UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Reduction of limonitic laterites by ferrous sulphate in ammoniacal solutions Zúñiga, Mariela 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_2016_february_zuniga_mariela.pdf [ 3.57MB ]
Metadata
JSON: 24-1.0220752.json
JSON-LD: 24-1.0220752-ld.json
RDF/XML (Pretty): 24-1.0220752-rdf.xml
RDF/JSON: 24-1.0220752-rdf.json
Turtle: 24-1.0220752-turtle.txt
N-Triples: 24-1.0220752-rdf-ntriples.txt
Original Record: 24-1.0220752-source.json
Full Text
24-1.0220752-fulltext.txt
Citation
24-1.0220752.ris

Full Text

REDUCTION OF LIMONITIC LATERITES BY FERROUS SULPHATE INAMMONIACAL SOLUTIONSbyMariela ZúñigaB.Sc., The University of Concepcion, 2010A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF APPLIED SCIENCEinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Materials Engineering)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)November 2015© Mariela Zuniga, 2015iiAbstractLeaching of a limonitic laterite (1.49% Ni, 0.33 % Co, 2.29% Mn, 51.8% Fe) usingferrous sulphate as reductant was studied in ammoniacal solutions. Studied parametersincluded temperature, agitation, ferrous sulphate concentration, solids content, ammoniaconcentration and a comparison between ammonium sulphate and chloride. Tests wereperformed in a batch cell with temperatures between 20 and 80ºC at atmospheric pressure.Leaching kinetics for nickel were different to cobalt and manganese.Nickel was favoured by temperature in ammonium sulphate reaching 66% extraction in 3hours, whereas in ammonium chloride it showed slower but steady kinetics with an extractionof 75% in 24 hours. Cobalt extraction in ammonium sulphate was low at high temperature andsolids content; however, at low solids content (4% w/w) and high temperature (80°C) cobaltextraction reached 80% in 4 hours. With 11% solids at the same temperature, extractiondecreased to 20%. In presence of ammonium chloride, temperature had a positive impact oncobalt reaching 80 % in 8 hours at 23% w/w solids. Cobalt appeared to re-precipitate inpresence of ammonium sulphate.Manganese behaved similarly to cobalt, however, in presence of ammonium sulphate, itonly reached 60% extraction after 8 hours at 80°C at low solids. In presence of ammoniumchloride, manganese reached 65% in 3 hours while increasing solids was not detrimental.Manganese suffered adsorption and/or co-precipitation in ammonium sulphate and chloride.This decrease in extraction exhibited by cobalt and manganese was attributed to co-precipitation and/or adsorption in iron oxides/hydroxides.Nickel extraction appeared aided by ferrous sulphate. Cobalt and manganese improved inpresence of higher ammonia/ammonium, agreeing with the observation that cobalt andmanganese are more prone to co-precipitation/adsorption; for which they prefer lower ferroussulphate and higher ammonia/ammonium in order to stabilize their ammines.iiiFeed and solid residue analyses using chemical assays, x-ray diffraction, scanningelectron microscopy and mineralogy, provided better understanding of limonite reduction. Themain phase in the residue was magnetiteDifferences in behaviour between nickel and cobalt/manganese suggest a two-stageprocess, and chloride solutions seem promising due to better stability of cobalt and manganese.ivPrefaceThis dissertation is original, unpublished, independent work by the author, M. Zuniga.vTable of contentsAbstract .......................................................................................................................................iiPreface ........................................................................................................................................ivTable of contents.........................................................................................................................vList of tables..............................................................................................................................viiList of figures ............................................................................................................................. ixAcknowledgments .....................................................................................................................xv1 INTRODUCTION ...............................................................................................................12 OBJECTIVES......................................................................................................................33 LITERATURE REVIEW ....................................................................................................43.1 Historical overview .......................................................................................................43.2 Nickel: physical and chemical properties......................................................................43.3 Nickel and cobalt: ores, main reserves, market and uses ..............................................63.4 Laterites: main deposits, formation and composition .................................................123.5 Main processes to recover nickel and cobalt............................................................... 143.6 Leaching of nickel laterites .........................................................................................153.6.1 Acid medium........................................................................................................153.6.2 Alkaline medium..................................................................................................213.6.3 General comparison of ammoniacal and acid routes ...........................................243.6.4 Ammoniacal leaching .......................................................................................... 263.7 Proposed leaching mechanism with ferrous sulphate .................................................374 EXPERIMENTAL PROCEDURES..................................................................................394.1 Sample preparation......................................................................................................394.2 Leaching tests ..............................................................................................................394.3 Solutions analysis ........................................................................................................424.4 Solids analysis .............................................................................................................424.5 Ore characterization ....................................................................................................435 RESULTS AND DISCUSSION........................................................................................485.1 Effect of agitation........................................................................................................485.2 Effect of solids content................................................................................................ 525.3 Effect of temperature...................................................................................................565.4 Effect of pH and ammonia concentration ...................................................................66vi5.5 Effect of ferrous sulphate concentration .....................................................................745.6 Comparison of chloride and sulphate solutions .......................................................... 815.7 Maximum expected extraction vs. adsorption and/or co-precipitation .......................956 CONCLUDING REMARKS........................................................................................... 1007 RECOMMENDATIONS FOR FUTURE RESEARCH..................................................102REFERENCES ........................................................................................................................104Others sources......................................................................................................................110APPENDICES .........................................................................................................................111Iron balance.......................................................................................................................... 111TIMA report.........................................................................................................................113Activation energy calculations............................................................................................. 135Proposed simplified flowsheet ............................................................................................. 150Total ammonia/ammonium assays.......................................................................................152Experimental data ................................................................................................................154viiList of tablesTable 3.1: Physical properties of nickel (Ballester, 2000)........................................................... 5Table 3.2: Chemical properties of nickel (Ballester, 2000) .........................................................5Table 3.3: Comparison of ammonia and acid routes for treating nickel laterites. .....................25Table 3.4: Properties of ammonia (Meng 1996) ........................................................................28Table 3.5: Partial pressure (kPa) of aqueous solutions of NH3..................................................32Table 4.1: Chemical analysis of the tested laterite (ICP)........................................................... 44Table 4.2: Quantitative phase analysis (wt %) using XRD .......................................................44Table 4.3: Quantitative phase analysis (wt %) using TIMA......................................................45Table 4.4: Metal deportment by mineral in head sample (from TIMA)....................................46Table 5.1: Solid residue phase analysis (XRD). Effect of temperature in ammonium sulphate.Total ammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4to laterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solutionvolume of 500 ml and a pH value adjusted to the ammonium/ammonia buffer pointcorresponding to each temperature. ........................................................................................... 59Table 5.2: Solid residue phase analysis. Effect of temperature in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................60Table 5.3: Solids residue phase content analysis after leaching in the presence of ammoniumsulphate. Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140g/l, initial mass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and asolution volume of 500 ml and 80°C working temperature.......................................................77Table 5.4: Solids residue phase content analysis after leaching in the presence of ammoniumchloride. Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 185g/l, initial mass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and asolution volume of 500 ml and 80°C working temperature.......................................................77Table 5.5: Ferrous sulphate to metal extracted ratio in the presence of ammonium sulphate. ..92Table 5.6: Ferrous sulphate to metal extracted ratio in the presence of ammonium chloride. ..93Table 5.7: Metal deportment by phase in the feed and leach residue (in µg) ............................ 97Table 5.8: Metal deportment by phase in the feed and leach residue (in w/w %) .....................97Table 5.9: Metal extraction per solid phase (* Ads. = possible adsorption and/or co-precipitation) .............................................................................................................................. 97Table 5.10: Maximum estimated extraction based on possible adsorbed and/or co-precipitated(Co-pp) metal plus metal dissolved during the leach................................................................. 98viiiTable9A.1: Sample mass and ICP assay..................................................................................111Table9A.2: Iron balance in head sample as per TIMA analysis. .............................................111Table9A.3: Iron balance in head sample as per Q-XRD analysis............................................111Table9A.4: Iron balance in leach residue as per TIMA analysis .............................................112Table9A.5: Iron balance in leach residue as per Q-XRD analysis ..........................................112Table9A.6: Iron balance comparison between ICP, Q-XRD and TIMA analyses. .................112Table9A.7: Iron balance comparison between ICP, Q-XRD and TIMA analyses using ICP asthe base case (100 %)...............................................................................................................112Table9A.8: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect oftemperature in the presence of ammonium sulphate................................................................ 154Table9A.9: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofagitation in the presence of ammonium sulphate.....................................................................154Table9A.10: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofsolids content in the presence of ammonium sulphate............................................................. 155Table9A.11: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect oftemperature in the presence of ammonium chloride................................................................ 155Table9A.12: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofagitation in the presence of ammonium chloride .....................................................................156Table9A.13: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofagitation in the presence of ammonium chloride .....................................................................156ixList of figuresFigure 3.1: Main nickel reserves in the world (www.infomine.com)..........................................7Figure 3.2: Nickel price during the last year (www.infomine.com). ...........................................8Figure 3.3: Cobalt price during the last year (www.infomine.com). ...........................................8Figure 3.4: Manganese price during the last year (www.infomine.com). ...................................9Figure 3.5: Main nickel world producers (www.infomine.com). ................................................9Figure 3.6: Main industrial uses of nickel (www.infomine.com)..............................................10Figure 3.7: World cobalt production. (www.infomine.com) .....................................................11Figure 3.8: Main industrial uses of cobalt. (www.infomine.com).............................................11Figure 3.9: Laterization zones schematic as presented by Krause (Krause, 2009)....................14Figure 3.10: Ammonia speciation diagram at 25oC (Asselin 2008) ..........................................30Figure 3.11: Nickel ammines speciation diagram at 25oC, 0.1 M Ni, 6 M total NH3 (Asselin2008). .........................................................................................................................................30Figure 3.12: Cobalt ammines speciation diagram at 25oC, 0.1 M Co, 6 M total NH3 (Asselin2008). .........................................................................................................................................31Figure 3.13: Iron ammines speciation diagram at 25oC, 10-3 M Fe, 6 M total NH3 (Asselin2008). .........................................................................................................................................31Figure 3.14: Ammonia partial pressure at 80oC with respect to different ammoniaconcentrations (Handbook of Chemistry and Physics 2004) .....................................................33Figure 3.15: Ammonium sulphate solubility at different temperatures (Data Handbook ofChemistry and Physics 2004).....................................................................................................34Figure 3.16: Eh-pH diagram Ni-NH3-H2O system at 25oC .......................................................35Figure 3.17: Ni-NH3-H2O Eh-pH diagram, 0.1 M Ni, 6 M total NH3 at 25oC (Asselin 2008). 35Figure 3.18: Co-NH3-H2O Eh-pH diagram, 0.1 M Co, 6 M total NH3 at 25oC (Asselin 2008).36Figure 3.19: Fe-NH3-H2O Eh-pH diagram, 10-3 M Fe, 6 M total NH3 at 25oC (Asselin 2008).36Figure 4.1: Experimental setup ..................................................................................................42Figure 4.2: Metal deportment by mineral in head sample (from TIMA)...................................46Figure 4.3: (a) and (b) Backscatter Electron Image (BEI) of composite particle of hematite,goethite, and asbolane in New Caledonia Head.........................................................................47Figure 5.1:Effect of agitation velocity on nickel removal in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................49xFigure 5.2: Effect of agitation velocity on nickel removal in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................50Figure 5.3: Effect of agitation velocity on cobalt removal in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................50Figure 5.4: Effect of agitation velocity on cobalt removal in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................51Figure 5.5: Effect of agitation velocity on manganese removal in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................51Figure 5.6: Effect of agitation velocity on manganese removal in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................52Figure 5.7: Effect of solids content on nickel extraction in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................53Figure 5.8: Effect of solids content on nickel extraction in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................54Figure 5.9: Effect of solids content on cobalt extraction in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................54Figure 5.10: Effect of solids content on cobalt extraction in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................55Figure 5.11: Effect of solids content on manganese extraction in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 150 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................55Figure 5.12: Effect of solids content on manganese extraction in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringxivelocity of 600 RPM, an F80 of 150 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0. ....................................................................56Figure 5.13: Effect of temperature on nickel extraction in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................60Figure 5.14: Effect of temperature on nickel extraction in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................61Figure 5.15: Effect of temperature on cobalt extraction in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................61Figure 5.16: Effect of temperature on cobalt extraction in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................62Figure 5.17: Effect of temperature on manganese extration in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................62Figure 5.18: Effect of temperature on manganese extraction in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................63Figure 5.19: Effect of temperature on iron precipitation in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................63Figure 5.20: Effect of temperature on iron precipitation in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................64xiiFigure 5.21: Effect of temperature on nickel extraction in ammonium sulphate in dilute slurry(4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 to lateriteratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volume of 500ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding to eachtemperature. ............................................................................................................................... 64Figure 5.22: Effect of temperature on cobalt extraction in ammonium sulphate in dilute slurry(4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 to lateriteratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volume of 500ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding to eachtemperature. ............................................................................................................................... 65Figure 5.23: Effect of temperature on manganese extraction in ammonium sulphate in diluteslurry (4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature. .......................................................................................................................65Figure 5.24: Effect of temperature on iron precipitation in ammonium sulphate in dilute slurry(4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 to lateriteratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volume of 500ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding to eachtemperature. ............................................................................................................................... 66Figure 5.25: Effect of pH on nickel extraction in ammonium sulphate and chloride solutions.69Figure 5.26: Effect of pH on cobalt extraction in ammonium sulphate and chloride solutions.70Figure 5.27: Effect of pH on manganese extraction in ammonium sulphate and chloridesolutions. ....................................................................................................................................70Figure 5.28: Effect of total ammonia/ammonium concentration on nickel extraction inammonium sulphate. ..................................................................................................................71Figure 5.29: Effect of total ammonia/ammonium concentration on nickel extraction inammonium chloride ...................................................................................................................71Figure 5.30: Effect of total ammonia/ammonium concentration on cobalt extraction inammonium sulphate ...................................................................................................................72Figure 5.31: Effect of total ammonia/ammonium concentration on cobalt extraction inammonium chloride ...................................................................................................................72Figure 5.32: Effect of total ammonia/ammonium concentration on manganese extraction inammonium sulphate ...................................................................................................................73Figure 5.33: Effect of total ammonia/ammonium concentration on manganese extraction inammonium chloride. ..................................................................................................................73Figure 5.34: Effect of ferrous sulphate addition on nickel extraction in ammonium sulphate.Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140 g/l, initialmass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and a solution volumeof 500 ml and 80°C working temperature. ................................................................................78xiiiFigure 5.35: Effect of ferrous sulphate addition on nickel extraction in ammonium chloride.Total ammonia/ammonium concentration at 185 g/l, initial mass sample of 60 grams, stirringvelocity of 600 RPM, an F80 of 32 μm and a solution volume of 500 ml and 80°C workingtemperature. ............................................................................................................................... 78Figure 5.36: Effect of ferrous sulphate addition on cobalt extraction in ammonium sulphate.Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140 g/l, initialmass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and a solution volumeof 500 ml and 80°C working temperature. ................................................................................79Figure 5.37: Effect of ferrous sulphate addition on cobalt extraction in ammonium chloride.Total ammonia/ammonium concentration at 185 g/l, initial mass sample of 60 grams, stirringvelocity of 600 RPM, an F80 of 32 μm and a solution volume of 500 ml and 80°C workingtemperature. ............................................................................................................................... 79Figure 5.38: Effect of ferrous sulphate addition on manganese extraction in ammoniumsulphate. Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140g/l, initial mass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and asolution volume of 500 ml and 80°C working temperature.......................................................80Figure 5.39: Effect of ferrous sulphate on manganese extraction in ammonium chloride. Totalammonia/ammonium concentration at 185 g/l, initial mass sample of 60 grams, stirringvelocity of 600 RPM, an F80 of 32 μm and a solution volume of 500 ml and 80°C workingtemperature. ............................................................................................................................... 80Figure 5.40: Solid residue micrograph after leaching in the presence of ammonium sulphate.83Figure 5.41: Solid residue micrograph after leaching in the presence of ammonium chloride .84Figure 5.42: Solid residue micrograph after leaching in the presence of ammonium sulphate. 84Figure 5.43: Solid residue micrograph after leaching in the presence of ammonium chloride .85Figure 5.44: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating ultra-fine, subhedral and euhedral cubic grains of likely magnetite and/or wustite.....................................................................................................................................................85Figure 5.45: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating remnant particle of hematite and rim of asbolane. ..................................................86Figure 5.46: Detail of partially leached particle from Figure5.38  depicting metal distribution.(a) Iron, (b) Nickel, (c) Cobalt, (d) Manganese. ........................................................................87Figure 5.47: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating martite/hematite with possible dissolution edges with absent nickel. .....................88Figure 5.48: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating a remnant composite particle of discontinuous rim of goethite on martite-hematitepseudomorph after magnetite.....................................................................................................88Figure 5.49: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating a remnant binary particle of mostly magnetite with subhedral pseudomorphs ofmartite-hematite. ........................................................................................................................89xivFigure 5.50: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating a remnant particle of martite-hematite with a discontinuous rim of goethite..........89Figure 5.51: Nickel extraction comparison in the presence of either ammonium sulphate orammonium chloride. ..................................................................................................................90Figure 5.52: Cobalt extraction comparison in the presence of either ammonium sulphate orammonium chloride ...................................................................................................................90Figure 5.53: Manganese extraction comparison in the presence of either ammonium sulphateor ammonium chloride. ..............................................................................................................91Figure 5.54: Two-Stage nickel leaching in the presence of ammonium sulphate .....................94Figure 5.55 Two-Stage nickel leaching in the presence of ammonium sulphate ......................94Figure 5.56: Two-Stage nickel leaching in the presence of ammonium sulphate .....................95Figure 5.57: Metal deportment in feed ......................................................................................98Figure 5.58: Metal deportment in leach residue ........................................................................99xvAcknowledgmentsI would like to thank my supervisor, Dr. Edouard Asselin for his support throughout thiswork and Dr. David Dreisinger and the Hydrometallurgy chair for the funding provided for thedevelopment of this project. I would also like to thank Elisabetta Pani and Matti Raudsep(UBC, Earth and Ocean Sciences) together with Geoff Lane (Process MineralogicalConsulting Ltd.) for their support and expertise in the feed and solid residue characterizationsection of this thesis and to Mr. Derek Barratt for his advice. Lastly, I would like to thank myfamily and friends and special thanks to Fernando for his support and help.11 INTRODUCTIONThe hydrometallurgical treatment of limonitic laterite minerals has established itself as aviable processing route complying with technical commercial and environmentalconsiderations, despite the challenges associated with some of the recent Australian and NewCaledonian Pressure Acid Leach (PAL) facilities. It is expected that, in the future, nickel andcobalt production will mostly arise from hydrometallurgical processes.Limonite contains iron in the form of goethite with the valuable metals locked in the ironmatrix as goethite, maghemite, hematite, etc. The presence of iron as the main element couldallow direct smelting of the ore, thus forming ferronickel alloys but, due to extremely fineparticle size, elements that could report to the slag and perhaps hinder phase separation, as wellas high costs associated with drying, smelting is not an economically attractive alternative forlimonites.Several studies have been performed in order to extract Ni and Co in ammoniacal mediaat near neutral pH. However, the goethite cannot be directly leached under these conditions asit must first be reduced in order to release nickel and cobalt to solution.The Caron process for extraction of nickel and cobalt in ammoniacal solution was firstused in 1942. This process includes a reductive roasting stage followed by oxidativeammoniacal leaching. This process presents some limitations such as high operational cost;mainly from the drying and roasting stages. It is also known to have poor Co and Niextractions (on the order of 50 % to 80%) probably associated with either the adsorption ofcobalt onto ferric oxy-hydroxides, which are precipitated during the leach, or the passivation ofthe Fe-Ni-Co alloys formed during the reduction step (Caron 1924, Caron 1950, Han 1993,Meng 1996, Dyer 2012).A large amount of research has been conducted on acid leaching. This includes PAL(general conditions 245 - 270ºC) or very high acidity under atmospheric conditions (~100ºC)to release Ni from goethite and processes including the addition of sea water and in situprecipitation of iron as jarosites such as the JAL (Jarosite Atmospheric Leach) or the Eramet2Process (Agin 2012). In general, acid leaching processes have high recoveries for nickel andcobalt but they are subject to high acid consumptions and leach liquors usually requireneutralization prior to further treatment. On the other hand, ammonia has several propertiesthat make it attractive as a leaching agent such as low toxicity, low cost, high selectivity fornickel and cobalt, the lack of a need for expensive reactor materials (such as titanium) and itoffers the potential of being regenerated for re-use in the process. Ammonia has beensuccessfully used in the processing of nickel and cobalt.This work reports experimental results obtained during the leaching of a limonitic lateritein ammoniacal solutions using ferrous sulphate as a reducing agent at atmospheric conditions.This work will first present a literature review to discuss the advantages anddisadvantages of the treatments currently used for leaching limonite, a discussion of theproposed leaching mechanism for the studied method of reductive ammoniacal leaching,followed by a description of the experimental procedures implemented throughout thisresearch work. This work includes a complete solids characterization (including head andleach residue samples), the effect of several variables that can affect the kinetics and otherexperimental results. The last section of this work includes conclusions and recommendationsfor future research.32 OBJECTIVESThe ultimate goal of this research work is to evaluate whether a limonitic laterite can bereduced using ferrous sulphate in ammoniacal solutions to release Ni, Co and Mn in a leachprocess under atmospheric pressure conditions and understand the variables affecting theleaching mechanisms of such an option and determine experimentally the best leachingconditions. If successful, leaching conditions are to be optimized in order to provide suitableparameters for further research and development.The studied variables included agitation, solids content in the slurry, temperature, initialtotal ammonia [NH3 + NH4+] concentration and ferrous sulphate concentration. A mediacomparison (sulphate and chloride) was also included.43 LITERATURE REVIEW3.1 Historical overviewNickel is a very abundant element on earth, but not in the crust. This element ranks 24thin order of abundance in the earth crust, with an estimated concentration of 0.008 %.Meteorites that landed on earth in the past, had high nickel concentration, thus natural nickel-iron alloys (over 30 % Ni concentration) have been found. Nickel metal was identified andisolated by A. Cronsted in Sweden in 1751 from an arsenical mineral. However, others believethat nickel was obtained in pure form in the late XVIII century (Ballester, 2000). Prior to this,nickel was found in copper mines and thought to be an “un-smeltable” copper ore. Primarynickel can resist corrosion and maintains its physical and mechanical properties even whenplaced under extreme temperatures. When these properties were recognized, the developmentof primary nickel began. It was found that by combining primary nickel with steel, even insmall quantities, the durability and strength of the steel increased significantly as did itsresistance to corrosion. This partnership has remained and the production of stainless steel isnow the single largest consumer of primary nickel today. Nickel is also used in the productionof many different metal alloys for specialized use.The Norwegians mined nickel sulphides during the XIX century. Subsequently oxideminerals from New Caledonia were mined. Later, nickel was extracted from sulphides inSudbury, Ontario (Canada). The Russians extracted nickel from Kola and Norilsk. Both ofthese Canadian and Russian deposits continue to be exploited. The massive use of nickelbegan early in the XX century, mainly as alloying in the steel industry (Ballester, 2000).3.2 Nickel: physical and chemical propertiesNickel is a white-grey metal, bright, hard, malleable and ductile, which shows magneticproperties at 353ºC. Table 3.1 summarizes the main physical properties and Table 3.2 showsits chemical properties (Ballester, 2000).5Table 3.1: Physical properties of nickel (Ballester, 2000)Properties Values UnitsDensity (20ºC) 8.9 g*cm-3Melting point 1453 ºCBoiling point 2732 ºCSpecific heat (25ºC) 0.468 J*ºC-1*g-1Thermal conductivity (25ºC) 87.3 W*m-1*K-1Curie point 353 ºCElectric resistivity (20ºC) 7.2 µΩ*cmYoung modulus 210 GPaBrinell Hardness 85Table 3.2: Chemical properties of nickel (Ballester, 2000)Properties Values UnitsAtomic mass 58.69 g*mol-1Electronic configuration 3s23p63d84s2Valence 2 and 3Atomic radius 0.125 nmIonic radio (Ni2+) 0.078 nmElectronegativity 1.8Atomic number 28Reduction potential (Ni2+- Ni) -0.23 V6Nickel cannot be oxidized with air at room temperature. This noble behaviour is due toits self-passivation capability, which is transmitted to alloyed metals like iron. Nickel isresistant to water. Even sea water results in a very slow attack on nickel. Attack by sulphuric,hydrochloric and hydrofluoric acid at concentrations below 10 % under mild temperatures isalso slow. However, nitric acid can attack nickel very strongly given its oxidizing nature.Nickel resists attack from alkalis either molten or in dissolution. Dissolved salts do not attacknickel except strong oxidants like FeCl3. At high temperatures, over 643ºC, SO2 and othersulphide gases can attack nickel metal to produce nickel sulphides (Ballester, 2000).CO reacts with nickel between 50 and 150ºC to produce nickel carbonyl, Ni(CO)4,reaction which is the basis for Inco’s carbonyl nickel refinery technology. Chlorine and otherhalogens do not attack nickel at high temperatures. Hydrogen gas easily diffuses throughnickel, and this tendency increases with temperature. Finally, nickel is a good catalyst inhydrogenation reactions; it is widely used in the petrochemical industry (Ballester, 2000).3.3 Nickel and cobalt: ores, main reserves, market and usesThe main nickel reserves are in the form of sulphides and oxides. The main sulphide oreis pentlandite (Ni,Fe)9S8, which accompanies other sulphides like millerite (NiS) and linnaeites((Fe,Co,Ni)3S4). Sulphide ores also often contain pyrite, pyrrhotite and chalcopyrite, whichmay also contain nickel and cobalt. Arsenic associated with nickel can also be found but it isless common. Oxide deposits can be found in tropical countries on magnesium silicate rockscontaining aluminum and iron. This type of deposit can contain approximately 0.2 % nickel.Metals in this kind of rock are dissolved, along with silica, and both precipitate in remote andlow areas of the deposit in the form hydrated nickel silicates called garnierites. If the rockshave been attacked by the weather, leading to dissolution of silica, such deposits are calledlaterites (Ballester, 2000). Another source of nickel that could become of interest in the futureis marine nodules with nickel contents between 0.3-0.9 % (Ballester, 2000).Nearly 60 % of world nickel production comes from high grade sulphidic ores whichrepresent only 30 % of the world reserves. Approximately 70 % of the world’s nickel reservesare found in low grade laterite deposits and this represents only approximately 40 % of the7world nickel production. Figure 3.1 shows the main nickel reserves in the world and therelationships between nickel laterite and sulphides (Krause, 2009).Figure 3.1: Main nickel reserves in the world (www.infomine.com).The price of metals is sensitive to the rate of economic growth in developing countries,especially in the case of nickel and cobalt. This occurs because the main application for thesemetals is to produce steel alloys. Steel consumption is directly related to developing countriesthrough its relationship to infrastructure. Figure 3.2 shows that the price has been decreasingsince September 2014. The price of cobalt is shown in Figure 3.3, and it is approximately 12USD/lb. It is important to mention that cobalt generally presents a price that is considerablyhigher than that of nickel, and it is often a by-product of nickel production. This fact makescobalt extraction a key issue when treating nickel laterites. Figure 3.4 shows the price ofmanganese (www.infomine.com, www.lme.com).01020304050607080Ni-laterites Ni-sulphidesNickel reserves (%)8Figure 3.2: Nickel price during the last year (www.infomine.com).Figure 3.3: Cobalt price during the last year (www.infomine.com).9Figure 3.4: Manganese price during the last year (www.infomine.com).Figure 3.5: Main nickel world producers (www.infomine.com).0510152025303540Europe Asia America Oceania AfricaNickel production(%)10In 2011 world nickel production reached 1.66 million tonnes. Figures 3.5 and 3.6 show asummary of the main world nickel producers and industrial nickel consumptions(www.infomine.com, www.lme.com).Figure 3.6: Main industrial uses of nickel (www.infomine.com).Cobalt is a grey, hard, lustrous metal, which was first discovered in 1739 and has a worldproduction of 82,200 tonnes (2011). Cobalt has many applications such as in the production ofcatalysts, batteries, chemicals, alloys and dyes. Figures 3.6 and 3.7 show a summary of themain world cobalt producers and its uses (www.infomine.com).010203040506070Stainless steel Other alloys Electroplating ChemicalConsumptions of nickel (%)11Figure 3.7: World cobalt production. (www.infomine.com)Figure 3.8: Main industrial uses of cobalt. (www.infomine.com)05101520253035404550Asia Europe America Oceania AfricaCobalt production(%)051015202530Batteries Superalloys Other Carbides &diamondtoolingColours &pigmentsCatalystsConsumptions of cobalt (%)123.4 Laterites: main deposits, formation and compositionLaterites consist mainly of iron oxides and hydrous magnesium silicates. They containtraces of valuable elements such as nickel and cobalt. Laterization is found generally nearequatorial regions, which are known for having a warm climate with abundant rain. Placeswhere laterites can be found include Cuba, Dominican Republic, Guatemala, New Caledonia,Indonesia, Philippines, Australia, Papua New Guinea, Brazil, Madagascar, Southwest Europe,Turkey, India, Southwest Asia and China (Monhemius 1987, Kerfoot 1988, Krause 2009).Laterization is a weathering chemical process with ultra-basic rocks, such as dunite orperidonite. Peridonite consist mainly in phosphoritic olivine and serpentinite which involvesmovement of elements in minerals by dissolution and precipitation (Krause 2009, Monhemius1987, Kerfoot 1988).In the structure of olivine and serpentinite, elements such as Ni, Fe and Mg can besubstituted as a consequence of their similar ionic radius. In the peridonite case, thecontribution of nickel is approximately 0.2-0.3%.During the laterization process, dissolution of carbon dioxide from the atmosphere andorganic acids acidify underground water, leading to gradual decomposition of peridonite, withpossible bacterial activity, which leads to percolation of elements such as Fe, Al, Cr, Ni, Mgand Si. Iron oxidizes quickly and precipitates, mostly at the surface, thus forming goethite (α-FeOOH). Aluminum also precipitates either in the goethite or as gibbsite (Al(OH)3).Chromium is found as a spinel associated with goethite. Cobalt and nickel precipitate togetherwith iron forming part of the crystalline lattice in the goethite (Krause 2009, Monhemius 1987,Kerfoot 1988). Next, a summary of the reactions taking place in the formation of laterites isshown in equations 3.1 to 3.3:Leaching: Peridonite      44242 ,,2,, SiONiMgFeSiONiMgFe 3.113Oxidation-Hydrolysis  HOHFeOOOHFe 4*25.032 222 3.2Exchange reactions: Serpentinite      3452324523 33 MgOHOSiNiNiOHOSiMg 3.3During Laterization, it is possible to find three different zones:1. Limonitic zone: this has abundant goethite, which has a low nickel concentration (Ni ≤1.5). Low crystallization and a very fine grain size. It also contains gibbsite, chromiteand asbolane that is a complex Mn IV oxide phase (~20% Ni + Co).2. Serpentine zone: this is located between limonitic and saprolitic zones; it containsseveral magnesium silicates and intermediate amounts of Ni, Fe, Mg and Si.3. Saprolitic zone: this is located at the bottom of the ore. It is rich in Ni (1.8 - 3.0 w/w %)and magnesium silicates.  It is very heterogeneous in mineralogical composition as wellas in chemical.Figure 3.9 shows laterization zones. Due to different mineralogy, laterite ores can betreated by different processing routes. The saprolite may be treated pyrometallurgically, forexample for ferronickel production or by atmospheric acid leaching (heap leaching). Thelimonite contains iron in the form of goethite with the valuable metals locked in the ironmatrix. It requires acid leaching under pressure (245 - 270ºC) or very high acid additions underatmospheric conditions (~100ºC) to release Ni from the goethite phase. However, the presenceof iron could allow direct smelting of the ore, thus forming ferronickel alloys but, sinceparticle size is extremely fine and considering the many elements forming in the slag, as wellas high costs coming from drying, all these drawbacks make this alternative economicallyunattractive (Monhemius 1987, Kerfoot 1988).14Figure 3.9: Laterization zones schematic as presented by Krause (Krause, 2009).3.5 Main processes to recover nickel and cobaltThe processes most commonly used to recover nickel are:Pyrometallurgical processes:i. Ferronickel (10 - 20 % Ni product)ii. Nickel matte (~70  % Ni)iii. Blast furnace pig iron (~4 % Ni) – Stainless steel feedPyro - Hydrometallurgical:Caron process (selective reductive roasting- ammoniacal leach)Hydrometallurgical:i. Pressure acid leaching (PAL)ii. Atmospheric acid leaching (AL)The PAL and Ferronickel processes are used most frequently (Krause, 2009).153.6 Leaching of nickel lateritesThe leaching of nickel laterites is divided into acid (PAL, HPAL, Eramet) or alkalinesolutions (Caron Process).3.6.1 Acid medium3.6.1.1 Sulphuric acidThe acid leaching of limonitic laterites has been under development through threedifferent alternatives; PAL and HPAL, atmospheric acid leaching and heap leaching.i. High temperature and pressure acid leaching (PAL, HPAL)Originally, this alternative was developed in Russia for Moa Bay Cuban laterites withlow magnesium content. It requires no drying or roasting of the ore, therefore this process haslow energy cost. In general, pressure acid leaching presents high extraction of nickel andcobalt (over 90 %) but it is not selective.  Below, a chronology of technology development canbe seen (Rubisov 2000, Georgiou 2004-2009, Krause, 2009):- First plant by Freeport (pressure Pachuca reactors) Moa Bay, Cuba (late 1959) and it remainsin operation today.In the 70s and 80s the AMAX and Nical Processes were developed for the Prony Projectin New Caledonia and the Gasquet Project in Northern of California respectively, which werenot successful due to low nickel prices. Only in the 90s when three plants were built inWestern Australia, the PAL process became a commercially feasible process (Taylor, 2013).- Western Australian plants (horizontal autoclaves: Bulong (currently closed), Cawse(September ´98, currently closed), Murrin Murrin (December ´98).- Second generation PAL plants: Rio Tuba (Philippines) in 2005, Ravensthorpe (WA) in 2008,Goro (New Caledonia) in 2009.- PAL plants nearing completion: Ramu (PNG), Ambatovy (Madagascar)16At high temperature and pressure (245-270ºC and ~5MPa) the dissociation of sulphuricacid is shown in equations 3.4 to 3.11:  442 HSOHSOH 3.4Metals dissolution at 250ºC happens mainly as bisulphates:OHHSONiSOHNiO 24242 22   3.5Once the solution is removed from the autoclave, it cools down and HSO4- converts tosulphate:  244 SOHHSO 3.6The higher the soluble divalent metals content in the ore (Mg, Ni, Mn, Co), the higherthe PAL acid requirements, the bisulphate concentration at temperature, and the higher the freeacid content found after leaching and cooling (Krause, 2009).The required free acid concentration at 250ºC (H+ activity in the autoclave) is the same(~0.1 m) for limonite ore leaching (200 kg acid/ton ore) or for leaching a limonite-saprolitefeed (400 kg acid/t ore). The main PAL reactions that take place are (Ballester, 2000, Krause,2009, Papangelakis, 2009):Goethite conversion to hematite:HOFeOHFeOHFeHOHFeO63223*3223233.7Gibbsite conversion to hydronium alunite:3.8  HOHSOOAlHOHSOAl 5)()(723 624332243 3.9Chrome-VI formation:OHAlHOHAl 233 33)(  17  HHCrOMnOHCrMnO 223223 42232 3.10Iron-II oxidation:  2222 222 FeOHMnHFeMnO 3.11In the autoclave, operational issues such as material build-up on agitators, sensors andvalves in the form of pellets or, simple accumulation (scaling) are common. One of the mainoperating problems encountered in pressure leaching is the formation of scales inside theautoclave and pipe lines. At Moa Bay, scales mainly contain alunite, formed due to lack ofgood agitation in the autoclaves.  As a result, it is necessary to shut down the plant for about aweek every month to clean the scales (Krause, 2009).Super-saturation with respect to Al3+, Fe3+, MgX+, etc. will affect: type of precipitateformed depends on the type of water used in the process (sodium content), solubility differentcompounds, rate of mineral dissolution and rate of precipitation.  The previous factors areinfluenced by temperature, acidity and rate of agitation (Krause, 2009).One of the main economic challenges with PAL is associated with acid consumptions; itdepends mainly on the magnesium content in the ore. At 250ºC, the solubility of MgSO4 isdramatically decreased, and magnesium precipitation is crucial to acid consumption (Krause2009, Papangelakis 2009). Other issues associated with the PAL process include the highcapital and maintenance cost associated, process conditions which are highly corrosive,downstream processing is complex and the length of the ramp-up time for the majority of theseprojects (Taylor 2013).The following operations currently employ the PAL process: Moa Bay (Cuba) producingaround 33,000 tpy, Murrin Murrin (Australia) 44,000 tpy, Coral Bay (Philippines) 20,000 tpyand Goro (New Caledonia) designed for 60,000 tpy. Projects which have been more recentlyfinished include Ambatovy (Madagascar) with a target of 60,000 tpy, Ramu (Papua NewGuinea) with a target of 31,000 tpy, and in construction Taganito (Philippines) with a target of30,000 tpy and Gordes (Turkey) with a target of 10,000 tpy (Taylor 2013).18ii. Atmospheric leach processes (AL)A summary of the work performed to develop atmospheric leach techniques is given asfollows: In the late 70s Amax-Cofremmi used (calcined) saprolite to neutralize freeacid after PAL leaching of limonitic ore. Early 90s BHP Billiton developed atmospheric leach process for limoniticplus saprolite leach under atmospheric conditions and limonite using PAL.The main advantages of combining PAL for limonite and AL for saproliteincludes reducing the net acid consumption for saprolite, reduction oflimestone consumption for neutralizing residual solutions from PAL anddecreased capital cost. The main issue in this process is the relativelycomplex process with two different leach processes (Taylor 2013). Skye resources developed the sulphation process Heap leaching of Southwestern European laterites Heap leaching of saprolite ores (Murrin Murrin, others) Processes incorporating addition of sea water and in situ precipitation ofjarosites in acid solutions.BHP Billiton’s atmospheric whole ore leach process: High acid additions (~1400 kg/tlimonite) are required to leach limonite within 5-6 h at 100ºC. Excess acid is neutralized withsaprolite (~10 h retention), causing added nickel dissolution and partial iron removal: asgoethite when fresh water is used or jarosite with saline water. Economics of the processdepend on limonite to saprolite ratio, acid and limestone consumptions and Ni/Co extractions(82 – 92 %), which is significantly lower than in PAL (Mcdonald 2008, Krause, 2009).The AL process has been tested by several companies, as an example recently Eramethas adopted atmospheric leaching for the Weda Bay Project in Indonesia. The AL processshows advantages from the point of view of reduced capital and maintenance cost as well as19reduced down time. However, high operational cost as a result of higher sulphuric acidconcentrations and higher iron and other metals as impurities can be present downstream. Inmany cases injection of sulphur dioxide is necessary in order to increase cobalt extraction(Taylor 2013).iii. Heap leachingSeveral companies have attempted to develop this technology in the 90’s (e.g. Yuanjiang(China), Caldag (Turkey), Murrin Murrin (WA) and others). In 2007, Murrin Murrin inWestern Australia treated grinding circuit scats but no standalone commercial operation hasyet been established (Taylor 2013).Specifically the Caldag project in Turkey, has a stand-alone HL (heap leach)demonstration plant, which was operated by European Nickel but the project was held up forpermit issues and in 2011 was sold to VTG Nickel of Turkey who is aiming for a commercialapplication. The ore contains 80 % goethite with 1.15 % nickel.  Nickel extraction reached ~78% after 550 days with an acid consumption of ~530 kg/t ore (Krause, 2009).Low metals recoveries are commonly obtained in the HL process. The main operationalissues in the HL process are associated with the large amount of the fines in laterite ores,where agglomeration is required.  The stability of the heap is affected in high rainfall locationsespecially with the breakdown of the agglomerates and dilution of the leach solution. Highacid consumption, especially for saprolite ores, and also the high iron and impurities present inthe solution require more difficult downstream purification processes (Taylor 2013).3.6.1.2 Hydrochloric acidThe main acid medium for treatment of nickel ore is sulphuric acid.  However,numerous processes have been developed in chloride medium. Chloride shows someadvantages; it is faster and shows a more complete leach of limonite than sulphuric acid in the85 to 100ºC range (Krause, 2009).20Many private interests have tested chloride solutions for laterites including Neomet,Nichromet, Process Research Ortech in Canada, SMS Siemag in Austria and Anglo Researchin South Africa. No commercial chloride leaching plants for laterites are yet operating.Bryn Harris (Intec Process, Anglo Process) used HCl in highly concentrated MgCl2brines for atmospheric leaching of nickel laterites. Nickel, iron and magnesium mineralsdissolve much faster than in sulfuric acid and iron oxidation rates with air increasesignificantly with increasing brine concentration. A comparison on acid consumptions showthat 90 % of nickel extraction is achieved with ~975 kg HCl /t ore while ~1315 kg H2SO4/t oreare necessary. For cobalt, with 37 kg of hydrochloric acid per ton of ore, 90 % extraction isachieved, while only 40 % is recovered with 37 kg of sulphuric acid per ton of ore, under thesame conditions (Krause, 2009).The MgCl2 brine can be subjected to pyro-hydrolysis at ~500ºC to regenerate HCl asshown in equation 3.12:HClMgOOHMgCl 222  3.12The HCl is recycled to leaching, MgO is used for iron precipitation (mostly as hematite)and impurities are removed. Anglo research has worked on a less energy intensive HClregeneration process; however, it has not been demonstrated on larger scale. Such problemscould be due to technical challenges, such as Cl- recovery and water balance issues (Ballester2000, Krause 2009, Papangelakis 2009).There can be several advantages when leaching in a chloride solution that could bementioned. This medium is applicable for both, limonite and saprolite ores, in agitated tanks atatmospheric pressure, lower capital cost in comparison with the PAL process, recoveries arehigh for nickel and cobalt and the lixiviant is regenerated and recycled. In the other hand,separation and recovery of metals is straightforward and there are commercially establishedmethods, solid-liquid separation can be better than in sulphuric acid and a secondneutralization stage to remove residual iron is not always needed. However, there can be issuesassociated to high corrosion, especially at elevated temperatures, formation of toxic HCl21vapours and the water balance can be an environmental issue due to the release of chloridecontaining solutions.3.6.1.3 Nitric acidSome studies have been developed using nitric acid for laterites. A recent example is theDNi process in Western Australia. Currently, there are no commercial plants for nickel andcobalt applications, however, research are exploring to use nitric at atmospheric and underpressure conditions.The nitric media show some advantages such as being applicable to both ores (limoniteand saprolite) in agitated tanks at atmospheric pressure. High nickel and cobalt recoveries, aswell as low capital cost in comparison with the PAL process. The lixiviant could beregenerated and recycled. The nitric solutions are corrosive but less than chloride, in generalstainless 304 or 316 are suitable for nitric solutions. Formation of toxic vapours is a personnelsafety issue for the industrial application and also the release of nitrate in solutions can becomean environmental concern (Taylor 2013).3.6.2 Alkaline mediumAlkaline solutions are used in the Caron Process, which uses ammonia as a lixiviant.Currently four installations use the Caron process: Nicaro (1944) by Freeport, Cuba; Yalubu(1974) by Freeport, Queensland Australia; Tocantins (1982) by Votorantim, Brazil and PuntaGorda (1986) by Cuban-Russian, Cuba (Taylor 2013).The Caron process, developed by M.H. Caron in the 1920’s, consists of four main steps(Caron 1924, Monhemius 1987, Kerfoot 1988):i. Ore grinding and dryingii. Reductive roastingiii. Leaching with NH3/(NH4)2CO3 solutioniv. Metal recovery from solution22i. Ore grinding and dryingOre with 0 - 1.2 % nickel with low magnesium content. Blended ore is dried in rotarykilns to about 2 to 3 % moisture. Since the moisture content of the raw ore is between 30 and50 wt %, drying requires considerable energy inputs.  Heavy fuel oil, or in some cases, lessexpensive coal, is used in the kilns (Monhemius 1987, Kerfoot 1988).ii. Reductive roastingIt aims to selectively reduce nickel and cobalt to their metallic state, minimizing thereduction of iron. Strongly reducing conditions are created in the roasters by burning gas withsub-stoichiometric amounts of air. Typically, about 10 % of the iron in the ore is also reduced,resulting in a ferronickel alloy containing about 15 % Fe, 80 % Ni, 2 % Co, and otherimpurities. The alloy is expected to form according to the following idealized reactions,equations 3.13 and 3.14 (Monhemius 1987, Kerfoot 1988):OHOFeFeNiHOFeNiO 243232 232  3.13OHOFeFeNiCOOFeNiO 24332 232  3.14Selective reduction of limonite ore is readily achieved at approximately 850ºC. Thenickel that is left un-reacted is insoluble in the subsequent leaching step. To minimize nickellosses in the leach residue, a careful temperature control is required to ensure that all the nickelhas been reduced. Heating rates of 5ºC/min or less require residence times of about 90 minutesin the roaster. Large multiple-hearth roasters provide the opportunity for close temperaturecontrol, and are universally used in Caron process plants. Is important to mention that forsaprolite and gardenite ores the reduction stage is more difficult, because in the range between750 to 850ºC, amorphous magnesium silicates re-crystallize to form fosterite (Mg2SiO4) andany nickel that has not been reduced to metal may stay in the fosterite lattice (Monhemius1987, Kerfoot 1988).iii. LeachingThe hot reduced ore leaving the roasters is cooled down to between 150 and 200ºC undera reducing atmosphere and discharged into a quench tank containing the NH3/(NH4)2CO323solution.  This solution, which is recycled from the leach residue washing circuit, containsabout 6.5 wt % NH3, 3.5 wt % CO2 and 1 wt % Ni. Leaching is carried out in a series ofaerated and agitated tanks, where the Fe-Ni alloy is oxidized and dissolved according toequation 3.15 (Monhemius 1987, Kerfoot 1988, Jandova 1994):  OHNHFeNHNiOHNHOFeNi 4)()(8 223263232 3.15Cobalt behaviour is identical to nickel in this reaction. Although the solution is at a pHof about 10, hydrolysis of nickel and cobalt is prevented by their strong affinities for dissolvedammonia, which favours the formation of ammine complexes.  Iron, on the other hand,although initially dissolving as ferrous ammine complexes, is rapidly oxidized to the ferricstate, which hydrolyses and precipitates as ferric hydroxide, eventually leaving the process inthe leach residue, however it has also been observed to passivate unless high amounts ofdissolved oxygen are present as shown in equation 3.16 (Monhemius 1987, Kerfoot 1988,Nikoloski 2003-2006, Nicol 2004, Anderson 2009):3322223 8)(428)(4 NHOHFeOHOHONHFe   3.16Following leaching, the pregnant leach solution is separated from the barren solids bycounter current decantation and washing in a series of thickeners, and further clarified byfiltration (Monhemius 1987, Kerfoot 1988).iv. Metal recovery from solutionRecovery of nickel and cobalt from the clarified ammonia leach liquors is typically donein two separate steps (Monhemius 1987, Kerfoot 1988): Sulphide precipitation of mixed CoS/NiS by-product Precipitation of basic nickel carbonate (BNC)Pregnant leach liquor is mixed with either H2S gas or a (NH4)2S solution, preferentiallycausing precipitation of cobalt sulphide, CoS, by equations 3.17 and 3.18:24  HCoSSHCo 222 3.17  4242 )( NHCoSSNHCo 3.18However, even though the Ni/Co ratio in solution is about 20:1, the precipitation ratio ofNiS to CoS is actually about 2:1.  This mixed precipitate is thickened, washed, and sold as is(Monhemius 1987, Kerfoot 1988).The Co free solution then goes into the Ni precipitation stage.  Here, NH3 and CO2 areremoved from solution by steam heating in stripping stills, causing precipitation of basic nickelcarbonate (BNC) according to equation 3.19 (Monhemius 1987, Kerfoot 1988):32223223263 3034*)(3*275)(5 NHCOOHOHNiNiCOOHCONHNi   3.19The gases are absorbed in water and recycled to the leaching circuit as ammoniumcarbonate solution. The precipitated solids are thickened and filtered (Monhemius 1987,Kerfoot 1988).The BNC filter cake is treated by one of three methods: Calcination to drive off CO2 and H2O, and sintering to give the final NiOproduct Re-dissolution in (NH4)2SO4 solution and precipitated with H2 gas as Nipowder.3.6.3 General comparison of ammoniacal and acid routesBoth treatment media for nickel laterites present advantages and disadvantages. In Table3.3 both alternatives can be compared on the basis of their technical aspects (Monhemius1987, Kerfoot 1988).25Table 3.3: Comparison of ammonia and acid routes for treating nickel laterites.Reductive roasting- Ammonia leaching Sulfuric acid pressureleachingOreHigh magnesium bearing ore will tend toform fosterite during roasting and lowernickel recovery ((Mg,Ni)2SiO4)Can accept 3-10 % magnesium.Less than 3% magnesiumwould need additionalneutralizing agent, while highmagnesium will consume moreacid.MetalExtractionNickel extraction is around 80 % for lowmagnesium ore, while cobalt extraction isabout 55 %Both nickel and cobalt can beextracted up to 95 % or moreEffluentEffluent contains small quantities of freeammonia, which needs to be dilutedbefore disposalEffluents contain dissolvedmetals like magnesium,manganese, aluminum, whichneed to be treated or diluted anddischarged into the seaEnergyConsumption ~630,000 MJ/ton of nickel. ~125,000MJ/ton of nickel.OperatingCostPressure leaching will have 40-50 % less operating cost than ammonialeaching, if credit for cobalt is considered.Usage of ammonia in the Caron process is limited to limonite and mixed ores with >35% of iron, this is a very sensitive process from the point of view of mineral composition andrequires careful mining and blending. On the other hand, the acid medium reports higherrecoveries; however, the low selectivity of acid, leads to leaching of other elements present inthe ore which increases acid consumptions, making it necessary to study other alternatives thatare more economically sustainable.From the above, alkaline solutions can be an attractive focus of research. The ammoniamedium presents several advantages such as high selectivity, low cost, low toxicity and it ispossible to regenerate it, but the iron content in limonitic ore (ferric state) is not directlysoluble in ammonia medium then it is necessary to use a reducing agent. Few studies havereported results in ammoniacal leaching using reducing agent. This work will show kineticleaching results in ammoniacal solutions using ferrous sulphate as a reducing agent. First, tounderstand ammoniacal leaching, it is necessary to study the thermodynamics features of thissystem.263.6.4 Ammoniacal leachingWhen studying ammoniacal systems, it is possible to divide it in two brief literaturereviews; first a description of the main studies and applications of ammonia as a lixiviant andsecondly, its main thermodynamic features.3.6.4.1 Studies and main applications of ammoniaAmmonia has a capacity of acting as a complexing agent. It surrounds a central ion thatcan be positive or negative, or even a neutral molecule. This feature means that metallic ionscan be kept and stabilized in solution (Park 2007).Ammonia has been widely used in a number of hydrometallurgical processes for manyyears due to its inherent advantages over alternative reagents. The main advantage is that thebasic leach solution decreases several corrosion problems encountered in acidic systems.Ammoniacal leaching minimizes the possibility of major wasteful components like iron, bytheir removal during the leaching step as insoluble oxy/hydroxyl compounds. This allowsselective extraction of the valuable metals (e.g. copper, cobalt, nickel) as soluble amminecomplexes, leading to higher solubility in most cases. Ammonia is also advantageous sinceammonia and ammonium act as a pH buffer in solution (Park 2007).Literature reveals that a considerable amount of research has been conducted on theammoniacal leaching of Cu, Ni and Co bearing oxide and sulphide ores, industrial scraps andwastes, over the past few decades.  Numerous review articles on such applications can befound in the literature (Kerfoot 1988, Arbiter 1993, Han 1993). The thermodynamics andreaction kinetics of the systems involving ammonia-copper, ammonia-nickel and ammonia-cobalt have been investigated in detail (Williams 1978, Boateng 1978, Vu and Han 1977).Extensive studies on the recovery of Cu, Ni and Co from ocean floor manganese nodules byreductive ammoniacal leaching technology have been reported (Han 1974, Das 1986-1995-1997, Acharya 1987-1989, Mukherjee 2005). Also there are studies on the transformation ofsynthetic goethite to magnetite in ammonia by Mohapatra and Das et al., 2002.It is important to mention that oxidative ammoniacal leaching has received specialattention, due to its potential applications. Oxygen, being the least expensive, is the most27widely studied of the oxidants. Several ammoniacal pressure leaching processes have beencommercialized successfully around the world.  As far back as 1947, direct ammonia leachingof Cu, Ni and Co sulphide ores employing high pressure and temperature in an autoclave withoxygen as an oxidant was developed by Sherrit-Gordon at its plant in Saskatchewan, Canada(Forward 1953). The Arbiter Process (Kuhn 1974), developed in 1970’s, is another example ofammonia as a lixiviant at 34.5 kPa with oxygen in a reaction vessel to extract Cu from copperconcentrates (Park 2007).Numerous studies have used ammonia as a lixiviant, but in order to understand theadvantages of this medium, it is necessary to study its thermodynamic foundations.3.6.4.2 Thermodynamic features of ammonia mediaThe molecular structure of ammonia reveals that ammonia consists of three hydrogenatoms bonded to one nitrogen atom. The internode angle is 107o which is close to thetetrahedral angle. Ammonia is colorless in solid, liquid and gaseous states. Its importantphysical constants are listed in Table 3.4 (Meng 1996).Ammonia is very soluble in water. For example a saturated solution containsapproximately 45 % by weight of ammonia at the freezing temperature of the solution andabout 30 % by weight of ammonia at standard conditions (Meng 1996).Ammonium dissolved in water forms a strongly alkaline solution of ammoniumhydroxide. The high solubility of ammonia in water is the result of the tendency of the twosubstances to interact (ammonia - ammonium) with each other through hydrogen-bondformation (Meng 1996).28Table 3.4: Properties of ammonia (Meng 1996)Properties Values UnitsMelting point -77.74 oCBoiling point -33.42 oCΔH (fusion at melting point) 5662 J/moleΔH (vaporization at boiling point) 233370 J/moleCritical temperature 132.90 oCCritical pressure 11.38 MPaDielectric constant (liquid at 60oC ±10oC) 26.70Density (liquid at -70oC) 0.72 g/cm3Density (liquid at -30oC) 0.68 g/cm3ΔHfo, 25oC -46.23 kJ/moleΔGfo, 25oC -16.65 kJ/moleEquilibrium constant of formation (log k) 2.91Entropy, So (25oC) 192.67 J/o moleHeat capacity, Cpo, 25oC 35.69 J/o moleViscosity (liquid at 25oC) 0.0013 poiseVapor pressure (liquid at -20oC) 190.22 kPaVapor pressure (liquid at 0oC) 429.94 kPaVapor pressure (liquid at 20oC) 857.05 kPaAmmonium sulphateWhenever an ammonium salt is dissolved in water, an equilibrium is established betweenammonium ions and aqueous molecular ammonia (NH3). In the case of (NH4)2SO4 theequilibrium reaction is shown by equation 3.20(Tromans 2000):3.20  HNHNH aq )(34    104)(3)298( 10*63.54 NHHNHK aqKNH29Ammonium hydroxideWhen gaseous ammonia is dissolved in water, an equilibrium is established betweenNH3aq and NH4+ as shown by equation 3.21 (Tromans 2000):3.21Dissolved ammonia in laterite leach liquors can be found in three forms: free neutralammonia molecules (NH3), ammonium ions (NH4+) and metal ammine complexes(Me(NH3)n2+). The concentrations of these aqueous species are independent and are describedby definitive equilibrium constants (Osseo-Asare 1979a).The ammonia-ammonium concentrations depend on pH, where the buffer point is 9.25 at25ºC. Below pH 9.25, free ammonia exists in solution mostly as NH4+, while above this pH,the neutral molecule NH3 predominates (Osseo-Asare 1979a).It is important to mention that ammonia and water are analogous molecules. Both formhigh associated liquids with hydrogen bonding between molecules, and both liquids areionizing solvents for inorganic electrolytes, including salts of cobalt, nickel and copper(Nicholls, 1979). Thus, a solution consisting primarily of water and ammonia is expected to actas a mixed solvent for oxygen. This is important because the presence of oxygen could bedetrimental when it is otherwise desirable to work in reducing conditions. Due to this fact,oxygen solubility will be analyzed in detail later according to (Tromans 2000).Ammonia (NH3) concentration depends mainly on pH. Figure 3.10 shows the speciationdiagram for ammonia and ammonium at 25oC. This graph shows that over the buffer point(pH= 9.25 at 25ºC) ammonia is the more stable species whereas below this point, ammoniumconcentrations are higher.  Figures 3.11, 3.12 and 3.13 show nickel, cobalt and iron speciationdiagrams (Asselin 2008).  OHNHOHNH aq 42)(3     5)(34298 10*78.1)(3 aqKNH NHOHNHKaq30Figure 3.10: Ammonia speciation diagram at 25oC (Asselin 2008)Figure 3.11: Nickel ammines speciation diagram at 25oC, 0.1 M Ni, 6 M total NH3 (Asselin2008).31Figure 3.12: Cobalt ammines speciation diagram at 25oC, 0.1 M Co, 6 M total NH3 (Asselin2008).Figure 3.13: Iron ammines speciation diagram at 25oC, 10-3 M Fe, 6 M total NH3 (Asselin2008).Ammonia solutions have relatively high vapour pressure. The relationship betweentemperature and the vapour pressure of ammonia in solutions is often considered by32metallurgical process designers. Table 3.5 lists the vapour pressure of ammonia solutions withvarious ammonia concentrations at temperatures up to 121oC. The predominance of ammoniais increased with pH, as it was shown in Figure 3.10, thus its volatility would be increased. Athigh pH it is expected that ammonia leaves the solution as gas. The solubility depends mainlyon temperature and pressure. Figure 3.14 shows ammonia partial pressure at 80oC with respectto a solution containing different ammonia concentrations (Meng 1996).Table 3.5: Partial pressure (kPa) of aqueous solutions of NH3.NH3 partial pressure in NH3 solutionT (oC)[NH3] (m) 2,9 6,2 9,8 13,9[NH3] wt (%) 4,74 9,5 14,29 19,1T (K) Partial pressure (kPa)0 273 1.79 3.59 6.21 10.44 277 2.28 4.55 7.86 13.210 283 3.24 6.14 10.3 17.416 289 4.27 8.20 13.8 22.121 294 5.72 10.5 17.9 29.527 300 7.17 13.7 23.0 37.632 305 9.38 17.4 29.3 47.438 311 11.9 22.1 36.8 59.343 316 14.8 27.6 45.9 73.149 322 18.4 34.1 56.6 90.354 327 22.6 42.0 69.6 11060 333 27.4 51.1 84.1 13266 339 33.0 61.5 101 15971 344 39.2 73.8 121 19077 350 46.5 87.6 144 22382 355 54.5 103 170 26388 361 63.6 121 199 30793 366 73.8 141 231 35699 372 84.8 163 268 412104 377 96.5 188 308 472110 383 110 214 352 538116 389 123 244 400 614121 394 139 276 453 69633Figure 3.14: Ammonia partial pressure at 80oC with respect to different ammoniaconcentrations (Handbook of Chemistry and Physics 2004)Changes in pressure have a small effect on the solubility of the solute, either solid orliquid, since both are difficult to compress, but in the case of gases, it is different, becausethese are very easy to compress, then their solubility can be enhanced when pressure isincreased.Ammonium salts are commonly used in industry. When an ammonium salt is dissolvedin water the pH obtained is approximately 6, then the ammonium ion is more stable (Osseo-Asare 1979a).  Ammonium sulphate presents high solubility, which increases withtemperature. Figure 3.15 shows the ammonium sulphate solubility at different temperatures.When ammonium hydroxide is dissolved in water it presents a pH over 12, thus it isexpected that ammonia molecules are predominant (Osseo-Asare 1979a). If only ammoniumhydroxide is used, high evaporation of NH3 would be expected due to its high pH. A differentscenario would be expected if only ammonium sulphate is used; in this case, the pH is belowthe NH4+/NH3 buffer point, then free ammonia would not be enough to complex metals. Thus,the mixture of both reactants could deliver better conditions for the stability and solubility ofthe ammonia, and hence, of dissolved metals.34Figure 3.15: Ammonium sulphate solubility at different temperatures (Data Handbook ofChemistry and Physics 2004)It has been observed that ammoniacal solutions have a high selectivity for certain metals.Copper, nickel and cobalt form very strong ammine complexes, as opposed to magnesium,manganese (2+) and iron (2+) (Osseo-Asare 1979a).To understand the behaviour of metals in ammoniacal solutions it is helpful to usePourbaix diagrams. These diagrams are maps that summarize chemical and electrochemicalequilibria. In 1973, a one century review of ammonia hydrometallurgy was presented by Engeland Hewedi. These authors calculated pH-potential relationships in metal-ammonia-ammonium carbonate solutions. Later, Power and Geiger also drew attention to the need fordeveloping phase diagrams for ammonia systems. In 1979-1983, Osseo-Asare estimated hightemperature thermodynamic data for Metal-NH3-H2O systems from 25-300ºC and Osseo-Asare and Fuerstenau presented a summary of the 25ºC equilibria for the systems Cu, Ni, Co-NH3-H2O (Osseo-Asare 1979a).The Eh-pH diagram for the nickel-water system is shown in Figure 3.16. This diagramindicates that nickelous oxide dissolves only if the pH is less than 6. However, as shown byFigures 3.17, 3.18 and 3.19, addition of ammonia introduces an ammine predominance areacentered around pH 9.25 at 25oC. It is the presence of this ammine stability region that makesit possible to leach reduced laterites with ammonia-ammonium solutions. For cobalt and iron,35the behaviour is similar. This is shown in Figures 3.17, 3.18, and 3.19 (Osseo-Asare 1979a,Asselin 2008).Figure 3.16: Eh-pH diagram Ni-NH3-H2O system at 25oCFigure 3.17: Ni-NH3-H2O Eh-pH diagram, 0.1 M Ni, 6 M total NH3 at 25oC (Asselin 2008).141210864202.01.51.00.50.0-0.5-1.0-1.5-2.0Ni - H2O  - System at 25.00 CC:\HSC5\EpH\Ni25.iep         pHEh (Volts)H 2O  L imitsNiNiONi(O H)3Ni(+2a)36Figure 3.18: Co-NH3-H2O Eh-pH diagram, 0.1 M Co, 6 M total NH3 at 25oC (Asselin 2008).Figure 3.19: Fe-NH3-H2O Eh-pH diagram, 10-3 M Fe, 6 M total NH3 at 25oC (Asselin 2008).In practical hydrometallurgical cases, anions such as CO32-, SO42- must be present.Thus, Osseo-Asare and Asihene studied Metal-NH3-H2O-SO4 systems. The use of ammoniumsalts adds other ions to the system. When sulphate ions are present, they affect the equilibriumand move the stability regions of metals. From the findings made by Osseo-Asare it is possibleto conclude that the Ni2+ and Co2+ regions are occupied by the ammonium double salt(NiSO4*(NH4)2SO4 and CoSO4*(NH4)2SO4). These compounds are formed below the bufferpoint of NH3/NH4+ (Osseo-Asare 1979a).373.7 Proposed leaching mechanism with ferrous sulphateThe proposed mechanism for laterite reductive leaching with ferrous sulphate isdescribed, where the overall chemical reaction can be written as seen in equation 3.22 for thereduction of goethite, which is commonly referred to as the main iron phase present in limonite(Zuniga 2013):)(424)(43)(3)(4)( )(22 aqsaqaqs SONHOFeNHFeSOOHFeO  ΔG° = -62.3 kJ/mol 3.22Where the main solid product is magnetite, which is a more reduced iron state thangoethite. This reaction is thermodynamically favourable as seen from the Gibbs free energy ofreaction value (∆Gº is negative).Other poorly defined phases containing manganese have been reported to be present inlimonites, such as FeMn-Wad which can contain nickel and cobalt as well. Chemically thisphase is expressed as K(Mn4+,Mn3+,Fe3+)8O16 or simplified as K(MnFe)8O16.Similarly asbolane has also been reported and is chemically expressed asMn(O,OH)4*nH2O or (Ni,Co)2-xMn4+(O,OH)4*nH2O or simplified as MnO2. These two phasescan contain nickel and cobalt in their structure as well as iron. Proposed reactions with ferroussulphate and ammonia are shown in equations 3.23 to 3.28  24)(432)(3)(43 )(4882 aqSONHOFeOHNHFeSOFe saqaq ΔG° = -384 kJ/mol 3.23In the previous reaction iron present originally as ferric and precipitates as magnetite andin the following two reactions, manganese is shown as being realised and ready to form asoluble ammine, or as forming a solid precipitate, which can instead, in reality, be adsorption.  2 )(4)(4)(432 )(2)(3)(4)(2 91638169 aqaqsaqaqaqs SONHOFeMnOHNHFeSOMnO ΔG° = -278kJ/mol3.24  )(424)(43)(432)(3)(4)(2 6261263 aqssaqaqs SONHOFeOMnOHNHFeSOMnO  ΔG°= -653kJ/mol3.2538In ammoniacal media, many reducing agents such as ferrous sulphate, glucose,thiosulphate, etc., have been tested to recover nickel and cobalt especially from deep seamanganese nodule ores. The best results have been reported when ferrous sulphate was usedwith extractions of over 90% achieved for nickel and cobalt (Peek 2009). Also, there are somestudies where synthetic goethite is directly reduced by ferrous sulphate in ammonia to producemagnetite, where the total transformation is achieved at around 130ºC (Park 2007). Recentstudies used metallic iron as a reducing agent to leach limonites. The results show a partialleach of the ore where the ferrous ion seems to play an important role on leaching kinetics(Meng 1996, Peng 2005, Wang 2008).It is important to mention that the leaching reactions show ammonium sulphate as one ofthe products, which can eventually build up if the solution is re-circulated. This couldpotentially be solved by having a bleed stream that removes some of the excess ammoniumsulphate which can later be crystallized, which can later be sold as a by-product.The advantage of having a reaction using ferrous sulphate is that it is produced in largequantities in China in the production of titanium dioxide. This could be beneficial as ferroussulphate in that process is a waste product. Also, the magnetite formed could be used as a by-product (Zuniga 2009, 2010 & 2013).When nickel and cobalt are released, these can form ammines, the reactions that can takeplace are:  23)(3)(2 )( naqaq NHNinNHNi 3.26  23)(3)(2 )( naqaq NHConNHCo 3.273.28Based on the Pourbaix diagrams for Ni-NH3-H2O and Co-NH3-H2O, ammines could varyfrom mono to hexa-ammines for nickel and cobalt (i.e. 1 ≤ n ≤ 6) (Asselin 2008, Osseo-Asare1979a, 1981a, 1981b, Isaev 1990, Zhong 1995).  23)(3)(2 )( naqaq NHMnnNHMn394 EXPERIMENTAL PROCEDURES4.1 Sample preparationThe samples were prepared as follows: 14 kg of limonitic laterite from New Caledoniacontaining approximately 37 wt % water were air dried. Limonitic laterites commonly exhibita bimodal size distribution with a large proportion of fines particles and small proportion ofcoarser ones. Particle size distribution was analyzed using a Malvern unit with ultrasound plusa dispersant to avoid natural agglomeration, where 100 % of particles were below 133 µm,with 80 % of particles in the feed below (F80) approximately 32 microns. Also, using TescanIntegrated Mineral Analysis (TIMA) microscopy, the grain size was estimated for a globalsample as well as for each main phase, so for the total head sample 80 % of particles werebelow approximately 150 µm . The coarser particle size distribution estimation form TIMAcould be due to natural agglomeration on the particles and perhaps limited samplerepresentativity due to its small size. For the leaching test the sample was homogenized byconing and quartering and divided in equal representative samples of approximately 100 geach.4.2 Leaching testsExperiments were done batch-wise in a 1.5 L glass jacketed cell with baffles andperformed under atmospheric pressure. The chemical reagents were of analytical grade andwere used as received including ferrous sulphate hepta-hydrate with a purity of 99.9 %, argongas 99.999 % pure, ammonium hydroxide from a 28 to 30 wt. % solution and ammoniumsulphate (or ammonium chloride) 99.9 % pure. Deionized water was used to make the leachsolutions. Then ferrous sulphate was dissolved in the mixture, the volume adjusted andsolution heating started. Temperature was controlled with a water bath thermostat.  The pHwas set above the buffer point between NH4+ and NH3(aq) (which changed with temperature), toensure the stability of aqueous ammonia, nickel, cobalt and iron ammines (50°C at  pH 8.7–9.2, 60°C at  pH 8.4–9.0, 70°C at pH 8.2–8.9  and  80°C at pH 8.0–8.8). The pH was alsoadjusted to reduce the influence of metal ion adsorption (especially cobalt) onto iron oxides40and hydroxides according to recommendations found in the literature (Esmadi 1995, Osseo-Asare 1979b, 1980a, 1980b, 1987). An Aplikon gel filled electrode was used to measure pHwhile a standard Ag/AgCl electrode against platinum was used to measure ORP and agitationwas done with an overhead pitched blade stirrer. After reaching the desired conditions thelaterite sample was fed to the reactor. The experiments were carried out under an argon ornitrogen gas atmosphere, to decrease the presence of oxygen and favour the formation of ironammines, rather than hydroxides, and a water chilled glass condenser was used to decreasesolution volume losses. Samples were taken throughout each test (14 ml approximately ofpulp). It is important to mention that, when the samples are not under reducing conditions theformation of iron hydroxides is favoured and magnetite precipitation could continue, thenconventional separation procedures are not applicable in this case. The samples were separatedquickly using a centrifuge followed by filtering using 1 micron particle size retention paper.The clean solution was then acidified with 2 M hydrochloric acid to avoid or minimize ferrichydroxide precipitation, and then it is further diluted in a 1 % w/w HCl solution and finallyanalyzed by atomic absorption spectroscopy (AAS). The solids, after drying, were ground,homogenized and sent for chemical assay (ICP).Different variables were studied including the effect of agitation, solids content,temperature, pH and total ammonia concentration, ferrous sulphate concentration andcomparison between ammonium sulphate vs. ammonium chloride. Oxidation-ReductionPotential (ORP) and pH were recorded for each sample point.The effect of agitation was tested at 600, 850 and 1200 RPM, while keeping thefollowing variables constant: Total ammonia/ammonium concentration at 140 g/l, a FeSO4 tolaterite ratio (g/g) of 0.738, an initial sample mass of 60 g, an F80 of 32 μm, solution volume of500 ml and 80ºC working temperature and a pH value of approximately 8.0.The effect of solids content was tested at 4, 11, 17 and 23 % w/w, while keeping thefollowing variables constant: Total ammonia/ammonium concentration at 140 g/l, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 150 μm, solution volume of500 ml and 80ºC working temperature and a pH value of approximately 8.0.41The effect of temperature was tested at 20, 45, 60 and 80ºC while keeping the followingvariables constant: Total ammonia/ammonium concentration at100 g/l, initial mass sample of60 grams, a FeSO4 to laterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32μm, a solution volume of 500 ml and a pH value adjusted to the ammonium/ammonia bufferpoint corresponding to each temperature.The effect of pH was tested between 5.5 and 8.8 while keeping the following variablesconstant:Ammonium sulphate: Total ammonia/ammonium concentration at 100 g/l, initial masssample of 60 grams, a FeSO4 to laterite ratio (g/g) of 0.738, stirring velocity of 600RPM, an F80 of 32 μm, a solution volume of 500 ml and 80°C working temperature.Ammonium chloride: Total ammonia/ammonium concentration at 140 g/l, initial masssample of 60 grams, a FeSO4 to laterite ratio (g/g) of 0.738, stirring velocity of 600RPM, an F80 of 32 μm, a solution volume of 500 ml and 80°C working temperature.The effect of total ammonia/ammonium concentration was tested at 140, 100, 90 and 64g/l in ammonium sulphate and 185, 162, 100 and 75 g/l in ammonium chloride, while keepingthe following variables constant: Initial mass sample of 60 grams FeSO4 to laterite ratio (g/g)of 0.738 or 1.485, stirring velocity of 600 RPM, an F80 of 150 μm, solution volume of 500 mland 80ºC working temperature and a pH value of approximately 8.0.The effect of ferrous sulphate was tested at 0.738 (100 % stoichiometric with respect toequation 3.23), 1.485 (200 % % stoichiometric with respect to equation 3.23) and 2.734 (300% % stoichiometric with respect to equation 3.23) FeSO4/Ore mass ratios. The rest of theconditions were as follows in the presence of ammonium sulphate and ammonium chloride:Ammonium sulphate: Total ammonia/ammonium concentration at 140 g/l, initial masssample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and a solutionvolume of 500 ml and 80°C working temperature.42Ammonium chloride: Total ammonia/ammonium concentration at 185 g/l, initial masssample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and a solutionvolume of 500 ml and 80°C working temperature.Figure 4.1 shows a schematic of the experimental setup including the glass jacketedreactor with heater, reagents, condenser, pH and ORP measurement.Figure 4.1: Experimental setup4.3 Solutions analysisThe concentrations of nickel, cobalt and iron were analyzed using atomic absorptionspectroscopy (AAS) and ICP.  For atomic absorption analysis, samples were diluted into a 1 %w/w hydrochloric acid solution.4.4 Solids analysisTo analyze the leach residues it was necessary to dry them at room temperature thenhomogenize them in order to take representatives samples. The samples were analyzed by ICP,43X-Ray Diffraction (XRD), scanning electron microscopy (SEM) and Tescan IntegratedMineral Analyser (TIMA).4.5 Ore characterizationThe limonitic laterite is a complex ore of nickel, cobalt and manganese. Usually manyiron oxides are present and the important metals are distributed throughout the differentphases, thus, different techniques were used to characterize feed and residue samples.The main elements present in the laterite are given Table 4.2 from ICP assays. Aquantitative phase analysis (QXRD) was carried out for the head sample and the results aresummarized in Table 4.3. Deeper and more sensitive analyses were necessary to identify themain phases to which nickel, cobalt and manganese are associated.  For this purpose thesample was prepared as polished block sections and examined using the Tescan IntegratedMineral Analyser (TIMA) equipped on a Tescan Vega 3 Scanning Electron Microscope whichincorporates an SEM equipped with the Oxford Inca Energy Dispersive X-ray (EDX)spectrometer and semi-quantitative elemental analysis software.The TIMA system incorporates the simultaneous acquisition of both Back ScatterElectrons (BSE) and EDX spectra to detect and analyze the mineral composition of a preparedsample. The acquisition time at each pixel is adjusted based upon the BSE & EDX signalstrength, significantly shortening data acquisition times. This technique offers superiorimaging and mineral detection with the use of an Yttrium Aluminum Garnet scintillator BSEdetector allowing for accurate mineral detection of ultra-fine-grained particles. Mineralabundance, mineral liberation and association analyses are carried out using this instrumentwhich are summarized in the Appendices.  Magnetite and hematite are difficult to reconcileusing the SEM so the ratios of hematite and magnetite were determined using the OpticalImage Analysis system equipped on the Leica petrographic microscope. The main phaseanalyses are shown in Tables 4.2 and 4.3 while metals deportment by mineral are summarizedin Table 4.4 and Figure 4.2. All of the TIMA work was done by Geoff Lane fromMineralogical Consultants in Maple Ridge, BC.44Table 4.1: Chemical analysis of the tested laterite (ICP)Elements Content (mass %)Iron 51.8Nickel 1.49Cobalt 0.33Manganese 2.29Table 4.2: Quantitative phase analysis (wt %) using XRDMinerals Ideal Formula Laterite (%)Goethite α-Fe 3+O(OH) 56.8Maghemite γ-Fe2O3 13.5Hematite 5.70Gibbsite Al(OH)3 1.30Kaolinite Al2SiO5(OH)4 2.90Quartz SiO2 0.40Asbolane (Co,Ni)1-y(Mn4+O2)2-x(OH)2-2y+2x nH2O 19.4The results of quantitative phase analysis by Rietveld refinements are given in Table 4.2.These amounts represent the relative amounts of crystalline phases normalized to 100%. Thus,the main phases obtained were goethite, maghemite and hematite. As the crystal structure forasbolane is unknown, published cell dimensions and hexagonal space group was used to fit thepeaks. Such a model (hkl phase) does not contain information about the atoms and thereforethe amount of asbolane cannot be directly determined.  However, by adding a known weight ofa well crystallized internal standard (CaF2) to the sample, the asbolane may be determined asan “amorphous component”. Note that the amount of asbolane would also include any otheramorphous component in the sample and is therefore a maximum value.One of the main objectives of doing mineralogical analyses was to identify the mainphases found in the laterite (XRD) but also trying to confirm or to obtain an indication inregards to what phases nickel, cobalt and manganese are associated to in the feed and leachresidues. From Tables 4.2 (XRD analysis) and 4.3 (TIMA analysis), some differences arefound, particularly in the amount goethite and a new phase named FeMn-Wad. Results of45mineralogical analyses can vary, particularly with materials like limonite that contain difficult,poorly crystalline, clay-like materials. Also, XRD will not detect amorphous phases or thosewith poor long range order (nano-sized crystals).  Trace elements are also better detected withTIMA. The sample size for TIMA is considerable smaller than for XRD. Furthermore, theTIMA results are difficult to match with the chemical assays, particularly in the case ofpotassium coming from FeMn-Wad, however, these phases can be non-stoichiometric andtherefore do not necessarily match their simplified chemical formulas. This can and should bestudied further, however, TIMA has provided valuable information that can help elucidate howcobalt and manganese (and to a lesser extent nickel) can co-precipitate or be adsorbed duringleaching tests.Table 4.3: Quantitative phase analysis (wt %) using TIMAMinerals Ideal Formula Laterite (%)Goethite α-Fe3+O(OH) 31.10MnFe-Wad K(Mn4+,Mn3+, Fe3+)8O16 30.00Hematite α-Fe2O3 22.50Magnetite Fe3O4 0.18Goethite_low Ni α-Fe3+O(OH) 10.60Asbolane (Co,Ni)1-y(Mn4+O2)2-x(OH)2-2y+2x nH2O 1.73Quartz SiO2 1.33Feldspar 0.54Other Silicates 0.47Other Ni 0.02Mica 0.02Olivine 0.01Pyroxene 0.01Other minerals 1.4346Table 4.4: Metal deportment by mineral in head sample (from TIMA)Phase Feed (%)Ni Co MnMagnetite or Wustite 0.00 0.00 0.00Goethite 5.00 0.00 0.00MnFe Wad 73.0 76.0 75.0Hematite 3.00 0.00 0.00Goethite-low Ni 11.0 0.00 1.90Asbolane 8.00 24.0 20.0Others 0.00 0.00 2.99Others Ni 0.00 0.00 0.90Total 100 100 100Figure 4.2: Metal deportment by mineral in head sample (from TIMA)From the TIMA results, phase distribution and distribution of metals of interest weredetected. The New Caledonia head is composed of complex intergrowths of asbolane, FeMn-wad (cryptomelane), goethite and hematite (martite). The particles complexes are intricatelyintergrown making distinction between phases difficult.  The major Ni and Co bearing phasespresent in the head sample are asbolane and Mn-Wad although Ni and occasional Co wasdetected in goethite and to a subordinate amount in hematite. The content of Ni and Co inthese minerals ranged dramatically, making the identification of particular phases difficult.Overall Mn distribution within the Head sample is present as FeMn-Wad with lower amounts0%10%20%30%40%50%60%70%80%90%100%Ni Co MnOthers NiOthersAsbolaneGoethite-low NiHematiteMnFe WadGoethiteMagnetite or WustiteMetal deportment by phase47present as asbolane and trace amounts present in goethite and hematite.  The hematite presentin the sample is as martite, which is a pseudomorph of magnetite having the oxidation state ofhematite with the crystallographic signature of magnetite.Figure 4.3 (a) and (b) shows the Backscattered Electron Image (BEI) of compositeparticle of goethite, hematite and asbolane in the New Caledonia deposit as typical limonitesample, in particular for this case the Fe/Mn ratio varies in the particles, for example in (a) aparticle rich in Mn is shown, where nickel is also present, but not associated with high ironconcentrations (such as goethite), which is often assumed to be case for limonite, spectra 1 and2 have the highest nickel concentrations. For particle (b) in spectra 1 and 3 the content of Fe ishigher than Mn (Fe > Mn), Spectrum 2 content of Fe = Mn, and Spectrum 4 the content of Mn> Fe. Some differences between QXRD results and TIMA are found in this sample. Morephases are included in the TIMA analysis because limit detection is higher in comparison toQXRD.(a) (b)Figure 4.3: (a) and (b) Backscatter Electron Image (BEI) of composite particle of hematite,goethite, and asbolane in New Caledonia Head.It is difficult to match the chemical assays results to these mineralogy observations,particularly if trying to close the potassium balance, however, these phases are often non-stoichiometric and do not match their simplified chemical formulas.Spectrum 1Spectrum 1Spectrum 2Spectrum 3Spectrum 4Spectrum 2485 RESULTS AND DISCUSSIONLeaching tests considered the study of the following leaching parameters:- Stirring velocity (agitation)- Solids content- Temperature- Total ammonia/ammonium concentration and pH- Ferrous sulphate concentrationThe source of ammonia, as explained in the procedures section, resulted from a mixtureof ammonium hydroxide with ammonium sulphate, or ammonium hydroxide with ammoniumchloride. All variables were, therefore, studied separately for the ammonium sulphate andammonium chloride cases.5.1 Effect of agitationWhen agitation has a considerable effect in leaching processes, in principle, kineticscould be improved by increasing stirring velocity to decrease mass transfer limitations.Whether the required agitation speed to reach such conditions is technically feasible is out ofthe scope of this work. In this case the agitation speed values selected were 600, 850 and 1200RPM in all cases.In the presence of ammonium sulphate nickel, cobalt and manganese appear to dissolvefaster at 850 RPM. The results, however, do not show a clear trend as, for instance, nickel andmanganese show lowest removal values at 1200 RPM while cobalt does at 600 RPM. This isshown in Figures 5.1, 5.3 and 5.5.Figures 5.2, 5.4 and 5.6 show the effect of agitation speeds for nickel, cobalt andmanganese, respectively, in the presence of ammonium chloride. Between 600 and 850 RPMthere is no clear effect on the dissolution of nickel, cobalt or manganese. At 1200 RPM there49seems to be a slight improvement for nickel and manganese, which disappears at the end ofeach test.In summary there does not appear to be a clear improvement as a result of increasingagitation velocity in either ammonium sulphate or ammonium chloride. Reproducibility wasnot confirmed based on the unclear effect of agitation on metal extraction. Perhaps it would beof interest to test the effect of agitation in conditions under which metal extraction is higherthan shown here, without being as highly affected by perhaps adsorption and/or co-precipitation.Figure 5.1:Effect of agitation velocity on nickel removal in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.05101520253035400 5 10 15 20 25600 RPM850 RPM1200 RPMTime (h)Nickel Extraction (%)50Figure 5.2: Effect of agitation velocity on nickel removal in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.Figure 5.3: Effect of agitation velocity on cobalt removal in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.0510152025300 5 10 15 20 25600 RPM850 RPM1200 RPMTime (h)Nickel  Extraction (%)05101520253035400 5 10 15 20 25600 RPM850 RPM1200 RPMTime (h)Cobalt Extraction(%)51Figure 5.4: Effect of agitation velocity on cobalt removal in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.Figure 5.5: Effect of agitation velocity on manganese removal in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.010203040506070800 5 10 15 20 25600 RPM850 RPM1200 RPMTime (h)Cobalt Extraction (%)051015202530350 5 10 15 20 25600 RPM850 RPM1200 RPMTime (h)Manganese Extraction(%)52Figure 5.6: Effect of agitation velocity on manganese removal in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, an initialsample mass of 60 g, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.5.2 Effect of solids contentIn the reductive leaching of limonite using ferrous sulphate, the aqueous ferrous ionsreact with the ferric species in the feed to form magnetite. It has been reported that metals suchas cobalt and/or manganese can precipitate or be adsorbed by magnetite and other iron oxidesand/or hydroxides, which has been observed by Osseo-Asare and Zuniga (Osseo-Asare 1979,Zuniga 2009-2010).A higher amount of solids would require a higher concentration of ferrous sulphate inorder to reduce a larger amount of limonite, which in turn means a larger amount of magnetiteand other iron species in the residue that could potentially lead to adsorption and/or co-precipitation of nickel, and more likely cobalt and manganese. Figures 5.7 to 5.12 show theeffect that different amounts of solids have in the leaching of nickel, cobalt and manganese inthe presence of ammonium sulphate and ammonium chloride.05101520253035400 5 10 15 20 25600 RPM850 RPM1200 RPMTime (h)Manganese Extraction (%)53Nickel appears to be favoured by the increase of solids in the leach in the presence ofammonium sulphate, perhaps due to the increased concentration of reducing agent. Howeverthe opposite is seen for cobalt and manganese, which are favoured by lower solids contents inthe leach. This behaviour is in agreement with previous research that shows that cobalt andmanganese can co-precipitate and/or be adsorbed by iron oxides or hydroxides that will beproduced in larger amounts as the initial amount of solids in the leach is increased (Zuniga2009, 2010 & 2013)The effect of solids content in the presence of ammonium chloride does not seem to havea considerable effect except for manganese. This might be due to the fact that iron (fromferrous sulphate) appears to last longer in solution, thus slowing down the formation ofmagnetite and other iron compounds, which in turn help to adsorb and/or co-precipitate cobaltand manganese. A very large impact is seen for manganese at the highest solids content, whichcould be re-evaluated for confirmation in future work. It is worth noting that due to theprecipitation of ferrous the mass of solids increases throughout the test.Figure 5.7: Effect of solids content on nickel extraction in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.0510152025303540450 5 10 15 20 254% sol11% sol17% sol23% solTime (h)Nickel Extraction(%)54Figure 5.8: Effect of solids content on nickel extraction in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.Figure 5.9: Effect of solids content on cobalt extraction in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.01020300 5 10 15 20 2511% sol17% sol23% solTime (h)Nickel Extraction (%)01020304050607080900 5 10 15 20 254 % sol11% sol17% sol23% solTime (h)CobaltExtraction(%)55Figure 5.10: Effect of solids content on cobalt extraction in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 32 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.Figure 5.11: Effect of solids content on manganese extraction in ammonium sulphate. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 150 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.01020304050607080900 5 10 15 20 2511% sol17% sol23% solTime (h)CobaltExtraction (%)01020304050600 5 10 15 20 254% sol11% sol17% sol23% solTime (h)ManganeseExtraction(%)56Figure 5.12: Effect of solids content on manganese extraction in ammonium chloride. Totalammonia/ammonium concentration at 140 g/l, a FeSO4 to laterite ratio (g/g) of 0.738, stirringvelocity of 600 RPM, an F80 of 150 μm, solution volume of 500 ml and 80ºC workingtemperature and a pH value of approximately 8.0.5.3 Effect of temperatureLeaching tests were carried out at 20, 45, 65 and 80ºC while keeping all other variablesconstant. Tests at higher temperatures were not possible as ammonia evaporation increased tothe point that the concentration in solution was not enough to keep metals in solution.Figures 5.13, 5.15 and 5.17 show the effect of temperature on the dissolution of nickel,cobalt and manganese respectively in a ferrous sulphate – ammonium sulphate solution.Increasing temperature has a clear beneficial effect on nickel removal, reaching a maximumvalue under five hours. For cobalt and manganese the effect of temperature is different andhighest extractions for both metals are reached at 45ºC. This is in agreement with the resultspresented in 2009 for the reductive leaching of limonite from Ivory Coast using metallic iron(Zuniga 2009).010203040500 5 10 15 20 2511 % sol17 % sol23% solTime (h)ManganeseExtraction (%)57It is important to mention that at high temperature and in the presence of ammoniumsulphate, cobalt extraction curves might be misleading, as it is possible that cobalt is initiallyextracted, but is then quickly adsorbed (or co-precipitated) by other iron oxides, especiallymagnetite, which is favoured by temperature. Figures 5.21, 5.23 and 5.24 show extractioncurves for nickel, cobalt and manganese, respectively, under the same conditions as the otherleaching tests in this work (detailed conditions in the experimental procedures section and inFigures captions), but with more dilute slurry containing only 4 % solids by weight. Fromthese curves it is clear that cobalt and to a lesser extent manganese extractions, are favoured byhigher temperatures if other conditions allow for them to remain in solution and exhibit lessadsorption and/or co-precipitation issues. Adsorption and/or co-precipitation of cobalt andmanganese still takes place under these conditions, however, highest extractions are achievedat the highest temperature at low solids loading.Figures 5.14, 5.16 and 5.18 show the effect of temperature on the dissolution of nickel,cobalt and manganese, respectively, in a ferrous sulphate – ammonium chloride solution, withan evident improvement on the removal of the three metals with increasing temperature.However, the three metals show different behaviours, with nickel having a slower startcompared to cobalt and manganese. Nickel removal reaches its highest value after 24 hours ofleaching.  However, the curve does not appear to have “flattened”, which suggests furtherdissolution might still be possible, whereas cobalt reaches its highest value at approximately12 hours, after which the curve seems to “flatten” with no further dissolution taking place.Manganese reaches its maximum extraction even quicker, between 4 to 6 hours, however afterthat, it drops considerably suggesting re-precipitation or adsorption.Cobalt and manganese reach higher extraction values in ammonium chloride, whichappears to lessen the effect of co-precipitation and/or adsorption onto the solid residue. Theremoval of nickel is faster and reaches higher values in ammonium sulphate than inammonium chloride. This can be attributed to the fact that ferrous ions take longer to reactwith the limonite in the presence of ammonium chloride, due perhaps to the fact that highertotal ammonia/ammonium concentrations can be maintained throughout the duration of theleaching test in this medium.  Thus ferrous ions might be able to remain longer in solution58instead of reacting with the limonite to form magnetite. Figures 5.19 and 5.20 show thebehaviour of iron in ammonium sulphate and ammonium chloride, respectively, where it canbe seen that iron tends to stay in solution longer in the presence of chloride.Tables 5.1 and 5.2 show a summary of the main iron phases found in the solid residue,where it is possible to see that ferric iron phases such as goethite and maghemite appear morestable at lower temperatures whereas magnetite appears generally at temperatures higher than65oC.  This suggests an explanation for the tendency of Co & Mn to co-precipitate or beadsorbed at higher temperatures and in the presence of sulphate. The hematite content doesnot change during the process. It is important to mention that the goethite mass in the residuein chloride solutions increases throughout the entire temperature range, while magnetiteformation is more substantial in the sulphate solutions. However, this last comment must beverified by comparing solid residues that presented similar nickel extractions (the magnetiteformation is mainly associated with nickel extraction rather than cobalt and manganese).These results suggest the following reactions could take place during the leaching in bothsolutions:424)(43)(3)(4)( )(2*2 SONHOFeNHFeSOOHFeO saqaqs  ΔG(80°C) = -57.4 kJ/mol 5.1424)(32)(3)(2)(4 )(4)(41084 SONHOHFeOHNHOFeSO saqaqaq  ΔG(80°C) = -533 kJ/mol 5.2OHOHFeOOHFe s 2)(3 *)(  ΔG(80°C) = -19.0 kJ/mol 5.3OHOFeOHFeO mags 2)(32)(*2  ΔG(80°C) = -1.88 kJ/mol 5.4The oxidation of ferrous ions was intended to be prevented or minimized by sparging aninert gas into the leach solution in order to displace oxygen. The initial ORP in the leachsolution without ferrous sulphate was approximately 0 mV vs. SHE. After adding the reductant(ferrous sulphate) this value decreased to between -300 and -500 mV vs. SHE. Despite thesereducing conditions, and oxygen being displaced from the system, the agitation and the systemnot being completely enclosed, perhaps allowed some oxygen to enter the solution and reactwith ferrous ions to form ferric compounds and precipitate, especially at lower temperatures.59As mentioned above, at lower temperatures nickel extraction is poor (especially below80°C), however, the ORP during the tests increased from -500 or -400 mV vs. SHE to valuesin the range of -100 to +100 mV vs. SHE, with low nickel extractions, suggesting that verylittle to no goethite was reduced at these conditions by the ferrous ions. These observationssuggest that ferrous ions are oxidized but not by the reaction with the goethite. Since theamounts of oxygen expected in solution are low, perhaps ferric iron phases are slowly formedduring the leach test, which lasted between 24 to 48 hours and/or during solids filtering andexposure to the environment while drying the leach residues. It is worth noting that when thesolid residues were dried at room temperature, a “reddish” colour layer formed over the blackmagnetite (when present) product shortly after being exposed to the environment.Table 5.1: Solid residue phase analysis (XRD). Effect of temperature in ammonium sulphate.Total ammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4to laterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solutionvolume of 500 ml and a pH value adjusted to the ammonium/ammonia buffer pointcorresponding to each temperature.Sample Mass (g)Goethite Maghemite C Magnetite HematiteHead 34.1 8.1 0.0 3.420oC 49.6 30.2 0.0 1.945oC 43.2 32.4 0.0 2.365oC 35.8 18.0 24.1 2.180oC 18.0 23.3 30.4 3.6Mass (%)Goethite Maghemite C Magnetite HematiteHead 56.8 13.5 0.0 5.720oC 57.7 35.2 0.0 2.245oC 51.5 38.7 0.0 2.765oC 43.2 21.8 29.1 2.580oC 22.4 29.0 37.9 4.460Table 5.2: Solid residue phase analysis. Effect of temperature in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.Sample Mass (g)Goethite Maghemite C Magnetite HematiteHead 34.1 8.1 0.0 3.420oC 48.7 29.1 0.0 2.545oC 45.0 32.3 0.0 1.965oC 45.7 0.0 30.1 4.280oC 45.6 12.4 18.5 2.9Mass (%)Temp Goethite Maghemite C Magnetite HematiteHead 56.8 13.5 0.0 5.720oC 56.4 33.7 0.0 2.945oC 53.0 33.7 0.0 2.365oC 53.9 0.0 35.5 4.980oC 55.1 15.0 22.4 3.5Figure 5.13: Effect of temperature on nickel extraction in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.0510152025303540450 5 10 15 20 2520 C 45 C65 C 80 CTime (h)Nickel Extractionn(%)61Figure 5.14: Effect of temperature on nickel extraction in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.Figure 5.15: Effect of temperature on cobalt extraction in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.01020300 5 10 15 20 2520 C 45 C65 C 80 CNickelExtraction (%)Time (h)(h)01020304050600 5 10 15 20 2520 C 45 C65 C 80 CTime (h)Cobalt Extraction(%)62Figure 5.16: Effect of temperature on cobalt extraction in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.Figure 5.17: Effect of temperature on manganese extration in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.010203040506070800 5 10 15 20 2520 C 45 C65 C 80 CCobalt Extraction (%)Time0510152025300 5 10 15 20 2520 C 45 C65 C 80 CTime (h)ManganeseExtraction (%)63Figure 5.18: Effect of temperature on manganese extraction in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.Figure 5.19: Effect of temperature on iron precipitation in ammonium sulphate. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.0102030400 5 10 15 20 2520 C 45 C65 C 80 CTime(h)ManganeseExtraction (%)0102030400 5 10 15 20 2520 C 45 C65 C 80 CTime (h)Ferrous in Solution(g/l)64Figure 5.20: Effect of temperature on iron precipitation in ammonium chloride. Totalammonia/ammonium concentration at100 g/l, initial mass sample of 60 grams, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.Figure 5.21: Effect of temperature on nickel extraction in ammonium sulphate in dilute slurry(4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 to lateriteratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volume of 500ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding to eachtemperature.0102030400 5 10 15 20 2520 C 45 C65 C 80 CFerrous in Solution (g/l)Time (h)0510152025300 5 10 15 20 2520 C45 C65 C80 CTime (h)Nickel Extraction (%)65Figure 5.22: Effect of temperature on cobalt extraction in ammonium sulphate in dilute slurry(4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 to lateriteratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volume of 500ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding to eachtemperature.Figure 5.23: Effect of temperature on manganese extraction in ammonium sulphate in diluteslurry (4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 tolaterite ratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volumeof 500 ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding toeach temperature.0.010.020.030.040.050.060.070.080.090.00 5 10 15 20 2520 C45 C65 C80 CTime (h)Cobalt Extraction(%)01020304050600 5 10 15 20 2520 C45 C65 C80 CTime (h)Manganese Extraction (%)66Figure 5.24: Effect of temperature on iron precipitation in ammonium sulphate in dilute slurry(4 % solids by weight). Total ammonia/ammonium concentration at100 g/l, a FeSO4 to lateriteratio (g/g) of 0.738, stirring velocity of 600 RPM, an F80 of 32 μm, a solution volume of 500ml and a pH value adjusted to the ammonium/ammonia buffer point corresponding to eachtemperature.5.4 Effect of pH and ammonia concentrationAmmonia is expected to form ammines with nickel, cobalt, manganese and iron. Theiron ammines (ferrous ammines formed from ferrous sulphate added as reductant) should reactwith the limonite forming mainly magnetite while releasing nickel, cobalt and manganese andforming ammines with them as well.The use of ammonia and ammonium salts is known to be a powerful lixiviant used inhydrometallurgical processes. Under reductive conditions the Ni, Fe and Co are released in the+2 oxidation state and stabilized in solution by the formation of ammine complexes with NH3ligands.The stability of metal ammines in solution is pH dependant. The buffer point at 80°C isapproximately 7.8, and according to the literature, higher ammine stability is found at hightemperatures close to the buffer point. Thus, leaching performance was evaluated between pH0246810120 5 10 15 20 2520 C45 C65 C80 CTime (h)Ferrous in Solution(g/l)67values of 5.5 to 8.8 at 80°C in order to establish an optimal pH range for each metal andcompare with the existing literature (Asselin 2008). Absorption and/or precipitation ontooxides as function of pH are well known for cobalt and manganese under alkaline conditions.When oxides are in contact with water their surface acts as an acid or base and this is also afunction of pH. Usually iron oxides (hematite, magnetite, goethite, etc.) have a negativelycharged surface above pH 6.5 and can attract cations from solution through ad/absorptionphenomena to fulfill the electroneutrality condition.  Thus, especially for cations with severaloxidation states, ab/adsorption or/and co-precipitation is more frequent.Figures 5.25 to 5.27 show metals extraction vs. pH for each metal at 80°C using 100 g/lof total ammonium/ammonia sulphate and 185 g/l of total ammonium/ammonia chloride. Fornickel, similar behaviour is observed in sulphate and chloride media, where pH values between7.5 and 8.5 are better for nickel ammines to remain in solution. It is important to mention thatthe buffer point between ammonium and ammonia at 80°C is around 7.8.Cobalt and manganese appear to be more sensitive to pH, with values over pH 8.2appearing to be detrimental for these metals. Specifically, at pH values below 7.5, cobaltshows very low extractions. This is probably due to low free ammonia concentrations at pHvalues below 7.8. Similarly to cobalt, Manganese is sensitive to pH changes, where at lowerpH it remains in solution but with increasing pH the stability appears to decrease, especially insulphate media where the total ammonia concentration was lower than in chloride. Thus, thebest pH found for the solution was close to the ammonia/ammonium buffer point to keep metalammines of interest in solution.Nickel and cobalt are affected differently by the amount of reductant (ferrous sulphate inthis case) and ammonia. Thus, total ammonia concentration was evaluated at differentreductant concentrations. In ammonium sulphate the total ammonia/ammonium concentrationwas changed from 45 to 140 g/l for a ferrous sulphate to laterite ratio of 0.738, and from 90 to140 g/l for a ferrous sulphate to laterite ratio of 1.485. In ammonium chloride the totalammonia/ammonium concentration was changed from 100 to 185 g/l and 75 to 185 g/l forferrous sulphate to laterite ratios of 0.738 and 1.485 respectively. It is important to mention68that the total ammonia/ammonium concentrations in sulphate solutions were always lower thanin chloride at the same pH. Results can be seen in Figures 5.28 to 5.33.In ammonium sulphate at low ferrous sulphate additions, the increase of totalammonia/ammonium concentration improves cobalt and manganese extractions morenoticeably than it does nickel. The opposite is seen for higher ferrous sulphate addition asnickel dissolution improves while cobalt, in fact, decreases with increasingammonia/ammonium concentration. It would appear that this is related to the nature of cobaltand manganese and their tendency to co-precipitate and/or be adsorbed in the presence of highamounts of ferrous sulphate, which leads them to be less stable in solution, whereas nickel isnot affected as much by precipitation or adsorption. Findings indicate that with a ferroussulphate to laterite ratio of 1.485, an ammonia concentration over 100 g/l does not appear to bebeneficial for nickel, as extraction does not improve further.From these results it can be recommended that for cobalt and manganese the totalammonia/ammonium concentration should be 140 g/l or greater, while for nickel it is sufficientto use a total ammonia/ammonium concentration of 100 g/l.In the presence of ammonium chloride and lower ferrous sulphate additions, nickelextraction does not improve with a total ammonia/ammonium concentration between 75 to 185g/l; however, at higher ferrous sulphate concentration and total ammonia/ammoniumconcentration no higher than 100 g/l is sufficient to get a maximum of 68 % of nickel insolution. On the other hand, for cobalt and manganese, higher ammonium/ammoniaconcentrations (around 162 g/l or greater) and ratios of 0.738 or 1.485 of ferrous sulphate/oreadditions, delivered similar cobalt extraction (around 70 %). This fact could suggest that cobaltleaching is more dependant on ammonia/ammonium concentration that reductantconcentration.When comparing both variables together, the results suggest that the reductant additionhas a more important effect than ammonia for nickel extraction. Thus, independent of thechloride or sulphate solution the final extraction for nickel is similar and around 65 to 68 %69extracted using a 1.485 ferrous sulphate to laterite ratio.  However, the kinetics look different;in the presence of ammonium sulphate, kinetics are faster and the maximum nickel extractionis achieved before 5 hours of residence time and after that a flat zone is seen in the extractioncurve.  In ammonium chloride, the maximum nickel extraction is achieved after 24 hours andthe curve looks to continue with a rising trend. These last facts are in agreement with thetemperature and ferrous sulphate effect observations, which show that in the presence ofammonium chloride, magnetite content in the solid residue is lower in comparison toammonium sulphate solutions. For cobalt and manganese, both variables play an importantrole; in the presence of ammonium chloride, it is possible to use relatively high ferroussulphate additions (maximum of 1.485 mass ratio of ferrous sulphate/ore) without adetrimental effect on the cobalt extraction at high ammonia concentrations, while in thepresence of ammonium sulphate, increasing ferrous sulphate additions have a negative impact.Figure 5.25: Effect of pH on nickel extraction in ammonium sulphate and chloride solutions.010203040506070804.5 5.5 6.5 7.5 8.5 9.5NH4Cl - NiNH42SO4 - NiNickelExtraction (%)pH @ 80 oC70Figure 5.26: Effect of pH on cobalt extraction in ammonium sulphate and chloride solutions.Figure 5.27: Effect of pH on manganese extraction in ammonium sulphate and chloridesolutions.010203040506070804.5 5.5 6.5 7.5 8.5 9.5NH4Cl - CoNH42SO4 - CoCobalt Extraction (%)pH @ 80 oC01020304050604.5 5.5 6.5 7.5 8.5 9.5NH4Cl - MnNH42SO4 - MnManganese Extraction (%)pH @ 80 oC71Figure 5.28: Effect of total ammonia/ammonium concentration on nickel extraction inammonium sulphate.Figure 5.29: Effect of total ammonia/ammonium concentration on nickel extraction inammonium chloride0102030405060700 5 10 15 20 25140 g/l - 1.485Red/Ore ratio100 g/l - 1.485Red/Ore ratio90 g/l - 1.485Red/Ore ratio140 g/l - 0.738Red/Ore ratio100 g/l - 0.738Red/Ore ratio64 g/l - 0.738Red/Ore ratio45 g/l - 0.738Red/Ore ratioNickelExtraction(%)Time (h)0102030405060700 5 10 15 20 25185 g/l - 1.485Red/Ore ratio100 g/l - 1.485Red/Ore ratio75 g/l - 1.485Red/Ore ratio185 g/l - 0.738Red/Ore ratio162 g/l - 0.738Red/Ore ratio100 g/l - 0.738Red/Ore ratio75 g/l - 0.738Red/Ore ratioNickel Extraction (%)Time (h)72Figure 5.30: Effect of total ammonia/ammonium concentration on cobalt extraction inammonium sulphateFigure 5.31: Effect of total ammonia/ammonium concentration on cobalt extraction inammonium chloride051015202530350 5 10 15 20 25140 g/l - 1.485Red/Ore ratio100 g/l - 1.485Red/Ore ratio90 g/l - 1.485Red/ore ratio140 g/l - 0.738Red/Ore ratio100 g/l - 0.738Red/Ore ratio64 g/l - 0.738Red/Ore ratioCobalt Extraction(%)Time (h)010203040506070800 5 10 15 20 25185 g/l - 1.485Red/Ore ratio100 g/l - 1.485Red/Ore ratio75 g/l - 1.485Red/Ore ratio185 g/l - 0.738Red/Ore ratio162 g/l - 0.738Red/Ore ratio100 g/l - 0.738Red/Ore ratioCobalt Extraction (%)Time (h)73Figure 5.32: Effect of total ammonia/ammonium concentration on manganese extraction inammonium sulphateFigure 5.33: Effect of total ammonia/ammonium concentration on manganese extraction inammonium chloride.024681012141618200 5 10 15 20 25140 g/l - 1.485Red/Ore ratio100 g/l - 1.485Red/Ore ratio90 g/l - 1.485Red/Ore ratio140 g/l - 0.738Red/Ore ratio100 g/l - 0.738Red/Ore ratio64 g/l - 0.738Red/Ore ratioTime (h)Manganese Extraction (%)0102030405060700 5 10 15 20 25185 g/l - 1.485Red/Ore ratio100 g/l - 1.485Red/Ore ratio75 g/l - 1.485Red/Ore ratio185 g/l - 0.738Red/Ore ratio162 g/l - 0.738Red/Ore ratio100 g/l - 0.738Red/Ore ratioManganese Extraction (%)Time (h)745.5 Effect of ferrous sulphate concentrationThe direct reductive leaching of limonite uses ferrous sulphate as its reducing agentwhich reacts with the ferric-based phases in the limonite to form magnetite while releasingmetals of interest.It would seem logical that increasing reagents would speed up the reaction and improveextractions. However, the leaching system described in this study has shown that nickelbehaves differently to cobalt and manganese, i.e. nickel appears to require higher ferroussulphate concentrations than cobalt and manganese. This may be due to the fact that nickel isbelieved to be distributed more evenly in all of the mineral phases found in the limonite,therefore more reductant (ferrous sulphate) would help its release from the ore. Also nickel hasnot been shown to co-precipitate and/or adsorb onto magnetite and other iron compoundspresent during the leach. It is important to mention that when in presence of an excess offerrous sulphate with respect to stoichiometry (equation 3.23), ferrous appears to continueprecipitating, however further precipitation also takes place during sampling.On the other hand, cobalt and manganese are known to have these issues with ironoxides and hydroxides (Zuniga 2009-2010), but they seem to be able to leach quicker thannickel, particularly in the presence of ammonium chloride with lower amounts of ferroussulphate. The fact that less reagent still enables cobalt and manganese to be leached agreeswith the observation that these two metals are associated mainly with MnFe-Wad andAsbolane and not as evenly distributed in the ore as nickel (MnFe-Wad, asbolane, goethite,hematite, etc.). It appears that ferrous sulphate takes longer to react with the limonite whenammonium chloride is present at high total ammonia/ammonium concentration, thus slowingdown the removal of nickel.Nickel was always positively affected by increasing ferrous sulphate additions.However, ferrous sulphate to laterite ratios over 1.485 do not result in an improvement inextractions in sulphate media, using approximately 140 g/l total ammonia/ammoniumconcentration achieving approximately 65 % nickel extraction in 5 hours. In ammoniumchloride, nickel extraction improves with increasing reductant additions at 185 g/l total75ammonia/ammonium concentration where approximately 75% of nickel extraction is achievedin 24 hours using a 2.73 ferrous sulphate to laterite ratio. It is important to mention thatsolid/liquid separation is an issue when reductant and total ammonia concentration are high inchloride solutions.In contrast, cobalt in sulphate media requires lower additions of ferrous sulphate, with aferrous sulphate to laterite ratio of 0.245 being optimal for a total ammonia/ammoniumconcentration of 140 g/l reaching 60 % cobalt extraction in 4 hours. Manganese in sulphatemedia behaves similarly to cobalt; however its highest extraction did not exceed 20 %.In chloride media, cobalt extraction increases faster reaching 80 % in 9 hours using a1.485 ferrous sulphate to laterite ratio with 185 g/l of total ammonia/ammonium. Similarextractions are observed using over 0.738 ferrous sulphate to laterite ratio after 24 hours.Further increasing ferrous sulphate concentrations lead to detrimental effect and slowerkinetics.Manganese extraction in chloride media is 65 % after 3 hours when the reductant isincreased up to 1.485 of ferrous sulphate to laterite ratio, it is clear that higher concentration offerrous is detrimental to maintaining manganese in solution. Also, in the cases where themanganese reached maximum extractions near 60 % its concentration then decreased to 45 %after further leaching, independent of reductant concentration.On the other hand, all tests carried out at different conditions show that independent ofthe total ferrous concentration, the maximum cobalt and manganese extraction is around 80%and 65%, respectively (followed by a decrease due to adsorption and/or co-precipitation) at185 g/l of ammonium/ammonia. Thus, these results likely point to the maximum amount ofcobalt and manganese that can be leached.Solid-liquid separation is an important point to consider for leaching processes. Inammoniacal solutions using ferrous sulphate as a reductant, the success of phase separationdepends not only on the particle size distribution but also on the ferrous concentration. Thus,76all tests were carried out over 24 hours to allow most ferrous ions to react and precipitateduring the leach in order to achieve better separation by minimizing the precipitation of ferrichydroxide during filtration. However, longer residence times are required to precipitate theremaining ferrous ions in solution and to allow for better filtration, especially in chloridesolutions when the ammonia/ammonium concentration was high (185 g/l) and ferrous sulphateto laterite ratio was higher than 1.74. Following the ORP throughout a test gives a goodindication as to when it is best to finish the test for better filtration results.Magnetite formation depends on temperature as well as on ferrous sulphateconcentration in both sulphate and chloride solutions. Tables 5.3 and 5.4 summarize acomparison of magnetite content between 0.738, 1.485 and 2.734 ferrous sulphate to lateriteratio using 140 g/l of ammonium/ammonia total concentration in sulphate media and 0.738 and1.485 ferrous sulphate to laterite ratio using 185 g/l ammonium/ammonia in chloride media.The results suggest that the magnetite content in the residue increases with ferrous sulphateconcentration, however, the formation of magnetite in sulphate is more extensive than inchloride media. Also, it is worth noting that in chloride, even at high magnetite contents in theresidue (for example over 63% by weight), cobalt and manganese appear to be maintained insolution when the ammonium/ammonia concentration is 185 g/l. The better metal amminestability in ammonium chloride could be due to the higher total ammonia concentration and/orchloride anion effect on the ammines.  With the present data it is possible to show advantages(or disadvantages) of ammonium sulphate and ammonium chloride, however, this needs to befurther studied in the future at constant total ammonia concentrations. Unfortunately, there areno XRD analysis results to compare both media at the same total ammonia/ammoniumconcentration to evaluate if the lower and/or slower magnetite formation in chloride media isrelated to ammonia concentration as well. Detailed results can be seen in Figures 5.34 to 5.39.It is unclear how ferrous iron precipitates in the cases in which an excess of ferroussulphate was added to the leaching test. Both XRD and TIMA analyses suggest that magnetiteis the most likely product formed while suggesting that wustite could also be present. Perhapsdue to the poor crystallinity of the residue it is difficult to accurately see which phase hasformed and it might be something in between wustite and magnetite.77Table 5.3: Solids residue phase content analysis after leaching in the presence of ammoniumsulphate. Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140g/l, initial mass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and asolution volume of 500 ml and 80°C working temperature.Ratio Red/Ore Mass (g)Goethite Maghemite C Magnetite HematiteHead 34.1 8.1 0.0 3.40.738 18.0 23.3 30.4 3.61.485 5.8 0.0 85.8 4.32.734 0.0 0.0 134.1 1.8Ratio Red/Ore Mass (%)Goethite Maghemite C Magnetite HematiteHead 56.8 13.5 0.0 5.70.738 22.4 29.0 37.9 4.41.485 5.6 0.0 83.5 4.22.734 0.0 0.0 94.6 1.3Table 5.4: Solids residue phase content analysis after leaching in the presence of ammoniumchloride. Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 185g/l, initial mass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and asolution volume of 500 ml and 80°C working temperature.Ratio Red/Ore Mass (g)Goethite Maghemite C Magnetite HematiteHead 34.1 8.1 0.0 3.40.738 45.6 12.4 18.5 2.91.485 30.1 0.0 66.4 5.1Ratio Red/Ore Mass (%)Goethite Maghemite C Magnetite HematiteHead 56.8 13.5 0.0 5.70.738 55.1 15.0 22.4 3.51.485 28.6 0.0 63.0 4.878Figure 5.34: Effect of ferrous sulphate addition on nickel extraction in ammonium sulphate.Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140 g/l, initialmass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and a solution volumeof 500 ml and 80°C working temperature.Figure 5.35: Effect of ferrous sulphate addition on nickel extraction in ammonium chloride.Total ammonia/ammonium concentration at 185 g/l, initial mass sample of 60 grams, stirringvelocity of 600 RPM, an F80 of 32 μm and a solution volume of 500 ml and 80°C workingtemperature.0102030405060700 5 10 15 20 252.734 Red/Ore ratio1.485 Red/Ore ratio0.738 Red/Ore ratio0.246 Red/Ore ratio0.123 Red/Ore ratioTime (h)Nickel Extraction (%)010203040506070800 5 10 15 20 250.49 Red/Ore ratiol0.74 Red/Ore ratio1.49 Red/Ore ratio2.73 Red/Ore ratioNickel Extraction (%)Time (h)79Figure 5.36: Effect of ferrous sulphate addition on cobalt extraction in ammonium sulphate.Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140 g/l, initialmass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and a solution volumeof 500 ml and 80°C working temperature.Figure 5.37: Effect of ferrous sulphate addition on cobalt extraction in ammonium chloride.Total ammonia/ammonium concentration at 185 g/l, initial mass sample of 60 grams, stirringvelocity of 600 RPM, an F80 of 32 μm and a solution volume of 500 ml and 80°C workingtemperature.0102030405060700 5 10 15 20 251.485 Red/Ore ratio0.738 Red/Ore ratio0.246 Red/Ore ratioTime (h)Cobalt Extraction (%)01020304050607080900 5 10 15 20 250.49 Red/Ore ratio0.74 Red/Ore ratio1.49 Red/Ore ratio2.73 Red/Ore ratioCobalt Extraction (%)Time (h)80Figure 5.38: Effect of ferrous sulphate addition on manganese extraction in ammoniumsulphate. Effect of ferrous sulphate addition. Total ammonia/ammonium concentration at 140g/l, initial mass sample of 60 grams, stirring velocity of 600 RPM, an F80 of 32 μm and asolution volume of 500 ml and 80°C working temperature.Figure 5.39: Effect of ferrous sulphate on manganese extraction in ammonium chloride. Totalammonia/ammonium concentration at 185 g/l, initial mass sample of 60 grams, stirringvelocity of 600 RPM, an F80 of 32 μm and a solution volume of 500 ml and 80°C workingtemperature.05101520250 5 10 15 20 251.485 Red/Ore ratio0.738 Red/Ore ratio0.246 Red/Ore ratio0.123 Red/Ore ratioTime (h)Manganese Extraction (%)0102030405060700 5 10 15 20 250.49 Red/Ore ratio0.74 Red/Ore ratio1.49 Red/Ore ratio2.73 Red/Ore ratioManganese Extraction (%)Time (h)815.6 Comparison of chloride and sulphate solutionsAmmonium sulphate and ammonium chloride media are included in this study and acomparison of the results is described below. For nickel, similar results are found in bothmedia where higher concentrations of ferrous sulphate benefit its extraction kinetics.From these results, it appears that for nickel extraction, ferrous sulphate addition has amore important role than total ammonia/ammonium concentration. On the other hand, cobaltand manganese leaching appears to depend mainly on ammonia stability and concentrationduring the test.In ammonium sulphate, the variation of total ammonia/ammonium concentration appearsto have a more important impact on leaching performance in comparison with ammoniumchloride where total ammonia/ammonium concentration was more stable during the test atconstant pH, i.e., less ammonia evaporated. This could be, at least in part, the reason thatbetter results for cobalt and manganese are found in chloride media. Experimentally it was notpossible to achieve higher concentrations of ammonium/ammonia in sulphate media, such as185 g/l, despite the solutions being prepared with the same amount of reagents due toevaporation, thus, a comparison at high concentration between both media was not possible tobe carried out.Solid residue analyses can help better understand the leaching process. As mentionedabove in previous sections, the magnetite content in the residue increases with temperature andferrous sulphate addition. Also, magnetite contents are always higher in sulphate, thisobservation coincides with the leaching behaviour of nickel, which is faster, reaching amaximum extraction value at 5 hours, time at which the curve stabilizes. However, the ORPvalue is still low after 5 hours of reaction (-400 mV vs. SHE approximately) thus, it isexpected that ferrous ions still remain in solution at this time. At 24 hours the Eh value is closeto 0 mV SHE but metal extraction does not increase further, which suggests that some ferrousis oxidized and precipitates.82In ammonium chloride, the ORP increases slowly in comparison to sulphate and theferrous ion concentration on the final leach solution is higher in these solutions. Thus, ORPvariations and ferrous concentration behaviour can describe the extent of magnetite formationand nickel extraction during the process.As mentioned before, in sulphate, ferrous ions precipitate during the process makingsolid/liquid separation simple. However, in chloride at high ferrous sulphate andammonia/ammonium concentration, settling and filtration became difficult even after 24 hoursof residence time. Iron appeared to start precipitating from solution once taken out of thereactor and in contact with oxygen.In chloride, in order to achieve the highest cobalt and nickel extractions, while at thesame time allowing for good solid liquid separation after the leaching tests, the followingconditions were found.  For cobalt, a 0.738 ferrous sulphate to laterite ratio with 185 g/l ofammonium/ammonia concentration is sufficient for cobalt extraction and good s/l separation.For nickel, a 1.485 ferrous sulphate to laterite ratio with 100 g/l of total ammonium/ammoniaconcentration allows for good solid liquid separation. These conditions were optimized forboth extraction and solid liquid separation. However, it is possible to achieve higherextractions with a better control of reagent addition, particularly ferrous sulphate. It isimportant to mention that it may be beneficial to add the reductant throughout the leachingtime (as opposed to all at once), as its concentration can affect cobalt and manganese stability.Small dosages of the reductant could keep the cobalt and manganese in solution and improvenickel extraction at the same time.Figures 5.40 to 5.50 show different leach residues exposed to ammonium sulphate andchloride. From these pictures it is possible to see that magnetite can precipitate around otherparticles, likely goethite. The precipitates have very small particle size (around 1 to 3 microns)and they are present preferably around particles rich in iron and manganese. No precipitate wasdetected around silica particles, for example. Through these pictures it is possible to see thatporous magnetite layers are formed. For the transformation of iron-rich phases, the P80 (80 %w/w of particles below a certain size in the final product) of the residue is approximately 2583µm but this could be misleading as many fine particles cover every unleached particle,especially goethite particles, which show particle sizes even as low as 10 microns in sulphatemedia.Figures 5.42 and 5.43 show the fine particles formed during the leaching process in bothmedia.  Specifically, in Figure 5.44, it is shown that particles can be below 2 µm.Figure 5.45 shows a partially leached particle followed by a mapping image for Fe, Ni,Co and Mn in Figure 5.46. Mn and Ni are associated in the ore as Asbolane, a phase whichwas not reduced during the process. In Figure 5.47 another partially leached particle is shownbeing rich in iron with an apparent low nickel content.Many iron phases are present in the head due to the nature of limonitic deposits and inthe solid residues.  In order to identify each phase it is possible to use microscopy, whichallows classification of iron phases by colour where goethite, hematite and magnetite areidentified as shown in Figures 5.48 to 5.50.Figure 5.40: Solid residue micrograph after leaching in the presence of ammonium sulphate84Figure 5.41: Solid residue micrograph after leaching in the presence of ammonium chlorideFigure 5.42: Solid residue micrograph after leaching in the presence of ammonium sulphate.85Figure 5.43: Solid residue micrograph after leaching in the presence of ammonium chlorideFigure 5.44: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating ultra-fine, subhedral and euhedral cubic grains of likely magnetite and/or wustite.86Figure 5.45: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating remnant particle of hematite and rim of asbolane.87(a) (b)(c) (d)Figure 5.46: Detail of partially leached particle from Figure5.38  depicting metal distribution.(a) Iron, (b) Nickel, (c) Cobalt, (d) Manganese.88Figure 5.47: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating martite/hematite with possible dissolution edges with absent nickel.Figure 5.48: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating a remnant composite particle of discontinuous rim of goethite on martite-hematitepseudomorph after magnetite.89Figure 5.49: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating a remnant binary particle of mostly magnetite with subhedral pseudomorphs ofmartite-hematite.Figure 5.50: Solid residue micrograph after leaching in the presence of ammonium sulphate,illustrating a remnant particle of martite-hematite with a discontinuous rim of goethite.90Figures 5.51 to 5.53 summarize the best conditions found to extract nickel, cobalt andmanganese in each the presence of ammonium sulphate or ammonium chloride.Figure 5.51: Nickel extraction comparison in the presence of either ammonium sulphate orammonium chloride.Figure 5.52: Cobalt extraction comparison in the presence of either ammonium sulphate orammonium chloride010203040506070800 5 10 15 20 25100 g/l NH3 tot - 0.74 Red/Sol - NH4Cl100 g/l NH3 tot - 0.74 Red/Sol - (NH4)2SO4100 g/l NH3 tot - 1.49 Red/Sol - NH4ClNickel Extraction (%)Time (h)01020304050607080900 5 10 15 20 25185 g/l NH3 tot - 0.74 Red/Sol - NH4Cl134 g/l NH3 tot - 0.74 Red/Sol - (NH4)2SO4185 g/l NH3 tot - 1.49 Red/Sol - NH4Cl134 g/l NH3 tot - 1.49 Red/Sol - (NH4)2SO4Cobalt  Extraction (%)Time (h)91Figure 5.53: Manganese extraction comparison in the presence of either ammonium sulphateor ammonium chloride.All these results discussed above suggest that leaching in stages could be a suitable wayto extract nickel, cobalt and manganese. Thus, some residues were re-leached in order todetermine the extent to which metal extraction could be increased by a second leaching stage.The following table summarizes results.0102030405060700 5 10 15 20 25185 g/l NH3 tot - 0.74 Red/Sol - NH4Cl134 g/l NH3 tot - 0.74 Red/Sol - (NH4)2SO4185 g/l NH3 tot - 1.49 Red/Sol - NH4Cl134 g/l NH3 tot - 1.49 Red/Sol - (NH4)2SO4Manganese  Extraction (%)Time (h)92Table 5.5: Ferrous sulphate to metal extracted ratio in the presence of ammonium sulphate.Stages ConsumptionI II III GlobalFeSO4/Metal extractedmass ratio (g/g)Red. (FeSO4)/Initial Oremass ratio (g/g) 1.49 0.738 2.22Extraction (%)Co 4.05 30.0 34.0 3271Ni 65.2 17.5 83.0 180Mn 8.68 26.4 35.0 317Red. (FeSO4)/Initial Oremass ratio (g/g) 1.49 0.74 2.22Extraction (%)Co 17.1 32.0 49.1 1390Ni 64.2 16.5 81.0 185Mn 6.96 21.4 28.4 342Red. (FeSO4)/Initial Oremass ratio (g/g) 0.74 0.74 1.48Extraction (%)Co 15.0 27.9 42.9 10576Ni 52.4 13.2 65.6 151Mn 0.00 1.36 1.36 Extraction too lowRed. (FeSO4)/Initial Oremass ratio (g/g) 0.25 0.25 0.74 1.23Extraction (%)Co 47.5 13.4 13.4 74.3 570Ni 19.7 2.61 29.3 51.7 159Mn 0.84 2.77 19.8 23.4 229From Table 5.5 it is possible to see that in sulphate media a first stage for cobalt andmanganese extraction could be an option to avoid exposing cobalt to high concentrations offerrous sulphate and to obtain approximately 74 % extraction. However, this would require atleast 3 stages. For nickel in sulphate media, it is possible to achieve over 80 % in two stageswith high concentrations of ferrous sulphate. Kinetics of the second stages are very slow andrecovery was not higher than 18 %.  It appears as though the product layer around the particlessignificantly influences the leaching process and limits further nickel extraction.93Table 5.6: Ferrous sulphate to metal extracted ratio in the presence of ammonium chloride.Stages ConsumptionI II III Global FeSO4/Metal mass ratio (g/g)Red. (FeSO4)/Initial Oremass ratio (g/g) 0.74 0.74Extraction (%)Co 74.7 74.7 299Ni 25.5 25.5 192Mn 8.22 8.22 387Red. (FeSO4)/Initial Oremass ratio (g/g) 1.49 1.49Extraction (%)Co 2.56 2.56 Extraction too lowNi 68.1 68.1 146Mn 1.09 1.09 Extraction too lowRed. (FeSO4)/Initial Oremass ratio (g/g) 1.49 1.49Extraction (%)Co 0.81 0.81 Extraction too lowNi 51.7 51.7 192Mn 0.90 0.90 Extraction too lowIn chloride solutions, Table 5.6, only one leaching stage was carried out.  However, asmentioned previously, when using an “intermediate” ferrous sulphate concentration, cobaltextraction reached approximately 75 %, which is much better than the results in the presenceof ammonium sulphate (300 FeSO4/metal extracted mass ratio). The last two tests in chlorideshown in Table 5.6 did not use high ammonia concentrations in order to improve s/lseparation. The best results for nickel in terms of extraction were achieved using 100 g/l ofammonium/ammonia total concentration and 1.49 ferrous sulphate to laterite ratio where 68 %of nickel was extracted.“Re-leach” kinetics in sulphate are shown in Figures 5.54 to 5.56 for nickel, cobalt andmanganese, respectively, at 100 and 140 g/l total ammonia/ammonium concentration. Thecharts show that 17% extraction is achieved very quickly in the second stage after an intensefirst stage (where nickel extraction was over 65 %).  This suggests that this amount of nickelwas probably leached during the first stage and precipitated and/or was absorbed on thesurface of the particles. Similar behaviours are observed for cobalt and manganese, thus, fromthese results maximum extractions in sulphate media are not expected to be higher than ca. 83%.94Figure 5.54: Two-Stage nickel leaching in the presence of ammonium sulphateFigure 5.55 Two-Stage nickel leaching in the presence of ammonium sulphate01020304050607080900 5 10 15 20 25 30 35 40 45 50140 g/l100 g/lTime (h)Nickel Extraction(%)Stage 1(0 to 24 hrs)Stage 2(24 to 48 hrs)0102030405060700 5 10 15 20 25 30 35 40 45 50140 g/l100 g/lTime (h)Cobalt Extraction(%)Stage 1(0 to 24 hrs)Stage 2(24 to 48 hrs)95Figure 5.56: Two-Stage nickel leaching in the presence of ammonium sulphate5.7 Maximum expected extraction vs. adsorption and/or co-precipitationIn this section a deeper discussion and analysis of feed vs. residues is presented.As mentioned before, nickel, cobalt and manganese behave differently during the leachprocess and each of these shows differences when in the presence of ammonium sulphate orammonium chloride. For instance, as seen throughout the results section, nickel has fasterextraction kinetics in the presence of ammonium sulphate, however its highest extraction isachieved in ammonium chloride. Cobalt also has faster extraction kinetics in ammoniumsulphate than in chloride (although not as evident as nickel), but it shows similar extractionextents in both media.  Despite the fact that it suffers from adsorption or co-precipitation in thepresence of ammonium sulphate. Finally manganese has similar extraction kinetics in bothmedia, reaching slightly higher extractions in the presence of ammonium chloride. Manganesealso suffers from adsorption and/or co-precipitation in ammonium sulphate and to a lesserextent in ammonium chloride. This adsorption and/or co-precipitation phenomenon wasincreased with higher solids content and higher ferrous sulphate additions.010203040506070800 5 10 15 20 25 30 35 40 45 50140 g/l100 g/lTime (h)Manganese Extraction(%)Stage 10 to 24 hrs) Stage 2(24 to 48 hrs)96In an attempt to better understand the adsorption and/or co-precipitation phenomenon,metal deportment data was produced using TIMA.  The sample chosen came from a test thatshowed high transformation of solid phases in the feed into new phases in the residue, but didnot have high extractions, particularly for cobalt and manganese. Since this phenomenon wasmuch more noticeable in ammonium sulphate, a leach residue treated in its presence wasselected.The idea is that this analysis provides information on new solid phases being produced,and if a new phase (for instance magnetite) contains nickel, cobalt or manganese, it can beconlcuded that adsorption and/or co-precipitation took place. This analysis will also provideinformation regarding the extent of extraction from each phase present in the feed for eachmetal of interest.Finally, by knowing how much metal (nickel, cobalt or manganese) was “transferred”from a solid phase in the feed to new phases in the residue, it can be assumed or suggested thatthe metal in question was, at some point during the leach, extracted, which can in theory givean approximate maximum expected extraction. It is worth mentioning that the sample inquestion (with a high intense first leach stage) had mass of 60 g, while the solid residue had amass of 142 g.Tables 5.7 and 5.8 show metal deportment in the feed and leach residue by mass andweight percentage, while Table 5.9 shows metal extraction by phase.  In the event that a solidphase in the residue has a higher metal content than in the feed, it is assumed that adsorptionand/or co-precipitation took place, which is calculated using the data from Table 5.7.Complementing this information, Figures 5.57 and 5.58 graphically depict the informationfrom Table 5.8.97Table 5.7: Metal deportment by phase in the feed and leach residue (in µg)Phase Feed (µg) Residue (µg)Ni Co Mn Ni Co MnMagnetite or Wustite 0.00 0.00 0.00 0.00 123 248Goethite 44.8 0.00 0.00 0.00 0.00 37.1MnFe Wad 654 149 1032 27.1 26.9 235Hematite 26.9 0.00 0.00 0.00 0.00 0.00Goethite-low Ni 98.60 0.00 26.1 246 0.00 446Asbolane 71.7 46.9 275 22.1 18.5 235Others Ni 0.00 0.00 12.4 0.00 0.00 0.00Others 0.00 0.00 41.1 0.00 0.00 37.1Total 896 196 1346 295 168 1201Table 5.8: Metal deportment by phase in the feed and leach residue (in w/w %)Phase Feed (%) Residue (%)Ni Co Mn Ni Co MnMagnetite or Wustite 0.00 0.00 0.00 0.00 73.0 20.0Goethite 5.00 0.00 0.00 0.00 0.00 3.00MnFe Wad 73.0 76.0 75.0 9.20 16.0 19.0Hematite 3.00 0.00 0.00 0.00 0.00 0.0Goethite-low Ni 11.0 0.00 1.90 83.3 0.00 36.0Asbolane 8.00 24.0 20.0 7.50 11.0 19.0Others Ni 0.00 0.00 0.90 0.00 0.00 0.00Others 0.00 0.00 2.99 0.00 0.00 3.00Total 100 100 100 100 100 100Table 5.9: Metal extraction per solid phase (* Ads. = possible adsorption and/or co-precipitation)Phase Extraction per phase (%)Ni Co MnMagnetite or Wustite - *Ads *AdsGoethite 100 - *AdsMnFe Wad 95.9 81.9 77.2Hematite 100 - -Goethite-low Ni *Ads - *AdsAsbolane 69.2 60.6 14.5Others Ni - - 100Others - - 9.7398Table 5.10: Maximum estimated extraction based on possible adsorbed and/or co-precipitated(Co-pp) metal plus metal dissolved during the leach.PhaseMetal adsorbed/Co-pp(%)Ni Co MnMagnetite or Wustite - 62.7 18.0Goethite - - 2.70MnFe Wad - - -Hematite - - -Goethite-low Ni 16.4 - 30.5Asbolane - - -Others Ni - - -Others - - -Total ads/co-pp 16.4 62.7 51.2Total extracted 67.1 14.1 10.8Possible total extraction 83.5 76.8 61.9Figure 5.57: Metal deportment in feed99Figure 5.58: Metal deportment in leach residueTable 5.10 gathers information from Table 5.7 and uses it to estimate how much metalmight have been adsorbed and/or co-precipitated. The result is interesting as when adding the“adsorbed” metal plus the amount extracted during the leach, the numbers obtained are in linewith the maximum extractions obtained in more optimized leaching tests (approximately 80%for nickel and cobalt and 65 % for manganese). This suggests that higher extractions might bedifficult through further optimization of the leaching conditions and perhaps a differentapproach is required, for instance, re-grinding the residue or autoclaving. This would logicallyincrease the cost of implementing this process; however it would be interesting to see its actualeffect, if any.Also, it is important to mention that the maximum extraction achieved in two stages insulphate media coincide with the maximum “possible extraction calculated” (approximately 83%). This suggests that in the presence of ammonium sulphate, the maximum leachable nickelcorresponds to 83 % and there is a fraction which probably will not leach in these conditions.Re-leaching in the presence of ammonium chloride with a previous first intense leach stagecould be attempted in order to corroborate this comment. Similarly, the maximum cobaltextraction does not exceed 80 % in both media, even at very dilute solid conditions, This againsuggests that that there is a portion of this metal, or of the phases to which it is associated, thatdo not react perhaps due to new phases that precipitate around the particles and slow down orstop the leaching process.1006 CONCLUDING REMARKSThe direct reductive leaching of limonite using ferrous sulphate in ammoniacal solutionsis shown to be effective to partially remove nickel, cobalt and manganese.  The leach producesa residue formed mostly by magnetite with high iron contents in an atmospheric leach at 80ºC.The highest extractions for nickel, cobalt and manganese are approximately 82, 80 and 65 %,respectively. However there is still room for improvement as the ferrous sulphateconsumptions should be optimized and evaporation of ammonia decreased.Temperature had a clear effect and best results are obtained at 80ºC. Further increases intemperature lead to excessive ammonia evaporation. This suggests that the process iscontrolled by the chemical reaction at the surface of the limonite particles.Total ammonia had a clear improvement on cobalt and manganese removal. However,the effect was not strong for nickel.  The opposite phenomenon was seen for ferrous sulphate,which improved nickel removal with higher concentrations, but did not work in the same wayfor cobalt and nickel, which is perhaps further evidence to support the negative effect that ironhas on the removal of cobalt and manganese. For nickel, ratios larger than 1.485 of ferroussulphate/ore with 100 g/l of total ammonia/ammonium concentration in both media it ispossible to obtain a nickel extraction of approximately 65 % in one stage.In the presence of ammonium sulphate, with a ratio of 0.0.738 of ferrous sulphate/ore acobalt extraction of approximately 80 % is achieved at high total ammonia/ammoniumconcentrations (100 g/l).  Cobalt is kept in solution only under low solid content conditionsand low magnetite formation. Ferrous sulphate ratios in the following range: 0.74 <reductant/ore < 1.485 in the presence of ammonium chloride result in over 70 % cobaltextraction with 185 g/l of total ammonia and a minimal effect on adsorption and/or co-precipitation.The addition of ammonium chloride rather than ammonium sulphate seemed to haveslowed nickel extraction.  However, it transformed the nickel extraction curve into a morestraight line (i.e. Figure 5.4 vs. Figure 5.13), which suggests nickel extraction could continue101for a longer period of time without passivation issues, as long as enough reductant andammonia are available for the leach to continue. Also, ammonium chloride helped to keepcobalt and manganese in solution; it decreased co-precipitation and/or adsorption issues.Mineralogical analyses of metal deportment in the feed and leach residue suggest thathigher extractions through further optimization of the leaching parameters is difficult orperhaps impossible. Mineralogical analyses suggest that goethite can be 100 % converted andthat it can adsorb and/or incorporate Mn in its lattice probably to replace the Ni removed.“Goethite low Ni” is not leached and it appears to adsorb and/or incorporate in its matrixmetals from solution (especially nickel), MnFe-Wad and asbolane are partially  leached duringthe process. The new phases formed (summarized as magnetite) adsorb and/or incorporatemetals from solution during the process, mainly cobalt and a small proportion of manganese.A comparison of sulphate and chloride solutions suggests that ammonium chloride couldbe more attractive due to the stability of cobalt and manganese, and at the same time similarextractions for nickel.Finally, this study provides a continuation from the previous study of direct reduction oflimonite with metallic iron (Zuniga 2009-2010), and proposes an alternative that couldpotentially match the performance of the Caron process. However, care must be taken whenlooking into the economics of this approach and optimization of reagent addition needs to belooked at.  This should include the efficiency of ferrous sulphate utilization and ammonialosses due to evaporation. Also the control of pH proved critical in this process and mustalways be taken into account.1027 RECOMMENDATIONS FOR FUTURE RESEARCHEven though this study has found that it can perform similarly to the Caron process interms of the removal percentages for nickel, cobalt and manganese, there are points that stillrequire further research.- Effect of agitation: It could be of interest in future work to look at the effect ofagitation in more detail. The effect of agitation in this work was carried out underconditions that lead to low goethite reduction and in general low metals extraction. Inorder to further confirm its effect, this could be tested in tests with higher extractions,in which the product layer in the residue is thicker.- Ferrous sulphate utilization: large amounts of ferrous sulphate were required and theefficiency was low compared to the proposed stoichiometry. It is possible that ferrousions end up precipitating as hydroxides, which is not uncommon in alkaline media,which would affect the performance of the process considerable- Ammonia: ammonia evaporation was a very noticeable issue during this study. It wasdifficult to maintain a constant concentration in the reactor which was not fully sealed.Also this would increase the cost substantially in case this process was considered forlarger scale.- Ammonium chloride could be an attractive medium to continue studying, with nickelexhibiting a continuously rising extraction curve, as well as better stability of cobaltand manganese. Optimization work could include extra additions of ferrous sulphateafter cobalt and manganese have reached a maximum extraction in order to see iffurther extraction is achieved; while at the same time continue with nickel dissolution.Also, re- leach (second stage) should be tested.- Autoclaving: as explained in the previous point and in the results, ammoniacalevaporation was a serious issue and increasing temperature was always beneficial butcould not be taken advantage of due to excessive evaporation. Autoclaving would103allow to increase temperature and perhaps improve the crystallinity of the solidresidues, thus improving kinetics and further treatment of the solids (thickening,filtering, etc.). However, the autoclave selected must allow for pH measuring andadjustment during the leaching test.- Further mineralogy analyses to confirm current findings. This would help elucidatewhether it is worth to continue pursuing higher extractions by optimizing leachparameters or by following a different route, such as re-grinding the leach residue orautoclaving. Also this might help elucidate how ferrous ions react and precipitate incases where an excess of ferrous sulphate was added to the leaching test.- Potential uses of the solid residue: considering that magnetite is the main phase foundin the solid residue, it could potentially be considered as an iron ore concentrate. Workshould be done in order to confirm whether it is possible to reach the necessary ironcontent in the residue to make it suitable for a blast furnace. In fact, the un-leachednickel and cobalt could potentially improve the quality of the product as added value.104REFERENCESAcharya, S., et al, Kinetics and mechanism of the reductive ammonia leaching of oceannodules by manganous ion. Hydrometallurgy, 1987. 1: p. 169-186.Acharya, S., et al, Ammonia leaching of ocean nodules using various reductants. Erzmetall.1989. 42(2): pp 66-73.Agin, J., et al, Process for the hydrometallurgical treatment of a lateritic nickel/cobalt ore andprocess for producing nickel and/or cobalt intermediate concentrates or commercialproducts using it, US Patent No. 8,287,827B2, 2012.Anderson, PA., et al, Reductive ammoniacal leaching of nickel and cobalt materials, USPatent No. 7,601,314 B2, 2009.Arbiter, N., et al, Copper hydrometallurgy – evolution and milestones. Hydrometallurgy –Fundamentals, technology and innovation, SME 1993: pp 549-565Asselin, E., Thermochemical aspects of the Fe, Ni and Co-NH3-H2O systems relevant to theCaron process. Proceedings of the Sixth International Symposium onHydrometallurgy, 2008.Ballester, A., Sancho, J., Verdeja, L., Extractive Metallurgy: Obtention process. 2: p. 296-317,2000.Boateng et al, The hydrometallurgy of nickel extraction from sulfide ores, I. Feed preparationand leaching systems. II. Solution purification and metal recovery. Minerals Sci. Eng.1978, 10: p. 155-171.Caron, MH., Process of recovering values from nickel and cobalt-nickel ores, US Patent1,487,145, 1924.105Das, RP., et al, Leaching of manganese nodules in ammoniacal medium using glucose asreductant. Hydrometallurgy, 1986. 16: p. 335-344.Das, PK et al, Behaviour of Ni(II) in the H2O-NH3-SO2-MnO2 system. Int. J. Miner. Process.1995. 43 pp 99-111.Das, PK et al, Minimization of Ni(II) precipitation in the Ni(II)-NH3-SO2-(NH4)2SO4-MnO2system. Int. J. Miner. Process. 1997. 50 pp 77-86.Das, R. and Anand, S., Precipitation of iron oxides from ammonia-ammonium sulphatesolutions. Hydrometallurgy, 1995. 38: p. 161-173.Dyer, L.,et al, Cobalt loss due to iron precipitation in ammoniacal carbonate solutions,Hydrometallurgy, 2012, Vol. 125-126, p. 144-147Esmadi, F. and Simm, J., Sorption of Cobalt (II) by amorphous ferric hydroxide. Colloids andSurfaces A: Physicochemical and Engineering Aspects, 1995. 104(2/3): p. 265-270.Forward, FA., et al, Ammonia pressure leach process for recovering copper, nickel and cobaltfrom Sherrit Gordon nickel sulfide concentrate. Can. Min. Metall. Bull. 1953, 499: p.677-684.Georgiou, D. and Papangelakis, V., Characterization of limonitic laterite and solids duringsulfuric acid pressure leaching using transmission electron microscopy. MineralsEngineering, 2004. 17: p. 461-463.Georgiou, D., Behaviour of cobalt during sulphuric acid pressure leaching of a limoniticlaterite, Hydrometallurgy, 2009. 100: p. 35-40Guskos, N. and Papadopoulos, G., Photoacoustic, EPR and electrical conductivityinvestigation of three synthetic minerals pigments: hematite, goethite and magnetite.Materials Research Bulletin, 2002. 37: p. 1051-1061.106Han, KN., et al, Ammonia-ammonium leaching of deep sea manganese nodules. Int. J. Miner.Process, 1974, 1: p. 215-230.Han, KN., et al, The leaching behaviour of nickel and cobalt from metals and ores – A review.The Paul E. Queneau International Symposium Extractive Metallurgy of Copper,Nickel and Cobalt. Vol. I: Fundamental aspects, TMS, 1993: p. 709-733.Isaev, I., et al., The formation of iron (II) ammines in aqueous solution. Russian journal ofInorganic Chemistry, 1990. 35(8): p. 1162-1164.Jandova, J. and Pedlik, M., Leaching behaviour of iron-nickel alloys in ammoniacal solution.Hydrometallurgy, 1994. 35(1): p. 123-128.Kerfoot, D. and Weir, D., The hydro and electrometallurgy of nickel and cobalt. Extractivemetallurgy of nickel and cobalt, 1988: p. 241-267.Krause, E., Hydrometallurgy of Ni Oxides and Silicates. Short Course on Nickel and CobaltHydrometallurgy, COM, 2009.Kuhn, MC., et al, Anaconda’s Arbiter process for copper. CIM Bulletin, February 1974: p. 62-73.McDonald, RG., et al, Atmospheric acid leaching of nickel laterites review. Part II. Chlorideand bio-technologies, Hydrometallurgy, 2008. 91: p. 56-69.McDonald, RG., et al, Atmospheric acid leaching of nickel laterites review. Part I. Sulphuricacid technologies, Hydrometallurgy, 2008. 50: p. 35-55.Meng, X. and Han, K., The principles and Applications of Ammonia Leaching of Metals.Mineral processing and extractive Metallurgy review, 1996. 16(1): p. 23-61.107Mohapatra, M., Anand, S., and Das R., Aqueous reduction of crystalline goethite underammoniacal conditions. Hydrometallurgy, 2002. 65(2-3): p. 227-235.Monhemius, A., Treatment of laterite ores of nickel to produce ferronickel, matte orprecipitated sulphide. Extractive Metallugy of Nickel, 1987. 17: p. 51-75.Mukherjee, A. et al, Dissolution of Cu, Co and Ni from ocean nodules by L-ascorbic acid,Chemical Engineering and Processing: Process intensification, 2005. 44(7): p. 754-759.Nicol, MJ., A fundamental study of the leaching reactions involved in the Caron process,International Laterite Nickel Symposium, 2004, 14 - 18 March, Charlotte, NC, USA.Nikoloski, AN., et al, The passivation of iron during the leaching of pre-reduced laterite oresin ammoniacal solutions, Electrochemical Society Proceedings Volume 2003-18: p.205-218.Nikoloski, AN., The electrochemistry of the leaching reactions in the Caron Process. I. AnodicProcesses, ECS Transactions, 2006, 2(3): p. 197-207.Osseo-Asare, K., Heterogeneous Equilibria in ammonia/laterite leaching systems.International Laterite Symposium 1979a, (30): p. 585-609.Osseo-Asare, K. and Fuerstenau, D., Adsorption Phenomena in hydrometallurgy, 1. Theuptake of copper, Nickel and Cobalt by Oxide Adsorbents in Aqueous AmmoniacalSolutions. International Journal of mineral processing, 1979b. 6(2): p. 85-104.Osseo-Asare, K. and Asihene, S., Heterogeneous equilibria in ammonia/laterite leachingsystems. International Laterite Symposium, 1979b: p. 585-609.108Osseo-Asare, K. and Fuerstenau, D., Adsorption Phenomena in Hydrometallurgy, 2.Electrophoretic behavior of oxides in aqueous Ammoniacal Solutions of Copper, Nickeland Cobalt. International Journal of mineral processing, 1980a. 7(2): p. 117-127.Osseo-Asare, K. and Fuerstenau, D., Adsorption Phenomena in Hydrometallurgy, 3.Model forCopper, Nickel and Cobalt uptake by oxides Adsorbents in aqueous AmmoniacalSolutions. International Journal of mineral processing, 1980b. 7(3): p. 219-234.Osseo-Asare, K., Application of activity-activity diagrams to ammonia hydrometallurgy II.The copper, nickel and Cobalt ammonia water systems at elevated temperatures.Process Fundament. Consid. Sel. Hydrometallurgy Syst., 1981a: p. 359-360.Osseo-Asare, K., Application of activity-activity diagrams to ammonia hydrometallurgy: Fe-NH3-H2O, Fe-NH3-H2O-CO3 and Fe-NH3-H2O-SO4 systems at 25ºC. Transactions ofMining and Metallurgy, 1981b. 90: p. 159-164.Osseo-Asare, K. and Fuerstenau, D., Adsorption of Copper, Nickel, and Cobalt by OxideAdsorbents from aqueous Ammoniacal Solutions. Journal of Colloid and interfaceScience, 1987. 118(2): p. 524-542.Papangelakis, V., Advances in solutions chemistry in Ni hydrometallurgy. Short Course onNickel and Cobalt Hydrometallurgy, COM, 2009.Park, K., et al., A study on the oxidative ammonia/ammonium sulphate leaching of a complex(Cu-Ni-Co-Fe) matte. Hydrometallurgy, 2007. 86( 3): p. 164-171.Peek, E., Chloride metallurgy- Industrial applications. short Course on Nickel and CobaltHydrometallurgy, COM, 2009.Peng, Z., Synthesis of magnetite nanorods through reduction of goethite. Chemistry Letters,2005. 34(5): p. 636-637.109Rubisov, DH., et al, Sulphuric acid pressure leaching of laterites – Universal kinetics of nickeldissolution for limonites and limonitic/saprolitic blends, Hydrometallurgy, 2000, 58: p.1-11.Rubisov, DH., et al, Sulphuric acid pressure leaching of laterites – A comprehensive model ofa continuous autoclave, Hydrometallurgy, 2000. 58: p. 89-101.Rudnik, E., Burzynska, L., and Gumowska, W., Hydrometallurgical recovery of copper andcobalt from reduction-roasted copper converter slag. Minerals Engineering, 2008.29(1): p. 88-95.Taylor, A., Laterites – Still a frontier of nickel process development. TMS 2013 142nd AnnualMeeting & Exhibition, San Antonio, Texas, USA.Tromans, D., Oxygen solubility modeling in ammoniacal leaching solutions: leaching ofsulphide concentrates. Mineral Engineering, 2000. 13(5): p. 497-515.Vu, C., et al, Leaching behaviour of cobalt in ammonia solutions. Trans. Instn. Min. MEtall.(Sec. C), 1977. 86: p. C119-121.Wang, J. and Deng, T., Studies on the behavior of Co2+ and Zn2+ in the conversion fromgoethite to magnetite in NH3(NaOH)-Fe2+, Co2+-H2O medium under hydrothermalconditions. Hydrometallurgy, 2008. 92: p. 107-114.Williams, RD., “Copper concentrate dissolution chemistry and kinetics in an ammonia-oxygenenvironment. In fundamental aspects of hydrometallurgical processes, AIChEsymposium series No 173, 74. 1978: p. 21-27.Zhong, S. and Hepworth, M., A Calculation method for determining equilibria in metal-ammonia-water systems. Hydrometallurgy, 1995. 38(1): p. 15-37.110Zuniga, M. et al, Reduction of limonitic laterite in ammoniacal solutions using metallic iron.Conference of Metallurgists, Sudbury ON, 2009, Canada.Zuniga, M. et al, Leaching of a limonitic laterite in ammoniacal solutions with metallic iron.Hydrometallurgy, 2010. 104: p. 260-267Zuniga, M. and Asselin, E., Reductive leaching of limonitic laterites using ferrous sulphate.TMS annual conference, San Antonio TX, 2013, USA.Others sources- Infomine website www.infomine.com, Commodities section.- London Metal Exchange website, www.lme.com.- Data Handbook of Chemistry and Physics, 85th Edition, 2004.111APPENDICESIron balanceTable0A.1: Sample mass and ICP assay.Sample Mass (g) ICP Iron assay(w/w %) Iron mass (g)Head 60 51.8 31.1Residue 142 63.2 89.6Table0A.2: Iron balance in head sample as per TIMA analysis.Mineral phase SimplifiedFormulaContent inSample (w/w %)Mass insample (g)Molecularmass (g)Iron contentin each phase(w/w %)Ironmass(g)Goethite FeO*OH 31.1 18.7 89.0 62.9 11.7MnFe-Wad KFe8O16 30.0 18.0 735 60.3 5.80Hematite Fe2O3 22.5 13.5 160 70 9.45Magnetite Fe3O4 0.18 0.11 232 72.4 0.08Goethite_low Ni FeO*OH 10.6 6.36 89.0 62.9 4.00Total Iron Mass (g) 31.1Table0A.3: Iron balance in head sample as per Q-XRD analysis.Mineral phase SimplifiedFormulaContent inSample (w/w %)Mass insample (g)Molecularmass (g)Iron content ineach phase(w/w %)Ironmass(g)Goethite FeO*OH 56.8 34.1 89.0 62.9 21.4Maghemite Fe2O3 13.5 8.10 160 70.0 5.67Hematite Fe2O3 5.70 3.42 160 70.0 2.39Total Iron Mass (g) 29.5112Table0A.4: Iron balance in leach residue as per TIMA analysisMineral phase SimplifiedFormulaContent inSample (w/w %)Mass insample (g)Molecularmass (g)Iron contentin each phase(w/w %)Ironmass(g)Goethite FeO*OH 0.10 0.14 89 62.9 0.09MnFe-Wad KFe8O16 0.48 0.68 735 60.3 0.08Hematite Fe2O3 12.7 18.0 160 70.0 12.6Magnetite Fe3O4 0.98 1.39 232 72.4 1.01Goethite_low Ni FeO*OH 9.44 13.4 89 62.9 8.42Wustite FeO 69.7 98.8 57 98.2 97.1Total Iron Mass (g) 119Table0A.5: Iron balance in leach residue as per Q-XRD analysisMineral phase SimplifiedFormulaContent inSample (w/w %)Mass insample (g)Molecularmass (g)Iron contentin each phase(w/w %)Ironmass(g)Hematite Fe2O3 1.30 1.84 160 70 1.29Magnetite Fe3O4 94.6 134 232 72.4 97.1Total Iron Mass (g) 98.4Table0A.6: Iron balance comparison between ICP, Q-XRD and TIMA analyses.Analysis Iron in head sample (g) Iron in residue (g)ICP 31.1 89.6XRD 29.5 98.4TIMA 31.1 119Table0A.7: Iron balance comparison between ICP, Q-XRD and TIMA analyses using ICP asthe base case (100 %)Analysis Head Sample (w/w %) Residue (w/w %)ICP 100 100XRD 95.0 110TIMA 100 133113TIMA report114115116117118119120121122123124125126127128129130131132133134135Activation energy calculationsActivation energy in the presence of ammonium sulphate1- CobaltLow ferrous sulphate concentration:136Chemical control model activation energy calculation: 41.1 kJ/molDiffusion on the solid layer model activation energy calculation: 61.3 kJ/molIntermediate ferrous sulphate concentration:137Chemical control model activation energy calculation: 24.3 kJ/molDiffusion in the solid layer model activation energy calculation: 41.0 kJ/mol138The chemical reaction control model fits data better. However the calculations suggest a mixedcontrol.For the intermediate ferrous sulphate concentration case at 80°C, cobalt appears in very lowconcentration in the presence of ammonium sulphate (perhaps adsorbed or/and precipitatedonto oxides and hydroxides) this is the reason this was not used for the calculations.Under high concentration of ferrous sulphate, the magnetite production is higher, which mightexplain/suggest a mixed control situation.2- NickelLow ferrous sulphate concentration:139Chemical control model activation energy calculation: 48.8 kJ/molDiffusion in the solid layer model activation energy calculation: 59.9 kJ/molThe diffusion in the solid layer fits better. However the calculations suggest chemical reactioncontrol.Perhaps, in the beginning the chemical reaction is the main control mechanism, but later itchanges, especially under high ferrous sulphate concentration and  high temperature ( forexample 65°C after 3 hours and 80°C after 1 hour).140For this section the activation energy calculation is shown below:Diffusion in the solid layer model activation energy calculation: 28.4 kJ/mol (it suggests mixedcontrol)Intermediate ferrous sulphate concentration:141Chemical control model activation energy calculation: 36.2 kJ/molDiffusion in the solid layer model activation energy calculation: 52.7 kJ/molApparently the chemical reaction and diffusion in the solid layer are the models that best fit thedata. However, the activation energy values suggest mixed control and/or chemical reaction.3- ManganeseFor the Manganese it is difficult to determine the activation energy, especially when theferrous concentration increases because this metal precipitates or/and gets absorbed quicklyonto the oxides.Low ferrous sulphate concentration:142Chemical Control model activation energy calculation: 39.9 kJ/molDiffusion in the solid layer model activation energy calculation: 58.3 kJ/mol143In this case, in the beginning there appears to be one type of control, later followed by adifferent type (this suggestion is due to the kinetics curve behaviour). However, the activationenergy calculation in both cases coincides with chemical control.Diffusion in the solid layer model activation energy calculation: 60.0 kJ/mol(activation energy value within chemical reaction control range)144Activation energy in the presence of ammonium chloride1. CobaltIntermediate ferrous sulphate concentration145The diffusion in the liquid film and chemical reaction models are the best fist for cobalt inammonium chloride. However, after the activation energy calculation the following isobtained:146Diffusion in the liquid film model activation energy calculation: 10.0 kJ/mol (diffusion inthe liquid film control = mass transfer control)Chemical reaction control model activation energy calculation: 0.16 kJ/mol2. NickelIntermediate ferrous sulphate concentration147The diffusional and chemical reaction models are the best fist for nickel in ammoniumchloride. However, after the activation energy calculation the following is obtained:148Diffusion in the liquid film model activation energy calculation: 3.16 kJ/mol (diffusion onthe liquid film control = mass transfer control)Chemical reaction control model activation energy calculation: 0.17 kJ/mol3. ManganeseIntermediate ferrous sulphate concentration149The diffusional and chemical reaction models are the best fits for manganese in ammoniumchloride. However, after the activation energy calculation the following is obtained:150Diffusion in the liquid film model activation energy calculation: 19.8 kJ/mol (diffusion inthe liquid film control = mass transfer control).Chemical reaction control model activation energy calculation: 0.29 kJ/molIn ammonium chloride media it is important to study the controlling mechanism under highferrous sulphate concentration where the layer of magnetite could play a more important rolein the leaching kinetics especially for nickel extraction. Agitation in ammonium chloride doesnot show an important effect in recoveries at the values studied. In order to better assess theeffect of agitation it might be necessary to run experiments at a wider range of agitationspeeds.  Only manganese appears to be slightly affected by changes in agitation. One issueseen under high agitation speeds was increased losses of ammonia.Proposed simplified flowsheet151152Total ammonia/ammonium assays153154Experimental dataTable0A.8: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect oftemperature in the presence of ammonium sulphateTemperatureTime(h)20C 45C 65C 80CAddedNH4OH (ml) pH Eh(mV)Fe(g/l)AddedNH4OH (ml) pH Eh(mV)Fe(g/l)AddedNH4OH(ml) pH Eh(mV)Fe(g/l)AddedNH4OH (ml) pH Eh(mV)Fe(g/l)0.00 29.7 29.7 30.4 30.40.25 50.0 10.4 -616 28.5 10.0 9.91 -724 28.5 0.00 9.50 -745 22.5 0.00 8.30 -711 18.00.50 10.0 10.4 26.3 10.0 9.91 24.0 0.00 9.50 19.5 0.00 8.30 15.01.00 10.0 10.4 24.8 10.0 9.91 21.0 0.00 9.50 16.5 0.00 8.30 12.02.00 10.0 10.4 21.8 10.0 9.91 18.0 0.00 9.50 12.8 0.00 8.30 -652 7.503.50 45.0 10.4 -579 17.3 10.0 9.87 -695 14.3 0.00 9.46 -723 9.75 0.00 8.40 -605 6.754.00 10.0 10.4 15.0 0.00 9.87 11.3 0.00 9.46 7.50 0.00 8.40 6.006.00 5.00 10.4 13.5 0.00 9.87 8.25 0.00 9.46 6.00 0.00 8.40 5.258.50 10.0 10.4 -313 11.3 10.0 9.83 -533 7.50 0.00 9.37 -148 5.25 0.00 8.46 -230 5.2512.0 10.0 10.4 8.25 10.0 9.83 6.75 0.00 9.37 5.25 16.7 8.46 5.2524.0 0.00 10.3 -136 3.12 0.00 9.80 20 2.78 37.00 9.20 49 2.62 7.13 8.31 123 1.98Wash 0.02 0.01 0.01 0.01Table0A.9: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofagitation in the presence of ammonium sulphateAgitationTime(h)600 RPM 850 RPM 1200 RPMAdded NH4OH (ml) pH Eh(mV) Fe (g/l) Added NH4OH (ml) pH Eh(mV) Fe (g/l) Added NH4OH (ml) pH Eh(mV) Fe (g/l)0.00 32.6 29.7 32.60.25 50.0 8.00 -456 25.0 8.03 -480 0.00 8.00 -5100.50 0.00 8.00 0.00 8.03 39.0 8.001.00 0.00 8.00 0.00 8.03 50.0 8.002.00 0.00 8.00 0.00 8.03 50.0 8.003.00 36.0 8.06 -624 11.0 21.0 8.06 -536 6.4 45.0 7.83 -480 1.354.00 0.00 8.06 0.00 8.06 45.0 7.836.00 0.00 8.06 0.00 8.06 0.00 7.839.00 0.00 7.96 -595 0.0 24.0 7.94 -205 0.2 45.0 7.83 -102 0.0811.0 15.0 7.96 50.0 7.94 0.00 7.9024.0 50.0 7.86 -205 0.05 50.0 8.01 -44 0.05 0.00 7.90 67.0 0.08Wash 0.01 0.00 0.00155Table0A.10: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofsolids content in the presence of ammonium sulphateSolids content in the slurryTime(h)3.80% 10.7% 16.7% 23.1%AddedNH4OH (ml) pH Eh(mV)Fe(g/l)AddedNH4OH (ml) pH Eh(mV)Fe(g/l)AddedNH4OH(ml) pH Eh(mV)Fe(g/l)AddedNH4OH (ml) pH Eh(mV) Fe (g/l)0.00 29.7 32.6 7.03 54.8 81.80.25 10.0 7.98 -617 28.5 50.0 8.00 -456 30.0 8.03 -439 50.0 7.85 -5100.50 5.00 7.94 -625 26.3 0.00 8.00 47.0 8.03 -616 50.0 7.851.00 10.0 7.95 -635 24.8 0.00 8.00 0.00 8.03 45.0 7.852.00 0.00 7.91 -630 21.8 0.00 8.00 0.00 8.03 30.0 7.853.00 45.0 7.88 -630 17.3 36.0 8.06 -624 11.0 16.0 7.97 -597 7.96 61.0 7.85 -586 17.34.00 0.00 8.06 15.0 0.00 8.06 17.0 8.06 0.0 7.836.50 45.0 8.06 -617 13.5 0.00 8.06 0.00 8.06 20.0 7.839.50 50.0 7.77 -609 11.3 0.00 7.96 -595 0.0 15.0 7.94 -520 1.38 20.0 7.98 -600 7.412.0 50.0 7.96 8.25 15.0 7.96 41.0 7.94 60.0 7.9824.0 35.0 7.96 -7.00 3.12 50.0 7.86 -205 0.05 85.0 8.01 -60 0.02 45.0 7.98 -165 0.01Wash 0.00 0.01 0.00 0.00Table0A.11: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect oftemperature in the presence of ammonium chloride.TemperatureTime(h)20C 45C 65C 80CAddedNH4OH (ml) pH Eh(mV)Fe(g/l)AddedNH4OH (ml) pH Eh(mV)Fe(g/l)AddedNH4OH (ml) pH Eh(mV)Fe(g/l)AddedNH4OH (ml) pH Eh(mV)Fe(g/l)0.00 32.6 32.6 30.4 30.40.25 69.0 9.8 -419 30.0 77.0 9.91 -724 30.8 50.0 9.08 -565 29.0 40.0 8.00 -480 28.90.50 0.00 -423 22.0 0.00 9.91 22.0 50.0 25.0 45.0 25.51.00 0.00 -476 15.0 0.00 9.91 16.5 50.0 22.5 0.00 21.32.00 0.00 11.0 0.00 9.91 12.1 50.0 19.0 0.00 17.03.00 0.00 10.3 -494 9.77 0.00 9.87 -695 8.36 26.0 9.20 -338 17.5 26.0 7.86 -536 14.44.00 4.00 7.00 0.00 9.87 4.92 0.00 15.2 0.00 12.86.00 0.00 2.00 0.00 9.87 2.86 0.00 12.5 0.00 11.29.50 5.00 10.5 -102 0.08 5.00 9.83 -533 1.10 0.00 9.30 -100 10.0 39.0 7.85 -515 9.0012.0 0.00 0.10 5.00 9.83 1.10 30.0 9.00 50.0 4.2524.0 0.00 10.7 15.0 0.10 0.00 9.80 20 0.68 33.0 9.10 16.0 0.94 50.0 7.72 112 0.03Wash 0.02 0.02 0.02 0.00156Table0A.12: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofagitation in the presence of ammonium chlorideAgitationTime(h)600 RPM 900 RPM 1200 RPMAdded NH4OH(ml) pH Eh(mV) Fe (g/l) Added NH4OH (ml) pH Eh(mV) Fe (g/l)Added NH4OH(ml) pH Eh(mV) Fe (g/l)0.00 30.4 33.5 35.70.25 40.0 8.00 -480 28.9 30.0 7.83 -486 10.0 7.83 -5970.50 45.0 25.5 27.0 8.02 10.0 8.00 -6331.00 0.00 21.3 0.00 10.0 8.032.00 0.00 17.0 0.00 7.94 -502 10.0 8.033.00 26.0 7.86 -536 14.4 18.0 0.0 10.0 8.09 -647 50.04.00 0.00 12.8 15.0 10.0 8.066.00 0.00 11.2 15.0 10.0 8.01 -6418.00 39.0 7.85 -515 9.0 30.0 7.97 -230 0.0 30.0 7.94 48.212.0 50.0 4.25 30.0 30.0 7.9224.0 50.0 7.72 112 0.03 40.0 8.00 0.00 0.06 10.0 7.92 -591 1.00Wash 0.00 0.00 . 0.02Table0A.13: Experimental data (NH4OH added as a 28% NH3 in H2O solution). Effect ofagitation in the presence of ammonium chlorideSolids content in the slurryTime(h)10.1% 17.8% 24.1%Added NH4OH(ml) pH Eh(mV) Fe (g/l)Added NH4OH(ml) pH Eh(mV) Fe (g/l)AddedNH4OH (ml) pH Eh(mV) Fe (g/l)0.00 30.4 59.3 7 86.60.25 40.0 8.00 -480 28.9 30.0 8.00 -498 50.0 8.03 -4390.50 45.0 25.5 30.0 8.10 -513 50.0 8.03 -6161.00 0.00 21.3 5.00 8.00 25.0 8.032.00 0.00 17.0 35.5 8.00 0.00 8.033.00 26.0 7.86 -536 14.4 35.5 8.06 -584 16.9 10.0 7.97 -597 39.04.00 0.00 12.8 10.0 8.06 10.0 8.066.00 0.00 11.2 20.0 8.06 10.0 8.068.00 39.0 7.85 -515 9.00 20.0 8.04 -582 16.8 10.0 7.94 -520 25.512.0 50.0 4.25 20.0 7.96 50.0 7.9424.0 50.0 7.72 112 0.03 0.0 7.86 -180 0.22 50.0 8.01 -60 0.34Wash 0.00 0.01 0.01

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0220752/manifest

Comment

Related Items