tested mixtures used for zeta potential distribution measurements.    Test Solution  Iron concentration in sample (mg/L)Hematite concentration in sample (mg/L)Hematite volume in sample (µl/L)Test 1 - pH 7 1 32.97 47.14 8.89Test 2 - pH 7 2 42.35 60.55 11.42Test 3 - pH 7 3 47.23 67.52 12.74Test 4 - pH 7 4 38.79 55.46 10.47Test 5 - pH 7 5 51.69 73.91 13.94Test 1 - pH10 6 43.70 62.48 11.79Test 2 - pH 10 7 49.30 70.49 13.30Test 3 - pH 10 8 55.31 79.08 14.92Test 4 - pH 10 9 46.90 67.06 12.65Test 5 - pH 10 10 56.24 80.40 15.17Test Solution Apatite volume (µl/L)Hematite volume (µl/L)Total  volume (µl/L)Hematite (%vol)Apatite (%vol)Test 1 - pH 7 1 1.20 8.89 10.09 88.14 11.86Test 2 - pH 7 2 1.96 11.42 13.39 85.35 14.65Test 3 - pH 7 3 0.70 12.74 13.44 94.83 5.17Test 4 - pH 7 4 1.99 10.47 12.45 84.02 15.98Test 5 - pH 7 5 1.26 13.94 15.20 91.73 8.27Test 1 - pH10 6 2.26 11.79 14.05 83.94 16.06Test 2 - pH 10 7 1.82 13.30 15.12 87.94 12.06Test 3 - pH 10 8 1.12 14.92 16.04 93.04 6.96Test 4 - pH 10 9 2.77 12.65 15.43 82.01 17.99Test 5 - pH 10 10 1.28 15.17 16.45 92.20 7.80  155 Appendix B  Calibration curves for calcium and iron concentrations using AAS Using AAS, iron and calcium ions were assayed at the wavelength of respectively 248.33 nm and 422.67 nm, the most sensitive and commonly used primary wavelengths. The calibration curves followed Beer’s Law at low concentrations of calcium and iron (0-5 mg/L). Deviations from linearity were observed in calibration curves at high concentrations.  Certified stock iron (Fe) and calcium solutions (1 g/L) for calibration were purchased from Fisher Scientific and used for preparing the standards. For direct measurements of hematite coating on coarse apatite crystal (slime coatings), deionized water was used for preparing iron standards. For determining the composition of tested mixtures used for zeta potential measurements using AAS, calcium and iron standards were made by 0.01 mol/L NaCl background electrolyte because the experiments were conducted in background electrolyte and therefore all samples contained 0.01 mol/L NaCl.     Figure B. 1 Absorbance versus concentration for calcium and iron at the wavelength of respectively 422.67 nm and 248.33 nm.     y = 0.082xR² = 0.99470.000.050.100.150.200.250.300.350 1 2 3 4 5AbsorbanceConcentration of standard solution (mg/L)Calcium- 422.67 nmy = 0.0606xR² = 0.99940.000.050.100.150.200.250.300.350 1 2 3 4 5AbsorbanceConcentration of standard solution (mg/L)Iron-248.33 nm

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0388358/manifest

Comment

Related Items