Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Physicochemical characterization of a carbon containing phosphate ore Sablok, Abhay 2019

Warning
You are currently on our download blacklist and unable to view media. You will be unbanned within an hour.
To un-ban yourself please visit the following link and solve the reCAPTCHA, we will then redirect you back here.

Item Metadata

Download

Media
24-ubc_2019_may_sablok_abhay.pdf [ 6.24MB ]
Metadata
JSON: 24-1.0376192.json
JSON-LD: 24-1.0376192-ld.json
RDF/XML (Pretty): 24-1.0376192-rdf.xml
RDF/JSON: 24-1.0376192-rdf.json
Turtle: 24-1.0376192-turtle.txt
N-Triples: 24-1.0376192-rdf-ntriples.txt
Original Record: 24-1.0376192-source.json
Full Text
24-1.0376192-fulltext.txt
Citation
24-1.0376192.ris

Full Text

PHYSICOCHEMICAL CHARACTERIZATION OF A CARBON CONTAINING PHOSPHATE ORE  by  Abhay Sablok  B.Technology., The University of Petroleum and Energy Studies, 2014  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Mining Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   January 2019  © Abhay Sablok, 2019 ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Physicochemical Characterization of a Carbon Containing Phosphate Ore  submitted by Abhay Sablok in partial fulfillment of the requirements for the degree of Master of Applied Science in Mining Engineering  Examining Committee: Dr. Marek Pawlik Supervisor  Dr. Sanja Miskovic Supervisory Committee Member  Dr. Maria Holuszko Supervisory Committee Member  University Examiner  University Examiner  Additional Supervisory Committee Members: Dr. Bern Klein Supervisory Committee Member  Supervisory Committee Member iii Abstract The current research project is aimed at investigating the occurrence of organic matter found in certain types of phosphate ores and the behavior of the ore under mechanical treatment. The presence of organic compounds, often collectively referred to as kerogen, complicates the froth flotation of such ores, and further advances in the processing of high-organics phosphate ores require a thorough understanding of the nature, occurrence, and distribution of organic compounds within phosphate ores. Among the most advanced methods, scanning electron microscopy (SEM) to generate maps of mineral components within grains of various sizes, and micro-Fourier transform infrared spectroscopy/microscopy (FTIR) to visualize the presence and distribution of various organic compounds on mineral particles were employed. These techniques provided information about associations of organics with different ore components. Finally, different mechanical treatment methods for pre-conditioning were tested in order to investigate the release of organics and other mineral phases as a possible alternative to indiscriminate crushing and grinding of the ore, and to selectively release clean phosphates of a higher grade.   iv Lay Summary Amongst its various uses, phosphates are a major raw product for the fertilizer industry. They constitute an important economic resource that has been mined for over a hundred years. The current work focuses on the processing problems of one such phosphate deposit in Idaho in the United States. The deposits have shown the presence of organic matter which is not well defined in terms of its nature and composition and inhibits optimum flotation of the mineral.  Characterization techniques are used to look at the ore both structurally and compositionally on the micro scale, and the information obtained is used to devise methods to maximize the recovery during processing. v Preface The characterization of the ore on a micro scale was thought of as an important link to gather information for devising an efficient processing flow-sheet. The selection of appropriate techniques and the testing procedures were developed with constant and indispensable guidance from Dr. Marek Pawlik and performed by the author, Abhay Sablok. All the experiments were performed by the author, either in the labs of Norman B. Keevil Institute of Mining Engineering or at facilities belonging to other departments at UBC. Chemical assays of samples obtained through lab testing were carried out by a commercial lab in the Vancouver area.  Chapters 1 and 2 provide a background to the research and provide an introduction leading in to the various experiments and scans conducted in the succeeding chapters – 3, 4 and 5. The appendix contain results from some supporting experiments that paved way for the research to move forward. vi Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables .............................................................................................................................. viii List of Figures ............................................................................................................................... ix List of Abbreviations ................................................................................................................... xi Acknowledgements ..................................................................................................................... xii Dedication ................................................................................................................................... xiii Chapter 1: Introduction ................................................................................................................1 1.1 Research Objectives ........................................................................................................ 2 1.1.1 Specific Objectives ................................................................................................. 3 Chapter 2: Literature Review .......................................................................................................5 Chapter 3: Material Information ...............................................................................................12 Chapter 4: Scanning Electron Microscopy ................................................................................14 4.1 Overview of EDX ......................................................................................................... 15 4.2 Sample Preparation ....................................................................................................... 16 4.3 Instrument Used ............................................................................................................ 17 4.4 Mineralogical Analysis using Image ............................................................................ 18 Chapter 5: Attrition Scrubbing ..................................................................................................30 5.1 Observations ................................................................................................................. 32 5.2 Attrition Scrubbing Tests .............................................................................................. 35 5.3 Mass Balance from Assays ........................................................................................... 39 vii 6.3 Scrubbing followed by crushing ......................................................................................... 49 Chapter 6: Results and Discussion .............................................................................................53 Chapter 7: Conclusion .................................................................................................................65 Chapter 8: Recommendations for Further Research ...............................................................67 Bibliography .................................................................................................................................68 Appendices ....................................................................................................................................70 Appendix A Infra-Red (FTIR) Spectroscopy ........................................................................... 70 A.1 Data acquisition and interpretation ........................................................................... 71 Appendix B Additional FT-IR Maps ........................................................................................ 75 Appendix C Single particle and batch tests .............................................................................. 77 Appendix D Particle Size Distributions of the Attrition Scrubbing Products .......................... 92 Appendix E Assay Result for the various Attrition Scrubbing Products .................................. 95 E.1 Assay Results for the Attrition followed by crushing tests: ..................................... 96 Appendix F XRD Report for the feed as received from the lab ............................................... 97  viii List of Tables Table 3.1 XRD analysis of the feed .............................................................................................. 12 Table 3.2 PSD of the feed ............................................................................................................. 13 Table 4.1 Element composition from spot analysis for sample#1 ................................................ 20 Table 4.2 Element composition data from spot analysis .............................................................. 22 Table 4.3 Semi-quantitative analysis using ImageJ ...................................................................... 26 Table 4.4 Semi-quantitative analysis using ImageJ ...................................................................... 28 Table 4.5 Summary of semi-quantitative analysis using ImageJ .................................................. 29 Table 5.1 Single particle tests in different solvents ...................................................................... 31 Table 5.2 Single particle tests in acidic media .............................................................................. 32 Table 5.3 Batch test in acidic media ............................................................................................. 33 Table 5.4 Testing conditions for attrition scrubbing ..................................................................... 36 Table 5.5 Assay results uncrushed feed ........................................................................................ 42 Table 5.6 Assay results – pH2.5, scrubbed for 20minutes ............................................................ 42 Table 5.7 Assay results – pH3.5, scrubbed for 20minutes ............................................................ 43 Table 5.8 Assay results – pH10.5, scrubbed for 20minutes .......................................................... 44 Table 5.9 Assay results – pH2.5, scrubbed for 10minutes ............................................................ 44 Table 5.10 Assay results – natural pH (5.98), scrubbed for 20minutes ........................................ 45 Table 5.11 Assay results – natural pH (6.06), scrubbed for 10minutes ........................................ 45 Table 5.12 Assay results for the re-scrubbed ore at natural pH for 20minutes ............................ 49 Table 6.1 Conditions for attrition scrubbing tests ......................................................................... 53  ix List of Figures Figure 4.1 Backscattered image with spot analysis for sample#1 ................................................ 19 Figure 4.2 Backscattered image with spot analysis for sample#2 ................................................ 21 Figure 4.3 Element maps for Phosphorous, Carbon, Aluminum and Silicon for the sample#1 ... 23 Figure 4.4 Backscattered image and element map for Phosphorous for sample#3 ...................... 24 Figure 4.5 Element maps for Phosphorous, Carbon, Silicon and Aluminum for sample#3 ......... 25 Figure 4.6 Black and White images for Silicon, Aluminum and the overlap between the two for sample#3 ....................................................................................................................................... 27 Figure 5.1 Denver D-12 Flotation Machine .................................................................................. 36 Figure 5.2 Particle Size Distribution for the attrition products in acidic media ........................... 37 Figure 5.3 Particle Size Distribution for the attrition products – longer attrition time ................. 38 Figure 5.4 Product size comparison for attrition scrubbing products ........................................... 39 Figure 5.5 P2O5 content variation with scrubbing time and conditions for (-0.600mm+0.075mm) size range ...................................................................................................................................... 46 Figure 5.6 Organic carbon vs P2O5 content variation for product obtained at natural pH and scrubbed for 20minutes ................................................................................................................. 46 Figure 5.7 Comparison of the product of the re-scrubbed ore with the original scrubbing tests .. 48 Figure 5.8 Attrition scrubbing followed by crushing – test#1 ...................................................... 50 Figure 5.9 Attrition scrubbing followed by crushing – test#2 ...................................................... 51 Figure 6.1 Main components of the +2.4mm fraction of the attrition products ........................... 54 Figure 6.2 Main components of the -2.4 + 0.600mm fraction of the attrition products ............... 55 Figure 6.3 Main components of the -0.600 + 0.075mm fraction of the attrition products ........... 56 Figure 6.4 Main components of the -0.075mm fraction of the attrition products ........................ 57 Figure 6.5 Scatterplot for silica vs phosphate grade variation ...................................................... 58 x Figure 6.6 Scatterplot for silica vs alumina grade variation ......................................................... 59 Figure 6.7 Scatterplot for phosphate vs alumina grade variation ................................................. 60 Figure 6.8 Scatterplot for carbon vs phosphate grade variation ................................................... 60 Figure 6.9 BSE image for the (-0.600 + 0.075mm) scrubbing product (above), and overlay of BSE, Al, Si, P (below) .................................................................................................................. 61 Figure 6.10 Overlay of SE and P for the -0.075mm scrubbing product (above), and overlay of SE, Al, Si(below) .......................................................................................................................... 63  xi List of Abbreviations CT: Computed Tomography FTIR: Fourier Transform Infra-Red (Microscopy/Spectroscopy) IR: Infra-Red (Microscopy/Spectroscopy) keV: kiloelectron volts MATLAB: Matrix Laboratory NTU: Nephelometric Turbidity Units OM: Organic Matter SEM: Scanning Electron Microscope/Microscopy  xii Acknowledgements I wish to express my deepest gratitude to Dr. Marek Pawlik for his constant supervision and willingness to help in all aspects of the research work. I would also like to thank Mr. Richard Pilon from Albatross Environmental & Process Consulting Inc. and MITACS for sponsoring the project.  I would also like to acknowledge Ryan MacIver who gave me invaluable background information and ideas based off his work on the same material conducted during a previous research internship.  I am also deeply indebted to Sally Finora, Jacob Kabel and other members of Dr. Pawlik’s research group for their willingness to help with the experimental and data collection part of this research. I am thankful to Carmen Jensen to have played a major part in helping me transition to a research degree. Lastly, I am extremely thankful to my family and friends who have consistently been a source of support in all possible forms. xiii Dedication I would like to dedicate my thesis to my beloved mom. 1 Chapter 1: Introduction The current work focuses on the phosphate deposits in the Idaho region of the USA. The mineralogy of the region varies widely, and each deposit creates a unique processing challenge. A low-grade ore from the area is studied in this thesis for its physical and chemical characteristics and to better understand the effect of mineralogy on processing options. Limited plant experience from the processing of this type of ore raises significant questions about the nature of the organic matter, the association of the organics with different minerals, and about the behavior of the organic phase under different treatment conditions, e.g., during mechanical attrition-scrubbing, conditioning at high and low pH values, varying temperature, or dissolution directly into process water. The organic compounds, which are often collectively referred to as kerogen in soil science, are very poorly characterized in terms of their source and behavior from a mineral processing viewpoint. As a result, the main motivation behind this research is to look at individual ore particles in their native state and employ microscopic techniques to understand the characteristics of these particles. Amongst advanced methods, the research employed methods scanning electron microscopy (SEM) to generate mineral maps of polished surfaces of the particles, and micro-Fourier transform infrared spectroscopy (FT-IR) to attain additional information and possibly, visualize the presence and distribution of various high carbon regions on the mineral surface. Using this information about the particles on the micro scale, a pre-conditioning step was proposed which was found to help in increasing the recovery of the apatite component from this ore. Conventionally, kerogen has been defined as insoluble, macromolecular organic matter (OM) dispersed in sedimentary rocks and is by far the most abundant form of OM on earth (Vandenbroucke & Largeau, 2007). 2 Kerogen generates petroleum and natural gas. Kerogen and petroleum are known to have complimentary characteristics (Vandenbroucke & Largeau, 2007). Petroleum (more generally bitumen) is soluble in usual organic solvents whereas kerogen is the sedimentary OM insoluble in those solvents. The definition has more recently been extended to all insoluble OM, not only those dispersed in sedimentary rocks but also ‘pure’ organic deposits such as humic and algal coals and various aliphatic substances as well as insoluble OM in recent sediments and even soils (Vandenbroucke & Largeau, 2007). Kerogen is a very stable substance. Elemental analysis shows that it is composed of C, H, N, O, S and possibly Fe from pyrite. Although there has been a significant  volume of research carried out on kerogen in recent years, there still remain certain important aspects that need to be researched upon and many significant advances expected in the near future (Keil & Mayer, 2013). The presence of kerogen has a wide scale implication in more than one industry and the understanding of its source, chemical composition and behavior becomes all the more important. This work focuses on the occurrence of the “high carbon” components in the ore, and on documenting the behavior of this material under attrition scrubbing. These apatite bearing ores are an important source of phosphorous, the key component in phosphorous-based fertilizers. 1.1 Research Objectives Previous work done with regards to flotation of these phosphorous bearing particles (apatite) has led to a stage where further progress in increasing recovery of phosphate depends upon the characterization of the organic matter, which is associated with these particles. This project was carried out to focus solely on the organic matter presence within these minerals from both a quantitative and qualitative viewpoint. As of now, the organic matter is believed to be a ‘humic’ substance with 5-6% w/w organic matter based on findings from the previous work. Its behavior 3 with different components in the sample, depending on the physical and chemical conditions, needed a more detailed investigation. The knowledge of the composition of the raw ore itself with regards to the organic matter contained in them, could lead to a better conditioning treatment. Thus, there was a definite potential for research about where this organic matter would be present both on a macro and a micro scale. That information, once obtained, can be used to determine the occurrence and associations of the organic matter in phosphate ores, and to document the behavior of this material under mechanical treatment.  1.1.1 Specific Objectives • To determine the occurrence of organic matter within phosphate ores as a function of particle size. The main research questions behind this objective were: where the organic matter actually resides – if it is primarily present on individual particle surfaces, or whether it functions more like a binder (glue) of different mineral particles and also how this occurrence would change as a result of chemical (e.g., high and low pH effects) and mechanical (e.g., attrition scrubbing) factors. • To determine the association and distribution of organics and various ore components (apatite, quartz, clay minerals). From the froth flotation point of view, it is essential to assess whether the organic phase is preferentially associated with the valuable phosphate minerals or with the gangue minerals (clay, quartz etc.). This type of information will suggest which mineral phase should be targeted in the flotation process, whether separation by flotation should proceed as a forward (phosphate) or a reverse (gangue) process. • To assess the behavior of the organic phase under different ore treatment conditions, e.g., during mechanical attrition and conditioning at various pH values. Since the organic phase 4 seems to be responsible for phosphate flotation problems, it is critical to understand whether a treatment process could be developed to completely remove/separate the organic phase from the mineral constituents ahead of the flotation stage. Such a process can only be successful if the release of organics from the ore under different conditions is better understood.  5 Chapter 2: Literature Review The literature about characterizing organic coatings on grains of phosphate ores that was found in (McClellan & Despujol, 1999), defines and distinguishes between ‘humate’, ‘humic’ and ‘humus’ as terms used for the various types of organic matter. Humus is defined as the accumulation of plant residues in an early stage of microbial and bacterial decay and is found at or within a few inches off the land surface. Soil humus is slow to decompose (degrade) under natural soil conditions. It can persist in the soil for several hundred years. Humus forms the major component of soil organic matter forming up to 75% of the total value. Humic substances are high molecular weight compounds that together form what is referred to as humus. They function to give the soil its structure, porosity and water holding capacity. Humic material consists of a number of fractions, such as fulvic, humic, and many other acids and salts described in the soil literature. Humates are metal (mineral) salts of humic or fulvic acids. Humate is the collective term used for the group of gel-like substances, carried in water suspension or solution that may precipitate or flocculate when exposed to various physical-chemical environments. Humate precipitates may form discrete grains in voids or occur as coatings on grains in the sediment.  The three categories of humic substances as stated above are arbitrarily based on the solubility of each fraction in water adjusted to different conditions (pH levels). They are summarized as below: 1. Fulvic acids: Mixture of weak aliphatic and aromatic organic acids that are soluble in water at all pH conditions (acidic, neutral and alkaline). 2. Humic acids: Mixture of weak aliphatic and aromatic organic acids which are not soluble in acidic conditions but are soluble in alkaline conditions. 3. Humin: These are a fraction of humic substances which are not soluble in water under any conditions and the humins present in the soil are most resistant to decomposition. 6 In immature coals, humin accounts for the major part of the humic compounds, in sharp contrast to the OM from soils, where humic and fulvic acids often dominate (Vandenbroucke & Largeau, 2007).  The very limited amount of literature dealing with organic matter in phosphate ores focuses more on coatings rather than on intrinsically found organic matter. However, the available works still partly relate to the present research. The presence of organics as coatings is well known in the Florida and Georgia regions where after initial wet milling, the phosphate ores are causticized with NaOH to remove the adhering organic compounds before stockpiling and dry milling. The hypothesis is that even though these coatings might be significantly removed by conditioning at high pH for a long enough time, they may interfere with the surface sorption of flotation reagents later, resulting in decreased selectivity of the froth flotation process (McClellan & Despujol, 1999).  In the Idaho region, which is where the tested ore originated from; the phosphate formation occurs in deeply buried formations. The deposits are contained in the marine sedimentary rocks that were deposited about 265 million years ago on the western margin of North America. These rocks consist primarily of organic carbon and phosphate rich mudstone, siltstone, phosphorite, carbonate, shale and chert deposited over a period of about 10 million years (U. S. Geological Survey, 2002).  A large area (760km2) in the Phosphoria formation in the Soda Springs area of south-eastern Idaho is estimated to be underlain by about 22 billion tons of phosphate rock that contains about 2.7 billion tons of phosphorus. About 44% of these resources is shallower than 600m, and the remainder is no deeper than about 2000m. (Gulbrandsen & Krier, 1980). Gulbrandson and Krier (1980), also analyzed the various components of the formation and classified it into three main categories based on the dominant mineral phase, namely: 1. apatite – the formation containing the apatite-based ore 7 2. OM-carb – a combination of the organic matter (OM) and the carbonate minerals (carb), such as calcite and dolomite; 3. Q-S – the remainder of the rock, chiefly quartz, potassium feldspar, buddingtonite, albite, muscovite-illite, kaolinite, montmorillonite, pyrite, and possible goethite. The phosphate minerals are found in igneous, metamorphic and sedimentary deposits around the world (P. Zhang, Khalek, El-Shall, & El-Mofty, 2003). A large proportion out of the world’s deposits are sedimentary deposits containing a significant amount of carbonate minerals (Amirech, Bouhenguel, & Kouachi, 2018). Phosphate ores are mainly associated with gangue minerals such as clays, silica, calcareous minerals (mainly calcite and dolomite), carbonaceous matter, and iron oxides (Amirech et al., 2018; P. Zhang et al., 2003). The beneficiation of low-grade sedimentary phosphate ores containing calcareous impurities, such as calcite and/or dolomite in addition to the siliceous gangues, is required to meet the increasing demand for phosphate (P. Zhang et al., 2003). This poses a complex problem as the carbonates and the phosphates have similar physicochemical properties relevant to froth flotation (City, 2000) and the valuable apatite component has similar bulk/physical properties like density, particle size, particle shape etc. to those of the associated gangue minerals (Amirech et al., 2018).  Various beneficiation schemes have been established for improving phosphate grade including scrubbing/washing and size classification, gravity separation, magnetic separation, and flotation (Amirech et al., 2018). The processing method is largely chosen based on the type of gangue associated with the phosphate. Flotation has been successfully applied to both the sedimentary and igneous phosphates with siliceous gangues while calcination was the only practiced technique for upgrading sedimentary phosphates with carbonate gangue (P. Zhang et al., 2003). The calcination process, however has drawbacks, such as high energy consumption and lower reactivity of the 8 products and hence is not very commonly used. Therefore it is imperative to find an alternative technique to upgrade sedimentary phosphates (P. Zhang et al., 2003). Both direct and reverse flotation has been successfully applied for phosphate beneficiation. In order to be considered successful; any beneficiation method must be able to increase the P2O5 content of the concentrate above 30% (Amirech et al., 2018).  Sedimentary apatites display large variation in physical forms and types of associated gangue owing to the their widely differing chemical compositions (P. Zhang et al., 2003). Sedimentary phosphates also show a wide variety of textures that reflect their complex geologic origins and histories (P. Zhang et al., 2003). The sedimentary phosphate ores fall into two main classes: consolidated and unconsolidated which can be further subdivided based on the cemented material for the consolidated ores and the size and/or structure of the apatite aggregates for the unconsolidated ores (P. Zhang et al., 2003). The sedimentary-carbonaceous ores are beneficiated using the flotation process which poses the problem that the physico-chemical properties of the phosphatic minerals and the carbonates are very similar (City, 2000). Most of the plants divide the beneficiation process into three main stages: washing, sizing, and flotation. Both direct and reverse flotation are successfully employed for the separation of carbonaceous gangue from sedimentary phosphate ores: 1. Direct flotation of phosphate with the depression of the carbonate gangue 2. Reverse flotation of the carbonate gangue with depression of the phosphates The Crago ‘double float’ process is a very commonly used flotation technique for the beneficiation of phosphates. In this process, the feed is dewatered and conditioned at about 70% or higher solids with fatty acid/fuel oil at a pH of about 9 for three minutes. The sized feed is first subjected to rougher flotation (P Zhang, Yu, & Bogan, 1997) . A significant amount (30-40%) of silica is also 9 floated in the rougher flotation step (P. Zhang et al., 2003; P Zhang et al., 1997). The rougher concentrate goes through a dewatering cyclone, an acid scrubber (pH 3-3.5) and a wash box to remove the reagents from the phosphate surfaces. After rinsing, the feed is transported into flotation cells where amine (often with diesel oil) is added and the pH is re-adjusted using sodium hydroxide. In this reverse flotation stage, the silica is floated from phosphate at neutral pH (P. Zhang et al., 2003; P Zhang et al., 1997). Thus, in this conventional process, about 30% to 40% (by weight) of the sands in the feed are floated twice, first by fatty acid and then by amine (P. Zhang et al., 2003). This process is, therefore, inefficient in terms of the utilization of collector. One of the major drawbacks of this process is the de-oiling process which consumes a significant amount of sulfuric acid, which in turn adds to the complexity of the process in terms of safety cautions and equipment maintenance (P. Zhang et al., 2003). The amine flotation step is also an issue since not only are amines more expensive than fatty acids, but they are also very sensitive to water quality and slimes content of the water (P. Zhang et al., 2003; P Zhang et al., 1997). Having been a relatively unchallenged process since the 1950s owing to the relative differences in cost of chemicals, availability and quality of the water used in processing, and the de-sliming technology in use; there are now variations to the standard Crago process. More recently, a variation called the ‘Reverse Crago’ process, has been employed. In this process, the phosphate is floated in the second step instead of the silica (P Zhang et al., 1997). It is worth noting that the Crago and the ‘Reverse Crago’ processes are not the only available processing options. In the processing of high-iron phosphate ores, a common occurrence as seen in the Kapuskasing deposit in Ontario, Canada as well as the Mt Weld deposit in Western Australia; the depression of iron minerals is usually achieved using starch and only a single rougher stage is generally used in the processing circuit (Nanthakumar, Grimm, & Pawlik, 2009; Wen Qi, Parentich, Little, & Warren, 1992). In contrast to the Crago process, these circuits do not include a reverse flotation of silica (Nanthakumar et al., 10 2009).  Fatty acid flotation of the Mt Weld high-iron deposit from Western Australia (38% Fe2O3, 22% P2O5) in the presence of sodium silicates of varying moduli (SiO2:Na2O ratio) as iron depressants (Nanthakumar et al., 2009; Wen Qi et al., 1992). The main iron minerals were hematite and goethite whose depression with starch did not give satisfactory results and the contamination of the concentrate by the iron minerals was found to be due to the ineffective depression of the well liberated iron oxides as well as due to the good flotation of the composite iron oxide-apatite particles (Nanthakumar et al., 2009; Wen Qi et al., 1992). For the Kapuskasing plant, sodium silicates were tried but without any significant benefit and a possible reason for the same was the high-calcium process water consuming the sodium silicates by precipitating the water-insoluble calcium silicates (Nanthakumar et al., 2009).  At present, there is no successful method developed specifically for phosphate ores high in carbon, as can be concluded from the available literature. The presence of organic carbon, or more loosely any ore which is high in carbon, is not well documented, and only anecdotal knowledge that the presence of this poorly defined phase has a negative effect on the flotation of such phosphate ores exists at this point. On the processing front, some work had been done on waste phosphate rock in Tunisia. The phosphate occurs in sedimentary rocks in that region as well. The waste rock generated has a relatively high P2O5 content (~12% P2O5) which is similar in grade to the ore for the current project although there was no mention of the presence of carbon in the Tunisian ore. The treated phosphate occurs with carbonate gangue, and hence direct flotation and thermal treatment are less attractive processes to separate the phosphates from carbonate gangue (Gallala et al., 2016). The aforementioned work also uses scrubbing and wet screening as a step to clean the phosphates and 11 remove the clays as much as possible. An increase in the P2O5 content was reported, mainly in the -315 + 40micron range, up from 16.24% to 21.24%, which also represented about 20% of the total weight. Attrition scrubbing was seen to significantly reduce the size of the particles and increase significantly the liberation of the fine particles from the coated coarse grains (Gallala et al., 2016).  A different study also found that lab scale attrition tests were found to increase the P2O5 recovery by 2.49% on a similar phosphate ore from Tunisia (Zidi, Babbou-Abdelmalek, Chaabani, & Abbassi, 2016). The attrition step was performed after a milling step and the product was used as a feed for the flotation stage. Thus, such an approach was attempted before with some significant promise, however, the effect of attrition alone on the size distribution and the chemical composition of the feed to the flotation was not elucidated clearly in these works. 12 Chapter 3: Material Information About 8 kilograms of the sample was provided by Albatross Environmental and Process Consulting Inc., and it came from a producing phosphate operation near Soda Springs, Idaho. The material was considered reject by the mining operation as it could not be currently processed. X-ray diffraction (XRD) analysis was performed to get the information about the key minerals in the sample. The tests were planned and performed with the knowledge of the quantity of material available and the nature of the information that was sought. The particle size distribution for the sample is also tabulated below along with the XRD analysis data. Mineral Ideal Formula Percentage Composition Fluorapatite Ca5(PO4)3F 54.3 Quartz SiO2 20.4 K-feldspar (Microcline, Orthoclase) KAlSi3O8 9.9 Illite-Muscovite 2M1 K0.65Al2.0Al0.65Si3.35O10(OH)2- KAl2AlSi3O10(OH)2 6.1 Dolomite CaMg(CO3)2 3.9 Plagioclase (Albite) NaAlSi3O8 – CaAl2Si2O8 2.9 Calcite CaCO3 1.1 Gypsum CaSO4·2H2O 0.8 Pyrite FeS2 0.6 Total  100.0 Table 3.1 XRD analysis of the feed The table above shows that the ore contains over 50% of fluorapatite, with quartz and aluminosilicates forming the main gangue minerals (a total of approximately 36%). There are also small amounts of carbonates (dolomite and calcite), on the order of 5% which are also noteworthy. 13 Because of the very limited amount of ore available for testing, larger-scale attrition scrubbing tests required careful planning so that sample wastage is minimized. The overall sample was homogenized and split into charges ranging from 400-600gm for each scrubbing test. More details about sample preparation are included within the respective sections.  Size Fractions (mm) Nominal Size (mm) Weight [%] Cumulative Retained [%] Cumulative Passing [%] +9.51 9.51 3.1 3.1 96.9 -9.51+6.7 6.70 11.1 14.2 85.8 -6.7+4.75 4.75 19.8 34.0 66.0 -4.75+3.35 3.35 13.2 47.3 52.7 -3.35+2.36 2.36 10.5 57.7 42.3 -2.36+1.70 1.70 7.1 64.8 35.2 -1.70+0.850 0.85 10.4 75.2 24.8 -0.850+0.600 0.60 4.1 79.3 20.7 -0.600+0.300 0.30 6.8 86.1 13.9 -0.300 0.30 13.9 100.0 - Total  100.0   Table 3.2 PSD of the feed The coarse nature of the sample should be noted, with some particles as coarse as 13mm (about half an inch) present in the sample. Practically all experiments in this thesis were carried out using this material as feed. Experiments on single particles were performed on hand-picked coarse grains from within the feed.   14 Chapter 4: Scanning Electron Microscopy The SEM operates by scanning an energetic, finely focused electron beam over an individual feature or a field of features. The primary electron beam interacts with the specimen producing a variety of secondary signals that can be monitored with appropriate detectors. These signals can be collected in synchronization with the position of the scanned electron beam to generate high-resolution images providing detailed spatial and composition information. Thus particle sizes, morphologies are well represented with this technique (Willis et al., 2002).  When the primary beam interacts with the specimen, the electrons undergo two types of collisions:  1. Elastic collisions where the energy of the incident electrons is unchanged, and; 2. Inelastic collisions in which the primary electrons lose energy in a succession of collisions, leaving the electrons with lower energy.  The elastic collisions give rise to the backscattered electron (BSE) signal, which forms the BSE image. Inelastic collisions give rise to the secondary electron (SE) signal, which forms the SE image. The SE and the BSE images provide different but complementary information. Secondary electrons are emitted from the atoms occupying the top surface and produce a readily interpretable image of the surface. The SE image possesses three-dimensional perspective, high depth of field, and the appearance of overhead illumination. A high-resolution image can be obtained because of the small diameter of the primary electron beam. Backscattered electrons are primary beam electrons that are “reflected” from atoms in the solid. The contrast in the BSE image is determined largely, though not exclusively, by the atomic number of the elements in the sample. The image can therefore show the distribution of different chemical phases in the sample. Because backscattered electrons are emitted from a depth in the sample, the resolution in the image is not as good as for secondary electrons (Willis et al., 2002).  15 Hence, the SE image is superior for displaying surface detail and particle morphology but does not show chemical heterogeneity. Changes in brightness within the BSE image correspond to changes in effective atomic number (Willis et al., 2002). Image magnification is defined as the linear dimension of the scan on the output device (monitor or printer) divided by the linear dimension of the scan on the sample. Therefore, magnification is not the best way to define and clearly convey the size of the features in the image. For that reason, a scale bar is generally included in the image to calibrate feature size (Willis et al., 2002).  4.1 Overview of EDX Interaction of the primary electron beam with atoms in the sample causes inner electron shell transitions, which result in the emission of X-rays. Two types of X-rays are generated: a) Bremsstrahlung or continuous X-rays, which generate a broad and slowly-varying background over the entire X-ray spectrum, and b) characteristic X-rays, which are narrow, discrete peaks in the spectrum whose energies are characteristic of specific elements present in the sample. A fraction of the X-rays emitted by the specimen are collected and analyzed by means of an EDX and/or WDX (Wavelength Dispersive X-Ray) analyzer. EDX analysis can be performed at two levels of sophistication: quantitative EDX or qualitative EDX. Under optimal conditions (appropriate samples, relevant standards, controlled experimental setup, and sophisticated data-reduction procedures), EDX can yield elemental compositions of flat, polished specimens with accuracies and precisions approaching 1% (Willis et al., 2002). Quantitative analysis is much more complicated than analyzing flat, polished samples. The difficulties largely result from the need to correct X-ray yields from particles with irregular surfaces and thicknesses less than the incident electron range. For particles that are larger than a 16 few micrometers, geometry-dependent absorption effects become severe, while for smaller particles, corrections for atomic number and thickness effects become critical (Willis et al., 2002). Although not a quantitative technique by any stretch, SEM can characterize a sample in terms of number percent by particle type or chemical class by applying image logic. 4.2 Sample Preparation Proper sample preparation is a prerequisite for successful SEM analysis. The method of sample preparation is governed by the nature of the sample and by the analytical objectives. Suitable samples include most solids that are stable under vacuum and exposure to an energetic electron beam (metals, ceramics, polymers and minerals) (Willis et al., 2002).  Usually the sample is set in epoxy and then polished to reveal a flattened and smooth surface. The sample is also coated with a thin conductive film to minimize sample charging problems. Carbon is the most common coating material because it is cheap and almost invisible to most x-rays. Gold or gold-palladium is preferred in cases where the sample has a very irregular surface. Both coatings are applied at a thickness of about 20nm.  In the present scenario, since one of the major considerations was to look for the presence of carbon in the native state, both setting in epoxy and coating with a conductive layer were avoided. The raw sample was polished to reveal a flat surface and was set on stage for the SEM analysis without any external coatings. Therefore, a technique called as environmental scanning electron microscope (ESEM) was used to scan the particles. The technique works in a way such that nonconductive samples do not necessarily have to be made conductive for analysis, thereby preserving their original characteristics. 17 As a first step, the particles were classified based on their varied appearance, and a sorting mechanism was developed for the same. The population of the particles was seen to have two broad categories: a ‘light’ phase and a ‘dark’ phase. The working assumption was that the dark phase was the one with the higher organic content, while the lighter phase had a lesser amount of naturally occurring organic matter. To avoid the subjectivity in this sorting mechanism, a MATLAB program based on image logic which would acquire images of particles in real time and sort them based on an intensity histogram was developed. The intensity thresholds were set based on visual observations to distinguish the ‘light’ and the ‘dark’ particles. The SEM scans were conducted bearing in mind this classification and the difference in compositional information that the scans could generate. 4.3 Instrument Used The SEM images were acquired at the Materials Engineering facility at the University of British Columbia using the FEI Quanta 650 microscope. The FEI Quanta 650 is an environmental scanning electron microscope. Utilizing a tetrode tungsten filament electron gun, the system is capable of accelerating voltages of up to 30kV. In high vacuum mode, the system is capable of imaging at a resolution of 3nm (8nm at 3kV). In ESEM mode, the system is capable of imaging at a resolution of 4nm using the environmental secondary electron detector. Environmental capability (ESEM) allows for the imaging of insulating samples that would otherwise require coating, as well as wet samples under very specific circumstances. In ESEM mode, only the backscattered and environmental secondary electron detectors are available. Also, present is a silicon drift energy dispersive x-ray (SDD-EDS/EDX) detector. This feature allows semi-quantitative and quantitative elemental compositional analysis with standards to be carried out. 18 For an overall idea of the elemental composition of the sample, spot analysis on all the visibly different areas was also conducted (grains, inter and intra-granular boundaries, cementing material between grains etc.). Individual back-scattered maps were then obtained for the various elements of interest.  These maps showed the variation in the occurrence of the different elements and thereby of the constituent mineral phases. 4.4 Mineralogical Analysis using Image To introduce an element of quantitative analysis to the SEM image analysis, a software called ImageJ was used to overlay single element maps and determine how they overlap with other elements that constitute a given mineral. For example, to verify the existence and distribution of clays, element maps of Al, K and Si were overlaid. This method showed systematic trends about the location and extent of different minerals present over the sample surface. Some of the examples for the same are illustrated in the form of images below. The results were tabulated to show recurring trends between several samples. 19  Figure 4.1 Backscattered image with spot analysis for sample#1 The image shows a sample back-scattered SEM image with selected spots for elemental analysis. The percentage composition for these points is tabulated below. Only the major elements that were detected with spot analysis are tabulated and hence the composition does not add up to 100% for all the tested points.    20   Elemental Composition (%) for sample#1 POINT ID C O F Al Si P S K Ca Sum  1 5.25 41.29 3.57 0.15 1.61 11.57 0.75 - 34.17 93.99 2 4.85 53.52 - 0.35 33.50 1.01 - - 5.73 102.87 3 5.80 49.83 - 0.36 34.76 1.28 - - 6.70 96.65 4 10.14 48.49 0.60 0.68 6.35 2.23 0.42 0.21 24.61 113.8 5 8.65 51.16 - 0.20 1.94 0.31 2.22 - 26.16 99.06 6 4.12 46.76 - 7.81 23.83 0.78 - 9.44 5.75 97.66 Table 4.1 Element composition from spot analysis for sample#1 Spot analysis is an effective way to quantify the elements present in the SEM image and to identify minerals in a small area of the image. It does give a fair representation of the presence or absence of a particular element, but the values cannot be taken as an absolute composition of a given spot.  The composition of fluorapatite, which is the phosphate source in the ore as confirmed by the XRD analysis; would suggest that there would be spots on the sample surface that would confirm the composition with the presence of major constituent elements- P, Ca, O and F. Point-1 confirms the existence of all of these elements. Another interesting observation from the data-set is the presence of carbon and possible nature of its existence. A thorough examination of individual spots with analysis of all elements present could yield information about the carbon being of the organic or inorganic nature at these spots. As an example, point-4 shows a relatively high composition of carbon (10.14%) but it is accompanied by a higher composition of calcium and oxygen as well pointing to the presence of a calcium-carbonate (potentially calcite) rather than an organic phase. Similarly, a closer look at the concentrations of Si and Al simultaneously gives an indication of the presence of quartz or alumino-silicates and the fact that they occur in the matrix and not in 21 the spots that occur on the grains themselves.  Figure 4.2 Backscattered image with spot analysis for sample#2 The image above is another example of spot analysis done on a different sample surface. The composition of the various spots shown is tabulated below.   22 Table 4.2 Element composition data from spot analysis Spots 8, 9 and 10; on the brightest areas of the image - show the presence of zinc sulfide (ZnS) which is a very unusual component and only rarely occurs in some of the samples that were scanned. Another interesting spot in the current sample is number 15 which seems to be located between larger grey domains (presumably coarse apatite particles). This point shows a large concentration of carbon along with oxygen. Since the spot does not contain a significant amount of calcium, the area could potentially be a high carbon region of organic nature rather than an inorganic carbonate mineral phase. As shown by these examples, association of various elements to form a mineral can be understood and visualized to a great degree by superimposing individual element maps. As stated above, this step in data analysis was achieved by using the ImageJ software. To demonstrate the full procedure, the following element maps acquired on the same sample surface will be used as an example. The maps correspond to the BSE image shown in Figure 4.1: Elemental Composition (%) for sample#2 Point ID C O F Al Si P S K Ca Zn Sum 8 22.75 8.5 - 0.38 2.89 1.00 20.25 - 2.36 35.12 93.26 9 16.87 8.95 - 0.81 3.70 1.68 21.16 - 3.36 39.49 96.02 10 12.79 7.40 - 0.83 2.51 1.88 21.88 - 3.99 41.05 92.35 11 8.72 58.83 - 16.32 18.59 1.27 0.54 5.05 3.65 - 112.97 12 9.19 40.48 4.59 0.48 1.95 13.96 1.24 - 32.89 1.11 105.89 13 10.95 40.72 4.16 0.46 1.79 12.99 1.54 0.10 31.34 0.87 104.92 14 8.76 61.33 - 0.63 40.90 1.37 0.63 - 3.07 1.24 117.92 15 51.55 24.98 - 3.11 7.77 3.35 4.69 1.34 8.81 - 105.60 23                 Figure 4.3 Element maps for Phosphorous, Carbon, Aluminum and Silicon for the sample#1 24 The images reflect well how the phosphorous grains are discrete in their occurrence while the carbon seems to occur everywhere on the sample surface, in addition to some well-defined brighter spots where the carbon concentration appears to be much higher than in the surrounding areas. The carbon map also shows darker areas, equivalent to low carbon areas, which in many cases correspond to the high-phosphorus grains. The silicon areas, which are indicative of quartz and aluminosilicate minerals, occur in between those high-phosphorous grains, almost as a filler material holding them together. It is noticed that the silica areas exhibit good correlation with the aluminum and potassium maps thereby confirming the presence of quartz and aluminosilicates. Another set of EDX data is shown in Figure 4.4 below.        Figure 4.4 Backscattered image and element map for Phosphorous for sample#3 25                 Figure 4.5 Element maps for Phosphorous, Carbon, Silicon and Aluminum for sample#3  26 Again, the phosphorous occurs as discrete grains, of a larger size this time (~600micron range) with the carbon in the interstices between the grains and also on the grain boundaries. The silica can be seen in between the grains as a cementing material. In order to quantify the overlap between specific elements of interest, the following steps were followed: 1. The individual element maps were first converted to a greyscale binary image (8-bit) for further analysis.  2. The individual element maps were then imported into the software. 3. These greyscale (black-and-white) images were then overlaid using the ImageJ software by applying the ‘AND’ filter available in the software. An image with only the areas that overlapped in both the images was obtained.  4. The surface area of the bright spots of the resulting map compared to the original areas of the element maps, give the amount of overlap between the elements to form minerals. An example of this type of transformation is shown below (Figure 4.6), presenting an overlay between Si and Al and the calculations for the overlap, done for sample#3. Slice (Sample Surface) Total Area (mm2) %Area Black/White Al 2.16 16.07 Black/White Si 0.87 6.46 Al & Si 0.45 3.35 Fraction of Al overlapping with Si = (0.45/2.16) = 0.21 Fraction of Si overlapping with Al = (0.45/0.87) = 0.52 Table 4.3 Semi-quantitative analysis using ImageJ 27   Figure 4.6 Black and White images for Silicon, Aluminum and the overlap between the two for sample#3 The analysis is useful in looking at associations between the different elements and the minerals they form. As an example, the fraction of Si that does not show an association with the Al (0.48 = 48% for the above example), represents the fraction that is present as SiO2 (quartz) while the fraction that shows association with Al (0.52 = 52%) is the aluminosilicate phase (clay fraction). The analysis of this nature was extended to multiple elements and surfaces, for example, to look at the associations between individual elements of interest (C and P) or minerals of interest with associated elements (P and SiO2, P and aluminosilicates). The table below shows the association for carbon and phosphorous for the same sample surface. It shows that even though a large fraction of the carbon present associates with the phosphorous, the remainder is essentially either present as inclusions within the phosphorous phase or in the surrounding matrix which is also verified with the visual observations.   28  Slice (Sample Surface) Total Area (mm2) %Area Black/White C  1.47 10.90 Black/White P 7.56 56.27 C & P 0.96 7.12 Fraction of P overlapping with C = (0.96/7.56) = 0.13 Fraction of C overlapping with P = (0.96/1.47) = 0.65 Table 4.4 Semi-quantitative analysis using ImageJ Thus, the SEM imaging for these samples provided some key mineralogical features for further analysis: 1. The phosphate (as seen from the phosphorous map occurs as discrete grains varying in size from about 50microns up to a size of 600microns. 2. The carbon in its native state is found to occur inside the phosphate grains and in the void spaces between the grains.   3. The silicon occurs together with aluminum and potassium, along grain boundaries, which suggests that clay (aluminosilicate) minerals act as a binding and a cementing material for the coarser phosphate particles, holding the aggregates together. The presence of “unassociated silicon”, presumably as quartz is also evident. 4. A fraction of organic carbon appears to be present in the same areas as the fine clay minerals, i.e., in the void spaces between the well-defined phosphate grains. 5. Overall, the large ore particles whose cross sections were analyzed by SEM-EDX are not large crystals or homogeneous rocks. They are aggregates of relatively coarser phosphate grains of varying sizes with spaces between them filled by finer silica and aluminosilicates. 6. The images and the overlays were done for 10 sample surfaces and were found to be consistent thereby validating the process and underlying image logic. The results for the 29 fraction of overlay between various elements of interest (averaged for all sample surfaces scanned) is tabulated below. Elements of Interest Average Fractional Value Standard Deviation Fraction of Phosphorous (P) overlapping with Carbon (C) 0.21 0.08 Fraction of Carbon (C) overlapping with Phosphorous (P) 0.57 0.12 Fraction of Aluminum (Al) overlapping with Silicon (Si) 0.28 0.09 Fraction of Silicon (Si) overlapping with Aluminum (Al)  0.39 0.14 Fraction of Phosphorous (P) overlapping with Silicon (Si) 0.04 0.02 Fraction of Silicon (Si) overlapping with Phosphorous (P) 0.19 0.06 Fraction of Sulphur (S) overlapping with Carbon (C) 0.33 0.21 Fraction of Carbon (C) overlapping with Sulphur (S) 0.31 0.18 Table 4.5 Summary of semi-quantitative analysis using ImageJ 30 Chapter 5: Attrition Scrubbing SEM-EDX testing of the coarse ore particles showed clearly that the particles are actually aggregates of finer phosphate particles, held together by a mixture of finer silica and clay (aluminosilicate) particles. The carbon was found to occur as inclusions inside the phosphate grains and as very fine material dispersed between phosphate particles. The latter carbon fraction appeared to fill the same areas as the clay particles. Based on this aggregate nature of the ore, the working assumption was that it should be possible to carefully disintegrate the aggregates (rather than crushing/grinding the entire feed to the same size range), and in doing so; liberating the phosphate grains without altering their original particle size. As a result, the response of the coarse particles to various conditions was investigated. A testing procedure involving minimal mechanical force was envisaged as a next possible step to reduce the particles in size, without necessarily crushing them all down to a sub-millimeter size range. The first step involved single particle tests and investigating the effect of chemicals on the physical state of the tested particles. The as-received coarse particles were placed in different solutions in a test-tube, at varying pH levels (highly acidic to highly basic), as well as in a range of organic solvents (acetone and toluene), and any changes in the appearance of the particles was observed by immersing them in the solvents for 24 hours without any mixing or stirring. Organic solvents were used as a dispersing medium in an attempt to elucidate the behavior of the carbon phase rather than to develop an ore treatment technique. Depending on the chemical nature of the organic phase in the particles (aliphatic, aromatic, oxidized, etc.), it was expected that the organic matter should differently respond to the tested solvents, and that the interparticle voids (filled up with clay and carbon material) would be wetted by solvents leading to partial or even complete disintegration of the coarse aggregate-type particles. The appearance of fines, crumpling of the particles, any spontaneous breakage/disintegration phenomena were noted.  In a more aggressive treatment, the 31 immersed particles were also placed in an ultrasonic bath for 60 minutes after having been left immersed in the solvent/solution for 24 hours and similar observations regarding the physical state of the particles were made. Further, the effect of temperature on the physical integrity of the articles was also monitored. In this round of tests, the particles (immersed in their respective solutions/solvents) and having gone through the ultrasonic bath stage, were put on an incubating shaking table and shaken at successively increasing temperatures (25°C, 30°C and 35°C), at 250 rpm for a period of 60 minutes at each temperature. Turbidity of the solution was used as a measure of the release of fines from the particles into the suspension and the measurements were made after each successive temperature setting. The actual test details and observations are added as an appendix to the thesis, but the turbidity values and the most significant findings can be summarized in the following tables. Note that each row represents one complete testing cycle performed on a single particle of the mentioned type and is independent of the preceding and/or succeeding rows. The turbidity values are noted in NTU (Nephelometric Turbidity Unit). Table 5.1 Single particle tests in different solvents  Test Conditions Particle Type After Ultrasonic Bath 60 minutes, 25°C 60 minutes, 30°C 60 minutes, 35°C Sulphuric Acid (98%) Dark 4000 5875 2504 3000 Sulphuric Acid (98%) Light 157 146 155 170 Acetone Dark 690 298 322 296 Acetone Light 110 64.5 149 137 32 Test Conditions Particle Type Ultrasonic Bath 60 minutes, 25°C 60 minutes, 30°C 60 minutes, 35°C Sulphuric Acid (pH-2) Dark 774 346 655 714 Sulphuric Acid (pH-2) Light 1228 1994 2582 3257 Sulphuric Acid (pH-3) Dark 271 89.1 50.3 135 Sulphuric Acid (pH-3) Light 233 653 1128 1310 Table 5.2 Single particle tests in acidic media 5.1 Observations 1. When left in highly concentrated sulfuric acid (98%), the dark particles showed the strongest tendency to disintegrate which was confirmed by the high turbidity values. The light particles, however, did not show turbidity values as high as the dark particles. 2. The particles showed little to no disintegration in organic solvents except acetone. All other organic solvents (toluene and n-hexane) had no discernable effect on the particles, but acetone did show some effect on the stability of the particles. Since the effect was not considerable, at least visually, the turbidity measurements were not recorded. 3. Similarly, higher pH values were also tested. The particles showed the highest Nephelometric Turbidity Unit (NTU) values at pH12 both for the light and the dark particles. 4. As for acidic conditions, the disintegration of both the light and dark particles at pH 2 was significantly more advanced than at pH 3. The tests gave an indication of the optimum conditions that worked for the spontaneous disintegration of these coarse aggregate particles. Although there was no chemical assay done on the solution after the tests, the low pH conditions seemed to offer a better environment for disintegration for single particles. 33 For the next stage of tests, a batch of particles instead of a single particle, was tested. Both the light and dark particles were subjected to similar conditions. A high solids ratio (greater than 33% by weight) was chosen to observe the effect of acid in conjunction with the rubbing/abrasion due to the now possible particle-particle collisions. A set of tests under acidic conditions (pH ranging between 2 and 3) were chosen but the overnight conditioning (soaking) in acid and the ultrasonic bath treatment were omitted. The samples were directly put through the various stages in the shaker with progressively increasing temperatures (stage-1 (25°C), stage-2 (30°C) and stage-3 (35°C)), at 250rpm for a period of 60 minutes each kept in a nalgene bottle. Turbidity values were measured and tabulated (in NTU) at the end of each stage on the shaking table and are shown below. Detailed results with the images of the solution after each stage are in the appendix, but the key observations from the tests are as follows: Test Conditions Particle Type After Stage-1  After Stage-2 After Stage-3 Sulphuric Acid (pH-2) Dark 1310 9580 7302 Sulphuric Acid (pH-2) Light out of range 8610 7295 Sulphuric Acid (pH-2.4) Dark 6870 out of range 5186 Sulphuric Acid (pH-2.4) Light 3362 8494 out of range Table 5.3 Batch test in acidic media The turbidity values were seen to be quite high for all the samples. These high values can be attributed to inter-particle rubbing and abrasion that causes them to break apart and disintegrate which was not seen as much in the individual particle tests. However, none of the tested conditions and their combinations resulted in complete disintegration of the original coarse particles. 34 Although the release of fines was substantial, as shown by the high turbidity values, relatively coarse particle cores remained unaffected.  The results from these tests along with the characterization studies done previously, were used to develop conditions for attrition-scrubbing of these samples at a high solids content. Attrition scrubbing is a common unit operation in industrial mineral processing, particularly in the phosphate and potash industries. The main purpose of the process is to release slimes from the surfaces of coarser grains under intense mixing conditions. The potash and phosphate industries target the recovery of as coarse particles as possible, and the release of slimes during attrition scrubbing facilitates de-sliming of the feed for subsequent froth flotation. Otherwise, the presence of unwanted ultrafine particles in flotation feed leads to slime-coating phenomena and very high reagent consumption rates. The main disadvantage of the process is the fact that very fine phosphate grains are also lost with slimes.  Nevertheless, attrition scrubbing was considered a meaningful conditioning step before further processing as it would avoid the more intensive steps of crushing and grinding and reducing the entire sample into the same size range which then makes the removal of the well-liberated phosphate particles difficult from the clay-silica matrix.  With the information about the samples obtained thus far based on SEM analysis, and the single particle test and batch tests, it was apparent that the coarse rocks are in fact aggregates of smaller phosphate particles “glued” together by alumino-silicate and carbon phases. The phosphate phase appears as coarse (600-microns and finer) well defined grains, with inter-grain spaces filled by alumino-silicates, silica and carbon. This type of mineral occurrence suggests that an attrition scrubbing approach would be useful to disintegrate the coarse particles rather than to grind the entire sample to a top size required for froth flotation (150-200 microns). Disintegration during 35 attrition was expected to release the well-defined phosphate grains from the silica-aluminosilicate-carbon matrix. 5.2 Attrition Scrubbing Tests A range of both high and low pH values were tested, and the resulting size fractions were assayed separately to test the degree of phosphate release in the different size ranges. To find an optimum combination of pH and attrition scrubbing time, a range of different conditions were investigated. The ore was homogenized and separated into charges ranging from 440gm to 600gm and was attrition scrubbed. Even though the weight of sample for the individual tests varied, the solids content (wt/wt%) in the attrition cell was always kept in the 50-60% range. The tests were carried out in the laboratory using a Denver D-12 flotation machine (Figure 5.1). The Denver D-12 has a 1/2HP, 1425rpm, 50Hz, 110/220 V, single-phase motor and is widely used for attrition scrubbing on a laboratory scale using the attrition scrubbing accessory designed for the equipment. All tests were conducted at 900rpm and with the other parameters as stated in Table 5.4. It must be noted that the rpm was chosen to be a relatively low value to test mild conditions for attrition scrubbing.   36  Figure 5.1 Denver D-12 Flotation Machine  The attrition-scrubbing stage was followed by wet-sieving for size separation. The particle size distributions of the products for all the tests are added as a part of the appendix but the mass balance calculations with assay results are tabulated in the following pages. A comparative graph with the effect of different attrition times and pH values is also added to show how the ideal set of conditions from our tests were fixed. Table 5.4 Testing conditions for attrition scrubbing S.No. Initial pH Final pH Attrition Time (minutes) Solids Percent (%) 1 2.5 4.85 5 56.7 2 2.5 4.81 10 54.4 3 2.5 4.73 20 51.5 4 3.5 5.6 20 54.3 5 10.5 8.1 20 54 6 Natural pH (5.98) 6.03 20 55 7 Natural pH (6.06) 6.1 10 55 37 After the attrition scrubbing step, the products were wet screened on an inclined screen with gradually finer sized screens. All the fractions were washed down with a water source to ensure the screening was performed efficiently and consistently for all tests. The undersize (-0.075mm) was pressure filtered, dried overnight and weighed after. The distribution of the products is shown below:   Figure 5.2 Particle Size Distribution for the attrition products in acidic media As the graph shows, the top size of the feed sample was around 12mm, with only about 14% of the material below 0.3mm. The shorter attrition times, on the order of 5-10 minutes result in a substantial change in the size distribution, and only a small further change is observed at 20 minutes. It does appear that the most significant changes in the particle size occur at shorter attrition times. The curves generally show that the scrubbing does not uniformly affect the size distribution of the tested sample. Although the fraction of fines, i.e., material finer than 0.3mm, increases from 15% to 50%, the product still contains substantial amounts of very coarse particles. This indicates that the finer particles are gradually released from the coarser particles while still 38 preserving the cores of the coarsest fraction without completely breaking them. As a result, all products are characterized by bimodal particle size distributions, as seen by the plateau areas in all the curves between 0.1 and 1mm. Such flat sections indicate the presence of only a very small amount of material in that size range. The curves show the presence of two dominant fractions:  1. About 50 - 60% of the product is coarser than 2.4mm, while between 25 - 40% is finer than 0.075mm. 2. Only between 10 - 20% lies in the -2.4 + 0.075mm size fraction  Figure 5.3 Particle Size Distribution for the attrition products – longer attrition time Even though single particle and batch tests on the shaker showed that pH had an effect on the disintegration of the coarse particles, the attrition-scrubbing time rather than the conditioning pH value seems to have a greater effect on the breakage of the feed. The experiments performed at different pH values gave very similar particle size distributions suggesting the same mechanism 39 of disintegration regardless of pH, as seen in the data presented in Figure 5.3. In all the cases, the particle size distributions of the products are bimodal, as noted before. It seems that the intense conditions of mixing in the attrition cell overcome any potential benefits of adjusting the pH of the feed slurry ahead of attrition.  Figure 5.4 Product size comparison for attrition scrubbing products Finally, Figure 5.4 shows that increasing the attrition time from 10 to 20 minutes does not change the particle size distribution to a significant extent, suggesting that the greatest changes occur in the early stages of attrition and that breakage of the particles ceases after about 10-20 minutes. As before, all the products show bimodal particle size distributions. 5.3 Mass Balance from Assays After the attrition scrubbing tests, the various size fractions were sent for assays to analyze the composition of each fraction and compare the differences between them. The objective was to 40 follow the distribution of the main ore components between different size fractions as a result of attrition-scrubbing, and to verify the overall assumption that the disintegration of the aggregate-type particles under attrition leads to a release of liberated phosphate grains at their natural sizes to the corresponding size fractions. Each size fraction underwent a standard x-ray fluorescence analysis (whole rock analysis by XRF) for 13 components, in addition to organic carbon and sulfur assays. The XRF results are reported as “oxide equivalents”, while the carbon and sulfur contents are the actual mass concentrations of those elements. The following sections focus only on the main ore components (P2O5, SiO2, Al2O3, and C), while the appendix includes the full analytical reports from the lab. The products were screened at the same 12 different size fractions after each test, but the different size ranges were blended together into four main sizes to be sent for assays to the lab. This decision was based on several factors: 1. A closer analysis of the bimodal particle size distributions of the attrition products, showed that the majority of the material was found in the +2.4mm range and in the -0.075mm range. These two size ranges were therefore used for assays. 2. The -0.600mm + 0.075mm fraction was of special interest to analyze the liberation of the phosphate in that range. Based on the SEM-EDX data, the phosphate grains in the tested coarse aggregate particles approximately occurred in that size range. 3. The intermediate size fraction of -2.4mm + 0.600mm was analyzed to obtain the complete set of data for mass balance calculations for the entire scrubbing product. The assay results with the mass balances are tabulated below. The recovery values (distribution) of the product were calculated in the following way: 41 1. The assay values (%) were multiplied by the weight percent of each fraction and divided by 100 to get the ‘units’ value for each oxide equivalent of interest and of the organic carbon. Units =  !""#$	&#'()	(%)	∗	.)/012	%344  2. Recovery (distribution) of each mineral and organic carbon phase was then calculated by dividing the individual units value by the total units of each mineral in the feed. Recovery (distribution) = 5678	9:;<=	>?	8@=	A7B=	?C:D87>6E>8:;	<678A	>?	8@=	F76=C:;   The calculation for the units has been shown for the first table as an example and has been done for the subsequent tests in the same way but has not been tabulated and only the final recovery (distribution) values have been shown.42 Test Result #1 – Uncrushed Feed  Table 5.5 Assay results uncrushed feed The table above shows the assay results for the feed. The P2O5 and organic carbon content remains fairly uniformly distributed for all size ranges even though the recovery is higher for the coarser fraction. The P2O5 grade of the feed can be back-calculated to be about 19.1%. The sample also contains 5.5% of organic carbon.  Test Result #2 – Particles scrubbed for 20-minutes at pH2.5  Table 5.6 Assay results – pH2.5, scrubbed for 20minutes Size Fractions Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 360.5 57.7 5.75 26.2 4.2 20.3 3.3 15.1 2.4 11.7 60.6 53.2 52.9 61.3-2.4 + 0.600 134.4 21.5 5.37 27.3 4.6 20.0 1.2 5.9 1.0 4.3 21.1 20.6 21.6 22.4-0.600 + 0.300 42.7 6.8 5.04 24.0 4.2 22.0 0.3 1.6 0.29 1.5 6.3 5.8 6.3 7.9-0.300 86.7 13.9 4.74 41.6 6.4 11.6 0.7 5.8 0.89 1.6 12.0 20.3 19.2 8.4Total 624.3 100.0 5.5 28.4 4.6 19.1Assay Units Recovery [%]Size Fractions Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 189.4 37.5 6.5 23.2 3.7 22.6 41.7 32.0 30.4 41.5-2.4 + 0.600 48.5 9.6 6.0 20.6 3.3 23.8 9.7 7.3 6.9 11.2-0.600 + 0.075 67.6 13.4 4.8 8.6 1.3 32.4 11.0 4.2 3.7 21.2-0.075 199.6 39.5 5.6 38.9 6.8 13.6 37.6 56.5 58.9 26.2Total 505 100.0Assay Distribution [%]43 Upon scrubbing at pH 2.5 for 20-minutes, the P2O5 and organic carbon contents are slightly higher in the coarsest fraction in the scrubbing product compared to the feed levels in the same size fraction, but there is now a marked increase in the P2O5 content in the (-0.600mm+0.075mm) range to about 32%, with a simultaneous decrease in the carbon content in the same size range.  Test Result #3 – Particles scrubbed for 20-minutes at pH3.5  Table 5.7 Assay results – pH3.5, scrubbed for 20minutes This scrubbing product tends to show a similar trend in terms of the assay results for the fraction of interest as marked on the table. The particle size distributions showed that the scrubbing product size was affected by the scrubbing time rather than by the pH of the solution, and that observation seems to be also reflected in the very similar assay results at pH 3.5 and at pH 10.5 (Table 6.8) at the same scrubbing time of 20-minutes.   Test Result #4 – Particles scrubbed for 20-minutes at pH10.5 Size Fractions Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 179.1 38.7 6.0 21.8 3.6 21.6 41.3 32.0 31.2 41.5-2.4 + 0.600 45.6 9.9 6.2 20.9 3.3 23.6 10.7 7.8 7.3 11.6-0.600 + 0.075 59.1 12.8 4.9 8.8 1.4 32.1 11.1 4.3 3.9 20.4-0.075 178.5 38.6 5.4 38.3 6.6 13.9 36.9 55.9 57.5 26.5Total 462 100.0Distribution [%]Assay 44  Table 5.8 Assay results – pH10.5, scrubbed for 20minutes Even at a high pH, the same trends are followed in the assay results. The (-0.600mm+0.075mm) size range consistently shows a high P2O5 content of over 32% and a relatively low organic carbon content. Test Result #5 – Particles scrubbed for 10-minutes at pH2.5  Table 5.9 Assay results – pH2.5, scrubbed for 10minutes This test was done to check the effect of scrubbing time and compare the difference between a 10-minute and a 20-minute scrubbing time at the same acidic conditions. It was observed again that the P2O5 grades were quite similar in the corresponding size ranges and the same size fraction (from -0.600mm to +0.075mm) gave a high P2O5 grade of over 30% in all the tests. Test Result #6 - Particles scrubbed for 20-minutes at natural pH (pH 6) Size Fractions Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 138.4 34.2 6.3 24.1 3.8 21.4 37.3 30.4 29.1 36.3-2.4 + 0.600 37.9 9.4 6.3 20.3 3.3 23.8 10.2 7.0 6.8 11.0-0.600 + 0.075 60.7 15.0 5.0 9.3 1.5 31.8 12.8 5.2 5.0 23.6-0.075 167.5 41.4 5.5 37.6 6.4 14.2 39.7 57.4 59.1 29.1Total 405 100.0Assay Distribution [%]Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 244.9 44.6 5.4 20.8 3.4 22.6 48.4 36.2 35.7 48.3-2.4 + 0.600 56.5 10.3 5.7 21.0 3.4 23.4 11.6 8.4 8.3 11.5-0.600 + 0.075 81.3 14.8 4.7 10.7 1.6 30.6 14.0 6.2 5.8 21.7-0.075 166 30.3 4.3 41.6 7.0 12.8 26.0 49.1 50.3 18.5Total 549 100.0Assay Distribution [%]45  Table 5.10 Assay results – natural pH (5.98), scrubbed for 20minutes As seen with the particle size distribution of the various product streams, the assay results reflect the same trend that the products of attrition are compositionally the same irrespective of the solution chemistry during the scrubbing test. This test also shows that the natural pH of the solution formed by immersing these particles in water, gives similar results for scrubbing products, compared to the tests done with much harsher chemical conditions. Test Result #7 - Particles scrubbed for 10-minutes at natural pH (pH 6.06) Table 5.11 Assay results – natural pH (6.06), scrubbed for 10minutes The scrubbing done for 10-minutes at natural pH also showed similar results as the test at a scrubbing time of 20-minutes. Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 183.2 42.5 5.6 25.2 3.7 22.2 46.9 39.2 36.5 45.5-2.4 + 0.600 45.1 10.5 5.7 22.2 3.5 22.9 11.7 8.5 8.6 11.6-0.600 + 0.075 66.9 15.5 4.8 11.0 1.7 30.4 14.6 6.2 6.0 22.8-0.075 135.4 31.4 4.4 40.0 6.7 13.3 26.8 46.0 48.9 20.2Total 431 100.0Assay Distribution [%]Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 184.2 44.5 5.5 26.1 4.0 21.0 47.4 42.3 40.5 45.7-2.4 + 0.600 35.4 8.5 5.6 20.5 3.2 23.7 9.3 6.4 6.3 9.9-0.600 + 0.075 53.5 12.9 4.5 8.6 1.3 32.0 11.3 4.1 3.7 20.3-0.075 141 34.0 4.9 38.0 6.3 14.4 32.1 47.2 49.5 24.1Total 414 100.0Assay Distribution [%]46  Figure 5.5 P2O5 content variation with scrubbing time and conditions for (-0.600mm+0.075mm) size range The graph above shows the variation in the P2O5 grade for the (-0.600mm+0.075mm) size range for the different test conditions compared to the total feed grade. The grade consistently lies above the 30% mark for all scrub times which is about a 10% upgrade compared to the raw feed.  Figure 5.6 Organic carbon vs P2O5 content variation for product obtained at natural pH and scrubbed for 20minutes 47 The graph above shows a correlation between the P2O5 content in different size fractions and the organic carbon content in the same fractions, obtained from an attrition-scrubbing test at natural pH for 20-minutes. The (-0.600mm+0.075mm) size range does show a small decrease in the amount of carbon present compared to the other size ranges, which points to a cleaner fraction of higher grade P2O5. It is also difficult to release carbon into the finest size fraction and the main components of that fraction are phosphate, clay and silica.  Other trends and correlations between the content of the various components and their variation in the different size fractions is looked at closely in the concluding section. Since the particle size distributions of the products did not significantly change after about 20 minutes of attrition-scrubbing, and the assays were very similar for the different testing conditions, it was rather clear that attrition-scrubbing reached a limit beyond which neither the P2O5 grade nor the particle size distribution of the products was significantly changing. Since the focus was on determining the level of disintegration of coarse particles, it was decided to re-scrub an oversize fraction to more clearly assess the behavior of that material alone. For this experiment, the +2.4 mm size fractions of all the products from the different tests (and conditions) were combined (total mass of 274.2gm) and attrition-scrubbed for another 20 minutes at a solids content of 55% at natural pH of the resulting solution (pH 6.3). The results of the particle size distribution and the assays on the various products are as shown below along with several other sets of data from other tests for comparison. 48  Figure 5.7 Comparison of the product of the re-scrubbed ore with the original scrubbing tests The plot shows how the curve flattens out at about 25% passing 1mm size. It also tends to fall more rapidly from the top size (~9.5mm) to the 1mm size. In this case, there was an insufficient amount of material in the characteristic (-0.60 + 0.075mm) size fraction for reliable assays, so a combined sample for the (-1.68+0.075mm) size fraction was assayed instead as an approximate measure of the composition of the (-0.60+0.075mm) fraction. The assay results for the products of this test are as: 49   Table 5.12 Assay results for the re-scrubbed ore at natural pH for 20minutes The broader size fraction from (-1.6mm to +0.075mm) still shows a higher grade of P2O5, equal to 26% compared to the other fractions. Since this fraction is a blend of the (-0.600 + 0.075mm) material with the (-1.6 + 0.600mm) fraction, it is entirely possible that the material in the narrower (-0.600 + 0.075mm) size range is of much higher grade since the coarser fractions usually contain about 20-23% P2O5, as seen in all the previous assays. However, under these conditions, only about 5% of the material is in the higher-grade size fraction, and this fraction recovers an additional 5.7% of P2O5. This test also gave a clearer idea about how the oversize material responds to re-scrubbing as opposed to using a longer scrubbing time for the full feed. Scrubbing alone reaches a limit of what can be achieved with the method. Although small increase in the overall recovery could be achieved, there is no distinctive benefit in re-scrubbing the oversize products. 6.3 Scrubbing followed by crushing To test the liberation characteristics of the difficult to scrub oversize material, two crushing tests were performed in which the +2.4mm fraction from two independent attrition scrubbing tests, were dried and then crushed in a jaw crusher and the products were then assayed by size fraction. The two crushing tests were conducted on the attrition scrubbing product in a way that they were 100% passing 2.4mm Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5+2.4 179.4 68.9 5.6 20.3 3.2 24.0 64.2 59.5 58.2 74.8-2.4 + 1.68 14.6 5.6 6.3 25.1 4.0 20.8 5.8 6.0 5.9 5.3-1.68 + 0.075 12.6 4.8 5.6 17.6 2.8 26.0 4.5 3.6 3.5 5.7-0.075 53.9 20.7 7.4 35.1 6.0 15.3 25.5 30.9 32.4 14.3Total 261 100.0Assay Distribution [%]50 in the first test and 100% passing 1.8mm in the second test. These sizes were chosen so that the crushing products would be comparable to the assay results already acquired from the previously conducted scrubbing tests.     Test -1:       Figure 5.8 Attrition scrubbing followed by crushing – test#1  Starting weight – 462.3gm, Natural pH (6.1) Attrition Scrubbed for 20min, 55% solids [wt/wt%] Crushed to 100% passing 2.4mm and assayed Oversize (+2.4mm) Undersize (-2.4mm): assayed Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5-2.4 + 0.600 37.2 9.2 5.8 21.4 3.3 23.1 0.5 2.0 0.3 2.1 9.9 7.7 7.3 10.0-0.600 + 0.075 40.5 10.1 4.6 7.7 1.1 33.1 0.5 0.8 0.1 3.3 8.6 3.0 2.7 15.6-0.075 151.7 37.7 5.2 37.5 6.4 14.8 1.9 14.1 2.4 5.6 36.0 54.8 57.7 26.3Total 402.3 5.4 25.8 4.2 21.3Assay Units Distribution [%]Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5-2.4 + 0.600 135 33.6 5.7 21.5 3.4 23.6 1.9 7.2 1.1 7.9 35.6 28.0 27.0 37.2-0.600 + 0.075 28.9 7.2 5.6 18.1 2.9 25.9 0.4 1.3 0.2 1.9 7.5 5.0 4.9 8.7-0.075 9 2.2 5.9 29.6 4.6 19.2 0.1 0.7 0.1 0.4 2.4 2.6 2.5 2.0Total 402.3 5.4 25.8 4.2 21.3Assay Units Distribution [%]51 Test 2:          Figure 5.9 Attrition scrubbing followed by crushing – test#2  Starting weight – 452.5gm, Natural pH (6.08) Attrition Scrubbed for 20min, 55% solids [wt/wt%] Crushed to 100% passing 1.18mm and assayed Oversize (+2.4mm) Undersize (-2.4mm): assayed Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5-2.4 + 0.600 47 11.3 5.9 21.3 3.4 23.9 0.7 2.4 0.4 2.7 12.3 8.8 8.5 13.3-0.600 + 0.075 48.6 11.7 4.4 8.4 1.3 32.5 0.5 1.0 0.1 3.8 9.5 3.6 3.3 18.7-0.075+0.045 86.2 20.8 4.9 37.5 6.9 14.5 1.0 7.8 1.4 3.0 19.0 28.3 31.7 14.7-0.045 56.6 13.6 4.3 44.8 7.0 11.5 0.6 6.1 1.0 1.6 10.9 22.2 21.3 7.7Total 415.0 5.4 27.5 4.5 20.4Assay Units Distribution [%]Size Fractions[mm] Weight [g] Weight [%] C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5 C [organic] SiO2 Al2O3 P2O5-1.18 + 0.600 79 19.0 5.9 23.7 3.8 21.9 1.1 4.5 0.7 4.2 20.9 16.4 16.1 20.5-0.600 + 0.075 72.1 17.4 6.0 21.1 3.4 23.7 1.0 3.7 0.6 4.1 19.3 13.3 13.2 20.2-0.075 25.5 6.1 6.3 32.8 5.2 16.6 0.4 2.0 0.3 1.0 7.2 7.3 7.1 5.0Total 415.0 5.4 27.5 4.5 20.4Assay Units Distribution [%]52 1. For both the tests, the scrubbing product was wet-screened and dried as with the previous tests, and a particle size distribution was conducted. The undersize (-2.4mm) was assayed for both tests. To get a better understanding of the composition of the -0.075mm fraction; the undersize for test-2 was split up into a (-0.075 + 0.045mm) and a (-0.045mm) fraction. 2. The oversize was crushed to 100% passing 2.4mm for the first test and 100% passing 1.18mm for the second test. These two sizes were chosen so as to not crush everything to a fine size while at the same time attempting to liberate the phosphate phase from the aggregates.  3. The undersize assays from the scrubbing product again shows the (-0.600 + 0.075mm) fraction with a high P2O5 grade confirming the results obtained from the previous tests. 4. Another benefit with the crushing is seen with the observation that not a very high mass of P2O5 is lost to the fines (-0.075mm). 5. Of the two scenarios, crushing to below 2.4mm (test-1) yields more favorable results in terms of the grade of P2O5 obtained with the crushing product and also the mass pull as a fraction of total P2O5 distribution of the product.  53 Chapter 6: Results and Discussion The testing conditions for the attrition tests are tabulated below to make referencing to the test conditions easier. Table 6.1 Conditions for attrition scrubbing tests The following graphs provide analysis of the products from the attrition tests and the trends that are observed in relation to the individual streams that were assayed.         S.No. Initial pH Final pH Attrition Time (minutes) Solids Percent (%) 1 2.5 4.73 20 51.5 2 3.5 5.6 20 54.3 3 10.5 8.1 20 54 4 2.5  10  5 Natural pH (5.98) 6.03 20 55 6 Natural pH (6.06) 6.1 10 55 7 Natural pH, combined oversize from previous tests, re-scrubbed 6.25 20 55 54  1. +2.4mm size fraction   Figure 6.1 Main components of the +2.4mm fraction of the attrition products Although not a very considerable difference, the P2O5 inversely correlates with the SiO2 content for this size fraction. From the separation point of view, this is a desirable effect since higher P2O5 grades can be expected with lower SiO2 contents, and vice versa. The trendline for Al2O3 follows the trend with the SiO2 contents, and, even though it is not as pronounced; it shows the occurrence of both silica and clays in the mineral matrix.       0.05.010.015.020.025.030.01 2 3 4 5 6 7Grade (%)Test numberMain components of the +2.4mm fraction of attrition productsSiO2P2O555  2. (-2.4+0.600mm) size fraction   Figure 6.2 Main components of the -2.4 + 0.600mm fraction of the attrition products The complimentary nature of the trendlines for P2O5 and SiO2 is less evident for the given size range and the grades do not show a great deal of variation for both. The carbon trendline closely matches the P2O5 line showing that the two phases are not liberated to a great deal in this size range and the interstitial carbon, that was observed in the SEM images still exists in the larger phosphorous grains.       0510152025301 2 3 4 5 6 7Grade (%)Test numberMain components of the (-2.4 + 0.60mm) fraction of attrition productsSiO2P2O556  3. (-0.600+0.075mm) size fraction   Figure 6.3 Main components of the -0.600 + 0.075mm fraction of the attrition products This particular size range showed the best liberation of the phosphate phase along with the lowest amount of silica. The trendlines show the phosphate grade consistently above 30%, up from around 20% in the feed while the SiO2 content in the concentrate decreases significantly from 27% in the feed to less than 10% in the concentrate. This means that the phosphate and the silica become highly liberate from one another and that after liberation they accumulate in different size fractions – silica goes to the fines while the phosphate stays in the coarse fraction which allows their separation to take place by simple screening. The next graph also shows the distribution clearly and supports this result. The carbon line stays around the 5% range meaning that this is the amount of carbon locked in with the phosphate phase and the difference between this value and the values from the coarser size ranges is the amount of carbon in the mineral matrix.  051015202530351 2 3 4 5 6 7Grade (%)Test numberMain components of the (-0.60 + 0.075mm) fraction of attrition productsSiO2P2O557  4. -0.075mm size fraction   Figure 6.4 Main components of the -0.075mm fraction of the attrition products The finest fraction shows the higher content of silica which largely constitute the slimes and the relatively low content of phosphate. Silica is now the dominant component with a content of about 40%. The carbon still exists in the same 5% range meaning that this percentage is the amount that is locked in the interstices and cannot be liberated without a more intensive mechanical treatment. To quantitatively analyze the liberation of the phosphates for the scrubbing products, the finer fractions (-0.600mm), were analyzed under SEM. The idea was to see the phosphate to be seen in the backscattered images as occurring independently and being well liberated from the clays and silica. The products of attrition were set in epoxy, polished and carbon coated for the scanning. Two different samples were prepared for the same. 0510152025303540451 2 3 4 5 6 7Grade (%)Test numberMain components of the -0.075mm fraction of attrition productsSiO2P2O558  Figure 6.5 Scatterplot for silica vs phosphate grade variation The plot above shows the variation of the silica content in comparison to the phosphate. It shows all the size fractions assayed for all the scrubbing tests along with the feed. The scatter of points is an interesting spread, highlighting the complimentary nature of the components mentioned above. The group of points on the top-left portion of the plot, represent the low-silica and high-phosphate content and are all points from the (-0.600 + 0.075mm) size fraction while the group of points in the bottom right portion of the curve, represent high-silica and low-phosphate content and are all from the (-0.075mm) size fraction of the scrubbing tests. Although even this finest size range also has a considerable amount of phosphate, the processing techniques for the two fractions should be envisaged differently. Some other correlations have been shown below.   0.05.010.015.020.025.030.035.00.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0P 2O 5(%)SiO2 (%)SiO2 vs P2O5 content variation FeedpH 2.5, 20minpH 3.5, 20minpH 10.5, 20minpH 2.5, 10minNatural pH, 20minNatural pH, 10minNatural pH, re-scrubbed: 20min59   Figure 6.6 Scatterplot for silica vs alumina grade variation The silica and alumina scatter plot shows a linear relation with a positive slope which is in accordance with the SEM results that also showed the occurrence of these phases together as alumino-silicates.  The plot overleaf in Figure 6.7, shows the variation of phosphate with alumina, and it follows a trend similar to the one seen between phosphate and silica in Figure 6.5. This again is in accordance as silica and alumina is seen to occur together as alumino-silicates which is seen both from the SEM results and also in their values for the respective assays. Figure 6.8 shows the plot between phosphate variation with regards to organic carbon. It is clear that there is no linear correlation that can be drawn from the plot and the points follow no distinct pattern.  60  Figure 6.7 Scatterplot for phosphate vs alumina grade variation  Figure 6.8 Scatterplot for carbon vs phosphate grade variation 0.05.010.015.020.025.030.035.00 1 2 3 4 5 6 7 8P2O5(%)C (%)C vs P2O5 content variation FeedpH 2.5, 20minpH 3.5, 20minpH 10.5, 20minpH 2.5, 10minNatural pH, 20minNatural pH, 10minNatural pH, re-scrubbed: 20min61  Figure 6.9 BSE image for the (-0.600 + 0.075mm) scrubbing product (above), and overlay of BSE, Al, Si, P (below) 62 The image on the top in Figure 6.9 is the backscattered image while the image on the bottom shows the overlay of the backscattered image with the element maps for phosphorous, aluminum and silicon. It shows the various phases: the clean and liberated phosphate, the silica and the clays. The high grade of phosphate from the assay results for this size fraction ( -0.600 + 0.075mm) seems to reflect in the SEM images as they show a large population of the grains well defined and distinct without the siliceous or clayey inclusions. Figure 6.10 overleaf shows the finer sized particles (-0.075mm) and represents, once more, the well liberated nature of the phosphate phase occurring independently from the clay and silica in the sample. 63  Figure 6.10 Overlay of SE and P for the -0.075mm scrubbing product (above), and overlay of SE, Al, Si(below) 64  Thus, the current work showed that the ore is amenable to mechanical treatment and carful pre-treatment rather than indiscriminate crushing is the suitable method to selectively screen the high-grade phosphate fraction after attrition scrubbing and treating only the fines with flotation. 65 Chapter 7: Conclusion • The aggregate nature of coarse ore particles suggested a physical processing approach in which the aggregates are disintegration into individual mineral particles while preserving their original, relatively coarse, sizes. The application of attrition-scrubbing – a technique widely used in the phosphate industry – resulted in very promising separation results.       • The occurrence of the main components of the high-carbon phosphate ore and their liberation characteristics were qualitatively analyzed using SEM-EDX and chemical assays by size fraction.  • The SEM images provided useful information about the degree of association of the different elements and minerals within the sample. The presence of silicon along with potassium and aluminum shows the presence of quartz and aluminosilicates along mineral boundaries and in between grains potentially as a cementing material. • The SEM scans also show that carbon occurs as separate well-defined carbon-rich domains, as well as within the phosphorous areas as inclusions. This shows that only a certain amount of the carbon can be removed from the material while the fraction that occurs as inclusions cannot be separated using purely mechanical means. Although the suspected hydrocarbon nature of these carbon inclusions could not be successfully confirmed by subsequent infra-red scanning, the SEM results demonstrate that the carbon phases in the ore are not associated with carbonate minerals. • There was a marked increase in the phosphate grade in the (-600 + 75micron) range for the attrition scrubbing product. It was up from about 22% in the feed to about 32% in all the product for the various tests performed. This result was noticed consistently under all testing conditions, i.e., varying pH values and also varying attrition scrubbing times. The clean product also shows the direct correlation with the observations from the scanning 66 electron microscopy (SEM) images where the phosphorous grains appeared to be in that size range. This tends to confirm the aggregate nature of the material where attrition helps in breaking up these grains and liberating them by the scrubbing and abrasion rather than external force. These are confirmed by analysing the scrubbing products with SEM again and the phosphorous is largely seen to occur as well liberated grains, along with the silica and quartz. The results also demonstrate that, using purely mechanical means; the carbon presence is not a major issue in obtaining a high-grade product. • Re-scrubbing of the product suggested that there was no further benefit in prolonged scrubbing beyond a certain time limit. At that point re-crushing of the oversize material should be considered after screening out the higher P2O5 grade undersize fractions to prevent over-crushing and P2O5 losses to fines.  • The crushing products still exhibited a relatively higher grade of phosphate (about 26% P2O5) in the (-600 + 75micron) range compared to the feed grade, which shows that further liberation of high P2O5 material is possible although crushing does not offer the same selectivity as does attrition scrubbing. • The tests show the amenability of the material to mechanical treatment (and spontaneous disintegration) rather than to crushing/grinding of the entire feed down to the same size.  • With this approach, only the finest fractions would require processing by froth flotation, while the coarser sizes could be processed by physical methods. 67 Chapter 8: Recommendations for Further Research As seen from the experimental work, attrition scrubbing yielded some positive results. However, specific attrition scrubbing parameters and characteristics like varying the solids content (over 55% as was used in this work), using a coarser top size for the feed etc. could be understood with greater detail. Another important aspect is understanding the chemical composition of the process water used in the attrition scrubbing step and assaying it for organic and inorganic carbon.  Detailed characterization of carbon using infra-red scanning technique in a different way or using more advanced chemical methods can also be employed to be get a better understanding of the material.     68 Bibliography Amirech, A., Bouhenguel, M., & Kouachi, S. (2018). Two-stage reverse flotation process for removal of carbonates and silicates from phosphate ore using anionic and cationic collectors. City, S. L. (2000). Evaluation of Flotation Strategies for Sedimentary Phosphates with siliceous and carbonates gangues. Science, 13(10), 991–998. Gallala, W., Herchi, F., Ali, I. Ben, Abbassi, L., Gaied, M. E., & Montacer, M. (2016). Beneficiation of Phosphate Solid Coarse Waste from Redayef (Gafsa Mining Basin) by Grinding and Flotation Techniques. Procedia Engineering, 138, 85–94. https://doi.org/10.1016/j.proeng.2016.02.065 Gulbrandsen, R. A., & Krier, D. J. (1980). Large and Rich Phosphorus Resources in the Phosphoria Formation in the Soda Springs Area, Southeastern Idaho, 1–24. Retrieved from https://pubs.usgs.gov/bul/1496/report.pdf Keil, R. G., & Mayer, L. M. (2013). Mineral Matrices and Organic Matter. Treatise on Geochemistry: Second Edition (2nd ed., Vol. 12). Elsevier Ltd. https://doi.org/10.1016/B978-0-08-095975-7.01024-X McClellan, G. H., & Despujol, B. (1999). Preliminary Characterization of Organic Coatings on Phosphate Ores. In P. Zhang, H. El-Shall, & R. Wiegl (Eds.), Beneficiation of Phosphates ADVANCES IN RESEARCH AND PRACTICE (pp. 261–268). SME. Nanthakumar, B., Grimm, D., & Pawlik, M. (2009). Anionic flotation of high-iron phosphate ores-Control of process water chemistry and depression of iron minerals by starch and guar gum. International Journal of Mineral Processing, 92(1–2), 49–57. https://doi.org/10.1016/j.minpro.2009.02.003 Stuart, B. (2004). Infrared Spectroscopy: Fundamentals and Applications. John Wiley and Sons 69 Ltd. U. S. Geological Survey. (2002, September). Western Phosphate Field , U . S . A .: Science in Support of Land Management, (September). Vandenbroucke, M., & Largeau, C. (2007). Kerogen origin, evolution and structure. Organic Geochemistry, 38(5), 719–833. https://doi.org/10.1016/j.orggeochem.2007.01.001 Wen Qi, G., Parentich, A., Little, L. H., & Warren, L. J. (1992). Selective flotation of apatite from iron oxides. International Journal of Mineral Processing, 34(1–2), 83–102. https://doi.org/10.1016/0301-7516(92)90017-Q Willis, R. D., Blanchard, F. T., & Connor, T. L. (2002). Guidelines for the Application of SEM / EDX Analytical Techniques to Particulate Matter Samples. Exposure, (November), 1–3. Zhang, P., Khalek, N. A., El-Shall, H., & El-Mofty, S. (2003). Beneficiation technology of phosphates: challenges and solutions. SME Annual Meeting, Feb 24-26, 2003, 1–12. Zhang, P., Yu, Y., & Bogan, M. (1997). CHALLENGING THE “CRAGO” DOUBLE FLOAT PROCESS II. AMINE-FATTY ACID FLOTATION OF SILICEOUS PHOSPHATES, 10(9), 983–994. https://doi.org/10.1016/S0892-6875(97)00078-2 Zidi, R., Babbou-Abdelmalek, C., Chaabani, F., & Abbassi, L. (2016). Enrichment of low-grade phosphate coarse particles by froth-flotation process, at the Kef-Eddur washing plant, Tunisia. Arabian Journal of Geosciences, 9(6). https://doi.org/10.1007/s12517-016-2495-6 70 Appendices Appendix A  Infra-Red (FTIR) Spectroscopy The infrared (IR) spectroscopy is an analytical method used to characterize the bonding structure of atoms based on the interaction of the IR radiation with matter and measures the frequencies of the radiation at which the substance absorbs, which leads to vibrations, stretching, and bending of bonds in molecules. IR spectroscopy provides a fast technique of identification and characterization of chemical structures to obtain information from biological to composite materials, from liquids to gaseous samples (Ţucureanu, Matei, & Avram, 2016). IR spectroscopy combines an old tool, the interferometer (developed by Albert Michelson in 1877) with an even older mathematical principle, the Fourier transform, to convert the output from an interferometer (interferogram) to a spectrum, all of which is greatly facilitated and fully automated today using computers. Using the computing capacity available, an interferogram is transformed instantaneously into a spectrum and a modern software algorithm allows the use of FTIR spectroscopy as a tool for qualitative and quantitative analysis. Generally, a FTIR spectrum is a graphical representation of the transmittance, in percent (T%) or absorbance, in units (A) versus IR frequency in terms of wavenumber (cm-1). In an IR spectrum the absorption bands are characterized by a wavenumber at which absorption occurs (corresponding to specific chemical bonds or chemical groups), and by the intensity of absorption that is proportional to the concentration of the substance in the sample (Ţucureanu et al., 2016).  It is a non-destructive and real-time measurement analytical method, enabling the identification of unknown materials (qualitative determination) and their concentration (quantitative determination) from organic and inorganic substances, from solid, liquid or gas samples. In some molecules during vibration, there is a change of the electric dipole moment. For the IR active substances, the absorption of the radiation corresponds to a change of the dipole moment. For IR inactive 71 substances the electric dipole moment is zero, no matter how long the bond is in the molecule (IR-active: polar bonds, asymmetrical molecules and IR inactive: non-polar bond, symmetrical molecule) (Krishna, Muthukumaran, Krishnamoorthy, & Nishat, 2013). In IR spectrometry, each chemical bond has a specific vibration frequency corresponding to an energy level. The energy is given by the equation: E=hv=hc/! where; h=Planck’s constant; v= wave number and c=speed of light. The IR absorption appears when radiant energy corresponds to energy of a molecular vibration. The vibrational mode of a molecule can involve a variation in inter-atomic bond length (stretching, v) or bond angle between two bonds (bending or deformation, d) (Stuart, 2004). A.1 Data acquisition and interpretation The IR maps were acquired at the Mining Engineering department at the University of British Columbia using a Thermo Scientific Nicolet iS50 spectrometer along with the OMNIC software package. The scans were performed on the same polished particles as the ones scanned with the SEM. Spots were chosen on the sample surface so as to represent all the characteristic areas- grains, grain boundaries, cementing materials etc. 72   IR spectra map for point#4 The image on the top shows the sample surface taken in reflected light using the optical microscope showing the background features with the testing points. The spectrum below it shows the IR scan 73 for point#4 on the sample surface. Point#4 in the image above shows a point in the matrix within the ‘dark areas’ of a given particle. The characteristic peak occurs at about 1050cm-1. Another example is shown overleaf for a different point on the same sample surface.  IR spectra map for point#5 74 Point#5 on the surface of the same sample is chosen on a high-phosphorous area (or a phosphate grain based on information obtained from SEM imaging). It shows the same characteristic peak around 1057cm-1. The scans did not reveal any bands and peaks characteristics of aliphatic hydrocarbon groups that could be associated with hydrocarbon-type organic matter. To notice similar trends, more points on the surface were analyzed. By repeatedly collecting data for a number of representative points, wavenumbers could be identified which were characteristic for a given type of feature (light area, dark area, interparticle spaces, etc.). The characteristic wavenumbers correlated to the clay phases in the sample but there was inconclusive evidence for the presence of organic or any areas with a characteristically high carbon concentration. Some additional scans have been added to the appendix of this work. 75 Appendix B  Additional FT-IR Maps Here are some more examples of Infra-Red maps on various sample surfaces.  76     77 Appendix C  Single particle and batch tests Effect of Sulfuric Acid, Acetone and Acetic Acid Both the light and the dark particles were individually subjected to similar test conditions as done previously with highly alkaline and organic solutions. A sample each of the light and dark phases were left in concentrated sulphuric acid (98%) and in acetone for 24 hours. They were then put in an ultrasonic bath for 60 minutes to notice any changes that may occur to their state.  To note the effect of temperature, the samples would be subject to elevated temperatures and constant vibratory motion on a shaking table. The same procedure would then be repeated for the ‘light’ phase of the sample space.  Next, the sample vials were put on a shaker and the temperature settings were sequentially incremented one after the other. The results (NTU values) were as follows: Stage1:   Stage2:    Stage3: RPM-250   RPM-250    RPM-250 Temperature-25°C  Temperature-30°C   Temperature-35°C Time-60minutes  Time-60minutes   Time-60minutes     78 Test Conditions  Ultrasonic Bath Stage-1 Stage-2 Stage-3 Sulphuric Acid (98%) Dark 4000 5875 2504 3000 Sulphuric Acid (98%) Light 157 146 155 170 Acetone Dark 690 298 322 296 Acetone Light 110 64.5 149 137  The images are in the following order: 1. After 60 minutes in the Ultrasonic Bath 2. After shaker Stage-1 3. After shaker Stage-2 4. After shaker Stage-3 79      Dark Particle in Sulphuric Acid  Light Particle in Sulphuric Acid  80      Dark Particle in Acetone  Light Particle in Acetone  81                              Observations and Discussion:  1. Dark particle in Sulphuric Acid: The particle dissociates to a great extent in the sulphuric acid solution. They show the highest values of turbidity as well. The resulting solution was also found to be the most viscous and ‘oily’. 2. Light particle in Sulphuric Acid: The particle dissociates to a great extent as well, but the turbidity values are rather (strangely) low not supplementing the resulting color of the solutions. The resulting solution is not as ‘oily’ as the one with the dark particle. 3. Dark particle in Acetone: The particles dissociate to a certain extent which is different than what was observed in the other organic solvents used. The turbidity values are high compared to the other solvents used previously. 4. Light particle in Acetone: The particle does not dissociate as much as the dark particle in acetone and no significant change in turbidity was observed either. Effect of Sulphuric acid (pH 2 and pH 3) and of Acetic acid: Particle weights at the time of immersion: 1. Sulphuric Acid pH 2:  Dark particle – 7.8gm Light particle – 1.70gm  pH 3:  Dark particle – 6.84gm Light particle – 2.74gm  2. Acetic Acid Dark particle – 3.45gm Light particle – 0.53gm  82 Test Conditions  Ultrasonic Bath Stage-1 Stage-2 Stage-3 Sulphuric Acid (pH-2) Dark 774 346 655 714 Sulphuric Acid (pH-2) Light 1228 1994 2582 3257 Sulphuric Acid (pH-3) Dark 271 89.1 50.3 135 Sulphuric Acid (pH-3) Light 233 653 1128 1310 Acetic Acid Dark 230 207 203 204 Acetic Acid Light 103 115 137 136  The images are in the following order: 1. After standing in solution for a day 2. After 60 minutes in the Ultrasonic Bath 3. After shaker Stage-1 4. After shaker Stage-2 5. After shaker Stage-3    83    Dark Particle in Sulphuric Acid (pH-2)   Dark Particle in Sulphuric Acid (pH-3) )   84    Light Particle in Sulphuric Acid (pH-2) )   Light Particle in Sulphuric Acid (pH-3) )   85     Dark Particle in Acetic Acid )   Light Particle in Acetic Acid )   86 Effect of Sulfuric Acid on a sample batch To study the effect of acid conditioning on a batch of the sample, the light and the dark particles were subjected to similar test conditions. A high solids ratio (greater than 33% by weight) was chosen to observe the effect of acid in conjunction with the rubbing/abrasion due to particle-particle interaction.   Two acid concentrations of pH-2 and pH 2.4 were chosen but the overnight conditioning and ultrasonic bath treatment was skipped. They were directly put through the various stages in the shaker. The solids content was kept in the 20-30% range for these tests.  To note the effect of temperature, the samples were also subjected to successive stages of elevated temperatures. The results (NTU values) of the supernatant were as follows: Stage1:   Stage2:    Stage3: RPM-250   RPM-250    RPM-250 Temperature-25°C  Temperature-30°C   Temperature-35°C Time-60minutes  Time-60minutes   Time-60minutes     87 Test Conditions  Stage-1 Stage-2 Stage-3 Sulphuric Acid (pH-2) Dark 1310 9580 7302 Sulphuric Acid (pH-2) Light Out of range  8610 7295 Sulphuric Acid (pH-2.4) Dark 6870 Out of range 5186 Sulphuric Acid (pH-2.4) Light 3362 8494 Out of range  The images are in the following order: 1st image: After shaker Stage-2 2nd image: After shaker Stage-3     88   “Dark Particles” in pH-2 solution “Light Particles” in pH-2 solution 89   “Light Particles” in pH-2.4 solution “Dark Particles” in pH-2.4 solution 90   Observations:  The turbidity values were quite high for all the samples observed. This test regime has yielded the most positive results done so far as there is definitely an amount of inter-particle rubbing and abrasion that is causing them to break apart and disintegrate which was not seen in the individual particle tests. Acidic conditions are also seen to be far better suited for conditioning compared to a high pH regime.  Effect of furnace heating on crushed sample Crushed sample of the ore (-300microns) was subject to furnace heating to observe the changes in weight and attribute them to the various components present in the sample. The sample was taken in four crucibles and each of them was placed in the furnace for a different amount of time.  The crucibles were taken out at the following temperatures: 1. 300 degrees Celsius 2. 500 degrees Celsius 3. 700 degrees Celsius 4. 950 degrees Celsius  The change in the weight was noted and then changed to specific weight loss (weight 91 per gram of sample). Once the specific weight loss for the different temperatures were calculated, the weight was subtracted from the previously obtained weight to get the weight loss for that specific stage. The table depicting the same is shown below:  Temperature © Time Taken (minutes) Initial Weight (gm)  Final Weight (gm) Weight change (gm) Weight change/gm of sample (gm/gm) or Specific weight change Successive weight change due to different phases (gm/gm) 1 300 9 2.2027 2.1955 0.0072 0.00327 0.00327 2 500 26 2.616 2.6045 0.0115 0.00439 0.00113 3 700 55 4.513 4.2086 0.3044 0.06745 0.06305 4 950 135 5.624 5.0489 0.5751 0.10226 0.03481   The data can be digested with the following explanation: a) Since we are observing the weight change/gm of sample, they are comparable between the different stages. b) The difference between the specific weight changes can be attributed to the changes occurring in the sample, both physically and chemically and hence give an idea about the overall state the sample is in. c) The first change can be attributed to the loss in the moisture content. d) The weight loss between 300 to 500 degrees Celsius can be attributed to the formation of carbon dioxide. e) The changes beyond that temperature are due to the phase changes within the sample.  92 Appendix D  Particle Size Distributions of the Attrition Scrubbing Products The various particle size distribution tables for the scrubbing products are as below:  Test#1. PSD for the feed   2. PSD for pH2.5, scrubbed for 20-minutes   3. PSD for pH3.5, scrubbed for 20-minutes   +9.51 9.51 19.3 3.1 3.1 96.9-9.51+6.7 6.70 69.5 11.1 14.2 85.8-6.7+4.75 4.75 123.9 19.8 34.0 66.0-4.75 + 3.35 3.35 82.5 13.2 47.3 52.7-3.35 + 2.36 2.36 65.4 10.5 57.7 42.3-2.36 +1.70 1.70 44.1 7.1 64.8 35.2-1.70 + 0.850 0.85 64.9 10.4 75.2 24.8-0.850 + 0.600 0.60 25.4 4.1 79.3 20.7-0.600 + 0.300 0.30 42.7 6.8 86.1 13.9-0.300 0.30 86.7 13.9 100.0 0.0Total 624.39 100.0Cummulative Passing [%]Size Fractions Nominal Size Weight [g] Weight [%]Cummulative Retained [%]+9.51 9.51 3.3 0.7 0.7 99.3-9.51+6.7 6.70 22.4 4.4 5.1 94.9-6.7+4.75 4.75 67.6 13.4 18.5 81.5-4.75+3.36 3.36 57.4 11.4 29.8 70.2-3.36+2.36 2.36 38.7 7.7 37.5 62.5-2.36+1.18 1.18 31.4 6.2 43.7 56.3-1.18+0.600 0.60 17.1 3.4 47.1 52.9-0.600+0.212 0.21 25.5 5.0 52.1 47.9-0.212+0.075 0.08 42.1 8.3 60.5 39.5-0.075 0.08 199.6 39.5 100.0 0.0Total 505.1 100.0Cummulative Passing [%]Size Fractions [mm] Nominal Size Weight [g] Weight [%]Cummulative Retained [%]+9.51 9.51 15.3 3.3 3.3 96.7-9.51+6.7 6.70 30.6 6.6 9.9 90.1-6.7+4.75 4.75 56.1 12.1 22.1 77.9-4.75+3.36 3.36 47.5 10.3 32.3 67.7-3.36+2.35 2.35 29.6 6.4 38.7 61.3-2.35+1.68 1.68 20.7 4.5 43.2 56.8-1.68+1.18 1.18 10.4 2.2 45.5 54.5-1.18+0.600 0.60 14.5 3.1 48.6 51.4-0.600+0.212 0.21 22.2 4.8 53.4 46.6-0.212+0.075 0.08 36.9 8.0 61.4 38.6-0.075 0.08 178.5 38.6 100.0 0.0Total 462.3 100.0Cummulative Passing [%]Size Fractions [mm] Nominal Size Weight [g] Weight [%]Cummulative Retained [%]93 4. PSD for pH10.5, scrubbed for 20-minutes   5. PSD for pH2.5, scrubbed for 10-minutes   6. PSD for natural pH (pH 6), scrubbed for 20-minutes   +9.51 9.51 3.5 0.9 0.9 99.1-9.51+6.7 6.70 19.0 4.7 5.6 94.4-6.7+4.75 4.75 48.7 12.0 17.6 82.4-4.75+3.36 3.36 42.7 10.6 28.2 71.8-3.36+2.35 2.35 24.5 6.1 34.2 65.8-2.35+1.68 1.68 16.2 4.0 38.2 61.8-1.68+1.18 1.18 8.8 2.2 40.4 59.6-1.18+0.600 0.60 12.9 3.2 43.6 56.4-0.600+0.212 0.21 21 5.2 48.8 51.2-0.212+0.075 0.08 39.7 9.8 58.6 41.4-0.075 0.08 167.5 41.4 100.0 0.0Total 404.5 100.0Cummulative Passing [%]Size Fractions [mm] Nominal Size Weight [g] Weight [%]Cummulative Retained [%]+9.51 9.51 20.6 3.8 3.8 96.2-9.51+6.7 6.70 48.4 8.8 12.6 87.4-6.7+3.35 3.35 132.3 24.1 36.7 63.3-3.35+2.36 2.36 43.6 7.9 44.6 55.4-2.36+1.18 1.18 36.6 6.7 51.3 48.7-1.18+0.850 0.85 9.3 1.7 53.0 47.0-0.850+0.600 0.60 10.6 1.9 54.9 45.1-0.600+0.425 0.43 11.3 2.1 57.0 43.0-0.425+0.212 0.21 21.9 4.0 61.0 39.0-0.212+0.106 0.11 32.2 5.9 66.8 33.2-0.106+0.075 0.08 15.9 2.9 69.7 30.3-0.075 0.08 166 30.3 100.0 0.0Total 548.7 100.0Cummulative Passing [%]Size Fractions [mm] Nominal Size Weight [g] Weight [%]Cummulative Retained [%]+9.51 9.51 14.6 3.4 3.4 96.6-9.51+6.7 6.70 40.6 9.4 12.8 87.2-6.7+4.75 4.75 60.5 14.1 26.9 73.1-4.75+3.35 3.35 41.1 9.5 36.4 63.6-3.35+2.36 2.36 26.4 6.1 42.5 57.5-2.36+1.68 1.68 19.9 4.6 47.2 52.8-1.68+1.18 1.18 9.3 2.2 49.3 50.7-1.18+0.600 0.60 15.9 3.7 53.0 47.0-0.600+0.212 0.21 28.5 6.6 59.6 40.4-0.212+0.075 0.08 38.4 8.9 68.6 31.4-0.075+0.045 0.05 19.8 4.6 73.2 26.8-0.045 0.05 115.6 26.8 100.0 0.0Total 430.6 100.0Cummulative Retained [%]Cummulative Passing [%]Size Fractions [mm] Nominal Size Weight [g] Weight [%]94 7. PSD for natural pH (pH 6.1), scrubbed for 10-minutes   8. PSD for re-scrubbing the oversize at natural pH (pH 6.25) for 20-minutes   +9.51 9.51 6.9 1.7 1.7 98.3-9.51+6.7 6.70 43.4 10.5 12.1 87.9-6.7+4.75 4.75 56.7 13.7 25.8 74.2-4.75+3.35 3.35 49.2 11.9 37.7 62.3-3.35+2.36 2.36 28 6.8 44.5 55.5-2.36+1.68 1.68 15.2 3.7 48.2 51.8-1.68+1.18 1.18 8.9 2.1 50.3 49.7-1.18+0.600 0.60 11.3 2.7 53.0 47.0-0.600+0.212 0.21 21.2 5.1 58.2 41.8-0.212+0.075 0.08 32.3 7.8 66.0 34.0-0.075+0.045 0.05 20.9 5.0 71.0 29.0-0.045 0.05 120.1 29.0 100.0 0.0Total 414.1 100.0Weight [%] Cummulative Retained [%]Cummulative Passing [%]Size Fractions [mm] Nominal Size Weight [g]+9.51 9.51 0 0.0 0.0 100.0-9.51+6.7 6.70 14.8 5.7 5.7 94.3-6.7+4.75 4.75 52.6 20.2 25.9 74.1-4.75+3.35 3.35 65 25.0 50.8 49.2-3.35+2.36 2.36 47 18.0 68.9 31.1-2.36+1.68 1.68 14.6 5.6 74.5 25.5-1.68+1.18 1.18 5.1 2.0 76.4 23.6-1.18+0.600 0.60 2.5 1.0 77.4 22.6-0.600+0.212 0.21 2.6 1.0 78.4 21.6-0.212+0.075 0.08 2.4 0.9 79.3 20.7-0.075+0.045 0.05 3.5 1.3 80.7 19.3-0.045 0.05 50.4 19.3 100.0 0.0Total 260.5 100.0Cummulative Retained [%]Cummulative Passing [%]Size Fractions [mm] Nominal Size Weight [g] Weight [%]95 Appendix E  Assay Result for the various Attrition Scrubbing Products   C-IR06a ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06SAMPLE C organic SiO2 Al2O3 Fe2O3 CaO MgO Na2O K2O Cr2O3 TiO2 MnO P2O5 SrO BaO LOI TotalDESCRIPTION % % % % % % % % % % % % % % % %Test#5 +2.4mm 5.44 20.81 3.37 1.25 33.82 1.46 0.36 0.89 0.11 0.22 0.01 22.612 0.08 0.03 13.45 98.47Test#5 2.4 + 0.600mm 5.65 21.02 3.38 1.41 33.74 0.99 0.37 0.95 0.11 0.22 0.01 23.444 0.08 0.02 12.4 98.14Test#5 -0.600 + 0.075mm 4.73 10.72 1.64 0.89 43 0.34 0.33 0.45 0.07 0.11 0.01 30.597 0.11 0.03 9.94 98.24Test#5 -0.075mm 4.32 41.62 7 2.83 18.58 0.72 0.44 1.93 0.21 0.46 0.02 12.779 0.05 0.03 12 98.67Test#6 +2.4mm 5.64 25.21 3.7 1.31 31.46 0.81 0.39 1.08 0.1 0.25 0.01 22.167 0.08 0.03 11.95 98.55Test#6 2.4 + 0.600mm 5.74 22.21 3.52 1.52 32.89 0.93 0.39 1 0.11 0.24 0.03 22.861 0.08 0.03 12.45 98.26Test#6 -0.600 + 0.075mm 4.82 10.98 1.66 0.89 42.68 0.36 0.33 0.45 0.07 0.12 0.02 30.384 0.11 0.03 10.2 98.28Test#6 -0.075mm 4.37 40.02 6.7 2.74 19.7 0.71 0.43 1.82 0.21 0.43 0.03 13.317 0.05 0.03 12.05 98.24Test#7 +2.4mm 5.49 26.1 3.96 1.58 30.36 0.96 0.41 1.08 0.11 0.27 0.01 20.979 0.07 0.03 12.2 98.12Test#7 2.4 + 0.600mm 5.58 20.46 3.21 1.38 34.34 1.05 0.37 0.9 0.11 0.22 0.01 23.716 0.09 0.03 12.4 98.29Test#7 -0.600 + 0.075mm 4.49 8.62 1.26 0.79 44.95 0.31 0.32 0.34 0.06 0.09 0.01 31.976 0.11 0.02 9.5 98.36Test#7 -0.075mm 4.86 38.02 6.33 2.65 21.15 0.73 0.45 1.73 0.19 0.41 0.02 14.424 0.06 0.03 12.4 98.59Test#8 +2.4mm 5.62 20.34 3.21 1.13 35.04 1.12 0.4 0.9 0.1 0.21 0.01 24.014 0.08 0.03 12 98.58Test#8 2.4 + 0.600mm 6.27 25.09 3.99 1.94 30.04 1.07 0.4 1.11 0.13 0.26 0.01 20.833 0.08 0.03 13.45 98.43Test#8 -0.600 + 0.075mm 5.62 17.64 2.76 1.79 36.77 0.58 0.37 0.76 0.1 0.18 0.01 26.004 0.1 0.03 11.65 98.74Test#8 -0.075mm 7.44 35.13 5.95 2.2 21.93 0.79 0.46 1.66 0.19 0.39 0.01 15.278 0.07 0.03 14.95 99.0496 E.1 Assay Results for the Attrition followed by crushing tests:         ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 ME-XRF06 C-IR06aSAMPLE SiO2 Al2O3 Fe2O3 CaO MgO Na2O K2O Cr2O3 TiO2 MnO P2O5 SrO BaO LOI Total C organicDESCRIPTION % % % % % % % % % % % % % % % %Test#1 Scrubbing Product -2.4 + 0.600mm 21.42 3.33 1.55 33.77 1.33 0.4 0.96 0.11 0.22 0.02 23.125 0.09 0.04 12.1 98.46 5.8Test#1 Scrubbing Product -0.600 + 0.075mm 7.69 1.11 0.75 46.37 0.33 0.34 0.29 0.05 0.07 0.01 33.053 0.12 0.03 8.58 98.79 4.6Test#1 Scrubbing Product -0.075mm 37.47 6.43 2.55 21.64 0.77 0.46 1.76 0.2 0.42 0.01 14.831 0.06 0.04 11.7 98.34 5.15Test#1 Crushing Product -2.4 + 0.600mm 21.53 3.38 1.25 34.26 1.22 0.38 0.94 0.11 0.23 <0.01 23.633 0.09 0.03 11.4 98.45 5.73Test#1 Crushing Product -0.600 + 0.075mm 18.05 2.85 1.16 37.07 0.97 0.35 0.79 0.1 0.19 <0.01 25.919 0.09 0.03 10.95 98.52 5.64Test#1 Crushing Product -0.075mm 29.58 4.61 1.68 28.06 1.23 0.46 1.23 0.14 0.32 <0.01 19.245 0.07 0.03 11.7 98.36 5.86Test#2 Scrubbing Product -2.4 + 0.600mm 21.3 3.39 1.29 34.38 0.99 0.39 0.94 0.11 0.23 <0.01 23.947 0.08 0.03 11.4 98.48 5.86Test#1 Scrubbing Product -0.600 + 0.075mm 8.38 1.26 0.74 45.67 0.31 0.32 0.33 0.06 0.08 0.01 32.534 0.11 0.03 8.97 98.8 4.4Test#1 Scrubbing Product -0.075 + 0.045mm 37.45 6.86 2.81 21.15 0.73 0.44 1.83 0.21 0.41 0.02 14.464 0.06 0.03 12.15 98.61 4.93Test#1 Scrubbing Product -0.045mm 44.79 7.02 2.67 17.46 0.76 0.52 1.93 0.2 0.51 0.01 11.546 0.05 0.04 10.8 98.31 4.33Test#1 Crushing Product -1.8 + 0.600mm 23.66 3.8 1.36 32.09 1.29 0.4 1.07 0.12 0.25 <0.01 21.932 0.08 0.03 12.35 98.43 5.94Test#1 Crushing Product -0.600 + 0.075mm 21.05 3.41 1.25 34.22 1.01 0.38 0.97 0.11 0.22 <0.01 23.74 0.08 0.03 11.9 98.37 5.99Test#1 Crushing Product  -0.075mm 32.84 5.22 1.79 24.88 1.29 0.48 1.46 0.16 0.36 <0.01 16.59 0.06 0.03 12.95 98.11 6.2997 Appendix F  XRD Report for the feed as received from the lab QUANTITATIVE PHASE ANALYSIS OF POWDER SAMPLE USING THE RIETVELD METHOD AND X-RAY POWDER DIFFRACTION DATA.   Abhay Sablok - Marek Pawlik Mining Engineering Dept. – UBC 5th Floor, 6350 Stores Road Vancouver, BC V6T 1Z4    Jacob Kabel, B.Sc. Elisabetta Pani, Ph.D. Edith Czech, M.Sc. Jenny Lai, B.Sc. Lan Kato, B.A.   Dept. of Earth, Ocean & Atmospheric Sciences The University of British Columbia  6339 Stores Road Vancouver, BC  V6T 1Z4   November 29, 2018 98 EXPERIMENTAL METHOD The sample was reduced to the optimum grain-size range for quantitative X-ray analysis (<10 µm) by grinding under ethanol in a vibratory McCrone Micronizing Mill for 7 minutes.  Continuous -scan X-ray powder-diffraction data was collected over a range 3-80°2q with CoKa radiation on a Bruker D8 Advance Bragg-Brentano diffractometer equipped with an Fe filter foil, 0.6 mm (0.3°) divergence slit, incident- and diffracted-beam Soller slits and a LynxEye-XE detector. The long fine-focus Co X-ray tube was operated at 35 kV and 40 mA, using a take-off angle of 6°.     RESULTS The X-ray diffractogram was analyzed using the International Centre for Diffraction Database PDF-4 and Search-Match software by Bruker. X-ray powder-diffraction data of the sample was refined with Rietveld program Topas 4.2 (Bruker AXS). The result of quantitative phase analysis by Rietveld refinement is given in Table 1. These amounts represent the relative amounts of crystalline phases normalized to 100%. The Rietveld refinement plot is shown in Figure 1.   99 Table 1. Results of quantitative phase analysis (wt.%) Mineral Ideal Formula Phosphate Ore  Calcite CaCO3 1.1 Dolomite CaMg(CO3)2 3.9 Fluorapatite Ca5(PO4)3F 54.3 Gypsum CaSO4·2H2O 0.8 Illite-Muscovite 2M1 K0.65Al2.0Al0.65Si3.35O10(OH)2- KAl2AlSi3O10(OH)2 6.1 K-feldspar (Microcline, Orthoclase) KAlSi3O8 9.9 Plagioclase (Albite) NaAlSi3O8 – CaAl2Si2O8 2.9 Pyrite FeS2 0.6 Quartz SiO2 20.4 Total  100.0  100   Figure 1. Rietveld refinement plot of sample Phosphate Ore (blue line - observed intensity at each step; red line - calculated pattern; solid grey line below –  difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections). Coloured lines are individual diffraction patterns of all phases.  1AS_PhosphateOre.raw2Th Degrees75706560555045403530252015105Sqrt(Counts)150100500Quartz low 20.43 %Fluorapatite 54.28 %Dolomite 3.92 %Illite/Muscovite 2M1 6.10 %Calcite 1.09 %Gypsum 0.82 %Albite low 2.92 %Orthoclase 3.34 %Pyrite 0.56 %Microcline intermediate 6.55 %

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0376192/manifest

Comment

Related Items