Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The removal of cobalt from zinc sulphate electrolytes using the copper-antimoney process Lew, Richard W. 1994

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

Item Metadata

Download

Media
831-ubc_1994-0133.pdf [ 8.66MB ]
Metadata
JSON: 831-1.0078522.json
JSON-LD: 831-1.0078522-ld.json
RDF/XML (Pretty): 831-1.0078522-rdf.xml
RDF/JSON: 831-1.0078522-rdf.json
Turtle: 831-1.0078522-turtle.txt
N-Triples: 831-1.0078522-rdf-ntriples.txt
Original Record: 831-1.0078522-source.json
Full Text
831-1.0078522-fulltext.txt
Citation
831-1.0078522.ris

Full Text

THE REMOVAL OF COBALT FROM ZINC SULPHATE ELECTROLYTES USING THE COPPER-ANTIMONY PROCESS by Richard W. Lew B.Sc, The University of British Columbia, 1985 B.A.SC., The University of British Columbia, 1990 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Metals and Materials Engineering We accept this thesis as conforming t© the required standard THE UNIVERSITY OF BRITISH COLUMBIA January 1994 © Richard W. Lew, 1994 in presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allov^ed without my written permission. Department of fe ^S'fcAO^Ig ^7)A Y}) t^ ' PgPT OF M E T A ^ S f f^/^efiALS t W e , The University of British Columbia Vancouver, Canada Date DE-6 (2/88) ABSTRACT At Cominco's zinc refinery located at Trail, British Columbia, impurities are removed fi^om zinc sulphate electrolyte by a two stage purification process using zinc dust. Copper, cadmium and thallium are removed in the first stage whereas cobalt is removed in the second stage. Cobalt removal kinetics are affected by soluble antimony and copper additions to the electrolyte. This study was focused on the kinetics of cobalt removal by zinc powder cementation. Experimental variables included copper and antimony concentrations, pH and temperature. Common organic additives employed in zinc plants were studied to determine potential interactions with the cobalt removal process. Cementation products obtained from laboratory experiments and plant practice were studied using the SEM. A number of significant findings were observed. 1) Cobalt removal was extremely slow in the absence of any activators in solution. Batch testing have shown that antimony activation performed better than copper activation, but the addition of both copper and antimony together increased the rate of cobalt removal significantly. Copper and antimony ions were essentially removed from solution during the early stages of the cementation reaction (within the first 15 minutes), presumably, establishing the cathode surface for cobalt removal. 2) The cobalt removal rate constant increased with increasing temperature, and the highest rate was obtained at 90°C (the highest temperature tested). An apparent energy of activation of 86.6 kJ/mol was calculated for this system, and this value was consistent with activation controlled processes. 3) pH significantly affected the cobalt cementation rate constant. Too low a pH (<4.0) increases the rate of the competing hydrogen reduction reaction and too high a pH (>4.0) can result in passivation of the residue's surface. In both these extreme ranges, cobalt removal was detrimentally affected. 4) A drop in the cobalt removal rates can be correlated with a rise in the E^ of the slurry. Lignin sulphonate, Percol 351 and animal glue at levels of 2 mg/L significantly impede cobalt cementation. 5) SEM analyses of the cementation residues have shown that the activation agents and the cobalt impurity was collected onto an impurity shell surrounding the dissolving zinc particle. This shell consisted mainly of Zn and minor constituents of Co, Cu and Sb. pagei TABLE OF CONTENTS 1. INTRODUCTION 1 2. LITERATURE REVIEW 3 2.1. Hydrometallurgical Zinc Extraction (overview) 3 2.2. Impurities in Zinc Electrolyte 7 2.3. Commercial Zinc Dust Purification Processes 10 2.4. Review: Cobalt Cementation 12 2.4.1. Chemical Properities of Zinc 13 2.4.2. Thermodynamic Analysis of Cobalt Cementation 13 2.4.3. Cobalt Cementation Rate Control 19 2.5. Unactivated Cobalt Cementation 23 2.6. Activation Agents 25 2.7. Mixed Potential for the Cementation Reaction 27 2.8. Hydrogen Evolution 30 2.9. Substrates for Cobalt Deposition (the Sb-Cu Activation System) 33 2.10. Organics 36 2.11. Conclusions 37 3. EXPERIMENTAL SET-UP AND CONDITIONS 38 4. RESULTS AND DISCUSSIONS 41 4.1. The Standard Experiment 41 4.2. Effect of Activation Agents 42 Antimony Activation 42 Copper Activation 44 Copper and Antimony (Cu-Sb) Activation 45 4.2.1. Extended Time Experiments 47 4.2.2. The Effect of Copper Concentration on Cu-Sb Activation 48 4.2.3. Effect of Antimony Concentration on Cu-Sb Activation 50 4.2.4. Effect ofHighSb and Cu Concentrations on Cu-Sb Activation.. 52 4.3. Effect of Pre-Coating Zinc Powders 56 4.4. Effect of Initial Cobalt Concentration.... 58 4.5. Effect of pH 60 page u 4.6. Effect of Temperature 62 4.7. Effect of Zinc Dust Loading 66 4.8. Cementation in an Initially Zinc-Free Solution 69 4.9. Effects of Organic Additives 70 4.10. Compositional and Morphological Analysis of Cementation Residues 73 Surface Morphology 73 Unactivated Cobalt Cementation using Zinc Dust 74 Antimony Activation 75 Copper Activation 76 Cu-Sb Activation 77 Cu-Sb Activation under High Copper Concentrations 78 Cross-Sectional Analysis 79 5. CONCLUSIONS 83 7. RECOMMENDATIONS 86 6. REFERENCES 87 APPENDIX A - Cobah Cementation Rate Calculations 92 APPENDIX B - Zinc Dust Surface Area Estimation 93 APPENDIX C - Criss and Cobble Method for Estimating Activity of Ionic Species at Elevated Temperatures 95 APPENDIX D - Listing of Thermodynamic Data Used to Generate Ej^ -pH Diagrams 97 APPENDIX E - X-ray Analysis and Resolution 98 page ill LIST OF TABLES Table 2-1 Impuritiy Levels in Various Zinc Plant Electrolyte 9 Table 2-2 Electrochemical Potentials at Standard and at Plant Conditions 14 Table 2-3 Activation Energies and Kinetic Rate Constants of Cobalt Cemetation 20 Table 2-4 Hydrogen Overpotentials in the Presence of IM H2SO4 at 25°C 32 Table 4-1 Summary of the Effects of Cu-Sb Activation Agents 55 Table 4-2 Effect of pH on Cementation Kinetics.. 61 Table 4-3 Effect of Temperature on Cobalt Removal 64 Table 4-4 Effect of Zinc Dust Loading on the Cobalt Cementation Rate 67 Table 4-5 Summary of the Surface Area Analysis by BET 67 Table 4-6 Effect of Organic Additives 73 Table 4-7 Summary of WDX Analysis of Cementation Residues 80 page IV LIST OF FIGURES Figure 2-1 Flowsheet of Cominco's Zinc Process 4 Figure 2-2 Effect of Impurities on Current Efficiency 8 Figure 2-3 Synergistic Effects of Arsenic and Cobalt Impurities on Current Efficiency 8 Figure 2-4 Flowsheet of Cominco's Zinc Dust Purification Circuit 11 Figure 2-5 ^-pH Diagrams for the Zinc-Water System a) Unit Activities at 73°C, and b) Plant Conditions at 73°C 16 Figure 2-6 E -^pH Diagram for Zinc-Water System at 73°C under Plant Conditions. [the ZnO species was not considered for the creation of this diagram]... 17 Figure 2-7 E -^pH Diagram of Zinc-Water System at 73°C, and under Plant Conditions. Basic zinc sulphate stability region was created using thermodynamic data fi"om Fountoulakis 18 Figure 2-8 Mass Transfer Coefficient versus Relative Power 22 Figure 2-9 Zinc Ion Inhibition of Cobalt Cementation 23 Figure 2-10 Schematic Diagram of Reactions Taking Place During Cobalt Cementation 26 Figure 2-11 Evans Diagram for Activation Controlled Processes. 28 Figure 2-12 Evans Diagram Displaying the Effect of Cathodic and Anodic Inhibition 29 Figure 2-13 Relationship between Metal-Hydrogen Bond Strengths and the Exchange Current Density for Hydrogen Evolution 31 Figure 2-14 Evans Diagram for Codeposition of Copper and Zinc to Form Brass 33 Figure 3-1 Schematic Diagram of the Experimental Setup 39 Figure 4-1 Cobalt Removal Profile for the Standard Experiment 41 Figure 4-2 Cupric Specific Ion Electrode Potential Profile for the Standard Experiment 41 Figure 4-3 Slurry Potential Profile for Standard Experiment 42 Figure 4-4 pH Profile for Standard Experiment 42 Figure 4-5 Effect of Antimony Activation on pH 43 page V Figure 4-6 Effect of Copper Activation on pH 44 Figure 4-7 Effect of Copper Activation on Cupric Specific Potential 44 Figure 4-8 Copper, Antimony and Cobalt Removal versus Time 45 Figure 4-9 Eflfect of the Copper and Antimony Activators on Cobalt Removal with Zinc Powder 46 Figure 4-10 Cobalt Removal for the Extended Time Experiment 47 Figure 4-11 Slurry Potential (vs. Ag/AgCl) Profile for the Extended Time Experiment 48 Figure 4-12 pH Profile for the Extended Time Experiment 48 Figure 4-13 Effect of Copper Concentration on the Cobalt Removal Rate in the Presence of 1.5 mg/L Antimony. 49 Figure 4-14 Effect of Copper Concentration on Cobalt Removal after 90 Minutes... 49 Figure 4-15 Effect of Copper Concentration on the pH Profile 50 Figure 4-16 Eflfect of Antimony Concentration on the Cobalt Removal Rate in the Presence of Copper 51 Figure 4-17 Effect of Antimony Concentration on Cobalt Removal in the Presence of Copper after 90 Minutes 51 Figure 4-18 Slurry Potential Profile for the Copper Series 52 Figure 4-19 pH Profile for the Copper Series 52 Figure 4-20 Eflfect of Copper Concentration on the Cobalt Removal Rate in the Presence of 3 mg/L Antimony 53 Figure 4-21 Effect of Copper Concentration on the Slurry Potential Profile in the Presence of 3 mg/L Antimony 53 Figure 4-22 Cupric Potential Profile for the Copper Series in the Presence of 3 mg/L Antimony... 54 Figure 4-23 pH Profile for the Copper Series in the Presence of 3 mg/L Antimony... 54 Figure 4-24 Eflfect of Pre-Coating Zinc Powder with 1.5 mg/L Sb and 46 mg/L Cu.. 56 Figure 4-25 Effect of Pre-Coating Zinc Powder with 1.5 mg/L Sb and 92 mg/L Cu.. 57 page VI Figure 4-26 Effect of Initial Cobalt Concentration on Cobalt Removal 58 Figure 4-27 Effect of Initial Cobalt Concentration on the pH Profile 58 Figure 4-28 Effect of Initial Cobalt Concentration on Cobalt Removal at 90 Minutes 59 Figure 4-29 Effect of Controlled pH on Slurry Potential 61 Figure 4-3 0 Efifect of Controlled pH on the Hydrogen Addition Rate 61 Figure 4-31 Effect of Controlled pH on Cobalt Removal 62 Figure 4-32 Effect of Controlled pH on Cobalt Concentration at 90 Minutes 62 Figure 4-33 Efifect of Temperature on the Cobalt Removal Rate 63 Figure 4-34 Efifect of Temperature on Cobalt Removal at 90 Minutes 65 Figure 4-3 5 Effect of Temperature on the Slurry Potential Profile 65 Figure 4-36 Arrhenius Plot for Cobalt Cementation using the Cu-Sb Activation System 65 Figure 4-37 Effect of Zinc Dust Loading on Cobalt Removal 66 Figure 4-38 The Efifect of Zinc Dust Loading on the Cobalt Removal Rate Constant 66 Figure 4-39 The Efifect of Zinc Dust Loading on the Cupric Potential Profile 68 Figure 4-40 The Efifect of Zinc Dust Loading on the Slurry Potential Profile 68 Figure 4-41 Comparison Between Cobalt Removal Profiles in Zinc Electrolyte and in an Initially Zinc-Free Solution 69 Figure 4-42 pH Profile for a Cementation Experiment using Initially Zinc-Free Solution 70 Figure 4-43 Slurry Potential Profile for a Cementation Experiment using Imtially Zinc-Free Solution 70 Figure 4-44 The Effect of Animal Glue on Cobalt Cementation 71 page vii Figure 4-45 The Effect of Percol 351 on Cobalt Cementation 72 Figure 4-46 The Effect of Lignin Sulphonic Acid on Cobalt Cementation 72 Figure 4-47 Surface Morphology of an Unactivated Zinc Dust Particle after 90 Minutes 74 Figure 4-48 Surface Morphology of an Antimony-Activated Zinc Dust Particle. (a) 5 minutes (b) 60 minutes 75 Figure 4-49 Surface Morphology of a Copper-Activated Zinc Dust Particle after 60 Minutes 76 Figure 4-50 Surface Morphology of a Sb-Cu Activated Zinc Dust Particle after 90 Minutes 77 Figure 4-51 Surface Morphology of a Sb-Cu Activated (High Copper) Zinc Dust Particleafter 30 Minutes 78 Figure 4-52 SEM Micrograph of Cross-sectioned Cementation Residue 80 Figure 4-53a Positioning of Compositional Line Scan Through a Cementation Residue 80 Figure 4-53b Compositional Line Scan for Cobalt in the Cementation Residue 82 Figure 4-53c Compositional Line Scan for Zinc in the Cementation Residue 82 Figure 4-53d Compositional Line Scan for Copper in the Cementation Residue 82 Figure 4-53e Compositional Line Scan for Antimony in the Cementation Residue 82 page viu LIST OF SYMBOLS Electrochemical Symbols: A aM2+ 3-M,alloy ba be E° ^ cell F I lo i io J K n r R T Tla Tic surface area of the electrode, m^  activity of metal ion species activity of metal in the alloy Tafel slope for the anodic reaction Tafel slope for the cathodic reaction standard potential, V cell potential under standard conditions, V Faraday constant, 96500 C/g-equiv current, A exchange current, A current density, A/m^ exchange current density, A/m^ flux, mol/m^ equilibrium constant number of electrons exchanged in the reaction radius of a particle, m gas constant, 8.314 J/mol-K temperature, °K anodic overvoltage, V cathodic overvoltage, V page IX Symbols used for the Cobalt Cementation Rate Calculations: A active surface area, m^ C(0) concentration at time zero, mol/cm^ C(t) concentration at time t, mol/cm^ k rate constant, m/sec kco cobalt removal rate constant, m/sec t time, sec V volume, m^ Symbols used for Criss and Cobble Calculations: S ^ entropy at Tj, cal Sxi entropy at Tj, cal AS° entropy change, cal Cp heat capacity aj intercept from plot of entropy at 298°K vs. entropy at an elevated temperature, Tj bx slope from plot of entropy at 298°K vs. entropy at an elevated temperature, T2 a values calculated from ax, cal/deg mol P values calculated from hj, cal/deg mol T temperature, °K pagex ACKNOWLEDGMENTS I am most appreciative to Cominco Ltd. for their invaluable technical assistance and anal5^ical support during the course of this project and for their permission to publish the results fi-om this research. This project was jointly funded by Cominco Ltd., and the National Science and Engineering Research Council of Canada. I wish to thank Mr. Peter Mclver for allowing me to perform the experimental portion of my thesis at Cominco Research (located at Trail, B.C.), and to Dr. Alberto Gonzalez-Dominguez for his insightful discussions, ideas and encouragement during this project. The technical staff at Cominco Research and the Assay Office also deserves credit for all their help, but I am especially indebted to Dilip Makwana and Chuck Maklon. Other people who deserve recognition include; my supervisor, Dr. David Dreisinger (for his guidance, his organisational abilities and advice). Dr. Ernie Peters (for his encouragement, and his entertaining and thought provoking inquiries into the meaning of life, and hydrometallurgy), and to my eclectic office-mates Paul West-Sells, Ben Saito and George Owusu. Of course, the office staff of the Metals and Materials Engineering department who are often forgotten but who makes everything behind the scene run smoothly. I am especially indebted to Joan Kitchen, and Carol Duerden for all their help over the years and whose efforts are so often taken for granted. Finally, but not least, I would also like to thank my parents and family for their support during this study. page XI 1 INTRODUCTION One of the most successful applications of hydrometallurgy is in the extraction and refining of zinc. Prior to the adoption of a hydrometallurgical process for zinc, the majority of the world's zinc production was based on the pyrometallurgical vertical retort technology. In this process, zinc oxide is reduced by carbon at a temperature above the boiling point of zinc (907°C). The metal vapour released consists of the pure metal which is quickly condensed and collected as liquid zinc [1]. This was an unhygienic and labour intensive operation, and the development of new technology and the increasing demand for zinc in the chemical and coating industries forced this process to be largely displaced by the middle of the 20th century. Over 80% of the world's zinc is currently produced by the Roast-Leach-Electrowinning (RLE) Process. Some of the advantages of the RLE process over pyrometallurgical alternatives are lower capital and energy costs, higher zinc recovery, and a higher purity product. A disadvantage of the RLE process is that sulphur dioxide is generated during roasting which is a necessary operation to produce a leachable calcine. The commercial-scale RLE process [2] was pioneered jointly by the Anaconda Company of the USA, by the Consolidated Mining and Smelting Company of Canada (later to be known as Cominco) and by the Electrolytic Zinc Company of Australasia. The first RLE plants (whose technology is based on a sulphate electrolyte) opened in 1915 in Anaconda (USA) and Trail (Canada). At that time, these plants produced 25 and 50 tonnes/day of zinc, respectively. Today, Cominco's zinc plant located at Trail, B.C., produces 272,000 tonnes per year of special high grade (>99.99%) zinc. The electrowinning of zinc requires solutions of extremely high purity. An impure electrolyte will decrease zinc electrowinning current efficiency and can decrease the quality of the final zinc product. Therefore, solution purification is a critical operation. Purification is a complex process that takes place in a number of steps. The bulk of the impurities in the electrolyte (Fe, Hg, Ga, Pb, In, Sb, Sn, Bi, Ba and Ag) are removed through calcine leaching and iron hydrolysis. This is followed by several zinc dust cementation stages in which most of the remaining metal impurities (Co, Cu, Cd, Ge, Ni, Ar, Tl and Sb) are removed. page 1 Of the impurities found in zinc electrolyte, cobalt can be particularly harmful to zinc electrowinning. Its presence can decrease the current efficiency and cause redissolution of the zinc towards the end of the electrowinning cycle. The difficulty in achieving cobalt removal is highlighted by the need for special purification steps. In many hydrometallurgical zinc plants in the world, cobalt is removed from the electrolyte during what is called "hot stage purification" using zinc powder. This step can take place between 72 and 95°C, using an excess amount (above the stoichiometric requirement) of zinc powder and additions of antimony and copper activating agents to catalyze cobalt removal. The role played by copper and antimony is poorly understood. This project was performed in collaboration with Cominco to examine cobalt purification from zinc electrolyte solution. The anticipated incorporation of a high cobalt-containing concentrate (Red Dog, Alaska) is a major concern to Cominco's zinc operations because it will result in higher cobalt concentrations in the electrolyte. The objective of this study was to advance the understanding of the kinetics of cobalt removal under conditions simulating the hot stage purification used at Cominco. The results of this study are expected to help to decrease zinc dust consumption and improve cobah removal rates. The presentation of this work is divided into 4 major sections. Chapter 2 deals with a review of the literature, providing a summary of the modem ideas behind cobalt removal by cementation. Chapter 3 describes the experimental procedure used in this study and the experimental parameters tested, while Chapter 4 provides a review and discussion of the experimental results. Chapter 5 presents the summary of this work. page 2 2 LITERATURE REVIEW The RLE process was originally conceived by Leon Letrange and was described in a German patent issued in 1881 [2]. The use of zinc dust purification was introduced in the early 1890's, though no industrial scale plants became operational until 1915. The RLE process is practised in Canada by Cominco Ltd. at Trail, B.C., by Hudson Bay Mining and Smelting in Flin Flon, Manitoba, by Falconbridge Kidd Creek at Timmins, Ontario, and by Canadian Electrolytic Zinc at Valleyfield, P.Q. The literature review begins with an overview of the industrial zinc process, covering the RLE and the relatively new zinc pressure leaching processes and the zinc dust purification circuit used at Cominco. This is followed by a review of the thermodynamics and kinetics of cobalt cementation. 2.1 Hydrometallurgica! Zinc Extraction (overview) Hydrometallurgical zinc extraction typically involves leaching the zinc calcines, purifying the leachate to remove any undesired impurities, and electrowinning to obtain pure zinc. However, there are many different routes to leaching and purification depending on the feed material and the desired quality of the final product(s). Figure 2-1 is a simplified overview of the hydrometallurgical route for zinc as practised at Cominco. Cominco practises two methods for leaching the concentrate metal values: (1) the RLE process and (2) the zinc pressure leach (ZPL) process. In the RLE process, zinc is leached in a zinc-containing sulphuric acid solution returned fi^om electrowinning (return electrolyte). However, because zinc concentrates containing ZnS minerals leach poorly in sulphuric acid at low temperatures, the ore is first roasted to an oxide to produce a leachable calcine according to equation (2.1.1). The roasting process takes place in air, at a temperature between 800 to 975°C, producing a calcine and sulphur dioxide gas. Sulphur dioxide is either recovered as liquid SOj or used to make sulphuric acid. ZnS + 3/2 02~> ZnO + S02(g) (roasting) (2.1.1) page 3 Roaster Calcine Sulphide Leach |» Plant (SLP) T Neutral Thickener Overflow (NTO) L/S ZnS cone (Red Dog) Oxide Fume from Smelter Return Acid Zinc Pressure Leach 3t: Na2C03 Leach (Halide Removal) Oxide Leach Plant PBJ ZnS04 Fe3+ Elemental Sulphur Zn dust Purification TMf Ge, In Plant Ge, In • t Lead Smelter PURMCATION cold-hot-polish Cadmium Plant Glue Addition S/L Gypsum Removal Stage ELECTROWINNING PLANT ZincMetal Retum Acid Zinc Dust Cu&Sb Additions Melt Shop Figure 2-1. Flowsheet of Cominco's Zinc Process page 4 The calcine is readily leached in sulphuric acid according to equation (2.1.2). ZnO + H2SO4 - > ZnS04 + H^O (leaching, RLE) (2.1.2) Certain impurities such as silver, bismuth, lead and barium form insoluble precipitates which remain in the leach residue. An alternative leaching technology, pioneered by Sherritt Gordon Ltd and Cominco, is the zinc pressure leach (ZPL) process [3]. This direct leach is conducted at 130 to 155°C using pressurized oxygen. Zinc is dissolved in this process and elemental sulphur is produced as a by-product (eq. 2. L 3 a): ZnS + I/2O2 +H2SO4 - > ZnSO^ + S° + H^O (overall leaching, ZPL) (2. L3a) ZnS leaching is actually catalyzed by soluble Fe^ ^ (eq. 2.L3b): 2FeS04 + H2SO4 + 1/2 O2 - > Fe2(S04)3 + H^O (2.1.3b) Fe2(S04)3 + ZnS - > 2FeS04 + ZnSO^ + S° (2.1.3c) At Cominco, the resulting combined electrolyte from the calcine leach and the ZPL is purified to remove impurities prior to zinc electrowinning. The impure electrolyte is treated by two purification steps: (1) coprecipitation of impurities with the precipitation of iron hydroxide and (2) cementation by zinc dust. Iron, mercury, gallium, indium, lead, antimony and tin impurities are removed by coprecipitation as hydroxides when the pH is adjusted to between 4.5 and 5.1 by return acid and calcine addition, as illustrated by reaction (eq. 2.1.4). 2Fe'^ + 3ZnO + SH^O - > 2Fe(0H)^ (ppt) + 3Zn'^ (2.1.4) Copper, nickel, cobalt, aluminum, arsenic and germanium are only partially co-precipitated [4] or adsorbed with iron, while cadmium, magnesium and chloride remain in solution. Significant amounts of impurities are still present in the electrolyte and flirther treatment is needed before zinc can be efficiently electrowon from solution. At Cominco, iron-free solution is referred to as "thickener overflow". This iron-free solution still contains relatively high levels of cobalt, copper, cadmium, germanium, nickel, arsenic, thallium page 5 and antimony. These impurities are subsequently removed by zinc dust purification which can be generally described by the metal displacement reaction (eq. 2.1.5). M^^  + Zn°-> M° + Zn^ ^ (2.1.5) Copper, cadmium, germanium, arsenic and nickel are easily cemented onto zinc dust, while cobalt requires the addition of "activators", either antimony, antimony-copper or arsenic-copper, for successful cementation. Zinc is recovered from the purified electrolyte by electrowinning, according to the overall electrochemical reaction (eq. 2.1.6). ZnSO^ + HjO ~> Zn + H2SO4 + I/2O2 (overall) (2.1.6) The electrowinning process requires a very pure zinc sulphate electrolyte to achieve acceptable current efficiencies and a high purity metal product. The electro won zinc is stripped, melted and cast into bars or ingots for shipment to market. The sulphuric acid generated (referred to as "return acid") is recycled to the leaching step. page 6 2.2 Impurities in Zinc Electrolyte Electrolj^e purification is a key operation for the economical extraction of zinc by the hydrometallurgical route. Impurities in the electrolyte can lead to: • a decrease in the current efficiency during electrowinning [5,6] • formation of dendrites on the deposited zinc, leading to short circuits and stripping difficulties • a decrease in the quality and purity of the final zinc product In the production of zinc, low levels of cobalt, copper, antimony, iron, cadmium, arsenic, germanium, tin, selenium, tellurium, silver, bismuth, and barium in the electrolyte can affect the quality of the zinc product and the current efficiency during zinc electrowinning [7,8]. These impurities may co-deposit with the zinc and become incorporated into the cathodic zinc, or can lower the hydrogen overpotential. Ohoyama and Morioka [9] ranked the effects of impurities by the amount of H2 evolved on the zinc cathode. Cobalt was considered to be one of the more detrimental impurities in the zinc electrolyte: bad impurities > worst impurities [Ga, Bi, Tl, Cd, Hg, In, Pb]->[As, Ni, Sn, Co, Fe, Ag]~> [Ge, Sb, Te, Se] At cobalt levels greater than 0.5 mg/L, cobalt alone affects zinc electrowinning current efficiency (Figure 2-2). However, the presence of cobalt can also enhance the effects of other impurities on current efficiency. For example, even though iron alone has an insignificant effect on the current efficiency, its combination with cobalt and arsenic was found to decrease dramatically the zinc electrowinning current efficiency [10] (Figure 2-3). page 7 100 90 I c 80 -70 10 I • rv . . -^m— Co As Fe ' 1 = 2 b a ^ • O • i I ^ H \V v. N! 1 f 1 • • - • % ^ \ O 1 1 -3 10-2 10" ' 10** 10' lO ' Concenlrotion (mg/L) 10^ Figure 2-2. Effect of Impurities on Current Efficiency. Conditions: Neutral leach solution containing 60 g/L Zn^ "", 200 g/L H2SO4 5 mg/L Dowfroth, Img/L Saponin, 0.015 mg/L Sb and 13 mg/L Na2Si03 500A/m^ 38°C, and 24 hour Zn deposition time [10] >-o z U J u. 11. UJ 1 -z UJ 0: oc —1 u 40 ?o As (mg/L) ""0 0.01 • O.Ol A 1.0 _ A 1.0 O 5.0 • 5.0 Co (mg/L) a i5 5.0 0.15 1.0 0.15 5.0 ft to 20 30 IRON (mg/L) Figure 2-3. Synergistic Effects of Arsenic and Cobalt Impurities on Current EflBciency. Conditions: Neutral leach solution containing 60 g/L Zn^ "^ , 200 g/L H2SO4, 5 mg/L Dowfroth, 1 mg/L Saponin, 0.015 mg/L Sb and 13 mg/L NajSiOj 500 A/m^ 38°C, and 24 hour Zn deposition time [10] page 8 These observations have been confirmed in industry. It is standard tankhouse practice to control cobalt (<0.5 mg/L) and arsenic (<0.5 mg/L) levels in the electrolyte, while iron concentrations are usually kept to below 5 mg/L. The tankhouse has a limited capability to offset some of these impurities with the addition of organics to the electrolyte. At Kidd Creek, for example, a mix of 5 mg/L Dowfroth, 1 mg/L Saponin, 0.015 mg/L Sb and 13 mg/L Na2Si03 was found to be an optimal combination to control the harmfial effects of electrolyte impurities, to produce smooth deposits and control acid misting. Table 2-1 [11,12,13,14] illustrates the range of impurities found in various zinc plant electrolytes. Even though the solutions are relatively free of impurities, the range in impurity concentrations is typical of the solution purity different plants require. Table 2-1. Impurity Levels in Various Zinc Plant Electrolytes mg/L Cobalt Copper Cadmium Iron Antimony Nickel Arsenic Thallium COMINCO [11] 0.3 0.2 0.3 <5.0 0.03 <0.05 -0.100 VMBALEN [13] 0.23 0.2 5 15 0.01 0.01 0.01 0.5-5 NATIONAL zmc [12] 0.7 <0.1 2.3 <0.1 0.01 0.3 <0.01 17.0 KIDD CREEK [14] 0.3 0.1 <0.1 5 <0.02 <0.10 0.02 -Very low cobalt levels are needed for Cominco's 72-hour electrowinning cycle, otherwise zinc redissolution takes place at significant rates towards the end of the cycle. In the presence of elements such as germanium and higher acid concentrations, the cobalt tolerance limit is reduced to 0.2 mg/L. With the recent incorporation of the high cobalt Alaskan Red Dog concentrate as its feed material, Cominco must deal with higher cobalt levels in its process streams. page 9 2.3 Commercial Zinc Dust Purification Processes In the hydrometallurgical zinc extraction process, zinc dust purification is the last opportunity to remove any residual impurities from the electrolyte prior to electrowinning. Relative to other impurities such as copper, antimony and thallium, cobalt is particularly difficult to remove from zinc electrolyte solutions by zinc dust purification. It requires high temperatures (>72°C), activating agents and a large excess of zinc dust. In the continous zinc dust purification process used by Cominco, (shown in Figure 2-4) impurity removal takes place over two cementation stages. The first stage (referred to as the "cold" stage), operates at 50 to 60°C, and removes copper, thallium, nickel and cadmium. The subsequent stage (referred to as the "hot stage"), primarily removes cobalt. This stage operates at 70 to 80°C, and at a pH of about 4.2. Additions of copper as copper sulphate in the range of 20 to 50 mg/L and antimony, as potassium antimony tartrate, in the range of 1 to 3 mg/L are necessary to promote the removal of cobalt. A third stage, known as the "polish" stage, is only occasionally used to remove any residual impurities in the electrolyte. One of the unique features of the Cominco purification circuit is the recycling of the cementation residues. Recycling is designed to reduce zinc dust consumption by separating and recycling the coarse zinc-containing particles to the hot stage purification. The main disadvantage is more complicated process control, since variables such as cementation residue size distribution and slurry aeration (a cause of cobalt redissolution) must also be considered. The solutions from zinc dust purification are filtered, gypsum is removed and the electrolyte is then forwarded to the tankhouse for the final metal recovery step. At Cominco, cementation residues are collected and treated to recover metal values [15]. Residues from the cold stage purification are sent to the cadmium plant for the recovery of the metal values, while residues from the hot stage purification are leached to recover zinc. Remaining metals, including cobalt, are sent to the smelter to be disposed of in the slag [16]. Operational experience [17] has led to the recognition of several key operating parameters required for the efficient operation of the "hot" stage zinc dust purification. Temperature was found to be important in controlling basic zinc sulphate formation, whereas pH control was necessary to minimize acid additions and prevent redissolution of the cementation residues. page 10 In summary, zinc dust purification is operationally relatively simple and the capital costs are comparatively low. The major process operating costs are the the energy required to operate at the relatively high temperatures and the amount of zinc dust consumed. With an annual production of 272,000 tonnes of zinc per year at Cominco, a reduction in the percentage of zinc dust required for purification is economically significant. Neutral Thickener Overflow Return Heat Exchanger HZ Zinc Dust COLD STAGE PURIFICATION Spiral Heat Exchanger Sperry Filters Cadmium Plant Storage Tank Return Acid Cu«b addition •gyctone DISTRIBUTOR TANK Zinc Dust Cu/Sb Addition HOT STAGE PURIFICATION recycle cyclone overflow u Return Acid& Zinc Dust y us Filters underflow purge overflow n Y Zinc Dust un<)efflgw POLISH STAGE PURIFICATION (2 TANKS) Cu/Sb Addition (optional) _t_ Polish Stage Purification (1 tank) Zinc Dust Return Pad • ' r HOT STAGE RESIDUE TANK RELEACH Return Add I Smelter [• Shriver Filters j Gyspum Removal Zinc Electrowinning Tankhouse NTO TANK Figure 2-4. Flowsheet of Cominco's Zinc Dust Purification Circuit. page 11 2.4 Review: Cobalt Cementation Cementation is used extensively for purifying process streams [18], such as the purification of zinc electrolyte, and in the recovery of metal values from waste solutions. Although the zinc dust purification process is industrially very important, fundamental understanding of the process is limited. Much of what is known of the industrial process was derived empirically through operating experience with full- and pilot-sized plants. One such discovery was that cobah cementation with zinc dust proceeded at acceptable rates only when the zinc powder was "activated" by the presence of antimony and copper or by arsenic and copper. During the course of cementation, many changes can take place on the zinc surface making the process quite complicated. Occurring on the electroactive surfaces are various competing reduction reactions, localized changes to the surface conditions, variations in the surface area, and compositional and reactivity changes on the anodic and cathodic surfaces. To better understand these complex phenomena, the thermodynamics, electrochemistry and mechanism by which cementation takes place must first be reviewed. page 12 2.4.1 Chemical Properties of Zinc Because of zinc's favourable redox characteristics, it is used extensively in the cementation of such metals as cadmium, copper and cobalt. In aqueous systems, ionization is favoured further by a high exothermic energy of ion hydration. This favourable chemistry makes the use of zinc powders an ideal choice as a reducing agent for cementation reactions. 2.4.2 TJiermodynamic Analysis of Cobalt Cementation According to the electrochemical series, zinc provides a strong thermodynamic driving force for the cementation of more noble metals such as copper and cobalt. The greater the potential difference, the greater the thermodynamic driving force for the reaction. Co'^ + 2e- = Co, E° = -280 mV -(Zn'^ + 2e- = Zn), E° = -(762.8) mV Co'^ + Zn"" ==> Co^ + Zn'^ ,E°,,i, =+482.8 mV (2.6.1) K = e x p ( ^ ] (2.6.2) Thermodynamic calculations (equations 2.6.1 and 2.6.2), under standard conditions, predicted that cobalt cementation is a favourable reaction [19,20] with an equilibrium constant, K, of 2x10'^. This shows that there is a strong thermodynamic driving force for the reduction of cobalt with zinc with negligible back reaction. Other species which can also be reduced by zinc include hydrogen, copper and antimony, and, compared to cobalt, an even stronger thermodynamic driving force for the reduction of these species exists. page 13 Most electrochemical series are based on standard conditions of 25°C and unit activity, but industrial processes operate under quite different conditions. At Cominco, the hot stage purification is conducted at a temperature of 73°C, a pH of 4.2, and with an electrolyte consisting of Zn 150 g/L Cu 33 mg/L Co 20 mg/L Sb 1.5 mg/L SO4'- 170 g/L The potentials listed in the Table 2-2 includes both standard state values, and corrected potentials for industrial conditions. Adjustment of the values under plant conditions was made for temperature using the Criss and Cobble method, and for concentration by the Nemst equation. As illustrated in Table 2-2, different conditions can change the thermodynamics of the system, with the hydrogen reduction reaction showing the largest change. As the pH of the solution increases, the hydrogen reaction becomes less noble, bringing it close to that of cobalt reduction. Table 2-2. Electrochemical Potentials at Standard and at Plant Conditions. Reduction Reactions Cu*" + 2e = Cu SbO* + 2 i r + 3e = Sb + H^O 2YL+ 2e=H,(g) Co^ + 2e- = Co Zn^ + 2e-= Zn' mV(SHE)at 25''C, and unit activity 340.2 204 0.00 -280 -762.8 mV(SHE) atJS'Cand unit activity 371.3 199 (50''C) 35 -247.7 -725.2 mV(SHE)at JS'Cand plant conditions 258.6 -115(50''C) -253 -366.8 -712.8 When zinc powder is contacted with an acidic zinc sulphate solution, it begins to dissolve and supplies electrons for a number of thermodynamically favourable reduction reactions. Some reactions are beneficial to the cementation of cobalt such as the deposition of activation agents page 14 onto the surface of the zinc powder which catalyze the reduction of cobalt. In contrast, other reactions are detrimental to cobalt cementation (for example, hydrogen reduction). Not only does hydrogen evolution compete with cobalt and addition agents for the available electrons, but it can cause localized surface changes in pH on the zinc powder surface which can result in the formation of precipitation products which dramatically impede the reaction rate. The Ejj-pH diagram is a useful tool for predicting the thermodynamically stable species present under a range of pH and electrode potential (E J conditions. One of the prime disadvantages of this diagram is that it fails to account for the kinetics of the reaction. An example of this is the kinetically slow hydrogen reaction on the surface of zinc during electrowinning. Researchers [20,22,46] have found that the formation of zinc oxides, hydroxides or salts on the surface of the zinc particle act to inhibit the cementation reaction by slowing the transportation of the cobalt ion to and the diffusion of zinc ions away from the zinc particle surface. The formation of these films or salts is enhanced with increasing ionic strengths and pH. It is not well established which film or salt actually forms on the surface of the particle during cementation, since all analysis to date has been performed on a dried specimen either by X-ray diffraction or by microanalysis under the scanning electron microscope. Figures 2-5 to 2-6 are examples of Ej^ -pH diagrams for the zinc-water system under standard and plant conditions. Under standard conditions, ZnO and Zn(0H)2 are the stable species when the pH approaches alkalinity, and this trend persists under plant conditions. However, under high ionic strength conditions in the presence of sulphate, a basic zinc sulphate salt (ZnS04*3Zn(0H)2 •4H2O) was found to be thermodynamically stable (Figure 2-7) in the pH range used under zinc dust purification [46]. Unfortunately, the thermodynamic data for this species are not well established. page 15 Eh (Volts) 2.0 Molality m mol /kg S l.OOOE+00 Zn l.OOOE+00 Pressure p bar S l.OOOE+00 Zn l.OOOE+00 Eh (Volts) 2.0 1.5 h 1.0 0.5 0.0 -0.5 h -1.0 -1.5 -2.0 1 \ 1 : — Zn(+2a) " ^ * • * • • -• 1 1 • • • ' 1 • — 1 1 1 1 1 ZnO * ' ' • — - , — "- . ^ ^ ^ ^ ^ ^ ^ — ^ ^ ^ " Zn • : 1 1 1 . 1 1 • 1 1 1 ' 1 1 Molality m mol / kg S 3.000E+00 Zn 1.500E+00 10 Pressure p bar S l.OOOE+00 Zn l.OOOE+OO 12 14 pH Figure 2-5. Ej^ -pH Diagrams for Zinc-Water System a) Unit Activities at 73°C, and b) Plant Conditions at 73°C page 16 Eh (Volts) 2.0 Molality m mol/kg S 3.000E+00 Zn 1.500E+00 10 Pressure p bar S l.OOOE+00 Zn l.OOOE+OO 12 14 PH Figure 2-6. E^-pH Diagram for Zinc-Water System at 73°C under Plant Conditions, [the ZnO species was not considered for the creation of this diagram] page 17 Eh (Voltj) 1.5 1.0 0.5 0.0 -05 -1.0 -1.5 ..._ ^.., 1 1 Zn(+2«) T • - • 1 — 1 '1 *** '"•"*^ " Zn(OH;2*ZnS04 ' ^ • " ' ^ ' ^ o * ^ ^ , ^ I ' 1 " I \ T ZnO ZD 1 1 1 1 1 1 1 1 1 1 1 1 1 -1 10 12 14 pH Molality m mol/kg S 3.000E+00 Zn 1.500E+00 Pressure p bar S l.OOOE+00 Zn l.OOOE+00 Figure 2-7. E^-pH Diagram of Zinc-Sulphate-Water System at 73°C, and under Plant Conditions. Basic zinc sulphate stability region was created using thermodynamic data from Fountoulakis[46]. (Note: the pH range for Zn(0H)2 • ZnSO^ is wider than commonly observed.) page 18 2.4.3 Cobalt Cementation Rate Control Cementation reactions can be broken down into a number of chemical and mass transfer steps: 1) Transport of the electroactive species fi-om the bulk solution to the surface. 2) Electron transfer from the dissolving metal species to the electroactive species. 3) Incorporation of the deposited metal into the crystal lattice. 4) Desorption of the metal ion. 5) Transport of the dissolved metal ion from the surface to the bulk solution. A number of researchers [20, 22, 33] have observed that although a majority of cementation reactions are transport controlled, cobalt cementation with zinc powder is under either mixed or surface/chemical control (see Table 2-3). Since cobalt removal is improved by the addition of certain activation agents while maintaining other conditions constant, it is possible that the surface reaction step of the system is changed. If the rate-controlling step is either step (1) or (5), then the reaction is under mass transfer control. Steps (2) to (4) encompass the surface activation controlling reactions. Researchers have found that cementation reactions usually follow a first order rate law dictated by equation 2.7.1 regardless of whether they are mass transfer controlled or surface activation controlled (Appendix A). page 19 Table 2-3: Activation Energies and Kinetic Rate Constants of Cobalt Cementation [21,22,23] Activation System TeO activation As activation lOmg/L As, 25mg/L Cu lOmg/L As, lOOmg/L Cu lOmg/L Sb, 25 -100 mg/LCu Sb Activation Sb Activation l.Omg/L Sb, 405mg/LCd, 3.8mg/LNi l.Omg/L Sb, 405mg/LCd, 3.8mg/L Ni Img/LSb Img/LSb Img/LSb T(°C) 95 90 90 90 85 75 75 75 85 95 Reference Lawson & Nhan [19] Lawson & Nhan [19] Tozawa [48] Tozawa [48] Tozawa [48] Kerby [33] Kerby [33] Adams & Chapman [22] Adams & Chapman [22] Dodson [23] Dodson [23] Dodson [23] Cathodic Substrate Zn Rotating Disc Zn Rotating Disc Zn Rotating Disc Zn Rotating Disc Zn Rotating Disc Mn Powder, 2g/L Zn Powder, 2g/L Zn Powder, 2g/L Zn Powder, 4g/L Zn Powder, 4g/L Zn Powder, 4g/L Zn Powder, 4g/L Activation Energy (kJ/moIe) 29-35 40-60 n/a n/a 53 18-62 n/a n/a n/a 115 115 115 Cobalt Removal Rate Constant, k X10-* (sec>) n/a 0.25 -1.0 X 10" (m/s) 150 300 67-83 1517 417 600 1150 78 225 438 The advantages of using a rotating disc are that it provides a uniformly accessible surface for cementation reactions, and that its mass transfer behaviour is well defined. Usually an enhanced kinetic region can be seen in rotating disc studies since mass transfer is affected by boundary layer disturbances such as morphological changes on the rotating surface. The formation of dendritic deposits roughens the reaction surface and improves mass transfer between the bulk solution and the deposit-solution interface. These surface effects are amplified with increasing temperature as changes in liquid density and viscosity will also occur. For a rotating disc set-up, an apparent activation energy of < 12.5 kJ/mole for a laminar flow regime and < 25.1 kJ/mole for a turbulent flow regime is considered to be mass transfer controlled, while values greater than these are either mixed or surface activation controlled. These systems are important in research studies as the boundary layer can be well controlled, and an accurate determination of the mass transfer coefficient can be calculated. page 20 While rotating disc and cylinder electrodes are useful tools for research purposes, industrial scale cementation processes are performed in stirred tank reactors with particulate metal powders in solution, and, therefore, suspended particle studies are more realistic commercially. Theoretical treatment of mass transfer to suspended particles in a stirred reactor has not yet been fiiUy developed due to the complicated solution flows surrounding the particle, non-uniformity of the particle shape and the difficulty of measuring local boundary conditions surrounding each agitated particle. A rudimentary mass transfer coefficient can be calculated for suspended particles from the Harriott [24] equation and from application of Stokes Law [25]. Complications may also arise in cementation systems associated with surface deposit formation, and in estimating an average available surface area for the zinc dust. A close physical analogy to cementation is the agitation of solids during leaching. Surface activated cementation processes probably also share the characteristics that solids "off-bottom" suspension is probably more important than agitation. Increasing agitation would not be of any benefit to increasing the reaction rate or conserving power for agitation (Figure 2-8). page 21 40 -8 2} CO mCO a:< 20 10 OFF BOTTOM SUSPENSION POINT 10 20 40 100 RELATIVE POWER 200 400 Figure 2-8. Mass Transfer Coefficient versus Relative Power [31]. A range of kinetic rates can be obtained using zinc dust (Table 2-3). These differences in kinetics can be attributed to zinc dust sizing, electrolyte composition, concentration of activators used, shape of the dust, zinc dust surface area, and the alloy content of the zinc dust [26]. page 22 2.5 Unactivated Cobalt Cementation The initial step in understanding the activation process for cobalt removal must involve an understanding of why the cobalt cementation rate is slow without activation. Figure 2-9 illustrates the inhibition of cobalt cementation with increasing zinc ion concentration in solution. The inhibition of cobalt precipitation on zinc dust is believed to arise from the precipitation of ZnO or Zn(0H)2 by the combined effect of adsorption of zinc ions in the double layer, and a localized increase in pH near the surface arising from hydrogen gas evolution on the cathode surface. 0.5 Time (min) ^°'^ Figure 2-9. Zinc Ion Inhibition of Cobalt Cementation [20 ] Higher zinc ion concentrations are expected in the double layer than in the bulk since the mass transfer coefficient of zinc ions is seven times smaller than that for hydrogen ions. The resulting charge imbalance caused by the rapid hydrogen ion arrival to the double layer and the slow departure of the zinc ion to the bulk solution result in a locally charged boundary layer. This charge imbalance will be equilibrated by the attraction of anions, possibly sulphate ions, and the repulsion of cations, such as H^. page 23 The discharge of hydrogen gas on the zinc surface creates a localized increased pH in the double layer and an increase in the hydroxide ion concentration according to equation (2.8.1). 2H2O + Zn° --> HjCg) + Zn'^ + 20H- (2.8.1) When the conditions of a sufficiently high local zinc ion concentration and high pH are satisfied by the dissolution of zinc, the formation of a zinc hydroxide or zinc oxide film can take place on the surface of the zinc powder. This film effectively decreases the available cathode surface area on which cobalt cementation can take place. Presumably, a surface or double layer adsorption mechanism is involved, and experimental observations showing that cobalt deposition can be inhibited by high zinc ion solution concentrations support this conclusion. Tozawa found that for zinc concentrations greater than 50g/L, cobalt removal was limited to only a few percent even after 3 hours. Lawson and Nhan [20], Yunus et al [27], and Xiong and Ritchie [28] have recorded similiar observations. page 24 2.6 Activation Agents In zinc dust purification, cobalt removal proceeds slowly unless it is catalysed by "activating agents". Activation using copper, antimony or arsenic alone does not improve cobalt cementation as significantly as having both antimony and copper or arsenic and copper present together. Activation agents are believed to act by cementing on the surface of the zinc powder to provide a more active cathode surface on which cobalt deposition can take place [29]. In addition, these substrates must act in some way to inhibit the formation of zinc hydroxide, zinc oxide or basic zinc sulphate before reasonable cobalt cementation kinetics can be attained. Figure 2-10 is a schematic diagram illustrating the primary electrochemical reduction reactions that take place on the shell of the activated zinc powder during cobalt cementation. These reactions can be either beneficial (such as cobalt deposition and deposition of activating agents), or detrimental (such as hydrogen evolution) to cementation. The degree to which this surface controls the detrimental hydrogen evolution reaction and improves the activity of cobalt deposition is a measure of its effectiveness. In summary, activation agents for cobalt cementation can improve cementation by acting to: • control or degrade the formation of any passivating film or salts on the zinc powder surface. • increase the cathodic surface area. • increase the hydrogen overpotential on zinc and on cobalt rich areas. • form intermetallic compounds or alloys that have a high hydrogen overpotential and a low overpotential for cobalt deposition. • form stable deposition products that are resistant to redissolution. page 25 Impurity Shell Co 2+ step 2 Activation Agents Figure 2-10. Schematic Diagram of Reactions Taking Place During Cobalt Cementation. page 26 2.7 Mixed Potential for the Cementation Reaction Cementation reactions can be thought to behave analogously to a short-circuited or galvanically coupled corrosion cell. Unlike electrolysis reactions where there is external control of the potential of the reaction, cementation systems have a mixed potential whose value depends on the combined rate of the anodic and cathodic reactions. This concept is summarized in the mixed potential theory, which states that the net current for the cathodic half-cell must be equal to the net current of the anodic half-cell. The mixed potential is not an equilibrium potential. It is a steady state potential whose magnitude arises fi-om the relative activation energy barriers for the cathodic (reduction) and anodic (oxidation) half-cell reactions. Assuming that there are insignificant internal and ohmic resistances, each half-cell (cathodic and anodic) sees the same mixed potential. The mixed potential may change during the course of the cementation reaction as the cathodic and anodic areas and compositions, and localized surface conditions vary with time. Given the fact that cobalt cementation is chemically or activation controlled, a relationship can be established between the mixed potential and the exchange current density. For small overvoltage values, the mixed potential can be described by the Tafel slopes, and by using the definition iA = I, the following relationships can be derived. Tlc=-bclogi^=-bclog-^ (2.7.1) lo io.c Tla=bal0gi^ = b a l 0 g i - (2.7.2) lo Ao,a where 110= the cathodic overvoltage, and Tia= the anodic overvoltage. The Evans diagram graphically illustrates this relationship (Figure 2-11). page 27 Cathodic potential '••-• c a> • * — • o a. Anodic potential exchange current density, io, for cathodic reaction cathodic overpotential anodic overpotential cementation potential exchange current density, io, for anodic reaction logi cementation current, I cementation Figure 2-11. Evans Diagram for Activation Controlled Processes. AtE cementation ' Icem i a — i c — icementation iA = I and since T|a — licementation ^ i ~ U a ' O g ' j . Tjc ~ •t'cementation l^c — DclOg j By combining the terms, the equation below can be derived: , ^ (Ec - E a ) + (balOg Io,a + b g l o g Ip.c) log!cementation — / L L \ (ba+bc) (2.7.3) (2.7.4) (2.7.5a) (2.7.5b) (2.7.6) This equation summarizes the important variables involved in activation controlled cementation. and that the value of I^ ^^ j^^ jj^ ^ can be increased by: increasing the potential difference increasing the anodic or cathodic exchange current density, io. increasing the total area of the anode or the cathodes [30]. page 28 One way to increase the potential difference between the anodic and cathodic reactions can be through the use of a stronger reducing agent. Manganese powder is more reducing than zinc, and it has been tested for cementation of cobalt and cadmium systems [31, 33]. It was found that during cobalt cementation, the removal rate for cobalt using manganese powder was improved by a factor of 3 compared to using similarly sized zinc powder. However, 50 times the stoichiometric requirement for manganese powder was needed, and the deposited impurities had a strong tendency to redissolve [32]. Manganese powder is not being used industrially for zinc electrolyte purification. Cathodic Inhibition * decreasing I , cementation * decreasing E,^ ^^^^^^ Anodic Inhibition * decreasing I .cementation * increasing E,^ ^^^ ,^„^ Figure 2-12. Evans Diagram Displaying the Effect of Cathodic and Anodic Inhibition. The total cementation current can be decreased by reducing the anodic or the cathodic exchange current, or both. The anodic and cathodic exchange currents can be decreased by reducing the exchange current density for the overall reaction, and by varying the areas of the anodic and cathodic sites. The Evans diagram (Figure 2-12) illustrates that by blocking the anodic or the cathodic sites with either a passivating film (such as, zinc hydroxide or basic zinc sulphate salts) or the adsorption of an organic inhibitor will decrease the cementation rate. Cathodic site inhibition will result in a decrease in the mixed potential, while an anodic inhibition will increase it. page 29 Similarly, the formation of a non-porous shell around the cementation residue can inhibit the cementation when the concentration of the dissolution products becomes high. Ingraham and Kerby [33] found such a mechanism taking place for cadmium cementation with zinc dust, and that Cd and Cd3Cu alloys were the primary contributors to the formation of a non-porous impurity shell. In contrast, the deposition of noble metals in the form of high surface area structures such as dendrites will increase the cementation reaction rate provided that the anodic reaction is not rate limiting. 2.8 Hydrogen Evolution Hydrogen evolution is the primary parasitic reaction of many electrodeposition reactions including cobalt cementation. This reaction takes place on the surface of the cathode substrate and different substrates have varying abilities to catalyse it. Certain substrates, such as zinc and antimony, possess a relatively high hydrogen overpotential. The hydrogen overpotential can be thought of as an additional potential or energy barrier, over that of the standard reversible potential, needed to evolve hydrogen at an appreciable rate. In other words, it is a measure of the difficulty in causing the hydrogen reduction reaction (2.8.1) to proceed in the forward direction [34]. 2ir + 2e- = H2(g) (2.8.1) Equation (2.8.1) only depicts an overall reaction, while the mechanism for this reaction taking place on the surface of the substrate is complicated. It is commonly believed that the formation of an adsorbed species (equation (2.8.2)) on the substrate surface is the first step, followed by two possible reactions to form hydrogen gas. H^ + e- = H^ (2.8.2) The two mechanisms that are generally believed to occur on the surface of the substrate are illustrated below by equations (2.8.3) and (2.8.4): 1) i r + e- + M~>M-H followed by 2M-H—>2M + H2 (2.8.3) and/or 2) H^ + e + M--> M-H followed by M-H + lT+ 6—>M + H2 (2.8.4) page 30 where M is a metal atom in the metal lattice. In both mechanisms, hydrogen is adsorbed on the surface, forms hydrogen-metal bonds, and cleavage of these bonds results in the formation of hydrogen gas. Therefore, if the cathodic surface is changed so that an increase in the free energy of adsorption favours the formation of the adsorbed species, it will have the opposite effect in the subsequent step in the overall reaction. As a result, there is a dependency between the hydrogen evolution exchange current density and the strength of the metal-hydrogen bond formation. The triangular-shaped plot (Figure 2-13) shows clearly that the maximum rate of hydrogen evolution occurs at an intermediate value of AG^. The strength of this M-H bond dictates whether the surface will have a high or a low hydrogen overpotential [35]. 10*'-10'*-10-^ -10-'-TlV A u / •In 1 10 APt • Ri Mo A \ T I Nb»^«To 1 20 M-H bond strength/kj mol"* Figure 2-13. Relationship between Metal-Hydrogen Bond Strengths and the Exchange Current Density for Hydrogen Evolution. Table 2-4 summarizes measured hydrogen overpotentials, and the exchange current densities for hydrogen evolution on the various substrates of interest. A higher hydrogen overpotential represents increased difficulty in evolving hydrogen on these substrates. The results in the second column are a collection of data from cyclic voltammetry experiments. They show that certain page 31 substrates (such as zinc) do not adsorb hydrogen atoms as well as platinum, copper or cobalt. These substrates require a higher current density (indicative of a higher hydrogen overpotential, ) to produce significant hydrogen evolution. From the group of metals Zn, Sb, Cu and Co, Zn possessed the highest hydrogen overpotential, followed by Sb and Cu, while Co had the lowest ^^wahie. Table 2-4: Hydrogen Overpotentials in the Presence oflMH2S04at25°C[36] Species Pt Co Cu Sb As Zn Hydrogen Overpotential (V) 0.000002 0.101* 0.190 0.233 0.369 0.415* * calculated values based on data presented by Fletcher [35] Antimony metal surfaces possess a higher hydrogen overpotential than either copper or cobalt. It has been shown through galvanically-coupled experiments, that increasing the antimony content in copper and zinc-alloyed cathodes decreased the rate of hydrogen evolution [37], and that this correlated with an improvement in the cobalt removal rate from solution. This strongly suggests that antimony enhances cobalt removal by providing cathodic sites with high hydrogen overpotentials. The behaviour of antimony, however, is difficult to understand. During cobalt cementation using zinc dust, an antimony metal surface behaves differently from antimony ions in solution. Aitkenhead [30] showed that cobalt can be removed using zinc dust from zinc electrolyte solutions using powdered antimony (-325 mesh) metal as an activator. He claimed to have achieved almost 100% cobalt removal at 90°C after 2 hours with an initial cobalt and zinc concentration of lOmg/L and 160g/L, respectively. In this system, the presence of copper ions in the solution was found to affect detrimentally cobalt removal. In contrast, other researchers have page 32 found that zinc dust activation with copper and antimony ions in solution, (antimony being introduced into solution as a potassium antimony tartrate salt), requires the presence of copper ions in order to attain acceptable cobah removal rates. 2.9 Substrates for Cobalt Deposition (the Sb-Cu Activation System) In the investigations of cementation mechanisms using a rotating disc and a cylinder, there is a growing body of experimental evidence which suggests that the deposited metals may actually form alloys, and that these act as the "activated" cathodic sites for cobalt deposition. It is already well established that hydrogen overpotential plays an important role in determining the suitability of a cathode substrate for cementation reactions. However, important consideration should be given to the dissolution stability of the deposit and the ability of the surface to provide a strong thermod5maniic driving force for cobalt deposition. Thermodynamically, alloying benefits deposition reactions. This can be explained using the Nemst equation (eq. 2.9.1) and the Evans diagram (Figure 2-14): E = E° + ^ ln RT ' M 2 + 2F "^ aj^o^alloy (2.9.1) c a. -m Figure 2-14. Evans Diagram for Codeposition of Copper and Zinc to Form Brass. The activity of a metal in an alloy is less than that of the pure metal which normally has an activity equal to one. Equation 2.9.1 predicts that there will be a higher electrochemical potential for page 33 metal deposition when an alloy phase is present compared to a pure metal due to the reduction in metal activities. This translates into a stronger thermodynamic driving force for metal deposition. Fisher-Bartelk et al [38] suggested that the thermodynamic driving force for cobah deposition was due to the formation of a cobalt-zinc alloy. Alloys have a more positive equilibrium potential, as a result of a lower activity of the alloy compared to the pure metal phase. They found that cobalt deposition occurred only in association with an alloy phase. It was postulated that the formation of strong cobah alloys, such as cobalt and zinc, imparts a more positive potential to the cementation reaction. They identified alpha-brass and gamma-cobalt-zinc phases in the cementation products by radiographic analysis. Four Co-Zn alloy phases were identified, Tl (0.006%Co), a(6.5-8.2%Co), Y(10-11%CO), and 5(13.5-23%Co). Results from kinetic studies suggest that zinc deposited with some copper presents a low overpotential surface for cobalt deposition, and that antimony and arsenic somehow activate the deposition process. Straumanis et al [39] showed that cementation deposits of copper contained co-deposited zinc, and suggested that these deposits form small cathodes on the zinc surface. This occurs by underpotential or anomolous co-deposition [40]. deBlander [41] performed work on ascertaining the role of copper and antimony on cobalt cementation. Electrodeposits were obtained at a controlled potential of -730 mV (SHE) from a zinc sulphate solution containing 10 ppm Co, 10 ppm Sb and 12 ppm Cu, using flat electrodes of Pt and Cu. It was postulated that the deposits consisted of a Cu-Sb-Co alloy of variable composition and that because the Co-Cu deposit contained more cobalt than a Sb-Co deposit, copper accelerated the cementation process. Antimony was considered to act as a stabiliser of the cementation product because dissolution of cobalt from the Cu-Sb-Co deposit was much slower than from the Cu-Co deposit. He proposed the following cathodic reaction (Equation 2.9.2). Co'^ + HSbOj + Cu'^ + 3 i r + 7e- = Cu-Co-Sb (alloy) + 2H2O (2.9.2) However, examination of the cementation products by X-ray diffraction and microscopic analysis failed to identify alloys of Cu-Sb-Co. Fontana and Winand [42] have shown that the addition of antimony was necessary to cement out cobalt from zinc-rich solutions, and identified CoSb and CoSb^ compounds in the cementation page 34 products by X-ray diffraction. Using single sweep voltammetry Coubot [43] showed that the potential difference between the cobalt-zinc couple increases with temperature and with the addition of antimony. He concluded that this provided a higher driving force for cobalt deposition. Kroleva [44] found that Cu-Sb activation enhances cobalt cementation on zinc dust, while it had the detrimental effect of depolarising the hydrogen and oxygen evolution reactions. He proposed the following mechanism. AtT=5(yCandpH>3: Zn + (CujSb) = Zn'^ + (Cu^Sb)'" (2.9.3) Co'"- + (Cu^Sbf = Co + (CujSb) (2.9.4) AtT=SCFCandpH<3: 2H" + (CujSb)'- = H^ + (Cu^Sb) (2.9.5) 2H^ + 1/2 O2 + (Cu^Sb)'- = H2O (Cu^Sb) (2.9.6) Gravel [45] found that an electrolytic Zn-Sb alloy was plated out at -500 mV (vs. SHE) from a solution of zinc sulphate and 1-2 g/L antimony, made up from antimony tartrate. Fountoulakis [46] found (using cyclic voltammetry) that cobalt cementation with zinc in the presence of copper and antimony led to cobalt precipitation as a Cu-Sb-Co alloy. This alloy was more stable than pure cobalt and had a higher hydrogen overpotential. He concluded that both dissolved antimony and copper are required for the efficient removal of cobalt by cementation. Antimony is essential for precipitation of the alloy while copper increases its stability. The role of zinc during cementation was ignored because Fountoulakis used a zinc-free electrolyte in his studies. The stability of the deposits is an important consideration in cementation studies, as an unstable deposit will tend to have a smaller overall reaction rate and a shorter residence time before it must be removed from solution. Miller [31] found experimentally that only a small amount of antimony was needed to stabilize the cement from cobalt redissolution. Other researchers have found similar results. Tozawa et al [47] developed Ej^ -pH diagrams illustrating that while all metal arsenides were more stable than the corresponding metals, only cobalt and nickel antimonides were thermodynamically more stable than their corresponding metals. page 35 Fountoulakis [47] determined the stability of substrates electrochemically and found that Co-Sb compounds were less stable and had a lower hydrogen overpotential than Cu-Sb compounds. A Cu-Sb-Co alloy was found to have properties similar to Cu-Sb and Co-Sb alloys. It had a high hydrogen overpotential (higher than pure cobalt), was more stable than pure cobalt, and it dissolved at nobler potentials than the three metal components. Because of the corrosion of zinc in the hydrogen evolution region, Fountoulakis concluded that zinc alloys would be very unstable and would make a poor cathode substrate. Ingraham and Kerby observed that Zn-Cu alloys dissolve readily [48] in sulphuric acid media. Classic cobalt cementation processes use either a copper - antimony (250 mg/L and 10 mg/L, respectively) or a copper -arsenic (250mg/L and 100 mg/L, respectively) system. The arsenic-copper activated process is a more straightforward and stronger activating system than the antimony-copper process. However, the evolution of toxic arsine gas and the higher operating temperatures required for this process have encouraged a number of plants to use the copper-antimony system. 2.10Organics Small amounts of organic material can have a detrimental effect on the kinetics of cobalt removal during zinc dust purification. Organics that may find their way to purification include flocculants fi-om thickeners, and glue and foaming agents used in the tankhouse. Even though most organics are assumed to degrade with time and temperature, there is evidence that some organics may enter the purification process and cause cementation problems. Houlachi et al [49] have shown that the presence of organics in amounts as low as 0.5 mg/L can have a detrimental effect on impurity removal during zinc dust purification. It has been suggested that a chelation effect with cobalt may take place slowing the reaction with zinc dust, or flocculation of the zinc particles during cementation. Krauss [50] believed that organic interference with zinc dust purification must be due to an adsorption on the surface of the zinc powder because the concentrations of organics is too small for a mechanism based on complex ion formation. An adsorption mechanism would be consistent with previous research on the ftmction of organics as corrosion inhibitors. page 36 2.11 Conclusions A large amount of published data is available dealing with the kinetics of the cobalt cementation process using both advanced electrochemical techniques such as rotating disc electrodes and relatively straightforward experiments using zinc dust. Experiments using rotating disc electrodes are particularly useful for studying mass transfer controlled processes because a constant boundary layer thickness across which mass transfer occurs can be maintained. However, because cobalt cementation is a surface activated phenomenon the advantages of this using technique cannot be fully exploited. Kinetic constants obtained from rotating disc electrode tests, however, cannot be extrapolated accurately to zinc dust systems because the surface area of the zinc dust is not well defined. The advantage of performing laboratory scale experiments using zinc dust is that such experiments provide more realistic kinetic data since the agitation, surface area and reactions occurring experimentally are representative of the commercial processes presently being practised. Even though zinc dust experiments have been performed and documented, not all of the experimental data can be compared effectively due to differences in experimental design. Morphological and compositional analyses of the cement residues have not been performed extensively in previous investigations. Such analyses may reveal important insights into the characteristics of the cathode surface for cobah reduction. Cobalt removal kinetics are influenced strongly by the chemical properties of the cathode site and the available surface area. The kinetics of the reaction may also be affected by the concentration of the activating agents, pH, temperature, the ionic strength of the electrolyte, and the presence of organics in the electrolyte. The main objective of this study is to examine and summarize the effects of these operating parameters, and to attempt to correlate these findings with cementation residue morphology and composition. An expectation is that the results can be applied directly to improving the kinetics of the hot stage purification process being practised at Cominco. page 37 3 EXPERIMENTAL SET-UP AND CONDITIONS Cementation tests were performed using zinc powders for kinetic studies since this approach is more representative of the industrial process. To improve the reproducibility, a batch reactor was used as opposed to a continuous reactor. Purified zinc sulphate solution from Cominco's Trail plant was used for all of the experiments. This solution contained 151 g/L Zn, <0.1 mg/L Co, <0.1 mg/L Cd, 0.01 mg/L Sb, and <0.1 mg/L Tl. A column of activated carbon was used to treat this solution prior to the cementation studies in an effort to remove any organics that might be present in the solution which could cause unreproducible results. The solution was then adjusted to the desired concentrations of cobah, copper and antimony using reagent grade cobalt sulphate (J.T. Baker), cupric sulphate and potassium antimony tartrate (BDH Chemicals Ltd.), respectively. Wet atomized zinc powder containing 1.0% Pb, 0.002% Fe, <0.01% Al and <0.001%Cu was obtained from Cominco and screened to -100+140 mesh. Other researchers have found that alloying zinc dust with lead improved the stability of the cemented cobalt deposit [27]. A 4L glass reactor (Figure 3-1) containing four stainless steel baffles was used for all experiments. The role of the baffles was to minimize the effects of vortexing which can introduce oxygen to the system. Oxygen is undesirable as it can cause resolution of the cementation residue. The cementation batch reactor was heated in a thermostated water bath and maintained to within TC. To start an experiment, three litres of the test solution was heated while sparging with nitrogen until the set temperature was reached. Sparging with an inert gas helps to eliminate oxygen from the solution [51,52]. The pH was adjusted to 3.7 with reagent grade sulphuric acid and zinc powder was added. The electrolyte was agitated at 830 RPM using a 64-mm diameter A310 impeller connected to a laboratory stirring motor through a stainless steel shaft. The agitation rate was checked periodically with an optical tachometer. An acrylic lid was used to reduce evaporative losses. The pH was controlled to within 0.2 pH units of the target value using a pH controller. During the test the solution pH, solution potential (by a Fisher Scientific bright platinum combination electrode referenced versus Ag-AgCl), and the activity of Cu^ ^ (by a cupric page 38 ion specific electrode referenced versus a Leeds and Northrup #117392 Ag-AgCl difiusion-type reference electrode) were monitored. Ag/AgCI reference electrode platinum combination electrode V nitrogen sparger baffle pH probe cupric specific ion electrode sampling port temperature probe A310 impeller Figure 3-1. Schematic Diagram of the Experimental Setup. Solution sampling was performed using a 50-mL pipette at intervals of 5, 15, 30, 45, 60 and 90 minutes. The sample was filtered through a Millipore filter under vacuum, and the filter residues were water washed, rinsed twice with dilute sulphuric acid and finally rinsed in water before being dried at room temperature in a vacuum desiccator. At the conclusion of each batch experiment the solution was filtered under a nitrogen atmosphere. The filter cake was rinsed twice with water, then with alcohol, and finally dried at room temperature in a vacuum desiccator. These tests were performed using a screened zinc powder in an effort to obtain better reproducibility by narrowing the size distribution of the zinc substrate. Most tests were performed using 4 g/L of "untreated" zinc powder where the activating agents and cobalt were all present in solution prior to zinc powder addition. In tests using "pre-coated" zinc powder, cobalt was absent fi"om the solution prior to the addition of zinc powder so that cementation with the activation agents copper and antimony could take place. Preliminary experiments showed that page 39 twenty minutes was an adequate amount of time to remove copper and antimony from the solution to concentrations determined by the detection limits of the assay method. Stirring of the reactor was interrupted, cobalt was added to the solution, and then the agitator was restarted. This point marks the onset of the cobalt cementation experiment (time zero). Selected residues were characterized by scanning electron microscope (SEM) and wavelength dispersive X-ray (WDX) analysis. The residues were mounted and examined using two different techniques. In the first technique, cement sponge was sprinkled onto fresh carbon paint overlaying a graphite stud. After allowing the sample to dry, any loose cement sponge was gently tapped off. In the second SEM technique, cross-sections of the cement sponge were analyzed. Samples were mounted in a low viscosity resin and polished with 6 micron alumina. X-ray diffraction (XRD) was also performed on these residues. However, because of the large zinc background, no distinct diffraction pattern other than that of zinc could be determined. The filtrates were analyzed by atomic absorption spectrophotometry for cobalt, copper and antimony. A colorimetric cobalt assay was also performed using Nitroso-R salt as a check. There was good correlation between the two methods. page 40 4 RESULTS AND DISCUSSION 4.1 THE STANDARD EXPERIMENT The "standard" experiment was performed under the following initial conditions: • 4 g/L loading with -100+140 mesh Cominco zinc dust • 73°C • "natural pH": the pH was allowed to come to its own equilibrium fi-om a starting pH of 3.7 measured at 73 °C • 26mg/LCo, 46mg/LCu, 1.5mg/LSb • Agitator running at 830 RPM • nitrogen sparging • carbon treated clarifier overflow spiked with the impurities Co, Cu and Sb to simulate the hot stage purification feed solution The cobalt removal, cupric specific ion electrode potential, slurry potential and pH profiles are illustrated in Figures 4-1, 4-2, 4-3 and 4-4, respectively. o g2 o O " -. -{ 1 0 20 1 1 40 60 time (minutes) 1 80 100 0) ouu o . 200 CO > E % 100 (D LU O S. (0 *k. a. ^ 1 r\f\ \ • \ • \ ' -w 1 Q -lUU 0 20 r , 1 , 1 40 60 8C Time (minutesO • 1C Figure 4-1. Cobalt Removal Profile for the Standard Experiment. Figure 4-2. Cupric Specific Ion Electrode Potential Profile for the Standard Experiment. page 41 -1000 20 40 60 80 Time (minutes) 100 t . o 4.6 4.4 ^ 4 . 2 4 3.8 • , ^ - « — s ' ^ * ^ * ' ^ -,^J^_JU— 1 20 40 60 Time (minutes) 80 100 Figure 4-3. Slurry Potential Profile for Standard Experiment. Figure 4-4. pH Profile for Standard Experiment. A cobalt removal rate constant of 788x10'* sec'^  and a final cobalt concentration of 0.72 mg/L afl;er 90 minutes were obtained in the standard run. The cupric ion specific electrode potential reached a low measured potential afl;er only 10 minutes, indicating that the cupric ion was removed rapidly from solution. The slurry potential remained relatively stable at around -800 mV (vs Ag/AgCl) throughout the experiment, while the pH increased rapidly after the addition of zinc dust. 4.2 EFFECT OF ACTIVATION AGENTS The role of the activation agents was examined using untreated zinc powder. The reactor was loaded with 4 g/L of zinc powder for these experiments and the pH was allowed to reach its own level (thus termed, natural pH) unless otherwise stated. The results below summarize the effects of antimony, copper and the combined effects of copper and antimony on cobalt removal with zinc powder. Antimony Activation The objective of this series of tests was to isolate the role of antimony in the activation of zinc powder. The experiments were performed by testing a solution containing only cobalt and antimony. It was found that antimony can activate the cementation of cobalt on its own. page 42 However, the final cobalt concentration after 90 minutes was much higher (10.3 mg/L) than when both activation agents were present (0.72 mg/L). In addition, the cobalt cementation rate was found to be slower. These results indicate that cobalt may cement on zinc, or on antimony or on a zinc-antimony intermetallic or alloyed substrate. 4.4 4.2 I Q. 3.8 3.6 3.4 0 mg/L Sb 1.5 mg/L Sb 20 40 60 80 Time (minutes) 100 Figure 4-5. Effect of Antimony Activation on pH. Initial Conditions: [Co]=26 mg/L, [Cu]=0 mg/L, [Sb]=0 to 1.5 mg/L, 73°C, 4g/L zinc powder loading, natural pH The graph of pH versus time (Figure 4-5) shows that the pH increases more slowly in the presence of Sb. The pH increase can be linked to hydrogen evolution, and therefore, it can be assumed that antimony, in the absence of copper, suppresses hydrogen evolution during cobalt cementation. The hydrogen evolution reaction competes with cobalt deposition for electrons made available from the dissolution of zinc. Antimony, therefore, plays a role in depressing the competing hydrogen reaction and in providing or preparing a deposition surface for cobalt. The hydrogen inhibiting behaviour by antimony is well known in corrosion protection [53]. page 43 Copper Activation In this series of tests, copper and cobalt were present initially in the solution prior to zinc powder addition. Copper activation of the zinc powder can remove cobalt from zinc electrolyte, but only by a small amount (20.6 mg/L after 90 minutes). The cobalt removal rate exhibited first order kinetics, but the rate constants were slower than that of antimony activation (Figure 4-8). According to Figure 4-6, copper (in the absence of antimony) does not appear to have a strong effect on hydrogen evolution. A pH of about 4.2 was reached after 90 minutes in both experiments. This suggests that copper acts primarily in providing a substrate for cobalt cementation, while antimony plays a role in both maintaining a high hydrogen overpotential for effective cobalt removal, and in providing or conditioning the surface for cobalt deposition. 20 40 60 Time (minutes) 80 100 46 mg/L Cu . 0 mg/L Cu I • I 1 L 20 40 60 80 Time (minutes) 100 Figure 4-6. Effect of Copper Activation on Figure 4-7. Effect of Copper Activation on pH. Initial Conditions: [Co]=26 mg/L, Cupric Specific Potential. Initial Conditions: [Cu]=0 to 46 mg/L, [Sb]=0 mg/L, 73°C, 4g/L [Co]=26 mg/L, [Cu]=0 to 46 mg/L, [Sb]=0 zinc powder loading, natural pH mg/L, 73 °C, 4g/L zinc powder loading, natural pH The monitoring of the solution with a cupric specific ion electrode provided a means to monitor the change in cupric ion activity with time (Figure 4-7). It was found that copper cemented from solution to low levels rapidly within the first 10 minutes, but it took slightly longer to reach a low page 44 potential than in the standard case. Based on the better cobalt removal, the cobalt deposition overpotential must be higher on this substrate than on the Sb-activated substrate. Copper and Antimony (Cu-Sb) Activation The concentrations of cobalt, copper and antimony in solution were followed during the course of the cementation experiment and the results are summarized in Figure 4-8. Antimony and copper are both removed from solution to low levels (0.01 - 0.02 mg/L and <0.1 mg/L, respectively) within the first ten minutes. Within the analytical limits of the assay method the concentration of these species in solution remained low and did not change significantly after this rapid initial removal. Cobalt precipitation occurred at a much lower rate and follows first-order reaction kinetics. 6 100 time (minutes) Figure 4-8. Copper, Antimony and Cobalt Removal versus Time. Initial Conditions: pH=3.6, 73°C, lOg/L zinc powder loading, [Co]=27 mg/L, [Cu]=36 mg/L, [Sb]=1.5 mg/L. page 45 Earlier results have shown that antimony has a role in both the deposition of cobalt and in the suppression of the hydrogen evolution reaction, while copper appears to be involved in providing a suitable substrate for the deposition of cobalt. The presence of both activating agents was more beneficial for cobalt removal than the presence of each activating agent on its own. The activation mechanism, therefore, is complicated with the two activators acting synergistically. Hydrogen evolution was more active during Cu-Sb activation (pHgo^= 4.3), than for either the Sb-activated (pH^omn^ 4.1) or the Cu-activated (pH9o^= 4.2) conditions. Figure 4-9 summarizes the cobalt removal behaviour for the various activating agents. 4 O O 'o ' O Cu & Sb Activation (46 mg/L Cu, 1.5 mg/L Sb) iK Unactivated Zinc Powder )^ Cu Activation (46 mg/L Cu) Sb Activation (1.5 mg/L Sb) 40 60 time (minutes) 100 Figure 4-9. Effect of the Copper and Antimony Activators on Cobalt Removal with Zinc Powder. Initial Conditions: [Co]=26 mg/L, 73°C, 4 g/L zinc powder, natural pH page 46 4.2.1 Extended Time Experiments: Extended (4-hour) experiments were performed using the initial conditions of 26 mg/L cobalt, 46 mg/L copper and 1.5 mg/L antimony. A cobalt concentration of 0.4 mg/L was reached after 120 minutes (Figure 4-10), after which the cobalt concentration remained relatively constant. This plateau in the cobalt removal rate coincided with a sudden rise in potential (Figure 4-11). It was postulated that at a pH (Figure 4-12) of 4.4, a passivating surface barrier begins to form on the surface of the cementation residue, acting to insulate the particle from the platinum electrode and causing a rise in slurry potential [54]. This observation may be usefijl in monitoring the progress of the cementation reaction in the lab and in the plant. 50 100 150 200 Time (minutes) 250 300 Figure 4-10: Cobalt Removal for the Extended Time Experiment. Initial Conditions: [Co]=26mg/L, [Sb]=1.5 m&O., [Cu]=46 mg/L, 73°C, natural pH page 47 50 100 150 200 Time (minutes) 300 H . O 4.4 4.3 4.2 i .4 .1 4 3.9 3.8 7 7 . ^ ^ ^ ^ " " " ^ ^ ^ ^ y ^ - / -1 \ , 1 , 1 , 50 100 150 200 Time (minutes) 250 300 Figure 4-11: Slurry Potential (vs. Ag/AgCl) Figure 4-12: pH Profile for the Extended Time Profile for the Extended Time Experiment. Experiment. Initial Conditions: [Co]=26 mg/L, [Sb]=1.5 Initial Conditions: [Co]=26 mg/L, [Sb]=1.5 mg/L, [Cu]=46 mg/L, 73°C, natural pH mg/L, [Cu]=46 mg^, 73°C, natural pH 4.2.2 The Effect of Copper Concentration on Cu-Sb Activation A copper series of experiments (Figure 4-13) was performed where initial copper concentrations were varied while cobalt and antimony levels were initially at 26 mg/L and 1.5 mg/L for each test, respectively. The presence of copper in solution improved significantly the cobalt cementation rate, but only up to 46 mg/L of initial copper. The initial rate of cobalt removal was almost identical for the two cases of 46 mg/L and 92 mg/L copper, but the 46 mg/L copper curve begins to deviate fi^om linearity after 30 minutes suggesting that some other mechanism is beginning to take effect. page 48 20 40 60 80 time (minutes) 100 20 40 60 [Cu] (mg/L) 80 100 Figure 4-13. Effect of Copper Figure 4-14. Effect of Copper Concentration Concentration on the Cobalt Removal Rate on Cobalt Removal after 90 Minutes. Initial in the Presence of 1.5 mg/L Antimony. Conditions: [Co]=26 mg/L, [Sb]=l.5 mg/L, Initial Conditions: [Co]=26 mg/L, [Sb]=l .5 [Cu]=0 to 92 mg/L, 73C, natural pH mg/L, [Cu]=0 to 92 mg/L, 73C, natural pH Figure 4-14 displays the relationship between initial copper concentration and the final cobalt removal after 90 minutes under normal conditions. The presence of copper in solution containing 1.5 mg/L Sb greatly improves cobalt removal, but additional copper above 46 mg/L Cu did not lead to further improvements. If the role of copper is strictly that of a substrate, then increasing this area should increase the cobalt cementation rate provided that other effects, such as surface passivation, can be controlled and that sufficient Sb is present to activate this surface. As the initial copper concentration is increased, the rate at which the pH rises is accelerated and the final pH is increased (Figure 4-15). This suggests that a copper substrate possesses a lower hydrogen overpotential than an antimony surface, and that increasing the initial copper concentration in solution increases this surface area. page 49 4.6 3.8 -3.6 Omg/LCu 18.4mgyLCu 46 mg/L Cu 92 mg/L Cu 20 40 60 80 Time (minutes) 100 Figure 4-15. Effect of Copper Concentration on the pH Profile. Initial Conditions: [Co]=26 mg/L, [Sb]=1.5 mg/L, [Cu]=0 to 92 mg/L, 73C, natural pH 4.2.3 Effect of Antimony Concentration on Cu-Sb Activation Cobalt cementation was also affected by the initial antimony concentration in solution. An antimony series (Figure 4-16) was generated by varying the initial level of antimony from 0 to 3.0 mg/L holding constant the copper and cobalt concentrations. The initial rates of cobalt removal were quite similar at antimony levels of 1.5 mg/L and 3.0 mg/L. Increasing the antimony concentration beyond 1.5 mg/L , in the presence of 46 mg/L copper, did not improve the ultimate cobalt removal (Figure 4-17), indicating that there may be some maximum amount of antimony that can deposit or adsorb onto the surface. In fact, excess antimony may remain in solution and may not participate further in catalyzing the cobalt reaction. page 50 20 40 60 80 time (minutes) 100 1 1.5 2 2.5 [Sb] (mg/L) Figure 4-16: EflFect of Antimony Figure 4-17. Effect of Antimony Concentration on the Cobalt Removal Rate Concentration on Cobalt Removal in the in the Presence of Copper. Initial PresenceofCopper after 90 Minutes. Initial Conditions: [Co]=26 mg/L, [Cu]=46 Conditions: 4g/L zinc dust loading, 26 mg/L mg/L, [Sb]=0 to 3.0 mg/L, 73°C, natural Co, 46 mg/L Cu, [Sb]=0 to 3.0 mg/L, 73°C, pH natural pH. With 3 mg/L Sb in solution, cobalt cementation stopped after 60 minutes coinciding with a rise in the slurry potential rise (Figure 4-18), and coinciding with the point at which the pH rose above 4.4 (Figure 4-19). This is believed to be due to the formation of a zinc hydroxide film or a basic zinc sulphate salt on the surface of the zinc particle. In the presence of copper in solution, increasing concentrations of antimony resulted in an increase in hydrogen generation. This suggests that the mechanism for cobalt removal is complicated, and that the role played by antimony and copper are concentration dependent. page 51 -0.3 -0.4 -0.5 CS £ ^ - 0 . . CO -0.8 -0.9 • (^ 3mg/LSb , iV' ' \ , -1 \ / Omg/LSb V / . . >A, y^iltt^i^fifJR^ "~~~ 1.5 mg/L Sb 1 1 r 1 20 40 60 time (minutes) 4.6 4.4 -4.2 Q. 80 100 3.8 3.6 3 mg/L Sb _ 1.5 mg/L Sb J ^ ' "T -- .V-- • • • / . ' ^ T-, - / 0 mg/L Sb it 1 1 , 1 , 1 . . ^ ; 1 — ; 20 40 60 80 Time (minutes) 100 Figure 4-18: Slurry Potential Profile for the Figure 4-19: pH Profile for the Copper Series. Copper Series. Initial Conditions: 4g/L zinc Initial Conditions: Initial Conditions: 4g/L dust loading, 26 mg/L Co, 46 mg/L Cu, zinc dust loading, 26 mg/L Co, 46 mg/L Cu, [Sb]=0 to 3.0 mg/L, 73°C, natural pH. [Sb]=0 to 3.0 mg/L, 73°C, natural pH. 4.2.4 Effect of High Sb and Cu Concentrations on Cu-Sb Activation It was hypothesized that increasing the concentration of both activation agents added to the solution should improve the cobalt removal kinetics. Increasing both the copper (to 92 mg/L) and the antimony (to 3 mg/L) concentrations improved the initial cobalt removal rate during the first 15 minutes of the reaction (Figure 4-20). Afl:er 15 minutes, the rate decreased and a slower kinetic region was established. page 52 46 mg/L Cu 20 40 60 80 time (minutes) 100 92 mg/L Cu 20 40 60 80 Time (minutes) 100 Figure 4-20: Effect of Copper Figure 4-21: Effect of Copper Concentration Concentration on the Cobalt Removal on the Slurry Potential Profile in the Presence Rate in the Presence of 3 mg/L Antimony, of 3 mg/L Antimony. Initial Conditions: Initial Conditions: [Co]=26 mg/L, [Co]=26 mg/L, [Sb]=3.0 mg/L, [Cu]=46 and [Sb]=3.0 mg/L, [Cu]=46 and 92 mg/L, 92 mg/L, 73°C 73°C The final cobalt concentrations (for initial antimony concentration of 3 mg/L) after 90 minutes for the 92 mg/L copper and 46 mg/L copper were 0.54 mg/L and 0.94 mg/L, respectively. Increasing levels of copper tend to reduce the final cobalt concentration in the solution, but cobalt removal ceased after 60 minutes for the experiment using 46 mg/L copper. The slurry potential versus time plot (Figure 4-21) showed that a rise in potential did not occur for the 92 mg/L copper case, but a potential rise was observed for the 46 mg/L copper case. The pH at 90 minutes were 4.4 and 4.2 for the low and high copper concentrations respectively (Figure 4-23). Copper cementation is a mass transfer controlled process [31]. This initial slow deposition is probably attributable to a nucleation period needed before copper can be rapidly deposited from solution. The removal of cobalt in the first 5 minutes of the reaction also occurred slowly, coinciding with this lag period for copper precipitation. page 53 The graph which monitored the cupric ion activity in the system appears in Figure 4-22. For the 92 mg/L copper and 3 mg/L antimony case, copper removal takes place in two stages with the majority of the copper removed within the first 10 minutes. > E 0.5 £ 0.4 o o> 5 £ I 0.1 , > 1 1 « • * 46 mg/L Cu 92 mg/L Cu 1 ^ 1 « 1 ' 1 , 1 ' I < 1 ( 1 1 . 1 . 1 . 1 . 20 40 60 80 Time (minutes) 100 92 mgA. Cu 46 mg/L Cu J I 1 I 1 L 40 60 Time (minutes) 100 Figure 4-22. Cupric Potential Profile for Figure 4-23. pH Profile for the Copper Series the Copper Series in the Presence of 3 in the Presence of 3 mg/L Antimony. Initial mg/L Antimony. Initial Conditions: Conditions: [Co]=26mg/L, [Cu]=46 to [Co]=26mg/L, [Cu]=46 to 92mg/L, 92mg/L, [Sb]=3.0 mg/L, 73°C, natural pH. [Sb]=3.0 mg/L, 73°C, natural pH. The non-linearity of the curve in Figure 4-20 indicates that an effect other than cobah deposition is occurring. The most probable explanation is that the cathode surface area was increasing during surface activation (probably due to copper deposition), and that there was enough antimony ion available in solution to activate this new surface. Table 4-1 summarizes the effects of activating agent concentration on the initial cobalt cementation rate constant, k, calculated fi'om the -ln([Co]/[Co]init) versus time graphs. The normalized k values quoted in the table represents the k values normalized to the value obtained for the standard condition of [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, and a temperature of 73 °C. In general, faster initial cobalt removal rates were observed with elevated initial antimony page 54 levels (3 mg/L), but the fastest rates were obtained with high concentrations of both antimony and copper in solution. Despite the various rates of cobah removal, the final cobalt concentrations were affected adversely by the rising pH of the system. Too high a pH promotes the formation of a passivating barrier on the surface consisting of either a basic zinc sulphate salt or a zinc hydroxide/oxide film which acts to inhibit effectively cobalt removal. Table 4-1: Summary of the Effects of Cu-Sb Activation Agents Conditions: 73°C, natural pH, 4 g/L zinc powder loading Initial [Co] mg/L 26 26 26 26 26 26 26 26 Initial [Cul mg/L 92 92 0 18.4 46 46 46 0 Initial [Sb] mg/L 1.5 3 1.5 1.5 1.5 0 3 0 Cementation Rate Constant, kxlO'' (sec*) 806 1053 172 596 788 50 895 31 Normalized Rate Constant 1.02 1.34 0.2 0.7 1 0.06 1.14 0.04 [Co] at 90 Minutes, (mg/L) 0.71 0.54 10.3 1.03 0.72 20.6 0.94 24.0 page 55 4.3 EFFECT OF PRE-COATING ZINC POWDERS It was hypothesised that using a zinc powder pre-coated with the activation agents, would give better cobalt removal kinetics since the area and the electrochemical characteristics of the cathode sites would be already established. For this series of experiments, pre-coated zinc powder was produced by contacting zinc powder with clarifier overflow solution containing 1.5 mg/L antimony and 46 mg/L copper under the test conditions. After allowing sufficient time for surface activation to occur, 26 mg/L cobalt was introduced into the solution. This was considered to be time zero. There were no significant differences in cobalt removal kinetics between pre-coated zinc powder and untreated zinc powder (Figure 4-24 and 4-25). The activation of the zinc powder by antimony and copper deposition v^thin the first 10 minutes eliminated any advantages imparted by precoating. Untreated Zinc Powder Pre-Coated Zinc Powder —0— 40 60 time (minutes) 100 Figure 4-24: Effect of Pre-Coating Zinc Powder with 1.5 mg/L Sb and 46 mg/L Cu. Initial Conditions: [Co]=26 mg/L, 73°C, natural pH page 56 Pre-Coated Zinc Untreated Zinc Powder Powder • —•&- ^ 40 60 time (minutes) 80 100 Figure 4-25: Effect of Pre-Coating Zinc Powder with 1.5 mg/L Sb and 92 mg/L Cu. Initial Conditions: [Co]=26 mg/L, 73°C, natural pH Pre-coating zinc powder with activators illustrates that the presence of activators in solution is not needed once they have precipitated on the surface of the zinc powder. It is expected, therefore, to find under SEM examination a Cu, Sb, Cu-Sb, Cu-Zn, Sb-Zn or Cu-Sb-Zn substrate or substrates on the activated surface. When cobalt deposits onto this surface, a layering effect should be observed. This pre-coating experiment also shows that co-deposition of cobalt with copper or antimony does not necessarily occur. page 57 4.4 EFFECT OF INITIAL COBALT CONCENTRATION As shown in Figure 4-26, the initial cobalt concentration does not affect the kinetics of the cobalt cementation reaction . Good reproducibility was attained and the pH monitoring (Figure 4-27) was consistent for experiments with different initial cobalt concentrations. Figure 4-28 is a graph of the final cobalt concentration after 90 minutes versus the initial cobalt concentration. 5 4 *—*. - t : •M,3 o O i ^ * "o* o '?2 # 1 ni 6.5 mg/L Co —B— 26 mg/L Co - • -^ J B ^ X * ^ , 1 , r 13 mg/L Co 39 mg/L Co - • -, 'yy ry/^^ ' / 4 l * X^^ * V * 1 ' • y^^ • 1 20 40 60 80 Time (minutes) 100 4.8 3.8 3.6 6.5 mg/L Co 13 mg/L Co - o -26 mg/L Co 39 mg/L Co — • — — e — 20 40 60 Time (minutes) 80 100 Figure 4-26: Effect of Initial Cobalt Figure 4-27: Effect of Initial Cobalt Concentration on Cobalt Removal. Initial Concentration on the pH Profile. Initial Conditions: [Cu]=46 mg/L, [Sb]=1.5 mg/L, Conditions: [Cu]=46 mg/L, [Sb]=L5 mg/L, [Co]=6.5 to 39 mg/L, 73°C, natural pH [Co]=6.5 to 39 mg/L, 73''C, natural pH These results show that cobalt removal takes place by first-order kinetics and that the rates are virtually independent of the initial cobalt concentration. page 58 1.6 1.4 -1.2 C3) E ~—' c E o o> «. (1) CO o O 1 0.8 0.6 0.4 0.2 Ofi 20 30 Initial [Co] (mg/L) 50 Figure 4-28: Effect of Initial Cobalt Concentration on Cobalt Removal at 90 Minutes. Initial Conditions: [Cu]=46 mg/L, [Sb]=1.5 mg/L, [Co]=6.5 to 39 mg/L, 73°C, 4g/L zinc dust, natural pH page 59 4.5 EFFECT OF PH This series of experiments examined the effects of pH on cobalt cementation. pH levels were controlled with a pH controller and the solutions were adjusted to a desired pH value by the addition of sulphuric acid. This is different from previous experiments which allowed pH levels to increase naturally during the course of the experimental run. Figures 4-29 to 4-32 and Table 4-2 summarize the results from these experiments. Maintaining a pH at a relatively high level apparently results in a passivating barrier on the surface of the zinc dust particle which decreases the kinetics of cobalt cementation. However, a high pH also lowers the hydrogen generation rate. A low pH appears to increase the competition between the hydrogen and cobalt reduction. This is reflected in a decrease in the cobalt cementation rate constant and an increase in the acid addition rate. Figures 4-29 and 4-30 together show that a high hydrogen evolution rate can be correlated to both a more positive slurry potential and a slower cobalt removal rate constant. Figure 4-32 indicates that there is an optimum pH range for cobalt cementation, but the natural pH gives the highest cobalt removal rate constant (Figure 4-31). page 60 250 -1000 20 40 60 80 Time (minutes) 100 100 time (minutes) Figure 4-29. Effect of Controlled pH on Slurry Figure 4-30. Effect of Controlled pH on the Potential. Initial Conditions: [Co]=26 mg/L, Hydrogen Addition Rate. Initial Conditions: [Cu]=46 mg/L, [Sb]=l .5 mg/L, 73°C, 4 g/L [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=l .5 mg/L, zinc powder loading 73°C, 4 g/L zinc powder loading, controlled pH, acid addition = 8.82 moles/L Table 4-2: Effect of pH on Cementation Kinetics. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L 73 °C, 4 g/L zinc powder loading, 90 minutes. pH 3 3.6 4 4.4 natural Cementation Rate Constant, kxlO * (sec-') 489 577 684 575 788 Normalized Rate Constant 0.62 0.73 0.87 0.73 1 [Col at 90 Minutes, (mg/L) 2.76 1.49 0.79 1.10 0.72 H* Addition Rate (moles/min) 2.29 0.85 0.10 0.01 0 page 61 pH=3.0 pH = 3.6 pH=4.0 pH = 4.4 natural pH — 0 — —•— 20 40 60 80 100 time (minutes) Figure 4-31: Effect of Controlled pH on Cobalt Removal. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, 73°C, 4g/L zinc dust loading. 2.8 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 pH Figure 4-32: Effect of Controlled pH on Cobalt Concentration at 90 Minutes. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg^, 73°C, 4g/L zinc dust loading. Because the electrolyte is strongly buffered, the change in pH during the natural pH experiments can only be a rough indication of the consumption of hydrogen ions in the solution. However, the hydrogen consumption measured from the controlled pH experiments shows that the rate of hydrogen evolution increased with increasing solution acidity. A low hydrogen consumption rate was experienced in the experiment at a controlled pH of 4.4. 4.6 EFFECT OF TEMPERATURE Cobalt removal was studied in the temperature range of 65 to 90°C. The results are summarized in Table 4-3 and Figures 4-33 through 4-35. The lowest cobalt concentrations achieved after 90 minutes and the fastest cobalt removal rates were experienced at 90°C, the highest temperature tested. page 62 6 -O O 'o' O 2 -65C 73C 80C 85C 90C —B • e • — - -X- • ^ ' y^'^~~ Z^^ ^ / / ^^"^^^^^ Wyy j ^ r - " " ' ^ ' ^ i^"^^ 1 1 1 1 - 5K 80 100 0 20 40 60 time (minutes) Figure 4-33: Effect of Temperature on the Cobalt Removal Rate. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, natural pH The cobalt removal rates and the final cobalt concentrations are both improved with increasing temperatures (Figure 4-33 and 4-34). Cobalt cementation appears to plateau after 60 minutes at temperatures higher than 80°C. page 63 Table 4-3: Effect of Temperature on Cobalt Removal Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L 73°C, natural pH, 4 g/L zinc powder loading TCC) 65 73 80 85 90 Cementation Rate Constant, k x l O ^ (sec"') 432 788 918 1238 1507 Normalized Rate Constant 0.55 1 1.16 1.57 1.91 [Co] at 90 Minutes (mg/L) 2.6 0.72 0.31 0.28 0.16 Since the solubility of basic zinc sulphate increases with increasing temperatures, it is unlikely that this kinetic plateau can result from basic zinc sulphate sah formation on the zinc powder surface. It is more probable that the hydrogen evolution reaction becomes more active at elevated temperatures, promoting the formation of a localized passivating barrier on the surface. The improved rates are reflected by the more reducing potential of the slurry at elevated temperatures (Figure 4-35). From the slope of an Arrhenius plot (Figure 4-36), an apparent activation energy of 86.6 kJ/mol was obtained. This relatively high value suggests that the cobalt removal process is chemically or electrochemically controlled. This value is, however, significantly lower than that reported by Adams and Chapman [23] for the Sb activated process (115 kJ/mole). page 64 60 65 70 75 80 85 90 95 Temperature (degrees Celsius) -0.65 40 60 80 Time (minutes) 100 Figure 4-34: Effect of Temperature on Cobalt Removal at 90 Minutes. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, natural pH Figure 4-35: Effect of Temperature on the Slurry Potential Profile. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, natural pH (6.4) (6.6) (6.8) _ (7) S (7.2) (7.4) (7.6) (7.8) Apparent Energy of Activation = 86.6 lU/mole SLOPE = 10.42 J I I I I I I , L 0.0027 0.0028 0.0029 0.003 0.00275 0.00285 0.00295 1/T(1/K) Figure 4-36. Arrhenius Plot for Cobalt Cementation using the Cu-Sb Activation System. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, natural pH, 4 g/L zinc powder page 65 4.7 EFFECT OF ZINC DUST LOADING The effects of zinc dust loading are summarized in Figures 4-37 through 4-40. A linear relationship exists between zinc dust loading and the cobalt removal rate but only up to 4g/L. The superficial explanation is that increasing the zinc dust loading increases the available surface area and, therefore, increases the rate of cobalt removal. However, as loading was increased to levels greater than 4g/L, the relationship began to deviate fi^om linearity. 0.0016 10 g/L zinc dust 2 g/L zinc dust 20 40 60 80 time (minutes) 100 2 4 6 8 10 Zinc Dust Loading (g/L) 12 Figure 4-37. Effect of Zinc Dust Figure 4-38. The Effect of Zinc Dust Loading Loading on Cobalt Removal. on the Cobalt Removal Rate Constant. Initial Conditions: -100+140 mesh, 26 Initial Conditions: 26 mg/L Co, 46 mg/L Cu, mg/L Co, 46 mg/L Cu, 1.5 mg/L Sb, 1.5 mg/L Sb, 73°C, natural pH. 73°C, natural pH. page 66 Table 4-4. The Effect of Zinc Dust Loading on the Cobalt Cementation Rate. Initial Conditions: 26 mg/L Co, 46 mg/L Cu, 1.5 mg/L Sb, 73°C, natural pH. Zinc Dust Loading (g/L) 2 4 10 Cobalt Cementation Rate Constant, kxlO*(sec*) 406 788 1355 [Col at 90 Minutes (mg/L) 2.98 0.72 0.10 For example, MacKinnon [55] determined a surface area of 440.2 ( ±4.4) cmVg by the BET analysis for his 105 to 149 micron sized zinc dust. A BET analysis was performed on the -100+140 mesh (126 to 174 micron) zinc powders used in these experiments (Table 4-5) and was found to be slightly higher at 668 to 785 cmVg. Table 4-5. Summary of the Surface Area Analysis by BET. mesh size -200, +270 -140, +200 -100, +140 -50, +100 average area using 10% nitrogen (cmVg) 1900 1548 668 655 average area using 30% nitrogen (cmVg) 1938 1500 785 803 If the -100+140 mesh particles were assumed to be monosized spheres with a diameter of 125 microns, a theoretical area of 67.6 cm^ per gram was calculated (see Appendix C). The large discrepancy between the measured and the theoretical areas can be attributed to the roughened surface and the non-spherical shape of the zinc powders tested. A roughened surface can greatly increase the surface area. The problems with correlating the BET or any rudimentary surface area analysis performed on zinc dust to a cobalt removal rate arise from the fact that the true cathodic surface area cannot be differentiated by these techniques. Removing cobaU to a concentration below 0.3 mg/L within 90 minutes at 73°C was difficult and a large concentration of zinc dust was needed to reach this concentration. Despite the appreciable cobalt removal improvement with increasing zinc powder loading up to 4g/L, subsequent improvement in the cobalt removal rate was not found to be proportional to the total available page 67 surface area of the zinc powder. Figure 4-38 suggests that beyond 4 g/L zinc powder loading, the ability to remove cobalt begins to plateau for an activation with 46 mg/L Cu and 1.5 mg/L Sb. This suggests that the available cathodic surface area for cobalt cementation must be proportional to the concentration of the activation agents in solution, implying that faster cobalt removal can be achieved by increasing both zinc powder loading and the concentration of the activation agents. The results from the cupric potential profile (Figure 4-39) show that copper was removed from solution faster with increasing zinc dust loading and this can be attributed to the larger surface area available with higher loading. 300 S 250 2 g/L zinc dust 4 g/L zinc dust 10 g/L zinc dust 5 10 15 20 Time (minutes) 25 > E O < 5 -200 -400 .5 -600 c B o ^ -800 _3 CO -1000 10 g/L zinc dust 4 g/L zinc dust / " ', ' . ' ' 2 g/L zinc dust 20 40 60 80 100 Time (minutes) Figure 4-39. The Effect of Zinc Dust Loading on the Cupric Potential Profile. Initial Conditions: 26 mg/L Co, 46 mg/L Cu, 1.5 mg/L Sb, 73°C, natural pH. Figure 4-40. The Effect of Zinc Dust Loading on the Slurry Potential Profile. Initial Conditions: 26 mg/L Co, 46 mg/L Cu, 1.5 mg/L Sb, 73°C, natural pH. One important feature in Figure 4-40 was that with 10 g/L zinc dust loading, the slurry potential increased by about 200 mV. This coincided with a slight decrease in the cobalt removal experienced after this time, suggesting the onset of basic zinc sulphate or some other passivating surface film. page 68 4.8 CEMENTATION IN AN INITIALLY ZINC-FREE SOLUTION Cobalt removal was performed with a solution made from distilled water and normal additions of impurities and activation agents. It has been assumed by many authors, such as Fountoulakis [46], that basic zinc sulphate formation inhibits cobalt removal. The objective of this test was to examine the cobalt removal kinetics in a solution where basic zinc sulphate would have great difficulty forming. The results from the experiment are summarized in Figure 4-41. Two distinct kinetic regions were observed. A cobalt removal rate constant of 1225 X 10"^  sec"' was calculated during the first 5 minutes, while a slower region followed with a removal rate of 47 X 10'* sec''. A cobalt removal of 31% was attained after only 5 minutes. A fiirther 15% of total cobalt removed was obtained during the following 85 minutes of the reaction. Despite the apparently high renioval rate constant, it may not be significantly faster than cementation in zinc electrolyte when experimental error is taken into account. 4 Zinc electrolyte' 20 40 60 80 Time (minute) Figure 4-41. Comparison Between Cobalt Removal Profiles in Zinc Electrolyte and in an Initially Zinc-Free Solution. Initial Conditions: 4g/L zinc dust loading, 26 mg/L Co, 46 mg/L Cu, 1.5 mg/L Sb, 73°C, naturd pH. A rise in slurry potential and solution pH (Figures 4-42 and 4-43, respectively) coincides with the cessation of the cobalt cementation reaction, as observed vAth experiments performed with clarifier overflow. The slow rates observed after 5 minutes can be attributed to the formation of a passivating zinc hydroxide, and not a basic zinc sulphate salt barrier, on the surface. The E^-pH page 69 diagram for the Zn-HjO system illustrates (Figures 2-6) that the formation of a zinc hydroxide passivating film on the surface of the zinc dust is thermodynamically possible in the absence of large quantities of SO/'. The result of this experiment shows that achieving both high rates and good removal of cobalt fi-om solution with Cu-Sb activation depends on controlling passivation of the anodic and cathodic surface sites. 4.8 • 4.6 • 4.4 -4.2 • Q. 4 • 3.8 3.6" O A • ^•'^0 ^ ^ -^ • ^ ^ ^ ^ 1 1 . 1 , 1 . 1 . 20 40 60 80 1C Time (minutes) 20 40 60 80 100 Time (minutes) Figure 4-42. pH Profile for a Cementation Figure 4-43. Slurry Potential Profile for a Experiment using Initially Zinc-Free Cementation Experiment using Initially Solution. Initial Conditions: 4g/L zinc Zinc-Free Solution. Initial Conditions: 4g/L dust loading, 26 mg/L Co, 46 mg/L Cu, 1.5 zinc dust loading, 26 mg/L Co, 46 mg/L Cu, mg/L Sb, 73°C, natural pH. 1.5 mg/L Sb, 73T, natural pH. 4.9 EFFECTS OF ORGANIC ADDITIVES Most of the organics added to zinc plant solutions and slurries are assumed to degrade with time and temperature or adsorb on precipitates before cementation. However, there is evidence that some organics may persist into the purification circuit and cause problems during cobalt cementation. A preliminary survey was conducted on the organic additives that are used in zinc plants and their effect on cobalt cementation kinetics. There are a number of possible ways that organics may interfere with cobalt removal which may include agglomeration and chelating effects. The most probable mechanism is that of surface page 70 adsorption because of the strong inhibiting effects that are possible with extremely small quantities of organics in solution. Animal glue, Percol 351 and lignin sulphonic acid were examined. The results are summarized in Table 4-6 and Figures 4-44, 4-45 and 4-46. It was found that even in the presence of 2 mg/L of organic, there was a significant reduction in the removal of cobalt by zinc dust cementation. Even though the plant applications for these organics are very different, their ability to affect cobalt cementation were similar. These results are consistent with the results published by Houlachi et al [49] who performed experimental work with animal glue, Percol 156 and Saponin and found that small amounts of organics (as low as 0.5 mg/L) can significantly decrease the cobalt cementation rate. .e 3 c o O 5 ^ 0 mg/L Animal Glue 200 mg/L 40 60 Tim e (m inutes) 100 Figure 4-44. The Effect of Animal Glue on Cobalt Cementation. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, 73°C, natural pH page 71 ^ 3 c O ^ 2 O 0 mg/L Percol 20 40 60 Time (minutes) 100 Figure 4-45: The Effect of Percol 351 on Cobalt Cementation. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, 73°C, natural pH S 3 c O "o 2 O . 0 ---mg/L . * 1 L ignin - ^ — Sulph( _ e ^ — 1  1 anic Acid y 2 mg/L \ _ 20 mg/L / 200 "^T^  ^ T , 1 mg/L —© L« 20 40 60 Time (minutes) 80 100 Figure 4-46. The Effect of Lignin Sulphonic Acid on Cobalt Cementation. Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L, 73°C, natural pH page 72 Table 4-6: Effect of Organic Additives Initial Conditions: [Co]=26 mg/L, [Cu]=46 mg/L, [Sb]=1.5 mg/L 73°C, natural pH, 4 g/L zinc powder loading Additive Lignin Sulphonic Acid Percol351 Animal Glue No Additives Cone. (mg/L) 2 20 200 2 20 200 2 20 200 0 Cementation Rate Constant, k xlO ** (sec') 316 n/a n/a 328 n/a n/a 252 n/a n/a 788 Normalized Rate Constant 0.4 n/a n/a 0.42 n/a n/a 0.32 n/a n/a 1 [Co] at 90 minutes (mg/L) 4.7 20 22 4.6 20 21 6.3 22 22 0.72 4.10 COMPOSITIONAL and MORPHOLOGICAL ANALYSIS OF CEMENTATION RESIDUES As described in the literature review, the composition and area of the cathode surface both play significant roles in cementation. Microscopic analysis of the cementation residues is important in establishing the correlation between morphology, composition, and the kinetics of cobalt reduction. SURFACE MORPHOLOGY Surface morphology plays an important role in establishing the available cathodic surface area for cobalt deposition and, therefore, affects the observable kinetic rate for cobalt reduction. Selected samples were placed under the SEM and the surface morphologies produced during the course of the cementation reaction were examined. page 73 Unactivated Cobalt Cementation using Zinc Dust: The surface characteristics of the unactivated sponge were consistent throughout the experiment. Microscopic examination revealed that the surface remained relatively smooth and featureless (Figure 4-47). No discernible deposit developed on the surface. This observation correlates well with the poor cobalt removal kinetics obtained under unactivated conditions. Elemental analysis by Energy Dispersive Spectroscopy of the surface of the cementation residue substantiates this observation. The analysis revealed that the surface consisted essentially of zinc, with no detectable levels of Co, Cu or Sb. Figure 4-47. Surface Morphology of an Unactivated Zinc Dust Particle after 90 Minutes. Initial Conditions: [Co]= 26 mg/L, [Cu]= 0 mg/L, [Sb]= 0 mg/L, 73 °C, natural pH, 4 g/L zinc powder loading page 74 Antimony Activation: Examination of the surface during antimony-activated cementation shows that some metal deposits on the surface of the zinc dust very rapidly and would appear to be stable against dissolution. During the first 5 minutes of the reaction, the surface of the residue does not show any distinct features unless extremely high magnification is used to observe selected localized regions. Some of these regions had a "cuspy" type deposit on the surface, but these local features gradually disappeared giving way to a more nodular morphology after 15 minutes. The outer surface of the residue appears to be very porous and the surface area seems to be enlarged. Even though the shell surrounding the residue was found to contain various amounts of Co, Zn and Sb, it is not obvious whether cobalt and zinc deposit along with antimony or on top of antimony during cementation. (a) 5 minutes (b) 60 minutes. Figure 4-48. Surface Morphology of an Antimony-Activated Zinc Dust Particle. Initial Conditions: [Co]= 26 mg/L, [Cu]= 0 mg/L, [Sb]= 1.5 mg/L, 73°C, natural pH, 4 g/L zinc powder loading page 75 Copper Activation: Unlike Sb-activation, the Cu-activated zinc powder began to develop general surface features within the first 5 minutes. Copper precipitated onto the zinc powder rapidly and after 15 minutes a distinct surface morphology was established. During the early stages of the cementation process (5 minutes), the deposits were spaced apart and possessed fine feathery and columnar characteristics. By the end of cementation (90 minutes), the surface had a roughened pebbled appearance with many nodular bumps. The impurity shell after 30 minutes of Cu-activation looked similar to that of Sb-activation; the shell was porous and had a slightly pebbled surface. However, the deposits in the case of Cu-activation were raised fi^om the surface significantly more and the pebbled features looked more roughened. An elemental spot analysis of this shell was found to contain Co, Cu and Zn in various amounts, but primarily Zn. Figure 4-49. Surface Morphology of a Copper-Activated Zinc Dust Particle after 60 Minutes. Initial Conditions: [Co]= 26 mg/L, [Cu]= 46 mg/L, [Sb]= 0 mg/L, 73°C, natural pH, 4 g/L zinc powder loading page 76 Cu-Sb Activation: For the Cu-Sb activated sponge, the surface morphology was, for the most part, similar to the Cu-activated situation. No dendrites were observed on the surface, and in general, the impurity shell had numerous holes and gaps throughout. A pebbled surface developed after 15 minutes of contact and persisted with no significant changes for the remainder of the test. It is difficult to assess how much of an area change is present as a result of the formation of this impurity shell for cobalt reduction, but microscopic observations suggest that this area probably does not change significantly with normal additions of activators. i t - i if. if ^ ^ , A; Figure 4-50. Surface Morphology of an Sb-Cu Activated Zinc Dust Particle after 90 Minutes. Initial Conditions: [Co]= 26 mg/L, [Cu]= 46 mg/L, [Sb]== 1.5 mg/L, 73°C, natural pH, 4 g/L zinc powder loading page 77 Cu-Sb Activation under High Copper Concentrations: After examining a large number of photographs of cementation sponge under various test conditions, it was found that they all had approximately similar surface morphologies. However, varying the copper concentrations in the presence of either 1.5 or 3.0 mg/L antimony gave distinct morphological features. Increasing the concentration of copper in the initial solution, increased the tendency to form dendritic features on the impurity shell. These dendrites were observable after 15 minutes, and persisted for the duration of the test. These formations would appear to have a role in increasing the available cathode surface area and improving the cobalt reduction rate. Figure 4-51. Surface Morphology of an Sb-Cu Activated (High Copper) Zinc Dust Particle after 30 Minutes. Initial Conditions: [Co]= 26 mg/L, [Cu]= 92 mg/L, [Sb]= 1.5 mg/L, 73°C, natural pH, 4 g/L zinc powder loading. page 78 CROSS-SECTIONAL ANALYSIS Cross-sections through the cementation residues were analyzed by SEM and WDX. Impurities were deposited from solution onto a thin and porous shell surrounding the dissolving zinc powder (Figure 4-52). The composition of this impurity shell should help to determine what phases are present that catalyze the cobalt cementation reaction. Compositional analysis of localized areas on the shell and the core of these samples are summarized in Table 4-7. Figure 4-53a represents the positioning of this analysis. Figures 4-53b, 4-53c, 4-53d and 4-53e represent compositional scans for zinc, cobalt, copper and antimony, respectively, along a line traversing across the cross-section of the cement residues. The results suggest that antimony, cobalt and copper form an alloy with zinc. The analysis indicated that cobalt and copper were present in a ratio ranging from approximately 1:1 to 2:1 and that zinc was always present throughout the thickness of the impurity shell. It is a reasonable suggestion that some zinc redeposition takes place on the surface [56]. The phases present may consist of either alloys or mixed phases of the analysed elements. Their exact nature could not be determined due largely to the resolution limits of the micro-analytical technique. page 79 Table 4-7: Summary of WDX Analysis of Cementation Residues [Cu] mg/L 0 0 46 46 46 92 46 Initially Zinc Free Solution 46 46 [Sb] mg/L 0 1.5 0 1.5 3.0 3.0 1.5 1.5 3,0 Zn at% 100 98.6 82 87.3 92.5 72 99.2 70.2 88.4 Cu at% 0.0 0.0 16.4 6.1 3.62 14 0.5 21.7 6.9 Sb at% 0.0 0.15-0.24 0.0 0.25 0.17 6.4 0.07 0.62 0.5 Co at% 0.0 1.3 1.3 6.4 3.7 7.1 0.13 8.1 4.2 Position shell shell shell shell shell shell dendrites dendrites Figure 4-52. SEM Micrograph of Cross-sectioned Cementation Residue, X 1500 Initial Conditions: [Co] = 26 mg/L, [Cu] = 46 mg/L and [Sb] = 1.5 mg/L, 73°C. Figure 4-53a. Positioning of Compositional Line Scan Through a Cementation Residue, X7000 Initial Conditions: [Co] = 26 mg/L, [Cu] = 46 mg/L and [Sb] = 1.5 mgA., 73°C. page 80 With antimony activation, the impurity shell was made up of a relatively small amount of Sb compared to the amount of cobalt deposited. This would also indicate that Sb is deposited in its metallic form on this impurity shell along with metallic forms of cobalt and zinc. With copper activation, a much larger amount of copper was deposited onto the impurity shell compared to Sb activation. However, the atomic percent of cobalt deposited in the shell in both cases was very similar. This is inconsistent with the result that a much faster removal rate for cobalt was found with Sb-activation compared to Cu-activation and that much smaller levels of Sb were required. It appears that the most probable surface is indeed an Sb-Zn surface for Sb-activation based solely on the composition of the shell and either a Cu or Cu-Zn surface for Cu-activation. With the combined Cu-Sb activated situation, much higher concentrations of cobalt were deposited onto this impurity shell. Analysis of the impurity shell by WDX revealed that the composition of the shell consists of Zn, Co, Cu and Sb. Cu-Sb activation with high copper concentrations also revealed an impurity shell containing detectable levels of Zn, Co, Cu and Sb. However, the concentration of copper in the shell was much higher than with activation using the normal initial concentration of activators in solution. From the SEM data alone, it remains unclear how the activators promote cobalt removal. Though the kinetic data would suggest that antimony and copper each play a specific role in the activation process, more substantial evidence is needed to establish the mechanism of cobalt deposition and the role of zinc, since pre-activation does not affect the kinetic rates for cobalt removal. It is unclear what form of the substrate is responsible for cobalt deposition and the options remain as either Sb, Cu, Cu-Sb, Sb-Zn, Cu-Zn, or Cu-Sb-Zn alloys or mixed metallic phases. page 81 resrn impurity shell &gap zinc dust core_ t -52St|* impurity shell &gap zinc dust core_ t Figure 4-53b. Compositional Line Scan for Cobalt in the Cementation Residue. Initial Conditions: [Co] = 26 mg/L, [Cu] = 46 mg/L and [Sb] = 1.5 mgA., 73°C. Figure 4-53c. Compositional Line Scan for Zinc in the Cementation Residue. Initial Conditions: [Co] = 26 mg/L, [Cu] = 46 mg/L and [Sb] = 1.5 mgA., ITC. impurity shell resin, & gap zmc dust core_ t resm impurity shell zinc dusti core. t Figure 4-53d. Compositional Line Scan for Copper in the Cementation Residue. Initial Conditions: [Co] = 26 mg/L, [Cu] = 46 mg/L and [Sb] = 1.5 mg/L, 73°C. Figure 4-53e. Compositional Line Scan for Antimony in the Cementation Residue. Initial Conditions: [Co] = 26 mg/L, [Cu] = 46 mg/L and [Sb] = 1.5 mgA., 73T. page 82 5 CONCLUSIONS Without the addition of activation agents, cobalt removal with only zinc dust proceeds at a very slow rate. Batch testing has shown that antimony activation performed better than copper activation, but the addition of both copper and antimony together increased the rate of cobalt removal significantly. These activation agents were found to precipitate onto the zinc powder surface within the first 10 minutes of the cementation reaction, and seem to play a role in controlling hydrogen evolution, maintaining the stability of the cobalt deposit and increasing the cathode surface area for cobalt reduction. Increasing the copper concentration increased cobalt removal in the presence of antimony, and the best cobalt reduction rates were obtained with 92 mg/L copper and 3.0 mg/L antimony (the highest concentration of activators tested). The cobalt removal rate constant using zinc dust activated with antimony and copper is influenced by the available zinc dust surface area, activator concentration, pH, temperature, and the presence of organics. Cobalt removal was improved with increasing zinc dust loading as this effectively increases the available surface area for the reaction. However, this increase in available surface area must occur in conjunction with an adequate amount of activators to promote cobalt cementation. The activation roles of antimony and of copper appear to be quite different. Batch testing suggests that antimony probably works by increasing the hydrogen overpotential on the surface, thereby decreasing the magnitude of this competitive reaction and/or by improving the stability of the cobalt deposit towards redissolution. In contrast, copper seems to form the cathode surface for cobalt deposition because, on its own, it had little effect on either the hydrogen overpotential or the cobalt deposition overpotential. The cemented copper surface may also act as the site for hydrogen evolution due to the surface's low hydrogen overpotential, and because it is the most likely site for cobalt deposition. pH plays a significant role in cobalt removal since the formation of hydroxides or basic zinc sulphate salts at elevated pH values can restrict the movement of ions to and fi-om the bulk and the reaction surface. It was found that an optimum pH range lies between 4.2 and 4.4, but "natural" pH experiments resulted in the fastest cementation rate constants. A low pH page 83 detrimentally affects cobalt cementation since the activity of the competing hydrogen reaction was increased. At high pH levels, the reaction was impeded most likely by the formation of a passivating film on the surface; either basic zinc sulphate, or zinc hydroxide. Increasing temperature improved the cobalt removal rate primarily by increasing the energy available to overcome the activation energy for the reaction. However, increasing temperature also increases the solubility of basic zinc sulphate, making it more difficult to form on the surface of the zinc powder and, therefore, postponing the onset of surface passivation. The best cobalt removal rate constants and lowest residual cobalt concentrations at 90 minutes, were obtained at 90°C, the highest temperature tested. An apparent activation energy of 86.6 kJ/mol was calculated for this reaction, which is consistent with the established range for activation or chemically rate controlled processes. The presence of the organic agents, animal glue, Percol 351 and lignin sulphonic acid can dramatically reduce the cobalt cementation rate, and the total cobalt removed. Very low concentrations (2 mg/L) resulted in a dramatic reduction in the cementation rate constant, suggesting a surface adsorption mechanism. Measuring the potential of the slurry during the course of cementation may be useful for determining the point of basic zinc sulphate or zinc hydroxide formation on the surface of the zinc dust particle. In practice, this may be usefiil as an indicator for pH adjustment or the point at which the cementation reaction ceases. WDX analysis indicated that an alloy or a mixed phase of zinc, cobalt, copper and antimony is present in the impurity shell surrounding the dissolving zinc powder core. It was surprising to discover that a significant amount of zinc deposits onto this shell during cobalt cementation. It is believed that this zinc alloy is catalytic towards cobalt removal by lowering the cobalt deposition overpotential and by increasing the hydrogen overpotential. This finding only highlights the complexity of the cementation system based on antimony and copper. It is entirely possible that the precipitation of Zn(OH)2or Zn(OH)2-ZnS04 onto the shell may also contribute to the high zinc content resulting fi-om the analysis since the method does not distinguish between zinc metal and zinc compounds, and that the conditions (high pH and high zinc concentrations) near the surface of the shell would promote the formation of these compounds. page 84 The Cu-Sb activation mechanism remains unclear. It has not been possible by WDX analysis to distinguish between alloys and separate single metal phases, or to determine if antimonide phases of copper or cobalt have been formed. This has been attributed primarily to the thinness of the impurity shell on the surface of the cementation residues. Batch tests have revealed that antimony plays an integral role in the activation of the zinc dust surface (either by suppressing hydrogen evolution on the copper surface, or by improving the stability of the cobalt deposit towards redissolution) and only a small amount of antimony is required to activate the system. How antimony performs its role during activition will be one of the keys to understanding the process in future studies. page 85 6 RECOMMENDATIONS Even though batch testing using screened zinc dust powders were very successful for determining the effects of temperature, pH, loading density, activation agent concentrations, and the effects of organic compounds in the electrolyte on the cobalt removal rate, analysis of the cementation residues by SEM provided only limited insight into the activation mechanism. It is therefore recommended that future work in this area should be divided between an academically oriented study on the mechanism of surface activation with copper and antimony, and an industrially oriented study with an emphasis on improving the cobalt removal rate, reducing zinc dust usage and increasing the stability of the deposited cobalt. This project has demostrated that the mechanism for cobalt removal is complicated, and that the role played by antimony and by zinc is not well understood. Further mechanistic studies using zinc dust would not be particularly helpfiil unless a better surface analysis technique (other than WDX, EDS, or XPS) can be found which will allow for compositional analysis along the thickness of the impurity shell surrounding the cementation residue. An alternative is to perform cobalt removal using more elaborate electrochemical techniques (such as rotating discs, galvanic cells, or flat plate electrodes), and by analysing the deposits on these electrodes by conventional means. There is considerable scope for improving the cobalt cementation process. Increased kinetics and reduced zinc dust consumption can be attained by increasing cathode surface area, improving deposit stability, and decreasing hydrogen evolution. Some suggestions for future studies using zinc dust systems may include studying the effects of different zinc dust morphologies, alloying additions to the zinc dust, recycling cementation residues and their activation, altering the impurity shell composition, applying potential control to the system, and by modeUng the process. page 86 1 HafiFord, B.C., Pepper, W.E., and Lloyd, T.B., Zinc Dust and Zinc Powder: Their Production, Properties and Applications, International Lead Zinc Research Organization, Inc., New York, New York, USA, 1982. 2 Bolton, G., 1987, Zinc Hydrometallurgy, m Hydrometallurgy: Theory and Practice, Course Notes (Eds. Dreisinger, D., and Peters, P.), Department of Metals and Materials Engineering, University of British Columbia, Vancouver, Canada 3 Gupta, C.K. and Mukherjee, T.K., Hydrometallurgy in Extraction Processes, Volume J., CRC Press, Boca Raton, Florida, USA, 1990. 4 Yaroslavtsev, A.S., Getskin, L.S., Usenov, A.E., and Margulis, E.V., "Behaviour of Impurities when Precipitating Iron from Sulphate Zinc Solutions", TSVETNYE MET ALLY/ Non-Ferrous Metals, pp. 40-41. 5 Ingraham T.R. and Kerby R., "Kinetics of Cadmium Cementation on Zinc in Buffered Sulfate Solutions", Transactions of the Metallurgical Society of AIME, vol. 245, 1969, pp. 17-21. 6 Kerby, R.C. and Ingraham, T.R., "Effect of Impurities on the Current Efficiency of Zinc Electrodepostion", Research Report R243, Dept. of Energy, Mines and Resources, Mines Branch, Ottawa, Canada, 1971. 7 Aitkenhead, William C, "Injurious Impurities in Zinc Electrolysing solutions and their Removal", Paper presented at the Metallurgical Section of the Northwest Mining Association, Spokane, Washington, December 2-3, 1960. 8 Clifford J. Krauss, "Effects of Minor Elements on the Production of Electrolytic Zinc from Electrolytic Zinc from Zinc Sulphide Concentrates", ZINC '85, Proceedings of International Symposium on Extractive Metallurgy of Zinc, (ed. K. Tazawa), Published by The Mining and Metallurgical Institute of Japan (MMIJ), October 14-16, 1985, Tokyo, Japan, pp. 467-481. 9 Ohyama, S., and Morioka, S., "Effect of some impurities on the electrowinning of zinc", ZINC '85, Proceedings of International Symposium on Extractive Metallurgy of Zinc, (ed. K. Tazawa), Published by The Mining and Metallurgical Institute of Japan (MMIJ), October 14-16, 1985, Tokyo, Japan, pp. 219-234. 10 Morrison, R.M., MacKinnon, D.J., Uceda, D.A., Warren, P.E., and Mouland, J.E., "The Effect of some Trace Metals Impurities on the Electrowinning of Zinc from Kidd Creek Electrolyte", Hydrometallurgy, 29, (1992), pp. 413-430. page 87 11 Connolly, M.L., Honey, R.N., and Krauss, C.J., "High-Productivity Zinc Electrowinning Plant", CIMBull, 70, (1977), pp. 144-151. 12 Knobler, R.R., Moore, T.I. and Capps, R.L., "The New Zinc Electrolysis and Residue Treatment Plant of the National Zinc Company", Erzmetall, 32, (1979), pp. 109-116. 13 Andre, J. A. and Delvoux, R.J., "Production of Electrolytic Zinc at the Balen Plant of S.A. Vielle-Montagne", World Symposium on Mining and Metallurgy of Lead and Zinc, C.H. Cotterill and J.M. Cigan, editors, AJME, New York, 1970, pp 178-197. 14 Morrison, R.M., MacKinnon, D.A., Uceda, P.E., Warren, P.E., and Mouland, I.E., "The ElBfect of Some Trace Metal Impurities on the Electrowinning of Zinc from Kidd Creek Electrolyte", Hydrometallurgy, 29, (1992), pp. 413-430. 15 Hamilton, Ernie. R., Retreatment of Zinc Plant Purification Precipitate, pp. 172-186. 16 Landucci, L.P., Cominco Technical Report #731-10, Sept 1 1992. 17 Mellor, J.W., A Comprehensive Treatise on Inorganic and Theoretical Chemistry, Vol 3, Longmans, Green and Co., London, 1923. 18 Weast, R.C (editor), CRC Handbook of Chemistry and Physics, 62nd Edition, CRC Press Inc., Boca Raton, Florida, USA, 1981. 19 Lawson, F., and Nhan, L.T., Kinetics of Removal of Cobalt from Zinc Sulphate Electrolytes by Cementation, Hydrometallurgy 81, Proceedings of the Society of the Chemical Industry Symposium, June 30 - July 3, 1981, Manchester, England, pp. G4-1 - G4-10. 20 Lawson, F. and Strickland, P.H., "Measurement and Interpretation of Cementation Rate Data", International Symposium on Hydrometallurgy, Chicago, Illinois, USA, 1973, Eds. Evans, D.J.I., and Shoemaker, R.S., p. 293. 21 Mackinnon D.J., "Kinetics of Cobalt Cementation on Zinc", Research Report R259, Dept of Energy, Mines and Resources, Mines Branch, Ottawa, Canada, 1973. 22 Adams, R.W. and Chapman, R.L., "New Developments in the Zinc Dust Purification of Zinc Plant Electrolyte", The Aus.I.M.M. Melbourne Branch, Symposium on "Extractive Metallurgy", November, 1984, pp. 169-178. 23 Dodson, F., "Method of Purifying Zinc Sulphate Solutions", Australian Patent 465,511., 1972. 24 Harriot, P., "Mass Transfer to Particles Part I., Suspended in Agitated Tanks", AIChEJ., 8 (1), 1962, pp. 93. page 88 25 Chemical Engineers' Handbook, 50th Edition, Editor: Robert H. Perry, McGraw-Hill Book Company, New York, 1984. 26 Kerby, R.C., "Influence of Atomized Zinc Alloys and of Antimony Valence States on Impurity Removal from Zinc Electrolyte by Cementation", Cominco Ltd., Tech Research Report, 780-8-02, 1983. 27 Yunus, M., Capel-Boute, C , and Decroly, C , "Inhibition Effect of Zinc on the Cathodic Deposition of Cobalt - 1 . Electrochemical and Structural Observations in Sulfate Solutions", Electrochim. Acta 10, pp. 885-900, 1965. 28 Xiong Jiang, and Ritchie I.M., "An Electrochemical Study of the Inhibition of the Co(n)/Zn Cementation Reaction", Proceedings of the First International Conference of Hydrometallurgy (ICHM '88), editors Yulian and Jiazhong, X., Pergamon Press, New York, USA, 1989, pp. 632-636. 29 Aitkenhead, William C, "Removal of Cobalt from Zinc Electrolysing Solutions by Antimony Metal", Paper presented at the Pacific Northwest Metals and Minerals Conference, AIME, April 26-28, 1962. 30 Miller, J.D., "Solution Concentration and Purification", Rate Process Extr. Metall, editors Sohn, H.Y. and Wadsworth, M.E., Plenum, New York, N.Y., USA, 1979, pp. 197-244. 31 Blaser, M. and O'Keefe, T.J., "Cementation of Cd from ZnS04 Solutions Using Zn and Mn Powder", Hydrometallurgy - Research, Development and Plant Practice, editors K. Osseo -Asare and J.D. Miller, Conference Proceedings of the Metallurgical Society of AIME, Atlanta, Georgia, USA, 1983, pp. 587-601. 32 Kerby, R.C., "Manganese Powder Purification of Zinc Electrolyte", Cominco Ltd., Technical Research Report, Zinc Plant Section 516, Report No. 2, 1976. 33 Ingraham T.R. and Kerby R., "Kinetics of Cadmium Cementation on Zinc in Buffered Sulfate Solutions", Transactions of the Metallurgical Society of AIME, vol 245, 1969, pp. 17-21. 34 Pourbaix, M., Atlas of Electrochemical Equilibria in Aqueous Solutions, National Association of Corrosion Engineers, Houston, Texas, USA, 1958, p. 112-121. 35 Pletcher, D. and Walsh, F.C., Industrial Electrochemistry, 2nd Edition, Chapman and Hall, New York, N.Y., USA, 1990, pp. 38-44. 36 Thiel, A., and Hammerschmidt, W., Z. anorg. allgem. Chem. 132, p. 15-35 (1923). 37 Tromans, D., MMAT 550 course notes. University of British Columbia, 1993. page 89 38 Fisher-Bartelk, C , Lange, A., and Schwabe, K., "Klarung der Ursachen Fur Die Schwierigkeiten Der Kobalt - Und Nickelentfemung Aus Zinksulfatelektrolyten Durch Zmkstauhzementation", Electrochimica Acta, 14, 1969, pp. 829-44. 39 Straumanis, M.E., and Fang, C.C, "The Structure of Metal Deposits Obtained by Electrochemical Displacement upon Zinc", J. Electrochem. Soc, 98, pp. 9-13, 1951. 40 Power, G.P., and Ritchie, M., "Metal Displacement Reactions", Modem Aspects of Electrochemistry, vol. II, p. 199, 1975. 41 DeBlander, F., and Winand, R., "Influence de L'Antimoine et du Cuivre sur la Cementation du Cobalt par le Zinc", Electrochimica Acta, 20, pp 839-852, 1975. 42 Fontana, A. and Winand, R., "Chemical Precipitation of Cobalt by MetalUc Zinc for Purification of Solutions Used for Production of Electrolytic Zinc. I. Effect of Antimony", Metallurgie XI (3), p. 162, 1971. 43 Goubot, Mireille, Libert Michael, Barbey G., and Caullet C , "Etue Voltamperometrique do la Cementation du Cobalt par le Zinc", C.R. AcadSci Paris, 1282, May 10, 1976, p. 883-886. 44 Kroleva, V., "Copper Antimonide as Activator for Cobalt Cementation with Zinc Dust", Metalurgiya, Sofia, 1980. 45 Gravel A., Wiss VeroflFentlich Seimera, Konzem, 13, 1934. pp. 61-67. 46 Fountoulakis, S.G., "Studies on the Cementation of Cobalt with Zinc in the Presence of Copper and Antimony Additives", PhD Thesis, Columbia University, 1983. 47 Tozawa, K., Nishimura, T., Akahori, M., and Malaga, M.A., Comparison Between Purification Processes for Zinc Leach Solutions with Arsenic and Antimony Trioxides, Hydrometallurgy, 1992, vol. 30, pp. 445-461. 48 Centnerszer, M. and Gonet, F., J. Am. Chem. Soc, vol. 60, pp. 435-440, 1938. 49 Houlachi, G., Belanger, F., and Principe, F., Effect of Organic Additives on the Kinetics of Cobalt Purification, Proc. Int. Symp. Electrometall Plant Pract. (Eds. Claessens, R., Harris, G.B.) pp. 177-190, 1989, Pergamon, New York, NY. 50 Krauss, C.J., "Zinc Dust Purification - Sulphide Leach Plant", Cominco Ltd., Technical Report, Zinc Plant Section 55, Report Number 28, January 21, 1965. 51 Lee, E.C., Lawson, F., and Han, K.N., "Precipitation of Cadmium with Zinc fi-om Dilute Aqueous Solutions: 1- Under an Inert Atmosphere", Trans. InstnMin. Metall, (Sect. C: Mineral Process. Extr. Metall.), 84, 1975, C87-91. page 90 52 Lee, E.C., Lawson, F., and Han, K.N., "Precipitation of Cadmium with Zinc from Dilute Aqueous Solutions: 2- Under Oxygen-Containing Atmospheres", Trans. InstnMin. Metall, (Sect. C: Mineral Process. Extr. Metall.), 84, 1975, C149-153. 53 Tromans, D., private communications. University of British Columbia, 1993. 54 Dreisinger, D., private communications. University of British Columbia, 1993. 55 MacKinnon, D.J., "Removal of Cobalt from Synthetic Zinc Sulphate Electrolyte by Cementation with Zinc Dust", Mines Branch Research Report R264, Ottawa, Canada, 1973. 56 Straumanis, M.E. and Fang, C.C, "The Structure of Metal Deposits Obtained by Electrochemical Displacement upon Zinc", J. Electrochem. Soc, vol. 98, pp. 9-13, 1951. 57 Kneule, F . , Chem Ing. Tech, vol 28, p. 222, 1956. 58 Bindell, J.B., "Elements of Scanning Electron Microscopy", Advanced Materials and Processes, March 1993. pp. 20 - 27. page 91 APPENDIX A: Cobalt Cementation Rate Calculations Kinetic Calculation for Batch Tests: ^ = -kC (First Order Equation) (a-1) at mass balance will give: you assume that the volume and the surface areas do not change, then the following equation would be adequate, and •"'^ ]= -kco t (a-3) by plotting equation (2) versus t, we can obtain a value for k, where k is the 1st order kinetic constant, giving the equation: ^^  ^ = -kco (a-4) At kco = k ( 0 ) (a-5) According to equation a-5, the experimentally determined cobalt removal rate can be strongly influenced by the surface area, A. This surface area can be increased by using a finer zinc dust for the same solids loading, increasing the solids loading in the reactor, and modifying the morphology of the zinc dust. page 92 APPENDIX B: Zinc Dust Surface Area Estimation Surface Area of Zinc Dust: Mesh Size 10 50 100 140 200 270 325 Maximum Opening Diameter (microns) 2135 337 174 126 91 66 57 By assuming a sphere and an "average" sized particle, the surface area of a single zinc dust particle can be estimated for the +100-140 mesh screened particles: Surface Area/particle = 47tr2 = 47C(125 * 10"*/2)^  = 4.91 x 10'^  m^  per particle The total surface area will be: Volume for single sphere = |icr^= 1.02 x 10"'^  m^  Weight of single particle = (vol.of single particle)*(density = 7.1g/cm )^ = 7.26 x 10"* g/particle Number of Particles per gram = 1.38x10^ particles/gram of sample Surface area = surface area per particle * number of particles = 6.76 x 10^ m^ per gram This corresponds to the BET analysis in a "poor" manner. The Quantasorb surface area analysis method (BET) is based on the idea that when a gas flow of nitrogen and helium passes across a solid held at liquid nitrogen temperature, nitrogen will physically adsorb onto the surface of the solid, and desorption occurs when the sample is warmed up. The surface area of the solid may be calculated once the quantity of nitrogen adsorbed is known as well as the cross-sectional area occupied by one molecule of nitrogen. page 93 Tests were performed using BET analysis for zinc dust of various mesh sizes. The following results using 4 grams of leaded zinc powder collected from the dewatering cones (wet atomized) on May 19, 1993: mesh size -200, +270 -140, +200 -100, +140 -50, +100 average area for 10% (cmVg) 1900 1548 668 655 average area for 30% (cmVg) 1938 1500 785 803 page 94 APPENDTX C: Criss and Cohhle Method for Estimating Activity of Ionic Species at Elevated Temperatures The Nemst Equation can be used to determine changes in the potential of a reaction due to temperature and activity: T7 T^o , RTi [reactant]" r- j •^ E = E° + —=- In ; 5f for reduction reactions. nF [product] The Criss and Cobble method is used to compensate for changes in activity at elevated temperatures for ionic species in solution. For a single ionic species, the Criss and Cobble method can be applied: G^2 = G^, + ATC^ITJ - AT(S?)abs-T2C;iT?ln ( 1 ^ ) Cplj^ =Average Cp between Tj and T2 for species. 7^|T2 _ a T - ( S ^ 9 8 ) a b s ( l - Q - b T ) _ Cp I Tj - J ^ - ttT + PT(a298)abs 1"(298 (S^98)abs = (S298)eo„v-5-0(z) ttT = —-p—r [cal/deg mol] p ^ ^ J L O - b j ) fj^al/degmol] In [298 For simple cations (Na"", Mg^*), the Criss and Cobble method uses empirically derived values of T CC) ar bx 25.000 0.000 1.000 60.000 3.900 0.955 100.000 10.300 0.876 150.000 16.200 0.792 200.000 23.300 0.711 The values at 73°C were determined by linear interpolation, and were found to be equal to: ax = 6.29518,bT = 0.92345. page 95 These values were used for the subsequent calculations for entropy, heat capacity and Gibbs Free Energy. The concentrations of metal species in solution were calculated to be representative of the levels found in the purification plant and have the values: • Zn = 150g/L = 2.29 moles/L • Cu = 33mg/L = 5.19X10^ moles/L • Co = 20mg/L = 3.39X10-^ moles/L • Sb = 1.5mg/L = 1.23X10-^  moles/L element Zn metal Zn2+ Cu metal Cu2+ Co metal Co2+ Temp (C) 73 73 73 73 73 73 (cal) S conv 9.95 -26.8 7.92 -23.8 7.18 -27 (cal) Sabs -0.05 -36.8 -2.08 -33.8 -2.82 -37 (cal) stdG.Tl 0 -35140 0 15660 0 -13000 Cp 42.18 61.01 43.22 59.48 43.6 61.12 (cal) stdG,T2 -152 -33598 -59.1 17063 -24.8 -11448 (cal) stdG,rxt 33445 -17123 11424 (volts) Erxt,T2 -0.73 0.37 -0.25 (volts) Erxt,Tl -0.76 0.34 -0.28 The calculated values are tabulated in Chapter 2, Table 2-2. page 96 APPENDIX D: Listing of Thermodynamic Data Used to Generate E^ -^pH Diagrams Temperature -Delta G of H20 Dielectric Constant -Ion Strenght •• pH min V min 0.000 -2.000 Element Molality Zn l.OOOE+00 Species Zn Zn (OH) 2 HZn02 <-a) Zn(+2a) Zn02(-2a) Zn(0H)2(a) ZnOH(+a) Zn(0H)3(-a) Zn(OH)4(-2a) 73.000 C -54.833 kcal/mol 62.843 0.000 pH max V max 14.000 2.000 Pressure l.OOOE+00 Delta G kcal/mol 0.000 -129.057 -77.671 -34.877 -60.390 -104.269 -66.996 -115.015 -135.132 Temperature Delta G of H20 --Dielectric Constant • Ion Strenght pH min 0.000 V -2 -min 000 Element Molality S 1 Zn 1 Species H2S2 H2S04 H2S04*6.5H20 H2S04«H20 H2S04«2H20 H2S04«3H20 H2S04«4H20 S . H2S(a) HS(-a) H2S03(a) H2S04(a) H2S204(a) H2S2oe(a) HS03(-a) HS04(-a) HS204(-a) S(-2a) S2(-2a) S3{-2a) S4(-2a) S5(-2a) S02(a) S03(a) S04(-2a) S203(-2a) S204(-2a) S20B(-2a) S406(-2a) Zn ZnO Zn (OH) 2 ZnO*2ZnS04 HZn02(-a) Zn(+2a) Zn02(-2a) Zn(0H)2(a) ZnOH('»a) Zn(0H)3(-a) Zn(OH)4(-2a) OOOE+00 OOOE+00 73 -54 62 0 pH max L4.000 000 C 833 koal/mol 843 000 V max 2.000 Pressure l.OOOE+00 l.OOOE-fOO Delta G kcal/mol -4.855 -160.163 -529.227 -220.370 -278.020 -334.545 570.884 0.000 -6.006 4.188 -125.750 -171.358 -46.222 -257.501 -122.399 -175.632 -96.453 22.602 20.976 19.524 18.415 17.577 -70.940 -121.356 -171.395 • -118.655 -137.563 -257.736 -241.712 0.000 -74.871 -129.057 -4B4.483 -77.671 -34.877 -60.390 -104.269 -66.996 -115.015 -135.132 Zinc -Water System Zinc - Sulphate - Water System page 97 APPENDIX E; X-rav Analysis and Resolution: When an electron of an appropriate energy interact with a sample, they cause the emission of X-rays whose energy and relative abundance depend upon the sample's composition. Each element will emit a unique and characteristic pattern of X-rays. The X-rays can be sorted out by two methods: 1) Energy Dispersive Microanalysis: the X-ray emissions are sorted electronically using ENERGY ANALYSER and data manipulation. 2) Wavelength Dispersive Microanalysis: X-ray emissions are sorted by means of a diffraction crystal (separating the X-ray into component wavelengths), which are then detected and measured. When a sample is excited by a beam of electrons, the volume of sample as small as a cubic micrometer and characteristic X-rays are emitted by the excited sample. Planck's equation shows the equivalence of wavelength and energy: 1 _ he where lambda = wavelength of the radiation (angstrom) c = speed of light h = Planck's constant E = energy of the radiation (keV) During electron-sample interactions, a number of emission events may occur: 1) secondary electrons 2) backscattered electrons 3) X-ray Continuum 4) Characteristic X-rays The elemental analysis using the EDS or the WDX, the characteristic X-ray is of interest to us. When a high energy beam interacts with a sample, an electron is ejected from the inner atomic page 98 shell giving an ion in an excited state. Relaxation or a return to its ground state results in the ion giving up energy. This usually occurs when an electron from the outer shell drops into a vacancy in an inner shell. The magnitude of this energy drop is the difference between the vacant shell and the shell contributing the electron, and this energy takes the form of a X-ray. Energy of the X-ray is characteristic of the element. The primary electrons can travel considerable distances into the specimen before they lose enough energy through collisions to no longer be able to excite X-ray emission. This means that a large volume of the specimen will produce X-rays for any position of the smaller primary beam. X-rays travel longer distances (unlike Auger or secondary electrons) through the sample than electrons, and therefore, escape from depths at which the primary electron beam has been widely spread. As a consequence. X-ray signal has poor spatial resolution compared to secondary electrons or backscattered electron signals. A typical value of minimum X-ray spatial resolution is about 1 micron at 30keV irrespective of the beam diameter [58]. Operating Parameters and Procedure: Prior to using the WDX analysis feature of the SEM, the instrument must be calibrated against known standards. The suggested settings are listed in Table, but the operating setting of beam current was a compromise value of 30 nA so that all the selected elements could be analysed together. Suggested WDX Instrument Settings for Analysis of Selected Elements: Element Co Cu Sb Zn Crystal Wavelength Ka Ka LiF Ka Beam Voltage (mV) 30 30 20 30 Beam Current (nA) 4 4 30-40 4 A program was set-up for this procedure under the name: Cominco. WDX analysis requires a flat surface for proper analysis, a 45° tilting of the sample stage, a 35mm working distance, and a beam current of 30 nA that is checked between each analysis using current measurement obtained with the Faraday Cup. Spot analysis were performed on the standard samples for calibration purposes at a magnification of 2000 times. page 99 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0078522/manifest

Comment

Related Items