Hydrophobic Surface State of Talc and How It Is

HYDROPHOBIC SURFACE STATE OF TALC AND HOW IT IS

INFLUENCED BY POLYSACCHARIDE ADSORPTION

Venkata Veeren Babu Atluri

A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Department of Metallurgical Engineering

The University of Utah

May 2019

The University of Utah Graduate School

STATEMENT OF DISSERTATION APPROVAL

The dissertation of Venkata Veeren Babu Atluri has been approved by the following supervisory committee members:

Jan D Miller , Chair 01/15/2019 Date Approved

Xuming Wang , Member 01/15/2019 Date Approved

Michael L. Free , Member 01/15/2019 Date Approved

Vladimir Hlady , Member 01/15/2019 Date Approved

York R. Smith , Member 01/15/2019 Date Approved

and by Michael Simpson , Chair/Dean of the Department/College/School of Metallurgical Engineering and by David B. Kieda, Dean of The Graduate School. ABSTRACT

Talc is both an important industrial mineral product recovered by flotation, and also in other cases, a gangue mineral of concern in the flotation of certain sulfide ores, such as the PGM ores in South Africa and in the United States. The talc face surface is naturally hydrophobic with a contact angle of nearly 80º, which accounts for its flotation recovery in one case, and its contamination of sulfide mineral concentrates in other instances. High- quality talc structures were investigated using surface analysis techniques including contact angle analysis, high-speed video bubble attachment measurements, atomic force microscopy, molecular dynamics simulation (MDS), microflotation, and film thickness measurements by Synchronized Triwavelength Reflection Interferometry Microscopy

(STRIM).

The presence of aluminum, which replaces silicon in the silica tetrahedral layer of the talc structure, results in a charge imbalance on the face surface because Si+4 is replaced by Al+3. Experimental sessile drop contact angles were found to decrease with increased aluminum content, decreasing from about 80º for no substitution (talc) to 0º for extensive substitution (phlogopite). For a hydrophilic phlogopite surface, the water film is stable with an equilibrium film thickness (he) of 25 nm. However, for a hydrophobic talc surface, air bubbles readily attach to the talc face surface with a critical rupture thickness (hc) of 56 nm.

Further, the wetting characteristics and water film stability at the talc surface have been studied, regarding the effect of polysaccharides such as guar gum, starch, and dextrin.

In the presence of polysaccharides, there was a significant increase in bubble attachment time at the talc surface but only a slight change in contact angle, which suggests that polysaccharide depression of talc was due to the slow rate of bubble attachment and not due to a change in contact angle. The adsorption state of the polysaccharides can be described as being due to a hydrophobic interaction between the nonpolar mineral surface and the hydrophobic portion of the polysaccharide molecule. Interestingly, it was found that the critical and equilibrium film thickness values do not change significantly with the polysaccharide type or concentration.

Dedicated to my family and friends

TABLE OF CONTENTS

ABSTRACT ...... iii

LIST OF TABLES ...... ix

LIST OF FIGURES ...... xi

ACKNOWLEDGMENTS ...... xiv

Chapters

1. INTRODUCTION ...... 1

1.1 Flotation of Sulfide Minerals ...... 2 1.2 Phyllosilicate Minerals ...... 6 1.3 Background on Talc ...... 11 1.4 Polysaccharide Depressants ...... 13 1.5 Research Objectives ...... 17 1.6 Dissertation Organization ...... 18

2. WETTING CHARACTERISTICS OF TALC ...... 21

2.1 Introduction ...... 21 2.2 Materials and Methods...... 24 2.2.1 Minerals and Reagents ...... 24 2.2.2 X-Ray Photoelectron Spectroscopy ...... 24 2.2.3 Experimental and Simulation Sessile Drop Contact Angle ...... 25 2.2.4 MDS Interfacial Water Analysis ...... 26 2.2.5 Experimental and Simulation Bubble Attachment Time ...... 26 2.2.6 Synchronized Triwavelength Reflection Interferometry Microscopy ... 29 2.2.7 Atomic Force Microscopy ...... 31 2.3 Sessile Drop Contact Angle Results ...... 31 2.3.1 Experimental ...... 32 2.3.2 MD Simulation...... 32 2.4 MDS Interfacial Water Analysis...... 34 2.4.1 Relative Number Density ...... 37 2.4.2 Water Dipole Moment and Hydrogen Bonding ...... 37 2.5 Bubble Attachment Results ...... 40 2.5.1 Experimental ...... 40 2.5.2 MD Simulation...... 42 2.6 Film Thickness Results ...... 44 2.6.1 Synchronized Triwavelength Reflection Interferometry Microscopy ... 44 2.6.2 Discussion ...... 46 2.7 Atomic Force Microscopy Results ...... 49 2.8 Summary ...... 51

3. SIGNIFICANCE OF ALUMINUM SUBSTITUTION...... 53

3.1 Introduction ...... 53 3.2 Materials and Methods...... 55 3.2.1 Minerals and Reagents ...... 55 3.2.2 X-Ray Photoelectron Spectroscopy ...... 57 3.2.3 Experimental and Simulation Sessile Drop Contact Angle ...... 57 3.2.4 MDS Interfacial Water Analysis ...... 58 3.2.5 Experimental and Simulation Bubble Attachment Time ...... 60 3.3 X-ray Photoelectron Spectroscopy Results...... 60 3.4 Sessile Drop Contact Angle Results ...... 62 3.4.1 Experimental ...... 62 3.4.2 MD Simulation...... 65 3.5 Interfacial Water Analysis ...... 67 3.5.1 Relative Number Density ...... 75 3.5.2 Water Dipole Moment and Hydrogen Position ...... 77 3.6 Bubble Attachment Results ...... 79 3.6.1 Experimental ...... 79 3.6.2 MD Simulation...... 81 3.7 Summary ...... 87

4. ADSORPTION OF POLYSACCHARIDES ON TALC ...... 90

4.1 Introduction ...... 90 4.2 Materials and Methods...... 93 4.2.1 Minerals and Reagents ...... 93 4.2.2 Captive Bubble Contact Angle ...... 94 4.2.3 Bubble Attachment Time ...... 94 4.2.4 Atomic Force Microscopy (AFM) ...... 95 4.2.5 Film Thickness Measurements by Synchronized Triwavelength Reflection Interferometry Microscopy (STRIM)...... 95 4.2.6 Microflotation ...... 96 4.3 Bubble Attachment Results ...... 96 4.3.1 Captive Bubble Contact Angle ...... 97 4.3.2 Bubble Attachment Time ...... 97 4.4 Film Thickness Results ...... 101 4.4.1 Synchronized Triwavelength Reflection Interferometry Microscopy . 101 4.4.2 Discussion ...... 109 4.5 Adsorption State ...... 110 4.5.1 Atomic Force Microscopy ...... 113 4.5.2 Microflotation ...... 117 4.6 Summary ...... 121

vii 5. CONCLUSIONS AND RECOMMENDATIONS ...... 125

REFERENCES ...... 129

viii LIST OF TABLES

Tables

1.1. Various ores treated by flotation and corresponding concentrates produced in the United States (in million metric tons) ...... 3

1.2. Tonnage of mineral resources treated by flotation and corresponding reagents used in the United States mining industry (in metric tons) ...... 5

1.3. World mine production of talc mineral products for 2016 ...... 14

2.1. CLAYFF force field parameters ...... 27

2.2. Contact angle and bubble attachment time for talc and phlogopite surfaces ...... 33

2.3. Contrast between water film thickness values at talc and phlogopite surfaces ...... 47

3.1. CLAYFF force field parameters ...... 59

3.2. XPS surface analysis of talc samples in atomic percent ...... 61

3.3. Experimental advancing sessile drop contact angles for talc samples and phlogopite ...... 64

3.4. MD simulated advancing sessile drop contact angles for crystals with different levels of aluminum substituted in the tetrahedral layer of the face surface ...... 66

3.5. Experimental bubble attachment time for crystals with different levels of aluminum substituted in the tetrahedral sheet of the face surface ...... 80

3.6. Summarized data on interfacial water analysis at the talc structures with increasing aluminum substitution ratio ...... 89

4.1. Bubble attachment with guar gum adsorbed at the talc face surface ...... 98

4.2. Bubble attachment with corn starch adsorbed at the talc face surface ...... 99

4.3. Bubble attachment with dextrin adsorbed at the talc face surface ...... 100

4.4. Characteristics of wetting films on talc surfaces at different guar gum concentrations ...... 105 4.5. Characteristics of wetting films on talc surfaces at different corn starch concentrations ...... 107

4.6. Characteristics of wetting films on talc surfaces at different dextrin concentrations ...... 108

4.7. Bubble attachment time for HOPG with increasing corn starch concentration ...... 114

4.8. Bubble attachment time for talc with increasing corn starch concentration ...... 115

LIST OF FIGURES

Figures

1.1. Crystal structure of antigorite (left) and kaolinite (right), talc (A) and pyrophyllite (B), muscovite (C), phlogopite (D) ...... 7

1.2. Crystal structure of talc...... 12

1.3. Schematic showing the structure of various polysaccharides used for flotation in the mineral processing industry...... 16

2.1. MDS snapshots of equilibrated water/talc surfaces. (A) Shows structure of water at basal plane, (B) and (C) are water structures at edge surfaces viewed from different angles...... 22

2.2. Schematic of the Synchronized Triwavelength Reflection Interferometry Microscopy (STRIM) setup...... 30

2.3. Snapshots of the advancing sessile drop contact angle at the surface with no aluminum substitution for silicon in the tetrahedral face surface for MDS Talc 1 at 0, 0.5, and 1 nanosecond...... 35

2.4. MDS snapshots (two different directions) for interfacial water analysis at the unsubstituted surface (MDS Talc 1), aluminum ratio = 0...... 36

2.5. Relative number density profiles at the unsubstituted surface of MDS Talc 1 (talc 001 surface), aluminum ratio = 0...... 38

2.6. Schematic illustration describing the orientation of a water dipole moment by the angle (α) and hydrogen position relative to the surface normal by the angle (β)...... 39

2.7. Water dipole moment and hydrogen position at the talc face surface...... 41

2.8. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of an unsubstituted ideal talc crystal (MDS Talc 1) for interaction times of 0, 0.1, 0.5, and 1 nanosecond...... 43

2.9. A comparison of spatiotemporal thickness profile of wetting films formed on a) fresh talc surface and b) fresh phlogopite surface in 1 mM NaCl solution...... 45

2.10. AFM nanobubbles at a fresh talc surface...... 50

3.1. Crystal structures of talc, impure talc, and phlogopite with aluminum/total tetrahedral site ratios ranging from 0 to 0.25...... 56

3.2. Snapshots of the advancing sessile drop contact angle at the surface with no aluminum substitution for silicon in the tetrahedral face surface for MDS Talc 1 at 0, 0.5, and 1 nanosecond...... 68

3.3. Snapshots of the advancing sessile drop contact angle at the surface with aluminum substitution (0.013) for silicon in the tetrahedral face surface for MDS Talc 2 at 0, 0.5, and 1 nanosecond...... 69

3.4. Snapshots of the advancing sessile drop contact angle at the surface with aluminum substitution (0.08) for silicon in the tetrahedral face surface for MDS Talc 3 at 0, 0.5, and 1 nanosecond...... 70

3.5. Snapshots of advancing sessile drop contact angle at the face surface of the crystal which has aluminum substituted for every fourth silicon (0.25) in the tetrahedral face surface for MDS Talc 4 (phlogopite) at 0, 0.5, and 1 nanosecond...... 71

3.6. Snapshots of advancing sessile drop contact angle at the face surface of the crystals, MDS Talc 1, MDS Talc 2, MDS Talc 3, and MDS Talc 4 with increasing aluminum substitution ratio at 1 nanosecond...... 72

3.7. Sessile drop contact angle as a function of aluminum lattice substitution in tetrahedral layer for the talc face surface as determined from experimental measurements and from molecular dynamics simulations...... 73

3.8. MDS snapshots (two different directions) for interfacial water analysis at MDS Talc 1, aluminum ratio = 0, MDS Talc 2, aluminum ratio = 0.013, MDS Talc 3, aluminum ratio = 0.08, and MDS Talc 4, phlogopite, aluminum ratio = 0.25...... 74

3.9. Relative number density profiles at the MDS Talc 1, aluminum ratio = 0, MDS Talc 2, aluminum ratio = 0.013, MDS Talc 3, aluminum ratio = 0.08, and MDS Talc 4, phlogopite, aluminum ratio = 0.25...... 76

3.10. Water dipole moment and hydrogen position of MDS Talc 1, aluminum ratio = 0, MDS Talc 2, aluminum ratio = 0.013, MDS Talc 3, aluminum ratio = 0.08, and MDS Talc 4, phlogopite, aluminum ratio = 0.25...... 78

3.11. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of an unsubstituted ideal talc crystal (MDS Talc 1) for interaction times of 0, 0.1, 0.5, and 1 nanosecond...... 82

3.12. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of MDS Talc 1, aluminum ratio = 0.013 for interaction times of 0, 0.1, 0.5, and 1 nanosecond...... 83

3.13. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of

xii MDS Talc 3, aluminum ratio = 0.08 for interaction times of 0, 0.1, 0.5, and 1 nanosecond...... 84

3.14. Sequence of MDS snapshots for nitrogen bubble at the face surface of phlogopite (MDS Talc 4), with extensive aluminum substitution for interaction times of 0, 0.1, 0.5, and 1 nanosecond...... 86

4.1. Adsorption isotherms for guar gum, starch, and dextrin adsorption by talc...... 102

4.2. Presence of water droplets inside the attached air bubble with dextrin adsorbed at the talc surface (magnification 10x) ...... 103

4.3. Captive bubble contact angle at HOPG surface with varying pH and varying corn starch concentration ...... 111

4.4. Captive bubble contact angle at the talc surface with varying pH and varying corn starch concentration ...... 112

4.5. TMAFM height and phase image of fresh HOPG (A, B) and HOPG conditioned with 50 ppm corn starch (C, D). Corn starch adsorbs at HOPG as a cellular branched chain structure...... 116

4.6. TMAFM height and phase image of fresh talc (A, B) and talc conditioned with 50 ppm corn starch (C, D). Corn starch adsorbs at the (001) talc surface in the form of elongated patches ...... 118

4.7. Contact mode height (A) and deflection error (B) images of fresh talc conditioned with 30 ppm guar gum. Guar gum adsorbs at the (001) talc surface in the form of cylindrical patches ...... 119

4.8. TMAFM height (A) and phase (B) images of fresh talc conditioned with 50 ppm dextrin. Dextrin adsorbs at the (001) talc surface in the form of small network of islands ...... 120

4.9. Flotation recovery of synthetic graphite in the presence and absence of corn starch with varying pH ...... 122

4.10. Flotation recovery of talc in the presence and absence of corn starch with varying pH ...... 123

xiii

ACKNOWLEDGMENTS

First, I am grateful to be under Dr. Jan D Miller’s tutelage during my dissertation research. Dr. Jan Miller is the epitome of hard work, dedication, and perseverance. His knowledge and experience in the field of mineral processing and flotation enriched my skills, and helped me pursue a research-oriented career. The amount of discussions with him will be cherished throughout my whole life. I have gained good traits, leadership qualities, and humility which helped me in my career development, and also being a better person.

Second, I would like to thank Dr. Xuming Wang for his amazing support throughout my experimental research work, teaching me policies and procedures in the lab.

His support was paramount in order to ensure smooth running of the experimental work.

Third, I am thankful to Yuesheng Gao and Dr. Lei Pan for their help with the instrumentation at the Michigan Technological University. I appreciate the help from Dr.

Jiaqi Jin in the simulation work and the collaboration from Dr. Kaustubh Shrimali for the work that was done at University of Utah. I would also like to thank Dr. Michael L. Free,

Dr. Vladimir Hlady, and Dr. York Smith who have agreed to be on my committee and give their critical observation when reviewing my work. I would like to take this opportunity to thank Dorrie Spurlock for her unending assistance in the proof reading and submissions of my manuscripts.

Fourth, I would like to thank the hard work of the administrative staff of the

Department of Metallurgical Engineering. Also, I was lucky to have an amazing group of people to work with in the lab and amazing friends that I had spent my time with during my dissertation study as a graduate student at University of Utah.

Last but not the least, I would like to thank my parents, my uncle Dr. Yalamanchili

Madhava Rao, and Dr. Veeramasuneni Srinivas for their trust in me. They have put faith in me and gave their continuous support and guidance throughout my dissertation study.

CHAPTER 1

INTRODUCTION

Mineral processing is an important part of extractive metallurgy. Minerals are natural inorganic substances, some of which are of significant economic value and are extracted from the earth’s crust. When these minerals exist in ore deposits in sufficient quantities, they can be profitably recovered with appropriate processing technology.

Flotation technology is one of the most important particle separation technologies used in the mineral processing industry.1 The flotation separation is based on the selective control of the surface state of the mineral components of the ore. Typically, the valuable mineral components are made hydrophobic, distinct from other components which are maintained in a hydrophilic state. These hydrophobic mineral particles attach to air bubbles in the flotation tank and float to the top, leaving behind the invaluable/gangue minerals. Most ores containing sulfide minerals are processed by flotation to recover the sulfide minerals which contain valuable metals like copper, iron, nickel, etc. Surface chemistry phenomena control the flotation separation process for mineral systems. The basis of flotation technology for particle separation is preferred due to the low cost, where thousands of tons of sulfide mineral are floated and recovered all over the world. The current research is focused on the fundamental issues associated with polysaccharide depression of the gangue mineral talc in the flotation of sulfide minerals, and its use in various industrial processes.

1.1 Flotation of Sulfide Minerals

Historically, there has been a continual need for reducing costs in copper ore production as well as in the processing of important metal sulfide ores. By 1980 there was a huge increase in the need for metals and other commodities. The increase in production of various sulfide minerals during the past decade is shown in Table 1.1.1 During the development of more recent flotation processes, use of various reagents like oils, coal, tar, petrol, and pine oils in sulfide flotation were eliminated. Due to the introduction of flotation, millions of tons of waste rock have been processed and converted into valuable metals. In the past several decades, sulfide minerals were the major mineral class that has been enriched by the flotation process.

To maximize the profitability and to increase grade and recovery, a number of different reagents are used in the flotation process.2 The flotation reagents can be classified into six categories: collectors, depressants, frothers, modifiers, activators, and flocculants.

Collectors adsorb at the mineral surfaces making them hydrophobic, which makes them easier to attach to the air bubbles and float to the top of the tank in the flotation process.

Depressants are the class of reagents that work opposite to the collectors, where they depress the unwanted gangue minerals in flotation, and valuable minerals in the case of reverse flotation. Frothers are the reagents added to improve the dispersion and stability of bubbles, and account for the formation of a stable froth phase. Modifiers are compounds added to the flotation tank to vary the pH and control the surface charge of minerals as some collectors and depressants are charge sensitive. Activators adsorb on the mineral surfaces to enhance the collector adsorption. Flocculants are generally water soluble macromolecules added to the flotation tank which results in the formation of flocs, which

Table 1.1. Various ores treated by flotation and corresponding concentrates produced in

the United States (in million metric tons).1

1926 1960 1980 Type of ore Treated Concentrate Treated Concentrate Treated Concentrate Copper 39.89 2.17 133.38 4.82 211.61 4.67 Lead-zinc 5.57 0.84 7.43 0.49 11.39 0.84 Gold- silver 0.48 0.03 0.12 0.003 0.1 0.005 Iron 1.39 0.54 37.88 21.48 Phosphate 19.03 6.37 108.7 26.63 Potash 10.87 2.83 12.93 2.99 Coal 3.73 2.54 11.7 6.86 Feldspar- mica 1.67 1.06 11.58 8.51 Industrial minerals 2.23 0.83 0.58 0.37

4 helps in settling, and also added to the dewatering plants for thickening. Examples of the flotation reagents used in the United States are presented in Table 1.2.1

Early research on the sulfide minerals was conducted particularly on sphalerite, chalcopyrite, pyrite, and chalcocite.3 Sulfide minerals are readily oxidized when they are exposed to the atmosphere. Thus, they become hydrophilic and requires collector addition to enhance the flotation recovery. But, in the case of sulfide mineral with fresh surfaces, they have a natural flotability due to the low polarity of the surface atoms and/or the crystal structure. For example, molecular crystal of sulfur, S8, is held together by van der Waals bonds.4 Two examples of sulfide minerals that exhibit natural hydrophobicity are stibnite and molybdenite which have a layered structure.5

Flotation chemistry research has been performed by various researchers all over the world during past decades to study the wettability of the mineral surfaces including sulfide minerals and layered silicate minerals.6-16 One of the main issues in the flotation of sulfide minerals from their ores is the presence of layered silicates which are hydrophobic minerals such as talc and pyrophyllite. Being hydrophobic, these minerals report to the sulfide concentrate, which causes problems in the downstream processes, and also reduce the grade and recovery of sulfide minerals. The economics of the plant will take a huge dent due to the presence of these hydrophobic layered silicates. Since these ores to be separated may contain such layered silicate minerals as talc, which is naturally hydrophobic, removing it from the sulfide concentrates is a main concern in most flotation plants around the world. Notable plants where the talc is a major issue are in PGM ores of South Africa, as well as in other plants in the United States and elsewhere. The current dissertation research focuses on the surface state of talc and wettability characteristics.

Table 1.2.Tonnage of mineral resources treated by flotation and corresponding reagents

used in the United States mining industry (in metric tons).1

1925 1980 Ore Treated 41,259,000 440,361,000

Reagent Use (metric tons) Frothers 2,195 12,489 Collectors Oils 8,818 115,218 Chemicals 1,875 108,883 Modifiers Acids 18,157 35,169 Alkalis 1,695 413,055 Other 28,735 Activators 3,210 3,925 Depressants 724 33,389 Flocculants 18,069

1.2 Phyllosilicate Minerals

Phyllosilicate minerals are also known as the layered silicates that contain a silica tetrahedral layer and a magnesia/alumina octahedral layer. Their classification is based on the ratio of silica and alumina/magnesia layers, namely 1:1 phyllosilicates (bi-layer) and

2:1 phyllosilicates (tri-layer). The crystal structure and surface state of the phyllosilicate minerals determine their wetting characteristics. The classification is based on the number of octahedra occupied, and the octahedral sheets are classified into dioctahedral and trioctahedral sheets. Most of these layered silicate minerals have perfect basal cleavage, i.e., the plane perpendicular to the crystal axis. Examples of the phyllosilicate mineral structures are presented in Figure 1.1.17 Some of these minerals are valuable to a number of manufacturing industries such as cosmetics, paper, textile, catalyst, rubber, and polymer.18-20 Conversely, a major portion of these phyllosilicate minerals present a problem in the mineral processing industry, mainly in flotation and downstream processes.

Phyllosilicates, such as talc and pyrophyllite, are hydrophobic and report to the concentrate reducing the grade and recovery of the valuable minerals. Although minerals have specific chemical composition and atomic structure, they exhibit isomorphism, where other atoms substitute for existing atoms or polymorphism, where the minerals have same elemental composition but different properties. For example, Si+4 in the tetrahedral layer of talc is substituted by Al+3. The magnitude of this substitution usually determines the surface state of the phyllosilicate minerals.21 Further information on the phyllosilicate minerals can be found in the literature.22, 23 The current dissertation research will provide a comprehensive study on talc, a phyllosilicate mineral which is an important gangue mineral in some industrial sulfide mineral flotation processes.

Figure 1.1. Crystal structure of antigorite (left) and kaolinite (right), talc (A) and pyrophyllite (B), muscovite (C), phlogopite (D). Atom codes: red, OH; blue, Al; yellow, O; green, Mg; black, Si.17

Figure 1.1. (Continued)

1.3 Background on Talc

Talc is a 2:1 trioctahedral phyllosilicate with tetrahedral layers composed of silicon and oxygen, and octahedral layers composed of magnesium, oxygen, and hydroxyls. Talc is among the softest phyllosilicates due the fact that these T-O-T layers are held together by weak van der Waals forces which in turn also imparts talc its natural hydrophobicity.

The color variation is extreme based on the impurities present in talc, mainly aluminum impurity which will be discussed in Chapter 3. Talc is hydrophobic and easily floatable as the hydroxyl groups and the active sites are absent on the basal plane face surfaces. In contrast to the basal plane or the face surface of talc, its edge surface has a hydrophilic nature. This hydrophilicity of the edges can be explained due to the presence of hydroxyl groups formed by the breakage of the Si-O or Mg-O bonds. Talc is a 2:1 phyllosilicate mineral containing layers of silica tetrahedron and magnesia octahedron and is described as a hydrated magnesium silicate having the chemical formula Mg3Si4O10(OH)2 with a structure as shown in Figure 1.2.

In the United States, 660 thousand tons of talc valued at $19.1 million was produced by three different companies that operated during 2016. The leading producer of talc was

Montana which was followed by Texas and Vermont. In regard to the sales of talc, 545 thousand tons of talc were sold domestically and exported with an estimated revenue of

$97.5 million, which was a slight decline from the sales made in 2015. The uses of talc that was produced in the United States are as follows:24

 Ceramics, including automotive catalytic converters (26%)

 Paper (18%)

 Paint (17%)

Figure 1.2. Crystal structure of talc. The atom codes are: empty circles, O; black circles, Si; gray circles, OH; black triangles, Mg; black squares, Al.

 Unclassified end uses (13%)

 Plastics (12%)

 Roofing (7%)

 Rubber (4%)

 Cosmetics (3%)

Also, an estimated 385 thousand tons of talc was imported in 2016 and 75% of it was consumed for cosmetics, paints and plastic. World production and reserves of talc are presented in Table 1.3.24 China (2,250,000 tons, includes other phyllosilicates), Brazil

(845,000 tons), and India (925,000 tons) are the largest producers of talc in the world.

The depression of talc is a significant issue in sulfide mineral flotation systems.

Minute quantities of talc that are present in the ore substantially reduce the sulfide mineral concentrate’s recovery and grade. High polysaccharide depressant addition produces smaller bubbles and a highly mineralized froth. The reduction in froth phase bubble size and stability results in lower gangue mineral recovery. In this regard, a fundamental study on the surface state of talc, significance of aluminum impurities, and the influence of polysaccharides for talc depression is focused in this research work.

1.4 Polysaccharide Depressants

The current dissertation research deals with the reagents used as depressants in the flotation of sulfide mineral systems such as dextrin, starch, CMC, and guar gum. The use of polysaccharides as depressants is not limited to the depression of talcaceous gangue; they are also used in the flotation of nonsulfide minerals and also in the reverse flotation of nonsulfide ores, such as iron ore. Although inorganic depressants have been traditionally

Table 1.3. World mine production of talc mineral products for 2016 (in thousand tons).24

Mine

production Origin 2016 United States (crude) 660 Brazil (crude and beneficiated)* 850 China (unspecified minerals) 2200 France (crude) 450 India* 925 Japan* 370 Korea* 610 Mexico 750 Other countries 1600 World total 8400 *Includes pyrophyllite

15 used as depressants, their toxic nature poses a threat to both humans and the environment.

Some examples are sodium hydrosulfide, sodium cyanide, arsenic trioxide, phosphorous pentasulfide, etc. Therefore, nowadays, organic depressants which are nontoxic and biodegradable are used when effective and in such cases have replaced inorganic depressants. They are relatively cheaper, can be stored in large quantities, and are more environmentally friendly. Some of the polysaccharides used in the depression of talc are dextrin, starch, carboxy methyl cellulose (CMC), and guar gum.

The term polysaccharide means the presence of numerous monomer sugars such as glucose, mannose and galactose as shown in Figure 1.3.25 In the case of starch, the basic building monomer is D-glucose. The D-glucose monomer is made up of five carbon atoms and one oxygen atom in a cyclical structure. Two orientations of this monomer are generally observed, where the ‘α’ and ‘β’ can be described as the “axial” and “equatorial” positions, respectively, which is based on the hydroxyl groups attached to carbon 1 (C1).

Starch is made up of α-D-glucose, and consists of two fractions called “amylose,” which has a straight chain, connected at C1 and C4, and “amylopectin,” which has both straight and branched chains connected at C1-C4 and C1-C6 (see Figure 1.3). The hydroxyl groups are also attached to carbons C2 and C3, and CH2OH to carbon C5 in the D-glucose monomer. Further, thermal degradation of starch under acidic conditions produces dextrin.

The molecular weight of dextrin is far less when compared to the molecular weight on starch since it is formed by the broken chains of starch and it represents amylopectin-like structure. In the case of guar gum, its structure containing β-D-mannose and α-D-galactose is related to a group of galactomannan. The straight chain of β-D-mannose is connected to

C1-C4 carbons which are in turn attached to α-D-glucose though a C1-C6 linkage.25 See

Figure 1.3. Schematic showing the structure of various polysaccharides used for flotation in the mineral processing industry.25 (Reprinted with permission)

Figure 1.3. The molecular weight of these polysaccharides also plays an important role in their depressing ability. Generally, higher molecular weight (MW) polysaccharides have better depressing ability as compared to the lower molecular weight (MW) polysaccharides. But high MW polysaccharides are less selective compared to low MW polysaccharides. Starch (MW 200,000 to 900,000 Daltons) and guar gum (MW 100,000 to

300,000 Daltons) have higher molecular weights as compared to dextrin (MW 6,000 to

15,000 Daltons) which is formed by the breakage of starch molecules.26-28 Due to the presence of carbon chains these polysaccharides are not fully soluble in water. Some of the methods used in the preparation of these solutions include acid or base addition and heating the solution to break down the molecules. Nevertheless, most of the depressants used for flotation in the mineral processing industry are engineered to have better solubility and/or dispersion in water. Most of these methods involve treating the polysaccharides with acid or base to hydrate them and improve their dispersion in water.

1.5 Research Objectives

The current dissertation research deals with the ideal talc face surface, the aluminum impurity in the silica tetrahedral layer of talc crystal, and the influence of polysaccharide adsorption on the wetting characteristics of the talc face surface, specifically, the polysaccharides such as starch, dextrin, and guar gum. The research consists of following objectives:

1. Experimental determination of contact angle, bubble attachment, and wetting

characteristics of the talc face surface with and without polysaccharides to evaluate

the effectiveness of different polysaccharides for talc depression. Of particular

interest is the nature of film rupture and displacement during bubble attachment

which has not been reported in the literature.

2. Molecular dynamics simulations (MDS) of the talc face surface to complement

traditional experimental results and to examine the effect of lattice defects,

specifically the effect of aluminum (Al) substitution in the silica tetrahedral layer,

on the wetting characteristics, which simulations have not been reported in the

literature.

3. Nature of the polysaccharide adsorption, particularly the influence on bubble

attachment and film thickness. This objective includes imaging the morphology of

polysaccharides at the talc surface using atomic force microscopy (AFM) to

characterize and describe the organization and structure of the adsorbed

polysaccharides. Although AFM images of dextrin and CMC at the talc surface

have been reported, not much imaging research is reported in the case of guar gum

and starch. Further, the influence of polysaccharides on bubble attachment time and

film thickness has not been reported in the literature.

1.6 Dissertation Organization

In Chapter 1, a basic introduction to the use of flotation in mineral processing is followed by a description and history of sulfide mineral flotation practice. The hydrophobic character of the sulfide minerals is explained along with the various reagents used in flotation processes. Then, the layered silicates, which are the gangue minerals in many sulfide flotation systems, are mentioned. This chapter concludes with a discussion on the background of talc and the use of polysaccharide talc depressants: starch, dextrin, and guar

19 gum.

In Chapter 2, the hydrophobic surface state of talc is studied using surface analysis techniques. XPS analysis is done to understand the quality of talc face surface.

Experimental contact angle and bubble attachment results are done and are complemented by molecular dynamics simulations (MDS) of water at the face surface of talc. Such simulations for high-quality talc surfaces have not been reported in the literature. Further,

AFM studies on the high-quality talc sample (Talc 1) mentions the presence of the nanobubbles at the face surface of the talc crystal. This is the first time the results from the

MDS nitrogen bubble attachments are presented at the talc surface.

In Chapter 3, molecular dynamics simulations (MDS) analysis of the talc face surface is presented along with experimental contact angle and bubble attachment time measurements at the surface of talc. These results are focused to help to examine the effect of lattice substitution. Specifically, the effect of aluminum substitution for silicon in the silica tetrahedral face on the wetting characteristics is reported. Such simulations for aluminum substituted talc samples have not been reported in the literature. It is expected that experimental results and simulation results will provide a further understanding of the hydrophobic surface state of talc as influenced by lattice substitution of aluminum in the tetrahedral layer. This is the first-time sessile drop contact angle simulation and bubble attachment simulation have been done on such aluminum substituted talc structures.

In Chapter 4, fundamental issues associated with film stability and bubble attachment at the talc face surface as influenced by polysaccharide depressants are presented. Examination of the effect of polysaccharides on water film thickness at the talc surface will help us understand talc depression and the efficient use of these depressants in

20 sulfide flotation systems. Many researchers have identified effective polysaccharides for talc depression. In this research work, the effect of polysaccharides on the talc surface chemistry is considered and further research including the measurement of film thickness using the recently developed STRIM analysis is presented. The results from this chapter help us understand the intricacies of film rupture and displacement during bubble attachment and explain some of the previous research reported in the literature on talc depressants. This is the first time film thickness measurements have been done at the talc face surface, phlogopite face surface, and at the talc face surface with adsorbed polysaccharides.

Conclusions and recommendations are presented in Chapter 5.

CHAPTER 2

WETTING CHARACTERISTICS OF TALC

2.1 Introduction

The hydrophobic surface state and the wetting characteristics of talc are of importance to understanding flotation response both when talc is floated for recovery as an industrial mineral product and when talc is depressed in the recovery of certain sulfide minerals as in the flotation of PGM ores from South Africa and United States. Flotation is a widely used method for concentrating sulfide ores, by separation and recovery of valuable sulfide mineral particles. As discussed in Chapter 1, talc is a 2:1 phyllosilicate mineral consisting of sheets of silica tetrahedra and magnesia octahedra and is described as a hydrated magnesium silicate having the chemical formula Mg3Si4O10(OH)2. Talc has a layered structure which is held together by weak van der Waals bonds. The basic structural element of talc is composed of one magnesia octahedral layer (brucite layer) sandwiched between two silica tetrahedral layers. The face surface of talc is of low polarity, and naturally hydrophobic with a contact angle of about 80º due to the absence of hydrogen bonding sites. The hydrophobic talc face surface has been established in previous research as reported in the literature. In contrast, the edge surface is less hydrophobic. Molecular dynamics simulation (MDS) reveals the difference in interfacial water structure at face and edge surfaces of talc as shown in Figure 2.1.29 Also, it is reported that lattice impurities

Figure 2.1. MDS snapshots of equilibrated water/talc surfaces. (A) Shows structure of water at basal plane, (B) and (C) are water structures at edge surfaces viewed from different angles. 29 (Reprinted with permission) 23 affect the wetting characteristics of the talc surface.30 This topic which will be discussed later in Chapter 3.

Traditional experimental analysis of the wetting characteristics of talc surfaces is accomplished by sessile drop and captive bubble contact angle measurements which provide a basic understanding of the surface state.31 Such results are presented for the high- quality talc surface (Talc 1) used in this research. These results are complemented by MD contact angle and bubble attachment simulations, as well as interfacial water analysis, which are not previously reported in the literature. In addition to these “equilibrium” measurements of contact angle, bubble attachment time measurements are reported to describe the hydrophobic surface state from a kinetic view. Finally, film thickness values of the thin liquid film present between the mineral surface and the approaching air bubble are reported in the case of bubble attachment.

The importance of thin liquid films in various fields including flotation chemistry is reported in the literature by various researchers.32-34 In addition to the wide application in flotation fundamentals, the thin film studies are particularly important in various research studies like bitumen extraction, nanochannel flow technology, and in medical drug delivery.35-39 Most film thickness measurements by researchers are accomplished by interferometry based on the passing of light through the film to produce interference fringe patterns. In the current research, an improved method which accurately describes the thickness of a thin liquid film called Synchronized Triwavelength Interferometry

Microscopy (STRIM) was utilized.

In addition, the mystery of the presence of nanobubbles on the surface of hydrophobic surfaces has been of interest to many researchers.40-42 One of the novel and 24 interesting works on the wetting characteristics of talc is the presence and imaging of nanobubbles. It appears for such hydrophobic surfaces, nanobubbles of different shape form and develop a contact angle which may differ from the traditional macrobubble contact angles.43 This research has a wide range of applications in various industries.44-46

Some of the notable research works on nanobubbles and nanodrops have been performed by a number of researchers.47, 48 In the present study nanobubbles are studied at the talc mineral surface using atomic force microscopy (AFM).

2.2 Materials and Methods

2.2.1 Minerals and Reagents

Fresh Talc 1 mineral samples were obtained from Argonaut Mine, Vermont, and a

Talc 4 (phlogopite) sample, was obtained from Department of Geology, College of Mines and Earth Sciences, University of Utah. Sample structures were confirmed by XRD analysis.

Sodium hydroxide (NaOH) and hydrochloric acid (HCl) obtained from Sigma-

Aldrich were used to adjust the pH. Potassium chloride (KCl), used in solution for ionic strength control, was obtained from Sigma-Aldrich. Deionized water was obtained from a

Millipore system in the laboratory with specific conductance of 18 MΩ.cm.

2.2.2 X-Ray Photoelectron Spectroscopy

The surface chemical compositions of the samples were obtained by XPS using the

Kratos Axis Ultra Model. The monochromatic aluminum K-alpha source (1486 keV) and detector was used at 15 kV and 10 mA (150 watts power). A spot size of 700 x 300 µm2

25 was used to collect the XPS spectra. The pressure during analysis was ~4x10-9 torr at room temperature. Sample heating during analysis is not expected to be significant. The binding energies were determined from the spectra of each sample to determine the elemental content of the surface. Under these conditions a penetration depth of 10 nm is expected and is indicative of the surface composition. Low-resolution survey scans were run from 1400 to 0 eV while high-resolution region scans covered the range of the peaks for specific elements.

2.2.3 Experimental and Simulation Sessile Drop Contact Angle

The sessile drop contact angles were measured using a RAMEHART goniometer.

These measurements provide initial wettability results and help to describe the hydrophobicity of the talc face surface. Fresh surfaces of Talc 1 and Talc 4 (phlogopite) were obtained by using scotch tape to remove top layers. All sessile drop contact angle measurements were obtained using 18 MΩ ultrapure Milli-Q water with a water drop size of 3 mm. At least 6 contact angle measurements were obtained, and they were averaged to obtain the reported contact angle value. Variation in contact angle measurements was ±2°.

Molecular dynamics simulation (MDS) consisting of the sample mineral (001) surface and a water box were combined in a periodic box to simulate advancing sessile drop contact angle. The talc crystal structure was obtained from the American

Mineralogical Crystal Structure Database. Crystal Maker software was used to prepare the talc crystal structure.

AMBER software was used for simulations involving the ideal talc surface.

This software was also used for the non-ideal talc surfaces containing aluminum impurities 26 which will be discussed in Chapter 3. With this information, sessile drop simulations were done using the SPC/E water model. The van der Waals energy term represented by

Lennard-Jones function (Eq. 1) and Columbic energy (Eq. 2) together with AMBER force field data from UFF and CLAYFF presented in Table 2.1 were used for atom-atom interaction.

12 6 푟푚,푖푗 푟푚,푖푗 퐸푉퐷푊 = ∑푖≠푗 휀푖푗 [( ) − 2 ( ) ] (1) 푟푖푗 푟푖푗

2 ⅇ 푞푖푞푗 퐸퐶표푙푢푚푏푖푐 = ∑ (2) 4훱휖 푟푖푗 푖≠푗

2.2.4 MDS Interfacial Water Analysis

Molecular dynamics simulation interfacial water analysis was done to describe the interaction between the water molecules with the surface atoms of the talc. The talc crystal structure consisted of atoms including Mg, Si, and O. The size of the talc surfaces in the periodic box was about 136Å x 150Å. Similar to the simulated sessile drop contact angle methods, water molecules along with the crystal structure were simulated in a periodic box.

This analysis of water structures close to the crystal surface helps to understand the polar or nonpolar nature of the talc face surface.

2.2.5 Experimental and Simulation Bubble Attachment Time

For the experimental bubble attachment time measurements, a high-speed camera,

KODAK EKTAPRO, which can record videos at 1000 frames/second, was used. Fresh surfaces of Talc 1 and Talc 4 (phlogopite) were obtained by using scotch tape to remove top layers. By checking the frames in the video, the time required for bubble attachment 27

Table 2.1. CLAYFF force field parameters.

Atom rm(A) Ɛ(kcal/mol) q(charge)

Si 3.704 1.84E-06 2.1

Mg 5.909 9.03E-07 1.36

Al 3.7064 1.84E-06 1.575

K 3.7423 1.00E-01 0.525

OT 3.5532 1.55E-01 -1.05

OC 3.5532 1.55E-01 -1.2825

OH 3.5532 1.55E-01 -0.95

H 0.425

28 was calculated, which includes film thinning time, and time required for expansion of the three-phase line of contact. At least 6 bubble attachment time measurements were obtained, and they were averaged to obtain the reported value. Variation in the bubble attachment time measurements was ±8 milliseconds.

Similar to the MDS sessile drop contact angle simulations, sample structures and surface data were obtained from the American Mineralogical Crystal Structure Database and created using Crystal Maker software. AMBER software was used, and bubble attachment simulation has been done using the methods mentioned in section 2.2.3.

Nitrogen gas was chosen for this simulation and a two-point model for nitrogen molecules was used.49, 50 The initial coordinates of a nitrogen bubble containing 906 nitrogen molecules in an aqueous phase containing about 100,000 water molecules was generated by the Xleap module of the Amber software packages. Then the isothermal- isobaric (NPT) ensemble was used to run the simulation for equilibration of the water and the nitrogen bubble with a simulation period of 500 ps. The amount (N), pressure (P), and temperature (T) were conserved. The simulation temperature was set as 298 K.

According to the Young-Laplace equation, pressure must increase as the bubble size decreases. For the nitrogen bubble with a diameter of 7 nm (R = 3.5 nm) at 298 K, the pressure inside of the bubble is 41.6 × 107 dyne/cm2. A previous study by Takahashi et al. has confirmed this relationship.51

At high pressure, the ideal gas law is not valid, so the corrected real gas law at high pressure is applied for the nitrogen bubble, as shown in Equation 3.

(푉−푛푏) = 푛푅푇 (3) where V, p, and n are the volume, pressure, and moles of nitrogen. For nitrogen gas the

29 constant b is 0.04 dm3/mol. T is 298 K, and R is the molar gas constant (8.314 J·K-1·mol-

1). From this equation, the number of molecules in the nitrogen bubble was found to be

1086 which is close to MD simulations.52

2.2.6 Synchronized Triwavelength Reflection Interferometry

Microscopy (STRIM)

The instrumentation and technique developed by Dr. Lei Pan at the Michigan

Technological University, called Synchronized Triwavelength Interferometry Microscopy

(STRIM), uses Red, Green, and Blue light wavelengths (λ = 460, 527, and 620 nm) simultaneously, to accurately measure fringes and calculate water film thickness. The film thickness at the hydrophobic mineral surface at the point of rupture is called the critical rupture thickness (hc) and at the hydrophilic surface where no water film rupture occurs, the thickness is called the equilibrium film thickness (he). The setup is shown in Figure 2.2.

This procedure uses RGB light wavelengths (λ = 460, 527, and 620 nm) to calculate the solution film thickness separating a bubble from a mineral surface. It consists of an inverted microscope, RGB light sources, and a glass cell with a hydrophobic base to hold the air bubble of 0.4 mm diameter. The glass cell rests on a piezo stage that moves up with a speed of approximately 1 µm/s. Film thickness is determined from interference fringes which form when the bubble moves close to the mineral surface. All solutions were prepared in

1 mM NaCl background electrolyte. The solution film thickness values were then calculated. The detailed experimental method can be found elsewhere.53 The obtained interference fringes were analyzed to determine the spatiotemporal thickness profile of the wetting film.

Figure 2.2. Schematic of the Synchronized Triwavelength Reflection Interferometry Microscopy (STRIM) setup.53 (Reprinted with permission)

2.2.7 Atomic Force Microscopy

AFM imaging of the nanobubbles was done using a Nanoscope III A (Digital

Instruments), and SNL-10 silicon probes (Bruker), which are specifically designed for tapping mode images in liquids. The tip had a resonant frequency varying between 10 and

100 Hz and had a spring constant varying from 20 N/m to 35 N/m. The cleaning procedure for the tip, as well as the cantilever, involved cleaning them with acetone and ethanol, rinsing with milli-Q water, followed by drying with ultrapure nitrogen. A solvent exchange procedure using ethanol and DI water was used to produce nanobubbles at the talc surface.

Initially, ethanol was pumped into the liquid cell using a freshly opened syringe. After 1 hour, the ethanol in the liquid cell was replaced by pumping in DI water using another freshly opened syringe. Further details on the solvent exchange procedure can be found in the literature.47, 48 The tapping mode images were taken in a liquid cell and subjected to

2nd-order flattening and 2nd-order low pass filtering. In the case of imaging nanobubbles, both height and phase images were taken at the talc face surface.

2.3 Sessile Drop Contact Angle Results

Sessile drop contact angle analysis was accomplished by traditional experimental contact angle measurements which are used to determine the behavior of the liquid on a solid surface. These measurements provide an initial idea on the wetting characteristics of the mineral surface. Further, molecular dynamics simulations (MDS) of sessile drop contact angles were done to provide further characterization, and complement the experimental analysis. 32

2.3.1 Experimental

Sessile drop contact angles were measured at various locations on the face surface of Talc 1 and Talc 4 (phlogopite) and show in Table 2.2. The results presented here are the advancing sessile drop contact angles. Also, the MD simulated angles are expected to be closely related to the advancing contact angles. As seen from XPS results the Talc 1 surface contains <0.1 % aluminum impurity and the surface is nonpolar. The Talc 1 has a sessile drop contact angle of 79° and Talc 4 (phlogopite), which represents an aluminum substitution ratio of 0.25, has a sessile drop contact angle of 0º. The aluminum substituted talc structures with aluminum substitution ratios 0.01 and 0.027 are later explained in

Chapter 3. Therefore, from the experimental sessile drop contact angle on Talc 1, it can be deduced that the surface is distinctly hydrophobic. Further analysis of the surface state of talc from molecular dynamics simulations (MDS) is provided in the following section.

2.3.2 MD Simulation

AMBER software was used for simulations involving the high-quality talc surface.

With this information, sessile drop simulations were done. SPC/E water model has been used for water molecule simulations. Lennard-Jones and Columbic potentials together with

AMBER force field data from UFF and CLAYFF were used for atom-atom interaction.

Molecular dynamics simulation (MDS) consisted of the talc mineral (001) surface and a water box which were combined together in a periodic box to simulate advancing sessile drop contact angles. The talc crystal structure was obtained from the American

Mineralogical Crystal Structure Database. Crystal Maker software was used to prepare the talc mineral surface. 33

Table 2.2. Contact angle and bubble attachment time for talc and phlogopite surfaces.

Mineral Sample XPS Aluminum, Contact Angle, Bubble Attachment

Atomic % Degrees Time, (seconds)

Talc <0.1 79 0.025

Phlogopite 7.0 0 NA

To complement the traditional experimental advancing sessile drop contact angle for Talc 1, a talc mineral surface, namely MDS Talc 1, was created with using Crystal

Maker software. The mineral crystal structure consists of atoms including Mg, Si, O, and is free of any impurities representing an ideal condition. The size of the mineral surfaces in the periodic box was about 136Å x 150Å. A simulation time of 1 nanosecond was considered, where 500 picoseconds was the equilibration time for the water drop to spread, and another 500 picoseconds was the analysis time to calculate the contact angle. To determine the MDS sessile drop contact angle, postprocessed densities of water molecules were plotted in two center planes: x-z plane and y-z plane. An analysis program was coded with Fortran 90 to explore the 2-dimensional number density distribution of water molecules for the water drop in the projection plane. The contact angles were measured in the x-z and y-z planes and then averaged.54, 55

Simulated sessile drop contact angles were determined for the MDS Talc 1 representing an ideal crystal. MDS snapshots of the simulated sessile drop contact angles are shown in Figure 2.3. The results again show that the talc face surface is hydrophobic with a with the contact angle of 70°, similar to the experimental results (Table 2.1). The surface state of the ideal talc face surface clearly exhibits a hydrophobic character.

2.4 MDS Interfacial Water Analysis

Snapshots from MDS interfacial analysis of MDS Talc 1 surface are shown in

Figure 2.4. There is an insignificant interaction between water molecules and the surface of talc structure. Later in Chapter 3, the significance of the interfacial water structure after considering the aluminum impurity the crystal structure of talc will be explained. In the

Figure 2.3. Snapshots of the advancing sessile drop contact angle at the surface with no aluminum substitution for silicon in the tetrahedral face surface for MDS Talc 1 at 0, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H.

Figure 2.4. MDS snapshots (two different directions) for interfacial water analysis at the unsubstituted surface (MDS Talc 1), aluminum ratio = 0. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H. 37 case of the MDS Talc 1, there is a gap between the water phase and the surface, and this gap is known as the “water-excluded volume.”54, 56, 57 In the case of MDS Talc 1, the water exclusion zone is present due to the insignificant interaction of water molecules with face surface atoms. The interfacial water analysis is of importance in understanding film thinning, rupture, displacement, and bubble attachment at talc surfaces.

2.4.1 Relative Number Density

The relative number density can be defined as the fraction of absolute density over an average number density for the bulk water. For oxygen atoms in the water phase, the relative number densities tend to remain at ‘1’ for distances greater than about 8 Å away from the surface, which suggests little influence of the surface atoms on the water density beyond this point. The relative number density distribution function for the oxygen atoms in water molecules along the surface normal for the talc crystal is shown in Figure 2.5. The water density peak higher than ‘1’ can be interpreted as the “primary water density peak.”54

From the analysis of these peaks, a gap of approximately 3.2 Å is observed for MDS Talc

1 with no surface impurities. This observation provides an insight into the interfacial water characteristics at a hydrophobic surface where there are insignificant interactions between the water molecules and the surface.

2.4.2 Water Dipole Moment and Hydrogen Position

The water dipole moment and the hydrogen position relative to the surface normal are illustrated in Figure 2.6. From this figure, ‘α’ can be described as the angle between water dipole moment and surface normal, and ‘β’ can be described as the acute angle between surface normal and the line connecting two hydrogen atoms. 38

Figure 2.5. Relative number density profiles at the unsubstituted surface of MDS Talc 1 (talc 001 surface), aluminum ratio = 0. 39

Figure 2.6. Schematic illustration describing the orientation of a water dipole moment by the angle (α) and hydrogen position relative to the surface normal by the angle (β).30 (Reprinted with permission) 40

From the relative number density (Figure 2.5), the observed gap to the first oxygen is nearly 3.2 Å. When we consider the water dipole orientations, the angle ‘α’ is around

90° to the surface normal line and is parallel to the face surface of the crystal (001 plane).

Similarly, for the hydrogen positions, the angle ‘β’ is also around 90° to the surface normal and is parallel to the surface as shown in Figure 2.7. Therefore, the water molecules are oriented parallel to the surface and have insignificant interactions with the surface. These results are in good correlation and consistent with the interfacial water analysis.

2.5 Bubble Attachment Results

Film thinning, rupture, displacement, and subsequent bubble attachment time have been studied using experimental and simulation methods at the talc surface.

2.5.1 Experimental

Experimental bubble attachment time measurements were performed at the talc surface and the results complemented the contact angle results. With <0.1% aluminum substitution ratio, the bubble readily attaches to the crystal surface as shown in Table 2.1.

It is interesting to note that the bubble attachment is nearly 28 milliseconds in

<0.1% aluminum substituted crystal sample. From the section on interfacial water analysis, it can be seen that there is an insignificant interaction between the water molecules and the surface atoms of the crystals. Therefore the thin liquid film offers no resistance to the air bubble which readily attaches to the surface. To explain this phenomenon at the molecular level, MD bubble attachment simulation has been performed using nitrogen gas to qualitatively explain the experimental result, and is presented in the next section. 41

Figure 2.7. Water dipole moment and hydrogen position at the talc face surface. 42

2.5.2 MD Simulation

For measuring MDS bubble attachment time, MDS Talc 1 was prepared with zero aluminum to silicon substitution ratio in the tetrahedral layer. The crystal structure consists of atoms including Mg, Si, and O. The nitrogen bubble along with the water molecules was put adjacent to the mineral (001) surface at a distance of 1 nanometer. The size of the mineral surfaces in the periodic box was about 136Å x 150Å. Simulation time of 1 nanosecond was considered, where 500 picoseconds was the equilibration time for the nitrogen bubble to move closer to the surface, and another 500 picoseconds analysis time to check the interaction.

Bubble attachment MD simulation snapshots for MDS Talc 1 are presented in

Figure 2.8. To better understand the interaction between the nitrogen bubble and the surface in the absence of aluminum substitution, the atoms in the talc crystal are represented by spheres with a diameter of 1.8 Å, nitrogen molecules with a diameter of 3.8 Å, and the water molecules with a diameter 0.5 Å. In the simulation, as time progresses, the nitrogen bubble moves closer to the surface of the zero aluminum substituted talc (MDS Talc 1), resulting in water film rupture and subsequent attachment. From the analysis of the nitrogen bubble as it moves closer to the talc surface, the water film starts to rupture and displacement occurs at 0.1 nanosecond and the diameter of the nitrogen bubble expands on the talc surface until an elapsed time of 1 nanosecond. The diameter of the three-phase line of contact gradually increases during this time. Also, as seen from the previous section on interfacial water analysis, there is an insignificant interaction between the water molecules and surface atoms (see Figure 2.5). Consequently, water film rupture and film displacement occur causing the nitrogen bubble to attach and expand on the face surface of talc.

Figure 2.8. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of an unsubstituted ideal talc crystal (MDS Talc 1) for interaction times of 0, 0.1, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H; Blue, N.

2.6 Film Thickness Results

The thin film phenomena, particularly the theory of film rupture and subsequent bubble attachment, have been discussed in the flotation literature.58 Bubble attachment is described by the well-known classical condition prescribed in equation 3.

훥퐺 = 훾푆퐺 − 훾퐿퐺 − 훾푆퐿 < 0 (3) where 훾푆퐺 is the surface free energy of the solid-gas, 훾퐿퐺 is the surface free energy of the

1 liquid-gas and 훾푆퐿 is the surface free energy of the solid-liquid. The introduction of the term “disjoining pressure”, Π, by Derjaguin, and is defined as a measure of change in the corresponding thermodynamic properties.59 This disjoining pressure has to be negative for the bubble attachment and is described in terms of surface energy in equation 4.

∞ 훾 = 훾 + 휋 ⅆℎ (4) 0 ∫ℎ where, γ is the specific-surface free energy of the thin liquid film, 훾0 is the specific-surface free energy of infinite thin liquid film, h is the thickness of the film, and Π is the disjoining pressure.

2.6.1 Synchronized Triwavelength Reflection Interferometry

Microscopy (STRIM)

STRIM experiments were conducted by bringing an air bubble towards a flat mineral surface at a constant nominal velocity of 1.1 μm/s, while monitoring the time evolution of separation distance for hydrophobic Talc 1 surface and hydrophilic Talc 4

(phlogopite) surface. See Table 2.1. Figure 2.9 shows spatiotemporal thickness profiles of the wetting films formed on a talc and a phlogopite surface in 1 mM NaCl solutions. The radii of the air bubbles are in the range of 0.35-0.45 mm.

Figure 2.9. A comparison of spatiotemporal thickness profile of wetting films formed on a) fresh talc surface and b) fresh phlogopite surface in 1 mM NaCl solution. 46

The results show that on both talc and phlogopite surfaces, the air bubble remained spherical at a separation distance of greater than 300 nm, where both hydrodynamic and surface forces were negligible. As the film continued to thin to a thickness of less than 300 nm, the bubble started to flatten due to an arising repulsive hydrodynamic force. The wetting film ruptured when the closest separation distance reached a critical rupture thickness (hc).The results show in the present work that the critical rupture thickness (hc) of 56-60 nm for fresh talc surfaces. On the contrary, the wetting film formed on the phlogopite surface is stable and the equilibrium film thickness (he) was determined to be

25 nm; bubble-mineral attachment did not occur. The equilibrium was reached when the arising disjoining pressure became equal to the capillary pressure (pc). Table 2.3 shows a summary of the film thickness measurements on talc and phlogopite surfaces. The present results confirm that wetting films are stable at hydrophilic surfaces, while being unstable at hydrophobic surfaces.

2.6.2 Discussion

Water film thickness values at the hydrophobic talc surface and the hydrophilic phlogopite surface provide a contrast in the critical rupture thickness in the former and equilibrium film thickness in the latter. The film thickness values from the STRIM measurements on talc and phlogopite have not been previously reported, and these results are significant to further research on the behavior of the thin liquid films at the surface of layered silicate minerals. The formation of the unstable film in the case of talc is related to the interfacial water analysis (see Figure 2.4), where there are insignificant interactions between the water molecules and the surface atoms. See Figure 2.7. This causes the thin 47

Table 2.3. Contrast between water film thickness values at talc and phlogopite surfaces.

Sample Talc Phlogopite Critical rupture thickness 56 nm Equilibrium film thickness 25 nm Air bubble Attachment No Attachment Surface state Hydrophobic Hydrophilic

48 film to rupture and be displaced during air bubble attachment. But in the case of phlogopite, the thin film is stable as the electrostatic interactions between the water molecules and the surface counter ions (K+) are significant. More discussion on the interfacial water analysis on phlogopite is presented in Chapter 3.

To put things in perspective, it is useful to look at the surface charge of various layered silicate minerals and how they affect film thickness and film stability. For the crystal structures of these layered silicates, refer to Figure 1.1 in Chapter 1. The importance of the influence of hydrophobic force, electrostatic force, and van der Waals force or a combination of these forces on the disjoining pressure and subsequent film rupture is discussed in the literature.60-67 The surface charge on the face surface of various layered silicates such as kaolinite, talc, muscovite, and chlorite has been studied by many researchers using atomic force microscopy force measurements.18, 68-70 Since the layered silicates exhibit isomorphic substitution, the face surface or the basal plane usually has a negative charge at neutral pH.18 The magnitude of this negative charge increases as the isomorphic substitution increases which determines the surface state of the layered silicate minerals. For example, from the literature it is reported that the hydrophobic talc face surface has a surface potential of -22 mV and the hydrophilic muscovite face surface has a surface potential of -88 mV at neutral pH.18 It should also be noted that an air bubble carries a negative charge which can be in the range of -20 mV to -74 mV.71 Therefore, film rupture at the hydrophobic talc face surface is facilitated by the low surface potential, insignificant interactions between the water molecules and face surface atoms (water exclusion zone), together with nanobubble formation, the conditions that account for the attractive hydrophobic force, film rupture and bubble attachment.72 But in the case of the hydrophilic

49 phlogopite surface, which has a similar structure to muscovite, higher surface potential increases the coulumbic repulsion between the face surface and the air bubble, and the interactions between interfacial water molecules and potassium ions at the faces surface cause the water film to be stable. Further research on the film thickness of layered silicates with respect to the surface state, pH, and salt concentration is warranted.

Some of the critical rupture thickness (hc) measurements in the literature have a large range of film thickness values ranging from 11 nm at a hydrophobic surface (90º) with 100 µm air bubbles, 60 nm to 140 nm at hydrophobic silica surfaces (75º to 103º) with

4 mm air bubbles, and 205 nm at a hydrophobic surface (62.5º) with 1 mm air bubbles.34,

73, 74 Therefore, the STRIM measurements aim to provide accurate film thickness values using three laser light sources (RGB).

2.7 Atomic Force Microscopy Results

The presence of micropancake-like nanobubbles have been seen in some other research papers. Contact angle measurements determine the surface state and wetting characteristics of a mineral sample. Usually sessile drop or captive bubble contact angles are measured in macroscale with drop/bubble sizes in the range of 2-4 mm. In the contact angle experiments using MD simulations, the drop/bubble sizes are in the range of 5-10 nm. These contact angles measured in the MD nanometer scale and the experimental millimeter scale correlate well with each other. AFM imaging is used to fill the gap between these drop/bubble sizes. Using solvent exchange method of ethanol and DI water nanobubbles are observed at the surface of talc as shown in Figure 2.10. The base dimensions for the nanobubbles are in the range of 100 to 500 nm, height dimension in the range of 7

Figure 2.10. AFM nanobubbles at a fresh talc surface. 51 to 10 nm, and the contact angle is estimated to be greater than 90°. These surface nanobubbles form pancake-like structures which have been observed in the literature.48

Also, as suggested by some researchers, the contact angle of the nanobubbles is higher than the macrocontact angles measured, and there is a wide distribution of contact angle data for the same surface.43 Literature is also available on the contact angle of nanodrops, particularly oils, whose contact angles are similar to those observed in nanobubbles.47 In addition, the stability of the nanobubbles on the hydrophobic surfaces is observed to be ranging from minutes, hours, and even days.75 Further experiments are necessary to determine the size and shape of these nanobubbles and nanodrops on hydrophobic mineral surfaces.

2.8 Summary

The importance of the hydrophobic surface state and the wetting characteristics of talc in understanding flotation response are presented with results from contact angle, bubble attachment, film thickness, and atomic force microscopy. The extent of aluminum substitution in the silica tetrahedral layer for Talc 1 and Talc 4 samples was examined by

XPS, and the corresponding experimental sessile drop contact angles were found to decrease with an increase in the aluminum substitution, decreasing from about 79° degrees for zero aluminum substituted talc to 0º for more extensive substitution (phlogopite). MD simulated sessile drop contact angles also showed similar results.

In addition to the contact angle measurements, the wettability characteristics of the talc face surface was studied by making bubble attachment time measurements. As anticipated the air bubble attached to the face surface of talc instantly with a time of 25 ms.

In contrast, no bubble attachment occurred when 0.30 sites are occupied by aluminum in

52 the case of phlogopite crystal.

These results indicate that interfacial water structure is an important factor which determines the film stability and bubble attachment. From interfacial water analysis by

MDS, there was an insignificant interaction between the water molecules and the surface atoms of talc. The water dipole and hydrogen bonding results also showed that there is no hydrogen bonding between the water molecules and the talc surface. These results are important to explain the instability of the water film at a hydrophobic surface. From the film thickness measurements by STRIM, a critical rupture thickness of 56 nm was observed at the talc surface and an equilibrium film thickness of 25 nm was observed at the phlogopite surface.

In addition, the presence of nanobubbles at the face surface of talc were shown by solvent exchange method. The presence of these nanobubbles provides additional proof to the hydrophobic surface of talc using atomic force microscopy and the interesting behavior with respect to shape and size which results in high contact angles.

CHAPTER 3

SIGNIFICANCE OF ALUMINUM SUBSTITUTION

3.1 Introduction

Most research studies on the surface chemistry of talc have been done on high- quality mineral samples which have little or no impurity.76-80 In the literature, talc samples with varying aluminum impurity level have not been extensively studied to understand the effect of lattice impurities on the surface state of talc. The research work in this chapter aims to improve the understanding of how the effect of aluminum impurities substituted in the silica tetrahedral layer influences the state of talc face surface. A few research articles were published and reported on the change in surface polarity/hydrophobicity of talc with aluminum substitution in the silica tetrahedral layer. Some of the research work in the case of lattice substitution has been done by Burdukova et al., Yin et al., and Mierczynska-

Vasilev and Beattie.18, 81, 82 It is reported that lattice impurities affect polysaccharide adsorption and the wetting characteristics of the talc surface. Thus, it is expected that aluminum atoms, which substitute for silicon atoms in the silica tetrahedral layer, will alter the surface state of talc.

Burdukova et al. reported that the basal plane of talc is negatively charged based on electrophoretic measurements and explained that this charge was due to the aluminum

(Al3+) substitution for silicon (Si4+) in the silica tetrahedral layer.81 This charge imbalance

54 in the tetrahedral layer is expected to be satisfied by exchangeable cations like K+. The results reported are based on a talc sample from New York with 0.4 wt % aluminum substitution for silicon in the silica tetrahedral layer.

Yin et al. studied the surface charge and wetting characteristics of various bilayer and trilayer phyllosilicate minerals and found that the wetting characteristics of the trilayer silicate sequence pyrophyllite, illite, and muscovite change with the aluminum substitution in the silica tetrahedral layer.18 Results from molecular dynamics simulations were done for crystals with varying degrees of aluminum substitution to determine the change in wetting characteristics and their surface polarity. From the analysis, it was found that the surface properties of the crystals changed from a hydrophobic surface state to a hydrophilic surface state with increasing substitution in the silica tetrahedral layer. Similar MD simulations have been performed by other researchers to study the wetting characteristics and surface state of various minerals.29, 54

Mierczynska-Vasilev and Beattie discussed the importance of aluminum substitution on the hydrophobicity of talc and the substitution effect on polysaccharide adsorption.82 From their study, the aluminum substitution alters the mineral basal plane surface and changes the surface resulting in different cleavage characteristics. Surface hydrophobicity also varies with the aluminum substitution. Further, polymer adsorption is influenced by substitution which results in a variation of the talc flotation response. Finally,

Shrimali et al. investigated the adsorption of starch at hydrophobic surfaces, including the talc face surface.83

The present chapter deals with the change in the surface properties of talc due to the presence of aluminum impurity in the silica tetrahedral layer. The primary focus is to 55 assess the lattice substitution influence on surface polarity and the corresponding wetting features, as well as film stability in the case of bubble attachment. The wettability characteristics and surface state of these substituted talc minerals were examined by sessile drop contact angle and captive bubble attachment time including experimental methods and molecular dynamics simulations. In the simulation studies, four talc crystal structures were prepared with varying level of aluminum substitution with the aluminum/total tetrahedral site ratios ranging from zero for ideal talc structure to 0.25 for the phlogopite structure. Examples of crystal structures are shown in Figure 3.1.

3.2 Materials and Methods

3.2.1 Minerals and Reagents

In addition to the high-quality Talc 1 sample, as discussed in Chapter 2, three additional mineral samples, including two talc samples and one phlogopite sample, with different elemental composition, were obtained from Department of Geology, College of

Mines and Earth Sciences, University of Utah (samples identified as Talc 2, 3 and 4).

Sample structures were confirmed by XRD analysis.

Sodium hydroxide (NaOH) and hydrochloric acid (HCl) obtained from Sigma-

Aldrich were used to adjust the pH. Potassium chloride (KCl), used in solution for ionic strength control, was obtained from Sigma-Aldrich. Deionized water was obtained from a

Millipore system in the laboratory with specific conductance of 18 MΩ.cm. Care was taken to make sure the experimental apparatus was free from impurities. 56

Figure 3.1 Crystal structures of talc, impure talc, and phlogopite with aluminum/total tetrahedral site ratios ranging from 0 to 0.25. The atom codes are: empty circles, O; black circles, Si; gray circles, OH; black triangles, Mg; black squares, Al. 57

3.2.2 X-Ray Photoelectron Spectroscopy

The surface chemical compositions of the samples were obtained by XPS using the

Kratos Axis Ultra Model. The monochromatic aluminum K-alpha source (1486 keV) and detector was used at 15 kV and 10 mA (150 watts power). A spot size of 700 x 300 mm was used to collect the XPS spectra. The pressure was ~4x10-9 torr at room temperature.

Sample heating during analysis is not expected to be significant. The binding energies were determined from the spectra of each sample to determine the elemental content of the surface. Under these conditions a penetration depth of 10 nanometers is expected and is indicative of the surface composition. Low-resolution survey scans were run from 1400 to

0 eV while high-resolution region scans covered the range of the peaks for specific elements.

3.2.3 Experimental and Simulation Sessile Drop Contact Angle

The sessile drop contact angles were measured using a RAMEHART goniometer.

These measurements provide initial wettability results and help to describe the hydrophobicity of the talc face surface. A fresh surface of Talc 1 and Talc 4 (phlogopite), was obtained by using scotch tape to remove top layers. Fresh surfaces for Talc 2 and Talc

3 were obtained by cleaning them with acetone, ethanol, and water. All sessile drop contact angle measurements including the results for Talc 1 reported in Chapter 2 were obtained using 18 MΩ ultrapure Milli-Q water with a water drop size of approximately 3 mm. At least 6 contact angle measurements were obtained, and they were averaged to obtain the reported contact angle value. Variation in contact angle measurements was ±2°.

Molecular dynamics simulation (MDS) consisting of the sample mineral (001) 58 surfaces and a water box were combined in a periodic box to simulate advancing sessile drop contact angles. The talc crystal structures were obtained from the American

Mineralogical Crystal Structure Database. Crystal Maker software was used to prepare the various face surfaces of the talc structures. AMBER software was used for simulations involving the ideal talc surface and the non-ideal talc surfaces containing aluminum impurities. With this information, sessile drop simulations were done using the SPC/E water model. The van der Waals energy term represented by

Lennard-Jones function (Eq. 1) and Columbic energy (Eq. 2) together with AMBER force field data from UFF and CLAYFF presented in Table 3.1 were used for atom- atom interaction.54, 84

3.2.4 MDS Interfacial Water Analysis

Molecular dynamics simulation interfacial water analysis was done to describe the interaction between the water molecules with the surface atoms of the talc structures. The talc crystal structure consists of atoms including Mg, Si, O, Al, and K. The size of the talc surfaces in the periodic box was about 136Å x 150Å. A simulation time of 1 nanosecond was considered, where 500 picoseconds was the equilibration time for the water to spread and 500 picoseconds for equilibrium. Similar to the simulated sessile drop contact angle methods, water molecules along with the crystal structure were simulated in a periodic box. This analysis of water structures close to the crystal surface helps to understand the change in surface polarity of various talc structures based on aluminum substitution for silicon in the silica tetrahedral layer. 59

Table 3.1. CLAYFF force field parameters.

Atom rm(A) Ɛ(kcal/mol) q(charge)

Si 3.704 1.84E-06 2.1

Mg 5.909 9.03E-07 1.36

Al 3.7064 1.84E-06 1.575

K 3.7423 1.00E-01 0.525

OT 3.5532 1.55E-01 -1.05

OC 3.5532 1.55E-01 -1.2825

OH 3.5532 1.55E-01 -0.95

H 0.425

3.2.5 Experimental and Simulation Bubble Attachment Time

For bubble attachment time measurements, a high-speed camera, KODAK

EKTAPRO, which can record videos at 1000 frames/second, was used. Fresh surfaces of

Talc 1 and Talc 4 were obtained by using scotch tape to remove top layers. Talc 2 and Talc

3 were cleaned using acetone, ethanol, and water. By checking the frames in the video, the time required for bubble attachment was calculated, which includes film thinning time, and time required for expansion of the three-phase line of contact. At least 6 bubble attachment time measurements were obtained, and they were averaged to obtain the reported value.

Variation in bubble attachment time measurements was ± 8 milliseconds.

3.3 X-Ray Photoelectron Spectroscopy Results

XPS analysis has been done on these samples to quantify the elemental composition of the surface. Sample identification from XPS analysis is presented in Table 3.2.

Theoretically, in the crystal structure of phlogopite, every fourth silicon atom is replaced by aluminum atom giving rise to a ratio of 0.25 aluminum substitution. Therefore, in the experimental studies, talc samples with varying aluminum impurity and one phlogopite sample were obtained to provide a comprehensive surface state analysis. The mineral samples are listed with increasing aluminum substitution for silicon in the silica tetrahedral layer. For the low aluminum substituted samples, the percentage of substitution ranged

Table 3.2. XPS surface analysis of talc samples in atomic percent.

Mineral Sample O K C Si Mg Al

Talc 1 (talc, light green) 56.5 2.4 24.2 13.4 <0.1

Talc 2 (talc, dark green) 46.4 19.5 20.2 13.6 0.3

Talc 3 (talc, white) 51.7 14.1 21.4 12.2 0.6

Talc 4 (phlogopite) 52.8 3.9 6.6 16.3 13.4 7.0

62 from <0.1 to 0.6 atomic percent. However, in the case of Talc 4 (phlogopite), as seen from the XPS analysis in Table 3.2, the face surface contains 7 atomic percent aluminum and 16 atomic percent silicon which composition is attributed to the tetrahedral layer. The aluminum atomic percent is divided by the total tetrahedral site atomic percent (7 divided by 23) for an aluminum substitution ratio of 0.30, and this ratio is close to the expected ratio of 0.25.

3.4 Sessile Drop Contact Angle Results

Trilayer silicate minerals consist of an octahedral sheet sandwiched between two tetrahedral sheets. In the case of talc, the magnesia octahedral layer is sandwiched between two silica tetrahedral layers. In studies by Yin et al., pyrophyllite (0% substitution), illite

(5%, 15% substitution), and muscovite (25% substitution) structures with aluminum substitution for silicon in the tetrahedral layer were simulated to obtain the change in wetting characteristics and surface polarity. The surface state of these crystals changed from a hydrophobic state in the case of pyrophyllite to a hydrophilic state in the case of muscovite, with increasing aluminum substitution.18 Current studies have been done using experimental and simulation methods to understand the change in surface state.

3.4.1 Experimental

Sessile drop contact angles were performed for the various aluminum substituted talc samples. The results presented here are the advancing sessile drop contact angles. Also, the MD simulated angles are expected to be closely related to the advancing contact angles.

The change in the contact angle of the samples from a strongly hydrophobic state (Talc 1)

63 to a lower hydrophobic state (Talc 2 and Talc 3), and to a hydrophilic state (Talc 4, phlogopite) is expected because of the substitution of aluminum in the silica tetrahedral layer, which can be seen in Table 3.3.

To understand the significance of aluminum substitution, a column for aluminum atoms/total tetrahedral site ratio has been added in Table 3.3. This ratio represents the number of aluminum sites to total sites in the tetrahedral layer. Similarly, as mentioned in

Section 3.3, in the case of Talc 2 and Talc 3, with 0.3 and 0.6 atomic percent aluminum, very few silicon atoms are substituted by aluminum atoms and the aluminum substitution ratios are 0.01 and 0.027, respectively. In the case of Talc 4 (phlogopite), the aluminum atomic percentage was divided by the total tetrahedral site atomic percent (7 divided by

23) for an aluminum ratio of 0.30, and this ratio is close to the expected ratio of 0.25.

As seen in Chapter 2, for the Talc 1 sample with <0.1% aluminum substitution, the contact angle was 79°. In the case of the Talc 2 and Talc 3 samples, the contact angles are

66° and 60° when the aluminum substitution ratio is 0.01 and 0.027, respectively. It is important to note that for Talc 4 sample (phlogopite) when the aluminum substitution ratio is 0.30, water wets and spreads on the surface resulting in a contact angle of 0°. It should also be noted that the surface of phlogopite contains counter ions like K+ to balance the charge instability arising from the aluminum substitution of silicon. Therefore, from the experimental sessile drop contact angles, it can be deduced that the surface state changes from a slightly hydrophobic state to a hydrophilic state, with increasing aluminum substitution. Further analysis of the change in surface state of talc from molecular dynamics simulations (MDS) is provided in the following section.

Table 3.3. Experimental advancing sessile drop contact angles for talc samples and phlogopite.

Mineral Sample XPS Aluminum, Aluminum Contact Angle,

Atomic % Sites/Total Sites Degrees

Talc 1(talc, light green) <0.1 0 79

Talc 2 (talc, dark green) 0.3 0.01 66

Talc 3 (talc, white) 0.6 0.027 60

Talc 4 (phlogopite) 7.0 0.30 0

3.4.2 MD Simulation

To study the effect of aluminum substitution in the silica tetrahedral layer, different talc mineral surfaces, MDS Talc 1, MDS Talc 2, MDS Talc 3, and MDS Talc 4 (phlogopite) were created with 0, 0.01, 0.08, and 0.25 aluminum substitution ratios, respectively. The mineral crystal structure consists of atoms including Mg, Si, O, Al, and K. The size of the mineral surfaces in the periodic box was about 136Å x 150Å. A simulation time of 1 nanosecond was considered, where 500 picoseconds was the equilibration time for the water drop to spread, and another 500 picoseconds was the analysis time to calculate the contact angle. To determine the MDS sessile drop contact angle, postprocessed densities of water molecules were plotted in two center planes: x-z plane and y-z plane. An analysis program was coded with Fortran 90 to explore the 2-dimensional number density distribution of water molecules for the water drop in the projection plane. The contact angles were measured in the x-z and y-z planes and then averaged.54, 55

Simulated sessile drop contact angles were determined for these four MDS Talc samples representing different levels of aluminum substitution. The results again show that aluminum substitution in the talc tetrahedral lattice sites leads to reduced hydrophobicity at the talc surface. Similar to the experimental results, the MDS results reported in Table

3.4, show the expected change in polarity and wetting characteristics for the surfaces with the contact angle changing from 70° for zero aluminum substituted talc surface to the polar phlogopite surface with a contact angle of 3°. The surface state of talc structures changes from a hydrophobic state in MDS Talc 1 to a slightly hydrophobic state in the case of MDS

Talc 2 and MDS Talc 3, and to a completely hydrophilic state in the case of phlogopite, which is MDS Talc 4. MDS snapshots of the simulated sessile drop contact angles are

Table 3.4. MD simulated advancing sessile drop contact angles for crystals with different levels of aluminum substituted in the tetrahedral layer of the face surface.

Crystal Aluminum Contact Angle,

Sites/Total Sites Degrees

MDS Talc 1 0 70

MDS Talc 2 0.013 45

MDS Talc 3 0.08 24

MDS Talc 4 (phlogopite) 0.25 3

67 shown in Figure 3.2, Figure 3.3, Figure 3.4, and Figure 3.5.

Further, snapshots of all the crystals at a simulation time of 1 nanosecond are shown in Figure 3.6. As the aluminum ratio in the crystal increases, the contact angle decreases.

Also, a comparison of the experimental and simulated sessile drop contact angles, based on aluminum substitution in the tetrahedral layer, is plotted in Figure 3.7, also considering the pure talc crystal results from Chapter 2. The MDS contact angle decreases from nearly

70° for the ideal unsubstituted talc crystal to 3° for the 0.25 aluminum substituted phlogopite crystal. Similarly, the experimental contact angle decreases from 79° to 0° with an increase in the aluminum substitution ratio. This significant change in contact angle can be attributed to the change in polarity of the talc face surface.

3.5 Interfacial Water Analysis

Snapshots from MDS interfacial analysis showing the interaction between water molecules and the surface of talc structures, MDS Talc 1, MDS Talc 2 and MDS Talc 3, as well as MDS Talc 4 surface (phlogopite) are shown in Figure 3.8. The crystal structure of the surfaces significantly influences the interfacial water structure. In the case of the

MDS Talc 1 (Chapter 2), MDS Talc 2, and MDS Talc 3 surfaces with zero and low aluminum substitution, the gap between the water phase and the surface is greater than for the MDS Talc 4 surface (0.25 aluminum substitution). The gap between the water phase and the crystal surface is known as the “water-excluded volume.”54, 56, 57

In the case of MDS Talc 4 (phlogopite), the water exclusion zone is reduced due to the interaction of water molecules with K+ ions present on the face surface created by the aluminum substitution. This change in surface state from a hydrophobic state, in the case

Figure 3.2. Snapshots of the advancing sessile drop contact angle at the surface with no aluminum substitution for silicon in the tetrahedral face surface for MDS Talc 1 at 0, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H.

Figure 3.3. Snapshots of the advancing sessile drop contact angle at the surface with aluminum substitution (0.013) for silicon in the tetrahedral face surface for MDS Talc 2 at 0, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H.

Figure 3.4. Snapshots of the advancing sessile drop contact angle at the surface with aluminum substitution (0.08) for silicon in the tetrahedral face surface for MDS Talc 3 at 0, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H.

Figure 3.5 Snapshots of advancing sessile drop contact angle at the face surface of the crystal which has aluminum substituted for every fourth silicon (0.25) in the tetrahedral face surface for MDS Talc 4 (phlogopite) at 0, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H.

Figure 3.6. Snapshots of advancing sessile drop contact angle at the face surface of the crystals, MDS Talc 1, MDS Talc 2, MDS Talc 3, and MDS Talc 4 with increasing aluminum substitution ratio at 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H.

Figure 3.7. Sessile drop contact angle as a function of aluminum lattice substitution in tetrahedral layer for the talc face surface as determined from experimental measurements and from molecular dynamics simulations.

Figure 3.8. MDS snapshots (two different directions) for interfacial water analysis at MDS Talc 1, aluminum ratio = 0, MDS Talc 2, aluminum ratio = 0.013, MDS Talc 3, aluminum ratio = 0.08, and MDS Talc 4, phlogopite, aluminum ratio = 0.25. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H.

75 of low aluminum substitution, to a hydrophilic state, in the case of a 0.25 aluminum substitution ratio, again illustrates the significance of aluminum impurity on the wetting characteristics as described at the molecular level. The water-excluded volumes in the case of extensive aluminum substituted crystals are smaller, which indicates stronger interaction of interfacial water with atoms of the face surface. The surface polarity of the talc structures changes from nonpolar to polar when the aluminum substitution increases, due to this interaction. The interfacial water analysis is of importance in understanding film thinning, rupture, displacement, and bubble attachment at talc surfaces, and how these flotation features are impacted by aluminum substitution in the tetrahedral layer.

3.5.1 Relative Number Density

The relative number density can be defined as the fraction of absolute density over an average number density for the bulk water. For oxygen atoms in the water phase, the relative number densities tend to remain at 1 for distances greater than about 8 Å away from the surface, which suggests little influence of the surface atoms on the water density beyond this point. The relative number density distribution function for the oxygen atoms in water molecules along the surface normal for the talc structures is shown in Figure 3.9.

The water density peak higher than ‘1’ can be interpreted as the “primary water density peak.”54 From the analysis of these peaks, a gap of approximately 3 Å is observed for MDS

Talc 2 and MDS Talc 3, which represent 0.013 and 0.08 aluminum substitution ratio, respectively, and 2.6 Å is observed for MDS Talc 4 (phlogopite, the hydrophilic surface state), which represents 0.25 extensive aluminum substitution ratio. As seen in Chapter 2, a gap of 3.2 Å was observed for MDS Talc 1 with zero aluminum substitution. This observation provides insight into the differences in interfacial water characteristics at a

Figure 3.9. Relative number density profiles at the MDS Talc 1, aluminum ratio = 0, MDS Talc 2, aluminum ratio = 0.013, MDS Talc 3, aluminum ratio = 0.08, and MDS Talc 4, phlogopite, aluminum ratio = 0.25.

77 hydrophobic surface where there are insignificant interactions between the water molecules, and hydrophilic surfaces where the interactions are significant.

3.5.2 Water Dipole Moment and Hydrogen Position

From the relative number density (Figure 3.8), the observed gap to the first oxygens is nearly 3 Å for low aluminum substitutions and 2.6 Å for extensive aluminum substitution. The water dipole orientations and hydrogen positions are similar for talc structures with low aluminum substitution but vary in the case of phlogopite with extensive aluminum substitution as shown in Figure 3.10. When we consider the water dipole orientations, the angle ‘α’ is around 90° to the surface normal line and is parallel to the face surface of the crystal (001 plane). Similarly, for the hydrogen positions, the angle ‘β’ is also around 90° to the surface normal and is parallel to the surface. Therefore, the water molecules are oriented in a plane parallel to the surface and have weak interactions with the surface. However, for phlogopite, the water dipole orientation, angle ‘α’, is around 130° to the surface normal, and the hydrogen positions, angle ‘β’ is around 35° to the surface normal, which suggests that the water molecules are in fact interacting with the face surface atoms. These results from the interfacial water analysis help to explain the observed variation in surface polarity.

Figure 3.10. Water dipole moment and hydrogen position of MDS Talc 1, aluminum ratio = 0, MDS Talc 2, aluminum ratio = 0.013, MDS Talc 3, aluminum ratio = 0.08, and MDS Talc 4, phlogopite, aluminum ratio = 0.25.

3.6 Bubble Attachment Results

Water film thinning, rupture, and displacement are expected to be impacted with change in the surface state of mineral surfaces.85-87 Bubble attachment time measurements have been done at the surface of talc structures with varying aluminum substitution.

Investigation of this film thinning, rupture, displacement, and subsequent bubble attachment has been studied using experimental and simulation methods to further understand how the change in surface state influences these fundamental flotation features.

In addition, film thickness measurements prior to film rupture have been measured and are reported Atluri et al.88 As seen in Chapter 2, the critical rupture thickness for talc sample is 56 nm, while the equilibrium film thickness at the phlogopite surface is 25 nm. Further discussion on the effect of polysaccharides on film thickness at a talc surface will be presented in Chapter 4

3.6.1 Experimental

To study the effect of aluminum substitution in the silica tetrahedral lattice, and its effect on water film stability, experimental and MDS bubble attachment time measurements were performed at the talc surface. Experimental bubble attachment time results complemented the contact angle results. With the low aluminum substitution ratio, the bubble readily attaches to the crystal surface and with the aluminum substitution ratio at 0.30, representing extensive substitution, there is no film rupture and no bubble attachment. Results are reported in Table 3.5.

It is interesting to note that the bubble attachment is nearly 50 milliseconds in 0.01 and 0.027 aluminum substituted crystal samples, and the bubble does not attach to the

Table 3.5. Experimental bubble attachment time for crystals with different levels of aluminum substituted in the tetrahedral sheet of the face surface.

Mineral Sample XPS Aluminum, Aluminum Bubble Attachment,

Atomic % Sites/Total Sites milliseconds

Talc 1(talc, light green) <0.1 0 28

Talc 2 (talc, dark green) 0.3 0.01 49

Talc 3 (talc, white) 0.6 0.027 52

Talc 4 (phlogopite) 7.0 0.30 No Attachment

81 surface of 0.30 aluminum substituted crystal which represents extensive substitution, while for zero aluminum substituted talc has a time of 28 milliseconds, as seen in Chapter 2. It is recognized that the water molecules interact with the surface atoms of the crystals in the case of high aluminum substitutions, preventing the bubble attachment on the surface of the crystal. The substitution also creates a charge imbalance when Al+3 is substituted for

Si+4. This causes the surface to change from a slightly nonpolar state to a polar state as aluminum substitution increases from zero to the substitution of 0.30 for Talc 4

(phlogopite). To explain this phenomenon at the molecular level, MD simulations have been performed with various aluminum substitutions presented in the next section.

3.6.2 MD Simulation

For measuring MDS bubble attachment time, different talc minerals, MDS Talc 1,

MDS Talc 2, MDS Talc 3, and MDS Talc 4 (phlogopite) were prepared with 0, 0.01, 0.08, and 0.25 percent aluminum to silicon substitution ratio in the tetrahedral layer. The crystal structure consists of atoms including Mg, Si, O, Al, and K. The nitrogen bubble along with the water molecules was put adjacent to the mineral (001) surface at a distance of 1 nanometer. The size of the mineral surfaces in the periodic box was about 136Å x 150Å.

Simulation time of 1 nanosecond was considered, where 500 picoseconds was the equilibration time for the nitrogen bubble to move closer to the surface, and another 500 picoseconds analysis time to check the interaction.

Bubble attachment MD simulation snapshots for MDS Talc 1, MDS Talc 2, and

MDS Talc 3 with varying aluminum substitution are presented in Figure 3.11, Figure 3.12, and Figure 3.13. To better understand the interaction between the nitrogen bubble and the surface in the absence of aluminum substitution, the atoms in the talc crystal are

Figure 3.11. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of an unsubstituted ideal talc crystal (MDS Talc 1) for interaction times of 0, 0.1, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H; Blue, N.

Figure 3.12. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of MDS Talc 1, aluminum ratio = 0.013 for interaction times of 0, 0.1, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H; Blue, N.

Figure 3.13. Sequence of MDS snapshots for nitrogen bubble attachment at the face surface of MDS Talc 3, aluminum ratio = 0.08 for interaction times of 0, 0.1, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H; Blue, N. 85 represented by spheres with a diameter of 1.8 Å, nitrogen molecules with a diameter of 3.8

Å, and the water molecules with a diameter 0.5 Å. In the simulation, as time progresses, the nitrogen bubble moves closer to the surface of the low aluminum substituted talc face surfaces, resulting in water film rupture and subsequent attachment. From the analysis of the nitrogen bubble as it moves closer to the talc surface, the water film starts to rupture and displacement occurs at 0.1 nanosecond and the diameter of the nitrogen bubble expands on the talc surface until an elapsed time of 1 nanosecond. The diameter of the three-phase line of contact gradually increases during this time. This three-phase line of contact formations is slower than the expansion for unsubstituted ideal talc as seen in

Chapter 2. Also, as seen from the previous section on interfacial water analysis, there is insignificant interaction between the water molecules and surface atoms (see Figure 3.7).

This results in water film rupture and displacement causing the nitrogen bubble to attach and expand at the surface of the ideal talc face surface where we can see increased resistance to bubble rupture and attachment.

Consideration of the nitrogen bubble attachment as the extent of aluminum substitution increases shows that the water film displacement decreases, and upon extensive aluminum substitution (phlogopite), attachment does not occur. For example,

MD simulated snapshots for bubble interaction in the case of extensive aluminum substitution (phlogopite surface) are presented in Figure 3.14. The nitrogen bubble does not attach to the surface of the phlogopite. From the analysis of the nitrogen bubble as it moves closer to the talc surface, there is no attachment of the nitrogen bubble at any time during the 1 nanosecond simulation. In this case, there is a significant interaction between the water molecules and surface atoms (see Figure 3.6), and the water film is found to be

Figure 3.14. Sequence of MDS snapshots for nitrogen bubble at the face surface of phlogopite (MDS Talc 4), with extensive aluminum substitution for interaction times of 0, 0.1, 0.5, and 1 nanosecond. The color codes for atoms are: yellow, K; cyan, Al; green, Si; purple, Mg; red, O; white, H; Blue, N.

87 stable at the phlogopite surface. The change in surface state from a hydrophobic state in the case of zero aluminum substitution to a hydrophilic state in the case of 0.25 aluminum substitution is a significant feature in the wetting characteristics of the talc face surfaces.

3.7 Summary

The surface properties and wetting characteristics of the talc face surface change with the substitution of aluminum for silicon in the tetrahedral layer of the talc crystal structure. This substitution results in a charge imbalance at the silica tetrahedral face surface, which is compensated by a univalent cation, usually potassium. The extent of aluminum substitution in the silica tetrahedral layer for different talc samples was examined by XPS, and the corresponding experimental sessile drop contact angles were found to decrease with an increase in the aluminum substitution, decreasing from about

66° degrees for low aluminum substituted talc to 0 degrees for more extensive substitution

(phlogopite). Similar results were found for MD simulated sessile drop contact angles.

In addition to sessile drop contact angles, the wettability characteristics of the talc face surface was studied by making bubble attachment time measurements. As anticipated the bubble attachment time increased with an increase in the fractional substitution of aluminum for silicon in the tetrahedral layer with 52 milliseconds when 0.027 sites are occupied by aluminum, and no bubble attachment when 0.30 sites are occupied by aluminum. These results indicate that interfacial water structure is an important factor which determines film stability and bubble attachment.

From interfacial water analysis by MDS, water molecules were found to interact with surface atoms at talc surfaces with more extensive aluminum substitution. However, insignificant interactions were observed between water molecules and surface atoms in the

88 case of the low aluminum talc structures. This is significant because the water film is stable in the case of an extensively aluminum substituted crystal which restricts the air bubble from attaching to the face surface. The variation in polarity of the talc structures as revealed by contact angle and bubble attachment time measurements can be explained by MD simulation of interfacial water analysis summarized in Table 3.6. It is interesting to see the primary water peak for the unsubstituted talc crystal was found to be at approximately 3.2

Å, gradually reducing to nearly 3 Å for low aluminum substituted crystals, and 2.6 Å for the extensive substituted phlogopite crystal. Similarly, the orientation of the water dipole moment and hydrogen position angles ‘α’ and ‘β’, respectively, change from a parallel orientation to the crystal surface in zero aluminum substituted crystal to having a slightly disordered orientation in low aluminum substituted crystals, and a highly disordered orientation in extensive aluminum substituted crystal. See Figure 3.9 and Figure 3.11. This change in orientation was due to the significant electrostatic interactions of the water molecules with the surface counter ions (K+) as the aluminum substitution increases. 89

Table 3.6. Summarized data on interfacial water analysis at the talc structures with increasing aluminum substitution ratio.

Interfacial Water Molecule Aluminum Exclusion Zone Crystal Orientation w.r.t Face Subtitution Ratio Thickness (Å) Surface MDS Talc 1 0 3.2 Parallel MDS Talc 2 0.01 3 Slightly Parallel MDS Talc 3 0.08 3 Slightly Parallel MDS Talc 4 0.25 2.6 Oblique, Diagonal

CHAPTER 4

ADSORPTION OF POLYSACCHARIDES ON TALC

4.1 Introduction

As discussed in Chapter 1, the sulfide minerals have a modest hydrophobic character making them easier to float. Frequently, the depression of talc is a significant problem in the flotation of PGM ores and other base metal sulfide ores.89 It is reported that the fast-floating gangue such as talc competes with valuable sulfide minerals for bubble space.90 This effect has a detrimental influence on flotation kinetics and concentrate grades.

One way of reducing the floatable gangue that reports to the concentrate is to use polysaccharide depressants such as starch, CMC, and guar gum which upon adsorption reduce the hydrophobic character of talc. Polysaccharide adsorption state at hydrophobic surfaces has received considerable attention in the literature. The selectivity and effectiveness of polysaccharide depressants for talc depression have been studied by many researchers.91-103 In this way, improved concentrate grade and sulfide mineral recovery is achieved. High molecular weight polysaccharides are better depressants for talc but are not selective and sometimes adsorb on the valuable sulfide minerals reducing the grade and recovery. In this chapter, fundamental issues associated with the thin liquid film stability and bubble attachment at the talc face surface are presented as influenced by polysaccharide depressants.

Research on the wetting characteristics of talc by Atluri et al. has been published recently.30 As discussed in Chapter 3, it was found that the surface properties and wetting characteristics of the talc face surface change with substitution of aluminum for silicon in the tetrahedral layer of the talc crystal structure. This substitution results in a charge imbalance at the silica tetrahedral face surface, which is compensated for by a univalent cation, usually potassium. The extent of aluminum substitution in the silica tetrahedral layer for different talc samples was examined by XPS, and the corresponding experimental sessile drop contact angles were found to decrease with an increase in the aluminum substitution, decreasing from about 80° for unsubstituted talc to 0° for more extensive substitution (phlogopite). The water film was found to be stable at the phlogopite surface due to the interaction between water molecules and the increased polarity of the surface state. This stable water film restricts the air bubble from attaching to such face surfaces.

However, in the absence of aluminum substitution, no interactions between the water molecules and the face surface were observed and the air bubble readily attaches to the face surface of talc. It is this hydrophobic surface state of talc that accounts for its flotation and the need for depressants such as the polysaccharides currently being used.

As mentioned, talc is a major gangue mineral in the flotation of platinum group metal (PGM) ores. Batch flotation studies on the Merensky ore by Wiese et al. and

O’Connor et al. indicated that the final grade and recovery of the sulfide ore is affected by the presence of naturally floatable minerals such as talc. The PGM ore contains base metal sulfide ores like pentlandite, pyrrhotite, chalcopyrite, and pyrite. These ores also contain a significant quantity of talc (1.2~1.3%) which is deleterious to flotation.104, 105 Hence, the depression of talc gangue during flotation is achieved by using polysaccharide depressants

92 such as modified guar gum and CMC to prevent the attachment of air bubbles at the talc particle surfaces.

Adsorption states of polysaccharides at various mineral surfaces can be due to electrostatic interactions, chemical and hydrogen bonding, and hydrophobic interactions.106-109 The nature of polysaccharide adsorption by talc has been explained as reported in the literature. Laskowski et al., Rath et al. and Liu et al. discussed the importance of metallic sites for the adsorption of polysaccharides by acid/base interaction.92, 100, 101 In contrast, Brossard et al., Du and Miller, Jenkins and Ralston, and

Steenberg and Harris suggested that the adsorption of polysaccharides on talc was due to hydrophobic interactions between the talc basal plane surface and the hydrophobic moieties of the polysaccharide molecule.29, 102, 103, 110 Polysaccharides frequently adsorb at naturally hydrophobic surfaces, which can be described by Langmuir adsorption isotherms. In the case of dextrin, adsorption is about the same regardless of the hydrophobic mineral surface which will be discussed later.

Recent studies were done by Jueying Wu et al. and Shrimali et al. on bubble-particle interactions at hydrophobic surfaces in the presence of polysaccharides such as CMC, and starch. The use of polysaccharides reduces the water contact angle at the talc surface and also increases the bubble attachment time.83, 87 As the concentration of the dissolved polysaccharide increases adsorption increases, film stability increases, and bubble attachment time increases reaching a point where the wetting film is stable and attachment does not occur. In this way, polysaccharides serve as talc depressants when needed in the flotation of sulfide ores. A particularly important parameter is the effect of polysaccharide on the bubble attachment time, which is significantly larger than the bubble attachment

93 time in the absence of polysaccharide. In addition to looking at the increase in bubble attachment time, water film stability plays a crucial role in determining bubble attachment at mineral surfaces. In this regard, water film thickness measurements can be made possible for many mineral surfaces using Synchronized Triwavelength Reflection Interferometry

Microscopy (STRIM).53 The results from such measurements are reported later in this chapter.

4.2 Materials and Methods

4.2.1 Minerals and Reagents

High-quality Talc 1 crystal samples obtained from Argonaut Mine, Vermont, were used in this research for contact angle, bubble attachment time, film thickness measurements, and AFM imaging. Fresh crystal surfaces were prepared by using scotch tape to remove the top layers and expose a fresh surface. Care was taken to assure that the surface was free from imperfections and that a flat surface was prepared for contact angle and AFM measurements. The rms roughness of the Talc 1 sample was found to be 1.6 nm.

Polysaccharides were obtained from Sigma Aldrich and include guar gum, corn starch, and dextrin. Stock solutions of these polysaccharides were prepared using the following methods. For guar gum, 0.05 g was added to a 100 ml solution of 1 mM KCl which is rapidly stirred in a beaker. For corn starch, 1 g was added to 10 ml of 1 M NaOH in a beaker. After a gel was formed, 15 ml of 1 mM KCl solution was added and the solution was stirred overnight. The solution was then diluted to make a stock solution of 1000 ppm.

For dextrin, 0.05 g was added to a 100 ml solution of 1 mM KCl and heated to 80°C. All the solutions were stirred overnight for complete dissolution, and were used within 3 days 94 of preparation.

Potassium chloride (KCl) used to control ionic strength was obtained from Sigma-

Aldrich. Deionized water was obtained from a Millipore system in the laboratory with specific conductance of 18 MΩ.cm.

4.2.2 Captive Bubble Contact Angle

Fresh talc surfaces were obtained by using scotch tape to remove the top layer and expose a fresh surface. The distance between the bubble needle and the sample surface was kept constant (5 mm) for each captive bubble contact angle measurement. All the measurements were done in 1 mM KCl solution. Before each contact angle measurement, the sample was conditioned in the polysaccharide solution for 15 minutes. The measurement was made in the same solution. At least 5 contact angle measurements were obtained at different locations and they were averaged to obtain the reported contact angle.

Variation in contact angle measurements was ±3°.

4.2.3 Bubble Attachment Time

For bubble attachment time measurements, a high-speed camera, KODAK

EKTAPRO, was used to record videos at 1000 frames/second. Fresh talc surfaces were conditioned for 15 minutes in the polysaccharide solution before the start of the experiment. Similar to contact angle measurements, the conditioned samples were placed in a glass cell and a captive bubble released 5 mm from the surface. By checking the frames in the video, the time required for bubble attachment was calculated, which includes film thinning time and time required for expansion of the three-phase line of contact. The

95 variation in attachment time was ± 8 ms at zero polysaccharide concentrations, ±20 ms in the case of dextrin, and ± 10 s with an increase in the concentration of corn starch and guar gum, and no attachment at high concentration upto 5 minutes of observation.

4.2.4 Atomic Force Microscopy (AFM)

AFM imaging was done using a Nanoscope III A (Digital Instruments), and SNL-

10 silicon probes (Bruker), which are specifically designed for tapping mode images in liquids. The tip had a resonant frequency varying between 10 and 100 Hz and had a spring constant varying from 20 N/m to 35 N/m. The cleaning procedure for the tip, as well as the cantilever, involved cleaning them with acetone and ethanol, rinsing with milli-Q water, followed by drying with ultrapure nitrogen. The contact mode images were taken in a liquid cell and subjected to 2nd-order flattening and 2nd-order low pass filtering. In the case of imaging polysaccharides, both height and deflection error images were taken to describe molecular organization at the talc face surface.

4.2.5 Film Thickness Measurements by Synchronized Triwavelength

Interferometry Microscopy (STRIM)

The STRIM deploys three synchronized high-speed cameras that record the interference fringes at three different wavelengths (λ = 460, 527, and 620nm) simultaneously. The procedure uses RGB light wavelengths to calculate the solution film thickness separating a bubble from a mineral surface. It consists of an inverted microscope,

RGB light sources, and a glass cell with a hydrophobic base to hold the air bubble of 0.4 mm diameter. The glass cell rests on a piezo stage that moves up with a speed of

96 approximately 1 µm/s. Film thickness is determined from interference fringes which form when the bubble moves close to the mineral surface. All solutions were prepared in 1 mM

NaCl background electrolyte. The solution film thickness values were then calculated. The detailed experimental method can be found elsewhere.53 The obtained interference fringes were analyzed to determine the spatiotemporal thickness profile of the wetting film for various surfaces.

4.2.6 Microflotation

The microflotation tests for talc with MIBC (7 ppm) as frother and corn starch (100 ppm) as depressant were conducted using a 112-ml column cell of diameter 2.5 cm with a porous sintered glass bottom for gas dispersion and a magnetic stirrer. Ultrapure nitrogen was used to produce the bubbles needed for the flotation process, and was maintained at

60 ml/min. The mineral sample (1 g) of 150x106 µm was conditioned by stirring for 5 minutes in corn starch solution (80 ml), followed by addition of MIBC. After that, the sample, together with solution, was transferred to the flotation cell and the flotation cell was filled with the same solution used for conditioning. The flotation process was continued for 5 minutes using nitrogen gas. The concentrate was collected after flotation using a filter paper, which was then dried and weighed to determine the recovery. Micro- flotation was carried out with 1 mM KCL as the background electrolyte.

4.3 Bubble Attachment Results

Film thinning, rupture, displacement, and subsequent bubble attachment time have been studied using experimental and simulation methods at the talc surface. 97

4.3.1 Captive Bubble Contact Angle

Captive bubble contact angle measurements in the presence of low concentrations of polysaccharide solutions show that there is a slight decrease in the contact angle with increasing concentrations of the polysaccharide as shown in Table 4.1, Table 4.2 and Table

4.3 for guar gum, corn starch, and dextrin, respectively. From the contact angle measurements at low concentrations of polysaccharides, we observe the contact angle changes from approximately 80° to 60°, in the case of all three polysaccharides. This reduction in contact angle is not as significant as the increase in bubble attachment time which is discussed in the next section.

4.3.2 Bubble Attachment Time

Bubble attachment time measurements have been done at the face surface of talc to complement the captive bubble contact angle measurements and to check the wetting efficiency of the three polysaccharides, guar gum, starch, and dextrin. The results are presented in Table 4.1, Table 4.2, and Table 4.3 for guar gum, corn starch, and dextrin, respectively. Contact angle measurements in the presence of 5 ppm corn starch solution show that there was a slight decrease in the contact angle but a significant increase in bubble attachment time values. It should be noted that there is a 1000-fold increase in the value of bubble attachment time, from milliseconds in the absence of corn starch and guar gum to seconds when they are present in the system. This increase in the bubble attachment time was similar to the observation made by Beaussart et al., where adsorption of dextrin- based polymers leads to a significant increase in bubble attachment time, but only a slight change in contact angle.111, 112

Table 4.1. Bubble attachment with guar gum adsorbed at the talc face surface.

Conc. (Guar Gum, Contact Angle Bubble Attachment Time

ppm) (degrees) (seconds)

0 81 0.03

5 65 120

15 NA NA

25 NA NA

100 NA NA

Table 4.2. Bubble attachment with corn starch adsorbed at the talc face surface.

Conc. Starch Contact Angle Bubble Attachment Time

(ppm) (degrees) (seconds)

0 79 0.025

5 72 110

25 56 200

50 NA NA

100 NA NA

100

Table 4.3. Bubble attachment with dextrin adsorbed at the talc face surface.

Contact Angle Bubble Attachment Time Conc. (Dextrin, ppm) (degrees) (seconds)

0 81 0.023

5 76 0.065

25 70 0.100

50 60 0.128

100 NA NA

101

It is interesting to note that the air bubble is prevented from attachment at different concentrations depending on the polysaccharides used. For example, the minimum concentrations to prevent attachment are 15 ppm for guar gum, 25 ppm for starch, and 100 ppm for dextrin. This order can be explained by the extent of polysaccharide adsorption, as described by the adsorption isotherms. When the surface is saturated with polysaccharides, bubble attachment is not possible, as is evident from the adsorption isotherms shown in Figure 4.1.99, 102 The decreasing order of polysaccharide effectiveness in preventing bubble attachment is as follows: guar gum>starch>dextrin, which is the same order observed for effectiveness as a depressant in flotation experiments.99, 100

Also, in the presence of polysaccharide dextrin (30 ppm), small droplets of water

(~80 μm) have been observed at the talc surface inside the attached air bubble, where the water displacement is not complete as shown in Figure 4.2. In the absence of dextrin, water is completely displaced and water microdrops were not observed. It is expected that adsorbed dextrin molecules are present as islands at the surface of talc.113 The water binds to adsorbed dextrin molecules and the water droplets are not displaced, but are stabilized at the talc surface inside the attached air bubble. This is an important observation, because it helps to explain the increase in bubble attachment time and three-phase line of contact.

4.4 Film Thickness Results

4.4.1 Synchronized Triwavelength Reflection Interferometry

Microscopy (STRIM)

The stability of wetting films formed between air bubbles and mineral surfaces were studied using a newly developed Synchronized Triwavelength Reflection

102

Figure 4.1. Adsorption isotherms for guar gum, starch, and dextrin adsorption by talc.99, 102

103

Figure 4.2. Presence of water droplets inside the attached air bubble with dextrin adsorbed at the talc surface (magnification 10x).

104

Interferometry Microscopy (STRIM) technique. The STRIM experiments on bubble attachment for determination of film thickness and stability were conducted by bringing an air bubble towards a flat mineral surface at a constant nominal velocity of 1.1 μm/s, while monitoring the time evolution of separation distance. At Re << 1, the bubble-mineral attachment process is considered as a quasistatic approach.114 The radii of the air bubbles are in the range of 0.35-0.45 mm.

The effect of the adsorbed polysaccharides like guar gum, starch, and dextrin on the stability of wetting films formed on hydrophobic talc surfaces was studied. Table 4.4 shows the result obtained with talc surfaces in the presence of 0-100 ppm guar gum. At guar gum concentrations less than 5 ppm, the wetting films formed on talc surfaces were unstable with a critical rupture thickness (hc) in the range of 56-58 nm. When the concentration of guar gum was increased to 15 ppm or above, the wetting films became stable with no bubble attachment. The present results confirm that the adsorption of guar gum on talc surfaces lowers the surface hydrophobicity of talc surfaces. The equilibrium film thickness (he) of 55-58 nm at 15-100 ppm guar gum concentrations is about the same as the critical rupture thickness at lower concentrations, and indicates that the adsorption of guar gum on talc surfaces does not significantly change the surface potentials.

The critical rupture thicknesses (hc) obtained at 0 and 5 ppm of guar gum were the same, despite the significant difference in bubble attachment times as shown in Table 4.1.

This discrepancy in bubble attachment time and water film thickness might be associated with different hydrodynamic conditions in the two experimental setups. For wetting film stability measurements (i.e., quasistatic measurements), in which the bubble approach velocity was 1.1 μm/s, the film rupture might be governed dominantly by the surface

105

Table 4.4. Characteristics of wetting films on talc surfaces at different guar gum concentrations.

Dosage (ppm) hc (nm) he (nm) Stability

0 56 - Unstable

5 58 - Unstable

15 - 58 Stable

25 - 55 Stable

100 - 56 Stable

106 hydrophobicity of substrates.61 In this regard, the talc surface switched from a hydrophobic state to a hydrophilic state when the guar gum dosage was increased from 5 ppm to 15 ppm. On the contrary, for dynamic bubble attachment time measurements, in which the free rising air bubble was impacting a flat solid plate, the terminal rising velocity before impact was estimated to be 0.3 m/s.115 Such high impacting velocity results in a large interfacial deformation and potentially a turbulence that benefits the film rupture.

Table 4.5 and Table 4.6 show the results of wetting film stability measurements obtained with starch and dextrin as the depressant. Starch molecules have larger molecular weight (MW) than dextrin. The wetting films formed on talc surfaces were unstable and ruptured at a thickness of 56-64 nm for starch concentrations less than 5 ppm. As the starch dosage was further increased to 25 ppm, the film became meta-stable. In this case, the wetting films ruptured in half of the total events while remained stable for the other half of the events. When the starch concentration was further increased to 50 ppm and 100 ppm, the wetting film became stable with an equilibrium film thickness (he) of 47-50 nm. It is evident that the inherent hydrophobicity of the fresh talc surface was altered by the adsorption of starch molecules at concentrations exceeding 25 ppm.

The results obtained with dextrin are slightly different from those obtained with starch. Dextrin exhibits a similar molecular structure to starch, but it has a lower weight.

The results show that the wetting films ruptured at a thickness of 55-66 nm but became stable when the dextrin dosage was increased from 50 to 100 ppm with an equilibrium film thickness of 51 nm. The critical dosage, at which the surface lost its hydrophobic character with stable wetting film, there is a slight difference between starch and dextrin.

107

Table 4.5. Characteristics of wetting films on talc surfaces at different corn starch concentrations.

Dosage (ppm) hc (nm) he (nm) Film Stability

0 56 - Unstable

5 64 - Unstable

25 56 - Metastable

50 - 50 Stable

100 - 47 Stable

108

Table 4.6. Characteristics of wetting films on talc surfaces at different dextrin concentrations.

Dosage (ppm) hc (nm) he (nm) Film Stability

0 56 - Unstable

5 61 - Unstable

25 66 - Unstable

50 55 - Unstable

100 - 51 Stable

109

4.4.2 Discussion

For guar gum, it can be seen from Figure 4.1 that the Langmuir adsorption isotherm plateaus at a concentration of about 20 ppm, and a saturation adsorption density of nearly

1.2 mg/m2 is observed. Under these conditions the water film is stable and the bubble does not attach at the talc face surface. See Table 4.1. Starch, being of higher molecular weight, exhibits an adsorption density of nearly 3 mg/m2 at about 30 ppm. See Figure 4.1. At such high adsorption density, the surface changes to a hydrophilic state, and the film becomes stable at 50 ppm; the bubble does not attach, as shown in Table 4.2. Dextrin exhibits

Langmuir adsorption as shown in Figure 4.1. Adsorption remains constant after 50 ppm, and the surface becomes more hydrophilic, with a stable film which prevents bubble attachment. See Table 4.6. At 50 ppm dextrin the talc surface with adsorbed dextrin still has some hydrophobicity, but at 100 ppm the surface is fully covered. The adsorption density is relatively low (0.6 mg/m2) at 50 ppm compared to guar gum and starch.

Thus, at higher concentrations and higher adsorption densities the talc surface state changes from a hydrophobic state to a hydrophilic state. These adsorption density results from the literature are in good agreement with results from air bubble attachment time measurements and with the results from water film thickness measurements.

Based on AFM measurements and other measurements, such as electrophoresis, the adsorbed polysaccharides are believed to extend, at most, 10 nm from the talc face surface into solution representing about 20% of the stable, equilibrium film thickness.83, 99, 102 The polysaccharide adsorption is considered to be due to hydrophobic interaction between the polymer and the naturally hydrophobic talc face surface.83, 102, 103 Of course, portions of the extended adsorbed polysaccharides are extensively hydrated, which facilitates H-bonding

110 with free water molecules and consequently stabilizes the film, creating a hydrophilic state at the talc surface.

4.5 Adsorption State

As discussed in the introduction to Chapter 4, the adsorption state of polysaccharides at the mineral surfaces can be due to electrostatic interactions, chemical and hydrogen bonding, or hydrophobic interactions. The description of electrostatic interactions in the case of talc is dismissed, since the face surface is nonpolar, and in some cases negatively charged, and the polysaccharides are neutral with no charge.

To provide further consideration on the nature of polysaccharide adsorption, corn starch has been used along with sample crystals of highly ordered pyrolytic graphite

(HOPG) and high-quality talc crystals (Talc 1). From the XPS analysis, no metallic impurities were identified on the HOPG crystal even after doing high resolution scans for

Fe, Ca and Mg. It contains 99.5% C and 0.5 % O by weight percent. Since no metallic impurities could be detected at the HOPG surface there should be no contribution to the adsorption of corn starch at the surface due to acid/base interaction.

As seen from Figure 4.3 and Figure 4.4, the fresh surfaces of talc and HOPG are hydrophobic at all pH values and have very small bubble attachment times ranging from

25 ms to 100 ms. Also, the contact angle and bubble attachment time measurements for talc and HOPG were independent of the solution pH, which suggests electrostatic interactions do not play a role in adsorption. Contact angle and bubble attachment time measurements in the presence of increasing corn starch solutions showed that there was a decrease in the contact angle and significant increase in bubble attachment time values for

111

Figure 4.3. Captive bubble contact angle at HOPG surface with varying pH and varying corn starch concentration.

112

Figure 4.4. Captive bubble contact angle at the talc surface with varying pH and varying corn starch concentration. 113 talc and HOPG at all pH values, as seen in Table 4.7 and Table 4.8. It is interesting to note that there is a 1000-fold increase in the value of bubble attachment time, similar to the results obtained for high-quality talc. These results suggest that hydrophobic interactions are the governing factor for the adsorption of polysaccharides at these ideal hydrophobic mineral surfaces.

4.5.1 Atomic Force Microscopy

Further, to visually confirm the corn starch adsorption at HOPG and high-quality talc surfaces tapping mode AFM imaging was done with corn starch solutions. The tapping mode AFM was done using a Nanoscope III A (Digital Instruments), and RTESP-300 silicon probes (Bruker), which are specifically designed for tapping mode in air. The tip had a resonant frequency varying between 200 and 400 Hz and had a spring constant varying from 20 N/m to 80 N/m. The cleaved surfaces of talc and HOPG were conditioned with 50 ppm corn starch (pH = 8) solution prepared in 1mM KCl solution for 30 minutes.

To remove the crystallized and excess nonadsorbed corn starch, the samples were cleaned with milli-Q water followed by gentle drying in a stream of ultrapure nitrogen gas.

As seen from Figure 4.5, the clean freshly cleaved surface of HOPG was smooth with a rms roughness of 0.3 nm and PTV (Peak to Valley) distance of 3.4 nm. Corn starch adsorbs at an HOPG surface in the form of a polymer network with the surface apparently exposed in cellular regions. These cells have a size ranging from 50 x 50 nm to 538 x 334 nm, and the percentage coverage was approximately 60%. The rms roughness as well as the PTV distance of HOPG significantly increased after the adsorption of corn starch. RMS roughness of the adsorbed HOPG was 1.6 nm and the PTV distance was 7.2 nm. Similarly, for freshly cleaved surface of talc, a rms roughness of 0.7 nm and PTV (Peak to Valley) 114

Table 4.7. Bubble attachment time for HOPG with increasing corn starch concentration.

Starch 0 ppm Starch 5 ppm Starch 10 ppm Starch 20 ppm Starch 30 ppm

Time Time Time Time Time pH pH pH pH pH (sec) (sec) (sec) (sec) (sec)

3.2 0.04 3.2 50 3.2 65 3.2 N.A. 3.2 N.A.

5.3 0.05 5.3 55 5.3 85 5.3 N.A. 5.3 N.A.

8.1 0.10 8.1 50 8.1 65 8.1 N.A. 8.1 N.A.

10.5 0.09 10.5 55 10.5 70 10.5 N.A. 10.5 N.A.

N.A. = No Attachment 115

Table 4.8. Bubble attachment time for talc with increasing corn starch concentration.

Starch 0 ppm Starch 5 ppm Starch 10 ppm Starch 20 ppm Starch 30 ppm

Time Time Time Time Time pH pH pH pH pH (sec) (sec) (sec) (sec) (sec)

3.2 0.44 3.2 100 3.2 70 3.2 180 3.2 N.A.

5.3 0.025 5.3 110 5.3 97 5.3 200 5.3 N.A.

8.1 0.55 8.1 120 8.1 117 8.1 210 8.1 N.A.

10.5 0.74 10.5 125 10.5 120 10.5 220 10.5 N.A.

N.A. = No Attachment

116

Figure 4.5. TMAFM height and phase image of fresh HOPG (A, B) and HOPG conditioned with 50 ppm corn starch (C, D). Corn starch adsorbs at HOPG as a cellular branched chain structure. 117 distance of 2.5 nm was observed. Corn starch adsorbs at talc in the form of elongated patches. These patches have a size ranging from 840 x 150 nm to 3000 x 300 nm, and the percentage coverage was approximately 55% as shown in Figure 4.6. The rms roughness as well as the PTV distance for talc significantly increased after the adsorption of corn starch. RMS roughness of the adsorbed talc was 8 nm and the PTV distance was 32 nm.

Further, high-quality talc crystals are conditioned in the solutions of guar gum and dextrin to observe the adsorption pattern. Images of the talc surface with 30 ppm guar gum are particularly interesting since an ordered structure is found as shown in Figure 4.7. The

100 nm by 100 nm images were subjected to second order flattening and second order low pass filtering. The clean freshly cleaved surface of talc is smooth with an rms roughness of

4 nm. The same surface with adsorbed guar gum reveals some ordering of the adsorbed guar gum molecules with a typical cylindrical structure size of approximately 7 nm in length and a height of about 0.7 nm as determined from a number of line profiles.

In the case of dextrin, small islands were formed on the surface of talc as shown in

Figure 4.8. These islands have a size of from the line analysis. The clean freshly cleaved surface of talc is smooth with an rms roughness of 1.4 nm. The same surface with adsorbed dextrin reveals some network of islands of the adsorbed dextrin molecules with a typical cylindrical structure size of approximately 50 nm in length and a height of about 0.6 nm as determined from line analysis.

4.5.2 Microflotation

Microflotation experiments were performed for talc and graphite particles (150 x

106 µm) in 100 ppm corn starch solution with 1 mM KCl as the background electrolyte, at 118

Figure 4.6. TMAFM height and phase image of fresh talc (A, B) and talc conditioned with 50 ppm corn starch (C, D). Corn starch adsorbs at the (001) talc surface in the form of elongated patches. 119

Figure 4.7. Contact mode height (A) and deflection error (B) images of fresh talc conditioned with 30 ppm guar gum. Guar gum adsorbs at the (001) talc surface in the form of cylindrical patches. 120

Figure 4.8. TMAFM height (A) and phase (B) images of fresh talc conditioned with 50 ppm dextrin. Dextrin adsorbs at the (001) talc surface in the form of small network of islands.

121 four pH values. These experiments were performed to analyze the effect of pH on flotation recovery. As shown in Figure 4.9 and Figure 4.10, flotation recoveries for graphite and talc in 1 mM KCl were 57 % and 99 %, respectively, and were independent of pH. These results are in agreement with contact angle results for fresh HOPG and fresh talc which showed that both surfaces were hydrophobic at all pH values. The relatively low flotation recovery of synthetic graphite in the presence of MIBC as a frother is consistent with previously published results.116

In the presence of 100 ppm corn starch, flotation recovery for graphite was below

10 % and flotation recovery for talc was below 22 % for all pH values. These results indicate that corn starch adsorbs at graphite and talc making them hydrophilic. As the XPS and XRD results showed no major impurities in graphite or talc, adsorption was not due to metallic impurities. These results also suggest the role of hydrophobic interactions, which account for corn starch adsorption at the graphite and talc surfaces.

4.6 Summary

Contact angle and bubble attachment time measurements have been presented for the talc surface and talc surface with adsorbed polysaccharides. In the absence of polysaccharides, the talc surface is hydrophobic, with a contact angle of 80°. Under these circumstances, the water film thins, ruptures, and the bubble attaches to form a three-phase line of contact. In contrast, as seen in Chapter 2 and Chapter 3, the hydrophilic phlogopite surface, with structure similar to that of talc, has a water contact angle of 0º and a stable water film.

Addition of polysaccharides has a significant influence on film stability and bubble 122

Figure 4.9. Flotation recovery of synthetic graphite in the presence and absence of corn starch with varying pH.

123

Figure 4.10. Flotation recovery of talc in the presence and absence of corn starch with varying pH.

124 attachment at the talc face surface. As the concentration of the polysaccharides increases, the contact angle decreases slightly, whereas the bubble attachment time increases significantly. Sometimes the increase in bubble attachment time is 1000-fold, as seen in the case of guar gum and starch. At high concentrations of the polysaccharides, after the talc surface is saturated, the surface state changes from a hydrophobic state to a hydrophilic state. The water film becomes stable and does not rupture, thus preventing the air bubble from attaching at the talc face surface. Also, the adsorption state of polysaccharides at high- quality hydrophobic surfaces like talc and HOPG appears to be due to the hydrophobic interaction between the nonpolar surface and the hydrophobic moieties of the polysaccharide molecule.

Film thickness measurements provide insights into bubble attachment at various mineral surfaces. The values of critical rupture thickness (hc) values were found to be about

55-60 nm, and in the case of stable films, equilibrium film thickness (he) values were about

54-65 nm. This similarity in the film thickness values at surfaces with adsorbed polysaccharides will need to be further investigated. In the case of talc surfaces with adsorbed polysaccharides, the wetting film gained stability and no bubble attachment occurred when the polysaccharide concentration reached a critical value. It has been found that guar gum is more effective than starch and dextrin in altering the surface hydrophobicity of talc surfaces.

CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS

In conclusion, talc, a trilayer phyllosilicate, has a nonpolar basal plane, results in a hydrophobic face surface, and generally accounts for its flotation behavior. However, based on the results presented and those found in the literature, it is evident that the wetting characteristics change and the flotation response of talc influenced by substitution of aluminum in the tetrahedral layer of the talc lattice. The effect of lattice substitution of aluminum for silicon in the tetrahedral layer of the talc face surface reduces the natural hydrophobicity. In the case of extensive aluminum substitution, where every fourth silicon atom is substituted by aluminum, the crystal structure is identified as phlogopite, and the surface becomes hydrophilic, as demonstrated from experimental and MDS studies.

The wettability of minerals is related to the interaction of interfacial water molecules with atoms exposed at the crystal surface. In the case of talc with no aluminum substitution, interfacial water analysis reveals an “exclusion zone” or “water-excluded volume,” which suggests relatively weak, insignificant interactions between interfacial water molecules and atoms of the talc face surface. However, in the case of phlogopite, with more extensive aluminum substitution of 0.25, a charge imbalance is created which is compensated by potassium counter ions at the crystal face surface. Consequently, relatively stronger interactions of interfacial water molecules occur with the potassium counter ions

126 that account for wetting of the phlogopite face surface. This interaction was observed from the exclusion zone thickness which decreased from 3.2 Å to 2.6 Å as the aluminum substitution ratio increased, from the unsubstituted talc crystal to the extensive substituted phlogopite, respectively. Also, the water molecules were oriented parallel to the face surface in the case of unsubstituted crystal, and as the aluminum substitution ratio increased, the orientation changed to slightly parallel and finally diagonal to the face surface of the crystals.

Experimental results associated with bubble attachment at the talc face surface in the presence of polysaccharides have been presented based on contact angle, bubble attachment time, and water film thickness measurements. Of course, bubble attachment consists of water film thinning, film rupture, attachment and formation of a three-phase line of contact, and these issues have been considered. In the absence of polysaccharides, the bubble attachment at the talc face surface was instantaneous with a time of approximately 25 milliseconds. As the concentration of the polysaccharides increased, the bubble attachment time increased to seconds, and finally at complete surface coverage, the bubble did not attach at the surface of talc. The film stability and the air bubble attachment from the film thickness and bubble attachment time measurements agreed well with each other qualitatively. But further experiments are warranted to relate the quasistatic conditions in film thickness and kinetic conditions in bubble attachment time. Also, from the AFM images, recall that guar gum is composed of galactose monomer rings having a radius of gyration of 1 nm similar to glucose. Consequently, the AFM cylindrical structures in Figure 4.7 appear to be single molecules of guar gum which has a molecular weight of

250000 Daltons.117,118

127

The results from bubble attachment time measurements and water film thickness measurements suggest that talc depression by polysaccharides is achieved by creation of a stable wetting film. The stability of the film and its thickness has been established under quasistatic conditions but, as expected, bubble attachment time measurements suggest that film rupture and attachment time will depend on the impact velocity. These results help us understand and describe the stability of water films at hydrophobic and hydrophilic surfaces. The results from this dissertation research will helps us in understanding further details of film rupture and displacement during bubble attachment. Further, these results and analyses provide a foundation for continued research, paving the way for the design of improved reagents for talc flotation as an industrial mineral product, and for talc depression in the recovery of sulfide mineral concentrates.

In the future, the following studies of talc surface chemistry will be extended to benefit research areas and fields such as chemical engineering and nanotechnology. The research results would contribute many novel procedures and opportunities to expand the field of study to keep up with the novel research on the layered silicate minerals such as talc, kaolinite, muscovite, etc. Future research topics include:

1. Nanobubbles and nanodrops imaging using atomic force microscopy by different

procedures including solvent exchange, milled AFM probes, and simple deposition

of a macro drop followed by blasting with a stream of ultrapure nitrogen gas to

leave some micro/nano drop residue at the surface.

2. The microscopic mechanisms associated with the effect of hydrodynamic

conditions on bubble attachment are still ambiguous, requiring future research on

film thickness and film stability.

128

3. Molecular dynamics simulation of polysaccharide molecules at the surface of talc,

different talc structures, and their influence on bubble attachment.

4. Characterizing the volume of the polysaccharide molecules in solution.

REFERENCES

1. Fuerstenau, M. C.; Jameson, G. J.; Yoon, R.-H., Froth Flotation: A Century of Innovation. SME: 2007.

2. Wills, B. A.; Finch, J., Wills' Mineral Processing Technology: An Introduction To The Practical Aspects Of Ore Treatment And Mineral Recovery. Butterworth-Heinemann: 2015.

3. Gaudin, A. M., Flotation Fundamentals. University of Utah: 1928.

4. Fuerstenau, M. C.; Chander, S.; Woods, R., Sulfide mineral flotation. Froth Flotation A Century of Innovation. Society for Mining, Metallurgy, Exploration, Inc., Littleton, Colorado 2007, 425-464.

5. Chander, S.; Fuerstenau, D., On the natural floatability of molybdenite. Transactions AIME 1972, 252, 62-69.

6. Tolley, W.; Kotlyar, D.; Van Wagoner, R., Fundamental electrochemical studies of sulfide mineral flotation. Minerals Engineering 1996, 9 (6), 603-637.

7. Richardson, P.; Stout Iii, J.; Proctor, C.; Walker, G., Electrochemical flotation of sulfides: Chalcocite-ethylxanthate interactions. International Journal of Mineral Processing 1984, 12 (1-3), 73-93.

8. Chander, S., Electrochemistry of sulfide flotation: Growth characteristics of surface coatings and their properties, with special reference to chalcopyrite and pyrite. International Journal of Mineral Processing 1991, 33 (1-4), 121-134.

9. Matis, K. A., Flotation Science and Engineering. CRC Press: 1994.

10. Chander, S., A brief review of pulp potentials in sulfide flotation. International Journal of Mineral Processing 2003, 72 (1-4), 141-150.

11. Rubio, J.; Capponi, F.; Rodrigues, R.; Matiolo, E., Enhanced flotation of sulfide fines using the emulsified oil extender technique. International Journal of Mineral Processing 2007, 84 (1-4), 41-50.

130

12. Alvarez-Silva, M.; Uribe-Salas, A.; Mirnezami, M.; Finch, J., The point of zero charge of phyllosilicate minerals using the Mular–Roberts titration technique. Minerals Engineering 2010, 23 (5), 383-389.

13. Schulze, D. G., An introduction to soil mineralogy. Soil Mineralogy with Environmental Applications 2002, (soilmineralogyw), 1-35.

14. Ndlovu, B.; Becker, M.; Forbes, E.; Deglon, D.; Franzidis, J.-P., The influence of phyllosilicate mineralogy on the rheology of mineral slurries. Minerals Engineering 2011, 24 (12), 1314-1322.

15. Glotch, T. D.; Rossman, G. R.; Aharonson, O., Mid-infrared (5–100 μm) reflectance spectra and optical constants of ten phyllosilicate minerals. Icarus 2007, 192 (2), 605-622.

16. Barron, P. F.; Slade, P.; Frost, R. L., Solid-state silicon-29 spin-lattice relaxation in several 2: 1 phyllosilicate minerals. The Journal of Physical Chemistry 1985, 89 (15), 3305-3310.

17. Abdul, B. Surface Chemistry of Talc Surfaces. Department of Metallurgical Engineering, University of Utah, 2008.

18. Yin, X.; Gupta, V.; Du, H.; Wang, X.; Miller, J. D., Surface charge and wetting characteristics of layered silicate minerals. Advances in Colloid and Interface Science 2012, 179-182, 43-50.

19. Harvey, C. C.; Murray, H. H., Industrial clays in the 21st century: A perspective of exploration, technology and utilization. Applied Clay Science 1997, 11 (5-6), 285-310.

20. Franco, F.; Pérez-Maqueda, L.; Pérez-Rodrıguez, J., The effect of ultrasound on the particle size and structural disorder of a well-ordered kaolinite. Journal of Colloid and Interface Science 2004, 274 (1), 107-117.

21. Hu, Y.; Wei, S.; Hao, J.; Miller, J.; Fa, K., The anomalous behavior of kaolinite flotation with dodecyl amine collector as explained from crystal structure considerations. International Journal of Mineral Processing 2005, 76 (3), 163-172.

22. Brown, G., Crystal structures of clay minerals and related phyllosilicates. Philosophical Transactions of the Royal Society of London, Series A 1984, 311 (1517), 221-240.

23. Hower, J.; Mowatt, T. C., The mineralogy of illites and mixed-layer illite/montmorillonites. American Mineralogist 1966, 51 (5-6).

24. U.S. Geological Survey, 2017, Mineral commodity summaries 2017: U.S. Geological Survey, 202 p., https://doi.org/10.3133/70180197. . 2017. 131

25. Liu, Q.; Zhang, Y.; Laskowski, J., The adsorption of polysaccharides onto mineral surfaces: An acid/base interaction. International Journal of Mineral Processing 2000, 60 (3-4), 229-245.

26. Sikora, M.; Schilling, C. H.; Tomasik, P.; Li, C., Dextrin plasticizers for aqueous colloidal processing of alumina. Journal of the European Ceramic Society 2002, 22 (5), 625-628.

27. Maier, H.; Anderson, M.; Karl, C.; Magnuson, K.; Whistler, R. L., Guar, locust bean, tara, and fenugreek gums. In Industrial Gums (Third Edition), Elsevier: 1993; pp 181-226.

28. Fiedorowicz, M.; Para, A., Structural and molecular properties of dialdehyde starch. Carbohydrate Polymers 2006, 63 (3), 360-366.

29. Du, H.; Miller, J., A molecular dynamics simulation study of water structure and adsorption states at talc surfaces. International Journal of Mineral Processing 2007, 84 (1- 4), 172-184.

30. Atluri, V.; Jin, J.; Shrimali, K.; Dang, L.; Wang, X.; Miller, J. D., The hydrophobic surface state of talc as influenced by aluminum substitution in the tetrahedral layer. Journal of Colloid and Interface Science 2019, 536, 737-748.

31. Wenzel, R. N., Surface roughness and contact angle. The Journal of Physical Chemistry 1949, 53 (9), 1466-1467.

32. Sheludko, A., Thin liquid films. Advances in Colloid and Interface Science 1967, 1 (4), 391-464.

33. Exerowa, D.; Zacharieva, M.; Cohen, R.; Platikanov, D., Dependence of the equilibrium thickness and double layer potential of foam films on the surfactant concentration. Colloid and Polymer Science 1979, 257 (10), 1089-1098.

34. Yoon, R.-H.; Yordan, J., The critical rupture thickness of thin water films on hydrophobic surfaces. Journal of Colloid and Interface Science 1991, 146 (2), 565-572.

35. Derjaguin, B.; Dukhin, S., Theory of flotation of small and medium-size particles. Progress in Surface Science 1993, 43 (1-4), 241-266.

36. Tchoukov, P.; Yang, F.; Xu, Z.; Dabros, T.; Czarnecki, J.; Sjöblom, J., Role of asphaltenes in stabilizing thin liquid emulsion films. Langmuir 2014, 30 (11), 3024-3033.

37. Masliyah, J.; Zhou, Z. J.; Xu, Z.; Czarnecki, J.; Hamza, H. J. T. C. J. o. C. E., Understanding water‐based bitumen extraction from Athabasca oil sands. The Canadian Journal of Chemical Engineering 2004, 82 (4), 628-654. 132

38. Pugh, R., Foaming, foam films, antifoaming and defoaming. Advances in Colloid and Interface Science 1996, 64, 67-142.

39. Schutt, E. G.; Klein, D. H.; Mattrey, R. M.; Riess, J. G., Injectable microbubbles as contrast agents for diagnostic ultrasound imaging: The key role of perfluorochemicals. Angewandte Chemie International Edition 2003, 42 (28), 3218-3235.

40. Lou, S.-T.; Ouyang, Z.-Q.; Zhang, Y.; Li, X.-J.; Hu, J.; Li, M.-Q.; Yang, F.-J., Nanobubbles on solid surface imaged by atomic force microscopy. Journal of Vacuum Science and Technology B: Microelectronics Nanometer Structures Processing, Measurement, Phenomena 2000, 18 (5), 2573-2575.

41. Ishida, N.; Inoue, T.; Miyahara, M.; Higashitani, K., Nano bubbles on a hydrophobic surface in water observed by tapping-mode atomic force microscopy. Langmuir 2000, 16 (16), 6377-6380.

42. Parker, J. L.; Claesson, P. M.; Attard, P., Bubbles, cavities, and the long-ranged attraction between hydrophobic surfaces. The Journal of Physical Chemistry 1994, 98 (34), 8468-8480.

43. Yang, J.; Duan, J.; Fornasiero, D.; Ralston, J., Very small bubble formation at the solid− water interface. The Journal of Physical Chemistry B 2003, 107 (25), 6139-6147.

44. Wildenschild, D.; Sheppard, A. P., X-ray imaging and analysis techniques for quantifying pore-scale structure and processes in subsurface porous medium systems. Advances in Water Resources 2013, 51, 217-246.

45. Marathe, R.; Turner, M. L.; Fogden, A., Pore-scale distribution of crude oil wettability in carbonate rocks. Energy and Fuels 2012, 26 (10), 6268-6281.

46. Pugh, R. J., Foaming, foam films, antifoaming and defoaming. Advances in Colloid and Interface Science 1996, 64, 67-142.

47. Lu, Z.; Lu, Z.; Peng, S.; Zhang, X.; Liu, Q., Microwetting of pH-sensitive surface and anisotropic MoS2 surfaces revealed by femtoliter sessile droplets. Langmuir 2016, 32 (43), 11273-11279.

48. An, H.; Liu, G.; Craig, V. S., Wetting of nanophases: Nanobubbles, nanodroplets and micropancakes on hydrophobic surfaces. Advances in Colloid and Interface Science 2015, 222, 9-17.

49. Rowlinson, J. S.; Swinton, F., Liquids and Liquid Mixtures: Butterworths Monographs in Chemistry. Butterworth-Heinemann: 2013. 133

50. Somasundaram, T.; In het Panhuis, M.; Lynden-Bell, R.; Patterson, C., A simulation study of the kinetics of passage of CO 2 and N 2 through the liquid/vapor interface of water. The Journal of Chemical Physics 1999, 111 (5), 2190-2199.

51. Takahashi, T.; Miyahara, T.; Mochizuki, H., Fundamental study of bubble formation in dissolved air pressure flotation. Journal of Chemical and Engineering of Japan 1979, 12 (4), 275-280.

52. Jin, J.; Dang, L. X.; Miller, J. D. J. P. P. o. M. P., Molecular dynamics simulations study of nano bubble attachment at hydrophobic surfaces. Physicochemical Problems of Mineral Processing 2018, 54 (PNNL-SA-130354).

53. Gao, Y.; Pan, L., Measurement of instability of thin liquid films by synchronized triwavelength reflection interferometry microscope. Langmuir 2018, 34 (47), 14215- 14225.

54. Jin, J.; Miller, J. D.; Dang, L. X., Molecular dynamics simulation and analysis of interfacial water at selected sulfide mineral surfaces under anaerobic conditions. International Journal of Mineral Processing 2014, 128, 55-67.

55. Shi, B.; Dhir, V. K., Molecular dynamics simulation of the contact angle of liquids on solid surfaces. The Journal of Chemical Physics 2009, 130 (3), 034705.

56. Abraham, F. F., The interfacial density profile of a Lennard‐Jones fluid in contact with a (100) Lennard‐Jones wall and its relationship to idealized fluid/wall systems: A Monte Carlo simulation. The Journal of Chemical Physics 1978, 68 (8), 3713-3716.

57. Yu, C.-J.; Richter, A.; Datta, A.; Durbin, M.; Dutta, P., Observation of molecular layering in thin liquid films using X-ray reflectivity. Physical Review Letters 1999, 82 (11), 2326.

58. Fuerstenau, D.; Raghavan, S., Some aspects of the thermodynamics of flotation. Flotation A.M. Gaudin Memorial 1976.

59. Derjaguin, B.; Shcherbakov, L., Effect of surface forces on phase equilibria of polymolecular layers and contact angles. Colloid Journal USSR 1961, 23, 33-43.

60. Pan, L.; Jung, S.; Yoon, R.-H., Effect of hydrophobicity on the stability of the wetting films of water formed on gold surfaces. Journal of Colloid and Interface Science 2011, 361 (1), 321-330.

61. Pan, L.; Yoon, R.-H., Measurement of hydrophobic forces in thin liquid films of water between bubbles and xanthate-treated gold surfaces. Minerals Engineering 2016, 98, 240-250. 134

62. Tabor, R. F.; Wu, C.; Grieser, F.; Dagastine, R. R.; Chan, D. Y. J. T. j. o. p. c. l., Measurement of the hydrophobic force in a soft matter system. The Journal of Physical Chemistry Letters 2013, 4 (22), 3872-3877.

63. Manica, R.; Connor, J. N.; Carnie, S. L.; Horn, R. G.; Chan, D. Y., Dynamics of interactions involving deformable drops: Hydrodynamic dimpling under attractive and repulsive electrical double layer interactions. Langmuir 2007, 23 (2), 626-637.

64. Jiang, L.; Krasowska, M.; Fornasiero, D.; Koh, P.; Ralston, J., Electrostatic attraction between a hydrophilic solid and a bubble. Physical Chemistry Chemical Physics 2010, 12 (43), 14527-14533.

65. Aronson, M.; Princen, H., Aqueous films on silica in the presence of cationic surfactants. Colloid and Polymer Science 1978, 256 (2), 140-149.

66. Alexandrova, L.; Pugh, R.; Tiberg, F.; Grigorov, L., Confirmation of the heterocoagulation theory of flotation. Langmuir 1999, 15 (22), 7464-7471.

67. Tchaliovska, S.; Herder, P.; Pugh, R.; Stenius, P.; Eriksson, J. C., Studies of the contact interaction between an air bubble and a mica surface submerged in dodecylammonium chloride solution. Langmuir 1990, 6 (10), 1535-1543.

68. Yin, X.; Yan, L.; Liu, J.; Xu, Z.; Miller, J. D., Anisotropic surface charging of chlorite surfaces. Clays and Clay Minerals 2013, 61 (2), 152-164.

69. Gupta, V.; Miller, J. D., Surface force measurements at the basal planes of ordered kaolinite particles. Journal of Colloid and Interface Science 2010, 344 (2), 362-371.

70. Yan, L.; Englert, A. H.; Masliyah, J. H.; Xu, Z., Determination of anisotropic surface characteristics of different phyllosilicates by direct force measurements. Langmuir 2011, 27 (21), 12996-13007.

71. McShea, J. A.; Callaghan, I., Electrokinetic potentials at the gas-aqueous interface by spinning cylinder electrophoresis. Colloid and Polymer Science 1983, 261 (9), 757-766.

72. Wang, X.; Yin, X.; Nalaskowski, J.; Du, H.; Miller, J. D., Molecular features of water films created with bubbles at silica surfaces. Surface Innovations 2014, 3 (1), 20-26.

73. Shi, C.; Cui, X.; Xie, L.; Liu, Q.; Chan, D. Y.; Israelachvili, J. N.; Zeng, H., Measuring forces and spatiotemporal evolution of thin water films between an air bubble and solid surfaces of different hydrophobicity. ACS Nano 2014, 9 (1), 95-104.

74. Zhang, X.; Tchoukov, P.; Manica, R.; Wang, L.; Liu, Q.; Xu, Z., Simultaneous measurement of dynamic force and spatial thin film thickness between deformable and solid surfaces by integrated thin liquid film force apparatus. Soft Matter 2016, 12 (44), 9105-9114. 135

75. Ducker, W. A., Contact angle and stability of interfacial nanobubbles. Langmuir 2009, 25 (16), 8907-8910.

76. Kursun, H.; Ulusoy, U., Influence of shape characteristics of talc mineral on the column flotation behavior. International Journal of Mineral Processing 2006, 78 (4), 262- 268.

77. Kho, C.-J.; Sohn, H.-J., Column flotation of talc. International Journal of Mineral Processing 1989, 27 (1-2), 157-167.

78. Yekeler, M.; Ulusoy, U.; Hiçyılmaz, C., Effect of particle shape and roughness of talc mineral ground by different mills on the wettability and floatability. Powder Technology 2004, 140 (1-2), 68-78.

79. Khraisheh, M.; Holland, C.; Creany, C.; Harris, P.; Parolis, L., Effect of molecular weight and concentration on the adsorption of CMC onto talc at different ionic strengths. International Journal of Mineral Processing 2005, 75 (3-4), 197-206.

80. Dellisanti, F.; Valdrè, G., Linear relationship between thermo-dehydroxylation and induced-strain by mechanical processing in vacuum: The case of industrial kaolinite, talc and montmorillonite. International Journal of Mineral Processing 2008, 88 (3-4), 94-99.

81. Burdukova, E.; Becker, M.; Bradshaw, D.; Laskowski, J., Presence of negative charge on the basal planes of New York talc. Journal of Colloid and Interface Science 2007, 315 (1), 337-342.

82. Mierczynska-Vasilev, A.; Beattie, D. A., The effect of impurities and cleavage characteristics on talc hydrophobicity and polymer adsorption. International Journal of Mineral Processing 2013, 118, 34-42.

83. Shrimali, K.; Atluri, V.; Wang, X.; Miller, J. D. J. C.; Physicochemical, S. A.; Aspects, E., Adsorption of corn starch molecules at hydrophobic mineral surfaces. Colloids and Surfaces A: Physhicochemical Engineering Aspects 2018, 546, 194-202.

84. Cygan, R. T.; Liang, J.-J.; Kalinichev, A. G., Molecular models of hydroxide, oxyhydroxide, and clay phases and the development of a general force field. The Journal of Physical Chemistry B 2004, 108 (4), 1255-1266.

85. Krasowska, M.; Sellapperumage, P. M.; Ralston, J.; Beattie, D. A., Influence of dissolved air on bubble attachment to highly oriented pyrolytic graphite. Physicochemical Problems of Mineral Processing 2018, 54 (1), 163-173.

86. Kor, M.; Korczyk, P. M.; Addai-Mensah, J.; Krasowska, M.; Beattie, D. A., Carboxymethylcellulose adsorption on molybdenite: The effect of electrolyte composition on adsorption, bubble–surface collisions, and flotation. Langmuir 2014, 30 (40), 11975- 11984. 136

87. Wu, J.; Delcheva, I.; Ngothai, Y.; Krasowska, M.; Beattie, D. A., Bubble–surface interactions with graphite in the presence of adsorbed carboxymethylcellulose. Soft Matter 2015, 11 (3), 587-599.

88. Atluri, V.; Gao, Y.; Wang, X.; Pan, L.; Miller, J. D., The influence of polysaccharides on film stability and bubble attachment at the talc surface. Mining, Metallurgy and Exploration 2019, 1-10.

89. Rao, S. R., Surface Chemistry of Froth Flotation: Volume 1: Fundamentals. Springer Science & Business Media: 2013.

90. Bulatovic, S. M., Handbook of Flotation Reagents: Chemistry, Theory and Practice: Volume 1: Flotation of Sulfide Ores. Elsevier: 2007.

91. Fuerstenau, D.; Huang, P. In Interfacial phenomena involved in talc flotation and depression, XXII International Mineral Processing Congress, South Africa, 2003; pp 1034- 1043.

92. Laskowski, J.; Liu, Q.; Bolin, N., Polysaccharides in flotation of sulphides. Part I. Adsorption of polysaccharides onto mineral surfaces. International Journal of Mineral Processing 1991, 33 (1-4), 223-234.

93. Liu, G.; Feng, Q.; Ou, L.; Lu, Y.; Zhang, G., Adsorption of polysaccharide onto talc. Minerals Engineering 2006, 19 (2), 147-153.

94. Laskowski, J.; Liu, Q.; O'Connor, C., Current understanding of the mechanism of polysaccharide adsorption at the mineral/aqueous solution interface. International Journal of Mineral Processing 2007, 84 (1-4), 59-68.

95. Mierczynska-Vasilev, A.; Kor, M.; Addai-Mensah, J.; Beattie, D. A., The influence of polymer chemistry on adsorption and flocculation of talc suspensions. Chemical Engineering Journal 2013, 220, 375-382.

96. Morris, G. E.; Fornasiero, D.; Ralston, J., Polymer depressants at the talc–water interface: Adsorption isotherm, microflotation and electrokinetic studies. International Journal of Mineral Processing 2002, 67 (1-4), 211-227.

97. Shortridge, P.; Harris, P.; Bradshaw, D.; Koopal, L., The effect of chemical composition and molecular weight of polysaccharide depressants on the flotation of talc. International Journal of Mineral Processing 2000, 59 (3), 215-224.

98. Kaggwa, G. B.; Huynh, L.; Ralston, J.; Bremmell, K., The influence of polymer structure and morphology on talc wettability. Langmuir 2006, 22 (7), 3221-3227.

137

99. Beattie, D. A.; Huynh, L.; Kaggwa, G. B.; Ralston, J., Influence of adsorbed polysaccharides and polyacrylamides on talc flotation. International Journal of Mineral Processing 2006, 78 (4), 238-249.

100. Rath, R. K.; Subramanian, S.; Laskowski, J., Adsorption of dextrin and guar gum onto talc. A comparative study. Langmuir 1997, 13 (23), 6260-6266.

101. Liu, Q.; Zhang, Y.; Laskowski, J., The adsorption of polysaccharides onto mineral surfaces: An acid/base interaction. International Journal of Mineral Processing 2000, 60 (3-4), 229-245.

102. Jenkins, P.; Ralston, J., The adsorption of a polysaccharide at the talc–aqueous solution interface. Colloids and Surfaces A: Physhiochemical Engineering Aspects 1998, 139 (1), 27-40.

103. Steenberg, E.; Harris, J., A study of the adsorption, zeta potential and dispersion characteristics of certain polymers in relation to their effect on the floatability of talc. South African Journal of Chemistry 1984, 37, 85.

104. Wiese, J.; Harris, P.; Bradshaw, D., The influence of the reagent suite on the flotation of ores from the Merensky reef. Minerals Engineering 2005, 18 (2), 189-198.

105. O'Connor, C. T., Investigations into the recovery of platinum group minerals from the platreef ore of the Bushveld Complex of South Africa. Platinum Metals Review 2013, 57 (4), 302-310.

106. Mackenzie, M.; Malhotra, D.; Riggs, W., Chemical reagents in the mineral processing industry. Society of Mineral Engineering. Colorado, USA 1986, 139-145.

107. Pugh, R., Macromolecular organic depressants in sulphide flotation—a review, 1. Principles, types and applications. International Journal of Mineral Processing 1989, 25 (1-2), 101-130.

108. Healy, T., Principles of adsorption of organics at solid-solution interfaces. Journal of Macromolecular Science - Chemistry 1974, 8 (3), 603-619.

109. Rath, R.; Subramanian, S., Studies on adsorption of guar gum onto biotite mica. Minerals Engineering 1997, 10 (12), 1405-1420.

110. Brossard, S. K.; Du, H.; Miller, J. D., Characteristics of dextrin adsorption by elemental sulfur. Journal of Colloid and Interface Science 2008, 317 (1), 18-25.

111. Beaussart, A.; Mierczynska-Vasilev, A.; Beattie, D. A., Adsorption of dextrin on hydrophobic minerals. Langmuir 2009, 25 (17), 9913-21. 138

112. Beaussart, A.; Parkinson, L.; Mierczynska-Vasilev, A.; Ralston, J.; Beattie, D. A., Effect of adsorbed polymers on bubble--particle attachment. Langmuir 2009, 25 (23), 13290-4.

113. Mierczynska-Vasilev, A.; Ralston, J.; Beattie, D. A., Adsorption of modified dextrins on talc: Effect of surface coverage and hydration water on hydrophobicity reduction. Langmuir 2008, 24 (12), 6121-6127.

114. Knaepen, B.; Kassinos, S.; Carati, D., Magnetohydrodynamic turbulence at moderate magnetic Reynolds number. Journal of Fluid Mechanics 2004, 513, 199-220.

115. Bozzano, G.; Dente, M., Shape and terminal velocity of single bubble motion: A novel approach. Computers and Chemical Engineering 2001, 25 (4-6), 571-576.

116. Pugh, R., Non-ionic polyethylene oxide frothers in graphite flotation. Minerals Engineering 2000, 13 (2), 151-162.

117. Donati, I., Coslovi, A., Gamini, A., Skjåk-Bræk, G., Vetere, A., Campa, C. Paoletti, S., Galactose-substituted alginate 2: Conformational aspects. Biomacromolecules 2004, 5(1), 186-196.

118. Frollini, E., Reed, W.F., Milas, M. Rinaudo, M., Polyelectrolytes from polysaccharides: Selective oxidation of guar gum—a revisited reaction. Carbohydrate Polymers 1995, 27(2), 129-135.