SCHOOL OF MATERIALS SCIENCE AND ENGINEERING
Carbothermal Reduction of Bauxite
A thesis in
Materials Science and Engineering
by
Chun-Hung Yeh
Submitted in partial fulfillment
of the requirements
for the Degree of
Master of Philosophy
2012
Certificate of Originality
I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, nor material which to a substantial extent has been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis.
I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project design and conception, or in style, presentation and linguistic expression is acknowledged.
------
Chun-Hung Yeh
UNSW Materials Science and Engineering School I
Acknowledgements
Many people assisted me both in the laboratory work and with the written aspects of this project. To these people I express deep appreciation.
Oleg Ostrovski, my supervisor, who always gave me constant encouragement and generous support;
Guangqing Zhang, my co-supervisor, who sparked my interest in this thesis project and who always shared his wealth of knowledge as well as his time;
Yan Li, who assisted with continual guidance and critical assessment of my work;
Xing Xing and Emily Wan for help with X-ray diffraction and LECO analysers, always without delay and with enthusiasm.
Emily Wan, who assisted with electron microscopy and made a good time table with me.
Maggie Zhang, appreciated for all things practical.
I also thank those people whose names have been omitted from this list.
UNSW Materials Science and Engineering School II
Abstract
The commercial technologies for aluminium production include production of alumina from bauxite by the Bayer process and smelting of alumina to produce aluminium by
Hall-Heroult process. The current technology is energy intensive, a major source of greenhouse gas emissions and harmful fluoride emissions. These issues have stimulated the interests in search for alternative technologies of aluminium production.
Carbothermal reduction of alumina is considered a promising alternative technology for aluminium production. However, the carbothermal process in investigation still needs pure alumina which does not solve the problems related to Bayer process, such as generation of harmful red mud, requires very high temperatures, and is overall still an energy intensive process.
This thesis is concerned with the development of an environmentally sustainable technology for aluminium production that will avoid the generation of environmentally negative red mud sludge. It is to investigate the feasibility of stepwise carbothermal reduction of bauxite at different temperatures with emphasis on the examination of the mechanisms and kinetics of reduction of different metal oxides in bauxite, and also on the deportment of impurities among the various phases formed. It is expected that, if successful, this technology will significantly decrease energy consumption and CO2 emissions compared to conventional carbothermal reduction for aluminium production without generation of red mud waste.
Understanding the kinetics and mechanisms of reduction of the different metal oxides in bauxite and the effects of operational parameters is essential for the achievement of the
UNSW Materials Science and Engineering School III
optimal conditions for the production of metallic aluminium and by-products such as ferroalloys and possibly titanium and silicon carbides.
This project investigated the carbothermal reduction of Western Australian and
Queensland bauxites in argon, carbon monoxide and hydrogen atmospheres.
Experiments were performed in a high temperature vertical tube furnace and the off-gas composition was monitored using an infra-red gas analyser. The phase composition of reduced samples was characterized by X-ray diffraction (XRD). Oxygen and carbon contents in reduced samples were determined by LECO analysers. The morphology of the surface and intersections was observed by Scanning Electron Microscopy (SEM).
The chemical compositions of the phases in the reduced samples were also detected by
Energy-dispersive X-ray spectroscopy (EDS).
The results of this study have proved the concept of stepwise reduction of bauxite ores in solid state by appropriate controlling reduction temperature. The products in reduced bauxites by temperature programmed reduction to 1600oC include ferroalloy of silicon and aluminium, carbides of titanium, silicon and aluminium, and unreacted alumina. It also showed that temperatures and the gas environment affected the extent of reduction.
These results are of importance to an explanation of the stepwise carbothermal reaction of bauxites, as well as providing aluminium industry with a better understanding of alternative ways to produce aluminium.
UNSW Materials Science and Engineering School IV
Table of Contents
Section Page
Certificate of Originality ...... I
Acknowledgements ...... II
Abstract ...... III
1 Introduction ...... I
1.1 Importance of Aluminium ...... 1
1.2 Current Aluminium Production Technologies...... 3
1.3 Environmental Issues Associated with Current Production Technologies ...... 4
1.4 Alternative Routes for Aluminium Production ...... 6
2 Literature Review ...... 8
2.1 Introduction to Bauxite, Alumina and Aluminium ...... 8
2.1.1 Bauxite...... 8 2.1.2 Alumina ...... 10 2.1.3 Aluminium ...... 17
2.2 Aluminium Alloy ...... 19
2.3 Metal Oxides Impurities in Bauxite Ores ...... 21
2.4 Carbothermal Reduction of Bauxite...... 22
2.4.1 Carbothermal Reduction of Alumina ...... 23 2.4.2 Carbothermal Reduction of Titania ...... 25 2.4.3 Carbothermal Reduction of Silica ...... 27 2.4.4 Carbothermal Reduction of Iron Oxides ...... 29
2.5 Thermodynamics and Kinetics of Carbothermal Reduction ...... 30
UNSW Materials Science and Engineering School V
2.5.1 Thermodynamics ...... 31 2.5.2 Kinetics ...... 35
2.6 Summary and Scope of the Thesis ...... 38
3 Experimental...... 40
3.1 Materials ...... 43
3.2 Gases ...... 44
3.3 Sample Preparation ...... 44
3.4 Experimental Set Up ...... 45
3.4.1 Experimental Furnace ...... 45 3.4.2 Reactor Set Up ...... 45 3.4.3 Gas System ...... 46
3.5 Experimental Procedure ...... 48
3.6 Sample Characterisation...... 49
3.6.1 X-ray Diffraction Analysis ...... 49 3.6.2 LECO Analyses ...... 50 3.6.3 Scanning Electron Microscopic (SEM) and Energy Dispersive X-ray Spectroscopy (EDS) ...... 51
3.7 Calculation of Extent of Reduction ...... 51
3.7.1 Calculation of Extent of Reduction Based on the Off-gas Analysis ...... 51 3.7.2 Calculation of Extent of Reduction Based on Content of Oxygen in the Reduced Sample ...... 52
4 Results and Discussion ...... 54
4.1 Progress of Carbothermal Reduction ...... 54
4.2 X-Ray Diffraction Analysis ...... 58
4.3 Scanning Electron Microscope Analysis of Reduced Samples ...... 63
UNSW Materials Science and Engineering School VI
4.3.1 Samples Reduced Until a Temperature of 1100oC ...... 63 4.3.2 Samples Reduced Up to 1600°C ...... 70 4.3.3 Segregation of Titanium, Silicon and Aluminium from Alloy Phase ...... 73
4.4 Extent of Reduction in Different Gas Atmospheres ...... 76
5 Conclusions and Recommendations for Future Research ...... 80
5.1 Conclusions ...... 80
5.2 Recommendations for Future Research ...... 81
List of References ...... 82
UNSW Materials Science and Engineering School VII
List of Figures
Number Title Page
1-1 Variation in global production of primary aluminium during
1887–1986 and that of secondary aluminium during 1940–1986 in
the Western world [2] ...... 1
2-1 Structure of crystal alumina [19] ...... 11
2-2 Schematic representation of the Bayer Process depicting its cyclic
nature [23] ...... 12
2-3 Fine white powder of aluminium oxide [24] ...... 13
2-4 Profile of a modern electrolytic cell used in Hall-Heroult process [37]
...... 19
2-5 Schematic of the reaction mechanism of the carbothermal reduction
of Al2O3 ...... 25
2-6 Schematic of the reaction mechanism of the carbothermal reduction
of TiO2 [50] ...... 26
2-7 Schematic of the reaction mechanism of the carbothermal reduction
of SiO2 [54] ...... 29
2-8 Iron phase change and carbon content as a function of temperature
[54] ...... 30
2-9 Illustration of the energy required to produce one kg of aluminium
via the carbothermal reduction of alumina [41] ...... 31
2-10 Standard Gibbs free energy changes of Reactions (21) – (28) ...... 34
2-11 The effect of C/Al2O3 weight ratio on the conversion of alumina at
1500oC [57] ...... 37
UNSW Materials Science and Engineering School VIII
3-1 Schematic diagram of the reaction system ...... 46
3-2 Schematic diagram of the gas supply system ...... 47
3-3 The setting point of a mass flow controller and corresponding gas
flow rate for Ar, CO and H2 ...... 48
4-1 Change in CO content in the off-gas during temperature
programmed reduction of Western Australian and Queensland
bauxite ores ...... 55
4-2 Changes in experimental temperature and CO content in the off gas
in step–by–step reduction experiments ...... 57
4-3 XRD patterns of Western Australian bauxite reduced at different
temperatures ...... 59
4-4 XRD patterns of Queensland bauxite reduced at different
temperatures ...... 60
4-5 XRD patterns of Western Australian bauxite samples reduced at
different steps ...... 61
4-6 XRD patterns of Queensland bauxite samples reduced at different
steps ...... 62
4-7 EDS elemental distribution of Queensland bauxite reduced until
1100oC...... 64
4-8 Point analysis of Queensland bauxite reduced until 1100oC ...... 66
4-9 SEM images and EDS elemental distribution of Western Australian
bauxite reduced until 1100oC ...... 67
4-10 EDS spectra and elemental composition of Western Australian
bauxite reduced until 1100°C ...... 69
4-11 SEM images and elemental distribution of Western Australian
bauxite reduced until 1600oC ...... 71 UNSW Materials Science and Engineering School IX
4-12 EDS spectra and elemental composition of Western Australian
bauxite reduced until 1600°C ...... 72
4-13 SEM images and elemental distribution of Western Australian
bauxite reduced to 1600oC ...... 74
4-14 SEM images and elemental distribution of Queensland bauxite
reduced to 1600oC ...... 75
UNSW Materials Science and Engineering School X
List of Tables
Number Title Page
1-1 Basic physical properties of aluminium [3] ...... 3
1-2 Reported annual production capacity of primary aluminium [8] ...... 4
2-1 Composition of other aluminium containing ores on a mass percent
basis [16] ...... 9
2-2 Geography of Aluminium Production and Consumption in 2000 [17]
...... 15
2-3 XRF analyses of red mud expressed as % oxide [29] ...... 16
2-4 Classification of aluminium alloys [39] ...... 21
3-1 Preliminary experiments of reduction of bauxites to identify
reduction stages ...... 41
3-2 Temperature programmed reduction experiments...... 41
3-3 Step-by-step reduction experiments ...... 42
3-4 Reduction of bauxites under different conditions ...... 42
3-5 Major components of bauxites by XRF, wt% ...... 43
3-6 XRD testing parameters ...... 50
4-1 Extent of reduction of Western Australian and Queensland bauxite
ores reduced in argon at different temperatures ...... 76
4-2 Contribution of different metal oxides, assuming their complete
reduction, to the extent of reduction of the two bauxite ores ...... 77
4-3 Extent of reduction of Western Australian and Queensland bauxites
by ramping from 850oC to different temperatures under CO
atmosphere ...... 78
UNSW Materials Science and Engineering School XI
1 Introduction
1.1 Importance of Aluminium
Aluminium industry plays a very important role in the global economic development and has developed rapidly during the last 100 years. The unique thermal and mechanical properties such as low density, high strength, and good malleability, make them suitable for a wide range of applications. [1] This metal can be formed into different shapes ranging from thick plates (250 mm) to thin foils (0.005 mm). Aluminium alloys can be modified in a range of tempers to achieve required of strength and formability, combined with low density. Modifications of other types of alloy have also been developed to improve their thermal and electrical conductivities. [2]
Figure 1-1 Variation in global production of primary aluminium during 1887–1986 and that of secondary aluminium during 1940–1986 in the Western world [2]
UNSW Materials Science and Engineering School 1
A glance at Figure 1-1 reveals some striking similarities between productions of primary aluminium during the period 1887–1986 and secondary aluminium during
1940–1986. Secondary aluminium is obtained by recycling primary aluminium products.
There was a steady increase in the world’s production of primary aluminium from 1950 to 1980. Aluminium production became 16 times higher in the 1980s than that in the
1890s. Production roughly doubled during the period 1950–1960 and again during
1960–1970. The past few decades have seen further gains. Aluminium production continues to be of significant importance in the 21st century.
Aluminium is a silvery white metal that is insoluble in water. It is very reactive and does not occur as a free element in nature. In nature, it is usually found in mineral forms.
Physical and chemical actions on aluminium–bearing rocks over a long period of time resulted in the formation of aluminous clays that are the basis of all refractory materials.
Further weathering action forms new compounds – for example, bauxite, a general term given to earthy materials with a high content of aluminium oxide.
Projections of future demands for aluminium are higher than that of other metals, which can be attributed to its increased use by the transportation, packaging and construction industries.
Table 1-1 Summarises some of the major physical properties of aluminium.
UNSW Materials Science and Engineering School 2
Table 1-1 Basic physical properties of aluminium [3]
Relative atomic mass 26.9815
o Melting point 660.37 C
Boiling point 2467oC
o 3 Density (at 20 C) 2.7 g/cm
Oxidation states 3
Electronegativity 1.47
Atomic radius 143.2 pm
Percentage mass of the earth’s crust 7.57
Sources from http://www.azom.com/article.aspx?ArticleID=2863
1.2 Current Aluminium Production Technologies
Research on aluminium production routes has increased over the past 50 years. Initially, aluminium was produced in a chemical reduction process, which was expensive and had low productivity. The electrolytic reduction method was established in 1886 which enhanced aluminium production and its subsequent application. [4]
Metallic aluminium is extracted from bauxite ore in a two–step process. In the first step, bauxite is chemically processed into alumina using the Bayer refining process.
Approximately 0.5 to 2.5 tonnes of bauxite are required to produce 1.0 tonne of alumina.
[5] In the second step, the alumina is smelted. During smelting, 1 tonne of aluminium is produced from approximately 2 tonnes of alumina by electrolysis employing the
Hall–Heroult process. This two-step process for producing aluminium is both energy and capital intensive. [6]
UNSW Materials Science and Engineering School 3
A typical aluminium smelter consists of around 300 pots, which can produce 125,000 tonnes of aluminium annually. However, some of the latest generations of smelters can produce aluminiumin the 350,000 – 400,000 tonnes range. [7]
Table 1-2 Reported annual production capacity of primary aluminium [8]
As at the end of: Reported primary aluminium annual production capacity
(thousands of metric tonnes)
Total
December 2010 27,295
June 2011 27,556
June 2012 28,302
June 2013 29,496
June 2014 30,294
The International Aluminium Institute foresees an average annual growth of 1.76% per year in the global aluminium demand for the next 2 years. Total production is currently increasing at a rate of approximately 1.2 million tonnes per annum. [8] These estimates indicate that aluminium demand, and therefore alumina and aluminium production, can as much as triple by early in the 22nd century.
1.3 Environmental Issues Associated with Current Production Technologies
An important point to be considered with regard to the production of aluminium is the major environmental problems associated with its production using Bayer process
UNSW Materials Science and Engineering School 4
followed by electrolytic reduction. Bauxite residue (red mud), whose high alkalinity is harmful to the surrounding environments, is the main by-product generated during aluminium production. At present, the bauxite residues are stored on land or at the bottom of the ocean in containers (marine disposal). However, potential problems of these disposals include leaching of alkaline solution from the container barrier, failure of retaining dams and damage of pipelines leading to leakage of bauxite residue slurry.
High capital investment and large areas of land are required for building bauxite residue dams. [9] Although the production process has existed for many years, only recently has the growing public pressure and environmental awareness prompted the development of solutions to these problems. The enormous quantity of red mud discharged by industries producing aluminium from bauxite poses serious environmental and economical problems. [10, 11] In order to understand these issues and develop better management solutions, it is useful to consider all aspects of aluminium production and its chemistry.
Many companies are embracing the environmental challenge by implementing changes to their production systems. Greenhouse gas emissions from aluminium production by electrolytic process contribute 2.5% of the CO2 emissions worldwide. Much effort has been focused on developing and enhancing the carbothermal reduction of Al2O3 to metallic aluminium, which then reacts with nitrogen to form AlN. Halmann et al. stated that replacement of the electrochemical process by carbothermal reduction of Al2O3 would decrease the total greenhouse gas emissions by at least 30%. [12] A fairly large body of literature exists on the recycling and reuse of bauxite residues, and their environment related issues. However, there is a surprising lack of information on alternative routes to produce aluminium, which can help avoid these environmental problems.
UNSW Materials Science and Engineering School 5
1.4 Alternative Routes for Aluminium Production
In recent years, research is focused on novel alternative routes for aluminium production. Compared with the Hall–Heroult process, carbothermal reduction of alumina offers merits of a simpler process, lower cost and lower requirement for raw materials. [13] Previous assessments showed that carbothermal reduction has the potential to reduce energy consumption by up to 38%, capital costs by more than 60%, overall operating costs by 25-30% and CO2 emissions by up to 30%, without any fluoride emission. [14]
The present research focuses on carbothermal reduction of bauxite with the major aim of investigating the effect of the gas atmosphere on the reduction kinetics. The proposed method involves two steps. Firstly, bauxite is reduced to a ferroalloy of titanium, silicon and aluminium as well as to their respective carbides. The reduced product is further heated to decompose aluminium carbide under vacuum or a suitable gas atmosphere to obtain aluminium vapour, which after condensation produces pure aluminium. Oxides and carbides of iron, silicon and titanium are separated out after the reduction of bauxite.
The relatively pure alumina obtained from residues can be further reduced to produce metallic aluminium.
This research thesis encompasses reduction experiments and subsequent characterisation of the reduced samples with the aim of understanding and enhancing the efficiency of the carbothermal reduction for aluminium production, which has the potential to lower the adverse impact on the environment.
UNSW Materials Science and Engineering School 6
This study may lead to a better understanding of aluminium production with reduced negative effect on the environment. Experimental results are of great interest for both applied and scientific research. This study may also be important for better understanding of bauxite reduction processes.
The thesis is structured as follows. Chapter 2 includes a survey of the published literature with some background information. It establishes a foundation for the current research and includes a statement of the specific research aims. Chapter 3 describes the experimental procedures and techniques employed in the characterization analysis of samples. Major experimental results, analysis and discussion are presented in Chapter 4.
Chapter 5 summarises the main conclusions of the research and recommendations are made for further research.
UNSW Materials Science and Engineering School 7
2 Literature Review
2.1 Introduction to Bauxite, Alumina and Aluminium
The present method for production of aluminium was discovered nearly simultaneously and completely independently in 1886 by Paul-Louis Heroult in France and Charles
Martin Hall in the United States.
Alumina or aluminium oxide is made from bauxite via the Bayer process followed by smelting using the Hall–Heroult process to produce primary aluminium. The Bayer process produces pure alumina from bauxite ore, and the Hall–Heroult process produces aluminium from alumina. The primary aluminium produced is 99.5–99.8 % pure. [15]
2.1.1 Bauxite
Aluminium begins as an ore, bauxite, which contains high concentrations of aluminium hydroxide minerals (40-60%) but there are many other ores from which aluminium can be manufactured. Some of the most common are alunite, kaolinite, illite, dawsonite and anorthosite. Compared with other ores, bauxite is the richest of the aluminium bearing ores and also contains the smallest percentage of impurities. [16] Table 2-1 gives the mass % of alumina in the major forms of aluminous containing materials.
UNSW Materials Science and Engineering School 8
Table 2-1 Composition of other aluminium containing ores on a mass percent basis [16]
Bauxite is converted to alumina via the Bayer process. Approximately 98% of primary aluminium production is based on bauxite. [17] Bauxite is a naturally occurring, heterogeneous material composed primarily of one and more aluminium hydroxide minerals plus different types of mixtures of silica, iron oxide, titania and other impurities.
Bauxite results from the decay and weathering of aluminium bearing rocks, often igneous but not necessarily so. It may form residual deposits replacing the original rock.
Large amounts of bauxite can be discovered in tropical and subtropical areas, for example, the West Indies, South America and Australia. [18]
Australia has huge reserves of bauxite. In 2000 it was the leading worldwide bauxite producer with a production share of 39%. Second in terms of bauxite production was
Guinea which has 13%, followed by Brazil (10%) and Jamaica (8%) (Table 2-2).
Generally the bauxite is extracted by open cast working methods. It is then washed and
UNSW Materials Science and Engineering School 9
screened to remove extraneous dirt. The lumps of washed ore may be treated locally at the mining place for the production of alumina, or the raw materials will be delivered to locations for processing the using Bayer process.
2.1.2 Alumina
The chemical formula for alumina is Al2O3. It is a ceramic oxide material that has been used as a biomedical implant material since the early 1970s. [19] It represents the intermediate stage between the naturally occurring ore, bauxite and metallic aluminium.
In material science and related fields, Al2O3 can refer to one of many chemical compounds including amorphous hydrous and anhydrous oxides, crystalline hydroxides and oxides, and alumina materials containing small amounts of alkali or alkali earth oxides. Thermodynamically, the most stable form of Al2O3 is the hexagonal crystal structure. The usual type of alumina encountered in biomedical engineering applications is either polycrystalline form (alumina) or single crystal form (sapphire). [20]
A number of solids, especially fluorides and oxides, possess a high melting point and are strongly bonded, as the bonds are intermediate ionic and covalent, the stability of the solids is due to the formation of huge molecules with uniform bonding. [21]
UNSW Materials Science and Engineering School 10
Figure 2-1 Structure of crystal alumina [19]
The aluminium and oxygen valence electrons are bonded in a complex manner by covalent homopolar bonding. Only Al3+ and O2-are present. There are no other atoms.
[22]
Alumina refining is a complex chemical process. The alumina must be purified before it can be refined to aluminium metal. Commercially, alumina is produced by the Bayer process, illustrated in Figure 2-2.
UNSW Materials Science and Engineering School 11
Figure 2-2 Schematic representation of the Bayer Process depicting its cyclic nature [23]
Mined bauxite ore is finely ground then crushed to a fine powder before chemical treatment. It is digested in caustic soda, NaOH, at high temperature (~250oC) and pressure (~3 atm) to form water soluble aluminium hydroxide. The aluminium hydroxide is then precipitated from the soda solution, washed and dried while the soda solution is recycled. The overall reactions are as follows:
- - Al2O3 (in bauxite) + 2OH + 3H2O → 2[Al(OH)] (1)
Impurities are removed via setting or filtration followed by cooling of the solution to room temperature in precipitator tanks where agitation with seed crystals generates the precipitation of alumina in a hydrated form. Subsequently it is collected, washed free of
UNSW Materials Science and Engineering School 12
caustic, dried, and lastly calcined at temperature up to ~1100oC in rotary kilns to produce alumina, a white powder.
By calcination, aluminium hydroxide loses water to form alumina:
2 Al(OH)3 → Al2O3 + 3 H2O (2)
After calcination the end product, aluminium oxide, is in a white powder form, shown below.
Figure 2-3 Fine white powder of aluminium oxide [24] Source from http://www.e-reful.com/products/white-fused-alumina.html
In China most of the local bauxite ores contain a high proportion of Si and therefore a combined Bayer process and bauxite calcination method is used for alumina refining.
[25] The Bayer process is capable of producing large quantities of powder inexpensively and with high purity. Raw material is readily available and the price is relatively stable in the market. [26]
UNSW Materials Science and Engineering School 13
Alumina can occur in the hydrate forms as trihydrate Al2O3．3 H2O (gibbsite) or the monohydrate Al2O3．H2O (boehmite) because of nature tendencies and the bauxite can contain either of these forms or a mixture. During the extraction of the alumina from bauxite with caustic soda, silica separates out with other insoluble residue producing red mud slurry. [27] The mud is diluted and sent for filter pressing. It is then pumped away as slurry into large man made lagoons. The by-product, red mud, still has no commercial use.
Table 2-2 shows a number of clear differences among countries with respected to bauxite, alumina and aluminium production. In 2000, Australia was the largest supplier globally of both bauxite and alumina. Australian alumina comprised some 31% of the
50.9 million tonnes of alumina produced worldwide. On the other hand, United States,
China and Jamaica produced 9%, 8% and 7% respectively.
The activity of primary industries often yields substantial amounts of by–products. Red mud is a by-product in the processing of bauxite to alumina through the Bayer process.
Its chemical and physical characteristics rely on the origin and treatment of the bauxite.
Red mud is composed largely of Fe and Ti oxides, behaving as chemically inert matter with variable percentages of nominal SiO2, Al2O3 and Na2O. The material is typically available as a watery mixture which settles slowly and may easily be conveyed from station to station by continuous fluid carrying machinery. [11]
UNSW Materials Science and Engineering School 14
Table 2-2 Geography of Aluminium Production and Consumption in 2000 [17]
Primary production (%)
Bauxite Alumina Primary Aluminium
Europe 9 22 33
France 0 1 2
Germany 0 2 3
Asia 13 11 21
China 6 8 12
India 5 0 3
Africa 13 1 5
Guinea 13 1 0
South Africa 0 0 3
America 26 33 34
Brazil 10 7 5
Canada 0 2 10
Jamaica 8 7 0
United States 0 9 15
Australia 39 31 7
Chemical analyses show that the characteristics of chromium, lead and sodium in the red mud are different from those normally found in sediments. The pH value of the liquid in red mud is about 12.8. The reason is that the presence of caustic remaining after Bayer processing. Dethlefsen and Rosenthal [28] found Al3+, Fe2+, Ca2+, Cl- and S2- in red mud and also measured the pH of the sea water of pH 0.2 after dumping it.
Consequently red mud is viewed as a corrosive substance requiring careful handling.
UNSW Materials Science and Engineering School 15
Table 2-3 XRF analyses of red mud expressed as % oxide [29]
Country Fe2O3 TiO2 CaO Al2O3
Brazil 45.6 4.3 15.6 15.1
Germany 44.8 12.3 5.4 16.2
Italy 15.2 6.2 18.6 24.7
Spain 37.5 11.5 4.4 21.2
USA 35.5 6.3 8.5 18.4
Australia [30] 47.0 5.0 3.7 20.2
Table 2-3 gives the information about the average percentage of oxides in red mud from different countries.
The major industry environmental impact is the large amounts of red mud waste at alumina production. There are various forms of waste materials in the alumina production processes which include sludge and solid residues. The exact composition of red mud depends on the source of the bauxite and its treatment during production.
Depending on the source of the bauxite, between 1.0 tons to 1.5 tonnes of red mud are produced per ton of alumina. [31] The traditional storage in nearby dumps can be impractical owing to the considerable masses involved and environmental restrictions.
[32] The previous findings on red mud toxicity are attributed to aluminium and iron being present in mud at high concentrations. [33] Other reports also focused on Al(III) and Fe(III) associated toxicity as well as the complex mixtures and their type of salts.
[34]
Several laboratory experiments were carried out to assess the effects of red mud on
UNSW Materials Science and Engineering School 16
marine organisms, by Dethlefsen and Rosenthal. [28] The results showed that red mud is harmful to a variety of organisms and the most predominant effect was agglutination of gill tissues. Dumping red mud in the sea would endanger the stocks of bottom ocean fauna and should therefore take place in areas without water life at depths greater than
3000 m. Liu et al. [9] indicated that environmental concerns have motivated the Bayer process industry in the west to increasingly adopt processes to neutralise the red mud before disposal.
2.1.3 Aluminium
Aluminum is a silvery white coloured metal having high reflectivity of light and heat.
Aluminium has an electrical conductivity high enough to permit its use as an electrical conductor. [35] Impurities such as titanium, vanadium and chromium have a negative effect on conductivity. Aluminium, normally combined with sufficient amounts of iron and silicon, has improved tensile strength and conductivity.
A flat clean surface of pure aluminium reflects approximately 80 to 85% incident visible radiation which has led to its wide use in lighting fixtures. [21] A flat aluminium surface absorbs less heat than other metals such as copper and steel. This property can be applied in the construction of various types of light and heat reflectors: for instance houses with roofs of this material are cooler in summer.
The most important applications for aluminium include building and construction, such as roof constructions, offshore installations, stairways, window frames and doors;
UNSW Materials Science and Engineering School 17
transportation, for example in airplanes, cars, trains, ships; packaging for food or beverages; and electrical technology, for instance, power transmission lines, transformers and panels. [36] The automotive industries prefer using aluminium where possible due to its light weight which improves fuel consumption and decreases pollution emissions.
In the Hall-Heroult process the starting materials are alumina, electrical energy and carbon. Smelting is a process of electrolytic reduction in a molten bath of natural and
o synthetic cryolite (Na3AlF6) to convert alumina to aluminium at 960 C. Normally it has a range of purity between 99.5 and 99.8 percent aluminium. The two main impurities are iron and silicon. [15]
The Hall–Heroult process is an electrolysis process so aluminium smelters use a prodigious amount of electricity. This is done in batteries of electric furnaces called reduction cells or pots. The reduction cells are shallow steel tanks lined with carbon.
Smelting is run as a batch process with the aluminium metal deposited at the bottom of the pots and periodically drained off. Figure 2-4 presents a schematic of an electrolytic cell for aluminium smelting.
The anodes are consumed during the process when they react with the oxygen from the alumina. Oxygen in the alumina separates and combines with the carbon from the anodes. Under constant power carbon dioxide is produced leaving molten aluminium to collect at the bottom of the cells.
The reaction is based on electrolysis of alumina in molten cryolite, the only medium into which alumina reasonably dissolves. Upon the flow of current, the following UNSW Materials Science and Engineering School 18
overall chemical reaction takes place for the production of primary aluminium:
2Al2O3 (dissolved) + 3C (solid, anode) → 4Al (molten) + 3CO2 (gas) (3)
Figure 2-4 Profile of a modern electrolytic cell used in Hall-Heroult process [37]
1. Steel shell; 2. Insulation; 3. Cathode; 4. Iron cathode bar; 5. Cathodic curry entry;
6. Liquid aluminium; 7. Electrolyte; 8. Prebaked anodes; 9. Steel nipple; 10. Anode beam;
11. Aluminium oxide hopper; 12. Gas suction unit; 13. Detachable covers; 14. Crust crusher.
2.2 Aluminium Alloy
The alloys of aluminium are generally of similar colour, some with a bluish tinge. The pure aluminium produced through the methods explained previously is a relatively weak material. Much larger increases in strength can be obtained by alloying it with small percentages of other metals such as copper, magnesium, manganese and zinc, usually in
UNSW Materials Science and Engineering School 19
combinations of two or more of these elements together with iron and silicon. Some of the alloys are further strengthened and hardened by heat treatments so that today aluminium alloys possessing tensile strengths approaching 100,000 pounds per square inch are available. [38] When the aluminium is applied in the construction or industrial fields the greater mechanical strength allows it to meet the requirements of a wide range of situations. The molten aluminium is transported to the cast house where it is alloyed in holding furnaces by the addition of other metals. Subsequently the metal is cast into ingots, rolled ingots or extrusions, depending on the proposed use for the aluminium or aluminium alloys. [7]
The molten aluminium is transported to the cast house where it is alloyed in holding furnaces by the addition of other metals. Subsequently the metal is cast into ingots, rolled ingots or extrusions, depending on the proposed use for the aluminium or aluminium alloys. [7] The alloys can be considered as three main groups: ingot derived from remelting, shaped casting; and mechanically worked (wrought) products. Heat treatable alloys assume the various forms by means of mechanical working. The material can be shaped into plates, sheets, foil, bars, extrusions, hollow sections, forging stock, forgings, tubes, wires, rivets and solid conductors. [35]
The alloys for castings will be referred to using the BS 1490 LM designations and the internationally agreed four digit system is used for wrought alloys. The first of the four digits in the designation demonstrates the major alloying element of alloys within the group, as Table 2-4.
UNSW Materials Science and Engineering School 20
Table 2-4 Classification of aluminium alloys [39]
1XXX Aluminium of 99.00% minimum purity
2XXX Copper
3XXX Manganese
4XXX Silicon
5XXX Magnesium
6XXX Magnesium plus silicon
7XXX Zinc
8XXX Other elements
9XXX Unused series
The increasing demands of engine makers led to major research projects to further improve aluminium alloys and in turn the resultant stronger materials able to withstand higher temperatures led to further new applications.
The first digit is representing the main alloy element in aluminium alloys. Second two digits indicate the different aluminium alloys (impurity or the adding special alloy modification). Last digit demonstrates the product forms.
2.3 Metal Oxides Impurities in Bauxite Ores
The differentiation of bauxites is mainly in terms of the age of the deposits and the weathering conditions to which they were exposed. Potassium (K), sodium (Na), UNSW Materials Science and Engineering School 21
manganese (Mg), and calcium (Ca) are the common impurities in bauxite. However, there are four main oxides, silicon dioxide (SiO2), titanium dioxide (TiO2), iron oxide
(Fe2O3) and aluminium oxide (Al2O3) occupying the large percentage in bauxite.
Following sections are only focused on these main four oxides.
2.4 Carbothermal Reduction of Bauxite
In order to achieve sustainable production it is very important to find technological alternatives to overcome the environmental problems, caused during alumina production. Few studies have been published on carbothermal reduction of bauxite.
Cochran [40] estimated that the costs for any new process would have to be almost half of the average Bayer/Hall-Heroult process to allow profitable operation through the lows in the price cycle for aluminium. Murray [41] provided extensive discussion of the applications which use high temperature solar process to replace the traditional methods of producing aluminium. Li et al [42], demonstrated that alumina can be reduced to aluminium carbide in a hydrogen atmosphere at temperatures as low as 1600oC to
1700oC. The carbothermal reaction strongly depends on the homogeneity of the bauxite-carbon mixture.
In the carbothermal reduction processes organic liquids are commonly used to homogenize and bind the starting powders with the carbon source. [43] Most of the alumina will be reduced to aluminium carbide due to the high content of alumina in the bauxite ore material and the relatively low solubility of aluminium in the molten phase.
Halmann et al. [12] aptly pointed out that utilising alumina, methane and oxygen as reagents, the coproduction of aluminium with syngas, to be converted to methanol, yields likely fuel savings of about 68% and CO2 emission decreases of about 91%.
UNSW Materials Science and Engineering School 22
Moreover there is a fuel saving of fuel saved of 66% and CO2 emission avoidance of
15% when carbon is used as reducing agent.
An excess of carbon is generally required in carbothermal reduction in order to:
(a) Increase the reaction rate;
(b) Help to complete the transformation;
(c) Improve the dispersion of the powder; and
(d) Control the powder aggregation.
Bauxites include four main components which are alumina, titania, silica and iron oxides. Following sections are going to introduce the situation for each oxides reduced by carbothermal reduction.
2.4.1 Carbothermal Reduction of Alumina
The overall reaction of carbothermal reduction for converting alumina to metallic aluminium is listed as below:
Al2O3 + 3C → 2Al +3CO (4)
This reaction occurs in two stages. The first is the reduction from alumina to aluminium carbide, Al4C3, shown as reaction (5). At the second stage, the Al4C3 reacts with excess alumina and metallic aluminium is formed according to reaction (6).
2Al2O3 + 9C →Al4C3 + 6CO (5)
UNSW Materials Science and Engineering School 23
Al4C3 +Al2O3 → 6Al +3CO (6)
Yuan et al. [4] reported that at the ALCOA Corporation a charge of Al2O3 and C was inserted into a high temperature upper reaction zone to form liquid mixtures of Al2O3 and Al4C3 which was then transferred to a lower reaction zone for the extraction of liquid Al.
Wang et al. [44] used bauxite as raw material of alumina, coal as reducing agent and anhydrous – AlCl3 as chloridising agent. The overall reaction can be represented by
Al2O3 +3C +AlCl3 → 3AlCl + 3CO (7)
3AlCl → 2Al +AlCl3 (8)
Both reactions are complicated by the formation of aluminium carbide, and oxycarbides in the carbothermal reduction process. Another issue is that the impurities from raw materials have uncertain effects on the carbothermic-chlorination processes.
Li et al. [45] mentioned that the internal mass transfer in the gas phase plays a major role in the reduction kinetics in a reaction cycle. Al vapour reacts with Al2O3 to form gaseous Al2O; the latter diffuses to graphite and reacts with carbon regenerating Al vapour; then the Al vapour diffuses to Al2O3 phase (Figure 2-5).
They also found that the aluminium carbide in hydrogen and helium was close to completion in 180 min at 1600oC and in 60 min at 1700oC. In argon a higher temperature and longer reaction time are required.
UNSW Materials Science and Engineering School 24
Carbothermal reduction for aluminium production has been found to be compatible with the induction shaft furnace electrochemical dissociation process [46] and the high temperature reduction-aluminium scrap cooling process. [47]
III
I II
Figure 2-5 Schematic of the reaction mechanism of the carbothermal reduction of Al2O3
o I. Al2O3 + 4Al(g) → 3Al2O(g), △G = -113.1 - 0.0084T (kJ)
o II. Al2O + C →2Al(g) + CO, △G = 689-0.264T (kJ)
o III. 2Al2O(g) + 5C → Al4C3 + 2CO, △G = -124 + 0.016T (kJ)
2.4.2 Carbothermal Reduction of Titania
Titanium carbide powder can be produced from titanium oxide from carbothermal reduction. Sen et al. [48] indicated that carbothermal reduction of titanium dioxide using a vacuum condition, with charcoal and plentiful TiC, can be found at 1550oC.
The overall carbothermal reduction reaction for titanium dioxide is:
TiO2 + 3C → TiC + 2CO (9)
UNSW Materials Science and Engineering School 25
This reaction proceeds thermodynamically at 1289oC. It can be separated into three parts:
4TiO2 + C → Ti4O7 + CO (10)
3Ti4O7 + C → 4Ti3O5 + CO (11)
Ti3O5 + 8C → 3TiC + 5CO (12)
In contrast, Woo et al. [49] strongly believe that the formation of TiCxOy was completed in a specimen which was reacted at 1500oC with carbon for five minutes. After five minutes the TiCxOy lost more oxygen as the purification of TiC proceeded. The similar conclusion also pointed out by Berger et al. [50] There are three reaction steps in the carbothermal reduction of titanium dioxide, schematically shown in Figure 2-6.
Figure 2-6 Schematic of the reaction mechanism of the carbothermal reduction of TiO2 [50]
UNSW Materials Science and Engineering School 26
The steps involved in the reduction are as follows:
I. CO, which is formed at the very beginning by a solid state reaction between the
oxide and carbon particles or by destruction of oxygen containing functional
groups at the carbon surface, acts as a reduction agent, resulting in the formation
of lower oxides (TinO2n-1).
II. CO acts as a reduction agent again and simultaneously is disproportionate at the
surface of TinO2n-1. The surface is rich in lattice defects which accommodates the
incorporation of carbon into the crystal lattice. CO2 is generated and convert to
CO on the solid carbon immediately. According to (CO2 + C → 2CO). Hence the
oxide particles will be the precursors for the oxycarbide formed.
III. Substitution of oxygen by carbon in TiCxOy. The mass transfer is realised by the
same reaction mechanism as in reaction step II.
2.4.3 Carbothermal Reduction of Silica
The carbothermal reduction reaction of SiO2 to the overall reaction is as follows:
SiO2 + C → SiO + CO (13)
SiO2 + CO → SiO + CO2 (14)
Reaction (13) includes a multiple step process. The first step is SiO2 reduction by carbon forming gaseous SiO with CO. Again CO can also be the reduction agent for
SiO2 and SiO can also be generated according to the above reaction (14).
UNSW Materials Science and Engineering School 27
CO2 + C → 2CO (15)
SiO + 2C → SiC + CO (16)
SiO + 3CO → SiC +CO2 (17)
Any created CO2 will react with the surrounding carbon immediately to form CO gas, followed by the gaseous SiO reacting with C and CO. CO2 from reaction (17) is followed by reaction (15) to synthesis CO, and the CO from reaction (15) can be utilised for reaction (14) where formed gas SiO and according to reaction (17), so that it becomes a cycle. [51] The generation rate of CO gas increases with rising temperature.
[52]
Weimer et al. [53] proposed a mechanism for the carbothermal reduction synthesis of
SiC. The schematic and sequence of reactions involved are shown in Figure 2-7.
I C(s) + SiO2(s) → SiO(g) + CO(g)
II SiO2(s) + CO(g) → SiO(g) + CO2(g)
III C(s) + CO2(g) → 2CO(g)
IV 2C(s) + SiO(g) → SiC(s) +CO(g)
By this mechanism, SiO is initially formed at the contact points of the carbon and silica particles (I). In reactions steps gaseous SiO as an intermediate product is formed and reacts as the surface of carbon particles (IV). SiC crystallites resembling the starting carbon crystallites in size and morphology prior to gain growth. Therefore the mass is mainly transferred from the starting SiO2 to the carbon particle which is the precursor of the SiC.
UNSW Materials Science and Engineering School 28
Figure 2-7 Schematic of the reaction mechanism of the carbothermal reduction of SiO2 [54]
2.4.4 Carbothermal Reduction of Iron Oxides
There are three forms of iron oxide: hematite (Fe2O3), magnetite (Fe3O4) and wustite
(FeO). These oxides are reduced in stages.
3Fe2O3 + CO → 2Fe3O4 + CO2 (18)
Fe3O4 + CO → 3FeO + CO2 (19)
FeO + CO → Fe + CO2 (20)
Figure 2-8 shows the quantitative iron phase and carbon percentages as a function of temperature. The analysis by Liu et al. [54] showed that when the Fe3+ started to reduce at 450oC the reduction was almost complete at 870oC. The percentage of Fe2+ climbed from 450oC, reaching the top point at 870oC and then falling with further heating up to
1100oC. The metallic iron rapidly increased from 870oC and reached up to 98.7% at
1200oC. They also demonstrated that the carbon reduced rapidly at temperatures above
800oC due to carbon gasification. This research helps to understand the reduction process of iron oxide from 3+ state to metallic iron in bauxite and carbon start to consume generally when the temperature increases.
UNSW Materials Science and Engineering School 29
Figure 2-8 Iron phase change and carbon content as a function of temperature [54]
2.5 Thermodynamics and Kinetics of Carbothermal Reduction
The carbothermal method for production of aluminium has been the subject of much research effort for almost half century, and has involved substantial investments by the major aluminium producers. In stark contrast, however, its development has not reached the industrial level yet.
During carbothermal reduction, the alumina and carbon fraction decrease generally with increasing reaction temperature. The carbothermal reaction strongly depends on the homogeneity of the alumina and carbon mixture. Furthermore, compressing the mixture and the eventual incorporation of a carbon generating binder can extend the solid-solid
UNSW Materials Science and Engineering School 30
interface. Alumina and carbon enable to be the most suitable starting mixture by chosen on the basis of such different criteria as purity, fineness, narrow particle size distribution, uniform particle shape, cost, and without hard agglomeration. [55]
2.5.1 Thermodynamics
The production of aluminium is an energy intensive process. The enthalpy change (ΔH) is the total amount of energy required in the reaction. Moreover an amount equal to the
Gibbs free energy change (ΔG) of the reaction is required, as high quality energy which is expensive. [56]
Figure 2-9 Illustration of the energy required to produce one kg of aluminium via the carbothermal reduction of alumina [41]
Figure 2-9 shows that while the total energy required remains fairly constant with increasing temperature, the amount of high quality (expensive energy) that must be
UNSW Materials Science and Engineering School 31
added decreases linearly.
Bauxite contains mainly hematite, alumina, silica and titania. Carbothermal reduction of these oxides and related Gibbs free energy changes and equilibrium temperature under standard condition are listed below:
Fe2O3 + 3C → 2Fe + 3CO,
△Go = 478.54 – 0.516T (kJ) (21)
Teq = 654 °C
Al2O3 +3C → 2Al + 3CO,
△Go = 1269.13 – 0.547T (kJ) (22)
Teq = 2047 °C
SiO2 + 2C = Si + 2CO,
△Go = 722.90 – 0.373T (kJ) (23)
Teq = 1665 °C
TiO2 + 2C → Ti + 2CO,
△Go = 708.49–0.348T (kJ) (24)
Teq = 1765 °C
From above equations, metal iron can be easily formed by carbothermal reduction of
Fe2O3. Formation of metallic aluminium, silicon, and titanium is much more difficult, with equilibrium temperatures under standard conditions of 2047, 1665 and 1765°C.
Even though, the formation of these metals is possible in experiments if they are UNSW Materials Science and Engineering School 32
dissolved in iron to form a ferroalloy. Another factor favouring their formation is that the partial pressure of CO in experiments can be significantly lower than 1 atm when the experiments are carried out in the atmosphere of another gas such as argon or hydrogen.
Because aluminium, titanium and silicon are highly reactive metals, carbothermal reduction of their pure oxides usually forms corresponding carbides:
Al2O3 + (9/4)C → (1/2)Al4C3 + 3CO,
△Go = 1198.88 – 0.528T (kJ) (25)
Teq = 1997 °C
SiO2 +3C → SiC + 2CO,
△Go = 589.19 – 0.327T (kJ) (26)
Teq = 1529 °C
TiO2 + 3C → TiC + 2CO,
△Go = 524.13 – 0.334T (kJ) (27)
Teq = 1569 °C
Carbothermal reduction of bauxite ores in this investigation is more complex than reduction of individual oxides, and the final products can be complex too, which needs careful characterisation. Furthermore, reduction of Fe2O3 and TiO2 can proceed in multiple steps, which makes the thermodynamics more complex. However, this complexity does not change the overall conclusions of above thermodynamic analysis.
UNSW Materials Science and Engineering School 33
1000
900 Eq.21 (Fe) Eq.22 (Al)
800 Eq.23 (Si) Eq.24 (Ti) Eq.25 (Al4C3) Eq.26 (SiC) 700
Eq.27 (TiC) Eq.28 (CO) 600
500
400
300
200
100
0
-100 Standard Gibbs free energy change Standard (kJ) -200
-300
-400
-500 500 700 900 1100 1300 1500 1700 1900 2100 2300 2500 Temperature (K)
Figure 2-10 Standard Gibbs free energy changes of Reactions (21) – (28)
Figure 2-10 summarises the standard Gibbs free energy changes of Reactions (21) –
(27). It is clear that hematite is the easiest to be reduced to metallic iron. The tendency to reduce titania and silica to titanium and silicon is very close, which is more difficult than reducing to their carbides. Reduction of alumina to metallic aluminium has the highest standard Gibbs free energy change, and needs the highest temperature to proceed. Relatively, formation of aluminium carbide is slightly easier than reduction to metallic aluminium, and so is the preferred product. However, as stated previously, ferroalloy of silicon, titanium and aluminium can be formed due to existence of metallic iron which will be saturated with carbon and melted above 1153 °C.
UNSW Materials Science and Engineering School 34
It is also strongly believed that the carbothermal reduction of solid oxides proceeds through the gas phase, when an oxide is reduced by CO which is regenerated by the
Boudouard reaction. [56]
C(s) + CO2(g)→ 2CO(g),
△Go = 160.77 – 0.168T (kJ) (28)
The standard Gibbs free energy change of Reaction (28) is also included in Figure 2-10.
Boudouard reaction is favoured at high temperatures. At lower reduction temperatures
CO2 is the favourite product, whereas CO is formed in higher temperature reactions.
2.5.2 Kinetics
Factors affecting the kinetics of carbothermal reduction of bauxite ores include ore composition, temperature, gas atmosphere, carbon to oxide ratio, particles size, gas flow rate, etc.
As demonstrated in the thermodynamic analysis, reduction of titanium, silicon and aluminium oxides from bauxite ores needs high temperature. Besides, high temperature enhances reaction kinetics according to Arrhenius law. However, direct heating of bauxite ores to a high temperature will melt the ores to form a slag phase of metal oxides, which decreases the activities of the metal oxides, and increases the resistance of the metal oxides to diffuse to the carbon phase for a reduction reaction to take place.
That is why this investigation reduces the bauxite ores in steps so that formation of a slag phase will be avoided, and the oxides are reduced in solid phase.
UNSW Materials Science and Engineering School 35
Gas atmosphere can affect the reduction kinetics in different ways. Firstly, the reaction of carbon and metal oxides in separate phases is through gas phase. This includes CO as a reaction intermediate, as previously mentioned. For the reduction of silica, vapour SiO plays an important role, while Al and Al2O vapours are involved in the reduction of alumina. Different gas atmosphere can change the diffusivity of above gas species in the gas phase, so as to change the mass transfer rate between two phases and so the reaction kinetics. When CO is used, it will promote the reduction of hematite as a reductant.
However, its high partial pressure will suppress the reduction of alumina, silica and titania thermodynamically. Using of hydrogen is expected to enhance the reaction kinetics, because it is a reductant, can increase the gas diffusivity, and may promote the transfer of carbon to the oxide phase by formation of methane.
High carbon content in the solid sample can provide more surface area for reaction with
Al2O3 and Al2O as well as accelerating nucleation. Consequently faster reduction can be reached. Figure 2-11 shows that the higher the weight ratio of C/Al2O3, the faster the reaction. [57]
Using carbon of finer particle size can achieve similar effect. Same amount of carbon with small particle size possesses large surface area and accelerates the reaction. When the particle size is small, the distance of the diffusion of reductive gas in the solid matrix is small, the intermediate gas species can diffuse easily from the reaction surface of a particle to another, and the reaction is accelerated.
UNSW Materials Science and Engineering School 36
o Figure 2-11 The effect of C/Al2O3 weight ratio on the conversion of alumina at 1500 C [57]
The reduction reactions can be beneficiated by mixing starting materials homogeneously.
Furthermore, compressing the mixture and eventual incorporation of a carbon generating binder can extend the solid-solid interface. The most suitable raw materials, bauxite and carbon, can be chosen on the basis of such different criteria as purity, fineness, narrow particle size distribution, uniform particle shape, cost, and with no hard agglomeration.
A higher gas flow rate in reduction experiments may favour reduction because it will reduce the partial pressure of CO, besides reducing external mass transfer resistance.
The reason is that the pressure of inlet gas is increased; simultaneously, the resistance of the gas film mass transfer is reduced. Motlagh [58] found that the rates of reduction are directly related to the gas flow rate from his direct reduction of Australian lump iron ore experiment.
UNSW Materials Science and Engineering School 37
When the reduction temperature is high enough, ferroalloy of titanium, silicon and aluminium will be produced. Fused iron may promote reduction by enhancing carbon transfer through the molten phase. However, too much molten phase will obstruct diffusion of CO generated in reduction and retard further reaction.
2.6 Summary and Scope of the Thesis
This review of academic literature shows that there is much interest in the discovery and development of reduction bauxite. Novel manufacture methods and cost effective carbide of silicon and titanium, alloys and aluminium are of particularly high appeal.
The proposed project involves the development of an environmentally sustainable technology for aluminium production that will avoid the generation of red mud sludge.
This project investigates stepwise carbothermal reduction of bauxite at different temperature levels. The ultimate aim of this project is to establish the fundamentals for a sustainable technology for aluminium production from bauxite ore, together with the production of valuable by-products in the process. The difference between this project and previous works is that the raw material, bauxite, will be used directly. Based on analysis of simple model systems, the reduction temperature is expected to be significantly lower, and the formation of a molten Al4O4C-Al4C3 phase will be avoided.
[59] Thus, many potential difficulties in commercialisation of the above processes will be avoided.
This thesis will describe the effect of different factors such as carbon/alumina ratio, gas composition, and flow rate. It will develop an understanding on the kinetics and
UNSW Materials Science and Engineering School 38
mechanisms of reduction of various metal oxides in bauxite via carbothermal reduction.
It will ascertain the optimal conditions for production of metallic aluminium and ferroalloy by-products.
UNSW Materials Science and Engineering School 39
3 Experimental
This study was to explore the feasibility of stepwise carbothermal reduction of bauxite.
The reduction experiments were carried out under argon and hydrogen atmospheres in a molybdenum disilicide high temperature vertical tube furnace using mixtures of bauxite ores (from Western Australia and Queensland) and synthetic graphite. The reducing temperature range was from 850°C to 1600°C. Effluent gas was analysed by a
CO/CO2/CH4 infra-red (IR) analyser. Ore samples without mixing with graphite were also reduced in CO gas to examine reduction of iron oxide impurities.
Carbothermal reduction of bauxite ore first was the temperature programmed experiments. CO/CO2 curve from here identified the reduction stages temperature zones.
Second, were the step-by-step reductions (Table 3-3) which hold the temperature after reach the target temperature for a period of time (30 mins) to let the samples reduced completely. Samples were stopped at different temperatures after the peak in reduction peak had occurred. The composition of the reduced samples was analysed using X-Ray
Diffraction (XRD). Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray
Spectroscopy (EDS) were also utilised to identify the components. Carbon and oxygen contents were analysed using LECO carbon and oxygen analysers. The extent of reduction was determined by means of gas and sample analyses.
Following tables list present project. Table 3-1 is showing the first heating of Western
Australian and Queensland bauxite mixed with carbon. The purpose here was obtaining the CO/CO2 curve for the whole reduction of bauxites.
UNSW Materials Science and Engineering School 40
WB/C: Western Australian bauxite mixing with carbon (C:O= 1.2:1)
QB/C: Queensland bauxite mixing with carbon (C:O= 1.2:1)
WB: Western Australian bauxite without mixing with carbon
QB: Queensland bauxite without mixing with carbon
Table 3-1 Preliminary experiments of reduction of bauxites to identify reduction stages
Sample Reduction temperature (oC) Atmosphere
Checked CO/CO2 Curve
WB/C 500-1600 Ar
QB/C 500-1600 Ar
Table 3-2 Temperature programmed reduction experiments
Sample Reduction temperature (oC) Atmosphere
Temperature Programmed Experiment
WB/C 850-1100 Ar
WB/C 850-1300 Ar
WB/C 850-1360 Ar
WB/C 850-1500 Ar
WB/C 850-1600 Ar
QB/C 850-1100 Ar
QB/C 850-1350 Ar
QB/C 850-1600 Ar
UNSW Materials Science and Engineering School 41
Table 3-3 Step-by-step reduction experiments
Sample Reduction temperature (oC) Atmosphere
Step-by-Step Experiments
WB/C 850-980 Ar
WB/C 850-1032 Ar
WB/C 850-1360 Ar
WB/C 850-1600 Ar
QB/C 850-1051 Ar
QB/C 850-1255 Ar
QB/C 850-1600 Ar
Table 3-4 Reduction of bauxites under different conditions
Sample Reduction temperature (oC) Atmosphere
Different gaseous reduction
WB/C 1600 CO
WB 1600 CO
WB/C 1100 CO
WB 1100 CO
QB/C 1600 CO
QB 1600 CO
QB/C 1100 CO
QB 1100 CO
WB/C 1600 H2
QB/C 1600 H2
UNSW Materials Science and Engineering School 42
3.1 Materials
The raw materials used in the project were:
(1) Metallurgical grade bauxite from Western Australia.
(2) Metallurgical grade bauxite from Queensland.
(3) Synthetic graphite with a powder size <20 µm. (99.97% purity)
Bauxite is a mixed ore which contains a lot of oxides. It is very difficult to calculate the oxygen contain certainly because the percentage of some oxides is very low. Luckily there are four main oxides in bauxite which are Al2O3, Fe2O3, SiO2 and TiO2. Current project will focus on these four oxides including analysis, calculation, XRD and EDS data. The major compositions of the two bauxite ores investigated in this project are presented in Table 3-5. Both bauxite ores were supplied by Rio Tinto Alcan and data are measured from chemistry laboratory UNSW by X-Ray Fluorescence (XRF).
Table 3-5 Major components of bauxites by XRF, wt%
(Alcoa), %
Component Western Australian Bauxite Queensland Bauxite
Al2O3 39.9 52.5
Fe2O3 19.3 13.6
SiO2 17.3 6.5
TiO2 1.3 2.6
UNSW Materials Science and Engineering School 43
Western Australian bauxite comprised approximately 40% aluminium oxide, one fifth of iron oxide, 17% silicon dioxide and ~ 1% of titanium dioxide. Compared with Western
Australian bauxite the most significant change to take place between the compositions in Queensland bauxite are the percentage of aluminium oxide and silicon dioxide. More than half the aluminium oxide exists in Queensland bauxite with 13% iron oxide and
6% silicon dioxide. Titanium dioxide is only 2.6% in Queensland bauxite. Graphite powder was supplied by Sigma-Aldrich, Australia.
3.2 Gases
The gases used in the project were argon and hydrogen, and they were supplied by Air
Liquid (Australia). Carbon monoxide was supplied by Core gas (Australia).
Specifications are as follow:
● Argon Gas, 99.999%,Ultra High Purity.
● Hydrogen Gas, 99.999%, Ultra High Purity.
● Carbon Monoxide Gas, 99.5%, High Purity.
3.3 Sample Preparation
The reduction mixture was prepared by mixing raw materials. The amount of graphite added was sufficient that to C:O = 1.2:1 (in Bauxite) molar ratio. A wet mixing method was used in order to obtain a uniform mixture. The powders were intimately mixed by addition of 0.3 wt% CMC (Carboxymethyl Cellulose) as a binder and 60 wt% (dry basis) of water. All these chemicals, together with 8-10 ceramic balls (φ 8.1mm), were placed into a 500mL plastic container using ball milling for about five hours to form a paste.
UNSW Materials Science and Engineering School 44
The paste was dried in an oven at 105°C for 12 hours. Then the powder was pressed into cylindrical pellets using a 10 tonnes hydraulic press, manufactured by Enerpac,
NSW, Australia. The pellets were pressed at 3kN for two minutes. The final pellet was
8mm in diameter and about 10 mm in height. The mass was approximately one gram.
3.4 Experimental Set Up
3.4.1 Experimental Furnace
The experiments were carried out in a molybdenum disilicide high temperature vertical tube furnace manufactured by Ceramic Engineering, Australia. The furnace was heated by four Kanthal Super 1800 molybdenum disilicide resistance heating elements. Power to the element was switched from the mains by a phase angle thyristor unit through a transformer. The thyristor unit was protected by a semiconductor fuse. Power to the element was controlled by the programmable temperature controller through the phase angle thyristor. The maximum temperature of 1700oC can be reached.
3.4.2 Reactor Set Up
The schematic of the reactor set up is shown in Figure 3-1. The system consists of an internal gas ducting tube and an external sheath, both being made of high purity recrystallised alumina. The internal tube dimensions are 8.6 mm internal diameter, 12 mm external diameter, and 860 mm in length. The external sheath is 19 mm internal diameter, 24 mm external diameter and 740 mm long. A sample pellet was loaded into a graphite crucible, which is located at the bottom of the alumina sheath. The temperature at the top of the pellet was measure by a type B thermocouple which was inserted
UNSW Materials Science and Engineering School 45
through the gas ducting tube. This temperature is referred to as the reduction temperature. The thermocouple consists of Pt-Rh wires and an insulator held in an alumina sheath of 6.4 mm OD and 4 mm ID to separate it from contact with reacting gas atmosphere.
740mm
24mm
Figure 3-1 Schematic diagram of the reaction system
3.4.3 Gas System
The gas, in which the reduction was studied, passed through the gas inlet into the reactor, through the sample and finally came out through the gas outlet which was connected
UNSW Materials Science and Engineering School 46
with the gas analyser. The furnace protecting gas from external to the reactor was continuously kept flowing from external to the reactor to avoid oxide formation on the sample and reactor. Gas flow was maintained until the end of the experiments.
The schematic diagram of the gas system is shown in Figure 3-2. The gases passed through 4A molecular sieve three columns for removal of moisture before being introduced into the reactor. The flow rate of the gases was controlled by Brooks mass-flow controllers (model 5850E mass flow meter and 0154E controlling electronics,
Brooks Instrument, Hatfield, PA). Compositions of effluent gases were monitored by a
CO/CO2/CH4 infrared (IR) analyser (Advanced Optima AO2020, ABB, Germany).
Figure 3-2 Schematic diagram of the gas supply system
The mass flow controller was calibrated for argon, hydrogen and carbon monoxide gases. Calibration curves obtained using bubble flow meters are shown in Figure 3-3.
UNSW Materials Science and Engineering School 47
Figure 3-3 The setting point of a mass flow controller and corresponding gas flow rate for Ar, CO
and H2
3.5 Experimental Procedure
A known mass of pellet of bauxite-carbon mixture (about 1 gram) was loaded in a graphite crucible and introduced to the bottom of the reaction tube. The thermocouple was placed into the reactor and the tube was sealed using fittings and O-rings.
The tests were composed of temperature programmed experiments and step by step reduction. The samples were flushed with argon gas and then positioned in the hot zone of the furnace. The reactor system was assembled, and purged with argon at room temperature for 10-15 minutes to remove air from the system. In temperature programmed experiments, the furnace temperature was ramped from 850oC until
1600oC at a heating rate of 2oC/min. For step by step reduction experiments, a range of
UNSW Materials Science and Engineering School 48
temperatures were set, between 1100oC to 1600oC, and the heating rate between the steps was 5oC/min.
After a sample was reduced for certain time the reactor assembly with sample was lifted from the hot zone of furnace very slowly and the sample was quenched in under the reaction atmosphere. Fast pulling out of the reactor was avoided to protect the external alumina sheath from thermal shock and cracking. The reaction gas atmosphere was still kept flowing during sample cooling to prevent the sample from reacting with air. When the temperature decreased to the room temperature, the sample and deposit were removed and weighed. They were kept in a closed desiccator for further analysis.
3.6 Sample Characterisation
The phase composition of samples in the progress of reduction was analysed by XRD.
The contents of oxygen and carbon in the samples were measured by LECO analysis.
3.6.1 X-ray Diffraction Analysis
After the experiments, the phase composition of the reduced samples was analysed using the powder X-Ray Diffraction (XRD) (Philips X’pert Multipurpose X-Ray
Diffraction System (MPD)). Before the XRD analysis, the reduced sample pellets were ground to a fine powder. Then the powder was packed in a hollow of 15mm diameter and 1mm depth on a steel sample holder. The surface of the sample powder was flat.
The phases were identified by XRD using monochromatic Cu Kα radiation.The parameters for XRD analysis were presented in Table 3-6.
UNSW Materials Science and Engineering School 49
Table 3-6 XRD testing parameters
2θ start angle (o) 20
2θ end angle (o) 80
Step size (o) 0.026
Time per step (s) 0.2
Scan Speed (o/s) 0.13
3.6.2 LECO Analyses
The oxygen content of the reduced samples was measured by LECO (Model
TC-436DR). Approximately 0.05-0.08 gram of the reduced sample powder was loaded into a tin capsule, and then the capsule was weighed and inserted in a nickel basket. The nickel basket with the capsule was loaded into a graphite crucible. After the system was purged by high purity helium, the crucible was heated to the combustion temperature.
The oxygen in the sample reacted with the carbon in the crucible to form carbon monoxide and carbon dioxide which were detected by thermal conduction detectors.
The carbon content of the reduced samples was detected by combustion method using a
LECO carbon and sulphur determinator (Model TC-444). A sample, about 0.1 gram in weight, was loaded into a ceramic boat, which was pushed into the hot zone of a furnace at 1300oC in a flowing oxygen gas. The carbon in the sample was oxidised into CO and
CO2 of which the amounts were detected by CO/CO2 detectors and the carbon content was calculated and outputted by the LECO analyser.
UNSW Materials Science and Engineering School 50
3.6.3 Scanning Electron Microscopic (SEM) and Energy Dispersive X-ray Spectroscopy (EDS)
The morphology of samples was observed using a Hitachi S-4500 Field Emission
Scanning Electron Microscope (FESEM) with Energy Dispersive Spectroscopy (EDS) attachment. The EDS detector was manufactured by Oxford Instrument ISIS. When the
EDS unit was enabled, the characteristic X-ray produced from electron bombardment were detected by an energy dispersive spectrometer, which is a solid state device that discriminates among X-ray energies. By using FESEM/EDS, the sample morphology and semi-quantitative composition can be deduced.
3.7 Calculation of Extent of Reduction
The extent of reduction was defined as a fraction of oxygen in bauxite removed from the sample in the process of reduction. The extent of reduction was calculated on the basis of the off-gas composition as measured by IR analyser and LECO analysis.
3.7.1 Calculation of Extent of Reduction Based on the Off-gas Analysis
Using data of the CO and CO2 concentrations in the off-gas measured by the IR analyser the extent of reduction was calculated as follows (for the gas flow rate
1L/min):
(44)
(45)
UNSW Materials Science and Engineering School 51
(46)
where VCO (%) and VCO2 (%) are the volume % of CO and CO2 in the off-gas and t is the reduction time.
Current thesis only focuses on the main oxides in bauxite (Fe2O3, SiO2, TiO2 and Al2O3).
Because the amounts of other oxides are too small compare with these four.
Total mass of O in sample =
3.7.2 Calculation of Extent of Reduction Based on Content of Oxygen in the Reduced Sample
The extent of reduction was calculated from the remaining oxygen content in the reduced sample analysed by LECO. The original samples (bauxite and graphite mixture) had a molar ratio of C:O (in bauxite) = 1.2:1. Therefore, the weight fraction of oxygen
UNSW Materials Science and Engineering School 52
in the original sample before the reduction is:
(47)
0 where f O is the weight fraction of oxygen in the pellet before the reduction, MWx is the molar weight of X, and fx is the weight fraction of component x in a bauxite ore. The
0 weight of O (WO ) in the pellet before reduction is:
(48)
0 where Ws is the weight of the original pellet before the reduction reaction. The weight of oxygen in a reduced sample (W t ) can be calculated as: O
(49)
t t where f O is weight fraction of oxygen in the pellet after reduction;Ws is the sample weight after reduction. Then the extent of reduction (X) is equal to:
(50)
UNSW Materials Science and Engineering School 53
4 Results and Discussion
Bauxite ores mined from Western Australian and Queensland were reduced using two reduction procedures. Temperature–programmed reduction was followed by sample characterisation to identify the stages and temperature range of reduction of different metal oxides. Further reduction experiments were carried out in different steps, with each step being at a constant temperature to demonstrate the feasibility of stepwise reduction of metal oxides from bauxite.
4.1 Progress of Carbothermal Reduction
Temperature–programmed reduction was carried out under a flowing argon gas atmosphere. The rate of reduction was monitored by detection of CO released from the reactions using a CO/CO2/CH4 IR analyser.
Figure 4-1 presents the change in CO content in the off–gas during reduction experiments of both bauxites. The reduction curves consist of multiple overlapped peaks, indicating that the metal oxides in the bauxite ores were reduced over a range of temperatures. The concentration of CO can be obtained from the CO/CO2 curve. When the reaction temperature was increased, an oxide was reduced by carbon, releasing CO.
The reduction curve of Western Australian bauxite is more complex than that of
Queensland bauxite owing to the presence of more impurities in the former ore.
UNSW Materials Science and Engineering School 54
Western Australian Bauxite 6000 1800
1600 5000 1400
CO level (ppm) 4000 1200 C Temperature, °C ° 1000 3000 800 CO levelCO (ppm) 2000 600 Temperature,
400 1000 200
0 0 0 100 200 300 400 500 Minute(s)
Queensland Bauxite 2500 1800
1600
2000 1400 CO level (ppm)
1200 ° C
Temperature, C ° 1500 1000
800 1000 CO levelCO (ppm) 600 Temperature,
500 400 200
0 0 0 100 200 300 400 500 Minute(s)
Figure 4-1 Change in CO content in the off-gas during temperature programmed reduction of
Western Australian and Queensland bauxite ores
UNSW Materials Science and Engineering School 55
Based on the major compositions (Table 3-5) and the reducibility of the metal oxides
(thermodynamic consideration), it can be seen that the first major reduction stage
(reduction of iron oxides) took place in the temperature range of 850–1100°C, while the second major reduction stage (reduction of silica and titania from bauxite) occurred in the range of 1300–1500°C. Further reduction was mainly attributed to that of alumina, which was not completely reduced at the end of the experiments.
In the step–by–step reduction experiments, reduction temperatures were decided on the basis of the peak temperatures of reduction curves in Figure 4-1. For Western Australian bauxite, 980°C , 1032°C , 1360°C and 1600°C, and for Queensland bauxite, 1051°C ,
1255°C and 1600°C were considered as the step temperature. The last temperature,
1600°C, was limited by the maximum operation temperature of the furnace used. The temperature and CO evolution curves are presented in Figure 4-2.
During the first step in the step–by–step reduction experiments, even though the furnace was heated to the designated temperatures in advance, the sample temperature took about 20–30 min to become stable. Therefore, although the reduction temperature is labelled constant in Figure 4-2, in reality, it was changing in the early stages. Different iron oxides have different reducibility; this explains the occurrence of multiple peaks in the CO curves of Western Australian bauxite ore. However, multiple peaks did not appear in the CO curve of Queensland bauxite, which may be because the compositions of the bauxite ores were different; the temperatures used in the reduction were also different.
UNSW Materials Science and Engineering School 56
Figure 4-2 Changes in experimental temperature and CO content in the off gas in step–by–step reduction experiments
UNSW Materials Science and Engineering School 57
It can be seen from the reduction curve of Western Australian bauxite that the reduction rate increased during 150–200 min of the reduction process, although the reduction temperature was remained constant. This phenomenon, which is rare in isothermal reduction experiments, could be attributed to the formation of new reaction interface along with reduction of iron oxides.
In the reduction curve of Queensland bauxite, a sharp CO peak appeared before the furnace temperature was stabilised at 1051oC. Increasing the temperature to 1255oC resulted in the formation of another CO peak because the rate of reduction increased according to Arrhenius law, following which the reaction rate slowed down due to consumption of reactants. The rate became increasing again due to an increase in temperature to 1600oC, and started to decrease after the temperature was hold at the temperature. .
4.2 X-Ray Diffraction Analysis
The temperature–programmed reduction experiments (Figure 4-1) were stopped at various temperatures and the reduced samples at different stages were analysed by
X-ray diffraction (XRD). The XRD patterns of reduced samples of Western Australian and Queensland bauxites are presented in Figures 4-3 and 4-4, respectively. XRD pattern database was from PDF22004 of X’pert Highscore Plus software.
As shown in Figure 4-3, original Western Australian bauxite contains hydrated alumina, goethite and hematite. Weak peaks of silica were also detected, but not those of titania due to its low content in the ore. When the reduction temperature was increased to
UNSW Materials Science and Engineering School 58
1100°C, dehydration of alumina took place and a peak corresponding to metallic iron appeared. Further reduction up to a temperature of 1300°C made the peaks of alumina stronger, reflecting higher content of alumina in the ore. An obvious shift of the iron peak is visible, which implies a significant change in its composition due to the formation of a ferroalloy. During the reduction experiments alloys such as Al4SiC4 and ferroalloy were formed. Both of them could contain varying amount of Si, Al and C.
Peaks of SiC could be detected in the samples reduced at a temperature of 1360°C and above. Meanwhile, only weak peaks of TiC can be observed at 1360oC, which may be attributed to low TiO2 content in the bauxite ore. In the samples reduced until 1500°C and 1600°C, peaks of aluminium carbide, Al4C3, were also detected.
Figure 4-3 XRD patterns of Western Australian bauxite reduced at different temperatures
UNSW Materials Science and Engineering School 59
Reduction of Queensland bauxite followed a route similar to that of Western Australian bauxite (Figure 4-4). Raw Queensland bauxite also possesses high content of hydrated alumina and hematite, with small amounts of silica and a trace amount of titania.
Hydrated alumina and hematite were converted to alumina and iron metal phase, respectively, when the temperature was increased to 1100°C. The peaks corresponding to silica imply that it cannot be converted to any other compound at low–temperature reduction conditions.
Figure 4-4 XRD patterns of Queensland bauxite reduced at different temperatures
Peaks corresponding to carbides of silicon and titanium were observed in the samples reduced at 1350°C, as the carbides started to form at this temperature. Shifting iron UNSW Materials Science and Engineering School 60
peaks indicates that it was converted to ferroalloy in the same way as in case of Western
Australian bauxite. An Al4SiC4 peak appeared at 1350°C; Al2O3 and SiO2 smelted to form Al4SiC4. Peaks of Al4C3 were displayed when samples were reduced at 1600°C.
Figure 4-5 XRD patterns of Western Australian bauxite samples reduced at different steps
In a step–by–step reduction, as shown in Figure 4-2, experiments were stopped after each step and the reduced samples were analysed by XRD. XRD patterns of the reduced
Western Australian and Queensland bauxite ores are presented in Figures 4-5 and 4-6, respectively.
UNSW Materials Science and Engineering School 61
Figure 4-6 XRD patterns of Queensland bauxite samples reduced at different steps
As shown in Figure 4-5, only hematite in the Western Australian bauxite was converted to metallic iron in the first stage at 980°C. Compared with other oxides, Al2O3 had higher peaks due to its high concentration in the raw material. The weaker peak of TiO2 was due to its low concentration in the raw material. Further increase in temperature to
1032°C led to the complete reduction of the iron. At 1360°C, new peaks, including those of SiC, TiC and Al4SiC4, appeared. Moreover, better crystallisation of Al2O3 gave stronger Al2O3 peaks, and a ferroalloy phase was also detected. In the reduction up to
1600°C , the peak of ferroalloy phase became very strong and also Al4C3 peaks appeared.
UNSW Materials Science and Engineering School 62
For Queensland bauxite (Figure 4-6), iron peaks were detected in the sample reduced at
1051°C ; the iron was transform ed to a ferroalloy at 1255°C with reduction of silica (and probably titania). Further reduction up to temperature of 1600°C caused stronger ferroalloy peak. Fe2O3 was the only oxide reduced at 1051°C. Peaks corresponding to carbides of silicon and titanium were noted at 1255°C and above. A small peak of
Al4SiC4 was also detected at this temperature.
Al4C3 was present only in the sample reduced at 1600°C . It was noted that the Al2O3 peaks at 1600°C were slightly weaker than those at 1255°C , suggesting that significant amount of alumina was reduced with formation of Fe–Si–Al alloy and Al4C3.
As titanium content was low in raw sample, the intensity of TiC peak in XRD patterns was very weak, making its detection extremely difficult in all the samples.
4.3 Scanning Electron Microscope Analysis of Reduced Samples
4.3.1 Samples Reduced Until a Temperature of 1100oC
Figure 4-7 shows the scanning electron microscope (SEM) images and the distribution of different elements in a Queensland bauxite sample reduced by the temperature–programmed reduction process (Figure 4-1) until 1100°C. Figure 4-8 presents the chemical composition of selected particles by Energy Dispersive
Spectroscopy (EDS) analysis. The bright dots distributed in the matrix of oxides correspond to metallic iron. The matrix (Points 17, 18 and 20) contains mainly alumina with low levels of silica, iron and titania. In the image of iron distribution, more bright
UNSW Materials Science and Engineering School 63
dots of iron are present at the lower right area than at any other area. The bright dots in the images of elemental silicon distribution represent quartz particles in bauxite that was not reduced until 1100oC. Although the particles at Point 19 contain mainly titania, low levels of silica, alumina and iron are also detected.
As mentioned before, EDS analysis was focused only on Al, Fe, Si and Ti elements; the missing percentages are related to the amount of impurities present.
1100oC Queensland Bauxite
Figure 4-7 EDS elemental distribution of Queensland bauxite reduced until 1100oC.
UNSW Materials Science and Engineering School 64
(a) Particles whose chemical compositions were detected by EDS.
Point 17 Point 18
Element Wt.% Norm. wt.% Norm. at.% Element Wt.% Norm. wt.% Norm. at.% O 31.52 39.92 45.73 O 25.63 36.95 41.30 Al 26.93 34.11 23.17 Al 23.31 33.59 22.27 C 14.42 18.26 27.87 C 15.93 22.96 34.19 Ti 2.84 3.59 1.37 Si 0.34 0.49 0.31 Fe 2.02 2.56 0.84 Fe 4.17 6.01 1.93 Si 1.23 1.56 1.02 Sum: 69.38 100.00 100.00
Sum: 78.96 100.00 100.00
(b) EDS spectroscopy and chemical composition of the selected points shown on image
UNSW Materials Science and Engineering School 65
Point 19 Point 20
Element Wt.% Norm. wt.% Norm. at.% Element Wt.% Norm. wt.% Norm. at.% Ti 44.38 45.23 22.63 O 27.06 39.73 45.38 O 33.77 34.42 51.53 Al 25.44 37.36 25.30 C 9.23 9.41 18.76 C 12.05 17.69 26.91 Fe 5.54 5.64 2.42 Fe 2.07 3.05 1.00 Al 4.18 4.26 3.78 Si 1.48 2.17 1.41 Si 1.02 1.04 0.88 Sum: 68.10 100.00 100.00
Sum: 98.11 100.00 100.00
Figure 4-8 Point analysis of Queensland bauxite reduced until 1100oC
Images in Figure 4-9 correspond to a Western Australian bauxite sample reduced until a temperature of 1100 °C. The EDS spectra and phase compositions by point detection are presented in Figure 4-10.
In case of Western Australian bauxite, the sample reduced at low temperature had much smaller bright dots compared with those reduced at high–temperature conditions. Iron and silica particles are easily identified by their brightness in the elemental distribution images.
The EDS images showed that amounts of Fe2O3 and SiO2 in the sample were more in
Western Australian bauxite (Figure 4-9) than in Queensland bauxite (Figure 4-7), which also match the information given in Table 3-5.
UNSW Materials Science and Engineering School 66
It can be seen from the EDS images that the oxides were widely distributed, which implies that the reduction reaction was far from completion. Different elements such as iron, silicon and titanium occupy different regions. No overlap between iron and silicon can be observed.
1100oC
Western Australian Bauxite
Figure 4-9 SEM images and EDS elemental distribution of Western Australian bauxite reduced until 1100oC
UNSW Materials Science and Engineering School 67
(a ) Particles whose chemical compositions were detected by EDS
Point 10 Point 11
Element Wt.% Norm. wt.% Norm. at.% Element Wt.% Norm. wt.% Norm. at.% O 33.04 47.20 53.93 Fe 27.77 28.97 10.76 Si 27.55 39.37 25.62 C 22.12 23.07 39.86 C 9.40 13.43 20.45 O 23.14 24.13 31.30 Sum: 69.99 100.00 100.00 Al 20.74 21.63 16.64
Si 1.52 1.58 1.17 Ti 0.59 0.62 0.27 Sum: 95.88 100.00 100.00
UNSW Materials Science and Engineering School 68
Point 12 Point 13
Element Wt.% Norm. wt.% Norm. at.% Element Wt.% Norm. wt.% Norm. at.% O 26.91 37.11 40.56 O 44.31 43.13 52.47 Al 25.58 35.28 22.86 Al 18.82 18.32 13.22 C 17.56 24.22 35.26 Si 17.76 17.29 11.98 Fe 1.85 2.55 0.80 C 10.10 9.83 15.93 Si 0.60 0.83 0.52 K 18.82 18.32 13.21 Sum: 72.51 100.00 100.00 Fe 2.58 2.52 0.88
Mg 1.85 1.80 1.45 Na 1.73 1.69 1.43 Ti 0.62 0.61 0.25 Sum: 102.75 100.00 100.00
(b) EDS spectroscopy and chemical composition of the selected points shown on image.
Figure 4-10 EDS spectra and elemental composition of Western Australian bauxite reduced until
1100°C
In Point 10, mainly silicon and oxygen elements were detected, other elements being negligible. Thus, it represents a quartz particle. In Point 11, a particle with white spots represents alumina and iron. Some of the iron phase started to grow and form a larger iron metal phase. The dark grey area (Point 12) has a large amount of alumina and a trace of iron and silicon. Point 13 is a special example that listed a number of impurities such as K, Mg and Na. These percentages of impurities do not show on other tables due to their small amounts in the bauxite ore. A high concentration of oxides existed in this area.
UNSW Materials Science and Engineering School 69
Reduction of iron in bauxite ores starts at a low temperature. In contrast to other oxides, iron oxide is reduced firstly in an argon atmosphere.
4.3.2 Samples Reduced Up to 1600°C
Figure 4-11 shows the SEM images and the distribution of different elements in a
Western Australian bauxite sample reduced up to 1600°C and further reduced at that temperature for 2h. Figure 4-12 presents the composition of the sample at selected points analysed by EDS. From the morphological point of view, although the sample consisted of sintered grains it was still porous, indicating that not all oxides were melted at the final temperature. The bright grains distributed in the matrix of oxides are ferroalloy containing a large amount of Fe and a small amount of Si and Al (Point 57 in
Figure 4-12). The oxide phase contains mainly alumina and a trace amount of residual iron oxide (Point 58 in Figure 4-12). Carbon existed in the system because sample preparation for SEM required carbon coating.
UNSW Materials Science and Engineering School 70
1600°C
Western Australian Bauxite
Figure 4-11 SEM images and elemental distribution of Western Australian bauxite reduced until 1600oC
UNSW Materials Science and Engineering School 71
(a ) Particles whose chemical compositions were detected by EDS
Point 57 Point 58
Element Wt.% Norm. wt.% Norm. at.% Element Wt.% Norm. wt.% Norm. at.% Fe 64.89 82.60 53.81 O 40.47 43.71 51.66 C 10.69 13.60 41.20 Al 41.90 45.25 31.71 Si 1.81 2.31 2.99 C 9.66 10.43 16.42 Al 1.17 1.49 2.01 Fe 0.57 0.62 0.21 Sum: 78.55 100.00 100.00 Sum: 92.60 100.00 100.00
(b) EDS spectroscopy and chemical composition of the selected points shown on image.
Figure 4-12 EDS spectra and elemental composition of Western Australian bauxite reduced until 1600°C
UNSW Materials Science and Engineering School 72
From SEM and EDS analyses, iron oxides in the bauxite ores were reduced at temperatures below approximately 1100oC. Reduction of alumina, silica and titania needs higher temperatures. At low temperature of 1100 °C, SEM images show that the particles are more dispersed to each other. On the other hand, at 1600 °C, the particles formed a cross linked net due to sintering. It seems that although generated metal iron was melted at high temperatures, the oxide phases were held as solid. This proves the concept of stepwise reduction of metal oxides from bauxite in solid phase.
4.3.3 Segregation of Titanium, Silicon and Aluminium from Alloy Phase
Figure 4-13 presents the SEM images and EDS mapping of elemental distribution of
Western Australian bauxite reduced to 1600oC. The big bright grain is Fe–Si alloy. The grey areas within the alloy grain represent aluminium oxide; titanium bright dots are distributed around and within the grain, which do not correspond to high oxygen content. It is assumed that this material represents titanium carbide. Titanium carbide does not have high solubility in ferroalloy phase. [60] Titanium oxides were reduced in the ferroalloy phase at high temperatures. During cooling, titanium was segregated and combined with carbon to form titanium carbide. [61] Figure 4-14 presents the SEM images and EDS mapping of elemental distribution of Queensland bauxite reduced to
1600oC. Silicon grains are surrounded by iron matrix. Most of the oxygen maps overlap with aluminium, indicating formation of Al2O3 that are seen at the outer layer of the grain. Fewer signals of titanium were observed as in case of Western Australian bauxite.
Presence of a lot of cracks on iron grains indicated that these silicon grains were formed by the segregation of ferroalloy phase during cooling. The growing silicon grains either broke the iron grain into pieces or formed fractures.
UNSW Materials Science and Engineering School 73
Figure 4-13 SEM images and elemental distribution of Western Australian bauxite reduced to
1600oC
UNSW Materials Science and Engineering School 74
Figure 4-14 SEM images and elemental distribution of Queensland bauxite reduced to 1600oC
UNSW Materials Science and Engineering School 75
4.4 Extent of Reduction in Different Gas Atmospheres
Bauxite samples reduced at different temperatures were tested by LECO analysers and the final extents of reduction (percentage measure), that is removal of oxygen from metal oxides, were calculated. Table 4-1 lists the extent of reduction at different temperatures.
Table 4-1 Extent of reduction of Western Australian and Queensland bauxite ores reduced in argon at different temperatures
Temperature (oC) Extent of reduction (%)
Western Australian bauxite
1100 26.8
1300 40.5
1360 43.6
1500 56.1
1600 60.2
Queensland bauxite
1100 17.7
1350 30.9
1600 40.8
Table 4-2 lists the contributions of each major oxide, when fully reduced, to the extent of reduction of the two bauxite ore. Here, it was assumed that oxygen in each oxide was fully removed due to reduction reaction. When Fe2O3 is fully reduced, it contributes
16.9% and 12.3% to the extent of reduction of Western Australian and Queensland
UNSW Materials Science and Engineering School 76
bauxite, respectively. Mainly four oxides (Fe2O3, SiO2, TiO2 and Al2O3) account for the total oxygen content in bauxite ores because oxygen concentration from the remaining oxides (impurities) is negligibly low.
Table 4-2 Contribution of different metal oxides, assuming their complete reduction, to the extent of reduction of the two bauxite ores
Metal oxide (%) Fe2O3 SiO2 TiO2 Al2O3
Western Australian bauxite 16.9 26.9 1.5 54.7
Queensland bauxite 12.3 10.3 3.1 74.3
Comparing Tables 4-1 and 4-2, it can be found that although iron oxide was not completely reduced at 1100°C, the extent of reduction was beyond the maximum possible contribution by Fe2O3. This indicates that reduction of SiO2 and TiO2 occurred at this relatively low temperature, which is a distinct difference from the reduction of pure oxides. Similarly at 1600°C, the extent of reduction of bauxite ores was higher than the sum of complete reduction of Fe, Si and Ti oxides. This illustrates that partial reduction of alumina took place.
Table 4-3 presents the extent of reduction of bauxite samples at different temperatures, with and without mixing with graphite when reduced under CO. Using the same
Western Australian bauxite-graphite or Queensland bauxite-graphite pellets, but in CO the extent of reduction achieved was 29% and 20%, respectively, which were higher than those achieved in argon. In fact, even without the addition of graphite, the extent of reduction reached under CO was 24% and 15%, respectively. These data illustrate that,
CO, if used, plays a major role in the reduction of iron oxides at temperatures below
UNSW Materials Science and Engineering School 77
1100oC.
Table 4-3 Extent of reduction of Western Australian and Queensland bauxites by ramping from
850oC to different temperatures under CO atmosphere
Pellet composition Temperature (oC) Extent of reduction (%)
Western Australian bauxite + 1100 29.2 graphite
Western Australian bauxite + 1600 40.0 graphite
Western Australian bauxite 1100 24.0
Western Australian bauxite 1600 33.7
Queensland bauxite + graphite 1100 19.9
Queensland bauxite + graphite 1600 28.8
Queensland bauxite 1100 14.8
Queensland bauxite 1600 21.4
The mole ratio of C:O was 1.2 : 1.
The effect of CO to reduction of bauxite is examined because it is not only a product of carbothermal reduction at high temperatures, but also a reductant, especially to iron oxides at low temperatures (below 1100 °C). According to Tables 4-1 and 4-3, at
1100oC, reduction in CO resulted in higher extent of reduction of iron oxides than in argon; as a result, FeO content in the oxide phase became reduced, making the sintering more difficult. When the temperature was increased to 1600 °C, the extent of reduction in CO was increased from 1100 °C for both bauxite ores and either mixed with carbon or not. However, the extent of reduction reached in CO at 1600 °C was much lower than
UNSW Materials Science and Engineering School 78
in argon.
This phenomenon pointed out that CO is an effective reductant of iron oxides, but not of
SiO2, TiO2 and especially Al2O3 because of thermodynamic limitation. High concentration of CO retards the reaction between alumina and carbon.
In the present project, the samples were heated to 1600oC and maintained at that temperature for 30 min. It is expected that increasing the reaction temperature or extend reduction time will increase the extent of reduction.
A possible route of bauxite processing is to separate ferroalloy formed by reduction from the oxide residue via magnetic separation. The remaining relatively pure alumina can undergo further carbothermal reduction to produce aluminium carbide which can be further converted to metallic aluminium. This way, the impurity oxides of Fe, Si and Ti in bauxite ores are also utilised in addition of aluminium production.
In the technology of carbothermal reduction of bauxites, there is no need to produce pure alumina, which avoids the consumption of a large amount of NaOH solution and generation of very harmful red mud waste. Furthermore, there is no fluorine compounds are involved in the process which so will avoid fluorine emission issues in aluminium production. Due to the relatively simple process and without consumption of a large amount of electricity for electrolysis, carbothermal reduction has a potential to decrease the production costs and CO2 emission in aluminium production. More detailed investigation of the process is in progress.
UNSW Materials Science and Engineering School 79
5 Conclusions and Recommendations for Future Research
5.1 Conclusions
The carbothermal reduction of bauxite from Western Australian and Queensland was carried out in atmospheres of Ar and CO at elevated temperatures. The major conclusions of this research are:
1. Bauxite ore can be directly reduced carbothermally by temperature programmed
reduction or stepwise reduction process. Appropriate control of reaction temperature
may maintain the residual oxide phases in solid state. The extent of reduction in
argon was 60% for Western Australia bauxite and 40% for Queensland bauxite.
2. The reduction sequence of the metal oxides in the bauxite ores is iron oxides then
silica and titania and then alumina. Metallic iron is formed at temperatures below
1100oC. At 1200oC, or above, a ferroalloy phase containing silicon and aluminium is
formed.
3. Carbides of titanium, silicon and aluminium were formed by carbothermal reduction.
The metals were formed and dissolved in the ferroalloy phase, which upon
saturation, were segregated as silicon carbide distributed inside alloy phase as
inclusions or around the alloy particles.
4. Carbothermal reduction of bauxite was faster in Ar than in CO; a higher extent of
reduction can be reached in Ar and using the sample mixed with carbon.
5. Carbon monoxide is effective in reduction of iron oxides, but shows a retarding
effect to reduction of silica, titania and alumina.
UNSW Materials Science and Engineering School 80
The reduction in this investigation was far from completion, a high percentage of unreacted alumina remained in the reduced samples. This was because of the limited temperature and time of reduction. A possible route of bauxite processing is to separate ferroalloy and further reduce relative pure alumina to produce metallic aluminium.
5.2 Recommendations for Future Research
The findings from this investigation proved the concept of stepwise reduction of bauxite as a promising alternative technology for aluminium production.
Suggested topics for future research include:
1. The flow rate may affect in the reduction of bauxite. The optimisation flow rate for
reduction reaction can be adjusted to reach better extent of reduction. The effect of
flow rate of argon can be monitored during the carbothermal reduction of bauxite.
2. Carbon plays an important role in carbothermal reduction. The ratio of
carbon/oxygen can reflect the extent of reduction directly. Thus, the effect of
carbon/oxygen ratio can be studied in carbothermal reduction of bauxite.
3. Effect of hydrogen on the carbothermal reduction of bauxite can be examined.
Hydrogen can enhance the reduction kinetics by methane formation from the
reaction of hydrogen and carbon.
UNSW Materials Science and Engineering School 81
List of References
1. Zhu FL, Yang B, Li Y, Yuan HB, Yu QC, Xu BQ, Dai YN: Behaviour Analysis
of Silica in Aluminium Production by Alumina Carbothermic
Reduction-Chlorination Process in Vacuum. Vacuum, 85(10):968-971.
2. King F: Aluminium and Its Alloys. Edited by West EG. Chichester: Ellis
Horwood Limited; 1987:241.
3. Aluminium-Specifications, Properties, Classifications and Calsses. England:
Aalco Metal Limited; 2011.
4. Yuan HB, Yang B, Xu BQ, Yu QC, Feng YB, Dai YN: Aluminium Production
by Carbothermo-Chlorination Reduction of Alumina in Vacuum.
Transactions of Nonferrous Metals Society of China, 20(8):1505-1510.
5. Wang X, Qu YY, Hu WW, Chen J, Zhao XY, Wu M: Particle Characteristics
and Rheological Constitutive Relations of High Concentration Red Mud.
Journal of China University of Mining and Technology 2008, 18(2):266-270.
6. McCoy JP: Bauxite Processing and the Role of Less Developed Countries.
The Social Science Journal 1989, 26(4):399-413.
7. Moors EHM: Technology Strategies for Sustainable Metals Production
Systems: A Case Study of Primary Aluminium Production in The
Netherlands and Norway. Journal of Cleaner Production 2006,
14(12-13):1121-1138.
8. Primary Aluminium Annual Production Capacity. International Aluminium
Institute; 22. Augest. 2011.
9. Liu W, Yang J, Xiao B: Review on Treatment and Utilization of Bauxite
Residues in China. International Journal of Mineral Processing 2009,
UNSW Materials Science and Engineering School 82
93(3-4):220-231.
10. Sglavo VM, Maurina S, Conci A, Salviati A, Carturan G, Cocco G: Bauxite
'Red Mud' in the Ceramic Industry. Part 2: Production of Clay-based
Ceramics. Journal of the European Ceramic Society 2000, 20(3):245-252.
11. Sglavo VM, Campostrini R, Maurina S, Carturan G, Monagheddu M, Budroni G,
Cocco G: Bauxite 'Red Mud' in the Ceramic Industry. Part 1: Thermal
Behaviour. Journal of the European Ceramic Society 2000, 20(3):235-244.
12. Halmann M, Frei A, Steinfeld A: Carbothermal Reduction of Alumina:
Thermochemical Equilibrium Calculations and Experimental Investigation.
Energy 2007, 32(12):2420-2427.
13. Yang D, Feng NX, Wang YW, Wu XI: Preparation of Primary Al-Si Alloy
from Bauxite Tailings by Carbothermal Reduction Process. Transactions of
Nonferrous Metals Society of China 2010, 20(1):147-152.
14. Myklebust H, Runde P: Greenhouse Gas Emission from Aluminium
Carbothermic Technology Compared to Hall-Heroult Technology. Light
Metal 2005:519-522.
15. Haupin W, Buschow KHJG, Robert WC, Merton CF, Bernard I, Edward JK,
Subhash M, Patrick V: Aluminium Production and Refining. In Encyclopedia
of Materials: Science and Technology. Oxford: Elsevier; 2001:132-141.
16. King F: Aluminium and Its Alloys. Edited by West EG. Chichester: Ellis
Horwood Limited; 1987:77.
17. Schwarz HG: Aluminum Production and Energy. In Encyclopedia of Energy.
New York; 2004.
18. Banks FE: Bauxite and Aluminum: An Introduction to the Economics of
Nonfuel Minerals. Toronto; 1979:189.
19. Bizot P, Sedel L: Alumina Bearings in Hip Replacement: Theoretical and UNSW Materials Science and Engineering School 83
Practical Aspects. Operative Techniques in Orthopaedics 2001, 11(4):263-269.
20. Standard OC: Wiley Encyclopedia of Biomedical Engineering. New York: John
Wiley and Sons; 2006.
21. MacKay KM, MacKay RA, Henderson W: Introduction to Modern Inorganic
Chemistry. Edited by 6. London; 2004:192.
22. Lancker Mv: Metallurgy of Aluminium Alloys. Michigan; 1967:18.
23. Hind AR, Bhargava SK, Grocott SC: The Surface Chemistry of Bayer Process
Solids: A Review. Colloids and Surfaces A: Physicochemical and Engineering
Aspects 1999, 146(1–3):359-374.
24. White Fused Alumina. China: Nine Nice Minerals International Group Limited;
2009.
25. Bi SW, Yu HY: Alumina Refining Technology. In Chemical Industry Press.
Volume 327. Beijing; 2006.
26. Gross GM, Seifert HJ, Aldinger F: Thermodynamic Assessment and
Experimental Check of Fluoride Sintering Aids for AlN. Journal of the
European Ceramic Society 1998, 18(7):871-877.
27. Blackman RAA, Wilson KW: Effects of Red Mud on Marine Animals. Marine
Pollution Bulletin 1973, 4(11):169-171.
28. Dethlefsen V, Rosenthal H: Problems with Dumping of Red Mud in Shallow
Waters. A Critical Review of Selected Literature. Aquaculture 1973,
2(0):267-280.
29. Snars K, Gilkes RJ: Evaluation of Bauxite Residues (red mud) of Different
Origins for Environmental Applications. Applied Clay Science 2009,
46(1):13-20.
30. Sushil S, Batra VS: Catalytic Applications of Red Mud, An Aluminium
Industry Waste: A Review. Applied Catalysis B: Environmental 2008, UNSW Materials Science and Engineering School 84
81(1–2):64-77.
31. Pérez-Villarejo L, Corpas-Iglesias FA, Martínez-Martínez S, Artiaga R,
Pascual-Cosp J: Manufacturing New Ceramic Materials from Clay and Red
Mud Derived from the Aluminium Industry. Construction and Building
Materials 2012, 35(0):656-665.
32. Pagano G, Meriç S, De Biase A, laccarino M, Petruzzelli D, Tünay O, Warnau M:
Toxicity of Bauxite Manufacturing By-products in Sea Urchin Embryos.
Ecotoxicology and Environmental Safety 2002, 51(1):28-34.
33. Trieff NM, Romana LA, Esposito A, Oral R: Effluent from Bauxite Factory
Induces Developmental and Reproductive Damage in Sea Urchins. Archives
of Environmental Contamination Toxicology 1995, 28:173-177.
34. Pagano G, His E, Beiras R, De Biase A, Korkina LG: Cytogenetic,
Developmental and Biochemical Effects of Aluminium, Iron and their
Mixture in Sea Urchins and Mussels. Archives of Environmental
Contamination and Toxicology 1996, 31:466-474.
35. King F: Aluminium and Its Alloys. Edited by West EG. Chichester: Ellis
Horwood Limited; 1987:58.
36. Friedrich O: TALAT - A Training Programme for Aluminium Application
Technologies in Europe. Materials Science and Engineering: A 1995,
199(1):73-77.
37. Kammer C: Fundamentals and Materials. In Aluminum Handbook. Volume 1.
London; 1999.
38. Aluminum Standard and Data. Volume 1. New York; 1972:209.
39. King F: Aluminium and Its Alloys. Edited by West EG. Chichester: Ellis
Horwood Limited; 1987:142.
40. Cochran CN: Alternate Smelting Processes for Aluminium. Light Metel UNSW Materials Science and Engineering School 85
1987:429-443.
41. Murray JP: Aluminium Production Using High-temperature Solar Process
Heat. Solar Energy 1999, 66(2):133-142.
42. Li J, Zhang G, Liu D, Ostrovski O: Low-temperature Synthesis of Aluminium
Carbide. Iron and Steel Institute of Japan International 2011, 51:870-877.
43. Corral MP, Moreno R, Requena J, Moya JS, Martínez R: Colloidal Dispersion
of Alumina-carbon Homogeneous Mixtures for Carbothermal Synthesis of
AlN. Journal of the European Ceramic Society 1991, 8(4):229-232.
44. Wang PY, Dai YN, Jiang SX, Zhong HQ: Study on Production of Aluminium
by Carbothermic Reduction and Subhalide Decomposition Under Vacuum.
Nonferrous Metals 2005, 2:11-13.
45. Li J, Zhang G, Liu D, Ostrovski O: Low-temperature synthesis of aluminium
carbide. Iron and Steel Institute of Japan International 2011, 51:870-877.
46. V. A. Dmitriev, Karasev SV. Russia; 2000.
47. Lacamera AF. Volume WO2000040767. 2000:18.
48. Sen W, Sun H, Yang B, Xu B, Ma W, Liu D, Dai Y: Preparation of Titanium
Carbide Powders by Carbothermal Reduction of Titania/Charcoal at
Vacuum Condition. International Journal of Refractory Metals and Hard
Materials 2010, 28(5):628-632.
49. Woo YC, Kang HJ, Kim DJ: Formation of TiC Particle During Carbothermal
Reduction of TiO2. Journal of the European Ceramic Society 2007,
27(2-3):719-722.
50. Berger LM, Gruner W, Langholf E, Stolle S: On the Mechanism of
Carbothermal Reduction Processes of TiO2 and ZrO2. International Journal
of Refractory Metals and Hard Materials 1999, 17(1-3):235-243.
51. Lee JS, Byeun YK, Lee SH, Choi SC: In Situ Growth of SiC Nanowires by UNSW Materials Science and Engineering School 86
Carbothermal Reduction Using a Mixture of Low-purity SiO2 and Carbon.
Journal of Alloys and Compounds 2008, 456(1-2):257-263.
52. Zheng Y, Zheng Y, Wang R, Wei K: Direct Determination of Carbothermal
Reduction Temperature for Preparing Silicon Carbide from the Vacuum
Furnace Thermobarogram. Vacuum 2007, 82(3):336-339.
53. Weimer A, Nilsen K, Cochran G, Roach R: Kinetic of Carbonthermal
Reduction Synthesis of beta Silicon Carbide. American Institute of Chemical
Engineers. 1993(39):493-503.
54. Liu GS, Strezov V, Lucas JA, Wibberley LJ: Thermal Investigations of Direct
Iron Ore Reduction with Coal. Thermochimica Acta 2004, 410(1-2):133-140.
55. O'Donnell RG, Trigg MB: The Mechanism of Conversion of Al2O3 to AlN via
Carbothermal Synthesis. Micron 1994, 25(6):575-579.
56. Cheng H, Reiser DB, Dean Jr S: On the Mechanism and Energetic of
Boudouard Reaction at FeO Surface: 2CO→C+CO2. Catalysis Today 1999,
50(3-4):579-588.
57. Chen HK, Lin CI, Lee C: Kinetics of the Reduction of Carbon/Alumina
Powder Mixture in a Flowing Nitrogen Stream. Journal of the American
Ceramic Society 1994, 77:1753-1756.
58. Motlagh M: Effect of Gas Flow on Simulataneous Carburization and
Reduction of Iron Ore. Ironmaking Steelmaking 1994, 21(4):297-302.
59. Yuan HB, Yue-Bin F, Yang B, Yu QC, Xu BQ, Wang PC, Dai YN: Thermal
Behaviour of Alumina in process of Carbothermic Reduction and Chloride
to Produce Aluminium. The Chinese Journal of Nonferrous Metals 2010,
20(4).
60. Mikhail SA, Webster AH: Chapter 13 Thermal Analysis in Metallurgy. In
Handbook of Thermal Analysis and Calorimetry. Volume Volume 2. Edited by UNSW Materials Science and Engineering School 87
Michael EB, Patrick KG: Elsevier Science B.V.; 2003:657-775.
61. Taylor KA: Solubility Products for Titanium-, Vanadium-, and
Niobium-Carbide in Ferrite. Scripta Metallurgica et Materialia 1995,
32(1):7-12.
UNSW Materials Science and Engineering School 88