Carbonylation of Nickel and Iron from Reduced Oxides and Laterite Ore

Home , Carbonylation, Cobalt extraction, Iron pentacarbonyl, Mond process, Nickel tetracarbonyl

The University of New South Wales

Faculty of Science

School of Materials Science and Engineering

Submitted in partial fulfilment of the requirements

for the degree of Doctor of Philosophy

Yongli Cui

December 2015 To my parents and wife COPYRIGHT STATEMENT

'I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only).

I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Date

AUTHENTICITY STATEMENT

'I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.'

Signed ......

Date ...... ORIGINALITY STATEMENT

'I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.'

Signed

Date

II THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: CUI

First name: YONGLI other name/s:

Abbreviation for degree as given In the University calendar: PhD

School: School of Materials Science and Engineering Faculty: Science

Title: Carbonylation of Nickel and Iron from Reduced Oxides and Laterite Ore

Abstract 350 words maximum: The PhD project was undertaken within the ARC Discovery Grant (Project No. DP1094880) which examined a novel approach to processing of laterite ores based on the selective reduction of the ore and extraction of nickel by carbonylation of the selectively reduced ore. The project undertook a systematic study of carbonylation of pure nickel and iron, and carbonylation of selectively reduced laterite ore to develop a further understanding of carbonylation reactions and feasibility of extraction of nickel from laterite ore by carbonylation. The project studied the effects of operational parameters on carbonylation of nickel and iron obtained by reduction of pure oxides and in the selectively reduced Australian laterite ores, including temperature, carbon monoxide pressure, gas flow rate, particle size, and the effect of sulphur-containing catalysts. The non-catalytic carbonylation of nickel at I 00 °C and CO pressnre 27 atm was close to completion in about 5.5 hours. Carbonylation of pure iron was slow; the extent of iron carbonylation at 100 °C and CO gauge pressure up to 55 atm was less than 5.0%. The extent of cobalt carbonylation nnder these conditions was less than 0.5%. Sulphur containing catalysts accelerated the carbonylation reaction. The time for a complete carbonylation of nickel was shortened from 5.5 hours in the non- catalytic reaction to 2 hours in the catalytic carbonylation at 100 °C and CO pressure 14 atm. The extent ofcarbonylation of nickel from the selectively reduced laterite ore at 100 °C and CO pressure 41 atm was below 50% (ore particle size 53-200 µm). The use of catalysts in the carbonylation of selectively reduced ore was inefficient. The rate of reaction increased significantly with decreasing particle size; the carbonylation of nickel in the ore with the particle size 38-53 and 75-90 µm, was close to completion after 4 hours reaction. Results of a systematic study of the carbonylation of nickel and selectively reduced latcrite ores are significant for further understanding of carbonylation reactions. Promising results were obtained for further development of technology of extraction of nickel from late rite ores by the carbonylation process.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only) .

Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and reauire the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS Acknowledgements

This project was made possible by funding from the Australian Research Council ( ARC )and the Commonwealth Scientific and Industrial Research Organisation ( CSIRO ). The project code is DP1094880.

Assistance during this project has come in many forms, both in technical and moral support. I am indebted to my supervisor Professor Oleg Ostrovski and co- supervisor Dr. Guangqing Zhang for believing in me and their advice and support during the project. I would like to thank Dr. Jianqiang Zhang and Sharif Jahanshahi for their advice during project reviews.

The technical staff was invaluable in their help and support for experimental work, Mr. John Sharp for equipment maintenance; Mr. Anthony Zhang and Dr. Rahmat Kartono for lab safety support; Dr. Yu Wang for XRD analysis; Dr. Deming Zhu and Dr. Leah Koloadin for SEM/EDS analysis; Ms. Rabeya Akter and Ms. Dorothy Yu for ICP-OES analysis, Dr. George Yang for metallography and lab assistant; Mr. Danny Kim and Ms. Jane Gao for I.T. support.

The administrative staff was generous with their time and kind works, helping with more than just paperwork, Mrs. Lana Strizhevsky, Ms. Qing Xia and Mrs. Judy Lim for their help with general administrative issues.

I have greatly enjoyed the opportunity to work with all the talented and dedicated people in our group as: Dr. Yan Li, Dr. Xiaohan Wan, Dr. Xing Xing, Mr. Le Yu, Mr. Jun Yang, and Mrs. Jing Zhang. Your encouragement and advice inspired this work.

Finally, I would like to thank all my family members especially my wife Mrs. Jieqing Gan for their mental support and constant encouragement throughout the whole course of my PhD study.

IV Abstract

Australia has abundant deposits of laterite ores which role in production of nickel is increasing with rising demand for nickel and depletion of sulphide reserves.

Laterite ores cannot be efficiently upgraded prior to pyrometallurgical or hydrometallurgical processing what leads to significant challenges in extraction of nickel. The PhD project was undertaken within the ARC (Australian Research Council) Discovery Grant (Project No. DP1094880) which examined a novel approach to processing of laterite ores based on the selective reduction of the ore and extraction of nickel by carbonylation of the selectively reduced ore. Selective reduction of the laterite ore was studied by J. Yang in his PhD project (Yang, 2014).

The ultimate aim of this project was to establish the feasibility of extraction of nickel by carbonylation of selectively reduced laterite ore. The project undertook a systematic study of carbonylation of pure nickel and iron, and carbonylation of selectively reduced laterite ore to develop a further understanding of carbonylation reactions and extraction of nickel from laterite ore by carbonylation.

Specific objectives of this study included:

1) to study the effects of reaction parameters on carbonylation of nickel and iron, including reaction temperature (80-100 oC), carbon monoxide (gauge) pressure (0-

56atm), gas flow rate (0.14-0.50 L·min-1), nickel mass (0.8-3.2g) and particle sizes (0.29

– 2.67 µm), and the effect of sulphur-containing catalysts (sulphur, iron sulphide and hydrogen sulphide);

2) to study the non-catalytic and catalytic carbonylation of laterite ores; the impacts of reduction conditions; and effects of reaction parameters;

3) to develop further understanding of kinetics and mechanisms of carbonylation processes.

Pure metals and nickel-iron mixture were prepared by the reduction of associated oxides by hydrogen at 500oC (gas flow rate 1.0 L·min-1); the degree of reduction of oxides was over 98%. Two types of Australian laterite ores supplied by CSIRO were examined (labelled by CSIRO): BCS ore containing 1.35 wt% Ni, 10.4 wt% Fe and 0.038 wt% Co, V particle size 53-200 µm; and MIN ore with particle sizes 38-53 µm, 75-90 µm, 140-200 µm, and 355-495 µm, containing 1.68-2.37 wt% Ni, 8.90-11.8 wt% Fe and 0.135-0.144 wt% Co). Selective reduction of laterite ores was conducted by CO-CO2 gas mixture (60 vol% CO) at 750 oC.

Carbonylation experiments were conducted in a flexible U – shaped reactor (max pressure 68 atm) immersed into the oil bath. Carbonylation was studied using CO at different pressures. Carbonyls were absorbed by aqua regia in two Dreschel bottles. Samples were taken from the aqua regia solution at different times and analysed by ICP-OES. The extent of carbonylation was calculated using results of the ICP-OES analysis.

The non-catalytic carbonylation of nickel at 100 oC and CO pressure 27 atm was close to completion (extent of reaction was 98%) in about 5.5 hours. Carbonylation of pure iron was slow; the extent of iron carbonylation at 100 oC and CO pressure up to 55 atm (gauge) was less than 5.0%. The extent of cobalt carbonylation under these conditions was less than 0.5%.

Sulphur containing catalysts accelerated the carbonylation reaction. The time for a complete carbonylation of nickel was shortened from 5.5 hours in the non-catalytic reaction to 2 hours in the catalytic carbonylation at 100 oC and CO pressure 14 atm.

The extent of non-catalytic carbonylation of nickel from the selectively reduced BCS laterite ore at 100 oC and CO pressure 41 atm was below 50%. The use of catalysts in the carbonylation of selectively reduced ore was inefficient. The major parameter affecting the rate of ore carbonylation was the particle size. The rate of reaction increased significantly with decreasing particle size; the carbonylation of nickel in MIN ore with the particle size 38-53 and 75-90 µm, was close to completion after 4 hours reaction.

Results of a systematic study of the carbonylation of nickel and selectively reduced laterite ores are significant for further understanding of carbonylation reactions. Promising results were obtained for further development of technology of extraction of nickel from laterite ores by the carbonylation process.

VI Contents

AUTHENTICITY STATEMENT ...... I

ORIGINALITY STATEMENT ...... II

Acknowledgements ...... IV

Abstract ...... V

Contents ...... VII

List of Figures ...... XI

List of Tables ...... XIX

Chapter 1 Introduction ...... 1

Chapter 2 Literature Review ...... 6

2.1 Nickel Deposits ...... 6

2.2 Extractive Metallurgical Processes for Laterite Ores ...... 10

2.2.1 Pyrometallurgical Processes ...... 11

2.2.2 Hydrometallurgical Processes ...... 12

2.2.3 Caron Process ...... 14

2.2.4 Bio-metallurgical Process ...... 16

2.3 Carbonylation ...... 16

2.3.1 Industrial Carbonylation Process ...... 16

2.3.2 Carbonylation of Nickel ...... 21

2.3.2 Carbonylation of Iron ...... 24

2.3.3 Carbonylation of Cobalt ...... 25

2.3.4 Factors Affecting the Carbonylation ...... 26 VII 2.4 Reduction of Metal oxides ...... 35

2.5 Summary and Project’s Objectives ...... 36

Chapter 3 Experimental ...... 37

3.1 Materials...... 37

3.2 Sample preparation ...... 39

3.3 Experimental Set-up and Procedures ...... 41

3.3.1 Reduction Experiments ...... 41

3.3.2 Carbonylation Experiments ...... 42

3.4 Analytical Equipment and Methodology ...... 44

3.4.1 X-Ray Diffraction (XRD) Analysis ...... 44

3.4.2 ICP-OES/MS Analysis ...... 45

3.4.3 Scanning Electron Microscopy (SEM)/Energy Dispersive X-ray Spectrometer (EDS) ...... 46

3.4.4 X-Ray Fluorescence (XRF) Spectroscopy ...... 46

3.5 Selection of the absorption reagent ...... 47

3.6 Data Analysis ...... 48

3.6.1 Definition of extent of carbonylation and the content of catalyst 48

3.6.2 Extent of metal carbonylation ...... 48

Chapter 4 Thermodynamic Analysis of Carbonylation ...... 51

4.1 Effect of CO Pressure ...... 53

4.2 Effect of Temperature ...... 55

Chapter 5 Experimental Results ...... 57

5.1 Reduction of Iron, Nickel and Cobalt Oxides ...... 57

VIII 5.2 Non-catalytic Carbonylation of Metal ...... 61

5.2.1 Non-Catalytic Carbonylation of Pure Nickel ...... 62

5.2.2 Non-Catalytic Carbonylation of Pure Iron ...... 74

5.2.3 Non-catalytic Carbonylation of Cobalt ...... 76

5.3 Catalytic Carbonylation of Nickel ...... 78

5.3.1 Effect of the Content of Iron Sulphide and Sulphur on the Carbonylation of Nickel ...... 79

5.3.2 Effect of Temperature ...... 81

5.3.3 Effect of Carbon Monoxide Pressure ...... 83

5.3.4 Effect of the Gas Flow Rate ...... 83

5.4 Carbonylation of Selectively Reduced Laterite Ores ...... 85

5.4.1 Non Catalytic Carbonylation of Selectively Reduced Laterite Ores ...... 86

5.4.2 Catalytic Carbonylation of Selectively Reduced Laterite Ore .... 93

Chapter 6 Discussion...... 107

6.1 Impurities in Nickel and Iron Oxides ...... 107

6.2 Reduction of Nickel, Iron and Cobalt Oxides ...... 108

6.3 Non-Catalytic Carbonylation of Nickel ...... 111

6.3.1 Effect of Temperature ...... 112

6.3.2 Effect of Carbon Monoxide Pressure ...... 116

6.3.3 Effect of Gas Flow Rate ...... 118

6.3.4 Effect of Nickel Mass ...... 120

6.3.5 Effect of Particle Size ...... 121

6.4 Non-catalytic Carbonylation of Iron and Cobalt ...... 123 IX 6.5 Non-catalytic Carbonylation of Nickel from Ni-Fe Mixture...... 124

6.6 Catalytic Carbonylation of Nickel ...... 125

6.6.1 Mechanism of catalytic carbonylation of nickel ...... 125

6.6.2 Catalytic Carbonylation of Nickel ...... 130

6.6.3 Effect of Gas Flow Rate ...... 134

6.6.4 Effect of CO Pressure ...... 135

6.7 Carbonylation of Selectively Reduced Laterite Ores ...... 136

Chapter 7 Conclusions and Recommendations for Further Work ...... 140

7.1 Conclusions ...... 140

7.2 Recommendations for Further Work ...... 142

Reference ...... 143

Appendix: ...... 159

Appendix A: Calibration of gas flow rate: ...... 159

Appendix B. Calibration of reduction furnace: ...... 160

Appendix C. Calibration of oil bath: ...... 161

Appendix D. Selection of absorption solution ...... 162

Appendix E. Statistics of nickel particles reduced at different temperature ...... 163

Appendix F: Reaction Rate Control in the Non-catalytic Carbonylation ...... 163

X List of Figures

Figure 1.1 World nickel mine production ...... 2 Figure 2.1 The trend in percentage of Ni produced from laterite ores ...... 9 Figure 2.2 Flow sheets of conventional metallurgical processes for laterite ores (Norgate et al. 2011) ...... 11 Figure 2.3 Schematic flow sheet of Caron process (Taylor 2013) ...... 16 Figure 2.4 Flow sheet of the extraction of Ni and Co using carbonylation process (Terekhov 2001) ...... 18 Figure 2.5 Flow sheet of nickel production using the atmospheric pressure carbonylation at INCO, Clydach (Teng-1 2006)...... 19 Figure 2.6 Flow sheet of nickel production using medium pressure carbonyl process at INCO, Clydach (Teng 2006) ...... 20

Figure 2.7 Effect of temperature on the Ni(CO)4 pressure at the reactor outlet at 3 -1 different CO flow rates (in cm s ): (a) 0.37, (b) 2.8, (c) 12.0. Broken line: Ni(CO)4 pressure at chemical equilibrium; Pco=1 atm (De Groot et al. 1980) ...... 22 Figure 2.8 Effect of CO pressure on the rate of nickel carbonylation at 20 oC (De Groot et al. 1980) ...... 23 Figure 2.9 Effect of temperature and pressure on synthesis of iron pentacarbonyl (Mond et al. 1922) ...... 25 Figure 2.10 Effect of CO pressure on catalytic carbonylation of nickel at 70 oC (2 wt% sulphur added before carbonylation); a: atmospheric pressure, b: 9.9 atm, c: 19.8 atm . 27 o Figure 2.11 The effect of H2S on carbonylation of nickel at 25 C and 1 atm (Heinicke et al. 1963) ...... 28 o Figure 2.12 Influence of H2S on iron and nickel carbonylation at 25 C and CO pressure 1.1 atm (Heinicke et al. 1970)...... 29

Figure 2.13 Chemical adsorption of H2S on nickel surface (Den Besten et al. 1962) ... 30 Figure 2.14 The effect of sulphur on the carbonylation of cobalt at 200 oC and CO pressure 200 atm for 10 hours (Kipnis et al. 1973) ...... 30 Figure 2.15 The effect of sulphur on iron carbonylation at 200 oC and CO pressure 200 atm (Hieber et al. 1950) ...... 31

XI Figure 2.16 Effect of sulphur on nickel carbonylation at 70oC and CO pressure of 10 atm (Wang et al., 2009) ...... 31 Figure 2.17 Effect of sulphur on extraction of iron from Nb-Fe-B via carbonylation ... 32 Figure 2.18 Effect of oxygen on the carbonylation of Ni (110) at 130oC. (a) 1ppm oxygen in carbon monoxide, (b) 3ppm in carbon monoxide (Lascelles et al. 1983) ...... 33 Figure 2.19 Carbonylation of nickel with pre-treatment by hydrogen at 150 and 250 oC (Wang et al. 2009) ...... 34 Figure 2.20 Effect of carbon content on cobalt carbonylation at 180 oC and CO gauge pressure 70 atm (Kozyrev et al. 2005) ...... 34 Figure 3.1 The standard Gibbs free energy ∆Go and equilibrium constant (presented as log(k)) for the reaction of iron sulphide with hydrogen ...... 40 Figure 3.2 Stages in the experimental study of carbonylation of Ni and Fe ...... 41 Figure 3.3 Schematic experimental set-up for reduction of Ni and Fe oxides and laterite ore ...... 41 Figure 3.4 Schematic setup for carbonylation experiment ...... 44 Figure 3.5 Extent of nickel carbonylation measured by the weight loss and by the ICP- OES analysis of nickel in aqua regia ...... 48 Figure 4.1 The standard Gibbs free energies ∆rGo and equilibrium constants (presented as log(K)) for reactions of carbonylation of nickel and iron ...... 52 Figure 4.2 Conversions of (a) nickel and (b) iron at different temperatures and CO pressures ...... 53

Figure 4.3 Effect of CO pressure on nickel conversion to Ni(CO)4 ...... 54

Figure 4.4 Effect of pressure on iron conversion to Fe(CO)5 ...... 54

Figure 4.5 Effect of temperature on nickel conversion to Ni(CO)4 ...... 55

Figure 4.6 Effect of temperature on iron conversion to Fe(CO)5...... 55 o Figure 5.1 XRD patterns of Fe2O3 samples reduced by hydrogen at 400 C for 1, 4 and 6 hours ...... 58 Figure 5.2 XRD patterns of NiO samples reduced by hydrogen at 500 oC ...... 59

Figure 5.3 XRD patterns of the NiO-Fe2O3 mixture before and after reduction by hydrogen at 500 oC for 6 hours ...... 60 o Figure 5.4 XRD patterns of Co3O4 sample reduced by hydrogen at 500 C for 6 hours ...... 61

XII Figure 5.5 Carbonylation of the Ni-Fe mixture (50 wt% Ni) at 80 oC and CO gauge pressures 14 and 54 atm ...... 64 Figure 5.6 Nickel carbonylation at 100 oC and CO gauge pressure 27 atm at different gas flow rates ...... 64 Figure 5.7 Effect of CO flow rate on nickel carbonylation at 100 oC and CO gauge pressure 27 atm ...... 65 Figure 5.8 Effect of temperature on the carbonylation of nickel at CO gauge pressure 27 atm ...... 66 Figure 5.9 Effect of temperature on nickel carbonylation at CO gauge pressure 0, 14 and 27 atm after 6.5 hours reaction ...... 66 Figure 5.10 Effect of CO pressure (gauge) on the carbonylation of nickel at 80, 90 and 100 oC...... 68 Figure 5.11 Extents of Ni carbonylation at 80, 90 and 100 oC at different CO gauge pressures after 6.5 hours reaction ...... 68 Figure 5.12 Effect of reduction temperature on nickel particle size ...... 69 Figure 5.13 Effect of particle size on the carbonylation of nickel at 100 oC and CO gauge pressure of 27 atm...... 70 Figure 5.14 Effect of particle size (reduction temperature) on the carbonylation of nickel at 100 oC and CO gauge pressure 27 atm after 4 hours reaction ...... 71 Figure 5.15 Carbonylation of nickel of different mass at 100 oC and CO gauge pressure of 27 atm...... 72 Figure 5.16 Carbonylation of nickel from the Ni-Fe mixtures of different compositions at 100 oC and CO gauge pressure 27 atm on mixture ...... 73 Figure 5.17 Carbonylation of nickel from the Ni-Fe mixtures of different compositions at 100 oC and CO gauge pressure 27 atm after 6.0 hour reaction...... 73 Figure 5.18 Carbonylation of iron at 80-100 oC and CO gauge pressure 0, 27 and 54 atm ...... 74 Figure 5.19 Carbonylation of iron at 80-100 oC and CO gauge pressure 0, 27 and 54 atm after 6.5-hour reaction ...... 75 Figure 5.20 Effect of CO gauge pressure on carbonylation of iron at 80-100 oC...... 75 Figure 5.21 Effect of CO pressure (gauge) on the carbonylation of iron at 80-100 oC after 6.5 hours reaction...... 76

XIII Figure 5.22 Effect of temperature on carbonylation of cobalt at CO gauge pressures 41 and 54 atm ...... 77 Figure 5.23 Effect of CO pressure on carbonylation of cobalt at 80, 90 and 100 oC ...... 78 Figure 5.24 Catalytic carbonylation of nickel at 100 oC and CO pressure (gauge) 14 atm with addition of different amounts of sulphur, introduced as S and FeS ...... 80 Figure 5.25 Catalytic carbonylation of nickel at 100 oC and CO pressure (gauge) 14 atm with addition of different amounts of sulphur in the form of S and FeS after 4 hours reaction ...... 81 Figure 5.26 Effect of temperature on the catalytic carbonylation of nickel ...... 82 Figure 5.27 Effect of temperature on the catalytic carbonylation of nickel after 4 hours reaction ...... 82 Figure 5.28 Catalytic nickel carbonylation at 100 oC and CO pressures (gauge) of 14 and 27 atm ...... 83 Figure 5.29 Effect of CO gas flow rate on the catalytic carbonylation of nickel at 100 oC and CO pressure (gauge) of 14 atm...... 84 Figure 5.30 Effect of gas flow rate on the catalytic nickel carbonylation at 100 oC and CO pressure (gauge) 14 atm after 3 hours reaction ...... 85 Figure 5.31 Effects of reduction conditions on carbonylation of nickel at 100 oC, CO gauge pressure 27 atm and gas flow rate of 0.5 L·min-1...... 87 Figure 5.32 Carbonylation of nickel and iron from the selectively reduced laterite ores at 80-100 oC and CO gauge pressure 27 atm ...... 88 Figure 5.33 Carbonylation of nickel and iron from laterite ore at 100 oC and CO pressure of 27 and 41 atm (gauge) ...... 89 Figure 5.34 Carbonylation of nickel and iron from selectively reduced laterite ore by CO-CO2 gas mixture (60 vol% CO) in comparison with carbonylation of ore o o reduced by H2 at 650 and 850 C; carbonylation temperature 100 C, CO gauge pressure 27 atm ...... 91 Figure 5.35 Carbonylation of nickel and iron from the selectively reduced laterite ore at 100 oC and CO pressure (gauge) 27 atm with flow rates 0.36 and 0.50 L·min-1 ...... 92 Figure 5.36 Catalytic carbonylation of nickel and iron from the selectively reduced laterite ore at 80, 90 and 100 oC; CO pressure (gauge) was 27 atm ...... 94 XIV Figure 5.37 Effect of temperature on the extents of catalytic carbonylation of Ni and Fe from the selectively reduced laterite ore at CO gauge pressure 27 atm after 4 hours reaction ...... 95 Figure 5.38 Catalytic carbonylation of nickel and iron from the selectively reduced laterite ore at 100 oC and CO gauge pressure 14, 27 and 41 atm ...... 96 Figure 5.39 Effect of CO gauge pressure on the extents of catalytic carbonylation of Ni and Fe from the selectively reduced laterite ores at 100 oC after 3 hours reaction ...... 96 Figure 5.40 Catalytic carbonylation of nickel and iron from the selectively reduced laterite ore at 100 oC and CO pressure (gauge) 27 atm at the gas flow rates 0.35 and 0.50 L·min-1...... 97 Figure 5.41 Catalytic carbonylation of nickel and iron from the selectively reduced laterite ore of different size at 100 oC and CO gauge pressure 27 atm ...... 99 Figure 5.42 Effect of particle size on the catalytic carbonylation of Ni and Fe from the selectively reduced laterite ore at 100 oC and CO gauge pressure 27 atm after 3.5 hours reaction...... 99 Figure 5.43 Reduction of Ni, Co and Fe oxides from the laterite ore with size (a) < 53 μm and (b) 53-200 μm in the gas atmosphere containing 60vol% CO and 40 vol%CO2, reduction time was 60 min (Yang, 2014) ...... 101 Figure 5.44 Catalytic carbonylation of nickel and iron from the selectively reduced laterite ore at 650, 750 and 850 oC; carbonylation temperature 100 oC, CO gauge pressure 27 atm ...... 101 Figure 5.45 Effect of the ore reduction temperature on the catalytic carbonylation of Ni and Fe at 100 oC and CO gauge pressure 27 atm ...... 101 Figure 5.46 Effect of CO partial pressure on reduction of Ni, Co and Fe oxides from 53-200 µm laterite ore at 740 oC for 60 min; gas flow rate was 700 ml∙min-1 ...... 102 Figure 5.47 Catalytic carbonylation of nickel and iron from laterite ore selectively reduced by CO-CO2 gas with different CO concentration. Carbonylation temperature was 100 oC, CO gauge pressure 27 atm ...... 103

Figure 5.48 Carbonylation of nickel and iron from laterite ore reduced by CO-CO2 gas o (60 vol% CO) with addition of 5 vol% H2S. Carbonylation temperature was 100 C; CO gauge pressure 27 atm; 1: Introduction of H2S during last 10 min of the reduction of XV laterite ore; 2: Introduction of H2S during last 30 min of the reduction of laterite ore; 3: catalytic carbonylation (with addition of 1 vol% H2S) of the selectively reduced laterite ore without catalyst; ...... 105 Figure 6.1 XRD patterns of reduced nickel and residual sample after carbonylation at 100 oC and CO pressure 27 atm at a gas flow rate of 0.5 L∙min-1 ...... 108 Figure 6.2 SEM images of nickel reduced at 600 and 700 oC for 6 and 12 hours ...... 110 Figure 6.3 SEM image of nickel reduced from nickel oxide by hydrogen at 500 oC .. 111 Figure 6.4 Schematic diagram of pure nickel carbonylation ...... 112 Figure 6.5 XRD patterns of residuals of samples after carbonylation at different temperatures at CO gauge pressure 27 atm (6.5 hours reaction) ...... 113 Figure 6.6 Effect of temperature on the carbonylation of nickel at CO pressure 56 atm ...... 114 Figure 6.7 Plots of ln k versus 1/T at temperatures between 80 and 100 oC and CO pressure 56 atm, gas flow rate 0.5 L∙min-1 ...... 115 Figure 6.8 Rate of non-catalytic carbonylation of nickel at 80, 90 and 100 oC and CO pressure 27 atm; gas flow rate 0.5 L∙min-1 ...... 116 Figure 6.9 XRD patterns of residual samples after non-catalytic carbonylation of nickel at 100oC at CO pressures of 14 and 27 atm ...... 117 Figure 6.10 Rate of non-catalytic carbonylation of nickel at 100oC and CO gauge pressures 14 and 27 atm (gas flow rate 0.5 L·min-1) ...... 118 Figure 6.11 XRD patterns of residuals of samples after non-catalytic carbonylation of nickel at 100 oC and CO pressure 27 atm at flow rates 0.14 and 0.50 L·min-1 after 6.5 hours reaction...... 119 Figure 6.12 Rates of nickel carbonylation at 100 oC and CO pressure 27 atm at gas flow rates of 0.23, 0.35 and 0.50 L∙min-1 ...... 120 Figure 6.13 Mass of carbonylated nickel at different concentrations of nickel tetracarbonyl in gaseous phase at a gas flow rate of 0.5 L·min-1 ...... 121 Figure 6.14 Rate of carbonylation of nickel with different particle size at 100 oC and CO pressure of 27 atm (gas flow rate 0.5 L∙min-1)...... 123 Figure 6.15 Rate of nickel carbonylation from Ni-Fe mixtures at 100oC, CO pressure 27 atm and gas flow rate 0.5 L·min-1 ...... 125

XVI Figure 6.16 XRD patterns of residual samples of Ni catalyzed by S and FeS (content of sulfur 5.0 wt%) at temperature 100 oC and CO pressure 14 atm; experiments were terminated at 10, 20 and 30 minutes in reactions with addition of sulphur, and 20 and 30 in reactions with addition of FeS. Samples marked as 00 min were not subjected to carbonylation; they were reactivated by heating in hydrogen and then cooled down to room temperature...... 126 Figure 6.17 XRD patterns of residual samples of nickel after carbonylation catalyzed by sulphur (3.5 and 7.5 wt%) in the form of FeS at 100 oC and CO pressure 14 atm after 3-4 hours reaction ...... 127 Figure 6.18 XRD patterns of residual samples of Ni-Fe mixture catalyzed by sulphur and iron sulphide FeS (content of sulfur 5.0%) at temperature 100 oC and pressure 27 atm; experiments were terminated at 10, 20 and 30 minutes...... 129 Figure 6.19 Schematic of the carbonylation of nickel catalyzed using elemental sulphur...... 129 Figure 6.20 Catalytic carbonylation of Ni at 100 oC and CO pressure 14 atm in comparison with non-catalytic carbonylation at CO pressure 14 atm (at 100 oC) and 27 atm (80-100 oC) ...... 131 Figure 6.21 Rates of catalytic nickel carbonylation at 100oC and CO pressure 14 atm (flow rate 0.5 L·min-1) with different contents of sulphur in the form of elemental sulphur and iron sulphide ...... 132 Figure 6.22 XRD patterns of residual samples after carbonylation of nickel catalysed by elemental sulphur (0.1 - 7.0 wt%) at 100 oC and CO pressure 14 atm (reaction time 3 hours) ...... 134 Figure 6.23 Rate of catalytic nickel carbonylation with addition of FeS (0.1 wt% S) at 100 oC and CO pressure 14 atm, and non-catalytic carbonylation at 100 oC and CO pressure 27 atm with different gas flow rates ...... 135 Figure 6.24 Extent and rate of catalytic nickel carbonylation with addition of FeS (0.1 wt% S) and non-catalytic carbonylation at 100 oC and CO pressure 14 and 27 atm (gas flow rate 0.5 L·min-1) ...... 136 Figure 6.25 Schematic diagram of extraction of nickel from laterites by carbonylation ...... 137 Figure 6.26 Extent and rate of non-catalytic and catalytic carbonylation of nickel in selectively reduced laterite ore at 100oC and CO pressure 27 atm ...... 138 XVII Appendix A. Calibration of gas flow meter (a & b gas flow meter for reduction. c. gas flow meter for carbonylation) ...... 160 Appendix B. Calibration of the isothermal zone and correlation between setup temperature and isothermal temperature ...... 161 Appendix C.2 Calibration of temperature of the oil bath ...... 162 Appendix D. Absorption of nickel and iron carbonyls in nitric acid and aqua regia (a and c: 50% - nitric acid solution, b and d: aqua regia) ...... 162 Appendix E. Statistical chart of nickel particles reduced at different temperature163 Table F1 Plots of X for gas film control, 1 – (1 – X)1/3 for the intrinsic control and 1 - 3(1-X)2/3 + 2(1-X) for the diffusion control in the first 2.5 hours carbonylation of Ni at 80 - 100 oC and CO pressure 55 atm with gas flow rate 0.5 L∙min-1 ...... 166

XVIII List of Tables

Table 1-1 2014 Ni production and total reserves by country (Statista, 2015a, Statista, 2015b) ...... 2 Table 2–1 Nickel bearing minerals and nickel contents (Boldt 1957) ...... 6 Table 2-2 Compositions of typical laterite ores, wt% (Li et al. 2007, Wang et al. 2008) 8 Table 2-3 Chemical composition and processing technology of laterite ores (Liu 2002) ...... 10 Table 2-4 Extents of metals extraction from laterite ores by HPAL process at Moa Bay Nickel Company (Peng, 2005) ...... 13 Table 3-1 Chemicals and gases ...... 37 Table 3-2 Composition of MIN and BCS garnieritic laterite ores (wt%) ...... 38 Table 3-3 Facilities and glasswares ...... 39 Table 3-4 Parameters in the reduction process ...... 42 Table 3-5 Parameters in carbonylation experiments ...... 43 Table 3-6 ICP-OES analysis parameters ...... 46

Table 5-1 Degree of reduction of Fe2O3 by hydrogen ...... 58 Table 5-2 Degree of reduction of NiO by hydrogen at 400 and 500 oC ...... 59 o Table 5-3 Degree of reduction of Fe2O3 – NiO mixture at 500 C ...... 60 Table 5-4 Parameters in a study of non-catalytic carbonylation of Ni, Fe, Co and Ni-Fe mixture ...... 62 Table 5-5 Particle size of Ni reduced at different temperatures ...... 69 Table 5-6 Experiment parameters in the catalytic carbonylation of nickel ...... 79 Table 5-7 Reaction parameters in non-catalytic carbonylation of laterite ores ...... 86 Table 5-8 Extents of carbonylation of nickel at 100 oC and CO pressure 27 atm (5.5 h reaction), produced by reduction of NiO by H2 and CO-CO2 gas mixture (60 vol% CO) ...... 87 Table 5-9 Extents of nickel and iron carbonylation from the selectively reduced laterite ores at 80-100 oC after 3.5 hours reaction ...... 88 Table 5-10 Effect of CO pressure on non-catalytic carbonylation of nickel and iron from the selectively reduced laterite ores at 100 oC ...... 90 Table 5-11 Extents of carbonylation of nickel and iron from selectively reduced laterite ore by CO-CO2 gas mixture (60 vol% CO) and from the ore reduced by H2 XIX at 650 and 850 oC (carbonylation temperature 100 oC, CO gauge pressure 27 atm) after 3.5 hours reaction ...... 91 Table 5-12 The extents of nickel and iron catalytic carbonylation from the selectively reduced laterite ore at 100 oC and CO pressure 27 atm at the gas flow rates 0.35 and 0.50 L·min-1 after 3.5 hours reaction...... 98 Table 5-13 Extents of catalytic carbonylation of nickel and iron from ores reduced by CO-CO2 gas with different contents of CO ...... 103 Table 5-14 Extents of carbonylation of Ni and Fe from laterite ore reduced with o addition of H2S. Carbonylation temperature was 100 C; CO pressure 27 atm; reaction time 2.5 hours ...... 105 Table 6-1 BET surface area and pore size of samples produced by reduction at different temperatures and residual samples after carbonylation ...... 122 Table 6-3 BET surface area and pore size of MIN ore with different particle size ..... 139 Appendix C.1 Calibration of the temperature (oC) in oil bath ...... 161

XX Chapter 1 Introduction

Nickel, a silvery-white lustrous metal, belongs to the transition metals of group IVB. It has relatively low electrical and thermal conductivities, and readily alloys with other metals to increase their strength and resistance to corrosion as well as their electrical, magnetic and heat resistant properties. Nickel is widely used in stainless steel, magnets, electroplating and other areas. About 65% of nickel is consumed in the manufacturing of stainless steel and heat-resisting steels for construction, military, marine, transport and aerospace industries. The great demand of stainless steel, especially in China, promotes the continuous increase of nickel production.

Nickel in ores exists in the form of sulphides and oxides. Laterites are classified as wet or tropical laterites (from Cuba and Indonesia) and dry laterites (Australia) Chen et al. (2004). Sulphide reserves are still the major resource of nickel production; they have nickel of high grade and can be easily upgraded by beneficiation. The demand and production of nickel has been increasing, but there are a few new sulphide deposits reported last decades. The increasing price of nickel and co-extraction of cobalt make it economically to produce nickel from laterites; the proportion of nickel production from laterite ores has been increasing. Figure 1.1 shows nickel mined from both laterite and sulphide ores from 2000 and the predicted production to 2030 by International Nickel Study Group (Miner 2015).

Australia is among the major suppliers and producers of nickel ore in the world. In 2014, Australia was the fifth largest nickel producer in the world (after Philippines, Russia, Indonesia and Canada). The economic reserve of nickel in Australia is about 19 million tonnes accounting for 35 per cent of the nickel resources in the world. Table 1-1 lists some statistical data in the year of 2014 (Statista-a 2015, Statista-b 2015).

Compared with nickel sulphide deposits that are found hundreds of metres under the ground, laterites are usually located nearer to or at the surface with abundant reserve, which renders it available to open mining. The major challenges in processing of laterite

1 ores include ores’ low and variable grades; besides, laterite ores cannot be concentrated by beneficiation.

Figure 1.1 World nickel mine production

Table 1-1 2014 Ni production and total reserves by country (Statista, 2015a, Statista, 2015b)

Country Production Country Reserves

Metric tons Million Metric tons

Philippines 440000 Australia 19

Russia 260,000 New Caledonia 12

Indonesia 240000 Brazil 9.1

Canada 233000 Russia 7.9

Australia 220000 Cuba 5.5

New Caledonia 165000 Indonesia 4.5

Brazil 126000 South Africa 3.7

China 100000 Philippines 3.1

Colombia 75000 China 3

2 Cuba 66000 Canada 2.9

South Aferica 54700 Madagascar 1.6

Madagascar 37800 Colombia 1.1

Dominican Republic 15800 Dominican Republic 0.93

United States 0.16

Other Countries 6.5

The conventional processes applied in extracting metals (nickel, iron and cobalt) from laterites are pyrometallurgical and hydrometallurgical routes. In the pyrometallurgical processing, ore is treated by drying, calcining and reducing at temperatures above 900 oC, followed by the smelting nickel and iron over 1300 oC. This process has a high fuel consumption; it is currently used in extracting nickel from saprolite ores rich in nickel. Pyrometallurgical processing of the low grade limonitic ore with nickel content less than 1.8% is not economical.

In hydrometallurgical routes, nickel, iron and cobalt are extracted from laterite ores at temperatures over 200 oC and high pressures above 100 atm; strong concentrated sulphuric acid is used to solve metals from ores lattice to enhance the reaction rate and degree of extraction, then nickel, iron and cobalt are recovered by precipitation. The residual solution must be neutralised before discharging into the environment. Nevertheless, acid and alkaline solutions which are used in the process cause the environmental concern. The operation of hydrometallurgical process is also influenced by iron and magnesium, which contents are much higher than contents of nickel and cobalt.

Laterite ores cannot be efficiently upgraded prior to pyrometallurgical or hydrometallurgical processing what leads to significant challenges in extraction of nickel and cobalt. Heating the ore containing a significant amount of impurities to high temperatures in pyrometallurgical route causes high fuel consumption and emission of carbon dioxide to the environment. High contents of iron and magnesium in the ores result in high consumption of acid applied in the hydrometallurgical route. Therefore application of conventional processes to extraction of nickel from laterites results in serious environment problems and high cost of nickel production. Therefore, new 3 technologies which lead to the reduction of the production cost and decrease environmental impact have been under development.

Promising results in extraction of nickel and cobalt from laterite ores were obtained using bio-technology and acid leaching by hydrochloric acid (Coto et al. 2008, McDonald et al. 2008). However, development of these technologies has not reached the industrial scale. The rate of bioleaching is very low, extents of nickel and iron extraction are less than 50%. The hydrochloric acid is highly corrosive to facilities; the use of the lower acid strength decreases efficiency of nickel and cobalt extraction from laterites (Peng 2005, Zhou 2005).

A major nickel refining process is electrowinning. Carbonyl process (based on the Mond process (Herrmann 1990)) is also broadly used for nickel refining. Nickel and iron form carbonyls by reactions (1-1) and (1-2), which are separated by distillation.

Ni + 4CO = Ni(CO)4 (1-1)

Fe + 5CO = Fe(CO)5 (1-2)

Then nickel and iron are recovered by thermal decomposition of carbonyls by reverse reactions of (1-1) and (1-2), and carbon monoxide is recycled.

Carbonylation technologies have been used in refining nickel and iron over 100 years; the carbonylation of nickel had been intensively studied in the first decades of 20th century. However, a study of carbonylation reactions was limited only in application to pure metals. No detailed investigation of kinetics and mechanisms of carbonylation of nickel and iron at elevated temperatures and medium pressures have been reported, especially in the carbonylation of reduced laterites.

Compared with pure nickel, the properties and chemical compositions of reduced laterites are more complex; their carbonylation needs a detailed investigation. This project will study a novel approach to the processing of laterite ores based on the selective reduction and carbonylation of reduced ore to extract nickel and iron. Reduction of laterite ore was examined in the PhD project “Selective Reduction of Laterite Ore” (Yang 2014).

4 In the project, prior to carbonylation, laterite ores are selectively reduced by CO-CO2 mixture; in selective reduction, most of nickel and only part of iron are reduced. Nickel and iron in the selectively reduced ore are extracted by carbonylation.

The ultimate aim of this project is to assess the feasibility of extraction of nickel and iron by carbonylation of selectively reduced laterite ore. The project undertakes a systematic study of carbonylation of pure nickel and iron, and carbonylation of selectively reduced laterite ore to develop a better understanding of carbonylation reactions and extraction of nickel from laterite ore by carbonylation. Specific objectives of this study are:

1) to study the effects of reaction parameters on carbonylation of nickel and iron, including reaction temperature, CO pressure, gas flow rate, particle size, sample mass, and the effect of sulphur-containing catalysts;

2) to study the non-catalytic and catalytic carbonylation of laterite ores; the impacts of reduction conditions; and effects of reaction parameters;

3) to develop further understanding of kinetics and mechanisms of carbonylation processes.

5 Chapter 2 Literature Review

Nickel is an important metal and alloying component applied in stainless steel, alloys, plating and numerous other applications in industry, among them stainless steel accounts for about two thirds of the annual nickel consumption (Zhao et al. 2012). Driven by the increasing production of stainless steel and superalloys, world nickel demand increased from 1.10 Mt in 2001 to 1.57 Mt in 2011, representing an annual growth rate of 4.2% on average (INSG-2 2013). The world primary nickel production reached 1.76 Mt in 2012, 1.96 Mt in 2013 and increased to 1.99 Mt in 2014 (INSG-1 2013, INSG-3 2015).

2.1 Nickel Deposits

Nickel is commonly present in two principal ore types - sulphide and laterite ores with a reserve of about 160 Mt, of which near 70% are contained in laterite ores (Valix et al. 2001, Zhou 2005, Norgate et al. 2011).

Table 2-1 lists the principal nickel-bearing minerals and their chemical compositions, in which pentlandite, (Ni, Fe)9S8 , accounts for about three-quarters of nickel mined in the past in the world (Boldt 1957).

Table 2–1 Nickel bearing minerals and nickel contents (Boldt 1957)

Formula Nickel Content, wt%

Sulfides

Pentlandite (Ni, Fe)9S8 34.2

Millerite NiS 64.7

Heazlewoodite Ni3S2 73.3

6 Linnaeite series (Fe, Co, Ni)3S4 variable

Polydymite Ni3S4 57.96

Violarite Ni2FeS4 38.9

Siegenite (Co, Ni)3S4 28.9

Gersdorffite NiAsS 35.4

Silicate and Oxide

Garnierite (Ni, Mg)6Si4O10(OH)8 Variable up to 47

Nickel-ferrous limonite (Fe, Ni)O(OH)·nH2O Variable

Sulphide deposits, with the primary ore mineral of pentlandite, are formed as the results of segregation and concentration of droplets of liquid sulphide from mafic or ultramafic magma (Naldrett 1999). This type ore can be beneficiated by flotation, gravity and magnetic separation to produce concentrates before extracting nickel from ores. This operation significantly reduces the capital and operation cost (Chen et al. 2004, Bo 2009). The sulphide deposits are mainly distributed in Canada, Russia, Australia, China, South Africa, Zimbabwe and Botswana (Cao 2005, Zhou 2005, Li et al. 2009, Liang et al. 2009).

Although over 50% of nickel is produced from sulphide minerals in the past, the major nickel bearing minerals are laterites; they are formed by chemical weathering of precursor ultramafic rocks and supergene enrichment of weathering products mostly under tropical climatic conditions (Oliveira et al. 2001, Chen et al. 2004, Norgate et al. 2011, Berger et al. 2013). Laterites can vary significantly in mineralogy according to location, climate and depth, for example, the clay-rich nickel laterites in Western Australia differ from Moa Bay (Cuba) laterites in mineralogy and have comparatively high silica and low chromium content. Nickel-bearing laterite deposits also contain 7 valuable cobalt minerals (Andersen et al. 2009). The main host minerals for nickel and cobalt can be either limonites (SiO2 ≈ 6%; MgO 3%; Ni 1.4%; Co 0.15% and Fe >40%) or saprolites (SiO2 ≈ 38%; MgO 25%; Ni 2.4%; Co 0.05% and Fe <15%) depending on the content of iron and magnesium (Whittington et al. 2000, Berger et al. 2013, Thubakgale et al. 2013). Large quantities of laterites are found in Australia, Cuba, Indonesia, New Caledonia and the Philippines (Moskalyk et al. 2002, USGS 2009). The typical laterite ores compositions are listed in Table 2-2 (Li et al. 2007, Wang et al. 2008).

Table 2-2 Compositions of typical laterite ores, wt% (Li et al. 2007, Wang et al. 2008)

Ni Co Fe MgO SiO2 CaO Al2O3 Cr

New Caledonia Ore 2.43 0.04 9.3 28.8 42.2

Indonesia Ore 2.6 0.1 14.47 25.5 36.4 0.75

International Nickel 2 19.3 17.4 33.3 Company (Canada)

2.3 0.07 21.57 15.77 35.9 1.24 Philippines Ore 1

Yuanjiang Ores (China) 1.24 0.08 24.6 19.4 31.8 0.34 6.9

Philippines Ore 2 1.15 0.09 38 0.6 10 1.5

Yabulu Ore (Australia) 1.57 0.1 30

Cawse Ore (Australia) 1.02 0.07 18 1.58 42.5

Bulong Ore (Australia) 1.9 0.08 18 1~2 42.5

Albania Ores I 0.89 0.06 42.9 2.67 13.9 2.31 5.9

Albania Ores II 0.96 0.06 50.4 1.33 6.48 2.46 3

In the recent past, the majority of nickel was produced from sulphide ores (Whittington et al. 2000), but with the declining sulphide ore reserves and the increasing demand for

8 nickel in the world, more nickel is extracted from laterite ores (Valix et al. 2001, Zhai et al. 2009, Norgate et al. 2011), which accounts for about 70% of nickel resources (Li et al. 2009). The proportion of nickel extracted from laterite ores is shown in Figure 2.1, the percentage has increased from about 15% in 1950 to about 50% in 2010 (Li et al. 2010). Besides the abundant reserves, nickel laterites are important sources of cobalt; they are of increasing significance for exploration and production (Andersen et al. 2009).

Proportion of laterite ores (%) of laterite Proportion 10

0 1950 1960 1970 1980 1990 2000 2010 Year

Figure 2.1 The trend in percentage of Ni produced from laterite ores

By now, more than 120 Ni-Co laterite deposits have been explored in the world. Different from the existing state of pentlandite in sulphide ores, nickel and cobalt in laterite ores present as cations substituted within manganese oxide, goethite and /or clays; the distribution of nickel throughout the lattice of the ore particles makes the beneficiation by flotation or gravity separation impractical, which results in intrinsically high capital and operating costs (Whittington et al. 2000, Norgate et al. 2011).

Nevertheless, there are many obvious benefits for mining and processing laterite ores. Firstly, nickel laterite ores are usually nearer to the surface and can be processed by open-cut mining (Chemlink 1997, Moskalyk et al. 2002), which reduces the high mining cost. Secondly, the reserve of laterite ores is abundant accounting for 70% of nickel reserve on land, which ensures the supply for nickel resource in future. Finally, cobalt, an important metal with eight-times higher value than nickel and representing up

9 to 40% of the total value of production, is often associated with nickel in laterite ores (Chemlink 1997, LI et al. 2009). The recovery of cobalt makes the laterite ores processing more practicable (Berger et al. 2013). For example, Anaconda Nickel NL commenced development of Ni laterite deposits for Ni and Co production in Murrin Murrin, WA (Wells et al. 2003).

2.2 Extractive Metallurgical Processes for Laterite Ores

The conventional metallurgical processes for extraction of nickel from laterite ores can be divided into three routes: pyrometallurgical route (ore smelting), hydrometallurgical route (ore leaching), and Caron-type processes (Berger et al. 2013). Because composition of laterites varies significantly in different deposits, and laterite ores cannot be effectively upgraded by beneficiation, the choice of the metallurgical technology depends on the ore composition and physical properties. Both pyro- and hydrometallurgical processes have been operated in nickel metallurgy for many decades, and are being improved. Table 2-3 lists the typical laterite ore types and their processing technology (Liu 2002).

New technologies for processing of laterites are under development, such as chloridizing segregation and bioleaching (McDonald et al. 2008).

Table 2-3 Chemical composition and processing technology of laterite ores (Liu 2002)

Chemical composition % Ore type Processing technology Ni Co Fe Cr2O3 MgO

Limonite 0.8-1.5 0.1-0.2 40-50 2-5 0.5-5 Hydro

1.5-1.8 0.02-0.1 25-40 1-2 5-15 Pyro or Hydro

Saprolite 1.8-3 0.02-0.1 10-25 1-2 15-35 Pyro

Flow sheets for processing of laterite ores are schematically shown in Figure 2.2 (Norgate et al. 2011). 10 Figure 2.2 Flow sheets of conventional metallurgical processes for laterite ores (Norgate et al. 2011)

Generally, saprolite ores with relatively high nickel, magnesia and silica contents are mainly processed by pyrometallurgical processes. Goethite-rich limonite ore which has a low content of nickel and high content of iron and cobalt is mainly processed by hydrometallurgical technology. The transition layer of laterite ore (saprolite-limonite ore) can be processed by pyrometallurgical or hydrometallurgical processes

2.2.1 Pyrometallurgical Processes

Pyrometallurgical process is the primary technology used in processing of sulphide ores (Moskalyk et al. 2002, Mudd 2009, Norgate et al. 2011). In application to laterite ore, it was firstly used to produce ferronickel in New Caledonia in 1879 (Norgate et al. 2011, Oxley et al. 2013). About 70% of mined laterite ores have been treated by pyrometallurgical processes (Xu 2005, Warner et al. 2006).

Pyrometallurgical processing of laterite ores is based on reduction of nickel oxide, which is less stable than other oxides in the ore. However, iron oxides are also partially reduced. Two major process routes are used in processing of laterite ores: production of ferronickel by selective reduction, and production of nickel matte from the selectively reduced laterite ore followed by conversion of nickel sulphide to nickel. 11 Production of ferronickel starts with processing of laterite ore in Rotary Kiln-Electric Furnace (RKEF) process (Whittington et al. 2000, Guo et al. 2009, Zhai et al. 2009, Zhao et al. 2012). Initially, the ores are screened, crushed and blended to produce feed for a consistent process plant with defined ratios of iron and nickel. Then the feed is charged into a rotary kiln where laterite ore is calcined and prereduced by coal and powdered coke. Afterwards the calcine and residual coke are charged into the electric furnace to produce crude ferronickel with nickel contents varying from 13 to 25%. The other components (mainly FeO, SiO2, MgO) are removed with slag (Christian Redl 2013).

Nickel matte smelting process (Whittington et al. 2000, Guo et al. 2009, Zhai et al. 2009, Zhao et al. 2012) is a variation of a standard flow sheet, in which sulphur is added to the kiln. The crude metal/matte is further processed/refined to produce nickel (Xu 2005).

Although pyrometallurgical processes present a mature technology, their disadvantages are also obvious. Because laterites cannot be efficiently concentrated before drying, the production of nickel from oxide ores by pyrometallurgical technologies consumes two to three times energy as processing sulphide ores with significant environmental impact. This technology is suitable only for saprolite ores. Finally, the valuable cobalt cannot be recovered (F.Habashi 1997, Whittington et al. 2000, Berger et al. 2013, Thubakgale et al. 2013).

2.2.2 Hydrometallurgical Processes Compared with pyrometallurgy, hydrometallurgical processes are more efficient in extraction of nickel and cobalt from low nickel grade laterite ores (Oxley et al. 2013). Hydrometallurgical processes are applied to the goethite-rich limonite of relatively low nickel content but high iron and cobalt contents (Whittington et al. 2000). Several hydrometallurgical processes have been developed, which include Atmospheric Leaching (AL), Acid Heap Leaching (AHP), Chloride Leaching (CL), Enhanced Pressure Acid Leaching (EPAL) and High Pressure Acid Leaching (HPAL), among which HPAL and EPAL processes are the only two processes reported to be used in industry (Xu 2005, McDonald et al. 2008, Zhai et al. 2009, Norgate et al. 2011).

12 In the laboratory study, AL was tested using sulphuric acid for limonite, smectite (clay) and saprolite laterites with advantages of lower capital and operation costs (McDonald et al. 2008). The leaching rate increased with temperature, acid concentration and the activity of the acid (McDonald et al. 2008).

The first commercial HPAL plant, operated at temperatures between 245 and 270 oC, started production of nickel and cobalt in 1959 at Moa Bay in Cuba and still continues Ni production (Oxley et al. 2013). HPAL process is also been used to extract nickel and cobalt from laterite ores in Western Australia (Whittington et al. 2003).

Hydrometallurgical processes include acid leaching and ammonia leaching technologies. The ammonia leaching was applied in the Caron process, which will be considered in the next section.

The acid leaching is implemented in titanium-lined autoclaves (Australian projects) or acid brick and lead-lined autoclaves (Moa Bay project) at approximately 250℃. The nickel and cobalt bound with the goethite or clays are released to the acid solution in the pressure leach at high pressure (>200 atm), nickel and cobalt can be extracted with high recoveries at 99.7%, and 98% correspondingly (Whittington et al. 2000, Jiang et al. 2004). The extraction of metals by of HPAL in Cuban Moa Bay Nickel Company is shown in Table 2-4 (Peng 2005).

Table 2-4 Extents of metals extraction from laterite ores by HPAL process at Moa Bay Nickel Company (Peng, 2005)

Component Leaching solution g·L-1 Leaching residue % Leaching rate % Ni 5.95 0.08 96

Co 0.04 0.008 95

Cu 0.1 - 100

Zn 0.2 - 100

Fe 0.8 51 0.4

Mn 2.0 0.4 57

13 Cr 0.3 - 3.0

Cr2O3 - 0.3 -

The main chemical (leaching) reactions involved in the hydrometallurgical process are given below (Jiang et al. 2009).

2 NiO 2H  Ni  H2 O (2-1)

2 2Ni2 O 4  12H  6Ni  6H2 O  [O 2 ] (2-2)

3 FeOOH 3H  Fe  2H2 O (2-3)

+ 3+ Fe2 Si 4 O 10 (OH) 2 + 6H  2Fe + 4SiO2 + 4H 2 O (2-4)

+ 2+ Co2 O 3 + 6H 2Co + 3H2 O (2-5)

Currently, sulphuric acid is used in most hydrometalligical processes for leaching nickel laterites. The use of hydrochloric acid for leaching process has also been examined (McDonald et al. 2008).

The advantage of hydro-metallurgy is high recoveries of both nickel and cobalt, but there are many challenges. One big challenge is the high capital and operation costs of the high pressure and temperature process and the scale formation in autoclaves. The use of sulphuric acid and ammonia for leaching potentially causes serious environmental problems. This route is not economic for processing ores containing acid consuming compounds, such as magnesia; in which the leaching of nearly all iron and magnesia results in a high acid consumption (Whittington et al. 2000, McDonald et al. 2008, Zhai et al. 2009).

2.2.3 Caron Process Caron process was developed in the 1920s for processing low grade laterite ores (Berger et al. 2013). Caron process is a combined pyro-/hydro-metallurgical process (Graaf et al. 1979, Whittington et al. 2000, Zhai et al. 2010); its schematic flow sheet is shown in Figure 2.3 (Taylor 2013).

In the Caron process, the first step is the reduction of nickel and cobalt oxides to metallic state at temperature above 700 oC, then nickel and cobalt are leached by the 14 ammonia-ammonium carbonate solution. After metals are transferred to the solution, ammonia is removed by boiling, and basic nickel carbonate precipitates which, upon calcination at 1200 oC, is converted to nickel oxide (Whittington et al. 2000, Nicol et al. 2004).

The main leaching and precipitation reactions in Caron process are listed as follows:

1 Ni O  6NH  CO  Ni(NH )22  CO (2-6) 2 2 3 2 3 6 3

22 Ni(NH3 ) 2  Ni 2NH3 (2-7)

22   6Ni 6OH  2CO3  3Ni(OH)2  2NiCO3 (2-8)

3Ni(OH)2  2NiCO3  5NiO  3H2 O  2CO2 (2-9)

3 2Co O  12NH  3CO  2Co(NH )32 3CO (2-10) 2 2 3 2 3 4 3

1 Fe O  nNH  CO  Fe(NH )22  CO (2-11) 2 2 3 2 3 3 3

1 2Fe(NH )2 5H O  O  2Fe(OH)   2(n  2)NH  4NH (2-12) 3 n 2 2 2 3 3 4

15 Figure 2.3 Schematic flow sheet of Caron process (Taylor 2013)

Caron process has disadvantages of the both pyro-and hydrometallurgical processes. Firstly, the initial ore drying step is characterised by high energy consumption; secondly, the cobalt recovery is low (usually less than 40%) and the extraction is particularly low from saprolite ores; finally it is sensitive to mineral composition and requires careful blending (Zhou 2005, Norgate et al. 2011).

2.2.4 Bio-metallurgical Process Bioleaching is a novel technology applicable to metal extraction from low-grade ores. During the bioleaching process, metals are dissolved from oxide and silicate ores into organic acids (citric acid, oxalic acid, gluconic acid, tartaric acid and pyruvic acid) produced from heterotrophic microorganisms. Aspergillus and Penicillium are two important filamentous fungi extensively studied for extraction of metals from oxide ores (Kar et al. 1996, Coto et al. 2008). In fungal bio-leaching of laterite ores for 21 days, nickel extraction reached only 36 %, cobalt 54% and iron 0.76 % (Valix et al. 2001). Bio-leaching is less efficient in comparison with extraction of metals by conventional technologies. There is a long way to go before bio-metallurgical extraction of Ni and CO can be commercially used in the industry.

2.3 Carbonylation

Carbonylation process was developed in application to nickel refining at the beginning of last century; it was also applied for production of ultra-pure iron (Mond et al. 1890, Herrmann 1990).

Nickel and iron carbonyls are formed at elevated temperatures and pressures, separated, isolated and thermally decomposed to yield pure metals and carbon monoxide gas. Cobalt which is often present with nickel and iron in laterite ores and tailings, forms nitrosyl carbonyls and can be extracted (Terekhov 2001).

2.3.1 Industrial Carbonylation Process The industrial application of carbonylation technology to nickel refining commenced in 1902 (Herrmann 1990).

16 The refining process was based on the different reaction rates and physical properties of metal carbonyls (different boiling points). Except for nickel, iron and cobalt, other metal carbonyls are not directly synthesized by reactions with carbon monoxide. The carbonylation rates of nickel, iron and cobalt vary greatly. Nickel reacts with CO rapidly and is completely converted into gaseous nickel tetracarbonyl, while iron and cobalt react with carbon monoxide relatively slowly (Rigin 1993). After carbonylation, nickel tetra-carbonyl and iron penta-carbonyl are separated by distillation from the carbonyls mixture. Cobalt carbonyl, having much lower vapour pressure than nickel and iron carbonyls, remains in the carbonyl reactor together with solid residues of the carbonylation reaction (Huang 1990, Terekhov 2001, Peng 2005).

The first patent (English patent No. 12626) on carbonylation was granted in 1890 (Herrmann 1990). In this patent, nickel ores were firstly roasted (calcined), then reduced at temperatures as low as possible using CO-H2 gas mixture, and finally nickel was converted to nickel tetra-carbonyl by reaction with carbon monoxide at 50 oC and 0.1 MPa. The mixture of the excess of CO and nickel tetracarbonyl was heated to 180 oC.

At this temperature, Ni(CO)4 was decomposed to pure metallic nickel and CO with nickel recovery of 92%. Based on this technology, Mond-nickel Co. Ltd produced 3000 tons of nickel in 1910, the Vale Limited started production of nickel carbonyl under pressure 7 MPa at 180 oC in 1973 (Herrmann 1990, Huang 1990).

In another patent (Terekhov 2001), cobalt and nickel carbonyls were formed at elevated temperatures and pressures, separated, isolated and thermally decomposed to yield pure metal pellets and carbon monoxide gas. Impure cobalt carbonyl reacted with a complex gaseous mixture of nitric oxide/carbon monoxide to produce cobalt nitrosyl carbonyl. The process was carried out at 60~100 oC, under 3 atm operation pressure. The reported recoveries of nickel and cobalt reached 97% and 94% correspondingly. Figure 2.4 shows the schematic diagram from this patent (Terekhov 2001).

17 Figure 2.4 Flow sheet of the extraction of Ni and Co using carbonylation process (Terekhov 2001)

Besides the plants mentioned above, carbonylation process to refine nickel or iron, has been operated by different companies. In Germany, BASF Aktiengesellschaft is o reported to annually produce approximately 10,000 tonnes of Fe(CO)5 at 200 C and CO pressure 204 atm (Herrmann 1990), which is used for matins iron and iron crude oxide pigments.

International Nickel Company (INCO, Canada) applied the medium pressure carbonylation process (7 MPa, 180 oC) to produce annually 45,000 tonnes of nickel (Peng 2005). Norilsk company (Russia), extracted nickel by high pressure carbonylation process (25 MPa, 280 oC) with a production of 5, 000 tons per year from 1963 (Peng 2005).

The carbonylation refining was operated under atmospheric, medium or high carbon monoxide pressure. Figure 2.5 shows the schematic diagram of atmospheric carbonylation at INCO of Clydach (Boldt 1957, Teng-1 2006, Teng 2006).

Under atmospheric pressure, the impurities in crude nickel metal did not enter the gaseous phase. Iron forms volatile iron pentacarbonyl with very low rate, cobalt can form tetracarbonyl, Co2(CO)8, and tricarbonyl, Co4(CO)12, but both carbonyls have low volatility. Therefore, the extraction of nickel as carbonyl from a crude feed is a highly selective process.

18 At INCO refinery, Clydash (Wales), the first step was the reduction of oxides by hydrogen-water gas at relatively low temperatures 343 - 427 oC, the degree of reduction at these temperatures was about 90%. Reduced nickel reduced was carbonylated at 38 – 60 oC by carbon monoxide; the extent of carbonylation was approximately 95% after a 4 days reaction.

NICKEL OXIDES GRANULES Water Gas

Combustion Gas Copper Liquor To Atmosphere Water

Gas Water REDUCER H2 (3 in Series) CO ABSORBER CO 300 psig ABSORBER Producer 300 psig Gas Gas

CO to DUST COMPRESSER 2 SCRUBBER Atmosphere CO Air Gas SEPARATORS Impure Nickel Sludge to Granules Thickeners Air

Nickel Carbonyl Gas

Granules Make-up CO Gas

Cooling Gas Water Filter

Recycle Pellets

VOLATILIZER (6 in Series) Overflow Pellets SCREEN Hot Gases -5/16" PREHEATER

CO Gas DECOMPOSER

Residue to Sulfating Roast & Leach

NICKEL PELLETS TO MARKET

Figure 2.5 Flow sheet of nickel production using the atmospheric pressure carbonylation at INCO, Clydach (Teng-1 2006)

In order to improve the carbonylation process, the operation pressure was increased to enhance the reaction rate and stabilize the carbonyl at temperatures over 100 oC. Liquid nickel tetra-carbonyl produced under medium pressures was then vaporized and purified

19 by distillation. Finally, ultrapure nickel powder was produced from the decomposition of nickel tetra-carbonyl at atmospheric pressure and temperature above 180 oC.

Figure 2.6 shows the schematic flow sheet of nickel carbonylation under medium pressures at INCO (Teng 2006).

NICKEL OXIDES Hydrogen Enriched GRANULES Water Gas Combustion Gas To Atmosphere

REDUCER Water Producer Gas

Inert Gas DUST SCRUBBER Gas Air Sludge to STORAGE Thickeners HOPPERS

Inert Gas Reaction Gas Liquid Nickel (CO & Carbonyls) Carbonyl BATCH VOLATILIZER Cooling Hot Water Cooling (300 PSIG) Water Water VAPORIZER CO DISTILLATION Liquid co COLUMN BLOWER Carbonyls DECOMPOSER SEPARATOR Vapor Hot Gas Residue to Return CO Sulfating Roast COMPRESSOR & Leach Liquid Hot Water CO from Copper Liquor Plant REBOILER

Iron Carbonyl Residue

NICKEL POWDER TO MARKET

Figure 2.6 Flow sheet of nickel production using medium pressure carbonyl process at INCO, Clydach (Teng 2006)

High pressure carbonylation was used by the Norilsk company (Teng 2006). The carbonylation reactions were carried out at temperatures of 150-250 oC and pressures of 225 atm. About 95% of nickel was carbonylated after 3 days operation. Iron and cobalt were also partially converted to carbonyls, nickel tetracarbonyl was purified by rectification. 20 2.3.2 Carbonylation of Nickel

As mentioned above, nickel tetracarbonyl, Ni(CO)4, was firstly synthesized in 1890 (Mond et al. 1890); fine metallic nickel reduced from nickel oxide reacted with CO at about 100 oC under atmospheric pressure (reaction (2-13)), then gaseous nickel carbonyl was collected in cold container. In order to recover nickel, nickel carbonyl was decomposed to Ni and CO at about 180 oC (reaction (2-14)) (Mond et al. 1891). In industry, nickel carbonylation is produced at elevated temperatures and pressures to improve the reaction rate and yield. Nickel is recovered by decomposition of Ni(CO)4 at temperatures from 150 to 316 oC (Peng 2005).

Ni + 4CO = Ni(CO)4 (2-13)

Ni(CO)4 = Ni + 4CO (2-14)

(Goldberger et al. 1963) studied the kinetics of nickel carbonyl formation from pure carbon monoxide and reduced nickel powder in a fixed bed at temperatures between 25 and 150 oC, CO pressures up to 4 atm, and gas velocities from 1.17 cm·s-1 to 1.92 cm·s-1.

The increase in CO pressure had a positive effect on the reaction; the maximum carbonylation rates were obtained at about 75 oC under all pressures. Surface chemical reaction was found to control the overall reaction rate. Experimental data were described by a modified first-order rate equation (2-22) o as ∗ In o = k (P − P )CO ∗ θ (2-15) as −x

Where, θ is reaction time, minute; P is partial pressure of CO, atm; P* is equilibrium

o partial pressure of CO, atm; x is conversion or fraction of nickel; a s is activity of nickel.

(Wang et al. 2009) discussed the effect of ore roasting, particle size, CO partial pressure, reaction temperature, flow rate, sulphur and hydrogen sulphide on the carbonylation of nickel. Ferronickel alloy re-activated with H2 is essential in the process, the optimum parameters for the carbonylation reaction were temperature 70oC and CO pressure of 9.9 atm; addition of sulphur and hydrogen sulphide greatly increased the carbonylation rate.

21 (Mazurek et al. 1982) established that the lowest CO pressure in production of Ni(CO)4 was 667 Pa, and increasing CO pressure could enhance the reaction rates. In order to check whether the reactivation by H2 was from the removal of O2 (the reactivation mechanism was supposed to remove oxygen adsorbed on the nickel surface before o carbonylation), samples in their work were heated at 400 C in H2 atmosphere for 4 hours prior to the carbonylation; once the carbonylation rate was steady, H2 was introduced to the system instead of CO for 30 minutes at carbonylation temperature to check whether the reaction rate is boosted again. It was suggested that the heat treatment of metal sample rather than the removal of oxygen from the surface promoted the carbonylation reaction.

(De Groot et al. 1980) in a study of nickel carbonylation at temperatures up to 160 oC observed the maximum reaction rate at 125oC. The effect of temperature on the formation of nickel carbonyl is shown in Figure 2.7; Figure 2.8 demonstrates the effect of CO pressure on the rate of formation of nickel carbonyl.

(111) SURFACE -3

-4 (a)

-5 (b) Log(Ni(CO)4 pressure) Log(Ni(CO)4

40 80 120 160 200 Temperature (oC)

The reaction rate increased with carbon monoxide pressure at an apparent order of 1.45 (Figure 2.8).

22 Figure 2.8 Effect of CO pressure on the rate of nickel carbonylation at 20 oC (De Groot et al. 1980)

A study of nickel carbonylation using chemisorption (Schmidt et al. 1982) showed that CO mainly adsorbs in the form of linearly and bridge-bonded molecules Ni(CO) and

Ni2(CO). The intermediate Ni(CO)2 complex was assumed to be formed by reaction (2-

16) when the coverage of nickel surface by CO, θ1 , was over 0.5, which represents the fraction of Ni surface atoms covered by linearly bonded CO.

3Ni(CO)  Ni(CO)22 + Ni (CO) (2-16)

 The adsorption equilibrium constant for the reaction Ni2 CO + CO 2Ni(CO) was

2 calculated as KP211 /CO (1 ) . A reaction order for this process was found equal to 1.5.

Further study confirmed that nickel carbonylation was more readily taken place on step atoms than perfect crystal surface atoms (Schmidt et al. 1982).

In a study of the carbonylation of Cu-Ni alloy (Teng et al. 2007) at temperatures between 120 and 150 oC under CO pressure in the range 7 ~ 18 MPa, the recovery of Ni after 42- hour reaction was about 95%; the reaction rate increased with the decrease of 23 particle size from 5 to 0.5 mm. The rate limited step was found to be the diffusion of nickel tetracarbonyl in the gas phase (Teng et al. 2007).

The maximum rate of the carbonylation reaction in work (Goldberger et al. 1963) was observed at 117 oC; the reaction rate was found to be a function of reaction time and surface composition (coverage by C, S and O).

In industry, the carbonylation is performed in a fixed bed reactor in vertical columns under a gauge pressure of 200-250 atm. For a complete recovery, the size of Ni grains should not exceed 10 ~ 15 µm (Kozyrev et al. 2005)

2.3.2 Carbonylation of Iron

Iron pentacarbonyl (Fe(CO)5) was synthesized by reaction (2-17) in 1891. Iron carbonyl decomposes to iron and CO at 180 oC by reaction (2-18) (Mond et al. 1891, Mond et al. 1891, Dewar et al. 1905).

Fe + 5CO = Fe(CO)5 (2-17)

Fe(CO)5 = Fe + 5CO (2-18)

Mond (Mond et al. 1891-1) targeted production of iron carbonyl by reaction of fine iron o reduced from iron oxide and reactivated by H2 at 300 C, and carbon monoxide, at a temperature of 180 ~ 220 oC under CO pressure of 150 ~ 250 atm for 24 hours. The reaction rate was very slow, only about 1% carbonyl was collected (Mond et al. 1891-1).

Effects of temperature (100-300 oC) and pressure (100 - 300 atm) on synthesis of iron pentacarbonyl studied by Mond and Wallis (1922) are shown in Figure 2.9. The o optimum temperature for production of Fe(CO)5 was found to be near 200 C, above which carbonyl decomposes to iron and CO (Mond et al. 1922).

24 15 14 13 12 11 10 9 300 8 7 200 6 5 150 4 3

Cubic centimetre of Fe(CO)5 100 2 1 0 100 150 200 250 300 Temperature Figure 2.9 Effect of temperature and pressure on synthesis of iron pentacarbonyl (Mond et al. 1922)

2.3.3 Carbonylation of Cobalt Cobalt carbonyl was synthesized by reaction (2-19) at 50 oC under CO pressure 150 atm in 1908 (Mond et al. 1908, Mond et al. 1908)

2Co 8CO Co28 (CO) (2-19) Cobalt carbonyls are large molecules, which are not readily vaporized under these experimental conditions; the carbonylation reaction was greatly enhanced by raising the pressure to 200 atm and the temperature to 200oC (Mond et al. 1908, Mond et al. 1908).

Compared with pure cobalt, the carbonylation rate of cobalt from the alloy with nickel (88-94 wt% Ni) increased 83 times. The extraction of cobalt was affected by nickel and iron carbonylation as a result of breakdown of the crystal lattice with removal of nickel and iron, or the catalytic and autocatalytic action of the formed carbonyls (Kipnis et al. 1973). Cobalt carbonyl nitrosyl, which is gaseous at room temperature, can be synthesized by reaction of cobalt tetra carbonyl with nitric oxide (reaction (2-20)) at 75-80 oC (Ivanova et al. 1999):

Co(CO)43 NO  Co(CO)  NO  CO (2-20)

25 Cobalt carbonyl nitrosyl can also be prepared by the method developed by Blanchard et al. (1934) in an alkaline suspension of cobaltous cyanide (Brockway et al. 1937) by the following reactions (Coleman et al. 1936, Ewens et al. 1939, Blanchard et al. 1940):

2CoCl2  11CO  12KOH  3K2 CO 3  2KCo(CO)4  4KCl (2-21)

1 KCo(CO) NO  H O  Co(Co) NO CO  KOH  H (2-22) 4 2 3 2 2

2.3.4 Factors Affecting the Carbonylation

The carbonylation processes are affected by temperature, carbon monoxide pressure, gas flow rate, particle/grain size. Sulphur containing compounds have a catalytic effect on the reaction. Carbon and reactivation of metals by hydrogen were also reported to influence the carbonylation reactions.

Effect of Temperature In the synthesis of Ni, Fe and Co carbonyls in Mond’s experiments, temperature was a critical parameter (Abel 1963). The formation of carbonyls was reported to be accelerated at elevated temperatures in many papers (Windsor et al. 1933, Hieber et al. 1950, Podall 1961, Abel 1963, Nametkin et al. 1975, Liang et al. 1983, Terekhov 2001, Kozyrev et al. 2005, Teng et al. 2007). Although the carbonylation of nickel, iron and cobalt could occur at temperatures below 100 oC, commercial carbonylation process is carried out at relatively high temperatures and pressures to increase the reaction rate and yield; in work by (Kipnis et al. 1973) carbonylation temperature was 200 oC under CO pressure 200 atm.

The effect of temperature on nickel carbonylation has been reported in many papers, the optimum temperature was found to be 70 oC at 10 atm (Wang et al. 2009), 75 oC (temperature investigated between 25 and 150 oC) at CO pressures up to 4 atm (Goldberger et al. 1963), 77 oC (temperature investigated between -3 and 147 oC) under pressures below 3*10-2 Pa(Liang et al. 1983), 117 oC(Goldberger et al. 1963) and 125 oC (in a study at temperatures up to 160 oC) (De Groot et al. 1980).

Effect of Carbon monoxide pressure

26 Carbon monoxide pressure was another important parameter in Mond’s experiments on synthesis of metal carbonyls (Abel 1963). The carbonylation reaction is favoured by the high CO gas pressure (Medvedev et al. 1998). High pressure of carbon monoxide was reported to increase the rate of carbonyl formation on a compact metal surface (Windsor et al. 1933). The increase of carbon monoxide pressure enhanced the reaction rate in work by Wang et al. (2009). Effect of CO pressure on catalytic carbonylation of nickel is shown in Figure 2.10 (Wang et al. 2009). Initial rate of carbonylation decreased with increasing CO pressure from 10 to 20 atm, although it was significantly higher at higher CO pressure after approximately 2.5 hours reaction. In industry, the carbonylation is performed under a gauge pressure of 200-250 atm (Kipnis et al. 1973, Kozyrev et al. 2005).

Figure 2.10 Effect of CO pressure on catalytic carbonylation of nickel at 70 oC (2 wt% sulphur added before carbonylation); a: atmospheric pressure, b: 9.9 atm, c: 19.8 atm

Catalytic effect of sulphur

Sulphur and inorganic sulphur compounds (FeS, NiS and H2S) can catalyse the carbonylation of nickel and iron (Heinicke et al. 1963, Heinicke et al. 1970).

Figure 2.11 shows the effect of hydrogen sulphide content in CO gas on the nickel carbonylation rate (Heinicke et al. 1963). The reaction rate increased sharply with increasing hydrogen sulphide content up to 1.0 vol%, further increase in the H2S concentration had a minor effect on the carbonylation rate.

27 100 ) -1 90

80 30 minutes 30  2 70 10  4 60

10 Reaction yield (mmol Ni(CO) (mmol yield Reaction 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

% H2S o Figure 2.11 The effect of H2S on carbonylation of nickel at 25 C and 1 atm (Heinicke et al. 1963)

Effect of hydrogen sulphide on the rate of carbonylation of nickel and iron studied by (Heinicke et al. 1970) is shown in Figure 2.12.

(Windsor et al. 1933) studied nickel carbonylation with addition of iron sulphide. They suggested that the catalytic effect of sulphur is related to formation of nickel tetracarbonyl from nickel sulphide and carbon monoxide.

28 25 Ni 20

15 CO

CO + 3 % H2S 10

0 0 5 10 15 20 25 30 35 40 45 50 55

14 Fe 12

Carbonyl yield (mmol/h) yield Carbonyl 10

8 CO

6 CO + 3 % H2S 4 2 0 0 5 10 15 20 25 30 35 40 45 50 55 Processing time (hour)

o Figure 2.12 Influence of H2S on iron and nickel carbonylation at 25 C and CO pressure 1.1 atm (Heinicke et al. 1970)

(Den Besten et al. 1962) suggested the following mechanisms of catalytic effect of sulphur on the nickel carbonylation. Hydrogen sulphide is chemisorbed on nickel surface and dissociated to hydrogen and sulphur (Figure 2.13) Hydrogen is adsorbed on available surface sites, but at high surface coverage, hydrogen is forced out to the gas phase. Sulphur reacts with nickel forming Ni-S complexes as shown in Figure 2.13, which is easily converted to nickel carbonyl.

29 Figure 2.13 Chemical adsorption of H2S on nickel surface (Den Besten et al. 1962)

However, the increase of sulphur content results in a decrease of the rate and extent of carbonylation (Kipnis et al. 1973, Miura et al. 2008). Conversion of cobalt to carbonyl decreased with increasing content of sulphide phases as shown in Figure 2.14 (Kipnis et al. 1973), what was related to the binding of cobalt in sulphide form.

100 quenched at 550 oC cast 80 quenched at 900 oC Alloys (Cu:Ni= 1:9) 200 oC 200 atm 10 h 60

20 Degree of Co extraction (%)

0 0 2 4 6 8 10 12 S content (%)

Figure 2.14 The effect of sulphur on the carbonylation of cobalt at 200 oC and CO pressure 200 atm for 10 hours (Kipnis et al. 1973)

In the synthesis of iron carbonyl in work (Hieber et al. 1950) , the addition of a small amount of sulphur (approximately 1 wt%) had a remarkable positive effect on the formation of iron pentacarbonyl at 200 oC and CO pressure of 200 atm, which gradually diminished with increasing sulphur addition. The effect of sulphur content on the carbonylation of iron is shown in Figure 2.15.

30 100

20 Extraction proportion of Fe (%)

200 atm; 200 oC; 15 hour 0 0 10 20 30 40 50 60 70 80 90 100 Addition of S (Atom - %)

Figure 2.15 The effect of sulphur on iron carbonylation at 200 oC and CO pressure 200 atm (Hieber et al. 1950)

A similar effect of sulphur addition was observed in carbonylation of nickel in the Ni- Fe alloy, which is shown in Figure 2.16 (Wang et al. 2009).

100

40 Yield (%)

2% S, CO 20 5% S, CO 10% S, CO 0% S,CO + H2S 0% S,CO + H2S (added intermittently) 0 0 4 8 12 16 20 24 Time (hr) Figure 2.16 Effect of sulphur on nickel carbonylation at 70oC and CO pressure of 10 atm (Wang et al., 2009)

Miura et al. (2008) used sulphur as a catalyst in the extraction of iron from the scrap (powder) of Nb-Fe-B sintered magnets via the carbonylation; reactions were implemented at 200 oC under CO pressure of 30 MPa for 24 h. Effect of sulphur addition on the carbonylation of iron is shown in Figure 2.17. The yield of iron carbonyl Fe(CO)5 was evaluated to be ~ 56% without pre-treatment. In other

31 experiments, iron produced by the disproportionation of Sm2Fe17 in hydrogen (HD) (Sugimoto et al. 2002), reacted with carbon monoxide under the same conditions, a maximum yield of 92% was achieved. Both the results are shown in Figure 2.17.

100

20 Extent of carbonylation (%) Extent As-obtained HD 0 0 2 4 6 8 10 12 14 16 18 20 S/[S+Fe] (%)

Figure 2.17 Effect of sulphur on extraction of iron from Nb-Fe-B via carbonylation

Effect of Oxygen

The effect of oxygen on nickel carbonylation is conflicting (Lascelles et al. 1983) reported that oxygen had a positive effect on nickel carbonylation as shown in Figure 2.18. It was suggested that oxygen is chemisorbed on the nickel surface and incorporated into the nickel lattice, which activated nickel atoms in the carbonylation reaction. However, Mazurek et al. (1982) reported that reactivation of metals by hydrogen before carbonylation removed oxygen.

32 6.0E-9 ) -1  min -1 4.0E-9

 cm (b)

2.0E-9

(a) Reaction (mole rate

0.0 0 20 40 60 80 100 Time (minutes) Figure 2.18 Effect of oxygen on the carbonylation of Ni (110) at 130oC. (a) 1ppm oxygen in carbon monoxide, (b) 3ppm in carbon monoxide (Lascelles et al. 1983)

Reactivation of Metal Surface by Hydrogen Hydrogen was used to reactivate the metal before carbonylation by heating metal in hydrogen (Windsor et al. 1933). (Wang et al. 2009) annealed a metal sample in hydrogen atmosphere at 400oC. Ni-Fe alloy was reactivated by hydrogen at 150oC for two hours before the introduction of carbon monoxide. Figure 2.19 shows the effect of reactivation of nickel by hydrogen at 150 and 250 oC on the reaction of nickel carbonylation (Wang et al. 2009). Heating of nickel in a reducing atmosphere removed adsorbed oxygen, which inhibits nickel carbonylation (Germer et al. 1962).

100

Yield (%) 40

t : reactivation temperature ℃ ℃ r tr 150 , tc 70 , CO+H2S

t 250 ℃ , t 70 ℃ , CO+H S tc: carbonylation temperature r c 2 0 0 5 10 15 20 Time (hr)

33 Figure 2.19 Carbonylation of nickel with pre-treatment by hydrogen at 150 and 250 oC (Wang et al. 2009)

Effect of Particle size The decrease of nickel particle size has a positive effect on the carbonylation of nickel. The extent of carbonyl formation on a compact metal surface was extremely small; very fine subdivision of the metal increased the yield of carbonylation (Windsor et al. 1933). In Mond’s experiments (Mond et al. 1891-1), iron carbonyl was prepared using iron fines reduced from iron oxide. The decrease of particle size increased the specific surface area, the contacting area between carbon monoxide and metal.

Effect of Carbon In carbonylation of metals at high temperature, carbon can be deposited on the metal surface as a result of partial decomposition of carbon monoxide. (Kozyrev et al. 2005), added carbon to nickel-copper-cobalt alloy to study the effect of carbon on cobalt carbonylation. It was reported that carbon inhibited carbonylation of cobalt. The effect of carbon on carbonylation of cobalt is shown in Figure 2.20 Increase in carbon concentration led to a dramatic decrease in the cobalt recovery from the alloy (Kozyrev et al. 2005).

Mass recovery of Co (%) 4

0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Mass content of carbon in alloy (%) Figure 2.20 Effect of carbon content on cobalt carbonylation at 180 oC and CO gauge pressure 70 atm (Kozyrev et al. 2005)

34 2.4 Reduction of Metal oxides

Metal oxides are reduced to the metallic state prior to carbonylation. This project included reduction of nickel, iron and cobalt oxides by hydrogen and selective reduction of laterite ore by the carbon monoxide - carbon dioxide mixture. However, a study of the reduction processes was outside the scope of this project; they were investigated in the PhD project “Selective reduction of laterite ore” by Jun Yang (Yang, 2014). This section presents a brief overview of literature on the reduction of nickel, iron and cobalt oxides by hydrogen, which was studied in numerous papers.

Reduction of oxides at low temperature was suggested by Mond et al. (1890)(Mond et al. 1890) to minimise sintering of reduced particles.

Complete reduction of iron oxide by hydrogen in work by (Kawasaki et al. 1962) was achieved within 60 minutes at 700 oC; at 345 oC, reduction took 120 min (Sastri et al. 1982). Reduction of hematite at temperatures below 570 oC occurred by the subsequent steps Fe2 O 3 Fe 3 O 4 Fe (Jozwiak et al. 2007). Hematite with grains smaller than 5μm were completely reduced in 10 minutes at temperatures above 550oC. Rising temperature accelerated the reaction rate up to 900oC (Wagner et al. 2008). Lin et al. (2003) first reduced hematite to magnetite at 297 oC, which was further converted to iron at 497 oC. Reduction of nickel oxide was achieved at temperatures from 250 to 350 oC (Rodriguez et al. 2002, Janković et al. 2008). Nickel oxide was completely reduced at 300 oC by hydrogen with gas flow rate of 3.0 L·min-1 for 24 hours (Parravano 1952);

Rashed and Rao (1997) reduced nickel oxide by hydrogen at temperatures over 340 oC. Jeangros et al. (2013) reported NiO reduction at temperatures between 350 and 400 oC; nickel grains size increased when the temperature exceeded 600 oC. The reduction rate increased with increasing temperature from 400 to 600 oC (Utigard et al. 2005), and decreased noticeably with further increase in the temperature; reduction became very fast at temperature above 950 oC.

35 2.5 Summary and Project’s Objectives

Industrial production of nickel from laterite ores is gaining momentum. Processing of laterite ores, started in Australia in 1998, has significantly enhanced Australian role on the international nickel market.

New technology development in processing of laterite ores has been mainly in the direction of hydrometallurgical processes. However, in general, hydrometallurgical processing of laterite ores is complex, involves high capital investment, consumes huge amount of hazardous chemical materials such as ammonia and sulphuric acid, and causes significant environmental concerns. Hydrometallurgical processing of laterite ores is economically sustainable only when cobalt value of the ore is also recovered.

An attractive approach to processing the low nickel grade laterite ores was undertaken in the ARC Discovery project DP1094880 “A Novel Approach to Processing of Australian Laterite Ores through Selective Reduction and Carbonylation”.

This PhD project studied carbonylation of selectively reduced laterite ore.

Carbonylation technologies have been used in the refinery of nickel for many decades. However, literature data on kinetics and mechanisms of carbonylation reactions, effects of temperature, CO pressure and other parameters, catalytic effect of sulphur are limited and often inconsistent. The aim of this project is a comprehensive study of carbonylation of nickel and iron reduced from pure oxides and selectively reduced laterite ore.

The project includes a study of non-catalytic carbonylation of pure Ni and Fe, and nickel-iron mixture; catalytic carbonylation of these metals with addition of FeS, S and

H2S; non-catalytic and catalytic carbonylation of reduced laterite ores.

36 Chapter 3 Experimental

The project involved an experimental study of reduction of nickel and iron oxides, carbonylation of nickel, iron and reduced laterite ores, measurements of metals concentrations in aqua regia by ICP-OES, and characterization of reduced samples and residue of samples after carbonylation by XRD, XRF and SEM/EDS.

3.1 Materials

Chemicals and gases used in the project are listed in Table 3-1.

Table 3-1 Chemicals and gases

Chemicals Specification Application

Carbonylation after reduction Nickel oxide (NiO) Powder, - 45μm, 99% by H2 Carbonylation after reduction Iron oxide (Fe2O3) Powder, < 5μm, > 99% by H2

Sulfur (S8) Powder, 99.9% Catalyst used in carbonylation Iron Sulfide (FeS) Powder, - 150μm Catalyst used in carbonylation Ultra high purity, Reaction gas for reduction; Hydrogen (H2) 99.999% activation gas for carbonylation Ultra high purity, Nitrogen (N2) Purging gas 99.999% Reactant gas for carbonylation Carbon monoxide (CO) High purity, 99.99% reducing gas for laterite ores

A gas component for selective Carbon dioxide (CO2) High purity, 99.99% reduction of laterite ores Hydrochloric acid Reagent for preparing aqua Concentration 32% wt% (HCl) regia

37 Reagent for preparing aqua Nitric acid (HNO3)) Concentration 70% wt% regia Sodium hydroxide Absorbent of corrosive gas from Pellets, lab research grade (NaOH) the outlet

-1 Deionized water (H2O) Resistivity 18 Ω∙cm Diluting reagent

Glycerol Analytical grade Fluid for the oil bath

Nickel nitrate ICP-OES standard 2% Standard in the ICP-OES

(Ni(NO3)2) HNO3 analysis ICP-OES standard 2% Standard in the ICP-OES Iron nitrate (Fe(NO3)3) HNO3 analysis Cobalt nitrate ICP-OES standard 2% Standard in the ICP-OES

(Co(NO3)3) HNO3 analysis ICP-OES standard 2% Standard in the ICP-OES Sulfuric acid (H2SO4) HNO3 analysis

BCS and MIN garnieritic laterite ores were provided by CSIRO Australia. The composition of the ores analysed by XRF and ICP-OES at the Analytical Center of UNSW is shown in Table 3-2. Effect of the particle size on carbonylation reactions was studied for the selectively reduced MIN ore; effects of other parameters (temperature, CO pressure and flow rate) on carbonylation of Ni and Fe was examined for the selectively reduced BCS ore.

Table 3-2 Composition of MIN and BCS garnieritic laterite ores (wt%)

BCS Ore MIN ore Element 53 ~ 200 μm <53 μm 38 ~ 53* 75 ~ 90* 140 ~ 200* 355 ~ 495* Ni 1.35 1.66 2.37 2.12 2.094 1.684 Mg 12.6 13 8.59 6.72 7.14 7.99 Si 20.5 19.4 - - - - Fe 10.4 11.7 11.83 10.68 10.57 8.90 Co 0.038 0.045 0.14 0.14 0.14 0.13

38 Al 2.2 2.1 - - - - Cr 0.5 0.5 - - - - Mn 0.2 0.2 - - - -

The experimental facilities and glassware used in the study of reduction and carbonylation processes are listed in Table 3-3.

Table 3-3 Facilities and glasswares

Facilities/ glassware Specification Application

Horizontal furnace Max. 1000 oC A furnace for reduction experiments Oil bath Max. 300 oC Carbonylation temperature control Dreschel bottle Vol. 125 ml, 250 ml Absorbent container Volumetric flask Vol. 1000 ml Diluting liquid sample

Measuring cylinder Vol. 200 ml Prepare aqua regia

Powder drill - Stirrer in reaction Swagelok flex tube SS-12BHT-36 Carbonylation reactor

Concentrate nitric acid solution and aqua regia were tested as the absorbent for nickel and iron carbonyls in preliminary experiments. Nitric acid solutions, 30 wt% and 50 wt%, were prepared by diluting concentrate nitric acid (70 wt%) by deionized water; aqua regia was obtained by mixing nitric acid (70 wt%) with hydrochloric acid (32 wt%) with the volume ratio of 1 to 3. Absorption solution (20 wt% sodium hydroxide solution) for the outlet gas, which corroded the brass connector, was made by dissolving 40- g pellet of sodium hydroxide into 160 ml of water.

3.2 Sample preparation

The project studied carbonylation of pure Ni and Fe, Ni-Fe mixture, Ni and Fe mixed with catalyst FeS and S, and reduced laterite ore.

Pure Ni and Fe were produced by reduction of metal oxides NiO and Fe2O3 by hydrogen. Nickel-iron mixture was obtained by mixing of Ni and Fe. Nickel oxide (NiO)

39 and iron oxide (Fe2O3) were reduced separately by hydrogen and then mixed in an agate mortar for about 20 minutes.

Iron sulphide was mixed with nickel and iron oxides in an agate mortar for about 15 minutes before reduction, degree of reduction of iron sulphide by hydrogen under experimental conditions is expected to be very low. Equilibrium constant for the reaction of iron sulphide with hydrogen, shown in Figure 3.1 is below 1.82*10-3 at 800 oC.

Sulphur has low melting and boiling temperatures, 115.21 oC and 444.6 oC respectively, therefore it could not be added to metal oxides before reduction. Sulphur was grinded and mixed with samples of metallic nickel or iron, or Ni-Fe mixture in an agate mortar for about 15 minutes.

68 -2

66 -4 64

62 -6 (kJ) o

G 60  Log(k) 58 -8

56 -10 54

52 -12 0 150 300 450 600 750 900 Temperature (oC)

Figure 3.1 The standard Gibbs free energy ∆Go and equilibrium constant (presented as log(k)) for the reaction of iron sulphide with hydrogen

Concentration of Ni and Fe in aqua regia was too high, more than 3000 mg·L-1, to be used in the ICP-OES analyses. In order to be analysed, the solution samples taken from the Dreschel bottles were diluted 10 to 50 times to 10 ml by 2 wt% aqua regia in a plastic tube. After carbonylation, the residual solution of aqua regia was diluted to 1000 ml in a volumetric flask and analysed by ICP-OES.

40 3.3 Experimental Set-up and Procedures

Stages in the experimental study of carbonylation of Ni and Fe are schematically shown in Figure 3.2. This section describes experimental set-ups and procedures for reduction and carbonylation processes.

Figure 3.2 Stages in the experimental study of carbonylation of Ni and Fe

3.3.1 Reduction Experiments Metal oxides were reduced by hydrogen in a horizontal furnace (Ceramic Engineering,

Sydney, Australia) at atmospheric pressure. In the reduction of laterite ore, the CO-CO2 gas mixture was used for selective reduction of metal oxides. Experimental setup is schematically shown in Figure 3.3.

Outlet

Flow meter Sample

Furnace

N2 H2

Figure 3.3 SchematicFig. experimental 1 Flow diagram set-up for for the reduction reduction of of Ni metal and oxides Fe oxides and laterite ore

The gas flow meter and thermocouple of the horizontal furnace were calibrated prior to the experiment study; the calibration results are shown in Appendixes A and B. Reduction was examined under conditions presented in Table 3-4.

41 Table 3-4 Parameters in the reduction process

Variable Unit Value

Temperature oC 400~900

Flow rate L·min-1 1

Content of CO in the CO- Vol% 60 CO2 gas mixture

6h in reduction of Ni and Fe Time hour oxides 4h in reduction of laterite ore

In the reduction, gas was supplied from gas cylinders of nitrogen, hydrogen, carbon monoxide and carbon dioxide. The gas flow rate was controlled by a needle valve and monitored by a gas flow meter. Oxides and ores were reduced in a horizontal furnace, which isothermal zone was determined in preliminary experiments by measuring the furnace temperature profile.

Samples in quartz boat were placed in the isothermal zone of the furnace, and heated in nitrogen (flow rate 0.5 L· min-1) at a rate of 5 oC · min-1. When the experimental temperature was reached, nitrogen was switched off and hydrogen or CO- CO2 gas mixture was introduced to the system. The reduction time was 6 hours in the reduction of Ni and Fe oxides, and 4 hours in the reduction of laterite ore. Samples were cooled to the room temperature in nitrogen.

3.3.2 Carbonylation Experiments Carbonylation experiments were conducted in a flexible U – shaped reactor (max pressure 68 atm) immersed into the oil bath. The schematic setup for carbonylation experiments is shown in Figure 3.4. Carbonylation was studied using CO at different pressures. The reaction temperature was controlled by the oil bath and monitored by three thermocouples placed in the centre of the reactor, outside the reactor and at the bottom of the oil bath. Calibration of the thermocouple for measurement of temperature

42 of the oil bath is shown in Appendix C. The outlet gas flow rate was adjusted by a needle valve and monitored by a gas flow meter. Pressure was regulated by clockwise rotating the adjustment screw/handle to increase the CO pressure to its desired value. Carbonyls were absorbed using two Dreschel bottles filled with aqua regia. The third absorption bottle filled with 20 wt% sodium hydroxide solution was used to treat the corrosive chlorine gas (Princeton 2014) from Dreschel bottles.

The carbonylation of metals was examined under conditions listed in Table 3-5.

Table 3-5 Parameters in carbonylation experiments

Variable Unit Value Temperature oC 80 ~ 100 Pressure atm 0 ~ 56

Gas flow rate L·min-1 0.1 ~ 0.5

Sample weight gram 0.8 ~ 3.2 Particle size μm 38 ~ 495

The flexible U-shape tube (internal layer: Teflon; outer layer: stainless steel) was purged by compressed air to remove residue powder and then by nitrogen to remove air prior to loading samples. Samples were inserted into the reactor, which was placed in the oil bath and connected to the gas system. A sample after reduction was exposed to air when it was moved to the carbonylation reactor. To reduce oxides which could be formed on the surface of reduced samples, they were heated to 150 oC in hydrogen for 2 hours. Cold glycerol (room temperature) was added into the oil bath to decrease the temperature to the experimental value for the carbonylation reaction. Afterwards, carbon monoxide was introduced into the reactor; CO pressure was raised to the experimental value at a specific flow rate. During the experiment, samples were taken from the solutions in Dreschel bottles at different times and diluted to 10 ml to be analysed by ICP-OES. When the experiment was finished, the CO gas was switched off

-1 and the gas flow rate of H2 was increased to 1L·min , hydrogen was introduced to the reactor to remove residue CO and carbonyls; the system was purged with hydrogen for

43 20 minutes at a flow rate of 1 L·min-1 at atmospheric pressure. Aqua regia in Dreschel bottles was diluted with ionised water to 1000±2 ml and examined by ICP-OES.

I II III

V b

a c IX VI N2 IV

VII Outlet e d

a: Oil Bath b: Stirrer c: Flow Meter d: Absorption Bottles e: Absorption bottle I ~ VIII: Valves I: CO Gas Cylinder Valve II: CO Gas Cylinder Valve Regulator III: CO Break Valve IV: N2 Break Valve V: Three Way Valve VI: Needle Valve VII: Three Way Valve

Figure 3.4 Schematic setup for carbonylation experiment

3.4 Analytical Equipment and Methodology

3.4.1 X-Ray Diffraction (XRD) Analysis The XRD analysis was conducted using the PANalytical Xpert Multipurpose X-ray Diffraction System (MPD), which is a theta to theta goniometer system used for powder samples.

The MPD system is equipped with the following components:

1) Incident beam path–Programmable divergence slit, which can be configured as a fixed slit or a variable slit, masks.

2) Diffracted beam path–Programmable anti-scattering slit and receiving slit, curved monochromators and proportional detector.

3) Automatic multi-sample changer (15 samples) can increase productivity and a sample spinner is to eliminate the preferred orientation.

44 The sample analysed by MPD needed to be grounded into powder of size less than 10μm in order to obtain a random orientation of material. The powdered sample was closely packed into the sample holder (10mm diameter) with flat surface, and then loaded into an automatic sample changer for XRD analysis.

The XRD analysis in this project was conducted with the following parameters:

 X-ray source: copper Kα radiation with λ=1.54 Å

 Voltage=45 kV; Current=40 mA

 Scan range (°): 10-100

 Scan step (°): 0.0131303

 Time per step (s): 25.50

 Net time per step (s): 23.97

 Scan speed (°/s): 0.131303

 No. of steps: 6855

 Total scanning time: 11 min 53 sec

 Divergence slit (°): 0.5

 Diffraction beam (°): 1

3.4.2 ICP-OES/MS Analysis An Inductively Coupled Plasma (ICP) is generated by heating a flow of argon gas through a torch fitted with a radio frequency induction coil. A dissolved sample is nebulised and fed into the plasma (at up to 10,000 K) where the elements in the sample ionise and emit characteristic radiation. The quantity of elements, Ni and Fe, can be simultaneously measured by detecting the characteristic emission of the element (ICP- AES).

Inductively Coupled Plasma - Optical Emission Spectroscopy (ICP-OES) (also known as ICP-AES Inductively Coupled Plasma - Atomic Emission Spectroscopy) is one of the most commonly used methods for the determination of trace concentrations of elements in samples. The detection limits are generally in the ppb range. The upper limit for a particular emission line is usually 104 to 106 times the detection limits. The precision of

45 analysis is usually in the 1-2% RSD (Relative Standard Deviation) range. Better precision can be obtained with trade-offs in speed. It is capable of multi-element, qualitative and quantitative analyses.

During the analysis, a work line was calibrated from standard solution in the range from

10-2 to 102 mg·L-1. Concentration of elements in the analysed sample was calculated based on the work line.

The operation parameters are listed in Table 3-6.

Table 3-6 ICP-OES analysis parameters

Nebulizer Auxiliary Plasma RF power Sample pump flow Plasma Nebulizer flow flow flow viewing W L·min-1 L·min-1 L·min-1 mL·min-1 Ar 1300 0.8 0.2 15 0.50 Axial

Background Processing Auto Read delay Rinse Spray Replicates correction mode integration chamber s s points Area Area 30 15 3 7 3

3.4.3 Scanning Electron Microscopy (SEM)/Energy Dispersive X-ray Spectrometer (EDS) The morphology of samples was examined by the HITACHI S3400 Field Emission Scanning Electron Microscope (FESEM). This instrument is fitted with secondary and backscatter electron detectors that allow for topographic and compositional (atomic number contrast) surface imaging of samples. It is used for imaging up to 3000X magnification. Elements in the samples can be identified by the linked X-ray Energy Dispersive Spectrometer (EDS).

Because the metal samples were conductive, they were adhered directly to the carbon tape and analysed. The samples of laterites were coated with gold and then tested.

3.4.4 X-Ray Fluorescence (XRF) Spectroscopy X-ray fluorescence (XRF) spectroscopy uses an X-ray source to eject core-shell electrons from an atom to create an excited state. The resulting cascade of electrons to 46 fill the holes results in emission of X-radiation from the atom (fluorescence) that has a characteristic wavelength/energy specific to each element. The fluorescence can be quantified to enable elemental analysis from the ppm level.

In this project, laterite ores samples were grinded to fine particles (< 300 mesh) and then dried at 200 oC for 24 hour; after the preparation, samples were submitted to the XRF Lab of the UNSW Analytical Centre for analysis.

3.5 Selection of the absorption reagent

The extent of nickel and iron carbonylation was calculated from the concentration of these metals in the absorbent measured by ICP – OES. Metal carbonyls are readily absorbed by an oxidative acid, such as concentrated sulphuric acid and nitric acids (Mond et al. 1891, Mond et al. 1891-1). Sulphuric acid was excluded in this project because sulphur containing catalysts were used in the carbonylation process. The alternative absorbents tested in this project were concentrated nitric acid and aqua regia. Results of the test presented in Appendix D, show that iron pentacarbonyl was completely absorbed by the concentrated nitric acid (50% wt) and aqua regia, while the absorption of nickel tetracarbonyl was more difficult. Two Dreschel bottles with aqua regia were needed for absorption of nickel carbonyl. To confirm this conclusion, the extent of nickel carbonylation was calculated from the weight loss and ICP – OES analysis. Results compared in Figure 3.5 demonstrate good agreement between two methods, what confirms that aqua regia is an appropriate absorbent for both, nickel and iron carbonyls.

47 100

Recovery (%) 40

20 From ICP-OES From weight loss

0 80 84 88 92 96 100 Temperature (oC)

Figure 3.5 Extent of nickel carbonylation measured by the weight loss and by the ICP- OES analysis of nickel in aqua regia

3.6 Data Analysis

3.6.1 Definition of extent of carbonylation and the content of catalyst Extent of metal carbonylation, y, is defined as the extent of metal converted to metal carbonyls (percentage)t, it was calculated using equation (3–1).

mass of converted metal y  * 100 (3–1) initial mass of metal

Another important factor in the catalytic carbonylation is the content of catalysts, it is defined as follows.

mass of sulfur element in catalyst Content = * 100 (3–2) mass of a sample

3.6.2 Extent of metal carbonylation As described, metal carbonyls were absorbed in two tandom Dreschel bottles, and about 7 ml of solution were taken from each bottle (I 125 ml, II 250 ml) during the experiment.

48 Metal in the absorption bottles and sample tubes taken out for ICP-OES were calculated, and the latter one should be accumulated. Because samples used in ICP-OES were all diluted to 10 ml, hence a dilution factor should be considered to get the concentration of metal in Dreschel bottles.

The extent of metal carbonylation was calculated from results of the ICP-OES analysis as follows:

(**ciI,,,,,, x iI V iI c iII **) x iII V iII  ( c,I  c ,II )*10  0 yMi,  *100 (3–3) mM

Where, (****)ciI,,,,,, x iI V iI c iII x iII V iII is the mass of metal in aqua regia in Dreschel

i th bottles when the i sample is taken; and (CC,I  ,II )*10 is the mass of metal in  0 sample tubes in ICP-OES.

M represents Ni, Fe or Co, mM is the initial mass of metal M, g, i is the serial number of samples taken at different times, 0 ~ 6,

I, II represent the tandem absorption bottles,

th -1 ci is the concentration of metal M in i sample for ICP-OES, g·ml , xi is the dilution factor in analysis of sample i, 10 ~ 50,

Vi is the volume of aqua regia in absorption bottles when sample i is taken, ml

th ci,,, I** x i I V i I is the mass of metal M absorbed in the first bottle when the i sample was taken, g

th ci,,, II** x i II V i II is the mass of metal M absorbed in the second bottle when the i sample was taken, g

49 i

 c ,I 10 is the mass of metal M in ICP-OES samples from bottle I  0

th accumulated to the i sample (  =0 means the sample was taken before carbon monoxide was introduced, actually this concentration was detected to be 0 or minus near 0); 10 is the volume of solution after dilution in sample tube, g

c ,II 10 is the mass of metal M in ICP-OES samples from bottle II  0 accumulated to the ith sample ( =0 means the sample was taken before carbon monoxide was introduced, actually this concentration was detected to be 0 or minus near 0); 10 is the volume of solution after dilution in sample tube, g yM,i is the extant of carbonylation of metal M when sample i is collected, %.

Degree of reduction of metal oxides:

Extent of reduction of nickel and iron oxides in the thesis was defined as the weight loss of a sample during the reduction process divided by the oxygen mass in Ni or Fe oxide:

푊푒𝑖푔ℎ푡 푙표푠푠 표푓 표푥𝑖푑푒푠 푠푎푚푝푙푒 퐷푒푔푟푒푒 표푓 푟푒푑푢푐푡푖표푛 = ∗ 100 퐶표푛푡푒푛푡 표푓 표푥푦푔푒푛 𝑖푛 표푥𝑖푑푒푠 푠푎푚푝푙푒 (3–4)

50 Chapter 4 Thermodynamic Analysis of Carbonylation

Carbonylation of nickel and iron can be presented by reactions (4-1) and (4-2):

Ni(s) + 4 CO(g) = Ni(CO)4(g) (4-1)

Fe(s) + 5 CO(g) = Fe(CO)5(g) (4-2)

Carbonylation reactions were analysed in Chapter 2. This chapter presents the thermodynamic analysis of above reactions to assess conditions for carbonylation of selectively reduced laterite ore.

The standard Gibbs free energies of reactions (4-1) and (4-2) were calculated by equations (4-3) and (4-4) using data from the HSC Chemistry database:

o o o o ∆G 4-1 = G Ni(CO)4 - 4G CO -G Ni (4-3)

o o o o ∆G 4-2 = G Fe(CO)5 - 5G CO -G Fe (4-4)

Calculated standard Gibbs free energies and equilibrium constants (log (K)) for reactions (4-1) and (4-2) are presented in Figure 4.1. Both the standard Gibbs free energies increase with temperature. The standard Gibbs free energy for nickel carbonylation is negative below 110 oC; while it is positive for iron carbonylation in the examined temperature range. This means that only formation of nickel carbonyl from CO and Ni is spontaneous under standard conditions; the equilibrium constant is high for the reaction of nickel carbonyl formation and low for the reaction of formation of iron carbonyl.

51 40 8

30 Ni + 4CO(g) = Ni(CO) (g) 4 6 Fe + 5CO(g) = Fe(CO)5(g) 20 4

10 2

(KJ) 0 0  G r  -10 -2 Log(K)

-4 -20 -6 -30 Ni + 4CO(g) = Ni(CO) (g) 4 -8 Fe + 5CO(g) = Fe(CO) (g) -40 5 40 50 60 70 80 90 100 110 Temperature (oC) Figure 4.1 The standard Gibbs free energies ∆rGo and equilibrium constants (presented as log(K)) for reactions of carbonylation of nickel and iron

Equilibrium constant for the reaction of nickel carbonyl formation decreases from 2.57*105 at 40 oC to 3.31 at 110 oC; the equilibrium constant for the reaction of iron carbonyl formation varies from 0.047 at 40 oC to 3.16 * 10-7 at 110 oC.

Conversions of nickel and iron to carbonyls by reactions (4-1) and (4-2) in a stoichiometric system were further calculated to examine the effects of temperature and CO pressure. Conversion was defined as a fraction of Ni or Fe in the form of carbonyl in the gas phase. Molar fractions of Ni carbonyl in the Ni(CO)4(g) –CO gas formed in reaction (4-1) and Fe carbonyl in the Fe(CO)5(g) –CO gas formed in reaction (4-2) were calculated as follows :

XNi(CO)4= 4PNi(CO)4/(4PNi(CO)4+PCO) (4-5)

XFe(CO)5= 5PFe(CO)5/(5PFe(CO)5+PCO) (4-6)

4 P Ni(CO)4 = K4-1*PCO (4-7)

5 P Fe(CO)5 = K4-2*PCO (4-8) where K4-1 and K4-2 are equilibrium constants for reactions (4-1) and (4-2).

52 Calculated conversions of Ni and Fe to carbonyls at different temperatures and CO pressures are presented in Figure 4.2. Both conversions increase with CO pressure and decrease with the increase of temperature. The carbonylation of iron (Figure 4.2 (b)), is more sensitive to reaction conditions. More than 90% of nickel is converted to nickel tetracarbonyl below 150 oC when the CO pressure is over 5 atm; at temperatures below 110oC, conversion of nickel to carbonyl is thermodynamically feasible at CO pressure below 1 atm. However, only about 10% of iron is converted to carbonyl at 100 oC when CO pressure is 10 atm.

150 0.10 150 0.0 ( a. ) ( b ) 0.19 0.10 135 135 0.28 0.20 120 0.37 0.30

120 ) ) ℃ ℃ 0.46 105 0.40 105 0.55 0.50 90 90 0.64 0.60 Temperature ( Temperature ( 75 0.73 0.70 75 0.82 60 0.80

60 0.91 0.90 Conversion (1.0 for 100%, 0.1 for 10%) 0.1 for 100%, Conversion (1.0 for

Conversion (1.0 for 100%, 0.1 for 10%) 0.1 for 100%, Conversion (1.0 for 45 1.0 1.0 0 10 20 30 40 50 60 10 20 30 40 50 60 CO partial pressure (atm) CO partial pressure (atm)

Figure 4.2 Conversions of (a) nickel and (b) iron at different temperatures and CO pressures

4.1 Effect of CO Pressure

Effect of CO pressure on conversion of nickel and iron to carbonyls is shown in Figures 4.3 and 4.4. More than 95% nickel is converted into carbonyls at temperatures below 105oC under CO pressure below 20 atm.

53 1

0.9

0.8 1

o 0.7 45 C 0.98 65 oC o 0.96

Conversion 85 C o 0.6 105 C 0.94 125 oC 0.92 145 oC 165 oC 0.9 0.5 0 20 40 60 80 0 15 30 45 60 75 Pressure (atm)

Figure 4.3 Effect of CO pressure on nickel conversion to Ni(CO)4

Conversion of iron under the same conditions is much lower, it is less than 50% at 100oC under CO pressure of 20 atm

1.0

0.8 50oC 100oC o 0.6 150 C 200oC 250oC 0.4 Conversion

0.2

0.0 0 10 20 30 40 50 60 70 80 Pressure (atm)

Figure 4.4 Effect of pressure on iron conversion to Fe(CO)5

54 4.2 Effect of Temperature

Figures 4.5 and 4.6 show the effect of temperature on conversions of nickel and iron to carbonyls. Conversions of both Ni and Fe decrease with the increasing temperature.

1.0

0.8 1 atm 0.6 5 atm 10 atm 20 atm 30 atm Conversion 0.4 40 atm 50 atm 0.2 60 atm 70 atm 80 atm 0.0 40 80 120 160 200 240 280 320 Temperature (oC)

Figure 4.5 Effect of temperature on nickel conversion to Ni(CO)4

Conversion of iron to Fe(CO)5 drops much faster with the increase of temperature; slightly over 80% of iron carbonyl can be obtained at 100oC at CO pressure 34 atm.

1.0 1.00 0.98 0.96 0.8 0.94 0.92

0.90 20 40 60 80 100 0.6 1 atm 7 atm

16 atm 0.4 25 atm

Conversion 34 atm 40 atm 50 atm 0.2 60 atm 70 atm

0.0 25 50 75 100 125 150 175 200 Temperature (oC)

Figure 4.6 Effect of temperature on iron conversion to Fe(CO)5 55 It can be concluded from the thermodynamic analysis that high extent of nickel carbonylation can be achieved at temperatures below 100 oC and CO pressure above 20 atm; carbonylation of iron at 100 oC requires higher pressure, above 30 atm. Kinetics of carbonylation reactions is affected by the thermodynamic driving force, which is much stronger in carbonylation of nickel in comparison with iron. The project will focus on carbonylation of Ni and Fe and selectively reduced laterite ore at 80-100 oC and CO pressure up to 55 atm.

56 Chapter 5 Experimental Results

As summarized in Chapter 2, the reaction rate and yield of nickel, iron and cobalt carbonylation, are affected by reaction temperature, carbon monoxide pressure, gas flow rate, particle size, and catalysts. This chapter focuses on the study of the effects of these parameters on the carbonylation of nickel, iron, nickel-iron mixture and laterite ores. The reaction of cobalt with carbon monoxide was also investigated to find out the extent of cobalt carbonylation under experimental conditions, which is expected to be low. The carbonylation process was proposed to be used in the application to reduced laterites, therefore, the project studied carbonylation of metals which were also reduced from associated oxides. The structure of this chapter is as follows: (1) The reduction of iron, nickel and cobalt oxides at different temperatures and various reaction times; (2) non- catalytic carbonylation of pure metals (nickel, cobalt and iron) and nickel-iron mixture; (3) catalytic carbonylation of nickel; and (4) non-catalytic and catalytic carbonylation of laterite ores (BCS laterite ore and MIN laterite ore).

5.1 Reduction of Iron, Nickel and Cobalt Oxides

In the reduction process, iron oxide (Fe2O3) was reduced by hydrogen at temperatures from 300 to 500 oC. The flow rate of hydrogen was 1.0 L·min-1. Degree of reduction of iron oxide measured by weight loss as a result of oxygen removal in the reduction experiments is listed in Table 5–1.

A reduction degree of 97.5% was achieved in a 6-hour experiment at 400 oC. Figure 5.1 shows the XRD patterns of samples reduced at 400 oC for 1, 4 and 6 hours. Iron oxide was not identified in the XRD pattern of the sample after 6-h reduction; hence it can be concluded that reduction of Fe2O3 to metallic iron was close to completion. The degree of reduction of iron oxides was enhanced to 98.8% after a 6-hour reaction at 500 oC.

57 Table 5-1 Degree of reduction of Fe2O3 by hydrogen

Degree of reduction of Temperature Time Fe2O3 oC Hour %

1 54.8

300 4 55.4

4 67.7 400 6 97.5

1 69.5 500 6 98.8

Fe 16384 Fe2O3 Fe3O4

8192

4096 6 hour 7500

6000 unt(cts)

Co 4500 4 hour 8000

6000

4000 1 hour 2000 20 30 40 50 60 70 80 Position (2)

o Figure 5.1 XRD patterns of Fe2O3 samples reduced by hydrogen at 400 C for 1, 4 and 6 hours

Based on the results obtained in the reduction of iron oxide, the reduction of nickel oxide was studied from 400 oC. Reduction of NiO was incomplete after 6-hour reaction at this temperature, therefore, the reduction temperature was increased to 500 oC. The reduction degree of NiO by hydrogen at 400 and 500 oC is presented in Table 5-2. The

58 degree of reduction of NiO after 6 hours reaction at 500 oC reached 97.4 %. The XRD analysis shown in Figure 5.2, did not detect nickel oxide residue in the sample.

Table 5-2 Degree of reduction of NiO by hydrogen at 400 and 500 oC

Temperature Time Degree of reduction of NiO

oC Hour %

4 86.1 400 6 91.8

500 6 97.4

2E5 NiO

1E5

5E4 NiO

3E5 t (cts)

2E5 Coun Ni

2E5

8E4 6 hour

20 30 40 50 60 70 80 90 Position (2) Figure 5.2 XRD patterns of NiO samples reduced by hydrogen at 500 oC

Reduction of Fe2O3 – NiO mixture by hydrogen was also tested. Nickel oxide and iron oxide were mixed in agate mortar for 30 minutes before reduction and then reacted with

59 hydrogen at 500oC for 4 and 6 hours. Degree of reduction calculated from the weight loss is given in Table 5-3.

o Table 5-3 Degree of reduction of Fe2O3 – NiO mixture at 500 C

Temperature Time Degree of reduction

oC Hour %

4 97.6 500 6 97.5

XRD analysis presented in Figure 5.3, confirmed that 6-hour reduction of oxide mixture at 500 oC was close to completion; nickel and iron oxides were not detected by the XRD analysis of the Fe2O3 – NiO mixture after reduction. Reduced samples contained nickel, iron and ferronickel. Some sintering of the metallic powder was observed after the reduction process.

NiO Fe2O3 10000 Ni Fe NiFe

8000

Reduced sample

6000 unt (cts)

Co 4000

2000 Raw mixture

0 20 30 40 50 60 70 80 90 Position (2)

Figure 5.3 XRD patterns of the NiO-Fe2O3 mixture before and after reduction by hydrogen at 500 oC for 6 hours

o Cobalt oxide (Co3O4) also was reduced by hydrogen at 500 C for 6 hours; the degree of reduction was 99.6%, indicating that reduction of cobalt oxide was close to completion.

60 The XRD patterns of Co3O4 and reduced sample are shown in Figure 5.4. Only metallic cobalt was detected in the reduced sample.

Co O 16384 3 4 Co

8192

Reduced Sample

4096 Count (cts)

2048 Raw oxide

1024 20 40 60 80 100 Position (2)

o Figure 5.4 XRD patterns of Co3O4 sample reduced by hydrogen at 500 C for 6 hours

The experimental results demonstrated that 6-hour reduction at 500 oC was sufficient for the reduction of oxides and oxides mixture. Therefore, all the metal samples (apart from samples in the study of the effect of particle size on the reduction of nickel oxide and laterite samples) were reduced by hydrogen at 500 oC for 6 hours.

5.2 Non-catalytic Carbonylation of Metal

This section presents results of a comprehensive study of the effect of reaction time, gas flow rate, reaction temperature, carbon monoxide pressure, particle size and other parameters on the non-catalytic carbonylation of nickel, iron, cobalt and Ni-Fe mixture. The reaction parameters are listed in Table 5-4.

61 Table 5-4 Parameters in a study of non-catalytic carbonylation of Ni, Fe, Co and Ni-Fe mixture

Ni-Fe Parameter/Metal Nickel Iron and Cobalt mixture Carbonylation 80, 90, 100 80, 90, 100 100 Temperature, oC 1, 28 55 for iron Gauge Pressure, atm 14, 27, 41, 55 27 44, 55 for cobalt

Gas Flow rate, L∙min-1 0.14, 0.23, 0.36, 0.50 0.50 0.50

Reduction temperature, 500, 600,700, 800,900 500 500 oC

Ni Sample mass, g 0.803, 1.52, 3.36 ~ 1.3 ~ 1.3

Nickel was also carbonylated at 100 oC and atmospheric pressure to evaluate the effect of high temperature and pressure.

5.2.1 Non-Catalytic Carbonylation of Pure Nickel The reported in literature reaction time needed for carbonylation of nickel and ferronickel alloy varied from seconds to several days (Liang et al. 1983, Herrmann 1990, Wang et al. 2009). The reaction time needed for carbonylation of metals under experimental conditions of this project (temperature range 80-100 oC) was assessed in a study of carbonylation of two samples of nickel-iron mixture (50 wt% Ni).

5.2.1.1 Assessment of Reaction Time for Carbonylation

Carbonylation of Ni-Fe mixture (mass 2 g; 50 wt% Ni) was examined at 80 oC (the lowest temperature in the experimental temperature range) at CO gauge pressures of 14 and 54 atm, with CO flow rate at 0.5 L•min-1. The Ni-Fe mixture was obtained by reduction of mixed nickel and iron oxides; the reduced samples were grinded in an agate mortar before being transferred to the reactor.

62 Carbonylation curves for nickel and iron from Ni-Fe mixtures are plotted in Figure 5.5. The rate of carbonylation of both, Ni and Fe increased with increasing CO pressure. Carbonylation of nickel had a much higher rate than that of iron, which is consistent with the literature (Mond et al. 1890, Mond et al. 1891).

The progress of carbonylation reaction became slow after 6 hours. The extent of carbonylation of Ni in experiments at 14 atm CO gauge pressure practically did not change after 6 hours of reaction; increase in the reaction time from 6 to 10 hours increased the extent of formation of nickel tetra-carbonyl at CO gauge pressure 54 atm from 50.1 to 55.9 %, Therefore, 6-hour reaction time was sufficient to study effects of temperature, CO pressure and other parameters on carbonylation of Ni, iron and their mixture.

14 atm 60 54 atm

10 Extent of Ni carbonylation (%)

0 0 2 4 6 8 10 Time (hour)

14 atm 60 54 atm

ent of iron carbonylation (%) 10 Ext 0 0 2 4 6 8 10 Time (hour)

63 Figure 5.5 Carbonylation of the Ni-Fe mixture (50 wt% Ni) at 80 oC and CO gauge pressures 14 and 54 atm

5.2.1.2 Effect of Gas Flow Rate

The effect of the gas flow rate on nickel carbonylation was studied at 100 oC and CO gauge pressure of 27 atm. Gas flow rate was changed from 0.14 to 0.50 L∙min-1; carbonylation curves are shown in Figures 5.6 and 5.7. The gas flow rate had a significant effect on the Ni carbonylation. The rate of carbonylation was constant when the gas flow rate was 0.14 L·min-1. Increase in the gas flow rate enhanced the reaction rate; carbonylation was close to completion in about 6 hours at the gas flow rate 0.5 L· min-1.

100

0.50 Lmin-1 -1 Extentcarbonylation Ni of (%) 20 0.35 Lmin -1

0.23 L? min 0.14 Lmin-1 0 0 123456 Time (Hour) Figure 5.6 Nickel carbonylation at 100 oC and CO gauge pressure 27 atm at different gas flow rates

Figure 5.7 shows the effect of gas flow rate on nickel carbonylation after 6-hours reaction. The carbonylation of pure nickel was greatly influenced by the gas flow rate; but the effect decreased with the increase of the CO flow rate. About 99% of nickel was converted to carbonyl in 6.0 hours when the gas flow rate was 0.5 L•min-1. A study of

64 the effects of other parameters on carbonylation of metals was operated at CO flow rate of 0.5 L•min-1.

100

40 tent of Ni carbonylation (%) 20 Ex

0 0.0 0.1 0.2 0.3 0.4 0.5 Gas flow rate (Lmin-1)

Figure 5.7 Effect of CO flow rate on nickel carbonylation at 100 oC and CO gauge pressure 27 atm

5.2.1.3 Effect of Temperature

Temperature is a significant factor in the nickel carbonylation (Abel 1963). The effect of temperature on carbonylation of nickel was examined in the range of 80 - 100 oC at CO gauge pressure 27 atm, the gas flow rate was 0.5 L•min-1.

Nickel carbonylation curves at 80, 90 and 100 oC are plotted in Figure 5.8. These plots include experimental data obtained by sampling the absorption solution (aqua regia, Chapter 3) at different times. These plots also include “ultimate” points, which were obtained after the completion of the experiment, and purging the reactor with hydrogen to remove Ni(CO)4 left in the reactor. The reaction system had a “dead” volume, which was estimated at about 0.2 L; as a result, a small fraction of Ni(CO)4 was trapped in the system.

65 100 80 oC 80 ultimate 90 oC 80 90 ultimate 100 oC 100 ultimate

40 Extent of Ni carbonylation (%) carbonylation of Ni Extent 20

0 0 1 2 3 4 5 6 7 Time (hour)

Figure 5.8 Effect of temperature on the carbonylation of nickel at CO gauge pressure 27 atm

The extents of nickel carbonylation after 6.5 hours reaction (“ultimate” points) at CO gauge pressure 0, 14 and 27 atm are presented in Figure 5.9. The extent of nickel carbonylation increased with temperature at constant CO pressure. Carbonylation of nickel at 100 oC and CO pressure 27 atm after 6.5 hours reaction was close to completion (extent of carbonylation was 99%).

100 0 atm 14 atm 80 27 atm

20 Extent of Ni carbonylation (%)

0 80 85 90 95 100 Temperature (oC) Figure 5.9 Effect of temperature on nickel carbonylation at CO gauge pressure 0, 14 and 27 atm after 6.5 hours reaction

66 5.2.1.4 Effect of Carbon Monoxide Pressure

The effect of CO pressure on nickel carbonylation was tested in experiments at 80-100 oC with carbon monoxide pressure (gauge pressure) up to 54 atm at gas flow rate 0.5 L·min-1.

Carbonylation curves are plotted in Figure 5.10. Figure 5.11 shows the effect of CO partial pressure on the extents of nickel carbonylation at 80, 90 and 100 oC after 6.5 hours reaction. Effect of CO pressure on nickel carbonylation depended on temperature. At all temperatures, the effect of CO pressure was very strong when the gauge pressure was increased from 0 to 14 atm, and weak when CO gauge pressure changed from 27 to 54 atm (within the experimental error, which was estimated to be 4.5%). The increase in CO pressure from 14 to 27 atm had a strong effect on the rate and extent of carbonylation at 80 and 100 oC and weaker at 90 oC.

The extents of Ni carbonylation at 80 and 90 oC were 47% and 75% at pressures over 27 atm after 6.5 hours reaction. A significant increase in the extent of nickel carbonylation to over 99% was observed when the temperature was increased to 100 oC. The extent of Ni carbonylation at 100 oC at CO pressure 14 atm after 6.5 hours reaction was 82.3%, which is higher than results at 80 and 90 oC at CO pressure above 27 atm.

100 100 80 oC 90 oC 80 80

60 60

40 40

20 20 Extent of(%) NiExtent carbonylation of(%) NiExtent carbonylation 0 0 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Time (hour) Time (hour)

100 100 oC

80 00 atm 60 14 atm

40 27 atm 43 atm 20 54 atm Extent of(%) NiExtent carbonylation 0 0 1 2 3 4 5 6 7 Time (hour)

67 Figure 5.10 Effect of CO pressure (gauge) on the carbonylation of nickel at 80, 90 and 100 oC.

100

20 80 oC Extent of nickel carbonylation (%) 90 oC 10 100 oC 0 0 5 10 15 20 25 30 35 40 45 50 55 CO gauge Pressure (atm) Figure 5.11 Extents of Ni carbonylation at 80, 90 and 100 oC at different CO gauge pressures after 6.5 hours reaction

5.2.1.5 Effect of Particle Size

Particle size of metallic nickel formed in the reduction of NiO depended on the reduction temperature. The particle size of metallic nickel subjected to carbonylation was adjusted in the reduction experiment by the change of the reduction temperature from 500 to 900 oC in 6 hour reduction.

Figure 5.12 shows the SEM images of nickel samples reduced at temperatures from 500 to 900 oC. Sintering at the elevated reduction temperatures coarsened the nickel particles; particle size increased with increasing reduction temperature. The mean particle sizes of samples produced at different temperatures are listed in Table 5-5; the statistical data are presented in Appendix E. The mean particle sizes of nickel were 0.29, 0.51, 1.10, 2.07 and 2.67 μm in samples reduced at 500, 600, 700, 800 and 900 oC respectively.

68 Figure 5.12 Effect of reduction temperature on nickel particle size

Effect of the nickel particle size on carbonylation was studied at 100 oC and CO gauge pressure of 27 atm, the gas flow rate was 0.5 L·min-1.

Table 5-5 Particle size of Ni reduced at different temperatures

Temperature, Number of counted Particle size oC particles Max. Min. Mean μm

500 61 0.46 0.17 0.29

600 86 0.89 0.02 0.51

700 86 2.22 0.61 1.10

800 63 3.58 1.05 2.07

69 900 77 4.35 1.45 2.67

The carbonylation curves obtained in experiments with nickel of different particle size are plotted in Figure 5.13. Nickel particles produced in the reduction at 500 and 600 oC were of sub-micron size, what made the nickel powder highly reactive. It can be suggested that after reduction, these particles were partially re-oxidised (from the surface), what retarded the carbonylation reaction. Figure 5.14 shows the effect of the reduction temperature (particle size) on the extents of nickel carbonylation after 6.5 hours reaction. Experimental data obtained in the first 3-hour carbonylation are rather scattered. However, they are consistent after 4 hours of reaction: the extent of carbonylation decreased with increasing particle size (Figure 5.14).

100

40 0.29 m 0.51 m 20 1.10 m

Extent of nickelExtent carbonylation (%) 2.07 m 2.67 m

0 0 1 2 3 4 5 6 Time (Hours)

Figure 5.13 Effect of particle size on the carbonylation of nickel at 100 oC and CO gauge pressure of 27 atm.

The effect of the particle size on the nickel carbonylation was strong when the mean particle size was above 1.1 μm (the reduction temperature increased over 700 oC).

70 100 o 500 C o o 700 C 600 C 800 oC 80

900 oC 60

20 Extent of nickel carbonylation (%) of nickel carbonylation Extent

0 0.5 1.0 1.5 2.0 2.5 3.0 Particle size (m)

Figure 5.14 Effect of particle size (reduction temperature) on the carbonylation of nickel at 100 oC and CO gauge pressure 27 atm after 4 hours reaction

5.2.1.6 Effect of Nickel Mass The effect of a sample mass on the nickel carbonylation was tested at 100 oC under CO gauge pressure 27 atm. The mass of nickel samples was 0.803, 1.52 and 3.36 g.

Carbonylation curves obtained in experiments with samples of different mass are shown in Figure 5.15. Effect of the sample mass on the nickel carbonylation was quite minor; the nickel carbonylation after 4-hour reaction was close to completion in all experiments.

71 100

40 0.803g 1.52g 20 3.36g Extent of Ni carbonylation (%) carbonylation Ni of Extent

0 0 1 2 3 4 5 6 Time (Hour)

Figure 5.15 Carbonylation of nickel of different mass at 100 oC and CO gauge pressure of 27 atm.

5.2.1.7 Effect of Iron on Nickel Carbonylation

The mixture of Ni and Fe was obtained by mixing nickel and iron reduced from Ni and iron oxides at 500.oC by hydrogen. Carbonylation of nickel in the Ni-Fe mixtures of different compositions was examined at 100 oC and CO gauge pressure of 27 atm. The proportion of iron in the nickel-iron mixture was 11, 20, 50, 80 and 89 wt%, the sample mass was 1.7 g.

The carbonylation curves are shown in Figure 5.16. The extents of the carbonylation reactions did not change after 3 hours for samples containing over 20 wt% of Fe, and within 4 hours for the sample with 11 wt% of Fe.

72 100 11% Fe 90 20% Fe 80 Pure nickel 50% Fe 80% Fe 70 89% Fe 60 50 40 30 20 10 Extent of Ni carbonylation (%) 0 0 1 2 3 4 5 6 7 Time (hour) Figure 5.16 Carbonylation of nickel from the Ni-Fe mixtures of different compositions at 100 oC and CO gauge pressure 27 atm on mixture

The extent of Ni carbonylation from mixtures with different iron contents at 6.0 hours reaction is shown in Figure 5.17; the extent carbonylation of pure nickel isalso drawn in the figure. The extent of nickel carbonylation decreased with the increase of iron content. The extent of nickel carbonylation dropped sharply from over 98% to 55.4% when 11 wt% of iron was added to nickel; it declined further to only 6.2% when iron content in the mixture increased to 89 wt%.

100 Pure Ni Ni

Extents of Extents Ni (%) carbonylation 20

0 0 20 40 60 80 100 Content of Fe in Ni-Fe mixture (%)

Figure 5.17 Carbonylation of nickel from the Ni-Fe mixtures of different compositions at 100 oC and CO gauge pressure 27 atm after 6.0 hour reaction.

73 5.2.2 Non-Catalytic Carbonylation of Pure Iron

The carbonylation of pure iron was studied at 80, 90 and 100 oC at CO gauge pressure of 0, 27 and 54 atm. The gas flow rate was 0.5 L·min-1, sample mass 1.3 g.

Carbonylation curves obtained in experiments at 80-100 oC and CO gauge pressure 0, 27 and 54 atm are shown in Figure 5.18.

10 10 0 atm 27 atm

8 8

6 6

4 4

2 2

Extent of Fe carbonylation (%) 0 Extent of Fe carbonylation (%) 0 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Time (hour) Time (hour) 10 54 atm o 8 80 C 90 oC o 6 100 C

Extent of Fe carbonylation (%) 0 0 1 2 3 4 5 6 7 Time (hour) Figure 5.18 Carbonylation of iron at 80-100 oC and CO gauge pressure 0, 27 and 54 atm

Extents of Fe carbonylation after 6.5-hour reaction are given in Figure 5.19. The rate and extents of iron carbonylation were much lower in comparison with carbonylation of pure nickel. The extent of iron carbonylation under atmospheric pressure increased only from 1.5% at 80 oC to 3.3% at 100 oC. The extent of carbonylation of pure iron at 80 oC increased with increasing CO pressure; however, the effect of temperature on the carbonylation of iron at CO gauge pressure 27 and 53 atm was insignificant. The extent of iron carbonylation at 100 oC under pressures 27 and 53 atm was 3.2-3.7%.

74 10 00 atm 27 atm 54 atm 8

2 Extent of iron carbonylation (%) of iron Extent

0 80 85 90 95 100 Temperature (oC)

Figure 5.19 Carbonylation of iron at 80-100 oC and CO gauge pressure 0, 27 and 54 atm after 6.5-hour reaction

The effect of CO pressure on the iron carbonylation at 80-100 oC is shown in Figures 5.20 and 5.21. The rate and extent of the iron carbonylation increased with increasing CO pressure at 80 and 90 oC; however, increase in CO pressure had no effect on the carbonylation of iron at 100 oC under experimental conditions.

10 10 o 80 C 90 oC

8 8

6 6

4 4

2 2 xtent of Fe carbonylation (%) Extent of Fe carbonylation (%) E 0 0 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Time (hour) Time (hour) 10 100 oC

8 00 atm 27 atm 6 54 atm

Extent of Fe carbonylation (%) 0 0 1 2 3 4 5 6 Time (hour)

Figure 5.20 Effect of CO gauge pressure on carbonylation of iron at 80-100 oC.

75 20 o 18 80 C 90oC 16 100oC 14 12 10 8 6 4 Extent of Fe carbonylation (%) Extent 2 0 0 5 10 15 20 25 30 35 40 45 50 55 CO pressure (atm)

Figure 5.21 Effect of CO pressure (gauge) on the carbonylation of iron at 80-100 oC after 6.5 hours reaction

5.2.3 Non-catalytic Carbonylation of Cobalt

Formation of cobalt carbonyl requires high temperature and high CO pressure which effect has been reported to be more critical. The experimental CO pressure in the project was up to 55 atm (gauge pressure 54 atm) and the temperature up to 100oC. Although formation of cobalt carbonyl under these experimental conditions is expected to be low, it is important to find the extent of Co carbonylation in carbonylation of reduced nickel laterites, which contain cobalt. The carbonylation of cobalt obtained from reduction of cobalt oxide was studied at temperatures 80, 90 and 100 oC and CO pressure (gauge) 43 and 54 atm; gas flow rate was 0.5 L·min-1.

Effects of temperature and CO pressure on the cobalt carbonylation are shown in Figures 5.22 and 5.23 respectively. The carbonylation of cobalt under given experimental conditions was very slow; less than 0.5% of cobalt was adsorbed in aqua regia. Change of reaction temperature did not affect the carbonylation of cobalt (Figure 5.22). The increase of carbon monoxide pressure from 43 to 56 atm slightly reduced the yield of cobalt carbonylation at 80 and 90 oC (Figure 5.23).

76 5 41 atm 80 oC 90 oC 4 o 100 C