"Applied Science, Faculty of"@en . "Materials Engineering, Department of"@en . "DSpace"@en . "UBCV"@en . "Malkhuuz, Ganbold"@en . "2010-01-12T22:15:29Z"@en . "2006"@en . "Master of Applied Science - MASc"@en . "University of British Columbia"@en . "The mixture of hydrochloric acid and magnesium chloride is a good lixiviant for the processing of sulfide minerals, concentrates and matte samples. The proton activity in this mixture deviates positively from ideal activity. This enhances the leaching power to break sulfide lattices down and to dissolve metals into solution. Magnesium (chloride) was chosen because it is one of a few reusable salts among alkali and alkali earth metal chlorides. The thermodynamic properties of this mixture are best characterized by activity coefficients of contributing ions in solution. The activity coefficient of hydrochloric acid in this mixture was measured at a total ionic strength of two at temperatures of 25, 35, and 45\u00C2\u00B0C applying the Electro Motive Force (EMF) measurement method. Further measurements at higher temperatures and higher ionic strengths were complicated due to unstable readings of the potentials. Therefore, a mathematical method published by Meissner was utilized to calculate the activity coefficients of hydrochloric acid and magnesium chloride in a mixture. Based on these calculations, individual ion activities were assigned using Bates\u00E2\u0080\u0099 equation and Jansz\u00E2\u0080\u0099s approach of applying variable water activities, hydration numbers and osmotic coefficients. Based on the individual ion activity, pH values were estimated for solutions where measurements were not applicable. As part of the thermodynamic studies of this mixture, the solubility limit of MgCl\u00E2\u0082\u0082 in hydrochloric acid solutions was investigated. The solubility of MgCl\u00E2\u0082\u0082 in water was measured as 485.6 g/l at 22\u00C2\u00B0C and 557 g/l at 82.5\u00C2\u00B0C. The solubility of MgCl\u00E2\u0082\u0082 in 6m HCl solutions was measured as 243 g/l at 22\u00C2\u00B0C and 452 g/l at 82.5\u00C2\u00B0C. Furthermore, the leaching chemistry of individual sulfide minerals-pyrite, millerite, troilite, heazelwoodite, violarite, and chalcopyrite were investigated in MgCl\u00E2\u0082\u0082-HC1 solutions. Pyrite was the most refractory mineral. About 6% iron was extracted in the mixture of 2m MgCl\u00E2\u0082\u0082 and 3-10m of HCl at 60\u00C2\u00B0C. Over 90% of iron was extracted from troilite in the mixture of 2m MgCl\u00E2\u0082\u0082 and 3m HCl. Millerite dissolved at acid concentrations greater than 6m HCl. At 60\u00C2\u00B0C, 60% of nickel was extracted in the mixture of 2m MgCl\u00E2\u0082\u0082 and 10m HCl. The dissolution results of these minerals were consistent with the thermodynamic predictions. About 20% of the nickel from violarite was extracted in mixtures with 1-6m HCI. The nickel extractions were increased up to 30% in mixtures of 10m HCl. About 60% of the heazelwoodite was dissolved in the mixture of 2m MgCl\u00E2\u0082\u0082 and 1m HCl. The heazelwoodite dissolution was consistent with the thermodynamic predictions described in section 2.2.6. In all above cases, the leaching time was 24 hours. Chalcopyrite partially leached (~22%) in the mixtures with 7m HCl at 100\u00C2\u00B0C. It dissolved forming cupric chloride, ferric chloride and the hydrogen sulfide gas (RXN 4.4 in section 4.4.3). No phase transformation (copper enriched product such as covellite) was observed. The results of individual mineral leaching experiments suggested the possibilities and conditions to process commercial sulfide products in this mixture. Two sulfide concentrates and a matte sample supplied by BHP Billiton were studied. Low MgO concentrate that consists of mainly pentlandite and pyrrhotite yielded 95% Ni, 87% Fe, 81% Co and 58% Cu extractions in solutions of 8m HCl and 2m MgCl\u00E2\u0082\u0082 with a retention time of 2 hours. The solid residue in this case contained mainly pyrite, talc and quartz. The high MgO concentrate that consists of mainly pentlandite yielded 95% Ni, 84% Fe, 75% Co and 19% Cu extractions in the mixtures of 6m HCl and 2m MgCl\u00E2\u0082\u0082 at 100\u00C2\u00B0C. The leach time was one hour. The leach residue in this case contained mostly quartz and talc. The addition of 0.5m cupric or ferric chloride to leach solutions of low MgO concentrate did not improve metal extractions due to the formation of copper and sulfur enriched layers on the particle surface. The reason is explained by the strong tendency of cupric or ferric ions to react with product gases such as H\u00E2\u0082\u0082S and H\u00E2\u0082\u0082 forming copper sulfide or elemental sulfur, respectively. The additions of either cupric or ferric chlorides to leach solutions of high MgO concentrate leaching also retarded metal dissolution. The reason of this low metal recovery is believed to be a formation of sulfur and oxidized layers on the surface of the particles. This low metal extraction is also explained by the same phenomenon as above in the case of ferric addition. The pentlandite, which is the main composition of the feed, remained substantially unleached. Nickel matte that mainly consisted of heazelwoodite yielded over 99% metal extractions in 6m acid solutions; however, the same metal extractions were obtained in 3m HCl mixtures with the exception of copper. The leach residue, where the highest metal extractions were obtained, consisted of 60% suredaite (a mixture of arsenic, sulfur and copper) in addition to 20% sulfur, according to the XRD and SEM-EDX. Low copper extraction from this sample was caused by the strong tendency of cupric ion to precipitate in the presence gases from heazelwoodite dissolution."@en . "https://circle.library.ubc.ca/rest/handle/2429/18061?expand=metadata"@en . "T H E USE OF STRONG BRINE AND HC1 SOLUTIONS TO PROCESS NICKEL SULFIDE CONCENTRATES by GANBOLD MALKHUUZ B . S c , Colorado School of Mines, 2000 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T F O R T H E D E G R E E OF M A S T E R ' O F A P P L I E D S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S Materials Engineering T H E UNIVERSITY O F BRITISH C O L U M B I A October 2006 \u00C2\u00A9 Ganbold Malkhuuz, 2006 A B S T R A C T The mixture of hydrochloric acid and magnesium chloride is a good lixiviant for the processing of sulfide minerals, concentrates and matte samples. The proton activity in this mixture deviates positively from ideal activity. This enhances the leaching power to break sulfide lattices down and to dissolve metals into solution. Magnesium (chloride) was chosen because it is one of a few reusable salts among alkali and alkali earth metal chlorides. The thermodynamic properties of this mixture are best characterized by activity coefficients of contributing ions in solution. The activity coefficient of hydrochloric acid in this mixture was measured at a total ionic strength of two at temperatures of 25, 35, and 45\u00C2\u00B0C applying the Electro Motive Force ( E M F ) measurement method. Further measurements at higher temperatures and higher ionic strengths were complicated due to unstable readings of the potentials. Therefore, a mathematical method published by Meissner was utilized to calculate the activity coefficients of hydrochloric acid and magnesium chloride in a mixture. Based on these calculations, individual ion activities were assigned using Bates' equation and Jansz's approach of applying variable water activities, hydration numbers and osmotic coefficients. Based on the individual ion activity, p H values were estimated for solutions where measurements were not applicable. A s part of the thermodynamic studies of this mixture, the solubility limit of MgCl2 in hydrochloric acid solutions was investigated. The solubility of M g C l 2 in water was measured as 485.6 g/1 at 22\u00C2\u00B0C and 557 g/1 at 82.5\u00C2\u00B0C. The solubility o f M g C l 2 in 6m HC1 solutions was measured as 243 g/1 at 22\u00C2\u00B0C and 452 g/1 at 82.5\u00C2\u00B0C. Furthermore, the leaching chemistry of individual sulfide minerals-pyrite, millerite, troilite, heazelwoodite, violarite, and chalcopyrite were investigated in M g C l 2 - H C l solutions. Pyrite was the most refractory mineral. About 6% iron was extracted in the mixture of 2m M g C l 2 and 3-10m of HC1 at 60\u00C2\u00B0C. Over 90% of iron was extracted from troilite in the mixture of 2m MgCl2 and 3m HC1. Millerite dissolved at acid concentrations greater than 6m HC1. A t 60\u00C2\u00B0C, 60% of nickel was extracted in the mixture of 2m M g C l 2 and 10m HC1. The dissolution results of these minerals were consistent with the thermodynamic predictions. About 20% of the nickel from violarite was extracted in mixtures with l -6m HC1. The nickel extractions were increased up to 30% in mixtures of 10m HC1. About 60% of the heazelwoodite was dissolved in the mixture of 2m M g C l 2 and l m HC1. The heazelwoodite dissolution was consistent with the thermodynamic predictions described in section 2.2.6. In all above cases, the leaching time was 24 hours. Chalcopyrite partially leached (-22%) in the mixtures with 7m HC1 at 100\u00C2\u00B0C. It dissolved forming cupric chloride, ferric i i chloride and the hydrogen sulfide gas ( R X N 4.4 in section 4.4.3). N o phase transformation (copper enriched product such as covellite) was observed. The results of individual mineral leaching experiments suggested the possibilities and conditions to process commercial sulfide products in this mixture. Two sulfide concentrates and a matte sample supplied by B H P Bi l l i ton were studied. L o w M g O concentrate that consists of mainly pentlandite and pyrrhotite yielded 95% N i , 87% Fe, 81%> Co and 58%> C u extractions in solutions of 8m HC1 and 2m M g C ^ with a retention time of 2 hours. The solid residue in this case contained mainly pyrite, talc and quartz. The high M g O concentrate that consists of mainly pentlandite yielded 95% N i , 84% Fe, 75% Co and 19% C u extractions in the mixtures of 6m HC1 and 2m MgCL. at 100\u00C2\u00B0C. The leach time was one hour. The leach residue in this case contained mostly quartz and talc. The addition of 0.5m cupric or ferric chloride to leach solutions of low M g O concentrate did not improve metal extractions due to the formation of copper and sulfur enriched layers on the particle surface. The reason is explained by the strong tendency of cupric or ferric ions to react with product gases such as H2S and H 2 forming copper sulfide or elemental sulfur, respectively. The additions of either cupric or ferric chlorides to leach solutions of high M g O concentrate leaching also retarded metal dissolution. The reason of this low metal recovery is believed to be a formation of sulfur and oxidized layers on the surface of the particles. This low metal extraction is also explained by the same phenomenon as above in the case of ferric addition. The pentlandite, which is the main composition of the feed, remained substantially unleached. Nickel matte that mainly consisted of heazelwoodite yielded over 99% metal extractions in 6m acid solutions; however, the same metal extractions were obtained in 3m HC1 mixtures with the exception of copper. The leach residue, where the highest metal extractions were obtained, consisted of 60% suredaite (a mixture of arsenic, sulfur and copper) in addition to 20% sulfur, according to the X R D and S E M - E D X . L o w copper extraction from this sample was caused by the strong tendency of cupric ion to precipitate in the presence gases from heazelwoodite dissolution. i i i T A B L E O F C O N T E N T S A B S T R A C T ii T A B L E O F C O N T E N T S iv LIST O F T A B L E S vii LIST O F FIGURES ix C H A P T E R 1 INTRODUCTION 1 C H A P T E R 2 L I T E R A T U R E R E V I E W 3 2.1 Nickel-cobalt production 3 2.1.1 General.:..: \u00E2\u0080\u00A2. 3 2.1.2 Processing Options 3 2.1.2.1 Sulfide Processing 4 2.1.2.2 Laterite Processing 5 2.1.3 Processes and Plant Practices.... 6 2.1.3.1 Ammonia Pressure Leach: The Sherritt Gordon Process 6 2.1.3.2 Ammonia-Ammonium Carbonate Atmospheric Leach: The Queensland Nickel Laterite Treatment Plant [87] 7 2.1.3.3 Total Pressure Oxidation P L A T S O L Process [84] 8 2.1.3.4 The Activox Process at Tati Nickel Mine in Botswana 10 2.1.3.5 Additional Processes for Sulfide Concentrates 11 2.1.3.6 High Pressure A c i d Leaching: M o a Bay Project 12 2.1.3.7 Cawse Project 13 2.1.3.8 Bulong Project 14 2.1.3.9 Mur r inMur r in ; ...15 2.1.3.10 Inco Laterite Process 16 2.1.3.11 Chloride Leach Process: Falconbridge Refinery 17 2.1.3.12 Atmospheric A c i d Chloride Leach ( A A L ) Process: Sechol Project 19 2.1.4 Technology Summary: Shortcomings of the Existing Technologies 21 2.1.5 Why M g C b ? Uniqueness of this mixture 22 2.2 Thermodynamics of aqueous chloride media 24 2.2.1 General 24 2.2.2 Fundamental expressions of aqueous chloride media 25 2.2.3 Estimation o f activity coefficients in aqueous chloride media 26 2.2.3.1 Estimation methods of activity coefficients in aqueous chloride media 26 2.2.3.2 Estimating activity coefficients by Meissner method 29 2.2.3.3 Experimental methods to estimate activity coefficients: E M F measurement 32 2.2.4 Corresponding water activity or osmotic coefficient 34 2.2.4.1 Expressions for water activity and osmotic coefficient 34 2.2.4.2 Estimating osmotic coefficients of HC1 and MgCl2 35 2.2.5 Individual ionic activities 36 2.2.5.1 Individual ionic activities in single/mixed electrolytes 36 2.2.5.2 Estimation of individual ionic activities and p H of a mixture 39 2.2.6 Thermodynamic predictions 44 2.2.7 Chloride complexes 47 iv 2.2.8 Solubility of chloride salts 50 2.2.9 Effect o f A1C1 3, N a C l and C a C l 2 on activities o f M g C l 2 + H C l mixture 51 2.2.10 Summary of thermodynamic review 54 C H A P T E R 3 E X P E R I M E N T A L ASPECTS 56 3.1 Experimental procedures and methods 56 3.2 Experimental instruments and set-up 57 3.3 Chemicals, minerals and concentrate samples 59 3.4 Solution preparation. 61 3.5 The basic principles of the S E M - E D X , X R D analyses 62 C H A P T E R 4 R E S U L T S AND DISCUSSIONS 64 4.1 Thermodynamic measurement 64 4.1.1 E M F measurement of the cell M g C l 2 - H C l - H 2 0 using P t / H 2 - A g / A g C l electrodes with no junction 64 4.1.2 Solubility of M g C i 2 in water and HC1 solutions. 67 4.1.3 Summary o f the thermodynamic measurements 68 4.2 Individual mineral leaching 69 4.2.1 Pyrite(P) 69 4.2.2 Milleri te (M) 70 4.2.3 Violarite (V) 72 4.2.4 Troilite (T) 74 4.2.5 Heazelwoodite (H) 75 4.2.6 Chalcopyrite (C) 77 4.2.7 Summary of individual mineral leaching 78 4.3 Commercial concentrates and matte leaching 79 4.3.1 L o w M g O concentrate 79 4.3.2 High M g O concentrate 81 4.3.3 Nickel matte 84 4.4 X R D , S E M - E D X analyses 87 4.4.1 Pyrite and leach residue 87 4.4.2 Millerite and leach residue 89 4.4.3 Violarite and leach residue 90 4.4.4 Chalcopyrite and leach residue 91 4.4.5 L o w M g O concentrate and its leach residue 93 4.4.6 High M g O concentrate and its leach residue 97 4.4.7 Nicke l matte and its leach residue 103 4.5 Summary of commercial concentrates leaching and X R D , S E M analyses 105 C H A P T E R 5 CONCLUSIONS AND R E C C O M M E N D A T I O N S 107 5.1 Conclusions 107 5.2 Recommendations 108 References 109 V Appendix 1 Certificates of analysis 116 Appendix 2 Summary of thermodynamic calculations 126 Appendix 3 Experimental aspects: Procedures and methods 135 Appendix 4 Mass balances of selected tests 144 vi LIST O F T A B L E S Table 2.1 Laterite Processing Plants based on the Caron process [1, 2,13] 5 Table 2. 2 Laterite Processing Plants based on H P A L [1, 2,13] 6 Table 2. 3 Hydration numbers of some simple salts and ions [28] 37 Table 2. 4 Hydration number at infinite dilution (ho) and (3 values [7] 38 Table 2. 5 Calculated hydrogen ion activity based on individual ion calculations 40 Table 2. 6 Comparison of pH values based on the two different approaches 40 Table 2. 7 Estimated pH of solution mixtures at temperatures 41 Table 2. 8 Calculated individual ionic activity coefficients for HC1 in relation to molal concentration and temperature 42 Table 2. 9 Calculated individual ionic activity coefficients for M g C h in relation to molal concentration and temperature 42 Table 2.10 Calculated individual ion activity coefficients for a mixture of HCI-MgCh in relation to molal concentrations of each species at 25\u00C2\u00B0C 43 Table 2.11 The A G 0 and estimated pH values for the dissolution of selected sulfide minerals 44 Table 2.12 Strengths of chloro-complexes according to periodicity [28] 47 Table 2.13 Resume of the common chloro-complexes [6] 48 Table 2.14 A G 0 reaction of selected chloride complexes at 25\u00C2\u00B0C, kJ/mole 50 Table 2.15 Effect of added salts on calculated HC1 activities at 25\u00C2\u00B0C 52 Table 3.1 Summary of mineral and concentrate samples 60 Table 3. 2 Solution preparation example for the E M F measurement 61 Table 3. 3 Solution preparation example for leaching tests 61 Table 4.1 Solution compositions, measured and corrected E M F , calculated activity coefficients for HC1 at 25\u00C2\u00B0C 64 Table 4. 2 Harned equation coefficients 65 Table 4. 3 Summary of M g C h solubility [g/1] in acid solutions 67 Table 4.4 Assays and mineralogical compositions of millerite sample 70 Table 4. 5 Dissolution of metals from millerite leaching at 60\u00C2\u00B0C 71 Table 4. 6 Assays and mineralogical compositions of violarite sample 72 vii Table 4. 7 Ni & Fe extractions from violarite sample at 60 C 73 Table 4. 8 Assays and mineralogical compositions of troilite sample 74 Table 4.9 Assay of heazelwoodite sample, % 75 Table 4.10 Summary of heazelwoodite leaching at 60\u00C2\u00B0C 76 Table 4.11 Assays and mineralogical compositions of chalcopyrite sample 77 Table 4.12 Summary results of chalcopyrite dissolution 77 Table 4.13 Metal extractions from matte sample at 60\u00C2\u00B0C 86 Table 4.14 Metal extractions from matte sample at 100\u00C2\u00B0C 86 Table 4.15 Composition of pyrite leach residue (10m HCl-2m MgCI 2 at 60\u00C2\u00B0C) 88 Table 4.16 Detailed mineral assay of low MgO concentrate (supplied from BHP Billiton).. 93 Table 4.17 Contents of low MgO concentrate residue (8m HCl-2m MgCl 2 ) 95 Table 4.18 Contents of low MgO concentrate leach residue (6m HCl-0.5m MgCl2-0.5m C u C l 2 a t l 0 0 \u00C2\u00B0 C ) 97 Table 4.19 Compositions of high MgO concentrate leach residues (6m HCl-2m M g C l 2 at 100\u00C2\u00B0C) 99 Table 4. 20 Compositions of high MgO concentrate leach residue (6m HCl-2m M g C l 2 -0.05m CuCl 2 at 100\u00C2\u00B0C) 101 Table 4. 21 Compositions of high MgO concentrate leach residue: (6m HCl-2m M g C l 2 -0.2m C u C l 2 at 100\u00C2\u00B0C) 102 Table 4. 22 Compositions of high MgO concentrate leach residue: (6m HCl-2m M g C l 2 -0.2m FeCl 3 at 100\u00C2\u00B0C) 103 Table 4. 23 Matte leaching solid residue compositions 105 viii LIST O F FIGURES Figure 2.1 Ammonia-Ammonium Carbonate Atmospheric Leach Flowsheet [87] 8 Figure 2. 2 Proposed P L A T S O L flowsheet for Cu-Ni-Co-PGM sulfide concentrates from NorthMet Mine [84] 9 Figure 2. 3 The Activox Process Flowsheet at Tati Nickel Mine [85] 10 Figure 2. 4 Pressure Acid Leaching Moa Bay Flowsheet [14] 12 Figure 2. 5 High Pressure Acid Leaching Cawse Flowsheet [87] 14 Figure 2. 6 High Pressure Acid Leaching Bulong Flowsheet [87] 15 Figure 2. 7 High Pressure Acid Leaching Murrin Murrin Flowsheet [87] 16 Figure 2. 8 High Pressure Acid Leaching Goro Flowsheet [87] 17 Figure 2. 9 Falconbridge Chloride Leaching Refinery Flowsheet [13] 18 Figure 2.10 Sechol Flowsheet for Laterite Treatment at Atmospheric Pressure [11,12] ...20 Figure 2.11 Calculated yHci of a mixture having I compared with experimental data from E M F measurement at temperatures [43] 31 Figure 2.12 Calculated activity coefficient vs. reference data [42] at 25\u00C2\u00B0C for single electrolytes 35 Figure 2.13 Calculated osmotic coefficients vs. reference data [42] at 25\u00C2\u00B0C for single electolytes 36 Figure 2.14 Predicted pH for the dissolution of selected minerals 45 Figure 2.15 Eh-Iog[Cl] diagram at 25\u00C2\u00B0C [28] 49 Figure 2.16 Effects of AIC13, NaCl and C a C l 2 on activities of MgCl 2 +HCl mixture 53 Figure 3.1 Experimental equipment set-up 57 Figure 3. 2 Experimental set-up for thermodynamic measurement 58 Figure 3. 3 Experimental set-up for leaching test 58 Figure 4.1 Log ynci values: experimental, calculated by the Harned equation vs. reference and the calculated values by the Meissner method 66 Figure 4. 2 Solubility dependence of M g C l 2 on acidity and the temperature 67 Figure 4. 3 Pyrite leaching: Effects of acid concentration, temperature and time 69 Figure 4. 4 Pyrite leaching: Effect of M g C l 2 concentration at 60\u00C2\u00B0C 69 Figure 4. 5 Millerite leaching: Ni extraction 71 Figure 4. 6 Millerite leaching: Fe extraction 71 ix Figure 4. 7 Violarite leaching: Ni extraction (2m MgCl 2 ) 73 Figure 4. 8 Violarite leaching: Co extraction 73 Figure 4. 9 Violarite leaching: Fe extraction 73 Figure 4.10 Troilite leaching: Fe extraction at 25\u00C2\u00B0C (2m MgCl 2 ) 75 Figure 4.11 Troilite leaching: Cu extraction at 25\u00C2\u00B0C 75 Figure 4.12 Troilite leaching: Fe extraction at 25 & 60\u00C2\u00B0C 75 Figure 4.13 Heazelwoodite leaching: Ni extraction (2m MgCI 2) 76 Figure 4.14 Chalcopyrite leaching: Metal extractions 77 Figure 4.15 Low MgO concentrate: Effect of leach time (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) 80 Figure 4.16 Effect of acid concentration (2m M g C l 2 , t=2 & 4 hr at 100\u00C2\u00B0C) 80 Figure 4.17 Effect of C u C l 2 addition (6m HC1, t=4 hr at 100\u00C2\u00B0C) 81 Figure 4.18 Effect of FeCl 3 addition (6m HC1, t=4 hr at 100\u00C2\u00B0C) 81 Figure 4.19 The dissolution of magnesium (6m HC1, t=4 hr at 100\u00C2\u00B0C) 81 Figure 4. 20 High MgO concentrate: Effect of leaching time (6m HC1 - 2m MgCI 2 at 100\u00C2\u00B0C) 82 Figure 4. 21 Effect of acid concentration (2m M g C l 2 , t=lhr at 100\u00C2\u00B0C) 82 Figure 4. 22 Effect of C u C l 2 addition (6m HC1 - 2m M g C l 2 , t=lhr at 100\u00C2\u00B0C) 82 Figure 4. 23 Effect of FeCl 3 addition (6m HC1 - 2m M g C l 2 , t=lhr at 100\u00C2\u00B0C) 82 Figure 4. 24 Dissolution of magnesium (6m HC1 - 2m M g C l 2 , t=lhr at 100\u00C2\u00B0C) 82 Figure 4. 25 Matte leaching: Ni extraction at 60\u00C2\u00B0C (2m MgCl 2 ) 85 Figure 4. 26 Matte leaching: Co extraction at 60\u00C2\u00B0C (2m MgCl 2 ) 85 Figure 4. 27 Matte leaching: Cu extraction at 60\u00C2\u00B0C (2m MgCl 2 ) 85 Figure 4. 28 Matte leaching: Fe extraction at 60\u00C2\u00B0C (2m MgCl 2 ) 85 Figure 4. 29 Matte leaching: Metal extractions at 100\u00C2\u00B0C (2m MgCl 2 ) 85 Figure 4. 30 X-ray patterns of pyrite and pyrite leach residue (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) 87 Figure 4. 31 S E M picture of pyrite leach residue (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) 88 Figure 4. 32 X-ray patterns of millerite and its leach residue (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) 89 Figure 4. 33 X R D patterns for violarite mineral sample 90 Figure 4. 34 Compared X R D patterns for violarite and leach residue 91 Figure 4. 35 X-ray patterns of chalcopyrite and its leach residue (7m HCl-2m M g C l 2 at 100\u00C2\u00B0C) 92 X Figure 4. 36 X-ray patterns of low MgO concentrate 93 Figure 4.37 Overlapped X-ray patterns of low MgO concentrate and its residue (8m HCl-2mMgCl 2 ) 94 Figure 4. 38 X-ray pattern of low MgO concentrate residue (8m HCI-2m M g C l 2 at 100\u00C2\u00B0C) 94 Figure 4. 39 S E M image of low MgO concentrate leach residue (8m HCl-2m M g C l 2 at 100\u00C2\u00B0C) 95 Figure 4. 40 X-ray pattern of low MgO concentrate and its leach residue (6m HCl-0.5m MgCl 2-0.5m CuCl 2 ) 96 Figure 4. 41 S E M of low MgO concentrate leach residue (6m HCl-0.5m MgCl 2-0.5m C u C l 2 a t l 0 0 \u00C2\u00B0 C ) 97 Figure 4. 42 X-ray patterns of high MgO concentrate and its leach residue (6m HC1 and 2m M g C l 2 a t l 0 0 \u00C2\u00B0 C ) 98 Figure 4. 43 S E M image of high MgO concentrate leach residue (6m HC1 and 2m M g C l 2 at 100\u00C2\u00B0C) 98 Figure 4. 44 S E M image of high MgO concentrate leach residue: Selected spots 99 Figure 4. 45 X-ray patterns of high MgO concentrate and its leach residue (6m HCl-2m MgCl 2-0.05m C u C l 2 at 100\u00C2\u00B0C) 100 Figure 4. 46 S E M image of high MgO concentrate: (6m HCl-2m MgCl 2-0.05m C u C l 2 at 100\u00C2\u00B0C) 100 Figure 4. 47 S E M image of high MgO concentrate leach residue: (6m HCl-2m M g C l 2 -0.2m C u C l 2 at 100\u00C2\u00B0C) 102 Figure 4. 48 Compositions of high MgO concentrate leach residue (6m HCI-2m MgCI 2 -0.2m FeCI 3 at 100\u00C2\u00B0C) 102 Figure 4. 49 X-ray patterns for matte and its leach residue (6m HCl-2m M g C l 2 at 100\u00C2\u00B0C) : 103 Figure 4. 50 S E M picture of matte leaching residue (6m HCl-2m M g C l 2 at 100\u00C2\u00B0C) 104 Figure 4. 51 S E M picture of matte leaching residue: Selected spots 104 xi A C K N O W L E D G E M E N T S I would like to express my sincere gratitude to my supervisor Dr. David Dreisinger for his enormous support o f my studies from the beginning to the end. This would never come to completion without his encouragement and thoughtful guidelines throughout these years. I am grateful for the opportunity to be a graduate student under his supervision at U B C . I would to extend my thanks to Dr. Eric Roche, Principal Research Scientist at the B H P Newcastle Technology Center in Australia, for his coordination of my research and encouragement. Special thanks are given to Dr. Berend Wassink for his everyday assistance in the lab, and to our hydrometallurgy group fellows, whom I have developed a good friendship with during these years, for always being ready to discuss my problems and giving me a hand when necessary. The other faculty and staff members at the Materials Engineering Department of U B C are friendly and welcoming, it is a nice community. Furthermore, the University of British Columbia is a great institution academically and an enjoyable place socially. The author also would like to acknowledge B H P Bil l i ton, Australia for their financial support and providing concentrate and matte samples and related mineral analyses. Finally, I would like to thank to my wife Khulan, son Bilguun and baby girl Nomuun for being \"brave\" by themselves not interrupting my studies at any time and supporting me as always. X l l C H A P T E R 1 INTRODUCTION Worldwide nickel production comes from two distinct natural resources: sulfide ores and laterite ores. Pyrometallurgy or hydrometallurgy have been utilized individually or in combination for the production o f nickel and associated metals from these resources. The sulfide ore processing consist of, traditional ore dressing to a concentrate, and the concentrate smelting to matte. Two pyrometallurgical processes have been commercialized to treat nickel laterites; smelting to ferronickel, and smelting in the presence of a sulfur-containing compound to produce nickel matte. Further, nickel matte may be processed hydrometallurgically to final metal products. Pyrometallurgy is well proven in the field of sulfide concentrate processing; however, it has limited applicability for processing low-grade and dirty concentrates due to energy cost and environmental restrictions. Moreover, it requires high capital cost. On the other hand, hydrometallurgical processing options for base metal production have several advantages when compared with pyrometallurgy. This route eliminates air pollution and yields higher product quality at lower capital and operating costs. Because of these advantages, two principal hydrometallurgical processes for the recovery of nickel from oxidized ores have been commercialized: Caron process and the High Pressure A c i d Leach ( H P A L ) process. The Caron process is preceded by a reduction roasting process. Five plants that utilize the Caron process were commercialized between 1952-1986. The major disadvantage of the Caron based processes is its front-end ore drying and calcining stage, which makes the process intensive with respect to energy cost. The first pressure acid leaching ( P A L ) plant for laterites was commercialized in 1959 in Cuba followed by three H P A L plants in Western Australia in mid 90's. Two new plants are expected in New Caledonia (Inco) and Australia (BHP Billi ton) for laterite resources. For sulfide concentrates, the processes such as P L A T S O L , C E S L and Activox are well developed. A l l these are based on pressure leaching. In addition, Inco is planning a pressure-leaching route for Voisey Bay's sulfide concentrate. The pressure leaching processes work reasonably well for most of the oxide ores and sulfide concentrates. The major disadvantage of these processes lies on the large amount of acid added (laterites) or produced (sulfides). Both solid residue and solution from leaching are neutralized. Therefore, P A L processes are mostly used to treat limonite (low acid consuming) ores rather than saprolite (high acid consuming). Clay content (high A l and MgO). in an ore is the limiting factor of the P A L usage. The modern operations successfully utilize the advanced methods such as solvent extraction, ion exchange, and pyrohydrolysis for solution treatment. Extraction reagents that 1 operate in any media have been developed. The shortcomings of the existing hydrometallurgical processes rest on the autoclave leaching unit for acid leaching processes, and on the ore drying and reductions stages for Caron based processes. A s mentioned above, feed characteristics are limiting criteria for those pressure-leaching circuits. The high-pressure vessels are the most expensive, energy-consuming and non-standard equipment in any hydrometallurgical circuit. To overcome shortcomings of the existing technologies, the chloride-processing option has recently entered a period of renewed interest and investigation because of its advantageous leaching power. The strong leaching power eliminates the use of a pressure vessel and its related costs. Second, processing in chloride media tolerates high clay content (high A l ) and high acid consuming (high Mg) feed. A l contributes to the increase of the proton activities, whilst M g is leached and recovered in subsequent stages. A t an industrial level, chloride processes proved their viabilities economically and technically. Three refineries worldwide are already operating to process nickel sulfide mattes based on chloride processes. Outokumpu developed the Hydrocopper process to treat chalcopyrite concentrates in cupric-sodium-chloride media. In the case of nickel, Jaguar Nicke l developed a process to treat laterites in the mixture o f magnesium chloride and hydrochloric acid at atmospheric pressure. This mixture calls for specific attention because o f the unique property of M g C b to be decomposed and recycled. While many excellent research works have been done on chloride metallurgy, little attention has been paid to the mixture of hydrochloric acid and the magnesium chloride. Recently, Jaguar Nicke l proposed the possibility of applying this mixture to process laterite resources. The gap to apply this mixture to process sulfide sources remains open. This application is attractive because the process takes place at atmospheric pressure and the reagents are recyclable. Magnesium chloride solutions can be pyrohydrolyzed to reagents hydrochloric acid and magnesium oxide for reuse. Therefore, this research is focused on treating sulfide concentrates of nickel/cobalt in the mixture of magnesium chloride and hydrochloric acid at atmospheric pressure. More specifically, the scope of this work is as follows: 1. To characterize thermodynamic properties of M g C ^ - H C l mixture by experimental and modeling approaches; 2. To study the dissolution chemistry of selected sulfide minerals and concentrates in this mixture as a function of acid concentration and temperature. 2 C H A P T E R 2 L I T E R A T U R E R E V I E W This chapter consists of two parts. The first part reviews existing processes and plant practices for nickel production with more emphasis on chloride based technologies. The second part reviews thermodynamics of aqueous chloride system focused on the mixture of magnesium chloride and hydrochloric acid. 2.1 Nickel-cobalt production 2.1.1 General Nickel is produced from two distinct natural resources: sulfide ores or laterite (oxidized) ores. Cobalt is produced as a byproduct of copper in Africa and as a byproduct of nickel elsewhere. Sulfide ores are generally' deep and hard-rock type deposit, which is mostly mined by open-pit and underground mining techniques, followed by concentrating, smelting and refining. About 30% of the world's nickel resource on land is in sulfide ore bodies; however, it accounts for about 60% of world nickel production. Laterite ores are found where prolonged tropical weathering (laterization) of \"ultramatic\" rocks occurred, forming clay-rich \"soft\" deposit. This type of deposit accounts for over 70% of the nickel resources; however only 40% of the nickel production comes from this source [1,2]. Distinct processes have been applied for the production of nickel and cobalt from both of these resources. In addition, a number o f processing options are developed to treat sulfide concentrates and matte products. The following sections present a brief summary of representative processes and plant practices. The hydrometallurgical route, more specifically chloride media at atmospheric pressure, draws more emphasis to the use of the mixture o f magnesium chloride and hydrochloric acid. 2.1.2 Processing Options The processes for metal production from natural ores fall into two general categories: pyrometallurgy and hydrometallurgy. Either individual or combination routes have been applied in the metal industry. Historically, pyrometallurgical route dominated the industry of extracting metals from their sulfide sources; recent developments o f hydrometallurgy enabled the treatment, 3 not only oxide sources, but also sulfide products. The hydrometallurgical processes can be classified in terms of media as ammonia, sulfate and chloride. 2.1.2.1 Sulfide Processing Both pyrometallurgy and hydrometallurgy routes have been used for nickel sulfide processing. Pyrometallurgical processing of sulfide ore is carried out in three stages: concentrating, smelting and refining. The concentrating stage comprises of crushing, grinding and ore dressing. Ore is crushed in several stages and fed into grinding mills along with process water. The m i l l discharge-slurry is classified by particle size and the fine fraction is fed into flotation cells, where reagents (collector, modifier, frother, air etc) are added. Based on the hydrophobic or hydrophilic properties of particle surfaces, certain sulfide minerals are attached to bubbles, separated from gangue and unwanted sulfide minerals. The concentrate, containing about 12 to 20% of nickel (or about 30% of Ni+Cu in Canadian operations), is thickened, dried and sent to smelters for further processing. Dry concentrate is mixed with flux (silica sand) and fed into the smelting furnace. In the furnace, the mixture reacts with oxygenated hot air and oxidizes almost instantaneously. Afterward, the smelted mixture is collected in a settler where the molten metal sulfide (matte) and the slag are separated. Nicke l matte sinks to the bottom of the settler, whereas the slag floats on top. Nicke l matte contains about 75% N i - C u (40-70% N i , and 5-35% Cu), up to 24% S, and less than 1%) of Fe. Typical slag contains about 0.1-1.0% of N i . A converter maybe utilized to upgrade nickel content of the matte by blowing oxygen into the matte converting residual sulfides to metal. Up to this point, the sulfide ore is processed by a combination o f traditional routes of ore dressing and smelting. Further refining of matte may be done hydrometallurgically. Matte refining comprises re-grinding, leaching and reduction. A n example of a refining plant w i l l be discussed in part 2.1.3. Despite smelting routes, a number of hydrometallurgical processes have been developed for sulfide concentrates. Sherritt Gordon (now Corefco) of Canada developed and commercialized a pressure leaching process in ammonia-ammonium sulfate media in 1954. The plant is still in successful operation in Fort Saskatchewan, Alberta treating sulfide concentrate with high cobalt content [14, 15]. Similar plants have been built at Kwinana (Australia) and Springs (South Africa). Hydrogen gas is used to reduce nickel ions and the reduced product precipitates as nickel metal. In 4 addition, a number of processes such as the P L A T S O L process by Polymet Min ing Company, H I K O by O U T O K U M P U , Act ivox process by Western Mineral Technology Pty Ltd ( W M T ) , C E S L nickel process by Cominco Engineering Service Ltd, Sumitomo, Nippon (all high pressure) and BIONTC (atmospheric pressure) process by B H P Bi l l i ton have been developed. Recently I N C O announced the startup o f a hydrometallurgical demonstration plant (pressure leaching in mixed sulphate-chloride media) in Argentia, Newfoundland for Voisey Bay sulfide concentrate. These pressure-leaching processes are a new approach to eliminating the conventional smelting routes for the processing of nickel sulfide concentrates and laterite ores. 2.1.2.2 Laterite Processing Because of their complex mineralogical compositions, there is no distinct processing option for laterites [20]. A variety of flowsheets are developed, which fall into two general categories o f pyrometallurgy and hydrometallurgy. Two pyrometallurgical processes have been commercialized to treat nickel laterites; smelting to ferronickel, and smelting in the presence of a sulfur-containing compound to produce nickel matte. Since the objective of this research is centered on hydrometallurgy, the pyrometallurgical routes w i l l not be discussed further. Table 2.1 Laterite Processing Plants based on the Caron process [1,2,13] Operation Company Country Capacity ktNi/yr Product Period Nicaro Nicaro Nickel Co. Cuba 23 N i oxide 1952-present Surigao Refinery at Nonoc Marinduque Mining & Industrial Co. Philippines 35 Briquettes 1974-1986 Greenvale/Yabulu Freeport/Metals Exp. Queensland Nickel/BHP Billiton Australia 18 Briquettes 1974-present 10 Briquettes Niquelandia/Sao Paulo Companhia Niquel Tocantins Brazil 17.5 electronickel 1981-present Punta Gorda Union del Niquel Cuba 31.5 N i oxide 1986-present Two principal hydrometallurgical processes for the recovery o f nickel from oxidized ores have been commercialized: Caron process and the High Pressure A c i d Leach ( H P A L ) process. A patent for the Caron process was granted to Professor Caron in 1924. The process leaches metal in ammonium media at atmospheric pressure, after an ore has been reduced by roasting. Nicaro Nicke l Co., Cuba operated the first plant that utilizes the Caron process. Subsequently, four plants that utilize the modified Caron process were commissioned in the early 70's and 80's (Table 2.1). 5 The Nonoc flowsheet, a variation of the Caron process, was actually modified by Sherritt Gordon of Canada. These are the only operations based on the Caron process, and are still in operation, with the exception of Nonoc. A s a representative of these operations, the Queensland Plant w i l l be covered in section 2.1.3. The pressure acid leaching of laterite ores with low magnesia was developed by Freeport for the M o a Bay deposit in 1959. This process utilizes technology licensed by Sherritt Gordon. The M o a Bay operation is the precursor of all o f the modern day H P A L operations for laterite ores. K e y H P A L operations in the past and future are presented in the following table. More details of published resources w i l l be introduced shortly. Table 2. 2 Laterite Processing Plants based on H P A L [1, 2,13] Operation Company Country Capacity .ktNi/yr Product Period Moa Bay debotllenecked Freeport Sulfur General Nickel/Sherritt JV Cuba 25 Mixed sulfide 1959-present 5 2000-present Murrin Murrin Anaconda Australia 40 N i Briquettes 1999-present Cawse Centaur Australia 9 Electronickel 1998-present Bulong Resolute/Preston Resources Australia 7 Electronickel 1999-2003 Goro Inco Ltd. New Caledonia 60 N i Oxide Projected in 2007 Ravensthorpe BHP Billiton Australia 40 N i Hydroxide 2006 Startup 2.1.3 Processes and Plant Practices A l l processes and operations described below are aimed to treat nickel sulfide concentrates, sulfide mattes and laterite ores, hydrometallurgically. They generally fall into three categories in terms of leach media: ammonia, sulfate and chloride. 2.1.3.1 Ammonia Pressure Leach: The Sherritt Gordon Process The Sheritt Gordon process is the first hydrometallurgical operation in the world for treatment of nickel sulfides. The process treats sulfide concentrate, blend of high-grade nickel concentrate and nickel matte with only minor process modifications. The first application of this process was in Fort Saskatchewan, Alberta in 1954. Since 1991 the flowsheet at Fort Saskatchewan has been modified to refine mixed sulfide feed of M o a Bay, Cuba with high cobalt 6 content. The final products are nickel briquettes, cobalt briquettes and all sulfur is recovered as an ammonium sulfate for fertilizer use. The modern Sherritt Gordon process flowsheet is extensively discussed in the literature, and interested readers are guided to references 13-15 of this work. This process was adopted by Western Min ing Co. , Australia at Kwinana plant for the treatment of nickel sulfide concentrates in 1970. Due to low nickel price, high-energy cost and the limited marketability of fertilizer in Western Australia during that time, the Kwinana plant opted for treatment of nickel matte [10]. The plant in Canada is still in successful operation with minor modifications. The process is very energy intensive (200GJ/t N i from concentrate), which is a major issue in some geographical locations [15]. 2.1.3.2 Ammonia-Ammonium Carbonate Atmospheric Leach: The Queensland Nicke l Laterite Treatment Plant [87] Queensland Nicke l in Australia and Marinduque Min ing and Industrial Corporation in the Philippines commissioned plants that utilize the Caron process in the early 1970's. The Queensland Nicke l laterite treatment plant has had major changes to the Caron process flowsheet. Figure 2.1 illustrates the simplified flowsheet of the modern plant. In both flowsheets, the reduced laterite ore is quenched in ammoniacal ammonium carbonate solution and leached in aerated tanks. The original flowsheet included a hydrogen sulfide precipitation stage where virtually all cobalt and about 10% of the nickel were recovered as a mixed sulfide. The mixed product was dried and shipped for further separation of cobalt and nickel. Cobalt-free nickel pregnant solution from the previous precipitation stage is subjected to stripping by ammonia and carbon dioxide where basic nickel carbonate (BNC) precipitates. The B N C product is recovered by thickening and filtration, and the wet filter cake is dried and calcined in a rotary k i ln to yield nickel oxide as a final product. In the modern Queensland nickel plant flowsheet (Figure 2.1), a solvent extraction circuit was introduced to improve plant performance on nickel recovery. The raffinate from the nickel extraction feeds into the cobalt precipitation stage. Extracted nickel is stripped by strong ammonia solution followed by nickel oxide production by the conventional steam strip and calcination process. A more recent improvement in the process involves recovering the cobalt sulfide precipitate and after pressure leaching and solvent extraction with D 2 E H P A and L L X 8 4 Q N i , converting the cobalt to cobalt oxy-hydroxide CoO(OH). The Caron based process, however, works well ; the only concern that has risen is its front-end ore drying and reduction circuit. This high-energy consuming unit hinders, the economic 7 viability o f this process. Figure 2.1 Ammonia-Ammonium Carbonate Atmospheric Leach Flowsheet [87] CO+Hydrogen Laterite ore {- f JCoal/Fuel Oil Ore Drying/ Reduction Air 1 < e s Aeration leach wash liquor CCD residue wash Steam Pregnant leach liquor Absorber Steam solution Tailings still ammonia/ carbon^ dioxide NH, Pre-boil stills Ni SX LIX84QNI Absorbers/ condensers [StrongNH, Solutioir Ni strip CoS ppt Residue to tailings I Strip liquor Product stills I T CoS Product NiC0 3 solution Calcine NiO Reduction/ cintering Ni products 2.1.3.3 Total Pressure Oxidation P L A T S O L Process [84] Polymet Min ing Company developed the P L A T S O L process to treat copper-nickel-cobalt-P G M sulfide concentrate from the Northmet deposit in Minnesota. Figure 2.2 illustrates the simplified flowsheet of the P L A T S O L process. The process includes a pressure autoclave that allows high levels of base and precious metal extractions from concentrate under total oxidation 8 conditions at 225\u00C2\u00B0C and about 100 psig of oxygen overpressure. 10-20 g/1 HC1 is added to the leach circuit to extract P G M and A u as chloride complexes. The solids are separated from the leach solution and washed. The washed solid residue is sent to waste disposal and the solution advances to metal recovery. Figure 2. 2 Proposed P L A T S O L flowsheet for Cu-Ni-Co-PGM sulfide concentrates from NorthMet Mine [84] e o 01 cu > s U S o s 05 Excess acid & Fe Cu-Ni-Co-PGM mixed sulfide concentrate Total Pressure Oxidation 225\u00C2\u00B0C, 100 psig I S/L Separation SO, 3 Solid Wash Fe 3 + to Fe 2 + reduction NaSH-PGM Precipitation limestone 3 Neutralization of excess acid Bleed stream for Ni, Co r recovery I Oxyhydrolysis I Cu SX IT: Cu EW Cu removal MgO T Cu cathode Ni-Co Precipitaion Ni-Co'mixed hydroxide waste disposal PGM Precipitate -Gypsum The precious metals are selectively precipitated with the addition of sulfide ion. The solution is treated with limestone to neutralize excess acid followed by SX-EW to produce cathode copper. A major portion of the raffinate from copper extraction is recycled to the autoclave to 9 maintain the water balance and build nickel and cobalt composition. The balance of the solution is bled to nickel and cobalt recovery. This solution is treated by oxyhydrolysis where excess acid and iron are removed followed by copper removal. The purified solution is treated with magnesium oxide to precipitate mixed nickel and cobalt hydroxide. The commercial application of the P L A T S O L process at NorthMet is expected in 2008. 2.1.3.4 The Activox Process at Tati Nicke l Mine in Botswana Western Minerals Technology ( W M T ) has made a significant step to commercialize its patented (Activox) technology. W M T commissioned a hydrometallurgical demonstration plant at the Tati Nicke l Mine in Botswana in 2004. Figure 2. 3 The Activox Process Flowsheet at Tati Nickel Mine [85] Sulfide concentrate i H 2 SQ 4 Ultra fine grinding HC1,0, Activox Leach 100-110\u00C2\u00B0C, 1100 kPa I S/L Separation I Solid Wash To PGM Recovery Cu SX 2 stages of Fe removal I Na 2 C0 3 Co SX Cu EW Cu cathode Co Precipitation Co Product ammonia 1, J ^ ^ ^ ^ ^ ^ ^ ^ I A m t r w Quicklime Ni SX 3\u00C2\u00A3 Ammonia Recovery N i E W Ni cathode Steam Ammonia Stripping Tailings Dam 10 The sulfide concentrate is re-ground to P80=10um. The ultra-fine ground feed for the autoclaves is diluted from 50% solids to 30% solids by the addition of copper raffinate. Sulfuric acid, oxygen and sodium chloride are added into the autoclave to catalyze the oxidation of sulfide minerals. After leaching, solids and liquids are separated, and the solids are subjected to C C D washing followed by P G M recovery. Copper is first solvent extracted from solution followed by electrowinning to produce cathode copper. Raffinate from copper solvent extraction is partially recycled into the leaching unit, whereas the remaining solution advances to cobalt solvent extraction. The cobalt is precipitated by sodium carbonate and recovered. The cobalt free solution from Co solvent extraction advances to N i solvent extraction followed by N i electro-winning. The barren solution is subjected to ammonia recovery with lime; recovered ammonia is recycled to the nickel and cobalt extraction circuits. 2.1.3.5 Additional Processes for Sulfide Concentrates In addition, several other processes have been developed for the treatment of sulfide nickel concentrates. OutoKumpu (Finland) developed a process called H I K O that consists of nickel sulfide ore benefication and concentrate leaching stages. This process was used for treatment of nickel sulfide concentrates at Hitura mine [86]. Cominco Engineering Service Ltd. (CESL) developed flowsheets, in which Pressure A c i d Leach ( P A L ) was involved, for both laterite ores and sulfide concentrates [21]. The C E S L nickel process was developed based on the success of the C E S L copper process, and is aimed to treat nickel, cobalt and copper containing sulfide concentrate. The C E S L nickel process begins with a pressure leaching step to put nickel, cobalt and copper into solution. Pressure leaching takes place at 150\u00C2\u00B0C, in solutions containing up to 60g/l free acid with the addition of up to 12g/l free chloride. The leaching is followed by solution purification to remove copper and other impurities such as iron and zinc. The copper is then treated by S X - E W to cathode copper. The purified solution is treated by lime to co-precipitate nickel and cobalt. The mixed hydroxide is re-dissolved in an ammonium sulfate media to produce a relatively pure solution o f nickel and cobalt as diammines. Cobalt, and then nickel are separately recovered from the diammine solution by individual S X - E W circuits, producing the respective metals as cathodes. Furthermore, the Sumitomo process to treat nickel sulfide concentrate and the Nippon process for cobalt sulfides are developed [10, 13]. A l l o f these processes are based on pressure acid leaching. Inco or its subsidiary V B N C (Voisey Bay Nicke l Company) commissioned m i l l and 11 concentrating plant earlier this year at the Voisey Bay deposit. V B N C is also planning hydrometallurgical (pressure leaching circuit) treatment for its concentrate to make final products. The V B N C hydrometallurgy process is currently being tested at a demonstration plant at Argentia, Newfoundland. B H P Bi l l i ton developed the BIONIC process for nickel sulfide concentrates as an alternative technology to high-pressure technologies [16]. This technology relies on the biological oxidation of ferrous to ferric. Ferric ion then oxidizes the sulfide minerals. This eliminates the use of a pressure vessel. 2.1.3.6 High Pressure A c i d Leaching: M o a Bay Project High-pressure acid leaching of laterite ore with low magnesia has been in operation since 1959. Freeport Sulfur built the M o a Bay plant with the purpose of finding a market for sulfur. The M o a Bay operation has faced several flowsheet changes, difficulties because of political reasons, before finally entering a period of stable operation since forming joint venture with Sherritt International in 1994. The recent flowsheet of the M o a Bay operation is presented in Figure 2.4. Figure 2. 4 Pressure Acid Leaching Moa Bay Flowsheet [14] INE SERPENT REJECT! TAILINGS O/F M I N E ORE| Ore Preparation LIMONITE Ore Thickeners SteamllH2SQ4 U/F L E A C H Ni-Co SULPHIDES Product Thickeners ^ 1 O/F| J H 2 S Precipitation J C C D Washing 1 U/F GYPSUM SLURRY 0/F Neutral. * 1 Thickeners * 1 Neutralization WASTE LIQUOR C r 6 + , Fe 3 + ^ 1 Reduction LIME STONE A series of washing and screening steps rejects high acid consuming serpentine before leaching. This rejection is based on the particle sizes since most of acid consuming serpentine is present as bigger particles. The leaching takes place in vertical Pachuca type reactors at 250\u00C2\u00B0C and 45 atm pressure. Slurry is directed to a seven stage C C D , and raw liquor is treated with hydrogen sulfide gas to reduce C r 6 + , F e 3 + and to precipitate copper as copper sulfide. Clarified solution is subjected to acid neutralization by Coral limestone, followed by nickel and cobalt precipitation with addition of hydrogen sulfide gas. Thickened product is filtered, dried and shipped to Fort Saskatchewan, Alberta for further refining. 12 Modern operations for laterite treatment Successful operation of M o a Bay has resulted in several additional pressure leach operations. The modern day H P A L operations are practiced in Western Australia at Bulong (now closed), Cawse and Murr in Murrin to treat laterite ores. B H P Bi l l i ton is planning hydrometallurgical processing of the upgraded laterite ore at its Ravensthorpe nickel project. This project uses the Enhanced Pressure A c i d Leach process to produce a mixed nickel and cobalt hydroxide as an intermediate product. C V R D of Brazi l announced the development of the Vermelho Nickel Project. The project w i l l use a high-pressure acid leaching to process nickel laterite. Inco owns a nickel laterite deposit at Goro in New Caledonia and are planning to utilize pressure acid leaching circuit. A s representatives of these processes, Western Australian and I N C O operations are briefly summarized in the next sections. Unl ike M o a Bay, these operations utilize more recent technologies such as solvent extraction and electro-winning to produce cobalt and nickel as final products. 2.1.3.7 Cawse Project The Cawse project is located 45 km northwest of Kalgoorlie, Australia. The plant uses 250000 tons of sulfuric acid to produce 8500 tons of LME-grade nickel and 1700 tons of cobalt as cobalt sulfide annually. A c i d consumption at this plant is about 360 kg of sulfuric acid per ton of ore. The acid is provided by B H P Bil l i ton 's Kalgoorlie nickel smelter and imported sulfur. The flowsheet of Cawse project is shown below. The leaching process takes place at 250\u00C2\u00B0C under 45 atm pressure to dissolve nickel and cobalt. After autoclaving, most impurities are removed by adding limestone. Magnesia is added to precipitate the nickel and cobalt as mixed hydroxide slurry. Unt i l this point, the solution treatment takes place in a sulfate media. Then the mixed hydroxide is leached in ammonium media. The pregnant leach solution is contacted with the organic phase, which contains the LIX84 , to extract nickel. The nickel is recovered as a cathode product after stripping the loaded organic and electro-winning. The raffinate from the N i extraction is subjected to CoS precipitation by treating the solution with ammonium sulfide. 13 Figure 2. 5 High Pressure Acid Leaching Cawse Flowsheet [87] Ni-Co ore Crushing Grinding Laterite ore \u00E2\u0080\u00A2 Screen upgrade I Residue \u00E2\u0080\u00A2< M g O - | Lime\u00E2\u0080\u0094H N H , H C C D Pressure leaching @250\u00C2\u00B0C Fe Precipitation H 2 S Q 4 Limestone Ni/Co Hydroxide Precipitation Ammonia leach Ni SX LIX84 I N i E W Ni Cathode ( N H 4 ) 2 S - ^ Co Precipitation Co Sulphide The Cawse project underwent some initial difficulties at the startup and was sold to the U S O M G group. The O M G Company ships the mixed N i - C o product to their refinery in Finland. 2.1.3.8 Bulong Project The Bulong process involves high temperature, acid pressure leaching followed by solvent extraction and electrowinning for nickel recovery. The plant consumed about 250000 tons of sulfuric acid to produce 9000 tons of nickel and 700 tons of cobalt annually, before shut down. The leach solid residue was neutralized for returning to the tailing. The simplified flowsheet of Bulong is illustrated below. After removing some impurities by adjusting p H and Eh , the whole process stream is subjected to Co solvent extraction using Cyanex 272. After the organic is stripped, the solution is treated by hydrogen sulfide to reject M n , M g and Ca. Cobalt precipitates as CoS. This product is pressure leached in an acidic environment. Other impurities such as Z n and C u are dissolved with Co. Z n is removed by D 2 E H P A extraction and C u is removed by ion exchange. Cobalt is electro-won from the purified solution. The raffinate from cobalt solvent extraction goes to the N i extraction circuit with Versatic acid extractant followed by N i electrowinning. The Bulong 14 flowsheet experienced major operating problems. During cobalt extraction, Cyanex 272 extractant leaked into the Versatic acid solvent extraction circiut. Cyanex 272 extracts calcium at p H 6.5, resulting in gypsum precipitation and the fouling of the solvent extraction in the recovery of nickel. Figure 2. 6 High Pressure Acid Leaching Bulong Flowsheet [87] Laterite ore \u00E2\u0080\u00A2 Ore Preparation Gypsum & Fe/Zn waste* N H , Neutralization Pressure leaching C C D Co SX CYANEX272 Co, M n , Mg, Ca, Zn, Cu^r Ni SX Versatic acid Sulphide M n , Mg, C a ^ \u00E2\u0080\u0094 | precipitation I Acid redissolution of CoS cake O Ni E W N H , Zn I Co E W Zn SX D2EHPA I Cu Removal IX - H 2 S 0 4 \u00E2\u0080\u0094 Limestone Residue - N H 3 \u00E2\u0080\u00A2 Ni Cathode \u00E2\u0080\u00A2 Co Cathode \u00E2\u0080\u00A2 Cu 2.1.3.9 Murr in Murr in The Murr in Murr in process requires 500000 tpa of imported sulfur to produce 45000 tons o f nickel and 3000 tons of cobalt annually. Murr in Murr in uses an acid leaching process followed by sulfide precipitation to make a mixed sulfide intermediate product followed by nickel and cobalt refining utilizing oxygen leach, solvent extraction and hydrogen reduction to metal products. The plant produces 150000 tons o f ammonium sulfate by-product annually for fertilizer use. 15 Figure 2.1 High Pressure Acid Leaching Murrin Murrin Flowsheet [87] Laterite Ore Ore Preparation I Gypsum & Fe/Zn waste H S gas-Neutralization I Pressure leaching CCD H 2 S 0 4 Calcrete Residue Ni/Co sulphide precipitation Pressure oxidation leach H 2 gas. Co Reduction I Sintering Co briquettes H , Co SX CYANEX272 ^ Ni stream Ni Reduction J - N H 3 ^Ammonium sulphate Sintering Ni briquettes The processing principle of Murr in Murr in is very similar to Cawse. The P A L leaching is followed by several stages o f solution purification treating acidic solution, and C o / N i are precipitated with hydrogen sulfide gas. This precipitated product is leached in an acidic environment, unlike the Cawse flowsheet. The Co is extracted with Cyanex 272. Each stream of Co and N i is subjected to hydrogen reduction to recover individual metals. Technologically, all H P A L operations have proved their viability; however, economically they face constraints. Both ore leaching and slurry neutralization are performed with significant cost. 2.1.3.10 Inco Laterite Process Inco owns the nickel laterite deposit at Goro in New Caledonia. The Goro project involves H P A L circuit in its flow. Like H P A L operations in Western Australia, the solid residue is neutralized by lime and calcium carbonate before it is discharged to tailings. The leach solution is subjected to several stages of solution purification for acid neutralization, iron removal and C u is removed by ion exchange. 16 Figure 2. 8 High Pressure Acid Leaching Goro Flowsheet [87] Saprolitic or limonitic ore Feed Preparaion f^-^water 1 Pressure leaching H 2 S 0 4 C C D H 2 S 0 4 Cu Removal IX i. HC1 Ni & Co SX Cyanex301 water \u00E2\u0080\u00A2 Zn Removal IX i water Ni & Co SX Alamine 336 NiCl 2 pyrohydrolysis Final neutralization i C a C 0 3 Ca(OH). N a 2 C Q 3 Co precipitation P Tailings C o C O , Ni powder Co and N i are bulk extracted by Cyanex 301 from clarified solution. Cyanex 301 extracts nickel and cobalt without p H adjustment. Co and N i along Z n are stripped using HC1 solution. Co and N i are separated with Alamine 336 extractant. The N i containing stream is subjected to pyrohydrolysis where N i is recovered as nickel oxide and hydrochloric acid is produced and recycled to the stripping circuit. 2.1.3.11 Chloride Leach Process: Falconbridge Refinery Falconbridge and Societe le Nicke l (SLN) have developed the hydrometallurgical treatment for nickel matte refining in chloride media and commercialized at Kristiansand, Norway and Sandouville-Le Havre, France, respectively. The flow diagram o f Falconbridge refinery is presented in Figure 2.9. The matte is leached in chloride solution near its boiling point at atmospheric pressure. 17 Figure 2. 9 Falconbridge Chloride Leaching Refinery Flowsheet [13] Ni-Cu Matte Ni-Cu Chloride '^^Solution Ni-Cu Matte Chlorine Ni Leach Copper removal Chlorine Air Residue f Ni Carbonate Iron removal Ni/Co chloride solution Roaster Ni-Cu Matte h ri rfc^e'rV knoiyte Copper Leach P G l J residue Cu EW T Cobalt Strip Cobalt extraction TIO amine Ni carbonate Cobalt Chlorine Nickel purificatioi Purificatio Chlorine Co EW Cu Cathode t Co Cathode Chlorine \u00E2\u0080\u0094 \ Pb/Mn Ni EW Anolyte to leach Ni Cathode Copper, iron and any other impurity metals are removed and recovered ahead of C o / N i extraction. Purified C o / N i solution is subjected to solvent extraction with trioctylamine, where cobalt is extracted leaving a purified nickel pregnant solution. The nickel solution is purified again by chlorine and nickel carbonate to remove any trace elements, and N i is recovered by direct electrowinning. The cobalt loaded organic is stripped by lean electrolyte solutions of cobalt, and the rich electrolyte is purified prior to cobalt electrowinning. The copper content in the feed, which was leached and subsequently precipitated in solid residue, is recovered as a cathode copper in a separate circuit. The S L N flowsheet is principally the same as that of Falconbridge, except for the source of the chloride lixiviant solution. In the S L N flowsheet, a ferric chloride solution is recycled to leaching from iron removal, whereas N i / C u chloride solution is recycled from the treatment of C u containing solid residue in Falconbridge. In addition to these two plants, The Sumitomo refinery in Japan utilizes chloride media for the treatment of nickel matte. These three refineries are the only commercial operations of nickel matte refining in chloride media. 18 2.1.3.12 Atmospheric A c i d Chloride Leach ( A A L ) Process: Sechol Project Jaguar Nickel Inc., investigated hydrometallurgical processing of nickel laterite at the Sechol Laterite Project in Guatemala. This process utilizes atmospheric leaching of laterite ore in acidic chloride media. The flowsheet consists of four main stages: leaching in acidic chloride media, purification, precipitation and pyrohydrolysis. In the first stage, nickel and cobalt are substantially leached, rejecting most of the iron and magnesium. In the next two stages, the impurity elements, such as calcium, aluminum and residual iron are discarded, and nickel and cobalt are recovered. In the final stage, hydrochloric acid and magnesia are recovered by pyrohydrolysis. The inventors of this process claim that all process steps are successfully illustrated in the scale of commercial application. For instance, Falconbridge's refineries, as wel l as Quebec Iron and Titanium plant, are excellent examples of successful operations in chloride media at industrial scale. The purification and precipitation stages are conventional hydrometallurgical processes and pyrohydrolysis is practiced in the steel pickling industries [22]. According to the developers of this process, it can be low in capital and operating costs since no pressure vessels are required and acid is recovered and reused with significant low cost, unlike other processes such as those in Western Australia in which the acid is neutralized with significant cost. O f course, energy is required for pyrohydrolysis and HC1 recovery. 19 Figure 2.10 Sechol Flowsheet for Laterite Treatment at Atmospheric Pressure [11,12] Laterite Ore Tailing to Disposal -Magnetic concentrate for sale i Benefication Non-Magnetic concentrate I Atmospheric Chloride Leach To Disosal Water MgO S/L Separation Solution purification S/L Separation MgO I Two-Stage Ni/Co precipitation Mixed Ni^Co Hydroxide for sale Energy Energy. Magnesia for sale or Recycle I Water S/L Separation I U , + u X MgCL Pre-Evaporation I MgCL Pyrohydrolysis H C 1 20 2.1.4 Technology Summary: Shortcomings of the Existing Technologies The review of existing technologies for nickel and cobalt production may now be summarized. A number of technologies have been running through several decades at industry level, while some others are just fresh from laboratory and pilot study and show promise to replace existing technologies. Pyrometallurgy is well proven in the field of sulfide concentrate processing; however it has started facing restrictions in the modern world due to emission of CO2, C O , SO2 gases to atmosphere and the limitation of processing low-grade and dirty concentrates with high content of A s , Pb, Hg . Moreover, it requires high capital cost. Smelting to ferronickel proved its applicability for high nickel grade (saprolite type) laterites, whereas it is an impractical option to process limonite sources due to higher energy consumption. Hence, the hydrometallurgical process options are deemed suitable for limonite sources. On the other hand, hydrometallurgical processing options for base metal production proved its several advantages over pyrometallurgy. This route eliminates air pollution and yields higher product quality at lower capital and operating costs. Despite its advantages, not all existing technologies in nickel production are suitable for global resources. The high-pressure acid technologies, however, work reasonably well for most of the oxide ores and sulfide concentrates, they still face acid consumption constraints especially for high acid consuming resources. Therefore, P A L processes are mostly used to treat limonite (low acid consuming) ores rather than saprolite (high acid consuming). Clay content (high A l and M g O ) in ore is a limiting factor of the P A L usage. Moreover, the solid residue and solution from the pressure leaching circuit is neutralized with significant cost. On the other hand, ammonium technology can effectively treat higher acid consuming feed (high M g O laterite ores) and sulfide concentrates. The major disadvantages of the Caron process are the ore preheating and drying stages. This makes Caron based processes very sensitive to energy cost. In general, it is unlikely that new ammonium hydrometallurgical plants would be put into operation. Existing plants and operations w i l l continue, since their capital is already invested. Recently, more attention has been put on P A L ( H P A L ) processes commercializing three plants in Western Australia. Two plants are expected in New Caledonia and Australia for laterite resources, and one in Argentia, N L for sulfide concentrate. The main advantage o f those modern operations is successful utilization of advanced methods, such as solvent extraction, ion exchange, 21 pyrohydrolysis for solution treatment. Though all processes work well , the shortcomings of the existing hydrometallurgical processes rest on the autoclave leaching unit for acid leaching processes, and on the ore drying and reductions stages for Caron based processes. The feed characteristics are limiting criteria for H P A L operations. Moreover, the high-pressure vessels are the most expensive and non-standard equipment in any hydrometallurgical circuit. To overcome shortcomings of the existing technologies, the chloride-processing option has recently entered a period of renewed interest and investigation, because of its advantageous leaching power. The strong leaching power eliminates the use of pressure vessels and related costs. Second, processing in chloride media tolerates high clay content and high acid consuming feed. A t an industrial level, chloride processes proved their viability economically and technically. Three refineries in the world are already operating to process nickel sulfide mattes based on chloride processes. Outokumpu developed the Hydrocopper process to treat chalcopyrite concentrates in cupric-sodium-chloride media. In the case o f nickel, Jaguar Nickel developed a process to treat laterites in the mixture of magnesium chloride and hydrochloric acid at atmospheric pressure. This mixture draws specific attention, which w i l l be described shortly. There could be the possibility to process sulfide minerals and concentrates in chloride media at atmospheric pressure. Therefore, current research focused to treat sulfide concentrates of nickel/cobalt in the mixture of magnesium chloride and hydrochloric acid at atmospheric pressure. 2.1.5 Why MgCh? Uniqueness of this mixture A mixture of strong magnesium chloride and hydrochloric acid creates a non-ideal aqueous solution. First, proton activity in this mixture increases from its ideal, which gives a strong leaching power. Second, water activity is reduced. Jaguar Nicke l has reported that iron is easily hydrolyzed and precipitated, which may allow selective leaching of cobalt and nickel over iron. Thermodynamically, HC1 solutions with divalent chlorides such as M g C l 2 create much higher proton activity than univalent salts such as N a C l . Trivalent salts such as AICI3 create an even higher activity than univalent or divalent metal chlorides. Therefore, M g C l 2 - H C l mixture is ideal for high M g and clay (high A l ) content feed, which is an economical constrain for any other above-mentioned processes. Whilst A l cations contribute to increase proton activity at the same time decreasing water activity, M g cation may minimize dissolution potential of magnesia from laterite lattices and M g containing concentrates due to common ion effects. Pyrohydrolysis is used to recover and recycle hydrochloric acid in strong chloride 22 processes such as in Inco and Sechol. During this process, MgCh is one of a few chloride salts that decompose unlike other alkali and alkali earth metal chlorides. The decomposed M g is recyclable as M g O [11, 12]. Excess M g O may be sold as a by-product. A few promising properties of this mixture are mentioned here. In order to utilize specific properties of chloride media, more detailed knowledge of aqueous chloride solution is required. Therefore, the next part of this chapter reviews the thermodynamics of strong brine and hydrochloric acid mixtures. 23 2.2 Thermodynamics of aqueous chloride media 2.2.1 General Thermodynamically, the mixture of hydrochloric acid and chloride salts of common base metals create very useful properties of aqueous solutions of relevance to hydrometallurgists. These mixtures can increase the proton activity by several orders of magnitude due to metal ion hydration by water, which reduces the activity of free water in solution. The reduced water activity may allow selective leaching of metals [11, 12]. On the other hand, the chloride complexation with metal cations can enhance the solubility limit of certain cations. These changes of proton and cation activities in solution are the particular interest of the hydrometallurgists. Recently, the mixture of strong magnesium chloride and hydrochloric acid solution is proposed to be a good lixiviant to selectively leach nickel from saprolite/limonite mixture with good rejection o f iron along with high extractions of nickel and cobalt [11, 12]. In order to utilize the specific properties of this mixture for the purpose of leaching sulfide minerals, proper understanding of thermodynamic properties of aqueous chloride solution is required. Many efforts have been made to measure and estimate activities of ions in concentrated electrolytes. Sets of data, applicable for aqueous chloride solutions, are provided by Robinson and Stokes [42] and Harned and Owen [54]. Unfortunately, all known thermodynamic properties of acid chloride solution are limited to dilute solutions with ionic strength of less than six at ambient temperature. Hydrometallurgical processes are expected to operate at the upper limits of chloride concentration. Estimation methods followed by experimental verification (when applicable) w i l l be used to quantify the thermodynamic properties of strong chloride solutions, more specifically focused on hydrochloric acid and its mixtures with magnesium chloride at 25-100\u00C2\u00B0C. The estimation procedure requires thorough knowledge of thermodynamic properties; namely: mean ionic activity coefficients; corresponding water activities; osmotic coefficients; and the hydration numbers. In addition, the formation of chloride complexes is another important contributor in the behavior of leaching system. A l l mentioned thermodynamic properties w i l l be defined as completely as possible in the following sections. 24 2.2.2 Fundamental expressions of aqueous chloride media The following expressions are commonly used for solute concentrations in aqueous solutions: mole fraction-x, molality-m and molarity-c. The molal (mol/kg solvent) scale is preferred over the others for thermodynamic calculation purpose because of its independence of temperature and pressure. Based on mass balance, the molar and molal scales can be related as follows: m, =c, . /( /?-0.00l]Tc,.^.) t 2 - 1 ) Where, p-density of the solution, Wj -the formula weight of the i t h solute, c and m are molar and molal concentrations, respectively. The ionic strength-I, relates to molal concentration by the following equation. / = 0 . 5 5 > , m K ( Z 2 ) For A that dissociates into: A <=> [v+B+ + v_C~] it becomes I = 0.5mA[v+zl+v_zt] ( 2 2 a ) where, v\u00C2\u00B1 are moles of respective cations and anions, and z\u00C2\u00B1 are the charge numbers. The reduced activity coefficient T\u00C2\u00B0 defined in Meissner method in the next section is related to the ionic strength of the solution. Furthermore, the mean ionic activity coefficient- y\u00C2\u00B1 is related to the reduced activity coefficient (r\u00C2\u00B0) as shown here: Y = (r\u00C2\u00B0) ( z + z ' (2.3) and the mean ionic activity a\u00C2\u00B1 itself is: a\u00C2\u00B1 = m\u00C2\u00B1y\u00C2\u00B1 (2.4) where m\u00C2\u00B1 is the mean ionic molality defined as: m\u00C2\u00B1n=(mv*mv)Uv (2-5) where, v = v+ + v. Knowing all these parameters, the partial molal free energy or the chemical potential Lij can be defined as: M . = / / , \u00C2\u00B0 + W ? r i n ( a \u00C2\u00B1 ) (2.6) More definitions and applications of those equations w i l l be covered in the following sections and in the attachments. 25 2.2.3 Estimation of activity coefficients in aqueous chloride media 2.2.3.1 Estimation methods of activity coefficients in aqueous chloride media To precisely analyze thermodynamic properties of strong chloride solution, it is important to start estimating/measuring accurate values for the participating ion activities. The initial effort to estimate the activity coefficients for electrolyte solutions was put forward by Lewis and Randall [30] (cited from 40) by relating the activity coefficient and the concentration of an aqueous electrolyte solution. They related the ionic strength (I) based on the molal concentration as follows: / = 0.5^>,.z, 2 ( 2 7 ) i Continuing this work, Debye & Hiickel developed an expression for mean activity coefficient (y\u00C2\u00B1) and the ionic strength: l o g / + = Az^zJ0'5 (2.8) A detailed description of the constant A is reported with a value of 0.5107 k g 1 7 2 mol\" 1 / 2 at 25\u00C2\u00B0C [40]. Guntelburg [cited from 40, 54] has made improvements on this expression, taking ion size into account. Nevertheless, both expressions give fair results at ionic activity up to 1=0.1, all later improvements are based on this equation. Further improvements to calculate the mean ionic activity coefficients have been studied extensively by Guggenheim [56], Stokes & Robinson [42]; however, the validity of those equations weakens in solutions of molality greater than one mol kg\" 1. Meissner [31-38] developed a more useful and straightforward method using a chart based on previously reported thermodynamic data. The Meissner chart is a family of curves constructed by plotting T\u00C2\u00B0 (reduced activity coefficient) against I (ionic strength) for various electrolytes at 25\u00C2\u00B0C. For a given solution, reduced activity coefficient (r\u00C2\u00B0) at any ionic strength can be approximated from these graphs. Besides this graphical method, they also provided the following expression relating the reduced activity coefficients with ionic strength for pure electrolytes. r\u00C2\u00B0 = [1 + B(\ + 0.U) q -B]*exp[-AIm / ( l + CTJ (2.9) 26 where, A(25\u00C2\u00B0C) = 0.5107, B=0.75 - 0.065q, C = 1 + 0.055q * exp(-0.023 I 3) and q is called Meissner q-value which has different values for each electrolyte and at each temperature. Average values for the q are reported in Meissner [37,38] and Dixon [6] for selected electrolytes at 25\u00C2\u00B0C. Knowing q\u00C2\u00B0 at 25 \u00C2\u00B0C, its value at other temperatures is obtained by the following empirical equation. q\u00C2\u00B0T = q\u00C2\u00B025[l-0.0027(T-25)/zlz2)] ( 2 1 0 ) where, T i s the temperature, and zj, Z2 are the charge numbers of the dissociated salt. Meissner et al [31-38] also derived semi-empirical expression applicable to calculate mean activity coefficients for multi-component aqueous solutions. For a binary system such as a target of this research work-the M g C V H C l mixture, the equation can be reduced to: ^ i o r w a = i o g l o r 0 w c / + o . 5 x 3 i o g 1 0 ( r ^ c / 2 / r ^ ) ( 2 - l l a ) ^gI0rMgCT2 = i o g l 0 r\u00C2\u00B0MgCl2 - o.5x, i o g l 0 (T\u00C2\u00B0MgCl21 r\u00C2\u00B0HCl) (2.1 ib) where, r \u00C2\u00B0 refers to the reduced activity coefficient T. in a pure component i having the same total ionic strength as the mixture, x i and x 3 represent the ionic fractions of the respective cations. Recently Dixon [6] provided the generalized equation to calculate the reduced activity coefficient of components for mixtures of electrolytes as follows: log^r^: = ( z , X / n ^ \u00E2\u0080\u009E i o g 1 0 r \u00C2\u00B0 +zJYJimvmJiogl0r\u00C2\u00B0mJ)/i(zi+zj) (2.12) where, ^ = 0 . 5 ( z . + z . f /2.z. In these equations, indices i and m refer to cations whereas j and n refer to anions. It is important to note that for equations 2.11-2.12, the reduced activity coefficients of each species are calculated as in equation 2.9 at the total ionic strength of the mixture. Calculations obtained by using equations 2.11 or 2.12, give slightly different result as w i l l be shown in sample calculations later. Meissner et al also provided an empirical expression for temperature dependences o f mean activity coefficients between 0-150\u00C2\u00B0C at total ionic strength o f 1=10. log 1 0 r\u00C2\u00A3( /_io) = [1 - 0-005(7 - 25)] * log 1 0 r\u00C2\u00B0 5 ( / = 1 0 ) (2.13) 27 Combination of this formula and the Meissner plot allows calculating mean activity coefficients at any temperature-T and at an ionic strength of other than 10. These authors provided sample calculation in their works. The expression in 2.13 assumed that the isotherm for a given electrolyte at 25\u00C2\u00B0C would coincide with that at some other temperatures, which is not true for most electrolytes. The dashed line on the Meissner's chart [37,38], designated by the subscript \" re f , w i l l be relatively unaffected by temperature. Applying this, the temperature effect can be estimated at any ionic strength below 20 by the following equations. log 1 0 r \u00C2\u00B0 =(1.125- 0.0057/) * log 1 0 T\u00C2\u00B0 5 - (0.125 - 0.0057) * log Y\u00C2\u00B0ref ( 2 ' 1 3 ) w h e r e tog10r^=-O.41/,/,/a + / , / i ) + p . 0 3 9 / f t 9 1 (2.14) Those expressions allow calculating the reduced activity coefficients (and then mean ionic activity coefficients) at any temperature not limited to ionic strength of one value. Pitzer and co-workers [68] developed an expression (model) accounting for the effect of ion-pair interaction and triple ion interaction. The latest modified Pitzer equation is more successful for higher concentrations up to six molal. Even though the Pitzer model is considered complex comparing to other models, it is considered the most precise fit to experimental data. Roy et al [43-44], as wel l as Khoo's work [49-50] clearly demonstrates the use of this equation for a variety of mixed chloride systems. Bromley [39] developed the simplified version of the Pitzer model. According to his work, the correlation between mean ionic activity coefficients of strong electrolytes and ionic strength can be illustrated as shown below: iogio r\u00C2\u00B12,22 =0.511* 7 \u00C2\u00B0 5 / ( l + 7 \u00C2\u00B0 5 ) + (0.06 + 0.6B)I / ( l +1.5/ / z+z_ f+BIIz+ (2.15) The Bromley B values are a function of temperature and are reported for selected electrolytes; when data is not available, it can be calculated [39]. However, modifications have been made on both of these models (Edwards [41] et al. extended this correlation up to 10-20 molal and between 0-170\u00C2\u00B0C for aqueous ammonia solutions), the limitation remains up to 6 m for the chloride system. 28 Analyzing all these equations, methods and modifications to estimate the mean ionic activity coefficients, the Meissner method based calculations (the method described in Meissner's original papers, or the generalized equations appear in Dixon 's paper [6]) gave a reasonable estimation for activity coefficients at higher ionic strengths and at temperatures for strong chloride aqueous electrolytes. Hence, the Meissner method was selected for use in this thesis. 2.2.3.2 Estimating activity coefficients by Meissner method Taking the Meissner's theory as a basis, Peters [4] calculated the activity coefficients of hydrochloric acid mixed with N a C l , C a C ^ and M g C b , respectively, at various ionic strengths. This calculation confirmed that the strong chloride salts enhance the activity of hydrochloric acid. In the same manner, strong solutions of hydrochloric acid also enhance the activity of dissolved salts, and even stronger acid solution results in salting out of metal chloride salts. This unique thermodynamic property of mixed chloride electrolytes was successfully utilized in the Falconbridge matte leaching and crystallization processes. Likewise, the salting-out property of strong hydrochloric acid is used to precipitate other non-associated salts such as FeCl2 and AICI3 as applied in the Pechiney process to recover aluminum from clay. In the present work, calculations were carried out to estimate activity coefficients of compounds in the mixture of MgCL. and HC1. The purpose of this estimation was to illustrate estimation techniques to compare experimentally obtained values and to apply the calculated values to characterize the leach solutions where measurement is not applicable due to the instrument limitation. Two techniques, one provided by Dixon [6], and the second described in Meissner's papers [31-38] have been utilized. Both of these are the same, except for different approaches to calculating reduced activity coefficients of components o f mixed electrolytes. The calculated mean activity coefficients of HC1 in mixture of H C l - M g C l 2 having various ionic strength, are plotted in Figures 2.11 a, 2.11b and 2.11c at temperatures 25, 35 and 45\u00C2\u00B0C, respectively, against reference data [43] at these ionic strengths and temperatures. A t the same time, activities of MgCL. in solution have been calculated. There are only two reference sources to compare calculated values of MgCL. . One reference was calculated by Roy et al [43] using Pitzer's equation. The second reference for MgCl2 in H C l - M g C b mixture is provided by Khoo etc [49], again calculated using Pitzer's equation. Calculated values of yno and YMgci2 as well as the reference values are summarized in the Table A2.3 and A2.4 (Appendix 2) for HC1 and M g C h , respectively. 29 A s shown in these sample calculations, the activity o f the species in strong brine and hydrochloric mixture can be estimated at any ionic strength and at any temperature. These calculations allow assigning individual ion activities. In addition, these calculated values could be used to characterize the ionic activities of high ionic strength solutions, where commercial instruments are not applicable. 30 Figure 2.11 Calculated ynci of a mixture having I compared with experimental data from E M F measurement at temperatures [43] a. 25\u00C2\u00B0C b. 35\u00C2\u00B0C Figure 45\u00C2\u00B0C 2.2.3.3 Experimental methods to estimate activity coefficients: E M F measurement Activi ty coefficient measurements in binary electrolyte mixtures of H C l + M g C l 2 + H 2 0 have been made by Khoo et al [49] at 25\u00C2\u00B0C up to ionic strength of 3.0. Roy and his coworkers [43] extended these measurements up to total ionic strength of 5.0 at temperatures of 5-45\u00C2\u00B0C. These are the only available activity coefficient data for this system. Both measurements utilized the method o f measuring the Electro Motive Force ( E M F ) of the cell A without junction: Pt I HUgasMm) \\HCl - MgCl2 - H20\\ Ag I AgCl (A) The reaction occurring in the cell A is: ' ^ ^ 2 ( g a s ) + AgCl{S)^Ag\u00C2\u00B0is)+H^) + Cl-aq) \u00E2\u0080\u00A2 Therefore, the cell potential is defined by Nernst equation as follows: E = E\u00C2\u00B0-2303RT/nF *logw a a (2.16) (2.17) W h P T * P 2 W = rH+rcrmH,mcr = YHClmH+mcr For mixtures of multiple electrolytes (e.g. HCI-MCI-MCI2-MCI3): mH. mcr = mHCI (mHCI + mMCl + 2mMCh + 3mMCh ) . In this case, assumptions are made that these are all strong electrolytes and dissociate completely. Therefore, the following equation can be used for mixtures of multiple electrolytes. log,0 W = l Q S'0 tic! * mHC, (Z mMCl + 2 E mMCh + 3 E mMCh + etC~) (2.1 8) where, mMCl is molality of H C l , N a C l etc, mMCh is molality of M g C h , C a C k etc and mMCh is molality of AICI3 etc. For specific case, the activity coefficient of H C l in the mixture o f H C l - M g C l 2 is calculated based on the measured potential of the cel l -A: logic YHCI = \" I / 2 [ (\u00C2\u00A3 -E\u00C2\u00B0)/k + log 1 0 mHCl{mHCl + 2mMgCh )],k = 2303RTIF (2.19) where, yHCl is activity coefficient of the hydrochloric acid E - cell potential (corrected by hydrogen pressure and liquid junction potential) E\u00C2\u00B0-reference electrode standard potential (varies with temperature) 32 mHCl, mMgCl2 -molal concentrations of hydrochloric acid and magnesium chloride, respectively. The next approach of measuring cell potential was to use commercial reference electrodes instead of the hydrogen electrode and adjust the potential by applying junction potential. This measurement approach has been discussed in M u i r and Senanayake's work for determining junction potential [29]. In this work, combination of S C E and hydrogen electrodes were used for hydrogen ion activity measurement, whilst S C E (with junction) and A g C l pasted A g electrode (no junction) was used for CI\" ion activity measurements. They proved that the Henderson equation, which is applicable to define junction potential, is valid up to ionic strength of six. Beyond this ionic strength, this equation is not applicable [29]. The next approach to measure activity coefficient in chloride mixture had been carried out by J i [19] at the University of British Columbia. Commercial p H glass electrode was chosen for the purpose of measurement. J i made p H measurement based on the following approximate principles. pH = - l o g 1 0 aH+ = - l o g 1 0 y+m+ \u00C2\u00AB - l o g , 0 y+CH+ ( 2 2 0 ) Concentrated hydrochloric acid solution was added into a known concentration of NiCl2 aqueous solution whose p H changes are measured constantly. Then a plot o f aH+ vs m(Hci) added was constructed for each N1CI2 concentration. The results show almost straight lines, whose reverse slope indicates the activity coefficient of that solution. c\u00E2\u0080\u009E.=c, +c 0=\u00C2\u00AB\u00E2\u0080\u009E./,\u00E2\u0080\u009E, =10-\"- <2-2\u00C2\u00BB The liquid junction potential o f the glass electrode was determined and in this way the true p H values were obtained. Because of its simplicity, the E M F measurement with no junction w i l l be tested in this work. 33 2.2.4 Corresponding water activity or osmotic coefficient 2.2.4.1 Expressions for water activity and osmotic coefficient Knowing mean ionic activities for an aqueous solution, the chemical potential p.; can be calculated using equation 2.6. Then the corresponding water activity may be obtained by solving the Gibbs-Duhem relation: ] \u00C2\u00A3 > M = 0 (2-22) or it can be written as: 1000/18dlnf l w + X[v /// /rfln(W)] = 0 ( 2 2 3 ) Further, for an aqueous solution of single electrolyte, it is simplified as: 1000 /18d In a + mAvd Ma.) = 0 W A V \u00C2\u00B1> ( 2 2 4 ) For 1:1 electrolyte, the corresponding water activity (a\u00C2\u00B0w) may be directly read from Meissner chart, whereas, it can be calculated as follows for non 1:1 electrolyte. l o g 1 0 ( ^ ) = 0 . 0 1 5 6 / ( l - l / z + z _ ) + l o g 1 0 K ' ) ( 2 - 2 5 ) For a mixture of electrolytes, the corresponding water activity is calculated as follows [6]: log.o aw = YLWv l Q g i o + r (2-26), where r = 0 . 0 1 5 6 [ / X 2 X / , . / . /(Z/\u00E2\u0080\u009E)-\u00C2\u00A3/,. Iz) - I / , / z j ] (2-27) and 34 In addition to water activity described here, the related quantity, the osmotic coefficient \u00C2\u00A7 is defined as follows: ^ (2-29) ^ = -55.511naw/2 Jv,.7w (. Generally, this equation defines the osmotic coefficients for pure electrolytes at 25\u00C2\u00B0C. More meaningful <|> value for mixed electrolytes at elevated temperatures can be determined from the activity coefficient data (equation 2.22) as described in Jansz [7].

\u00C2\u00AB w \u00C2\u00AB \u00C2\u00A3 v,w,. In y. - X vi Jf m rM (2.30) The integral part o f this equation may be graphically evaluated more conveniently. 2.2.4.2 Estimating osmotic coefficients of H C l and M g C l 2 To illustrate validity of equation 2.30, and further to use it for mixed electrolytes, the osmotic coefficient calculation has been carried out for pure H C l and M g C l 2 electrolytes, separately. (Figure 2.12a and 2.12b, and Figure 2.13a and 2.13b) Figure 2.12 Calculated activity coefficient vs. reference data [42] at 25\u00C2\u00B0C for single electrolytes a) H C l 1.8 1.4 1 0 \u00C2\u00A3 0 . 6 \u00E2\u0080\u00A23 0.2 -0.2 -0.6 -1 2 4 m HCl b) M g C l 2 1 1 1 1 1 o ref 142] \u00E2\u0080\u00A2cal c 3 2.2 <-> 1.4 DC s \u00C2\u00ABa 0.6 -0.2 -1 1 1 1 1 A ref[42] Mi i A 2 4 m MgCI2 35 The calculated mean ionic activity coefficient data compared with reference data show an excellent agreement for HC1, while M g C b data is a little bit off, but it can still be used for further calculation. The curves on Figure 2.12 are graphically integrated. In consequence, osmotic coefficients for HC1 and M g C h are evaluated. This calculated data shows an excellent agreement with reference data compiled by Robinson & Stokes [42] as shown in Figure 2.13a and 2.13b. Those calculated and reference values are summarized in table A2.2 in the Appendix 2. Figure 2.13 Calculated osmotic coefficients vs. reference data [42] at 25\u00C2\u00B0C for single electolytes a) HC1 b) M g C l 2 The calculated mean ionic activity coefficients for pure electrolytes o f HC1 and MgCb. are compiled in Table 2.6a and 2.7a, respectively. 2.2.5 Individual ionic activities 2.2.5.1 Individual ionic activities in single/mixed electrolytes There is no direct method to calculate the individual ionic activities. However, they may be estimated by a number of different methods. The most reliable methods are acidity functions (Ho), E M F measurement, hydration treatment and the Ferrocence methods [29]. From these, the E M F measurement method stands out as the fastest and most convenient. E M F measurement in 36 appropriate cell while knowing the liquid-junction potential has been applied to determine the activity of hydrochloric acid. Determining the junction potential of cell is the difficult part of this method [28]. Stokes and Robinson [42] proposed an expression for mean ionic coefficients for unassociated strong electrolytes as follows: lny\u00C2\u00B1 = [z+z~ ] In fDH-hlv In a w - l n [ l + 0.01 S(v-h)m] (2.31) where,//)//- is the electrostatic contribution expressed as an activity coefficient. / = -Al0'51(1 + Ba\u00C2\u00B0I05) can be estimated by Debye-Huckel theory giving correct values for the activity coefficient of hydrated ions on the mole fraction scale, h-hydration number (the number of moles of water bound to one mole of solute). Values of hydration numbers are provided for some electrolytes and specific molality in Table 2.3. Table 2. 3 Hydration numbers of some simple salts and ions [28] Salt m h h. HC1 8 7 1\u00C2\u00B11 NaCl 0 3.4 5 1\u00C2\u00B11 1 4.6 2 3.8 5.1 3.2 KC1 1 4.1 CaCl 2 0 12 12 1\u00C2\u00B11 1 8.2 2 6.9 4.1 5 M g C l 2 0 13.7 14 1\u00C2\u00B11 1 7.8 2.1 7.4 4.2 5.9 This approach is based on the fixed hydration number independent of concentration. In concentrated solution, especially for highly hydrated salts, where a large number of water bound to ions in solution, the calculated activity coefficient may have misleading results. To overcome this problem, Jansz brought up an empirical equation for an \"average hydration number\" as a function of fixed hydration number and the activity. 37 log,,, h = log 1 0 h0 + (5 log 1 0 aw (2.32) This equation is valid for water activity of 0.4-0.9, where log-log plot of calculated hydration number against water activity shows linear relation. Jansz provided experimental values for ho and P for H C l and M g C l 2 (Table 2.4). Table 2. 4 Hydration number at infinite dilution (ho) and P values [7] ho P H C l 6.4 0.51 M g C l 2 12 0.36 U p to this point, we have defined mean ionic activity o f a compound, hydration number and osmotic coefficients, which are the main parameters to estimate single ion activities. Bates et al. stated expressions for separating mean salt activities into the contributions of individual ionic species by extending hydration theory. For univalent salts (MCI): log 1 0 V = l o S i o Y\u00C2\u00B1 + O.OO782A/H0 (2.33a) log 1 0 7CR = log 1 0 Y\u00C2\u00B1 - 0.00782/zm^ ( 2 - 3 3 b ) For unassociated sa l ts -MC^: l\u00C2\u00B0gio 7 > = 2 log I 0 y\u00C2\u00B1 + 0 . 0 0 7 8 2 / ^ + log 1 0 [ l + 0.018(3 - h)m] ^ l o g 1 0 y c r =log 1 0 ^ \u00C2\u00B1 +0.00782/z/T2^-log[l + 0.018(3 - 4 ) w ] (2.34b) Further Robinson and Bates expressions can be extended to calculate single-ion activity coefficients for two 1-1 electrolytes, A and B at constant ionic strength were used: logio 7CR = rA l og 1 0 fA + YB log.o YB - 0 . 0 0 7 8 2 / ^ ( 2 3 5 a ) log.o YH, = 2 log 1 0 YA ~ log 1 0 YCR (2.35b) For mixture of 1-1 and 1-2 electrolytes: (this is important in view of the possible use of H C l / M g C b solutions for the leaching of sulfides.) 38 (THO + 2YMtp,.) lQgio 7cr = THCI 1 o 8 . O fHa + rMgc,2 log I 0 r ^a , \" 0 . 0 0 7 8 2 ^ ^ * ^ / 2 + y ^ h ^ 13)(2mHCI + 3mMgC,2) YHCI log 1 0 [ l + 0-018(2- ^ a ) m \u00E2\u0080\u009E a ] - 2 ^ log 1 0 [ l + 0.018(3 - hMgCl2 )mMgCh ] + log 1 0 [ l + 0.018{(2 - hHCl)mHCl + (3 - hMgCl2 )mMgCh}] (2.36a) The H + activity coefficient is calculated as follows: log 1 0 y^=21ogy*- logy c r ( 2 3 6 b ) Note that all parameters in single ionic activity calculation equations y\u00C2\u00B1, h, (j) are not fixed numbers, they all are variables as a function of concentration. Keeping this in mind, it is now possible to estimate single ion activities in a pure electrolyte and as well as for the mixed electrolytes. 2.2.5.2 Estimation of individual ionic activities and p H of a mixture Activities of H + and CI\" ions are calculated using equation 2.33a and 2.33b for solutions of l -6m HC1 at various temperatures. These results are summarized in Table 2.8. Individual ionic activities of M g 2 + and CI\" are calculated using equation 2.34a and 2.34b for aqueous solutions of MgCl2 in relation to molality and temperature, as well . The results are shown in Table 2.9. A s mentioned before the interest of hydrometallurgist is generally not in pure electrolytes. The mixture of electrolytes is much more interesting. Therefore, individual ionic activities are calculated for mixtures of HC1 and M g C b at various ionic strength and temperature using equation 2.36a and 2.36b. The results are summarized in table 2.7 at 25\u00C2\u00B0C. More results of this calculation at higher temperatures are compiled in Tables A2.5 and A2.6 (Appendix 2) at temperatures of 60 and 100\u00C2\u00B0C, respectively. The highlighted rows in these tables are the mixtures that maybe used in the leaching experiments of this work. Note that the osmotic coefficient for mixture electrolytes are evaluated using equation 2.30 finding the area (the integral part) under the curve of respective species on lny+.(jn mixture) vs. molality diagram. Water activity of mixture is calculated as described in equation 2.26. Based on this water activity value, the corresponding hydration numbers were 39 calculated. Using this approach, the individual ion activities of the mixture can be evaluated. For the osmotic coefficient calculation using equation 2.24, the integral part was calculated as follows; it was assumed that logy of either H C l or M g C l 2 obey the Harned equation. The general form of the Harned equation [53] is shown below. l o g 1 0 7net = log r\u00C2\u00B0Ho - aifyMSa2 - Pfrlgck (2.37) where, yMgCl represents the ionic strength fraction of the magnesium chloride in the mixture. A comparison has been made in terms of hydrogen ion activity for the pure electrolyte versus mixture electrolytes as shown below. Table 2. 5 Calculated hydrogen ion activity based on individual ion calculations m H C l M M R C I 2 0 2 3 1 0.87 -- 7.87 4 13.15 6621 \u00E2\u0080\u00A2 -These results confirm that the hydrogen ion activity in l m H C l increases 9 times with the addition o f 3m M g C l 2 . This would yield p H of -0.9 (pH=0.06 at V =0.87) at only l m of acid concentration. This shows the potential to develop high leaching power in mixed electrolytes. A s seen here, the calculation of the individual ion activity coefficient determines p H of a mixture. Note that the individual ion activities are calculated based on the mean ion activity coefficients as described in section 2.2.3.1. For the mean activity coefficient calculations, two approaches were tested. One was described in Meissner's papers, and the second was the modified method by Dixon for mixed electrolytes. The results by both approaches differ slightly as shown in Figure 2.11. Therefore, p H values were calculated for one set of data (I t otai =5 at 25\u00C2\u00B0C in Figure 2.11) and results are summarized in Table 2.6. There is not a big difference between p H values based on both approaches; hence one of these approaches w i l l be utilized further to characterize solution properties. Table 2. 6 Comparison of pH values based on the two different approaches m H c i m MgC12 I total Calculated pH Dixon Meisner 5.00 0.00 5.00 -1.30 -1.31 3.58 0.47 5.00 -0.90 -1.02 2.40 0.87 5.00 -0.65 -0.71 1.77 1.08 5.00 -0.46 -0.49 0.86 1.38 5.00 -0.04 -0.05 40 The p H values for the mixtures highlighted in Table 2.10 and ,in Tables A2.5-A2.6 (Appendix 2) were calculated and summarized in Table 2.7. The similar mixtures w i l l be used for the experiments o f this work. There are no commercial electrodes to measure the p H of these strong acid solutions; therefore, calculated p H values are applicable to characterize solution properties when measurement is limited. Table 2. 7 Estimated pH of solution mixtures at temperatures m H c i m MgCI2 pH at temperatures 25\u00C2\u00B0C 60\u00C2\u00B0C 100\u00C2\u00B0C 2.00 2.00 -1.05 -0.78 -0.53 4.00 2.00 -1.92 -1.57 -1.23 6.00 2.00 -2.62 -2.18 -1.76 7.00 2.00 -2.94 -2.46 -2.00 8.00 2.00 -3.23 -2.71 -2.21 10.00 . 2.00 -3.77 -3.17 -2.61 6.00 2.30 -2.74 -2.29 -1.86 6.00 2.40 -2.78 -2.32 -1.89 8.00 2.40 -3.36 -2.83 -2.33 41 Table 2. 8 Calculated individual ionic activity coefficients for H C l in relation to molal concentration and temperature m H C l 3w u 25\u00C2\u00B0C 40\u00C2\u00B0C 50\u00C2\u00B0C 60\u00C2\u00B0C 70\u00C2\u00B0C 80\u00C2\u00B0C 90\u00C2\u00B0C 100\u00C2\u00B0C n HCl Yci- Y H + Yci- Y H + Yci- Y H + Yci- Y H + Yci- Y H + Yci- Y H + Yci- Y H + Yci- Y H + 1.0 0.970 6.301 1.024 0.693 0.874 0.684 0.863 0.678 0.855 0.672 0.848 0.666 0.840 0.660 0.833 0.654 0.825 0.649 0.818 2.0 0.910 6.099 1.185 0.753 1.267 0.726 1.221 0.708 1.192 0.691 1.163 0.675 1.135 0.659 1.109 0.644 1.084 0.630 1.060 3.0 0.850 5.891 1.360 0.839 1.993 0.785 1.865 0.752 1.786 0.720 1.712 0.691 1.642 0.664 1.577 0.638 1.517 0.615 1.461 4.0 0.800 5.712 1.532 0.932 3.288 0.845 2.979 0.792 2.794 0.745 2.626 0.701 2.472 0.661 2.332 0.625 2.205 0.593 2.091 5.0 0.720 5.413 1.717 1.056 5.630 0.924 4.922 0.847 4.514 0.779 4.152 0.719 3.830 0.665 3.544 0.617 3.291 0.575 3.066 6.0 0.600 4.932 1.920 1.249 9.667 1.051 8.136 0.941 7.285 0.846 6.549 0.764 5.912 0.693 5.360 0.631 4.881 0.577 4.465 Table 2. 9 Calculated individual ionic activity coefficients for M g C l 2 in relation to molal concentration and temperature m M g C 1 2 3w h M g C 1 2 <() 25\u00C2\u00B0C 40\u00C2\u00B0C 50\u00C2\u00B0C 60\u00C2\u00B0C 70\u00C2\u00B0C 80\u00C2\u00B0C 90\u00C2\u00B0C 100\u00C2\u00B0C Y Cl- Y Mg2+ Y Cl- Y Mg2+ Y Cl- Y Mg2+ Y Cl- Y Mg2+ Y Cl- Y Mg2+ Y Cl- Y Mg2+ Y Cl- Y Mg2+ Y Cl- Y Mg2+ 1.0 0.929 11.685 1.111 2.0 0.835 11.247 1.517 3.0 0.705 10.582 2.023 4.0 0.620 10.104 2.527 5.0 0.524 9.507 3.043 0.518 0.541 0.707 3.157 0.885 25.140 0.912 207.07 0.813 1637.4 0.479 0.462 0.619 2.418 0.735 17.345 0.724 130.76 0.623 960.34 0.455 0.418 0.569 2.042 0.654 13.714 0.626 97.766 0.526 685.33 0.434 0.380 0.525 1.739 0.584 10.964 0.545 74.093 0.448 496.81 0.415 0.347 0.486 1.495 0.526 8.871 0.478 56.965 0.385 366.17 0.398 0.320 0.453 1.297 0.476 7.269 0.422 44.469 0.333 274.63 0.384 0.297 0.424 1.138 0.434 6.037 0.376 35.278 0.291 209.78 0.371 0.277 0.400 1.009 0.398 5.087 0.338 28.464 0.257 163.35 Table 2.10 Calculated individual ion activity coefficients for a mixture of HCl-MgCl2 in relation to molal concentrations of each species at 25\u00C2\u00B0C m H C I m MgC12 I total Dixon Meisners, Jansz et al. n MgC12 \" H C I ' yCl - yH+ yCl - yH+ TEMPERATURE: 25\u00C2\u00B0C 8.00 0.00 8.0 0.500 9.350 4.494 1.000 3.775 28.673 1.022 3.628 29.834 5.50 0.83 8.0 0.518 9.473 4.578 0.923 1.947 11.228 0.929 2.048 16.287 3.94 1.35 8.0 0.561 9.748 4.767 0.837 1.615 8.802 0.833 1.707 10.428 2.00 2.00 8.0 0.641 10.227 5.103 0.666 1.516 5.488 0.640 1.581 5.652 1.04 2.32 8.0 0.691 10.504 5.300 0.541 1.567 4.072 0.497 1.617 4.054 10.00 0.00 10.0 0.380 8.470 3.907 1.000 6.888 96.394 1.019 6.641 99.981 5.01 1.66 10.0 0.457 9.049 4.291 0.781 2.335 22.914 0.790 2.477 28.282 4.00 2.00 10.0 0.491 9.287 4.451 0.702 2.166 18.544 0.707 2.286 20.987 1.28 2.91 10.0 0.611 10.052 4.980 0.385 2.170 8.553 0.358 2.236 8.530 12.00 0.00 12.0 0.300 7.779 3.463 1.000 12.074 296.804 1.033 11.537 310.597 10.79 0.40 12.0 0.303 7.804 3.479 0.981 8.028 88.503 0.996 8.182 237.508 7.68 1.44 12.0 0.341 8.147 3.697 0.884. 4.001 73.282 0.864 4.361 109.998 6.00 2.00 12.0 0.379 8.465 3.903 0.796 3.205 56.711 0.757 3.512 69.950 3.95 2.68 12.0 0.443 8.950 4.225 0.637 2.783 36.411 0.573 3.025 38.401 1.37 3.54 12.0 0.550 9.678 4.719 0.323 2.897 16.800 0.215 3.088 16.134 13.00 0.00 13.0 0.280 7.589 3.344 1.000 15.533 511.733 1.043 14.716 540.157 11.20 0.60 13.0 0.286 7.647 3.380 0.971 8.754 141.214 0.984 9.077 360.537 9.75 1.08 13.0 0.301 7.785 3.467 0.933 6.178 132.639 0.925 6.659 253.694 7.00 2.00 13.0 0.351 8.228 3.750 0.813 3.927 95.502 0.770 4.346 124.419 3.02 3.33 13.0 0.477 9.191 4.386 0.478 3.148 38.491 0.381 3.434 38.045 1.06 3.98 13.0 0.564 9.767 4.781 0.198 3.504 19.824 0.065 3.763 18.716 14.00 0.00 14.0 0.240 7.179 3.091 1.000 20.456 837.344 1.041 19.486 879.029 12.20 0.60 14.0 0.246 7.240 3.128 0.972 11.587 213.733 0.985 12.029 588.716 10.49 1.17 14.0 0.261 7.400 3.227 0.928 7.679 199.253 0.919 8.325 392.419 8.00 2.00 14.0 0.300 7.782 3.465 0.827 4.993 151.778 0.789 5.546 210.331 3.39 3.54 14.0 0.429 8.850 4.157 0.465 3.628 57.019 0.368 3.981 56.480 2.23 3.92 14.0 0.475 9.178 4.378 0.319 3.739 39.847 0.202 4.070 38.090 16.00 0.00 16.0 0.200 6.723 2.816 1.000 33.211 2208.874 1.039 31.747 2310.718 14.63 0.46 16.0 0.202 6.752 2.834 0.981 21.475 487.828 1.000 21.947 1715.600 11.25 1.58 16.0 0.228 7.051 3.013 0.897 9.704 423.856 0.876 10.745 787.759 10.00 2.00 16.0 0.245 7.231 3.123 0.849 7.815 371.895 0.815 8.751 582.743 4.60 3.80 16.0 0.366 8.354 3.831 0.484 4.760 136.748 0.388 . 5.310 138.113 3.20 4.27 16.0 0.412 8.722 4.073 0.330 4.781 92.390 0.213 5.296 88.622 12.90 0.00 12.9 0.280 7.589 3.344 1.000 15.190 483.867 1.031 14.530 505.829 8.33 1.52 12.9 0.322 7.979 3.590 0.850 4.808 108.893 0.856 5.181 170.961 6.00 2.30 12.9 0.375 8.429 3.880 0.710 3.594 75.254 0.704 3.874 91.712 2.95 3.32 12.9 0.478 9.201 4.393 0.421 3.190 35.672 0.388 3.374 36.284 2.25 3.55 12.9 0.508 9.403 4.530 0.330 3.253 28.642 0.287 3.424 28.495 13.20 0.00 13.2 0.280 7.589 3.344 1.000 16.238 572.063 1.021 15.631 594.282 11.98 0.41 13.2 0.281 7.603 3.353 0.979 10.833 154.771 0.986 11.094 454.987 6.00 2.40 13.2 0.373 8.416 3.872 0.724 3.653 84.259 0.677 4.026 99.552 3.73 3.16 13.2 0.444 8.959 4.230 0.524 3.273 49.497 0.454 3.560 50.686 1.89 3.77 13.2 0.518 9.467 4.574 0.292 3.420 28.082 0.196 3.665 27.088 15.20 0.00 15.2 0.200 6.723 2.816 1.000 28.196 1472.131 1.028 27.251 1523.199 8.00 2.40 15.2 0.292 7.701 3.414 0.711 5.753 213.189 0.719 6.232 288.545 5.83 3.12 15.2 0.351 8.228 3.750 0.546 4.725 141.570 0.547 5.063 161.939 3.30 3.97 15.2 0.442 8.946 4.222 0.279 4.485 73.548 0.263 4.720 74.911 2.18 4.34 15.2 0.492 9.293 4.455 0.127 4.728 51.060 0.098 4.937 50.557 43 2.2.6 Thermodynamic predictions a) Dissolution of sulfide minerals The individual ion activity calculations allowed for estimates of the p H values of any mixtures as described in the previous section. To select a proper mixture of H C l and MgCh to dissolve certain sulfide minerals, it is useful to predict p H values for each sulfide minerals where dissolution occurs. Based on the Gibbs free energy of corresponding dissolution reactions, these p H values can be estimated. For example, the dissolution of millerite is predicted as follows: N i S + 2 H + ( a ) O N i 2 + ( a ) + H 2 S( g ) at equilibrium, AG\u00C2\u00B0(25\u00C2\u00B0C) =-2.14 kJ/mole (RXN2.1) Assuming unit activities for solid compounds and 1 arm pressure for gas phases, and assigning a unit activity for the dissolved metal cations (aMe\u00E2\u0080\u009Et =l) the solution p H relates to the Gibbs free energy as follows: H ( a = i ) = A G \u00C2\u00B0 ( J ) where m is the number of moles of the hydrogen ion involved in the sulfide mineral dissolution. In this specific example m=2. Applying the same principle, p H values for the selected minerals were estimated at different temperatures. The results are summarized in Table 2.11 and plotted on Figure 2.14. Note that the predicted p H values are calculated at a unit activity of the dissolved metal cations. It was assumed that at the unit activity of dissolved metal cations, an acceptable dissolution w i l l occur. Table 2.11 The AG\u00C2\u00B0 and estimated pH values for the dissolution ol ' selected sulfide minerals Mineral Equilibrium Reactions AG0 , kJ, at \u00C2\u00B0C pH at equilibrium when a(Me n +)=l 25 60 100 25 60 100 Ni 3 S 2 N i 3 S 2 + 6HCl ( a ) = 3NiCl 2 ( a ) + 2H 2 S ( g ) + H 2 ( g ) -18.32 -25.70 -28.29 0.53 0.67 0.66 NiS NiS + 2 H C l ( a ) = N i C l 2 ( a ) + H 2 S ( g ) -2.14 -5.17 -6.66 0.19 0.41 0.47 FeS FeS + 2HCl ( a ) = FeCl 2 ( a ) + 15.66 5.48 -5.50 -1.37 -0.43 0.38 PbS PbS + 2HCl ( a ) = PbCl 2 ( a ) + H 2 S ( g ) 20.97 11.48 1.65 -1.84 -0.90 -0.12 CuFeS 2 CuFeS2+ 2HCl ( a ) = CuS +FeCl 2 ( a )+H 2S ( g ) 47.58 37.30 26.17 -4.17 -2.92 -1.83 FeS 2 FeS2 + 2HC1 W = FeCl 2 ( a ) + H 2 S f e ) + S 72.52 60.94 48.26 -6.35 -4.78 -3.38 A G 0 data from HSC Version 5.1 44 Figure 2.14 Predicted p H for the dissolution of selected minerals Based on these thermodynamic calculations, the dissolutions of the selected minerals are summarized as follows: Heazelwoodite and millerite would be expected to dissolve in acid solutions with low acid concentration in the p H range of 0.0-0.5. Troilite and galena would be directly attacked in strong acid. In both cases, increasing temperature up to 100\u00C2\u00B0C has a beneficial effect to increase the dissolution p H of the minerals. Unlike these minerals, chalcopyrite dissolution requires a significant amount of acid addition to be dissolved. However, higher temperature increases the required p H up to around -2.0. According to Table 2.7, the mixture of 7m HC1 and 2m M g C l 2 w i l l have -2.0 of p H value. Among the sulfide minerals considered here, pyrite would be expected to have the most refractory behavior in the direct attack of acid. The mixture of 10m HC1 and 2m MgCl2 has a p H of about -2.76 (Table 2.7); therefore creating a solution with p H of about -3.5 (at 100\u00C2\u00B0C) to dissolve pyrite may not be practical. The previous section estimated the p H values of any mixtures. This section predicted the p H values where certain minerals could be dissolved. Combinations of these predictions allow selecting a suitable mixture of HC1 and MgCl2 to process certain minerals and commercial concentrates. 45 b) Chemistry of dissolution products Let us revisit the prediction of sulfide minerals dissolution (Table 2.11). In strong acid solution, the sulfide minerals o f common base metals dissolve forming H2S gas in addition to their corresponding metal chloride salts. The H2S gas is a strong reducing agent (as well as H2 gas in case of heazelwoodite dissolution). A n interesting question is how w i l l H2S gas react with the metal cations in solution. It is especially a concern, as illustrated below, i f the solution contains ferric and cupric ions. Cupric and ferric reduction by H2S gas is favorable in terms of Gibbs free energy. CuCl 2 ( a) + H 2 S ( g ) = CuS + 2 H C l ( a ) AG\u00C2\u00B0(25\u00C2\u00B0C) = -84.68 kJ/mol ( R X N 2.2) 2 F e C l 3 + 3 H 2 S ( g ) = 2FeS + S\u00C2\u00B0 + 6 H C l ( a ) AG\u00C2\u00B0(25\u00C2\u00B0C) = -202.9 kJ/mol ( R X N 2.3) The solid products formed by these reactions (especially elemental sulfur and copper sulfide) may have a deleterious effect on the leaching of minerals due to the formation o f passive layers. Furthermore, i f the feed contains copper, its extraction into solution w i l l always be lower than the other metals since the dissolved copper more likely precipitates according to the above reaction ( R X N 2.2). If feed contains both heazelwoodite and copper minerals, copper extraction w i l l be much lower than nickel since both H 2 S and H2 gases reduce dissolved copper. The reduction of cupric ion by H2 gas is also very favorable compared to other cations as illustrated by the following reactions. CuCl 2 ( a) + H 2( g) =Cu+ 2 H C l ( a ) AG\u00C2\u00B0(25\u00C2\u00B0C) = -61.4 kJ/mol ( R X N 2.4) FeCl 2 ( a) + H 2 ( g) = Fe+ 2 H C l ( a ) AG\u00C2\u00B0(25\u00C2\u00B0C) = +52.98 kJ/mol ( R X N 2.5) N i C l 2 ( a ) + H 2 ( g ) = Ni+ 2 H C l ( a ) AG\u00C2\u00B0(25\u00C2\u00B0C) = +54 kJ/mol ( R X N 2.6) Although these predictions were made at 25 \u00C2\u00B0C, we expect the same results at higher temperatures; therefore, for the processing of mixed metal sulfides, the addition or presence of cupric or ferric ions may have a negative impact on metal extraction. 46 2.2.7 Chloride complexes The chloride system presents many important features of relevance to the hydrometallurgist in addition to increased ionic activities. One of these is the formation of chloride complexes with the common metal cations. This feature allows formation of ionic species that are not present in other media such as in sulfate. A good example of this is the stabilization of cuprous chloride complexes along with cupric chloride. The combination of these ions can create a strong leaching environment to effectively leach sulfide minerals, while allowing copper metal to be recovered from its lower valence state. There are several observed phenomena for the chloride complex system. First, this type of complexation is weaker than the aquo complexes. Second, the metal cations that are complexed poorly by water, form strong chloride complexes. Table 2.12 summarizes a general categorization of metals according to their tendency of chloride complex formation. Table 2.12 Strengths of chloro-complexes according to periodicity [28] W E A K Lanthanides LogK3 Coinage metals ( C u + < A g + < A u 3 + ) Volatile metals ( Z n 2 + < C d 2 + \u00C2\u00AB H g 2 + ) Precious metals ( O s 3 + \u00C2\u00AB I r 3 + \u00C2\u00AB P d 2 + , P t 2 + , P t 4 + ) Group V ( A s 3 + \u00C2\u00AB Sb 3 + < B i 3 + ) The transition metals in the second row form stronger chloride complexes than the first row metals, especially i f in their lower valence state. In addition to general rules for metals, chloride complex formation is also strongly related to chloride ion concentration. Table 2.13 summarizes the chloride complexes of common metal cations that may be present in the solution of sulfide concentrate leaching. 47 Table 2.13 Resume of the common chloro-complexes [6] Low C l - concentration High C l - concentration Cu(II) C u 2 + C u C f C u C l 2 CuCl 3 \" CuCl 4 2 \" Cu(I) CuCl 2 ' CuCl 3 2 \" C u C l 4 3 \" Fe(III) F e 3 + - F e C l 2 + F e C l 2 + Fe(II) Fe 2 + F e C f Zn Zn Z n C f ZnCl 2 ZnCl 3 \" ZnCl 4 2 \" Pb P b C f PbCl 2 PbCl 3 \" PbCl 4 2 \" N i N i 2 + N i C l + Co C o 2 + C o C f M n M n 2 + M n C f Cd C d 3 + C d C l + C d C l 2 C d C V CdCl 4 2 \" Sb SbCl 2 + SbCl 2 + SbCl 3 SbCl 4\" SbCl 5 2 - SbCl 6 3 \" B i B i C l 2 + B i C l 2 + B i C l 3 B i C l 4 \" B i C l 5 2 \" B i C l 6 3 \" As A s C l 3 Ag A g c r A g C l 3 2 \" Hg H g C l+ HgCl 2 HgCl 3 - HgCl 4 2 \" In general, the complexation is represented by the following reaction. Me\"+ + iCl <^> MeCir (2.38) The maximum ligand number-i is often four in chloride system. The corresponding stepwise stability constants are defined as: K 0 = aMeCir (2.39) The cumulative stability constant is: (2.40) The complexation changes free energy of a metal ion as written below, which in turn changes its reduction potential. AG 0 c o m p,ex = AG\u00C2\u00B0m etai -2 .303RT Log Pi (2.41) Therefore, the effect of complexation on the reduction potential can be observed by the Nersnt equation (e.g. for reaction 2.38). 48 F = F * Me\"* I Me\" 2303RT \u00E2\u0080\u009E , A fli *log 0_i cr (2.42) MeCir The increase in background [Cl\"] generally decreases E h for metal ion/metal complexes depending on the Pi and complexing number-n. Considering the effects of complexation on the reduction potential, M u i r and co-workers constructed the Eh-p[Cl\"] diagram for various base metals (Figure 2.15). Figure 2.15 Eh-log[Cl] diagram at 25\u00C2\u00B0C [28] (Calculated using p\u00E2\u0080\u009E assuming y = 1; Pb(II) and Bi(III) = 10\"4 M , Ag(I) and Au(III) = 10\"5 M , others 10\"1 M , pH = 1) This diagram provides useful thermodynamic information of chloride complex species and their region of stability. It is known that the metals with lower E h in this diagram can displace (cement out) metals with higher E h from their solutions. It is detrimental i f impurities are cemented on to the final metal product, but advantageous in the solution purification and recovery of some metals from their chloride solutions. log[Cn 49 2.2.8 Solubility of chloride salts One of the multiple advantages of chloride leaching media is an increased solubility of metals compared to sulfate media. Solubility data of metal chlorides that are commonly found in the leaching of sulfide minerals are compiled extensively in Linke [47]. Even solubility of sparingly soluble salts is increased by several orders in chloride media [3, 7]. A s described in the literature, temperature has a significant effect on the solubility of chloride salts. Decreasing temperature may allow precipitation o f chloride salts such as A g C l and P b C ^ . Calculations of individual ion activities in section 2.2.5 confirmed that the mixtures of chloride salts with hydrochloric acid create high individual ion activities including hydrogen and chloride ion. The addition of salts that form a high background concentration of chloride has a significant effect on the solubility. Whilst hydrogen ion breaks the sulfide mineral lattices down forming metal cations, the chloride ions form complexes with these cations increasing their solubility. In addition, it is observed that the solubility limits vary, not only with the total chloride concentrations, but also with the type of chloride salts added to the solution (or contents of the commercial product that is processed in strong brine). Table 2.14 A G 0 reaction of selected chloride complexes at 25\u00C2\u00B0C, kJ/mole Complexing reactions n (complexing numbers) 1 2 3 4 C u + + nCl\" O CuCl,, 1\"\" -38.74 -30.24 -32.53 C u 2 + + nCl\" <\u00C2\u00BB CuCl n 2 \" n -2.71 23.24 13.08 26.21 N i 2 + + nCl\" O N i C l n 2 \" n 5.75 49.15 C o 2 + + nCl\" \u00C2\u00BB CoCl n 2 \" n -119.57 M g 2 + + nCl - O MgCl\u00E2\u0080\u009E 2- n 0.99 121.77 C a 2 + + nCl\" O CaCl n 2 \" n -0.82 Fe 2 + + nCl\" O FeCl n 2 \" n 0.93 91.53 Fe 3 + + nCl\" <\u00C2\u00BB FeCl n 3 \" n -8.44 -81.24 78.80 Table 2.14 shows the Gibbs free energies for the selected metal cation-chloride complexes that may form in chloride media during leaching of sulfide minerals. What is expected from this information? Cuprous, ferric and cobaltous cations have great affinity, whereas cupric and calcium ions have less tendencies to form complexes with chloride ions. Divalent cations such as nickel and magnesium have no tendency to form chloride complexes. These complexation behaviors have positive or negative effects on metal dissolutions. The weak complexation of magnesium leaves chloride ions free in solution resulting in an increased concentration of this ion (in case of using the 50 mixture of M g C l 2 and H O ) . The free chlorides form complexes with the dissolved metal cations enhancing their solubility. In addition to dissolved metal cations (in this specific case N i 2 + , C o 2 + , C u 2 + ) , i f a solution contains metal cations that form strong chloride complexes such as C u + and Z n 2 + , the complexation w i l l have deleterious effect on the solubility of the weak complexing metals. This phenomenon may easily be illustrated by the following example. Say that a leach solution contains ferric ion in addition to dissolved copper the following reaction w i l l most likely take place. C u C f + F e 3 + + 3C1\" => CuCl 2( S) + FeCl 2 + ( a q) ( R X N 2.7) The salts that form strong chloro-complexes depress the solubility of salts that form weak chloro-complexes. However, CuCl 2 ( S ) w i l l not precipitate in dilute solutions as shown here; this may retard dissolution kinetics i f the solution is concentrated. 2.2.9 Effect of A1C13, NaCl and C a C l 2 on activities of MgCl 2 +HCl mixture A leach system contains a variety of minerals, metal cations and chemical species; therefore, it is useful to estimate the effects of each component on the activities of each species. The effects of A l 3 + and N a + cations on proton activities were estimated (an interest of project supporters). In addition, the effects of M g 2 + and C a 2 + were considered since these metals are encountered in any type of ore and are dissolved in a chloride leach system. The effect M g 2 + in solution draws a particular interest in order to process high magnesium containing feeds in this mixture. Again, taking the Meissner method as a guide for mixed electrolytes, the activity coefficients of each salt were estimated assuming they were in the chloride salt form and dissociated completely. The results are shown in Figures 2.16a to 2.16d for the case of increasing concentrations of single salts (the remaining salt concentrations were kept constant). The effects o f double salts are plotted in Figures A2.1a to A2.1d (Appendix 2). Complete results are summarized in Table A2.7 (Appendix 2). Selected results at higher concentrations are shown in Table 2.15. A s shown in Table 2.15, AICI3 has the most impact on increased activity coefficient of HC1. In contrast, N a C l has no advantageous effect on increased HC1 activity coefficient. C a C l 2 has a noticeable effect on it, more importantly M g C l 2 has a higher effect than C a C l 2 . This shows that the high clay (high A l ) and M g O content feed mixed with an acid has the potential for creating high acid activities. The activity coefficients of HC1 in the presence or absence of added salts (58 vs. 7, respectively) are presented in the fifth and last row of this table. This tells the advantages of using mixtures instead of pure acid solution. Rows 6 to 8 of this table summarized the effects increasing 51 concentrations of double salts. Increasing molalities of AICI3 and M g C h up to 4m results in mean ionic activity of 88 for HC1 while the rest of salt ( M g C ^ , N a C l & HC1) molalities are kept at constant of l m . Double salts of M g C ^ and.CaCh create HC1 activity coefficient of 61 (vs. 88 in previous case), which is excellent combination for the purpose of increased HC1 activity coefficient. Figure 2.16a displays the effect of increasing AICI3 concentrations on the activity of HC1 while other salt concentrations remained constant. Each curve on this figure represents the activity coefficient changes of corresponding salts including HC1. In addition, a total ionic strength is included in this plot. Figures 2.16b-2.16d displays the same information, the only difference is the concentration o f salt, o f which the effect that was considered, is increasing. Figures A2.1a-A2.1d (Appendix 2) contain the same information, in these cases the effects o f double salts were estimated. Overall, more salt addition with increasing concentration results in increased HC1 activity coefficients in the mixture. Therefore, in terms of increased HC1 activity coefficient, the mixture of HC1 and M g C h is suitable to process high clay (high A l ) and M g O containing materials. Table 2.15 Effect of added salts on calculated HC1 activities at 25\u00C2\u00B0C Molal concentrations of each salt I total Activity of HC1 MgCl 2 NaCl A1C13 CaCl 2 HC1 1 1 1 4 1 32 34.81 1 1 1 1 4 23 20.97 1 1 4 1 1 17 10.53 1 4 1 1 1 23 24.48 1 1 1 1 17 58.47 1 4 1 4 1 41 88.46 1 4 1 1 4 32 61.12 1 4 0 4 0 37 62.14 4 0 0 0 0 4 7.00 52 Figure 2.16 Effects of A1C13, NaCl and C a C l 2 on activities of MgCl 2 +HCl mixture a) The effect of AICI3 b) The effect of G a C l 2 c) The effect of N a C l 20 0 1 2 3 4 Molality of NaCl - \u00E2\u0080\u00A2 - H C l -D-MgC12 \u00E2\u0080\u0094O-NaCI -0-A1C13 - \u00C2\u00A3 - C a C 1 2 1 total d) The effect of M g C l 2 0 1 2 3 Molality of M g C l 2 - \u00E2\u0080\u00A2 - H C l - D - M g C 1 2 - \u00E2\u0080\u00A2 - N a C l -0 -A1CI3 - 6 - C a C 1 2 1 total 53 2.2.10 Summary of thermodynamic review Reviewing the thermodynamics of strong electrolyte mixtures, it is clear that the activity coefficients are important parameters that need to be estimated. Among several techniques developed to estimate activity coefficient, the Meissner method is considered the simplest and most convenient to use. In the present work, activity coefficients of HC1 and M g C l 2 in the mixture of HCl - M g C l 2 -H20 have been calculated utilizing Meissner's method at various total ionic strengths and temperatures at various ionic strength fractions. Available reference data for this mixture is limited to total ionic strength of five and temperatures of up to 45\u00C2\u00B0C. This data was obtained from the E M F (Electro Motive Force) measurement method. The E M F measurement method is considered the simplest experimental method and was selected to be tested in this work. This method allows determining activity coefficients of HC1 in a mixture. Based on this, the activity coefficients of MgCl2 are calculated using very complex equation of Pitzer with multiple parameters. These parameters are not easily obtained, and are dependant on both ionic strength and temperature; therefore, current work w i l l only focus on HC1 activity coefficients. There is no experimental technique to determine individual ion activity coefficients in a mixed electrolyte. However, the individual ion activity coefficients can be estimated using expressions provided by Stokes, Robinson and Bates. More accurate estimation can be done applying variable hydration numbers, water activities and osmotic coefficients as suggested by Jansz. The individual ion activity allows the assigning of the p H of a solution. This calculation followed the mean activity calculations by the Meissner method. Two different approaches for mean activity coefficient estimation were used; however, both gave almost the same p H values for a mixture. Therefore, only one method was selected for further calculations of p H . There were no distinct criteria to select one approach over another. The thermodynamic predictions for dissolutions of sulfide minerals showed that the minerals such as heazelwoodite and millerite can be dissolved in acid solutions with low acid concentrations with pH~+0.5. The decomposition of troilite and galena by the direct attack of a strong acid at p H \u00E2\u0080\u0094 1.0, are favorable. In both cases, the increasing temperature of up to 100\u00C2\u00B0C has a beneficial effect on the increase of the dissolution p H of the minerals. Unlike these minerals, chalcopyrite dissolution requires a significant amount of acid addition (pH<-2.0) to be dissolved; however, higher temperature increases the required p H above -2. Pyrite shows the most refractory behavior in the 54 direct attack of acid. A n increasing temperature has a weak effect to increase the dissolution p H of this mineral. The chemistry of dissolution products (H2S and H 2 gases), with the metal cations in solution, was explored here. If solution contains an excess amount of cupric or ferric ions, these gases would react with these metal cations, and may cause formation of a protective layer of CuS or S, which may prevent further mineral dissolution. In addition, solubility and complexation behavior of chloride mixtures are reviewed here. Both o f them have positive or negative effects on the dissolution of sulfides depending on the compositions of leach media. A s metals form strong complexes with chlorides, their solubility increases. The complexation mostly decreases the reduction potential o f common base metals. The metals with lower potential on the Eh-Log(Cf) diagram tend to cement out the metals with higher potential. Some cations (Cu + , Fe 3 + ) have a strong tendency to be complexed, while others ( N i 2 + , C u 2 + ) have less tendency. The metals that form strong complexes with chlorides tend to precipitate the metals that form weak chloro-complexes. A l l these could lead to lower metal extractions and/or slower leaching kinetics. These may be deleterious effects of the complexation on the leaching of some metals such as N i , Co and Cu , which are the objectives of this investigation. The present review established the fundamentals of studying thermodynamics of these mixtures. The review calculations also provided several important points relevant to the processing of base metal sulfides in this mixture; furthermore, this review suggested studying of the leaching chemistries of individual minerals and commercial sulfide products in this mixture. Many useful leaching parameters, such as leaching time, temperature, and acid concentrations, w i l l be explored in this work. 55 C H A P T E R 3 E X P E R I M E N T A L ASPECTS The experimental program of this work covered the following two areas: 1. Thermodynamic measurement in the mixture of H C l and M g C l 2 . Measurement of H C l activity coefficient . Measurement of M g C l 2 solubility in H C l solution 2. Study of leaching chemistry of individual sulfide minerals and the concentrates in the mixture of H C l and M g C l 2 3.1 Experimental procedures and methods The first objective of the experimental part was to measure the E M F of the cell o f interest. The experimental procedure of the E M F measurement method is summarized in A3.1 (Appendix 3). This experimental technique allows direct and true measurement of the E M F of the cell avoiding junction potential; however, the electrodes are not commercially available and need to be prepared as described here. Preparation methods for hydrogen/platinum and silver-silver chloride electrodes were adopted from the work of Bates [52] and are summarized in A3.2 and A3.3 , respectively (Appendix 3). The glassware design used for the Harned cell measurement (the cell uses hydrogen electrode) is taken from Harned, Bates & Robinson [45-48] and is summarized in A3.5 (Appendix 3), The experimental method to measure the solubility o f M g C l 2 in hydrochloric acid solutions was adopted from Demopoulos et al, [63]. Excess amount of M g C l 2 was brought in contact with solution. After adequate stirring, the liquid sample was taken at specific temperatures. The following procedure solved the initial difficulty of solid re-crystallization on sampling. A fritted glass tube was inserted into the system deep enough to be at the same temperature as that o f the solution. A t a certain sampling temperature, the vacuum applied to the fritted glass tube allows for collecting clear solution inside the tube, from where sample was taken and diluted for further analysis. Beforehand, both the fritted glass tube and pipette were kept in an oven at certain temperatures in order to eliminate salt re-crystallization due to temperature changes. In each leaching test, a certain amount of either individual mineral or a concentrate sample charge was leached in 150 or 500 m l solutions of mixture. The acid concentration varies, whereas M g C l 2 was kept at 2m. Literature [64] suggests that a minimum of 200g/l total chloride is required for the processing of sulfide minerals in this mixture. Initial tests o f this work confirmed that increased total chloride concentration has no effect on dissolution. Therefore, through the leaching 56 tests 2-2.4m of M g C l 2 was added to insure enough chloride in the solution. Solution samples were withdrawn at specific time intervals, weighed, and converted to volume scale assuming no density change due to the dissolved metals. Samples were diluted and prepared for analysis. A t the end of leaching, solids and liquids were separated by a vacuum filter, and solids were washed with DI water. Solution volumes and solid weights were recorded and sent for analysis. Selected solid residues were subjected to X R D and S E M - E D X . A n assay of M g C l 2 stock solution was determined by E D T A titration method as described in attachment A3.4 . A n assay of HC1 stock solution was determined by potentiometric titration as illustrated in attachment A3.6. Free acid concentrations of leach and wash solutions were determined by a method described in Wassink [65]. This method is most suitable for acidic solutions with an excess amount of hydrolysable cations. 3.2 Experimental instruments and set-up The experimental set-up is similar to the setup shown in Figure 3.1, however minor modifications have been made from test to test. Figure 3.1 Experimental equipment set-up 1. Reaction vessel 2. Stirrer 3. P t / H 2 electrode 4. A g / A g C l reference electrode 5. Thermosensor 6. Fritted glass tube for sampling 7. Condenser 8. Hydroseal to prevent air entrance 9. Glass l id with standard openings 10. Barometer 11. Temperature controller o f heating equipment 12. Water bath or heating mantle 57 Figure 3. 2 Experimental set-up for thermodynamic measurement Figure 3. 3 Experimental set-up for leaching test Leaching tests at lower temperatures (60\u00C2\u00B0C) were carried out in a controlled temperature orbital shaker, which is shown on the far left o f Figure 3.3. 58 A voltmeter, with precision of O . lmV (Denver Instrument pH/mV) , was used to measure cell potential for the thermodynamic measurements. A number of reference electrodes (single and double junction) were used to substitute A g / A g C l reference electrodes. The commercial electrodes were supplied from Thermoelectron. Scientific Glassware Co Ltd., Richmond, B C made the bases of the Pt/H2 and A g / A g C l electrodes, glass feature, and the 250ml reaction vessel with five standard openings according to specifications and drawings. Pt bases for hydrogen electrodes are platinum-blacked before use. Pt wire base for A g / A g C l electrode was coated with A g first, and then coated by A g C l . For the fhermoelectrical coating purposes Harrison 6203B B C power supplier by Hewlett Packard instrument was used. Due to complication of this method, the latest measurements were made using A g / A g C l bases of the used p H electrodes. For the initial leaching tests, mineral samples of pyrite, millerite, pentlandite, violarite, heazelwoodite, troilite and chalcopyrite were sized to minus 200 mesh and leached in 250ml-baffled flasks with continuous shaking. S/L ratio was kept 1/150 during the tests, neglecting the amount of liquid samples withdrawn for sampling. The temperature control and continuous stirring was provided by Environ-Shaker by Lab-Line Instrument Inc. The equipment shaking speed was kept in the range of 220-240 rpm. For those shaking tests, the system was kept closed to minimize evaporation losses. Solution samples were withdrawn inserting Teflon tubes through a rubber stopper. Tests at higher temperatures (concentrate samples and some minerals) were carried out in a one-liter reaction vessel sitting on a heating mantle. A n overhead-plastic stirrer was used to aid in enhanced mass-heat transfer and to prevent solution contamination. The stirrer was inserted through a septum stopper into the reaction vessel. The contacting areas o f the stirrer rod and septum, and any other joints of glassware were greased by silica oi l in order to prevent air entrance into solution and acid losses into surroundings. Double condensers (stacked) were used to minimize evaporation losses. Samples were withdrawn via Teflon tubes using a syringe. The temperature sensor connected to D M C type Dataplate 520 temperature controller controlled power input of the heating mantle. The water bath temperature was regulated by its own controller. The actual temperature of the leach system was controlled by DigiSense thermometer. 3.3 Chemicals, minerals and concentrate samples Hydrochloric acid of analytical grade supplied by Fisher was used as a stock solution for an acid supply. The concentration of the stock acid solution was determined by repeated potentiometric 59 base titrations and the end point was defined as shown in example A3.6 (Appendix 3). Stock solution of magnesium chloride was prepared by dissolving an analytical grade salt supplied by also Fisher, and the concentration was determined by E D T A titration. Densities of stock solutions were measured to enable preparation of mixture of solutions at exact ionic strength fractions (YM gci2 and Y H C i ) o f M g C l 2 a n d HC1. C u C l 2 and F e C ^ solutions were prepared by weight percent from reagent grade of respective salts supplied by Fisher. In addition, 0.1N N a O H , 0.1N HC1 and 0.1N AgNC>3 standard solutions were supplied by Fisher. A reagent grade H 2 P t C l 6 * 6 H 2 0 (dihydrogen hexachloroplatinate-IV hexahydrate) supplied by A l f a Aesar was used for Pt-black purpose. 0 .05M E D T A standard solution was bought from LabChem Inc. A l l individual mineral samples, except pyrite, were bought from Mineralogical Research Co. , San Diego, U S A . Crystal fragments o f each mineral sample were dry ground in a laboratory mi l l . Minor inclusion of pentlandite in very rich chalcopyrite sample was removed by hand after crushing mineral particles in a laboratory cone crusher. The nickel concentrate samples with low (two identical samples) and high (two identical samples) M g O content were supplied from B H P Bil l i ton as well as matte samples (four identical samples) from its Kalgoorlie smelter. Only three representative samples out of a total of eight were tested. These are indexed as 1BHP (IB) , 3 B H P (3B) and 5 B H P (5B) throughout the test work. A l l commercial samples were dried. N o further sample preparation was required for concentrate samples in terms of particle size. In contrast, a matte sample was re-ground in powderizing m i l l to pass P80 -200 um prior to leaching. Mineral characterization involved ICP for 30 elements, A A for common base metals and L E C O for total sulfur. International Plasma Laboratory in Vancouver performed all analyses. Some of sample characteristics are summarized in Table 3.1. Trace elements are excluded in this table. Table 3.1 Summary of mineral and concentrate samples Sample character Assay, % Mineral samples Test ID Origin S(tot) Co Cu Fe Mg Ni Si02 MgO Pentlandite P Sudbury, Ontario 30,41 0,10 3,22 63,01 0,01 4,07 \u00E2\u0080\u0094 \u00E2\u0080\u0094 Troilite T Alta Mine, Del Norte County, California 34,59 0,05 8,68 57,23 0,41 0,27 -- --Millerite M McCreedy West Mine, Levack, Ontario 32,81 0,24 0,52 13,78 0,03 59,29 - -Heazelwoodite H Lord Brassey Nickel, Tasmania, Australia 4,50 0,30 0,14 15,12 17,04 10,48 - -Violarite V Perseverance Mine, Agnew. WA, Australia 33,39 0,32 0,43 36,90 0,33 19,03 23,78 -Chalcopyrite C Frood Mine, Sudbury, Ontaria, Canada 33,29 0,00 35,30 33,75 0,01 0,28 1,31 -Pyrite P Unknown 54,06 0,06 0,04 49,96 0,01 0,01 - -Concentrate samples --Low MgO concentrate 1BHP(1B) WMC Resources, BHP Billiton, Australia 29,92 0,28 0,32 38,49 4,55 13,50 - 8,76 High MgO concentrate 3BHP (3B) WMC Resources, BHP Billiton, Australia 20,23 0,68 0,12 21,19 9,33 25,75 - 17,4 Matte 5BHP (5B) Kalgoorlie Smelter, BHP Billiton, Australia 24,74 1,02 2,33 5,48 0,03 74,20 -- --\u00E2\u0080\u0094 not analyzed 60 3.4 Solution preparation Each of the stock solutions, of which concentrations and densities were previously determined, were weighed to prepare solutions for the thermodynamic measurement and leaching tests. Solutions for the activity coefficient measurement were prepared at the exact same ionic strength fractions as per the references. A n example is shown in the next table. Table 3. 2 Solution preparation example for the E M F measurement m H C l m MgC12 I total YHCI YMgC12 V (HCl) V (MgCl 2 ) water, ml Sol.V, ml 2.00 0.00 2.00 1.000 0.000 29.26 0.00 70.74 100 1.74 0.09 2.00 0.871 0.129 25.50 2.51 71.99 100 0.20 0.60 2.00 0.100 0.900 2.93 17.58 79.49 100 1.00 0.33 2.00 0.499 0.501 14.60 9.79 75.61 100 0.48 0.51 2.00 0.242 0.758 7.09 14.80 78.11 100 0.32 0.56 2.00 0.159 0.841 4.66 16.42 78.92 100 Table 3.3 illustrates the solution preparation for the leaching tests. The volume of each compound represents the volume of each stock solutions required for a total of 500 ml solution. In both tables (Tables 3.2 and 3.3), m represents the molal concentrations of each salt, Y represents the ionic strength fractions of each salt, and V represents the volume of each stock solution required to prepare the specific volume of a mixture solution. Table 3. 3 Solution preparation example for leaching tests m H c i m MgC12 m CuCI2 mFeC13 I total YHCI YMgC12 YcuC12 Y F e C B V H C . VMgCl2 VcuCl2 ^ F e C B H 2 0 , ml 4.0 2.4 0.2 0.0 11.8 0.34 0.20 0.02 0.00 120.26 223.56 100.00 0.00 56.18 4.0 2.4 0.0 0:2 12.4 0.32 0.19 0.00 0.02 120.26 223.56 0.00 100.00 56.18 6.0 2.4 0.2 0.0 13.7 0.44 0.17 0.01 0.00 180.39 223.56 90.12 0.00 5.93 61 3.5 The basic principles of the S E M - E D X , X R D analyses These analyses were utilized to characterize the solid samples before and after leaching. The composition tables in section 4.4 ( X R D , S E M - E D X analyses) were created based on the S E M - E D X results. Note that this method has an accuracy o f \u00C2\u00B1 5 % . These analyses are operated based on the following principles. For the S E M analysis, the sample of interest is bombarded by beam of electrons. The interaction of the electron beam and the sample produces a series of electron scatters with different energy level, which are analyzed by a sophisticated microprocessor. The electrons produced are classified as either elastic or inelastic in terms of energy level. The inelastic electrons are low energy electrons deflected from the surface of sample. They mostly have electron energies of less than 50eV, and give information of the surface topography and a black and white, three-dimensional image of the sample. The elastic electrons are any electrons that interact with the primary electron beam to produce a specific energy from the collision and retain most of its energy. These electrons are categorized as: \u00E2\u0080\u00A2 Backscattered electrons-yielding topological, compositional and crystallographical surface information. \u00E2\u0080\u00A2 Absorbed current-which enables the study of the internal structure of semi-conductors or (EBIC). \u00E2\u0080\u00A2 Auger electrons-contain elemental and chemical information of the surface layers. \u00E2\u0080\u00A2 Characteristic X-ray Radiation-yields microanalysis and distribution o f elements of a given sample. A typical S E M has the ability to analyze a particular sample utilizing any of the above mentioned methods. Unfortunately, each type of analysis considered is an additional peripheral accessory for the S E M . The most common accessory equipped with a S E M is the energy dispersive x-ray detector or E D X (sometimes referred to as EDS) . This type o f detector allows a user to analyze a samples molecular composition. In an S E M , x-rays are produced by accelerating the primary electron beam with enough current to pass through the sample thereby interacting with the elements inner core electrons. When enough high-velocity electron bombardment contacts the inner most electron shell o f an atom, it forces the orbiting electron to be kicked out. Subsequently, this results in the neighboring outer electrons to move into the vacant inner electron shell. The release of energy from the escaping 62 electrons from the inner most orbiting shell or core electrons are analyzed and measured based on their classification type. For the X R D analyses, the three-dimensional structure of non-amorphous materials, such as minerals, is defined by regular, repeating planes o f atoms that form a crystal lattice. When a focused X-ray beam interacts with these planes o f atoms, part of the beam is transmitted, part is absorbed by the sample, part is refracted and scattered, and part is diffracted. X-rays are diffracted by each mineral differently, depending on what atoms make up the crystal lattice and how these atoms are arranged. A detector detects the X-ray signal; the signal is then processed either by a microprocessor or electronically, converting the signal to a count rate. When an X-ray beam hits a sample and is diffracted, we can measure the distances between the planes of the atoms mat constitute the sample by applying Bragg's Law. The Bragg's Law is nX = 2 d sind where the integer n is the order of the diffracted beam, X is the wavelength of the incident X-ray beam, d is the distance between adjacent planes of atoms (the J-spacings), and 9 is the angle of incidence o f the X-ray beam. Since we know X and we can measured, and can calculate the rf-spacings. The geometry of an X R D unit is designed to accommodate this measurement. The characteristic set of d-spacings generated in a typical X-ray scan provides a unique \"fingerprint\" of the mineral or minerals present in the sample. When properly interpreted, by comparison with standard reference patterns and measurements, this \"fingerprint\" allows for identification of the material. 63 C H A P T E R 4 R E S U L T S AND DISCUSSIONS 4.1 Thermodynamic measurement 4.1.1 E M F measurement of the cell M g C l 2 - H C l - H 2 0 using Pt/H 2 - Ag/AgCl electrodes with no junction Table 4.1 lists the experimental E M F and corrected E M F to latm hydrogen pressure for constant ionic strength of two and at each ionic strength fractions of magnesium chloride-YM gci2-Activi ty coefficients of H C l in each mixture were calculated using equation 2.19 (or equation reported in Appendix 3). The last column in Table 4.1 lists the activity coefficients of H C l determined applying the Harned rule comparing to experimentally determined logynci by least square method. The measured potential is corrected to the hydrogen pressure of latm as follows: Ecorr=Emws+^-where,LE^RTI2F\n[16QlpHi] = RT[160 (4.1) P - barometric pressure, p H 0 - water pressure, and h - depth of jet in mm. The water vapor pressure is taken from thermodynamic reference [54] and attached as A3.7 . Table 4.1 Solution compositions, measured and corrected E M F , calculated activity coefficients for H C l at 25\u00C2\u00B0C m Hci I total YHCI p ^ meas ^ corr. *log Y HCl **H.E log y HCI 2.00 2.00 1.00 0.1885 0.1889 -0.0167 -0.0186 1.74 2.00 0.87 0.1942 0.1946 -0.0255 -0.0272 1.50 2.00 0.75 0.2000 0.2004 -0.0318 -0.0354 1.00 2.00 0.50 0.2163 0.2169 -0.0627 -0.0521 0.48 2.00 0.24 0.2404 0.2410 -0.0856 -0.0693 0.32 2.00 0.16 0.2485 0.2491 -0.0552 -0.0748 * ** Calculated activity coefficients based on measured E M F Calculated activity coefficients based on Harned equation Figure 4.1 presents the log values of activity coefficients for H C l . The activity coefficients-experimentally measured, determined by the Harned equation, the references [43] and the calculated values by the Meissner method by both approaches, are plotted in this figure at temperatures 25, 35 and 45\u00C2\u00B0C. Additional measurements were made at temperatures of 50 and of 60\u00C2\u00B0C (intended to make measurements up to 100\u00C2\u00B0C); however, it was difficult to obtain a stable E M F reading above 60\u00C2\u00B0C. The Harned rule has been convenient to express this kind of series measurements at a constant total ionic strength. It expresses the experimental activity coefficients as a function of concentration: 64 i\u00C2\u00B0gio YHO = log rlc, - aJyusch - PJy2MgCll ( 4 2 ) where YHCI is the activity coefficients of H C l in the mixture; YHCI0 is the activity coefficients of H C l when it is present in its pure solutions at the same total ionic strength as the mixed solution; a and /? are so called Harned coefficients; and yMgCl is the ionic strength fraction of the magnesium chloride in the mixture, which is given by: JW\u00C2\u00BB 2 = 3mMga2 '{mHci + mMgck ) (43) Table 4.2 compiled the Harned equation coefficients for each set of measurement at different temperatures. It is clear from Figure 4.1 the calculated activity coefficients based on the Harned coefficients give better results than the experimentally determined values. One way to cross check validity of these coefficients is to compare H C l activity coefficient with its pure state at the same total ionic strength. A t I =2.0, YHCI = 1.009 at 25\u00C2\u00B0C [43], in our case YHCI = 0.98, which means the Harned values agree with the experimentally obtained values. Table 4. 2 Harned equation coefficients Temp, \u00C2\u00B0C logY\u00C2\u00B0Hci a A P A A 1A 2 (error) 25 -0.0186 -0.0334 0.0000 0.0008 35 -0.0285 -0.0380 0.0000 0.0004 45 -0.0399 -0.0487 0.0090 0.0001 The results obtained during these measurements tell us that the current equipment set-up, instruments and experimental methods are applicable to get reasonable estimation of y H ci- The E M F measurement sounds so simple, but in reality, it is not straightforward. It is very sensitive to any small change in test conditions such as solution composition, temperature, and as well as the electrode condition. When temperature gets higher, one rarely obtains a stable reading using H 2 -A g / A g C l couple with no junction. It is also expected that the same difficulty w i l l be encountered in solutions with higher ionic strength. The interests of hydrometallurgists are the higher temperatures and the upper limits of the concentration. Through these E M F measurements variable combinations of electrodes (pH probe, p H half cell-Calomel/Saturated KC1 , H 2 - A g / A g C l , H2(gaS)-Ag/AgCl/Saturated KC1) have been utilized in order to choose the best combinations in terms of convenience in use and accuracy of the results. The H2(gas)-Ag/AgCl/Saturated KC1 combination gives rather stable reading; however, sensitivity of H 2 electrode disables the use of these electrodes for the leach solutions. Therefore, regular p H combination electrode is preferred for the purpose of p H and redox potential measurement of the leach solution. 65 Figure 4.1 Log y H ci values: experimental, calculated b y the Harned equation vs. reference and the calculated values b y the Meissner method a) 25\u00C2\u00B0C b) 35\u00C2\u00B0C c) 45\u00C2\u00B0C 0.05 -0.09 0 0.5 1 yMgcn-Ionic strength fraction 0.05 0.03 0.01 -0.01 u a >-g> -0.03 -0.05 -0.07 -0.09 -0.11 Roy [43] Calc [Dixon] \u00E2\u0080\u00A2Calc [Meisner] \u00E2\u0080\u00A2 1=2 Experimental \u00E2\u0080\u00A21=2 Harned Eq. 0 0.5 yMgci2-Ionic strength fraction -0.13 0 0.5 1 yMgci2-Ionic strength fraction 4.1.2 Solubility of M g C h in water and H C l solutions The next part o f the solution property measurement was the solubility measurement of MgCl2 salt in acid solutions with varying concentrations. The solubility of MgCl2 is determined as follows (Table 4.3 and Figure 4.2) as a function of acid concentration and the temperature. Table 4. 3 Summary of M g C h solubility [g/1] in acid solutions H C l Molarity Temperature, \u00C2\u00B0C 22 50 75 82.75 6 242.8 333.2 428.4 452.2 3 371.3 428.4 485.6 499.9 1.5 423.7 471.3 0.75 455.8 499.9 523.7 537.9 0 (H 2 0) 485.6 504.6 523.7 557.0 A s acid concentration increases, the solubility of M g C l 2 decreases due to the limited availability of water molecules because of hydration. Overall, solubility of M g C ^ is determined to be within range of 243-557 g/1 over this specific acid concentration and temperature ranges. Figure 4. 2 Solubility dependence of MgCb on acidity and the temperature 600 200 -I 1 1 1 1 0 25 50 75 100 Temperature, \u00C2\u00B0C 67 4.1.3 Summary of the thermodynamic measurements The E M F measurement of the cell o f H C l - M g C l 2 - H 2 0 allowed determining the activity coefficients of H C l in this mixture. Unfortunately, the activity coefficients of the H C l are not sufficient to assign individual ion activities (e.g. H + activity) since these require both acid and salt activities. Therefore, for high ionic strength solutions at high temperature, the individual ion activities can only be estimated based on the Meissner's method followed by the approaches described in Jansz (Note that extensive amount of calculations were carried out in the literature review including the mixtures used in these experiments). The reference electrodes with no junction effect ( H 2 ( g a s / A s - A g C l couple) were limited especially in concentrated solutions at higher temperatures due to unstable reading of E M F . Although, the commercial electrodes (single and double junction electrodes whose compartments were filled concentrated KC1 solution) provided a stable reading potential, the junction potential must be eliminated as accurately as possible. Nevertheless, the Henderson equation, which is applicable for calculating junction potential, is only valid up to an ionic strength of six. Therefore, the Meissner's method followed by individual ion calculation best estimates the property of strong solutions at higher temperatures. The solubility measurement of MgCh in aqueous solutions o f hydrochloric acid provides valuable data for the use of mixture of M g C k and H C l . A n increasing temperature results an increased solubility, whereas an increasing acid concentration results in a decreased solubility of MgCl2. The solubility limits of M g C l 2 in water are determined to be 485.6 and 557 g/1 at 22 and 82.5\u00C2\u00B0C, respectively. These limits decreased to 243 and 452 g/1 in 6m acid solutions at 22 and 82.5\u00C2\u00B0C, respectively. Exceeding these limits w i l l result in \"salting out\" effects. Therefore, the M g C b concentration should be kept according to this range in order to the get full effect of the solution for the further leaching tests. 68 4.2 Individual mineral leaching The thermodynamic calculations in section 2.2.6 predicted the dissolution behavior for the selected sulfide minerals in the mixture of H C l and M g C ^ . To prove the consistency of these predictions, a few of the sulfide minerals were tested here. 4.2.1 Pyrite (P) A high purity pyrite sample, which displays peaks for only pyrite on its X-ray patterns, was used in this study. The chemical analysis of the mineral sample showed an iron content of 49.86 wt pet and a sulfur content of 54.06 wt pet. Experiments were performed to study the influence of the acid concentration, total chloride concentration, temperature and the leaching time on the dissolution of iron from this mineral. Figure 4. 3 Pyrite leaching: Effects of Figure 4. 4 Pyrite leaching: Effect of MgCh acid concentration, temperature and time concentration at 60\u00C2\u00B0C a) Effect of acid concentration: To determine the effect of the acid concentration on the dissolution of pyrite mineral, experiments were carried out with acid concentrations in the range 0 to 10m. The results are shown in Figure 4.3, where we can observe that the acid concentration has no effect at 25\u00C2\u00B0C (the lowest curve on this figure). This is what was predicted for pyrite in section 2.2.6 (Figure 2.14). On the other hand, an increasing acid concentration has little promoting effect on iron dissolution in the acid range 1 to 6m; thereafter iron dissolution declines in the acid concentration range from 6 to 10m at 69 60\u00C2\u00B0C (upper 3 curves on Figure 4.3) b) Effect of total chloride concentration: To determine the effect of total chloride on the dissolution of pyrite, experiments were carried out in solutions with 3m H C l mixed with varying concentration of M g C ^ in the range of 0 to 2.75m. The results are shown in Figure 4.4 from where we observe that the increasing total chloride has no effect on iron dissolution from pyrite. c) Experiments were carried out at 25 and 60\u00C2\u00B0C to determine the effect of temperature on iron dissolution from pyrite. A t room temperature the pyrite mineral did not leach yielding less that one percent iron extraction, whilst at 60\u00C2\u00B0C about 6% of iron dissolved. Pyrite shows very refractory leaching behavior in this mixture at all temperatures. This is consistent with the thermodynamic predictions in section 2.2.6 for pyrite. d) To determine the effect of leaching time, samples were taken at 4, 8 and 24-hour intervals. The results suggest that the longer retention time could result in some small degree of increased iron dissolution from pyrite mineral. e) Generally, the results of pyrite dissolution are consistent with the thermodynamic calculations as predicted in section 2.2.6 (Figure 2.14). I.e. in no case was there a sufficient proton activity to promote the significant break down of pyrite. 4.2.2 Millerite (M) Chemical assays and calculated mineral compositions of the millerite sample used in this study are summarized in Table 4.4. The mineral sample contains about 97% of millerite and about 2% of pentlandite in the form of Ni4.5Fe4.5Sg. These two minerals were detected on X-ray diffraction patterns as shown in section 4.4.2; however, a trace amount of chalcopyrite was present based on elemental analysis. A n amount of iron assay in this sample belonged to the pentlandite and chalcopyrite. The trace element compositions were neither included in this table nor used in metal extraction calculations. Table 4. 4 Assays and mineralogical compositions of millerite sample Elements Assay,% S(tot) N i C u Fe Ca Co Pb M g 32.00 59.29 0.52 13.78 0.09 0.24 0.13 0.03 Minerals Assay, % M S (Millerite) (NiFe) 9 S 8 CuFeS 2 97.40 2.39 0.21 70 Throughout the experiments, the effects of acid concentration, temperature and the leaching time were studied. Figures 4.5 and 4.6 show experimental results of the millerite mineral dissolution in the mixture of M g C ^ and H C l . Note that the metal extractions were calculated based on the solution sample analysis and the head grade of the solid feed sample. Figure 4. 5 extraction Millerite leaching: Ni 8 = a ft^\u00E2\u0080\u0094-^\u00E2\u0080\u0094q 0 4-=\u00C2\u00B0= 0 5 10 Acid Concentration, Molal 4hrat60C \u00E2\u0080\u0094O\u00E2\u0080\u00948hr at 60C -6\u00E2\u0080\u009424hr at 60C \u00E2\u0080\u0094o\u00E2\u0080\u009424hr at 25C| Figure 4. 6 extraction Millerite leaching: Fe 0 5 10 Acid Concentration, Molal -4hrat60C \u00E2\u0080\u00A224hrat 60C \u00E2\u0080\u00A28hr at 60C \u00E2\u0080\u00A224hrat 25C Table 4. 5 Dissolution of metals from millerite leaching at 60\u00C2\u00B0C Acid cone, molal N i extraction, % Fe extraction, % Leaching time, hr 4 8 24 4 8 24 1 4.55 6.70 12.45 7.45 10.73 17.32 3 3.66 5.28 13.82 6.01 8.32 19.03 6 3.82 5.46 15.59 6.46 8.91 18.18 10 6.02 6.84 56.72 6.76 7.64 67.74 These results are summarized in Table 4.5. The summary o f results is as follows: a) Effect of temperature: Temperature has a promoting effect on millerite dissolution. There is no dissolution of millerite at room temperature, whereas dissolution increases by 15-20 times at 60\u00C2\u00B0C below 6m H C l in mixture. A t 10m H C l in mixture, this increase reached up to 60 times. These can be seen comparing the lowest curves on Figure 4.5 and 4.6 against upper curves. The effect of temperature is consistent with the thermodynamic predictions in section 2.2.6 for millerite (Figure 2.14). b) Effect of acid concentration: Hydrochloric acid concentration in the range of 1 to 6m does not have positive effect on increased dissolution. Thereafter, nickel dissolution 71 increases up to 40% in the acid range of 6 to 10m. Thermodynamically, it was predicted that millerite would dissolve at p H of above zero as discussed in section 2.2.6 (Figure 2.14). The leach solution of 2m M g C l 2 and 6m H C l creates pH~-2.0 at 60\u00C2\u00B0C (Table 2.7) according to the estimated p H values for mixtures. Many factors, such as kinetic factors, could cause the discrepancy between the prediction and experimental results, c) Effect of leaching time: We can observe that increased leaching time plays a major effect on millerite dissolution combined with increasing acid concentration. Sharp increase of dissolution occurs just after 6m H C l in mixture with 24 hours of retention time. Further, the dissolution is expected to increase as acid increases. Hence, it is clear that the strong acid and longer leaching time are beneficial for millerite dissolution. 4.2.3 Violarite (V) Chemical assays and the dominant mineral compositions of the violarite sample used in this study are summarized in Table 4.6. The sample contained about 50% violarite and 23% pyrite. The violarite weight percent was calculated based on the nickel assay. The remaining sulfur contributed to form pyrite mineral since the mineral supplier noted it as a possible association. Since all sulfur is used up for these two minerals, the remaining iron (17%) occurred probably not in the form of sulfides. Note that 46% o f the total iron was associated in neither one of the two main sulfide minerals. Table 4. 6 Assays and mineralogical compositions of violarite sample Elements assay, % S(tot) N i Fe Co Cu M g 33.39 19.26 36.93 0.35 0.45 0.33 . Mineral assay, % FeS 2 FeNi 2 S 4 * 23.10 49.46 17.02 0.35 0.45 0.33 * remaining iron, not included in pyrite or in violarite Experiments were conducted to study the effects of acid concentration, temperature and leach time on direct non-oxidative leaching of the violarite sample. The results are graphically represented in Figures 4.7, 4.8 and 4.9 in terms of N i , Co and Fe dissolutions, respectively. Selected results are summarized in Table 4.7. 72 Figure 4. 7 Violarite leaching: Ni extraction (2m MgCl 2 ) Figure 4. 8 Violarite leaching: Co extraction Figure 4. 9 Violarite leaching: Fe extraction 40 30 20 10 0 1 1 1 1 0 2 4 6 8 10 Acid Concentration, Molal \u00E2\u0080\u00A24hrat 60C \u00E2\u0080\u00A224hrat 60C \u00E2\u0080\u00A28hr at 60C \u00E2\u0080\u00A224hrat 25C i 1 1 r 0 2 4 6 8 10 Acid Concentration, Molal \u00C2\u00BB4hrat 60C \u00E2\u0080\u00A224hr at 60C \u00E2\u0080\u00A28hrat 60C \u00E2\u0080\u00A224hrat 25C I 2 4 6 8 10 Acid Concentration, Molal \u00E2\u0080\u00A24hrat60C \u00E2\u0080\u0094O\u00E2\u0080\u00948hr at 60C \u00E2\u0080\u00A224hr at 60C \u00E2\u0080\u0094o\u00E2\u0080\u0094 24hr at 25C Table 4. 7 Ni & Fe extractions from violarite sample at 60 C Acid cone. Ni extraction, % Fe extraction, % Leaching time, hr molal 4 8 24 4 8 24 1 8.67 9.00 16.00 37.24 39.05 50.05 3 4.64 6.70 19.61 37.36 37.68 53.38 6 7.81 9.28 18.76 36.48 38.68 53.26 10 7.35 10.31 32.30 34.36 35.83 58.27 Based on these results, the following effects are observed: a) Effect of acid concentration: Below 6m H C l in the mixture, violarite dissolution is unaffected by acid concentration at both 25 and 60\u00C2\u00B0C. A n increase o f acid concentration up to 10m results in a 22% increase of N i extraction from violarite. The same results were found for Co. On the other hand, Fe shows a different behavior. It is clear that this mineral sample contained acid soluble iron that dissolved rapidly in a very low acid concentration (Figure 4.9). Thereafter, Fe dissolution slightly increases as violarite dissolves as shown by the upper curve on Figure 4.9. A s mentioned before, 46% of the total iron is associated in neither one of two main sulfide minerals; b) Effect of temperature: A s always, an increasing temperature promotes violarite leaching. It is expected that even higher temperatures (up to 100\u00C2\u00B0C or just below the boiling point of a leach solution) could result in higher metal extractions; c) Effect of leaching time: Prolonged leaching time seems to promote mineral dissolution. 73 However, an increased temperature and a strong acid solution could result in high dissolution within a short leaching period. The violarite feed contained about 17% of iron that was associated neither in violarite nor in pyrite. The extra iron in this sample could be in the form of Fe20\"3 or FeO(OH). If it were hematite, the following reaction would take a place. F e 2 0 3 + 6HC1 = 2 F e C l 3 + 3H 2 0 ( R X N 4.1) It could cause FeNi 2 S4 leaching by: . F e N i S 4 + 6 F e C l 3 = 2 N i C l 2 + 7 F e C l 2 +4S ( R X N 4.2) In this case, violarite is leached by the effect of a ferric ion not by the effect of higher acid concentration. Therefore, feed and solid residue were subjected to X R D in section 4.4 to clarify the leaching chemistry of violarite. 4.2.4 Troilite (T) Chemical and mineral assays of troilite sample are summarized in Table 4.8. Based on the iron content, 90% of the total sample represented troilite mineral. Remaining sulfur (-0.018wt%) was associated with other base metals in this sample. Table 4. 8 Assays and mineralogical compositions of troilite sample Elements assay, % S(tot) 34.59 Fe 57.23 Co 0.05 Cu 8.68 M g 0.41 Ni 027 Minerals assay, % FeS (Troilite) 90.09 0.05 8.68 0.41 0.27 The effects of the acid concentration, temperature and the leaching time on troilite dissolution were investigated and the results are presented in Figures 4.10-4.12. Based on leaching kinetic curves of troilite leaching, the following kinetic evidence can be drawn: a) Effect of acid concentration: A t room temperature, iron dissolution of troilite steadily increases as acid concentration increases from 0 to 5m. Thereafter, dissolution of iron almost doubled in the range o f 5 to 6m as seen in Figure 4.10. A steady increase of C u extraction is observed as shown in Figure 4.11. On the other hand, 95% of iron dissolved within the acid range of 1 to 3m (Figure 4.12) at 60\u00C2\u00B0C. The solutions in this case w i l l have pH\u00E2\u0080\u00941 as it was predicted in section 2.2.5.2 (Table 2.7). The thermodynamic calculation in section 2.2.6 (Figure 2.14) predicted that the troilite leaches in the p H range of -1.0 to -0.5 at these temperatures. Therefore, the experimental results are 74 consistent with both thermodynamic predictions. b) Effect of temperature: A n increasing temperature results in a much higher dissolution of troilite even at lower acid concentration. c) Effect of leaching time. Dissolution percent increases with longer retention time but leaching time has a lesser effect than the acid concentration and the temperature. A low acid concentration at a higher temperature w i l l result in near complete dissolution of troilite mineral. Figure 4. 10 Troilite Figure 4.11 Troilite Figure 4.12 Troilite leaching: leaching: Fe extraction at leaching: Cu extraction at Fe extraction at 25 & 60\u00C2\u00B0C 25\u00C2\u00B0C (2m MgCI 2) 25\u00C2\u00B0C 4.2.5 Heazelwoodite (H) Dominant elemental assays in the heazelwoodite sample are shown in Table 4.9. Heazelwoodite (NisS 2 ) mineral composition in this sample is calculated based on N i assay of this sample. Assuming all nickel assay forms heazelwoodite minerals, about 13.5 percent of total weight accounts for this mineral. Table 4.9 Assay of heazelwoodite sample, % S(tot) S i 0 2 MgO A l Ca Co Cu Fe M g N i 4.51 23.78 0.00 1.63 1.29 0.31 0.13 14.58 15.83 9.92 The effects o f the acid concentration, temperature and the leaching time were investigated, 75 and only dissolution results of N i from this sample are presented here. Some o f selected results of heazelwoodite leaching are summarized in Table 4.10. More studies w i l l be covered in matte leaching which is an excellent representative of the heazelwoodite mineral. Figure 4.13 Heazelwoodite leaching: Ni extraction (2m MgCh) 80 a \u00C2\u00A9 60 X 40 W \"3 Nic 20 0 0 2 4 6 8 10 Acid Concentration, Molal \u00E2\u0080\u00A24hrat 60C \u00E2\u0080\u00A224hrat 60C \u00E2\u0080\u00A28hrat 60C \u00E2\u0080\u00A224hrat 25C a) Effect of acid concentration: Heazelwoodite started to dissolve at the beginning stage of acid addition as low as l m . Thereafter, there is no significant effect o f acid addition on N i extraction; b) Effect of temperature: Although, heazelwoodite leaches at room temperature, the temperature increase from 25 to 60\u00C2\u00B0C resulted in 6 times faster leaching kinetics (two overlapped lower curves). c) It seems that the retention time did not play as significant a role as the temperature on the heazelwoodite dissolution and kinetics. d) The heazelwoodite dissolution was predicted in section 2.2.6. According to Figure 2.14, a significant amount of heazelwoodite (a[Ni 2 + ]=l) dissolves at p H of about 2.0. Table 2.7 (section 2.2.5.2) estimates the p H of this solution (2m MgCl2 -2mHCl) is about -0.7 at 60\u00C2\u00B0C. The discrepancy between the prediction and the experimental results may be resulted by kinetic factors and the composition o f this sample. Table 4.10 Summary of heazelwoodite leaching at 60\u00C2\u00B0C Acid cone. N i extraction, % Fe extraction, % Leaching time, hr molal 4 8 24 4 8 24 1 29.80 34.94 55.06 55.73 70.03 79.07 3 26.95 31.73 53.22 72.12 72.57 82.65 6 27.70 32.93 60.35 66.02 66.15 79.85 10 25.21 31.41 64.76 60.76 62.42 74.56 76 4.2.6 Chalcopyrite (C) Based on the element assays, possible mineralogical compositions of this sample were calculated as shown in the next table along with assays of the predominant elements. On the X-ray diffraction analysis, peaks were only detected for chalcopyrite. Table 4.11 Assays and mineralogical compositions of chalcopyrite sample Element assay, % S(tot) 33.29 \u00E2\u0080\u00A2 Cu 35.30 Fe 33.75 Ni-0.28 Mineral assay, % CuFeS 2 (Chalcopyrite) 99.5 Ni4.5Fe4.5Sg (Pentlandite) 0.5 The mineral was subjected to leaching in the mixture of strong acid and magnesium chloride solutions to determine the effect o f acid concentration, and the leach temperature. A l l tests were run with four hours of retention time. The leaching results are presented in Figure 4.14 and summarized in Table 4.12. Figure 4.14 Chalcopyrite leaching: Metal extractions 25 20 15 10 5 0 0 2 4 6 8! Acid concentration, m * - C u @60C, Cl-total =284g/l -o\u00E2\u0080\u0094 Fe @ 60C, Cl-total = 284g/l \u00E2\u0080\u00A2 Cu @ 100C, Cl-total=390g/l \u00E2\u0080\u00A2 Fe @ 100C, Cl-total = 390g/l Table 4.12 Summary results of chalcopyrite dissolution Acid cone. Cu extraction, % Fe extraction, % molal 60\u00C2\u00B0C 100\u00C2\u00B0C 60\u00C2\u00B0C 100\u00C2\u00B0C 1 0.7 1.65 3 1.44 3.27 5 1.51 3.57 7 1.51 22.18 4.02 22.71 Dissolution characteristics of the most abundant copper mineral are as follows: a) Effect of acid concentration: A t low temperature (60\u00C2\u00B0C), an increasing acid concentration has a weak effect on metal dissolution. It is observed that the iron extraction is higher than copper extraction. This evidence suggested a selective leaching of iron over copper, 77 leaving copper enriched solid residue as expected, thermodynamically (Table 2.11 reaction for chalcopyrite). To study further, a single test was carried out in mixtures with 7m H C l and 2m M g C l 2 at 100\u00C2\u00B0C. b) Effect of temperature. A t a higher temperature, both temperature and the acid concentration resulted in increased metal dissolutions. The results at 100\u00C2\u00B0C clearly indicate that the same amount of copper and iron dissolved meaning that there was no phase transformation from chalcopyrite to its copper enriched phases Cu2- xS. The solid residue was then subjected to X R D to address i f there was any phase transformation. In contrast to the prediction in section 2.2.6, chalcopyrite could be directly attacked by high acid at high temperature forming cupric and ferric chlorides. This needs more study and investigation. 4.2.7 Summary of individual mineral leaching A few of the sulfide minerals were tested here in mixtures of H C l and MgCl2. Each mineral showed a unique dissolution behavior in this mixture depending on acid concentration and the temperature. Pyrite is not leachable in this mixture. Increases of H C l concentration, up to 10m, and the temperature increase from 25 to 60\u00C2\u00B0C, have resulted in only 6% iron dissolution from pyrite. In contrast, troilite - another iron mineral leached reasonably well yielding 90% iron extraction at 60\u00C2\u00B0C in mixtures with 3m H C l . The chalcopyrite showed a refractory behavior at 60\u00C2\u00B0C in solutions with varying H C l concentrations. However, 22% of both copper and iron were extracted at 100\u00C2\u00B0C in mixtures with 7m H C l . Among the nickel sulfides, heazelwoodite showed a highly favorable dissolution behavior in mixtures with H C l concentrations as low as l m within a short leaching period. Millerite dissolution depends on acid concentration, above 6m of acid, 60% of nickel leached into solution at 60\u00C2\u00B0C. The behavior of violarite depends on acid level; however, only 30% of nickel transferred into solution in mixtures with 6m of acid at 60\u00C2\u00B0C. A n increasing temperature plays a significant role on increased dissolution, e.g. 15-20 times increase in case of millerite. Overall, both temperature and the acid concentration have significant effects on metal extraction and kinetics for the selected sulfide minerals. Most of the leaching results of individual minerals were consistent with those predicted thermodynamically (in sections 2.2.5-2.2.8) in terms of acid concentration and temperature. 78 4.3 Commercial concentrates and matte leaching The leaching results o f individual minerals suggest possibilities and conditions of processing commercial sulfide concentrates in this mixture. The objectives of these leaching experiments were to achieve high metal extractions, to investigate leach parameters such as leach time, temperature, acid concentration and the addition of CuCl2 or FeCL, to the leach system. Equipment set-up for this commercial sample leaching was the same as for the individual minerals. The only differences were solid/liquid ratio and the temperature. S/L ratio of 25g/500ml was maintained for these tests. A l l tests were run at 100\u00C2\u00B0C except initial tests for a low M g O concentrate. 4.3.1 Low MgO concentrate One of two identical concentrate samples with low M g O content was tested. Several key results are outlined below. Note that all test conditions are in the absence of oxygen. a) Effect of leach time and the temperature: Initial leaching tests (at 60\u00C2\u00B0C) were carried out in solutions with 10m H C l mixed with 2m MgCL. to get an idea of retention time. The results show high metal recoveries can be achieved within 4 hours leaching time (Figure 4.15). The present goal is also to achieve high recoveries in solutions with lower acid concentration. Therefore, subsequent test were carried out with a retention time of 4 hours and an increasing temperature up to 100\u00C2\u00B0C. b) Effect of acid concentration: In order to determine the effect of acid concentration on dissolution, tests were performed in the mixtures of H C l - M g C l 2 with acid concentrations of 2, 6 and 8m. Results show that metal dissolutions are strongly acid dependent yielding around 75% extractions of N i , C u and Fe and about 52% extraction for Co at 6m H C l in mixture with a retention time of 4 hours. The results at 8m H C l in mixture were obtained with a retention time of 2 hours (The points at 8m H C l on Figure 4.16). N i , Fe extractions reached up to 95%, and Co extractions increased up to 75% (Figure 4.16). c) Effect of C u C ^ : To determine the effect of C u C l 2 addition, the concentrate sample was leached in mixtures o f 6m H C l - 2 m MgCL. at 100\u00C2\u00B0C with a retention time of 4 hours. 0.5m C u C b was added into solution. This addition resulted in a decrease of extraction for both N i and Fe of 10%, an 18% decrease for Co, and a 22% increase for C u compared to metal extractions with no copper addition (Figure 4.17). Thus, addition of CuCL. has a negative effect on increased dissolution of N i and Co. There are two possible reasons for 79 these low metal extractions. The C u ion tends to get complexed with chloride ions than N i 2 + ion (Table 2.12). Therefore, cupric ion in solution w i l l have more affinity to precipitate nickel chlorides as discussed in section 2.2.7 (as illustrated by the example in section 2.2.7, R X N 2.7). The second possible reason is explained by strong tendencies of C u reduction by H2S gas as discussed in section 2.2.6b ( R X N 2.2). The higher C u concentration favors this reduction. Therefore, there was no beneficial effect of the C u 2 + ion on metal extraction; instead, the reduced copper (or copper sulfide may have formed) covered the mineral surface. d) Effect of FeCb : The leach conditions for these tests were the same as before, instead of adding CuCl2, 0.5m FeCLj was added into leach solution. Addit ion of FeCb resulted in a 31, 33 and 40% decrease of Co, N i and C u extractions, respectively. The addition of this salt did not promote extractions (Figure 4.18). The reason is explained by the same phenomena as before. Even F e 3 + has more tendencies to form chloro-complexes, which may precipitate (or decreasing extraction) other metal cations such as N i 2 + , C o 2 + and C u 2 + . In addition, as discussed in section 2.2.6b, the F e 3 + has a strong tendency to be reduced by H2S gas ( R X N 2.3) forming elemental sulfur. Elemental sulfur may form a passive layer on the particle surface preventing further reaction. e) Figure 4.19 shows the dissolution of magnesium from this concentrate. 63% of magnesium was dissolved with no addition of cupric and ferric chlorides. The addition of 0.5m of either salts resulted in about 96% magnesium extraction from this concentrate. Figure 4.15 Low MgO concentrate: Effect Figure 4.16 Effect of acid concentration of leach time (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) (2m MgCI 2 , t=2 & 4 hr at 100\u00C2\u00B0C) 80 Figure 4.17 Effect of C u C l 2 Figure 4.18 Effect of FeCl 3 addition (6m H C l , t=4 hr at addition (6m H C l , t=4 hr at 100\u00C2\u00B0C) 100\u00C2\u00B0C) Figure 4.19 The dissolution of magnesium (6m H C l , t=4 hr at 100\u00C2\u00B0C) 0.0 0.1 0.2 0.3 0.4 0.5| Added salt concentration, m Generally, high metal extractions from low M g O concentrate can be obtained in mixtures with higher than 8m acid concentrations with a retention time o f 2 hours at 100\u00C2\u00B0C. The strong oxidants such as cupric and ferric chlorides do not promote metal extractions in this specific test condition (oxygen has been excluded). The decreased metal extractions were caused by complexation and solubility behaviors of metals that present in solution. Also , interaction of dissolution product gases such as H 2 S and H 2 with the metal cations in solution has a deleterious effect on metal extractions as explained above. 4.3.2 High MgO concentrate One of two identical concentrate samples with high M g O content was tested. The effects of leach time, acid concentration and the addition of strong oxidants were studied and the following results were obtained. a) Effect of leach time and temperature: The results of individual minerals and previous concentrate sample have suggested a leach temperature of 100\u00C2\u00B0C for the subsequent tests for high metal recoveries. So, all tests on this sample were run at this temperature. To determine the effect of leach time, the concentrate sample was leached in the mixture of 6m H C l - 2 m M g C l 2 . Unl ike low M g O concentrate, the high M g O concentrate yielded high metal extractions within retention time o f one hour (Figure 4.20). Subsequent tests were carried out with a retention time of one hour. 81 Figure 4. 20 High MgO concentrate: Effect of leaching time (6m H C l - 2m M g C l 2 at 100\u00C2\u00B0C) Figure 4. 21 Effect of acid concentration (2m MgCI 2 , t=lhr at 100\u00C2\u00B0C) Figure 4. 22 Effect of C u C l 2 addition (6m H C l -2m M g C l 2 , t=lhr at 100\u00C2\u00B0C) Figure 4. 23 Effect of FeCI 3 addition (6m H C l - 2m M g C l 2 , t=lhr at 100\u00C2\u00B0C) Figure 4. 24 Dissolution of magnesium (6m H C l - 2m M g C l 2 , t=lhr at 100\u00C2\u00B0C) 100 0.00 0.25 0.50[ CuCh concentration, m 100 0.0 0.1 0.2 0.3 0.4 0.5| Added salt concentration, m b) Effect of acid concentration: The mixture with 2m H C l in mixture has resulted poor metal extractions. In contrast, 95% of N i , 91% of Fe, 76% of Co and 19% of C u extractions were obtained from leaching in mixtures with 6m acid concentrations (Figure 4.21). c) Effect of C u C l 2 addition: To determine the effect o f C u C l 2 addition on metal dissolutions, tests were performed in mixtures of 6m H C l - 2 m M g C l 2 with the addition of 82 0.05, 0.2 and 0.5m of C u C l 2 . The addition o f a small amount o f C u C l 2 (0.05m) has resulted in the decrease of metal extractions of 66.5, 90, 46 and 72% for Co, Cu , Fe and N i , respectively. Thereafter, addition of 0.2 and 0.5m o f C u C l 2 has resulted in the increase of metal extractions; however, the metal extractions were lower than the extractions obtained with no addition of C u C l 2 (Figure 4.22). One possible reason for this low extraction may be explained by the discussion in section 2.2.6b. The C u 2 + ion is more favorable to be precipitated by H 2 S gas, which is the product of nickel sulfide dissolution by the direct attack of acid ( R X N 2.2). Hence, the reduced product of copper prevents further dissolution o f minerals. d) Effect of FeCb : To determine the effect of FeCl3 on metal dissolutions, the sample was leached in the mixture of 6m H C l - 2 m M g C l 2 with the addition of 0.2 and 0.5m of F e C ^ . A s FeCb increases, extractions of N i , Co and C u were decreasing. The addition of 0.5m FeCi3 has resulted in the decrease o f metal extractions o f 60-64% for N i , Co and Cu, and the increase of 8% for iron extraction (Figure 4.23). The reason for low metal extractions with the addition of ferric ion can be explained as follows: the ferric reacts with H 2 S gas as in R X N 2.3 (section 2.2.6b); as determined before, FeS is easily leached in strong acid solution; and the elemental sulfur covers mineral surface causing decreased mineral dissolution. e) Figure 4.24 shows the dissolution o f magnesium from this concentrate. About 83% of magnesium was dissolved with no addition of cupric and ferric chlorides. The addition of 0.5m of either salts resulted in about 96% magnesium extraction from this concentrate. Generally, the high M g O concentrate is leachable in mixtures with 6m acid concentrations with a retention time of one hour at temperatures of 100\u00C2\u00B0C at ambient pressure. The high M g (the most problematic substance during P A L ) was dissolved with a significant extent (83%). In a process flowsheet this magnesium could later be recovered as described by pyrohydrolysis process. The addition of C u C l 2 decreased N i , Co and Fe extractions. Similarly, addition of FeCh promoted Fe extraction and depressed N i , Co, C u extractions from this concentrate sample (Figure 4.22). The reasons of lower metal extractions with the addition o f ferric and cupric chlorides were explained by the same phenomena described in the previous section. 83 4.3.3 Nickel matte One of four identical matte samples supplied by B H P Bi l l i ton was studied here. The effects of temperature, acid concentration and the leaching times were investigated. a) Effect of temperature: Temperature has a significant effect on increased extraction and faster kinetics. Temperature increase from 60 to 100\u00C2\u00B0C has resulted in the following: \u00E2\u0080\u00A2 This accelerated the dissolution kinetics of nickel by almost four times. The same extractions were obtained in 8 and 2 hours o f the leaching period at 60 and 100\u00C2\u00B0C, respectively (Upper curve on Figure 4.25 against upper curve on Figure 4.29). \u00E2\u0080\u00A2 Faster kinetics and even higher cobalt extractions were observed due to the temperature increase (compare upper curve on Figure 4.26 against second curve from the top on Figure 4.29). Similar phenomena were observed for iron extraction from this matte sample. \u00E2\u0080\u00A2 Copper extraction from this sample increased by 9% in a four times shorter leaching period as a result of temperature increase from 60 to 100\u00C2\u00B0C. 83% C u extraction at 60\u00C2\u00B0C in 8 hours vs. 92% in 2 hours at 100\u00C2\u00B0C (Figure 4.27 against lower curve on Figure 4.29). b) Effect of acid concentration: Most of the metals except copper dissolved to a significant extent in solutions with acid concentrations as low as 3m (Compare Figures 4.25-4.28). In contrast, copper in this matte sample started to dissolve in solutions with 6m H C l in the mixture. This suggested leaching matte samples at 100\u00C2\u00B0C in solutions with 6m of acid in order to get higher metal extractions (including copper). The low copper extraction is explained by the following phenomena. A s discussed in section 2.2.6a, heazelwoodite dissolution in strong acid produces both H 2 S and H 2 gases (Table 2.11). Both gases are strong reducing agents. Among metal cations in solution ( N i 2 + , C o 2 + , C u 2 + ) , cupric ion has a strong tendency to be reduced by any one of these gases ( R X N 2.2 and R X N 2.4 in section 2.26b). This reduction results in a lower copper extraction. c) Effect of leaching time: Tests at 60\u00C2\u00B0C showed that leach time increase resulted in higher metal extractions (Figure 4.25 - Figure 4.28). On the other hand, there was almost no effect o f leaching time increase from 2 to 4h at 100\u00C2\u00B0C. Most o f the valuable metals of this matte sample were easily extracted at 100\u00C2\u00B0C in solutions with 6m H C l within one hour. 84 Figure 4. 25 Matte leaching: Figure 4. 26 Matte leaching: Figure 4. 27 Matte Ni extraction at 60\u00C2\u00B0C (2m Co extraction at 60\u00C2\u00B0C (2m leaching: Cu extraction at MgCl 2 ) MgCl 2 ) 60\u00C2\u00B0C (2m MgCl 2 ) Figure 4. 28 Matte leaching: Fe extraction at 60\u00C2\u00B0C (2m MgCl 2 ) Figure 4. 29 Matte leaching: Metal extractions at 100\u00C2\u00B0C (2m MgCl 2 ) 0 2 4 6 8 10 Acid Concentration, Molal \u00E2\u0080\u00A22hr at 60C \u00E2\u0080\u00A28hr at 60C \u00E2\u0080\u00A24hr at 60C \u00E2\u0080\u009E100 a .2 95 w OS S 90 W 3 85 80 Test cond: 6m HC1-2mMgC12, 100C, 1 2 3 4 Leach Time, hr - N i - a \u00E2\u0080\u0094 C o - C u \u00E2\u0080\u0094o\u00E2\u0080\u0094Fe Overall, nickel matte can be easily processed in the mixture of H C l and M g C l 2 just below the boiling point of the solution (~100\u00C2\u00B0C). A t this temperature, complete dissolution of nickel matte is expected within one hour of leaching in solutions with 6m of hydrochloric acid. 6m of acid is required in order to achieve higher copper extraction. 85 Table 4.13 Metal extractions from matte sample at 60\u00C2\u00B0C id traion, lal Metal extractions, % id traion, lal Co Cu Fe Ni O C o < n 1 Leach time, hr Leach time, hr Leach time, hr Leach time, hr o o 2 4 8 2 4 8 2 4 8 2 4 8 1 3 6 10 37.18 48.44 63.66 58.92 70.69 78.49 70.32 71.23 80.04 64.57 63.82 78.44 0.50 0.75 0.59 1.82 3.78 5.63 8.79 38.97 83.43 41.99 46.48 91.09 52.23 63.87 8038 63.22 71.12 79.72 70.97 70.69 81.75 64.28 63.48 80.44 30.91 41.90 64.16 59.65 73.19 94.57 73.47 75.36 94.36 66.82 66.05 96.05 Table 4.14 Metal extractions from matte sample at 100\u00C2\u00B0C L.time, hr Co Cu Fe Ni 2.00 97.41 91.83 96.46 99.92 4.00 97.31 90.20 96.79 100.00 4.4 XRD, S E M - E D X analyses The leaching results described in the previous sections give rise to a number of questions. Among them are how individual minerals and the commercial samples were dissolved in this mixture. To address these issues, X R D and S E M - E D X analyses were performed on the selected solid samples before and after leaching. 4.4.1 Pyrite and leach residue Figure 4.30 displays X-ray patterns of pyrite and its leach residue. Both patterns show exact peaks for natural pyrite. Figure 4. 30 X-ray patterns of pyrite and pyrite leach residue (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) inoline < | v | ? [ B [ B | l T g SCAN:40.0;80.0/0.05/1.5(sec)Xu(40kV.20inftM[ma>i)\u00C2\u00B03676.01/24/0607:02p <40.0Q-80.0Q> P PDF P CPS P 21(0) 10.0 j ] Sj u Overlaid patterns I i 1 \, It BBnBDBBBBBIIIBB 1 Jj|042-1340> Pyrite-FeS2 T|-M*l\u00C2\u00BBH*irTiiRFll 2 ~ ~ a. J J Pyrite \u00E2\u0080\u00A2 FeS2 >-. Q_ ^ Residue \u00C2\u00BB \u00E2\u0080\u00A2 \u00C2\u00AB \u00C2\u00A3 >> >> >< CL CL Q-\u00C2\u00A3 .-2 Feed >s 40 J Jl Jl 60 > a. 70 80 Thermodynamically, the following reaction is predicted for the direct dissolution of pyrite: FeS 2 + 2 H C l ( g ) = F e C l 2 ( a q ) + S\u00C2\u00B0 + H 2 S ( g ) ( R X N 4.3) Since the leach residue did not show any peaks corresponding to elemental sulfur, a S E M -E D X was performed to investigate the composition of the leach residue. 87 The following image was obtained from the S E M - E D X analysis of the pyrite leach residue. The image contains mostly pyrite (Spot2) and sulfur covered pyrite particles (Spotl), based on the S E M - E D X analysis. Figure 4.31 S E M picture of pyrite leach residue (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) Table 4.15 Composition of pyrite leach residue (10m HCI-2m MgCI 2 at 60\u00C2\u00B0C) Overall Area 1 Spot 1 Spot 2 Spot 3 wt % as Fe S 0.495 0.505 Fe S 0.497 0.503 Fe S 0.455 0.545 Fe S 0.491 0.509 Fe S 0.505 0.495 wt % as Fe FeS 2 0.055 0.945 Fe FeS 2 0.060 0.940 FeS 2 S 0.977 0.023 Fe FeS 2 0.048 0.952 Fe FeS 2 0.075 0.925 Table 4.15 summarizes the composition of the spots identified on Figure 4.31. Since only about 6% of pyrite was leached, the reaction product (elemental sulfur) cannot be identified on the X R D patterns. This explains why Figure 4.30 displays the peaks only for pyrite. Hence, the qualitative values obtained by S E M - E D X were used to calculate the weight fractions o f pyrite and elemental sulfur as shown in Table 4.15. In most cases, the product is iron enriched, except for Spot l . Spotl consists of about 98% pyrite and 2% elemental sulfur. Therefore, pyrite dissolution can be concluded as follows, based on the leaching, X R D and S E M - E D X results: \u00E2\u0080\u00A2 Pyrite is the refractory mineral in this mixture; yielded only about 6% iron extraction for 24 hours of leaching time in a mixture of 10m H C l and 2m M g C l 2 ; 88 \u00E2\u0080\u00A2 Elemental sulfur seems to be the dissolution product as expected ( R X N 4.3); however, this trace amount of elemental sulfur was hardly detected on the X R D ; \u00E2\u0080\u00A2 About 5% o f iron enrichment occurred on several areas (mostly black covered particles on Figure 4.31) of the leach residue. 4.4.2 Millerite and leach residue Figure 4.32 displays X-ray patterns of the millerite sample and the leach residue. A s reported before, this sample contains over 95% millerite and about 3% pentlandite. Both minerals were detected on the X-ray patterns o f the feed. The patterns of the leach residue contained exactly the same peaks of millerite and pentlandite as before leaching. N o peaks disappeared and no new peaks appeared as a result of leaching; however dissolution was only partual. Therefore, millerite, as predicted thermodynamically dissolved, as shown below. N i S + 2 H C l ( g ) - N i C l 2 ( a q ) + H 2 S ( g ) AG\u00C2\u00B0 = -15.76 kcal/mol ( R X N 4.4) Figure 4. 32 X-ray patterns of millerite and its leach residue (10m HCl-2m M g C l 2 at 60\u00C2\u00B0C) 89 4.4.3 Violarite and leach residue In section 4.2.3, a question was posed regarding whether the violarite mineral was leached by the effect of the ferric ion or by the effect of a strong acid. To answer this question, X R D patterns of the feed and the solid residue are compared below. The peaks on the X R D pattern of the feed (Figure 4.33) were identified as violarite and pyrite. These are consistent with those found in section 4.2.3. The extra iron in the feed was identified as siderite (FeCO^), in addition to a trace amount of hydromolysite (FeCl3*6H20). N o peaks for hematite were found, meaning that the violarite was leached by the effect o f increased acid concentration. Figure 4.34 shows compared peaks of violarite feed and the leach residue. A s a result of the 30% N i and about 60% Fe extraction, the peaks for violarite and pyrite remained. This is consistent with the fact that pyrite is refractory as described before. Some peaks, belonging to siderite, disappeared resulting in about 60% iron extraction. This proves that the violarite was leached by the direct attack of the strong acid in the mixture. Figure 4. 33 X R D patterns for violarite mineral sample Sr^:\u00C2\u00AB.0 /120.nrt .05/1 .5(stcLQ44<^ POFT CPS T~ 2\u00C2\u00AB0)|o.O * sj Counts ill X III \u00C2\u00B11 ||fe ~K\ii.M. i t H P>\ VMt-FMX* | i | s | c | n | * | h | t i | i i | * M i i i | x | h i r - M | \u00C2\u00BB | > l * l i l Siderite \u00E2\u0080\u00A2 FeC03 Hydromo s^ile \u00E2\u0080\u00A2 FeC]3!6H20 | ? c 0) c ] s \u00E2\u0080\u00A2 i ^ HF D : | ill 1 2 * v \ i \u00C2\u00A7 I P J i i l l 1 ^ f 1 * 1 7 m 5 ) 0 90 Figure 4. 34 Compared X R D patterns for violarite and leach residue 4.4.4 Chalcopyrite and leach residue X R D patterns of chalcopyrite and its leach residue both show peaks for chalcopyrite (Figure 4.35). A s reported before, about 22% of this mineral was dissolved at 100\u00C2\u00B0C. According to the non-oxidative dissolution of chalcopyrite, the leaching product is supposed to contain some covellite-CuS (or copper enriched products-Cu2.xS of CuFeS2), as predicted thermodynamically. CuFeS 2 + 2 H C l ( g ) = CuS + F e C l 2 + H 2 S ( g ) A G 0 = -3.8 kcal/mol ( R X N 4.5) Since X R D o f the leach residue did not show any peaks corresponding to CuS, and displays peaks for only chalcopyrite after 22% of dissolution, the probable dissolution of chalcopyrite is expected as follows: CuFeS 2 + 4 H C l ( g ) = CuCl 2 (a q ) + FeCl 2 ( aq) + 2H 2S( g) AG\u00C2\u00B0=47kJ/mol at 100\u00C2\u00B0C ( R X N 4.6) However, Gibbs' free energy o f this reaction is positive, the leach solution and the X-ray patterns of leach residue suggested that the dissolution o f chalcopyrite follows this reaction [ R X N 4.6]. 91 Figure 4. 35 100\u00C2\u00B0C) X-ray patterns of chalcopyrite and its leach residue (7m HCl-2m M g C l 2 at 92 4.4.5 Low MgO concentrate and its leach residue The detailed mineral assays of the low M g O concentrate were kindly supplied by B H P Bil l i ton (Table 4.16). X-ray patterns of the concentrate sample were consistent with the B H P Bil l i ton 's mineral assays (Figure 4.36). Note that the Ni-sulfide was identified as pentlandite and the sample contained a substantial amount of pyrrhotite. Table 4.16 Detailed mineral assay of low MgO concentrate (supplied from BHP Billiton) Ni-sulphide 32.58% Ni-arsenide 0.03% Pyrite 11.81% Pyrrhotite 29.38% Chalcopyrite 0.55% Tochilinite 3.00% Magnetite 2.72% Serpentine 5.51% Talc 9.93% Magnesite 0.36% Dolomite 0.84% Mg-silicate 2.48% Hydr-carb 0.13% Felsic 0.68% Total 99.98% Figure 4. 36 X-ray patterns of low MgO concentrate P\u00C2\u00ABillatti\u00C2\u00BBe-|F\u00C2\u00AB.Ni|933 PyrrtioSte W Fel.jiS Tafc2M-Ms33i*010i0H|2 PytSe \u00E2\u0080\u00A2 FeS2 Feed Overlapped patterns of the low M g O concentrate before and after leaching shows diminished peaks for residue (Figure 4.37). X-ray patterns of the solid residue, where the highest metal recoveries were obtained, shows peaks for talc, pyrite and quartz (Figure 4.38). The pentlandite and 93 pyrrhotite peaks disappeared. This confirms that pyrrhotite and pentlandite are soluble in this mixture. Figure 4. 37 Overlapped X-ray patterns of low MgO concentrate and its residue (8m HCl-2m MgCl 2 ) irolme < | : | \u00E2\u0080\u00A2> 1 B | B | I H SCAN 40 0/100.0/0.05/1.5(sec], Cu(40kV.20n-A). I[m\u00C2\u00AB>0-1198. 02/03/06 02:20p lu Overlaid patterns r ppFr\" C P S r~ 2fl(o) joo |1B5-8.2-2.raw] 185-8.2-2 \u00E2\u0080\u00A2 Gartold i | T | - M - k l \u00C2\u00BB I H \u00C2\u00AB l * h l H l x l H ^ [ j i a ; ] K i i M ^ M k \u00C2\u00AB , f f . ' H Q Residue Feed - i \u00E2\u0080\u0094 i \u00E2\u0080\u0094 i \u00E2\u0080\u0094 i \u00E2\u0080\u0094 + -Figure 4. 38 X-ray pattern of low MgO concentrate residue (8m HCl-2m M g C l 2 at 100\u00C2\u00B0C) 1\u00C2\u00AB1 ' 1: I I \" I 111 042-1340> Pyrite \u00E2\u0080\u00A2 FeS2 Millerite - NiS Pyrite - FeS2 Quartz - Si02 Talc-2M -Mg3Si4O10(DH)2 94 Based on the element assays determined by the S E M - E D X , the major mineral assays were calculated (Table 4.17) for the selected spots on the S E M image (Figure 4.39). The calculated mineral assays were consistent with those on the X-ray patterns (Figure 4.38). The dark and light dark spots on the S E M image had no distinction in terms of elements and minerals. This means that it is unlikely that a product layer formed on the surface of particles. Therefore, the low M g O concentrate leached in strong acid solutions leaving a solid residue of pyrite, talc and quartz. Figure 4. 39 S E M image of low MgO concentrate leach residue (8m HCl-2m M g C l 2 at 100\u00C2\u00B0C) Table 4.17 Contents of low MgO concentrate residue (8m HCl-2m MgCl 2 ) 1B5-8.2-2 wt. fraction of elements based on SEM-EDX ID on SEM Image s Si O Fe Ni Mg Na Sum S SiO, O FeS2 Ni Mg 3 Si 4 O 1 0 (OH) 2 Sum Overall 0.177 0.183 0.392 0.139 0.042 0.062 0.002 0.996 0.017 0.240 0.078 0.299 0.042 0.318 0.994 Area 1 0.159 0.199 0.410 0.129 0.032 0.065 0.005 1.000 0.011 0.266 0.070 0.277 0.032 0.338 0.995 Spot 1 0.162 0.200 0.410 0.130 0.033 0.063 0.003 1.000 0.013 0.277 0.068 0.278 0.033 0.328 0.997 Spot 2 0.166 0.192 0.400 0.137 0.033 0.064 0.009 1.000 0.009 0.252 0.072 0.293 0.033 0.332 0.991 wt. fraction of elements and calculated minerals Addition of 0.5m C u C h into leach solutions of 6m H C l and 2m M g C h did not improve metal extractions (Figure 4.17). However, peaks for pentlandite disappeared as a result of 60% N i and 66% Fe dissolution. Peaks for pyrite and talc appear on the X-ray patterns o f the leach residue. New peaks for quartz also appear because of leaching (Figure 4.40). To address the lower metal 95 dissolutions, leach residue was subjected to S E M - E X D . S E M image displays particles covered with flaky type dark layer (Figure 4.41). The calculated mineral assays based on the S E M - E D X are consistent with the peaks appear on the X-ray patterns. In addition, a significant amount of copper assay was detected. The darker area (e.g. Area 4 in Figure 4.41) contains more copper as seen on the last row of a summary table (Table 4.18). Therefore, an excess amount of copper in the leach solution formed an insoluble layer (probably combined with leaching products) on the surface of particles, and retarded the dissolution of the minerals further. The S E M - E D X results confirm the reasons for low metal extractions as explained in section 4.2.1. This was actually predicted in section 2.2.6b ( R X N 2.2). Addition of FeCi3 showed similar retarding effects on the dissolution. This is most likely due to formation of a sulfur-enriched surface layer that may be expected to retard non-oxidative leaching. Figure 4. 40 X-ray pattern of low MgO concentrate and its leach residue (6m HCl-0.5m MgCl 2 -0.5m CuCl 2 ) JJV|?|B[B|I|'-'1 SCAN: 40.0/100.0/0.05/1.5(sec), Cu[40kV,20rrA), l{max)\u00C2\u00AB1199.02/03/06 02:2 <40.00-1QO.0Q> f\"~ P D F P CPS l ~ 28(0) |0.0 j j | C [1B5-0.05CuCI2.tawI 1B5-6m HCI-0_lj_?l | t | t | i\ : | B | X | - j :||008-0090> Pentlandite-|Fe,Ni)3S8 Overlaid patterns ItUlal'H c JS C JL8 T O rx3 -u a JO +-> c c ai Residue \u00E2\u0080\u0094 C _2 5 CL \L4 TJ C 03 \u00C2\u00AE C L Pentlandite \u00E2\u0080\u00A2 (Fe.Ni)9S8 Pyrrhotite-4H-Fe1-xS 40 50 TJO cr: (DO) T3 C 03 Q) Q. X! C 03 Feed \u00E2\u0080\u0094 \u00C2\u00A3= JO a 60 70 80 90 100 96 Table 4.18 Contents of low MgO concentrate leach residue (6m HCl-0.5m MgCl 2-0.5m C u C l 2 a t l 0 0 \u00C2\u00B0 C ) I lB5-6.05-0.5CuCl, wt. fraction of elements based on SEM-EDX ID on SEM Image S Si 0 Fe Ni Mg Cu Sum S Si02 O FeS2 Ni Mg 3Si 40,\u00E2\u0080\u009E(OH) 2 Cu Sum Overall 0.266 0.126 0.328 0.119 0.044 0.050 0.057 0.991 0.129 0.131 0.113 0.257 0.044 0.260 0.057 0.991 Area 1 0.267 0.127 0.332 0.115 0.042 0.050 0.059 0.992 0.135 0.136 0.115 0.247 0.042 0.258 0.059 0.992 Area 2 0.267 0.128 0.333 0.116 0.042 0.049 0.056 0.991 0.134 0.143 0.114 0.249 0.042 0.252 0.056 0.991 Area 3 0.268 0.128 0.332 0.117 0.043 0.049 0.055 0.992 0.134 0.143 0.113 0.250 0.043 0.254 0.055 0.992 Area 4 0.268 0.128 0.330 0.116 0.041 0.049 0.063 0.995 0.135 0.143 0.111 0.249 0.041 0.253 0.063 0.995 wt. fraction of elements and calculated minerals based on SEM-EDX 4.4.6 High MgO concentrate and its leach residue Major peaks on the X-ray patterns of the high M g O concentrate feed (Figure 4.42) were identified as pentlandite. The middle pattern on this figure corresponds to the solid residue of this sample, where high metal extractions were obtained in the mixture o f 6m H C l and 2m M g C l 2 with no additions of either C u C l 2 or FeCb . A s a result o f high metal extractions, pentlantite peaks disappeared in the leach residue pattern. The residue was then subjected to S E M for further investigation. 97 Figure 4. 42 X-ray patterns of high MgO concentrate and its leach residue (6m H C l and 2m M g C l 2 atlOO\u00C2\u00B0C) u * I v I ? I B | B| 11 -1 SCAN: 40.0/'80.0/0.05/1.5(scc), Cu[40kV.20irA), Umax).3980,02/03/06 03:07p r PDF r CPS r zm |o.o jj g gj Overlaid patterns [3B5-6.2-1 law] 3B5-6.2-1 \u00E2\u0080\u00A2 Ganbol l | \"I M - | + l \" I \u00E2\u0080\u00A2 I \u00C2\u00AB H - I B| x |-r | jj|0QB-0090> Pentlandite \u00E2\u0080\u00A2 [Fe.Ni)9S8 ?| - M -'I \u00C2\u00BBl \" 1 \u00C2\u00BB 1 i | g | x V 1 *>j(^\u00C2\u00A3fri j a \u00E2\u0080\u00A2I 121 \"' Jg ro itfcl a>o> a) t_ J) 2<\u00C2\u00A5j Residue m _5>0) a) \u00C2\u00ABB r Pentlandite - (Fe.Ni)9S8 Millerite - NiS Quartz-Si02 0^-Pictures taken from S E M - E D X are shown below. The overall picture displays mostly fine, black-colored particles with a few white particles (Figure 4.43 and Figure 4.44). Figure 4. 43 S E M image of high MgO concentrate leach residue (6m H C l and 2m M g C l 2 at 100\u00C2\u00B0C) 98 Figure 4. 44 S E M image of high MgO concentrate leach residue: Selected spots Based on the X R D and S E M - E D X analyses, the compositions of the leach residue were determined for the selected spots on Figure 4.44 and tabulated below. Table 4.19 Compositions of high MgO concentrate leach residues (6m HCl-2m M g C l 2 at 100\u00C2\u00B0C) Spots weight fraction as S U M Fe S Si O Cu N i Co M g S i 0 2 Overal 0.063 0.082 ~ 0.143 0.000 0.055 0.006 0.054 0.601 1.003 Areal 0.008 0.013 \u00E2\u0080\u0094 0.055 0.000 0.000 0.000 0.015 0.907 0.998 SpotlW 0.243 0.340 - -- 0.000 0.310 0.000 0.012 0.093 0.997 Spot2D 0.010 0.020 - ~ 0.000 0.000 0.000 0.009 0.951 0.989 Spot3D 0.000 0.022 0.036 - 0.000 0.000 0.000 0.009 0.934 1.000 Note: \u00E2\u0080\u0094 Stochiometric amount of either Si or O presents. Accounted in S i0 2 . According to this table, the leach residue mostly consists of quartz. Spotl (a few white dots on overall picture) represents un-leached nickel minerals. A minor amount o f iron, sulfur and magnesium was reported in the residue. The X R D , S E M - E D X results were very consistent to the dissolution results reported in section 4.3.2. Leaching results with the addition o f C u C l 2 are reported in Figure 4.22. To address the low recoveries, the leach residue was subjected to S E M - E D X . The addition of 0.05m C u C l 2 has resulted in the lowest metal recoveries. Major peaks on the X-ray pattern of the leach residue were identified as pentlandite (the middle pattern on Figure 4.45). Note that major peaks o f the feed were identified as pentlandite. 99 Figure 4. 45 X-ray patterns of high MgO concentrate and its leach residue (6m HCl-2m MgCl 2-0.05m C u C l 2 at 100\u00C2\u00B0C) r P D F r C P S r swiloo U w =J :) J I $ | ? j Bj B | 11 -\u00E2\u0080\u00A2 I SCAN: 40.0/80.0/0.05/1.5(sec). Cu(40kV,20ftA), l[max)>3980,02/03/06 03:07p I I Overlaid patterns [3B 5-6.2-005CuD2-60\raw] 3B 5-6. i i i ? 4i )+(<\u00C2\u00BB! 11 11 J I : | B | X H ;l|073-0515> Pentlandite-Ni4.5Fe4.5S8 5 i S 5 \u00C2\u00ABJ to Residue 3 0> : Q.Q.I T Millerite-NiS Pentlandite \u00E2\u0080\u00A2 Ni4.5Fe4.5S8 5 Feed U 1 Solution 5 <0.01 <0.2 0.7 O.l 0.17 7.07 O.01 FeC13 Solution - - - - - \u00E2\u0080\u0094 __ __ CuC12 Solution - - - - - \u00E2\u0080\u0094 ._ _. RE 1B5-2.2-100-4 Repeat 7 0.50 0.2 0.2 999 2.2 O.l 0.72 9.25 0.21 -Minimum detection 1 0.01 0.1 0.1 0.01 0.011 0.01 Maximum detection 9999 999 999 9999 999 9999: 999 Method ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 Certificated: 06E0997 Sample Name Al Sb As Ba Bi Cd Ca Cr: Co Cu Fe La Pb me/L Mg mg/L mg/L mg/L mg/L O01 mg/L mg/L mg/L mg/L! mg/L mg/L mg/L mg/L ms/L 1.91 69262 7 C4-1-3.5 <0.2 0.3 <0 2 O.l O.01 3.8 O.01! O.01 34.2 77.12 O.05 C4-3-2.5 <0.2 0.3 0.2 0.01 O.l O.01 3.3 O.01; O.01 70.33 152.4 O.05 1.86 51320.5 C4-5-1.5 <0.2 0.3 <0.2 0.05 O.l O.01 3.1 O.01 O.01 73.68 166.58 O.05! 1.98 30932.2 C4-7-0.5 0.2 0.4 <0.2 0.01 O.l 0.25 19 O.01 O.01 73.69 188.24 0.06 2.04 11501 RE C4-1-3.5 <0.2 0.4 <0.2 <0.01 O.l O.01 4.1 O.01 O.01 37.74 78.63 O.05 1.94 70762.2 Minimum detection 0.2 0.1 0.2 0.01 0.1 0.01 0.1 0.01 0.01 0.01 0.03 0.05 0.05 0.1 Maximum detection 9999 9999 9999 999 9999 ICPH20 999 9999 9999 9999 ICPH20 9999 9999 999 9999 9999 Method ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 Na Sample Name Mn Hg Mo I Ni P K Sc Ag Tl Ti W Zn Zr mgrt- m.g5r. mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L C4-1-3.5 0.13 <0.05 0.39 0.93 O.l <2 0.02 0.1 7 0.2 O.l O.l 9.2 1.38 C4-3-2.5 0.16 <0.05 0.51 3.27 O.l <2 0.01 0.25 6 0.2 O.l O.l 17.19 1.6 C4-5-1.5 0.15 <0.05 0.4 3.96 O.l <2 0.01 0.37 6 0.2 O.l O.l 22.6 1.04 C4-7-0.5 0.19 <0.05 0.27: 4.49 1.1 <2 0.02 0.48 6 0.2 O.l O.l 27.35 0.4 REC4-1-3 5 0.14 O.05 0.36 i 1.01 O.l <2 0.01 0.1: 7 0.2 O.l O.l 9.45 1.38 0.01 9999' Minimum detection 0.01 0.05 0.02 0.02: 9999 0.1 2 0.01 0 02 1 0.2 0.1 0.1 \"\"\"9999 0.01 Maximum detection 999 9999 9999! 9999 9999 100 999: 9999 999 999 999 Method ICPH20 ICPH20I ICPH20! ICPH20! ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 ICPH20 PH g i CN 1.98 1.92 00 00 2.88 0.39 0.94 0.36 Os o CO 0.05 Os Os Os Os ICPH20 N s <0.01 0.08 0.57 0.75 <0.01 <0.01 0.03 O.01 <0.01 0.01 Os Os Os ICPH20 g Vi o o V vo o o V VO CD O V VO CD CD V o d V vo o CD V VO O CD V vo o CD V VO P CD V vo p CD Os Os Os O CN X PL, O a N g VO VO \u00E2\u0080\u00A2^ r vo 0 0 CN vo vq CD CO CN VQ vd Os \u00E2\u0080\u00A2o CN VO CN Os CO CN Os vo CN VO p CD Os Os Os Os O CN X Cu O 4> U. g vo co Os o 00 00 vo Os CN Os O vb Os vo CO 00 vo $ CO vo <1 vd vo 00 00 CN vq CN vo vo 00 CN Os CO CO 00 CO o oo CO O CD Os Os Os Os o CN X a, o > g VO CD OS vq CD CN CO p CO vo d CD CD V CO d VO o CD q^ CD O CD Os Os Os O CN PC CL, O 3 O g co co vo CN CN cn vo 9 00 o CO CD CN \"st vo o>' vs. CD VO vo VO o CO CO OS CN CN CN CO 0 0 oo Os CN CO O CD Os Os Os Os O CN X CL, o g CD V o V O V o V o V O V o V o V o V CD Os Os Os Os O CN ac CU o o O g 00 Os c-^ Os 00 o 0 0 CN vb vo CN os CO 0 0 t-^ VO CN VO 0 0 vb vo r-; vb CN 0 0 CO CO ^ Os O CD Os Os Os Os o CN X PL, O H g 0 0 CO CO VO CN CN CN CO CO CD CN CO CD 00 CO CD Os Os Os o CN X Cu O V, o g cn o CN o CO 0 0 CN CN OS CD CN CO p CN O CD Os Os Os Os o CN X PL, o P g CN CD V CN CD V CN CD V CN d V CN d V CN CD V CN d V CN CD V CN CD V CN CD OS Os Os O CN X Cu O o \u00E2\u0080\u00A2a g VO vb Os vo VO CO vo CO Os Os Os CN Os CO CN f-v CD Os Os Os Os 8 CL, o w g CO O CD 00 CD CD Os d O CD V 00 d O CD V CD O CD Os Os Os O CN a CL, o \u00E2\u0080\u00A2a O g p o V p o V p CD V p d V p CD V O CD V O d V p CD V O CD V O CD Os Os Os o CN X PH o g CO CO VO CO V V Os Os Os Os o CN X cu O 3 g o V o V o V o V o V o V o V CD V O V CD Os Os Os Os o CN X PL, u S g 9 $ OS *T Os CN VO CN od CO vo CN 00 CD VO CN \u00E2\u0080\u00A2<* i vo CO *q vb ^ vo CD CN CN CO VO vo CN CO OS CN CD CD Os Os Os Os o CN X Cu O PQ \u00E2\u0080\u00A2 i g p o V p CD V CD VD p CD V p CD V o d V P CD V p CD V O CD Os Os Os o CN CL, o o s \u00E2\u0080\u00A2 i g vp CD 00 vq CD Os VO CD VO vq CD vo d CD CD 0 0 CD vo vq CD CN p CD Os Os Os Os O CN X Cu o < i g 00 vb CN \u00E2\u0080\u00A2o o vp VO 00 CO 00 vb Os CD Os vb 0 0 vb CN CS CD Os Os Os Os O CN X Cu O OD \u00E2\u0080\u00A2 i g VO p CD V VO O CD V VO O CD V VO p CD V o d V VO O CD V VO O CD V VO O CD V vo p CD V VO O CD Os Os Os Os O CN X Cu o JO Vi \u00E2\u0080\u00A2A o V o V o V o V o V O V o V o V CD V CD Os Os Os Os O CN DC 0H o C s s CO [--; \u00C2\u00BBo' CN CN vq VO 00 vo vb CO \u00C2\u00ABo CN O CN Os 0 0 CN vo 0 0 CN 00 \u00C2\u00BBo CN O CD Os Os Os O CN ac 0H o < g 0 0 \u00E2\u0080\u00A2n 00 Os VO \u00C2\u00B0 0 ( N CN Os VO VO \u00C2\u00BB-> CN vb \"O 00 \u00E2\u0080\u00A2O CN CD Os Os Os Os o CN X Cu o 2 Os CN s 00 CD Os CN r-. vo o VO 0 0 vb CN c--CN Os 3' 0 0 CO vb CD CN CO Os CO VO VO CO vb o CN vo CD Os Os Os Os o CN ac PH o to a o 1 o CO c O I o to a O 1 o in a o 1 o Vi G o I O CO C o 1 o in C o 1 o to c O 1 o to OH ft H a> c o I o w c o 1 o C O | o Vi c o 1 o Vi c o 1 O w G o 1 o w c O 1 O CO c O 1 o Vi \"S3 a ft! CN vo tZ VO O 5tt S u o z w| o NO o vo o \u00E2\u0080\u00A2o o vo vo PQ CO d VO d O vo CD vo O VO vo m o vo o VO CD vo o vd VO PQ CO d vo a o a> tu d P vd \u00E2\u0080\u00A2A s o \u00C2\u00AB l CN vb VO m CO CD VO (6 CN vb vo PQ CO o vo a o VO O CD CN vb vo PQ CO o vo a O VO CD CD CN vb VO PQ CO d o d \u00C2\u00BBA o vd vA PQ CO s C O o V <1> 1 a O u \"8 \u00E2\u0080\u00A2o o a a> s a) 55 4> 1 OT| o vo a o \u00C2\u00ABo d vA o vd \u00E2\u0080\u00A2A PJ CO d vo O VO CD i vo O vb i vo 2 d a o a> U. d vA o vd PQ CO d VO d o U. d \u00E2\u0080\u00A2A o vd \u00E2\u0080\u00A2o m CD vo (6 CN vb VO PQ CO & d vo <$\u00E2\u0080\u00A2 CN vd \u00E2\u0080\u00A2A PQ CO o VO o vo o CD CN \b vo PQ CO O VO d o vo o CD CN vb 1 vo CQ CO d u \"i d vA o vd v~\u00C2\u00BB PQ CO C O \u00E2\u0080\u00A2j* o S3 O \u00E2\u0080\u00A2a 4> ~* 4> \u00E2\u0080\u00A2o o S Appendix 2 Summary of thermodynamic calculations Table A2.1 Best fit equation* values for lnynci & lnyMgci2 for integral part o f the osmotic coefficient calculation aj ns 3i H C l M g C l 2 3.000 -0.007 3.506 -0.013 2.000 0.085 2.253 0.002 1.000 -0.005 1.388 0.645 0.000 -0.306 0.062 -0.888 0.000 0.007 0.000 0.033 y\u00C2\u00B1 where rrij molal of species j , e.g. H C l or M g C l 2 . Table A2.2 Comparative table of calculated osmotic coefficient and mean ionic activity coefficient values against reference [42] values for pure H C l or pure M g C b electrolyte at 25\u00C2\u00B0C. r\u00C2\u00B1 4> r\u00C2\u00B1 m M g C l 2 H C l M g C l 2 H C l M g C l 2 H C l M g C l 2 H C l R E F E R E N C E V A L U E S C A L C U L A T E D V A L U E S 0.1 0.861 0.943 0.528 0.796 0.837 0.995 0.502 0.788 0.2 0.877 0.945 0.488 0.767 0.895 0.986 0.466 0.754 0.3 0.895 0.952 0.476 0.756 0.914 0.979 0.464 0.740 0.4 0.919 0.963 0.474 0.755 0.934 0.976 0.476 0.735 0.5 0.947 0.974 0.480 0.757 0.958 0.978 0.495 0.735 0.6 0.976 0.986 0.490 0.763 0.986 0.983 0.520 0.739 0.7 1.004 0.998 0.505 0.772 1.017 0.991 0.549 0.746 0.8 1.036 1.011 0.521 0.783 1.048 1.001 0.581 0.755 0.9 1.071 1.025 0.543 0.795 1.079 1.012 0.616 0.766 1 1.108 1.039 0.569 0.809 1.111 1.024 0.654 0.778 1.2 1.184 1.067 0.630 0.840 1.176 1.052 0.739 0.808 1.4 1.264 1.096 0.708 0.876 1.247 1.083 0.841 0.843 1.6 1.347 1.126 0.802 0.916 1.329 1.116 0.966 0.883 1.8 1.434 1.157 0.914 0.960 1.419 1.150 1.120 0.928 2 1.523 1.188 1.051 1.009 1.517 1.185 1.307 0.977 2.5 1.762 1.266 1.538 1.147 1.769 1.273 1.926 1.120 3 2.010 1.348 2.320 1.316 2.023 1.360 2.812 1.293 3.5 2.264 1.431 3.550 1.518 2.275 1.445 4.046 1.500 4 2.521 1.517 5.530 1.765 2.527 1.532 5.734 1.751 4.5 2.783 1.598 8.720 2.040 2.783 1.622 8.001 2.059 5 3.048 1.680 13.920 2.380 3.043 1.717 10.997 2.439 5.5 \u00E2\u0080\u0094 1.763 2.770 1.817 14.900 2.907 6 - 1.845 3.220 1.920 19.917 3.475 126 Table A2.3 Calculated mean ionic activity coefficients of H C l ( ^ \u00C2\u00B1 ( / / c / ) ) in mixture of H C l - M g C k compared with reference values T tntnl Temp 25.0 \u00C2\u00B0C Temp 35.0 \u00C2\u00B0C Temp 45.0 \u00C2\u00B0C X lUlal Y MRC12 Reff[43] Calc.D Calc. J Y MgC12 Reff[43] Calc.D Calc. J Y MRCL2 Reff [43] Calc.D Calc. J Calc.D and Calc.J - applied approaches described in Dixon [61 and Jansz [71, Meisner [31-381 papers, respectively 0.10 0.000 -0.097 -0.103 -0.103 0.000 -0.099 -0.103 -0.103 0.000 -0.101 -0.103 -0.103 0.10 0.313 -0.098 -0.108 -0.105 0.155 -0.100 -0.105 -0.104 0.155 -0.102 -0.105 -0.104 0.10 0.508 -0.099 -0.111 -0.106 0.508 -0.102 -0.111 -0.106 0.313 -0.103 -0.108 -0.105 0.10 0.765 -0.100 -0.114 -0.107 0.765 -0.103 -0.115 -0.107 0.508 -0.104 -0.111 -0.106 0.10 0.870 -0.101 -0.116 -0.107 0.870 -0.103 -0.116 -0.107 0.870 -0.106 -0.117 -0.108 0.25 0.000 -0.120 -0.128 -0.128 0.000 -0.123 -0.128 -0.128 0.000 -0.126 -0.128 -0.128 0.25 0.499 -0.125 -0.139 -0.132 0.331 -0.126 -0.136 -0.132 0.331 -0.129 -0.136 -0.132 0.25 0.640 -0.126 -0.142 -0.133 0.499 -0.127 -0.140 -0.133 0.499 -0.131 -0.141 -0.134 0.25 0.872 -0.129 -0.148 -0.135 0.872 -0.131 -0.149 -0.136 0.872 -0.135 -0.150 -0.137 0.50 0.000 -0.122 -0.134 -0.134 0.000 -0.126 -0.135 -0.135 0.000 -0.130 -0.136 -0.136 0.50 0.101 -0.125 -0.137 -0.136 0.101 -0.128 -0.138 -0.137 0.101 -0.133 -0.140 -0.138 0.50 0.360 -0.128 -0.146 -0.140 0.360 -0.133 -0.147 -0.142 0.360 -0.137 -0.149 -0.143 0.50 0.504 -0.132 -0.150 -0.142 0.504 -0.136 -0.152 -0.144 0.504 -0.141 -0.154 -0.146 0.50 0.671 -0.135 -0.156 -0.144 0.671 -0.139 -0.158 -0.146 0.671 -0.144 -0.160 -0.149 0.50 0.886 -0.140 -0.163 -0.147 0.886 -0.143 -0.165 -0.149 0.886 -0.148 -0.168 -0.152 1.50 0.000 -0.048 -0.064 -0.064 0.000 -0.055 -0.071 -0.071 0.000 -0.063 -0.078 -0.078 1.50 0.138 -0.058 -0.074 -0.072 0.138 -0.065 -0.081 -0.079 0.138 -0.072 -0.088 -0.086 1.50 0.250 -0.064 -0.082 -0.078 0.250 -0.072 -0.089 -0.085 0.250 -0.080 -0.096 -0.092 1.50 0.477 -0.081 ' -0.098 -0.089 0.477 -0.088 -0.105 -0.096 0.477 -0.098 -0.112 -0.102 1.50 0.768 -0.100 --0.119 -0.100 0.768 -0.108 -0.126 -0.107 0.768 -0.115 -0.133 -0.114 1.50 0.919 -0.112 -0.130 -0.106 0.919 -0.120 -0.137 -0.113 0.919 -0.126 -0.144 -0.120 2.00 0.000 0.004 -0.010 -0.010 0.000 -0.009 -0.021 -0.021 0.000 -0.020 -0.031 -0.031 2.00 0.129 -0.006 -0.022 -0.020 0.129 -0.020 -0.033 -0.031 0.129 -0.032 -0.043 -0.041 2.00 0.251 -0.019 -0.034 -0.029 0.251 -0.031 -0.044 -0.040 0.251 -0.043 -0.054 -0.050 2.00 0.501 -0.042 -0.057 -0.046 0.501 -0.055 -0.067 -0.056 0.501 -0.068 -0.077 -0.066 2.00 0.758 -0.064 -0.081 -0.060 0.758 -0.079 -0.091 -0.070 0.758 -0.085 -0.100 -0.080 2.00 0.841 -0.072 -0.089 -0.064 0.841 -0.080 -0.099 -0.074 0.841 -0.100 -0.108 -0.084 2.50 0.000 0.059 0.049 0.049 0.000 0.046 0.034 0.034 0.000 0.028 0.020 0.020 2.50 0.086 0.050 0.039 0.040 0.086 0.035 0.025 0.026 0.086 0.019 0.010 0.011 2.50 0.297 0.027 0.014 0.020 0.297 0.013 0.000 0.006 0.297 -0.005 -0.013 -0.008 2.50 0.509 0.003 -0.011 0.002 0.509 -0.012 -0.024 -0.011 0.509 -0.028 -0.037 -0.025 2.50 0.713 -0.018 -0.035 -0.013 0.713 -0.034 -0.048 -0.026 0.713 -0.060 -0.039 2.50 0.889 -0.041 -0.056 -0.025 0.889 -0.055 -0.068 -0.038 0.889 -0.070 -0.080 -0.050 3.00 0.000 0.119 0.112 0.112 0.000 0.109 0.092 0.092 0.000 0.093 0.073 0.073 3.00 0.355 0.071 0.061 0.069 0.355 0.043 0.051 0.355 0.026 0.034 3.00 0.517 0.051 0.037 0.052 0.517 0.036 0.021 0.035 0.517 0.021 0.004 0.018 3.00 0.771 0.014 0.001 0.028 0.771 0.001 -0.015 0.012 0.771 -0.018 -0.030 -0.004 3.00 0.826 0.006 -0.007 0.024 0.826 -0.007 -0.022 0.008 0.826 -0.022 -0.037 -0.008 3.50 0.000 0.181 0.176 0.176 0.000 0.160 0.152 0.152 0.000 0.138 0.129 0.129 3.50 0.092 0.167 0.160 0.161 0.092 0.144 0.137 0.138 0.092 0.127 0.114 0.115 3.50 0.513 0.102 0.089 0.104 0.513 0.082 0.069 0.084 0.513 0.062 0.049 0.063 3.50 0.717 0.063 0.054 0.081 0.717 0.048 0.035 0.062 0.717 0.037 0.017 0.042 3.50 0.857 0.044 0.030 0.067 0.857 0.030 0.012 0.048 0.857 0.015 -0.005 0.029 4.00 0.000 0.252 0.243 0.243 0.000 0.231 0.215 0.215 0.000 0.208 0.186 0.186 4.00 0.509 0.158 0.143 0.159 0.509 0.133 0.119 0.135 0.509 0.109 0.095 0.111 4.00 0.617 0.141 0.121 0.144 0.617 0.112 0.098 0.120 0.617 0.092 0.076 0.097 4.00 0.783 0.106 0.089 0.123 0.783 0.087 0.067 0.100 0.783 0.063 0.046 0.078 4.00 0.857 0.093 0.074 0.114 0.857 0.068 0.053 0.092 0.857 0.033 0.070 127 Table A2.3 Continues 4.50 0.000 0.306 0.314 0.314 0.000 0.280 0.280 0.280 0.000 0.258 0.247 0.247 4.50 0.389 0.228 0.226 0.236 0.389 0.206 0.196 0.207 0.389 0.184 0.168 0.178 4.50 0.501 0.205 0.200 0.217 0.501 0.180 0.172 0.189 0.501 0.158 0.145 0.161 4.50 0.640 0.179 0.169 0.195 0.640 0.155 0.142 0.168 0.640 0.135 0.117 0.141 4.50 0.783 0.148 0.137 0.174 0.783 0.129 0.112 0.148 0.783 0.107 0.088 0.123 5.00 0.000 0.377 0.387 0.387 0.000 0.364 0.348 0.348 0.000 0.343 0.310 0.310 5.00 0.284 0.315 0.315 0.321 0.284 0.296 0.279 0.285 0.284 0.272 0.245 0.251 5.00 0.520 0.261 0.254 0.274 0.520 0.235 0.222 0.241 0.520 0.191 0.210 5.00 0.647 0.231 0.222 0.251 0.647 0.205 0.192 0.220 0.647 0.184 0.162 0.189 5.00 0.829 0.189 0.175 0.221 0.829 0.167 0.148 0.191 0.829 0.146 0.121 0.163 6.00 0.000 0.541 0.541 0.000 0.491 0.491 0.000 0.442 0.442 6.00 0.150 0.493 0.495 0.150 0.446 0.448 0.150 0.400 0.402 6.00 0.300 0.446 0.453 0.300 0.401 0.408 0.300 0.358 0.365 6.00 0.450 0.398 0.415 0.450 0.357 0.373 0.450 0.316 0.332 6.00 0.600 0.351 0.380 0.600 0.312 0.340 0.600 0.275 0.301 6.00 0.750 0.303 0.348 0.750 0.267 0.310 0.750 0.233 0.273 6.00 0.900 0.256 0.318 0.900 0.223 0.282 0.900 0.191 0.247 6.00 1.000 0.300 0.300 1.000 0.265 0.265 1.000 0.231 0.231 9.00 0.000 0.993 0.993 0.000 0.911 0.911 0.000 0.831 0.831 9.00 0.150 0.919 0.920 0.150 0.842 0.843 0.150 0.768 0.769 9.00 0.300 0.845 0.853 0.300 0.773 0.781 0.300 0.704 0.711 9.00 0.450 0.771 0.793 0.450 0.704 0.725 0.450 0.640 0.659 9.00 0.600 0.697 0.737 0.600 0.635 0.673 0.600 0.576 0.611 9.00 0.750 0.623 0.686 0.750 0.566 0.625 0.750 0.512 0.567 9.00 0.900 0.549 0.638 0.900 0.498 0.582 0.900 0.448 0.527 9.00 0.970 0.514 0.618 1.000 0.554 0.554 1.000 0.501 0.501 I total Temp 25.0 \u00C2\u00B0C Temp 35.0 \u00C2\u00B0C Temp 45.0 \u00C2\u00B0C Y MRC12 Reff [49] Calc.D Calc. J Y MgC12 Reff [43] Calc.D Calc. J V MsC12 Reff [43] Calc.D Calc. J Calc.D and Calc.J - applied approaches described in Dixon [6] and Jansz [7], Meisner [31-38] papers, respectively 3.00 0.000 0.119 0.112 0.112 0.000 0.092 0.092 0.000 0.073 0.073 3.00 0.191 0.094 0.084 0.087 0.191 0.066 0.069 0.191 0.048 0.051 3.00 0.411 0.065 0.053 0.062 0.411 0.035 0.045 0.411 0.018 0.028 3.00 0.610 0.039 0.024 0.043 0.610 0.008 0.026 0.610 -0.008 0.010 3.00 0.815 0.010 -0.005 0.025 0.815 -0.021 0.009 0.815 -0.036 -0.007 3.00 0.921 -0.004 -0.021 0.016 0.921 -0.035 0.000 0.921 -0.050 -0.015 2.00 0.000 0.005 -0.010 -0.010 0.000 -0.021 -0.021 0.000 -0.031 -0.031 2.00 0.199 -0.012 -0.029 -0.026 0.199 -0.039 -0.036 0.199 -0.050 -0.046 2.00 0.412 -0.032 -0.049 -0.040 0.412 -0.059 -0.050 0.412 -0.069 -0.060 2.00 0.616 -0.050 -0.068 -0.052 0.616 -0.078 -0.062 0.616 -0.088 -0.072 2.00 0.818 -0.068 -0.087 -0.063 0.818 -0.097 -0.073 0.818 -0.106 -0.083 2.00 0.905 -0.076 -0.095 -0.068 0.905 -0.105 -0.077 0.905 -0.114 -0.087 1.00 0.000 -0.091 -0.109 -0.109 0.000 -0.113 -0.113 0.000 -0.116 -0.116 1.00 0.193 -0.099 -0.119 -0.116 0.193 -0.123 -0.120 0.193 -0.127 -0.124 1.00 0.388 -0.108 -0.129 -0.122 0.388 -0.133 -0.126 0.388 -0.137 -0.130 1.00 0.600 -0.117 -0.140 -0.128 0.600 -0.144 -0.132 0.600 -0.148 -0.136 1.00 0.805 -0.126 -0.150 -0.133 0.805 -0.155 -0.137 0.805 -0.159 -0.142 1.00 0.904 -0.131 -0.155 -0.135 0.904 -0.160 -0.140 0.904 -0.165 -0.144 128 Table A2.4 . Calculated mean ionic activity coefficients of M g C k (y\u00C2\u00B1MgCi2) m mixture o f H C l - M g C b compared with reference values T trvrnl Temp 25.0 \u00C2\u00B0C Temp 35.0 \u00C2\u00B0C Temp 45.0 \u00C2\u00B0C X lUlai Y MgCI2 Reff [43] Calc.D Calc. J V MgC12 Reff [43] Calc.D Calc. J Y MgCL2 Reff [43] Calc.D Calc. J Calc.D and Calc.J - applied approaches described in Dixon [61 and Jansz [71, Meisner [31-381 papers, res aectively 0.10 0.000 -0.197 -0.214 -0.214 0.000 -0.202 -0.216 -0.216 0.000 -0.206 -0.217 -0.217 0.10 0.313 -0.201 -0.231 -0.217 0.155 -0.205 -0.235 -0.218 0.155 -0.209 -0.237 -0.219 0.10 0.508 -0.203 -0.229 -0.219 0.508 -0.207 -0.231 -0.221 0.313 -0.211 -0.235 -0.221 0.10 0.765 -0.205 -0.225 -0.221 0.765 -0.209 -0.228 -0.223 0.508 -0.214 -0.233 -0.223 0.10 0.870 -0.206 -0.224 -0.221 0.870 -0.210 -0.227 -0.224 0.870 -0.215 -0.229 -0.227 0.25 0.000 -0.240 -0.272 -0.272 0.000 -0.246 -0.274 -0.274 0.000 -0.251 -0.277 -0.277 0.25 0.499 -0.253 -0.294 -0.281 0.331 -0.259 -0.300 -0.282 0.331 -0.265 -0.303 -0.285 0.25 0.640 -0.257 -0.293 -0.283 0.499 -0.263 -0.298 -0.285 0.499 -0.269 -0.302 -0.289 0.25 0.872 -0.263 -0.289 -0.286 0.872 -0.268 -0.294 -0.291 0.872 -0.274 -0.299 -0.295 0.50 0.000 -0.252 -0.297 -0.297 0.000 -0.259 -0.301 -0.301 0.000 -0.266 -0.306 -0.306 0.50 0.101 -0.258 -0.333 -0.300 0.101 -0.265 -0.338 -0.306 0.101 -0.272 -0.343 -0.311 0.50 0.360 -0.272 -0.331 -0.309 0.360 -0.279 -0.337 -0.315 0.360 -0.287 -0.342 -0.321 0.50 0.504 -0.279 -0.330 -0.313 0.504 -0.287 -0.336 -0.320 0.504 -0.294 -0.342 -0.326 0.50 0.671 -0.288 -0.328 -0.318 0.671 -0.295 -0.335 -0.325 0.671 -0.303 -0.341 -0.331 0.50 0.886 -0.298 -0.327 -0.323 0.886 -0.305 -0.334 -0.330 0.886 -0.313 -0.340 -0.337 1.50 0.000 -0.169 -0.217 -0.217 0.000 -0.182 -0.231 -0.231: 0.000 -0.192 -0.245 -0.245 1.50 0.138 -0.192 -0.276 -0.232 0.138 -0.205 -0.290 -0.247 0.138 -0.215 -0.304 -0.260 1.50 0.250 -0.210 -0.280 -0.244 0.250 -0.223 -0.294 -0.258 0.250 -0.234 -0.308 -0.272 1.50 0.477 -0.245 -0.288 -0.265 0.477 -0.257 -0.302 -0.280 0.477 -0.269 -0.316 -0.293 1.50 0.768 -0.287 -0.297 -0.289 0.768 -0.299 -0.312 -0.303 0.768 -0.311 -0.325 -0.317 1.50 0.919 -0.308 -0.302 -0.300 0.919 -0.319 -0.317 -0.314 0.919 -0.332 -0.331 -0.328 2.00 0.000 -0.104 -0.144 -0.144 0.000 -0.120 -0.164 -0.164 0.000 -0.132 -0.183 -0.183 2.00 0.129 -0.132 -0.216 -0.165 0.129 -0.147 -0.234 -0.184 0.129 -0.161 -0.252 -0.202 2.00 0.251 -0.157 -0.223 -0.182 0.251 -0.173 -0.242 -0.201 0.251 -0.186 -0.260 -0.220 2.00 0.501 -0.208 -0.238 -0.215 0.501 -0.223 -0.256 -0.234 0.501 -0.237 -0.274 -0.251 2.00 0.758 -0.257 -0.254 -0.244 0.758 -0.271 -0.271 -0.262 0.758 -0.286 -0.289 -0.279 2.00 0.841 -0.272 -0.259 -0.253 0.841 -0.286 -0.276 -0.271 0.841 -0.300 -0.293 -0.288 2.50 0.000 -0.031 -0.064 -0.064 0.000 -0.050 -0.089 -0.089 0.000 -0.066 -0.114 -0.114 2.50 0.086 -0.054 -0.145 -0.082 0.086 -0.073 -0.169 -0.107 0.086 -0.089 -0.192 -0.131 2.50 0.297 -0.108 -0.164 -0.123 0.297 -0.127 -0.187 -0.146 0.297 -0.144 -0.209' -0.169 2.50 0.509 -0.161 -0.183 -0.159 0.509 -0.179 -0.205 -0.181 0.509 -0.196 -0.226 -0.203 2.50 0.713 -0.210 -0.201 -0.189 0.713 -0.226 -0.223 -0.211 0.713 -0.243 -0.243 -0.231 2.50 0.889 -0.250 -0.217 -0.213 0.889 -0.266 -0.238 -0.234 0.889 -0.282 -0.258 -0.254 3.00 0.000 0.048 0.019 0.019 0.000 0.025 -0.011 -0.011 0.000 0.005 -0.041 -0.041 3.00 0.355 -0.061 -0.107 -0.067 0.355 -0.083 -0.133 -0.094 0.355 -0.104 -0.159 -0.121 3.00 0.517 -0.109 -0.126 -0.100 0.517 -0.130 -0.152 -0.127 0.517 -0.151 -0.177 -0.152 3.00 0.77.1 -0.181 -0.157 -0.147 0.771 -0.200 -0.181 -0.172 0.771 -0.219 -0.205 -0.195 3.00 0.826 -0.196 -0.163 -0.156 0.826 -0.214 -0.187 -0.181 0.826 -0.234 -0.211 -0.204 3.50 0.000 0.131 0.106 0.106 0.000 0.104 0.069 0.069 0.000 0.079 0.033 0.033 3.50 0.092 0.097 -0.002 0.077 0.092 0.071 -0.035 0.041 0.092 0.046 -0.067 0.007 3.50 0.513 -0.049 -0.066 -0.038 0.513 -0.073 -0.096 -0.068 0.513 -0.098 -0.125 -0.098 3.50 0.717 -0.116 -0.097 -0.084 0.717 -0.138 -0.125 -0.112 0.717 -0.162 -0.152 -0.140 3.50 0.857 -0.161 -0.118 -0.113 0.857 -0.182 -0.145 -0.140 0.857 -0.203 -0.171 -0.166 4.00 0.000 0.218 0.196 0.196 0.000 0.186 0.153 0.153 0.000 0.156 0.112 0.112 4.00 0.509 0.015 -0.003 0.028 0.509 -0.013 -0.037 -0.007 0.509 -0.042 -0.069 -0.040 4.00 0.617 -0.025 -0.023 -0.002 0.617 -0.052 -0.056 -0.035 0.617 -0.080 -0.087 -0.067 4.00 0.783 -0.088 -0.054 -0.044 0.783 -0.111 -0.085 -0.075 0.783 -0.137 -0.114 -0.105 4.00 0.857 -0.114 -0.068 -0.062 0.857 -0.137 -0.098 -0.092 0.857 -0.161 -0.126 -0.121 129 Table A2.4 Continues 4.50 0.000 0.308 0.291 0.291 0.000 0.271 0.242 0.242 0.000 0.235 0.194 0.194 4.50 0.389 0.134 0.089 0.136 0.389 0.100 0.050 0.096 0.389 0.064 0.012 0.056 4.50 0.501 0.085 0.064 0.098 0.501 0.053 0.027 0.059 0.501 0.018 -0.010 0.022 4.50 0.640 0.026 0.034 0.054 0.640 -0.004 -0.002 0.018 0.640 -0.036 -0.037 -0.017 4.50 0.783 -0.034 0.002 0.012 . 0.783 -0.061 -0.032 . -0.022 0.783 -0.090 -0.064 -0.055 5.00 0.000 0.401 0.390 0.390 0.000 0.359 0.334 0.334 0.000 0.315 0.280 0.280 5.00 0.284 0.259 0.190 0.258 0.284 0.220 0.144 0.209 0.284 0.176 0.100 0.163 5.00 0.520 0.146 0.129 0.163 0.520 0.110 0.088 0.121 0.520 0.070 0.048 0.079 5.00 0.647 0.086 0.097 0.118 0.647 0.052 0.057 0.078 0.647 0.016 0.019 0.039 5.00 0.829 0.002 0.050 0.058 0.829 -0.027 0.014 0.021 0.829 -0.059 -0.021 -0.014 I total Temp 25.0 \u00C2\u00B0C Temp 35.0 \u00C2\u00B0C Temp 45.0 \u00C2\u00B0C Y MSC12 Reff [49] Calc.D Calc. J Y MRCI2 Reff [43] Calc.D Calc. J Y MgC12 Reff [43] CalcD Calc. J Calc.D and Calc.J - applied approaches described in Dixon [6] and Jansz [7], Meisner [31-38] papers, respectively 3.00 1.000 -0.050 0.019 0.019 0.000 -0.207 -0.207 0.000 -0.230 -0.230 3.00 0.809 -0.075 -0.087 -0.029 0.191 -0.186 -0.178 0.191 -0.209 -0.201 3.00 0.589 -0.111 -0.113 -0.079 0.411 -0.160 -0.140 0.411 -0.185 -0.165 3.00 0.390 -0.151 -0.137 -0.118 0.610 -0.137 -0.102 0.610 -0.163 -0.128 3.00 0.185 -0.198 -0.162 -0.155 0.815 -0.114 -0.057 0.815 -0.141 -0.085 3.00 0.079 -0.224 -0.175 -0.172 0.921 -0.102 -0.031 0.921 -0.129 -0.061 2.00 1.000 -0.169 -0.144 -0.144 0.000 -0.286 -0.286 0.000 -0.303 -0.303 2.00 0.801 -0.192 -0.220 -0.175 0.199 -0.274 -0.267 0.199 -0.291 -0.284 2.00 0.588 -0.218 -0.233 -0.204 0.412 -0.261 -0.244 0.412 -0.279 -0.261 2.00 0.384 -0.246 -0.245 -0.229 0.616 -0.249 -0.219 0.616 -0.267 -0.237 2.00 0.182 -0.275 -0.257 -0.251 0.818 -0.238 -0.192 0.818 -0.256 -0.210 2.00 0.095 -0.288 -0.263 -0.259 0.905 -0.232 -0.179 0.905 -0.251 -0.197 1.00 1.000 -0.256 -0.274 -0.274 0.000 -0.342 -0.342 0.000 -0.352 -0.352 1.00 0.807 -0.270 -0.323 -0.288 0.193 -0.340 -0.333 0.193 -0.350 -0.343 1.00 0.612 -0.283 -0.325 -0.300 0.388 -0.337 -0.323 0.388 -0.347 -0.333 1.00 0.400 -0.298 -0.327 -0.312 0.600 -0.335 -0.311 0.600 -0.344 -0.321 1.00 0.195 -0.312 -0.329 -0.322 0.805 -0.332 -0.298 0.805 -0.342 -0.307 1.00 0.096 -0.319 -0.330 -0.327 0.904 -0.331 -0.291 0.904 -0.340 -0.300 130 Table A2.5 Calculated individual ion activity coefficients for a mixture o f H C l - M g C ^ in relation to molal concentrations of each species at 60\u00C2\u00B0C m H C I m MuC12 I total Dixon Meisners, Jansz et al. n M8CI2 h H C l \u00E2\u0080\u00A2 yCl- yH+ \u00E2\u0080\u00A2 yCI- yH+ TEMPERATURE: 60\u00C2\u00B0C 8.00 0.00 8.0 0.500 9.350 4.494 1.000 2.166 11.274 1.017 2.089 11.691 5.50 0.83 8.0 0.518 9.473 4.578 0.940 1.310 5.382 0.945 1.363 7.281 3.94 1.35 8.0 0.561 9.748 4.767 0.875 1.160 4.397 0.872 1.213 5.062 2.00 2.00 8.0 0.641 10.227 5.103 0.745 1.155 2.951 0.725 1.195 3.032 1.04 2.32 8.0 0.691 10.504 5.300 0.651 1.221 2.286 0.615 1.253 2.282 10.00 0.00 10.0 0.380 8.470 3.907 1.000 3.397 30.312 1.014 3.285 31.340 5.01 1.66 10.0 0.457 9.049 4.291 0.834 1.534 9.891 0.842 1.607 11.720 4.00 2.00 10.0 0.491 9.287 4.451 0.774 1.474 8.313 0.779 1.538 9.202 1.28 2.91 10.0 0.611 10.052 4.980 0.537 1.584 4.345 0.515 1.623 4.344 12.00 0.00 12.0 0.300 7.779 3.463 1.000 5.193 76.368 1.025 4.993 79.425 10.79 0.40 12.0 0.303 7.804 3.479 0.985 3.812 30.330 0.997 3.857 64.992 7.68 1.44 12.0 0.341 8.147 3.697 0.913 2.283 26.222 0.898 2.438 36.020 6.00 2.00 12;0 0.379 8.465 3.903 0.847 1.962 21.381 0.818 2.105 25.253 3.95 2.68 12.0 0.443 , 8.950 4.225 0.728 1.821 14.922 0.680 1.942 15.609 1.37 3.54 12.0 0.550 9.678 4.719 0.495 2.011 7.832 0.411 2.113 7.603 13.00 0.00 13.0 0.280 7.589 3.344 1.000 6.272 120.043 1.032 5.992 125.646 11.20 0.60 13.0 0.286 \u00E2\u0080\u00A2 7.647 3.380 0.978 4.071 45.197 0.988 4.175 92.869 9.75 1.08 13.0 0.301 7.785 3.467 0.949 3.145 43.116 0.944 3.325 71.111 7.00 2.00 13.0 0.351 8.228 3.750 0.860 2.277 33.330 0.828 2.460 41.009 3.02 3.33 13.0 0.477 9.191 4.386 0.611 2.059 15.935 0.537 2.201 15.841 1.06 3.98 13.0 0.564 9.767 4.781 0.403 2.378 9.126 0.301 2.513 8.741 14.00 0.00 14.0 0.240 7.179 3.091 1.000 7.777 179.947 1.030 7.468 187.397 12.20 0.60 14.0 0.246 7.240 3.128 0.979 5.079 64.046 0.989 5.214 138.862 10.49 1.17 14.0 0.261 7.400 3.227 0.946 3.745 60.762 0.939 3.978 102.181 8.00 2.00 14.0 0.300 7.782 3.465 0.871 2.747 49.214 0.842 2.974 63.381 3.39 3.54 14.0 0.429 8.850 4.157 0.602 2.302 22.417 0.528 2.472 22.319 2.23 3.92 14.0 0.475 9.178 4.378 0.494 2.429 16.656 0.404 2.592 16.124 16.00 0.00 16.0 0.200 6.723 2.816 1.000 11.289 403.688 1.029 10.879 418.924 14.63 0.46 16.0 0.202- 6.752 2.834 0.986 8.158 129.545 1.000 8.274 335.395 11.25 1.58 16.0 0.228 7.051 3.013 0.923 4.539 116.538 0.908 4.897 186.773 10.00 2.00 16.0 0.245 7.231 3.123 0.888 3.879 105.456 0.862 4.221 148.622 4.60 3.80 16.0 0.366 8.354 3.831 0.617 2.820 47.928 0.545 3.063 48.427 3.20 4.27 16.0 0.412 8.722 4.073 0.503 2.915 34.797 0.414 3.150 33.788 12.90 0.00 12.9 0.280 7.589 3.344 1.000 6.171 114.535 1.023 5.939 119.011 8.33 1.52 12.9 0.322 7.979 3.590 0.887 2.624 36.847 0.892 2.776 52.293 6.00 2.30 12.9 0.375 8.429 3.880 0.783 2.154 27.497 0.779 2.280 32.143 2.95 3.32 12.9 0.478 9.201 4.393 0.567 2.082 14.956 0.542 2.174 15.195 2.25 3.55 12.9 0.508 9.403 4.530 0.500 2.155 12.455 0.467 2.242 12.433 13.20 0.00 13.2 0.280 7.589 3.344 1.000 6.477 131.815 1.016 6.261 136.347 11.98 0.41 13.2 0.281 7.603 3.353 0.984 4.772 48.870 0.989 4.843 111.724 6.00 2.40 13.2 0.373 8.416 3.872 0.794 2.185 30.332 0.758 2.353 34.620 3.73 3.16 13.2 0.444 8.959 4.230 0.645 2.090 19.693 0.592 2.229 20.124 1.89 3.77 13.2 0.518 9.467 4.574 0.472 2.271 12.292 0.399 2.396 11.980 15.20 0.00 15.2 0.200 6.723 2.816 1.000 10.011 286.480 1.021 9.728 294.815 8.00 2.40 15.2 0.292 7.701 3.414 0.784 3.088 66.858 0.791 3.280 84.448 5.83 3.12 15.2 0.351 8.228 3.750 0.662 2.717 48.498 0.663 2.863 53.900 3.30 3.97 15.2 0.442 8.946 4.222 0.463 2.735 28.544 0.451 2.844 29.012 2.18 4.34 15.2 0.492 9.293 4.455 0.350 2.939 21.075 0.328 3.039 20.949 131 Table A2.6 Calculated individual ion activity coefficients for a mixture of H C l - M g C h in relation to molal concentrations of each species at 100\u00C2\u00B0C m H C l m MgC12 I total a* Dixon Meisners, Jansz et al. n MgCI2 n H C l \u00E2\u0080\u00A2 yCl- yH+ \u00E2\u0080\u00A2 yCl- yH+ TEMPERATURE: 100\u00C2\u00B0C 8.00 0.00 8.0 0.500 9.350 4.494 1.000 1.262 4.636 1.012 1.221 4.791 5.50 0.83 8.0 0.518 9.473 4.578 0.957 0.896 2.705 0.962 0.921 3.402 3.94 1.35 8.0 0.561 9.748 4.767 0.912 0.846 2.301 0.911 0.875 2.569 2.00 2.00 8.0 0.641 10.227 5.103 0.824 0.894 1.659 0.809 0.917 1.700 1.04 2.32 8.0 0.691 10.504 5.300 0.760 0.966 1.341 0.733 0.985 1.343 10.00 0.00 10.0 0.380 8.470 3.907 1.000 1.714 10.076 1.010 1.663 10.385 5.01 1.66 10.0 0.457 9.049 4.291 0.885 1.028 4.500 0.892 1.063 5.114 4.00 2.00 10.0 0.491 9.287 4.451 0.845 1.022 3.924 0.849 1.054 4.245 1.28 2.91 10.0 0.611 10.052 4.980 0.686 1.177 2.318 0.670 1.199 2.322 12.00 0.00 12.0 0.300 7.779 3.463 1.000 2.302 20.980 1.017 2.227 21.688 10.79 0.40 12.0 0.303 7.804 3.479 0.990 1.861 11.057 0.998 1.870 18.964 7.68 1.44 12.0 0.341 8.147 3.697 0.941 1.335 9.964 0.931 1.396 12.530 6.00 2.00 12.0 0.379 8.465 3.903 0:897 1.229 8.547 0.877 1.291 9.665 3.95 2.68 12.0 0.443 8.950 4.225 0.818 1.217 6.470 0.785 1.274 6.710 1.37 3.54 12.0 0.550 9.678 4.719 0.663 1.425 3.851 0.604 1.476 3.778 13.00 0.00 13.0 0.280 7.589 3.344 1.000 2.618 30.209 1.021 2.522 31.359 11.20 0.60 13.0 0.286 7.647 3.380 0.985 1.951 15.433 0.992 1.979 25.607 9.75 1.08 13.0 0.301 7.785 3.467 0.966 1.646 14.942 0.962 1.708 21.295 7.00 2.00 13.0 0.351 8.228 3.750 0.906 1.354 12.375 0.885 1.428 14.387 3.02 3.33 13.0 0.477 9.191 4.386 0.740 1.377 6.987 0.690 1.443 6.983 1.06 3.98 13.0 0.564 9.767 4.781 0.603 1.649 4.437 0.531 1.715 4.311 14.00 0.00 14.0 0.240 7.179 3.091 1.000 3.066 41.677 1.020 2.967 43.065 12.20 0.60 14.0 0.246 7.240 3.128 0.985 2.300 20.537 0.992 2.336 35.218 10.49 1.17 14.0 0.261 7.400 3.227 0.964 1.883 19.815 0.960 1.960 28.539 8.00 2.00 14.0 0.300 7.782 3.465 0.913 1.553 17.034 0.895 1.639 20.412 3.39 3.54 14.0 0.429 8.850 4.157 0.735 1.496 9.358 0.685 1.572 9.363 2.23 3.92 14.0 0.475 9.178 4.378 0.664 1.615 7.381 0.602 1.690 7.234 16.00 0.00 16.0 0.200 6.723 2.816 1.000 4.002 80.230 1.019 3.887 82.611 14.63 0.46 16.0 0.202 6.752 2.834 0.991 3.222 37.032 1.000 3.244 71.172 11.25 1.58 16.0 0.228 7.051 3.013 0.949 2.195 34.458 0.939 2.309 47.812 10.00 2.00 16.0 0.245 7.231 3.123 0.925 1.987 32.130 0.908 2.104 40.843 4.60 3.80 16.0 0.366 8.354 3.831 0.746 1.715 17.943 0.697 1.816 18.137 3.20 4.27 16.0 0.412 8.722 4.073 0.671 1.824 13.971 0.611 1.924 13.728 12.90 0.00 12.9 0.280 7.589 3.344 1.000 2.591 29.070 1.016 2.508 30.028 8.33 1.52 12.9 0.322 7.979 3.590 0.923 1.471 13.276 0.928 1.528 17.051 6.00 2.30 12.9 0.375 8.429 3.880 0.853 1.322 10.675 0.852 1.376 11.972 2.95 3.32 12.9 0.478 9.201 4.393 0.710 1.389 6.638 0.693 1.433 6.735 2.25 3.55 12.9 0.508 9.403 4.530 0.665 1.459 5.729 0.642 1.501 5.736 13.20 0.00 13.2 0.280 7.589 3.344 1.000 2.672 32.613 1.011 2.594 33.593 11.98 0.41 13.2 0.281 7.603 3.353 0.989 2.169 16.474 0.993 2.183 29.414 6.00 2.40 13.2 0.373 8.416 3.872 0.861 1.339 11.610 0.838 1.410 12.804 3.73 3.16 13.2 0.444 8.959 4.230 0.763 1.365 8.309 0.727 1.429 8.470 1.89 3.77 13.2 0.518 9.467 4.574 0.648 1.542 5.692 0.597 1.601 5.603 15.20 0.00 15.2 0.200 6.723 2.816 1.000 3.699 60.411 1.014 3.613 61.841 8.00 2.40 15.2 0.292 7.701 3.414 0.855 1.707 22.440 0.860 1.777 26.492 5.83 3.12 15.2 0.351 8.228 3.750 0.774 1.606 17.738 0.775 1.664 19.165 3.30 3.97 15.2 0.442 8.946 4.222 0.642 1.711 11.787 0.634 1.758 11.955 2.18 4.34 15.2 0.492 9.293 4.455 0.568 1.873 9.240 0.552 1.918 9.219 132 Table A2.7 Effect of A1C1 3 , N a C l and C a C l 2 on calculated activity coefficents of M g C l 2 + H C l mixture at 25 \u00C2\u00B0C Molal concentration of each salt Activities of each salt HCl MgCl 2 NaCl A1C13 CaCl 2 1 total HCl MgCl 2 NaCl A1C13 CaCl 2 1.0 1.0 1.0 0.0 1.0 8.0 2.771 2.279 1.385 0.000 1.787 1.0 1.0 1.0 1.0 1.0 14.0 6.248 6.536 2.255 5.344 4.615 1.0 1.0 1.0 2.0 1.0 20.0 12.195 15.677 3.465 31.084 10.205 1.0 1.0 1.0 3.0 1.0 26.0 21.419 32.870 5.028 114.504 20.038 1.0 1.0 1.0 4.0 1.0 32.0 34.814 62.325 6.964 331.496 35.961 1.0 1.0 1.0 1.0 0.0 11.0 3.800 3.542 1.686 2.622 0.000 1.0 1.0 1.0 1.0 1.0 14.0 6.248 6.536 2.255 5.344 4.615 1.0 1.0 1.0 1.0 2.0 17.0 9.744 11.392 2.932 10.249 15.125 1.0 1.0 1.0 1.0 3.0 20.0 14.554 18.905 3.724 18.630 35.620 1.0 1.0 1.0 1.0 4.0 23.0 20.972 30.081 4.636 32.318 71.876 1.0 1.0 0.0 1.0 1.0 13.0 5.180 5.283 0.000 4.247 3.863 1.0 1.0 1.0 1.0 1.0 14.0 6.248 6.536 2.255 5.344 4.615 1.0 1.0 2.0 1.0 1.0 15.0 7.484 8.041 4.887 6.705 5.487 1.0 1.0 3.0 1.0 1.0 16.0 8.906 9.836 7.918 8.382 6.495 1.0 1.0 4.0 1.0 1.0 17.0 10.535 11.966 11.370 10.436 7.653 1.0 0.0 1.0 1.0 . 1.0 11.0 3.580 . 0.000 1.588 2.397 2.473 1.0 1.0 1.0 1.0 1.0 14.0 6.248 6.536 2.255 5.344 4.615 1.0 2.0 1.0 1.0 1.0 17.0 10.304 24.546 3.101 11.145 8.147 1.0 3.0 1.0 1.0 1.0 20.0 16.203 65.440 4.146 21.884 13.700 1.0 4.0 1.0 1.0 1.0 23.0 24.477 147.862 5.411 40.751 . 22.081 0.0 1.0 1.0 1.0 1.0 13.0 0.000 4.536 1.850 3.578 3.317 1.0 1.0 1.0 1.0 1.0 14.0 6.248 6.536 2.255 5.344 4.615 2.0 1.0 1.0 1.0 1.0 15.0 16.740 9.335 2.733 7.931 6.370 3.0 1.0 1.0 1.0 1.0 16.0 33.332 13.209 3.292 11.679 8.722 4.0 1.0 1.0 1.0 1.0 17.0 58.467 18.519 3.944 17.057 11.844 1.0 0.0 1.0 0.0 1.0 5.0 1.450 0.000 0.947 0.000 0.877 1.0 0.5 1.0 0.5 1.0 9.5 3.156 1.371 1.480 0.937 2.109 1.0 1.0 1.0 1.0 1.0 14.0 6.248 6.536 2.255 5.344 4.615 1.0 1.5 1.0 1.5 1.0 18.5 11.226 20.858 3.279 19.830 9.137 1.0 2.0 1.0 2.0 1.0 23.0 18.699 53.929 4.571 58.517 16.670 1.0 2.5 1.0 2.5 1.0 27.5 29.358 121.358 6.152 148.203 28.464 1.0 3.0 1.0 3.0 1.0 32.0 43.967 246.853 8.042 335.497 46.048 1.0 3.5 1.0 3.5 1.0 36.5 63.364 464.627 10.261 696.282 71.253 1.0 4.0 1.0 4.0 1.0 41.0 88.459 822.103 12.829 1347.860 106.227 1.0 0.0 1.0 1.0 0.0 8.0 2.041 0.000 1.145 1.099 0.000 1.0 0.5 1.0 1.0 0.5 11.0 3.688 1.702 1.636 2.507 1.287 1.0 1.0 1.0 1.0 1.0 14.0 6.248 6.536 2.255 5.344 4.615 1.0 1.5 1.0 1.0 1.5 17.0 10.020 17.736 3.016 10.688 11.774 1.0 2.0 1.0 1.0 2.0 20.0 15.356 40.614 3.929 20.191 25.508 1.0 2.5 1.0 1.0 2.5 23.0 22.657 83.365 5.008 36.290 49.798 1.0 3.0 1.0 1.0 3.0 26.0 32.373 158.037 6.264 62.446 90.187 1.0 3.5 1.0 1.0 3.5 29.0 45.007 281.669 7.707 103.436 154.139 1.0 4.0 1.0 1.0 4.0 32.0 61.115 477.611 9.348 165.692 251.439 1.0 0.0 0.0 0.0 0.0 1.0 0.778 0.000 0.000 0.000 0.000 1.0 1.0 0.0 1.0 0.0 10.0 3.084 2.809 0.000 2.057 0.000 1.0 2.0 0.0 2.0 0.0 19.0 10.991 28.057 0.000 27.529 0.000 1.0 2.5 0.0 2.5 0.0 23.5 18.316 67.685 0.000 75.181 0.000 1.0 3.0 0.0 3.0 0.0 28.0 28.768 145.717 0.000 181.149 0.000 1.0 3.5 0.0 3.5 0.0 32.5 43.101 287.437 0.000 396.082 0.000 1.0 4.0 0.0 4.0 0.0 37.0 62.143 528.968 0.000 801.306 0.000 133 Figure A2.1 The effects of added salts on the calculated activity coefficients a) The effect of H C l b) The effect o f A1C1 3 & C a C l 2 c) The effects M g C l 2 & C a C l 2 ( lm) 0 1 2 3 Molality of MgCl 2 & CaCl 2 - \u00E2\u0080\u00A2 - H C l -CJ-MgC12 \u00E2\u0080\u0094O-NaCl -0-A1C13 -A-CaC12 1 total d) The effects M g C l 2 & C a C l 2 (Gm) 0 1 2 3 4 Molality of MgCl 2 & A1CI3 - \u00E2\u0080\u00A2 - H C l -0 -MgC12 \u00E2\u0080\u0094O-NaCl - O \u00E2\u0080\u0094 A1C13 - A - C a C 1 2 1 total 134 Appendix 3 Experimental aspects: Procedures and methods A3.1 Determining activity coefficients of H C l in the mixture o f H C l by measuring E M F of the cell [43, 49] Reagents: Hydrochloric acid of reagent grade and M g C l 2 ( A C S certified reagent grade) Stock solutions of each electrolyte are prepared and their molalities are determined by potentiometric titration for H C l and E D T A titration for M g C l 2 . Equipment: Denver Instrument 250 (Reasearch grade p H meter pH-ISE-conductivity) Electrodes: The procedures for preparation of hydrogen electrodes and A g / A g C l electrodes are summarized in attachments A3.2-A3.3 . Measurement: Dissolved air was removed from the solution by bubbling hydrogen gas through it before the cell was filled. The cell was immersed in a constant temperature water bath regulated to 0.1 \u00C2\u00B0C with the aid of Digi-Sense ThermoLogR R T D thermometer. The electrodes were immersed and the hydrogen gas wass bubbled through the Pt bases of the Pt/H2 electrode. When the reading stabilized, the measurement was taken. The E M F measurement corresponded to a cell o f the type (A): Pt, H 2 (g, lat) | H C l (m A ), M g C l 2 (m B), in H 2 0 | A g C l , A g (A) A l l the E M F readings were corrected to a hydrogen partial pressure of la tm (use Table 9-1. Bates) Defining m A for the molality of H C l , and me for the molality of M g C l 2 , the corresponding ionic strength contribution was calculated as: y B = 3m B / (m A + 3m B ) y A = m A / ( m A + 3m B ) The E M F of cell (A) is given by: E=E\u00C2\u00B0 -0.0591 m A ( m A +2m B )y A 2 From this the y A or the activity coefficient of the hydrochloric acid in this mixture is calculated. Table A3.1 provides the standard potential of the A g / A g C l reference electrode. Table A3 .1 . Standard potentials E\u00C2\u00B0 o f the A g - A g C l electrode at temperatures [52] t ,\u00C2\u00B0C 5 10 15 20 25 30 35 40 E\u00C2\u00B0,V 0.234 0.231 0.229 0.226 0.222 0.219 0.216 0.212 t, \u00C2\u00B0C 45 50 55 60 70 80 90 95 E\u00C2\u00B0,V 0.208 0.204 0.201 0.196 0.188 0.179 0.170 0.165 135 A3.2 A g - A g C l Electrode preparation Overview The silver-silver chloride electrode is a highly reproducible and reliable electrode, and it is certainly a convenient electrode to construct and use. More notably the thermal electrolytic type is commonly used and the base of this electrode is prepared by following manner. Materials & reagents: Hel ix o f #26 Pt wire, (L=7 mm, <|>=2), flint glass tube, HNO3, H C l , AgNO 3 -400g , NaOH-1 OOg, Equipment, instrument, and glassware: crucible furnace, small electrolysis cell with low voltage supplier, and common lab glassware Procedures: The base for this electrode was prepared as follows. A helix of #26 Pt wire about 7 mm in length and 2 mm in diameter was sealed in a tube of flint glass. The bases were cleaned in warm 6 M nitric acid and a thick paste of well washed silver oxide and water was applied to each helix. The electrodes were suspended in a crucible furnace heated to about 500\u00C2\u00B0C and allowed to remain there about 10 minutes or until they were completely white. A second layer of silver was formed in a similar manner with a slightly thinner paste to make the surface smooth. The silver on each electrode weighted about 150-200 mg. Each silver electrode was mounted in a cell o f modified U-tube design and electrolyzed in a 1 M solution of twice-distilled hydrochloric acid for 45 minutes at a current of 10mA. Silver was the positive electrode, and the platinum wire served as the negative electrode. If the current efficiency was 100%, 15-20 percent of the silver would be converted to silver chloride. Thick coats of silver chloride tend to make the electrodes sluggish and should be avoided. The completed electrodes were placed in a 0.05M solution of hydrochloric acid overnight. The potentials of each electrodes were then inter-compared. Individual electrodes that differ from the average o f the group by more that 0.1 mv were rejected. The purity of hydrochloric acid and the washing o f the silver oxide are important criteria for fabricating the best electrodes. The best silver and silver chloride electrodes are light gray to white. Preparation of silver oxide Dissolve 338 g (2 moles) of silver nitrate in 3 liter of water. Dissolve slightly less that two moles (80g) of sodium hydroxide in 400 m l of water, and add the solution drop by drop to the vigorously stirred solution of silver nitrate. Silver should be present in slight excess at the end of the precipitation. The product should be washed thirty to forty times with distilled water. 136 A3.3 Hydrogen electrode preparation Overview: The hydrogen electrode is the primary reference electrode used to define an internationally accepted scale of standard potentials in aqueous solutions. This electrode commonly consists of a platinum foil, the surface of which is able to catalyze the reaction: H + + e O \u00C2\u00BB/2 H 2 The base of this electrode is prepared as follows. Materials & reagents: Sheet of Pt about 0.125 mm thick, Hel ix of #26 Pt wire, (L=2 mm, <)>=2), asbestos board, flint glass, flint glass tubing of 5 mm outside diameter, H C l , HNO3, lead acetate trihydrate, scrap platinum Equipment, instrument, and glassware: small gas-oxygen flame, Bunsen flame, and steam bath Procedure: Sheet of platinum about 0.125 mm thick was cut into pieces of about one cm 2 . A piece of Pt wire 2 cm in length was spot welded to the foil near the center o f one edge. The welding was accomplished by placing the foil, with the wire in place, on a piece of asbestos board and heating the spot to be welded with a small gas-oxygen flame. A sharp blow with a hammer joined the two pieces o f white-hot metal. The flint glass was melted over the wire and the edge o f the foil to form a bead about four mm in diameter. This bead was then sealed into the end of an 8 cm length of flint glass tubing of 5 mm outside diameter. New electrodes were cleaned before use by brief immersion in a cleaning mixture prepared by combining three volumes of 12M H C l with one volume of 16M HNO3 and four volumes of water. For best results, the base metal surface has to be smooth. For this reason, the foil is sometimes polished with emery. Platinization is the best means of activating the surface of the electrode. The platinum black is best deposited from a 1 to 3 per cent solution of chloroplatinic acid (H 2 PtC l 6 ) containing a little lead acetate (0.005% lead acetate trihydrate). A current of 200 to 400 m A was passed for 1-3 minutes in such a direction that the electrode to be coated is negative. A similar platinum foil served as the positive electrode, (j = 10-20 m A / c m 2 in 2% chloroplatinic acid and 2 M H C l ) The finished electrodes should be stored in water. Dry electrodes exposed to air lose catalytic activity, and must be replatinized before use. Preparation of Pt solution: Scrap platinum of about 1.5 g is cleaned in hot concentrated nitric acid, rinsed with water, and ignited to red heat in a Bunsen flame. The metal is cut into small fragments to facilitate solution and is digested in warm aqua regia (3 volume of concentrated H C l + 137 1 volume of concentrated HNO3) until completely dissolved. The acid platinum mixture is evaporated to dryness on a steam bath and the residue taken up in about 20 m l of concentrated hydrochloric acid. The evaporation and addition of hydrochloric acid are repeated twice. The final (fourth) evaporation should be stopped before the crystals are completely dry. The residue of chloro-platinic acid hexahydrate, H2PtCl6*6H20, remaining after the final evaporation is dissolved in 100 ml of distilled water, and 80 mg of lead acetate trihydrate is added, preferably. A3.4 Determination of magnesium ion in aqueous solution M g 2 + can be directly titrated by standard solution of E D T A . Pipette 25mL Magnesium ion solution (c.a. 0.01M) into a 250 m l conical flask and dilute to about 100 ml with de-ionized water. Adjust the p H of the solution to 10 by addition of 10ml 1 M aqueous NH4CI solution, and then concentrated NH4OH solution drop wise until the p H about 10. A d d solochrome black and titrate with standard (0.01M) E D T A solution until the color changes from Red to Blue. Prepare a 1 M solution of ammonium chloride by dissolving 26.75g of the analytical grade solid in de-ionized water making up to 500ml in a graduated flask. Use concentrated ammonia solution. Preparing indicator solution: 0.2 g dyestaff in 15 m l triethanolamine, plus 5 ml absolute ethanol. 138 A3.5a Electrode design: Base for A g / A g C l electrode I copper wire i s -1 1 L \ * /flint glass tube Pt wire A3.5b Electrode design: Base for Pt/Eb electrode copper wire to potentiometer flint glass tube flint glass bead about 4 mm in diameter Pt wire, L=20mm Pt foil (10*10,1.25 mm thick) A3.5c Electrode design: Actual design of Hildebrand bell-type Pt/H2 electrode flint glass tube copper wire to potentiometer 1 H gas outside glass compartment o o flint glass bead about 4 mm in diameter \u00C2\u00A9 \u00C2\u00A9 Pt foil (10*10,1.25 \u00C2\u00A3 mm thick) 2 2* - 6 holes in diameter 2 mm A.3.6 Example of acid determination by potentiometric titration Table A3.6 Measurements and derivatives Figure A3.6.1 Potential vs. Solution added Sol. Added E AE/AV A 2 E / A V 2 ml mV mV mL\"1 mV mL\"2 4.6 198.6 -80 4.65 194.6 -92 4.7 190 -440 -114 4.75 184.3 -640 -146 4.8 .177- -1040 -198 4.85 167.1 -2640 -330 4.9 150.6 -11920 -926 4.95 104.3 -11960 -1524 5 28.1 -31720 -3110 5.05 -127.4 52320 -494 5.1 -152.1 4240 -282 5.15 -166.2 1000 -232 5.2 -177.8 4640 Acid determination 250 200 150 100 ^ 50 S 0 -50' -100 -150 -200 ._ J b 4 6 4 / 4 8 4 9 b 1 b 2 b Solution added, ml Figure A3.6.2 Second derivatives vs. Solution volume Acid determination 60000 50000 40000 j 30000 | 20000 j l 10000 > < 0 HI . ^ -10000 -20000 -30000 -40000 \ \ \ \ \ \u00E2\u0080\u00A2 6 4 7 4 9 5 1 5 2 5 \ | \u00E2\u0080\u00A2 Solution volume, ml 142 A3.7 Water vapor pressure values for the barometric pressure correction [54] t, oc Vapor pressure mm. Hg at 0\u00C2\u00B0 0 4,6 5 6,5 10 9,2 15 12,8 20 17,5 25 23,8 30 31,8 35 42,2 40 55,3 45 71,9 50 92,5 55 118,1 60 149,4 65 189,1 70 233,7 75 289,1 80 355,2 85 433,6 90 525,9 95 634 100 760 Appendix 4 Mass balances of selected tests MASS B A L A N C E FOR TEST ID: jc5-7.2-4100 Testcond: mHCl mCuCB t,hr T,\u00C2\u00B0C (chalcopyrite mineral) 7.00 2.40 0.00 0.00 1.00 100 1 1 1 1 INPUT AS SAY, (% for solids, g/1 for solutions) MASS, grams DISTRrBUTTON S(tot) | Co Cu Fe M g Ni sol+tra H20 HCl CI (exc HCl) TOTAL S(tot) Co Cu Fe Mg N i .sol+tra H20 1 HCl Cl-Cexc HCt) TOTAL Co Cu Fe M g Ni Solid Cone. 31.20 I 0.00 35.30 33.75 0.00 0.00 -0.25 100.00 1.56 0.00 1.77 1.69 0.00 0.00 -0.01 5.00 1.00 1.00 0.00 Solution 0.00 0.00 4.00 0.00 0.00 68.76 15.59 11.65 100.00 0.00 0.00 24.22 416.81 94.53 70.63 606.20 0.00 0.00 1.00 TotlNl 1.56 0.00 1.77 I 1.69 24.22 0.00 I -0.01 416.81 94.53 70.63 611.20 OUTPUT Solid Res. 3534 0.00 34.60 32.09 0.01 0.03 -197 100.00 1.18 0.00 1.163 1.08 0.00 0.00 -0.07 0.00 0.00 0.00 336 0.74 0.73 0.00 LeachSol 0.00 0.84 48.61 0.00 0.00 162.17 32.84 0.00 0.00 0391 038 22.60 0.00 0.00 381.81 75.41 70.63 55133 0.25 036 0.96 WashSol 0.00 0.05 0.05 2.33 0.00 0.00 10.75 0.00 0.00 0.00 0.020 0.02 1.05 0.00 0.00 0.00 4.84 0.00 5.93 0.01 0.02 0.04 OUT 1.18 0.00 1.57 1.49 23.65 0.00 -0.07 381.81 80.25 70.63 560.52 | | \ S olution V olume, ml Accountability mass.g -0.38 0.00 -0.19 -0.20 -0.57 0.00 -0.05 -35.00 -14.29 0.00 -50.67 Feed Leach Wash % -24% -11% -12% -2% 430% -8% -15% 0% -8% 500.00 465.00 450.00 ! i RECOVERY,% Solid based: 34.13 36.10 So l Based: 23.34 24.15 97.65 M A S S B A L A N C E FOR TEST ID: 1B5-6.2-4-100 Test cond: mHCl mMgCU mcuCU rtiFeCB t,nr T,\u00C2\u00B0C (Low M gO c one entrate) 6.00 240 0.00 0.00 1.00 100 1 1 1 INPUT ASSAY, (\u00C2\u00B0/b for solids, g/1 for solutions) MASS,grams DISTRIBUTION S(tot) Co Cu Fe M g Ni sol+tra H20 HCl CHexc HCl) TOTAL S(tot) Co Cu | Fe Mg Ni isol+tra H20 H a Cl-(exc HCl) TOTAL Co Cu Fe M g Ni Solid Cone. 2952 0.28 032 38.49 4.55 1350 12.94 100.00 7.48 0.07 0.08 9.62 1.14 337 3.24 25.00 1.00 1.00 1.00 0.04 1.00 Solution 0.00 0.00 4.03 0.00 0.00 70.71 13.49 11.76 100.00 0.00 0.00 24.22 424.60 81.03 70.63 600.48 0.00 0.00 0.00 0.96 0.00 TotIN 7.48 0.07 0.08 9.62 25.36 337 3.24 424.60 81.03 70.63 625.48 OUTPUT Solid Res. 2650 0.38 0.22 25.22 4.94 11.61 31.13 100.00 2.28 0.03 0.019 2.17 0.43 1.00 2.68 0D0 0.00 0.00 8.62 0.46 0.33 0.21 0.02 0.29 Leach.Sol 0.08 0.08 16.08 52.09 4.86 0.00 12792 33.48 0.00 0.04 0.037 7.62 24.69 2.30 0.00 398.60 60.63 70.63 564.56 0.51 0.63 0.75 0.94 0.67 Wash.Sol 0.01 0.01 0.90 l~2.86 0 3 f 1 0.00 9.09 0.00 0.00 0.00 0.002 035 1.12 0.12 0.00 0.00 354 0.00 5.14 0.03 0.04 0.03 0.04 0.04 OUT 2.28 0.07 0.06 10.15 26.23 3.43 2.68 398 60 64.18 70.63 578.32 Solution Volume, ml Accountability: mass,g -5.20 0.00 -0.02 053 0.88 0.05 -0.55 -26.00 -16.85 0.00 -47.17 Feed Leach Wash % -69% 2% -27% 5% 3% 1% -17% -6% -21% 0% -8% 500.00 474.00 390.00 I i | RECOVERY,% Solid based: 52.86 76J00 7741 62.56 7034 Sol. Based: 55.25 49.02 82.87 102 71.84 MASS BALANCE FOR TEST ID: 1B5-8.2-2 Test cond: K>HC1 ttlMgCB mcucc t,hr T,\u00C2\u00B0C 1 (Low MgO concentrate) 8.00 2.40 0.00 0.00 2.00 100 | 1 I | INPUT ASSAY, (% for solids, g/1 for solutions) MASS, grams DISTRIBUTION S(tot) Co Cu Fe M g Ni isol+tra H20 HCl Cl-Cexc HCl) TOTAL SCtot) Co Cu Fe I Mg Ni sol-Kta H20 HCl Cl-(exc HCl) TOTAL Co Cu Fe M g Ni Solid Cone. 2952 0.28 032 38.49 455 13.50 1 12.94 ! 100.00 7.48 0.07 0.08 9.62 | 1.14 337 3.24 25.00 1.00 1.00 1.00 0.04 1.00 Solution 0.00 0.00 3.96 0.00 I 0.00 66.84 17.66 11.54 100.00 0.00 0.00 I 24.22 409.02 108.04 70.63 61151 0.00 0.00 0.00 0.96 0.00 |TotIN 7.48 0.07 1 0.08 9.62 ! 25.36 337 3.24 409.02 108.04 70.63 63651 OUTPUT Solid Res. 2538 0.24 0.28 22.29 653 4.12 40.76 100.00 1.62 0.02 0.018 1.43 0.44 026 2.61 0.00 0.00 0.00 6.40 0.22 031 0.14 0.02 0.07 Leach.Sol 0.11 0.08 17.81 51.40 6.82 0.00 156.65 33.20 0.00 0.05 0.038 837 24.16 321 0.00 379.02 73.63 70.63 559.10 0.73 0.64 0.82 0.93 0.88 WashSol 0.01 0.01 1.15 3.11 0.43 0.00 11.25 0.00 0.00 0.00 0.003 0.46 1.24 0.17 0.00 0.00 4.50 0.00 638 0.04 0.05 0.04 0.05 0.05 OUT 1.62 I 0.07 0.06 1025 | 25.85 3.64 2.61 379.02 78.13 70.63 571.88 1 j \ | Solution Volume, ml I Accountability: mass.g -5.86 0.00 -0.02 0.63 0.49 027 -0.63 -30.00 -2951 0.00 -o\"5.03 Feed Leach Wash % -78% 0% -27% 7% 2% 8% -19% -7% -28% 0% -10% 500.00 470.00 400.00 1 RECOVERY,% Solid based: 7754 77153 85.18 61 DO 92.19 Sol. Based: 77.56 5053 91.73 100.19 100.11 MASS BALANCE FOR TEST ID: 3B5-6.2-1-100 Test cond: \"ma mntgCE mcucu mjFeCB t,hr T,\u00C2\u00B0C (high MgO concentrate) 6.00 2.40 0.00 0.00 1.00 100 T 1 INPUT ASSAY, ( for solids, g/1 for solutions) MASS,grams DISTRIBUTION S(tot) Co Cu Fe M g Ni Insl+tr H20 HCl Cl-Cexc HCl) TOTAL S(tot) Co Cu | Fe Mg Ni InsWr H20 HCl ClHC1 raKjCQ mcaCU my,cB t,hr T,\u00C2\u00B0C (high M gO c one entrate) 6.00 2.40 0.00 0.00 1.00 100 ! | INPUT ASSAY, (% for solids, g/1 for solutions) MASS, grams DISTRIBUTION S(tot) Co Cu I Fe M g Ni isoHtia H20 ! HCl Cl-(exc HCl) TOTAL S(tot) Co Cu Fe Mg Ni soB-tra H20 : HCl Cl "Thesis/Dissertation"@en . "2006-11"@en . "10.14288/1.0079242"@en . "eng"@en . "Materials Engineering"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "The use of strong brine and HCl solutions to process nickel sulfide concentrates"@en . "Text"@en . "http://hdl.handle.net/2429/18061"@en .