Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Metal recovery from medium temperature pressure leach residues Ward, Tim 2015

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

Item Metadata

Download

Media
24-ubc_2015_november_ward_tim.pdf [ 10.2MB ]
Metadata
JSON: 24-1.0166767.json
JSON-LD: 24-1.0166767-ld.json
RDF/XML (Pretty): 24-1.0166767-rdf.xml
RDF/JSON: 24-1.0166767-rdf.json
Turtle: 24-1.0166767-turtle.txt
N-Triples: 24-1.0166767-rdf-ntriples.txt
Original Record: 24-1.0166767-source.json
Full Text
24-1.0166767-fulltext.txt
Citation
24-1.0166767.ris

Full Text

Metal recovery from medium temperature pressure leach residues by  Tim Ward  M.A.Sc, The University of British Columbia, 2015  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 (Materials Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   October 2015  © Tim Ward, 2015   ii Abstract Residues of the CESL Cu and Vale Ni sulphide medium temperature pressure leach contain elevated levels of the interest Cu and Ni metals. Metal losses are disproportionately found within a poorly formed amorphous and semi-crystalline iron oxide phase analogous to ferrihydrite. This phase results from incomplete ferric hydrolysis and precipitation that occurs simultaneously with the leaching of the sulphide minerals. An investigation was conducted by ageing the residues at 95°C in water to assess the extraction of lost Cu and Ni and the transformation of ferrihydrite to the more crystalline goethite and hematite species. It was found that 22% of the Cu could be removed from the CESL residue in 24 hours. Repeated ageing was only capable of extracting 38% of the Cu. Extractions of 22% of the Ni and 6% of the Cu were achieved in 24 hours from the Vale residue. Recovery of the lost Cu and Ni was mainly attributed to the washing of entrained and adsorbed Cu and Ni sulphates evenly distributed throughout the residues. A small amount of Cu and Ni was gained from ferrihydrite dissolution during the ageing process. Further investigations by SEM-EDS revealed the presence of unleached Cu and Ni sulphides that also contribute notable metal losses. EMP analysis showed that all hematite precipitated in the Vale residue contained Cu, Ni and SO42−. The average hematite particle consisted of 66.1% Fe, 0.85% Ni, 0.73% Cu and 1.19% SO42− for an equivalent of 30% Ni and 25% Cu losses to the residue. QXRD analysis of the aged residue showed little evidence of ferrihydrite transformation though ageing stabilised the residues by reducing the amount of X-ray amorphous material present. Improved stability was confirmed by a reduction in mass loss following treatment by a sequential extraction procedure utilising acidified hydroxylamine hydrochloride. This treatment dissolves the ferrihydrite present in the residues and is used as a proxy to assess residue stability and losses of Cu and Ni to the non-crystalline iron oxide phase.    iii Preface The thesis that follows is my contribution to a larger project within the Hydrometallurgy group in the Materials Engineering Department at UBC. This larger project investigates medium temperature pressure leach residues in conjunction with industry sponsors. The UBC project team consists of myself, two PhD students, Baseer Abdul and Tasawar Javed, and is supervised by Associate Professor Edouard Asselin. Edouard Asselin devised the basis for my research into recovering metals from these residues by a simple ageing process as this was considered a gap in the knowledge of the project. The design and optimisation of the experimental and research programme was devised by myself with Ed’s occasional input and support. The performance and accuracy of the experimental results are purely my responsibility as I conducted the experiments and chose the samples. The chemical analysis was conducted by others utilising established and robust analytical procedures that are accurate and repeatable. Processing of the analytical data and results was conducted by myself and presented according to my preference. Assays of the original residues were provided to me by the project team prior to commencing my research and were used throughout my experimental programme. I am proud of my work and believe it is a true and fair reflection of my effort and ability towards fulfilling the requirements of a Masters of Applied Science in Materials Engineering from UBC.      iv Table of Contents  Abstract .......................................................................................................................................... ii!Preface ........................................................................................................................................... iii!Table of Contents ......................................................................................................................... iv!List of Tables .............................................................................................................................. viii!List of Figures ................................................................................................................................. x!Acknowledgements ..................................................................................................................... xii!Dedication ................................................................................................................................... xiii!Chapter 1: Introduction ................................................................................................................1!1.1! Project background ............................................................................................................ 1!1.1.1! Project aim .................................................................................................................. 4!1.2! Process conditions for producing CESL and Vale leach residues ..................................... 5!1.2.1! Concentrate producing S1 ........................................................................................... 5!1.2.2! Concentrate producing S3 ........................................................................................... 6!1.2.3! Initial residue analysis ................................................................................................. 7!Chapter 2: Literature review ......................................................................................................12!2.1! Hydrometallurgical treatment of sulphide concentrates .................................................. 12!2.1.1! CESL and VALE process for Cu and Ni sulphide concentrates ............................... 14!2.1.2! CESL copper process ................................................................................................ 14!2.1.3! Vale nickel process ................................................................................................... 17!2.2! Iron control in CESL and Vale pressure leach processes ................................................ 20!2.2.1! Iron hydrolysis .......................................................................................................... 22!  v 2.2.2! Hematite and goethite precipitation .......................................................................... 24!2.2.3! Ferrihydrite in hydrometallurgical residues .............................................................. 26!2.2.4! Metal adsorption to crystalline iron oxides and ferrihydrite ..................................... 28!2.2.5! Metal sulphate adsorption and occlusion to iron oxides ........................................... 31!2.3! Characterisation of the CESL and Vale leach residues ................................................... 33!2.3.1! Identifying ferrihydrite in CESL and Vale leach residues ........................................ 34!2.4! Ageing of the CESL and Vale residues ........................................................................... 37!2.4.1! Transformation of ferrihydrite .................................................................................. 38!2.4.2! Effect of dissolved ions during ferrihydrite ageing .................................................. 42!Chapter 3: Experimental Materials and Methods ....................................................................46!3.1! Residue collection ............................................................................................................ 46!3.2! Sample analysis ................................................................................................................ 47!3.2.1! Sequential extraction ................................................................................................. 47!3.2.2! Quantitative X-ray diffraction analysis ..................................................................... 48!3.2.3! SEM/EDS analysis .................................................................................................... 48!3.2.4! Electron microprobe analysis .................................................................................... 49!3.3! Ageing experiments ......................................................................................................... 49!3.3.1! Preliminary ageing experiments ............................................................................... 49!3.3.2! Kinetics of ageing at 150°C ...................................................................................... 52!3.3.3! Ageing with ferrous displacement ............................................................................ 54!3.3.4! Repeated ageing and extractions ............................................................................... 56!3.3.5! Enhanced atmospheric leach ..................................................................................... 56!3.3.6! Solid transformation study ........................................................................................ 57!  vi Chapter 4: Results and Discussion .............................................................................................58!4.1! QXRD and ICP analysis of aged residues ....................................................................... 58!4.1.1! Residue phase analysis .............................................................................................. 58!4.1.2! Fe phase analysis ....................................................................................................... 60!4.1.3! Stability of aged residues .......................................................................................... 63!4.1.4! Extraction of Cu and Ni ............................................................................................ 65!4.1.4.1! Metal sulphate extraction by ageing .................................................................. 68!4.1.4.2! Comparison of ageing to sequential extraction .................................................. 70!4.1.4.3! Ni and Cu associations in aged and sequentially extracted residues ................. 73!4.2! Residue investigation by SEM-EDS and EMPA ............................................................. 75!4.2.1! S1 investigation by scanning electron microscope ................................................... 76!4.2.2! S3 investigation by scanning electron microscope ................................................... 81!4.2.3! S3 investigation by electron microprobe .................................................................. 86!4.3! Cu and Ni extraction with high temperature ageing ........................................................ 88!4.3.1! Residue phase analysis of high temperature versus low temperature ageing ........... 91!4.4! Fe2+ displacement tests .................................................................................................... 94!4.4.1! Cu2+ and Ni2+ desorption with Fe2+ from residues .................................................... 95!4.4.2! Speciation of Cu and Ni losses ................................................................................. 99!4.5! Repeated ageing stages .................................................................................................. 102!4.5.1! Cu extractions at different %solids by ageing ........................................................ 102!4.5.2! Cu extraction after repeated ageing of S1 residue .................................................. 103!4.6! Ageing an alternative CESL residue .............................................................................. 107!4.7! Alternative residue treatment processes ......................................................................... 109!  vii 4.7.1! Enhanced atmospheric leach ................................................................................... 110!4.7.2! Residue pre-treatment by dry heating ..................................................................... 111!Chapter 5: Conclusion ...............................................................................................................113!5.1! Summary of results ........................................................................................................ 113!5.2! Significance of results .................................................................................................... 116!5.3! Strengths and limitations of research ............................................................................. 118!5.4! Applications of research and future work ...................................................................... 119!Bibliography ...............................................................................................................................121!Appendices ..................................................................................................................................134!Appendix A ............................................................................................................................. 134!Appendix B ............................................................................................................................. 135!B.1! Preliminary experiments data .................................................................................... 135!B.2! Fe displacements experiment data ............................................................................. 139!B.3! Sequential extraction data ......................................................................................... 142!B.4! Aged then Fe displacement experiment data ............................................................. 144!B.5! Repeated ageing experiments data ............................................................................ 146!Appendix C ............................................................................................................................. 148!C.1! Repeated ageing calculations .................................................................................... 148!C.2! Residue mass balances .............................................................................................. 152!   viii List of Tables Table 1: Leach conditions for sulphide concentrates ...................................................................... 5!Table 2: Mineralogical composition of Cu concentrate that produces the S1 residue .................... 6!Table 3: Pertinent elemental analysis of Cu concentrate that produces the S1 residue .................. 6!Table 4: Pertinent elemental analysis of Ni concentrate that produces the S3 residue ................... 7!Table 5: Comparison of amounts of X-ray amorphous and HaHC soluble phases in residue samples ............................................................................................................................................ 9!Table 6: Residue phase analysis by QXRD (wt%) ....................................................................... 10!Table 7: Pertinent residue chemical analysis by ICP-MS (wt.%) ................................................. 11!Table 8: XRD phase analysis S1, S1 R1, S3 and S3 R1 before and after ageing for 24 hours at 95°C (wt.%) .................................................................................................................................. 59!Table 9: Fe mass and phase balance by ICP and QXRD for aged residues, 24 hours, 95°C, 2%wt. solids ............................................................................................................................................. 61!Table 10: Comparison of total Fe distribution between original and aged samples by ICP, QXRD and stage 1 extraction .................................................................................................................... 63!Table 11: XRD phase analysis of 24 hour aged samples following further stage 1 extraction (wt.%) ............................................................................................................................................ 64!Table 12: Fe removed by acidified hydroxylamine hydrochloride of aged samples .................... 65!Table 13: Metal extractions and residue grades from ICP-MS analysis after 24 hours of ageing at 95°C .............................................................................................................................................. 66!Table 14: Metal extractions and residue grades from ICP-MS analysis after 1 hour of ageing at 95°C, 2%w/w solids ...................................................................................................................... 67!Table 15: Proportion of residue mass loss contributed to Fe2+, Cu2+, Ni2+ and Ca2+ as sulphates 69!  ix Table 16: Mass distributions for Ca, Cu, Fe, Ni of aged and sequentially extracted residues ..... 72!Table 17: Comparison of grade, recovery and mass losses in aged and stage 1 sequential extracted residues .......................................................................................................................... 73!Table 18: Mineral identification and composition in S1 residue determined by EDS analysis ... 79!Table 19: Composition of four hematite particles and average composition of 18 hematite particles analysed by EMPA ......................................................................................................... 87!Table 20: Phase analysis comparison between residues aged at 95°C and 150°C before stage 1 extraction (wt.%) ........................................................................................................................... 92!Table 21: Extraction of Cu and Ni from S1 and S3 residues by ageing in H2O and with added Fe2+ ................................................................................................................................................ 95!Table 22: Extraction of Cu and Ni by each process (%). Ageing at 95°C pH 2, 2% solids; sequential extraction; Fe displacement at 95°C, 2% solids .......................................................... 99!Table 23: Distribution of Cu and Ni by extraction process in S1 and S3 residues ..................... 101!Table 24: Cu, Fe, SO4 stage extractions by ageing and sequential extraction ............................ 107!Table 25: Composition of washed and unwashed S4 residue ..................................................... 108!Table 26: Metal extractions and residue grades from ICP-MS analysis after 24 hours of ageing at 95°C, and sequential extraction with 0.25 M hydroxylamine hydrochloride, 0.25 M HCl at 50°C for 30 minutes ............................................................................................................................. 109!Table 27: Proportion of residue mass loss contributed to Fe, Cu, Ni and Ca as sulphates ......... 109!Table 28: Enhanced atmospheric leach of S1 and S4 residues. 5 g/L Cl-, 70°C, 2 h, pH 1.5, 15%w/w solids ............................................................................................................................ 110!   x List of Figures Figure 1: Sequential extraction scheme for determination of HaHC soluble phases and crystalline iron oxides ....................................................................................................................................... 8!Figure 2: CESL Cu process flowsheet .......................................................................................... 17!Figure 3: Mechanism of ferrihydrite precipitation and growth .................................................... 42!Figure 4: Ageing experiments equipment set-up .......................................................................... 51!Figure 5: Parr Instruments Model 4520 Bench Top Reactor System ........................................... 53!Figure 6: Experimental flowchart for ageing in Fe2+ .................................................................... 55!Figure 7: Comparison of Fe and SO4 ratios to Cu and Ni for each process ................................. 75!Figure 8: Large, unleached enargite particle (1) with slight evidence of product layer ............... 78!Figure 9: Cu-rich, As-rich arsenopyrite/pyrite (2) particle with no apparent surface product layer....................................................................................................................................................... 78!Figure 10: Cu-poor, partially leached tennantite (3) surrounded by matrix of amorphous and poorly-formed particles and small crystalline minerals ................................................................ 78!Figure 11: Unleached Cu-rich tennantite (5) with small intrusion of Cu-poor tennantite mineral (6) in a matrix of amorphous and crystalline material .................................................................. 78!Figure 12: Partially leached enargite  (4) with amorphous and poorly formed particles nearby .. 79!Figure 13: Partially reacted chalcopyrite (7) and pyrite (8) .......................................................... 81!Figure 14: Large, porous hematite particle liberated from sulphides ........................................... 81!Figure 15: Pyrite particle surrounded by amorphous phases and other particles of unknown composition ................................................................................................................................... 82!Figure 16: Hematite particle with Cu and S in the centre, ongoing hematite formation is noted by the arrow ....................................................................................................................................... 82!  xi Figure 17: Close up of uncoated, homogenous hematite particle ................................................. 84!Figure 18: Hematite particle coated by a sulphurous mass containing some Cu .......................... 84!Figure 19: Large, oversized, unleached pentlandite particle ........................................................ 84!Figure 20: Large, partially leached pentlandite particle ............................................................... 84!Figure 21: Cu, Ni and Fe extraction from S1 and S3 residues at 150°C, 16% w/w solids, H2O .. 89!Figure 22: Cu, Fe extractions after single ageing stage at 95°C, 2, 6, 10, 15%w/w solids, 1 hour..................................................................................................................................................... 103!Figure 23: Cu stage extraction from S1 after each ageing stage. 95°C, 1 hour, average of 6, 10, 15%w/w solids ............................................................................................................................ 104!Figure 24: Total Cu grade and extraction after each stage of ageing and sequential extraction 105!Figure 25: Ratio of average Cu extraction versus average Fe and SO4 extraction for each stage. 5 ageing stages at 95°C for 1 hour followed by 2 sequential extractions with 0.25 M HCl and 0.25 M hydroxylamine hydrochloride ................................................................................................ 106!   xii Acknowledgements I offer my enduring gratitude to the faculty, staff and my fellow students at UBC, who have made my two years of study both highly enjoyable and smooth. I owe particular thanks to Associate Professor Edouard Asselin for his supervision and constant beneficial feedback on my thesis. My colleagues and fellow project members Baseer Abdul and Tasawar Javed are owed a great deal of appreciation for their assistance and mentoring during the course of my studies, it would have been much more difficult without them. Huge thanks must also be extended to Mr. Henry Salomon De-Friedberg from CESL for devoting significant amounts of his time to meet the project team and myself to review and discuss progress on the project. His probing questions and insightful input greatly contributed to the success of my research. This project would not have been possible without the massive amount of assistance from John MacIntosh and his team at CESL in analysing my solution and residue samples. For this I am truly grateful and his help was a major contribution to the success of my project. Input on the project from Dr. Indje Mihaylov from Vale was also appreciated and I would like to thank him for his time in attending our project presentations and answering some questions promptly over email. Special thanks are owed to Professor Mati Raudsepp, Edith Czech and Lan Kato at the Department of Earth and Ocean Sciences at UBC for their assistance and diligence in achieving excellent results from electron microprobe and quantitative X-ray diffraction analyses. The assistance of Jacob Kabel on the scanning electron microscope is greatly appreciated and contributed a lot to my research. I would also like to thank the Examining Team at UBC of Professors Warren Poole, Edouard Asselin, Wenying Liu and Marik Pawlik as well as Mary Janespar for organising the examination. Lastly and by no means least, I would like to extend my thanks and appreciation to the project sponsors for funding my research, these are Vale Ltd, Teck Ltd and NSERC.    xiii Dedication  I would like to dedicate my thesis to my father, Chris Ward for his encouragement, mentoring and assistance in enabling me to come to UBC for my MASc. I would not have had the desire or knowledge of the benefits of coming to UBC without him.   1 Chapter 1: Introduction 1.1 Project background The depletion of primary copper and nickel sulphide orebodies along with the implementation of more stringent environmental regulations by governments worldwide has led to the declining appeal of constructing new smelters for pyrometallurgically treating sulphide concentrates. A number of alternative hydrometallurgical processes for treating these concentrates have been developed to overcome the dual problem of declining grades and more refractory ores while meeting environmental goals. This technology can provide opportunities for mine-to-mill optimization, more scale dependent economics and enhanced metal supply security (Dreisinger 2006). McDonald and Muir (2007a) and Dreisinger (2006) have reviewed these processes extensively.   Cominco Engineering Services Ltd (CESL), now a subsidiary of Teck Resources Ltd, developed the CESL process to extract and recover copper from chalcopyrite by a purely hydrometallurgical route. This novel technology and approach was later expanded to include processes for precious metal and nickel recovery. While the large majority of the hydrometallurgical processing technologies have not gone further than the pilot scale, the CESL nickel process has been adapted and utilized for Vale’s new integrated hydrometallurgical refinery operation in Long Harbour, Newfoundland, Canada. This facility began operations in 2014 after a two-year construction phase and will process a bulk nickel-cobalt (>20% Ni and 1% Co) concentrate primarily from the Voisey’s Bay operation in Labrador, Canada to produce a high-quality nickel, copper and cobalt products.   2 The Vale Long Harbour process leaches the Voisey’s Bay nickel concentrate, consisting of pentlandite, pyrrhotite and some chalcopyrite, under pressure oxidative conditions at 150°C in a H2SO4-NiSO4/NiCl2 solution (Chen et al. 2006). Nickel, cobalt and iron from the pentlandite and pyrrhotite along with some chalcopyrite dissolve during the leach and the sulphides are converted to elemental sulphur. Iron is removed from solution by precipitation to hematite. Hematite is the desired product as it is the most stable of the iron precipitates and is suitable for long-term, environmentally sound, disposal in a tailings storage facility.   The commercialisation of the CESL copper process was investigated as a partnership between Vale and Teck Resources. The copper process is utilised by CESL for various pilot plant trials and studies at their facility in Richmond, BC. Vale has also developed a 10,000 t/y demonstration plant to treat various concentrates from the Carajás region in Brazil using CESL process technology. It was initially envisaged that the successful completion of the demonstration plant would lead to the construction of a full-scale 250,000 t/y processing plant in Brazil to treat other concentrates. Teck was also considering the application of the CESL process to its existing Highland Valley Copper Mine in BC. Further details are provided elsewhere (Defreyne et al. 2006).   Neither project has currently progressed since the completion of the demonstration plant trialling due to a combination of factors. Several of these may be related to the drawbacks associated with hydrometallurgical treatment of sulphide concentrates discussed by Dreisinger (2006) and that relate to the present study. In the leach residues of both processes, metal losses are noticed due to co-precipitation with iron hydrolysis products and there is difficulty in creating a stable waste  3 residue product. Characterisation of medium temperature pressure leach residues obtained by CESL Cu processing revealed the residues contain significant amounts of amorphous and semi-amorphous iron oxide precipitates (Sahu and Asselin 2011). These precipitates have been shown to occlude Cu and Ni and degrade over time in conditions of a tailings storage facility. These iron oxide precipitates are known as ferrihydrite and result from ferric hydrolysis. Ferrihydrite is metastable and transforms to more structurally ordered hematite and goethite over time. This ferrihydrite phase was shown to contain five times as much Cu as the crystalline iron oxides also present in the residue and as a result warrants further attention to recover Cu lost to this phase. The properties and occurrences of ferrihydrite are discussed in detail elsewhere (Jambor and Dutrizac 1998). The leach residue produced from the Vale Ni process also contains ferrihydrite and significant losses of Cu and Ni. Recovering these values can improve the economics of the Long Harbour project though the major concern is the long-term storage of the residue in an impoundment facility. The degradation of ferrihydrite within this residue is capable of releasing various toxic elements to solution that are an environmental concern to the project operators.   Treating both residues by an ageing process shows promise to both remove some lost Cu and Ni and provide conditions to promote transformation of ferrihydrite to more stable, crystalline iron oxides. Ferrihydrite eventually ages to hematite with sufficient time in aqueous solution though kinetics are greatly enhanced by increased temperatures as discussed by Cornell, Giovanoli, and Schneider (1989). Significant Cu and Ni are found as adsorbed and occluded sulphate species on the crystalline hematite and goethite precipitates, analogous to the hematite produced from similar conditions by Dutrizac and Chen (2012b). Specific adsorption of Cu2+ and Ni2+ to the iron oxides in the residue may also occur though the process conditions remain unfavourable.  4 The ageing process is capable of removing some of these species by dissolution as agitation of the residue also acts as a washing mechanism. Continued ageing of the residues continues to release Cu and Ni and dissolve ferrihydrite though the performance of this process is not as effective as selectively dissolving this phase with a sequential extraction treatment with acidified hydroxylamine hydrochloride.   1.1.1 Project aim The present study forms part of a larger project run at the Materials Engineering Department of the University of British Columbia with funding support and collaboration with industry partners, Vale and Teck Resources Ltd. The project aims to ultimately find the optimum practical process conditions to improve the CESL and Vale medium temperature pressure leach processes. Maximising the stability of the waste residues from the CESL and Vale processes and minimising the loss of Cu and Ni to this residue gauge achievement of the project goals. Meeting these goals largely depends on the ability to minimise the formation of metastable amorphous and semi-crystalline iron oxide precipitates known as ferrihydrite and to deport the leached Fe as a stable hematite precipitate. This meets both project goals as ferrihydrite disproportionately accounts for Cu and Ni losses to the residue and hematite is the most thermodynamically stable iron oxide.   Treatment of these residues to recover lost metal values and improve the stability of the iron oxide precipitates represents a significant economic and environmental opportunity for the project. This study will investigate an ageing process that serves the purpose of recovering occluded metal values as surface sulphate and adsorbed species and improves the stability of the  5 residue by dissolving and transforming some of the ferrihydrite phases. As part of this project it is important to characterise the residues and assess the nature of the lost Cu and Ni. Achievement of the project goals is determined by finding a maximum practical metal recovery from each residue by a simple atmospheric ageing process that could be cheaply implemented within a process flowsheet using existing or additional process equipment.   1.2 Process conditions for producing CESL and Vale leach residues  Three leach residues concern the present study, an older CESL Cu pilot-scale leach residue composite (S1), a typical Vale Ni demonstration-scale leach residue (S3) and a more recent CESL Cu pilot-scale leach residue (S4). S1 and S3 are studied extensively and compared with the more recently produced S4 residue. All 3 residues are produced from different concentrates by medium temperature pressure leaching under similar conditions. The residues were formed by the leach conditions given in Table 1. Table 1: Leach conditions for sulphide concentrates Operating Parameters Unit S1 S3 S4 Total pressure kPa(g) 1298 1264 1378 Temperature °C 150 150 150 Retention time min 91 80 90 Autoclave free acid concentration g/L 10.7 10-12 10-15 Solids loading % 10.2 5 10 [Cl-] g/L 11.2 5 12  1.2.1 Concentrate producing S1  The S1 residue is produced from a composite of several concentrates rich in mainly chalcopyrite and enargite with significant pyrite also present. S1 was produced from a CESL Cu pilot trial run at the CESL facilities in Richmond, BC during 2010 (Salomon de Friedberg 2015). The mineralogical composition of the concentrate that produces the S1 residue is given in Table 2.  6 Table 2: Mineralogical composition of Cu concentrate that produces the S1 residue Mineral Formula %Composition Bornite Cu5FeS4 7 Covellite CuS 5 Chalcopyrite CuFeS2 38 Tennantite Cu12As4S13 2 Luzonite/Enargite Cu3AsS4 20 Pyrite FeS2 18 Sphalerite ZnS 2 Gangue  8  The pertinent elemental chemical analysis of the concentrate producing S1 by full digestion in aqua-regia followed by ICP-MS analysis is given in Table 3.  Table 3: Pertinent elemental analysis of Cu concentrate that produces the S1 residue Cu % 29.5 Fe % 19.6 S % 33.5 As % 4.6 Au ppm 5 Ag ppm 105  1.2.2 Concentrate producing S3 Vale’s Voisey’s Bay concentrator situated on the northeast coast of Labrador, Canada, produces an estimated 50,000 tonnes of nickel, 38,600 tonnes of copper and 2,300 tonnes of cobalt from three concentrates. The concentrator generates a high-grade nickel concentrate (25% Ni, 1.2% Co), a high-grade copper concentrate and a low-grade Ni-Cu middlings concentrate. A portion of the high-grade Ni concentrate and the middlings concentrate (combined grade 17% Ni, 3% Cu and 1% Co) is currently smelted in Sudbury, Ontario. Upon completion of the Long Harbour facility, the Voisey’s Bay concentrator will be modified to produce a single nickel concentrate and a single copper concentrate (Chen et al. 2006).   7 The Vale Ni concentrate that is leached at the Long Harbour facility to produce S3 is analogous to the residue discussed by Chen et al. (2006) and consists mainly of pentlandite ((Fe,Ni)9S8), pyrrhotite (Fe1-xS) and troilite (FeS) with small amounts of entrained chalcopyrite (CuFeS2). The concentrate is fed to the autoclave at p80 20 µm at the conditions given above. Pertinent elemental chemical analysis of the concentrate producing S3 is given in Table 4. Elemental compositions are obtained by aqua regia digestion and ICP-MS analysis of the resulting solution.  Table 4: Pertinent elemental analysis of Ni concentrate that produces the S3 residue Ni % 20.6 Cu % 2.0 Co % 0.96 Fe % 39.0 S % 32.0  1.2.3 Initial residue analysis  Previous study of residues analogous to S1 and S3 revealed significant differences between the elemental composition calculated from quantitative X-ray diffraction (QXRD) data and aqua regia total digestion followed by ICP-MS (Sahu and Asselin 2011). The differences were found to be due to the presence of the amorphous and semi-crystalline iron oxide precipitates known as ferrihydrites. It is these amorphous Fe phases that are disproportionally responsible for most of the Cu and Ni losses to the residue and impart instability on the residues. A two-stage sequential extraction procedure was developed to assess the distribution of Cu and Ni with the ferrihydrite and crystalline iron oxide phases (Sahu and Asselin 2011). The ferrihydrite phase was found to be soluble by reaction with 0.25 M hydroxylamine hydrochloride (HAHC) and 0.25 M hydrochloric acid according to the procedure of Chao and Zhou (1983). Residues treated by this process are denoted by an R1 descriptor so that S1 residue treated by the first stage of sequential extraction is detailed as S1 R1, likewise for S3 R1. The remaining residue from the first stage of  8 sequential extraction is treated by 4 M HCl to dissolve the remaining crystalline iron oxide phases. The residue from this second stage of sequential extraction is then denoted as ‘R2’ in addition to S1 or S3. Metals associated and the proportion of Fe in each is determined by chemical analysis of the solution by ICP-MS. A schematic of the sequential extraction procedure is given in Figure 1 and is from Javed et al. (2015).  Figure 1: Sequential extraction scheme for determination of HaHC soluble phases and crystalline iron oxides (Javed et al. 2015)  9 Previous work to assess the stability of the original, as delivered S1 and S3 residues is presented in Table 5 (Javed et al. 2015). Residue mass loss during the first stage of extraction determines the %HaHC soluble phase and mostly represents dissolution of the semi-amorphous Fe phases and crystalline calcium phases (Sahu and Asselin 2011). This fraction is a subset of the overall amorphous phase as determined by XRD analysis. A higher mass loss from the first stage of extraction represents higher ferrihydrite content and less stability. Reducing the ‘HaHC soluble phase’ will lessen the X-ray amorphous phase present in the residues resulting in more stable residues and the release of Fe, Cu and Ni to solution.   Table 5: Comparison of amounts of X-ray amorphous and HaHC soluble phases in residue samples  (Javed et al. 2015) Sample %HaHC soluble phases %X-ray amorphous phase S1 35.5 52.7 S3 3.9 30.5  The QXRD phase analysis data for both the original S1 and S3 samples and the residues following the first stage of sequential extraction with HaHC (labelled as R1) are given in Table 6.  Phase and chemical analyses presented in Table 6 and 7 are from Javed et al. (2015).           10 Table 6: Residue phase analysis by QXRD (wt%)     Original Sample compositions Phase Formula S1 S1 R1 S3 S3 R1 Albite low NaAlSi3O8 0.9     Alunite KAl3(SO4)2(OH)6       Boehmite AlO(OH)       Calcite CaCO3       Chalcopyrite CuFeS2    0.2   Goethite α-Fe3+O(OH)    8.0 10.5 Gypsum CaSO4.2H2O 3.2  1.0 0.9 Hematite α-Fe2O3    35.0 35.2 Jarosite KFe3(SO4)2(OH)6 13.5 13.4 1.0   Lizardite Mg3Si2O5(OH)4 1.3 1.2    Magnetite Fe3O4       Muscovite KAl2(SiAl)O10(OH,F)2 3.9 2.2    Pyrite FeS2 1.5 1.7    Quartz low  SiO2 4.2 6.5    Sulphur S8 16.3 28.3 24.2 24.6 Talc Mg3Si4O10(OH)2 1.8     Tennanite (Cu,Fe)12As4S13 0.5 0.9    Amorphous  52.7 45.8 30.5 28.8 Total   100.0 100.0 100.0 100.0  Elemental analysis by QXRD is considered semi-quantitative due to the presence of large amounts of amorphous material and is therefore only useful for phase analysis. The stability of the residues is therefore indicated by the use of the first stage of sequential extraction. Chemical analysis determines the remaining Cu, Ni and Fe in the residues. Pertinent chemical analysis of the S1 and S3 residues is provided in Table 7.     11 Table 7: Pertinent residue chemical analysis by ICP-MS (wt.%) Element S1 S3 S 28.5 31.3 Fe 22.6 40.0 Ca 1.01 0.26 Cu 1.1 0.57 As 6.1 <0.01 Si 6.6 0.23 Al 0.83 0.07 Mg 0.02 0.01 K 0.23 0.01 Na 0.22 0.26 Ni <0.01 0.9 Pb 0.48 0.013 Sb 0.26 <0.01 Ti 0.04 <0.01   12 Chapter 2: Literature review 2.1 Hydrometallurgical treatment of sulphide concentrates Hydrometallurgical processes for the treatment of sulphide concentrates are based on either a sulphate or chloride system. The sulphate system is currently the more preferred option due to necessity of enhanced corrosion protection associated with chloride systems (Dreisinger 2006). Sulphate leaching systems can be divided into low, medium or high temperature sub-categories and are extensively reviewed elsewhere (McDonald and Muir 2007a; McDonald and Muir 2007b; Dreisinger 2006). Low temperature (typically 90-120°C) regimes typically require super fine grinding coupled with bacterial oxidation or high oxygen overpressure to overcome slow reaction kinetics. Medium temperature leaching processes utilize fine grinding, higher temperatures (typically 125-150°C) but require the addition of specific surfactant reagents to minimize wetting of the sulphide surface by molten sulphur that would otherwise slow the leaching kinetics.   Aside from the reaction kinetics these processes also vary in the leach products, particularly in the deportment of waste Fe and S particles to the residue. Low temperature processes such as the ACTIVOX® process typically favour conversion of Fe to hydronium jarosite and goethite (Palmer and Johnson 2005). Sulphide partially converts to elemental sulphur with conversion largely dependent on the mineralogy. High temperature processes such as the commercial Total Pressure Oxidation process of Phelps-Dodge (now Freeport McMoran Corporation) in Arizona, USA use temperatures of 200-230°C to convert all sulphide minerals to sulphate and sulphuric acid. The higher temperatures utilised by this process precipitates iron as hematite that can be recycled for precious metal recovery, disposed of or recycled back to the leach as a seeding agent  13 to produce more hematite and improve copper recoveries (Marsden et al. 2002; Marsden et al. 2010). Medium temperature pressure leach processes utilise high purity oxygen to effectively oxidise all the common sulphide minerals but is selective for sulphur. Sulphides are converted to elemental sulphur with minimal oxidation to sulphate. The typical conditions of pressure oxidation (150°C, pH 3) convert Fe to hematite by hydrolysis (Defreyne et al. 2006). Characteristics of two medium temperature pressure leach processes; the CESL Cu process and Vale Ni process concern the present study and will be discussed in the following section.   Medium temperature leach processes such as the CESL Cu process are capable of recovering 95-99% Cu from the concentrate. The source of the losses is often unclear though ~2% chalcopyrite has been found in residues leached under CESL conditions (McDonald and Muir 2007a). Further Cu losses can occur due to co-precipitation with iron hydrolysis products (Dreisinger 2006). Both processes face the technical and economic difficulty of fixing toxic by-products in a stable waste product. A metastable residue is prone to re-leach toxic metals to the environment with the potential to cause harm. It is also desirable to produce a waste product high in Fe, such as hematite, as this can significantly reduce the volume of the tailings storage facility required to hold the residue, saving significantly money over the life of mine (Dutrizac and Chen 1993). The nickel sulphide concentrate refining process at Long Harbour, Newfoundland, Canada operated by Vale would significantly benefit economically and environmentally from progress in this area.    14 2.1.1 CESL and VALE process for Cu and Ni sulphide concentrates Cominco Engineering Services Limited (CESL) has developed a proprietary hydrometallurgical process for base metal concentrates, notably those of copper and nickel. As discussed by several authors (Dreisinger 2006; Defreyne et al. 2006; McDonald and Muir 2007a) the CESL process achieves maximum oxidation of base metal sulphides while minimizing the oxidation of sulphur to sulphate and the leaching of iron. Medium temperature pressure leaching conditions balances the advantageous reduction in oxygen usage by conversion of sulphide to sulphur and reduced wetting of the sulphide surface against the rapid kinetics and better iron sulphide conversion to hematite of the high temperature processes. In medium temperature pressure leach systems, the oxidation of sulphide to elemental sulphur is ~70-80% at 150°C and is higher in the presence of chloride ion and at higher initial acidity. Hematite precipitation is generally favoured at higher temperatures (≥150°C), low acidity and low to moderate salinity. Goethite formation is favoured at lower temperatures (<150°C), low acidity and low salinity (McDonald and Muir 2007b). Jarosite forms in conditions of moderate to high acidity and formation is enhanced in the presence of sodium ions (McDonald and Muir 2007a). The CESL process ideally balances these factors to favour the precipitation of stable and easily treatable hematite over other precipitation products while minimizing many of the costly inputs into the process.   2.1.2 CESL copper process The CESL copper process is a two-stage leaching process consisting of a pressure leach and atmospheric leach stage, followed by solid-liquid separation, metal recovery and tailings disposal. The tailings produced by this process are indicative of the CESL leach residue studied in the present work. The process is notable for its selective nature with respect to sulphur despite  15 the very effective oxidation environment for base metals (Defreyne et al. 2006). The first stage of the process utilizes pressure oxidation of finely ground (~45 µm) sulphide minerals at elevated temperatures (150°C) in mixed sulphate-chloride liquor at pH 3. Low initial oxygen partial pressures (~200 kPa), 12 g/L Cl− as CuCl2 and ~25 g/L SO42− are used to facilitate the leaching process. All copper minerals, such as chalcopyrite (CuFeS2), are converted to basic copper salts (BCS) similar to antlerite (CuSO4.2Cu(OH)2), elemental sulphur and hematite. Almost all other base metal sulphide minerals (if present) are solubilized. The unique feature of oxygen gas as the oxidant rather than the ferric/ferrous couple is aided by the presence of chloride and copper ions in relatively dilute concentrations (Defreyne et al. 2006). The exact mechanism of chloride and cuprous catalysis in this environment is not fully understood. It is thought that the addition of Cl− ions as cupric chloride limits the production of acid-consuming hydronium jarosite during the initial stages of the leach (McDonald and Muir 2007a). This limits the increase in free acidity after the initial stage of leaching and favours chalcopyrite leaching and precipitation of Fe as hematite.    Chalcopyrite is converted to a basic copper sulphate (BCS) similar to antlerite by reaction 1 in the pressure oxidation stage. Similar equations can be written for the other primary copper sulphides. Details of the CESL Cu process and the role of iron are discussed by Defreyne et al. (2006).  12!"#$!! + 15!! + 4!!! + 4!!!"! → 4!"#$!. 2!"(!")! + 6!"!!! + 24!!![1]   16 The reaction shows that iron is not required in the feed liquor to catalyse the reaction. Iron sulphide minerals deport almost entirely to the solid residue as hematite and the sulphide deports as crystalline elemental sulphur. A lignosulfonate surfactant is used to disperse molten sulphur from the surface of the sulphide particles to prevent passivation (or coating) of the chalcopyrite. Oxygen utilisation and the heat energy required by the reaction are lessened by the partial oxidation of sulphur to elemental sulphur rather than sulphate. Iron precipitates from the feed liquor under the conditions of pressure oxidation by hydrolysis as per the reaction below.  !"!(!"!)! + 3!!! → !"!!! + 3!!!"! [2]  A second atmospheric leaching stage follows pressure leaching that dissolves the BCS using mildly acidic conditions at ambient temperature (~40°C) and pressure.  4!"#$!. 2!"(!")! + 2!!!"! → 3!"#$! + 4!!! [3]  Raffinate is added to the circuit to supply excess acid to maximize BCS leaching. A pH of 1.4-1.8 is maintained to limit the re-dissolution of the hematite by the reverse of reaction 2.  Any re-dissolution of hematite at this stage is immaterial as in continuous operation the iron concentration reaches equilibrium around 1-3 g/L so there is little to no net dissolution of iron. Leached copper is then recovered from solution following counter-current decantation (CCD), solvent extraction (SX) and electrowinning (EW). The process does not require neutralisation of excess acid prior to solid-liquid separation and metal recovery. The solids residue that leaves the fifth and final counter-current decantation is indicative of the S1 residue concerning the present study.   17  Figure 2: CESL Cu process flowsheet (Defreyne 2006)  2.1.3 Vale nickel process  The Vale nickel process utilizes technical aspects of the CESL nickel process with specifically adjusted Vale process chemistry conditions to produce refined nickel and cobalt metal from the Voisey’s Bay nickel concentrate. This process is used at the Long Harbour refining facility in Newfoundland, Canada. The Vale Long Harbour process leaches the nickel concentrate in H2SO4-NiSO4-NiCl2 solution at 150°C under pressure oxidative conditions of approximately 700 kPa oxygen overpressure.  The presence of chloride ion assists in the leaching of the nickel sulphide by dramatically assisting the dissolution rate and minimizing the conversion of sulphide to sulphate as per the CESL process. Typically only 25% of the sulphur will dissolve as sulphate,  18 limiting the oxygen requirement for the metal sulphide dissolution (Chen et al. 2006). Vale conditions utilize higher amounts of acid (50 g/L H2SO4), metals (50 g/L Ni, 5-15 g/L Na) and lower chloride (4-8 g/L Cl- as NiCl2) than the CESL process to affect the sulphide dissolution. The pressure leaching autoclave operates between 5-10%w/w slurried solids and 98% of the Ni and Co values are leached after 20-30 minutes at 150°C (Kerfoot et al. 2002). Nickel sulphides oxidise effectively under these pressure oxidation conditions such that a second atmospheric leaching stage is not required as per the CESL copper process.   Leaching reactions for the oxidative pressure leach of a pyrrhotite/pentlandite concentrate by the Vale Long Harbour process are given below and are taken from Chen et al. (2006). (!",!")!!! + 9 2!! !+ 9!!!"! → 9 !",!" !!! + 8!! + 9!!! [4] !"!!!! + (1− !) 2!! + 1− ! !!!"! → 1− ! !"!"! + !! + (1− !)!!! [5] Leached Fe sulphates precipitate to hematite or goethite releasing further acid for the leach.  2!"!"! + 2!!! + 1 2!! → !"!!! + 2!!!"!![6] 2!"#$! + 3!!! + 1 2!! → 2!"##$ + 2!!!"! [7] Overall reactions of pentlandite are represented by equations 8 and 9.  (!",!")!!! + 27− 9! 4!! + 9!!!!"! → 9!"#!"! + 8!! + 9!"!! +(9− 9!) 2!"!!!![8] (!",!")!!! + 27− 9! 4!! + 9!!!!"! → 9!"#!"! + 8!! + 27! − 9 2!!! +9− 9! !"##$ [9]  Solid-liquid separation following oxidative pressure leaching removes the solid leach residue from the leach solution. Most of the iron present re-precipitates as hematite with significant  19 goethite also noticed. Hematite and goethite are precipitated as either tiny (1-2 µm), discrete spheroids; larger, bubble shaped particles; or coarse irregular grains. Elemental sulphur precipitates as spheroids and irregular masses that bond to the other particles to form agglomerates (Chen et al. 2006). Sodium jarosite is rarely found in the residue but can possibly be formed under these conditions. The pressure leach residue produced following solid-liquid separation by counter-current decantation is stage is indicative of the S3 residue concerning the present work.   Concentrate fed to pressure leaching at Long Harbour typically contains 1-2% Cu as chalcopyrite that is leached by the same conditions as pentlandite. Copper is leached to solution according to reaction 10 (Kerfoot et al. 2002).  !"#$%! + !! + 2!!!"! → !"!"! + !"!"! + 2!!! + 2!! [10]  Chalcopyrite is more difficult to leach than pentlandite by pressure leaching and partially leached chalcopyrite minerals are typically found in the residues (Chen et al. 2006). Residual copper in the leach solution is removed from solution by solvent extraction or precipitation as a copper sulphide intermediate product prior to the neutralization and iron removal stage. An additional limestone neutralisation stage at controlled pH is included to remove both the remaining iron as ferric hydroxide and calcium as gypsum from solution, and to raise the solution pH. Cobalt is separated from the neutralized nickel-cobalt rich solution by solvent extraction before the nickel is won from solution as nickel cathode by electrowinning. Full details of the process are found in the US Patent 6,428,604 B1 (Kerfoot et al. 2002) and the residue is characterized by Chen et al. (2006).   20 Sodium and/or hydronium jarosite can also form in the residue under the same conditions as hematite/goethite, though it was rarely found in the residues studied (Chen et al. 2006). Sodium jarosite forms by hydrolysis with ferric.  !"!!"! + 3!"!(!"!)! + 12!!!! → 2!"!"!(!"!)! !" ! + 6!!!"! [11]  Jarosite species are formed in the absence of any hematite seed as a synthetic solution under the same conditions yielded iron precipitates as jarosite. The addition of hematite seed promoted hematite formation at the expense of jarosite. From this it is evident that hematite present in the early stages of leaching plays a key role in promoting the precipitation of hematite (Chen et al. 2006).  2.2 Iron control in CESL and Vale pressure leach processes Impurity Fe leached with Cu and Ni in the autoclave needs to be removed by precipitation prior to the Cu and Ni recovery stages. Iron precipitation generally occurs primarily and simply, through a process of oxidation and neutralization that effectively removes all iron from solution. The resulting gel-like precipitate is exceedingly difficult to filter and it occludes significant amounts of the processing solution. Furthermore, the precipitate is nearly impossible to wash effectively, and its environmental stability is poor because of the incorporation of many of the impurities present in the original processing solution (Dutrizac and Chen 1993). As a result of these limitations, the oxidation and neutralization processes are carefully controlled to produce suitable iron precipitates. These precipitates are typically more crystalline with minimal impurities, easy handling characteristics and more stable when disposed of.  Both the CESL and Vale pressure leaching processes aim to remove soluble iron by precipitation to hematite in a  21 stable leach residue for disposal in a tailings storage facility (TSF). The low free acid content of the leach liquor promotes the precipitation of thermodynamically stable and easily treatable hematite over other precipitation products and minimizes many of the costly inputs into the process. Its stability ensures that any co-precipitated impurities do not leach from the residue impoundment area. Hematite yields the lowest residue volume for a given amount of iron and has some potential for cement manufacture, and even for iron making (Dutrizac and Chen 2011b).    In the CESL process, almost all iron present in solution following pressure oxidation is converted to ferric and precipitated as hematite by hydrolysis. Ferric hydrolysis is complex and there may be other possible precipitation products present in the pressure oxidation discharge.!Along with hematite, goethite and jarosite, other semi-crystalline iron oxyhydroxide products may also be present (Loan et al. 2002b). These semi-crystalline iron oxyhydroxides precipitates are common in hydrometallurgy and were often referred to as colloidal ‘hydrous ferric oxides’ or amorphous iron oxyhydroxides (Jambor and Dutrizac 1998). A mineralogical characterisation of this material found that it was poorly crystalline and had similar cell parameters to hematite (Towe and Bradley 1967). This material was dubbed ferrihydrite and is responsible for many of the difficulties encountered in treating precipitated Fe residues. Ferrihydrite is metastable and capable of adsorbing metals from solution (Loan et al. 2002a). Precipitation of ferrihydrite is also partly responsible for some of the challenges facing sulphide concentrate treatment by hydrometallurgy. As discussed by Dreisinger (2006) copper losses can be due to co-precipitation with iron hydrolysis products and there is difficulty in fixing waste products to a stable phase.    22 Some of the Cu and Ni losses in the CESL and Vale process leach residues are associated with this ferrihydrite phase. Ferrihydrite has been shown to adsorb Cu more strongly than other iron oxides (Li 2008). This is due to the high reactivity and large surface area of ferrihydrite (Cornell, Schwertmann, and Editors. 1996) and the availability of groups of binding sites on the surface (Benjamin and Leckie 1981). Co-precipitation of Cu with iron hydroxides often occurs (Martinez and McBride 1998a) and during the transformation to more crystalline goethite and hematite products becomes incorporated within the crystal lattice (Cornell and Giovanoli 1988). It was found that approximately 5 times more Cu was associated with the ferrihydrite phase than with the crystalline iron oxide precipitates in a similar residue to the ones studied by the present work (Sahu and Asselin 2011).   2.2.1 Iron hydrolysis Iron hydrolysis is important to consider, as this process is responsible for the speciation of the Fe precipitates in the residues concerning the present study. Ferric solutions have low solubility and except at very low pH immediately undergo hydrolytic polymerisation from the original hexa-aquo ion, Fe(H2O)63+ to hydroxo and oxo species. The pathway from soluble Fe3+ ions to the various solid polymeric Fe(III) oxides of goethite and hematite occurs through a series of intermediate soluble and poorly-crystalline solid phases. The immediate solid polynuclear species are poorly crystalline Fe(III)-oxyhydroxide salts and oxydroxides, termed ferrihydrites.  These phases transform to either solid hematite or goethite by two competing mechanisms that are favoured depending on the rate and conditions of ferric hydrolysis (Schwertmann, Friedl, and Stanjek 1999). The rate of ferric hydrolysis is affected by pH, rate of OH- addition and temperature (Cornell, Giovanoli, and Schneider 1989). These factors can be manipulated in  23 hydrometallurgical leach solutions to produce various Fe-rich residues from the hydrolysis product of ferric sulphate. Jarosite, goethite and hematite rich residues can all be produced from the same leach solution based on the conditions used for precipitation. In the CESL and Vale process, Fe is hydrolysed and precipitated as hematite contemporaneously with dissolution of the Cu and Ni (Defreyne et al. 2006; Kerfoot et al. 2002).   Hydrolysis of ferric sulphate at various acidities proceeds in one of three ways (Umetsu, Tozawa, and Sasaki 1977). At low acidity: !"!(!"!)! + 3!!! → !"!!! + 3!!!"! [12]  At moderate acidity: 3!"!(!"!)! + 14!!! → 2 !!! !"!(!")!(!"!)! + 5!!!!! [13] 2 !!! !"!(!")!(!"!)! + 3!!!!! → 4!"#$!"! + !"!(!"!)! + 10!!! [14] 2!"#$!"! + !!! → !"!!! + 2!!!"! [15] At high acidity: !"!(!"!)! + !!! → 2!"#$!"! + !!!"! [16]  The series of reactions above show that the hydrolysis process changes with increasing initial iron or sulphuric acid concentration. The precipitated species changes with the variation in initial reactant concentration from hematite (Fe2O3) to basic iron sulphate (FeOHSO4). These authors showed that lower concentrations of sulphuric acid and higher temperatures favoured hematite precipitation over basic iron sulphate. This agrees with industrial practice where higher  24 temperatures (>180°C) are used when eliminating the maximum amount of Fe by hematite precipitation is preferred (Dutrizac and Chen 2011a).   2.2.2 Hematite and goethite precipitation  Hematite can be precipitated from ferric solutions at temperatures between 80°C and 250°C. The solubility of ferric oxide decreases significantly with an increase in temperature, favouring precipitation (Umetsu, Tozawa, and Sasaki 1977). In all hematite precipitation processes, precipitation proceeds by firstly reducing ferric to ferrous utilising SO2 gas or sulphide concentrate. Solution pH is adjusted and oxidation of the solution slowly produces further ferric. These ions hydrolyse and as the ferric solubility is low, hematite precipitation occurs readily.  The CESL and Vale sulphide concentrate leach processes precipitate hematite at temperatures of 145 to 155°C at pH 3. In sulphate solution, indirect precipitation of hematite occurs by the equations below (Dutrizac and Chen 2011b).   2!"!"! + !!!! + !!!"! → !"!(!"!)! + !!!!![17] 2!"!(!"!)! + 3!!! → !"!!! + 3!!!"!!![18] 2!"!"! + !!!! + 2!!!! → !"!!! + 2!!!"! [19]  At all temperatures, hematite seeding is a critical factor for the precipitation of hematite (Dutrizac and Chen 2011b). Seeding provides growth sites for the crystallisation of hematite from solution in acid media (Cornell, Giovanoli, and Schneider 1989). By neutralizing the acid formed from the hydrolysis reaction, hematite can be produced over goethite with minimal co-precipitation of other ferric oxides/hydroxides (Arima et al. 2006). The proportions of each precipitated product are dependent on the various conditions and chemistry of the process used.! 25 The favourability of hematite/goethite precipitation in low/reduced acid environments is detailed elsewhere (McDonald and Muir 2007a). The chemistry of ferric precipitation relevant to hydrometallurgy is explored in significant detail by other authors (Dutrizac 1979; Umetsu, Tozawa, and Sasaki 1977).  The initial ferric reduction stage is important in determining the final crystalline Fe end-member. The increased solubility of ferrous over ferric allows the pH (pH >2.5) to slowly rise to benefit hematite precipitation without precipitating complex ferric oxyhydroxide-sulphate species such as basic iron sulphate (Fe(SO4)(OH) and Fe4(SO4)(OH)10) and ferrihydrite (Dutrizac and Chen 2011b). These reactions appear relatively straightforward however the mechanism is quite complex and controlling ferric hydrolysis to produce suitable iron precipitates is difficult on an industrial scale.  The difficulty in controlling the ferric concentration of metallurgical leach liquors often leads to rapid hydrolysis of Fe3+. This rapid hydrolysis of Fe3+ solutions commonly leads to the formation of ferrihydrite and has been obtained by precipitation from Fe3+ solutions containing various anions (Jambor and Dutrizac 1998). Pathways to the formation of ferrihydrite from Fe3+ polynuclear species are reviewed by Cornell, Giovanoli, and Schneider (1989) and Schwertmann, Friedl, and Stanjek (1999) and will be discussed further in a following section.   To precipitate goethite (FeOOH) instead of hematite or hydronium jarosite (H3O)Fe3(OH)6(SO4)2 the ferric concentration has to be minimized by reduction to ferrous. Goethite is precipitated directly from ferrous iron in an environment low in ferric iron. Goethite precipitation occurs under the same conditions as jarosite and hematite precipitation (pH 1.5 – 3.0, >100°C) and is  26 often present in majority hematite residues Industrially, goethite precipitation occurs from ferrous iron as follows (Lahtinen, Svens, and Lehtinen 2006). 6!"!"! + 1/2!! + 3!!! → 2!"##$ + 2!!!!! [20]  2.2.3 Ferrihydrite in hydrometallurgical residues  The low solubility of Fe3+ inherently creates a high supersaturation environment favouring the precipitation of nanoscale, poorly crystalline and metastable phases (Loan et al. 2005). These species are now recognized as ferrihydrite that can be further classified as ‘2-line ferrihydrite for the material that exhibits little crystallinity and ‘6-line ferrihydrite’ for the more crystalline species (Jambor and Dutrizac 1998). Ferrihydrite is considered a complex, disordered, metastable iron oxy-hydroxide compound and is the first product precipitated in ferric acid media. Addition of sufficient base (OH:Fe > 3) to ferric solution immediately leads to the precipitation of a poorly ordered ferric hydroxides (Cornell, Giovanoli, and Schneider 1989). It subsequently proceeds to more crystalline goethite and hematite under various conditions during precipitation (Loan et al. 2002a). It is now recognised that most iron-based residues are likely to contain ferrihydrite or schwertmannite (Loan et al. 2005). Poor agreement between total chemical analysis by ICP-MS and those calculated from QXRD indicated the presence of ferrihydrite in the CESL (S1) and Vale (S3) residues concerning the present study (Sahu and Asselin 2011). Characterisation of these residues showed that significant Cu losses were associated with this phase. This will be further discussed in the following sections.   The ferrihydrite precipitation mechanism is dominated by rapid nucleation (inherent high supersaturation due to low solubility) and almost negligible growth. The level of this  27 supersaturation has a direct effect on the crystal growth rate (Loan et al. 2004; Loan et al. 2008; Loan et al. 2006b). Supersaturation is generically expressed as the quotient of the solute concentration and equilibrium solubility of the solute.  It is a function of ferric concentration, pH, mean residence time and the nucleation and growth kinetics of the system. At lower levels of supersaturation and fast rates of precipitation, less ordered iron oxyhydroxide precipitates form, mainly schwertmannite (Fe16O16(SO4)2(OH)12) and 2-line ferrihydrite (Richmond et al. 2004; Loan et al. 2006a). The less ordered 2-line ferrihydrite crystallites are typically smaller (1-3 nm) than the more ordered 6-line ferrihydrite crystallites (3-7 nm) that form from slower rates of crystallization. These larger-sized 6-line ferrihydrite crystallites are considered the limit of primary crystal growth though the crystallites can agglomerate and aggregate to form bulk ferrihydrite particles as large as 100 µm (Loan et al. 2002b).   The nucleation and agglomeration of ferrihydrite can be controlled on an industrial scale to promote formation to the more crystalline ferric oxides, goethite and hematite by altering the conditions of supersaturation within the leach solution. Controlling the ferric concentration below the level of supersaturation produces the more crystalline iron oxides as expressed above.  These crystalline species are normally precipitated from the ferrous state following reduction of the ferric ion, whereas direct precipitation from the ferric state leads to jarosite and ferrihydrite precipitation. This is an integral feature of the industrial Paragoethite process where continuous addition of concentrated acidic ferric liquor acts to promote dilution and negates the need for an iron reduction step in a zinc refinery (Loan et al. 2006a). Addition of hematite seed to a Fe precipitation process also aids in reducing the amount of poorly formed hematite and ferrihydrite in sulphate leach solutions (Dutrizac and Chen 2011a). This is practiced industrially at the  28 Morenci Cu operation, AZ, USA. Seeding of the pressure leaching autoclave with hematite from the residue provides a ‘preferential nucleation site for newly formed or liberated hematite, minimising ferrihydrite formation (Marsden et al. 2002).   Ferrihydrite is capable of incorporating metal values from hydrometallurgical leach solutions by co-precipitation and adsorption as discussed by several authors (Martinez and McBride 1998a; Loan et al. 2005; Benjamin and Leckie 1981). Ferrihydrite has a high affinity to adsorb Cu2+ and other divalent metal ions to its surface (Li 2008). The largely amorphous nature of ferrihydrite generates a large surface area (typically 200 m2 g-1 to 700 m2 g-1) and a high reactivity, making it especially amenable to adsorption (Jambor and Dutrizac 1998). Adsorption capacity can be even larger for freshly precipitated ferrihydrite over aged ferrihydrite and when adsorption occurs with precipitation (Fuller, Davis, and Waychunas 1993). Cu2+ is found to co-precipitate to a stable solid phase with Fe3+ during the formation of ferrihydrite (Martinez and McBride 1998a). These authors suggest that Cu is segregated within aggregates of the Fe oxides due to the closeness in ionic radius of Cu2+ and Fe3+ though homogeneous distribution throughout the structure is also possible. It is likely that Cu can be structurally incorporated at the expense of Fe in the aged crystalline product resulting from ferrihydrite. True solid solution of Cu in ferrihydrite in a soil-type system (23°C, ambient pressure) was not observed with the take-up of Cu2+ being the result of segregation and adsorption (Martinez and McBride 1998b).  2.2.4 Metal adsorption to crystalline iron oxides and ferrihydrite  Adsorption of common base metal cations to iron oxides in soil and wastewater aqueous systems is well established though scant information is available for metallurgical systems. Studies  29 conducted by several authors (Rodda, Johnson, and Wells 1993; Moon and Peacock 2013; Li 2008; Benjamin and Leckie 1981) have modelled the adsorption of divalent metal cations to the surface of iron oxyhydroxides. Adsorption of Cu2+ to amorphous iron oxyhydroxide (ferrihydrite) was shown to occur readily without a maximum adsorption limit being achieved (Benjamin and Leckie 1981). These authors suggest that the binding of Cu2+ to the surface followed a heterogeneous adsorption model that releases two protons and consists of bidentate bonding [21] or simultaneous adsorption and hydrolysis [22]. Ni2+ was not studied in this study but can be assumed to exhibit similar behaviour.  !"#$%&'&$!!"#$%#&:!!!!!2!"# +!"!! = !!!!!" + !2!! [21] !"#$%&'($) ℎ!"#$%!&'&:!!"# +!"!! + !!! = !"#$"% + 2!!![22]  SOH represents a protonated surface site and S2O2Me and SOMeOH represent surface sites occupied by a metal ion.  Solution pH and temperature are always a crucial factor during metal adsorption. Cu2+ showed a low tendency to adsorb at pH 4 and higher pH favoured adsorption with pH 6 achieving 100% adsorption (Benjamin and Leckie 1981). Similar results were achieved in a stratiform copper deposit where no adsorption of Cu2+ was noticed below pH 5 (Rose and Bianchi-Mosquera 1993). Cu2+ from low strength solutions (10-4 M) did not adsorb to synthetic goethite at pH 3 at 70°C though it was suggested that higher temperatures noticeably lower the pH of the adsorption edge (Rodda, Johnson, and Wells 1993). These studies are conducted in ambient temperature unsaturated natural soil systems though do discuss fundamentals that may be applicable to a hydrometallurgical application. Adsorption reactions for many divalent cations are endothermic so the reaction product (adsorbed species) is favoured at higher temperatures. Furthermore, the  30 surface charge for goethite at a given pH decreases and hydrolysis of the cations proceeds to a greater extent as temperature increases (Rodda, Johnson, and Wells 1993). It follows that an increase in temperature will reduce the electrostatic repulsion between the surface and the adsorbing species, allowing adsorption to occur more readily.   There is a strong correlation between an increase in Cu2+ adsorption to goethite with increasing ionic strength (Jiang, Xu, and Li 2010), a factor that is significant in the supersaturated environment of the CESL and Vale pressure leach. The correlation between adsorption and ionic strength is somewhat counter-intuitive as under acidic conditions goethite and other Fe oxides have a net positive surface charge that increases with increasing ionic strength (Jung, Cho, and Hahn 1998). Increasing Cu2+ adsorption therefore cannot be explained by an increase in electrostatic attraction between Cu2+ ions and the goethite surface. In fact, the reverse is true as a decrease in the magnitude of the positive surface charge results in increased Cu2+ adsorption.  The model initially proposed by Barrow et al. (1980) for anionic adsorption can explain the observed phenomena. An increase in the counter ion in the Cu2+ adsorption plane causes a decrease in the potential at the adsorption plane. The decrease in potential in the surface and diffuse layers favours specific adsorption of Cu2+ as it reduces the repulsive effect of the positively charged goethite surface.    Surface complexation modelling can largely explain the adsorption of Cu2+ and other metal cations to crystalline iron oxides and ferrihydrites in conditions that do not fundamentally favour adsorption. Cu2+ forms a strongly bound coordinative complex with surface O− sites on iron oxides (Jung, Cho, and Hahn 1998). These can be bidentate and analogous to equation 21 or  31 unidentate. This is confirmed by the modelling of Moon and Peacock (2013) who found ~3.5% Cu in ferrihydrite was adsorbed to inner-sphere surface (Δz0) as bidentate edge sharing complexes with two FeOH functional groups. The remaining adsorption occurred to the same complex at the electrostatic 1-plane (Δz1). These planes are inner-sphere and within the ferrihydrite-solution interface (Δz2).   A combination of these fundamental studies of adsorption in dilute, low temperature aqueous systems can be applied to a hydrothermal leaching process like the CESL/Vale pressure leach. Conditions of higher temperatures, supersaturation, favourability of hydrolysis and large-scale availability of surface sites from abundant hematite, goethite and ferrihydrite indicate that adsorption of Cu2+ (and Ni2+) can occur. The adsorption is likely tenacious as the presence of surface sulphate on precipitated hematite may provide further adsorption sites for Cu2+ and CuSO4 is thought to be strongly occluded to hematite and other iron oxide precipitates (Dutrizac and Chen 1993; Dutrizac and Chen 2012b).  2.2.5 Metal sulphate adsorption and occlusion to iron oxides Sulphate has been shown to tenaciously adsorb to iron oxides precipitated from ferric sulphate solutions.  Electron microscopy shows that SO42− is uniformly distributed through precipitated hematite crystallites and amorphous ferric oxides. Incorporation by adsorption is confirmed by displacement with strong anionic solutions (Dutrizac and Chen 1993). Around 1% SO42− is found uniformly adsorbed on all hematite precipitates from Fe2(SO4)3-CuSO4 and Fe2(SO4)3-NiSO4 solutions reacted in a pressure leaching environment (Dutrizac and Chen 2012b). Significant Cu and Ni were associated with the precipitated hematite and are likely the result of a combination  32 of incorporation within the crystal structure and occlusion of the metal sulphates. Tiny dispersed particles of nickel sulphate monohydrate (NiSO4.H2O) were noticed though no evidence of CuSO4 was noticed due to analytical difficulties. It was suggested by these authors that Cu2+ was not absorbed at pH <3 from other studies (Rose and Bianchi-Mosquera 1993). As suggested earlier these studies are conducted under markedly different conditions and Cu2+ and Ni2+ adsorption may be possible. The effect of sulphate adsorption on the iron oxide surface chemistry may allow for adsorption sites for Cu2+ and Ni2+ or the metal sulphates present in the leach liquor could adsorb as discrete occluded species on the individual crystallites and/or within the agglomerates as suggested by Dutrizac and Chen (2012b).   Sulphate is adsorbed to goethite and hematite by ligand exchange with exposed surface FeOH groups. The mechanism of adsorption is taken from Parfitt and Smart (1978). The reaction occurs when at pH <8 where FeOH are protonated to give FeOH+. The sulphate exchanges only with one coordinated FeOH and FeOH2+ group and each sulphate replaces two of these groups to give the Fe-O-S(O2)-O-Fe complex on the surface of the goethite, hematite and ferrihydrite. The mechanism is given in equation 23 (Parfitt and Smart 1978).  [23] The formation of the Fe sulphate surface complex is important in the context of Cu/Ni sulphide medium temperature pressure leaching. The structure of the complex suggests that a positively charged surface site and a neutral (or δ+) site are replaced with a single complex with two neutral or (δ-) surface sites. These sites may be suitable for Cu2+ or Ni2+ specific adsorption though it is suggested that the complex is uncharged (Parfitt and Smart 1977). Adsorption of sulphate on  33 hydrous oxides shifts the point of zero charge to higher values (Breeuwsma and Lyklema 1973). Such a shift also favours a reduction in specific adsorption of Cu2+ and Ni2+.  This same study suggested that a restabilisation region of Ba(NO3)2 was observed where equivalent adsorption of Ba2+ also occurred. This was not noticed in K2SO4 or monovalent electrolytes though the restabilisation of hematite surfaces with both phosphate and a divalent ion suggests that restabilisation is likely possible by divalent Cu and Ni sulphates. The exchange phenomena of surface OH− and H2O with SO4 likely explain the occlusion of CuSO4 and NiSO4.H2O noticed by Dutrizac and Chen (2012b). Specific adsorption of individual Cu2+ and Ni2+ is not found specifically in the literature under sulphide hydrothermal leach conditions though it may also occur with the noted chemisorption and co-precipitation discussed earlier.   2.3 Characterisation of the CESL and Vale leach residues  Residues similar to S1 and S3 were characterised to determine the source of Cu losses from the CESL medium temperature leach process (Sahu and Asselin 2011). Poor agreement between total chemical analyses by inductively coupled plasma-mass spectrometry (ICP-MS) and those calculated from mineralogical analysis by quantitative X-ray diffraction (QXRD) indicated the presence of iron-rich amorphous phases. SEM-EDX mapping of the residues indicated that the gelatinous, amorphous iron-rich oxide phases contained 5 to 8% Cu and therefore contribute significantly to copper losses in the CESL residue. This Fe-amorphous phase can be assumed to be ferrihydrite as the amorphous ferric oxyhydroxide produced in a number of hydrometallurgical operations has been suggested to be ferrihydrite (Jambor and Dutrizac 1998). The remaining losses are likely a combination of surface adsorption, unleached sulphides,   34 interstitial incorporation in crystalline iron precipitates and occlusion of metal sulphates. The mechanism for inclusion within iron oxides is discussed in the previous section.   Preliminary investigations of the CESL copper residue found Cu was associated with both the amorphous ferrihydrite and hematite phases. The exact nature of this association was unable to be determined due to the inherent limitations with the spectroscopic equipment used. It appears that precipitated iron hydroxides coat the surface of some chalcopyrite particles causing them to become entrapped inside iron oxide phases and preventing leaching (Sahu and Asselin 2011). SEM-EDX mapping of the amorphous iron oxyhydroxide phases revealed the presence of further co-precipitated copper. To overcome the limitations of spectroscopic methods for determining the association of copper in the CESL and Vale residues a chemical extraction method was used. This method revealed there was a five-fold increase in the copper associated with the amorphous ferrihydrite phase than the crystalline Fe phase. The low proportion of goethite in the residues suggests that the presence of copper hinders the formation of goethite from ferrihydrite. This occurs due to structural incorporation of Cu in ferrihydrite that stabilizes the precipitation and hinders the subsequent transformation to goethite. The transformation of ferrihydrite is discussed in the next section.   2.3.1 Identifying ferrihydrite in CESL and Vale leach residues  The presence of other amorphous and crystalline phases in hydrometallurgical residues masks the low intensity XRD pattern of ferrihydrite making it easily overlooked (Loan et al. 2002a). A rigorous, multi-faceted approach is therefore needed to identify the presence of ferrihydrite. This approach utilizes a variety of instrumental techniques such as X-ray diffraction (XRD),   35 Mössbauer spectroscopy, scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Only approximate quantities can be generated utilizing these techniques and there is no way to indicate the distribution of any recoverable metals that may be incorporated or adsorbed in the residue (Loan et al. 2002a). To accurately quantify and assess the distribution of valuable metals within the CESL and Vale residues that overcomes the inherent inaccuracies of spectroscopic methods a two-stage selective extraction procedure was developed (Sahu and Asselin 2011). This technique has been used to systemically detail the characteristics and components of the CESL and Vale residues and account for the presence of ferrihydrite prior to the commencement of the present study.   The first extraction stage uses a solution of 0.25 M hydroxylamine hydrochloride (NH2OH.HCl) and 0.25 M hydrochloric acid (HCl) at 50°C to dissolve the ferrihydrite phase but not the crystalline iron phases (Chao and Zhou 1983). This reagent also dissolves most of the gypsum (CaSO4.2H2O) and anhydrite (CaSO4). The residue remaining from the first stage is treated with a 4 M hydrochloric acid solution at 95°C to dissolve all the crystalline iron oxides (Leinz et al. 2000). The extraction time is 30 minutes for both stages. Analysing the residue and reactant solution from each extraction stage for Ca, Fe, Cu and Ni by ICP-MS distinguishes the adsorbed and co-precipitated metals associated with each phase. Copper is also released into solution when these phases are dissolved. The major crystalline iron phase associated with copper losses is hematite rather than goethite.  Utilising the calcium analysis enables the calculation of the amount of gypsum and anhydrite dissolved. Coupled with the mass loss of the residues following the first treatment stage, the total   36 amorphous iron content (Feam(%)) can be calculated. This is given by the equation below:  !"!" % = 100 ∗ (! − !! − !)/! [24] where R is the total mass of the residue, R1 is the mass of the residue produced after the 1st stage of extraction and C is the calculated total mass of gypsum and anhydrite dissolved in the first stage of extraction.   The combination of 0.25 M hydroxylamine hydrochloride and 0.25 M hydrochloric acid at 50°C with 30 minutes of extraction time is the most desirable extractant for amorphous iron oxides. Less than 1% of the total iron is extracted from crystalline iron oxides and shows close agreement with the traditional use of Tamm’s reagent (0.175 M (NH4)2C2O4 – 0.100 M H2C2O4) in darkness and is a simpler process (Chao and Zhou 1983). Acidified hydroxylamine hydrochloride solution acts to both reduce and dissolve the more readily soluble species, represented by the more reactive iron oxide species (Chao 1972). The large surface area of ferrihydrite makes it the most reactive iron phase for dissolution.   Dissolution of ferrihydrite by hydroxylamine hydrochloride occurs as a reduction of Fe3+ to Fe2+ similar to the theoretical reaction mechanism posited for the dissolution of manganese oxide by reduction of Mn3+ to Mn2+ with hydroxylamine hydrochloride (Neaman et al. 2004).  !"!!!!!! + 2!"!!"! + 2!! → 2!"!! + !! + 5!!! [25] Protons consumed by the reaction can be produced by hydrolysis of NH2OH-HCl in water. !"!!"! + !!! → !"!!!" + !! [26] The addition of hydrochloric acid to the first extractant stage ensures plentiful H+ for the reaction without the need for hydroxylamine hydrochloride hydrolysis that would deplete the reactant.   37 We can assume that ferrihydrite dissolves with hydroxylamine hydrochloride by the same theoretical mechanism as proposed above.  !"!!!. 9!!! + 2!"!!"! + 2!! → 2!"!! + !! + 14!!! [27]  Acidified hydroxylamine hydrochloride is selective for poorly crystalline and amorphous iron oxides due to factors related to the mineral itself, such as the mineralogical form, degree of crystallinity and particle size or surface area (Chao 1972). Stronger concentrations of hydroxylamine are therefore less selective, hence the need for appropriate acidification. The second stage of extraction uses strong hydrochloric acid to dissolve the crystalline iron phases. Hydrochloric can dissolve iron oxides of varying degrees of crystallinity, depending on the temperature and concentration of the acid. It is also useful at dissolving any carbonate minerals and replaces any metal ions adsorbed on inorganic or organic colloids (Chao 1984).  2.4 Ageing of the CESL and Vale residues Ageing is a transformation process where solid materials become more structurally ordered over time and is often conducted at elevated temperatures. Ferrihydrite ageing involves the use of elevated temperatures, sufficient reaction time and a variety of solution conditions for the initially formed colloidal polynuclears of ferrihydrite to morph into crystalline goethite and hematite. Elevated temperatures greatly assist the ageing process as although ordering may be thermodynamically favoured, the kinetics are often slow. Ageing for 970 days at room temperature at pH <7 left 2-19% ferrihydrite remaining and less than 2% at pH >7 (Schwertmann and Murad 1983).  The remaining ferrihydrite was converted to either goethite or hematite. At 92°C, 116 hours were needed for conversion of ferrihydrite to hematite (Johnston and Lewis   38 1983). The ferrihydrite species (2-line and 6-line) and the FeO(OH) phases resulting from ferrihydrite transformation are the outcome of specific crystallization conditions determined by kinetics (Schwertmann, Friedl, and Stanjek 1999).   Utilizing an ageing process may release some of the Cu and Ni associated with the ferrihydrite in the CESL and Vale residues during the transformation process to hematite or goethite. Ferrihydrite transformation to hematite and goethite incorporates Cu into the crystal lattice, though the presence of adsorbed foreign ions such as Cu2+ determines the end-member of the transformation (Cornell and Giovanoli 1988). The presence of Cu2+ in ferrihydrite hinders the formation of goethite by limiting the dissolution prior to its re-precipitation as goethite. This favours hematite by an alternate mechanism. Sahu and Asselin (2011) suggest this produces significantly more hematite than goethite in residues produced under CESL conditions. Hematite is formed by dissolution and rearrangement of ferrihydrite and goethite (Cornell, Giovanoli, and Schneider 1989). During this process Cu can be released into solution and also incorporated into the hematite (Sahu and Asselin 2011). Complete dissolution of ferrihydrite is likely required to prevent incorporation in the crystalline iron oxides species and this may not be achieved by ageing.   2.4.1 Transformation of ferrihydrite  Ferrihydrite is thermodynamically unstable and transforms into the more stable, crystalline oxides, goethite (α-FeO(OH)) and hematite (α-Fe2O3) with time. It is well known that hematite (α-Fe2O3) is the end product of the thermal decomposition of ferrihydrite, iron-oxyhydroxides or iron salts in an oxidative atmosphere (Cornell, Schwertmann, and Editors. 1996). Goethite   39 (FeOOH) transforms in acid media into α-Fe2O3 with sufficient time and temperature (Cornell, Giovanoli, and Schneider 1989). Transformation to hematite by ageing requires a week of hydrothermal treatment at 160-180°C (Schwertmann, Friedl, and Stanjek 1999).    The transformation of ferrihydrite to goethite and/or hematite is a first-order reaction where the rate of transformation is proportional to the amount of ferrihydrite left for transformation (Cornell, Giovanoli, and Schneider 1989). Ferrihydrite can be considered to exist in equilibrium with monovalent Fe(III) ions where the activity of these ions determines the product of transformation. The solubility product of ferrihydrite (Ksp = 10-39) is higher than those of goethite (Ksp = 10-41) and hematite (Ksp = 10-43) and should spontaneously transform to these minerals (Cudennec and Lecerf 2006). This occurs in salt-free aqueous systems below 100°C. Solubility products for the iron oxyhydroxides are calculated from the activities of Fe3+ and OH− according to the work of Langmuir (1969). Ferrihydrite can also transform in the solid-state by dry heating at higher temperatures (Ristić et al. 2007).  The transformation process to each of these iron oxides occurs by competing mechanisms as both species contest a limited number of growth sites. As such, favourable conditions for hematite formation limit the formation of goethite and vice versa (Cornell, Giovanoli, and Schneider 1989). Hematite is produced by initial rearrangement and dehydration within ferrihydrite aggregates while goethite formation occurs by a dissolution and re-precipitation process from monovalent Fe3+ ions such as Fe(OH)2+ and Fe(OH)4− (Liu et al. 2007; Schwertmann and Murad 1983). Goethite is favoured in conditions where the Fe3+ equilibrium solubility with ferrihydrite is maximised and hematite is favoured when Fe3+ solubility is   40 minimised. Higher temperatures favour hematite as this promotes dehydration of the H2O molecules from ferrihydrite (Schwertmann and Murad 1983).   The dissolution of ferrihydrite is decisive in favouring one ferrihydrite transformation mechanism over the other. The dissolution of ferrihydrite forms various soluble ionic species depending on the solution pH that act as precursors for obtaining goethite, hematite or both. At low pH (pH <2) and pH (pH 7-8) hematite formation is favoured while moderate acid (pH 2-7) and basic pH favours goethite (Cudennec and Lecerf 2006). At moderate acidity of pH 2-7 the formation of the species, Fe(OH)(H2O)52+ and Fe(OH)2(H2O)4+ is favoured. Fe3+ solubility is maximised at pH 3-4 leading to higher concentrations of these two species and maximizing the yield of goethite precipitated. At basic pH iron dissolves to form the soluble species of Fe(OH)4(H2O)2−, Fe(OH)5(H2O)2− and Fe(OH)63− which also form goethite. Above pH 12, only goethite is formed due to the presence of large amounts of OH−. The presence of aquahydroxo and hydroxo ions leads to goethite as olation and oxolation processes of OH− and H2O ligands form the necessary Fe-OH-Fe and Fe-O-Fe bridges required for goethite formation. The authors (Cudennec and Lecerf 2006) believe that these five species are responsible for the formation of goethite and partially explain the inhibition of hematite formation. Schemes for the dissolution of ferrihydrite in acidic and basic solution prior to the crystallization of goethite are proposed by these authors and presented below. Acidic solutions: 0 ≤ n < 3 5!"!!!. 9!!! + 30− 10! !!!! + 6+ 10! !!! ↔ 10[!" !" !(!!!)(!!!)] !!! ! [28]!Alkaline solutions: 3 < n ≤ 6 5!"!!!. 9!!! + 10! − 30 !"! + 66− 10! !!! ↔ 10[!" !" !(!!!) !!! ] !!! !![29]   41 Hematite produced from ferrihydrite is maximised at pH values where goethite formation is inhibited and can also be explained in terms of solubility. The presence of monovalent Fe3+ ions was most effective for goethite formation while hematite was favoured at the ferrihydrite point of zero charge (p.z.c pH ~8) where Fe3+ concentrations were minimized (Schwertmann and Murad 1983). At neutral pH (pH 7-8) the concentration of H3O+ and OH− concentrations are too weak to dissolve sufficient ferrihydrite for goethite formation. The main dissolved ferrihydrite entity is the complex Fe(OH)3(H2O)3 but its concentration is too low for precipitation so solid state transformation to hematite remains the only option (Cudennec and Lecerf 2006). At low pH (pH < 2) iron dissolves to the complex hexaaqua iron(III) form, Fe(H2O)63+ that does not bridge with OH− and H2O to form goethite and instead favours hematite formation by the competing mechanism. The transformation of ferrihydrite (Fe2O3.9H2O) to hematite (αFe2O3) occurs through the reagent scheme given below.  5!"!!!. 9!!!! ↔ 5!!!"!!! + 9!!! [30] Nucleation and growth of α-Fe2O3 from ferrihydrite occurs as both species have a similar hexagonal close-packed anion sublattice structure. Prior to nucleation, aggregation of ferrihydrite particles into larger masses forms the hexagonal crystal shape of α-Fe2O3 and promotes ordering. A combination of dehydration and condensation reactions occurs to facilitate the nucleation and growth from these large aggregated masses. The mechanism involves the loss of a proton from an OH group followed by the migration of the proton to another OH group to form a linkage and elimination of a water molecule. Charge balance is restored by the redistribution of Fe3+ within the cation sublattice to form α-Fe2O3 (Cornell, Giovanoli, and Schneider 1989). This direct transformation of ferrihydrite to hematite (no transitional phase) has been observed by other   42 authors (Johnston and Lewis 1983; Schwertmann, Friedl, and Stanjek 1999) and is largely facilitated by the formation of precursor aggregates. These aggregates form the hematite nuclei and are then fed by more precursors by a short-range solution process. The flocculating effect of solution pH close to the p.z.c and by higher ionic strengths improves the aggregation and promotes further hematite formation.   Figure 3: Mechanism of ferrihydrite precipitation and growth (Dousma, den Ottelander and de Bruyn 1979)  2.4.2 Effect of dissolved ions during ferrihydrite ageing  Co-precipitated Cu with ferrihydrite stabilizes the dissolution and therefore favours transformation to goethite far more than if the Cu is added to the system following ferrihydrite precipitation (Martinez and McBride 1998b). This is due to the incorporation of Cu in the ferrihydrite structure rather than by surface adsorption following ferrihydrite precipitation. The structural incorporation of Cu dramatically slows the dissolution of ferrihydrite and therefore the transformation to goethite. Goethite forms readily from ferrihydrite to which Cu is added indicating that soluble Cu species do not interfere in the nucleation of goethite to any extent (Cornell and Giovanoli 1988). Goethite and hematite form by competing mechanisms so the retardation of ferrihydrite dissolution prior to the formation of goethite indirectly favours hematite. The mechanism of hematite growth by internal rearrangement is unaffected by the presence of incorporated Cu.    43 The kinetics of the ferrihydrite transformation mechanisms is affected by the presence of certain metal ions and other additives. The majority of these additives retard the crystallization mechanism of ferrihydrite resulting in an increase of transformed hematite relative to goethite by the competing rearrangement mechanism. Additives that adsorb on ferrihydrite have a far stronger retarding effect than those that act only in solution (Cornell, Giovanoli, and Schneider 1989). Adsorbing species retard goethite formation by slowing the dissolution of ferrihydrite and thus lessening the Fe3+ in solution for nucleation. Metal ions such as Cu2+, Ni2+ and Zn2+ have been shown to adsorb and occlude within precipitated iron oxides (Rodda, Johnson, and Wells 1993; Chen et al. 2006). Formation of hematite is slowed by adsorbing species either linking ferrihydrite particles to form an immobile chain or through reducing particle aggregation by increasing the negative surface charge with anionic surface complexes. Sulphates have been shown to adsorb to the surface of precipitated hematite (Dutrizac and Chen 1993) and the occurrence of a tenaciously bound surface anion may affect both the precipitation of hematite and ferrihydrite transformation mechanisms.   The presence of sulphate ions promotes hematite formation from ferric oxyhydroxides in solution. The presence of sulphate increases the formation rate of the dimers (FeOH)24+ and lowers the pH of the first formation of precipitated iron oxide phases (Dousma, den Ottelander, and de Bruyn 1979). The mechanism of precipitation proposed below is affected by sulphate as the formation of iron-sulphate complexes lowers the concentration of Fe3+ promoting nucleation and growth at lower supersaturations (steps 2 and 3). The authors also detail that the presence of sulphate prevents the oxolation reactions that turn ferrihydrite to goethite, though the precise mechanism for this is not known. Chloride has also been shown to complex with Fe3+ during   44 ferrihydrite precipitation and become incorporated in the colloidal FeOOH phase (Dousma, Van den Hoven, and De Bruyn 1978).   At lower pH values and shorter residence times, where Fe3+ hydrolysis is not complete, the mechanism of formation of the various oxides or the material produced is less predictable. The mechanism of Fe3+ hydrolysis and the end products of ferrihydrite, goethite and hematite are sensitive to temperature. Johnston and Lewis (1983) suggest 85°C represents a critical temperature where all three products are made. At this temperature several diverse processes are occurring and it is difficult to determine whether ferrihydrite is an intermediate product for the end members goethite and hematite. At 92°C no goethite is formed so it is possible to study the transformation of ferrihydrite to hematite in aqueous medium at temperatures greater than this. Ageing temperatures of 95°C were chosen for this reason and because the experiments could be conducted at atmospheric pressure.  The catalyzing effect of Fe2+ on the formation of goethite and hematite from ferrihydrite is detailed by Liu et al. (2008) and Liu et al. (2007).!The presence of adsorbed Fe2+ aids in the dissolution of ferrihydrite to the monomers Fe(OH)2+ and Fe(OH)2+. These monomers increase the supersaturation of Fe3+ that precipitates to lepidocrocite before transformation to goethite. A temperature dependent mechanism for ferrihydrite conversion to goethite is given by Liu et al. (2007).! !"##$ℎ!"#$%&! !"!! !"(!")!!!"(!")!! !"#$%& ! − !"##$!"#$%&'( !"! !" !!! ↔ !"!(!")!!! ↔ !"#$%&'()*+(",↔ !"!!!(!")!!!!!!!! ! → ! − !"##$  [31]!  45 In the proposed catalytic dissolution of ferrihydrite, ferric species are dissolved to the two monomers and the subsequent two precipitation processes compete with each other. At low temperature the direct formation of γ-FeOOH dominates. High temperatures favour polymerization and formation of α-FeOOH by the indirect mechanism. Even higher temperatures (~100°C) favour the formation of hematite by solid-state transformation of goethite and dissolution/reprecipitation of ferrihydrite. The results show that a high temperature of around 100°C, a pH range of 5-9 in the presence of Fe2+ leads to rapid hematite precipitation. Lower temperatures and pH leads to goethite transformation from ferrihydrite by either of the two mechanisms (Liu et al. 2007). The presence of Fe2+ in the ageing solution may also desorb adsorbed Cu2+ and Ni2+ from the residue as well as promote hematite or goethite formation.    46 Chapter 3: Experimental Materials and Methods 3.1 Residue collection Sulphide pressure leach residue samples of the CESL Cu process (S1) and the Vale Long Harbour Ni process (S3) were provided and analysed. The S1 sample is a composite solid residue of several Teck Cu flotation process plant concentrates treated by the pilot plant at the CESL facilities in Richmond, BC, Canada during 2010. Grab samples were taken from the 5th and last counter-current decantation (CCD) stage and filtered before being sent as moist filter cakes to UBC. The samples were provided at the beginning of the project in April 2011 and left as a moist cake in a glass jar away from sunlight in a cupboard below the laboratory bench top. The S3 sample was collected from the discharge of the last CCD stage of the demonstration plant for the Vale Long Harbour Ni process in Argentia, Newfoundland, Canada. Activities of the demonstration plant ended in 2008 and the sample was collected during one of the final runs. It was filtered and dried at UBC during 2012. The dry sample was left in a glass jar away of sunlight below the laboratory bench top. The Munsell colour of S1 was 7.5YR 5/6 (strong brown-dry) and sample S3 was 10R 3/6 (dark red-dry). The samples are expected to have aged over the 4 and 3 years, respectively, that they were kept in storage between collection and use for experiments. This means it is likely that some of the metastable ferrihydrite material had degraded and some of the associated metal losses may have been more readily available for recovery from the bulk residue.  Although, given the results presented below, the extent to which this occurred is expected to be very minor.     47 3.2 Sample analysis Chemical analysis of the solid residues was conducted by CESL at their facilities in Richmond, BC. Samples were dried overnight in an oven at 60°C. Dried samples were lightly ground with mortar and pestle and a portion was digested in aqua regia prior to analysis by inductively coupled plasma mass spectrometry (ICP-MS). This method gives complete elemental analysis of the residue and is the only way to definitively determine the precise elemental constituents of the residues. Ageing solution samples were sent to CESL in 10 mL sealed, plastic vials and analysed utilising the same ICP-MS spectrometer as the dissolved residue solution samples. Solution samples were diluted with the appropriate matrix as required to meet the detection band of the analyser.   3.2.1 Sequential extraction The extractions were conducted by stirring residue to a 1:50 solids to solution ratio in acidified hydroxylamine hydrochloride solution at 50°C for 30 minutes. Usually 4 g of residue was agitated with a stirrer bar in 200 mL of solution. The solution was made up to a concentration of 0.25 M with 1 N hydrochloric acid provided by VWR Analytical and deionised water. Solid (~3.5 g) granulated hydroxylamine hydrochloride was dissolved into the solution to make a 0.25 M concentration. Solid hydroxylamine was supplied by VWR from EMD Millipore AA and is >96% pure. A maximum of 0.005% sulphur as SO4 and 5 ppm Fe is present in the hydroxylamine hydrochloride solids provided.    48 3.2.2 Quantitative X-ray diffraction analysis Quantitative phase analysis using QXRD was conducted at the Centre of Earth and Ocean Sciences at UBC. Samples were ground in ethanol to a fine powder (<10 µm) optimum for X-ray analysis using a vibratory McCrone Micronising Mill for 7 minutes. Corundum was added to the ground powder to a quantity of 10% by mass prior to analysis. Corundum acts an internal standard for quantification of the relative amounts of crystalline and amorphous phases. Higher quantities of corundum could be used to generate more accurate results and better differentiate the amorphous phase though this level of accuracy was deemed unnecessary. Continuous-scan X-ray powder diffraction data were collected over a range 3-80° 2θ with CoKα radiation on a Bruker D8 Focus Bragg-Brentano diffractometer. The International Centre for Diffraction Database PDF-4 and Search-Match software by Bruker AXS were used to analyse the diffractograms. Raw X-ray diffraction data was refined using the Rietveld method by the Rietveld program Topas 4.2. The Rietveld method for analysis is considered necessary since the amorphous phases present in the leach residues do not contribute to the diffraction peaks in QXRD analysis. The proportion of the crystalline phases is corrected by renormalizing the data by Rietveld phase analysis (Sahu and Asselin 2011).   3.2.3 SEM/EDS analysis Original residue samples were investigated by scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS). SEM was used to identify the mineralogy and source of Cu and Ni losses within the residues. For these analyses the residues were mounted onto 26 x 46 mm, 200 µm thick glass slides with an epoxy resin, polished and carbon coated. A Hitachi S-300N scanning electron microscope was used for the analysis. EDS point scans were utilised to assess   49 the amount of Cu, Ni, Fe and S present in the residues. Relative accuracy of the EDS point scans was quoted as +/- 15% (Kabel 2015).  C and O were subtracted from the analysis as these are present in the epoxy and affect the relative proportions of elements at a particular point.   3.2.4 Electron microprobe analysis  Hematite particles of the Vale residue (S3) were investigated to determine their elemental composition and association with Cu and Ni by electron-probe microanalysis (EMPA). Analysis was conducted at the Centre for Earth and Ocean Sciences at UBC. A collection of 20 random particles of varying sizes was chosen for the analysis, 7 of these particles were hematite rimmed with lighter coloured structures that appeared to be sulphur based and the remaining 13 were single, homogeneous hematite grains with no appearance of any rims. Electron-probe micro-analyses of iron oxides were done on a fully automated CAMECA SX-50 instrument, operating in the wavelength-dispersion mode with the following operating conditions:  excitation voltage, 20 kV; beam current, 40 nA; peak count time, 20 s; background count-time, 10 s; spot diameter, 1 µm.  Data reduction was done using the 'PAP' φ(ρZ) method (Pouchou and Pichoir 1985). For the elements considered, the following standards, X-ray lines and crystals were used: arsenopyrite, SKα, PET; hematite, FeKα, LIF; synthetic Ni2SiO4, NiKα, LIF; Cu element, CuKα, LIF.   3.3 Ageing experiments 3.3.1 Preliminary ageing experiments Initially the residues were aged in water at 95°C for 24 hours to assess the effect of elevated temperature and extended time on the residue phase composition. Preliminary ageing   50 experiments were conducted on dry 4 g samples of S1, S1 R1, S3 and S3 R1 in 200 mL of distilled water. S1 R1 and S3 R1 represent acidified hydroxylamine hydrochloride extracted (stage 1 sequential extraction) CESL Cu-rich (S1) and Vale Ni-rich (S3) residues. The initial pH of the slurry at the beginning of the experiments was pH 2.3 for S1, pH 2.9 for S3, pH 1.9 for S1 R1 and pH 2.1 for S3 R1. A low solids concentration of 2% w/w was utilised to provide direct comparison with the sequential extraction procedure and ensure maximum washing of the residues. A glass jacketed reactor vessel connected to a hot water bath contained the slurry. Agitation of the slurry was by a magnetic stirrer bar rotated by a Corning 120 V stirrer. A rubber bung with a thermometer sealed the reactor vessel to minimise evaporation and heat loss. At the completion of the ageing time the slurry was filtered by gravity using a ceramic filter cone and 55 mm diameter Whatman® No 42 ashless filter paper. After filtration, the moist weight of the sample and paper were taken to determine the solution lost to moisture before the solids were dried overnight in an oven at 60°C. Dry sample weights were taken to two decimal places on a balance and the residue mass loss was calculated by difference from the original residue mass. This set-up is shown in Figure 4 below and is common for all further ageing experiments though an extra reactor was added for the later batch ageing experiments to double the experimental turnover time. Ageing was repeated once for each sample. An average composition of the samples following ageing was calculated from QXRD analysis.   Ageing was repeated under the same conditions and the filtered and dried solid samples were subsequently treated by sequential extraction with acidified hydroxylamine hydrochloride to assess any improvements in residue stability. These solids were filtered, dried and weighed again. Solid aged and solid sequentially extracted residue samples were sent for QXRD phase   51 analysis. Phase analysis by QXRD was conducted to determine the nature of the effect ageing had on the residues compared with removing the amorphous Fe and ferrihydrite phases with acidified hydroxylamine hydrochloride. Chemical concentrations of Cu, Ni, Fe, Ca and S in the resultant ageing solution were gained from ICP-MS analysis.  A pH measurement was recorded before and after ageing using an Oakton Instruments pH 5+ series pH meter fitted with an Oakton Instruments glass, refillable, double-junction, pH electrode. The electrode has built in automatic temperature correction to correctly read the pH at temperatures up to 100°C and uses an Ag/AgCl reference cell. The electrode was regularly filled with KCl reference solution and calibrated daily to ensure accurate readings.           A temperature of 95°C was chosen as the ageing experiments could be conducted at atmospheric pressure utilising basic laboratory equipment. This temperature was considered sufficient as it is above the 92°C utilized in ageing experiments by Johnston and Lewis (1983). In those experiments some ferrihydrite transformed in solution to hematite within one hour. At 92°C, 116  Figure 4: Ageing experiments equipment set-up   52 hours were needed by Johnston and Lewis to convert all the ferrihydrite to hematite. Despite this, an ageing time of 24 hours was considered reasonable since the cost of tankage for 116 hours residence time would be too excessive for an industrial operation such as Vale Long Harbour to consider.   3.3.2 Kinetics of ageing at 150°C To better understand the kinetics of the ageing process an investigation was conducted at 150°C in a laboratory scale autoclave. The ageing time was varied to assess the ageing equilibrium. A 1000 mL, 4520 model, stirred, bench top Pressure Reactor system from Parr Instruments was used for the ageing experiments. The reactor system includes dual 2.28” 6-blade impellers, cooling coils, a titanium thermowell with a T316 stainless steel thermocouple for temperature measurements and a 4.5” manual pressure gauge for pressure readings. All internal fittings are made of corrosion resistant titanium. The 1 L titanium reactor vessel is lined with a 1 L glass, cylindrical liner for the ageing experiments. A 6-compression bolt, split ring with a FKM O-ring encloses the reactor vessel to the head. A variable speed drive turns the impeller, though for the ageing experiments the impeller revolutions were kept constant at 580 rpm. The reactor system is controlled with a Model 4848 Modular controller, also from Parr Instruments. This model controller comes with temperature control module and a tachometer control module that can be set in either closed or open loops. During the ageing experiments, temperature was operated in closed loop and stirring speed in open loop. Stirring speed did not alter during the course of the experiment so closed loop control was not necessary.    53  Figure 5: Parr Instruments Model 4520 Bench Top Reactor System (Image available online at parrinst.com)  Residue slurries were made from 32 g of solids in 200 mL of water to produce a 16% w/w solids slurry. The slurry was heated to 150°C in the autoclave reactor system described above. Heating from room temperature to 150°C took 45 minutes before the timer was started. The slurry was mixed at 580 rpm for 2, 4, 6, 8 and 12 hours. At the completion of the experiment, the timer was stopped and the slurry was allowed to cool to room temperature. Cooling took 20 minutes. The slurry was filtered, dried and weighed before solid and solution samples were sent for analysis at CESL. Solid and solutions samples were measured for Cu, Ni, Fe, Ca and S by ICP-MS using the procedure described above. Sequential extraction was conducted on the 12 hour aged sample and the solution was sent for ICP-MS analysis and the solid for QXRD phase analysis.    54 3.3.3 Ageing with ferrous displacement  The potential presence of adsorbed Cu and Ni as sulphates or as divalent Cu2+, Ni2+ on the surface of precipitated iron oxyhydroxides is discussed in Section 2.2.5. An attempt was made to gauge the presence and proportion of these metal losses by ageing the residues with Fe2+ added to the ageing solution. Ferrous may adsorb more strongly to the residue than Cu2+ and Ni2+ has been shown to adsorb more strongly than many other common cations (Rodda, Johnson, and Wells 1993). During the adsorption, the crowding of surface sites may desorb extra Cu2+ and Ni2+ from the residues than just ageing in water. To achieve this, 0.1 and 1 g/L Fe2+ solutions were made up by dissolving solid, hydrated ferrous sulphate Fe(SO4).5H2O in water. The hydrated ferrous sulphate is reagent grade, measured 99.9% purity and supplied by Fisher Scientific Ltd. The reagent was shipped to UBC in 2013.   For each experiment 200 mL of Fe2+ solution was added to the reactor set-up shown in Figure 4 and the solution was heated to 95°C and stirred. Once the temperature was reached, 4 g of S1 or S3 was added to the reactor vessel and a pH measurement was taken. The pH was adjusted to pH 2 using dropwise addition of 1 N H2SO4 or if the pH was too low, dropwise additions of 1 N NaOH were used. Once the pH was stable at pH 2, the timer was started. Ageing in ferrous solution was conducted for 1 and 24 hours and repeated twice for S1 and S3 in each of the two ferrous solutions.   To differentiate the replacement of Cu2+ and Ni2+ with Fe2+ it was necessary to remove other potential sources of Cu and Ni in the residues. A sequential ageing procedure in ferrous solution was devised to achieve this. Residues were aged in water at 95°C for 24 hours prior to repeated   55 ageing in ferrous solution. Samples were also aged in water at 95°C for 24 hours and then sequentially extracted with hydroxylamine hydrochloride prior to ageing in ferrous solution. These were compared with ‘blank’ results from ageing in water, sequential extraction and the ageing of sequentially extracted residues S1 R1 and S3 R1 in ferrous solution. Figure 6 shows the experimental design for the course of experiments.    Figure 6: Experimental flowchart for ageing in Fe2+  Solids and solutions from each experiment were collected, filtered, and dried before being sent to CESL for analysis of Cu, Ni, Fe, S and Ca by digestion and ICP-MS.  From the results a mass balance was calculated to determine the relative amounts of Cu and Ni affected by each of the three separate processes (H2O ageing, stage 1 extraction with HaHC and ageing with ferrous).   S1 and S3 samples Ageing with Fe2+ Treated with HaHC Ageing with Fe2+ Ageing in H2O Treated with HaHC Ageing with Fe2+   56 3.3.4 Repeated ageing and extractions Repeating ageing of the S1 residue was conducted to establish the minimum amount of Cu that remains entrained in the residue and the limits of its metastability.  Four samples of S1 residue were aged for one hour at 95°C in deionised water as per the standard ageing procedure described previously. Initial solids concentrations were varied with samples aged at a solids concentration of 2, 6, 10 and 15%. Each sample was filtered, dried and weighed before the ageing procedure was repeated 4 more times.  Following ageing the samples were twice treated by the first stage of the sequential extraction process with acidified hydroxylamine hydrochloride. Solid and solution sub-samples were collected from each stage and sent to CESL for chemical analysis of Cu, Ni, Fe, S and Ca.   3.3.5 Enhanced atmospheric leach The CESL laboratory in Richmond, BC utilises an alternative process to account for removable Cu and Ni lost to the residue, termed an ‘Enhanced Atmospheric Leach’. In this process, residues are stirred at 70°C for two hours residence time in a 5 g/L KCl synthetic solution. Solid KCl was 99.9% pure and obtained from VWR Scientific Ltd. The pH is kept at 1.5 by addition of drops of 1 N H2SO4. Solids content is not critical but is generally 15-20% w/w and was 15% w/w for our experiments. An enhanced atmospheric leach was conducted in the jacketed reactor set-up described in chapter 3.3.1 on 30 g samples of CESL (S1) and CESL 2 (S4) to compare with results gained from water ageing and Fe displacement.     57 3.3.6 Solid transformation study  An attempt was made to transform the amorphous Fe material to more crystalline forms by a pre-treatment heating process prior to ageing the residues and recovering any copper. Samples of aged and fresh S1 residues were mixed with silica sand found in the laboratory and placed in a ceramic crucible. A sample of original S1 residue was mixed to a proportion of 1:1 with silica, a sample of aged and HAHC treated S1 was mixed to a ratio 1:3 and a sample previously treated by EAL was mixed 1:1 also.  Unmixed samples with no silica sand added were also placed in a crucible for comparison. The samples were then dry heated in an immersion furnace at 150°C for 4 hours. A non-mercury thermometer was suspended inside the furnace with the bulb situated just above the surface of the solids inside the crucible. The temperature setting of the furnace was adjusted to read 150 ± 5°C on the thermometer as a significant temperature gradient exists within the furnace. The correct temperature was reached after 90 minutes at which time the timer was started. The solids mixture was not agitated or stirred to maintain contact between the residues and the dehydrating silica solids. At the completion of the 4 hour heating time the samples were immediately removed from the furnace and allowed to cool to room temperature.   After cooling, 4 g sub-samples of the heated residues were aged in 200 mL of deionised water at 95°C for an hour in the glass, jacketed reactor vessel shown in Figure 4 using the same method as described in 3.3.1. Solid and solution sub-samples were sent to CESL for chemical analysis.     58 Chapter 4: Results and Discussion 4.1 QXRD and ICP analysis of aged residues Ageing improves the residues by partially removing the metastable ferrihydrite and amorphous Fe phases and extracting lost Cu and Ni to a recoverable solution. These Fe phases are disproportionately responsible for Cu and Ni losses to the residue (Sahu and Asselin 2011). Additional sources of Cu and Ni extraction is likely realised from re-dissolving occluded Cu and Ni sulphates that are tenaciously bound within the residue (Chen et al. 2006; Dutrizac and Chen 2012b). The ageing process utilised in the present study can therefore be considered to improve the residue by two mechanisms. The process increases the proportion of crystalline phases by dissolving and partially altering the amorphous phase resulting in an increase in the proportion of crystalline phases. In addition, agitation of the residue in low solid content slurry acts as a washing process to dissolve and remove the occluded Cu and Ni sulphates and Cu and Ni contained in amorphous Fe phases. Conditions of extended time and elevated temperature allow for partial transformation of ferrihydrite to more ordered species such as maghemite (γ-Fe2O3) (Jambor and Dutrizac 1998) and hematite by hydrolysis (Schwertmann, Friedl, and Stanjek 1999). This phenomenon is considered minor compared to the washing and dissolution processes imparted by ageing.  4.1.1 Residue phase analysis Analysis by QXRD reveals that ageing alters the residues by lessening the X-ray amorphous and ferrihydrite phases and dissolves all of the crystalline Ca, Na and Mg phases. Loss of the calcium, sodium and magnesium phases is responsible for a large portion of the mass loss from the residues (Sahu and Asselin 2011). Considering all Ca2+ in solution is from gypsum   59 dissolution 42% of the S1 mass loss and 20% of the S3 mass loss is from this species. Ca extractions greater than 95% are achieved from both residues. Reductions in the amorphous content generate residues with improved stability as a result of the higher proportion of crystalline phases remaining. Aged residue data presented in Table 8 is adjusted to reflect the mass of the original sample to elucidate any potential phase changes more accurately. Numbers are given as weight percent (wt.%).   Table 8: XRD phase analysis S1, S1 R1, S3 and S3 R1 before and after ageing for 24 hours at 95°C (wt.%) Compound Ideal Formula Before ageing After ageing Before ageing After ageing S1 S1 R1 S1 S1 R1 S3 S3 R1 S3 S3 R1 Albite low NaAlSi3O8 0.9        Beaverite Pb(Fe3+,Cu)3(SO4)2(OH)6   1.0 0.4     Chalcopyrite CuFeS2     0.2    Goethite α-Fe3+O(OH)     8.0 10.5 6.6 6.9 Gypsum CaSO4.2H2O 3.2    1.0 0.9   Hematite α-Fe2O3     35.0 35.2 36.4 35.6 Jarosite MFe33+(SO4)2(OH)6 13.5 13.4 11.1 10.9 1.0    Kaolinite KAl2(AlSi3O10)(F,OH)2   0.9 1.4     Lizardite Mg3Si2O5(OH)4 1.3 1.2       Maghemite γ-Fe2O3   0.8 0.8     Magnetite Fe3O4         Molybdenite MoS2    0.1     Muscovite KAl2(Si,Al)O10(OH,F)2 3.9 2.2 1.7 2.4     Pyrite FeS2 1.5 1.7 0.8 0.8     Quartz low SiO2 4.2 6.5 4.4 6.2   0.2 0.6 Sulphur S8 16.3 28.3 24.9 38.6 24.2 24.6 31.1 32.2 Talc Mg3Si4O10(OH)2 1.8        Tennantite (Cu,Fe)12As4S13 0.5 0.9 0.3 0.3     Amorphous  52.7 45.8 47.6 32.5 30.5 28.8 20.5 20.4 Mass loss    6.5 5.6   5.2 4.3 Total  100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0  A subset of the amorphous phases in the residues before and after ageing is ferrihydrite that is susceptible to dissolution with acidified hydroxylamine hydrochloride. The remainder of the amorphous phase largely consists of further poorly crystalline iron phases and altered sulphide minerals that are not susceptible to acidified hydroxylamine hydrochloride dissolution and are unable to be detected by QXRD (Javed et al. 2015).    60 Ageing is able to reduce the amorphous content to the same level for the S3 residues (20.5% versus 20.4%) but is not as effective at reducing the S1 amorphous content to the same level as the S1 R1 residue (47.6% versus 32.5%). Aged S3 residue is much improved over S3 R1 as a higher crystalline content and lower amorphous content is noticed. Reductions in the amorphous phase composition of previously extracted residues (S1 R1 and S3 R1) due to ageing indicate that maximum dissolution of metastable amorphous phases; particularly ferrihydrite requires additional time and elevated temperature. Long ageing times, even at elevated temperatures, are needed for ferrihydrite to overcome the initial induction phase during transformation (Cornell, Giovanoli, and Schneider 1989). Dissolution of ferrihydrite is found to be highly sensitive to equilibration time, even during selective extraction with hydroxylamine hydrochloride (Chao 1972). Longer ageing times allow the induction time to pass and ferrihydrite to dissolve by exploiting its inherent metastability without the use of hydroxylamine hydrochloride. The rate of Fe dissolution by hydroxylamine hydrochloride, among many other factors, is affected by mineralogical structure (Chao 1972) so the much higher amorphous content of the S1 residues requires more intensive treatment and more time than S3 to remove and/or stabilise this phase, and ageing is not sufficient to stabilise S1.   4.1.2 Fe phase analysis  Ferrihydrite dissolution is only marginally responsible for improving residue stability by ageing. Solids mass loss due to soluble Fe (considered as Fe2+ or Fe3+) contributes up to 9.7% and 1.9% of the total mass loss for the S1 and S3 residues respectively. If it is considered that all of the solubilised Fe in solution is ferrihydrite (Fe5HO8.4H2O, (Jambor and Dutrizac 1998)), the mass loss attributable to Fe increases to 16.7% for S1 and 3.2% for S3. Mass losses to Fe are therefore   61 relatively minor compared with Ca dissolution. Fe mass distributions of aged and sequentially extracted residues, given in Table 9, show that only small amounts of Fe from ferrihydrite are dissolved during ageing. Comparison of the Fe concentration obtained by ICP with that obtained by phase proportions (QXRD) generates the value of amorphous Fe extracted to the water.   Table 9: Fe mass and phase balance by ICP and QXRD for aged residues, 24 hours, 95°C, 2%wt. solids  S1 S1 R1 S3 S3 R1 Before ageing ICP QXRD ICP QXRD ICP QXRD ICP QXRD Fe as Hematite     63.6% 82.0% 64.8% 78.9% Fe as goethite     13.1% 16.8% 17.4% 21.1% Fe as jarosite 19.4% 92.3% 26.8% 92.2% 0.9% 1.1%   Fe as pyrite 1.6% 7.7% 2.3% 7.8%     Fe as amorphous 79.0%  70.9%  22.5%  17.8%  Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% After ageing         Fe as Hematite     66.1% 85.9% 67.2% 84.9% Fe as goethite     10.8% 14.1% 11.9% 15.1% Fe as jarosite 15.9% 81.3% 22.0% 80.0% 0.0% 0.0%   Fe as pyrite 1.4% 7.0% 2.1% 7.5%     Fe as maghemite 2.3% 11.6% 3.4% 12.5%     Fe as amorphous 79.2%  70.2%  22.8%  18.6%  leached  1.2%  2.3%  0.2%  2.2%  Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Fe in mass loss 9.7%  6.8%  1.9%  10.4%   The Fe amorphous phase appears unaffected by ageing despite the extraction of Fe from the residue. This suggests that solution Fe is sourced from partially soluble jarosite or insufficient ferrihydrite is dissolved to noticeably affect the Fe distribution to the amorphous phase. Jarosite is moderately affected by ageing with approximately 20% of the jarosite in the S1 and all 1% of the jarosite in S3 residues dissolving. Jarosite is stable in aqueous solution though partial dissolution would be possible in unsaturated, neutral ageing conditions at 95°C as temperatures of around 100°C and 0.3 M acid has been shown to dissolve sodium and lead jarosites with   62 sufficient time (Dutrizac and Jambor 2000). These conditions are more acidic (pH <1) than those of ageing though ageing appears to similarly affect the jarosite species. Mössbauer spectrometry of the residues reveals that jarosite is an unlikely source of the Cu losses to the poorly crystalline phase present in the residues (Abdul et al. 2014) and the effect of jarosite will not be considered further.  Insufficient detectable ferrihydrite may be explained as the transformation of ferrihydrite to hematite occurs through a process of solid-state internal dehydration and rearrangement (Cornell, Giovanoli, and Schneider 1989).  The nucleation of α-Fe2O3 requires significant aggregation of ferrihydrite so it is feasible that solubilised Fe is aggregated to form hematite growth units that are detected as amorphous and are not in solution. The large and complex nature of the residue also poses difficulties in establishing changes in ferrihydrite content.    Trace maghemite (γ-Fe2O3) found in the S1 QXRD diffractograms is evidence of ferrihydrite transformation as maghemite is considered a precursor to hematite (Jambor and Dutrizac 1998; Cudennec and Lecerf 2006). Definitively proving the existence of any transformation is difficult in metallurgical leach residues given the impure, imperfect and heterogeneous nature of the precipitates produced from a hydrometallurgical environment (Loan et al. 2005) and beyond the means of the present study.  Differences between Fe analysis by QXRD and ICP for aged samples, shown in Table 10 are similar to those of the original residue samples suggesting that negligible transformation to more ordered Fe oxyhydroxides occurs. This is not unexpected as it has been shown that ferrihydrite   63 dissolution and subsequent transformation can occur over a matter of years at atmospheric temperatures (Schwertmann and Murad 1983). Dissolution is further slowed as the ferrihydrite concentration reduces and at low pH. The Fe distribution does show evidence of transformation as the mass of Fe as goethite is matched by a comparative increase in Fe due to hematite following ageing. Evidence does exist to suggest that goethite is capable of transforming to hematite at conditions above 92°C in water, even after a few hours (Johnston and Lewis 1983), however, the transformation is nanoscale and the data is likely erroneous as the stability of goethite requires hydrothermal treatment at 160-180°C to transform into hematite after 1 week of ageing (Schwertmann, Friedl, and Stanjek 1999).   Table 10: Comparison of total Fe distribution between original and aged samples by ICP, QXRD and stage 1 extraction (Javed et al. 2015) Total Fe % Original X-ray normalised QXRD ICP Differential 1st Stage S1 5.3 23.3 17.3 12.1 S1 R1 4.9 16.7 11.8  S3 29.9 40.0 10.1 1.1 S3 R1 31.2 38.0 6.8  Aged      S1 5.3 26.6 21.3 1.7 S1 R1 4.9 17.5 12.6 19.8 S3 31.2 40.5 9.3 1.0 S3 R1 31.2 39.0 7.8 1.1  4.1.3 Stability of aged residues Aged residues show improved stability when subjected to further sequential extraction. Residues aged for 24 hours at 95°C in water were subjected to sequential extraction to assess the extent and permanency of any improvement in stability. Mass losses due to sequential extraction are significantly reduced as a result of the prior dissolution of the Ca, Mg and Na phases during   64 ageing. The low solution sulphur assays suggest that further entrained sulphates are not washed from the residue by sequential extraction. Table 11 shows the effect of an additional treatment with acidified hydroxylamine hydrochloride on the aged residues presented in Table 8. Chemical solution analysis confirms that Fe is the major metal species of the suite analysed, proving that the mass loss is the result of ferrihydrite that was not dissolved by ageing.   Table 11: XRD phase analysis of 24 hour aged samples following further stage 1 extraction (wt.%) Compound Ideal Formula Aged then stage 1 extraction with HAHC S1 S1 R1 S3 S3 R1 Albite low NaAlSi3O8     Beaverite Pb(Fe3+,Cu)3(SO4)2(OH)6 0.8 0.8   Chalcopyrite CuFeS2   0.3 0.1 Goethite α-Fe3+O(OH)   6.5 7.1 Gypsum CaSO4.2H2O     Hematite α-Fe2O3   37.3 36.1 Jarosite MFe33+(SO4)2(OH)6 15.3 13.1   Kaolinite KAl2(AlSi3O10)(F,OH)2 1.2 1.7   Lizardite Mg3Si2O5(OH)4     Maghemite γ-Fe2O3 0.7 0.6   Magnetite Fe3O4     Molybdenite MoS2 0.2 0.2   Muscovite KAl2(Si,Al)O10(OH,F)2 2.2 3.0   Pyrite FeS2 1.3 1.3   Quartz low SiO2 5.8 7.8   Sulphur S8 31.5 47.8 31.5 33.4 Talc Mg3Si4O10(OH)2     Tennantite (Cu,Fe)12As4S13 0.5 0.7   Amorphous  27.9 18.5 24.1 23.3 Mass loss  12.7 4.5 0.4 0 Total  100.0 100.0 100.0 100.0  S3 residues are effectively stabilised by ageing as negligible further mass losses occurs with sequential extraction. Small amounts of ferrihydrite remain following ageing as approximately 1% of the Fe is extracted to solution by sequential extraction. Significant mass losses from the S1 residues indicate lessening of ferrihydrite and amorphous phases though the residue remains metastable. Prior ageing is likely to have exposed the ferrihydrite phases in S1 R1 to react with   65 the hydroxylamine hydrochloride and accounts for the continued dissolution. Extended ageing and multiple sequential extractions are likely needed to completely stabilise the S1 residue. Ferrihydrite has been shown to take at least 116 hours to completely transform in similar conditions (Johnston and Lewis 1983). Fe species can still be unstable after 400 hours and, depending on pH, 24 hours may be the required the induction time (Cornell, Giovanoli, and Schneider 1989).   Table 12: Fe removed by acidified hydroxylamine hydrochloride of aged samples  S1 S1 R1 S3 S3 R1 %Fe leached 1.7 19.8 1.0 1.1 %S leached 2.1 0.3 0.2 0.2  The appearance of chalcopyrite in the aged and extracted residues is due to the removal of overlaying phases that brings the proportion of chalcopyrite above the minimum level for detection by XRD. Unleached chalcopyrite is notable in Vale pressure leach residues as it is more difficult to leach and remains even after 60 minutes of batch leaching (Chen et al. 2006).    4.1.4 Extraction of Cu and Ni  Ageing extracts Cu from the original, as-received S1 and Cu and Ni from the original, as-received S3 residues. Sequentially extracted S1 and S3 residues (S1 R1 and S3 R1) also release Cu and Ni by ageing, indicating that the sequential extraction is unable to remove all the recoverable metals. Ageing extracts up to 22% of the Cu from S1 and 6% Cu and 22% Ni from S3 after 24 hours of ageing at 95°C. A further 15% and 25% Cu is extracted from the S1 R1 and S3 R1 samples under the same conditions. Ni extractions of 20% are realised from ageing S3 R1.    66 Ageing for one hour shows similar Fe, Cu, Ni and S grades and extractions by mass balance from the average residue grades are given in Table 13.   Table 13: Metal extractions and residue grades from ICP-MS analysis after 24 hours of ageing at 95°C  S1 S1 R1 S3 S3 R1 Fe     Initial Fe grade (wt%) 23.3 16.7 38.5 38.0 Final Fe grade (wt%) 24.7 17.7 40.5 39.6 Fe extracted to solution (%) 0.5 2 0.1 0.2 Cu     Initial Cu grade (wt%) 1.08 0.83 0.58 0.49 Final Cu grade (wt%) 0.88 0.76 0.57 0.38 Cu extracted to solution (%) 22 15 6 25 Ni     Initial Ni grade (wt%)    0.83 0.76 Final Ni grade (wt%)   0.68 0.62 Ni extracted to solution (%)   22 20 S     Initial S grade (wt%)  29.0 42.8 31.5 32.3 Final S grade (wt%) 28.4 45.6 32.0 33.2 S as SO4-2 extracted to solution (%) 6 1 1 1  Increasing the ageing time noticeably improved recovery of Ni from S3 but Cu extractions from both residues remained comparable. After 24 hours more Ni is extracted by ageing by sequential extraction suggesting that Ni is not or only weakly associated with the ferrihydrite phase. Ni may be bound tenaciously to the surface of residue hematite (Rose and Bianchi-Mosquera 1993) and the resulting desorption process requires extended time frames. Alternatively Ni may be occluded as tiny, micron-sized NiSO4.H2O particles that form under autoclave conditions that are dissolved very slowly, even in hot water (Dutrizac and Chen 2012b). Incorporation of solid solution Ni in S3 hematite is also likely as heating of Ni-containing ferrihydrite can incorporate significant Ni in the resulting hematite (Wells, Gilkes, and Fitzpatrick 2001). The small amount of ferrihydrite present in S3 contains some Ni though this is easily removed during ageing and   67 the majority of Ni lost to ferrihydrite is likely to have already transformed to solid solution Ni in hematite.   Table 14: Metal extractions and residue grades from ICP-MS analysis after 1 hour of ageing at 95°C, 2%w/w solids  S1 S3 Cu   Initial Cu grade (wt%) 1.08 0.58 Final Cu grade (wt%) 0.88 0.56 Cu extracted to solution (%) 23.9 10 Ni   Initial Ni grade (wt%)   0.83 Final Ni grade (wt%)  0.80 Ni extracted to solution (%)  9.3 S   Initial S grade (wt%) 29.0 31.5 Final S grade (wt%) 29.9 32.0 S extracted to solution (%) 4.3 2  Ageing extracts significant Cu and Ni from the residues that are unrelated to the amorphous and semi-crystalline Fe phases that are dissolved by sequential extraction. The low distribution of Fe to solution suggests that extracted Cu and Ni likely results from other sources. These sources could be adsorbed Cu2+ and Ni2+ (Jiang, Xu, and Li 2010; Jung, Cho, and Hahn 1998; Rose and Bianchi-Mosquera 1993) and/or occluded sulphate species such as NiSO4.H2O, CuSO4 and basic copper sulphate, CuSO4.2Cu(OH)2. These sulphate species are difficult, if not impossible, to detect by SEM-EDS and EMP though they can be considered to be uniformly distributed throughout hematite residues (Dutrizac and Chen 2012a; Chen et al. 2006). Continued extraction of Cu and Ni from S1 R1 and S3 R1 proves that Cu and Ni are associated both with the dissolved ferrihydrite and the crystalline Fe phases that are not removed by sequential extraction. Removal of a large portion of the amorphous phase is likely to expose tenaciously bound Cu and Ni and   68 entrained sulphates to further extraction by the ageing solution in the sequentially extracted residues.   4.1.4.1 Metal sulphate extraction by ageing The dissolution or desorption of sulphate species is a major component of the mass losses from ageing. Considering all S in solution as SO42− shows that 48.5% of the mass lost in ageing S1 and 31.8% in ageing S3 is from sulphate. The calcium sulphates, gypsum (CaSO4.2H2O) and anhydrite (CaSO4), contribute most to sulphate mass loss (42% in S1, 20% in S3) from aged species. Mass lost to sulphate species from residues aged after pre-treatment by sequential extraction is less than in the original residues. Table 15 shows the distribution of the metal ions when sulphate is considered the counter-ion. The metal sulphate species are assumed to be simple sulphates though numerous other complex species are likely present as well, including FeOHSO4 (Umetsu, Tozawa, and Sasaki 1977), or adsorbed metal sulphate species such as SO−-CuOH+ (Davis and Leckie 1978). Complex metal chlorides may also be occluded or adsorbed to goethite and hematite and other Fe phases (Rose and Bianchi-Mosquera 1993) but as chloride catalytically aids the leach process (Defreyne et al. 2006) metal losses to chloride species are not considered. A balance of the sulphate species reveals that there are additional counter ions that are not Fe, Cu, Ni or Ca as the addition of these species does not account for all the sulphate extracted to solution. The remaining sulphate may be an adsorbed surface species or another metal ion that has not been accounted for. Fe2+, Cu2+, Ni2+ and Ca2+ counter ions account for 85.3% of the sulphate in S1 and 68.8% in S3.     69 Table 15: Proportion of residue mass loss contributed to Fe2+, Cu2+, Ni2+ and Ca2+ as sulphates  As %MSO4 of total mass loss As %MSO4 of total sulphate  S1 S1 R1 S3  S3 R1 S1 S1 R1 S3 S as SO4-2 48.5 15.0 31.8 15.0 100.0 100.0 100.0 FeSO4 8.6 10.2 1.3 4.7 11.2 42.8 2.6 CuSO4 15.8 4.3 3.2 10.1 19.7 17.3 6.0 NiSO4   1.3 12.6   21.5 CaSO4 37.4 4.6 17.4 1.5 54.4 21.5 38.7 Total 61.8 19.0 32.7 28.9 85.3 81.6 68.8  The sorption of sulphates to crystalline iron oxides is widely demonstrated (Dutrizac and Chen 1993; Dutrizac and Chen 2012b; Peak, Elzinga, and Sparks 2001) and coexisting sulphate salts become intertwined with the products of ferric hydrolysis (Tozawa and Sasaki 1986). The dispersed nature and high specific surface area of amorphous and semi-crystalline Fe oxyhydroxides are capable of removing large amounts of sulphate from aqueous solutions (Boukhalfa et al. 2007). Cu2+ and Ni2+ show an affinity to adsorb to the surface of iron oxyhydroxides (Li 2008) though it is likely that these ions attach to the negatively charged surface sites supplied by the bound sulphates (Peacock and Sherman 2004; Rodda, Johnson, and Wells 1993). A full sulphate speciation and examination of the residue surfaces by EXAFS would likely reveal the precise nature of the occlusion and adsorption of Cu, Ni and Fe sulphates. This would allow for a more accurate account of the mass lost to these species however this is beyond the scope of the present study.   A modest reduction in the Cu and Ni grade is noticed after ageing though the residues still contain significant metal values that may be unrecoverable. Some of the unrecovered values are encased in the sulphides demonstrated by SEM-EDS analysis in the following section and there are likely further occluded Cu and Ni values that require extensive and uneconomic processing to recover. The low ageing extractions of Fe and S indicate that ageing has minimal effect on the   70 stable S and Fe phases. Soluble Fe is sourced from amorphous Fe that is assumed to be ferrihydrite (Sahu and Asselin 2011) though the presence of sparingly soluble ferric sulphate salts, such as FeOHSO4, within the residue surfaces may also contribute (Boukhalfa et al. 2007). These species are the precursors of ferric hydrolysis (Umetsu, Tozawa, and Sasaki 1977; Dutrizac 1979) and the incomplete Fe hydrolysis exhibited by ferrihydrite may occlude these species amongst the residue. Sulphur is present as sulphate and is partially the result of dissolving metal salts including gypsum (CaSO4.H2O), anhydrite (CaSO4) and some of the occluded Cu and Ni sulphates and sulphur species. In the S3 residue, these species were noticed by SEM-EDS and EMPA to be bound to crystalline Fe phases and are likely distributed throughout S1 but are not able to be detected. A more detailed metal-sulphate balance is conducted further on in this study.   4.1.4.2 Comparison of ageing to sequential extraction  Ageing is not as effective or as selective at removing ferrihydrite from the residues as sequential extraction with acidified hydroxylamine hydrochloride is. The reduction in ferrihydrite dissolution results in a lower Cu extraction due to ageing (22% versus 50% for S1), as Cu is disproportionally lost to the ferrihydrite and other amorphous phases (Sahu and Asselin 2011). Cu distribution to the solution from sequential extraction is twice that of the resulting ageing solution. Cu association with this ferrihydrite phase clearly requires a strong reductive process, as provided by the hydroxylamine reactant, to dissolve the Fe3+ ion in ferrihydrite and release the Cu. The close ionic radius of Cu2+ and Fe3+ likely means that Cu is structurally incorporated in hematite and its precursor ferrihydrite that requires the structural matrix of both to be broken and dissolved (Martinez and McBride 1998a). The remaining Fe left in both residues following   71 sequential extraction may also contain Cu and is unlikely ferrihydrite. This agrees with Cu2+ and Ni2+ adsorption (Jung, Cho, and Hahn 1998; Li 2008) and sulphate occlusion (Dutrizac and Chen 2012a; Chen et al. 2006) that occurs on crystalline iron oxyhydroxide surfaces and is not affected by hydroxylamine hydrochloride sequential extraction.   The mass distributions of Ca, Fe, Cu and Ni to the treated residues and solutions are given in Table 16. Data is analysed by a mass balance conducted on the digested and ICP analysed S1 and S3 residues from each process. A second stage of extraction was not conducted on the residues so the phases that are not recovered by sequential extraction or ageing are completely dissolved by the analytical process to determine the total Ni, Cu, Fe, Ca left over. Grade and mass loss data for aged and sequentially extracted residues combine data from the present study and (Javed et al. 2015) and are given in Table 17.   Utilising equation 32 (Sahu and Asselin 2011) the percentage of amorphous Fe left in the residue following each process can be compared. R1 is considered as mass loss due to ageing or sequential extraction as appropriate.   %!"!" = !100 ∗ ! − !! − ! /!       [32]        72 Table 16: Mass distributions for Ca, Cu, Fe, Ni of aged and sequentially extracted residues Sample %Ca res. %Ca soln. %Total Ca %Fe res. %Fe soln. %Total Fe %Feam %Cu res. %Cu soln. %Total Cu %Ni res. %Ni soln. %Total Ni S1, 1 hour 0.06 0.88 0.93 21.63 0.12 21.75 2.59 0.88 0.24 1.12    S1, 24 hours 0.35 0.53 0.87 23.52 0.15 23.67 2.76 0.84 0.30 1.14    S1, sequential extraction 0.04 0.89 0.93 12.14 9.92 22.07 24.30 0.65 0.50 1.15    S3, 1 hour 0.01 0.26 0.26 34.40 0.09 34.49 3.46 0.53 0.06 0.59 0.77 0.08 0.84 S3, 24 hours 0.01 0.23 0.25 38.65 0.02 38.67 3.67 0.55 0.06 0.61 0.65 0.19 0.84 S3, sequential extraction 0.01 0.28 0.28 33.76 1.04 34.80 1.36 0.53 0.13 0.66 0.77 0.04 0.81  The difference in the effect of ageing and sequential extraction on the residues is best demonstrated by the %Fe in solution and %Feam. Ageing is able to extract approximately half the Cu in S1 as sequential extraction is, but with a seventy-fold decrease in the extraction of Fe to solution. A similar trend is noticed for S3 where a ten to fifty-fold decrease in Fe extraction removes half the Cu. Ageing appears to be able to remove the surface bound and occluded Cu in both residues rather than the ferrihydrite associated Cu that is targeted by hydroxylamine hydrochloride. This assertion is supported by the low values of %Feam. These suggest that for the similar losses of Ca and mass, less amorphous Fe is dissolved than by sequential extraction. The remaining mass loss that generates a low %Feam is likely the Cu and Ni sulphate species.   Ni lost to the S3 residues provides further evidence of the different nature of both residues and the effect of both treatment processes. Ageing removes more Ni from S3 than sequential   73 extraction, suggesting that Ni losses are less associated with the ferrihydrite phase. Ni, likely as a sulphate, is more effectively washed out from all phases in the residue rather than extracted by selective dissolution of the ferrihydrite phases. Since washing is a function of time and repetition, sequential extraction becomes ineffective for Ni after all the ferrihydrite is dissolved. The longer reaction times and higher temperatures of ageing that dissolve the Ni species influence the lower value of %Feam and the lower Fe extraction suggest that it is not an appropriate measure of amorphous Fe for this residue.  Table 17: Comparison of grade, recovery and mass losses in aged and stage 1 sequential extracted residues  1st stage sequential extraction S1 S3 Fe extracted to solution (%) 53.8 5.1 %Mass loss due to Fe 23.0 29.1 Initial Fe grade (%) 23.3 38.5 Fe grade after process (%) 16.7 38.0 Cu extracted to solution (%) 50.3 19.2 %Mass loss due to Cu 1.1 0.5 Initial Cu grade (%) 1.1 0.58 Cu grade after process (%) 0.88 0.58 Ni extracted to solution (%)  12.3 %Mass loss due to Ni  0.6 Initial Ni grade (%)  0.90 Ni grade after process (%)  0.76 Ageing   Fe extracted to solution (%) 0.8 0.1 %Mass loss due to Fe 3.2 0.5 Initial Fe grade (%) 23.3 38.5 Fe grade after process (%) 24.7 40.5 Cu extracted to solution (%) 22 6 %Mass loss due to Cu 6.3 1.3 Initial Cu grade (%) 1.1 0.58 Cu grade after process (%) 0.88 0.57 Ni extracted to solution (%)  22 %Mass loss due to Ni (%)  4.2 Initial Ni grade (%)  0.90 Ni grade after process (%)  0.68  4.1.4.3 Ni and Cu associations in aged and sequentially extracted residues Examining the ratios of Cu and Ni to Fe and SO4 further evidences the different nature of the metal losses to both residues. For the S3 residue, ~800% more Ni is in solution than Fe   74 compared with ~300% more Cu (a ratio of 1:27 Ni:Fe to 1:9 Cu:Fe).  Extracted Ni appears more likely related to sulphate as a ratio of 1:5 and 1:14 Ni:SO4 are noticed in 1 and 24 hour aged samples. The large increase between 1 and 24 hours suggests that the Ni-sulphates are tenaciously adsorbed to other phases that are not ferrihydrite. Ageing for 1 hour and sequential extraction remove the Ni-sulphates occluded with the ferrihydrite phases as identical ratios of Ni:SO4 (1:14) are noticed in both samples. The large increase in the Cu:SO4 ratio noticed by sequentially extracted S3 against the aged samples proves that the extra Fe dissolved by hydroxylamine hydrochloride releases significantly more Cu than is possible by washing the sulphates out of the residue.   Further evidence of the strong association between Fe and Cu is noted in S1 as the ratios of Cu:SO4 stay relatively similar (~1:5 to 1:10) for a large change in the Cu:Fe ratios imparted by the sequential extraction. This agrees with earlier studies (Chen et al. 2006; Dutrizac and Chen 2012b; Dutrizac and Chen 2012a) that show metal and sulphate sorption to hematite that can be removed from the surface. Cu can also enter the structure of ferrihydrite and remains incorporated during the transition to hematite (Cornell and Giovanoli 1988; Jambor and Dutrizac 1998).    75  Figure 7: Comparison of Fe and SO4 ratios to Cu and Ni for each process  4.2 Residue investigation by SEM-EDS and EMPA Visual investigation of the original S1 and S3 residues by scanning electron microscopy (SEM) and elemental analysis by energy-dispersive X-ray spectroscopy (EDS) reveals the residues are a mix of crystalline gangue material, partially formed precipitated Fe species, partially leached and unleached sulphides and some amorphous phases. In general, amorphous phases were difficult to detect and distinguish as these phases appear as a gelatinous mass interspersed with a matrix of more crystalline particles that are several microns in size. These crystalline particles are often rimmed with amorphous material or are hybrid particles that contain amorphous and crystalline components. In S1, Cu can be found in the agglomerated, matrix of crystalline and amorphous phases and in partially leached sulphides such as enargite and chalcopyrite.  Cu in S3 is found in   76 partially leached chalcopyrite particles and is sometimes noticed with sulphur rims on hematite particles. Ni is not noticed with the amorphous phases or with hematite particles but can be found with partially leached and unleached pentlandite minerals.   Analysing individual hematite particles of the S3 residue by electron microprobe analysis (EMPA) found significant Cu, Ni and SO4 in each of hematite particles analysed. An average composition of the 18 hematite particles analysed returned a mineral containing 66.1 ± 1.4% Fe, 0.73 ± 0.46% Cu, 0.85 ± 0.60% Ni and 3.57 ± 1.07% S (or 1.19 ± 0.36% SO4). Each particle showed no distinct difference or patterning under the SEM electron beam, suggesting that the Cu and Ni inclusion is uniformly distributed throughout the hematite. Evidence of Cu inclusions in hematite is widespread (Martinez and McBride 1998a; Dutrizac and Chen 2012b; Li 2008) though as suggested by these authors EMPA is unable to distinguish between incorporated and surface bound species. Studies of Ni residues (Chen et al. 2006) saw only unleached pentlandite as a source of losses though the similar ionic radii between Cu and Ni mean it is likely to also be incorporated with hematite. It is likely that the Cu and Ni adsorbed and occluded as surface species is recovered by ageing.   4.2.1 S1 investigation by scanning electron microscope Visual inspection of the S1 residue with SEM shows a mix of agglomerations of phases that are likely amorphous, interspersed amongst a matrix of several crystalline particles. The gelatinous mass that constitutes the amorphous and semi-crystalline Fe phases is termed ferrihydrite (Sahu and Asselin 2011) and is revealed to contain Cu by EDS field scan. The ferrihydrite particles are likely nanoscale and can agglomerate to form larger masses that are visible by SEM (Loan et al.   77 2002a), though categorically determining discrete ferrihydrite masses amongst these particles was not possible. The associations of incorporated Cu with other minerals and elements could not be determined by SEM-EDS. Cu does not appear to be evenly dispersed throughout the gelatinous mass as Cu was found by EDS point scan in discrete areas of the matrix but was difficult to locate and random.   The S1 residue shows evidence of incomplete leaching as there are widespread unleached and partially leached Cu-sulphide species. These species appear mostly as Cu deficient crystalline structures with aspects of amorphous morphology that is imparted by the leaching process. Cu-sulphides appear mostly as arsenic-rich tennantite of varying Cu/Fe ratio with significant enargite/luzonite also observed. Tennanite and chalcopyrite appear from the QXRD diffractograms of the original and aged residues though no evidence of enargite is found by QXRD. The main diffraction peaks of enargite are perfectly overlapped by major sulphur peaks so the detection limit of enargite in the S1 sample is optimistically as low as ~1% (M Raudsepp, personal communication 2015). Copper losses to enargite have not been determined but even 1% unleached enargite in the residue may account for ~40% of the Cu losses. These losses are unrecoverable by ageing. Further Cu was found with minerals that appear to be pyrite or arsenopyrite though could be Cu and Fe deficient tennantite or enargite as the abundance of sulphur and arsenic indicates that it is not chalcopyrite. Copper losses to these partially leached sulphide minerals is noteworthy but they contribute only a small mass fraction (~0.5% by QXRD) of the original residue. The remaining portion of the Cu losses must therefore be associated with the agglomerations of amorphous masses, ferrihydrite and in the matrix of other phases surrounding these partially leached sulphide minerals that are not visible by SEM-EDS.   78 Some of the Cu present in the agglomerated and amorphous phases is likely to be recoverable by ageing.     Figure 8: Large, unleached enargite particle (1) with slight evidence of product layer  Figure 9: Cu-rich, As-rich arsenopyrite/pyrite (2) particle with no apparent surface product layer     Figure 10: Cu-poor, partially leached tennantite (3) surrounded by matrix of amorphous and poorly-formed particles and small crystalline minerals Figure 11: Unleached Cu-rich tennantite (5) with small intrusion of Cu-poor tennantite mineral (6) in a matrix of amorphous and crystalline material   79   Figure 12: Partially leached enargite  (4) with amorphous and poorly formed particles nearby   Table 18: Mineral identification and composition in S1 residue determined by EDS analysis    Elemental composition (%) Mineral # Name Formula Cu Fe S As Total 1 Enargite Cu3AsS4 45 0.2 31.8 23 100 2 Pyrite?/Arsenopyrite? FeS2/FeAsS 5.5 41 36.5 11.5 94.5 3 Tennantite (Cu,Fe)12As4S13 12.5 37 28 21 98.5 4 Enargite Cu3AsS4 12 17 35 29 93 5 Tennantite (Cu,Fe)12As4S13 12 34 23 18 87 6 Tennantite (Cu,Fe)12As4S13 7 37 25 16 85  The SEM micrographs shown in Figure 8 to Figure 12 represent a cross section of the Cu losses to the S1 residue. Some losses are the result of unleached and partially leached primary Cu minerals like enargite (mineral #1 and #4) shown in Figure 8 and Figure 12 and Cu containing lesser minerals like tennantite (#3) in Figure 10. The partially leached and unleached minerals are unable to be completely leached in the autoclave as they are larger than than the p80 ~25 µm passing size of the leach feed as a result of poor classification.  Minerals #2, 3, 5 and 6 evidence further Cu-sulphide losses. Mineral 2 contains a small amount of Cu (~5.5%) and large amounts of sulphur (36.5%) suggesting that the mineral is most likely pyrite or pyrrhotite rather than chalcopyrite. Low Cu values on these sulphide minerals may be due to CuSO4 wetting the mineral surface to distort the analysis or these minerals may be Cu deficient sulphide minerals   80 undergoing leaching. Surface Cu or occluded CuSO4 may be recoverable by washing the residue with the ageing process.   Tennantite is a complex mineral that often passes unleached through the autoclave. The matrix of crystalline particles and gelatinous amorphous masses that sometimes surround these particles likely further hinders the diffusion processes necessary for leaching of tennantite and other sulphide minerals. This is demonstrated in Figure 10 where a 10 µm tennantite particle (#3) is surrounded by a mixture of small crystalline particles and masses that appear to be the amorphous and semi-crystalline Fe phases resulting from precipitation. A magnified micrograph of two tennantite particles (#5 and #6) in Figure 11 shows that the surrounding phases are not porous and mass transport of the leached Cu species is likely impeded by these phases. The agglomeration of the precipitated nanoscale ferrihydrite is likely to enhance this effect by occluding significant Cu2+ while in proximity to any leached species.   The mass transfer phenomenon of S1 may differ from the pressure leaching of a largely pentlandite concentrate, similar to the precursor concentrate for S3, discussed by Chen et al. (2006). In that study, porous hematite precipitating on the surface of pentlandite allowed the mass transfer of Ni through the reaction surface layer product. The amorphous and semi-crystalline Fe phases that form during the leaching of the precursor copper sulphides for the S1 residue do not appear to be as porous. The reduced porosity likely limits the mass transfer of Cu away from the sulphide surfaces and prevents some of the Cu from entering the bulk solution. This would likely cause the wetting of the surrounding sulphide particles and amorphous Fe phases with the leached CuSO4 and other soluble Cu species. Sulphates are difficult to determine   81 using SEM-EDS though the nature of the material surrounding the sulphide particles does suggest that entrainment of leached Cu can easily occur. Partially dissolving these amorphous phases and washing the entrained Cu and sulphate species intertwined within the matrices of tiny crystalline phases is likely achievable by sufficient ageing and agitation of the residue.   4.2.2 S3 investigation by scanning electron microscope S3 consists of a mixture of crystalline hematite and goethite with some partially leached and unleached sulphide minerals and quartz occasionally noticed by SEM. Some of the hematite and goethite particles are rimmed with sulphur or contain sulphur intrusions that often bear Cu and Ni. Phases that are QXRD amorphous are difficult to locate and may be attached to other particles or are defect structures of incompletely leached sulphides and poorly precipitated Fe and S species. In general, the minerals appear well liberated though instances of incomplete leaching and matrices of tiny crystalline particles and amorphous masses are present throughout the residue.    Figure 13: Partially reacted chalcopyrite (7) and pyrite (8) Figure 14: Large, porous hematite particle liberated from sulphides   82   Figure 15: Pyrite particle surrounded by amorphous phases and other particles of unknown composition Figure 16: Hematite particle with Cu and S in the centre, ongoing hematite formation is noted by the arrow  Figure 13 to Figure 16 show the difference in the typical mineral associations between the S1 and S3 residues. S3 consists of the crystalline Fe oxides, hematite and goethite, that are well formed and cleanly liberated from the sulphide minerals. In S1, the Fe precipitates were typically agglomerations of amorphous and semi-crystalline Fe phases interspersed with tiny crystalline phases surrounding the sulphide particles. A large, porous hematite mineral such as #9 are common throughout the residue and are unlikely to further impede sulphide leaching. The matrix of minerals surrounding partially leached pyrite (#10) are less common than in S1 and often do not contain Cu and Ni.  Unleached chalcopyrite (#7) is the source of some of the Cu losses and often remains in the residue even after 80 minutes of batch leaching (Chen et al. 2006). Chalcopyrite in small concentrations (0.2%) is noticed in the QXRD diffractograms of the original residue. Assuming S3 typically contains 0.2% chalcopyrite, this would account for ~24% of the residue Cu losses. Some of the remaining Cu and possibly Ni losses are likely the result of the process exhibited in Figure 16. This image shows hematite (#11) forming at the edges of a sulphide particle that is   83 likely chalcopyrite. The porosity exhibited by hematite and the continued formation of the mineral (shown by the arrow) indicates that leaching is continuing and that the chalcopyrite is not passivated by the surrounding hematite. Since the mineral is in the process of being leached it is not likely to be revealed by QXRD, and therefore the analysis likely overestimates losses to other phases.   Significant Cu (and likely Ni) is lost to the crystalline Fe phases and is likely typical of the Cu and sulphur associated hematite demonstrated in Figure 18. This image shows a hematite particle (dark patches) surrounded by a sulphurous mass that contains 1.6% Cu. Targeted EDS analysis showed that the hematite itself was devoid of Cu but an increasing concentration gradient of Cu was noticed at the edges of the particle. Concentration of Cu on the periphery of the hematite suggests that Cu and S may be surface adsorbed or occluded as large mass of Cu sulphate to hematite rather than hematite forming from partially leached chalcopyrite. In this instance the hematite would be noted at the particle boundaries as shown by Figure 16. Adsorption and Cu/Ni sulphate occlusion is practically impossible to differentiate and categorically determine from SEM-EDS analysis. The heterogeneous particle does indicate the strong and potentially widespread association of the crystalline iron phases with Cu losses. Large masses of sulphur inextricable to hematite are relatively ubiquitous throughout the residue though not all contain Cu. A uniform, uncoated hematite particle is shown on the left in Figure 17 for comparison to the heterogeneous Cu containing hematite particle shown on the right.    84   Figure 17: Close up of uncoated, homogenous hematite particle Figure 18: Hematite particle coated by a sulphurous mass containing some Cu  Ni losses are observed from unleached and partially leached primary sulphides, which are mainly pentlandite. Visual inspection and elemental analysis with SEM-EDS did not notice any Ni losses to the amorphous Fe phases, the tiny crystalline particle matrix or associated with the crystalline Fe phases. In general it was more difficult to find lost Ni than Cu despite the higher concentration of Ni contained in the residue. This may be down to pure chance or suggests that the Ni is distributed more widely and in lower concentrations that may not be visible by SEM-EDS. A collection of partially leached and unleached pentlandite particles present in S3 are shown in Figure 19 and Figure 20.   Figure 19: Large, oversized, unleached pentlandite particle Figure 20: Large, partially leached pentlandite particle    85 The sources of Ni lost to S3 found by SEM are typical of minerals 14 and 15. These particles are larger than the desired p80 20 µm leach feed particle size and as a result are not completely dissolved by pressure leaching. Mineral 14 is pentlandite that appears to have passed through the autoclave unaltered while mineral 15 shows evidence of partial dissolution of the Ni from pentlandite and the diffusion of the hematite product away from the mineral surface. It is unclear whether mineral 15 would appear in QXRD analysis due to its altered nature and pentlandite similar to mineral 14 is likely of insufficient quantity to reach the detection limit of the QXRD beam analysis software. Assuming that unleached pentlandite minerals reach the minimum QXRD detection limit of 0.1%, Ni losses to this phase account for only 4% of the total Ni losses to the residue. Further losses of Cu and Ni may be found with the hematite particles and likely accounts for a significant portion of the total residue losses that are unlikely to be detected by SEM-EDS.   Ni is associated with the hematite and goethite phases as a result of pentlandite leaching and hematite precipitation process described by Chen et al. (2006). SEM analysis of the pentlandite leach residue showed that a large portion of hematite ‘shells’ precipitated on the surface of pentlandite during the dissolution. The porous shells of hematite allowed leaching to continue as the hematite precipitate proceeded to grow and become agglomerated by sulphur. The study showed that these shells are angular-shaped and represent the original surfaces of the pentlandite particles. The close association of the pentlandite minerals and the product hematite particles reveal the source of Ni losses may be the result of the incomplete diffusion of Ni away from the hematite product layer. This would likely leave tiny particulates of NiSO4.2H2O on the surfaces of the hematite product that are largely responsible for Ni losses in ferric sulphate solutions   86 (Dutrizac and Chen 2012b). These particulates are likely to be partially extracted from the S3 residue by ageing.   The wide difference in the Cu extractions between the two residues is related to the residual Cu concentration in each residue (1.1% Cu in S1 and 0.6% Cu in S3). S1 has more ferrihydrite and less crystalline Fe as a result of incomplete Fe hydrolysis. This may stem from the difficulty in leaching chalcopyrite over pentlandite as suggested by Chen et al. (2006). Pentlandite leaching occurs more readily than chalcopyrite in the Vale conditions studied. The porous and well-formed hematite would likely sorb Ni and some sulphate, possibly as NiSO4, to the surface rather than become incorporated with ferrihydrite. S1 is produced from a Cu concentrate with significant enargite (20%) and chalcopyrite (38%) compared to the trace Cu (~6% chalcopyrite) found in the largely pentlandite concentrate that produces S3. The close ionic radius of Cu2+ and Fe3+ means that Cu is adsorbed and/or segregated on/within ferrihydrite (Martinez and McBride 1998b). Cu in S3 is likely to be associated with crystalline Fe phases as a surface bound species (Dutrizac and Chen 2012a) or interstitially incorporated in hematite (Martinez and McBride 1998a). Less Cu may occur as a ‘trapped’ Cu2+ ion within the ferrihydrite (Cornell and Giovanoli 1988) due to the lower ferrihydrite concentration in S3 than S1.   4.2.3 S3 investigation by electron microprobe  Significant residue losses of Cu and Ni to the crystalline hematite phase were noticed by use of an electron microprobe that were not evident by SEM-EDS. All 18 hematite minerals analysed contained significant concentrations of Ni, Cu and S. Of the 2 other minerals analysed, 1 resembled sulphur-rich goethite and 1 is a Cu-Fe-S mineral that may be partially leached   87 chalcopyrite. No pure hematite particles were found, with 67.6% Fe representing the richest Fe content of the particles investigated. Metal inclusions in the hematite minerals ranged from 0.7% to 2.7% Ni, 0.2 % to 1.9% Cu and 2.6% to 6.3% S (0.9% to 2.1% SO4). The remainder of the mineral was considered as oxygen. An average of the 18 hematite particles yields a mineral of 66.1 ± 1.4% Fe, 0.73 ± 0.46% Cu, 0.85 ± 0.60% Ni and 3.57 ± 1.07% S (or 1.19 ± 0.36% SO4). Given that the residue constitutes 35% hematite, an average particle such as this represents 30% of Ni losses and 25% of the Cu losses, or a grade of 0.25% Ni and 0.15% Cu of the total residue. These analyses agree with the published literature as Ni and Cu is regularly noted to occur with crystalline iron oxides precipitated from sulphate leach solutions (Dutrizac and Chen 1993; Rodda, Johnson, and Wells 1993; Li 2008). The values of Ni and Cu inclusion are higher than expressed by Chen et al. (2006) who found 0.33wt% Ni and 0.11wt% Cu in a similar residue though this was produced from a continuous mini-pilot scale operation and was sulphur free rather than produced from by a large-scale pilot operation like S3.   Table 19: Composition of four hematite particles and average composition of 18 hematite particles analysed by EMPA Mineral description %Fe %Cu %Ni %SO4 Average composition 66.1 0.73 0.85 1.19 Cu-rich hematite 63.4 1.87 1.43 1.74 Low Cu hematite 67.4 0.21 0.08 1.11 Ni-rich hematite 62.1 1.21 2.69 2.11 Low Ni hematite Highest Fe hematite 64.8 67.6 1.54 0.13 <0.01 0.34 1.80 0.90  The reduction in the Fe content in hematite from the pure mineral concentration is due to the additional Cu, Ni and S that is inextricable from the Fe analysis. An oxygen analysis is not able to determine the nature of the association as O is the balance element of the Cu, Ni, S and Fe analysed by EMPA. Close agreement is found between the difference in the Fe mole balance of   88 pure hematite with the analysed moles of Fe in hematite and the moles of Cu, Ni found in the hematite. A positive value indicates that excess Fe sites are available in the matrix for incorporation and a negative value indicates that more Cu and Ni is sorbed than there are Fe sites available and indicates surface binding. A balance of sulphate with Cu and Ni shows that CuSO4 and NiSO4 are present in all hematite particles. Small quantities of Cu and Ni that are not countered by sulphate are likely surface adsorbed as divalent ions or may incorporated into the hematite crystals. It is not possible to determine the quantities or proportions of these occurrences utilising EMPA only, though calculating the elemental composition of the hematite particles in this way does indicate the nature of the inclusion. The occurrence of both methods of Cu and Ni incorporation, surface binding/occlusion and incorporation agrees with published literature (Dutrizac and Chen 1993; Dutrizac and Chen 2012b; Li 2008; Martinez and McBride 1998b) though these authors were also unable to categorically determine the proportions of and exact nature of the inclusions.  4.3 Cu and Ni extraction with high temperature ageing   Ageing the residues at 150°C improves the extraction of Cu from both residues and improves Ni extraction from the S3 residue. Extraction increased marginally from 2 to 6 hours in both residues with the large majority occurring after 4 hours of ageing. Figure 21 shows the extraction of Cu, Ni and Fe from the residues for each ageing time.      89  Figure 21: Cu, Ni and Fe extraction from S1 and S3 residues at 150°C, 16% w/w solids, H2O  Extraction of Cu from the S1 samples showed marked improvement at higher temperatures. After 6 hours 40% of the Cu was removed compared with 22% Cu extraction from samples aged at 95°C for 24 hours. S1 residues aged for 12 hours at 150°C lose a further 7% mass and ~40% Fe on subsequent sequential extraction suggesting that higher ageing temperatures are not sufficient to stabilise the S1 residue. An additional 39% Cu extraction is realised with the sequential extraction after ageing to give a total recovery for both stages of 58%. Release of additional Cu would be expected if transformation to hematite were occurring (Marsden et al. 2002) though hematite formation is not noticed in the QXRD diffractograms. Existing pressure leaching operations use recycled residue to seed the transformation of ferrihydrite to hematite at 150°C and increase the Cu extraction (Marsden et al. 2002). The lack of the required crystalline iron oxide precursors in the S1 residues likely limits the formation of hematite end members. As   90 a result the increased extraction of Cu may be due to the better washing imparted by the dual autoclave impellers and increased sulphate solubility at higher ageing temperatures. The large increase in Cu extraction at higher temperatures is further evidence that a high proportion of Cu is contained in the amorphous and semi-crystalline Fe phases in the S1 residues. This is different to the mostly occluded and interstitially incorporated Cu that characterises the S3 residues.   Cu extraction in the S3 samples of ~5% is comparable to the extractions experienced at 95°C suggesting that recovery of the easily removable Cu is close to maximum. Treatment of the 12 hour aged S3 residue by sequential extraction achieved 8% further Cu recovery suggesting that the Cu locked in the crystalline iron oxide phases and unleached chalcopyrite accounts for the remaining ~87% of Cu losses and is likely unrecoverable by ageing. Stability of the S3 residues is enhanced after ageing at 150°C for all ageing times. Sequential extraction of the S3 residues accounted for a further ~1% Fe dissolution and no further mass loss indicating that the iron phases are effectively stabilised after 12 hours.   Higher nickel extractions at 150°C again reveal the different nature and relative proportions of Ni inclusion in the residue over Cu. Extractions of 46% are achieved after 4 and 6 hours of ageing at 150°C. This compares favourably with the ~22% Ni extraction from samples aged for 24 hours at 95°C. The stability of the Fe phases under these conditions evidences that the extra Ni extracted may be due to the improved washing from improved stirring in the autoclave and increased sulphate solubility imparted by higher temperatures. The levelling out of the recovery after 6 hours suggests that a maximum amount of Ni has been extracted at these conditions and that the remaining ~55% is locked within the crystalline Fe phases and as unleached pentlandite   91 and is likely unrecoverable. These results are similar to those found in a similar Voisey’s Bay leach residue investigated by Chen et al. (2006). EMPA of that residue found 0.33% Ni associated with hematite particles in a similar pilot-scale residue, this accounts for 20% of the Ni losses. EMPA of our residue suggests that ~30% of the Ni is locked in hematite and the remaining 25% may be as unleached pentlandite if all Ni in hematite is interstitially incorporated.   4.3.1 Residue phase analysis of high temperature versus low temperature ageing  Phase analysis by QXRD was conducted on the 12 hour aged then sequentially extracted, low (95°C) and high (150°C) temperature residues to assess the phase changes that are responsible for the noted increase in metal extractions. The increased temperature of the ageing process greatly enhances the proportion of the crystalline phases in the S3 residue. Additional crystalline phases were not noticed in the S1 residue as the residue contained a higher amorphous content from ageing at 150°C than 95°C. This is likely the result of the partial breakdown of solid sulphur as the temperature approaches or overshoots 155°C where sulphur becomes viscous. Temporarily overshooting the 150°C target temperature and possibly causing the breakdown of solid sulphur in this way is possibly due to experimental procedural error. The residues presented in Table 20 were aged at 95°C and 150°C for 12 hours. Low temperature (95°C) ageing was conducted at 2% w/w solids and the high temperature ageing at 16% w/w solids to match the content of the established lab scale (ageing) and plant scale (pressure leaching) processes. Following ageing the residue samples were treated by sequential extraction to ensure the Fe phases formed were stable.    92 Table 20: Phase analysis comparison between residues aged at 95°C and 150°C before stage 1 extraction (wt.%) Compound Ideal Formula 12 h, 95°C aged, sequentially extracted samples 12h, 150°C aged, sequentially extracted samples S1 S3 S1 S3 Albite low NaAlSi3O8     Beaverite Pb(Fe3+,Cu)3(SO4)2(OH)6 0.6  0.6  Chalcopyrite CuFeS2  0.2  0.3 Goethite α-Fe3+O(OH)  7.7   Gypsum CaSO4.2H2O     Hematite α-Fe2O3  37.6  53.4 Jarosite MFe33+(SO4)2(OH)6 9.4  8.7  Kaolinite KAl2(AlSi3O10)(F,OH)2 1.5  1.2  Lizardite Mg3Si2O5(OH)4     Maghemite γ-Fe2O3 0.7  0.9  Magnetite Fe3O4     Molybdenite MoS2 0.2  0.1  Muscovite KAl2(Si,Al)O10(OH,F)2 2.5  1.6  Pyrite FeS2 1.4  1.1  Quartz low SiO2 6.6  5.3  Sulphur S8 34.3 33.3 26.7 30.5 Talc Mg3Si4O10(OH)2     Tennantite (Cu,Fe)12As4S13 0.4  0.3  Amorphous  34.0 21.3 46.7 15.9 Mass loss  8.6  6.8  Total  100.0 100.0 100.0 100.0  The phase analysis comparison between the S3 residues reveals that the higher temperatures are capable of converting goethite and some of the Fe amorphous phases to hematite. The increase in hematite content from the 95°C to the 150°C residue samples corresponds with the disappearance of the goethite phase and reduction in the amorphous phase. The transformation of ferrihydrite to hematite has been shown to occur after 4 days at temperatures above 92°C in water at neutral pH (Johnston and Lewis 1983). Selection of this process temperature was to minimise the formation of goethite though goethite appears to be an intermediary during the transformation. Non-neutral pH values favour goethite formation over hematite from ferrihydrite so the natural S3 residue pH of ~3 indicates that transformation of goethite should begin   93 instantaneously. The high process temperature and low acidity conditions (Umetsu, Tozawa, and Sasaki 1977) overcomes the unfavourable pH as hematite is favoured by higher temperatures that aid both the dissolution and solid-state transformation by dehydration and internal rearrangement of ferrihydrite to form hematite (Manceau and Drits 1993; Jambor and Dutrizac 1998).   The formation of hematite from goethite is perplexing as 160-180°C hydrothermal treatment for a week was shown to be required for the transformation to occur (Schwertmann, Friedl, and Stanjek 1999). The similarity in the stability of hematite and goethite makes the direct transformation via dissolution and recrystallization not likely (Blesa and Matijevic 1989). The absence of hematite seeding and insufficient precipitation time forms akaganéite (β-FeO.OH) from ferric sulphate leach solutions and is metastable in relation to hematite (Dutrizac and Riveros 1999). The availability of seed hematite and the extended 12 hour reaction time at 150°C in high temperature ageing would cause metastable akaganéite to transform to hematite whereas goethite would require a much longer timeframe. The difference between well-formed crystalline goethite and akaganéite is easily determined by QXRD. Distinguishing between these two crystalline Fe phases becomes more difficult if the crystallites are nanoscale and there is sufficient background scatter in the QXRD diffractograms (Raudsepp 2015).  The combined QXRD phase analysis and ICP chemical analysis of the residues show that recycling the residue through the autoclave is likely to improve recovery of Cu and Ni and lead to more crystalline Fe in the residue. Recycling of copper pressure leach residues is industrially practiced and increases recovery by providing surfaces sites for hematite precipitation and subsequently releases Cu from the poorly formed amorphous phases (Marsden et al. 2002).    94 Treating the fresh S1 and S3 residues at elevated temperatures with a new process stage, as these set of high temperature ageing experiments replicate, is not practical. The physical limits of the installed equipment and process flowsheet mean further investigation is confined recycling residue to as a seed and to further wash Cu and Ni from the residue. An investigation into seeding with leached residue has been conducted by another member of the project team.   4.4 Fe2+ displacement tests Numerous studies have shown the ability of ferrihdyrite, hematite and goethite to sorb Cu2+ to their surfaces as a dentate and bidentate complex (Meng et al. 2014; Moon and Peacock 2013; Peacock and Sherman 2004; Rodda, Johnson, and Wells 1993). Ni2+ has also been shown to adsorb to a lesser extent than Cu2+ to goethite and hematite (Rose and Bianchi-Mosquera 1993). Cu2+ adsorption to hematite and goethite has been shown to occur independent of ionic-strength (Jung, Cho, and Hahn 1998) and occur more strongly in ferrihydrite (Li 2008). Cu2+ is more strongly adsorbed to goethite than other common cations such as Zn2+ and Pb2+ at similar pH (Rodda, Johnson, and Wells 1993). Low pH (pH < 3) allows H3O+ to compete with Cu2+ to reduce the adsorption (Meng et al. 2014). Presence of Fe2+ has been shown to reduce Cu2+ adsorption (Rose and Bianchi-Mosquera 1993) and may be able to replace surface bound Cu2+ by a desorption and replacement process due to their similar ionic radius and valence.   The addition of free Fe2+ ions to the ageing solution reveals the strongly adsorptive capacity of the residues. Fe ions are adsorbed from solution to increase the Fe content of the residues and release additional Cu2+. The relation between the adsorption of Fe and the synergistic extraction or desorption of Cu is unable to be definitively proven due to the complex nature of the system.   95 Both processes are likely to occur separately as the adsorption of Fe is much higher than desorption of Cu.   4.4.1 Cu2+ and Ni2+ desorption with Fe2+ from residues An enhancement in the extraction of Cu and Ni from the residues due to displacement of the Cu2+ off both residues and Ni2+ off S3 was noticed by the addition of Fe2+ to the ageing solution. This may be incidental as it has been reported that Cu2+ does not adsorb to hematite at pH values less than 3 (Rose and Bianchi-Mosquera 1993; Dutrizac and Chen 2012b). A mass balance from the ICP chemical analysis of the solids and resultant ageing solution with additional Fe2+ is presented in Table 21.   Table 21: Extraction of Cu and Ni from S1 and S3 residues by ageing in H2O and with added Fe2+  Results of the Fe displacement experiments presented by Table 21 suggest that the addition of Fe2+ to the ageing solution has a synergistic effect that increases Cu and Ni extractions. In all solutions there was a noted adsorption of Fe2+ to the residue, particularly for 1 g/L Fe2+ solutions where adsorption of Fe2+ was greater than the Fe released by ageing. The 5% and 7% improvement in S1 Cu extractions after 1 hour suggests that the adsorption of Fe2+ enables additional Cu2+ extraction by replacing Cu2+ on surface binding sites. The process may be slow Treatment Method H2O, pH 2 1 hour 0.1 g/L Fe  1 hour 1 g/L Fe 1 hour H2O, pH 2 24 hours 0.1 g/L Fe  24 hours 1 g/L Fe 24 hours  S1 %Cu extraction 18.0 23.0 25.2 22.2 40.1 38.5 %Fe adsorbed  0.1 1.4  0.1 5.4  S3 %Cu extraction 8.7 8.5 5.5 5.9 51.3 19.2 %Ni extraction 7.6 6.9 8.4 22.0 25.0 29.8 %Fe adsorbed  0.0 0.3  6.0 0.7   96 as nearly twice the Cu is extracted after 24 hours. Comparison between the molar concentrations of Fe2+ adsorbed and Cu2+ released is inconclusive however and adsorption analysis of such a complex residue is difficult and may be impossible.   These results agree somewhat with published literature. Meng et al. (2014) showed that small quantities of Cu could be desorbed from artificial ferrihydrites though the adsorption ‘is essentially irreversible within the time frame of these experiments’. Desorption lines did not coincide with the adsorption isotherms and the ‘low initial slopes’ presented by Cu desorption isotherms are evidence that desorption does occur but to a ‘low degree’. Lowering the pH facilitated the partial desorption as the availability of H3O+ competed with Cu2+ for surface sites. Ageing with Fe2+ for 24 hours shows similar, if slightly enhanced extractions so both ions may be competing with Cu2+ for surface sites. The replacement with H3O+ mechanism was unable to be determined for these experiments as ageing extracts other sources of hydrolysis that contributed to a decrease in pH of ~0.3 to pH 1.7 in both residues. It follows that H3O+ adsorption would rather increase the pH of the solution so the interfering acid producing species cloud any analysis of H+ mass transfer.  The increase in contact time between the residues and the Fe2+ solution increased Cu extraction in S1. Improved desorption with increasing time may not necessarily be responsible for the improvement in Cu and Ni extractions. It has been shown that increasing ageing times increased the release of adsorbed Cu2+ at constant pH for one of the ferrihydrites investigated (Meng et al. 2014). The two remaining ferrihydrites saw a reduction in Cu2+ in solution at the commencement of ageing and remained stable, indicating further adsorption to an unsaturated ferrihydrite   97 surface. No evidence of transformation by QXRD was noticed in any of the ferrihydrites studied. Improved extraction may be related to improved washing and dissolution conditions for the occluded Cu and Ni sulphates, the apparent slow desorption kinetics, undetectable transformation or a combination of all three. The use of QXRD by the author to determine the transformation may be inappropriate. Ageing ferrihydrite for 6 weeks at room temperature is likely to result in the appearance of goethite or hematite at a nanoscale level. These would only be detectable by Mössbauer spectroscopy and neutron scattering and expressing change in terms of ferrihydrite disappearance by chemical analysis may be more appropriate (Cornell, Giovanoli, and Schneider 1989).   Fe2+ addition has a different effect on the S3 residue providing further evidence that the nature of Cu inclusion is different to that of S1. Cu adsorption to hematite is tenacious and often occurs with related sulphate (Dutrizac and Chen 1993). Cu extraction from S3 shows negligible variation from that achieved without Fe2+ addition after 1 hour. The much lower proportion of ferrihydrite provides a lower specific surface area for adsorption of Cu2+ so it follows that less Cu is surface bound within this phase and can not differentiated from the occluded Cu removed as sulphates by the ageing process. Extraction is unlikely to be as high as 51%, this result is most likely erroneous, but the near 100% increase in extraction with 1 g/L Fe after 24 hours over ageing with H2O evidences that additional Cu can be crowded out of the residues with Fe2+.   Analysis reveals an increase in the amount of sulphur as sulphate leached to solution. Additional Cu removed by Fe2+ is more likely associated with the tenaciously bound surface CuSO4 and SO42− on hematite and relaxed complex Cu sulphates salts than adsorbed as Cu2+ to available   98 anionic sites (Dutrizac and Chen 2012b). Sulphate adsorption to hematite has been shown to be partially reversible by the addition of strong, concentrated counter anions such as F− and OH− (Dutrizac and Chen 1993) but no evidence suggests that Fe2+ is capable of removing adsorbed SO42−.   The extraction and desorption of Ni from S3 is slow and is only slightly enhanced by the addition of Fe2+. After 1 hour, the extraction of Ni2+ with 0.1 and 1 g/L Fe2+ shows only slight variation from the Ni extraction with H2O only. Ageing times of 24 hours produce increases in Ni extraction of ~2% with 0.1 g/L Fe2+ and ~7% with 1 g/L Fe2+. The limited improvement in ageing with Fe2+ suggests that the large majority of available Ni is extracted by ageing with H2O only. The remaining Ni is likely unavailable as unleached pentlandite and interstitially incorporated within the hematite lattice. Chen et al. (2006) suggests that 0.3 - 0.4% Ni and 5.4 - 5.8% SO4 are detected by electron microprobe in the hematite and goethite phase of a Vale (S3) residue. Solid solution Ni within the hematite lattice would be indiscernible from adsorption and occlusion with electron microprobe.  Utilising a complexing anion such as Cl− may achieve similar or even better results. In this instance the anion would replace SO42− and complex any available Cu and Ni. This new chloride species may be less inclined to adsorb to the residue and be released to solution.  This has been tested using an enhanced atmospheric leach with 5 g/L Cl− at 70°C for 2 hours. Results of these experiments are presented further in the present study.    99 4.4.2 Speciation of Cu and Ni losses  A combination of ageing, sequential extraction and Fe displacement experiments were used to determine and quantify the proportions of Cu and Ni lost to surface adsorption/occlusion, ferrihydrite association and incorporated within crystalline Fe phases. The extraction processes aim to target different forms of Cu and Ni present in the residues. Sequential extraction removes Cu and Ni associated with the ferrihydrite phases, ageing dissolves some ferrihydrite and occluded Cu and Ni sulphates and the addition of Fe2+ to the ageing solution targets adsorbed Cu2+ and Ni2+. Stage extractions represent the proportion of metal removed from the solids entering that stage. Total extractions are back calculated to the original residues after each process stage.   Table 22: Extraction of Cu and Ni by each process (%). Ageing at 95°C pH 2, 2% solids; sequential extraction; Fe displacement at 95°C, 2% solids  Residue S1 S3 Process  Cu Stage Cu total Cu stage Cu total Ni stage  Ni total 24 hour ageing  22.2  5.9  22.0 24 hour ageing + 1 g/L Fe treatment 8.9 31.1 3.2 6.2 3.3 23.6 24 hour ageing + 0.1 g/L Fe treatment 13.0 31.0 4.7 10.2 8.2 29.1 24 hour ageing + sequential extraction 28.0 47.1 6.3 14.0 0.0 22.8 24 hour ageing + sequential extraction + 1 g/L Fe treatment 8.0 51.4 0.0 14.0 5.0 26.7 Sequential extraction  50.3  19.2  12.2 Sequential extraction + 24 hour ageing 15.3 57.9 25.2 39.5 20.7 30.4 Stage 1 extraction + 24 hour ageing + sequential extraction 7.8 61.2 3.9 41.9 0.0 30.4  Tenaciously bound Cu and Ni is continually extracted from the residues by further process stages. Removal of significant Cu and Ni from sequentially extracted residues reveals that these tenaciously bound species may be uniformly distributed among Fe phases. These findings match   100 earlier results and published literature, which suggests Cu and Ni are both structurally incorporated in precipitated hematite and uniformly distributed as a surface adsorbed species throughout the leach residue (Dutrizac and Chen 2012b; Dutrizac and Chen 1993). The dissolution of ferrihydrite and crystalline Ca phases by sequential extraction reveals that exposure plays a fundamental role in the metal extraction from both residues. Removing the Fe amorphous phases leaves the tenaciously adsorbed Cu and Ni and occluded Ni and Cu sulphates more susceptible to desorption and dissolution by the ageing and Fe2+ solutions. The reverse process (ageing then sequential extraction) extracts less Cu and Ni than sequential extraction alone. This suggests that Cu and Ni associated with the amorphous Fe phases are closely proximate with the adsorbed and occluded species to the crystalline phases. The removal of the high surface area Fe amorphous phases allows these species to be dissolved with subsequent ageing.    Consecutive treatment of the S3 residue by ageing and sequential extraction further evidences the different nature and compositions of Cu and Ni losses to the residue. No further Ni is extracted when treated by sequential extraction following ageing whereas Cu extraction continues even after two sequential extraction stages and an ageing stage. Little evidence for Ni incorporation with ferrihydrite exists in the literature though Cu2+ has been shown to be most likely to adsorb and/or segregate with the ferrihydrite structure due to it having the ionic radius closest to Fe3+ (Martinez and McBride 1998b). If this is the case, it explains that Ni is more likely associated with the crystalline Fe phases as a surface adsorbed or occluded NiSO4.2H2O species as suggested by Dutrizac and Chen (2012b). Solid solution Ni in the study of Dutrizac   101 and Chen (2012b) was small relative to the bulk Ni content (0.22%) in the hematite. Continued ageing and surface desorption is therefore likely to extract further Ni from the residue.   By differential subtraction of the metal extracted by each process, an indication of the associations of Cu and Ni in each residue was generated. This data is presented in Table 23. The total amount of Cu and Ni extracted are based on a mass balance constructed from the batch treatment of the residues by a single stage of each process. The batch extraction data is back calculated to simulate a continuous process that where each stage is 100% selective for proposed form of association with the residues. The total Cu and Ni represent the expected recovery of all three stages used consecutively. Assuming that the metal remaining is locked inside the crystalline mineral lattices and unrecoverable, the relative contribution of each process to the achievable extraction is obtained.   Table 23: Distribution of Cu and Ni by extraction process in S1 and S3 residues  S1 S3  % Cu % of extract % Cu % of extract % Ni % of extract Ageing 29.7 57.3 2.3 16.6 23.7 86.1 Sequential extraction 18.7 36.1 11.7 83.4 0.0 0.0 Fe displacement 3.4 6.6 0.0 0.0 3.8 13.9 Total 51.8 100.0 14.0 100.0 27.5 100.0  Removal of the occluded sulphate phases by ageing contributes substantially to the metal extractions from both residues. Data analysis by this method indicates that 57.3% of the recoverable Cu in S1 is in a form that does not require the selectivity of the acidified hydroxylamine hydrochloride to remove and can be washed from solution with hot water by ageing the residue. This is not true of Cu contained within S3 as ~83% of the ‘recoverable’ Cu   102 located within the soluble Fe amorphous phase is specifically dissolved by acidified hydroxylamine hydrochloride. Ferrihydrite phases soluble by sequential extraction contribute substantially to the Cu losses in both residues (18.7% and 11.7%) but do not contribute to nickel losses in S3. Ni association within S3 is likely associated with the crystalline iron oxide phases as 3.8% Ni is recovered by desorption with Fe2+ and 23.7% is washed out from the crystalline phases by the ageing solution.  4.5 Repeated ageing stages  Consecutive treatments by a combination of ageing, sequential extraction and Fe2+ displacement revealed that significant Cu and Ni was still extracted from the residues after repeated process stages. S3 appears to reach maximum of 41.9% Cu extraction and 30.4% Ni after two stages of sequential extraction and a single ageing stage (Table 22). Cu extraction of 61.2% was realised from S1 after the same process stages though the incremental 7.8% stage extraction suggests that extraction is not maximised and further Cu can be removed by repeated treatment. After 5 stages of ageing followed by two stages of sequential extraction, 63.6% Cu extractions were achieved. Large incremental stage extractions of 16% and 9% by each sequential extraction stage suggest the residues are still metastable and the significant Cu-containing ferrihydrite is still present in S1. These extractions were reached by averaging extractions of repeatedly recycled S1 residue aged for 1 hour at 95°C at 6, 10 and 15 %w/w solids.   4.5.1 Cu extractions at different %solids by ageing  Ageing the residues at 2, 6, 10 and 15%w/w solids in batch experiments demonstrates variability in Cu extraction as values range from 20.2% to 23.9% and do not appear to show a noticeable   103 trend. The most dilute slurry (2% w/w) extracts the most Cu as expected though the residue aged at 15% w/w solids extracts the next most. These results are close enough to suggest that variability in grab sampling and random experimental error is responsible for the fluctuations. The closeness in the extraction data %solids have minimal, if any, effect on the extractions in the range tested.    Figure 22: Cu, Fe extractions after single ageing stage at 95°C, 2, 6, 10, 15%w/w solids, 1 hour  4.5.2 Cu extraction after repeated ageing of S1 residue Repeated ageing of the S1 residues continues to extract Cu at each stage. Extraction is much reduced after the first stage (22% to 7%) though stays relatively consistent (~5%) after each repetition. Extractions of Fe and S released to the solution remain the same suggesting that the   104 metastable ferrihydrite and associated Cu sulphate species are being gradually washed out of the residue. Residue copper grade gradually decreases to an average of 0.86% Cu after ageing and to 0.73% after two stages of sequential extraction. The average of the total Cu extraction from 4 samples of S1 residue, aged 5 times before two stages of sequential extraction is 63.6%.      Figure 23: Cu stage extraction from S1 after each ageing stage. 95°C, 1 hour, average of 6, 10, 15%w/w solids    105  Figure 24: Total Cu grade and extraction after each stage of ageing and sequential extraction  It is expected that repeated ageing is unable to recover all of the Cu located within ferrihydrite as earlier experiments within the present study showed that ferrihydrite was only marginally affected by the ageing process. Indeed several authors have shown that phase transformations and dissolution of ferrihydrite can take several days, weeks or even years to occur at room temperature and elevated atmospheric temperatures (Johnston and Lewis 1983; Schwertmann and Murad 1983).   Ratios of Cu extraction to Fe and SO4 reveal that it is mostly occluded Cu sulphates and small amounts of ferrihydrite that are dissolved during the ageing process. Similar proportions of Cu to SO4 indicate that similar species are continually removed. The large decrease in the Cu/Fe ratio after the first stage of ageing is due to the decrease in Cu extraction as the easily removed   106 occluded Cu sulphate species have been dissolved. Incremental Cu extractions are the result of further washing of tenaciously adsorbed species and the partial and gradual dissolution of the ferrihydrite phases. Significant ferrihydrite remains as two stages of sequential extraction remove an additional 26.9% and 35.6% Fe respectively at each stage. Dissolution of these species result in significant further Cu extraction.   Figure 25: Ratio of average Cu extraction versus average Fe and SO4 extraction for each stage. 5 ageing stages at 95°C for 1 hour followed by 2 sequential extractions with 0.25 M HCl and 0.25 M hydroxylamine hydrochloride   107  Table 24: Cu, Fe, SO4 stage extractions by ageing and sequential extraction Stage extractions (%) 1 2 3 4 5 Sequential extraction 1 Sequential extraction 2 Cu 23.0 6.9 4.2 3.5 5.4 22.8 20.8 SO4 3.3 1.0 0.9 1.0 1.4 1.9 1.5 Fe 0.2 0.2 0.3 0.3 0.4 26.9 35.6  The continued dissolution of ferrihydrite by acidified hydroxylamine hydrochloride after a single sequential extraction stage is not apparent from a study on similar residues (Sahu and Asselin 2011). This study detailed that all Cu dissolved by a single stage of sequential extraction was due to ferrihydrite and all of the amorphous Fe phases were dissolved by a single extraction. Hydroxylamine hydrochloride is typically a sensitive reagent and small increases in reaction time or process temperature lead to dissolution of additional iron oxides (Chao 1972; Chao and Zhou 1983). The inherent metastability of ferrihydrite may also be responsible as ferrihydrite equilibrium solubility is never reached and exhibits only short-term localised equilibrium. Further ferrihydrite may degrade and dissolve to release additional Cu beyond the timeframe of the experiments. Dry and moist residue samples similar to S1 have been known to release additional Cu after being left in the laboratory for several years (Salomon de Friedberg 2015). Continued ferrihydrite dissolution with acidified hydroxylamine and/or ageing may continue until all Fe is removed from the residue.  4.6 Ageing an alternative CESL residue  A recently produced CESL medium temperature pressure leach residue (S4) from a pilot plant operation was treated by ageing for comparison to the composite S1 residue produced in 2010. Ageing washed S4 residue samples extracted similar proportions of Cu as S1 though the Cu   108 grade of S4 is significantly higher grade. Original S4 samples were insufficiently washed prior to delivery to UBC and contained significant residual CuSO4 and CaSO4 that was removed by rinsing with cold water. The similarity in the properties and Cu extractions of S1 and S4 indicate that any recent process changes or use of different Cu concentrates are not sufficient to overcome the inherent problems that lead to both ferrihydrite formation and excess Cu losses. An average composition of the washed and original, unwashed S4 residues is given in Table 25.   Table 25: Composition of washed and unwashed S4 residue  Washed S4 Unwashed S4 Assay Fe (%) 23.7 21.5 Assay Ni (ppm) 0 334 Assay Cu (%) 1.76 3.23 Assay Ca (ppm) 1915 6206 Assay S (%) 22.5 21.6  A mass balance conducted on the S4 residues shows similarity between Cu and Fe extractions with S1 by ageing and sequential extraction. Results of ageing the original S4 residue are an unfair comparison as significant occluded CuSO4 remains throughout the residue that is not removed during the pilot scale counter current decantation and filtration processes at the plant. Sequential extraction of the washed S4 dissolves similar Cu and Fe as S1 suggesting that the nature and proportion of ferrihydrite is similar in both residues. Ageing data suggests there is no reason for the increase in Cu grade (1.76% versus 1.09%) between S1 and S4 but may be due to incomplete washing of the original S4 residue or additional unleached Cu sulphdies left in the residue.     109 Table 26: Metal extractions and residue grades from ICP-MS analysis after 24 hours of ageing at 95°C, and sequential extraction with 0.25 M hydroxylamine hydrochloride, 0.25 M HCl at 50°C for 30 minutes  Washed S4 Unwashed S4 Sequentially extracted S4 Fe    Initial Fe grade (wt%) 23.7 21.5 23.7 Final Fe grade (wt%) 24.2 24.1 19.1 Fe extracted to solution (%) 0.3 0.5 33.0 Cu    Initial Cu grade (wt%)  1.76 3.23 1.76 Final Cu grade (wt%) 1.22 1.37 1.09 Cu extracted to solution (%) 27.0 65.5 51.6 S    Initial S grade (wt%)  22.5 21.6 22.5 Final S grade (wt%) 22.7 22.8 28.1 S as SO42− extracted to solution (%) 1.9 8.0 3.5  Copper extracted from washed S4 appears more likely related to the dissolution of occluded Cu sulphates than Cu extractions from S1. The results may be skewed due to incomplete cold water washing of easily removed CuSO4 from the S4 residues. The dissolution of Fe by sequential extraction suggests that similar ferrihydrite and therefore Cu in ferrihydrite is similar is present in the two residues.  Table 27: Proportion of residue mass loss contributed to Fe, Cu, Ni and Ca as sulphates  As %MSO4 of total mass loss As %MSO4 of total sulphate  S1 S4 S1 S4  MSO4 48.5 25.9 100.0 100.0 FeSO4 8.6 3.9 2.6 9.6 CuSO4 15.8 24.0 6.0 55.9 CaSO4 37.4 3.8 38.7 10.5 Total 61.8 31.7 68.8 75.9  4.7 Alternative residue treatment processes Alternatives treatment processes to ageing were explored to assess whether significant additional Cu could be extracted from the residue. These processes may not be applicable in a metallurgical   110 processing environment but potentially reveal future avenues for more research or reveal the nature of Cu associations and extractions. An enhanced atmospheric leach (EAL) process shows the most promise for further study. Results of pre-treating the residue by dry heating in a furnace prior to ageing were inconclusive but show evidence of encapsulating rather than releasing Cu to solution.   4.7.1 Enhanced atmospheric leach An enhanced atmospheric leach is used by CESL to assess the efficiency of the pressure leaching process and closely resembles the atmospheric leach stage practised industrially. At the laboratory level addition of 5 g/L Cl− to a buffered pH 1.5 solution at 70°C for two hours is used to simulate the atmospheric leaching process. EAL extracts an additional ~2-3% Cu and ~0.2-0.3% Fe from the S1 and S4 residues than ageing does.   Table 28: Enhanced atmospheric leach of S1 and S4 residues. 5 g/L Cl-, 70°C, 2 h, pH 1.5, 15%w/w solids Residue S1 S4 %Cu extraction 25.6 28.1 %Cu grade after EAL 0.90 1.16 %Fe extraction  0.5 0.6 %Fe grade after EAL 25.3 24.1 %S extraction 3.9 2.5 %S grade after EAL 28.2 22.7 %Metal sulphates/mass loss 94.7 72.3 %CuSO4 in total sulphate 12.2 39.7  Enhanced acid leaching improves the extraction of Cu and SO4 from the residue compared with ageing. A greater proportion of the solids mass loss from EAL is due to metal sulphates, which is likely the source of additional Cu extraction. Assuming Ca, Fe, and Cu sulphates (as CuSO4,   111 CaSO4 and FeSO4) are responsible for S extraction, these species contribute 94.7% and 72.3% of the mass loss for S1 and S4 respectively. This compares with 61.8% and 68.8% of the mass loss due to these same sulphates for S1 and S4 from ageing. The improvement in extraction of sulphates is could be due to the presence of the chloride ion or the reduced pH of 1.5 from the ageing and natural pH of the S1 residue of pH >2. Acid conditions of pH 1.4 - 1.8 maximise the leaching of soluble basic copper sulphate (CuSO4.2Cu(OH)2) while minimising the leaching of hematite in the industrial CESL process (Defreyne et al. 2006).   Chloride ion is added to CESL pressure leach process and acts as a catalyst to reduce the need to ultrafine grind the concentrate for fast kinetics (McDonald and Muir 2007a; Defreyne et al. 2006). Addition of Cl− ions as cupric chloride prevents the formation of hydronium jarosite that consumes a significant fraction of the initial acid during the leach. Potassium chloride used in EAL helps to buffer the solution pH during the leach though may assist in removing Cu sulphates from the residues. Addition of a strong anion has been shown to remove sulphate adhered to iron oxides (Dutrizac and Chen 1993) that may also contain Cu occluded as CuSO4. Further work is required to definitively determine whether addition of strong anions such as Cl− to the residue improves the Cu extraction.   4.7.2 Residue pre-treatment by dry heating Dry heating the S1 residue at 150°C prior to ageing was conducted to dehydrate the residue and remove any water from the ferrihydrite structure to facilitate transformation. Transformation is not expected to release Cu as Cu gets incorporated into the crystal lattice of hematite and goethite (Cornell and Giovanoli 1988). It is well known that α-Fe2O3 is the end-product of   112 thermal decomposition of ferrihydrite, iron oxyhydroxides or iron salts in oxidative atmosphere (Cornell, Schwertmann, and Editors. 1996). Slight ferrihydrite transformation to hematite occurs by dry heating at 245°C and full conversion is noted at 325°C (Ristić et al. 2007). Absence of water during dry heating leads to gradual structural ordering within ferrihydrite towards the hematite structure. Increased temperatures enhance the crystallinity of the species (Schwertmann, Friedl, and Stanjek 1999).   Results of ageing heated residues are quite inconclusive though it does appear that ~3 times less Cu is extracted to solution than if the residues are unheated. Transformation is unlikely at the low temperatures of heating though heating may aid in ferrihydrite dissolution as ~3 times more Fe is extracted by ageing. Another mechanism may be responsible for the reduction of Cu in the residue, perhaps Cu is tenaciously re-adsorbed to the surface of other phases as the ferrihydrite degrades. Further work is needed in this area to confirm the nature of the changes in the residue by heating and dehydration.        113 Chapter 5: Conclusion 5.1 Summary of results Sulphide residues of the CESL Cu (S1) and Vale Ni (S3) medium temperature pressure leach process contain Cu and Ni and a significant X-ray amorphous phase that is largely metastable ferrihydrite. Formation of ferrihydrite in the residues results from the incomplete nucleation and growth of soluble ferric species during the leaching of the sulphide minerals and simultaneous precipitation of waste Fe as hematite. The inherent metastability of ferrihydrite means that it will transform to hematite with sufficient time. This process is slow and the ferrihydrite and other poorly formed precipitates are more likely to degrade during impoundment in a tailings storage facility. Labile ferrihydrite limits the overall stability of the process residues and may potentially release toxic elements such as arsenic to the tailings storage facility and environment.   Loss of valuable metals to the residue represents a significant lost economic opportunity. Some of the metal losses are associated with the ferrihydrite phase due to co-precipitation as surface adsorbed species or occluded within the individual crystallites and larger agglomerates. Adsorption to ferrihydrite is facilitated due to its specific large surface area (200-700 m2/g) and availability of surface binding sites (Jambor and Dutrizac 1998; Benjamin and Leckie 1981). Co-precipitation of Cu and Ni with the iron oxide minerals affects the transformation of ferrihydrite to hematite and often traps Cu and Ni within the crystal matrix (Martinez and McBride 1998a). Dissolution of the ferrihydrite phase by sequential extraction with hydroxylamine hydrochloride releases 50.3% of the Cu from the S1 residue and 19.2% of the Cu and 12.3% of the Ni from the S3 residue revealing that minimising the formation of this phase will have a beneficial economic effect on the process.    114 Some of the Cu and Ni lost to the residues are attributable to other species found in the residue. The ageing process utilised by the present study removed 22% Cu from S1 and 6% Cu and 22% Ni from S3 after 24 hours without dissolving significant ferrihydrite. Cu and Ni removed by ageing is essentially washed from the residue and is likely source from adsorbed and occluded CuSO4, CuSO4.2Cu(OH)2, and NiSO4.H2O attached to the surfaces of the crystalline iron oxides. These species were not noticed by SEM-EDS analysis of the residues but EMP analysis of hematite particles in the S3 residues showed the presence of Cu, Ni and S in all 18 hematite particles analysed. Average grade of the hematite particles analysed from a wide selection of particle sizes was 66.1% Fe, 0.73% Cu, 0.85% Ni and 1.19% SO4. Given that the S3 residue constitutes 35% hematite, an average particle such as this represents 30% of Ni losses and 25% of the Cu losses, or a grade of 0.25% Ni and 0.15% Cu of the total residue. This agrees with published literature (Dutrizac and Chen 2012b) that suggests Cu and Ni are lost to surface bound Cu and Ni sulphates on hematite. The authors suggest that metal losses to incorporation within the hematite crystal structure is insignificant relative to the bulk losses though this was not able to be determined by EMPA in the present study. Losses of Cu in S1 are also likely to be related to surface bound sulphate species evenly distributed throughout the residue though the lack of defined crystalline Fe phases makes determining the nature of this association difficult.   Repeated 1 hour ageing of the S1 residue continued to release Cu and small amounts of Fe to solution. After ageing the residue five times 38% of the Cu was extracted to solution. A further 24% Cu was extracted with an additional two sequential extraction stages. Continued incremental Cu extraction by ageing suggests the tenacious nature of the association and even seven stages of treatment were unable to reach the maximum possible extraction of Cu. Utilising   115 hydroxylamine hydrochloride as a reagent is impractical outside of bench scale so 38% represents the maximum extraction of Cu from the residue and even this may not be economically practical. Extraction of Cu and Ni from S3 shows only marginal improvement by sequential extraction following ageing suggesting that extraction of 6% Cu and 22% Ni from a single ageing stage is fairly close to a practical maximum. Up to 42% Cu and 31% Ni were removed from treatment with two sequential extraction stages and ageing suggesting that the assumptions from EMP analysis are fairly accurate.   Cu and Ni losses not associated with ferrihydrite or with the crystalline iron oxides, either structurally incorporated or surface bound, are unleached or partially leached sulphides. Evidence of unleached pentlandite and chalcopyrite were found in the S3 residues. S1 residue contained widespread enargite and tennantite minerals either liberated or within a complex matrix of amorphous agglomerations of other phases that include ferrihydrite. Losses to these phases are unlikely to appear in QXRD diffractograms due to their low concentrations and prevalence of interference from background scatter. These minerals are not extracted by ageing and would require recycling to the pressure leach stage in an autoclave to recover the lost metal values.   Evidence of transformation of ferrihydrite to the more crystalline iron oxides, hematite and goethite was not conclusively evidenced following ageing. Ageing was able to dissolve approximately 0.3% of the Fe from the residue that is most likely ferrihydrite. This contributed to improved stability of the residues by dramatically reducing the X-ray amorphous phase. Mass loss to sequential extraction was greatly lessened due to this reduction. It was unclear whether   116 additional hematite or goethite was formed in the aged residues when the residues were analysed by QXRD. High temperature ageing at 150°C for 12 hours appeared to produce additional hematite at the expense of goethite and ferrihydrite though it was difficult to categorically determine this was the case and the result was not repeated elsewhere. Ageing likely dissolves the more unstable ferrihydrite phases and partially stabilises the remaining ferrihydrite and poorly formed iron oxides though it is not clear whether this results in increased crystallinity of the resulting phases.    5.2 Significance of results  Residues of the CESL and Vale pressure leach contain significant Cu and Ni that are difficult to recover by simple, conventional hydrometallurgical treatment and are likely uneconomic by the methods used in the present study. This study did show that properly precipitated crystalline hematite species contain important Cu and Ni concentrations. This is significant as Cu and Ni losses to the residue may be unavoidable and unrecoverable. Transformation of the Cu and Ni occluding ferrihydrite species by an ageing process does not appear to release significant amounts of the lost Cu and Ni. This transformation process may even further encapsulate the metals within the matrix of the transformed species. It would be of greater benefit to study methods of reducing the deportment of Cu and Ni to the iron oxide precipitates during precipitation than attempting to remove the lost metals from these phases. This is the goal of the larger project of which this present study is a small part. This project largely achieved the goals of finding practical limits of recovery but further work is required to definitively allocate the losses of each metal to each species within the residue.     117 Improving the CESL and Vale sulphide pressure leaching processes to improve metal recovery and reduce losses to the precipitated iron oxide residues requires further work. A better understanding of ferric hydrolysis to create a more stable, more pure hematite precipitate that contains much less Cu and Ni is required. This would achieve numerous economic benefits that would greatly improve the attractiveness of these processes. It may not be possible to create a hematite product completely devoid of sulphate, Cu and Ni such that it can be resold though understanding the process by which these are reduced would be highly beneficial to hydrometallurgy. This is likely related to creating better hematite precipitates over ferrihydrite so transformation studies of ferrihydrite to hematite in a hydrometallurgical environment may not be required.   The presence of significant sulphate inextricable to the iron oxide precipitates is a major finding of the present study. Ni2+ and Cu2+ often counter the sulphates as discrete particulates that adsorb and occlude both to the well-formed hematite species and are bound within ferrihydrite agglomerates. Losses to the residue of these species is likely both a function of slow mass transfer of leached CuSO4 and NiSO4 away from the surface of the leached minerals and also a result of ‘robbing’ these species from the bulk solution by physical and chemical adsorption. These processes occur independently of ferrihydrite formation though the presence of significant ferrihydrite allows for greater amounts of these sulphate species to be lost to the residue. This is due to the increased surface area of the ferrihydrite agglomerates that provides ample adsorption sites and likely further limits the mass transfer of the leached species away from the mineral surfaces. Improving the Fe3+ hydrolysis process to reduce ferrihydrite formation is an important   118 focus for further hydrometallurgical research though further work in exploring and reversing the mechanisms of metal sulphate occlusion in iron oxide residues is also worth investigating.   5.3 Strengths and limitations of research  The project benefited from using simple experimental and analytical techniques to enhance the understanding of the nature of the Cu and Ni losses to each of the residues. Using a simple washing and ageing set-up found that significant losses were due to sulphates and other adsorbed and occluded species that are difficult, if not, impossible to determine by microscopic methods. This will benefit the future direction of the project by focussing on ways to improve washing of these species from the residue or to reduce adsorption and occlusion of these phases.   It was beneficial to assess the stability of ferrihydrite at more commonly encountered hydrometallurgical conditions rather than use sophisticated spectroscopic techniques such as EXAFS and TEM. Gaining structural and improved characterisation information by these techniques does not give an indication of the behaviour and stability of this phase nor how much Cu and Ni can be feasibly released using conventional hydrometallurgy.   The major limitation of the project was the inability to determine the exact speciation of the Cu and Ni associated with the residues. It was assumed by mass balance and literature review that the adsorbed and occluded species that did not appear to be associated with ferrihydrite were sulphates. These could either be physically occluded CuSO4.2(OH)2 that are not dissolved by atmospheric leaching or CuSO4 and NiSO4.H2O that are improperly washed or ‘robbed’ from solution by adsorption to the residues. Determining the exact speciation of the metal sulphate   119 species is difficult and requires more complex analysis or more comprehensive chemical studies. The only suggestion in the literature was that these species were simple Cu and Ni sulphates that were found by close examination with SEM-EDS and EMPA though these are not exact. No method for finding the exact speciation of sulphates associated with hematite residues was found in the literature.   5.4 Applications of research and future work Ageing studies were beneficial in further characterising the residue though the application of a standalone ageing process within the CESL and Vale process flowsheet is not practical. The information gathered regarding the nature of the Cu and Ni losses should be utilised to prevent and/or minimise the association of Cu and Ni sulphates with the precipitated crystalline iron oxides and ferrihydrite. This could be conducted in conjunction with studies that focus on washing Cu and Ni sulphates from the residue and transforming the ferrihydrite in conditions of the CESL and Vale leach. This is best achieved by recycling the residue to the start of the pressure leach. This is both of academic interest and a practical solution to minimise the need for additional process equipment and stages. Recycling the residue has the benefit of providing a seed for the growth of hematite from the leach solution, provided sufficient hematite can be initially generated from the CESL pressure leach. Vale residue recycle would likely form hematite from the ferrihydrite present and recover some of the Cu and Ni associated with this phase. Sulphide pressure leach residue recycle is already practiced industrially as discussed by Marsden et al. (2002) and could be optimised for CESL and Vale process conditions. It is uncertain whether this process would wash some of the metal sulphates from the hematite though this could be established with more fundamental studies utilising CESL conditions.    120 Fundamental studies of the transformation of ferrihydrite in CESL conditions are needed to assess whether or not Cu and Ni are released by the change in phase to hematite. This is discussed in soil systems (Martinez and McBride 1998a) but has not been explored in hydrometallurgy. This may require formation of ferrihydrite in artificial solutions without the leaching of concentrates.     121 Bibliography Abdul, B., M. Pernechele, D. H. Ryan, and E. Asselin. 2014. Mossbauer studies on synthetic jarosites doped with copper. In: Asselin, E., Dixon, D.G., Doyle, F.M., Dreisinger, D.B., Jeffrey, M.I., Moats, M.S. (Eds.), Hydrometallurgy 2014. Canadian Institute of Mining, Metallurgy and Petroleum, Montreal, Canada, 1, 113 – 126.  Ali, Muhammad A., and David A. Dzombak. 1996. Competitive sorption of simple organic acids and sulfate on goethite. Environmental Science and Technology 30 (4): 1061-71.  ———. 1996. Effects of simple organic acids on sorption of Cu2+ and Ca2+ on goethite. Geochimica Et Cosmochimica Acta 60 (2): 291-304.  ———. 1996. Interactions of copper, organic acids, and sulfate in goethite suspensions. Geochimica Et Cosmochimica Acta 60 (24): 5045-53.  Arai, Yuji. 2008. Spectroscopic evidence for ni(II) surface speciation at the iron oxyhydroxides-water interface. Environmental Science & Technology 42 (4): 1151-6.  Arima, H., T. Aichi, Y. Kudo, K. Saruta, M. Kanno, and R. Togashi. 2006. Recent improvement in the hematite precipitation process at the akita zinc company. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 123-134. Canadian Institute of Mining, Metallurgy and Petroleum  Baron, Dirk, and Carl D. Palmer. 1996. Solubility of jarosite at 4-35°C. Geochimica Et Cosmochimica Acta 60 (2): 185-95.  Barrow, N. J., J. W. Bowden, A. M. Posner, and J. P. Quirk. 1980. Describing the effects of electrolyte on adsorption of phosphate by a variable charge surface. Soil Research 18 (4) (01/01): 395-404.  Beaulieu, R., G. Gagne, M. Nasmyth, G. Cooper, and C. Inostroza. 2006. Iron control and management in the zinc industry. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 45-55. Canadian Institute of Mining, Metallurgy and Petroleum  Benjamin, Mark M., and James O. Leckie. 1981. Multiple-site adsorption of cadmium, copper, zinc, and lead on amorphous iron oxyhydroxide. Journal of Colloid and Interface Science 79 (1): 209-21.  Bigham, J. M., L. Carlson, and E. Murad. 1994. Schwertmannite, a new iron oxyhydroxy-sulfate from pyhasalmi, finland, and other localities. Mineralogical Magazine 58 (393): 641-8.    122 Blesa, Miguel A., and Egon Matijevic. 1989. Phase transformations of iron oxides, oxohydroxides, and hydrous oxides in aqueous media. Advances in Colloid and Interface Science 29 (3-4): 173-221.  Boukhalfa, Chahrazed. 2010. Sulfate removal from aqueous solutions by hydrous iron oxide in the presence of heavy metals and competitive anions: Macroscopic and spectroscopic analyses. Desalination 250 (1) (1/1): 428-32.  Boukhalfa, Chahrazed, Ammar Mennour, Laurence Reinert, and Herve Fuzellier. 2007. Sulfate removal from aqueous solutions by hydrous iron oxide macroscopic, thermal and spectroscopic analyses. Desalination 214 (1–3) (8/15): 38-48.  Breeuwsma, A., and J. Lyklema. 1973. Physical and chemical adsorption of ions in the electrical double layer on hematite (α-iron(III) oxide). Journal of Colloid and Interface Science 43 (2): 437-48.  Chao, T. T., and Liyi Zhou. 1983. Extraction techniques for selective dissolution of amorphous iron oxides from soils and sediments. Soil Science Society of America Journal 47 (2): 225-32.  Chao, T. T. 1984. Use of partial dissolution techniques in geochemical exploration. Journal of Geochemical Exploration 20 (2) (4): 101-35.  ———. 1972. Selective dissolution of manganese oxides from soils and sediments with acidified hydroxylamine hydrochloride. Soil Sci.Soc.Am.J. 36 (5) (1972): 764-8.  Chao, T. T., and P. K. Theobald J. 1976. The significance of secondary iron and manganese oxides in geochemical exploration. Economic Geology and the Bulletin of the Society of Economic Geologists 71 (8): 1560-9.  Charlet, Laurent, Nancy Dise, and Werner Stumm. 1993. Sulfate adsorption on a variable charge soil and on reference minerals. Agriculture, Ecosystems & Environment 47 (2): 87-102.  Chen, T. T., J. E. Dutrizac, G. Poirier, D. G. Kerfoot, and A. Singhal. 2006. Characterization of the iron-rich residues generated during the pressure oxidative leaching of voisey's bay nickel sulphide concentrate. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 429-444. Canadian Institute of Mining, Metallurgy and Petroleum  Cheng, C. Y., G. H. Kelsall, and D. Pilone. 2005. Modeling potentials, concentrations and current densities in porous electrodes for metal recovery from dilute aqueous effluents. Journal of Applied Electrochemistry 35 (12): 1191-202.  Chester, R., and M. J. Hughes. 1967. A chemical technique for the separation of ferro-manganese minerals, carbonate minerals and adsorbed trace elements from pelagic sediments. Chemical Geology 2 (0): 249-62.    123 Cornell, R. M., R. Giovanoli, and W. Schneider. 1989. Review of the hydrolysis of iron (III) and the crystallization of amorphous iron (III) hydroxide hydrate. Journal of Chemical Technology and Biotechnology 46 (2): 115-34.  Cornell, R. M., and R. Giovanoli. 1988. The influence of copper on the transformation of ferrihydrite (5Fe2O3·9H2O) into crystalline products in alkaline media. Polyhedron 7 (5): 385-91.  Cornell, R. M., U. Schwertmann, and Editors. 1996. The iron oxides: Structure, properties, reactions, occurrence and uses.  Cudennec, Yannick, and Andre Lecerf. 2006. The transformation of ferrihydrite into goethite or hematite, revisited. Journal of Solid State Chemistry 179 (3): 716-22.  Davis, James A., and James O. Leckie. 1980. Surface ionization and complexation at the oxide/water interface. 3. adsorption of anions. Journal of Colloid and Interface Science 74 (1): 32-43.  ———. 1979. Speciation of adsorbed ions at the oxide/water interface. ACS Symposium Series 93 : 299-317.  ———. 1978. Surface ionization and complexation at the oxide/water interface. 2. surface properties of amorphous iron oxyhydroxide and adsorption of metal ions. Journal of Colloid and Interface Science 67 (1): 90-107.  Defreyne, J., W. Grieve, D. L. Jones, and K. Mayhew. 2006. The role of iron in the CESL process. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 205-220. Canadian Institute of Mining, Metallurgy and Petroleum  Dousma, J., D. den Ottelander, and P. L. de Bruyn. 1979. The influence of sulfate ions on the formation of iron(III) oxides. Journal of Inorganic and Nuclear Chemistry 41 (11): 1565-8.  Dousma, J., T. J. Van den Hoven, and P. L. De Bruyn. 1978. The influence of chloride ions on the formation of iron(III) oxyhydroxide. Journal of Inorganic and Nuclear Chemistry 40 (6): 1089-93.  Dreisinger, David. 2006. Copper leaching from primary sulfides: Options for biological and chemical extraction of copper. Hydrometallurgy 83 (1-4): 10-20.  Drits, V. A., B. A. Sakharov, A. L. Salyn, and A. Manceau. 1993. Structural model for ferrihydrite. Clay Minerals 28 (2): 185-207.  Dutrizac, J. E., and T. T. Chen. 2011. Hematite precipitation from ferric sulphate solutions at 225 and 100°C. Eur. Metall. Conf. 2011 (3): 895-918, GDMG Informations GmBH   124 ———. 1993. Surface and structural impurity incorporation in iron precipitates. Emerging Separation Technologies, Metallurgy, Fuels Process Symposium 1993: 183-198, The Mining, Metallurgy and Materials Society. Dutrizac, J. E., and P. A. Riveros. 1999. The precipitation of hematite from ferric chloride media at atmospheric pressure. Metallurgical and Materials Transactions B: Process Metallurgy and Materials Processing Science 30B (6): 993-1001.  Dutrizac, J. E. 1999. The effectiveness of jarosite species for precipitating sodium jarosite. JOM 51 (12): 30-2.  ———. 1979. The physical chemistry of iron precipitation in the zinc industry. Lead-Zinc-Tin '80, Proc.World Symp.Metall.Environ.Control: 532-564. Metall. Soc. AIME. Dutrizac, J. E., and A. Kuiper. 2006. The solubility of calcium sulphate in simulated nickel sulphate–chloride processing solutions. Hydrometallurgy 82 (1–2) (7): 13-31.  Dutrizac, J. E., and A. Sunyer. 2012. Hematite formation from jarosite type compounds by hydrothermal conversion. Canadian Metallurgical Quarterly 51 (1): 11-23.  Dutrizac, John E., and Tzong T. Chen. 2012. Behaviour of various impurities during the precipitation of hematite from ferric sulphate solutions at 225°C. T.T.Chen Honorary Symp.Hydrometall., Electrometall.Mater.Charact., Proc: 489-499. John Wiley & Sons, Inc ———. 2012. Impurity incorporation in hematite precipitated from ferric sulphate solutions. World of Metallurgy--Erzmetall 65 (1): 31-47.  ———. 2011. Precipitation of hematite directly from ferric sulphate solutions. World of Metallurgy--Erzmetall 64 (3): 134-50.  ———. 2010. Precipitation of Fe4(SO4)(OH)10 and hydronium jarosite, (H3O)Fe3(SO4)2(OH)6, in ferric sulphate-sulphuric acid media. World of Metallurgy--Erzmetall 63 (4): 181-96.  Dutrizac, John E., and John L. Jambor. 2000. Jarosites and their application in hydrometallurgy. Reviews in Mineralogy & Geochemistry 40 : 405-52.  Dyer, James A., Paras Trivedi, Noel C. Scrivner, and Donald L. Sparks. 2004. Surface complexation modeling of zinc sorption onto ferrihydrite. Journal of Colloid and Interface Science 270 (1): 56-65.  Eggleston, Carrick M., Stephan Hug, Werner Stumm, Barbara Sulzberger, and Maria Dos Santos Afonso. 1998. Surface complexation of sulfate by hematite surfaces: FTIR and STM observations. Geochimica Et Cosmochimica Acta 62 (4) (2): 585-93.    125 Eggleton, Richard A., and Robert W. Fitzpatrick. 1988. New data and a revised structural model for ferrihydrite. Clays and Clay Minerals 36 (2): 111-24.  Elzinga, E. J., D. Peak, and D. L. Sparks. 2001. Spectroscopic studies of pb(II)-sulfate interactions at the goethite-water interface. Geochimica Et Cosmochimica Acta 65 (14) (7/1): 2219-30.  Elzinga, E. J, D. L. Sparks, and D, Peak. 2001. Understanding sulfate adsorption mechanisms on iron (III) oxides and hydroxides. In Heavy Metals Release in Soils: 167-190. CRC Press.  Erdemoğlu, Murat, and Musa Sarıkaya. 2006. Effects of heavy metals and oxalate on the zeta potential of magnetite. Journal of Colloid and Interface Science 300 (2) (8/15): 795-804.  Evanko, Cynthia R., and David A. Dzombak. 1999. Surface complexation modeling of organic acid sorption to goethite. Journal of Colloid and Interface Science 214 (2) (6/15): 189-206.  Ferron, C. J. 2006. Iron control in hydrometallurgy: The positive side of the coin. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 429-444. Canadian Institute of Mining, Metallurgy and Petroleum.  Fredrickson, James K., John M. Zachara, Ravi K. Kukkadapu, Yuri A. Gorby, Steven C. Smith, and Christopher F. Brown. 2001. Biotransformation of ni-substituted hydrous ferric oxide by an fe(III)-reducing bacterium. Environmental Science and Technology 35 (4): 703-12.  Fuller, Christopher C., James A. Davis, and Glenn A. Waychunas. 1993. Surface chemistry of ferrihydrite: Part 2. kinetics of arsenate adsorption and coprecipitation. Geochimica Et Cosmochimica Acta 57 (10): 2271-82.  Gathje, J. C. 2006. Iron reduction in copper leach liquors using controlled precipitation of sulfate species during high temperature pressure oxidation of base metal ores. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 231-246. Canadian Institute of Mining, Metallurgy and Petroleum. Gräfe, Markus, David A. Beattie, Euan Smith, William M. Skinner, and Balwant Singh. 2008. Copper and arsenate co-sorption at the mineral–water interfaces of goethite and jarosite. Journal of Colloid and Interface Science 322 (2) (6/15): 399-413.  Haavanlammi, L., O. Hyvarinen, and J. Karonen. 2006. Iron behavior in the hydrocopper process. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 221-229. Canadian Institute of Mining, Metallurgy and Petroleum. Hall, G. E. M., J. E. Vaive, R. Beer, and M. Hoashi. 1996. Selective leaches revisited, with emphasis on the amorphous fe oxyhydroxide phase extraction. Journal of Geochemical Exploration 56 (1) (6): 59-78.    126 Harris, G. B., C. W. White, and G. P. Demopoulos. 2006. Iron control in high-concentration chloride leaching processes. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 445-464. Canadian Institute of Mining, Metallurgy and Petroleum. Hug, Stephan J. 1997. In situ fourier transform infrared measurements of sulfate adsorption on hematite in aqueous solutions. Journal of Colloid and Interface Science 188 (2): 415-22.  Hug, Stephan J., and Barbara Sulzberger. 1994. In situ fourier transform infrared spectroscopic evidence for the formation of several different surface complexes of oxalate on TiO2 in the aqueous phase. Langmuir 10 (10) (10/01; 2014/04): 3587-97.  Jambor, John L., and John E. Dutrizac. 1998. Occurrence and constitution of natural and synthetic ferrihydrite, a widespread iron oxyhydroxide. Chemical Reviews (Washington, D.C.) 98 (7): 2549-85.  Janney, Dawn E., J. M. Cowley, and Peter R. Buseck. 2000. Structure of synthetic 2-line ferrihydrite by electron nanodiffraction. American Mineralogist 85 (9): 1180-7.  Janney, Dawn E., John M. Cowley, and Peter R. Buseck. 2000. Transmission electron microscopy of synthetic 2- and 6-line ferrihydrite. Clays and Clay Minerals 48 (1): 111-9.  Javed, T., B. Abdul, M. Raudsepp, D. Ryan, and Asselin. E. 2015. Amorphous iron phases in medium temperature sulphide concentrate leach residues from pilot and demonstration plants. International Journal of Mineral Processing (unpublished).  Jiang, Jun, Ren-Kou Xu, and Su-Zhen Li. 2010. Effect of ionic strength and mechanism of Cu(II) adsorption by goethite and α-Al2O3. Journal of Chemical & Engineering Data 55 (12): 5547-52.  Johnston, J. H., and D. G. Lewis. 1983. A detailed study of the transformation of ferrihydrite to hematite in an aqueous medium at 92°C. Geochimica Et Cosmochimica Acta 47 (11): 1823-31.  Jung, Jinho, Young-Hwan Cho, and Pilsoo Hahn. 1998. Comparative study of Cu2+ adsorption of goethite, hematite and kaolinite: Mechanistic modeling approach. Bulletin of the Korean Chemical Society 19 (3): 324-7.  Kabel, J. 2015. Personal communication Kerfoot, D. G. E., E. Krause, B. J. Love, and A. Singhal. 2002. Hydrometallurgical process for the recovery of nickel and cobalt values from a sulfidic flotation concentrate. Google Patents.    127 Lahtinen, M., K. Svens, and L. Lehtinen. 2006. Hematite versus jarosite precipitation in zinc production. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 93-107. Canadian Institute of Mining, Metallurgy and Petroleum.   Langmuir, Donald. 1969. Gibbs free energies of substances in the system Fe-O2=H2O-CO2 at 25°C. U.S.Geol.Surv, Prof.Pap. 650: 180-184. Li, Yu. 2008. Cu2+ and Zn2+ adsorption to synthetic iron oxides and natural iron ore powder. 2nd International Conference on Bioinformatics and Biomedical Engineering 2008. Liu, Hui, Hui Guo, Ping Li, and Yu Wei. 2008. The transformation of ferrihydrite in the presence of trace Fe(II): The effect of the anionic media. Journal of Solid State Chemistry 181 (10): 2666-71.  Liu, Hui, Ping Li, Meiying Zhu, Yu Wei, and Yuhan Sun. 2007. Fe(II)-induced transformation from ferrihydrite to lepidocrocite and goethite. Journal of Solid State Chemistry 180 (7): 2121-8.  Loan, M., O. M. G. Newman, R. M. G. Cooper, J. B. Farrow, and G. M. Parkinson. 2006. Defining the paragoethite process for iron removal in zinc hydrometallurgy. Hydrometallurgy 81 (2): 104-29.  Loan, M., W. R. Richmond, and G. M. Parkinson. 2005. On the crystal growth of nanoscale schwertmannite. Journal of Crystal Growth 275 (1-2): e1875-81.  Loan, M., Pierre St. T. G., G. M. Parkinson, O. G. M. Newman, and J. B. Farrow. 2002. Identifying nanoscale ferrihydrite in hydrometallurgical residues. Jom 54 (12): 40-3.  Loan, Mitch, John M. Cowley, Robert Hart, and Gordon M. Parkinson. 2004. Evidence on the structure of synthetic schwertmannite. American Mineralogist 89 (11-12): 1735-42.  Loan, Mitch, O. G. Mike Newman, John B. Farrow, and Gordon M. Parkinson. 2008. Effect of rate of crystallization on the continuous reactive crystallization of nanoscale 6-line ferrihydrite. Crystal Growth & Design 8 (4): 1384-9.  Loan, Mitch, O. M. G. Newman, John B. Farrow, and Gordon M. Parkinson. 2006. Continuous reactive crystallization of nanoscale six-line ferrihydrite. Crystal Growth & Design 6 (1): 79-86.  Loan, Mitch, Gordon M. Parkinson, and William R. Richmond. 2005. The effect of zinc sulfide on phase transformations of ferrihydrite. American Mineralogist 90 (1): 258-61.  Loan, Mitch, Gordon Parkinson, Mike Newman, and John Farrow. 2002. Iron oxy-hydroxide crystallization in a hydrometallurgical residue. Journal of Crystal Growth 235 (1-4): 482-8.    128 Loan, Mitch, William R. Richmond, James Hockridge, Pierre St. Timothy G., Robert Hart, O. G. Mike Newman, John B. Farrow, and Gordon M. Parkinson. 2005. Characterization of ironIII oxyhydroxides in hydrometallurgical residues. EPD Congr.2005, Proc.Sess.Symp.TMS Annu.Meet. Minerals, Metals & Materials Society. Manceau, A. 2009. Evaluation of the structural model for ferrihydrite derived from real-space modelling of high-energy X-ray diffraction data. Clay Minerals 44 (1): 19-34.  Manceau, A., J. M. Combes, and G. Calas. 1990. New data and a revised structural model for ferrihydrite: Comment. Clays and Clay Minerals 38 (3): 331-4.  Manceau, A., and V. A. Drits. 1993. Local structure of ferrihydrite and feroxyhite feroxyhyte] by EXAFS spectroscopy. Clay Minerals 28 (2): 165-84.  Manceau, A., and W. P. Gates. 1997. Surface structural model for ferrihydrite. Clays and Clay Minerals 45 (3): 448-60.  Marsden, J. O., R. E. Brewer, S. R. Brewer, J. M. Robertson, D. R. Baughman, P. Thompson, W. W. Hazen, and C. M. A. Bemelmans. 2010. Process for recovery of copper from copper-bearing material using pressure leaching, direct electrowinning and solvent/solution extraction. Google Patents.  Marsden, J. O., R. E. Brewer, J. M. Robertson, D. R. Baughman, P. Thompson, W. W. Hazen, and R. Schmidt. 2002. Method for improving metals recovery using high temperature leaching. Google Patents.  Martinez, Carmen Enid, and Murray B. McBride. 1998. Coprecipitates of cd, cu, pb and zn in iron oxides: Solid phase transformation and metal solubility after aging and thermal treatment. Clays and Clay Minerals 46 (5): 537-45.  ———. 1998. Solubility of Cd2+, Cu2+, Pb2+, and Zn2+ in aged coprecipitates with amorphous iron hydroxides or oxides. Environmental Science and Technology 32 (6): 743-8.  McDonald, R. G., and D. M. Muir. 2007. Pressure oxidation leaching of chalcopyrite. part I. comparison of high and low temperature reaction kinetics and products. Hydrometallurgy 86 (3–4) (5): 191-205.  ———. 2007. Pressure oxidation leaching of chalcopyrite: Part II: Comparison of medium temperature kinetics and products and effect of chloride ion. Hydrometallurgy 86 (3–4) (5): 206-20.  Meng, Shan, Huanling Wang, Hui Liu, Caihong Yang, Yu Wei, and Denglu Hou. 2014. Evaluation of the ability of ferrihydrite to bind heavy metal ions: Based on formation environment, adsorption reversibility and ageing. Applied Geochemistry 45: 114-9.    129 Mohapatra, M., L. Mohapatra, D. Hariprasad, S. Anand, and B. K. Mishra. 2012. Nano-structured mg-doped Fe2O3-ferrihydrite powder - a new adsorbent for cation removal from aqueous solutions. Environmental Technology 33 (15): 1717-26.  Moon, Ellen M., and Caroline L. Peacock. 2013. Modelling cu(II) adsorption to ferrihydrite and ferrihydrite-bacteria composites: Deviation from additive adsorption in the composite sorption system. Geochimica Et Cosmochimica Acta 104 : 148-64.  Muller, Katharina, and Gregory Lefebvre. 2011. Vibrational characteristics of outer-sphere surface complexes: Example of sulfate ions adsorbed onto metal (hydr)oxides. Langmuir 27 (11) (06/07; 2014/05): 6830-5.  Neaman, A., B. Waller, F. Mouélé, F. Trolard, and G. Bourrié. 2004. Improved methods for selective dissolution of manganese oxides from soils and rocks. European Journal of Soil Science 55 (1) (03/01): 47-54.  Nelson, Hanna, Staffan Sjoberg, and Lars Lovgren. 2013. Surface complexation modelling of arsenate and copper adsorbed at the goethite/water interface. Applied Geochemistry 35: 64-74.  Okita, Y., A. Singhal, and J.J Perraud. 2006. Iron control in the Goro nickel process. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 635-651. Canadian Institute of Mining, Metallurgy and Petroleum.   Onoda, G. Y., J., and P. L. De Bruyn. 1966. Proton adsorption at the ferric oxide-aqueous solution interface. I. kinetic study of adsorption. Surface Science 4 (1): 48-63.  Ostergren, John D., Gordon E. Brown Jr., George A. Parks, and Per Persson. 2000. Inorganic ligand effects on pb(II) sorption to goethite (α-FeOOH) II. sulfate. Journal of Colloid and Interface Science 225 (2): 483-93.  Palmer, C. M., and G. D. Johnson. 2005. The activox process: Growing significance in the nickel industry. Jom 57 (7): 40-7.  Parfitt, R. L., and R. S. Smart. 1977. Infrared spectra from binuclear bridging complexes of sulfate adsorbed on goethite (α-FeOOH). Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases 73 (5): 796-802.  Parfitt, Roger L., and Roger St C. Smart. 1978. The mechanism of sulfate adsorption on iron oxides. Soil Science Society of America Journal 42 (1): 48-50.  Peacock, Caroline L., and David M. Sherman. 2004. Copper(II) sorption onto goethite, hematite and lepidocrocite: A surface complexation model based on abinitio molecular geometries and EXAFS spectroscopy. Geochimica Et Cosmochimica Acta 68 (12) (6/15): 2623-37.    130 Peak, Derek, Evert J. Elzinga, and Donald L. Sparks. 2001. Understanding sulfate adsorption mechanisms on iron(III) oxides and hydroxides: Results from ATR-FTIR spectroscopy. In Heavy Metals Release in Soils: 167-190. Lewis Publishers. Peak, Derek, Robert G. Ford, and Donald L. Sparks. 1999. An in situ ATR-FTIR investigation of sulfate bonding mechanisms on goethite. Journal of Colloid and Interface Science 218 (1): 289-99.  Pouchou, J. L., and F. Pichoir. 1985. PAP f(rZ) procedure for improved quantitative microanalysis. Microbeam Analysis: 104-106.  Powers, Dana A., George R. Rossman, Harvey J. Schugar, and Harry B. Gray. 1975. Magnetic behavior and infrared spectra of jarosite, basic iron sulfate, and their chromate analogs. Journal of Solid State Chemistry 13 (1-2): 1-13.  Rancourt, Denis G., Danielle Fortin, Thomas Pichler, Pierre-Jean Thibault, Gilles Lamarche, Richard V. Morris, and Patrick H. J. Mercier. 2001. Mineralogy of a natural as-rich hydrous ferric oxide coprecipitate formed by mixing of hydrothermal fluid and seawater: Implications regarding surface complexation and color banding in ferrihydrite deposits. American Mineralogist 86 (7-8): 834-51.  Raudsepp, M. 2015. Personal communication Reid, M., and V. G. Papangelakis. 2006. New data on hematite solubility in sulphuric acid solutions from 130 to 270°C. Iron Control Technologies, 3rd Proc.Int.Symp.Iron Control Hydrometallurgy: 673-686. Canadian Institute of Mining, Metallurgy and Petroleum.   Richmond, William R., Mitch Loan, Jonathon Morton, and Gordon M. Parkinson. 2004. Arsenic removal from aqueous solution via ferrihydrite crystallization control. Environmental Science and Technology 38 (8): 2368-72.  Richmond, William R., Mitch Loan, Mike Newman, and Gordon M. Parkinson. 2005. Zinc sulfide as a solid phase additive for improving the processing characteristics of ferrihydrite residues. Hydrometallurgy 78 (3–4) (8): 172-9.  Ristić, M., E. De Grave, S. Musić, S. Popović, and Z. Orehovec. 2007. Transformation of low crystalline ferrihydrite to α-Fe2O3 in the solid state. Journal of Molecular Structure; MOLECULAR SPECTROSCOPY AND MOLECULAR STRUCTURE 2006 A Collection of Papers Presented at the XXVIIIth European Congress on Molecular Spectroscopy, Istanbul, Turkey - September 3-8, 2006 (834-836): 454-60.  Riveros, P. A., and J. E. Dutrizac. 1997. The precipitation of hematite from ferric chloride media. Hydrometallurgy 46 (1-2): 85-104.    131 Rodda, Darren P., Bruce B. Johnson, and John D. Wells. 1993. The effect of temperature and pH on the adsorption of copper(II), lead(II), and zinc(II) onto goethite. Journal of Colloid and Interface Science 161 (1): 57-62.  Roonasi, Payman, and Allan Holmgren. 2009. An ATR-FTIR study of sulphate sorption on magnetite; rate of adsorption, surface speciation, and effect of calcium ions. Journal of Colloid and Interface Science 333 (1): 27-32.  Rose, Arthur W., and Gino Bianchi-Mosquera. 1993. Adsorption of copper, lead, zinc, cobalt, nickel, and silver on goethite and hematite: A control of metal mobilization from red beds into stratiform copper deposits. Economic Geology and the Bulletin of the Society of Economic Geologists 88 (5): 1226-36.  Ruiz, M. C., J. Zapata, and R. Padilla. 2007. Effect of variables on the quality of hematite precipitated from sulfate solutions. Hydrometallurgy 89 (1–2) (9): 32-9.  Sahu, S. K., and E. Asselin. 2011. Characterization of residue generated during medium temperature leaching of chalcopyrite concentrate under CESL conditions. Hydrometallurgy 110: 107-14.  Salomon de Friedberg, H. 2015. Personal communication Sasaki, Kinichi, Kenji Ootsuka, and Kazuteru Tozawa. 1994. Hydrometallurgical studies on hydrolysis of ferric sulfate solutions at elevated temperatures. III. the effect of addition of magnesium sulfate on hydrolysis of ferric sulfate solutions at elevated temperatures. Shigen to Sozai 110 (8): 643-52.  Schwertmann, U., and E. Murad. 1983. Effect of pH on the formation of goethite and hematite from ferrihydrite. Clays and Clay Minerals 31 (4): 277-84.  Schwertmann, U., D. G. Schulze, and E. Murad. 1982. Identification of ferrihydrite in soils by dissolution kinetics, differential x-ray diffraction, and moessbauer spectroscopy. Soil Science Society of America Journal 46 (4): 869-75.  Schwertmann, Udo, Josef Friedl, and Helge Stanjek. 1999. From Fe(III) ions to ferrihydrite and then to hematite. Journal of Colloid and Interface Science 209 (1): 215-23.  Singh, Balwant, D. M. Sherman, R. J. Gilkes, M. Wells, and J. F. W. Mosselmans. 2000. Structural chemistry of Fe, Mn, and Ni in synthetic hematites as determined by extended X-ray absorption fine structure spectroscopy. Clays and Clay Minerals 48 (5): 521-7.  Sugimoto, Tadao, and Yinsheng Wang. 1998. Mechanism of the shape and structure control of monodispersed α-Fe2O3 particles by sulfate ions. Journal of Colloid and Interface Science 207 (1): 137-49.    132 Sun, Zhong-Xi, Fen-Wei Su, Willis Forsling, and Per-Olof Samskog. 1998. Surface characteristics of magnetite in aqueous suspension. Journal of Colloid and Interface Science 197 (1) (1/1): 151-9.  Towe, Kenneth M., and William F. Bradley. 1967. Mineralogical constitution of colloidal "hydrous ferric oxides". Journal of Colloid and Interface Science 24 (3): 384-92.  Tozawa, K., and K. Sasaki. 1986. Effect of coexisting sulfates on precipitation of ferric oxide from ferric sulfate solutions at elevated temperatures. Iron Control Hydrometall., Int.Symp. 1986: 454-76. Turner, L. J., and J. R. Kramer. 1991. Sulfate ion binding on goethite and hematite. Soil Science 152 (3): 226-30.  Umetsu, Yoshiyuki, Kazuteru Tozawa, and Kinichi Sasaki. 1977. The hydrolysis of ferric sulfate solutions at elevated temperatures. Canadian Metallurgical Quarterly 16 (1-4): 111-17.  Vallina, B., J. Rodriguez-Blanco, A. P. Brown, L. G. Benning, and J. A. Blanco. 2014. Enhanced magnetic coercivity of α-Fe2O3 obtained from carbonated 2-line ferrihydrite. Journal of Nanoparticle Research 16 (3): 2322/1, 2322/13.  Van, der Woude, and P. L. De Bruyn. 1984. Formation of colloidal dispersions from supersaturated iron(III) nitrate solutions. V. synthesis of monodisperse goethite sols. Colloids and Surfaces 12 (1-2): 179-88.  ———. 1983. Formation of colloidal dispersions from supersaturated iron(III) nitrate solutions. I. precipitation of amorphous iron hydroxide. Colloids and Surfaces 8 (1): 55-78.  Van, der Woude, P. L. De Bruyn, and J. Pieters. 1984. Formation of colloidal dispersions from supersaturated iron(III) nitrate solutions. III. development of goethite at room temperature. Colloids and Surfaces 9 (2): 173-88.  Weert, Gus Van, and Vincent Kok. 1995. Impurity levels in hematite produced by autoclave hydrolysis of ferric nitrate. Sep. Processes Proc. Symp 1995: 225-233. Wells, M. A., R. J. Gilkes, and R. W. Fitzpatrick. 2001. Properties and acid dissolution of metal-substituted hematites. Clays and Clay Minerals 49 (1): 60-72.  Wijnja, Hotze, and Cristian P. Schulthess. 2000. Vibrational spectroscopy study of selenate and sulfate adsorption mechanisms on fe and al (hydr)oxide surfaces. Journal of Colloid and Interface Science 229 (1): 286-97.  Yang, Yu-huan, Hao Chen, and Gang Pan. 2007. Particle concentration effect in adsorption/desorption of zn(II) on anatase type nano TiO2. Journal of Environmental Sciences (Beijing, China) 19 (12): 1442-5.    133 Zhao, Jianmin, Frank E. Huggins, Zhen Feng, and Gerald P. Huffman. 1994. Ferrihydrite: Surface structure and its effects on phase transformation. Clays and Clay Minerals 42 (6): 737-46.     134 Appendices  Appendix A    RESIDUE ASSAYS  S1 S1 R1 S3 S3 R1 S4 S4 R1 Fe (ppm) 233000 167000 385000 380000 214800 237400 Fe (%) 23.3 16.7 38.5 38.0 21.5 23.7 Ni (ppm) <1 <1 8281 7556 334 100 Cu (ppm) 10750 8282 5843 4911 32280 17580 Ca (ppm) 8732 673 2277 314 6206 1915 S (%) 29 42.8 31.5 32.3 21.6 22.5 S (ppm) 290000 428000 315000 323000 216000 224600 Mass loss (%)   35.5   3.9     Ni recovery (%)   -   12.3     Cu recovery (%)   50.3   19.2     Fe recovery (%)   53.8   5.1     Fe% mass loss   23.0   29.1     %Mass loss to Cu   1.1   0.5     %Mass loss to Ni       0.6        assay values   calculated values   backcalculated data   135 Appendix B     assay values   calculated values   backcalculated data B.1 Preliminary experiments data Preliminary Experiments S1 water 24 hours In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.3675 3.9741 266.8 9.0% - Assay Fe (%) 23.3 25.4% 29 1.02 1.01 0.01 0.76% 0.0% Assay Ni (ppm) <1   -  0.00 - - Assay Cu (ppm) 10750 7853 59 0.05 0.03 0.02 33.5% 0.0% Assay Ca (ppm) 8732 -877 156 0.04 0.00 0.04 109.1% 0.0% Assay S (%) 29 30.3% 239 1.27 1.20 0.06 5.0% 0.00 S1 water 24 hours, HAHC In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 3.988 3.3261 244.9 16.6% - Assay Fe (%) 18.0% 21.2 1523 0.72 0.71 0.01 1.7% 0.0% Assay Ni (ppm)  <1 <1      Assay Cu (ppm) 9156 8033 40 0.04 0.03 0.01 26.8% 0.0% Assay Ca (ppm) 569 535 2 0.00 0.00 0.00 21.6% 0.0% Assay S (%) 29.7 34.9 102 1.19 1.16 0.02 2.1% 0.0% S1R1 water 24 hours In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 3.9927 3.7126 209.3 7.0%  Assay Fe (%) 16.7 17.68% 50 0.67 0.66 0.01 1.6% 0.0% Assay Ni (ppm) <1  <1      Assay Cu (ppm) 8282 7610 23 0.03 0.03 0.00 14.6% 0.0% Assay Ca (ppm) 673 -291 18 0.00 0.00 0.00 140.2% 0.0% Assay S (%) 42.8 45.65% 67 1.71 1.69 0.01 0.8% 0.0% S1 R1 water 24 hours, HAHC In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)     3.6245 3.3375 237.8 7.9%   Assay Fe (%) 18.6% 16.2 560 0.67 0.54 0.13 19.8% 0.0% Assay Ni (ppm) 0 <1 <1 0.00    - Assay Cu (ppm) 7544 7552 9 0.03 0.03 0.002 7.8% 0.0%   136  In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Assay Ca (ppm) 561 609 <1 0.00 0.00  0.0% 0.0% Assay S (%) 43.0% 46.6 20 1.56 1.56 0.005 0.3% 0.0% S3 water 24 hours In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)   - - 4.0505 3.8637 263.8 4.6% - Assay Fe (%) 38.5 40.3% 3 1.56 1.56 0.00 0.1% 0.0% Assay Ni (ppm) 8281 6564.8 31 0.03 0.03 0.01 24.4% 0.0% Assay Cu (ppm) 5843 5715.8 6 0.02 0.02 0.00 6.7% 0.0% Assay Ca (ppm) 2277 -70.9 36 0.01 0.00 0.01 103.0% 0.0% Assay S (%) 31.5 32.5% 82 1.28 1.25 0.02 1.7% 0.0% S3 water 24 hours, HAHC In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 3.7414 3.7027 318.1 1.0% - Assay Fe (%) 40.6 40.6 49 1.52 1.50 0.02 1.0% 0.0% Assay Ni (ppm) 6490 6558 <1 0.02 0.02 0.00 0.0% 0.0% Assay Cu (ppm) 5509 5169 4 0.02 0.02 0.00 6.2% 0.0% Assay Ca (ppm) 553 559 <1 0.00 0.00 0.00 0.0% 0.0% Assay S (%) 31.9 32.2 7 1.19 1.19 0.00 0.2% 0.0%          S3 R1 water 24 hours In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.1615 3.9909 244.7 4.1% - Assay Fe (%) 38.0 39.6% 12 1.58 1.58 0.00 0.2% 0.0% Assay Ni (ppm) 7556 6268 34 0.03 0.03 0.01 20.5% 6.0% Assay Cu (ppm) 4911 3831 28 0.02 0.02 0.01 25.2% 8.3% Assay Ca (ppm) 314 184.6 3 0.00 0.00 0.00 43.6% 12.6% Assay S (%) 32.3 33.2% 35 1.34 1.32 0.01 1.5% -0.8% S3 R1 water 24 hours, HAHC In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 3.8544 3.8677 266.5 -0.3% - Assay Fe (%) 41.2% 40.6 67 1.59 1.57 0.02 1.1% 0.0% Assay Ni (ppm) 6268 6246 <1 0.02 0.02  0.0% 0.0% Assay Cu (ppm) 3831 3680 2 0.01 0.01 0.00 3.6% 0.0% Assay Ca (ppm) 185 184 <1 0.00 0.00  0.0% 0.0% Assay S (%) 33.2% 33 10 1.28 1.28 0.00 0.2% 0.0%   137           Water ageing  S1 aged in water, 24 hours In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 40.20 38.28 394.10 4.8% - Assay Fe (%) 23.3 24.7 154.0 9.37 9.46 0.06 -0.9% 1.6% Assay Ni (ppm) <1 <20 3.0 - - - - - Assay Cu (ppm) 10750.0 8780.0 307.0 0.43 0.34 0.12 22.2% 5.8% Assay Ca (ppm) 8732.0 3655.0 536.0 0.35 0.14 0.21 60.1% 0.0% Assay S (%) 29.0 28.4 788.0 11.66 10.87 0.31 6.7% 4.1% S3 aged in water, 24 hours In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 40.26 38.42 680 4.6% - Assay Fe (%) 38.5 40.5 13 15.500 15.560 0.009 -0.4% 0.4% Assay Ni (ppm) 8281 6767 113 0.33 0.26 0.08 22.0% 1.0% Assay Cu (ppm) 5843 5761 34 0.24 0.22 0.02 5.9% 3.9% Assay Ca (ppm) 2277 117 139 0.09 0.00 0.09 95.1% 8.0% Assay S (%) 31.5 31.8 287 12.68 12.22 0.20 3.7% -2.1% S1 in water at pH 2 at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.06 7.58 219.20 6.0% - Assay Fe (%) 23.3 23.0 45 1.88 1.74 0.010 0.5% -6.6% Assay Ni (ppm) <1 0 0 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 9388 88 0.09 0.07 0.02 24% 4.4% Assay Ca (ppm) 8732 589 322 0.07 0.00 0.07 94% 6.6% Assay S (%) 29 28.9 398 2.34 2.19 0.09 4% -2.5% S1 in water at pH 2 at 95°C for 1 hour #2 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.07 3.83 288.1 5.9% - Assay Fe (%) 23.3 27.2 23 0.95 1.04 0.01 -9.9% 10.6% Assay Ni (ppm) <1 0 0 - - - - - Assay Cu (ppm) 10750 10230 91 0.04 0.04 0.03 10.4% 49.5% Assay Ca (ppm) 8732 1167 169 0.04 0.00 0.05 87.4% 49.6% Assay S (%) 29 28.7 244 1.18 1.10 0.07 6.9% -0.9%               138  S3 in water at pH 2 at 95°C for 1 hour #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.05 3.87 240.3 4.4% - Assay Fe (%) 38.5 36.0 15 1.56 1.39 0.0036 10.6% -10.4% Assay Ni (ppm) 8281 8010 13 0.03 0.03 0.00 7.6% 1.7% Assay Cu (ppm) 5843 5583 10 0.02 0.02 0.00 8.7% 1.5% Assay Ca (ppm) 2277 71 43 0.01 0.00 0.01 97.0% 15.0% Assay S (%) 31.5 32.0 105 1.28 1.24 0.03 2.9% -0.9% S3 in water at pH 2 at 95°C for 1 hour #2 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.07 3.83 288.1 5.9% - Assay Fe (%) 38.5 42.5 20 1.57 1.63 0.01 -3.9% 4.2% Assay Ni (ppm) 8281 8391 30 0.03 0.03 0.01 3.7% 22.0% Assay Cu (ppm) 5843 5967 21 0.02 0.02 0.01 2.9% 22.5% Assay Ca (ppm) 2277 525 36 0.01 0.00 0.01 78.1% 33.8% Assay S (%) 31.5 32.4 122 1.28 1.25 0.04 2.2% 0.5%     139 B.2 Fe displacements experiment data 1 hour Fe displacements         S3 in 1 g/L Fe at pH 2 at 95°C for 1 hour #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.16 8.01 438.42 1.9% - Assay Fe (%) 38.5 39.1 397 3.14 3.13 -0.02 -0.5% 0.9% Assay Ni (ppm) 8281 7755 13 0.07 0.06 0.01 8.4% 0.3% Assay Cu (ppm) 5843 5369 6 0.05 0.04 0.00 5.5% 4.3% Assay Ca (ppm) 2277 253 39 0.02 0.00 0.02 92.0% 2.9% Assay S (%) 31.5 31.7 376 2.57 2.54 0.16 6.4% 5.2% S3 in 1 g/L Fe at pH 2 at 95°C for 1 hour #2 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.09 7.94 487.93 1.8% - Assay Fe (%) 38.5 39.4 372 3.11 3.13 -0.01 -0.3% 0.1% Assay Ni (ppm) 8281 7787 11 0.07 0.06 0.01 8.0% 0.3% Assay Cu (ppm) 5843 5287 4 0.05 0.04 0.00 4.1% 7.1% Assay Ca (ppm) 2277 339 36 0.02 0.00 0.02 95.4% 10.0% Assay S (%) 31.5 32.0 345 2.55 2.54 0.17 6.6% 6.3% S3 in 0.1 g/L Fe at pH 2 at 95°C for 1 hour #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.09 7.74 410.04 4.3% - Assay Fe (%) 38.5 39.3 50 3.11 3.04 0.00 0.0% 2.3% Assay Ni (ppm) 8281 7818 12 0.07 0.06 0.00 7.3% 2.3% Assay Cu (ppm) 5843 5026 9 0.05 0.04 0.00 7.8% 9.9% Assay Ca (ppm) 2277 310 41 0.02 0.00 0.02 91.3% 4.3% Assay S (%) 31.5 32.0 143 2.55 2.48 0.06 2.3% 0.5%          S3 in 0.1 g/L Fe at pH 2 at 95°C for 1 hour #2 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.03 7.68 286.50 4.4% - Assay Fe (%) 38.5 39.6 72 3.09 3.04 0.00043 0.0% 1.6% Assay Ni (ppm) 8281 7948 16 0.07 0.06 0.005 6.9% 1.3% Assay Cu (ppm) 5843 5132 14 0.05 0.04 0.00 8.5% 7.4% Assay Ca (ppm) 2277 248 58 0.02 0.00 0.02 90.9% 1.3% Assay S (%) 31.5 31.8 210 2.53 2.44 0.06 2.4% 1.1%            140 S1 in 1 g/L Fe at pH 2 at 95°C for 1 hour #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.11 7.79 285.75 4.0% - Assay Fe (%) 23.3 24.9 574 1.89 1.94 -0.03 -1.4% 0.9% Assay Ni (ppm) <1 <1 <1 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8790 77 0.09 0.07 0.02 25.2% 0.8% Assay Ca (ppm) 8732 670 241 0.07 0.01 0.07 97.2% 2.0% Assay S (%) 29 29.0 696 2.35 2.26 0.20 8.5% 2.4% S1 in 1 g/L Fe at pH 2 at 95°C for 1 hour #2 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.05 7.49 322.95 7.0% - Assay Fe (%) 23.3 25.1 530 1.88 1.88 -0.02 -1.1% 0.9% Assay Ni (ppm) <1 <1 <1 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8766 67 0.09 0.07 0.02 25.0% 0.8% Assay Ca (ppm) 8732 559 209 0.07 0.00 0.07 96.0% 2.0% Assay S (%) 29 29.3 607 2.34 2.19 0.20 8.4% 2.4% S1 in 0.1 g/L Fe at pH 2 at 95°C for 1 hour #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.00 7.92 438.20 1.0% - Assay Fe (%) 23.3 24.8 46 1.86 1.96 0.00 0.0% 5.4% Assay Ni (ppm) <1 <1 <1 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8933 47 0.09 0.07 0.02 23.9% 6.2% Assay Ca (ppm) 8732 732 148 0.07 0.01 0.06 92.8% 1.1% Assay S (%) 29 29.7 227 2.32 2.35 0.10 4.3% 5.7% S1 in 0.1 g/L Fe at pH 2 at 95°C for 1 hour #2 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.08 7.38 186.80 8.7% - Assay Fe (%) 23.3 25.1 119 1.88 1.85 0.00 0.1% 1.5% Assay Ni (ppm) <1 <1 <1 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8885 107 0.09 0.07 0.02 23.0% 1.5% Assay Ca (ppm) 8732 731 381 0.07 0.01 0.07 100.9% 8.5% Assay S (%) 29 29.2 570 2.34 2.15 0.11 4.5% 3.5%                 141 24 hour Fe displacements         S1 in 1 g/L Fe at pH 2 at 95°C for 24 hours  In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.16 3.94 209.70 5.3% - Assay Fe (%) 23.3 23.8 743 0.97 0.94 -0.05 -5.4% 8.7% Assay Ni (ppm) <1 0 0 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8114 82 0.04 0.03 0.02 38.5% 9.9% Assay Ca (ppm) 8732 506 177 0.04 0.00 0.04 102.2% 7.7% Assay S (%) 29 27.8 573 1.21 1.10 0.12 10.0% 0.8% S1 in 0.1 g/L Fe at pH 2 at 95°C for 24 hours  In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.53 4.21 279.10 7.1% - Assay Fe (%) 23.3 22.6 68 1.06 0.95 0.0012 0.1% 9.7% Assay Ni (ppm) <1 0 0 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8526 70 0.05 0.04 0.02 40.1% 13.8% Assay Ca (ppm) 8732 338 154 0.04 0.00 0.04 108.7% 12.3% Assay S (%) 29 28.4 364 1.31 1.20 0.10 7.7% 1.3% S3 in 0.1 g/L Fe at pH 2 at 95°C for 24 hours  In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.18 4.07 432.20 2.6% - Assay Fe (%) 38.5 36.6 260 1.61 1.49 -0.10 -6.0% 13.4% Assay Ni (ppm) 8281 6440 20 0.03 0.03 0.01 25.0% 0.7% Assay Cu (ppm) 5843 3542 29 0.02 0.01 0.01 51.3% 10.3% Assay Ca (ppm) 2277 49 31 0.01 0.00 0.01 140.8% 42.9% Assay S (%) 31.5 30.4 271 1.32 1.24 0.12 8.9% 2.9% S3 in 1 g/L Fe at pH 2 at 95°C for 24 hours  In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.07 3.77 304.20 7.4% - Assay Fe (%) 38.5 35.5 20 1.57 1.34 -0.01 -0.7% 15.3% Assay Ni (ppm) 8281 6460 33 0.03 0.02 0.01 29.8% 2.0% Assay Cu (ppm) 5843 5054 15 0.02 0.02 0.00 19.2% 0.7% Assay Ca (ppm) 2277 350 43 0.01 0.00 0.01 141.1% 55.4% Assay S (%) 31.5 31.6 122 1.28 1.19 0.04 2.9% 4.2%     142 B.3 Sequential extraction data HAHC dissolution         S1 HAHC extraction #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.15 3.00 215.50 27.7% - Assay Fe (%) 23.3 16.8 1911 0.97 0.50 0.41 47.9% -5.3% Assay Ni (ppm) <1 0 0 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8998 96 0.04 0.03 0.02 39.5% 6.9% Assay Ca (ppm) 8732 578 171 0.04 0.00 0.04 95.2% 6.5% Assay S (%) 29 38.8 268 1.20 1.16 0.06 3.3% 1.5% S1 HAHC extraction #2 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.11 2.90 242.00 29.4% - Assay Fe (%) 23.3 16.2 1781 0.96 0.49 0.43 49.1% -4.1% Assay Ni (ppm) <1 0 0 0.00 0.00 0.00 - - Assay Cu (ppm) 10750 8690 107 0.04 0.03 0.03 43.0% 15.6% Assay Ca (ppm) 8732 568 146 0.04 0.00 0.04 95.4% 3.0% Assay S (%) 29 39.5 245 1.19 1.15 0.06 3.9% 1.1% S3 HAHC extraction #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.12 4.02 337.40 2.4% - Assay Fe (%) 38.5 34.6 127 1.59 1.39 0.04 12.3% -9.6% Assay Ni (ppm) 8281 7900 5 0.03 0.03 0.00 6.9% -2.0% Assay Cu (ppm) 5843 5409 16 0.02 0.02 0.01 9.7% 12.8% Assay Ca (ppm) 2277 56 34 0.01 0.00 0.01 97.6% 24.7% Assay S (%) 31.5 32.3 38 1.30 1.30 0.01 -0.1% 1.0% Aged then HAHC         Sample 50 (S1) treated by HAHC method In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.01 6.13 447.90 23.5% - Assay Fe (%) 24.7 20.7 1585 1.98 1.27 0.71 35.9% 0.0% Assay Ni (ppm) <20 0.0 <1 0.00 0.00 0.00 - - Assay Cu (ppm) 8780 8258 44 0.07 0.05 0.02 28.0% 0.0% Assay Ca (ppm) 3655 -46.5 66 0.03 0.00 0.03 101.0% 0.0% Assay S (%) 28.4 35.9 164 2.27 2.20 0.07 3.2% 0.0%            143 Sample 51 (S3) treated by HAHC method In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.05 7.86 583.10 2.4% - Assay Fe (%) 40.5 41.2 36 3.26 3.24 0.02 0.6% 0.0% Assay Ni (ppm) 6767 6930.6 0 0.05 0.05 0.00 0.0% 0.0% Assay Cu (ppm) 5761 5529 5 0.05 0.04 0.00 6.3% 0.0% Assay Ca (ppm) 117 -325.3 6 0.00 0.00 0.00 371.5% 0.0% Assay S (%) 31.8 32.5 5 2.56 2.56 0.00 0.1% 0.0%     144 B.4 Aged then Fe displacement experiment data Aged then Fe Displacements Sample 50 (aged S1) in 1 g/L Fe solution at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.01 7.85 393.05 2.0% - Assay Fe (%) 24.7 25.3 492 1.98 1.99 0.19 -1.8% -1.4% Assay Ni (ppm) <20 <20 <1 0.00 0.00 0.00 - - Assay Cu (ppm) 8780 8400 16 0.07 0.07 0.01 8.9% 2.7% Assay Ca (ppm) 3655 476.0 82 0.03 0.00 0.03 110.1% 22.9% Assay S (%) 28.4 28.7 461 2.27 2.25 0.18 8.0% 7.0% Sample 51 (aged S3) in 1 g/L Fe at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.02 7.91 296.60 1.4% - Assay Fe (%) 40.5 40.8 595 3.25 3.23 0.18 -1.6% -2.3% Assay Ni (ppm) 6767 6794 6 0.05 0.05 0.00 3.3% 2.3% Assay Cu (ppm) 5761 5814 5 0.05 0.05 0.00 3.2% 2.7% Assay Ca (ppm) 117 230 6 0.00 0.00 0.00 189.7% 283.5% Assay S (%) 31.8 31.9 481 2.55 2.52 0.14 5.6% 4.5% Sample 50 (aged S1) in 0.1 g/L Fe at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 8.05 7.72 215.20 4.1% - Assay Fe (%) 24.7 23.3 124 1.99 1.80 0.05 2.7% -7.7% Assay Ni (ppm) <20 0 0 0.00 0.00 0.00 - - Assay Cu (ppm) 8780 8878 21 0.07 0.07 0.01 13.0% 10.0% Assay Ca (ppm) 3655 419 173 0.03 0.00 0.08 257.7% 168.6% Assay S (%) 28.4 28.9 267 2.29 2.23 0.12 5.1% 2.7% Sample 51 (aged S3) in 0.1 g/L Fe at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.12 3.94 277.20 4.4% - Assay Fe (%) 40.5 35.8 89 1.67 1.47 0.02 1.5% -11.2% Assay Ni (ppm) 6767 6330 8 0.03 0.03 0.00 8.0% 1.5% Assay Cu (ppm) 5761 5676 4 0.02 0.02 0.00 4.7% 3.2% Assay Ca (ppm) 117 84 3 0.00 0.00 0.00 172.5% 144.3% Assay S (%) 31.8 32.0 104 1.31 1.32 0.03 2.2% 2.8%              145 S1R1 in 0.1 g/L Fe for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.12 3.94 277.20 4.4% - Assay Fe (%) 16.7 15.9 127 0.64 0.63 0.06 8.7% -25.9% Assay Ni (ppm) <1 0 0 0.00 0.00 0.00 - - Assay Cu (ppm) 8282 7751 22 0.03 0.03 0.01 30.4% 26.7% Assay Ca (ppm) 673 469 4 0.00 0.00 0.00 68.0% 39.7% Assay S (%) 42.8 42.1 126 1.64 1.66 0.06 3.4% 4.6% S3R1 in 0.1 g/L Fe for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.02 3.85 226.00 4.2% - Assay Fe (%) 38.0 35.2 90 1.46 1.36 0.04 2.7% -18.9% Assay Ni (ppm) 7556.0 7120 18 0.03 0.03 0.01 27.1% 21.4% Assay Cu (ppm) 4911.0 4154 19 0.02 0.02 0.01 44.1% 28.6% Assay Ca (ppm) 314.0 59 2 0.00 0.00 0.00 72.5% -8.7% Assay S (%) 32.3 32.9 52 1.24 1.27 0.02 1.8% 3.7% Aged then HAHC then Fe         Sample 56 (S1) treated with 1 g/L Fe for 1 hour at 95°C In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 4.87 4.79 356.20 1.6% - Assay Fe (%) 20.7 20.3 550 1.01 0.97 0.20 -3.3% 5.6% Assay Ni (ppm) 0.0 <20 <1 0.00 0.00 0.00 - - Assay Cu (ppm) 8257.8 7558 9 0.04 0.04 0.00 8.0% 2.0% Assay Ca (ppm) -46.5 421 4 0.02 0.00 0.00 6.1% 85.3% Assay S (%) 35.9 36.6 349 1.75 1.75 0.12 7.1% 7.4% Sample 57 (S3) treated with 1 g/L Fe for 1 hour at 95°C In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 7.65 7.63 376.67 0.3% - Assay Fe (%) 41.2 40.3 454 3.15 3.07 0.17 -1.2% 3.4% Assay Ni (ppm) 6930.6 6555 7 0.05 0.05 0.00 5.0% 0.7% Assay Cu (ppm) 5529.3 5476 <1 0.04 0.04 0.00 0.0% 1.2% Assay Ca (ppm) -325.3 60 3 0.00 0.00 0.00 -45.4% -163.8% Assay S (%) 32.5 32.6 267 2.49 2.49 0.10 4.0% 4.0% !!  146 B.5 Repeated ageing experiments data Repeated ageing         S1 in water at 95°C for 1 hour (6% solids) In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 12.12 11.38 292.8 6.1%  Assay Fe (%,ppm) 23.3 26.4 24 2.82 3.00 0.01 0.2% 6.6% Assay Ni (%,ppm) <1 < 0.01 < 1 0.00 0.00 0.00 0.0% 0.0% Assay Cu (%,ppm) 10750 1.04 105 0.13 0.12 0.03 23.6% 14.4% Assay Ca (ppm) 8732 1320 414 0.11 0.02 0.12 114.5% 28.7% Assay S (%,ppm) 29 28.9 472 3.51 3.29 0.14 3.9% 2.5% S1 (sample 90) in water at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g) - - - 10.65 10.36 257.2 2.7%  Assay Fe (%,ppm) 26.4 26.0 31 2.81 2.69 0.01 0.3% 3.9% Assay Ni (%,ppm) 0 < 0.01 < 1 0.00 0.00 0.00 0.0% 0.0% Assay Cu (%,ppm) 1.04 0.9 48 0.11 0.10 0.01 11.1% 0.9% Assay Ca (ppm) 1320 504.0 16 0.01 0.01 0.00 29.3% 33.6% Assay S (%,ppm) 28.9 28.8 114 3.08 2.98 0.03 1.0% 2.1% S1 (sample 94) in water at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)    9.51 9.29 289.4 2.3%  Assay Fe (%,ppm) 26.0 26.4 30 2.47 2.45 0.01 0.4% 0.5% Assay Ni (%,ppm) 0.00 < 0.01 < 1 0.00 0.00 0.00 0.0% 0.0% Assay Cu (%,ppm) 0.94 0.91 14 0.09 0.08 0.00 4.5% 0.9% Assay Ca (ppm) 504 585 4 0.00 0.01 0.00 24.2% 37.5% Assay S (%,ppm) 28.8 28.4 117 2.74 2.64 0.03 1.2% 2.4% S1 (sample 98) in water at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)    8.69 8.42 320.4 3.1%  Assay Fe (%,ppm) 26.4 27.1 26 2.29 2.29 0.01 0.4% 0.0% Assay Ni (%,ppm) 0 < 0.01 < 1 0.00 0.00 0.00 0.0% 0.0% Assay Cu (%,ppm) 0.91 0.9 10 0.08 0.08 0.00 4.1% 0.0% Assay Ca (ppm) 585 -8.7 38 0.01 -0.01 0.01 239.5% 0.0% Assay S (%,ppm) 28.4 28.9 113 2.47 2.43 0.04 1.5% 0.0%            147 S1 (sample 102) in water at 95°C for 1 hour In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)    8.25 7.99 257 3.2%  Assay Fe (%,ppm) 27.1 26.7 31 2.24 2.13 0.01 0.4% 4.4% Assay Ni (%,ppm) 0.0 < 0.01 < 1 0.00 0.00 0.00 0.0% 0.0% Assay Cu (%,ppm) 0.90 0.86 28 0.07 0.07 0.01 9.7% 2.1% Assay Ca (ppm) 585 583 14 0.00 0.00 0.00 74.6% 71.1% Assay S (%,ppm) 28.9 28.3 145 2.38 2.26 0.04 1.6% 3.5% S1 (sample 106) HAHC #1 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)    3.55 3.04 235.6 14.4%  Assay Fe (%,ppm) 26.7 19.9 1030 0.95 0.60 0.24 25.6% 10.6% Assay Ni (%,ppm) 0.0 < 0.01 < 1 0.00 0.00 0.00 0.0% 0.0% Assay Cu (%,ppm) 0.86 0.76 35 0.03 0.02 0.01 27.0% 2.7% Assay Ca (ppm) 583 652 3 0.00 0.00 0.00 34.2% 29.9% Assay S (%,ppm) 28.3 31.4 78 1.00 0.95 0.02 1.8% 3.2%  S1 (sample 116) HAHC #2, composite of 110 and 111 In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)    3.52 2.99 222.7 15.1%  Assay Fe (%,ppm) 20.8 15 1090 0.73 0.45 0.24 33.2% 5.6% Assay Ni (%,ppm) 0.0 < 0.01 < 1 0.00   0.0%  Assay Cu (%,ppm) 0.76 0.73 22 0.03 0.02 0.00 18.3% 0.1% Assay Ca (ppm) 723 741 3 0.00 0.00 0.00 26.3% 13.3% Assay S (%,ppm) 30.5 37.8 71 1.07 1.13 0.02 1.5% 6.7% S1 in water at 95°C for 1 hour (10% solids) In  solids Out solids  Out  solution Mass In solids (g) Mass Out Solids (g) Mass out  solution (g) %extraction %error Sample mass (g)    20.51 19.17 289.20 6.5%  Assay Fe (%,ppm) 23.3 25.8 39 4.78 4.95 0.01 0.2% 3.7% Assay Ni (%,ppm) 0.0 < 0.01 < 1 0.00 0.00 0.00 0.0% 0.0% Assay Cu (%,ppm) 1.08 1.01 169 0.22 0.19 0.05 22.2% 10.0% Assay Ca (ppm) 8732 2640 567 0.18 0.05 0.16 91.6% 19.8% Assay S (%,ppm) 29.0 28.8 669 5.95 5.52 0.19 3.3% 3.9%      148 Appendix C   C.1 Repeated ageing calculations   REPEATED AGEING      1st stage %Solids  2% w/w solids 6% w/w solids 10% w/w solids 15% w/w solids Average  %Cu extraction 23.9% 20.6% 20.2% 21.2% 21.5%  %Cu grade 0.88 1.04 1.01 0.97 0.98  %Fe extraction 0.3% 0.2% 0.2% 0.3% 0.3%  %Fe grade 25.0 26.4 25.8 25.6 25.7  %mass loss 7.2% 6.1% 6.5% 5.1% 6.2%  %S extraction 4.3% 4.0% 3.4% 2.8% 3.6%  %grade 29.90 28.9 28.8 28.8 29.10        2nd stage %Cu extraction 9.5% 11.1% 4.7% 4.8% 7.5%  %Cu grade 0.95 0.94 0.93 0.93 0.94  %Fe extraction 0.5% 0.3% 0.1% 0.2% 0.3%  %Fe grade 26.9 26 25.9 26.7 26.4  %mass loss 8.8% 2.7% 3.0% 3.6% 4.5%        3rd stage %Cu extraction 5.5% 4.5% 4.3% 3.9% 4.6%  %Cu grade 0.93 0.91 0.9 0.9 0.91  %Fe extraction 0.5% 0.4% 0.2% 0.3% 0.3%  %Fe grade 26.9 26.4 26.7 26.2 26.55  %mass loss 3.1% 2.3% 3.0% 1.2% 2.4%        4th stage %Cu extraction 18.5% 4.1% 3.8% 2.7% 7.3%  %Cu grade 0.87 0.90 0.88 0.88 0.88  %Fe extraction 0.5% 0.4% 0.4% 0.1% 0.3%  %Fe grade 26.4 27.1 27.2 26.2 26.7  %mass loss 8.1% 3.1% 2.0% 2.7% 4.0%        5th stage %Cu extraction 4.3% 9.7% 3.0% 3.7% 5.1%  %Cu grade 0.89 0.86 0.87 0.86 0.87  %Fe extraction 0.4% 0.4% 0.3% 0.5% 0.4%  %Fe grade 26.0 26.7 26.4 25.3 26.1   149   2% w/w solids 6% w/w solids 10% w/w solids 15% w/w solids Average    %mass loss 6.4% 3.2% 2.2% 1.3% 3.3%  HAHC stage 1 %Cu extraction  27.0% 21.8% 21.1% 23.3%  %Cu grade  0.76 0.77 0.75 0.76  %Fe extraction  25.6% 28.2% 28.5% 27.4%  %Fe grade  19.9 20.2 20.5 20.2  %mass loss  14.4% 15.0% 16.4% 15.2%        HAHC stage 2 %Cu extraction  18.3% 21.8% 22.2% 20.8%  %Cu grade  0.73 0.74 0.72 0.73  %Fe extraction  33.2% 38.0% 35.8% 35.6%  %Fe grade  15.0 15.2 16.2 15.5  %mass loss  15.1% 19.0% 18.7% 17.6%            150  Total recovery calculations 2% w/w solids 6% w/w solids 10% w/w solids 15% w/w solids Average 1st stage Mass In  12.12 20.51 30.20   Cu In  0.149 0.242 0.353   Grade  1.23% 1.18% 1.17% 1.19%  Mass out  11.38 19.17 28.66   Cu out  0.118 0.194 0.278   Grade  1.04% 1.01% 0.97% 1.01%  Stage recovery  20.6% 20.2% 21.2% 20.7%  Total Recovery  20.6% 20.2% 21.2% 20.7%        2nd stage Mass In  11.38 19.17 28.66   Cu In  0.118 0.194 0.278   Grade  1.04% 1.01% 0.97% 1.01%  Mass out  11.07 18.60 27.64   Cu out  0.104 0.173 0.257   Grade  0.94% 0.93% 0.93% 0.93%  Stage recovery  11.1% 4.7% 4.77% 6.9%  Total Recovery  30.2% 28.7% 27.2% 28.7%        3rd stage Mass In  11.07 18.60 27.64   Cu In  0.104 0.173 0.257   Grade  0.94% 0.93% 0.93% 0.93%  Mass out  10.81 18.22 27.30   Cu out  0.098 0.164 0.246   Grade  0.91% 0.90% 0.90% 0.90%  Stage recovery  4.5% 4.3% 3.88% 4.2%  Total Recovery  34.0% 32.4% 30.4% 32.3%        4th stage Mass In  10.81 18.22 27.30   Cu In  0.098 0.164 0.246   Grade  0.91% 0.90% 0.90% 0.90%  Mass out  10.48 17.82 26.56   Cu out  0.094 0.158 0.234   Grade  0.90% 0.88% 0.88% 0.89%  Stage recovery  4.1% 3.8% 2.72% 3.5%  Total Recovery  36.7% 35.0% 33.8% 35.1%   151  Total recovery calculations 2% w/w solids 6% w/w solids 10% w/w solids 15% w/w solids Average 5th stage Mass In  10.48 17.82 26.56   Cu In  0.094 0.158 0.234   Grade  0.90% 0.88% 0.88% 0.89%  Mass out  10.15 17.43 26.21   Cu out  0.087 0.152 0.225   Grade  0.86% 0.87% 0.86% 0.86%  Stage recovery  9.7% 3.0% 3.70% 5.4%  Total Recovery  41.5% 37.5% 36.2% 38.4%        HAHC 1 Mass In  10.15 17.43 26.21   Cu In  0.087 0.152 0.225   Grade  0.86% 0.87% 0.86% 0.86%  Mass out  8.69 14.82 21.90   Cu out  0.066 0.114 0.164   Grade  0.76% 0.77% 0.75% 0.76%  Stage recovery  27.0% 21.8% 21.09% 23.3%  Total Recovery  55.7% 52.9% 53.5% 54.0%        HAHC 1 Mass In  8.69 14.82 21.90   Cu In  0.066 0.114 0.164   Grade  0.76% 0.77% 0.75% 0.76%  Mass out  7.38 12.00 17.81   Cu out  0.054 0.089 0.129   Grade  0.73% 0.74% 0.72% 0.73%  Stage recovery  18.3% 21.8% 22.22% 20.8%  Total Recovery  63.9% 63.4% 63.6% 63.6%       152  C.2 Residue mass balances CESL RESIDUE MASS BALANCE Description mass in (g) mass out (g) mass loss Cu in (g) grade (%) Cu out (g) grade (%) %solids recovery  %liquids recovery S1 aged for 24 hours 100.00 93.11 6.9% 1.100 1.100% 0.788 0.846% 71.7% 28.3% Aged then HAHC 93.11 71.26 23.5% 0.788 0.846% 0.567 0.796% 72.0% 28.0% Aged then HAHC then Fe  71.26 70.08 1.6% 0.567 0.796% 0.530 0.756% 91.9% 8.1%           Aged in 1 g/L Fe for 1 hour 100.00 94.50 5.5% 1.100 1.100% 0.813 0.860% 73.9% 26.1% Aged in 1 g/L Fe for 24 hours 100.00 94.71 5.29% 1.100 1.100% 0.715 0.755% 65.0% 35.0%           Aged in 0.1 g/L Fe for 1 hour 100.00 95.17 4.8% 1.100 1.100% 0.847 0.890% 77.0% 23.0% Aged in 0.1 g/L Fe for 24 hours 100.00 92.94 7.1% 1.100 1.10% 0.712 0.766% 64.8% 35.2%           Aged in water for 1 hour 100.00 93.21 6.8% 1.100 1.10% 0.806 0.865% 73.3% 26.7%           S1 HaHC extraction 100.00 71.42 28.6% 1.100 1.100% 0.582 0.815% 53.0% 47.0%           Aged then 1 g/L Fe for 1 hour 93.11 91.25 2.0% 0.835 0.897% 0.762 0.835% 91.3% 8.7% Aged then 0.1 g/L Fe for 1 hour 93.11 89.29 4.1% 0.835 0.897% 0.783 0.877% 93.8% 6.2%           S1 HAHC then aged 100.00 92.98 7.0% 0.822 0.822% 0.702 0.754% 85.3% 14.7% S1 HaHC aged then HaHC 92.98 85.62 7.9% 0.702 0.754% 0.647 0.755% 92.2% 7.8% S1 HAHC then 0.1 g/L Fe 100.00 93.87 6.1% 0.822 0.822% 0.712 0.758% 86.6% 13.4%  mass % of Cu % of Rec Cu       S1 removable by ageing only 0.33 29.7% 57.3%       S1 removable by HaHC only 0.206 18.70% 36.1%       *S1 removable by Fe displacement only 0.037 3.4% 6.6%       * needs further confirmation 0.57 51.8% 100.0%          153                   VALE RESIDUE MASS BALANCE Description mass in (g) mass out (g) mass loss Cu in (g) grade (%) Cu out (g) grade (%) %solids recovery  S3 aged for 24 hours 100.00 95.41 4.6% 0.58 0.58% 0.54 0.56% 91.8% Aged then HAHC 95.41 93.16 2.4% 0.54 0.56% 0.50 0.54% 93.7% Aged then HAHC then Fe  93.16 92.91 0.3% 0.50 0.54% 0.50 0.54% 100.0%          Aged in 1 g/L Fe for 1 hour 100.00 98.15 1.8% 0.58 0.58% 0.55 0.56% 94.9% Aged in 1 g/L Fe for 24 hours 100.00 92.63 7.4% 0.58 0.58% 0.31 0.34% 53.5%          Aged in 0.1 g/L Fe for 1 hour 100.00 95.66 4.3% 0.58 0.58% 0.53 0.56% 91.0% Aged in 0.1 g/L Fe for 24 hours 100.00 92.63 7.4% 0.58 0.58% 0.47 0.51% 80.7%          Aged in water for 1 hour 100.00 94.83 5.2% 0.58 0.58% 0.49 0.52% 84.5%          S3 HaHC extraction 100.00 97.57 2.4% 0.58 0.58% 0.47 0.48% 80.1%          Aged then 1 g/L Fe for 1 hour 100.00 98.63 1.4% 0.58 0.58% 0.57 0.57% 96.9% Aged then 0.1 g/L Fe for 1 hour 100.00 95.63 4.4% 0.58 0.58% 0.56 0.58% 95.3%          S3 HaHC then aged  97.57 93.57 4.1% 0.52 0.53% 0.36 0.38% 69.1% S3 HaHC aged then HaHC 93.57 93.90 -0.3% 0.36 0.38% 0.35 0.37% 96.4% S3 HaHC then 0.1 g/L Fe  97.57 93.45 4.2% 0.52 0.53% 0.41 0.44% 78.8%   154              mass % of Cu % of Rec Cu mass % of Ni % of rec Ni Removable by ageing only 0.01 2.3% 16.6% 0.20 23.7% 86.1% Removable by HaHC only 0.068 11.7% 83.4% 0.000 0.0%  *Removable by Fe displacement only 0.000 0.0% 0.0% 0.032 3.8% 13.9% * needs further confirmation 0.08 14.0% 100.0% 0.23 27.5% 100.0%   Description mass in (g) mass out (g) mass loss Ni in (g) grade (%) Ni out (g) grade (%) %solids recovery  S3 aged for 24 hours 100.00 95.41 4.6% 0.83 0.83% 0.63 0.66% 76.3% Aged then HAHC 95.41 93.16 2.4% 0.63 0.66% 0.63 0.68% 100.0% Aged then HAHC then Fe  93.16 92.91 0.3% 0.63 0.68% 0.60 0.65% 95.0%          Aged in 1 g/L Fe for 1 hour 100.00 98.15 1.8% 0.83 0.83% 0.76 0.78% 91.8% Aged in 1 g/L Fe for 24 hours 100.00 92.63 7.4% 0.83 0.83% 0.62 0.67% 75.2%          Aged in 0.1 g/L Fe for 1 hour 100.00 95.66 4.3% 0.83 0.83% 0.77 0.80% 92.7% Aged in 0.1 g/L Fe for 24 hours 100.00 92.63 7.4% 0.83 0.83% 0.59 0.63% 70.8%          Aged in water for 1 hour 100.00 94.83 5.2% 0.83 0.83% 0.70 0.74% 84.8%          S3 HaHC extraction 100.00 97.57 2.4% 0.83 0.83% 0.79 0.81% 95.0%          Aged then 1 g/L Fe for 1 hour 100.00 98.63 1.4% 0.83 0.83% 0.80 0.81% 96.8% Aged then 0.1 g/L Fe for 1 hour 100.00 95.63 4.4% 0.83 0.83% 0.76 0.80% 91.8%          S3 HaHC then aged  97.57 93.57 4.1% 0.78 0.80% 0.58 0.62% 75.0% S3 HaHC aged then HaHC 93.57 93.90 -0.3% 0.58 0.62% 0.58 0.62% 100.0% S3 HaHC then 0.1 g/L Fe  97.57 93.45 4.2% 0.78 0.80% 0.68 0.73% 87.1% 

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:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0166767/manifest

Comment

Related Items