close

Вход

Забыли?

вход по аккаунту

?

A study of the microwave carbothermic reduction of chromite ore

код для вставкиСкачать
A Study of the Microwave Carbothermic
Reduction of Chromite Ore
by
Kristian David Mackowiak
A thesis submitted to
The Robert M. Buchan Department of Mining
in conformity with the requirements for
the degree of Master of Applied Science
Queen’s University
Kingston, Ontario, Canada
November 2016
c Kristian David Mackowiak, 2016
Copyright ProQuest Number: 10588824
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
ProQuest 10588824
Published by ProQuest LLC (2017 ). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
Abstract
Microwave reduction testing using activated charcoal as a reducing agent was performed on a sample of Black Thor chromite ore from the Ring of Fire deposit in
Northern Ontario. First, a thermodynamic model was constructed for the system.
Activity coefficients for several species were found in the literature. The model
predicted chromium grades of 61.60% and recoveries of 93.43% for a 15% carbon
addition. Next, reduction testing on the chromite ore was performed. Tests were
performed at increasing power levels and reduction times. Testing atmospheres used
were air, argon, and vacuum. The reduced product had maximum grades of 72.89%
and recoveries of 80.37%. These maximum values were obtained in the same test
where an argon atmosphere was used, with a carbon addition of 15%, optimal power
level of 1200 W (actual 1171 W), and a time of 400 seconds. During this test, 17.53%
of the initial mass was lost as gas, a carbon grade of 1.95% was found for the sintered
core product. Additional work is recommended to try and purify the sintered core
product as well as reduce more of the initial sample. Changing reagent schemes or a
two step reduction / separation process could be implemented.
i
Acknowledgments
I would like to thank my supervisor Dr. Chris Pickles for his guidance and support
throughout this thesis. His ideas and comments have made a positive impact on all
aspects of my work and this thesis would not be possible without him.
Thanks to my parents, Krzysztof and Jadwiga Mackowiak, for their continued encouragement of my studies.
I would also like to thank Larissa Smith and Maritza Bailey for their assistance in
the labs. Their tireless work ensured that all of my experiments and analysis could
proceed on time. Thanks to Agatha Dobosz who allowed me to use the analytical
equipment of the Geology Department and to Charlie Cooney for allowing me to use
his lab and provide vital insight into the properties of the materials I was analyzing.
Finally, I would like to thank Ron Hutcheon for conducting permittivity analysis and
for the discussions which allowed me to gain a better understanding of my work.
Next, thank you to all graduate students for their friendship over the duration of
this thesis. Special thanks to John Forster, Richard Elliott, Alexander Cushing, and
Denver Cowan for their assistance in many aspects of my work.
Finally, I would like to thank all of my friends for their continued support. Special
thanks to Mark Nardi, Ross Manson, Cody Stewart, Chris Carrick, and Ryan Chew.
ii
I could not have completed this thesis without you. Lastly, thank you to the staff of
Sir Johns Public House for their service and hospitality each Thursday over the past
2 years.
Special thanks to Noront Resources for providing the ore used in this thesis.
iii
Contents
Abstract
i
Acknowledgments
ii
Contents
iv
List of Tables
viii
List of Figures
ix
Chapter 1:
Introduction
1.1 General Overview . . . . . . . . . . . . . . . . .
1.2 Uses of Chromium . . . . . . . . . . . . . . . .
1.3 Economics of Chromium . . . . . . . . . . . . .
1.4 Chromium Deposits . . . . . . . . . . . . . . . .
1.5 Chromium Processing . . . . . . . . . . . . . .
1.6 Research Scope and Objectives of Current Work
1.7 Thesis Organization . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 2:
Literature Review
2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.1 Iron-Chromium Phase Diagram . . . . . . . . .
2.2 Reduction Behaviour of Chromite Ores . . . . . . . . .
2.2.1 Thermodynamic Studies of Chromite Reduction
2.2.2 Kinetic Studies of Chromite Reduction . . . . .
2.3 Conventional Heating . . . . . . . . . . . . . . . . . . .
2.3.1 Other Ferrochrome Production Methods . . . .
2.4 Microwave Heating . . . . . . . . . . . . . . . . . . . .
2.5 Vacuum Processing . . . . . . . . . . . . . . . . . . . .
2.5.1 Energy Savings . . . . . . . . . . . . . . . . . .
iv
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1
1
3
5
7
8
9
.
.
.
.
.
.
.
.
.
.
11
11
12
12
18
22
25
27
31
35
38
2.6
2.7
Hydrometallurgy . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 3:
Microwave Theory
3.1 Microwave Background . . . . . . . . .
3.2 Microwave Properties . . . . . . . . . .
3.3 Microwave Heating Effects . . . . . . .
3.4 Microwave Advantages and Limitations
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 4:
Experimental Methods and Material Used
4.1 Ore Preparation . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2 Material Classification . . . . . . . . . . . . . . . . . . . . . .
4.2.1 SEM Analysis . . . . . . . . . . . . . . . . . . . . . . .
4.2.1.1 As-Received Ore . . . . . . . . . . . . . . . .
4.2.1.2 Size Analysis . . . . . . . . . . . . . . . . . .
4.2.2 XRD . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.3 Assay . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3 Experimental Equipment . . . . . . . . . . . . . . . . . . . . .
4.3.1 Sample Preparation . . . . . . . . . . . . . . . . . . . .
4.3.2 Microwave Setup . . . . . . . . . . . . . . . . . . . . .
4.3.2.1 Mode Stirrer . . . . . . . . . . . . . . . . . .
4.3.2.2 Vacuum . . . . . . . . . . . . . . . . . . . . .
4.3.3 Microwave Heating Tests . . . . . . . . . . . . . . . . .
4.4 Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.5 Analysis Methods and Equipment . . . . . . . . . . . . . . . .
4.5.1 Thermogravimetric and Differential Thermal Analysis
(TGA/DTA) . . . . . . . . . . . . . . . . . . . . . . .
4.5.1.1 Proximate Analysis of Charcoal . . . . . . . .
4.5.2 Carbon-Sulphur Analysis . . . . . . . . . . . . . . . . .
4.5.3 X-Ray Diffraction (XRD) . . . . . . . . . . . . . . . .
4.5.4 X-Ray Fluorescence (XRF) . . . . . . . . . . . . . . .
4.5.5 Environmental Scanning Electron Microscope (ESEM)
4.5.5.1 Mineral Liberation Analysis (MLA) . . . . . .
4.5.6 Electron Microprobe (EMP) . . . . . . . . . . . . . . .
4.5.7 Cavity Perturbation Technique . . . . . . . . . . . . .
4.5.8 Particle Size Analysis . . . . . . . . . . . . . . . . . . .
4.5.9 Metallography . . . . . . . . . . . . . . . . . . . . . . .
4.6 Variables Investigated . . . . . . . . . . . . . . . . . . . . . .
v
.
.
.
.
.
.
.
.
.
.
.
.
39
46
.
.
.
.
48
48
49
51
55
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
57
57
58
58
58
61
65
65
67
67
70
74
75
75
76
76
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
76
77
79
80
81
82
83
83
84
85
85
86
4.7
4.8
Recovery Calculations . . . . . . . . . . . . . . . . . . . . . . . . . .
Error Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 5:
Thermodynamics
5.1 Thermodynamic Considerations
5.2 Model Limitations . . . . . . .
5.3 Species and Phases Considered
5.4 Activity Coefficients . . . . . .
5.5 Atmospheric Pressure Model . .
5.5.1 Effect of Carbon . . . .
5.6 Reduced Pressure Model . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 6:
Results and Discussion
6.1 Permittivity . . . . . . . . . . . . . . . . . .
6.1.1 Effect of Frequency . . . . . . . . . .
6.1.2 Effect of Carbon . . . . . . . . . . .
6.2 TGA/DTA Analysis . . . . . . . . . . . . .
6.2.1 TGA/DTA with Permittivty Overlay
6.3 Initial Heating Tests . . . . . . . . . . . . .
6.4 Black Thor Chromite Results . . . . . . . .
6.4.1 Effect of Carbon . . . . . . . . . . .
6.4.2 Effect of Energy . . . . . . . . . . . .
6.4.3 Effect of Atmospheric Composition .
6.5 Discussion . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 7:
Conclusions and Recommended Future
7.1 Thermodynamic Model . . . . . . . . . . . . . . . .
7.2 Black Thor Chromite . . . . . . . . . . . . . . . . .
7.2.1 Effect of Carbon . . . . . . . . . . . . . . .
7.2.2 Effect of Energy . . . . . . . . . . . . . . . .
7.2.3 Effect of Atmosphere . . . . . . . . . . . . .
7.3 Future Work . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Work
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
88
90
.
.
.
.
.
.
.
91
91
92
92
94
96
106
109
.
.
.
.
.
.
.
.
.
.
.
115
115
115
117
122
126
129
133
134
140
163
167
.
.
.
.
.
.
170
170
171
172
172
173
175
Bibliography
177
Appendix A: Experiments Conducted
188
Appendix B: Scanning Electron Microscopy
191
vi
Appendix C: Additional Results Data
199
C.1 Additional Permittivity Data . . . . . . . . . . . . . . . . . . . . . . 199
C.2 Additional Heating Test Data . . . . . . . . . . . . . . . . . . . . . . 209
vii
List of Tables
1.1
Chromite ore grades [Hock et al., 1986, Geovic, 2010]. . . . . . . . . .
2.1
Energy requirements and savings comparision between three conven-
2
tional FeCr processes and vacuum-reduction [Pahlman et al., 1981]. .
39
4.1
Whole rock analysis results from SGS Lakefield. . . . . . . . . . . . .
66
4.2
Elemental assay results from SGS Lakefield. . . . . . . . . . . . . . .
67
4.3
Composition of activated charcoal determined using a modified proximate analysis method on a Jupiter STA 449 TGA. . . . . . . . . . .
79
4.4
All experimental variables. . . . . . . . . . . . . . . . . . . . . . . . .
86
4.5
Time, input power, and total energy for the reduction tests. . . . . .
87
5.1
Species and phases considered in the thermodynamic model, as entered
into the HSC Equilibrium Module. . . . . . . . . . . . . . . . . . . .
6.1
93
Summary of results for reduction testing. . . . . . . . . . . . . . . . . 167
A.1 All main series experiments performed. . . . . . . . . . . . . . . . . . 188
viii
List of Figures
1.1
Historical end use distribution of chromium from 1975 to 2003 [USGS,
2014]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2
World production and apparent consumption of chromium from 1900
to 2014 [USGS, 2014]. . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3
21
The Perrin Process, from Electric Smelting Processes, 1973 [Robiette,
1973]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.5
19
Partial pressure-temperature plot for the reduction of FeCr2 O4 , including various reactions [Hino et al., 1998]. . . . . . . . . . . . . . .
2.4
12
Isothermal section of the Fe-Cr-C phase diagram at 1473 K [Hino
et al., 1998]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3
5
The iron-chromium binary phase diagram at atmospheric pressure
[Durand-Charre, 2004]. . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2
4
Price per tonne of chromium from 1900 to 2014 adjusted to 1998 USD
[USGS, 2014]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1
3
28
The flowsheet for the Outokumpu Process [Relander and Honkaniemi,
1985]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ix
30
2.6
VOD effect on carbon and chromium over time, adapted from [Nair
et al., 2013]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.7
31
The effect of holding time on carbon content at various temperatures
[Hao et al., 2014]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
34
2.8
Effect of pressure on total reduction at 1300 ◦ C [Pahlman et al., 1981]. 36
2.9
Effect of reducing material on total reduction at 1270◦ C and 1 torr
[Pahlman et al., 1981]. . . . . . . . . . . . . . . . . . . . . . . . . . .
37
2.10 Effect of temperature on total reduction at 1 torr and using graphite
as a carbon source [Pahlman et al., 1981]. . . . . . . . . . . . . . . .
38
2.11 Effect of leaching time on chromium extraction for various temperatures [Amer, 1992]. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
41
2.12 Effect of leaching time on extraction of chromium for various particle
sizes [Zhao et al., 2014]. . . . . . . . . . . . . . . . . . . . . . . . . .
44
3.1
A diagram of an electromagnetic wave [ploufandsplash, 2007]. . . . .
49
3.2
Expanding core model of microwave heating. . . . . . . . . . . . . . .
52
3.3
The effect of dielectric loss on power absorbed per unit volume [Thostenson and Chou, 1999]. . . . . . . . . . . . . . . . . . . . . . . . . . . .
54
4.1
Size distribution of pulverized chromite ore. . . . . . . . . . . . . . .
58
4.2
BSE image and elemental mapping of as-received chromite ore. . . . .
60
4.3
BSE image and elemental mapping of the +105 µm size fraction. . . .
62
4.4
BSE image and elemental mapping of the +44 -53 µm size fraction. .
63
4.5
BSE image and elemental mapping of the -20 µm size fraction. . . . .
64
x
4.6
XRD spectrum of pulverized chromite ore. . . . . . . . . . . . . . . .
65
4.7
Sample setup for air atmosphere reduction tests. . . . . . . . . . . . .
68
4.8
Sample setup for argon atmosphere reduction tests. . . . . . . . . . .
69
4.9
Sample setup for vacuum reduction tests. . . . . . . . . . . . . . . . .
70
4.10 1.2 kW variable magnetron used for the current research [GAE, 2015]. 71
4.11 3-port circulator used to control microwave propagation [GAE, 2015].
71
4.12 Dummy load used to dissipate reflected microwaves [GAE, 2015].
72
. .
4.13 Dual power monitor that allowed for the measurement of absorbed
power [GAE, 2015]. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
72
4.14 3-stub impedance tuner located after the power meters and before the
applicator [GAE, 2015]. . . . . . . . . . . . . . . . . . . . . . . . . . .
73
4.15 Microwave applicator where samples were processed [GAE, 2015]. . .
73
4.16 Overhead view of the connected microwave system. . . . . . . . . . .
74
4.17 A visual representation of Bragg’s Law [Hadjiantonis, 2013]. . . . . .
81
4.18 X-ray fluorescence schematic [Calvero., 2016]. . . . . . . . . . . . . .
82
4.19 Schematic diagram of the apparatus used in the cavity perturbation
technique [Hutcheon et al., 1992]. . . . . . . . . . . . . . . . . . . . .
5.1
Chromium species distribution between 200 and 1400 ◦ C for a 15%
carbon addition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2
97
Iron species distribution between 200 and 1400 ◦ C for a 15% carbon
addition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.3
85
98
Chromium grade of the ferrochrome alloy as a function of temperature
for various initial carbon additions. . . . . . . . . . . . . . . . . . . .
xi
99
5.4
Chromium recovery as a function of temperature for various initial
carbon additions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.5
Iron grade of the ferrochrome alloy as a function of temperature for
various initial carbon additions. . . . . . . . . . . . . . . . . . . . . . 101
5.6
Iron recovery as a function of temperature for various initial carbon
additions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.7
Carbon grade of the ferrochrome alloy as a function of temperature
for various initial carbon additions. . . . . . . . . . . . . . . . . . . . 103
5.8
Unreacted carbon percentage as a function of temperature for various
initial carbon additons. . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.9
Gaseous species distribution between 200 and 1400 ◦ C for a 15% carbon addition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.10 Gaseous species distribution, without CO, between 200 and 1400 ◦ C
for a 15% carbon addition. . . . . . . . . . . . . . . . . . . . . . . . . 106
5.11 Chromium, iron, and carbon grades as a function of input carbon at
1400 ◦ C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.12 Chromium and iron recoveries as a function of input carbon at 1400 ◦ C.108
5.13 Unreacted carbon as a function of input carbon at 1400 ◦ C. . . . . . . 109
5.14 Chromium grade as a function of temperature and pressure for 15%
input carbon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.15 Chromium recovery as a function of temperature and pressure for 15%
input carbon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
xii
5.16 Iron grade as a function of temperature and pressure for 15% input
carbon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
5.17 Iron recovery as a function of temperature and pressure for 15% input
carbon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5.18 Carbon grade as a function of temperature and pressure for 15% input
carbon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6.1
The effect of temperature on the real permittivity of chromite ore for
various frequencies. The sample density was 3.01 g/cm3 . . . . . . . . 116
6.2
The effect of temperature on the imaginary permittivity of chromite
ore for various frequencies. The sample density was 3.01 g/cm3 . . . . 117
6.3
Comparison of real permittivities from 30 to 1430 ◦ C at 2466 MHz for
chromite ore and chromite with 8 and 15% carbon additions. Sample
densities were 3.01 g/cm3 for chromite, 2.18 g/cm3 for chromite with
8% carbon, and 2.36 g/cm3 for chromite with 15% carbon. . . . . . . 118
6.4
Comparison of imaginary permittivities from 30 to 1430 ◦ C at 2466
MHz for chromite ore and chromite with 8 and 15% carbon additions. Sample densities were 3.01 g/cm3 for chromite, 2.18 g/cm3 for
chromite with 8% carbon, and 2.36 g/cm3 for chromite with 15% carbon.119
6.5
Comparison of loss tangents from 30 to 1430 ◦ C at 2466 MHz for
chromite ore and chromite with 8 and 15% carbon additions. Sample
densities were 3.01 g/cm3 for chromite, 2.18 g/cm3 for chromite with
8% carbon, and 2.36 g/cm3 for chromite with 15% carbon. . . . . . . 120
xiii
6.6
Comparison of penetration depths from 30 to 1430 ◦ C at 2466 MHz for
chromite ore and chromite with 8 and 15% carbon additions. Sample
densities were 3.01 g/cm3 for chromite, 2.18 g/cm3 for chromite with
8% carbon, and 2.36 g/cm3 for chromite with 15% carbon. . . . . . . 122
6.7
Thermogravimetric comparison of chromite ore, and chromite with an
8 and 15% carbon addition. . . . . . . . . . . . . . . . . . . . . . . . 124
6.8
Differential thermal analysis comparison of chromite ore, and chromite
with an 8 and 15% carbon addition. . . . . . . . . . . . . . . . . . . . 125
6.9
TGA, DTA, and permittivity data for chromite ore. Test conditions
were identical with an argon atmosphere and a maximum temperature
of 1400 ◦ C. Sample density for the permittivity test was 3.01 g/cm3 . . 127
6.10 TGA, DTA, and permittivity data for chromite ore with an 8% carbon
addition. The atmosphere used in both tests was argon with a maximum temperature of 1400 ◦ C. Sample density for the permittivity test
was 2.18 g/cm3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.11 TGA, DTA, and permittivity data for chromite ore with a 15% carbon
addition. The atmosphere used in both tests was argon with a maximum temperature of 1400 ◦ C. Sample density for the permittivity test
was 2.36 g/cm3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
6.12 Absorbed power over time for both mode stirrer and non mode stirrer
tests over 900 seconds. End temperature after each run was approximately 1360 ◦ C. Input power was 800 W in an argon atmosphere. . . 131
xiv
6.13 Absorbed power over time for non mode stirrer tests for 300, 600, and
900 seconds. Input power was 800 W in an argon atmosphere. . . . . 132
6.14 Double exponentially smoothed absorbed power over time for mode
stirrer tests for 300, 600, and 900 seconds. Input power was 800 W in
an argon atmosphere. . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
6.15 Absorbed energy as a function of time for chromite ore and chromite
ore with various initial carbon additions. Tests were run in an argon
atmosphere with an input energy of 800 W. . . . . . . . . . . . . . . 135
6.16 The effect of carbon on the mass loss and the total absorbed power.
Tests were 600 seconds long in an argon atmosphere. The input power
was 800 W. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.17 Chromium grade as a function of initial carbon percentage. Tests were
run in argon with an input power of 800 W for 600 seconds. . . . . . 137
6.18 Overall and core recoveries of chromium and iron as a function of
initial carbon amount. Tests were run in an argon atmosphere with
an input power of 800 W for 600 seconds. . . . . . . . . . . . . . . . . 138
6.19 SEM image of the polished reduced core from the 10% input carbon
test. This sample was reduced for 600 seconds in an argon atmosphere
with 800 W input power. . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.20 SEM image of the polished reduced core from the 15% input carbon
test. This sample was reduced for 600 seconds in an argon atmosphere
with 800 W input power. . . . . . . . . . . . . . . . . . . . . . . . . . 140
xv
6.21 Rate of mass loss as a function of input power for various atmospheric
and vacuum compositions. Tests varied in duration from 150 to 600
seconds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6.22 Reduced core mass as a function of absorbed energy for various atmospheric and vacuum compositions. Tests varied between 150 and 600
seconds with input energies between 400 and 1200 W. . . . . . . . . . 144
6.23 Carbon in the reduced core as a function of absorbed energy for various
atmospheric and vacuum compositions. Tests varied between 150 and
600 seconds with input energies of 800 and 1200 W. . . . . . . . . . . 145
6.24 Chromium grade as a function of absorbed energy for various atmospheric and vacuum compositions. Reaction times varied from 150 to
600 seconds with an input power range of 400 to 1200 W. . . . . . . . 147
6.25 Overall chromium recovery as a function of absorbed energy for various
atmospheric and vacuum compositions. Reaction times varied from
150 to 600 seconds with an input power range of 400 to 1200 W. . . . 148
6.26 Core chromium recovery as a function of absorbed energy for various
atmospheric and vacuum compositions. Reaction times varied from
150 to 600 seconds with an input power range of 400 to 1200 W. . . . 150
6.27 Overall iron recovery as a function of absorbed energy for various
atmospheric and vacuum compositions. Reaction times varied from
150 to 600 seconds with an input power range of 400 to 1200 W. . . . 151
xvi
6.28 Core iron recovery as a function of absorbed energy for various atmospheric and vacuum compositions. Reaction times varied from 150 to
600 seconds with an input power range of 400 to 1200 W. . . . . . . . 152
6.29 The input versus actual absorbed energy for each test run in all three
atmospheric and vacuum conditions. The bold line indicates 100%
energy absorption. Included are regressions for all three atmospheric
compositions. Reaction times varied from 150 to 600 seconds with
input powers ranging from 400 to 1200 W. . . . . . . . . . . . . . . . 153
6.30 SEM image of the sintered core from the 300 second, 800 W air test.
155
6.31 SEM image of the sintered core from the 200 second, 1200 W air test. 156
6.32 SEM image of the sintered core from the 600 second, 800 W air test.
157
6.33 SEM image of the sintered core from the 400 second, 1200 W air test. 158
6.34 SEM image from a part of the boundary between powder and reacted
core from the 600 second, 800 W test. . . . . . . . . . . . . . . . . . . 159
6.35 SEM image of the sintered core from the 300 second, 800 W argon test.160
6.36 SEM image of the sintered core from the 200 second, 1200 W argon
test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
6.37 SEM image of the sintered core from the 600 second, 800 W argon test.162
6.38 SEM image of the sintered core from the 400 second, 1200 W argon
test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.39 Highest obtained chromium grades for various atmospheric and vacuum conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
xvii
6.40 Highest obtained chromium and iron recoveries for various atmospheric
and vacuum conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . 165
6.41 Chromium grade as a function of particle area for various atmospheric
and vacuum conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . 166
B.1 BSE image and elemental mapping of as-received chromite ore. . . . . 192
B.2 BSE image and elemental mapping of as-received chromite ore. . . . . 193
B.3 BSE image and elemental mapping of as-received chromite ore. . . . . 194
B.4 BSE image and elemental mapping of the +74 -105 µm size fraction.
195
B.5 BSE image and elemental mapping of the +53 -74 µm size fraction. . 196
B.6 BSE image and elemental mapping of the +37 -44 µm size fraction. . 197
B.7 BSE image and elemental mapping of the +25 -37 µm size fraction. . 198
C.1 Effect of frequency on loss tangent for chromite ore. Sample density
was 3.01 g/cm3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
C.2 Effect of frequency on penetration depth for chromite ore. Sample
density was 3.01 g/cm3 . . . . . . . . . . . . . . . . . . . . . . . . . . 201
C.3 Effect of frequency on real permittivity for chromite ore with 8% carbon. Sample was 2.18 g/cm3 . . . . . . . . . . . . . . . . . . . . . . . 202
C.4 Effect of frequency on imaginary permittivity for chromite ore with
8% carbon. Sample was 2.18 g/cm3 . . . . . . . . . . . . . . . . . . . . 203
C.5 Effect of frequency on loss tangent for chromite ore with 8% carbon.
Sample density was 2.18 g/cm3 . . . . . . . . . . . . . . . . . . . . . . 204
xviii
C.6 Effect of frequency on penetration depth for chromite ore with 8%
carbon. Sample density was 2.18 g/cm3 . . . . . . . . . . . . . . . . . 205
C.7 Effect of frequency on real permittivity for chromite ore with 15%
carbon. Sample density was 2.36 g/cm3 . . . . . . . . . . . . . . . . . 206
C.8 Effect of frequency on imaginary permittivity for chromite ore with
15% carbon. Sample density was 2.36 g/cm3 . . . . . . . . . . . . . . . 207
C.9 Effect of frequency on loss tangent for chromite ore with 15% carbon.
Sample density was 2.36 g/cm3 . . . . . . . . . . . . . . . . . . . . . . 208
C.10 Effect of frequency on penetration depth for chromite ore with 15%
carbon. Sample density was 2.36 g/cm3 . . . . . . . . . . . . . . . . . 209
C.11 Absorbed power over time plots for the 300 second mode stirrer and
non mode stirrer heating tests. Samples were heated in an argon atmosphere with 800 W input power. Final temperatures were 129.5◦ C
for the non mode stirrer test and 130.2◦ C for the mode stirrer test. . 210
C.12 Absorbed power over time plots for the 600 second mode stirrer and
non mode stirrer heating tests. Samples were heated in an argon
atmosphere with 800 W input power. Final temperatures were 211◦ C
for the non mode stirrer test and 574◦ C for the mode stirrer test. . . 211
xix
1
Chapter 1
Introduction
1.1
General Overview
In this thesis, the viability of microwave processing of a chromite ore powder was
studied. Incentive for this research was to study the effects and determine the viability of microwave processing of chromite ores, specifically those from the Black
Thor deposit in the Northern Ontario Ring of Fire. Microwave processing, in relation
to other processing methods, occurs faster and has unique reduction characteristics
which may present it as an alternative method of pyrometallurgical processing. The
following sections will discuss the uses of chromium, economics, the deposits and
processing methods used, and the scope and objectives of this thesis.
1.2
Uses of Chromium
In 2013, the world produced approximately 28,800,000 metric tons of chromite ore
with 9,570,000 metric tons of ferrochromium being produced [Papp, 2013]. According to a report by the International Stainless Steel Forum [ISSF, 2015], in 2014
1.2. USES OF CHROMIUM
2
chromium ores could be divided into four grades. 96% of chromium ore was metallurgical grade, 2% for both chemical and foundry grade, and 0.2% for refractory
grade. Differences between grades of chromite relate to their chemical composition.
The grade differences can be found in Table 1.1.
Table 1.1: Chromite ore grades [Hock et al., 1986, Geovic, 2010].
Grade
Chromium Oxide
Alumina
Other
Metallurgical
45-53%
12-18%
–
Chemical
>45%
low
Cr/Fe = 1.5
Foundry
>44%
low
<27% Fe, Low Si, Mg
Refractory
30-44%
20-30%
–
Metallurgical grade chromium is typically converted into three primary types of
ferrochrome for steel production. These types and amounts are 94% high-carbon ferrochrome, 2% medium-carbon, and 4% low carbon. Metallurgical ferrochrome has
the following use breakdown: 77% stainless steel, 19% engineering and alloy steel,
and 4% other steels [ISSF, 2015]. Applications of stainless steel are vast due to its
corrosion resistance; such applications include cutlery, machinery parts, automotive,
aerospace, oil and gas, renewable energy, and construction [JFE, 2010]. Chemical
grade chromite is used in chromium chemicals, super alloys, chrome plating, pigments, and leather tanning. Foundry grade chromite is used in casting moulds, and
finally refractory grade ores are used for refractory bricks, cement kiln, and other
high temperature applications [ISSF, 2015].
Figure 1.1 from the United States Geological Survey (USGS) data shows historical
1.3. ECONOMICS OF CHROMIUM
3
values for end uses of chromium from 1975 to 2003. ‘Metallurgical’ refers to stainless
steels and other alloys while ‘other’ refers to all non-alloying applications such as
those listed above.
Figure 1.1: Historical end use distribution of chromium from 1975 to 2003 [USGS,
2014].
1.3
Economics of Chromium
Historic chromium prices and production/consumption data from 1900 to 2104 was
obtained from the USGS. Figure 1.2 shows the world production and consumption
of chromium, in kilotonnes. It can be seen that the production follows an almost exponential growth trend. Large dips from the increasing production can be attributed
1.3. ECONOMICS OF CHROMIUM
4
to economic depressions such as those that occurred in 1995 and 2008. Consumption
of chromium increases until 1980, at which point it dips before a slight rise until
the 2000s. Increases in the production can be seen in the 1940s due to military
requirements of the Second World War, and in the 1960s during the Cold War.
Figure 1.2: World production and apparent consumption of chromium from 1900 to
2014 [USGS, 2014].
Figure 1.3 shows the price of one tonne of chromium from 1900 to 2014 adjusted
to 1998 US Dollars. Prices can be seen to increase towards the end of the First World
War, and quickly decreasing after that due to a economic recession. Prices stayed low
until 1930 where they increased to 500$ (1998)/tonne due to the Great Depression.
Prices stayed relatively the same, increasing during the Second World War. The
1.4. CHROMIUM DEPOSITS
5
prices continued to stay relatively constant until the 1973 Depression and after this
gradually declined, but they are marked with peaks due to other economic events.
Steel use saw a boom in 2000, leading to increased prices until the 2008 economic
crisis. The year 2008 saw the highest steel price on record, which is reflected in the
chromium price [OECD, 2009].
Figure 1.3: Price per tonne of chromium from 1900 to 2014 adjusted to 1998 USD
[USGS, 2014].
1.4
Chromium Deposits
Chromium can be found in four types of deposits: stratiform, podiform, placer, and
laterite [Schulte et al., 2012, Gujar et al., 2010, Berger and Frei, 2014]. Chromium
1.4. CHROMIUM DEPOSITS
6
primarily comes from stratiform and podiform deposit types.
The Ring of Fire deposits, from where the ore for the current research originates,
are stratiform deposits [Ontario, 2011]. These deposits are characterized by massive
chromitite, of grades greater than 90% chromite, in seams found in mafic-ultramafic
intrusions. The chromitite seams are typically found in the lower portion of the
intrusion and as such these deposits are typically mined via underground methods.
The host intrusions vary in size from 2 to 180 km in strike length. Intrusion thicknesses can be upwards of 15 km, with the chromite seams varying in thickness from
1 cm to 8 m [Schulte et al., 2010].
The best-studied chromite deposit is that of the Bushveld Complex, located in
South Africa. Other stratiform deposits include the Great Dyke, Zimbabwe, Muskox,
Nunavut, and the Stillwater Complex, Montana. Platinum Group Elements (PGEs)
are typically associated with stratiform deposits, and can feature PGE deposits
within chromite seams of stratiform deposits. This occurs in the Stillwater Complex and the Bushveld Complex. Chromium to iron ratios between 1 and 2.1 are
typical for stratifrom deposits, making them ideal candidates for ferrochrome production [Schulte et al., 2012].
Podiform deposits are named due to the chromite being present in large pods or
lenses which can reach up to 100 m long. These pods, like stratiform deposits, are
hosted in ultramafic rock. Formation of the pods is still unclear, however some researchers suggest partial melting and reforming of chromite, or magma injection into
the ultramafic rock. Unlike stratiform deposits, exploration for podiform deposits
is difficult due to the almost random occurrence of the chromite pods. Similar to
1.5. CHROMIUM PROCESSING
7
stratiform deposits, podiform deposits can contain various PGEs. Ore from podiform
deposits is typically metallurgical or refractory grade [Mosier et al., 2012].
1.5
Chromium Processing
Chromite ore has traditionally been processed using pyrometallurgical methods to
obtain low carbon ferrochrome. Today, this is typically done using submerged arc
furnaces. Chromite ore is added along with a carbon reducing agent and a flux,
typically silica. The mixture is then smelted in the submerged arc furnace. This
process is energy intensive, requiring 4 MWh per tonne of material. Ferrochrome is
then poured out and crushed to the desired size [ICDA, 2011, Lyakishev and Gasik,
1998].
The submerged arc furnace alone does not produce ferrochrome adequate for
steel making. Due to variances in ore grade and composition, additional refining is
required. Ladle refining via the Argon-Oxygen Decarburization (AOD) or VacuumOxygen Decarburization (VOD) process allows for the removal of carbon, nitrogen,
and other impurities. Certain steelmakers can use these processes to produce steel
with less than 100 ppm carbon and nitrogen [Xu et al., 2009]. Both of these decarburization processes involve blowing oxygen into the molten ferrochrome through
tuyeres, with argon in the AOD process, in order to react it with any impurities.
The use of vacuum in the VOD process or inert gas lowers the partial pressure of
gases such as carbon monoxide, improving decarburization [Wei and Zhu, 2002].
Pre-sintering of materials before smelting in a submerged arc furnace is also used
in industry. Pellets of chromite, coke, and flux are premixed and sintered causing
1.6. RESEARCH SCOPE AND OBJECTIVES OF CURRENT WORK8
some reduction to take place. These sintering kilns are typically located above the
submerged arc furnace and make use of the rising hot gas to sinter the pellets [ICDA,
2011].
1.6
Research Scope and Objectives of Current Work
The current work examines the ability of a chromite ore to be reduced using microwaves with carbon as a reducing agent. Few, if any, studies have been done on
this reduction method. There have been previous studies on the decarburization of
ferrochrome using microwaves [Chen et al., 2009, Hao et al., 2014]. Vacuum microwave reduction has been studied by previous researchers on different ore types
[Forster, 2015]. Similar methods are implemented and adapted for the chromite ore
under research. The primary objectives of the current work are to:
1) Construct a thermodynamic model to be the basis of a reduction investigation.
This model would predict the effects of temperature, pressure, and carbon
content on the grade and recovery of chromium and iron.
2) Investigate the effects of initial carbon level, time, power, and atmosphere on the
reduction of chromite ore.
3) Compare the reduced sample to the thermodynamic model and account for any
discrepancies.
4) Recommend future work and optimization of the current system.
1.7. THESIS ORGANIZATION
1.7
9
Thesis Organization
This thesis contains six chapters, excluding the introduction:
Chapter 2: Literature Review A review of all literature pertaining to the extraction and processing of chromite ore. Studies on thermodynamics, reduction,
and kinetics are included.
Chapter 3: Microwave Theory An overview of microwaves and their metallurgical applications. This includes a description of the four key properties to
microwave heating as well as advantages and limitations.
Chapter 4: Experimental Methods and Material Used This chapter outlines
the experimental methods and analytical equipment used during the current
research. A description of the microwave system used for reduction testing can
be found here. The classification of the chromite ore using SEM and x-ray
methods can also be found here. The key variables examined in the current
research are reported.
Chapter 5: Thermodynamics A thermodynamic model was constructed for the
chromite reduction system at atmospheric and reduced pressures between temperatures of 200 and 1400 ◦ C. Activity coefficients were obtained from literature
sources. Several plots are shown displaying the model outputs.
Chapter 6: Results and Discussion The results of the heating and reduction
tests can be found here. Provided are the microwave absorbed power plots as
well as mass loss, energy absorption, grades, and recoveries.
1.7. THESIS ORGANIZATION
10
Chapter 7: Conclusions and Recommended Future Work This chapter shows
the significant findings from the thermodynamic model and reduction studies,
correlating the two and discussing possible discrepancies. Future recommended
work is discussed.
11
Chapter 2
Literature Review
2.1
Overview
Due to the importance of chromium in the stainless steel and alloying industries,
there is a significant amount of research on the reduction of chromite and also ferrochrome decarburization. Various reduction studies have been conducted on the
reduction of chromite ore. Boericke (1944) was one of the first researchers to determine the reduction pathways for chromite ore. These were expanded upon by
various other researchers over time. Thermodynamic and kinetic studies have also
been performed on the ores. Reduction testing in vacuum has also been studied, and
the use of microwave heating on both chromite and ferrochrome has been studied.
Hydrometallurgical methods for the production of chromium compounds are also
explored.
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
2.1.1
12
Iron-Chromium Phase Diagram
Figure 2.1 shows the binary iron-chromium phase diagram at atmospheric pressure.
The phase diagram allows for the prediction of phases present in the final product
and during sample cooling. In the current research a final product of α + α0 is
expected.
Figure 2.1: The iron-chromium binary phase diagram at atmospheric pressure
[Durand-Charre, 2004].
2.2
Reduction Behaviour of Chromite Ores
The reduction behaviour of chromite has been thoroughly studied in literature. Boericke in 1944 found four reactions to describe the reduction of chromium oxide through
carbon [Boericke, 1944]. His original reactions featured a chromium carbide species
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
13
with the equation Cr4 C, and this was corrected in 1963 to Cr23 C6 by J.H. Downing
[Pahlman et al., 1981].
1/3 Cr2 O3 + 13/9 C → 2/9 Cr3 C2 + CO
(2.1)
1/3 Cr2 O3 + 27/15 Cr3 C2 → 13/15 Cr7 C3 + CO
(2.2)
1/3 Cr2 O3 + Cr7 C3 → 1/3 Cr23 C6 + CO
(2.3)
1/3 Cr2 O3 + 1/6 Cr23 C6 → 27/6 Cr + CO
(2.4)
The overall reaction for chromium(III) oxide reduction can be written as:
1/3 Cr2 O3 + C → 2/3 Cr + CO
(2.5)
Boericke determined that reactions 2.1 and 2.2, at atmospheric pressure, began
at a temperature just under 1300 ◦ C. He also found that reactions 2.3 and 2.4 occur
at higher temperatures and require the removal or dilution of CO [Boericke, 1944,
Pahlman et al., 1981].
Wang et al. (2014) performed kinetic analysis on synthetic chromite reduction.
Additional reduction reactions between carbon and oxide species for the formation
of carbide and chromium metal are provided [Wang et al., 2014].
21 Cr2 O3 + 81 C → 14 Cr7 C3 + 63 CO
(2.6)
21 Cr2 O3 + 81 Fe3 C → 14 Cr7 C3 + 243 Fe + 63 CO
(2.7)
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
21 Cr2 O3 + 63 C → 42 Cr + 63 CO
14
(2.8)
Samarin and Vertman (1957) found that the reduction rate of chromium(III)
oxide increased as the temperature increases from 870 to 1370 ◦ C, and found that the
rate of reaction at higher temperatures occurred faster at lower pressures [Samarin
and Vertman, 1957, Pahlman et al., 1981].
Chromium ore is typically found as a spinel with either MgO, FeO, and/or Al2 O3 .
These extra species are important to consider as they can have a large impact on the
reduction process.
Katayama and Tokuda in 1979 studied the reduction behaviour of synthetic
chromite containing both magnesium and iron spinels. They found that pure iron
chromite (FeCr2 O4 ) reduced into FeO and Cr2 O3 at temperatures around 1150 ◦ C.
FeO would quickly reduce with carbon to form metallic iron by the reaction below
[Katayama and Tokuda, 1979].
FeCr2 O4 + C → Fe + Cr2 O3 + CO
(2.9)
This iron would then be converted into an iron carbide with a similar structure to
those chromium carbides seen in reactions 2.1 to 2.3. They proposed the following
change in metallic products during reduction [Katayama and Tokuda, 1979].
Austenite → Austenite+(Fe,Cr)3 C → (Fe,Cr)3 C+(Fe,Cr)7 C3 → (Fe,Cr)7 C3 (2.10)
Katayama and Tokuda showed two potential reduction pathways for magnesium
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
15
chromite. They stated that the temperature required to begin reduction was 1200◦ C,
increasing in rate at temperatures over 1250◦ C. Reactions 2.11 and 2.12 show the
reduction of magnesium chromite [Katayama and Tokuda, 1979].
3 MgCr2 O4 + 17 CO → 2 Cr3 C2 + 3 MgO + 13 CO2
(2.11)
7 MgCr2 O4 + 33 CO → 2 Cr7 C3 + 7 MgO + 27 CO2
(2.12)
Complex chromites, those containing both spinels of iron and magnesium, were
found to have two distinct reduction stages, the first being the iron reduction phase
to FeO occurring at 1150 ◦ C, and the second being the magnesium reduction phase
at 1250 ◦ C. Should the complex spinels contain some alumina (Al2 O3 ), the reduction
temperature was pushed up to 1330 ◦ C [Katayama and Tokuda, 1979].
Finally, the two researchers determined that the final metallic product of the
complex spinels was (Fe,Cr)7 C3 , however those without iron, magnesium and magnesium aluminum chromite, produced Cr3 C2 as a final product [Katayama and Tokuda,
1979].
Lekatou and Walker (1995) studied the mechanisms of solid-state chromite reduction. Researchers studied a Greek chromite ore using reduction temperatures
between 1100 and 1470 ◦ C. Reduction was carried out for 210 minutes. A maximum reduction of 75% was obtained for the 1470 ◦ C degree reduction temperature,
decreasing to a minimum of 10% at 1100 ◦ C. Researchers were able to determine a
six stage reduction process, including reduction reactions for iron. The first stage is
characterized by nucleation of iron on the chromite grain. Researchers determined
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
16
the following reactions for the formation of an iron phase [Lekatou and Walker, 1995].
3 Fe2 O3 + C → 2 Fe3 O4 + CO
(2.13)
Fe3 O4 + C → 3 FeO + CO
(2.14)
Fe3 O4 + CO → 3 FeO + CO2
(2.15)
FeO + C → Fe + CO
(2.16)
FeO + CO → Fe + CO2
(2.17)
CO2 + C → 2 CO
(2.18)
3 Fe + C → Fe3 C
(2.19)
3 Fe + 2 CO → Fe3 C + CO2
(2.20)
The second reduction stage is characterized by the reduction and growth of
chromium and iron in the chromite particles. A metal phase first forms in cracks
where CO or C can enter the grain and reduce the spinel. This stage occurs at
1100 ◦ C and accelerates as the temperature increases. Stage three is formation of a
metallic rim at the boundary of the chromite core and the spinel shell. The fourth
stage involves metallization of the chromite core. The already reduced carbides act
as reducing agents for the oxides present in the chromite core. The fifth stage involves the formation of a slag phase at the spinel boundaries. This phase dissolves
the spinels, allowing for additional reduction of iron and chromium. The final phase
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
17
is a disintegration of the spinels due to cracks. Liquid metal and CO can then penetrate the grains allowing for final reduction of the remaining chromite [Lekatou and
Walker, 1995].
Nafziger et al. (1979) studied the reduction of two chromite ores from the United
States. The two ores used are a metallurgical grade ore from High Plateau, California,
and a high iron chromite from Mouat, Montana. Five potential reducing agents were
used in their experiments, and these included coal char (72.9% C), coke breeze (86.5%
C), metallurgical coke (78.6% C), petroleum coke (89.1% C), and shell carbon (97.5%
C). Researchers found that both chromite ores behaved similarly in their reduction.
Between 1100 and 1300 ◦ C coal char provides the highest degree of reduction for
both ore types. At higher temperatures, approximately 1500 ◦ C, metallurgical coke
is ideal for the high iron chromite from Mouat, while coke breeze is preferred for the
metallurgical chromite from High Plateau [Nafziger et al., 1979].
The researchers stated that the first 15 minutes of reduction have the highest rate
of reduction, with increasing temperatures leading to higher overall reduction and
metallization. They reported sample fusion at temperatures over 1300 ◦ C. Carbon
in the reduced product was reported to be inversely proportional to the degree of
reduction for high iron chromite, and metallurgical chromite was reported to have a
less pronounced trend. Finally, they noted that the ferrochrome product begins as
small beads on the surface of the ore particle, agglomerating to form larger particles
as reduction progresses [Nafziger et al., 1979].
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
2.2.1
18
Thermodynamic Studies of Chromite Reduction
Hino et al. (1998) performed a comprehensive thermodynamic analysis on the reduction of chromite ore with carbon. Researchers first studied the chromite ore by
determining the activity coefficients of FeCr2 O4 and MgCr2 O4 . They used these
to study the effect of both magnesium and aluminium on the activity of chromium
oxide. They found that the chromite spinel system has a negative deviation from
ideality that increases as MgCr2 O4 is replaced by MgAl2 O4 [Hino et al., 1998].
Next, researchers studied the iron-chromium-carbon system. They commented
that few studies have been done on the iron-chromium-carbon system at temperatures between 1300 and 1700 K. It is within this temperature range that several
of the primary reduction reactions occur. They used previous modelling work done
by J.-O. Andersson (1988) [Andersson, 1988] to construct isothermal plots of the
Fe-Cr-C phase diagram. The isothermal plot for 1473 K can be found in Figure 2.2.
These plots were used to study the formation of various phases and their equilibrium
with other species [Hino et al., 1998].
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
19
Figure 2.2: Isothermal section of the Fe-Cr-C phase diagram at 1473 K [Hino et al.,
1998].
Using the previously seen work on the thermodynamics of the chromite ore and
the Fe-Cr-C analysis the researchers were able to study the reduction behaviour
of the chromite ore by carbon. The reduction of FeCr2 O4 was first studied. A
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
20
temperature-partial pressure plot was constructed to show the reduction, and can
be seen in Figure 2.3. Included above the plot are several reactions expected to
occur at various points throughout the reaction. The thick solid line seen next to
the striped phase indicates reduction by carbon materials rather than gas. From the
figure, researchers stated that, in order to fully reduce all oxide, the CO2 /CO gas
ratio must be kept below 10−2.6 . They were also able to determine that the lowest
temperature that could reduce Cr2 O3 in the system was 1390 K. End products of
reduction at 1873 K according to the researchers are liquid iron and Cr3 C2 [Hino
et al., 1998].
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
21
Figure 2.3: Partial pressure-temperature plot for the reduction of FeCr2 O4 , including
various reactions [Hino et al., 1998].
A similar plot was constructed to show the reduction of an iron-magnesium
chromite. Trends were reported to be similar to those seen in Figure 2.3. Due to the
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
22
presence of MgCr2 O4 , however, there are some differences. The liquid metal/carbide
phases shift towards lower CO2 /CO ratios, and reduction of iron chromite in this
case requires a CO2 /CO gas ratio less than 10−2.2 . The lowest temperature to reduce
magnesium chromite was reported to be 1470 K [Hino et al., 1998].
2.2.2
Kinetic Studies of Chromite Reduction
Kekkonen et al. (1995) studied the solid-state kinetics of chromite pellet reduction.
First, pellets of Kemi chromite ore (≈ 40% Cr2 O3 ) were created containing either
0, 2.9, or 5.1% carbon, added in the form of coke powder. Bentonite was added as
a binding agent. Experiments took place in a thermogravimetric analysis (TGA)
machine with either argon or carbon monoxide gas. Tests were carried out at 1420,
1520, and 1595 ◦ C. Researchers found an increase in reduction rate in argon when
increasing the temperature from 1420 to 1520 ◦ C, and a final increase to 1595 ◦ C did
not have this effect. The final mass losses for all argon tests were below the theoretical
maximum. With CO gas, the temperature had no effect on the reduction rate, as all
three temperatures had nearly identical rates. Optical microscope analysis revealed
that reduced samples had two oxide zones, a center zone with high iron content and
an outer layer with lower iron content. Researchers concluded using the unreacted
core model that carbon dioxide diffusion through the product layer was the rate
controlling step for samples reduced in argon [Kekkonen et al., 1995].
Niayesh and Dippenaar (1992) studied the kinetics of chromite reduction using
primarily CO gas. The researchers chose CO gas due to its availability in coaloxygen smelters as an off gas. They stated that the CO gas could be used for
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
23
pre-reduction of chromite ore. Experiments carried out in a TGA used samples of
10 g heated in an inert atmosphere at 1300 ◦ C for 2 hours. Pellets consisting of ore
and carbon were heated in nitrogen gas at various flow rates to determine the degree
of reduction. It was found that increasing the flow rate of the purge gas resulted in
lower reduction degrees and a slower rate of reduction. This was attributed to the
purge gas carrying away the gaseous reduction products, namely CO gas. Composite
pellets were then created by layering graphite between two layers of chromite ore.
Researchers determined that although the reduction of chromite using carbon occurs
with an intermediate gas step, the use of CO gas alone is not enough for reduction.
The researchers found three distinct reduction stages of composite pellets under
isothermal reduction at 1300 ◦ C. In the first stage, the supply of CO gas by carbon
gasification was determined to be the rate-limiting step. The activation energy was
found to be 125 kJ/mol, with 60 to 85% of reduction occurring during this stage. The
second stage occurs when the CO gas reaches an equilibrium. During this step the
metal begins to form around the oxide particles. Finally, in the last stage productlayer diffusion was determined to be the rate-limiting step. The activation energy at
this step was 238 kJ/mol [Niayesh and Dippenaar, 1992].
Wang et al. (2014) studied the kinetics of carbothermic reduction of a synthetic chromite ore. Pellets of synthetic FeCr2 O4 were produced by first combining
Fe powder with Fe3 O4 in pellets pressed at 12 MPa and heating them at 1100 ◦ C
for 10 hours. This process formed FeO, which was later combined with Cr2 O3 under the same conditions for 48 hours to form FeCr2 O4 . Graphite was used as the
2.2. REDUCTION BEHAVIOUR OF CHROMITE ORES
24
reducing agent. Researchers conducted both isothermal and non-isothermal experiments. Isothermal experiments were conducted in a molybdenum-wire furnace at
temperatures between 1473 and 1673 K. Non-isothermal tests were done in a Netzsch
STA449C TGA with temperatures up to 1773 K. Both sets of tests were conducted
in an argon atmosphere. Reduction in non-isothermal conditions reached 90% at
1773 K. Under isothermal conditions, reductions only reached 80% at 1673 K, decreasing to a minimum of 50% at 1473 K, both after 40 minutes of reducing. Higher
temperatures resulted in higher rates of reaction. Upon studying the effect of carbon content, researchers found that carbon additions of 1.1 and 1.2 times the molar
ratio to chromite led to higher reductions. With a carbon ratio of 1.2 the reduction
amount reached 95%. Researchers stated that the carbon was both reducing the
sample and aiding in the formation of metal carbides which allowed for more alloy
to be formed. It was found that iron was reduced first, either forming iron or iron
carbide. Chromium oxide was then reduced using either carbon or the iron carbide to form chromium carbide. Upon carbon depletion the reactions cease, leaving
metallic iron and chromium, chromium and iron carbides, and unreacted oxide. A
kinetic analysis was performed, and the first stage was determined to be controlled
by nucleation and had an activation energy of 120 kJ/mol, while the second stage
was determined to be limited by crystallographic transformation with an activation
energy of 288 kJ/mol [Wang et al., 2014].
2.3. CONVENTIONAL HEATING
2.3
25
Conventional Heating
Chromite processing with conventional heating is covered in Section 1.5 in the Introduction Chapter. There is however a growing body of research on conventional
heating of chromite and ferrochrome due to its importance in the iron and steel
industries.
Ranganathan (1998) performed thermochemical analysis on ferrochromium produced in submerged arc furnaces (SAF). In the reduction process of chromite ores,
iron and chromium are completely reduced in the SAF while any magnesium, calcium, and aluminium oxides transfer into the slag. Silica partially reduces and is
present in the metal as silicon metal and in the slag as silica. The reduction order is
typically iron, followed by chromium, and finally by silicon. Two layers are present in
the furnace, a lower temperature area where kinetics are the primary reducing factor,
and a hotter area where thermodynamic equilibrium can be expected. Ranganathan
focused on the high temperature area, namely the slag-metal interface. As charges
descend in the furnace, they encounter gases produced lower in the furnace. This
gas alters the metal and slag equilibrium. The metal equilibrium is not affected until
exposed to temperatures over 2000◦ C, at which point silicon reports to the metal
solution. This temperature is a boundary between two thermochemical zones. At
temperatures above 2000 ◦ C, silicon is expected to report to the slag phase if it is in
contact with it. Should two thermochemical zones of different temperature exist, it
becomes difficult to remove silicon as it will not be in contact with the slag. As such
it is important to remove or minimize the lower temperature zone to remove silicon
from solution. Ranganathan then analyzed mass balances of the system, noting that
2.3. CONVENTIONAL HEATING
26
the amount of carbon used in the system was significantly higher than that which
was accounted for by the reduction of iron, chromium, and silicon and any carbon
in the metal phase. A carbon balance showed that additional chromium and iron
are reduced and then trapped in the slag phase. Ranganathan confirmed this using
XRD which showed chromium in the slag phase. The thermochemical model constructed was then compared to the actual output and carbon balance predictions.
Both the model prediction and carbon balance produce higher values then the actual production. The amount of chromium in the slag phase was also higher than
predicted by the model. The model can be further used to better understand the
higher temperature region of ferrochrome reduction [Ranganathan, 1998].
Privalov et al. (2008) studied characteristics of the Aksai Ferroalloy Plant. Theoretical research on the reduction of oxides by carbon has been ongoing for 15 years
with the goal of improving technology. A table of data from 2005-07 was compiled
including power consumption, extraction, grades, yield, and input materials. Researchers then examined the two years worth of data to determine improvements
that could be made to the plant. It was found that 3.18% of chromium was lost
due to being trapped in the slag, and 3.14% was lost as dust. The grades across all
furnaces were found to be within 1.2% of each other. Researchers suggested that slag
must be further processed to remove more trapped chromium, and that additional
research is required to minimize dust and other chromium losses [Privalov et al.,
2008].
Statnykh et al. (2013) studied ways to improve production of low-carbon ferrochrome, namely looking at the OAO Serovskii Zavod Ferrosplavov. This plant
2.3. CONVENTIONAL HEATING
27
utilized aluminothermic reduction to achieve ferrochrome with 75% Cr, 0.03% C,
and 0.4% Al. Aluminium is added to the system as aluminium powder in the primary charge. This method is not common as it suffers from low productivity and
high cost, as the aluminium used is lost during the enrichment and is expensive
to replace. Researchers compared three ferrochrome production methods used elsewhere to determine if an improved process could be used. The ideal process based
on the compared processes was a two stage process with variable slag composition
between each stage. In this process, ferrosilicochrome is added into a ladle with an
ore-lime melt. Additional chromite ore is added into this ladle. The alloy from this
first stage is 4% Si and 0.1% C. The second stage has an increase in slag basicity,
which allows for low silicon grades as it is removed in the slag phase. Comparing
this process to the original process, the researchers determined that the productivity
could be increased by 7.2%, while reducing power consumption by 9.9% [Statnykh
et al., 2013].
2.3.1
Other Ferrochrome Production Methods
In addition to the common use of a submerged arc furnace for the reduction and
production of ferrochrome, several other methods have been developed. Aspects of
some of these methods have been applied to current methods, and some of the listed
methods are in use in ferroalloy plants around the world.
The Perrin Process, seen in Figure 2.4, is an alternative process for the production
of low-carbon ferrochrome. The process involves two stages. The first stage involves
the smelting of chromite ore with carbon and quartz to create silicochromium. This
2.3. CONVENTIONAL HEATING
28
is then transferred to a ladle for refining. At the same time, a chromium rich slag
is created using chromite and lime. This slag, along with the product from the first
ladle, are mixed in a second ladle. Silicon is removed in this ladle by reacting with
the chromium oxide slag introduced, forming silica, which reports to the slag phase.
The slag from this second ladle is recycled back to the first. The product from the
second ladle is ferrochromium with less than 0.5% carbon. Waste slag from the first
ladle contains less than 1% chromium oxide [Östberg, 2003, Robiette, 1973].
Figure 2.4: The Perrin Process, from Electric Smelting Processes, 1973 [Robiette,
1973].
The Perrin Process works on the principle of silicothermic reduction. Increasing
silicon in an alloy decreases the solubility of carbon, allowing for very low carbon
grades to be produced. Smelting is typically done in furnaces lined with magnesium
oxide [Bhonde et al., 2007].
2.3. CONVENTIONAL HEATING
29
The Duplex Process, another silicothermic reduction process, is a method for
the production of low carbon ferrochrome. The process begins with the smelting of
silicochromium in an electric arc furnace using coke and quartz. The alloy is then
poured, cooled, and crushed. This crushed SiCr is then placed into a ladle for further
refining. The slag used is premade in a slag furnace using chromite ore and lime.
After refining, the alloy contains between 12 and 14% Si. This alloy is transferred to
another ladle while the slag is taken off. The second ladle uses slag from the same
source as the first ladle. The final alloy is 65-70% Cr, <1% Si, 0.03% C. The waste
slag contains 3% chromium oxide. This process uses 3200-3500 kWh/MT of ore, not
including the energy required to produce the initial SiCr [Ghose et al., 1983].
The Outokumpu Process was developed in 1968 in response to a low-grade chromite
deposit discovered in Northern Finland in 1959. A flowsheet of the process can be
seen in Figure 2.5. The low-grade ore is concentrated using gravity separation methods to 40 to 47% chromite. This concentrate is combined with recycled fines from
the sintering process. The concentrate and fines are then ground and pelletized using
a bentonite binder. They are then passed into a column sintering furnace. Fuel from
the furnace comes from the off-gases of the submerged arc furnace seen later in the
process and dust from coke drying. The pellets are then mixed with coke and flux
and are fed into a rotary kiln with off-gas from the submerged arc furnace providing
heat for preheating. It is estimated that the sintering of pellets and the pre-reduction
saves 800 kWh/t of ferrochrome produced. Energy use for the submerged arc furnace is 2600-2800 kWh/t ferrochrome. Final grades for the produced ferrochrome
are 53.5% Cr, 2.5% Si, and 7% C, the remainder being Fe [Relander and Honkaniemi,
2.3. CONVENTIONAL HEATING
30
1985].
Figure 2.5: The flowsheet for the Outokumpu Process [Relander and Honkaniemi,
1985].
The Triplex Process is primarily a method for the production of high-grade steels.
This process involves the use of an electric arc furnace, a converter, and finally
vacuum-oxygen decarburization (VOD) to produce steels with extremely low impurity contents. Feed into the electric arc furnace is high carbon ferrochrome, scrap,
and stainless steel. Phosphorus and sulphur are both controlled during this stage.
This stage and the next both take 60 minutes to complete. The molten alloy is then
transferred to an oxygen top and bottom blown converter, where carbon is reduced
to 0.4%. The contents are then moved to a VOD furnace for final refining. This
further reduced the carbon to 0.03%. The effect of time on chromium and carbon in
the VOD furnace can be seen in Figure 2.6. After 70 minutes, the melt is rinsed with
argon and nitrogen to control other inclusions. Nitrogen is picked up by the alloy
2.4. MICROWAVE HEATING
31
during this process. The use of vacuum ensures that as little gases as possible are
trapped in the alloy. Nitrogen grades are typically around 0.17% after 180 minutes
of decarburization. Additional elements such as Ni, Cu, Mo, or Ti can be added
during the VOD process [Nair et al., 2013].
Figure 2.6: VOD effect on carbon and chromium over time, adapted from [Nair et al.,
2013].
2.4
Microwave Heating
There have been few studies done on the use of microwaves on chromite ore or its
products.
Bayat et al. (2013) studied the effects of microwave pre-treatment on chromite
ores. They examined the effect of time and power on the fineness of ground chromite
2.4. MICROWAVE HEATING
32
ore. Researchers found that microwaving chromite ores gave a 32.20% decrease in the
50% cumulative undersize [Bayat et al., 2013]. Similar studies have been conducted
on other ore types as well. Amankwah and Ofori-Sarpong examined microwave
pre-treatment for gold ore. They found that microwaved ore had a reduced crushing
strength of 31.2%, and increased extractions after leaching, achieving 95% extraction
after 12 hours compared to 22 hours for non-microwaved ore [Amankwah and OforiSarpong, 2011]. Finally, Walkiewicz et al. in 1988 and 1993 performed various
microwave heating tests. In 1988 they studied the heating effects on various samples,
including ores, liquids, and metal powder. They discovered that magnetite is an
extremely efficient absorber of microwaves, going so far as to suggest doping other
ores with magnetite to increase its microwave absorption. They also discovered
that the mechanism behind finer grinding was fracturing induced by heat stress
[Walkiewicz et al., 1988]. In 1993, Walkiewicz et al. found that pre-treatment of a
taconite ore resulted in a decrease of 13% in the work index of the ore [Walkiewicz
et al., 1993].
Chen et al. (2009) used microwave heating as a way to produce sponge ferrochrome with a low carbon content. Researchers combined Indian chromite, anthracite (carbon reducing agent), and lime, totalling 1 kg, in specific proportions
and placed the mixture in a metallurgical microwave oven. Microwaves used were
2.45 GHz with a power level of 10 kW. Reaction temperatures were 1273 to 1573 K.
Researchers determined that increasing the reaction temperature caused an increase
in both chromium and carbon found in the ferrochrome. Increasing reaction time
caused chromium grades to increase while decreasing carbon grade. They were able
2.4. MICROWAVE HEATING
33
to obtain carbon grades as low as 4.68% in the ferrochrome at 1273 K [Chen et al.,
2009].
Hao et al. (2014) studied microwave decarburization of high carbon ferrochrome.
Ferrochrome with a carbon content of 8.16% was crushed and combined with calcium
carbonate in a 1:1.4 ratio. The combined powder sample was placed in a metallurgical
microwave oven. The microwaves used had a frequency of 2.45 GHz with a power
of 10 kW. Reaction temperatures were 1173 to 1473 K with reaction times of 20,
40, and 60 minutes. Researchers found that there was less carbon in the sample at
higher temperatures, but noted that the rate of carbon loss decreased with time at
higher temperatures. Figure 2.7 shows the effect of reaction time and temperature
on carbon content. Researchers were able to determine an activation energy for
the reaction of 72.80 kJ/mol. This energy is stated to be lower than that seen in
conventional heating. They also determined that carbon dioxide first reacts with
(Cr,Fe)7 C3 before other carbides [Hao et al., 2014].
2.4. MICROWAVE HEATING
34
Figure 2.7: The effect of holding time on carbon content at various temperatures
[Hao et al., 2014].
Li et al. (2014) improved upon the work done by Hao et al. (2014) by determining the dielectric properties of high carbon ferrochrome during decarburization.
Materials and equipment used by Li et al. are the same as those used by Hao et al.
A vector network analyzer was used to measure permittivity and permeability over a
range of 1 to 18 GHz. They stated that in the carbide materials the bonds between
carbon and chromium or iron vibrate in reaction to the microwaves. This weakens the
bonds by stretching them resulting in lower activation energies. Researchers found
that as samples decarburized a change was seen in the dielectric constants. Permittivities were found to increase as carburization occurred, followed by a decrease and
eventual leveling-off associated by the final decarburization step of carbide to pure
metal. Researchers concluded that the change in these electromagnetic properties
reflect internal changes during the decarburization process [Li et al., 2014].
2.5. VACUUM PROCESSING
2.5
35
Vacuum Processing
In 1981, researchers Pahlman et al. from the Bureau of Mines conducted experiments
on chromite and carbon material pellets in vacuum at high temperatures. They
conducted tests at pressures of 0.1 to 1 torr with temperatures ranging from 1230
to 1320 ◦ C and were able to achieve ferrochromium alloys with less than 2% carbon
[Pahlman et al., 1981].
Three of the factors studied by Pahlman et al. were the effect of pressure, carbon
reductant, and temperature. Figures 2.8 to Figure 2.10, from their report, outline
their results. Unless otherwise stated, the samples in each experiment were pellets
made from -400 mesh graphite and chromite concentrate [Pahlman et al., 1981].
Figure 2.8 shows the effects of pressure on reduction. It can be seen that higher
levels of reductions occur at lower pressures. Curve ‘A’ goes above the 100% reduction line due to chromium vaporization at those temperatures and pressures.
Due to this phenomenon, the researchers suggest a minimum reduction pressure 10
times that of the vapour pressure of chromium metal for the reduction temperature
[Pahlman et al., 1981].
2.5. VACUUM PROCESSING
36
Figure 2.8: Effect of pressure on total reduction at 1300 ◦ C [Pahlman et al., 1981].
Researchers tested four different carbon materials as a reducing agent, and their
effects on total reduction can be seen in Figure 2.9. It is seen that foundry coke
followed by carbon black are the two best reduction agents. No explanation was
given by the researchers as to why the coke and carbon black were superior [Pahlman
et al., 1981].
2.5. VACUUM PROCESSING
37
Figure 2.9: Effect of reducing material on total reduction at 1270◦ C and 1 torr
[Pahlman et al., 1981].
The effect of temperature on reduction time can be seen in Figure 2.10. As
expected, increasing the temperature leads to an increased reaction time. It can be
seen that the rate of reduction increases quickly in the first 100 minutes of reduction,
but, after this the reduction rate for all temperatures decreases. After 800 minutes
of reduction, the 1320 and 1310 ◦ C tests both had achieved 100% reduction, while
at lower temperatures a reduction of just under 70% was achieved.
2.5. VACUUM PROCESSING
38
Figure 2.10: Effect of temperature on total reduction at 1 torr and using graphite as
a carbon source [Pahlman et al., 1981].
2.5.1
Energy Savings
Pahlman et al. compared the energy used in three low-carbon ferrochrome producing
processes to an estimation of the energy required to create the same amount of
ferrochrome using a vacuum reduction method. The three methods examined are the
Simplex, Perrin, and submerged-arc smelting and AOD Processes. First, researchers
determined that an estimated 4900 to 7200 kWh is required to smelt 1 long ton (LT)
of low-carbon ferrochrome. Through private communications with industry sources,
the researchers obtained the energy requirements for each of the three processes
stated earlier. Table 2.1 illustrates each process, the energy requirements stated by
industry, and the potential energy savings [Pahlman et al., 1981].
2.6. HYDROMETALLURGY
39
Table 2.1: Energy requirements and savings comparision between three conventional
FeCr processes and vacuum-reduction [Pahlman et al., 1981].
Method
Energy (kWh/LT)
Savings (kWh/LT)
Simplex Process
8500 - 12400
3600 - 5200
Perrin Process
8950
1750 - 4050
Submerged-arc / AOD Processes
5400 - 6300 (high carbon only)
AOD decarburization
Researchers noted, that for the Simplex process, additional energy savings of
450 to 900 kWh/LT could be obtained through adding iron to the feed to produce
stainless steel directly. For the submerged arc / AOD processes, their energy expenditures to produce high-carbon ferrochrome are similar to that of the proposed
vacuum reduction. Savings would be limited to any additional decarburization required [Pahlman et al., 1981].
2.6
Hydrometallurgy
Hydrometallurgical processes for chromium extraction from chromite are used in the
preparation of chromium compounds.
The conventional preparation method for the formation of chromium chemicals
first involves the roasting of chromite ore with soda ash to form sodium chromate
[Vardar et al., 1994]. This step is done to convert insoluble Cr(III) into soluble
Cr(VI). This process is done using alkali oxidation of chromite, as seen in the following reactions [Zhang et al., 2016].
FeCr2 O4 + 2 Na2 CO3 + 7/4 O2 → 2 Na2 CrO4 + 1/2 Fe2 O3 + 2 CO2
(2.21)
2.6. HYDROMETALLURGY
40
Fe2 O3 + Na2 CO3 → 2 NaFeO2 + CO2
(2.22)
FeCr2 O4 + 4 Na2 CO3 + 7/4 O2 → 2 Na2 CrO4 + 5/2 Fe2 O3
(2.23)
The use of sulphuric acid leaching leads to extraction of chromium in its Cr(III)
state. Cr(VI)is environmentally dangerous and carcinogenic in humans [Zhang et al.,
2016]. Limiting its production using acid leaching is therefore desirable.
Amer (1992) studied the extraction of chromite from an Egyptian ore using sulphuric acid. The reaction for chromite leaching using sulphuric acid is given by
reaction 2.25 [Amer, 1992, Geveci et al., 2002].
FeCr2 O4 + 4 H2 SO4 → Cr2 (SO4 )3 + FeSO4 + 4 H2 O
(2.24)
Ore was ground in a ball mill to size fractions of 64, 80, and 250 µm. Factors
studied by Amer were the temperature, acid concentration, grain size, solid/liquid
ratio, and leach time. The effect of leaching time and temperature can be seen in
Figure 2.11. It was discovered that high temperatures (250 ◦ C) had extraction values
of 90% Cr. This decreased to 30% at 180 ◦ C. Leaching times for both of these tests
were 120 minutes [Amer, 1992].
2.6. HYDROMETALLURGY
41
Figure 2.11: Effect of leaching time on chromium extraction for various temperatures
[Amer, 1992].
Using the ideal test criteria as before, the effect of particle size was determined.
Smaller particle sizes, at -64 µm led to recoveries of 90% while larger sizes, at 250 µm, had recoveries of 30%. The addition of manganese oxide to the system
led to a decrease of dissolved iron in solution. A MnO2 /FeCr2 O4 ratio increase
from 0.04 to 0.4 led to a decrease in dissolved iron of 15% from 27 to 12% after 15
minutes of leaching. This trend continues to approximately 8% dissolved iron after
120 minutes for the 0.4 MnO2 /FeCr2 O4 ratio. It was also noted that aluminium
leaches with temperatures up to 180 ◦ C. Above this temperature the leaching of
aluminum decreased, and this was assumed to be due to the hydrolysis of aluminium
sulphate [Amer, 1992].
2.6. HYDROMETALLURGY
42
Vardar et al. (1994) studied leaching of Bushveld complex chromite using sulphuric acid. Samples were crushed to a size of -90+75 µm, with a constant solid/liquid
ratio of 1.25 for each experiment. Temperatures varied from 140 to 210 ◦ C, leach
times varied from 2 to 6 hours, and sulphuric acid concentrations varied from 60 to
90% by weight. Higher temperatures led to higher extraction of chromium. At 210
◦
C extraction was 65% after 6 hours. This decreased to a minimum of 28% for 170
◦
C. The effect of sulphuric acid was less pronounced. The difference between 90%
and 65% sulphuric acid was less than 10% after 6 hours. Perchloric acid could be
added to improve the extraction of chromite. With a perchloric acid/chromite ratio
of 1/2 the extraction of chromite could be increased to almost 100% after 6 hours at
210 ◦ C [Vardar et al., 1994].
Geveci et al. (2002) studied leaching of a Turkish chromite concentrate with
sulphuric acid and perchloric acid. As with the previous researchers, the ore was
crushed and had the temperature, duration, acid concentration, and perchloric acid
ratio varied. Increasing the sulphuric acid concentration from 70 to 90% by volume
decreased the chromium recovery from 58 to 47%. Researchers attributed this to a
decrease in reactivity of the acid at higher concentrations. Increasing the temperature
from 140 to 175 ◦ C increased the recovery from 10 to 45%, however further increasing
the temperature did not have any effect on the recovery. Finally, the addition of
perchloric acid increased the chromium recovery. A ratio of acid/ore of 1/2 led to
recoveries of 83% after 2 hours at 175 ◦ C. Compared to the results of Vardar et
al. (1994), the results here are lower. This was attributed to sample chemistry, as
the ore currently in question has additional magnesia and silica which could reduce
2.6. HYDROMETALLURGY
43
chromite extraction [Geveci et al., 2002].
Zhao et al. (2014) performed a detailed analysis of a sulphuric acid leach with
an oxidant on a South African chromite. They tested both powdered chromite and
lump chromite, which was polished prior to testing. The temperature during testing
was 160 ◦ C unless otherwise stated. Analysis of the silica gangue phase over the leach
duration indicated that aluminium and magnesium leached out at the same time,
leaving behind a porous silicate. The chromite spinel phase decreased in volume as
the leach proceeded. If the spinel was coated in the silicate phase then the leach
proceeded very slowly for that chromite. This could be reduced by lowering the
particle size. Figure 2.12 shows the effect of leach time and particle size on the
extraction of chromite. It can be seen that lower particle sizes, at -39 µm, were able
to achieve extractions of 95% after 90 minutes, while the largest size fraction, +125
µm, reached 45% extraction in the same leach time [Zhao et al., 2014].
2.6. HYDROMETALLURGY
44
Figure 2.12: Effect of leaching time on extraction of chromium for various particle
sizes [Zhao et al., 2014].
Zhao et al. also showed the typical leaching behaviour of a chromite lump with
sulphuric acid. Chromite found in the silicate matrix gradually shrinks while in
contact with acid, leaving a void once the spinel fully dissolves. Smaller chromite
grains may fall out of the silicate once they react enough, again leaving voids. These
voids allow for chromite grains deeper in the sample to leach out. Finally, an analysis
of metal leaching order determined that all four primary metals (Fe, Cr, Al, Mg)
leached into solution simultaneously [Zhao et al., 2014].
Amer and Ibrahim (1996) studied the leaching effects of sodium hydroxide on a
low-grade Egyptian chromite ore. Ore was ground in sodium hydroxide solution for
between 0 and 25 minutes. Maximum chromite extraction of 90% was obtained for
a 25 minute grind followed by a 90 minute leach at 240 ◦ C with a sodium hydroxide
2.6. HYDROMETALLURGY
45
concentration of 200 g/L. Oxygen partial pressure was 10 bar. Researchers were able
to determine the reaction constant and constructed an Arrhenius plot to calculate an
activation energy of 6 kJ/mol. The reaction was found to be first-order. Researchers
stated that this suggested that the leaching of chromite is controlled by diffusion
through a boundary layer [Amer and Ibrahim, 1996].
The chromite leaching reaction can be seen in reaction 2.25. Prior to this reaction,
it was stated that the chromium oxide dissociates into ions at the oxide/solution interface, followed by the sample reacting with dissolved oxygen to become Cr2 O5 . The
final oxidation reaction produces the sodium chromate product [Amer and Ibrahim,
1996].
Cr2 O5 + 1/2 O2 + 4 NaOH → 2 Na2 CrO4 + 2 H2 O
(2.25)
Xu et al. (2005) examined the leaching of a Vietnamese ore using potassium
hydroxide. Leaching reactions can be seen in reactions 2.26 through 2.28. The
leaching temperature was 300 ◦ C with the ore being crushed to -70 µm. Leaching
tests were conducted between 0.5 and 6 hours. The conversion of chromium was found
to increase with increasing temperature. The primary reaction temperature of 300 ◦ C
led to a chromium conversion of 95% after 6 hours. Further increases in temperatures
led to higher conversions. The minimum conversion achieved was 60% after 6 hours
at a temperature of 260 ◦ C. The solubility of potassium chromate was found to
be low as the concentration of potassium hydroxide increased. This caused the
potassium chromate to precipitate out as crystals, and these could be separated from
the remainder of the leach residue using gravity separation methods. Researchers also
2.7. SUMMARY
46
conducted a kinetic analysis, and an activation energy of 52.5 kJ/mol was determined.
The process was determined to be chemically rate controlled. It was also found
that aluminium and silicon would dissolve into solution, however they would form a
potassium aluminosilicate and quickly precipitate out of solution [Xu et al., 2005].
1/2 FeCr2 O4 + 2 KOH + 7/8 O2 → 1/4 Fe2 O3 + K2 CrO4 + H2 O
(2.26)
1/2 MgCr2 O4 + 2 KOH + 3/4 O2 → 1/2 MgO + K2 CrO4 + H2 O
(2.27)
1/2 Cr2 O3 + 2 KOH + 3/4 O2 → K2 CrO4 + H2 O
2.7
(2.28)
Summary
The reduction of chromite ore is a complex process due to the spinels that make
up the ore. The minimum thermodynamic temperature required to reduce iron
chromite is 1116 ◦ C. This temperature increases to 1197 ◦ C for magnesium chromite.
Reduction of chromite is carried out using carbon monoxide or a carbon reducing
agent such as metallurgical coal or coke. The use of CO gas by itself is not enough
as the reduction reaction requires carbon in contact with the spinel. Iron is the first
species to be reduced, forming either metallic iron or iron carbide as a reduction
product. After almost all of the iron leaves the system, chromium begins to reduce
to chromium carbide. The final reduction product for the reduction of chromite is
typically metallic iron and chromium carbide. This ferrochrome can then be sold to
steel plants for use in stainless steels.
2.7. SUMMARY
47
Vacuum processing was suggested in 1981, however no additional work has been
done on the process. This is assumed to be due to the capital costs required and
the development of other reduction methods including silicothermic reduction that
can achieve similar carbon grades without the need for a vacuum system. Microwave
heating has also been studied, however there have been very few studies actually conducted. Tests have been conducted on the use of microwaves in sponge ferrochrome
production from chromite fines. Researchers have also conducted tests on the use of
microwaves as a method to decarburize ferrochrome, and energy requirements were
stated to be lower than those of industry however the decarburization levels achieved
did not come close to those seen in AOD/VOD processes. No literature was found
on the microwave processing of chromite ores as a primary heating method.
Finally, hydrometallurgical methods involving sulphuric and perchloric acid, as
well as sodium and potassium hydroxide, were examined. Traditional chromite processing to produce chromium chemicals involves oxygen roasting to produce sodium
chromite, a species with dangerous Cr(VI). The use of sulphuric acid, sometimes with
the addition of perchloric acid, allows for the leaching of chromite into chromium sulphate, which contains the safer Cr(III). The addition of perchloric acid can increase
extractions to over 90%. The use of sodium and potassium hydroxides do produce
species with hexavalent chromium however they are able to achieve extractions of
over 95%.
48
Chapter 3
Microwave Theory
3.1
Microwave Background
Microwaves are a form of electromagnetic radiation, a diagram of which can be seen
in Figure 3.1, with wavelengths, λ, from 1 m to 1 mm. These wavelengths correspond
to frequencies of 300 MHz to 300 GHz. As part of the electromagnetic spectrum, microwaves obey the Laws of Optics, meaning that they can be absorbed, transmitted,
and reflected by various materials. Frequencies typically used by microwaves are 915
MHz and 2.45 GHz in both domestic and industrial applications [Chandrasekaran
et al., 2012]. The microwave used for the current research had a frequency of 2.45
GHz and was supplied by Gerling Applied Engineering.
3.2. MICROWAVE PROPERTIES
49
Figure 3.1: A diagram of an electromagnetic wave [ploufandsplash, 2007].
3.2
Microwave Properties
Heating from microwaves differs from conventional heating methods due to the heat
being generated from complex interactions between the microwaves and the sample
itself [Gupta and Wong, 2007]. Conventional heating relies on conduction and convection to heat a sample. In microwave heating, the heat is generated inside of the
sample.
Four important factors determine how well a material will be heated by microwaves. These are the real permittivity, 0 , the imaginary permittivity, 00 , the loss
tangent, tan δ, and the penetration depth, Dp .
Permittivities, or dielectric properties, are important in determining heating capabilities of the material in question. The real permittivity, 0 , or dielectric constant
3.2. MICROWAVE PROPERTIES
50
represents the ability of microwaves to penetration the material. Imaginary permittivity, 00 , or dielectric loss, represents how well the material is able to convert
microwave energy into heat [Chandrasekaran et al., 2012, Gupta and Wong, 2007].
These dielectric properties are a function of material temperature and microwave
frequency.
The loss tangent, tan δ, is related to the dielectric properties based on Equation
3.1. The loss tangent represents the efficiency of the material to convert any absorbed
energy into heat.
tan δ =
00
0
(3.1)
The penetration depth is the distance to where the microwave energy has been
reduced to 1/e (≈ 0.3679) of its energy on the surface. The penetration depth can
be calculated using Equation 3.2.
Dp =
1
2α
(3.2)
The attenuation factor, α, can be determined using Equation 3.3.
r
α = 2πf
µ0 µ0 0 0
·
2
qp
1 + (tan δ)2 − 1
(3.3)
where: f is the frequency, µ0 is the permeability of free space (4π × 10−7 H/m),
µ0 is the permeability, and 0 is the permittivity of free space (8.8541 × 10−12 F/m)
[Gupta and Wong, 2007].
The rate at which the temperature of the material rises can also be expressed via
3.3. MICROWAVE HEATING EFFECTS
51
Equation 3.4. This can be extremely important as the temperature of a microwave
reaction may not be easy to measure directly. By knowing how a material will behave
during processing, estimations of internal temperature can be made.
2πf 0 00eff E2rms
∆T
=
∆t
ρcp
(3.4)
Where 00eff is the effective relative dielectric loss factor, E is the electric field
strength, ρ the sample density, and cp the specific heat of the sample [Gupta and
Wong, 2007].
3.3
Microwave Heating Effects
Microwaves heat samples internally as opposed to conventional heating which heats
them from the outside in. Figure 3.2 shows the expanding core model present in
microwave heating. In this model, A, the unreacted material, gets reduced as the
core, B, gradually expands. The temperature of the core, T2 , is higher than that of
the starting material, T1 . The core expands either as the initial material absorbs
enough microwaves to reach critical temperature, or conductive heating occurs on
the boundary of the core and unreacted sample. This core is produced as a result of
the microwave properties described in the previous section. The penetration depth
typically decreases as the sample heats, and this leads to greater absorption in the
hot core of the system.
3.3. MICROWAVE HEATING EFFECTS
52
Figure 3.2: Expanding core model of microwave heating.
These heating effects are generated through the interaction between the electromagnetic field and the sample. There are two methods in which heating occurs.
Dipolar rotation, or dielectric heating, occurs when polar molecules, such as water,
interact with the changing electromagnetic field. The molecules attempt to align
themselves to the field which alternates at the frequency of the electromagnetic wave.
In the microwave system used for the current research, 2.45 GHz corresponds to the
field changing alignment 2450 million times per second. Polar molecules attempting
to align to the field will generate internal friction within the sample, causing it to
heat up. The second method is called ion heating or AC resistance heating. This
occurs when charged particles in the sample, such as electrons, shift to adjust to the
electromagnetic field. Ohmic heating in the sample occurs as a current generated
within the material. This phenomenon occurs in carbon. Due to carbon’s ability to
reduce oxides and other materials, it is an excellent choice for microwave processing
applications [Pickles, 2009, Menendez et al., 2010].
Every material behaves differently under microwave stimulation. Materials with
3.3. MICROWAVE HEATING EFFECTS
53
low dielectric losses, such as quartz and aluminium oxide, have dielectric loss values of
0.001 and 0.009, respectively. These materials have extremely large half power penetration depths, essentially making them invisible to microwaves [Gupta and Wong,
2007]. The reverse is true for materials with high dielectric loss values such as metals.
Thostenson and Chou, 1999, stated that materials with high conductivity and low
capacitance such as metals act as microwave reflectors due to an extremely small
power penetration depth. This allows for microwave applicators and waveguides to
be constructed from metal without the risk of arcing or absorbing many microwaves.
Figure 3.3 shows the power absorbed per unit volume as a function of dielectric loss
factor. Thostenson and Chou stated that materials in the center of the dielectric loss
portion are ideal for microwave heating, while materials with high dielectric loss are
better heated by conventional methods [Thostenson and Chou, 1999].
3.3. MICROWAVE HEATING EFFECTS
54
Figure 3.3: The effect of dielectric loss on power absorbed per unit volume [Thostenson and Chou, 1999].
Walkiewicz, Kazonich, and McGill in 1988 tested several substances in powder
form including metals and ores for their microwave heating characteristics. They
found powdered metals to be good absorbers of microwave energy, and able to quickly
reach temperatures over 600◦ C. They also found that microwave heating of certain
materials caused stress cracks in materials with matrix minerals that did not absorb
microwaves. These cracks in minerals allowed for increased extraction of copper from
a chalcopyrite ore. These cracks also decrease the work index of an ore, suggesting
that the use of a microwave pretreatment step could save energy during crushing and
grinding. The researchers stated that these experiments would need to be scaled up
to determine any true benefit [Walkiewicz et al., 1988, Walkiewicz et al., 1993].
3.4. MICROWAVE ADVANTAGES AND LIMITATIONS
3.4
55
Microwave Advantages and Limitations
Microwave heating has various advantages and limitations compared to conventional
heating. Microwave heating is capable of rapid, penetrating, selective heating that
is able to self limit. Microwave heating tends to be more rapid then conventional
heating due to the heat being generated inside the sample itself, and the majority of
samples become better microwave absorbers the hotter they get due to increases in
dielectric properties at higher temperatures. This leads to a positive feedback loop
which causes samples to quickly heat up. The penetrative heating aspect allows for
microwaves to be used in a variety of atmospheres and testing conditions. Crucibles
and material holders can be used that do not interact with microwaves, allowing for
complex setups within an applicator that are not possible with conventional heating.
In the current research, quartz crucibles were used to hold the sample for processing,
which is placed on a refractory material comprised of silica and alumina. Both
of these materials are invisible to microwaves and do not impede the microwaves,
allowing all of the microwaves to interact only with the sample. For vacuum and
argon testing, a glass bell jar was used. This glass is invisible to microwaves and
allows transmission of the microwaves to the sample. Microwave heating is also selflimiting, meaning that at high temperatures the dielectric properties, namely the
dielectric loss, drops and makes the material less susceptible to microwaves [Gupta
and Wong, 2007].
Microwaves are also generated from electricity, requiring no hydrocarbons or other
fuel to be burned to generate heat. This allows microwave processing facilities to be
built in areas where renewable power sources such as wind, solar, or hydroelectric
3.4. MICROWAVE ADVANTAGES AND LIMITATIONS
56
are possible. The absence of fuel allows the microwaves to be a cleaner alternative
heating source.
Limitations to microwave heating are also present, and these include temperature
control, thermal runaway, hotspots, and the ability to scale up. It is difficult to
moderate the temperature inside a microwave reaction. Due to the fact that the
microwave heating profile generates and expanding core model, it can be very difficult
to get direct readings of the reaction temperature. Once the surface of the sample
gets hot the inside of the sample could be hundreds of degrees hotter. Microwaves can
also cause a phenomenon known as thermal runaway, where a sample experiences a
sudden change in dielectric constants which can lead to greater microwave absorption.
Temperatures can increase extremely quickly and can lead to undesired reactions
or damage to the sample holder or microwave apparatus. Plasma generation from
certain systems is also possible. Plasmas are excellent absorbers of microwaves and
will absorb any microwaves which should have been absorbed by the sample. In
addition to this, any system not expecting plasma generation may be damaged by the
high heat generated. Hotspots can be created in materials improperly mixed. This
leads to uneven heating which can lead to poor processing results. Microwave tests
done typically involve batch testing of a small sample in a controlled atmosphere.
In order for microwave heating to be used industrially, scale up and the ability to
convert a batch process to a continuous process are essential [Chandrasekaran et al.,
2012, Gupta and Wong, 2007].
57
Chapter 4
Experimental Methods and Material Used
4.1
Ore Preparation
The chromite sample used for testing was provided by Noront Resources from the
Black Thor deposit from the Northern Ontario Ring of Fire. One rock sample weighing 16.4 kg was delivered. Being too big to directly fit into a jaw crusher, the sample
was first hammered to remove some pieces followed by wet circular saw cutting. All
pieces were passed through the jaw crusher, coming out approximately gravel sized.
10 kg of the gravel sample was put aside, and the remaining 6.4 kg were crushed
down and pulverized to 80% passing 100 µm as seen in Figure 4.1.
4.2. MATERIAL CLASSIFICATION
58
Figure 4.1: Size distribution of pulverized chromite ore.
4.2
Material Classification
4.2.1
SEM Analysis
Both as-received and pulverized ore were analyzed using scanning electron microscopy.
4.2.1.1
As-Received Ore
Small pieces of crushed chromite ore (gravel, 12 mm size) were mounted in epoxy and
polished using progressively finer sandpaper, diamond wheels, and finally colloidal
silica, as described in the metallography section (Section 4.5.9). The 25 mm diameter
epoxy samples were carbon coated and put under a high vacuum (2.5e-5 torr) for
analysis. Figure 4.2 shows the back-scatter electron image and elemental mapping
4.2. MATERIAL CLASSIFICATION
results for Cr, Fe, Mg, Si.
59
4.2. MATERIAL CLASSIFICATION
BSE
Cr
Fe
Mg
Si
Figure 4.2: BSE image and elemental mapping of as-received chromite ore.
60
4.2. MATERIAL CLASSIFICATION
61
It can be seen from the elemental mapping that the chromium and iron occur in the same grains within the ore. This corresponds to the mineral chromite
((Mg,Fe)Cr2 O4 ) as found in XRD analysis (Section 4.2.2). The darker areas in the
BSE image, where Mg and Si are found in the elemental map, correspond to the
mineral clinochlore ((Mg,Fe)5 Al(Si3 Al)O10 (OH)8 ).
From the images it can be seen that the majority of the chromite grains are over
400 µm in size. Certain grains appear fragmented due to the crushing required to
get the 12 mm particle size for polishing.
4.2.1.2
Size Analysis
Each size fraction was analyzed using SEM analysis to ensure sufficient mineral liberation for processing. Figures 4.3, 4.4, and 4.5 shows the BSE images and elemental
mapping results for three different size fractions. Additional SEM and elemental
mapping images can be found in Appendix B. In the +105 µm fraction it can be
seen that the majority of the chromite particles have sufficient liberation, though a
few still have a small amount of clinochlore attached.
4.2. MATERIAL CLASSIFICATION
BSE
Cr
Fe
Mg
Si
Figure 4.3: BSE image and elemental mapping of the +105 µm size fraction.
62
4.2. MATERIAL CLASSIFICATION
63
BSE
Cr
Fe
Mg
Si
Figure 4.4: BSE image and elemental mapping of the +44 -53 µm size fraction.
4.2. MATERIAL CLASSIFICATION
BSE
Cr
Fe
Mg
Si
Figure 4.5: BSE image and elemental mapping of the -20 µm size fraction.
64
4.2. MATERIAL CLASSIFICATION
4.2.2
65
XRD
Sample purity and initial analysis determinations were conducted using x-ray diffraction (Section 4.5.3), and the results are shown in Figure 4.6.
Figure 4.6: XRD spectrum of pulverized chromite ore.
It can be seen that the two primary species in the ore are chromite ((Mg,Fe)Cr2 O4 )
and clinochlore ((Mg,Fe)5 Al(Si3 Al)O10 (OH)8 ). The chromite species were shown as
the brighter phases in the SEM images previously seen, while the clinochlore corresponds to the darker matrix phase.
4.2.3
Assay
Chromite ore assaying was performed by SGS Lakefield. Whole rock analysis was
performed using the borate fused disc and XRF methods. Elemental analysis was
4.2. MATERIAL CLASSIFICATION
66
done using one of three methods, 30 g Fire Assay followed by ICP-AES (GE FAI313),
pyrosulphate fusion followed by XRF (GO XRF77B), and by internal standard XRF
(GO XRF75F). The following two tables, 4.1 and 4.2, show the assay results. Values
of 0 indicate that the value was 0 or below detection range.
Table 4.1: Whole rock analysis results from SGS Lakefield.
Species
Assay (%)
SiO2
7.04
Al2 O3
13.3
Fe2 O3
19.8
MgO
14.5
CaO
0.06
K2 O
0.02
Na2 O
0.03
TiO2
0.36
MnO
0.19
P2 O5
0
Cr2 O3
43.0
V2 O5
0.15
LOI
1.55
Total
100
4.3. EXPERIMENTAL EQUIPMENT
67
Table 4.2: Elemental assay results from SGS Lakefield.
4.3
Element
Assay (ppb or %)
Method
Au
12 ppb
GE FAI313
Pt
232 ppb
GE FAI313
Pd
138 ppb
GE FAI313
Co
0.02 %
GO XRF77B
Cu
0
GO XRF77B
Ni
0.13 %
GO XRF77B
Pb
0
GO XRF77B
As
0
GO XRF75F
Experimental Equipment
This section describes the experimental equipment used for sample processing.
4.3.1
Sample Preparation
Samples were prepared from pulverized as-received ore. A bulk 2 kg sample was
prepared by mixing pulverized ore with 15% activated charcoal. The mixed sample
was then placed onto roll mixers to ensure proper mixing. Prior to reduction testing,
samples were dehydrated for 24 h in drying ovens to remove as much moisture as
possible. 30 g of sample were placed into the bottom of a quartz crucible. The sample
mass was chosen based on previous microwave reduction tests on different materials
[Forster, 2015]. Quartz was selected due to its high melting point and microwave
4.3. EXPERIMENTAL EQUIPMENT
68
transparency.
Tests which did not use 15% carbon were prepared separately. The corresponding
amounts of ore and charcoal were added to weighboats and mechanically mixed. Once
proper mixing was assured the samples were placed into quartz crucibles.
For air reduction testing the loaded crucibles were simply placed into the applicator on top of a heat resistant, microwave transparent material as seen in Figure
4.7.
Figure 4.7: Sample setup for air atmosphere reduction tests.
Tests performed in argon had a setup shown in Figure 4.8. Argon was fed through
the rear of the applicator through a port and connected to a Teflon base. A glass
bell jar with a tube coming off of the top was then placed on the Teflon base. Argon
gas flowed through a hole in the bottom of the base into the enclosed space created
by the bell jar. The tube at the top of the jar allowed for gases to escape the system
and allowed for purging the system with argon prior to each test.
4.3. EXPERIMENTAL EQUIPMENT
69
Figure 4.8: Sample setup for argon atmosphere reduction tests.
The setup for a vacuum system was similar to that for argon atmosphere reduction
tests, however the glass bell jar did not have a tube at the top. A diagram can be
seen in Figure 4.9. Loaded crucibles were prepared similar to air and argon testing.
A 25 mm pressed alumina briquette was placed on top of the sample in an attempt
to minimize sample loss from the crucible. Glass wool was finally packed into the
remaining crucible space and secured with glass string to hold the alumina and
sample in place.
4.3. EXPERIMENTAL EQUIPMENT
70
Figure 4.9: Sample setup for vacuum reduction tests.
4.3.2
Microwave Setup
The microwave and its necessary components were purchased from Gerling Applied
Engineering Inc. Each component had similar waveguide connections allowing for
ease of assembly. The parts and their features are outlined in the following paragraphs.
The magnetron used is a model GA4248 as shown in Figure 4.10. It has a
maximum output power of 1200 W which could be continuously varied in intensity.
A control knob on the power supply was used to control power.
4.3. EXPERIMENTAL EQUIPMENT
71
Figure 4.10: 1.2 kW variable magnetron used for the current research [GAE, 2015].
The magnetron was connected to a GA1121 3-port circulator, shown in Figure
4.11. This allowed microwaves to travel from the magnetron into the applicator, while
redirecting any reflected microwaves towards a dummy load. This was essential in
order to avoid magnetron overheating or malfunction.
Figure 4.11: 3-port circulator used to control microwave propagation [GAE, 2015].
The reflected microwaves that were redirected by the circulator were sent into the
dummy load, GA1210, as shown in Figure 4.12. This had a water load attached which
absorbed and removed reflected microwaves from the system. This was connected to
the interlock system which did not allow the magnetron to function without sufficient
water flow.
4.3. EXPERIMENTAL EQUIPMENT
72
Figure 4.12: Dummy load used to dissipate reflected microwaves [GAE, 2015].
The other component that was attached to the circulator was a GA3002 dual
power monitor, shown in Figure 4.13. This device used crystals to measure the
forward and reflected microwaves and output a DC signal. The detectors were connected to an analog meter to see power levels during testing and to a digital multimeter which recorded the voltage to a text file which was then analyzed in Microsoft
Excel.
Figure 4.13: Dual power monitor that allowed for the measurement of absorbed
power [GAE, 2015].
A GA1003 3-stub tuner was located after the power monitor, as shown in Figure
4.14. This device could be used to match load impedance for high power systems
[GAE, 2015]. It was not used during the reduction testing.
4.3. EXPERIMENTAL EQUIPMENT
73
Figure 4.14: 3-stub impedance tuner located after the power meters and before the
applicator [GAE, 2015].
Finally, the applicator was located after the tuner. The applicator was a GA6201,
as shown in Figure 4.15. It features a mode stirrer, described in detail in Section
4.3.2.1. The rear featured 2 Swagelok tube fittings and a Swagelok VCO feed thru
port. These three ports allowed for argon and vacuum atmosphere testing. Access
to the inside of the applicator was through a door on the front. Interlocks in the
door ensured that the magnetron would not run with the door open.
Figure 4.15: Microwave applicator where samples were processed [GAE, 2015].
Figure 4.16 shows an overhead schematic of the microwave system that was used.
The samples used are those shown in schematic views in Figures 4.7 through 4.9.
4.3. EXPERIMENTAL EQUIPMENT
74
Figure 4.16: Overhead view of the connected microwave system.
4.3.2.1
Mode Stirrer
The applicator that was used featured an aluminium mode stirrer. This device was
used to ensure proper sample heating. It featured four rotating blades, that when
spinning, changed the cavity geometry and the standing wave pattern. This created a
dynamic homogeneous field which improves sample heating. Plaza-González et al. in
2005 studied the effects of mode stirrer configuration on sample heating. They found
that the use of a mode stirrer helped to form uniform electric fields within a lowpermittivity sample [Plaza-González et al., 2005]. Increased field uniformity allows
4.3. EXPERIMENTAL EQUIPMENT
75
for the negation of hotspot development in samples and allows for faster heating.
4.3.2.2
Vacuum
The vacuum pump utilized was a General Electric rotary vane pump (No. 5KC37PG433X)
connected to a 250V GE motor. This pump was connected to a vacuum gauge before
entering the applicator through a Swagelok VCO feed-thru port on the back of the
applicator via a polyethylene tube. This polyethylene tube was connected to the
vacuum chamber base by a polypropylene compression fitting. By using this vacuum
system, pressures between 0.1 and 0.05 bar were achieved.
4.3.3
Microwave Heating Tests
Initial heating tests were preformed on pulverized as-received chromite ore. Two
sets of heating tests were conducted, one with the use of the mode stirrer in the
applicator and one without. Three tests of increasing duration were conducted for
each set at times of 300, 600, and 900 seconds. The input energy for each test was
800 W. Samples for these heating tests had masses of 30 g and were not placed in
dewatering furnaces prior to testing.
For both sets of testing an argon atmosphere was used. Immediately after each
test concluded the loaded crucible was removed from the argon vessel and its temperature measured with a thermocouple.
4.4. SAFETY
4.4
76
Safety
Several safety measures were taken to ensure that no harm came to any person or
equipment while conducting experiments. A labcoat, safety glasses, and nitrile gloves
were worn at all times while preparing samples. Heat resistant gloves were worn when
handling samples immediately after processing or to remove a tray of samples from
the drying ovens. Hot crucibles were removed from the microwave using crucible
tongs. A handheld carbon monoxide probe was also used during any microwave runs
as the samples could give off carbon monoxide as a by-product. A microwave leakage
detector was used periodically to ensure that no microwaves leaked from the unit.
When dealing with any fine particles, a 3M N95 respirator was worn. Safety showers
and eye wash stations were located and clearly marked in all areas where experiments
were conducted.
4.5
Analysis Methods and Equipment
Several different analysis methods and types of equipment were utilized and are
described in this section.
4.5.1
Thermogravimetric and Differential Thermal Analysis
(TGA/DTA)
A Jupiter STA 449 F3 was used for thermogravimetric and differential thermal analysis. This instrument is capable of simultaneously measuring the thermogravimetric
and differential thermal signals in a variety of atmospheres and over a range of heating rates [NETZSCH, 2012].
4.5. ANALYSIS METHODS AND EQUIPMENT
77
The furnace used was a silicon carbide furnace capable of heating up to 1550
◦
C, with heating rates between 0.001 ◦ C/min to 50 ◦ C/min. Nitrogen was used
as a protective gas at a flow rate of 20 mL/min. Other gases such as oxygen and
argon could be introduced through the two purge gas ports at various flow rates
[NETZSCH, 2012]. Argon gas was used for the analysis of chromite ore, while an
oxygen/nitrogen atmosphere was used for the proximate analysis of charcoal.
A balance system at the base of the instrument recorded the mass changes over
the duration of the tests. The sensitivity of the balance was 0.1 µg. Differential
thermal analysis was conducted using an inert, empty reference crucible next to
the crucible containing a sample. A thermocouple in the sample carrier measured
the differences in temperature between the inert reference and sample crucible and
recorded the difference as µV/mg [NETZSCH, 2012].
4.5.1.1
Proximate Analysis of Charcoal
In order to determine the composition of the activated charcoal reducing agent, a
proximate analysis technique using TGA was used. Proximate analysis of coal and
coke is defined by ASTM International in designation D3172-13 as “an assay of the
moisture, ash and volatile matter as determine by prescribed methods and the calculation of fixed carbon by difference” [ASTM, 2013]. ASTM Standard D7582-15
defines standard test methods for proximate analysis using macro thermogravimetry
[ASTM, 2015]. This method is not applicable to the TGA unit available, so a modified version applicable to micro thermogravimetry was used. This method, similar
to that used by Donahue and Rais (2009) to teach proximate analysis on coal using
4.5. ANALYSIS METHODS AND EQUIPMENT
78
thermogravimetry, does not require the sample to be covered midway through testing
or cooled down as the ASTM D7582-15 method does [Donahue and Rais, 2009]. In
addition it allows for smaller sample sizes (30 mg) compared to the 1 gram required
by the ASTM method.
Samples were first heated in a nitrogen atmosphere to 110 ◦ C, and held for 10
minutes for any moisture to be removed. Samples were then heated from 110 to 950
◦
C at a rate of 25 ◦ C/min, at this point any volatile matter was removed from the
sample. The sample was held isothermally for 15 minutes before a 50% oxygen atmosphere was produced. This removes the fixed carbon in the sample, the remaining
mass is ash.
Proximate analysis using thermogravimetric analysis are shown in the results
of Table 4.3. Six tests were conducted using identical heating profiles. The data
obtained from the proximate analysis was used in the thermodynamic model as
various input species.
4.5. ANALYSIS METHODS AND EQUIPMENT
79
Table 4.3: Composition of activated charcoal determined using a modified proximate
analysis method on a Jupiter STA 449 TGA.
Test
Moisture (%)
Volatiles (%)
Fixed Carbon (%)
Ash (%)
1
3.42
5.32
86.27
4.98
2
3.87
4.60
–
–
3
2.83
5.95
87.98
3.24
4
3.94
5.24
87.29
3.52
5
3.42
5.56
86.29
4.72
6
3.57
5.11
88.07
3.25
Avg.
3.51
5.30
87.18
3.94
Std. Dev.
0.37
0.41
0.78
0.75
4.5.2
Carbon-Sulphur Analysis
An Eltra CS-2000 Carbon Sulphur Determinator was used to analyze the as-received
ore and the reduction products for their carbon and sulphur contents. Samples are
loaded into alumina crucibles 25 mm in diameter. Iron and tungsten chips are added
as accelerators. The loaded crucible is placed in the induction furnace which heats
samples to over 2000 ◦ C [ELTRA, 2016]. Off gases are then carried through the
system to the infra-red detectors. Four of these detectors are present in the system,
two for carbon and two for sulphur, and each element has both a high and low range
detector. The determinator has an accuracy of 0.1 ppm for both carbon and sulphur
for samples of 500 mg size. The accuracy of both carbon and sulphur for all detectors
is ±0.5% of the element present [ELTRA, 2010].
4.5. ANALYSIS METHODS AND EQUIPMENT
4.5.3
80
X-Ray Diffraction (XRD)
Philips X’Pert Pro MPD (Multi Purpose Diffractometer) with an X’celerator detector
was used for initial sample and product characterization. Samples were crushed into
a powder form and placed on a metal disk. Methanol was used to spread the sample
evenly across the surface of the metal plate. Samples were then placed inside the
analysis chamber and exposed to x-rays emitted from an x-ray tube. The x-ray
tube slowly rotates around the sample, changing the angle, θ, as it rotates. As the
angle changes the x-rays diffract within the sample. At many angles the x-rays will
destructively interfere with each other and few will make it to the detector, however
at certain angles the x-rays will constructively interfere and a large amount will make
it to the detector. Bragg’s Law, illustrated in Figure 4.17 and described in Equation
4.1 can be used to determine the crystal structure of the sample. Databases are
constructed with thousands of minerals tested which can be compared against the
sample to determine its composition.
nλ = 2d sin θ
(4.1)
4.5. ANALYSIS METHODS AND EQUIPMENT
81
Figure 4.17: A visual representation of Bragg’s Law [Hadjiantonis, 2013].
4.5.4
X-Ray Fluorescence (XRF)
An Oxford Instruments X-Supreme 8000 X-Ray Fluorescence machine was used for
material and product analysis. Samples are loaded into 25 mm sample holders with a
transparent film placed over the bottom. Samples were typically powdered however
larger rocks and even liquids can be analyzed. These loaded sample holders are then
placed into the analysis machine. Air or helium gas is present between the x-ray
emitter, sample, and detector. Helium is used for analysis of lighter elements (magnesium and lighter). The sample is then exposed to x-rays (hν), and will generate
secondary x-rays (hν’) as electrons in the sample change energy levels, as seen in
Figure 4.18. The secondary electron generated has a characteristic frequency that
is unique to that element. As a result, the secondary electrons can be counted and
the elements present identified. In addition, if a known concentration of sample is
analyzed then the count of secondary x-rays can be quantified. This allows for the
use of XRF as a fast way to assay samples.
4.5. ANALYSIS METHODS AND EQUIPMENT
82
Figure 4.18: X-ray fluorescence schematic [Calvero., 2016].
The electron microscope and microprobe that are discussed in the following sections have XRF detectors equipped to them.
4.5.5
Environmental Scanning Electron Microscope (ESEM)
A FEI Quanta 650 FEG Environmental SEM was used for detailed imaging and
elemental mapping. Polished samples, mounted in epoxy, were placed into 25 mm
sample holders and then placed on a sample tray. This tray could hold up to 14
samples at a time. The tray was then loaded into the electron microscope chamber.
The chamber was then closed and pressurized. The model used was an ESEM,
meaning that water vapour could be used as the gas present in the sample chamber
and lower pressures were able to be used. Samples were analyzed at 50 Pa of pressure
with a water vapour atmosphere. The water permitted for electrons to be carried
away from the sample into the grounded chamber. Lower pressures, down to 6e−4
Pa, were used for samples which were carbon coated [FEI, 2011].
Elemental mapping and imaging was done by the ESEM. Imaging was done using
4.5. ANALYSIS METHODS AND EQUIPMENT
83
a backscatter electron detector. These electrons are generated in the field emission
gun and travel to the sample where they then interact with the sample and scatter
out. They are picked up by the backscatter detector and used to create an image.
Brighter phases in the image represent higher atomic weight species. Elemental
mapping was also done using XRF detectors present in the chamber. These two
detectors detect secondary x-rays and could be used to determine grades and map
where certain elements were present in a sample [FEI, 2011].
4.5.5.1
Mineral Liberation Analysis (MLA)
Mineral Liberation Analysis could be performed in conjunction with ESEM analysis. The microscope would take a composite picture of the sample and identify any
particles present based on input criteria. The microscope would then take an x-ray
measurement of each particle identified. Particles could be grouped and grades assigned to each of those groups. From this, the grade of a sample or section of sample
could be determined.
4.5.6
Electron Microprobe (EMP)
A JEOL JXA-8230 Electron Microprobe was used for quantitative analysis of individual grains of the processed samples. The microprobe worked on a similar principle
to that of the ESEM. X-ray detectors on the microprobe are much more sensitive
and can therefore be tuned to look for a specific element as opposed to the entire
range seen by the ESEM. Analyzing crystals are present in the path of the x-ray
which are tuned for a specific x-ray range. This allows for selectivity and precision
4.5. ANALYSIS METHODS AND EQUIPMENT
84
in samples. As with the ESEM, an electron beam is generated in an electron gun.
This passes through electromagnetic lenses before interacting with the sample. Upon
interaction, the sample gives off x-rays which are then analyzed. Standards are typically run before each test to ensure correct x-ray counts. Samples are run under
high vacuum and must be carbon coated prior to analysis.
4.5.7
Cavity Perturbation Technique
The dielectric properties of the chromite ore and chromite with a carbon addition
were investigated by Microwave Properties North using the cavity perturbation technique. Hutcheon et al. (1992) developed the technique. It involves placing a heated
sample into a cooled tube whose resonant frequency and Q factor (quality factor) are
known. A schematic diagram of the apparatus used can be seen in Figure 4.19. The
introduction of this sample changes the frequency and Q factor, which both relate to
the permittivity () and permeability (µ) of the sample. Samples can be measured
up to a temperature of 1400 ◦ C with frequencies from 50 to 2450 MHz [Hutcheon
et al., 1992].
4.5. ANALYSIS METHODS AND EQUIPMENT
85
Figure 4.19: Schematic diagram of the apparatus used in the cavity perturbation
technique [Hutcheon et al., 1992].
4.5.8
Particle Size Analysis
Particle size analysis was carried out using a Fritsch Analysette 3 Pro sieve shaker.
A series of screens ranging from 20 mesh (841 µm) to 400 mesh (37 µm) were used.
The shaker had the capability of automatic amplitude control and had a built-in
timer.
4.5.9
Metallography
Product characterization was done using metallographic techniques. A processed
sample was either pulverized in a ring pulverizer or placed whole into a 25 mm
4.6. VARIABLES INVESTIGATED
86
plastic mould. West Systems 105 Epoxy Resin and 206 Slow Hardener were mixed
together in specified amounts and poured into the moulds. The slow hardener was
selected to allow ample time for sample re-positioning prior to epoxy hardening.
Once the epoxy hardened it was removed from the mould and was then polished
using progressively finer grits of sandpaper. The final polishing was done using a
6 micron diamond wheel, and if further polishing was required 0.5 micron colloidal
silica was used. Optical microscopy was conducted on the samples to ensure that
the polishing done was adequate and no significant polishing defects were present in
the samples.
4.6
Variables Investigated
Table 4.4 shows the variables that were the subject of investigation in the current
work.
Table 4.4: All experimental variables.
Variable
Values
Carbon Level
0, 2.5, 5, 7.5, 10, 15%
Time
150, 200, 300, 400, 600, 900 seconds
Power Level
400, 800, 1200 watts
Atmosphere
Argon, Air, Vacuum (5 kPa)
Tests were conduced with the following conditions for all three atmosphere types:
400 and 800 W input power tests with 150, 300, and 600 second times, and 1200
W tests with 200 and 400 second times. The reasoning for these times and input
4.6. VARIABLES INVESTIGATED
87
powers was to study the effects of energy input rate. Several tests have identical
total ideal input energies, meaning that the only variables affecting the reduction
are atmospheric type, time, and power level. Table 4.5 shows the time, input power,
and total energy used for each test. Energy, in joules, is calculated according to
Equation 4.2. This value was calculated for each test from the absorbed power
measurements, taken once per second. The area under any absorbed power graph
gives the total energy absorbed by the system.
E=P×s
(4.2)
where P is the input power in watts and s is the reaction time in seconds.
Table 4.5: Time, input power, and total energy for the reduction tests.
Time (sec)
Input Power (W)
Total Energy (kJ)
150
400
60
300
400
120
600
400
240
150
800
120
300
800
240
600
800
480
200
1200
240
400
1200
480
From the table it can be seen that several tests have the same total input energy.
Any differences in results are therefore a result of the input power level and time.
4.7. RECOVERY CALCULATIONS
4.7
88
Recovery Calculations
Two recovery values for both chromium and iron were calculated for each ferrochrome
producing test. Replacing chromium with iron in the following equations allows one
to determine the iron recoveries.
Overall Recovery, RO This is the recovery of chromium or iron into the ferroalloy
for the total chromium or iron present in the initial sample. The overall recovery
can be seen in Equation 4.3.
RO
Cr =
CrAlloy
CrT otal
(4.3)
where CrAlloy is the chromium present in the ferroalloy and CrT otal is the total
amount of chromium present in the initial ore sample. CrAlloy can be calculated
using Equation 4.4.
CrAlloy = (mcore · XAlloy ) · (%CrAlloy )
(4.4)
where mcore is the mass of the reduced core, XAlloy is the fraction of the reduced
core which is ferroalloy, and (%Cr)Alloy is the chromium grade of the alloy. The
reduced core fraction, X, can be calculated using Equation 4.5.
XAlloy =
AAlloy
Atotal
core
(4.5)
where A is the area of the alloy phase, or the total core area which is the sum
4.7. RECOVERY CALCULATIONS
89
of alloy and oxide areas. This was calculated using image analysis techniques
using the SEM. Using the magnification scale, a relationship of µm/pixel is
determined and then the areas of each particle are calculated using the amount
of pixels shown on the image.
Crtotal can be calculated using Equation 4.6
Crtotal = msample · (%Cr)ore · 1 − (%C)initial
(4.6)
where msample is the initial mass of the sample, (%Cr)ore is the chromium grade
of the ore and (%C)initial is the initial carbon addition.
Core Recovery, RC Also known as metallization. This is the recovery of chromium
or iron into the ferroalloy assuming only the iron or chromium found within
the reduced core of the system. Equation 4.7 defines this recovery.
RC
Cr =
CrAlloy
CrCore
(4.7)
CrAlloy has previously been defined in Equation 4.4. Crcore can be calculated
using Equation 4.8.
CrCore = mcore ·
X
(%Cr)phase · Xphase
(4.8)
where (%Cr)phase is the chromium grade of the phase and Xphase is the area
fraction of that phase. There are typically three phases present in the reduced
4.8. ERROR ANALYSIS
90
core: the ferroalloy and two reduced oxides.
4.8
Error Analysis
Duplicate reduction tests and analyses were performed in order to determine errors.
Assays performed by SGS Lakefield were repeated and found a maximum variance
of ±0.45%. This variance can be attributed to variability in the mineralogy.
Thermogravimetric analysis performed on the chromite ore showed variances in
mass loss percentage of ±0.5%. This is similar to variances done during previous
TGA work [Elliot, 2015]. Proximate analysis of coal using TGA had a maximum
variance of ±0.90%. This is higher than previously seen variance however it can be
explained by the volatile nature of the substances analyzed.
Repeat reduction testing found a maximum variance of ±3.5% in total absorbed
power. Reduced core masses had a variance of ±0.78 g. Total mass losses had
variances of ±0.92%. Ferrochromium grades were found to vary by ±0.45%, while
recoveries could vary by ±6.42%.
91
Chapter 5
Thermodynamics
5.1
Thermodynamic Considerations
A thermodynamic model was developed using the Equilibrium and Gibbs Solver
modules of HSC Chemistry 6.1 [Roine, 2006]. The model was constructed to determine the amount of reducing material necessary and final grades and recoveries of
both chromium and iron. The carbon content of the ferroalloy could also be determined. The model was run at atmospheric and reduced pressures to determine the
effects of pressure. Selected species were entered into the equilibrium module along
with various activity coefficients found in literature. A Gibbs free energy minimization technique was then used to determine the equilibrium concentrations at various
temperatures and pressures. The Gibbs free energy of a system is given by Equation
5.1 [Koukkari, 2014].
G=
XX
α
nαk µαk
(5.1)
k
where G is the Gibbs free energy of the system, α is the phase in question, k is
5.2. MODEL LIMITATIONS
92
the species, n is the amount, and µ is the constant mole fraction. Species should be
stable over the temperature and pressure range. A minimum Gibbs free energy can
be determined using computer methods, such as those present in the Gibbs Solver
module of HSC [Roine, 2006, Koukkari, 2014].
5.2
Model Limitations
The thermodynamic model was constructed prior to reduction testing to determine
ideal values for species in the system. However, this model has some limitations,
namely with regards to kinetics and the vacuum system. First, this model does
not take kinetics into account, only the minimization of Gibbs free energy. Kinetic
dependent reactions will not be appropriately modeled and as such their products
will not be accurately represented during actual reduction. Second, the vacuum
system used for a portion of the reduction tests removes gaseous species from the
system, something which this model does not do. Products such as carbon monoxide
are almost immediately removed from the system, which will affect the equilibrium
of the system according to the Le Châtelier’s Principle.
5.3
Species and Phases Considered
The species and phases in Table 5.1 were considered for the thermodynamic model.
Twenty-seven species were selected from a list of 306 potential candidates from elements seen in the assays on the sample, as shown in Table 4.1. Species were considered from the following elements: H, C, O, Mg, Al, Si, Cr, and Fe. Viable species
were determined on their stability across the selected temperature range, 200 to 1400
5.3. SPECIES AND PHASES CONSIDERED
◦
93
C. Two species, FeO*OH and Fe3 C, had to have their data extrapolated to work
over the entire range of the model, as data for both species was only reported to
1227 ◦ C, 173 ◦ C below that of the maximum temperature used by the model.
Input for the system was 100 kg of ore whose contents were based on both the
assay of the ore, and proximate analysis (Section 4.5.1.1) done on the charcoal. An
Excel spreadsheet was created to calculate compositions of input materials based on
specified carbon input.
Phases were determined primarily based on the group that the species belonged
to. The three phases used were gases, oxides, and alloy. For display purposes the
spinel oxides are listed in a separate column in Table 5.1, however in the model
both the oxides and spinels were grouped under oxides. The alloy phase contains all
species which are present in a ferrochrome solid solution.
Table 5.1: Species and phases considered in the thermodynamic model, as entered
into the HSC Equilibrium Module.
Gases
Oxides
Spinels (Oxides)
Alloy
Carbon
CH4
Al2 O3
Cr2 FeO4
Cr
C
CO
Cr2 O3
Cr2 MgO4
Cr23 C6
CO2
Fe2 O3
Fe3 O4
Cr3 C2
H2
FeO
Mg2 SiO4
Cr7 C3
H2 O
FeO*OH
MgFe2 O4
Fe
O2
MgO
MgO*Al2 O3
Fe3 C
MgSiO3
SiO2
5.4. ACTIVITY COEFFICIENTS
5.4
94
Activity Coefficients
This section details the activity coefficients used for the thermodynamic model. Most
of these activity coefficients were reported at one temperature, and, as such the
regular solution model, expressed by Equation 5.2, was used to obtain values for
activity coefficients at other temperatures.
γ1 =
γ2 T2
T1
(5.2)
The activity coefficients for both chromium and iron were adapted from Mazandarany and Pehlke (1973) [Mazandarany and Pehlke, 1973]. Chromium activities
were reported for various concentrations and at temperatures of 900, 1000, 1100,
and 1200 ◦ C. Each data set was plotted and equations were generated for the linear
portions. The slope and intercept data for each equation was then plotted against
its respective temperature and equations generated from that data. Those equations
were then used to determine a generalized equation for the activity coefficient for
chromium, which can be seen in Equation 5.3.
lnγCr = (−0.0013 · T + 2.3851) + (0.0014 · T − 2.4865) · XCr
(5.3)
To determine the activity coefficient for iron the Gibbs-Duhem equation was used
as seen in Equation 5.4.
Z
Z
d lnγF e = −
XCr
· d lnγCr
XF e
(5.4)
5.4. ACTIVITY COEFFICIENTS
95
The Gibbs-Duhem equation was then graphically evaluated using the trapezoidal
rule to obtain the equation for the activity coefficient of iron found in Equation
5.5. Iron is the only species which required the use of the Gibbs-Duhem equation to
determine its activity coefficient.
lnγF e =(−1 · 10−6 · T 2 + 0.0035 · T − 3.0) · XF2 e
+ (1 · 10−7 · T 2 − 0.0004 · T + 0.2344) · XF e
(5.5)
+ (8 · 10−7 · T 2 − 0.003 · T + 2.7866)
Five activity coefficients were reported at a single temperature, Equations 5.6
through 5.10, and as such the regular solution model (Equation 5.2) was used. Each
journal article reported the measured activity values at various concentrations. The
activity coefficients were then calculated using γ = a/x. The natural logarithm
of each activity coefficient value was then taken and plotted against concentration.
Sigmaplot 12 was then used to determine equations for activity using a built-in
nonlinear regression tool.
The activities for Fe3 O4 and FeCr2 O4 were adapted from Petric and Jacob (1982).
Values were reported at 1673 K [Petric and Jacob, 1982]. The equations for both
activity coefficients are as follow:
lnγF e3 O4 =
−0.803 · XF e3 O4
−3.208 · XF e3 O4 1673
+
·
−0.075 + XF e3 O4 −4.724 + XF e3 O4
T
lnγCr2 F eO4 =
−0.8396 · XF e3 O4
1673
·
1 + (−0.7192 · XF e3 O4 )
T
(5.6)
(5.7)
5.5. ATMOSPHERIC PRESSURE MODEL
96
The activity of MgFe2 O4 was adapted by Katayama and Iseda (2002). Values were
reported as 1273 K, and the activity coefficient is shown in Equation 5.8 [Katayama
and Iseda, 2002].
lnγM gF e2 O4 =
1.0319 + (−0.9159 · XM gF e2 O4 ) 1273
·
1 + (4.6090 · XM gF e2 O4 )
T
(5.8)
Activities for MgCr2 O4 and MgAl2 O4 were adapted from Jacob and Behera
(2000). Values were reported at 1473 K, and the equations for the activity coefficients are as follow [Jacob and Behera, 2000]:
lnγCr2 M gO4 = (−0.0869 + 1.2356 · exp(−3.4652 · XCr2 M gO4 )) ·
1473
T
2
lnγM gO∗Al2 O3 =(0.0161 − 0.5747 · XCr2 M gO4 + 3.8823 · XCr
2 M gO4
− 3.1114 ·
5.5
3
XCr
)
2 M gO4
1473
·
T
(5.9)
(5.10)
Atmospheric Pressure Model
The model was initially run at atmospheric pressure, 1 bar, at temperatures between
200 and 1400 ◦ C, with input carbon levels of 2.5, 5, 7.5, 10, 15, and 20%.
The chromium species distribution as a function of temperature for an initial carbon input of 15% can be seen in Figure 5.1. The initial species containing chromium
are iron and magnesium chromite. As the reaction temperature increases these
spinels break apart, forming Feo, MgO, and Cr2 O3 . This is seen in an increase
of Cr2 O3 over the temperature range, significantly increasing after 800 ◦ C when the
5.5. ATMOSPHERIC PRESSURE MODEL
97
iron chromite breaks apart. This chromium oxide is the primary chromium species
until 1100 ◦ C, when reduction begins to occur producing chromium carbide and
metallic chromium. It can be seen that at 1100 ◦ C the primary chromium species
switches from Cr2 O3 to Cr3 C2 as carbothermic reduction begins. As the temperature
increases, this carbide species reacts with the remaining oxides to decarburize into
Cr7 C3 , Cr2 3C6 , or Cr. Cr2 3C6 exists in the model in very small quantities.
Figure 5.1: Chromium species distribution between 200 and 1400 ◦ C for a 15% carbon addition.
The iron species distribution as a function of temperature for an initial carbon
input of 15% can be found in Figure 5.2. The only input species that contains iron is
iron chromite. Between 200 and 800 ◦ C some Cr2 FeO4 decomposes to form FeO and
5.5. ATMOSPHERIC PRESSURE MODEL
98
Cr2 O3 . After 800 ◦ C reduction of the Cr2 FeO4 and FeO begins, creating primarily
metallic Fe and Fe3 C. The Fe3 C amount increases between 800 and 1100 ◦ C until
decreasing as it is decarburized by reacting with oxide species. The Fe3 O4 amount
is extremely low over the entire temperature range.
Figure 5.2: Iron species distribution between 200 and 1400 ◦ C for a 15% carbon
addition.
Figure 5.3 shows the chromium grade of the ferrochrome as a function of temperature for various initial carbon inputs. It can be seen that for all initial input carbon
grades no chromium is present until approximately 1100 ◦ C. Increasing the input
carbon leads to a higher chromium grade, reaching a maximum at 15% input carbon
of 62% chromium. Increasing the carbon to 20% leads to a lower chromium grade,
5.5. ATMOSPHERIC PRESSURE MODEL
99
61.3%. This is due to a higher amount of carbide species present in the ferrochrome,
increasing the carbon grade and decreasing the iron and chromium grades.
Figure 5.3: Chromium grade of the ferrochrome alloy as a function of temperature
for various initial carbon additions.
The chromium recovery can be seen in Figure 5.4. As with the grade seen in
the previous plot, the recovery increases as additional carbon is added. A maximum
recovery of 100% chromium is obtained with 20% carbon input. A carbon input of
15% leads to a recovery of 96.3% chromium. It can be seen that the increase in
recovery is approximately 18% chromium for every 2.5% input of carbon.
5.5. ATMOSPHERIC PRESSURE MODEL
100
Figure 5.4: Chromium recovery as a function of temperature for various initial carbon additions.
The iron grade as a function of temperature for various initial carbon inputs can
be seen in Figure 5.5. Iron is reduced at a lower temperature than chromium and
as such is the primary component of the ferrochrome. This can be seen for all input
carbon amounts at 850 ◦ C. The iron grade decreases at 1100 ◦ C when the chromium
begins to be reduced and decreases to a minimum grade of 28.9% at 20% input
carbon.
5.5. ATMOSPHERIC PRESSURE MODEL
101
Figure 5.5: Iron grade of the ferrochrome alloy as a function of temperature for
various initial carbon additions.
Figure 5.6 shows the iron recovery as a function of temperature for various initial
carbon inputs. High iron recoveries are seen at carbon inputs of 5% or higher.
Recoveries of 99.6% and 100% iron can be achieved with carbon inputs of 15 and
20%, respectively.
5.5. ATMOSPHERIC PRESSURE MODEL
102
Figure 5.6: Iron recovery as a function of temperature for various initial carbon additions.
Figure 5.7 shows the carbon grade of the ferrochrome alloy as a function of temperature for various initial carbon inputs. Carbon is only present in the ferrochrome
alloy once iron reduces to form an alloy. Carbon is present in the alloy at this time in
the form of Fe3 C. Carbon grades increase once chromium reduction begins. The primary species generated by the reduction of chromium are chromium carbides, which
increase the carbon grade. The drop after the increase is due to the use of carbides as
reducing agents, decarburizing them to form lower carbon species or metallic species.
For 2.5% initial carbon, the carbon level decreases due to the limited reduction of
5.5. ATMOSPHERIC PRESSURE MODEL
103
chromium and subsequent decarburization of carbides. In the 20% input carbon series, the carbon grades do not decrease after the increase due to chromium reduction.
This is due to the carbon being in excess and, as a result, the carbides do not react
with oxide species.
Figure 5.7: Carbon grade of the ferrochrome alloy as a function of temperature for
various initial carbon additions.
The unreacted carbon percentage as a function of temperature for various initial
carbon inputs can be seen in Figure 5.8. This is defined as a percentage of the
input carbon still present in the system as carbon. Carbon in a carbide form or in
a gas is not considered. It can be seen that almost all of the input carbon for each
run is present in the system until iron reduction begins at 850 ◦ C. Increasing the
5.5. ATMOSPHERIC PRESSURE MODEL
104
input carbon amount leads to additional unreacted carbon being present at higher
temperatures. This is attributed to the ratio of iron and chromium to carbon. If
there is less carbon compared to iron and chromium it is completely reacted at lower
temperatures. As this ratio increases the amount of unreacted carbon increases as
there is less material available to be reduced. It can be seen that for each run with
unreacted carbon remaining at 1100 ◦ C, the temperature when chromium reduces
drops to 0 in the next 100 ◦ C. The 20% test has unreacted carbon present at the
final reduction temperature due to the system being in excess of carbon. No more
material is able to be reduced and as such the remaining carbon cannot react.
Figure 5.8: Unreacted carbon percentage as a function of temperature for various
initial carbon additons.
5.5. ATMOSPHERIC PRESSURE MODEL
105
Figures 5.9 and 5.10 show the gas composition as a function of temperature for
an input carbon content of 15%. Figure 5.10 shows the gas composition without
CO. It can be seen that once reduction of iron and chromium begin the CO gas
amount increases significantly. This is the primary reduction product and as such
is expected to be generated in large volumes. Hydrogen gas is also present from
the decomposition of methane and water. Methane and water gases are added as
components of the charcoal. The concentration of these gas species decreases as the
temperature increases. Carbon dioxide is generated at lower temperatures, reaching
levels of almost zero after 800 ◦ C. Little oxygen gas is generated throughout the
entire reduction.
Figure 5.9: Gaseous species distribution between 200 and 1400 ◦ C for a 15% carbon
addition.
5.5. ATMOSPHERIC PRESSURE MODEL
106
Figure 5.10: Gaseous species distribution, without CO, between 200 and 1400 ◦ C for
a 15% carbon addition.
5.5.1
Effect of Carbon
A variant of the atmospheric pressure model was run to determine the effect of carbon
on iron and chromium. The model temperature was fixed at 1400 ◦ C.
Figure 5.11 shows the grades of chromium, iron, and carbon as a function of
input carbon at 1400 ◦ C. As the initial carbon content increases the iron grade decreases while the carbon and chromium grades increase. This is due to preferential
iron reduction. As more carbon is added chromium oxide begins to reduce, forming
chromium carbide and metallic chromium. An input carbon amount of 16% or over
5.5. ATMOSPHERIC PRESSURE MODEL
107
led to the same grades for all three species. This carbon level indicates completed reduction as no more oxide species are available to form metal species or to decarburize
carbide species.
Figure 5.11: Chromium, iron, and carbon grades as a function of input carbon at
1400 ◦ C.
The iron and carbon recoveries as a function of input carbon at 1400 ◦ C can be
seen in Figure 5.12. It can be seen that iron reaches higher recovery levels than
chromium for lower input carbon amounts. The iron recovery increases significantly
between 1 and 4% input carbon, then does not increase as quickly once chromium
reduction begins. This is due to preferential reduction of iron. Enough carbon must
be present in the system to reduce the iron before beginning to reduce chromium. The
5.5. ATMOSPHERIC PRESSURE MODEL
108
chromium recovery increase after 4% carbon input can be seen to increase linearly
per unit input. Maximum recoveries are obtained at approximately 15.5% input of
carbon.
Figure 5.12: Chromium and iron recoveries as a function of input carbon at 1400 ◦ C.
The unreacted carbon as a function of input carbon at 1400 ◦ C can be seen in
Figure 5.13. It can be seen that there is no unreacted carbon present until an input
carbon amount of 16%. For this amount of input carbon it is expected that 100% of
the chromium and iron will be recovered. Input carbon amounts over 16% will lead
to unreacted carbon being present in the sample which will dilute the final product.
5.6. REDUCED PRESSURE MODEL
109
Figure 5.13: Unreacted carbon as a function of input carbon at 1400 ◦ C.
5.6
Reduced Pressure Model
The model was run at reduced pressures between 0.2 and 0.01 bar in order to determine the effect of pressure. The model results between 1 and 0.2 bar are almost
identical, and there is less than 0.25% difference between results.
Figure 5.14 shows the chromium grade of the ferrochrome alloy as a function of
pressure and temperature. The grade increases as the pressure decreases for the same
temperature. This leads to chromium grades of 62% being obtained at temperatures
150 ◦ C lower. The lower pressure also leads to higher maximum grades of +0.2% at
0.01 bar.
5.6. REDUCED PRESSURE MODEL
110
Figure 5.14: Chromium grade as a function of temperature and pressure for 15%
input carbon.
Chromium recovery as a function of temperature and pressure can be seen in
Figure 5.15. As with the grade, the recovery increases as pressure decreases for the
same temperature. Similar recovery values are obtained at temperatures 150 ◦ C
lower at 0.01 bar than at 0.2 bar.
5.6. REDUCED PRESSURE MODEL
111
Figure 5.15: Chromium recovery as a function of temperature and pressure for 15%
input carbon.
The iron grade of the ferrochrome alloy as a function of temperature and pressure
can be seen in Figure 5.16. The iron grade increases from 0 to 95% between 650 and
800 ◦ C. Lower pressures result in higher grades at lower temperatures. This same
trend applies when chromium reduction begins and the iron grade decreases.
5.6. REDUCED PRESSURE MODEL
112
Figure 5.16: Iron grade as a function of temperature and pressure for 15% input
carbon.
The iron recovery as a function of temperature and pressure can be seen in Figure
5.17. Unlike the chromium recovery, the iron recovery increases to a maximum
of 99.9% over 200 ◦ C. Similar to previous trends, lower pressures lead to higher
recoveries at the same temperature. Similar recoveries are obtained 150 ◦ C lower at
0.01 bar than 0.2 bar.
5.6. REDUCED PRESSURE MODEL
113
Figure 5.17: Iron recovery as a function of temperature and pressure for 15% input
carbon.
Figure 5.18 shows the carbon grade of the ferrochrome alloy as a function of
temperature and pressure. Carbon is only present in the ferrochrome alloy after
iron reduction begins. This occurs at lower temperatures and at lower pressures.
The carbon grade then increases when chromium reduction begins, between 850 and
1000 ◦ C. As temperatures increase the carbon grade drops due to decarburization of
carbide species.
5.6. REDUCED PRESSURE MODEL
114
Figure 5.18: Carbon grade as a function of temperature and pressure for 15% input
carbon.
115
Chapter 6
Results and Discussion
6.1
Permittivity
Permittivity analysis of chromite and chromite ore with 8 and 15% carbon additions
are discussed in the two following subsections.
6.1.1
Effect of Frequency
The cavity perturbation technique used to analyze the samples allows for a variety of
frequency results to be reported. Figures 6.1 and 6.2 show the effect of frequency on
the real and imaginary permittivities for the chromite ore. It can be seen that both
the real and imaginary permittivities decrease with frequency. This trend can also
be seen in loss tangent and penetration depth data, as well as in ore with a carbon
addition. Plots demonstrating this can be found in Appendix C.
6.1. PERMITTIVITY
116
Figure 6.1: The effect of temperature on the real permittivity of chromite ore for
various frequencies. The sample density was 3.01 g/cm3 .
6.1. PERMITTIVITY
117
Figure 6.2: The effect of temperature on the imaginary permittivity of chromite ore
for various frequencies. The sample density was 3.01 g/cm3 .
6.1.2
Effect of Carbon
The real (0 ) and imaginary (”) permittivities for chromite and chromite with 8
and 15% carbon can be seen in Figures 6.3 and 6.4. The data reported is for 2466
MHz. The 15% carbon test has such a large discrepancy compared to the other
tests due to the sample being above the carbon percolation threshold. The sample
is sufficiently saturated in carbon to allow enough particles to touch and essentially
act as a conductor from one side of the sample to the other. This effect dissipates
once the sample is hot enough for the carbon to begin to react with the oxides and
6.1. PERMITTIVITY
118
leave the system. This occurs at 1200 ◦ C. The permittivity values for the 8 and 15%
tests are similar, however. They follow a similar trend and have end points that are
almost the same. This implies that both samples have a similar final composition.
The chromite ore sees a steady increase of both permittivities over the entire heating
range. This implies that, as the ore heats up, it is more likely to absorb and convert
microwaves to heat.
Figure 6.3: Comparison of real permittivities from 30 to 1430 ◦ C at 2466 MHz for
chromite ore and chromite with 8 and 15% carbon additions. Sample
densities were 3.01 g/cm3 for chromite, 2.18 g/cm3 for chromite with 8%
carbon, and 2.36 g/cm3 for chromite with 15% carbon.
6.1. PERMITTIVITY
119
Figure 6.4: Comparison of imaginary permittivities from 30 to 1430 ◦ C at 2466 MHz
for chromite ore and chromite with 8 and 15% carbon additions. Sample
densities were 3.01 g/cm3 for chromite, 2.18 g/cm3 for chromite with 8%
carbon, and 2.36 g/cm3 for chromite with 15% carbon.
The loss tangent comparison is shown in Figure 6.5. As seen previously, the 15%
test has a significantly different value for the duration of the test due to the percolation effect. The 8% carbon test begins higher than the chromite ore, implying
that the addition of the carbon allows for better microwave heat conversion. This
increases over the duration of the test, decreasing slightly at 800 ◦ C and then increasing again. The chromite ore can be seen to have a gradually increasing loss
tangent. As previously stated, as the ore heats up it will become a better absorber
6.1. PERMITTIVITY
120
of microwaves, further increasing in temperature. All three tests have a decrease
in their loss tangent values after approximately 1300 ◦ C. This is indicative of microwaves self-limiting behaviour, where the tests will absorb fewer microwaves after
this temperature and therefore decrease their rate of temperature increase.
Figure 6.5: Comparison of loss tangents from 30 to 1430 ◦ C at 2466 MHz for chromite
ore and chromite with 8 and 15% carbon additions. Sample densities were
3.01 g/cm3 for chromite, 2.18 g/cm3 for chromite with 8% carbon, and
2.36 g/cm3 for chromite with 15% carbon.
The penetration depth comparison can be seen in Figure 6.6. The low penetration
depth of the 15% test is attributed to its high carbon content. Microwaves cannot
penetrate the sample as they are quickly absorbed by the carbon. The 8% test has
6.1. PERMITTIVITY
121
a lower penetration depth than the chromite ore by itself for the lower temperature
range. This implies that a smaller sample is required to absorb enough microwaves to
begin increasing temperature. This trend reverses at 400 ◦ C, where the chromite ore
has a lower penetration depth for the remainder of the test. The low value obtained
implies that, at high temperatures, the microwaves will not penetrate too far into a
molten sample core, heating only the outside of the core and potentially expanding
it. The 15% test sees an increase in the penetration depth after 1200 ◦ C, when the
carbon is reacted and removed from the system. The 8% test, after 500 ◦ C, has little
change in the penetration depth until 900 ◦ C, where it increases before decreasing
once reduction reactions begin in the sample. The increase in the penetration depth
is attributed to chemical changes within the sample such as spinel decomposition.
6.2. TGA/DTA ANALYSIS
122
Figure 6.6: Comparison of penetration depths from 30 to 1430 ◦ C at 2466 MHz for
chromite ore and chromite with 8 and 15% carbon additions. Sample
densities were 3.01 g/cm3 for chromite, 2.18 g/cm3 for chromite with 8%
carbon, and 2.36 g/cm3 for chromite with 15% carbon.
6.2
TGA/DTA Analysis
Figure 6.7 shows the TGA data for chromite and chromite with carbon. It can be seen
that the pure chromite sample exhibits no change until approximately 500 ◦ C, where
it then loses approximately 2.5% of its mass. Another drop of about 0.5% occurs at
650 ◦ C. After this the mass stays relatively constant for the remainder of the test,
and a small increase in mass is seen. In contrast, the chromite with 15% carbon saw
significantly more mass change. There is an initial drop of approximately 2% which
6.2. TGA/DTA ANALYSIS
123
corresponds to moisture leaving the sample. There is then a gradual decrease until
500 ◦ C, where the volatile component of the charcoal is being removed and reacting
with some of the chromite. The drop at 500 ◦ C is the same as with the chromite
without carbon, indicating that the ore itself is undergoing a reaction leading to
the release of gas. After this, the primary reduction reaction begins and quickly
accelerates as the sample is heated to over 1000 ◦ C degrees. As the sample heats up
additional reduction reactions occur, releasing more CO and CO2 gas. The chromite
with 8% carbon follows trends similar to that of the chromite with 15% carbon, the
primary difference is that the mass loss at each stage is less than that of the chromite
with 15% carbon.
6.2. TGA/DTA ANALYSIS
124
Figure 6.7: Thermogravimetric comparison of chromite ore, and chromite with an 8
and 15% carbon addition.
6.2. TGA/DTA ANALYSIS
125
Figure 6.8 shows all differential thermal data for chromite and chromite with
carbon additions. It can be seen that there is not much deviation from one data set
to the other for the majority of the run. This slight variance can be explained by the
additional carbon. The primary difference occur at approximately 1250 to 1300 ◦ C,
where both chromite with carbon sets increase as the reduction reactions involving
chromite accelerate. At this time the majority of the carbon is used and converted
into carbides and carbon monoxide. Additional reactions involving carbides reacting
with oxide species begin to occur at this time. It is assumed that the chromite with
8% carbon sees this change in DTA 50 ◦ C later due to its reduced carbon amount.
Figure 6.8: Differential thermal analysis comparison of chromite ore, and chromite
with an 8 and 15% carbon addition.
6.2. TGA/DTA ANALYSIS
6.2.1
126
TGA/DTA with Permittivty Overlay
The permittivity data was combined with TGA/DTA data to determine if any relationship exists. The following three Figures, 6.9 through 6.11 show the data for the
chromite and the chromite with carbon additions.
Figure 6.9 shows the TGA/DTA data in conjunction with the real and imaginary
permittivities. It can be seen that the imaginary permittivity increases with the DTA
data. When the DTA data plateaus at 600 ◦ C the imaginary permittivity is seen to
have a decrease in rate. The rate of imaginary permittivity then begins to increase
after 900◦ C, when the DTA curve slope starts to be negative. Real permittivity
increases gradually throughout the entire test, and a drop at 1250 ◦ C is seen to
correspond to a drop in the DTA data at this same temperature.
6.2. TGA/DTA ANALYSIS
127
Figure 6.9: TGA, DTA, and permittivity data for chromite ore. Test conditions were
identical with an argon atmosphere and a maximum temperature of 1400
◦
C. Sample density for the permittivity test was 3.01 g/cm3 .
Figure 6.10 shows the chromite with 8% carbon TGA/DTA data as well as the real
and imaginary permittivities. It can be seen that the imaginary permittivity follows
the trends of the DTA plot very closely. An increase in the imaginary permittivty
is seen to correspond with a drop in the TGA curve at 1100 ◦ C. The imaginary
permittivty then drops off after the DTA data rises and begins to fall. The real
permittivity gradually increases throughout the duration of the test, and the largest
increase comes at approximately 400 to 800 ◦ C, the same point at which the TGA
data experiences a significant mass loss.
6.2. TGA/DTA ANALYSIS
128
Figure 6.10: TGA, DTA, and permittivity data for chromite ore with an 8% carbon
addition. The atmosphere used in both tests was argon with a maximum
temperature of 1400 ◦ C. Sample density for the permittivity test was
2.18 g/cm3 .
Figure 6.11 shows the chromite with 15% carbon TGA/DTA as well as the real
and imaginary permittivity values. Both the real and imaginary permittivities follow
a similar trend to that of the DTA curve. The primary difference compared to the
8% test, however, is the drop in both permittivities at 1200 ◦ C. This is stated as the
minimum reduction temperature in literature, seen in the TGA data as a mass drop
of approximately 15%. The mass is being lost as CO gas due to chromite reduction.
6.3. INITIAL HEATING TESTS
129
Figure 6.11: TGA, DTA, and permittivity data for chromite ore with a 15% carbon
addition. The atmosphere used in both tests was argon with a maximum
temperature of 1400 ◦ C. Sample density for the permittivity test was
2.36 g/cm3 .
6.3
Initial Heating Tests
Initial heating test data can be found in Figures 6.12 through 6.13. Additional plots
for 300 and 600 second tests can be found in Appendix C.
Figure 6.12 shows the absorbed power for the 600 second static and mode stirrer
heating tests. It can be seen that the mode stirrer obtained higher overall absorbed
power throughout the majority of the test. The no mode stirrer test can be seen to
have an increase in absorbed power rate after 700 seconds, and at 850 seconds the
6.3. INITIAL HEATING TESTS
130
absorbed power jumped 45% to 95%, before gradually decreasing to the end of the
run. A similar jump can be seen in the mode stirrer data. The gradual increase
seen at 700 seconds for the no mode stirrer tests occurs 100 seconds sooner using the
mode stirrer at 600 seconds, and the jump occurs 120 seconds sooner at 720 seconds.
This indicates that the use of a mode stirrer leads to maximum temperatures being
achieved sooner. Both of these jumps can be attributed to the dielectric properties
of the samples described previously in Section 6.1. It can be seen in Figures 6.3 and
6.4 that after reaching approximately 600 ◦ C the imaginary permittivity begins to
increase. This increase in turn allows for more microwave absorption, which increases
the heating rate. The real permittivity increases after 1250 ◦ C, and this is assumed
to correlate to the significant jumps in absorbed power seen in both the mode stirrer
and non mode stirrer tests.
6.3. INITIAL HEATING TESTS
131
Figure 6.12: Absorbed power over time for both mode stirrer and non mode stirrer
tests over 900 seconds. End temperature after each run was approximately 1360 ◦ C. Input power was 800 W in an argon atmosphere.
Figure 6.13 shows all three non mode stirrer tests. The absorbed energy increase
and drop at the beginning of each test is attributed to water evaporation. The
samples used in these tests were not dehydrated prior to heating. It can be seen that
all three tests follow a similar trend, being within 7% of each other for the duration
of their runs. Variances in absorbed power can be attributed to position within the
applicator cavity and slight variances in sample composition or packing. The final
temperatures for each test are 129 ◦ C for the 300 second test, 211 ◦ C for the 600
second, and 1357 ◦ C for the 900 second test.
6.3. INITIAL HEATING TESTS
132
Figure 6.13: Absorbed power over time for non mode stirrer tests for 300, 600, and
900 seconds. Input power was 800 W in an argon atmosphere.
Figure 6.14 shows double exponentially smoothed data for absorbed power over
time for the mode stirrer tests. The initial drops are due to water loss from the
sample. Tests follow similar trends for the duration of the testing. The 600 second
test shows an increase in the rate of absorbed power towards the end of its test. This
leads to a final temperature of 574 ◦ C. The final temperature for the 300 second test
was 130.2 ◦ C and 1365 ◦ C for the 900 second test.
6.4. BLACK THOR CHROMITE RESULTS
133
Figure 6.14: Double exponentially smoothed absorbed power over time for mode
stirrer tests for 300, 600, and 900 seconds. Input power was 800 W
in an argon atmosphere.
6.4
Black Thor Chromite Results
This section outlines the results obtained from reduction testing on the Black Thor
chromite ore.
6.4. BLACK THOR CHROMITE RESULTS
6.4.1
134
Effect of Carbon
Figure 6.15 shows the effect of time on absorbed energy for various initial carbon
amounts. It can be seen that, with no initial carbon, the sample will absorb approximately 300 kJ of energy during the duration of the 600 second run. The addition
of 2.5, 5, and 7.5% initial carbon causes the total absorbed energy to decrease by
approximately 12 kJ. This decrease diminished as additional carbon is added. This
is due to a change in the bulk density of the sample. Although carbon is an excellent
absorber of microwaves, enough of it must be present in a sample to significantly
affect absorption of microwaves. The carbon used in these tests has a lower density
than the ore, and as such an addition of carbon leads to an overall drop in bulk
density leading to a lower total absorbed energy. The 10% initial carbon addition
has a higher total absorbed energy than the lower carbon tests. At this point enough
carbon has been added to absorb more of the microwaves. At approximately 400
seconds it can be seen that the rate of energy rise increases implying that the sample had reached a temperature that allowed it to absorb more energy as reduction
reactions began. The total absorbed energy increases further with a 15% addition.
It can be seen that after 100 seconds the 15% initial carbon test is absorbing more
energy. This trend continues and ends with the 15% carbon sample absorbing 90 kJ
more than the carbon free ore. While there is no specific point at which the absorbed
energy rate suddenly increases, as seen in the 10% carbon test at 400 seconds, the
rate is gradually increasing over the duration of the 15% carbon run.
6.4. BLACK THOR CHROMITE RESULTS
135
Figure 6.15: Absorbed energy as a function of time for chromite ore and chromite
ore with various initial carbon additions. Tests were run in an argon
atmosphere with an input energy of 800 W.
Figure 6.16 shows the effect of input carbon on the mass loss and total absorbed
energy percent. Increasing the initial carbon leads to higher mass loss percentages,
increasing significantly at 10%. The absorbed energy decreases with low carbon
additions, increasing at 10% input carbon. This is due to a change in the bulk
density of the sample. Low amounts of input carbon lower the density of the sample
and reduce the absorption of the sample more than the benefit given by the very
absorbent carbon. Once enough carbon has been added, at 10% or higher, the sample
is capable of absorbing more microwave energy and reaching higher temperatures.
6.4. BLACK THOR CHROMITE RESULTS
136
Figure 6.16: The effect of carbon on the mass loss and the total absorbed power.
Tests were 600 seconds long in an argon atmosphere. The input power
was 800 W.
Figure 6.17 shows the effect of initial carbon content on the chromium grade of
the ferrochrome. No ferrochrome was formed during any 600 second test for carbon
levels below 10%. It can be seen that the chromium grade increases from 45.50%
for 10% initial carbon to 71.77% at 15%. This increase in grade is due to a higher
total absorbed energy and more available reducing material. As seen in Figure 6.15
the 15% carbon test absorbed 60 kJ more than the 10% carbon test, allowing it to
reach higher reaction temperatures and maintain that temperature. An increase in
the carbon amount allows for more reduction reactions to take place as the chromite
has more reducing material in contact with it.
6.4. BLACK THOR CHROMITE RESULTS
137
Figure 6.17: Chromium grade as a function of initial carbon percentage. Tests were
run in argon with an input power of 800 W for 600 seconds.
Figure 6.18 shows the effect of initial carbon on the overall and core recoveries
for both chromium and iron. It can be seen that the overall recovery of chromium
and iron are almost identical for the 10% initial carbon test. Both of these recoveries
increase and vary for the 15% carbon overall recovery. Core recoveries for iron are
higher for both 10 and 15% initial carbon contents due to the preferential reduction
of iron over chromium.
6.4. BLACK THOR CHROMITE RESULTS
138
Figure 6.18: Overall and core recoveries of chromium and iron as a function of initial
carbon amount. Tests were run in an argon atmosphere with an input
power of 800 W for 600 seconds.
The two following Figures, 6.19 and 6.20, show sections of the reduced core at
approximately the same scale. It can be seen that the 10% input carbon reduced core
has some small beads of ferrochrome present, however much of the core is partially
reduced oxide. In comparison, the 15% input carbon sintered core has much larger
and more numerous ferrochrome beads present. The large particle seen in Figure 6.20
is approximately 1 mm long and can easily be seen without the aid of microscopes.
6.4. BLACK THOR CHROMITE RESULTS
139
Figure 6.19: SEM image of the polished reduced core from the 10% input carbon
test. This sample was reduced for 600 seconds in an argon atmosphere
with 800 W input power.
6.4. BLACK THOR CHROMITE RESULTS
140
Figure 6.20: SEM image of the polished reduced core from the 15% input carbon
test. This sample was reduced for 600 seconds in an argon atmosphere
with 800 W input power.
6.4.2
Effect of Energy
This subsection details the effect of energy on the reduction of the chromite ore.
Figure 6.21 shows the mass loss rate as a function of input power for the three
atmospheric compositions. It can be seen that the air and argon tests have rates
6.4. BLACK THOR CHROMITE RESULTS
141
that increase linearly with input power. The argon mass loss rate increases faster
than that of both air and vacuum. In air testing this lower mass loss rate is a result
of the reducing carbon reacting with oxygen in the air to form carbon monoxide /
carbon dioxide instead of reacting with the oxide species in the sample. Vacuum
testing sees the highest rate of mass loss for 400 W. The mass loss rate increases for
the 800W tests but decreases for the 1200 W tests. This is due to the fluidization
of the sample as it is heated. As CO gas is generated and removed the particles
hot enough to undergo reduction are displaced from their original positions due to
the fluidization. These particles move heat away from the center of the sample and
required it to be reheated. This effect repeats until some particles get hot enough to
sinter together and from a reduced core. Further reduction can then commence.
6.4. BLACK THOR CHROMITE RESULTS
142
Figure 6.21: Rate of mass loss as a function of input power for various atmospheric
and vacuum compositions. Tests varied in duration from 150 to 600
seconds.
The reduced core mass trend as a function of absorbed energy for various atmospheric compositions can be seen in Figure 6.22. It can be seen that an absorbed
energy of 140 kJ is the lowest amount that produced a reduced core. Air and argon
testing were able to produce 4 reduced cores each, while vacuum in comparison only
produced one. Regression analysis shows that increasing the absorbed energy in air
and argon increases the reduced core mass. The reduced core masses at the higher
6.4. BLACK THOR CHROMITE RESULTS
143
absorbed energy amounts are also closer in mass to each other than those at lower
absorbed energies. This discrepancy is due to a difference in reaction time. The
vacuum tests at absorbed energies of approximately 400 kJ are affected the most by
the difference in reduction time. The test at 380 kJ with a reduced core mass of
approximately 7% was microwaved for 600 seconds and formed a reduced core. In
comparison the test at 410 kJ was only microwaved for 400 seconds, however it had
an input power level of 1200 W. This shorter reduction time did not allow for the
formation of a reduced core as due to the fluidization effect previously described.
The lack of a reduced core leads to no ferrochrome being produced, which can be
seen in the subsequent figures.
6.4. BLACK THOR CHROMITE RESULTS
144
Figure 6.22: Reduced core mass as a function of absorbed energy for various atmospheric and vacuum compositions. Tests varied between 150 and 600
seconds with input energies between 400 and 1200 W.
Figure 6.23 shows the effect of carbon percentage in the reduced core as a function
of absorbed energy for various atmospheric compositions. Samples which did not
produce a reduced core are not shown. A first-order regression is also shown. It
can be seen that the carbon in the reduced core decreases with increasing absorbed
energy. This can be attributed to the reduction proceeding further as more energy
is absorbed.
6.4. BLACK THOR CHROMITE RESULTS
145
Figure 6.23: Carbon in the reduced core as a function of absorbed energy for various
atmospheric and vacuum compositions. Tests varied between 150 and
600 seconds with input energies of 800 and 1200 W.
Figure 6.24 shows the effect of absorbed energy on the chromium grade for various
atmospheric compositions. It can be seen that the majority of tests that produced
ferrochrome feature a final chromium grade of over 60%. This can be attributed to
microwave heating effects. These tests all experienced a rapid increase in temperature
prior to reduction, as seen in the heating tests towards the end of their runs. The
increase in temperature is enough to begin reduction, which allows for both iron and
6.4. BLACK THOR CHROMITE RESULTS
146
chromium to reduce at the same time. Iron is preferentially reduced, which is the
reason why the lower absorbed energy tests have more variance in their chromium
grades compared to the higher absorbed energy tests. These low absorbed energy
tests are short in duration and may only experience reduction for a fraction of that
of the longer tests, this is enough time for only part of the chromium to reduce.
Longer tests which result in higher absorbed energy have more time for reduction
to occur, and can move towards an equilibrium and maintain it over the duration of
their reduction.
6.4. BLACK THOR CHROMITE RESULTS
147
Figure 6.24: Chromium grade as a function of absorbed energy for various atmospheric and vacuum compositions. Reaction times varied from 150 to
600 seconds with an input power range of 400 to 1200 W.
Figure 6.25 shows the overall recovery of chromium as a function of absorbed energy. It can be seen that for lower absorbed energies the recoveries are low, increasing
as the absorbed energy increases. Vacuum recovery is low for high absorbed energy.
This can be attributed to short reduction times due to fluidization of the sample or a
lack of reducing materials present in the core. The fluidization does not allow for the
formation of a reduced core until particles get hot enough to sinter together. This
6.4. BLACK THOR CHROMITE RESULTS
148
is not expected to occur until later in the test when the sample has reached a high
enough temperature. During the fluidization the carbon is also used to form CO
which leads to the effect occurring. The use of carbon during this stage can lead to
less carbon being present when the core forms to reduce the oxide species. Maximum
overall recoveries of 57% can be seen at high absorbed energies for an argon test.
Figure 6.25: Overall chromium recovery as a function of absorbed energy for various
atmospheric and vacuum compositions. Reaction times varied from 150
to 600 seconds with an input power range of 400 to 1200 W.
Figure 6.26 shows the core recovery of chromium as a function of absorbed energy
6.4. BLACK THOR CHROMITE RESULTS
149
for various atmospheric compositions. Vacuum recoveries were low, only reaching
22%, due to the effects described in the summary of the previous plot. Maximum
recoveries of 81% are seen for an argon test. Similar to the chromium grade seen in
Figure 6.24, the recovery jumps from 0 to between 38 and 62% at lower absorbed
energy levels. This is indicative of a high temperature which allows for chromium
and iron to reduce simultaneously. The higher absorbed energy recovery values are
also similar to the trend seen in the grade. The longer reaction time allows for more
material to reduce and begins to reach an equilibrium.
6.4. BLACK THOR CHROMITE RESULTS
150
Figure 6.26: Core chromium recovery as a function of absorbed energy for various
atmospheric and vacuum compositions. Reaction times varied from 150
to 600 seconds with an input power range of 400 to 1200 W.
Overall iron recoveries as a function of absorbed energy for various atmospheric
compositions can be seen in Figure 6.27. It can be seen that at lower absorbed
energies air testing provides higher recoveries over argon and vacuum. This changes
at higher absorbed energy levels, those over 340 kJ, when argon recoveries are higher.
Recoveries for vacuum systems are the lowest of all three tests.
6.4. BLACK THOR CHROMITE RESULTS
151
Figure 6.27: Overall iron recovery as a function of absorbed energy for various atmospheric and vacuum compositions. Reaction times varied from 150
to 600 seconds with an input power range of 400 to 1200 W.
Figure 6.28 shows the core iron recovery as a function of absorbed energy. Core
iron recoveries at lower absorbed energies are higher than those seen at the same
absorbed energy for chromium. This implies that iron is preferentially reduced over
chromium. As with chromium, iron recoveries in vacuum are the lowest. The highest
iron recoveries were 83% during an argon test.
6.4. BLACK THOR CHROMITE RESULTS
152
Figure 6.28: Core iron recovery as a function of absorbed energy for various atmospheric and vacuum compositions. Reaction times varied from 150 to
600 seconds with an input power range of 400 to 1200 W.
Figure 6.29 shows the actual absorbed energy versus the input energy for various
atmospheric compositions. The thick black line indicates 100% absorption. Through
regression analysis it can be seen that vacuum tests have the highest absorbed energies of the three atmospheric test types conducted despite not achieving reducing
temperatures in many tests. This was attributed to the fluidization phenomenon
seen in vacuum testing. Argon tests were slightly less efficient than vacuum tests,
6.4. BLACK THOR CHROMITE RESULTS
153
with air being the lowest of the three. This was previously attributed to undesired
reactions between carbon and oxygen forming CO and CO2 .
Figure 6.29: The input versus actual absorbed energy for each test run in all three atmospheric and vacuum conditions. The bold line indicates 100% energy
absorption. Included are regressions for all three atmospheric compositions. Reaction times varied from 150 to 600 seconds with input powers
ranging from 400 to 1200 W.
The following SEM images, Figures 6.30 through 6.38, illustrate the effect of
energy on the sintered core product. Images are displayed one after another where
each has the same total ideal absorbed energy.
6.4. BLACK THOR CHROMITE RESULTS
154
Figures 6.30 and 6.31 show the sintered core for samples from two air tests with
an input energy of 240 kJ. In actuality these samples absorbed 160 and 144 kJ for
the 300 and 200 second tests, respectively. It can be seen that, although the 200
second had less reaction time, it was able to form much larger ferrochrome particles.
The 300 second test has many more smaller ferrochrome particles forming around the
edges of the partially reacted oxide. This is the growth behaviour expected. With
additional time these small beads would have agglomerated into larger particles.
6.4. BLACK THOR CHROMITE RESULTS
155
Figure 6.30: SEM image of the sintered core from the 300 second, 800 W air test.
6.4. BLACK THOR CHROMITE RESULTS
156
Figure 6.31: SEM image of the sintered core from the 200 second, 1200 W air test.
Figures 6.32 and 6.33 show the sintered core for two air tests with an input
energy of 480 kJ. These tests actually absorbed 331 and 337 kJ for the 600 and 400
second tests, respectively. Unlike the previous two images, both of these contain large
particles of ferrochrome, which are visible without the aid of a microscope. However,
as before, the 400 second test still contains larger particles than its 600 second
6.4. BLACK THOR CHROMITE RESULTS
157
counterpart. This indicates that the rate of energy input is critical to ferrochrome
formation. Higher energy input rates allow for the sample to heat much faster at its
core and quickly rise to reaction temperature. Lower rates do not heat the sample as
quickly allowing for some of the heat generated to dissipate throughout the sample
or leave the sample entirely.
Figure 6.32: SEM image of the sintered core from the 600 second, 800 W air test.
6.4. BLACK THOR CHROMITE RESULTS
158
Figure 6.33: SEM image of the sintered core from the 400 second, 1200 W air test.
Figure 6.34 shows a small section of the unreacted powder that was located in
the boundary between the reduced core and unreacted sample. Chromium particles
can be seen forming on the edges and within cracks in the chromite particles. These
ferrochrome particles are similar to those described by Lekatou and Walker (1995).
They are primarily iron, and will increase in size as chromium species are reduced
6.4. BLACK THOR CHROMITE RESULTS
159
and report to the alloy. These small particles will eventually agglomerate to form
the larger particles seen in the previous images.
Figure 6.34: SEM image from a part of the boundary between powder and reacted
core from the 600 second, 800 W test.
Figures 6.35 and 6.36 show the sintered core of two argon tests with an input
energy of 240 kJ. These tests have an actual absorbed energy of 171 and 183 kJ for
the 200 and 300 second tests, respectively. It can be seen that the 300 second test has
6.4. BLACK THOR CHROMITE RESULTS
160
fewer, larger particles than the 200 second test. It also contains more intermediate
sized particles which would agglomerate to form larger particles. Substantially more
ferroalloy can be seen inside the crack in the partially reduced oxide in the 300 second
test compared to the 200 second test.
Figure 6.35: SEM image of the sintered core from the 300 second, 800 W argon test.
6.4. BLACK THOR CHROMITE RESULTS
161
Figure 6.36: SEM image of the sintered core from the 200 second, 1200 W argon
test.
Figures 6.37 and 6.38 show sintered cores from two argon tests with an ideal total
input energy of 480 kJ. Both of these tests formed large ferrochrome beads over 500
µm in size. Chromium grades between these samples varied less than 1%, however
the chromium recovery for the 400 second test was 80.37%, 7% higher than for the
600 second test.
6.4. BLACK THOR CHROMITE RESULTS
162
Figure 6.37: SEM image of the sintered core from the 600 second, 800 W argon test.
6.4. BLACK THOR CHROMITE RESULTS
163
Figure 6.38: SEM image of the sintered core from the 400 second, 1200 W argon
test.
6.4.3
Effect of Atmospheric Composition
Figures 6.39 and 6.40 show the maximum grades and recoveries for each atmospheric
composition. Grades and recovery values were taken from the same test. Maximum
grades and recoveries for both air and argon were obtained during the 400 second
6.4. BLACK THOR CHROMITE RESULTS
164
1200 W test, while the vacuum test was conducted for 600 seconds and at 800 W
input.
Figure 6.39: Highest obtained chromium grades for various atmospheric and vacuum
conditions.
6.4. BLACK THOR CHROMITE RESULTS
165
Figure 6.40: Highest obtained chromium and iron recoveries for various atmospheric
and vacuum conditions.
Figure 6.41 shows the effect of particle size on chromium grade for all three
atmospheric types. It can be seen that for all atmospheric types the chromium grade
increases with particle area. This is due to smaller particles being predominantly iron
as iron is the first metal species to reduce and form an alloy phase. As chromium
begins to reduce the particles increase in both size and chromium grade.
6.4. BLACK THOR CHROMITE RESULTS
166
Figure 6.41: Chromium grade as a function of particle area for various atmospheric
and vacuum conditions.
6.5. DISCUSSION
6.5
167
Discussion
Table 6.1 shows a summary of results for each atmospheric test type conducted.
Table 6.1: Summary of results for reduction testing.
Criteria
Air
Argon
Vacuum
Grade
71.04%
72.89%
65.03%
Recovery (Core)
76.02% Cr, 82.83% Fe
80.37% Cr, 83.30% Fe
21.78% Cr, 26.31% Fe
Mass Loss Rate (all)
0.78 / 6.33 / 11.9 mg/s
1.58 / 10.2 / 18.7 mg/s
2.76 / 7.98 / 3.8 mg/s
Reduced Core Mass
13.2 g
16.29 g
7.17 g
Absorbed Energy (avg)
65.49%
73.77%
77.28%
Carbon in Reduced Core
2.88%
1.95%
2.13%
The chromium grade does not vary significantly between air and argon testing. In
all three atmospheric compositions, the chromium grade is higher than the thermodynamically predicted 62%. Similar air and argon tests produced the highest grades
seen, for a 400 second duration and 1200 W power reduction test. Vacuum testing
required more time, as the 600 second, 800 W test was the only test to produce
ferrochrome. This is attributed to fluidization occurring within the sample. This
fluidization does not allow for the formation of a reduced core until particles are
hot enough to sinter together. The recovery values found for chromium have more
variance than the grade, and are within the error of ±6.42%. Vacuum has very low
recoveries for both chromium and iron. This can again be attributed to the delayed
formation of the reduced core.
Mass loss rates are on average highest for argon testing. In air tests the carbon
6.5. DISCUSSION
168
is able to react with oxygen present in the air instead of the oxides in the sample,
leading to less mass loss in the sample. Vacuum samples are not able to easily form
a reduced core to reduce oxide species, lowering their mass loss rate.
The reduced core masses are high for both air and argon tests, making up over
50% of the remaining sample mass in some cases. The reduced core of the vacuum
test is the smallest of all three atmospheric types. This can be attributed to a delayed
core formation due to fluidization.
Vacuum tests had the highest absorbed energies of the three atmospheric types,
despite fluidization and no reduced core. This is a result of fluidization. Heat generated in one area of the sample is spread throughout the entire sample by the particle
movement. This causes the entire sample to heat up in a more uniform fashion than
for other atmospheric types, however it means that additional energy is required to
reach reaction temperatures. The overall increase in sample temperature also leads
to an increase in the dielectric properties which led to increased absorption. Neither
air nor argon saw this fluidization effect. Gases generated within the sample would
not violently exit due to either air or argon being present at 1 bar of pressure. It
was expected that the vacuum test would react faster due to the removal of CO gas,
therefore reducing its partial pressure and disrupting the reaction equilibrium. Such
an effect could have been present in argon testing as the flowing argon would have
carried away some of the generated CO.
The carbon in the reduced core is the highest for air testing, followed by vacuum
testing, and has the lowest amount in argon testing, lower than air by 0.93%. Air
testing is expected to have the lowest amount of carbon as the oxygen in the air
6.5. DISCUSSION
169
can react with it and remove it. The carbon is primarily located within the dense
reduced core as a carbide species, and it is possible that oxygen could not enter the
core to react with the remaining carbon. It is believed that the argon testing has
the lowest carbon amount due to a higher degree of reduction. Argon tests absorbed
more energy than air tests, and, as a result, it was assumed that the additional
energy allowed for the decarburization of the ferroalloy by reacting with oxides. The
vacuum test has a low amount of carbon for a different reason. The low recovery
of both iron and chromium implies that the majority of the core is comprised of
partially reduced oxides. This leads to a lower amount of ferrochrome and therefore
carbide species present, resulting in a lower carbon amount.
170
Chapter 7
Conclusions and Recommended Future Work
7.1
Thermodynamic Model
A thermodynamic model was constructed for initial study of the chromite reduction
process. A potential list of 306 species was narrowed down to 27 based on their stability across the 200 to 1400 ◦ C temperature range. Activity coefficients for species
in the model were obtained from literature. Assays of the chromite ore provided the
ratios of input species to add and proximate analysis on activated charcoal provided
the reducing agent quantities. The model was run at atmospheric pressure to determine the effects of temperature and input carbon grade on several factors including
chromium and iron grades, recoveries, and CO gas evolution. The model was again
run at reduced pressures between 0.2 and 0.01 bar to determine the effect of pressure.
The ideal amount of initial reducing material was determined to be 15% carbon.
For this initial carbon level at 1400 ◦ C the chromium grade and recovery were determined to be 61.32% and 100%. Reduction tests under argon produced a maximum
chromium grade of 72.89% and recovery of 80.37%. It is presumed that if the reaction
7.2. BLACK THOR CHROMITE
171
was able to go on for an extended period of time then the recovery would increase
and the grade would drop as more iron is reduced to dilute the product. This would
put the reduction tests more in line with the thermodynamic predictions. Iron recoveries were lower than the expected 99.99%, and the same reduction test as mentioned
previously was able to achieve an iron recovery of 83.30%. Again, with additional
time, the reduction system is expected to have results similar to those predicted by
thermodynamics.
The effect of pressure was determined to only play a factor at pressures lower than
0.2 bar. It was found that lower pressures could be used to achieve higher grades
and recoveries compared to atmospheric pressure at the same temperature. Overall,
it was found that lower pressures were able to achieve grade and recovery values
at temperatures 150 ◦ C lower than at higher pressures. Final grades and recoveries
did not benefit as much from the reduced pressure, only increasing by 0.3% in some
cases.
7.2
Black Thor Chromite
Several tests were performed on pulverized chromite ore. Ore was first pulverized
to 80% passing 100 mesh, and the activated charcoal was pulverized to 80% passing
200 mesh. 30 g samples were placed into quartz crucibles and heated via microwave
energy. Analysis was performed using a scanning electron microscope and mineral
liberation analysis techniques.
7.2. BLACK THOR CHROMITE
7.2.1
172
Effect of Carbon
The effect of carbon addition was studied using five samples with carbon grades
varying from 2.5, 5, 7.5, 10, and 15%. These tests were conducted in an argon
atmosphere, with 800 watts input power for 10 minutes. Tests with 2.5 through
7.5% carbon addition did not reach reduction temperature during the test time.
The 10 and 15% carbon tests reached reduction temperature, and both produced a
sintered core with ferrochrome present. The grades obtained were 45.53 and 71.77%
for the 10 and 15% addition tests, respectively. Recoveries for chromium were 12.27%
for the 10% carbon addition and 73.19% for the 15% carbon addition. Iron recoveries
were 28.27 and 82.42% for the 10 and 15% additions, respectively. The total mass
loss for the 15% carbon test was 7.2% higher than for the 10% test. These results
indicate that additional carbon leads to reaction times occurring sooner and is able
to progress farther and affect more material.
7.2.2
Effect of Energy
Several tests were conducted with the same maximum ideal absorbed energy, with
variances in the rate at which the energy was added into the system. Three tests
had an ideal absorbed energy of 240 kJ while two had 480 kJ.
For the 240 kJ testing, one which had the energy supplied slowly, 400 watts for
600 seconds, did not reach reduction temperature. The other two, which had 800 W
supplied for 300 seconds and 1200 W supplied for 200 seconds did reach reduction
temperature and formed ferrochrome in both air and argon testing. In air testing, the
1200 W test featured a larger sintered core than the 800 W test, at 8.75 g compared
7.2. BLACK THOR CHROMITE
173
to 3.86 g. Chromium grades for the 1200 W test were 10% higher than for the 800
W test. The chromium recovery was 22.30% higher at 63.20% compared to 40.90%.
Iron recovery was 14.5% higher at 67.14%. The argon tests showed different trends
than the air tests. Both the grades and recoveries were higher for the 800 W test.
Grades were 74.65% for the 800 W test compared to 53.68% for the 1200 W test.
Chromium recoveries were 51.66% and 35.88%.
Both of the 480 kJ tests produced ferrochrome for air and argon. In air testing,
the grades differed by less than 2%, at 69.30% in the 800 W test and 71.04% in the
1200 W test. The chromium recovery was 9% higher at 76.02% for the 1200 W test.
The iron recovery was 8.8% higher at 82.83% for the 1200 W test.
Overall, the indication is that a faster rate of energy addition (higher power level)
leads to better grades and recoveries after a certain amount of time. The higher power
into the system allows for the samples to heat up and reach reduction temperatures
faster than for lower power levels. In addition to this, heat flow out of the system
has a much larger impact on the lower power systems, as this lost heat is not able
to be replenished as fast. Microwave heating accelerates in this sample the hotter
it gets, up to approximately 1350 ◦ C. The faster a sample gets hotter the faster it
is able to absorb more energy, in turn heating up faster and allowing reduction to
happen.
7.2.3
Effect of Atmosphere
Three different atmospheric compositions were used during the reduction test. Air
and argon tests were able to easily form ferrochrome, while the vacuum test was only
7.2. BLACK THOR CHROMITE
174
able to form ferrochrome during one test. Air and argon testing had results similar
to each other, however the highest grade and recoveries were present during argon
testing. A maximum grade of 72.89% was found during a 400 second test at 1200 W.
The corresponding test in air, which lead to the highest air grades and recoveries,
had a chromium grade slightly lower at 71.04%. Recoveries for the argon test were
80.37% for chromium and 83.30% for iron. This differed from the air test by only
4.35% and 0.47% for chromium and iron, respectively. The difference in grade and
recovery can be attributed to a re-oxidation effect in the air environment. As the
sample cools the chromium is preferentially oxidized by oxygen present in the air
back into an oxide, lowering both grade and recovery.
The vacuum testing produced a chromium grade of 65.03%, with recoveries
of chromium and iron of 21.78 and 26.31%. Only one test in vacuum produced
chromium. This can be attributed to the evolution and immediate removal of gases
from the system. In literature the reduction reaction is stated to require an intermediate gasification step which requires contact of carbon to the chromite in order
to generate CO. This CO is then used for further reduction, however, if it is immediately removed from the system, reduction will take longer. The fluidization effect
also moves heat around the sample. In the air and argon testing a reduced core of
high temperature material was able to form, however, if the sample is fluidized, then
no reduced core would be able to form.
The atmospheric composition is extremely important in the reduction of chromite.
Air and argon testing both produced high grades and recoveries, while the vacuum
did not. Should this process be scaled up, argon or another inert gas is suggested
7.3. FUTURE WORK
175
to avoid the oxidation of already reduced chromite. Vacuum reduction on a powder
can be used, however enough time is needed to ensure reduction.
7.3
Future Work
From the work done on this thesis several future work recommendations can be
made. First, the incorporation of a binding agent should be attempted to allow for
vacuum testing to proceed. In this thesis, bentonite was attempted to be used in
small quantities, to no avail. It should be noted that the addition of a binder could
change the reduction chemistry. It is vital to use a binder with little to no effect on
the final product. Using such agents will need to be investigated, and if enough of
one is used a new thermodynamic model will need to be constructed to account for
other species or elements present.
Second, different types of chromite ores should be tested in a similar manner.
The ore seen here is metallurgical grade, thus other compositions involving varying
amounts of chromium, iron, magnesium, silicon, and aluminium should be used. In
addition, the sample received for this thesis was one large portion, and variances in
the previously mentioned element ratios can cause significant changes in the reduction
characteristics. Thermodynamic modelling can be performed to try to minimize the
changes, however small scale testing should be performed to ensure quality.
Third, different reducing agents should also be tested to see if they can provide
improved grades or recoveries. Activated charcoal is readily available however the
composition of carbon, moisture, and volatiles can change which can lead to different
reducing behaviours. Sulphur, in the form of pyrite, should be studied by itself or in
7.3. FUTURE WORK
176
conjunction with a carbon agent.
Fourth, different gas atmospheres should be tested. Related to the previous point,
the use of carbon monoxide gas or methane (or a mixture of the two) could be studied
as those gases could replace a solid reducing agent.
Finally, a second step purification or separation step could be used to increase
both the grades and recoveries. Unreacted powder could simply be reprocessed,
however the sintered core should be treated again to raise quality. This could be
performed by a second reduction using carbon or another reducing agent after pulverizing the core, or if a continuous process is devised a reducing gas such as carbon
monoxide could replace the other gas to allow for continued core reduction. Separation of the ferrochrome could also be performed. Magnetic separation of a crushed
core would not work as effectively as some of the oxides in the sintered core are
magnetic or paramagnetic. Gravity separation or possibly flotation could be utilized
on a crushed sintered core.
BIBLIOGRAPHY
177
Bibliography
[Amankwah and Ofori-Sarpong, 2011] Amankwah, R. K. and Ofori-Sarpong, G.
(2011). Microwave heating of gold ores for enhanced grindability and cyanide
amenability. Minerals Engineering, 24:541–544.
[Amer, 1992] Amer, A. M. (1992). Processing of ras-shait chromite deposits. Hydrometallurgy, 28:29–43.
[Amer and Ibrahim, 1996] Amer, A. M. and Ibrahim, I. A. (1996). Leaching of a low
grade egyptian chromite ore. Hydrometallurgy, 43:307–316.
[Andersson, 1988] Andersson, J. O. (1988). A thermodynamic evaluation of the fecr-c system. METALLURGICAL TRANSACTIONS A, 19:627–636.
[ASTM, 2013] ASTM (2013). Standard Practice for Proximate Analysis of Coal and
Coke. ASTM International, 100 Barr Harbor Drive, West Conshohocken, PA, 13
edition.
[ASTM, 2015] ASTM (2015). Standard Test Methods for Proximate Analysis of Coal
and Coke by Macro Thermogravimetric Analysis. ASTM International, 100 Barr
Harbor Drive, West Conshohocken, PA, 15 edition.
BIBLIOGRAPHY
178
[Bayat et al., 2013] Bayat, O., Altiner, A., Altincelep, Z., and Kaymakoglu, B.
(2013). Processings of the xv balkan mineral processing congress. In Effects of
microwave pre-heating on grindability and liberalization degree of chromite ores.
[Berger and Frei, 2014] Berger, A. and Frei, R. (2014). The fate of chromium during tropical weathering: A laterite profile from central madagascar. Geoderma,
(213):521–532.
[Bhonde et al., 2007] Bhonde, P. J., Ghodgaonkar, A. M., and Angal, R. D. (2007).
Infacon xi. In Various techniques to produce low carbon ferrochrome, New Delhi,
India. The Indian Ferro Alloy Producers Association.
[Boericke, 1944] Boericke, F. (1944). Equilibria in the reduction of chromic oxide
by carbon and their relation to the decarburization of chromium and ferrochrome.
Technical report, Bureau of Mines.
[Calvero., 2016] Calvero. (2016).
X-ray fluorescence simple figure.
URL:
https://commons.wikimedia.org/wiki/File:X-ray fluorescence simple figure.svg.
[Chandrasekaran et al., 2012] Chandrasekaran, S., Ramanathan, S., and T., B.
(2012). Microwave material processing - a review. Fluid Mechanics and Transport
Phenomena, 58(2):330–363.
[Chen et al., 2009] Chen, J. amd Zhao, J., Zhang, M., Yan, H., and Zhou, J. X.
(2009). Carburization of ferrochromium metals in chromium ore fines containing
coal during voluminal reduction by microwave heating. Journal of Central South
University of Technology, 16(43-48).
BIBLIOGRAPHY
179
[Donahue and Rais, 2009] Donahue, C. J. and Rais, E. A. (2009). Proximate analysis
of coal. Journal of Chemical Education, 86(2):222–224.
[Durand-Charre, 2004] Durand-Charre, M. (2004). The Microstructure of Steels and
Cast Irons. Springer.
[Elliot, 2015] Elliot, R. S. B. (2015). A study on the role of sulphur in the thermal
upgrading of nickeliferous laterite ores. Master’s thesis, Queen’s University.
[ELTRA, 2010] ELTRA (2010). CS-200 Dual Furnace Double Dual Range Carbon
/ Sulfur Determinator. ELTRA GmbH, Mainstr. 85 Block 20 D-41469 Neuss,
Germany.
[ELTRA, 2016] ELTRA (2016). Carbon Sulfur Analyzer CS-2000. ELTRA GmbH,
Mainstr. 85 Block 20 D-41469 Neuss, Germany.
[FEI, 2011] FEI (2011). Quanta 650 FEG. FEI, 5350 NE Dawson Creek Drive
Hillsboro, Oregon 97124 USA.
[Forster, 2015] Forster, J. H. (2015). Investigation of the microwave vacuum carbothermic reduction and sulphidation of a low grade nickeliferous silicate laterite
ore. Master’s thesis, Queen’s University.
[GAE, 2015] GAE
(2015).
Gerling
applied
engineering.
URL:
http://www.2450mhz.com/.
[Geovic, 2010] Geovic (2010). Chromite market summary. Technical report, Geovic
Mining Corp.
BIBLIOGRAPHY
180
[Geveci et al., 2002] Geveci, A., Topkaya, Y., and Ayhan, E. (2002). Sulfuric acid
leaching of turkish chromite concentrate. Minerals Engineering, 15:885–888.
[Ghose et al., 1983] Ghose, S., Nanda, J. K., and Patel, B. B. (1983). Proceedings of
the seminar on problems and prospects of ferro-alloy industry in india. In Duplex
process for production of low carbon ferrochrome, Jamshedpur, India.
[Gujar et al., 2010] Gujar, A. R., Ambre, N. V., Iyer, S. D., Mislankar, P. G., and
J., L. V. (2010). Placer chromite along south maharashtra, central west coast of
india. CURRENT SCIENCE, 99(4):492–499.
[Gupta and Wong, 2007] Gupta, M. and Wong, E. (2007). Microwaves and Metals.
Wiley.
[Hadjiantonis, 2013] Hadjiantonis,
M.
(2013).
Bragg’s
law.
URL:
https://commons.wikimedia.org/wiki/File:Bragg’s law.svg.
[Hao et al., 2014] Hao, J. J., Chen, J., Han, P. d., Li, K., and Sun, H. f. (2014).
Solid phase decarburization kinetics of highcarbon ferrochrome powders in the
microwave field. Steel Research International, 85(3):461–465.
[Hino et al., 1998] Hino, M., Higuchi, K. I., Nagasaka, T., and Ban-Ya, S. (1998).
Thermodynamic estimation on the reduction behaviour of iron-chromium ore with
carbon. Metallurgical and Materials Transactions B, 29B:351–360.
[Hock et al., 1986] Hock, M., Friedrich, G., Pluger, W. I., and Wichowski, A.
(1986). Refractory- and metallurgical-type chromite ores, zambales ophiolite, luzon, philippines. Mineralium Deposita, 21:190–199.
BIBLIOGRAPHY
181
[Hutcheon et al., 1992] Hutcheon, R., de Jong, M., and Adams, F. (1992). A system
for rapid measurements of rf and microwave properties up to 1400◦ c. Journal of
Microwave Power and Electromagnetic Energy, 27(2):87–92.
[ICDA, 2011] ICDA (2011). Discover chromium - ore processing. Technical report,
International Chromium Development Association.
[ISSF, 2015] ISSF (2015). Stainless steel in figures 2015.
[Jacob and Behera, 2000] Jacob, K. T. and Behera, C. K. (2000). Spinel-corundum
equilibria and activities in the system mgo-al2o3-cr2o3 at 1473 k. Metallurgical
and Materials Transactions B, 31(B):1323–1332.
[JFE, 2010] JFE (2010). Stainless steels. Online, Hibiya Kokusai Building, 2-3
Uchisaiwaicho 2-chome, Chiyodaku, Tokyo 100-0011, Japan.
[Katayama and Tokuda, 1979] Katayama, H. and Tokuda, M. (1979). The reduction
behavior of synthetic chromites by carbon. Tetsu-toHagane, 65(3):331–340.
[Katayama and Iseda, 2002] Katayama, I. and Iseda, A. (2002). Thermodynamic
study of spinel-type solid solutions of the fe3o4-mgfe2o4 coexisting with fe2o3 by
emf method. Scandanavian Journal of Metallurgy, 31:374–378.
[Kekkonen et al., 1995] Kekkonen, M., Xiao, Y., and Holappa, L. (1995). Sula seminar. In Kinetic Study on Solid State Reduction of Chromite Pellets, pages 215–222,
Trondheim, Norway. Finnish Steel and Metal Producers’ Association.
BIBLIOGRAPHY
[Koukkari, 2014] Koukkari, P. (2014).
182
Introduction to constrained Gibbs energy
methods in process and materials research. VTT Technical Research Centre of
Finland, Tekniikantie 4 A, Espoo.
[Lekatou and Walker, 1995] Lekatou, A. and Walker, R. D. (1995). Mechanism of
solid state reduction of chromite concentrate. Ironmaking and Steelmaking, pages
393–404.
[Li et al., 2014] Li, W., Chen, J., Guo, L., Hao, J., Han, P., and Liu, J. (2014).
Electromagnetic properties of high-carbon ferrochrome powders decarburized in
solid phase by microwave heating. Materials Science and Engineering B, (189):58–
63.
[Lyakishev and Gasik, 1998] Lyakishev, N. P. and Gasik, M. I. (1998). Metallurgy
of Chromium. Allerton Press, 150 Fifth Avenue, New York, NY 10011 USA.
[Mazandarany and Pehlke, 1973] Mazandarany, F. N. and Pehlke, R. D. (1973).
Thermodynamic properties of solid alloys of chromium with nickel and iron. Metallurgical Transactions, 4:2067–2076.
[Menendez et al., 2010] Menendez, J. A., Arenillas, A., Fidalgo, B., Fernandez, Y.,
Zubizarreta, L., and Calvo, E. G. (2010). Microwave heating processes involving
carbon materials. Fuel Processing Technology, 91(1):1–8.
[Mosier et al., 2012] Mosier, A. L., Singer, D. A., Moring, B. C., and Galloway, J. P.
(2012). Podiform chromite depositsdatabase and grade and tonnage models. Scientific Investigations Report 2012-5157, U.S. Geological Survey, Reston, Virginia.
BIBLIOGRAPHY
183
[Nafziger et al., 1979] Nafziger, R. H., Tress, J. E., and Paige, J. I. (1979). Carbothermic reduction of domestic chromites. Metallurgical Transactions B, 10(B):5–14.
[Nair et al., 2013] Nair, S., Savio, D., Srinivas, Y. M., and Srinivasan, K. R. (2013).
Manufacture of specialty stainless steels through triplex route at mukand. Advanced Materials Research, 794:90–102.
[NETZSCH, 2012] NETZSCH (2012). Simultaneous Thermal Analysis: Method,
Technique, Applications. NETZSCH-Gerätebau GmbH, Wittelsbacherstraße 42,
95100 Selb, Germany.
[Niayesh and Dippenaar, 1992] Niayesh, M. J. and Dippenaar, R. J. (1992). International ferroalloys congress. In The Solid-state Reduction of Chromite, volume 1,
pages 57–63, Johannesburg, South Africa.
[OECD, 2009] OECD
(2009).
Oecd
steel
committee.
URL:
https://www.oecd.org/sti/ind/43312347.pdf.
[Ontario, 2011] Ontario (2011). Chromite mineralization and exploration in ontario.
URL: http://www.thunderbay.ca/Assets/CEDC/docs/Chromite+Mineralization
+and+Exploration+in+Ontario.pdf.
[Östberg, 2003] Östberg, G. (2003). On the application of the perrin process for
ferro-alloy production. Scandinavian Journal of Metallurgy, 32:157–170.
[Pahlman et al., 1981] Pahlman, J. E., Boucher, M. L., Payne, S. L., and Khalafalla,
S. E. (1981). Method for preparing ferrochromium alloys by vacuum reduction
BIBLIOGRAPHY
184
and simple melting of reduced products. Report of Investigations 8514, Bureau of
Mines, United States Department of the Interior.
[Papp, 2013] Papp, J. F. (2013). Chromium. Technical report, United States Geological Survey.
[Petric and Jacob, 1982] Petric, A. and Jacob, K. T. (1982). Thermodynamic properties of fe3o4-fev2o4 and fe3o4-fecr2o4 spinel solid solutions. Journal of the American Ceramic Society, 65(2):117–123.
[Pickles, 2009] Pickles, C. A. (2009). Microwaves in extractive metallurgy: Part 1.
Minerals Engineering, pages 1102–1111.
[Plaza-González et al., 2005] Plaza-González, P., Monzó-Cabrera, J., Catalá-Civera,
J. M., and Sánchez-Hernández, D. (2005). Effect of mode-stirrer configurations
on dielectric heating performance in multimode microwave applicators. IEEE
Transactions on Microwave Theory and Techniques, 53(5):1699–1706.
[ploufandsplash, 2007] ploufandsplash (2007). Onde electromagnétique.png. URL:
https://commons.wikimedia.org/wiki/File:Onde electromagnétique.png.
[Privalov et al., 2008] Privalov, O. E., Kulinich, V. I., Golovachev, N. P., Suslov,
A. V., and Prokop’ev, L. Y. (2008).
Theoretical characteristics of carbon-
ferrochrome smelting in furnaces produced by aksai ferroalloy plant. Stal, 38(2):40–
43. Translated from Russian.
BIBLIOGRAPHY
185
[Ranganathan, 1998] Ranganathan, S. (1998). Thermochemical analysis of production of ferrochromium in submerged arc furnace. Ironmaking and Steelmaking,
25(6):446–471.
[Relander and Honkaniemi, 1985] Relander, J. and Honkaniemi, M. (1985). Production of ferrochrome from low-grade ores. International Journal of Mineral
Processing, 15:83–88.
[Robiette, 1973] Robiette, A. G. E. (1973). Electric Smelting Processes. John Wiley
& Sons.
[Roine, 2006] Roine, A. (2006). HSC Chemistry 6.0 - User’s Guide. Outokumpu,
Poir, Finland.
[Samarin and Vertman, 1957] Samarin, A. M. and Vertman, A. A. (1957). Production of chromium and carbon free ferrochrome by the method of vacuum sintering.
Tr Inst Met, (1):60–66.
[Schulte et al., 2010] Schulte, R. F., Taylor, R. D., Piatak, N. M., and Seal II, R. R.
(2010). Stratiform chromite deposit model. Open File Report 20101232, U.S.
Geological Survey, Reston, Virginia.
[Schulte et al., 2012] Schulte, R. F., Taylor, R. D., Piatak, N. M., and Seal II,
R. R. (2012). Stratiform chromite deposit model. Scientific Investigations Report 20105070E, U.S. Geological Survey, Reston, Virginia.
BIBLIOGRAPHY
186
[Statnykh et al., 2013] Statnykh, R. N., Ostrovskii, Y. I., Tsernikel, A. A., and
Zhuchkov, V. I. (2013). Improving the production of low-carbon ferrochrome.
Stal, 43(5):309–312. Translated from Russian.
[Thostenson and Chou, 1999] Thostenson, E. T. and Chou, T. W. (1999). Microwave
processing: fundamentals and applications. Composites Part A: Applied Science
and Manufacturing, 30:1055–1071.
[USGS, 2014] USGS (2014). Historical statistics for mineral and material commodities in the united states. URL: http://minerals.usgs.gov/minerals/pubs/historicalstatistics/.
[Vardar et al., 1994] Vardar, E., Eric, R. H., and Letowski, F. K. (1994). Acid leaching of chromite. Minerals Engineering, 7(5/6):605–617.
[Walkiewicz et al., 1988] Walkiewicz, J. W., Kazonich, G., and McGill, S. L. (1988).
Microwave heating characteristics of selected minerals and compounds. Minerals
and Metallurgical Processing, pages 39–42.
[Walkiewicz et al., 1993] Walkiewicz, J. W., Lindroth, D. P., and Clarck, A. E.
(1993). Grindability of taconite rod mill feed enhanced by microwave induced
cracking. In SME Annual Meeting, Reno, Nevada. SME.
[Wang et al., 2014] Wang, Y., Wang, L., Yu, J., and Chou, K. C. (2014). Kinetics of
carbothermic reduction of synthetic chromite. Journal of Mining and Metallurgy,
Section B: Metallurgy, 51(1):15–21.
BIBLIOGRAPHY
187
[Wei and Zhu, 2002] Wei, J. H. and Zhu, D. P. (2002). Mathematical modeling
of the argon-oxygen decarburization refining process of stainless steel: Part i.
mathematical model of the process. Metallurgical and Materials Transactions B,
33(B):111–119.
[Xu et al., 2005] Xu, H. B., Zheng, S. L., Zhang, L., Li, Z. H., and Wang, Z. K.
(2005). Oxidative leaching of a vietnamese chromite ore in highly concentrated
potassium hydroxide aqueous solution at 300◦ c and atmospheric pressure. Minerals
Engineering, 18:527–535.
[Xu et al., 2009] Xu, Y., Chen, Z., and Zhang, G. (2009). Kinetic model of decarburization and denitrogenation in vacuum oxygen decarburization process for ferritic
stainless steel. Metallurgical and Materials Transactions B, 40(B):345–352.
[Zhang et al., 2016] Zhang, B., Shi, P., and Jiang, M. (2016). Advances towards a
clean hydrometallurgical process for chromite. Minerals, 6(7).
[Zhao et al., 2014] Zhao, A., Liu, C., Shi, P., Zhang, B., Jiang, M., Zhang, Q., Saxen,
H., and Zevenhoven, R. (2014). Sulfuric acid leaching of south african chromite.
part 1: Study on leaching behavior. International Journal of Mineral Processing,
130:95–101.
188
Appendix A
Experiments Conducted
Table A.1 shows all main series experiments conducted. Duplicates of several experiments were performed.
Table A.1: All main series experiments performed.
ID
Atmosphere
Carbon (%)
Time (sec)
Input Power (W)
HT1
Argon
0
300
800
HT2
Argon
0
600
800
HT3
Argon
0
900
800
HT4
Argon
0
300
800
HT5
Argon
0
600
800
HT6
Argon
0
900
800
Ar-2.5C-10-800
Argon
2.5
600
800
Ar-5C-10-800
Argon
5
600
800
Ar-7.5C-10-800
Argon
7.5
600
800
Ar-10C-10-800
Argon
10
600
800
189
R-2.5-400
Air
15
150
400
R-5-400
Air
15
300
400
R-10-400
Air
15
600
400
R-2.5-800
Air
15
150
800
R-5-800
Air
15
300
800
R-10-800
Air
15
600
800
R-3.33-1200
Air
15
200
1200
R-6.67-1200
Air
15
400
1200
Vac-2.5-400
Vacuum
15
150
400
Vac-5-400
Vacuum
15
300
400
Vac-10-400
Vacuum
15
600
400
Vac-2.5-800
Vacuum
15
150
800
Vac-5-800
Vacuum
15
300
800
Vac-10-800
Vacuum
15
600
800
Vac-3.33-1200
Vacuum
15
200
1200
Vac-6.67-1200
Vacuum
15
400
1200
Ar-2.5-400
Argon
15
150
400
Ar-5-400
Argon
15
300
400
Ar-10-400
Argon
15
600
400
Ar-2.5-800
Argon
15
150
800
Ar-5-800
Argon
15
300
800
Ar-10-800
Argon
15
600
800
Ar-3.33-1200
Argon
15
200
1200
190
Ar-6.67-1200
Argon
15
400
1200
191
Appendix B
Scanning Electron Microscopy
This Appendix contains additional scanning electron microscope and elemental mapping images of the chromite ore. Included are images of the as-received ore and size
fractioned ore.
192
BSE
Cr
Fe
Mg
Si
Figure B.1: BSE image and elemental mapping of as-received chromite ore.
193
BSE
Cr
Fe
Ca
Si
Figure B.2: BSE image and elemental mapping of as-received chromite ore.
194
BSE
Cr
Fe
Mg
Si
Figure B.3: BSE image and elemental mapping of as-received chromite ore.
195
BSE
Cr
Fe
Mg
Si
Figure B.4: BSE image and elemental mapping of the +74 -105 µm size fraction.
196
BSE
Cr
Fe
Mg
Si
Figure B.5: BSE image and elemental mapping of the +53 -74 µm size fraction.
197
BSE
Cr
Fe
Mg
Si
Figure B.6: BSE image and elemental mapping of the +37 -44 µm size fraction.
198
BSE
Cr
Fe
Mg
Si
Figure B.7: BSE image and elemental mapping of the +25 -37 µm size fraction.
199
Appendix C
Additional Results Data
C.1
Additional Permittivity Data
Additional plots showing the effect of frequency are shown.
C.1. ADDITIONAL PERMITTIVITY DATA
200
Figure C.1: Effect of frequency on loss tangent for chromite ore. Sample density was
3.01 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
201
Figure C.2: Effect of frequency on penetration depth for chromite ore. Sample density was 3.01 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
202
Figure C.3: Effect of frequency on real permittivity for chromite ore with 8% carbon.
Sample was 2.18 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
203
Figure C.4: Effect of frequency on imaginary permittivity for chromite ore with 8%
carbon. Sample was 2.18 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
204
Figure C.5: Effect of frequency on loss tangent for chromite ore with 8% carbon.
Sample density was 2.18 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
205
Figure C.6: Effect of frequency on penetration depth for chromite ore with 8% carbon. Sample density was 2.18 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
206
Figure C.7: Effect of frequency on real permittivity for chromite ore with 15% carbon. Sample density was 2.36 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
207
Figure C.8: Effect of frequency on imaginary permittivity for chromite ore with 15%
carbon. Sample density was 2.36 g/cm3 .
C.1. ADDITIONAL PERMITTIVITY DATA
208
Figure C.9: Effect of frequency on loss tangent for chromite ore with 15% carbon.
Sample density was 2.36 g/cm3 .
C.2. ADDITIONAL HEATING TEST DATA
209
Figure C.10: Effect of frequency on penetration depth for chromite ore with 15%
carbon. Sample density was 2.36 g/cm3 .
C.2
Additional Heating Test Data
Additional plots showing the absorbed powers for the 300 and 600 second tests are
shown.
C.2. ADDITIONAL HEATING TEST DATA
210
Figure C.11: Absorbed power over time plots for the 300 second mode stirrer and non
mode stirrer heating tests. Samples were heated in an argon atmosphere
with 800 W input power. Final temperatures were 129.5◦ C for the non
mode stirrer test and 130.2◦ C for the mode stirrer test.
C.2. ADDITIONAL HEATING TEST DATA
211
Figure C.12: Absorbed power over time plots for the 600 second mode stirrer and non
mode stirrer heating tests. Samples were heated in an argon atmosphere
with 800 W input power. Final temperatures were 211◦ C for the non
mode stirrer test and 574◦ C for the mode stirrer test.
Документ
Категория
Без категории
Просмотров
0
Размер файла
42 573 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа